06 - MNIST cwiczenie.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9f9fbf0d",
   "metadata": {},
   "source": [
    "<img src=\"data:image/png;base64,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\" width=\"150\" style=\"float:left\">\n",
    "<div style=\"clear:both\"></div>\n",
    "Rafał Nowak<br/>\n",
    "<a href=\"https://github.com/rnoxy\">github.com/rnoxy</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b754d8c",
   "metadata": {},
   "source": [
    "> `!pip install torch`<br/>\n",
    "> `!pip install torchvision`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2385168b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torchvision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "10370e80",
   "metadata": {},
   "outputs": [],
   "source": [
    "from torchvision.datasets import MNIST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e1c4d0e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_train = MNIST(root='datasets/', download=True, train=True)\n",
    "dataset_test  = MNIST(root='datasets/', download=False, train=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8a2e84c0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 10000)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(dataset_train), len(dataset_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "880adc48",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = dataset_train.data / 255\n",
    "X_test = dataset_test.data / 255\n",
    "\n",
    "X_train = torch.flatten( X_train, start_dim=1 )\n",
    "X_test = torch.flatten( X_test, start_dim=1 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8728c950",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([60000, 784]), torch.Size([10000, 784]))"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "28e0dda5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "9921c80a",
   "metadata": {},
   "outputs": [],
   "source": [
    "W1 = torch.randn( 784, 100, requires_grad=True )\n",
    "b1 = torch.randn( 100, requires_grad=True )\n",
    "\n",
    "W2 = torch.randn( 100, 10, requires_grad=True )\n",
    "b2 = torch.randn( 10, requires_grad=True )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "24ccb500",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "046cc05d",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = dataset_train.targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "b63dc274",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cross entropy losss\n",
    "loss = torch.nn.functional.cross_entropy( Z2, y_train )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "35b23e87",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  0,  step=   0,  loss=19.017,  accuracy=0.062\n",
      "Epoch  0,  step=   1,  loss=14.729,  accuracy=0.125\n",
      "Epoch  0,  step=   2,  loss=15.514,  accuracy=0.094\n",
      "Epoch  0,  step=   3,  loss=15.559,  accuracy=0.062\n",
      "Epoch  0,  step=   4,  loss=9.644,  accuracy=0.094\n",
      "Epoch  0,  step=   5,  loss=11.971,  accuracy=0.156\n",
      "Epoch  0,  step=   6,  loss=11.573,  accuracy=0.094\n",
      "Epoch  0,  step=   7,  loss=11.651,  accuracy=0.062\n",
      "Epoch  0,  step=   8,  loss=9.478,  accuracy=0.344\n",
      "Epoch  0,  step=   9,  loss=10.129,  accuracy=0.250\n",
      "Epoch  0,  step=  10,  loss=9.583,  accuracy=0.188\n",
      "Epoch  0,  step=  11,  loss=7.953,  accuracy=0.281\n",
      "Epoch  0,  step=  12,  loss=9.822,  accuracy=0.219\n",
      "Epoch  0,  step=  13,  loss=11.110,  accuracy=0.219\n",
      "Epoch  0,  step=  14,  loss=7.975,  accuracy=0.219\n",
      "Epoch  0,  step=  15,  loss=10.232,  accuracy=0.125\n",
      "Epoch  0,  step=  16,  loss=9.602,  accuracy=0.156\n",
      "Epoch  0,  step=  17,  loss=7.410,  accuracy=0.312\n",
      "Epoch  0,  step=  18,  loss=7.978,  accuracy=0.219\n",
      "Epoch  0,  step=  19,  loss=13.428,  accuracy=0.156\n",
      "Epoch  0,  step=  20,  loss=9.715,  accuracy=0.156\n",
      "Epoch  0,  step=  21,  loss=5.844,  accuracy=0.375\n",
      "Epoch  0,  step=  22,  loss=6.118,  accuracy=0.312\n",
      "Epoch  0,  step=  23,  loss=8.083,  accuracy=0.219\n",
      "Epoch  0,  step=  24,  loss=5.565,  accuracy=0.469\n",
      "Epoch  0,  step=  25,  loss=4.800,  accuracy=0.312\n",
      "Epoch  0,  step=  26,  loss=5.256,  accuracy=0.250\n",
      "Epoch  0,  step=  27,  loss=7.046,  accuracy=0.188\n",
      "Epoch  0,  step=  28,  loss=7.264,  accuracy=0.219\n",
      "Epoch  0,  step=  29,  loss=6.497,  accuracy=0.312\n",
      "Epoch  0,  step=  30,  loss=5.484,  accuracy=0.250\n",
      "Epoch  0,  step=  31,  loss=5.919,  accuracy=0.312\n",
      "Epoch  0,  step=  32,  loss=6.959,  accuracy=0.250\n",
      "Epoch  0,  step=  33,  loss=6.514,  accuracy=0.188\n",
      "Epoch  0,  step=  34,  loss=6.727,  accuracy=0.344\n",
      "Epoch  0,  step=  35,  loss=7.741,  accuracy=0.219\n",
      "Epoch  0,  step=  36,  loss=7.598,  accuracy=0.344\n",
      "Epoch  0,  step=  37,  loss=4.908,  accuracy=0.469\n",
      "Epoch  0,  step=  38,  loss=4.819,  accuracy=0.375\n",
      "Epoch  0,  step=  39,  loss=4.928,  accuracy=0.406\n",
      "Epoch  0,  step=  40,  loss=4.866,  accuracy=0.406\n",
      "Epoch  0,  step=  41,  loss=5.532,  accuracy=0.375\n",
      "Epoch  0,  step=  42,  loss=6.103,  accuracy=0.375\n",
      "Epoch  0,  step=  43,  loss=7.317,  accuracy=0.344\n",
      "Epoch  0,  step=  44,  loss=6.039,  accuracy=0.219\n",
      "Epoch  0,  step=  45,  loss=3.945,  accuracy=0.406\n",
      "Epoch  0,  step=  46,  loss=5.030,  accuracy=0.438\n",
      "Epoch  0,  step=  47,  loss=5.069,  accuracy=0.406\n",
      "Epoch  0,  step=  48,  loss=5.794,  accuracy=0.312\n",
      "Epoch  0,  step=  49,  loss=5.256,  accuracy=0.375\n",
      "Epoch  0,  step=  50,  loss=6.014,  accuracy=0.375\n",
      "Epoch  0,  step=  51,  loss=3.892,  accuracy=0.406\n",
      "Epoch  0,  step=  52,  loss=3.747,  accuracy=0.531\n",
      "Epoch  0,  step=  53,  loss=4.060,  accuracy=0.438\n",
      "Epoch  0,  step=  54,  loss=3.849,  accuracy=0.594\n",
      "Epoch  0,  step=  55,  loss=4.064,  accuracy=0.531\n",
      "Epoch  0,  step=  56,  loss=4.296,  accuracy=0.375\n",
      "Epoch  0,  step=  57,  loss=4.544,  accuracy=0.469\n",
      "Epoch  0,  step=  58,  loss=3.576,  accuracy=0.438\n",
      "Epoch  0,  step=  59,  loss=5.007,  accuracy=0.438\n",
      "Epoch  0,  step=  60,  loss=4.182,  accuracy=0.469\n",
      "Epoch  0,  step=  61,  loss=4.242,  accuracy=0.375\n",
      "Epoch  0,  step=  62,  loss=3.982,  accuracy=0.500\n",
      "Epoch  0,  step=  63,  loss=4.752,  accuracy=0.406\n",
      "Epoch  0,  step=  64,  loss=6.635,  accuracy=0.344\n",
      "Epoch  0,  step=  65,  loss=2.230,  accuracy=0.562\n",
      "Epoch  0,  step=  66,  loss=3.952,  accuracy=0.469\n",
      "Epoch  0,  step=  67,  loss=2.134,  accuracy=0.719\n",
      "Epoch  0,  step=  68,  loss=3.125,  accuracy=0.594\n",
      "Epoch  0,  step=  69,  loss=3.966,  accuracy=0.594\n",
      "Epoch  0,  step=  70,  loss=5.073,  accuracy=0.406\n",
      "Epoch  0,  step=  71,  loss=4.639,  accuracy=0.406\n",
      "Epoch  0,  step=  72,  loss=4.189,  accuracy=0.406\n",
      "Epoch  0,  step=  73,  loss=3.467,  accuracy=0.469\n",
      "Epoch  0,  step=  74,  loss=5.126,  accuracy=0.312\n",
      "Epoch  0,  step=  75,  loss=4.298,  accuracy=0.438\n",
      "Epoch  0,  step=  76,  loss=4.960,  accuracy=0.406\n",
      "Epoch  0,  step=  77,  loss=3.834,  accuracy=0.469\n",
      "Epoch  0,  step=  78,  loss=3.055,  accuracy=0.594\n",
      "Epoch  0,  step=  79,  loss=3.089,  accuracy=0.594\n",
      "Epoch  0,  step=  80,  loss=3.234,  accuracy=0.562\n",
      "Epoch  0,  step=  81,  loss=2.662,  accuracy=0.469\n",
      "Epoch  0,  step=  82,  loss=2.962,  accuracy=0.562\n",
      "Epoch  0,  step=  83,  loss=3.485,  accuracy=0.625\n",
      "Epoch  0,  step=  84,  loss=4.904,  accuracy=0.594\n",
      "Epoch  0,  step=  85,  loss=5.716,  accuracy=0.438\n",
      "Epoch  0,  step=  86,  loss=4.078,  accuracy=0.562\n",
      "Epoch  0,  step=  87,  loss=5.532,  accuracy=0.500\n",
      "Epoch  0,  step=  88,  loss=3.373,  accuracy=0.531\n",
      "Epoch  0,  step=  89,  loss=2.766,  accuracy=0.594\n",
      "Epoch  0,  step=  90,  loss=3.624,  accuracy=0.500\n",
      "Epoch  0,  step=  91,  loss=3.240,  accuracy=0.688\n",
      "Epoch  0,  step=  92,  loss=2.410,  accuracy=0.562\n",
      "Epoch  0,  step=  93,  loss=3.601,  accuracy=0.562\n",
      "Epoch  0,  step=  94,  loss=3.504,  accuracy=0.531\n",
      "Epoch  0,  step=  95,  loss=3.688,  accuracy=0.562\n",
      "Epoch  0,  step=  96,  loss=2.945,  accuracy=0.656\n",
      "Epoch  0,  step=  97,  loss=2.427,  accuracy=0.719\n",
      "Epoch  0,  step=  98,  loss=4.120,  accuracy=0.656\n",
      "Epoch  0,  step=  99,  loss=3.539,  accuracy=0.500\n",
      "Epoch  0,  step= 100,  loss=2.686,  accuracy=0.594\n",
      "Epoch  0,  step= 101,  loss=3.634,  accuracy=0.500\n",
      "Epoch  0,  step= 102,  loss=2.990,  accuracy=0.531\n",
      "Epoch  0,  step= 103,  loss=2.774,  accuracy=0.500\n",
      "Epoch  0,  step= 104,  loss=4.539,  accuracy=0.500\n",
      "Epoch  0,  step= 105,  loss=3.993,  accuracy=0.500\n",
      "Epoch  0,  step= 106,  loss=3.043,  accuracy=0.531\n",
      "Epoch  0,  step= 107,  loss=2.156,  accuracy=0.531\n",
      "Epoch  0,  step= 108,  loss=1.860,  accuracy=0.750\n",
      "Epoch  0,  step= 109,  loss=3.238,  accuracy=0.562\n",
      "Epoch  0,  step= 110,  loss=3.384,  accuracy=0.500\n",
      "Epoch  0,  step= 111,  loss=4.497,  accuracy=0.500\n",
      "Epoch  0,  step= 112,  loss=2.534,  accuracy=0.688\n",
      "Epoch  0,  step= 113,  loss=2.790,  accuracy=0.656\n",
      "Epoch  0,  step= 114,  loss=3.419,  accuracy=0.594\n",
      "Epoch  0,  step= 115,  loss=2.393,  accuracy=0.594\n",
      "Epoch  0,  step= 116,  loss=2.991,  accuracy=0.531\n",
      "Epoch  0,  step= 117,  loss=4.402,  accuracy=0.375\n",
      "Epoch  0,  step= 118,  loss=4.569,  accuracy=0.531\n",
      "Epoch  0,  step= 119,  loss=3.512,  accuracy=0.406\n",
      "Epoch  0,  step= 120,  loss=2.040,  accuracy=0.688\n",
      "Epoch  0,  step= 121,  loss=2.400,  accuracy=0.719\n",
      "Epoch  0,  step= 122,  loss=1.969,  accuracy=0.625\n",
      "Epoch  0,  step= 123,  loss=2.097,  accuracy=0.688\n",
      "Epoch  0,  step= 124,  loss=4.415,  accuracy=0.500\n",
      "Epoch  0,  step= 125,  loss=2.684,  accuracy=0.625\n",
      "Epoch  0,  step= 126,  loss=1.696,  accuracy=0.656\n",
      "Epoch  0,  step= 127,  loss=1.941,  accuracy=0.719\n",
      "Epoch  0,  step= 128,  loss=2.407,  accuracy=0.688\n",
      "Epoch  0,  step= 129,  loss=2.623,  accuracy=0.500\n",
      "Epoch  0,  step= 130,  loss=2.710,  accuracy=0.594\n",
      "Epoch  0,  step= 131,  loss=1.973,  accuracy=0.750\n",
      "Epoch  0,  step= 132,  loss=1.262,  accuracy=0.781\n",
      "Epoch  0,  step= 133,  loss=4.081,  accuracy=0.469\n",
      "Epoch  0,  step= 134,  loss=2.061,  accuracy=0.688\n",
      "Epoch  0,  step= 135,  loss=2.691,  accuracy=0.656\n",
      "Epoch  0,  step= 136,  loss=2.777,  accuracy=0.688\n",
      "Epoch  0,  step= 137,  loss=1.361,  accuracy=0.688\n",
      "Epoch  0,  step= 138,  loss=2.554,  accuracy=0.594\n",
      "Epoch  0,  step= 139,  loss=3.666,  accuracy=0.438\n",
      "Epoch  0,  step= 140,  loss=1.941,  accuracy=0.625\n",
      "Epoch  0,  step= 141,  loss=1.448,  accuracy=0.719\n",
      "Epoch  0,  step= 142,  loss=1.096,  accuracy=0.781\n",
      "Epoch  0,  step= 143,  loss=2.692,  accuracy=0.656\n",
      "Epoch  0,  step= 144,  loss=3.952,  accuracy=0.562\n",
      "Epoch  0,  step= 145,  loss=3.004,  accuracy=0.500\n",
      "Epoch  0,  step= 146,  loss=2.694,  accuracy=0.438\n",
      "Epoch  0,  step= 147,  loss=3.709,  accuracy=0.562\n",
      "Epoch  0,  step= 148,  loss=3.836,  accuracy=0.531\n",
      "Epoch  0,  step= 149,  loss=2.781,  accuracy=0.531\n",
      "Epoch  0,  step= 150,  loss=4.327,  accuracy=0.375\n",
      "Epoch  0,  step= 151,  loss=2.415,  accuracy=0.594\n",
      "Epoch  0,  step= 152,  loss=1.625,  accuracy=0.688\n",
      "Epoch  0,  step= 153,  loss=2.258,  accuracy=0.625\n",
      "Epoch  0,  step= 154,  loss=2.883,  accuracy=0.625\n",
      "Epoch  0,  step= 155,  loss=2.658,  accuracy=0.500\n",
      "Epoch  0,  step= 156,  loss=1.993,  accuracy=0.750\n",
      "Epoch  0,  step= 157,  loss=3.223,  accuracy=0.594\n",
      "Epoch  0,  step= 158,  loss=3.487,  accuracy=0.469\n",
      "Epoch  0,  step= 159,  loss=2.910,  accuracy=0.531\n",
      "Epoch  0,  step= 160,  loss=2.558,  accuracy=0.594\n",
      "Epoch  0,  step= 161,  loss=4.195,  accuracy=0.344\n",
      "Epoch  0,  step= 162,  loss=2.543,  accuracy=0.562\n",
      "Epoch  0,  step= 163,  loss=1.521,  accuracy=0.688\n",
      "Epoch  0,  step= 164,  loss=2.456,  accuracy=0.625\n",
      "Epoch  0,  step= 165,  loss=3.707,  accuracy=0.469\n",
      "Epoch  0,  step= 166,  loss=2.809,  accuracy=0.531\n",
      "Epoch  0,  step= 167,  loss=1.932,  accuracy=0.688\n",
      "Epoch  0,  step= 168,  loss=2.443,  accuracy=0.625\n",
      "Epoch  0,  step= 169,  loss=1.141,  accuracy=0.812\n",
      "Epoch  0,  step= 170,  loss=1.142,  accuracy=0.719\n",
      "Epoch  0,  step= 171,  loss=1.769,  accuracy=0.812\n",
      "Epoch  0,  step= 172,  loss=2.044,  accuracy=0.656\n",
      "Epoch  0,  step= 173,  loss=2.619,  accuracy=0.562\n",
      "Epoch  0,  step= 174,  loss=1.049,  accuracy=0.844\n",
      "Epoch  0,  step= 175,  loss=2.784,  accuracy=0.594\n",
      "Epoch  0,  step= 176,  loss=3.263,  accuracy=0.531\n",
      "Epoch  0,  step= 177,  loss=2.453,  accuracy=0.625\n",
      "Epoch  0,  step= 178,  loss=3.218,  accuracy=0.562\n",
      "Epoch  0,  step= 179,  loss=2.198,  accuracy=0.531\n",
      "Epoch  0,  step= 180,  loss=2.669,  accuracy=0.688\n",
      "Epoch  0,  step= 181,  loss=2.441,  accuracy=0.562\n",
      "Epoch  0,  step= 182,  loss=2.742,  accuracy=0.531\n",
      "Epoch  0,  step= 183,  loss=2.140,  accuracy=0.656\n",
      "Epoch  0,  step= 184,  loss=1.900,  accuracy=0.625\n",
      "Epoch  0,  step= 185,  loss=3.666,  accuracy=0.625\n",
      "Epoch  0,  step= 186,  loss=2.754,  accuracy=0.688\n",
      "Epoch  0,  step= 187,  loss=1.234,  accuracy=0.812\n",
      "Epoch  0,  step= 188,  loss=0.763,  accuracy=0.812\n",
      "Epoch  0,  step= 189,  loss=0.957,  accuracy=0.844\n",
      "Epoch  0,  step= 190,  loss=2.280,  accuracy=0.750\n",
      "Epoch  0,  step= 191,  loss=1.135,  accuracy=0.875\n",
      "Epoch  0,  step= 192,  loss=1.118,  accuracy=0.719\n",
      "Epoch  0,  step= 193,  loss=1.119,  accuracy=0.875\n",
      "Epoch  0,  step= 194,  loss=1.307,  accuracy=0.656\n",
      "Epoch  0,  step= 195,  loss=2.301,  accuracy=0.625\n",
      "Epoch  0,  step= 196,  loss=2.223,  accuracy=0.594\n",
      "Epoch  0,  step= 197,  loss=2.591,  accuracy=0.719\n",
      "Epoch  0,  step= 198,  loss=1.573,  accuracy=0.750\n",
      "Epoch  0,  step= 199,  loss=1.699,  accuracy=0.750\n",
      "Epoch  0,  step= 200,  loss=1.704,  accuracy=0.656\n",
      "Epoch  0,  step= 201,  loss=1.973,  accuracy=0.625\n",
      "Epoch  0,  step= 202,  loss=2.299,  accuracy=0.688\n",
      "Epoch  0,  step= 203,  loss=2.632,  accuracy=0.688\n",
      "Epoch  0,  step= 204,  loss=1.524,  accuracy=0.781\n",
      "Epoch  0,  step= 205,  loss=2.298,  accuracy=0.719\n",
      "Epoch  0,  step= 206,  loss=0.791,  accuracy=0.906\n",
      "Epoch  0,  step= 207,  loss=1.011,  accuracy=0.812\n",
      "Epoch  0,  step= 208,  loss=1.205,  accuracy=0.750\n",
      "Epoch  0,  step= 209,  loss=2.138,  accuracy=0.719\n",
      "Epoch  0,  step= 210,  loss=1.836,  accuracy=0.781\n",
      "Epoch  0,  step= 211,  loss=1.179,  accuracy=0.688\n",
      "Epoch  0,  step= 212,  loss=2.334,  accuracy=0.594\n",
      "Epoch  0,  step= 213,  loss=2.662,  accuracy=0.688\n",
      "Epoch  0,  step= 214,  loss=2.265,  accuracy=0.594\n",
      "Epoch  0,  step= 215,  loss=2.549,  accuracy=0.531\n",
      "Epoch  0,  step= 216,  loss=2.585,  accuracy=0.688\n",
      "Epoch  0,  step= 217,  loss=2.111,  accuracy=0.594\n",
      "Epoch  0,  step= 218,  loss=1.812,  accuracy=0.594\n",
      "Epoch  0,  step= 219,  loss=3.327,  accuracy=0.469\n",
      "Epoch  0,  step= 220,  loss=2.054,  accuracy=0.594\n",
      "Epoch  0,  step= 221,  loss=2.770,  accuracy=0.594\n",
      "Epoch  0,  step= 222,  loss=1.017,  accuracy=0.812\n",
      "Epoch  0,  step= 223,  loss=1.319,  accuracy=0.812\n",
      "Epoch  0,  step= 224,  loss=2.038,  accuracy=0.531\n",
      "Epoch  0,  step= 225,  loss=3.546,  accuracy=0.594\n",
      "Epoch  0,  step= 226,  loss=3.551,  accuracy=0.500\n",
      "Epoch  0,  step= 227,  loss=3.412,  accuracy=0.469\n",
      "Epoch  0,  step= 228,  loss=2.470,  accuracy=0.562\n",
      "Epoch  0,  step= 229,  loss=3.958,  accuracy=0.531\n",
      "Epoch  0,  step= 230,  loss=2.319,  accuracy=0.719\n",
      "Epoch  0,  step= 231,  loss=0.606,  accuracy=0.844\n",
      "Epoch  0,  step= 232,  loss=3.135,  accuracy=0.594\n",
      "Epoch  0,  step= 233,  loss=1.751,  accuracy=0.625\n",
      "Epoch  0,  step= 234,  loss=2.077,  accuracy=0.625\n",
      "Epoch  0,  step= 235,  loss=2.487,  accuracy=0.562\n",
      "Epoch  0,  step= 236,  loss=1.062,  accuracy=0.750\n",
      "Epoch  0,  step= 237,  loss=3.407,  accuracy=0.531\n",
      "Epoch  0,  step= 238,  loss=4.837,  accuracy=0.469\n",
      "Epoch  0,  step= 239,  loss=1.882,  accuracy=0.688\n",
      "Epoch  0,  step= 240,  loss=0.824,  accuracy=0.750\n",
      "Epoch  0,  step= 241,  loss=2.642,  accuracy=0.531\n",
      "Epoch  0,  step= 242,  loss=2.321,  accuracy=0.594\n",
      "Epoch  0,  step= 243,  loss=2.450,  accuracy=0.719\n",
      "Epoch  0,  step= 244,  loss=1.634,  accuracy=0.750\n",
      "Epoch  0,  step= 245,  loss=0.928,  accuracy=0.781\n",
      "Epoch  0,  step= 246,  loss=3.339,  accuracy=0.562\n",
      "Epoch  0,  step= 247,  loss=2.275,  accuracy=0.656\n",
      "Epoch  0,  step= 248,  loss=2.180,  accuracy=0.750\n",
      "Epoch  0,  step= 249,  loss=2.243,  accuracy=0.688\n",
      "Epoch  0,  step= 250,  loss=2.743,  accuracy=0.500\n",
      "Epoch  0,  step= 251,  loss=0.854,  accuracy=0.781\n",
      "Epoch  0,  step= 252,  loss=1.440,  accuracy=0.750\n",
      "Epoch  0,  step= 253,  loss=2.624,  accuracy=0.688\n",
      "Epoch  0,  step= 254,  loss=1.583,  accuracy=0.719\n",
      "Epoch  0,  step= 255,  loss=0.921,  accuracy=0.812\n",
      "Epoch  0,  step= 256,  loss=2.590,  accuracy=0.594\n",
      "Epoch  0,  step= 257,  loss=2.503,  accuracy=0.656\n",
      "Epoch  0,  step= 258,  loss=1.670,  accuracy=0.656\n",
      "Epoch  0,  step= 259,  loss=1.391,  accuracy=0.719\n",
      "Epoch  0,  step= 260,  loss=1.447,  accuracy=0.656\n",
      "Epoch  0,  step= 261,  loss=1.159,  accuracy=0.750\n",
      "Epoch  0,  step= 262,  loss=1.787,  accuracy=0.688\n",
      "Epoch  0,  step= 263,  loss=1.552,  accuracy=0.719\n",
      "Epoch  0,  step= 264,  loss=2.871,  accuracy=0.531\n",
      "Epoch  0,  step= 265,  loss=1.377,  accuracy=0.625\n",
      "Epoch  0,  step= 266,  loss=0.955,  accuracy=0.812\n",
      "Epoch  0,  step= 267,  loss=1.553,  accuracy=0.719\n",
      "Epoch  0,  step= 268,  loss=1.046,  accuracy=0.781\n",
      "Epoch  0,  step= 269,  loss=2.417,  accuracy=0.625\n",
      "Epoch  0,  step= 270,  loss=2.171,  accuracy=0.594\n",
      "Epoch  0,  step= 271,  loss=3.897,  accuracy=0.562\n",
      "Epoch  0,  step= 272,  loss=2.904,  accuracy=0.688\n",
      "Epoch  0,  step= 273,  loss=2.442,  accuracy=0.594\n",
      "Epoch  0,  step= 274,  loss=1.549,  accuracy=0.625\n",
      "Epoch  0,  step= 275,  loss=1.226,  accuracy=0.750\n",
      "Epoch  0,  step= 276,  loss=1.752,  accuracy=0.562\n",
      "Epoch  0,  step= 277,  loss=2.433,  accuracy=0.562\n",
      "Epoch  0,  step= 278,  loss=2.803,  accuracy=0.625\n",
      "Epoch  0,  step= 279,  loss=0.649,  accuracy=0.781\n",
      "Epoch  0,  step= 280,  loss=2.837,  accuracy=0.781\n",
      "Epoch  0,  step= 281,  loss=1.534,  accuracy=0.719\n",
      "Epoch  0,  step= 282,  loss=1.012,  accuracy=0.750\n",
      "Epoch  0,  step= 283,  loss=1.314,  accuracy=0.781\n",
      "Epoch  0,  step= 284,  loss=1.660,  accuracy=0.750\n",
      "Epoch  0,  step= 285,  loss=1.569,  accuracy=0.594\n",
      "Epoch  0,  step= 286,  loss=1.167,  accuracy=0.750\n",
      "Epoch  0,  step= 287,  loss=1.741,  accuracy=0.781\n",
      "Epoch  0,  step= 288,  loss=1.035,  accuracy=0.750\n",
      "Epoch  0,  step= 289,  loss=2.365,  accuracy=0.750\n",
      "Epoch  0,  step= 290,  loss=2.515,  accuracy=0.594\n",
      "Epoch  0,  step= 291,  loss=1.912,  accuracy=0.625\n",
      "Epoch  0,  step= 292,  loss=1.166,  accuracy=0.750\n",
      "Epoch  0,  step= 293,  loss=1.741,  accuracy=0.750\n",
      "Epoch  0,  step= 294,  loss=1.238,  accuracy=0.781\n",
      "Epoch  0,  step= 295,  loss=1.744,  accuracy=0.688\n",
      "Epoch  0,  step= 296,  loss=1.391,  accuracy=0.750\n",
      "Epoch  0,  step= 297,  loss=1.840,  accuracy=0.688\n",
      "Epoch  0,  step= 298,  loss=2.301,  accuracy=0.688\n",
      "Epoch  0,  step= 299,  loss=1.805,  accuracy=0.656\n",
      "Epoch  0,  step= 300,  loss=1.680,  accuracy=0.688\n",
      "Epoch  0,  step= 301,  loss=1.743,  accuracy=0.719\n",
      "Epoch  0,  step= 302,  loss=2.088,  accuracy=0.625\n",
      "Epoch  0,  step= 303,  loss=0.663,  accuracy=0.844\n",
      "Epoch  0,  step= 304,  loss=2.115,  accuracy=0.781\n",
      "Epoch  0,  step= 305,  loss=2.408,  accuracy=0.688\n",
      "Epoch  0,  step= 306,  loss=1.865,  accuracy=0.750\n",
      "Epoch  0,  step= 307,  loss=0.966,  accuracy=0.812\n",
      "Epoch  0,  step= 308,  loss=1.577,  accuracy=0.719\n",
      "Epoch  0,  step= 309,  loss=0.922,  accuracy=0.875\n",
      "Epoch  0,  step= 310,  loss=1.252,  accuracy=0.750\n",
      "Epoch  0,  step= 311,  loss=0.746,  accuracy=0.844\n",
      "Epoch  0,  step= 312,  loss=1.216,  accuracy=0.750\n",
      "Epoch  0,  step= 313,  loss=2.453,  accuracy=0.656\n",
      "Epoch  0,  step= 314,  loss=1.168,  accuracy=0.750\n",
      "Epoch  0,  step= 315,  loss=1.724,  accuracy=0.719\n",
      "Epoch  0,  step= 316,  loss=1.010,  accuracy=0.812\n",
      "Epoch  0,  step= 317,  loss=1.993,  accuracy=0.688\n",
      "Epoch  0,  step= 318,  loss=2.742,  accuracy=0.656\n",
      "Epoch  0,  step= 319,  loss=1.519,  accuracy=0.688\n",
      "Epoch  0,  step= 320,  loss=2.019,  accuracy=0.625\n",
      "Epoch  0,  step= 321,  loss=1.853,  accuracy=0.688\n",
      "Epoch  0,  step= 322,  loss=1.101,  accuracy=0.844\n",
      "Epoch  0,  step= 323,  loss=1.435,  accuracy=0.812\n",
      "Epoch  0,  step= 324,  loss=1.097,  accuracy=0.719\n",
      "Epoch  0,  step= 325,  loss=1.856,  accuracy=0.688\n",
      "Epoch  0,  step= 326,  loss=2.139,  accuracy=0.562\n",
      "Epoch  0,  step= 327,  loss=1.005,  accuracy=0.875\n",
      "Epoch  0,  step= 328,  loss=1.930,  accuracy=0.688\n",
      "Epoch  0,  step= 329,  loss=0.816,  accuracy=0.781\n",
      "Epoch  0,  step= 330,  loss=0.797,  accuracy=0.844\n",
      "Epoch  0,  step= 331,  loss=0.952,  accuracy=0.812\n",
      "Epoch  0,  step= 332,  loss=1.256,  accuracy=0.781\n",
      "Epoch  0,  step= 333,  loss=1.153,  accuracy=0.656\n",
      "Epoch  0,  step= 334,  loss=1.289,  accuracy=0.781\n",
      "Epoch  0,  step= 335,  loss=2.202,  accuracy=0.688\n",
      "Epoch  0,  step= 336,  loss=1.537,  accuracy=0.594\n",
      "Epoch  0,  step= 337,  loss=1.173,  accuracy=0.750\n",
      "Epoch  0,  step= 338,  loss=1.002,  accuracy=0.844\n",
      "Epoch  0,  step= 339,  loss=1.770,  accuracy=0.656\n",
      "Epoch  0,  step= 340,  loss=0.575,  accuracy=0.875\n",
      "Epoch  0,  step= 341,  loss=1.290,  accuracy=0.719\n",
      "Epoch  0,  step= 342,  loss=0.505,  accuracy=0.844\n",
      "Epoch  0,  step= 343,  loss=1.206,  accuracy=0.875\n",
      "Epoch  0,  step= 344,  loss=1.346,  accuracy=0.812\n",
      "Epoch  0,  step= 345,  loss=1.379,  accuracy=0.781\n",
      "Epoch  0,  step= 346,  loss=0.916,  accuracy=0.625\n",
      "Epoch  0,  step= 347,  loss=1.359,  accuracy=0.625\n",
      "Epoch  0,  step= 348,  loss=1.058,  accuracy=0.750\n",
      "Epoch  0,  step= 349,  loss=0.523,  accuracy=0.812\n",
      "Epoch  0,  step= 350,  loss=1.746,  accuracy=0.719\n",
      "Epoch  0,  step= 351,  loss=1.835,  accuracy=0.656\n",
      "Epoch  0,  step= 352,  loss=0.923,  accuracy=0.781\n",
      "Epoch  0,  step= 353,  loss=1.026,  accuracy=0.812\n",
      "Epoch  0,  step= 354,  loss=0.561,  accuracy=0.812\n",
      "Epoch  0,  step= 355,  loss=2.000,  accuracy=0.719\n",
      "Epoch  0,  step= 356,  loss=1.561,  accuracy=0.812\n",
      "Epoch  0,  step= 357,  loss=0.384,  accuracy=0.875\n",
      "Epoch  0,  step= 358,  loss=0.848,  accuracy=0.844\n",
      "Epoch  0,  step= 359,  loss=2.094,  accuracy=0.750\n",
      "Epoch  0,  step= 360,  loss=0.598,  accuracy=0.906\n",
      "Epoch  0,  step= 361,  loss=2.270,  accuracy=0.656\n",
      "Epoch  0,  step= 362,  loss=2.308,  accuracy=0.656\n",
      "Epoch  0,  step= 363,  loss=2.259,  accuracy=0.562\n",
      "Epoch  0,  step= 364,  loss=2.417,  accuracy=0.688\n",
      "Epoch  0,  step= 365,  loss=2.721,  accuracy=0.625\n",
      "Epoch  0,  step= 366,  loss=1.130,  accuracy=0.781\n",
      "Epoch  0,  step= 367,  loss=1.373,  accuracy=0.750\n",
      "Epoch  0,  step= 368,  loss=2.062,  accuracy=0.656\n",
      "Epoch  0,  step= 369,  loss=1.137,  accuracy=0.750\n",
      "Epoch  0,  step= 370,  loss=1.524,  accuracy=0.750\n",
      "Epoch  0,  step= 371,  loss=1.008,  accuracy=0.781\n",
      "Epoch  0,  step= 372,  loss=0.756,  accuracy=0.781\n",
      "Epoch  0,  step= 373,  loss=1.108,  accuracy=0.719\n",
      "Epoch  0,  step= 374,  loss=1.707,  accuracy=0.688\n",
      "Epoch  0,  step= 375,  loss=1.292,  accuracy=0.750\n",
      "Epoch  0,  step= 376,  loss=1.764,  accuracy=0.656\n",
      "Epoch  0,  step= 377,  loss=1.327,  accuracy=0.688\n",
      "Epoch  0,  step= 378,  loss=2.456,  accuracy=0.562\n",
      "Epoch  0,  step= 379,  loss=0.978,  accuracy=0.750\n",
      "Epoch  0,  step= 380,  loss=0.809,  accuracy=0.688\n",
      "Epoch  0,  step= 381,  loss=2.219,  accuracy=0.625\n",
      "Epoch  0,  step= 382,  loss=1.476,  accuracy=0.781\n",
      "Epoch  0,  step= 383,  loss=2.409,  accuracy=0.625\n",
      "Epoch  0,  step= 384,  loss=1.192,  accuracy=0.625\n",
      "Epoch  0,  step= 385,  loss=1.209,  accuracy=0.688\n",
      "Epoch  0,  step= 386,  loss=1.801,  accuracy=0.719\n",
      "Epoch  0,  step= 387,  loss=0.948,  accuracy=0.750\n",
      "Epoch  0,  step= 388,  loss=2.060,  accuracy=0.562\n",
      "Epoch  0,  step= 389,  loss=1.905,  accuracy=0.688\n",
      "Epoch  0,  step= 390,  loss=2.340,  accuracy=0.656\n",
      "Epoch  0,  step= 391,  loss=1.365,  accuracy=0.688\n",
      "Epoch  0,  step= 392,  loss=2.871,  accuracy=0.594\n",
      "Epoch  0,  step= 393,  loss=1.996,  accuracy=0.719\n",
      "Epoch  0,  step= 394,  loss=1.362,  accuracy=0.688\n",
      "Epoch  0,  step= 395,  loss=2.359,  accuracy=0.688\n",
      "Epoch  0,  step= 396,  loss=3.289,  accuracy=0.656\n",
      "Epoch  0,  step= 397,  loss=0.569,  accuracy=0.812\n",
      "Epoch  0,  step= 398,  loss=2.141,  accuracy=0.656\n",
      "Epoch  0,  step= 399,  loss=2.090,  accuracy=0.594\n",
      "Epoch  0,  step= 400,  loss=1.020,  accuracy=0.688\n",
      "Epoch  0,  step= 401,  loss=2.143,  accuracy=0.719\n",
      "Epoch  0,  step= 402,  loss=2.151,  accuracy=0.625\n",
      "Epoch  0,  step= 403,  loss=1.533,  accuracy=0.719\n",
      "Epoch  0,  step= 404,  loss=2.113,  accuracy=0.688\n",
      "Epoch  0,  step= 405,  loss=1.761,  accuracy=0.594\n",
      "Epoch  0,  step= 406,  loss=2.959,  accuracy=0.469\n",
      "Epoch  0,  step= 407,  loss=2.112,  accuracy=0.688\n",
      "Epoch  0,  step= 408,  loss=3.258,  accuracy=0.469\n",
      "Epoch  0,  step= 409,  loss=2.286,  accuracy=0.688\n",
      "Epoch  0,  step= 410,  loss=1.453,  accuracy=0.719\n",
      "Epoch  0,  step= 411,  loss=1.100,  accuracy=0.656\n",
      "Epoch  0,  step= 412,  loss=1.935,  accuracy=0.750\n",
      "Epoch  0,  step= 413,  loss=0.698,  accuracy=0.781\n",
      "Epoch  0,  step= 414,  loss=2.199,  accuracy=0.688\n",
      "Epoch  0,  step= 415,  loss=1.573,  accuracy=0.656\n",
      "Epoch  0,  step= 416,  loss=1.597,  accuracy=0.781\n",
      "Epoch  0,  step= 417,  loss=2.400,  accuracy=0.688\n",
      "Epoch  0,  step= 418,  loss=1.947,  accuracy=0.656\n",
      "Epoch  0,  step= 419,  loss=0.954,  accuracy=0.781\n",
      "Epoch  0,  step= 420,  loss=0.969,  accuracy=0.750\n",
      "Epoch  0,  step= 421,  loss=0.663,  accuracy=0.906\n",
      "Epoch  0,  step= 422,  loss=1.611,  accuracy=0.719\n",
      "Epoch  0,  step= 423,  loss=0.929,  accuracy=0.875\n",
      "Epoch  0,  step= 424,  loss=0.483,  accuracy=0.844\n",
      "Epoch  0,  step= 425,  loss=1.658,  accuracy=0.719\n",
      "Epoch  0,  step= 426,  loss=2.921,  accuracy=0.594\n",
      "Epoch  0,  step= 427,  loss=2.121,  accuracy=0.625\n",
      "Epoch  0,  step= 428,  loss=1.960,  accuracy=0.625\n",
      "Epoch  0,  step= 429,  loss=1.473,  accuracy=0.688\n",
      "Epoch  0,  step= 430,  loss=1.193,  accuracy=0.781\n",
      "Epoch  0,  step= 431,  loss=1.044,  accuracy=0.812\n",
      "Epoch  0,  step= 432,  loss=2.266,  accuracy=0.625\n",
      "Epoch  0,  step= 433,  loss=1.802,  accuracy=0.594\n",
      "Epoch  0,  step= 434,  loss=0.926,  accuracy=0.688\n",
      "Epoch  0,  step= 435,  loss=2.141,  accuracy=0.781\n",
      "Epoch  0,  step= 436,  loss=0.867,  accuracy=0.781\n",
      "Epoch  0,  step= 437,  loss=2.416,  accuracy=0.594\n",
      "Epoch  0,  step= 438,  loss=1.445,  accuracy=0.719\n",
      "Epoch  0,  step= 439,  loss=2.749,  accuracy=0.688\n",
      "Epoch  0,  step= 440,  loss=0.745,  accuracy=0.844\n",
      "Epoch  0,  step= 441,  loss=2.762,  accuracy=0.594\n",
      "Epoch  0,  step= 442,  loss=1.177,  accuracy=0.844\n",
      "Epoch  0,  step= 443,  loss=1.510,  accuracy=0.719\n",
      "Epoch  0,  step= 444,  loss=1.933,  accuracy=0.656\n",
      "Epoch  0,  step= 445,  loss=2.249,  accuracy=0.625\n",
      "Epoch  0,  step= 446,  loss=1.149,  accuracy=0.750\n",
      "Epoch  0,  step= 447,  loss=1.349,  accuracy=0.781\n",
      "Epoch  0,  step= 448,  loss=3.013,  accuracy=0.594\n",
      "Epoch  0,  step= 449,  loss=2.472,  accuracy=0.562\n",
      "Epoch  0,  step= 450,  loss=1.450,  accuracy=0.688\n",
      "Epoch  0,  step= 451,  loss=0.929,  accuracy=0.781\n",
      "Epoch  0,  step= 452,  loss=1.673,  accuracy=0.719\n",
      "Epoch  0,  step= 453,  loss=2.661,  accuracy=0.656\n",
      "Epoch  0,  step= 454,  loss=1.988,  accuracy=0.656\n",
      "Epoch  0,  step= 455,  loss=1.047,  accuracy=0.781\n",
      "Epoch  0,  step= 456,  loss=2.121,  accuracy=0.594\n",
      "Epoch  0,  step= 457,  loss=2.504,  accuracy=0.594\n",
      "Epoch  0,  step= 458,  loss=1.980,  accuracy=0.594\n",
      "Epoch  0,  step= 459,  loss=2.553,  accuracy=0.531\n",
      "Epoch  0,  step= 460,  loss=1.995,  accuracy=0.531\n",
      "Epoch  0,  step= 461,  loss=2.819,  accuracy=0.594\n",
      "Epoch  0,  step= 462,  loss=2.116,  accuracy=0.656\n",
      "Epoch  0,  step= 463,  loss=2.044,  accuracy=0.656\n",
      "Epoch  0,  step= 464,  loss=0.818,  accuracy=0.906\n",
      "Epoch  0,  step= 465,  loss=1.505,  accuracy=0.688\n",
      "Epoch  0,  step= 466,  loss=1.071,  accuracy=0.844\n",
      "Epoch  0,  step= 467,  loss=0.945,  accuracy=0.781\n",
      "Epoch  0,  step= 468,  loss=1.021,  accuracy=0.781\n",
      "Epoch  0,  step= 469,  loss=0.740,  accuracy=0.812\n",
      "Epoch  0,  step= 470,  loss=1.366,  accuracy=0.750\n",
      "Epoch  0,  step= 471,  loss=0.612,  accuracy=0.844\n",
      "Epoch  0,  step= 472,  loss=2.723,  accuracy=0.625\n",
      "Epoch  0,  step= 473,  loss=2.008,  accuracy=0.812\n",
      "Epoch  0,  step= 474,  loss=1.740,  accuracy=0.750\n",
      "Epoch  0,  step= 475,  loss=1.656,  accuracy=0.656\n",
      "Epoch  0,  step= 476,  loss=2.131,  accuracy=0.688\n",
      "Epoch  0,  step= 477,  loss=1.439,  accuracy=0.812\n",
      "Epoch  0,  step= 478,  loss=0.795,  accuracy=0.719\n",
      "Epoch  0,  step= 479,  loss=1.464,  accuracy=0.750\n",
      "Epoch  0,  step= 480,  loss=0.878,  accuracy=0.719\n",
      "Epoch  0,  step= 481,  loss=1.276,  accuracy=0.844\n",
      "Epoch  0,  step= 482,  loss=1.475,  accuracy=0.719\n",
      "Epoch  0,  step= 483,  loss=0.710,  accuracy=0.781\n",
      "Epoch  0,  step= 484,  loss=1.571,  accuracy=0.656\n",
      "Epoch  0,  step= 485,  loss=1.459,  accuracy=0.688\n",
      "Epoch  0,  step= 486,  loss=0.826,  accuracy=0.781\n",
      "Epoch  0,  step= 487,  loss=1.164,  accuracy=0.781\n",
      "Epoch  0,  step= 488,  loss=0.865,  accuracy=0.844\n",
      "Epoch  0,  step= 489,  loss=1.181,  accuracy=0.781\n",
      "Epoch  0,  step= 490,  loss=0.844,  accuracy=0.812\n",
      "Epoch  0,  step= 491,  loss=2.873,  accuracy=0.625\n",
      "Epoch  0,  step= 492,  loss=2.531,  accuracy=0.594\n",
      "Epoch  0,  step= 493,  loss=1.481,  accuracy=0.719\n",
      "Epoch  0,  step= 494,  loss=2.041,  accuracy=0.594\n",
      "Epoch  0,  step= 495,  loss=2.405,  accuracy=0.719\n",
      "Epoch  0,  step= 496,  loss=2.111,  accuracy=0.781\n",
      "Epoch  0,  step= 497,  loss=1.005,  accuracy=0.844\n",
      "Epoch  0,  step= 498,  loss=1.211,  accuracy=0.750\n",
      "Epoch  0,  step= 499,  loss=1.996,  accuracy=0.719\n",
      "Epoch  0,  step= 500,  loss=1.544,  accuracy=0.781\n",
      "Epoch  0,  step= 501,  loss=2.143,  accuracy=0.625\n",
      "Epoch  0,  step= 502,  loss=1.224,  accuracy=0.750\n",
      "Epoch  0,  step= 503,  loss=2.019,  accuracy=0.719\n",
      "Epoch  0,  step= 504,  loss=0.745,  accuracy=0.750\n",
      "Epoch  0,  step= 505,  loss=1.330,  accuracy=0.656\n",
      "Epoch  0,  step= 506,  loss=2.367,  accuracy=0.625\n",
      "Epoch  0,  step= 507,  loss=0.326,  accuracy=0.969\n",
      "Epoch  0,  step= 508,  loss=1.596,  accuracy=0.781\n",
      "Epoch  0,  step= 509,  loss=0.777,  accuracy=0.812\n",
      "Epoch  0,  step= 510,  loss=1.014,  accuracy=0.750\n",
      "Epoch  0,  step= 511,  loss=1.070,  accuracy=0.781\n",
      "Epoch  0,  step= 512,  loss=0.902,  accuracy=0.781\n",
      "Epoch  0,  step= 513,  loss=1.482,  accuracy=0.719\n",
      "Epoch  0,  step= 514,  loss=0.738,  accuracy=0.812\n",
      "Epoch  0,  step= 515,  loss=0.756,  accuracy=0.750\n",
      "Epoch  0,  step= 516,  loss=1.055,  accuracy=0.844\n",
      "Epoch  0,  step= 517,  loss=1.185,  accuracy=0.750\n",
      "Epoch  0,  step= 518,  loss=0.976,  accuracy=0.812\n",
      "Epoch  0,  step= 519,  loss=0.982,  accuracy=0.844\n",
      "Epoch  0,  step= 520,  loss=1.039,  accuracy=0.844\n",
      "Epoch  0,  step= 521,  loss=1.966,  accuracy=0.562\n",
      "Epoch  0,  step= 522,  loss=0.852,  accuracy=0.844\n",
      "Epoch  0,  step= 523,  loss=1.985,  accuracy=0.656\n",
      "Epoch  0,  step= 524,  loss=2.914,  accuracy=0.562\n",
      "Epoch  0,  step= 525,  loss=1.733,  accuracy=0.688\n",
      "Epoch  0,  step= 526,  loss=1.317,  accuracy=0.781\n",
      "Epoch  0,  step= 527,  loss=1.314,  accuracy=0.750\n",
      "Epoch  0,  step= 528,  loss=1.224,  accuracy=0.812\n",
      "Epoch  0,  step= 529,  loss=1.524,  accuracy=0.750\n",
      "Epoch  0,  step= 530,  loss=1.986,  accuracy=0.688\n",
      "Epoch  0,  step= 531,  loss=1.892,  accuracy=0.594\n",
      "Epoch  0,  step= 532,  loss=1.286,  accuracy=0.875\n",
      "Epoch  0,  step= 533,  loss=0.650,  accuracy=0.812\n",
      "Epoch  0,  step= 534,  loss=1.414,  accuracy=0.688\n",
      "Epoch  0,  step= 535,  loss=1.133,  accuracy=0.781\n",
      "Epoch  0,  step= 536,  loss=1.327,  accuracy=0.688\n",
      "Epoch  0,  step= 537,  loss=1.826,  accuracy=0.781\n",
      "Epoch  0,  step= 538,  loss=0.830,  accuracy=0.812\n",
      "Epoch  0,  step= 539,  loss=0.798,  accuracy=0.812\n",
      "Epoch  0,  step= 540,  loss=0.562,  accuracy=0.812\n",
      "Epoch  0,  step= 541,  loss=1.177,  accuracy=0.750\n",
      "Epoch  0,  step= 542,  loss=0.772,  accuracy=0.750\n",
      "Epoch  0,  step= 543,  loss=1.257,  accuracy=0.656\n",
      "Epoch  0,  step= 544,  loss=0.179,  accuracy=0.938\n",
      "Epoch  0,  step= 545,  loss=1.302,  accuracy=0.719\n",
      "Epoch  0,  step= 546,  loss=1.357,  accuracy=0.781\n",
      "Epoch  0,  step= 547,  loss=1.125,  accuracy=0.719\n",
      "Epoch  0,  step= 548,  loss=2.583,  accuracy=0.688\n",
      "Epoch  0,  step= 549,  loss=0.986,  accuracy=0.781\n",
      "Epoch  0,  step= 550,  loss=1.067,  accuracy=0.719\n",
      "Epoch  0,  step= 551,  loss=2.429,  accuracy=0.625\n",
      "Epoch  0,  step= 552,  loss=1.224,  accuracy=0.625\n",
      "Epoch  0,  step= 553,  loss=2.180,  accuracy=0.625\n",
      "Epoch  0,  step= 554,  loss=0.866,  accuracy=0.781\n",
      "Epoch  0,  step= 555,  loss=1.466,  accuracy=0.812\n",
      "Epoch  0,  step= 556,  loss=2.388,  accuracy=0.594\n",
      "Epoch  0,  step= 557,  loss=1.636,  accuracy=0.719\n",
      "Epoch  0,  step= 558,  loss=1.170,  accuracy=0.688\n",
      "Epoch  0,  step= 559,  loss=1.368,  accuracy=0.719\n",
      "Epoch  0,  step= 560,  loss=0.740,  accuracy=0.812\n",
      "Epoch  0,  step= 561,  loss=0.788,  accuracy=0.844\n",
      "Epoch  0,  step= 562,  loss=0.884,  accuracy=0.656\n",
      "Epoch  0,  step= 563,  loss=0.527,  accuracy=0.844\n",
      "Epoch  0,  step= 564,  loss=1.629,  accuracy=0.719\n",
      "Epoch  0,  step= 565,  loss=1.665,  accuracy=0.750\n",
      "Epoch  0,  step= 566,  loss=0.634,  accuracy=0.875\n",
      "Epoch  0,  step= 567,  loss=0.718,  accuracy=0.844\n",
      "Epoch  0,  step= 568,  loss=0.081,  accuracy=1.000\n",
      "Epoch  0,  step= 569,  loss=0.529,  accuracy=0.812\n",
      "Epoch  0,  step= 570,  loss=1.343,  accuracy=0.750\n",
      "Epoch  0,  step= 571,  loss=0.307,  accuracy=0.906\n",
      "Epoch  0,  step= 572,  loss=0.662,  accuracy=0.781\n",
      "Epoch  0,  step= 573,  loss=0.853,  accuracy=0.719\n",
      "Epoch  0,  step= 574,  loss=0.784,  accuracy=0.750\n",
      "Epoch  0,  step= 575,  loss=1.278,  accuracy=0.656\n",
      "Epoch  0,  step= 576,  loss=0.978,  accuracy=0.812\n",
      "Epoch  0,  step= 577,  loss=2.267,  accuracy=0.750\n",
      "Epoch  0,  step= 578,  loss=1.761,  accuracy=0.688\n",
      "Epoch  0,  step= 579,  loss=0.949,  accuracy=0.781\n",
      "Epoch  0,  step= 580,  loss=1.894,  accuracy=0.656\n",
      "Epoch  0,  step= 581,  loss=0.759,  accuracy=0.875\n",
      "Epoch  0,  step= 582,  loss=1.486,  accuracy=0.812\n",
      "Epoch  0,  step= 583,  loss=0.625,  accuracy=0.844\n",
      "Epoch  0,  step= 584,  loss=0.728,  accuracy=0.781\n",
      "Epoch  0,  step= 585,  loss=0.955,  accuracy=0.750\n",
      "Epoch  0,  step= 586,  loss=0.900,  accuracy=0.844\n",
      "Epoch  0,  step= 587,  loss=0.604,  accuracy=0.844\n",
      "Epoch  0,  step= 588,  loss=1.165,  accuracy=0.781\n",
      "Epoch  0,  step= 589,  loss=0.896,  accuracy=0.812\n",
      "Epoch  0,  step= 590,  loss=0.770,  accuracy=0.812\n",
      "Epoch  0,  step= 591,  loss=1.184,  accuracy=0.781\n",
      "Epoch  0,  step= 592,  loss=0.403,  accuracy=0.875\n",
      "Epoch  0,  step= 593,  loss=1.605,  accuracy=0.688\n",
      "Epoch  0,  step= 594,  loss=1.028,  accuracy=0.719\n",
      "Epoch  0,  step= 595,  loss=0.478,  accuracy=0.875\n",
      "Epoch  0,  step= 596,  loss=1.361,  accuracy=0.812\n",
      "Epoch  0,  step= 597,  loss=1.543,  accuracy=0.781\n",
      "Epoch  0,  step= 598,  loss=1.303,  accuracy=0.719\n",
      "Epoch  0,  step= 599,  loss=1.414,  accuracy=0.750\n",
      "Epoch  0,  step= 600,  loss=1.549,  accuracy=0.719\n",
      "Epoch  0,  step= 601,  loss=1.165,  accuracy=0.812\n",
      "Epoch  0,  step= 602,  loss=2.082,  accuracy=0.750\n",
      "Epoch  0,  step= 603,  loss=1.366,  accuracy=0.719\n",
      "Epoch  0,  step= 604,  loss=1.245,  accuracy=0.688\n",
      "Epoch  0,  step= 605,  loss=1.498,  accuracy=0.719\n",
      "Epoch  0,  step= 606,  loss=1.086,  accuracy=0.812\n",
      "Epoch  0,  step= 607,  loss=0.593,  accuracy=0.812\n",
      "Epoch  0,  step= 608,  loss=0.868,  accuracy=0.812\n",
      "Epoch  0,  step= 609,  loss=1.361,  accuracy=0.844\n",
      "Epoch  0,  step= 610,  loss=1.769,  accuracy=0.656\n",
      "Epoch  0,  step= 611,  loss=1.344,  accuracy=0.656\n",
      "Epoch  0,  step= 612,  loss=1.276,  accuracy=0.688\n",
      "Epoch  0,  step= 613,  loss=1.263,  accuracy=0.750\n",
      "Epoch  0,  step= 614,  loss=0.802,  accuracy=0.844\n",
      "Epoch  0,  step= 615,  loss=0.594,  accuracy=0.812\n",
      "Epoch  0,  step= 616,  loss=0.182,  accuracy=0.906\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  0,  step= 617,  loss=0.186,  accuracy=0.969\n",
      "Epoch  0,  step= 618,  loss=0.734,  accuracy=0.875\n",
      "Epoch  0,  step= 619,  loss=2.576,  accuracy=0.688\n",
      "Epoch  0,  step= 620,  loss=1.358,  accuracy=0.812\n",
      "Epoch  0,  step= 621,  loss=0.287,  accuracy=0.906\n",
      "Epoch  0,  step= 622,  loss=1.841,  accuracy=0.688\n",
      "Epoch  0,  step= 623,  loss=1.433,  accuracy=0.812\n",
      "Epoch  0,  step= 624,  loss=0.930,  accuracy=0.781\n",
      "Epoch  0,  step= 625,  loss=1.465,  accuracy=0.750\n",
      "Epoch  0,  step= 626,  loss=1.952,  accuracy=0.594\n",
      "Epoch  0,  step= 627,  loss=1.643,  accuracy=0.688\n",
      "Epoch  0,  step= 628,  loss=1.003,  accuracy=0.688\n",
      "Epoch  0,  step= 629,  loss=0.956,  accuracy=0.781\n",
      "Epoch  0,  step= 630,  loss=1.742,  accuracy=0.719\n",
      "Epoch  0,  step= 631,  loss=1.046,  accuracy=0.750\n",
      "Epoch  0,  step= 632,  loss=0.856,  accuracy=0.875\n",
      "Epoch  0,  step= 633,  loss=1.023,  accuracy=0.719\n",
      "Epoch  0,  step= 634,  loss=0.832,  accuracy=0.812\n",
      "Epoch  0,  step= 635,  loss=1.147,  accuracy=0.812\n",
      "Epoch  0,  step= 636,  loss=0.789,  accuracy=0.875\n",
      "Epoch  0,  step= 637,  loss=1.114,  accuracy=0.812\n",
      "Epoch  0,  step= 638,  loss=0.541,  accuracy=0.875\n",
      "Epoch  0,  step= 639,  loss=0.567,  accuracy=0.844\n",
      "Epoch  0,  step= 640,  loss=0.586,  accuracy=0.906\n",
      "Epoch  0,  step= 641,  loss=1.103,  accuracy=0.875\n",
      "Epoch  0,  step= 642,  loss=2.178,  accuracy=0.812\n",
      "Epoch  0,  step= 643,  loss=0.933,  accuracy=0.812\n",
      "Epoch  0,  step= 644,  loss=1.288,  accuracy=0.750\n",
      "Epoch  0,  step= 645,  loss=1.411,  accuracy=0.750\n",
      "Epoch  0,  step= 646,  loss=0.247,  accuracy=0.969\n",
      "Epoch  0,  step= 647,  loss=1.743,  accuracy=0.781\n",
      "Epoch  0,  step= 648,  loss=0.966,  accuracy=0.844\n",
      "Epoch  0,  step= 649,  loss=1.962,  accuracy=0.750\n",
      "Epoch  0,  step= 650,  loss=1.518,  accuracy=0.781\n",
      "Epoch  0,  step= 651,  loss=1.356,  accuracy=0.688\n",
      "Epoch  0,  step= 652,  loss=1.564,  accuracy=0.812\n",
      "Epoch  0,  step= 653,  loss=1.362,  accuracy=0.781\n",
      "Epoch  0,  step= 654,  loss=1.405,  accuracy=0.688\n",
      "Epoch  0,  step= 655,  loss=1.815,  accuracy=0.719\n",
      "Epoch  0,  step= 656,  loss=1.247,  accuracy=0.719\n",
      "Epoch  0,  step= 657,  loss=0.714,  accuracy=0.781\n",
      "Epoch  0,  step= 658,  loss=0.719,  accuracy=0.875\n",
      "Epoch  0,  step= 659,  loss=0.867,  accuracy=0.812\n",
      "Epoch  0,  step= 660,  loss=2.290,  accuracy=0.781\n",
      "Epoch  0,  step= 661,  loss=0.648,  accuracy=0.875\n",
      "Epoch  0,  step= 662,  loss=2.028,  accuracy=0.750\n",
      "Epoch  0,  step= 663,  loss=0.411,  accuracy=0.906\n",
      "Epoch  0,  step= 664,  loss=0.592,  accuracy=0.844\n",
      "Epoch  0,  step= 665,  loss=1.594,  accuracy=0.750\n",
      "Epoch  0,  step= 666,  loss=1.113,  accuracy=0.812\n",
      "Epoch  0,  step= 667,  loss=1.431,  accuracy=0.812\n",
      "Epoch  0,  step= 668,  loss=1.103,  accuracy=0.750\n",
      "Epoch  0,  step= 669,  loss=1.654,  accuracy=0.781\n",
      "Epoch  0,  step= 670,  loss=0.445,  accuracy=0.812\n",
      "Epoch  0,  step= 671,  loss=0.604,  accuracy=0.875\n",
      "Epoch  0,  step= 672,  loss=0.849,  accuracy=0.781\n",
      "Epoch  0,  step= 673,  loss=0.950,  accuracy=0.719\n",
      "Epoch  0,  step= 674,  loss=1.393,  accuracy=0.781\n",
      "Epoch  0,  step= 675,  loss=1.632,  accuracy=0.750\n",
      "Epoch  0,  step= 676,  loss=0.830,  accuracy=0.844\n",
      "Epoch  0,  step= 677,  loss=0.823,  accuracy=0.844\n",
      "Epoch  0,  step= 678,  loss=1.223,  accuracy=0.750\n",
      "Epoch  0,  step= 679,  loss=0.845,  accuracy=0.719\n",
      "Epoch  0,  step= 680,  loss=1.023,  accuracy=0.875\n",
      "Epoch  0,  step= 681,  loss=0.119,  accuracy=0.938\n",
      "Epoch  0,  step= 682,  loss=0.595,  accuracy=0.875\n",
      "Epoch  0,  step= 683,  loss=0.498,  accuracy=0.938\n",
      "Epoch  0,  step= 684,  loss=1.114,  accuracy=0.719\n",
      "Epoch  0,  step= 685,  loss=0.879,  accuracy=0.781\n",
      "Epoch  0,  step= 686,  loss=0.246,  accuracy=0.906\n",
      "Epoch  0,  step= 687,  loss=0.296,  accuracy=0.938\n",
      "Epoch  0,  step= 688,  loss=1.309,  accuracy=0.688\n",
      "Epoch  0,  step= 689,  loss=1.079,  accuracy=0.750\n",
      "Epoch  0,  step= 690,  loss=1.200,  accuracy=0.781\n",
      "Epoch  0,  step= 691,  loss=1.330,  accuracy=0.719\n",
      "Epoch  0,  step= 692,  loss=1.194,  accuracy=0.688\n",
      "Epoch  0,  step= 693,  loss=1.607,  accuracy=0.719\n",
      "Epoch  0,  step= 694,  loss=2.844,  accuracy=0.750\n",
      "Epoch  0,  step= 695,  loss=0.949,  accuracy=0.750\n",
      "Epoch  0,  step= 696,  loss=1.800,  accuracy=0.719\n",
      "Epoch  0,  step= 697,  loss=0.522,  accuracy=0.875\n",
      "Epoch  0,  step= 698,  loss=0.555,  accuracy=0.875\n",
      "Epoch  0,  step= 699,  loss=0.616,  accuracy=0.844\n",
      "Epoch  0,  step= 700,  loss=0.564,  accuracy=0.781\n",
      "Epoch  0,  step= 701,  loss=0.219,  accuracy=0.938\n",
      "Epoch  0,  step= 702,  loss=2.049,  accuracy=0.750\n",
      "Epoch  0,  step= 703,  loss=2.510,  accuracy=0.469\n",
      "Epoch  0,  step= 704,  loss=1.941,  accuracy=0.656\n",
      "Epoch  0,  step= 705,  loss=2.017,  accuracy=0.594\n",
      "Epoch  0,  step= 706,  loss=1.535,  accuracy=0.719\n",
      "Epoch  0,  step= 707,  loss=1.335,  accuracy=0.781\n",
      "Epoch  0,  step= 708,  loss=1.072,  accuracy=0.719\n",
      "Epoch  0,  step= 709,  loss=1.119,  accuracy=0.844\n",
      "Epoch  0,  step= 710,  loss=0.574,  accuracy=0.812\n",
      "Epoch  0,  step= 711,  loss=1.354,  accuracy=0.688\n",
      "Epoch  0,  step= 712,  loss=1.590,  accuracy=0.719\n",
      "Epoch  0,  step= 713,  loss=1.082,  accuracy=0.750\n",
      "Epoch  0,  step= 714,  loss=0.247,  accuracy=0.906\n",
      "Epoch  0,  step= 715,  loss=0.987,  accuracy=0.719\n",
      "Epoch  0,  step= 716,  loss=0.750,  accuracy=0.750\n",
      "Epoch  0,  step= 717,  loss=0.255,  accuracy=0.875\n",
      "Epoch  0,  step= 718,  loss=0.257,  accuracy=0.875\n",
      "Epoch  0,  step= 719,  loss=0.881,  accuracy=0.875\n",
      "Epoch  0,  step= 720,  loss=0.857,  accuracy=0.812\n",
      "Epoch  0,  step= 721,  loss=1.341,  accuracy=0.719\n",
      "Epoch  0,  step= 722,  loss=0.637,  accuracy=0.906\n",
      "Epoch  0,  step= 723,  loss=1.946,  accuracy=0.719\n",
      "Epoch  0,  step= 724,  loss=0.996,  accuracy=0.812\n",
      "Epoch  0,  step= 725,  loss=0.644,  accuracy=0.844\n",
      "Epoch  0,  step= 726,  loss=1.081,  accuracy=0.812\n",
      "Epoch  0,  step= 727,  loss=0.773,  accuracy=0.781\n",
      "Epoch  0,  step= 728,  loss=0.803,  accuracy=0.812\n",
      "Epoch  0,  step= 729,  loss=0.703,  accuracy=0.844\n",
      "Epoch  0,  step= 730,  loss=1.257,  accuracy=0.781\n",
      "Epoch  0,  step= 731,  loss=1.099,  accuracy=0.750\n",
      "Epoch  0,  step= 732,  loss=1.026,  accuracy=0.812\n",
      "Epoch  0,  step= 733,  loss=1.204,  accuracy=0.781\n",
      "Epoch  0,  step= 734,  loss=1.045,  accuracy=0.844\n",
      "Epoch  0,  step= 735,  loss=0.264,  accuracy=0.906\n",
      "Epoch  0,  step= 736,  loss=0.550,  accuracy=0.844\n",
      "Epoch  0,  step= 737,  loss=1.040,  accuracy=0.812\n",
      "Epoch  0,  step= 738,  loss=0.830,  accuracy=0.750\n",
      "Epoch  0,  step= 739,  loss=0.725,  accuracy=0.781\n",
      "Epoch  0,  step= 740,  loss=1.401,  accuracy=0.656\n",
      "Epoch  0,  step= 741,  loss=2.387,  accuracy=0.625\n",
      "Epoch  0,  step= 742,  loss=0.952,  accuracy=0.750\n",
      "Epoch  0,  step= 743,  loss=0.814,  accuracy=0.781\n",
      "Epoch  0,  step= 744,  loss=2.052,  accuracy=0.719\n",
      "Epoch  0,  step= 745,  loss=1.171,  accuracy=0.750\n",
      "Epoch  0,  step= 746,  loss=1.283,  accuracy=0.656\n",
      "Epoch  0,  step= 747,  loss=1.035,  accuracy=0.812\n",
      "Epoch  0,  step= 748,  loss=1.028,  accuracy=0.875\n",
      "Epoch  0,  step= 749,  loss=1.034,  accuracy=0.844\n",
      "Epoch  0,  step= 750,  loss=0.965,  accuracy=0.781\n",
      "Epoch  0,  step= 751,  loss=0.785,  accuracy=0.781\n",
      "Epoch  0,  step= 752,  loss=0.178,  accuracy=0.875\n",
      "Epoch  0,  step= 753,  loss=1.375,  accuracy=0.594\n",
      "Epoch  0,  step= 754,  loss=0.326,  accuracy=0.906\n",
      "Epoch  0,  step= 755,  loss=1.304,  accuracy=0.875\n",
      "Epoch  0,  step= 756,  loss=1.325,  accuracy=0.781\n",
      "Epoch  0,  step= 757,  loss=1.298,  accuracy=0.719\n",
      "Epoch  0,  step= 758,  loss=2.618,  accuracy=0.719\n",
      "Epoch  0,  step= 759,  loss=0.985,  accuracy=0.844\n",
      "Epoch  0,  step= 760,  loss=0.447,  accuracy=0.844\n",
      "Epoch  0,  step= 761,  loss=2.143,  accuracy=0.781\n",
      "Epoch  0,  step= 762,  loss=1.123,  accuracy=0.750\n",
      "Epoch  0,  step= 763,  loss=0.984,  accuracy=0.812\n",
      "Epoch  0,  step= 764,  loss=1.096,  accuracy=0.750\n",
      "Epoch  0,  step= 765,  loss=1.404,  accuracy=0.750\n",
      "Epoch  0,  step= 766,  loss=1.608,  accuracy=0.656\n",
      "Epoch  0,  step= 767,  loss=0.609,  accuracy=0.812\n",
      "Epoch  0,  step= 768,  loss=1.301,  accuracy=0.750\n",
      "Epoch  0,  step= 769,  loss=1.992,  accuracy=0.719\n",
      "Epoch  0,  step= 770,  loss=1.478,  accuracy=0.781\n",
      "Epoch  0,  step= 771,  loss=0.807,  accuracy=0.781\n",
      "Epoch  0,  step= 772,  loss=1.764,  accuracy=0.625\n",
      "Epoch  0,  step= 773,  loss=1.473,  accuracy=0.750\n",
      "Epoch  0,  step= 774,  loss=1.145,  accuracy=0.781\n",
      "Epoch  0,  step= 775,  loss=1.851,  accuracy=0.688\n",
      "Epoch  0,  step= 776,  loss=0.460,  accuracy=0.844\n",
      "Epoch  0,  step= 777,  loss=1.453,  accuracy=0.750\n",
      "Epoch  0,  step= 778,  loss=0.910,  accuracy=0.781\n",
      "Epoch  0,  step= 779,  loss=1.152,  accuracy=0.844\n",
      "Epoch  0,  step= 780,  loss=1.223,  accuracy=0.781\n",
      "Epoch  0,  step= 781,  loss=0.615,  accuracy=0.812\n",
      "Epoch  0,  step= 782,  loss=0.457,  accuracy=0.906\n",
      "Epoch  0,  step= 783,  loss=0.829,  accuracy=0.844\n",
      "Epoch  0,  step= 784,  loss=1.318,  accuracy=0.844\n",
      "Epoch  0,  step= 785,  loss=1.312,  accuracy=0.656\n",
      "Epoch  0,  step= 786,  loss=1.674,  accuracy=0.625\n",
      "Epoch  0,  step= 787,  loss=1.969,  accuracy=0.750\n",
      "Epoch  0,  step= 788,  loss=0.897,  accuracy=0.844\n",
      "Epoch  0,  step= 789,  loss=1.102,  accuracy=0.812\n",
      "Epoch  0,  step= 790,  loss=2.083,  accuracy=0.625\n",
      "Epoch  0,  step= 791,  loss=1.375,  accuracy=0.750\n",
      "Epoch  0,  step= 792,  loss=0.193,  accuracy=0.906\n",
      "Epoch  0,  step= 793,  loss=0.630,  accuracy=0.875\n",
      "Epoch  0,  step= 794,  loss=0.118,  accuracy=0.906\n",
      "Epoch  0,  step= 795,  loss=1.567,  accuracy=0.719\n",
      "Epoch  0,  step= 796,  loss=1.398,  accuracy=0.750\n",
      "Epoch  0,  step= 797,  loss=0.433,  accuracy=0.906\n",
      "Epoch  0,  step= 798,  loss=1.163,  accuracy=0.812\n",
      "Epoch  0,  step= 799,  loss=0.731,  accuracy=0.750\n",
      "Epoch  0,  step= 800,  loss=0.529,  accuracy=0.781\n",
      "Epoch  0,  step= 801,  loss=0.937,  accuracy=0.812\n",
      "Epoch  0,  step= 802,  loss=1.557,  accuracy=0.812\n",
      "Epoch  0,  step= 803,  loss=0.573,  accuracy=0.812\n",
      "Epoch  0,  step= 804,  loss=1.379,  accuracy=0.781\n",
      "Epoch  0,  step= 805,  loss=0.862,  accuracy=0.781\n",
      "Epoch  0,  step= 806,  loss=1.300,  accuracy=0.781\n",
      "Epoch  0,  step= 807,  loss=0.747,  accuracy=0.875\n",
      "Epoch  0,  step= 808,  loss=0.375,  accuracy=0.875\n",
      "Epoch  0,  step= 809,  loss=0.687,  accuracy=0.812\n",
      "Epoch  0,  step= 810,  loss=2.028,  accuracy=0.562\n",
      "Epoch  0,  step= 811,  loss=1.587,  accuracy=0.719\n",
      "Epoch  0,  step= 812,  loss=0.352,  accuracy=0.906\n",
      "Epoch  0,  step= 813,  loss=0.878,  accuracy=0.781\n",
      "Epoch  0,  step= 814,  loss=1.327,  accuracy=0.688\n",
      "Epoch  0,  step= 815,  loss=0.211,  accuracy=0.875\n",
      "Epoch  0,  step= 816,  loss=0.480,  accuracy=0.875\n",
      "Epoch  0,  step= 817,  loss=0.611,  accuracy=0.844\n",
      "Epoch  0,  step= 818,  loss=0.516,  accuracy=0.906\n",
      "Epoch  0,  step= 819,  loss=1.144,  accuracy=0.844\n",
      "Epoch  0,  step= 820,  loss=0.729,  accuracy=0.906\n",
      "Epoch  0,  step= 821,  loss=1.442,  accuracy=0.844\n",
      "Epoch  0,  step= 822,  loss=0.405,  accuracy=0.844\n",
      "Epoch  0,  step= 823,  loss=1.140,  accuracy=0.781\n",
      "Epoch  0,  step= 824,  loss=1.456,  accuracy=0.844\n",
      "Epoch  0,  step= 825,  loss=1.175,  accuracy=0.781\n",
      "Epoch  0,  step= 826,  loss=1.489,  accuracy=0.719\n",
      "Epoch  0,  step= 827,  loss=0.913,  accuracy=0.750\n",
      "Epoch  0,  step= 828,  loss=0.994,  accuracy=0.625\n",
      "Epoch  0,  step= 829,  loss=1.219,  accuracy=0.750\n",
      "Epoch  0,  step= 830,  loss=0.923,  accuracy=0.844\n",
      "Epoch  0,  step= 831,  loss=2.168,  accuracy=0.656\n",
      "Epoch  0,  step= 832,  loss=1.855,  accuracy=0.719\n",
      "Epoch  0,  step= 833,  loss=0.577,  accuracy=0.875\n",
      "Epoch  0,  step= 834,  loss=1.029,  accuracy=0.750\n",
      "Epoch  0,  step= 835,  loss=2.482,  accuracy=0.656\n",
      "Epoch  0,  step= 836,  loss=1.046,  accuracy=0.688\n",
      "Epoch  0,  step= 837,  loss=1.082,  accuracy=0.781\n",
      "Epoch  0,  step= 838,  loss=0.597,  accuracy=0.812\n",
      "Epoch  0,  step= 839,  loss=1.488,  accuracy=0.781\n",
      "Epoch  0,  step= 840,  loss=1.280,  accuracy=0.781\n",
      "Epoch  0,  step= 841,  loss=0.757,  accuracy=0.844\n",
      "Epoch  0,  step= 842,  loss=0.576,  accuracy=0.812\n",
      "Epoch  0,  step= 843,  loss=0.523,  accuracy=0.844\n",
      "Epoch  0,  step= 844,  loss=0.772,  accuracy=0.844\n",
      "Epoch  0,  step= 845,  loss=0.614,  accuracy=0.781\n",
      "Epoch  0,  step= 846,  loss=1.090,  accuracy=0.750\n",
      "Epoch  0,  step= 847,  loss=0.653,  accuracy=0.688\n",
      "Epoch  0,  step= 848,  loss=1.357,  accuracy=0.875\n",
      "Epoch  0,  step= 849,  loss=2.456,  accuracy=0.656\n",
      "Epoch  0,  step= 850,  loss=1.924,  accuracy=0.656\n",
      "Epoch  0,  step= 851,  loss=0.691,  accuracy=0.750\n",
      "Epoch  0,  step= 852,  loss=0.518,  accuracy=0.906\n",
      "Epoch  0,  step= 853,  loss=1.157,  accuracy=0.781\n",
      "Epoch  0,  step= 854,  loss=0.739,  accuracy=0.844\n",
      "Epoch  0,  step= 855,  loss=0.877,  accuracy=0.750\n",
      "Epoch  0,  step= 856,  loss=0.857,  accuracy=0.812\n",
      "Epoch  0,  step= 857,  loss=1.359,  accuracy=0.719\n",
      "Epoch  0,  step= 858,  loss=0.811,  accuracy=0.812\n",
      "Epoch  0,  step= 859,  loss=1.481,  accuracy=0.781\n",
      "Epoch  0,  step= 860,  loss=0.656,  accuracy=0.875\n",
      "Epoch  0,  step= 861,  loss=0.148,  accuracy=0.938\n",
      "Epoch  0,  step= 862,  loss=0.504,  accuracy=0.812\n",
      "Epoch  0,  step= 863,  loss=0.973,  accuracy=0.750\n",
      "Epoch  0,  step= 864,  loss=0.273,  accuracy=0.906\n",
      "Epoch  0,  step= 865,  loss=1.499,  accuracy=0.781\n",
      "Epoch  0,  step= 866,  loss=0.958,  accuracy=0.844\n",
      "Epoch  0,  step= 867,  loss=0.292,  accuracy=0.906\n",
      "Epoch  0,  step= 868,  loss=1.090,  accuracy=0.594\n",
      "Epoch  0,  step= 869,  loss=0.998,  accuracy=0.875\n",
      "Epoch  0,  step= 870,  loss=0.574,  accuracy=0.875\n",
      "Epoch  0,  step= 871,  loss=1.772,  accuracy=0.688\n",
      "Epoch  0,  step= 872,  loss=0.390,  accuracy=0.844\n",
      "Epoch  0,  step= 873,  loss=1.156,  accuracy=0.750\n",
      "Epoch  0,  step= 874,  loss=0.724,  accuracy=0.812\n",
      "Epoch  0,  step= 875,  loss=1.465,  accuracy=0.781\n",
      "Epoch  0,  step= 876,  loss=0.420,  accuracy=0.938\n",
      "Epoch  0,  step= 877,  loss=1.333,  accuracy=0.812\n",
      "Epoch  0,  step= 878,  loss=0.669,  accuracy=0.875\n",
      "Epoch  0,  step= 879,  loss=1.292,  accuracy=0.750\n",
      "Epoch  0,  step= 880,  loss=0.634,  accuracy=0.875\n",
      "Epoch  0,  step= 881,  loss=1.517,  accuracy=0.750\n",
      "Epoch  0,  step= 882,  loss=1.104,  accuracy=0.875\n",
      "Epoch  0,  step= 883,  loss=0.722,  accuracy=0.781\n",
      "Epoch  0,  step= 884,  loss=1.207,  accuracy=0.844\n",
      "Epoch  0,  step= 885,  loss=0.385,  accuracy=0.844\n",
      "Epoch  0,  step= 886,  loss=2.393,  accuracy=0.750\n",
      "Epoch  0,  step= 887,  loss=0.715,  accuracy=0.906\n",
      "Epoch  0,  step= 888,  loss=0.368,  accuracy=0.875\n",
      "Epoch  0,  step= 889,  loss=0.670,  accuracy=0.812\n",
      "Epoch  0,  step= 890,  loss=1.483,  accuracy=0.750\n",
      "Epoch  0,  step= 891,  loss=1.515,  accuracy=0.750\n",
      "Epoch  0,  step= 892,  loss=0.382,  accuracy=0.906\n",
      "Epoch  0,  step= 893,  loss=0.860,  accuracy=0.719\n",
      "Epoch  0,  step= 894,  loss=1.932,  accuracy=0.750\n",
      "Epoch  0,  step= 895,  loss=2.379,  accuracy=0.688\n",
      "Epoch  0,  step= 896,  loss=0.283,  accuracy=0.906\n",
      "Epoch  0,  step= 897,  loss=1.154,  accuracy=0.750\n",
      "Epoch  0,  step= 898,  loss=0.425,  accuracy=0.875\n",
      "Epoch  0,  step= 899,  loss=1.045,  accuracy=0.781\n",
      "Epoch  0,  step= 900,  loss=0.766,  accuracy=0.812\n",
      "Epoch  0,  step= 901,  loss=0.950,  accuracy=0.781\n",
      "Epoch  0,  step= 902,  loss=0.575,  accuracy=0.875\n",
      "Epoch  0,  step= 903,  loss=0.880,  accuracy=0.781\n",
      "Epoch  0,  step= 904,  loss=1.384,  accuracy=0.656\n",
      "Epoch  0,  step= 905,  loss=0.724,  accuracy=0.906\n",
      "Epoch  0,  step= 906,  loss=1.122,  accuracy=0.750\n",
      "Epoch  0,  step= 907,  loss=0.519,  accuracy=0.844\n",
      "Epoch  0,  step= 908,  loss=1.433,  accuracy=0.812\n",
      "Epoch  0,  step= 909,  loss=1.388,  accuracy=0.688\n",
      "Epoch  0,  step= 910,  loss=2.207,  accuracy=0.656\n",
      "Epoch  0,  step= 911,  loss=1.525,  accuracy=0.594\n",
      "Epoch  0,  step= 912,  loss=1.252,  accuracy=0.688\n",
      "Epoch  0,  step= 913,  loss=0.627,  accuracy=0.750\n",
      "Epoch  0,  step= 914,  loss=0.637,  accuracy=0.844\n",
      "Epoch  0,  step= 915,  loss=1.348,  accuracy=0.688\n",
      "Epoch  0,  step= 916,  loss=1.542,  accuracy=0.781\n",
      "Epoch  0,  step= 917,  loss=0.522,  accuracy=0.875\n",
      "Epoch  0,  step= 918,  loss=0.411,  accuracy=0.906\n",
      "Epoch  0,  step= 919,  loss=0.810,  accuracy=0.844\n",
      "Epoch  0,  step= 920,  loss=0.737,  accuracy=0.781\n",
      "Epoch  0,  step= 921,  loss=1.009,  accuracy=0.750\n",
      "Epoch  0,  step= 922,  loss=0.475,  accuracy=0.938\n",
      "Epoch  0,  step= 923,  loss=0.960,  accuracy=0.656\n",
      "Epoch  0,  step= 924,  loss=1.137,  accuracy=0.812\n",
      "Epoch  0,  step= 925,  loss=0.754,  accuracy=0.781\n",
      "Epoch  0,  step= 926,  loss=1.508,  accuracy=0.625\n",
      "Epoch  0,  step= 927,  loss=1.228,  accuracy=0.812\n",
      "Epoch  0,  step= 928,  loss=1.867,  accuracy=0.688\n",
      "Epoch  0,  step= 929,  loss=0.680,  accuracy=0.875\n",
      "Epoch  0,  step= 930,  loss=0.721,  accuracy=0.750\n",
      "Epoch  0,  step= 931,  loss=0.590,  accuracy=0.812\n",
      "Epoch  0,  step= 932,  loss=2.065,  accuracy=0.688\n",
      "Epoch  0,  step= 933,  loss=1.909,  accuracy=0.656\n",
      "Epoch  0,  step= 934,  loss=1.517,  accuracy=0.688\n",
      "Epoch  0,  step= 935,  loss=1.864,  accuracy=0.656\n",
      "Epoch  0,  step= 936,  loss=1.995,  accuracy=0.688\n",
      "Epoch  0,  step= 937,  loss=0.532,  accuracy=0.844\n",
      "Epoch  0,  step= 938,  loss=1.209,  accuracy=0.781\n",
      "Epoch  0,  step= 939,  loss=1.170,  accuracy=0.688\n",
      "Epoch  0,  step= 940,  loss=1.533,  accuracy=0.688\n",
      "Epoch  0,  step= 941,  loss=1.662,  accuracy=0.688\n",
      "Epoch  0,  step= 942,  loss=1.786,  accuracy=0.688\n",
      "Epoch  0,  step= 943,  loss=1.437,  accuracy=0.781\n",
      "Epoch  0,  step= 944,  loss=0.704,  accuracy=0.844\n",
      "Epoch  0,  step= 945,  loss=0.627,  accuracy=0.875\n",
      "Epoch  0,  step= 946,  loss=0.074,  accuracy=0.938\n",
      "Epoch  0,  step= 947,  loss=1.135,  accuracy=0.719\n",
      "Epoch  0,  step= 948,  loss=0.882,  accuracy=0.812\n",
      "Epoch  0,  step= 949,  loss=0.860,  accuracy=0.875\n",
      "Epoch  0,  step= 950,  loss=0.675,  accuracy=0.812\n",
      "Epoch  0,  step= 951,  loss=1.210,  accuracy=0.688\n",
      "Epoch  0,  step= 952,  loss=0.895,  accuracy=0.812\n",
      "Epoch  0,  step= 953,  loss=1.840,  accuracy=0.594\n",
      "Epoch  0,  step= 954,  loss=1.441,  accuracy=0.750\n",
      "Epoch  0,  step= 955,  loss=0.626,  accuracy=0.906\n",
      "Epoch  0,  step= 956,  loss=2.322,  accuracy=0.625\n",
      "Epoch  0,  step= 957,  loss=0.290,  accuracy=0.906\n",
      "Epoch  0,  step= 958,  loss=1.046,  accuracy=0.750\n",
      "Epoch  0,  step= 959,  loss=1.374,  accuracy=0.719\n",
      "Epoch  0,  step= 960,  loss=1.149,  accuracy=0.750\n",
      "Epoch  0,  step= 961,  loss=0.417,  accuracy=0.938\n",
      "Epoch  0,  step= 962,  loss=0.432,  accuracy=0.812\n",
      "Epoch  0,  step= 963,  loss=1.173,  accuracy=0.812\n",
      "Epoch  0,  step= 964,  loss=1.140,  accuracy=0.750\n",
      "Epoch  0,  step= 965,  loss=2.187,  accuracy=0.688\n",
      "Epoch  0,  step= 966,  loss=0.447,  accuracy=0.906\n",
      "Epoch  0,  step= 967,  loss=0.689,  accuracy=0.844\n",
      "Epoch  0,  step= 968,  loss=0.788,  accuracy=0.812\n",
      "Epoch  0,  step= 969,  loss=1.346,  accuracy=0.750\n",
      "Epoch  0,  step= 970,  loss=0.739,  accuracy=0.844\n",
      "Epoch  0,  step= 971,  loss=1.284,  accuracy=0.781\n",
      "Epoch  0,  step= 972,  loss=1.684,  accuracy=0.750\n",
      "Epoch  0,  step= 973,  loss=0.696,  accuracy=0.844\n",
      "Epoch  0,  step= 974,  loss=1.167,  accuracy=0.812\n",
      "Epoch  0,  step= 975,  loss=0.899,  accuracy=0.781\n",
      "Epoch  0,  step= 976,  loss=0.492,  accuracy=0.844\n",
      "Epoch  0,  step= 977,  loss=2.438,  accuracy=0.625\n",
      "Epoch  0,  step= 978,  loss=1.555,  accuracy=0.625\n",
      "Epoch  0,  step= 979,  loss=1.260,  accuracy=0.719\n",
      "Epoch  0,  step= 980,  loss=1.790,  accuracy=0.656\n",
      "Epoch  0,  step= 981,  loss=1.440,  accuracy=0.781\n",
      "Epoch  0,  step= 982,  loss=1.244,  accuracy=0.812\n",
      "Epoch  0,  step= 983,  loss=1.221,  accuracy=0.719\n",
      "Epoch  0,  step= 984,  loss=0.872,  accuracy=0.844\n",
      "Epoch  0,  step= 985,  loss=0.763,  accuracy=0.781\n",
      "Epoch  0,  step= 986,  loss=0.806,  accuracy=0.875\n",
      "Epoch  0,  step= 987,  loss=0.919,  accuracy=0.812\n",
      "Epoch  0,  step= 988,  loss=1.368,  accuracy=0.812\n",
      "Epoch  0,  step= 989,  loss=2.459,  accuracy=0.656\n",
      "Epoch  0,  step= 990,  loss=2.018,  accuracy=0.594\n",
      "Epoch  0,  step= 991,  loss=1.402,  accuracy=0.719\n",
      "Epoch  0,  step= 992,  loss=0.911,  accuracy=0.844\n",
      "Epoch  0,  step= 993,  loss=0.652,  accuracy=0.781\n",
      "Epoch  0,  step= 994,  loss=0.316,  accuracy=0.844\n",
      "Epoch  0,  step= 995,  loss=1.331,  accuracy=0.812\n",
      "Epoch  0,  step= 996,  loss=0.806,  accuracy=0.812\n",
      "Epoch  0,  step= 997,  loss=0.646,  accuracy=0.875\n",
      "Epoch  0,  step= 998,  loss=1.477,  accuracy=0.719\n",
      "Epoch  0,  step= 999,  loss=0.505,  accuracy=0.906\n",
      "Epoch  0,  step=1000,  loss=0.846,  accuracy=0.812\n",
      "Epoch  0,  step=1001,  loss=0.788,  accuracy=0.844\n",
      "Epoch  0,  step=1002,  loss=0.791,  accuracy=0.906\n",
      "Epoch  0,  step=1003,  loss=0.460,  accuracy=0.750\n",
      "Epoch  0,  step=1004,  loss=1.582,  accuracy=0.625\n",
      "Epoch  0,  step=1005,  loss=1.175,  accuracy=0.781\n",
      "Epoch  0,  step=1006,  loss=0.495,  accuracy=0.875\n",
      "Epoch  0,  step=1007,  loss=0.991,  accuracy=0.781\n",
      "Epoch  0,  step=1008,  loss=1.141,  accuracy=0.750\n",
      "Epoch  0,  step=1009,  loss=1.216,  accuracy=0.844\n",
      "Epoch  0,  step=1010,  loss=1.860,  accuracy=0.656\n",
      "Epoch  0,  step=1011,  loss=1.228,  accuracy=0.781\n",
      "Epoch  0,  step=1012,  loss=1.105,  accuracy=0.781\n",
      "Epoch  0,  step=1013,  loss=1.891,  accuracy=0.656\n",
      "Epoch  0,  step=1014,  loss=1.509,  accuracy=0.750\n",
      "Epoch  0,  step=1015,  loss=1.083,  accuracy=0.688\n",
      "Epoch  0,  step=1016,  loss=0.589,  accuracy=0.812\n",
      "Epoch  0,  step=1017,  loss=1.070,  accuracy=0.719\n",
      "Epoch  0,  step=1018,  loss=0.653,  accuracy=0.875\n",
      "Epoch  0,  step=1019,  loss=1.144,  accuracy=0.750\n",
      "Epoch  0,  step=1020,  loss=1.081,  accuracy=0.812\n",
      "Epoch  0,  step=1021,  loss=0.495,  accuracy=0.812\n",
      "Epoch  0,  step=1022,  loss=0.991,  accuracy=0.750\n",
      "Epoch  0,  step=1023,  loss=0.582,  accuracy=0.875\n",
      "Epoch  0,  step=1024,  loss=1.316,  accuracy=0.750\n",
      "Epoch  0,  step=1025,  loss=1.089,  accuracy=0.812\n",
      "Epoch  0,  step=1026,  loss=0.919,  accuracy=0.812\n",
      "Epoch  0,  step=1027,  loss=0.688,  accuracy=0.844\n",
      "Epoch  0,  step=1028,  loss=1.129,  accuracy=0.844\n",
      "Epoch  0,  step=1029,  loss=0.696,  accuracy=0.875\n",
      "Epoch  0,  step=1030,  loss=0.164,  accuracy=0.938\n",
      "Epoch  0,  step=1031,  loss=0.349,  accuracy=0.906\n",
      "Epoch  0,  step=1032,  loss=1.150,  accuracy=0.625\n",
      "Epoch  0,  step=1033,  loss=0.766,  accuracy=0.875\n",
      "Epoch  0,  step=1034,  loss=0.483,  accuracy=0.875\n",
      "Epoch  0,  step=1035,  loss=0.812,  accuracy=0.781\n",
      "Epoch  0,  step=1036,  loss=0.513,  accuracy=0.812\n",
      "Epoch  0,  step=1037,  loss=1.531,  accuracy=0.750\n",
      "Epoch  0,  step=1038,  loss=1.113,  accuracy=0.844\n",
      "Epoch  0,  step=1039,  loss=0.510,  accuracy=0.875\n",
      "Epoch  0,  step=1040,  loss=1.231,  accuracy=0.812\n",
      "Epoch  0,  step=1041,  loss=0.690,  accuracy=0.781\n",
      "Epoch  0,  step=1042,  loss=1.665,  accuracy=0.688\n",
      "Epoch  0,  step=1043,  loss=0.774,  accuracy=0.844\n",
      "Epoch  0,  step=1044,  loss=1.014,  accuracy=0.812\n",
      "Epoch  0,  step=1045,  loss=1.451,  accuracy=0.688\n",
      "Epoch  0,  step=1046,  loss=0.774,  accuracy=0.844\n",
      "Epoch  0,  step=1047,  loss=0.801,  accuracy=0.781\n",
      "Epoch  0,  step=1048,  loss=0.200,  accuracy=0.906\n",
      "Epoch  0,  step=1049,  loss=1.331,  accuracy=0.750\n",
      "Epoch  0,  step=1050,  loss=0.717,  accuracy=0.812\n",
      "Epoch  0,  step=1051,  loss=0.470,  accuracy=0.906\n",
      "Epoch  0,  step=1052,  loss=0.814,  accuracy=0.750\n",
      "Epoch  0,  step=1053,  loss=0.507,  accuracy=0.844\n",
      "Epoch  0,  step=1054,  loss=0.510,  accuracy=0.906\n",
      "Epoch  0,  step=1055,  loss=1.869,  accuracy=0.656\n",
      "Epoch  0,  step=1056,  loss=0.794,  accuracy=0.750\n",
      "Epoch  0,  step=1057,  loss=0.566,  accuracy=0.875\n",
      "Epoch  0,  step=1058,  loss=0.576,  accuracy=0.875\n",
      "Epoch  0,  step=1059,  loss=0.534,  accuracy=0.875\n",
      "Epoch  0,  step=1060,  loss=0.421,  accuracy=0.906\n",
      "Epoch  0,  step=1061,  loss=0.467,  accuracy=0.875\n",
      "Epoch  0,  step=1062,  loss=0.547,  accuracy=0.875\n",
      "Epoch  0,  step=1063,  loss=0.751,  accuracy=0.781\n",
      "Epoch  0,  step=1064,  loss=1.100,  accuracy=0.656\n",
      "Epoch  0,  step=1065,  loss=1.288,  accuracy=0.688\n",
      "Epoch  0,  step=1066,  loss=0.958,  accuracy=0.812\n",
      "Epoch  0,  step=1067,  loss=0.202,  accuracy=0.938\n",
      "Epoch  0,  step=1068,  loss=0.815,  accuracy=0.875\n",
      "Epoch  0,  step=1069,  loss=0.461,  accuracy=0.875\n",
      "Epoch  0,  step=1070,  loss=0.405,  accuracy=0.906\n",
      "Epoch  0,  step=1071,  loss=0.559,  accuracy=0.781\n",
      "Epoch  0,  step=1072,  loss=0.726,  accuracy=0.844\n",
      "Epoch  0,  step=1073,  loss=0.840,  accuracy=0.781\n",
      "Epoch  0,  step=1074,  loss=0.381,  accuracy=0.812\n",
      "Epoch  0,  step=1075,  loss=0.986,  accuracy=0.750\n",
      "Epoch  0,  step=1076,  loss=1.344,  accuracy=0.688\n",
      "Epoch  0,  step=1077,  loss=1.032,  accuracy=0.875\n",
      "Epoch  0,  step=1078,  loss=1.148,  accuracy=0.812\n",
      "Epoch  0,  step=1079,  loss=1.749,  accuracy=0.625\n",
      "Epoch  0,  step=1080,  loss=0.548,  accuracy=0.844\n",
      "Epoch  0,  step=1081,  loss=0.920,  accuracy=0.719\n",
      "Epoch  0,  step=1082,  loss=0.932,  accuracy=0.781\n",
      "Epoch  0,  step=1083,  loss=1.116,  accuracy=0.781\n",
      "Epoch  0,  step=1084,  loss=1.329,  accuracy=0.688\n",
      "Epoch  0,  step=1085,  loss=0.937,  accuracy=0.812\n",
      "Epoch  0,  step=1086,  loss=0.544,  accuracy=0.938\n",
      "Epoch  0,  step=1087,  loss=0.630,  accuracy=0.812\n",
      "Epoch  0,  step=1088,  loss=1.954,  accuracy=0.625\n",
      "Epoch  0,  step=1089,  loss=0.736,  accuracy=0.781\n",
      "Epoch  0,  step=1090,  loss=0.648,  accuracy=0.812\n",
      "Epoch  0,  step=1091,  loss=0.370,  accuracy=0.875\n",
      "Epoch  0,  step=1092,  loss=0.285,  accuracy=0.906\n",
      "Epoch  0,  step=1093,  loss=0.473,  accuracy=0.844\n",
      "Epoch  0,  step=1094,  loss=0.204,  accuracy=0.875\n",
      "Epoch  0,  step=1095,  loss=0.650,  accuracy=0.844\n",
      "Epoch  0,  step=1096,  loss=0.684,  accuracy=0.938\n",
      "Epoch  0,  step=1097,  loss=0.997,  accuracy=0.781\n",
      "Epoch  0,  step=1098,  loss=1.482,  accuracy=0.688\n",
      "Epoch  0,  step=1099,  loss=0.953,  accuracy=0.750\n",
      "Epoch  0,  step=1100,  loss=1.371,  accuracy=0.812\n",
      "Epoch  0,  step=1101,  loss=1.179,  accuracy=0.781\n",
      "Epoch  0,  step=1102,  loss=0.512,  accuracy=0.844\n",
      "Epoch  0,  step=1103,  loss=0.691,  accuracy=0.812\n",
      "Epoch  0,  step=1104,  loss=0.313,  accuracy=0.906\n",
      "Epoch  0,  step=1105,  loss=0.420,  accuracy=0.906\n",
      "Epoch  0,  step=1106,  loss=0.761,  accuracy=0.875\n",
      "Epoch  0,  step=1107,  loss=1.079,  accuracy=0.750\n",
      "Epoch  0,  step=1108,  loss=2.222,  accuracy=0.719\n",
      "Epoch  0,  step=1109,  loss=0.742,  accuracy=0.844\n",
      "Epoch  0,  step=1110,  loss=0.525,  accuracy=0.812\n",
      "Epoch  0,  step=1111,  loss=0.721,  accuracy=0.875\n",
      "Epoch  0,  step=1112,  loss=0.413,  accuracy=0.906\n",
      "Epoch  0,  step=1113,  loss=1.812,  accuracy=0.656\n",
      "Epoch  0,  step=1114,  loss=0.276,  accuracy=0.875\n",
      "Epoch  0,  step=1115,  loss=1.169,  accuracy=0.781\n",
      "Epoch  0,  step=1116,  loss=0.418,  accuracy=0.844\n",
      "Epoch  0,  step=1117,  loss=0.679,  accuracy=0.875\n",
      "Epoch  0,  step=1118,  loss=0.682,  accuracy=0.812\n",
      "Epoch  0,  step=1119,  loss=0.500,  accuracy=0.875\n",
      "Epoch  0,  step=1120,  loss=1.001,  accuracy=0.875\n",
      "Epoch  0,  step=1121,  loss=0.545,  accuracy=0.781\n",
      "Epoch  0,  step=1122,  loss=0.635,  accuracy=0.875\n",
      "Epoch  0,  step=1123,  loss=1.089,  accuracy=0.750\n",
      "Epoch  0,  step=1124,  loss=1.269,  accuracy=0.812\n",
      "Epoch  0,  step=1125,  loss=1.235,  accuracy=0.781\n",
      "Epoch  0,  step=1126,  loss=1.075,  accuracy=0.812\n",
      "Epoch  0,  step=1127,  loss=1.419,  accuracy=0.719\n",
      "Epoch  0,  step=1128,  loss=1.630,  accuracy=0.750\n",
      "Epoch  0,  step=1129,  loss=0.351,  accuracy=0.875\n",
      "Epoch  0,  step=1130,  loss=1.187,  accuracy=0.781\n",
      "Epoch  0,  step=1131,  loss=0.349,  accuracy=0.875\n",
      "Epoch  0,  step=1132,  loss=0.930,  accuracy=0.781\n",
      "Epoch  0,  step=1133,  loss=0.770,  accuracy=0.812\n",
      "Epoch  0,  step=1134,  loss=0.063,  accuracy=1.000\n",
      "Epoch  0,  step=1135,  loss=0.706,  accuracy=0.812\n",
      "Epoch  0,  step=1136,  loss=0.331,  accuracy=0.812\n",
      "Epoch  0,  step=1137,  loss=0.912,  accuracy=0.750\n",
      "Epoch  0,  step=1138,  loss=1.610,  accuracy=0.719\n",
      "Epoch  0,  step=1139,  loss=1.370,  accuracy=0.781\n",
      "Epoch  0,  step=1140,  loss=1.352,  accuracy=0.750\n",
      "Epoch  0,  step=1141,  loss=0.605,  accuracy=0.781\n",
      "Epoch  0,  step=1142,  loss=0.208,  accuracy=0.938\n",
      "Epoch  0,  step=1143,  loss=0.642,  accuracy=0.875\n",
      "Epoch  0,  step=1144,  loss=0.093,  accuracy=0.969\n",
      "Epoch  0,  step=1145,  loss=1.275,  accuracy=0.688\n",
      "Epoch  0,  step=1146,  loss=0.252,  accuracy=0.906\n",
      "Epoch  0,  step=1147,  loss=0.727,  accuracy=0.812\n",
      "Epoch  0,  step=1148,  loss=0.606,  accuracy=0.812\n",
      "Epoch  0,  step=1149,  loss=0.606,  accuracy=0.844\n",
      "Epoch  0,  step=1150,  loss=0.623,  accuracy=0.906\n",
      "Epoch  0,  step=1151,  loss=0.772,  accuracy=0.688\n",
      "Epoch  0,  step=1152,  loss=0.578,  accuracy=0.812\n",
      "Epoch  0,  step=1153,  loss=0.520,  accuracy=0.812\n",
      "Epoch  0,  step=1154,  loss=0.635,  accuracy=0.812\n",
      "Epoch  0,  step=1155,  loss=0.446,  accuracy=0.875\n",
      "Epoch  0,  step=1156,  loss=0.250,  accuracy=0.875\n",
      "Epoch  0,  step=1157,  loss=0.967,  accuracy=0.812\n",
      "Epoch  0,  step=1158,  loss=0.991,  accuracy=0.750\n",
      "Epoch  0,  step=1159,  loss=0.974,  accuracy=0.812\n",
      "Epoch  0,  step=1160,  loss=0.450,  accuracy=0.875\n",
      "Epoch  0,  step=1161,  loss=0.896,  accuracy=0.750\n",
      "Epoch  0,  step=1162,  loss=0.977,  accuracy=0.781\n",
      "Epoch  0,  step=1163,  loss=0.278,  accuracy=0.906\n",
      "Epoch  0,  step=1164,  loss=1.255,  accuracy=0.719\n",
      "Epoch  0,  step=1165,  loss=0.992,  accuracy=0.781\n",
      "Epoch  0,  step=1166,  loss=1.091,  accuracy=0.750\n",
      "Epoch  0,  step=1167,  loss=1.501,  accuracy=0.781\n",
      "Epoch  0,  step=1168,  loss=2.020,  accuracy=0.719\n",
      "Epoch  0,  step=1169,  loss=1.228,  accuracy=0.719\n",
      "Epoch  0,  step=1170,  loss=1.738,  accuracy=0.719\n",
      "Epoch  0,  step=1171,  loss=0.943,  accuracy=0.844\n",
      "Epoch  0,  step=1172,  loss=0.982,  accuracy=0.750\n",
      "Epoch  0,  step=1173,  loss=1.725,  accuracy=0.656\n",
      "Epoch  0,  step=1174,  loss=1.474,  accuracy=0.656\n",
      "Epoch  0,  step=1175,  loss=0.994,  accuracy=0.812\n",
      "Epoch  0,  step=1176,  loss=0.924,  accuracy=0.781\n",
      "Epoch  0,  step=1177,  loss=0.507,  accuracy=0.844\n",
      "Epoch  0,  step=1178,  loss=1.028,  accuracy=0.750\n",
      "Epoch  0,  step=1179,  loss=1.109,  accuracy=0.812\n",
      "Epoch  0,  step=1180,  loss=1.061,  accuracy=0.781\n",
      "Epoch  0,  step=1181,  loss=1.101,  accuracy=0.781\n",
      "Epoch  0,  step=1182,  loss=0.889,  accuracy=0.781\n",
      "Epoch  0,  step=1183,  loss=0.655,  accuracy=0.781\n",
      "Epoch  0,  step=1184,  loss=0.773,  accuracy=0.781\n",
      "Epoch  0,  step=1185,  loss=1.231,  accuracy=0.750\n",
      "Epoch  0,  step=1186,  loss=0.444,  accuracy=0.938\n",
      "Epoch  0,  step=1187,  loss=0.811,  accuracy=0.844\n",
      "Epoch  0,  step=1188,  loss=0.922,  accuracy=0.812\n",
      "Epoch  0,  step=1189,  loss=1.108,  accuracy=0.750\n",
      "Epoch  0,  step=1190,  loss=0.495,  accuracy=0.875\n",
      "Epoch  0,  step=1191,  loss=0.421,  accuracy=0.875\n",
      "Epoch  0,  step=1192,  loss=1.067,  accuracy=0.656\n",
      "Epoch  0,  step=1193,  loss=0.501,  accuracy=0.938\n",
      "Epoch  0,  step=1194,  loss=0.838,  accuracy=0.844\n",
      "Epoch  0,  step=1195,  loss=1.408,  accuracy=0.750\n",
      "Epoch  0,  step=1196,  loss=0.580,  accuracy=0.844\n",
      "Epoch  0,  step=1197,  loss=0.558,  accuracy=0.875\n",
      "Epoch  0,  step=1198,  loss=0.349,  accuracy=0.844\n",
      "Epoch  0,  step=1199,  loss=1.360,  accuracy=0.719\n",
      "Epoch  0,  step=1200,  loss=0.432,  accuracy=0.906\n",
      "Epoch  0,  step=1201,  loss=0.829,  accuracy=0.750\n",
      "Epoch  0,  step=1202,  loss=0.172,  accuracy=0.906\n",
      "Epoch  0,  step=1203,  loss=1.474,  accuracy=0.781\n",
      "Epoch  0,  step=1204,  loss=0.504,  accuracy=0.938\n",
      "Epoch  0,  step=1205,  loss=1.268,  accuracy=0.750\n",
      "Epoch  0,  step=1206,  loss=0.790,  accuracy=0.875\n",
      "Epoch  0,  step=1207,  loss=0.639,  accuracy=0.812\n",
      "Epoch  0,  step=1208,  loss=0.957,  accuracy=0.812\n",
      "Epoch  0,  step=1209,  loss=0.708,  accuracy=0.812\n",
      "Epoch  0,  step=1210,  loss=0.346,  accuracy=0.938\n",
      "Epoch  0,  step=1211,  loss=0.955,  accuracy=0.812\n",
      "Epoch  0,  step=1212,  loss=0.494,  accuracy=0.844\n",
      "Epoch  0,  step=1213,  loss=0.297,  accuracy=0.938\n",
      "Epoch  0,  step=1214,  loss=0.229,  accuracy=0.875\n",
      "Epoch  0,  step=1215,  loss=0.430,  accuracy=0.875\n",
      "Epoch  0,  step=1216,  loss=0.366,  accuracy=0.875\n",
      "Epoch  0,  step=1217,  loss=0.252,  accuracy=0.938\n",
      "Epoch  0,  step=1218,  loss=1.067,  accuracy=0.719\n",
      "Epoch  0,  step=1219,  loss=0.383,  accuracy=0.844\n",
      "Epoch  0,  step=1220,  loss=0.622,  accuracy=0.906\n",
      "Epoch  0,  step=1221,  loss=0.229,  accuracy=0.875\n",
      "Epoch  0,  step=1222,  loss=1.229,  accuracy=0.781\n",
      "Epoch  0,  step=1223,  loss=0.800,  accuracy=0.844\n",
      "Epoch  0,  step=1224,  loss=0.418,  accuracy=0.875\n",
      "Epoch  0,  step=1225,  loss=1.280,  accuracy=0.812\n",
      "Epoch  0,  step=1226,  loss=0.901,  accuracy=0.812\n",
      "Epoch  0,  step=1227,  loss=0.729,  accuracy=0.812\n",
      "Epoch  0,  step=1228,  loss=2.556,  accuracy=0.656\n",
      "Epoch  0,  step=1229,  loss=1.328,  accuracy=0.719\n",
      "Epoch  0,  step=1230,  loss=0.656,  accuracy=0.812\n",
      "Epoch  0,  step=1231,  loss=1.443,  accuracy=0.812\n",
      "Epoch  0,  step=1232,  loss=1.119,  accuracy=0.750\n",
      "Epoch  0,  step=1233,  loss=0.592,  accuracy=0.938\n",
      "Epoch  0,  step=1234,  loss=0.667,  accuracy=0.812\n",
      "Epoch  0,  step=1235,  loss=0.872,  accuracy=0.844\n",
      "Epoch  0,  step=1236,  loss=0.593,  accuracy=0.906\n",
      "Epoch  0,  step=1237,  loss=1.255,  accuracy=0.812\n",
      "Epoch  0,  step=1238,  loss=0.859,  accuracy=0.812\n",
      "Epoch  0,  step=1239,  loss=1.306,  accuracy=0.750\n",
      "Epoch  0,  step=1240,  loss=1.220,  accuracy=0.844\n",
      "Epoch  0,  step=1241,  loss=0.535,  accuracy=0.781\n",
      "Epoch  0,  step=1242,  loss=1.445,  accuracy=0.750\n",
      "Epoch  0,  step=1243,  loss=1.332,  accuracy=0.688\n",
      "Epoch  0,  step=1244,  loss=0.695,  accuracy=0.812\n",
      "Epoch  0,  step=1245,  loss=1.141,  accuracy=0.781\n",
      "Epoch  0,  step=1246,  loss=1.420,  accuracy=0.719\n",
      "Epoch  0,  step=1247,  loss=1.370,  accuracy=0.750\n",
      "Epoch  0,  step=1248,  loss=1.108,  accuracy=0.781\n",
      "Epoch  0,  step=1249,  loss=1.077,  accuracy=0.781\n",
      "Epoch  0,  step=1250,  loss=0.571,  accuracy=0.812\n",
      "Epoch  0,  step=1251,  loss=0.751,  accuracy=0.750\n",
      "Epoch  0,  step=1252,  loss=0.728,  accuracy=0.781\n",
      "Epoch  0,  step=1253,  loss=0.633,  accuracy=0.844\n",
      "Epoch  0,  step=1254,  loss=0.871,  accuracy=0.844\n",
      "Epoch  0,  step=1255,  loss=0.497,  accuracy=0.812\n",
      "Epoch  0,  step=1256,  loss=0.575,  accuracy=0.812\n",
      "Epoch  0,  step=1257,  loss=0.912,  accuracy=0.844\n",
      "Epoch  0,  step=1258,  loss=0.768,  accuracy=0.812\n",
      "Epoch  0,  step=1259,  loss=0.784,  accuracy=0.750\n",
      "Epoch  0,  step=1260,  loss=0.551,  accuracy=0.812\n",
      "Epoch  0,  step=1261,  loss=0.469,  accuracy=0.875\n",
      "Epoch  0,  step=1262,  loss=0.969,  accuracy=0.688\n",
      "Epoch  0,  step=1263,  loss=0.288,  accuracy=0.844\n",
      "Epoch  0,  step=1264,  loss=0.587,  accuracy=0.812\n",
      "Epoch  0,  step=1265,  loss=1.171,  accuracy=0.719\n",
      "Epoch  0,  step=1266,  loss=0.922,  accuracy=0.812\n",
      "Epoch  0,  step=1267,  loss=0.489,  accuracy=0.844\n",
      "Epoch  0,  step=1268,  loss=1.209,  accuracy=0.781\n",
      "Epoch  0,  step=1269,  loss=0.516,  accuracy=0.875\n",
      "Epoch  0,  step=1270,  loss=1.550,  accuracy=0.750\n",
      "Epoch  0,  step=1271,  loss=0.437,  accuracy=0.844\n",
      "Epoch  0,  step=1272,  loss=0.603,  accuracy=0.875\n",
      "Epoch  0,  step=1273,  loss=0.408,  accuracy=0.906\n",
      "Epoch  0,  step=1274,  loss=0.130,  accuracy=0.906\n",
      "Epoch  0,  step=1275,  loss=0.574,  accuracy=0.781\n",
      "Epoch  0,  step=1276,  loss=0.264,  accuracy=0.875\n",
      "Epoch  0,  step=1277,  loss=0.390,  accuracy=0.844\n",
      "Epoch  0,  step=1278,  loss=0.249,  accuracy=0.938\n",
      "Epoch  0,  step=1279,  loss=0.459,  accuracy=0.875\n",
      "Epoch  0,  step=1280,  loss=0.989,  accuracy=0.812\n",
      "Epoch  0,  step=1281,  loss=1.200,  accuracy=0.750\n",
      "Epoch  0,  step=1282,  loss=0.339,  accuracy=0.875\n",
      "Epoch  0,  step=1283,  loss=0.543,  accuracy=0.875\n",
      "Epoch  0,  step=1284,  loss=0.506,  accuracy=0.906\n",
      "Epoch  0,  step=1285,  loss=0.474,  accuracy=0.906\n",
      "Epoch  0,  step=1286,  loss=0.474,  accuracy=0.906\n",
      "Epoch  0,  step=1287,  loss=0.322,  accuracy=0.844\n",
      "Epoch  0,  step=1288,  loss=1.069,  accuracy=0.750\n",
      "Epoch  0,  step=1289,  loss=1.468,  accuracy=0.844\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  0,  step=1290,  loss=1.260,  accuracy=0.781\n",
      "Epoch  0,  step=1291,  loss=1.261,  accuracy=0.781\n",
      "Epoch  0,  step=1292,  loss=0.955,  accuracy=0.688\n",
      "Epoch  0,  step=1293,  loss=0.849,  accuracy=0.719\n",
      "Epoch  0,  step=1294,  loss=1.742,  accuracy=0.688\n",
      "Epoch  0,  step=1295,  loss=1.330,  accuracy=0.781\n",
      "Epoch  0,  step=1296,  loss=0.865,  accuracy=0.719\n",
      "Epoch  0,  step=1297,  loss=0.923,  accuracy=0.781\n",
      "Epoch  0,  step=1298,  loss=0.860,  accuracy=0.781\n",
      "Epoch  0,  step=1299,  loss=1.147,  accuracy=0.750\n",
      "Epoch  0,  step=1300,  loss=0.905,  accuracy=0.812\n",
      "Epoch  0,  step=1301,  loss=0.164,  accuracy=0.906\n",
      "Epoch  0,  step=1302,  loss=0.506,  accuracy=0.750\n",
      "Epoch  0,  step=1303,  loss=0.736,  accuracy=0.781\n",
      "Epoch  0,  step=1304,  loss=0.894,  accuracy=0.719\n",
      "Epoch  0,  step=1305,  loss=1.202,  accuracy=0.750\n",
      "Epoch  0,  step=1306,  loss=1.073,  accuracy=0.781\n",
      "Epoch  0,  step=1307,  loss=0.619,  accuracy=0.875\n",
      "Epoch  0,  step=1308,  loss=1.192,  accuracy=0.781\n",
      "Epoch  0,  step=1309,  loss=1.239,  accuracy=0.750\n",
      "Epoch  0,  step=1310,  loss=1.024,  accuracy=0.844\n",
      "Epoch  0,  step=1311,  loss=0.993,  accuracy=0.750\n",
      "Epoch  0,  step=1312,  loss=1.132,  accuracy=0.781\n",
      "Epoch  0,  step=1313,  loss=1.036,  accuracy=0.750\n",
      "Epoch  0,  step=1314,  loss=0.807,  accuracy=0.844\n",
      "Epoch  0,  step=1315,  loss=0.500,  accuracy=0.844\n",
      "Epoch  0,  step=1316,  loss=0.903,  accuracy=0.688\n",
      "Epoch  0,  step=1317,  loss=0.481,  accuracy=0.875\n",
      "Epoch  0,  step=1318,  loss=0.930,  accuracy=0.781\n",
      "Epoch  0,  step=1319,  loss=0.773,  accuracy=0.750\n",
      "Epoch  0,  step=1320,  loss=0.855,  accuracy=0.812\n",
      "Epoch  0,  step=1321,  loss=0.305,  accuracy=0.875\n",
      "Epoch  0,  step=1322,  loss=1.763,  accuracy=0.719\n",
      "Epoch  0,  step=1323,  loss=1.676,  accuracy=0.656\n",
      "Epoch  0,  step=1324,  loss=0.502,  accuracy=0.906\n",
      "Epoch  0,  step=1325,  loss=1.018,  accuracy=0.781\n",
      "Epoch  0,  step=1326,  loss=0.916,  accuracy=0.781\n",
      "Epoch  0,  step=1327,  loss=1.176,  accuracy=0.812\n",
      "Epoch  0,  step=1328,  loss=1.382,  accuracy=0.750\n",
      "Epoch  0,  step=1329,  loss=0.544,  accuracy=0.875\n",
      "Epoch  0,  step=1330,  loss=0.373,  accuracy=0.875\n",
      "Epoch  0,  step=1331,  loss=0.848,  accuracy=0.844\n",
      "Epoch  0,  step=1332,  loss=0.901,  accuracy=0.812\n",
      "Epoch  0,  step=1333,  loss=1.279,  accuracy=0.812\n",
      "Epoch  0,  step=1334,  loss=0.757,  accuracy=0.719\n",
      "Epoch  0,  step=1335,  loss=0.568,  accuracy=0.906\n",
      "Epoch  0,  step=1336,  loss=0.830,  accuracy=0.875\n",
      "Epoch  0,  step=1337,  loss=0.930,  accuracy=0.844\n",
      "Epoch  0,  step=1338,  loss=1.022,  accuracy=0.812\n",
      "Epoch  0,  step=1339,  loss=1.512,  accuracy=0.812\n",
      "Epoch  0,  step=1340,  loss=1.063,  accuracy=0.750\n",
      "Epoch  0,  step=1341,  loss=0.864,  accuracy=0.844\n",
      "Epoch  0,  step=1342,  loss=1.204,  accuracy=0.719\n",
      "Epoch  0,  step=1343,  loss=0.817,  accuracy=0.781\n",
      "Epoch  0,  step=1344,  loss=1.333,  accuracy=0.719\n",
      "Epoch  0,  step=1345,  loss=0.707,  accuracy=0.781\n",
      "Epoch  0,  step=1346,  loss=1.058,  accuracy=0.781\n",
      "Epoch  0,  step=1347,  loss=0.931,  accuracy=0.844\n",
      "Epoch  0,  step=1348,  loss=0.212,  accuracy=0.938\n",
      "Epoch  0,  step=1349,  loss=0.563,  accuracy=0.812\n",
      "Epoch  0,  step=1350,  loss=2.102,  accuracy=0.688\n",
      "Epoch  0,  step=1351,  loss=0.616,  accuracy=0.812\n",
      "Epoch  0,  step=1352,  loss=0.684,  accuracy=0.750\n",
      "Epoch  0,  step=1353,  loss=0.199,  accuracy=0.938\n",
      "Epoch  0,  step=1354,  loss=0.144,  accuracy=0.938\n",
      "Epoch  0,  step=1355,  loss=0.735,  accuracy=0.906\n",
      "Epoch  0,  step=1356,  loss=0.561,  accuracy=0.875\n",
      "Epoch  0,  step=1357,  loss=0.861,  accuracy=0.812\n",
      "Epoch  0,  step=1358,  loss=0.512,  accuracy=0.844\n",
      "Epoch  0,  step=1359,  loss=0.481,  accuracy=0.844\n",
      "Epoch  0,  step=1360,  loss=0.958,  accuracy=0.719\n",
      "Epoch  0,  step=1361,  loss=0.757,  accuracy=0.844\n",
      "Epoch  0,  step=1362,  loss=0.459,  accuracy=0.812\n",
      "Epoch  0,  step=1363,  loss=0.505,  accuracy=0.906\n",
      "Epoch  0,  step=1364,  loss=1.004,  accuracy=0.781\n",
      "Epoch  0,  step=1365,  loss=0.677,  accuracy=0.812\n",
      "Epoch  0,  step=1366,  loss=0.284,  accuracy=0.875\n",
      "Epoch  0,  step=1367,  loss=0.404,  accuracy=0.938\n",
      "Epoch  0,  step=1368,  loss=0.555,  accuracy=0.875\n",
      "Epoch  0,  step=1369,  loss=1.217,  accuracy=0.688\n",
      "Epoch  0,  step=1370,  loss=0.909,  accuracy=0.812\n",
      "Epoch  0,  step=1371,  loss=0.541,  accuracy=0.812\n",
      "Epoch  0,  step=1372,  loss=0.169,  accuracy=0.938\n",
      "Epoch  0,  step=1373,  loss=1.614,  accuracy=0.719\n",
      "Epoch  0,  step=1374,  loss=0.702,  accuracy=0.812\n",
      "Epoch  0,  step=1375,  loss=0.128,  accuracy=0.969\n",
      "Epoch  0,  step=1376,  loss=0.713,  accuracy=0.906\n",
      "Epoch  0,  step=1377,  loss=0.721,  accuracy=0.750\n",
      "Epoch  0,  step=1378,  loss=0.915,  accuracy=0.750\n",
      "Epoch  0,  step=1379,  loss=1.488,  accuracy=0.750\n",
      "Epoch  0,  step=1380,  loss=0.757,  accuracy=0.781\n",
      "Epoch  0,  step=1381,  loss=0.706,  accuracy=0.844\n",
      "Epoch  0,  step=1382,  loss=0.980,  accuracy=0.781\n",
      "Epoch  0,  step=1383,  loss=1.255,  accuracy=0.812\n",
      "Epoch  0,  step=1384,  loss=0.828,  accuracy=0.844\n",
      "Epoch  0,  step=1385,  loss=0.727,  accuracy=0.719\n",
      "Epoch  0,  step=1386,  loss=1.151,  accuracy=0.875\n",
      "Epoch  0,  step=1387,  loss=0.332,  accuracy=0.844\n",
      "Epoch  0,  step=1388,  loss=0.830,  accuracy=0.844\n",
      "Epoch  0,  step=1389,  loss=0.424,  accuracy=0.906\n",
      "Epoch  0,  step=1390,  loss=0.751,  accuracy=0.906\n",
      "Epoch  0,  step=1391,  loss=0.524,  accuracy=0.938\n",
      "Epoch  0,  step=1392,  loss=0.136,  accuracy=0.938\n",
      "Epoch  0,  step=1393,  loss=0.166,  accuracy=0.906\n",
      "Epoch  0,  step=1394,  loss=0.319,  accuracy=0.844\n",
      "Epoch  0,  step=1395,  loss=0.946,  accuracy=0.750\n",
      "Epoch  0,  step=1396,  loss=0.686,  accuracy=0.844\n",
      "Epoch  0,  step=1397,  loss=0.630,  accuracy=0.812\n",
      "Epoch  0,  step=1398,  loss=0.724,  accuracy=0.781\n",
      "Epoch  0,  step=1399,  loss=0.225,  accuracy=0.906\n",
      "Epoch  0,  step=1400,  loss=0.765,  accuracy=0.875\n",
      "Epoch  0,  step=1401,  loss=0.846,  accuracy=0.812\n",
      "Epoch  0,  step=1402,  loss=1.139,  accuracy=0.812\n",
      "Epoch  0,  step=1403,  loss=0.963,  accuracy=0.750\n",
      "Epoch  0,  step=1404,  loss=0.864,  accuracy=0.812\n",
      "Epoch  0,  step=1405,  loss=0.603,  accuracy=0.844\n",
      "Epoch  0,  step=1406,  loss=1.076,  accuracy=0.812\n",
      "Epoch  0,  step=1407,  loss=1.781,  accuracy=0.719\n",
      "Epoch  0,  step=1408,  loss=0.674,  accuracy=0.844\n",
      "Epoch  0,  step=1409,  loss=1.088,  accuracy=0.781\n",
      "Epoch  0,  step=1410,  loss=0.558,  accuracy=0.844\n",
      "Epoch  0,  step=1411,  loss=0.693,  accuracy=0.844\n",
      "Epoch  0,  step=1412,  loss=0.493,  accuracy=0.906\n",
      "Epoch  0,  step=1413,  loss=0.463,  accuracy=0.875\n",
      "Epoch  0,  step=1414,  loss=0.364,  accuracy=0.875\n",
      "Epoch  0,  step=1415,  loss=0.130,  accuracy=0.969\n",
      "Epoch  0,  step=1416,  loss=0.472,  accuracy=0.906\n",
      "Epoch  0,  step=1417,  loss=0.276,  accuracy=0.844\n",
      "Epoch  0,  step=1418,  loss=0.282,  accuracy=0.938\n",
      "Epoch  0,  step=1419,  loss=1.274,  accuracy=0.781\n",
      "Epoch  0,  step=1420,  loss=1.141,  accuracy=0.750\n",
      "Epoch  0,  step=1421,  loss=1.998,  accuracy=0.688\n",
      "Epoch  0,  step=1422,  loss=0.932,  accuracy=0.781\n",
      "Epoch  0,  step=1423,  loss=0.450,  accuracy=0.875\n",
      "Epoch  0,  step=1424,  loss=0.750,  accuracy=0.781\n",
      "Epoch  0,  step=1425,  loss=0.882,  accuracy=0.875\n",
      "Epoch  0,  step=1426,  loss=1.291,  accuracy=0.688\n",
      "Epoch  0,  step=1427,  loss=0.861,  accuracy=0.844\n",
      "Epoch  0,  step=1428,  loss=0.061,  accuracy=0.969\n",
      "Epoch  0,  step=1429,  loss=0.658,  accuracy=0.812\n",
      "Epoch  0,  step=1430,  loss=0.787,  accuracy=0.906\n",
      "Epoch  0,  step=1431,  loss=0.888,  accuracy=0.656\n",
      "Epoch  0,  step=1432,  loss=1.017,  accuracy=0.719\n",
      "Epoch  0,  step=1433,  loss=1.155,  accuracy=0.781\n",
      "Epoch  0,  step=1434,  loss=1.101,  accuracy=0.719\n",
      "Epoch  0,  step=1435,  loss=0.781,  accuracy=0.781\n",
      "Epoch  0,  step=1436,  loss=0.594,  accuracy=0.875\n",
      "Epoch  0,  step=1437,  loss=0.955,  accuracy=0.812\n",
      "Epoch  0,  step=1438,  loss=0.809,  accuracy=0.781\n",
      "Epoch  0,  step=1439,  loss=0.737,  accuracy=0.812\n",
      "Epoch  0,  step=1440,  loss=1.217,  accuracy=0.688\n",
      "Epoch  0,  step=1441,  loss=0.525,  accuracy=0.906\n",
      "Epoch  0,  step=1442,  loss=0.707,  accuracy=0.781\n",
      "Epoch  0,  step=1443,  loss=0.521,  accuracy=0.844\n",
      "Epoch  0,  step=1444,  loss=0.341,  accuracy=0.906\n",
      "Epoch  0,  step=1445,  loss=2.275,  accuracy=0.688\n",
      "Epoch  0,  step=1446,  loss=1.073,  accuracy=0.812\n",
      "Epoch  0,  step=1447,  loss=0.955,  accuracy=0.750\n",
      "Epoch  0,  step=1448,  loss=0.370,  accuracy=0.906\n",
      "Epoch  0,  step=1449,  loss=0.399,  accuracy=0.812\n",
      "Epoch  0,  step=1450,  loss=1.133,  accuracy=0.656\n",
      "Epoch  0,  step=1451,  loss=0.587,  accuracy=0.844\n",
      "Epoch  0,  step=1452,  loss=0.437,  accuracy=0.812\n",
      "Epoch  0,  step=1453,  loss=0.325,  accuracy=0.938\n",
      "Epoch  0,  step=1454,  loss=0.520,  accuracy=0.875\n",
      "Epoch  0,  step=1455,  loss=0.282,  accuracy=0.906\n",
      "Epoch  0,  step=1456,  loss=0.705,  accuracy=0.781\n",
      "Epoch  0,  step=1457,  loss=0.843,  accuracy=0.781\n",
      "Epoch  0,  step=1458,  loss=0.201,  accuracy=0.938\n",
      "Epoch  0,  step=1459,  loss=0.607,  accuracy=0.844\n",
      "Epoch  0,  step=1460,  loss=1.503,  accuracy=0.688\n",
      "Epoch  0,  step=1461,  loss=0.495,  accuracy=0.812\n",
      "Epoch  0,  step=1462,  loss=0.429,  accuracy=0.844\n",
      "Epoch  0,  step=1463,  loss=0.153,  accuracy=0.906\n",
      "Epoch  0,  step=1464,  loss=0.574,  accuracy=0.844\n",
      "Epoch  0,  step=1465,  loss=0.781,  accuracy=0.844\n",
      "Epoch  0,  step=1466,  loss=0.463,  accuracy=0.938\n",
      "Epoch  0,  step=1467,  loss=0.199,  accuracy=0.906\n",
      "Epoch  0,  step=1468,  loss=0.759,  accuracy=0.812\n",
      "Epoch  0,  step=1469,  loss=1.229,  accuracy=0.844\n",
      "Epoch  0,  step=1470,  loss=0.864,  accuracy=0.781\n",
      "Epoch  0,  step=1471,  loss=0.508,  accuracy=0.938\n",
      "Epoch  0,  step=1472,  loss=0.127,  accuracy=0.969\n",
      "Epoch  0,  step=1473,  loss=0.902,  accuracy=0.750\n",
      "Epoch  0,  step=1474,  loss=0.546,  accuracy=0.906\n",
      "Epoch  0,  step=1475,  loss=1.499,  accuracy=0.781\n",
      "Epoch  0,  step=1476,  loss=0.925,  accuracy=0.781\n",
      "Epoch  0,  step=1477,  loss=0.704,  accuracy=0.844\n",
      "Epoch  0,  step=1478,  loss=1.252,  accuracy=0.656\n",
      "Epoch  0,  step=1479,  loss=1.009,  accuracy=0.812\n",
      "Epoch  0,  step=1480,  loss=0.869,  accuracy=0.906\n",
      "Epoch  0,  step=1481,  loss=0.987,  accuracy=0.812\n",
      "Epoch  0,  step=1482,  loss=0.792,  accuracy=0.750\n",
      "Epoch  0,  step=1483,  loss=0.879,  accuracy=0.656\n",
      "Epoch  0,  step=1484,  loss=0.249,  accuracy=0.906\n",
      "Epoch  0,  step=1485,  loss=0.432,  accuracy=0.938\n",
      "Epoch  0,  step=1486,  loss=0.731,  accuracy=0.875\n",
      "Epoch  0,  step=1487,  loss=1.507,  accuracy=0.688\n",
      "Epoch  0,  step=1488,  loss=0.513,  accuracy=0.844\n",
      "Epoch  0,  step=1489,  loss=0.657,  accuracy=0.719\n",
      "Epoch  0,  step=1490,  loss=0.772,  accuracy=0.875\n",
      "Epoch  0,  step=1491,  loss=0.607,  accuracy=0.875\n",
      "Epoch  0,  step=1492,  loss=1.234,  accuracy=0.719\n",
      "Epoch  0,  step=1493,  loss=0.432,  accuracy=0.906\n",
      "Epoch  0,  step=1494,  loss=0.397,  accuracy=0.969\n",
      "Epoch  0,  step=1495,  loss=0.426,  accuracy=0.906\n",
      "Epoch  0,  step=1496,  loss=0.407,  accuracy=0.906\n",
      "Epoch  0,  step=1497,  loss=0.913,  accuracy=0.875\n",
      "Epoch  0,  step=1498,  loss=1.275,  accuracy=0.781\n",
      "Epoch  0,  step=1499,  loss=0.217,  accuracy=0.906\n",
      "Epoch  0,  step=1500,  loss=0.556,  accuracy=0.875\n",
      "Epoch  0,  step=1501,  loss=0.262,  accuracy=0.844\n",
      "Epoch  0,  step=1502,  loss=0.690,  accuracy=0.812\n",
      "Epoch  0,  step=1503,  loss=0.629,  accuracy=0.844\n",
      "Epoch  0,  step=1504,  loss=0.316,  accuracy=0.875\n",
      "Epoch  0,  step=1505,  loss=0.517,  accuracy=0.844\n",
      "Epoch  0,  step=1506,  loss=0.210,  accuracy=0.938\n",
      "Epoch  0,  step=1507,  loss=0.129,  accuracy=0.938\n",
      "Epoch  0,  step=1508,  loss=0.500,  accuracy=0.812\n",
      "Epoch  0,  step=1509,  loss=0.100,  accuracy=0.938\n",
      "Epoch  0,  step=1510,  loss=0.846,  accuracy=0.875\n",
      "Epoch  0,  step=1511,  loss=1.231,  accuracy=0.656\n",
      "Epoch  0,  step=1512,  loss=1.088,  accuracy=0.812\n",
      "Epoch  0,  step=1513,  loss=0.126,  accuracy=0.969\n",
      "Epoch  0,  step=1514,  loss=0.533,  accuracy=0.875\n",
      "Epoch  0,  step=1515,  loss=0.670,  accuracy=0.844\n",
      "Epoch  0,  step=1516,  loss=0.497,  accuracy=0.844\n",
      "Epoch  0,  step=1517,  loss=0.467,  accuracy=0.875\n",
      "Epoch  0,  step=1518,  loss=0.373,  accuracy=0.844\n",
      "Epoch  0,  step=1519,  loss=0.294,  accuracy=0.938\n",
      "Epoch  0,  step=1520,  loss=1.019,  accuracy=0.875\n",
      "Epoch  0,  step=1521,  loss=0.659,  accuracy=0.781\n",
      "Epoch  0,  step=1522,  loss=0.404,  accuracy=0.812\n",
      "Epoch  0,  step=1523,  loss=0.212,  accuracy=0.875\n",
      "Epoch  0,  step=1524,  loss=0.254,  accuracy=0.875\n",
      "Epoch  0,  step=1525,  loss=0.336,  accuracy=0.844\n",
      "Epoch  0,  step=1526,  loss=0.535,  accuracy=0.844\n",
      "Epoch  0,  step=1527,  loss=0.655,  accuracy=0.812\n",
      "Epoch  0,  step=1528,  loss=1.389,  accuracy=0.812\n",
      "Epoch  0,  step=1529,  loss=1.395,  accuracy=0.656\n",
      "Epoch  0,  step=1530,  loss=1.249,  accuracy=0.844\n",
      "Epoch  0,  step=1531,  loss=1.985,  accuracy=0.656\n",
      "Epoch  0,  step=1532,  loss=0.950,  accuracy=0.812\n",
      "Epoch  0,  step=1533,  loss=1.311,  accuracy=0.812\n",
      "Epoch  0,  step=1534,  loss=1.242,  accuracy=0.656\n",
      "Epoch  0,  step=1535,  loss=1.168,  accuracy=0.719\n",
      "Epoch  0,  step=1536,  loss=0.602,  accuracy=0.812\n",
      "Epoch  0,  step=1537,  loss=2.016,  accuracy=0.656\n",
      "Epoch  0,  step=1538,  loss=1.140,  accuracy=0.781\n",
      "Epoch  0,  step=1539,  loss=0.248,  accuracy=0.938\n",
      "Epoch  0,  step=1540,  loss=0.419,  accuracy=0.906\n",
      "Epoch  0,  step=1541,  loss=0.269,  accuracy=0.875\n",
      "Epoch  0,  step=1542,  loss=0.260,  accuracy=0.938\n",
      "Epoch  0,  step=1543,  loss=0.382,  accuracy=0.875\n",
      "Epoch  0,  step=1544,  loss=1.975,  accuracy=0.688\n",
      "Epoch  0,  step=1545,  loss=0.330,  accuracy=0.875\n",
      "Epoch  0,  step=1546,  loss=1.380,  accuracy=0.750\n",
      "Epoch  0,  step=1547,  loss=0.958,  accuracy=0.719\n",
      "Epoch  0,  step=1548,  loss=1.296,  accuracy=0.719\n",
      "Epoch  0,  step=1549,  loss=1.334,  accuracy=0.844\n",
      "Epoch  0,  step=1550,  loss=0.865,  accuracy=0.844\n",
      "Epoch  0,  step=1551,  loss=1.474,  accuracy=0.719\n",
      "Epoch  0,  step=1552,  loss=0.605,  accuracy=0.781\n",
      "Epoch  0,  step=1553,  loss=0.377,  accuracy=0.875\n",
      "Epoch  0,  step=1554,  loss=0.569,  accuracy=0.844\n",
      "Epoch  0,  step=1555,  loss=0.360,  accuracy=0.875\n",
      "Epoch  0,  step=1556,  loss=0.725,  accuracy=0.844\n",
      "Epoch  0,  step=1557,  loss=1.028,  accuracy=0.781\n",
      "Epoch  0,  step=1558,  loss=0.342,  accuracy=0.906\n",
      "Epoch  0,  step=1559,  loss=1.270,  accuracy=0.750\n",
      "Epoch  0,  step=1560,  loss=0.578,  accuracy=0.719\n",
      "Epoch  0,  step=1561,  loss=0.423,  accuracy=0.844\n",
      "Epoch  0,  step=1562,  loss=0.912,  accuracy=0.719\n",
      "Epoch  0,  step=1563,  loss=0.749,  accuracy=0.750\n",
      "Epoch  0,  step=1564,  loss=0.380,  accuracy=0.875\n",
      "Epoch  0,  step=1565,  loss=0.458,  accuracy=0.844\n",
      "Epoch  0,  step=1566,  loss=0.304,  accuracy=0.938\n",
      "Epoch  0,  step=1567,  loss=1.127,  accuracy=0.812\n",
      "Epoch  0,  step=1568,  loss=0.285,  accuracy=0.938\n",
      "Epoch  0,  step=1569,  loss=1.597,  accuracy=0.781\n",
      "Epoch  0,  step=1570,  loss=0.620,  accuracy=0.906\n",
      "Epoch  0,  step=1571,  loss=0.591,  accuracy=0.875\n",
      "Epoch  0,  step=1572,  loss=1.123,  accuracy=0.781\n",
      "Epoch  0,  step=1573,  loss=1.303,  accuracy=0.781\n",
      "Epoch  0,  step=1574,  loss=1.234,  accuracy=0.656\n",
      "Epoch  0,  step=1575,  loss=2.286,  accuracy=0.594\n",
      "Epoch  0,  step=1576,  loss=0.871,  accuracy=0.719\n",
      "Epoch  0,  step=1577,  loss=0.271,  accuracy=0.938\n",
      "Epoch  0,  step=1578,  loss=1.246,  accuracy=0.844\n",
      "Epoch  0,  step=1579,  loss=0.298,  accuracy=0.875\n",
      "Epoch  0,  step=1580,  loss=1.751,  accuracy=0.719\n",
      "Epoch  0,  step=1581,  loss=0.602,  accuracy=0.781\n",
      "Epoch  0,  step=1582,  loss=1.589,  accuracy=0.812\n",
      "Epoch  0,  step=1583,  loss=0.396,  accuracy=0.938\n",
      "Epoch  0,  step=1584,  loss=0.891,  accuracy=0.781\n",
      "Epoch  0,  step=1585,  loss=0.989,  accuracy=0.781\n",
      "Epoch  0,  step=1586,  loss=0.625,  accuracy=0.875\n",
      "Epoch  0,  step=1587,  loss=0.556,  accuracy=0.844\n",
      "Epoch  0,  step=1588,  loss=0.648,  accuracy=0.781\n",
      "Epoch  0,  step=1589,  loss=0.399,  accuracy=0.906\n",
      "Epoch  0,  step=1590,  loss=0.829,  accuracy=0.781\n",
      "Epoch  0,  step=1591,  loss=0.303,  accuracy=0.906\n",
      "Epoch  0,  step=1592,  loss=0.354,  accuracy=0.875\n",
      "Epoch  0,  step=1593,  loss=0.215,  accuracy=0.906\n",
      "Epoch  0,  step=1594,  loss=0.274,  accuracy=0.844\n",
      "Epoch  0,  step=1595,  loss=0.884,  accuracy=0.781\n",
      "Epoch  0,  step=1596,  loss=0.457,  accuracy=0.812\n",
      "Epoch  0,  step=1597,  loss=0.583,  accuracy=0.875\n",
      "Epoch  0,  step=1598,  loss=0.525,  accuracy=0.906\n",
      "Epoch  0,  step=1599,  loss=1.154,  accuracy=0.750\n",
      "Epoch  0,  step=1600,  loss=0.804,  accuracy=0.719\n",
      "Epoch  0,  step=1601,  loss=0.935,  accuracy=0.625\n",
      "Epoch  0,  step=1602,  loss=0.909,  accuracy=0.844\n",
      "Epoch  0,  step=1603,  loss=1.153,  accuracy=0.844\n",
      "Epoch  0,  step=1604,  loss=0.791,  accuracy=0.750\n",
      "Epoch  0,  step=1605,  loss=0.338,  accuracy=0.812\n",
      "Epoch  0,  step=1606,  loss=0.272,  accuracy=0.875\n",
      "Epoch  0,  step=1607,  loss=0.395,  accuracy=0.812\n",
      "Epoch  0,  step=1608,  loss=0.304,  accuracy=0.906\n",
      "Epoch  0,  step=1609,  loss=0.567,  accuracy=0.781\n",
      "Epoch  0,  step=1610,  loss=0.871,  accuracy=0.844\n",
      "Epoch  0,  step=1611,  loss=0.576,  accuracy=0.812\n",
      "Epoch  0,  step=1612,  loss=0.109,  accuracy=0.906\n",
      "Epoch  0,  step=1613,  loss=0.650,  accuracy=0.781\n",
      "Epoch  0,  step=1614,  loss=0.314,  accuracy=0.906\n",
      "Epoch  0,  step=1615,  loss=0.573,  accuracy=0.906\n",
      "Epoch  0,  step=1616,  loss=0.897,  accuracy=0.750\n",
      "Epoch  0,  step=1617,  loss=0.695,  accuracy=0.844\n",
      "Epoch  0,  step=1618,  loss=0.895,  accuracy=0.781\n",
      "Epoch  0,  step=1619,  loss=0.104,  accuracy=0.969\n",
      "Epoch  0,  step=1620,  loss=0.565,  accuracy=0.750\n",
      "Epoch  0,  step=1621,  loss=0.323,  accuracy=0.906\n",
      "Epoch  0,  step=1622,  loss=0.656,  accuracy=0.812\n",
      "Epoch  0,  step=1623,  loss=1.243,  accuracy=0.750\n",
      "Epoch  0,  step=1624,  loss=1.553,  accuracy=0.844\n",
      "Epoch  0,  step=1625,  loss=0.256,  accuracy=0.906\n",
      "Epoch  0,  step=1626,  loss=0.170,  accuracy=0.906\n",
      "Epoch  0,  step=1627,  loss=1.195,  accuracy=0.812\n",
      "Epoch  0,  step=1628,  loss=1.259,  accuracy=0.719\n",
      "Epoch  0,  step=1629,  loss=2.079,  accuracy=0.688\n",
      "Epoch  0,  step=1630,  loss=1.486,  accuracy=0.750\n",
      "Epoch  0,  step=1631,  loss=0.556,  accuracy=0.812\n",
      "Epoch  0,  step=1632,  loss=0.990,  accuracy=0.750\n",
      "Epoch  0,  step=1633,  loss=0.615,  accuracy=0.906\n",
      "Epoch  0,  step=1634,  loss=0.297,  accuracy=0.875\n",
      "Epoch  0,  step=1635,  loss=0.617,  accuracy=0.875\n",
      "Epoch  0,  step=1636,  loss=0.714,  accuracy=0.875\n",
      "Epoch  0,  step=1637,  loss=0.743,  accuracy=0.875\n",
      "Epoch  0,  step=1638,  loss=0.381,  accuracy=0.938\n",
      "Epoch  0,  step=1639,  loss=0.621,  accuracy=0.812\n",
      "Epoch  0,  step=1640,  loss=0.288,  accuracy=0.875\n",
      "Epoch  0,  step=1641,  loss=0.255,  accuracy=0.844\n",
      "Epoch  0,  step=1642,  loss=0.535,  accuracy=0.906\n",
      "Epoch  0,  step=1643,  loss=0.260,  accuracy=0.938\n",
      "Epoch  0,  step=1644,  loss=0.395,  accuracy=0.906\n",
      "Epoch  0,  step=1645,  loss=1.331,  accuracy=0.750\n",
      "Epoch  0,  step=1646,  loss=1.223,  accuracy=0.812\n",
      "Epoch  0,  step=1647,  loss=1.046,  accuracy=0.750\n",
      "Epoch  0,  step=1648,  loss=0.561,  accuracy=0.812\n",
      "Epoch  0,  step=1649,  loss=0.660,  accuracy=0.875\n",
      "Epoch  0,  step=1650,  loss=0.836,  accuracy=0.781\n",
      "Epoch  0,  step=1651,  loss=0.228,  accuracy=0.938\n",
      "Epoch  0,  step=1652,  loss=1.027,  accuracy=0.844\n",
      "Epoch  0,  step=1653,  loss=1.054,  accuracy=0.812\n",
      "Epoch  0,  step=1654,  loss=1.413,  accuracy=0.688\n",
      "Epoch  0,  step=1655,  loss=0.813,  accuracy=0.719\n",
      "Epoch  0,  step=1656,  loss=0.646,  accuracy=0.812\n",
      "Epoch  0,  step=1657,  loss=0.815,  accuracy=0.750\n",
      "Epoch  0,  step=1658,  loss=0.775,  accuracy=0.750\n",
      "Epoch  0,  step=1659,  loss=0.619,  accuracy=0.844\n",
      "Epoch  0,  step=1660,  loss=0.818,  accuracy=0.781\n",
      "Epoch  0,  step=1661,  loss=1.331,  accuracy=0.719\n",
      "Epoch  0,  step=1662,  loss=0.424,  accuracy=0.844\n",
      "Epoch  0,  step=1663,  loss=1.048,  accuracy=0.812\n",
      "Epoch  0,  step=1664,  loss=0.244,  accuracy=0.906\n",
      "Epoch  0,  step=1665,  loss=0.498,  accuracy=0.844\n",
      "Epoch  0,  step=1666,  loss=0.573,  accuracy=0.875\n",
      "Epoch  0,  step=1667,  loss=0.313,  accuracy=0.875\n",
      "Epoch  0,  step=1668,  loss=0.480,  accuracy=0.844\n",
      "Epoch  0,  step=1669,  loss=0.372,  accuracy=0.938\n",
      "Epoch  0,  step=1670,  loss=0.873,  accuracy=0.812\n",
      "Epoch  0,  step=1671,  loss=0.522,  accuracy=0.844\n",
      "Epoch  0,  step=1672,  loss=0.633,  accuracy=0.875\n",
      "Epoch  0,  step=1673,  loss=0.816,  accuracy=0.844\n",
      "Epoch  0,  step=1674,  loss=0.923,  accuracy=0.812\n",
      "Epoch  0,  step=1675,  loss=0.716,  accuracy=0.844\n",
      "Epoch  0,  step=1676,  loss=0.555,  accuracy=0.781\n",
      "Epoch  0,  step=1677,  loss=0.420,  accuracy=0.844\n",
      "Epoch  0,  step=1678,  loss=0.256,  accuracy=0.938\n",
      "Epoch  0,  step=1679,  loss=1.134,  accuracy=0.781\n",
      "Epoch  0,  step=1680,  loss=0.261,  accuracy=0.906\n",
      "Epoch  0,  step=1681,  loss=0.548,  accuracy=0.906\n",
      "Epoch  0,  step=1682,  loss=0.557,  accuracy=0.812\n",
      "Epoch  0,  step=1683,  loss=0.825,  accuracy=0.844\n",
      "Epoch  0,  step=1684,  loss=0.997,  accuracy=0.844\n",
      "Epoch  0,  step=1685,  loss=0.618,  accuracy=0.812\n",
      "Epoch  0,  step=1686,  loss=1.550,  accuracy=0.750\n",
      "Epoch  0,  step=1687,  loss=0.882,  accuracy=0.844\n",
      "Epoch  0,  step=1688,  loss=0.584,  accuracy=0.875\n",
      "Epoch  0,  step=1689,  loss=0.617,  accuracy=0.781\n",
      "Epoch  0,  step=1690,  loss=1.179,  accuracy=0.688\n",
      "Epoch  0,  step=1691,  loss=0.177,  accuracy=0.938\n",
      "Epoch  0,  step=1692,  loss=0.124,  accuracy=0.938\n",
      "Epoch  0,  step=1693,  loss=0.875,  accuracy=0.688\n",
      "Epoch  0,  step=1694,  loss=0.468,  accuracy=0.875\n",
      "Epoch  0,  step=1695,  loss=0.778,  accuracy=0.781\n",
      "Epoch  0,  step=1696,  loss=0.538,  accuracy=0.812\n",
      "Epoch  0,  step=1697,  loss=0.173,  accuracy=0.875\n",
      "Epoch  0,  step=1698,  loss=0.410,  accuracy=0.812\n",
      "Epoch  0,  step=1699,  loss=1.331,  accuracy=0.719\n",
      "Epoch  0,  step=1700,  loss=0.274,  accuracy=0.906\n",
      "Epoch  0,  step=1701,  loss=0.995,  accuracy=0.844\n",
      "Epoch  0,  step=1702,  loss=0.339,  accuracy=0.844\n",
      "Epoch  0,  step=1703,  loss=0.892,  accuracy=0.812\n",
      "Epoch  0,  step=1704,  loss=0.840,  accuracy=0.781\n",
      "Epoch  0,  step=1705,  loss=0.758,  accuracy=0.781\n",
      "Epoch  0,  step=1706,  loss=0.408,  accuracy=0.875\n",
      "Epoch  0,  step=1707,  loss=0.793,  accuracy=0.844\n",
      "Epoch  0,  step=1708,  loss=0.366,  accuracy=0.875\n",
      "Epoch  0,  step=1709,  loss=0.197,  accuracy=0.938\n",
      "Epoch  0,  step=1710,  loss=0.080,  accuracy=0.969\n",
      "Epoch  0,  step=1711,  loss=0.784,  accuracy=0.781\n",
      "Epoch  0,  step=1712,  loss=0.739,  accuracy=0.812\n",
      "Epoch  0,  step=1713,  loss=0.760,  accuracy=0.875\n",
      "Epoch  0,  step=1714,  loss=0.577,  accuracy=0.781\n",
      "Epoch  0,  step=1715,  loss=1.065,  accuracy=0.781\n",
      "Epoch  0,  step=1716,  loss=1.830,  accuracy=0.625\n",
      "Epoch  0,  step=1717,  loss=1.476,  accuracy=0.719\n",
      "Epoch  0,  step=1718,  loss=0.879,  accuracy=0.781\n",
      "Epoch  0,  step=1719,  loss=0.670,  accuracy=0.844\n",
      "Epoch  0,  step=1720,  loss=1.202,  accuracy=0.812\n",
      "Epoch  0,  step=1721,  loss=0.496,  accuracy=0.875\n",
      "Epoch  0,  step=1722,  loss=0.609,  accuracy=0.906\n",
      "Epoch  0,  step=1723,  loss=0.240,  accuracy=0.938\n",
      "Epoch  0,  step=1724,  loss=0.600,  accuracy=0.875\n",
      "Epoch  0,  step=1725,  loss=0.448,  accuracy=0.875\n",
      "Epoch  0,  step=1726,  loss=0.227,  accuracy=0.875\n",
      "Epoch  0,  step=1727,  loss=0.325,  accuracy=0.906\n",
      "Epoch  0,  step=1728,  loss=0.917,  accuracy=0.781\n",
      "Epoch  0,  step=1729,  loss=0.808,  accuracy=0.875\n",
      "Epoch  0,  step=1730,  loss=0.088,  accuracy=0.938\n",
      "Epoch  0,  step=1731,  loss=0.196,  accuracy=0.938\n",
      "Epoch  0,  step=1732,  loss=0.746,  accuracy=0.812\n",
      "Epoch  0,  step=1733,  loss=0.713,  accuracy=0.875\n",
      "Epoch  0,  step=1734,  loss=1.095,  accuracy=0.844\n",
      "Epoch  0,  step=1735,  loss=0.831,  accuracy=0.812\n",
      "Epoch  0,  step=1736,  loss=0.114,  accuracy=0.938\n",
      "Epoch  0,  step=1737,  loss=0.534,  accuracy=0.844\n",
      "Epoch  0,  step=1738,  loss=0.564,  accuracy=0.844\n",
      "Epoch  0,  step=1739,  loss=0.489,  accuracy=0.844\n",
      "Epoch  0,  step=1740,  loss=0.203,  accuracy=0.906\n",
      "Epoch  0,  step=1741,  loss=1.346,  accuracy=0.812\n",
      "Epoch  0,  step=1742,  loss=0.323,  accuracy=0.906\n",
      "Epoch  0,  step=1743,  loss=0.398,  accuracy=0.969\n",
      "Epoch  0,  step=1744,  loss=0.149,  accuracy=0.938\n",
      "Epoch  0,  step=1745,  loss=0.980,  accuracy=0.906\n",
      "Epoch  0,  step=1746,  loss=0.349,  accuracy=0.844\n",
      "Epoch  0,  step=1747,  loss=0.485,  accuracy=0.844\n",
      "Epoch  0,  step=1748,  loss=0.130,  accuracy=0.906\n",
      "Epoch  0,  step=1749,  loss=0.607,  accuracy=0.812\n",
      "Epoch  0,  step=1750,  loss=0.603,  accuracy=0.844\n",
      "Epoch  0,  step=1751,  loss=0.223,  accuracy=0.875\n",
      "Epoch  0,  step=1752,  loss=0.276,  accuracy=0.938\n",
      "Epoch  0,  step=1753,  loss=0.297,  accuracy=0.906\n",
      "Epoch  0,  step=1754,  loss=0.709,  accuracy=0.844\n",
      "Epoch  0,  step=1755,  loss=0.577,  accuracy=0.812\n",
      "Epoch  0,  step=1756,  loss=1.191,  accuracy=0.875\n",
      "Epoch  0,  step=1757,  loss=0.202,  accuracy=0.969\n",
      "Epoch  0,  step=1758,  loss=0.753,  accuracy=0.750\n",
      "Epoch  0,  step=1759,  loss=1.037,  accuracy=0.750\n",
      "Epoch  0,  step=1760,  loss=0.519,  accuracy=0.781\n",
      "Epoch  0,  step=1761,  loss=0.680,  accuracy=0.781\n",
      "Epoch  0,  step=1762,  loss=0.386,  accuracy=0.875\n",
      "Epoch  0,  step=1763,  loss=0.589,  accuracy=0.844\n",
      "Epoch  0,  step=1764,  loss=2.435,  accuracy=0.688\n",
      "Epoch  0,  step=1765,  loss=0.512,  accuracy=0.844\n",
      "Epoch  0,  step=1766,  loss=0.150,  accuracy=0.906\n",
      "Epoch  0,  step=1767,  loss=0.089,  accuracy=0.969\n",
      "Epoch  0,  step=1768,  loss=0.690,  accuracy=0.812\n",
      "Epoch  0,  step=1769,  loss=0.843,  accuracy=0.844\n",
      "Epoch  0,  step=1770,  loss=1.092,  accuracy=0.781\n",
      "Epoch  0,  step=1771,  loss=0.421,  accuracy=0.875\n",
      "Epoch  0,  step=1772,  loss=0.180,  accuracy=0.906\n",
      "Epoch  0,  step=1773,  loss=0.210,  accuracy=0.969\n",
      "Epoch  0,  step=1774,  loss=0.376,  accuracy=0.906\n",
      "Epoch  0,  step=1775,  loss=0.534,  accuracy=0.812\n",
      "Epoch  0,  step=1776,  loss=0.366,  accuracy=0.812\n",
      "Epoch  0,  step=1777,  loss=0.150,  accuracy=0.938\n",
      "Epoch  0,  step=1778,  loss=0.480,  accuracy=0.875\n",
      "Epoch  0,  step=1779,  loss=0.147,  accuracy=0.938\n",
      "Epoch  0,  step=1780,  loss=0.706,  accuracy=0.875\n",
      "Epoch  0,  step=1781,  loss=0.535,  accuracy=0.844\n",
      "Epoch  0,  step=1782,  loss=0.943,  accuracy=0.875\n",
      "Epoch  0,  step=1783,  loss=0.966,  accuracy=0.812\n",
      "Epoch  0,  step=1784,  loss=0.143,  accuracy=0.969\n",
      "Epoch  0,  step=1785,  loss=0.413,  accuracy=0.875\n",
      "Epoch  0,  step=1786,  loss=0.681,  accuracy=0.812\n",
      "Epoch  0,  step=1787,  loss=0.385,  accuracy=0.875\n",
      "Epoch  0,  step=1788,  loss=0.691,  accuracy=0.719\n",
      "Epoch  0,  step=1789,  loss=0.438,  accuracy=0.812\n",
      "Epoch  0,  step=1790,  loss=0.956,  accuracy=0.719\n",
      "Epoch  0,  step=1791,  loss=0.888,  accuracy=0.781\n",
      "Epoch  0,  step=1792,  loss=0.448,  accuracy=0.938\n",
      "Epoch  0,  step=1793,  loss=0.450,  accuracy=0.844\n",
      "Epoch  0,  step=1794,  loss=0.324,  accuracy=0.906\n",
      "Epoch  0,  step=1795,  loss=0.270,  accuracy=0.906\n",
      "Epoch  0,  step=1796,  loss=0.815,  accuracy=0.781\n",
      "Epoch  0,  step=1797,  loss=0.606,  accuracy=0.844\n",
      "Epoch  0,  step=1798,  loss=0.841,  accuracy=0.688\n",
      "Epoch  0,  step=1799,  loss=0.897,  accuracy=0.750\n",
      "Epoch  0,  step=1800,  loss=0.907,  accuracy=0.781\n",
      "Epoch  0,  step=1801,  loss=0.783,  accuracy=0.781\n",
      "Epoch  0,  step=1802,  loss=0.939,  accuracy=0.812\n",
      "Epoch  0,  step=1803,  loss=1.074,  accuracy=0.781\n",
      "Epoch  0,  step=1804,  loss=1.173,  accuracy=0.750\n",
      "Epoch  0,  step=1805,  loss=0.353,  accuracy=0.844\n",
      "Epoch  0,  step=1806,  loss=0.659,  accuracy=0.875\n",
      "Epoch  0,  step=1807,  loss=0.343,  accuracy=0.875\n",
      "Epoch  0,  step=1808,  loss=0.313,  accuracy=0.906\n",
      "Epoch  0,  step=1809,  loss=0.498,  accuracy=0.875\n",
      "Epoch  0,  step=1810,  loss=0.478,  accuracy=0.875\n",
      "Epoch  0,  step=1811,  loss=0.681,  accuracy=0.844\n",
      "Epoch  0,  step=1812,  loss=0.257,  accuracy=0.906\n",
      "Epoch  0,  step=1813,  loss=0.359,  accuracy=0.875\n",
      "Epoch  0,  step=1814,  loss=0.682,  accuracy=0.906\n",
      "Epoch  0,  step=1815,  loss=1.198,  accuracy=0.750\n",
      "Epoch  0,  step=1816,  loss=0.305,  accuracy=0.875\n",
      "Epoch  0,  step=1817,  loss=0.121,  accuracy=0.969\n",
      "Epoch  0,  step=1818,  loss=0.035,  accuracy=1.000\n",
      "Epoch  0,  step=1819,  loss=0.115,  accuracy=0.906\n",
      "Epoch  0,  step=1820,  loss=0.231,  accuracy=0.906\n",
      "Epoch  0,  step=1821,  loss=0.166,  accuracy=0.938\n",
      "Epoch  0,  step=1822,  loss=0.309,  accuracy=0.875\n",
      "Epoch  0,  step=1823,  loss=0.057,  accuracy=0.969\n",
      "Epoch  0,  step=1824,  loss=0.392,  accuracy=0.906\n",
      "Epoch  0,  step=1825,  loss=0.581,  accuracy=0.906\n",
      "Epoch  0,  step=1826,  loss=0.276,  accuracy=0.938\n",
      "Epoch  0,  step=1827,  loss=0.411,  accuracy=0.875\n",
      "Epoch  0,  step=1828,  loss=0.496,  accuracy=0.875\n",
      "Epoch  0,  step=1829,  loss=0.468,  accuracy=0.875\n",
      "Epoch  0,  step=1830,  loss=0.503,  accuracy=0.875\n",
      "Epoch  0,  step=1831,  loss=0.453,  accuracy=0.844\n",
      "Epoch  0,  step=1832,  loss=0.408,  accuracy=0.906\n",
      "Epoch  0,  step=1833,  loss=0.232,  accuracy=0.938\n",
      "Epoch  0,  step=1834,  loss=0.143,  accuracy=0.938\n",
      "Epoch  0,  step=1835,  loss=0.102,  accuracy=0.938\n",
      "Epoch  0,  step=1836,  loss=0.213,  accuracy=0.875\n",
      "Epoch  0,  step=1837,  loss=0.852,  accuracy=0.750\n",
      "Epoch  0,  step=1838,  loss=0.412,  accuracy=0.906\n",
      "Epoch  0,  step=1839,  loss=0.838,  accuracy=0.906\n",
      "Epoch  0,  step=1840,  loss=0.155,  accuracy=0.938\n",
      "Epoch  0,  step=1841,  loss=0.034,  accuracy=1.000\n",
      "Epoch  0,  step=1842,  loss=0.097,  accuracy=0.969\n",
      "Epoch  0,  step=1843,  loss=0.459,  accuracy=0.906\n",
      "Epoch  0,  step=1844,  loss=0.156,  accuracy=0.938\n",
      "Epoch  0,  step=1845,  loss=0.509,  accuracy=0.844\n",
      "Epoch  0,  step=1846,  loss=0.348,  accuracy=0.906\n",
      "Epoch  0,  step=1847,  loss=0.190,  accuracy=0.906\n",
      "Epoch  0,  step=1848,  loss=0.399,  accuracy=0.875\n",
      "Epoch  0,  step=1849,  loss=0.493,  accuracy=0.906\n",
      "Epoch  0,  step=1850,  loss=0.375,  accuracy=0.844\n",
      "Epoch  0,  step=1851,  loss=0.401,  accuracy=0.906\n",
      "Epoch  0,  step=1852,  loss=0.647,  accuracy=0.781\n",
      "Epoch  0,  step=1853,  loss=1.279,  accuracy=0.812\n",
      "Epoch  0,  step=1854,  loss=1.158,  accuracy=0.812\n",
      "Epoch  0,  step=1855,  loss=0.808,  accuracy=0.719\n",
      "Epoch  0,  step=1856,  loss=0.643,  accuracy=0.750\n",
      "Epoch  0,  step=1857,  loss=0.275,  accuracy=0.812\n",
      "Epoch  0,  step=1858,  loss=0.271,  accuracy=0.875\n",
      "Epoch  0,  step=1859,  loss=0.220,  accuracy=0.906\n",
      "Epoch  0,  step=1860,  loss=0.196,  accuracy=0.969\n",
      "Epoch  0,  step=1861,  loss=0.315,  accuracy=0.875\n",
      "Epoch  0,  step=1862,  loss=0.162,  accuracy=0.906\n",
      "Epoch  0,  step=1863,  loss=0.173,  accuracy=0.969\n",
      "Epoch  0,  step=1864,  loss=0.452,  accuracy=0.906\n",
      "Epoch  0,  step=1865,  loss=0.479,  accuracy=0.906\n",
      "Epoch  0,  step=1866,  loss=2.407,  accuracy=0.719\n",
      "Epoch  0,  step=1867,  loss=0.302,  accuracy=0.906\n",
      "Epoch  0,  step=1868,  loss=0.682,  accuracy=0.688\n",
      "Epoch  0,  step=1869,  loss=0.288,  accuracy=0.938\n",
      "Epoch  0,  step=1870,  loss=0.098,  accuracy=0.938\n",
      "Epoch  0,  step=1871,  loss=0.092,  accuracy=0.969\n",
      "Epoch  0,  step=1872,  loss=1.757,  accuracy=0.750\n",
      "Epoch  0,  step=1873,  loss=0.250,  accuracy=0.906\n",
      "Epoch  0,  step=1874,  loss=0.269,  accuracy=0.938\n",
      "Epoch  1,  step=   0,  loss=0.668,  accuracy=0.844\n",
      "Epoch  1,  step=   1,  loss=0.680,  accuracy=0.906\n",
      "Epoch  1,  step=   2,  loss=0.789,  accuracy=0.906\n",
      "Epoch  1,  step=   3,  loss=0.387,  accuracy=0.875\n",
      "Epoch  1,  step=   4,  loss=0.725,  accuracy=0.844\n",
      "Epoch  1,  step=   5,  loss=0.794,  accuracy=0.906\n",
      "Epoch  1,  step=   6,  loss=0.184,  accuracy=0.938\n",
      "Epoch  1,  step=   7,  loss=1.013,  accuracy=0.812\n",
      "Epoch  1,  step=   8,  loss=0.926,  accuracy=0.781\n",
      "Epoch  1,  step=   9,  loss=0.177,  accuracy=0.938\n",
      "Epoch  1,  step=  10,  loss=0.496,  accuracy=0.906\n",
      "Epoch  1,  step=  11,  loss=0.431,  accuracy=0.844\n",
      "Epoch  1,  step=  12,  loss=0.459,  accuracy=0.875\n",
      "Epoch  1,  step=  13,  loss=0.916,  accuracy=0.875\n",
      "Epoch  1,  step=  14,  loss=0.330,  accuracy=0.875\n",
      "Epoch  1,  step=  15,  loss=0.775,  accuracy=0.812\n",
      "Epoch  1,  step=  16,  loss=0.377,  accuracy=0.938\n",
      "Epoch  1,  step=  17,  loss=0.364,  accuracy=0.906\n",
      "Epoch  1,  step=  18,  loss=0.510,  accuracy=0.844\n",
      "Epoch  1,  step=  19,  loss=1.189,  accuracy=0.750\n",
      "Epoch  1,  step=  20,  loss=0.633,  accuracy=0.812\n",
      "Epoch  1,  step=  21,  loss=0.313,  accuracy=0.906\n",
      "Epoch  1,  step=  22,  loss=0.459,  accuracy=0.938\n",
      "Epoch  1,  step=  23,  loss=0.414,  accuracy=0.812\n",
      "Epoch  1,  step=  24,  loss=0.774,  accuracy=0.875\n",
      "Epoch  1,  step=  25,  loss=0.374,  accuracy=0.906\n",
      "Epoch  1,  step=  26,  loss=1.387,  accuracy=0.781\n",
      "Epoch  1,  step=  27,  loss=0.928,  accuracy=0.812\n",
      "Epoch  1,  step=  28,  loss=1.440,  accuracy=0.844\n",
      "Epoch  1,  step=  29,  loss=0.468,  accuracy=0.906\n",
      "Epoch  1,  step=  30,  loss=0.312,  accuracy=0.875\n",
      "Epoch  1,  step=  31,  loss=0.847,  accuracy=0.781\n",
      "Epoch  1,  step=  32,  loss=0.834,  accuracy=0.812\n",
      "Epoch  1,  step=  33,  loss=1.636,  accuracy=0.750\n",
      "Epoch  1,  step=  34,  loss=0.698,  accuracy=0.781\n",
      "Epoch  1,  step=  35,  loss=1.711,  accuracy=0.656\n",
      "Epoch  1,  step=  36,  loss=0.358,  accuracy=0.906\n",
      "Epoch  1,  step=  37,  loss=0.375,  accuracy=0.906\n",
      "Epoch  1,  step=  38,  loss=1.436,  accuracy=0.812\n",
      "Epoch  1,  step=  39,  loss=0.863,  accuracy=0.781\n",
      "Epoch  1,  step=  40,  loss=0.384,  accuracy=0.875\n",
      "Epoch  1,  step=  41,  loss=0.517,  accuracy=0.875\n",
      "Epoch  1,  step=  42,  loss=1.373,  accuracy=0.781\n",
      "Epoch  1,  step=  43,  loss=1.042,  accuracy=0.844\n",
      "Epoch  1,  step=  44,  loss=0.513,  accuracy=0.844\n",
      "Epoch  1,  step=  45,  loss=0.716,  accuracy=0.812\n",
      "Epoch  1,  step=  46,  loss=0.364,  accuracy=0.906\n",
      "Epoch  1,  step=  47,  loss=1.239,  accuracy=0.750\n",
      "Epoch  1,  step=  48,  loss=0.536,  accuracy=0.875\n",
      "Epoch  1,  step=  49,  loss=0.973,  accuracy=0.719\n",
      "Epoch  1,  step=  50,  loss=0.661,  accuracy=0.875\n",
      "Epoch  1,  step=  51,  loss=0.462,  accuracy=0.844\n",
      "Epoch  1,  step=  52,  loss=0.437,  accuracy=0.875\n",
      "Epoch  1,  step=  53,  loss=0.194,  accuracy=0.938\n",
      "Epoch  1,  step=  54,  loss=0.348,  accuracy=0.906\n",
      "Epoch  1,  step=  55,  loss=0.203,  accuracy=0.906\n",
      "Epoch  1,  step=  56,  loss=0.778,  accuracy=0.875\n",
      "Epoch  1,  step=  57,  loss=0.682,  accuracy=0.844\n",
      "Epoch  1,  step=  58,  loss=0.239,  accuracy=0.906\n",
      "Epoch  1,  step=  59,  loss=0.175,  accuracy=0.938\n",
      "Epoch  1,  step=  60,  loss=0.343,  accuracy=0.906\n",
      "Epoch  1,  step=  61,  loss=0.435,  accuracy=0.781\n",
      "Epoch  1,  step=  62,  loss=0.738,  accuracy=0.812\n",
      "Epoch  1,  step=  63,  loss=0.385,  accuracy=0.906\n",
      "Epoch  1,  step=  64,  loss=0.555,  accuracy=0.875\n",
      "Epoch  1,  step=  65,  loss=0.509,  accuracy=0.938\n",
      "Epoch  1,  step=  66,  loss=0.478,  accuracy=0.906\n",
      "Epoch  1,  step=  67,  loss=0.180,  accuracy=0.969\n",
      "Epoch  1,  step=  68,  loss=0.313,  accuracy=0.906\n",
      "Epoch  1,  step=  69,  loss=0.769,  accuracy=0.875\n",
      "Epoch  1,  step=  70,  loss=0.479,  accuracy=0.812\n",
      "Epoch  1,  step=  71,  loss=0.368,  accuracy=0.906\n",
      "Epoch  1,  step=  72,  loss=0.352,  accuracy=0.875\n",
      "Epoch  1,  step=  73,  loss=0.118,  accuracy=0.969\n",
      "Epoch  1,  step=  74,  loss=0.198,  accuracy=0.938\n",
      "Epoch  1,  step=  75,  loss=1.312,  accuracy=0.781\n",
      "Epoch  1,  step=  76,  loss=0.345,  accuracy=0.906\n",
      "Epoch  1,  step=  77,  loss=0.194,  accuracy=0.938\n",
      "Epoch  1,  step=  78,  loss=0.163,  accuracy=0.906\n",
      "Epoch  1,  step=  79,  loss=0.653,  accuracy=0.844\n",
      "Epoch  1,  step=  80,  loss=0.320,  accuracy=0.812\n",
      "Epoch  1,  step=  81,  loss=0.315,  accuracy=0.938\n",
      "Epoch  1,  step=  82,  loss=0.706,  accuracy=0.844\n",
      "Epoch  1,  step=  83,  loss=1.516,  accuracy=0.781\n",
      "Epoch  1,  step=  84,  loss=0.791,  accuracy=0.875\n",
      "Epoch  1,  step=  85,  loss=0.342,  accuracy=0.812\n",
      "Epoch  1,  step=  86,  loss=0.639,  accuracy=0.906\n",
      "Epoch  1,  step=  87,  loss=0.897,  accuracy=0.781\n",
      "Epoch  1,  step=  88,  loss=0.611,  accuracy=0.812\n",
      "Epoch  1,  step=  89,  loss=0.361,  accuracy=0.875\n",
      "Epoch  1,  step=  90,  loss=0.477,  accuracy=0.906\n",
      "Epoch  1,  step=  91,  loss=0.747,  accuracy=0.750\n",
      "Epoch  1,  step=  92,  loss=0.591,  accuracy=0.812\n",
      "Epoch  1,  step=  93,  loss=0.514,  accuracy=0.875\n",
      "Epoch  1,  step=  94,  loss=0.475,  accuracy=0.906\n",
      "Epoch  1,  step=  95,  loss=0.244,  accuracy=0.938\n",
      "Epoch  1,  step=  96,  loss=0.260,  accuracy=0.906\n",
      "Epoch  1,  step=  97,  loss=0.508,  accuracy=0.906\n",
      "Epoch  1,  step=  98,  loss=0.561,  accuracy=0.875\n",
      "Epoch  1,  step=  99,  loss=0.182,  accuracy=0.906\n",
      "Epoch  1,  step= 100,  loss=0.382,  accuracy=0.875\n",
      "Epoch  1,  step= 101,  loss=0.544,  accuracy=0.812\n",
      "Epoch  1,  step= 102,  loss=0.760,  accuracy=0.812\n",
      "Epoch  1,  step= 103,  loss=0.178,  accuracy=0.938\n",
      "Epoch  1,  step= 104,  loss=0.173,  accuracy=0.906\n",
      "Epoch  1,  step= 105,  loss=0.713,  accuracy=0.875\n",
      "Epoch  1,  step= 106,  loss=0.743,  accuracy=0.906\n",
      "Epoch  1,  step= 107,  loss=0.371,  accuracy=0.875\n",
      "Epoch  1,  step= 108,  loss=0.432,  accuracy=0.875\n",
      "Epoch  1,  step= 109,  loss=0.831,  accuracy=0.875\n",
      "Epoch  1,  step= 110,  loss=0.498,  accuracy=0.906\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  1,  step= 111,  loss=0.369,  accuracy=0.875\n",
      "Epoch  1,  step= 112,  loss=0.292,  accuracy=0.938\n",
      "Epoch  1,  step= 113,  loss=0.438,  accuracy=0.875\n",
      "Epoch  1,  step= 114,  loss=1.112,  accuracy=0.844\n",
      "Epoch  1,  step= 115,  loss=0.945,  accuracy=0.844\n",
      "Epoch  1,  step= 116,  loss=0.947,  accuracy=0.812\n",
      "Epoch  1,  step= 117,  loss=0.184,  accuracy=0.938\n",
      "Epoch  1,  step= 118,  loss=0.835,  accuracy=0.781\n",
      "Epoch  1,  step= 119,  loss=0.790,  accuracy=0.844\n",
      "Epoch  1,  step= 120,  loss=0.190,  accuracy=0.969\n",
      "Epoch  1,  step= 121,  loss=0.513,  accuracy=0.938\n",
      "Epoch  1,  step= 122,  loss=0.136,  accuracy=0.938\n",
      "Epoch  1,  step= 123,  loss=0.214,  accuracy=0.906\n",
      "Epoch  1,  step= 124,  loss=0.932,  accuracy=0.844\n",
      "Epoch  1,  step= 125,  loss=0.631,  accuracy=0.844\n",
      "Epoch  1,  step= 126,  loss=0.451,  accuracy=0.938\n",
      "Epoch  1,  step= 127,  loss=0.609,  accuracy=0.906\n",
      "Epoch  1,  step= 128,  loss=0.359,  accuracy=0.875\n",
      "Epoch  1,  step= 129,  loss=0.925,  accuracy=0.719\n",
      "Epoch  1,  step= 130,  loss=0.871,  accuracy=0.844\n",
      "Epoch  1,  step= 131,  loss=0.520,  accuracy=0.969\n",
      "Epoch  1,  step= 132,  loss=0.333,  accuracy=0.906\n",
      "Epoch  1,  step= 133,  loss=0.620,  accuracy=0.844\n",
      "Epoch  1,  step= 134,  loss=0.402,  accuracy=0.844\n",
      "Epoch  1,  step= 135,  loss=0.306,  accuracy=0.938\n",
      "Epoch  1,  step= 136,  loss=0.815,  accuracy=0.844\n",
      "Epoch  1,  step= 137,  loss=0.206,  accuracy=0.906\n",
      "Epoch  1,  step= 138,  loss=0.240,  accuracy=0.906\n",
      "Epoch  1,  step= 139,  loss=1.238,  accuracy=0.719\n",
      "Epoch  1,  step= 140,  loss=0.508,  accuracy=0.906\n",
      "Epoch  1,  step= 141,  loss=0.065,  accuracy=0.969\n",
      "Epoch  1,  step= 142,  loss=0.125,  accuracy=0.969\n",
      "Epoch  1,  step= 143,  loss=0.827,  accuracy=0.812\n",
      "Epoch  1,  step= 144,  loss=0.443,  accuracy=0.812\n",
      "Epoch  1,  step= 145,  loss=1.010,  accuracy=0.750\n",
      "Epoch  1,  step= 146,  loss=0.368,  accuracy=0.906\n",
      "Epoch  1,  step= 147,  loss=0.454,  accuracy=0.844\n",
      "Epoch  1,  step= 148,  loss=0.567,  accuracy=0.906\n",
      "Epoch  1,  step= 149,  loss=0.442,  accuracy=0.844\n",
      "Epoch  1,  step= 150,  loss=0.737,  accuracy=0.781\n",
      "Epoch  1,  step= 151,  loss=0.139,  accuracy=0.969\n",
      "Epoch  1,  step= 152,  loss=0.337,  accuracy=0.938\n",
      "Epoch  1,  step= 153,  loss=0.399,  accuracy=0.938\n",
      "Epoch  1,  step= 154,  loss=0.542,  accuracy=0.875\n",
      "Epoch  1,  step= 155,  loss=1.001,  accuracy=0.812\n",
      "Epoch  1,  step= 156,  loss=0.461,  accuracy=0.812\n",
      "Epoch  1,  step= 157,  loss=0.917,  accuracy=0.844\n",
      "Epoch  1,  step= 158,  loss=0.998,  accuracy=0.844\n",
      "Epoch  1,  step= 159,  loss=0.703,  accuracy=0.781\n",
      "Epoch  1,  step= 160,  loss=0.838,  accuracy=0.781\n",
      "Epoch  1,  step= 161,  loss=1.170,  accuracy=0.812\n",
      "Epoch  1,  step= 162,  loss=0.354,  accuracy=0.875\n",
      "Epoch  1,  step= 163,  loss=0.171,  accuracy=0.938\n",
      "Epoch  1,  step= 164,  loss=0.664,  accuracy=0.875\n",
      "Epoch  1,  step= 165,  loss=1.430,  accuracy=0.812\n",
      "Epoch  1,  step= 166,  loss=0.791,  accuracy=0.906\n",
      "Epoch  1,  step= 167,  loss=0.756,  accuracy=0.781\n",
      "Epoch  1,  step= 168,  loss=0.412,  accuracy=0.875\n",
      "Epoch  1,  step= 169,  loss=0.390,  accuracy=0.938\n",
      "Epoch  1,  step= 170,  loss=0.239,  accuracy=0.875\n",
      "Epoch  1,  step= 171,  loss=0.522,  accuracy=0.906\n",
      "Epoch  1,  step= 172,  loss=0.233,  accuracy=0.938\n",
      "Epoch  1,  step= 173,  loss=0.830,  accuracy=0.844\n",
      "Epoch  1,  step= 174,  loss=0.207,  accuracy=0.938\n",
      "Epoch  1,  step= 175,  loss=0.814,  accuracy=0.781\n",
      "Epoch  1,  step= 176,  loss=0.640,  accuracy=0.875\n",
      "Epoch  1,  step= 177,  loss=0.807,  accuracy=0.781\n",
      "Epoch  1,  step= 178,  loss=0.890,  accuracy=0.781\n",
      "Epoch  1,  step= 179,  loss=0.227,  accuracy=0.938\n",
      "Epoch  1,  step= 180,  loss=0.750,  accuracy=0.812\n",
      "Epoch  1,  step= 181,  loss=0.234,  accuracy=0.938\n",
      "Epoch  1,  step= 182,  loss=1.018,  accuracy=0.750\n",
      "Epoch  1,  step= 183,  loss=0.411,  accuracy=0.875\n",
      "Epoch  1,  step= 184,  loss=0.423,  accuracy=0.906\n",
      "Epoch  1,  step= 185,  loss=0.677,  accuracy=0.875\n",
      "Epoch  1,  step= 186,  loss=0.346,  accuracy=0.844\n",
      "Epoch  1,  step= 187,  loss=0.121,  accuracy=0.938\n",
      "Epoch  1,  step= 188,  loss=0.366,  accuracy=0.844\n",
      "Epoch  1,  step= 189,  loss=0.127,  accuracy=0.906\n",
      "Epoch  1,  step= 190,  loss=0.781,  accuracy=0.875\n",
      "Epoch  1,  step= 191,  loss=0.397,  accuracy=0.938\n",
      "Epoch  1,  step= 192,  loss=0.234,  accuracy=0.906\n",
      "Epoch  1,  step= 193,  loss=0.122,  accuracy=0.938\n",
      "Epoch  1,  step= 194,  loss=0.431,  accuracy=0.875\n",
      "Epoch  1,  step= 195,  loss=1.093,  accuracy=0.844\n",
      "Epoch  1,  step= 196,  loss=0.646,  accuracy=0.844\n",
      "Epoch  1,  step= 197,  loss=0.747,  accuracy=0.812\n",
      "Epoch  1,  step= 198,  loss=0.444,  accuracy=0.875\n",
      "Epoch  1,  step= 199,  loss=0.288,  accuracy=0.938\n",
      "Epoch  1,  step= 200,  loss=0.687,  accuracy=0.812\n",
      "Epoch  1,  step= 201,  loss=0.706,  accuracy=0.906\n",
      "Epoch  1,  step= 202,  loss=0.516,  accuracy=0.906\n",
      "Epoch  1,  step= 203,  loss=0.557,  accuracy=0.844\n",
      "Epoch  1,  step= 204,  loss=0.575,  accuracy=0.781\n",
      "Epoch  1,  step= 205,  loss=0.565,  accuracy=0.844\n",
      "Epoch  1,  step= 206,  loss=0.472,  accuracy=0.906\n",
      "Epoch  1,  step= 207,  loss=0.278,  accuracy=0.938\n",
      "Epoch  1,  step= 208,  loss=0.213,  accuracy=0.938\n",
      "Epoch  1,  step= 209,  loss=0.225,  accuracy=0.906\n",
      "Epoch  1,  step= 210,  loss=0.539,  accuracy=0.875\n",
      "Epoch  1,  step= 211,  loss=0.168,  accuracy=0.906\n",
      "Epoch  1,  step= 212,  loss=0.578,  accuracy=0.750\n",
      "Epoch  1,  step= 213,  loss=0.689,  accuracy=0.750\n",
      "Epoch  1,  step= 214,  loss=0.514,  accuracy=0.875\n",
      "Epoch  1,  step= 215,  loss=0.607,  accuracy=0.844\n",
      "Epoch  1,  step= 216,  loss=1.302,  accuracy=0.750\n",
      "Epoch  1,  step= 217,  loss=0.458,  accuracy=0.844\n",
      "Epoch  1,  step= 218,  loss=0.249,  accuracy=0.875\n",
      "Epoch  1,  step= 219,  loss=1.174,  accuracy=0.812\n",
      "Epoch  1,  step= 220,  loss=0.443,  accuracy=0.844\n",
      "Epoch  1,  step= 221,  loss=0.319,  accuracy=0.844\n",
      "Epoch  1,  step= 222,  loss=0.294,  accuracy=0.875\n",
      "Epoch  1,  step= 223,  loss=0.717,  accuracy=0.906\n",
      "Epoch  1,  step= 224,  loss=0.310,  accuracy=0.906\n",
      "Epoch  1,  step= 225,  loss=0.948,  accuracy=0.844\n",
      "Epoch  1,  step= 226,  loss=1.007,  accuracy=0.719\n",
      "Epoch  1,  step= 227,  loss=0.789,  accuracy=0.781\n",
      "Epoch  1,  step= 228,  loss=0.618,  accuracy=0.812\n",
      "Epoch  1,  step= 229,  loss=0.915,  accuracy=0.812\n",
      "Epoch  1,  step= 230,  loss=0.499,  accuracy=0.875\n",
      "Epoch  1,  step= 231,  loss=0.125,  accuracy=0.938\n",
      "Epoch  1,  step= 232,  loss=0.554,  accuracy=0.875\n",
      "Epoch  1,  step= 233,  loss=0.070,  accuracy=1.000\n",
      "Epoch  1,  step= 234,  loss=0.215,  accuracy=0.906\n",
      "Epoch  1,  step= 235,  loss=0.841,  accuracy=0.781\n",
      "Epoch  1,  step= 236,  loss=0.147,  accuracy=0.969\n",
      "Epoch  1,  step= 237,  loss=0.985,  accuracy=0.750\n",
      "Epoch  1,  step= 238,  loss=1.340,  accuracy=0.750\n",
      "Epoch  1,  step= 239,  loss=0.327,  accuracy=0.844\n",
      "Epoch  1,  step= 240,  loss=0.302,  accuracy=0.875\n",
      "Epoch  1,  step= 241,  loss=0.984,  accuracy=0.750\n",
      "Epoch  1,  step= 242,  loss=0.679,  accuracy=0.812\n",
      "Epoch  1,  step= 243,  loss=0.740,  accuracy=0.750\n",
      "Epoch  1,  step= 244,  loss=0.407,  accuracy=0.906\n",
      "Epoch  1,  step= 245,  loss=0.436,  accuracy=0.875\n",
      "Epoch  1,  step= 246,  loss=1.078,  accuracy=0.875\n",
      "Epoch  1,  step= 247,  loss=0.476,  accuracy=0.906\n",
      "Epoch  1,  step= 248,  loss=0.349,  accuracy=0.906\n",
      "Epoch  1,  step= 249,  loss=0.659,  accuracy=0.875\n",
      "Epoch  1,  step= 250,  loss=0.651,  accuracy=0.812\n",
      "Epoch  1,  step= 251,  loss=0.333,  accuracy=0.875\n",
      "Epoch  1,  step= 252,  loss=0.356,  accuracy=0.875\n",
      "Epoch  1,  step= 253,  loss=0.738,  accuracy=0.812\n",
      "Epoch  1,  step= 254,  loss=0.270,  accuracy=0.938\n",
      "Epoch  1,  step= 255,  loss=0.251,  accuracy=0.844\n",
      "Epoch  1,  step= 256,  loss=1.348,  accuracy=0.781\n",
      "Epoch  1,  step= 257,  loss=0.566,  accuracy=0.812\n",
      "Epoch  1,  step= 258,  loss=0.285,  accuracy=0.906\n",
      "Epoch  1,  step= 259,  loss=0.370,  accuracy=0.844\n",
      "Epoch  1,  step= 260,  loss=0.291,  accuracy=0.906\n",
      "Epoch  1,  step= 261,  loss=0.177,  accuracy=0.906\n",
      "Epoch  1,  step= 262,  loss=0.337,  accuracy=0.844\n",
      "Epoch  1,  step= 263,  loss=0.860,  accuracy=0.812\n",
      "Epoch  1,  step= 264,  loss=1.023,  accuracy=0.844\n",
      "Epoch  1,  step= 265,  loss=0.407,  accuracy=0.812\n",
      "Epoch  1,  step= 266,  loss=0.496,  accuracy=0.875\n",
      "Epoch  1,  step= 267,  loss=0.514,  accuracy=0.875\n",
      "Epoch  1,  step= 268,  loss=0.275,  accuracy=0.875\n",
      "Epoch  1,  step= 269,  loss=0.745,  accuracy=0.812\n",
      "Epoch  1,  step= 270,  loss=0.465,  accuracy=0.812\n",
      "Epoch  1,  step= 271,  loss=1.690,  accuracy=0.688\n",
      "Epoch  1,  step= 272,  loss=1.690,  accuracy=0.719\n",
      "Epoch  1,  step= 273,  loss=0.786,  accuracy=0.750\n",
      "Epoch  1,  step= 274,  loss=0.670,  accuracy=0.750\n",
      "Epoch  1,  step= 275,  loss=0.670,  accuracy=0.844\n",
      "Epoch  1,  step= 276,  loss=0.616,  accuracy=0.812\n",
      "Epoch  1,  step= 277,  loss=0.947,  accuracy=0.781\n",
      "Epoch  1,  step= 278,  loss=1.158,  accuracy=0.719\n",
      "Epoch  1,  step= 279,  loss=0.626,  accuracy=0.906\n",
      "Epoch  1,  step= 280,  loss=0.525,  accuracy=0.906\n",
      "Epoch  1,  step= 281,  loss=0.541,  accuracy=0.875\n",
      "Epoch  1,  step= 282,  loss=0.223,  accuracy=0.906\n",
      "Epoch  1,  step= 283,  loss=0.415,  accuracy=0.906\n",
      "Epoch  1,  step= 284,  loss=0.823,  accuracy=0.875\n",
      "Epoch  1,  step= 285,  loss=0.710,  accuracy=0.844\n",
      "Epoch  1,  step= 286,  loss=0.530,  accuracy=0.875\n",
      "Epoch  1,  step= 287,  loss=0.527,  accuracy=0.875\n",
      "Epoch  1,  step= 288,  loss=0.250,  accuracy=0.906\n",
      "Epoch  1,  step= 289,  loss=1.059,  accuracy=0.844\n",
      "Epoch  1,  step= 290,  loss=0.714,  accuracy=0.906\n",
      "Epoch  1,  step= 291,  loss=0.347,  accuracy=0.938\n",
      "Epoch  1,  step= 292,  loss=0.473,  accuracy=0.875\n",
      "Epoch  1,  step= 293,  loss=0.988,  accuracy=0.812\n",
      "Epoch  1,  step= 294,  loss=0.529,  accuracy=0.844\n",
      "Epoch  1,  step= 295,  loss=0.594,  accuracy=0.906\n",
      "Epoch  1,  step= 296,  loss=0.216,  accuracy=0.938\n",
      "Epoch  1,  step= 297,  loss=0.482,  accuracy=0.781\n",
      "Epoch  1,  step= 298,  loss=0.381,  accuracy=0.844\n",
      "Epoch  1,  step= 299,  loss=0.406,  accuracy=0.875\n",
      "Epoch  1,  step= 300,  loss=0.579,  accuracy=0.906\n",
      "Epoch  1,  step= 301,  loss=0.578,  accuracy=0.844\n",
      "Epoch  1,  step= 302,  loss=0.772,  accuracy=0.844\n",
      "Epoch  1,  step= 303,  loss=0.067,  accuracy=0.969\n",
      "Epoch  1,  step= 304,  loss=0.402,  accuracy=0.938\n",
      "Epoch  1,  step= 305,  loss=0.960,  accuracy=0.781\n",
      "Epoch  1,  step= 306,  loss=0.789,  accuracy=0.844\n",
      "Epoch  1,  step= 307,  loss=0.169,  accuracy=0.906\n",
      "Epoch  1,  step= 308,  loss=0.405,  accuracy=0.812\n",
      "Epoch  1,  step= 309,  loss=0.202,  accuracy=0.938\n",
      "Epoch  1,  step= 310,  loss=0.425,  accuracy=0.875\n",
      "Epoch  1,  step= 311,  loss=0.442,  accuracy=0.875\n",
      "Epoch  1,  step= 312,  loss=0.265,  accuracy=0.875\n",
      "Epoch  1,  step= 313,  loss=1.185,  accuracy=0.812\n",
      "Epoch  1,  step= 314,  loss=0.375,  accuracy=0.875\n",
      "Epoch  1,  step= 315,  loss=0.593,  accuracy=0.875\n",
      "Epoch  1,  step= 316,  loss=0.522,  accuracy=0.844\n",
      "Epoch  1,  step= 317,  loss=1.123,  accuracy=0.750\n",
      "Epoch  1,  step= 318,  loss=0.954,  accuracy=0.812\n",
      "Epoch  1,  step= 319,  loss=0.710,  accuracy=0.844\n",
      "Epoch  1,  step= 320,  loss=0.866,  accuracy=0.688\n",
      "Epoch  1,  step= 321,  loss=1.089,  accuracy=0.688\n",
      "Epoch  1,  step= 322,  loss=0.434,  accuracy=0.969\n",
      "Epoch  1,  step= 323,  loss=0.616,  accuracy=0.906\n",
      "Epoch  1,  step= 324,  loss=0.242,  accuracy=0.969\n",
      "Epoch  1,  step= 325,  loss=0.346,  accuracy=0.938\n",
      "Epoch  1,  step= 326,  loss=0.458,  accuracy=0.812\n",
      "Epoch  1,  step= 327,  loss=0.249,  accuracy=0.906\n",
      "Epoch  1,  step= 328,  loss=0.551,  accuracy=0.844\n",
      "Epoch  1,  step= 329,  loss=0.106,  accuracy=1.000\n",
      "Epoch  1,  step= 330,  loss=0.351,  accuracy=0.938\n",
      "Epoch  1,  step= 331,  loss=0.229,  accuracy=0.906\n",
      "Epoch  1,  step= 332,  loss=0.462,  accuracy=0.875\n",
      "Epoch  1,  step= 333,  loss=0.349,  accuracy=0.875\n",
      "Epoch  1,  step= 334,  loss=0.414,  accuracy=0.875\n",
      "Epoch  1,  step= 335,  loss=0.998,  accuracy=0.844\n",
      "Epoch  1,  step= 336,  loss=0.653,  accuracy=0.844\n",
      "Epoch  1,  step= 337,  loss=0.341,  accuracy=0.875\n",
      "Epoch  1,  step= 338,  loss=0.378,  accuracy=0.875\n",
      "Epoch  1,  step= 339,  loss=0.566,  accuracy=0.906\n",
      "Epoch  1,  step= 340,  loss=0.310,  accuracy=0.938\n",
      "Epoch  1,  step= 341,  loss=0.196,  accuracy=0.938\n",
      "Epoch  1,  step= 342,  loss=0.151,  accuracy=0.969\n",
      "Epoch  1,  step= 343,  loss=0.596,  accuracy=0.875\n",
      "Epoch  1,  step= 344,  loss=0.397,  accuracy=0.906\n",
      "Epoch  1,  step= 345,  loss=0.406,  accuracy=0.844\n",
      "Epoch  1,  step= 346,  loss=0.192,  accuracy=0.969\n",
      "Epoch  1,  step= 347,  loss=0.350,  accuracy=0.875\n",
      "Epoch  1,  step= 348,  loss=0.313,  accuracy=0.812\n",
      "Epoch  1,  step= 349,  loss=0.180,  accuracy=0.906\n",
      "Epoch  1,  step= 350,  loss=0.787,  accuracy=0.844\n",
      "Epoch  1,  step= 351,  loss=0.715,  accuracy=0.812\n",
      "Epoch  1,  step= 352,  loss=0.375,  accuracy=0.844\n",
      "Epoch  1,  step= 353,  loss=0.364,  accuracy=0.906\n",
      "Epoch  1,  step= 354,  loss=0.310,  accuracy=0.906\n",
      "Epoch  1,  step= 355,  loss=0.557,  accuracy=0.844\n",
      "Epoch  1,  step= 356,  loss=0.442,  accuracy=0.844\n",
      "Epoch  1,  step= 357,  loss=0.276,  accuracy=0.906\n",
      "Epoch  1,  step= 358,  loss=0.199,  accuracy=0.938\n",
      "Epoch  1,  step= 359,  loss=0.566,  accuracy=0.844\n",
      "Epoch  1,  step= 360,  loss=0.431,  accuracy=0.906\n",
      "Epoch  1,  step= 361,  loss=1.230,  accuracy=0.750\n",
      "Epoch  1,  step= 362,  loss=0.685,  accuracy=0.781\n",
      "Epoch  1,  step= 363,  loss=0.987,  accuracy=0.781\n",
      "Epoch  1,  step= 364,  loss=0.549,  accuracy=0.875\n",
      "Epoch  1,  step= 365,  loss=1.148,  accuracy=0.750\n",
      "Epoch  1,  step= 366,  loss=0.510,  accuracy=0.844\n",
      "Epoch  1,  step= 367,  loss=0.950,  accuracy=0.875\n",
      "Epoch  1,  step= 368,  loss=0.584,  accuracy=0.844\n",
      "Epoch  1,  step= 369,  loss=0.302,  accuracy=0.875\n",
      "Epoch  1,  step= 370,  loss=0.650,  accuracy=0.812\n",
      "Epoch  1,  step= 371,  loss=0.301,  accuracy=0.875\n",
      "Epoch  1,  step= 372,  loss=0.182,  accuracy=0.875\n",
      "Epoch  1,  step= 373,  loss=0.414,  accuracy=0.906\n",
      "Epoch  1,  step= 374,  loss=0.529,  accuracy=0.812\n",
      "Epoch  1,  step= 375,  loss=0.593,  accuracy=0.844\n",
      "Epoch  1,  step= 376,  loss=0.722,  accuracy=0.844\n",
      "Epoch  1,  step= 377,  loss=0.531,  accuracy=0.812\n",
      "Epoch  1,  step= 378,  loss=0.531,  accuracy=0.875\n",
      "Epoch  1,  step= 379,  loss=0.231,  accuracy=0.906\n",
      "Epoch  1,  step= 380,  loss=0.440,  accuracy=0.781\n",
      "Epoch  1,  step= 381,  loss=0.506,  accuracy=0.906\n",
      "Epoch  1,  step= 382,  loss=0.326,  accuracy=0.906\n",
      "Epoch  1,  step= 383,  loss=1.083,  accuracy=0.750\n",
      "Epoch  1,  step= 384,  loss=0.468,  accuracy=0.875\n",
      "Epoch  1,  step= 385,  loss=0.504,  accuracy=0.844\n",
      "Epoch  1,  step= 386,  loss=0.700,  accuracy=0.781\n",
      "Epoch  1,  step= 387,  loss=0.145,  accuracy=0.969\n",
      "Epoch  1,  step= 388,  loss=0.764,  accuracy=0.750\n",
      "Epoch  1,  step= 389,  loss=0.625,  accuracy=0.844\n",
      "Epoch  1,  step= 390,  loss=1.149,  accuracy=0.781\n",
      "Epoch  1,  step= 391,  loss=0.403,  accuracy=0.875\n",
      "Epoch  1,  step= 392,  loss=1.299,  accuracy=0.781\n",
      "Epoch  1,  step= 393,  loss=0.988,  accuracy=0.750\n",
      "Epoch  1,  step= 394,  loss=0.858,  accuracy=0.750\n",
      "Epoch  1,  step= 395,  loss=0.878,  accuracy=0.844\n",
      "Epoch  1,  step= 396,  loss=1.826,  accuracy=0.688\n",
      "Epoch  1,  step= 397,  loss=0.159,  accuracy=0.938\n",
      "Epoch  1,  step= 398,  loss=0.501,  accuracy=0.812\n",
      "Epoch  1,  step= 399,  loss=0.642,  accuracy=0.875\n",
      "Epoch  1,  step= 400,  loss=0.358,  accuracy=0.875\n",
      "Epoch  1,  step= 401,  loss=0.569,  accuracy=0.781\n",
      "Epoch  1,  step= 402,  loss=0.615,  accuracy=0.781\n",
      "Epoch  1,  step= 403,  loss=0.355,  accuracy=0.844\n",
      "Epoch  1,  step= 404,  loss=1.138,  accuracy=0.750\n",
      "Epoch  1,  step= 405,  loss=0.714,  accuracy=0.750\n",
      "Epoch  1,  step= 406,  loss=1.098,  accuracy=0.750\n",
      "Epoch  1,  step= 407,  loss=1.175,  accuracy=0.750\n",
      "Epoch  1,  step= 408,  loss=1.615,  accuracy=0.656\n",
      "Epoch  1,  step= 409,  loss=0.624,  accuracy=0.750\n",
      "Epoch  1,  step= 410,  loss=0.666,  accuracy=0.812\n",
      "Epoch  1,  step= 411,  loss=0.389,  accuracy=0.875\n",
      "Epoch  1,  step= 412,  loss=0.917,  accuracy=0.750\n",
      "Epoch  1,  step= 413,  loss=0.109,  accuracy=0.969\n",
      "Epoch  1,  step= 414,  loss=1.018,  accuracy=0.812\n",
      "Epoch  1,  step= 415,  loss=0.679,  accuracy=0.781\n",
      "Epoch  1,  step= 416,  loss=0.384,  accuracy=0.906\n",
      "Epoch  1,  step= 417,  loss=0.424,  accuracy=0.844\n",
      "Epoch  1,  step= 418,  loss=0.773,  accuracy=0.812\n",
      "Epoch  1,  step= 419,  loss=0.348,  accuracy=0.875\n",
      "Epoch  1,  step= 420,  loss=0.474,  accuracy=0.875\n",
      "Epoch  1,  step= 421,  loss=0.050,  accuracy=1.000\n",
      "Epoch  1,  step= 422,  loss=0.695,  accuracy=0.906\n",
      "Epoch  1,  step= 423,  loss=0.611,  accuracy=0.875\n",
      "Epoch  1,  step= 424,  loss=0.071,  accuracy=0.969\n",
      "Epoch  1,  step= 425,  loss=0.347,  accuracy=0.875\n",
      "Epoch  1,  step= 426,  loss=1.108,  accuracy=0.781\n",
      "Epoch  1,  step= 427,  loss=0.859,  accuracy=0.750\n",
      "Epoch  1,  step= 428,  loss=0.956,  accuracy=0.781\n",
      "Epoch  1,  step= 429,  loss=0.639,  accuracy=0.750\n",
      "Epoch  1,  step= 430,  loss=0.162,  accuracy=0.938\n",
      "Epoch  1,  step= 431,  loss=0.277,  accuracy=0.906\n",
      "Epoch  1,  step= 432,  loss=1.075,  accuracy=0.812\n",
      "Epoch  1,  step= 433,  loss=0.448,  accuracy=0.844\n",
      "Epoch  1,  step= 434,  loss=0.340,  accuracy=0.875\n",
      "Epoch  1,  step= 435,  loss=0.382,  accuracy=0.875\n",
      "Epoch  1,  step= 436,  loss=0.715,  accuracy=0.812\n",
      "Epoch  1,  step= 437,  loss=0.850,  accuracy=0.781\n",
      "Epoch  1,  step= 438,  loss=0.366,  accuracy=0.844\n",
      "Epoch  1,  step= 439,  loss=1.071,  accuracy=0.812\n",
      "Epoch  1,  step= 440,  loss=0.324,  accuracy=0.938\n",
      "Epoch  1,  step= 441,  loss=1.043,  accuracy=0.750\n",
      "Epoch  1,  step= 442,  loss=0.352,  accuracy=0.875\n",
      "Epoch  1,  step= 443,  loss=0.630,  accuracy=0.750\n",
      "Epoch  1,  step= 444,  loss=0.995,  accuracy=0.875\n",
      "Epoch  1,  step= 445,  loss=0.783,  accuracy=0.750\n",
      "Epoch  1,  step= 446,  loss=0.464,  accuracy=0.844\n",
      "Epoch  1,  step= 447,  loss=0.508,  accuracy=0.875\n",
      "Epoch  1,  step= 448,  loss=1.793,  accuracy=0.750\n",
      "Epoch  1,  step= 449,  loss=0.785,  accuracy=0.781\n",
      "Epoch  1,  step= 450,  loss=0.364,  accuracy=0.906\n",
      "Epoch  1,  step= 451,  loss=0.460,  accuracy=0.875\n",
      "Epoch  1,  step= 452,  loss=0.334,  accuracy=0.844\n",
      "Epoch  1,  step= 453,  loss=0.590,  accuracy=0.844\n",
      "Epoch  1,  step= 454,  loss=0.964,  accuracy=0.750\n",
      "Epoch  1,  step= 455,  loss=0.585,  accuracy=0.844\n",
      "Epoch  1,  step= 456,  loss=0.755,  accuracy=0.781\n",
      "Epoch  1,  step= 457,  loss=0.691,  accuracy=0.750\n",
      "Epoch  1,  step= 458,  loss=0.867,  accuracy=0.812\n",
      "Epoch  1,  step= 459,  loss=0.822,  accuracy=0.781\n",
      "Epoch  1,  step= 460,  loss=1.313,  accuracy=0.656\n",
      "Epoch  1,  step= 461,  loss=1.807,  accuracy=0.719\n",
      "Epoch  1,  step= 462,  loss=0.823,  accuracy=0.844\n",
      "Epoch  1,  step= 463,  loss=1.044,  accuracy=0.844\n",
      "Epoch  1,  step= 464,  loss=0.423,  accuracy=0.844\n",
      "Epoch  1,  step= 465,  loss=0.424,  accuracy=0.875\n",
      "Epoch  1,  step= 466,  loss=0.386,  accuracy=0.906\n",
      "Epoch  1,  step= 467,  loss=0.271,  accuracy=0.875\n",
      "Epoch  1,  step= 468,  loss=0.579,  accuracy=0.812\n",
      "Epoch  1,  step= 469,  loss=0.146,  accuracy=0.938\n",
      "Epoch  1,  step= 470,  loss=0.627,  accuracy=0.938\n",
      "Epoch  1,  step= 471,  loss=0.160,  accuracy=0.969\n",
      "Epoch  1,  step= 472,  loss=1.740,  accuracy=0.750\n",
      "Epoch  1,  step= 473,  loss=0.891,  accuracy=0.875\n",
      "Epoch  1,  step= 474,  loss=1.142,  accuracy=0.875\n",
      "Epoch  1,  step= 475,  loss=0.437,  accuracy=0.812\n",
      "Epoch  1,  step= 476,  loss=0.262,  accuracy=0.875\n",
      "Epoch  1,  step= 477,  loss=0.588,  accuracy=0.875\n",
      "Epoch  1,  step= 478,  loss=0.112,  accuracy=0.969\n",
      "Epoch  1,  step= 479,  loss=0.685,  accuracy=0.781\n",
      "Epoch  1,  step= 480,  loss=0.372,  accuracy=0.875\n",
      "Epoch  1,  step= 481,  loss=0.437,  accuracy=0.875\n",
      "Epoch  1,  step= 482,  loss=0.706,  accuracy=0.906\n",
      "Epoch  1,  step= 483,  loss=0.353,  accuracy=0.875\n",
      "Epoch  1,  step= 484,  loss=0.744,  accuracy=0.812\n",
      "Epoch  1,  step= 485,  loss=0.655,  accuracy=0.812\n",
      "Epoch  1,  step= 486,  loss=0.334,  accuracy=0.875\n",
      "Epoch  1,  step= 487,  loss=0.647,  accuracy=0.844\n",
      "Epoch  1,  step= 488,  loss=0.368,  accuracy=0.875\n",
      "Epoch  1,  step= 489,  loss=0.443,  accuracy=0.875\n",
      "Epoch  1,  step= 490,  loss=0.437,  accuracy=0.938\n",
      "Epoch  1,  step= 491,  loss=1.288,  accuracy=0.656\n",
      "Epoch  1,  step= 492,  loss=0.977,  accuracy=0.750\n",
      "Epoch  1,  step= 493,  loss=0.643,  accuracy=0.812\n",
      "Epoch  1,  step= 494,  loss=0.443,  accuracy=0.844\n",
      "Epoch  1,  step= 495,  loss=1.184,  accuracy=0.781\n",
      "Epoch  1,  step= 496,  loss=0.507,  accuracy=0.875\n",
      "Epoch  1,  step= 497,  loss=0.483,  accuracy=0.844\n",
      "Epoch  1,  step= 498,  loss=0.722,  accuracy=0.844\n",
      "Epoch  1,  step= 499,  loss=0.908,  accuracy=0.812\n",
      "Epoch  1,  step= 500,  loss=0.993,  accuracy=0.781\n",
      "Epoch  1,  step= 501,  loss=1.050,  accuracy=0.719\n",
      "Epoch  1,  step= 502,  loss=0.611,  accuracy=0.844\n",
      "Epoch  1,  step= 503,  loss=1.058,  accuracy=0.875\n",
      "Epoch  1,  step= 504,  loss=0.263,  accuracy=0.906\n",
      "Epoch  1,  step= 505,  loss=0.222,  accuracy=0.906\n",
      "Epoch  1,  step= 506,  loss=0.863,  accuracy=0.781\n",
      "Epoch  1,  step= 507,  loss=0.083,  accuracy=0.969\n",
      "Epoch  1,  step= 508,  loss=0.789,  accuracy=0.844\n",
      "Epoch  1,  step= 509,  loss=0.307,  accuracy=0.906\n",
      "Epoch  1,  step= 510,  loss=0.142,  accuracy=0.938\n",
      "Epoch  1,  step= 511,  loss=0.704,  accuracy=0.812\n",
      "Epoch  1,  step= 512,  loss=0.207,  accuracy=0.875\n",
      "Epoch  1,  step= 513,  loss=0.497,  accuracy=0.812\n",
      "Epoch  1,  step= 514,  loss=0.412,  accuracy=0.938\n",
      "Epoch  1,  step= 515,  loss=0.396,  accuracy=0.906\n",
      "Epoch  1,  step= 516,  loss=0.518,  accuracy=0.875\n",
      "Epoch  1,  step= 517,  loss=0.870,  accuracy=0.812\n",
      "Epoch  1,  step= 518,  loss=0.489,  accuracy=0.875\n",
      "Epoch  1,  step= 519,  loss=0.420,  accuracy=0.906\n",
      "Epoch  1,  step= 520,  loss=0.681,  accuracy=0.844\n",
      "Epoch  1,  step= 521,  loss=0.805,  accuracy=0.781\n",
      "Epoch  1,  step= 522,  loss=0.338,  accuracy=0.906\n",
      "Epoch  1,  step= 523,  loss=0.681,  accuracy=0.812\n",
      "Epoch  1,  step= 524,  loss=1.275,  accuracy=0.719\n",
      "Epoch  1,  step= 525,  loss=0.519,  accuracy=0.812\n",
      "Epoch  1,  step= 526,  loss=0.490,  accuracy=0.844\n",
      "Epoch  1,  step= 527,  loss=0.593,  accuracy=0.875\n",
      "Epoch  1,  step= 528,  loss=0.559,  accuracy=0.781\n",
      "Epoch  1,  step= 529,  loss=0.775,  accuracy=0.781\n",
      "Epoch  1,  step= 530,  loss=1.421,  accuracy=0.719\n",
      "Epoch  1,  step= 531,  loss=0.997,  accuracy=0.719\n",
      "Epoch  1,  step= 532,  loss=0.558,  accuracy=0.906\n",
      "Epoch  1,  step= 533,  loss=0.318,  accuracy=0.906\n",
      "Epoch  1,  step= 534,  loss=0.572,  accuracy=0.781\n",
      "Epoch  1,  step= 535,  loss=0.410,  accuracy=0.875\n",
      "Epoch  1,  step= 536,  loss=0.540,  accuracy=0.906\n",
      "Epoch  1,  step= 537,  loss=1.102,  accuracy=0.812\n",
      "Epoch  1,  step= 538,  loss=0.627,  accuracy=0.812\n",
      "Epoch  1,  step= 539,  loss=0.411,  accuracy=0.906\n",
      "Epoch  1,  step= 540,  loss=0.088,  accuracy=1.000\n",
      "Epoch  1,  step= 541,  loss=0.103,  accuracy=0.969\n",
      "Epoch  1,  step= 542,  loss=0.198,  accuracy=0.906\n",
      "Epoch  1,  step= 543,  loss=0.465,  accuracy=0.844\n",
      "Epoch  1,  step= 544,  loss=0.132,  accuracy=0.969\n",
      "Epoch  1,  step= 545,  loss=0.240,  accuracy=0.875\n",
      "Epoch  1,  step= 546,  loss=1.063,  accuracy=0.906\n",
      "Epoch  1,  step= 547,  loss=0.386,  accuracy=0.812\n",
      "Epoch  1,  step= 548,  loss=1.002,  accuracy=0.812\n",
      "Epoch  1,  step= 549,  loss=0.785,  accuracy=0.781\n",
      "Epoch  1,  step= 550,  loss=0.250,  accuracy=0.906\n",
      "Epoch  1,  step= 551,  loss=1.056,  accuracy=0.844\n",
      "Epoch  1,  step= 552,  loss=0.517,  accuracy=0.844\n",
      "Epoch  1,  step= 553,  loss=1.124,  accuracy=0.719\n",
      "Epoch  1,  step= 554,  loss=0.381,  accuracy=0.875\n",
      "Epoch  1,  step= 555,  loss=0.485,  accuracy=0.938\n",
      "Epoch  1,  step= 556,  loss=1.021,  accuracy=0.656\n",
      "Epoch  1,  step= 557,  loss=0.600,  accuracy=0.875\n",
      "Epoch  1,  step= 558,  loss=0.389,  accuracy=0.844\n",
      "Epoch  1,  step= 559,  loss=0.363,  accuracy=0.875\n",
      "Epoch  1,  step= 560,  loss=0.405,  accuracy=0.906\n",
      "Epoch  1,  step= 561,  loss=0.412,  accuracy=0.938\n",
      "Epoch  1,  step= 562,  loss=0.306,  accuracy=0.875\n",
      "Epoch  1,  step= 563,  loss=0.410,  accuracy=0.906\n",
      "Epoch  1,  step= 564,  loss=0.523,  accuracy=0.938\n",
      "Epoch  1,  step= 565,  loss=0.817,  accuracy=0.906\n",
      "Epoch  1,  step= 566,  loss=0.089,  accuracy=0.969\n",
      "Epoch  1,  step= 567,  loss=0.506,  accuracy=0.906\n",
      "Epoch  1,  step= 568,  loss=0.139,  accuracy=0.938\n",
      "Epoch  1,  step= 569,  loss=0.143,  accuracy=0.938\n",
      "Epoch  1,  step= 570,  loss=0.741,  accuracy=0.812\n",
      "Epoch  1,  step= 571,  loss=0.285,  accuracy=0.875\n",
      "Epoch  1,  step= 572,  loss=0.286,  accuracy=0.875\n",
      "Epoch  1,  step= 573,  loss=0.246,  accuracy=0.875\n",
      "Epoch  1,  step= 574,  loss=0.382,  accuracy=0.906\n",
      "Epoch  1,  step= 575,  loss=0.674,  accuracy=0.750\n",
      "Epoch  1,  step= 576,  loss=0.308,  accuracy=0.875\n",
      "Epoch  1,  step= 577,  loss=0.976,  accuracy=0.781\n",
      "Epoch  1,  step= 578,  loss=0.816,  accuracy=0.812\n",
      "Epoch  1,  step= 579,  loss=0.333,  accuracy=0.875\n",
      "Epoch  1,  step= 580,  loss=0.622,  accuracy=0.750\n",
      "Epoch  1,  step= 581,  loss=0.306,  accuracy=0.938\n",
      "Epoch  1,  step= 582,  loss=0.941,  accuracy=0.844\n",
      "Epoch  1,  step= 583,  loss=0.365,  accuracy=0.875\n",
      "Epoch  1,  step= 584,  loss=0.170,  accuracy=0.969\n",
      "Epoch  1,  step= 585,  loss=0.476,  accuracy=0.875\n",
      "Epoch  1,  step= 586,  loss=0.093,  accuracy=0.969\n",
      "Epoch  1,  step= 587,  loss=0.200,  accuracy=0.938\n",
      "Epoch  1,  step= 588,  loss=0.539,  accuracy=0.844\n",
      "Epoch  1,  step= 589,  loss=0.265,  accuracy=0.906\n",
      "Epoch  1,  step= 590,  loss=0.267,  accuracy=0.906\n",
      "Epoch  1,  step= 591,  loss=0.578,  accuracy=0.844\n",
      "Epoch  1,  step= 592,  loss=0.079,  accuracy=0.969\n",
      "Epoch  1,  step= 593,  loss=0.412,  accuracy=0.906\n",
      "Epoch  1,  step= 594,  loss=0.254,  accuracy=0.938\n",
      "Epoch  1,  step= 595,  loss=0.206,  accuracy=0.938\n",
      "Epoch  1,  step= 596,  loss=0.882,  accuracy=0.875\n",
      "Epoch  1,  step= 597,  loss=0.837,  accuracy=0.844\n",
      "Epoch  1,  step= 598,  loss=0.868,  accuracy=0.781\n",
      "Epoch  1,  step= 599,  loss=0.402,  accuracy=0.844\n",
      "Epoch  1,  step= 600,  loss=0.524,  accuracy=0.844\n",
      "Epoch  1,  step= 601,  loss=0.584,  accuracy=0.812\n",
      "Epoch  1,  step= 602,  loss=0.850,  accuracy=0.812\n",
      "Epoch  1,  step= 603,  loss=0.598,  accuracy=0.844\n",
      "Epoch  1,  step= 604,  loss=0.488,  accuracy=0.844\n",
      "Epoch  1,  step= 605,  loss=0.639,  accuracy=0.812\n",
      "Epoch  1,  step= 606,  loss=0.429,  accuracy=0.844\n",
      "Epoch  1,  step= 607,  loss=0.194,  accuracy=0.906\n",
      "Epoch  1,  step= 608,  loss=0.303,  accuracy=0.906\n",
      "Epoch  1,  step= 609,  loss=0.763,  accuracy=0.812\n",
      "Epoch  1,  step= 610,  loss=0.712,  accuracy=0.844\n",
      "Epoch  1,  step= 611,  loss=0.464,  accuracy=0.875\n",
      "Epoch  1,  step= 612,  loss=0.475,  accuracy=0.875\n",
      "Epoch  1,  step= 613,  loss=0.502,  accuracy=0.844\n",
      "Epoch  1,  step= 614,  loss=0.161,  accuracy=0.938\n",
      "Epoch  1,  step= 615,  loss=0.143,  accuracy=0.938\n",
      "Epoch  1,  step= 616,  loss=0.114,  accuracy=0.969\n",
      "Epoch  1,  step= 617,  loss=0.101,  accuracy=0.969\n",
      "Epoch  1,  step= 618,  loss=0.574,  accuracy=0.875\n",
      "Epoch  1,  step= 619,  loss=1.343,  accuracy=0.812\n",
      "Epoch  1,  step= 620,  loss=0.541,  accuracy=0.875\n",
      "Epoch  1,  step= 621,  loss=0.114,  accuracy=0.938\n",
      "Epoch  1,  step= 622,  loss=1.035,  accuracy=0.781\n",
      "Epoch  1,  step= 623,  loss=0.565,  accuracy=0.844\n",
      "Epoch  1,  step= 624,  loss=0.360,  accuracy=0.875\n",
      "Epoch  1,  step= 625,  loss=0.723,  accuracy=0.750\n",
      "Epoch  1,  step= 626,  loss=0.664,  accuracy=0.750\n",
      "Epoch  1,  step= 627,  loss=0.474,  accuracy=0.906\n",
      "Epoch  1,  step= 628,  loss=0.559,  accuracy=0.812\n",
      "Epoch  1,  step= 629,  loss=0.448,  accuracy=0.875\n",
      "Epoch  1,  step= 630,  loss=1.217,  accuracy=0.719\n",
      "Epoch  1,  step= 631,  loss=0.281,  accuracy=0.844\n",
      "Epoch  1,  step= 632,  loss=0.721,  accuracy=0.875\n",
      "Epoch  1,  step= 633,  loss=0.313,  accuracy=0.812\n",
      "Epoch  1,  step= 634,  loss=0.255,  accuracy=0.906\n",
      "Epoch  1,  step= 635,  loss=0.602,  accuracy=0.906\n",
      "Epoch  1,  step= 636,  loss=0.248,  accuracy=0.906\n",
      "Epoch  1,  step= 637,  loss=0.267,  accuracy=0.938\n",
      "Epoch  1,  step= 638,  loss=0.101,  accuracy=1.000\n",
      "Epoch  1,  step= 639,  loss=0.221,  accuracy=0.906\n",
      "Epoch  1,  step= 640,  loss=0.070,  accuracy=0.969\n",
      "Epoch  1,  step= 641,  loss=0.620,  accuracy=0.875\n",
      "Epoch  1,  step= 642,  loss=1.089,  accuracy=0.781\n",
      "Epoch  1,  step= 643,  loss=0.289,  accuracy=0.906\n",
      "Epoch  1,  step= 644,  loss=0.353,  accuracy=0.938\n",
      "Epoch  1,  step= 645,  loss=0.721,  accuracy=0.812\n",
      "Epoch  1,  step= 646,  loss=0.341,  accuracy=0.938\n",
      "Epoch  1,  step= 647,  loss=1.180,  accuracy=0.750\n",
      "Epoch  1,  step= 648,  loss=0.339,  accuracy=0.938\n",
      "Epoch  1,  step= 649,  loss=1.076,  accuracy=0.875\n",
      "Epoch  1,  step= 650,  loss=0.863,  accuracy=0.844\n",
      "Epoch  1,  step= 651,  loss=0.294,  accuracy=0.875\n",
      "Epoch  1,  step= 652,  loss=0.639,  accuracy=0.844\n",
      "Epoch  1,  step= 653,  loss=0.780,  accuracy=0.750\n",
      "Epoch  1,  step= 654,  loss=0.922,  accuracy=0.844\n",
      "Epoch  1,  step= 655,  loss=0.866,  accuracy=0.719\n",
      "Epoch  1,  step= 656,  loss=0.662,  accuracy=0.844\n",
      "Epoch  1,  step= 657,  loss=0.256,  accuracy=0.875\n",
      "Epoch  1,  step= 658,  loss=0.334,  accuracy=0.938\n",
      "Epoch  1,  step= 659,  loss=0.382,  accuracy=0.844\n",
      "Epoch  1,  step= 660,  loss=1.040,  accuracy=0.812\n",
      "Epoch  1,  step= 661,  loss=0.427,  accuracy=0.938\n",
      "Epoch  1,  step= 662,  loss=0.802,  accuracy=0.781\n",
      "Epoch  1,  step= 663,  loss=0.102,  accuracy=0.938\n",
      "Epoch  1,  step= 664,  loss=0.184,  accuracy=0.938\n",
      "Epoch  1,  step= 665,  loss=0.926,  accuracy=0.844\n",
      "Epoch  1,  step= 666,  loss=0.384,  accuracy=0.969\n",
      "Epoch  1,  step= 667,  loss=0.392,  accuracy=0.906\n",
      "Epoch  1,  step= 668,  loss=0.532,  accuracy=0.875\n",
      "Epoch  1,  step= 669,  loss=0.904,  accuracy=0.812\n",
      "Epoch  1,  step= 670,  loss=0.356,  accuracy=0.844\n",
      "Epoch  1,  step= 671,  loss=0.237,  accuracy=0.906\n",
      "Epoch  1,  step= 672,  loss=0.363,  accuracy=0.844\n",
      "Epoch  1,  step= 673,  loss=0.323,  accuracy=0.875\n",
      "Epoch  1,  step= 674,  loss=0.714,  accuracy=0.875\n",
      "Epoch  1,  step= 675,  loss=0.769,  accuracy=0.812\n",
      "Epoch  1,  step= 676,  loss=0.483,  accuracy=0.875\n",
      "Epoch  1,  step= 677,  loss=0.320,  accuracy=0.875\n",
      "Epoch  1,  step= 678,  loss=0.704,  accuracy=0.844\n",
      "Epoch  1,  step= 679,  loss=0.357,  accuracy=0.875\n",
      "Epoch  1,  step= 680,  loss=0.407,  accuracy=0.938\n",
      "Epoch  1,  step= 681,  loss=0.137,  accuracy=0.969\n",
      "Epoch  1,  step= 682,  loss=0.179,  accuracy=0.938\n",
      "Epoch  1,  step= 683,  loss=0.214,  accuracy=0.938\n",
      "Epoch  1,  step= 684,  loss=0.375,  accuracy=0.875\n",
      "Epoch  1,  step= 685,  loss=0.509,  accuracy=0.844\n",
      "Epoch  1,  step= 686,  loss=0.309,  accuracy=0.969\n",
      "Epoch  1,  step= 687,  loss=0.348,  accuracy=0.906\n",
      "Epoch  1,  step= 688,  loss=0.179,  accuracy=0.906\n",
      "Epoch  1,  step= 689,  loss=0.444,  accuracy=0.875\n",
      "Epoch  1,  step= 690,  loss=0.455,  accuracy=0.844\n",
      "Epoch  1,  step= 691,  loss=0.711,  accuracy=0.781\n",
      "Epoch  1,  step= 692,  loss=0.513,  accuracy=0.812\n",
      "Epoch  1,  step= 693,  loss=0.815,  accuracy=0.812\n",
      "Epoch  1,  step= 694,  loss=1.764,  accuracy=0.844\n",
      "Epoch  1,  step= 695,  loss=0.210,  accuracy=0.906\n",
      "Epoch  1,  step= 696,  loss=0.843,  accuracy=0.750\n",
      "Epoch  1,  step= 697,  loss=0.208,  accuracy=0.969\n",
      "Epoch  1,  step= 698,  loss=0.051,  accuracy=1.000\n",
      "Epoch  1,  step= 699,  loss=0.139,  accuracy=0.969\n",
      "Epoch  1,  step= 700,  loss=0.250,  accuracy=0.844\n",
      "Epoch  1,  step= 701,  loss=0.071,  accuracy=1.000\n",
      "Epoch  1,  step= 702,  loss=1.240,  accuracy=0.719\n",
      "Epoch  1,  step= 703,  loss=0.827,  accuracy=0.844\n",
      "Epoch  1,  step= 704,  loss=1.025,  accuracy=0.812\n",
      "Epoch  1,  step= 705,  loss=0.955,  accuracy=0.719\n",
      "Epoch  1,  step= 706,  loss=0.801,  accuracy=0.719\n",
      "Epoch  1,  step= 707,  loss=0.428,  accuracy=0.875\n",
      "Epoch  1,  step= 708,  loss=0.479,  accuracy=0.781\n",
      "Epoch  1,  step= 709,  loss=0.824,  accuracy=0.812\n",
      "Epoch  1,  step= 710,  loss=0.427,  accuracy=0.844\n",
      "Epoch  1,  step= 711,  loss=0.715,  accuracy=0.844\n",
      "Epoch  1,  step= 712,  loss=0.752,  accuracy=0.844\n",
      "Epoch  1,  step= 713,  loss=0.468,  accuracy=0.844\n",
      "Epoch  1,  step= 714,  loss=0.164,  accuracy=0.906\n",
      "Epoch  1,  step= 715,  loss=0.276,  accuracy=0.906\n",
      "Epoch  1,  step= 716,  loss=0.197,  accuracy=0.969\n",
      "Epoch  1,  step= 717,  loss=0.026,  accuracy=1.000\n",
      "Epoch  1,  step= 718,  loss=0.095,  accuracy=0.969\n",
      "Epoch  1,  step= 719,  loss=0.586,  accuracy=0.844\n",
      "Epoch  1,  step= 720,  loss=0.381,  accuracy=0.875\n",
      "Epoch  1,  step= 721,  loss=0.764,  accuracy=0.781\n",
      "Epoch  1,  step= 722,  loss=0.284,  accuracy=0.906\n",
      "Epoch  1,  step= 723,  loss=0.661,  accuracy=0.781\n",
      "Epoch  1,  step= 724,  loss=0.441,  accuracy=0.906\n",
      "Epoch  1,  step= 725,  loss=0.455,  accuracy=0.875\n",
      "Epoch  1,  step= 726,  loss=0.413,  accuracy=0.906\n",
      "Epoch  1,  step= 727,  loss=0.283,  accuracy=0.906\n",
      "Epoch  1,  step= 728,  loss=0.506,  accuracy=0.781\n",
      "Epoch  1,  step= 729,  loss=0.307,  accuracy=0.969\n",
      "Epoch  1,  step= 730,  loss=0.817,  accuracy=0.844\n",
      "Epoch  1,  step= 731,  loss=0.424,  accuracy=0.844\n",
      "Epoch  1,  step= 732,  loss=0.425,  accuracy=0.875\n",
      "Epoch  1,  step= 733,  loss=0.914,  accuracy=0.812\n",
      "Epoch  1,  step= 734,  loss=0.514,  accuracy=0.906\n",
      "Epoch  1,  step= 735,  loss=0.142,  accuracy=0.938\n",
      "Epoch  1,  step= 736,  loss=0.260,  accuracy=0.906\n",
      "Epoch  1,  step= 737,  loss=0.806,  accuracy=0.875\n",
      "Epoch  1,  step= 738,  loss=0.533,  accuracy=0.812\n",
      "Epoch  1,  step= 739,  loss=0.294,  accuracy=0.938\n",
      "Epoch  1,  step= 740,  loss=0.654,  accuracy=0.812\n",
      "Epoch  1,  step= 741,  loss=1.330,  accuracy=0.719\n",
      "Epoch  1,  step= 742,  loss=0.214,  accuracy=0.938\n",
      "Epoch  1,  step= 743,  loss=0.233,  accuracy=0.906\n",
      "Epoch  1,  step= 744,  loss=0.873,  accuracy=0.812\n",
      "Epoch  1,  step= 745,  loss=0.676,  accuracy=0.812\n",
      "Epoch  1,  step= 746,  loss=0.502,  accuracy=0.812\n",
      "Epoch  1,  step= 747,  loss=0.348,  accuracy=0.906\n",
      "Epoch  1,  step= 748,  loss=0.736,  accuracy=0.875\n",
      "Epoch  1,  step= 749,  loss=0.325,  accuracy=0.875\n",
      "Epoch  1,  step= 750,  loss=0.326,  accuracy=0.844\n",
      "Epoch  1,  step= 751,  loss=0.604,  accuracy=0.781\n",
      "Epoch  1,  step= 752,  loss=0.137,  accuracy=0.938\n",
      "Epoch  1,  step= 753,  loss=0.805,  accuracy=0.719\n",
      "Epoch  1,  step= 754,  loss=0.201,  accuracy=0.938\n",
      "Epoch  1,  step= 755,  loss=0.625,  accuracy=0.875\n",
      "Epoch  1,  step= 756,  loss=0.897,  accuracy=0.781\n",
      "Epoch  1,  step= 757,  loss=0.288,  accuracy=0.938\n",
      "Epoch  1,  step= 758,  loss=1.303,  accuracy=0.750\n",
      "Epoch  1,  step= 759,  loss=0.370,  accuracy=0.938\n",
      "Epoch  1,  step= 760,  loss=0.161,  accuracy=0.906\n",
      "Epoch  1,  step= 761,  loss=0.646,  accuracy=0.875\n",
      "Epoch  1,  step= 762,  loss=0.361,  accuracy=0.906\n",
      "Epoch  1,  step= 763,  loss=0.332,  accuracy=0.875\n",
      "Epoch  1,  step= 764,  loss=0.479,  accuracy=0.812\n",
      "Epoch  1,  step= 765,  loss=0.493,  accuracy=0.875\n",
      "Epoch  1,  step= 766,  loss=0.852,  accuracy=0.750\n",
      "Epoch  1,  step= 767,  loss=0.346,  accuracy=0.812\n",
      "Epoch  1,  step= 768,  loss=0.864,  accuracy=0.812\n",
      "Epoch  1,  step= 769,  loss=1.011,  accuracy=0.750\n",
      "Epoch  1,  step= 770,  loss=1.060,  accuracy=0.781\n",
      "Epoch  1,  step= 771,  loss=0.500,  accuracy=0.875\n",
      "Epoch  1,  step= 772,  loss=0.898,  accuracy=0.719\n",
      "Epoch  1,  step= 773,  loss=0.597,  accuracy=0.812\n",
      "Epoch  1,  step= 774,  loss=0.866,  accuracy=0.906\n",
      "Epoch  1,  step= 775,  loss=1.038,  accuracy=0.719\n",
      "Epoch  1,  step= 776,  loss=0.279,  accuracy=0.906\n",
      "Epoch  1,  step= 777,  loss=0.605,  accuracy=0.844\n",
      "Epoch  1,  step= 778,  loss=0.393,  accuracy=0.906\n",
      "Epoch  1,  step= 779,  loss=0.816,  accuracy=0.812\n",
      "Epoch  1,  step= 780,  loss=0.535,  accuracy=0.875\n",
      "Epoch  1,  step= 781,  loss=0.224,  accuracy=0.938\n",
      "Epoch  1,  step= 782,  loss=0.281,  accuracy=0.906\n",
      "Epoch  1,  step= 783,  loss=0.162,  accuracy=0.875\n",
      "Epoch  1,  step= 784,  loss=0.851,  accuracy=0.875\n",
      "Epoch  1,  step= 785,  loss=0.754,  accuracy=0.781\n",
      "Epoch  1,  step= 786,  loss=0.779,  accuracy=0.750\n",
      "Epoch  1,  step= 787,  loss=0.800,  accuracy=0.812\n",
      "Epoch  1,  step= 788,  loss=0.482,  accuracy=0.906\n",
      "Epoch  1,  step= 789,  loss=0.306,  accuracy=0.938\n",
      "Epoch  1,  step= 790,  loss=0.760,  accuracy=0.750\n",
      "Epoch  1,  step= 791,  loss=0.833,  accuracy=0.812\n",
      "Epoch  1,  step= 792,  loss=0.112,  accuracy=0.969\n",
      "Epoch  1,  step= 793,  loss=0.317,  accuracy=0.938\n",
      "Epoch  1,  step= 794,  loss=0.026,  accuracy=1.000\n",
      "Epoch  1,  step= 795,  loss=0.696,  accuracy=0.844\n",
      "Epoch  1,  step= 796,  loss=0.641,  accuracy=0.812\n",
      "Epoch  1,  step= 797,  loss=0.272,  accuracy=0.938\n",
      "Epoch  1,  step= 798,  loss=0.729,  accuracy=0.875\n",
      "Epoch  1,  step= 799,  loss=0.300,  accuracy=0.781\n",
      "Epoch  1,  step= 800,  loss=0.284,  accuracy=0.875\n",
      "Epoch  1,  step= 801,  loss=0.557,  accuracy=0.844\n",
      "Epoch  1,  step= 802,  loss=0.534,  accuracy=0.906\n",
      "Epoch  1,  step= 803,  loss=0.393,  accuracy=0.844\n",
      "Epoch  1,  step= 804,  loss=0.631,  accuracy=0.781\n",
      "Epoch  1,  step= 805,  loss=0.464,  accuracy=0.938\n",
      "Epoch  1,  step= 806,  loss=1.059,  accuracy=0.781\n",
      "Epoch  1,  step= 807,  loss=0.309,  accuracy=0.906\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  1,  step= 808,  loss=0.098,  accuracy=0.969\n",
      "Epoch  1,  step= 809,  loss=0.509,  accuracy=0.844\n",
      "Epoch  1,  step= 810,  loss=0.576,  accuracy=0.688\n",
      "Epoch  1,  step= 811,  loss=0.611,  accuracy=0.875\n",
      "Epoch  1,  step= 812,  loss=0.108,  accuracy=0.938\n",
      "Epoch  1,  step= 813,  loss=0.238,  accuracy=0.906\n",
      "Epoch  1,  step= 814,  loss=0.791,  accuracy=0.812\n",
      "Epoch  1,  step= 815,  loss=0.073,  accuracy=0.969\n",
      "Epoch  1,  step= 816,  loss=0.263,  accuracy=0.875\n",
      "Epoch  1,  step= 817,  loss=0.155,  accuracy=0.938\n",
      "Epoch  1,  step= 818,  loss=0.452,  accuracy=0.906\n",
      "Epoch  1,  step= 819,  loss=0.370,  accuracy=0.906\n",
      "Epoch  1,  step= 820,  loss=0.246,  accuracy=0.906\n",
      "Epoch  1,  step= 821,  loss=0.801,  accuracy=0.812\n",
      "Epoch  1,  step= 822,  loss=0.243,  accuracy=0.906\n",
      "Epoch  1,  step= 823,  loss=0.986,  accuracy=0.750\n",
      "Epoch  1,  step= 824,  loss=1.058,  accuracy=0.906\n",
      "Epoch  1,  step= 825,  loss=0.388,  accuracy=0.906\n",
      "Epoch  1,  step= 826,  loss=0.472,  accuracy=0.875\n",
      "Epoch  1,  step= 827,  loss=0.481,  accuracy=0.875\n",
      "Epoch  1,  step= 828,  loss=0.459,  accuracy=0.781\n",
      "Epoch  1,  step= 829,  loss=0.577,  accuracy=0.812\n",
      "Epoch  1,  step= 830,  loss=0.506,  accuracy=0.844\n",
      "Epoch  1,  step= 831,  loss=0.660,  accuracy=0.781\n",
      "Epoch  1,  step= 832,  loss=1.199,  accuracy=0.812\n",
      "Epoch  1,  step= 833,  loss=0.490,  accuracy=0.906\n",
      "Epoch  1,  step= 834,  loss=0.283,  accuracy=0.781\n",
      "Epoch  1,  step= 835,  loss=1.311,  accuracy=0.719\n",
      "Epoch  1,  step= 836,  loss=0.567,  accuracy=0.781\n",
      "Epoch  1,  step= 837,  loss=0.657,  accuracy=0.812\n",
      "Epoch  1,  step= 838,  loss=0.269,  accuracy=0.844\n",
      "Epoch  1,  step= 839,  loss=0.926,  accuracy=0.812\n",
      "Epoch  1,  step= 840,  loss=0.887,  accuracy=0.812\n",
      "Epoch  1,  step= 841,  loss=0.274,  accuracy=0.906\n",
      "Epoch  1,  step= 842,  loss=0.205,  accuracy=0.875\n",
      "Epoch  1,  step= 843,  loss=0.277,  accuracy=0.938\n",
      "Epoch  1,  step= 844,  loss=0.383,  accuracy=0.844\n",
      "Epoch  1,  step= 845,  loss=0.132,  accuracy=0.938\n",
      "Epoch  1,  step= 846,  loss=0.551,  accuracy=0.719\n",
      "Epoch  1,  step= 847,  loss=0.268,  accuracy=0.906\n",
      "Epoch  1,  step= 848,  loss=0.664,  accuracy=0.875\n",
      "Epoch  1,  step= 849,  loss=1.528,  accuracy=0.688\n",
      "Epoch  1,  step= 850,  loss=0.691,  accuracy=0.875\n",
      "Epoch  1,  step= 851,  loss=0.324,  accuracy=0.875\n",
      "Epoch  1,  step= 852,  loss=0.342,  accuracy=0.938\n",
      "Epoch  1,  step= 853,  loss=0.501,  accuracy=0.844\n",
      "Epoch  1,  step= 854,  loss=0.336,  accuracy=0.875\n",
      "Epoch  1,  step= 855,  loss=0.535,  accuracy=0.812\n",
      "Epoch  1,  step= 856,  loss=0.451,  accuracy=0.875\n",
      "Epoch  1,  step= 857,  loss=0.619,  accuracy=0.844\n",
      "Epoch  1,  step= 858,  loss=0.196,  accuracy=0.938\n",
      "Epoch  1,  step= 859,  loss=0.606,  accuracy=0.875\n",
      "Epoch  1,  step= 860,  loss=0.178,  accuracy=0.969\n",
      "Epoch  1,  step= 861,  loss=0.110,  accuracy=0.938\n",
      "Epoch  1,  step= 862,  loss=0.375,  accuracy=0.875\n",
      "Epoch  1,  step= 863,  loss=0.452,  accuracy=0.906\n",
      "Epoch  1,  step= 864,  loss=0.212,  accuracy=0.938\n",
      "Epoch  1,  step= 865,  loss=1.338,  accuracy=0.750\n",
      "Epoch  1,  step= 866,  loss=0.718,  accuracy=0.750\n",
      "Epoch  1,  step= 867,  loss=0.087,  accuracy=1.000\n",
      "Epoch  1,  step= 868,  loss=0.340,  accuracy=0.844\n",
      "Epoch  1,  step= 869,  loss=0.513,  accuracy=0.875\n",
      "Epoch  1,  step= 870,  loss=0.507,  accuracy=0.906\n",
      "Epoch  1,  step= 871,  loss=0.998,  accuracy=0.750\n",
      "Epoch  1,  step= 872,  loss=0.099,  accuracy=0.969\n",
      "Epoch  1,  step= 873,  loss=0.535,  accuracy=0.875\n",
      "Epoch  1,  step= 874,  loss=0.393,  accuracy=0.969\n",
      "Epoch  1,  step= 875,  loss=0.933,  accuracy=0.844\n",
      "Epoch  1,  step= 876,  loss=0.099,  accuracy=0.938\n",
      "Epoch  1,  step= 877,  loss=0.635,  accuracy=0.875\n",
      "Epoch  1,  step= 878,  loss=0.370,  accuracy=0.938\n",
      "Epoch  1,  step= 879,  loss=0.525,  accuracy=0.812\n",
      "Epoch  1,  step= 880,  loss=0.412,  accuracy=0.906\n",
      "Epoch  1,  step= 881,  loss=0.684,  accuracy=0.844\n",
      "Epoch  1,  step= 882,  loss=0.540,  accuracy=0.875\n",
      "Epoch  1,  step= 883,  loss=0.201,  accuracy=0.906\n",
      "Epoch  1,  step= 884,  loss=0.556,  accuracy=0.781\n",
      "Epoch  1,  step= 885,  loss=0.058,  accuracy=0.969\n",
      "Epoch  1,  step= 886,  loss=1.673,  accuracy=0.750\n",
      "Epoch  1,  step= 887,  loss=0.484,  accuracy=0.938\n",
      "Epoch  1,  step= 888,  loss=0.108,  accuracy=0.938\n",
      "Epoch  1,  step= 889,  loss=0.326,  accuracy=0.875\n",
      "Epoch  1,  step= 890,  loss=0.755,  accuracy=0.906\n",
      "Epoch  1,  step= 891,  loss=0.781,  accuracy=0.875\n",
      "Epoch  1,  step= 892,  loss=0.353,  accuracy=0.969\n",
      "Epoch  1,  step= 893,  loss=0.337,  accuracy=0.938\n",
      "Epoch  1,  step= 894,  loss=0.946,  accuracy=0.781\n",
      "Epoch  1,  step= 895,  loss=1.591,  accuracy=0.719\n",
      "Epoch  1,  step= 896,  loss=0.117,  accuracy=0.969\n",
      "Epoch  1,  step= 897,  loss=0.486,  accuracy=0.875\n",
      "Epoch  1,  step= 898,  loss=0.127,  accuracy=0.938\n",
      "Epoch  1,  step= 899,  loss=0.535,  accuracy=0.781\n",
      "Epoch  1,  step= 900,  loss=0.392,  accuracy=0.844\n",
      "Epoch  1,  step= 901,  loss=0.468,  accuracy=0.812\n",
      "Epoch  1,  step= 902,  loss=0.297,  accuracy=0.875\n",
      "Epoch  1,  step= 903,  loss=0.636,  accuracy=0.812\n",
      "Epoch  1,  step= 904,  loss=0.511,  accuracy=0.781\n",
      "Epoch  1,  step= 905,  loss=0.493,  accuracy=0.906\n",
      "Epoch  1,  step= 906,  loss=0.533,  accuracy=0.844\n",
      "Epoch  1,  step= 907,  loss=0.387,  accuracy=0.781\n",
      "Epoch  1,  step= 908,  loss=0.647,  accuracy=0.875\n",
      "Epoch  1,  step= 909,  loss=0.718,  accuracy=0.781\n",
      "Epoch  1,  step= 910,  loss=1.100,  accuracy=0.656\n",
      "Epoch  1,  step= 911,  loss=0.496,  accuracy=0.750\n",
      "Epoch  1,  step= 912,  loss=0.501,  accuracy=0.844\n",
      "Epoch  1,  step= 913,  loss=0.397,  accuracy=0.781\n",
      "Epoch  1,  step= 914,  loss=0.312,  accuracy=0.906\n",
      "Epoch  1,  step= 915,  loss=0.718,  accuracy=0.781\n",
      "Epoch  1,  step= 916,  loss=0.919,  accuracy=0.812\n",
      "Epoch  1,  step= 917,  loss=0.325,  accuracy=0.938\n",
      "Epoch  1,  step= 918,  loss=0.254,  accuracy=0.938\n",
      "Epoch  1,  step= 919,  loss=0.558,  accuracy=0.938\n",
      "Epoch  1,  step= 920,  loss=0.240,  accuracy=0.844\n",
      "Epoch  1,  step= 921,  loss=0.419,  accuracy=0.812\n",
      "Epoch  1,  step= 922,  loss=0.355,  accuracy=0.906\n",
      "Epoch  1,  step= 923,  loss=0.472,  accuracy=0.812\n",
      "Epoch  1,  step= 924,  loss=0.709,  accuracy=0.844\n",
      "Epoch  1,  step= 925,  loss=0.236,  accuracy=0.906\n",
      "Epoch  1,  step= 926,  loss=0.538,  accuracy=0.812\n",
      "Epoch  1,  step= 927,  loss=0.521,  accuracy=0.875\n",
      "Epoch  1,  step= 928,  loss=0.938,  accuracy=0.812\n",
      "Epoch  1,  step= 929,  loss=0.473,  accuracy=0.844\n",
      "Epoch  1,  step= 930,  loss=0.252,  accuracy=0.875\n",
      "Epoch  1,  step= 931,  loss=0.262,  accuracy=0.969\n",
      "Epoch  1,  step= 932,  loss=1.342,  accuracy=0.781\n",
      "Epoch  1,  step= 933,  loss=0.796,  accuracy=0.781\n",
      "Epoch  1,  step= 934,  loss=0.378,  accuracy=0.844\n",
      "Epoch  1,  step= 935,  loss=0.896,  accuracy=0.719\n",
      "Epoch  1,  step= 936,  loss=1.028,  accuracy=0.812\n",
      "Epoch  1,  step= 937,  loss=0.329,  accuracy=0.938\n",
      "Epoch  1,  step= 938,  loss=0.677,  accuracy=0.781\n",
      "Epoch  1,  step= 939,  loss=0.643,  accuracy=0.875\n",
      "Epoch  1,  step= 940,  loss=0.845,  accuracy=0.812\n",
      "Epoch  1,  step= 941,  loss=0.913,  accuracy=0.781\n",
      "Epoch  1,  step= 942,  loss=0.850,  accuracy=0.688\n",
      "Epoch  1,  step= 943,  loss=0.728,  accuracy=0.812\n",
      "Epoch  1,  step= 944,  loss=0.414,  accuracy=0.938\n",
      "Epoch  1,  step= 945,  loss=0.323,  accuracy=0.938\n",
      "Epoch  1,  step= 946,  loss=0.081,  accuracy=0.969\n",
      "Epoch  1,  step= 947,  loss=0.547,  accuracy=0.812\n",
      "Epoch  1,  step= 948,  loss=0.470,  accuracy=0.875\n",
      "Epoch  1,  step= 949,  loss=0.393,  accuracy=0.844\n",
      "Epoch  1,  step= 950,  loss=0.308,  accuracy=0.844\n",
      "Epoch  1,  step= 951,  loss=0.722,  accuracy=0.750\n",
      "Epoch  1,  step= 952,  loss=0.419,  accuracy=0.875\n",
      "Epoch  1,  step= 953,  loss=0.870,  accuracy=0.656\n",
      "Epoch  1,  step= 954,  loss=0.836,  accuracy=0.812\n",
      "Epoch  1,  step= 955,  loss=0.307,  accuracy=0.938\n",
      "Epoch  1,  step= 956,  loss=1.242,  accuracy=0.781\n",
      "Epoch  1,  step= 957,  loss=0.235,  accuracy=0.938\n",
      "Epoch  1,  step= 958,  loss=0.448,  accuracy=0.812\n",
      "Epoch  1,  step= 959,  loss=0.565,  accuracy=0.781\n",
      "Epoch  1,  step= 960,  loss=0.530,  accuracy=0.812\n",
      "Epoch  1,  step= 961,  loss=0.192,  accuracy=0.938\n",
      "Epoch  1,  step= 962,  loss=0.457,  accuracy=0.844\n",
      "Epoch  1,  step= 963,  loss=0.838,  accuracy=0.844\n",
      "Epoch  1,  step= 964,  loss=0.774,  accuracy=0.844\n",
      "Epoch  1,  step= 965,  loss=1.067,  accuracy=0.812\n",
      "Epoch  1,  step= 966,  loss=0.249,  accuracy=0.875\n",
      "Epoch  1,  step= 967,  loss=0.509,  accuracy=0.875\n",
      "Epoch  1,  step= 968,  loss=0.591,  accuracy=0.844\n",
      "Epoch  1,  step= 969,  loss=0.509,  accuracy=0.781\n",
      "Epoch  1,  step= 970,  loss=0.551,  accuracy=0.938\n",
      "Epoch  1,  step= 971,  loss=0.863,  accuracy=0.875\n",
      "Epoch  1,  step= 972,  loss=1.084,  accuracy=0.719\n",
      "Epoch  1,  step= 973,  loss=0.549,  accuracy=0.906\n",
      "Epoch  1,  step= 974,  loss=1.152,  accuracy=0.844\n",
      "Epoch  1,  step= 975,  loss=0.310,  accuracy=0.906\n",
      "Epoch  1,  step= 976,  loss=0.241,  accuracy=0.906\n",
      "Epoch  1,  step= 977,  loss=1.260,  accuracy=0.750\n",
      "Epoch  1,  step= 978,  loss=1.106,  accuracy=0.719\n",
      "Epoch  1,  step= 979,  loss=0.741,  accuracy=0.781\n",
      "Epoch  1,  step= 980,  loss=0.994,  accuracy=0.719\n",
      "Epoch  1,  step= 981,  loss=1.201,  accuracy=0.812\n",
      "Epoch  1,  step= 982,  loss=0.760,  accuracy=0.812\n",
      "Epoch  1,  step= 983,  loss=0.512,  accuracy=0.812\n",
      "Epoch  1,  step= 984,  loss=0.381,  accuracy=0.875\n",
      "Epoch  1,  step= 985,  loss=0.264,  accuracy=0.906\n",
      "Epoch  1,  step= 986,  loss=0.467,  accuracy=0.906\n",
      "Epoch  1,  step= 987,  loss=0.493,  accuracy=0.906\n",
      "Epoch  1,  step= 988,  loss=0.703,  accuracy=0.844\n",
      "Epoch  1,  step= 989,  loss=1.624,  accuracy=0.719\n",
      "Epoch  1,  step= 990,  loss=0.739,  accuracy=0.781\n",
      "Epoch  1,  step= 991,  loss=1.014,  accuracy=0.750\n",
      "Epoch  1,  step= 992,  loss=0.384,  accuracy=0.875\n",
      "Epoch  1,  step= 993,  loss=0.359,  accuracy=0.812\n",
      "Epoch  1,  step= 994,  loss=0.150,  accuracy=0.906\n",
      "Epoch  1,  step= 995,  loss=0.657,  accuracy=0.844\n",
      "Epoch  1,  step= 996,  loss=0.374,  accuracy=0.938\n",
      "Epoch  1,  step= 997,  loss=0.416,  accuracy=0.875\n",
      "Epoch  1,  step= 998,  loss=0.688,  accuracy=0.844\n",
      "Epoch  1,  step= 999,  loss=0.265,  accuracy=0.906\n",
      "Epoch  1,  step=1000,  loss=0.466,  accuracy=0.875\n",
      "Epoch  1,  step=1001,  loss=0.491,  accuracy=0.812\n",
      "Epoch  1,  step=1002,  loss=0.443,  accuracy=0.906\n",
      "Epoch  1,  step=1003,  loss=0.232,  accuracy=0.938\n",
      "Epoch  1,  step=1004,  loss=0.677,  accuracy=0.812\n",
      "Epoch  1,  step=1005,  loss=0.754,  accuracy=0.875\n",
      "Epoch  1,  step=1006,  loss=0.391,  accuracy=0.875\n",
      "Epoch  1,  step=1007,  loss=0.622,  accuracy=0.875\n",
      "Epoch  1,  step=1008,  loss=0.559,  accuracy=0.938\n",
      "Epoch  1,  step=1009,  loss=0.612,  accuracy=0.844\n",
      "Epoch  1,  step=1010,  loss=0.972,  accuracy=0.688\n",
      "Epoch  1,  step=1011,  loss=0.786,  accuracy=0.875\n",
      "Epoch  1,  step=1012,  loss=0.498,  accuracy=0.812\n",
      "Epoch  1,  step=1013,  loss=0.741,  accuracy=0.781\n",
      "Epoch  1,  step=1014,  loss=0.695,  accuracy=0.875\n",
      "Epoch  1,  step=1015,  loss=0.463,  accuracy=0.781\n",
      "Epoch  1,  step=1016,  loss=0.254,  accuracy=0.906\n",
      "Epoch  1,  step=1017,  loss=0.525,  accuracy=0.844\n",
      "Epoch  1,  step=1018,  loss=0.346,  accuracy=0.906\n",
      "Epoch  1,  step=1019,  loss=0.682,  accuracy=0.812\n",
      "Epoch  1,  step=1020,  loss=0.552,  accuracy=0.875\n",
      "Epoch  1,  step=1021,  loss=0.307,  accuracy=0.906\n",
      "Epoch  1,  step=1022,  loss=0.432,  accuracy=0.812\n",
      "Epoch  1,  step=1023,  loss=0.198,  accuracy=0.906\n",
      "Epoch  1,  step=1024,  loss=0.802,  accuracy=0.812\n",
      "Epoch  1,  step=1025,  loss=0.641,  accuracy=0.875\n",
      "Epoch  1,  step=1026,  loss=0.579,  accuracy=0.844\n",
      "Epoch  1,  step=1027,  loss=0.635,  accuracy=0.875\n",
      "Epoch  1,  step=1028,  loss=0.694,  accuracy=0.812\n",
      "Epoch  1,  step=1029,  loss=0.356,  accuracy=0.906\n",
      "Epoch  1,  step=1030,  loss=0.105,  accuracy=0.938\n",
      "Epoch  1,  step=1031,  loss=0.151,  accuracy=0.938\n",
      "Epoch  1,  step=1032,  loss=0.511,  accuracy=0.812\n",
      "Epoch  1,  step=1033,  loss=0.667,  accuracy=0.938\n",
      "Epoch  1,  step=1034,  loss=0.216,  accuracy=0.906\n",
      "Epoch  1,  step=1035,  loss=0.391,  accuracy=0.875\n",
      "Epoch  1,  step=1036,  loss=0.324,  accuracy=0.938\n",
      "Epoch  1,  step=1037,  loss=0.580,  accuracy=0.875\n",
      "Epoch  1,  step=1038,  loss=0.683,  accuracy=0.875\n",
      "Epoch  1,  step=1039,  loss=0.220,  accuracy=0.875\n",
      "Epoch  1,  step=1040,  loss=0.693,  accuracy=0.938\n",
      "Epoch  1,  step=1041,  loss=0.240,  accuracy=0.938\n",
      "Epoch  1,  step=1042,  loss=0.496,  accuracy=0.781\n",
      "Epoch  1,  step=1043,  loss=0.327,  accuracy=0.906\n",
      "Epoch  1,  step=1044,  loss=0.439,  accuracy=0.781\n",
      "Epoch  1,  step=1045,  loss=0.617,  accuracy=0.719\n",
      "Epoch  1,  step=1046,  loss=0.356,  accuracy=0.875\n",
      "Epoch  1,  step=1047,  loss=0.424,  accuracy=0.875\n",
      "Epoch  1,  step=1048,  loss=0.103,  accuracy=0.969\n",
      "Epoch  1,  step=1049,  loss=0.911,  accuracy=0.844\n",
      "Epoch  1,  step=1050,  loss=0.319,  accuracy=0.906\n",
      "Epoch  1,  step=1051,  loss=0.380,  accuracy=0.938\n",
      "Epoch  1,  step=1052,  loss=0.392,  accuracy=0.906\n",
      "Epoch  1,  step=1053,  loss=0.148,  accuracy=0.875\n",
      "Epoch  1,  step=1054,  loss=0.326,  accuracy=0.906\n",
      "Epoch  1,  step=1055,  loss=1.112,  accuracy=0.750\n",
      "Epoch  1,  step=1056,  loss=0.465,  accuracy=0.875\n",
      "Epoch  1,  step=1057,  loss=0.282,  accuracy=0.906\n",
      "Epoch  1,  step=1058,  loss=0.111,  accuracy=0.969\n",
      "Epoch  1,  step=1059,  loss=0.296,  accuracy=0.875\n",
      "Epoch  1,  step=1060,  loss=0.384,  accuracy=0.844\n",
      "Epoch  1,  step=1061,  loss=0.190,  accuracy=0.938\n",
      "Epoch  1,  step=1062,  loss=0.421,  accuracy=0.938\n",
      "Epoch  1,  step=1063,  loss=0.315,  accuracy=0.906\n",
      "Epoch  1,  step=1064,  loss=0.547,  accuracy=0.906\n",
      "Epoch  1,  step=1065,  loss=0.736,  accuracy=0.719\n",
      "Epoch  1,  step=1066,  loss=0.704,  accuracy=0.875\n",
      "Epoch  1,  step=1067,  loss=0.162,  accuracy=0.938\n",
      "Epoch  1,  step=1068,  loss=0.471,  accuracy=0.938\n",
      "Epoch  1,  step=1069,  loss=0.240,  accuracy=0.906\n",
      "Epoch  1,  step=1070,  loss=0.170,  accuracy=0.969\n",
      "Epoch  1,  step=1071,  loss=0.142,  accuracy=0.938\n",
      "Epoch  1,  step=1072,  loss=0.469,  accuracy=0.812\n",
      "Epoch  1,  step=1073,  loss=0.448,  accuracy=0.875\n",
      "Epoch  1,  step=1074,  loss=0.131,  accuracy=0.969\n",
      "Epoch  1,  step=1075,  loss=0.602,  accuracy=0.812\n",
      "Epoch  1,  step=1076,  loss=0.743,  accuracy=0.844\n",
      "Epoch  1,  step=1077,  loss=0.553,  accuracy=0.844\n",
      "Epoch  1,  step=1078,  loss=0.739,  accuracy=0.844\n",
      "Epoch  1,  step=1079,  loss=1.249,  accuracy=0.750\n",
      "Epoch  1,  step=1080,  loss=0.245,  accuracy=0.906\n",
      "Epoch  1,  step=1081,  loss=0.534,  accuracy=0.781\n",
      "Epoch  1,  step=1082,  loss=0.498,  accuracy=0.875\n",
      "Epoch  1,  step=1083,  loss=0.777,  accuracy=0.781\n",
      "Epoch  1,  step=1084,  loss=0.667,  accuracy=0.844\n",
      "Epoch  1,  step=1085,  loss=0.551,  accuracy=0.875\n",
      "Epoch  1,  step=1086,  loss=0.307,  accuracy=0.938\n",
      "Epoch  1,  step=1087,  loss=0.469,  accuracy=0.875\n",
      "Epoch  1,  step=1088,  loss=1.025,  accuracy=0.688\n",
      "Epoch  1,  step=1089,  loss=0.480,  accuracy=0.844\n",
      "Epoch  1,  step=1090,  loss=0.463,  accuracy=0.906\n",
      "Epoch  1,  step=1091,  loss=0.314,  accuracy=0.875\n",
      "Epoch  1,  step=1092,  loss=0.158,  accuracy=0.906\n",
      "Epoch  1,  step=1093,  loss=0.169,  accuracy=0.906\n",
      "Epoch  1,  step=1094,  loss=0.161,  accuracy=0.938\n",
      "Epoch  1,  step=1095,  loss=0.530,  accuracy=0.781\n",
      "Epoch  1,  step=1096,  loss=0.373,  accuracy=0.875\n",
      "Epoch  1,  step=1097,  loss=0.488,  accuracy=0.875\n",
      "Epoch  1,  step=1098,  loss=0.789,  accuracy=0.750\n",
      "Epoch  1,  step=1099,  loss=0.390,  accuracy=0.875\n",
      "Epoch  1,  step=1100,  loss=0.558,  accuracy=0.906\n",
      "Epoch  1,  step=1101,  loss=0.744,  accuracy=0.781\n",
      "Epoch  1,  step=1102,  loss=0.138,  accuracy=0.969\n",
      "Epoch  1,  step=1103,  loss=0.237,  accuracy=0.938\n",
      "Epoch  1,  step=1104,  loss=0.111,  accuracy=0.906\n",
      "Epoch  1,  step=1105,  loss=0.363,  accuracy=0.938\n",
      "Epoch  1,  step=1106,  loss=0.316,  accuracy=0.906\n",
      "Epoch  1,  step=1107,  loss=0.630,  accuracy=0.781\n",
      "Epoch  1,  step=1108,  loss=1.458,  accuracy=0.750\n",
      "Epoch  1,  step=1109,  loss=0.381,  accuracy=0.906\n",
      "Epoch  1,  step=1110,  loss=0.288,  accuracy=0.875\n",
      "Epoch  1,  step=1111,  loss=0.438,  accuracy=0.906\n",
      "Epoch  1,  step=1112,  loss=0.225,  accuracy=0.906\n",
      "Epoch  1,  step=1113,  loss=1.149,  accuracy=0.781\n",
      "Epoch  1,  step=1114,  loss=0.146,  accuracy=0.906\n",
      "Epoch  1,  step=1115,  loss=0.598,  accuracy=0.844\n",
      "Epoch  1,  step=1116,  loss=0.201,  accuracy=0.938\n",
      "Epoch  1,  step=1117,  loss=0.323,  accuracy=0.875\n",
      "Epoch  1,  step=1118,  loss=0.271,  accuracy=0.875\n",
      "Epoch  1,  step=1119,  loss=0.291,  accuracy=0.906\n",
      "Epoch  1,  step=1120,  loss=0.623,  accuracy=0.875\n",
      "Epoch  1,  step=1121,  loss=0.284,  accuracy=0.812\n",
      "Epoch  1,  step=1122,  loss=0.415,  accuracy=0.844\n",
      "Epoch  1,  step=1123,  loss=0.500,  accuracy=0.781\n",
      "Epoch  1,  step=1124,  loss=0.690,  accuracy=0.906\n",
      "Epoch  1,  step=1125,  loss=0.749,  accuracy=0.781\n",
      "Epoch  1,  step=1126,  loss=0.699,  accuracy=0.875\n",
      "Epoch  1,  step=1127,  loss=0.688,  accuracy=0.812\n",
      "Epoch  1,  step=1128,  loss=0.956,  accuracy=0.812\n",
      "Epoch  1,  step=1129,  loss=0.299,  accuracy=0.875\n",
      "Epoch  1,  step=1130,  loss=0.518,  accuracy=0.906\n",
      "Epoch  1,  step=1131,  loss=0.250,  accuracy=0.938\n",
      "Epoch  1,  step=1132,  loss=0.485,  accuracy=0.875\n",
      "Epoch  1,  step=1133,  loss=0.428,  accuracy=0.844\n",
      "Epoch  1,  step=1134,  loss=0.046,  accuracy=1.000\n",
      "Epoch  1,  step=1135,  loss=0.337,  accuracy=0.844\n",
      "Epoch  1,  step=1136,  loss=0.156,  accuracy=0.938\n",
      "Epoch  1,  step=1137,  loss=0.436,  accuracy=0.812\n",
      "Epoch  1,  step=1138,  loss=0.803,  accuracy=0.844\n",
      "Epoch  1,  step=1139,  loss=0.803,  accuracy=0.812\n",
      "Epoch  1,  step=1140,  loss=0.652,  accuracy=0.812\n",
      "Epoch  1,  step=1141,  loss=0.440,  accuracy=0.875\n",
      "Epoch  1,  step=1142,  loss=0.156,  accuracy=0.938\n",
      "Epoch  1,  step=1143,  loss=0.355,  accuracy=0.938\n",
      "Epoch  1,  step=1144,  loss=0.041,  accuracy=1.000\n",
      "Epoch  1,  step=1145,  loss=0.630,  accuracy=0.812\n",
      "Epoch  1,  step=1146,  loss=0.145,  accuracy=0.938\n",
      "Epoch  1,  step=1147,  loss=0.409,  accuracy=0.906\n",
      "Epoch  1,  step=1148,  loss=0.291,  accuracy=0.875\n",
      "Epoch  1,  step=1149,  loss=0.163,  accuracy=0.938\n",
      "Epoch  1,  step=1150,  loss=0.530,  accuracy=0.906\n",
      "Epoch  1,  step=1151,  loss=0.250,  accuracy=0.844\n",
      "Epoch  1,  step=1152,  loss=0.402,  accuracy=0.906\n",
      "Epoch  1,  step=1153,  loss=0.191,  accuracy=0.938\n",
      "Epoch  1,  step=1154,  loss=0.266,  accuracy=0.875\n",
      "Epoch  1,  step=1155,  loss=0.199,  accuracy=0.906\n",
      "Epoch  1,  step=1156,  loss=0.154,  accuracy=0.969\n",
      "Epoch  1,  step=1157,  loss=0.549,  accuracy=0.812\n",
      "Epoch  1,  step=1158,  loss=0.510,  accuracy=0.812\n",
      "Epoch  1,  step=1159,  loss=0.659,  accuracy=0.906\n",
      "Epoch  1,  step=1160,  loss=0.420,  accuracy=0.844\n",
      "Epoch  1,  step=1161,  loss=0.476,  accuracy=0.812\n",
      "Epoch  1,  step=1162,  loss=0.512,  accuracy=0.812\n",
      "Epoch  1,  step=1163,  loss=0.112,  accuracy=0.969\n",
      "Epoch  1,  step=1164,  loss=0.951,  accuracy=0.812\n",
      "Epoch  1,  step=1165,  loss=0.520,  accuracy=0.844\n",
      "Epoch  1,  step=1166,  loss=0.534,  accuracy=0.781\n",
      "Epoch  1,  step=1167,  loss=0.936,  accuracy=0.812\n",
      "Epoch  1,  step=1168,  loss=1.436,  accuracy=0.781\n",
      "Epoch  1,  step=1169,  loss=0.925,  accuracy=0.750\n",
      "Epoch  1,  step=1170,  loss=0.968,  accuracy=0.844\n",
      "Epoch  1,  step=1171,  loss=0.404,  accuracy=0.906\n",
      "Epoch  1,  step=1172,  loss=0.461,  accuracy=0.844\n",
      "Epoch  1,  step=1173,  loss=0.985,  accuracy=0.750\n",
      "Epoch  1,  step=1174,  loss=0.846,  accuracy=0.750\n",
      "Epoch  1,  step=1175,  loss=0.569,  accuracy=0.844\n",
      "Epoch  1,  step=1176,  loss=0.472,  accuracy=0.844\n",
      "Epoch  1,  step=1177,  loss=0.307,  accuracy=0.906\n",
      "Epoch  1,  step=1178,  loss=0.672,  accuracy=0.844\n",
      "Epoch  1,  step=1179,  loss=0.930,  accuracy=0.844\n",
      "Epoch  1,  step=1180,  loss=0.517,  accuracy=0.875\n",
      "Epoch  1,  step=1181,  loss=0.777,  accuracy=0.875\n",
      "Epoch  1,  step=1182,  loss=0.860,  accuracy=0.812\n",
      "Epoch  1,  step=1183,  loss=0.259,  accuracy=0.906\n",
      "Epoch  1,  step=1184,  loss=0.434,  accuracy=0.750\n",
      "Epoch  1,  step=1185,  loss=0.634,  accuracy=0.844\n",
      "Epoch  1,  step=1186,  loss=0.287,  accuracy=0.969\n",
      "Epoch  1,  step=1187,  loss=0.531,  accuracy=0.844\n",
      "Epoch  1,  step=1188,  loss=0.416,  accuracy=0.938\n",
      "Epoch  1,  step=1189,  loss=0.635,  accuracy=0.719\n",
      "Epoch  1,  step=1190,  loss=0.314,  accuracy=0.938\n",
      "Epoch  1,  step=1191,  loss=0.371,  accuracy=0.906\n",
      "Epoch  1,  step=1192,  loss=0.437,  accuracy=0.812\n",
      "Epoch  1,  step=1193,  loss=0.205,  accuracy=0.969\n",
      "Epoch  1,  step=1194,  loss=0.560,  accuracy=0.844\n",
      "Epoch  1,  step=1195,  loss=0.886,  accuracy=0.781\n",
      "Epoch  1,  step=1196,  loss=0.270,  accuracy=0.906\n",
      "Epoch  1,  step=1197,  loss=0.357,  accuracy=0.812\n",
      "Epoch  1,  step=1198,  loss=0.124,  accuracy=0.969\n",
      "Epoch  1,  step=1199,  loss=1.111,  accuracy=0.781\n",
      "Epoch  1,  step=1200,  loss=0.225,  accuracy=0.906\n",
      "Epoch  1,  step=1201,  loss=0.492,  accuracy=0.844\n",
      "Epoch  1,  step=1202,  loss=0.112,  accuracy=0.938\n",
      "Epoch  1,  step=1203,  loss=0.984,  accuracy=0.812\n",
      "Epoch  1,  step=1204,  loss=0.257,  accuracy=0.906\n",
      "Epoch  1,  step=1205,  loss=0.604,  accuracy=0.844\n",
      "Epoch  1,  step=1206,  loss=0.635,  accuracy=0.875\n",
      "Epoch  1,  step=1207,  loss=0.268,  accuracy=0.844\n",
      "Epoch  1,  step=1208,  loss=0.461,  accuracy=0.906\n",
      "Epoch  1,  step=1209,  loss=0.285,  accuracy=0.875\n",
      "Epoch  1,  step=1210,  loss=0.238,  accuracy=0.938\n",
      "Epoch  1,  step=1211,  loss=0.705,  accuracy=0.812\n",
      "Epoch  1,  step=1212,  loss=0.379,  accuracy=0.875\n",
      "Epoch  1,  step=1213,  loss=0.111,  accuracy=0.969\n",
      "Epoch  1,  step=1214,  loss=0.067,  accuracy=1.000\n",
      "Epoch  1,  step=1215,  loss=0.265,  accuracy=0.938\n",
      "Epoch  1,  step=1216,  loss=0.225,  accuracy=0.938\n",
      "Epoch  1,  step=1217,  loss=0.206,  accuracy=0.969\n",
      "Epoch  1,  step=1218,  loss=0.500,  accuracy=0.844\n",
      "Epoch  1,  step=1219,  loss=0.165,  accuracy=0.906\n",
      "Epoch  1,  step=1220,  loss=0.447,  accuracy=0.875\n",
      "Epoch  1,  step=1221,  loss=0.144,  accuracy=0.906\n",
      "Epoch  1,  step=1222,  loss=0.834,  accuracy=0.812\n",
      "Epoch  1,  step=1223,  loss=0.572,  accuracy=0.844\n",
      "Epoch  1,  step=1224,  loss=0.333,  accuracy=0.938\n",
      "Epoch  1,  step=1225,  loss=0.648,  accuracy=0.938\n",
      "Epoch  1,  step=1226,  loss=0.638,  accuracy=0.812\n",
      "Epoch  1,  step=1227,  loss=0.405,  accuracy=0.906\n",
      "Epoch  1,  step=1228,  loss=1.778,  accuracy=0.719\n",
      "Epoch  1,  step=1229,  loss=0.913,  accuracy=0.781\n",
      "Epoch  1,  step=1230,  loss=0.341,  accuracy=0.812\n",
      "Epoch  1,  step=1231,  loss=1.096,  accuracy=0.812\n",
      "Epoch  1,  step=1232,  loss=0.612,  accuracy=0.844\n",
      "Epoch  1,  step=1233,  loss=0.493,  accuracy=0.938\n",
      "Epoch  1,  step=1234,  loss=0.298,  accuracy=0.875\n",
      "Epoch  1,  step=1235,  loss=0.560,  accuracy=0.906\n",
      "Epoch  1,  step=1236,  loss=0.377,  accuracy=0.938\n",
      "Epoch  1,  step=1237,  loss=0.803,  accuracy=0.844\n",
      "Epoch  1,  step=1238,  loss=0.466,  accuracy=0.812\n",
      "Epoch  1,  step=1239,  loss=0.952,  accuracy=0.812\n",
      "Epoch  1,  step=1240,  loss=0.891,  accuracy=0.844\n",
      "Epoch  1,  step=1241,  loss=0.315,  accuracy=0.875\n",
      "Epoch  1,  step=1242,  loss=0.879,  accuracy=0.812\n",
      "Epoch  1,  step=1243,  loss=0.924,  accuracy=0.719\n",
      "Epoch  1,  step=1244,  loss=0.444,  accuracy=0.906\n",
      "Epoch  1,  step=1245,  loss=0.672,  accuracy=0.875\n",
      "Epoch  1,  step=1246,  loss=0.644,  accuracy=0.812\n",
      "Epoch  1,  step=1247,  loss=0.739,  accuracy=0.812\n",
      "Epoch  1,  step=1248,  loss=0.612,  accuracy=0.812\n",
      "Epoch  1,  step=1249,  loss=0.629,  accuracy=0.875\n",
      "Epoch  1,  step=1250,  loss=0.380,  accuracy=0.875\n",
      "Epoch  1,  step=1251,  loss=0.490,  accuracy=0.812\n",
      "Epoch  1,  step=1252,  loss=0.443,  accuracy=0.875\n",
      "Epoch  1,  step=1253,  loss=0.382,  accuracy=0.875\n",
      "Epoch  1,  step=1254,  loss=0.378,  accuracy=0.906\n",
      "Epoch  1,  step=1255,  loss=0.354,  accuracy=0.906\n",
      "Epoch  1,  step=1256,  loss=0.379,  accuracy=0.938\n",
      "Epoch  1,  step=1257,  loss=0.519,  accuracy=0.875\n",
      "Epoch  1,  step=1258,  loss=0.357,  accuracy=0.875\n",
      "Epoch  1,  step=1259,  loss=0.551,  accuracy=0.875\n",
      "Epoch  1,  step=1260,  loss=0.299,  accuracy=0.844\n",
      "Epoch  1,  step=1261,  loss=0.129,  accuracy=0.906\n",
      "Epoch  1,  step=1262,  loss=0.431,  accuracy=0.781\n",
      "Epoch  1,  step=1263,  loss=0.266,  accuracy=0.844\n",
      "Epoch  1,  step=1264,  loss=0.137,  accuracy=0.938\n",
      "Epoch  1,  step=1265,  loss=0.676,  accuracy=0.781\n",
      "Epoch  1,  step=1266,  loss=0.534,  accuracy=0.812\n",
      "Epoch  1,  step=1267,  loss=0.351,  accuracy=0.875\n",
      "Epoch  1,  step=1268,  loss=0.493,  accuracy=0.812\n",
      "Epoch  1,  step=1269,  loss=0.333,  accuracy=0.938\n",
      "Epoch  1,  step=1270,  loss=0.966,  accuracy=0.781\n",
      "Epoch  1,  step=1271,  loss=0.191,  accuracy=0.906\n",
      "Epoch  1,  step=1272,  loss=0.297,  accuracy=0.938\n",
      "Epoch  1,  step=1273,  loss=0.255,  accuracy=0.906\n",
      "Epoch  1,  step=1274,  loss=0.121,  accuracy=0.938\n",
      "Epoch  1,  step=1275,  loss=0.144,  accuracy=0.938\n",
      "Epoch  1,  step=1276,  loss=0.168,  accuracy=0.906\n",
      "Epoch  1,  step=1277,  loss=0.324,  accuracy=0.875\n",
      "Epoch  1,  step=1278,  loss=0.115,  accuracy=0.969\n",
      "Epoch  1,  step=1279,  loss=0.287,  accuracy=0.938\n",
      "Epoch  1,  step=1280,  loss=0.551,  accuracy=0.812\n",
      "Epoch  1,  step=1281,  loss=0.931,  accuracy=0.781\n",
      "Epoch  1,  step=1282,  loss=0.279,  accuracy=0.969\n",
      "Epoch  1,  step=1283,  loss=0.306,  accuracy=0.906\n",
      "Epoch  1,  step=1284,  loss=0.305,  accuracy=0.875\n",
      "Epoch  1,  step=1285,  loss=0.290,  accuracy=0.938\n",
      "Epoch  1,  step=1286,  loss=0.467,  accuracy=0.812\n",
      "Epoch  1,  step=1287,  loss=0.226,  accuracy=0.969\n",
      "Epoch  1,  step=1288,  loss=0.429,  accuracy=0.812\n",
      "Epoch  1,  step=1289,  loss=0.912,  accuracy=0.812\n",
      "Epoch  1,  step=1290,  loss=0.832,  accuracy=0.844\n",
      "Epoch  1,  step=1291,  loss=0.677,  accuracy=0.812\n",
      "Epoch  1,  step=1292,  loss=0.452,  accuracy=0.781\n",
      "Epoch  1,  step=1293,  loss=0.412,  accuracy=0.906\n",
      "Epoch  1,  step=1294,  loss=0.904,  accuracy=0.812\n",
      "Epoch  1,  step=1295,  loss=0.781,  accuracy=0.812\n",
      "Epoch  1,  step=1296,  loss=0.401,  accuracy=0.812\n",
      "Epoch  1,  step=1297,  loss=0.437,  accuracy=0.844\n",
      "Epoch  1,  step=1298,  loss=0.621,  accuracy=0.812\n",
      "Epoch  1,  step=1299,  loss=0.779,  accuracy=0.812\n",
      "Epoch  1,  step=1300,  loss=0.518,  accuracy=0.844\n",
      "Epoch  1,  step=1301,  loss=0.091,  accuracy=0.938\n",
      "Epoch  1,  step=1302,  loss=0.237,  accuracy=0.938\n",
      "Epoch  1,  step=1303,  loss=0.409,  accuracy=0.875\n",
      "Epoch  1,  step=1304,  loss=0.596,  accuracy=0.812\n",
      "Epoch  1,  step=1305,  loss=0.574,  accuracy=0.844\n",
      "Epoch  1,  step=1306,  loss=0.646,  accuracy=0.781\n",
      "Epoch  1,  step=1307,  loss=0.307,  accuracy=0.906\n",
      "Epoch  1,  step=1308,  loss=0.627,  accuracy=0.844\n",
      "Epoch  1,  step=1309,  loss=0.659,  accuracy=0.781\n",
      "Epoch  1,  step=1310,  loss=0.679,  accuracy=0.906\n",
      "Epoch  1,  step=1311,  loss=0.659,  accuracy=0.812\n",
      "Epoch  1,  step=1312,  loss=0.697,  accuracy=0.812\n",
      "Epoch  1,  step=1313,  loss=0.666,  accuracy=0.812\n",
      "Epoch  1,  step=1314,  loss=0.564,  accuracy=0.875\n",
      "Epoch  1,  step=1315,  loss=0.342,  accuracy=0.906\n",
      "Epoch  1,  step=1316,  loss=0.487,  accuracy=0.812\n",
      "Epoch  1,  step=1317,  loss=0.206,  accuracy=0.938\n",
      "Epoch  1,  step=1318,  loss=0.526,  accuracy=0.844\n",
      "Epoch  1,  step=1319,  loss=0.341,  accuracy=0.875\n",
      "Epoch  1,  step=1320,  loss=0.497,  accuracy=0.844\n",
      "Epoch  1,  step=1321,  loss=0.131,  accuracy=0.938\n",
      "Epoch  1,  step=1322,  loss=1.074,  accuracy=0.750\n",
      "Epoch  1,  step=1323,  loss=0.963,  accuracy=0.750\n",
      "Epoch  1,  step=1324,  loss=0.379,  accuracy=0.938\n",
      "Epoch  1,  step=1325,  loss=0.796,  accuracy=0.875\n",
      "Epoch  1,  step=1326,  loss=0.474,  accuracy=0.875\n",
      "Epoch  1,  step=1327,  loss=0.564,  accuracy=0.844\n",
      "Epoch  1,  step=1328,  loss=1.013,  accuracy=0.781\n",
      "Epoch  1,  step=1329,  loss=0.360,  accuracy=0.875\n",
      "Epoch  1,  step=1330,  loss=0.272,  accuracy=0.938\n",
      "Epoch  1,  step=1331,  loss=0.367,  accuracy=0.844\n",
      "Epoch  1,  step=1332,  loss=0.465,  accuracy=0.875\n",
      "Epoch  1,  step=1333,  loss=0.676,  accuracy=0.781\n",
      "Epoch  1,  step=1334,  loss=0.700,  accuracy=0.781\n",
      "Epoch  1,  step=1335,  loss=0.241,  accuracy=0.938\n",
      "Epoch  1,  step=1336,  loss=0.613,  accuracy=0.906\n",
      "Epoch  1,  step=1337,  loss=0.608,  accuracy=0.906\n",
      "Epoch  1,  step=1338,  loss=0.738,  accuracy=0.875\n",
      "Epoch  1,  step=1339,  loss=0.937,  accuracy=0.875\n",
      "Epoch  1,  step=1340,  loss=0.607,  accuracy=0.844\n",
      "Epoch  1,  step=1341,  loss=0.524,  accuracy=0.906\n",
      "Epoch  1,  step=1342,  loss=0.532,  accuracy=0.844\n",
      "Epoch  1,  step=1343,  loss=0.304,  accuracy=0.844\n",
      "Epoch  1,  step=1344,  loss=0.917,  accuracy=0.812\n",
      "Epoch  1,  step=1345,  loss=0.304,  accuracy=0.844\n",
      "Epoch  1,  step=1346,  loss=0.726,  accuracy=0.812\n",
      "Epoch  1,  step=1347,  loss=0.178,  accuracy=0.875\n",
      "Epoch  1,  step=1348,  loss=0.127,  accuracy=0.938\n",
      "Epoch  1,  step=1349,  loss=0.629,  accuracy=0.812\n",
      "Epoch  1,  step=1350,  loss=1.229,  accuracy=0.750\n",
      "Epoch  1,  step=1351,  loss=0.258,  accuracy=0.906\n",
      "Epoch  1,  step=1352,  loss=0.340,  accuracy=0.875\n",
      "Epoch  1,  step=1353,  loss=0.063,  accuracy=0.969\n",
      "Epoch  1,  step=1354,  loss=0.071,  accuracy=1.000\n",
      "Epoch  1,  step=1355,  loss=0.392,  accuracy=0.875\n",
      "Epoch  1,  step=1356,  loss=0.266,  accuracy=0.875\n",
      "Epoch  1,  step=1357,  loss=0.555,  accuracy=0.906\n",
      "Epoch  1,  step=1358,  loss=0.364,  accuracy=0.875\n",
      "Epoch  1,  step=1359,  loss=0.226,  accuracy=0.906\n",
      "Epoch  1,  step=1360,  loss=0.396,  accuracy=0.875\n",
      "Epoch  1,  step=1361,  loss=0.317,  accuracy=0.906\n",
      "Epoch  1,  step=1362,  loss=0.369,  accuracy=0.875\n",
      "Epoch  1,  step=1363,  loss=0.340,  accuracy=0.938\n",
      "Epoch  1,  step=1364,  loss=0.764,  accuracy=0.812\n",
      "Epoch  1,  step=1365,  loss=0.365,  accuracy=0.875\n",
      "Epoch  1,  step=1366,  loss=0.111,  accuracy=0.906\n",
      "Epoch  1,  step=1367,  loss=0.226,  accuracy=0.969\n",
      "Epoch  1,  step=1368,  loss=0.369,  accuracy=0.906\n",
      "Epoch  1,  step=1369,  loss=0.715,  accuracy=0.781\n",
      "Epoch  1,  step=1370,  loss=0.463,  accuracy=0.875\n",
      "Epoch  1,  step=1371,  loss=0.370,  accuracy=0.906\n",
      "Epoch  1,  step=1372,  loss=0.139,  accuracy=0.938\n",
      "Epoch  1,  step=1373,  loss=0.918,  accuracy=0.750\n",
      "Epoch  1,  step=1374,  loss=0.383,  accuracy=0.906\n",
      "Epoch  1,  step=1375,  loss=0.068,  accuracy=1.000\n",
      "Epoch  1,  step=1376,  loss=0.517,  accuracy=0.906\n",
      "Epoch  1,  step=1377,  loss=0.478,  accuracy=0.812\n",
      "Epoch  1,  step=1378,  loss=0.440,  accuracy=0.875\n",
      "Epoch  1,  step=1379,  loss=0.712,  accuracy=0.844\n",
      "Epoch  1,  step=1380,  loss=0.371,  accuracy=0.844\n",
      "Epoch  1,  step=1381,  loss=0.314,  accuracy=0.906\n",
      "Epoch  1,  step=1382,  loss=0.667,  accuracy=0.781\n",
      "Epoch  1,  step=1383,  loss=0.933,  accuracy=0.844\n",
      "Epoch  1,  step=1384,  loss=0.504,  accuracy=0.844\n",
      "Epoch  1,  step=1385,  loss=0.301,  accuracy=0.906\n",
      "Epoch  1,  step=1386,  loss=0.730,  accuracy=0.938\n",
      "Epoch  1,  step=1387,  loss=0.151,  accuracy=0.938\n",
      "Epoch  1,  step=1388,  loss=0.516,  accuracy=0.906\n",
      "Epoch  1,  step=1389,  loss=0.263,  accuracy=0.938\n",
      "Epoch  1,  step=1390,  loss=0.573,  accuracy=0.844\n",
      "Epoch  1,  step=1391,  loss=0.187,  accuracy=0.969\n",
      "Epoch  1,  step=1392,  loss=0.116,  accuracy=0.938\n",
      "Epoch  1,  step=1393,  loss=0.089,  accuracy=0.938\n",
      "Epoch  1,  step=1394,  loss=0.104,  accuracy=0.969\n",
      "Epoch  1,  step=1395,  loss=0.597,  accuracy=0.812\n",
      "Epoch  1,  step=1396,  loss=0.493,  accuracy=0.875\n",
      "Epoch  1,  step=1397,  loss=0.266,  accuracy=0.906\n",
      "Epoch  1,  step=1398,  loss=0.665,  accuracy=0.781\n",
      "Epoch  1,  step=1399,  loss=0.151,  accuracy=0.969\n",
      "Epoch  1,  step=1400,  loss=0.455,  accuracy=0.906\n",
      "Epoch  1,  step=1401,  loss=0.510,  accuracy=0.875\n",
      "Epoch  1,  step=1402,  loss=0.969,  accuracy=0.875\n",
      "Epoch  1,  step=1403,  loss=0.598,  accuracy=0.844\n",
      "Epoch  1,  step=1404,  loss=0.337,  accuracy=0.875\n",
      "Epoch  1,  step=1405,  loss=0.322,  accuracy=0.906\n",
      "Epoch  1,  step=1406,  loss=0.533,  accuracy=0.875\n",
      "Epoch  1,  step=1407,  loss=1.108,  accuracy=0.781\n",
      "Epoch  1,  step=1408,  loss=0.349,  accuracy=0.906\n",
      "Epoch  1,  step=1409,  loss=0.954,  accuracy=0.812\n",
      "Epoch  1,  step=1410,  loss=0.388,  accuracy=0.906\n",
      "Epoch  1,  step=1411,  loss=0.617,  accuracy=0.844\n",
      "Epoch  1,  step=1412,  loss=0.500,  accuracy=0.906\n",
      "Epoch  1,  step=1413,  loss=0.312,  accuracy=0.875\n",
      "Epoch  1,  step=1414,  loss=0.266,  accuracy=0.875\n",
      "Epoch  1,  step=1415,  loss=0.114,  accuracy=0.969\n",
      "Epoch  1,  step=1416,  loss=0.288,  accuracy=0.938\n",
      "Epoch  1,  step=1417,  loss=0.129,  accuracy=0.938\n",
      "Epoch  1,  step=1418,  loss=0.135,  accuracy=0.969\n",
      "Epoch  1,  step=1419,  loss=0.773,  accuracy=0.781\n",
      "Epoch  1,  step=1420,  loss=0.780,  accuracy=0.812\n",
      "Epoch  1,  step=1421,  loss=1.224,  accuracy=0.781\n",
      "Epoch  1,  step=1422,  loss=0.704,  accuracy=0.781\n",
      "Epoch  1,  step=1423,  loss=0.221,  accuracy=0.938\n",
      "Epoch  1,  step=1424,  loss=0.394,  accuracy=0.844\n",
      "Epoch  1,  step=1425,  loss=0.553,  accuracy=0.875\n",
      "Epoch  1,  step=1426,  loss=0.706,  accuracy=0.781\n",
      "Epoch  1,  step=1427,  loss=0.606,  accuracy=0.875\n",
      "Epoch  1,  step=1428,  loss=0.029,  accuracy=1.000\n",
      "Epoch  1,  step=1429,  loss=0.349,  accuracy=0.844\n",
      "Epoch  1,  step=1430,  loss=0.576,  accuracy=0.938\n",
      "Epoch  1,  step=1431,  loss=0.446,  accuracy=0.844\n",
      "Epoch  1,  step=1432,  loss=0.565,  accuracy=0.812\n",
      "Epoch  1,  step=1433,  loss=0.785,  accuracy=0.875\n",
      "Epoch  1,  step=1434,  loss=0.698,  accuracy=0.812\n",
      "Epoch  1,  step=1435,  loss=0.503,  accuracy=0.812\n",
      "Epoch  1,  step=1436,  loss=0.576,  accuracy=0.812\n",
      "Epoch  1,  step=1437,  loss=0.620,  accuracy=0.906\n",
      "Epoch  1,  step=1438,  loss=0.482,  accuracy=0.781\n",
      "Epoch  1,  step=1439,  loss=0.563,  accuracy=0.906\n",
      "Epoch  1,  step=1440,  loss=0.743,  accuracy=0.844\n",
      "Epoch  1,  step=1441,  loss=0.223,  accuracy=0.938\n",
      "Epoch  1,  step=1442,  loss=0.313,  accuracy=0.875\n",
      "Epoch  1,  step=1443,  loss=0.231,  accuracy=0.969\n",
      "Epoch  1,  step=1444,  loss=0.122,  accuracy=0.938\n",
      "Epoch  1,  step=1445,  loss=1.436,  accuracy=0.781\n",
      "Epoch  1,  step=1446,  loss=0.873,  accuracy=0.750\n",
      "Epoch  1,  step=1447,  loss=0.615,  accuracy=0.781\n",
      "Epoch  1,  step=1448,  loss=0.337,  accuracy=0.844\n",
      "Epoch  1,  step=1449,  loss=0.231,  accuracy=0.938\n",
      "Epoch  1,  step=1450,  loss=0.676,  accuracy=0.812\n",
      "Epoch  1,  step=1451,  loss=0.249,  accuracy=0.875\n",
      "Epoch  1,  step=1452,  loss=0.143,  accuracy=0.969\n",
      "Epoch  1,  step=1453,  loss=0.223,  accuracy=0.906\n",
      "Epoch  1,  step=1454,  loss=0.293,  accuracy=0.906\n",
      "Epoch  1,  step=1455,  loss=0.181,  accuracy=0.906\n",
      "Epoch  1,  step=1456,  loss=0.283,  accuracy=0.906\n",
      "Epoch  1,  step=1457,  loss=0.486,  accuracy=0.844\n",
      "Epoch  1,  step=1458,  loss=0.133,  accuracy=0.938\n",
      "Epoch  1,  step=1459,  loss=0.309,  accuracy=0.875\n",
      "Epoch  1,  step=1460,  loss=1.027,  accuracy=0.750\n",
      "Epoch  1,  step=1461,  loss=0.364,  accuracy=0.906\n",
      "Epoch  1,  step=1462,  loss=0.315,  accuracy=0.875\n",
      "Epoch  1,  step=1463,  loss=0.102,  accuracy=0.969\n",
      "Epoch  1,  step=1464,  loss=0.282,  accuracy=0.875\n",
      "Epoch  1,  step=1465,  loss=0.510,  accuracy=0.781\n",
      "Epoch  1,  step=1466,  loss=0.360,  accuracy=0.938\n",
      "Epoch  1,  step=1467,  loss=0.151,  accuracy=0.969\n",
      "Epoch  1,  step=1468,  loss=0.393,  accuracy=0.844\n",
      "Epoch  1,  step=1469,  loss=0.909,  accuracy=0.844\n",
      "Epoch  1,  step=1470,  loss=0.519,  accuracy=0.844\n",
      "Epoch  1,  step=1471,  loss=0.316,  accuracy=0.969\n",
      "Epoch  1,  step=1472,  loss=0.154,  accuracy=0.969\n",
      "Epoch  1,  step=1473,  loss=0.379,  accuracy=0.875\n",
      "Epoch  1,  step=1474,  loss=0.272,  accuracy=0.906\n",
      "Epoch  1,  step=1475,  loss=1.044,  accuracy=0.812\n",
      "Epoch  1,  step=1476,  loss=0.712,  accuracy=0.844\n",
      "Epoch  1,  step=1477,  loss=0.371,  accuracy=0.906\n",
      "Epoch  1,  step=1478,  loss=0.858,  accuracy=0.812\n",
      "Epoch  1,  step=1479,  loss=0.743,  accuracy=0.938\n",
      "Epoch  1,  step=1480,  loss=0.640,  accuracy=0.906\n",
      "Epoch  1,  step=1481,  loss=0.764,  accuracy=0.844\n",
      "Epoch  1,  step=1482,  loss=0.544,  accuracy=0.844\n",
      "Epoch  1,  step=1483,  loss=0.480,  accuracy=0.750\n",
      "Epoch  1,  step=1484,  loss=0.063,  accuracy=1.000\n",
      "Epoch  1,  step=1485,  loss=0.298,  accuracy=0.875\n",
      "Epoch  1,  step=1486,  loss=0.628,  accuracy=0.906\n",
      "Epoch  1,  step=1487,  loss=0.787,  accuracy=0.781\n",
      "Epoch  1,  step=1488,  loss=0.326,  accuracy=0.906\n",
      "Epoch  1,  step=1489,  loss=0.322,  accuracy=0.844\n",
      "Epoch  1,  step=1490,  loss=0.572,  accuracy=0.844\n",
      "Epoch  1,  step=1491,  loss=0.491,  accuracy=0.906\n",
      "Epoch  1,  step=1492,  loss=0.609,  accuracy=0.781\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  1,  step=1493,  loss=0.343,  accuracy=0.906\n",
      "Epoch  1,  step=1494,  loss=0.312,  accuracy=0.969\n",
      "Epoch  1,  step=1495,  loss=0.186,  accuracy=0.969\n",
      "Epoch  1,  step=1496,  loss=0.300,  accuracy=0.906\n",
      "Epoch  1,  step=1497,  loss=0.734,  accuracy=0.906\n",
      "Epoch  1,  step=1498,  loss=0.555,  accuracy=0.844\n",
      "Epoch  1,  step=1499,  loss=0.208,  accuracy=0.906\n",
      "Epoch  1,  step=1500,  loss=0.327,  accuracy=0.875\n",
      "Epoch  1,  step=1501,  loss=0.136,  accuracy=0.938\n",
      "Epoch  1,  step=1502,  loss=0.415,  accuracy=0.875\n",
      "Epoch  1,  step=1503,  loss=0.456,  accuracy=0.844\n",
      "Epoch  1,  step=1504,  loss=0.136,  accuracy=0.969\n",
      "Epoch  1,  step=1505,  loss=0.198,  accuracy=0.938\n",
      "Epoch  1,  step=1506,  loss=0.100,  accuracy=0.969\n",
      "Epoch  1,  step=1507,  loss=0.078,  accuracy=0.969\n",
      "Epoch  1,  step=1508,  loss=0.229,  accuracy=0.906\n",
      "Epoch  1,  step=1509,  loss=0.112,  accuracy=0.938\n",
      "Epoch  1,  step=1510,  loss=0.520,  accuracy=0.875\n",
      "Epoch  1,  step=1511,  loss=0.715,  accuracy=0.750\n",
      "Epoch  1,  step=1512,  loss=0.851,  accuracy=0.781\n",
      "Epoch  1,  step=1513,  loss=0.051,  accuracy=0.969\n",
      "Epoch  1,  step=1514,  loss=0.321,  accuracy=0.844\n",
      "Epoch  1,  step=1515,  loss=0.473,  accuracy=0.875\n",
      "Epoch  1,  step=1516,  loss=0.287,  accuracy=0.875\n",
      "Epoch  1,  step=1517,  loss=0.274,  accuracy=0.906\n",
      "Epoch  1,  step=1518,  loss=0.254,  accuracy=0.906\n",
      "Epoch  1,  step=1519,  loss=0.064,  accuracy=1.000\n",
      "Epoch  1,  step=1520,  loss=0.700,  accuracy=0.875\n",
      "Epoch  1,  step=1521,  loss=0.389,  accuracy=0.844\n",
      "Epoch  1,  step=1522,  loss=0.211,  accuracy=0.938\n",
      "Epoch  1,  step=1523,  loss=0.236,  accuracy=0.906\n",
      "Epoch  1,  step=1524,  loss=0.182,  accuracy=0.938\n",
      "Epoch  1,  step=1525,  loss=0.186,  accuracy=0.938\n",
      "Epoch  1,  step=1526,  loss=0.408,  accuracy=0.844\n",
      "Epoch  1,  step=1527,  loss=0.414,  accuracy=0.844\n",
      "Epoch  1,  step=1528,  loss=0.950,  accuracy=0.844\n",
      "Epoch  1,  step=1529,  loss=0.891,  accuracy=0.688\n",
      "Epoch  1,  step=1530,  loss=0.871,  accuracy=0.844\n",
      "Epoch  1,  step=1531,  loss=1.705,  accuracy=0.719\n",
      "Epoch  1,  step=1532,  loss=0.511,  accuracy=0.812\n",
      "Epoch  1,  step=1533,  loss=0.895,  accuracy=0.812\n",
      "Epoch  1,  step=1534,  loss=0.535,  accuracy=0.812\n",
      "Epoch  1,  step=1535,  loss=0.786,  accuracy=0.812\n",
      "Epoch  1,  step=1536,  loss=0.341,  accuracy=0.812\n",
      "Epoch  1,  step=1537,  loss=1.481,  accuracy=0.688\n",
      "Epoch  1,  step=1538,  loss=0.578,  accuracy=0.812\n",
      "Epoch  1,  step=1539,  loss=0.221,  accuracy=0.938\n",
      "Epoch  1,  step=1540,  loss=0.249,  accuracy=0.906\n",
      "Epoch  1,  step=1541,  loss=0.166,  accuracy=0.938\n",
      "Epoch  1,  step=1542,  loss=0.129,  accuracy=0.969\n",
      "Epoch  1,  step=1543,  loss=0.150,  accuracy=0.906\n",
      "Epoch  1,  step=1544,  loss=1.172,  accuracy=0.719\n",
      "Epoch  1,  step=1545,  loss=0.299,  accuracy=0.906\n",
      "Epoch  1,  step=1546,  loss=0.985,  accuracy=0.719\n",
      "Epoch  1,  step=1547,  loss=0.804,  accuracy=0.781\n",
      "Epoch  1,  step=1548,  loss=0.937,  accuracy=0.781\n",
      "Epoch  1,  step=1549,  loss=0.968,  accuracy=0.875\n",
      "Epoch  1,  step=1550,  loss=0.505,  accuracy=0.844\n",
      "Epoch  1,  step=1551,  loss=0.987,  accuracy=0.750\n",
      "Epoch  1,  step=1552,  loss=0.523,  accuracy=0.812\n",
      "Epoch  1,  step=1553,  loss=0.257,  accuracy=0.938\n",
      "Epoch  1,  step=1554,  loss=0.309,  accuracy=0.938\n",
      "Epoch  1,  step=1555,  loss=0.216,  accuracy=0.938\n",
      "Epoch  1,  step=1556,  loss=0.415,  accuracy=0.844\n",
      "Epoch  1,  step=1557,  loss=0.473,  accuracy=0.844\n",
      "Epoch  1,  step=1558,  loss=0.153,  accuracy=0.969\n",
      "Epoch  1,  step=1559,  loss=0.957,  accuracy=0.781\n",
      "Epoch  1,  step=1560,  loss=0.414,  accuracy=0.781\n",
      "Epoch  1,  step=1561,  loss=0.168,  accuracy=0.938\n",
      "Epoch  1,  step=1562,  loss=0.481,  accuracy=0.812\n",
      "Epoch  1,  step=1563,  loss=0.482,  accuracy=0.719\n",
      "Epoch  1,  step=1564,  loss=0.381,  accuracy=0.875\n",
      "Epoch  1,  step=1565,  loss=0.278,  accuracy=0.875\n",
      "Epoch  1,  step=1566,  loss=0.280,  accuracy=0.906\n",
      "Epoch  1,  step=1567,  loss=0.690,  accuracy=0.844\n",
      "Epoch  1,  step=1568,  loss=0.171,  accuracy=0.906\n",
      "Epoch  1,  step=1569,  loss=0.967,  accuracy=0.812\n",
      "Epoch  1,  step=1570,  loss=0.671,  accuracy=0.875\n",
      "Epoch  1,  step=1571,  loss=0.252,  accuracy=0.938\n",
      "Epoch  1,  step=1572,  loss=0.717,  accuracy=0.812\n",
      "Epoch  1,  step=1573,  loss=0.737,  accuracy=0.781\n",
      "Epoch  1,  step=1574,  loss=0.814,  accuracy=0.656\n",
      "Epoch  1,  step=1575,  loss=1.440,  accuracy=0.594\n",
      "Epoch  1,  step=1576,  loss=0.556,  accuracy=0.719\n",
      "Epoch  1,  step=1577,  loss=0.096,  accuracy=1.000\n",
      "Epoch  1,  step=1578,  loss=0.801,  accuracy=0.844\n",
      "Epoch  1,  step=1579,  loss=0.164,  accuracy=0.938\n",
      "Epoch  1,  step=1580,  loss=1.283,  accuracy=0.781\n",
      "Epoch  1,  step=1581,  loss=0.241,  accuracy=0.844\n",
      "Epoch  1,  step=1582,  loss=1.164,  accuracy=0.812\n",
      "Epoch  1,  step=1583,  loss=0.287,  accuracy=0.969\n",
      "Epoch  1,  step=1584,  loss=0.603,  accuracy=0.875\n",
      "Epoch  1,  step=1585,  loss=0.584,  accuracy=0.875\n",
      "Epoch  1,  step=1586,  loss=0.399,  accuracy=0.906\n",
      "Epoch  1,  step=1587,  loss=0.355,  accuracy=0.875\n",
      "Epoch  1,  step=1588,  loss=0.358,  accuracy=0.875\n",
      "Epoch  1,  step=1589,  loss=0.236,  accuracy=0.906\n",
      "Epoch  1,  step=1590,  loss=0.384,  accuracy=0.875\n",
      "Epoch  1,  step=1591,  loss=0.117,  accuracy=0.938\n",
      "Epoch  1,  step=1592,  loss=0.223,  accuracy=0.906\n",
      "Epoch  1,  step=1593,  loss=0.199,  accuracy=0.875\n",
      "Epoch  1,  step=1594,  loss=0.143,  accuracy=0.906\n",
      "Epoch  1,  step=1595,  loss=0.528,  accuracy=0.844\n",
      "Epoch  1,  step=1596,  loss=0.281,  accuracy=0.906\n",
      "Epoch  1,  step=1597,  loss=0.414,  accuracy=0.875\n",
      "Epoch  1,  step=1598,  loss=0.348,  accuracy=0.906\n",
      "Epoch  1,  step=1599,  loss=0.739,  accuracy=0.844\n",
      "Epoch  1,  step=1600,  loss=0.319,  accuracy=0.844\n",
      "Epoch  1,  step=1601,  loss=0.512,  accuracy=0.812\n",
      "Epoch  1,  step=1602,  loss=0.590,  accuracy=0.844\n",
      "Epoch  1,  step=1603,  loss=0.778,  accuracy=0.875\n",
      "Epoch  1,  step=1604,  loss=0.403,  accuracy=0.844\n",
      "Epoch  1,  step=1605,  loss=0.199,  accuracy=0.906\n",
      "Epoch  1,  step=1606,  loss=0.254,  accuracy=0.906\n",
      "Epoch  1,  step=1607,  loss=0.197,  accuracy=0.906\n",
      "Epoch  1,  step=1608,  loss=0.164,  accuracy=0.969\n",
      "Epoch  1,  step=1609,  loss=0.249,  accuracy=0.844\n",
      "Epoch  1,  step=1610,  loss=0.638,  accuracy=0.844\n",
      "Epoch  1,  step=1611,  loss=0.430,  accuracy=0.812\n",
      "Epoch  1,  step=1612,  loss=0.066,  accuracy=1.000\n",
      "Epoch  1,  step=1613,  loss=0.429,  accuracy=0.906\n",
      "Epoch  1,  step=1614,  loss=0.137,  accuracy=0.969\n",
      "Epoch  1,  step=1615,  loss=0.398,  accuracy=0.938\n",
      "Epoch  1,  step=1616,  loss=0.696,  accuracy=0.906\n",
      "Epoch  1,  step=1617,  loss=0.505,  accuracy=0.875\n",
      "Epoch  1,  step=1618,  loss=0.481,  accuracy=0.844\n",
      "Epoch  1,  step=1619,  loss=0.110,  accuracy=0.969\n",
      "Epoch  1,  step=1620,  loss=0.384,  accuracy=0.844\n",
      "Epoch  1,  step=1621,  loss=0.249,  accuracy=0.938\n",
      "Epoch  1,  step=1622,  loss=0.322,  accuracy=0.875\n",
      "Epoch  1,  step=1623,  loss=0.778,  accuracy=0.781\n",
      "Epoch  1,  step=1624,  loss=1.045,  accuracy=0.812\n",
      "Epoch  1,  step=1625,  loss=0.133,  accuracy=0.969\n",
      "Epoch  1,  step=1626,  loss=0.122,  accuracy=0.969\n",
      "Epoch  1,  step=1627,  loss=0.724,  accuracy=0.812\n",
      "Epoch  1,  step=1628,  loss=0.989,  accuracy=0.750\n",
      "Epoch  1,  step=1629,  loss=1.417,  accuracy=0.750\n",
      "Epoch  1,  step=1630,  loss=0.860,  accuracy=0.781\n",
      "Epoch  1,  step=1631,  loss=0.292,  accuracy=0.875\n",
      "Epoch  1,  step=1632,  loss=0.658,  accuracy=0.781\n",
      "Epoch  1,  step=1633,  loss=0.382,  accuracy=0.906\n",
      "Epoch  1,  step=1634,  loss=0.183,  accuracy=0.906\n",
      "Epoch  1,  step=1635,  loss=0.362,  accuracy=0.938\n",
      "Epoch  1,  step=1636,  loss=0.585,  accuracy=0.906\n",
      "Epoch  1,  step=1637,  loss=0.557,  accuracy=0.906\n",
      "Epoch  1,  step=1638,  loss=0.327,  accuracy=0.906\n",
      "Epoch  1,  step=1639,  loss=0.234,  accuracy=0.875\n",
      "Epoch  1,  step=1640,  loss=0.216,  accuracy=0.938\n",
      "Epoch  1,  step=1641,  loss=0.156,  accuracy=0.875\n",
      "Epoch  1,  step=1642,  loss=0.243,  accuracy=0.906\n",
      "Epoch  1,  step=1643,  loss=0.196,  accuracy=0.938\n",
      "Epoch  1,  step=1644,  loss=0.182,  accuracy=0.969\n",
      "Epoch  1,  step=1645,  loss=0.854,  accuracy=0.844\n",
      "Epoch  1,  step=1646,  loss=0.779,  accuracy=0.844\n",
      "Epoch  1,  step=1647,  loss=0.653,  accuracy=0.781\n",
      "Epoch  1,  step=1648,  loss=0.446,  accuracy=0.812\n",
      "Epoch  1,  step=1649,  loss=0.375,  accuracy=0.875\n",
      "Epoch  1,  step=1650,  loss=0.521,  accuracy=0.844\n",
      "Epoch  1,  step=1651,  loss=0.198,  accuracy=0.938\n",
      "Epoch  1,  step=1652,  loss=0.596,  accuracy=0.844\n",
      "Epoch  1,  step=1653,  loss=0.744,  accuracy=0.844\n",
      "Epoch  1,  step=1654,  loss=0.905,  accuracy=0.781\n",
      "Epoch  1,  step=1655,  loss=0.469,  accuracy=0.781\n",
      "Epoch  1,  step=1656,  loss=0.512,  accuracy=0.875\n",
      "Epoch  1,  step=1657,  loss=0.508,  accuracy=0.812\n",
      "Epoch  1,  step=1658,  loss=0.417,  accuracy=0.875\n",
      "Epoch  1,  step=1659,  loss=0.476,  accuracy=0.875\n",
      "Epoch  1,  step=1660,  loss=0.497,  accuracy=0.844\n",
      "Epoch  1,  step=1661,  loss=0.997,  accuracy=0.812\n",
      "Epoch  1,  step=1662,  loss=0.225,  accuracy=0.906\n",
      "Epoch  1,  step=1663,  loss=0.836,  accuracy=0.844\n",
      "Epoch  1,  step=1664,  loss=0.150,  accuracy=0.906\n",
      "Epoch  1,  step=1665,  loss=0.356,  accuracy=0.906\n",
      "Epoch  1,  step=1666,  loss=0.227,  accuracy=0.938\n",
      "Epoch  1,  step=1667,  loss=0.111,  accuracy=0.969\n",
      "Epoch  1,  step=1668,  loss=0.371,  accuracy=0.906\n",
      "Epoch  1,  step=1669,  loss=0.207,  accuracy=0.938\n",
      "Epoch  1,  step=1670,  loss=0.566,  accuracy=0.844\n",
      "Epoch  1,  step=1671,  loss=0.402,  accuracy=0.844\n",
      "Epoch  1,  step=1672,  loss=0.453,  accuracy=0.875\n",
      "Epoch  1,  step=1673,  loss=0.559,  accuracy=0.844\n",
      "Epoch  1,  step=1674,  loss=0.537,  accuracy=0.875\n",
      "Epoch  1,  step=1675,  loss=0.378,  accuracy=0.906\n",
      "Epoch  1,  step=1676,  loss=0.309,  accuracy=0.906\n",
      "Epoch  1,  step=1677,  loss=0.417,  accuracy=0.906\n",
      "Epoch  1,  step=1678,  loss=0.150,  accuracy=0.969\n",
      "Epoch  1,  step=1679,  loss=0.813,  accuracy=0.750\n",
      "Epoch  1,  step=1680,  loss=0.058,  accuracy=0.969\n",
      "Epoch  1,  step=1681,  loss=0.425,  accuracy=0.906\n",
      "Epoch  1,  step=1682,  loss=0.459,  accuracy=0.812\n",
      "Epoch  1,  step=1683,  loss=0.513,  accuracy=0.875\n",
      "Epoch  1,  step=1684,  loss=0.775,  accuracy=0.844\n",
      "Epoch  1,  step=1685,  loss=0.384,  accuracy=0.875\n",
      "Epoch  1,  step=1686,  loss=1.026,  accuracy=0.781\n",
      "Epoch  1,  step=1687,  loss=0.502,  accuracy=0.844\n",
      "Epoch  1,  step=1688,  loss=0.438,  accuracy=0.844\n",
      "Epoch  1,  step=1689,  loss=0.347,  accuracy=0.875\n",
      "Epoch  1,  step=1690,  loss=0.992,  accuracy=0.750\n",
      "Epoch  1,  step=1691,  loss=0.132,  accuracy=0.938\n",
      "Epoch  1,  step=1692,  loss=0.090,  accuracy=0.969\n",
      "Epoch  1,  step=1693,  loss=0.477,  accuracy=0.844\n",
      "Epoch  1,  step=1694,  loss=0.218,  accuracy=0.938\n",
      "Epoch  1,  step=1695,  loss=0.403,  accuracy=0.844\n",
      "Epoch  1,  step=1696,  loss=0.400,  accuracy=0.844\n",
      "Epoch  1,  step=1697,  loss=0.155,  accuracy=0.938\n",
      "Epoch  1,  step=1698,  loss=0.230,  accuracy=0.906\n",
      "Epoch  1,  step=1699,  loss=0.910,  accuracy=0.875\n",
      "Epoch  1,  step=1700,  loss=0.258,  accuracy=0.938\n",
      "Epoch  1,  step=1701,  loss=0.697,  accuracy=0.844\n",
      "Epoch  1,  step=1702,  loss=0.185,  accuracy=0.938\n",
      "Epoch  1,  step=1703,  loss=0.530,  accuracy=0.875\n",
      "Epoch  1,  step=1704,  loss=0.560,  accuracy=0.812\n",
      "Epoch  1,  step=1705,  loss=0.640,  accuracy=0.844\n",
      "Epoch  1,  step=1706,  loss=0.252,  accuracy=0.875\n",
      "Epoch  1,  step=1707,  loss=0.477,  accuracy=0.875\n",
      "Epoch  1,  step=1708,  loss=0.261,  accuracy=0.875\n",
      "Epoch  1,  step=1709,  loss=0.140,  accuracy=0.969\n",
      "Epoch  1,  step=1710,  loss=0.061,  accuracy=0.969\n",
      "Epoch  1,  step=1711,  loss=0.494,  accuracy=0.844\n",
      "Epoch  1,  step=1712,  loss=0.513,  accuracy=0.812\n",
      "Epoch  1,  step=1713,  loss=0.355,  accuracy=0.938\n",
      "Epoch  1,  step=1714,  loss=0.472,  accuracy=0.875\n",
      "Epoch  1,  step=1715,  loss=0.655,  accuracy=0.844\n",
      "Epoch  1,  step=1716,  loss=1.085,  accuracy=0.719\n",
      "Epoch  1,  step=1717,  loss=0.792,  accuracy=0.812\n",
      "Epoch  1,  step=1718,  loss=0.350,  accuracy=0.906\n",
      "Epoch  1,  step=1719,  loss=0.433,  accuracy=0.906\n",
      "Epoch  1,  step=1720,  loss=0.730,  accuracy=0.812\n",
      "Epoch  1,  step=1721,  loss=0.420,  accuracy=0.875\n",
      "Epoch  1,  step=1722,  loss=0.514,  accuracy=0.906\n",
      "Epoch  1,  step=1723,  loss=0.073,  accuracy=1.000\n",
      "Epoch  1,  step=1724,  loss=0.356,  accuracy=0.938\n",
      "Epoch  1,  step=1725,  loss=0.234,  accuracy=0.906\n",
      "Epoch  1,  step=1726,  loss=0.238,  accuracy=0.875\n",
      "Epoch  1,  step=1727,  loss=0.208,  accuracy=0.906\n",
      "Epoch  1,  step=1728,  loss=0.567,  accuracy=0.812\n",
      "Epoch  1,  step=1729,  loss=0.571,  accuracy=0.938\n",
      "Epoch  1,  step=1730,  loss=0.099,  accuracy=0.938\n",
      "Epoch  1,  step=1731,  loss=0.146,  accuracy=0.938\n",
      "Epoch  1,  step=1732,  loss=0.385,  accuracy=0.875\n",
      "Epoch  1,  step=1733,  loss=0.356,  accuracy=0.844\n",
      "Epoch  1,  step=1734,  loss=0.835,  accuracy=0.875\n",
      "Epoch  1,  step=1735,  loss=0.563,  accuracy=0.781\n",
      "Epoch  1,  step=1736,  loss=0.040,  accuracy=1.000\n",
      "Epoch  1,  step=1737,  loss=0.276,  accuracy=0.938\n",
      "Epoch  1,  step=1738,  loss=0.150,  accuracy=0.938\n",
      "Epoch  1,  step=1739,  loss=0.268,  accuracy=0.906\n",
      "Epoch  1,  step=1740,  loss=0.244,  accuracy=0.938\n",
      "Epoch  1,  step=1741,  loss=1.069,  accuracy=0.812\n",
      "Epoch  1,  step=1742,  loss=0.206,  accuracy=0.938\n",
      "Epoch  1,  step=1743,  loss=0.292,  accuracy=0.938\n",
      "Epoch  1,  step=1744,  loss=0.114,  accuracy=0.938\n",
      "Epoch  1,  step=1745,  loss=0.697,  accuracy=0.906\n",
      "Epoch  1,  step=1746,  loss=0.329,  accuracy=0.875\n",
      "Epoch  1,  step=1747,  loss=0.356,  accuracy=0.906\n",
      "Epoch  1,  step=1748,  loss=0.140,  accuracy=0.969\n",
      "Epoch  1,  step=1749,  loss=0.285,  accuracy=0.875\n",
      "Epoch  1,  step=1750,  loss=0.343,  accuracy=0.906\n",
      "Epoch  1,  step=1751,  loss=0.180,  accuracy=0.969\n",
      "Epoch  1,  step=1752,  loss=0.187,  accuracy=0.906\n",
      "Epoch  1,  step=1753,  loss=0.099,  accuracy=0.969\n",
      "Epoch  1,  step=1754,  loss=0.503,  accuracy=0.875\n",
      "Epoch  1,  step=1755,  loss=0.248,  accuracy=0.938\n",
      "Epoch  1,  step=1756,  loss=0.984,  accuracy=0.875\n",
      "Epoch  1,  step=1757,  loss=0.152,  accuracy=0.969\n",
      "Epoch  1,  step=1758,  loss=0.351,  accuracy=0.875\n",
      "Epoch  1,  step=1759,  loss=0.575,  accuracy=0.812\n",
      "Epoch  1,  step=1760,  loss=0.299,  accuracy=0.844\n",
      "Epoch  1,  step=1761,  loss=0.435,  accuracy=0.844\n",
      "Epoch  1,  step=1762,  loss=0.327,  accuracy=0.812\n",
      "Epoch  1,  step=1763,  loss=0.378,  accuracy=0.906\n",
      "Epoch  1,  step=1764,  loss=1.680,  accuracy=0.719\n",
      "Epoch  1,  step=1765,  loss=0.446,  accuracy=0.906\n",
      "Epoch  1,  step=1766,  loss=0.065,  accuracy=1.000\n",
      "Epoch  1,  step=1767,  loss=0.051,  accuracy=1.000\n",
      "Epoch  1,  step=1768,  loss=0.437,  accuracy=0.812\n",
      "Epoch  1,  step=1769,  loss=0.536,  accuracy=0.844\n",
      "Epoch  1,  step=1770,  loss=0.870,  accuracy=0.781\n",
      "Epoch  1,  step=1771,  loss=0.337,  accuracy=0.875\n",
      "Epoch  1,  step=1772,  loss=0.117,  accuracy=0.969\n",
      "Epoch  1,  step=1773,  loss=0.089,  accuracy=1.000\n",
      "Epoch  1,  step=1774,  loss=0.268,  accuracy=0.906\n",
      "Epoch  1,  step=1775,  loss=0.296,  accuracy=0.875\n",
      "Epoch  1,  step=1776,  loss=0.232,  accuracy=0.875\n",
      "Epoch  1,  step=1777,  loss=0.136,  accuracy=0.969\n",
      "Epoch  1,  step=1778,  loss=0.374,  accuracy=0.906\n",
      "Epoch  1,  step=1779,  loss=0.070,  accuracy=0.969\n",
      "Epoch  1,  step=1780,  loss=0.347,  accuracy=0.906\n",
      "Epoch  1,  step=1781,  loss=0.524,  accuracy=0.844\n",
      "Epoch  1,  step=1782,  loss=0.738,  accuracy=0.906\n",
      "Epoch  1,  step=1783,  loss=0.719,  accuracy=0.875\n",
      "Epoch  1,  step=1784,  loss=0.094,  accuracy=0.969\n",
      "Epoch  1,  step=1785,  loss=0.205,  accuracy=0.938\n",
      "Epoch  1,  step=1786,  loss=0.458,  accuracy=0.844\n",
      "Epoch  1,  step=1787,  loss=0.309,  accuracy=0.906\n",
      "Epoch  1,  step=1788,  loss=0.446,  accuracy=0.844\n",
      "Epoch  1,  step=1789,  loss=0.338,  accuracy=0.875\n",
      "Epoch  1,  step=1790,  loss=0.523,  accuracy=0.812\n",
      "Epoch  1,  step=1791,  loss=0.662,  accuracy=0.812\n",
      "Epoch  1,  step=1792,  loss=0.205,  accuracy=0.938\n",
      "Epoch  1,  step=1793,  loss=0.284,  accuracy=0.938\n",
      "Epoch  1,  step=1794,  loss=0.165,  accuracy=0.969\n",
      "Epoch  1,  step=1795,  loss=0.202,  accuracy=0.938\n",
      "Epoch  1,  step=1796,  loss=0.535,  accuracy=0.875\n",
      "Epoch  1,  step=1797,  loss=0.455,  accuracy=0.844\n",
      "Epoch  1,  step=1798,  loss=0.519,  accuracy=0.781\n",
      "Epoch  1,  step=1799,  loss=0.483,  accuracy=0.844\n",
      "Epoch  1,  step=1800,  loss=0.513,  accuracy=0.906\n",
      "Epoch  1,  step=1801,  loss=0.696,  accuracy=0.781\n",
      "Epoch  1,  step=1802,  loss=0.576,  accuracy=0.781\n",
      "Epoch  1,  step=1803,  loss=0.735,  accuracy=0.781\n",
      "Epoch  1,  step=1804,  loss=0.748,  accuracy=0.844\n",
      "Epoch  1,  step=1805,  loss=0.143,  accuracy=0.938\n",
      "Epoch  1,  step=1806,  loss=0.368,  accuracy=0.844\n",
      "Epoch  1,  step=1807,  loss=0.387,  accuracy=0.875\n",
      "Epoch  1,  step=1808,  loss=0.299,  accuracy=0.906\n",
      "Epoch  1,  step=1809,  loss=0.217,  accuracy=0.938\n",
      "Epoch  1,  step=1810,  loss=0.370,  accuracy=0.906\n",
      "Epoch  1,  step=1811,  loss=0.390,  accuracy=0.906\n",
      "Epoch  1,  step=1812,  loss=0.160,  accuracy=0.906\n",
      "Epoch  1,  step=1813,  loss=0.224,  accuracy=0.938\n",
      "Epoch  1,  step=1814,  loss=0.618,  accuracy=0.875\n",
      "Epoch  1,  step=1815,  loss=0.833,  accuracy=0.781\n",
      "Epoch  1,  step=1816,  loss=0.104,  accuracy=0.938\n",
      "Epoch  1,  step=1817,  loss=0.049,  accuracy=1.000\n",
      "Epoch  1,  step=1818,  loss=0.054,  accuracy=0.969\n",
      "Epoch  1,  step=1819,  loss=0.041,  accuracy=1.000\n",
      "Epoch  1,  step=1820,  loss=0.196,  accuracy=0.906\n",
      "Epoch  1,  step=1821,  loss=0.095,  accuracy=0.969\n",
      "Epoch  1,  step=1822,  loss=0.247,  accuracy=0.875\n",
      "Epoch  1,  step=1823,  loss=0.034,  accuracy=1.000\n",
      "Epoch  1,  step=1824,  loss=0.232,  accuracy=0.906\n",
      "Epoch  1,  step=1825,  loss=0.377,  accuracy=0.906\n",
      "Epoch  1,  step=1826,  loss=0.138,  accuracy=0.938\n",
      "Epoch  1,  step=1827,  loss=0.271,  accuracy=0.906\n",
      "Epoch  1,  step=1828,  loss=0.458,  accuracy=0.906\n",
      "Epoch  1,  step=1829,  loss=0.384,  accuracy=0.875\n",
      "Epoch  1,  step=1830,  loss=0.279,  accuracy=0.969\n",
      "Epoch  1,  step=1831,  loss=0.324,  accuracy=0.844\n",
      "Epoch  1,  step=1832,  loss=0.241,  accuracy=0.906\n",
      "Epoch  1,  step=1833,  loss=0.189,  accuracy=0.938\n",
      "Epoch  1,  step=1834,  loss=0.102,  accuracy=0.938\n",
      "Epoch  1,  step=1835,  loss=0.054,  accuracy=1.000\n",
      "Epoch  1,  step=1836,  loss=0.154,  accuracy=0.938\n",
      "Epoch  1,  step=1837,  loss=0.461,  accuracy=0.812\n",
      "Epoch  1,  step=1838,  loss=0.251,  accuracy=0.938\n",
      "Epoch  1,  step=1839,  loss=0.642,  accuracy=0.938\n",
      "Epoch  1,  step=1840,  loss=0.107,  accuracy=0.969\n",
      "Epoch  1,  step=1841,  loss=0.028,  accuracy=1.000\n",
      "Epoch  1,  step=1842,  loss=0.047,  accuracy=0.969\n",
      "Epoch  1,  step=1843,  loss=0.257,  accuracy=0.875\n",
      "Epoch  1,  step=1844,  loss=0.061,  accuracy=1.000\n",
      "Epoch  1,  step=1845,  loss=0.284,  accuracy=0.938\n",
      "Epoch  1,  step=1846,  loss=0.242,  accuracy=0.844\n",
      "Epoch  1,  step=1847,  loss=0.155,  accuracy=0.938\n",
      "Epoch  1,  step=1848,  loss=0.316,  accuracy=0.906\n",
      "Epoch  1,  step=1849,  loss=0.271,  accuracy=0.938\n",
      "Epoch  1,  step=1850,  loss=0.278,  accuracy=0.938\n",
      "Epoch  1,  step=1851,  loss=0.285,  accuracy=0.969\n",
      "Epoch  1,  step=1852,  loss=0.463,  accuracy=0.844\n",
      "Epoch  1,  step=1853,  loss=0.934,  accuracy=0.844\n",
      "Epoch  1,  step=1854,  loss=0.858,  accuracy=0.812\n",
      "Epoch  1,  step=1855,  loss=0.452,  accuracy=0.781\n",
      "Epoch  1,  step=1856,  loss=0.316,  accuracy=0.875\n",
      "Epoch  1,  step=1857,  loss=0.268,  accuracy=0.844\n",
      "Epoch  1,  step=1858,  loss=0.121,  accuracy=0.938\n",
      "Epoch  1,  step=1859,  loss=0.105,  accuracy=0.938\n",
      "Epoch  1,  step=1860,  loss=0.128,  accuracy=0.969\n",
      "Epoch  1,  step=1861,  loss=0.214,  accuracy=0.906\n",
      "Epoch  1,  step=1862,  loss=0.076,  accuracy=0.969\n",
      "Epoch  1,  step=1863,  loss=0.090,  accuracy=0.969\n",
      "Epoch  1,  step=1864,  loss=0.394,  accuracy=0.938\n",
      "Epoch  1,  step=1865,  loss=0.482,  accuracy=0.906\n",
      "Epoch  1,  step=1866,  loss=1.699,  accuracy=0.688\n",
      "Epoch  1,  step=1867,  loss=0.238,  accuracy=0.906\n",
      "Epoch  1,  step=1868,  loss=0.512,  accuracy=0.781\n",
      "Epoch  1,  step=1869,  loss=0.155,  accuracy=0.938\n",
      "Epoch  1,  step=1870,  loss=0.065,  accuracy=0.969\n",
      "Epoch  1,  step=1871,  loss=0.066,  accuracy=1.000\n",
      "Epoch  1,  step=1872,  loss=1.324,  accuracy=0.812\n",
      "Epoch  1,  step=1873,  loss=0.200,  accuracy=0.938\n",
      "Epoch  1,  step=1874,  loss=0.195,  accuracy=0.938\n",
      "Epoch  2,  step=   0,  loss=0.368,  accuracy=0.906\n",
      "Epoch  2,  step=   1,  loss=0.621,  accuracy=0.844\n",
      "Epoch  2,  step=   2,  loss=0.537,  accuracy=0.938\n",
      "Epoch  2,  step=   3,  loss=0.272,  accuracy=0.906\n",
      "Epoch  2,  step=   4,  loss=0.710,  accuracy=0.906\n",
      "Epoch  2,  step=   5,  loss=0.572,  accuracy=0.875\n",
      "Epoch  2,  step=   6,  loss=0.057,  accuracy=1.000\n",
      "Epoch  2,  step=   7,  loss=0.715,  accuracy=0.875\n",
      "Epoch  2,  step=   8,  loss=0.604,  accuracy=0.812\n",
      "Epoch  2,  step=   9,  loss=0.111,  accuracy=0.906\n",
      "Epoch  2,  step=  10,  loss=0.361,  accuracy=0.969\n",
      "Epoch  2,  step=  11,  loss=0.288,  accuracy=0.875\n",
      "Epoch  2,  step=  12,  loss=0.225,  accuracy=0.938\n",
      "Epoch  2,  step=  13,  loss=0.530,  accuracy=0.875\n",
      "Epoch  2,  step=  14,  loss=0.305,  accuracy=0.906\n",
      "Epoch  2,  step=  15,  loss=0.631,  accuracy=0.906\n",
      "Epoch  2,  step=  16,  loss=0.234,  accuracy=0.938\n",
      "Epoch  2,  step=  17,  loss=0.296,  accuracy=0.906\n",
      "Epoch  2,  step=  18,  loss=0.277,  accuracy=0.875\n",
      "Epoch  2,  step=  19,  loss=0.835,  accuracy=0.812\n",
      "Epoch  2,  step=  20,  loss=0.312,  accuracy=0.906\n",
      "Epoch  2,  step=  21,  loss=0.170,  accuracy=0.938\n",
      "Epoch  2,  step=  22,  loss=0.384,  accuracy=0.938\n",
      "Epoch  2,  step=  23,  loss=0.309,  accuracy=0.875\n",
      "Epoch  2,  step=  24,  loss=0.494,  accuracy=0.875\n",
      "Epoch  2,  step=  25,  loss=0.295,  accuracy=0.906\n",
      "Epoch  2,  step=  26,  loss=1.071,  accuracy=0.781\n",
      "Epoch  2,  step=  27,  loss=0.609,  accuracy=0.844\n",
      "Epoch  2,  step=  28,  loss=1.091,  accuracy=0.875\n",
      "Epoch  2,  step=  29,  loss=0.394,  accuracy=0.938\n",
      "Epoch  2,  step=  30,  loss=0.121,  accuracy=0.938\n",
      "Epoch  2,  step=  31,  loss=0.569,  accuracy=0.875\n",
      "Epoch  2,  step=  32,  loss=0.623,  accuracy=0.844\n",
      "Epoch  2,  step=  33,  loss=1.217,  accuracy=0.781\n",
      "Epoch  2,  step=  34,  loss=0.336,  accuracy=0.875\n",
      "Epoch  2,  step=  35,  loss=1.151,  accuracy=0.688\n",
      "Epoch  2,  step=  36,  loss=0.201,  accuracy=0.906\n",
      "Epoch  2,  step=  37,  loss=0.187,  accuracy=0.969\n",
      "Epoch  2,  step=  38,  loss=1.138,  accuracy=0.812\n",
      "Epoch  2,  step=  39,  loss=0.662,  accuracy=0.812\n",
      "Epoch  2,  step=  40,  loss=0.284,  accuracy=0.875\n",
      "Epoch  2,  step=  41,  loss=0.483,  accuracy=0.812\n",
      "Epoch  2,  step=  42,  loss=0.859,  accuracy=0.812\n",
      "Epoch  2,  step=  43,  loss=0.746,  accuracy=0.781\n",
      "Epoch  2,  step=  44,  loss=0.320,  accuracy=0.906\n",
      "Epoch  2,  step=  45,  loss=0.402,  accuracy=0.812\n",
      "Epoch  2,  step=  46,  loss=0.290,  accuracy=0.906\n",
      "Epoch  2,  step=  47,  loss=0.969,  accuracy=0.781\n",
      "Epoch  2,  step=  48,  loss=0.375,  accuracy=0.875\n",
      "Epoch  2,  step=  49,  loss=0.612,  accuracy=0.812\n",
      "Epoch  2,  step=  50,  loss=0.400,  accuracy=0.906\n",
      "Epoch  2,  step=  51,  loss=0.356,  accuracy=0.844\n",
      "Epoch  2,  step=  52,  loss=0.250,  accuracy=0.938\n",
      "Epoch  2,  step=  53,  loss=0.102,  accuracy=0.969\n",
      "Epoch  2,  step=  54,  loss=0.246,  accuracy=0.906\n",
      "Epoch  2,  step=  55,  loss=0.234,  accuracy=0.906\n",
      "Epoch  2,  step=  56,  loss=0.568,  accuracy=0.906\n",
      "Epoch  2,  step=  57,  loss=0.585,  accuracy=0.875\n",
      "Epoch  2,  step=  58,  loss=0.181,  accuracy=0.938\n",
      "Epoch  2,  step=  59,  loss=0.137,  accuracy=0.938\n",
      "Epoch  2,  step=  60,  loss=0.302,  accuracy=0.906\n",
      "Epoch  2,  step=  61,  loss=0.393,  accuracy=0.781\n",
      "Epoch  2,  step=  62,  loss=0.420,  accuracy=0.844\n",
      "Epoch  2,  step=  63,  loss=0.268,  accuracy=0.906\n",
      "Epoch  2,  step=  64,  loss=0.342,  accuracy=0.875\n",
      "Epoch  2,  step=  65,  loss=0.367,  accuracy=0.938\n",
      "Epoch  2,  step=  66,  loss=0.400,  accuracy=0.875\n",
      "Epoch  2,  step=  67,  loss=0.119,  accuracy=0.969\n",
      "Epoch  2,  step=  68,  loss=0.224,  accuracy=0.906\n",
      "Epoch  2,  step=  69,  loss=0.369,  accuracy=0.938\n",
      "Epoch  2,  step=  70,  loss=0.317,  accuracy=0.844\n",
      "Epoch  2,  step=  71,  loss=0.281,  accuracy=0.969\n",
      "Epoch  2,  step=  72,  loss=0.173,  accuracy=0.938\n",
      "Epoch  2,  step=  73,  loss=0.095,  accuracy=1.000\n",
      "Epoch  2,  step=  74,  loss=0.093,  accuracy=0.969\n",
      "Epoch  2,  step=  75,  loss=0.857,  accuracy=0.812\n",
      "Epoch  2,  step=  76,  loss=0.274,  accuracy=0.875\n",
      "Epoch  2,  step=  77,  loss=0.154,  accuracy=0.969\n",
      "Epoch  2,  step=  78,  loss=0.111,  accuracy=0.969\n",
      "Epoch  2,  step=  79,  loss=0.373,  accuracy=0.875\n",
      "Epoch  2,  step=  80,  loss=0.198,  accuracy=0.906\n",
      "Epoch  2,  step=  81,  loss=0.272,  accuracy=0.906\n",
      "Epoch  2,  step=  82,  loss=0.421,  accuracy=0.938\n",
      "Epoch  2,  step=  83,  loss=1.208,  accuracy=0.844\n",
      "Epoch  2,  step=  84,  loss=0.425,  accuracy=0.906\n",
      "Epoch  2,  step=  85,  loss=0.184,  accuracy=0.969\n",
      "Epoch  2,  step=  86,  loss=0.505,  accuracy=0.875\n",
      "Epoch  2,  step=  87,  loss=0.729,  accuracy=0.844\n",
      "Epoch  2,  step=  88,  loss=0.598,  accuracy=0.844\n",
      "Epoch  2,  step=  89,  loss=0.252,  accuracy=0.906\n",
      "Epoch  2,  step=  90,  loss=0.324,  accuracy=0.938\n",
      "Epoch  2,  step=  91,  loss=0.447,  accuracy=0.844\n",
      "Epoch  2,  step=  92,  loss=0.353,  accuracy=0.812\n",
      "Epoch  2,  step=  93,  loss=0.331,  accuracy=0.938\n",
      "Epoch  2,  step=  94,  loss=0.354,  accuracy=0.938\n",
      "Epoch  2,  step=  95,  loss=0.135,  accuracy=0.969\n",
      "Epoch  2,  step=  96,  loss=0.189,  accuracy=0.938\n",
      "Epoch  2,  step=  97,  loss=0.379,  accuracy=0.906\n",
      "Epoch  2,  step=  98,  loss=0.190,  accuracy=0.906\n",
      "Epoch  2,  step=  99,  loss=0.123,  accuracy=0.969\n",
      "Epoch  2,  step= 100,  loss=0.316,  accuracy=0.906\n",
      "Epoch  2,  step= 101,  loss=0.308,  accuracy=0.906\n",
      "Epoch  2,  step= 102,  loss=0.685,  accuracy=0.844\n",
      "Epoch  2,  step= 103,  loss=0.131,  accuracy=0.969\n",
      "Epoch  2,  step= 104,  loss=0.110,  accuracy=0.969\n",
      "Epoch  2,  step= 105,  loss=0.530,  accuracy=0.812\n",
      "Epoch  2,  step= 106,  loss=0.578,  accuracy=0.938\n",
      "Epoch  2,  step= 107,  loss=0.263,  accuracy=0.875\n",
      "Epoch  2,  step= 108,  loss=0.274,  accuracy=0.906\n",
      "Epoch  2,  step= 109,  loss=0.620,  accuracy=0.906\n",
      "Epoch  2,  step= 110,  loss=0.335,  accuracy=0.906\n",
      "Epoch  2,  step= 111,  loss=0.225,  accuracy=0.844\n",
      "Epoch  2,  step= 112,  loss=0.079,  accuracy=0.938\n",
      "Epoch  2,  step= 113,  loss=0.323,  accuracy=0.875\n",
      "Epoch  2,  step= 114,  loss=0.734,  accuracy=0.875\n",
      "Epoch  2,  step= 115,  loss=0.815,  accuracy=0.812\n",
      "Epoch  2,  step= 116,  loss=0.731,  accuracy=0.875\n",
      "Epoch  2,  step= 117,  loss=0.120,  accuracy=0.969\n",
      "Epoch  2,  step= 118,  loss=0.501,  accuracy=0.906\n",
      "Epoch  2,  step= 119,  loss=0.582,  accuracy=0.844\n",
      "Epoch  2,  step= 120,  loss=0.151,  accuracy=0.938\n",
      "Epoch  2,  step= 121,  loss=0.372,  accuracy=0.969\n",
      "Epoch  2,  step= 122,  loss=0.084,  accuracy=0.969\n",
      "Epoch  2,  step= 123,  loss=0.125,  accuracy=0.969\n",
      "Epoch  2,  step= 124,  loss=0.515,  accuracy=0.906\n",
      "Epoch  2,  step= 125,  loss=0.445,  accuracy=0.875\n",
      "Epoch  2,  step= 126,  loss=0.325,  accuracy=0.906\n",
      "Epoch  2,  step= 127,  loss=0.354,  accuracy=0.875\n",
      "Epoch  2,  step= 128,  loss=0.268,  accuracy=0.844\n",
      "Epoch  2,  step= 129,  loss=0.817,  accuracy=0.750\n",
      "Epoch  2,  step= 130,  loss=0.537,  accuracy=0.844\n",
      "Epoch  2,  step= 131,  loss=0.367,  accuracy=0.969\n",
      "Epoch  2,  step= 132,  loss=0.255,  accuracy=0.938\n",
      "Epoch  2,  step= 133,  loss=0.445,  accuracy=0.875\n",
      "Epoch  2,  step= 134,  loss=0.218,  accuracy=0.875\n",
      "Epoch  2,  step= 135,  loss=0.165,  accuracy=0.969\n",
      "Epoch  2,  step= 136,  loss=0.469,  accuracy=0.906\n",
      "Epoch  2,  step= 137,  loss=0.110,  accuracy=0.938\n",
      "Epoch  2,  step= 138,  loss=0.177,  accuracy=0.906\n",
      "Epoch  2,  step= 139,  loss=0.691,  accuracy=0.812\n",
      "Epoch  2,  step= 140,  loss=0.359,  accuracy=0.938\n",
      "Epoch  2,  step= 141,  loss=0.058,  accuracy=0.969\n",
      "Epoch  2,  step= 142,  loss=0.092,  accuracy=0.969\n",
      "Epoch  2,  step= 143,  loss=0.471,  accuracy=0.844\n",
      "Epoch  2,  step= 144,  loss=0.244,  accuracy=0.906\n",
      "Epoch  2,  step= 145,  loss=0.780,  accuracy=0.781\n",
      "Epoch  2,  step= 146,  loss=0.221,  accuracy=0.938\n",
      "Epoch  2,  step= 147,  loss=0.271,  accuracy=0.906\n",
      "Epoch  2,  step= 148,  loss=0.430,  accuracy=0.906\n",
      "Epoch  2,  step= 149,  loss=0.349,  accuracy=0.875\n",
      "Epoch  2,  step= 150,  loss=0.394,  accuracy=0.812\n",
      "Epoch  2,  step= 151,  loss=0.078,  accuracy=1.000\n",
      "Epoch  2,  step= 152,  loss=0.216,  accuracy=0.938\n",
      "Epoch  2,  step= 153,  loss=0.348,  accuracy=0.938\n",
      "Epoch  2,  step= 154,  loss=0.425,  accuracy=0.875\n",
      "Epoch  2,  step= 155,  loss=0.801,  accuracy=0.906\n",
      "Epoch  2,  step= 156,  loss=0.304,  accuracy=0.906\n",
      "Epoch  2,  step= 157,  loss=0.691,  accuracy=0.844\n",
      "Epoch  2,  step= 158,  loss=0.729,  accuracy=0.844\n",
      "Epoch  2,  step= 159,  loss=0.549,  accuracy=0.875\n",
      "Epoch  2,  step= 160,  loss=0.540,  accuracy=0.844\n",
      "Epoch  2,  step= 161,  loss=0.966,  accuracy=0.844\n",
      "Epoch  2,  step= 162,  loss=0.352,  accuracy=0.844\n",
      "Epoch  2,  step= 163,  loss=0.115,  accuracy=0.969\n",
      "Epoch  2,  step= 164,  loss=0.519,  accuracy=0.844\n",
      "Epoch  2,  step= 165,  loss=1.137,  accuracy=0.844\n",
      "Epoch  2,  step= 166,  loss=0.547,  accuracy=0.875\n",
      "Epoch  2,  step= 167,  loss=0.445,  accuracy=0.844\n",
      "Epoch  2,  step= 168,  loss=0.253,  accuracy=0.906\n",
      "Epoch  2,  step= 169,  loss=0.287,  accuracy=0.938\n",
      "Epoch  2,  step= 170,  loss=0.144,  accuracy=0.938\n",
      "Epoch  2,  step= 171,  loss=0.271,  accuracy=0.906\n",
      "Epoch  2,  step= 172,  loss=0.146,  accuracy=0.906\n",
      "Epoch  2,  step= 173,  loss=0.514,  accuracy=0.906\n",
      "Epoch  2,  step= 174,  loss=0.124,  accuracy=0.938\n",
      "Epoch  2,  step= 175,  loss=0.472,  accuracy=0.781\n",
      "Epoch  2,  step= 176,  loss=0.352,  accuracy=0.906\n",
      "Epoch  2,  step= 177,  loss=0.616,  accuracy=0.875\n",
      "Epoch  2,  step= 178,  loss=0.543,  accuracy=0.844\n",
      "Epoch  2,  step= 179,  loss=0.157,  accuracy=0.906\n",
      "Epoch  2,  step= 180,  loss=0.489,  accuracy=0.906\n",
      "Epoch  2,  step= 181,  loss=0.131,  accuracy=0.969\n",
      "Epoch  2,  step= 182,  loss=0.564,  accuracy=0.781\n",
      "Epoch  2,  step= 183,  loss=0.305,  accuracy=0.906\n",
      "Epoch  2,  step= 184,  loss=0.355,  accuracy=0.906\n",
      "Epoch  2,  step= 185,  loss=0.410,  accuracy=0.938\n",
      "Epoch  2,  step= 186,  loss=0.185,  accuracy=0.938\n",
      "Epoch  2,  step= 187,  loss=0.076,  accuracy=0.938\n",
      "Epoch  2,  step= 188,  loss=0.221,  accuracy=0.875\n",
      "Epoch  2,  step= 189,  loss=0.118,  accuracy=0.938\n",
      "Epoch  2,  step= 190,  loss=0.593,  accuracy=0.906\n",
      "Epoch  2,  step= 191,  loss=0.342,  accuracy=0.938\n",
      "Epoch  2,  step= 192,  loss=0.194,  accuracy=0.906\n",
      "Epoch  2,  step= 193,  loss=0.175,  accuracy=0.938\n",
      "Epoch  2,  step= 194,  loss=0.314,  accuracy=0.906\n",
      "Epoch  2,  step= 195,  loss=0.824,  accuracy=0.812\n",
      "Epoch  2,  step= 196,  loss=0.525,  accuracy=0.875\n",
      "Epoch  2,  step= 197,  loss=0.420,  accuracy=0.906\n",
      "Epoch  2,  step= 198,  loss=0.268,  accuracy=0.906\n",
      "Epoch  2,  step= 199,  loss=0.245,  accuracy=0.906\n",
      "Epoch  2,  step= 200,  loss=0.527,  accuracy=0.875\n",
      "Epoch  2,  step= 201,  loss=0.538,  accuracy=0.906\n",
      "Epoch  2,  step= 202,  loss=0.283,  accuracy=0.938\n",
      "Epoch  2,  step= 203,  loss=0.349,  accuracy=0.906\n",
      "Epoch  2,  step= 204,  loss=0.345,  accuracy=0.844\n",
      "Epoch  2,  step= 205,  loss=0.331,  accuracy=0.906\n",
      "Epoch  2,  step= 206,  loss=0.274,  accuracy=0.906\n",
      "Epoch  2,  step= 207,  loss=0.262,  accuracy=0.938\n",
      "Epoch  2,  step= 208,  loss=0.226,  accuracy=0.938\n",
      "Epoch  2,  step= 209,  loss=0.220,  accuracy=0.938\n",
      "Epoch  2,  step= 210,  loss=0.296,  accuracy=0.938\n",
      "Epoch  2,  step= 211,  loss=0.087,  accuracy=1.000\n",
      "Epoch  2,  step= 212,  loss=0.367,  accuracy=0.875\n",
      "Epoch  2,  step= 213,  loss=0.642,  accuracy=0.812\n",
      "Epoch  2,  step= 214,  loss=0.514,  accuracy=0.875\n",
      "Epoch  2,  step= 215,  loss=0.541,  accuracy=0.875\n",
      "Epoch  2,  step= 216,  loss=0.971,  accuracy=0.812\n",
      "Epoch  2,  step= 217,  loss=0.343,  accuracy=0.906\n",
      "Epoch  2,  step= 218,  loss=0.171,  accuracy=0.938\n",
      "Epoch  2,  step= 219,  loss=0.810,  accuracy=0.844\n",
      "Epoch  2,  step= 220,  loss=0.350,  accuracy=0.875\n",
      "Epoch  2,  step= 221,  loss=0.135,  accuracy=0.938\n",
      "Epoch  2,  step= 222,  loss=0.251,  accuracy=0.875\n",
      "Epoch  2,  step= 223,  loss=0.603,  accuracy=0.906\n",
      "Epoch  2,  step= 224,  loss=0.219,  accuracy=0.938\n",
      "Epoch  2,  step= 225,  loss=0.735,  accuracy=0.750\n",
      "Epoch  2,  step= 226,  loss=0.818,  accuracy=0.781\n",
      "Epoch  2,  step= 227,  loss=0.628,  accuracy=0.812\n",
      "Epoch  2,  step= 228,  loss=0.293,  accuracy=0.938\n",
      "Epoch  2,  step= 229,  loss=0.581,  accuracy=0.844\n",
      "Epoch  2,  step= 230,  loss=0.291,  accuracy=0.906\n",
      "Epoch  2,  step= 231,  loss=0.048,  accuracy=1.000\n",
      "Epoch  2,  step= 232,  loss=0.412,  accuracy=0.875\n",
      "Epoch  2,  step= 233,  loss=0.079,  accuracy=1.000\n",
      "Epoch  2,  step= 234,  loss=0.209,  accuracy=0.875\n",
      "Epoch  2,  step= 235,  loss=0.635,  accuracy=0.875\n",
      "Epoch  2,  step= 236,  loss=0.083,  accuracy=0.969\n",
      "Epoch  2,  step= 237,  loss=0.772,  accuracy=0.750\n",
      "Epoch  2,  step= 238,  loss=0.803,  accuracy=0.812\n",
      "Epoch  2,  step= 239,  loss=0.288,  accuracy=0.875\n",
      "Epoch  2,  step= 240,  loss=0.287,  accuracy=0.906\n",
      "Epoch  2,  step= 241,  loss=0.696,  accuracy=0.812\n",
      "Epoch  2,  step= 242,  loss=0.433,  accuracy=0.844\n",
      "Epoch  2,  step= 243,  loss=0.341,  accuracy=0.844\n",
      "Epoch  2,  step= 244,  loss=0.308,  accuracy=0.906\n",
      "Epoch  2,  step= 245,  loss=0.400,  accuracy=0.906\n",
      "Epoch  2,  step= 246,  loss=0.648,  accuracy=0.906\n",
      "Epoch  2,  step= 247,  loss=0.459,  accuracy=0.875\n",
      "Epoch  2,  step= 248,  loss=0.252,  accuracy=0.906\n",
      "Epoch  2,  step= 249,  loss=0.450,  accuracy=0.875\n",
      "Epoch  2,  step= 250,  loss=0.474,  accuracy=0.844\n",
      "Epoch  2,  step= 251,  loss=0.212,  accuracy=0.906\n",
      "Epoch  2,  step= 252,  loss=0.302,  accuracy=0.875\n",
      "Epoch  2,  step= 253,  loss=0.681,  accuracy=0.812\n",
      "Epoch  2,  step= 254,  loss=0.172,  accuracy=0.906\n",
      "Epoch  2,  step= 255,  loss=0.097,  accuracy=0.969\n",
      "Epoch  2,  step= 256,  loss=1.036,  accuracy=0.781\n",
      "Epoch  2,  step= 257,  loss=0.497,  accuracy=0.906\n",
      "Epoch  2,  step= 258,  loss=0.325,  accuracy=0.906\n",
      "Epoch  2,  step= 259,  loss=0.250,  accuracy=0.906\n",
      "Epoch  2,  step= 260,  loss=0.274,  accuracy=0.906\n",
      "Epoch  2,  step= 261,  loss=0.160,  accuracy=0.938\n",
      "Epoch  2,  step= 262,  loss=0.239,  accuracy=0.906\n",
      "Epoch  2,  step= 263,  loss=0.551,  accuracy=0.875\n",
      "Epoch  2,  step= 264,  loss=0.822,  accuracy=0.875\n",
      "Epoch  2,  step= 265,  loss=0.367,  accuracy=0.812\n",
      "Epoch  2,  step= 266,  loss=0.343,  accuracy=0.875\n",
      "Epoch  2,  step= 267,  loss=0.321,  accuracy=0.875\n",
      "Epoch  2,  step= 268,  loss=0.194,  accuracy=0.906\n",
      "Epoch  2,  step= 269,  loss=0.559,  accuracy=0.812\n",
      "Epoch  2,  step= 270,  loss=0.288,  accuracy=0.875\n",
      "Epoch  2,  step= 271,  loss=1.112,  accuracy=0.719\n",
      "Epoch  2,  step= 272,  loss=1.475,  accuracy=0.719\n",
      "Epoch  2,  step= 273,  loss=0.382,  accuracy=0.844\n",
      "Epoch  2,  step= 274,  loss=0.466,  accuracy=0.781\n",
      "Epoch  2,  step= 275,  loss=0.517,  accuracy=0.906\n",
      "Epoch  2,  step= 276,  loss=0.578,  accuracy=0.844\n",
      "Epoch  2,  step= 277,  loss=0.732,  accuracy=0.781\n",
      "Epoch  2,  step= 278,  loss=0.802,  accuracy=0.750\n",
      "Epoch  2,  step= 279,  loss=0.406,  accuracy=0.906\n",
      "Epoch  2,  step= 280,  loss=0.419,  accuracy=0.906\n",
      "Epoch  2,  step= 281,  loss=0.370,  accuracy=0.875\n",
      "Epoch  2,  step= 282,  loss=0.161,  accuracy=0.938\n",
      "Epoch  2,  step= 283,  loss=0.345,  accuracy=0.938\n",
      "Epoch  2,  step= 284,  loss=0.663,  accuracy=0.844\n",
      "Epoch  2,  step= 285,  loss=0.620,  accuracy=0.906\n",
      "Epoch  2,  step= 286,  loss=0.496,  accuracy=0.844\n",
      "Epoch  2,  step= 287,  loss=0.331,  accuracy=0.906\n",
      "Epoch  2,  step= 288,  loss=0.253,  accuracy=0.938\n",
      "Epoch  2,  step= 289,  loss=0.911,  accuracy=0.781\n",
      "Epoch  2,  step= 290,  loss=0.424,  accuracy=0.906\n",
      "Epoch  2,  step= 291,  loss=0.248,  accuracy=0.906\n",
      "Epoch  2,  step= 292,  loss=0.304,  accuracy=0.906\n",
      "Epoch  2,  step= 293,  loss=0.813,  accuracy=0.781\n",
      "Epoch  2,  step= 294,  loss=0.364,  accuracy=0.906\n",
      "Epoch  2,  step= 295,  loss=0.495,  accuracy=0.875\n",
      "Epoch  2,  step= 296,  loss=0.081,  accuracy=0.969\n",
      "Epoch  2,  step= 297,  loss=0.346,  accuracy=0.844\n",
      "Epoch  2,  step= 298,  loss=0.357,  accuracy=0.906\n",
      "Epoch  2,  step= 299,  loss=0.339,  accuracy=0.906\n",
      "Epoch  2,  step= 300,  loss=0.329,  accuracy=0.938\n",
      "Epoch  2,  step= 301,  loss=0.328,  accuracy=0.875\n",
      "Epoch  2,  step= 302,  loss=0.429,  accuracy=0.875\n",
      "Epoch  2,  step= 303,  loss=0.052,  accuracy=1.000\n",
      "Epoch  2,  step= 304,  loss=0.217,  accuracy=0.938\n",
      "Epoch  2,  step= 305,  loss=0.757,  accuracy=0.812\n",
      "Epoch  2,  step= 306,  loss=0.482,  accuracy=0.875\n",
      "Epoch  2,  step= 307,  loss=0.047,  accuracy=1.000\n",
      "Epoch  2,  step= 308,  loss=0.265,  accuracy=0.906\n",
      "Epoch  2,  step= 309,  loss=0.076,  accuracy=0.969\n",
      "Epoch  2,  step= 310,  loss=0.323,  accuracy=0.906\n",
      "Epoch  2,  step= 311,  loss=0.237,  accuracy=0.906\n",
      "Epoch  2,  step= 312,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step= 313,  loss=0.870,  accuracy=0.844\n",
      "Epoch  2,  step= 314,  loss=0.263,  accuracy=0.906\n",
      "Epoch  2,  step= 315,  loss=0.409,  accuracy=0.875\n",
      "Epoch  2,  step= 316,  loss=0.417,  accuracy=0.875\n",
      "Epoch  2,  step= 317,  loss=0.771,  accuracy=0.750\n",
      "Epoch  2,  step= 318,  loss=0.731,  accuracy=0.812\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  2,  step= 319,  loss=0.532,  accuracy=0.844\n",
      "Epoch  2,  step= 320,  loss=0.801,  accuracy=0.812\n",
      "Epoch  2,  step= 321,  loss=0.762,  accuracy=0.781\n",
      "Epoch  2,  step= 322,  loss=0.344,  accuracy=0.969\n",
      "Epoch  2,  step= 323,  loss=0.424,  accuracy=0.938\n",
      "Epoch  2,  step= 324,  loss=0.209,  accuracy=0.906\n",
      "Epoch  2,  step= 325,  loss=0.161,  accuracy=0.938\n",
      "Epoch  2,  step= 326,  loss=0.326,  accuracy=0.875\n",
      "Epoch  2,  step= 327,  loss=0.214,  accuracy=0.938\n",
      "Epoch  2,  step= 328,  loss=0.304,  accuracy=0.906\n",
      "Epoch  2,  step= 329,  loss=0.057,  accuracy=1.000\n",
      "Epoch  2,  step= 330,  loss=0.224,  accuracy=0.938\n",
      "Epoch  2,  step= 331,  loss=0.124,  accuracy=0.969\n",
      "Epoch  2,  step= 332,  loss=0.173,  accuracy=0.906\n",
      "Epoch  2,  step= 333,  loss=0.223,  accuracy=0.969\n",
      "Epoch  2,  step= 334,  loss=0.274,  accuracy=0.875\n",
      "Epoch  2,  step= 335,  loss=0.699,  accuracy=0.875\n",
      "Epoch  2,  step= 336,  loss=0.469,  accuracy=0.875\n",
      "Epoch  2,  step= 337,  loss=0.266,  accuracy=0.938\n",
      "Epoch  2,  step= 338,  loss=0.279,  accuracy=0.906\n",
      "Epoch  2,  step= 339,  loss=0.477,  accuracy=0.906\n",
      "Epoch  2,  step= 340,  loss=0.205,  accuracy=0.938\n",
      "Epoch  2,  step= 341,  loss=0.166,  accuracy=0.906\n",
      "Epoch  2,  step= 342,  loss=0.210,  accuracy=0.906\n",
      "Epoch  2,  step= 343,  loss=0.492,  accuracy=0.906\n",
      "Epoch  2,  step= 344,  loss=0.246,  accuracy=0.969\n",
      "Epoch  2,  step= 345,  loss=0.307,  accuracy=0.906\n",
      "Epoch  2,  step= 346,  loss=0.150,  accuracy=0.969\n",
      "Epoch  2,  step= 347,  loss=0.275,  accuracy=0.938\n",
      "Epoch  2,  step= 348,  loss=0.200,  accuracy=0.906\n",
      "Epoch  2,  step= 349,  loss=0.161,  accuracy=0.938\n",
      "Epoch  2,  step= 350,  loss=0.670,  accuracy=0.844\n",
      "Epoch  2,  step= 351,  loss=0.522,  accuracy=0.844\n",
      "Epoch  2,  step= 352,  loss=0.190,  accuracy=0.938\n",
      "Epoch  2,  step= 353,  loss=0.283,  accuracy=0.906\n",
      "Epoch  2,  step= 354,  loss=0.200,  accuracy=0.938\n",
      "Epoch  2,  step= 355,  loss=0.365,  accuracy=0.812\n",
      "Epoch  2,  step= 356,  loss=0.160,  accuracy=0.969\n",
      "Epoch  2,  step= 357,  loss=0.271,  accuracy=0.906\n",
      "Epoch  2,  step= 358,  loss=0.128,  accuracy=0.969\n",
      "Epoch  2,  step= 359,  loss=0.295,  accuracy=0.812\n",
      "Epoch  2,  step= 360,  loss=0.258,  accuracy=0.906\n",
      "Epoch  2,  step= 361,  loss=0.865,  accuracy=0.750\n",
      "Epoch  2,  step= 362,  loss=0.374,  accuracy=0.812\n",
      "Epoch  2,  step= 363,  loss=0.753,  accuracy=0.844\n",
      "Epoch  2,  step= 364,  loss=0.340,  accuracy=0.906\n",
      "Epoch  2,  step= 365,  loss=0.633,  accuracy=0.781\n",
      "Epoch  2,  step= 366,  loss=0.287,  accuracy=0.906\n",
      "Epoch  2,  step= 367,  loss=0.659,  accuracy=0.906\n",
      "Epoch  2,  step= 368,  loss=0.381,  accuracy=0.812\n",
      "Epoch  2,  step= 369,  loss=0.207,  accuracy=0.844\n",
      "Epoch  2,  step= 370,  loss=0.466,  accuracy=0.844\n",
      "Epoch  2,  step= 371,  loss=0.251,  accuracy=0.906\n",
      "Epoch  2,  step= 372,  loss=0.129,  accuracy=0.938\n",
      "Epoch  2,  step= 373,  loss=0.374,  accuracy=0.906\n",
      "Epoch  2,  step= 374,  loss=0.273,  accuracy=0.906\n",
      "Epoch  2,  step= 375,  loss=0.422,  accuracy=0.906\n",
      "Epoch  2,  step= 376,  loss=0.507,  accuracy=0.844\n",
      "Epoch  2,  step= 377,  loss=0.428,  accuracy=0.844\n",
      "Epoch  2,  step= 378,  loss=0.330,  accuracy=0.938\n",
      "Epoch  2,  step= 379,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step= 380,  loss=0.362,  accuracy=0.875\n",
      "Epoch  2,  step= 381,  loss=0.399,  accuracy=0.906\n",
      "Epoch  2,  step= 382,  loss=0.248,  accuracy=0.938\n",
      "Epoch  2,  step= 383,  loss=0.725,  accuracy=0.812\n",
      "Epoch  2,  step= 384,  loss=0.322,  accuracy=0.875\n",
      "Epoch  2,  step= 385,  loss=0.382,  accuracy=0.875\n",
      "Epoch  2,  step= 386,  loss=0.621,  accuracy=0.812\n",
      "Epoch  2,  step= 387,  loss=0.082,  accuracy=0.969\n",
      "Epoch  2,  step= 388,  loss=0.435,  accuracy=0.844\n",
      "Epoch  2,  step= 389,  loss=0.472,  accuracy=0.875\n",
      "Epoch  2,  step= 390,  loss=0.774,  accuracy=0.812\n",
      "Epoch  2,  step= 391,  loss=0.405,  accuracy=0.844\n",
      "Epoch  2,  step= 392,  loss=1.067,  accuracy=0.812\n",
      "Epoch  2,  step= 393,  loss=0.838,  accuracy=0.844\n",
      "Epoch  2,  step= 394,  loss=0.708,  accuracy=0.875\n",
      "Epoch  2,  step= 395,  loss=0.772,  accuracy=0.844\n",
      "Epoch  2,  step= 396,  loss=1.336,  accuracy=0.719\n",
      "Epoch  2,  step= 397,  loss=0.112,  accuracy=0.969\n",
      "Epoch  2,  step= 398,  loss=0.387,  accuracy=0.875\n",
      "Epoch  2,  step= 399,  loss=0.426,  accuracy=0.906\n",
      "Epoch  2,  step= 400,  loss=0.206,  accuracy=0.938\n",
      "Epoch  2,  step= 401,  loss=0.316,  accuracy=0.875\n",
      "Epoch  2,  step= 402,  loss=0.512,  accuracy=0.812\n",
      "Epoch  2,  step= 403,  loss=0.312,  accuracy=0.906\n",
      "Epoch  2,  step= 404,  loss=0.903,  accuracy=0.844\n",
      "Epoch  2,  step= 405,  loss=0.446,  accuracy=0.812\n",
      "Epoch  2,  step= 406,  loss=0.846,  accuracy=0.844\n",
      "Epoch  2,  step= 407,  loss=0.681,  accuracy=0.812\n",
      "Epoch  2,  step= 408,  loss=1.198,  accuracy=0.750\n",
      "Epoch  2,  step= 409,  loss=0.450,  accuracy=0.781\n",
      "Epoch  2,  step= 410,  loss=0.456,  accuracy=0.844\n",
      "Epoch  2,  step= 411,  loss=0.193,  accuracy=0.875\n",
      "Epoch  2,  step= 412,  loss=0.654,  accuracy=0.781\n",
      "Epoch  2,  step= 413,  loss=0.094,  accuracy=0.969\n",
      "Epoch  2,  step= 414,  loss=0.782,  accuracy=0.875\n",
      "Epoch  2,  step= 415,  loss=0.372,  accuracy=0.844\n",
      "Epoch  2,  step= 416,  loss=0.195,  accuracy=0.938\n",
      "Epoch  2,  step= 417,  loss=0.240,  accuracy=0.938\n",
      "Epoch  2,  step= 418,  loss=0.416,  accuracy=0.844\n",
      "Epoch  2,  step= 419,  loss=0.303,  accuracy=0.906\n",
      "Epoch  2,  step= 420,  loss=0.396,  accuracy=0.969\n",
      "Epoch  2,  step= 421,  loss=0.042,  accuracy=1.000\n",
      "Epoch  2,  step= 422,  loss=0.598,  accuracy=0.875\n",
      "Epoch  2,  step= 423,  loss=0.384,  accuracy=0.906\n",
      "Epoch  2,  step= 424,  loss=0.049,  accuracy=1.000\n",
      "Epoch  2,  step= 425,  loss=0.176,  accuracy=0.906\n",
      "Epoch  2,  step= 426,  loss=0.710,  accuracy=0.844\n",
      "Epoch  2,  step= 427,  loss=0.604,  accuracy=0.812\n",
      "Epoch  2,  step= 428,  loss=0.665,  accuracy=0.844\n",
      "Epoch  2,  step= 429,  loss=0.514,  accuracy=0.781\n",
      "Epoch  2,  step= 430,  loss=0.095,  accuracy=0.969\n",
      "Epoch  2,  step= 431,  loss=0.169,  accuracy=0.938\n",
      "Epoch  2,  step= 432,  loss=0.798,  accuracy=0.812\n",
      "Epoch  2,  step= 433,  loss=0.371,  accuracy=0.906\n",
      "Epoch  2,  step= 434,  loss=0.224,  accuracy=0.938\n",
      "Epoch  2,  step= 435,  loss=0.312,  accuracy=0.938\n",
      "Epoch  2,  step= 436,  loss=0.529,  accuracy=0.875\n",
      "Epoch  2,  step= 437,  loss=0.586,  accuracy=0.844\n",
      "Epoch  2,  step= 438,  loss=0.342,  accuracy=0.812\n",
      "Epoch  2,  step= 439,  loss=0.670,  accuracy=0.875\n",
      "Epoch  2,  step= 440,  loss=0.265,  accuracy=0.938\n",
      "Epoch  2,  step= 441,  loss=0.775,  accuracy=0.781\n",
      "Epoch  2,  step= 442,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step= 443,  loss=0.421,  accuracy=0.844\n",
      "Epoch  2,  step= 444,  loss=0.799,  accuracy=0.875\n",
      "Epoch  2,  step= 445,  loss=0.513,  accuracy=0.875\n",
      "Epoch  2,  step= 446,  loss=0.403,  accuracy=0.938\n",
      "Epoch  2,  step= 447,  loss=0.368,  accuracy=0.875\n",
      "Epoch  2,  step= 448,  loss=1.325,  accuracy=0.750\n",
      "Epoch  2,  step= 449,  loss=0.519,  accuracy=0.812\n",
      "Epoch  2,  step= 450,  loss=0.230,  accuracy=0.938\n",
      "Epoch  2,  step= 451,  loss=0.316,  accuracy=0.938\n",
      "Epoch  2,  step= 452,  loss=0.257,  accuracy=0.906\n",
      "Epoch  2,  step= 453,  loss=0.313,  accuracy=0.875\n",
      "Epoch  2,  step= 454,  loss=0.695,  accuracy=0.812\n",
      "Epoch  2,  step= 455,  loss=0.457,  accuracy=0.812\n",
      "Epoch  2,  step= 456,  loss=0.631,  accuracy=0.812\n",
      "Epoch  2,  step= 457,  loss=0.553,  accuracy=0.812\n",
      "Epoch  2,  step= 458,  loss=0.671,  accuracy=0.781\n",
      "Epoch  2,  step= 459,  loss=0.690,  accuracy=0.812\n",
      "Epoch  2,  step= 460,  loss=0.956,  accuracy=0.688\n",
      "Epoch  2,  step= 461,  loss=1.395,  accuracy=0.750\n",
      "Epoch  2,  step= 462,  loss=0.613,  accuracy=0.844\n",
      "Epoch  2,  step= 463,  loss=0.791,  accuracy=0.812\n",
      "Epoch  2,  step= 464,  loss=0.375,  accuracy=0.906\n",
      "Epoch  2,  step= 465,  loss=0.323,  accuracy=0.906\n",
      "Epoch  2,  step= 466,  loss=0.316,  accuracy=0.875\n",
      "Epoch  2,  step= 467,  loss=0.190,  accuracy=0.938\n",
      "Epoch  2,  step= 468,  loss=0.424,  accuracy=0.844\n",
      "Epoch  2,  step= 469,  loss=0.112,  accuracy=0.969\n",
      "Epoch  2,  step= 470,  loss=0.562,  accuracy=0.969\n",
      "Epoch  2,  step= 471,  loss=0.122,  accuracy=0.938\n",
      "Epoch  2,  step= 472,  loss=1.393,  accuracy=0.781\n",
      "Epoch  2,  step= 473,  loss=0.763,  accuracy=0.875\n",
      "Epoch  2,  step= 474,  loss=0.778,  accuracy=0.875\n",
      "Epoch  2,  step= 475,  loss=0.329,  accuracy=0.938\n",
      "Epoch  2,  step= 476,  loss=0.260,  accuracy=0.906\n",
      "Epoch  2,  step= 477,  loss=0.448,  accuracy=0.906\n",
      "Epoch  2,  step= 478,  loss=0.104,  accuracy=0.969\n",
      "Epoch  2,  step= 479,  loss=0.533,  accuracy=0.844\n",
      "Epoch  2,  step= 480,  loss=0.180,  accuracy=0.938\n",
      "Epoch  2,  step= 481,  loss=0.381,  accuracy=0.875\n",
      "Epoch  2,  step= 482,  loss=0.573,  accuracy=0.906\n",
      "Epoch  2,  step= 483,  loss=0.187,  accuracy=0.906\n",
      "Epoch  2,  step= 484,  loss=0.501,  accuracy=0.875\n",
      "Epoch  2,  step= 485,  loss=0.539,  accuracy=0.875\n",
      "Epoch  2,  step= 486,  loss=0.265,  accuracy=0.906\n",
      "Epoch  2,  step= 487,  loss=0.440,  accuracy=0.938\n",
      "Epoch  2,  step= 488,  loss=0.297,  accuracy=0.875\n",
      "Epoch  2,  step= 489,  loss=0.323,  accuracy=0.875\n",
      "Epoch  2,  step= 490,  loss=0.367,  accuracy=0.938\n",
      "Epoch  2,  step= 491,  loss=1.013,  accuracy=0.719\n",
      "Epoch  2,  step= 492,  loss=0.768,  accuracy=0.781\n",
      "Epoch  2,  step= 493,  loss=0.560,  accuracy=0.812\n",
      "Epoch  2,  step= 494,  loss=0.257,  accuracy=0.875\n",
      "Epoch  2,  step= 495,  loss=1.057,  accuracy=0.750\n",
      "Epoch  2,  step= 496,  loss=0.279,  accuracy=0.906\n",
      "Epoch  2,  step= 497,  loss=0.343,  accuracy=0.906\n",
      "Epoch  2,  step= 498,  loss=0.527,  accuracy=0.938\n",
      "Epoch  2,  step= 499,  loss=0.582,  accuracy=0.812\n",
      "Epoch  2,  step= 500,  loss=0.834,  accuracy=0.812\n",
      "Epoch  2,  step= 501,  loss=0.900,  accuracy=0.750\n",
      "Epoch  2,  step= 502,  loss=0.457,  accuracy=0.875\n",
      "Epoch  2,  step= 503,  loss=0.700,  accuracy=0.875\n",
      "Epoch  2,  step= 504,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step= 505,  loss=0.165,  accuracy=0.938\n",
      "Epoch  2,  step= 506,  loss=0.505,  accuracy=0.781\n",
      "Epoch  2,  step= 507,  loss=0.053,  accuracy=0.969\n",
      "Epoch  2,  step= 508,  loss=0.617,  accuracy=0.844\n",
      "Epoch  2,  step= 509,  loss=0.327,  accuracy=0.875\n",
      "Epoch  2,  step= 510,  loss=0.098,  accuracy=0.969\n",
      "Epoch  2,  step= 511,  loss=0.561,  accuracy=0.844\n",
      "Epoch  2,  step= 512,  loss=0.138,  accuracy=0.969\n",
      "Epoch  2,  step= 513,  loss=0.366,  accuracy=0.875\n",
      "Epoch  2,  step= 514,  loss=0.257,  accuracy=0.938\n",
      "Epoch  2,  step= 515,  loss=0.388,  accuracy=0.875\n",
      "Epoch  2,  step= 516,  loss=0.355,  accuracy=0.938\n",
      "Epoch  2,  step= 517,  loss=0.719,  accuracy=0.844\n",
      "Epoch  2,  step= 518,  loss=0.301,  accuracy=0.875\n",
      "Epoch  2,  step= 519,  loss=0.222,  accuracy=0.938\n",
      "Epoch  2,  step= 520,  loss=0.577,  accuracy=0.844\n",
      "Epoch  2,  step= 521,  loss=0.650,  accuracy=0.812\n",
      "Epoch  2,  step= 522,  loss=0.202,  accuracy=0.969\n",
      "Epoch  2,  step= 523,  loss=0.572,  accuracy=0.875\n",
      "Epoch  2,  step= 524,  loss=0.931,  accuracy=0.781\n",
      "Epoch  2,  step= 525,  loss=0.328,  accuracy=0.844\n",
      "Epoch  2,  step= 526,  loss=0.299,  accuracy=0.906\n",
      "Epoch  2,  step= 527,  loss=0.524,  accuracy=0.875\n",
      "Epoch  2,  step= 528,  loss=0.379,  accuracy=0.844\n",
      "Epoch  2,  step= 529,  loss=0.601,  accuracy=0.812\n",
      "Epoch  2,  step= 530,  loss=0.933,  accuracy=0.781\n",
      "Epoch  2,  step= 531,  loss=0.572,  accuracy=0.750\n",
      "Epoch  2,  step= 532,  loss=0.423,  accuracy=0.906\n",
      "Epoch  2,  step= 533,  loss=0.223,  accuracy=0.906\n",
      "Epoch  2,  step= 534,  loss=0.313,  accuracy=0.906\n",
      "Epoch  2,  step= 535,  loss=0.272,  accuracy=0.938\n",
      "Epoch  2,  step= 536,  loss=0.424,  accuracy=0.906\n",
      "Epoch  2,  step= 537,  loss=0.882,  accuracy=0.875\n",
      "Epoch  2,  step= 538,  loss=0.420,  accuracy=0.875\n",
      "Epoch  2,  step= 539,  loss=0.261,  accuracy=0.969\n",
      "Epoch  2,  step= 540,  loss=0.109,  accuracy=0.969\n",
      "Epoch  2,  step= 541,  loss=0.062,  accuracy=1.000\n",
      "Epoch  2,  step= 542,  loss=0.201,  accuracy=0.906\n",
      "Epoch  2,  step= 543,  loss=0.378,  accuracy=0.844\n",
      "Epoch  2,  step= 544,  loss=0.145,  accuracy=0.938\n",
      "Epoch  2,  step= 545,  loss=0.186,  accuracy=0.875\n",
      "Epoch  2,  step= 546,  loss=0.798,  accuracy=0.906\n",
      "Epoch  2,  step= 547,  loss=0.270,  accuracy=0.906\n",
      "Epoch  2,  step= 548,  loss=0.696,  accuracy=0.844\n",
      "Epoch  2,  step= 549,  loss=0.551,  accuracy=0.812\n",
      "Epoch  2,  step= 550,  loss=0.127,  accuracy=1.000\n",
      "Epoch  2,  step= 551,  loss=0.770,  accuracy=0.844\n",
      "Epoch  2,  step= 552,  loss=0.451,  accuracy=0.844\n",
      "Epoch  2,  step= 553,  loss=0.842,  accuracy=0.812\n",
      "Epoch  2,  step= 554,  loss=0.242,  accuracy=0.906\n",
      "Epoch  2,  step= 555,  loss=0.314,  accuracy=0.938\n",
      "Epoch  2,  step= 556,  loss=0.677,  accuracy=0.719\n",
      "Epoch  2,  step= 557,  loss=0.360,  accuracy=0.875\n",
      "Epoch  2,  step= 558,  loss=0.236,  accuracy=0.875\n",
      "Epoch  2,  step= 559,  loss=0.199,  accuracy=0.906\n",
      "Epoch  2,  step= 560,  loss=0.329,  accuracy=0.906\n",
      "Epoch  2,  step= 561,  loss=0.285,  accuracy=0.906\n",
      "Epoch  2,  step= 562,  loss=0.236,  accuracy=0.938\n",
      "Epoch  2,  step= 563,  loss=0.358,  accuracy=0.875\n",
      "Epoch  2,  step= 564,  loss=0.463,  accuracy=0.938\n",
      "Epoch  2,  step= 565,  loss=0.513,  accuracy=0.875\n",
      "Epoch  2,  step= 566,  loss=0.069,  accuracy=1.000\n",
      "Epoch  2,  step= 567,  loss=0.337,  accuracy=0.906\n",
      "Epoch  2,  step= 568,  loss=0.165,  accuracy=0.938\n",
      "Epoch  2,  step= 569,  loss=0.152,  accuracy=0.969\n",
      "Epoch  2,  step= 570,  loss=0.651,  accuracy=0.844\n",
      "Epoch  2,  step= 571,  loss=0.268,  accuracy=0.938\n",
      "Epoch  2,  step= 572,  loss=0.223,  accuracy=0.906\n",
      "Epoch  2,  step= 573,  loss=0.194,  accuracy=0.938\n",
      "Epoch  2,  step= 574,  loss=0.320,  accuracy=0.938\n",
      "Epoch  2,  step= 575,  loss=0.525,  accuracy=0.781\n",
      "Epoch  2,  step= 576,  loss=0.175,  accuracy=0.938\n",
      "Epoch  2,  step= 577,  loss=0.651,  accuracy=0.906\n",
      "Epoch  2,  step= 578,  loss=0.545,  accuracy=0.812\n",
      "Epoch  2,  step= 579,  loss=0.166,  accuracy=0.938\n",
      "Epoch  2,  step= 580,  loss=0.562,  accuracy=0.844\n",
      "Epoch  2,  step= 581,  loss=0.340,  accuracy=0.906\n",
      "Epoch  2,  step= 582,  loss=0.646,  accuracy=0.844\n",
      "Epoch  2,  step= 583,  loss=0.309,  accuracy=0.906\n",
      "Epoch  2,  step= 584,  loss=0.162,  accuracy=0.938\n",
      "Epoch  2,  step= 585,  loss=0.330,  accuracy=0.875\n",
      "Epoch  2,  step= 586,  loss=0.057,  accuracy=0.969\n",
      "Epoch  2,  step= 587,  loss=0.165,  accuracy=0.875\n",
      "Epoch  2,  step= 588,  loss=0.391,  accuracy=0.844\n",
      "Epoch  2,  step= 589,  loss=0.232,  accuracy=0.906\n",
      "Epoch  2,  step= 590,  loss=0.195,  accuracy=0.938\n",
      "Epoch  2,  step= 591,  loss=0.437,  accuracy=0.875\n",
      "Epoch  2,  step= 592,  loss=0.092,  accuracy=0.969\n",
      "Epoch  2,  step= 593,  loss=0.314,  accuracy=0.906\n",
      "Epoch  2,  step= 594,  loss=0.144,  accuracy=0.969\n",
      "Epoch  2,  step= 595,  loss=0.166,  accuracy=0.969\n",
      "Epoch  2,  step= 596,  loss=0.647,  accuracy=0.938\n",
      "Epoch  2,  step= 597,  loss=0.660,  accuracy=0.844\n",
      "Epoch  2,  step= 598,  loss=0.656,  accuracy=0.781\n",
      "Epoch  2,  step= 599,  loss=0.247,  accuracy=0.906\n",
      "Epoch  2,  step= 600,  loss=0.335,  accuracy=0.906\n",
      "Epoch  2,  step= 601,  loss=0.430,  accuracy=0.938\n",
      "Epoch  2,  step= 602,  loss=0.414,  accuracy=0.906\n",
      "Epoch  2,  step= 603,  loss=0.327,  accuracy=0.938\n",
      "Epoch  2,  step= 604,  loss=0.440,  accuracy=0.906\n",
      "Epoch  2,  step= 605,  loss=0.578,  accuracy=0.812\n",
      "Epoch  2,  step= 606,  loss=0.377,  accuracy=0.906\n",
      "Epoch  2,  step= 607,  loss=0.096,  accuracy=1.000\n",
      "Epoch  2,  step= 608,  loss=0.260,  accuracy=0.906\n",
      "Epoch  2,  step= 609,  loss=0.699,  accuracy=0.812\n",
      "Epoch  2,  step= 610,  loss=0.398,  accuracy=0.812\n",
      "Epoch  2,  step= 611,  loss=0.387,  accuracy=0.844\n",
      "Epoch  2,  step= 612,  loss=0.349,  accuracy=0.906\n",
      "Epoch  2,  step= 613,  loss=0.359,  accuracy=0.906\n",
      "Epoch  2,  step= 614,  loss=0.123,  accuracy=0.969\n",
      "Epoch  2,  step= 615,  loss=0.163,  accuracy=0.938\n",
      "Epoch  2,  step= 616,  loss=0.064,  accuracy=0.969\n",
      "Epoch  2,  step= 617,  loss=0.091,  accuracy=0.969\n",
      "Epoch  2,  step= 618,  loss=0.388,  accuracy=0.906\n",
      "Epoch  2,  step= 619,  loss=1.124,  accuracy=0.812\n",
      "Epoch  2,  step= 620,  loss=0.383,  accuracy=0.906\n",
      "Epoch  2,  step= 621,  loss=0.088,  accuracy=1.000\n",
      "Epoch  2,  step= 622,  loss=0.686,  accuracy=0.812\n",
      "Epoch  2,  step= 623,  loss=0.294,  accuracy=0.938\n",
      "Epoch  2,  step= 624,  loss=0.318,  accuracy=0.875\n",
      "Epoch  2,  step= 625,  loss=0.456,  accuracy=0.812\n",
      "Epoch  2,  step= 626,  loss=0.431,  accuracy=0.812\n",
      "Epoch  2,  step= 627,  loss=0.347,  accuracy=0.906\n",
      "Epoch  2,  step= 628,  loss=0.404,  accuracy=0.875\n",
      "Epoch  2,  step= 629,  loss=0.350,  accuracy=0.875\n",
      "Epoch  2,  step= 630,  loss=0.866,  accuracy=0.750\n",
      "Epoch  2,  step= 631,  loss=0.207,  accuracy=0.938\n",
      "Epoch  2,  step= 632,  loss=0.611,  accuracy=0.906\n",
      "Epoch  2,  step= 633,  loss=0.227,  accuracy=0.906\n",
      "Epoch  2,  step= 634,  loss=0.181,  accuracy=0.906\n",
      "Epoch  2,  step= 635,  loss=0.517,  accuracy=0.875\n",
      "Epoch  2,  step= 636,  loss=0.167,  accuracy=0.969\n",
      "Epoch  2,  step= 637,  loss=0.193,  accuracy=0.906\n",
      "Epoch  2,  step= 638,  loss=0.113,  accuracy=1.000\n",
      "Epoch  2,  step= 639,  loss=0.156,  accuracy=0.969\n",
      "Epoch  2,  step= 640,  loss=0.039,  accuracy=1.000\n",
      "Epoch  2,  step= 641,  loss=0.421,  accuracy=0.875\n",
      "Epoch  2,  step= 642,  loss=0.813,  accuracy=0.844\n",
      "Epoch  2,  step= 643,  loss=0.219,  accuracy=0.906\n",
      "Epoch  2,  step= 644,  loss=0.307,  accuracy=0.938\n",
      "Epoch  2,  step= 645,  loss=0.495,  accuracy=0.906\n",
      "Epoch  2,  step= 646,  loss=0.319,  accuracy=0.938\n",
      "Epoch  2,  step= 647,  loss=0.917,  accuracy=0.719\n",
      "Epoch  2,  step= 648,  loss=0.265,  accuracy=0.938\n",
      "Epoch  2,  step= 649,  loss=0.894,  accuracy=0.875\n",
      "Epoch  2,  step= 650,  loss=0.640,  accuracy=0.875\n",
      "Epoch  2,  step= 651,  loss=0.163,  accuracy=0.906\n",
      "Epoch  2,  step= 652,  loss=0.464,  accuracy=0.906\n",
      "Epoch  2,  step= 653,  loss=0.732,  accuracy=0.781\n",
      "Epoch  2,  step= 654,  loss=0.593,  accuracy=0.844\n",
      "Epoch  2,  step= 655,  loss=0.693,  accuracy=0.719\n",
      "Epoch  2,  step= 656,  loss=0.571,  accuracy=0.844\n",
      "Epoch  2,  step= 657,  loss=0.191,  accuracy=0.938\n",
      "Epoch  2,  step= 658,  loss=0.294,  accuracy=0.969\n",
      "Epoch  2,  step= 659,  loss=0.258,  accuracy=0.906\n",
      "Epoch  2,  step= 660,  loss=0.708,  accuracy=0.844\n",
      "Epoch  2,  step= 661,  loss=0.339,  accuracy=0.938\n",
      "Epoch  2,  step= 662,  loss=0.677,  accuracy=0.844\n",
      "Epoch  2,  step= 663,  loss=0.115,  accuracy=0.969\n",
      "Epoch  2,  step= 664,  loss=0.124,  accuracy=0.938\n",
      "Epoch  2,  step= 665,  loss=0.644,  accuracy=0.844\n",
      "Epoch  2,  step= 666,  loss=0.281,  accuracy=0.969\n",
      "Epoch  2,  step= 667,  loss=0.343,  accuracy=0.906\n",
      "Epoch  2,  step= 668,  loss=0.284,  accuracy=0.969\n",
      "Epoch  2,  step= 669,  loss=0.576,  accuracy=0.906\n",
      "Epoch  2,  step= 670,  loss=0.261,  accuracy=0.906\n",
      "Epoch  2,  step= 671,  loss=0.190,  accuracy=0.969\n",
      "Epoch  2,  step= 672,  loss=0.289,  accuracy=0.875\n",
      "Epoch  2,  step= 673,  loss=0.214,  accuracy=0.938\n",
      "Epoch  2,  step= 674,  loss=0.621,  accuracy=0.875\n",
      "Epoch  2,  step= 675,  loss=0.613,  accuracy=0.875\n",
      "Epoch  2,  step= 676,  loss=0.373,  accuracy=0.906\n",
      "Epoch  2,  step= 677,  loss=0.217,  accuracy=0.938\n",
      "Epoch  2,  step= 678,  loss=0.541,  accuracy=0.844\n",
      "Epoch  2,  step= 679,  loss=0.264,  accuracy=0.875\n",
      "Epoch  2,  step= 680,  loss=0.320,  accuracy=0.938\n",
      "Epoch  2,  step= 681,  loss=0.052,  accuracy=1.000\n",
      "Epoch  2,  step= 682,  loss=0.080,  accuracy=0.969\n",
      "Epoch  2,  step= 683,  loss=0.104,  accuracy=0.969\n",
      "Epoch  2,  step= 684,  loss=0.324,  accuracy=0.906\n",
      "Epoch  2,  step= 685,  loss=0.357,  accuracy=0.844\n",
      "Epoch  2,  step= 686,  loss=0.252,  accuracy=0.969\n",
      "Epoch  2,  step= 687,  loss=0.258,  accuracy=0.938\n",
      "Epoch  2,  step= 688,  loss=0.126,  accuracy=0.969\n",
      "Epoch  2,  step= 689,  loss=0.318,  accuracy=0.875\n",
      "Epoch  2,  step= 690,  loss=0.358,  accuracy=0.844\n",
      "Epoch  2,  step= 691,  loss=0.536,  accuracy=0.844\n",
      "Epoch  2,  step= 692,  loss=0.486,  accuracy=0.844\n",
      "Epoch  2,  step= 693,  loss=0.564,  accuracy=0.875\n",
      "Epoch  2,  step= 694,  loss=1.368,  accuracy=0.875\n",
      "Epoch  2,  step= 695,  loss=0.090,  accuracy=1.000\n",
      "Epoch  2,  step= 696,  loss=0.579,  accuracy=0.844\n",
      "Epoch  2,  step= 697,  loss=0.176,  accuracy=0.938\n",
      "Epoch  2,  step= 698,  loss=0.042,  accuracy=1.000\n",
      "Epoch  2,  step= 699,  loss=0.112,  accuracy=0.969\n",
      "Epoch  2,  step= 700,  loss=0.210,  accuracy=0.906\n",
      "Epoch  2,  step= 701,  loss=0.067,  accuracy=1.000\n",
      "Epoch  2,  step= 702,  loss=0.898,  accuracy=0.781\n",
      "Epoch  2,  step= 703,  loss=0.625,  accuracy=0.875\n",
      "Epoch  2,  step= 704,  loss=0.819,  accuracy=0.781\n",
      "Epoch  2,  step= 705,  loss=0.774,  accuracy=0.719\n",
      "Epoch  2,  step= 706,  loss=0.616,  accuracy=0.750\n",
      "Epoch  2,  step= 707,  loss=0.340,  accuracy=0.906\n",
      "Epoch  2,  step= 708,  loss=0.305,  accuracy=0.844\n",
      "Epoch  2,  step= 709,  loss=0.636,  accuracy=0.844\n",
      "Epoch  2,  step= 710,  loss=0.327,  accuracy=0.844\n",
      "Epoch  2,  step= 711,  loss=0.585,  accuracy=0.906\n",
      "Epoch  2,  step= 712,  loss=0.576,  accuracy=0.844\n",
      "Epoch  2,  step= 713,  loss=0.278,  accuracy=0.875\n",
      "Epoch  2,  step= 714,  loss=0.189,  accuracy=0.938\n",
      "Epoch  2,  step= 715,  loss=0.250,  accuracy=0.906\n",
      "Epoch  2,  step= 716,  loss=0.133,  accuracy=0.969\n",
      "Epoch  2,  step= 717,  loss=0.025,  accuracy=1.000\n",
      "Epoch  2,  step= 718,  loss=0.088,  accuracy=0.969\n",
      "Epoch  2,  step= 719,  loss=0.492,  accuracy=0.875\n",
      "Epoch  2,  step= 720,  loss=0.297,  accuracy=0.906\n",
      "Epoch  2,  step= 721,  loss=0.606,  accuracy=0.781\n",
      "Epoch  2,  step= 722,  loss=0.169,  accuracy=0.938\n",
      "Epoch  2,  step= 723,  loss=0.356,  accuracy=0.875\n",
      "Epoch  2,  step= 724,  loss=0.279,  accuracy=0.938\n",
      "Epoch  2,  step= 725,  loss=0.429,  accuracy=0.875\n",
      "Epoch  2,  step= 726,  loss=0.278,  accuracy=0.906\n",
      "Epoch  2,  step= 727,  loss=0.186,  accuracy=0.906\n",
      "Epoch  2,  step= 728,  loss=0.352,  accuracy=0.844\n",
      "Epoch  2,  step= 729,  loss=0.260,  accuracy=0.969\n",
      "Epoch  2,  step= 730,  loss=0.706,  accuracy=0.906\n",
      "Epoch  2,  step= 731,  loss=0.278,  accuracy=0.844\n",
      "Epoch  2,  step= 732,  loss=0.253,  accuracy=0.938\n",
      "Epoch  2,  step= 733,  loss=0.676,  accuracy=0.812\n",
      "Epoch  2,  step= 734,  loss=0.362,  accuracy=0.906\n",
      "Epoch  2,  step= 735,  loss=0.109,  accuracy=0.938\n",
      "Epoch  2,  step= 736,  loss=0.266,  accuracy=0.906\n",
      "Epoch  2,  step= 737,  loss=0.704,  accuracy=0.875\n",
      "Epoch  2,  step= 738,  loss=0.374,  accuracy=0.875\n",
      "Epoch  2,  step= 739,  loss=0.224,  accuracy=0.938\n",
      "Epoch  2,  step= 740,  loss=0.497,  accuracy=0.844\n",
      "Epoch  2,  step= 741,  loss=1.062,  accuracy=0.812\n",
      "Epoch  2,  step= 742,  loss=0.156,  accuracy=0.969\n",
      "Epoch  2,  step= 743,  loss=0.222,  accuracy=0.906\n",
      "Epoch  2,  step= 744,  loss=0.563,  accuracy=0.875\n",
      "Epoch  2,  step= 745,  loss=0.520,  accuracy=0.812\n",
      "Epoch  2,  step= 746,  loss=0.339,  accuracy=0.875\n",
      "Epoch  2,  step= 747,  loss=0.215,  accuracy=0.875\n",
      "Epoch  2,  step= 748,  loss=0.637,  accuracy=0.938\n",
      "Epoch  2,  step= 749,  loss=0.109,  accuracy=0.969\n",
      "Epoch  2,  step= 750,  loss=0.266,  accuracy=0.875\n",
      "Epoch  2,  step= 751,  loss=0.409,  accuracy=0.812\n",
      "Epoch  2,  step= 752,  loss=0.140,  accuracy=0.938\n",
      "Epoch  2,  step= 753,  loss=0.608,  accuracy=0.750\n",
      "Epoch  2,  step= 754,  loss=0.164,  accuracy=0.938\n",
      "Epoch  2,  step= 755,  loss=0.420,  accuracy=0.875\n",
      "Epoch  2,  step= 756,  loss=0.761,  accuracy=0.812\n",
      "Epoch  2,  step= 757,  loss=0.198,  accuracy=0.938\n",
      "Epoch  2,  step= 758,  loss=0.840,  accuracy=0.812\n",
      "Epoch  2,  step= 759,  loss=0.208,  accuracy=0.938\n",
      "Epoch  2,  step= 760,  loss=0.118,  accuracy=0.938\n",
      "Epoch  2,  step= 761,  loss=0.335,  accuracy=0.875\n",
      "Epoch  2,  step= 762,  loss=0.203,  accuracy=0.969\n",
      "Epoch  2,  step= 763,  loss=0.240,  accuracy=0.938\n",
      "Epoch  2,  step= 764,  loss=0.346,  accuracy=0.875\n",
      "Epoch  2,  step= 765,  loss=0.337,  accuracy=0.906\n",
      "Epoch  2,  step= 766,  loss=0.620,  accuracy=0.750\n",
      "Epoch  2,  step= 767,  loss=0.260,  accuracy=0.875\n",
      "Epoch  2,  step= 768,  loss=0.571,  accuracy=0.844\n",
      "Epoch  2,  step= 769,  loss=0.821,  accuracy=0.750\n",
      "Epoch  2,  step= 770,  loss=0.854,  accuracy=0.812\n",
      "Epoch  2,  step= 771,  loss=0.421,  accuracy=0.906\n",
      "Epoch  2,  step= 772,  loss=0.679,  accuracy=0.750\n",
      "Epoch  2,  step= 773,  loss=0.444,  accuracy=0.844\n",
      "Epoch  2,  step= 774,  loss=0.779,  accuracy=0.875\n",
      "Epoch  2,  step= 775,  loss=0.681,  accuracy=0.750\n",
      "Epoch  2,  step= 776,  loss=0.186,  accuracy=0.938\n",
      "Epoch  2,  step= 777,  loss=0.429,  accuracy=0.875\n",
      "Epoch  2,  step= 778,  loss=0.341,  accuracy=0.906\n",
      "Epoch  2,  step= 779,  loss=0.715,  accuracy=0.812\n",
      "Epoch  2,  step= 780,  loss=0.405,  accuracy=0.812\n",
      "Epoch  2,  step= 781,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step= 782,  loss=0.250,  accuracy=0.906\n",
      "Epoch  2,  step= 783,  loss=0.107,  accuracy=1.000\n",
      "Epoch  2,  step= 784,  loss=0.706,  accuracy=0.906\n",
      "Epoch  2,  step= 785,  loss=0.432,  accuracy=0.844\n",
      "Epoch  2,  step= 786,  loss=0.604,  accuracy=0.844\n",
      "Epoch  2,  step= 787,  loss=0.455,  accuracy=0.844\n",
      "Epoch  2,  step= 788,  loss=0.406,  accuracy=0.938\n",
      "Epoch  2,  step= 789,  loss=0.224,  accuracy=0.938\n",
      "Epoch  2,  step= 790,  loss=0.516,  accuracy=0.844\n",
      "Epoch  2,  step= 791,  loss=0.583,  accuracy=0.875\n",
      "Epoch  2,  step= 792,  loss=0.086,  accuracy=0.969\n",
      "Epoch  2,  step= 793,  loss=0.294,  accuracy=0.938\n",
      "Epoch  2,  step= 794,  loss=0.035,  accuracy=1.000\n",
      "Epoch  2,  step= 795,  loss=0.538,  accuracy=0.875\n",
      "Epoch  2,  step= 796,  loss=0.527,  accuracy=0.875\n",
      "Epoch  2,  step= 797,  loss=0.220,  accuracy=0.938\n",
      "Epoch  2,  step= 798,  loss=0.552,  accuracy=0.875\n",
      "Epoch  2,  step= 799,  loss=0.235,  accuracy=0.875\n",
      "Epoch  2,  step= 800,  loss=0.254,  accuracy=0.938\n",
      "Epoch  2,  step= 801,  loss=0.351,  accuracy=0.875\n",
      "Epoch  2,  step= 802,  loss=0.380,  accuracy=0.938\n",
      "Epoch  2,  step= 803,  loss=0.277,  accuracy=0.906\n",
      "Epoch  2,  step= 804,  loss=0.476,  accuracy=0.812\n",
      "Epoch  2,  step= 805,  loss=0.448,  accuracy=0.938\n",
      "Epoch  2,  step= 806,  loss=0.993,  accuracy=0.781\n",
      "Epoch  2,  step= 807,  loss=0.229,  accuracy=0.906\n",
      "Epoch  2,  step= 808,  loss=0.089,  accuracy=0.938\n",
      "Epoch  2,  step= 809,  loss=0.465,  accuracy=0.906\n",
      "Epoch  2,  step= 810,  loss=0.410,  accuracy=0.812\n",
      "Epoch  2,  step= 811,  loss=0.379,  accuracy=0.875\n",
      "Epoch  2,  step= 812,  loss=0.118,  accuracy=0.969\n",
      "Epoch  2,  step= 813,  loss=0.135,  accuracy=0.938\n",
      "Epoch  2,  step= 814,  loss=0.489,  accuracy=0.844\n",
      "Epoch  2,  step= 815,  loss=0.088,  accuracy=0.969\n",
      "Epoch  2,  step= 816,  loss=0.196,  accuracy=0.938\n",
      "Epoch  2,  step= 817,  loss=0.101,  accuracy=0.969\n",
      "Epoch  2,  step= 818,  loss=0.324,  accuracy=0.938\n",
      "Epoch  2,  step= 819,  loss=0.279,  accuracy=0.938\n",
      "Epoch  2,  step= 820,  loss=0.167,  accuracy=0.938\n",
      "Epoch  2,  step= 821,  loss=0.554,  accuracy=0.844\n",
      "Epoch  2,  step= 822,  loss=0.178,  accuracy=0.938\n",
      "Epoch  2,  step= 823,  loss=0.833,  accuracy=0.781\n",
      "Epoch  2,  step= 824,  loss=0.899,  accuracy=0.906\n",
      "Epoch  2,  step= 825,  loss=0.303,  accuracy=0.938\n",
      "Epoch  2,  step= 826,  loss=0.310,  accuracy=0.906\n",
      "Epoch  2,  step= 827,  loss=0.383,  accuracy=0.906\n",
      "Epoch  2,  step= 828,  loss=0.378,  accuracy=0.875\n",
      "Epoch  2,  step= 829,  loss=0.418,  accuracy=0.844\n",
      "Epoch  2,  step= 830,  loss=0.439,  accuracy=0.844\n",
      "Epoch  2,  step= 831,  loss=0.479,  accuracy=0.875\n",
      "Epoch  2,  step= 832,  loss=1.024,  accuracy=0.812\n",
      "Epoch  2,  step= 833,  loss=0.332,  accuracy=0.938\n",
      "Epoch  2,  step= 834,  loss=0.163,  accuracy=0.938\n",
      "Epoch  2,  step= 835,  loss=0.985,  accuracy=0.781\n",
      "Epoch  2,  step= 836,  loss=0.492,  accuracy=0.875\n",
      "Epoch  2,  step= 837,  loss=0.463,  accuracy=0.844\n",
      "Epoch  2,  step= 838,  loss=0.220,  accuracy=0.844\n",
      "Epoch  2,  step= 839,  loss=0.720,  accuracy=0.844\n",
      "Epoch  2,  step= 840,  loss=0.725,  accuracy=0.812\n",
      "Epoch  2,  step= 841,  loss=0.235,  accuracy=0.938\n",
      "Epoch  2,  step= 842,  loss=0.164,  accuracy=0.969\n",
      "Epoch  2,  step= 843,  loss=0.199,  accuracy=0.906\n",
      "Epoch  2,  step= 844,  loss=0.261,  accuracy=0.844\n",
      "Epoch  2,  step= 845,  loss=0.120,  accuracy=0.938\n",
      "Epoch  2,  step= 846,  loss=0.447,  accuracy=0.781\n",
      "Epoch  2,  step= 847,  loss=0.247,  accuracy=0.906\n",
      "Epoch  2,  step= 848,  loss=0.517,  accuracy=0.906\n",
      "Epoch  2,  step= 849,  loss=1.284,  accuracy=0.719\n",
      "Epoch  2,  step= 850,  loss=0.477,  accuracy=0.875\n",
      "Epoch  2,  step= 851,  loss=0.255,  accuracy=0.906\n",
      "Epoch  2,  step= 852,  loss=0.272,  accuracy=0.969\n",
      "Epoch  2,  step= 853,  loss=0.357,  accuracy=0.906\n",
      "Epoch  2,  step= 854,  loss=0.254,  accuracy=0.875\n",
      "Epoch  2,  step= 855,  loss=0.420,  accuracy=0.844\n",
      "Epoch  2,  step= 856,  loss=0.323,  accuracy=0.906\n",
      "Epoch  2,  step= 857,  loss=0.562,  accuracy=0.844\n",
      "Epoch  2,  step= 858,  loss=0.113,  accuracy=0.938\n",
      "Epoch  2,  step= 859,  loss=0.474,  accuracy=0.875\n",
      "Epoch  2,  step= 860,  loss=0.140,  accuracy=0.969\n",
      "Epoch  2,  step= 861,  loss=0.129,  accuracy=0.938\n",
      "Epoch  2,  step= 862,  loss=0.304,  accuracy=0.844\n",
      "Epoch  2,  step= 863,  loss=0.318,  accuracy=0.969\n",
      "Epoch  2,  step= 864,  loss=0.189,  accuracy=0.938\n",
      "Epoch  2,  step= 865,  loss=1.079,  accuracy=0.750\n",
      "Epoch  2,  step= 866,  loss=0.585,  accuracy=0.875\n",
      "Epoch  2,  step= 867,  loss=0.076,  accuracy=1.000\n",
      "Epoch  2,  step= 868,  loss=0.203,  accuracy=0.969\n",
      "Epoch  2,  step= 869,  loss=0.384,  accuracy=0.875\n",
      "Epoch  2,  step= 870,  loss=0.462,  accuracy=0.906\n",
      "Epoch  2,  step= 871,  loss=0.729,  accuracy=0.781\n",
      "Epoch  2,  step= 872,  loss=0.073,  accuracy=1.000\n",
      "Epoch  2,  step= 873,  loss=0.358,  accuracy=0.906\n",
      "Epoch  2,  step= 874,  loss=0.319,  accuracy=0.938\n",
      "Epoch  2,  step= 875,  loss=0.751,  accuracy=0.844\n",
      "Epoch  2,  step= 876,  loss=0.080,  accuracy=0.969\n",
      "Epoch  2,  step= 877,  loss=0.464,  accuracy=0.906\n",
      "Epoch  2,  step= 878,  loss=0.286,  accuracy=0.938\n",
      "Epoch  2,  step= 879,  loss=0.309,  accuracy=0.875\n",
      "Epoch  2,  step= 880,  loss=0.333,  accuracy=0.906\n",
      "Epoch  2,  step= 881,  loss=0.495,  accuracy=0.938\n",
      "Epoch  2,  step= 882,  loss=0.487,  accuracy=0.938\n",
      "Epoch  2,  step= 883,  loss=0.148,  accuracy=0.938\n",
      "Epoch  2,  step= 884,  loss=0.337,  accuracy=0.906\n",
      "Epoch  2,  step= 885,  loss=0.047,  accuracy=0.969\n",
      "Epoch  2,  step= 886,  loss=1.513,  accuracy=0.781\n",
      "Epoch  2,  step= 887,  loss=0.334,  accuracy=0.938\n",
      "Epoch  2,  step= 888,  loss=0.085,  accuracy=0.969\n",
      "Epoch  2,  step= 889,  loss=0.234,  accuracy=0.875\n",
      "Epoch  2,  step= 890,  loss=0.737,  accuracy=0.906\n",
      "Epoch  2,  step= 891,  loss=0.523,  accuracy=0.906\n",
      "Epoch  2,  step= 892,  loss=0.312,  accuracy=0.969\n",
      "Epoch  2,  step= 893,  loss=0.239,  accuracy=0.969\n",
      "Epoch  2,  step= 894,  loss=0.599,  accuracy=0.750\n",
      "Epoch  2,  step= 895,  loss=1.243,  accuracy=0.812\n",
      "Epoch  2,  step= 896,  loss=0.116,  accuracy=0.969\n",
      "Epoch  2,  step= 897,  loss=0.285,  accuracy=0.938\n",
      "Epoch  2,  step= 898,  loss=0.111,  accuracy=0.969\n",
      "Epoch  2,  step= 899,  loss=0.414,  accuracy=0.812\n",
      "Epoch  2,  step= 900,  loss=0.246,  accuracy=0.906\n",
      "Epoch  2,  step= 901,  loss=0.345,  accuracy=0.812\n",
      "Epoch  2,  step= 902,  loss=0.232,  accuracy=0.875\n",
      "Epoch  2,  step= 903,  loss=0.381,  accuracy=0.844\n",
      "Epoch  2,  step= 904,  loss=0.345,  accuracy=0.844\n",
      "Epoch  2,  step= 905,  loss=0.401,  accuracy=0.906\n",
      "Epoch  2,  step= 906,  loss=0.460,  accuracy=0.812\n",
      "Epoch  2,  step= 907,  loss=0.315,  accuracy=0.875\n",
      "Epoch  2,  step= 908,  loss=0.428,  accuracy=0.875\n",
      "Epoch  2,  step= 909,  loss=0.576,  accuracy=0.844\n",
      "Epoch  2,  step= 910,  loss=0.794,  accuracy=0.719\n",
      "Epoch  2,  step= 911,  loss=0.453,  accuracy=0.812\n",
      "Epoch  2,  step= 912,  loss=0.356,  accuracy=0.875\n",
      "Epoch  2,  step= 913,  loss=0.368,  accuracy=0.875\n",
      "Epoch  2,  step= 914,  loss=0.199,  accuracy=0.938\n",
      "Epoch  2,  step= 915,  loss=0.561,  accuracy=0.781\n",
      "Epoch  2,  step= 916,  loss=0.688,  accuracy=0.812\n",
      "Epoch  2,  step= 917,  loss=0.225,  accuracy=0.906\n",
      "Epoch  2,  step= 918,  loss=0.223,  accuracy=0.938\n",
      "Epoch  2,  step= 919,  loss=0.494,  accuracy=0.906\n",
      "Epoch  2,  step= 920,  loss=0.189,  accuracy=0.906\n",
      "Epoch  2,  step= 921,  loss=0.218,  accuracy=0.938\n",
      "Epoch  2,  step= 922,  loss=0.279,  accuracy=0.906\n",
      "Epoch  2,  step= 923,  loss=0.240,  accuracy=0.906\n",
      "Epoch  2,  step= 924,  loss=0.556,  accuracy=0.875\n",
      "Epoch  2,  step= 925,  loss=0.139,  accuracy=0.938\n",
      "Epoch  2,  step= 926,  loss=0.305,  accuracy=0.844\n",
      "Epoch  2,  step= 927,  loss=0.357,  accuracy=0.906\n",
      "Epoch  2,  step= 928,  loss=0.682,  accuracy=0.844\n",
      "Epoch  2,  step= 929,  loss=0.412,  accuracy=0.906\n",
      "Epoch  2,  step= 930,  loss=0.163,  accuracy=0.969\n",
      "Epoch  2,  step= 931,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step= 932,  loss=0.829,  accuracy=0.781\n",
      "Epoch  2,  step= 933,  loss=0.574,  accuracy=0.906\n",
      "Epoch  2,  step= 934,  loss=0.258,  accuracy=0.906\n",
      "Epoch  2,  step= 935,  loss=0.641,  accuracy=0.781\n",
      "Epoch  2,  step= 936,  loss=0.688,  accuracy=0.812\n",
      "Epoch  2,  step= 937,  loss=0.299,  accuracy=0.938\n",
      "Epoch  2,  step= 938,  loss=0.404,  accuracy=0.844\n",
      "Epoch  2,  step= 939,  loss=0.525,  accuracy=0.906\n",
      "Epoch  2,  step= 940,  loss=0.598,  accuracy=0.875\n",
      "Epoch  2,  step= 941,  loss=0.715,  accuracy=0.875\n",
      "Epoch  2,  step= 942,  loss=0.619,  accuracy=0.719\n",
      "Epoch  2,  step= 943,  loss=0.523,  accuracy=0.844\n",
      "Epoch  2,  step= 944,  loss=0.355,  accuracy=0.969\n",
      "Epoch  2,  step= 945,  loss=0.275,  accuracy=0.938\n",
      "Epoch  2,  step= 946,  loss=0.085,  accuracy=0.969\n",
      "Epoch  2,  step= 947,  loss=0.363,  accuracy=0.906\n",
      "Epoch  2,  step= 948,  loss=0.303,  accuracy=0.875\n",
      "Epoch  2,  step= 949,  loss=0.287,  accuracy=0.875\n",
      "Epoch  2,  step= 950,  loss=0.278,  accuracy=0.875\n",
      "Epoch  2,  step= 951,  loss=0.555,  accuracy=0.781\n",
      "Epoch  2,  step= 952,  loss=0.283,  accuracy=0.906\n",
      "Epoch  2,  step= 953,  loss=0.569,  accuracy=0.750\n",
      "Epoch  2,  step= 954,  loss=0.640,  accuracy=0.844\n",
      "Epoch  2,  step= 955,  loss=0.284,  accuracy=0.938\n",
      "Epoch  2,  step= 956,  loss=0.962,  accuracy=0.781\n",
      "Epoch  2,  step= 957,  loss=0.222,  accuracy=0.906\n",
      "Epoch  2,  step= 958,  loss=0.362,  accuracy=0.844\n",
      "Epoch  2,  step= 959,  loss=0.418,  accuracy=0.812\n",
      "Epoch  2,  step= 960,  loss=0.443,  accuracy=0.844\n",
      "Epoch  2,  step= 961,  loss=0.093,  accuracy=0.969\n",
      "Epoch  2,  step= 962,  loss=0.426,  accuracy=0.844\n",
      "Epoch  2,  step= 963,  loss=0.632,  accuracy=0.844\n",
      "Epoch  2,  step= 964,  loss=0.483,  accuracy=0.875\n",
      "Epoch  2,  step= 965,  loss=0.839,  accuracy=0.812\n",
      "Epoch  2,  step= 966,  loss=0.156,  accuracy=0.938\n",
      "Epoch  2,  step= 967,  loss=0.546,  accuracy=0.844\n",
      "Epoch  2,  step= 968,  loss=0.390,  accuracy=0.875\n",
      "Epoch  2,  step= 969,  loss=0.348,  accuracy=0.844\n",
      "Epoch  2,  step= 970,  loss=0.523,  accuracy=0.938\n",
      "Epoch  2,  step= 971,  loss=0.769,  accuracy=0.875\n",
      "Epoch  2,  step= 972,  loss=0.827,  accuracy=0.781\n",
      "Epoch  2,  step= 973,  loss=0.447,  accuracy=0.906\n",
      "Epoch  2,  step= 974,  loss=0.922,  accuracy=0.844\n",
      "Epoch  2,  step= 975,  loss=0.166,  accuracy=0.969\n",
      "Epoch  2,  step= 976,  loss=0.211,  accuracy=0.938\n",
      "Epoch  2,  step= 977,  loss=1.015,  accuracy=0.719\n",
      "Epoch  2,  step= 978,  loss=0.995,  accuracy=0.719\n",
      "Epoch  2,  step= 979,  loss=0.581,  accuracy=0.812\n",
      "Epoch  2,  step= 980,  loss=0.720,  accuracy=0.719\n",
      "Epoch  2,  step= 981,  loss=1.076,  accuracy=0.812\n",
      "Epoch  2,  step= 982,  loss=0.533,  accuracy=0.875\n",
      "Epoch  2,  step= 983,  loss=0.313,  accuracy=0.906\n",
      "Epoch  2,  step= 984,  loss=0.287,  accuracy=0.906\n",
      "Epoch  2,  step= 985,  loss=0.142,  accuracy=0.969\n",
      "Epoch  2,  step= 986,  loss=0.350,  accuracy=0.906\n",
      "Epoch  2,  step= 987,  loss=0.382,  accuracy=0.906\n",
      "Epoch  2,  step= 988,  loss=0.472,  accuracy=0.844\n",
      "Epoch  2,  step= 989,  loss=1.334,  accuracy=0.750\n",
      "Epoch  2,  step= 990,  loss=0.423,  accuracy=0.875\n",
      "Epoch  2,  step= 991,  loss=0.830,  accuracy=0.812\n",
      "Epoch  2,  step= 992,  loss=0.255,  accuracy=0.938\n",
      "Epoch  2,  step= 993,  loss=0.304,  accuracy=0.844\n",
      "Epoch  2,  step= 994,  loss=0.143,  accuracy=0.938\n",
      "Epoch  2,  step= 995,  loss=0.589,  accuracy=0.844\n",
      "Epoch  2,  step= 996,  loss=0.207,  accuracy=0.938\n",
      "Epoch  2,  step= 997,  loss=0.339,  accuracy=0.906\n",
      "Epoch  2,  step= 998,  loss=0.545,  accuracy=0.875\n",
      "Epoch  2,  step= 999,  loss=0.214,  accuracy=0.969\n",
      "Epoch  2,  step=1000,  loss=0.371,  accuracy=0.875\n",
      "Epoch  2,  step=1001,  loss=0.343,  accuracy=0.812\n",
      "Epoch  2,  step=1002,  loss=0.329,  accuracy=0.906\n",
      "Epoch  2,  step=1003,  loss=0.167,  accuracy=0.969\n",
      "Epoch  2,  step=1004,  loss=0.408,  accuracy=0.875\n",
      "Epoch  2,  step=1005,  loss=0.621,  accuracy=0.875\n",
      "Epoch  2,  step=1006,  loss=0.378,  accuracy=0.906\n",
      "Epoch  2,  step=1007,  loss=0.497,  accuracy=0.844\n",
      "Epoch  2,  step=1008,  loss=0.414,  accuracy=0.938\n",
      "Epoch  2,  step=1009,  loss=0.453,  accuracy=0.875\n",
      "Epoch  2,  step=1010,  loss=0.742,  accuracy=0.656\n",
      "Epoch  2,  step=1011,  loss=0.683,  accuracy=0.875\n",
      "Epoch  2,  step=1012,  loss=0.350,  accuracy=0.844\n",
      "Epoch  2,  step=1013,  loss=0.521,  accuracy=0.781\n",
      "Epoch  2,  step=1014,  loss=0.482,  accuracy=0.875\n",
      "Epoch  2,  step=1015,  loss=0.318,  accuracy=0.875\n",
      "Epoch  2,  step=1016,  loss=0.183,  accuracy=0.938\n",
      "Epoch  2,  step=1017,  loss=0.459,  accuracy=0.875\n",
      "Epoch  2,  step=1018,  loss=0.232,  accuracy=0.938\n",
      "Epoch  2,  step=1019,  loss=0.423,  accuracy=0.844\n",
      "Epoch  2,  step=1020,  loss=0.451,  accuracy=0.875\n",
      "Epoch  2,  step=1021,  loss=0.288,  accuracy=0.938\n",
      "Epoch  2,  step=1022,  loss=0.223,  accuracy=0.938\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  2,  step=1023,  loss=0.140,  accuracy=0.906\n",
      "Epoch  2,  step=1024,  loss=0.710,  accuracy=0.875\n",
      "Epoch  2,  step=1025,  loss=0.511,  accuracy=0.875\n",
      "Epoch  2,  step=1026,  loss=0.488,  accuracy=0.812\n",
      "Epoch  2,  step=1027,  loss=0.583,  accuracy=0.875\n",
      "Epoch  2,  step=1028,  loss=0.566,  accuracy=0.875\n",
      "Epoch  2,  step=1029,  loss=0.249,  accuracy=0.969\n",
      "Epoch  2,  step=1030,  loss=0.086,  accuracy=0.969\n",
      "Epoch  2,  step=1031,  loss=0.092,  accuracy=0.969\n",
      "Epoch  2,  step=1032,  loss=0.410,  accuracy=0.875\n",
      "Epoch  2,  step=1033,  loss=0.572,  accuracy=0.938\n",
      "Epoch  2,  step=1034,  loss=0.156,  accuracy=0.938\n",
      "Epoch  2,  step=1035,  loss=0.339,  accuracy=0.906\n",
      "Epoch  2,  step=1036,  loss=0.281,  accuracy=0.938\n",
      "Epoch  2,  step=1037,  loss=0.334,  accuracy=0.875\n",
      "Epoch  2,  step=1038,  loss=0.522,  accuracy=0.906\n",
      "Epoch  2,  step=1039,  loss=0.206,  accuracy=0.906\n",
      "Epoch  2,  step=1040,  loss=0.551,  accuracy=0.844\n",
      "Epoch  2,  step=1041,  loss=0.197,  accuracy=0.969\n",
      "Epoch  2,  step=1042,  loss=0.320,  accuracy=0.844\n",
      "Epoch  2,  step=1043,  loss=0.248,  accuracy=0.938\n",
      "Epoch  2,  step=1044,  loss=0.321,  accuracy=0.875\n",
      "Epoch  2,  step=1045,  loss=0.361,  accuracy=0.812\n",
      "Epoch  2,  step=1046,  loss=0.303,  accuracy=0.875\n",
      "Epoch  2,  step=1047,  loss=0.360,  accuracy=0.875\n",
      "Epoch  2,  step=1048,  loss=0.061,  accuracy=1.000\n",
      "Epoch  2,  step=1049,  loss=0.789,  accuracy=0.812\n",
      "Epoch  2,  step=1050,  loss=0.201,  accuracy=0.906\n",
      "Epoch  2,  step=1051,  loss=0.274,  accuracy=0.906\n",
      "Epoch  2,  step=1052,  loss=0.367,  accuracy=0.906\n",
      "Epoch  2,  step=1053,  loss=0.114,  accuracy=0.938\n",
      "Epoch  2,  step=1054,  loss=0.291,  accuracy=0.969\n",
      "Epoch  2,  step=1055,  loss=0.852,  accuracy=0.750\n",
      "Epoch  2,  step=1056,  loss=0.357,  accuracy=0.875\n",
      "Epoch  2,  step=1057,  loss=0.165,  accuracy=0.969\n",
      "Epoch  2,  step=1058,  loss=0.076,  accuracy=0.969\n",
      "Epoch  2,  step=1059,  loss=0.171,  accuracy=0.938\n",
      "Epoch  2,  step=1060,  loss=0.206,  accuracy=0.906\n",
      "Epoch  2,  step=1061,  loss=0.126,  accuracy=0.969\n",
      "Epoch  2,  step=1062,  loss=0.356,  accuracy=0.938\n",
      "Epoch  2,  step=1063,  loss=0.212,  accuracy=0.938\n",
      "Epoch  2,  step=1064,  loss=0.499,  accuracy=0.906\n",
      "Epoch  2,  step=1065,  loss=0.524,  accuracy=0.781\n",
      "Epoch  2,  step=1066,  loss=0.542,  accuracy=0.906\n",
      "Epoch  2,  step=1067,  loss=0.092,  accuracy=0.969\n",
      "Epoch  2,  step=1068,  loss=0.364,  accuracy=0.938\n",
      "Epoch  2,  step=1069,  loss=0.146,  accuracy=0.938\n",
      "Epoch  2,  step=1070,  loss=0.152,  accuracy=0.969\n",
      "Epoch  2,  step=1071,  loss=0.089,  accuracy=0.938\n",
      "Epoch  2,  step=1072,  loss=0.297,  accuracy=0.906\n",
      "Epoch  2,  step=1073,  loss=0.334,  accuracy=0.875\n",
      "Epoch  2,  step=1074,  loss=0.085,  accuracy=0.969\n",
      "Epoch  2,  step=1075,  loss=0.510,  accuracy=0.812\n",
      "Epoch  2,  step=1076,  loss=0.574,  accuracy=0.844\n",
      "Epoch  2,  step=1077,  loss=0.335,  accuracy=0.875\n",
      "Epoch  2,  step=1078,  loss=0.596,  accuracy=0.844\n",
      "Epoch  2,  step=1079,  loss=1.141,  accuracy=0.750\n",
      "Epoch  2,  step=1080,  loss=0.182,  accuracy=0.938\n",
      "Epoch  2,  step=1081,  loss=0.367,  accuracy=0.844\n",
      "Epoch  2,  step=1082,  loss=0.313,  accuracy=0.875\n",
      "Epoch  2,  step=1083,  loss=0.575,  accuracy=0.844\n",
      "Epoch  2,  step=1084,  loss=0.520,  accuracy=0.844\n",
      "Epoch  2,  step=1085,  loss=0.344,  accuracy=0.906\n",
      "Epoch  2,  step=1086,  loss=0.250,  accuracy=0.938\n",
      "Epoch  2,  step=1087,  loss=0.425,  accuracy=0.875\n",
      "Epoch  2,  step=1088,  loss=0.700,  accuracy=0.750\n",
      "Epoch  2,  step=1089,  loss=0.334,  accuracy=0.875\n",
      "Epoch  2,  step=1090,  loss=0.416,  accuracy=0.906\n",
      "Epoch  2,  step=1091,  loss=0.265,  accuracy=0.875\n",
      "Epoch  2,  step=1092,  loss=0.161,  accuracy=0.906\n",
      "Epoch  2,  step=1093,  loss=0.145,  accuracy=0.938\n",
      "Epoch  2,  step=1094,  loss=0.136,  accuracy=0.938\n",
      "Epoch  2,  step=1095,  loss=0.391,  accuracy=0.812\n",
      "Epoch  2,  step=1096,  loss=0.292,  accuracy=0.938\n",
      "Epoch  2,  step=1097,  loss=0.377,  accuracy=0.906\n",
      "Epoch  2,  step=1098,  loss=0.578,  accuracy=0.781\n",
      "Epoch  2,  step=1099,  loss=0.298,  accuracy=0.906\n",
      "Epoch  2,  step=1100,  loss=0.393,  accuracy=0.906\n",
      "Epoch  2,  step=1101,  loss=0.647,  accuracy=0.750\n",
      "Epoch  2,  step=1102,  loss=0.097,  accuracy=1.000\n",
      "Epoch  2,  step=1103,  loss=0.222,  accuracy=0.875\n",
      "Epoch  2,  step=1104,  loss=0.110,  accuracy=0.938\n",
      "Epoch  2,  step=1105,  loss=0.293,  accuracy=0.969\n",
      "Epoch  2,  step=1106,  loss=0.226,  accuracy=0.938\n",
      "Epoch  2,  step=1107,  loss=0.426,  accuracy=0.844\n",
      "Epoch  2,  step=1108,  loss=1.158,  accuracy=0.781\n",
      "Epoch  2,  step=1109,  loss=0.322,  accuracy=0.875\n",
      "Epoch  2,  step=1110,  loss=0.236,  accuracy=0.906\n",
      "Epoch  2,  step=1111,  loss=0.369,  accuracy=0.875\n",
      "Epoch  2,  step=1112,  loss=0.132,  accuracy=0.969\n",
      "Epoch  2,  step=1113,  loss=0.907,  accuracy=0.719\n",
      "Epoch  2,  step=1114,  loss=0.117,  accuracy=0.969\n",
      "Epoch  2,  step=1115,  loss=0.397,  accuracy=0.938\n",
      "Epoch  2,  step=1116,  loss=0.124,  accuracy=0.969\n",
      "Epoch  2,  step=1117,  loss=0.162,  accuracy=0.938\n",
      "Epoch  2,  step=1118,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step=1119,  loss=0.209,  accuracy=0.938\n",
      "Epoch  2,  step=1120,  loss=0.475,  accuracy=0.875\n",
      "Epoch  2,  step=1121,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step=1122,  loss=0.315,  accuracy=0.906\n",
      "Epoch  2,  step=1123,  loss=0.314,  accuracy=0.875\n",
      "Epoch  2,  step=1124,  loss=0.546,  accuracy=0.875\n",
      "Epoch  2,  step=1125,  loss=0.544,  accuracy=0.812\n",
      "Epoch  2,  step=1126,  loss=0.678,  accuracy=0.875\n",
      "Epoch  2,  step=1127,  loss=0.521,  accuracy=0.844\n",
      "Epoch  2,  step=1128,  loss=0.770,  accuracy=0.844\n",
      "Epoch  2,  step=1129,  loss=0.198,  accuracy=0.938\n",
      "Epoch  2,  step=1130,  loss=0.405,  accuracy=0.906\n",
      "Epoch  2,  step=1131,  loss=0.197,  accuracy=0.875\n",
      "Epoch  2,  step=1132,  loss=0.338,  accuracy=0.875\n",
      "Epoch  2,  step=1133,  loss=0.314,  accuracy=0.906\n",
      "Epoch  2,  step=1134,  loss=0.034,  accuracy=1.000\n",
      "Epoch  2,  step=1135,  loss=0.288,  accuracy=0.812\n",
      "Epoch  2,  step=1136,  loss=0.099,  accuracy=1.000\n",
      "Epoch  2,  step=1137,  loss=0.275,  accuracy=0.844\n",
      "Epoch  2,  step=1138,  loss=0.679,  accuracy=0.844\n",
      "Epoch  2,  step=1139,  loss=0.576,  accuracy=0.812\n",
      "Epoch  2,  step=1140,  loss=0.398,  accuracy=0.875\n",
      "Epoch  2,  step=1141,  loss=0.360,  accuracy=0.844\n",
      "Epoch  2,  step=1142,  loss=0.172,  accuracy=0.938\n",
      "Epoch  2,  step=1143,  loss=0.252,  accuracy=0.938\n",
      "Epoch  2,  step=1144,  loss=0.049,  accuracy=1.000\n",
      "Epoch  2,  step=1145,  loss=0.491,  accuracy=0.812\n",
      "Epoch  2,  step=1146,  loss=0.098,  accuracy=0.969\n",
      "Epoch  2,  step=1147,  loss=0.349,  accuracy=0.938\n",
      "Epoch  2,  step=1148,  loss=0.261,  accuracy=0.875\n",
      "Epoch  2,  step=1149,  loss=0.125,  accuracy=0.969\n",
      "Epoch  2,  step=1150,  loss=0.457,  accuracy=0.906\n",
      "Epoch  2,  step=1151,  loss=0.179,  accuracy=0.938\n",
      "Epoch  2,  step=1152,  loss=0.315,  accuracy=0.906\n",
      "Epoch  2,  step=1153,  loss=0.106,  accuracy=1.000\n",
      "Epoch  2,  step=1154,  loss=0.153,  accuracy=0.938\n",
      "Epoch  2,  step=1155,  loss=0.133,  accuracy=0.938\n",
      "Epoch  2,  step=1156,  loss=0.095,  accuracy=0.969\n",
      "Epoch  2,  step=1157,  loss=0.512,  accuracy=0.875\n",
      "Epoch  2,  step=1158,  loss=0.457,  accuracy=0.906\n",
      "Epoch  2,  step=1159,  loss=0.461,  accuracy=0.906\n",
      "Epoch  2,  step=1160,  loss=0.371,  accuracy=0.906\n",
      "Epoch  2,  step=1161,  loss=0.380,  accuracy=0.906\n",
      "Epoch  2,  step=1162,  loss=0.359,  accuracy=0.844\n",
      "Epoch  2,  step=1163,  loss=0.105,  accuracy=0.969\n",
      "Epoch  2,  step=1164,  loss=0.816,  accuracy=0.812\n",
      "Epoch  2,  step=1165,  loss=0.420,  accuracy=0.875\n",
      "Epoch  2,  step=1166,  loss=0.357,  accuracy=0.844\n",
      "Epoch  2,  step=1167,  loss=0.748,  accuracy=0.875\n",
      "Epoch  2,  step=1168,  loss=1.190,  accuracy=0.844\n",
      "Epoch  2,  step=1169,  loss=0.827,  accuracy=0.844\n",
      "Epoch  2,  step=1170,  loss=0.739,  accuracy=0.844\n",
      "Epoch  2,  step=1171,  loss=0.278,  accuracy=0.938\n",
      "Epoch  2,  step=1172,  loss=0.354,  accuracy=0.875\n",
      "Epoch  2,  step=1173,  loss=0.701,  accuracy=0.781\n",
      "Epoch  2,  step=1174,  loss=0.596,  accuracy=0.719\n",
      "Epoch  2,  step=1175,  loss=0.495,  accuracy=0.844\n",
      "Epoch  2,  step=1176,  loss=0.313,  accuracy=0.875\n",
      "Epoch  2,  step=1177,  loss=0.250,  accuracy=0.906\n",
      "Epoch  2,  step=1178,  loss=0.610,  accuracy=0.875\n",
      "Epoch  2,  step=1179,  loss=0.712,  accuracy=0.875\n",
      "Epoch  2,  step=1180,  loss=0.369,  accuracy=0.906\n",
      "Epoch  2,  step=1181,  loss=0.704,  accuracy=0.875\n",
      "Epoch  2,  step=1182,  loss=0.622,  accuracy=0.844\n",
      "Epoch  2,  step=1183,  loss=0.232,  accuracy=0.906\n",
      "Epoch  2,  step=1184,  loss=0.328,  accuracy=0.875\n",
      "Epoch  2,  step=1185,  loss=0.487,  accuracy=0.906\n",
      "Epoch  2,  step=1186,  loss=0.231,  accuracy=0.969\n",
      "Epoch  2,  step=1187,  loss=0.370,  accuracy=0.906\n",
      "Epoch  2,  step=1188,  loss=0.368,  accuracy=0.969\n",
      "Epoch  2,  step=1189,  loss=0.433,  accuracy=0.812\n",
      "Epoch  2,  step=1190,  loss=0.305,  accuracy=0.938\n",
      "Epoch  2,  step=1191,  loss=0.342,  accuracy=0.906\n",
      "Epoch  2,  step=1192,  loss=0.276,  accuracy=0.844\n",
      "Epoch  2,  step=1193,  loss=0.152,  accuracy=0.969\n",
      "Epoch  2,  step=1194,  loss=0.379,  accuracy=0.906\n",
      "Epoch  2,  step=1195,  loss=0.663,  accuracy=0.781\n",
      "Epoch  2,  step=1196,  loss=0.175,  accuracy=0.938\n",
      "Epoch  2,  step=1197,  loss=0.248,  accuracy=0.812\n",
      "Epoch  2,  step=1198,  loss=0.122,  accuracy=0.969\n",
      "Epoch  2,  step=1199,  loss=0.890,  accuracy=0.750\n",
      "Epoch  2,  step=1200,  loss=0.217,  accuracy=0.938\n",
      "Epoch  2,  step=1201,  loss=0.370,  accuracy=0.844\n",
      "Epoch  2,  step=1202,  loss=0.077,  accuracy=0.969\n",
      "Epoch  2,  step=1203,  loss=0.881,  accuracy=0.844\n",
      "Epoch  2,  step=1204,  loss=0.201,  accuracy=0.938\n",
      "Epoch  2,  step=1205,  loss=0.433,  accuracy=0.875\n",
      "Epoch  2,  step=1206,  loss=0.623,  accuracy=0.844\n",
      "Epoch  2,  step=1207,  loss=0.249,  accuracy=0.906\n",
      "Epoch  2,  step=1208,  loss=0.373,  accuracy=0.906\n",
      "Epoch  2,  step=1209,  loss=0.212,  accuracy=0.938\n",
      "Epoch  2,  step=1210,  loss=0.176,  accuracy=0.938\n",
      "Epoch  2,  step=1211,  loss=0.553,  accuracy=0.844\n",
      "Epoch  2,  step=1212,  loss=0.348,  accuracy=0.875\n",
      "Epoch  2,  step=1213,  loss=0.064,  accuracy=0.969\n",
      "Epoch  2,  step=1214,  loss=0.063,  accuracy=1.000\n",
      "Epoch  2,  step=1215,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step=1216,  loss=0.209,  accuracy=0.906\n",
      "Epoch  2,  step=1217,  loss=0.171,  accuracy=0.969\n",
      "Epoch  2,  step=1218,  loss=0.357,  accuracy=0.906\n",
      "Epoch  2,  step=1219,  loss=0.147,  accuracy=0.938\n",
      "Epoch  2,  step=1220,  loss=0.317,  accuracy=0.906\n",
      "Epoch  2,  step=1221,  loss=0.107,  accuracy=0.938\n",
      "Epoch  2,  step=1222,  loss=0.663,  accuracy=0.875\n",
      "Epoch  2,  step=1223,  loss=0.427,  accuracy=0.875\n",
      "Epoch  2,  step=1224,  loss=0.287,  accuracy=0.969\n",
      "Epoch  2,  step=1225,  loss=0.506,  accuracy=0.938\n",
      "Epoch  2,  step=1226,  loss=0.526,  accuracy=0.844\n",
      "Epoch  2,  step=1227,  loss=0.339,  accuracy=0.938\n",
      "Epoch  2,  step=1228,  loss=1.301,  accuracy=0.781\n",
      "Epoch  2,  step=1229,  loss=0.743,  accuracy=0.781\n",
      "Epoch  2,  step=1230,  loss=0.304,  accuracy=0.844\n",
      "Epoch  2,  step=1231,  loss=0.851,  accuracy=0.844\n",
      "Epoch  2,  step=1232,  loss=0.398,  accuracy=0.844\n",
      "Epoch  2,  step=1233,  loss=0.472,  accuracy=0.938\n",
      "Epoch  2,  step=1234,  loss=0.224,  accuracy=0.875\n",
      "Epoch  2,  step=1235,  loss=0.436,  accuracy=0.906\n",
      "Epoch  2,  step=1236,  loss=0.348,  accuracy=0.938\n",
      "Epoch  2,  step=1237,  loss=0.617,  accuracy=0.844\n",
      "Epoch  2,  step=1238,  loss=0.361,  accuracy=0.875\n",
      "Epoch  2,  step=1239,  loss=0.809,  accuracy=0.844\n",
      "Epoch  2,  step=1240,  loss=0.739,  accuracy=0.844\n",
      "Epoch  2,  step=1241,  loss=0.243,  accuracy=0.906\n",
      "Epoch  2,  step=1242,  loss=0.670,  accuracy=0.812\n",
      "Epoch  2,  step=1243,  loss=0.668,  accuracy=0.781\n",
      "Epoch  2,  step=1244,  loss=0.301,  accuracy=0.938\n",
      "Epoch  2,  step=1245,  loss=0.474,  accuracy=0.906\n",
      "Epoch  2,  step=1246,  loss=0.434,  accuracy=0.844\n",
      "Epoch  2,  step=1247,  loss=0.551,  accuracy=0.875\n",
      "Epoch  2,  step=1248,  loss=0.452,  accuracy=0.875\n",
      "Epoch  2,  step=1249,  loss=0.443,  accuracy=0.875\n",
      "Epoch  2,  step=1250,  loss=0.338,  accuracy=0.875\n",
      "Epoch  2,  step=1251,  loss=0.328,  accuracy=0.844\n",
      "Epoch  2,  step=1252,  loss=0.316,  accuracy=0.844\n",
      "Epoch  2,  step=1253,  loss=0.349,  accuracy=0.875\n",
      "Epoch  2,  step=1254,  loss=0.352,  accuracy=0.875\n",
      "Epoch  2,  step=1255,  loss=0.340,  accuracy=0.906\n",
      "Epoch  2,  step=1256,  loss=0.313,  accuracy=0.906\n",
      "Epoch  2,  step=1257,  loss=0.509,  accuracy=0.875\n",
      "Epoch  2,  step=1258,  loss=0.170,  accuracy=0.906\n",
      "Epoch  2,  step=1259,  loss=0.459,  accuracy=0.875\n",
      "Epoch  2,  step=1260,  loss=0.230,  accuracy=0.875\n",
      "Epoch  2,  step=1261,  loss=0.071,  accuracy=1.000\n",
      "Epoch  2,  step=1262,  loss=0.242,  accuracy=0.938\n",
      "Epoch  2,  step=1263,  loss=0.245,  accuracy=0.844\n",
      "Epoch  2,  step=1264,  loss=0.097,  accuracy=0.969\n",
      "Epoch  2,  step=1265,  loss=0.615,  accuracy=0.875\n",
      "Epoch  2,  step=1266,  loss=0.355,  accuracy=0.875\n",
      "Epoch  2,  step=1267,  loss=0.289,  accuracy=0.906\n",
      "Epoch  2,  step=1268,  loss=0.288,  accuracy=0.812\n",
      "Epoch  2,  step=1269,  loss=0.252,  accuracy=0.938\n",
      "Epoch  2,  step=1270,  loss=0.723,  accuracy=0.781\n",
      "Epoch  2,  step=1271,  loss=0.118,  accuracy=1.000\n",
      "Epoch  2,  step=1272,  loss=0.269,  accuracy=0.938\n",
      "Epoch  2,  step=1273,  loss=0.248,  accuracy=0.906\n",
      "Epoch  2,  step=1274,  loss=0.106,  accuracy=0.969\n",
      "Epoch  2,  step=1275,  loss=0.115,  accuracy=0.969\n",
      "Epoch  2,  step=1276,  loss=0.155,  accuracy=0.938\n",
      "Epoch  2,  step=1277,  loss=0.240,  accuracy=0.969\n",
      "Epoch  2,  step=1278,  loss=0.084,  accuracy=0.969\n",
      "Epoch  2,  step=1279,  loss=0.187,  accuracy=0.969\n",
      "Epoch  2,  step=1280,  loss=0.507,  accuracy=0.906\n",
      "Epoch  2,  step=1281,  loss=0.680,  accuracy=0.781\n",
      "Epoch  2,  step=1282,  loss=0.214,  accuracy=0.969\n",
      "Epoch  2,  step=1283,  loss=0.234,  accuracy=0.938\n",
      "Epoch  2,  step=1284,  loss=0.267,  accuracy=0.844\n",
      "Epoch  2,  step=1285,  loss=0.188,  accuracy=0.969\n",
      "Epoch  2,  step=1286,  loss=0.328,  accuracy=0.844\n",
      "Epoch  2,  step=1287,  loss=0.230,  accuracy=0.906\n",
      "Epoch  2,  step=1288,  loss=0.263,  accuracy=0.844\n",
      "Epoch  2,  step=1289,  loss=0.825,  accuracy=0.812\n",
      "Epoch  2,  step=1290,  loss=0.705,  accuracy=0.844\n",
      "Epoch  2,  step=1291,  loss=0.498,  accuracy=0.906\n",
      "Epoch  2,  step=1292,  loss=0.304,  accuracy=0.906\n",
      "Epoch  2,  step=1293,  loss=0.344,  accuracy=0.906\n",
      "Epoch  2,  step=1294,  loss=0.458,  accuracy=0.875\n",
      "Epoch  2,  step=1295,  loss=0.543,  accuracy=0.875\n",
      "Epoch  2,  step=1296,  loss=0.310,  accuracy=0.875\n",
      "Epoch  2,  step=1297,  loss=0.328,  accuracy=0.906\n",
      "Epoch  2,  step=1298,  loss=0.502,  accuracy=0.875\n",
      "Epoch  2,  step=1299,  loss=0.624,  accuracy=0.812\n",
      "Epoch  2,  step=1300,  loss=0.379,  accuracy=0.844\n",
      "Epoch  2,  step=1301,  loss=0.050,  accuracy=1.000\n",
      "Epoch  2,  step=1302,  loss=0.177,  accuracy=0.969\n",
      "Epoch  2,  step=1303,  loss=0.292,  accuracy=0.906\n",
      "Epoch  2,  step=1304,  loss=0.544,  accuracy=0.875\n",
      "Epoch  2,  step=1305,  loss=0.387,  accuracy=0.875\n",
      "Epoch  2,  step=1306,  loss=0.479,  accuracy=0.844\n",
      "Epoch  2,  step=1307,  loss=0.278,  accuracy=0.906\n",
      "Epoch  2,  step=1308,  loss=0.428,  accuracy=0.875\n",
      "Epoch  2,  step=1309,  loss=0.497,  accuracy=0.844\n",
      "Epoch  2,  step=1310,  loss=0.529,  accuracy=0.906\n",
      "Epoch  2,  step=1311,  loss=0.544,  accuracy=0.875\n",
      "Epoch  2,  step=1312,  loss=0.584,  accuracy=0.875\n",
      "Epoch  2,  step=1313,  loss=0.500,  accuracy=0.812\n",
      "Epoch  2,  step=1314,  loss=0.417,  accuracy=0.938\n",
      "Epoch  2,  step=1315,  loss=0.327,  accuracy=0.938\n",
      "Epoch  2,  step=1316,  loss=0.347,  accuracy=0.844\n",
      "Epoch  2,  step=1317,  loss=0.172,  accuracy=0.938\n",
      "Epoch  2,  step=1318,  loss=0.406,  accuracy=0.875\n",
      "Epoch  2,  step=1319,  loss=0.285,  accuracy=0.875\n",
      "Epoch  2,  step=1320,  loss=0.330,  accuracy=0.875\n",
      "Epoch  2,  step=1321,  loss=0.094,  accuracy=0.938\n",
      "Epoch  2,  step=1322,  loss=0.877,  accuracy=0.844\n",
      "Epoch  2,  step=1323,  loss=0.841,  accuracy=0.812\n",
      "Epoch  2,  step=1324,  loss=0.268,  accuracy=0.938\n",
      "Epoch  2,  step=1325,  loss=0.586,  accuracy=0.906\n",
      "Epoch  2,  step=1326,  loss=0.381,  accuracy=0.875\n",
      "Epoch  2,  step=1327,  loss=0.365,  accuracy=0.844\n",
      "Epoch  2,  step=1328,  loss=0.841,  accuracy=0.812\n",
      "Epoch  2,  step=1329,  loss=0.259,  accuracy=0.906\n",
      "Epoch  2,  step=1330,  loss=0.230,  accuracy=0.906\n",
      "Epoch  2,  step=1331,  loss=0.219,  accuracy=0.844\n",
      "Epoch  2,  step=1332,  loss=0.351,  accuracy=0.875\n",
      "Epoch  2,  step=1333,  loss=0.488,  accuracy=0.812\n",
      "Epoch  2,  step=1334,  loss=0.584,  accuracy=0.844\n",
      "Epoch  2,  step=1335,  loss=0.134,  accuracy=0.969\n",
      "Epoch  2,  step=1336,  loss=0.513,  accuracy=0.906\n",
      "Epoch  2,  step=1337,  loss=0.511,  accuracy=0.875\n",
      "Epoch  2,  step=1338,  loss=0.726,  accuracy=0.906\n",
      "Epoch  2,  step=1339,  loss=0.804,  accuracy=0.875\n",
      "Epoch  2,  step=1340,  loss=0.450,  accuracy=0.875\n",
      "Epoch  2,  step=1341,  loss=0.427,  accuracy=0.938\n",
      "Epoch  2,  step=1342,  loss=0.424,  accuracy=0.875\n",
      "Epoch  2,  step=1343,  loss=0.247,  accuracy=0.906\n",
      "Epoch  2,  step=1344,  loss=0.748,  accuracy=0.844\n",
      "Epoch  2,  step=1345,  loss=0.293,  accuracy=0.906\n",
      "Epoch  2,  step=1346,  loss=0.589,  accuracy=0.906\n",
      "Epoch  2,  step=1347,  loss=0.122,  accuracy=0.969\n",
      "Epoch  2,  step=1348,  loss=0.098,  accuracy=0.969\n",
      "Epoch  2,  step=1349,  loss=0.417,  accuracy=0.875\n",
      "Epoch  2,  step=1350,  loss=0.917,  accuracy=0.719\n",
      "Epoch  2,  step=1351,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step=1352,  loss=0.235,  accuracy=0.938\n",
      "Epoch  2,  step=1353,  loss=0.045,  accuracy=0.969\n",
      "Epoch  2,  step=1354,  loss=0.071,  accuracy=0.969\n",
      "Epoch  2,  step=1355,  loss=0.282,  accuracy=0.938\n",
      "Epoch  2,  step=1356,  loss=0.148,  accuracy=0.938\n",
      "Epoch  2,  step=1357,  loss=0.471,  accuracy=0.938\n",
      "Epoch  2,  step=1358,  loss=0.313,  accuracy=0.906\n",
      "Epoch  2,  step=1359,  loss=0.186,  accuracy=0.906\n",
      "Epoch  2,  step=1360,  loss=0.344,  accuracy=0.875\n",
      "Epoch  2,  step=1361,  loss=0.247,  accuracy=0.906\n",
      "Epoch  2,  step=1362,  loss=0.272,  accuracy=0.906\n",
      "Epoch  2,  step=1363,  loss=0.293,  accuracy=0.938\n",
      "Epoch  2,  step=1364,  loss=0.609,  accuracy=0.844\n",
      "Epoch  2,  step=1365,  loss=0.271,  accuracy=0.906\n",
      "Epoch  2,  step=1366,  loss=0.058,  accuracy=1.000\n",
      "Epoch  2,  step=1367,  loss=0.174,  accuracy=0.969\n",
      "Epoch  2,  step=1368,  loss=0.335,  accuracy=0.906\n",
      "Epoch  2,  step=1369,  loss=0.555,  accuracy=0.781\n",
      "Epoch  2,  step=1370,  loss=0.294,  accuracy=0.875\n",
      "Epoch  2,  step=1371,  loss=0.345,  accuracy=0.906\n",
      "Epoch  2,  step=1372,  loss=0.139,  accuracy=0.938\n",
      "Epoch  2,  step=1373,  loss=0.691,  accuracy=0.812\n",
      "Epoch  2,  step=1374,  loss=0.414,  accuracy=0.906\n",
      "Epoch  2,  step=1375,  loss=0.075,  accuracy=1.000\n",
      "Epoch  2,  step=1376,  loss=0.385,  accuracy=0.906\n",
      "Epoch  2,  step=1377,  loss=0.339,  accuracy=0.844\n",
      "Epoch  2,  step=1378,  loss=0.393,  accuracy=0.844\n",
      "Epoch  2,  step=1379,  loss=0.586,  accuracy=0.844\n",
      "Epoch  2,  step=1380,  loss=0.258,  accuracy=0.906\n",
      "Epoch  2,  step=1381,  loss=0.235,  accuracy=0.938\n",
      "Epoch  2,  step=1382,  loss=0.481,  accuracy=0.812\n",
      "Epoch  2,  step=1383,  loss=0.726,  accuracy=0.844\n",
      "Epoch  2,  step=1384,  loss=0.354,  accuracy=0.844\n",
      "Epoch  2,  step=1385,  loss=0.297,  accuracy=0.906\n",
      "Epoch  2,  step=1386,  loss=0.645,  accuracy=0.938\n",
      "Epoch  2,  step=1387,  loss=0.140,  accuracy=0.969\n",
      "Epoch  2,  step=1388,  loss=0.426,  accuracy=0.938\n",
      "Epoch  2,  step=1389,  loss=0.211,  accuracy=0.938\n",
      "Epoch  2,  step=1390,  loss=0.477,  accuracy=0.906\n",
      "Epoch  2,  step=1391,  loss=0.145,  accuracy=0.969\n",
      "Epoch  2,  step=1392,  loss=0.074,  accuracy=1.000\n",
      "Epoch  2,  step=1393,  loss=0.068,  accuracy=0.969\n",
      "Epoch  2,  step=1394,  loss=0.127,  accuracy=0.969\n",
      "Epoch  2,  step=1395,  loss=0.385,  accuracy=0.875\n",
      "Epoch  2,  step=1396,  loss=0.397,  accuracy=0.906\n",
      "Epoch  2,  step=1397,  loss=0.204,  accuracy=0.906\n",
      "Epoch  2,  step=1398,  loss=0.711,  accuracy=0.781\n",
      "Epoch  2,  step=1399,  loss=0.111,  accuracy=0.969\n",
      "Epoch  2,  step=1400,  loss=0.328,  accuracy=0.906\n",
      "Epoch  2,  step=1401,  loss=0.481,  accuracy=0.812\n",
      "Epoch  2,  step=1402,  loss=0.792,  accuracy=0.875\n",
      "Epoch  2,  step=1403,  loss=0.496,  accuracy=0.875\n",
      "Epoch  2,  step=1404,  loss=0.242,  accuracy=0.906\n",
      "Epoch  2,  step=1405,  loss=0.243,  accuracy=0.938\n",
      "Epoch  2,  step=1406,  loss=0.412,  accuracy=0.938\n",
      "Epoch  2,  step=1407,  loss=0.939,  accuracy=0.812\n",
      "Epoch  2,  step=1408,  loss=0.245,  accuracy=0.906\n",
      "Epoch  2,  step=1409,  loss=0.778,  accuracy=0.812\n",
      "Epoch  2,  step=1410,  loss=0.308,  accuracy=0.938\n",
      "Epoch  2,  step=1411,  loss=0.464,  accuracy=0.875\n",
      "Epoch  2,  step=1412,  loss=0.374,  accuracy=0.906\n",
      "Epoch  2,  step=1413,  loss=0.273,  accuracy=0.906\n",
      "Epoch  2,  step=1414,  loss=0.199,  accuracy=0.906\n",
      "Epoch  2,  step=1415,  loss=0.104,  accuracy=0.969\n",
      "Epoch  2,  step=1416,  loss=0.233,  accuracy=0.906\n",
      "Epoch  2,  step=1417,  loss=0.145,  accuracy=0.938\n",
      "Epoch  2,  step=1418,  loss=0.117,  accuracy=0.969\n",
      "Epoch  2,  step=1419,  loss=0.618,  accuracy=0.844\n",
      "Epoch  2,  step=1420,  loss=0.693,  accuracy=0.812\n",
      "Epoch  2,  step=1421,  loss=0.984,  accuracy=0.844\n",
      "Epoch  2,  step=1422,  loss=0.539,  accuracy=0.844\n",
      "Epoch  2,  step=1423,  loss=0.190,  accuracy=0.938\n",
      "Epoch  2,  step=1424,  loss=0.315,  accuracy=0.875\n",
      "Epoch  2,  step=1425,  loss=0.464,  accuracy=0.875\n",
      "Epoch  2,  step=1426,  loss=0.521,  accuracy=0.781\n",
      "Epoch  2,  step=1427,  loss=0.456,  accuracy=0.906\n",
      "Epoch  2,  step=1428,  loss=0.028,  accuracy=1.000\n",
      "Epoch  2,  step=1429,  loss=0.296,  accuracy=0.844\n",
      "Epoch  2,  step=1430,  loss=0.464,  accuracy=0.875\n",
      "Epoch  2,  step=1431,  loss=0.354,  accuracy=0.906\n",
      "Epoch  2,  step=1432,  loss=0.489,  accuracy=0.812\n",
      "Epoch  2,  step=1433,  loss=0.733,  accuracy=0.875\n",
      "Epoch  2,  step=1434,  loss=0.580,  accuracy=0.844\n",
      "Epoch  2,  step=1435,  loss=0.463,  accuracy=0.844\n",
      "Epoch  2,  step=1436,  loss=0.585,  accuracy=0.781\n",
      "Epoch  2,  step=1437,  loss=0.510,  accuracy=0.906\n",
      "Epoch  2,  step=1438,  loss=0.387,  accuracy=0.812\n",
      "Epoch  2,  step=1439,  loss=0.454,  accuracy=0.906\n",
      "Epoch  2,  step=1440,  loss=0.644,  accuracy=0.875\n",
      "Epoch  2,  step=1441,  loss=0.218,  accuracy=0.938\n",
      "Epoch  2,  step=1442,  loss=0.217,  accuracy=0.938\n",
      "Epoch  2,  step=1443,  loss=0.185,  accuracy=0.969\n",
      "Epoch  2,  step=1444,  loss=0.108,  accuracy=0.969\n",
      "Epoch  2,  step=1445,  loss=1.223,  accuracy=0.781\n",
      "Epoch  2,  step=1446,  loss=0.738,  accuracy=0.781\n",
      "Epoch  2,  step=1447,  loss=0.549,  accuracy=0.781\n",
      "Epoch  2,  step=1448,  loss=0.342,  accuracy=0.875\n",
      "Epoch  2,  step=1449,  loss=0.242,  accuracy=0.906\n",
      "Epoch  2,  step=1450,  loss=0.554,  accuracy=0.844\n",
      "Epoch  2,  step=1451,  loss=0.180,  accuracy=0.906\n",
      "Epoch  2,  step=1452,  loss=0.131,  accuracy=0.969\n",
      "Epoch  2,  step=1453,  loss=0.159,  accuracy=0.969\n",
      "Epoch  2,  step=1454,  loss=0.225,  accuracy=0.906\n",
      "Epoch  2,  step=1455,  loss=0.150,  accuracy=0.906\n",
      "Epoch  2,  step=1456,  loss=0.244,  accuracy=0.938\n",
      "Epoch  2,  step=1457,  loss=0.375,  accuracy=0.906\n",
      "Epoch  2,  step=1458,  loss=0.115,  accuracy=0.969\n",
      "Epoch  2,  step=1459,  loss=0.251,  accuracy=0.906\n",
      "Epoch  2,  step=1460,  loss=0.903,  accuracy=0.750\n",
      "Epoch  2,  step=1461,  loss=0.341,  accuracy=0.938\n",
      "Epoch  2,  step=1462,  loss=0.248,  accuracy=0.875\n",
      "Epoch  2,  step=1463,  loss=0.111,  accuracy=0.969\n",
      "Epoch  2,  step=1464,  loss=0.209,  accuracy=0.906\n",
      "Epoch  2,  step=1465,  loss=0.337,  accuracy=0.906\n",
      "Epoch  2,  step=1466,  loss=0.306,  accuracy=0.906\n",
      "Epoch  2,  step=1467,  loss=0.162,  accuracy=0.938\n",
      "Epoch  2,  step=1468,  loss=0.257,  accuracy=0.969\n",
      "Epoch  2,  step=1469,  loss=0.775,  accuracy=0.844\n",
      "Epoch  2,  step=1470,  loss=0.470,  accuracy=0.875\n",
      "Epoch  2,  step=1471,  loss=0.271,  accuracy=0.938\n",
      "Epoch  2,  step=1472,  loss=0.145,  accuracy=0.969\n",
      "Epoch  2,  step=1473,  loss=0.241,  accuracy=0.938\n",
      "Epoch  2,  step=1474,  loss=0.274,  accuracy=0.906\n",
      "Epoch  2,  step=1475,  loss=0.859,  accuracy=0.812\n",
      "Epoch  2,  step=1476,  loss=0.642,  accuracy=0.781\n",
      "Epoch  2,  step=1477,  loss=0.317,  accuracy=0.938\n",
      "Epoch  2,  step=1478,  loss=0.708,  accuracy=0.812\n",
      "Epoch  2,  step=1479,  loss=0.621,  accuracy=0.906\n",
      "Epoch  2,  step=1480,  loss=0.560,  accuracy=0.875\n",
      "Epoch  2,  step=1481,  loss=0.649,  accuracy=0.844\n",
      "Epoch  2,  step=1482,  loss=0.421,  accuracy=0.875\n",
      "Epoch  2,  step=1483,  loss=0.372,  accuracy=0.812\n",
      "Epoch  2,  step=1484,  loss=0.071,  accuracy=1.000\n",
      "Epoch  2,  step=1485,  loss=0.243,  accuracy=0.875\n",
      "Epoch  2,  step=1486,  loss=0.476,  accuracy=0.906\n",
      "Epoch  2,  step=1487,  loss=0.654,  accuracy=0.844\n",
      "Epoch  2,  step=1488,  loss=0.253,  accuracy=0.906\n",
      "Epoch  2,  step=1489,  loss=0.250,  accuracy=0.906\n",
      "Epoch  2,  step=1490,  loss=0.462,  accuracy=0.875\n",
      "Epoch  2,  step=1491,  loss=0.372,  accuracy=0.938\n",
      "Epoch  2,  step=1492,  loss=0.478,  accuracy=0.812\n",
      "Epoch  2,  step=1493,  loss=0.288,  accuracy=0.906\n",
      "Epoch  2,  step=1494,  loss=0.268,  accuracy=0.969\n",
      "Epoch  2,  step=1495,  loss=0.142,  accuracy=0.969\n",
      "Epoch  2,  step=1496,  loss=0.275,  accuracy=0.906\n",
      "Epoch  2,  step=1497,  loss=0.664,  accuracy=0.906\n",
      "Epoch  2,  step=1498,  loss=0.475,  accuracy=0.875\n",
      "Epoch  2,  step=1499,  loss=0.217,  accuracy=0.906\n",
      "Epoch  2,  step=1500,  loss=0.255,  accuracy=0.906\n",
      "Epoch  2,  step=1501,  loss=0.174,  accuracy=0.906\n",
      "Epoch  2,  step=1502,  loss=0.361,  accuracy=0.875\n",
      "Epoch  2,  step=1503,  loss=0.356,  accuracy=0.906\n",
      "Epoch  2,  step=1504,  loss=0.127,  accuracy=0.969\n",
      "Epoch  2,  step=1505,  loss=0.157,  accuracy=0.938\n",
      "Epoch  2,  step=1506,  loss=0.072,  accuracy=0.969\n",
      "Epoch  2,  step=1507,  loss=0.066,  accuracy=0.969\n",
      "Epoch  2,  step=1508,  loss=0.206,  accuracy=0.938\n",
      "Epoch  2,  step=1509,  loss=0.104,  accuracy=0.938\n",
      "Epoch  2,  step=1510,  loss=0.390,  accuracy=0.875\n",
      "Epoch  2,  step=1511,  loss=0.577,  accuracy=0.781\n",
      "Epoch  2,  step=1512,  loss=0.726,  accuracy=0.781\n",
      "Epoch  2,  step=1513,  loss=0.037,  accuracy=1.000\n",
      "Epoch  2,  step=1514,  loss=0.248,  accuracy=0.844\n",
      "Epoch  2,  step=1515,  loss=0.322,  accuracy=0.938\n",
      "Epoch  2,  step=1516,  loss=0.268,  accuracy=0.906\n",
      "Epoch  2,  step=1517,  loss=0.216,  accuracy=0.938\n",
      "Epoch  2,  step=1518,  loss=0.245,  accuracy=0.938\n",
      "Epoch  2,  step=1519,  loss=0.042,  accuracy=1.000\n",
      "Epoch  2,  step=1520,  loss=0.567,  accuracy=0.906\n",
      "Epoch  2,  step=1521,  loss=0.305,  accuracy=0.875\n",
      "Epoch  2,  step=1522,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step=1523,  loss=0.261,  accuracy=0.938\n",
      "Epoch  2,  step=1524,  loss=0.175,  accuracy=0.938\n",
      "Epoch  2,  step=1525,  loss=0.166,  accuracy=0.906\n",
      "Epoch  2,  step=1526,  loss=0.410,  accuracy=0.844\n",
      "Epoch  2,  step=1527,  loss=0.358,  accuracy=0.875\n",
      "Epoch  2,  step=1528,  loss=0.802,  accuracy=0.844\n",
      "Epoch  2,  step=1529,  loss=0.821,  accuracy=0.750\n",
      "Epoch  2,  step=1530,  loss=0.732,  accuracy=0.875\n",
      "Epoch  2,  step=1531,  loss=1.483,  accuracy=0.750\n",
      "Epoch  2,  step=1532,  loss=0.343,  accuracy=0.844\n",
      "Epoch  2,  step=1533,  loss=0.750,  accuracy=0.844\n",
      "Epoch  2,  step=1534,  loss=0.388,  accuracy=0.844\n",
      "Epoch  2,  step=1535,  loss=0.667,  accuracy=0.812\n",
      "Epoch  2,  step=1536,  loss=0.325,  accuracy=0.875\n",
      "Epoch  2,  step=1537,  loss=1.234,  accuracy=0.688\n",
      "Epoch  2,  step=1538,  loss=0.330,  accuracy=0.844\n",
      "Epoch  2,  step=1539,  loss=0.185,  accuracy=0.938\n",
      "Epoch  2,  step=1540,  loss=0.215,  accuracy=0.938\n",
      "Epoch  2,  step=1541,  loss=0.181,  accuracy=0.938\n",
      "Epoch  2,  step=1542,  loss=0.073,  accuracy=0.969\n",
      "Epoch  2,  step=1543,  loss=0.086,  accuracy=0.969\n",
      "Epoch  2,  step=1544,  loss=0.902,  accuracy=0.750\n",
      "Epoch  2,  step=1545,  loss=0.232,  accuracy=0.938\n",
      "Epoch  2,  step=1546,  loss=0.923,  accuracy=0.750\n",
      "Epoch  2,  step=1547,  loss=0.689,  accuracy=0.781\n",
      "Epoch  2,  step=1548,  loss=0.806,  accuracy=0.812\n",
      "Epoch  2,  step=1549,  loss=0.855,  accuracy=0.875\n",
      "Epoch  2,  step=1550,  loss=0.433,  accuracy=0.875\n",
      "Epoch  2,  step=1551,  loss=0.791,  accuracy=0.812\n",
      "Epoch  2,  step=1552,  loss=0.447,  accuracy=0.875\n",
      "Epoch  2,  step=1553,  loss=0.170,  accuracy=0.938\n",
      "Epoch  2,  step=1554,  loss=0.264,  accuracy=0.938\n",
      "Epoch  2,  step=1555,  loss=0.174,  accuracy=0.969\n",
      "Epoch  2,  step=1556,  loss=0.372,  accuracy=0.812\n",
      "Epoch  2,  step=1557,  loss=0.316,  accuracy=0.844\n",
      "Epoch  2,  step=1558,  loss=0.089,  accuracy=0.969\n",
      "Epoch  2,  step=1559,  loss=0.803,  accuracy=0.812\n",
      "Epoch  2,  step=1560,  loss=0.283,  accuracy=0.844\n",
      "Epoch  2,  step=1561,  loss=0.154,  accuracy=0.969\n",
      "Epoch  2,  step=1562,  loss=0.398,  accuracy=0.875\n",
      "Epoch  2,  step=1563,  loss=0.321,  accuracy=0.844\n",
      "Epoch  2,  step=1564,  loss=0.342,  accuracy=0.875\n",
      "Epoch  2,  step=1565,  loss=0.173,  accuracy=0.906\n",
      "Epoch  2,  step=1566,  loss=0.223,  accuracy=0.938\n",
      "Epoch  2,  step=1567,  loss=0.525,  accuracy=0.906\n",
      "Epoch  2,  step=1568,  loss=0.195,  accuracy=0.938\n",
      "Epoch  2,  step=1569,  loss=0.798,  accuracy=0.844\n",
      "Epoch  2,  step=1570,  loss=0.636,  accuracy=0.875\n",
      "Epoch  2,  step=1571,  loss=0.109,  accuracy=0.969\n",
      "Epoch  2,  step=1572,  loss=0.574,  accuracy=0.812\n",
      "Epoch  2,  step=1573,  loss=0.536,  accuracy=0.812\n",
      "Epoch  2,  step=1574,  loss=0.639,  accuracy=0.750\n",
      "Epoch  2,  step=1575,  loss=1.128,  accuracy=0.656\n",
      "Epoch  2,  step=1576,  loss=0.383,  accuracy=0.875\n",
      "Epoch  2,  step=1577,  loss=0.102,  accuracy=0.938\n",
      "Epoch  2,  step=1578,  loss=0.620,  accuracy=0.906\n",
      "Epoch  2,  step=1579,  loss=0.127,  accuracy=0.969\n",
      "Epoch  2,  step=1580,  loss=1.076,  accuracy=0.812\n",
      "Epoch  2,  step=1581,  loss=0.198,  accuracy=0.875\n",
      "Epoch  2,  step=1582,  loss=0.970,  accuracy=0.844\n",
      "Epoch  2,  step=1583,  loss=0.208,  accuracy=0.969\n",
      "Epoch  2,  step=1584,  loss=0.488,  accuracy=0.844\n",
      "Epoch  2,  step=1585,  loss=0.399,  accuracy=0.875\n",
      "Epoch  2,  step=1586,  loss=0.327,  accuracy=0.906\n",
      "Epoch  2,  step=1587,  loss=0.308,  accuracy=0.906\n",
      "Epoch  2,  step=1588,  loss=0.307,  accuracy=0.844\n",
      "Epoch  2,  step=1589,  loss=0.168,  accuracy=0.938\n",
      "Epoch  2,  step=1590,  loss=0.312,  accuracy=0.844\n",
      "Epoch  2,  step=1591,  loss=0.074,  accuracy=1.000\n",
      "Epoch  2,  step=1592,  loss=0.208,  accuracy=0.906\n",
      "Epoch  2,  step=1593,  loss=0.198,  accuracy=0.875\n",
      "Epoch  2,  step=1594,  loss=0.108,  accuracy=0.938\n",
      "Epoch  2,  step=1595,  loss=0.474,  accuracy=0.875\n",
      "Epoch  2,  step=1596,  loss=0.239,  accuracy=0.938\n",
      "Epoch  2,  step=1597,  loss=0.337,  accuracy=0.875\n",
      "Epoch  2,  step=1598,  loss=0.306,  accuracy=0.875\n",
      "Epoch  2,  step=1599,  loss=0.566,  accuracy=0.844\n",
      "Epoch  2,  step=1600,  loss=0.218,  accuracy=0.906\n",
      "Epoch  2,  step=1601,  loss=0.377,  accuracy=0.844\n",
      "Epoch  2,  step=1602,  loss=0.513,  accuracy=0.844\n",
      "Epoch  2,  step=1603,  loss=0.534,  accuracy=0.906\n",
      "Epoch  2,  step=1604,  loss=0.294,  accuracy=0.906\n",
      "Epoch  2,  step=1605,  loss=0.166,  accuracy=0.938\n",
      "Epoch  2,  step=1606,  loss=0.219,  accuracy=0.906\n",
      "Epoch  2,  step=1607,  loss=0.170,  accuracy=0.875\n",
      "Epoch  2,  step=1608,  loss=0.125,  accuracy=0.969\n",
      "Epoch  2,  step=1609,  loss=0.143,  accuracy=0.938\n",
      "Epoch  2,  step=1610,  loss=0.526,  accuracy=0.875\n",
      "Epoch  2,  step=1611,  loss=0.319,  accuracy=0.844\n",
      "Epoch  2,  step=1612,  loss=0.059,  accuracy=1.000\n",
      "Epoch  2,  step=1613,  loss=0.317,  accuracy=0.938\n",
      "Epoch  2,  step=1614,  loss=0.115,  accuracy=0.969\n",
      "Epoch  2,  step=1615,  loss=0.331,  accuracy=0.938\n",
      "Epoch  2,  step=1616,  loss=0.586,  accuracy=0.906\n",
      "Epoch  2,  step=1617,  loss=0.383,  accuracy=0.906\n",
      "Epoch  2,  step=1618,  loss=0.308,  accuracy=0.906\n",
      "Epoch  2,  step=1619,  loss=0.074,  accuracy=0.969\n",
      "Epoch  2,  step=1620,  loss=0.312,  accuracy=0.844\n",
      "Epoch  2,  step=1621,  loss=0.258,  accuracy=0.875\n",
      "Epoch  2,  step=1622,  loss=0.273,  accuracy=0.875\n",
      "Epoch  2,  step=1623,  loss=0.575,  accuracy=0.781\n",
      "Epoch  2,  step=1624,  loss=0.798,  accuracy=0.812\n",
      "Epoch  2,  step=1625,  loss=0.092,  accuracy=0.969\n",
      "Epoch  2,  step=1626,  loss=0.143,  accuracy=0.938\n",
      "Epoch  2,  step=1627,  loss=0.513,  accuracy=0.844\n",
      "Epoch  2,  step=1628,  loss=0.837,  accuracy=0.719\n",
      "Epoch  2,  step=1629,  loss=1.169,  accuracy=0.812\n",
      "Epoch  2,  step=1630,  loss=0.648,  accuracy=0.844\n",
      "Epoch  2,  step=1631,  loss=0.241,  accuracy=0.906\n",
      "Epoch  2,  step=1632,  loss=0.516,  accuracy=0.812\n",
      "Epoch  2,  step=1633,  loss=0.325,  accuracy=0.844\n",
      "Epoch  2,  step=1634,  loss=0.150,  accuracy=0.969\n",
      "Epoch  2,  step=1635,  loss=0.304,  accuracy=0.969\n",
      "Epoch  2,  step=1636,  loss=0.462,  accuracy=0.969\n",
      "Epoch  2,  step=1637,  loss=0.428,  accuracy=0.906\n",
      "Epoch  2,  step=1638,  loss=0.265,  accuracy=0.938\n",
      "Epoch  2,  step=1639,  loss=0.116,  accuracy=0.969\n",
      "Epoch  2,  step=1640,  loss=0.186,  accuracy=0.938\n",
      "Epoch  2,  step=1641,  loss=0.125,  accuracy=0.906\n",
      "Epoch  2,  step=1642,  loss=0.114,  accuracy=0.938\n",
      "Epoch  2,  step=1643,  loss=0.166,  accuracy=0.969\n",
      "Epoch  2,  step=1644,  loss=0.118,  accuracy=0.969\n",
      "Epoch  2,  step=1645,  loss=0.690,  accuracy=0.844\n",
      "Epoch  2,  step=1646,  loss=0.584,  accuracy=0.875\n",
      "Epoch  2,  step=1647,  loss=0.560,  accuracy=0.781\n",
      "Epoch  2,  step=1648,  loss=0.387,  accuracy=0.875\n",
      "Epoch  2,  step=1649,  loss=0.262,  accuracy=0.875\n",
      "Epoch  2,  step=1650,  loss=0.478,  accuracy=0.844\n",
      "Epoch  2,  step=1651,  loss=0.161,  accuracy=0.906\n",
      "Epoch  2,  step=1652,  loss=0.439,  accuracy=0.812\n",
      "Epoch  2,  step=1653,  loss=0.589,  accuracy=0.875\n",
      "Epoch  2,  step=1654,  loss=0.790,  accuracy=0.812\n",
      "Epoch  2,  step=1655,  loss=0.356,  accuracy=0.844\n",
      "Epoch  2,  step=1656,  loss=0.424,  accuracy=0.875\n",
      "Epoch  2,  step=1657,  loss=0.440,  accuracy=0.844\n",
      "Epoch  2,  step=1658,  loss=0.340,  accuracy=0.906\n",
      "Epoch  2,  step=1659,  loss=0.420,  accuracy=0.844\n",
      "Epoch  2,  step=1660,  loss=0.367,  accuracy=0.906\n",
      "Epoch  2,  step=1661,  loss=0.839,  accuracy=0.812\n",
      "Epoch  2,  step=1662,  loss=0.201,  accuracy=0.906\n",
      "Epoch  2,  step=1663,  loss=0.601,  accuracy=0.906\n",
      "Epoch  2,  step=1664,  loss=0.105,  accuracy=0.938\n",
      "Epoch  2,  step=1665,  loss=0.275,  accuracy=0.938\n",
      "Epoch  2,  step=1666,  loss=0.147,  accuracy=0.938\n",
      "Epoch  2,  step=1667,  loss=0.079,  accuracy=0.969\n",
      "Epoch  2,  step=1668,  loss=0.299,  accuracy=0.906\n",
      "Epoch  2,  step=1669,  loss=0.099,  accuracy=0.969\n",
      "Epoch  2,  step=1670,  loss=0.424,  accuracy=0.844\n",
      "Epoch  2,  step=1671,  loss=0.401,  accuracy=0.875\n",
      "Epoch  2,  step=1672,  loss=0.418,  accuracy=0.875\n",
      "Epoch  2,  step=1673,  loss=0.427,  accuracy=0.844\n",
      "Epoch  2,  step=1674,  loss=0.454,  accuracy=0.906\n",
      "Epoch  2,  step=1675,  loss=0.284,  accuracy=0.906\n",
      "Epoch  2,  step=1676,  loss=0.277,  accuracy=0.906\n",
      "Epoch  2,  step=1677,  loss=0.410,  accuracy=0.906\n",
      "Epoch  2,  step=1678,  loss=0.101,  accuracy=0.969\n",
      "Epoch  2,  step=1679,  loss=0.660,  accuracy=0.750\n",
      "Epoch  2,  step=1680,  loss=0.036,  accuracy=1.000\n",
      "Epoch  2,  step=1681,  loss=0.386,  accuracy=0.938\n",
      "Epoch  2,  step=1682,  loss=0.386,  accuracy=0.844\n",
      "Epoch  2,  step=1683,  loss=0.374,  accuracy=0.938\n",
      "Epoch  2,  step=1684,  loss=0.718,  accuracy=0.875\n",
      "Epoch  2,  step=1685,  loss=0.249,  accuracy=0.938\n",
      "Epoch  2,  step=1686,  loss=0.852,  accuracy=0.750\n",
      "Epoch  2,  step=1687,  loss=0.428,  accuracy=0.844\n",
      "Epoch  2,  step=1688,  loss=0.331,  accuracy=0.906\n",
      "Epoch  2,  step=1689,  loss=0.271,  accuracy=0.938\n",
      "Epoch  2,  step=1690,  loss=0.809,  accuracy=0.781\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  2,  step=1691,  loss=0.085,  accuracy=1.000\n",
      "Epoch  2,  step=1692,  loss=0.100,  accuracy=0.969\n",
      "Epoch  2,  step=1693,  loss=0.357,  accuracy=0.906\n",
      "Epoch  2,  step=1694,  loss=0.146,  accuracy=0.969\n",
      "Epoch  2,  step=1695,  loss=0.301,  accuracy=0.906\n",
      "Epoch  2,  step=1696,  loss=0.367,  accuracy=0.812\n",
      "Epoch  2,  step=1697,  loss=0.135,  accuracy=0.938\n",
      "Epoch  2,  step=1698,  loss=0.165,  accuracy=0.906\n",
      "Epoch  2,  step=1699,  loss=0.818,  accuracy=0.844\n",
      "Epoch  2,  step=1700,  loss=0.243,  accuracy=0.938\n",
      "Epoch  2,  step=1701,  loss=0.553,  accuracy=0.875\n",
      "Epoch  2,  step=1702,  loss=0.106,  accuracy=1.000\n",
      "Epoch  2,  step=1703,  loss=0.431,  accuracy=0.875\n",
      "Epoch  2,  step=1704,  loss=0.383,  accuracy=0.812\n",
      "Epoch  2,  step=1705,  loss=0.556,  accuracy=0.812\n",
      "Epoch  2,  step=1706,  loss=0.148,  accuracy=0.906\n",
      "Epoch  2,  step=1707,  loss=0.313,  accuracy=0.906\n",
      "Epoch  2,  step=1708,  loss=0.186,  accuracy=0.906\n",
      "Epoch  2,  step=1709,  loss=0.130,  accuracy=0.969\n",
      "Epoch  2,  step=1710,  loss=0.064,  accuracy=0.969\n",
      "Epoch  2,  step=1711,  loss=0.394,  accuracy=0.875\n",
      "Epoch  2,  step=1712,  loss=0.390,  accuracy=0.844\n",
      "Epoch  2,  step=1713,  loss=0.248,  accuracy=0.969\n",
      "Epoch  2,  step=1714,  loss=0.397,  accuracy=0.906\n",
      "Epoch  2,  step=1715,  loss=0.594,  accuracy=0.844\n",
      "Epoch  2,  step=1716,  loss=0.717,  accuracy=0.750\n",
      "Epoch  2,  step=1717,  loss=0.646,  accuracy=0.844\n",
      "Epoch  2,  step=1718,  loss=0.272,  accuracy=0.906\n",
      "Epoch  2,  step=1719,  loss=0.372,  accuracy=0.906\n",
      "Epoch  2,  step=1720,  loss=0.596,  accuracy=0.844\n",
      "Epoch  2,  step=1721,  loss=0.364,  accuracy=0.906\n",
      "Epoch  2,  step=1722,  loss=0.464,  accuracy=0.906\n",
      "Epoch  2,  step=1723,  loss=0.042,  accuracy=1.000\n",
      "Epoch  2,  step=1724,  loss=0.263,  accuracy=0.938\n",
      "Epoch  2,  step=1725,  loss=0.244,  accuracy=0.875\n",
      "Epoch  2,  step=1726,  loss=0.209,  accuracy=0.906\n",
      "Epoch  2,  step=1727,  loss=0.166,  accuracy=0.906\n",
      "Epoch  2,  step=1728,  loss=0.449,  accuracy=0.844\n",
      "Epoch  2,  step=1729,  loss=0.484,  accuracy=0.938\n",
      "Epoch  2,  step=1730,  loss=0.095,  accuracy=0.938\n",
      "Epoch  2,  step=1731,  loss=0.110,  accuracy=0.969\n",
      "Epoch  2,  step=1732,  loss=0.240,  accuracy=0.906\n",
      "Epoch  2,  step=1733,  loss=0.281,  accuracy=0.906\n",
      "Epoch  2,  step=1734,  loss=0.610,  accuracy=0.875\n",
      "Epoch  2,  step=1735,  loss=0.400,  accuracy=0.812\n",
      "Epoch  2,  step=1736,  loss=0.038,  accuracy=1.000\n",
      "Epoch  2,  step=1737,  loss=0.206,  accuracy=0.969\n",
      "Epoch  2,  step=1738,  loss=0.103,  accuracy=0.938\n",
      "Epoch  2,  step=1739,  loss=0.251,  accuracy=0.906\n",
      "Epoch  2,  step=1740,  loss=0.217,  accuracy=0.938\n",
      "Epoch  2,  step=1741,  loss=0.820,  accuracy=0.812\n",
      "Epoch  2,  step=1742,  loss=0.216,  accuracy=0.906\n",
      "Epoch  2,  step=1743,  loss=0.287,  accuracy=0.938\n",
      "Epoch  2,  step=1744,  loss=0.127,  accuracy=0.938\n",
      "Epoch  2,  step=1745,  loss=0.599,  accuracy=0.906\n",
      "Epoch  2,  step=1746,  loss=0.347,  accuracy=0.875\n",
      "Epoch  2,  step=1747,  loss=0.296,  accuracy=0.906\n",
      "Epoch  2,  step=1748,  loss=0.124,  accuracy=0.969\n",
      "Epoch  2,  step=1749,  loss=0.231,  accuracy=0.875\n",
      "Epoch  2,  step=1750,  loss=0.245,  accuracy=0.906\n",
      "Epoch  2,  step=1751,  loss=0.142,  accuracy=0.969\n",
      "Epoch  2,  step=1752,  loss=0.169,  accuracy=0.875\n",
      "Epoch  2,  step=1753,  loss=0.089,  accuracy=0.938\n",
      "Epoch  2,  step=1754,  loss=0.404,  accuracy=0.938\n",
      "Epoch  2,  step=1755,  loss=0.176,  accuracy=0.938\n",
      "Epoch  2,  step=1756,  loss=0.793,  accuracy=0.875\n",
      "Epoch  2,  step=1757,  loss=0.133,  accuracy=0.969\n",
      "Epoch  2,  step=1758,  loss=0.298,  accuracy=0.875\n",
      "Epoch  2,  step=1759,  loss=0.461,  accuracy=0.844\n",
      "Epoch  2,  step=1760,  loss=0.238,  accuracy=0.906\n",
      "Epoch  2,  step=1761,  loss=0.338,  accuracy=0.844\n",
      "Epoch  2,  step=1762,  loss=0.328,  accuracy=0.875\n",
      "Epoch  2,  step=1763,  loss=0.283,  accuracy=0.938\n",
      "Epoch  2,  step=1764,  loss=1.341,  accuracy=0.719\n",
      "Epoch  2,  step=1765,  loss=0.432,  accuracy=0.875\n",
      "Epoch  2,  step=1766,  loss=0.048,  accuracy=1.000\n",
      "Epoch  2,  step=1767,  loss=0.040,  accuracy=1.000\n",
      "Epoch  2,  step=1768,  loss=0.346,  accuracy=0.875\n",
      "Epoch  2,  step=1769,  loss=0.377,  accuracy=0.906\n",
      "Epoch  2,  step=1770,  loss=0.664,  accuracy=0.844\n",
      "Epoch  2,  step=1771,  loss=0.305,  accuracy=0.969\n",
      "Epoch  2,  step=1772,  loss=0.065,  accuracy=1.000\n",
      "Epoch  2,  step=1773,  loss=0.075,  accuracy=1.000\n",
      "Epoch  2,  step=1774,  loss=0.238,  accuracy=0.906\n",
      "Epoch  2,  step=1775,  loss=0.257,  accuracy=0.906\n",
      "Epoch  2,  step=1776,  loss=0.239,  accuracy=0.875\n",
      "Epoch  2,  step=1777,  loss=0.112,  accuracy=0.969\n",
      "Epoch  2,  step=1778,  loss=0.294,  accuracy=0.938\n",
      "Epoch  2,  step=1779,  loss=0.055,  accuracy=0.969\n",
      "Epoch  2,  step=1780,  loss=0.274,  accuracy=0.938\n",
      "Epoch  2,  step=1781,  loss=0.495,  accuracy=0.844\n",
      "Epoch  2,  step=1782,  loss=0.628,  accuracy=0.906\n",
      "Epoch  2,  step=1783,  loss=0.635,  accuracy=0.906\n",
      "Epoch  2,  step=1784,  loss=0.081,  accuracy=0.969\n",
      "Epoch  2,  step=1785,  loss=0.148,  accuracy=0.938\n",
      "Epoch  2,  step=1786,  loss=0.365,  accuracy=0.875\n",
      "Epoch  2,  step=1787,  loss=0.224,  accuracy=0.906\n",
      "Epoch  2,  step=1788,  loss=0.360,  accuracy=0.875\n",
      "Epoch  2,  step=1789,  loss=0.296,  accuracy=0.875\n",
      "Epoch  2,  step=1790,  loss=0.404,  accuracy=0.844\n",
      "Epoch  2,  step=1791,  loss=0.577,  accuracy=0.812\n",
      "Epoch  2,  step=1792,  loss=0.130,  accuracy=0.938\n",
      "Epoch  2,  step=1793,  loss=0.258,  accuracy=0.938\n",
      "Epoch  2,  step=1794,  loss=0.145,  accuracy=0.906\n",
      "Epoch  2,  step=1795,  loss=0.150,  accuracy=0.969\n",
      "Epoch  2,  step=1796,  loss=0.404,  accuracy=0.906\n",
      "Epoch  2,  step=1797,  loss=0.383,  accuracy=0.844\n",
      "Epoch  2,  step=1798,  loss=0.444,  accuracy=0.812\n",
      "Epoch  2,  step=1799,  loss=0.387,  accuracy=0.812\n",
      "Epoch  2,  step=1800,  loss=0.537,  accuracy=0.875\n",
      "Epoch  2,  step=1801,  loss=0.575,  accuracy=0.781\n",
      "Epoch  2,  step=1802,  loss=0.357,  accuracy=0.781\n",
      "Epoch  2,  step=1803,  loss=0.581,  accuracy=0.844\n",
      "Epoch  2,  step=1804,  loss=0.566,  accuracy=0.875\n",
      "Epoch  2,  step=1805,  loss=0.100,  accuracy=0.938\n",
      "Epoch  2,  step=1806,  loss=0.269,  accuracy=0.875\n",
      "Epoch  2,  step=1807,  loss=0.374,  accuracy=0.875\n",
      "Epoch  2,  step=1808,  loss=0.261,  accuracy=0.875\n",
      "Epoch  2,  step=1809,  loss=0.152,  accuracy=0.938\n",
      "Epoch  2,  step=1810,  loss=0.291,  accuracy=0.906\n",
      "Epoch  2,  step=1811,  loss=0.244,  accuracy=0.906\n",
      "Epoch  2,  step=1812,  loss=0.139,  accuracy=0.969\n",
      "Epoch  2,  step=1813,  loss=0.178,  accuracy=0.938\n",
      "Epoch  2,  step=1814,  loss=0.559,  accuracy=0.844\n",
      "Epoch  2,  step=1815,  loss=0.680,  accuracy=0.844\n",
      "Epoch  2,  step=1816,  loss=0.065,  accuracy=0.969\n",
      "Epoch  2,  step=1817,  loss=0.041,  accuracy=1.000\n",
      "Epoch  2,  step=1818,  loss=0.043,  accuracy=0.969\n",
      "Epoch  2,  step=1819,  loss=0.038,  accuracy=1.000\n",
      "Epoch  2,  step=1820,  loss=0.165,  accuracy=0.938\n",
      "Epoch  2,  step=1821,  loss=0.097,  accuracy=0.969\n",
      "Epoch  2,  step=1822,  loss=0.250,  accuracy=0.938\n",
      "Epoch  2,  step=1823,  loss=0.032,  accuracy=1.000\n",
      "Epoch  2,  step=1824,  loss=0.195,  accuracy=0.906\n",
      "Epoch  2,  step=1825,  loss=0.216,  accuracy=0.906\n",
      "Epoch  2,  step=1826,  loss=0.103,  accuracy=0.969\n",
      "Epoch  2,  step=1827,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step=1828,  loss=0.344,  accuracy=0.906\n",
      "Epoch  2,  step=1829,  loss=0.337,  accuracy=0.906\n",
      "Epoch  2,  step=1830,  loss=0.167,  accuracy=0.969\n",
      "Epoch  2,  step=1831,  loss=0.264,  accuracy=0.875\n",
      "Epoch  2,  step=1832,  loss=0.192,  accuracy=0.906\n",
      "Epoch  2,  step=1833,  loss=0.174,  accuracy=0.938\n",
      "Epoch  2,  step=1834,  loss=0.096,  accuracy=0.969\n",
      "Epoch  2,  step=1835,  loss=0.042,  accuracy=1.000\n",
      "Epoch  2,  step=1836,  loss=0.138,  accuracy=0.938\n",
      "Epoch  2,  step=1837,  loss=0.323,  accuracy=0.812\n",
      "Epoch  2,  step=1838,  loss=0.246,  accuracy=0.969\n",
      "Epoch  2,  step=1839,  loss=0.497,  accuracy=0.938\n",
      "Epoch  2,  step=1840,  loss=0.105,  accuracy=0.969\n",
      "Epoch  2,  step=1841,  loss=0.025,  accuracy=1.000\n",
      "Epoch  2,  step=1842,  loss=0.029,  accuracy=1.000\n",
      "Epoch  2,  step=1843,  loss=0.206,  accuracy=0.906\n",
      "Epoch  2,  step=1844,  loss=0.036,  accuracy=1.000\n",
      "Epoch  2,  step=1845,  loss=0.235,  accuracy=0.938\n",
      "Epoch  2,  step=1846,  loss=0.233,  accuracy=0.875\n",
      "Epoch  2,  step=1847,  loss=0.134,  accuracy=0.969\n",
      "Epoch  2,  step=1848,  loss=0.293,  accuracy=0.906\n",
      "Epoch  2,  step=1849,  loss=0.225,  accuracy=0.938\n",
      "Epoch  2,  step=1850,  loss=0.183,  accuracy=0.938\n",
      "Epoch  2,  step=1851,  loss=0.256,  accuracy=0.969\n",
      "Epoch  2,  step=1852,  loss=0.468,  accuracy=0.812\n",
      "Epoch  2,  step=1853,  loss=0.891,  accuracy=0.844\n",
      "Epoch  2,  step=1854,  loss=0.700,  accuracy=0.844\n",
      "Epoch  2,  step=1855,  loss=0.280,  accuracy=0.906\n",
      "Epoch  2,  step=1856,  loss=0.253,  accuracy=0.938\n",
      "Epoch  2,  step=1857,  loss=0.219,  accuracy=0.875\n",
      "Epoch  2,  step=1858,  loss=0.093,  accuracy=0.969\n",
      "Epoch  2,  step=1859,  loss=0.105,  accuracy=0.969\n",
      "Epoch  2,  step=1860,  loss=0.104,  accuracy=0.969\n",
      "Epoch  2,  step=1861,  loss=0.149,  accuracy=0.906\n",
      "Epoch  2,  step=1862,  loss=0.084,  accuracy=0.969\n",
      "Epoch  2,  step=1863,  loss=0.069,  accuracy=1.000\n",
      "Epoch  2,  step=1864,  loss=0.357,  accuracy=0.938\n",
      "Epoch  2,  step=1865,  loss=0.367,  accuracy=0.906\n",
      "Epoch  2,  step=1866,  loss=1.434,  accuracy=0.719\n",
      "Epoch  2,  step=1867,  loss=0.196,  accuracy=0.906\n",
      "Epoch  2,  step=1868,  loss=0.404,  accuracy=0.781\n",
      "Epoch  2,  step=1869,  loss=0.114,  accuracy=0.938\n",
      "Epoch  2,  step=1870,  loss=0.049,  accuracy=1.000\n",
      "Epoch  2,  step=1871,  loss=0.060,  accuracy=1.000\n",
      "Epoch  2,  step=1872,  loss=1.123,  accuracy=0.812\n",
      "Epoch  2,  step=1873,  loss=0.188,  accuracy=0.938\n",
      "Epoch  2,  step=1874,  loss=0.144,  accuracy=0.938\n",
      "Epoch  3,  step=   0,  loss=0.269,  accuracy=0.906\n",
      "Epoch  3,  step=   1,  loss=0.578,  accuracy=0.875\n",
      "Epoch  3,  step=   2,  loss=0.466,  accuracy=0.969\n",
      "Epoch  3,  step=   3,  loss=0.235,  accuracy=0.938\n",
      "Epoch  3,  step=   4,  loss=0.611,  accuracy=0.906\n",
      "Epoch  3,  step=   5,  loss=0.531,  accuracy=0.875\n",
      "Epoch  3,  step=   6,  loss=0.047,  accuracy=1.000\n",
      "Epoch  3,  step=   7,  loss=0.587,  accuracy=0.875\n",
      "Epoch  3,  step=   8,  loss=0.473,  accuracy=0.844\n",
      "Epoch  3,  step=   9,  loss=0.090,  accuracy=0.938\n",
      "Epoch  3,  step=  10,  loss=0.298,  accuracy=0.969\n",
      "Epoch  3,  step=  11,  loss=0.197,  accuracy=0.906\n",
      "Epoch  3,  step=  12,  loss=0.147,  accuracy=0.969\n",
      "Epoch  3,  step=  13,  loss=0.376,  accuracy=0.875\n",
      "Epoch  3,  step=  14,  loss=0.261,  accuracy=0.906\n",
      "Epoch  3,  step=  15,  loss=0.639,  accuracy=0.875\n",
      "Epoch  3,  step=  16,  loss=0.166,  accuracy=0.969\n",
      "Epoch  3,  step=  17,  loss=0.277,  accuracy=0.906\n",
      "Epoch  3,  step=  18,  loss=0.228,  accuracy=0.906\n",
      "Epoch  3,  step=  19,  loss=0.728,  accuracy=0.781\n",
      "Epoch  3,  step=  20,  loss=0.216,  accuracy=0.906\n",
      "Epoch  3,  step=  21,  loss=0.115,  accuracy=0.969\n",
      "Epoch  3,  step=  22,  loss=0.381,  accuracy=0.938\n",
      "Epoch  3,  step=  23,  loss=0.297,  accuracy=0.875\n",
      "Epoch  3,  step=  24,  loss=0.396,  accuracy=0.875\n",
      "Epoch  3,  step=  25,  loss=0.189,  accuracy=0.938\n",
      "Epoch  3,  step=  26,  loss=0.817,  accuracy=0.844\n",
      "Epoch  3,  step=  27,  loss=0.526,  accuracy=0.844\n",
      "Epoch  3,  step=  28,  loss=0.937,  accuracy=0.875\n",
      "Epoch  3,  step=  29,  loss=0.335,  accuracy=0.938\n",
      "Epoch  3,  step=  30,  loss=0.086,  accuracy=1.000\n",
      "Epoch  3,  step=  31,  loss=0.479,  accuracy=0.875\n",
      "Epoch  3,  step=  32,  loss=0.533,  accuracy=0.844\n",
      "Epoch  3,  step=  33,  loss=0.999,  accuracy=0.781\n",
      "Epoch  3,  step=  34,  loss=0.255,  accuracy=0.906\n",
      "Epoch  3,  step=  35,  loss=0.920,  accuracy=0.688\n",
      "Epoch  3,  step=  36,  loss=0.183,  accuracy=0.875\n",
      "Epoch  3,  step=  37,  loss=0.161,  accuracy=0.969\n",
      "Epoch  3,  step=  38,  loss=1.059,  accuracy=0.812\n",
      "Epoch  3,  step=  39,  loss=0.537,  accuracy=0.812\n",
      "Epoch  3,  step=  40,  loss=0.212,  accuracy=0.906\n",
      "Epoch  3,  step=  41,  loss=0.395,  accuracy=0.812\n",
      "Epoch  3,  step=  42,  loss=0.641,  accuracy=0.875\n",
      "Epoch  3,  step=  43,  loss=0.624,  accuracy=0.781\n",
      "Epoch  3,  step=  44,  loss=0.240,  accuracy=0.875\n",
      "Epoch  3,  step=  45,  loss=0.252,  accuracy=0.938\n",
      "Epoch  3,  step=  46,  loss=0.203,  accuracy=0.906\n",
      "Epoch  3,  step=  47,  loss=0.761,  accuracy=0.812\n",
      "Epoch  3,  step=  48,  loss=0.288,  accuracy=0.875\n",
      "Epoch  3,  step=  49,  loss=0.449,  accuracy=0.875\n",
      "Epoch  3,  step=  50,  loss=0.332,  accuracy=0.906\n",
      "Epoch  3,  step=  51,  loss=0.307,  accuracy=0.844\n",
      "Epoch  3,  step=  52,  loss=0.197,  accuracy=0.938\n",
      "Epoch  3,  step=  53,  loss=0.073,  accuracy=0.969\n",
      "Epoch  3,  step=  54,  loss=0.199,  accuracy=0.906\n",
      "Epoch  3,  step=  55,  loss=0.222,  accuracy=0.906\n",
      "Epoch  3,  step=  56,  loss=0.466,  accuracy=0.938\n",
      "Epoch  3,  step=  57,  loss=0.524,  accuracy=0.875\n",
      "Epoch  3,  step=  58,  loss=0.171,  accuracy=0.938\n",
      "Epoch  3,  step=  59,  loss=0.124,  accuracy=0.969\n",
      "Epoch  3,  step=  60,  loss=0.291,  accuracy=0.906\n",
      "Epoch  3,  step=  61,  loss=0.357,  accuracy=0.812\n",
      "Epoch  3,  step=  62,  loss=0.276,  accuracy=0.844\n",
      "Epoch  3,  step=  63,  loss=0.213,  accuracy=0.938\n",
      "Epoch  3,  step=  64,  loss=0.300,  accuracy=0.875\n",
      "Epoch  3,  step=  65,  loss=0.313,  accuracy=0.938\n",
      "Epoch  3,  step=  66,  loss=0.307,  accuracy=0.875\n",
      "Epoch  3,  step=  67,  loss=0.091,  accuracy=0.969\n",
      "Epoch  3,  step=  68,  loss=0.141,  accuracy=0.969\n",
      "Epoch  3,  step=  69,  loss=0.246,  accuracy=0.938\n",
      "Epoch  3,  step=  70,  loss=0.227,  accuracy=0.938\n",
      "Epoch  3,  step=  71,  loss=0.239,  accuracy=0.969\n",
      "Epoch  3,  step=  72,  loss=0.130,  accuracy=0.969\n",
      "Epoch  3,  step=  73,  loss=0.099,  accuracy=0.969\n",
      "Epoch  3,  step=  74,  loss=0.069,  accuracy=1.000\n",
      "Epoch  3,  step=  75,  loss=0.729,  accuracy=0.812\n",
      "Epoch  3,  step=  76,  loss=0.174,  accuracy=0.875\n",
      "Epoch  3,  step=  77,  loss=0.129,  accuracy=0.969\n",
      "Epoch  3,  step=  78,  loss=0.144,  accuracy=0.969\n",
      "Epoch  3,  step=  79,  loss=0.311,  accuracy=0.875\n",
      "Epoch  3,  step=  80,  loss=0.159,  accuracy=0.938\n",
      "Epoch  3,  step=  81,  loss=0.286,  accuracy=0.875\n",
      "Epoch  3,  step=  82,  loss=0.329,  accuracy=0.938\n",
      "Epoch  3,  step=  83,  loss=1.027,  accuracy=0.875\n",
      "Epoch  3,  step=  84,  loss=0.296,  accuracy=0.906\n",
      "Epoch  3,  step=  85,  loss=0.124,  accuracy=0.969\n",
      "Epoch  3,  step=  86,  loss=0.509,  accuracy=0.875\n",
      "Epoch  3,  step=  87,  loss=0.641,  accuracy=0.844\n",
      "Epoch  3,  step=  88,  loss=0.563,  accuracy=0.875\n",
      "Epoch  3,  step=  89,  loss=0.144,  accuracy=0.938\n",
      "Epoch  3,  step=  90,  loss=0.241,  accuracy=0.969\n",
      "Epoch  3,  step=  91,  loss=0.344,  accuracy=0.875\n",
      "Epoch  3,  step=  92,  loss=0.307,  accuracy=0.875\n",
      "Epoch  3,  step=  93,  loss=0.212,  accuracy=0.938\n",
      "Epoch  3,  step=  94,  loss=0.311,  accuracy=0.938\n",
      "Epoch  3,  step=  95,  loss=0.111,  accuracy=0.969\n",
      "Epoch  3,  step=  96,  loss=0.172,  accuracy=0.969\n",
      "Epoch  3,  step=  97,  loss=0.346,  accuracy=0.906\n",
      "Epoch  3,  step=  98,  loss=0.121,  accuracy=0.969\n",
      "Epoch  3,  step=  99,  loss=0.097,  accuracy=0.969\n",
      "Epoch  3,  step= 100,  loss=0.304,  accuracy=0.875\n",
      "Epoch  3,  step= 101,  loss=0.256,  accuracy=0.906\n",
      "Epoch  3,  step= 102,  loss=0.593,  accuracy=0.844\n",
      "Epoch  3,  step= 103,  loss=0.102,  accuracy=0.969\n",
      "Epoch  3,  step= 104,  loss=0.105,  accuracy=0.969\n",
      "Epoch  3,  step= 105,  loss=0.413,  accuracy=0.875\n",
      "Epoch  3,  step= 106,  loss=0.471,  accuracy=0.938\n",
      "Epoch  3,  step= 107,  loss=0.171,  accuracy=0.906\n",
      "Epoch  3,  step= 108,  loss=0.250,  accuracy=0.906\n",
      "Epoch  3,  step= 109,  loss=0.509,  accuracy=0.906\n",
      "Epoch  3,  step= 110,  loss=0.289,  accuracy=0.906\n",
      "Epoch  3,  step= 111,  loss=0.185,  accuracy=0.906\n",
      "Epoch  3,  step= 112,  loss=0.044,  accuracy=1.000\n",
      "Epoch  3,  step= 113,  loss=0.210,  accuracy=0.938\n",
      "Epoch  3,  step= 114,  loss=0.479,  accuracy=0.875\n",
      "Epoch  3,  step= 115,  loss=0.699,  accuracy=0.812\n",
      "Epoch  3,  step= 116,  loss=0.574,  accuracy=0.875\n",
      "Epoch  3,  step= 117,  loss=0.109,  accuracy=0.969\n",
      "Epoch  3,  step= 118,  loss=0.354,  accuracy=0.906\n",
      "Epoch  3,  step= 119,  loss=0.462,  accuracy=0.844\n",
      "Epoch  3,  step= 120,  loss=0.109,  accuracy=0.969\n",
      "Epoch  3,  step= 121,  loss=0.345,  accuracy=0.969\n",
      "Epoch  3,  step= 122,  loss=0.075,  accuracy=0.969\n",
      "Epoch  3,  step= 123,  loss=0.100,  accuracy=0.969\n",
      "Epoch  3,  step= 124,  loss=0.376,  accuracy=0.906\n",
      "Epoch  3,  step= 125,  loss=0.364,  accuracy=0.906\n",
      "Epoch  3,  step= 126,  loss=0.230,  accuracy=0.906\n",
      "Epoch  3,  step= 127,  loss=0.254,  accuracy=0.906\n",
      "Epoch  3,  step= 128,  loss=0.228,  accuracy=0.844\n",
      "Epoch  3,  step= 129,  loss=0.661,  accuracy=0.781\n",
      "Epoch  3,  step= 130,  loss=0.364,  accuracy=0.844\n",
      "Epoch  3,  step= 131,  loss=0.294,  accuracy=0.969\n",
      "Epoch  3,  step= 132,  loss=0.208,  accuracy=0.969\n",
      "Epoch  3,  step= 133,  loss=0.334,  accuracy=0.875\n",
      "Epoch  3,  step= 134,  loss=0.156,  accuracy=0.969\n",
      "Epoch  3,  step= 135,  loss=0.130,  accuracy=0.969\n",
      "Epoch  3,  step= 136,  loss=0.392,  accuracy=0.875\n",
      "Epoch  3,  step= 137,  loss=0.113,  accuracy=0.969\n",
      "Epoch  3,  step= 138,  loss=0.139,  accuracy=0.938\n",
      "Epoch  3,  step= 139,  loss=0.531,  accuracy=0.844\n",
      "Epoch  3,  step= 140,  loss=0.307,  accuracy=0.938\n",
      "Epoch  3,  step= 141,  loss=0.061,  accuracy=0.969\n",
      "Epoch  3,  step= 142,  loss=0.074,  accuracy=0.969\n",
      "Epoch  3,  step= 143,  loss=0.382,  accuracy=0.875\n",
      "Epoch  3,  step= 144,  loss=0.194,  accuracy=0.906\n",
      "Epoch  3,  step= 145,  loss=0.687,  accuracy=0.844\n",
      "Epoch  3,  step= 146,  loss=0.188,  accuracy=0.969\n",
      "Epoch  3,  step= 147,  loss=0.184,  accuracy=0.906\n",
      "Epoch  3,  step= 148,  loss=0.360,  accuracy=0.875\n",
      "Epoch  3,  step= 149,  loss=0.270,  accuracy=0.875\n",
      "Epoch  3,  step= 150,  loss=0.270,  accuracy=0.906\n",
      "Epoch  3,  step= 151,  loss=0.063,  accuracy=1.000\n",
      "Epoch  3,  step= 152,  loss=0.166,  accuracy=0.938\n",
      "Epoch  3,  step= 153,  loss=0.269,  accuracy=0.906\n",
      "Epoch  3,  step= 154,  loss=0.387,  accuracy=0.875\n",
      "Epoch  3,  step= 155,  loss=0.694,  accuracy=0.812\n",
      "Epoch  3,  step= 156,  loss=0.263,  accuracy=0.906\n",
      "Epoch  3,  step= 157,  loss=0.495,  accuracy=0.844\n",
      "Epoch  3,  step= 158,  loss=0.527,  accuracy=0.844\n",
      "Epoch  3,  step= 159,  loss=0.410,  accuracy=0.875\n",
      "Epoch  3,  step= 160,  loss=0.410,  accuracy=0.906\n",
      "Epoch  3,  step= 161,  loss=0.831,  accuracy=0.844\n",
      "Epoch  3,  step= 162,  loss=0.318,  accuracy=0.906\n",
      "Epoch  3,  step= 163,  loss=0.093,  accuracy=0.969\n",
      "Epoch  3,  step= 164,  loss=0.397,  accuracy=0.844\n",
      "Epoch  3,  step= 165,  loss=1.025,  accuracy=0.875\n",
      "Epoch  3,  step= 166,  loss=0.477,  accuracy=0.906\n",
      "Epoch  3,  step= 167,  loss=0.315,  accuracy=0.906\n",
      "Epoch  3,  step= 168,  loss=0.228,  accuracy=0.906\n",
      "Epoch  3,  step= 169,  loss=0.208,  accuracy=0.938\n",
      "Epoch  3,  step= 170,  loss=0.093,  accuracy=0.969\n",
      "Epoch  3,  step= 171,  loss=0.192,  accuracy=0.938\n",
      "Epoch  3,  step= 172,  loss=0.121,  accuracy=0.969\n",
      "Epoch  3,  step= 173,  loss=0.419,  accuracy=0.938\n",
      "Epoch  3,  step= 174,  loss=0.092,  accuracy=0.938\n",
      "Epoch  3,  step= 175,  loss=0.363,  accuracy=0.750\n",
      "Epoch  3,  step= 176,  loss=0.235,  accuracy=0.938\n",
      "Epoch  3,  step= 177,  loss=0.496,  accuracy=0.875\n",
      "Epoch  3,  step= 178,  loss=0.464,  accuracy=0.875\n",
      "Epoch  3,  step= 179,  loss=0.130,  accuracy=0.969\n",
      "Epoch  3,  step= 180,  loss=0.432,  accuracy=0.906\n",
      "Epoch  3,  step= 181,  loss=0.111,  accuracy=0.969\n",
      "Epoch  3,  step= 182,  loss=0.402,  accuracy=0.844\n",
      "Epoch  3,  step= 183,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step= 184,  loss=0.309,  accuracy=0.938\n",
      "Epoch  3,  step= 185,  loss=0.349,  accuracy=0.938\n",
      "Epoch  3,  step= 186,  loss=0.138,  accuracy=0.969\n",
      "Epoch  3,  step= 187,  loss=0.063,  accuracy=0.969\n",
      "Epoch  3,  step= 188,  loss=0.166,  accuracy=0.906\n",
      "Epoch  3,  step= 189,  loss=0.084,  accuracy=0.969\n",
      "Epoch  3,  step= 190,  loss=0.536,  accuracy=0.906\n",
      "Epoch  3,  step= 191,  loss=0.324,  accuracy=0.969\n",
      "Epoch  3,  step= 192,  loss=0.171,  accuracy=0.938\n",
      "Epoch  3,  step= 193,  loss=0.177,  accuracy=0.938\n",
      "Epoch  3,  step= 194,  loss=0.207,  accuracy=0.938\n",
      "Epoch  3,  step= 195,  loss=0.618,  accuracy=0.812\n",
      "Epoch  3,  step= 196,  loss=0.413,  accuracy=0.875\n",
      "Epoch  3,  step= 197,  loss=0.304,  accuracy=0.906\n",
      "Epoch  3,  step= 198,  loss=0.206,  accuracy=0.938\n",
      "Epoch  3,  step= 199,  loss=0.209,  accuracy=0.938\n",
      "Epoch  3,  step= 200,  loss=0.445,  accuracy=0.844\n",
      "Epoch  3,  step= 201,  loss=0.436,  accuracy=0.969\n",
      "Epoch  3,  step= 202,  loss=0.251,  accuracy=0.938\n",
      "Epoch  3,  step= 203,  loss=0.306,  accuracy=0.938\n",
      "Epoch  3,  step= 204,  loss=0.249,  accuracy=0.875\n",
      "Epoch  3,  step= 205,  loss=0.248,  accuracy=0.938\n",
      "Epoch  3,  step= 206,  loss=0.167,  accuracy=0.938\n",
      "Epoch  3,  step= 207,  loss=0.235,  accuracy=0.938\n",
      "Epoch  3,  step= 208,  loss=0.219,  accuracy=0.938\n",
      "Epoch  3,  step= 209,  loss=0.167,  accuracy=0.938\n",
      "Epoch  3,  step= 210,  loss=0.246,  accuracy=0.969\n",
      "Epoch  3,  step= 211,  loss=0.068,  accuracy=1.000\n",
      "Epoch  3,  step= 212,  loss=0.258,  accuracy=0.875\n",
      "Epoch  3,  step= 213,  loss=0.576,  accuracy=0.844\n",
      "Epoch  3,  step= 214,  loss=0.503,  accuracy=0.875\n",
      "Epoch  3,  step= 215,  loss=0.537,  accuracy=0.875\n",
      "Epoch  3,  step= 216,  loss=0.743,  accuracy=0.812\n",
      "Epoch  3,  step= 217,  loss=0.309,  accuracy=0.906\n",
      "Epoch  3,  step= 218,  loss=0.137,  accuracy=0.969\n",
      "Epoch  3,  step= 219,  loss=0.700,  accuracy=0.812\n",
      "Epoch  3,  step= 220,  loss=0.252,  accuracy=0.906\n",
      "Epoch  3,  step= 221,  loss=0.110,  accuracy=0.969\n",
      "Epoch  3,  step= 222,  loss=0.203,  accuracy=0.906\n",
      "Epoch  3,  step= 223,  loss=0.505,  accuracy=0.906\n",
      "Epoch  3,  step= 224,  loss=0.174,  accuracy=0.938\n",
      "Epoch  3,  step= 225,  loss=0.614,  accuracy=0.812\n",
      "Epoch  3,  step= 226,  loss=0.728,  accuracy=0.781\n",
      "Epoch  3,  step= 227,  loss=0.539,  accuracy=0.781\n",
      "Epoch  3,  step= 228,  loss=0.243,  accuracy=0.969\n",
      "Epoch  3,  step= 229,  loss=0.498,  accuracy=0.906\n",
      "Epoch  3,  step= 230,  loss=0.219,  accuracy=0.938\n",
      "Epoch  3,  step= 231,  loss=0.037,  accuracy=1.000\n",
      "Epoch  3,  step= 232,  loss=0.330,  accuracy=0.875\n",
      "Epoch  3,  step= 233,  loss=0.082,  accuracy=1.000\n",
      "Epoch  3,  step= 234,  loss=0.142,  accuracy=0.938\n",
      "Epoch  3,  step= 235,  loss=0.569,  accuracy=0.875\n",
      "Epoch  3,  step= 236,  loss=0.054,  accuracy=1.000\n",
      "Epoch  3,  step= 237,  loss=0.615,  accuracy=0.781\n",
      "Epoch  3,  step= 238,  loss=0.510,  accuracy=0.812\n",
      "Epoch  3,  step= 239,  loss=0.250,  accuracy=0.906\n",
      "Epoch  3,  step= 240,  loss=0.254,  accuracy=0.906\n",
      "Epoch  3,  step= 241,  loss=0.553,  accuracy=0.812\n",
      "Epoch  3,  step= 242,  loss=0.302,  accuracy=0.875\n",
      "Epoch  3,  step= 243,  loss=0.283,  accuracy=0.875\n",
      "Epoch  3,  step= 244,  loss=0.238,  accuracy=0.906\n",
      "Epoch  3,  step= 245,  loss=0.384,  accuracy=0.875\n",
      "Epoch  3,  step= 246,  loss=0.479,  accuracy=0.938\n",
      "Epoch  3,  step= 247,  loss=0.394,  accuracy=0.906\n",
      "Epoch  3,  step= 248,  loss=0.162,  accuracy=0.969\n",
      "Epoch  3,  step= 249,  loss=0.390,  accuracy=0.875\n",
      "Epoch  3,  step= 250,  loss=0.432,  accuracy=0.875\n",
      "Epoch  3,  step= 251,  loss=0.179,  accuracy=0.906\n",
      "Epoch  3,  step= 252,  loss=0.291,  accuracy=0.875\n",
      "Epoch  3,  step= 253,  loss=0.624,  accuracy=0.812\n",
      "Epoch  3,  step= 254,  loss=0.129,  accuracy=0.938\n",
      "Epoch  3,  step= 255,  loss=0.082,  accuracy=0.969\n",
      "Epoch  3,  step= 256,  loss=0.937,  accuracy=0.812\n",
      "Epoch  3,  step= 257,  loss=0.449,  accuracy=0.906\n",
      "Epoch  3,  step= 258,  loss=0.238,  accuracy=0.938\n",
      "Epoch  3,  step= 259,  loss=0.223,  accuracy=0.938\n",
      "Epoch  3,  step= 260,  loss=0.202,  accuracy=0.906\n",
      "Epoch  3,  step= 261,  loss=0.150,  accuracy=0.969\n",
      "Epoch  3,  step= 262,  loss=0.189,  accuracy=0.875\n",
      "Epoch  3,  step= 263,  loss=0.422,  accuracy=0.844\n",
      "Epoch  3,  step= 264,  loss=0.779,  accuracy=0.844\n",
      "Epoch  3,  step= 265,  loss=0.339,  accuracy=0.844\n",
      "Epoch  3,  step= 266,  loss=0.254,  accuracy=0.875\n",
      "Epoch  3,  step= 267,  loss=0.275,  accuracy=0.875\n",
      "Epoch  3,  step= 268,  loss=0.180,  accuracy=0.938\n",
      "Epoch  3,  step= 269,  loss=0.500,  accuracy=0.812\n",
      "Epoch  3,  step= 270,  loss=0.224,  accuracy=0.906\n",
      "Epoch  3,  step= 271,  loss=0.909,  accuracy=0.750\n",
      "Epoch  3,  step= 272,  loss=1.260,  accuracy=0.750\n",
      "Epoch  3,  step= 273,  loss=0.237,  accuracy=0.906\n",
      "Epoch  3,  step= 274,  loss=0.342,  accuracy=0.875\n",
      "Epoch  3,  step= 275,  loss=0.508,  accuracy=0.875\n",
      "Epoch  3,  step= 276,  loss=0.558,  accuracy=0.875\n",
      "Epoch  3,  step= 277,  loss=0.683,  accuracy=0.781\n",
      "Epoch  3,  step= 278,  loss=0.637,  accuracy=0.812\n",
      "Epoch  3,  step= 279,  loss=0.323,  accuracy=0.906\n",
      "Epoch  3,  step= 280,  loss=0.361,  accuracy=0.938\n",
      "Epoch  3,  step= 281,  loss=0.250,  accuracy=0.875\n",
      "Epoch  3,  step= 282,  loss=0.138,  accuracy=0.938\n",
      "Epoch  3,  step= 283,  loss=0.273,  accuracy=0.938\n",
      "Epoch  3,  step= 284,  loss=0.574,  accuracy=0.844\n",
      "Epoch  3,  step= 285,  loss=0.546,  accuracy=0.906\n",
      "Epoch  3,  step= 286,  loss=0.437,  accuracy=0.844\n",
      "Epoch  3,  step= 287,  loss=0.255,  accuracy=0.938\n",
      "Epoch  3,  step= 288,  loss=0.243,  accuracy=0.938\n",
      "Epoch  3,  step= 289,  loss=0.741,  accuracy=0.812\n",
      "Epoch  3,  step= 290,  loss=0.332,  accuracy=0.938\n",
      "Epoch  3,  step= 291,  loss=0.212,  accuracy=0.906\n",
      "Epoch  3,  step= 292,  loss=0.233,  accuracy=0.906\n",
      "Epoch  3,  step= 293,  loss=0.672,  accuracy=0.844\n",
      "Epoch  3,  step= 294,  loss=0.310,  accuracy=0.875\n",
      "Epoch  3,  step= 295,  loss=0.369,  accuracy=0.875\n",
      "Epoch  3,  step= 296,  loss=0.056,  accuracy=1.000\n",
      "Epoch  3,  step= 297,  loss=0.289,  accuracy=0.875\n",
      "Epoch  3,  step= 298,  loss=0.357,  accuracy=0.875\n",
      "Epoch  3,  step= 299,  loss=0.310,  accuracy=0.906\n",
      "Epoch  3,  step= 300,  loss=0.269,  accuracy=0.938\n",
      "Epoch  3,  step= 301,  loss=0.247,  accuracy=0.906\n",
      "Epoch  3,  step= 302,  loss=0.324,  accuracy=0.875\n",
      "Epoch  3,  step= 303,  loss=0.038,  accuracy=1.000\n",
      "Epoch  3,  step= 304,  loss=0.145,  accuracy=0.938\n",
      "Epoch  3,  step= 305,  loss=0.613,  accuracy=0.844\n",
      "Epoch  3,  step= 306,  loss=0.356,  accuracy=0.875\n",
      "Epoch  3,  step= 307,  loss=0.027,  accuracy=1.000\n",
      "Epoch  3,  step= 308,  loss=0.239,  accuracy=0.938\n",
      "Epoch  3,  step= 309,  loss=0.035,  accuracy=1.000\n",
      "Epoch  3,  step= 310,  loss=0.280,  accuracy=0.906\n",
      "Epoch  3,  step= 311,  loss=0.166,  accuracy=0.969\n",
      "Epoch  3,  step= 312,  loss=0.193,  accuracy=0.938\n",
      "Epoch  3,  step= 313,  loss=0.708,  accuracy=0.844\n",
      "Epoch  3,  step= 314,  loss=0.261,  accuracy=0.906\n",
      "Epoch  3,  step= 315,  loss=0.297,  accuracy=0.875\n",
      "Epoch  3,  step= 316,  loss=0.336,  accuracy=0.906\n",
      "Epoch  3,  step= 317,  loss=0.645,  accuracy=0.781\n",
      "Epoch  3,  step= 318,  loss=0.570,  accuracy=0.875\n",
      "Epoch  3,  step= 319,  loss=0.445,  accuracy=0.875\n",
      "Epoch  3,  step= 320,  loss=0.760,  accuracy=0.812\n",
      "Epoch  3,  step= 321,  loss=0.607,  accuracy=0.750\n",
      "Epoch  3,  step= 322,  loss=0.287,  accuracy=0.969\n",
      "Epoch  3,  step= 323,  loss=0.371,  accuracy=0.969\n",
      "Epoch  3,  step= 324,  loss=0.200,  accuracy=0.906\n",
      "Epoch  3,  step= 325,  loss=0.121,  accuracy=0.938\n",
      "Epoch  3,  step= 326,  loss=0.256,  accuracy=0.906\n",
      "Epoch  3,  step= 327,  loss=0.196,  accuracy=0.938\n",
      "Epoch  3,  step= 328,  loss=0.240,  accuracy=0.906\n",
      "Epoch  3,  step= 329,  loss=0.047,  accuracy=1.000\n",
      "Epoch  3,  step= 330,  loss=0.191,  accuracy=0.938\n",
      "Epoch  3,  step= 331,  loss=0.077,  accuracy=0.969\n",
      "Epoch  3,  step= 332,  loss=0.123,  accuracy=0.938\n",
      "Epoch  3,  step= 333,  loss=0.180,  accuracy=0.969\n",
      "Epoch  3,  step= 334,  loss=0.221,  accuracy=0.906\n",
      "Epoch  3,  step= 335,  loss=0.585,  accuracy=0.938\n",
      "Epoch  3,  step= 336,  loss=0.417,  accuracy=0.906\n",
      "Epoch  3,  step= 337,  loss=0.263,  accuracy=0.938\n",
      "Epoch  3,  step= 338,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step= 339,  loss=0.443,  accuracy=0.875\n",
      "Epoch  3,  step= 340,  loss=0.162,  accuracy=0.969\n",
      "Epoch  3,  step= 341,  loss=0.140,  accuracy=0.906\n",
      "Epoch  3,  step= 342,  loss=0.238,  accuracy=0.938\n",
      "Epoch  3,  step= 343,  loss=0.457,  accuracy=0.906\n",
      "Epoch  3,  step= 344,  loss=0.216,  accuracy=0.969\n",
      "Epoch  3,  step= 345,  loss=0.274,  accuracy=0.906\n",
      "Epoch  3,  step= 346,  loss=0.135,  accuracy=0.969\n",
      "Epoch  3,  step= 347,  loss=0.273,  accuracy=0.938\n",
      "Epoch  3,  step= 348,  loss=0.178,  accuracy=0.938\n",
      "Epoch  3,  step= 349,  loss=0.136,  accuracy=0.938\n",
      "Epoch  3,  step= 350,  loss=0.584,  accuracy=0.844\n",
      "Epoch  3,  step= 351,  loss=0.432,  accuracy=0.812\n",
      "Epoch  3,  step= 352,  loss=0.137,  accuracy=0.938\n",
      "Epoch  3,  step= 353,  loss=0.239,  accuracy=0.938\n",
      "Epoch  3,  step= 354,  loss=0.120,  accuracy=0.938\n",
      "Epoch  3,  step= 355,  loss=0.324,  accuracy=0.844\n",
      "Epoch  3,  step= 356,  loss=0.135,  accuracy=0.938\n",
      "Epoch  3,  step= 357,  loss=0.258,  accuracy=0.906\n",
      "Epoch  3,  step= 358,  loss=0.166,  accuracy=0.969\n",
      "Epoch  3,  step= 359,  loss=0.259,  accuracy=0.812\n",
      "Epoch  3,  step= 360,  loss=0.216,  accuracy=0.906\n",
      "Epoch  3,  step= 361,  loss=0.692,  accuracy=0.781\n",
      "Epoch  3,  step= 362,  loss=0.318,  accuracy=0.875\n",
      "Epoch  3,  step= 363,  loss=0.670,  accuracy=0.875\n",
      "Epoch  3,  step= 364,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step= 365,  loss=0.431,  accuracy=0.906\n",
      "Epoch  3,  step= 366,  loss=0.216,  accuracy=0.938\n",
      "Epoch  3,  step= 367,  loss=0.532,  accuracy=0.906\n",
      "Epoch  3,  step= 368,  loss=0.351,  accuracy=0.875\n",
      "Epoch  3,  step= 369,  loss=0.148,  accuracy=0.969\n",
      "Epoch  3,  step= 370,  loss=0.388,  accuracy=0.875\n",
      "Epoch  3,  step= 371,  loss=0.216,  accuracy=0.906\n",
      "Epoch  3,  step= 372,  loss=0.122,  accuracy=0.969\n",
      "Epoch  3,  step= 373,  loss=0.341,  accuracy=0.906\n",
      "Epoch  3,  step= 374,  loss=0.160,  accuracy=0.938\n",
      "Epoch  3,  step= 375,  loss=0.304,  accuracy=0.906\n",
      "Epoch  3,  step= 376,  loss=0.405,  accuracy=0.875\n",
      "Epoch  3,  step= 377,  loss=0.363,  accuracy=0.844\n",
      "Epoch  3,  step= 378,  loss=0.250,  accuracy=0.938\n",
      "Epoch  3,  step= 379,  loss=0.104,  accuracy=0.938\n",
      "Epoch  3,  step= 380,  loss=0.292,  accuracy=0.906\n",
      "Epoch  3,  step= 381,  loss=0.362,  accuracy=0.875\n",
      "Epoch  3,  step= 382,  loss=0.189,  accuracy=0.969\n",
      "Epoch  3,  step= 383,  loss=0.570,  accuracy=0.812\n",
      "Epoch  3,  step= 384,  loss=0.305,  accuracy=0.938\n",
      "Epoch  3,  step= 385,  loss=0.270,  accuracy=0.938\n",
      "Epoch  3,  step= 386,  loss=0.576,  accuracy=0.875\n",
      "Epoch  3,  step= 387,  loss=0.056,  accuracy=0.969\n",
      "Epoch  3,  step= 388,  loss=0.315,  accuracy=0.875\n",
      "Epoch  3,  step= 389,  loss=0.420,  accuracy=0.938\n",
      "Epoch  3,  step= 390,  loss=0.600,  accuracy=0.844\n",
      "Epoch  3,  step= 391,  loss=0.335,  accuracy=0.875\n",
      "Epoch  3,  step= 392,  loss=0.905,  accuracy=0.781\n",
      "Epoch  3,  step= 393,  loss=0.763,  accuracy=0.844\n",
      "Epoch  3,  step= 394,  loss=0.648,  accuracy=0.875\n",
      "Epoch  3,  step= 395,  loss=0.686,  accuracy=0.844\n",
      "Epoch  3,  step= 396,  loss=1.066,  accuracy=0.750\n",
      "Epoch  3,  step= 397,  loss=0.093,  accuracy=0.969\n",
      "Epoch  3,  step= 398,  loss=0.358,  accuracy=0.875\n",
      "Epoch  3,  step= 399,  loss=0.384,  accuracy=0.906\n",
      "Epoch  3,  step= 400,  loss=0.163,  accuracy=0.969\n",
      "Epoch  3,  step= 401,  loss=0.228,  accuracy=0.875\n",
      "Epoch  3,  step= 402,  loss=0.427,  accuracy=0.875\n",
      "Epoch  3,  step= 403,  loss=0.270,  accuracy=0.938\n",
      "Epoch  3,  step= 404,  loss=0.781,  accuracy=0.844\n",
      "Epoch  3,  step= 405,  loss=0.337,  accuracy=0.875\n",
      "Epoch  3,  step= 406,  loss=0.662,  accuracy=0.844\n",
      "Epoch  3,  step= 407,  loss=0.471,  accuracy=0.875\n",
      "Epoch  3,  step= 408,  loss=0.971,  accuracy=0.781\n",
      "Epoch  3,  step= 409,  loss=0.372,  accuracy=0.844\n",
      "Epoch  3,  step= 410,  loss=0.389,  accuracy=0.875\n",
      "Epoch  3,  step= 411,  loss=0.166,  accuracy=0.906\n",
      "Epoch  3,  step= 412,  loss=0.547,  accuracy=0.781\n",
      "Epoch  3,  step= 413,  loss=0.077,  accuracy=0.969\n",
      "Epoch  3,  step= 414,  loss=0.669,  accuracy=0.938\n",
      "Epoch  3,  step= 415,  loss=0.275,  accuracy=0.938\n",
      "Epoch  3,  step= 416,  loss=0.200,  accuracy=0.938\n",
      "Epoch  3,  step= 417,  loss=0.175,  accuracy=0.938\n",
      "Epoch  3,  step= 418,  loss=0.337,  accuracy=0.875\n",
      "Epoch  3,  step= 419,  loss=0.287,  accuracy=0.938\n",
      "Epoch  3,  step= 420,  loss=0.361,  accuracy=0.969\n",
      "Epoch  3,  step= 421,  loss=0.035,  accuracy=1.000\n",
      "Epoch  3,  step= 422,  loss=0.564,  accuracy=0.875\n",
      "Epoch  3,  step= 423,  loss=0.311,  accuracy=0.906\n",
      "Epoch  3,  step= 424,  loss=0.054,  accuracy=1.000\n",
      "Epoch  3,  step= 425,  loss=0.112,  accuracy=0.969\n",
      "Epoch  3,  step= 426,  loss=0.583,  accuracy=0.844\n",
      "Epoch  3,  step= 427,  loss=0.543,  accuracy=0.812\n",
      "Epoch  3,  step= 428,  loss=0.512,  accuracy=0.844\n",
      "Epoch  3,  step= 429,  loss=0.405,  accuracy=0.812\n",
      "Epoch  3,  step= 430,  loss=0.092,  accuracy=0.969\n",
      "Epoch  3,  step= 431,  loss=0.152,  accuracy=0.969\n",
      "Epoch  3,  step= 432,  loss=0.757,  accuracy=0.812\n",
      "Epoch  3,  step= 433,  loss=0.281,  accuracy=0.906\n",
      "Epoch  3,  step= 434,  loss=0.155,  accuracy=0.969\n",
      "Epoch  3,  step= 435,  loss=0.279,  accuracy=0.938\n",
      "Epoch  3,  step= 436,  loss=0.510,  accuracy=0.844\n",
      "Epoch  3,  step= 437,  loss=0.487,  accuracy=0.844\n",
      "Epoch  3,  step= 438,  loss=0.301,  accuracy=0.875\n",
      "Epoch  3,  step= 439,  loss=0.571,  accuracy=0.875\n",
      "Epoch  3,  step= 440,  loss=0.249,  accuracy=0.938\n",
      "Epoch  3,  step= 441,  loss=0.657,  accuracy=0.812\n",
      "Epoch  3,  step= 442,  loss=0.126,  accuracy=0.969\n",
      "Epoch  3,  step= 443,  loss=0.327,  accuracy=0.906\n",
      "Epoch  3,  step= 444,  loss=0.668,  accuracy=0.875\n",
      "Epoch  3,  step= 445,  loss=0.428,  accuracy=0.906\n",
      "Epoch  3,  step= 446,  loss=0.249,  accuracy=0.969\n",
      "Epoch  3,  step= 447,  loss=0.290,  accuracy=0.906\n",
      "Epoch  3,  step= 448,  loss=1.093,  accuracy=0.750\n",
      "Epoch  3,  step= 449,  loss=0.453,  accuracy=0.844\n",
      "Epoch  3,  step= 450,  loss=0.188,  accuracy=0.938\n",
      "Epoch  3,  step= 451,  loss=0.223,  accuracy=0.938\n",
      "Epoch  3,  step= 452,  loss=0.210,  accuracy=0.906\n",
      "Epoch  3,  step= 453,  loss=0.205,  accuracy=0.938\n",
      "Epoch  3,  step= 454,  loss=0.541,  accuracy=0.844\n",
      "Epoch  3,  step= 455,  loss=0.421,  accuracy=0.938\n",
      "Epoch  3,  step= 456,  loss=0.545,  accuracy=0.812\n",
      "Epoch  3,  step= 457,  loss=0.423,  accuracy=0.906\n",
      "Epoch  3,  step= 458,  loss=0.546,  accuracy=0.781\n",
      "Epoch  3,  step= 459,  loss=0.631,  accuracy=0.781\n",
      "Epoch  3,  step= 460,  loss=0.788,  accuracy=0.750\n",
      "Epoch  3,  step= 461,  loss=1.267,  accuracy=0.750\n",
      "Epoch  3,  step= 462,  loss=0.545,  accuracy=0.844\n",
      "Epoch  3,  step= 463,  loss=0.653,  accuracy=0.875\n",
      "Epoch  3,  step= 464,  loss=0.369,  accuracy=0.906\n",
      "Epoch  3,  step= 465,  loss=0.307,  accuracy=0.875\n",
      "Epoch  3,  step= 466,  loss=0.276,  accuracy=0.875\n",
      "Epoch  3,  step= 467,  loss=0.169,  accuracy=0.906\n",
      "Epoch  3,  step= 468,  loss=0.345,  accuracy=0.844\n",
      "Epoch  3,  step= 469,  loss=0.125,  accuracy=0.969\n",
      "Epoch  3,  step= 470,  loss=0.545,  accuracy=0.969\n",
      "Epoch  3,  step= 471,  loss=0.105,  accuracy=0.969\n",
      "Epoch  3,  step= 472,  loss=1.171,  accuracy=0.781\n",
      "Epoch  3,  step= 473,  loss=0.697,  accuracy=0.875\n",
      "Epoch  3,  step= 474,  loss=0.503,  accuracy=0.875\n",
      "Epoch  3,  step= 475,  loss=0.277,  accuracy=0.938\n",
      "Epoch  3,  step= 476,  loss=0.219,  accuracy=0.906\n",
      "Epoch  3,  step= 477,  loss=0.387,  accuracy=0.906\n",
      "Epoch  3,  step= 478,  loss=0.107,  accuracy=0.969\n",
      "Epoch  3,  step= 479,  loss=0.443,  accuracy=0.844\n",
      "Epoch  3,  step= 480,  loss=0.150,  accuracy=0.938\n",
      "Epoch  3,  step= 481,  loss=0.329,  accuracy=0.938\n",
      "Epoch  3,  step= 482,  loss=0.508,  accuracy=0.969\n",
      "Epoch  3,  step= 483,  loss=0.144,  accuracy=0.938\n",
      "Epoch  3,  step= 484,  loss=0.440,  accuracy=0.938\n",
      "Epoch  3,  step= 485,  loss=0.422,  accuracy=0.875\n",
      "Epoch  3,  step= 486,  loss=0.191,  accuracy=0.938\n",
      "Epoch  3,  step= 487,  loss=0.402,  accuracy=0.938\n",
      "Epoch  3,  step= 488,  loss=0.275,  accuracy=0.875\n",
      "Epoch  3,  step= 489,  loss=0.228,  accuracy=0.875\n",
      "Epoch  3,  step= 490,  loss=0.304,  accuracy=0.906\n",
      "Epoch  3,  step= 491,  loss=0.874,  accuracy=0.750\n",
      "Epoch  3,  step= 492,  loss=0.637,  accuracy=0.781\n",
      "Epoch  3,  step= 493,  loss=0.449,  accuracy=0.844\n",
      "Epoch  3,  step= 494,  loss=0.185,  accuracy=0.969\n",
      "Epoch  3,  step= 495,  loss=0.939,  accuracy=0.750\n",
      "Epoch  3,  step= 496,  loss=0.246,  accuracy=0.938\n",
      "Epoch  3,  step= 497,  loss=0.255,  accuracy=0.906\n",
      "Epoch  3,  step= 498,  loss=0.483,  accuracy=0.938\n",
      "Epoch  3,  step= 499,  loss=0.520,  accuracy=0.875\n",
      "Epoch  3,  step= 500,  loss=0.768,  accuracy=0.781\n",
      "Epoch  3,  step= 501,  loss=0.749,  accuracy=0.781\n",
      "Epoch  3,  step= 502,  loss=0.368,  accuracy=0.875\n",
      "Epoch  3,  step= 503,  loss=0.526,  accuracy=0.844\n",
      "Epoch  3,  step= 504,  loss=0.152,  accuracy=0.938\n",
      "Epoch  3,  step= 505,  loss=0.114,  accuracy=0.969\n",
      "Epoch  3,  step= 506,  loss=0.347,  accuracy=0.812\n",
      "Epoch  3,  step= 507,  loss=0.028,  accuracy=1.000\n",
      "Epoch  3,  step= 508,  loss=0.513,  accuracy=0.875\n",
      "Epoch  3,  step= 509,  loss=0.285,  accuracy=0.906\n",
      "Epoch  3,  step= 510,  loss=0.084,  accuracy=1.000\n",
      "Epoch  3,  step= 511,  loss=0.412,  accuracy=0.875\n",
      "Epoch  3,  step= 512,  loss=0.128,  accuracy=0.938\n",
      "Epoch  3,  step= 513,  loss=0.307,  accuracy=0.875\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  3,  step= 514,  loss=0.218,  accuracy=0.938\n",
      "Epoch  3,  step= 515,  loss=0.365,  accuracy=0.875\n",
      "Epoch  3,  step= 516,  loss=0.253,  accuracy=0.938\n",
      "Epoch  3,  step= 517,  loss=0.596,  accuracy=0.844\n",
      "Epoch  3,  step= 518,  loss=0.236,  accuracy=0.875\n",
      "Epoch  3,  step= 519,  loss=0.174,  accuracy=0.969\n",
      "Epoch  3,  step= 520,  loss=0.489,  accuracy=0.844\n",
      "Epoch  3,  step= 521,  loss=0.549,  accuracy=0.812\n",
      "Epoch  3,  step= 522,  loss=0.170,  accuracy=0.969\n",
      "Epoch  3,  step= 523,  loss=0.536,  accuracy=0.906\n",
      "Epoch  3,  step= 524,  loss=0.773,  accuracy=0.812\n",
      "Epoch  3,  step= 525,  loss=0.294,  accuracy=0.875\n",
      "Epoch  3,  step= 526,  loss=0.281,  accuracy=0.906\n",
      "Epoch  3,  step= 527,  loss=0.460,  accuracy=0.875\n",
      "Epoch  3,  step= 528,  loss=0.371,  accuracy=0.844\n",
      "Epoch  3,  step= 529,  loss=0.460,  accuracy=0.812\n",
      "Epoch  3,  step= 530,  loss=0.741,  accuracy=0.812\n",
      "Epoch  3,  step= 531,  loss=0.387,  accuracy=0.844\n",
      "Epoch  3,  step= 532,  loss=0.384,  accuracy=0.875\n",
      "Epoch  3,  step= 533,  loss=0.191,  accuracy=0.938\n",
      "Epoch  3,  step= 534,  loss=0.244,  accuracy=0.875\n",
      "Epoch  3,  step= 535,  loss=0.222,  accuracy=0.938\n",
      "Epoch  3,  step= 536,  loss=0.377,  accuracy=0.906\n",
      "Epoch  3,  step= 537,  loss=0.802,  accuracy=0.844\n",
      "Epoch  3,  step= 538,  loss=0.374,  accuracy=0.906\n",
      "Epoch  3,  step= 539,  loss=0.171,  accuracy=0.969\n",
      "Epoch  3,  step= 540,  loss=0.102,  accuracy=0.969\n",
      "Epoch  3,  step= 541,  loss=0.058,  accuracy=1.000\n",
      "Epoch  3,  step= 542,  loss=0.199,  accuracy=0.938\n",
      "Epoch  3,  step= 543,  loss=0.336,  accuracy=0.844\n",
      "Epoch  3,  step= 544,  loss=0.141,  accuracy=0.938\n",
      "Epoch  3,  step= 545,  loss=0.156,  accuracy=0.969\n",
      "Epoch  3,  step= 546,  loss=0.598,  accuracy=0.875\n",
      "Epoch  3,  step= 547,  loss=0.235,  accuracy=0.875\n",
      "Epoch  3,  step= 548,  loss=0.481,  accuracy=0.812\n",
      "Epoch  3,  step= 549,  loss=0.449,  accuracy=0.812\n",
      "Epoch  3,  step= 550,  loss=0.108,  accuracy=1.000\n",
      "Epoch  3,  step= 551,  loss=0.622,  accuracy=0.875\n",
      "Epoch  3,  step= 552,  loss=0.447,  accuracy=0.812\n",
      "Epoch  3,  step= 553,  loss=0.782,  accuracy=0.812\n",
      "Epoch  3,  step= 554,  loss=0.196,  accuracy=0.906\n",
      "Epoch  3,  step= 555,  loss=0.264,  accuracy=0.938\n",
      "Epoch  3,  step= 556,  loss=0.482,  accuracy=0.844\n",
      "Epoch  3,  step= 557,  loss=0.239,  accuracy=0.875\n",
      "Epoch  3,  step= 558,  loss=0.168,  accuracy=0.906\n",
      "Epoch  3,  step= 559,  loss=0.178,  accuracy=0.906\n",
      "Epoch  3,  step= 560,  loss=0.293,  accuracy=0.906\n",
      "Epoch  3,  step= 561,  loss=0.229,  accuracy=0.938\n",
      "Epoch  3,  step= 562,  loss=0.214,  accuracy=0.938\n",
      "Epoch  3,  step= 563,  loss=0.299,  accuracy=0.906\n",
      "Epoch  3,  step= 564,  loss=0.438,  accuracy=0.938\n",
      "Epoch  3,  step= 565,  loss=0.411,  accuracy=0.906\n",
      "Epoch  3,  step= 566,  loss=0.070,  accuracy=1.000\n",
      "Epoch  3,  step= 567,  loss=0.271,  accuracy=0.938\n",
      "Epoch  3,  step= 568,  loss=0.168,  accuracy=0.938\n",
      "Epoch  3,  step= 569,  loss=0.128,  accuracy=0.969\n",
      "Epoch  3,  step= 570,  loss=0.536,  accuracy=0.812\n",
      "Epoch  3,  step= 571,  loss=0.237,  accuracy=0.938\n",
      "Epoch  3,  step= 572,  loss=0.162,  accuracy=0.938\n",
      "Epoch  3,  step= 573,  loss=0.207,  accuracy=0.938\n",
      "Epoch  3,  step= 574,  loss=0.283,  accuracy=0.906\n",
      "Epoch  3,  step= 575,  loss=0.446,  accuracy=0.781\n",
      "Epoch  3,  step= 576,  loss=0.127,  accuracy=0.938\n",
      "Epoch  3,  step= 577,  loss=0.556,  accuracy=0.906\n",
      "Epoch  3,  step= 578,  loss=0.450,  accuracy=0.844\n",
      "Epoch  3,  step= 579,  loss=0.115,  accuracy=0.938\n",
      "Epoch  3,  step= 580,  loss=0.524,  accuracy=0.844\n",
      "Epoch  3,  step= 581,  loss=0.384,  accuracy=0.938\n",
      "Epoch  3,  step= 582,  loss=0.468,  accuracy=0.875\n",
      "Epoch  3,  step= 583,  loss=0.215,  accuracy=0.906\n",
      "Epoch  3,  step= 584,  loss=0.158,  accuracy=0.938\n",
      "Epoch  3,  step= 585,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step= 586,  loss=0.039,  accuracy=1.000\n",
      "Epoch  3,  step= 587,  loss=0.149,  accuracy=0.906\n",
      "Epoch  3,  step= 588,  loss=0.299,  accuracy=0.875\n",
      "Epoch  3,  step= 589,  loss=0.189,  accuracy=0.938\n",
      "Epoch  3,  step= 590,  loss=0.163,  accuracy=0.938\n",
      "Epoch  3,  step= 591,  loss=0.392,  accuracy=0.906\n",
      "Epoch  3,  step= 592,  loss=0.080,  accuracy=0.969\n",
      "Epoch  3,  step= 593,  loss=0.240,  accuracy=0.906\n",
      "Epoch  3,  step= 594,  loss=0.103,  accuracy=0.969\n",
      "Epoch  3,  step= 595,  loss=0.141,  accuracy=0.969\n",
      "Epoch  3,  step= 596,  loss=0.560,  accuracy=0.906\n",
      "Epoch  3,  step= 597,  loss=0.549,  accuracy=0.875\n",
      "Epoch  3,  step= 598,  loss=0.513,  accuracy=0.812\n",
      "Epoch  3,  step= 599,  loss=0.237,  accuracy=0.875\n",
      "Epoch  3,  step= 600,  loss=0.265,  accuracy=0.906\n",
      "Epoch  3,  step= 601,  loss=0.371,  accuracy=0.969\n",
      "Epoch  3,  step= 602,  loss=0.265,  accuracy=0.938\n",
      "Epoch  3,  step= 603,  loss=0.263,  accuracy=0.906\n",
      "Epoch  3,  step= 604,  loss=0.424,  accuracy=0.906\n",
      "Epoch  3,  step= 605,  loss=0.579,  accuracy=0.812\n",
      "Epoch  3,  step= 606,  loss=0.318,  accuracy=0.906\n",
      "Epoch  3,  step= 607,  loss=0.087,  accuracy=0.969\n",
      "Epoch  3,  step= 608,  loss=0.231,  accuracy=0.938\n",
      "Epoch  3,  step= 609,  loss=0.647,  accuracy=0.812\n",
      "Epoch  3,  step= 610,  loss=0.305,  accuracy=0.875\n",
      "Epoch  3,  step= 611,  loss=0.353,  accuracy=0.844\n",
      "Epoch  3,  step= 612,  loss=0.297,  accuracy=0.906\n",
      "Epoch  3,  step= 613,  loss=0.338,  accuracy=0.938\n",
      "Epoch  3,  step= 614,  loss=0.136,  accuracy=0.969\n",
      "Epoch  3,  step= 615,  loss=0.153,  accuracy=0.969\n",
      "Epoch  3,  step= 616,  loss=0.051,  accuracy=1.000\n",
      "Epoch  3,  step= 617,  loss=0.094,  accuracy=0.969\n",
      "Epoch  3,  step= 618,  loss=0.320,  accuracy=0.906\n",
      "Epoch  3,  step= 619,  loss=0.989,  accuracy=0.812\n",
      "Epoch  3,  step= 620,  loss=0.357,  accuracy=0.938\n",
      "Epoch  3,  step= 621,  loss=0.101,  accuracy=0.969\n",
      "Epoch  3,  step= 622,  loss=0.562,  accuracy=0.844\n",
      "Epoch  3,  step= 623,  loss=0.207,  accuracy=0.938\n",
      "Epoch  3,  step= 624,  loss=0.325,  accuracy=0.875\n",
      "Epoch  3,  step= 625,  loss=0.287,  accuracy=0.844\n",
      "Epoch  3,  step= 626,  loss=0.360,  accuracy=0.875\n",
      "Epoch  3,  step= 627,  loss=0.300,  accuracy=0.906\n",
      "Epoch  3,  step= 628,  loss=0.290,  accuracy=0.906\n",
      "Epoch  3,  step= 629,  loss=0.245,  accuracy=0.906\n",
      "Epoch  3,  step= 630,  loss=0.702,  accuracy=0.781\n",
      "Epoch  3,  step= 631,  loss=0.156,  accuracy=0.938\n",
      "Epoch  3,  step= 632,  loss=0.560,  accuracy=0.906\n",
      "Epoch  3,  step= 633,  loss=0.174,  accuracy=0.969\n",
      "Epoch  3,  step= 634,  loss=0.153,  accuracy=0.938\n",
      "Epoch  3,  step= 635,  loss=0.442,  accuracy=0.875\n",
      "Epoch  3,  step= 636,  loss=0.154,  accuracy=0.938\n",
      "Epoch  3,  step= 637,  loss=0.174,  accuracy=0.938\n",
      "Epoch  3,  step= 638,  loss=0.101,  accuracy=0.969\n",
      "Epoch  3,  step= 639,  loss=0.169,  accuracy=0.969\n",
      "Epoch  3,  step= 640,  loss=0.029,  accuracy=1.000\n",
      "Epoch  3,  step= 641,  loss=0.363,  accuracy=0.906\n",
      "Epoch  3,  step= 642,  loss=0.726,  accuracy=0.844\n",
      "Epoch  3,  step= 643,  loss=0.188,  accuracy=0.906\n",
      "Epoch  3,  step= 644,  loss=0.270,  accuracy=0.938\n",
      "Epoch  3,  step= 645,  loss=0.375,  accuracy=0.906\n",
      "Epoch  3,  step= 646,  loss=0.277,  accuracy=0.938\n",
      "Epoch  3,  step= 647,  loss=0.825,  accuracy=0.781\n",
      "Epoch  3,  step= 648,  loss=0.247,  accuracy=0.938\n",
      "Epoch  3,  step= 649,  loss=0.839,  accuracy=0.906\n",
      "Epoch  3,  step= 650,  loss=0.485,  accuracy=0.906\n",
      "Epoch  3,  step= 651,  loss=0.139,  accuracy=0.938\n",
      "Epoch  3,  step= 652,  loss=0.364,  accuracy=0.938\n",
      "Epoch  3,  step= 653,  loss=0.670,  accuracy=0.812\n",
      "Epoch  3,  step= 654,  loss=0.452,  accuracy=0.875\n",
      "Epoch  3,  step= 655,  loss=0.607,  accuracy=0.812\n",
      "Epoch  3,  step= 656,  loss=0.557,  accuracy=0.875\n",
      "Epoch  3,  step= 657,  loss=0.158,  accuracy=0.938\n",
      "Epoch  3,  step= 658,  loss=0.283,  accuracy=0.969\n",
      "Epoch  3,  step= 659,  loss=0.209,  accuracy=0.906\n",
      "Epoch  3,  step= 660,  loss=0.616,  accuracy=0.875\n",
      "Epoch  3,  step= 661,  loss=0.298,  accuracy=0.938\n",
      "Epoch  3,  step= 662,  loss=0.642,  accuracy=0.844\n",
      "Epoch  3,  step= 663,  loss=0.079,  accuracy=1.000\n",
      "Epoch  3,  step= 664,  loss=0.105,  accuracy=0.969\n",
      "Epoch  3,  step= 665,  loss=0.470,  accuracy=0.844\n",
      "Epoch  3,  step= 666,  loss=0.240,  accuracy=0.969\n",
      "Epoch  3,  step= 667,  loss=0.325,  accuracy=0.906\n",
      "Epoch  3,  step= 668,  loss=0.256,  accuracy=0.969\n",
      "Epoch  3,  step= 669,  loss=0.404,  accuracy=0.906\n",
      "Epoch  3,  step= 670,  loss=0.230,  accuracy=0.906\n",
      "Epoch  3,  step= 671,  loss=0.168,  accuracy=0.969\n",
      "Epoch  3,  step= 672,  loss=0.262,  accuracy=0.875\n",
      "Epoch  3,  step= 673,  loss=0.184,  accuracy=0.938\n",
      "Epoch  3,  step= 674,  loss=0.579,  accuracy=0.875\n",
      "Epoch  3,  step= 675,  loss=0.536,  accuracy=0.906\n",
      "Epoch  3,  step= 676,  loss=0.280,  accuracy=0.906\n",
      "Epoch  3,  step= 677,  loss=0.211,  accuracy=0.938\n",
      "Epoch  3,  step= 678,  loss=0.414,  accuracy=0.875\n",
      "Epoch  3,  step= 679,  loss=0.227,  accuracy=0.906\n",
      "Epoch  3,  step= 680,  loss=0.239,  accuracy=0.938\n",
      "Epoch  3,  step= 681,  loss=0.052,  accuracy=1.000\n",
      "Epoch  3,  step= 682,  loss=0.058,  accuracy=1.000\n",
      "Epoch  3,  step= 683,  loss=0.078,  accuracy=1.000\n",
      "Epoch  3,  step= 684,  loss=0.313,  accuracy=0.906\n",
      "Epoch  3,  step= 685,  loss=0.262,  accuracy=0.844\n",
      "Epoch  3,  step= 686,  loss=0.209,  accuracy=0.969\n",
      "Epoch  3,  step= 687,  loss=0.226,  accuracy=0.938\n",
      "Epoch  3,  step= 688,  loss=0.163,  accuracy=0.969\n",
      "Epoch  3,  step= 689,  loss=0.265,  accuracy=0.906\n",
      "Epoch  3,  step= 690,  loss=0.276,  accuracy=0.906\n",
      "Epoch  3,  step= 691,  loss=0.467,  accuracy=0.844\n",
      "Epoch  3,  step= 692,  loss=0.451,  accuracy=0.844\n",
      "Epoch  3,  step= 693,  loss=0.471,  accuracy=0.844\n",
      "Epoch  3,  step= 694,  loss=1.171,  accuracy=0.906\n",
      "Epoch  3,  step= 695,  loss=0.068,  accuracy=1.000\n",
      "Epoch  3,  step= 696,  loss=0.432,  accuracy=0.875\n",
      "Epoch  3,  step= 697,  loss=0.144,  accuracy=0.938\n",
      "Epoch  3,  step= 698,  loss=0.035,  accuracy=1.000\n",
      "Epoch  3,  step= 699,  loss=0.128,  accuracy=0.969\n",
      "Epoch  3,  step= 700,  loss=0.223,  accuracy=0.906\n",
      "Epoch  3,  step= 701,  loss=0.063,  accuracy=1.000\n",
      "Epoch  3,  step= 702,  loss=0.822,  accuracy=0.781\n",
      "Epoch  3,  step= 703,  loss=0.589,  accuracy=0.875\n",
      "Epoch  3,  step= 704,  loss=0.712,  accuracy=0.844\n",
      "Epoch  3,  step= 705,  loss=0.666,  accuracy=0.812\n",
      "Epoch  3,  step= 706,  loss=0.537,  accuracy=0.781\n",
      "Epoch  3,  step= 707,  loss=0.302,  accuracy=0.906\n",
      "Epoch  3,  step= 708,  loss=0.279,  accuracy=0.875\n",
      "Epoch  3,  step= 709,  loss=0.522,  accuracy=0.844\n",
      "Epoch  3,  step= 710,  loss=0.293,  accuracy=0.844\n",
      "Epoch  3,  step= 711,  loss=0.514,  accuracy=0.906\n",
      "Epoch  3,  step= 712,  loss=0.517,  accuracy=0.875\n",
      "Epoch  3,  step= 713,  loss=0.189,  accuracy=0.938\n",
      "Epoch  3,  step= 714,  loss=0.201,  accuracy=0.906\n",
      "Epoch  3,  step= 715,  loss=0.232,  accuracy=0.938\n",
      "Epoch  3,  step= 716,  loss=0.110,  accuracy=0.969\n",
      "Epoch  3,  step= 717,  loss=0.024,  accuracy=1.000\n",
      "Epoch  3,  step= 718,  loss=0.077,  accuracy=0.938\n",
      "Epoch  3,  step= 719,  loss=0.447,  accuracy=0.875\n",
      "Epoch  3,  step= 720,  loss=0.206,  accuracy=0.938\n",
      "Epoch  3,  step= 721,  loss=0.570,  accuracy=0.844\n",
      "Epoch  3,  step= 722,  loss=0.114,  accuracy=0.969\n",
      "Epoch  3,  step= 723,  loss=0.268,  accuracy=0.938\n",
      "Epoch  3,  step= 724,  loss=0.222,  accuracy=0.969\n",
      "Epoch  3,  step= 725,  loss=0.431,  accuracy=0.875\n",
      "Epoch  3,  step= 726,  loss=0.219,  accuracy=0.938\n",
      "Epoch  3,  step= 727,  loss=0.159,  accuracy=0.938\n",
      "Epoch  3,  step= 728,  loss=0.268,  accuracy=0.875\n",
      "Epoch  3,  step= 729,  loss=0.258,  accuracy=0.969\n",
      "Epoch  3,  step= 730,  loss=0.662,  accuracy=0.906\n",
      "Epoch  3,  step= 731,  loss=0.240,  accuracy=0.844\n",
      "Epoch  3,  step= 732,  loss=0.172,  accuracy=0.938\n",
      "Epoch  3,  step= 733,  loss=0.512,  accuracy=0.844\n",
      "Epoch  3,  step= 734,  loss=0.325,  accuracy=0.906\n",
      "Epoch  3,  step= 735,  loss=0.098,  accuracy=0.938\n",
      "Epoch  3,  step= 736,  loss=0.241,  accuracy=0.938\n",
      "Epoch  3,  step= 737,  loss=0.593,  accuracy=0.875\n",
      "Epoch  3,  step= 738,  loss=0.309,  accuracy=0.906\n",
      "Epoch  3,  step= 739,  loss=0.184,  accuracy=0.938\n",
      "Epoch  3,  step= 740,  loss=0.414,  accuracy=0.812\n",
      "Epoch  3,  step= 741,  loss=0.952,  accuracy=0.812\n",
      "Epoch  3,  step= 742,  loss=0.123,  accuracy=0.969\n",
      "Epoch  3,  step= 743,  loss=0.204,  accuracy=0.906\n",
      "Epoch  3,  step= 744,  loss=0.465,  accuracy=0.906\n",
      "Epoch  3,  step= 745,  loss=0.443,  accuracy=0.844\n",
      "Epoch  3,  step= 746,  loss=0.264,  accuracy=0.938\n",
      "Epoch  3,  step= 747,  loss=0.183,  accuracy=0.875\n",
      "Epoch  3,  step= 748,  loss=0.602,  accuracy=0.938\n",
      "Epoch  3,  step= 749,  loss=0.091,  accuracy=0.969\n",
      "Epoch  3,  step= 750,  loss=0.244,  accuracy=0.875\n",
      "Epoch  3,  step= 751,  loss=0.320,  accuracy=0.844\n",
      "Epoch  3,  step= 752,  loss=0.114,  accuracy=0.938\n",
      "Epoch  3,  step= 753,  loss=0.514,  accuracy=0.812\n",
      "Epoch  3,  step= 754,  loss=0.152,  accuracy=0.938\n",
      "Epoch  3,  step= 755,  loss=0.403,  accuracy=0.875\n",
      "Epoch  3,  step= 756,  loss=0.649,  accuracy=0.812\n",
      "Epoch  3,  step= 757,  loss=0.127,  accuracy=1.000\n",
      "Epoch  3,  step= 758,  loss=0.662,  accuracy=0.844\n",
      "Epoch  3,  step= 759,  loss=0.180,  accuracy=0.938\n",
      "Epoch  3,  step= 760,  loss=0.127,  accuracy=0.938\n",
      "Epoch  3,  step= 761,  loss=0.174,  accuracy=0.875\n",
      "Epoch  3,  step= 762,  loss=0.181,  accuracy=0.969\n",
      "Epoch  3,  step= 763,  loss=0.196,  accuracy=0.938\n",
      "Epoch  3,  step= 764,  loss=0.266,  accuracy=0.875\n",
      "Epoch  3,  step= 765,  loss=0.246,  accuracy=0.906\n",
      "Epoch  3,  step= 766,  loss=0.496,  accuracy=0.844\n",
      "Epoch  3,  step= 767,  loss=0.204,  accuracy=0.938\n",
      "Epoch  3,  step= 768,  loss=0.462,  accuracy=0.844\n",
      "Epoch  3,  step= 769,  loss=0.694,  accuracy=0.812\n",
      "Epoch  3,  step= 770,  loss=0.726,  accuracy=0.812\n",
      "Epoch  3,  step= 771,  loss=0.365,  accuracy=0.906\n",
      "Epoch  3,  step= 772,  loss=0.655,  accuracy=0.781\n",
      "Epoch  3,  step= 773,  loss=0.321,  accuracy=0.875\n",
      "Epoch  3,  step= 774,  loss=0.760,  accuracy=0.875\n",
      "Epoch  3,  step= 775,  loss=0.494,  accuracy=0.781\n",
      "Epoch  3,  step= 776,  loss=0.147,  accuracy=0.938\n",
      "Epoch  3,  step= 777,  loss=0.368,  accuracy=0.844\n",
      "Epoch  3,  step= 778,  loss=0.263,  accuracy=0.906\n",
      "Epoch  3,  step= 779,  loss=0.656,  accuracy=0.875\n",
      "Epoch  3,  step= 780,  loss=0.314,  accuracy=0.812\n",
      "Epoch  3,  step= 781,  loss=0.175,  accuracy=0.938\n",
      "Epoch  3,  step= 782,  loss=0.215,  accuracy=0.938\n",
      "Epoch  3,  step= 783,  loss=0.089,  accuracy=1.000\n",
      "Epoch  3,  step= 784,  loss=0.627,  accuracy=0.906\n",
      "Epoch  3,  step= 785,  loss=0.400,  accuracy=0.844\n",
      "Epoch  3,  step= 786,  loss=0.554,  accuracy=0.875\n",
      "Epoch  3,  step= 787,  loss=0.322,  accuracy=0.844\n",
      "Epoch  3,  step= 788,  loss=0.386,  accuracy=0.938\n",
      "Epoch  3,  step= 789,  loss=0.179,  accuracy=0.938\n",
      "Epoch  3,  step= 790,  loss=0.456,  accuracy=0.844\n",
      "Epoch  3,  step= 791,  loss=0.424,  accuracy=0.906\n",
      "Epoch  3,  step= 792,  loss=0.085,  accuracy=0.969\n",
      "Epoch  3,  step= 793,  loss=0.277,  accuracy=0.938\n",
      "Epoch  3,  step= 794,  loss=0.037,  accuracy=1.000\n",
      "Epoch  3,  step= 795,  loss=0.490,  accuracy=0.938\n",
      "Epoch  3,  step= 796,  loss=0.471,  accuracy=0.875\n",
      "Epoch  3,  step= 797,  loss=0.200,  accuracy=0.938\n",
      "Epoch  3,  step= 798,  loss=0.508,  accuracy=0.938\n",
      "Epoch  3,  step= 799,  loss=0.240,  accuracy=0.875\n",
      "Epoch  3,  step= 800,  loss=0.219,  accuracy=0.938\n",
      "Epoch  3,  step= 801,  loss=0.290,  accuracy=0.875\n",
      "Epoch  3,  step= 802,  loss=0.357,  accuracy=0.938\n",
      "Epoch  3,  step= 803,  loss=0.213,  accuracy=0.938\n",
      "Epoch  3,  step= 804,  loss=0.372,  accuracy=0.812\n",
      "Epoch  3,  step= 805,  loss=0.385,  accuracy=0.938\n",
      "Epoch  3,  step= 806,  loss=0.954,  accuracy=0.781\n",
      "Epoch  3,  step= 807,  loss=0.210,  accuracy=0.938\n",
      "Epoch  3,  step= 808,  loss=0.083,  accuracy=0.938\n",
      "Epoch  3,  step= 809,  loss=0.419,  accuracy=0.906\n",
      "Epoch  3,  step= 810,  loss=0.370,  accuracy=0.875\n",
      "Epoch  3,  step= 811,  loss=0.337,  accuracy=0.875\n",
      "Epoch  3,  step= 812,  loss=0.115,  accuracy=0.938\n",
      "Epoch  3,  step= 813,  loss=0.090,  accuracy=1.000\n",
      "Epoch  3,  step= 814,  loss=0.371,  accuracy=0.875\n",
      "Epoch  3,  step= 815,  loss=0.097,  accuracy=0.969\n",
      "Epoch  3,  step= 816,  loss=0.180,  accuracy=0.938\n",
      "Epoch  3,  step= 817,  loss=0.059,  accuracy=1.000\n",
      "Epoch  3,  step= 818,  loss=0.269,  accuracy=0.969\n",
      "Epoch  3,  step= 819,  loss=0.227,  accuracy=0.906\n",
      "Epoch  3,  step= 820,  loss=0.154,  accuracy=0.938\n",
      "Epoch  3,  step= 821,  loss=0.439,  accuracy=0.844\n",
      "Epoch  3,  step= 822,  loss=0.160,  accuracy=0.938\n",
      "Epoch  3,  step= 823,  loss=0.663,  accuracy=0.781\n",
      "Epoch  3,  step= 824,  loss=0.848,  accuracy=0.906\n",
      "Epoch  3,  step= 825,  loss=0.277,  accuracy=0.938\n",
      "Epoch  3,  step= 826,  loss=0.281,  accuracy=0.906\n",
      "Epoch  3,  step= 827,  loss=0.304,  accuracy=0.938\n",
      "Epoch  3,  step= 828,  loss=0.350,  accuracy=0.875\n",
      "Epoch  3,  step= 829,  loss=0.351,  accuracy=0.844\n",
      "Epoch  3,  step= 830,  loss=0.411,  accuracy=0.844\n",
      "Epoch  3,  step= 831,  loss=0.405,  accuracy=0.875\n",
      "Epoch  3,  step= 832,  loss=0.884,  accuracy=0.812\n",
      "Epoch  3,  step= 833,  loss=0.218,  accuracy=0.938\n",
      "Epoch  3,  step= 834,  loss=0.123,  accuracy=0.969\n",
      "Epoch  3,  step= 835,  loss=0.846,  accuracy=0.781\n",
      "Epoch  3,  step= 836,  loss=0.448,  accuracy=0.875\n",
      "Epoch  3,  step= 837,  loss=0.356,  accuracy=0.844\n",
      "Epoch  3,  step= 838,  loss=0.172,  accuracy=0.938\n",
      "Epoch  3,  step= 839,  loss=0.556,  accuracy=0.844\n",
      "Epoch  3,  step= 840,  loss=0.630,  accuracy=0.844\n",
      "Epoch  3,  step= 841,  loss=0.209,  accuracy=0.938\n",
      "Epoch  3,  step= 842,  loss=0.142,  accuracy=0.969\n",
      "Epoch  3,  step= 843,  loss=0.158,  accuracy=0.938\n",
      "Epoch  3,  step= 844,  loss=0.236,  accuracy=0.875\n",
      "Epoch  3,  step= 845,  loss=0.111,  accuracy=0.938\n",
      "Epoch  3,  step= 846,  loss=0.411,  accuracy=0.812\n",
      "Epoch  3,  step= 847,  loss=0.260,  accuracy=0.906\n",
      "Epoch  3,  step= 848,  loss=0.470,  accuracy=0.906\n",
      "Epoch  3,  step= 849,  loss=1.136,  accuracy=0.688\n",
      "Epoch  3,  step= 850,  loss=0.419,  accuracy=0.875\n",
      "Epoch  3,  step= 851,  loss=0.233,  accuracy=0.906\n",
      "Epoch  3,  step= 852,  loss=0.259,  accuracy=0.969\n",
      "Epoch  3,  step= 853,  loss=0.245,  accuracy=0.875\n",
      "Epoch  3,  step= 854,  loss=0.215,  accuracy=0.938\n",
      "Epoch  3,  step= 855,  loss=0.342,  accuracy=0.906\n",
      "Epoch  3,  step= 856,  loss=0.282,  accuracy=0.938\n",
      "Epoch  3,  step= 857,  loss=0.529,  accuracy=0.906\n",
      "Epoch  3,  step= 858,  loss=0.092,  accuracy=0.938\n",
      "Epoch  3,  step= 859,  loss=0.410,  accuracy=0.906\n",
      "Epoch  3,  step= 860,  loss=0.127,  accuracy=0.969\n",
      "Epoch  3,  step= 861,  loss=0.122,  accuracy=0.938\n",
      "Epoch  3,  step= 862,  loss=0.246,  accuracy=0.844\n",
      "Epoch  3,  step= 863,  loss=0.250,  accuracy=0.969\n",
      "Epoch  3,  step= 864,  loss=0.186,  accuracy=0.938\n",
      "Epoch  3,  step= 865,  loss=0.887,  accuracy=0.781\n",
      "Epoch  3,  step= 866,  loss=0.510,  accuracy=0.844\n",
      "Epoch  3,  step= 867,  loss=0.067,  accuracy=1.000\n",
      "Epoch  3,  step= 868,  loss=0.167,  accuracy=0.969\n",
      "Epoch  3,  step= 869,  loss=0.307,  accuracy=0.906\n",
      "Epoch  3,  step= 870,  loss=0.392,  accuracy=0.906\n",
      "Epoch  3,  step= 871,  loss=0.649,  accuracy=0.812\n",
      "Epoch  3,  step= 872,  loss=0.056,  accuracy=1.000\n",
      "Epoch  3,  step= 873,  loss=0.298,  accuracy=0.906\n",
      "Epoch  3,  step= 874,  loss=0.287,  accuracy=0.938\n",
      "Epoch  3,  step= 875,  loss=0.629,  accuracy=0.875\n",
      "Epoch  3,  step= 876,  loss=0.080,  accuracy=0.969\n",
      "Epoch  3,  step= 877,  loss=0.372,  accuracy=0.906\n",
      "Epoch  3,  step= 878,  loss=0.248,  accuracy=0.938\n",
      "Epoch  3,  step= 879,  loss=0.244,  accuracy=0.906\n",
      "Epoch  3,  step= 880,  loss=0.280,  accuracy=0.938\n",
      "Epoch  3,  step= 881,  loss=0.438,  accuracy=0.938\n",
      "Epoch  3,  step= 882,  loss=0.433,  accuracy=0.938\n",
      "Epoch  3,  step= 883,  loss=0.148,  accuracy=0.938\n",
      "Epoch  3,  step= 884,  loss=0.273,  accuracy=0.938\n",
      "Epoch  3,  step= 885,  loss=0.039,  accuracy=1.000\n",
      "Epoch  3,  step= 886,  loss=1.399,  accuracy=0.781\n",
      "Epoch  3,  step= 887,  loss=0.291,  accuracy=0.969\n",
      "Epoch  3,  step= 888,  loss=0.095,  accuracy=0.969\n",
      "Epoch  3,  step= 889,  loss=0.204,  accuracy=0.906\n",
      "Epoch  3,  step= 890,  loss=0.693,  accuracy=0.906\n",
      "Epoch  3,  step= 891,  loss=0.407,  accuracy=0.906\n",
      "Epoch  3,  step= 892,  loss=0.304,  accuracy=0.969\n",
      "Epoch  3,  step= 893,  loss=0.218,  accuracy=0.969\n",
      "Epoch  3,  step= 894,  loss=0.426,  accuracy=0.781\n",
      "Epoch  3,  step= 895,  loss=1.088,  accuracy=0.781\n",
      "Epoch  3,  step= 896,  loss=0.113,  accuracy=0.969\n",
      "Epoch  3,  step= 897,  loss=0.208,  accuracy=0.906\n",
      "Epoch  3,  step= 898,  loss=0.091,  accuracy=0.969\n",
      "Epoch  3,  step= 899,  loss=0.372,  accuracy=0.844\n",
      "Epoch  3,  step= 900,  loss=0.197,  accuracy=0.938\n",
      "Epoch  3,  step= 901,  loss=0.291,  accuracy=0.875\n",
      "Epoch  3,  step= 902,  loss=0.218,  accuracy=0.906\n",
      "Epoch  3,  step= 903,  loss=0.263,  accuracy=0.875\n",
      "Epoch  3,  step= 904,  loss=0.268,  accuracy=0.906\n",
      "Epoch  3,  step= 905,  loss=0.312,  accuracy=0.938\n",
      "Epoch  3,  step= 906,  loss=0.399,  accuracy=0.844\n",
      "Epoch  3,  step= 907,  loss=0.276,  accuracy=0.938\n",
      "Epoch  3,  step= 908,  loss=0.325,  accuracy=0.875\n",
      "Epoch  3,  step= 909,  loss=0.478,  accuracy=0.844\n",
      "Epoch  3,  step= 910,  loss=0.628,  accuracy=0.812\n",
      "Epoch  3,  step= 911,  loss=0.410,  accuracy=0.812\n",
      "Epoch  3,  step= 912,  loss=0.325,  accuracy=0.906\n",
      "Epoch  3,  step= 913,  loss=0.319,  accuracy=0.906\n",
      "Epoch  3,  step= 914,  loss=0.159,  accuracy=0.938\n",
      "Epoch  3,  step= 915,  loss=0.521,  accuracy=0.781\n",
      "Epoch  3,  step= 916,  loss=0.557,  accuracy=0.906\n",
      "Epoch  3,  step= 917,  loss=0.217,  accuracy=0.906\n",
      "Epoch  3,  step= 918,  loss=0.213,  accuracy=0.938\n",
      "Epoch  3,  step= 919,  loss=0.467,  accuracy=0.938\n",
      "Epoch  3,  step= 920,  loss=0.164,  accuracy=0.938\n",
      "Epoch  3,  step= 921,  loss=0.161,  accuracy=0.969\n",
      "Epoch  3,  step= 922,  loss=0.212,  accuracy=0.938\n",
      "Epoch  3,  step= 923,  loss=0.145,  accuracy=0.938\n",
      "Epoch  3,  step= 924,  loss=0.517,  accuracy=0.906\n",
      "Epoch  3,  step= 925,  loss=0.156,  accuracy=0.938\n",
      "Epoch  3,  step= 926,  loss=0.249,  accuracy=0.906\n",
      "Epoch  3,  step= 927,  loss=0.295,  accuracy=0.875\n",
      "Epoch  3,  step= 928,  loss=0.577,  accuracy=0.844\n",
      "Epoch  3,  step= 929,  loss=0.386,  accuracy=0.906\n",
      "Epoch  3,  step= 930,  loss=0.153,  accuracy=0.969\n",
      "Epoch  3,  step= 931,  loss=0.165,  accuracy=0.969\n",
      "Epoch  3,  step= 932,  loss=0.654,  accuracy=0.781\n",
      "Epoch  3,  step= 933,  loss=0.476,  accuracy=0.906\n",
      "Epoch  3,  step= 934,  loss=0.169,  accuracy=0.969\n",
      "Epoch  3,  step= 935,  loss=0.572,  accuracy=0.844\n",
      "Epoch  3,  step= 936,  loss=0.512,  accuracy=0.812\n",
      "Epoch  3,  step= 937,  loss=0.284,  accuracy=0.906\n",
      "Epoch  3,  step= 938,  loss=0.299,  accuracy=0.875\n",
      "Epoch  3,  step= 939,  loss=0.448,  accuracy=0.906\n",
      "Epoch  3,  step= 940,  loss=0.460,  accuracy=0.875\n",
      "Epoch  3,  step= 941,  loss=0.605,  accuracy=0.875\n",
      "Epoch  3,  step= 942,  loss=0.499,  accuracy=0.812\n",
      "Epoch  3,  step= 943,  loss=0.453,  accuracy=0.875\n",
      "Epoch  3,  step= 944,  loss=0.324,  accuracy=0.969\n",
      "Epoch  3,  step= 945,  loss=0.232,  accuracy=0.938\n",
      "Epoch  3,  step= 946,  loss=0.069,  accuracy=0.969\n",
      "Epoch  3,  step= 947,  loss=0.319,  accuracy=0.906\n",
      "Epoch  3,  step= 948,  loss=0.191,  accuracy=0.906\n",
      "Epoch  3,  step= 949,  loss=0.221,  accuracy=0.938\n",
      "Epoch  3,  step= 950,  loss=0.240,  accuracy=0.938\n",
      "Epoch  3,  step= 951,  loss=0.467,  accuracy=0.812\n",
      "Epoch  3,  step= 952,  loss=0.229,  accuracy=0.938\n",
      "Epoch  3,  step= 953,  loss=0.439,  accuracy=0.781\n",
      "Epoch  3,  step= 954,  loss=0.549,  accuracy=0.875\n",
      "Epoch  3,  step= 955,  loss=0.289,  accuracy=0.938\n",
      "Epoch  3,  step= 956,  loss=0.770,  accuracy=0.812\n",
      "Epoch  3,  step= 957,  loss=0.215,  accuracy=0.906\n",
      "Epoch  3,  step= 958,  loss=0.333,  accuracy=0.844\n",
      "Epoch  3,  step= 959,  loss=0.366,  accuracy=0.844\n",
      "Epoch  3,  step= 960,  loss=0.428,  accuracy=0.906\n",
      "Epoch  3,  step= 961,  loss=0.088,  accuracy=0.969\n",
      "Epoch  3,  step= 962,  loss=0.355,  accuracy=0.875\n",
      "Epoch  3,  step= 963,  loss=0.522,  accuracy=0.844\n",
      "Epoch  3,  step= 964,  loss=0.306,  accuracy=0.906\n",
      "Epoch  3,  step= 965,  loss=0.716,  accuracy=0.812\n",
      "Epoch  3,  step= 966,  loss=0.127,  accuracy=0.938\n",
      "Epoch  3,  step= 967,  loss=0.564,  accuracy=0.875\n",
      "Epoch  3,  step= 968,  loss=0.288,  accuracy=0.906\n",
      "Epoch  3,  step= 969,  loss=0.279,  accuracy=0.969\n",
      "Epoch  3,  step= 970,  loss=0.509,  accuracy=0.938\n",
      "Epoch  3,  step= 971,  loss=0.623,  accuracy=0.875\n",
      "Epoch  3,  step= 972,  loss=0.681,  accuracy=0.844\n",
      "Epoch  3,  step= 973,  loss=0.370,  accuracy=0.906\n",
      "Epoch  3,  step= 974,  loss=0.791,  accuracy=0.875\n",
      "Epoch  3,  step= 975,  loss=0.129,  accuracy=0.938\n",
      "Epoch  3,  step= 976,  loss=0.200,  accuracy=0.938\n",
      "Epoch  3,  step= 977,  loss=0.841,  accuracy=0.719\n",
      "Epoch  3,  step= 978,  loss=0.867,  accuracy=0.750\n",
      "Epoch  3,  step= 979,  loss=0.499,  accuracy=0.812\n",
      "Epoch  3,  step= 980,  loss=0.643,  accuracy=0.812\n",
      "Epoch  3,  step= 981,  loss=0.906,  accuracy=0.844\n",
      "Epoch  3,  step= 982,  loss=0.435,  accuracy=0.875\n",
      "Epoch  3,  step= 983,  loss=0.223,  accuracy=0.906\n",
      "Epoch  3,  step= 984,  loss=0.187,  accuracy=0.906\n",
      "Epoch  3,  step= 985,  loss=0.138,  accuracy=0.969\n",
      "Epoch  3,  step= 986,  loss=0.277,  accuracy=0.906\n",
      "Epoch  3,  step= 987,  loss=0.362,  accuracy=0.906\n",
      "Epoch  3,  step= 988,  loss=0.407,  accuracy=0.844\n",
      "Epoch  3,  step= 989,  loss=1.140,  accuracy=0.750\n",
      "Epoch  3,  step= 990,  loss=0.302,  accuracy=0.844\n",
      "Epoch  3,  step= 991,  loss=0.735,  accuracy=0.875\n",
      "Epoch  3,  step= 992,  loss=0.210,  accuracy=0.938\n",
      "Epoch  3,  step= 993,  loss=0.275,  accuracy=0.844\n",
      "Epoch  3,  step= 994,  loss=0.142,  accuracy=0.938\n",
      "Epoch  3,  step= 995,  loss=0.545,  accuracy=0.844\n",
      "Epoch  3,  step= 996,  loss=0.191,  accuracy=0.938\n",
      "Epoch  3,  step= 997,  loss=0.316,  accuracy=0.938\n",
      "Epoch  3,  step= 998,  loss=0.471,  accuracy=0.906\n",
      "Epoch  3,  step= 999,  loss=0.195,  accuracy=0.969\n",
      "Epoch  3,  step=1000,  loss=0.331,  accuracy=0.938\n",
      "Epoch  3,  step=1001,  loss=0.301,  accuracy=0.844\n",
      "Epoch  3,  step=1002,  loss=0.223,  accuracy=0.906\n",
      "Epoch  3,  step=1003,  loss=0.157,  accuracy=0.938\n",
      "Epoch  3,  step=1004,  loss=0.314,  accuracy=0.906\n",
      "Epoch  3,  step=1005,  loss=0.548,  accuracy=0.875\n",
      "Epoch  3,  step=1006,  loss=0.358,  accuracy=0.906\n",
      "Epoch  3,  step=1007,  loss=0.429,  accuracy=0.844\n",
      "Epoch  3,  step=1008,  loss=0.333,  accuracy=0.969\n",
      "Epoch  3,  step=1009,  loss=0.402,  accuracy=0.875\n",
      "Epoch  3,  step=1010,  loss=0.526,  accuracy=0.812\n",
      "Epoch  3,  step=1011,  loss=0.576,  accuracy=0.906\n",
      "Epoch  3,  step=1012,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step=1013,  loss=0.340,  accuracy=0.906\n",
      "Epoch  3,  step=1014,  loss=0.339,  accuracy=0.906\n",
      "Epoch  3,  step=1015,  loss=0.240,  accuracy=0.938\n",
      "Epoch  3,  step=1016,  loss=0.148,  accuracy=0.938\n",
      "Epoch  3,  step=1017,  loss=0.398,  accuracy=0.906\n",
      "Epoch  3,  step=1018,  loss=0.157,  accuracy=0.938\n",
      "Epoch  3,  step=1019,  loss=0.229,  accuracy=0.906\n",
      "Epoch  3,  step=1020,  loss=0.423,  accuracy=0.875\n",
      "Epoch  3,  step=1021,  loss=0.283,  accuracy=0.938\n",
      "Epoch  3,  step=1022,  loss=0.154,  accuracy=0.938\n",
      "Epoch  3,  step=1023,  loss=0.117,  accuracy=0.906\n",
      "Epoch  3,  step=1024,  loss=0.676,  accuracy=0.875\n",
      "Epoch  3,  step=1025,  loss=0.440,  accuracy=0.875\n",
      "Epoch  3,  step=1026,  loss=0.424,  accuracy=0.812\n",
      "Epoch  3,  step=1027,  loss=0.529,  accuracy=0.875\n",
      "Epoch  3,  step=1028,  loss=0.538,  accuracy=0.875\n",
      "Epoch  3,  step=1029,  loss=0.193,  accuracy=0.969\n",
      "Epoch  3,  step=1030,  loss=0.065,  accuracy=0.969\n",
      "Epoch  3,  step=1031,  loss=0.082,  accuracy=0.969\n",
      "Epoch  3,  step=1032,  loss=0.367,  accuracy=0.875\n",
      "Epoch  3,  step=1033,  loss=0.494,  accuracy=0.938\n",
      "Epoch  3,  step=1034,  loss=0.159,  accuracy=0.906\n",
      "Epoch  3,  step=1035,  loss=0.332,  accuracy=0.906\n",
      "Epoch  3,  step=1036,  loss=0.238,  accuracy=0.938\n",
      "Epoch  3,  step=1037,  loss=0.258,  accuracy=0.906\n",
      "Epoch  3,  step=1038,  loss=0.466,  accuracy=0.906\n",
      "Epoch  3,  step=1039,  loss=0.160,  accuracy=0.906\n",
      "Epoch  3,  step=1040,  loss=0.453,  accuracy=0.844\n",
      "Epoch  3,  step=1041,  loss=0.192,  accuracy=0.969\n",
      "Epoch  3,  step=1042,  loss=0.299,  accuracy=0.844\n",
      "Epoch  3,  step=1043,  loss=0.261,  accuracy=0.938\n",
      "Epoch  3,  step=1044,  loss=0.221,  accuracy=0.906\n",
      "Epoch  3,  step=1045,  loss=0.253,  accuracy=0.875\n",
      "Epoch  3,  step=1046,  loss=0.225,  accuracy=0.906\n",
      "Epoch  3,  step=1047,  loss=0.305,  accuracy=0.875\n",
      "Epoch  3,  step=1048,  loss=0.049,  accuracy=1.000\n",
      "Epoch  3,  step=1049,  loss=0.716,  accuracy=0.812\n",
      "Epoch  3,  step=1050,  loss=0.155,  accuracy=0.938\n",
      "Epoch  3,  step=1051,  loss=0.213,  accuracy=0.906\n",
      "Epoch  3,  step=1052,  loss=0.340,  accuracy=0.906\n",
      "Epoch  3,  step=1053,  loss=0.099,  accuracy=0.969\n",
      "Epoch  3,  step=1054,  loss=0.275,  accuracy=0.969\n",
      "Epoch  3,  step=1055,  loss=0.712,  accuracy=0.781\n",
      "Epoch  3,  step=1056,  loss=0.304,  accuracy=0.906\n",
      "Epoch  3,  step=1057,  loss=0.137,  accuracy=0.969\n",
      "Epoch  3,  step=1058,  loss=0.065,  accuracy=0.969\n",
      "Epoch  3,  step=1059,  loss=0.132,  accuracy=0.969\n",
      "Epoch  3,  step=1060,  loss=0.153,  accuracy=0.938\n",
      "Epoch  3,  step=1061,  loss=0.103,  accuracy=0.969\n",
      "Epoch  3,  step=1062,  loss=0.258,  accuracy=0.938\n",
      "Epoch  3,  step=1063,  loss=0.162,  accuracy=0.938\n",
      "Epoch  3,  step=1064,  loss=0.473,  accuracy=0.875\n",
      "Epoch  3,  step=1065,  loss=0.437,  accuracy=0.812\n",
      "Epoch  3,  step=1066,  loss=0.470,  accuracy=0.906\n",
      "Epoch  3,  step=1067,  loss=0.061,  accuracy=0.969\n",
      "Epoch  3,  step=1068,  loss=0.312,  accuracy=0.938\n",
      "Epoch  3,  step=1069,  loss=0.113,  accuracy=0.969\n",
      "Epoch  3,  step=1070,  loss=0.133,  accuracy=0.969\n",
      "Epoch  3,  step=1071,  loss=0.073,  accuracy=0.969\n",
      "Epoch  3,  step=1072,  loss=0.205,  accuracy=0.938\n",
      "Epoch  3,  step=1073,  loss=0.268,  accuracy=0.875\n",
      "Epoch  3,  step=1074,  loss=0.071,  accuracy=0.969\n",
      "Epoch  3,  step=1075,  loss=0.442,  accuracy=0.812\n",
      "Epoch  3,  step=1076,  loss=0.482,  accuracy=0.875\n",
      "Epoch  3,  step=1077,  loss=0.204,  accuracy=0.938\n",
      "Epoch  3,  step=1078,  loss=0.469,  accuracy=0.875\n",
      "Epoch  3,  step=1079,  loss=1.026,  accuracy=0.812\n",
      "Epoch  3,  step=1080,  loss=0.174,  accuracy=0.938\n",
      "Epoch  3,  step=1081,  loss=0.291,  accuracy=0.875\n",
      "Epoch  3,  step=1082,  loss=0.230,  accuracy=0.906\n",
      "Epoch  3,  step=1083,  loss=0.478,  accuracy=0.875\n",
      "Epoch  3,  step=1084,  loss=0.462,  accuracy=0.844\n",
      "Epoch  3,  step=1085,  loss=0.285,  accuracy=0.906\n",
      "Epoch  3,  step=1086,  loss=0.224,  accuracy=0.938\n",
      "Epoch  3,  step=1087,  loss=0.377,  accuracy=0.875\n",
      "Epoch  3,  step=1088,  loss=0.546,  accuracy=0.812\n",
      "Epoch  3,  step=1089,  loss=0.277,  accuracy=0.906\n",
      "Epoch  3,  step=1090,  loss=0.402,  accuracy=0.906\n",
      "Epoch  3,  step=1091,  loss=0.219,  accuracy=0.875\n",
      "Epoch  3,  step=1092,  loss=0.143,  accuracy=0.938\n",
      "Epoch  3,  step=1093,  loss=0.134,  accuracy=0.938\n",
      "Epoch  3,  step=1094,  loss=0.114,  accuracy=0.938\n",
      "Epoch  3,  step=1095,  loss=0.333,  accuracy=0.844\n",
      "Epoch  3,  step=1096,  loss=0.265,  accuracy=0.938\n",
      "Epoch  3,  step=1097,  loss=0.302,  accuracy=0.875\n",
      "Epoch  3,  step=1098,  loss=0.491,  accuracy=0.812\n",
      "Epoch  3,  step=1099,  loss=0.243,  accuracy=0.938\n",
      "Epoch  3,  step=1100,  loss=0.308,  accuracy=0.938\n",
      "Epoch  3,  step=1101,  loss=0.566,  accuracy=0.750\n",
      "Epoch  3,  step=1102,  loss=0.075,  accuracy=0.969\n",
      "Epoch  3,  step=1103,  loss=0.211,  accuracy=0.875\n",
      "Epoch  3,  step=1104,  loss=0.099,  accuracy=0.969\n",
      "Epoch  3,  step=1105,  loss=0.248,  accuracy=0.969\n",
      "Epoch  3,  step=1106,  loss=0.193,  accuracy=0.938\n",
      "Epoch  3,  step=1107,  loss=0.327,  accuracy=0.906\n",
      "Epoch  3,  step=1108,  loss=1.016,  accuracy=0.812\n",
      "Epoch  3,  step=1109,  loss=0.298,  accuracy=0.938\n",
      "Epoch  3,  step=1110,  loss=0.190,  accuracy=0.938\n",
      "Epoch  3,  step=1111,  loss=0.310,  accuracy=0.875\n",
      "Epoch  3,  step=1112,  loss=0.102,  accuracy=0.969\n",
      "Epoch  3,  step=1113,  loss=0.780,  accuracy=0.719\n",
      "Epoch  3,  step=1114,  loss=0.092,  accuracy=0.969\n",
      "Epoch  3,  step=1115,  loss=0.338,  accuracy=0.938\n",
      "Epoch  3,  step=1116,  loss=0.093,  accuracy=1.000\n",
      "Epoch  3,  step=1117,  loss=0.141,  accuracy=0.938\n",
      "Epoch  3,  step=1118,  loss=0.154,  accuracy=0.969\n",
      "Epoch  3,  step=1119,  loss=0.171,  accuracy=0.969\n",
      "Epoch  3,  step=1120,  loss=0.402,  accuracy=0.938\n",
      "Epoch  3,  step=1121,  loss=0.167,  accuracy=0.938\n",
      "Epoch  3,  step=1122,  loss=0.235,  accuracy=0.938\n",
      "Epoch  3,  step=1123,  loss=0.223,  accuracy=0.875\n",
      "Epoch  3,  step=1124,  loss=0.476,  accuracy=0.938\n",
      "Epoch  3,  step=1125,  loss=0.444,  accuracy=0.812\n",
      "Epoch  3,  step=1126,  loss=0.629,  accuracy=0.875\n",
      "Epoch  3,  step=1127,  loss=0.424,  accuracy=0.844\n",
      "Epoch  3,  step=1128,  loss=0.680,  accuracy=0.875\n",
      "Epoch  3,  step=1129,  loss=0.170,  accuracy=0.938\n",
      "Epoch  3,  step=1130,  loss=0.351,  accuracy=0.906\n",
      "Epoch  3,  step=1131,  loss=0.199,  accuracy=0.875\n",
      "Epoch  3,  step=1132,  loss=0.258,  accuracy=0.875\n",
      "Epoch  3,  step=1133,  loss=0.271,  accuracy=0.969\n",
      "Epoch  3,  step=1134,  loss=0.033,  accuracy=1.000\n",
      "Epoch  3,  step=1135,  loss=0.241,  accuracy=0.844\n",
      "Epoch  3,  step=1136,  loss=0.102,  accuracy=0.969\n",
      "Epoch  3,  step=1137,  loss=0.242,  accuracy=0.906\n",
      "Epoch  3,  step=1138,  loss=0.614,  accuracy=0.844\n",
      "Epoch  3,  step=1139,  loss=0.455,  accuracy=0.812\n",
      "Epoch  3,  step=1140,  loss=0.323,  accuracy=0.875\n",
      "Epoch  3,  step=1141,  loss=0.306,  accuracy=0.875\n",
      "Epoch  3,  step=1142,  loss=0.176,  accuracy=0.938\n",
      "Epoch  3,  step=1143,  loss=0.204,  accuracy=0.906\n",
      "Epoch  3,  step=1144,  loss=0.052,  accuracy=0.969\n",
      "Epoch  3,  step=1145,  loss=0.430,  accuracy=0.875\n",
      "Epoch  3,  step=1146,  loss=0.075,  accuracy=0.969\n",
      "Epoch  3,  step=1147,  loss=0.325,  accuracy=0.938\n",
      "Epoch  3,  step=1148,  loss=0.246,  accuracy=0.875\n",
      "Epoch  3,  step=1149,  loss=0.122,  accuracy=0.969\n",
      "Epoch  3,  step=1150,  loss=0.402,  accuracy=0.906\n",
      "Epoch  3,  step=1151,  loss=0.171,  accuracy=0.938\n",
      "Epoch  3,  step=1152,  loss=0.298,  accuracy=0.938\n",
      "Epoch  3,  step=1153,  loss=0.107,  accuracy=0.969\n",
      "Epoch  3,  step=1154,  loss=0.108,  accuracy=0.969\n",
      "Epoch  3,  step=1155,  loss=0.104,  accuracy=0.938\n",
      "Epoch  3,  step=1156,  loss=0.079,  accuracy=0.969\n",
      "Epoch  3,  step=1157,  loss=0.499,  accuracy=0.906\n",
      "Epoch  3,  step=1158,  loss=0.426,  accuracy=0.938\n",
      "Epoch  3,  step=1159,  loss=0.355,  accuracy=0.906\n",
      "Epoch  3,  step=1160,  loss=0.334,  accuracy=0.875\n",
      "Epoch  3,  step=1161,  loss=0.351,  accuracy=0.906\n",
      "Epoch  3,  step=1162,  loss=0.306,  accuracy=0.875\n",
      "Epoch  3,  step=1163,  loss=0.094,  accuracy=0.969\n",
      "Epoch  3,  step=1164,  loss=0.673,  accuracy=0.875\n",
      "Epoch  3,  step=1165,  loss=0.348,  accuracy=0.875\n",
      "Epoch  3,  step=1166,  loss=0.265,  accuracy=0.844\n",
      "Epoch  3,  step=1167,  loss=0.701,  accuracy=0.875\n",
      "Epoch  3,  step=1168,  loss=1.092,  accuracy=0.906\n",
      "Epoch  3,  step=1169,  loss=0.709,  accuracy=0.844\n",
      "Epoch  3,  step=1170,  loss=0.673,  accuracy=0.844\n",
      "Epoch  3,  step=1171,  loss=0.220,  accuracy=0.938\n",
      "Epoch  3,  step=1172,  loss=0.285,  accuracy=0.844\n",
      "Epoch  3,  step=1173,  loss=0.623,  accuracy=0.781\n",
      "Epoch  3,  step=1174,  loss=0.485,  accuracy=0.750\n",
      "Epoch  3,  step=1175,  loss=0.406,  accuracy=0.844\n",
      "Epoch  3,  step=1176,  loss=0.242,  accuracy=0.875\n",
      "Epoch  3,  step=1177,  loss=0.217,  accuracy=0.906\n",
      "Epoch  3,  step=1178,  loss=0.544,  accuracy=0.844\n",
      "Epoch  3,  step=1179,  loss=0.611,  accuracy=0.875\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  3,  step=1180,  loss=0.307,  accuracy=0.875\n",
      "Epoch  3,  step=1181,  loss=0.637,  accuracy=0.906\n",
      "Epoch  3,  step=1182,  loss=0.516,  accuracy=0.906\n",
      "Epoch  3,  step=1183,  loss=0.212,  accuracy=0.906\n",
      "Epoch  3,  step=1184,  loss=0.306,  accuracy=0.812\n",
      "Epoch  3,  step=1185,  loss=0.361,  accuracy=0.906\n",
      "Epoch  3,  step=1186,  loss=0.213,  accuracy=0.938\n",
      "Epoch  3,  step=1187,  loss=0.277,  accuracy=0.906\n",
      "Epoch  3,  step=1188,  loss=0.346,  accuracy=0.969\n",
      "Epoch  3,  step=1189,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step=1190,  loss=0.266,  accuracy=0.938\n",
      "Epoch  3,  step=1191,  loss=0.324,  accuracy=0.906\n",
      "Epoch  3,  step=1192,  loss=0.225,  accuracy=0.906\n",
      "Epoch  3,  step=1193,  loss=0.144,  accuracy=0.969\n",
      "Epoch  3,  step=1194,  loss=0.266,  accuracy=0.938\n",
      "Epoch  3,  step=1195,  loss=0.514,  accuracy=0.844\n",
      "Epoch  3,  step=1196,  loss=0.137,  accuracy=0.969\n",
      "Epoch  3,  step=1197,  loss=0.185,  accuracy=0.875\n",
      "Epoch  3,  step=1198,  loss=0.144,  accuracy=0.969\n",
      "Epoch  3,  step=1199,  loss=0.738,  accuracy=0.750\n",
      "Epoch  3,  step=1200,  loss=0.215,  accuracy=0.938\n",
      "Epoch  3,  step=1201,  loss=0.321,  accuracy=0.844\n",
      "Epoch  3,  step=1202,  loss=0.060,  accuracy=0.969\n",
      "Epoch  3,  step=1203,  loss=0.782,  accuracy=0.844\n",
      "Epoch  3,  step=1204,  loss=0.148,  accuracy=0.938\n",
      "Epoch  3,  step=1205,  loss=0.351,  accuracy=0.906\n",
      "Epoch  3,  step=1206,  loss=0.616,  accuracy=0.844\n",
      "Epoch  3,  step=1207,  loss=0.235,  accuracy=0.938\n",
      "Epoch  3,  step=1208,  loss=0.356,  accuracy=0.906\n",
      "Epoch  3,  step=1209,  loss=0.167,  accuracy=0.969\n",
      "Epoch  3,  step=1210,  loss=0.152,  accuracy=0.938\n",
      "Epoch  3,  step=1211,  loss=0.473,  accuracy=0.844\n",
      "Epoch  3,  step=1212,  loss=0.303,  accuracy=0.875\n",
      "Epoch  3,  step=1213,  loss=0.056,  accuracy=0.969\n",
      "Epoch  3,  step=1214,  loss=0.065,  accuracy=1.000\n",
      "Epoch  3,  step=1215,  loss=0.157,  accuracy=0.938\n",
      "Epoch  3,  step=1216,  loss=0.186,  accuracy=0.906\n",
      "Epoch  3,  step=1217,  loss=0.141,  accuracy=0.969\n",
      "Epoch  3,  step=1218,  loss=0.314,  accuracy=0.875\n",
      "Epoch  3,  step=1219,  loss=0.132,  accuracy=0.969\n",
      "Epoch  3,  step=1220,  loss=0.251,  accuracy=0.906\n",
      "Epoch  3,  step=1221,  loss=0.082,  accuracy=0.969\n",
      "Epoch  3,  step=1222,  loss=0.523,  accuracy=0.844\n",
      "Epoch  3,  step=1223,  loss=0.339,  accuracy=0.875\n",
      "Epoch  3,  step=1224,  loss=0.232,  accuracy=0.969\n",
      "Epoch  3,  step=1225,  loss=0.431,  accuracy=0.938\n",
      "Epoch  3,  step=1226,  loss=0.426,  accuracy=0.906\n",
      "Epoch  3,  step=1227,  loss=0.268,  accuracy=0.969\n",
      "Epoch  3,  step=1228,  loss=1.144,  accuracy=0.812\n",
      "Epoch  3,  step=1229,  loss=0.662,  accuracy=0.781\n",
      "Epoch  3,  step=1230,  loss=0.275,  accuracy=0.906\n",
      "Epoch  3,  step=1231,  loss=0.735,  accuracy=0.844\n",
      "Epoch  3,  step=1232,  loss=0.306,  accuracy=0.969\n",
      "Epoch  3,  step=1233,  loss=0.449,  accuracy=0.938\n",
      "Epoch  3,  step=1234,  loss=0.210,  accuracy=0.938\n",
      "Epoch  3,  step=1235,  loss=0.374,  accuracy=0.875\n",
      "Epoch  3,  step=1236,  loss=0.307,  accuracy=0.938\n",
      "Epoch  3,  step=1237,  loss=0.488,  accuracy=0.844\n",
      "Epoch  3,  step=1238,  loss=0.293,  accuracy=0.906\n",
      "Epoch  3,  step=1239,  loss=0.726,  accuracy=0.875\n",
      "Epoch  3,  step=1240,  loss=0.649,  accuracy=0.844\n",
      "Epoch  3,  step=1241,  loss=0.223,  accuracy=0.906\n",
      "Epoch  3,  step=1242,  loss=0.532,  accuracy=0.812\n",
      "Epoch  3,  step=1243,  loss=0.547,  accuracy=0.812\n",
      "Epoch  3,  step=1244,  loss=0.258,  accuracy=0.938\n",
      "Epoch  3,  step=1245,  loss=0.384,  accuracy=0.906\n",
      "Epoch  3,  step=1246,  loss=0.344,  accuracy=0.875\n",
      "Epoch  3,  step=1247,  loss=0.489,  accuracy=0.875\n",
      "Epoch  3,  step=1248,  loss=0.396,  accuracy=0.906\n",
      "Epoch  3,  step=1249,  loss=0.386,  accuracy=0.875\n",
      "Epoch  3,  step=1250,  loss=0.312,  accuracy=0.875\n",
      "Epoch  3,  step=1251,  loss=0.249,  accuracy=0.906\n",
      "Epoch  3,  step=1252,  loss=0.271,  accuracy=0.844\n",
      "Epoch  3,  step=1253,  loss=0.306,  accuracy=0.875\n",
      "Epoch  3,  step=1254,  loss=0.317,  accuracy=0.875\n",
      "Epoch  3,  step=1255,  loss=0.315,  accuracy=0.906\n",
      "Epoch  3,  step=1256,  loss=0.255,  accuracy=0.906\n",
      "Epoch  3,  step=1257,  loss=0.474,  accuracy=0.875\n",
      "Epoch  3,  step=1258,  loss=0.121,  accuracy=0.969\n",
      "Epoch  3,  step=1259,  loss=0.432,  accuracy=0.875\n",
      "Epoch  3,  step=1260,  loss=0.196,  accuracy=0.906\n",
      "Epoch  3,  step=1261,  loss=0.054,  accuracy=1.000\n",
      "Epoch  3,  step=1262,  loss=0.195,  accuracy=0.938\n",
      "Epoch  3,  step=1263,  loss=0.221,  accuracy=0.844\n",
      "Epoch  3,  step=1264,  loss=0.118,  accuracy=0.969\n",
      "Epoch  3,  step=1265,  loss=0.534,  accuracy=0.875\n",
      "Epoch  3,  step=1266,  loss=0.322,  accuracy=0.906\n",
      "Epoch  3,  step=1267,  loss=0.225,  accuracy=0.906\n",
      "Epoch  3,  step=1268,  loss=0.235,  accuracy=0.875\n",
      "Epoch  3,  step=1269,  loss=0.257,  accuracy=0.938\n",
      "Epoch  3,  step=1270,  loss=0.588,  accuracy=0.781\n",
      "Epoch  3,  step=1271,  loss=0.082,  accuracy=1.000\n",
      "Epoch  3,  step=1272,  loss=0.257,  accuracy=0.938\n",
      "Epoch  3,  step=1273,  loss=0.241,  accuracy=0.906\n",
      "Epoch  3,  step=1274,  loss=0.107,  accuracy=0.969\n",
      "Epoch  3,  step=1275,  loss=0.100,  accuracy=1.000\n",
      "Epoch  3,  step=1276,  loss=0.149,  accuracy=0.969\n",
      "Epoch  3,  step=1277,  loss=0.188,  accuracy=0.969\n",
      "Epoch  3,  step=1278,  loss=0.060,  accuracy=0.969\n",
      "Epoch  3,  step=1279,  loss=0.132,  accuracy=0.969\n",
      "Epoch  3,  step=1280,  loss=0.471,  accuracy=0.906\n",
      "Epoch  3,  step=1281,  loss=0.552,  accuracy=0.781\n",
      "Epoch  3,  step=1282,  loss=0.171,  accuracy=0.969\n",
      "Epoch  3,  step=1283,  loss=0.199,  accuracy=0.906\n",
      "Epoch  3,  step=1284,  loss=0.244,  accuracy=0.844\n",
      "Epoch  3,  step=1285,  loss=0.149,  accuracy=0.969\n",
      "Epoch  3,  step=1286,  loss=0.294,  accuracy=0.875\n",
      "Epoch  3,  step=1287,  loss=0.232,  accuracy=0.906\n",
      "Epoch  3,  step=1288,  loss=0.179,  accuracy=0.938\n",
      "Epoch  3,  step=1289,  loss=0.715,  accuracy=0.875\n",
      "Epoch  3,  step=1290,  loss=0.701,  accuracy=0.844\n",
      "Epoch  3,  step=1291,  loss=0.412,  accuracy=0.938\n",
      "Epoch  3,  step=1292,  loss=0.224,  accuracy=0.906\n",
      "Epoch  3,  step=1293,  loss=0.306,  accuracy=0.938\n",
      "Epoch  3,  step=1294,  loss=0.321,  accuracy=0.906\n",
      "Epoch  3,  step=1295,  loss=0.471,  accuracy=0.875\n",
      "Epoch  3,  step=1296,  loss=0.246,  accuracy=0.906\n",
      "Epoch  3,  step=1297,  loss=0.282,  accuracy=0.906\n",
      "Epoch  3,  step=1298,  loss=0.428,  accuracy=0.875\n",
      "Epoch  3,  step=1299,  loss=0.507,  accuracy=0.875\n",
      "Epoch  3,  step=1300,  loss=0.285,  accuracy=0.906\n",
      "Epoch  3,  step=1301,  loss=0.033,  accuracy=1.000\n",
      "Epoch  3,  step=1302,  loss=0.146,  accuracy=0.969\n",
      "Epoch  3,  step=1303,  loss=0.236,  accuracy=0.906\n",
      "Epoch  3,  step=1304,  loss=0.488,  accuracy=0.875\n",
      "Epoch  3,  step=1305,  loss=0.294,  accuracy=0.906\n",
      "Epoch  3,  step=1306,  loss=0.380,  accuracy=0.844\n",
      "Epoch  3,  step=1307,  loss=0.252,  accuracy=0.906\n",
      "Epoch  3,  step=1308,  loss=0.257,  accuracy=0.938\n",
      "Epoch  3,  step=1309,  loss=0.427,  accuracy=0.844\n",
      "Epoch  3,  step=1310,  loss=0.457,  accuracy=0.906\n",
      "Epoch  3,  step=1311,  loss=0.443,  accuracy=0.875\n",
      "Epoch  3,  step=1312,  loss=0.519,  accuracy=0.875\n",
      "Epoch  3,  step=1313,  loss=0.385,  accuracy=0.844\n",
      "Epoch  3,  step=1314,  loss=0.334,  accuracy=0.938\n",
      "Epoch  3,  step=1315,  loss=0.321,  accuracy=0.938\n",
      "Epoch  3,  step=1316,  loss=0.298,  accuracy=0.844\n",
      "Epoch  3,  step=1317,  loss=0.142,  accuracy=0.938\n",
      "Epoch  3,  step=1318,  loss=0.335,  accuracy=0.906\n",
      "Epoch  3,  step=1319,  loss=0.273,  accuracy=0.906\n",
      "Epoch  3,  step=1320,  loss=0.275,  accuracy=0.875\n",
      "Epoch  3,  step=1321,  loss=0.088,  accuracy=0.969\n",
      "Epoch  3,  step=1322,  loss=0.771,  accuracy=0.812\n",
      "Epoch  3,  step=1323,  loss=0.770,  accuracy=0.844\n",
      "Epoch  3,  step=1324,  loss=0.169,  accuracy=0.938\n",
      "Epoch  3,  step=1325,  loss=0.475,  accuracy=0.906\n",
      "Epoch  3,  step=1326,  loss=0.350,  accuracy=0.906\n",
      "Epoch  3,  step=1327,  loss=0.280,  accuracy=0.875\n",
      "Epoch  3,  step=1328,  loss=0.704,  accuracy=0.844\n",
      "Epoch  3,  step=1329,  loss=0.212,  accuracy=0.906\n",
      "Epoch  3,  step=1330,  loss=0.198,  accuracy=0.969\n",
      "Epoch  3,  step=1331,  loss=0.166,  accuracy=0.875\n",
      "Epoch  3,  step=1332,  loss=0.297,  accuracy=0.875\n",
      "Epoch  3,  step=1333,  loss=0.423,  accuracy=0.844\n",
      "Epoch  3,  step=1334,  loss=0.544,  accuracy=0.844\n",
      "Epoch  3,  step=1335,  loss=0.116,  accuracy=0.969\n",
      "Epoch  3,  step=1336,  loss=0.448,  accuracy=0.906\n",
      "Epoch  3,  step=1337,  loss=0.463,  accuracy=0.875\n",
      "Epoch  3,  step=1338,  loss=0.710,  accuracy=0.906\n",
      "Epoch  3,  step=1339,  loss=0.702,  accuracy=0.875\n",
      "Epoch  3,  step=1340,  loss=0.403,  accuracy=0.875\n",
      "Epoch  3,  step=1341,  loss=0.379,  accuracy=0.938\n",
      "Epoch  3,  step=1342,  loss=0.370,  accuracy=0.906\n",
      "Epoch  3,  step=1343,  loss=0.219,  accuracy=0.875\n",
      "Epoch  3,  step=1344,  loss=0.600,  accuracy=0.812\n",
      "Epoch  3,  step=1345,  loss=0.298,  accuracy=0.906\n",
      "Epoch  3,  step=1346,  loss=0.552,  accuracy=0.906\n",
      "Epoch  3,  step=1347,  loss=0.123,  accuracy=0.969\n",
      "Epoch  3,  step=1348,  loss=0.094,  accuracy=0.969\n",
      "Epoch  3,  step=1349,  loss=0.305,  accuracy=0.875\n",
      "Epoch  3,  step=1350,  loss=0.750,  accuracy=0.719\n",
      "Epoch  3,  step=1351,  loss=0.154,  accuracy=0.938\n",
      "Epoch  3,  step=1352,  loss=0.218,  accuracy=0.938\n",
      "Epoch  3,  step=1353,  loss=0.042,  accuracy=0.969\n",
      "Epoch  3,  step=1354,  loss=0.071,  accuracy=1.000\n",
      "Epoch  3,  step=1355,  loss=0.227,  accuracy=0.938\n",
      "Epoch  3,  step=1356,  loss=0.111,  accuracy=0.969\n",
      "Epoch  3,  step=1357,  loss=0.426,  accuracy=0.938\n",
      "Epoch  3,  step=1358,  loss=0.268,  accuracy=0.906\n",
      "Epoch  3,  step=1359,  loss=0.189,  accuracy=0.938\n",
      "Epoch  3,  step=1360,  loss=0.285,  accuracy=0.906\n",
      "Epoch  3,  step=1361,  loss=0.207,  accuracy=0.906\n",
      "Epoch  3,  step=1362,  loss=0.216,  accuracy=0.906\n",
      "Epoch  3,  step=1363,  loss=0.257,  accuracy=0.938\n",
      "Epoch  3,  step=1364,  loss=0.509,  accuracy=0.875\n",
      "Epoch  3,  step=1365,  loss=0.235,  accuracy=0.906\n",
      "Epoch  3,  step=1366,  loss=0.035,  accuracy=1.000\n",
      "Epoch  3,  step=1367,  loss=0.146,  accuracy=0.969\n",
      "Epoch  3,  step=1368,  loss=0.319,  accuracy=0.938\n",
      "Epoch  3,  step=1369,  loss=0.449,  accuracy=0.844\n",
      "Epoch  3,  step=1370,  loss=0.238,  accuracy=0.906\n",
      "Epoch  3,  step=1371,  loss=0.350,  accuracy=0.875\n",
      "Epoch  3,  step=1372,  loss=0.132,  accuracy=0.938\n",
      "Epoch  3,  step=1373,  loss=0.576,  accuracy=0.812\n",
      "Epoch  3,  step=1374,  loss=0.371,  accuracy=0.906\n",
      "Epoch  3,  step=1375,  loss=0.071,  accuracy=1.000\n",
      "Epoch  3,  step=1376,  loss=0.283,  accuracy=0.906\n",
      "Epoch  3,  step=1377,  loss=0.260,  accuracy=0.938\n",
      "Epoch  3,  step=1378,  loss=0.359,  accuracy=0.844\n",
      "Epoch  3,  step=1379,  loss=0.488,  accuracy=0.844\n",
      "Epoch  3,  step=1380,  loss=0.213,  accuracy=0.969\n",
      "Epoch  3,  step=1381,  loss=0.185,  accuracy=0.938\n",
      "Epoch  3,  step=1382,  loss=0.396,  accuracy=0.844\n",
      "Epoch  3,  step=1383,  loss=0.590,  accuracy=0.875\n",
      "Epoch  3,  step=1384,  loss=0.325,  accuracy=0.875\n",
      "Epoch  3,  step=1385,  loss=0.257,  accuracy=0.906\n",
      "Epoch  3,  step=1386,  loss=0.583,  accuracy=0.938\n",
      "Epoch  3,  step=1387,  loss=0.143,  accuracy=0.969\n",
      "Epoch  3,  step=1388,  loss=0.386,  accuracy=0.938\n",
      "Epoch  3,  step=1389,  loss=0.188,  accuracy=0.938\n",
      "Epoch  3,  step=1390,  loss=0.406,  accuracy=0.875\n",
      "Epoch  3,  step=1391,  loss=0.124,  accuracy=0.969\n",
      "Epoch  3,  step=1392,  loss=0.060,  accuracy=1.000\n",
      "Epoch  3,  step=1393,  loss=0.069,  accuracy=0.969\n",
      "Epoch  3,  step=1394,  loss=0.135,  accuracy=0.938\n",
      "Epoch  3,  step=1395,  loss=0.289,  accuracy=0.938\n",
      "Epoch  3,  step=1396,  loss=0.330,  accuracy=0.938\n",
      "Epoch  3,  step=1397,  loss=0.171,  accuracy=0.938\n",
      "Epoch  3,  step=1398,  loss=0.640,  accuracy=0.781\n",
      "Epoch  3,  step=1399,  loss=0.097,  accuracy=0.969\n",
      "Epoch  3,  step=1400,  loss=0.251,  accuracy=0.906\n",
      "Epoch  3,  step=1401,  loss=0.487,  accuracy=0.812\n",
      "Epoch  3,  step=1402,  loss=0.708,  accuracy=0.875\n",
      "Epoch  3,  step=1403,  loss=0.425,  accuracy=0.875\n",
      "Epoch  3,  step=1404,  loss=0.217,  accuracy=0.906\n",
      "Epoch  3,  step=1405,  loss=0.203,  accuracy=0.938\n",
      "Epoch  3,  step=1406,  loss=0.387,  accuracy=0.938\n",
      "Epoch  3,  step=1407,  loss=0.845,  accuracy=0.844\n",
      "Epoch  3,  step=1408,  loss=0.209,  accuracy=0.938\n",
      "Epoch  3,  step=1409,  loss=0.670,  accuracy=0.812\n",
      "Epoch  3,  step=1410,  loss=0.271,  accuracy=0.938\n",
      "Epoch  3,  step=1411,  loss=0.370,  accuracy=0.906\n",
      "Epoch  3,  step=1412,  loss=0.314,  accuracy=0.938\n",
      "Epoch  3,  step=1413,  loss=0.245,  accuracy=0.906\n",
      "Epoch  3,  step=1414,  loss=0.178,  accuracy=0.938\n",
      "Epoch  3,  step=1415,  loss=0.094,  accuracy=0.969\n",
      "Epoch  3,  step=1416,  loss=0.194,  accuracy=0.906\n",
      "Epoch  3,  step=1417,  loss=0.163,  accuracy=0.938\n",
      "Epoch  3,  step=1418,  loss=0.105,  accuracy=0.969\n",
      "Epoch  3,  step=1419,  loss=0.576,  accuracy=0.844\n",
      "Epoch  3,  step=1420,  loss=0.603,  accuracy=0.812\n",
      "Epoch  3,  step=1421,  loss=0.866,  accuracy=0.875\n",
      "Epoch  3,  step=1422,  loss=0.480,  accuracy=0.875\n",
      "Epoch  3,  step=1423,  loss=0.175,  accuracy=0.938\n",
      "Epoch  3,  step=1424,  loss=0.253,  accuracy=0.906\n",
      "Epoch  3,  step=1425,  loss=0.434,  accuracy=0.875\n",
      "Epoch  3,  step=1426,  loss=0.419,  accuracy=0.844\n",
      "Epoch  3,  step=1427,  loss=0.376,  accuracy=0.875\n",
      "Epoch  3,  step=1428,  loss=0.026,  accuracy=1.000\n",
      "Epoch  3,  step=1429,  loss=0.278,  accuracy=0.844\n",
      "Epoch  3,  step=1430,  loss=0.428,  accuracy=0.875\n",
      "Epoch  3,  step=1431,  loss=0.323,  accuracy=0.938\n",
      "Epoch  3,  step=1432,  loss=0.461,  accuracy=0.812\n",
      "Epoch  3,  step=1433,  loss=0.692,  accuracy=0.875\n",
      "Epoch  3,  step=1434,  loss=0.516,  accuracy=0.875\n",
      "Epoch  3,  step=1435,  loss=0.429,  accuracy=0.875\n",
      "Epoch  3,  step=1436,  loss=0.567,  accuracy=0.781\n",
      "Epoch  3,  step=1437,  loss=0.453,  accuracy=0.906\n",
      "Epoch  3,  step=1438,  loss=0.331,  accuracy=0.875\n",
      "Epoch  3,  step=1439,  loss=0.398,  accuracy=0.906\n",
      "Epoch  3,  step=1440,  loss=0.569,  accuracy=0.875\n",
      "Epoch  3,  step=1441,  loss=0.196,  accuracy=0.906\n",
      "Epoch  3,  step=1442,  loss=0.175,  accuracy=0.938\n",
      "Epoch  3,  step=1443,  loss=0.154,  accuracy=0.969\n",
      "Epoch  3,  step=1444,  loss=0.120,  accuracy=0.969\n",
      "Epoch  3,  step=1445,  loss=1.111,  accuracy=0.750\n",
      "Epoch  3,  step=1446,  loss=0.637,  accuracy=0.781\n",
      "Epoch  3,  step=1447,  loss=0.498,  accuracy=0.781\n",
      "Epoch  3,  step=1448,  loss=0.329,  accuracy=0.875\n",
      "Epoch  3,  step=1449,  loss=0.242,  accuracy=0.906\n",
      "Epoch  3,  step=1450,  loss=0.462,  accuracy=0.875\n",
      "Epoch  3,  step=1451,  loss=0.157,  accuracy=0.906\n",
      "Epoch  3,  step=1452,  loss=0.101,  accuracy=0.969\n",
      "Epoch  3,  step=1453,  loss=0.140,  accuracy=0.969\n",
      "Epoch  3,  step=1454,  loss=0.179,  accuracy=0.906\n",
      "Epoch  3,  step=1455,  loss=0.133,  accuracy=0.906\n",
      "Epoch  3,  step=1456,  loss=0.236,  accuracy=0.938\n",
      "Epoch  3,  step=1457,  loss=0.331,  accuracy=0.938\n",
      "Epoch  3,  step=1458,  loss=0.090,  accuracy=1.000\n",
      "Epoch  3,  step=1459,  loss=0.233,  accuracy=0.906\n",
      "Epoch  3,  step=1460,  loss=0.835,  accuracy=0.781\n",
      "Epoch  3,  step=1461,  loss=0.320,  accuracy=0.938\n",
      "Epoch  3,  step=1462,  loss=0.216,  accuracy=0.875\n",
      "Epoch  3,  step=1463,  loss=0.104,  accuracy=0.969\n",
      "Epoch  3,  step=1464,  loss=0.181,  accuracy=0.906\n",
      "Epoch  3,  step=1465,  loss=0.274,  accuracy=0.906\n",
      "Epoch  3,  step=1466,  loss=0.246,  accuracy=0.906\n",
      "Epoch  3,  step=1467,  loss=0.145,  accuracy=0.938\n",
      "Epoch  3,  step=1468,  loss=0.215,  accuracy=0.969\n",
      "Epoch  3,  step=1469,  loss=0.682,  accuracy=0.844\n",
      "Epoch  3,  step=1470,  loss=0.440,  accuracy=0.875\n",
      "Epoch  3,  step=1471,  loss=0.245,  accuracy=0.938\n",
      "Epoch  3,  step=1472,  loss=0.107,  accuracy=0.969\n",
      "Epoch  3,  step=1473,  loss=0.207,  accuracy=0.938\n",
      "Epoch  3,  step=1474,  loss=0.269,  accuracy=0.906\n",
      "Epoch  3,  step=1475,  loss=0.728,  accuracy=0.812\n",
      "Epoch  3,  step=1476,  loss=0.559,  accuracy=0.781\n",
      "Epoch  3,  step=1477,  loss=0.307,  accuracy=0.906\n",
      "Epoch  3,  step=1478,  loss=0.600,  accuracy=0.844\n",
      "Epoch  3,  step=1479,  loss=0.533,  accuracy=0.969\n",
      "Epoch  3,  step=1480,  loss=0.511,  accuracy=0.906\n",
      "Epoch  3,  step=1481,  loss=0.510,  accuracy=0.844\n",
      "Epoch  3,  step=1482,  loss=0.393,  accuracy=0.875\n",
      "Epoch  3,  step=1483,  loss=0.326,  accuracy=0.844\n",
      "Epoch  3,  step=1484,  loss=0.082,  accuracy=1.000\n",
      "Epoch  3,  step=1485,  loss=0.219,  accuracy=0.875\n",
      "Epoch  3,  step=1486,  loss=0.425,  accuracy=0.938\n",
      "Epoch  3,  step=1487,  loss=0.629,  accuracy=0.812\n",
      "Epoch  3,  step=1488,  loss=0.205,  accuracy=0.906\n",
      "Epoch  3,  step=1489,  loss=0.193,  accuracy=0.938\n",
      "Epoch  3,  step=1490,  loss=0.425,  accuracy=0.875\n",
      "Epoch  3,  step=1491,  loss=0.286,  accuracy=0.938\n",
      "Epoch  3,  step=1492,  loss=0.411,  accuracy=0.875\n",
      "Epoch  3,  step=1493,  loss=0.249,  accuracy=0.938\n",
      "Epoch  3,  step=1494,  loss=0.243,  accuracy=0.938\n",
      "Epoch  3,  step=1495,  loss=0.101,  accuracy=0.969\n",
      "Epoch  3,  step=1496,  loss=0.248,  accuracy=0.906\n",
      "Epoch  3,  step=1497,  loss=0.599,  accuracy=0.906\n",
      "Epoch  3,  step=1498,  loss=0.424,  accuracy=0.875\n",
      "Epoch  3,  step=1499,  loss=0.200,  accuracy=0.938\n",
      "Epoch  3,  step=1500,  loss=0.238,  accuracy=0.906\n",
      "Epoch  3,  step=1501,  loss=0.177,  accuracy=0.906\n",
      "Epoch  3,  step=1502,  loss=0.345,  accuracy=0.875\n",
      "Epoch  3,  step=1503,  loss=0.302,  accuracy=0.938\n",
      "Epoch  3,  step=1504,  loss=0.131,  accuracy=0.969\n",
      "Epoch  3,  step=1505,  loss=0.129,  accuracy=0.938\n",
      "Epoch  3,  step=1506,  loss=0.065,  accuracy=1.000\n",
      "Epoch  3,  step=1507,  loss=0.061,  accuracy=1.000\n",
      "Epoch  3,  step=1508,  loss=0.203,  accuracy=0.938\n",
      "Epoch  3,  step=1509,  loss=0.088,  accuracy=0.938\n",
      "Epoch  3,  step=1510,  loss=0.322,  accuracy=0.875\n",
      "Epoch  3,  step=1511,  loss=0.510,  accuracy=0.844\n",
      "Epoch  3,  step=1512,  loss=0.671,  accuracy=0.781\n",
      "Epoch  3,  step=1513,  loss=0.034,  accuracy=1.000\n",
      "Epoch  3,  step=1514,  loss=0.222,  accuracy=0.844\n",
      "Epoch  3,  step=1515,  loss=0.267,  accuracy=0.969\n",
      "Epoch  3,  step=1516,  loss=0.242,  accuracy=0.906\n",
      "Epoch  3,  step=1517,  loss=0.191,  accuracy=0.938\n",
      "Epoch  3,  step=1518,  loss=0.217,  accuracy=0.938\n",
      "Epoch  3,  step=1519,  loss=0.038,  accuracy=1.000\n",
      "Epoch  3,  step=1520,  loss=0.509,  accuracy=0.906\n",
      "Epoch  3,  step=1521,  loss=0.260,  accuracy=0.875\n",
      "Epoch  3,  step=1522,  loss=0.171,  accuracy=0.938\n",
      "Epoch  3,  step=1523,  loss=0.215,  accuracy=0.906\n",
      "Epoch  3,  step=1524,  loss=0.160,  accuracy=0.938\n",
      "Epoch  3,  step=1525,  loss=0.147,  accuracy=0.906\n",
      "Epoch  3,  step=1526,  loss=0.387,  accuracy=0.844\n",
      "Epoch  3,  step=1527,  loss=0.344,  accuracy=0.875\n",
      "Epoch  3,  step=1528,  loss=0.716,  accuracy=0.844\n",
      "Epoch  3,  step=1529,  loss=0.759,  accuracy=0.750\n",
      "Epoch  3,  step=1530,  loss=0.624,  accuracy=0.875\n",
      "Epoch  3,  step=1531,  loss=1.348,  accuracy=0.781\n",
      "Epoch  3,  step=1532,  loss=0.253,  accuracy=0.844\n",
      "Epoch  3,  step=1533,  loss=0.647,  accuracy=0.812\n",
      "Epoch  3,  step=1534,  loss=0.325,  accuracy=0.875\n",
      "Epoch  3,  step=1535,  loss=0.618,  accuracy=0.844\n",
      "Epoch  3,  step=1536,  loss=0.356,  accuracy=0.875\n",
      "Epoch  3,  step=1537,  loss=1.002,  accuracy=0.688\n",
      "Epoch  3,  step=1538,  loss=0.229,  accuracy=0.875\n",
      "Epoch  3,  step=1539,  loss=0.172,  accuracy=0.938\n",
      "Epoch  3,  step=1540,  loss=0.198,  accuracy=0.938\n",
      "Epoch  3,  step=1541,  loss=0.187,  accuracy=0.938\n",
      "Epoch  3,  step=1542,  loss=0.044,  accuracy=1.000\n",
      "Epoch  3,  step=1543,  loss=0.057,  accuracy=1.000\n",
      "Epoch  3,  step=1544,  loss=0.835,  accuracy=0.750\n",
      "Epoch  3,  step=1545,  loss=0.190,  accuracy=0.938\n",
      "Epoch  3,  step=1546,  loss=0.847,  accuracy=0.812\n",
      "Epoch  3,  step=1547,  loss=0.614,  accuracy=0.750\n",
      "Epoch  3,  step=1548,  loss=0.768,  accuracy=0.812\n",
      "Epoch  3,  step=1549,  loss=0.827,  accuracy=0.875\n",
      "Epoch  3,  step=1550,  loss=0.430,  accuracy=0.875\n",
      "Epoch  3,  step=1551,  loss=0.676,  accuracy=0.844\n",
      "Epoch  3,  step=1552,  loss=0.397,  accuracy=0.875\n",
      "Epoch  3,  step=1553,  loss=0.128,  accuracy=0.938\n",
      "Epoch  3,  step=1554,  loss=0.270,  accuracy=0.938\n",
      "Epoch  3,  step=1555,  loss=0.151,  accuracy=0.969\n",
      "Epoch  3,  step=1556,  loss=0.321,  accuracy=0.938\n",
      "Epoch  3,  step=1557,  loss=0.267,  accuracy=0.906\n",
      "Epoch  3,  step=1558,  loss=0.062,  accuracy=0.969\n",
      "Epoch  3,  step=1559,  loss=0.756,  accuracy=0.812\n",
      "Epoch  3,  step=1560,  loss=0.202,  accuracy=0.938\n",
      "Epoch  3,  step=1561,  loss=0.146,  accuracy=0.969\n",
      "Epoch  3,  step=1562,  loss=0.357,  accuracy=0.844\n",
      "Epoch  3,  step=1563,  loss=0.265,  accuracy=0.906\n",
      "Epoch  3,  step=1564,  loss=0.324,  accuracy=0.875\n",
      "Epoch  3,  step=1565,  loss=0.125,  accuracy=0.938\n",
      "Epoch  3,  step=1566,  loss=0.190,  accuracy=0.969\n",
      "Epoch  3,  step=1567,  loss=0.466,  accuracy=0.875\n",
      "Epoch  3,  step=1568,  loss=0.191,  accuracy=0.969\n",
      "Epoch  3,  step=1569,  loss=0.730,  accuracy=0.844\n",
      "Epoch  3,  step=1570,  loss=0.576,  accuracy=0.875\n",
      "Epoch  3,  step=1571,  loss=0.091,  accuracy=0.969\n",
      "Epoch  3,  step=1572,  loss=0.465,  accuracy=0.812\n",
      "Epoch  3,  step=1573,  loss=0.450,  accuracy=0.812\n",
      "Epoch  3,  step=1574,  loss=0.557,  accuracy=0.812\n",
      "Epoch  3,  step=1575,  loss=1.008,  accuracy=0.750\n",
      "Epoch  3,  step=1576,  loss=0.337,  accuracy=0.875\n",
      "Epoch  3,  step=1577,  loss=0.097,  accuracy=0.938\n",
      "Epoch  3,  step=1578,  loss=0.515,  accuracy=0.906\n",
      "Epoch  3,  step=1579,  loss=0.098,  accuracy=0.969\n",
      "Epoch  3,  step=1580,  loss=0.928,  accuracy=0.844\n",
      "Epoch  3,  step=1581,  loss=0.169,  accuracy=0.938\n",
      "Epoch  3,  step=1582,  loss=0.847,  accuracy=0.844\n",
      "Epoch  3,  step=1583,  loss=0.148,  accuracy=0.969\n",
      "Epoch  3,  step=1584,  loss=0.444,  accuracy=0.844\n",
      "Epoch  3,  step=1585,  loss=0.297,  accuracy=0.906\n",
      "Epoch  3,  step=1586,  loss=0.288,  accuracy=0.906\n",
      "Epoch  3,  step=1587,  loss=0.288,  accuracy=0.938\n",
      "Epoch  3,  step=1588,  loss=0.280,  accuracy=0.875\n",
      "Epoch  3,  step=1589,  loss=0.129,  accuracy=0.938\n",
      "Epoch  3,  step=1590,  loss=0.279,  accuracy=0.938\n",
      "Epoch  3,  step=1591,  loss=0.064,  accuracy=1.000\n",
      "Epoch  3,  step=1592,  loss=0.195,  accuracy=0.938\n",
      "Epoch  3,  step=1593,  loss=0.195,  accuracy=0.938\n",
      "Epoch  3,  step=1594,  loss=0.099,  accuracy=0.969\n",
      "Epoch  3,  step=1595,  loss=0.457,  accuracy=0.875\n",
      "Epoch  3,  step=1596,  loss=0.223,  accuracy=0.969\n",
      "Epoch  3,  step=1597,  loss=0.293,  accuracy=0.875\n",
      "Epoch  3,  step=1598,  loss=0.286,  accuracy=0.875\n",
      "Epoch  3,  step=1599,  loss=0.477,  accuracy=0.844\n",
      "Epoch  3,  step=1600,  loss=0.176,  accuracy=0.938\n",
      "Epoch  3,  step=1601,  loss=0.323,  accuracy=0.844\n",
      "Epoch  3,  step=1602,  loss=0.474,  accuracy=0.875\n",
      "Epoch  3,  step=1603,  loss=0.441,  accuracy=0.906\n",
      "Epoch  3,  step=1604,  loss=0.262,  accuracy=0.938\n",
      "Epoch  3,  step=1605,  loss=0.138,  accuracy=0.938\n",
      "Epoch  3,  step=1606,  loss=0.189,  accuracy=0.906\n",
      "Epoch  3,  step=1607,  loss=0.162,  accuracy=0.906\n",
      "Epoch  3,  step=1608,  loss=0.119,  accuracy=0.969\n",
      "Epoch  3,  step=1609,  loss=0.107,  accuracy=0.969\n",
      "Epoch  3,  step=1610,  loss=0.457,  accuracy=0.906\n",
      "Epoch  3,  step=1611,  loss=0.253,  accuracy=0.844\n",
      "Epoch  3,  step=1612,  loss=0.050,  accuracy=1.000\n",
      "Epoch  3,  step=1613,  loss=0.297,  accuracy=0.938\n",
      "Epoch  3,  step=1614,  loss=0.108,  accuracy=0.969\n",
      "Epoch  3,  step=1615,  loss=0.304,  accuracy=0.906\n",
      "Epoch  3,  step=1616,  loss=0.527,  accuracy=0.938\n",
      "Epoch  3,  step=1617,  loss=0.334,  accuracy=0.906\n",
      "Epoch  3,  step=1618,  loss=0.238,  accuracy=0.938\n",
      "Epoch  3,  step=1619,  loss=0.059,  accuracy=1.000\n",
      "Epoch  3,  step=1620,  loss=0.265,  accuracy=0.875\n",
      "Epoch  3,  step=1621,  loss=0.252,  accuracy=0.875\n",
      "Epoch  3,  step=1622,  loss=0.220,  accuracy=0.938\n",
      "Epoch  3,  step=1623,  loss=0.487,  accuracy=0.875\n",
      "Epoch  3,  step=1624,  loss=0.693,  accuracy=0.844\n",
      "Epoch  3,  step=1625,  loss=0.084,  accuracy=0.969\n",
      "Epoch  3,  step=1626,  loss=0.159,  accuracy=0.938\n",
      "Epoch  3,  step=1627,  loss=0.435,  accuracy=0.906\n",
      "Epoch  3,  step=1628,  loss=0.735,  accuracy=0.750\n",
      "Epoch  3,  step=1629,  loss=1.027,  accuracy=0.812\n",
      "Epoch  3,  step=1630,  loss=0.526,  accuracy=0.875\n",
      "Epoch  3,  step=1631,  loss=0.208,  accuracy=0.938\n",
      "Epoch  3,  step=1632,  loss=0.435,  accuracy=0.875\n",
      "Epoch  3,  step=1633,  loss=0.292,  accuracy=0.844\n",
      "Epoch  3,  step=1634,  loss=0.105,  accuracy=0.969\n",
      "Epoch  3,  step=1635,  loss=0.289,  accuracy=0.969\n",
      "Epoch  3,  step=1636,  loss=0.440,  accuracy=0.938\n",
      "Epoch  3,  step=1637,  loss=0.368,  accuracy=0.906\n",
      "Epoch  3,  step=1638,  loss=0.237,  accuracy=0.938\n",
      "Epoch  3,  step=1639,  loss=0.090,  accuracy=0.969\n",
      "Epoch  3,  step=1640,  loss=0.167,  accuracy=0.938\n",
      "Epoch  3,  step=1641,  loss=0.095,  accuracy=0.938\n",
      "Epoch  3,  step=1642,  loss=0.078,  accuracy=1.000\n",
      "Epoch  3,  step=1643,  loss=0.137,  accuracy=0.969\n",
      "Epoch  3,  step=1644,  loss=0.090,  accuracy=0.969\n",
      "Epoch  3,  step=1645,  loss=0.598,  accuracy=0.875\n",
      "Epoch  3,  step=1646,  loss=0.469,  accuracy=0.875\n",
      "Epoch  3,  step=1647,  loss=0.506,  accuracy=0.844\n",
      "Epoch  3,  step=1648,  loss=0.328,  accuracy=0.875\n",
      "Epoch  3,  step=1649,  loss=0.208,  accuracy=0.875\n",
      "Epoch  3,  step=1650,  loss=0.425,  accuracy=0.844\n",
      "Epoch  3,  step=1651,  loss=0.141,  accuracy=0.938\n",
      "Epoch  3,  step=1652,  loss=0.320,  accuracy=0.844\n",
      "Epoch  3,  step=1653,  loss=0.530,  accuracy=0.875\n",
      "Epoch  3,  step=1654,  loss=0.749,  accuracy=0.844\n",
      "Epoch  3,  step=1655,  loss=0.286,  accuracy=0.906\n",
      "Epoch  3,  step=1656,  loss=0.362,  accuracy=0.906\n",
      "Epoch  3,  step=1657,  loss=0.396,  accuracy=0.844\n",
      "Epoch  3,  step=1658,  loss=0.289,  accuracy=0.906\n",
      "Epoch  3,  step=1659,  loss=0.393,  accuracy=0.844\n",
      "Epoch  3,  step=1660,  loss=0.299,  accuracy=0.906\n",
      "Epoch  3,  step=1661,  loss=0.742,  accuracy=0.844\n",
      "Epoch  3,  step=1662,  loss=0.174,  accuracy=0.906\n",
      "Epoch  3,  step=1663,  loss=0.449,  accuracy=0.906\n",
      "Epoch  3,  step=1664,  loss=0.092,  accuracy=0.938\n",
      "Epoch  3,  step=1665,  loss=0.231,  accuracy=0.906\n",
      "Epoch  3,  step=1666,  loss=0.112,  accuracy=0.969\n",
      "Epoch  3,  step=1667,  loss=0.070,  accuracy=1.000\n",
      "Epoch  3,  step=1668,  loss=0.226,  accuracy=0.906\n",
      "Epoch  3,  step=1669,  loss=0.064,  accuracy=1.000\n",
      "Epoch  3,  step=1670,  loss=0.361,  accuracy=0.906\n",
      "Epoch  3,  step=1671,  loss=0.392,  accuracy=0.844\n",
      "Epoch  3,  step=1672,  loss=0.363,  accuracy=0.875\n",
      "Epoch  3,  step=1673,  loss=0.352,  accuracy=0.875\n",
      "Epoch  3,  step=1674,  loss=0.413,  accuracy=0.906\n",
      "Epoch  3,  step=1675,  loss=0.232,  accuracy=0.906\n",
      "Epoch  3,  step=1676,  loss=0.265,  accuracy=0.938\n",
      "Epoch  3,  step=1677,  loss=0.441,  accuracy=0.875\n",
      "Epoch  3,  step=1678,  loss=0.079,  accuracy=0.969\n",
      "Epoch  3,  step=1679,  loss=0.544,  accuracy=0.750\n",
      "Epoch  3,  step=1680,  loss=0.030,  accuracy=1.000\n",
      "Epoch  3,  step=1681,  loss=0.358,  accuracy=0.906\n",
      "Epoch  3,  step=1682,  loss=0.338,  accuracy=0.875\n",
      "Epoch  3,  step=1683,  loss=0.316,  accuracy=0.938\n",
      "Epoch  3,  step=1684,  loss=0.677,  accuracy=0.844\n",
      "Epoch  3,  step=1685,  loss=0.165,  accuracy=0.938\n",
      "Epoch  3,  step=1686,  loss=0.795,  accuracy=0.750\n",
      "Epoch  3,  step=1687,  loss=0.394,  accuracy=0.844\n",
      "Epoch  3,  step=1688,  loss=0.284,  accuracy=0.906\n",
      "Epoch  3,  step=1689,  loss=0.225,  accuracy=0.938\n",
      "Epoch  3,  step=1690,  loss=0.692,  accuracy=0.812\n",
      "Epoch  3,  step=1691,  loss=0.060,  accuracy=1.000\n",
      "Epoch  3,  step=1692,  loss=0.107,  accuracy=0.969\n",
      "Epoch  3,  step=1693,  loss=0.328,  accuracy=0.875\n",
      "Epoch  3,  step=1694,  loss=0.117,  accuracy=0.969\n",
      "Epoch  3,  step=1695,  loss=0.260,  accuracy=0.906\n",
      "Epoch  3,  step=1696,  loss=0.340,  accuracy=0.844\n",
      "Epoch  3,  step=1697,  loss=0.114,  accuracy=0.938\n",
      "Epoch  3,  step=1698,  loss=0.137,  accuracy=0.938\n",
      "Epoch  3,  step=1699,  loss=0.745,  accuracy=0.906\n",
      "Epoch  3,  step=1700,  loss=0.220,  accuracy=0.938\n",
      "Epoch  3,  step=1701,  loss=0.481,  accuracy=0.906\n",
      "Epoch  3,  step=1702,  loss=0.079,  accuracy=1.000\n",
      "Epoch  3,  step=1703,  loss=0.395,  accuracy=0.906\n",
      "Epoch  3,  step=1704,  loss=0.298,  accuracy=0.844\n",
      "Epoch  3,  step=1705,  loss=0.496,  accuracy=0.812\n",
      "Epoch  3,  step=1706,  loss=0.096,  accuracy=0.969\n",
      "Epoch  3,  step=1707,  loss=0.278,  accuracy=0.906\n",
      "Epoch  3,  step=1708,  loss=0.177,  accuracy=0.938\n",
      "Epoch  3,  step=1709,  loss=0.142,  accuracy=0.969\n",
      "Epoch  3,  step=1710,  loss=0.068,  accuracy=0.969\n",
      "Epoch  3,  step=1711,  loss=0.375,  accuracy=0.875\n",
      "Epoch  3,  step=1712,  loss=0.301,  accuracy=0.906\n",
      "Epoch  3,  step=1713,  loss=0.203,  accuracy=0.969\n",
      "Epoch  3,  step=1714,  loss=0.371,  accuracy=0.906\n",
      "Epoch  3,  step=1715,  loss=0.595,  accuracy=0.844\n",
      "Epoch  3,  step=1716,  loss=0.541,  accuracy=0.812\n",
      "Epoch  3,  step=1717,  loss=0.551,  accuracy=0.844\n",
      "Epoch  3,  step=1718,  loss=0.243,  accuracy=0.906\n",
      "Epoch  3,  step=1719,  loss=0.344,  accuracy=0.875\n",
      "Epoch  3,  step=1720,  loss=0.507,  accuracy=0.844\n",
      "Epoch  3,  step=1721,  loss=0.311,  accuracy=0.906\n",
      "Epoch  3,  step=1722,  loss=0.435,  accuracy=0.906\n",
      "Epoch  3,  step=1723,  loss=0.037,  accuracy=1.000\n",
      "Epoch  3,  step=1724,  loss=0.215,  accuracy=0.938\n",
      "Epoch  3,  step=1725,  loss=0.231,  accuracy=0.875\n",
      "Epoch  3,  step=1726,  loss=0.172,  accuracy=0.938\n",
      "Epoch  3,  step=1727,  loss=0.150,  accuracy=0.906\n",
      "Epoch  3,  step=1728,  loss=0.370,  accuracy=0.844\n",
      "Epoch  3,  step=1729,  loss=0.429,  accuracy=0.938\n",
      "Epoch  3,  step=1730,  loss=0.089,  accuracy=0.938\n",
      "Epoch  3,  step=1731,  loss=0.096,  accuracy=0.969\n",
      "Epoch  3,  step=1732,  loss=0.210,  accuracy=0.906\n",
      "Epoch  3,  step=1733,  loss=0.251,  accuracy=0.938\n",
      "Epoch  3,  step=1734,  loss=0.469,  accuracy=0.875\n",
      "Epoch  3,  step=1735,  loss=0.282,  accuracy=0.906\n",
      "Epoch  3,  step=1736,  loss=0.035,  accuracy=1.000\n",
      "Epoch  3,  step=1737,  loss=0.175,  accuracy=0.969\n",
      "Epoch  3,  step=1738,  loss=0.099,  accuracy=0.969\n",
      "Epoch  3,  step=1739,  loss=0.230,  accuracy=0.906\n",
      "Epoch  3,  step=1740,  loss=0.184,  accuracy=0.938\n",
      "Epoch  3,  step=1741,  loss=0.708,  accuracy=0.812\n",
      "Epoch  3,  step=1742,  loss=0.224,  accuracy=0.938\n",
      "Epoch  3,  step=1743,  loss=0.300,  accuracy=0.938\n",
      "Epoch  3,  step=1744,  loss=0.139,  accuracy=0.938\n",
      "Epoch  3,  step=1745,  loss=0.548,  accuracy=0.906\n",
      "Epoch  3,  step=1746,  loss=0.351,  accuracy=0.844\n",
      "Epoch  3,  step=1747,  loss=0.268,  accuracy=0.906\n",
      "Epoch  3,  step=1748,  loss=0.105,  accuracy=0.969\n",
      "Epoch  3,  step=1749,  loss=0.204,  accuracy=0.906\n",
      "Epoch  3,  step=1750,  loss=0.211,  accuracy=0.969\n",
      "Epoch  3,  step=1751,  loss=0.092,  accuracy=0.969\n",
      "Epoch  3,  step=1752,  loss=0.172,  accuracy=0.875\n",
      "Epoch  3,  step=1753,  loss=0.079,  accuracy=0.938\n",
      "Epoch  3,  step=1754,  loss=0.330,  accuracy=0.938\n",
      "Epoch  3,  step=1755,  loss=0.145,  accuracy=0.969\n",
      "Epoch  3,  step=1756,  loss=0.663,  accuracy=0.875\n",
      "Epoch  3,  step=1757,  loss=0.122,  accuracy=0.969\n",
      "Epoch  3,  step=1758,  loss=0.286,  accuracy=0.875\n",
      "Epoch  3,  step=1759,  loss=0.413,  accuracy=0.844\n",
      "Epoch  3,  step=1760,  loss=0.203,  accuracy=0.969\n",
      "Epoch  3,  step=1761,  loss=0.253,  accuracy=0.844\n",
      "Epoch  3,  step=1762,  loss=0.300,  accuracy=0.875\n",
      "Epoch  3,  step=1763,  loss=0.217,  accuracy=0.938\n",
      "Epoch  3,  step=1764,  loss=1.184,  accuracy=0.719\n",
      "Epoch  3,  step=1765,  loss=0.427,  accuracy=0.875\n",
      "Epoch  3,  step=1766,  loss=0.043,  accuracy=1.000\n",
      "Epoch  3,  step=1767,  loss=0.044,  accuracy=1.000\n",
      "Epoch  3,  step=1768,  loss=0.316,  accuracy=0.875\n",
      "Epoch  3,  step=1769,  loss=0.332,  accuracy=0.906\n",
      "Epoch  3,  step=1770,  loss=0.487,  accuracy=0.844\n",
      "Epoch  3,  step=1771,  loss=0.277,  accuracy=0.969\n",
      "Epoch  3,  step=1772,  loss=0.050,  accuracy=1.000\n",
      "Epoch  3,  step=1773,  loss=0.071,  accuracy=1.000\n",
      "Epoch  3,  step=1774,  loss=0.194,  accuracy=0.938\n",
      "Epoch  3,  step=1775,  loss=0.221,  accuracy=0.906\n",
      "Epoch  3,  step=1776,  loss=0.221,  accuracy=0.875\n",
      "Epoch  3,  step=1777,  loss=0.093,  accuracy=1.000\n",
      "Epoch  3,  step=1778,  loss=0.248,  accuracy=0.938\n",
      "Epoch  3,  step=1779,  loss=0.045,  accuracy=0.969\n",
      "Epoch  3,  step=1780,  loss=0.233,  accuracy=0.938\n",
      "Epoch  3,  step=1781,  loss=0.464,  accuracy=0.844\n",
      "Epoch  3,  step=1782,  loss=0.560,  accuracy=0.938\n",
      "Epoch  3,  step=1783,  loss=0.565,  accuracy=0.906\n",
      "Epoch  3,  step=1784,  loss=0.075,  accuracy=0.969\n",
      "Epoch  3,  step=1785,  loss=0.115,  accuracy=0.938\n",
      "Epoch  3,  step=1786,  loss=0.308,  accuracy=0.875\n",
      "Epoch  3,  step=1787,  loss=0.178,  accuracy=0.969\n",
      "Epoch  3,  step=1788,  loss=0.323,  accuracy=0.875\n",
      "Epoch  3,  step=1789,  loss=0.270,  accuracy=0.844\n",
      "Epoch  3,  step=1790,  loss=0.358,  accuracy=0.875\n",
      "Epoch  3,  step=1791,  loss=0.519,  accuracy=0.844\n",
      "Epoch  3,  step=1792,  loss=0.087,  accuracy=0.938\n",
      "Epoch  3,  step=1793,  loss=0.265,  accuracy=0.938\n",
      "Epoch  3,  step=1794,  loss=0.119,  accuracy=0.938\n",
      "Epoch  3,  step=1795,  loss=0.157,  accuracy=0.938\n",
      "Epoch  3,  step=1796,  loss=0.340,  accuracy=0.906\n",
      "Epoch  3,  step=1797,  loss=0.354,  accuracy=0.906\n",
      "Epoch  3,  step=1798,  loss=0.350,  accuracy=0.844\n",
      "Epoch  3,  step=1799,  loss=0.348,  accuracy=0.844\n",
      "Epoch  3,  step=1800,  loss=0.536,  accuracy=0.875\n",
      "Epoch  3,  step=1801,  loss=0.497,  accuracy=0.812\n",
      "Epoch  3,  step=1802,  loss=0.288,  accuracy=0.938\n",
      "Epoch  3,  step=1803,  loss=0.478,  accuracy=0.875\n",
      "Epoch  3,  step=1804,  loss=0.489,  accuracy=0.906\n",
      "Epoch  3,  step=1805,  loss=0.085,  accuracy=0.969\n",
      "Epoch  3,  step=1806,  loss=0.213,  accuracy=0.906\n",
      "Epoch  3,  step=1807,  loss=0.351,  accuracy=0.875\n",
      "Epoch  3,  step=1808,  loss=0.228,  accuracy=0.938\n",
      "Epoch  3,  step=1809,  loss=0.111,  accuracy=0.938\n",
      "Epoch  3,  step=1810,  loss=0.216,  accuracy=0.938\n",
      "Epoch  3,  step=1811,  loss=0.190,  accuracy=0.906\n",
      "Epoch  3,  step=1812,  loss=0.122,  accuracy=0.938\n",
      "Epoch  3,  step=1813,  loss=0.158,  accuracy=0.938\n",
      "Epoch  3,  step=1814,  loss=0.478,  accuracy=0.875\n",
      "Epoch  3,  step=1815,  loss=0.601,  accuracy=0.906\n",
      "Epoch  3,  step=1816,  loss=0.053,  accuracy=0.969\n",
      "Epoch  3,  step=1817,  loss=0.038,  accuracy=1.000\n",
      "Epoch  3,  step=1818,  loss=0.036,  accuracy=0.969\n",
      "Epoch  3,  step=1819,  loss=0.053,  accuracy=1.000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  3,  step=1820,  loss=0.133,  accuracy=0.969\n",
      "Epoch  3,  step=1821,  loss=0.094,  accuracy=0.969\n",
      "Epoch  3,  step=1822,  loss=0.228,  accuracy=0.938\n",
      "Epoch  3,  step=1823,  loss=0.030,  accuracy=1.000\n",
      "Epoch  3,  step=1824,  loss=0.192,  accuracy=0.938\n",
      "Epoch  3,  step=1825,  loss=0.183,  accuracy=0.906\n",
      "Epoch  3,  step=1826,  loss=0.102,  accuracy=0.969\n",
      "Epoch  3,  step=1827,  loss=0.156,  accuracy=0.938\n",
      "Epoch  3,  step=1828,  loss=0.298,  accuracy=0.938\n",
      "Epoch  3,  step=1829,  loss=0.322,  accuracy=0.938\n",
      "Epoch  3,  step=1830,  loss=0.098,  accuracy=0.969\n",
      "Epoch  3,  step=1831,  loss=0.226,  accuracy=0.906\n",
      "Epoch  3,  step=1832,  loss=0.192,  accuracy=0.938\n",
      "Epoch  3,  step=1833,  loss=0.173,  accuracy=0.938\n",
      "Epoch  3,  step=1834,  loss=0.081,  accuracy=0.969\n",
      "Epoch  3,  step=1835,  loss=0.034,  accuracy=1.000\n",
      "Epoch  3,  step=1836,  loss=0.133,  accuracy=0.938\n",
      "Epoch  3,  step=1837,  loss=0.258,  accuracy=0.844\n",
      "Epoch  3,  step=1838,  loss=0.262,  accuracy=0.906\n",
      "Epoch  3,  step=1839,  loss=0.432,  accuracy=0.938\n",
      "Epoch  3,  step=1840,  loss=0.094,  accuracy=0.938\n",
      "Epoch  3,  step=1841,  loss=0.020,  accuracy=1.000\n",
      "Epoch  3,  step=1842,  loss=0.020,  accuracy=1.000\n",
      "Epoch  3,  step=1843,  loss=0.183,  accuracy=0.906\n",
      "Epoch  3,  step=1844,  loss=0.028,  accuracy=1.000\n",
      "Epoch  3,  step=1845,  loss=0.208,  accuracy=0.938\n",
      "Epoch  3,  step=1846,  loss=0.207,  accuracy=0.875\n",
      "Epoch  3,  step=1847,  loss=0.108,  accuracy=0.969\n",
      "Epoch  3,  step=1848,  loss=0.255,  accuracy=0.906\n",
      "Epoch  3,  step=1849,  loss=0.187,  accuracy=0.969\n",
      "Epoch  3,  step=1850,  loss=0.158,  accuracy=0.938\n",
      "Epoch  3,  step=1851,  loss=0.189,  accuracy=0.969\n",
      "Epoch  3,  step=1852,  loss=0.460,  accuracy=0.844\n",
      "Epoch  3,  step=1853,  loss=0.837,  accuracy=0.844\n",
      "Epoch  3,  step=1854,  loss=0.630,  accuracy=0.844\n",
      "Epoch  3,  step=1855,  loss=0.235,  accuracy=0.906\n",
      "Epoch  3,  step=1856,  loss=0.233,  accuracy=0.969\n",
      "Epoch  3,  step=1857,  loss=0.184,  accuracy=0.875\n",
      "Epoch  3,  step=1858,  loss=0.084,  accuracy=1.000\n",
      "Epoch  3,  step=1859,  loss=0.098,  accuracy=0.938\n",
      "Epoch  3,  step=1860,  loss=0.090,  accuracy=0.969\n",
      "Epoch  3,  step=1861,  loss=0.115,  accuracy=0.938\n",
      "Epoch  3,  step=1862,  loss=0.083,  accuracy=0.969\n",
      "Epoch  3,  step=1863,  loss=0.065,  accuracy=1.000\n",
      "Epoch  3,  step=1864,  loss=0.326,  accuracy=0.938\n",
      "Epoch  3,  step=1865,  loss=0.306,  accuracy=0.938\n",
      "Epoch  3,  step=1866,  loss=1.197,  accuracy=0.719\n",
      "Epoch  3,  step=1867,  loss=0.172,  accuracy=0.938\n",
      "Epoch  3,  step=1868,  loss=0.335,  accuracy=0.750\n",
      "Epoch  3,  step=1869,  loss=0.094,  accuracy=0.969\n",
      "Epoch  3,  step=1870,  loss=0.043,  accuracy=1.000\n",
      "Epoch  3,  step=1871,  loss=0.046,  accuracy=1.000\n",
      "Epoch  3,  step=1872,  loss=0.971,  accuracy=0.875\n",
      "Epoch  3,  step=1873,  loss=0.164,  accuracy=0.938\n",
      "Epoch  3,  step=1874,  loss=0.123,  accuracy=0.938\n",
      "Epoch  4,  step=   0,  loss=0.202,  accuracy=0.969\n",
      "Epoch  4,  step=   1,  loss=0.547,  accuracy=0.875\n",
      "Epoch  4,  step=   2,  loss=0.439,  accuracy=0.969\n",
      "Epoch  4,  step=   3,  loss=0.240,  accuracy=0.938\n",
      "Epoch  4,  step=   4,  loss=0.529,  accuracy=0.906\n",
      "Epoch  4,  step=   5,  loss=0.523,  accuracy=0.875\n",
      "Epoch  4,  step=   6,  loss=0.045,  accuracy=1.000\n",
      "Epoch  4,  step=   7,  loss=0.506,  accuracy=0.875\n",
      "Epoch  4,  step=   8,  loss=0.401,  accuracy=0.844\n",
      "Epoch  4,  step=   9,  loss=0.077,  accuracy=0.969\n",
      "Epoch  4,  step=  10,  loss=0.251,  accuracy=0.969\n",
      "Epoch  4,  step=  11,  loss=0.142,  accuracy=0.938\n",
      "Epoch  4,  step=  12,  loss=0.128,  accuracy=0.969\n",
      "Epoch  4,  step=  13,  loss=0.314,  accuracy=0.875\n",
      "Epoch  4,  step=  14,  loss=0.220,  accuracy=0.906\n",
      "Epoch  4,  step=  15,  loss=0.611,  accuracy=0.875\n",
      "Epoch  4,  step=  16,  loss=0.133,  accuracy=0.969\n",
      "Epoch  4,  step=  17,  loss=0.274,  accuracy=0.906\n",
      "Epoch  4,  step=  18,  loss=0.183,  accuracy=0.938\n",
      "Epoch  4,  step=  19,  loss=0.676,  accuracy=0.812\n",
      "Epoch  4,  step=  20,  loss=0.254,  accuracy=0.906\n",
      "Epoch  4,  step=  21,  loss=0.093,  accuracy=0.969\n",
      "Epoch  4,  step=  22,  loss=0.370,  accuracy=0.938\n",
      "Epoch  4,  step=  23,  loss=0.288,  accuracy=0.875\n",
      "Epoch  4,  step=  24,  loss=0.352,  accuracy=0.875\n",
      "Epoch  4,  step=  25,  loss=0.117,  accuracy=0.938\n",
      "Epoch  4,  step=  26,  loss=0.624,  accuracy=0.844\n",
      "Epoch  4,  step=  27,  loss=0.479,  accuracy=0.906\n",
      "Epoch  4,  step=  28,  loss=0.833,  accuracy=0.875\n",
      "Epoch  4,  step=  29,  loss=0.309,  accuracy=0.938\n",
      "Epoch  4,  step=  30,  loss=0.074,  accuracy=1.000\n",
      "Epoch  4,  step=  31,  loss=0.437,  accuracy=0.875\n",
      "Epoch  4,  step=  32,  loss=0.460,  accuracy=0.844\n",
      "Epoch  4,  step=  33,  loss=0.880,  accuracy=0.812\n",
      "Epoch  4,  step=  34,  loss=0.231,  accuracy=0.906\n",
      "Epoch  4,  step=  35,  loss=0.768,  accuracy=0.719\n",
      "Epoch  4,  step=  36,  loss=0.174,  accuracy=0.875\n",
      "Epoch  4,  step=  37,  loss=0.153,  accuracy=0.969\n",
      "Epoch  4,  step=  38,  loss=1.013,  accuracy=0.812\n",
      "Epoch  4,  step=  39,  loss=0.440,  accuracy=0.844\n",
      "Epoch  4,  step=  40,  loss=0.199,  accuracy=0.906\n",
      "Epoch  4,  step=  41,  loss=0.330,  accuracy=0.875\n",
      "Epoch  4,  step=  42,  loss=0.570,  accuracy=0.875\n",
      "Epoch  4,  step=  43,  loss=0.537,  accuracy=0.812\n",
      "Epoch  4,  step=  44,  loss=0.202,  accuracy=0.906\n",
      "Epoch  4,  step=  45,  loss=0.186,  accuracy=0.969\n",
      "Epoch  4,  step=  46,  loss=0.129,  accuracy=0.938\n",
      "Epoch  4,  step=  47,  loss=0.661,  accuracy=0.875\n",
      "Epoch  4,  step=  48,  loss=0.241,  accuracy=0.875\n",
      "Epoch  4,  step=  49,  loss=0.381,  accuracy=0.875\n",
      "Epoch  4,  step=  50,  loss=0.302,  accuracy=0.906\n",
      "Epoch  4,  step=  51,  loss=0.282,  accuracy=0.875\n",
      "Epoch  4,  step=  52,  loss=0.187,  accuracy=0.906\n",
      "Epoch  4,  step=  53,  loss=0.063,  accuracy=0.969\n",
      "Epoch  4,  step=  54,  loss=0.159,  accuracy=0.969\n",
      "Epoch  4,  step=  55,  loss=0.196,  accuracy=0.906\n",
      "Epoch  4,  step=  56,  loss=0.411,  accuracy=0.938\n",
      "Epoch  4,  step=  57,  loss=0.473,  accuracy=0.875\n",
      "Epoch  4,  step=  58,  loss=0.164,  accuracy=0.969\n",
      "Epoch  4,  step=  59,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step=  60,  loss=0.279,  accuracy=0.906\n",
      "Epoch  4,  step=  61,  loss=0.308,  accuracy=0.844\n",
      "Epoch  4,  step=  62,  loss=0.226,  accuracy=0.906\n",
      "Epoch  4,  step=  63,  loss=0.190,  accuracy=0.906\n",
      "Epoch  4,  step=  64,  loss=0.283,  accuracy=0.906\n",
      "Epoch  4,  step=  65,  loss=0.299,  accuracy=0.938\n",
      "Epoch  4,  step=  66,  loss=0.229,  accuracy=0.906\n",
      "Epoch  4,  step=  67,  loss=0.075,  accuracy=0.969\n",
      "Epoch  4,  step=  68,  loss=0.092,  accuracy=1.000\n",
      "Epoch  4,  step=  69,  loss=0.187,  accuracy=0.938\n",
      "Epoch  4,  step=  70,  loss=0.201,  accuracy=0.906\n",
      "Epoch  4,  step=  71,  loss=0.225,  accuracy=0.969\n",
      "Epoch  4,  step=  72,  loss=0.102,  accuracy=0.969\n",
      "Epoch  4,  step=  73,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step=  74,  loss=0.056,  accuracy=1.000\n",
      "Epoch  4,  step=  75,  loss=0.651,  accuracy=0.812\n",
      "Epoch  4,  step=  76,  loss=0.149,  accuracy=0.938\n",
      "Epoch  4,  step=  77,  loss=0.121,  accuracy=0.969\n",
      "Epoch  4,  step=  78,  loss=0.138,  accuracy=0.969\n",
      "Epoch  4,  step=  79,  loss=0.287,  accuracy=0.875\n",
      "Epoch  4,  step=  80,  loss=0.152,  accuracy=0.938\n",
      "Epoch  4,  step=  81,  loss=0.264,  accuracy=0.906\n",
      "Epoch  4,  step=  82,  loss=0.272,  accuracy=0.938\n",
      "Epoch  4,  step=  83,  loss=0.910,  accuracy=0.875\n",
      "Epoch  4,  step=  84,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step=  85,  loss=0.124,  accuracy=0.938\n",
      "Epoch  4,  step=  86,  loss=0.481,  accuracy=0.875\n",
      "Epoch  4,  step=  87,  loss=0.558,  accuracy=0.844\n",
      "Epoch  4,  step=  88,  loss=0.532,  accuracy=0.875\n",
      "Epoch  4,  step=  89,  loss=0.108,  accuracy=0.969\n",
      "Epoch  4,  step=  90,  loss=0.199,  accuracy=0.969\n",
      "Epoch  4,  step=  91,  loss=0.309,  accuracy=0.875\n",
      "Epoch  4,  step=  92,  loss=0.270,  accuracy=0.844\n",
      "Epoch  4,  step=  93,  loss=0.154,  accuracy=0.938\n",
      "Epoch  4,  step=  94,  loss=0.279,  accuracy=0.938\n",
      "Epoch  4,  step=  95,  loss=0.094,  accuracy=1.000\n",
      "Epoch  4,  step=  96,  loss=0.168,  accuracy=0.969\n",
      "Epoch  4,  step=  97,  loss=0.339,  accuracy=0.906\n",
      "Epoch  4,  step=  98,  loss=0.087,  accuracy=1.000\n",
      "Epoch  4,  step=  99,  loss=0.086,  accuracy=0.969\n",
      "Epoch  4,  step= 100,  loss=0.307,  accuracy=0.875\n",
      "Epoch  4,  step= 101,  loss=0.231,  accuracy=0.906\n",
      "Epoch  4,  step= 102,  loss=0.501,  accuracy=0.844\n",
      "Epoch  4,  step= 103,  loss=0.073,  accuracy=0.969\n",
      "Epoch  4,  step= 104,  loss=0.104,  accuracy=0.969\n",
      "Epoch  4,  step= 105,  loss=0.356,  accuracy=0.938\n",
      "Epoch  4,  step= 106,  loss=0.399,  accuracy=0.938\n",
      "Epoch  4,  step= 107,  loss=0.124,  accuracy=0.969\n",
      "Epoch  4,  step= 108,  loss=0.247,  accuracy=0.906\n",
      "Epoch  4,  step= 109,  loss=0.444,  accuracy=0.906\n",
      "Epoch  4,  step= 110,  loss=0.264,  accuracy=0.906\n",
      "Epoch  4,  step= 111,  loss=0.151,  accuracy=0.938\n",
      "Epoch  4,  step= 112,  loss=0.045,  accuracy=1.000\n",
      "Epoch  4,  step= 113,  loss=0.163,  accuracy=0.938\n",
      "Epoch  4,  step= 114,  loss=0.353,  accuracy=0.875\n",
      "Epoch  4,  step= 115,  loss=0.638,  accuracy=0.812\n",
      "Epoch  4,  step= 116,  loss=0.505,  accuracy=0.906\n",
      "Epoch  4,  step= 117,  loss=0.106,  accuracy=0.969\n",
      "Epoch  4,  step= 118,  loss=0.289,  accuracy=0.906\n",
      "Epoch  4,  step= 119,  loss=0.384,  accuracy=0.906\n",
      "Epoch  4,  step= 120,  loss=0.097,  accuracy=0.969\n",
      "Epoch  4,  step= 121,  loss=0.326,  accuracy=0.969\n",
      "Epoch  4,  step= 122,  loss=0.074,  accuracy=0.969\n",
      "Epoch  4,  step= 123,  loss=0.089,  accuracy=0.969\n",
      "Epoch  4,  step= 124,  loss=0.312,  accuracy=0.906\n",
      "Epoch  4,  step= 125,  loss=0.313,  accuracy=0.906\n",
      "Epoch  4,  step= 126,  loss=0.165,  accuracy=0.969\n",
      "Epoch  4,  step= 127,  loss=0.220,  accuracy=0.938\n",
      "Epoch  4,  step= 128,  loss=0.194,  accuracy=0.938\n",
      "Epoch  4,  step= 129,  loss=0.572,  accuracy=0.781\n",
      "Epoch  4,  step= 130,  loss=0.265,  accuracy=0.875\n",
      "Epoch  4,  step= 131,  loss=0.235,  accuracy=0.969\n",
      "Epoch  4,  step= 132,  loss=0.194,  accuracy=0.969\n",
      "Epoch  4,  step= 133,  loss=0.303,  accuracy=0.875\n",
      "Epoch  4,  step= 134,  loss=0.152,  accuracy=0.969\n",
      "Epoch  4,  step= 135,  loss=0.126,  accuracy=0.969\n",
      "Epoch  4,  step= 136,  loss=0.348,  accuracy=0.938\n",
      "Epoch  4,  step= 137,  loss=0.113,  accuracy=0.969\n",
      "Epoch  4,  step= 138,  loss=0.115,  accuracy=0.938\n",
      "Epoch  4,  step= 139,  loss=0.453,  accuracy=0.844\n",
      "Epoch  4,  step= 140,  loss=0.253,  accuracy=0.938\n",
      "Epoch  4,  step= 141,  loss=0.065,  accuracy=0.969\n",
      "Epoch  4,  step= 142,  loss=0.073,  accuracy=0.969\n",
      "Epoch  4,  step= 143,  loss=0.318,  accuracy=0.906\n",
      "Epoch  4,  step= 144,  loss=0.167,  accuracy=0.906\n",
      "Epoch  4,  step= 145,  loss=0.602,  accuracy=0.844\n",
      "Epoch  4,  step= 146,  loss=0.170,  accuracy=0.969\n",
      "Epoch  4,  step= 147,  loss=0.140,  accuracy=0.906\n",
      "Epoch  4,  step= 148,  loss=0.324,  accuracy=0.875\n",
      "Epoch  4,  step= 149,  loss=0.227,  accuracy=0.938\n",
      "Epoch  4,  step= 150,  loss=0.211,  accuracy=0.938\n",
      "Epoch  4,  step= 151,  loss=0.064,  accuracy=1.000\n",
      "Epoch  4,  step= 152,  loss=0.135,  accuracy=0.938\n",
      "Epoch  4,  step= 153,  loss=0.228,  accuracy=0.906\n",
      "Epoch  4,  step= 154,  loss=0.357,  accuracy=0.875\n",
      "Epoch  4,  step= 155,  loss=0.609,  accuracy=0.844\n",
      "Epoch  4,  step= 156,  loss=0.242,  accuracy=0.906\n",
      "Epoch  4,  step= 157,  loss=0.390,  accuracy=0.844\n",
      "Epoch  4,  step= 158,  loss=0.426,  accuracy=0.906\n",
      "Epoch  4,  step= 159,  loss=0.326,  accuracy=0.875\n",
      "Epoch  4,  step= 160,  loss=0.350,  accuracy=0.844\n",
      "Epoch  4,  step= 161,  loss=0.714,  accuracy=0.844\n",
      "Epoch  4,  step= 162,  loss=0.303,  accuracy=0.906\n",
      "Epoch  4,  step= 163,  loss=0.089,  accuracy=0.969\n",
      "Epoch  4,  step= 164,  loss=0.309,  accuracy=0.875\n",
      "Epoch  4,  step= 165,  loss=0.941,  accuracy=0.875\n",
      "Epoch  4,  step= 166,  loss=0.433,  accuracy=0.938\n",
      "Epoch  4,  step= 167,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step= 168,  loss=0.219,  accuracy=0.906\n",
      "Epoch  4,  step= 169,  loss=0.178,  accuracy=0.969\n",
      "Epoch  4,  step= 170,  loss=0.068,  accuracy=0.969\n",
      "Epoch  4,  step= 171,  loss=0.151,  accuracy=0.938\n",
      "Epoch  4,  step= 172,  loss=0.102,  accuracy=0.969\n",
      "Epoch  4,  step= 173,  loss=0.367,  accuracy=0.938\n",
      "Epoch  4,  step= 174,  loss=0.072,  accuracy=0.969\n",
      "Epoch  4,  step= 175,  loss=0.337,  accuracy=0.812\n",
      "Epoch  4,  step= 176,  loss=0.178,  accuracy=0.938\n",
      "Epoch  4,  step= 177,  loss=0.417,  accuracy=0.875\n",
      "Epoch  4,  step= 178,  loss=0.423,  accuracy=0.875\n",
      "Epoch  4,  step= 179,  loss=0.117,  accuracy=0.969\n",
      "Epoch  4,  step= 180,  loss=0.393,  accuracy=0.906\n",
      "Epoch  4,  step= 181,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step= 182,  loss=0.323,  accuracy=0.875\n",
      "Epoch  4,  step= 183,  loss=0.290,  accuracy=0.906\n",
      "Epoch  4,  step= 184,  loss=0.276,  accuracy=0.938\n",
      "Epoch  4,  step= 185,  loss=0.318,  accuracy=0.938\n",
      "Epoch  4,  step= 186,  loss=0.113,  accuracy=0.969\n",
      "Epoch  4,  step= 187,  loss=0.063,  accuracy=1.000\n",
      "Epoch  4,  step= 188,  loss=0.127,  accuracy=0.938\n",
      "Epoch  4,  step= 189,  loss=0.061,  accuracy=0.969\n",
      "Epoch  4,  step= 190,  loss=0.512,  accuracy=0.906\n",
      "Epoch  4,  step= 191,  loss=0.313,  accuracy=0.969\n",
      "Epoch  4,  step= 192,  loss=0.138,  accuracy=0.969\n",
      "Epoch  4,  step= 193,  loss=0.164,  accuracy=0.906\n",
      "Epoch  4,  step= 194,  loss=0.134,  accuracy=0.969\n",
      "Epoch  4,  step= 195,  loss=0.493,  accuracy=0.812\n",
      "Epoch  4,  step= 196,  loss=0.349,  accuracy=0.875\n",
      "Epoch  4,  step= 197,  loss=0.251,  accuracy=0.906\n",
      "Epoch  4,  step= 198,  loss=0.174,  accuracy=0.938\n",
      "Epoch  4,  step= 199,  loss=0.172,  accuracy=0.938\n",
      "Epoch  4,  step= 200,  loss=0.393,  accuracy=0.875\n",
      "Epoch  4,  step= 201,  loss=0.394,  accuracy=0.969\n",
      "Epoch  4,  step= 202,  loss=0.248,  accuracy=0.938\n",
      "Epoch  4,  step= 203,  loss=0.241,  accuracy=0.938\n",
      "Epoch  4,  step= 204,  loss=0.194,  accuracy=0.906\n",
      "Epoch  4,  step= 205,  loss=0.169,  accuracy=0.938\n",
      "Epoch  4,  step= 206,  loss=0.124,  accuracy=0.938\n",
      "Epoch  4,  step= 207,  loss=0.206,  accuracy=0.969\n",
      "Epoch  4,  step= 208,  loss=0.209,  accuracy=0.938\n",
      "Epoch  4,  step= 209,  loss=0.123,  accuracy=0.969\n",
      "Epoch  4,  step= 210,  loss=0.202,  accuracy=0.969\n",
      "Epoch  4,  step= 211,  loss=0.060,  accuracy=1.000\n",
      "Epoch  4,  step= 212,  loss=0.202,  accuracy=0.906\n",
      "Epoch  4,  step= 213,  loss=0.518,  accuracy=0.844\n",
      "Epoch  4,  step= 214,  loss=0.472,  accuracy=0.906\n",
      "Epoch  4,  step= 215,  loss=0.524,  accuracy=0.875\n",
      "Epoch  4,  step= 216,  loss=0.585,  accuracy=0.812\n",
      "Epoch  4,  step= 217,  loss=0.286,  accuracy=0.906\n",
      "Epoch  4,  step= 218,  loss=0.128,  accuracy=0.969\n",
      "Epoch  4,  step= 219,  loss=0.628,  accuracy=0.844\n",
      "Epoch  4,  step= 220,  loss=0.190,  accuracy=0.906\n",
      "Epoch  4,  step= 221,  loss=0.105,  accuracy=0.969\n",
      "Epoch  4,  step= 222,  loss=0.164,  accuracy=0.906\n",
      "Epoch  4,  step= 223,  loss=0.422,  accuracy=0.906\n",
      "Epoch  4,  step= 224,  loss=0.153,  accuracy=0.938\n",
      "Epoch  4,  step= 225,  loss=0.522,  accuracy=0.844\n",
      "Epoch  4,  step= 226,  loss=0.631,  accuracy=0.812\n",
      "Epoch  4,  step= 227,  loss=0.450,  accuracy=0.812\n",
      "Epoch  4,  step= 228,  loss=0.234,  accuracy=0.969\n",
      "Epoch  4,  step= 229,  loss=0.412,  accuracy=0.906\n",
      "Epoch  4,  step= 230,  loss=0.173,  accuracy=0.938\n",
      "Epoch  4,  step= 231,  loss=0.038,  accuracy=1.000\n",
      "Epoch  4,  step= 232,  loss=0.260,  accuracy=0.875\n",
      "Epoch  4,  step= 233,  loss=0.083,  accuracy=0.969\n",
      "Epoch  4,  step= 234,  loss=0.110,  accuracy=0.969\n",
      "Epoch  4,  step= 235,  loss=0.506,  accuracy=0.906\n",
      "Epoch  4,  step= 236,  loss=0.047,  accuracy=1.000\n",
      "Epoch  4,  step= 237,  loss=0.524,  accuracy=0.781\n",
      "Epoch  4,  step= 238,  loss=0.353,  accuracy=0.875\n",
      "Epoch  4,  step= 239,  loss=0.188,  accuracy=0.906\n",
      "Epoch  4,  step= 240,  loss=0.217,  accuracy=0.906\n",
      "Epoch  4,  step= 241,  loss=0.450,  accuracy=0.844\n",
      "Epoch  4,  step= 242,  loss=0.241,  accuracy=0.906\n",
      "Epoch  4,  step= 243,  loss=0.274,  accuracy=0.875\n",
      "Epoch  4,  step= 244,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step= 245,  loss=0.370,  accuracy=0.875\n",
      "Epoch  4,  step= 246,  loss=0.423,  accuracy=0.938\n",
      "Epoch  4,  step= 247,  loss=0.351,  accuracy=0.938\n",
      "Epoch  4,  step= 248,  loss=0.126,  accuracy=0.969\n",
      "Epoch  4,  step= 249,  loss=0.353,  accuracy=0.875\n",
      "Epoch  4,  step= 250,  loss=0.398,  accuracy=0.844\n",
      "Epoch  4,  step= 251,  loss=0.167,  accuracy=0.906\n",
      "Epoch  4,  step= 252,  loss=0.268,  accuracy=0.875\n",
      "Epoch  4,  step= 253,  loss=0.587,  accuracy=0.812\n",
      "Epoch  4,  step= 254,  loss=0.102,  accuracy=0.969\n",
      "Epoch  4,  step= 255,  loss=0.070,  accuracy=0.969\n",
      "Epoch  4,  step= 256,  loss=0.875,  accuracy=0.812\n",
      "Epoch  4,  step= 257,  loss=0.383,  accuracy=0.906\n",
      "Epoch  4,  step= 258,  loss=0.191,  accuracy=0.938\n",
      "Epoch  4,  step= 259,  loss=0.190,  accuracy=0.938\n",
      "Epoch  4,  step= 260,  loss=0.143,  accuracy=0.938\n",
      "Epoch  4,  step= 261,  loss=0.125,  accuracy=0.969\n",
      "Epoch  4,  step= 262,  loss=0.155,  accuracy=0.969\n",
      "Epoch  4,  step= 263,  loss=0.349,  accuracy=0.875\n",
      "Epoch  4,  step= 264,  loss=0.718,  accuracy=0.844\n",
      "Epoch  4,  step= 265,  loss=0.304,  accuracy=0.844\n",
      "Epoch  4,  step= 266,  loss=0.217,  accuracy=0.875\n",
      "Epoch  4,  step= 267,  loss=0.232,  accuracy=0.875\n",
      "Epoch  4,  step= 268,  loss=0.169,  accuracy=0.938\n",
      "Epoch  4,  step= 269,  loss=0.473,  accuracy=0.812\n",
      "Epoch  4,  step= 270,  loss=0.195,  accuracy=0.938\n",
      "Epoch  4,  step= 271,  loss=0.824,  accuracy=0.781\n",
      "Epoch  4,  step= 272,  loss=1.151,  accuracy=0.781\n",
      "Epoch  4,  step= 273,  loss=0.183,  accuracy=0.906\n",
      "Epoch  4,  step= 274,  loss=0.302,  accuracy=0.875\n",
      "Epoch  4,  step= 275,  loss=0.488,  accuracy=0.875\n",
      "Epoch  4,  step= 276,  loss=0.552,  accuracy=0.875\n",
      "Epoch  4,  step= 277,  loss=0.645,  accuracy=0.781\n",
      "Epoch  4,  step= 278,  loss=0.518,  accuracy=0.844\n",
      "Epoch  4,  step= 279,  loss=0.267,  accuracy=0.906\n",
      "Epoch  4,  step= 280,  loss=0.323,  accuracy=0.938\n",
      "Epoch  4,  step= 281,  loss=0.202,  accuracy=0.875\n",
      "Epoch  4,  step= 282,  loss=0.130,  accuracy=0.938\n",
      "Epoch  4,  step= 283,  loss=0.215,  accuracy=0.938\n",
      "Epoch  4,  step= 284,  loss=0.526,  accuracy=0.875\n",
      "Epoch  4,  step= 285,  loss=0.496,  accuracy=0.906\n",
      "Epoch  4,  step= 286,  loss=0.393,  accuracy=0.875\n",
      "Epoch  4,  step= 287,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step= 288,  loss=0.239,  accuracy=0.969\n",
      "Epoch  4,  step= 289,  loss=0.651,  accuracy=0.844\n",
      "Epoch  4,  step= 290,  loss=0.283,  accuracy=0.969\n",
      "Epoch  4,  step= 291,  loss=0.200,  accuracy=0.906\n",
      "Epoch  4,  step= 292,  loss=0.212,  accuracy=0.906\n",
      "Epoch  4,  step= 293,  loss=0.584,  accuracy=0.844\n",
      "Epoch  4,  step= 294,  loss=0.303,  accuracy=0.875\n",
      "Epoch  4,  step= 295,  loss=0.281,  accuracy=0.906\n",
      "Epoch  4,  step= 296,  loss=0.050,  accuracy=1.000\n",
      "Epoch  4,  step= 297,  loss=0.250,  accuracy=0.875\n",
      "Epoch  4,  step= 298,  loss=0.346,  accuracy=0.906\n",
      "Epoch  4,  step= 299,  loss=0.292,  accuracy=0.875\n",
      "Epoch  4,  step= 300,  loss=0.231,  accuracy=0.938\n",
      "Epoch  4,  step= 301,  loss=0.223,  accuracy=0.906\n",
      "Epoch  4,  step= 302,  loss=0.249,  accuracy=0.875\n",
      "Epoch  4,  step= 303,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step= 304,  loss=0.114,  accuracy=0.938\n",
      "Epoch  4,  step= 305,  loss=0.515,  accuracy=0.844\n",
      "Epoch  4,  step= 306,  loss=0.290,  accuracy=0.875\n",
      "Epoch  4,  step= 307,  loss=0.021,  accuracy=1.000\n",
      "Epoch  4,  step= 308,  loss=0.221,  accuracy=0.938\n",
      "Epoch  4,  step= 309,  loss=0.023,  accuracy=1.000\n",
      "Epoch  4,  step= 310,  loss=0.259,  accuracy=0.938\n",
      "Epoch  4,  step= 311,  loss=0.121,  accuracy=0.969\n",
      "Epoch  4,  step= 312,  loss=0.184,  accuracy=0.938\n",
      "Epoch  4,  step= 313,  loss=0.629,  accuracy=0.844\n",
      "Epoch  4,  step= 314,  loss=0.241,  accuracy=0.906\n",
      "Epoch  4,  step= 315,  loss=0.245,  accuracy=0.906\n",
      "Epoch  4,  step= 316,  loss=0.299,  accuracy=0.906\n",
      "Epoch  4,  step= 317,  loss=0.579,  accuracy=0.812\n",
      "Epoch  4,  step= 318,  loss=0.455,  accuracy=0.875\n",
      "Epoch  4,  step= 319,  loss=0.408,  accuracy=0.844\n",
      "Epoch  4,  step= 320,  loss=0.724,  accuracy=0.812\n",
      "Epoch  4,  step= 321,  loss=0.518,  accuracy=0.781\n",
      "Epoch  4,  step= 322,  loss=0.271,  accuracy=0.969\n",
      "Epoch  4,  step= 323,  loss=0.351,  accuracy=0.969\n",
      "Epoch  4,  step= 324,  loss=0.188,  accuracy=0.906\n",
      "Epoch  4,  step= 325,  loss=0.103,  accuracy=0.938\n",
      "Epoch  4,  step= 326,  loss=0.220,  accuracy=0.938\n",
      "Epoch  4,  step= 327,  loss=0.183,  accuracy=0.938\n",
      "Epoch  4,  step= 328,  loss=0.198,  accuracy=0.969\n",
      "Epoch  4,  step= 329,  loss=0.040,  accuracy=1.000\n",
      "Epoch  4,  step= 330,  loss=0.157,  accuracy=0.938\n",
      "Epoch  4,  step= 331,  loss=0.050,  accuracy=1.000\n",
      "Epoch  4,  step= 332,  loss=0.109,  accuracy=0.938\n",
      "Epoch  4,  step= 333,  loss=0.163,  accuracy=0.969\n",
      "Epoch  4,  step= 334,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step= 335,  loss=0.544,  accuracy=0.906\n",
      "Epoch  4,  step= 336,  loss=0.388,  accuracy=0.906\n",
      "Epoch  4,  step= 337,  loss=0.254,  accuracy=0.938\n",
      "Epoch  4,  step= 338,  loss=0.293,  accuracy=0.906\n",
      "Epoch  4,  step= 339,  loss=0.418,  accuracy=0.875\n",
      "Epoch  4,  step= 340,  loss=0.145,  accuracy=0.969\n",
      "Epoch  4,  step= 341,  loss=0.123,  accuracy=0.906\n",
      "Epoch  4,  step= 342,  loss=0.221,  accuracy=0.938\n",
      "Epoch  4,  step= 343,  loss=0.412,  accuracy=0.906\n",
      "Epoch  4,  step= 344,  loss=0.214,  accuracy=0.969\n",
      "Epoch  4,  step= 345,  loss=0.258,  accuracy=0.938\n",
      "Epoch  4,  step= 346,  loss=0.139,  accuracy=0.969\n",
      "Epoch  4,  step= 347,  loss=0.256,  accuracy=0.938\n",
      "Epoch  4,  step= 348,  loss=0.150,  accuracy=0.938\n",
      "Epoch  4,  step= 349,  loss=0.113,  accuracy=0.969\n",
      "Epoch  4,  step= 350,  loss=0.545,  accuracy=0.906\n",
      "Epoch  4,  step= 351,  loss=0.397,  accuracy=0.844\n",
      "Epoch  4,  step= 352,  loss=0.112,  accuracy=0.938\n",
      "Epoch  4,  step= 353,  loss=0.225,  accuracy=0.906\n",
      "Epoch  4,  step= 354,  loss=0.071,  accuracy=1.000\n",
      "Epoch  4,  step= 355,  loss=0.320,  accuracy=0.844\n",
      "Epoch  4,  step= 356,  loss=0.130,  accuracy=0.969\n",
      "Epoch  4,  step= 357,  loss=0.248,  accuracy=0.906\n",
      "Epoch  4,  step= 358,  loss=0.171,  accuracy=0.969\n",
      "Epoch  4,  step= 359,  loss=0.249,  accuracy=0.875\n",
      "Epoch  4,  step= 360,  loss=0.222,  accuracy=0.906\n",
      "Epoch  4,  step= 361,  loss=0.575,  accuracy=0.781\n",
      "Epoch  4,  step= 362,  loss=0.272,  accuracy=0.875\n",
      "Epoch  4,  step= 363,  loss=0.623,  accuracy=0.875\n",
      "Epoch  4,  step= 364,  loss=0.241,  accuracy=0.969\n",
      "Epoch  4,  step= 365,  loss=0.370,  accuracy=0.906\n",
      "Epoch  4,  step= 366,  loss=0.191,  accuracy=0.938\n",
      "Epoch  4,  step= 367,  loss=0.431,  accuracy=0.938\n",
      "Epoch  4,  step= 368,  loss=0.364,  accuracy=0.844\n",
      "Epoch  4,  step= 369,  loss=0.131,  accuracy=0.969\n",
      "Epoch  4,  step= 370,  loss=0.347,  accuracy=0.875\n",
      "Epoch  4,  step= 371,  loss=0.191,  accuracy=0.906\n",
      "Epoch  4,  step= 372,  loss=0.117,  accuracy=0.969\n",
      "Epoch  4,  step= 373,  loss=0.300,  accuracy=0.906\n",
      "Epoch  4,  step= 374,  loss=0.118,  accuracy=0.938\n",
      "Epoch  4,  step= 375,  loss=0.240,  accuracy=0.906\n",
      "Epoch  4,  step= 376,  loss=0.345,  accuracy=0.906\n",
      "Epoch  4,  step= 377,  loss=0.295,  accuracy=0.875\n",
      "Epoch  4,  step= 378,  loss=0.215,  accuracy=0.938\n",
      "Epoch  4,  step= 379,  loss=0.072,  accuracy=0.969\n",
      "Epoch  4,  step= 380,  loss=0.247,  accuracy=0.906\n",
      "Epoch  4,  step= 381,  loss=0.322,  accuracy=0.875\n",
      "Epoch  4,  step= 382,  loss=0.176,  accuracy=0.969\n",
      "Epoch  4,  step= 383,  loss=0.479,  accuracy=0.844\n",
      "Epoch  4,  step= 384,  loss=0.296,  accuracy=0.906\n",
      "Epoch  4,  step= 385,  loss=0.253,  accuracy=0.938\n",
      "Epoch  4,  step= 386,  loss=0.506,  accuracy=0.812\n",
      "Epoch  4,  step= 387,  loss=0.045,  accuracy=1.000\n",
      "Epoch  4,  step= 388,  loss=0.280,  accuracy=0.875\n",
      "Epoch  4,  step= 389,  loss=0.410,  accuracy=0.938\n",
      "Epoch  4,  step= 390,  loss=0.494,  accuracy=0.875\n",
      "Epoch  4,  step= 391,  loss=0.289,  accuracy=0.906\n",
      "Epoch  4,  step= 392,  loss=0.821,  accuracy=0.812\n",
      "Epoch  4,  step= 393,  loss=0.707,  accuracy=0.844\n",
      "Epoch  4,  step= 394,  loss=0.584,  accuracy=0.875\n",
      "Epoch  4,  step= 395,  loss=0.576,  accuracy=0.844\n",
      "Epoch  4,  step= 396,  loss=0.920,  accuracy=0.812\n",
      "Epoch  4,  step= 397,  loss=0.079,  accuracy=0.969\n",
      "Epoch  4,  step= 398,  loss=0.339,  accuracy=0.875\n",
      "Epoch  4,  step= 399,  loss=0.394,  accuracy=0.906\n",
      "Epoch  4,  step= 400,  loss=0.148,  accuracy=0.969\n",
      "Epoch  4,  step= 401,  loss=0.170,  accuracy=0.906\n",
      "Epoch  4,  step= 402,  loss=0.381,  accuracy=0.844\n",
      "Epoch  4,  step= 403,  loss=0.238,  accuracy=0.938\n",
      "Epoch  4,  step= 404,  loss=0.663,  accuracy=0.844\n",
      "Epoch  4,  step= 405,  loss=0.292,  accuracy=0.875\n",
      "Epoch  4,  step= 406,  loss=0.576,  accuracy=0.844\n",
      "Epoch  4,  step= 407,  loss=0.396,  accuracy=0.938\n",
      "Epoch  4,  step= 408,  loss=0.807,  accuracy=0.844\n",
      "Epoch  4,  step= 409,  loss=0.342,  accuracy=0.844\n",
      "Epoch  4,  step= 410,  loss=0.357,  accuracy=0.875\n",
      "Epoch  4,  step= 411,  loss=0.153,  accuracy=0.938\n",
      "Epoch  4,  step= 412,  loss=0.496,  accuracy=0.812\n",
      "Epoch  4,  step= 413,  loss=0.071,  accuracy=0.969\n",
      "Epoch  4,  step= 414,  loss=0.595,  accuracy=0.938\n",
      "Epoch  4,  step= 415,  loss=0.229,  accuracy=0.938\n",
      "Epoch  4,  step= 416,  loss=0.169,  accuracy=0.938\n",
      "Epoch  4,  step= 417,  loss=0.153,  accuracy=0.969\n",
      "Epoch  4,  step= 418,  loss=0.267,  accuracy=0.875\n",
      "Epoch  4,  step= 419,  loss=0.277,  accuracy=0.938\n",
      "Epoch  4,  step= 420,  loss=0.328,  accuracy=0.969\n",
      "Epoch  4,  step= 421,  loss=0.030,  accuracy=1.000\n",
      "Epoch  4,  step= 422,  loss=0.511,  accuracy=0.875\n",
      "Epoch  4,  step= 423,  loss=0.264,  accuracy=0.906\n",
      "Epoch  4,  step= 424,  loss=0.056,  accuracy=1.000\n",
      "Epoch  4,  step= 425,  loss=0.088,  accuracy=0.969\n",
      "Epoch  4,  step= 426,  loss=0.543,  accuracy=0.875\n",
      "Epoch  4,  step= 427,  loss=0.479,  accuracy=0.875\n",
      "Epoch  4,  step= 428,  loss=0.426,  accuracy=0.844\n",
      "Epoch  4,  step= 429,  loss=0.321,  accuracy=0.875\n",
      "Epoch  4,  step= 430,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step= 431,  loss=0.136,  accuracy=0.969\n",
      "Epoch  4,  step= 432,  loss=0.750,  accuracy=0.812\n",
      "Epoch  4,  step= 433,  loss=0.227,  accuracy=0.938\n",
      "Epoch  4,  step= 434,  loss=0.137,  accuracy=0.969\n",
      "Epoch  4,  step= 435,  loss=0.242,  accuracy=0.938\n",
      "Epoch  4,  step= 436,  loss=0.503,  accuracy=0.875\n",
      "Epoch  4,  step= 437,  loss=0.368,  accuracy=0.844\n",
      "Epoch  4,  step= 438,  loss=0.258,  accuracy=0.906\n",
      "Epoch  4,  step= 439,  loss=0.513,  accuracy=0.875\n",
      "Epoch  4,  step= 440,  loss=0.234,  accuracy=0.938\n",
      "Epoch  4,  step= 441,  loss=0.574,  accuracy=0.875\n",
      "Epoch  4,  step= 442,  loss=0.116,  accuracy=0.969\n",
      "Epoch  4,  step= 443,  loss=0.272,  accuracy=0.906\n",
      "Epoch  4,  step= 444,  loss=0.605,  accuracy=0.875\n",
      "Epoch  4,  step= 445,  loss=0.383,  accuracy=0.906\n",
      "Epoch  4,  step= 446,  loss=0.153,  accuracy=0.969\n",
      "Epoch  4,  step= 447,  loss=0.258,  accuracy=0.906\n",
      "Epoch  4,  step= 448,  loss=0.926,  accuracy=0.750\n",
      "Epoch  4,  step= 449,  loss=0.436,  accuracy=0.844\n",
      "Epoch  4,  step= 450,  loss=0.162,  accuracy=0.938\n",
      "Epoch  4,  step= 451,  loss=0.171,  accuracy=0.938\n",
      "Epoch  4,  step= 452,  loss=0.179,  accuracy=0.969\n",
      "Epoch  4,  step= 453,  loss=0.149,  accuracy=0.969\n",
      "Epoch  4,  step= 454,  loss=0.459,  accuracy=0.875\n",
      "Epoch  4,  step= 455,  loss=0.422,  accuracy=0.906\n",
      "Epoch  4,  step= 456,  loss=0.488,  accuracy=0.812\n",
      "Epoch  4,  step= 457,  loss=0.353,  accuracy=0.938\n",
      "Epoch  4,  step= 458,  loss=0.444,  accuracy=0.812\n",
      "Epoch  4,  step= 459,  loss=0.604,  accuracy=0.750\n",
      "Epoch  4,  step= 460,  loss=0.644,  accuracy=0.781\n",
      "Epoch  4,  step= 461,  loss=1.197,  accuracy=0.781\n",
      "Epoch  4,  step= 462,  loss=0.503,  accuracy=0.844\n",
      "Epoch  4,  step= 463,  loss=0.601,  accuracy=0.875\n",
      "Epoch  4,  step= 464,  loss=0.356,  accuracy=0.906\n",
      "Epoch  4,  step= 465,  loss=0.273,  accuracy=0.875\n",
      "Epoch  4,  step= 466,  loss=0.236,  accuracy=0.906\n",
      "Epoch  4,  step= 467,  loss=0.156,  accuracy=0.938\n",
      "Epoch  4,  step= 468,  loss=0.289,  accuracy=0.875\n",
      "Epoch  4,  step= 469,  loss=0.129,  accuracy=0.969\n",
      "Epoch  4,  step= 470,  loss=0.533,  accuracy=0.969\n",
      "Epoch  4,  step= 471,  loss=0.090,  accuracy=0.969\n",
      "Epoch  4,  step= 472,  loss=1.019,  accuracy=0.781\n",
      "Epoch  4,  step= 473,  loss=0.645,  accuracy=0.875\n",
      "Epoch  4,  step= 474,  loss=0.351,  accuracy=0.875\n",
      "Epoch  4,  step= 475,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step= 476,  loss=0.161,  accuracy=0.906\n",
      "Epoch  4,  step= 477,  loss=0.336,  accuracy=0.906\n",
      "Epoch  4,  step= 478,  loss=0.095,  accuracy=0.969\n",
      "Epoch  4,  step= 479,  loss=0.407,  accuracy=0.844\n",
      "Epoch  4,  step= 480,  loss=0.144,  accuracy=0.969\n",
      "Epoch  4,  step= 481,  loss=0.282,  accuracy=0.938\n",
      "Epoch  4,  step= 482,  loss=0.476,  accuracy=0.969\n",
      "Epoch  4,  step= 483,  loss=0.126,  accuracy=0.938\n",
      "Epoch  4,  step= 484,  loss=0.383,  accuracy=0.938\n",
      "Epoch  4,  step= 485,  loss=0.347,  accuracy=0.875\n",
      "Epoch  4,  step= 486,  loss=0.136,  accuracy=0.969\n",
      "Epoch  4,  step= 487,  loss=0.397,  accuracy=0.938\n",
      "Epoch  4,  step= 488,  loss=0.261,  accuracy=0.844\n",
      "Epoch  4,  step= 489,  loss=0.174,  accuracy=0.906\n",
      "Epoch  4,  step= 490,  loss=0.259,  accuracy=0.938\n",
      "Epoch  4,  step= 491,  loss=0.818,  accuracy=0.781\n",
      "Epoch  4,  step= 492,  loss=0.544,  accuracy=0.844\n",
      "Epoch  4,  step= 493,  loss=0.388,  accuracy=0.844\n",
      "Epoch  4,  step= 494,  loss=0.153,  accuracy=0.969\n",
      "Epoch  4,  step= 495,  loss=0.812,  accuracy=0.781\n",
      "Epoch  4,  step= 496,  loss=0.243,  accuracy=0.969\n",
      "Epoch  4,  step= 497,  loss=0.186,  accuracy=0.938\n",
      "Epoch  4,  step= 498,  loss=0.439,  accuracy=0.938\n",
      "Epoch  4,  step= 499,  loss=0.494,  accuracy=0.875\n",
      "Epoch  4,  step= 500,  loss=0.713,  accuracy=0.781\n",
      "Epoch  4,  step= 501,  loss=0.650,  accuracy=0.781\n",
      "Epoch  4,  step= 502,  loss=0.316,  accuracy=0.875\n",
      "Epoch  4,  step= 503,  loss=0.402,  accuracy=0.875\n",
      "Epoch  4,  step= 504,  loss=0.155,  accuracy=0.906\n",
      "Epoch  4,  step= 505,  loss=0.099,  accuracy=0.969\n",
      "Epoch  4,  step= 506,  loss=0.269,  accuracy=0.844\n",
      "Epoch  4,  step= 507,  loss=0.025,  accuracy=1.000\n",
      "Epoch  4,  step= 508,  loss=0.488,  accuracy=0.875\n",
      "Epoch  4,  step= 509,  loss=0.291,  accuracy=0.906\n",
      "Epoch  4,  step= 510,  loss=0.074,  accuracy=1.000\n",
      "Epoch  4,  step= 511,  loss=0.318,  accuracy=0.875\n",
      "Epoch  4,  step= 512,  loss=0.124,  accuracy=0.938\n",
      "Epoch  4,  step= 513,  loss=0.274,  accuracy=0.906\n",
      "Epoch  4,  step= 514,  loss=0.199,  accuracy=0.938\n",
      "Epoch  4,  step= 515,  loss=0.327,  accuracy=0.875\n",
      "Epoch  4,  step= 516,  loss=0.191,  accuracy=0.938\n",
      "Epoch  4,  step= 517,  loss=0.514,  accuracy=0.844\n",
      "Epoch  4,  step= 518,  loss=0.224,  accuracy=0.875\n",
      "Epoch  4,  step= 519,  loss=0.161,  accuracy=0.969\n",
      "Epoch  4,  step= 520,  loss=0.394,  accuracy=0.875\n",
      "Epoch  4,  step= 521,  loss=0.491,  accuracy=0.844\n",
      "Epoch  4,  step= 522,  loss=0.150,  accuracy=0.969\n",
      "Epoch  4,  step= 523,  loss=0.505,  accuracy=0.906\n",
      "Epoch  4,  step= 524,  loss=0.629,  accuracy=0.875\n",
      "Epoch  4,  step= 525,  loss=0.286,  accuracy=0.906\n",
      "Epoch  4,  step= 526,  loss=0.253,  accuracy=0.906\n",
      "Epoch  4,  step= 527,  loss=0.411,  accuracy=0.906\n",
      "Epoch  4,  step= 528,  loss=0.276,  accuracy=0.906\n",
      "Epoch  4,  step= 529,  loss=0.409,  accuracy=0.812\n",
      "Epoch  4,  step= 530,  loss=0.610,  accuracy=0.875\n",
      "Epoch  4,  step= 531,  loss=0.326,  accuracy=0.844\n",
      "Epoch  4,  step= 532,  loss=0.327,  accuracy=0.875\n",
      "Epoch  4,  step= 533,  loss=0.170,  accuracy=0.969\n",
      "Epoch  4,  step= 534,  loss=0.216,  accuracy=0.906\n",
      "Epoch  4,  step= 535,  loss=0.205,  accuracy=0.938\n",
      "Epoch  4,  step= 536,  loss=0.336,  accuracy=0.906\n",
      "Epoch  4,  step= 537,  loss=0.737,  accuracy=0.844\n",
      "Epoch  4,  step= 538,  loss=0.341,  accuracy=0.906\n",
      "Epoch  4,  step= 539,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step= 540,  loss=0.086,  accuracy=1.000\n",
      "Epoch  4,  step= 541,  loss=0.061,  accuracy=1.000\n",
      "Epoch  4,  step= 542,  loss=0.207,  accuracy=0.938\n",
      "Epoch  4,  step= 543,  loss=0.308,  accuracy=0.844\n",
      "Epoch  4,  step= 544,  loss=0.136,  accuracy=0.938\n",
      "Epoch  4,  step= 545,  loss=0.135,  accuracy=0.969\n",
      "Epoch  4,  step= 546,  loss=0.515,  accuracy=0.875\n",
      "Epoch  4,  step= 547,  loss=0.213,  accuracy=0.906\n",
      "Epoch  4,  step= 548,  loss=0.394,  accuracy=0.812\n",
      "Epoch  4,  step= 549,  loss=0.396,  accuracy=0.875\n",
      "Epoch  4,  step= 550,  loss=0.099,  accuracy=1.000\n",
      "Epoch  4,  step= 551,  loss=0.534,  accuracy=0.875\n",
      "Epoch  4,  step= 552,  loss=0.452,  accuracy=0.812\n",
      "Epoch  4,  step= 553,  loss=0.760,  accuracy=0.812\n",
      "Epoch  4,  step= 554,  loss=0.159,  accuracy=0.969\n",
      "Epoch  4,  step= 555,  loss=0.246,  accuracy=0.969\n",
      "Epoch  4,  step= 556,  loss=0.392,  accuracy=0.844\n",
      "Epoch  4,  step= 557,  loss=0.176,  accuracy=0.906\n",
      "Epoch  4,  step= 558,  loss=0.121,  accuracy=0.969\n",
      "Epoch  4,  step= 559,  loss=0.174,  accuracy=0.875\n",
      "Epoch  4,  step= 560,  loss=0.259,  accuracy=0.906\n",
      "Epoch  4,  step= 561,  loss=0.199,  accuracy=0.938\n",
      "Epoch  4,  step= 562,  loss=0.191,  accuracy=0.938\n",
      "Epoch  4,  step= 563,  loss=0.280,  accuracy=0.906\n",
      "Epoch  4,  step= 564,  loss=0.407,  accuracy=0.938\n",
      "Epoch  4,  step= 565,  loss=0.357,  accuracy=0.906\n",
      "Epoch  4,  step= 566,  loss=0.076,  accuracy=1.000\n",
      "Epoch  4,  step= 567,  loss=0.239,  accuracy=0.906\n",
      "Epoch  4,  step= 568,  loss=0.156,  accuracy=0.938\n",
      "Epoch  4,  step= 569,  loss=0.121,  accuracy=0.969\n",
      "Epoch  4,  step= 570,  loss=0.485,  accuracy=0.812\n",
      "Epoch  4,  step= 571,  loss=0.205,  accuracy=0.938\n",
      "Epoch  4,  step= 572,  loss=0.133,  accuracy=0.969\n",
      "Epoch  4,  step= 573,  loss=0.209,  accuracy=0.969\n",
      "Epoch  4,  step= 574,  loss=0.265,  accuracy=0.906\n",
      "Epoch  4,  step= 575,  loss=0.376,  accuracy=0.781\n",
      "Epoch  4,  step= 576,  loss=0.102,  accuracy=0.938\n",
      "Epoch  4,  step= 577,  loss=0.513,  accuracy=0.875\n",
      "Epoch  4,  step= 578,  loss=0.418,  accuracy=0.875\n",
      "Epoch  4,  step= 579,  loss=0.098,  accuracy=1.000\n",
      "Epoch  4,  step= 580,  loss=0.494,  accuracy=0.844\n",
      "Epoch  4,  step= 581,  loss=0.366,  accuracy=0.938\n",
      "Epoch  4,  step= 582,  loss=0.389,  accuracy=0.875\n",
      "Epoch  4,  step= 583,  loss=0.166,  accuracy=0.938\n",
      "Epoch  4,  step= 584,  loss=0.137,  accuracy=0.938\n",
      "Epoch  4,  step= 585,  loss=0.275,  accuracy=0.906\n",
      "Epoch  4,  step= 586,  loss=0.032,  accuracy=1.000\n",
      "Epoch  4,  step= 587,  loss=0.123,  accuracy=0.906\n",
      "Epoch  4,  step= 588,  loss=0.259,  accuracy=0.906\n",
      "Epoch  4,  step= 589,  loss=0.171,  accuracy=0.938\n",
      "Epoch  4,  step= 590,  loss=0.140,  accuracy=0.938\n",
      "Epoch  4,  step= 591,  loss=0.349,  accuracy=0.906\n",
      "Epoch  4,  step= 592,  loss=0.064,  accuracy=0.969\n",
      "Epoch  4,  step= 593,  loss=0.186,  accuracy=0.969\n",
      "Epoch  4,  step= 594,  loss=0.084,  accuracy=1.000\n",
      "Epoch  4,  step= 595,  loss=0.135,  accuracy=0.969\n",
      "Epoch  4,  step= 596,  loss=0.507,  accuracy=0.906\n",
      "Epoch  4,  step= 597,  loss=0.445,  accuracy=0.875\n",
      "Epoch  4,  step= 598,  loss=0.416,  accuracy=0.812\n",
      "Epoch  4,  step= 599,  loss=0.225,  accuracy=0.906\n",
      "Epoch  4,  step= 600,  loss=0.248,  accuracy=0.906\n",
      "Epoch  4,  step= 601,  loss=0.336,  accuracy=0.969\n",
      "Epoch  4,  step= 602,  loss=0.215,  accuracy=0.969\n",
      "Epoch  4,  step= 603,  loss=0.223,  accuracy=0.906\n",
      "Epoch  4,  step= 604,  loss=0.390,  accuracy=0.906\n",
      "Epoch  4,  step= 605,  loss=0.560,  accuracy=0.812\n",
      "Epoch  4,  step= 606,  loss=0.282,  accuracy=0.906\n",
      "Epoch  4,  step= 607,  loss=0.084,  accuracy=0.969\n",
      "Epoch  4,  step= 608,  loss=0.213,  accuracy=0.938\n",
      "Epoch  4,  step= 609,  loss=0.612,  accuracy=0.812\n",
      "Epoch  4,  step= 610,  loss=0.276,  accuracy=0.875\n",
      "Epoch  4,  step= 611,  loss=0.326,  accuracy=0.875\n",
      "Epoch  4,  step= 612,  loss=0.285,  accuracy=0.906\n",
      "Epoch  4,  step= 613,  loss=0.322,  accuracy=0.938\n",
      "Epoch  4,  step= 614,  loss=0.135,  accuracy=0.969\n",
      "Epoch  4,  step= 615,  loss=0.143,  accuracy=0.969\n",
      "Epoch  4,  step= 616,  loss=0.048,  accuracy=1.000\n",
      "Epoch  4,  step= 617,  loss=0.085,  accuracy=0.969\n",
      "Epoch  4,  step= 618,  loss=0.284,  accuracy=0.906\n",
      "Epoch  4,  step= 619,  loss=0.885,  accuracy=0.812\n",
      "Epoch  4,  step= 620,  loss=0.345,  accuracy=0.938\n",
      "Epoch  4,  step= 621,  loss=0.111,  accuracy=0.969\n",
      "Epoch  4,  step= 622,  loss=0.481,  accuracy=0.844\n",
      "Epoch  4,  step= 623,  loss=0.158,  accuracy=0.969\n",
      "Epoch  4,  step= 624,  loss=0.325,  accuracy=0.875\n",
      "Epoch  4,  step= 625,  loss=0.235,  accuracy=0.906\n",
      "Epoch  4,  step= 626,  loss=0.329,  accuracy=0.875\n",
      "Epoch  4,  step= 627,  loss=0.276,  accuracy=0.906\n",
      "Epoch  4,  step= 628,  loss=0.201,  accuracy=0.938\n",
      "Epoch  4,  step= 629,  loss=0.218,  accuracy=0.906\n",
      "Epoch  4,  step= 630,  loss=0.642,  accuracy=0.781\n",
      "Epoch  4,  step= 631,  loss=0.135,  accuracy=0.969\n",
      "Epoch  4,  step= 632,  loss=0.506,  accuracy=0.906\n",
      "Epoch  4,  step= 633,  loss=0.144,  accuracy=0.969\n",
      "Epoch  4,  step= 634,  loss=0.147,  accuracy=0.969\n",
      "Epoch  4,  step= 635,  loss=0.432,  accuracy=0.938\n",
      "Epoch  4,  step= 636,  loss=0.153,  accuracy=0.938\n",
      "Epoch  4,  step= 637,  loss=0.142,  accuracy=0.938\n",
      "Epoch  4,  step= 638,  loss=0.091,  accuracy=0.969\n",
      "Epoch  4,  step= 639,  loss=0.174,  accuracy=0.969\n",
      "Epoch  4,  step= 640,  loss=0.024,  accuracy=1.000\n",
      "Epoch  4,  step= 641,  loss=0.326,  accuracy=0.906\n",
      "Epoch  4,  step= 642,  loss=0.605,  accuracy=0.844\n",
      "Epoch  4,  step= 643,  loss=0.169,  accuracy=0.906\n",
      "Epoch  4,  step= 644,  loss=0.243,  accuracy=0.938\n",
      "Epoch  4,  step= 645,  loss=0.296,  accuracy=0.906\n",
      "Epoch  4,  step= 646,  loss=0.245,  accuracy=0.969\n",
      "Epoch  4,  step= 647,  loss=0.780,  accuracy=0.812\n",
      "Epoch  4,  step= 648,  loss=0.242,  accuracy=0.938\n",
      "Epoch  4,  step= 649,  loss=0.782,  accuracy=0.906\n",
      "Epoch  4,  step= 650,  loss=0.420,  accuracy=0.906\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  4,  step= 651,  loss=0.130,  accuracy=0.969\n",
      "Epoch  4,  step= 652,  loss=0.317,  accuracy=0.938\n",
      "Epoch  4,  step= 653,  loss=0.632,  accuracy=0.812\n",
      "Epoch  4,  step= 654,  loss=0.410,  accuracy=0.906\n",
      "Epoch  4,  step= 655,  loss=0.552,  accuracy=0.844\n",
      "Epoch  4,  step= 656,  loss=0.528,  accuracy=0.938\n",
      "Epoch  4,  step= 657,  loss=0.149,  accuracy=0.938\n",
      "Epoch  4,  step= 658,  loss=0.263,  accuracy=0.969\n",
      "Epoch  4,  step= 659,  loss=0.184,  accuracy=0.938\n",
      "Epoch  4,  step= 660,  loss=0.564,  accuracy=0.875\n",
      "Epoch  4,  step= 661,  loss=0.283,  accuracy=0.938\n",
      "Epoch  4,  step= 662,  loss=0.610,  accuracy=0.875\n",
      "Epoch  4,  step= 663,  loss=0.063,  accuracy=1.000\n",
      "Epoch  4,  step= 664,  loss=0.086,  accuracy=0.969\n",
      "Epoch  4,  step= 665,  loss=0.372,  accuracy=0.906\n",
      "Epoch  4,  step= 666,  loss=0.217,  accuracy=0.969\n",
      "Epoch  4,  step= 667,  loss=0.305,  accuracy=0.938\n",
      "Epoch  4,  step= 668,  loss=0.249,  accuracy=0.969\n",
      "Epoch  4,  step= 669,  loss=0.330,  accuracy=0.969\n",
      "Epoch  4,  step= 670,  loss=0.216,  accuracy=0.906\n",
      "Epoch  4,  step= 671,  loss=0.146,  accuracy=0.969\n",
      "Epoch  4,  step= 672,  loss=0.233,  accuracy=0.875\n",
      "Epoch  4,  step= 673,  loss=0.174,  accuracy=0.938\n",
      "Epoch  4,  step= 674,  loss=0.536,  accuracy=0.875\n",
      "Epoch  4,  step= 675,  loss=0.489,  accuracy=0.938\n",
      "Epoch  4,  step= 676,  loss=0.225,  accuracy=0.938\n",
      "Epoch  4,  step= 677,  loss=0.165,  accuracy=0.938\n",
      "Epoch  4,  step= 678,  loss=0.324,  accuracy=0.875\n",
      "Epoch  4,  step= 679,  loss=0.194,  accuracy=0.938\n",
      "Epoch  4,  step= 680,  loss=0.219,  accuracy=0.938\n",
      "Epoch  4,  step= 681,  loss=0.053,  accuracy=1.000\n",
      "Epoch  4,  step= 682,  loss=0.052,  accuracy=1.000\n",
      "Epoch  4,  step= 683,  loss=0.070,  accuracy=1.000\n",
      "Epoch  4,  step= 684,  loss=0.301,  accuracy=0.906\n",
      "Epoch  4,  step= 685,  loss=0.214,  accuracy=0.906\n",
      "Epoch  4,  step= 686,  loss=0.189,  accuracy=0.969\n",
      "Epoch  4,  step= 687,  loss=0.203,  accuracy=0.938\n",
      "Epoch  4,  step= 688,  loss=0.163,  accuracy=0.938\n",
      "Epoch  4,  step= 689,  loss=0.213,  accuracy=0.906\n",
      "Epoch  4,  step= 690,  loss=0.239,  accuracy=0.906\n",
      "Epoch  4,  step= 691,  loss=0.422,  accuracy=0.844\n",
      "Epoch  4,  step= 692,  loss=0.409,  accuracy=0.906\n",
      "Epoch  4,  step= 693,  loss=0.405,  accuracy=0.844\n",
      "Epoch  4,  step= 694,  loss=1.003,  accuracy=0.906\n",
      "Epoch  4,  step= 695,  loss=0.054,  accuracy=1.000\n",
      "Epoch  4,  step= 696,  loss=0.356,  accuracy=0.906\n",
      "Epoch  4,  step= 697,  loss=0.125,  accuracy=0.938\n",
      "Epoch  4,  step= 698,  loss=0.030,  accuracy=1.000\n",
      "Epoch  4,  step= 699,  loss=0.135,  accuracy=0.938\n",
      "Epoch  4,  step= 700,  loss=0.229,  accuracy=0.906\n",
      "Epoch  4,  step= 701,  loss=0.058,  accuracy=1.000\n",
      "Epoch  4,  step= 702,  loss=0.789,  accuracy=0.812\n",
      "Epoch  4,  step= 703,  loss=0.566,  accuracy=0.844\n",
      "Epoch  4,  step= 704,  loss=0.603,  accuracy=0.875\n",
      "Epoch  4,  step= 705,  loss=0.592,  accuracy=0.781\n",
      "Epoch  4,  step= 706,  loss=0.472,  accuracy=0.781\n",
      "Epoch  4,  step= 707,  loss=0.280,  accuracy=0.906\n",
      "Epoch  4,  step= 708,  loss=0.255,  accuracy=0.875\n",
      "Epoch  4,  step= 709,  loss=0.433,  accuracy=0.844\n",
      "Epoch  4,  step= 710,  loss=0.289,  accuracy=0.844\n",
      "Epoch  4,  step= 711,  loss=0.463,  accuracy=0.906\n",
      "Epoch  4,  step= 712,  loss=0.472,  accuracy=0.875\n",
      "Epoch  4,  step= 713,  loss=0.152,  accuracy=0.969\n",
      "Epoch  4,  step= 714,  loss=0.209,  accuracy=0.906\n",
      "Epoch  4,  step= 715,  loss=0.218,  accuracy=0.938\n",
      "Epoch  4,  step= 716,  loss=0.099,  accuracy=0.969\n",
      "Epoch  4,  step= 717,  loss=0.024,  accuracy=1.000\n",
      "Epoch  4,  step= 718,  loss=0.070,  accuracy=0.969\n",
      "Epoch  4,  step= 719,  loss=0.432,  accuracy=0.875\n",
      "Epoch  4,  step= 720,  loss=0.184,  accuracy=0.938\n",
      "Epoch  4,  step= 721,  loss=0.554,  accuracy=0.844\n",
      "Epoch  4,  step= 722,  loss=0.104,  accuracy=0.969\n",
      "Epoch  4,  step= 723,  loss=0.228,  accuracy=0.938\n",
      "Epoch  4,  step= 724,  loss=0.189,  accuracy=0.969\n",
      "Epoch  4,  step= 725,  loss=0.421,  accuracy=0.875\n",
      "Epoch  4,  step= 726,  loss=0.191,  accuracy=0.906\n",
      "Epoch  4,  step= 727,  loss=0.137,  accuracy=0.938\n",
      "Epoch  4,  step= 728,  loss=0.206,  accuracy=0.938\n",
      "Epoch  4,  step= 729,  loss=0.255,  accuracy=0.969\n",
      "Epoch  4,  step= 730,  loss=0.628,  accuracy=0.906\n",
      "Epoch  4,  step= 731,  loss=0.223,  accuracy=0.875\n",
      "Epoch  4,  step= 732,  loss=0.138,  accuracy=1.000\n",
      "Epoch  4,  step= 733,  loss=0.447,  accuracy=0.844\n",
      "Epoch  4,  step= 734,  loss=0.329,  accuracy=0.906\n",
      "Epoch  4,  step= 735,  loss=0.108,  accuracy=0.938\n",
      "Epoch  4,  step= 736,  loss=0.180,  accuracy=0.938\n",
      "Epoch  4,  step= 737,  loss=0.513,  accuracy=0.906\n",
      "Epoch  4,  step= 738,  loss=0.269,  accuracy=0.938\n",
      "Epoch  4,  step= 739,  loss=0.161,  accuracy=0.938\n",
      "Epoch  4,  step= 740,  loss=0.371,  accuracy=0.844\n",
      "Epoch  4,  step= 741,  loss=0.874,  accuracy=0.812\n",
      "Epoch  4,  step= 742,  loss=0.104,  accuracy=0.969\n",
      "Epoch  4,  step= 743,  loss=0.190,  accuracy=0.938\n",
      "Epoch  4,  step= 744,  loss=0.432,  accuracy=0.906\n",
      "Epoch  4,  step= 745,  loss=0.401,  accuracy=0.875\n",
      "Epoch  4,  step= 746,  loss=0.238,  accuracy=0.938\n",
      "Epoch  4,  step= 747,  loss=0.165,  accuracy=0.906\n",
      "Epoch  4,  step= 748,  loss=0.580,  accuracy=0.938\n",
      "Epoch  4,  step= 749,  loss=0.086,  accuracy=1.000\n",
      "Epoch  4,  step= 750,  loss=0.229,  accuracy=0.906\n",
      "Epoch  4,  step= 751,  loss=0.282,  accuracy=0.875\n",
      "Epoch  4,  step= 752,  loss=0.086,  accuracy=0.969\n",
      "Epoch  4,  step= 753,  loss=0.444,  accuracy=0.844\n",
      "Epoch  4,  step= 754,  loss=0.144,  accuracy=0.969\n",
      "Epoch  4,  step= 755,  loss=0.375,  accuracy=0.875\n",
      "Epoch  4,  step= 756,  loss=0.577,  accuracy=0.844\n",
      "Epoch  4,  step= 757,  loss=0.092,  accuracy=1.000\n",
      "Epoch  4,  step= 758,  loss=0.570,  accuracy=0.844\n",
      "Epoch  4,  step= 759,  loss=0.173,  accuracy=0.969\n",
      "Epoch  4,  step= 760,  loss=0.129,  accuracy=0.906\n",
      "Epoch  4,  step= 761,  loss=0.127,  accuracy=0.938\n",
      "Epoch  4,  step= 762,  loss=0.179,  accuracy=0.969\n",
      "Epoch  4,  step= 763,  loss=0.175,  accuracy=0.938\n",
      "Epoch  4,  step= 764,  loss=0.234,  accuracy=0.875\n",
      "Epoch  4,  step= 765,  loss=0.191,  accuracy=0.938\n",
      "Epoch  4,  step= 766,  loss=0.402,  accuracy=0.844\n",
      "Epoch  4,  step= 767,  loss=0.195,  accuracy=0.938\n",
      "Epoch  4,  step= 768,  loss=0.404,  accuracy=0.844\n",
      "Epoch  4,  step= 769,  loss=0.598,  accuracy=0.844\n",
      "Epoch  4,  step= 770,  loss=0.634,  accuracy=0.844\n",
      "Epoch  4,  step= 771,  loss=0.334,  accuracy=0.906\n",
      "Epoch  4,  step= 772,  loss=0.640,  accuracy=0.781\n",
      "Epoch  4,  step= 773,  loss=0.235,  accuracy=0.938\n",
      "Epoch  4,  step= 774,  loss=0.761,  accuracy=0.875\n",
      "Epoch  4,  step= 775,  loss=0.392,  accuracy=0.812\n",
      "Epoch  4,  step= 776,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step= 777,  loss=0.316,  accuracy=0.875\n",
      "Epoch  4,  step= 778,  loss=0.215,  accuracy=0.906\n",
      "Epoch  4,  step= 779,  loss=0.598,  accuracy=0.875\n",
      "Epoch  4,  step= 780,  loss=0.247,  accuracy=0.875\n",
      "Epoch  4,  step= 781,  loss=0.160,  accuracy=0.938\n",
      "Epoch  4,  step= 782,  loss=0.169,  accuracy=0.969\n",
      "Epoch  4,  step= 783,  loss=0.075,  accuracy=1.000\n",
      "Epoch  4,  step= 784,  loss=0.589,  accuracy=0.906\n",
      "Epoch  4,  step= 785,  loss=0.400,  accuracy=0.844\n",
      "Epoch  4,  step= 786,  loss=0.504,  accuracy=0.875\n",
      "Epoch  4,  step= 787,  loss=0.285,  accuracy=0.875\n",
      "Epoch  4,  step= 788,  loss=0.366,  accuracy=0.938\n",
      "Epoch  4,  step= 789,  loss=0.138,  accuracy=0.938\n",
      "Epoch  4,  step= 790,  loss=0.441,  accuracy=0.812\n",
      "Epoch  4,  step= 791,  loss=0.334,  accuracy=0.906\n",
      "Epoch  4,  step= 792,  loss=0.077,  accuracy=0.969\n",
      "Epoch  4,  step= 793,  loss=0.264,  accuracy=0.938\n",
      "Epoch  4,  step= 794,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step= 795,  loss=0.454,  accuracy=0.938\n",
      "Epoch  4,  step= 796,  loss=0.430,  accuracy=0.875\n",
      "Epoch  4,  step= 797,  loss=0.190,  accuracy=0.938\n",
      "Epoch  4,  step= 798,  loss=0.489,  accuracy=0.938\n",
      "Epoch  4,  step= 799,  loss=0.248,  accuracy=0.875\n",
      "Epoch  4,  step= 800,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step= 801,  loss=0.248,  accuracy=0.906\n",
      "Epoch  4,  step= 802,  loss=0.335,  accuracy=0.969\n",
      "Epoch  4,  step= 803,  loss=0.184,  accuracy=0.938\n",
      "Epoch  4,  step= 804,  loss=0.292,  accuracy=0.969\n",
      "Epoch  4,  step= 805,  loss=0.346,  accuracy=0.969\n",
      "Epoch  4,  step= 806,  loss=0.866,  accuracy=0.781\n",
      "Epoch  4,  step= 807,  loss=0.172,  accuracy=0.938\n",
      "Epoch  4,  step= 808,  loss=0.074,  accuracy=0.969\n",
      "Epoch  4,  step= 809,  loss=0.357,  accuracy=0.906\n",
      "Epoch  4,  step= 810,  loss=0.343,  accuracy=0.875\n",
      "Epoch  4,  step= 811,  loss=0.316,  accuracy=0.875\n",
      "Epoch  4,  step= 812,  loss=0.120,  accuracy=0.938\n",
      "Epoch  4,  step= 813,  loss=0.074,  accuracy=0.969\n",
      "Epoch  4,  step= 814,  loss=0.314,  accuracy=0.875\n",
      "Epoch  4,  step= 815,  loss=0.102,  accuracy=0.969\n",
      "Epoch  4,  step= 816,  loss=0.175,  accuracy=0.938\n",
      "Epoch  4,  step= 817,  loss=0.045,  accuracy=1.000\n",
      "Epoch  4,  step= 818,  loss=0.247,  accuracy=0.969\n",
      "Epoch  4,  step= 819,  loss=0.172,  accuracy=0.906\n",
      "Epoch  4,  step= 820,  loss=0.151,  accuracy=0.938\n",
      "Epoch  4,  step= 821,  loss=0.374,  accuracy=0.844\n",
      "Epoch  4,  step= 822,  loss=0.148,  accuracy=0.938\n",
      "Epoch  4,  step= 823,  loss=0.543,  accuracy=0.781\n",
      "Epoch  4,  step= 824,  loss=0.822,  accuracy=0.906\n",
      "Epoch  4,  step= 825,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step= 826,  loss=0.294,  accuracy=0.906\n",
      "Epoch  4,  step= 827,  loss=0.247,  accuracy=0.938\n",
      "Epoch  4,  step= 828,  loss=0.326,  accuracy=0.906\n",
      "Epoch  4,  step= 829,  loss=0.312,  accuracy=0.844\n",
      "Epoch  4,  step= 830,  loss=0.396,  accuracy=0.875\n",
      "Epoch  4,  step= 831,  loss=0.373,  accuracy=0.875\n",
      "Epoch  4,  step= 832,  loss=0.810,  accuracy=0.812\n",
      "Epoch  4,  step= 833,  loss=0.188,  accuracy=0.938\n",
      "Epoch  4,  step= 834,  loss=0.118,  accuracy=0.969\n",
      "Epoch  4,  step= 835,  loss=0.798,  accuracy=0.875\n",
      "Epoch  4,  step= 836,  loss=0.417,  accuracy=0.906\n",
      "Epoch  4,  step= 837,  loss=0.295,  accuracy=0.875\n",
      "Epoch  4,  step= 838,  loss=0.139,  accuracy=0.938\n",
      "Epoch  4,  step= 839,  loss=0.473,  accuracy=0.875\n",
      "Epoch  4,  step= 840,  loss=0.572,  accuracy=0.875\n",
      "Epoch  4,  step= 841,  loss=0.203,  accuracy=0.906\n",
      "Epoch  4,  step= 842,  loss=0.127,  accuracy=0.969\n",
      "Epoch  4,  step= 843,  loss=0.116,  accuracy=0.969\n",
      "Epoch  4,  step= 844,  loss=0.226,  accuracy=0.875\n",
      "Epoch  4,  step= 845,  loss=0.095,  accuracy=0.938\n",
      "Epoch  4,  step= 846,  loss=0.369,  accuracy=0.844\n",
      "Epoch  4,  step= 847,  loss=0.267,  accuracy=0.906\n",
      "Epoch  4,  step= 848,  loss=0.429,  accuracy=0.906\n",
      "Epoch  4,  step= 849,  loss=1.011,  accuracy=0.719\n",
      "Epoch  4,  step= 850,  loss=0.384,  accuracy=0.875\n",
      "Epoch  4,  step= 851,  loss=0.210,  accuracy=0.906\n",
      "Epoch  4,  step= 852,  loss=0.237,  accuracy=0.969\n",
      "Epoch  4,  step= 853,  loss=0.172,  accuracy=0.875\n",
      "Epoch  4,  step= 854,  loss=0.181,  accuracy=0.938\n",
      "Epoch  4,  step= 855,  loss=0.266,  accuracy=0.906\n",
      "Epoch  4,  step= 856,  loss=0.258,  accuracy=0.938\n",
      "Epoch  4,  step= 857,  loss=0.491,  accuracy=0.938\n",
      "Epoch  4,  step= 858,  loss=0.071,  accuracy=1.000\n",
      "Epoch  4,  step= 859,  loss=0.361,  accuracy=0.906\n",
      "Epoch  4,  step= 860,  loss=0.136,  accuracy=0.969\n",
      "Epoch  4,  step= 861,  loss=0.101,  accuracy=0.938\n",
      "Epoch  4,  step= 862,  loss=0.216,  accuracy=0.906\n",
      "Epoch  4,  step= 863,  loss=0.209,  accuracy=0.969\n",
      "Epoch  4,  step= 864,  loss=0.179,  accuracy=0.938\n",
      "Epoch  4,  step= 865,  loss=0.775,  accuracy=0.812\n",
      "Epoch  4,  step= 866,  loss=0.425,  accuracy=0.844\n",
      "Epoch  4,  step= 867,  loss=0.060,  accuracy=1.000\n",
      "Epoch  4,  step= 868,  loss=0.163,  accuracy=0.938\n",
      "Epoch  4,  step= 869,  loss=0.260,  accuracy=0.938\n",
      "Epoch  4,  step= 870,  loss=0.333,  accuracy=0.906\n",
      "Epoch  4,  step= 871,  loss=0.620,  accuracy=0.812\n",
      "Epoch  4,  step= 872,  loss=0.047,  accuracy=1.000\n",
      "Epoch  4,  step= 873,  loss=0.266,  accuracy=0.906\n",
      "Epoch  4,  step= 874,  loss=0.257,  accuracy=0.938\n",
      "Epoch  4,  step= 875,  loss=0.551,  accuracy=0.906\n",
      "Epoch  4,  step= 876,  loss=0.080,  accuracy=0.969\n",
      "Epoch  4,  step= 877,  loss=0.303,  accuracy=0.938\n",
      "Epoch  4,  step= 878,  loss=0.206,  accuracy=0.938\n",
      "Epoch  4,  step= 879,  loss=0.221,  accuracy=0.938\n",
      "Epoch  4,  step= 880,  loss=0.264,  accuracy=0.938\n",
      "Epoch  4,  step= 881,  loss=0.394,  accuracy=0.938\n",
      "Epoch  4,  step= 882,  loss=0.356,  accuracy=0.938\n",
      "Epoch  4,  step= 883,  loss=0.134,  accuracy=0.969\n",
      "Epoch  4,  step= 884,  loss=0.228,  accuracy=0.938\n",
      "Epoch  4,  step= 885,  loss=0.039,  accuracy=1.000\n",
      "Epoch  4,  step= 886,  loss=1.299,  accuracy=0.781\n",
      "Epoch  4,  step= 887,  loss=0.290,  accuracy=0.969\n",
      "Epoch  4,  step= 888,  loss=0.106,  accuracy=0.969\n",
      "Epoch  4,  step= 889,  loss=0.182,  accuracy=0.938\n",
      "Epoch  4,  step= 890,  loss=0.648,  accuracy=0.906\n",
      "Epoch  4,  step= 891,  loss=0.360,  accuracy=0.906\n",
      "Epoch  4,  step= 892,  loss=0.299,  accuracy=0.969\n",
      "Epoch  4,  step= 893,  loss=0.203,  accuracy=0.969\n",
      "Epoch  4,  step= 894,  loss=0.350,  accuracy=0.781\n",
      "Epoch  4,  step= 895,  loss=1.000,  accuracy=0.781\n",
      "Epoch  4,  step= 896,  loss=0.104,  accuracy=0.969\n",
      "Epoch  4,  step= 897,  loss=0.160,  accuracy=0.938\n",
      "Epoch  4,  step= 898,  loss=0.074,  accuracy=1.000\n",
      "Epoch  4,  step= 899,  loss=0.355,  accuracy=0.812\n",
      "Epoch  4,  step= 900,  loss=0.168,  accuracy=0.938\n",
      "Epoch  4,  step= 901,  loss=0.253,  accuracy=0.906\n",
      "Epoch  4,  step= 902,  loss=0.227,  accuracy=0.906\n",
      "Epoch  4,  step= 903,  loss=0.215,  accuracy=0.875\n",
      "Epoch  4,  step= 904,  loss=0.209,  accuracy=0.938\n",
      "Epoch  4,  step= 905,  loss=0.240,  accuracy=0.938\n",
      "Epoch  4,  step= 906,  loss=0.373,  accuracy=0.875\n",
      "Epoch  4,  step= 907,  loss=0.256,  accuracy=0.938\n",
      "Epoch  4,  step= 908,  loss=0.265,  accuracy=0.875\n",
      "Epoch  4,  step= 909,  loss=0.400,  accuracy=0.844\n",
      "Epoch  4,  step= 910,  loss=0.593,  accuracy=0.875\n",
      "Epoch  4,  step= 911,  loss=0.381,  accuracy=0.812\n",
      "Epoch  4,  step= 912,  loss=0.320,  accuracy=0.906\n",
      "Epoch  4,  step= 913,  loss=0.295,  accuracy=0.906\n",
      "Epoch  4,  step= 914,  loss=0.143,  accuracy=0.938\n",
      "Epoch  4,  step= 915,  loss=0.483,  accuracy=0.781\n",
      "Epoch  4,  step= 916,  loss=0.493,  accuracy=0.906\n",
      "Epoch  4,  step= 917,  loss=0.212,  accuracy=0.906\n",
      "Epoch  4,  step= 918,  loss=0.214,  accuracy=0.938\n",
      "Epoch  4,  step= 919,  loss=0.452,  accuracy=0.938\n",
      "Epoch  4,  step= 920,  loss=0.149,  accuracy=0.938\n",
      "Epoch  4,  step= 921,  loss=0.130,  accuracy=0.969\n",
      "Epoch  4,  step= 922,  loss=0.176,  accuracy=0.969\n",
      "Epoch  4,  step= 923,  loss=0.103,  accuracy=1.000\n",
      "Epoch  4,  step= 924,  loss=0.487,  accuracy=0.906\n",
      "Epoch  4,  step= 925,  loss=0.171,  accuracy=0.938\n",
      "Epoch  4,  step= 926,  loss=0.239,  accuracy=0.906\n",
      "Epoch  4,  step= 927,  loss=0.269,  accuracy=0.906\n",
      "Epoch  4,  step= 928,  loss=0.495,  accuracy=0.875\n",
      "Epoch  4,  step= 929,  loss=0.372,  accuracy=0.906\n",
      "Epoch  4,  step= 930,  loss=0.141,  accuracy=0.969\n",
      "Epoch  4,  step= 931,  loss=0.151,  accuracy=0.969\n",
      "Epoch  4,  step= 932,  loss=0.581,  accuracy=0.781\n",
      "Epoch  4,  step= 933,  loss=0.407,  accuracy=0.938\n",
      "Epoch  4,  step= 934,  loss=0.108,  accuracy=1.000\n",
      "Epoch  4,  step= 935,  loss=0.526,  accuracy=0.875\n",
      "Epoch  4,  step= 936,  loss=0.442,  accuracy=0.844\n",
      "Epoch  4,  step= 937,  loss=0.267,  accuracy=0.906\n",
      "Epoch  4,  step= 938,  loss=0.266,  accuracy=0.875\n",
      "Epoch  4,  step= 939,  loss=0.392,  accuracy=0.906\n",
      "Epoch  4,  step= 940,  loss=0.349,  accuracy=0.875\n",
      "Epoch  4,  step= 941,  loss=0.521,  accuracy=0.875\n",
      "Epoch  4,  step= 942,  loss=0.449,  accuracy=0.875\n",
      "Epoch  4,  step= 943,  loss=0.412,  accuracy=0.875\n",
      "Epoch  4,  step= 944,  loss=0.313,  accuracy=0.938\n",
      "Epoch  4,  step= 945,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step= 946,  loss=0.053,  accuracy=1.000\n",
      "Epoch  4,  step= 947,  loss=0.301,  accuracy=0.906\n",
      "Epoch  4,  step= 948,  loss=0.169,  accuracy=0.906\n",
      "Epoch  4,  step= 949,  loss=0.174,  accuracy=0.938\n",
      "Epoch  4,  step= 950,  loss=0.240,  accuracy=0.906\n",
      "Epoch  4,  step= 951,  loss=0.442,  accuracy=0.844\n",
      "Epoch  4,  step= 952,  loss=0.195,  accuracy=0.969\n",
      "Epoch  4,  step= 953,  loss=0.343,  accuracy=0.844\n",
      "Epoch  4,  step= 954,  loss=0.480,  accuracy=0.906\n",
      "Epoch  4,  step= 955,  loss=0.286,  accuracy=0.938\n",
      "Epoch  4,  step= 956,  loss=0.666,  accuracy=0.875\n",
      "Epoch  4,  step= 957,  loss=0.199,  accuracy=0.906\n",
      "Epoch  4,  step= 958,  loss=0.315,  accuracy=0.844\n",
      "Epoch  4,  step= 959,  loss=0.307,  accuracy=0.875\n",
      "Epoch  4,  step= 960,  loss=0.415,  accuracy=0.906\n",
      "Epoch  4,  step= 961,  loss=0.086,  accuracy=0.969\n",
      "Epoch  4,  step= 962,  loss=0.301,  accuracy=0.906\n",
      "Epoch  4,  step= 963,  loss=0.408,  accuracy=0.844\n",
      "Epoch  4,  step= 964,  loss=0.216,  accuracy=0.906\n",
      "Epoch  4,  step= 965,  loss=0.609,  accuracy=0.875\n",
      "Epoch  4,  step= 966,  loss=0.120,  accuracy=0.938\n",
      "Epoch  4,  step= 967,  loss=0.553,  accuracy=0.875\n",
      "Epoch  4,  step= 968,  loss=0.232,  accuracy=0.906\n",
      "Epoch  4,  step= 969,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step= 970,  loss=0.487,  accuracy=0.938\n",
      "Epoch  4,  step= 971,  loss=0.533,  accuracy=0.875\n",
      "Epoch  4,  step= 972,  loss=0.590,  accuracy=0.844\n",
      "Epoch  4,  step= 973,  loss=0.283,  accuracy=0.938\n",
      "Epoch  4,  step= 974,  loss=0.735,  accuracy=0.875\n",
      "Epoch  4,  step= 975,  loss=0.123,  accuracy=0.969\n",
      "Epoch  4,  step= 976,  loss=0.203,  accuracy=0.906\n",
      "Epoch  4,  step= 977,  loss=0.706,  accuracy=0.781\n",
      "Epoch  4,  step= 978,  loss=0.753,  accuracy=0.750\n",
      "Epoch  4,  step= 979,  loss=0.425,  accuracy=0.812\n",
      "Epoch  4,  step= 980,  loss=0.553,  accuracy=0.812\n",
      "Epoch  4,  step= 981,  loss=0.802,  accuracy=0.844\n",
      "Epoch  4,  step= 982,  loss=0.379,  accuracy=0.875\n",
      "Epoch  4,  step= 983,  loss=0.152,  accuracy=0.906\n",
      "Epoch  4,  step= 984,  loss=0.150,  accuracy=0.938\n",
      "Epoch  4,  step= 985,  loss=0.127,  accuracy=0.969\n",
      "Epoch  4,  step= 986,  loss=0.236,  accuracy=0.969\n",
      "Epoch  4,  step= 987,  loss=0.353,  accuracy=0.906\n",
      "Epoch  4,  step= 988,  loss=0.380,  accuracy=0.844\n",
      "Epoch  4,  step= 989,  loss=0.987,  accuracy=0.750\n",
      "Epoch  4,  step= 990,  loss=0.271,  accuracy=0.875\n",
      "Epoch  4,  step= 991,  loss=0.686,  accuracy=0.875\n",
      "Epoch  4,  step= 992,  loss=0.187,  accuracy=0.938\n",
      "Epoch  4,  step= 993,  loss=0.264,  accuracy=0.844\n",
      "Epoch  4,  step= 994,  loss=0.139,  accuracy=0.938\n",
      "Epoch  4,  step= 995,  loss=0.494,  accuracy=0.844\n",
      "Epoch  4,  step= 996,  loss=0.182,  accuracy=0.938\n",
      "Epoch  4,  step= 997,  loss=0.299,  accuracy=0.938\n",
      "Epoch  4,  step= 998,  loss=0.424,  accuracy=0.906\n",
      "Epoch  4,  step= 999,  loss=0.186,  accuracy=0.969\n",
      "Epoch  4,  step=1000,  loss=0.298,  accuracy=0.938\n",
      "Epoch  4,  step=1001,  loss=0.276,  accuracy=0.875\n",
      "Epoch  4,  step=1002,  loss=0.169,  accuracy=0.938\n",
      "Epoch  4,  step=1003,  loss=0.163,  accuracy=0.938\n",
      "Epoch  4,  step=1004,  loss=0.256,  accuracy=0.906\n",
      "Epoch  4,  step=1005,  loss=0.480,  accuracy=0.875\n",
      "Epoch  4,  step=1006,  loss=0.313,  accuracy=0.906\n",
      "Epoch  4,  step=1007,  loss=0.396,  accuracy=0.844\n",
      "Epoch  4,  step=1008,  loss=0.287,  accuracy=0.969\n",
      "Epoch  4,  step=1009,  loss=0.356,  accuracy=0.906\n",
      "Epoch  4,  step=1010,  loss=0.461,  accuracy=0.844\n",
      "Epoch  4,  step=1011,  loss=0.470,  accuracy=0.906\n",
      "Epoch  4,  step=1012,  loss=0.291,  accuracy=0.938\n",
      "Epoch  4,  step=1013,  loss=0.269,  accuracy=0.969\n",
      "Epoch  4,  step=1014,  loss=0.276,  accuracy=0.906\n",
      "Epoch  4,  step=1015,  loss=0.188,  accuracy=0.938\n",
      "Epoch  4,  step=1016,  loss=0.132,  accuracy=0.969\n",
      "Epoch  4,  step=1017,  loss=0.348,  accuracy=0.906\n",
      "Epoch  4,  step=1018,  loss=0.122,  accuracy=0.938\n",
      "Epoch  4,  step=1019,  loss=0.154,  accuracy=0.906\n",
      "Epoch  4,  step=1020,  loss=0.417,  accuracy=0.875\n",
      "Epoch  4,  step=1021,  loss=0.284,  accuracy=0.938\n",
      "Epoch  4,  step=1022,  loss=0.130,  accuracy=0.938\n",
      "Epoch  4,  step=1023,  loss=0.108,  accuracy=0.969\n",
      "Epoch  4,  step=1024,  loss=0.649,  accuracy=0.906\n",
      "Epoch  4,  step=1025,  loss=0.394,  accuracy=0.875\n",
      "Epoch  4,  step=1026,  loss=0.371,  accuracy=0.844\n",
      "Epoch  4,  step=1027,  loss=0.495,  accuracy=0.906\n",
      "Epoch  4,  step=1028,  loss=0.544,  accuracy=0.875\n",
      "Epoch  4,  step=1029,  loss=0.133,  accuracy=0.969\n",
      "Epoch  4,  step=1030,  loss=0.044,  accuracy=1.000\n",
      "Epoch  4,  step=1031,  loss=0.078,  accuracy=0.969\n",
      "Epoch  4,  step=1032,  loss=0.330,  accuracy=0.875\n",
      "Epoch  4,  step=1033,  loss=0.443,  accuracy=0.938\n",
      "Epoch  4,  step=1034,  loss=0.168,  accuracy=0.906\n",
      "Epoch  4,  step=1035,  loss=0.326,  accuracy=0.906\n",
      "Epoch  4,  step=1036,  loss=0.217,  accuracy=0.938\n",
      "Epoch  4,  step=1037,  loss=0.218,  accuracy=0.906\n",
      "Epoch  4,  step=1038,  loss=0.433,  accuracy=0.906\n",
      "Epoch  4,  step=1039,  loss=0.116,  accuracy=0.938\n",
      "Epoch  4,  step=1040,  loss=0.356,  accuracy=0.906\n",
      "Epoch  4,  step=1041,  loss=0.178,  accuracy=0.969\n",
      "Epoch  4,  step=1042,  loss=0.254,  accuracy=0.844\n",
      "Epoch  4,  step=1043,  loss=0.245,  accuracy=0.938\n",
      "Epoch  4,  step=1044,  loss=0.172,  accuracy=0.906\n",
      "Epoch  4,  step=1045,  loss=0.185,  accuracy=0.875\n",
      "Epoch  4,  step=1046,  loss=0.186,  accuracy=0.906\n",
      "Epoch  4,  step=1047,  loss=0.264,  accuracy=0.938\n",
      "Epoch  4,  step=1048,  loss=0.042,  accuracy=1.000\n",
      "Epoch  4,  step=1049,  loss=0.618,  accuracy=0.875\n",
      "Epoch  4,  step=1050,  loss=0.150,  accuracy=0.938\n",
      "Epoch  4,  step=1051,  loss=0.154,  accuracy=0.938\n",
      "Epoch  4,  step=1052,  loss=0.330,  accuracy=0.906\n",
      "Epoch  4,  step=1053,  loss=0.098,  accuracy=0.969\n",
      "Epoch  4,  step=1054,  loss=0.266,  accuracy=0.969\n",
      "Epoch  4,  step=1055,  loss=0.636,  accuracy=0.812\n",
      "Epoch  4,  step=1056,  loss=0.236,  accuracy=0.906\n",
      "Epoch  4,  step=1057,  loss=0.136,  accuracy=0.938\n",
      "Epoch  4,  step=1058,  loss=0.063,  accuracy=1.000\n",
      "Epoch  4,  step=1059,  loss=0.124,  accuracy=0.969\n",
      "Epoch  4,  step=1060,  loss=0.129,  accuracy=0.938\n",
      "Epoch  4,  step=1061,  loss=0.103,  accuracy=0.969\n",
      "Epoch  4,  step=1062,  loss=0.187,  accuracy=0.969\n",
      "Epoch  4,  step=1063,  loss=0.142,  accuracy=0.938\n",
      "Epoch  4,  step=1064,  loss=0.438,  accuracy=0.906\n",
      "Epoch  4,  step=1065,  loss=0.374,  accuracy=0.844\n",
      "Epoch  4,  step=1066,  loss=0.438,  accuracy=0.906\n",
      "Epoch  4,  step=1067,  loss=0.048,  accuracy=1.000\n",
      "Epoch  4,  step=1068,  loss=0.261,  accuracy=0.938\n",
      "Epoch  4,  step=1069,  loss=0.092,  accuracy=0.969\n",
      "Epoch  4,  step=1070,  loss=0.108,  accuracy=0.969\n",
      "Epoch  4,  step=1071,  loss=0.062,  accuracy=0.969\n",
      "Epoch  4,  step=1072,  loss=0.168,  accuracy=0.938\n",
      "Epoch  4,  step=1073,  loss=0.228,  accuracy=0.906\n",
      "Epoch  4,  step=1074,  loss=0.065,  accuracy=0.969\n",
      "Epoch  4,  step=1075,  loss=0.397,  accuracy=0.844\n",
      "Epoch  4,  step=1076,  loss=0.409,  accuracy=0.875\n",
      "Epoch  4,  step=1077,  loss=0.162,  accuracy=0.938\n",
      "Epoch  4,  step=1078,  loss=0.386,  accuracy=0.969\n",
      "Epoch  4,  step=1079,  loss=0.912,  accuracy=0.844\n",
      "Epoch  4,  step=1080,  loss=0.174,  accuracy=0.938\n",
      "Epoch  4,  step=1081,  loss=0.254,  accuracy=0.906\n",
      "Epoch  4,  step=1082,  loss=0.175,  accuracy=0.938\n",
      "Epoch  4,  step=1083,  loss=0.429,  accuracy=0.844\n",
      "Epoch  4,  step=1084,  loss=0.435,  accuracy=0.844\n",
      "Epoch  4,  step=1085,  loss=0.243,  accuracy=0.938\n",
      "Epoch  4,  step=1086,  loss=0.222,  accuracy=0.938\n",
      "Epoch  4,  step=1087,  loss=0.320,  accuracy=0.938\n",
      "Epoch  4,  step=1088,  loss=0.445,  accuracy=0.875\n",
      "Epoch  4,  step=1089,  loss=0.255,  accuracy=0.938\n",
      "Epoch  4,  step=1090,  loss=0.389,  accuracy=0.906\n",
      "Epoch  4,  step=1091,  loss=0.199,  accuracy=0.906\n",
      "Epoch  4,  step=1092,  loss=0.125,  accuracy=0.938\n",
      "Epoch  4,  step=1093,  loss=0.112,  accuracy=0.969\n",
      "Epoch  4,  step=1094,  loss=0.088,  accuracy=1.000\n",
      "Epoch  4,  step=1095,  loss=0.308,  accuracy=0.844\n",
      "Epoch  4,  step=1096,  loss=0.240,  accuracy=0.969\n",
      "Epoch  4,  step=1097,  loss=0.262,  accuracy=0.906\n",
      "Epoch  4,  step=1098,  loss=0.434,  accuracy=0.844\n",
      "Epoch  4,  step=1099,  loss=0.201,  accuracy=0.938\n",
      "Epoch  4,  step=1100,  loss=0.262,  accuracy=0.938\n",
      "Epoch  4,  step=1101,  loss=0.514,  accuracy=0.750\n",
      "Epoch  4,  step=1102,  loss=0.069,  accuracy=1.000\n",
      "Epoch  4,  step=1103,  loss=0.192,  accuracy=0.906\n",
      "Epoch  4,  step=1104,  loss=0.092,  accuracy=0.969\n",
      "Epoch  4,  step=1105,  loss=0.216,  accuracy=0.969\n",
      "Epoch  4,  step=1106,  loss=0.172,  accuracy=0.938\n",
      "Epoch  4,  step=1107,  loss=0.299,  accuracy=0.906\n",
      "Epoch  4,  step=1108,  loss=0.954,  accuracy=0.844\n",
      "Epoch  4,  step=1109,  loss=0.274,  accuracy=0.906\n",
      "Epoch  4,  step=1110,  loss=0.166,  accuracy=0.938\n",
      "Epoch  4,  step=1111,  loss=0.272,  accuracy=0.906\n",
      "Epoch  4,  step=1112,  loss=0.083,  accuracy=0.969\n",
      "Epoch  4,  step=1113,  loss=0.697,  accuracy=0.750\n",
      "Epoch  4,  step=1114,  loss=0.080,  accuracy=0.969\n",
      "Epoch  4,  step=1115,  loss=0.302,  accuracy=0.938\n",
      "Epoch  4,  step=1116,  loss=0.075,  accuracy=1.000\n",
      "Epoch  4,  step=1117,  loss=0.131,  accuracy=0.969\n",
      "Epoch  4,  step=1118,  loss=0.136,  accuracy=0.969\n",
      "Epoch  4,  step=1119,  loss=0.145,  accuracy=0.969\n",
      "Epoch  4,  step=1120,  loss=0.334,  accuracy=0.938\n",
      "Epoch  4,  step=1121,  loss=0.153,  accuracy=0.938\n",
      "Epoch  4,  step=1122,  loss=0.185,  accuracy=0.969\n",
      "Epoch  4,  step=1123,  loss=0.177,  accuracy=1.000\n",
      "Epoch  4,  step=1124,  loss=0.434,  accuracy=0.938\n",
      "Epoch  4,  step=1125,  loss=0.380,  accuracy=0.844\n",
      "Epoch  4,  step=1126,  loss=0.603,  accuracy=0.875\n",
      "Epoch  4,  step=1127,  loss=0.339,  accuracy=0.844\n",
      "Epoch  4,  step=1128,  loss=0.616,  accuracy=0.875\n",
      "Epoch  4,  step=1129,  loss=0.143,  accuracy=0.969\n",
      "Epoch  4,  step=1130,  loss=0.322,  accuracy=0.906\n",
      "Epoch  4,  step=1131,  loss=0.190,  accuracy=0.875\n",
      "Epoch  4,  step=1132,  loss=0.214,  accuracy=0.875\n",
      "Epoch  4,  step=1133,  loss=0.258,  accuracy=0.969\n",
      "Epoch  4,  step=1134,  loss=0.034,  accuracy=1.000\n",
      "Epoch  4,  step=1135,  loss=0.208,  accuracy=0.906\n",
      "Epoch  4,  step=1136,  loss=0.093,  accuracy=0.969\n",
      "Epoch  4,  step=1137,  loss=0.218,  accuracy=0.906\n",
      "Epoch  4,  step=1138,  loss=0.572,  accuracy=0.844\n",
      "Epoch  4,  step=1139,  loss=0.381,  accuracy=0.844\n",
      "Epoch  4,  step=1140,  loss=0.281,  accuracy=0.938\n",
      "Epoch  4,  step=1141,  loss=0.279,  accuracy=0.875\n",
      "Epoch  4,  step=1142,  loss=0.166,  accuracy=0.938\n",
      "Epoch  4,  step=1143,  loss=0.172,  accuracy=0.938\n",
      "Epoch  4,  step=1144,  loss=0.048,  accuracy=0.969\n",
      "Epoch  4,  step=1145,  loss=0.386,  accuracy=0.875\n",
      "Epoch  4,  step=1146,  loss=0.055,  accuracy=0.969\n",
      "Epoch  4,  step=1147,  loss=0.304,  accuracy=0.938\n",
      "Epoch  4,  step=1148,  loss=0.252,  accuracy=0.875\n",
      "Epoch  4,  step=1149,  loss=0.125,  accuracy=0.969\n",
      "Epoch  4,  step=1150,  loss=0.361,  accuracy=0.938\n",
      "Epoch  4,  step=1151,  loss=0.166,  accuracy=0.938\n",
      "Epoch  4,  step=1152,  loss=0.282,  accuracy=0.938\n",
      "Epoch  4,  step=1153,  loss=0.090,  accuracy=0.969\n",
      "Epoch  4,  step=1154,  loss=0.103,  accuracy=0.969\n",
      "Epoch  4,  step=1155,  loss=0.080,  accuracy=1.000\n",
      "Epoch  4,  step=1156,  loss=0.075,  accuracy=0.969\n",
      "Epoch  4,  step=1157,  loss=0.490,  accuracy=0.906\n",
      "Epoch  4,  step=1158,  loss=0.403,  accuracy=0.938\n",
      "Epoch  4,  step=1159,  loss=0.302,  accuracy=0.938\n",
      "Epoch  4,  step=1160,  loss=0.326,  accuracy=0.875\n",
      "Epoch  4,  step=1161,  loss=0.347,  accuracy=0.906\n",
      "Epoch  4,  step=1162,  loss=0.275,  accuracy=0.875\n",
      "Epoch  4,  step=1163,  loss=0.094,  accuracy=0.938\n",
      "Epoch  4,  step=1164,  loss=0.585,  accuracy=0.844\n",
      "Epoch  4,  step=1165,  loss=0.321,  accuracy=0.906\n",
      "Epoch  4,  step=1166,  loss=0.212,  accuracy=0.875\n",
      "Epoch  4,  step=1167,  loss=0.696,  accuracy=0.875\n",
      "Epoch  4,  step=1168,  loss=1.039,  accuracy=0.906\n",
      "Epoch  4,  step=1169,  loss=0.628,  accuracy=0.844\n",
      "Epoch  4,  step=1170,  loss=0.661,  accuracy=0.844\n",
      "Epoch  4,  step=1171,  loss=0.198,  accuracy=0.938\n",
      "Epoch  4,  step=1172,  loss=0.240,  accuracy=0.875\n",
      "Epoch  4,  step=1173,  loss=0.579,  accuracy=0.750\n",
      "Epoch  4,  step=1174,  loss=0.435,  accuracy=0.844\n",
      "Epoch  4,  step=1175,  loss=0.296,  accuracy=0.938\n",
      "Epoch  4,  step=1176,  loss=0.196,  accuracy=0.906\n",
      "Epoch  4,  step=1177,  loss=0.194,  accuracy=0.938\n",
      "Epoch  4,  step=1178,  loss=0.488,  accuracy=0.875\n",
      "Epoch  4,  step=1179,  loss=0.556,  accuracy=0.906\n",
      "Epoch  4,  step=1180,  loss=0.260,  accuracy=0.875\n",
      "Epoch  4,  step=1181,  loss=0.584,  accuracy=0.906\n",
      "Epoch  4,  step=1182,  loss=0.455,  accuracy=0.906\n",
      "Epoch  4,  step=1183,  loss=0.190,  accuracy=0.906\n",
      "Epoch  4,  step=1184,  loss=0.289,  accuracy=0.812\n",
      "Epoch  4,  step=1185,  loss=0.273,  accuracy=0.906\n",
      "Epoch  4,  step=1186,  loss=0.196,  accuracy=0.938\n",
      "Epoch  4,  step=1187,  loss=0.211,  accuracy=0.906\n",
      "Epoch  4,  step=1188,  loss=0.329,  accuracy=0.969\n",
      "Epoch  4,  step=1189,  loss=0.237,  accuracy=0.938\n",
      "Epoch  4,  step=1190,  loss=0.219,  accuracy=0.938\n",
      "Epoch  4,  step=1191,  loss=0.298,  accuracy=0.906\n",
      "Epoch  4,  step=1192,  loss=0.198,  accuracy=0.906\n",
      "Epoch  4,  step=1193,  loss=0.142,  accuracy=0.969\n",
      "Epoch  4,  step=1194,  loss=0.203,  accuracy=0.938\n",
      "Epoch  4,  step=1195,  loss=0.416,  accuracy=0.906\n",
      "Epoch  4,  step=1196,  loss=0.125,  accuracy=0.969\n",
      "Epoch  4,  step=1197,  loss=0.160,  accuracy=0.938\n",
      "Epoch  4,  step=1198,  loss=0.156,  accuracy=0.969\n",
      "Epoch  4,  step=1199,  loss=0.641,  accuracy=0.750\n",
      "Epoch  4,  step=1200,  loss=0.210,  accuracy=0.938\n",
      "Epoch  4,  step=1201,  loss=0.278,  accuracy=0.906\n",
      "Epoch  4,  step=1202,  loss=0.065,  accuracy=0.969\n",
      "Epoch  4,  step=1203,  loss=0.694,  accuracy=0.844\n",
      "Epoch  4,  step=1204,  loss=0.096,  accuracy=0.969\n",
      "Epoch  4,  step=1205,  loss=0.300,  accuracy=0.906\n",
      "Epoch  4,  step=1206,  loss=0.569,  accuracy=0.875\n",
      "Epoch  4,  step=1207,  loss=0.222,  accuracy=0.938\n",
      "Epoch  4,  step=1208,  loss=0.357,  accuracy=0.906\n",
      "Epoch  4,  step=1209,  loss=0.141,  accuracy=0.969\n",
      "Epoch  4,  step=1210,  loss=0.143,  accuracy=0.969\n",
      "Epoch  4,  step=1211,  loss=0.429,  accuracy=0.875\n",
      "Epoch  4,  step=1212,  loss=0.262,  accuracy=0.875\n",
      "Epoch  4,  step=1213,  loss=0.056,  accuracy=0.969\n",
      "Epoch  4,  step=1214,  loss=0.065,  accuracy=1.000\n",
      "Epoch  4,  step=1215,  loss=0.135,  accuracy=0.938\n",
      "Epoch  4,  step=1216,  loss=0.180,  accuracy=0.938\n",
      "Epoch  4,  step=1217,  loss=0.133,  accuracy=0.969\n",
      "Epoch  4,  step=1218,  loss=0.280,  accuracy=0.875\n",
      "Epoch  4,  step=1219,  loss=0.119,  accuracy=0.969\n",
      "Epoch  4,  step=1220,  loss=0.217,  accuracy=0.906\n",
      "Epoch  4,  step=1221,  loss=0.061,  accuracy=1.000\n",
      "Epoch  4,  step=1222,  loss=0.421,  accuracy=0.844\n",
      "Epoch  4,  step=1223,  loss=0.281,  accuracy=0.875\n",
      "Epoch  4,  step=1224,  loss=0.194,  accuracy=0.969\n",
      "Epoch  4,  step=1225,  loss=0.387,  accuracy=0.938\n",
      "Epoch  4,  step=1226,  loss=0.345,  accuracy=0.938\n",
      "Epoch  4,  step=1227,  loss=0.243,  accuracy=0.938\n",
      "Epoch  4,  step=1228,  loss=1.071,  accuracy=0.812\n",
      "Epoch  4,  step=1229,  loss=0.615,  accuracy=0.781\n",
      "Epoch  4,  step=1230,  loss=0.266,  accuracy=0.875\n",
      "Epoch  4,  step=1231,  loss=0.662,  accuracy=0.844\n",
      "Epoch  4,  step=1232,  loss=0.265,  accuracy=0.969\n",
      "Epoch  4,  step=1233,  loss=0.436,  accuracy=0.938\n",
      "Epoch  4,  step=1234,  loss=0.197,  accuracy=0.938\n",
      "Epoch  4,  step=1235,  loss=0.338,  accuracy=0.906\n",
      "Epoch  4,  step=1236,  loss=0.272,  accuracy=0.938\n",
      "Epoch  4,  step=1237,  loss=0.380,  accuracy=0.875\n",
      "Epoch  4,  step=1238,  loss=0.270,  accuracy=0.906\n",
      "Epoch  4,  step=1239,  loss=0.662,  accuracy=0.906\n",
      "Epoch  4,  step=1240,  loss=0.570,  accuracy=0.844\n",
      "Epoch  4,  step=1241,  loss=0.219,  accuracy=0.906\n",
      "Epoch  4,  step=1242,  loss=0.458,  accuracy=0.812\n",
      "Epoch  4,  step=1243,  loss=0.463,  accuracy=0.875\n",
      "Epoch  4,  step=1244,  loss=0.236,  accuracy=0.938\n",
      "Epoch  4,  step=1245,  loss=0.336,  accuracy=0.906\n",
      "Epoch  4,  step=1246,  loss=0.291,  accuracy=0.906\n",
      "Epoch  4,  step=1247,  loss=0.469,  accuracy=0.875\n",
      "Epoch  4,  step=1248,  loss=0.364,  accuracy=0.906\n",
      "Epoch  4,  step=1249,  loss=0.353,  accuracy=0.875\n",
      "Epoch  4,  step=1250,  loss=0.275,  accuracy=0.906\n",
      "Epoch  4,  step=1251,  loss=0.209,  accuracy=0.906\n",
      "Epoch  4,  step=1252,  loss=0.256,  accuracy=0.844\n",
      "Epoch  4,  step=1253,  loss=0.275,  accuracy=0.906\n",
      "Epoch  4,  step=1254,  loss=0.286,  accuracy=0.875\n",
      "Epoch  4,  step=1255,  loss=0.298,  accuracy=0.906\n",
      "Epoch  4,  step=1256,  loss=0.204,  accuracy=0.938\n",
      "Epoch  4,  step=1257,  loss=0.431,  accuracy=0.906\n",
      "Epoch  4,  step=1258,  loss=0.103,  accuracy=1.000\n",
      "Epoch  4,  step=1259,  loss=0.415,  accuracy=0.875\n",
      "Epoch  4,  step=1260,  loss=0.187,  accuracy=0.969\n",
      "Epoch  4,  step=1261,  loss=0.047,  accuracy=1.000\n",
      "Epoch  4,  step=1262,  loss=0.173,  accuracy=0.938\n",
      "Epoch  4,  step=1263,  loss=0.193,  accuracy=0.906\n",
      "Epoch  4,  step=1264,  loss=0.116,  accuracy=0.938\n",
      "Epoch  4,  step=1265,  loss=0.465,  accuracy=0.875\n",
      "Epoch  4,  step=1266,  loss=0.307,  accuracy=0.938\n",
      "Epoch  4,  step=1267,  loss=0.161,  accuracy=0.938\n",
      "Epoch  4,  step=1268,  loss=0.214,  accuracy=0.906\n",
      "Epoch  4,  step=1269,  loss=0.251,  accuracy=0.938\n",
      "Epoch  4,  step=1270,  loss=0.526,  accuracy=0.812\n",
      "Epoch  4,  step=1271,  loss=0.065,  accuracy=1.000\n",
      "Epoch  4,  step=1272,  loss=0.240,  accuracy=0.938\n",
      "Epoch  4,  step=1273,  loss=0.252,  accuracy=0.906\n",
      "Epoch  4,  step=1274,  loss=0.096,  accuracy=0.969\n",
      "Epoch  4,  step=1275,  loss=0.104,  accuracy=0.969\n",
      "Epoch  4,  step=1276,  loss=0.134,  accuracy=0.969\n",
      "Epoch  4,  step=1277,  loss=0.162,  accuracy=0.969\n",
      "Epoch  4,  step=1278,  loss=0.052,  accuracy=1.000\n",
      "Epoch  4,  step=1279,  loss=0.096,  accuracy=0.969\n",
      "Epoch  4,  step=1280,  loss=0.450,  accuracy=0.938\n",
      "Epoch  4,  step=1281,  loss=0.501,  accuracy=0.844\n",
      "Epoch  4,  step=1282,  loss=0.143,  accuracy=0.969\n",
      "Epoch  4,  step=1283,  loss=0.182,  accuracy=0.875\n",
      "Epoch  4,  step=1284,  loss=0.232,  accuracy=0.875\n",
      "Epoch  4,  step=1285,  loss=0.143,  accuracy=0.969\n",
      "Epoch  4,  step=1286,  loss=0.277,  accuracy=0.875\n",
      "Epoch  4,  step=1287,  loss=0.233,  accuracy=0.906\n",
      "Epoch  4,  step=1288,  loss=0.176,  accuracy=0.938\n",
      "Epoch  4,  step=1289,  loss=0.612,  accuracy=0.875\n",
      "Epoch  4,  step=1290,  loss=0.706,  accuracy=0.844\n",
      "Epoch  4,  step=1291,  loss=0.366,  accuracy=0.938\n",
      "Epoch  4,  step=1292,  loss=0.203,  accuracy=0.969\n",
      "Epoch  4,  step=1293,  loss=0.286,  accuracy=0.938\n",
      "Epoch  4,  step=1294,  loss=0.294,  accuracy=0.906\n",
      "Epoch  4,  step=1295,  loss=0.418,  accuracy=0.875\n",
      "Epoch  4,  step=1296,  loss=0.202,  accuracy=0.906\n",
      "Epoch  4,  step=1297,  loss=0.254,  accuracy=0.906\n",
      "Epoch  4,  step=1298,  loss=0.375,  accuracy=0.906\n",
      "Epoch  4,  step=1299,  loss=0.433,  accuracy=0.844\n",
      "Epoch  4,  step=1300,  loss=0.226,  accuracy=0.938\n",
      "Epoch  4,  step=1301,  loss=0.029,  accuracy=1.000\n",
      "Epoch  4,  step=1302,  loss=0.123,  accuracy=0.969\n",
      "Epoch  4,  step=1303,  loss=0.209,  accuracy=0.906\n",
      "Epoch  4,  step=1304,  loss=0.434,  accuracy=0.875\n",
      "Epoch  4,  step=1305,  loss=0.249,  accuracy=0.906\n",
      "Epoch  4,  step=1306,  loss=0.323,  accuracy=0.844\n",
      "Epoch  4,  step=1307,  loss=0.227,  accuracy=0.938\n",
      "Epoch  4,  step=1308,  loss=0.171,  accuracy=0.969\n",
      "Epoch  4,  step=1309,  loss=0.387,  accuracy=0.906\n",
      "Epoch  4,  step=1310,  loss=0.392,  accuracy=0.906\n",
      "Epoch  4,  step=1311,  loss=0.383,  accuracy=0.906\n",
      "Epoch  4,  step=1312,  loss=0.469,  accuracy=0.906\n",
      "Epoch  4,  step=1313,  loss=0.329,  accuracy=0.844\n",
      "Epoch  4,  step=1314,  loss=0.273,  accuracy=0.938\n",
      "Epoch  4,  step=1315,  loss=0.315,  accuracy=0.938\n",
      "Epoch  4,  step=1316,  loss=0.270,  accuracy=0.844\n",
      "Epoch  4,  step=1317,  loss=0.120,  accuracy=0.938\n",
      "Epoch  4,  step=1318,  loss=0.303,  accuracy=0.906\n",
      "Epoch  4,  step=1319,  loss=0.259,  accuracy=0.906\n",
      "Epoch  4,  step=1320,  loss=0.205,  accuracy=0.906\n",
      "Epoch  4,  step=1321,  loss=0.090,  accuracy=0.969\n",
      "Epoch  4,  step=1322,  loss=0.668,  accuracy=0.812\n",
      "Epoch  4,  step=1323,  loss=0.692,  accuracy=0.844\n",
      "Epoch  4,  step=1324,  loss=0.127,  accuracy=0.938\n",
      "Epoch  4,  step=1325,  loss=0.401,  accuracy=0.906\n",
      "Epoch  4,  step=1326,  loss=0.336,  accuracy=0.906\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  4,  step=1327,  loss=0.241,  accuracy=0.906\n",
      "Epoch  4,  step=1328,  loss=0.614,  accuracy=0.844\n",
      "Epoch  4,  step=1329,  loss=0.171,  accuracy=0.906\n",
      "Epoch  4,  step=1330,  loss=0.187,  accuracy=0.969\n",
      "Epoch  4,  step=1331,  loss=0.134,  accuracy=0.938\n",
      "Epoch  4,  step=1332,  loss=0.259,  accuracy=0.906\n",
      "Epoch  4,  step=1333,  loss=0.385,  accuracy=0.844\n",
      "Epoch  4,  step=1334,  loss=0.536,  accuracy=0.812\n",
      "Epoch  4,  step=1335,  loss=0.113,  accuracy=0.969\n",
      "Epoch  4,  step=1336,  loss=0.405,  accuracy=0.906\n",
      "Epoch  4,  step=1337,  loss=0.419,  accuracy=0.875\n",
      "Epoch  4,  step=1338,  loss=0.702,  accuracy=0.906\n",
      "Epoch  4,  step=1339,  loss=0.638,  accuracy=0.875\n",
      "Epoch  4,  step=1340,  loss=0.375,  accuracy=0.875\n",
      "Epoch  4,  step=1341,  loss=0.346,  accuracy=0.938\n",
      "Epoch  4,  step=1342,  loss=0.334,  accuracy=0.906\n",
      "Epoch  4,  step=1343,  loss=0.183,  accuracy=0.938\n",
      "Epoch  4,  step=1344,  loss=0.511,  accuracy=0.812\n",
      "Epoch  4,  step=1345,  loss=0.275,  accuracy=0.875\n",
      "Epoch  4,  step=1346,  loss=0.512,  accuracy=0.906\n",
      "Epoch  4,  step=1347,  loss=0.119,  accuracy=0.969\n",
      "Epoch  4,  step=1348,  loss=0.094,  accuracy=0.938\n",
      "Epoch  4,  step=1349,  loss=0.216,  accuracy=0.906\n",
      "Epoch  4,  step=1350,  loss=0.663,  accuracy=0.750\n",
      "Epoch  4,  step=1351,  loss=0.149,  accuracy=0.938\n",
      "Epoch  4,  step=1352,  loss=0.209,  accuracy=0.938\n",
      "Epoch  4,  step=1353,  loss=0.037,  accuracy=0.969\n",
      "Epoch  4,  step=1354,  loss=0.066,  accuracy=1.000\n",
      "Epoch  4,  step=1355,  loss=0.192,  accuracy=0.938\n",
      "Epoch  4,  step=1356,  loss=0.089,  accuracy=1.000\n",
      "Epoch  4,  step=1357,  loss=0.397,  accuracy=0.938\n",
      "Epoch  4,  step=1358,  loss=0.236,  accuracy=0.906\n",
      "Epoch  4,  step=1359,  loss=0.190,  accuracy=0.938\n",
      "Epoch  4,  step=1360,  loss=0.235,  accuracy=0.906\n",
      "Epoch  4,  step=1361,  loss=0.184,  accuracy=0.906\n",
      "Epoch  4,  step=1362,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step=1363,  loss=0.208,  accuracy=0.938\n",
      "Epoch  4,  step=1364,  loss=0.473,  accuracy=0.875\n",
      "Epoch  4,  step=1365,  loss=0.204,  accuracy=0.906\n",
      "Epoch  4,  step=1366,  loss=0.024,  accuracy=1.000\n",
      "Epoch  4,  step=1367,  loss=0.131,  accuracy=0.969\n",
      "Epoch  4,  step=1368,  loss=0.306,  accuracy=0.938\n",
      "Epoch  4,  step=1369,  loss=0.388,  accuracy=0.875\n",
      "Epoch  4,  step=1370,  loss=0.199,  accuracy=0.938\n",
      "Epoch  4,  step=1371,  loss=0.340,  accuracy=0.875\n",
      "Epoch  4,  step=1372,  loss=0.129,  accuracy=0.938\n",
      "Epoch  4,  step=1373,  loss=0.518,  accuracy=0.844\n",
      "Epoch  4,  step=1374,  loss=0.330,  accuracy=0.906\n",
      "Epoch  4,  step=1375,  loss=0.069,  accuracy=1.000\n",
      "Epoch  4,  step=1376,  loss=0.208,  accuracy=0.938\n",
      "Epoch  4,  step=1377,  loss=0.232,  accuracy=0.938\n",
      "Epoch  4,  step=1378,  loss=0.337,  accuracy=0.906\n",
      "Epoch  4,  step=1379,  loss=0.426,  accuracy=0.875\n",
      "Epoch  4,  step=1380,  loss=0.178,  accuracy=0.969\n",
      "Epoch  4,  step=1381,  loss=0.163,  accuracy=0.938\n",
      "Epoch  4,  step=1382,  loss=0.352,  accuracy=0.844\n",
      "Epoch  4,  step=1383,  loss=0.514,  accuracy=0.938\n",
      "Epoch  4,  step=1384,  loss=0.299,  accuracy=0.906\n",
      "Epoch  4,  step=1385,  loss=0.226,  accuracy=0.906\n",
      "Epoch  4,  step=1386,  loss=0.524,  accuracy=0.938\n",
      "Epoch  4,  step=1387,  loss=0.147,  accuracy=0.969\n",
      "Epoch  4,  step=1388,  loss=0.365,  accuracy=0.938\n",
      "Epoch  4,  step=1389,  loss=0.172,  accuracy=0.938\n",
      "Epoch  4,  step=1390,  loss=0.363,  accuracy=0.875\n",
      "Epoch  4,  step=1391,  loss=0.103,  accuracy=0.969\n",
      "Epoch  4,  step=1392,  loss=0.057,  accuracy=1.000\n",
      "Epoch  4,  step=1393,  loss=0.071,  accuracy=0.969\n",
      "Epoch  4,  step=1394,  loss=0.129,  accuracy=1.000\n",
      "Epoch  4,  step=1395,  loss=0.229,  accuracy=0.969\n",
      "Epoch  4,  step=1396,  loss=0.275,  accuracy=0.938\n",
      "Epoch  4,  step=1397,  loss=0.160,  accuracy=0.938\n",
      "Epoch  4,  step=1398,  loss=0.559,  accuracy=0.844\n",
      "Epoch  4,  step=1399,  loss=0.083,  accuracy=0.969\n",
      "Epoch  4,  step=1400,  loss=0.218,  accuracy=0.938\n",
      "Epoch  4,  step=1401,  loss=0.483,  accuracy=0.844\n",
      "Epoch  4,  step=1402,  loss=0.643,  accuracy=0.875\n",
      "Epoch  4,  step=1403,  loss=0.373,  accuracy=0.875\n",
      "Epoch  4,  step=1404,  loss=0.206,  accuracy=0.906\n",
      "Epoch  4,  step=1405,  loss=0.186,  accuracy=0.938\n",
      "Epoch  4,  step=1406,  loss=0.385,  accuracy=0.938\n",
      "Epoch  4,  step=1407,  loss=0.797,  accuracy=0.844\n",
      "Epoch  4,  step=1408,  loss=0.196,  accuracy=0.938\n",
      "Epoch  4,  step=1409,  loss=0.577,  accuracy=0.812\n",
      "Epoch  4,  step=1410,  loss=0.231,  accuracy=0.969\n",
      "Epoch  4,  step=1411,  loss=0.329,  accuracy=0.906\n",
      "Epoch  4,  step=1412,  loss=0.282,  accuracy=0.969\n",
      "Epoch  4,  step=1413,  loss=0.221,  accuracy=0.906\n",
      "Epoch  4,  step=1414,  loss=0.170,  accuracy=0.906\n",
      "Epoch  4,  step=1415,  loss=0.080,  accuracy=0.969\n",
      "Epoch  4,  step=1416,  loss=0.166,  accuracy=0.906\n",
      "Epoch  4,  step=1417,  loss=0.167,  accuracy=0.938\n",
      "Epoch  4,  step=1418,  loss=0.102,  accuracy=0.969\n",
      "Epoch  4,  step=1419,  loss=0.560,  accuracy=0.875\n",
      "Epoch  4,  step=1420,  loss=0.553,  accuracy=0.812\n",
      "Epoch  4,  step=1421,  loss=0.782,  accuracy=0.906\n",
      "Epoch  4,  step=1422,  loss=0.428,  accuracy=0.875\n",
      "Epoch  4,  step=1423,  loss=0.161,  accuracy=0.938\n",
      "Epoch  4,  step=1424,  loss=0.235,  accuracy=0.938\n",
      "Epoch  4,  step=1425,  loss=0.420,  accuracy=0.875\n",
      "Epoch  4,  step=1426,  loss=0.386,  accuracy=0.844\n",
      "Epoch  4,  step=1427,  loss=0.314,  accuracy=0.875\n",
      "Epoch  4,  step=1428,  loss=0.024,  accuracy=1.000\n",
      "Epoch  4,  step=1429,  loss=0.259,  accuracy=0.875\n",
      "Epoch  4,  step=1430,  loss=0.391,  accuracy=0.906\n",
      "Epoch  4,  step=1431,  loss=0.301,  accuracy=0.938\n",
      "Epoch  4,  step=1432,  loss=0.427,  accuracy=0.812\n",
      "Epoch  4,  step=1433,  loss=0.623,  accuracy=0.875\n",
      "Epoch  4,  step=1434,  loss=0.463,  accuracy=0.906\n",
      "Epoch  4,  step=1435,  loss=0.396,  accuracy=0.875\n",
      "Epoch  4,  step=1436,  loss=0.494,  accuracy=0.812\n",
      "Epoch  4,  step=1437,  loss=0.414,  accuracy=0.906\n",
      "Epoch  4,  step=1438,  loss=0.299,  accuracy=0.875\n",
      "Epoch  4,  step=1439,  loss=0.362,  accuracy=0.938\n",
      "Epoch  4,  step=1440,  loss=0.506,  accuracy=0.875\n",
      "Epoch  4,  step=1441,  loss=0.180,  accuracy=0.906\n",
      "Epoch  4,  step=1442,  loss=0.148,  accuracy=0.938\n",
      "Epoch  4,  step=1443,  loss=0.130,  accuracy=0.969\n",
      "Epoch  4,  step=1444,  loss=0.121,  accuracy=0.969\n",
      "Epoch  4,  step=1445,  loss=1.034,  accuracy=0.781\n",
      "Epoch  4,  step=1446,  loss=0.582,  accuracy=0.781\n",
      "Epoch  4,  step=1447,  loss=0.467,  accuracy=0.812\n",
      "Epoch  4,  step=1448,  loss=0.308,  accuracy=0.875\n",
      "Epoch  4,  step=1449,  loss=0.244,  accuracy=0.906\n",
      "Epoch  4,  step=1450,  loss=0.399,  accuracy=0.875\n",
      "Epoch  4,  step=1451,  loss=0.135,  accuracy=0.938\n",
      "Epoch  4,  step=1452,  loss=0.075,  accuracy=1.000\n",
      "Epoch  4,  step=1453,  loss=0.126,  accuracy=0.969\n",
      "Epoch  4,  step=1454,  loss=0.143,  accuracy=0.938\n",
      "Epoch  4,  step=1455,  loss=0.117,  accuracy=0.938\n",
      "Epoch  4,  step=1456,  loss=0.244,  accuracy=0.906\n",
      "Epoch  4,  step=1457,  loss=0.299,  accuracy=0.938\n",
      "Epoch  4,  step=1458,  loss=0.072,  accuracy=1.000\n",
      "Epoch  4,  step=1459,  loss=0.217,  accuracy=0.938\n",
      "Epoch  4,  step=1460,  loss=0.776,  accuracy=0.812\n",
      "Epoch  4,  step=1461,  loss=0.283,  accuracy=0.938\n",
      "Epoch  4,  step=1462,  loss=0.202,  accuracy=0.969\n",
      "Epoch  4,  step=1463,  loss=0.101,  accuracy=0.969\n",
      "Epoch  4,  step=1464,  loss=0.160,  accuracy=0.906\n",
      "Epoch  4,  step=1465,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step=1466,  loss=0.206,  accuracy=0.938\n",
      "Epoch  4,  step=1467,  loss=0.116,  accuracy=0.969\n",
      "Epoch  4,  step=1468,  loss=0.195,  accuracy=0.969\n",
      "Epoch  4,  step=1469,  loss=0.591,  accuracy=0.844\n",
      "Epoch  4,  step=1470,  loss=0.405,  accuracy=0.906\n",
      "Epoch  4,  step=1471,  loss=0.228,  accuracy=0.938\n",
      "Epoch  4,  step=1472,  loss=0.080,  accuracy=1.000\n",
      "Epoch  4,  step=1473,  loss=0.181,  accuracy=0.938\n",
      "Epoch  4,  step=1474,  loss=0.258,  accuracy=0.906\n",
      "Epoch  4,  step=1475,  loss=0.646,  accuracy=0.812\n",
      "Epoch  4,  step=1476,  loss=0.482,  accuracy=0.812\n",
      "Epoch  4,  step=1477,  loss=0.291,  accuracy=0.938\n",
      "Epoch  4,  step=1478,  loss=0.511,  accuracy=0.844\n",
      "Epoch  4,  step=1479,  loss=0.467,  accuracy=0.969\n",
      "Epoch  4,  step=1480,  loss=0.461,  accuracy=0.906\n",
      "Epoch  4,  step=1481,  loss=0.389,  accuracy=0.875\n",
      "Epoch  4,  step=1482,  loss=0.382,  accuracy=0.844\n",
      "Epoch  4,  step=1483,  loss=0.309,  accuracy=0.875\n",
      "Epoch  4,  step=1484,  loss=0.086,  accuracy=1.000\n",
      "Epoch  4,  step=1485,  loss=0.200,  accuracy=0.875\n",
      "Epoch  4,  step=1486,  loss=0.408,  accuracy=0.906\n",
      "Epoch  4,  step=1487,  loss=0.598,  accuracy=0.812\n",
      "Epoch  4,  step=1488,  loss=0.177,  accuracy=0.906\n",
      "Epoch  4,  step=1489,  loss=0.164,  accuracy=0.938\n",
      "Epoch  4,  step=1490,  loss=0.405,  accuracy=0.938\n",
      "Epoch  4,  step=1491,  loss=0.223,  accuracy=0.938\n",
      "Epoch  4,  step=1492,  loss=0.383,  accuracy=0.844\n",
      "Epoch  4,  step=1493,  loss=0.237,  accuracy=0.938\n",
      "Epoch  4,  step=1494,  loss=0.231,  accuracy=0.938\n",
      "Epoch  4,  step=1495,  loss=0.084,  accuracy=0.969\n",
      "Epoch  4,  step=1496,  loss=0.220,  accuracy=0.938\n",
      "Epoch  4,  step=1497,  loss=0.531,  accuracy=0.875\n",
      "Epoch  4,  step=1498,  loss=0.351,  accuracy=0.875\n",
      "Epoch  4,  step=1499,  loss=0.179,  accuracy=0.938\n",
      "Epoch  4,  step=1500,  loss=0.220,  accuracy=0.906\n",
      "Epoch  4,  step=1501,  loss=0.156,  accuracy=0.969\n",
      "Epoch  4,  step=1502,  loss=0.309,  accuracy=0.875\n",
      "Epoch  4,  step=1503,  loss=0.284,  accuracy=0.938\n",
      "Epoch  4,  step=1504,  loss=0.129,  accuracy=0.969\n",
      "Epoch  4,  step=1505,  loss=0.120,  accuracy=0.906\n",
      "Epoch  4,  step=1506,  loss=0.059,  accuracy=1.000\n",
      "Epoch  4,  step=1507,  loss=0.060,  accuracy=1.000\n",
      "Epoch  4,  step=1508,  loss=0.202,  accuracy=0.938\n",
      "Epoch  4,  step=1509,  loss=0.080,  accuracy=0.969\n",
      "Epoch  4,  step=1510,  loss=0.272,  accuracy=0.938\n",
      "Epoch  4,  step=1511,  loss=0.459,  accuracy=0.844\n",
      "Epoch  4,  step=1512,  loss=0.630,  accuracy=0.781\n",
      "Epoch  4,  step=1513,  loss=0.032,  accuracy=1.000\n",
      "Epoch  4,  step=1514,  loss=0.198,  accuracy=0.969\n",
      "Epoch  4,  step=1515,  loss=0.244,  accuracy=0.969\n",
      "Epoch  4,  step=1516,  loss=0.216,  accuracy=0.906\n",
      "Epoch  4,  step=1517,  loss=0.182,  accuracy=0.938\n",
      "Epoch  4,  step=1518,  loss=0.193,  accuracy=0.938\n",
      "Epoch  4,  step=1519,  loss=0.038,  accuracy=1.000\n",
      "Epoch  4,  step=1520,  loss=0.447,  accuracy=0.906\n",
      "Epoch  4,  step=1521,  loss=0.231,  accuracy=0.875\n",
      "Epoch  4,  step=1522,  loss=0.170,  accuracy=0.969\n",
      "Epoch  4,  step=1523,  loss=0.150,  accuracy=0.906\n",
      "Epoch  4,  step=1524,  loss=0.150,  accuracy=0.938\n",
      "Epoch  4,  step=1525,  loss=0.131,  accuracy=0.938\n",
      "Epoch  4,  step=1526,  loss=0.366,  accuracy=0.844\n",
      "Epoch  4,  step=1527,  loss=0.316,  accuracy=0.875\n",
      "Epoch  4,  step=1528,  loss=0.684,  accuracy=0.875\n",
      "Epoch  4,  step=1529,  loss=0.692,  accuracy=0.812\n",
      "Epoch  4,  step=1530,  loss=0.565,  accuracy=0.875\n",
      "Epoch  4,  step=1531,  loss=1.252,  accuracy=0.781\n",
      "Epoch  4,  step=1532,  loss=0.200,  accuracy=0.906\n",
      "Epoch  4,  step=1533,  loss=0.582,  accuracy=0.875\n",
      "Epoch  4,  step=1534,  loss=0.253,  accuracy=0.906\n",
      "Epoch  4,  step=1535,  loss=0.559,  accuracy=0.844\n",
      "Epoch  4,  step=1536,  loss=0.371,  accuracy=0.875\n",
      "Epoch  4,  step=1537,  loss=0.822,  accuracy=0.750\n",
      "Epoch  4,  step=1538,  loss=0.177,  accuracy=0.938\n",
      "Epoch  4,  step=1539,  loss=0.165,  accuracy=0.938\n",
      "Epoch  4,  step=1540,  loss=0.181,  accuracy=0.938\n",
      "Epoch  4,  step=1541,  loss=0.192,  accuracy=0.938\n",
      "Epoch  4,  step=1542,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step=1543,  loss=0.047,  accuracy=1.000\n",
      "Epoch  4,  step=1544,  loss=0.755,  accuracy=0.750\n",
      "Epoch  4,  step=1545,  loss=0.177,  accuracy=0.969\n",
      "Epoch  4,  step=1546,  loss=0.794,  accuracy=0.812\n",
      "Epoch  4,  step=1547,  loss=0.554,  accuracy=0.781\n",
      "Epoch  4,  step=1548,  loss=0.763,  accuracy=0.812\n",
      "Epoch  4,  step=1549,  loss=0.791,  accuracy=0.844\n",
      "Epoch  4,  step=1550,  loss=0.413,  accuracy=0.906\n",
      "Epoch  4,  step=1551,  loss=0.601,  accuracy=0.844\n",
      "Epoch  4,  step=1552,  loss=0.376,  accuracy=0.875\n",
      "Epoch  4,  step=1553,  loss=0.113,  accuracy=0.969\n",
      "Epoch  4,  step=1554,  loss=0.276,  accuracy=0.938\n",
      "Epoch  4,  step=1555,  loss=0.134,  accuracy=0.969\n",
      "Epoch  4,  step=1556,  loss=0.276,  accuracy=0.938\n",
      "Epoch  4,  step=1557,  loss=0.251,  accuracy=0.906\n",
      "Epoch  4,  step=1558,  loss=0.051,  accuracy=0.969\n",
      "Epoch  4,  step=1559,  loss=0.754,  accuracy=0.812\n",
      "Epoch  4,  step=1560,  loss=0.162,  accuracy=0.938\n",
      "Epoch  4,  step=1561,  loss=0.139,  accuracy=0.969\n",
      "Epoch  4,  step=1562,  loss=0.329,  accuracy=0.844\n",
      "Epoch  4,  step=1563,  loss=0.242,  accuracy=0.906\n",
      "Epoch  4,  step=1564,  loss=0.324,  accuracy=0.906\n",
      "Epoch  4,  step=1565,  loss=0.114,  accuracy=0.938\n",
      "Epoch  4,  step=1566,  loss=0.164,  accuracy=0.969\n",
      "Epoch  4,  step=1567,  loss=0.425,  accuracy=0.906\n",
      "Epoch  4,  step=1568,  loss=0.175,  accuracy=0.969\n",
      "Epoch  4,  step=1569,  loss=0.688,  accuracy=0.844\n",
      "Epoch  4,  step=1570,  loss=0.510,  accuracy=0.875\n",
      "Epoch  4,  step=1571,  loss=0.078,  accuracy=0.969\n",
      "Epoch  4,  step=1572,  loss=0.402,  accuracy=0.844\n",
      "Epoch  4,  step=1573,  loss=0.399,  accuracy=0.844\n",
      "Epoch  4,  step=1574,  loss=0.530,  accuracy=0.844\n",
      "Epoch  4,  step=1575,  loss=0.915,  accuracy=0.750\n",
      "Epoch  4,  step=1576,  loss=0.319,  accuracy=0.906\n",
      "Epoch  4,  step=1577,  loss=0.098,  accuracy=0.938\n",
      "Epoch  4,  step=1578,  loss=0.462,  accuracy=0.906\n",
      "Epoch  4,  step=1579,  loss=0.093,  accuracy=1.000\n",
      "Epoch  4,  step=1580,  loss=0.811,  accuracy=0.844\n",
      "Epoch  4,  step=1581,  loss=0.171,  accuracy=0.938\n",
      "Epoch  4,  step=1582,  loss=0.752,  accuracy=0.844\n",
      "Epoch  4,  step=1583,  loss=0.091,  accuracy=0.969\n",
      "Epoch  4,  step=1584,  loss=0.426,  accuracy=0.844\n",
      "Epoch  4,  step=1585,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step=1586,  loss=0.258,  accuracy=0.906\n",
      "Epoch  4,  step=1587,  loss=0.279,  accuracy=0.938\n",
      "Epoch  4,  step=1588,  loss=0.260,  accuracy=0.875\n",
      "Epoch  4,  step=1589,  loss=0.101,  accuracy=0.969\n",
      "Epoch  4,  step=1590,  loss=0.249,  accuracy=0.938\n",
      "Epoch  4,  step=1591,  loss=0.068,  accuracy=1.000\n",
      "Epoch  4,  step=1592,  loss=0.193,  accuracy=0.906\n",
      "Epoch  4,  step=1593,  loss=0.184,  accuracy=0.938\n",
      "Epoch  4,  step=1594,  loss=0.096,  accuracy=0.969\n",
      "Epoch  4,  step=1595,  loss=0.421,  accuracy=0.906\n",
      "Epoch  4,  step=1596,  loss=0.218,  accuracy=0.969\n",
      "Epoch  4,  step=1597,  loss=0.269,  accuracy=0.875\n",
      "Epoch  4,  step=1598,  loss=0.274,  accuracy=0.875\n",
      "Epoch  4,  step=1599,  loss=0.410,  accuracy=0.875\n",
      "Epoch  4,  step=1600,  loss=0.154,  accuracy=0.938\n",
      "Epoch  4,  step=1601,  loss=0.302,  accuracy=0.844\n",
      "Epoch  4,  step=1602,  loss=0.454,  accuracy=0.875\n",
      "Epoch  4,  step=1603,  loss=0.384,  accuracy=0.906\n",
      "Epoch  4,  step=1604,  loss=0.252,  accuracy=0.938\n",
      "Epoch  4,  step=1605,  loss=0.117,  accuracy=0.938\n",
      "Epoch  4,  step=1606,  loss=0.160,  accuracy=0.938\n",
      "Epoch  4,  step=1607,  loss=0.152,  accuracy=0.906\n",
      "Epoch  4,  step=1608,  loss=0.117,  accuracy=0.969\n",
      "Epoch  4,  step=1609,  loss=0.096,  accuracy=0.969\n",
      "Epoch  4,  step=1610,  loss=0.405,  accuracy=0.938\n",
      "Epoch  4,  step=1611,  loss=0.206,  accuracy=0.875\n",
      "Epoch  4,  step=1612,  loss=0.050,  accuracy=1.000\n",
      "Epoch  4,  step=1613,  loss=0.301,  accuracy=0.938\n",
      "Epoch  4,  step=1614,  loss=0.107,  accuracy=0.969\n",
      "Epoch  4,  step=1615,  loss=0.286,  accuracy=0.906\n",
      "Epoch  4,  step=1616,  loss=0.488,  accuracy=0.938\n",
      "Epoch  4,  step=1617,  loss=0.292,  accuracy=0.906\n",
      "Epoch  4,  step=1618,  loss=0.205,  accuracy=0.969\n",
      "Epoch  4,  step=1619,  loss=0.054,  accuracy=1.000\n",
      "Epoch  4,  step=1620,  loss=0.241,  accuracy=0.906\n",
      "Epoch  4,  step=1621,  loss=0.237,  accuracy=0.906\n",
      "Epoch  4,  step=1622,  loss=0.185,  accuracy=0.938\n",
      "Epoch  4,  step=1623,  loss=0.434,  accuracy=0.906\n",
      "Epoch  4,  step=1624,  loss=0.593,  accuracy=0.906\n",
      "Epoch  4,  step=1625,  loss=0.086,  accuracy=0.969\n",
      "Epoch  4,  step=1626,  loss=0.158,  accuracy=0.938\n",
      "Epoch  4,  step=1627,  loss=0.392,  accuracy=0.906\n",
      "Epoch  4,  step=1628,  loss=0.642,  accuracy=0.781\n",
      "Epoch  4,  step=1629,  loss=0.906,  accuracy=0.812\n",
      "Epoch  4,  step=1630,  loss=0.434,  accuracy=0.844\n",
      "Epoch  4,  step=1631,  loss=0.188,  accuracy=0.938\n",
      "Epoch  4,  step=1632,  loss=0.378,  accuracy=0.906\n",
      "Epoch  4,  step=1633,  loss=0.265,  accuracy=0.875\n",
      "Epoch  4,  step=1634,  loss=0.071,  accuracy=1.000\n",
      "Epoch  4,  step=1635,  loss=0.283,  accuracy=0.969\n",
      "Epoch  4,  step=1636,  loss=0.434,  accuracy=0.938\n",
      "Epoch  4,  step=1637,  loss=0.309,  accuracy=0.938\n",
      "Epoch  4,  step=1638,  loss=0.225,  accuracy=0.938\n",
      "Epoch  4,  step=1639,  loss=0.081,  accuracy=0.969\n",
      "Epoch  4,  step=1640,  loss=0.150,  accuracy=0.938\n",
      "Epoch  4,  step=1641,  loss=0.071,  accuracy=1.000\n",
      "Epoch  4,  step=1642,  loss=0.065,  accuracy=1.000\n",
      "Epoch  4,  step=1643,  loss=0.118,  accuracy=0.969\n",
      "Epoch  4,  step=1644,  loss=0.074,  accuracy=0.969\n",
      "Epoch  4,  step=1645,  loss=0.554,  accuracy=0.875\n",
      "Epoch  4,  step=1646,  loss=0.406,  accuracy=0.875\n",
      "Epoch  4,  step=1647,  loss=0.478,  accuracy=0.875\n",
      "Epoch  4,  step=1648,  loss=0.260,  accuracy=0.906\n",
      "Epoch  4,  step=1649,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step=1650,  loss=0.387,  accuracy=0.844\n",
      "Epoch  4,  step=1651,  loss=0.120,  accuracy=0.969\n",
      "Epoch  4,  step=1652,  loss=0.261,  accuracy=0.906\n",
      "Epoch  4,  step=1653,  loss=0.510,  accuracy=0.875\n",
      "Epoch  4,  step=1654,  loss=0.722,  accuracy=0.844\n",
      "Epoch  4,  step=1655,  loss=0.260,  accuracy=0.906\n",
      "Epoch  4,  step=1656,  loss=0.324,  accuracy=0.906\n",
      "Epoch  4,  step=1657,  loss=0.341,  accuracy=0.844\n",
      "Epoch  4,  step=1658,  loss=0.267,  accuracy=0.906\n",
      "Epoch  4,  step=1659,  loss=0.360,  accuracy=0.844\n",
      "Epoch  4,  step=1660,  loss=0.248,  accuracy=0.906\n",
      "Epoch  4,  step=1661,  loss=0.690,  accuracy=0.844\n",
      "Epoch  4,  step=1662,  loss=0.140,  accuracy=0.906\n",
      "Epoch  4,  step=1663,  loss=0.385,  accuracy=0.938\n",
      "Epoch  4,  step=1664,  loss=0.085,  accuracy=0.938\n",
      "Epoch  4,  step=1665,  loss=0.202,  accuracy=0.906\n",
      "Epoch  4,  step=1666,  loss=0.089,  accuracy=0.969\n",
      "Epoch  4,  step=1667,  loss=0.061,  accuracy=1.000\n",
      "Epoch  4,  step=1668,  loss=0.163,  accuracy=0.906\n",
      "Epoch  4,  step=1669,  loss=0.053,  accuracy=1.000\n",
      "Epoch  4,  step=1670,  loss=0.335,  accuracy=0.906\n",
      "Epoch  4,  step=1671,  loss=0.353,  accuracy=0.844\n",
      "Epoch  4,  step=1672,  loss=0.331,  accuracy=0.875\n",
      "Epoch  4,  step=1673,  loss=0.315,  accuracy=0.875\n",
      "Epoch  4,  step=1674,  loss=0.390,  accuracy=0.906\n",
      "Epoch  4,  step=1675,  loss=0.195,  accuracy=0.906\n",
      "Epoch  4,  step=1676,  loss=0.250,  accuracy=0.938\n",
      "Epoch  4,  step=1677,  loss=0.453,  accuracy=0.844\n",
      "Epoch  4,  step=1678,  loss=0.068,  accuracy=0.969\n",
      "Epoch  4,  step=1679,  loss=0.469,  accuracy=0.781\n",
      "Epoch  4,  step=1680,  loss=0.026,  accuracy=1.000\n",
      "Epoch  4,  step=1681,  loss=0.329,  accuracy=0.906\n",
      "Epoch  4,  step=1682,  loss=0.317,  accuracy=0.875\n",
      "Epoch  4,  step=1683,  loss=0.297,  accuracy=0.938\n",
      "Epoch  4,  step=1684,  loss=0.651,  accuracy=0.844\n",
      "Epoch  4,  step=1685,  loss=0.109,  accuracy=0.969\n",
      "Epoch  4,  step=1686,  loss=0.746,  accuracy=0.750\n",
      "Epoch  4,  step=1687,  loss=0.376,  accuracy=0.875\n",
      "Epoch  4,  step=1688,  loss=0.253,  accuracy=0.906\n",
      "Epoch  4,  step=1689,  loss=0.207,  accuracy=0.938\n",
      "Epoch  4,  step=1690,  loss=0.609,  accuracy=0.844\n",
      "Epoch  4,  step=1691,  loss=0.051,  accuracy=1.000\n",
      "Epoch  4,  step=1692,  loss=0.123,  accuracy=0.969\n",
      "Epoch  4,  step=1693,  loss=0.310,  accuracy=0.875\n",
      "Epoch  4,  step=1694,  loss=0.099,  accuracy=0.969\n",
      "Epoch  4,  step=1695,  loss=0.228,  accuracy=0.906\n",
      "Epoch  4,  step=1696,  loss=0.310,  accuracy=0.906\n",
      "Epoch  4,  step=1697,  loss=0.095,  accuracy=0.938\n",
      "Epoch  4,  step=1698,  loss=0.122,  accuracy=0.969\n",
      "Epoch  4,  step=1699,  loss=0.682,  accuracy=0.938\n",
      "Epoch  4,  step=1700,  loss=0.190,  accuracy=0.938\n",
      "Epoch  4,  step=1701,  loss=0.443,  accuracy=0.906\n",
      "Epoch  4,  step=1702,  loss=0.064,  accuracy=1.000\n",
      "Epoch  4,  step=1703,  loss=0.379,  accuracy=0.906\n",
      "Epoch  4,  step=1704,  loss=0.241,  accuracy=0.875\n",
      "Epoch  4,  step=1705,  loss=0.462,  accuracy=0.844\n",
      "Epoch  4,  step=1706,  loss=0.081,  accuracy=0.969\n",
      "Epoch  4,  step=1707,  loss=0.247,  accuracy=0.906\n",
      "Epoch  4,  step=1708,  loss=0.184,  accuracy=0.969\n",
      "Epoch  4,  step=1709,  loss=0.142,  accuracy=0.969\n",
      "Epoch  4,  step=1710,  loss=0.064,  accuracy=0.969\n",
      "Epoch  4,  step=1711,  loss=0.361,  accuracy=0.906\n",
      "Epoch  4,  step=1712,  loss=0.257,  accuracy=0.906\n",
      "Epoch  4,  step=1713,  loss=0.171,  accuracy=0.969\n",
      "Epoch  4,  step=1714,  loss=0.365,  accuracy=0.906\n",
      "Epoch  4,  step=1715,  loss=0.578,  accuracy=0.875\n",
      "Epoch  4,  step=1716,  loss=0.446,  accuracy=0.875\n",
      "Epoch  4,  step=1717,  loss=0.469,  accuracy=0.906\n",
      "Epoch  4,  step=1718,  loss=0.214,  accuracy=0.906\n",
      "Epoch  4,  step=1719,  loss=0.317,  accuracy=0.875\n",
      "Epoch  4,  step=1720,  loss=0.448,  accuracy=0.844\n",
      "Epoch  4,  step=1721,  loss=0.259,  accuracy=0.906\n",
      "Epoch  4,  step=1722,  loss=0.412,  accuracy=0.906\n",
      "Epoch  4,  step=1723,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step=1724,  loss=0.186,  accuracy=0.938\n",
      "Epoch  4,  step=1725,  loss=0.195,  accuracy=0.906\n",
      "Epoch  4,  step=1726,  loss=0.141,  accuracy=0.969\n",
      "Epoch  4,  step=1727,  loss=0.136,  accuracy=0.906\n",
      "Epoch  4,  step=1728,  loss=0.285,  accuracy=0.875\n",
      "Epoch  4,  step=1729,  loss=0.392,  accuracy=0.938\n",
      "Epoch  4,  step=1730,  loss=0.080,  accuracy=0.969\n",
      "Epoch  4,  step=1731,  loss=0.088,  accuracy=1.000\n",
      "Epoch  4,  step=1732,  loss=0.197,  accuracy=0.938\n",
      "Epoch  4,  step=1733,  loss=0.223,  accuracy=0.969\n",
      "Epoch  4,  step=1734,  loss=0.406,  accuracy=0.875\n",
      "Epoch  4,  step=1735,  loss=0.246,  accuracy=0.906\n",
      "Epoch  4,  step=1736,  loss=0.034,  accuracy=1.000\n",
      "Epoch  4,  step=1737,  loss=0.152,  accuracy=0.969\n",
      "Epoch  4,  step=1738,  loss=0.098,  accuracy=1.000\n",
      "Epoch  4,  step=1739,  loss=0.216,  accuracy=0.938\n",
      "Epoch  4,  step=1740,  loss=0.158,  accuracy=0.969\n",
      "Epoch  4,  step=1741,  loss=0.638,  accuracy=0.812\n",
      "Epoch  4,  step=1742,  loss=0.216,  accuracy=0.938\n",
      "Epoch  4,  step=1743,  loss=0.310,  accuracy=0.938\n",
      "Epoch  4,  step=1744,  loss=0.144,  accuracy=0.938\n",
      "Epoch  4,  step=1745,  loss=0.513,  accuracy=0.906\n",
      "Epoch  4,  step=1746,  loss=0.338,  accuracy=0.844\n",
      "Epoch  4,  step=1747,  loss=0.251,  accuracy=0.938\n",
      "Epoch  4,  step=1748,  loss=0.080,  accuracy=0.969\n",
      "Epoch  4,  step=1749,  loss=0.173,  accuracy=0.938\n",
      "Epoch  4,  step=1750,  loss=0.204,  accuracy=0.969\n",
      "Epoch  4,  step=1751,  loss=0.062,  accuracy=1.000\n",
      "Epoch  4,  step=1752,  loss=0.174,  accuracy=0.906\n",
      "Epoch  4,  step=1753,  loss=0.073,  accuracy=0.938\n",
      "Epoch  4,  step=1754,  loss=0.279,  accuracy=0.938\n",
      "Epoch  4,  step=1755,  loss=0.119,  accuracy=0.969\n",
      "Epoch  4,  step=1756,  loss=0.545,  accuracy=0.875\n",
      "Epoch  4,  step=1757,  loss=0.112,  accuracy=0.969\n",
      "Epoch  4,  step=1758,  loss=0.285,  accuracy=0.875\n",
      "Epoch  4,  step=1759,  loss=0.374,  accuracy=0.875\n",
      "Epoch  4,  step=1760,  loss=0.190,  accuracy=0.969\n",
      "Epoch  4,  step=1761,  loss=0.200,  accuracy=0.875\n",
      "Epoch  4,  step=1762,  loss=0.275,  accuracy=0.875\n",
      "Epoch  4,  step=1763,  loss=0.174,  accuracy=0.938\n",
      "Epoch  4,  step=1764,  loss=1.066,  accuracy=0.750\n",
      "Epoch  4,  step=1765,  loss=0.407,  accuracy=0.906\n",
      "Epoch  4,  step=1766,  loss=0.041,  accuracy=1.000\n",
      "Epoch  4,  step=1767,  loss=0.048,  accuracy=1.000\n",
      "Epoch  4,  step=1768,  loss=0.304,  accuracy=0.875\n",
      "Epoch  4,  step=1769,  loss=0.318,  accuracy=0.906\n",
      "Epoch  4,  step=1770,  loss=0.389,  accuracy=0.875\n",
      "Epoch  4,  step=1771,  loss=0.249,  accuracy=0.969\n",
      "Epoch  4,  step=1772,  loss=0.045,  accuracy=1.000\n",
      "Epoch  4,  step=1773,  loss=0.066,  accuracy=1.000\n",
      "Epoch  4,  step=1774,  loss=0.176,  accuracy=0.938\n",
      "Epoch  4,  step=1775,  loss=0.201,  accuracy=0.938\n",
      "Epoch  4,  step=1776,  loss=0.196,  accuracy=0.938\n",
      "Epoch  4,  step=1777,  loss=0.084,  accuracy=1.000\n",
      "Epoch  4,  step=1778,  loss=0.221,  accuracy=0.938\n",
      "Epoch  4,  step=1779,  loss=0.040,  accuracy=1.000\n",
      "Epoch  4,  step=1780,  loss=0.210,  accuracy=0.938\n",
      "Epoch  4,  step=1781,  loss=0.427,  accuracy=0.812\n",
      "Epoch  4,  step=1782,  loss=0.508,  accuracy=0.938\n",
      "Epoch  4,  step=1783,  loss=0.505,  accuracy=0.906\n",
      "Epoch  4,  step=1784,  loss=0.072,  accuracy=0.969\n",
      "Epoch  4,  step=1785,  loss=0.091,  accuracy=0.969\n",
      "Epoch  4,  step=1786,  loss=0.260,  accuracy=0.875\n",
      "Epoch  4,  step=1787,  loss=0.164,  accuracy=0.938\n",
      "Epoch  4,  step=1788,  loss=0.299,  accuracy=0.906\n",
      "Epoch  4,  step=1789,  loss=0.255,  accuracy=0.875\n",
      "Epoch  4,  step=1790,  loss=0.333,  accuracy=0.844\n",
      "Epoch  4,  step=1791,  loss=0.457,  accuracy=0.844\n",
      "Epoch  4,  step=1792,  loss=0.070,  accuracy=1.000\n",
      "Epoch  4,  step=1793,  loss=0.266,  accuracy=0.938\n",
      "Epoch  4,  step=1794,  loss=0.106,  accuracy=0.969\n",
      "Epoch  4,  step=1795,  loss=0.152,  accuracy=0.938\n",
      "Epoch  4,  step=1796,  loss=0.302,  accuracy=0.906\n",
      "Epoch  4,  step=1797,  loss=0.352,  accuracy=0.906\n",
      "Epoch  4,  step=1798,  loss=0.257,  accuracy=0.906\n",
      "Epoch  4,  step=1799,  loss=0.327,  accuracy=0.844\n",
      "Epoch  4,  step=1800,  loss=0.528,  accuracy=0.875\n",
      "Epoch  4,  step=1801,  loss=0.455,  accuracy=0.812\n",
      "Epoch  4,  step=1802,  loss=0.261,  accuracy=0.938\n",
      "Epoch  4,  step=1803,  loss=0.405,  accuracy=0.906\n",
      "Epoch  4,  step=1804,  loss=0.468,  accuracy=0.875\n",
      "Epoch  4,  step=1805,  loss=0.086,  accuracy=0.969\n",
      "Epoch  4,  step=1806,  loss=0.184,  accuracy=0.906\n",
      "Epoch  4,  step=1807,  loss=0.331,  accuracy=0.875\n",
      "Epoch  4,  step=1808,  loss=0.209,  accuracy=0.938\n",
      "Epoch  4,  step=1809,  loss=0.094,  accuracy=1.000\n",
      "Epoch  4,  step=1810,  loss=0.169,  accuracy=0.938\n",
      "Epoch  4,  step=1811,  loss=0.176,  accuracy=0.906\n",
      "Epoch  4,  step=1812,  loss=0.101,  accuracy=0.969\n",
      "Epoch  4,  step=1813,  loss=0.145,  accuracy=0.938\n",
      "Epoch  4,  step=1814,  loss=0.401,  accuracy=0.875\n",
      "Epoch  4,  step=1815,  loss=0.555,  accuracy=0.906\n",
      "Epoch  4,  step=1816,  loss=0.044,  accuracy=1.000\n",
      "Epoch  4,  step=1817,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step=1818,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step=1819,  loss=0.071,  accuracy=0.969\n",
      "Epoch  4,  step=1820,  loss=0.113,  accuracy=0.938\n",
      "Epoch  4,  step=1821,  loss=0.076,  accuracy=1.000\n",
      "Epoch  4,  step=1822,  loss=0.215,  accuracy=0.938\n",
      "Epoch  4,  step=1823,  loss=0.030,  accuracy=1.000\n",
      "Epoch  4,  step=1824,  loss=0.194,  accuracy=0.938\n",
      "Epoch  4,  step=1825,  loss=0.168,  accuracy=0.938\n",
      "Epoch  4,  step=1826,  loss=0.112,  accuracy=0.969\n",
      "Epoch  4,  step=1827,  loss=0.148,  accuracy=0.938\n",
      "Epoch  4,  step=1828,  loss=0.282,  accuracy=0.938\n",
      "Epoch  4,  step=1829,  loss=0.304,  accuracy=0.938\n",
      "Epoch  4,  step=1830,  loss=0.072,  accuracy=0.969\n",
      "Epoch  4,  step=1831,  loss=0.191,  accuracy=0.906\n",
      "Epoch  4,  step=1832,  loss=0.189,  accuracy=0.938\n",
      "Epoch  4,  step=1833,  loss=0.158,  accuracy=0.938\n",
      "Epoch  4,  step=1834,  loss=0.074,  accuracy=0.969\n",
      "Epoch  4,  step=1835,  loss=0.028,  accuracy=1.000\n",
      "Epoch  4,  step=1836,  loss=0.133,  accuracy=0.969\n",
      "Epoch  4,  step=1837,  loss=0.231,  accuracy=0.844\n",
      "Epoch  4,  step=1838,  loss=0.275,  accuracy=0.906\n",
      "Epoch  4,  step=1839,  loss=0.399,  accuracy=0.938\n",
      "Epoch  4,  step=1840,  loss=0.091,  accuracy=0.938\n",
      "Epoch  4,  step=1841,  loss=0.017,  accuracy=1.000\n",
      "Epoch  4,  step=1842,  loss=0.017,  accuracy=1.000\n",
      "Epoch  4,  step=1843,  loss=0.164,  accuracy=0.938\n",
      "Epoch  4,  step=1844,  loss=0.026,  accuracy=1.000\n",
      "Epoch  4,  step=1845,  loss=0.172,  accuracy=0.938\n",
      "Epoch  4,  step=1846,  loss=0.188,  accuracy=0.906\n",
      "Epoch  4,  step=1847,  loss=0.085,  accuracy=0.969\n",
      "Epoch  4,  step=1848,  loss=0.215,  accuracy=0.938\n",
      "Epoch  4,  step=1849,  loss=0.154,  accuracy=0.969\n",
      "Epoch  4,  step=1850,  loss=0.121,  accuracy=0.938\n",
      "Epoch  4,  step=1851,  loss=0.123,  accuracy=0.969\n",
      "Epoch  4,  step=1852,  loss=0.416,  accuracy=0.812\n",
      "Epoch  4,  step=1853,  loss=0.794,  accuracy=0.844\n",
      "Epoch  4,  step=1854,  loss=0.603,  accuracy=0.844\n",
      "Epoch  4,  step=1855,  loss=0.209,  accuracy=0.938\n",
      "Epoch  4,  step=1856,  loss=0.222,  accuracy=0.969\n",
      "Epoch  4,  step=1857,  loss=0.171,  accuracy=0.906\n",
      "Epoch  4,  step=1858,  loss=0.080,  accuracy=1.000\n",
      "Epoch  4,  step=1859,  loss=0.090,  accuracy=0.938\n",
      "Epoch  4,  step=1860,  loss=0.081,  accuracy=0.969\n",
      "Epoch  4,  step=1861,  loss=0.095,  accuracy=0.969\n",
      "Epoch  4,  step=1862,  loss=0.080,  accuracy=0.938\n",
      "Epoch  4,  step=1863,  loss=0.067,  accuracy=1.000\n",
      "Epoch  4,  step=1864,  loss=0.307,  accuracy=0.938\n",
      "Epoch  4,  step=1865,  loss=0.287,  accuracy=0.938\n",
      "Epoch  4,  step=1866,  loss=1.058,  accuracy=0.719\n",
      "Epoch  4,  step=1867,  loss=0.169,  accuracy=0.938\n",
      "Epoch  4,  step=1868,  loss=0.293,  accuracy=0.844\n",
      "Epoch  4,  step=1869,  loss=0.089,  accuracy=0.969\n",
      "Epoch  4,  step=1870,  loss=0.037,  accuracy=1.000\n",
      "Epoch  4,  step=1871,  loss=0.035,  accuracy=1.000\n",
      "Epoch  4,  step=1872,  loss=0.854,  accuracy=0.906\n",
      "Epoch  4,  step=1873,  loss=0.142,  accuracy=0.938\n",
      "Epoch  4,  step=1874,  loss=0.117,  accuracy=0.969\n"
     ]
    }
   ],
   "source": [
    "\n",
    "W1 = torch.randn( 784, 100, requires_grad=True )\n",
    "b1 = torch.randn( 100, requires_grad=True )\n",
    "\n",
    "W2 = torch.randn( 100, 10, requires_grad=True )\n",
    "b2 = torch.randn( 10, requires_grad=True )\n",
    "\n",
    "nr_epochs = 5\n",
    "alpha = torch.tensor(0.1)\n",
    "batch_size = 32\n",
    "\n",
    "for epoch in range(nr_epochs):\n",
    "\n",
    "    for batch_nr in range(len(X_train) // batch_size):\n",
    "        \n",
    "        x_batch = X_train[ batch_nr*batch_size :  (batch_nr+1)*batch_size ]\n",
    "        y_batch = y_train[ batch_nr*batch_size :  (batch_nr+1)*batch_size ]\n",
    "    \n",
    "        # Forward propagation\n",
    "        Z1 = torch.matmul( x_batch , W1 ) + b1\n",
    "        A1 = torch.tanh( Z1 )\n",
    "\n",
    "        Z2 = torch.matmul( A1, W2 ) + b2 # logits\n",
    "        A2 = torch.softmax( Z2, dim=1 )\n",
    "\n",
    "        loss = torch.nn.functional.cross_entropy( Z2, y_batch )\n",
    "\n",
    "        with torch.no_grad():\n",
    "            acc_train = torch.sum( torch.argmax( A2, dim=1 ) == y_batch ) / len(y_batch) \n",
    "        print(f\"Epoch {epoch:2d},  step={batch_nr:4d},  loss={loss.item():.3f},  accuracy={acc_train:.3f}\")\n",
    "\n",
    "        # Backward propagation\n",
    "        loss.backward()\n",
    "\n",
    "        with torch.no_grad():\n",
    "            W1 -= alpha * W1.grad\n",
    "            W2 -= alpha * W2.grad\n",
    "            b1 -= alpha * b1.grad\n",
    "            b2 -= alpha * b2.grad\n",
    "\n",
    "        W1.grad.zero_()\n",
    "        W2.grad.zero_()\n",
    "        b1.grad.zero_()\n",
    "        b2.grad.zero_()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "2abc447e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  4,  loss=0.277,  accuracy=0.917\n"
     ]
    }
   ],
   "source": [
    "with torch.no_grad():\n",
    "    Z1 = torch.matmul( X_train , W1 ) + b1\n",
    "    A1 = torch.tanh( Z1 )\n",
    "\n",
    "    Z2 = torch.matmul( A1, W2 ) + b2 # logits\n",
    "    A2 = torch.softmax( Z2, dim=1 )\n",
    "\n",
    "    loss = torch.nn.functional.cross_entropy( Z2, y_train )\n",
    "    acc_train = torch.sum( torch.argmax( A2, dim=1 ) == y_train ) / len(y_train) \n",
    "\n",
    "print(f\"Epoch {epoch:2d},  loss={loss.item():.3f},  accuracy={acc_train:.3f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "5c437887",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test = dataset_test.targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "d20366ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  4,  loss=0.327,  accuracy=0.904\n"
     ]
    }
   ],
   "source": [
    "with torch.no_grad():\n",
    "    Z1 = torch.matmul( X_test , W1 ) + b1\n",
    "    A1 = torch.tanh( Z1 )\n",
    "\n",
    "    Z2 = torch.matmul( A1, W2 ) + b2 # logits\n",
    "    A2 = torch.softmax( Z2, dim=1 )\n",
    "\n",
    "    loss = torch.nn.functional.cross_entropy( Z2, y_test )\n",
    "    acc_train = torch.sum( torch.argmax( A2, dim=1 ) == y_test ) / len(y_test) \n",
    "\n",
    "print(f\"Epoch {epoch:2d},  loss={loss.item():.3f},  accuracy={acc_train:.3f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "808fd083",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}