DenseNet.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "Hr6uu3lMzh0M"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.python.training.moving_averages import assign_moving_average\n",
    "import numpy as np\n",
    "import os, pdb\n",
    "import cv2\n",
    "import numpy as np\n",
    "import random as rn\n",
    "import tensorflow as tf\n",
    "import threading\n",
    "import time\n",
    "\n",
    "global n_classes, count\n",
    "n_classes = 50\n",
    "count=0 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "tCNUC9U1zh0U"
   },
   "outputs": [],
   "source": [
    "#==========================================================================\n",
    "#=============Reading data in multithreading manner========================\n",
    "#==========================================================================\n",
    "def read_labeled_image_list(image_list_file, training_img_dir):\n",
    "    \"\"\"Reads a .txt file containing pathes and labeles\n",
    "    Args:\n",
    "       image_list_file: a .txt file with one /path/to/image per line\n",
    "       label: optionally, if set label will be pasted after each line\n",
    "    Returns:\n",
    "       List with all filenames in file image_list_file\n",
    "    \"\"\"\n",
    "    f = open(image_list_file, 'r')\n",
    "    filenames = []\n",
    "    labels = []\n",
    "\n",
    "    for line in f:\n",
    "        filename, label = line[:-1].split(' ')\n",
    "        filename = training_img_dir+filename\n",
    "        filenames.append(filename)\n",
    "        labels.append(int(label))\n",
    "        \n",
    "    return filenames, labels\n",
    "    \n",
    "    \n",
    "def read_images_from_disk(input_queue, size1=256):\n",
    "    \"\"\"Consumes a single filename and label as a ' '-delimited string.\n",
    "    Args:\n",
    "      filename_and_label_tensor: A scalar string tensor.\n",
    "    Returns:\n",
    "      Two tensors: the decoded image, and the string label.\n",
    "    \"\"\"\n",
    "    label = input_queue[1]\n",
    "    fn=input_queue[0]\n",
    "    file_contents = tf.read_file(input_queue[0])\n",
    "    example = tf.image.decode_jpeg(file_contents, channels=3)\n",
    "    \n",
    "    #example = tf.image.decode_png(file_contents, channels=3, name=\"dataset_image\") # png fo rlfw\n",
    "    example=tf.image.resize_images(example, [size1,size1])\n",
    "    return example, label, fn\n",
    "\n",
    "def setup_inputs(sess, filenames, training_img_dir, image_size=256, crop_size=224, isTest=False, batch_size=64):\n",
    "    \n",
    "    # Read each image file\n",
    "    image_list, label_list = read_labeled_image_list(filenames, training_img_dir)\n",
    "\n",
    "    images = tf.cast(image_list, tf.string)\n",
    "    labels = tf.cast(label_list, tf.int64)\n",
    "     # Makes an input queue\n",
    "    if isTest is False:\n",
    "        isShuffle = True\n",
    "        numThr = 4\n",
    "    else:\n",
    "        isShuffle = False\n",
    "        numThr = 1\n",
    "        \n",
    "    input_queue = tf.train.slice_input_producer([images, labels], shuffle=isShuffle)\n",
    "    image, y,fn = read_images_from_disk(input_queue)\n",
    "\n",
    "    channels = 3\n",
    "    image.set_shape([None, None, channels])\n",
    "        \n",
    "    # Crop and other random augmentations\n",
    "    if isTest is False:\n",
    "        image = tf.image.random_flip_left_right(image)\n",
    "        image = tf.image.random_flip_up_down(image)\n",
    "        image = tf.image.random_hue(image,0.05)\n",
    "        image = tf.image.random_saturation(image, .95, 1.05)\n",
    "        image = tf.image.random_brightness(image, .05)\n",
    "        image = tf.image.random_contrast(image, .95, 1.05)\n",
    "        \n",
    "\n",
    "    image = tf.random_crop(image, [crop_size, crop_size, 3])\n",
    "    image = tf.cast(image, tf.float32)/255.0\n",
    "    \n",
    "    image, y,fn = tf.train.batch([image, y, fn], batch_size=batch_size, capacity=batch_size*3, num_threads=numThr, name='labels_and_images')\n",
    "\n",
    "    tf.train.start_queue_runners(sess=sess)\n",
    "\n",
    "    return image, y, fn, len(label_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def activation(x,name=\"activation\"):\n",
    "    return tf.nn.swish(x, name=name)\n",
    "    \n",
    "def conv2d(name, l_input, w, b, s, p):\n",
    "    l_input = tf.nn.conv2d(l_input, w, strides=[1,s,s,1], padding=p, name=name)\n",
    "    l_input = l_input+b\n",
    "\n",
    "    return l_input\n",
    "\n",
    "def batchnorm(conv, isTraining, name='bn'):\n",
    "    return tf.layers.batch_normalization(conv, training=isTraining, name=\"bn\"+name)\n",
    "\n",
    "def initializer(in_filters, out_filters, name):\n",
    "    w1 = tf.get_variable(name+\"W\", [3, 3, in_filters, out_filters], initializer=tf.truncated_normal_initializer())\n",
    "    b1 = tf.get_variable(name+\"B\", [out_filters], initializer=tf.truncated_normal_initializer())\n",
    "    return w1, b1\n",
    "\n",
    "def merge(in_x, growth_rate, isTraining, name):\n",
    "    channels = int(in_x.get_shape()[-1])\n",
    "    w1, b1 = initializer(channels, growth_rate, name+\"merge_unit\")\n",
    "    x = batchnorm(in_x, isTraining, name=name+'mer')\n",
    "    x = activation(x)\n",
    "    x = conv2d(name+'r1', x, w1, b1, 1, \"SAME\")\n",
    "    return x\n",
    "\n",
    "def block_unit(in_x, growth_rate,isTraining, name):\n",
    "    channels = int(in_x.get_shape()[-1])\n",
    "    w1, b1 = initializer(channels, 4*growth_rate, name+\"block_unit\")\n",
    "    x = batchnorm(in_x, isTraining, name=name+'unit')\n",
    "    x = activation(x)\n",
    "    x = conv2d(name+'r1', x, w1, b1, 1, \"SAME\")\n",
    "    x = merge(x, growth_rate, isTraining, name) #NOTE: output/input_tensor_depth is different from input\n",
    "    return x\n",
    "    \n",
    "def transition_layer(in_x,isTraining, theta, name):\n",
    "    channels = int(in_x.get_shape()[-1])\n",
    "    x = batchnorm(in_x, isTraining, name=name+'tra')\n",
    "    w1, b1 = initializer(channels, channels*theta, name+\"transition_lay\")\n",
    "    x = conv2d(name+'r1', x, w1, b1, 1, \"SAME\")\n",
    "    x = tf.nn.avg_pool(value=x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME', name=name+'avg_pool_2x2')\n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "zyjBt-MqVce6"
   },
   "outputs": [],
   "source": [
    "\n",
    "def dense_block(in_x, growth_rate,isTraining, name):\n",
    "    channels = int(in_x.get_shape()[-1])\n",
    "    with tf.name_scope(\"block1\") as scope:\n",
    "        x = block_unit(in_x, growth_rate, isTraining, name+\"block1\")#NOTE:64 is 2k, here k = 32, 128 is 4k, output is k = 32\n",
    "        x = tf.concat(values=[in_x, x], axis=3, name=name+'stack0')# 96\n",
    "        print(\"Denseblock1 with channels %d\"%(x.get_shape().as_list()[-1]))\n",
    "\n",
    "    with tf.name_scope(\"block2\") as scope:\n",
    "        out_x = block_unit(x, growth_rate, isTraining, name+\"block2\")\n",
    "        x = tf.concat(values=[x, out_x], axis=3, name=name+'stack1')\n",
    "        print(\"Denseblock2 with channels %d\"%(x.get_shape().as_list()[-1]))\n",
    "\n",
    "    with tf.name_scope(\"block3\") as scope:\n",
    "        out_x = block_unit(x, growth_rate, isTraining, name+\"block3\")\n",
    "        x = tf.concat(values=[x, out_x], axis=3, name=name+'stack2')\n",
    "        print(\"Denseblock3 with channels %d\"%(x.get_shape().as_list()[-1]))\n",
    "\n",
    "    with tf.name_scope(\"block4\") as scope:\n",
    "        out_x = block_unit(x, growth_rate, isTraining, name+\"block4\")\n",
    "        x = tf.concat(values=[x, out_x], axis=3, name=name+'stack3')\n",
    "        print(\"Denseblock4 with channels %d\"%(x.get_shape().as_list()[-1]))\n",
    "\n",
    "    return x\n",
    "\n",
    "      \n",
    "def DenseNet(_X, isTraining,growth_rate):\n",
    "    global n_classes\n",
    "    w1 = tf.get_variable(\"initW\", [7, 7, 3, 96], initializer=tf.truncated_normal_initializer())\n",
    "    b1 = tf.get_variable(\"initB\", [96], initializer=tf.truncated_normal_initializer())\n",
    "    x = conv2d('conv1', _X, w1, b1, 4, \"VALID\")\n",
    "    num_blocks = 4\n",
    "    for j in range(num_blocks):\n",
    "        x =dense_block(x, growth_rate, isTraining, \"denb%d\"%(j))\n",
    "        \n",
    "        x =transition_layer(x, isTraining, 0.5, \"transb%d\"%(j))\n",
    "\n",
    "    channels = int(x.get_shape()[-1])\n",
    "    wo, bo=initializer(channels, n_classes, \"FinalOutput\")\n",
    "    x = conv2d('final', x, wo, bo, 1, \"SAME\")\n",
    "    \n",
    "    x = activation(x)\n",
    "    x = batchnorm(x, isTraining, name='FinalBn')\n",
    "    \n",
    "    x = tf.reduce_mean(x, [1,2])\n",
    "    W = tf.get_variable(\"FinalW\", [n_classes, n_classes], initializer=tf.truncated_normal_initializer())\n",
    "    b = tf.get_variable(\"FinalB\", [n_classes], initializer=tf.truncated_normal_initializer())\n",
    "    \n",
    "    out = tf.matmul(x, W) + b\n",
    "                            \n",
    "\n",
    "    return out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "h7AtFd4czh0W"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GO!!\n"
     ]
    }
   ],
   "source": [
    "batch_size = 32\n",
    "display_step = 20\n",
    "learning_rate = tf.placeholder(tf.float32)      # Learning rate to be fed\n",
    "lr = 1e-3                          # Learning rate start\n",
    "print('GO!!')\n",
    "with open('val.txt','r') as fp:\n",
    "    tlen1 = len(fp.readlines())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "7LCc7xsQzh0Y",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing the training & validation data...\n",
      "WARNING:tensorflow:From <ipython-input-2-d5c7c5a21643>:56: slice_input_producer (from tensorflow.python.training.input) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Queue-based input pipelines have been replaced by `tf.data`. Use `tf.data.Dataset.from_tensor_slices(tuple(tensor_list)).shuffle(tf.shape(input_tensor, out_type=tf.int64)[0]).repeat(num_epochs)`. If `shuffle=False`, omit the `.shuffle(...)`.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\input.py:374: range_input_producer (from tensorflow.python.training.input) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Queue-based input pipelines have been replaced by `tf.data`. Use `tf.data.Dataset.range(limit).shuffle(limit).repeat(num_epochs)`. If `shuffle=False`, omit the `.shuffle(...)`.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\input.py:320: input_producer (from tensorflow.python.training.input) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Queue-based input pipelines have been replaced by `tf.data`. Use `tf.data.Dataset.from_tensor_slices(input_tensor).shuffle(tf.shape(input_tensor, out_type=tf.int64)[0]).repeat(num_epochs)`. If `shuffle=False`, omit the `.shuffle(...)`.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\input.py:190: limit_epochs (from tensorflow.python.training.input) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Queue-based input pipelines have been replaced by `tf.data`. Use `tf.data.Dataset.from_tensors(tensor).repeat(num_epochs)`.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\input.py:199: QueueRunner.__init__ (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "To construct input pipelines, use the `tf.data` module.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\input.py:199: add_queue_runner (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "To construct input pipelines, use the `tf.data` module.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\input.py:202: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\ops\\control_flow_ops.py:3632: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n",
      "WARNING:tensorflow:From <ipython-input-2-d5c7c5a21643>:75: batch (from tensorflow.python.training.input) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Queue-based input pipelines have been replaced by `tf.data`. Use `tf.data.Dataset.batch(batch_size)` (or `padded_batch(...)` if `dynamic_pad=True`).\n",
      "WARNING:tensorflow:From <ipython-input-2-d5c7c5a21643>:77: start_queue_runners (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "To construct input pipelines, use the `tf.data` module.\n",
      "Preparing the training model with learning rate = 0.00100...\n",
      "WARNING:tensorflow:From <ipython-input-3-556ebf206a92>:11: batch_normalization (from tensorflow.python.layers.normalization) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use keras.layers.batch_normalization instead.\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\framework\\function.py:1007: calling Graph.create_op (from tensorflow.python.framework.ops) with compute_shapes is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Shapes are always computed; don't use the compute_shapes as it has no effect.\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "Denseblock1 with channels 120\n",
      "Denseblock2 with channels 144\n",
      "Denseblock3 with channels 168\n",
      "Denseblock4 with channels 192\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\ops\\math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n"
     ]
    }
   ],
   "source": [
    "# Setup the tensorflow...\n",
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth = True\n",
    "sess = tf.Session(config=config)\n",
    "\n",
    "print(\"Preparing the training & validation data...\")\n",
    "train_data, train_labels, filelist1, glen1 = setup_inputs(sess, \"train.txt\", \"\", batch_size=batch_size)\n",
    "val_data, val_labels, filelist2, tlen1 = setup_inputs(sess, \"val.txt\", \"\",isTest=True, batch_size=batch_size)\n",
    "\n",
    "max_iter = glen1*400\n",
    "print(\"Preparing the training model with learning rate = %.5f...\" % (lr))\n",
    "\n",
    "\n",
    "with tf.variable_scope(\"DenseNet\") as scope:\n",
    "  pred = DenseNet(train_data, True, 24)\n",
    "  scope.reuse_variables()\n",
    "  valpred = DenseNet(val_data, False, 24)\n",
    "\n",
    "with tf.name_scope('Loss_and_Accuracy'):\n",
    "  update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)\n",
    "  with tf.control_dependencies(update_ops):\n",
    "    cost = tf.losses.sparse_softmax_cross_entropy(labels=train_labels, logits=pred)\n",
    "    optimizer = tf.train.AdamOptimizer(learning_rate).minimize(cost)\n",
    "    \n",
    "  correct_prediction = tf.equal(tf.argmax(pred, 1), train_labels)\n",
    "  accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "\n",
    "  \n",
    "  correct_prediction2 = tf.equal(tf.argmax(valpred,1), val_labels)\n",
    "  accuracy2 = tf.reduce_mean(tf.cast(correct_prediction2, tf.float32))\n",
    "  \n",
    "  tf.summary.scalar('Loss', cost)\n",
    "  tf.summary.scalar('Training_Accuracy', accuracy)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "3CGfQ_Cbzh0c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We are going to train the ImageNet model based on denseNet!!!\n",
      "Iter=672/epoch=0, Loss=7.807419, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=1312/epoch=0, Loss=6.838737, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=1952/epoch=0, Loss=5.792522, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=2592/epoch=0, Loss=5.675241, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=3232/epoch=0, Loss=5.670481, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=3872/epoch=0, Loss=6.387673, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=4512/epoch=0, Loss=6.495638, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=5152/epoch=0, Loss=5.403708, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=5792/epoch=0, Loss=6.241925, Training Accuracy=0.031250, lr=0.001000\n",
      "\n",
      "Iter=6400/epoch=0, Validation Accuracy=0.024554\n",
      "Iter=6432/epoch=0, Loss=4.965719, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=7072/epoch=0, Loss=5.233444, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=7712/epoch=0, Loss=5.229833, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=8352/epoch=0, Loss=5.713189, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=8992/epoch=0, Loss=4.284764, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=9632/epoch=0, Loss=5.326404, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=10272/epoch=0, Loss=5.459352, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=10912/epoch=0, Loss=4.641769, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=11552/epoch=0, Loss=5.013316, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=12192/epoch=0, Loss=4.759883, Training Accuracy=0.000000, lr=0.001000\n",
      "\n",
      "Iter=12800/epoch=0, Validation Accuracy=0.050223\n",
      "Iter=12832/epoch=0, Loss=4.921036, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=13472/epoch=0, Loss=4.801101, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=14112/epoch=0, Loss=4.469523, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=14752/epoch=0, Loss=4.824238, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=15392/epoch=0, Loss=5.098574, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=16032/epoch=0, Loss=4.291076, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=16672/epoch=0, Loss=4.370993, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=17312/epoch=0, Loss=4.837467, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=17952/epoch=0, Loss=4.750093, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=18592/epoch=0, Loss=4.753622, Training Accuracy=0.031250, lr=0.001000\n",
      "\n",
      "Iter=19200/epoch=0, Validation Accuracy=0.056920\n",
      "Iter=19232/epoch=0, Loss=4.668517, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=19872/epoch=0, Loss=4.242735, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=20512/epoch=0, Loss=4.797921, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=21152/epoch=0, Loss=4.422206, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=21792/epoch=0, Loss=4.890947, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=22432/epoch=0, Loss=4.701829, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=23072/epoch=0, Loss=4.297168, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=23712/epoch=0, Loss=4.937227, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=24352/epoch=0, Loss=4.433956, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=24992/epoch=0, Loss=4.525991, Training Accuracy=0.031250, lr=0.001000\n",
      "\n",
      "Iter=25600/epoch=0, Validation Accuracy=0.059152\n",
      "Iter=25632/epoch=0, Loss=4.354241, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=26272/epoch=0, Loss=4.620420, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=26912/epoch=0, Loss=4.162121, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=27552/epoch=0, Loss=4.514984, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=28192/epoch=0, Loss=4.377581, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=28832/epoch=0, Loss=4.312712, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=29472/epoch=0, Loss=4.676558, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=30112/epoch=0, Loss=4.729206, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=30752/epoch=0, Loss=4.564183, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=31392/epoch=0, Loss=4.729299, Training Accuracy=0.031250, lr=0.001000\n",
      "\n",
      "Iter=32000/epoch=0, Validation Accuracy=0.064732\n",
      "Model saved in file at iteration 32032: ./tf_densenet_model_iter1001.ckpt\n",
      "Iter=32032/epoch=0, Loss=4.725856, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=32672/epoch=0, Loss=4.045053, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=33312/epoch=0, Loss=4.311452, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=33952/epoch=0, Loss=4.524888, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=34592/epoch=0, Loss=4.760416, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=35232/epoch=0, Loss=4.253124, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=35872/epoch=0, Loss=4.283075, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=36512/epoch=0, Loss=3.952878, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=37152/epoch=0, Loss=3.841104, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=37792/epoch=0, Loss=4.119085, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=38400/epoch=0, Validation Accuracy=0.078125\n",
      "Iter=38432/epoch=0, Loss=4.337504, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=39072/epoch=0, Loss=4.116633, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=39712/epoch=0, Loss=4.125237, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=40352/epoch=0, Loss=4.121910, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=40992/epoch=0, Loss=4.026015, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=41632/epoch=0, Loss=4.145577, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=42272/epoch=0, Loss=4.141063, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=42912/epoch=0, Loss=4.290124, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=43552/epoch=0, Loss=4.240186, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=44192/epoch=0, Loss=4.005371, Training Accuracy=0.093750, lr=0.001000\n",
      "\n",
      "Iter=44800/epoch=0, Validation Accuracy=0.073661\n",
      "Iter=44832/epoch=0, Loss=4.403804, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=45472/epoch=0, Loss=3.774621, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=46112/epoch=0, Loss=4.312628, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=46752/epoch=0, Loss=3.902541, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=47392/epoch=0, Loss=4.140368, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=48032/epoch=0, Loss=4.088691, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=48672/epoch=0, Loss=3.903686, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=49312/epoch=0, Loss=4.260918, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=49952/epoch=0, Loss=4.208343, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=50592/epoch=0, Loss=4.162849, Training Accuracy=0.062500, lr=0.001000\n",
      "\n",
      "Iter=51200/epoch=0, Validation Accuracy=0.097098\n",
      "Iter=51232/epoch=0, Loss=4.058324, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=51872/epoch=0, Loss=4.080213, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=52512/epoch=0, Loss=4.372366, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=53152/epoch=0, Loss=4.037142, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=53792/epoch=0, Loss=4.208378, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=54432/epoch=0, Loss=4.318551, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=55072/epoch=0, Loss=4.413860, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=55712/epoch=0, Loss=4.097141, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=56352/epoch=0, Loss=4.397923, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=56992/epoch=0, Loss=4.359499, Training Accuracy=0.093750, lr=0.001000\n",
      "\n",
      "Iter=57600/epoch=0, Validation Accuracy=0.083705\n",
      "Iter=57632/epoch=0, Loss=4.174773, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=58272/epoch=0, Loss=3.808521, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=58912/epoch=0, Loss=4.158280, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=59552/epoch=0, Loss=3.940917, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=60192/epoch=0, Loss=4.331827, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=60832/epoch=0, Loss=4.131415, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=61472/epoch=0, Loss=4.502131, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=62112/epoch=0, Loss=4.111420, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=62752/epoch=0, Loss=3.797695, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=63392/epoch=0, Loss=4.183283, Training Accuracy=0.093750, lr=0.001000\n",
      "\n",
      "Iter=64000/epoch=0, Validation Accuracy=0.099330\n",
      "Model saved in file at iteration 64032: ./tf_densenet_model_iter2001.ckpt\n",
      "Iter=64032/epoch=0, Loss=3.895403, Training Accuracy=0.093750, lr=0.001000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=64672/epoch=0, Loss=3.961715, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=65312/epoch=0, Loss=4.295077, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=65952/epoch=0, Loss=3.946056, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=66592/epoch=0, Loss=4.034994, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=67232/epoch=0, Loss=3.856010, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=67872/epoch=0, Loss=3.950647, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=68512/epoch=0, Loss=3.880178, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=69152/epoch=0, Loss=4.138126, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=69792/epoch=0, Loss=4.262732, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=70400/epoch=0, Validation Accuracy=0.109375\n",
      "Iter=70432/epoch=0, Loss=3.981267, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=71072/epoch=0, Loss=3.637785, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=71712/epoch=0, Loss=3.676419, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=72352/epoch=0, Loss=4.160649, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=72992/epoch=0, Loss=3.888603, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=73632/epoch=0, Loss=3.901175, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=74272/epoch=0, Loss=3.924283, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=74912/epoch=0, Loss=3.680551, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=75552/epoch=0, Loss=3.839955, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=76192/epoch=0, Loss=3.990792, Training Accuracy=0.000000, lr=0.001000\n",
      "\n",
      "Iter=76800/epoch=0, Validation Accuracy=0.118304\n",
      "Iter=76832/epoch=0, Loss=3.797090, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=77472/epoch=0, Loss=4.015787, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=78112/epoch=0, Loss=3.629948, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=78752/epoch=0, Loss=3.971567, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=79392/epoch=0, Loss=4.127312, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=80032/epoch=0, Loss=3.890290, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=80672/epoch=0, Loss=3.485484, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=81312/epoch=0, Loss=4.409786, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=81952/epoch=0, Loss=3.847260, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=82592/epoch=0, Loss=3.581558, Training Accuracy=0.156250, lr=0.001000\n",
      "\n",
      "Iter=83200/epoch=0, Validation Accuracy=0.112723\n",
      "Iter=83232/epoch=0, Loss=3.946722, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=83872/epoch=0, Loss=3.719910, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=84512/epoch=0, Loss=3.782498, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=85152/epoch=0, Loss=3.681016, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=85792/epoch=0, Loss=4.082478, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=86432/epoch=0, Loss=4.215271, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=87072/epoch=0, Loss=4.162940, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=87712/epoch=0, Loss=3.716555, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=88352/epoch=0, Loss=3.796972, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=88992/epoch=0, Loss=4.144063, Training Accuracy=0.031250, lr=0.001000\n",
      "\n",
      "Iter=89600/epoch=0, Validation Accuracy=0.107143\n",
      "Iter=89632/epoch=0, Loss=4.061411, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=90272/epoch=0, Loss=4.031908, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=90912/epoch=0, Loss=3.576889, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=91552/epoch=0, Loss=4.032420, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=92192/epoch=0, Loss=4.042075, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=92832/epoch=0, Loss=3.751393, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=93472/epoch=0, Loss=3.725445, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=94112/epoch=0, Loss=3.800338, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=94752/epoch=0, Loss=4.170784, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=95392/epoch=0, Loss=3.784307, Training Accuracy=0.093750, lr=0.001000\n",
      "\n",
      "Iter=96000/epoch=0, Validation Accuracy=0.123884\n",
      "Model saved in file at iteration 96032: ./tf_densenet_model_iter3001.ckpt\n",
      "Iter=96032/epoch=0, Loss=3.841678, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=96672/epoch=0, Loss=4.013052, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=97312/epoch=0, Loss=3.848417, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=97952/epoch=0, Loss=3.665232, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=98592/epoch=0, Loss=3.892403, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=99232/epoch=0, Loss=3.368933, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=99872/epoch=0, Loss=3.759321, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=100512/epoch=0, Loss=4.136856, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=101152/epoch=0, Loss=3.789052, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=101792/epoch=0, Loss=3.493559, Training Accuracy=0.156250, lr=0.001000\n",
      "\n",
      "Iter=102400/epoch=0, Validation Accuracy=0.130580\n",
      "Iter=102432/epoch=0, Loss=3.504606, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=103072/epoch=0, Loss=3.808476, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=103712/epoch=0, Loss=3.666580, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=104352/epoch=0, Loss=3.743345, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=104992/epoch=0, Loss=3.741717, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=105632/epoch=0, Loss=3.566626, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=106272/epoch=0, Loss=3.781474, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=106912/epoch=0, Loss=3.579478, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=107552/epoch=0, Loss=3.562063, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=108192/epoch=0, Loss=3.243082, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=108800/epoch=0, Validation Accuracy=0.126116\n",
      "Iter=108832/epoch=0, Loss=3.893528, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=109472/epoch=0, Loss=3.653457, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=110112/epoch=0, Loss=3.872071, Training Accuracy=0.000000, lr=0.001000\n",
      "Iter=110752/epoch=0, Loss=3.989881, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=111392/epoch=0, Loss=3.761811, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=112032/epoch=0, Loss=3.456960, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=112672/epoch=0, Loss=3.508530, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=113312/epoch=0, Loss=3.759619, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=113952/epoch=0, Loss=3.514550, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=114592/epoch=0, Loss=3.555320, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=115200/epoch=0, Validation Accuracy=0.116071\n",
      "Iter=115232/epoch=0, Loss=3.205969, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=115872/epoch=0, Loss=3.492421, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=116512/epoch=0, Loss=3.484452, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=117152/epoch=0, Loss=3.394533, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=117792/epoch=0, Loss=3.752932, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=118432/epoch=0, Loss=3.619023, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=119072/epoch=0, Loss=3.133247, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=119712/epoch=0, Loss=3.514554, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=120352/epoch=0, Loss=3.619226, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=120992/epoch=0, Loss=4.194405, Training Accuracy=0.031250, lr=0.001000\n",
      "\n",
      "Iter=121600/epoch=0, Validation Accuracy=0.131696\n",
      "Iter=121632/epoch=0, Loss=3.348486, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=122272/epoch=0, Loss=4.077876, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=122912/epoch=0, Loss=3.952117, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=123552/epoch=0, Loss=3.493709, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=124192/epoch=0, Loss=3.418821, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=124832/epoch=0, Loss=3.517826, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=125472/epoch=0, Loss=3.725889, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=126112/epoch=0, Loss=3.240968, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=126752/epoch=0, Loss=3.611830, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=127392/epoch=1, Loss=3.922557, Training Accuracy=0.062500, lr=0.001000\n",
      "\n",
      "Iter=128000/epoch=1, Validation Accuracy=0.125000\n",
      "Model saved in file at iteration 128032: ./tf_densenet_model_iter4001.ckpt\n",
      "Iter=128032/epoch=1, Loss=3.386881, Training Accuracy=0.187500, lr=0.001000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=128672/epoch=1, Loss=3.258851, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=129312/epoch=1, Loss=3.434563, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=129952/epoch=1, Loss=3.269869, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=130592/epoch=1, Loss=3.854667, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=131232/epoch=1, Loss=3.422405, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=131872/epoch=1, Loss=3.471973, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=132512/epoch=1, Loss=3.101473, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=133152/epoch=1, Loss=3.819543, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=133792/epoch=1, Loss=3.627036, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=134400/epoch=1, Validation Accuracy=0.092634\n",
      "Iter=134432/epoch=1, Loss=3.840149, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=135072/epoch=1, Loss=4.063134, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=135712/epoch=1, Loss=3.370374, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=136352/epoch=1, Loss=3.531100, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=136992/epoch=1, Loss=3.471380, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=137632/epoch=1, Loss=3.511551, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=138272/epoch=1, Loss=3.699402, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=138912/epoch=1, Loss=3.924515, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=139552/epoch=1, Loss=3.657443, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=140192/epoch=1, Loss=3.538504, Training Accuracy=0.062500, lr=0.001000\n",
      "\n",
      "Iter=140800/epoch=1, Validation Accuracy=0.117188\n",
      "Iter=140832/epoch=1, Loss=3.785200, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=141472/epoch=1, Loss=3.712661, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=142112/epoch=1, Loss=3.694923, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=142752/epoch=1, Loss=3.581161, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=143392/epoch=1, Loss=3.471185, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=144032/epoch=1, Loss=3.650493, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=144672/epoch=1, Loss=3.162307, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=145312/epoch=1, Loss=3.602658, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=145952/epoch=1, Loss=2.992759, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=146592/epoch=1, Loss=3.314279, Training Accuracy=0.218750, lr=0.001000\n",
      "\n",
      "Iter=147200/epoch=1, Validation Accuracy=0.157366\n",
      "Iter=147232/epoch=1, Loss=3.329196, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=147872/epoch=1, Loss=3.253393, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=148512/epoch=1, Loss=3.527359, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=149152/epoch=1, Loss=3.379435, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=149792/epoch=1, Loss=3.790520, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=150432/epoch=1, Loss=3.683619, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=151072/epoch=1, Loss=3.494073, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=151712/epoch=1, Loss=3.268517, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=152352/epoch=1, Loss=3.109628, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=152992/epoch=1, Loss=3.286792, Training Accuracy=0.218750, lr=0.001000\n",
      "\n",
      "Iter=153600/epoch=1, Validation Accuracy=0.166295\n",
      "Iter=153632/epoch=1, Loss=3.251158, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=154272/epoch=1, Loss=3.525706, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=154912/epoch=1, Loss=3.382718, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=155552/epoch=1, Loss=3.603830, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=156192/epoch=1, Loss=3.184028, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=156832/epoch=1, Loss=2.798703, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=157472/epoch=1, Loss=3.432558, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=158112/epoch=1, Loss=3.700767, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=158752/epoch=1, Loss=3.676165, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=159392/epoch=1, Loss=2.928436, Training Accuracy=0.343750, lr=0.001000\n",
      "\n",
      "Iter=160000/epoch=1, Validation Accuracy=0.171875\n",
      "Model saved in file at iteration 160032: ./tf_densenet_model_iter5001.ckpt\n",
      "Iter=160032/epoch=1, Loss=3.901898, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=160672/epoch=1, Loss=3.656874, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=161312/epoch=1, Loss=3.496585, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=161952/epoch=1, Loss=3.536664, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=162592/epoch=1, Loss=3.488956, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=163232/epoch=1, Loss=3.596457, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=163872/epoch=1, Loss=3.321608, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=164512/epoch=1, Loss=3.428037, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=165152/epoch=1, Loss=3.387696, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=165792/epoch=1, Loss=3.553836, Training Accuracy=0.093750, lr=0.001000\n",
      "\n",
      "Iter=166400/epoch=1, Validation Accuracy=0.172991\n",
      "Iter=166432/epoch=1, Loss=3.903831, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=167072/epoch=1, Loss=3.841862, Training Accuracy=0.031250, lr=0.001000\n",
      "Iter=167712/epoch=1, Loss=3.395181, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=168352/epoch=1, Loss=3.329964, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=168992/epoch=1, Loss=3.299862, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=169632/epoch=1, Loss=3.335734, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=170272/epoch=1, Loss=3.475551, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=170912/epoch=1, Loss=3.284023, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=171552/epoch=1, Loss=3.502463, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=172192/epoch=1, Loss=3.305453, Training Accuracy=0.250000, lr=0.001000\n",
      "\n",
      "Iter=172800/epoch=1, Validation Accuracy=0.164062\n",
      "Iter=172832/epoch=1, Loss=3.023344, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=173472/epoch=1, Loss=3.570679, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=174112/epoch=1, Loss=3.432480, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=174752/epoch=1, Loss=3.530351, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=175392/epoch=1, Loss=2.988152, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=176032/epoch=1, Loss=3.036402, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=176672/epoch=1, Loss=3.207147, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=177312/epoch=1, Loss=3.489642, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=177952/epoch=1, Loss=2.888950, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=178592/epoch=1, Loss=3.530626, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=179200/epoch=1, Validation Accuracy=0.176339\n",
      "Iter=179232/epoch=1, Loss=3.606812, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=179872/epoch=1, Loss=3.201516, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=180512/epoch=1, Loss=3.884178, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=181152/epoch=1, Loss=3.549574, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=181792/epoch=1, Loss=3.895373, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=182432/epoch=1, Loss=3.136438, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=183072/epoch=1, Loss=3.134775, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=183712/epoch=1, Loss=3.233005, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=184352/epoch=1, Loss=3.193731, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=184992/epoch=1, Loss=3.246778, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=185600/epoch=1, Validation Accuracy=0.152902\n",
      "Iter=185632/epoch=1, Loss=2.915269, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=186272/epoch=1, Loss=3.296908, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=186912/epoch=1, Loss=2.681667, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=187552/epoch=1, Loss=2.773536, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=188192/epoch=1, Loss=3.425950, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=188832/epoch=1, Loss=3.438919, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=189472/epoch=1, Loss=3.321185, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=190112/epoch=1, Loss=3.195099, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=190752/epoch=1, Loss=3.001124, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=191392/epoch=1, Loss=3.353971, Training Accuracy=0.187500, lr=0.001000\n",
      "\n",
      "Iter=192000/epoch=1, Validation Accuracy=0.157366\n",
      "WARNING:tensorflow:From C:\\Users\\user\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\tensorflow\\python\\training\\saver.py:966: remove_checkpoint (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use standard file APIs to delete files with this prefix.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model saved in file at iteration 192032: ./tf_densenet_model_iter6001.ckpt\n",
      "Iter=192032/epoch=1, Loss=3.845389, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=192672/epoch=1, Loss=3.081644, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=193312/epoch=1, Loss=3.295479, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=193952/epoch=1, Loss=3.086167, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=194592/epoch=1, Loss=3.245038, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=195232/epoch=1, Loss=3.329552, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=195872/epoch=1, Loss=2.729349, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=196512/epoch=1, Loss=3.110675, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=197152/epoch=1, Loss=3.773790, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=197792/epoch=1, Loss=3.745154, Training Accuracy=0.156250, lr=0.001000\n",
      "\n",
      "Iter=198400/epoch=1, Validation Accuracy=0.181920\n",
      "Iter=198432/epoch=1, Loss=3.074841, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=199072/epoch=1, Loss=3.189715, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=199712/epoch=1, Loss=3.259954, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=200352/epoch=1, Loss=3.099657, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=200992/epoch=1, Loss=3.196986, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=201632/epoch=1, Loss=3.163409, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=202272/epoch=1, Loss=3.255792, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=202912/epoch=1, Loss=2.531886, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=203552/epoch=1, Loss=2.944850, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=204192/epoch=1, Loss=3.197668, Training Accuracy=0.156250, lr=0.001000\n",
      "\n",
      "Iter=204800/epoch=1, Validation Accuracy=0.195312\n",
      "Iter=204832/epoch=1, Loss=3.426024, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=205472/epoch=1, Loss=3.236972, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=206112/epoch=1, Loss=3.198573, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=206752/epoch=1, Loss=3.157194, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=207392/epoch=1, Loss=3.310381, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=208032/epoch=1, Loss=2.990987, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=208672/epoch=1, Loss=2.637287, Training Accuracy=0.437500, lr=0.001000\n",
      "Iter=209312/epoch=1, Loss=3.369094, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=209952/epoch=1, Loss=3.423112, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=210592/epoch=1, Loss=3.017698, Training Accuracy=0.187500, lr=0.001000\n",
      "\n",
      "Iter=211200/epoch=1, Validation Accuracy=0.214286\n",
      "Iter=211232/epoch=1, Loss=3.309091, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=211872/epoch=1, Loss=4.178893, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=212512/epoch=1, Loss=3.114468, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=213152/epoch=1, Loss=3.039879, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=213792/epoch=1, Loss=3.835371, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=214432/epoch=1, Loss=2.669108, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=215072/epoch=1, Loss=3.303701, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=215712/epoch=1, Loss=3.366412, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=216352/epoch=1, Loss=3.663122, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=216992/epoch=1, Loss=3.368467, Training Accuracy=0.218750, lr=0.001000\n",
      "\n",
      "Iter=217600/epoch=1, Validation Accuracy=0.169643\n",
      "Iter=217632/epoch=1, Loss=2.804214, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=218272/epoch=1, Loss=3.364888, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=218912/epoch=1, Loss=3.432431, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=219552/epoch=1, Loss=3.279835, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=220192/epoch=1, Loss=2.867708, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=220832/epoch=1, Loss=3.142808, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=221472/epoch=1, Loss=3.071920, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=222112/epoch=1, Loss=3.230754, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=222752/epoch=1, Loss=3.105862, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=223392/epoch=1, Loss=2.906186, Training Accuracy=0.218750, lr=0.001000\n",
      "\n",
      "Iter=224000/epoch=1, Validation Accuracy=0.205357\n",
      "Model saved in file at iteration 224032: ./tf_densenet_model_iter7001.ckpt\n",
      "Iter=224032/epoch=1, Loss=3.309080, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=224672/epoch=1, Loss=3.400197, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=225312/epoch=1, Loss=3.111405, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=225952/epoch=1, Loss=3.765427, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=226592/epoch=1, Loss=3.321681, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=227232/epoch=1, Loss=3.030101, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=227872/epoch=1, Loss=3.148149, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=228512/epoch=1, Loss=3.012776, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=229152/epoch=1, Loss=2.764656, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=229792/epoch=1, Loss=3.320062, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=230400/epoch=1, Validation Accuracy=0.161830\n",
      "Iter=230432/epoch=1, Loss=3.707209, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=231072/epoch=1, Loss=3.182227, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=231712/epoch=1, Loss=2.749638, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=232352/epoch=1, Loss=2.822049, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=232992/epoch=1, Loss=3.355917, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=233632/epoch=1, Loss=3.449802, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=234272/epoch=1, Loss=2.939870, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=234912/epoch=1, Loss=3.102799, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=235552/epoch=1, Loss=2.816547, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=236192/epoch=1, Loss=3.082721, Training Accuracy=0.187500, lr=0.001000\n",
      "\n",
      "Iter=236800/epoch=1, Validation Accuracy=0.241071\n",
      "Iter=236832/epoch=1, Loss=3.149577, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=237472/epoch=1, Loss=3.267509, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=238112/epoch=1, Loss=2.784268, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=238752/epoch=1, Loss=3.642448, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=239392/epoch=1, Loss=2.707600, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=240032/epoch=1, Loss=2.919468, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=240672/epoch=1, Loss=3.188915, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=241312/epoch=1, Loss=3.209289, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=241952/epoch=1, Loss=3.192889, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=242592/epoch=1, Loss=3.221736, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=243200/epoch=1, Validation Accuracy=0.187500\n",
      "Iter=243232/epoch=1, Loss=3.213493, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=243872/epoch=1, Loss=2.790478, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=244512/epoch=1, Loss=3.113554, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=245152/epoch=1, Loss=3.232084, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=245792/epoch=1, Loss=3.070524, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=246432/epoch=1, Loss=3.164067, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=247072/epoch=1, Loss=3.319020, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=247712/epoch=1, Loss=3.036296, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=248352/epoch=1, Loss=3.145922, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=248992/epoch=1, Loss=3.384703, Training Accuracy=0.187500, lr=0.001000\n",
      "\n",
      "Iter=249600/epoch=1, Validation Accuracy=0.260045\n",
      "Iter=249632/epoch=1, Loss=3.124356, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=250272/epoch=1, Loss=2.784527, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=250912/epoch=1, Loss=3.425006, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=251552/epoch=1, Loss=3.408993, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=252192/epoch=1, Loss=2.596301, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=252832/epoch=1, Loss=2.920010, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=253472/epoch=1, Loss=3.070472, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=254112/epoch=1, Loss=2.910526, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=254752/epoch=2, Loss=2.890491, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=255392/epoch=2, Loss=2.891498, Training Accuracy=0.218750, lr=0.001000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=256000/epoch=2, Validation Accuracy=0.239955\n",
      "Model saved in file at iteration 256032: ./tf_densenet_model_iter8001.ckpt\n",
      "Iter=256032/epoch=2, Loss=2.712805, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=256672/epoch=2, Loss=3.619224, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=257312/epoch=2, Loss=3.186782, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=257952/epoch=2, Loss=3.537404, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=258592/epoch=2, Loss=2.809527, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=259232/epoch=2, Loss=3.210867, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=259872/epoch=2, Loss=3.202291, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=260512/epoch=2, Loss=2.646619, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=261152/epoch=2, Loss=3.260927, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=261792/epoch=2, Loss=3.285800, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=262400/epoch=2, Validation Accuracy=0.237723\n",
      "Iter=262432/epoch=2, Loss=3.209051, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=263072/epoch=2, Loss=3.100965, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=263712/epoch=2, Loss=2.570869, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=264352/epoch=2, Loss=3.173733, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=264992/epoch=2, Loss=3.401520, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=265632/epoch=2, Loss=2.892606, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=266272/epoch=2, Loss=2.843310, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=266912/epoch=2, Loss=3.444857, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=267552/epoch=2, Loss=2.799148, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=268192/epoch=2, Loss=2.643621, Training Accuracy=0.187500, lr=0.001000\n",
      "\n",
      "Iter=268800/epoch=2, Validation Accuracy=0.204241\n",
      "Iter=268832/epoch=2, Loss=2.798560, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=269472/epoch=2, Loss=2.748488, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=270112/epoch=2, Loss=3.117337, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=270752/epoch=2, Loss=3.200663, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=271392/epoch=2, Loss=2.761144, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=272032/epoch=2, Loss=3.012728, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=272672/epoch=2, Loss=3.034260, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=273312/epoch=2, Loss=3.171494, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=273952/epoch=2, Loss=2.826682, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=274592/epoch=2, Loss=3.196479, Training Accuracy=0.125000, lr=0.001000\n",
      "\n",
      "Iter=275200/epoch=2, Validation Accuracy=0.214286\n",
      "Iter=275232/epoch=2, Loss=2.740762, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=275872/epoch=2, Loss=2.700983, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=276512/epoch=2, Loss=3.206404, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=277152/epoch=2, Loss=3.176987, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=277792/epoch=2, Loss=3.008806, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=278432/epoch=2, Loss=2.554902, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=279072/epoch=2, Loss=3.260259, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=279712/epoch=2, Loss=3.000967, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=280352/epoch=2, Loss=2.918662, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=280992/epoch=2, Loss=3.262589, Training Accuracy=0.187500, lr=0.001000\n",
      "\n",
      "Iter=281600/epoch=2, Validation Accuracy=0.236607\n",
      "Iter=281632/epoch=2, Loss=3.241615, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=282272/epoch=2, Loss=2.925075, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=282912/epoch=2, Loss=3.155075, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=283552/epoch=2, Loss=2.994397, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=284192/epoch=2, Loss=3.055983, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=284832/epoch=2, Loss=2.918343, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=285472/epoch=2, Loss=2.948720, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=286112/epoch=2, Loss=2.828590, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=286752/epoch=2, Loss=3.094099, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=287392/epoch=2, Loss=2.470838, Training Accuracy=0.375000, lr=0.001000\n",
      "\n",
      "Iter=288000/epoch=2, Validation Accuracy=0.217634\n",
      "Model saved in file at iteration 288032: ./tf_densenet_model_iter9001.ckpt\n",
      "Iter=288032/epoch=2, Loss=3.231261, Training Accuracy=0.062500, lr=0.001000\n",
      "Iter=288672/epoch=2, Loss=2.837618, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=289312/epoch=2, Loss=2.991821, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=289952/epoch=2, Loss=2.648443, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=290592/epoch=2, Loss=3.075876, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=291232/epoch=2, Loss=2.959150, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=291872/epoch=2, Loss=2.976560, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=292512/epoch=2, Loss=2.603429, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=293152/epoch=2, Loss=2.939975, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=293792/epoch=2, Loss=2.943151, Training Accuracy=0.218750, lr=0.001000\n",
      "\n",
      "Iter=294400/epoch=2, Validation Accuracy=0.268973\n",
      "Iter=294432/epoch=2, Loss=2.650672, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=295072/epoch=2, Loss=2.560809, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=295712/epoch=2, Loss=2.768432, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=296352/epoch=2, Loss=2.933746, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=296992/epoch=2, Loss=3.244787, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=297632/epoch=2, Loss=2.733278, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=298272/epoch=2, Loss=2.736500, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=298912/epoch=2, Loss=2.586385, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=299552/epoch=2, Loss=2.892445, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=300192/epoch=2, Loss=2.783143, Training Accuracy=0.250000, lr=0.001000\n",
      "\n",
      "Iter=300800/epoch=2, Validation Accuracy=0.268973\n",
      "Iter=300832/epoch=2, Loss=2.472568, Training Accuracy=0.468750, lr=0.001000\n",
      "Iter=301472/epoch=2, Loss=3.117623, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=302112/epoch=2, Loss=2.550237, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=302752/epoch=2, Loss=2.826977, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=303392/epoch=2, Loss=2.933370, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=304032/epoch=2, Loss=2.878418, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=304672/epoch=2, Loss=3.439523, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=305312/epoch=2, Loss=2.880692, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=305952/epoch=2, Loss=2.821931, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=306592/epoch=2, Loss=2.738441, Training Accuracy=0.312500, lr=0.001000\n",
      "\n",
      "Iter=307200/epoch=2, Validation Accuracy=0.255580\n",
      "Iter=307232/epoch=2, Loss=3.303233, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=307872/epoch=2, Loss=2.853981, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=308512/epoch=2, Loss=2.918653, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=309152/epoch=2, Loss=2.459334, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=309792/epoch=2, Loss=3.014177, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=310432/epoch=2, Loss=2.834737, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=311072/epoch=2, Loss=2.498058, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=311712/epoch=2, Loss=2.773469, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=312352/epoch=2, Loss=2.597807, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=312992/epoch=2, Loss=2.487081, Training Accuracy=0.281250, lr=0.001000\n",
      "\n",
      "Iter=313600/epoch=2, Validation Accuracy=0.293527\n",
      "Iter=313632/epoch=2, Loss=2.642715, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=314272/epoch=2, Loss=2.792479, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=314912/epoch=2, Loss=3.084155, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=315552/epoch=2, Loss=2.840651, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=316192/epoch=2, Loss=2.959316, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=316832/epoch=2, Loss=2.856776, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=317472/epoch=2, Loss=2.890734, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=318112/epoch=2, Loss=3.121591, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=318752/epoch=2, Loss=2.792088, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=319392/epoch=2, Loss=3.083352, Training Accuracy=0.250000, lr=0.001000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=320000/epoch=2, Validation Accuracy=0.251116\n",
      "Model saved in file at iteration 320032: ./tf_densenet_model_iter10001.ckpt\n",
      "Iter=320032/epoch=2, Loss=2.468014, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=320672/epoch=2, Loss=2.552375, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=321312/epoch=2, Loss=3.003588, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=321952/epoch=2, Loss=3.287302, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=322592/epoch=2, Loss=2.502260, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=323232/epoch=2, Loss=3.108982, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=323872/epoch=2, Loss=2.828356, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=324512/epoch=2, Loss=3.010932, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=325152/epoch=2, Loss=2.621475, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=325792/epoch=2, Loss=2.818868, Training Accuracy=0.250000, lr=0.001000\n",
      "\n",
      "Iter=326400/epoch=2, Validation Accuracy=0.220982\n",
      "Iter=326432/epoch=2, Loss=2.763047, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=327072/epoch=2, Loss=2.417635, Training Accuracy=0.468750, lr=0.001000\n",
      "Iter=327712/epoch=2, Loss=2.732744, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=328352/epoch=2, Loss=2.490680, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=328992/epoch=2, Loss=2.972492, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=329632/epoch=2, Loss=3.199054, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=330272/epoch=2, Loss=2.685219, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=330912/epoch=2, Loss=2.671717, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=331552/epoch=2, Loss=2.572359, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=332192/epoch=2, Loss=3.000678, Training Accuracy=0.218750, lr=0.001000\n",
      "\n",
      "Iter=332800/epoch=2, Validation Accuracy=0.255580\n",
      "Iter=332832/epoch=2, Loss=3.072679, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=333472/epoch=2, Loss=2.675075, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=334112/epoch=2, Loss=2.764626, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=334752/epoch=2, Loss=2.935227, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=335392/epoch=2, Loss=2.689819, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=336032/epoch=2, Loss=3.024802, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=336672/epoch=2, Loss=3.170734, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=337312/epoch=2, Loss=2.944457, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=337952/epoch=2, Loss=3.033484, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=338592/epoch=2, Loss=2.775870, Training Accuracy=0.250000, lr=0.001000\n",
      "\n",
      "Iter=339200/epoch=2, Validation Accuracy=0.280134\n",
      "Iter=339232/epoch=2, Loss=2.903728, Training Accuracy=0.093750, lr=0.001000\n",
      "Iter=339872/epoch=2, Loss=2.913396, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=340512/epoch=2, Loss=3.281185, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=341152/epoch=2, Loss=2.580835, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=341792/epoch=2, Loss=3.207012, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=342432/epoch=2, Loss=3.062247, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=343072/epoch=2, Loss=2.889205, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=343712/epoch=2, Loss=2.316994, Training Accuracy=0.468750, lr=0.001000\n",
      "Iter=344352/epoch=2, Loss=2.566151, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=344992/epoch=2, Loss=2.657679, Training Accuracy=0.312500, lr=0.001000\n",
      "\n",
      "Iter=345600/epoch=2, Validation Accuracy=0.294643\n",
      "Iter=345632/epoch=2, Loss=3.036753, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=346272/epoch=2, Loss=2.720121, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=346912/epoch=2, Loss=2.753469, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=347552/epoch=2, Loss=2.440527, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=348192/epoch=2, Loss=3.016146, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=348832/epoch=2, Loss=2.903301, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=349472/epoch=2, Loss=2.877284, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=350112/epoch=2, Loss=2.919276, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=350752/epoch=2, Loss=2.412592, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=351392/epoch=2, Loss=2.776072, Training Accuracy=0.281250, lr=0.001000\n",
      "\n",
      "Iter=352000/epoch=2, Validation Accuracy=0.293527\n",
      "Model saved in file at iteration 352032: ./tf_densenet_model_iter11001.ckpt\n",
      "Iter=352032/epoch=2, Loss=2.831048, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=352672/epoch=2, Loss=2.683491, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=353312/epoch=2, Loss=2.657282, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=353952/epoch=2, Loss=2.719460, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=354592/epoch=2, Loss=2.574986, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=355232/epoch=2, Loss=3.021316, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=355872/epoch=2, Loss=3.326837, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=356512/epoch=2, Loss=2.510460, Training Accuracy=0.437500, lr=0.001000\n",
      "Iter=357152/epoch=2, Loss=3.039958, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=357792/epoch=2, Loss=2.401360, Training Accuracy=0.375000, lr=0.001000\n",
      "\n",
      "Iter=358400/epoch=2, Validation Accuracy=0.261161\n",
      "Iter=358432/epoch=2, Loss=2.672437, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=359072/epoch=2, Loss=2.651940, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=359712/epoch=2, Loss=2.373332, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=360352/epoch=2, Loss=2.620598, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=360992/epoch=2, Loss=2.922402, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=361632/epoch=2, Loss=3.465619, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=362272/epoch=2, Loss=3.184386, Training Accuracy=0.156250, lr=0.001000\n",
      "Iter=362912/epoch=2, Loss=2.822191, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=363552/epoch=2, Loss=2.958672, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=364192/epoch=2, Loss=2.684108, Training Accuracy=0.250000, lr=0.001000\n",
      "\n",
      "Iter=364800/epoch=2, Validation Accuracy=0.283482\n",
      "Iter=364832/epoch=2, Loss=2.734939, Training Accuracy=0.437500, lr=0.001000\n",
      "Iter=365472/epoch=2, Loss=2.620491, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=366112/epoch=2, Loss=2.811651, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=366752/epoch=2, Loss=2.940875, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=367392/epoch=2, Loss=2.815888, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=368032/epoch=2, Loss=2.580204, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=368672/epoch=2, Loss=2.689505, Training Accuracy=0.437500, lr=0.001000\n",
      "Iter=369312/epoch=2, Loss=2.834620, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=369952/epoch=2, Loss=2.850165, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=370592/epoch=2, Loss=2.549710, Training Accuracy=0.312500, lr=0.001000\n",
      "\n",
      "Iter=371200/epoch=2, Validation Accuracy=0.294643\n",
      "Iter=371232/epoch=2, Loss=2.430827, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=371872/epoch=2, Loss=2.696929, Training Accuracy=0.281250, lr=0.001000\n",
      "Iter=372512/epoch=2, Loss=2.669277, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=373152/epoch=2, Loss=2.657809, Training Accuracy=0.437500, lr=0.001000\n",
      "Iter=373792/epoch=2, Loss=3.022964, Training Accuracy=0.187500, lr=0.001000\n",
      "Iter=374432/epoch=2, Loss=2.480549, Training Accuracy=0.312500, lr=0.001000\n",
      "Iter=375072/epoch=2, Loss=2.891202, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=375712/epoch=2, Loss=2.445305, Training Accuracy=0.406250, lr=0.001000\n",
      "Iter=376352/epoch=2, Loss=3.235053, Training Accuracy=0.250000, lr=0.001000\n",
      "Iter=376992/epoch=2, Loss=2.763513, Training Accuracy=0.250000, lr=0.001000\n",
      "\n",
      "Iter=377600/epoch=2, Validation Accuracy=0.285714\n",
      "Iter=377632/epoch=2, Loss=2.698906, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=378272/epoch=2, Loss=2.871022, Training Accuracy=0.125000, lr=0.001000\n",
      "Iter=378912/epoch=2, Loss=2.695306, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=379552/epoch=2, Loss=2.615795, Training Accuracy=0.375000, lr=0.001000\n",
      "Iter=380192/epoch=2, Loss=2.484511, Training Accuracy=0.343750, lr=0.001000\n",
      "Iter=380832/epoch=2, Loss=2.264219, Training Accuracy=0.468750, lr=0.001000\n",
      "Iter=381472/epoch=2, Loss=3.036209, Training Accuracy=0.218750, lr=0.001000\n",
      "Iter=382112/epoch=3, Loss=2.732061, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=382752/epoch=3, Loss=2.679893, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=383392/epoch=3, Loss=3.150170, Training Accuracy=0.250000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=384000/epoch=3, Validation Accuracy=0.332589\n",
      "Model saved in file at iteration 384032: ./tf_densenet_model_iter12001.ckpt\n",
      "Iter=384032/epoch=3, Loss=3.240562, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=384672/epoch=3, Loss=3.094224, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=385312/epoch=3, Loss=2.781794, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=385952/epoch=3, Loss=2.718500, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=386592/epoch=3, Loss=2.690510, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=387232/epoch=3, Loss=3.076213, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=387872/epoch=3, Loss=3.273485, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=388512/epoch=3, Loss=2.025408, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=389152/epoch=3, Loss=3.193532, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=389792/epoch=3, Loss=2.414224, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=390400/epoch=3, Validation Accuracy=0.335938\n",
      "Iter=390432/epoch=3, Loss=2.353145, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=391072/epoch=3, Loss=2.166430, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=391712/epoch=3, Loss=3.232145, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=392352/epoch=3, Loss=2.527706, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=392992/epoch=3, Loss=2.376145, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=393632/epoch=3, Loss=3.082211, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=394272/epoch=3, Loss=2.583871, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=394912/epoch=3, Loss=2.401156, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=395552/epoch=3, Loss=2.439414, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=396192/epoch=3, Loss=2.536188, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=396800/epoch=3, Validation Accuracy=0.350446\n",
      "Iter=396832/epoch=3, Loss=2.719761, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=397472/epoch=3, Loss=2.518339, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=398112/epoch=3, Loss=2.496928, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=398752/epoch=3, Loss=2.445545, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=399392/epoch=3, Loss=2.675471, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=400032/epoch=3, Loss=2.579286, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=400672/epoch=3, Loss=2.630961, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=401312/epoch=3, Loss=2.351579, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=401952/epoch=3, Loss=1.992301, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=402592/epoch=3, Loss=2.746652, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=403200/epoch=3, Validation Accuracy=0.359375\n",
      "Iter=403232/epoch=3, Loss=2.448929, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=403872/epoch=3, Loss=2.170345, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=404512/epoch=3, Loss=2.270520, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=405152/epoch=3, Loss=2.618832, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=405792/epoch=3, Loss=2.591852, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=406432/epoch=3, Loss=2.350605, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=407072/epoch=3, Loss=2.280569, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=407712/epoch=3, Loss=2.896200, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=408352/epoch=3, Loss=2.412775, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=408992/epoch=3, Loss=2.321846, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=409600/epoch=3, Validation Accuracy=0.366071\n",
      "Iter=409632/epoch=3, Loss=2.655217, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=410272/epoch=3, Loss=2.478267, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=410912/epoch=3, Loss=2.173938, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=411552/epoch=3, Loss=2.541460, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=412192/epoch=3, Loss=2.185504, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=412832/epoch=3, Loss=2.398122, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=413472/epoch=3, Loss=2.331795, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=414112/epoch=3, Loss=2.277320, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=414752/epoch=3, Loss=3.007721, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=415392/epoch=3, Loss=2.871077, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=416000/epoch=3, Validation Accuracy=0.361607\n",
      "Model saved in file at iteration 416032: ./tf_densenet_model_iter13001.ckpt\n",
      "Iter=416032/epoch=3, Loss=2.270290, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=416672/epoch=3, Loss=2.235392, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=417312/epoch=3, Loss=2.931277, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=417952/epoch=3, Loss=2.237205, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=418592/epoch=3, Loss=2.738974, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=419232/epoch=3, Loss=3.019298, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=419872/epoch=3, Loss=2.773190, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=420512/epoch=3, Loss=2.554896, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=421152/epoch=3, Loss=2.797178, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=421792/epoch=3, Loss=2.535848, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=422400/epoch=3, Validation Accuracy=0.367188\n",
      "Iter=422432/epoch=3, Loss=2.299597, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=423072/epoch=3, Loss=2.874764, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=423712/epoch=3, Loss=2.306525, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=424352/epoch=3, Loss=2.539982, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=424992/epoch=3, Loss=2.167048, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=425632/epoch=3, Loss=2.298765, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=426272/epoch=3, Loss=2.608562, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=426912/epoch=3, Loss=2.894596, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=427552/epoch=3, Loss=2.292707, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=428192/epoch=3, Loss=2.184864, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=428800/epoch=3, Validation Accuracy=0.367188\n",
      "Iter=428832/epoch=3, Loss=2.871123, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=429472/epoch=3, Loss=2.605392, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=430112/epoch=3, Loss=2.611673, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=430752/epoch=3, Loss=2.362236, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=431392/epoch=3, Loss=2.425467, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=432032/epoch=3, Loss=2.622193, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=432672/epoch=3, Loss=2.193992, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=433312/epoch=3, Loss=2.879449, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=433952/epoch=3, Loss=2.755300, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=434592/epoch=3, Loss=2.191644, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=435200/epoch=3, Validation Accuracy=0.353795\n",
      "Iter=435232/epoch=3, Loss=3.082272, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=435872/epoch=3, Loss=2.559917, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=436512/epoch=3, Loss=2.424799, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=437152/epoch=3, Loss=3.223350, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=437792/epoch=3, Loss=2.983816, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=438432/epoch=3, Loss=2.645320, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=439072/epoch=3, Loss=2.680043, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=439712/epoch=3, Loss=2.144868, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=440352/epoch=3, Loss=2.009805, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=440992/epoch=3, Loss=3.197638, Training Accuracy=0.125000, lr=0.000100\n",
      "\n",
      "Iter=441600/epoch=3, Validation Accuracy=0.359375\n",
      "Iter=441632/epoch=3, Loss=2.188564, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=442272/epoch=3, Loss=2.772827, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=442912/epoch=3, Loss=2.397021, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=443552/epoch=3, Loss=2.635271, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=444192/epoch=3, Loss=2.323504, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=444832/epoch=3, Loss=2.457520, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=445472/epoch=3, Loss=2.607131, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=446112/epoch=3, Loss=2.458658, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=446752/epoch=3, Loss=2.483151, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=447392/epoch=3, Loss=2.371247, Training Accuracy=0.375000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=448000/epoch=3, Validation Accuracy=0.368304\n",
      "Model saved in file at iteration 448032: ./tf_densenet_model_iter14001.ckpt\n",
      "Iter=448032/epoch=3, Loss=2.642589, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=448672/epoch=3, Loss=2.545541, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=449312/epoch=3, Loss=3.060283, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=449952/epoch=3, Loss=2.134020, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=450592/epoch=3, Loss=2.014796, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=451232/epoch=3, Loss=2.535649, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=451872/epoch=3, Loss=2.192397, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=452512/epoch=3, Loss=2.533953, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=453152/epoch=3, Loss=2.373854, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=453792/epoch=3, Loss=2.569548, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=454400/epoch=3, Validation Accuracy=0.370536\n",
      "Iter=454432/epoch=3, Loss=2.974117, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=455072/epoch=3, Loss=2.361479, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=455712/epoch=3, Loss=2.326222, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=456352/epoch=3, Loss=2.344018, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=456992/epoch=3, Loss=2.179005, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=457632/epoch=3, Loss=2.598984, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=458272/epoch=3, Loss=2.457473, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=458912/epoch=3, Loss=2.777373, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=459552/epoch=3, Loss=2.714305, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=460192/epoch=3, Loss=2.572522, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=460800/epoch=3, Validation Accuracy=0.369420\n",
      "Iter=460832/epoch=3, Loss=2.276255, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=461472/epoch=3, Loss=2.473701, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=462112/epoch=3, Loss=2.776007, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=462752/epoch=3, Loss=2.767525, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=463392/epoch=3, Loss=3.075352, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=464032/epoch=3, Loss=2.695993, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=464672/epoch=3, Loss=2.270859, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=465312/epoch=3, Loss=2.624127, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=465952/epoch=3, Loss=2.215516, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=466592/epoch=3, Loss=2.603240, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=467200/epoch=3, Validation Accuracy=0.373884\n",
      "Iter=467232/epoch=3, Loss=2.188007, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=467872/epoch=3, Loss=2.412287, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=468512/epoch=3, Loss=2.578598, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=469152/epoch=3, Loss=2.289297, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=469792/epoch=3, Loss=2.146995, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=470432/epoch=3, Loss=2.280329, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=471072/epoch=3, Loss=2.355457, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=471712/epoch=3, Loss=2.620205, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=472352/epoch=3, Loss=2.072971, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=472992/epoch=3, Loss=2.242880, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=473600/epoch=3, Validation Accuracy=0.369420\n",
      "Iter=473632/epoch=3, Loss=3.031860, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=474272/epoch=3, Loss=2.320809, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=474912/epoch=3, Loss=2.403758, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=475552/epoch=3, Loss=2.350588, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=476192/epoch=3, Loss=2.704400, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=476832/epoch=3, Loss=2.934514, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=477472/epoch=3, Loss=2.305274, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=478112/epoch=3, Loss=2.969754, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=478752/epoch=3, Loss=2.561551, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=479392/epoch=3, Loss=2.054063, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=480000/epoch=3, Validation Accuracy=0.366071\n",
      "Model saved in file at iteration 480032: ./tf_densenet_model_iter15001.ckpt\n",
      "Iter=480032/epoch=3, Loss=2.435621, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=480672/epoch=3, Loss=2.418766, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=481312/epoch=3, Loss=2.426343, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=481952/epoch=3, Loss=2.380073, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=482592/epoch=3, Loss=2.593779, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=483232/epoch=3, Loss=2.345564, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=483872/epoch=3, Loss=2.791337, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=484512/epoch=3, Loss=2.622301, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=485152/epoch=3, Loss=2.581647, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=485792/epoch=3, Loss=2.746125, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=486400/epoch=3, Validation Accuracy=0.372768\n",
      "Iter=486432/epoch=3, Loss=2.189088, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=487072/epoch=3, Loss=2.894537, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=487712/epoch=3, Loss=2.159830, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=488352/epoch=3, Loss=3.028611, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=488992/epoch=3, Loss=1.832070, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=489632/epoch=3, Loss=2.965736, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=490272/epoch=3, Loss=3.108195, Training Accuracy=0.125000, lr=0.000100\n",
      "Iter=490912/epoch=3, Loss=2.392298, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=491552/epoch=3, Loss=2.623605, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=492192/epoch=3, Loss=2.615482, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=492800/epoch=3, Validation Accuracy=0.378348\n",
      "Iter=492832/epoch=3, Loss=2.512810, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=493472/epoch=3, Loss=2.343602, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=494112/epoch=3, Loss=1.950940, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=494752/epoch=3, Loss=2.118573, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=495392/epoch=3, Loss=2.358985, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=496032/epoch=3, Loss=2.366083, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=496672/epoch=3, Loss=2.028632, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=497312/epoch=3, Loss=2.359130, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=497952/epoch=3, Loss=2.410834, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=498592/epoch=3, Loss=2.704103, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=499200/epoch=3, Validation Accuracy=0.388393\n",
      "Iter=499232/epoch=3, Loss=2.810932, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=499872/epoch=3, Loss=1.977881, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=500512/epoch=3, Loss=2.729884, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=501152/epoch=3, Loss=2.479000, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=501792/epoch=3, Loss=2.571560, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=502432/epoch=3, Loss=2.448356, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=503072/epoch=3, Loss=2.696967, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=503712/epoch=3, Loss=2.072639, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=504352/epoch=3, Loss=2.090761, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=504992/epoch=3, Loss=2.354886, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=505600/epoch=3, Validation Accuracy=0.378348\n",
      "Iter=505632/epoch=3, Loss=2.585581, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=506272/epoch=3, Loss=2.177588, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=506912/epoch=3, Loss=2.571385, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=507552/epoch=3, Loss=2.210031, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=508192/epoch=3, Loss=2.789227, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=508832/epoch=3, Loss=2.719957, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=509472/epoch=4, Loss=3.325101, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=510112/epoch=4, Loss=2.350459, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=510752/epoch=4, Loss=2.486197, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=511392/epoch=4, Loss=2.330370, Training Accuracy=0.375000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=512000/epoch=4, Validation Accuracy=0.362723\n",
      "Model saved in file at iteration 512032: ./tf_densenet_model_iter16001.ckpt\n",
      "Iter=512032/epoch=4, Loss=2.650642, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=512672/epoch=4, Loss=2.140047, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=513312/epoch=4, Loss=2.604771, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=513952/epoch=4, Loss=1.972137, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=514592/epoch=4, Loss=3.159562, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=515232/epoch=4, Loss=3.386153, Training Accuracy=0.125000, lr=0.000100\n",
      "Iter=515872/epoch=4, Loss=2.640430, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=516512/epoch=4, Loss=2.593185, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=517152/epoch=4, Loss=1.882890, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=517792/epoch=4, Loss=2.175430, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=518400/epoch=4, Validation Accuracy=0.371652\n",
      "Iter=518432/epoch=4, Loss=2.414929, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=519072/epoch=4, Loss=2.296572, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=519712/epoch=4, Loss=2.616544, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=520352/epoch=4, Loss=2.308474, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=520992/epoch=4, Loss=1.920268, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=521632/epoch=4, Loss=1.954757, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=522272/epoch=4, Loss=2.426271, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=522912/epoch=4, Loss=2.263801, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=523552/epoch=4, Loss=2.498245, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=524192/epoch=4, Loss=2.273616, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=524800/epoch=4, Validation Accuracy=0.373884\n",
      "Iter=524832/epoch=4, Loss=2.704600, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=525472/epoch=4, Loss=2.438611, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=526112/epoch=4, Loss=1.907553, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=526752/epoch=4, Loss=2.607484, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=527392/epoch=4, Loss=2.514813, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=528032/epoch=4, Loss=2.555714, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=528672/epoch=4, Loss=2.300801, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=529312/epoch=4, Loss=2.583167, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=529952/epoch=4, Loss=2.761115, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=530592/epoch=4, Loss=2.499114, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=531200/epoch=4, Validation Accuracy=0.377232\n",
      "Iter=531232/epoch=4, Loss=1.809929, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=531872/epoch=4, Loss=3.101861, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=532512/epoch=4, Loss=2.706950, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=533152/epoch=4, Loss=2.749449, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=533792/epoch=4, Loss=2.823308, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=534432/epoch=4, Loss=2.532061, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=535072/epoch=4, Loss=2.339578, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=535712/epoch=4, Loss=2.616215, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=536352/epoch=4, Loss=2.464333, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=536992/epoch=4, Loss=2.333618, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=537600/epoch=4, Validation Accuracy=0.375000\n",
      "Iter=537632/epoch=4, Loss=2.324922, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=538272/epoch=4, Loss=2.186890, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=538912/epoch=4, Loss=2.751983, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=539552/epoch=4, Loss=2.341264, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=540192/epoch=4, Loss=1.845186, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=540832/epoch=4, Loss=1.918831, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=541472/epoch=4, Loss=2.519985, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=542112/epoch=4, Loss=2.546512, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=542752/epoch=4, Loss=2.183450, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=543392/epoch=4, Loss=2.979004, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=544000/epoch=4, Validation Accuracy=0.375000\n",
      "Model saved in file at iteration 544032: ./tf_densenet_model_iter17001.ckpt\n",
      "Iter=544032/epoch=4, Loss=2.760445, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=544672/epoch=4, Loss=2.664366, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=545312/epoch=4, Loss=1.597552, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=545952/epoch=4, Loss=2.143924, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=546592/epoch=4, Loss=2.668156, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=547232/epoch=4, Loss=2.212775, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=547872/epoch=4, Loss=2.588582, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=548512/epoch=4, Loss=2.627794, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=549152/epoch=4, Loss=3.052268, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=549792/epoch=4, Loss=2.441299, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=550400/epoch=4, Validation Accuracy=0.386161\n",
      "Iter=550432/epoch=4, Loss=2.449933, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=551072/epoch=4, Loss=2.521532, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=551712/epoch=4, Loss=2.762793, Training Accuracy=0.125000, lr=0.000100\n",
      "Iter=552352/epoch=4, Loss=2.821656, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=552992/epoch=4, Loss=2.308981, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=553632/epoch=4, Loss=2.784488, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=554272/epoch=4, Loss=2.710343, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=554912/epoch=4, Loss=2.334604, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=555552/epoch=4, Loss=2.738605, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=556192/epoch=4, Loss=3.167056, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=556800/epoch=4, Validation Accuracy=0.367188\n",
      "Iter=556832/epoch=4, Loss=1.919824, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=557472/epoch=4, Loss=2.408501, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=558112/epoch=4, Loss=2.107957, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=558752/epoch=4, Loss=2.411694, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=559392/epoch=4, Loss=2.008229, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=560032/epoch=4, Loss=2.153971, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=560672/epoch=4, Loss=2.709745, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=561312/epoch=4, Loss=1.709596, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=561952/epoch=4, Loss=2.430285, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=562592/epoch=4, Loss=2.824526, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=563200/epoch=4, Validation Accuracy=0.364955\n",
      "Iter=563232/epoch=4, Loss=1.868149, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=563872/epoch=4, Loss=2.800943, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=564512/epoch=4, Loss=2.555287, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=565152/epoch=4, Loss=2.814980, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=565792/epoch=4, Loss=2.236104, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=566432/epoch=4, Loss=2.229898, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=567072/epoch=4, Loss=2.946438, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=567712/epoch=4, Loss=2.245672, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=568352/epoch=4, Loss=2.308069, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=568992/epoch=4, Loss=2.307662, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=569600/epoch=4, Validation Accuracy=0.378348\n",
      "Iter=569632/epoch=4, Loss=2.781594, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=570272/epoch=4, Loss=2.382946, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=570912/epoch=4, Loss=1.776206, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=571552/epoch=4, Loss=2.320590, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=572192/epoch=4, Loss=2.172871, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=572832/epoch=4, Loss=2.304036, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=573472/epoch=4, Loss=2.411703, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=574112/epoch=4, Loss=3.066831, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=574752/epoch=4, Loss=2.100757, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=575392/epoch=4, Loss=1.777597, Training Accuracy=0.468750, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=576000/epoch=4, Validation Accuracy=0.369420\n",
      "Model saved in file at iteration 576032: ./tf_densenet_model_iter18001.ckpt\n",
      "Iter=576032/epoch=4, Loss=2.604901, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=576672/epoch=4, Loss=2.454968, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=577312/epoch=4, Loss=2.161022, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=577952/epoch=4, Loss=2.634142, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=578592/epoch=4, Loss=2.953367, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=579232/epoch=4, Loss=2.540370, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=579872/epoch=4, Loss=2.593835, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=580512/epoch=4, Loss=2.323925, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=581152/epoch=4, Loss=2.516685, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=581792/epoch=4, Loss=2.528563, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=582400/epoch=4, Validation Accuracy=0.393973\n",
      "Iter=582432/epoch=4, Loss=2.928885, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=583072/epoch=4, Loss=2.429404, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=583712/epoch=4, Loss=2.417090, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=584352/epoch=4, Loss=2.635261, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=584992/epoch=4, Loss=2.463434, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=585632/epoch=4, Loss=2.284193, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=586272/epoch=4, Loss=2.659994, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=586912/epoch=4, Loss=2.889861, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=587552/epoch=4, Loss=2.561912, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=588192/epoch=4, Loss=2.583756, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=588800/epoch=4, Validation Accuracy=0.395089\n",
      "Iter=588832/epoch=4, Loss=2.632690, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=589472/epoch=4, Loss=1.973092, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=590112/epoch=4, Loss=2.305745, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=590752/epoch=4, Loss=2.774832, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=591392/epoch=4, Loss=2.239215, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=592032/epoch=4, Loss=2.593978, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=592672/epoch=4, Loss=2.477957, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=593312/epoch=4, Loss=2.607829, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=593952/epoch=4, Loss=2.605489, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=594592/epoch=4, Loss=2.435650, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=595200/epoch=4, Validation Accuracy=0.386161\n",
      "Iter=595232/epoch=4, Loss=2.182197, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=595872/epoch=4, Loss=2.501416, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=596512/epoch=4, Loss=2.220173, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=597152/epoch=4, Loss=2.034554, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=597792/epoch=4, Loss=2.008189, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=598432/epoch=4, Loss=2.345621, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=599072/epoch=4, Loss=2.478488, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=599712/epoch=4, Loss=2.027343, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=600352/epoch=4, Loss=2.164270, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=600992/epoch=4, Loss=2.339369, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=601600/epoch=4, Validation Accuracy=0.383929\n",
      "Iter=601632/epoch=4, Loss=2.152701, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=602272/epoch=4, Loss=2.755162, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=602912/epoch=4, Loss=2.102418, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=603552/epoch=4, Loss=1.944664, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=604192/epoch=4, Loss=2.562583, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=604832/epoch=4, Loss=2.214838, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=605472/epoch=4, Loss=2.192322, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=606112/epoch=4, Loss=1.958669, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=606752/epoch=4, Loss=2.879102, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=607392/epoch=4, Loss=1.978736, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=608000/epoch=4, Validation Accuracy=0.382812\n",
      "Model saved in file at iteration 608032: ./tf_densenet_model_iter19001.ckpt\n",
      "Iter=608032/epoch=4, Loss=1.694657, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=608672/epoch=4, Loss=2.515968, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=609312/epoch=4, Loss=2.223976, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=609952/epoch=4, Loss=2.112113, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=610592/epoch=4, Loss=2.609779, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=611232/epoch=4, Loss=2.805880, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=611872/epoch=4, Loss=2.438545, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=612512/epoch=4, Loss=2.404473, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=613152/epoch=4, Loss=2.369046, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=613792/epoch=4, Loss=2.252760, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=614400/epoch=4, Validation Accuracy=0.388393\n",
      "Iter=614432/epoch=4, Loss=2.327287, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=615072/epoch=4, Loss=2.107055, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=615712/epoch=4, Loss=2.769276, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=616352/epoch=4, Loss=2.097450, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=616992/epoch=4, Loss=2.649919, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=617632/epoch=4, Loss=2.662286, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=618272/epoch=4, Loss=2.512889, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=618912/epoch=4, Loss=2.677535, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=619552/epoch=4, Loss=2.839581, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=620192/epoch=4, Loss=2.174437, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=620800/epoch=4, Validation Accuracy=0.391741\n",
      "Iter=620832/epoch=4, Loss=3.028445, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=621472/epoch=4, Loss=2.334918, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=622112/epoch=4, Loss=2.209870, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=622752/epoch=4, Loss=2.414366, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=623392/epoch=4, Loss=2.094622, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=624032/epoch=4, Loss=2.738513, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=624672/epoch=4, Loss=1.940375, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=625312/epoch=4, Loss=1.999539, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=625952/epoch=4, Loss=2.150099, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=626592/epoch=4, Loss=2.426355, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=627200/epoch=4, Validation Accuracy=0.373884\n",
      "Iter=627232/epoch=4, Loss=1.946394, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=627872/epoch=4, Loss=2.216502, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=628512/epoch=4, Loss=2.059793, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=629152/epoch=4, Loss=2.580113, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=629792/epoch=4, Loss=2.001489, Training Accuracy=0.687500, lr=0.000100\n",
      "Iter=630432/epoch=4, Loss=2.498834, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=631072/epoch=4, Loss=2.991414, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=631712/epoch=4, Loss=2.279289, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=632352/epoch=4, Loss=2.035419, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=632992/epoch=4, Loss=2.382865, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=633600/epoch=4, Validation Accuracy=0.373884\n",
      "Iter=633632/epoch=4, Loss=2.657886, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=634272/epoch=4, Loss=2.562093, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=634912/epoch=4, Loss=2.608135, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=635552/epoch=4, Loss=2.011950, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=636192/epoch=4, Loss=2.532576, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=636832/epoch=5, Loss=2.579460, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=637472/epoch=5, Loss=2.650733, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=638112/epoch=5, Loss=2.444703, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=638752/epoch=5, Loss=2.345553, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=639392/epoch=5, Loss=2.075950, Training Accuracy=0.312500, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=640000/epoch=5, Validation Accuracy=0.393973\n",
      "Model saved in file at iteration 640032: ./tf_densenet_model_iter20001.ckpt\n",
      "Iter=640032/epoch=5, Loss=2.788638, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=640672/epoch=5, Loss=2.444649, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=641312/epoch=5, Loss=2.043705, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=641952/epoch=5, Loss=2.184963, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=642592/epoch=5, Loss=2.611818, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=643232/epoch=5, Loss=2.354341, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=643872/epoch=5, Loss=2.970813, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=644512/epoch=5, Loss=2.230697, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=645152/epoch=5, Loss=2.412669, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=645792/epoch=5, Loss=2.112513, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=646400/epoch=5, Validation Accuracy=0.402902\n",
      "Iter=646432/epoch=5, Loss=2.621755, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=647072/epoch=5, Loss=2.500920, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=647712/epoch=5, Loss=2.059959, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=648352/epoch=5, Loss=2.650572, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=648992/epoch=5, Loss=2.308145, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=649632/epoch=5, Loss=1.933024, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=650272/epoch=5, Loss=2.368513, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=650912/epoch=5, Loss=2.348664, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=651552/epoch=5, Loss=2.393675, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=652192/epoch=5, Loss=2.540975, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=652800/epoch=5, Validation Accuracy=0.379464\n",
      "Iter=652832/epoch=5, Loss=2.261144, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=653472/epoch=5, Loss=2.671363, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=654112/epoch=5, Loss=1.857904, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=654752/epoch=5, Loss=2.836237, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=655392/epoch=5, Loss=2.427572, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=656032/epoch=5, Loss=2.324716, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=656672/epoch=5, Loss=2.240281, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=657312/epoch=5, Loss=2.290199, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=657952/epoch=5, Loss=1.952037, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=658592/epoch=5, Loss=2.038895, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=659200/epoch=5, Validation Accuracy=0.389509\n",
      "Iter=659232/epoch=5, Loss=2.390978, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=659872/epoch=5, Loss=2.378781, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=660512/epoch=5, Loss=2.547553, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=661152/epoch=5, Loss=2.762353, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=661792/epoch=5, Loss=2.720460, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=662432/epoch=5, Loss=2.447148, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=663072/epoch=5, Loss=2.438237, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=663712/epoch=5, Loss=2.328552, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=664352/epoch=5, Loss=2.445029, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=664992/epoch=5, Loss=2.381036, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=665600/epoch=5, Validation Accuracy=0.386161\n",
      "Iter=665632/epoch=5, Loss=2.347924, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=666272/epoch=5, Loss=2.246518, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=666912/epoch=5, Loss=2.614026, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=667552/epoch=5, Loss=2.686952, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=668192/epoch=5, Loss=1.967986, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=668832/epoch=5, Loss=3.060777, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=669472/epoch=5, Loss=2.611742, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=670112/epoch=5, Loss=2.573946, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=670752/epoch=5, Loss=2.581257, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=671392/epoch=5, Loss=2.702179, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=672000/epoch=5, Validation Accuracy=0.395089\n",
      "Model saved in file at iteration 672032: ./tf_densenet_model_iter21001.ckpt\n",
      "Iter=672032/epoch=5, Loss=2.698609, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=672672/epoch=5, Loss=2.232648, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=673312/epoch=5, Loss=2.796723, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=673952/epoch=5, Loss=2.077486, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=674592/epoch=5, Loss=2.618124, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=675232/epoch=5, Loss=3.141181, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=675872/epoch=5, Loss=3.078457, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=676512/epoch=5, Loss=2.160752, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=677152/epoch=5, Loss=2.147718, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=677792/epoch=5, Loss=2.537578, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=678400/epoch=5, Validation Accuracy=0.377232\n",
      "Iter=678432/epoch=5, Loss=2.206812, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=679072/epoch=5, Loss=2.262860, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=679712/epoch=5, Loss=2.685210, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=680352/epoch=5, Loss=2.347384, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=680992/epoch=5, Loss=2.967117, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=681632/epoch=5, Loss=2.567189, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=682272/epoch=5, Loss=2.406229, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=682912/epoch=5, Loss=2.636151, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=683552/epoch=5, Loss=2.284916, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=684192/epoch=5, Loss=2.728686, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=684800/epoch=5, Validation Accuracy=0.380580\n",
      "Iter=684832/epoch=5, Loss=2.688462, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=685472/epoch=5, Loss=2.418978, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=686112/epoch=5, Loss=2.427492, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=686752/epoch=5, Loss=2.399149, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=687392/epoch=5, Loss=2.186076, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=688032/epoch=5, Loss=2.491776, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=688672/epoch=5, Loss=2.482533, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=689312/epoch=5, Loss=2.511796, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=689952/epoch=5, Loss=2.524869, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=690592/epoch=5, Loss=2.391242, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=691200/epoch=5, Validation Accuracy=0.391741\n",
      "Iter=691232/epoch=5, Loss=2.175407, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=691872/epoch=5, Loss=2.923581, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=692512/epoch=5, Loss=2.788894, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=693152/epoch=5, Loss=2.603664, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=693792/epoch=5, Loss=2.687148, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=694432/epoch=5, Loss=2.058173, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=695072/epoch=5, Loss=2.192085, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=695712/epoch=5, Loss=2.470236, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=696352/epoch=5, Loss=1.994770, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=696992/epoch=5, Loss=2.376163, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=697600/epoch=5, Validation Accuracy=0.382812\n",
      "Iter=697632/epoch=5, Loss=2.791288, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=698272/epoch=5, Loss=2.622119, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=698912/epoch=5, Loss=2.742243, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=699552/epoch=5, Loss=2.571439, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=700192/epoch=5, Loss=2.419565, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=700832/epoch=5, Loss=2.727877, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=701472/epoch=5, Loss=2.975063, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=702112/epoch=5, Loss=2.552325, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=702752/epoch=5, Loss=2.379421, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=703392/epoch=5, Loss=2.748461, Training Accuracy=0.343750, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=704000/epoch=5, Validation Accuracy=0.381696\n",
      "Model saved in file at iteration 704032: ./tf_densenet_model_iter22001.ckpt\n",
      "Iter=704032/epoch=5, Loss=2.179396, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=704672/epoch=5, Loss=2.697342, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=705312/epoch=5, Loss=2.410536, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=705952/epoch=5, Loss=2.085660, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=706592/epoch=5, Loss=2.046515, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=707232/epoch=5, Loss=2.523496, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=707872/epoch=5, Loss=2.567817, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=708512/epoch=5, Loss=2.406390, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=709152/epoch=5, Loss=2.464212, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=709792/epoch=5, Loss=2.375628, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=710400/epoch=5, Validation Accuracy=0.395089\n",
      "Iter=710432/epoch=5, Loss=2.572289, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=711072/epoch=5, Loss=2.260611, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=711712/epoch=5, Loss=2.345065, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=712352/epoch=5, Loss=2.094581, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=712992/epoch=5, Loss=2.551266, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=713632/epoch=5, Loss=2.550829, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=714272/epoch=5, Loss=2.322653, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=714912/epoch=5, Loss=2.326746, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=715552/epoch=5, Loss=2.367615, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=716192/epoch=5, Loss=2.275948, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=716800/epoch=5, Validation Accuracy=0.405134\n",
      "Iter=716832/epoch=5, Loss=2.529175, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=717472/epoch=5, Loss=2.174326, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=718112/epoch=5, Loss=2.350970, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=718752/epoch=5, Loss=2.469637, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=719392/epoch=5, Loss=3.007115, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=720032/epoch=5, Loss=2.452747, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=720672/epoch=5, Loss=2.633526, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=721312/epoch=5, Loss=2.382144, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=721952/epoch=5, Loss=2.390892, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=722592/epoch=5, Loss=2.321816, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=723200/epoch=5, Validation Accuracy=0.389509\n",
      "Iter=723232/epoch=5, Loss=2.224826, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=723872/epoch=5, Loss=1.656192, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=724512/epoch=5, Loss=2.149348, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=725152/epoch=5, Loss=2.447712, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=725792/epoch=5, Loss=2.170824, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=726432/epoch=5, Loss=2.520820, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=727072/epoch=5, Loss=3.038827, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=727712/epoch=5, Loss=2.545162, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=728352/epoch=5, Loss=2.007637, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=728992/epoch=5, Loss=2.473614, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=729600/epoch=5, Validation Accuracy=0.375000\n",
      "Iter=729632/epoch=5, Loss=2.051969, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=730272/epoch=5, Loss=2.659120, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=730912/epoch=5, Loss=2.323422, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=731552/epoch=5, Loss=2.180088, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=732192/epoch=5, Loss=2.294824, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=732832/epoch=5, Loss=2.413200, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=733472/epoch=5, Loss=2.172311, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=734112/epoch=5, Loss=2.388088, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=734752/epoch=5, Loss=2.371064, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=735392/epoch=5, Loss=2.666050, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=736000/epoch=5, Validation Accuracy=0.386161\n",
      "Model saved in file at iteration 736032: ./tf_densenet_model_iter23001.ckpt\n",
      "Iter=736032/epoch=5, Loss=2.568282, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=736672/epoch=5, Loss=2.204037, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=737312/epoch=5, Loss=2.513285, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=737952/epoch=5, Loss=2.199679, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=738592/epoch=5, Loss=2.852245, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=739232/epoch=5, Loss=2.429914, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=739872/epoch=5, Loss=2.776832, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=740512/epoch=5, Loss=2.215076, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=741152/epoch=5, Loss=2.630781, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=741792/epoch=5, Loss=2.470316, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=742400/epoch=5, Validation Accuracy=0.379464\n",
      "Iter=742432/epoch=5, Loss=2.048705, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=743072/epoch=5, Loss=2.047408, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=743712/epoch=5, Loss=1.850343, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=744352/epoch=5, Loss=2.384034, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=744992/epoch=5, Loss=2.419872, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=745632/epoch=5, Loss=2.289120, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=746272/epoch=5, Loss=2.318714, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=746912/epoch=5, Loss=2.282054, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=747552/epoch=5, Loss=2.159389, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=748192/epoch=5, Loss=2.052891, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=748800/epoch=5, Validation Accuracy=0.389509\n",
      "Iter=748832/epoch=5, Loss=2.233442, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=749472/epoch=5, Loss=2.366113, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=750112/epoch=5, Loss=2.807357, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=750752/epoch=5, Loss=2.145699, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=751392/epoch=5, Loss=2.161527, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=752032/epoch=5, Loss=2.782051, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=752672/epoch=5, Loss=2.583723, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=753312/epoch=5, Loss=2.911896, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=753952/epoch=5, Loss=2.196874, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=754592/epoch=5, Loss=2.514433, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=755200/epoch=5, Validation Accuracy=0.381696\n",
      "Iter=755232/epoch=5, Loss=2.902274, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=755872/epoch=5, Loss=2.746535, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=756512/epoch=5, Loss=2.025856, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=757152/epoch=5, Loss=2.496594, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=757792/epoch=5, Loss=2.776608, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=758432/epoch=5, Loss=2.655720, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=759072/epoch=5, Loss=2.500195, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=759712/epoch=5, Loss=3.060468, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=760352/epoch=5, Loss=2.823855, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=760992/epoch=5, Loss=2.576659, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=761600/epoch=5, Validation Accuracy=0.392857\n",
      "Iter=761632/epoch=5, Loss=2.424827, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=762272/epoch=5, Loss=2.389621, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=762912/epoch=5, Loss=1.989298, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=763552/epoch=5, Loss=2.503504, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=764192/epoch=6, Loss=2.292435, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=764832/epoch=6, Loss=2.834091, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=765472/epoch=6, Loss=2.145636, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=766112/epoch=6, Loss=2.094623, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=766752/epoch=6, Loss=1.838855, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=767392/epoch=6, Loss=2.226321, Training Accuracy=0.437500, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=768000/epoch=6, Validation Accuracy=0.389509\n",
      "Model saved in file at iteration 768032: ./tf_densenet_model_iter24001.ckpt\n",
      "Iter=768032/epoch=6, Loss=2.086977, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=768672/epoch=6, Loss=2.728605, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=769312/epoch=6, Loss=2.128515, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=769952/epoch=6, Loss=2.565263, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=770592/epoch=6, Loss=3.102008, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=771232/epoch=6, Loss=2.107795, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=771872/epoch=6, Loss=2.058918, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=772512/epoch=6, Loss=2.144597, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=773152/epoch=6, Loss=1.992313, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=773792/epoch=6, Loss=2.723257, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=774400/epoch=6, Validation Accuracy=0.382812\n",
      "Iter=774432/epoch=6, Loss=2.204040, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=775072/epoch=6, Loss=1.999275, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=775712/epoch=6, Loss=2.609144, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=776352/epoch=6, Loss=2.045927, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=776992/epoch=6, Loss=2.904796, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=777632/epoch=6, Loss=2.421360, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=778272/epoch=6, Loss=2.395568, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=778912/epoch=6, Loss=2.974360, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=779552/epoch=6, Loss=2.751951, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=780192/epoch=6, Loss=2.494861, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=780800/epoch=6, Validation Accuracy=0.383929\n",
      "Iter=780832/epoch=6, Loss=2.240456, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=781472/epoch=6, Loss=1.960495, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=782112/epoch=6, Loss=2.382286, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=782752/epoch=6, Loss=1.984547, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=783392/epoch=6, Loss=2.203699, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=784032/epoch=6, Loss=2.997441, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=784672/epoch=6, Loss=2.085784, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=785312/epoch=6, Loss=1.858604, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=785952/epoch=6, Loss=2.584304, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=786592/epoch=6, Loss=2.509299, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=787200/epoch=6, Validation Accuracy=0.383929\n",
      "Iter=787232/epoch=6, Loss=2.506270, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=787872/epoch=6, Loss=2.574123, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=788512/epoch=6, Loss=2.339279, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=789152/epoch=6, Loss=2.614519, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=789792/epoch=6, Loss=2.826548, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=790432/epoch=6, Loss=2.442731, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=791072/epoch=6, Loss=2.328585, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=791712/epoch=6, Loss=2.609146, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=792352/epoch=6, Loss=2.712946, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=792992/epoch=6, Loss=1.940450, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=793600/epoch=6, Validation Accuracy=0.389509\n",
      "Iter=793632/epoch=6, Loss=2.045944, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=794272/epoch=6, Loss=2.841563, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=794912/epoch=6, Loss=2.496274, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=795552/epoch=6, Loss=1.757067, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=796192/epoch=6, Loss=2.921997, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=796832/epoch=6, Loss=2.663373, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=797472/epoch=6, Loss=2.255620, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=798112/epoch=6, Loss=2.621518, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=798752/epoch=6, Loss=1.699551, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=799392/epoch=6, Loss=2.180006, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=800000/epoch=6, Validation Accuracy=0.392857\n",
      "Model saved in file at iteration 800032: ./tf_densenet_model_iter25001.ckpt\n",
      "Iter=800032/epoch=6, Loss=2.478435, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=800672/epoch=6, Loss=2.081411, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=801312/epoch=6, Loss=2.394409, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=801952/epoch=6, Loss=2.783006, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=802592/epoch=6, Loss=2.318087, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=803232/epoch=6, Loss=2.597782, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=803872/epoch=6, Loss=1.952398, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=804512/epoch=6, Loss=1.957708, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=805152/epoch=6, Loss=2.439942, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=805792/epoch=6, Loss=2.438063, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=806400/epoch=6, Validation Accuracy=0.396205\n",
      "Iter=806432/epoch=6, Loss=2.714617, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=807072/epoch=6, Loss=2.113569, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=807712/epoch=6, Loss=2.627131, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=808352/epoch=6, Loss=2.530613, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=808992/epoch=6, Loss=1.918432, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=809632/epoch=6, Loss=2.921161, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=810272/epoch=6, Loss=2.426965, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=810912/epoch=6, Loss=2.619974, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=811552/epoch=6, Loss=2.595863, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=812192/epoch=6, Loss=2.375327, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=812800/epoch=6, Validation Accuracy=0.383929\n",
      "Iter=812832/epoch=6, Loss=1.721274, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=813472/epoch=6, Loss=2.244042, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=814112/epoch=6, Loss=2.659982, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=814752/epoch=6, Loss=2.371405, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=815392/epoch=6, Loss=3.000001, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=816032/epoch=6, Loss=1.744279, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=816672/epoch=6, Loss=2.462521, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=817312/epoch=6, Loss=2.277568, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=817952/epoch=6, Loss=3.036795, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=818592/epoch=6, Loss=1.890941, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=819200/epoch=6, Validation Accuracy=0.388393\n",
      "Iter=819232/epoch=6, Loss=2.641106, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=819872/epoch=6, Loss=2.091195, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=820512/epoch=6, Loss=2.348312, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=821152/epoch=6, Loss=2.696564, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=821792/epoch=6, Loss=2.017974, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=822432/epoch=6, Loss=2.646784, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=823072/epoch=6, Loss=2.727923, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=823712/epoch=6, Loss=2.710746, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=824352/epoch=6, Loss=2.079175, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=824992/epoch=6, Loss=1.970752, Training Accuracy=0.593750, lr=0.000100\n",
      "\n",
      "Iter=825600/epoch=6, Validation Accuracy=0.399554\n",
      "Iter=825632/epoch=6, Loss=2.291294, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=826272/epoch=6, Loss=2.927532, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=826912/epoch=6, Loss=1.823050, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=827552/epoch=6, Loss=2.031436, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=828192/epoch=6, Loss=2.559504, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=828832/epoch=6, Loss=1.847709, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=829472/epoch=6, Loss=2.171859, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=830112/epoch=6, Loss=2.757259, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=830752/epoch=6, Loss=2.396037, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=831392/epoch=6, Loss=1.931997, Training Accuracy=0.531250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=832000/epoch=6, Validation Accuracy=0.399554\n",
      "Model saved in file at iteration 832032: ./tf_densenet_model_iter26001.ckpt\n",
      "Iter=832032/epoch=6, Loss=2.785538, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=832672/epoch=6, Loss=2.042018, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=833312/epoch=6, Loss=3.065811, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=833952/epoch=6, Loss=2.393185, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=834592/epoch=6, Loss=2.053288, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=835232/epoch=6, Loss=2.943906, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=835872/epoch=6, Loss=2.326264, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=836512/epoch=6, Loss=2.372478, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=837152/epoch=6, Loss=2.818012, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=837792/epoch=6, Loss=2.382511, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=838400/epoch=6, Validation Accuracy=0.399554\n",
      "Iter=838432/epoch=6, Loss=2.249272, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=839072/epoch=6, Loss=2.394058, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=839712/epoch=6, Loss=2.499821, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=840352/epoch=6, Loss=3.029176, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=840992/epoch=6, Loss=2.213738, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=841632/epoch=6, Loss=2.332532, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=842272/epoch=6, Loss=2.720841, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=842912/epoch=6, Loss=2.462427, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=843552/epoch=6, Loss=2.385048, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=844192/epoch=6, Loss=2.397126, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=844800/epoch=6, Validation Accuracy=0.393973\n",
      "Iter=844832/epoch=6, Loss=2.809055, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=845472/epoch=6, Loss=2.569055, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=846112/epoch=6, Loss=2.430678, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=846752/epoch=6, Loss=2.238751, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=847392/epoch=6, Loss=2.306244, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=848032/epoch=6, Loss=2.411685, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=848672/epoch=6, Loss=1.880250, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=849312/epoch=6, Loss=2.351543, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=849952/epoch=6, Loss=2.205078, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=850592/epoch=6, Loss=2.560853, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=851200/epoch=6, Validation Accuracy=0.400670\n",
      "Iter=851232/epoch=6, Loss=2.583019, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=851872/epoch=6, Loss=2.165913, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=852512/epoch=6, Loss=2.313117, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=853152/epoch=6, Loss=2.733491, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=853792/epoch=6, Loss=2.044068, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=854432/epoch=6, Loss=2.325316, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=855072/epoch=6, Loss=2.227815, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=855712/epoch=6, Loss=2.186085, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=856352/epoch=6, Loss=2.286050, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=856992/epoch=6, Loss=2.233201, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=857600/epoch=6, Validation Accuracy=0.393973\n",
      "Iter=857632/epoch=6, Loss=2.037924, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=858272/epoch=6, Loss=2.597071, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=858912/epoch=6, Loss=2.054791, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=859552/epoch=6, Loss=2.234493, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=860192/epoch=6, Loss=2.140998, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=860832/epoch=6, Loss=2.052247, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=861472/epoch=6, Loss=2.158262, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=862112/epoch=6, Loss=2.286462, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=862752/epoch=6, Loss=2.351797, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=863392/epoch=6, Loss=2.368200, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=864000/epoch=6, Validation Accuracy=0.387277\n",
      "Model saved in file at iteration 864032: ./tf_densenet_model_iter27001.ckpt\n",
      "Iter=864032/epoch=6, Loss=2.560365, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=864672/epoch=6, Loss=2.660085, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=865312/epoch=6, Loss=2.476535, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=865952/epoch=6, Loss=2.433209, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=866592/epoch=6, Loss=2.457327, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=867232/epoch=6, Loss=2.441787, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=867872/epoch=6, Loss=2.241057, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=868512/epoch=6, Loss=2.052262, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=869152/epoch=6, Loss=2.405882, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=869792/epoch=6, Loss=2.782227, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=870400/epoch=6, Validation Accuracy=0.415179\n",
      "Iter=870432/epoch=6, Loss=1.844565, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=871072/epoch=6, Loss=2.326070, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=871712/epoch=6, Loss=2.320170, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=872352/epoch=6, Loss=1.981937, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=872992/epoch=6, Loss=2.280644, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=873632/epoch=6, Loss=2.313608, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=874272/epoch=6, Loss=2.492247, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=874912/epoch=6, Loss=2.990732, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=875552/epoch=6, Loss=2.574618, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=876192/epoch=6, Loss=2.167449, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=876800/epoch=6, Validation Accuracy=0.390625\n",
      "Iter=876832/epoch=6, Loss=2.279134, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=877472/epoch=6, Loss=2.245868, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=878112/epoch=6, Loss=2.284457, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=878752/epoch=6, Loss=2.589780, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=879392/epoch=6, Loss=3.122826, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=880032/epoch=6, Loss=2.372459, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=880672/epoch=6, Loss=2.665440, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=881312/epoch=6, Loss=2.891191, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=881952/epoch=6, Loss=2.337057, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=882592/epoch=6, Loss=2.609811, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=883200/epoch=6, Validation Accuracy=0.392857\n",
      "Iter=883232/epoch=6, Loss=2.105537, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=883872/epoch=6, Loss=2.591030, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=884512/epoch=6, Loss=2.088394, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=885152/epoch=6, Loss=1.955916, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=885792/epoch=6, Loss=2.458665, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=886432/epoch=6, Loss=2.421537, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=887072/epoch=6, Loss=2.687350, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=887712/epoch=6, Loss=2.671164, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=888352/epoch=6, Loss=2.637080, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=888992/epoch=6, Loss=1.694024, Training Accuracy=0.562500, lr=0.000100\n",
      "\n",
      "Iter=889600/epoch=6, Validation Accuracy=0.402902\n",
      "Iter=889632/epoch=6, Loss=2.286485, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=890272/epoch=6, Loss=1.726093, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=890912/epoch=6, Loss=1.640151, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=891552/epoch=7, Loss=2.415168, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=892192/epoch=7, Loss=1.985923, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=892832/epoch=7, Loss=2.629621, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=893472/epoch=7, Loss=2.509436, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=894112/epoch=7, Loss=2.847669, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=894752/epoch=7, Loss=2.195190, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=895392/epoch=7, Loss=2.151695, Training Accuracy=0.406250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=896000/epoch=7, Validation Accuracy=0.387277\n",
      "Model saved in file at iteration 896032: ./tf_densenet_model_iter28001.ckpt\n",
      "Iter=896032/epoch=7, Loss=2.654547, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=896672/epoch=7, Loss=2.927577, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=897312/epoch=7, Loss=2.281344, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=897952/epoch=7, Loss=2.025857, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=898592/epoch=7, Loss=2.048821, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=899232/epoch=7, Loss=2.004059, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=899872/epoch=7, Loss=2.602887, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=900512/epoch=7, Loss=2.410299, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=901152/epoch=7, Loss=2.355588, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=901792/epoch=7, Loss=2.053110, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=902400/epoch=7, Validation Accuracy=0.387277\n",
      "Iter=902432/epoch=7, Loss=2.150433, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=903072/epoch=7, Loss=3.048988, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=903712/epoch=7, Loss=2.210833, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=904352/epoch=7, Loss=2.689263, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=904992/epoch=7, Loss=2.326141, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=905632/epoch=7, Loss=2.419965, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=906272/epoch=7, Loss=2.305569, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=906912/epoch=7, Loss=1.970929, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=907552/epoch=7, Loss=2.319307, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=908192/epoch=7, Loss=2.526742, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=908800/epoch=7, Validation Accuracy=0.397321\n",
      "Iter=908832/epoch=7, Loss=2.449563, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=909472/epoch=7, Loss=2.496133, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=910112/epoch=7, Loss=1.876882, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=910752/epoch=7, Loss=2.253047, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=911392/epoch=7, Loss=2.428266, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=912032/epoch=7, Loss=2.042232, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=912672/epoch=7, Loss=2.110263, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=913312/epoch=7, Loss=2.242961, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=913952/epoch=7, Loss=2.277640, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=914592/epoch=7, Loss=2.719669, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=915200/epoch=7, Validation Accuracy=0.402902\n",
      "Iter=915232/epoch=7, Loss=2.589394, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=915872/epoch=7, Loss=2.278225, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=916512/epoch=7, Loss=2.828756, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=917152/epoch=7, Loss=2.521023, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=917792/epoch=7, Loss=2.656065, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=918432/epoch=7, Loss=2.725206, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=919072/epoch=7, Loss=1.839502, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=919712/epoch=7, Loss=2.488919, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=920352/epoch=7, Loss=2.619690, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=920992/epoch=7, Loss=2.268705, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=921600/epoch=7, Validation Accuracy=0.398438\n",
      "Iter=921632/epoch=7, Loss=3.329562, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=922272/epoch=7, Loss=2.420417, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=922912/epoch=7, Loss=1.944136, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=923552/epoch=7, Loss=2.465834, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=924192/epoch=7, Loss=1.851185, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=924832/epoch=7, Loss=1.986092, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=925472/epoch=7, Loss=2.287047, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=926112/epoch=7, Loss=1.947728, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=926752/epoch=7, Loss=2.425709, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=927392/epoch=7, Loss=2.335522, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=928000/epoch=7, Validation Accuracy=0.397321\n",
      "Model saved in file at iteration 928032: ./tf_densenet_model_iter29001.ckpt\n",
      "Iter=928032/epoch=7, Loss=2.519867, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=928672/epoch=7, Loss=2.180702, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=929312/epoch=7, Loss=1.982214, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=929952/epoch=7, Loss=2.692381, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=930592/epoch=7, Loss=2.021945, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=931232/epoch=7, Loss=2.514180, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=931872/epoch=7, Loss=2.539788, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=932512/epoch=7, Loss=2.131139, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=933152/epoch=7, Loss=1.984230, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=933792/epoch=7, Loss=2.412225, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=934400/epoch=7, Validation Accuracy=0.393973\n",
      "Iter=934432/epoch=7, Loss=2.198640, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=935072/epoch=7, Loss=2.151107, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=935712/epoch=7, Loss=1.957834, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=936352/epoch=7, Loss=2.418852, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=936992/epoch=7, Loss=2.509144, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=937632/epoch=7, Loss=2.022909, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=938272/epoch=7, Loss=2.127585, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=938912/epoch=7, Loss=2.734453, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=939552/epoch=7, Loss=2.555774, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=940192/epoch=7, Loss=2.857445, Training Accuracy=0.187500, lr=0.000100\n",
      "\n",
      "Iter=940800/epoch=7, Validation Accuracy=0.391741\n",
      "Iter=940832/epoch=7, Loss=2.153715, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=941472/epoch=7, Loss=2.518564, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=942112/epoch=7, Loss=1.887805, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=942752/epoch=7, Loss=2.171760, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=943392/epoch=7, Loss=1.929494, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=944032/epoch=7, Loss=2.467203, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=944672/epoch=7, Loss=2.441893, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=945312/epoch=7, Loss=2.125134, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=945952/epoch=7, Loss=2.526375, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=946592/epoch=7, Loss=2.643237, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=947200/epoch=7, Validation Accuracy=0.404018\n",
      "Iter=947232/epoch=7, Loss=2.558964, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=947872/epoch=7, Loss=2.431532, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=948512/epoch=7, Loss=2.454013, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=949152/epoch=7, Loss=2.902420, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=949792/epoch=7, Loss=2.262883, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=950432/epoch=7, Loss=1.879160, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=951072/epoch=7, Loss=2.143222, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=951712/epoch=7, Loss=2.277903, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=952352/epoch=7, Loss=2.042334, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=952992/epoch=7, Loss=2.007620, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=953600/epoch=7, Validation Accuracy=0.397321\n",
      "Iter=953632/epoch=7, Loss=2.303722, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=954272/epoch=7, Loss=2.443005, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=954912/epoch=7, Loss=2.234326, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=955552/epoch=7, Loss=2.625837, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=956192/epoch=7, Loss=2.330377, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=956832/epoch=7, Loss=2.879282, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=957472/epoch=7, Loss=2.303052, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=958112/epoch=7, Loss=2.207343, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=958752/epoch=7, Loss=2.860729, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=959392/epoch=7, Loss=2.639758, Training Accuracy=0.312500, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=960000/epoch=7, Validation Accuracy=0.395089\n",
      "Model saved in file at iteration 960032: ./tf_densenet_model_iter30001.ckpt\n",
      "Iter=960032/epoch=7, Loss=2.036014, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=960672/epoch=7, Loss=2.041389, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=961312/epoch=7, Loss=2.207312, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=961952/epoch=7, Loss=2.397024, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=962592/epoch=7, Loss=2.576560, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=963232/epoch=7, Loss=2.622556, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=963872/epoch=7, Loss=2.488340, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=964512/epoch=7, Loss=2.052926, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=965152/epoch=7, Loss=2.472287, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=965792/epoch=7, Loss=2.403474, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=966400/epoch=7, Validation Accuracy=0.393973\n",
      "Iter=966432/epoch=7, Loss=2.171427, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=967072/epoch=7, Loss=2.209481, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=967712/epoch=7, Loss=2.654367, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=968352/epoch=7, Loss=2.289461, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=968992/epoch=7, Loss=2.042243, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=969632/epoch=7, Loss=2.666521, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=970272/epoch=7, Loss=2.402390, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=970912/epoch=7, Loss=2.457185, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=971552/epoch=7, Loss=2.364823, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=972192/epoch=7, Loss=2.257429, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=972800/epoch=7, Validation Accuracy=0.395089\n",
      "Iter=972832/epoch=7, Loss=2.294804, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=973472/epoch=7, Loss=2.323037, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=974112/epoch=7, Loss=2.804722, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=974752/epoch=7, Loss=2.960488, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=975392/epoch=7, Loss=1.479098, Training Accuracy=0.656250, lr=0.000100\n",
      "Iter=976032/epoch=7, Loss=2.648744, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=976672/epoch=7, Loss=2.443084, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=977312/epoch=7, Loss=2.398080, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=977952/epoch=7, Loss=1.913678, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=978592/epoch=7, Loss=2.749927, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=979200/epoch=7, Validation Accuracy=0.401786\n",
      "Iter=979232/epoch=7, Loss=2.746428, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=979872/epoch=7, Loss=2.144559, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=980512/epoch=7, Loss=2.378759, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=981152/epoch=7, Loss=2.918593, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=981792/epoch=7, Loss=2.463634, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=982432/epoch=7, Loss=2.673700, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=983072/epoch=7, Loss=2.103622, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=983712/epoch=7, Loss=2.215987, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=984352/epoch=7, Loss=2.331112, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=984992/epoch=7, Loss=2.373573, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=985600/epoch=7, Validation Accuracy=0.404018\n",
      "Iter=985632/epoch=7, Loss=1.875677, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=986272/epoch=7, Loss=2.108052, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=986912/epoch=7, Loss=2.228743, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=987552/epoch=7, Loss=2.018131, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=988192/epoch=7, Loss=2.550761, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=988832/epoch=7, Loss=2.112799, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=989472/epoch=7, Loss=2.749472, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=990112/epoch=7, Loss=2.180204, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=990752/epoch=7, Loss=1.972517, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=991392/epoch=7, Loss=2.251778, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=992000/epoch=7, Validation Accuracy=0.418527\n",
      "Model saved in file at iteration 992032: ./tf_densenet_model_iter31001.ckpt\n",
      "Iter=992032/epoch=7, Loss=2.496073, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=992672/epoch=7, Loss=2.138474, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=993312/epoch=7, Loss=2.325437, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=993952/epoch=7, Loss=2.263449, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=994592/epoch=7, Loss=2.111445, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=995232/epoch=7, Loss=2.868462, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=995872/epoch=7, Loss=1.622746, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=996512/epoch=7, Loss=2.055565, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=997152/epoch=7, Loss=1.977927, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=997792/epoch=7, Loss=2.679912, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=998400/epoch=7, Validation Accuracy=0.404018\n",
      "Iter=998432/epoch=7, Loss=2.359768, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=999072/epoch=7, Loss=2.364254, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=999712/epoch=7, Loss=1.901084, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1000352/epoch=7, Loss=2.805382, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1000992/epoch=7, Loss=2.503711, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1001632/epoch=7, Loss=2.763529, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1002272/epoch=7, Loss=2.695495, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1002912/epoch=7, Loss=2.764418, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1003552/epoch=7, Loss=2.461689, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1004192/epoch=7, Loss=2.391073, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1004800/epoch=7, Validation Accuracy=0.398438\n",
      "Iter=1004832/epoch=7, Loss=2.348441, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1005472/epoch=7, Loss=2.121242, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1006112/epoch=7, Loss=2.137320, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1006752/epoch=7, Loss=2.123751, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1007392/epoch=7, Loss=2.084833, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1008032/epoch=7, Loss=2.428255, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1008672/epoch=7, Loss=2.096532, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1009312/epoch=7, Loss=2.403971, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1009952/epoch=7, Loss=2.128132, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1010592/epoch=7, Loss=2.608786, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1011200/epoch=7, Validation Accuracy=0.408482\n",
      "Iter=1011232/epoch=7, Loss=2.178080, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1011872/epoch=7, Loss=2.161280, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1012512/epoch=7, Loss=2.596303, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1013152/epoch=7, Loss=2.275398, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1013792/epoch=7, Loss=2.049358, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1014432/epoch=7, Loss=1.933132, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1015072/epoch=7, Loss=2.223854, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1015712/epoch=7, Loss=2.513096, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1016352/epoch=7, Loss=2.070327, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1016992/epoch=7, Loss=1.886559, Training Accuracy=0.562500, lr=0.000100\n",
      "\n",
      "Iter=1017600/epoch=7, Validation Accuracy=0.388393\n",
      "Iter=1017632/epoch=7, Loss=2.513039, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1018272/epoch=7, Loss=2.128200, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1018912/epoch=8, Loss=2.114247, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1019552/epoch=8, Loss=2.051713, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1020192/epoch=8, Loss=2.225075, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1020832/epoch=8, Loss=2.692894, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1021472/epoch=8, Loss=2.697207, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1022112/epoch=8, Loss=2.864005, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1022752/epoch=8, Loss=2.122537, Training Accuracy=0.531250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1023392/epoch=8, Loss=2.104619, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1024000/epoch=8, Validation Accuracy=0.406250\n",
      "Model saved in file at iteration 1024032: ./tf_densenet_model_iter32001.ckpt\n",
      "Iter=1024032/epoch=8, Loss=1.984391, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1024672/epoch=8, Loss=2.812429, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1025312/epoch=8, Loss=2.445618, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1025952/epoch=8, Loss=2.057242, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1026592/epoch=8, Loss=2.698906, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1027232/epoch=8, Loss=2.031300, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1027872/epoch=8, Loss=2.355309, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1028512/epoch=8, Loss=2.018578, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1029152/epoch=8, Loss=2.451777, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1029792/epoch=8, Loss=2.051547, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1030400/epoch=8, Validation Accuracy=0.404018\n",
      "Iter=1030432/epoch=8, Loss=2.209914, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1031072/epoch=8, Loss=2.318210, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1031712/epoch=8, Loss=2.489340, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1032352/epoch=8, Loss=2.547235, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1032992/epoch=8, Loss=2.737175, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1033632/epoch=8, Loss=2.184680, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1034272/epoch=8, Loss=2.202700, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1034912/epoch=8, Loss=2.365524, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1035552/epoch=8, Loss=2.591029, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1036192/epoch=8, Loss=2.091371, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1036800/epoch=8, Validation Accuracy=0.405134\n",
      "Iter=1036832/epoch=8, Loss=2.603769, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1037472/epoch=8, Loss=2.518112, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1038112/epoch=8, Loss=2.520238, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1038752/epoch=8, Loss=3.174590, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1039392/epoch=8, Loss=2.130879, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1040032/epoch=8, Loss=2.472008, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1040672/epoch=8, Loss=2.128147, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1041312/epoch=8, Loss=2.388504, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1041952/epoch=8, Loss=2.341722, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1042592/epoch=8, Loss=1.841659, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1043200/epoch=8, Validation Accuracy=0.390625\n",
      "Iter=1043232/epoch=8, Loss=2.299490, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1043872/epoch=8, Loss=2.297598, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1044512/epoch=8, Loss=2.346369, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1045152/epoch=8, Loss=2.300631, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1045792/epoch=8, Loss=2.316872, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1046432/epoch=8, Loss=2.325572, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1047072/epoch=8, Loss=2.102838, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1047712/epoch=8, Loss=1.922028, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1048352/epoch=8, Loss=1.995743, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1048992/epoch=8, Loss=2.394164, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1049600/epoch=8, Validation Accuracy=0.420759\n",
      "Iter=1049632/epoch=8, Loss=1.927550, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1050272/epoch=8, Loss=2.217581, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1050912/epoch=8, Loss=2.241442, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1051552/epoch=8, Loss=2.105227, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1052192/epoch=8, Loss=1.946609, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1052832/epoch=8, Loss=1.923711, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1053472/epoch=8, Loss=2.365168, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1054112/epoch=8, Loss=2.708147, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1054752/epoch=8, Loss=2.715032, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1055392/epoch=8, Loss=1.675997, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=1056000/epoch=8, Validation Accuracy=0.412946\n",
      "Model saved in file at iteration 1056032: ./tf_densenet_model_iter33001.ckpt\n",
      "Iter=1056032/epoch=8, Loss=2.034086, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1056672/epoch=8, Loss=2.229232, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1057312/epoch=8, Loss=2.110469, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1057952/epoch=8, Loss=2.163174, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1058592/epoch=8, Loss=2.215899, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1059232/epoch=8, Loss=2.749228, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1059872/epoch=8, Loss=2.046755, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1060512/epoch=8, Loss=2.773697, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1061152/epoch=8, Loss=2.088335, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1061792/epoch=8, Loss=1.998335, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1062400/epoch=8, Validation Accuracy=0.414062\n",
      "Iter=1062432/epoch=8, Loss=1.978040, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1063072/epoch=8, Loss=2.722878, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1063712/epoch=8, Loss=2.126194, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1064352/epoch=8, Loss=2.283795, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1064992/epoch=8, Loss=2.549425, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1065632/epoch=8, Loss=1.772904, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1066272/epoch=8, Loss=2.426110, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1066912/epoch=8, Loss=2.467168, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1067552/epoch=8, Loss=2.822242, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1068192/epoch=8, Loss=2.743573, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1068800/epoch=8, Validation Accuracy=0.416295\n",
      "Iter=1068832/epoch=8, Loss=2.355636, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1069472/epoch=8, Loss=2.544452, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1070112/epoch=8, Loss=2.561666, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1070752/epoch=8, Loss=1.972899, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1071392/epoch=8, Loss=1.936986, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1072032/epoch=8, Loss=2.525851, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1072672/epoch=8, Loss=2.276144, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1073312/epoch=8, Loss=2.414974, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1073952/epoch=8, Loss=2.114372, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1074592/epoch=8, Loss=2.691688, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1075200/epoch=8, Validation Accuracy=0.412946\n",
      "Iter=1075232/epoch=8, Loss=2.763699, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1075872/epoch=8, Loss=2.381708, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1076512/epoch=8, Loss=2.280552, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1077152/epoch=8, Loss=2.815754, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1077792/epoch=8, Loss=1.866177, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1078432/epoch=8, Loss=1.718852, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1079072/epoch=8, Loss=1.717285, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1079712/epoch=8, Loss=2.465988, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1080352/epoch=8, Loss=2.368637, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1080992/epoch=8, Loss=2.128771, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1081600/epoch=8, Validation Accuracy=0.418527\n",
      "Iter=1081632/epoch=8, Loss=2.064545, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1082272/epoch=8, Loss=1.827734, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1082912/epoch=8, Loss=2.619326, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1083552/epoch=8, Loss=2.335540, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1084192/epoch=8, Loss=2.306587, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1084832/epoch=8, Loss=2.367400, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1085472/epoch=8, Loss=2.775290, Training Accuracy=0.218750, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1086112/epoch=8, Loss=2.029784, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1086752/epoch=8, Loss=2.529513, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1087392/epoch=8, Loss=2.371602, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1088000/epoch=8, Validation Accuracy=0.396205\n",
      "Model saved in file at iteration 1088032: ./tf_densenet_model_iter34001.ckpt\n",
      "Iter=1088032/epoch=8, Loss=2.413143, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1088672/epoch=8, Loss=2.029887, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1089312/epoch=8, Loss=2.684589, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1089952/epoch=8, Loss=2.970746, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1090592/epoch=8, Loss=2.192987, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1091232/epoch=8, Loss=2.193183, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1091872/epoch=8, Loss=2.431785, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1092512/epoch=8, Loss=2.735785, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1093152/epoch=8, Loss=2.585459, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1093792/epoch=8, Loss=2.656695, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1094400/epoch=8, Validation Accuracy=0.411830\n",
      "Iter=1094432/epoch=8, Loss=2.341198, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1095072/epoch=8, Loss=2.235460, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1095712/epoch=8, Loss=2.406866, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1096352/epoch=8, Loss=2.402262, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1096992/epoch=8, Loss=2.585958, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1097632/epoch=8, Loss=2.483275, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1098272/epoch=8, Loss=2.295172, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1098912/epoch=8, Loss=2.183546, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1099552/epoch=8, Loss=2.267903, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1100192/epoch=8, Loss=2.654619, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1100800/epoch=8, Validation Accuracy=0.399554\n",
      "Iter=1100832/epoch=8, Loss=2.357641, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1101472/epoch=8, Loss=2.360579, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1102112/epoch=8, Loss=2.513265, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1102752/epoch=8, Loss=2.519385, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1103392/epoch=8, Loss=1.867244, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1104032/epoch=8, Loss=2.293097, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1104672/epoch=8, Loss=2.319317, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1105312/epoch=8, Loss=2.449020, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1105952/epoch=8, Loss=2.363979, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1106592/epoch=8, Loss=2.386247, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1107200/epoch=8, Validation Accuracy=0.408482\n",
      "Iter=1107232/epoch=8, Loss=1.986050, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1107872/epoch=8, Loss=2.119561, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1108512/epoch=8, Loss=2.745083, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1109152/epoch=8, Loss=2.002323, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1109792/epoch=8, Loss=2.494155, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1110432/epoch=8, Loss=2.334573, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1111072/epoch=8, Loss=1.936535, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1111712/epoch=8, Loss=2.052833, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1112352/epoch=8, Loss=2.004508, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1112992/epoch=8, Loss=2.065157, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1113600/epoch=8, Validation Accuracy=0.414062\n",
      "Iter=1113632/epoch=8, Loss=2.631335, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1114272/epoch=8, Loss=2.448671, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1114912/epoch=8, Loss=2.816519, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1115552/epoch=8, Loss=2.368030, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1116192/epoch=8, Loss=2.365702, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1116832/epoch=8, Loss=2.284606, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1117472/epoch=8, Loss=1.883933, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1118112/epoch=8, Loss=2.481910, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1118752/epoch=8, Loss=2.352475, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1119392/epoch=8, Loss=2.350888, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1120000/epoch=8, Validation Accuracy=0.406250\n",
      "Model saved in file at iteration 1120032: ./tf_densenet_model_iter35001.ckpt\n",
      "Iter=1120032/epoch=8, Loss=2.532931, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1120672/epoch=8, Loss=2.405023, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1121312/epoch=8, Loss=2.114371, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1121952/epoch=8, Loss=2.596337, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1122592/epoch=8, Loss=1.847553, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1123232/epoch=8, Loss=1.932944, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1123872/epoch=8, Loss=2.241164, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1124512/epoch=8, Loss=2.394366, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1125152/epoch=8, Loss=2.130517, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1125792/epoch=8, Loss=2.048627, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1126400/epoch=8, Validation Accuracy=0.414062\n",
      "Iter=1126432/epoch=8, Loss=2.146428, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1127072/epoch=8, Loss=2.558893, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1127712/epoch=8, Loss=2.243351, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1128352/epoch=8, Loss=3.182829, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1128992/epoch=8, Loss=2.208806, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1129632/epoch=8, Loss=2.535684, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1130272/epoch=8, Loss=2.634152, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1130912/epoch=8, Loss=2.390960, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1131552/epoch=8, Loss=2.536449, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1132192/epoch=8, Loss=2.267317, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1132800/epoch=8, Validation Accuracy=0.410714\n",
      "Iter=1132832/epoch=8, Loss=1.934994, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1133472/epoch=8, Loss=2.761257, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1134112/epoch=8, Loss=2.438239, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1134752/epoch=8, Loss=2.428209, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1135392/epoch=8, Loss=2.061040, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1136032/epoch=8, Loss=2.114301, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1136672/epoch=8, Loss=2.591118, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1137312/epoch=8, Loss=2.427848, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1137952/epoch=8, Loss=2.368096, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1138592/epoch=8, Loss=2.255635, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1139200/epoch=8, Validation Accuracy=0.400670\n",
      "Iter=1139232/epoch=8, Loss=2.011126, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1139872/epoch=8, Loss=2.405863, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1140512/epoch=8, Loss=2.101622, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1141152/epoch=8, Loss=2.120974, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1141792/epoch=8, Loss=2.096800, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1142432/epoch=8, Loss=2.397397, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1143072/epoch=8, Loss=2.355075, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1143712/epoch=8, Loss=2.103904, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1144352/epoch=8, Loss=2.201416, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1144992/epoch=8, Loss=2.504725, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1145600/epoch=8, Validation Accuracy=0.416295\n",
      "Iter=1145632/epoch=8, Loss=2.815791, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1146272/epoch=9, Loss=2.004910, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1146912/epoch=9, Loss=2.754177, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1147552/epoch=9, Loss=2.172886, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1148192/epoch=9, Loss=2.222424, Training Accuracy=0.375000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1148832/epoch=9, Loss=1.960442, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1149472/epoch=9, Loss=2.810583, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1150112/epoch=9, Loss=2.252636, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1150752/epoch=9, Loss=2.555544, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1151392/epoch=9, Loss=1.958781, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1152000/epoch=9, Validation Accuracy=0.405134\n",
      "Model saved in file at iteration 1152032: ./tf_densenet_model_iter36001.ckpt\n",
      "Iter=1152032/epoch=9, Loss=2.220589, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1152672/epoch=9, Loss=2.027325, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1153312/epoch=9, Loss=2.219506, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1153952/epoch=9, Loss=2.295664, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1154592/epoch=9, Loss=2.360525, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1155232/epoch=9, Loss=2.595899, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1155872/epoch=9, Loss=2.292702, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1156512/epoch=9, Loss=2.089555, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1157152/epoch=9, Loss=1.837743, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1157792/epoch=9, Loss=2.589059, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1158400/epoch=9, Validation Accuracy=0.418527\n",
      "Iter=1158432/epoch=9, Loss=2.317671, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1159072/epoch=9, Loss=2.024036, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1159712/epoch=9, Loss=2.498923, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1160352/epoch=9, Loss=2.563967, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1160992/epoch=9, Loss=1.969573, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1161632/epoch=9, Loss=1.710834, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1162272/epoch=9, Loss=1.722276, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1162912/epoch=9, Loss=2.645335, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1163552/epoch=9, Loss=2.522863, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1164192/epoch=9, Loss=2.214605, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=1164800/epoch=9, Validation Accuracy=0.405134\n",
      "Iter=1164832/epoch=9, Loss=2.404844, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1165472/epoch=9, Loss=2.465387, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1166112/epoch=9, Loss=1.820402, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1166752/epoch=9, Loss=1.624399, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1167392/epoch=9, Loss=1.897246, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1168032/epoch=9, Loss=2.658737, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1168672/epoch=9, Loss=1.978607, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1169312/epoch=9, Loss=2.424902, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1169952/epoch=9, Loss=2.085427, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1170592/epoch=9, Loss=2.043641, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1171200/epoch=9, Validation Accuracy=0.411830\n",
      "Iter=1171232/epoch=9, Loss=2.666655, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1171872/epoch=9, Loss=2.347904, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1172512/epoch=9, Loss=2.258675, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1173152/epoch=9, Loss=1.701566, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1173792/epoch=9, Loss=2.939858, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1174432/epoch=9, Loss=2.660960, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1175072/epoch=9, Loss=1.754894, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1175712/epoch=9, Loss=2.312567, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1176352/epoch=9, Loss=2.651918, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1176992/epoch=9, Loss=2.268299, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1177600/epoch=9, Validation Accuracy=0.402902\n",
      "Iter=1177632/epoch=9, Loss=2.047490, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1178272/epoch=9, Loss=2.084307, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1178912/epoch=9, Loss=2.058638, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1179552/epoch=9, Loss=1.752794, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1180192/epoch=9, Loss=2.411956, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1180832/epoch=9, Loss=2.652884, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1181472/epoch=9, Loss=2.835310, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1182112/epoch=9, Loss=2.118968, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1182752/epoch=9, Loss=2.621787, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1183392/epoch=9, Loss=2.703557, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1184000/epoch=9, Validation Accuracy=0.415179\n",
      "Model saved in file at iteration 1184032: ./tf_densenet_model_iter37001.ckpt\n",
      "Iter=1184032/epoch=9, Loss=2.426757, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1184672/epoch=9, Loss=2.406139, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1185312/epoch=9, Loss=2.255248, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1185952/epoch=9, Loss=2.171354, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1186592/epoch=9, Loss=2.415984, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1187232/epoch=9, Loss=2.221272, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1187872/epoch=9, Loss=2.058193, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1188512/epoch=9, Loss=2.704903, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1189152/epoch=9, Loss=2.053970, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1189792/epoch=9, Loss=2.077004, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1190400/epoch=9, Validation Accuracy=0.425223\n",
      "Iter=1190432/epoch=9, Loss=2.581685, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1191072/epoch=9, Loss=2.351983, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1191712/epoch=9, Loss=2.371868, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1192352/epoch=9, Loss=2.260620, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1192992/epoch=9, Loss=1.883513, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1193632/epoch=9, Loss=2.011134, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1194272/epoch=9, Loss=2.548605, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1194912/epoch=9, Loss=2.307552, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1195552/epoch=9, Loss=2.184173, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1196192/epoch=9, Loss=2.699188, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1196800/epoch=9, Validation Accuracy=0.418527\n",
      "Iter=1196832/epoch=9, Loss=2.833944, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1197472/epoch=9, Loss=2.637133, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1198112/epoch=9, Loss=2.558177, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1198752/epoch=9, Loss=2.175562, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1199392/epoch=9, Loss=2.739142, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1200032/epoch=9, Loss=2.477873, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1200672/epoch=9, Loss=2.340963, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1201312/epoch=9, Loss=2.427891, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1201952/epoch=9, Loss=2.172601, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1202592/epoch=9, Loss=2.205246, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1203200/epoch=9, Validation Accuracy=0.417411\n",
      "Iter=1203232/epoch=9, Loss=2.464829, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1203872/epoch=9, Loss=2.474190, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1204512/epoch=9, Loss=2.232263, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1205152/epoch=9, Loss=2.217697, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1205792/epoch=9, Loss=2.271909, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1206432/epoch=9, Loss=2.029428, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1207072/epoch=9, Loss=2.831731, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1207712/epoch=9, Loss=2.245379, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1208352/epoch=9, Loss=2.339681, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1208992/epoch=9, Loss=2.708048, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=1209600/epoch=9, Validation Accuracy=0.414062\n",
      "Iter=1209632/epoch=9, Loss=2.520564, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1210272/epoch=9, Loss=3.164261, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1210912/epoch=9, Loss=2.918930, Training Accuracy=0.375000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1211552/epoch=9, Loss=1.719877, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1212192/epoch=9, Loss=2.978200, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1212832/epoch=9, Loss=2.433479, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1213472/epoch=9, Loss=2.123429, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1214112/epoch=9, Loss=2.192956, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1214752/epoch=9, Loss=1.594854, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1215392/epoch=9, Loss=2.762779, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1216000/epoch=9, Validation Accuracy=0.414062\n",
      "Model saved in file at iteration 1216032: ./tf_densenet_model_iter38001.ckpt\n",
      "Iter=1216032/epoch=9, Loss=2.375354, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1216672/epoch=9, Loss=2.238612, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1217312/epoch=9, Loss=2.272044, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1217952/epoch=9, Loss=2.275033, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1218592/epoch=9, Loss=2.578112, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1219232/epoch=9, Loss=2.921975, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1219872/epoch=9, Loss=2.467648, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1220512/epoch=9, Loss=2.539250, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1221152/epoch=9, Loss=1.820906, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1221792/epoch=9, Loss=2.380675, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1222400/epoch=9, Validation Accuracy=0.429688\n",
      "Iter=1222432/epoch=9, Loss=1.972736, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1223072/epoch=9, Loss=2.352024, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1223712/epoch=9, Loss=1.645132, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1224352/epoch=9, Loss=1.859380, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1224992/epoch=9, Loss=2.211720, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1225632/epoch=9, Loss=2.162467, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1226272/epoch=9, Loss=2.759198, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=1226912/epoch=9, Loss=2.639073, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1227552/epoch=9, Loss=2.074592, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1228192/epoch=9, Loss=2.120713, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1228800/epoch=9, Validation Accuracy=0.438616\n",
      "Iter=1228832/epoch=9, Loss=1.743292, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1229472/epoch=9, Loss=2.292688, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1230112/epoch=9, Loss=2.699993, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1230752/epoch=9, Loss=2.458471, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1231392/epoch=9, Loss=2.634084, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1232032/epoch=9, Loss=2.940590, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1232672/epoch=9, Loss=2.311374, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1233312/epoch=9, Loss=2.430242, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1233952/epoch=9, Loss=1.896813, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1234592/epoch=9, Loss=2.019986, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1235200/epoch=9, Validation Accuracy=0.418527\n",
      "Iter=1235232/epoch=9, Loss=2.463653, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1235872/epoch=9, Loss=2.194860, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1236512/epoch=9, Loss=2.173644, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1237152/epoch=9, Loss=2.640895, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1237792/epoch=9, Loss=2.409357, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1238432/epoch=9, Loss=2.549388, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1239072/epoch=9, Loss=2.023898, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1239712/epoch=9, Loss=2.791169, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1240352/epoch=9, Loss=2.047799, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1240992/epoch=9, Loss=1.642688, Training Accuracy=0.562500, lr=0.000100\n",
      "\n",
      "Iter=1241600/epoch=9, Validation Accuracy=0.415179\n",
      "Iter=1241632/epoch=9, Loss=2.352834, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1242272/epoch=9, Loss=2.088162, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1242912/epoch=9, Loss=2.570949, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1243552/epoch=9, Loss=2.445882, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1244192/epoch=9, Loss=2.491436, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1244832/epoch=9, Loss=2.836141, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1245472/epoch=9, Loss=2.217619, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1246112/epoch=9, Loss=2.420444, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1246752/epoch=9, Loss=2.184647, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1247392/epoch=9, Loss=2.080748, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1248000/epoch=9, Validation Accuracy=0.407366\n",
      "Model saved in file at iteration 1248032: ./tf_densenet_model_iter39001.ckpt\n",
      "Iter=1248032/epoch=9, Loss=2.161067, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1248672/epoch=9, Loss=2.163943, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1249312/epoch=9, Loss=1.965675, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1249952/epoch=9, Loss=2.606614, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1250592/epoch=9, Loss=2.104908, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1251232/epoch=9, Loss=1.744689, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1251872/epoch=9, Loss=2.270685, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1252512/epoch=9, Loss=1.922970, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1253152/epoch=9, Loss=2.264518, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1253792/epoch=9, Loss=2.227113, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1254400/epoch=9, Validation Accuracy=0.409598\n",
      "Iter=1254432/epoch=9, Loss=2.254468, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1255072/epoch=9, Loss=2.455210, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1255712/epoch=9, Loss=2.202770, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1256352/epoch=9, Loss=2.245886, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1256992/epoch=9, Loss=1.890490, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1257632/epoch=9, Loss=2.227712, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1258272/epoch=9, Loss=2.165225, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1258912/epoch=9, Loss=2.759696, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1259552/epoch=9, Loss=2.156688, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1260192/epoch=9, Loss=2.254972, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1260800/epoch=9, Validation Accuracy=0.405134\n",
      "Iter=1260832/epoch=9, Loss=2.605356, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1261472/epoch=9, Loss=2.396070, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1262112/epoch=9, Loss=2.335217, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1262752/epoch=9, Loss=2.728346, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1263392/epoch=9, Loss=2.171535, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1264032/epoch=9, Loss=2.246370, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1264672/epoch=9, Loss=2.242483, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1265312/epoch=9, Loss=1.904123, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1265952/epoch=9, Loss=2.107769, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1266592/epoch=9, Loss=2.012304, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1267200/epoch=9, Validation Accuracy=0.412946\n",
      "Iter=1267232/epoch=9, Loss=2.491652, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1267872/epoch=9, Loss=2.287519, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1268512/epoch=9, Loss=1.942093, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1269152/epoch=9, Loss=2.365821, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1269792/epoch=9, Loss=2.327682, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1270432/epoch=9, Loss=2.470983, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1271072/epoch=9, Loss=2.284834, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1271712/epoch=9, Loss=2.014218, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1272352/epoch=9, Loss=2.211730, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1272992/epoch=9, Loss=2.881210, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1273600/epoch=10, Validation Accuracy=0.404018\n",
      "Iter=1273632/epoch=10, Loss=2.405805, Training Accuracy=0.343750, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1274272/epoch=10, Loss=2.106030, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1274912/epoch=10, Loss=2.220591, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1275552/epoch=10, Loss=2.410861, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1276192/epoch=10, Loss=2.364122, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1276832/epoch=10, Loss=1.575766, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1277472/epoch=10, Loss=2.245997, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1278112/epoch=10, Loss=2.373123, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1278752/epoch=10, Loss=2.202060, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1279392/epoch=10, Loss=2.455319, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1280000/epoch=10, Validation Accuracy=0.425223\n",
      "Model saved in file at iteration 1280032: ./tf_densenet_model_iter40001.ckpt\n",
      "Iter=1280032/epoch=10, Loss=2.183280, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1280672/epoch=10, Loss=2.508535, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1281312/epoch=10, Loss=2.385595, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1281952/epoch=10, Loss=2.095073, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1282592/epoch=10, Loss=2.365843, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1283232/epoch=10, Loss=2.815714, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1283872/epoch=10, Loss=1.904822, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1284512/epoch=10, Loss=2.412863, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1285152/epoch=10, Loss=2.565029, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1285792/epoch=10, Loss=1.744583, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1286400/epoch=10, Validation Accuracy=0.407366\n",
      "Iter=1286432/epoch=10, Loss=1.977009, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1287072/epoch=10, Loss=2.820594, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1287712/epoch=10, Loss=2.544807, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1288352/epoch=10, Loss=1.958723, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1288992/epoch=10, Loss=2.601672, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1289632/epoch=10, Loss=2.767305, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1290272/epoch=10, Loss=2.887527, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1290912/epoch=10, Loss=1.955584, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1291552/epoch=10, Loss=2.463018, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1292192/epoch=10, Loss=2.030718, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1292800/epoch=10, Validation Accuracy=0.419643\n",
      "Iter=1292832/epoch=10, Loss=2.646669, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1293472/epoch=10, Loss=2.434583, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1294112/epoch=10, Loss=2.216544, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1294752/epoch=10, Loss=2.080606, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1295392/epoch=10, Loss=2.051196, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1296032/epoch=10, Loss=2.788918, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1296672/epoch=10, Loss=2.075963, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1297312/epoch=10, Loss=2.069988, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1297952/epoch=10, Loss=2.342505, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1298592/epoch=10, Loss=2.156212, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1299200/epoch=10, Validation Accuracy=0.407366\n",
      "Iter=1299232/epoch=10, Loss=2.459605, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1299872/epoch=10, Loss=2.064092, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1300512/epoch=10, Loss=2.300888, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1301152/epoch=10, Loss=2.117337, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1301792/epoch=10, Loss=1.935918, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1302432/epoch=10, Loss=2.213682, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1303072/epoch=10, Loss=2.268905, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1303712/epoch=10, Loss=2.031307, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1304352/epoch=10, Loss=2.616314, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1304992/epoch=10, Loss=1.915463, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1305600/epoch=10, Validation Accuracy=0.422991\n",
      "Iter=1305632/epoch=10, Loss=1.842181, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1306272/epoch=10, Loss=2.066252, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1306912/epoch=10, Loss=2.157304, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1307552/epoch=10, Loss=2.863307, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1308192/epoch=10, Loss=2.126848, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1308832/epoch=10, Loss=2.264827, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1309472/epoch=10, Loss=2.759686, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1310112/epoch=10, Loss=2.340239, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1310752/epoch=10, Loss=2.560195, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1311392/epoch=10, Loss=2.316429, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1312000/epoch=10, Validation Accuracy=0.426339\n",
      "Model saved in file at iteration 1312032: ./tf_densenet_model_iter41001.ckpt\n",
      "Iter=1312032/epoch=10, Loss=1.995320, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1312672/epoch=10, Loss=2.098848, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1313312/epoch=10, Loss=3.464017, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1313952/epoch=10, Loss=2.345634, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1314592/epoch=10, Loss=2.175801, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1315232/epoch=10, Loss=1.954089, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1315872/epoch=10, Loss=1.850281, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1316512/epoch=10, Loss=2.253623, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1317152/epoch=10, Loss=2.607349, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1317792/epoch=10, Loss=2.534416, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1318400/epoch=10, Validation Accuracy=0.406250\n",
      "Iter=1318432/epoch=10, Loss=1.553486, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1319072/epoch=10, Loss=2.169842, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1319712/epoch=10, Loss=1.807236, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1320352/epoch=10, Loss=2.356277, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1320992/epoch=10, Loss=2.094180, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1321632/epoch=10, Loss=2.194831, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1322272/epoch=10, Loss=2.035758, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1322912/epoch=10, Loss=2.026752, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1323552/epoch=10, Loss=2.108095, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1324192/epoch=10, Loss=2.765684, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1324800/epoch=10, Validation Accuracy=0.422991\n",
      "Iter=1324832/epoch=10, Loss=2.671321, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1325472/epoch=10, Loss=2.501959, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1326112/epoch=10, Loss=1.915166, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1326752/epoch=10, Loss=2.135096, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1327392/epoch=10, Loss=2.812173, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1328032/epoch=10, Loss=2.303942, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1328672/epoch=10, Loss=1.832290, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1329312/epoch=10, Loss=2.194732, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1329952/epoch=10, Loss=2.939545, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1330592/epoch=10, Loss=2.452446, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1331200/epoch=10, Validation Accuracy=0.435268\n",
      "Iter=1331232/epoch=10, Loss=2.833692, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1331872/epoch=10, Loss=1.972834, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1332512/epoch=10, Loss=2.477807, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1333152/epoch=10, Loss=2.076959, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1333792/epoch=10, Loss=1.800578, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1334432/epoch=10, Loss=2.114859, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1335072/epoch=10, Loss=2.170965, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1335712/epoch=10, Loss=2.210145, Training Accuracy=0.406250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1336352/epoch=10, Loss=1.851460, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1336992/epoch=10, Loss=3.008770, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=1337600/epoch=10, Validation Accuracy=0.427455\n",
      "Iter=1337632/epoch=10, Loss=2.098440, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1338272/epoch=10, Loss=2.741660, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1338912/epoch=10, Loss=1.795908, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1339552/epoch=10, Loss=2.187419, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1340192/epoch=10, Loss=1.922998, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1340832/epoch=10, Loss=2.471927, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1341472/epoch=10, Loss=2.451619, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1342112/epoch=10, Loss=2.345044, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1342752/epoch=10, Loss=2.770950, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=1343392/epoch=10, Loss=2.367604, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1344000/epoch=10, Validation Accuracy=0.434152\n",
      "Model saved in file at iteration 1344032: ./tf_densenet_model_iter42001.ckpt\n",
      "Iter=1344032/epoch=10, Loss=2.661378, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1344672/epoch=10, Loss=2.436922, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1345312/epoch=10, Loss=2.467596, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1345952/epoch=10, Loss=2.352662, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1346592/epoch=10, Loss=2.667243, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1347232/epoch=10, Loss=2.274582, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1347872/epoch=10, Loss=2.379058, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1348512/epoch=10, Loss=2.041725, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1349152/epoch=10, Loss=1.757897, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1349792/epoch=10, Loss=2.264190, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1350400/epoch=10, Validation Accuracy=0.430804\n",
      "Iter=1350432/epoch=10, Loss=2.881163, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1351072/epoch=10, Loss=2.322335, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1351712/epoch=10, Loss=2.701812, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1352352/epoch=10, Loss=2.069803, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1352992/epoch=10, Loss=2.637963, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1353632/epoch=10, Loss=2.276640, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1354272/epoch=10, Loss=2.335205, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1354912/epoch=10, Loss=1.968467, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1355552/epoch=10, Loss=2.484289, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1356192/epoch=10, Loss=2.678668, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=1356800/epoch=10, Validation Accuracy=0.416295\n",
      "Iter=1356832/epoch=10, Loss=1.953055, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1357472/epoch=10, Loss=2.378359, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1358112/epoch=10, Loss=1.979099, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1358752/epoch=10, Loss=2.169005, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1359392/epoch=10, Loss=2.116647, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1360032/epoch=10, Loss=2.234944, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1360672/epoch=10, Loss=2.290768, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1361312/epoch=10, Loss=2.293112, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1361952/epoch=10, Loss=2.490526, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1362592/epoch=10, Loss=2.573525, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1363200/epoch=10, Validation Accuracy=0.400670\n",
      "Iter=1363232/epoch=10, Loss=1.662016, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1363872/epoch=10, Loss=2.344186, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1364512/epoch=10, Loss=2.228849, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1365152/epoch=10, Loss=1.703269, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1365792/epoch=10, Loss=2.071370, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1366432/epoch=10, Loss=1.937029, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1367072/epoch=10, Loss=2.305683, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1367712/epoch=10, Loss=2.296128, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1368352/epoch=10, Loss=2.124023, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1368992/epoch=10, Loss=2.091677, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1369600/epoch=10, Validation Accuracy=0.411830\n",
      "Iter=1369632/epoch=10, Loss=1.954389, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1370272/epoch=10, Loss=2.757907, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1370912/epoch=10, Loss=2.699566, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1371552/epoch=10, Loss=2.198690, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1372192/epoch=10, Loss=2.004753, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1372832/epoch=10, Loss=2.178251, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1373472/epoch=10, Loss=2.093132, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1374112/epoch=10, Loss=2.448431, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1374752/epoch=10, Loss=2.437779, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1375392/epoch=10, Loss=2.047860, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=1376000/epoch=10, Validation Accuracy=0.420759\n",
      "Model saved in file at iteration 1376032: ./tf_densenet_model_iter43001.ckpt\n",
      "Iter=1376032/epoch=10, Loss=2.382674, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1376672/epoch=10, Loss=2.199241, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1377312/epoch=10, Loss=1.923860, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1377952/epoch=10, Loss=2.184878, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1378592/epoch=10, Loss=2.038818, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1379232/epoch=10, Loss=2.029533, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1379872/epoch=10, Loss=1.881581, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1380512/epoch=10, Loss=1.789195, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1381152/epoch=10, Loss=1.994651, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1381792/epoch=10, Loss=2.496203, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1382400/epoch=10, Validation Accuracy=0.417411\n",
      "Iter=1382432/epoch=10, Loss=2.528152, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1383072/epoch=10, Loss=2.821745, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1383712/epoch=10, Loss=2.020782, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1384352/epoch=10, Loss=2.199822, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1384992/epoch=10, Loss=1.818104, Training Accuracy=0.656250, lr=0.000100\n",
      "Iter=1385632/epoch=10, Loss=2.220634, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1386272/epoch=10, Loss=2.315652, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1386912/epoch=10, Loss=2.650396, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1387552/epoch=10, Loss=2.164324, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1388192/epoch=10, Loss=2.006703, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1388800/epoch=10, Validation Accuracy=0.422991\n",
      "Iter=1388832/epoch=10, Loss=2.163458, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1389472/epoch=10, Loss=2.382340, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1390112/epoch=10, Loss=2.839266, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1390752/epoch=10, Loss=2.578767, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1391392/epoch=10, Loss=2.457654, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1392032/epoch=10, Loss=2.642829, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1392672/epoch=10, Loss=2.215390, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1393312/epoch=10, Loss=2.275463, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1393952/epoch=10, Loss=2.555321, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1394592/epoch=10, Loss=2.145637, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1395200/epoch=10, Validation Accuracy=0.408482\n",
      "Iter=1395232/epoch=10, Loss=2.834419, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1395872/epoch=10, Loss=2.466240, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1396512/epoch=10, Loss=2.721323, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1397152/epoch=10, Loss=2.318759, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1397792/epoch=10, Loss=2.329385, Training Accuracy=0.312500, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1398432/epoch=10, Loss=1.969702, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1399072/epoch=10, Loss=1.978315, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1399712/epoch=10, Loss=2.406100, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1400352/epoch=10, Loss=2.017407, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1400992/epoch=11, Loss=1.673555, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1401600/epoch=11, Validation Accuracy=0.404018\n",
      "Iter=1401632/epoch=11, Loss=2.088916, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1402272/epoch=11, Loss=2.166985, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1402912/epoch=11, Loss=2.555427, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1403552/epoch=11, Loss=2.907404, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1404192/epoch=11, Loss=2.591931, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1404832/epoch=11, Loss=2.699966, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1405472/epoch=11, Loss=1.941634, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1406112/epoch=11, Loss=2.013743, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1406752/epoch=11, Loss=1.909411, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1407392/epoch=11, Loss=2.084587, Training Accuracy=0.562500, lr=0.000100\n",
      "\n",
      "Iter=1408000/epoch=11, Validation Accuracy=0.426339\n",
      "Model saved in file at iteration 1408032: ./tf_densenet_model_iter44001.ckpt\n",
      "Iter=1408032/epoch=11, Loss=1.775453, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1408672/epoch=11, Loss=2.259557, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1409312/epoch=11, Loss=2.330610, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1409952/epoch=11, Loss=1.994988, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1410592/epoch=11, Loss=2.222963, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1411232/epoch=11, Loss=2.505873, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1411872/epoch=11, Loss=1.955455, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1412512/epoch=11, Loss=2.777436, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1413152/epoch=11, Loss=1.862678, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1413792/epoch=11, Loss=2.570381, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1414400/epoch=11, Validation Accuracy=0.424107\n",
      "Iter=1414432/epoch=11, Loss=2.870917, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1415072/epoch=11, Loss=2.085346, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1415712/epoch=11, Loss=2.241242, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1416352/epoch=11, Loss=2.037819, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1416992/epoch=11, Loss=2.028596, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1417632/epoch=11, Loss=2.328862, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1418272/epoch=11, Loss=2.884066, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1418912/epoch=11, Loss=2.228366, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1419552/epoch=11, Loss=2.550875, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1420192/epoch=11, Loss=2.798650, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1420800/epoch=11, Validation Accuracy=0.416295\n",
      "Iter=1420832/epoch=11, Loss=2.256294, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1421472/epoch=11, Loss=2.288494, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1422112/epoch=11, Loss=1.689018, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1422752/epoch=11, Loss=2.506130, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1423392/epoch=11, Loss=2.103666, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1424032/epoch=11, Loss=2.035877, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1424672/epoch=11, Loss=2.303586, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1425312/epoch=11, Loss=2.572802, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1425952/epoch=11, Loss=2.254490, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1426592/epoch=11, Loss=2.888977, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=1427200/epoch=11, Validation Accuracy=0.421875\n",
      "Iter=1427232/epoch=11, Loss=2.019782, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1427872/epoch=11, Loss=2.103650, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1428512/epoch=11, Loss=3.354867, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1429152/epoch=11, Loss=2.381541, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1429792/epoch=11, Loss=2.217840, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1430432/epoch=11, Loss=2.255054, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1431072/epoch=11, Loss=2.171548, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1431712/epoch=11, Loss=2.057657, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1432352/epoch=11, Loss=2.720360, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1432992/epoch=11, Loss=2.092105, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1433600/epoch=11, Validation Accuracy=0.424107\n",
      "Iter=1433632/epoch=11, Loss=2.252794, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1434272/epoch=11, Loss=2.444035, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1434912/epoch=11, Loss=2.202770, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1435552/epoch=11, Loss=1.996853, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1436192/epoch=11, Loss=2.282891, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1436832/epoch=11, Loss=2.754932, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1437472/epoch=11, Loss=2.576136, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1438112/epoch=11, Loss=2.444992, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1438752/epoch=11, Loss=2.431242, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1439392/epoch=11, Loss=2.463091, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1440000/epoch=11, Validation Accuracy=0.407366\n",
      "Model saved in file at iteration 1440032: ./tf_densenet_model_iter45001.ckpt\n",
      "Iter=1440032/epoch=11, Loss=2.449700, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1440672/epoch=11, Loss=1.553727, Training Accuracy=0.781250, lr=0.000100\n",
      "Iter=1441312/epoch=11, Loss=2.029744, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1441952/epoch=11, Loss=1.561696, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1442592/epoch=11, Loss=1.982505, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1443232/epoch=11, Loss=1.610772, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1443872/epoch=11, Loss=2.142609, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1444512/epoch=11, Loss=1.937284, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1445152/epoch=11, Loss=2.342043, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1445792/epoch=11, Loss=2.459044, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1446400/epoch=11, Validation Accuracy=0.417411\n",
      "Iter=1446432/epoch=11, Loss=1.873677, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1447072/epoch=11, Loss=2.328827, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1447712/epoch=11, Loss=2.645361, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1448352/epoch=11, Loss=2.469144, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1448992/epoch=11, Loss=1.749982, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1449632/epoch=11, Loss=2.502122, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1450272/epoch=11, Loss=1.938549, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1450912/epoch=11, Loss=2.249522, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1451552/epoch=11, Loss=1.810375, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1452192/epoch=11, Loss=1.829081, Training Accuracy=0.562500, lr=0.000100\n",
      "\n",
      "Iter=1452800/epoch=11, Validation Accuracy=0.434152\n",
      "Iter=1452832/epoch=11, Loss=2.099720, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1453472/epoch=11, Loss=2.168637, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1454112/epoch=11, Loss=2.194297, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1454752/epoch=11, Loss=2.149038, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1455392/epoch=11, Loss=2.031435, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1456032/epoch=11, Loss=2.480369, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1456672/epoch=11, Loss=2.510253, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1457312/epoch=11, Loss=2.212018, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1457952/epoch=11, Loss=1.792418, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1458592/epoch=11, Loss=2.255650, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1459200/epoch=11, Validation Accuracy=0.421875\n",
      "Iter=1459232/epoch=11, Loss=2.228672, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1459872/epoch=11, Loss=2.804007, Training Accuracy=0.250000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1460512/epoch=11, Loss=2.092072, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1461152/epoch=11, Loss=2.223129, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1461792/epoch=11, Loss=1.678465, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1462432/epoch=11, Loss=2.354891, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1463072/epoch=11, Loss=2.525780, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1463712/epoch=11, Loss=1.912765, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1464352/epoch=11, Loss=2.038299, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1464992/epoch=11, Loss=2.247514, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1465600/epoch=11, Validation Accuracy=0.430804\n",
      "Iter=1465632/epoch=11, Loss=2.319964, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1466272/epoch=11, Loss=2.107102, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1466912/epoch=11, Loss=2.086537, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1467552/epoch=11, Loss=2.462043, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1468192/epoch=11, Loss=1.941029, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1468832/epoch=11, Loss=2.131187, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1469472/epoch=11, Loss=1.730271, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1470112/epoch=11, Loss=2.119730, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1470752/epoch=11, Loss=2.334773, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1471392/epoch=11, Loss=2.860893, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1472000/epoch=11, Validation Accuracy=0.417411\n",
      "Model saved in file at iteration 1472032: ./tf_densenet_model_iter46001.ckpt\n",
      "Iter=1472032/epoch=11, Loss=2.385539, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1472672/epoch=11, Loss=2.245165, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1473312/epoch=11, Loss=2.651286, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1473952/epoch=11, Loss=2.481030, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1474592/epoch=11, Loss=1.909236, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1475232/epoch=11, Loss=2.081296, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1475872/epoch=11, Loss=2.561347, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1476512/epoch=11, Loss=2.058681, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1477152/epoch=11, Loss=2.055959, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1477792/epoch=11, Loss=2.944150, Training Accuracy=0.187500, lr=0.000100\n",
      "\n",
      "Iter=1478400/epoch=11, Validation Accuracy=0.419643\n",
      "Iter=1478432/epoch=11, Loss=2.384508, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1479072/epoch=11, Loss=2.082546, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1479712/epoch=11, Loss=2.489841, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1480352/epoch=11, Loss=1.855536, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1480992/epoch=11, Loss=2.155268, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1481632/epoch=11, Loss=2.135570, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1482272/epoch=11, Loss=2.233835, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1482912/epoch=11, Loss=2.722361, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1483552/epoch=11, Loss=3.021221, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1484192/epoch=11, Loss=1.936424, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1484800/epoch=11, Validation Accuracy=0.407366\n",
      "Iter=1484832/epoch=11, Loss=2.223192, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1485472/epoch=11, Loss=2.665215, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1486112/epoch=11, Loss=1.977639, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1486752/epoch=11, Loss=2.587023, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1487392/epoch=11, Loss=2.278255, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1488032/epoch=11, Loss=2.338419, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1488672/epoch=11, Loss=1.982992, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1489312/epoch=11, Loss=2.375399, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1489952/epoch=11, Loss=1.637359, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1490592/epoch=11, Loss=2.307857, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1491200/epoch=11, Validation Accuracy=0.415179\n",
      "Iter=1491232/epoch=11, Loss=1.957599, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1491872/epoch=11, Loss=2.171346, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1492512/epoch=11, Loss=2.669726, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1493152/epoch=11, Loss=2.831065, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1493792/epoch=11, Loss=2.235899, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1494432/epoch=11, Loss=2.345534, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1495072/epoch=11, Loss=1.970137, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1495712/epoch=11, Loss=2.059825, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1496352/epoch=11, Loss=2.009688, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1496992/epoch=11, Loss=2.183439, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1497600/epoch=11, Validation Accuracy=0.420759\n",
      "Iter=1497632/epoch=11, Loss=2.837389, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1498272/epoch=11, Loss=2.418700, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1498912/epoch=11, Loss=2.261612, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1499552/epoch=11, Loss=2.293634, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1500192/epoch=11, Loss=2.492270, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1500832/epoch=11, Loss=1.911685, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1501472/epoch=11, Loss=2.687850, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1502112/epoch=11, Loss=1.711753, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1502752/epoch=11, Loss=2.033822, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1503392/epoch=11, Loss=1.730848, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1504000/epoch=11, Validation Accuracy=0.426339\n",
      "Model saved in file at iteration 1504032: ./tf_densenet_model_iter47001.ckpt\n",
      "Iter=1504032/epoch=11, Loss=2.061427, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1504672/epoch=11, Loss=2.165866, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1505312/epoch=11, Loss=2.337769, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1505952/epoch=11, Loss=2.089085, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1506592/epoch=11, Loss=1.983392, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1507232/epoch=11, Loss=1.923385, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1507872/epoch=11, Loss=2.017793, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1508512/epoch=11, Loss=1.960698, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1509152/epoch=11, Loss=2.299835, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1509792/epoch=11, Loss=1.783708, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1510400/epoch=11, Validation Accuracy=0.418527\n",
      "Iter=1510432/epoch=11, Loss=2.517137, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1511072/epoch=11, Loss=2.081599, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1511712/epoch=11, Loss=2.136952, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1512352/epoch=11, Loss=1.869757, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1512992/epoch=11, Loss=2.258902, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1513632/epoch=11, Loss=1.618169, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1514272/epoch=11, Loss=2.418752, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1514912/epoch=11, Loss=2.089438, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1515552/epoch=11, Loss=2.634447, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1516192/epoch=11, Loss=2.210819, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=1516800/epoch=11, Validation Accuracy=0.429688\n",
      "Iter=1516832/epoch=11, Loss=2.155523, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1517472/epoch=11, Loss=2.619149, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1518112/epoch=11, Loss=2.626670, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1518752/epoch=11, Loss=2.444860, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1519392/epoch=11, Loss=2.104126, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1520032/epoch=11, Loss=2.497248, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1520672/epoch=11, Loss=2.302576, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1521312/epoch=11, Loss=1.899563, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1521952/epoch=11, Loss=2.193469, Training Accuracy=0.406250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1522592/epoch=11, Loss=3.007765, Training Accuracy=0.156250, lr=0.000100\n",
      "\n",
      "Iter=1523200/epoch=11, Validation Accuracy=0.444196\n",
      "Iter=1523232/epoch=11, Loss=2.231587, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1523872/epoch=11, Loss=1.977923, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1524512/epoch=11, Loss=2.456143, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1525152/epoch=11, Loss=2.659019, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1525792/epoch=11, Loss=2.021969, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1526432/epoch=11, Loss=2.757872, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1527072/epoch=11, Loss=1.883367, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1527712/epoch=11, Loss=2.074158, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1528352/epoch=12, Loss=2.495477, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1528992/epoch=12, Loss=2.450176, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1529600/epoch=12, Validation Accuracy=0.428571\n",
      "Iter=1529632/epoch=12, Loss=2.436236, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1530272/epoch=12, Loss=2.509936, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1530912/epoch=12, Loss=2.071642, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1531552/epoch=12, Loss=2.672606, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1532192/epoch=12, Loss=2.365960, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1532832/epoch=12, Loss=2.095706, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1533472/epoch=12, Loss=1.856139, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1534112/epoch=12, Loss=1.714676, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1534752/epoch=12, Loss=2.501549, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1535392/epoch=12, Loss=2.059428, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1536000/epoch=12, Validation Accuracy=0.420759\n",
      "Model saved in file at iteration 1536032: ./tf_densenet_model_iter48001.ckpt\n",
      "Iter=1536032/epoch=12, Loss=2.412930, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1536672/epoch=12, Loss=2.182364, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1537312/epoch=12, Loss=1.834137, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1537952/epoch=12, Loss=2.123724, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1538592/epoch=12, Loss=2.252321, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1539232/epoch=12, Loss=2.141999, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1539872/epoch=12, Loss=2.313540, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1540512/epoch=12, Loss=2.700753, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1541152/epoch=12, Loss=2.325675, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1541792/epoch=12, Loss=2.167842, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1542400/epoch=12, Validation Accuracy=0.417411\n",
      "Iter=1542432/epoch=12, Loss=1.725999, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1543072/epoch=12, Loss=2.828053, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1543712/epoch=12, Loss=2.106383, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1544352/epoch=12, Loss=1.662070, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1544992/epoch=12, Loss=2.238627, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1545632/epoch=12, Loss=2.136851, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1546272/epoch=12, Loss=2.156752, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1546912/epoch=12, Loss=1.807374, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1547552/epoch=12, Loss=1.857752, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1548192/epoch=12, Loss=3.039217, Training Accuracy=0.218750, lr=0.000100\n",
      "\n",
      "Iter=1548800/epoch=12, Validation Accuracy=0.426339\n",
      "Iter=1548832/epoch=12, Loss=2.153332, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1549472/epoch=12, Loss=2.184777, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1550112/epoch=12, Loss=2.062180, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1550752/epoch=12, Loss=2.603780, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1551392/epoch=12, Loss=2.523927, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1552032/epoch=12, Loss=2.420053, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1552672/epoch=12, Loss=2.290602, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1553312/epoch=12, Loss=1.945109, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1553952/epoch=12, Loss=2.111684, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1554592/epoch=12, Loss=2.099733, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1555200/epoch=12, Validation Accuracy=0.443080\n",
      "Iter=1555232/epoch=12, Loss=2.623128, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1555872/epoch=12, Loss=2.445916, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1556512/epoch=12, Loss=1.929935, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1557152/epoch=12, Loss=2.635240, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1557792/epoch=12, Loss=1.762298, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1558432/epoch=12, Loss=2.117821, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1559072/epoch=12, Loss=2.074459, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1559712/epoch=12, Loss=2.284766, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1560352/epoch=12, Loss=3.002992, Training Accuracy=0.156250, lr=0.000100\n",
      "Iter=1560992/epoch=12, Loss=1.799732, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1561600/epoch=12, Validation Accuracy=0.424107\n",
      "Iter=1561632/epoch=12, Loss=2.048885, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1562272/epoch=12, Loss=2.170264, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1562912/epoch=12, Loss=2.221319, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1563552/epoch=12, Loss=2.339711, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1564192/epoch=12, Loss=2.236489, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1564832/epoch=12, Loss=2.261503, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1565472/epoch=12, Loss=2.021657, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1566112/epoch=12, Loss=2.254101, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1566752/epoch=12, Loss=2.612715, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1567392/epoch=12, Loss=2.394563, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1568000/epoch=12, Validation Accuracy=0.441964\n",
      "Model saved in file at iteration 1568032: ./tf_densenet_model_iter49001.ckpt\n",
      "Iter=1568032/epoch=12, Loss=2.530097, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1568672/epoch=12, Loss=2.433970, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1569312/epoch=12, Loss=2.291620, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1569952/epoch=12, Loss=2.278193, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1570592/epoch=12, Loss=2.548073, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1571232/epoch=12, Loss=2.711814, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1571872/epoch=12, Loss=1.994152, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1572512/epoch=12, Loss=1.796356, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1573152/epoch=12, Loss=1.910981, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1573792/epoch=12, Loss=2.375377, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1574400/epoch=12, Validation Accuracy=0.448661\n",
      "Iter=1574432/epoch=12, Loss=2.169132, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1575072/epoch=12, Loss=2.357258, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1575712/epoch=12, Loss=2.519034, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1576352/epoch=12, Loss=2.503300, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1576992/epoch=12, Loss=2.211658, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1577632/epoch=12, Loss=1.979842, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1578272/epoch=12, Loss=2.424822, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1578912/epoch=12, Loss=2.191591, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1579552/epoch=12, Loss=1.981602, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1580192/epoch=12, Loss=2.016027, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1580800/epoch=12, Validation Accuracy=0.434152\n",
      "Iter=1580832/epoch=12, Loss=2.042734, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1581472/epoch=12, Loss=1.887570, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1582112/epoch=12, Loss=2.380763, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1582752/epoch=12, Loss=2.049095, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1583392/epoch=12, Loss=2.173288, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1584032/epoch=12, Loss=1.809887, Training Accuracy=0.531250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1584672/epoch=12, Loss=2.452800, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1585312/epoch=12, Loss=2.052290, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1585952/epoch=12, Loss=2.288027, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1586592/epoch=12, Loss=2.147201, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1587200/epoch=12, Validation Accuracy=0.436384\n",
      "Iter=1587232/epoch=12, Loss=2.228517, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1587872/epoch=12, Loss=1.985860, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1588512/epoch=12, Loss=2.669905, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1589152/epoch=12, Loss=2.385065, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1589792/epoch=12, Loss=2.021894, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1590432/epoch=12, Loss=1.729329, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1591072/epoch=12, Loss=2.130864, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1591712/epoch=12, Loss=2.132607, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1592352/epoch=12, Loss=1.755453, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1592992/epoch=12, Loss=2.364585, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1593600/epoch=12, Validation Accuracy=0.417411\n",
      "Iter=1593632/epoch=12, Loss=2.432054, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1594272/epoch=12, Loss=2.018770, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1594912/epoch=12, Loss=2.533292, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1595552/epoch=12, Loss=2.101064, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1596192/epoch=12, Loss=2.394236, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1596832/epoch=12, Loss=1.820263, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1597472/epoch=12, Loss=1.606642, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1598112/epoch=12, Loss=2.923078, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1598752/epoch=12, Loss=1.949154, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1599392/epoch=12, Loss=2.143899, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1600000/epoch=12, Validation Accuracy=0.436384\n",
      "Model saved in file at iteration 1600032: ./tf_densenet_model_iter50001.ckpt\n",
      "Iter=1600032/epoch=12, Loss=2.584969, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1600672/epoch=12, Loss=1.893191, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1601312/epoch=12, Loss=2.072024, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1601952/epoch=12, Loss=2.117221, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1602592/epoch=12, Loss=2.214858, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1603232/epoch=12, Loss=2.002045, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1603872/epoch=12, Loss=2.167990, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1604512/epoch=12, Loss=2.842277, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1605152/epoch=12, Loss=2.030045, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1605792/epoch=12, Loss=2.249832, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1606400/epoch=12, Validation Accuracy=0.433036\n",
      "Iter=1606432/epoch=12, Loss=2.723270, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1607072/epoch=12, Loss=1.898194, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1607712/epoch=12, Loss=1.945402, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1608352/epoch=12, Loss=2.476061, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1608992/epoch=12, Loss=2.221677, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1609632/epoch=12, Loss=2.163568, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1610272/epoch=12, Loss=2.247947, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1610912/epoch=12, Loss=2.259803, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1611552/epoch=12, Loss=1.938026, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1612192/epoch=12, Loss=2.309381, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1612800/epoch=12, Validation Accuracy=0.421875\n",
      "Iter=1612832/epoch=12, Loss=2.109785, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1613472/epoch=12, Loss=1.803302, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1614112/epoch=12, Loss=2.225589, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1614752/epoch=12, Loss=2.245117, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1615392/epoch=12, Loss=2.488537, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1616032/epoch=12, Loss=1.936712, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1616672/epoch=12, Loss=2.144445, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1617312/epoch=12, Loss=2.056010, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1617952/epoch=12, Loss=2.473473, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1618592/epoch=12, Loss=1.896046, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1619200/epoch=12, Validation Accuracy=0.425223\n",
      "Iter=1619232/epoch=12, Loss=2.200074, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1619872/epoch=12, Loss=1.832465, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1620512/epoch=12, Loss=2.097297, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1621152/epoch=12, Loss=2.237214, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1621792/epoch=12, Loss=1.745229, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1622432/epoch=12, Loss=2.434006, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1623072/epoch=12, Loss=2.136789, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1623712/epoch=12, Loss=2.830859, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1624352/epoch=12, Loss=2.656157, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1624992/epoch=12, Loss=2.056336, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1625600/epoch=12, Validation Accuracy=0.422991\n",
      "Iter=1625632/epoch=12, Loss=1.896982, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1626272/epoch=12, Loss=2.339784, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1626912/epoch=12, Loss=2.424990, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1627552/epoch=12, Loss=2.176076, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1628192/epoch=12, Loss=2.000228, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1628832/epoch=12, Loss=1.815997, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1629472/epoch=12, Loss=1.749203, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1630112/epoch=12, Loss=1.977147, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1630752/epoch=12, Loss=2.001944, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1631392/epoch=12, Loss=2.598257, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1632000/epoch=12, Validation Accuracy=0.434152\n",
      "Model saved in file at iteration 1632032: ./tf_densenet_model_iter51001.ckpt\n",
      "Iter=1632032/epoch=12, Loss=2.064240, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1632672/epoch=12, Loss=2.236240, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1633312/epoch=12, Loss=2.817019, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1633952/epoch=12, Loss=1.736060, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1634592/epoch=12, Loss=2.938285, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1635232/epoch=12, Loss=2.402670, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1635872/epoch=12, Loss=2.267941, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1636512/epoch=12, Loss=2.086138, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1637152/epoch=12, Loss=2.440427, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1637792/epoch=12, Loss=2.185485, Training Accuracy=0.250000, lr=0.000100\n",
      "\n",
      "Iter=1638400/epoch=12, Validation Accuracy=0.436384\n",
      "Iter=1638432/epoch=12, Loss=1.691355, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1639072/epoch=12, Loss=2.269432, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1639712/epoch=12, Loss=1.780952, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1640352/epoch=12, Loss=1.945242, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1640992/epoch=12, Loss=2.234295, Training Accuracy=0.218750, lr=0.000100\n",
      "Iter=1641632/epoch=12, Loss=2.384713, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1642272/epoch=12, Loss=2.187764, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1642912/epoch=12, Loss=2.744557, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1643552/epoch=12, Loss=1.709409, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1644192/epoch=12, Loss=2.388062, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1644800/epoch=12, Validation Accuracy=0.429688\n",
      "Iter=1644832/epoch=12, Loss=2.766066, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1645472/epoch=12, Loss=2.158225, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1646112/epoch=12, Loss=2.038149, Training Accuracy=0.468750, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1646752/epoch=12, Loss=2.439831, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1647392/epoch=12, Loss=2.184305, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1648032/epoch=12, Loss=2.585201, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1648672/epoch=12, Loss=2.384521, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1649312/epoch=12, Loss=1.937416, Training Accuracy=0.656250, lr=0.000100\n",
      "Iter=1649952/epoch=12, Loss=2.343494, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1650592/epoch=12, Loss=2.188044, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=1651200/epoch=12, Validation Accuracy=0.431920\n",
      "Iter=1651232/epoch=12, Loss=2.297215, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1651872/epoch=12, Loss=2.071112, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1652512/epoch=12, Loss=1.656175, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1653152/epoch=12, Loss=1.962408, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1653792/epoch=12, Loss=1.882871, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1654432/epoch=12, Loss=2.566921, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1655072/epoch=12, Loss=2.566872, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1655712/epoch=13, Loss=2.166618, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1656352/epoch=13, Loss=2.204250, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1656992/epoch=13, Loss=2.348972, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1657600/epoch=13, Validation Accuracy=0.431920\n",
      "Iter=1657632/epoch=13, Loss=2.185487, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1658272/epoch=13, Loss=2.536325, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1658912/epoch=13, Loss=2.113722, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1659552/epoch=13, Loss=2.446890, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1660192/epoch=13, Loss=1.615613, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1660832/epoch=13, Loss=2.206395, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1661472/epoch=13, Loss=1.997027, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1662112/epoch=13, Loss=1.848260, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1662752/epoch=13, Loss=1.937970, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1663392/epoch=13, Loss=2.150474, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1664000/epoch=13, Validation Accuracy=0.436384\n",
      "Model saved in file at iteration 1664032: ./tf_densenet_model_iter52001.ckpt\n",
      "Iter=1664032/epoch=13, Loss=1.944148, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1664672/epoch=13, Loss=2.105521, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1665312/epoch=13, Loss=2.043499, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1665952/epoch=13, Loss=2.077933, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1666592/epoch=13, Loss=2.221962, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1667232/epoch=13, Loss=2.156698, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1667872/epoch=13, Loss=2.151647, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1668512/epoch=13, Loss=2.734488, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1669152/epoch=13, Loss=1.750458, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1669792/epoch=13, Loss=2.388588, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1670400/epoch=13, Validation Accuracy=0.425223\n",
      "Iter=1670432/epoch=13, Loss=2.237025, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1671072/epoch=13, Loss=2.254469, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1671712/epoch=13, Loss=1.762872, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1672352/epoch=13, Loss=2.096229, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1672992/epoch=13, Loss=2.013216, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1673632/epoch=13, Loss=2.059999, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1674272/epoch=13, Loss=2.486408, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1674912/epoch=13, Loss=1.942506, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1675552/epoch=13, Loss=2.570251, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1676192/epoch=13, Loss=2.097572, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1676800/epoch=13, Validation Accuracy=0.425223\n",
      "Iter=1676832/epoch=13, Loss=2.059281, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1677472/epoch=13, Loss=2.347206, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1678112/epoch=13, Loss=2.123390, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1678752/epoch=13, Loss=1.888274, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1679392/epoch=13, Loss=1.614250, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1680032/epoch=13, Loss=2.078818, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1680672/epoch=13, Loss=1.801693, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1681312/epoch=13, Loss=2.816914, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1681952/epoch=13, Loss=2.208519, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1682592/epoch=13, Loss=2.971641, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1683200/epoch=13, Validation Accuracy=0.427455\n",
      "Iter=1683232/epoch=13, Loss=2.529589, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1683872/epoch=13, Loss=2.418847, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1684512/epoch=13, Loss=2.111510, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1685152/epoch=13, Loss=2.233946, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1685792/epoch=13, Loss=1.961865, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1686432/epoch=13, Loss=2.152937, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1687072/epoch=13, Loss=2.196831, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1687712/epoch=13, Loss=2.461027, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1688352/epoch=13, Loss=2.157058, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1688992/epoch=13, Loss=2.552706, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1689600/epoch=13, Validation Accuracy=0.430804\n",
      "Iter=1689632/epoch=13, Loss=2.021303, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1690272/epoch=13, Loss=2.277832, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1690912/epoch=13, Loss=1.741996, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1691552/epoch=13, Loss=1.727672, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1692192/epoch=13, Loss=2.123837, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1692832/epoch=13, Loss=1.937285, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1693472/epoch=13, Loss=1.918500, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1694112/epoch=13, Loss=2.272409, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1694752/epoch=13, Loss=2.291479, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1695392/epoch=13, Loss=1.819261, Training Accuracy=0.312500, lr=0.000100\n",
      "\n",
      "Iter=1696000/epoch=13, Validation Accuracy=0.419643\n",
      "Model saved in file at iteration 1696032: ./tf_densenet_model_iter53001.ckpt\n",
      "Iter=1696032/epoch=13, Loss=2.519642, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1696672/epoch=13, Loss=2.444852, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1697312/epoch=13, Loss=2.344819, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1697952/epoch=13, Loss=2.437329, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1698592/epoch=13, Loss=1.973737, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1699232/epoch=13, Loss=2.305348, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1699872/epoch=13, Loss=1.706394, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1700512/epoch=13, Loss=2.205816, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1701152/epoch=13, Loss=2.111618, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1701792/epoch=13, Loss=1.780858, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=1702400/epoch=13, Validation Accuracy=0.430804\n",
      "Iter=1702432/epoch=13, Loss=2.007641, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1703072/epoch=13, Loss=1.805086, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1703712/epoch=13, Loss=2.163961, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1704352/epoch=13, Loss=1.722604, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1704992/epoch=13, Loss=2.668919, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1705632/epoch=13, Loss=2.315640, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1706272/epoch=13, Loss=2.269992, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1706912/epoch=13, Loss=2.421027, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1707552/epoch=13, Loss=2.606828, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1708192/epoch=13, Loss=1.826479, Training Accuracy=0.406250, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iter=1708800/epoch=13, Validation Accuracy=0.450893\n",
      "Iter=1708832/epoch=13, Loss=2.437619, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1709472/epoch=13, Loss=2.164026, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1710112/epoch=13, Loss=1.963714, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1710752/epoch=13, Loss=1.679650, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1711392/epoch=13, Loss=1.923410, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1712032/epoch=13, Loss=3.049421, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1712672/epoch=13, Loss=1.825582, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1713312/epoch=13, Loss=2.275497, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1713952/epoch=13, Loss=2.023242, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1714592/epoch=13, Loss=2.571919, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1715200/epoch=13, Validation Accuracy=0.426339\n",
      "Iter=1715232/epoch=13, Loss=2.357147, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1715872/epoch=13, Loss=2.428443, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1716512/epoch=13, Loss=2.152878, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1717152/epoch=13, Loss=1.987950, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1717792/epoch=13, Loss=2.241944, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1718432/epoch=13, Loss=2.213686, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1719072/epoch=13, Loss=1.740054, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1719712/epoch=13, Loss=2.324636, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1720352/epoch=13, Loss=2.857011, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1720992/epoch=13, Loss=2.040951, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1721600/epoch=13, Validation Accuracy=0.446429\n",
      "Iter=1721632/epoch=13, Loss=2.253922, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1722272/epoch=13, Loss=2.325170, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1722912/epoch=13, Loss=2.212377, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1723552/epoch=13, Loss=2.065577, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1724192/epoch=13, Loss=1.851467, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1724832/epoch=13, Loss=2.144767, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1725472/epoch=13, Loss=2.756350, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1726112/epoch=13, Loss=2.732496, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1726752/epoch=13, Loss=2.069514, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1727392/epoch=13, Loss=1.817729, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1728000/epoch=13, Validation Accuracy=0.418527\n",
      "Model saved in file at iteration 1728032: ./tf_densenet_model_iter54001.ckpt\n",
      "Iter=1728032/epoch=13, Loss=2.130386, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1728672/epoch=13, Loss=2.448395, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1729312/epoch=13, Loss=1.837467, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1729952/epoch=13, Loss=2.646404, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1730592/epoch=13, Loss=2.272952, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1731232/epoch=13, Loss=2.336970, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1731872/epoch=13, Loss=3.280895, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1732512/epoch=13, Loss=2.204154, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1733152/epoch=13, Loss=1.736674, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1733792/epoch=13, Loss=2.252904, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1734400/epoch=13, Validation Accuracy=0.420759\n",
      "Iter=1734432/epoch=13, Loss=2.227356, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1735072/epoch=13, Loss=1.957652, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1735712/epoch=13, Loss=2.058103, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1736352/epoch=13, Loss=1.960914, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1736992/epoch=13, Loss=1.811459, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1737632/epoch=13, Loss=2.120144, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1738272/epoch=13, Loss=2.156881, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1738912/epoch=13, Loss=2.577771, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1739552/epoch=13, Loss=2.581970, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1740192/epoch=13, Loss=2.450705, Training Accuracy=0.281250, lr=0.000100\n",
      "\n",
      "Iter=1740800/epoch=13, Validation Accuracy=0.448661\n",
      "Iter=1740832/epoch=13, Loss=2.511968, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1741472/epoch=13, Loss=1.852635, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1742112/epoch=13, Loss=1.824126, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1742752/epoch=13, Loss=2.444853, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1743392/epoch=13, Loss=1.905691, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1744032/epoch=13, Loss=2.394064, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1744672/epoch=13, Loss=2.626417, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1745312/epoch=13, Loss=1.632055, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1745952/epoch=13, Loss=2.465330, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1746592/epoch=13, Loss=2.078802, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1747200/epoch=13, Validation Accuracy=0.433036\n",
      "Iter=1747232/epoch=13, Loss=2.306995, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1747872/epoch=13, Loss=2.024294, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1748512/epoch=13, Loss=2.656395, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1749152/epoch=13, Loss=1.631614, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1749792/epoch=13, Loss=1.853487, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1750432/epoch=13, Loss=2.027363, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1751072/epoch=13, Loss=2.291260, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1751712/epoch=13, Loss=2.921785, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1752352/epoch=13, Loss=2.020457, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1752992/epoch=13, Loss=2.194981, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1753600/epoch=13, Validation Accuracy=0.424107\n",
      "Iter=1753632/epoch=13, Loss=2.011492, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1754272/epoch=13, Loss=1.646107, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1754912/epoch=13, Loss=2.081270, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1755552/epoch=13, Loss=2.051682, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1756192/epoch=13, Loss=2.187046, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1756832/epoch=13, Loss=2.377464, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1757472/epoch=13, Loss=2.197072, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1758112/epoch=13, Loss=2.329443, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1758752/epoch=13, Loss=1.532876, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1759392/epoch=13, Loss=2.491320, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1760000/epoch=13, Validation Accuracy=0.427455\n",
      "Model saved in file at iteration 1760032: ./tf_densenet_model_iter55001.ckpt\n",
      "Iter=1760032/epoch=13, Loss=2.293705, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1760672/epoch=13, Loss=2.257756, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1761312/epoch=13, Loss=2.211643, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1761952/epoch=13, Loss=2.402281, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1762592/epoch=13, Loss=2.044493, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1763232/epoch=13, Loss=2.374553, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1763872/epoch=13, Loss=2.127327, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1764512/epoch=13, Loss=2.333350, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1765152/epoch=13, Loss=1.903506, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1765792/epoch=13, Loss=2.295331, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1766400/epoch=13, Validation Accuracy=0.436384\n",
      "Iter=1766432/epoch=13, Loss=1.916330, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1767072/epoch=13, Loss=2.801631, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1767712/epoch=13, Loss=2.284485, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1768352/epoch=13, Loss=1.910454, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1768992/epoch=13, Loss=1.840539, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1769632/epoch=13, Loss=2.238065, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1770272/epoch=13, Loss=2.163220, Training Accuracy=0.437500, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1770912/epoch=13, Loss=1.713547, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1771552/epoch=13, Loss=2.297004, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1772192/epoch=13, Loss=2.329874, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1772800/epoch=13, Validation Accuracy=0.435268\n",
      "Iter=1772832/epoch=13, Loss=2.274540, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1773472/epoch=13, Loss=2.488975, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1774112/epoch=13, Loss=2.510805, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1774752/epoch=13, Loss=1.973266, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1775392/epoch=13, Loss=2.258759, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1776032/epoch=13, Loss=2.057205, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1776672/epoch=13, Loss=2.490748, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1777312/epoch=13, Loss=2.324169, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1777952/epoch=13, Loss=2.321272, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1778592/epoch=13, Loss=2.132040, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1779200/epoch=13, Validation Accuracy=0.435268\n",
      "Iter=1779232/epoch=13, Loss=2.103663, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1779872/epoch=13, Loss=1.728341, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1780512/epoch=13, Loss=2.349241, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1781152/epoch=13, Loss=2.062026, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1781792/epoch=13, Loss=2.424800, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1782432/epoch=13, Loss=2.507471, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1783072/epoch=14, Loss=1.694760, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1783712/epoch=14, Loss=1.970868, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1784352/epoch=14, Loss=2.670508, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1784992/epoch=14, Loss=1.754182, Training Accuracy=0.562500, lr=0.000100\n",
      "\n",
      "Iter=1785600/epoch=14, Validation Accuracy=0.439732\n",
      "Iter=1785632/epoch=14, Loss=2.350018, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1786272/epoch=14, Loss=1.978303, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1786912/epoch=14, Loss=2.269833, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1787552/epoch=14, Loss=2.298070, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1788192/epoch=14, Loss=2.126567, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1788832/epoch=14, Loss=2.740981, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1789472/epoch=14, Loss=1.999588, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1790112/epoch=14, Loss=2.150920, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1790752/epoch=14, Loss=2.202737, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1791392/epoch=14, Loss=2.315800, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1792000/epoch=14, Validation Accuracy=0.439732\n",
      "Model saved in file at iteration 1792032: ./tf_densenet_model_iter56001.ckpt\n",
      "Iter=1792032/epoch=14, Loss=2.108458, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1792672/epoch=14, Loss=2.470693, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1793312/epoch=14, Loss=1.652929, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1793952/epoch=14, Loss=2.421848, Training Accuracy=0.281250, lr=0.000100\n",
      "Iter=1794592/epoch=14, Loss=2.106205, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1795232/epoch=14, Loss=2.112573, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1795872/epoch=14, Loss=1.724812, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1796512/epoch=14, Loss=2.072581, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1797152/epoch=14, Loss=2.094612, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1797792/epoch=14, Loss=2.306466, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1798400/epoch=14, Validation Accuracy=0.439732\n",
      "Iter=1798432/epoch=14, Loss=2.605653, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1799072/epoch=14, Loss=2.078801, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1799712/epoch=14, Loss=2.480072, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1800352/epoch=14, Loss=2.093231, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1800992/epoch=14, Loss=2.371948, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1801632/epoch=14, Loss=2.207170, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1802272/epoch=14, Loss=1.927149, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1802912/epoch=14, Loss=1.952185, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1803552/epoch=14, Loss=1.960009, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1804192/epoch=14, Loss=2.275917, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1804800/epoch=14, Validation Accuracy=0.422991\n",
      "Iter=1804832/epoch=14, Loss=2.377626, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1805472/epoch=14, Loss=1.759488, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1806112/epoch=14, Loss=2.730167, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1806752/epoch=14, Loss=2.281193, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1807392/epoch=14, Loss=2.111404, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1808032/epoch=14, Loss=1.735195, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1808672/epoch=14, Loss=1.846619, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1809312/epoch=14, Loss=2.441906, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1809952/epoch=14, Loss=1.564959, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1810592/epoch=14, Loss=2.106189, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1811200/epoch=14, Validation Accuracy=0.447545\n",
      "Iter=1811232/epoch=14, Loss=1.877964, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1811872/epoch=14, Loss=2.150213, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1812512/epoch=14, Loss=1.708906, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1813152/epoch=14, Loss=2.452080, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1813792/epoch=14, Loss=2.096209, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1814432/epoch=14, Loss=2.344285, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1815072/epoch=14, Loss=2.069189, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1815712/epoch=14, Loss=2.068225, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1816352/epoch=14, Loss=2.106174, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1816992/epoch=14, Loss=2.595858, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1817600/epoch=14, Validation Accuracy=0.450893\n",
      "Iter=1817632/epoch=14, Loss=2.096037, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1818272/epoch=14, Loss=2.562979, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1818912/epoch=14, Loss=2.446650, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1819552/epoch=14, Loss=2.285189, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1820192/epoch=14, Loss=2.089210, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1820832/epoch=14, Loss=2.001756, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1821472/epoch=14, Loss=2.924448, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1822112/epoch=14, Loss=2.085861, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1822752/epoch=14, Loss=1.672055, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1823392/epoch=14, Loss=2.186515, Training Accuracy=0.500000, lr=0.000100\n",
      "\n",
      "Iter=1824000/epoch=14, Validation Accuracy=0.441964\n",
      "Model saved in file at iteration 1824032: ./tf_densenet_model_iter57001.ckpt\n",
      "Iter=1824032/epoch=14, Loss=2.508690, Training Accuracy=0.187500, lr=0.000100\n",
      "Iter=1824672/epoch=14, Loss=2.117084, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1825312/epoch=14, Loss=2.452715, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1825952/epoch=14, Loss=2.477093, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1826592/epoch=14, Loss=2.036937, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1827232/epoch=14, Loss=2.501077, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1827872/epoch=14, Loss=1.870266, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1828512/epoch=14, Loss=2.141863, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1829152/epoch=14, Loss=1.936417, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1829792/epoch=14, Loss=2.075151, Training Accuracy=0.437500, lr=0.000100\n",
      "\n",
      "Iter=1830400/epoch=14, Validation Accuracy=0.438616\n",
      "Iter=1830432/epoch=14, Loss=1.893106, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1831072/epoch=14, Loss=2.384802, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1831712/epoch=14, Loss=2.154975, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1832352/epoch=14, Loss=2.265452, Training Accuracy=0.375000, lr=0.000100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter=1832992/epoch=14, Loss=1.812457, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1833632/epoch=14, Loss=2.419190, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1834272/epoch=14, Loss=2.134389, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1834912/epoch=14, Loss=1.854194, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1835552/epoch=14, Loss=2.357246, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1836192/epoch=14, Loss=2.188159, Training Accuracy=0.468750, lr=0.000100\n",
      "\n",
      "Iter=1836800/epoch=14, Validation Accuracy=0.439732\n",
      "Iter=1836832/epoch=14, Loss=1.800017, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1837472/epoch=14, Loss=2.186322, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1838112/epoch=14, Loss=1.914082, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1838752/epoch=14, Loss=1.648366, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1839392/epoch=14, Loss=2.076680, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1840032/epoch=14, Loss=2.097458, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1840672/epoch=14, Loss=1.855616, Training Accuracy=0.625000, lr=0.000100\n",
      "Iter=1841312/epoch=14, Loss=2.302958, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1841952/epoch=14, Loss=1.667224, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1842592/epoch=14, Loss=2.054458, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1843200/epoch=14, Validation Accuracy=0.449777\n",
      "Iter=1843232/epoch=14, Loss=1.638397, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1843872/epoch=14, Loss=2.009560, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1844512/epoch=14, Loss=2.245741, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1845152/epoch=14, Loss=1.964350, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1845792/epoch=14, Loss=2.205620, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1846432/epoch=14, Loss=1.727983, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1847072/epoch=14, Loss=2.546029, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1847712/epoch=14, Loss=2.081465, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1848352/epoch=14, Loss=1.711375, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1848992/epoch=14, Loss=1.773780, Training Accuracy=0.531250, lr=0.000100\n",
      "\n",
      "Iter=1849600/epoch=14, Validation Accuracy=0.444196\n",
      "Iter=1849632/epoch=14, Loss=2.119015, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1850272/epoch=14, Loss=2.649953, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1850912/epoch=14, Loss=1.875042, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1851552/epoch=14, Loss=2.137797, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1852192/epoch=14, Loss=2.114021, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1852832/epoch=14, Loss=2.725522, Training Accuracy=0.250000, lr=0.000100\n",
      "Iter=1853472/epoch=14, Loss=2.026319, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1854112/epoch=14, Loss=2.034854, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1854752/epoch=14, Loss=2.288017, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1855392/epoch=14, Loss=1.895196, Training Accuracy=0.406250, lr=0.000100\n",
      "\n",
      "Iter=1856000/epoch=14, Validation Accuracy=0.437500\n",
      "Model saved in file at iteration 1856032: ./tf_densenet_model_iter58001.ckpt\n",
      "Iter=1856032/epoch=14, Loss=1.895243, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1856672/epoch=14, Loss=2.719090, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1857312/epoch=14, Loss=1.757398, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1857952/epoch=14, Loss=1.746099, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1858592/epoch=14, Loss=2.208261, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1859232/epoch=14, Loss=1.824275, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1859872/epoch=14, Loss=2.349862, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1860512/epoch=14, Loss=2.496064, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1861152/epoch=14, Loss=1.606251, Training Accuracy=0.593750, lr=0.000100\n",
      "Iter=1861792/epoch=14, Loss=2.354206, Training Accuracy=0.375000, lr=0.000100\n",
      "\n",
      "Iter=1862400/epoch=14, Validation Accuracy=0.435268\n",
      "Iter=1862432/epoch=14, Loss=1.826603, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1863072/epoch=14, Loss=2.024496, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1863712/epoch=14, Loss=2.321797, Training Accuracy=0.406250, lr=0.000100\n",
      "Iter=1864352/epoch=14, Loss=1.628615, Training Accuracy=0.531250, lr=0.000100\n",
      "Iter=1864992/epoch=14, Loss=1.489515, Training Accuracy=0.562500, lr=0.000100\n",
      "Iter=1865632/epoch=14, Loss=2.276030, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1866272/epoch=14, Loss=2.066870, Training Accuracy=0.468750, lr=0.000100\n",
      "Iter=1866912/epoch=14, Loss=2.113498, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1867552/epoch=14, Loss=2.212448, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1868192/epoch=14, Loss=2.414884, Training Accuracy=0.343750, lr=0.000100\n",
      "\n",
      "Iter=1868800/epoch=14, Validation Accuracy=0.438616\n",
      "Iter=1868832/epoch=14, Loss=1.844159, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1869472/epoch=14, Loss=2.104906, Training Accuracy=0.437500, lr=0.000100\n",
      "Iter=1870112/epoch=14, Loss=2.086513, Training Accuracy=0.500000, lr=0.000100\n",
      "Iter=1870752/epoch=14, Loss=2.306700, Training Accuracy=0.343750, lr=0.000100\n",
      "Iter=1871392/epoch=14, Loss=2.312342, Training Accuracy=0.375000, lr=0.000100\n",
      "Iter=1872032/epoch=14, Loss=2.384843, Training Accuracy=0.312500, lr=0.000100\n",
      "Iter=1872672/epoch=14, Loss=2.304941, Training Accuracy=0.343750, lr=0.000100\n"
     ]
    }
   ],
   "source": [
    "saver = tf.train.Saver()\n",
    "init = tf.global_variables_initializer()\n",
    "sess.run(init)\n",
    "step = 0\n",
    "writer = tf.summary.FileWriter(\"log2\", sess.graph)\n",
    "summaries = tf.summary.merge_all()\n",
    "\n",
    "print(\"We are going to train the ImageNet model based on denseNet!!!\")\n",
    "while (step * batch_size) < max_iter:\n",
    "    epoch1=np.floor((step*batch_size)/glen1)\n",
    "    if (((step*batch_size)%glen1 < batch_size) & (lr==1e-3) & (epoch1 >2)):\n",
    "        lr /= 10\n",
    "\n",
    "    sess.run(optimizer,  feed_dict={learning_rate: lr})\n",
    "\n",
    "    if (step % 1000==1) & (step>1000):\n",
    "        save_path = saver.save(sess, \"./tf_densenet_model_iter\" + str(step) + \".ckpt\")\n",
    "        print(\"Model saved in file at iteration %d: %s\" % (step*batch_size,save_path))\n",
    "\n",
    "    if step>1 and step % display_step == 1:\n",
    "        # calculate the loss\n",
    "        loss, acc, summaries_string = sess.run([cost, accuracy, summaries])\n",
    "        print(\"Iter=%d/epoch=%d, Loss=%.6f, Training Accuracy=%.6f, lr=%f\" % (step*batch_size, epoch1 ,loss, acc,  lr))\n",
    "        writer.add_summary(summaries_string, step)\n",
    "        \n",
    "    if step>1 and (step % (display_step*10) == 0):\n",
    "        rounds = tlen1 // batch_size\n",
    "        \n",
    "        valacc=[]\n",
    "        for k in range(rounds):\n",
    "            acc= sess.run(accuracy2)\n",
    "            valacc.append(acc)\n",
    "        print(\"\\nIter=%d/epoch=%d, Validation Accuracy=%.6f\" % (step*batch_size, epoch1 , np.mean(valacc)))\n",
    "\n",
    "  \n",
    "    step += 1\n",
    "print(\"Optimization Finished!\")\n",
    "save_path = saver.save(sess, \"./tfdensenet_model.ckpt\")\n",
    "print(\"Model saved in file: %s\" % save_path)\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "id": "AOQOeejKzh0f"
   },
   "outputs": [],
   "source": [
    "exit()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "default_view": {},
   "name": "ResNet_init.ipynb",
   "provenance": [
    {
     "file_id": "1VL1sK4Hz_IsuhkyQRc8h2SD3w90OF1yp",
     "timestamp": 1526642217104
    }
   ],
   "version": "0.3.2",
   "views": {}
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}