ParkinsonDNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x7fe1c1862610>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from torch import from_numpy, optim\n",
    "import numpy as np\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "torch.manual_seed(1234)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalize(X, mu=None, sigma=None):\n",
    "    if mu is None and sigma is None:\n",
    "        mu = X.mean(axis=0)\n",
    "        sigma = X.std(axis=0)\n",
    "    return (X - mu)/sigma, mu, sigma"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ParkinsonsTrainDataset(Dataset):\n",
    "    def __init__(self):\n",
    "        xy = np.loadtxt('./parkinsons_updrs.data',\n",
    "                        delimiter=',', skiprows=1, dtype=np.float32)\n",
    "        train_len = int(xy.shape[0] * 0.7)\n",
    "        train_xy = xy[:train_len]\n",
    "        self.len = train_xy.shape[0]\n",
    "        y_label_indx = [5, 6]\n",
    "        self.train_x = np.delete(train_xy, y_label_indx, axis=1)\n",
    "        self.train_x = torch.from_numpy(self.train_x)\n",
    "        self.train_y = torch.from_numpy(train_xy[:, y_label_indx[0]: y_label_indx\n",
    "        [1] + 1])\n",
    "        \n",
    "        self.train_x, mu, sigma = normalize(self.train_x)\n",
    "        self.train_y, mu, sigma = normalize(self.train_y)\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        return self.train_x[index], self.train_y[index]\n",
    "\n",
    "    def __len__(self):\n",
    "        return self.len\n",
    "\n",
    "class ParkinsonsTestDataset(Dataset):\n",
    "    def __init__(self):\n",
    "        xy = np.loadtxt('./parkinsons_updrs.data',\n",
    "                        delimiter=',', skiprows=1, dtype=np.float32)\n",
    "        train_len = int(xy.shape[0] * 0.7)\n",
    "        test_xy = xy[train_len:]\n",
    "        self.len = test_xy.shape[0]\n",
    "\n",
    "        y_label_indx = [5, 6]\n",
    "        self.test_x = np.delete(test_xy, y_label_indx, axis=1)\n",
    "        self.test_x = torch.from_numpy(self.test_x)\n",
    "        self.test_y = torch.from_numpy(test_xy[:, y_label_indx[0]: y_label_indx\n",
    "        [1] + 1])\n",
    "        \n",
    "        self.test_x, mu, sigma = normalize(self.test_x)\n",
    "        self.test_y, mu, sigma = normalize(self.test_y)\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        return self.test_x[index], self.test_y[index]\n",
    "\n",
    "    def __len__(self):\n",
    "        return self.len\n",
    "\n",
    "train_dataset = ParkinsonsTrainDataset()\n",
    "test_dataset = ParkinsonsTestDataset()\n",
    "\n",
    "\n",
    "train_loader = DataLoader(dataset=train_dataset,\n",
    "                          batch_size=32,\n",
    "                          shuffle=True,\n",
    "                          num_workers=0, drop_last=True)\n",
    "test_loader = DataLoader(dataset=test_dataset,\n",
    "                          batch_size=32,\n",
    "                          num_workers=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Model(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Model, self).__init__()\n",
    "        self.l1 = nn.Linear(20, 300) \n",
    "        self.relu = nn.ReLU()\n",
    "        self.l2 = nn.Linear(300, 2)  \n",
    "    \n",
    "    def forward(self, x):\n",
    "        out1 = self.l1(x)\n",
    "        out2 = self.relu(out1)\n",
    "        out = self.l2(out2)\n",
    "        return out\n",
    "\n",
    "model = Model()\n",
    "\n",
    "criterion = nn.MSELoss()\n",
    "optimizer = optim.SGD(model.parameters(), lr=0.0001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 | Step: 1 | Loss: 3.1491\n",
      "Epoch 1 | Step: 2 | Loss: 2.7927\n",
      "Epoch 1 | Step: 3 | Loss: 4.5199\n",
      "Epoch 1 | Step: 4 | Loss: 2.6819\n",
      "Epoch 1 | Step: 5 | Loss: 2.5432\n",
      "Epoch 1 | Step: 6 | Loss: 3.4362\n",
      "Epoch 1 | Step: 7 | Loss: 6.3725\n",
      "Epoch 1 | Step: 8 | Loss: 2.4417\n",
      "Epoch 1 | Step: 9 | Loss: 3.5389\n",
      "Epoch 1 | Step: 10 | Loss: 5.9428\n",
      "Epoch 1 | Step: 11 | Loss: 2.5269\n",
      "Epoch 1 | Step: 12 | Loss: 3.3163\n",
      "Epoch 1 | Step: 13 | Loss: 3.3028\n",
      "Epoch 1 | Step: 14 | Loss: 3.9892\n",
      "Epoch 1 | Step: 15 | Loss: 2.7228\n",
      "Epoch 1 | Step: 16 | Loss: 3.5326\n",
      "Epoch 1 | Step: 17 | Loss: 3.0878\n",
      "Epoch 1 | Step: 18 | Loss: 2.7876\n",
      "Epoch 1 | Step: 19 | Loss: 4.4158\n",
      "Epoch 1 | Step: 20 | Loss: 5.2297\n",
      "Epoch 1 | Step: 21 | Loss: 14.7799\n",
      "Epoch 1 | Step: 22 | Loss: 3.7823\n",
      "Epoch 1 | Step: 23 | Loss: 4.0783\n",
      "Epoch 1 | Step: 24 | Loss: 3.1925\n",
      "Epoch 1 | Step: 25 | Loss: 2.6304\n",
      "Epoch 1 | Step: 26 | Loss: 3.4002\n",
      "Epoch 1 | Step: 27 | Loss: 3.3471\n",
      "Epoch 1 | Step: 28 | Loss: 3.2222\n",
      "Epoch 1 | Step: 29 | Loss: 3.9894\n",
      "Epoch 1 | Step: 30 | Loss: 3.5883\n",
      "Epoch 1 | Step: 31 | Loss: 4.0440\n",
      "Epoch 1 | Step: 32 | Loss: 3.0499\n",
      "Epoch 1 | Step: 33 | Loss: 2.0652\n",
      "Epoch 1 | Step: 34 | Loss: 3.8008\n",
      "Epoch 1 | Step: 35 | Loss: 2.5974\n",
      "Epoch 1 | Step: 36 | Loss: 2.7250\n",
      "Epoch 1 | Step: 37 | Loss: 2.6650\n",
      "Epoch 1 | Step: 38 | Loss: 3.7006\n",
      "Epoch 1 | Step: 39 | Loss: 3.6613\n",
      "Epoch 1 | Step: 40 | Loss: 15.4882\n",
      "Epoch 1 | Step: 41 | Loss: 2.0713\n",
      "Epoch 1 | Step: 42 | Loss: 2.8575\n",
      "Epoch 1 | Step: 43 | Loss: 2.8508\n",
      "Epoch 1 | Step: 44 | Loss: 3.6543\n",
      "Epoch 1 | Step: 45 | Loss: 2.7392\n",
      "Epoch 1 | Step: 46 | Loss: 3.0461\n",
      "Epoch 1 | Step: 47 | Loss: 3.5267\n",
      "Epoch 1 | Step: 48 | Loss: 4.2570\n",
      "Epoch 1 | Step: 49 | Loss: 3.1295\n",
      "Epoch 1 | Step: 50 | Loss: 5.7125\n",
      "Epoch 1 | Step: 51 | Loss: 2.3748\n",
      "Epoch 1 | Step: 52 | Loss: 3.7704\n",
      "Epoch 1 | Step: 53 | Loss: 3.8687\n",
      "Epoch 1 | Step: 54 | Loss: 4.1699\n",
      "Epoch 1 | Step: 55 | Loss: 3.1487\n",
      "Epoch 1 | Step: 56 | Loss: 3.0735\n",
      "Epoch 1 | Step: 57 | Loss: 2.7303\n",
      "Epoch 1 | Step: 58 | Loss: 3.9484\n",
      "Epoch 1 | Step: 59 | Loss: 2.4057\n",
      "Epoch 1 | Step: 60 | Loss: 3.1545\n",
      "Epoch 1 | Step: 61 | Loss: 3.5582\n",
      "Epoch 1 | Step: 62 | Loss: 3.3306\n",
      "Epoch 1 | Step: 63 | Loss: 2.6580\n",
      "Epoch 1 | Step: 64 | Loss: 3.8107\n",
      "Epoch 1 | Step: 65 | Loss: 3.1345\n",
      "Epoch 1 | Step: 66 | Loss: 2.9124\n",
      "Epoch 1 | Step: 67 | Loss: 2.6974\n",
      "Epoch 1 | Step: 68 | Loss: 2.7672\n",
      "Epoch 1 | Step: 69 | Loss: 2.7809\n",
      "Epoch 1 | Step: 70 | Loss: 3.5183\n",
      "Epoch 1 | Step: 71 | Loss: 3.2698\n",
      "Epoch 1 | Step: 72 | Loss: 4.1315\n",
      "Epoch 1 | Step: 73 | Loss: 2.6200\n",
      "Epoch 1 | Step: 74 | Loss: 12.5259\n",
      "Epoch 1 | Step: 75 | Loss: 2.7673\n",
      "Epoch 1 | Step: 76 | Loss: 3.7645\n",
      "Epoch 1 | Step: 77 | Loss: 3.0926\n",
      "Epoch 1 | Step: 78 | Loss: 2.5710\n",
      "Epoch 1 | Step: 79 | Loss: 2.9754\n",
      "Epoch 1 | Step: 80 | Loss: 2.9197\n",
      "Epoch 1 | Step: 81 | Loss: 3.3640\n",
      "Epoch 1 | Step: 82 | Loss: 2.6934\n",
      "Epoch 1 | Step: 83 | Loss: 3.2064\n",
      "Epoch 1 | Step: 84 | Loss: 2.9576\n",
      "Epoch 1 | Step: 85 | Loss: 3.2099\n",
      "Epoch 1 | Step: 86 | Loss: 3.4495\n",
      "Epoch 1 | Step: 87 | Loss: 3.0708\n",
      "Epoch 1 | Step: 88 | Loss: 2.2381\n",
      "Epoch 1 | Step: 89 | Loss: 14.1389\n",
      "Epoch 1 | Step: 90 | Loss: 2.3359\n",
      "Epoch 1 | Step: 91 | Loss: 3.2830\n",
      "Epoch 1 | Step: 92 | Loss: 2.3161\n",
      "Epoch 1 | Step: 93 | Loss: 2.6687\n",
      "Epoch 1 | Step: 94 | Loss: 3.8600\n",
      "Epoch 1 | Step: 95 | Loss: 3.4422\n",
      "Epoch 1 | Step: 96 | Loss: 3.1516\n",
      "Epoch 1 | Step: 97 | Loss: 1.9694\n",
      "Epoch 1 | Step: 98 | Loss: 2.9577\n",
      "Epoch 1 | Step: 99 | Loss: 2.7476\n",
      "Epoch 1 | Step: 100 | Loss: 2.9113\n",
      "Epoch 1 | Step: 101 | Loss: 2.4265\n",
      "Epoch 1 | Step: 102 | Loss: 2.3578\n",
      "Epoch 1 | Step: 103 | Loss: 2.9733\n",
      "Epoch 1 | Step: 104 | Loss: 2.4230\n",
      "Epoch 1 | Step: 105 | Loss: 3.2049\n",
      "Epoch 1 | Step: 106 | Loss: 2.2637\n",
      "Epoch 1 | Step: 107 | Loss: 2.5150\n",
      "Epoch 1 | Step: 108 | Loss: 2.6471\n",
      "Epoch 1 | Step: 109 | Loss: 2.4336\n",
      "Epoch 1 | Step: 110 | Loss: 3.5618\n",
      "Epoch 1 | Step: 111 | Loss: 2.8541\n",
      "Epoch 1 | Step: 112 | Loss: 2.8093\n",
      "Epoch 1 | Step: 113 | Loss: 2.7895\n",
      "Epoch 1 | Step: 114 | Loss: 8.2300\n",
      "Epoch 1 | Step: 115 | Loss: 2.8547\n",
      "Epoch 1 | Step: 116 | Loss: 2.5135\n",
      "Epoch 1 | Step: 117 | Loss: 3.5861\n",
      "Epoch 1 | Step: 118 | Loss: 2.0153\n",
      "Epoch 1 | Step: 119 | Loss: 2.9943\n",
      "Epoch 1 | Step: 120 | Loss: 3.0765\n",
      "Epoch 1 | Step: 121 | Loss: 4.3528\n",
      "Epoch 1 | Step: 122 | Loss: 2.3173\n",
      "Epoch 1 | Step: 123 | Loss: 2.6067\n",
      "Epoch 1 | Step: 124 | Loss: 2.4265\n",
      "Epoch 1 | Step: 125 | Loss: 2.5583\n",
      "Epoch 1 | Step: 126 | Loss: 3.8893\n",
      "Epoch 1 | Step: 127 | Loss: 2.8553\n",
      "Epoch 1 | Step: 128 | Loss: 3.2531\n",
      "Epoch 2 | Step: 1 | Loss: 2.3994\n",
      "Epoch 2 | Step: 2 | Loss: 2.2395\n",
      "Epoch 2 | Step: 3 | Loss: 2.8834\n",
      "Epoch 2 | Step: 4 | Loss: 9.6050\n",
      "Epoch 2 | Step: 5 | Loss: 2.9113\n",
      "Epoch 2 | Step: 6 | Loss: 2.1967\n",
      "Epoch 2 | Step: 7 | Loss: 2.5488\n",
      "Epoch 2 | Step: 8 | Loss: 2.5927\n",
      "Epoch 2 | Step: 9 | Loss: 2.3498\n",
      "Epoch 2 | Step: 10 | Loss: 3.2220\n",
      "Epoch 2 | Step: 11 | Loss: 4.6003\n",
      "Epoch 2 | Step: 12 | Loss: 2.5002\n",
      "Epoch 2 | Step: 13 | Loss: 3.8499\n",
      "Epoch 2 | Step: 14 | Loss: 1.8291\n",
      "Epoch 2 | Step: 15 | Loss: 2.3327\n",
      "Epoch 2 | Step: 16 | Loss: 2.4004\n",
      "Epoch 2 | Step: 17 | Loss: 4.4127\n",
      "Epoch 2 | Step: 18 | Loss: 2.3318\n",
      "Epoch 2 | Step: 19 | Loss: 1.5390\n",
      "Epoch 2 | Step: 20 | Loss: 2.2244\n",
      "Epoch 2 | Step: 21 | Loss: 1.8055\n",
      "Epoch 2 | Step: 22 | Loss: 6.2182\n",
      "Epoch 2 | Step: 23 | Loss: 2.5792\n",
      "Epoch 2 | Step: 24 | Loss: 3.0547\n",
      "Epoch 2 | Step: 25 | Loss: 3.1098\n",
      "Epoch 2 | Step: 26 | Loss: 3.0736\n",
      "Epoch 2 | Step: 27 | Loss: 2.1050\n",
      "Epoch 2 | Step: 28 | Loss: 2.2663\n",
      "Epoch 2 | Step: 29 | Loss: 1.9556\n",
      "Epoch 2 | Step: 30 | Loss: 2.4101\n",
      "Epoch 2 | Step: 31 | Loss: 2.7332\n",
      "Epoch 2 | Step: 32 | Loss: 2.4186\n",
      "Epoch 2 | Step: 33 | Loss: 2.3951\n",
      "Epoch 2 | Step: 34 | Loss: 1.6341\n",
      "Epoch 2 | Step: 35 | Loss: 3.2913\n",
      "Epoch 2 | Step: 36 | Loss: 2.9194\n",
      "Epoch 2 | Step: 37 | Loss: 2.3545\n",
      "Epoch 2 | Step: 38 | Loss: 3.5913\n",
      "Epoch 2 | Step: 39 | Loss: 2.8739\n",
      "Epoch 2 | Step: 40 | Loss: 2.5327\n",
      "Epoch 2 | Step: 41 | Loss: 2.8237\n",
      "Epoch 2 | Step: 42 | Loss: 1.9928\n",
      "Epoch 2 | Step: 43 | Loss: 3.0277\n",
      "Epoch 2 | Step: 44 | Loss: 10.0803\n",
      "Epoch 2 | Step: 45 | Loss: 1.8382\n",
      "Epoch 2 | Step: 46 | Loss: 2.7924\n",
      "Epoch 2 | Step: 47 | Loss: 2.6860\n",
      "Epoch 2 | Step: 48 | Loss: 1.8377\n",
      "Epoch 2 | Step: 49 | Loss: 2.7683\n",
      "Epoch 2 | Step: 50 | Loss: 2.2748\n",
      "Epoch 2 | Step: 51 | Loss: 2.6142\n",
      "Epoch 2 | Step: 52 | Loss: 1.4956\n",
      "Epoch 2 | Step: 53 | Loss: 2.0990\n",
      "Epoch 2 | Step: 54 | Loss: 3.1781\n",
      "Epoch 2 | Step: 55 | Loss: 2.8250\n",
      "Epoch 2 | Step: 56 | Loss: 2.4672\n",
      "Epoch 2 | Step: 57 | Loss: 2.6099\n",
      "Epoch 2 | Step: 58 | Loss: 2.7316\n",
      "Epoch 2 | Step: 59 | Loss: 1.9534\n",
      "Epoch 2 | Step: 60 | Loss: 2.3320\n",
      "Epoch 2 | Step: 61 | Loss: 3.0835\n",
      "Epoch 2 | Step: 62 | Loss: 1.9652\n",
      "Epoch 2 | Step: 63 | Loss: 2.9415\n",
      "Epoch 2 | Step: 64 | Loss: 2.4665\n",
      "Epoch 2 | Step: 65 | Loss: 1.9688\n",
      "Epoch 2 | Step: 66 | Loss: 2.2703\n",
      "Epoch 2 | Step: 67 | Loss: 2.6220\n",
      "Epoch 2 | Step: 68 | Loss: 2.9062\n",
      "Epoch 2 | Step: 69 | Loss: 2.7043\n",
      "Epoch 2 | Step: 70 | Loss: 2.2602\n",
      "Epoch 2 | Step: 71 | Loss: 3.0783\n",
      "Epoch 2 | Step: 72 | Loss: 2.9447\n",
      "Epoch 2 | Step: 73 | Loss: 2.7956\n",
      "Epoch 2 | Step: 74 | Loss: 2.7801\n",
      "Epoch 2 | Step: 75 | Loss: 2.2092\n",
      "Epoch 2 | Step: 76 | Loss: 2.9720\n",
      "Epoch 2 | Step: 77 | Loss: 2.0500\n",
      "Epoch 2 | Step: 78 | Loss: 2.9206\n",
      "Epoch 2 | Step: 79 | Loss: 10.1242\n",
      "Epoch 2 | Step: 80 | Loss: 2.2937\n",
      "Epoch 2 | Step: 81 | Loss: 2.5135\n",
      "Epoch 2 | Step: 82 | Loss: 2.4456\n",
      "Epoch 2 | Step: 83 | Loss: 1.8439\n",
      "Epoch 2 | Step: 84 | Loss: 1.8096\n",
      "Epoch 2 | Step: 85 | Loss: 2.6129\n",
      "Epoch 2 | Step: 86 | Loss: 2.1145\n",
      "Epoch 2 | Step: 87 | Loss: 2.3052\n",
      "Epoch 2 | Step: 88 | Loss: 7.1961\n",
      "Epoch 2 | Step: 89 | Loss: 2.6765\n",
      "Epoch 2 | Step: 90 | Loss: 1.5792\n",
      "Epoch 2 | Step: 91 | Loss: 2.1899\n",
      "Epoch 2 | Step: 92 | Loss: 1.8881\n",
      "Epoch 2 | Step: 93 | Loss: 2.0725\n",
      "Epoch 2 | Step: 94 | Loss: 2.0662\n",
      "Epoch 2 | Step: 95 | Loss: 1.9894\n",
      "Epoch 2 | Step: 96 | Loss: 2.0533\n",
      "Epoch 2 | Step: 97 | Loss: 1.8594\n",
      "Epoch 2 | Step: 98 | Loss: 2.0997\n",
      "Epoch 2 | Step: 99 | Loss: 2.5401\n",
      "Epoch 2 | Step: 100 | Loss: 3.3211\n",
      "Epoch 2 | Step: 101 | Loss: 3.5941\n",
      "Epoch 2 | Step: 102 | Loss: 2.7443\n",
      "Epoch 2 | Step: 103 | Loss: 1.8989\n",
      "Epoch 2 | Step: 104 | Loss: 2.9082\n",
      "Epoch 2 | Step: 105 | Loss: 1.9458\n",
      "Epoch 2 | Step: 106 | Loss: 1.7383\n",
      "Epoch 2 | Step: 107 | Loss: 2.1737\n",
      "Epoch 2 | Step: 108 | Loss: 1.5850\n",
      "Epoch 2 | Step: 109 | Loss: 1.8827\n",
      "Epoch 2 | Step: 110 | Loss: 2.3363\n",
      "Epoch 2 | Step: 111 | Loss: 2.7936\n",
      "Epoch 2 | Step: 112 | Loss: 2.6481\n",
      "Epoch 2 | Step: 113 | Loss: 3.0601\n",
      "Epoch 2 | Step: 114 | Loss: 3.0344\n",
      "Epoch 2 | Step: 115 | Loss: 2.8810\n",
      "Epoch 2 | Step: 116 | Loss: 1.4449\n",
      "Epoch 2 | Step: 117 | Loss: 1.8657\n",
      "Epoch 2 | Step: 118 | Loss: 2.8296\n",
      "Epoch 2 | Step: 119 | Loss: 1.6915\n",
      "Epoch 2 | Step: 120 | Loss: 2.5267\n",
      "Epoch 2 | Step: 121 | Loss: 2.3539\n",
      "Epoch 2 | Step: 122 | Loss: 2.0836\n",
      "Epoch 2 | Step: 123 | Loss: 1.6686\n",
      "Epoch 2 | Step: 124 | Loss: 1.6452\n",
      "Epoch 2 | Step: 125 | Loss: 1.5393\n",
      "Epoch 2 | Step: 126 | Loss: 1.9739\n",
      "Epoch 2 | Step: 127 | Loss: 1.5093\n",
      "Epoch 2 | Step: 128 | Loss: 1.7751\n",
      "Epoch 3 | Step: 1 | Loss: 2.1249\n",
      "Epoch 3 | Step: 2 | Loss: 2.6815\n",
      "Epoch 3 | Step: 3 | Loss: 1.0042\n",
      "Epoch 3 | Step: 4 | Loss: 2.4350\n",
      "Epoch 3 | Step: 5 | Loss: 2.1184\n",
      "Epoch 3 | Step: 6 | Loss: 3.1188\n",
      "Epoch 3 | Step: 7 | Loss: 2.7498\n",
      "Epoch 3 | Step: 8 | Loss: 2.4727\n",
      "Epoch 3 | Step: 9 | Loss: 2.1907\n",
      "Epoch 3 | Step: 10 | Loss: 2.0780\n",
      "Epoch 3 | Step: 11 | Loss: 2.6849\n",
      "Epoch 3 | Step: 12 | Loss: 2.0877\n",
      "Epoch 3 | Step: 13 | Loss: 2.3094\n",
      "Epoch 3 | Step: 14 | Loss: 1.6508\n",
      "Epoch 3 | Step: 15 | Loss: 1.6176\n",
      "Epoch 3 | Step: 16 | Loss: 2.1248\n",
      "Epoch 3 | Step: 17 | Loss: 1.9411\n",
      "Epoch 3 | Step: 18 | Loss: 1.6850\n",
      "Epoch 3 | Step: 19 | Loss: 1.9802\n",
      "Epoch 3 | Step: 20 | Loss: 2.1636\n",
      "Epoch 3 | Step: 21 | Loss: 1.9577\n",
      "Epoch 3 | Step: 22 | Loss: 2.3193\n",
      "Epoch 3 | Step: 23 | Loss: 1.9209\n",
      "Epoch 3 | Step: 24 | Loss: 2.0321\n",
      "Epoch 3 | Step: 25 | Loss: 1.5831\n",
      "Epoch 3 | Step: 26 | Loss: 1.6760\n",
      "Epoch 3 | Step: 27 | Loss: 2.0254\n",
      "Epoch 3 | Step: 28 | Loss: 1.8809\n",
      "Epoch 3 | Step: 29 | Loss: 1.7026\n",
      "Epoch 3 | Step: 30 | Loss: 1.3671\n",
      "Epoch 3 | Step: 31 | Loss: 3.1092\n",
      "Epoch 3 | Step: 32 | Loss: 1.8515\n",
      "Epoch 3 | Step: 33 | Loss: 2.0948\n",
      "Epoch 3 | Step: 34 | Loss: 2.0605\n",
      "Epoch 3 | Step: 35 | Loss: 1.6790\n",
      "Epoch 3 | Step: 36 | Loss: 1.8550\n",
      "Epoch 3 | Step: 37 | Loss: 4.5864\n",
      "Epoch 3 | Step: 38 | Loss: 1.9882\n",
      "Epoch 3 | Step: 39 | Loss: 1.5265\n",
      "Epoch 3 | Step: 40 | Loss: 2.2650\n",
      "Epoch 3 | Step: 41 | Loss: 1.9421\n",
      "Epoch 3 | Step: 42 | Loss: 1.8881\n",
      "Epoch 3 | Step: 43 | Loss: 3.3307\n",
      "Epoch 3 | Step: 44 | Loss: 2.1031\n",
      "Epoch 3 | Step: 45 | Loss: 2.1711\n",
      "Epoch 3 | Step: 46 | Loss: 2.0037\n",
      "Epoch 3 | Step: 47 | Loss: 2.5097\n",
      "Epoch 3 | Step: 48 | Loss: 2.3870\n",
      "Epoch 3 | Step: 49 | Loss: 1.7322\n",
      "Epoch 3 | Step: 50 | Loss: 2.3367\n",
      "Epoch 3 | Step: 51 | Loss: 2.0395\n",
      "Epoch 3 | Step: 52 | Loss: 2.2244\n",
      "Epoch 3 | Step: 53 | Loss: 1.7039\n",
      "Epoch 3 | Step: 54 | Loss: 2.3002\n",
      "Epoch 3 | Step: 55 | Loss: 2.3030\n",
      "Epoch 3 | Step: 56 | Loss: 1.9684\n",
      "Epoch 3 | Step: 57 | Loss: 2.0390\n",
      "Epoch 3 | Step: 58 | Loss: 1.6103\n",
      "Epoch 3 | Step: 59 | Loss: 1.7588\n",
      "Epoch 3 | Step: 60 | Loss: 1.7142\n",
      "Epoch 3 | Step: 61 | Loss: 1.8973\n",
      "Epoch 3 | Step: 62 | Loss: 2.2088\n",
      "Epoch 3 | Step: 63 | Loss: 1.6294\n",
      "Epoch 3 | Step: 64 | Loss: 2.6709\n",
      "Epoch 3 | Step: 65 | Loss: 8.3148\n",
      "Epoch 3 | Step: 66 | Loss: 1.4829\n",
      "Epoch 3 | Step: 67 | Loss: 2.0455\n",
      "Epoch 3 | Step: 68 | Loss: 2.1861\n",
      "Epoch 3 | Step: 69 | Loss: 2.1448\n",
      "Epoch 3 | Step: 70 | Loss: 1.4197\n",
      "Epoch 3 | Step: 71 | Loss: 1.9546\n",
      "Epoch 3 | Step: 72 | Loss: 1.8628\n",
      "Epoch 3 | Step: 73 | Loss: 1.6590\n",
      "Epoch 3 | Step: 74 | Loss: 1.5571\n",
      "Epoch 3 | Step: 75 | Loss: 1.7197\n",
      "Epoch 3 | Step: 76 | Loss: 1.6537\n",
      "Epoch 3 | Step: 77 | Loss: 1.2607\n",
      "Epoch 3 | Step: 78 | Loss: 2.5162\n",
      "Epoch 3 | Step: 79 | Loss: 1.7129\n",
      "Epoch 3 | Step: 80 | Loss: 1.6022\n",
      "Epoch 3 | Step: 81 | Loss: 1.7105\n",
      "Epoch 3 | Step: 82 | Loss: 1.6383\n",
      "Epoch 3 | Step: 83 | Loss: 1.8828\n",
      "Epoch 3 | Step: 84 | Loss: 1.7792\n",
      "Epoch 3 | Step: 85 | Loss: 2.0953\n",
      "Epoch 3 | Step: 86 | Loss: 5.9308\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 3 | Step: 87 | Loss: 1.9410\n",
      "Epoch 3 | Step: 88 | Loss: 1.8121\n",
      "Epoch 3 | Step: 89 | Loss: 1.6975\n",
      "Epoch 3 | Step: 90 | Loss: 2.2283\n",
      "Epoch 3 | Step: 91 | Loss: 1.8878\n",
      "Epoch 3 | Step: 92 | Loss: 1.9148\n",
      "Epoch 3 | Step: 93 | Loss: 2.2554\n",
      "Epoch 3 | Step: 94 | Loss: 1.7951\n",
      "Epoch 3 | Step: 95 | Loss: 1.3730\n",
      "Epoch 3 | Step: 96 | Loss: 1.5003\n",
      "Epoch 3 | Step: 97 | Loss: 2.1048\n",
      "Epoch 3 | Step: 98 | Loss: 1.7808\n",
      "Epoch 3 | Step: 99 | Loss: 1.7954\n",
      "Epoch 3 | Step: 100 | Loss: 2.1036\n",
      "Epoch 3 | Step: 101 | Loss: 1.1415\n",
      "Epoch 3 | Step: 102 | Loss: 1.5767\n",
      "Epoch 3 | Step: 103 | Loss: 1.2394\n",
      "Epoch 3 | Step: 104 | Loss: 1.4115\n",
      "Epoch 3 | Step: 105 | Loss: 2.4677\n",
      "Epoch 3 | Step: 106 | Loss: 8.1301\n",
      "Epoch 3 | Step: 107 | Loss: 1.4481\n",
      "Epoch 3 | Step: 108 | Loss: 1.2065\n",
      "Epoch 3 | Step: 109 | Loss: 1.9491\n",
      "Epoch 3 | Step: 110 | Loss: 2.0507\n",
      "Epoch 3 | Step: 111 | Loss: 2.1518\n",
      "Epoch 3 | Step: 112 | Loss: 1.7572\n",
      "Epoch 3 | Step: 113 | Loss: 1.6113\n",
      "Epoch 3 | Step: 114 | Loss: 1.4970\n",
      "Epoch 3 | Step: 115 | Loss: 1.6638\n",
      "Epoch 3 | Step: 116 | Loss: 5.1477\n",
      "Epoch 3 | Step: 117 | Loss: 2.2727\n",
      "Epoch 3 | Step: 118 | Loss: 1.6361\n",
      "Epoch 3 | Step: 119 | Loss: 1.8627\n",
      "Epoch 3 | Step: 120 | Loss: 2.3830\n",
      "Epoch 3 | Step: 121 | Loss: 1.9688\n",
      "Epoch 3 | Step: 122 | Loss: 1.6530\n",
      "Epoch 3 | Step: 123 | Loss: 1.7259\n",
      "Epoch 3 | Step: 124 | Loss: 2.0495\n",
      "Epoch 3 | Step: 125 | Loss: 1.4551\n",
      "Epoch 3 | Step: 126 | Loss: 1.9075\n",
      "Epoch 3 | Step: 127 | Loss: 1.4238\n",
      "Epoch 3 | Step: 128 | Loss: 1.2563\n",
      "Epoch 4 | Step: 1 | Loss: 2.0650\n",
      "Epoch 4 | Step: 2 | Loss: 2.2573\n",
      "Epoch 4 | Step: 3 | Loss: 1.4310\n",
      "Epoch 4 | Step: 4 | Loss: 1.6567\n",
      "Epoch 4 | Step: 5 | Loss: 1.9581\n",
      "Epoch 4 | Step: 6 | Loss: 1.6552\n",
      "Epoch 4 | Step: 7 | Loss: 2.0632\n",
      "Epoch 4 | Step: 8 | Loss: 2.1051\n",
      "Epoch 4 | Step: 9 | Loss: 1.8077\n",
      "Epoch 4 | Step: 10 | Loss: 1.6284\n",
      "Epoch 4 | Step: 11 | Loss: 1.5138\n",
      "Epoch 4 | Step: 12 | Loss: 1.8620\n",
      "Epoch 4 | Step: 13 | Loss: 1.5378\n",
      "Epoch 4 | Step: 14 | Loss: 1.3818\n",
      "Epoch 4 | Step: 15 | Loss: 1.1963\n",
      "Epoch 4 | Step: 16 | Loss: 1.3502\n",
      "Epoch 4 | Step: 17 | Loss: 1.7680\n",
      "Epoch 4 | Step: 18 | Loss: 1.1570\n",
      "Epoch 4 | Step: 19 | Loss: 1.4306\n",
      "Epoch 4 | Step: 20 | Loss: 1.1765\n",
      "Epoch 4 | Step: 21 | Loss: 1.7468\n",
      "Epoch 4 | Step: 22 | Loss: 1.9647\n",
      "Epoch 4 | Step: 23 | Loss: 2.1079\n",
      "Epoch 4 | Step: 24 | Loss: 1.3640\n",
      "Epoch 4 | Step: 25 | Loss: 1.1762\n",
      "Epoch 4 | Step: 26 | Loss: 1.9766\n",
      "Epoch 4 | Step: 27 | Loss: 1.0019\n",
      "Epoch 4 | Step: 28 | Loss: 1.9184\n",
      "Epoch 4 | Step: 29 | Loss: 1.7837\n",
      "Epoch 4 | Step: 30 | Loss: 1.1264\n",
      "Epoch 4 | Step: 31 | Loss: 1.1922\n",
      "Epoch 4 | Step: 32 | Loss: 1.9439\n",
      "Epoch 4 | Step: 33 | Loss: 4.9674\n",
      "Epoch 4 | Step: 34 | Loss: 1.5409\n",
      "Epoch 4 | Step: 35 | Loss: 1.5803\n",
      "Epoch 4 | Step: 36 | Loss: 1.2657\n",
      "Epoch 4 | Step: 37 | Loss: 1.8006\n",
      "Epoch 4 | Step: 38 | Loss: 1.0849\n",
      "Epoch 4 | Step: 39 | Loss: 1.3771\n",
      "Epoch 4 | Step: 40 | Loss: 1.2090\n",
      "Epoch 4 | Step: 41 | Loss: 1.4718\n",
      "Epoch 4 | Step: 42 | Loss: 1.5752\n",
      "Epoch 4 | Step: 43 | Loss: 1.8764\n",
      "Epoch 4 | Step: 44 | Loss: 1.3305\n",
      "Epoch 4 | Step: 45 | Loss: 1.5076\n",
      "Epoch 4 | Step: 46 | Loss: 1.5242\n",
      "Epoch 4 | Step: 47 | Loss: 1.4391\n",
      "Epoch 4 | Step: 48 | Loss: 2.1724\n",
      "Epoch 4 | Step: 49 | Loss: 6.4722\n",
      "Epoch 4 | Step: 50 | Loss: 1.5054\n",
      "Epoch 4 | Step: 51 | Loss: 1.3330\n",
      "Epoch 4 | Step: 52 | Loss: 1.6076\n",
      "Epoch 4 | Step: 53 | Loss: 1.5117\n",
      "Epoch 4 | Step: 54 | Loss: 1.8942\n",
      "Epoch 4 | Step: 55 | Loss: 1.4920\n",
      "Epoch 4 | Step: 56 | Loss: 1.5078\n",
      "Epoch 4 | Step: 57 | Loss: 1.5427\n",
      "Epoch 4 | Step: 58 | Loss: 1.3179\n",
      "Epoch 4 | Step: 59 | Loss: 1.9268\n",
      "Epoch 4 | Step: 60 | Loss: 1.7697\n",
      "Epoch 4 | Step: 61 | Loss: 1.6980\n",
      "Epoch 4 | Step: 62 | Loss: 2.2931\n",
      "Epoch 4 | Step: 63 | Loss: 2.2356\n",
      "Epoch 4 | Step: 64 | Loss: 1.7362\n",
      "Epoch 4 | Step: 65 | Loss: 1.8598\n",
      "Epoch 4 | Step: 66 | Loss: 1.9454\n",
      "Epoch 4 | Step: 67 | Loss: 2.0176\n",
      "Epoch 4 | Step: 68 | Loss: 6.4245\n",
      "Epoch 4 | Step: 69 | Loss: 1.8633\n",
      "Epoch 4 | Step: 70 | Loss: 2.1548\n",
      "Epoch 4 | Step: 71 | Loss: 1.5636\n",
      "Epoch 4 | Step: 72 | Loss: 1.0596\n",
      "Epoch 4 | Step: 73 | Loss: 1.2975\n",
      "Epoch 4 | Step: 74 | Loss: 1.2468\n",
      "Epoch 4 | Step: 75 | Loss: 1.4885\n",
      "Epoch 4 | Step: 76 | Loss: 1.6138\n",
      "Epoch 4 | Step: 77 | Loss: 1.8351\n",
      "Epoch 4 | Step: 78 | Loss: 1.2667\n",
      "Epoch 4 | Step: 79 | Loss: 1.0998\n",
      "Epoch 4 | Step: 80 | Loss: 1.7498\n",
      "Epoch 4 | Step: 81 | Loss: 1.1567\n",
      "Epoch 4 | Step: 82 | Loss: 1.3392\n",
      "Epoch 4 | Step: 83 | Loss: 1.8584\n",
      "Epoch 4 | Step: 84 | Loss: 1.3343\n",
      "Epoch 4 | Step: 85 | Loss: 2.2011\n",
      "Epoch 4 | Step: 86 | Loss: 1.5706\n",
      "Epoch 4 | Step: 87 | Loss: 1.4771\n",
      "Epoch 4 | Step: 88 | Loss: 1.4258\n",
      "Epoch 4 | Step: 89 | Loss: 1.3326\n",
      "Epoch 4 | Step: 90 | Loss: 2.1494\n",
      "Epoch 4 | Step: 91 | Loss: 0.9390\n",
      "Epoch 4 | Step: 92 | Loss: 1.3280\n",
      "Epoch 4 | Step: 93 | Loss: 1.3930\n",
      "Epoch 4 | Step: 94 | Loss: 2.2797\n",
      "Epoch 4 | Step: 95 | Loss: 1.5016\n",
      "Epoch 4 | Step: 96 | Loss: 2.1862\n",
      "Epoch 4 | Step: 97 | Loss: 1.2928\n",
      "Epoch 4 | Step: 98 | Loss: 1.8153\n",
      "Epoch 4 | Step: 99 | Loss: 1.1804\n",
      "Epoch 4 | Step: 100 | Loss: 1.1686\n",
      "Epoch 4 | Step: 101 | Loss: 1.5760\n",
      "Epoch 4 | Step: 102 | Loss: 1.1753\n",
      "Epoch 4 | Step: 103 | Loss: 1.6046\n",
      "Epoch 4 | Step: 104 | Loss: 1.4451\n",
      "Epoch 4 | Step: 105 | Loss: 1.5456\n",
      "Epoch 4 | Step: 106 | Loss: 1.2031\n",
      "Epoch 4 | Step: 107 | Loss: 1.5669\n",
      "Epoch 4 | Step: 108 | Loss: 1.7940\n",
      "Epoch 4 | Step: 109 | Loss: 1.4400\n",
      "Epoch 4 | Step: 110 | Loss: 1.6588\n",
      "Epoch 4 | Step: 111 | Loss: 1.0600\n",
      "Epoch 4 | Step: 112 | Loss: 1.0708\n",
      "Epoch 4 | Step: 113 | Loss: 1.4710\n",
      "Epoch 4 | Step: 114 | Loss: 1.8635\n",
      "Epoch 4 | Step: 115 | Loss: 1.1534\n",
      "Epoch 4 | Step: 116 | Loss: 1.2914\n",
      "Epoch 4 | Step: 117 | Loss: 1.8091\n",
      "Epoch 4 | Step: 118 | Loss: 1.6168\n",
      "Epoch 4 | Step: 119 | Loss: 1.9245\n",
      "Epoch 4 | Step: 120 | Loss: 1.6040\n",
      "Epoch 4 | Step: 121 | Loss: 1.5323\n",
      "Epoch 4 | Step: 122 | Loss: 3.9233\n",
      "Epoch 4 | Step: 123 | Loss: 1.2494\n",
      "Epoch 4 | Step: 124 | Loss: 3.1393\n",
      "Epoch 4 | Step: 125 | Loss: 2.3394\n",
      "Epoch 4 | Step: 126 | Loss: 1.1392\n",
      "Epoch 4 | Step: 127 | Loss: 1.4666\n",
      "Epoch 4 | Step: 128 | Loss: 1.0036\n",
      "Epoch 5 | Step: 1 | Loss: 1.4813\n",
      "Epoch 5 | Step: 2 | Loss: 1.4959\n",
      "Epoch 5 | Step: 3 | Loss: 1.6113\n",
      "Epoch 5 | Step: 4 | Loss: 5.7352\n",
      "Epoch 5 | Step: 5 | Loss: 1.3893\n",
      "Epoch 5 | Step: 6 | Loss: 0.9928\n",
      "Epoch 5 | Step: 7 | Loss: 1.3315\n",
      "Epoch 5 | Step: 8 | Loss: 0.8711\n",
      "Epoch 5 | Step: 9 | Loss: 0.9891\n",
      "Epoch 5 | Step: 10 | Loss: 1.6739\n",
      "Epoch 5 | Step: 11 | Loss: 1.3796\n",
      "Epoch 5 | Step: 12 | Loss: 1.4214\n",
      "Epoch 5 | Step: 13 | Loss: 1.0841\n",
      "Epoch 5 | Step: 14 | Loss: 1.6789\n",
      "Epoch 5 | Step: 15 | Loss: 1.5800\n",
      "Epoch 5 | Step: 16 | Loss: 2.0209\n",
      "Epoch 5 | Step: 17 | Loss: 0.9746\n",
      "Epoch 5 | Step: 18 | Loss: 1.1162\n",
      "Epoch 5 | Step: 19 | Loss: 1.5620\n",
      "Epoch 5 | Step: 20 | Loss: 1.1576\n",
      "Epoch 5 | Step: 21 | Loss: 1.6299\n",
      "Epoch 5 | Step: 22 | Loss: 1.1855\n",
      "Epoch 5 | Step: 23 | Loss: 1.8421\n",
      "Epoch 5 | Step: 24 | Loss: 1.4316\n",
      "Epoch 5 | Step: 25 | Loss: 1.3051\n",
      "Epoch 5 | Step: 26 | Loss: 1.2208\n",
      "Epoch 5 | Step: 27 | Loss: 2.0864\n",
      "Epoch 5 | Step: 28 | Loss: 1.4712\n",
      "Epoch 5 | Step: 29 | Loss: 0.9503\n",
      "Epoch 5 | Step: 30 | Loss: 0.8262\n",
      "Epoch 5 | Step: 31 | Loss: 3.9768\n",
      "Epoch 5 | Step: 32 | Loss: 1.3175\n",
      "Epoch 5 | Step: 33 | Loss: 1.8288\n",
      "Epoch 5 | Step: 34 | Loss: 1.5099\n",
      "Epoch 5 | Step: 35 | Loss: 1.6032\n",
      "Epoch 5 | Step: 36 | Loss: 1.3521\n",
      "Epoch 5 | Step: 37 | Loss: 1.9436\n",
      "Epoch 5 | Step: 38 | Loss: 1.2330\n",
      "Epoch 5 | Step: 39 | Loss: 1.4092\n",
      "Epoch 5 | Step: 40 | Loss: 1.6954\n",
      "Epoch 5 | Step: 41 | Loss: 1.5735\n",
      "Epoch 5 | Step: 42 | Loss: 1.0936\n",
      "Epoch 5 | Step: 43 | Loss: 1.2183\n",
      "Epoch 5 | Step: 44 | Loss: 1.2088\n",
      "Epoch 5 | Step: 45 | Loss: 1.8614\n",
      "Epoch 5 | Step: 46 | Loss: 1.0058\n",
      "Epoch 5 | Step: 47 | Loss: 1.4139\n",
      "Epoch 5 | Step: 48 | Loss: 1.1769\n",
      "Epoch 5 | Step: 49 | Loss: 1.3753\n",
      "Epoch 5 | Step: 50 | Loss: 1.5058\n",
      "Epoch 5 | Step: 51 | Loss: 1.5563\n",
      "Epoch 5 | Step: 52 | Loss: 1.2576\n",
      "Epoch 5 | Step: 53 | Loss: 1.1344\n",
      "Epoch 5 | Step: 54 | Loss: 1.1569\n",
      "Epoch 5 | Step: 55 | Loss: 1.3159\n",
      "Epoch 5 | Step: 56 | Loss: 1.2952\n",
      "Epoch 5 | Step: 57 | Loss: 1.3087\n",
      "Epoch 5 | Step: 58 | Loss: 1.1700\n",
      "Epoch 5 | Step: 59 | Loss: 1.0986\n",
      "Epoch 5 | Step: 60 | Loss: 1.3782\n",
      "Epoch 5 | Step: 61 | Loss: 1.2383\n",
      "Epoch 5 | Step: 62 | Loss: 1.2691\n",
      "Epoch 5 | Step: 63 | Loss: 1.3155\n",
      "Epoch 5 | Step: 64 | Loss: 1.1792\n",
      "Epoch 5 | Step: 65 | Loss: 1.4649\n",
      "Epoch 5 | Step: 66 | Loss: 1.7837\n",
      "Epoch 5 | Step: 67 | Loss: 2.5132\n",
      "Epoch 5 | Step: 68 | Loss: 0.9064\n",
      "Epoch 5 | Step: 69 | Loss: 1.3118\n",
      "Epoch 5 | Step: 70 | Loss: 1.3135\n",
      "Epoch 5 | Step: 71 | Loss: 0.9332\n",
      "Epoch 5 | Step: 72 | Loss: 1.3188\n",
      "Epoch 5 | Step: 73 | Loss: 1.1626\n",
      "Epoch 5 | Step: 74 | Loss: 1.1333\n",
      "Epoch 5 | Step: 75 | Loss: 1.5730\n",
      "Epoch 5 | Step: 76 | Loss: 1.1316\n",
      "Epoch 5 | Step: 77 | Loss: 1.0645\n",
      "Epoch 5 | Step: 78 | Loss: 1.3292\n",
      "Epoch 5 | Step: 79 | Loss: 1.4304\n",
      "Epoch 5 | Step: 80 | Loss: 1.3810\n",
      "Epoch 5 | Step: 81 | Loss: 1.1864\n",
      "Epoch 5 | Step: 82 | Loss: 1.3670\n",
      "Epoch 5 | Step: 83 | Loss: 1.4562\n",
      "Epoch 5 | Step: 84 | Loss: 1.1114\n",
      "Epoch 5 | Step: 85 | Loss: 1.9653\n",
      "Epoch 5 | Step: 86 | Loss: 1.3027\n",
      "Epoch 5 | Step: 87 | Loss: 0.7636\n",
      "Epoch 5 | Step: 88 | Loss: 2.2920\n",
      "Epoch 5 | Step: 89 | Loss: 1.8885\n",
      "Epoch 5 | Step: 90 | Loss: 1.4481\n",
      "Epoch 5 | Step: 91 | Loss: 1.1865\n",
      "Epoch 5 | Step: 92 | Loss: 1.0149\n",
      "Epoch 5 | Step: 93 | Loss: 1.4082\n",
      "Epoch 5 | Step: 94 | Loss: 0.9242\n",
      "Epoch 5 | Step: 95 | Loss: 1.1131\n",
      "Epoch 5 | Step: 96 | Loss: 1.0033\n",
      "Epoch 5 | Step: 97 | Loss: 0.9262\n",
      "Epoch 5 | Step: 98 | Loss: 1.4741\n",
      "Epoch 5 | Step: 99 | Loss: 3.4141\n",
      "Epoch 5 | Step: 100 | Loss: 4.3325\n",
      "Epoch 5 | Step: 101 | Loss: 1.3445\n",
      "Epoch 5 | Step: 102 | Loss: 1.5685\n",
      "Epoch 5 | Step: 103 | Loss: 1.4158\n",
      "Epoch 5 | Step: 104 | Loss: 0.8650\n",
      "Epoch 5 | Step: 105 | Loss: 1.2705\n",
      "Epoch 5 | Step: 106 | Loss: 1.2785\n",
      "Epoch 5 | Step: 107 | Loss: 0.7136\n",
      "Epoch 5 | Step: 108 | Loss: 0.9958\n",
      "Epoch 5 | Step: 109 | Loss: 1.0495\n",
      "Epoch 5 | Step: 110 | Loss: 1.1492\n",
      "Epoch 5 | Step: 111 | Loss: 1.2290\n",
      "Epoch 5 | Step: 112 | Loss: 1.0218\n",
      "Epoch 5 | Step: 113 | Loss: 1.4589\n",
      "Epoch 5 | Step: 114 | Loss: 0.9431\n",
      "Epoch 5 | Step: 115 | Loss: 1.1583\n",
      "Epoch 5 | Step: 116 | Loss: 1.2564\n",
      "Epoch 5 | Step: 117 | Loss: 1.4276\n",
      "Epoch 5 | Step: 118 | Loss: 1.0793\n",
      "Epoch 5 | Step: 119 | Loss: 1.2617\n",
      "Epoch 5 | Step: 120 | Loss: 1.1253\n",
      "Epoch 5 | Step: 121 | Loss: 1.0203\n",
      "Epoch 5 | Step: 122 | Loss: 1.6563\n",
      "Epoch 5 | Step: 123 | Loss: 1.2507\n",
      "Epoch 5 | Step: 124 | Loss: 1.2238\n",
      "Epoch 5 | Step: 125 | Loss: 1.2591\n",
      "Epoch 5 | Step: 126 | Loss: 0.9814\n",
      "Epoch 5 | Step: 127 | Loss: 0.7309\n",
      "Epoch 5 | Step: 128 | Loss: 1.5076\n",
      "Epoch 6 | Step: 1 | Loss: 1.6058\n",
      "Epoch 6 | Step: 2 | Loss: 1.4001\n",
      "Epoch 6 | Step: 3 | Loss: 1.1249\n",
      "Epoch 6 | Step: 4 | Loss: 1.3203\n",
      "Epoch 6 | Step: 5 | Loss: 0.8709\n",
      "Epoch 6 | Step: 6 | Loss: 1.2647\n",
      "Epoch 6 | Step: 7 | Loss: 1.1077\n",
      "Epoch 6 | Step: 8 | Loss: 1.1026\n",
      "Epoch 6 | Step: 9 | Loss: 0.8538\n",
      "Epoch 6 | Step: 10 | Loss: 1.1017\n",
      "Epoch 6 | Step: 11 | Loss: 1.0162\n",
      "Epoch 6 | Step: 12 | Loss: 1.3978\n",
      "Epoch 6 | Step: 13 | Loss: 1.2621\n",
      "Epoch 6 | Step: 14 | Loss: 1.3524\n",
      "Epoch 6 | Step: 15 | Loss: 1.1243\n",
      "Epoch 6 | Step: 16 | Loss: 1.2521\n",
      "Epoch 6 | Step: 17 | Loss: 0.9411\n",
      "Epoch 6 | Step: 18 | Loss: 0.8517\n",
      "Epoch 6 | Step: 19 | Loss: 1.5070\n",
      "Epoch 6 | Step: 20 | Loss: 1.5188\n",
      "Epoch 6 | Step: 21 | Loss: 1.0023\n",
      "Epoch 6 | Step: 22 | Loss: 0.7439\n",
      "Epoch 6 | Step: 23 | Loss: 0.9842\n",
      "Epoch 6 | Step: 24 | Loss: 0.9619\n",
      "Epoch 6 | Step: 25 | Loss: 4.4065\n",
      "Epoch 6 | Step: 26 | Loss: 1.0611\n",
      "Epoch 6 | Step: 27 | Loss: 0.9638\n",
      "Epoch 6 | Step: 28 | Loss: 0.9180\n",
      "Epoch 6 | Step: 29 | Loss: 1.0049\n",
      "Epoch 6 | Step: 30 | Loss: 1.5384\n",
      "Epoch 6 | Step: 31 | Loss: 1.2569\n",
      "Epoch 6 | Step: 32 | Loss: 1.5044\n",
      "Epoch 6 | Step: 33 | Loss: 0.7060\n",
      "Epoch 6 | Step: 34 | Loss: 1.0483\n",
      "Epoch 6 | Step: 35 | Loss: 0.5006\n",
      "Epoch 6 | Step: 36 | Loss: 2.4933\n",
      "Epoch 6 | Step: 37 | Loss: 1.4499\n",
      "Epoch 6 | Step: 38 | Loss: 1.1903\n",
      "Epoch 6 | Step: 39 | Loss: 1.1751\n",
      "Epoch 6 | Step: 40 | Loss: 1.2514\n",
      "Epoch 6 | Step: 41 | Loss: 1.6004\n",
      "Epoch 6 | Step: 42 | Loss: 1.5558\n",
      "Epoch 6 | Step: 43 | Loss: 0.9958\n",
      "Epoch 6 | Step: 44 | Loss: 1.0055\n",
      "Epoch 6 | Step: 45 | Loss: 2.8817\n",
      "Epoch 6 | Step: 46 | Loss: 0.8293\n",
      "Epoch 6 | Step: 47 | Loss: 1.1782\n",
      "Epoch 6 | Step: 48 | Loss: 1.0891\n",
      "Epoch 6 | Step: 49 | Loss: 1.0107\n",
      "Epoch 6 | Step: 50 | Loss: 0.8704\n",
      "Epoch 6 | Step: 51 | Loss: 1.1197\n",
      "Epoch 6 | Step: 52 | Loss: 1.0434\n",
      "Epoch 6 | Step: 53 | Loss: 1.5553\n",
      "Epoch 6 | Step: 54 | Loss: 1.2134\n",
      "Epoch 6 | Step: 55 | Loss: 1.0639\n",
      "Epoch 6 | Step: 56 | Loss: 1.0720\n",
      "Epoch 6 | Step: 57 | Loss: 1.1405\n",
      "Epoch 6 | Step: 58 | Loss: 1.5323\n",
      "Epoch 6 | Step: 59 | Loss: 1.1136\n",
      "Epoch 6 | Step: 60 | Loss: 0.9357\n",
      "Epoch 6 | Step: 61 | Loss: 1.0037\n",
      "Epoch 6 | Step: 62 | Loss: 0.9937\n",
      "Epoch 6 | Step: 63 | Loss: 1.1117\n",
      "Epoch 6 | Step: 64 | Loss: 0.9329\n",
      "Epoch 6 | Step: 65 | Loss: 1.3252\n",
      "Epoch 6 | Step: 66 | Loss: 1.2253\n",
      "Epoch 6 | Step: 67 | Loss: 1.2872\n",
      "Epoch 6 | Step: 68 | Loss: 1.1292\n",
      "Epoch 6 | Step: 69 | Loss: 1.0314\n",
      "Epoch 6 | Step: 70 | Loss: 0.7818\n",
      "Epoch 6 | Step: 71 | Loss: 0.9144\n",
      "Epoch 6 | Step: 72 | Loss: 0.8272\n",
      "Epoch 6 | Step: 73 | Loss: 1.1867\n",
      "Epoch 6 | Step: 74 | Loss: 1.3248\n",
      "Epoch 6 | Step: 75 | Loss: 1.0545\n",
      "Epoch 6 | Step: 76 | Loss: 1.0894\n",
      "Epoch 6 | Step: 77 | Loss: 1.3241\n",
      "Epoch 6 | Step: 78 | Loss: 1.4272\n",
      "Epoch 6 | Step: 79 | Loss: 0.7460\n",
      "Epoch 6 | Step: 80 | Loss: 0.6082\n",
      "Epoch 6 | Step: 81 | Loss: 1.0674\n",
      "Epoch 6 | Step: 82 | Loss: 1.4643\n",
      "Epoch 6 | Step: 83 | Loss: 0.8023\n",
      "Epoch 6 | Step: 84 | Loss: 1.0447\n",
      "Epoch 6 | Step: 85 | Loss: 0.9188\n",
      "Epoch 6 | Step: 86 | Loss: 1.0150\n",
      "Epoch 6 | Step: 87 | Loss: 0.9451\n",
      "Epoch 6 | Step: 88 | Loss: 1.1767\n",
      "Epoch 6 | Step: 89 | Loss: 0.8135\n",
      "Epoch 6 | Step: 90 | Loss: 1.5116\n",
      "Epoch 6 | Step: 91 | Loss: 1.4355\n",
      "Epoch 6 | Step: 92 | Loss: 4.7046\n",
      "Epoch 6 | Step: 93 | Loss: 0.9015\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 6 | Step: 94 | Loss: 0.6471\n",
      "Epoch 6 | Step: 95 | Loss: 1.0454\n",
      "Epoch 6 | Step: 96 | Loss: 0.8400\n",
      "Epoch 6 | Step: 97 | Loss: 0.9036\n",
      "Epoch 6 | Step: 98 | Loss: 1.0402\n",
      "Epoch 6 | Step: 99 | Loss: 0.7784\n",
      "Epoch 6 | Step: 100 | Loss: 0.9806\n",
      "Epoch 6 | Step: 101 | Loss: 0.8733\n",
      "Epoch 6 | Step: 102 | Loss: 1.5086\n",
      "Epoch 6 | Step: 103 | Loss: 1.0038\n",
      "Epoch 6 | Step: 104 | Loss: 1.0515\n",
      "Epoch 6 | Step: 105 | Loss: 0.8477\n",
      "Epoch 6 | Step: 106 | Loss: 2.5423\n",
      "Epoch 6 | Step: 107 | Loss: 1.4057\n",
      "Epoch 6 | Step: 108 | Loss: 0.9667\n",
      "Epoch 6 | Step: 109 | Loss: 0.9984\n",
      "Epoch 6 | Step: 110 | Loss: 1.0484\n",
      "Epoch 6 | Step: 111 | Loss: 1.4752\n",
      "Epoch 6 | Step: 112 | Loss: 1.2457\n",
      "Epoch 6 | Step: 113 | Loss: 0.8885\n",
      "Epoch 6 | Step: 114 | Loss: 1.3874\n",
      "Epoch 6 | Step: 115 | Loss: 0.9904\n",
      "Epoch 6 | Step: 116 | Loss: 1.3448\n",
      "Epoch 6 | Step: 117 | Loss: 1.0075\n",
      "Epoch 6 | Step: 118 | Loss: 1.0800\n",
      "Epoch 6 | Step: 119 | Loss: 1.1774\n",
      "Epoch 6 | Step: 120 | Loss: 0.9172\n",
      "Epoch 6 | Step: 121 | Loss: 1.3328\n",
      "Epoch 6 | Step: 122 | Loss: 0.9145\n",
      "Epoch 6 | Step: 123 | Loss: 1.0117\n",
      "Epoch 6 | Step: 124 | Loss: 0.9127\n",
      "Epoch 6 | Step: 125 | Loss: 1.0304\n",
      "Epoch 6 | Step: 126 | Loss: 0.8509\n",
      "Epoch 6 | Step: 127 | Loss: 0.9275\n",
      "Epoch 6 | Step: 128 | Loss: 1.6597\n",
      "Epoch 7 | Step: 1 | Loss: 0.9141\n",
      "Epoch 7 | Step: 2 | Loss: 0.6065\n",
      "Epoch 7 | Step: 3 | Loss: 1.7745\n",
      "Epoch 7 | Step: 4 | Loss: 1.6659\n",
      "Epoch 7 | Step: 5 | Loss: 1.2432\n",
      "Epoch 7 | Step: 6 | Loss: 0.8270\n",
      "Epoch 7 | Step: 7 | Loss: 0.9429\n",
      "Epoch 7 | Step: 8 | Loss: 1.1129\n",
      "Epoch 7 | Step: 9 | Loss: 0.6210\n",
      "Epoch 7 | Step: 10 | Loss: 1.1645\n",
      "Epoch 7 | Step: 11 | Loss: 1.0599\n",
      "Epoch 7 | Step: 12 | Loss: 0.9234\n",
      "Epoch 7 | Step: 13 | Loss: 1.1191\n",
      "Epoch 7 | Step: 14 | Loss: 1.0320\n",
      "Epoch 7 | Step: 15 | Loss: 0.8366\n",
      "Epoch 7 | Step: 16 | Loss: 1.1342\n",
      "Epoch 7 | Step: 17 | Loss: 1.0449\n",
      "Epoch 7 | Step: 18 | Loss: 0.9400\n",
      "Epoch 7 | Step: 19 | Loss: 0.8957\n",
      "Epoch 7 | Step: 20 | Loss: 0.8840\n",
      "Epoch 7 | Step: 21 | Loss: 1.0213\n",
      "Epoch 7 | Step: 22 | Loss: 0.8676\n",
      "Epoch 7 | Step: 23 | Loss: 0.8937\n",
      "Epoch 7 | Step: 24 | Loss: 0.7348\n",
      "Epoch 7 | Step: 25 | Loss: 0.9404\n",
      "Epoch 7 | Step: 26 | Loss: 0.6855\n",
      "Epoch 7 | Step: 27 | Loss: 0.8796\n",
      "Epoch 7 | Step: 28 | Loss: 0.8513\n",
      "Epoch 7 | Step: 29 | Loss: 0.7555\n",
      "Epoch 7 | Step: 30 | Loss: 0.6587\n",
      "Epoch 7 | Step: 31 | Loss: 0.8522\n",
      "Epoch 7 | Step: 32 | Loss: 1.0262\n",
      "Epoch 7 | Step: 33 | Loss: 1.1623\n",
      "Epoch 7 | Step: 34 | Loss: 1.1278\n",
      "Epoch 7 | Step: 35 | Loss: 0.9288\n",
      "Epoch 7 | Step: 36 | Loss: 0.8923\n",
      "Epoch 7 | Step: 37 | Loss: 1.3202\n",
      "Epoch 7 | Step: 38 | Loss: 0.8905\n",
      "Epoch 7 | Step: 39 | Loss: 1.1652\n",
      "Epoch 7 | Step: 40 | Loss: 0.6131\n",
      "Epoch 7 | Step: 41 | Loss: 1.0760\n",
      "Epoch 7 | Step: 42 | Loss: 1.2460\n",
      "Epoch 7 | Step: 43 | Loss: 0.9568\n",
      "Epoch 7 | Step: 44 | Loss: 0.8456\n",
      "Epoch 7 | Step: 45 | Loss: 0.9397\n",
      "Epoch 7 | Step: 46 | Loss: 0.8924\n",
      "Epoch 7 | Step: 47 | Loss: 1.3347\n",
      "Epoch 7 | Step: 48 | Loss: 0.8840\n",
      "Epoch 7 | Step: 49 | Loss: 0.7082\n",
      "Epoch 7 | Step: 50 | Loss: 0.8978\n",
      "Epoch 7 | Step: 51 | Loss: 1.0303\n",
      "Epoch 7 | Step: 52 | Loss: 1.1227\n",
      "Epoch 7 | Step: 53 | Loss: 1.2426\n",
      "Epoch 7 | Step: 54 | Loss: 1.1354\n",
      "Epoch 7 | Step: 55 | Loss: 1.0182\n",
      "Epoch 7 | Step: 56 | Loss: 0.9510\n",
      "Epoch 7 | Step: 57 | Loss: 0.6779\n",
      "Epoch 7 | Step: 58 | Loss: 0.9712\n",
      "Epoch 7 | Step: 59 | Loss: 0.5737\n",
      "Epoch 7 | Step: 60 | Loss: 1.1128\n",
      "Epoch 7 | Step: 61 | Loss: 0.5377\n",
      "Epoch 7 | Step: 62 | Loss: 1.0788\n",
      "Epoch 7 | Step: 63 | Loss: 0.9748\n",
      "Epoch 7 | Step: 64 | Loss: 1.5961\n",
      "Epoch 7 | Step: 65 | Loss: 0.9551\n",
      "Epoch 7 | Step: 66 | Loss: 0.7751\n",
      "Epoch 7 | Step: 67 | Loss: 0.8290\n",
      "Epoch 7 | Step: 68 | Loss: 0.5310\n",
      "Epoch 7 | Step: 69 | Loss: 0.5819\n",
      "Epoch 7 | Step: 70 | Loss: 0.9468\n",
      "Epoch 7 | Step: 71 | Loss: 0.8403\n",
      "Epoch 7 | Step: 72 | Loss: 1.0947\n",
      "Epoch 7 | Step: 73 | Loss: 0.8888\n",
      "Epoch 7 | Step: 74 | Loss: 1.0720\n",
      "Epoch 7 | Step: 75 | Loss: 0.8120\n",
      "Epoch 7 | Step: 76 | Loss: 0.9058\n",
      "Epoch 7 | Step: 77 | Loss: 1.0848\n",
      "Epoch 7 | Step: 78 | Loss: 1.4823\n",
      "Epoch 7 | Step: 79 | Loss: 0.9191\n",
      "Epoch 7 | Step: 80 | Loss: 0.7756\n",
      "Epoch 7 | Step: 81 | Loss: 0.7593\n",
      "Epoch 7 | Step: 82 | Loss: 0.8990\n",
      "Epoch 7 | Step: 83 | Loss: 0.9767\n",
      "Epoch 7 | Step: 84 | Loss: 0.7267\n",
      "Epoch 7 | Step: 85 | Loss: 0.8779\n",
      "Epoch 7 | Step: 86 | Loss: 0.8348\n",
      "Epoch 7 | Step: 87 | Loss: 0.9347\n",
      "Epoch 7 | Step: 88 | Loss: 1.1102\n",
      "Epoch 7 | Step: 89 | Loss: 1.3461\n",
      "Epoch 7 | Step: 90 | Loss: 0.8688\n",
      "Epoch 7 | Step: 91 | Loss: 1.0767\n",
      "Epoch 7 | Step: 92 | Loss: 1.2922\n",
      "Epoch 7 | Step: 93 | Loss: 1.2211\n",
      "Epoch 7 | Step: 94 | Loss: 0.7512\n",
      "Epoch 7 | Step: 95 | Loss: 0.8350\n",
      "Epoch 7 | Step: 96 | Loss: 0.7557\n",
      "Epoch 7 | Step: 97 | Loss: 1.1679\n",
      "Epoch 7 | Step: 98 | Loss: 0.6959\n",
      "Epoch 7 | Step: 99 | Loss: 1.0350\n",
      "Epoch 7 | Step: 100 | Loss: 1.0694\n",
      "Epoch 7 | Step: 101 | Loss: 0.9487\n",
      "Epoch 7 | Step: 102 | Loss: 1.0586\n",
      "Epoch 7 | Step: 103 | Loss: 0.5256\n",
      "Epoch 7 | Step: 104 | Loss: 1.4845\n",
      "Epoch 7 | Step: 105 | Loss: 0.8472\n",
      "Epoch 7 | Step: 106 | Loss: 1.0990\n",
      "Epoch 7 | Step: 107 | Loss: 1.1551\n",
      "Epoch 7 | Step: 108 | Loss: 1.0901\n",
      "Epoch 7 | Step: 109 | Loss: 0.9437\n",
      "Epoch 7 | Step: 110 | Loss: 0.8811\n",
      "Epoch 7 | Step: 111 | Loss: 1.0097\n",
      "Epoch 7 | Step: 112 | Loss: 0.8895\n",
      "Epoch 7 | Step: 113 | Loss: 2.4213\n",
      "Epoch 7 | Step: 114 | Loss: 4.4310\n",
      "Epoch 7 | Step: 115 | Loss: 0.9446\n",
      "Epoch 7 | Step: 116 | Loss: 0.9092\n",
      "Epoch 7 | Step: 117 | Loss: 0.7587\n",
      "Epoch 7 | Step: 118 | Loss: 0.9067\n",
      "Epoch 7 | Step: 119 | Loss: 0.7222\n",
      "Epoch 7 | Step: 120 | Loss: 0.6231\n",
      "Epoch 7 | Step: 121 | Loss: 0.8250\n",
      "Epoch 7 | Step: 122 | Loss: 0.9256\n",
      "Epoch 7 | Step: 123 | Loss: 1.1933\n",
      "Epoch 7 | Step: 124 | Loss: 1.0823\n",
      "Epoch 7 | Step: 125 | Loss: 1.0023\n",
      "Epoch 7 | Step: 126 | Loss: 0.8536\n",
      "Epoch 7 | Step: 127 | Loss: 3.3366\n",
      "Epoch 7 | Step: 128 | Loss: 1.1276\n",
      "Epoch 8 | Step: 1 | Loss: 0.8768\n",
      "Epoch 8 | Step: 2 | Loss: 1.0252\n",
      "Epoch 8 | Step: 3 | Loss: 0.6361\n",
      "Epoch 8 | Step: 4 | Loss: 0.8973\n",
      "Epoch 8 | Step: 5 | Loss: 0.6981\n",
      "Epoch 8 | Step: 6 | Loss: 0.4987\n",
      "Epoch 8 | Step: 7 | Loss: 0.4470\n",
      "Epoch 8 | Step: 8 | Loss: 1.2338\n",
      "Epoch 8 | Step: 9 | Loss: 0.7337\n",
      "Epoch 8 | Step: 10 | Loss: 1.0415\n",
      "Epoch 8 | Step: 11 | Loss: 1.2915\n",
      "Epoch 8 | Step: 12 | Loss: 0.9116\n",
      "Epoch 8 | Step: 13 | Loss: 0.7352\n",
      "Epoch 8 | Step: 14 | Loss: 0.7676\n",
      "Epoch 8 | Step: 15 | Loss: 0.7620\n",
      "Epoch 8 | Step: 16 | Loss: 0.7347\n",
      "Epoch 8 | Step: 17 | Loss: 0.9759\n",
      "Epoch 8 | Step: 18 | Loss: 0.8453\n",
      "Epoch 8 | Step: 19 | Loss: 0.5123\n",
      "Epoch 8 | Step: 20 | Loss: 0.9215\n",
      "Epoch 8 | Step: 21 | Loss: 1.0845\n",
      "Epoch 8 | Step: 22 | Loss: 1.1166\n",
      "Epoch 8 | Step: 23 | Loss: 0.9208\n",
      "Epoch 8 | Step: 24 | Loss: 0.6586\n",
      "Epoch 8 | Step: 25 | Loss: 0.8526\n",
      "Epoch 8 | Step: 26 | Loss: 0.8336\n",
      "Epoch 8 | Step: 27 | Loss: 0.6872\n",
      "Epoch 8 | Step: 28 | Loss: 1.3192\n",
      "Epoch 8 | Step: 29 | Loss: 1.1407\n",
      "Epoch 8 | Step: 30 | Loss: 1.3722\n",
      "Epoch 8 | Step: 31 | Loss: 0.6018\n",
      "Epoch 8 | Step: 32 | Loss: 0.9291\n",
      "Epoch 8 | Step: 33 | Loss: 0.7004\n",
      "Epoch 8 | Step: 34 | Loss: 0.5027\n",
      "Epoch 8 | Step: 35 | Loss: 1.4810\n",
      "Epoch 8 | Step: 36 | Loss: 0.6362\n",
      "Epoch 8 | Step: 37 | Loss: 0.9918\n",
      "Epoch 8 | Step: 38 | Loss: 1.1352\n",
      "Epoch 8 | Step: 39 | Loss: 0.6689\n",
      "Epoch 8 | Step: 40 | Loss: 0.5121\n",
      "Epoch 8 | Step: 41 | Loss: 0.8847\n",
      "Epoch 8 | Step: 42 | Loss: 0.7385\n",
      "Epoch 8 | Step: 43 | Loss: 1.0867\n",
      "Epoch 8 | Step: 44 | Loss: 1.1470\n",
      "Epoch 8 | Step: 45 | Loss: 0.6149\n",
      "Epoch 8 | Step: 46 | Loss: 0.6941\n",
      "Epoch 8 | Step: 47 | Loss: 0.8085\n",
      "Epoch 8 | Step: 48 | Loss: 0.8109\n",
      "Epoch 8 | Step: 49 | Loss: 0.9881\n",
      "Epoch 8 | Step: 50 | Loss: 0.8039\n",
      "Epoch 8 | Step: 51 | Loss: 0.7933\n",
      "Epoch 8 | Step: 52 | Loss: 0.6402\n",
      "Epoch 8 | Step: 53 | Loss: 0.6720\n",
      "Epoch 8 | Step: 54 | Loss: 0.9875\n",
      "Epoch 8 | Step: 55 | Loss: 0.9169\n",
      "Epoch 8 | Step: 56 | Loss: 1.2800\n",
      "Epoch 8 | Step: 57 | Loss: 0.9456\n",
      "Epoch 8 | Step: 58 | Loss: 0.6294\n",
      "Epoch 8 | Step: 59 | Loss: 0.9628\n",
      "Epoch 8 | Step: 60 | Loss: 0.9048\n",
      "Epoch 8 | Step: 61 | Loss: 0.6387\n",
      "Epoch 8 | Step: 62 | Loss: 0.7533\n",
      "Epoch 8 | Step: 63 | Loss: 0.7379\n",
      "Epoch 8 | Step: 64 | Loss: 0.9275\n",
      "Epoch 8 | Step: 65 | Loss: 0.8020\n",
      "Epoch 8 | Step: 66 | Loss: 0.4325\n",
      "Epoch 8 | Step: 67 | Loss: 0.9329\n",
      "Epoch 8 | Step: 68 | Loss: 3.2398\n",
      "Epoch 8 | Step: 69 | Loss: 0.8915\n",
      "Epoch 8 | Step: 70 | Loss: 0.7494\n",
      "Epoch 8 | Step: 71 | Loss: 0.8130\n",
      "Epoch 8 | Step: 72 | Loss: 0.8954\n",
      "Epoch 8 | Step: 73 | Loss: 0.6289\n",
      "Epoch 8 | Step: 74 | Loss: 0.8247\n",
      "Epoch 8 | Step: 75 | Loss: 0.7341\n",
      "Epoch 8 | Step: 76 | Loss: 0.7907\n",
      "Epoch 8 | Step: 77 | Loss: 1.0447\n",
      "Epoch 8 | Step: 78 | Loss: 0.7630\n",
      "Epoch 8 | Step: 79 | Loss: 1.0135\n",
      "Epoch 8 | Step: 80 | Loss: 0.8976\n",
      "Epoch 8 | Step: 81 | Loss: 0.7626\n",
      "Epoch 8 | Step: 82 | Loss: 0.8075\n",
      "Epoch 8 | Step: 83 | Loss: 0.7931\n",
      "Epoch 8 | Step: 84 | Loss: 1.1796\n",
      "Epoch 8 | Step: 85 | Loss: 0.5551\n",
      "Epoch 8 | Step: 86 | Loss: 1.0392\n",
      "Epoch 8 | Step: 87 | Loss: 0.9843\n",
      "Epoch 8 | Step: 88 | Loss: 0.6411\n",
      "Epoch 8 | Step: 89 | Loss: 0.7869\n",
      "Epoch 8 | Step: 90 | Loss: 1.0963\n",
      "Epoch 8 | Step: 91 | Loss: 0.5741\n",
      "Epoch 8 | Step: 92 | Loss: 0.6927\n",
      "Epoch 8 | Step: 93 | Loss: 0.9515\n",
      "Epoch 8 | Step: 94 | Loss: 0.8517\n",
      "Epoch 8 | Step: 95 | Loss: 0.9355\n",
      "Epoch 8 | Step: 96 | Loss: 0.7594\n",
      "Epoch 8 | Step: 97 | Loss: 0.6290\n",
      "Epoch 8 | Step: 98 | Loss: 0.7903\n",
      "Epoch 8 | Step: 99 | Loss: 0.6523\n",
      "Epoch 8 | Step: 100 | Loss: 0.9173\n",
      "Epoch 8 | Step: 101 | Loss: 0.9471\n",
      "Epoch 8 | Step: 102 | Loss: 3.3275\n",
      "Epoch 8 | Step: 103 | Loss: 0.6563\n",
      "Epoch 8 | Step: 104 | Loss: 0.6222\n",
      "Epoch 8 | Step: 105 | Loss: 0.9881\n",
      "Epoch 8 | Step: 106 | Loss: 0.9862\n",
      "Epoch 8 | Step: 107 | Loss: 0.9119\n",
      "Epoch 8 | Step: 108 | Loss: 0.8953\n",
      "Epoch 8 | Step: 109 | Loss: 0.4553\n",
      "Epoch 8 | Step: 110 | Loss: 1.1377\n",
      "Epoch 8 | Step: 111 | Loss: 0.7229\n",
      "Epoch 8 | Step: 112 | Loss: 0.7227\n",
      "Epoch 8 | Step: 113 | Loss: 0.8866\n",
      "Epoch 8 | Step: 114 | Loss: 1.8235\n",
      "Epoch 8 | Step: 115 | Loss: 0.9191\n",
      "Epoch 8 | Step: 116 | Loss: 0.9896\n",
      "Epoch 8 | Step: 117 | Loss: 0.4767\n",
      "Epoch 8 | Step: 118 | Loss: 0.7781\n",
      "Epoch 8 | Step: 119 | Loss: 2.0833\n",
      "Epoch 8 | Step: 120 | Loss: 0.5656\n",
      "Epoch 8 | Step: 121 | Loss: 0.7512\n",
      "Epoch 8 | Step: 122 | Loss: 0.6362\n",
      "Epoch 8 | Step: 123 | Loss: 0.7194\n",
      "Epoch 8 | Step: 124 | Loss: 0.6293\n",
      "Epoch 8 | Step: 125 | Loss: 0.7736\n",
      "Epoch 8 | Step: 126 | Loss: 0.5333\n",
      "Epoch 8 | Step: 127 | Loss: 1.0672\n",
      "Epoch 8 | Step: 128 | Loss: 0.7504\n",
      "Epoch 9 | Step: 1 | Loss: 0.7971\n",
      "Epoch 9 | Step: 2 | Loss: 0.6236\n",
      "Epoch 9 | Step: 3 | Loss: 0.7650\n",
      "Epoch 9 | Step: 4 | Loss: 0.7949\n",
      "Epoch 9 | Step: 5 | Loss: 0.5493\n",
      "Epoch 9 | Step: 6 | Loss: 2.0758\n",
      "Epoch 9 | Step: 7 | Loss: 0.6603\n",
      "Epoch 9 | Step: 8 | Loss: 0.7046\n",
      "Epoch 9 | Step: 9 | Loss: 1.1030\n",
      "Epoch 9 | Step: 10 | Loss: 0.6416\n",
      "Epoch 9 | Step: 11 | Loss: 0.5889\n",
      "Epoch 9 | Step: 12 | Loss: 0.8649\n",
      "Epoch 9 | Step: 13 | Loss: 0.9758\n",
      "Epoch 9 | Step: 14 | Loss: 0.8638\n",
      "Epoch 9 | Step: 15 | Loss: 1.0203\n",
      "Epoch 9 | Step: 16 | Loss: 2.6684\n",
      "Epoch 9 | Step: 17 | Loss: 0.7250\n",
      "Epoch 9 | Step: 18 | Loss: 0.9408\n",
      "Epoch 9 | Step: 19 | Loss: 0.7823\n",
      "Epoch 9 | Step: 20 | Loss: 0.9241\n",
      "Epoch 9 | Step: 21 | Loss: 0.7044\n",
      "Epoch 9 | Step: 22 | Loss: 0.6304\n",
      "Epoch 9 | Step: 23 | Loss: 0.7783\n",
      "Epoch 9 | Step: 24 | Loss: 0.6750\n",
      "Epoch 9 | Step: 25 | Loss: 1.6250\n",
      "Epoch 9 | Step: 26 | Loss: 0.8814\n",
      "Epoch 9 | Step: 27 | Loss: 0.7675\n",
      "Epoch 9 | Step: 28 | Loss: 0.9292\n",
      "Epoch 9 | Step: 29 | Loss: 0.8644\n",
      "Epoch 9 | Step: 30 | Loss: 0.6312\n",
      "Epoch 9 | Step: 31 | Loss: 0.4388\n",
      "Epoch 9 | Step: 32 | Loss: 0.9956\n",
      "Epoch 9 | Step: 33 | Loss: 1.1619\n",
      "Epoch 9 | Step: 34 | Loss: 1.0378\n",
      "Epoch 9 | Step: 35 | Loss: 0.8651\n",
      "Epoch 9 | Step: 36 | Loss: 0.5954\n",
      "Epoch 9 | Step: 37 | Loss: 0.7013\n",
      "Epoch 9 | Step: 38 | Loss: 1.1631\n",
      "Epoch 9 | Step: 39 | Loss: 0.7962\n",
      "Epoch 9 | Step: 40 | Loss: 0.7349\n",
      "Epoch 9 | Step: 41 | Loss: 0.8603\n",
      "Epoch 9 | Step: 42 | Loss: 0.7628\n",
      "Epoch 9 | Step: 43 | Loss: 0.8629\n",
      "Epoch 9 | Step: 44 | Loss: 0.6296\n",
      "Epoch 9 | Step: 45 | Loss: 0.7310\n",
      "Epoch 9 | Step: 46 | Loss: 0.6302\n",
      "Epoch 9 | Step: 47 | Loss: 1.2398\n",
      "Epoch 9 | Step: 48 | Loss: 0.5155\n",
      "Epoch 9 | Step: 49 | Loss: 0.5544\n",
      "Epoch 9 | Step: 50 | Loss: 0.4006\n",
      "Epoch 9 | Step: 51 | Loss: 0.5687\n",
      "Epoch 9 | Step: 52 | Loss: 0.9299\n",
      "Epoch 9 | Step: 53 | Loss: 0.6083\n",
      "Epoch 9 | Step: 54 | Loss: 0.5221\n",
      "Epoch 9 | Step: 55 | Loss: 0.6191\n",
      "Epoch 9 | Step: 56 | Loss: 0.7908\n",
      "Epoch 9 | Step: 57 | Loss: 0.5574\n",
      "Epoch 9 | Step: 58 | Loss: 0.3493\n",
      "Epoch 9 | Step: 59 | Loss: 0.8566\n",
      "Epoch 9 | Step: 60 | Loss: 0.7639\n",
      "Epoch 9 | Step: 61 | Loss: 0.5453\n",
      "Epoch 9 | Step: 62 | Loss: 0.6193\n",
      "Epoch 9 | Step: 63 | Loss: 0.6363\n",
      "Epoch 9 | Step: 64 | Loss: 0.6464\n",
      "Epoch 9 | Step: 65 | Loss: 0.5064\n",
      "Epoch 9 | Step: 66 | Loss: 0.8228\n",
      "Epoch 9 | Step: 67 | Loss: 0.7893\n",
      "Epoch 9 | Step: 68 | Loss: 0.9243\n",
      "Epoch 9 | Step: 69 | Loss: 0.8426\n",
      "Epoch 9 | Step: 70 | Loss: 0.6824\n",
      "Epoch 9 | Step: 71 | Loss: 0.6773\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 9 | Step: 72 | Loss: 0.9018\n",
      "Epoch 9 | Step: 73 | Loss: 0.5139\n",
      "Epoch 9 | Step: 74 | Loss: 0.6945\n",
      "Epoch 9 | Step: 75 | Loss: 0.6291\n",
      "Epoch 9 | Step: 76 | Loss: 0.8004\n",
      "Epoch 9 | Step: 77 | Loss: 0.6284\n",
      "Epoch 9 | Step: 78 | Loss: 0.3861\n",
      "Epoch 9 | Step: 79 | Loss: 0.7713\n",
      "Epoch 9 | Step: 80 | Loss: 0.7491\n",
      "Epoch 9 | Step: 81 | Loss: 0.8894\n",
      "Epoch 9 | Step: 82 | Loss: 2.9692\n",
      "Epoch 9 | Step: 83 | Loss: 0.7551\n",
      "Epoch 9 | Step: 84 | Loss: 0.6270\n",
      "Epoch 9 | Step: 85 | Loss: 1.0066\n",
      "Epoch 9 | Step: 86 | Loss: 0.7857\n",
      "Epoch 9 | Step: 87 | Loss: 0.6606\n",
      "Epoch 9 | Step: 88 | Loss: 0.6030\n",
      "Epoch 9 | Step: 89 | Loss: 0.8566\n",
      "Epoch 9 | Step: 90 | Loss: 0.8610\n",
      "Epoch 9 | Step: 91 | Loss: 0.6785\n",
      "Epoch 9 | Step: 92 | Loss: 0.6245\n",
      "Epoch 9 | Step: 93 | Loss: 0.3425\n",
      "Epoch 9 | Step: 94 | Loss: 0.7321\n",
      "Epoch 9 | Step: 95 | Loss: 0.9674\n",
      "Epoch 9 | Step: 96 | Loss: 1.2982\n",
      "Epoch 9 | Step: 97 | Loss: 0.5649\n",
      "Epoch 9 | Step: 98 | Loss: 0.9320\n",
      "Epoch 9 | Step: 99 | Loss: 0.6247\n",
      "Epoch 9 | Step: 100 | Loss: 0.7636\n",
      "Epoch 9 | Step: 101 | Loss: 0.3980\n",
      "Epoch 9 | Step: 102 | Loss: 0.5947\n",
      "Epoch 9 | Step: 103 | Loss: 0.9188\n",
      "Epoch 9 | Step: 104 | Loss: 0.6096\n",
      "Epoch 9 | Step: 105 | Loss: 0.7279\n",
      "Epoch 9 | Step: 106 | Loss: 0.5883\n",
      "Epoch 9 | Step: 107 | Loss: 0.4852\n",
      "Epoch 9 | Step: 108 | Loss: 0.6643\n",
      "Epoch 9 | Step: 109 | Loss: 0.4817\n",
      "Epoch 9 | Step: 110 | Loss: 0.8086\n",
      "Epoch 9 | Step: 111 | Loss: 0.6593\n",
      "Epoch 9 | Step: 112 | Loss: 0.8036\n",
      "Epoch 9 | Step: 113 | Loss: 0.7430\n",
      "Epoch 9 | Step: 114 | Loss: 0.4920\n",
      "Epoch 9 | Step: 115 | Loss: 0.9182\n",
      "Epoch 9 | Step: 116 | Loss: 0.7969\n",
      "Epoch 9 | Step: 117 | Loss: 0.6430\n",
      "Epoch 9 | Step: 118 | Loss: 0.8228\n",
      "Epoch 9 | Step: 119 | Loss: 0.7997\n",
      "Epoch 9 | Step: 120 | Loss: 0.7326\n",
      "Epoch 9 | Step: 121 | Loss: 0.7049\n",
      "Epoch 9 | Step: 122 | Loss: 0.4679\n",
      "Epoch 9 | Step: 123 | Loss: 0.7369\n",
      "Epoch 9 | Step: 124 | Loss: 0.9284\n",
      "Epoch 9 | Step: 125 | Loss: 0.5786\n",
      "Epoch 9 | Step: 126 | Loss: 0.6866\n",
      "Epoch 9 | Step: 127 | Loss: 0.7102\n",
      "Epoch 9 | Step: 128 | Loss: 0.7296\n",
      "Epoch 10 | Step: 1 | Loss: 0.6372\n",
      "Epoch 10 | Step: 2 | Loss: 0.6089\n",
      "Epoch 10 | Step: 3 | Loss: 0.9874\n",
      "Epoch 10 | Step: 4 | Loss: 0.8053\n",
      "Epoch 10 | Step: 5 | Loss: 0.5322\n",
      "Epoch 10 | Step: 6 | Loss: 0.9178\n",
      "Epoch 10 | Step: 7 | Loss: 0.3748\n",
      "Epoch 10 | Step: 8 | Loss: 0.5472\n",
      "Epoch 10 | Step: 9 | Loss: 0.4904\n",
      "Epoch 10 | Step: 10 | Loss: 0.7814\n",
      "Epoch 10 | Step: 11 | Loss: 0.6414\n",
      "Epoch 10 | Step: 12 | Loss: 0.7533\n",
      "Epoch 10 | Step: 13 | Loss: 0.5350\n",
      "Epoch 10 | Step: 14 | Loss: 0.7912\n",
      "Epoch 10 | Step: 15 | Loss: 0.6935\n",
      "Epoch 10 | Step: 16 | Loss: 0.5601\n",
      "Epoch 10 | Step: 17 | Loss: 0.6444\n",
      "Epoch 10 | Step: 18 | Loss: 0.5436\n",
      "Epoch 10 | Step: 19 | Loss: 0.6750\n",
      "Epoch 10 | Step: 20 | Loss: 0.8519\n",
      "Epoch 10 | Step: 21 | Loss: 0.5024\n",
      "Epoch 10 | Step: 22 | Loss: 0.8107\n",
      "Epoch 10 | Step: 23 | Loss: 0.4747\n",
      "Epoch 10 | Step: 24 | Loss: 0.5479\n",
      "Epoch 10 | Step: 25 | Loss: 0.7278\n",
      "Epoch 10 | Step: 26 | Loss: 0.7321\n",
      "Epoch 10 | Step: 27 | Loss: 0.6519\n",
      "Epoch 10 | Step: 28 | Loss: 0.7245\n",
      "Epoch 10 | Step: 29 | Loss: 0.4322\n",
      "Epoch 10 | Step: 30 | Loss: 0.4868\n",
      "Epoch 10 | Step: 31 | Loss: 0.5849\n",
      "Epoch 10 | Step: 32 | Loss: 0.6145\n",
      "Epoch 10 | Step: 33 | Loss: 0.5342\n",
      "Epoch 10 | Step: 34 | Loss: 0.8137\n",
      "Epoch 10 | Step: 35 | Loss: 0.6285\n",
      "Epoch 10 | Step: 36 | Loss: 2.6765\n",
      "Epoch 10 | Step: 37 | Loss: 0.4479\n",
      "Epoch 10 | Step: 38 | Loss: 1.0130\n",
      "Epoch 10 | Step: 39 | Loss: 1.0262\n",
      "Epoch 10 | Step: 40 | Loss: 0.6068\n",
      "Epoch 10 | Step: 41 | Loss: 0.6721\n",
      "Epoch 10 | Step: 42 | Loss: 0.7578\n",
      "Epoch 10 | Step: 43 | Loss: 0.7107\n",
      "Epoch 10 | Step: 44 | Loss: 1.3477\n",
      "Epoch 10 | Step: 45 | Loss: 0.6898\n",
      "Epoch 10 | Step: 46 | Loss: 1.0330\n",
      "Epoch 10 | Step: 47 | Loss: 0.5911\n",
      "Epoch 10 | Step: 48 | Loss: 0.6038\n",
      "Epoch 10 | Step: 49 | Loss: 0.3766\n",
      "Epoch 10 | Step: 50 | Loss: 0.6978\n",
      "Epoch 10 | Step: 51 | Loss: 0.8935\n",
      "Epoch 10 | Step: 52 | Loss: 0.7102\n",
      "Epoch 10 | Step: 53 | Loss: 0.7439\n",
      "Epoch 10 | Step: 54 | Loss: 0.6450\n",
      "Epoch 10 | Step: 55 | Loss: 0.7169\n",
      "Epoch 10 | Step: 56 | Loss: 0.5197\n",
      "Epoch 10 | Step: 57 | Loss: 0.5487\n",
      "Epoch 10 | Step: 58 | Loss: 0.5636\n",
      "Epoch 10 | Step: 59 | Loss: 0.5462\n",
      "Epoch 10 | Step: 60 | Loss: 0.7529\n",
      "Epoch 10 | Step: 61 | Loss: 0.6701\n",
      "Epoch 10 | Step: 62 | Loss: 0.7160\n",
      "Epoch 10 | Step: 63 | Loss: 0.3371\n",
      "Epoch 10 | Step: 64 | Loss: 0.7192\n",
      "Epoch 10 | Step: 65 | Loss: 0.5776\n",
      "Epoch 10 | Step: 66 | Loss: 0.5939\n",
      "Epoch 10 | Step: 67 | Loss: 0.5990\n",
      "Epoch 10 | Step: 68 | Loss: 0.5967\n",
      "Epoch 10 | Step: 69 | Loss: 0.7989\n",
      "Epoch 10 | Step: 70 | Loss: 0.4540\n",
      "Epoch 10 | Step: 71 | Loss: 0.6908\n",
      "Epoch 10 | Step: 72 | Loss: 0.8467\n",
      "Epoch 10 | Step: 73 | Loss: 0.7781\n",
      "Epoch 10 | Step: 74 | Loss: 0.8930\n",
      "Epoch 10 | Step: 75 | Loss: 1.5884\n",
      "Epoch 10 | Step: 76 | Loss: 1.1305\n",
      "Epoch 10 | Step: 77 | Loss: 0.7214\n",
      "Epoch 10 | Step: 78 | Loss: 0.5945\n",
      "Epoch 10 | Step: 79 | Loss: 0.7644\n",
      "Epoch 10 | Step: 80 | Loss: 0.7323\n",
      "Epoch 10 | Step: 81 | Loss: 0.5666\n",
      "Epoch 10 | Step: 82 | Loss: 0.5730\n",
      "Epoch 10 | Step: 83 | Loss: 0.8763\n",
      "Epoch 10 | Step: 84 | Loss: 0.6094\n",
      "Epoch 10 | Step: 85 | Loss: 0.5959\n",
      "Epoch 10 | Step: 86 | Loss: 0.7729\n",
      "Epoch 10 | Step: 87 | Loss: 0.8158\n",
      "Epoch 10 | Step: 88 | Loss: 0.4599\n",
      "Epoch 10 | Step: 89 | Loss: 0.4605\n",
      "Epoch 10 | Step: 90 | Loss: 0.5777\n",
      "Epoch 10 | Step: 91 | Loss: 0.5093\n",
      "Epoch 10 | Step: 92 | Loss: 0.8516\n",
      "Epoch 10 | Step: 93 | Loss: 0.3546\n",
      "Epoch 10 | Step: 94 | Loss: 0.5616\n",
      "Epoch 10 | Step: 95 | Loss: 0.3328\n",
      "Epoch 10 | Step: 96 | Loss: 0.6221\n",
      "Epoch 10 | Step: 97 | Loss: 0.4236\n",
      "Epoch 10 | Step: 98 | Loss: 1.0947\n",
      "Epoch 10 | Step: 99 | Loss: 0.6946\n",
      "Epoch 10 | Step: 100 | Loss: 0.5959\n",
      "Epoch 10 | Step: 101 | Loss: 0.5757\n",
      "Epoch 10 | Step: 102 | Loss: 0.3259\n",
      "Epoch 10 | Step: 103 | Loss: 0.8366\n",
      "Epoch 10 | Step: 104 | Loss: 0.8404\n",
      "Epoch 10 | Step: 105 | Loss: 0.7095\n",
      "Epoch 10 | Step: 106 | Loss: 0.6332\n",
      "Epoch 10 | Step: 107 | Loss: 0.7219\n",
      "Epoch 10 | Step: 108 | Loss: 0.7824\n",
      "Epoch 10 | Step: 109 | Loss: 0.7639\n",
      "Epoch 10 | Step: 110 | Loss: 0.4752\n",
      "Epoch 10 | Step: 111 | Loss: 0.6127\n",
      "Epoch 10 | Step: 112 | Loss: 0.6168\n",
      "Epoch 10 | Step: 113 | Loss: 0.7367\n",
      "Epoch 10 | Step: 114 | Loss: 0.4861\n",
      "Epoch 10 | Step: 115 | Loss: 0.6058\n",
      "Epoch 10 | Step: 116 | Loss: 0.4879\n",
      "Epoch 10 | Step: 117 | Loss: 1.0524\n",
      "Epoch 10 | Step: 118 | Loss: 0.6133\n",
      "Epoch 10 | Step: 119 | Loss: 0.6757\n",
      "Epoch 10 | Step: 120 | Loss: 0.5887\n",
      "Epoch 10 | Step: 121 | Loss: 0.8685\n",
      "Epoch 10 | Step: 122 | Loss: 0.4767\n",
      "Epoch 10 | Step: 123 | Loss: 0.7495\n",
      "Epoch 10 | Step: 124 | Loss: 2.3227\n",
      "Epoch 10 | Step: 125 | Loss: 0.7437\n",
      "Epoch 10 | Step: 126 | Loss: 0.4764\n",
      "Epoch 10 | Step: 127 | Loss: 0.5002\n",
      "Epoch 10 | Step: 128 | Loss: 0.6725\n",
      "Epoch 11 | Step: 1 | Loss: 0.6495\n",
      "Epoch 11 | Step: 2 | Loss: 0.6607\n",
      "Epoch 11 | Step: 3 | Loss: 0.3832\n",
      "Epoch 11 | Step: 4 | Loss: 0.5053\n",
      "Epoch 11 | Step: 5 | Loss: 0.6218\n",
      "Epoch 11 | Step: 6 | Loss: 0.6964\n",
      "Epoch 11 | Step: 7 | Loss: 0.6776\n",
      "Epoch 11 | Step: 8 | Loss: 0.5266\n",
      "Epoch 11 | Step: 9 | Loss: 0.4485\n",
      "Epoch 11 | Step: 10 | Loss: 0.3769\n",
      "Epoch 11 | Step: 11 | Loss: 0.7913\n",
      "Epoch 11 | Step: 12 | Loss: 0.6964\n",
      "Epoch 11 | Step: 13 | Loss: 2.1816\n",
      "Epoch 11 | Step: 14 | Loss: 0.8161\n",
      "Epoch 11 | Step: 15 | Loss: 0.5670\n",
      "Epoch 11 | Step: 16 | Loss: 0.6825\n",
      "Epoch 11 | Step: 17 | Loss: 0.5902\n",
      "Epoch 11 | Step: 18 | Loss: 0.6792\n",
      "Epoch 11 | Step: 19 | Loss: 0.2999\n",
      "Epoch 11 | Step: 20 | Loss: 0.3644\n",
      "Epoch 11 | Step: 21 | Loss: 0.7880\n",
      "Epoch 11 | Step: 22 | Loss: 0.4807\n",
      "Epoch 11 | Step: 23 | Loss: 0.7761\n",
      "Epoch 11 | Step: 24 | Loss: 0.6389\n",
      "Epoch 11 | Step: 25 | Loss: 0.7232\n",
      "Epoch 11 | Step: 26 | Loss: 0.5433\n",
      "Epoch 11 | Step: 27 | Loss: 0.5815\n",
      "Epoch 11 | Step: 28 | Loss: 0.5782\n",
      "Epoch 11 | Step: 29 | Loss: 0.5167\n",
      "Epoch 11 | Step: 30 | Loss: 0.4154\n",
      "Epoch 11 | Step: 31 | Loss: 0.5615\n",
      "Epoch 11 | Step: 32 | Loss: 0.5711\n",
      "Epoch 11 | Step: 33 | Loss: 0.5911\n",
      "Epoch 11 | Step: 34 | Loss: 0.6760\n",
      "Epoch 11 | Step: 35 | Loss: 0.8574\n",
      "Epoch 11 | Step: 36 | Loss: 0.9092\n",
      "Epoch 11 | Step: 37 | Loss: 0.4255\n",
      "Epoch 11 | Step: 38 | Loss: 0.4990\n",
      "Epoch 11 | Step: 39 | Loss: 0.7119\n",
      "Epoch 11 | Step: 40 | Loss: 0.5265\n",
      "Epoch 11 | Step: 41 | Loss: 0.8472\n",
      "Epoch 11 | Step: 42 | Loss: 0.6400\n",
      "Epoch 11 | Step: 43 | Loss: 0.5699\n",
      "Epoch 11 | Step: 44 | Loss: 0.4572\n",
      "Epoch 11 | Step: 45 | Loss: 0.8706\n",
      "Epoch 11 | Step: 46 | Loss: 0.6052\n",
      "Epoch 11 | Step: 47 | Loss: 0.6189\n",
      "Epoch 11 | Step: 48 | Loss: 0.6788\n",
      "Epoch 11 | Step: 49 | Loss: 0.6368\n",
      "Epoch 11 | Step: 50 | Loss: 0.5622\n",
      "Epoch 11 | Step: 51 | Loss: 0.5422\n",
      "Epoch 11 | Step: 52 | Loss: 0.4949\n",
      "Epoch 11 | Step: 53 | Loss: 0.5836\n",
      "Epoch 11 | Step: 54 | Loss: 0.6587\n",
      "Epoch 11 | Step: 55 | Loss: 0.4828\n",
      "Epoch 11 | Step: 56 | Loss: 0.6716\n",
      "Epoch 11 | Step: 57 | Loss: 0.4509\n",
      "Epoch 11 | Step: 58 | Loss: 0.5404\n",
      "Epoch 11 | Step: 59 | Loss: 0.5917\n",
      "Epoch 11 | Step: 60 | Loss: 0.5225\n",
      "Epoch 11 | Step: 61 | Loss: 0.4166\n",
      "Epoch 11 | Step: 62 | Loss: 0.7883\n",
      "Epoch 11 | Step: 63 | Loss: 0.6207\n",
      "Epoch 11 | Step: 64 | Loss: 0.8239\n",
      "Epoch 11 | Step: 65 | Loss: 0.4477\n",
      "Epoch 11 | Step: 66 | Loss: 0.5741\n",
      "Epoch 11 | Step: 67 | Loss: 0.4407\n",
      "Epoch 11 | Step: 68 | Loss: 0.6836\n",
      "Epoch 11 | Step: 69 | Loss: 0.8049\n",
      "Epoch 11 | Step: 70 | Loss: 0.7092\n",
      "Epoch 11 | Step: 71 | Loss: 0.5186\n",
      "Epoch 11 | Step: 72 | Loss: 0.6200\n",
      "Epoch 11 | Step: 73 | Loss: 0.8290\n",
      "Epoch 11 | Step: 74 | Loss: 0.4551\n",
      "Epoch 11 | Step: 75 | Loss: 0.8152\n",
      "Epoch 11 | Step: 76 | Loss: 0.6974\n",
      "Epoch 11 | Step: 77 | Loss: 0.6441\n",
      "Epoch 11 | Step: 78 | Loss: 0.4612\n",
      "Epoch 11 | Step: 79 | Loss: 0.4509\n",
      "Epoch 11 | Step: 80 | Loss: 0.8488\n",
      "Epoch 11 | Step: 81 | Loss: 0.5336\n",
      "Epoch 11 | Step: 82 | Loss: 0.5509\n",
      "Epoch 11 | Step: 83 | Loss: 0.5114\n",
      "Epoch 11 | Step: 84 | Loss: 0.7745\n",
      "Epoch 11 | Step: 85 | Loss: 0.5183\n",
      "Epoch 11 | Step: 86 | Loss: 0.6260\n",
      "Epoch 11 | Step: 87 | Loss: 0.4655\n",
      "Epoch 11 | Step: 88 | Loss: 0.5725\n",
      "Epoch 11 | Step: 89 | Loss: 0.5508\n",
      "Epoch 11 | Step: 90 | Loss: 0.6822\n",
      "Epoch 11 | Step: 91 | Loss: 0.5446\n",
      "Epoch 11 | Step: 92 | Loss: 2.1756\n",
      "Epoch 11 | Step: 93 | Loss: 0.6398\n",
      "Epoch 11 | Step: 94 | Loss: 0.7580\n",
      "Epoch 11 | Step: 95 | Loss: 0.6963\n",
      "Epoch 11 | Step: 96 | Loss: 0.3806\n",
      "Epoch 11 | Step: 97 | Loss: 0.4534\n",
      "Epoch 11 | Step: 98 | Loss: 0.4480\n",
      "Epoch 11 | Step: 99 | Loss: 0.4768\n",
      "Epoch 11 | Step: 100 | Loss: 0.6035\n",
      "Epoch 11 | Step: 101 | Loss: 0.5339\n",
      "Epoch 11 | Step: 102 | Loss: 0.3728\n",
      "Epoch 11 | Step: 103 | Loss: 0.7039\n",
      "Epoch 11 | Step: 104 | Loss: 0.6312\n",
      "Epoch 11 | Step: 105 | Loss: 0.7719\n",
      "Epoch 11 | Step: 106 | Loss: 0.5714\n",
      "Epoch 11 | Step: 107 | Loss: 0.7350\n",
      "Epoch 11 | Step: 108 | Loss: 0.6081\n",
      "Epoch 11 | Step: 109 | Loss: 0.6345\n",
      "Epoch 11 | Step: 110 | Loss: 0.4742\n",
      "Epoch 11 | Step: 111 | Loss: 0.3610\n",
      "Epoch 11 | Step: 112 | Loss: 0.6103\n",
      "Epoch 11 | Step: 113 | Loss: 0.5349\n",
      "Epoch 11 | Step: 114 | Loss: 2.1181\n",
      "Epoch 11 | Step: 115 | Loss: 0.4743\n",
      "Epoch 11 | Step: 116 | Loss: 0.4907\n",
      "Epoch 11 | Step: 117 | Loss: 0.5083\n",
      "Epoch 11 | Step: 118 | Loss: 0.6653\n",
      "Epoch 11 | Step: 119 | Loss: 0.5422\n",
      "Epoch 11 | Step: 120 | Loss: 0.7014\n",
      "Epoch 11 | Step: 121 | Loss: 0.4445\n",
      "Epoch 11 | Step: 122 | Loss: 0.5467\n",
      "Epoch 11 | Step: 123 | Loss: 0.4170\n",
      "Epoch 11 | Step: 124 | Loss: 0.5218\n",
      "Epoch 11 | Step: 125 | Loss: 0.7462\n",
      "Epoch 11 | Step: 126 | Loss: 0.6383\n",
      "Epoch 11 | Step: 127 | Loss: 0.6723\n",
      "Epoch 11 | Step: 128 | Loss: 0.6265\n",
      "Epoch 12 | Step: 1 | Loss: 0.7189\n",
      "Epoch 12 | Step: 2 | Loss: 0.5481\n",
      "Epoch 12 | Step: 3 | Loss: 0.8052\n",
      "Epoch 12 | Step: 4 | Loss: 0.5306\n",
      "Epoch 12 | Step: 5 | Loss: 0.4566\n",
      "Epoch 12 | Step: 6 | Loss: 0.5503\n",
      "Epoch 12 | Step: 7 | Loss: 0.5222\n",
      "Epoch 12 | Step: 8 | Loss: 1.1054\n",
      "Epoch 12 | Step: 9 | Loss: 0.5093\n",
      "Epoch 12 | Step: 10 | Loss: 0.4042\n",
      "Epoch 12 | Step: 11 | Loss: 0.5139\n",
      "Epoch 12 | Step: 12 | Loss: 0.6179\n",
      "Epoch 12 | Step: 13 | Loss: 0.3859\n",
      "Epoch 12 | Step: 14 | Loss: 0.4196\n",
      "Epoch 12 | Step: 15 | Loss: 0.5660\n",
      "Epoch 12 | Step: 16 | Loss: 0.6784\n",
      "Epoch 12 | Step: 17 | Loss: 0.7841\n",
      "Epoch 12 | Step: 18 | Loss: 0.5755\n",
      "Epoch 12 | Step: 19 | Loss: 0.4411\n",
      "Epoch 12 | Step: 20 | Loss: 0.2042\n",
      "Epoch 12 | Step: 21 | Loss: 0.4979\n",
      "Epoch 12 | Step: 22 | Loss: 0.6784\n",
      "Epoch 12 | Step: 23 | Loss: 0.8519\n",
      "Epoch 12 | Step: 24 | Loss: 0.6685\n",
      "Epoch 12 | Step: 25 | Loss: 0.7261\n",
      "Epoch 12 | Step: 26 | Loss: 0.5448\n",
      "Epoch 12 | Step: 27 | Loss: 0.7586\n",
      "Epoch 12 | Step: 28 | Loss: 0.5031\n",
      "Epoch 12 | Step: 29 | Loss: 0.6171\n",
      "Epoch 12 | Step: 30 | Loss: 0.5106\n",
      "Epoch 12 | Step: 31 | Loss: 0.6055\n",
      "Epoch 12 | Step: 32 | Loss: 0.4693\n",
      "Epoch 12 | Step: 33 | Loss: 0.4844\n",
      "Epoch 12 | Step: 34 | Loss: 0.6579\n",
      "Epoch 12 | Step: 35 | Loss: 0.2758\n",
      "Epoch 12 | Step: 36 | Loss: 0.7594\n",
      "Epoch 12 | Step: 37 | Loss: 0.6150\n",
      "Epoch 12 | Step: 38 | Loss: 0.5294\n",
      "Epoch 12 | Step: 39 | Loss: 0.3816\n",
      "Epoch 12 | Step: 40 | Loss: 0.4068\n",
      "Epoch 12 | Step: 41 | Loss: 0.4952\n",
      "Epoch 12 | Step: 42 | Loss: 0.5956\n",
      "Epoch 12 | Step: 43 | Loss: 1.2944\n",
      "Epoch 12 | Step: 44 | Loss: 0.5757\n",
      "Epoch 12 | Step: 45 | Loss: 0.4070\n",
      "Epoch 12 | Step: 46 | Loss: 0.6139\n",
      "Epoch 12 | Step: 47 | Loss: 1.0278\n",
      "Epoch 12 | Step: 48 | Loss: 0.4968\n",
      "Epoch 12 | Step: 49 | Loss: 0.3996\n",
      "Epoch 12 | Step: 50 | Loss: 0.6488\n",
      "Epoch 12 | Step: 51 | Loss: 0.5332\n",
      "Epoch 12 | Step: 52 | Loss: 0.4157\n",
      "Epoch 12 | Step: 53 | Loss: 0.3666\n",
      "Epoch 12 | Step: 54 | Loss: 0.3650\n",
      "Epoch 12 | Step: 55 | Loss: 0.5749\n",
      "Epoch 12 | Step: 56 | Loss: 0.4646\n",
      "Epoch 12 | Step: 57 | Loss: 0.4418\n",
      "Epoch 12 | Step: 58 | Loss: 0.5251\n",
      "Epoch 12 | Step: 59 | Loss: 0.4111\n",
      "Epoch 12 | Step: 60 | Loss: 0.8091\n",
      "Epoch 12 | Step: 61 | Loss: 0.4029\n",
      "Epoch 12 | Step: 62 | Loss: 0.5604\n",
      "Epoch 12 | Step: 63 | Loss: 0.6230\n",
      "Epoch 12 | Step: 64 | Loss: 0.4701\n",
      "Epoch 12 | Step: 65 | Loss: 0.6718\n",
      "Epoch 12 | Step: 66 | Loss: 0.7618\n",
      "Epoch 12 | Step: 67 | Loss: 0.6253\n",
      "Epoch 12 | Step: 68 | Loss: 0.5100\n",
      "Epoch 12 | Step: 69 | Loss: 0.3708\n",
      "Epoch 12 | Step: 70 | Loss: 0.5873\n",
      "Epoch 12 | Step: 71 | Loss: 0.5080\n",
      "Epoch 12 | Step: 72 | Loss: 1.8984\n",
      "Epoch 12 | Step: 73 | Loss: 0.4953\n",
      "Epoch 12 | Step: 74 | Loss: 2.0668\n",
      "Epoch 12 | Step: 75 | Loss: 0.5081\n",
      "Epoch 12 | Step: 76 | Loss: 0.4869\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 12 | Step: 77 | Loss: 0.3458\n",
      "Epoch 12 | Step: 78 | Loss: 0.4804\n",
      "Epoch 12 | Step: 79 | Loss: 0.4586\n",
      "Epoch 12 | Step: 80 | Loss: 0.3274\n",
      "Epoch 12 | Step: 81 | Loss: 0.7879\n",
      "Epoch 12 | Step: 82 | Loss: 0.6206\n",
      "Epoch 12 | Step: 83 | Loss: 0.6166\n",
      "Epoch 12 | Step: 84 | Loss: 0.7940\n",
      "Epoch 12 | Step: 85 | Loss: 0.4207\n",
      "Epoch 12 | Step: 86 | Loss: 0.4954\n",
      "Epoch 12 | Step: 87 | Loss: 0.5776\n",
      "Epoch 12 | Step: 88 | Loss: 0.5918\n",
      "Epoch 12 | Step: 89 | Loss: 0.5202\n",
      "Epoch 12 | Step: 90 | Loss: 0.6109\n",
      "Epoch 12 | Step: 91 | Loss: 0.5363\n",
      "Epoch 12 | Step: 92 | Loss: 0.5529\n",
      "Epoch 12 | Step: 93 | Loss: 0.3735\n",
      "Epoch 12 | Step: 94 | Loss: 0.4226\n",
      "Epoch 12 | Step: 95 | Loss: 0.6790\n",
      "Epoch 12 | Step: 96 | Loss: 0.5075\n",
      "Epoch 12 | Step: 97 | Loss: 0.4877\n",
      "Epoch 12 | Step: 98 | Loss: 0.5220\n",
      "Epoch 12 | Step: 99 | Loss: 0.6894\n",
      "Epoch 12 | Step: 100 | Loss: 0.4619\n",
      "Epoch 12 | Step: 101 | Loss: 0.4381\n",
      "Epoch 12 | Step: 102 | Loss: 0.4561\n",
      "Epoch 12 | Step: 103 | Loss: 0.6081\n",
      "Epoch 12 | Step: 104 | Loss: 0.4152\n",
      "Epoch 12 | Step: 105 | Loss: 0.3783\n",
      "Epoch 12 | Step: 106 | Loss: 0.7943\n",
      "Epoch 12 | Step: 107 | Loss: 0.7140\n",
      "Epoch 12 | Step: 108 | Loss: 0.5294\n",
      "Epoch 12 | Step: 109 | Loss: 0.3436\n",
      "Epoch 12 | Step: 110 | Loss: 0.6810\n",
      "Epoch 12 | Step: 111 | Loss: 0.6202\n",
      "Epoch 12 | Step: 112 | Loss: 0.6184\n",
      "Epoch 12 | Step: 113 | Loss: 0.4651\n",
      "Epoch 12 | Step: 114 | Loss: 0.5433\n",
      "Epoch 12 | Step: 115 | Loss: 0.5569\n",
      "Epoch 12 | Step: 116 | Loss: 0.8606\n",
      "Epoch 12 | Step: 117 | Loss: 0.4853\n",
      "Epoch 12 | Step: 118 | Loss: 0.2032\n",
      "Epoch 12 | Step: 119 | Loss: 0.5341\n",
      "Epoch 12 | Step: 120 | Loss: 0.5813\n",
      "Epoch 12 | Step: 121 | Loss: 0.5397\n",
      "Epoch 12 | Step: 122 | Loss: 0.6923\n",
      "Epoch 12 | Step: 123 | Loss: 0.4564\n",
      "Epoch 12 | Step: 124 | Loss: 0.5008\n",
      "Epoch 12 | Step: 125 | Loss: 0.6316\n",
      "Epoch 12 | Step: 126 | Loss: 0.2716\n",
      "Epoch 12 | Step: 127 | Loss: 0.4899\n",
      "Epoch 12 | Step: 128 | Loss: 0.7264\n",
      "Epoch 13 | Step: 1 | Loss: 0.4462\n",
      "Epoch 13 | Step: 2 | Loss: 0.5904\n",
      "Epoch 13 | Step: 3 | Loss: 0.3675\n",
      "Epoch 13 | Step: 4 | Loss: 1.4776\n",
      "Epoch 13 | Step: 5 | Loss: 0.4287\n",
      "Epoch 13 | Step: 6 | Loss: 0.6506\n",
      "Epoch 13 | Step: 7 | Loss: 0.4590\n",
      "Epoch 13 | Step: 8 | Loss: 0.9359\n",
      "Epoch 13 | Step: 9 | Loss: 0.5293\n",
      "Epoch 13 | Step: 10 | Loss: 0.6513\n",
      "Epoch 13 | Step: 11 | Loss: 0.5280\n",
      "Epoch 13 | Step: 12 | Loss: 0.3081\n",
      "Epoch 13 | Step: 13 | Loss: 0.6421\n",
      "Epoch 13 | Step: 14 | Loss: 0.5179\n",
      "Epoch 13 | Step: 15 | Loss: 0.2272\n",
      "Epoch 13 | Step: 16 | Loss: 0.4845\n",
      "Epoch 13 | Step: 17 | Loss: 0.5320\n",
      "Epoch 13 | Step: 18 | Loss: 0.4817\n",
      "Epoch 13 | Step: 19 | Loss: 0.4904\n",
      "Epoch 13 | Step: 20 | Loss: 0.6442\n",
      "Epoch 13 | Step: 21 | Loss: 0.9812\n",
      "Epoch 13 | Step: 22 | Loss: 0.5060\n",
      "Epoch 13 | Step: 23 | Loss: 0.6319\n",
      "Epoch 13 | Step: 24 | Loss: 0.3472\n",
      "Epoch 13 | Step: 25 | Loss: 0.6254\n",
      "Epoch 13 | Step: 26 | Loss: 0.3683\n",
      "Epoch 13 | Step: 27 | Loss: 0.7445\n",
      "Epoch 13 | Step: 28 | Loss: 0.5029\n",
      "Epoch 13 | Step: 29 | Loss: 0.4345\n",
      "Epoch 13 | Step: 30 | Loss: 0.4329\n",
      "Epoch 13 | Step: 31 | Loss: 0.4926\n",
      "Epoch 13 | Step: 32 | Loss: 0.6200\n",
      "Epoch 13 | Step: 33 | Loss: 0.6573\n",
      "Epoch 13 | Step: 34 | Loss: 1.9290\n",
      "Epoch 13 | Step: 35 | Loss: 0.8349\n",
      "Epoch 13 | Step: 36 | Loss: 0.7204\n",
      "Epoch 13 | Step: 37 | Loss: 0.4405\n",
      "Epoch 13 | Step: 38 | Loss: 0.3606\n",
      "Epoch 13 | Step: 39 | Loss: 0.5600\n",
      "Epoch 13 | Step: 40 | Loss: 0.3645\n",
      "Epoch 13 | Step: 41 | Loss: 0.5968\n",
      "Epoch 13 | Step: 42 | Loss: 0.4717\n",
      "Epoch 13 | Step: 43 | Loss: 0.4079\n",
      "Epoch 13 | Step: 44 | Loss: 0.4788\n",
      "Epoch 13 | Step: 45 | Loss: 0.5253\n",
      "Epoch 13 | Step: 46 | Loss: 0.3386\n",
      "Epoch 13 | Step: 47 | Loss: 0.4687\n",
      "Epoch 13 | Step: 48 | Loss: 0.4889\n",
      "Epoch 13 | Step: 49 | Loss: 0.5152\n",
      "Epoch 13 | Step: 50 | Loss: 0.7854\n",
      "Epoch 13 | Step: 51 | Loss: 0.3334\n",
      "Epoch 13 | Step: 52 | Loss: 0.5523\n",
      "Epoch 13 | Step: 53 | Loss: 0.5137\n",
      "Epoch 13 | Step: 54 | Loss: 0.6111\n",
      "Epoch 13 | Step: 55 | Loss: 0.4268\n",
      "Epoch 13 | Step: 56 | Loss: 0.4120\n",
      "Epoch 13 | Step: 57 | Loss: 0.4111\n",
      "Epoch 13 | Step: 58 | Loss: 0.4320\n",
      "Epoch 13 | Step: 59 | Loss: 0.6032\n",
      "Epoch 13 | Step: 60 | Loss: 0.3609\n",
      "Epoch 13 | Step: 61 | Loss: 0.6822\n",
      "Epoch 13 | Step: 62 | Loss: 0.5224\n",
      "Epoch 13 | Step: 63 | Loss: 0.5141\n",
      "Epoch 13 | Step: 64 | Loss: 0.5210\n",
      "Epoch 13 | Step: 65 | Loss: 0.5296\n",
      "Epoch 13 | Step: 66 | Loss: 0.4373\n",
      "Epoch 13 | Step: 67 | Loss: 0.5402\n",
      "Epoch 13 | Step: 68 | Loss: 0.6482\n",
      "Epoch 13 | Step: 69 | Loss: 0.2279\n",
      "Epoch 13 | Step: 70 | Loss: 0.5910\n",
      "Epoch 13 | Step: 71 | Loss: 1.0867\n",
      "Epoch 13 | Step: 72 | Loss: 0.5748\n",
      "Epoch 13 | Step: 73 | Loss: 0.5314\n",
      "Epoch 13 | Step: 74 | Loss: 0.5788\n",
      "Epoch 13 | Step: 75 | Loss: 0.4411\n",
      "Epoch 13 | Step: 76 | Loss: 0.2292\n",
      "Epoch 13 | Step: 77 | Loss: 0.7589\n",
      "Epoch 13 | Step: 78 | Loss: 0.3946\n",
      "Epoch 13 | Step: 79 | Loss: 0.3997\n",
      "Epoch 13 | Step: 80 | Loss: 0.4874\n",
      "Epoch 13 | Step: 81 | Loss: 0.4449\n",
      "Epoch 13 | Step: 82 | Loss: 0.5962\n",
      "Epoch 13 | Step: 83 | Loss: 0.7931\n",
      "Epoch 13 | Step: 84 | Loss: 0.7820\n",
      "Epoch 13 | Step: 85 | Loss: 0.3621\n",
      "Epoch 13 | Step: 86 | Loss: 0.5778\n",
      "Epoch 13 | Step: 87 | Loss: 0.3669\n",
      "Epoch 13 | Step: 88 | Loss: 0.4850\n",
      "Epoch 13 | Step: 89 | Loss: 0.6868\n",
      "Epoch 13 | Step: 90 | Loss: 0.4737\n",
      "Epoch 13 | Step: 91 | Loss: 0.6307\n",
      "Epoch 13 | Step: 92 | Loss: 0.4591\n",
      "Epoch 13 | Step: 93 | Loss: 0.3585\n",
      "Epoch 13 | Step: 94 | Loss: 0.4628\n",
      "Epoch 13 | Step: 95 | Loss: 0.6001\n",
      "Epoch 13 | Step: 96 | Loss: 0.3879\n",
      "Epoch 13 | Step: 97 | Loss: 0.4967\n",
      "Epoch 13 | Step: 98 | Loss: 0.5731\n",
      "Epoch 13 | Step: 99 | Loss: 0.5068\n",
      "Epoch 13 | Step: 100 | Loss: 0.4759\n",
      "Epoch 13 | Step: 101 | Loss: 0.5518\n",
      "Epoch 13 | Step: 102 | Loss: 0.5258\n",
      "Epoch 13 | Step: 103 | Loss: 0.4487\n",
      "Epoch 13 | Step: 104 | Loss: 0.5498\n",
      "Epoch 13 | Step: 105 | Loss: 0.2674\n",
      "Epoch 13 | Step: 106 | Loss: 0.4298\n",
      "Epoch 13 | Step: 107 | Loss: 0.4844\n",
      "Epoch 13 | Step: 108 | Loss: 0.2173\n",
      "Epoch 13 | Step: 109 | Loss: 0.4855\n",
      "Epoch 13 | Step: 110 | Loss: 0.7947\n",
      "Epoch 13 | Step: 111 | Loss: 0.6968\n",
      "Epoch 13 | Step: 112 | Loss: 0.5130\n",
      "Epoch 13 | Step: 113 | Loss: 0.6081\n",
      "Epoch 13 | Step: 114 | Loss: 0.3937\n",
      "Epoch 13 | Step: 115 | Loss: 0.4560\n",
      "Epoch 13 | Step: 116 | Loss: 0.3714\n",
      "Epoch 13 | Step: 117 | Loss: 0.5063\n",
      "Epoch 13 | Step: 118 | Loss: 0.5307\n",
      "Epoch 13 | Step: 119 | Loss: 0.4973\n",
      "Epoch 13 | Step: 120 | Loss: 0.4026\n",
      "Epoch 13 | Step: 121 | Loss: 0.4138\n",
      "Epoch 13 | Step: 122 | Loss: 0.5623\n",
      "Epoch 13 | Step: 123 | Loss: 0.3203\n",
      "Epoch 13 | Step: 124 | Loss: 0.3650\n",
      "Epoch 13 | Step: 125 | Loss: 0.4307\n",
      "Epoch 13 | Step: 126 | Loss: 0.4765\n",
      "Epoch 13 | Step: 127 | Loss: 0.3190\n",
      "Epoch 13 | Step: 128 | Loss: 0.3949\n",
      "Epoch 14 | Step: 1 | Loss: 0.5781\n",
      "Epoch 14 | Step: 2 | Loss: 0.2628\n",
      "Epoch 14 | Step: 3 | Loss: 0.4082\n",
      "Epoch 14 | Step: 4 | Loss: 0.6881\n",
      "Epoch 14 | Step: 5 | Loss: 0.3428\n",
      "Epoch 14 | Step: 6 | Loss: 0.3957\n",
      "Epoch 14 | Step: 7 | Loss: 0.6146\n",
      "Epoch 14 | Step: 8 | Loss: 0.4593\n",
      "Epoch 14 | Step: 9 | Loss: 0.5942\n",
      "Epoch 14 | Step: 10 | Loss: 0.4667\n",
      "Epoch 14 | Step: 11 | Loss: 0.4555\n",
      "Epoch 14 | Step: 12 | Loss: 0.5662\n",
      "Epoch 14 | Step: 13 | Loss: 0.3652\n",
      "Epoch 14 | Step: 14 | Loss: 0.2463\n",
      "Epoch 14 | Step: 15 | Loss: 0.4793\n",
      "Epoch 14 | Step: 16 | Loss: 0.4224\n",
      "Epoch 14 | Step: 17 | Loss: 0.6890\n",
      "Epoch 14 | Step: 18 | Loss: 0.3832\n",
      "Epoch 14 | Step: 19 | Loss: 0.4375\n",
      "Epoch 14 | Step: 20 | Loss: 0.4187\n",
      "Epoch 14 | Step: 21 | Loss: 0.3334\n",
      "Epoch 14 | Step: 22 | Loss: 0.4231\n",
      "Epoch 14 | Step: 23 | Loss: 0.5369\n",
      "Epoch 14 | Step: 24 | Loss: 0.5467\n",
      "Epoch 14 | Step: 25 | Loss: 0.4303\n",
      "Epoch 14 | Step: 26 | Loss: 0.4793\n",
      "Epoch 14 | Step: 27 | Loss: 0.5184\n",
      "Epoch 14 | Step: 28 | Loss: 0.5342\n",
      "Epoch 14 | Step: 29 | Loss: 0.4848\n",
      "Epoch 14 | Step: 30 | Loss: 0.4918\n",
      "Epoch 14 | Step: 31 | Loss: 0.3920\n",
      "Epoch 14 | Step: 32 | Loss: 0.4790\n",
      "Epoch 14 | Step: 33 | Loss: 0.5808\n",
      "Epoch 14 | Step: 34 | Loss: 0.5436\n",
      "Epoch 14 | Step: 35 | Loss: 0.5287\n",
      "Epoch 14 | Step: 36 | Loss: 0.3558\n",
      "Epoch 14 | Step: 37 | Loss: 0.3645\n",
      "Epoch 14 | Step: 38 | Loss: 0.4446\n",
      "Epoch 14 | Step: 39 | Loss: 0.4536\n",
      "Epoch 14 | Step: 40 | Loss: 0.3864\n",
      "Epoch 14 | Step: 41 | Loss: 0.4469\n",
      "Epoch 14 | Step: 42 | Loss: 0.3204\n",
      "Epoch 14 | Step: 43 | Loss: 0.6367\n",
      "Epoch 14 | Step: 44 | Loss: 0.4633\n",
      "Epoch 14 | Step: 45 | Loss: 0.6296\n",
      "Epoch 14 | Step: 46 | Loss: 0.6796\n",
      "Epoch 14 | Step: 47 | Loss: 0.4257\n",
      "Epoch 14 | Step: 48 | Loss: 0.4013\n",
      "Epoch 14 | Step: 49 | Loss: 0.3578\n",
      "Epoch 14 | Step: 50 | Loss: 0.2882\n",
      "Epoch 14 | Step: 51 | Loss: 0.5665\n",
      "Epoch 14 | Step: 52 | Loss: 0.7325\n",
      "Epoch 14 | Step: 53 | Loss: 0.3590\n",
      "Epoch 14 | Step: 54 | Loss: 1.5910\n",
      "Epoch 14 | Step: 55 | Loss: 0.5618\n",
      "Epoch 14 | Step: 56 | Loss: 0.3971\n",
      "Epoch 14 | Step: 57 | Loss: 0.4653\n",
      "Epoch 14 | Step: 58 | Loss: 0.4801\n",
      "Epoch 14 | Step: 59 | Loss: 0.3825\n",
      "Epoch 14 | Step: 60 | Loss: 0.6958\n",
      "Epoch 14 | Step: 61 | Loss: 0.1752\n",
      "Epoch 14 | Step: 62 | Loss: 0.3979\n",
      "Epoch 14 | Step: 63 | Loss: 0.5510\n",
      "Epoch 14 | Step: 64 | Loss: 0.4187\n",
      "Epoch 14 | Step: 65 | Loss: 0.8098\n",
      "Epoch 14 | Step: 66 | Loss: 0.7491\n",
      "Epoch 14 | Step: 67 | Loss: 0.3852\n",
      "Epoch 14 | Step: 68 | Loss: 0.2877\n",
      "Epoch 14 | Step: 69 | Loss: 0.3641\n",
      "Epoch 14 | Step: 70 | Loss: 0.3493\n",
      "Epoch 14 | Step: 71 | Loss: 0.5647\n",
      "Epoch 14 | Step: 72 | Loss: 0.4150\n",
      "Epoch 14 | Step: 73 | Loss: 0.4356\n",
      "Epoch 14 | Step: 74 | Loss: 0.4844\n",
      "Epoch 14 | Step: 75 | Loss: 0.3759\n",
      "Epoch 14 | Step: 76 | Loss: 0.5315\n",
      "Epoch 14 | Step: 77 | Loss: 0.3177\n",
      "Epoch 14 | Step: 78 | Loss: 0.4390\n",
      "Epoch 14 | Step: 79 | Loss: 0.4749\n",
      "Epoch 14 | Step: 80 | Loss: 0.4811\n",
      "Epoch 14 | Step: 81 | Loss: 0.5158\n",
      "Epoch 14 | Step: 82 | Loss: 0.4593\n",
      "Epoch 14 | Step: 83 | Loss: 0.4474\n",
      "Epoch 14 | Step: 84 | Loss: 0.3211\n",
      "Epoch 14 | Step: 85 | Loss: 0.5301\n",
      "Epoch 14 | Step: 86 | Loss: 0.3239\n",
      "Epoch 14 | Step: 87 | Loss: 0.5988\n",
      "Epoch 14 | Step: 88 | Loss: 0.4023\n",
      "Epoch 14 | Step: 89 | Loss: 0.2953\n",
      "Epoch 14 | Step: 90 | Loss: 0.5919\n",
      "Epoch 14 | Step: 91 | Loss: 0.5631\n",
      "Epoch 14 | Step: 92 | Loss: 0.5108\n",
      "Epoch 14 | Step: 93 | Loss: 0.4669\n",
      "Epoch 14 | Step: 94 | Loss: 0.7659\n",
      "Epoch 14 | Step: 95 | Loss: 0.4849\n",
      "Epoch 14 | Step: 96 | Loss: 0.5962\n",
      "Epoch 14 | Step: 97 | Loss: 0.7395\n",
      "Epoch 14 | Step: 98 | Loss: 0.2708\n",
      "Epoch 14 | Step: 99 | Loss: 0.2622\n",
      "Epoch 14 | Step: 100 | Loss: 0.3011\n",
      "Epoch 14 | Step: 101 | Loss: 0.5874\n",
      "Epoch 14 | Step: 102 | Loss: 0.6715\n",
      "Epoch 14 | Step: 103 | Loss: 0.2716\n",
      "Epoch 14 | Step: 104 | Loss: 0.3164\n",
      "Epoch 14 | Step: 105 | Loss: 0.4917\n",
      "Epoch 14 | Step: 106 | Loss: 0.6011\n",
      "Epoch 14 | Step: 107 | Loss: 0.6990\n",
      "Epoch 14 | Step: 108 | Loss: 0.4560\n",
      "Epoch 14 | Step: 109 | Loss: 0.3434\n",
      "Epoch 14 | Step: 110 | Loss: 0.4624\n",
      "Epoch 14 | Step: 111 | Loss: 0.5187\n",
      "Epoch 14 | Step: 112 | Loss: 1.7174\n",
      "Epoch 14 | Step: 113 | Loss: 0.4729\n",
      "Epoch 14 | Step: 114 | Loss: 0.6070\n",
      "Epoch 14 | Step: 115 | Loss: 0.3907\n",
      "Epoch 14 | Step: 116 | Loss: 0.2918\n",
      "Epoch 14 | Step: 117 | Loss: 0.4181\n",
      "Epoch 14 | Step: 118 | Loss: 0.6676\n",
      "Epoch 14 | Step: 119 | Loss: 0.9189\n",
      "Epoch 14 | Step: 120 | Loss: 0.3948\n",
      "Epoch 14 | Step: 121 | Loss: 0.4956\n",
      "Epoch 14 | Step: 122 | Loss: 0.5259\n",
      "Epoch 14 | Step: 123 | Loss: 0.7305\n",
      "Epoch 14 | Step: 124 | Loss: 0.5284\n",
      "Epoch 14 | Step: 125 | Loss: 0.4130\n",
      "Epoch 14 | Step: 126 | Loss: 0.4530\n",
      "Epoch 14 | Step: 127 | Loss: 0.5539\n",
      "Epoch 14 | Step: 128 | Loss: 0.2624\n",
      "Epoch 15 | Step: 1 | Loss: 0.4159\n",
      "Epoch 15 | Step: 2 | Loss: 0.6448\n",
      "Epoch 15 | Step: 3 | Loss: 0.3341\n",
      "Epoch 15 | Step: 4 | Loss: 0.5395\n",
      "Epoch 15 | Step: 5 | Loss: 0.5692\n",
      "Epoch 15 | Step: 6 | Loss: 0.1457\n",
      "Epoch 15 | Step: 7 | Loss: 0.4252\n",
      "Epoch 15 | Step: 8 | Loss: 0.4907\n",
      "Epoch 15 | Step: 9 | Loss: 0.4103\n",
      "Epoch 15 | Step: 10 | Loss: 0.5807\n",
      "Epoch 15 | Step: 11 | Loss: 0.2377\n",
      "Epoch 15 | Step: 12 | Loss: 0.3485\n",
      "Epoch 15 | Step: 13 | Loss: 0.4780\n",
      "Epoch 15 | Step: 14 | Loss: 0.4494\n",
      "Epoch 15 | Step: 15 | Loss: 0.4270\n",
      "Epoch 15 | Step: 16 | Loss: 0.4023\n",
      "Epoch 15 | Step: 17 | Loss: 0.5737\n",
      "Epoch 15 | Step: 18 | Loss: 0.3636\n",
      "Epoch 15 | Step: 19 | Loss: 0.5852\n",
      "Epoch 15 | Step: 20 | Loss: 0.7765\n",
      "Epoch 15 | Step: 21 | Loss: 0.5170\n",
      "Epoch 15 | Step: 22 | Loss: 0.3886\n",
      "Epoch 15 | Step: 23 | Loss: 0.2622\n",
      "Epoch 15 | Step: 24 | Loss: 0.3410\n",
      "Epoch 15 | Step: 25 | Loss: 0.5079\n",
      "Epoch 15 | Step: 26 | Loss: 0.4579\n",
      "Epoch 15 | Step: 27 | Loss: 0.3617\n",
      "Epoch 15 | Step: 28 | Loss: 0.4347\n",
      "Epoch 15 | Step: 29 | Loss: 0.4001\n",
      "Epoch 15 | Step: 30 | Loss: 0.4452\n",
      "Epoch 15 | Step: 31 | Loss: 0.4510\n",
      "Epoch 15 | Step: 32 | Loss: 0.2387\n",
      "Epoch 15 | Step: 33 | Loss: 0.5575\n",
      "Epoch 15 | Step: 34 | Loss: 0.3030\n",
      "Epoch 15 | Step: 35 | Loss: 0.9244\n",
      "Epoch 15 | Step: 36 | Loss: 0.4947\n",
      "Epoch 15 | Step: 37 | Loss: 0.4284\n",
      "Epoch 15 | Step: 38 | Loss: 0.6953\n",
      "Epoch 15 | Step: 39 | Loss: 0.5641\n",
      "Epoch 15 | Step: 40 | Loss: 0.3300\n",
      "Epoch 15 | Step: 41 | Loss: 0.4950\n",
      "Epoch 15 | Step: 42 | Loss: 0.3008\n",
      "Epoch 15 | Step: 43 | Loss: 0.4920\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 15 | Step: 44 | Loss: 0.3060\n",
      "Epoch 15 | Step: 45 | Loss: 0.2543\n",
      "Epoch 15 | Step: 46 | Loss: 1.3441\n",
      "Epoch 15 | Step: 47 | Loss: 0.4164\n",
      "Epoch 15 | Step: 48 | Loss: 0.4319\n",
      "Epoch 15 | Step: 49 | Loss: 0.7102\n",
      "Epoch 15 | Step: 50 | Loss: 0.5415\n",
      "Epoch 15 | Step: 51 | Loss: 0.3426\n",
      "Epoch 15 | Step: 52 | Loss: 0.3037\n",
      "Epoch 15 | Step: 53 | Loss: 0.2325\n",
      "Epoch 15 | Step: 54 | Loss: 0.5181\n",
      "Epoch 15 | Step: 55 | Loss: 0.4200\n",
      "Epoch 15 | Step: 56 | Loss: 0.3928\n",
      "Epoch 15 | Step: 57 | Loss: 0.4443\n",
      "Epoch 15 | Step: 58 | Loss: 0.8784\n",
      "Epoch 15 | Step: 59 | Loss: 0.3521\n",
      "Epoch 15 | Step: 60 | Loss: 0.4297\n",
      "Epoch 15 | Step: 61 | Loss: 0.4909\n",
      "Epoch 15 | Step: 62 | Loss: 0.3034\n",
      "Epoch 15 | Step: 63 | Loss: 0.6421\n",
      "Epoch 15 | Step: 64 | Loss: 0.2314\n",
      "Epoch 15 | Step: 65 | Loss: 0.4493\n",
      "Epoch 15 | Step: 66 | Loss: 0.4415\n",
      "Epoch 15 | Step: 67 | Loss: 0.3488\n",
      "Epoch 15 | Step: 68 | Loss: 0.3026\n",
      "Epoch 15 | Step: 69 | Loss: 0.3936\n",
      "Epoch 15 | Step: 70 | Loss: 0.3344\n",
      "Epoch 15 | Step: 71 | Loss: 0.5185\n",
      "Epoch 15 | Step: 72 | Loss: 0.2060\n",
      "Epoch 15 | Step: 73 | Loss: 0.5173\n",
      "Epoch 15 | Step: 74 | Loss: 0.4784\n",
      "Epoch 15 | Step: 75 | Loss: 0.5630\n",
      "Epoch 15 | Step: 76 | Loss: 0.5386\n",
      "Epoch 15 | Step: 77 | Loss: 0.3415\n",
      "Epoch 15 | Step: 78 | Loss: 0.3247\n",
      "Epoch 15 | Step: 79 | Loss: 0.4971\n",
      "Epoch 15 | Step: 80 | Loss: 0.5057\n",
      "Epoch 15 | Step: 81 | Loss: 0.4381\n",
      "Epoch 15 | Step: 82 | Loss: 0.6078\n",
      "Epoch 15 | Step: 83 | Loss: 0.4890\n",
      "Epoch 15 | Step: 84 | Loss: 0.2804\n",
      "Epoch 15 | Step: 85 | Loss: 1.6793\n",
      "Epoch 15 | Step: 86 | Loss: 0.3951\n",
      "Epoch 15 | Step: 87 | Loss: 0.3420\n",
      "Epoch 15 | Step: 88 | Loss: 0.4917\n",
      "Epoch 15 | Step: 89 | Loss: 0.3863\n",
      "Epoch 15 | Step: 90 | Loss: 0.3968\n",
      "Epoch 15 | Step: 91 | Loss: 0.6800\n",
      "Epoch 15 | Step: 92 | Loss: 0.3523\n",
      "Epoch 15 | Step: 93 | Loss: 0.5038\n",
      "Epoch 15 | Step: 94 | Loss: 0.3139\n",
      "Epoch 15 | Step: 95 | Loss: 0.4983\n",
      "Epoch 15 | Step: 96 | Loss: 0.4616\n",
      "Epoch 15 | Step: 97 | Loss: 0.4390\n",
      "Epoch 15 | Step: 98 | Loss: 0.4584\n",
      "Epoch 15 | Step: 99 | Loss: 0.6616\n",
      "Epoch 15 | Step: 100 | Loss: 0.3492\n",
      "Epoch 15 | Step: 101 | Loss: 0.4163\n",
      "Epoch 15 | Step: 102 | Loss: 0.2886\n",
      "Epoch 15 | Step: 103 | Loss: 0.4872\n",
      "Epoch 15 | Step: 104 | Loss: 0.3929\n",
      "Epoch 15 | Step: 105 | Loss: 0.6237\n",
      "Epoch 15 | Step: 106 | Loss: 0.3989\n",
      "Epoch 15 | Step: 107 | Loss: 0.3616\n",
      "Epoch 15 | Step: 108 | Loss: 0.7121\n",
      "Epoch 15 | Step: 109 | Loss: 0.4997\n",
      "Epoch 15 | Step: 110 | Loss: 0.4587\n",
      "Epoch 15 | Step: 111 | Loss: 0.4621\n",
      "Epoch 15 | Step: 112 | Loss: 0.3856\n",
      "Epoch 15 | Step: 113 | Loss: 0.4343\n",
      "Epoch 15 | Step: 114 | Loss: 0.4114\n",
      "Epoch 15 | Step: 115 | Loss: 0.5828\n",
      "Epoch 15 | Step: 116 | Loss: 0.4449\n",
      "Epoch 15 | Step: 117 | Loss: 0.4602\n",
      "Epoch 15 | Step: 118 | Loss: 0.4522\n",
      "Epoch 15 | Step: 119 | Loss: 0.2468\n",
      "Epoch 15 | Step: 120 | Loss: 0.5124\n",
      "Epoch 15 | Step: 121 | Loss: 0.5473\n",
      "Epoch 15 | Step: 122 | Loss: 0.4491\n",
      "Epoch 15 | Step: 123 | Loss: 0.3734\n",
      "Epoch 15 | Step: 124 | Loss: 0.3291\n",
      "Epoch 15 | Step: 125 | Loss: 0.4551\n",
      "Epoch 15 | Step: 126 | Loss: 0.4111\n",
      "Epoch 15 | Step: 127 | Loss: 0.4361\n",
      "Epoch 15 | Step: 128 | Loss: 0.4570\n",
      "Epoch 16 | Step: 1 | Loss: 0.4598\n",
      "Epoch 16 | Step: 2 | Loss: 0.2468\n",
      "Epoch 16 | Step: 3 | Loss: 0.3574\n",
      "Epoch 16 | Step: 4 | Loss: 0.3703\n",
      "Epoch 16 | Step: 5 | Loss: 0.2338\n",
      "Epoch 16 | Step: 6 | Loss: 0.3664\n",
      "Epoch 16 | Step: 7 | Loss: 0.4105\n",
      "Epoch 16 | Step: 8 | Loss: 0.2872\n",
      "Epoch 16 | Step: 9 | Loss: 0.2211\n",
      "Epoch 16 | Step: 10 | Loss: 0.3809\n",
      "Epoch 16 | Step: 11 | Loss: 0.4300\n",
      "Epoch 16 | Step: 12 | Loss: 0.3412\n",
      "Epoch 16 | Step: 13 | Loss: 0.3450\n",
      "Epoch 16 | Step: 14 | Loss: 0.5324\n",
      "Epoch 16 | Step: 15 | Loss: 0.4550\n",
      "Epoch 16 | Step: 16 | Loss: 0.3473\n",
      "Epoch 16 | Step: 17 | Loss: 0.5187\n",
      "Epoch 16 | Step: 18 | Loss: 0.3961\n",
      "Epoch 16 | Step: 19 | Loss: 0.3331\n",
      "Epoch 16 | Step: 20 | Loss: 0.3547\n",
      "Epoch 16 | Step: 21 | Loss: 0.4390\n",
      "Epoch 16 | Step: 22 | Loss: 0.7038\n",
      "Epoch 16 | Step: 23 | Loss: 0.2678\n",
      "Epoch 16 | Step: 24 | Loss: 0.5983\n",
      "Epoch 16 | Step: 25 | Loss: 0.4935\n",
      "Epoch 16 | Step: 26 | Loss: 0.3005\n",
      "Epoch 16 | Step: 27 | Loss: 0.4879\n",
      "Epoch 16 | Step: 28 | Loss: 0.5079\n",
      "Epoch 16 | Step: 29 | Loss: 0.3716\n",
      "Epoch 16 | Step: 30 | Loss: 0.4172\n",
      "Epoch 16 | Step: 31 | Loss: 0.3132\n",
      "Epoch 16 | Step: 32 | Loss: 0.4043\n",
      "Epoch 16 | Step: 33 | Loss: 0.4965\n",
      "Epoch 16 | Step: 34 | Loss: 0.3044\n",
      "Epoch 16 | Step: 35 | Loss: 0.3026\n",
      "Epoch 16 | Step: 36 | Loss: 0.5318\n",
      "Epoch 16 | Step: 37 | Loss: 0.3029\n",
      "Epoch 16 | Step: 38 | Loss: 0.6121\n",
      "Epoch 16 | Step: 39 | Loss: 0.2264\n",
      "Epoch 16 | Step: 40 | Loss: 1.3280\n",
      "Epoch 16 | Step: 41 | Loss: 0.3894\n",
      "Epoch 16 | Step: 42 | Loss: 0.3949\n",
      "Epoch 16 | Step: 43 | Loss: 0.4306\n",
      "Epoch 16 | Step: 44 | Loss: 0.6699\n",
      "Epoch 16 | Step: 45 | Loss: 0.4635\n",
      "Epoch 16 | Step: 46 | Loss: 0.3772\n",
      "Epoch 16 | Step: 47 | Loss: 0.3349\n",
      "Epoch 16 | Step: 48 | Loss: 0.3707\n",
      "Epoch 16 | Step: 49 | Loss: 0.4333\n",
      "Epoch 16 | Step: 50 | Loss: 0.4078\n",
      "Epoch 16 | Step: 51 | Loss: 0.4163\n",
      "Epoch 16 | Step: 52 | Loss: 0.3595\n",
      "Epoch 16 | Step: 53 | Loss: 0.6343\n",
      "Epoch 16 | Step: 54 | Loss: 0.6346\n",
      "Epoch 16 | Step: 55 | Loss: 0.4144\n",
      "Epoch 16 | Step: 56 | Loss: 0.4385\n",
      "Epoch 16 | Step: 57 | Loss: 0.4422\n",
      "Epoch 16 | Step: 58 | Loss: 0.4513\n",
      "Epoch 16 | Step: 59 | Loss: 0.3674\n",
      "Epoch 16 | Step: 60 | Loss: 0.3865\n",
      "Epoch 16 | Step: 61 | Loss: 0.3929\n",
      "Epoch 16 | Step: 62 | Loss: 0.5078\n",
      "Epoch 16 | Step: 63 | Loss: 0.4766\n",
      "Epoch 16 | Step: 64 | Loss: 0.5200\n",
      "Epoch 16 | Step: 65 | Loss: 0.5742\n",
      "Epoch 16 | Step: 66 | Loss: 0.4149\n",
      "Epoch 16 | Step: 67 | Loss: 0.4302\n",
      "Epoch 16 | Step: 68 | Loss: 0.6364\n",
      "Epoch 16 | Step: 69 | Loss: 0.6479\n",
      "Epoch 16 | Step: 70 | Loss: 0.3887\n",
      "Epoch 16 | Step: 71 | Loss: 0.9454\n",
      "Epoch 16 | Step: 72 | Loss: 0.5004\n",
      "Epoch 16 | Step: 73 | Loss: 0.4476\n",
      "Epoch 16 | Step: 74 | Loss: 0.4689\n",
      "Epoch 16 | Step: 75 | Loss: 0.4518\n",
      "Epoch 16 | Step: 76 | Loss: 0.4518\n",
      "Epoch 16 | Step: 77 | Loss: 0.4005\n",
      "Epoch 16 | Step: 78 | Loss: 0.4172\n",
      "Epoch 16 | Step: 79 | Loss: 0.3618\n",
      "Epoch 16 | Step: 80 | Loss: 0.4247\n",
      "Epoch 16 | Step: 81 | Loss: 0.5999\n",
      "Epoch 16 | Step: 82 | Loss: 0.2104\n",
      "Epoch 16 | Step: 83 | Loss: 0.3589\n",
      "Epoch 16 | Step: 84 | Loss: 0.3542\n",
      "Epoch 16 | Step: 85 | Loss: 0.4785\n",
      "Epoch 16 | Step: 86 | Loss: 0.2858\n",
      "Epoch 16 | Step: 87 | Loss: 0.2166\n",
      "Epoch 16 | Step: 88 | Loss: 0.3854\n",
      "Epoch 16 | Step: 89 | Loss: 0.3434\n",
      "Epoch 16 | Step: 90 | Loss: 0.4085\n",
      "Epoch 16 | Step: 91 | Loss: 0.6190\n",
      "Epoch 16 | Step: 92 | Loss: 0.3606\n",
      "Epoch 16 | Step: 93 | Loss: 0.3581\n",
      "Epoch 16 | Step: 94 | Loss: 0.3882\n",
      "Epoch 16 | Step: 95 | Loss: 0.3613\n",
      "Epoch 16 | Step: 96 | Loss: 0.4055\n",
      "Epoch 16 | Step: 97 | Loss: 0.3691\n",
      "Epoch 16 | Step: 98 | Loss: 0.4354\n",
      "Epoch 16 | Step: 99 | Loss: 0.3749\n",
      "Epoch 16 | Step: 100 | Loss: 0.4345\n",
      "Epoch 16 | Step: 101 | Loss: 0.5902\n",
      "Epoch 16 | Step: 102 | Loss: 0.5132\n",
      "Epoch 16 | Step: 103 | Loss: 0.4280\n",
      "Epoch 16 | Step: 104 | Loss: 0.1976\n",
      "Epoch 16 | Step: 105 | Loss: 0.2121\n",
      "Epoch 16 | Step: 106 | Loss: 0.6852\n",
      "Epoch 16 | Step: 107 | Loss: 0.3609\n",
      "Epoch 16 | Step: 108 | Loss: 0.2610\n",
      "Epoch 16 | Step: 109 | Loss: 0.4661\n",
      "Epoch 16 | Step: 110 | Loss: 0.2440\n",
      "Epoch 16 | Step: 111 | Loss: 0.4938\n",
      "Epoch 16 | Step: 112 | Loss: 0.3336\n",
      "Epoch 16 | Step: 113 | Loss: 0.7446\n",
      "Epoch 16 | Step: 114 | Loss: 0.4658\n",
      "Epoch 16 | Step: 115 | Loss: 0.4742\n",
      "Epoch 16 | Step: 116 | Loss: 0.5495\n",
      "Epoch 16 | Step: 117 | Loss: 0.2956\n",
      "Epoch 16 | Step: 118 | Loss: 0.4849\n",
      "Epoch 16 | Step: 119 | Loss: 0.3277\n",
      "Epoch 16 | Step: 120 | Loss: 0.4343\n",
      "Epoch 16 | Step: 121 | Loss: 0.4125\n",
      "Epoch 16 | Step: 122 | Loss: 0.2543\n",
      "Epoch 16 | Step: 123 | Loss: 1.4572\n",
      "Epoch 16 | Step: 124 | Loss: 0.3454\n",
      "Epoch 16 | Step: 125 | Loss: 0.4125\n",
      "Epoch 16 | Step: 126 | Loss: 0.4283\n",
      "Epoch 16 | Step: 127 | Loss: 0.4088\n",
      "Epoch 16 | Step: 128 | Loss: 0.3068\n",
      "Epoch 17 | Step: 1 | Loss: 0.4361\n",
      "Epoch 17 | Step: 2 | Loss: 0.2332\n",
      "Epoch 17 | Step: 3 | Loss: 0.4143\n",
      "Epoch 17 | Step: 4 | Loss: 0.4865\n",
      "Epoch 17 | Step: 5 | Loss: 0.4473\n",
      "Epoch 17 | Step: 6 | Loss: 0.3878\n",
      "Epoch 17 | Step: 7 | Loss: 0.4475\n",
      "Epoch 17 | Step: 8 | Loss: 0.3334\n",
      "Epoch 17 | Step: 9 | Loss: 0.2845\n",
      "Epoch 17 | Step: 10 | Loss: 0.8350\n",
      "Epoch 17 | Step: 11 | Loss: 0.3733\n",
      "Epoch 17 | Step: 12 | Loss: 0.4234\n",
      "Epoch 17 | Step: 13 | Loss: 0.5468\n",
      "Epoch 17 | Step: 14 | Loss: 0.3777\n",
      "Epoch 17 | Step: 15 | Loss: 0.4224\n",
      "Epoch 17 | Step: 16 | Loss: 0.4839\n",
      "Epoch 17 | Step: 17 | Loss: 0.4846\n",
      "Epoch 17 | Step: 18 | Loss: 0.3537\n",
      "Epoch 17 | Step: 19 | Loss: 0.2468\n",
      "Epoch 17 | Step: 20 | Loss: 0.3729\n",
      "Epoch 17 | Step: 21 | Loss: 0.3199\n",
      "Epoch 17 | Step: 22 | Loss: 0.3810\n",
      "Epoch 17 | Step: 23 | Loss: 0.5029\n",
      "Epoch 17 | Step: 24 | Loss: 0.2809\n",
      "Epoch 17 | Step: 25 | Loss: 0.3659\n",
      "Epoch 17 | Step: 26 | Loss: 0.3798\n",
      "Epoch 17 | Step: 27 | Loss: 0.2693\n",
      "Epoch 17 | Step: 28 | Loss: 0.2439\n",
      "Epoch 17 | Step: 29 | Loss: 0.4229\n",
      "Epoch 17 | Step: 30 | Loss: 0.5245\n",
      "Epoch 17 | Step: 31 | Loss: 1.3843\n",
      "Epoch 17 | Step: 32 | Loss: 0.4787\n",
      "Epoch 17 | Step: 33 | Loss: 0.3663\n",
      "Epoch 17 | Step: 34 | Loss: 0.4181\n",
      "Epoch 17 | Step: 35 | Loss: 0.5144\n",
      "Epoch 17 | Step: 36 | Loss: 0.3337\n",
      "Epoch 17 | Step: 37 | Loss: 0.8165\n",
      "Epoch 17 | Step: 38 | Loss: 0.2520\n",
      "Epoch 17 | Step: 39 | Loss: 0.2877\n",
      "Epoch 17 | Step: 40 | Loss: 0.3427\n",
      "Epoch 17 | Step: 41 | Loss: 1.1291\n",
      "Epoch 17 | Step: 42 | Loss: 0.3044\n",
      "Epoch 17 | Step: 43 | Loss: 0.2858\n",
      "Epoch 17 | Step: 44 | Loss: 0.4635\n",
      "Epoch 17 | Step: 45 | Loss: 0.5106\n",
      "Epoch 17 | Step: 46 | Loss: 0.3079\n",
      "Epoch 17 | Step: 47 | Loss: 0.3481\n",
      "Epoch 17 | Step: 48 | Loss: 0.3628\n",
      "Epoch 17 | Step: 49 | Loss: 0.4203\n",
      "Epoch 17 | Step: 50 | Loss: 0.4152\n",
      "Epoch 17 | Step: 51 | Loss: 0.4402\n",
      "Epoch 17 | Step: 52 | Loss: 0.3685\n",
      "Epoch 17 | Step: 53 | Loss: 0.5644\n",
      "Epoch 17 | Step: 54 | Loss: 0.2830\n",
      "Epoch 17 | Step: 55 | Loss: 0.5689\n",
      "Epoch 17 | Step: 56 | Loss: 0.3621\n",
      "Epoch 17 | Step: 57 | Loss: 0.4139\n",
      "Epoch 17 | Step: 58 | Loss: 0.3585\n",
      "Epoch 17 | Step: 59 | Loss: 0.3689\n",
      "Epoch 17 | Step: 60 | Loss: 0.2715\n",
      "Epoch 17 | Step: 61 | Loss: 0.3480\n",
      "Epoch 17 | Step: 62 | Loss: 0.4141\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 17 | Step: 63 | Loss: 0.4146\n",
      "Epoch 17 | Step: 64 | Loss: 0.3831\n",
      "Epoch 17 | Step: 65 | Loss: 0.3455\n",
      "Epoch 17 | Step: 66 | Loss: 0.2951\n",
      "Epoch 17 | Step: 67 | Loss: 0.4150\n",
      "Epoch 17 | Step: 68 | Loss: 0.3194\n",
      "Epoch 17 | Step: 69 | Loss: 0.4601\n",
      "Epoch 17 | Step: 70 | Loss: 0.3228\n",
      "Epoch 17 | Step: 71 | Loss: 0.2965\n",
      "Epoch 17 | Step: 72 | Loss: 0.4270\n",
      "Epoch 17 | Step: 73 | Loss: 0.3511\n",
      "Epoch 17 | Step: 74 | Loss: 0.2599\n",
      "Epoch 17 | Step: 75 | Loss: 0.4752\n",
      "Epoch 17 | Step: 76 | Loss: 0.2448\n",
      "Epoch 17 | Step: 77 | Loss: 0.2722\n",
      "Epoch 17 | Step: 78 | Loss: 0.3543\n",
      "Epoch 17 | Step: 79 | Loss: 0.2692\n",
      "Epoch 17 | Step: 80 | Loss: 0.4294\n",
      "Epoch 17 | Step: 81 | Loss: 0.4695\n",
      "Epoch 17 | Step: 82 | Loss: 0.3710\n",
      "Epoch 17 | Step: 83 | Loss: 0.5614\n",
      "Epoch 17 | Step: 84 | Loss: 0.3651\n",
      "Epoch 17 | Step: 85 | Loss: 0.2494\n",
      "Epoch 17 | Step: 86 | Loss: 0.3321\n",
      "Epoch 17 | Step: 87 | Loss: 0.4805\n",
      "Epoch 17 | Step: 88 | Loss: 0.5176\n",
      "Epoch 17 | Step: 89 | Loss: 0.1590\n",
      "Epoch 17 | Step: 90 | Loss: 0.4137\n",
      "Epoch 17 | Step: 91 | Loss: 0.3188\n",
      "Epoch 17 | Step: 92 | Loss: 0.4253\n",
      "Epoch 17 | Step: 93 | Loss: 0.4769\n",
      "Epoch 17 | Step: 94 | Loss: 0.4461\n",
      "Epoch 17 | Step: 95 | Loss: 0.4088\n",
      "Epoch 17 | Step: 96 | Loss: 0.5394\n",
      "Epoch 17 | Step: 97 | Loss: 0.3954\n",
      "Epoch 17 | Step: 98 | Loss: 0.4415\n",
      "Epoch 17 | Step: 99 | Loss: 0.5001\n",
      "Epoch 17 | Step: 100 | Loss: 0.3957\n",
      "Epoch 17 | Step: 101 | Loss: 0.2780\n",
      "Epoch 17 | Step: 102 | Loss: 0.4085\n",
      "Epoch 17 | Step: 103 | Loss: 0.2643\n",
      "Epoch 17 | Step: 104 | Loss: 0.5373\n",
      "Epoch 17 | Step: 105 | Loss: 0.3962\n",
      "Epoch 17 | Step: 106 | Loss: 0.5016\n",
      "Epoch 17 | Step: 107 | Loss: 0.2731\n",
      "Epoch 17 | Step: 108 | Loss: 0.6223\n",
      "Epoch 17 | Step: 109 | Loss: 0.5492\n",
      "Epoch 17 | Step: 110 | Loss: 0.4765\n",
      "Epoch 17 | Step: 111 | Loss: 0.3145\n",
      "Epoch 17 | Step: 112 | Loss: 0.4667\n",
      "Epoch 17 | Step: 113 | Loss: 0.5717\n",
      "Epoch 17 | Step: 114 | Loss: 0.4471\n",
      "Epoch 17 | Step: 115 | Loss: 0.4707\n",
      "Epoch 17 | Step: 116 | Loss: 0.3256\n",
      "Epoch 17 | Step: 117 | Loss: 0.3784\n",
      "Epoch 17 | Step: 118 | Loss: 0.4756\n",
      "Epoch 17 | Step: 119 | Loss: 0.4172\n",
      "Epoch 17 | Step: 120 | Loss: 0.4982\n",
      "Epoch 17 | Step: 121 | Loss: 0.4233\n",
      "Epoch 17 | Step: 122 | Loss: 0.3475\n",
      "Epoch 17 | Step: 123 | Loss: 0.5027\n",
      "Epoch 17 | Step: 124 | Loss: 0.2740\n",
      "Epoch 17 | Step: 125 | Loss: 0.3344\n",
      "Epoch 17 | Step: 126 | Loss: 0.2726\n",
      "Epoch 17 | Step: 127 | Loss: 0.3960\n",
      "Epoch 17 | Step: 128 | Loss: 0.3946\n",
      "Epoch 18 | Step: 1 | Loss: 0.1831\n",
      "Epoch 18 | Step: 2 | Loss: 0.3968\n",
      "Epoch 18 | Step: 3 | Loss: 0.4946\n",
      "Epoch 18 | Step: 4 | Loss: 0.2551\n",
      "Epoch 18 | Step: 5 | Loss: 0.3041\n",
      "Epoch 18 | Step: 6 | Loss: 0.3073\n",
      "Epoch 18 | Step: 7 | Loss: 0.4341\n",
      "Epoch 18 | Step: 8 | Loss: 0.3668\n",
      "Epoch 18 | Step: 9 | Loss: 0.3376\n",
      "Epoch 18 | Step: 10 | Loss: 0.2954\n",
      "Epoch 18 | Step: 11 | Loss: 0.1967\n",
      "Epoch 18 | Step: 12 | Loss: 0.4866\n",
      "Epoch 18 | Step: 13 | Loss: 0.3013\n",
      "Epoch 18 | Step: 14 | Loss: 0.4071\n",
      "Epoch 18 | Step: 15 | Loss: 0.3778\n",
      "Epoch 18 | Step: 16 | Loss: 0.3178\n",
      "Epoch 18 | Step: 17 | Loss: 0.4618\n",
      "Epoch 18 | Step: 18 | Loss: 0.3294\n",
      "Epoch 18 | Step: 19 | Loss: 0.5106\n",
      "Epoch 18 | Step: 20 | Loss: 0.5949\n",
      "Epoch 18 | Step: 21 | Loss: 0.4969\n",
      "Epoch 18 | Step: 22 | Loss: 0.3828\n",
      "Epoch 18 | Step: 23 | Loss: 0.4944\n",
      "Epoch 18 | Step: 24 | Loss: 0.2415\n",
      "Epoch 18 | Step: 25 | Loss: 0.2920\n",
      "Epoch 18 | Step: 26 | Loss: 0.5205\n",
      "Epoch 18 | Step: 27 | Loss: 0.3696\n",
      "Epoch 18 | Step: 28 | Loss: 0.3518\n",
      "Epoch 18 | Step: 29 | Loss: 0.5235\n",
      "Epoch 18 | Step: 30 | Loss: 0.2577\n",
      "Epoch 18 | Step: 31 | Loss: 0.4552\n",
      "Epoch 18 | Step: 32 | Loss: 0.2353\n",
      "Epoch 18 | Step: 33 | Loss: 0.4348\n",
      "Epoch 18 | Step: 34 | Loss: 0.3910\n",
      "Epoch 18 | Step: 35 | Loss: 0.4016\n",
      "Epoch 18 | Step: 36 | Loss: 0.3137\n",
      "Epoch 18 | Step: 37 | Loss: 0.5091\n",
      "Epoch 18 | Step: 38 | Loss: 0.5304\n",
      "Epoch 18 | Step: 39 | Loss: 0.2626\n",
      "Epoch 18 | Step: 40 | Loss: 0.4376\n",
      "Epoch 18 | Step: 41 | Loss: 0.4945\n",
      "Epoch 18 | Step: 42 | Loss: 0.4672\n",
      "Epoch 18 | Step: 43 | Loss: 0.4336\n",
      "Epoch 18 | Step: 44 | Loss: 0.2337\n",
      "Epoch 18 | Step: 45 | Loss: 0.4966\n",
      "Epoch 18 | Step: 46 | Loss: 0.2097\n",
      "Epoch 18 | Step: 47 | Loss: 0.3069\n",
      "Epoch 18 | Step: 48 | Loss: 0.2853\n",
      "Epoch 18 | Step: 49 | Loss: 0.3570\n",
      "Epoch 18 | Step: 50 | Loss: 0.4315\n",
      "Epoch 18 | Step: 51 | Loss: 1.0373\n",
      "Epoch 18 | Step: 52 | Loss: 0.4458\n",
      "Epoch 18 | Step: 53 | Loss: 0.4239\n",
      "Epoch 18 | Step: 54 | Loss: 0.3567\n",
      "Epoch 18 | Step: 55 | Loss: 0.3124\n",
      "Epoch 18 | Step: 56 | Loss: 0.2296\n",
      "Epoch 18 | Step: 57 | Loss: 0.3834\n",
      "Epoch 18 | Step: 58 | Loss: 0.4972\n",
      "Epoch 18 | Step: 59 | Loss: 0.3365\n",
      "Epoch 18 | Step: 60 | Loss: 0.2779\n",
      "Epoch 18 | Step: 61 | Loss: 0.3202\n",
      "Epoch 18 | Step: 62 | Loss: 0.4061\n",
      "Epoch 18 | Step: 63 | Loss: 0.4646\n",
      "Epoch 18 | Step: 64 | Loss: 0.2915\n",
      "Epoch 18 | Step: 65 | Loss: 0.3043\n",
      "Epoch 18 | Step: 66 | Loss: 0.4510\n",
      "Epoch 18 | Step: 67 | Loss: 0.5325\n",
      "Epoch 18 | Step: 68 | Loss: 0.3620\n",
      "Epoch 18 | Step: 69 | Loss: 0.5547\n",
      "Epoch 18 | Step: 70 | Loss: 0.4305\n",
      "Epoch 18 | Step: 71 | Loss: 0.4417\n",
      "Epoch 18 | Step: 72 | Loss: 0.4856\n",
      "Epoch 18 | Step: 73 | Loss: 0.3322\n",
      "Epoch 18 | Step: 74 | Loss: 0.3163\n",
      "Epoch 18 | Step: 75 | Loss: 0.3433\n",
      "Epoch 18 | Step: 76 | Loss: 0.3175\n",
      "Epoch 18 | Step: 77 | Loss: 0.3311\n",
      "Epoch 18 | Step: 78 | Loss: 0.4359\n",
      "Epoch 18 | Step: 79 | Loss: 0.4666\n",
      "Epoch 18 | Step: 80 | Loss: 0.5958\n",
      "Epoch 18 | Step: 81 | Loss: 0.4614\n",
      "Epoch 18 | Step: 82 | Loss: 0.5890\n",
      "Epoch 18 | Step: 83 | Loss: 0.2706\n",
      "Epoch 18 | Step: 84 | Loss: 0.5657\n",
      "Epoch 18 | Step: 85 | Loss: 0.1556\n",
      "Epoch 18 | Step: 86 | Loss: 0.3153\n",
      "Epoch 18 | Step: 87 | Loss: 0.3872\n",
      "Epoch 18 | Step: 88 | Loss: 0.4479\n",
      "Epoch 18 | Step: 89 | Loss: 0.2774\n",
      "Epoch 18 | Step: 90 | Loss: 0.4265\n",
      "Epoch 18 | Step: 91 | Loss: 0.3259\n",
      "Epoch 18 | Step: 92 | Loss: 0.5487\n",
      "Epoch 18 | Step: 93 | Loss: 0.3463\n",
      "Epoch 18 | Step: 94 | Loss: 0.2080\n",
      "Epoch 18 | Step: 95 | Loss: 0.3055\n",
      "Epoch 18 | Step: 96 | Loss: 0.2018\n",
      "Epoch 18 | Step: 97 | Loss: 0.5193\n",
      "Epoch 18 | Step: 98 | Loss: 0.3508\n",
      "Epoch 18 | Step: 99 | Loss: 0.2358\n",
      "Epoch 18 | Step: 100 | Loss: 0.3910\n",
      "Epoch 18 | Step: 101 | Loss: 0.3807\n",
      "Epoch 18 | Step: 102 | Loss: 0.2698\n",
      "Epoch 18 | Step: 103 | Loss: 0.4026\n",
      "Epoch 18 | Step: 104 | Loss: 0.3175\n",
      "Epoch 18 | Step: 105 | Loss: 0.3822\n",
      "Epoch 18 | Step: 106 | Loss: 0.4407\n",
      "Epoch 18 | Step: 107 | Loss: 1.1820\n",
      "Epoch 18 | Step: 108 | Loss: 0.5622\n",
      "Epoch 18 | Step: 109 | Loss: 0.4017\n",
      "Epoch 18 | Step: 110 | Loss: 0.3650\n",
      "Epoch 18 | Step: 111 | Loss: 0.2756\n",
      "Epoch 18 | Step: 112 | Loss: 0.3234\n",
      "Epoch 18 | Step: 113 | Loss: 0.4370\n",
      "Epoch 18 | Step: 114 | Loss: 0.2160\n",
      "Epoch 18 | Step: 115 | Loss: 0.2470\n",
      "Epoch 18 | Step: 116 | Loss: 0.5816\n",
      "Epoch 18 | Step: 117 | Loss: 0.4549\n",
      "Epoch 18 | Step: 118 | Loss: 0.4435\n",
      "Epoch 18 | Step: 119 | Loss: 0.3709\n",
      "Epoch 18 | Step: 120 | Loss: 0.2352\n",
      "Epoch 18 | Step: 121 | Loss: 0.5932\n",
      "Epoch 18 | Step: 122 | Loss: 0.5696\n",
      "Epoch 18 | Step: 123 | Loss: 0.4879\n",
      "Epoch 18 | Step: 124 | Loss: 0.3115\n",
      "Epoch 18 | Step: 125 | Loss: 0.3367\n",
      "Epoch 18 | Step: 126 | Loss: 0.3144\n",
      "Epoch 18 | Step: 127 | Loss: 0.4910\n",
      "Epoch 18 | Step: 128 | Loss: 0.2432\n",
      "Epoch 19 | Step: 1 | Loss: 0.4385\n",
      "Epoch 19 | Step: 2 | Loss: 0.4133\n",
      "Epoch 19 | Step: 3 | Loss: 0.3647\n",
      "Epoch 19 | Step: 4 | Loss: 0.3825\n",
      "Epoch 19 | Step: 5 | Loss: 0.2766\n",
      "Epoch 19 | Step: 6 | Loss: 0.2952\n",
      "Epoch 19 | Step: 7 | Loss: 0.5289\n",
      "Epoch 19 | Step: 8 | Loss: 0.2756\n",
      "Epoch 19 | Step: 9 | Loss: 0.3401\n",
      "Epoch 19 | Step: 10 | Loss: 0.2988\n",
      "Epoch 19 | Step: 11 | Loss: 0.5459\n",
      "Epoch 19 | Step: 12 | Loss: 0.5335\n",
      "Epoch 19 | Step: 13 | Loss: 0.4277\n",
      "Epoch 19 | Step: 14 | Loss: 0.3930\n",
      "Epoch 19 | Step: 15 | Loss: 0.3350\n",
      "Epoch 19 | Step: 16 | Loss: 0.3855\n",
      "Epoch 19 | Step: 17 | Loss: 0.4349\n",
      "Epoch 19 | Step: 18 | Loss: 0.3761\n",
      "Epoch 19 | Step: 19 | Loss: 0.3676\n",
      "Epoch 19 | Step: 20 | Loss: 0.4642\n",
      "Epoch 19 | Step: 21 | Loss: 0.1748\n",
      "Epoch 19 | Step: 22 | Loss: 0.4359\n",
      "Epoch 19 | Step: 23 | Loss: 0.4072\n",
      "Epoch 19 | Step: 24 | Loss: 0.3215\n",
      "Epoch 19 | Step: 25 | Loss: 0.2955\n",
      "Epoch 19 | Step: 26 | Loss: 0.3188\n",
      "Epoch 19 | Step: 27 | Loss: 0.5142\n",
      "Epoch 19 | Step: 28 | Loss: 1.2206\n",
      "Epoch 19 | Step: 29 | Loss: 0.4943\n",
      "Epoch 19 | Step: 30 | Loss: 0.2433\n",
      "Epoch 19 | Step: 31 | Loss: 0.2365\n",
      "Epoch 19 | Step: 32 | Loss: 0.2299\n",
      "Epoch 19 | Step: 33 | Loss: 0.3458\n",
      "Epoch 19 | Step: 34 | Loss: 0.4246\n",
      "Epoch 19 | Step: 35 | Loss: 0.3138\n",
      "Epoch 19 | Step: 36 | Loss: 0.2460\n",
      "Epoch 19 | Step: 37 | Loss: 0.2961\n",
      "Epoch 19 | Step: 38 | Loss: 0.4692\n",
      "Epoch 19 | Step: 39 | Loss: 0.2371\n",
      "Epoch 19 | Step: 40 | Loss: 0.3342\n",
      "Epoch 19 | Step: 41 | Loss: 0.3951\n",
      "Epoch 19 | Step: 42 | Loss: 0.5595\n",
      "Epoch 19 | Step: 43 | Loss: 0.2611\n",
      "Epoch 19 | Step: 44 | Loss: 0.3819\n",
      "Epoch 19 | Step: 45 | Loss: 0.6154\n",
      "Epoch 19 | Step: 46 | Loss: 0.5068\n",
      "Epoch 19 | Step: 47 | Loss: 0.1564\n",
      "Epoch 19 | Step: 48 | Loss: 0.2848\n",
      "Epoch 19 | Step: 49 | Loss: 0.2039\n",
      "Epoch 19 | Step: 50 | Loss: 0.2100\n",
      "Epoch 19 | Step: 51 | Loss: 0.4558\n",
      "Epoch 19 | Step: 52 | Loss: 0.3589\n",
      "Epoch 19 | Step: 53 | Loss: 0.3623\n",
      "Epoch 19 | Step: 54 | Loss: 0.3028\n",
      "Epoch 19 | Step: 55 | Loss: 0.4286\n",
      "Epoch 19 | Step: 56 | Loss: 0.4967\n",
      "Epoch 19 | Step: 57 | Loss: 0.3040\n",
      "Epoch 19 | Step: 58 | Loss: 0.3815\n",
      "Epoch 19 | Step: 59 | Loss: 0.2421\n",
      "Epoch 19 | Step: 60 | Loss: 0.2246\n",
      "Epoch 19 | Step: 61 | Loss: 0.3432\n",
      "Epoch 19 | Step: 62 | Loss: 0.3279\n",
      "Epoch 19 | Step: 63 | Loss: 0.4600\n",
      "Epoch 19 | Step: 64 | Loss: 0.2724\n",
      "Epoch 19 | Step: 65 | Loss: 0.3234\n",
      "Epoch 19 | Step: 66 | Loss: 0.3527\n",
      "Epoch 19 | Step: 67 | Loss: 0.4203\n",
      "Epoch 19 | Step: 68 | Loss: 0.3116\n",
      "Epoch 19 | Step: 69 | Loss: 0.3086\n",
      "Epoch 19 | Step: 70 | Loss: 0.3648\n",
      "Epoch 19 | Step: 71 | Loss: 0.4280\n",
      "Epoch 19 | Step: 72 | Loss: 0.3170\n",
      "Epoch 19 | Step: 73 | Loss: 0.4374\n",
      "Epoch 19 | Step: 74 | Loss: 0.4257\n",
      "Epoch 19 | Step: 75 | Loss: 0.5361\n",
      "Epoch 19 | Step: 76 | Loss: 0.5717\n",
      "Epoch 19 | Step: 77 | Loss: 0.5923\n",
      "Epoch 19 | Step: 78 | Loss: 0.4220\n",
      "Epoch 19 | Step: 79 | Loss: 0.4252\n",
      "Epoch 19 | Step: 80 | Loss: 0.2658\n",
      "Epoch 19 | Step: 81 | Loss: 0.3724\n",
      "Epoch 19 | Step: 82 | Loss: 0.3479\n",
      "Epoch 19 | Step: 83 | Loss: 0.4120\n",
      "Epoch 19 | Step: 84 | Loss: 0.3934\n",
      "Epoch 19 | Step: 85 | Loss: 0.5104\n",
      "Epoch 19 | Step: 86 | Loss: 0.2710\n",
      "Epoch 19 | Step: 87 | Loss: 0.5070\n",
      "Epoch 19 | Step: 88 | Loss: 0.4567\n",
      "Epoch 19 | Step: 89 | Loss: 0.5092\n",
      "Epoch 19 | Step: 90 | Loss: 0.3394\n",
      "Epoch 19 | Step: 91 | Loss: 0.3521\n",
      "Epoch 19 | Step: 92 | Loss: 0.2980\n",
      "Epoch 19 | Step: 93 | Loss: 0.3608\n",
      "Epoch 19 | Step: 94 | Loss: 0.5100\n",
      "Epoch 19 | Step: 95 | Loss: 0.3524\n",
      "Epoch 19 | Step: 96 | Loss: 0.2862\n",
      "Epoch 19 | Step: 97 | Loss: 0.4216\n",
      "Epoch 19 | Step: 98 | Loss: 0.4291\n",
      "Epoch 19 | Step: 99 | Loss: 0.3356\n",
      "Epoch 19 | Step: 100 | Loss: 0.3243\n",
      "Epoch 19 | Step: 101 | Loss: 0.3970\n",
      "Epoch 19 | Step: 102 | Loss: 0.2322\n",
      "Epoch 19 | Step: 103 | Loss: 0.3266\n",
      "Epoch 19 | Step: 104 | Loss: 0.3318\n",
      "Epoch 19 | Step: 105 | Loss: 0.4038\n",
      "Epoch 19 | Step: 106 | Loss: 0.2523\n",
      "Epoch 19 | Step: 107 | Loss: 0.3217\n",
      "Epoch 19 | Step: 108 | Loss: 0.4154\n",
      "Epoch 19 | Step: 109 | Loss: 0.4240\n",
      "Epoch 19 | Step: 110 | Loss: 0.3823\n",
      "Epoch 19 | Step: 111 | Loss: 0.4685\n",
      "Epoch 19 | Step: 112 | Loss: 0.8254\n",
      "Epoch 19 | Step: 113 | Loss: 0.3204\n",
      "Epoch 19 | Step: 114 | Loss: 0.2305\n",
      "Epoch 19 | Step: 115 | Loss: 0.3014\n",
      "Epoch 19 | Step: 116 | Loss: 0.5380\n",
      "Epoch 19 | Step: 117 | Loss: 0.2697\n",
      "Epoch 19 | Step: 118 | Loss: 0.2897\n",
      "Epoch 19 | Step: 119 | Loss: 0.3057\n",
      "Epoch 19 | Step: 120 | Loss: 0.3269\n",
      "Epoch 19 | Step: 121 | Loss: 0.3481\n",
      "Epoch 19 | Step: 122 | Loss: 0.2482\n",
      "Epoch 19 | Step: 123 | Loss: 0.2702\n",
      "Epoch 19 | Step: 124 | Loss: 0.3340\n",
      "Epoch 19 | Step: 125 | Loss: 0.3123\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19 | Step: 126 | Loss: 0.5180\n",
      "Epoch 19 | Step: 127 | Loss: 0.3844\n",
      "Epoch 19 | Step: 128 | Loss: 0.2928\n",
      "Epoch 20 | Step: 1 | Loss: 0.2893\n",
      "Epoch 20 | Step: 2 | Loss: 0.3254\n",
      "Epoch 20 | Step: 3 | Loss: 0.2691\n",
      "Epoch 20 | Step: 4 | Loss: 0.3492\n",
      "Epoch 20 | Step: 5 | Loss: 0.4164\n",
      "Epoch 20 | Step: 6 | Loss: 0.3221\n",
      "Epoch 20 | Step: 7 | Loss: 0.3383\n",
      "Epoch 20 | Step: 8 | Loss: 0.6690\n",
      "Epoch 20 | Step: 9 | Loss: 0.2939\n",
      "Epoch 20 | Step: 10 | Loss: 0.3686\n",
      "Epoch 20 | Step: 11 | Loss: 0.4336\n",
      "Epoch 20 | Step: 12 | Loss: 0.4856\n",
      "Epoch 20 | Step: 13 | Loss: 0.2380\n",
      "Epoch 20 | Step: 14 | Loss: 0.3682\n",
      "Epoch 20 | Step: 15 | Loss: 0.2214\n",
      "Epoch 20 | Step: 16 | Loss: 0.2837\n",
      "Epoch 20 | Step: 17 | Loss: 0.3866\n",
      "Epoch 20 | Step: 18 | Loss: 0.5149\n",
      "Epoch 20 | Step: 19 | Loss: 0.3290\n",
      "Epoch 20 | Step: 20 | Loss: 0.3740\n",
      "Epoch 20 | Step: 21 | Loss: 0.2726\n",
      "Epoch 20 | Step: 22 | Loss: 0.3925\n",
      "Epoch 20 | Step: 23 | Loss: 0.3415\n",
      "Epoch 20 | Step: 24 | Loss: 0.3897\n",
      "Epoch 20 | Step: 25 | Loss: 0.3681\n",
      "Epoch 20 | Step: 26 | Loss: 0.5825\n",
      "Epoch 20 | Step: 27 | Loss: 0.3760\n",
      "Epoch 20 | Step: 28 | Loss: 1.1274\n",
      "Epoch 20 | Step: 29 | Loss: 0.3276\n",
      "Epoch 20 | Step: 30 | Loss: 0.2914\n",
      "Epoch 20 | Step: 31 | Loss: 0.4340\n",
      "Epoch 20 | Step: 32 | Loss: 0.2494\n",
      "Epoch 20 | Step: 33 | Loss: 0.3837\n",
      "Epoch 20 | Step: 34 | Loss: 0.3739\n",
      "Epoch 20 | Step: 35 | Loss: 0.3320\n",
      "Epoch 20 | Step: 36 | Loss: 0.2025\n",
      "Epoch 20 | Step: 37 | Loss: 0.3630\n",
      "Epoch 20 | Step: 38 | Loss: 0.4617\n",
      "Epoch 20 | Step: 39 | Loss: 0.4723\n",
      "Epoch 20 | Step: 40 | Loss: 0.2168\n",
      "Epoch 20 | Step: 41 | Loss: 0.1982\n",
      "Epoch 20 | Step: 42 | Loss: 0.2259\n",
      "Epoch 20 | Step: 43 | Loss: 0.3506\n",
      "Epoch 20 | Step: 44 | Loss: 0.3053\n",
      "Epoch 20 | Step: 45 | Loss: 0.3658\n",
      "Epoch 20 | Step: 46 | Loss: 0.2050\n",
      "Epoch 20 | Step: 47 | Loss: 0.4448\n",
      "Epoch 20 | Step: 48 | Loss: 0.3873\n",
      "Epoch 20 | Step: 49 | Loss: 0.3349\n",
      "Epoch 20 | Step: 50 | Loss: 0.4185\n",
      "Epoch 20 | Step: 51 | Loss: 0.4258\n",
      "Epoch 20 | Step: 52 | Loss: 0.2738\n",
      "Epoch 20 | Step: 53 | Loss: 0.3637\n",
      "Epoch 20 | Step: 54 | Loss: 0.2923\n",
      "Epoch 20 | Step: 55 | Loss: 0.3631\n",
      "Epoch 20 | Step: 56 | Loss: 0.3213\n",
      "Epoch 20 | Step: 57 | Loss: 0.2449\n",
      "Epoch 20 | Step: 58 | Loss: 0.2110\n",
      "Epoch 20 | Step: 59 | Loss: 1.1295\n",
      "Epoch 20 | Step: 60 | Loss: 0.4806\n",
      "Epoch 20 | Step: 61 | Loss: 0.4124\n",
      "Epoch 20 | Step: 62 | Loss: 0.4042\n",
      "Epoch 20 | Step: 63 | Loss: 0.3162\n",
      "Epoch 20 | Step: 64 | Loss: 0.4632\n",
      "Epoch 20 | Step: 65 | Loss: 0.3369\n",
      "Epoch 20 | Step: 66 | Loss: 0.3020\n",
      "Epoch 20 | Step: 67 | Loss: 0.3048\n",
      "Epoch 20 | Step: 68 | Loss: 0.3509\n",
      "Epoch 20 | Step: 69 | Loss: 0.4410\n",
      "Epoch 20 | Step: 70 | Loss: 0.4274\n",
      "Epoch 20 | Step: 71 | Loss: 0.2717\n",
      "Epoch 20 | Step: 72 | Loss: 0.4883\n",
      "Epoch 20 | Step: 73 | Loss: 0.4171\n",
      "Epoch 20 | Step: 74 | Loss: 0.4546\n",
      "Epoch 20 | Step: 75 | Loss: 0.2585\n",
      "Epoch 20 | Step: 76 | Loss: 0.4635\n",
      "Epoch 20 | Step: 77 | Loss: 0.4146\n",
      "Epoch 20 | Step: 78 | Loss: 0.3313\n",
      "Epoch 20 | Step: 79 | Loss: 0.3269\n",
      "Epoch 20 | Step: 80 | Loss: 0.3236\n",
      "Epoch 20 | Step: 81 | Loss: 0.3047\n",
      "Epoch 20 | Step: 82 | Loss: 0.4400\n",
      "Epoch 20 | Step: 83 | Loss: 0.2116\n",
      "Epoch 20 | Step: 84 | Loss: 0.2344\n",
      "Epoch 20 | Step: 85 | Loss: 0.4009\n",
      "Epoch 20 | Step: 86 | Loss: 0.3532\n",
      "Epoch 20 | Step: 87 | Loss: 0.2783\n",
      "Epoch 20 | Step: 88 | Loss: 0.2633\n",
      "Epoch 20 | Step: 89 | Loss: 0.6416\n",
      "Epoch 20 | Step: 90 | Loss: 0.3352\n",
      "Epoch 20 | Step: 91 | Loss: 0.4324\n",
      "Epoch 20 | Step: 92 | Loss: 0.3829\n",
      "Epoch 20 | Step: 93 | Loss: 0.5716\n",
      "Epoch 20 | Step: 94 | Loss: 0.2339\n",
      "Epoch 20 | Step: 95 | Loss: 0.2602\n",
      "Epoch 20 | Step: 96 | Loss: 0.4698\n",
      "Epoch 20 | Step: 97 | Loss: 0.3740\n",
      "Epoch 20 | Step: 98 | Loss: 0.3070\n",
      "Epoch 20 | Step: 99 | Loss: 0.2102\n",
      "Epoch 20 | Step: 100 | Loss: 0.3529\n",
      "Epoch 20 | Step: 101 | Loss: 0.2480\n",
      "Epoch 20 | Step: 102 | Loss: 0.3091\n",
      "Epoch 20 | Step: 103 | Loss: 0.1935\n",
      "Epoch 20 | Step: 104 | Loss: 0.3245\n",
      "Epoch 20 | Step: 105 | Loss: 0.3958\n",
      "Epoch 20 | Step: 106 | Loss: 0.1808\n",
      "Epoch 20 | Step: 107 | Loss: 0.3596\n",
      "Epoch 20 | Step: 108 | Loss: 0.4008\n",
      "Epoch 20 | Step: 109 | Loss: 0.4595\n",
      "Epoch 20 | Step: 110 | Loss: 0.5425\n",
      "Epoch 20 | Step: 111 | Loss: 0.3791\n",
      "Epoch 20 | Step: 112 | Loss: 0.3795\n",
      "Epoch 20 | Step: 113 | Loss: 0.4470\n",
      "Epoch 20 | Step: 114 | Loss: 0.3893\n",
      "Epoch 20 | Step: 115 | Loss: 0.5621\n",
      "Epoch 20 | Step: 116 | Loss: 0.2474\n",
      "Epoch 20 | Step: 117 | Loss: 0.2703\n",
      "Epoch 20 | Step: 118 | Loss: 0.2425\n",
      "Epoch 20 | Step: 119 | Loss: 0.3196\n",
      "Epoch 20 | Step: 120 | Loss: 0.4401\n",
      "Epoch 20 | Step: 121 | Loss: 0.3594\n",
      "Epoch 20 | Step: 122 | Loss: 0.3071\n",
      "Epoch 20 | Step: 123 | Loss: 0.2685\n",
      "Epoch 20 | Step: 124 | Loss: 0.3148\n",
      "Epoch 20 | Step: 125 | Loss: 0.1752\n",
      "Epoch 20 | Step: 126 | Loss: 0.3348\n",
      "Epoch 20 | Step: 127 | Loss: 0.1438\n",
      "Epoch 20 | Step: 128 | Loss: 0.4318\n",
      "Epoch 21 | Step: 1 | Loss: 0.3237\n",
      "Epoch 21 | Step: 2 | Loss: 0.3765\n",
      "Epoch 21 | Step: 3 | Loss: 0.3475\n",
      "Epoch 21 | Step: 4 | Loss: 0.2829\n",
      "Epoch 21 | Step: 5 | Loss: 0.4023\n",
      "Epoch 21 | Step: 6 | Loss: 0.4945\n",
      "Epoch 21 | Step: 7 | Loss: 0.3453\n",
      "Epoch 21 | Step: 8 | Loss: 1.0564\n",
      "Epoch 21 | Step: 9 | Loss: 0.4870\n",
      "Epoch 21 | Step: 10 | Loss: 0.2614\n",
      "Epoch 21 | Step: 11 | Loss: 0.2370\n",
      "Epoch 21 | Step: 12 | Loss: 0.2474\n",
      "Epoch 21 | Step: 13 | Loss: 0.2753\n",
      "Epoch 21 | Step: 14 | Loss: 0.2709\n",
      "Epoch 21 | Step: 15 | Loss: 0.2707\n",
      "Epoch 21 | Step: 16 | Loss: 0.9260\n",
      "Epoch 21 | Step: 17 | Loss: 0.5141\n",
      "Epoch 21 | Step: 18 | Loss: 0.5292\n",
      "Epoch 21 | Step: 19 | Loss: 0.5005\n",
      "Epoch 21 | Step: 20 | Loss: 0.3962\n",
      "Epoch 21 | Step: 21 | Loss: 0.4562\n",
      "Epoch 21 | Step: 22 | Loss: 0.5893\n",
      "Epoch 21 | Step: 23 | Loss: 0.2333\n",
      "Epoch 21 | Step: 24 | Loss: 0.5047\n",
      "Epoch 21 | Step: 25 | Loss: 0.3333\n",
      "Epoch 21 | Step: 26 | Loss: 0.3230\n",
      "Epoch 21 | Step: 27 | Loss: 0.3226\n",
      "Epoch 21 | Step: 28 | Loss: 0.3018\n",
      "Epoch 21 | Step: 29 | Loss: 0.2206\n",
      "Epoch 21 | Step: 30 | Loss: 0.2910\n",
      "Epoch 21 | Step: 31 | Loss: 0.2136\n",
      "Epoch 21 | Step: 32 | Loss: 0.3197\n",
      "Epoch 21 | Step: 33 | Loss: 0.3075\n",
      "Epoch 21 | Step: 34 | Loss: 0.2848\n",
      "Epoch 21 | Step: 35 | Loss: 0.2029\n",
      "Epoch 21 | Step: 36 | Loss: 0.2738\n",
      "Epoch 21 | Step: 37 | Loss: 0.2193\n",
      "Epoch 21 | Step: 38 | Loss: 0.2849\n",
      "Epoch 21 | Step: 39 | Loss: 0.3128\n",
      "Epoch 21 | Step: 40 | Loss: 0.2184\n",
      "Epoch 21 | Step: 41 | Loss: 0.3220\n",
      "Epoch 21 | Step: 42 | Loss: 0.2833\n",
      "Epoch 21 | Step: 43 | Loss: 0.3252\n",
      "Epoch 21 | Step: 44 | Loss: 0.5780\n",
      "Epoch 21 | Step: 45 | Loss: 0.2147\n",
      "Epoch 21 | Step: 46 | Loss: 0.3028\n",
      "Epoch 21 | Step: 47 | Loss: 0.2749\n",
      "Epoch 21 | Step: 48 | Loss: 0.5496\n",
      "Epoch 21 | Step: 49 | Loss: 0.4258\n",
      "Epoch 21 | Step: 50 | Loss: 0.3976\n",
      "Epoch 21 | Step: 51 | Loss: 0.3814\n",
      "Epoch 21 | Step: 52 | Loss: 0.4125\n",
      "Epoch 21 | Step: 53 | Loss: 0.4516\n",
      "Epoch 21 | Step: 54 | Loss: 0.4241\n",
      "Epoch 21 | Step: 55 | Loss: 0.2185\n",
      "Epoch 21 | Step: 56 | Loss: 0.2901\n",
      "Epoch 21 | Step: 57 | Loss: 0.2904\n",
      "Epoch 21 | Step: 58 | Loss: 0.3840\n",
      "Epoch 21 | Step: 59 | Loss: 0.3677\n",
      "Epoch 21 | Step: 60 | Loss: 0.4195\n",
      "Epoch 21 | Step: 61 | Loss: 0.2585\n",
      "Epoch 21 | Step: 62 | Loss: 0.4014\n",
      "Epoch 21 | Step: 63 | Loss: 0.4268\n",
      "Epoch 21 | Step: 64 | Loss: 0.2966\n",
      "Epoch 21 | Step: 65 | Loss: 0.1245\n",
      "Epoch 21 | Step: 66 | Loss: 0.4052\n",
      "Epoch 21 | Step: 67 | Loss: 0.3057\n",
      "Epoch 21 | Step: 68 | Loss: 0.1673\n",
      "Epoch 21 | Step: 69 | Loss: 0.2819\n",
      "Epoch 21 | Step: 70 | Loss: 0.4227\n",
      "Epoch 21 | Step: 71 | Loss: 0.2976\n",
      "Epoch 21 | Step: 72 | Loss: 0.4523\n",
      "Epoch 21 | Step: 73 | Loss: 0.2734\n",
      "Epoch 21 | Step: 74 | Loss: 0.2708\n",
      "Epoch 21 | Step: 75 | Loss: 0.2967\n",
      "Epoch 21 | Step: 76 | Loss: 0.3859\n",
      "Epoch 21 | Step: 77 | Loss: 0.2993\n",
      "Epoch 21 | Step: 78 | Loss: 0.2833\n",
      "Epoch 21 | Step: 79 | Loss: 0.3626\n",
      "Epoch 21 | Step: 80 | Loss: 0.5535\n",
      "Epoch 21 | Step: 81 | Loss: 0.2486\n",
      "Epoch 21 | Step: 82 | Loss: 0.4563\n",
      "Epoch 21 | Step: 83 | Loss: 0.2892\n",
      "Epoch 21 | Step: 84 | Loss: 0.3783\n",
      "Epoch 21 | Step: 85 | Loss: 0.3714\n",
      "Epoch 21 | Step: 86 | Loss: 0.4727\n",
      "Epoch 21 | Step: 87 | Loss: 0.3940\n",
      "Epoch 21 | Step: 88 | Loss: 0.3850\n",
      "Epoch 21 | Step: 89 | Loss: 0.3004\n",
      "Epoch 21 | Step: 90 | Loss: 0.2829\n",
      "Epoch 21 | Step: 91 | Loss: 0.3957\n",
      "Epoch 21 | Step: 92 | Loss: 0.3343\n",
      "Epoch 21 | Step: 93 | Loss: 0.1625\n",
      "Epoch 21 | Step: 94 | Loss: 0.3892\n",
      "Epoch 21 | Step: 95 | Loss: 0.3930\n",
      "Epoch 21 | Step: 96 | Loss: 0.4083\n",
      "Epoch 21 | Step: 97 | Loss: 0.3460\n",
      "Epoch 21 | Step: 98 | Loss: 0.5099\n",
      "Epoch 21 | Step: 99 | Loss: 0.4331\n",
      "Epoch 21 | Step: 100 | Loss: 0.3688\n",
      "Epoch 21 | Step: 101 | Loss: 0.2269\n",
      "Epoch 21 | Step: 102 | Loss: 0.2935\n",
      "Epoch 21 | Step: 103 | Loss: 0.2540\n",
      "Epoch 21 | Step: 104 | Loss: 0.5141\n",
      "Epoch 21 | Step: 105 | Loss: 0.3044\n",
      "Epoch 21 | Step: 106 | Loss: 0.2495\n",
      "Epoch 21 | Step: 107 | Loss: 0.2195\n",
      "Epoch 21 | Step: 108 | Loss: 0.3303\n",
      "Epoch 21 | Step: 109 | Loss: 0.4134\n",
      "Epoch 21 | Step: 110 | Loss: 0.1933\n",
      "Epoch 21 | Step: 111 | Loss: 0.3741\n",
      "Epoch 21 | Step: 112 | Loss: 0.3826\n",
      "Epoch 21 | Step: 113 | Loss: 0.2326\n",
      "Epoch 21 | Step: 114 | Loss: 0.3378\n",
      "Epoch 21 | Step: 115 | Loss: 0.3849\n",
      "Epoch 21 | Step: 116 | Loss: 0.3487\n",
      "Epoch 21 | Step: 117 | Loss: 0.4998\n",
      "Epoch 21 | Step: 118 | Loss: 0.2263\n",
      "Epoch 21 | Step: 119 | Loss: 0.4263\n",
      "Epoch 21 | Step: 120 | Loss: 0.1847\n",
      "Epoch 21 | Step: 121 | Loss: 0.2918\n",
      "Epoch 21 | Step: 122 | Loss: 0.2724\n",
      "Epoch 21 | Step: 123 | Loss: 0.3363\n",
      "Epoch 21 | Step: 124 | Loss: 0.3623\n",
      "Epoch 21 | Step: 125 | Loss: 0.3600\n",
      "Epoch 21 | Step: 126 | Loss: 0.2838\n",
      "Epoch 21 | Step: 127 | Loss: 0.3333\n",
      "Epoch 21 | Step: 128 | Loss: 0.3920\n",
      "Epoch 22 | Step: 1 | Loss: 0.2548\n",
      "Epoch 22 | Step: 2 | Loss: 0.3265\n",
      "Epoch 22 | Step: 3 | Loss: 0.2953\n",
      "Epoch 22 | Step: 4 | Loss: 0.3446\n",
      "Epoch 22 | Step: 5 | Loss: 0.2712\n",
      "Epoch 22 | Step: 6 | Loss: 0.4545\n",
      "Epoch 22 | Step: 7 | Loss: 0.4182\n",
      "Epoch 22 | Step: 8 | Loss: 0.2874\n",
      "Epoch 22 | Step: 9 | Loss: 0.5563\n",
      "Epoch 22 | Step: 10 | Loss: 0.3418\n",
      "Epoch 22 | Step: 11 | Loss: 0.3817\n",
      "Epoch 22 | Step: 12 | Loss: 0.2412\n",
      "Epoch 22 | Step: 13 | Loss: 0.2318\n",
      "Epoch 22 | Step: 14 | Loss: 0.3709\n",
      "Epoch 22 | Step: 15 | Loss: 0.2603\n",
      "Epoch 22 | Step: 16 | Loss: 0.3992\n",
      "Epoch 22 | Step: 17 | Loss: 0.2555\n",
      "Epoch 22 | Step: 18 | Loss: 0.3586\n",
      "Epoch 22 | Step: 19 | Loss: 0.2917\n",
      "Epoch 22 | Step: 20 | Loss: 0.2345\n",
      "Epoch 22 | Step: 21 | Loss: 0.3439\n",
      "Epoch 22 | Step: 22 | Loss: 0.3672\n",
      "Epoch 22 | Step: 23 | Loss: 0.2290\n",
      "Epoch 22 | Step: 24 | Loss: 0.4034\n",
      "Epoch 22 | Step: 25 | Loss: 0.3419\n",
      "Epoch 22 | Step: 26 | Loss: 0.2571\n",
      "Epoch 22 | Step: 27 | Loss: 0.3121\n",
      "Epoch 22 | Step: 28 | Loss: 0.3780\n",
      "Epoch 22 | Step: 29 | Loss: 0.3748\n",
      "Epoch 22 | Step: 30 | Loss: 0.2843\n",
      "Epoch 22 | Step: 31 | Loss: 0.4560\n",
      "Epoch 22 | Step: 32 | Loss: 0.3470\n",
      "Epoch 22 | Step: 33 | Loss: 0.2393\n",
      "Epoch 22 | Step: 34 | Loss: 0.3760\n",
      "Epoch 22 | Step: 35 | Loss: 0.4057\n",
      "Epoch 22 | Step: 36 | Loss: 0.1910\n",
      "Epoch 22 | Step: 37 | Loss: 0.3543\n",
      "Epoch 22 | Step: 38 | Loss: 0.2994\n",
      "Epoch 22 | Step: 39 | Loss: 0.2930\n",
      "Epoch 22 | Step: 40 | Loss: 0.4616\n",
      "Epoch 22 | Step: 41 | Loss: 0.3655\n",
      "Epoch 22 | Step: 42 | Loss: 0.2888\n",
      "Epoch 22 | Step: 43 | Loss: 0.4049\n",
      "Epoch 22 | Step: 44 | Loss: 0.3541\n",
      "Epoch 22 | Step: 45 | Loss: 0.2968\n",
      "Epoch 22 | Step: 46 | Loss: 0.4465\n",
      "Epoch 22 | Step: 47 | Loss: 0.4352\n",
      "Epoch 22 | Step: 48 | Loss: 0.4715\n",
      "Epoch 22 | Step: 49 | Loss: 0.3882\n",
      "Epoch 22 | Step: 50 | Loss: 0.3251\n",
      "Epoch 22 | Step: 51 | Loss: 0.3034\n",
      "Epoch 22 | Step: 52 | Loss: 0.2721\n",
      "Epoch 22 | Step: 53 | Loss: 0.2164\n",
      "Epoch 22 | Step: 54 | Loss: 0.2752\n",
      "Epoch 22 | Step: 55 | Loss: 0.4818\n",
      "Epoch 22 | Step: 56 | Loss: 0.2502\n",
      "Epoch 22 | Step: 57 | Loss: 0.1831\n",
      "Epoch 22 | Step: 58 | Loss: 0.2222\n",
      "Epoch 22 | Step: 59 | Loss: 0.4015\n",
      "Epoch 22 | Step: 60 | Loss: 0.4709\n",
      "Epoch 22 | Step: 61 | Loss: 0.3036\n",
      "Epoch 22 | Step: 62 | Loss: 0.2995\n",
      "Epoch 22 | Step: 63 | Loss: 0.3714\n",
      "Epoch 22 | Step: 64 | Loss: 0.3007\n",
      "Epoch 22 | Step: 65 | Loss: 0.3746\n",
      "Epoch 22 | Step: 66 | Loss: 0.2886\n",
      "Epoch 22 | Step: 67 | Loss: 0.3326\n",
      "Epoch 22 | Step: 68 | Loss: 0.2397\n",
      "Epoch 22 | Step: 69 | Loss: 0.3622\n",
      "Epoch 22 | Step: 70 | Loss: 0.4638\n",
      "Epoch 22 | Step: 71 | Loss: 0.3789\n",
      "Epoch 22 | Step: 72 | Loss: 0.2772\n",
      "Epoch 22 | Step: 73 | Loss: 0.3289\n",
      "Epoch 22 | Step: 74 | Loss: 0.2965\n",
      "Epoch 22 | Step: 75 | Loss: 0.7660\n",
      "Epoch 22 | Step: 76 | Loss: 0.2166\n",
      "Epoch 22 | Step: 77 | Loss: 0.2854\n",
      "Epoch 22 | Step: 78 | Loss: 0.3582\n",
      "Epoch 22 | Step: 79 | Loss: 0.2886\n",
      "Epoch 22 | Step: 80 | Loss: 0.2883\n",
      "Epoch 22 | Step: 81 | Loss: 0.3230\n",
      "Epoch 22 | Step: 82 | Loss: 0.3777\n",
      "Epoch 22 | Step: 83 | Loss: 0.2944\n",
      "Epoch 22 | Step: 84 | Loss: 0.5138\n",
      "Epoch 22 | Step: 85 | Loss: 0.4437\n",
      "Epoch 22 | Step: 86 | Loss: 0.2638\n",
      "Epoch 22 | Step: 87 | Loss: 0.3394\n",
      "Epoch 22 | Step: 88 | Loss: 0.2564\n",
      "Epoch 22 | Step: 89 | Loss: 0.5427\n",
      "Epoch 22 | Step: 90 | Loss: 0.3773\n",
      "Epoch 22 | Step: 91 | Loss: 0.3837\n",
      "Epoch 22 | Step: 92 | Loss: 0.3562\n",
      "Epoch 22 | Step: 93 | Loss: 0.2047\n",
      "Epoch 22 | Step: 94 | Loss: 0.2510\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 22 | Step: 95 | Loss: 0.2285\n",
      "Epoch 22 | Step: 96 | Loss: 0.2844\n",
      "Epoch 22 | Step: 97 | Loss: 0.9673\n",
      "Epoch 22 | Step: 98 | Loss: 0.2625\n",
      "Epoch 22 | Step: 99 | Loss: 0.2992\n",
      "Epoch 22 | Step: 100 | Loss: 0.2436\n",
      "Epoch 22 | Step: 101 | Loss: 0.2070\n",
      "Epoch 22 | Step: 102 | Loss: 0.2195\n",
      "Epoch 22 | Step: 103 | Loss: 0.3799\n",
      "Epoch 22 | Step: 104 | Loss: 0.3394\n",
      "Epoch 22 | Step: 105 | Loss: 0.3877\n",
      "Epoch 22 | Step: 106 | Loss: 0.4475\n",
      "Epoch 22 | Step: 107 | Loss: 0.3320\n",
      "Epoch 22 | Step: 108 | Loss: 0.4559\n",
      "Epoch 22 | Step: 109 | Loss: 0.2523\n",
      "Epoch 22 | Step: 110 | Loss: 0.2953\n",
      "Epoch 22 | Step: 111 | Loss: 0.2616\n",
      "Epoch 22 | Step: 112 | Loss: 0.3503\n",
      "Epoch 22 | Step: 113 | Loss: 0.2194\n",
      "Epoch 22 | Step: 114 | Loss: 0.4401\n",
      "Epoch 22 | Step: 115 | Loss: 0.3202\n",
      "Epoch 22 | Step: 116 | Loss: 0.3643\n",
      "Epoch 22 | Step: 117 | Loss: 0.4544\n",
      "Epoch 22 | Step: 118 | Loss: 0.2525\n",
      "Epoch 22 | Step: 119 | Loss: 0.1723\n",
      "Epoch 22 | Step: 120 | Loss: 0.3111\n",
      "Epoch 22 | Step: 121 | Loss: 0.3770\n",
      "Epoch 22 | Step: 122 | Loss: 0.2993\n",
      "Epoch 22 | Step: 123 | Loss: 0.5271\n",
      "Epoch 22 | Step: 124 | Loss: 0.4442\n",
      "Epoch 22 | Step: 125 | Loss: 0.3486\n",
      "Epoch 22 | Step: 126 | Loss: 0.3588\n",
      "Epoch 22 | Step: 127 | Loss: 0.3119\n",
      "Epoch 22 | Step: 128 | Loss: 0.3407\n",
      "Epoch 23 | Step: 1 | Loss: 0.3604\n",
      "Epoch 23 | Step: 2 | Loss: 0.3844\n",
      "Epoch 23 | Step: 3 | Loss: 0.2416\n",
      "Epoch 23 | Step: 4 | Loss: 0.2386\n",
      "Epoch 23 | Step: 5 | Loss: 0.3046\n",
      "Epoch 23 | Step: 6 | Loss: 0.2790\n",
      "Epoch 23 | Step: 7 | Loss: 0.2761\n",
      "Epoch 23 | Step: 8 | Loss: 0.3079\n",
      "Epoch 23 | Step: 9 | Loss: 0.3505\n",
      "Epoch 23 | Step: 10 | Loss: 0.3241\n",
      "Epoch 23 | Step: 11 | Loss: 0.3454\n",
      "Epoch 23 | Step: 12 | Loss: 0.3459\n",
      "Epoch 23 | Step: 13 | Loss: 0.3287\n",
      "Epoch 23 | Step: 14 | Loss: 0.2909\n",
      "Epoch 23 | Step: 15 | Loss: 0.3439\n",
      "Epoch 23 | Step: 16 | Loss: 0.2794\n",
      "Epoch 23 | Step: 17 | Loss: 0.3413\n",
      "Epoch 23 | Step: 18 | Loss: 0.2330\n",
      "Epoch 23 | Step: 19 | Loss: 0.2346\n",
      "Epoch 23 | Step: 20 | Loss: 0.2326\n",
      "Epoch 23 | Step: 21 | Loss: 0.2694\n",
      "Epoch 23 | Step: 22 | Loss: 0.3826\n",
      "Epoch 23 | Step: 23 | Loss: 0.3726\n",
      "Epoch 23 | Step: 24 | Loss: 0.2372\n",
      "Epoch 23 | Step: 25 | Loss: 0.3061\n",
      "Epoch 23 | Step: 26 | Loss: 0.3055\n",
      "Epoch 23 | Step: 27 | Loss: 0.2616\n",
      "Epoch 23 | Step: 28 | Loss: 0.4257\n",
      "Epoch 23 | Step: 29 | Loss: 0.2941\n",
      "Epoch 23 | Step: 30 | Loss: 0.2770\n",
      "Epoch 23 | Step: 31 | Loss: 0.3982\n",
      "Epoch 23 | Step: 32 | Loss: 0.3620\n",
      "Epoch 23 | Step: 33 | Loss: 0.2414\n",
      "Epoch 23 | Step: 34 | Loss: 0.3403\n",
      "Epoch 23 | Step: 35 | Loss: 0.3637\n",
      "Epoch 23 | Step: 36 | Loss: 0.7209\n",
      "Epoch 23 | Step: 37 | Loss: 0.2287\n",
      "Epoch 23 | Step: 38 | Loss: 0.2526\n",
      "Epoch 23 | Step: 39 | Loss: 0.3793\n",
      "Epoch 23 | Step: 40 | Loss: 0.4962\n",
      "Epoch 23 | Step: 41 | Loss: 0.3079\n",
      "Epoch 23 | Step: 42 | Loss: 0.3487\n",
      "Epoch 23 | Step: 43 | Loss: 0.5408\n",
      "Epoch 23 | Step: 44 | Loss: 0.3426\n",
      "Epoch 23 | Step: 45 | Loss: 0.2173\n",
      "Epoch 23 | Step: 46 | Loss: 0.2892\n",
      "Epoch 23 | Step: 47 | Loss: 0.3802\n",
      "Epoch 23 | Step: 48 | Loss: 0.3555\n",
      "Epoch 23 | Step: 49 | Loss: 0.3361\n",
      "Epoch 23 | Step: 50 | Loss: 0.3839\n",
      "Epoch 23 | Step: 51 | Loss: 0.3845\n",
      "Epoch 23 | Step: 52 | Loss: 0.4943\n",
      "Epoch 23 | Step: 53 | Loss: 0.2570\n",
      "Epoch 23 | Step: 54 | Loss: 0.3930\n",
      "Epoch 23 | Step: 55 | Loss: 0.3042\n",
      "Epoch 23 | Step: 56 | Loss: 0.2589\n",
      "Epoch 23 | Step: 57 | Loss: 0.4034\n",
      "Epoch 23 | Step: 58 | Loss: 0.3138\n",
      "Epoch 23 | Step: 59 | Loss: 0.3975\n",
      "Epoch 23 | Step: 60 | Loss: 0.3395\n",
      "Epoch 23 | Step: 61 | Loss: 0.9314\n",
      "Epoch 23 | Step: 62 | Loss: 0.2163\n",
      "Epoch 23 | Step: 63 | Loss: 0.4782\n",
      "Epoch 23 | Step: 64 | Loss: 0.2624\n",
      "Epoch 23 | Step: 65 | Loss: 0.6444\n",
      "Epoch 23 | Step: 66 | Loss: 0.2791\n",
      "Epoch 23 | Step: 67 | Loss: 0.2237\n",
      "Epoch 23 | Step: 68 | Loss: 0.1950\n",
      "Epoch 23 | Step: 69 | Loss: 0.3560\n",
      "Epoch 23 | Step: 70 | Loss: 0.2624\n",
      "Epoch 23 | Step: 71 | Loss: 0.3111\n",
      "Epoch 23 | Step: 72 | Loss: 0.3112\n",
      "Epoch 23 | Step: 73 | Loss: 0.3429\n",
      "Epoch 23 | Step: 74 | Loss: 0.3464\n",
      "Epoch 23 | Step: 75 | Loss: 0.3965\n",
      "Epoch 23 | Step: 76 | Loss: 0.2926\n",
      "Epoch 23 | Step: 77 | Loss: 0.2360\n",
      "Epoch 23 | Step: 78 | Loss: 0.2860\n",
      "Epoch 23 | Step: 79 | Loss: 0.3948\n",
      "Epoch 23 | Step: 80 | Loss: 0.2506\n",
      "Epoch 23 | Step: 81 | Loss: 0.2141\n",
      "Epoch 23 | Step: 82 | Loss: 0.2737\n",
      "Epoch 23 | Step: 83 | Loss: 0.2269\n",
      "Epoch 23 | Step: 84 | Loss: 0.3633\n",
      "Epoch 23 | Step: 85 | Loss: 0.3401\n",
      "Epoch 23 | Step: 86 | Loss: 0.4294\n",
      "Epoch 23 | Step: 87 | Loss: 0.4020\n",
      "Epoch 23 | Step: 88 | Loss: 0.2604\n",
      "Epoch 23 | Step: 89 | Loss: 0.4238\n",
      "Epoch 23 | Step: 90 | Loss: 0.2775\n",
      "Epoch 23 | Step: 91 | Loss: 0.3532\n",
      "Epoch 23 | Step: 92 | Loss: 0.4153\n",
      "Epoch 23 | Step: 93 | Loss: 0.3259\n",
      "Epoch 23 | Step: 94 | Loss: 0.2921\n",
      "Epoch 23 | Step: 95 | Loss: 0.4582\n",
      "Epoch 23 | Step: 96 | Loss: 0.4021\n",
      "Epoch 23 | Step: 97 | Loss: 0.3014\n",
      "Epoch 23 | Step: 98 | Loss: 0.2795\n",
      "Epoch 23 | Step: 99 | Loss: 0.3052\n",
      "Epoch 23 | Step: 100 | Loss: 0.3286\n",
      "Epoch 23 | Step: 101 | Loss: 0.3893\n",
      "Epoch 23 | Step: 102 | Loss: 0.2922\n",
      "Epoch 23 | Step: 103 | Loss: 0.3193\n",
      "Epoch 23 | Step: 104 | Loss: 0.3449\n",
      "Epoch 23 | Step: 105 | Loss: 0.3211\n",
      "Epoch 23 | Step: 106 | Loss: 0.3374\n",
      "Epoch 23 | Step: 107 | Loss: 0.3220\n",
      "Epoch 23 | Step: 108 | Loss: 0.3979\n",
      "Epoch 23 | Step: 109 | Loss: 0.2111\n",
      "Epoch 23 | Step: 110 | Loss: 0.3989\n",
      "Epoch 23 | Step: 111 | Loss: 0.3048\n",
      "Epoch 23 | Step: 112 | Loss: 0.2629\n",
      "Epoch 23 | Step: 113 | Loss: 0.2511\n",
      "Epoch 23 | Step: 114 | Loss: 0.3438\n",
      "Epoch 23 | Step: 115 | Loss: 0.4868\n",
      "Epoch 23 | Step: 116 | Loss: 0.2287\n",
      "Epoch 23 | Step: 117 | Loss: 0.2783\n",
      "Epoch 23 | Step: 118 | Loss: 0.2393\n",
      "Epoch 23 | Step: 119 | Loss: 0.3497\n",
      "Epoch 23 | Step: 120 | Loss: 0.3462\n",
      "Epoch 23 | Step: 121 | Loss: 0.3858\n",
      "Epoch 23 | Step: 122 | Loss: 0.1963\n",
      "Epoch 23 | Step: 123 | Loss: 0.4347\n",
      "Epoch 23 | Step: 124 | Loss: 0.2701\n",
      "Epoch 23 | Step: 125 | Loss: 0.3518\n",
      "Epoch 23 | Step: 126 | Loss: 0.2546\n",
      "Epoch 23 | Step: 127 | Loss: 0.2023\n",
      "Epoch 23 | Step: 128 | Loss: 0.3698\n",
      "Epoch 24 | Step: 1 | Loss: 0.1754\n",
      "Epoch 24 | Step: 2 | Loss: 0.3454\n",
      "Epoch 24 | Step: 3 | Loss: 0.3599\n",
      "Epoch 24 | Step: 4 | Loss: 0.3836\n",
      "Epoch 24 | Step: 5 | Loss: 0.2114\n",
      "Epoch 24 | Step: 6 | Loss: 0.3716\n",
      "Epoch 24 | Step: 7 | Loss: 0.2825\n",
      "Epoch 24 | Step: 8 | Loss: 0.2568\n",
      "Epoch 24 | Step: 9 | Loss: 0.3345\n",
      "Epoch 24 | Step: 10 | Loss: 0.5861\n",
      "Epoch 24 | Step: 11 | Loss: 0.4061\n",
      "Epoch 24 | Step: 12 | Loss: 0.1909\n",
      "Epoch 24 | Step: 13 | Loss: 0.2732\n",
      "Epoch 24 | Step: 14 | Loss: 0.4978\n",
      "Epoch 24 | Step: 15 | Loss: 0.3058\n",
      "Epoch 24 | Step: 16 | Loss: 0.2012\n",
      "Epoch 24 | Step: 17 | Loss: 0.3052\n",
      "Epoch 24 | Step: 18 | Loss: 0.9671\n",
      "Epoch 24 | Step: 19 | Loss: 0.3174\n",
      "Epoch 24 | Step: 20 | Loss: 0.1639\n",
      "Epoch 24 | Step: 21 | Loss: 0.2947\n",
      "Epoch 24 | Step: 22 | Loss: 0.4790\n",
      "Epoch 24 | Step: 23 | Loss: 0.3795\n",
      "Epoch 24 | Step: 24 | Loss: 0.2185\n",
      "Epoch 24 | Step: 25 | Loss: 0.2035\n",
      "Epoch 24 | Step: 26 | Loss: 0.3055\n",
      "Epoch 24 | Step: 27 | Loss: 0.4651\n",
      "Epoch 24 | Step: 28 | Loss: 0.2889\n",
      "Epoch 24 | Step: 29 | Loss: 0.3063\n",
      "Epoch 24 | Step: 30 | Loss: 0.5067\n",
      "Epoch 24 | Step: 31 | Loss: 0.2543\n",
      "Epoch 24 | Step: 32 | Loss: 0.3604\n",
      "Epoch 24 | Step: 33 | Loss: 0.3362\n",
      "Epoch 24 | Step: 34 | Loss: 0.3888\n",
      "Epoch 24 | Step: 35 | Loss: 0.5062\n",
      "Epoch 24 | Step: 36 | Loss: 0.3304\n",
      "Epoch 24 | Step: 37 | Loss: 0.2056\n",
      "Epoch 24 | Step: 38 | Loss: 0.2667\n",
      "Epoch 24 | Step: 39 | Loss: 0.2959\n",
      "Epoch 24 | Step: 40 | Loss: 0.3511\n",
      "Epoch 24 | Step: 41 | Loss: 0.3354\n",
      "Epoch 24 | Step: 42 | Loss: 0.4120\n",
      "Epoch 24 | Step: 43 | Loss: 0.2929\n",
      "Epoch 24 | Step: 44 | Loss: 0.3220\n",
      "Epoch 24 | Step: 45 | Loss: 0.3443\n",
      "Epoch 24 | Step: 46 | Loss: 0.2469\n",
      "Epoch 24 | Step: 47 | Loss: 0.3098\n",
      "Epoch 24 | Step: 48 | Loss: 0.3177\n",
      "Epoch 24 | Step: 49 | Loss: 0.2512\n",
      "Epoch 24 | Step: 50 | Loss: 0.2199\n",
      "Epoch 24 | Step: 51 | Loss: 0.2911\n",
      "Epoch 24 | Step: 52 | Loss: 0.4941\n",
      "Epoch 24 | Step: 53 | Loss: 0.3608\n",
      "Epoch 24 | Step: 54 | Loss: 0.2694\n",
      "Epoch 24 | Step: 55 | Loss: 0.3613\n",
      "Epoch 24 | Step: 56 | Loss: 0.4712\n",
      "Epoch 24 | Step: 57 | Loss: 0.2867\n",
      "Epoch 24 | Step: 58 | Loss: 0.3778\n",
      "Epoch 24 | Step: 59 | Loss: 0.3601\n",
      "Epoch 24 | Step: 60 | Loss: 0.4157\n",
      "Epoch 24 | Step: 61 | Loss: 0.2268\n",
      "Epoch 24 | Step: 62 | Loss: 0.4126\n",
      "Epoch 24 | Step: 63 | Loss: 0.3052\n",
      "Epoch 24 | Step: 64 | Loss: 0.3765\n",
      "Epoch 24 | Step: 65 | Loss: 0.3152\n",
      "Epoch 24 | Step: 66 | Loss: 0.4650\n",
      "Epoch 24 | Step: 67 | Loss: 0.4222\n",
      "Epoch 24 | Step: 68 | Loss: 0.3340\n",
      "Epoch 24 | Step: 69 | Loss: 0.3691\n",
      "Epoch 24 | Step: 70 | Loss: 0.2634\n",
      "Epoch 24 | Step: 71 | Loss: 0.6171\n",
      "Epoch 24 | Step: 72 | Loss: 0.2993\n",
      "Epoch 24 | Step: 73 | Loss: 0.3360\n",
      "Epoch 24 | Step: 74 | Loss: 0.2801\n",
      "Epoch 24 | Step: 75 | Loss: 0.2712\n",
      "Epoch 24 | Step: 76 | Loss: 0.3160\n",
      "Epoch 24 | Step: 77 | Loss: 0.2703\n",
      "Epoch 24 | Step: 78 | Loss: 0.2410\n",
      "Epoch 24 | Step: 79 | Loss: 0.3609\n",
      "Epoch 24 | Step: 80 | Loss: 0.1904\n",
      "Epoch 24 | Step: 81 | Loss: 0.3751\n",
      "Epoch 24 | Step: 82 | Loss: 0.4212\n",
      "Epoch 24 | Step: 83 | Loss: 0.3859\n",
      "Epoch 24 | Step: 84 | Loss: 0.2168\n",
      "Epoch 24 | Step: 85 | Loss: 0.2370\n",
      "Epoch 24 | Step: 86 | Loss: 0.3065\n",
      "Epoch 24 | Step: 87 | Loss: 0.3111\n",
      "Epoch 24 | Step: 88 | Loss: 0.3185\n",
      "Epoch 24 | Step: 89 | Loss: 0.2721\n",
      "Epoch 24 | Step: 90 | Loss: 0.4645\n",
      "Epoch 24 | Step: 91 | Loss: 0.3458\n",
      "Epoch 24 | Step: 92 | Loss: 0.3285\n",
      "Epoch 24 | Step: 93 | Loss: 0.2774\n",
      "Epoch 24 | Step: 94 | Loss: 0.2492\n",
      "Epoch 24 | Step: 95 | Loss: 0.3050\n",
      "Epoch 24 | Step: 96 | Loss: 0.2424\n",
      "Epoch 24 | Step: 97 | Loss: 0.3113\n",
      "Epoch 24 | Step: 98 | Loss: 0.1889\n",
      "Epoch 24 | Step: 99 | Loss: 0.3526\n",
      "Epoch 24 | Step: 100 | Loss: 0.1891\n",
      "Epoch 24 | Step: 101 | Loss: 0.3189\n",
      "Epoch 24 | Step: 102 | Loss: 0.2917\n",
      "Epoch 24 | Step: 103 | Loss: 0.4949\n",
      "Epoch 24 | Step: 104 | Loss: 0.1426\n",
      "Epoch 24 | Step: 105 | Loss: 0.3045\n",
      "Epoch 24 | Step: 106 | Loss: 0.3182\n",
      "Epoch 24 | Step: 107 | Loss: 0.2723\n",
      "Epoch 24 | Step: 108 | Loss: 0.3234\n",
      "Epoch 24 | Step: 109 | Loss: 0.2510\n",
      "Epoch 24 | Step: 110 | Loss: 0.4865\n",
      "Epoch 24 | Step: 111 | Loss: 0.2108\n",
      "Epoch 24 | Step: 112 | Loss: 0.3697\n",
      "Epoch 24 | Step: 113 | Loss: 0.3030\n",
      "Epoch 24 | Step: 114 | Loss: 0.2853\n",
      "Epoch 24 | Step: 115 | Loss: 0.1482\n",
      "Epoch 24 | Step: 116 | Loss: 0.2927\n",
      "Epoch 24 | Step: 117 | Loss: 0.3277\n",
      "Epoch 24 | Step: 118 | Loss: 0.1826\n",
      "Epoch 24 | Step: 119 | Loss: 0.1879\n",
      "Epoch 24 | Step: 120 | Loss: 0.2690\n",
      "Epoch 24 | Step: 121 | Loss: 0.2805\n",
      "Epoch 24 | Step: 122 | Loss: 0.2834\n",
      "Epoch 24 | Step: 123 | Loss: 0.3978\n",
      "Epoch 24 | Step: 124 | Loss: 0.3348\n",
      "Epoch 24 | Step: 125 | Loss: 0.2366\n",
      "Epoch 24 | Step: 126 | Loss: 0.3029\n",
      "Epoch 24 | Step: 127 | Loss: 0.3995\n",
      "Epoch 24 | Step: 128 | Loss: 0.3133\n",
      "Epoch 25 | Step: 1 | Loss: 0.3106\n",
      "Epoch 25 | Step: 2 | Loss: 0.3624\n",
      "Epoch 25 | Step: 3 | Loss: 0.3030\n",
      "Epoch 25 | Step: 4 | Loss: 0.3040\n",
      "Epoch 25 | Step: 5 | Loss: 0.3062\n",
      "Epoch 25 | Step: 6 | Loss: 0.1993\n",
      "Epoch 25 | Step: 7 | Loss: 0.2409\n",
      "Epoch 25 | Step: 8 | Loss: 0.3453\n",
      "Epoch 25 | Step: 9 | Loss: 0.3330\n",
      "Epoch 25 | Step: 10 | Loss: 0.2850\n",
      "Epoch 25 | Step: 11 | Loss: 0.2380\n",
      "Epoch 25 | Step: 12 | Loss: 0.4182\n",
      "Epoch 25 | Step: 13 | Loss: 0.2583\n",
      "Epoch 25 | Step: 14 | Loss: 0.2505\n",
      "Epoch 25 | Step: 15 | Loss: 0.2335\n",
      "Epoch 25 | Step: 16 | Loss: 0.2666\n",
      "Epoch 25 | Step: 17 | Loss: 0.3204\n",
      "Epoch 25 | Step: 18 | Loss: 0.2762\n",
      "Epoch 25 | Step: 19 | Loss: 0.3324\n",
      "Epoch 25 | Step: 20 | Loss: 0.2766\n",
      "Epoch 25 | Step: 21 | Loss: 0.1816\n",
      "Epoch 25 | Step: 22 | Loss: 0.3917\n",
      "Epoch 25 | Step: 23 | Loss: 0.3278\n",
      "Epoch 25 | Step: 24 | Loss: 0.3245\n",
      "Epoch 25 | Step: 25 | Loss: 0.2639\n",
      "Epoch 25 | Step: 26 | Loss: 0.2309\n",
      "Epoch 25 | Step: 27 | Loss: 0.3934\n",
      "Epoch 25 | Step: 28 | Loss: 0.2673\n",
      "Epoch 25 | Step: 29 | Loss: 0.2275\n",
      "Epoch 25 | Step: 30 | Loss: 0.1267\n",
      "Epoch 25 | Step: 31 | Loss: 0.3352\n",
      "Epoch 25 | Step: 32 | Loss: 0.2760\n",
      "Epoch 25 | Step: 33 | Loss: 0.3247\n",
      "Epoch 25 | Step: 34 | Loss: 0.4709\n",
      "Epoch 25 | Step: 35 | Loss: 0.2282\n",
      "Epoch 25 | Step: 36 | Loss: 0.4569\n",
      "Epoch 25 | Step: 37 | Loss: 0.4988\n",
      "Epoch 25 | Step: 38 | Loss: 0.3914\n",
      "Epoch 25 | Step: 39 | Loss: 0.2729\n",
      "Epoch 25 | Step: 40 | Loss: 0.3098\n",
      "Epoch 25 | Step: 41 | Loss: 0.4872\n",
      "Epoch 25 | Step: 42 | Loss: 0.2195\n",
      "Epoch 25 | Step: 43 | Loss: 0.2810\n",
      "Epoch 25 | Step: 44 | Loss: 0.4003\n",
      "Epoch 25 | Step: 45 | Loss: 0.3079\n",
      "Epoch 25 | Step: 46 | Loss: 0.3087\n",
      "Epoch 25 | Step: 47 | Loss: 0.3609\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 25 | Step: 48 | Loss: 0.3961\n",
      "Epoch 25 | Step: 49 | Loss: 0.2870\n",
      "Epoch 25 | Step: 50 | Loss: 0.3034\n",
      "Epoch 25 | Step: 51 | Loss: 0.1868\n",
      "Epoch 25 | Step: 52 | Loss: 0.4043\n",
      "Epoch 25 | Step: 53 | Loss: 0.2632\n",
      "Epoch 25 | Step: 54 | Loss: 0.2298\n",
      "Epoch 25 | Step: 55 | Loss: 0.3109\n",
      "Epoch 25 | Step: 56 | Loss: 0.2878\n",
      "Epoch 25 | Step: 57 | Loss: 0.2978\n",
      "Epoch 25 | Step: 58 | Loss: 0.2894\n",
      "Epoch 25 | Step: 59 | Loss: 0.2839\n",
      "Epoch 25 | Step: 60 | Loss: 0.3843\n",
      "Epoch 25 | Step: 61 | Loss: 0.2164\n",
      "Epoch 25 | Step: 62 | Loss: 0.4448\n",
      "Epoch 25 | Step: 63 | Loss: 0.3921\n",
      "Epoch 25 | Step: 64 | Loss: 0.2527\n",
      "Epoch 25 | Step: 65 | Loss: 0.2287\n",
      "Epoch 25 | Step: 66 | Loss: 0.2443\n",
      "Epoch 25 | Step: 67 | Loss: 0.2997\n",
      "Epoch 25 | Step: 68 | Loss: 0.1715\n",
      "Epoch 25 | Step: 69 | Loss: 0.5430\n",
      "Epoch 25 | Step: 70 | Loss: 0.2376\n",
      "Epoch 25 | Step: 71 | Loss: 0.3423\n",
      "Epoch 25 | Step: 72 | Loss: 0.3032\n",
      "Epoch 25 | Step: 73 | Loss: 0.2261\n",
      "Epoch 25 | Step: 74 | Loss: 0.2956\n",
      "Epoch 25 | Step: 75 | Loss: 0.2599\n",
      "Epoch 25 | Step: 76 | Loss: 0.3929\n",
      "Epoch 25 | Step: 77 | Loss: 0.4289\n",
      "Epoch 25 | Step: 78 | Loss: 0.3353\n",
      "Epoch 25 | Step: 79 | Loss: 0.3255\n",
      "Epoch 25 | Step: 80 | Loss: 0.3165\n",
      "Epoch 25 | Step: 81 | Loss: 0.7209\n",
      "Epoch 25 | Step: 82 | Loss: 0.2139\n",
      "Epoch 25 | Step: 83 | Loss: 0.2123\n",
      "Epoch 25 | Step: 84 | Loss: 0.2209\n",
      "Epoch 25 | Step: 85 | Loss: 0.3781\n",
      "Epoch 25 | Step: 86 | Loss: 0.2425\n",
      "Epoch 25 | Step: 87 | Loss: 0.2315\n",
      "Epoch 25 | Step: 88 | Loss: 0.2418\n",
      "Epoch 25 | Step: 89 | Loss: 0.4064\n",
      "Epoch 25 | Step: 90 | Loss: 0.3929\n",
      "Epoch 25 | Step: 91 | Loss: 0.4945\n",
      "Epoch 25 | Step: 92 | Loss: 0.4397\n",
      "Epoch 25 | Step: 93 | Loss: 0.2601\n",
      "Epoch 25 | Step: 94 | Loss: 0.4638\n",
      "Epoch 25 | Step: 95 | Loss: 0.3276\n",
      "Epoch 25 | Step: 96 | Loss: 0.3466\n",
      "Epoch 25 | Step: 97 | Loss: 0.2975\n",
      "Epoch 25 | Step: 98 | Loss: 0.1817\n",
      "Epoch 25 | Step: 99 | Loss: 0.2998\n",
      "Epoch 25 | Step: 100 | Loss: 0.2708\n",
      "Epoch 25 | Step: 101 | Loss: 0.4261\n",
      "Epoch 25 | Step: 102 | Loss: 0.1594\n",
      "Epoch 25 | Step: 103 | Loss: 0.1550\n",
      "Epoch 25 | Step: 104 | Loss: 0.2248\n",
      "Epoch 25 | Step: 105 | Loss: 0.3253\n",
      "Epoch 25 | Step: 106 | Loss: 0.4371\n",
      "Epoch 25 | Step: 107 | Loss: 0.1705\n",
      "Epoch 25 | Step: 108 | Loss: 0.2447\n",
      "Epoch 25 | Step: 109 | Loss: 0.2475\n",
      "Epoch 25 | Step: 110 | Loss: 0.3148\n",
      "Epoch 25 | Step: 111 | Loss: 0.3249\n",
      "Epoch 25 | Step: 112 | Loss: 0.5887\n",
      "Epoch 25 | Step: 113 | Loss: 0.4390\n",
      "Epoch 25 | Step: 114 | Loss: 0.3143\n",
      "Epoch 25 | Step: 115 | Loss: 0.2919\n",
      "Epoch 25 | Step: 116 | Loss: 0.2886\n",
      "Epoch 25 | Step: 117 | Loss: 0.3307\n",
      "Epoch 25 | Step: 118 | Loss: 0.2296\n",
      "Epoch 25 | Step: 119 | Loss: 0.3521\n",
      "Epoch 25 | Step: 120 | Loss: 0.3045\n",
      "Epoch 25 | Step: 121 | Loss: 0.2742\n",
      "Epoch 25 | Step: 122 | Loss: 0.3643\n",
      "Epoch 25 | Step: 123 | Loss: 0.3450\n",
      "Epoch 25 | Step: 124 | Loss: 0.3246\n",
      "Epoch 25 | Step: 125 | Loss: 0.2576\n",
      "Epoch 25 | Step: 126 | Loss: 0.3781\n",
      "Epoch 25 | Step: 127 | Loss: 0.7847\n",
      "Epoch 25 | Step: 128 | Loss: 0.3405\n",
      "Epoch 26 | Step: 1 | Loss: 0.3064\n",
      "Epoch 26 | Step: 2 | Loss: 0.2882\n",
      "Epoch 26 | Step: 3 | Loss: 0.3587\n",
      "Epoch 26 | Step: 4 | Loss: 0.3077\n",
      "Epoch 26 | Step: 5 | Loss: 0.4726\n",
      "Epoch 26 | Step: 6 | Loss: 0.2827\n",
      "Epoch 26 | Step: 7 | Loss: 0.4322\n",
      "Epoch 26 | Step: 8 | Loss: 0.2040\n",
      "Epoch 26 | Step: 9 | Loss: 0.3440\n",
      "Epoch 26 | Step: 10 | Loss: 0.3037\n",
      "Epoch 26 | Step: 11 | Loss: 0.3387\n",
      "Epoch 26 | Step: 12 | Loss: 0.2526\n",
      "Epoch 26 | Step: 13 | Loss: 0.4134\n",
      "Epoch 26 | Step: 14 | Loss: 0.3496\n",
      "Epoch 26 | Step: 15 | Loss: 0.3159\n",
      "Epoch 26 | Step: 16 | Loss: 0.2855\n",
      "Epoch 26 | Step: 17 | Loss: 0.2377\n",
      "Epoch 26 | Step: 18 | Loss: 0.3002\n",
      "Epoch 26 | Step: 19 | Loss: 0.2000\n",
      "Epoch 26 | Step: 20 | Loss: 0.1711\n",
      "Epoch 26 | Step: 21 | Loss: 0.2165\n",
      "Epoch 26 | Step: 22 | Loss: 0.2005\n",
      "Epoch 26 | Step: 23 | Loss: 0.2551\n",
      "Epoch 26 | Step: 24 | Loss: 0.3605\n",
      "Epoch 26 | Step: 25 | Loss: 0.3502\n",
      "Epoch 26 | Step: 26 | Loss: 0.2727\n",
      "Epoch 26 | Step: 27 | Loss: 0.2011\n",
      "Epoch 26 | Step: 28 | Loss: 0.4232\n",
      "Epoch 26 | Step: 29 | Loss: 0.3192\n",
      "Epoch 26 | Step: 30 | Loss: 0.3759\n",
      "Epoch 26 | Step: 31 | Loss: 0.3977\n",
      "Epoch 26 | Step: 32 | Loss: 0.1897\n",
      "Epoch 26 | Step: 33 | Loss: 0.3320\n",
      "Epoch 26 | Step: 34 | Loss: 0.2159\n",
      "Epoch 26 | Step: 35 | Loss: 0.1648\n",
      "Epoch 26 | Step: 36 | Loss: 0.2709\n",
      "Epoch 26 | Step: 37 | Loss: 0.2315\n",
      "Epoch 26 | Step: 38 | Loss: 0.4033\n",
      "Epoch 26 | Step: 39 | Loss: 0.3771\n",
      "Epoch 26 | Step: 40 | Loss: 0.3289\n",
      "Epoch 26 | Step: 41 | Loss: 0.3165\n",
      "Epoch 26 | Step: 42 | Loss: 0.3372\n",
      "Epoch 26 | Step: 43 | Loss: 0.4034\n",
      "Epoch 26 | Step: 44 | Loss: 0.3866\n",
      "Epoch 26 | Step: 45 | Loss: 0.2583\n",
      "Epoch 26 | Step: 46 | Loss: 0.3364\n",
      "Epoch 26 | Step: 47 | Loss: 0.8572\n",
      "Epoch 26 | Step: 48 | Loss: 0.2878\n",
      "Epoch 26 | Step: 49 | Loss: 0.2602\n",
      "Epoch 26 | Step: 50 | Loss: 0.3092\n",
      "Epoch 26 | Step: 51 | Loss: 0.4270\n",
      "Epoch 26 | Step: 52 | Loss: 0.2305\n",
      "Epoch 26 | Step: 53 | Loss: 0.2226\n",
      "Epoch 26 | Step: 54 | Loss: 0.3314\n",
      "Epoch 26 | Step: 55 | Loss: 0.2572\n",
      "Epoch 26 | Step: 56 | Loss: 0.2674\n",
      "Epoch 26 | Step: 57 | Loss: 0.3620\n",
      "Epoch 26 | Step: 58 | Loss: 0.3355\n",
      "Epoch 26 | Step: 59 | Loss: 0.2781\n",
      "Epoch 26 | Step: 60 | Loss: 0.3331\n",
      "Epoch 26 | Step: 61 | Loss: 0.3006\n",
      "Epoch 26 | Step: 62 | Loss: 0.3687\n",
      "Epoch 26 | Step: 63 | Loss: 0.2849\n",
      "Epoch 26 | Step: 64 | Loss: 0.2993\n",
      "Epoch 26 | Step: 65 | Loss: 0.2235\n",
      "Epoch 26 | Step: 66 | Loss: 0.5181\n",
      "Epoch 26 | Step: 67 | Loss: 0.2776\n",
      "Epoch 26 | Step: 68 | Loss: 0.2860\n",
      "Epoch 26 | Step: 69 | Loss: 0.3941\n",
      "Epoch 26 | Step: 70 | Loss: 0.3210\n",
      "Epoch 26 | Step: 71 | Loss: 0.3585\n",
      "Epoch 26 | Step: 72 | Loss: 0.3105\n",
      "Epoch 26 | Step: 73 | Loss: 0.2093\n",
      "Epoch 26 | Step: 74 | Loss: 0.3163\n",
      "Epoch 26 | Step: 75 | Loss: 0.2537\n",
      "Epoch 26 | Step: 76 | Loss: 0.3275\n",
      "Epoch 26 | Step: 77 | Loss: 0.2754\n",
      "Epoch 26 | Step: 78 | Loss: 0.3100\n",
      "Epoch 26 | Step: 79 | Loss: 0.5320\n",
      "Epoch 26 | Step: 80 | Loss: 0.3949\n",
      "Epoch 26 | Step: 81 | Loss: 0.2479\n",
      "Epoch 26 | Step: 82 | Loss: 0.3534\n",
      "Epoch 26 | Step: 83 | Loss: 0.4444\n",
      "Epoch 26 | Step: 84 | Loss: 0.2913\n",
      "Epoch 26 | Step: 85 | Loss: 0.1976\n",
      "Epoch 26 | Step: 86 | Loss: 0.1723\n",
      "Epoch 26 | Step: 87 | Loss: 0.1801\n",
      "Epoch 26 | Step: 88 | Loss: 0.2866\n",
      "Epoch 26 | Step: 89 | Loss: 0.4152\n",
      "Epoch 26 | Step: 90 | Loss: 0.3022\n",
      "Epoch 26 | Step: 91 | Loss: 0.2741\n",
      "Epoch 26 | Step: 92 | Loss: 0.1642\n",
      "Epoch 26 | Step: 93 | Loss: 0.2970\n",
      "Epoch 26 | Step: 94 | Loss: 0.4212\n",
      "Epoch 26 | Step: 95 | Loss: 0.3352\n",
      "Epoch 26 | Step: 96 | Loss: 0.2135\n",
      "Epoch 26 | Step: 97 | Loss: 0.1866\n",
      "Epoch 26 | Step: 98 | Loss: 0.5097\n",
      "Epoch 26 | Step: 99 | Loss: 0.3354\n",
      "Epoch 26 | Step: 100 | Loss: 0.2381\n",
      "Epoch 26 | Step: 101 | Loss: 0.3869\n",
      "Epoch 26 | Step: 102 | Loss: 0.2543\n",
      "Epoch 26 | Step: 103 | Loss: 0.2224\n",
      "Epoch 26 | Step: 104 | Loss: 0.3445\n",
      "Epoch 26 | Step: 105 | Loss: 0.3010\n",
      "Epoch 26 | Step: 106 | Loss: 0.3997\n",
      "Epoch 26 | Step: 107 | Loss: 0.2768\n",
      "Epoch 26 | Step: 108 | Loss: 0.2834\n",
      "Epoch 26 | Step: 109 | Loss: 0.3342\n",
      "Epoch 26 | Step: 110 | Loss: 0.3994\n",
      "Epoch 26 | Step: 111 | Loss: 0.3149\n",
      "Epoch 26 | Step: 112 | Loss: 0.3438\n",
      "Epoch 26 | Step: 113 | Loss: 0.2475\n",
      "Epoch 26 | Step: 114 | Loss: 0.1564\n",
      "Epoch 26 | Step: 115 | Loss: 0.2522\n",
      "Epoch 26 | Step: 116 | Loss: 0.2426\n",
      "Epoch 26 | Step: 117 | Loss: 0.6373\n",
      "Epoch 26 | Step: 118 | Loss: 0.3119\n",
      "Epoch 26 | Step: 119 | Loss: 0.2195\n",
      "Epoch 26 | Step: 120 | Loss: 0.2334\n",
      "Epoch 26 | Step: 121 | Loss: 0.3043\n",
      "Epoch 26 | Step: 122 | Loss: 0.3600\n",
      "Epoch 26 | Step: 123 | Loss: 0.3786\n",
      "Epoch 26 | Step: 124 | Loss: 0.3445\n",
      "Epoch 26 | Step: 125 | Loss: 0.2137\n",
      "Epoch 26 | Step: 126 | Loss: 0.3716\n",
      "Epoch 26 | Step: 127 | Loss: 0.2534\n",
      "Epoch 26 | Step: 128 | Loss: 0.2659\n",
      "Epoch 27 | Step: 1 | Loss: 0.2830\n",
      "Epoch 27 | Step: 2 | Loss: 0.4715\n",
      "Epoch 27 | Step: 3 | Loss: 0.3269\n",
      "Epoch 27 | Step: 4 | Loss: 0.3155\n",
      "Epoch 27 | Step: 5 | Loss: 0.2639\n",
      "Epoch 27 | Step: 6 | Loss: 0.3331\n",
      "Epoch 27 | Step: 7 | Loss: 0.2006\n",
      "Epoch 27 | Step: 8 | Loss: 0.3690\n",
      "Epoch 27 | Step: 9 | Loss: 0.3608\n",
      "Epoch 27 | Step: 10 | Loss: 0.3131\n",
      "Epoch 27 | Step: 11 | Loss: 0.3092\n",
      "Epoch 27 | Step: 12 | Loss: 0.2670\n",
      "Epoch 27 | Step: 13 | Loss: 0.2336\n",
      "Epoch 27 | Step: 14 | Loss: 0.2487\n",
      "Epoch 27 | Step: 15 | Loss: 0.2933\n",
      "Epoch 27 | Step: 16 | Loss: 0.2438\n",
      "Epoch 27 | Step: 17 | Loss: 0.3042\n",
      "Epoch 27 | Step: 18 | Loss: 0.3999\n",
      "Epoch 27 | Step: 19 | Loss: 0.2111\n",
      "Epoch 27 | Step: 20 | Loss: 0.2284\n",
      "Epoch 27 | Step: 21 | Loss: 0.4397\n",
      "Epoch 27 | Step: 22 | Loss: 0.2279\n",
      "Epoch 27 | Step: 23 | Loss: 0.4883\n",
      "Epoch 27 | Step: 24 | Loss: 0.3100\n",
      "Epoch 27 | Step: 25 | Loss: 0.3336\n",
      "Epoch 27 | Step: 26 | Loss: 0.2343\n",
      "Epoch 27 | Step: 27 | Loss: 0.3565\n",
      "Epoch 27 | Step: 28 | Loss: 0.3527\n",
      "Epoch 27 | Step: 29 | Loss: 0.2493\n",
      "Epoch 27 | Step: 30 | Loss: 0.3473\n",
      "Epoch 27 | Step: 31 | Loss: 0.2991\n",
      "Epoch 27 | Step: 32 | Loss: 0.3173\n",
      "Epoch 27 | Step: 33 | Loss: 0.7975\n",
      "Epoch 27 | Step: 34 | Loss: 0.3991\n",
      "Epoch 27 | Step: 35 | Loss: 0.2160\n",
      "Epoch 27 | Step: 36 | Loss: 0.1846\n",
      "Epoch 27 | Step: 37 | Loss: 0.2342\n",
      "Epoch 27 | Step: 38 | Loss: 0.3472\n",
      "Epoch 27 | Step: 39 | Loss: 0.2690\n",
      "Epoch 27 | Step: 40 | Loss: 0.2879\n",
      "Epoch 27 | Step: 41 | Loss: 0.2865\n",
      "Epoch 27 | Step: 42 | Loss: 0.3335\n",
      "Epoch 27 | Step: 43 | Loss: 0.4075\n",
      "Epoch 27 | Step: 44 | Loss: 0.2146\n",
      "Epoch 27 | Step: 45 | Loss: 0.3044\n",
      "Epoch 27 | Step: 46 | Loss: 0.2128\n",
      "Epoch 27 | Step: 47 | Loss: 0.3796\n",
      "Epoch 27 | Step: 48 | Loss: 0.2720\n",
      "Epoch 27 | Step: 49 | Loss: 0.2289\n",
      "Epoch 27 | Step: 50 | Loss: 0.3119\n",
      "Epoch 27 | Step: 51 | Loss: 0.2055\n",
      "Epoch 27 | Step: 52 | Loss: 0.4150\n",
      "Epoch 27 | Step: 53 | Loss: 0.3386\n",
      "Epoch 27 | Step: 54 | Loss: 0.2436\n",
      "Epoch 27 | Step: 55 | Loss: 0.3662\n",
      "Epoch 27 | Step: 56 | Loss: 0.4315\n",
      "Epoch 27 | Step: 57 | Loss: 0.5548\n",
      "Epoch 27 | Step: 58 | Loss: 0.3100\n",
      "Epoch 27 | Step: 59 | Loss: 0.2221\n",
      "Epoch 27 | Step: 60 | Loss: 0.3939\n",
      "Epoch 27 | Step: 61 | Loss: 0.2427\n",
      "Epoch 27 | Step: 62 | Loss: 0.2982\n",
      "Epoch 27 | Step: 63 | Loss: 0.2639\n",
      "Epoch 27 | Step: 64 | Loss: 0.2198\n",
      "Epoch 27 | Step: 65 | Loss: 0.1870\n",
      "Epoch 27 | Step: 66 | Loss: 0.2139\n",
      "Epoch 27 | Step: 67 | Loss: 0.2058\n",
      "Epoch 27 | Step: 68 | Loss: 0.2307\n",
      "Epoch 27 | Step: 69 | Loss: 0.2172\n",
      "Epoch 27 | Step: 70 | Loss: 0.1966\n",
      "Epoch 27 | Step: 71 | Loss: 0.3453\n",
      "Epoch 27 | Step: 72 | Loss: 0.1879\n",
      "Epoch 27 | Step: 73 | Loss: 0.2950\n",
      "Epoch 27 | Step: 74 | Loss: 0.3197\n",
      "Epoch 27 | Step: 75 | Loss: 0.2720\n",
      "Epoch 27 | Step: 76 | Loss: 0.2279\n",
      "Epoch 27 | Step: 77 | Loss: 0.2880\n",
      "Epoch 27 | Step: 78 | Loss: 0.3245\n",
      "Epoch 27 | Step: 79 | Loss: 0.3158\n",
      "Epoch 27 | Step: 80 | Loss: 0.3282\n",
      "Epoch 27 | Step: 81 | Loss: 0.1894\n",
      "Epoch 27 | Step: 82 | Loss: 0.1788\n",
      "Epoch 27 | Step: 83 | Loss: 0.6398\n",
      "Epoch 27 | Step: 84 | Loss: 0.2206\n",
      "Epoch 27 | Step: 85 | Loss: 0.2444\n",
      "Epoch 27 | Step: 86 | Loss: 0.2977\n",
      "Epoch 27 | Step: 87 | Loss: 0.2448\n",
      "Epoch 27 | Step: 88 | Loss: 0.2331\n",
      "Epoch 27 | Step: 89 | Loss: 0.2935\n",
      "Epoch 27 | Step: 90 | Loss: 0.2644\n",
      "Epoch 27 | Step: 91 | Loss: 0.4713\n",
      "Epoch 27 | Step: 92 | Loss: 0.2732\n",
      "Epoch 27 | Step: 93 | Loss: 0.4010\n",
      "Epoch 27 | Step: 94 | Loss: 0.3031\n",
      "Epoch 27 | Step: 95 | Loss: 0.3076\n",
      "Epoch 27 | Step: 96 | Loss: 0.3561\n",
      "Epoch 27 | Step: 97 | Loss: 0.2482\n",
      "Epoch 27 | Step: 98 | Loss: 0.2048\n",
      "Epoch 27 | Step: 99 | Loss: 0.2980\n",
      "Epoch 27 | Step: 100 | Loss: 0.2514\n",
      "Epoch 27 | Step: 101 | Loss: 0.4022\n",
      "Epoch 27 | Step: 102 | Loss: 0.2904\n",
      "Epoch 27 | Step: 103 | Loss: 0.3280\n",
      "Epoch 27 | Step: 104 | Loss: 0.2986\n",
      "Epoch 27 | Step: 105 | Loss: 0.3814\n",
      "Epoch 27 | Step: 106 | Loss: 0.2822\n",
      "Epoch 27 | Step: 107 | Loss: 0.2752\n",
      "Epoch 27 | Step: 108 | Loss: 0.4049\n",
      "Epoch 27 | Step: 109 | Loss: 0.2757\n",
      "Epoch 27 | Step: 110 | Loss: 0.3886\n",
      "Epoch 27 | Step: 111 | Loss: 0.3714\n",
      "Epoch 27 | Step: 112 | Loss: 0.1877\n",
      "Epoch 27 | Step: 113 | Loss: 0.4090\n",
      "Epoch 27 | Step: 114 | Loss: 0.3913\n",
      "Epoch 27 | Step: 115 | Loss: 0.3606\n",
      "Epoch 27 | Step: 116 | Loss: 0.3002\n",
      "Epoch 27 | Step: 117 | Loss: 0.3422\n",
      "Epoch 27 | Step: 118 | Loss: 0.2295\n",
      "Epoch 27 | Step: 119 | Loss: 0.3240\n",
      "Epoch 27 | Step: 120 | Loss: 0.2156\n",
      "Epoch 27 | Step: 121 | Loss: 0.2876\n",
      "Epoch 27 | Step: 122 | Loss: 0.2958\n",
      "Epoch 27 | Step: 123 | Loss: 0.3240\n",
      "Epoch 27 | Step: 124 | Loss: 0.3499\n",
      "Epoch 27 | Step: 125 | Loss: 0.3306\n",
      "Epoch 27 | Step: 126 | Loss: 0.2305\n",
      "Epoch 27 | Step: 127 | Loss: 0.4141\n",
      "Epoch 27 | Step: 128 | Loss: 0.2474\n",
      "Epoch 28 | Step: 1 | Loss: 0.2960\n",
      "Epoch 28 | Step: 2 | Loss: 0.3252\n",
      "Epoch 28 | Step: 3 | Loss: 0.2465\n",
      "Epoch 28 | Step: 4 | Loss: 0.3546\n",
      "Epoch 28 | Step: 5 | Loss: 0.3266\n",
      "Epoch 28 | Step: 6 | Loss: 0.2664\n",
      "Epoch 28 | Step: 7 | Loss: 0.3117\n",
      "Epoch 28 | Step: 8 | Loss: 0.3182\n",
      "Epoch 28 | Step: 9 | Loss: 0.2693\n",
      "Epoch 28 | Step: 10 | Loss: 0.2521\n",
      "Epoch 28 | Step: 11 | Loss: 0.4061\n",
      "Epoch 28 | Step: 12 | Loss: 0.3779\n",
      "Epoch 28 | Step: 13 | Loss: 0.2137\n",
      "Epoch 28 | Step: 14 | Loss: 0.3943\n",
      "Epoch 28 | Step: 15 | Loss: 0.1989\n",
      "Epoch 28 | Step: 16 | Loss: 0.3659\n",
      "Epoch 28 | Step: 17 | Loss: 0.3209\n",
      "Epoch 28 | Step: 18 | Loss: 0.2270\n",
      "Epoch 28 | Step: 19 | Loss: 0.3052\n",
      "Epoch 28 | Step: 20 | Loss: 0.2551\n",
      "Epoch 28 | Step: 21 | Loss: 0.2445\n",
      "Epoch 28 | Step: 22 | Loss: 0.3079\n",
      "Epoch 28 | Step: 23 | Loss: 0.2768\n",
      "Epoch 28 | Step: 24 | Loss: 0.2585\n",
      "Epoch 28 | Step: 25 | Loss: 0.2876\n",
      "Epoch 28 | Step: 26 | Loss: 0.3034\n",
      "Epoch 28 | Step: 27 | Loss: 0.3351\n",
      "Epoch 28 | Step: 28 | Loss: 0.1617\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 28 | Step: 29 | Loss: 0.4996\n",
      "Epoch 28 | Step: 30 | Loss: 0.5298\n",
      "Epoch 28 | Step: 31 | Loss: 0.2798\n",
      "Epoch 28 | Step: 32 | Loss: 0.3899\n",
      "Epoch 28 | Step: 33 | Loss: 0.2209\n",
      "Epoch 28 | Step: 34 | Loss: 0.3816\n",
      "Epoch 28 | Step: 35 | Loss: 0.2094\n",
      "Epoch 28 | Step: 36 | Loss: 0.3501\n",
      "Epoch 28 | Step: 37 | Loss: 0.3143\n",
      "Epoch 28 | Step: 38 | Loss: 0.2310\n",
      "Epoch 28 | Step: 39 | Loss: 0.2341\n",
      "Epoch 28 | Step: 40 | Loss: 0.3205\n",
      "Epoch 28 | Step: 41 | Loss: 0.2151\n",
      "Epoch 28 | Step: 42 | Loss: 0.3428\n",
      "Epoch 28 | Step: 43 | Loss: 0.2458\n",
      "Epoch 28 | Step: 44 | Loss: 0.2248\n",
      "Epoch 28 | Step: 45 | Loss: 0.2679\n",
      "Epoch 28 | Step: 46 | Loss: 0.3575\n",
      "Epoch 28 | Step: 47 | Loss: 0.2555\n",
      "Epoch 28 | Step: 48 | Loss: 0.3015\n",
      "Epoch 28 | Step: 49 | Loss: 0.1572\n",
      "Epoch 28 | Step: 50 | Loss: 0.1598\n",
      "Epoch 28 | Step: 51 | Loss: 0.3699\n",
      "Epoch 28 | Step: 52 | Loss: 0.3314\n",
      "Epoch 28 | Step: 53 | Loss: 0.2363\n",
      "Epoch 28 | Step: 54 | Loss: 0.2840\n",
      "Epoch 28 | Step: 55 | Loss: 0.2707\n",
      "Epoch 28 | Step: 56 | Loss: 0.2823\n",
      "Epoch 28 | Step: 57 | Loss: 0.3869\n",
      "Epoch 28 | Step: 58 | Loss: 0.2331\n",
      "Epoch 28 | Step: 59 | Loss: 0.4171\n",
      "Epoch 28 | Step: 60 | Loss: 0.2449\n",
      "Epoch 28 | Step: 61 | Loss: 0.2011\n",
      "Epoch 28 | Step: 62 | Loss: 0.2611\n",
      "Epoch 28 | Step: 63 | Loss: 0.3709\n",
      "Epoch 28 | Step: 64 | Loss: 0.3348\n",
      "Epoch 28 | Step: 65 | Loss: 0.2924\n",
      "Epoch 28 | Step: 66 | Loss: 0.6777\n",
      "Epoch 28 | Step: 67 | Loss: 0.2889\n",
      "Epoch 28 | Step: 68 | Loss: 0.1080\n",
      "Epoch 28 | Step: 69 | Loss: 0.2794\n",
      "Epoch 28 | Step: 70 | Loss: 0.2663\n",
      "Epoch 28 | Step: 71 | Loss: 0.2634\n",
      "Epoch 28 | Step: 72 | Loss: 0.1664\n",
      "Epoch 28 | Step: 73 | Loss: 0.2019\n",
      "Epoch 28 | Step: 74 | Loss: 0.3943\n",
      "Epoch 28 | Step: 75 | Loss: 0.3023\n",
      "Epoch 28 | Step: 76 | Loss: 0.3072\n",
      "Epoch 28 | Step: 77 | Loss: 0.7496\n",
      "Epoch 28 | Step: 78 | Loss: 0.3579\n",
      "Epoch 28 | Step: 79 | Loss: 0.2889\n",
      "Epoch 28 | Step: 80 | Loss: 0.2363\n",
      "Epoch 28 | Step: 81 | Loss: 0.4201\n",
      "Epoch 28 | Step: 82 | Loss: 0.3915\n",
      "Epoch 28 | Step: 83 | Loss: 0.3817\n",
      "Epoch 28 | Step: 84 | Loss: 0.3118\n",
      "Epoch 28 | Step: 85 | Loss: 0.2947\n",
      "Epoch 28 | Step: 86 | Loss: 0.3083\n",
      "Epoch 28 | Step: 87 | Loss: 0.2749\n",
      "Epoch 28 | Step: 88 | Loss: 0.5140\n",
      "Epoch 28 | Step: 89 | Loss: 0.2877\n",
      "Epoch 28 | Step: 90 | Loss: 0.3988\n",
      "Epoch 28 | Step: 91 | Loss: 0.3222\n",
      "Epoch 28 | Step: 92 | Loss: 0.3431\n",
      "Epoch 28 | Step: 93 | Loss: 0.2862\n",
      "Epoch 28 | Step: 94 | Loss: 0.3108\n",
      "Epoch 28 | Step: 95 | Loss: 0.2674\n",
      "Epoch 28 | Step: 96 | Loss: 0.2376\n",
      "Epoch 28 | Step: 97 | Loss: 0.1672\n",
      "Epoch 28 | Step: 98 | Loss: 0.3550\n",
      "Epoch 28 | Step: 99 | Loss: 0.2939\n",
      "Epoch 28 | Step: 100 | Loss: 0.3364\n",
      "Epoch 28 | Step: 101 | Loss: 0.4095\n",
      "Epoch 28 | Step: 102 | Loss: 0.2939\n",
      "Epoch 28 | Step: 103 | Loss: 0.2439\n",
      "Epoch 28 | Step: 104 | Loss: 0.2662\n",
      "Epoch 28 | Step: 105 | Loss: 0.3814\n",
      "Epoch 28 | Step: 106 | Loss: 0.3373\n",
      "Epoch 28 | Step: 107 | Loss: 0.2945\n",
      "Epoch 28 | Step: 108 | Loss: 0.2819\n",
      "Epoch 28 | Step: 109 | Loss: 0.2714\n",
      "Epoch 28 | Step: 110 | Loss: 0.4601\n",
      "Epoch 28 | Step: 111 | Loss: 0.2107\n",
      "Epoch 28 | Step: 112 | Loss: 0.3480\n",
      "Epoch 28 | Step: 113 | Loss: 0.1860\n",
      "Epoch 28 | Step: 114 | Loss: 0.2374\n",
      "Epoch 28 | Step: 115 | Loss: 0.2172\n",
      "Epoch 28 | Step: 116 | Loss: 0.3132\n",
      "Epoch 28 | Step: 117 | Loss: 0.3955\n",
      "Epoch 28 | Step: 118 | Loss: 0.2456\n",
      "Epoch 28 | Step: 119 | Loss: 0.2411\n",
      "Epoch 28 | Step: 120 | Loss: 0.3651\n",
      "Epoch 28 | Step: 121 | Loss: 0.2989\n",
      "Epoch 28 | Step: 122 | Loss: 0.1406\n",
      "Epoch 28 | Step: 123 | Loss: 0.2165\n",
      "Epoch 28 | Step: 124 | Loss: 0.4125\n",
      "Epoch 28 | Step: 125 | Loss: 0.1888\n",
      "Epoch 28 | Step: 126 | Loss: 0.3177\n",
      "Epoch 28 | Step: 127 | Loss: 0.2458\n",
      "Epoch 28 | Step: 128 | Loss: 0.2473\n",
      "Epoch 29 | Step: 1 | Loss: 0.3325\n",
      "Epoch 29 | Step: 2 | Loss: 0.2845\n",
      "Epoch 29 | Step: 3 | Loss: 0.4044\n",
      "Epoch 29 | Step: 4 | Loss: 0.2635\n",
      "Epoch 29 | Step: 5 | Loss: 0.4745\n",
      "Epoch 29 | Step: 6 | Loss: 0.2003\n",
      "Epoch 29 | Step: 7 | Loss: 0.3391\n",
      "Epoch 29 | Step: 8 | Loss: 0.3434\n",
      "Epoch 29 | Step: 9 | Loss: 0.3424\n",
      "Epoch 29 | Step: 10 | Loss: 0.2967\n",
      "Epoch 29 | Step: 11 | Loss: 0.2721\n",
      "Epoch 29 | Step: 12 | Loss: 0.2817\n",
      "Epoch 29 | Step: 13 | Loss: 0.1876\n",
      "Epoch 29 | Step: 14 | Loss: 0.2783\n",
      "Epoch 29 | Step: 15 | Loss: 0.2679\n",
      "Epoch 29 | Step: 16 | Loss: 0.2771\n",
      "Epoch 29 | Step: 17 | Loss: 0.2470\n",
      "Epoch 29 | Step: 18 | Loss: 0.2384\n",
      "Epoch 29 | Step: 19 | Loss: 0.2034\n",
      "Epoch 29 | Step: 20 | Loss: 0.3950\n",
      "Epoch 29 | Step: 21 | Loss: 0.1645\n",
      "Epoch 29 | Step: 22 | Loss: 0.2659\n",
      "Epoch 29 | Step: 23 | Loss: 0.1538\n",
      "Epoch 29 | Step: 24 | Loss: 0.2856\n",
      "Epoch 29 | Step: 25 | Loss: 0.1721\n",
      "Epoch 29 | Step: 26 | Loss: 0.2933\n",
      "Epoch 29 | Step: 27 | Loss: 0.1756\n",
      "Epoch 29 | Step: 28 | Loss: 0.2599\n",
      "Epoch 29 | Step: 29 | Loss: 0.2551\n",
      "Epoch 29 | Step: 30 | Loss: 0.4885\n",
      "Epoch 29 | Step: 31 | Loss: 0.2855\n",
      "Epoch 29 | Step: 32 | Loss: 0.2776\n",
      "Epoch 29 | Step: 33 | Loss: 0.2600\n",
      "Epoch 29 | Step: 34 | Loss: 0.3616\n",
      "Epoch 29 | Step: 35 | Loss: 0.2432\n",
      "Epoch 29 | Step: 36 | Loss: 0.2066\n",
      "Epoch 29 | Step: 37 | Loss: 0.2913\n",
      "Epoch 29 | Step: 38 | Loss: 0.1317\n",
      "Epoch 29 | Step: 39 | Loss: 0.1747\n",
      "Epoch 29 | Step: 40 | Loss: 0.4340\n",
      "Epoch 29 | Step: 41 | Loss: 0.2740\n",
      "Epoch 29 | Step: 42 | Loss: 0.3175\n",
      "Epoch 29 | Step: 43 | Loss: 0.1806\n",
      "Epoch 29 | Step: 44 | Loss: 0.3703\n",
      "Epoch 29 | Step: 45 | Loss: 0.3506\n",
      "Epoch 29 | Step: 46 | Loss: 0.4202\n",
      "Epoch 29 | Step: 47 | Loss: 0.3121\n",
      "Epoch 29 | Step: 48 | Loss: 0.4318\n",
      "Epoch 29 | Step: 49 | Loss: 0.3683\n",
      "Epoch 29 | Step: 50 | Loss: 0.3281\n",
      "Epoch 29 | Step: 51 | Loss: 0.2402\n",
      "Epoch 29 | Step: 52 | Loss: 0.2803\n",
      "Epoch 29 | Step: 53 | Loss: 0.2472\n",
      "Epoch 29 | Step: 54 | Loss: 0.2720\n",
      "Epoch 29 | Step: 55 | Loss: 0.2646\n",
      "Epoch 29 | Step: 56 | Loss: 0.2937\n",
      "Epoch 29 | Step: 57 | Loss: 0.3049\n",
      "Epoch 29 | Step: 58 | Loss: 0.1916\n",
      "Epoch 29 | Step: 59 | Loss: 0.2971\n",
      "Epoch 29 | Step: 60 | Loss: 0.3172\n",
      "Epoch 29 | Step: 61 | Loss: 0.3601\n",
      "Epoch 29 | Step: 62 | Loss: 0.2690\n",
      "Epoch 29 | Step: 63 | Loss: 0.3270\n",
      "Epoch 29 | Step: 64 | Loss: 0.2998\n",
      "Epoch 29 | Step: 65 | Loss: 0.2701\n",
      "Epoch 29 | Step: 66 | Loss: 0.2042\n",
      "Epoch 29 | Step: 67 | Loss: 0.4679\n",
      "Epoch 29 | Step: 68 | Loss: 0.2773\n",
      "Epoch 29 | Step: 69 | Loss: 0.2626\n",
      "Epoch 29 | Step: 70 | Loss: 0.3750\n",
      "Epoch 29 | Step: 71 | Loss: 0.2426\n",
      "Epoch 29 | Step: 72 | Loss: 0.3278\n",
      "Epoch 29 | Step: 73 | Loss: 0.3246\n",
      "Epoch 29 | Step: 74 | Loss: 0.3228\n",
      "Epoch 29 | Step: 75 | Loss: 0.2011\n",
      "Epoch 29 | Step: 76 | Loss: 0.3933\n",
      "Epoch 29 | Step: 77 | Loss: 0.2090\n",
      "Epoch 29 | Step: 78 | Loss: 0.1868\n",
      "Epoch 29 | Step: 79 | Loss: 0.3002\n",
      "Epoch 29 | Step: 80 | Loss: 0.3675\n",
      "Epoch 29 | Step: 81 | Loss: 0.3976\n",
      "Epoch 29 | Step: 82 | Loss: 0.3040\n",
      "Epoch 29 | Step: 83 | Loss: 0.3492\n",
      "Epoch 29 | Step: 84 | Loss: 0.2825\n",
      "Epoch 29 | Step: 85 | Loss: 0.3756\n",
      "Epoch 29 | Step: 86 | Loss: 0.3167\n",
      "Epoch 29 | Step: 87 | Loss: 0.2481\n",
      "Epoch 29 | Step: 88 | Loss: 0.3569\n",
      "Epoch 29 | Step: 89 | Loss: 0.2942\n",
      "Epoch 29 | Step: 90 | Loss: 0.2876\n",
      "Epoch 29 | Step: 91 | Loss: 0.2901\n",
      "Epoch 29 | Step: 92 | Loss: 0.3397\n",
      "Epoch 29 | Step: 93 | Loss: 0.2034\n",
      "Epoch 29 | Step: 94 | Loss: 0.3782\n",
      "Epoch 29 | Step: 95 | Loss: 0.6001\n",
      "Epoch 29 | Step: 96 | Loss: 0.2719\n",
      "Epoch 29 | Step: 97 | Loss: 0.2607\n",
      "Epoch 29 | Step: 98 | Loss: 0.2329\n",
      "Epoch 29 | Step: 99 | Loss: 0.2492\n",
      "Epoch 29 | Step: 100 | Loss: 0.3021\n",
      "Epoch 29 | Step: 101 | Loss: 0.2930\n",
      "Epoch 29 | Step: 102 | Loss: 0.2862\n",
      "Epoch 29 | Step: 103 | Loss: 0.3041\n",
      "Epoch 29 | Step: 104 | Loss: 0.3506\n",
      "Epoch 29 | Step: 105 | Loss: 0.2978\n",
      "Epoch 29 | Step: 106 | Loss: 0.2755\n",
      "Epoch 29 | Step: 107 | Loss: 0.7757\n",
      "Epoch 29 | Step: 108 | Loss: 0.3644\n",
      "Epoch 29 | Step: 109 | Loss: 0.2158\n",
      "Epoch 29 | Step: 110 | Loss: 0.3784\n",
      "Epoch 29 | Step: 111 | Loss: 0.3138\n",
      "Epoch 29 | Step: 112 | Loss: 0.3389\n",
      "Epoch 29 | Step: 113 | Loss: 0.3875\n",
      "Epoch 29 | Step: 114 | Loss: 0.2529\n",
      "Epoch 29 | Step: 115 | Loss: 0.2196\n",
      "Epoch 29 | Step: 116 | Loss: 0.2210\n",
      "Epoch 29 | Step: 117 | Loss: 0.3228\n",
      "Epoch 29 | Step: 118 | Loss: 0.2814\n",
      "Epoch 29 | Step: 119 | Loss: 0.2324\n",
      "Epoch 29 | Step: 120 | Loss: 0.2938\n",
      "Epoch 29 | Step: 121 | Loss: 0.2897\n",
      "Epoch 29 | Step: 122 | Loss: 0.2824\n",
      "Epoch 29 | Step: 123 | Loss: 0.3303\n",
      "Epoch 29 | Step: 124 | Loss: 0.2282\n",
      "Epoch 29 | Step: 125 | Loss: 0.2743\n",
      "Epoch 29 | Step: 126 | Loss: 0.2026\n",
      "Epoch 29 | Step: 127 | Loss: 0.3365\n",
      "Epoch 29 | Step: 128 | Loss: 0.2817\n",
      "Epoch 30 | Step: 1 | Loss: 0.3426\n",
      "Epoch 30 | Step: 2 | Loss: 0.1487\n",
      "Epoch 30 | Step: 3 | Loss: 0.2668\n",
      "Epoch 30 | Step: 4 | Loss: 0.2681\n",
      "Epoch 30 | Step: 5 | Loss: 0.2197\n",
      "Epoch 30 | Step: 6 | Loss: 0.2465\n",
      "Epoch 30 | Step: 7 | Loss: 0.3141\n",
      "Epoch 30 | Step: 8 | Loss: 0.2848\n",
      "Epoch 30 | Step: 9 | Loss: 0.2615\n",
      "Epoch 30 | Step: 10 | Loss: 0.5438\n",
      "Epoch 30 | Step: 11 | Loss: 0.2412\n",
      "Epoch 30 | Step: 12 | Loss: 0.2700\n",
      "Epoch 30 | Step: 13 | Loss: 0.4042\n",
      "Epoch 30 | Step: 14 | Loss: 0.3091\n",
      "Epoch 30 | Step: 15 | Loss: 0.2385\n",
      "Epoch 30 | Step: 16 | Loss: 0.2882\n",
      "Epoch 30 | Step: 17 | Loss: 0.3459\n",
      "Epoch 30 | Step: 18 | Loss: 0.2157\n",
      "Epoch 30 | Step: 19 | Loss: 0.2548\n",
      "Epoch 30 | Step: 20 | Loss: 0.2676\n",
      "Epoch 30 | Step: 21 | Loss: 0.3895\n",
      "Epoch 30 | Step: 22 | Loss: 0.3359\n",
      "Epoch 30 | Step: 23 | Loss: 0.3675\n",
      "Epoch 30 | Step: 24 | Loss: 0.3212\n",
      "Epoch 30 | Step: 25 | Loss: 0.2330\n",
      "Epoch 30 | Step: 26 | Loss: 0.3491\n",
      "Epoch 30 | Step: 27 | Loss: 0.2117\n",
      "Epoch 30 | Step: 28 | Loss: 0.1929\n",
      "Epoch 30 | Step: 29 | Loss: 0.3641\n",
      "Epoch 30 | Step: 30 | Loss: 0.2364\n",
      "Epoch 30 | Step: 31 | Loss: 0.2948\n",
      "Epoch 30 | Step: 32 | Loss: 0.2896\n",
      "Epoch 30 | Step: 33 | Loss: 0.2982\n",
      "Epoch 30 | Step: 34 | Loss: 0.2533\n",
      "Epoch 30 | Step: 35 | Loss: 0.3088\n",
      "Epoch 30 | Step: 36 | Loss: 0.2238\n",
      "Epoch 30 | Step: 37 | Loss: 0.3854\n",
      "Epoch 30 | Step: 38 | Loss: 0.3801\n",
      "Epoch 30 | Step: 39 | Loss: 0.3668\n",
      "Epoch 30 | Step: 40 | Loss: 0.4427\n",
      "Epoch 30 | Step: 41 | Loss: 0.2672\n",
      "Epoch 30 | Step: 42 | Loss: 0.3003\n",
      "Epoch 30 | Step: 43 | Loss: 0.3043\n",
      "Epoch 30 | Step: 44 | Loss: 0.2110\n",
      "Epoch 30 | Step: 45 | Loss: 0.2239\n",
      "Epoch 30 | Step: 46 | Loss: 0.2349\n",
      "Epoch 30 | Step: 47 | Loss: 0.2858\n",
      "Epoch 30 | Step: 48 | Loss: 0.1796\n",
      "Epoch 30 | Step: 49 | Loss: 0.2756\n",
      "Epoch 30 | Step: 50 | Loss: 0.2355\n",
      "Epoch 30 | Step: 51 | Loss: 0.3870\n",
      "Epoch 30 | Step: 52 | Loss: 0.2951\n",
      "Epoch 30 | Step: 53 | Loss: 0.2661\n",
      "Epoch 30 | Step: 54 | Loss: 0.2474\n",
      "Epoch 30 | Step: 55 | Loss: 0.3027\n",
      "Epoch 30 | Step: 56 | Loss: 0.1753\n",
      "Epoch 30 | Step: 57 | Loss: 0.2732\n",
      "Epoch 30 | Step: 58 | Loss: 0.3003\n",
      "Epoch 30 | Step: 59 | Loss: 0.3070\n",
      "Epoch 30 | Step: 60 | Loss: 0.3258\n",
      "Epoch 30 | Step: 61 | Loss: 0.4057\n",
      "Epoch 30 | Step: 62 | Loss: 0.1936\n",
      "Epoch 30 | Step: 63 | Loss: 0.2300\n",
      "Epoch 30 | Step: 64 | Loss: 0.2712\n",
      "Epoch 30 | Step: 65 | Loss: 0.2664\n",
      "Epoch 30 | Step: 66 | Loss: 0.3021\n",
      "Epoch 30 | Step: 67 | Loss: 0.4327\n",
      "Epoch 30 | Step: 68 | Loss: 0.2367\n",
      "Epoch 30 | Step: 69 | Loss: 0.3190\n",
      "Epoch 30 | Step: 70 | Loss: 0.2017\n",
      "Epoch 30 | Step: 71 | Loss: 0.2441\n",
      "Epoch 30 | Step: 72 | Loss: 0.2435\n",
      "Epoch 30 | Step: 73 | Loss: 0.1586\n",
      "Epoch 30 | Step: 74 | Loss: 0.3039\n",
      "Epoch 30 | Step: 75 | Loss: 0.2381\n",
      "Epoch 30 | Step: 76 | Loss: 0.3216\n",
      "Epoch 30 | Step: 77 | Loss: 0.2405\n",
      "Epoch 30 | Step: 78 | Loss: 0.2901\n",
      "Epoch 30 | Step: 79 | Loss: 0.3140\n",
      "Epoch 30 | Step: 80 | Loss: 0.4118\n",
      "Epoch 30 | Step: 81 | Loss: 0.1605\n",
      "Epoch 30 | Step: 82 | Loss: 0.2646\n",
      "Epoch 30 | Step: 83 | Loss: 0.3718\n",
      "Epoch 30 | Step: 84 | Loss: 0.3035\n",
      "Epoch 30 | Step: 85 | Loss: 0.2577\n",
      "Epoch 30 | Step: 86 | Loss: 0.2790\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 30 | Step: 87 | Loss: 0.4822\n",
      "Epoch 30 | Step: 88 | Loss: 0.2121\n",
      "Epoch 30 | Step: 89 | Loss: 0.2996\n",
      "Epoch 30 | Step: 90 | Loss: 0.2331\n",
      "Epoch 30 | Step: 91 | Loss: 0.2600\n",
      "Epoch 30 | Step: 92 | Loss: 0.4091\n",
      "Epoch 30 | Step: 93 | Loss: 0.2010\n",
      "Epoch 30 | Step: 94 | Loss: 0.3474\n",
      "Epoch 30 | Step: 95 | Loss: 0.2558\n",
      "Epoch 30 | Step: 96 | Loss: 0.2976\n",
      "Epoch 30 | Step: 97 | Loss: 0.2444\n",
      "Epoch 30 | Step: 98 | Loss: 0.5026\n",
      "Epoch 30 | Step: 99 | Loss: 0.2477\n",
      "Epoch 30 | Step: 100 | Loss: 0.2361\n",
      "Epoch 30 | Step: 101 | Loss: 0.4338\n",
      "Epoch 30 | Step: 102 | Loss: 0.2400\n",
      "Epoch 30 | Step: 103 | Loss: 0.3347\n",
      "Epoch 30 | Step: 104 | Loss: 0.2383\n",
      "Epoch 30 | Step: 105 | Loss: 0.2924\n",
      "Epoch 30 | Step: 106 | Loss: 0.2542\n",
      "Epoch 30 | Step: 107 | Loss: 0.2650\n",
      "Epoch 30 | Step: 108 | Loss: 0.3975\n",
      "Epoch 30 | Step: 109 | Loss: 0.2828\n",
      "Epoch 30 | Step: 110 | Loss: 0.2157\n",
      "Epoch 30 | Step: 111 | Loss: 0.2746\n",
      "Epoch 30 | Step: 112 | Loss: 0.3259\n",
      "Epoch 30 | Step: 113 | Loss: 0.3263\n",
      "Epoch 30 | Step: 114 | Loss: 0.3054\n",
      "Epoch 30 | Step: 115 | Loss: 0.3925\n",
      "Epoch 30 | Step: 116 | Loss: 0.2834\n",
      "Epoch 30 | Step: 117 | Loss: 0.1902\n",
      "Epoch 30 | Step: 118 | Loss: 0.3337\n",
      "Epoch 30 | Step: 119 | Loss: 0.3453\n",
      "Epoch 30 | Step: 120 | Loss: 0.3902\n",
      "Epoch 30 | Step: 121 | Loss: 0.3575\n",
      "Epoch 30 | Step: 122 | Loss: 0.1707\n",
      "Epoch 30 | Step: 123 | Loss: 0.2403\n",
      "Epoch 30 | Step: 124 | Loss: 0.7500\n",
      "Epoch 30 | Step: 125 | Loss: 0.1925\n",
      "Epoch 30 | Step: 126 | Loss: 0.2913\n",
      "Epoch 30 | Step: 127 | Loss: 0.4192\n",
      "Epoch 30 | Step: 128 | Loss: 0.2351\n"
     ]
    }
   ],
   "source": [
    "for epoch in range(30):\n",
    "    for i, data in enumerate(train_loader):\n",
    "        inputs, labels = data\n",
    "\n",
    "        y_pred = model(inputs)\n",
    "\n",
    "        motor_updrs_mse = criterion(y_pred[:, 0], labels[:, 0])\n",
    "        total_updrs_mse= criterion(y_pred[:, 1], labels[:, 1])\n",
    "\n",
    "        total_loss = 3 * motor_updrs_mse + total_updrs_mse\n",
    "        \n",
    "        print(f'Epoch {epoch + 1} | Step: {i+1} | Loss: {total_loss.item():.4f}')\n",
    "\n",
    "        optimizer.zero_grad()\n",
    "        total_loss.backward()\n",
    "        optimizer.step()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss: 0.0754\tmotor_updrs_mse: 0.1102\ttotal_updrs_mse: 0.0406\n",
      "loss: 0.0834\tmotor_updrs_mse: 0.1341\ttotal_updrs_mse: 0.0326\n",
      "loss: 0.0761\tmotor_updrs_mse: 0.1285\ttotal_updrs_mse: 0.0238\n",
      "loss: 0.0985\tmotor_updrs_mse: 0.1527\ttotal_updrs_mse: 0.0442\n",
      "loss: 0.0877\tmotor_updrs_mse: 0.1584\ttotal_updrs_mse: 0.0171\n",
      "loss: 0.0873\tmotor_updrs_mse: 0.1601\ttotal_updrs_mse: 0.0144\n",
      "loss: 0.0518\tmotor_updrs_mse: 0.0986\ttotal_updrs_mse: 0.0050\n",
      "loss: 0.0632\tmotor_updrs_mse: 0.1185\ttotal_updrs_mse: 0.0079\n",
      "loss: 0.0198\tmotor_updrs_mse: 0.0337\ttotal_updrs_mse: 0.0059\n",
      "loss: 0.0907\tmotor_updrs_mse: 0.1756\ttotal_updrs_mse: 0.0057\n",
      "loss: 0.1541\tmotor_updrs_mse: 0.3004\ttotal_updrs_mse: 0.0078\n",
      "loss: 0.1065\tmotor_updrs_mse: 0.2032\ttotal_updrs_mse: 0.0097\n",
      "loss: 0.1255\tmotor_updrs_mse: 0.2427\ttotal_updrs_mse: 0.0083\n",
      "loss: 0.0664\tmotor_updrs_mse: 0.0736\ttotal_updrs_mse: 0.0592\n",
      "loss: 0.0480\tmotor_updrs_mse: 0.0177\ttotal_updrs_mse: 0.0782\n",
      "loss: 0.0423\tmotor_updrs_mse: 0.0168\ttotal_updrs_mse: 0.0677\n",
      "loss: 0.0323\tmotor_updrs_mse: 0.0155\ttotal_updrs_mse: 0.0492\n",
      "loss: 0.0263\tmotor_updrs_mse: 0.0135\ttotal_updrs_mse: 0.0390\n",
      "loss: 0.3664\tmotor_updrs_mse: 0.7212\ttotal_updrs_mse: 0.0117\n",
      "loss: 0.4214\tmotor_updrs_mse: 0.8336\ttotal_updrs_mse: 0.0091\n",
      "loss: 0.4142\tmotor_updrs_mse: 0.8172\ttotal_updrs_mse: 0.0112\n",
      "loss: 0.3829\tmotor_updrs_mse: 0.7577\ttotal_updrs_mse: 0.0080\n",
      "loss: 0.3624\tmotor_updrs_mse: 0.7167\ttotal_updrs_mse: 0.0081\n",
      "loss: 0.2000\tmotor_updrs_mse: 0.2962\ttotal_updrs_mse: 0.1037\n",
      "loss: 0.1359\tmotor_updrs_mse: 0.1398\ttotal_updrs_mse: 0.1319\n",
      "loss: 0.1148\tmotor_updrs_mse: 0.0728\ttotal_updrs_mse: 0.1568\n",
      "loss: 0.1127\tmotor_updrs_mse: 0.0531\ttotal_updrs_mse: 0.1723\n",
      "loss: 0.0874\tmotor_updrs_mse: 0.0791\ttotal_updrs_mse: 0.0957\n",
      "loss: 0.0769\tmotor_updrs_mse: 0.0947\ttotal_updrs_mse: 0.0591\n",
      "loss: 0.0631\tmotor_updrs_mse: 0.0946\ttotal_updrs_mse: 0.0315\n",
      "loss: 0.0689\tmotor_updrs_mse: 0.1271\ttotal_updrs_mse: 0.0106\n",
      "loss: 0.0559\tmotor_updrs_mse: 0.1047\ttotal_updrs_mse: 0.0072\n",
      "loss: 0.0156\tmotor_updrs_mse: 0.0219\ttotal_updrs_mse: 0.0093\n",
      "loss: 0.0194\tmotor_updrs_mse: 0.0289\ttotal_updrs_mse: 0.0099\n",
      "loss: 0.0181\tmotor_updrs_mse: 0.0246\ttotal_updrs_mse: 0.0116\n",
      "loss: 0.0168\tmotor_updrs_mse: 0.0248\ttotal_updrs_mse: 0.0087\n",
      "loss: 0.0467\tmotor_updrs_mse: 0.0536\ttotal_updrs_mse: 0.0398\n",
      "loss: 0.0286\tmotor_updrs_mse: 0.0482\ttotal_updrs_mse: 0.0091\n",
      "loss: 0.0309\tmotor_updrs_mse: 0.0520\ttotal_updrs_mse: 0.0098\n",
      "loss: 0.0288\tmotor_updrs_mse: 0.0404\ttotal_updrs_mse: 0.0172\n",
      "loss: 0.0281\tmotor_updrs_mse: 0.0490\ttotal_updrs_mse: 0.0071\n",
      "loss: 0.0358\tmotor_updrs_mse: 0.0512\ttotal_updrs_mse: 0.0203\n",
      "loss: 0.0315\tmotor_updrs_mse: 0.0425\ttotal_updrs_mse: 0.0205\n",
      "loss: 0.0474\tmotor_updrs_mse: 0.0696\ttotal_updrs_mse: 0.0252\n",
      "loss: 0.0376\tmotor_updrs_mse: 0.0581\ttotal_updrs_mse: 0.0172\n",
      "loss: 0.0386\tmotor_updrs_mse: 0.0576\ttotal_updrs_mse: 0.0196\n",
      "loss: 0.0163\tmotor_updrs_mse: 0.0197\ttotal_updrs_mse: 0.0128\n",
      "loss: 0.0186\tmotor_updrs_mse: 0.0238\ttotal_updrs_mse: 0.0135\n",
      "loss: 0.0196\tmotor_updrs_mse: 0.0155\ttotal_updrs_mse: 0.0236\n",
      "loss: 0.0207\tmotor_updrs_mse: 0.0250\ttotal_updrs_mse: 0.0165\n",
      "loss: 0.0250\tmotor_updrs_mse: 0.0364\ttotal_updrs_mse: 0.0135\n",
      "loss: 0.0385\tmotor_updrs_mse: 0.0644\ttotal_updrs_mse: 0.0126\n",
      "loss: 0.0412\tmotor_updrs_mse: 0.0776\ttotal_updrs_mse: 0.0048\n",
      "loss: 0.0487\tmotor_updrs_mse: 0.0881\ttotal_updrs_mse: 0.0093\n",
      "loss: 0.0684\tmotor_updrs_mse: 0.1300\ttotal_updrs_mse: 0.0069\n",
      "loss: 0.0970\tmotor_updrs_mse: 0.1871\ttotal_updrs_mse: 0.0069\n",
      "\n",
      "\n",
      "[RESULT]\taverage all_updrs_mse: 0.0912\n",
      "\t\taverage motor_updrs_mse: 0.1519\n",
      "\t\taverage total_updrs_mse: 0.0305\n"
     ]
    }
   ],
   "source": [
    "pred_arr = []\n",
    "labels_arr = []\n",
    "\n",
    "avg_all_updrs_mse = 0\n",
    "avg_motor_updrs_mse = 0\n",
    "avg_total_updrs_mse = 0\n",
    "    \n",
    "for i, data in enumerate(test_loader):\n",
    "    inputs, labels = data\n",
    "\n",
    "    test_pred = model(inputs)\n",
    "    \n",
    "    pred_arr.append(test_pred.detach().numpy())\n",
    "    labels_arr.append(labels.detach().numpy())\n",
    "    \n",
    "    motor_updrs_preds = test_pred[:, 0]\n",
    "    motor_updrs_labels = labels[:, 0]\n",
    "        \n",
    "    total_updrs_preds = test_pred[:, 1]\n",
    "    total_updrs_labels = labels[:, 1]\n",
    "    \n",
    "    motor_updrs_mse = criterion(motor_updrs_preds, motor_updrs_labels)\n",
    "    total_updrs_mse = criterion(total_updrs_preds, total_updrs_labels)  \n",
    "    all_updrs_mse = criterion(test_pred, labels)\n",
    "\n",
    "    avg_all_updrs_mse += all_updrs_mse.item()\n",
    "    avg_motor_updrs_mse += motor_updrs_mse.item()\n",
    "    avg_total_updrs_mse += total_updrs_mse.item()\n",
    "    \n",
    "    print(f'loss: {all_updrs_mse.item():.4f}\\tmotor_updrs_mse: {motor_updrs_mse.item():.4f}\\ttotal_updrs_mse: {total_updrs_mse.item():.4f}')\n",
    "        \n",
    "        \n",
    "\n",
    "        \n",
    "print(f\"\\n\\n[RESULT]\\taverage all_updrs_mse: {avg_all_updrs_mse/i:.4f}\")\n",
    "print(f\"\\t\\taverage motor_updrs_mse: {avg_motor_updrs_mse/i:.4f}\")\n",
    "print(f\"\\t\\taverage total_updrs_mse: {avg_total_updrs_mse/i:.4f}\")  \n",
    "\n",
    "test_pred = np.array(pred_arr[:-1]).reshape(-1, 2)\n",
    "labels = np.array(labels_arr[:-1]).reshape(-1, 2)\n",
    "#avg = all_updrs_mse / 32.0    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def draw_figure(pred, label):\n",
    "    plt.figure(figsize=(15,15))\n",
    "    t = np.arange(0, len(pred), 1)\n",
    "    plt.plot(t, pred, c='blue', label='predict_updrs')\n",
    "    plt.plot(t, label, c='red', label='real_updrs')\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#motor_updrs\n",
    "draw_figure(test_pred[:, 0], labels[:, 0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#total_updrs\n",
    "draw_figure(test_pred[:, 1], labels[:, 1])"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}