One_Student_Temp_Loss_Weight.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# lambda * loss1 + (1 - lambda) * loss2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "from __future__ import absolute_import\n",
    "from __future__ import division\n",
    "from __future__ import print_function\n",
    "\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "\n",
    "tf.logging.set_verbosity(tf.logging.INFO)\n",
    "\n",
    "def cnn_model_fn(features, labels, mode):\n",
    "  \"\"\"Model function for CNN.\"\"\"\n",
    "  # Input Layer\n",
    "  # Reshape X to 4-D tensor: [batch_size, width, height, channels]\n",
    "  # MNIST images are 28x28 pixels, and have one color channel\n",
    "  input_layer = tf.reshape(features[\"x\"], [-1, 28, 28, 1])\n",
    "  print(\"_THIS_\",type(features))\n",
    "  # Convolutional Layer #1\n",
    "  # Computes 32 features using a 5x5 filter with ReLU activation.\n",
    "  # Padding is added to preserve width and height.\n",
    "  # Input Tensor Shape: [batch_size, 28, 28, 1]\n",
    "  # Output Tensor Shape: [batch_size, 28, 28, 32]\n",
    "  conv1 = tf.layers.conv2d(\n",
    "      inputs=input_layer,\n",
    "      filters=32,\n",
    "      kernel_size=[5, 5],\n",
    "      padding=\"same\",\n",
    "      activation=tf.nn.relu)\n",
    "\n",
    "  # Pooling Layer #1\n",
    "  # First max pooling layer with a 2x2 filter and stride of 2\n",
    "  # Input Tensor Shape: [batch_size, 28, 28, 32]\n",
    "  # Output Tensor Shape: [batch_size, 14, 14, 32]\n",
    "  pool1 = tf.layers.max_pooling2d(inputs=conv1, pool_size=[2, 2], strides=2)\n",
    "\n",
    "  # Convolutional Layer #2\n",
    "  # Computes 64 features using a 5x5 filter.\n",
    "  # Padding is added to preserve width and height.\n",
    "  # Input Tensor Shape: [batch_size, 14, 14, 32]\n",
    "  # Output Tensor Shape: [batch_size, 14, 14, 64]\n",
    "  conv2 = tf.layers.conv2d(\n",
    "      inputs=pool1,\n",
    "      filters=64,\n",
    "      kernel_size=[5, 5],\n",
    "      padding=\"same\",\n",
    "      activation=tf.nn.relu)\n",
    "\n",
    "  # Pooling Layer #2\n",
    "  # Second max pooling layer with a 2x2 filter and stride of 2\n",
    "  # Input Tensor Shape: [batch_size, 14, 14, 64]\n",
    "  # Output Tensor Shape: [batch_size, 7, 7, 64]\n",
    "  pool2 = tf.layers.max_pooling2d(inputs=conv2, pool_size=[2, 2], strides=2)\n",
    "\n",
    "  # Flatten tensor into a batch of vectors\n",
    "  # Input Tensor Shape: [batch_size, 7, 7, 64]\n",
    "  # Output Tensor Shape: [batch_size, 7 * 7 * 64]\n",
    "  pool2_flat = tf.reshape(pool2, [-1, 7 * 7 * 64])\n",
    "\n",
    "  # Dense Layer\n",
    "  # Densely connected layer with 1024 neurons\n",
    "  # Input Tensor Shape: [batch_size, 7 * 7 * 64]\n",
    "  # Output Tensor Shape: [batch_size, 1024]\n",
    "  dense = tf.layers.dense(inputs=pool2_flat, units=1024, activation=tf.nn.relu)\n",
    "\n",
    "  # Add dropout operation; 0.6 probability that element will be kept\n",
    "  dropout = tf.layers.dropout(\n",
    "      inputs=dense, rate=0.4, training=mode == tf.estimator.ModeKeys.TRAIN)\n",
    "\n",
    "  # Logits layer\n",
    "  # Input Tensor Shape: [batch_size, 1024]\n",
    "  # Output Tensor Shape: [batch_size, 10]\n",
    "  logits = tf.layers.dense(inputs=dropout, units=10)\n",
    "\n",
    "  predictions = {\n",
    "      # Generate predictions (for PREDICT and EVAL mode)\n",
    "      \"classes\": tf.argmax(input=logits, axis=1),\n",
    "      # Add `softmax_tensor` to the graph. It is used for PREDICT and by the\n",
    "      # `logging_hook`.\n",
    "      \"probabilities\": tf.nn.softmax(logits/temprature, name=\"softmax_tensor\")\n",
    "  }\n",
    "  if mode == tf.estimator.ModeKeys.PREDICT:\n",
    "    return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions)\n",
    "\n",
    "  # Calculate Loss (for both TRAIN and EVAL modes)\n",
    "  loss = tf.losses.sparse_softmax_cross_entropy(labels=labels, logits=logits)\n",
    "\n",
    "  # Configure the Training Op (for TRAIN mode)\n",
    "  if mode == tf.estimator.ModeKeys.TRAIN:\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.001)\n",
    "    train_op = optimizer.minimize(\n",
    "        loss=loss,\n",
    "        global_step=tf.train.get_global_step())\n",
    "    return tf.estimator.EstimatorSpec(mode=mode, loss=loss, train_op=train_op)\n",
    "\n",
    "  # Add evaluation metrics (for EVAL mode)\n",
    "  eval_metric_ops = {\n",
    "      \"accuracy\": tf.metrics.accuracy(\n",
    "          labels=labels, predictions=predictions[\"classes\"]),\n",
    "  }\n",
    "  return tf.estimator.EstimatorSpec(\n",
    "      mode=mode, loss=loss, eval_metric_ops=eval_metric_ops)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#cross entropy between two logits with temperature applied\n",
    "#returns a 1d array with loss for each row in the inputs\n",
    "def cross_entropy2(tar_soft_t, pred_logits):\n",
    "   pred_soft_t = tf.nn.softmax(pred_logits/temprature)\n",
    "#    tar_soft = tf.nn.softmax(tar_logits_t/temprature)\n",
    "   pred_log = -1 * tf.log(pred_soft_t)\n",
    "   product =tf.multiply(tar_soft_t,pred_log)\n",
    "#    print(\"Swapnil\", pred_log,tar_soft_t,product)\n",
    "   return tf.reduce_mean(product)\n",
    "    \n",
    "def custom_loss(y_true, pred_logits, tar_soft_t):\n",
    "#    print(tf.losses.sparse_softmax_cross_entropy(labels = y_true, logits = pred_logits))\n",
    "#    print(l * cross_entropy2(tar_soft_t, pred_logits_t))\n",
    "#    return loss_weight * tf.losses.sparse_softmax_cross_entropy(labels = y_true, logits = pred_logits) + (1-loss_weight) * cross_entropy2(tar_soft_t, pred_logits_t)\n",
    "  return  loss_weight * tf.losses.sparse_softmax_cross_entropy(labels = y_true, logits = pred_logits)+(1-loss_weight)*cross_entropy2(tar_soft_t,pred_logits)\n",
    "#     one_hot=tf.one_hot(y_true,10)\n",
    "#     sess2=tf.InteractiveSession()\n",
    "#     tf.train.start_queue_runners(sess2)\n",
    "#     print(\"S_\",tar_soft_t.eval())\n",
    "#     print(\"T_\",pred_soft_t.eval())\n",
    "#     sess2.close()\n",
    "#     return tf.losses.mean_squared_error(labels=tar_soft_t,predictions=pred_soft_t)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def getFilterData(f,l):\n",
    "  sess=tf.InteractiveSession()\n",
    "  tf.train.start_queue_runners(sess)\n",
    "  data_s=f['x'].eval()\n",
    "  out_s=l.eval()\n",
    "  sess.close()\n",
    "  return data_s,out_s\n",
    "\n",
    "\n",
    "def student_model_fn(features, labels, mode):\n",
    "\n",
    "  print(features,labels)\n",
    "  data_swap,out_swap=getFilterData(features,labels)\n",
    "  # Input Layer\n",
    "  # Reshape X to 4-D tensor: [batch_size, width, height, channels]\n",
    "  # MNIST images are 28x28 pixels, and have one color channel\n",
    "  input_layer = tf.reshape(features[\"x\"], [-1, 28, 28, 1])\n",
    "  # Convolutional Layer #1\n",
    "  # Computes 32 features using a 5x5 filter with ReLU activation.\n",
    "  # Padding is added to preserve width and height.\n",
    "  # Input Tensor Shape: [batch_size, 28, 28, 1]\n",
    "  # Output Tensor Shape: [batch_size, 28, 28, 32]\n",
    "  if mode == tf.estimator.ModeKeys.TRAIN:\n",
    "      eval_teacher_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "          x={\"x\":data_swap},\n",
    "          y=out_swap,\n",
    "          batch_size=100,\n",
    "          shuffle=False)\n",
    "\n",
    "      eval_teacher=mnist_classifier.evaluate(input_fn=eval_teacher_fn)\n",
    "      outlog.write('%f;' % eval_teacher['accuracy'])\n",
    "      outlog.write('%f;' % eval_teacher['loss'])\n",
    "      outlog.write('\\n')\n",
    "      predictions=mnist_classifier.predict(input_fn=eval_teacher_fn)\n",
    "      teacher_pred=list(predictions)\n",
    "      teacher_soft=[ p['probabilities'] for p in teacher_pred]\n",
    "      temp=np.array(teacher_soft)\n",
    "      \n",
    "      teacher_soft_t=tf.convert_to_tensor(temp)\n",
    "\n",
    "\n",
    "  conv1 = tf.layers.conv2d(\n",
    "      inputs=input_layer,\n",
    "      filters=8,\n",
    "      kernel_size=[5, 5],\n",
    "      padding=\"same\",\n",
    "      activation=tf.nn.relu)\n",
    "\n",
    "  # Pooling Layer #1\n",
    "  # First max pooling layer with a 2x2 filter and stride of 2\n",
    "  # Input Tensor Shape: [batch_size, 28, 28, 32]\n",
    "  # Output Tensor Shape: [batch_size, 14, 14, 32]\n",
    "  pool1 = tf.layers.max_pooling2d(inputs=conv1, pool_size=[2, 2], strides=2)\n",
    "\n",
    "  # Convolutional Layer #2\n",
    "  # Computes 64 features using a 5x5 filter.\n",
    "  # Padding is added to preserve width and height.\n",
    "  # Input Tensor Shape: [batch_size, 14, 14, 32]\n",
    "  # Output Tensor Shape: [batch_size, 14, 14, 64]\n",
    "  conv2 = tf.layers.conv2d(\n",
    "      inputs=pool1,\n",
    "      filters=8,\n",
    "      kernel_size=[5, 5],\n",
    "      padding=\"same\",\n",
    "      activation=tf.nn.relu)\n",
    "\n",
    "  # Pooling Layer #2\n",
    "  # Second max pooling layer with a 2x2 filter and stride of 2\n",
    "  # Input Tensor Shape: [batch_size, 14, 14, 64]\n",
    "  # Output Tensor Shape: [batch_size, 7, 7, 64]\n",
    "  pool2 = tf.layers.max_pooling2d(inputs=conv2, pool_size=[2, 2], strides=2)\n",
    "\n",
    "  # Flatten tensor into a batch of vectors\n",
    "  # Input Tensor Shape: [batch_size, 7, 7, 64]\n",
    "  # Output Tensor Shape: [batch_size, 7 * 7 * 64]\n",
    "  pool2_flat = tf.reshape(pool2, [-1, 7 * 7 * 8])\n",
    "\n",
    "  # Dense Layer\n",
    "  # Densely connected layer with 1024 neurons\n",
    "  # Input Tensor Shape: [batch_size, 7 * 7 * 64]\n",
    "  # Output Tensor Shape: [batch_size, 1024]\n",
    "  dense = tf.layers.dense(inputs=pool2_flat, units=256, activation=tf.nn.relu)\n",
    "\n",
    "  # Add dropout operation; 0.6 probability that element will be kept\n",
    "  dropout = tf.layers.dropout(\n",
    "      inputs=dense, rate=0.4, training=mode == tf.estimator.ModeKeys.TRAIN)\n",
    "\n",
    "  # Logits layer\n",
    "  # Input Tensor Shape: [batch_size, 1024]\n",
    "  # Output Tensor Shape: [batch_size, 10]\n",
    "  logits = tf.layers.dense(inputs=dropout, units=10)\n",
    "\n",
    "  predictions = {\n",
    "      # Generate predictions (for PREDICT and EVAL mode)\n",
    "      \"classes\": tf.argmax(input=logits, axis=1),\n",
    "      # Add `softmax_tensor` to the graph. It is used for PREDICT and by the\n",
    "      # `logging_hook`.\n",
    "      \"probabilities\": tf.nn.softmax(logits, name=\"softmax_tensor\")\n",
    "  }\n",
    "  if mode == tf.estimator.ModeKeys.PREDICT:\n",
    "    return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions)\n",
    "\n",
    "  # Calculate Loss (for both TRAIN and EVAL modes)\n",
    "  #   loss = tf.losses.sparse_softmax_cross_entropy(labels=labels, logits=logits)\n",
    "  \n",
    "    \n",
    "#   student_pred=predictions['probabilities']\n",
    "#   sess1=tf.InteractiveSession()\n",
    "#   tf.train.start_queue_runners(sess1)\n",
    "#   student_soft=student_pred.eval()\n",
    "#   sess1.close()  \n",
    "  if mode == tf.estimator.ModeKeys.TRAIN:\n",
    "#     pred_soft_t=tf.nn.softmax(logits, name=\"softmax_tensor\")\n",
    "#     print('a',labels,'b', logits,'c', teacher_soft_t ,'d', pred_logits_t)\n",
    "    loss=custom_loss(labels, logits,teacher_soft_t)\n",
    "  else:\n",
    "    loss = tf.losses.sparse_softmax_cross_entropy(labels=labels, logits=logits)\n",
    "  \n",
    "    \n",
    "  # Configure the Training Op (for TRAIN mode)\n",
    "  if mode == tf.estimator.ModeKeys.TRAIN:\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.001)\n",
    "    train_op = optimizer.minimize(\n",
    "        loss=loss,\n",
    "        global_step=tf.train.get_global_step())\n",
    "    return tf.estimator.EstimatorSpec(mode=mode, loss=loss, train_op=train_op)\n",
    "\n",
    "  # Add evaluation metrics (for EVAL mode)\n",
    "  eval_metric_ops = {\n",
    "      \"accuracy\": tf.metrics.accuracy(\n",
    "          labels=labels, predictions=predictions[\"classes\"]),\n",
    "  }\n",
    "  return tf.estimator.EstimatorSpec(\n",
    "      mode=mode, loss=loss, eval_metric_ops=eval_metric_ops)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259779c390>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': '/tmp/mnist_convnet_model_test1', '_save_summary_steps': 100}\n",
      "Extracting MNIST-data/train-images-idx3-ubyte.gz\n",
      "Extracting MNIST-data/train-labels-idx1-ubyte.gz\n",
      "Extracting MNIST-data/t10k-images-idx3-ubyte.gz\n",
      "Extracting MNIST-data/t10k-labels-idx1-ubyte.gz\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f2596dea910>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_1_w_0', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:48:43\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:48:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.97, global_step = 20000, loss = 0.09232447\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0/model.ckpt-10020\n",
      "INFO:tensorflow:Saving checkpoints for 10021 into ./mnist_convnet_student_mod_t_1_w_0/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.19398823, step = 10021\n",
      "INFO:tensorflow:global_step/sec: 167.04\n",
      "INFO:tensorflow:loss = 0.1987425, step = 10121 (0.598 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.579\n",
      "INFO:tensorflow:loss = 0.1951468, step = 10221 (0.539 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.789\n",
      "INFO:tensorflow:loss = 0.19410446, step = 10321 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.192\n",
      "INFO:tensorflow:loss = 0.19487633, step = 10421 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.241\n",
      "INFO:tensorflow:loss = 0.1955357, step = 10521 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.281\n",
      "INFO:tensorflow:loss = 0.19140543, step = 10621 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.04\n",
      "INFO:tensorflow:loss = 0.18959509, step = 10721 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.046\n",
      "INFO:tensorflow:loss = 0.18665151, step = 10821 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.181\n",
      "INFO:tensorflow:loss = 0.1844777, step = 10921 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.384\n",
      "INFO:tensorflow:loss = 0.19076332, step = 11021 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.99\n",
      "INFO:tensorflow:loss = 0.18424754, step = 11121 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 177.417\n",
      "INFO:tensorflow:loss = 0.18396184, step = 11221 (0.558 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.271\n",
      "INFO:tensorflow:loss = 0.18703231, step = 11321 (0.528 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.084\n",
      "INFO:tensorflow:loss = 0.18450232, step = 11421 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.719\n",
      "INFO:tensorflow:loss = 0.18378916, step = 11521 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.258\n",
      "INFO:tensorflow:loss = 0.1814044, step = 11621 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.372\n",
      "INFO:tensorflow:loss = 0.1754249, step = 11721 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.919\n",
      "INFO:tensorflow:loss = 0.17631948, step = 11821 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.296\n",
      "INFO:tensorflow:loss = 0.17804728, step = 11921 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.18\n",
      "INFO:tensorflow:loss = 0.18235749, step = 12021 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.873\n",
      "INFO:tensorflow:loss = 0.17650278, step = 12121 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.149\n",
      "INFO:tensorflow:loss = 0.17260483, step = 12221 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.345\n",
      "INFO:tensorflow:loss = 0.17551845, step = 12321 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.067\n",
      "INFO:tensorflow:loss = 0.17242403, step = 12421 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.172\n",
      "INFO:tensorflow:loss = 0.16793816, step = 12521 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.108\n",
      "INFO:tensorflow:loss = 0.17531049, step = 12621 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.724\n",
      "INFO:tensorflow:loss = 0.17175491, step = 12721 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.374\n",
      "INFO:tensorflow:loss = 0.16483389, step = 12821 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.038\n",
      "INFO:tensorflow:loss = 0.167335, step = 12921 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.547\n",
      "INFO:tensorflow:loss = 0.16727018, step = 13021 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.448\n",
      "INFO:tensorflow:loss = 0.16824523, step = 13121 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.573\n",
      "INFO:tensorflow:loss = 0.16318339, step = 13221 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.965\n",
      "INFO:tensorflow:loss = 0.16338886, step = 13321 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.919\n",
      "INFO:tensorflow:loss = 0.16123928, step = 13421 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.506\n",
      "INFO:tensorflow:loss = 0.16111875, step = 13521 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.142\n",
      "INFO:tensorflow:loss = 0.16225067, step = 13621 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.669\n",
      "INFO:tensorflow:loss = 0.15708558, step = 13721 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.753\n",
      "INFO:tensorflow:loss = 0.16710617, step = 13821 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.533\n",
      "INFO:tensorflow:loss = 0.16094604, step = 13921 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.232\n",
      "INFO:tensorflow:loss = 0.15983906, step = 14021 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.354\n",
      "INFO:tensorflow:loss = 0.15933946, step = 14121 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.321\n",
      "INFO:tensorflow:loss = 0.1573923, step = 14221 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.3\n",
      "INFO:tensorflow:loss = 0.1474517, step = 14321 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.302\n",
      "INFO:tensorflow:loss = 0.15153055, step = 14421 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.036\n",
      "INFO:tensorflow:loss = 0.1500033, step = 14521 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.635\n",
      "INFO:tensorflow:loss = 0.14352494, step = 14621 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.341\n",
      "INFO:tensorflow:loss = 0.14691484, step = 14721 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.223\n",
      "INFO:tensorflow:loss = 0.15280002, step = 14821 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.793\n",
      "INFO:tensorflow:loss = 0.1514114, step = 14921 (0.511 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15020 into ./mnist_convnet_student_mod_t_1_w_0/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.14711124.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:49:10\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0/model.ckpt-15020\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1073:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:49:12\n",
      "INFO:tensorflow:Saving dict for global step 15020: accuracy = 0.5368207, global_step = 15020, loss = 1.5864601\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:49:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0/model.ckpt-15020\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:49:13\n",
      "INFO:tensorflow:Saving dict for global step 15020: accuracy = 0.5261023, global_step = 15020, loss = 1.579772\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f25977ebc90>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_1_w_0.25', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:49:13\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:49:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.96, global_step = 20000, loss = 0.17090626\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.25/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_1_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.085419044, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 175.083\n",
      "INFO:tensorflow:loss = 0.10168868, step = 10101 (0.572 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.815\n",
      "INFO:tensorflow:loss = 0.09136477, step = 10201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.673\n",
      "INFO:tensorflow:loss = 0.10917812, step = 10301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.825\n",
      "INFO:tensorflow:loss = 0.10660365, step = 10401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.861\n",
      "INFO:tensorflow:loss = 0.09908785, step = 10501 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.405\n",
      "INFO:tensorflow:loss = 0.09895761, step = 10601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.298\n",
      "INFO:tensorflow:loss = 0.097034395, step = 10701 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.411\n",
      "INFO:tensorflow:loss = 0.084896445, step = 10801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.673\n",
      "INFO:tensorflow:loss = 0.08662553, step = 10901 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.815\n",
      "INFO:tensorflow:loss = 0.088229015, step = 11001 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.226\n",
      "INFO:tensorflow:loss = 0.08562848, step = 11101 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.059\n",
      "INFO:tensorflow:loss = 0.100398555, step = 11201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.199\n",
      "INFO:tensorflow:loss = 0.09110738, step = 11301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.505\n",
      "INFO:tensorflow:loss = 0.07138172, step = 11401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.756\n",
      "INFO:tensorflow:loss = 0.08646821, step = 11501 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 186.966\n",
      "INFO:tensorflow:loss = 0.07591409, step = 11601 (0.535 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.617\n",
      "INFO:tensorflow:loss = 0.08543021, step = 11701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.852\n",
      "INFO:tensorflow:loss = 0.07586555, step = 11801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.933\n",
      "INFO:tensorflow:loss = 0.08778496, step = 11901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.648\n",
      "INFO:tensorflow:loss = 0.0712606, step = 12001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.579\n",
      "INFO:tensorflow:loss = 0.08352283, step = 12101 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.814\n",
      "INFO:tensorflow:loss = 0.08483952, step = 12201 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.225\n",
      "INFO:tensorflow:loss = 0.086542346, step = 12301 (0.528 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.823\n",
      "INFO:tensorflow:loss = 0.07492749, step = 12401 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.714\n",
      "INFO:tensorflow:loss = 0.07605912, step = 12501 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.853\n",
      "INFO:tensorflow:loss = 0.07410201, step = 12601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.526\n",
      "INFO:tensorflow:loss = 0.06950593, step = 12701 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.274\n",
      "INFO:tensorflow:loss = 0.06612766, step = 12801 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.182\n",
      "INFO:tensorflow:loss = 0.07839294, step = 12901 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.435\n",
      "INFO:tensorflow:loss = 0.0787436, step = 13001 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.748\n",
      "INFO:tensorflow:loss = 0.069014564, step = 13101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.293\n",
      "INFO:tensorflow:loss = 0.06783873, step = 13201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.284\n",
      "INFO:tensorflow:loss = 0.06785157, step = 13301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.951\n",
      "INFO:tensorflow:loss = 0.07304537, step = 13401 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.609\n",
      "INFO:tensorflow:loss = 0.063855514, step = 13501 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.81\n",
      "INFO:tensorflow:loss = 0.074826404, step = 13601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.191\n",
      "INFO:tensorflow:loss = 0.074860096, step = 13701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.951\n",
      "INFO:tensorflow:loss = 0.077660814, step = 13801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.999\n",
      "INFO:tensorflow:loss = 0.07812798, step = 13901 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.534\n",
      "INFO:tensorflow:loss = 0.07996583, step = 14001 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.57\n",
      "INFO:tensorflow:loss = 0.066004544, step = 14101 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.564\n",
      "INFO:tensorflow:loss = 0.06346787, step = 14201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.408\n",
      "INFO:tensorflow:loss = 0.060134854, step = 14301 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.691\n",
      "INFO:tensorflow:loss = 0.061296664, step = 14401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.805\n",
      "INFO:tensorflow:loss = 0.07096471, step = 14501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.347\n",
      "INFO:tensorflow:loss = 0.07009504, step = 14601 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.115\n",
      "INFO:tensorflow:loss = 0.056405324, step = 14701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.054\n",
      "INFO:tensorflow:loss = 0.056214724, step = 14801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.819\n",
      "INFO:tensorflow:loss = 0.071995735, step = 14901 (0.513 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_1_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.06231386.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:49:40\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:49:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7589168, global_step = 15000, loss = 0.8178701\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:49:42\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.25/model.ckpt-15000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1086:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:49:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7688228, global_step = 15000, loss = 0.7950935\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259a9a5750>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_1_w_0.5', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:49:43\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:49:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.98, global_step = 20000, loss = 0.09350607\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.5/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_1_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.04598814, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 171.785\n",
      "INFO:tensorflow:loss = 0.062028185, step = 10101 (0.583 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.958\n",
      "INFO:tensorflow:loss = 0.045925662, step = 10201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.906\n",
      "INFO:tensorflow:loss = 0.045128886, step = 10301 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.095\n",
      "INFO:tensorflow:loss = 0.045122012, step = 10401 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.931\n",
      "INFO:tensorflow:loss = 0.042792518, step = 10501 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.985\n",
      "INFO:tensorflow:loss = 0.05014988, step = 10601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.941\n",
      "INFO:tensorflow:loss = 0.045345843, step = 10701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.621\n",
      "INFO:tensorflow:loss = 0.04294441, step = 10801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.265\n",
      "INFO:tensorflow:loss = 0.06372536, step = 10901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.922\n",
      "INFO:tensorflow:loss = 0.043695167, step = 11001 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.271\n",
      "INFO:tensorflow:loss = 0.07192924, step = 11101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.897\n",
      "INFO:tensorflow:loss = 0.050533734, step = 11201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.016\n",
      "INFO:tensorflow:loss = 0.056846514, step = 11301 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.464\n",
      "INFO:tensorflow:loss = 0.042438664, step = 11401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.511\n",
      "INFO:tensorflow:loss = 0.042844903, step = 11501 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.139\n",
      "INFO:tensorflow:loss = 0.051221747, step = 11601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.065\n",
      "INFO:tensorflow:loss = 0.039683595, step = 11701 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.583\n",
      "INFO:tensorflow:loss = 0.04035174, step = 11801 (0.525 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.35\n",
      "INFO:tensorflow:loss = 0.039960727, step = 11901 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.191\n",
      "INFO:tensorflow:loss = 0.039399777, step = 12001 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.428\n",
      "INFO:tensorflow:loss = 0.04184037, step = 12101 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.111\n",
      "INFO:tensorflow:loss = 0.049494706, step = 12201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.729\n",
      "INFO:tensorflow:loss = 0.041263163, step = 12301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.756\n",
      "INFO:tensorflow:loss = 0.04072638, step = 12401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.538\n",
      "INFO:tensorflow:loss = 0.037041314, step = 12501 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.863\n",
      "INFO:tensorflow:loss = 0.040748328, step = 12601 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.45\n",
      "INFO:tensorflow:loss = 0.035221774, step = 12701 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.454\n",
      "INFO:tensorflow:loss = 0.045820475, step = 12801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.44\n",
      "INFO:tensorflow:loss = 0.039090663, step = 12901 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.771\n",
      "INFO:tensorflow:loss = 0.0395632, step = 13001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.395\n",
      "INFO:tensorflow:loss = 0.037789732, step = 13101 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.967\n",
      "INFO:tensorflow:loss = 0.028507374, step = 13201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.695\n",
      "INFO:tensorflow:loss = 0.026791565, step = 13301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.438\n",
      "INFO:tensorflow:loss = 0.04059595, step = 13401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.073\n",
      "INFO:tensorflow:loss = 0.036292948, step = 13501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.629\n",
      "INFO:tensorflow:loss = 0.029706996, step = 13601 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.91\n",
      "INFO:tensorflow:loss = 0.03626395, step = 13701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.466\n",
      "INFO:tensorflow:loss = 0.033601575, step = 13801 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.645\n",
      "INFO:tensorflow:loss = 0.03824542, step = 13901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 181.348\n",
      "INFO:tensorflow:loss = 0.034694646, step = 14001 (0.551 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.677\n",
      "INFO:tensorflow:loss = 0.03867152, step = 14101 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.291\n",
      "INFO:tensorflow:loss = 0.028897684, step = 14201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.29\n",
      "INFO:tensorflow:loss = 0.026496863, step = 14301 (0.506 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.136\n",
      "INFO:tensorflow:loss = 0.030768776, step = 14401 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.108\n",
      "INFO:tensorflow:loss = 0.030050296, step = 14501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.084\n",
      "INFO:tensorflow:loss = 0.029914696, step = 14601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.937\n",
      "INFO:tensorflow:loss = 0.02959666, step = 14701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.837\n",
      "INFO:tensorflow:loss = 0.035775304, step = 14801 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.659\n",
      "INFO:tensorflow:loss = 0.02984367, step = 14901 (0.514 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_1_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.030072035.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1093:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:50:10\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.5/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:50:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.75503516, global_step = 15000, loss = 0.95766884\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:50:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.5/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:50:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.76320714, global_step = 15000, loss = 0.9395068\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f2597751810>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_1_w_0.75', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:50:12\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:50:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.96, global_step = 20000, loss = 0.10454481\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.75/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_1_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.028023131, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 169.496\n",
      "INFO:tensorflow:loss = 0.036143497, step = 10101 (0.596 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.04\n",
      "INFO:tensorflow:loss = 0.041561984, step = 10201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.823\n",
      "INFO:tensorflow:loss = 0.026761634, step = 10301 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.142\n",
      "INFO:tensorflow:loss = 0.02836914, step = 10401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.148\n",
      "INFO:tensorflow:loss = 0.03235422, step = 10501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.225\n",
      "INFO:tensorflow:loss = 0.027083304, step = 10601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.171\n",
      "INFO:tensorflow:loss = 0.020047802, step = 10701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.433\n",
      "INFO:tensorflow:loss = 0.028279025, step = 10801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.075\n",
      "INFO:tensorflow:loss = 0.04465786, step = 10901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.097\n",
      "INFO:tensorflow:loss = 0.034880716, step = 11001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.267\n",
      "INFO:tensorflow:loss = 0.028439252, step = 11101 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.896\n",
      "INFO:tensorflow:loss = 0.02858825, step = 11201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.787\n",
      "INFO:tensorflow:loss = 0.030178927, step = 11301 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.679\n",
      "INFO:tensorflow:loss = 0.036760982, step = 11401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.63\n",
      "INFO:tensorflow:loss = 0.035045676, step = 11501 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.759\n",
      "INFO:tensorflow:loss = 0.027393164, step = 11601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.522\n",
      "INFO:tensorflow:loss = 0.023953266, step = 11701 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.385\n",
      "INFO:tensorflow:loss = 0.019870903, step = 11801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.11\n",
      "INFO:tensorflow:loss = 0.02293478, step = 11901 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.431\n",
      "INFO:tensorflow:loss = 0.041897472, step = 12001 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.436\n",
      "INFO:tensorflow:loss = 0.026123129, step = 12101 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.83\n",
      "INFO:tensorflow:loss = 0.034020036, step = 12201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.237\n",
      "INFO:tensorflow:loss = 0.026115436, step = 12301 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.557\n",
      "INFO:tensorflow:loss = 0.02449587, step = 12401 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.627\n",
      "INFO:tensorflow:loss = 0.028359644, step = 12501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.656\n",
      "INFO:tensorflow:loss = 0.027916992, step = 12601 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.535\n",
      "INFO:tensorflow:loss = 0.02036189, step = 12701 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.957\n",
      "INFO:tensorflow:loss = 0.023799907, step = 12801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.539\n",
      "INFO:tensorflow:loss = 0.023999326, step = 12901 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.382\n",
      "INFO:tensorflow:loss = 0.021430809, step = 13001 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.625\n",
      "INFO:tensorflow:loss = 0.01962875, step = 13101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.627\n",
      "INFO:tensorflow:loss = 0.031247593, step = 13201 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.793\n",
      "INFO:tensorflow:loss = 0.02423081, step = 13301 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.669\n",
      "INFO:tensorflow:loss = 0.018689215, step = 13401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.807\n",
      "INFO:tensorflow:loss = 0.03348569, step = 13501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.123\n",
      "INFO:tensorflow:loss = 0.03375595, step = 13601 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.762\n",
      "INFO:tensorflow:loss = 0.02385984, step = 13701 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.139\n",
      "INFO:tensorflow:loss = 0.019468468, step = 13801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.57\n",
      "INFO:tensorflow:loss = 0.018486954, step = 13901 (0.531 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.868\n",
      "INFO:tensorflow:loss = 0.028486453, step = 14001 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.282\n",
      "INFO:tensorflow:loss = 0.020491226, step = 14101 (0.530 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.416\n",
      "INFO:tensorflow:loss = 0.0297768, step = 14201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.583\n",
      "INFO:tensorflow:loss = 0.023954941, step = 14301 (0.533 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.477\n",
      "INFO:tensorflow:loss = 0.017585585, step = 14401 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.974\n",
      "INFO:tensorflow:loss = 0.019068185, step = 14501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.623\n",
      "INFO:tensorflow:loss = 0.022787748, step = 14601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.502\n",
      "INFO:tensorflow:loss = 0.022283059, step = 14701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.467\n",
      "INFO:tensorflow:loss = 0.017775178, step = 14801 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.109\n",
      "INFO:tensorflow:loss = 0.020824134, step = 14901 (0.520 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_1_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.021539059.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:50:40\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1103:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:50:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7696463, global_step = 15000, loss = 0.9567322\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:50:42\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:50:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7829659, global_step = 15000, loss = 0.9030881\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259a9ad6d0>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_1_w_1', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:50:42\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:50:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 1.0, global_step = 20000, loss = 0.04139505\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_1/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_1_w_1/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.009364049, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 157.322\n",
      "INFO:tensorflow:loss = 0.009907476, step = 10101 (0.636 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.796\n",
      "INFO:tensorflow:loss = 0.025256416, step = 10201 (0.532 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.679\n",
      "INFO:tensorflow:loss = 0.009372017, step = 10301 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.916\n",
      "INFO:tensorflow:loss = 0.015690176, step = 10401 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.411\n",
      "INFO:tensorflow:loss = 0.015124615, step = 10501 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 182.515\n",
      "INFO:tensorflow:loss = 0.018405372, step = 10601 (0.547 sec)\n",
      "INFO:tensorflow:global_step/sec: 181.741\n",
      "INFO:tensorflow:loss = 0.008470565, step = 10701 (0.551 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.103\n",
      "INFO:tensorflow:loss = 0.014994104, step = 10801 (0.533 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.209\n",
      "INFO:tensorflow:loss = 0.00762406, step = 10901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.782\n",
      "INFO:tensorflow:loss = 0.012816416, step = 11001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.654\n",
      "INFO:tensorflow:loss = 0.011575109, step = 11101 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.99\n",
      "INFO:tensorflow:loss = 0.026720962, step = 11201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.62\n",
      "INFO:tensorflow:loss = 0.013203171, step = 11301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.468\n",
      "INFO:tensorflow:loss = 0.014477697, step = 11401 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.529\n",
      "INFO:tensorflow:loss = 0.010196055, step = 11501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.909\n",
      "INFO:tensorflow:loss = 0.017310334, step = 11601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.711\n",
      "INFO:tensorflow:loss = 0.007922872, step = 11701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.89\n",
      "INFO:tensorflow:loss = 0.012603415, step = 11801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.895\n",
      "INFO:tensorflow:loss = 0.0063970396, step = 11901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.931\n",
      "INFO:tensorflow:loss = 0.021343185, step = 12001 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.519\n",
      "INFO:tensorflow:loss = 0.018848091, step = 12101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.007\n",
      "INFO:tensorflow:loss = 0.015065355, step = 12201 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.065\n",
      "INFO:tensorflow:loss = 0.01383496, step = 12301 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.997\n",
      "INFO:tensorflow:loss = 0.007010012, step = 12401 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.901\n",
      "INFO:tensorflow:loss = 0.012975062, step = 12501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.01\n",
      "INFO:tensorflow:loss = 0.0077166385, step = 12601 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.757\n",
      "INFO:tensorflow:loss = 0.010194306, step = 12701 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.437\n",
      "INFO:tensorflow:loss = 0.009730759, step = 12801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.427\n",
      "INFO:tensorflow:loss = 0.008755504, step = 12901 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.11\n",
      "INFO:tensorflow:loss = 0.018535938, step = 13001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.552\n",
      "INFO:tensorflow:loss = 0.015727, step = 13101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.955\n",
      "INFO:tensorflow:loss = 0.015566425, step = 13201 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.834\n",
      "INFO:tensorflow:loss = 0.009982677, step = 13301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.314\n",
      "INFO:tensorflow:loss = 0.026937578, step = 13401 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.24\n",
      "INFO:tensorflow:loss = 0.0060207522, step = 13501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.849\n",
      "INFO:tensorflow:loss = 0.010530978, step = 13601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.455\n",
      "INFO:tensorflow:loss = 0.025096979, step = 13701 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.441\n",
      "INFO:tensorflow:loss = 0.0039074128, step = 13801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.72\n",
      "INFO:tensorflow:loss = 0.007928337, step = 13901 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.951\n",
      "INFO:tensorflow:loss = 0.008324084, step = 14001 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.996\n",
      "INFO:tensorflow:loss = 0.009884636, step = 14101 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.593\n",
      "INFO:tensorflow:loss = 0.009679354, step = 14201 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.099\n",
      "INFO:tensorflow:loss = 0.01053746, step = 14301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.227\n",
      "INFO:tensorflow:loss = 0.008898196, step = 14401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.69\n",
      "INFO:tensorflow:loss = 0.015050705, step = 14501 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.215\n",
      "INFO:tensorflow:loss = 0.017301746, step = 14601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.266\n",
      "INFO:tensorflow:loss = 0.011064504, step = 14701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.903\n",
      "INFO:tensorflow:loss = 0.011580366, step = 14801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.234\n",
      "INFO:tensorflow:loss = 0.007287562, step = 14901 (0.512 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_1_w_1/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.009046206.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:51:10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1113:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:51:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7829208, global_step = 15000, loss = 0.97952604\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:51:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_1_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:51:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7896215, global_step = 15000, loss = 0.9558685\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f2597786d10>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_2_w_0', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:51:12\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:51:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.97, global_step = 20000, loss = 0.08945859\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_2_w_0/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.22744296, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 168.333\n",
      "INFO:tensorflow:loss = 0.22781067, step = 10101 (0.595 sec)\n",
      "INFO:tensorflow:global_step/sec: 173.025\n",
      "INFO:tensorflow:loss = 0.22723697, step = 10201 (0.580 sec)\n",
      "INFO:tensorflow:global_step/sec: 182.063\n",
      "INFO:tensorflow:loss = 0.22711772, step = 10301 (0.548 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.255\n",
      "INFO:tensorflow:loss = 0.22688942, step = 10401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.427\n",
      "INFO:tensorflow:loss = 0.22738239, step = 10501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.09\n",
      "INFO:tensorflow:loss = 0.22773941, step = 10601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 173.527\n",
      "INFO:tensorflow:loss = 0.22723547, step = 10701 (0.576 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.276\n",
      "INFO:tensorflow:loss = 0.22735077, step = 10801 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.427\n",
      "INFO:tensorflow:loss = 0.2266465, step = 10901 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.9\n",
      "INFO:tensorflow:loss = 0.22640207, step = 11001 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.135\n",
      "INFO:tensorflow:loss = 0.22696877, step = 11101 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.478\n",
      "INFO:tensorflow:loss = 0.2261745, step = 11201 (0.506 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.363\n",
      "INFO:tensorflow:loss = 0.22753051, step = 11301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.501\n",
      "INFO:tensorflow:loss = 0.22642873, step = 11401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.827\n",
      "INFO:tensorflow:loss = 0.2271044, step = 11501 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.992\n",
      "INFO:tensorflow:loss = 0.22680509, step = 11601 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.652\n",
      "INFO:tensorflow:loss = 0.22688827, step = 11701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.358\n",
      "INFO:tensorflow:loss = 0.22670965, step = 11801 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.307\n",
      "INFO:tensorflow:loss = 0.22664163, step = 11901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.26\n",
      "INFO:tensorflow:loss = 0.22664285, step = 12001 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.544\n",
      "INFO:tensorflow:loss = 0.22606179, step = 12101 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.198\n",
      "INFO:tensorflow:loss = 0.22668347, step = 12201 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.628\n",
      "INFO:tensorflow:loss = 0.22662356, step = 12301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.742\n",
      "INFO:tensorflow:loss = 0.22588904, step = 12401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.035\n",
      "INFO:tensorflow:loss = 0.22523032, step = 12501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.508\n",
      "INFO:tensorflow:loss = 0.22717819, step = 12601 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.558\n",
      "INFO:tensorflow:loss = 0.22635739, step = 12701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.353\n",
      "INFO:tensorflow:loss = 0.22604527, step = 12801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.255\n",
      "INFO:tensorflow:loss = 0.22562993, step = 12901 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.032\n",
      "INFO:tensorflow:loss = 0.2249108, step = 13001 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.712\n",
      "INFO:tensorflow:loss = 0.22597092, step = 13101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.23\n",
      "INFO:tensorflow:loss = 0.22590832, step = 13201 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.811\n",
      "INFO:tensorflow:loss = 0.22489554, step = 13301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.651\n",
      "INFO:tensorflow:loss = 0.22692409, step = 13401 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.868\n",
      "INFO:tensorflow:loss = 0.22618455, step = 13501 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.578\n",
      "INFO:tensorflow:loss = 0.22523974, step = 13601 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.009\n",
      "INFO:tensorflow:loss = 0.22567318, step = 13701 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.672\n",
      "INFO:tensorflow:loss = 0.22578806, step = 13801 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.982\n",
      "INFO:tensorflow:loss = 0.22578554, step = 13901 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.432\n",
      "INFO:tensorflow:loss = 0.22530487, step = 14001 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.847\n",
      "INFO:tensorflow:loss = 0.22507434, step = 14101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.976\n",
      "INFO:tensorflow:loss = 0.22510347, step = 14201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.413\n",
      "INFO:tensorflow:loss = 0.22539443, step = 14301 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.654\n",
      "INFO:tensorflow:loss = 0.22556898, step = 14401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.78\n",
      "INFO:tensorflow:loss = 0.22514984, step = 14501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.834\n",
      "INFO:tensorflow:loss = 0.2255833, step = 14601 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.778\n",
      "INFO:tensorflow:loss = 0.22456682, step = 14701 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.125\n",
      "INFO:tensorflow:loss = 0.22534482, step = 14801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.025\n",
      "INFO:tensorflow:loss = 0.22371751, step = 14901 (0.508 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_2_w_0/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.22503504.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:51:40\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:51:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.17509522, global_step = 15000, loss = 2.2570999\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:51:42\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:51:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.17148502, global_step = 15000, loss = 2.2575088\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259a9a3b50>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_2_w_0.25', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:51:43\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:51:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.99, global_step = 20000, loss = 0.082206175\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.25/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_2_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.16458434, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 157.907\n",
      "INFO:tensorflow:loss = 0.1537739, step = 10101 (0.634 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.312\n",
      "INFO:tensorflow:loss = 0.14654519, step = 10201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.985\n",
      "INFO:tensorflow:loss = 0.16745916, step = 10301 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 180.746\n",
      "INFO:tensorflow:loss = 0.14412725, step = 10401 (0.555 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.667\n",
      "INFO:tensorflow:loss = 0.15619019, step = 10501 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.571\n",
      "INFO:tensorflow:loss = 0.14592046, step = 10601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.551\n",
      "INFO:tensorflow:loss = 0.15122291, step = 10701 (0.528 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.411\n",
      "INFO:tensorflow:loss = 0.13578552, step = 10801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.537\n",
      "INFO:tensorflow:loss = 0.15087777, step = 10901 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.672\n",
      "INFO:tensorflow:loss = 0.14071539, step = 11001 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.617\n",
      "INFO:tensorflow:loss = 0.142955, step = 11101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.027\n",
      "INFO:tensorflow:loss = 0.1399867, step = 11201 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.506\n",
      "INFO:tensorflow:loss = 0.13120922, step = 11301 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.71\n",
      "INFO:tensorflow:loss = 0.121412754, step = 11401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.35\n",
      "INFO:tensorflow:loss = 0.13787127, step = 11501 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 194\n",
      "INFO:tensorflow:loss = 0.1360311, step = 11601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.214\n",
      "INFO:tensorflow:loss = 0.121490076, step = 11701 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.803\n",
      "INFO:tensorflow:loss = 0.1353574, step = 11801 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.736\n",
      "INFO:tensorflow:loss = 0.135443, step = 11901 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.805\n",
      "INFO:tensorflow:loss = 0.12372546, step = 12001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.003\n",
      "INFO:tensorflow:loss = 0.12883686, step = 12101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.429\n",
      "INFO:tensorflow:loss = 0.11721739, step = 12201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.658\n",
      "INFO:tensorflow:loss = 0.11959256, step = 12301 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.68\n",
      "INFO:tensorflow:loss = 0.12793362, step = 12401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.542\n",
      "INFO:tensorflow:loss = 0.11161507, step = 12501 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.438\n",
      "INFO:tensorflow:loss = 0.13139036, step = 12601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.381\n",
      "INFO:tensorflow:loss = 0.11662248, step = 12701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196\n",
      "INFO:tensorflow:loss = 0.119355164, step = 12801 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.762\n",
      "INFO:tensorflow:loss = 0.10481945, step = 12901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.941\n",
      "INFO:tensorflow:loss = 0.12051803, step = 13001 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.116\n",
      "INFO:tensorflow:loss = 0.1094599, step = 13101 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.839\n",
      "INFO:tensorflow:loss = 0.11107132, step = 13201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.408\n",
      "INFO:tensorflow:loss = 0.12095395, step = 13301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.265\n",
      "INFO:tensorflow:loss = 0.105830915, step = 13401 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.821\n",
      "INFO:tensorflow:loss = 0.10991823, step = 13501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.581\n",
      "INFO:tensorflow:loss = 0.11656584, step = 13601 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.762\n",
      "INFO:tensorflow:loss = 0.105075106, step = 13701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.481\n",
      "INFO:tensorflow:loss = 0.113842405, step = 13801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.783\n",
      "INFO:tensorflow:loss = 0.097089656, step = 13901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.181\n",
      "INFO:tensorflow:loss = 0.10608746, step = 14001 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.688\n",
      "INFO:tensorflow:loss = 0.102240816, step = 14101 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 186.978\n",
      "INFO:tensorflow:loss = 0.11018354, step = 14201 (0.535 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.328\n",
      "INFO:tensorflow:loss = 0.1077016, step = 14301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.93\n",
      "INFO:tensorflow:loss = 0.10938245, step = 14401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.68\n",
      "INFO:tensorflow:loss = 0.11359585, step = 14501 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.27\n",
      "INFO:tensorflow:loss = 0.10239108, step = 14601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.702\n",
      "INFO:tensorflow:loss = 0.09947026, step = 14701 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.203\n",
      "INFO:tensorflow:loss = 0.105790794, step = 14801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.938\n",
      "INFO:tensorflow:loss = 0.101771444, step = 14901 (0.510 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_2_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.102835745.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:52:10\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:52:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7926249, global_step = 15000, loss = 0.6561564\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:52:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:52:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7952371, global_step = 15000, loss = 0.6387563\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f25904560d0>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_2_w_0.5', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:52:13\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:52:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.96, global_step = 20000, loss = 0.09689247\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.5/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_2_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.0835414, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 161.466\n",
      "INFO:tensorflow:loss = 0.08628471, step = 10101 (0.623 sec)\n",
      "INFO:tensorflow:global_step/sec: 181.562\n",
      "INFO:tensorflow:loss = 0.07516888, step = 10201 (0.549 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.602\n",
      "INFO:tensorflow:loss = 0.07640487, step = 10301 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.651\n",
      "INFO:tensorflow:loss = 0.0767674, step = 10401 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.147\n",
      "INFO:tensorflow:loss = 0.08841418, step = 10501 (0.539 sec)\n",
      "INFO:tensorflow:global_step/sec: 182.785\n",
      "INFO:tensorflow:loss = 0.07962565, step = 10601 (0.547 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.307\n",
      "INFO:tensorflow:loss = 0.09351416, step = 10701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.711\n",
      "INFO:tensorflow:loss = 0.084316194, step = 10801 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.5\n",
      "INFO:tensorflow:loss = 0.06483824, step = 10901 (0.534 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.558\n",
      "INFO:tensorflow:loss = 0.08159761, step = 11001 (0.529 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.66\n",
      "INFO:tensorflow:loss = 0.07687772, step = 11101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.017\n",
      "INFO:tensorflow:loss = 0.07941669, step = 11201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.873\n",
      "INFO:tensorflow:loss = 0.08717035, step = 11301 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.97\n",
      "INFO:tensorflow:loss = 0.06615187, step = 11401 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.028\n",
      "INFO:tensorflow:loss = 0.07475668, step = 11501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.079\n",
      "INFO:tensorflow:loss = 0.080374494, step = 11601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.144\n",
      "INFO:tensorflow:loss = 0.06350586, step = 11701 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.497\n",
      "INFO:tensorflow:loss = 0.08780302, step = 11801 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.462\n",
      "INFO:tensorflow:loss = 0.06677789, step = 11901 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.684\n",
      "INFO:tensorflow:loss = 0.06520638, step = 12001 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.141\n",
      "INFO:tensorflow:loss = 0.072801635, step = 12101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.408\n",
      "INFO:tensorflow:loss = 0.060661364, step = 12201 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.526\n",
      "INFO:tensorflow:loss = 0.07388802, step = 12301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.169\n",
      "INFO:tensorflow:loss = 0.0587679, step = 12401 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.474\n",
      "INFO:tensorflow:loss = 0.060421295, step = 12501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.099\n",
      "INFO:tensorflow:loss = 0.058360443, step = 12601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.4\n",
      "INFO:tensorflow:loss = 0.06420273, step = 12701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.698\n",
      "INFO:tensorflow:loss = 0.06938212, step = 12801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.332\n",
      "INFO:tensorflow:loss = 0.060431413, step = 12901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.373\n",
      "INFO:tensorflow:loss = 0.06439998, step = 13001 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.372\n",
      "INFO:tensorflow:loss = 0.06621039, step = 13101 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.752\n",
      "INFO:tensorflow:loss = 0.056045774, step = 13201 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.976\n",
      "INFO:tensorflow:loss = 0.05865808, step = 13301 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.602\n",
      "INFO:tensorflow:loss = 0.05791655, step = 13401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.168\n",
      "INFO:tensorflow:loss = 0.073629916, step = 13501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.723\n",
      "INFO:tensorflow:loss = 0.062489808, step = 13601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.881\n",
      "INFO:tensorflow:loss = 0.06631659, step = 13701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.367\n",
      "INFO:tensorflow:loss = 0.060956486, step = 13801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.038\n",
      "INFO:tensorflow:loss = 0.06146245, step = 13901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.892\n",
      "INFO:tensorflow:loss = 0.07143617, step = 14001 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.007\n",
      "INFO:tensorflow:loss = 0.06378675, step = 14101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.274\n",
      "INFO:tensorflow:loss = 0.06534417, step = 14201 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.487\n",
      "INFO:tensorflow:loss = 0.063214995, step = 14301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.013\n",
      "INFO:tensorflow:loss = 0.06243536, step = 14401 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.778\n",
      "INFO:tensorflow:loss = 0.055248562, step = 14501 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.829\n",
      "INFO:tensorflow:loss = 0.06424664, step = 14601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.537\n",
      "INFO:tensorflow:loss = 0.06853526, step = 14701 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.256\n",
      "INFO:tensorflow:loss = 0.055595327, step = 14801 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.46\n",
      "INFO:tensorflow:loss = 0.05851335, step = 14901 (0.512 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_2_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.06105393.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:52:40\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.5/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:52:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7822982, global_step = 15000, loss = 0.82740337\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:52:42\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.5/model.ckpt-15000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1146:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:52:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7843178, global_step = 15000, loss = 0.816922\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f2596df0890>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_2_w_0.75', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:52:43\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:52:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.95, global_step = 20000, loss = 0.20539017\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.75/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_2_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.058300003, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 159.615\n",
      "INFO:tensorflow:loss = 0.042362683, step = 10101 (0.626 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.6\n",
      "INFO:tensorflow:loss = 0.04957994, step = 10201 (0.533 sec)\n",
      "INFO:tensorflow:global_step/sec: 179.939\n",
      "INFO:tensorflow:loss = 0.039557405, step = 10301 (0.558 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.685\n",
      "INFO:tensorflow:loss = 0.053157814, step = 10401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.714\n",
      "INFO:tensorflow:loss = 0.048722424, step = 10501 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.501\n",
      "INFO:tensorflow:loss = 0.04891356, step = 10601 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.82\n",
      "INFO:tensorflow:loss = 0.047081944, step = 10701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.267\n",
      "INFO:tensorflow:loss = 0.04151934, step = 10801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.861\n",
      "INFO:tensorflow:loss = 0.0419416, step = 10901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.651\n",
      "INFO:tensorflow:loss = 0.052858207, step = 11001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.797\n",
      "INFO:tensorflow:loss = 0.05199843, step = 11101 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.418\n",
      "INFO:tensorflow:loss = 0.044026874, step = 11201 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.862\n",
      "INFO:tensorflow:loss = 0.03749102, step = 11301 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.887\n",
      "INFO:tensorflow:loss = 0.0625353, step = 11401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.065\n",
      "INFO:tensorflow:loss = 0.03789474, step = 11501 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.658\n",
      "INFO:tensorflow:loss = 0.039474748, step = 11601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.327\n",
      "INFO:tensorflow:loss = 0.043827306, step = 11701 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.242\n",
      "INFO:tensorflow:loss = 0.04621254, step = 11801 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.219\n",
      "INFO:tensorflow:loss = 0.053233944, step = 11901 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.274\n",
      "INFO:tensorflow:loss = 0.03881769, step = 12001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.868\n",
      "INFO:tensorflow:loss = 0.0401132, step = 12101 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.175\n",
      "INFO:tensorflow:loss = 0.035697214, step = 12201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.065\n",
      "INFO:tensorflow:loss = 0.04369209, step = 12301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.262\n",
      "INFO:tensorflow:loss = 0.057504773, step = 12401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.697\n",
      "INFO:tensorflow:loss = 0.039745454, step = 12501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.747\n",
      "INFO:tensorflow:loss = 0.039658334, step = 12601 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.441\n",
      "INFO:tensorflow:loss = 0.039571986, step = 12701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.939\n",
      "INFO:tensorflow:loss = 0.040150724, step = 12801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.721\n",
      "INFO:tensorflow:loss = 0.050563037, step = 12901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.185\n",
      "INFO:tensorflow:loss = 0.041568935, step = 13001 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.225\n",
      "INFO:tensorflow:loss = 0.05801519, step = 13101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.34\n",
      "INFO:tensorflow:loss = 0.048490204, step = 13201 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.091\n",
      "INFO:tensorflow:loss = 0.042252984, step = 13301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.003\n",
      "INFO:tensorflow:loss = 0.03611902, step = 13401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.637\n",
      "INFO:tensorflow:loss = 0.036406606, step = 13501 (0.533 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.666\n",
      "INFO:tensorflow:loss = 0.040717863, step = 13601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.263\n",
      "INFO:tensorflow:loss = 0.039090857, step = 13701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.141\n",
      "INFO:tensorflow:loss = 0.044635363, step = 13801 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.152\n",
      "INFO:tensorflow:loss = 0.050035607, step = 13901 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.507\n",
      "INFO:tensorflow:loss = 0.044271126, step = 14001 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.087\n",
      "INFO:tensorflow:loss = 0.039097633, step = 14101 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.411\n",
      "INFO:tensorflow:loss = 0.03470307, step = 14201 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.705\n",
      "INFO:tensorflow:loss = 0.038225375, step = 14301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.22\n",
      "INFO:tensorflow:loss = 0.052313387, step = 14401 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.052\n",
      "INFO:tensorflow:loss = 0.03432124, step = 14501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.136\n",
      "INFO:tensorflow:loss = 0.042986207, step = 14601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.706\n",
      "INFO:tensorflow:loss = 0.03330175, step = 14701 (0.528 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.27\n",
      "INFO:tensorflow:loss = 0.03942462, step = 14801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.519\n",
      "INFO:tensorflow:loss = 0.036524687, step = 14901 (0.513 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_2_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.03976357.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:53:10\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:53:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.77248424, global_step = 15000, loss = 0.95264703\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:53:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:53:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.780574, global_step = 15000, loss = 0.8989817\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f25991364d0>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_2_w_1', '_save_summary_steps': 100}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:53:13\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:53:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.99, global_step = 20000, loss = 0.045277234\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_1/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_2_w_1/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.008933035, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 155.204\n",
      "INFO:tensorflow:loss = 0.0155206155, step = 10101 (0.646 sec)\n",
      "INFO:tensorflow:global_step/sec: 180.963\n",
      "INFO:tensorflow:loss = 0.02674498, step = 10201 (0.551 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.97\n",
      "INFO:tensorflow:loss = 0.013609977, step = 10301 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.762\n",
      "INFO:tensorflow:loss = 0.013429604, step = 10401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.971\n",
      "INFO:tensorflow:loss = 0.00892911, step = 10501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.805\n",
      "INFO:tensorflow:loss = 0.019408638, step = 10601 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.524\n",
      "INFO:tensorflow:loss = 0.01127132, step = 10701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.949\n",
      "INFO:tensorflow:loss = 0.00795492, step = 10801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.421\n",
      "INFO:tensorflow:loss = 0.011291659, step = 10901 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.007\n",
      "INFO:tensorflow:loss = 0.015241169, step = 11001 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.101\n",
      "INFO:tensorflow:loss = 0.012748895, step = 11101 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.076\n",
      "INFO:tensorflow:loss = 0.016644774, step = 11201 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.528\n",
      "INFO:tensorflow:loss = 0.030463042, step = 11301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.034\n",
      "INFO:tensorflow:loss = 0.0143309245, step = 11401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.207\n",
      "INFO:tensorflow:loss = 0.018284937, step = 11501 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.616\n",
      "INFO:tensorflow:loss = 0.02570161, step = 11601 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.16\n",
      "INFO:tensorflow:loss = 0.015882198, step = 11701 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.886\n",
      "INFO:tensorflow:loss = 0.021568352, step = 11801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.502\n",
      "INFO:tensorflow:loss = 0.014207812, step = 11901 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.623\n",
      "INFO:tensorflow:loss = 0.044553183, step = 12001 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.914\n",
      "INFO:tensorflow:loss = 0.00952029, step = 12101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.848\n",
      "INFO:tensorflow:loss = 0.015975224, step = 12201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.252\n",
      "INFO:tensorflow:loss = 0.015628643, step = 12301 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.597\n",
      "INFO:tensorflow:loss = 0.00896713, step = 12401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.181\n",
      "INFO:tensorflow:loss = 0.0062792674, step = 12501 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.745\n",
      "INFO:tensorflow:loss = 0.012610858, step = 12601 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.758\n",
      "INFO:tensorflow:loss = 0.010149906, step = 12701 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.028\n",
      "INFO:tensorflow:loss = 0.015719442, step = 12801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.683\n",
      "INFO:tensorflow:loss = 0.011025094, step = 12901 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.265\n",
      "INFO:tensorflow:loss = 0.009710056, step = 13001 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.017\n",
      "INFO:tensorflow:loss = 0.01568345, step = 13101 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.442\n",
      "INFO:tensorflow:loss = 0.0075634625, step = 13201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.015\n",
      "INFO:tensorflow:loss = 0.007860344, step = 13301 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 183.895\n",
      "INFO:tensorflow:loss = 0.009154947, step = 13401 (0.544 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.246\n",
      "INFO:tensorflow:loss = 0.007473769, step = 13501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.891\n",
      "INFO:tensorflow:loss = 0.0060703615, step = 13601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.501\n",
      "INFO:tensorflow:loss = 0.0075354096, step = 13701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.342\n",
      "INFO:tensorflow:loss = 0.0139397085, step = 13801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.542\n",
      "INFO:tensorflow:loss = 0.008175357, step = 13901 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.841\n",
      "INFO:tensorflow:loss = 0.024888545, step = 14001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.396\n",
      "INFO:tensorflow:loss = 0.016194327, step = 14101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.09\n",
      "INFO:tensorflow:loss = 0.013034037, step = 14201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.592\n",
      "INFO:tensorflow:loss = 0.0073614246, step = 14301 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.418\n",
      "INFO:tensorflow:loss = 0.00880801, step = 14401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.898\n",
      "INFO:tensorflow:loss = 0.0068433303, step = 14501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.098\n",
      "INFO:tensorflow:loss = 0.009453688, step = 14601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.55\n",
      "INFO:tensorflow:loss = 0.010051349, step = 14701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.851\n",
      "INFO:tensorflow:loss = 0.008245758, step = 14801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.793\n",
      "INFO:tensorflow:loss = 0.008449255, step = 14901 (0.508 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_2_w_1/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.01021336.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:53:40\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:53:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.819595, global_step = 15000, loss = 0.74800706\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:53:42\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_2_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:53:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.82674706, global_step = 15000, loss = 0.7363451\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259044bf90>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_5_w_0', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:53:43\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:53:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.99, global_step = 20000, loss = 0.05171888\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0/model.ckpt-10000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_5_w_0/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.22982433, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 175.688\n",
      "INFO:tensorflow:loss = 0.22959398, step = 10101 (0.571 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.407\n",
      "INFO:tensorflow:loss = 0.2296048, step = 10201 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.822\n",
      "INFO:tensorflow:loss = 0.22954701, step = 10301 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.784\n",
      "INFO:tensorflow:loss = 0.22968288, step = 10401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 184.525\n",
      "INFO:tensorflow:loss = 0.22976574, step = 10501 (0.544 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.199\n",
      "INFO:tensorflow:loss = 0.22949417, step = 10601 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.987\n",
      "INFO:tensorflow:loss = 0.22964174, step = 10701 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.843\n",
      "INFO:tensorflow:loss = 0.22947635, step = 10801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.618\n",
      "INFO:tensorflow:loss = 0.22948743, step = 10901 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.625\n",
      "INFO:tensorflow:loss = 0.22990374, step = 11001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.542\n",
      "INFO:tensorflow:loss = 0.2296075, step = 11101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.011\n",
      "INFO:tensorflow:loss = 0.22977452, step = 11201 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.735\n",
      "INFO:tensorflow:loss = 0.22956309, step = 11301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.946\n",
      "INFO:tensorflow:loss = 0.22962846, step = 11401 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.222\n",
      "INFO:tensorflow:loss = 0.22942923, step = 11501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.754\n",
      "INFO:tensorflow:loss = 0.22963215, step = 11601 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 198.579\n",
      "INFO:tensorflow:loss = 0.22960062, step = 11701 (0.504 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.443\n",
      "INFO:tensorflow:loss = 0.22967513, step = 11801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.178\n",
      "INFO:tensorflow:loss = 0.22962561, step = 11901 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.172\n",
      "INFO:tensorflow:loss = 0.22961655, step = 12001 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.639\n",
      "INFO:tensorflow:loss = 0.22972539, step = 12101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.345\n",
      "INFO:tensorflow:loss = 0.22954763, step = 12201 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.67\n",
      "INFO:tensorflow:loss = 0.22985786, step = 12301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.824\n",
      "INFO:tensorflow:loss = 0.22932078, step = 12401 (0.505 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.723\n",
      "INFO:tensorflow:loss = 0.22963014, step = 12501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.857\n",
      "INFO:tensorflow:loss = 0.22938246, step = 12601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.825\n",
      "INFO:tensorflow:loss = 0.22963774, step = 12701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.13\n",
      "INFO:tensorflow:loss = 0.2294832, step = 12801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.776\n",
      "INFO:tensorflow:loss = 0.22957005, step = 12901 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.529\n",
      "INFO:tensorflow:loss = 0.22954233, step = 13001 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.04\n",
      "INFO:tensorflow:loss = 0.22941443, step = 13101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.15\n",
      "INFO:tensorflow:loss = 0.22946396, step = 13201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.021\n",
      "INFO:tensorflow:loss = 0.2293909, step = 13301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.615\n",
      "INFO:tensorflow:loss = 0.2292771, step = 13401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.543\n",
      "INFO:tensorflow:loss = 0.2295018, step = 13501 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.592\n",
      "INFO:tensorflow:loss = 0.22970693, step = 13601 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.629\n",
      "INFO:tensorflow:loss = 0.22934707, step = 13701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.965\n",
      "INFO:tensorflow:loss = 0.22947614, step = 13801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.19\n",
      "INFO:tensorflow:loss = 0.2296148, step = 13901 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.179\n",
      "INFO:tensorflow:loss = 0.22952735, step = 14001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.019\n",
      "INFO:tensorflow:loss = 0.22959577, step = 14101 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.817\n",
      "INFO:tensorflow:loss = 0.22951956, step = 14201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.781\n",
      "INFO:tensorflow:loss = 0.22966924, step = 14301 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.052\n",
      "INFO:tensorflow:loss = 0.2291945, step = 14401 (0.525 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.167\n",
      "INFO:tensorflow:loss = 0.22957407, step = 14501 (0.530 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.287\n",
      "INFO:tensorflow:loss = 0.22941133, step = 14601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.368\n",
      "INFO:tensorflow:loss = 0.22966498, step = 14701 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.801\n",
      "INFO:tensorflow:loss = 0.22950412, step = 14801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.34\n",
      "INFO:tensorflow:loss = 0.22947589, step = 14901 (0.528 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_5_w_0/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.22933662.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:54:11\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:54:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.20461038, global_step = 15000, loss = 2.2709026\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:54:12\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1176:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:54:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.20902662, global_step = 15000, loss = 2.267397\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f252c767190>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_5_w_0.25', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:54:13\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:54:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.98, global_step = 20000, loss = 0.119127065\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.25/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_5_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.20999964, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 168.685\n",
      "INFO:tensorflow:loss = 0.2152474, step = 10101 (0.595 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.306\n",
      "INFO:tensorflow:loss = 0.20445111, step = 10201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.8\n",
      "INFO:tensorflow:loss = 0.19421661, step = 10301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.782\n",
      "INFO:tensorflow:loss = 0.19570391, step = 10401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.207\n",
      "INFO:tensorflow:loss = 0.1990067, step = 10501 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.753\n",
      "INFO:tensorflow:loss = 0.2065534, step = 10601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.741\n",
      "INFO:tensorflow:loss = 0.21588916, step = 10701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.251\n",
      "INFO:tensorflow:loss = 0.20252098, step = 10801 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.458\n",
      "INFO:tensorflow:loss = 0.18881452, step = 10901 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.716\n",
      "INFO:tensorflow:loss = 0.18561932, step = 11001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.348\n",
      "INFO:tensorflow:loss = 0.18685065, step = 11101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.577\n",
      "INFO:tensorflow:loss = 0.18846679, step = 11201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.238\n",
      "INFO:tensorflow:loss = 0.18134426, step = 11301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.425\n",
      "INFO:tensorflow:loss = 0.18750185, step = 11401 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.345\n",
      "INFO:tensorflow:loss = 0.18749882, step = 11501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.593\n",
      "INFO:tensorflow:loss = 0.17582592, step = 11601 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.706\n",
      "INFO:tensorflow:loss = 0.17995028, step = 11701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.151\n",
      "INFO:tensorflow:loss = 0.18882492, step = 11801 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 182.8\n",
      "INFO:tensorflow:loss = 0.1778626, step = 11901 (0.547 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.879\n",
      "INFO:tensorflow:loss = 0.18492323, step = 12001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.39\n",
      "INFO:tensorflow:loss = 0.18139683, step = 12101 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.685\n",
      "INFO:tensorflow:loss = 0.17212701, step = 12201 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.997\n",
      "INFO:tensorflow:loss = 0.1747424, step = 12301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.208\n",
      "INFO:tensorflow:loss = 0.18190028, step = 12401 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.38\n",
      "INFO:tensorflow:loss = 0.17631549, step = 12501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.126\n",
      "INFO:tensorflow:loss = 0.1717966, step = 12601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.928\n",
      "INFO:tensorflow:loss = 0.17080444, step = 12701 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.9\n",
      "INFO:tensorflow:loss = 0.17473255, step = 12801 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.274\n",
      "INFO:tensorflow:loss = 0.16922663, step = 12901 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.756\n",
      "INFO:tensorflow:loss = 0.18557456, step = 13001 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.314\n",
      "INFO:tensorflow:loss = 0.17682442, step = 13101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.741\n",
      "INFO:tensorflow:loss = 0.17292844, step = 13201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.773\n",
      "INFO:tensorflow:loss = 0.16959569, step = 13301 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.793\n",
      "INFO:tensorflow:loss = 0.16587076, step = 13401 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.524\n",
      "INFO:tensorflow:loss = 0.16657743, step = 13501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.763\n",
      "INFO:tensorflow:loss = 0.16459805, step = 13601 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.194\n",
      "INFO:tensorflow:loss = 0.16679557, step = 13701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.405\n",
      "INFO:tensorflow:loss = 0.16034563, step = 13801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.446\n",
      "INFO:tensorflow:loss = 0.16001649, step = 13901 (0.526 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.765\n",
      "INFO:tensorflow:loss = 0.16121873, step = 14001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.217\n",
      "INFO:tensorflow:loss = 0.16434018, step = 14101 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.265\n",
      "INFO:tensorflow:loss = 0.16253154, step = 14201 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.269\n",
      "INFO:tensorflow:loss = 0.16276419, step = 14301 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.418\n",
      "INFO:tensorflow:loss = 0.16833414, step = 14401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.391\n",
      "INFO:tensorflow:loss = 0.15811938, step = 14501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.214\n",
      "INFO:tensorflow:loss = 0.169734, step = 14601 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.369\n",
      "INFO:tensorflow:loss = 0.15319645, step = 14701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.548\n",
      "INFO:tensorflow:loss = 0.16422327, step = 14801 (0.506 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.805\n",
      "INFO:tensorflow:loss = 0.15981203, step = 14901 (0.521 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_5_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.16407032.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:54:41\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:54:42\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.71739054, global_step = 15000, loss = 0.9694052\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:54:42\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1186:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:54:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.734297, global_step = 15000, loss = 0.9131765\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f25860ad610>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_5_w_0.5', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:54:43\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:54:43\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.97, global_step = 20000, loss = 0.13422287\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.5/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_5_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.14635149, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 169.378\n",
      "INFO:tensorflow:loss = 0.1294744, step = 10101 (0.593 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.173\n",
      "INFO:tensorflow:loss = 0.1350053, step = 10201 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.72\n",
      "INFO:tensorflow:loss = 0.13022286, step = 10301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.359\n",
      "INFO:tensorflow:loss = 0.12774079, step = 10401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.935\n",
      "INFO:tensorflow:loss = 0.14994413, step = 10501 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.282\n",
      "INFO:tensorflow:loss = 0.13255665, step = 10601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.156\n",
      "INFO:tensorflow:loss = 0.13306843, step = 10701 (0.529 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.022\n",
      "INFO:tensorflow:loss = 0.13062555, step = 10801 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.091\n",
      "INFO:tensorflow:loss = 0.13169333, step = 10901 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.932\n",
      "INFO:tensorflow:loss = 0.13967235, step = 11001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.91\n",
      "INFO:tensorflow:loss = 0.13658227, step = 11101 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.135\n",
      "INFO:tensorflow:loss = 0.12549232, step = 11201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.246\n",
      "INFO:tensorflow:loss = 0.1308191, step = 11301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.611\n",
      "INFO:tensorflow:loss = 0.11958561, step = 11401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.76\n",
      "INFO:tensorflow:loss = 0.13372293, step = 11501 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.44\n",
      "INFO:tensorflow:loss = 0.1219646, step = 11601 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.605\n",
      "INFO:tensorflow:loss = 0.12476532, step = 11701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.19\n",
      "INFO:tensorflow:loss = 0.13535473, step = 11801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.959\n",
      "INFO:tensorflow:loss = 0.13168654, step = 11901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.249\n",
      "INFO:tensorflow:loss = 0.1361407, step = 12001 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.838\n",
      "INFO:tensorflow:loss = 0.117920265, step = 12101 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 186.967\n",
      "INFO:tensorflow:loss = 0.12649052, step = 12201 (0.536 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.072\n",
      "INFO:tensorflow:loss = 0.11779684, step = 12301 (0.528 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.939\n",
      "INFO:tensorflow:loss = 0.11613238, step = 12401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.874\n",
      "INFO:tensorflow:loss = 0.11676202, step = 12501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.245\n",
      "INFO:tensorflow:loss = 0.12584266, step = 12601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.202\n",
      "INFO:tensorflow:loss = 0.11735775, step = 12701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.702\n",
      "INFO:tensorflow:loss = 0.13304323, step = 12801 (0.539 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.038\n",
      "INFO:tensorflow:loss = 0.12568317, step = 12901 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.644\n",
      "INFO:tensorflow:loss = 0.12177897, step = 13001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.229\n",
      "INFO:tensorflow:loss = 0.13535775, step = 13101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.121\n",
      "INFO:tensorflow:loss = 0.11794785, step = 13201 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.314\n",
      "INFO:tensorflow:loss = 0.12963058, step = 13301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.548\n",
      "INFO:tensorflow:loss = 0.12385906, step = 13401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.327\n",
      "INFO:tensorflow:loss = 0.12649983, step = 13501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.359\n",
      "INFO:tensorflow:loss = 0.11940746, step = 13601 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.303\n",
      "INFO:tensorflow:loss = 0.11087996, step = 13701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.008\n",
      "INFO:tensorflow:loss = 0.115555465, step = 13801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.682\n",
      "INFO:tensorflow:loss = 0.12770584, step = 13901 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.448\n",
      "INFO:tensorflow:loss = 0.124820784, step = 14001 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.58\n",
      "INFO:tensorflow:loss = 0.121070966, step = 14101 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.482\n",
      "INFO:tensorflow:loss = 0.110134974, step = 14201 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.464\n",
      "INFO:tensorflow:loss = 0.120972544, step = 14301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.302\n",
      "INFO:tensorflow:loss = 0.11435372, step = 14401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.886\n",
      "INFO:tensorflow:loss = 0.116796, step = 14501 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.785\n",
      "INFO:tensorflow:loss = 0.11729307, step = 14601 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.599\n",
      "INFO:tensorflow:loss = 0.12181191, step = 14701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.744\n",
      "INFO:tensorflow:loss = 0.11646122, step = 14801 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.541\n",
      "INFO:tensorflow:loss = 0.11873439, step = 14901 (0.517 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_5_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.12558416.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:55:11\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.5/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:55:12\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7903545, global_step = 15000, loss = 0.7193802\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:55:12\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1196:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.5/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:55:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.80199665, global_step = 15000, loss = 0.67773306\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259aad9910>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_5_w_0.75', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:55:13\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:55:13\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.97, global_step = 20000, loss = 0.08445575\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.75/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_5_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.07805783, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 147.973\n",
      "INFO:tensorflow:loss = 0.072785564, step = 10101 (0.679 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.588\n",
      "INFO:tensorflow:loss = 0.08002365, step = 10201 (0.539 sec)\n",
      "INFO:tensorflow:global_step/sec: 181.863\n",
      "INFO:tensorflow:loss = 0.06442702, step = 10301 (0.548 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.661\n",
      "INFO:tensorflow:loss = 0.066070765, step = 10401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.713\n",
      "INFO:tensorflow:loss = 0.07509412, step = 10501 (0.530 sec)\n",
      "INFO:tensorflow:global_step/sec: 176.899\n",
      "INFO:tensorflow:loss = 0.0656589, step = 10601 (0.566 sec)\n",
      "INFO:tensorflow:global_step/sec: 181.192\n",
      "INFO:tensorflow:loss = 0.076400295, step = 10701 (0.552 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.045\n",
      "INFO:tensorflow:loss = 0.06341462, step = 10801 (0.533 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.417\n",
      "INFO:tensorflow:loss = 0.07224209, step = 10901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.873\n",
      "INFO:tensorflow:loss = 0.062823616, step = 11001 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 178.259\n",
      "INFO:tensorflow:loss = 0.06496046, step = 11101 (0.560 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.675\n",
      "INFO:tensorflow:loss = 0.06728366, step = 11201 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 183.904\n",
      "INFO:tensorflow:loss = 0.06525113, step = 11301 (0.544 sec)\n",
      "INFO:tensorflow:global_step/sec: 182.163\n",
      "INFO:tensorflow:loss = 0.070543155, step = 11401 (0.549 sec)\n",
      "INFO:tensorflow:global_step/sec: 180.42\n",
      "INFO:tensorflow:loss = 0.07130021, step = 11501 (0.553 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.852\n",
      "INFO:tensorflow:loss = 0.06781314, step = 11601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.133\n",
      "INFO:tensorflow:loss = 0.06508673, step = 11701 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.789\n",
      "INFO:tensorflow:loss = 0.07773789, step = 11801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.287\n",
      "INFO:tensorflow:loss = 0.07188814, step = 11901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.188\n",
      "INFO:tensorflow:loss = 0.058726363, step = 12001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.464\n",
      "INFO:tensorflow:loss = 0.06472294, step = 12101 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.622\n",
      "INFO:tensorflow:loss = 0.06372656, step = 12201 (0.538 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.458\n",
      "INFO:tensorflow:loss = 0.06643268, step = 12301 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.769\n",
      "INFO:tensorflow:loss = 0.06150389, step = 12401 (0.538 sec)\n",
      "INFO:tensorflow:global_step/sec: 185.748\n",
      "INFO:tensorflow:loss = 0.07166449, step = 12501 (0.541 sec)\n",
      "INFO:tensorflow:global_step/sec: 180.782\n",
      "INFO:tensorflow:loss = 0.06316779, step = 12601 (0.553 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.669\n",
      "INFO:tensorflow:loss = 0.061277315, step = 12701 (0.531 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.073\n",
      "INFO:tensorflow:loss = 0.061578006, step = 12801 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.939\n",
      "INFO:tensorflow:loss = 0.061643347, step = 12901 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.313\n",
      "INFO:tensorflow:loss = 0.06282696, step = 13001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.556\n",
      "INFO:tensorflow:loss = 0.067008175, step = 13101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.404\n",
      "INFO:tensorflow:loss = 0.06326947, step = 13201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.684\n",
      "INFO:tensorflow:loss = 0.059521608, step = 13301 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.683\n",
      "INFO:tensorflow:loss = 0.07738202, step = 13401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.337\n",
      "INFO:tensorflow:loss = 0.059889454, step = 13501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.853\n",
      "INFO:tensorflow:loss = 0.05965277, step = 13601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.37\n",
      "INFO:tensorflow:loss = 0.06290438, step = 13701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.644\n",
      "INFO:tensorflow:loss = 0.05645845, step = 13801 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 174.178\n",
      "INFO:tensorflow:loss = 0.057307065, step = 13901 (0.570 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.305\n",
      "INFO:tensorflow:loss = 0.057438895, step = 14001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.989\n",
      "INFO:tensorflow:loss = 0.060624674, step = 14101 (0.530 sec)\n",
      "INFO:tensorflow:global_step/sec: 182.645\n",
      "INFO:tensorflow:loss = 0.062201817, step = 14201 (0.548 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.654\n",
      "INFO:tensorflow:loss = 0.0599547, step = 14301 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.026\n",
      "INFO:tensorflow:loss = 0.060534474, step = 14401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.792\n",
      "INFO:tensorflow:loss = 0.060433477, step = 14501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.229\n",
      "INFO:tensorflow:loss = 0.064591244, step = 14601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 183.313\n",
      "INFO:tensorflow:loss = 0.06530923, step = 14701 (0.545 sec)\n",
      "INFO:tensorflow:global_step/sec: 186.18\n",
      "INFO:tensorflow:loss = 0.060856216, step = 14801 (0.537 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.441\n",
      "INFO:tensorflow:loss = 0.059040032, step = 14901 (0.530 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_5_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.063351296.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:55:42\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:55:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7920756, global_step = 15000, loss = 0.76848805\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:55:43\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_0.75/model.ckpt-15000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1206:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:55:44\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.80532444, global_step = 15000, loss = 0.73938185\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259aad9e90>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_5_w_1', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:55:44\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:55:44\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.95, global_step = 20000, loss = 0.1377061\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_1/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_5_w_1/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.03656557, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 159.842\n",
      "INFO:tensorflow:loss = 0.020596815, step = 10101 (0.626 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.112\n",
      "INFO:tensorflow:loss = 0.017523376, step = 10201 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.944\n",
      "INFO:tensorflow:loss = 0.048041955, step = 10301 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.305\n",
      "INFO:tensorflow:loss = 0.021254992, step = 10401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.966\n",
      "INFO:tensorflow:loss = 0.029588781, step = 10501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.792\n",
      "INFO:tensorflow:loss = 0.033282116, step = 10601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.276\n",
      "INFO:tensorflow:loss = 0.026163394, step = 10701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.912\n",
      "INFO:tensorflow:loss = 0.0142207, step = 10801 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.661\n",
      "INFO:tensorflow:loss = 0.016327845, step = 10901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.874\n",
      "INFO:tensorflow:loss = 0.014171834, step = 11001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.749\n",
      "INFO:tensorflow:loss = 0.023405077, step = 11101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.556\n",
      "INFO:tensorflow:loss = 0.026812192, step = 11201 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.281\n",
      "INFO:tensorflow:loss = 0.019678058, step = 11301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.609\n",
      "INFO:tensorflow:loss = 0.012603638, step = 11401 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.331\n",
      "INFO:tensorflow:loss = 0.022826066, step = 11501 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.647\n",
      "INFO:tensorflow:loss = 0.021779293, step = 11601 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.687\n",
      "INFO:tensorflow:loss = 0.010369688, step = 11701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.622\n",
      "INFO:tensorflow:loss = 0.01548953, step = 11801 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.977\n",
      "INFO:tensorflow:loss = 0.019031877, step = 11901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.318\n",
      "INFO:tensorflow:loss = 0.014811146, step = 12001 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.235\n",
      "INFO:tensorflow:loss = 0.02297677, step = 12101 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.588\n",
      "INFO:tensorflow:loss = 0.028150262, step = 12201 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.232\n",
      "INFO:tensorflow:loss = 0.013164568, step = 12301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.911\n",
      "INFO:tensorflow:loss = 0.015873415, step = 12401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.635\n",
      "INFO:tensorflow:loss = 0.015044808, step = 12501 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.664\n",
      "INFO:tensorflow:loss = 0.020452589, step = 12601 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.8\n",
      "INFO:tensorflow:loss = 0.021351099, step = 12701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.367\n",
      "INFO:tensorflow:loss = 0.013895404, step = 12801 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.143\n",
      "INFO:tensorflow:loss = 0.013979789, step = 12901 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.946\n",
      "INFO:tensorflow:loss = 0.01966403, step = 13001 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.676\n",
      "INFO:tensorflow:loss = 0.012925937, step = 13101 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.693\n",
      "INFO:tensorflow:loss = 0.009737084, step = 13201 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.153\n",
      "INFO:tensorflow:loss = 0.010356203, step = 13301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.712\n",
      "INFO:tensorflow:loss = 0.01166746, step = 13401 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.024\n",
      "INFO:tensorflow:loss = 0.011002014, step = 13501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.201\n",
      "INFO:tensorflow:loss = 0.015085202, step = 13601 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.396\n",
      "INFO:tensorflow:loss = 0.01306295, step = 13701 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.213\n",
      "INFO:tensorflow:loss = 0.011887683, step = 13801 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.764\n",
      "INFO:tensorflow:loss = 0.0109419795, step = 13901 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.609\n",
      "INFO:tensorflow:loss = 0.0097478535, step = 14001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.944\n",
      "INFO:tensorflow:loss = 0.010837432, step = 14101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.477\n",
      "INFO:tensorflow:loss = 0.010760882, step = 14201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.539\n",
      "INFO:tensorflow:loss = 0.0065399692, step = 14301 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.366\n",
      "INFO:tensorflow:loss = 0.0051380508, step = 14401 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.206\n",
      "INFO:tensorflow:loss = 0.018033952, step = 14501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.702\n",
      "INFO:tensorflow:loss = 0.0096151205, step = 14601 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.115\n",
      "INFO:tensorflow:loss = 0.012686099, step = 14701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.288\n",
      "INFO:tensorflow:loss = 0.011181757, step = 14801 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.424\n",
      "INFO:tensorflow:loss = 0.014814437, step = 14901 (0.512 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_5_w_1/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.01803711.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:56:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:56:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7814926, global_step = 15000, loss = 1.0389446\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:56:13\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_5_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:56:14\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.79617304, global_step = 15000, loss = 0.97291976\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f2597761610>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_10_w_0', '_save_summary_steps': 100}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:56:14\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:56:14\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 1.0, global_step = 20000, loss = 0.02928738\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_10_w_0/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.2302427, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 160.701\n",
      "INFO:tensorflow:loss = 0.23029666, step = 10101 (0.623 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.894\n",
      "INFO:tensorflow:loss = 0.23029791, step = 10201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.192\n",
      "INFO:tensorflow:loss = 0.23034175, step = 10301 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.368\n",
      "INFO:tensorflow:loss = 0.23027842, step = 10401 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.291\n",
      "INFO:tensorflow:loss = 0.23028609, step = 10501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.708\n",
      "INFO:tensorflow:loss = 0.23028228, step = 10601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.271\n",
      "INFO:tensorflow:loss = 0.23030066, step = 10701 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.013\n",
      "INFO:tensorflow:loss = 0.23032443, step = 10801 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.065\n",
      "INFO:tensorflow:loss = 0.23031011, step = 10901 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.383\n",
      "INFO:tensorflow:loss = 0.2303014, step = 11001 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.271\n",
      "INFO:tensorflow:loss = 0.23029953, step = 11101 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.553\n",
      "INFO:tensorflow:loss = 0.23026432, step = 11201 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.515\n",
      "INFO:tensorflow:loss = 0.23029573, step = 11301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.427\n",
      "INFO:tensorflow:loss = 0.23032072, step = 11401 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.995\n",
      "INFO:tensorflow:loss = 0.23030679, step = 11501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.079\n",
      "INFO:tensorflow:loss = 0.23031273, step = 11601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.681\n",
      "INFO:tensorflow:loss = 0.2302703, step = 11701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.512\n",
      "INFO:tensorflow:loss = 0.23027836, step = 11801 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.437\n",
      "INFO:tensorflow:loss = 0.23031387, step = 11901 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.928\n",
      "INFO:tensorflow:loss = 0.23028153, step = 12001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.266\n",
      "INFO:tensorflow:loss = 0.2303084, step = 12101 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.591\n",
      "INFO:tensorflow:loss = 0.23022994, step = 12201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.985\n",
      "INFO:tensorflow:loss = 0.23035893, step = 12301 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.148\n",
      "INFO:tensorflow:loss = 0.23028222, step = 12401 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.663\n",
      "INFO:tensorflow:loss = 0.23030932, step = 12501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.305\n",
      "INFO:tensorflow:loss = 0.23028098, step = 12601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.428\n",
      "INFO:tensorflow:loss = 0.23026524, step = 12701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.464\n",
      "INFO:tensorflow:loss = 0.23030522, step = 12801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.964\n",
      "INFO:tensorflow:loss = 0.23034196, step = 12901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.461\n",
      "INFO:tensorflow:loss = 0.23031405, step = 13001 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.471\n",
      "INFO:tensorflow:loss = 0.23026876, step = 13101 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.472\n",
      "INFO:tensorflow:loss = 0.23032188, step = 13201 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.887\n",
      "INFO:tensorflow:loss = 0.2302929, step = 13301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.852\n",
      "INFO:tensorflow:loss = 0.23025556, step = 13401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.167\n",
      "INFO:tensorflow:loss = 0.23030142, step = 13501 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.033\n",
      "INFO:tensorflow:loss = 0.2302747, step = 13601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.273\n",
      "INFO:tensorflow:loss = 0.23029155, step = 13701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.173\n",
      "INFO:tensorflow:loss = 0.23024791, step = 13801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.199\n",
      "INFO:tensorflow:loss = 0.23027872, step = 13901 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.507\n",
      "INFO:tensorflow:loss = 0.2302835, step = 14001 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.121\n",
      "INFO:tensorflow:loss = 0.2302745, step = 14101 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.638\n",
      "INFO:tensorflow:loss = 0.23029178, step = 14201 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.515\n",
      "INFO:tensorflow:loss = 0.23025325, step = 14301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.725\n",
      "INFO:tensorflow:loss = 0.23029315, step = 14401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.339\n",
      "INFO:tensorflow:loss = 0.2302745, step = 14501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.65\n",
      "INFO:tensorflow:loss = 0.23027222, step = 14601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.907\n",
      "INFO:tensorflow:loss = 0.23030192, step = 14701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.296\n",
      "INFO:tensorflow:loss = 0.2303119, step = 14801 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.963\n",
      "INFO:tensorflow:loss = 0.23024341, step = 14901 (0.521 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_10_w_0/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.23033419.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:56:41\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:56:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.13010839, global_step = 15000, loss = 2.319237\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:56:43\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:56:44\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.12551996, global_step = 15000, loss = 2.320594\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f25866b6250>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_10_w_0.25', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:56:44\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:56:44\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.97, global_step = 20000, loss = 0.10695127\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.25/model.ckpt-10000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_10_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.35144985, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 174.183\n",
      "INFO:tensorflow:loss = 0.32945758, step = 10101 (0.575 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.883\n",
      "INFO:tensorflow:loss = 0.32711464, step = 10201 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.592\n",
      "INFO:tensorflow:loss = 0.31635344, step = 10301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.908\n",
      "INFO:tensorflow:loss = 0.31822282, step = 10401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.329\n",
      "INFO:tensorflow:loss = 0.33519042, step = 10501 (0.529 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.606\n",
      "INFO:tensorflow:loss = 0.31088817, step = 10601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.616\n",
      "INFO:tensorflow:loss = 0.3240525, step = 10701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.34\n",
      "INFO:tensorflow:loss = 0.30363333, step = 10801 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.105\n",
      "INFO:tensorflow:loss = 0.31868708, step = 10901 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.126\n",
      "INFO:tensorflow:loss = 0.31593227, step = 11001 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.135\n",
      "INFO:tensorflow:loss = 0.27353743, step = 11101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.38\n",
      "INFO:tensorflow:loss = 0.2944389, step = 11201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.894\n",
      "INFO:tensorflow:loss = 0.29087844, step = 11301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.261\n",
      "INFO:tensorflow:loss = 0.29591876, step = 11401 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.588\n",
      "INFO:tensorflow:loss = 0.28439513, step = 11501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.668\n",
      "INFO:tensorflow:loss = 0.29671884, step = 11601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.294\n",
      "INFO:tensorflow:loss = 0.28747505, step = 11701 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.812\n",
      "INFO:tensorflow:loss = 0.272158, step = 11801 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.446\n",
      "INFO:tensorflow:loss = 0.28053755, step = 11901 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.195\n",
      "INFO:tensorflow:loss = 0.27900323, step = 12001 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.08\n",
      "INFO:tensorflow:loss = 0.2799347, step = 12101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.751\n",
      "INFO:tensorflow:loss = 0.27227372, step = 12201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.562\n",
      "INFO:tensorflow:loss = 0.26822528, step = 12301 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.224\n",
      "INFO:tensorflow:loss = 0.27416617, step = 12401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.816\n",
      "INFO:tensorflow:loss = 0.26567662, step = 12501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.822\n",
      "INFO:tensorflow:loss = 0.27067626, step = 12601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.429\n",
      "INFO:tensorflow:loss = 0.26506785, step = 12701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.246\n",
      "INFO:tensorflow:loss = 0.2557365, step = 12801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.932\n",
      "INFO:tensorflow:loss = 0.26395854, step = 12901 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.165\n",
      "INFO:tensorflow:loss = 0.2399247, step = 13001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.535\n",
      "INFO:tensorflow:loss = 0.24215487, step = 13101 (0.525 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.076\n",
      "INFO:tensorflow:loss = 0.25704417, step = 13201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.801\n",
      "INFO:tensorflow:loss = 0.2419834, step = 13301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.464\n",
      "INFO:tensorflow:loss = 0.25074604, step = 13401 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.211\n",
      "INFO:tensorflow:loss = 0.23982121, step = 13501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.787\n",
      "INFO:tensorflow:loss = 0.24150217, step = 13601 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.234\n",
      "INFO:tensorflow:loss = 0.2566658, step = 13701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.151\n",
      "INFO:tensorflow:loss = 0.2407999, step = 13801 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.636\n",
      "INFO:tensorflow:loss = 0.24465232, step = 13901 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.615\n",
      "INFO:tensorflow:loss = 0.2237285, step = 14001 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.078\n",
      "INFO:tensorflow:loss = 0.23061153, step = 14101 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.259\n",
      "INFO:tensorflow:loss = 0.22564183, step = 14201 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.752\n",
      "INFO:tensorflow:loss = 0.2426891, step = 14301 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.5\n",
      "INFO:tensorflow:loss = 0.23066261, step = 14401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.163\n",
      "INFO:tensorflow:loss = 0.22766887, step = 14501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.828\n",
      "INFO:tensorflow:loss = 0.24597591, step = 14601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.615\n",
      "INFO:tensorflow:loss = 0.22244298, step = 14701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.938\n",
      "INFO:tensorflow:loss = 0.23050547, step = 14801 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.29\n",
      "INFO:tensorflow:loss = 0.23576373, step = 14901 (0.521 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_10_w_0.25/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.21977699.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:57:12\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:57:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7559689, global_step = 15000, loss = 0.75968164\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:57:13\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.25/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:57:14\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7544717, global_step = 15000, loss = 0.73076737\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259aaa1f10>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_10_w_0.5', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:57:14\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:57:14\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.98, global_step = 20000, loss = 0.09247019\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.5/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_10_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.14957927, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 175.477\n",
      "INFO:tensorflow:loss = 0.1480473, step = 10101 (0.570 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.724\n",
      "INFO:tensorflow:loss = 0.16232257, step = 10201 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.359\n",
      "INFO:tensorflow:loss = 0.13733286, step = 10301 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.383\n",
      "INFO:tensorflow:loss = 0.14741358, step = 10401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.121\n",
      "INFO:tensorflow:loss = 0.14282282, step = 10501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.292\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:loss = 0.15368144, step = 10601 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.863\n",
      "INFO:tensorflow:loss = 0.14181226, step = 10701 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.727\n",
      "INFO:tensorflow:loss = 0.14353433, step = 10801 (0.522 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.261\n",
      "INFO:tensorflow:loss = 0.14192408, step = 10901 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.819\n",
      "INFO:tensorflow:loss = 0.13606174, step = 11001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.496\n",
      "INFO:tensorflow:loss = 0.14402232, step = 11101 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.059\n",
      "INFO:tensorflow:loss = 0.15437758, step = 11201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.65\n",
      "INFO:tensorflow:loss = 0.14781255, step = 11301 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.24\n",
      "INFO:tensorflow:loss = 0.14099224, step = 11401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.731\n",
      "INFO:tensorflow:loss = 0.13844961, step = 11501 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.59\n",
      "INFO:tensorflow:loss = 0.13654394, step = 11601 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.918\n",
      "INFO:tensorflow:loss = 0.13374616, step = 11701 (0.529 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.116\n",
      "INFO:tensorflow:loss = 0.13328469, step = 11801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.188\n",
      "INFO:tensorflow:loss = 0.13795292, step = 11901 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.983\n",
      "INFO:tensorflow:loss = 0.13830495, step = 12001 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.2\n",
      "INFO:tensorflow:loss = 0.13593519, step = 12101 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.307\n",
      "INFO:tensorflow:loss = 0.15027994, step = 12201 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.796\n",
      "INFO:tensorflow:loss = 0.13424353, step = 12301 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.687\n",
      "INFO:tensorflow:loss = 0.13693355, step = 12401 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.637\n",
      "INFO:tensorflow:loss = 0.13381934, step = 12501 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.674\n",
      "INFO:tensorflow:loss = 0.13403171, step = 12601 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.096\n",
      "INFO:tensorflow:loss = 0.13524127, step = 12701 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.157\n",
      "INFO:tensorflow:loss = 0.12350246, step = 12801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.729\n",
      "INFO:tensorflow:loss = 0.12888177, step = 12901 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.345\n",
      "INFO:tensorflow:loss = 0.12459424, step = 13001 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.775\n",
      "INFO:tensorflow:loss = 0.13185707, step = 13101 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.253\n",
      "INFO:tensorflow:loss = 0.1292159, step = 13201 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.285\n",
      "INFO:tensorflow:loss = 0.13884404, step = 13301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.947\n",
      "INFO:tensorflow:loss = 0.12895952, step = 13401 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.227\n",
      "INFO:tensorflow:loss = 0.12983157, step = 13501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.125\n",
      "INFO:tensorflow:loss = 0.12432665, step = 13601 (0.523 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.414\n",
      "INFO:tensorflow:loss = 0.13328066, step = 13701 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.535\n",
      "INFO:tensorflow:loss = 0.135057, step = 13801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.765\n",
      "INFO:tensorflow:loss = 0.14151773, step = 13901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.889\n",
      "INFO:tensorflow:loss = 0.12754649, step = 14001 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.748\n",
      "INFO:tensorflow:loss = 0.13837168, step = 14101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.06\n",
      "INFO:tensorflow:loss = 0.13313909, step = 14201 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.28\n",
      "INFO:tensorflow:loss = 0.13136499, step = 14301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.558\n",
      "INFO:tensorflow:loss = 0.124394216, step = 14401 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.155\n",
      "INFO:tensorflow:loss = 0.13175091, step = 14501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.797\n",
      "INFO:tensorflow:loss = 0.12775871, step = 14601 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.03\n",
      "INFO:tensorflow:loss = 0.121732876, step = 14701 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.595\n",
      "INFO:tensorflow:loss = 0.12623926, step = 14801 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.964\n",
      "INFO:tensorflow:loss = 0.12674162, step = 14901 (0.526 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_10_w_0.5/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.12694696.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:57:41\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.5/model.ckpt-15000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1243:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:57:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.8022008, global_step = 15000, loss = 0.7228941\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:57:43\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.5/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:57:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.8143719, global_step = 15000, loss = 0.66707253\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f259ab9f4d0>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_10_w_0.75', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:57:44\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:57:44\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.93, global_step = 20000, loss = 0.2841084\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.75/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_10_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.14965421, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 153.453\n",
      "INFO:tensorflow:loss = 0.1264433, step = 10101 (0.654 sec)\n",
      "INFO:tensorflow:global_step/sec: 173.302\n",
      "INFO:tensorflow:loss = 0.10711712, step = 10201 (0.574 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.903\n",
      "INFO:tensorflow:loss = 0.13087815, step = 10301 (0.529 sec)\n",
      "INFO:tensorflow:global_step/sec: 190.753\n",
      "INFO:tensorflow:loss = 0.121764556, step = 10401 (0.524 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.8\n",
      "INFO:tensorflow:loss = 0.12464474, step = 10501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.183\n",
      "INFO:tensorflow:loss = 0.098118775, step = 10601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.593\n",
      "INFO:tensorflow:loss = 0.10983523, step = 10701 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.487\n",
      "INFO:tensorflow:loss = 0.10088968, step = 10801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.286\n",
      "INFO:tensorflow:loss = 0.083215535, step = 10901 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.768\n",
      "INFO:tensorflow:loss = 0.1119114, step = 11001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.328\n",
      "INFO:tensorflow:loss = 0.100256376, step = 11101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.039\n",
      "INFO:tensorflow:loss = 0.096023, step = 11201 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.09\n",
      "INFO:tensorflow:loss = 0.098194346, step = 11301 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.625\n",
      "INFO:tensorflow:loss = 0.0996468, step = 11401 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.58\n",
      "INFO:tensorflow:loss = 0.11543636, step = 11501 (0.531 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.098\n",
      "INFO:tensorflow:loss = 0.093187705, step = 11601 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.577\n",
      "INFO:tensorflow:loss = 0.09077543, step = 11701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.977\n",
      "INFO:tensorflow:loss = 0.09448737, step = 11801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.48\n",
      "INFO:tensorflow:loss = 0.09752302, step = 11901 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.58\n",
      "INFO:tensorflow:loss = 0.09522472, step = 12001 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.167\n",
      "INFO:tensorflow:loss = 0.08506856, step = 12101 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.717\n",
      "INFO:tensorflow:loss = 0.080875896, step = 12201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.83\n",
      "INFO:tensorflow:loss = 0.094108686, step = 12301 (0.519 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.205\n",
      "INFO:tensorflow:loss = 0.080534585, step = 12401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.466\n",
      "INFO:tensorflow:loss = 0.08210379, step = 12501 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.366\n",
      "INFO:tensorflow:loss = 0.08676193, step = 12601 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.161\n",
      "INFO:tensorflow:loss = 0.076817036, step = 12701 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.55\n",
      "INFO:tensorflow:loss = 0.085825495, step = 12801 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.634\n",
      "INFO:tensorflow:loss = 0.083284274, step = 12901 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.306\n",
      "INFO:tensorflow:loss = 0.091642834, step = 13001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.007\n",
      "INFO:tensorflow:loss = 0.083203085, step = 13101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.019\n",
      "INFO:tensorflow:loss = 0.08116089, step = 13201 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.234\n",
      "INFO:tensorflow:loss = 0.08220953, step = 13301 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.564\n",
      "INFO:tensorflow:loss = 0.07570483, step = 13401 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.947\n",
      "INFO:tensorflow:loss = 0.08060023, step = 13501 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.303\n",
      "INFO:tensorflow:loss = 0.079865195, step = 13601 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.107\n",
      "INFO:tensorflow:loss = 0.07341011, step = 13701 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.744\n",
      "INFO:tensorflow:loss = 0.08147618, step = 13801 (0.508 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.967\n",
      "INFO:tensorflow:loss = 0.08581245, step = 13901 (0.527 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.961\n",
      "INFO:tensorflow:loss = 0.07954316, step = 14001 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 189.376\n",
      "INFO:tensorflow:loss = 0.084716275, step = 14101 (0.528 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.969\n",
      "INFO:tensorflow:loss = 0.08059792, step = 14201 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.873\n",
      "INFO:tensorflow:loss = 0.08350904, step = 14301 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.763\n",
      "INFO:tensorflow:loss = 0.08009761, step = 14401 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.631\n",
      "INFO:tensorflow:loss = 0.08686681, step = 14501 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.945\n",
      "INFO:tensorflow:loss = 0.086272225, step = 14601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.516\n",
      "INFO:tensorflow:loss = 0.07341735, step = 14701 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 187.253\n",
      "INFO:tensorflow:loss = 0.07507533, step = 14801 (0.534 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.302\n",
      "INFO:tensorflow:loss = 0.07658172, step = 14901 (0.515 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_10_w_0.75/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.07530098.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:58:11\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1253:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:58:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.8176908, global_step = 15000, loss = 0.75486296\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:58:13\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_0.75/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:58:13\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.82570714, global_step = 15000, loss = 0.7271139\n",
      "INFO:tensorflow:Using default config.\n",
      "INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_task_type': 'worker', '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f25904b3ad0>, '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_service': None, '_num_ps_replicas': 0, '_tf_random_seed': None, '_master': '', '_num_worker_replicas': 1, '_task_id': 0, '_log_step_count_steps': 100, '_model_dir': './mnist_convnet_student_mod_t_10_w_1', '_save_summary_steps': 100}\n",
      "{'x': <tf.Tensor 'Const:0' shape=(100, 784) dtype=float32>} Tensor(\"Const_1:0\", shape=(100,), dtype=int32, device=/device:CPU:0)\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:58:13\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:58:14\n",
      "INFO:tensorflow:Saving dict for global step 20000: accuracy = 0.97, global_step = 20000, loss = 0.09986788\n",
      "_THIS_ <type 'dict'>\n",
      "INFO:tensorflow:Restoring parameters from /tmp/mnist_convnet_model_test1/model.ckpt-20000\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_1/model.ckpt-10000\n",
      "INFO:tensorflow:Saving checkpoints for 10001 into ./mnist_convnet_student_mod_t_10_w_1/model.ckpt.\n",
      "INFO:tensorflow:loss = 0.024309572, step = 10001\n",
      "INFO:tensorflow:global_step/sec: 148.168\n",
      "INFO:tensorflow:loss = 0.041315097, step = 10101 (0.676 sec)\n",
      "INFO:tensorflow:global_step/sec: 183.922\n",
      "INFO:tensorflow:loss = 0.054080665, step = 10201 (0.544 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.757\n",
      "INFO:tensorflow:loss = 0.048089046, step = 10301 (0.517 sec)\n",
      "INFO:tensorflow:global_step/sec: 186.349\n",
      "INFO:tensorflow:loss = 0.027929101, step = 10401 (0.537 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.889\n",
      "INFO:tensorflow:loss = 0.02586631, step = 10501 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.908\n",
      "INFO:tensorflow:loss = 0.017971704, step = 10601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.865\n",
      "INFO:tensorflow:loss = 0.035953015, step = 10701 (0.505 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.075\n",
      "INFO:tensorflow:loss = 0.033260044, step = 10801 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.878\n",
      "INFO:tensorflow:loss = 0.024114655, step = 10901 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.058\n",
      "INFO:tensorflow:loss = 0.020636711, step = 11001 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.571\n",
      "INFO:tensorflow:loss = 0.045679346, step = 11101 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.594\n",
      "INFO:tensorflow:loss = 0.014061174, step = 11201 (0.506 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.84\n",
      "INFO:tensorflow:loss = 0.018608864, step = 11301 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.646\n",
      "INFO:tensorflow:loss = 0.036862407, step = 11401 (0.506 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.924\n",
      "INFO:tensorflow:loss = 0.031744827, step = 11501 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.908\n",
      "INFO:tensorflow:loss = 0.036127295, step = 11601 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.844\n",
      "INFO:tensorflow:loss = 0.02762023, step = 11701 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.295\n",
      "INFO:tensorflow:loss = 0.020935982, step = 11801 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.512\n",
      "INFO:tensorflow:loss = 0.022290535, step = 11901 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.142\n",
      "INFO:tensorflow:loss = 0.022578472, step = 12001 (0.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.176\n",
      "INFO:tensorflow:loss = 0.022925314, step = 12101 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.637\n",
      "INFO:tensorflow:loss = 0.012604851, step = 12201 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 197.168\n",
      "INFO:tensorflow:loss = 0.016512152, step = 12301 (0.507 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.326\n",
      "INFO:tensorflow:loss = 0.033166945, step = 12401 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.328\n",
      "INFO:tensorflow:loss = 0.032219686, step = 12501 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.528\n",
      "INFO:tensorflow:loss = 0.012321171, step = 12601 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.252\n",
      "INFO:tensorflow:loss = 0.03036734, step = 12701 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.222\n",
      "INFO:tensorflow:loss = 0.016168606, step = 12801 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 188.865\n",
      "INFO:tensorflow:loss = 0.025682725, step = 12901 (0.530 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.858\n",
      "INFO:tensorflow:loss = 0.0063237376, step = 13001 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.109\n",
      "INFO:tensorflow:loss = 0.012255849, step = 13101 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.946\n",
      "INFO:tensorflow:loss = 0.016867895, step = 13201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.372\n",
      "INFO:tensorflow:loss = 0.0164818, step = 13301 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.075\n",
      "INFO:tensorflow:loss = 0.015938083, step = 13401 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.358\n",
      "INFO:tensorflow:loss = 0.017238773, step = 13501 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.096\n",
      "INFO:tensorflow:loss = 0.025633885, step = 13601 (0.521 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.541\n",
      "INFO:tensorflow:loss = 0.026803197, step = 13701 (0.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 192.186\n",
      "INFO:tensorflow:loss = 0.039834704, step = 13801 (0.520 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.684\n",
      "INFO:tensorflow:loss = 0.009973731, step = 13901 (0.516 sec)\n",
      "INFO:tensorflow:global_step/sec: 194.492\n",
      "INFO:tensorflow:loss = 0.016570527, step = 14001 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.935\n",
      "INFO:tensorflow:loss = 0.010094966, step = 14101 (0.515 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.558\n",
      "INFO:tensorflow:loss = 0.010418793, step = 14201 (0.511 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.095\n",
      "INFO:tensorflow:loss = 0.009279108, step = 14301 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.315\n",
      "INFO:tensorflow:loss = 0.012164373, step = 14401 (0.509 sec)\n",
      "INFO:tensorflow:global_step/sec: 195.165\n",
      "INFO:tensorflow:loss = 0.012105482, step = 14501 (0.512 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.009\n",
      "INFO:tensorflow:loss = 0.029262094, step = 14601 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 193.017\n",
      "INFO:tensorflow:loss = 0.010652008, step = 14701 (0.518 sec)\n",
      "INFO:tensorflow:global_step/sec: 196.095\n",
      "INFO:tensorflow:loss = 0.01611799, step = 14801 (0.510 sec)\n",
      "INFO:tensorflow:global_step/sec: 191.671\n",
      "INFO:tensorflow:loss = 0.0099984035, step = 14901 (0.522 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 15000 into ./mnist_convnet_student_mod_t_10_w_1/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 0.031792164.\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:58:41\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1263:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:58:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.7854658, global_step = 15000, loss = 0.8922863\n",
      "{'x': <tf.Tensor 'fifo_queue_DequeueUpTo:1' shape=(?, 784) dtype=float32>} Tensor(\"fifo_queue_DequeueUpTo:2\", shape=(?,), dtype=int32, device=/device:CPU:0)\n",
      "ERROR:tensorflow:Exception in QueueRunner: Attempted to use a closed Session.\n",
      "INFO:tensorflow:Starting evaluation at 2018-05-06-18:58:43\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception in thread Thread-1266:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 801, in __bootstrap_inner\n",
      "    self.run()\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/threading.py\", line 754, in run\n",
      "    self.__target(*self.__args, **self.__kwargs)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/estimator/inputs/queues/feeding_queue_runner.py\", line 94, in _run\n",
      "    sess.run(enqueue_op, feed_dict=feed_dict)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 895, in run\n",
      "    run_metadata_ptr)\n",
      "  File \"/home/sanjay/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1053, in _run\n",
      "    raise RuntimeError('Attempted to use a closed Session.')\n",
      "RuntimeError: Attempted to use a closed Session.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./mnist_convnet_student_mod_t_10_w_1/model.ckpt-15000\n",
      "INFO:tensorflow:Finished evaluation at 2018-05-06-18:58:43\n",
      "INFO:tensorflow:Saving dict for global step 15000: accuracy = 0.79242927, global_step = 15000, loss = 0.8824561\n"
     ]
    }
   ],
   "source": [
    "mnist_classifier = tf.estimator.Estimator(\n",
    "  model_fn=cnn_model_fn, model_dir=\"./models/mnist_convnet_model_test1\")\n",
    "\n",
    "mnist = tf.contrib.learn.datasets.load_dataset(\"mnist\")\n",
    "train_data = mnist.train.images  # Returns np.array\n",
    "train_labels = np.asarray(mnist.train.labels, dtype=np.int32)\n",
    "eval_data = mnist.test.images  # Returns np.array\n",
    "eval_labels = np.asarray(mnist.test.labels, dtype=np.int32)\n",
    "temp_list=[1,2,5,10]\n",
    "loss_weight_list=[0,0.25,0.5,0.75,1]\n",
    "p=-1\n",
    "l=-1\n",
    "def getdatap():\n",
    "    global p\n",
    "    p += 1\n",
    "    if p>len(train_labels)/100:\n",
    "        p=0\n",
    "    data=train_data[p*100:(p+1)*100]\n",
    "    labels=train_labels[p*100:(p+1)*100]\n",
    "    \n",
    "    datatensor=tf.convert_to_tensor(data)\n",
    "    labeltensor=tf.convert_to_tensor(labels)\n",
    "    \n",
    "    \n",
    "    return {'x':datatensor},labeltensor\n",
    "\n",
    "out = open('swapout4.csv', 'w')\n",
    "outlog = open('out_log.csv', 'w')\n",
    "\n",
    "for i in temp_list:\n",
    "    for j in loss_weight_list:\n",
    "        temprature=i\n",
    "        loss_weight=j\n",
    "        # Create the Estimator\n",
    "        student_classifier = tf.estimator.Estimator(\n",
    "          model_fn=student_model_fn, model_dir=\"./models/mnist_convnet_student_mod_t_\"+str(i)+\"_w_\"+str(j))\n",
    "\n",
    "        \n",
    "#         eval_train_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "#           x={\"x\": train_data[:100]},\n",
    "#           y=train_labels[:100],\n",
    "#           num_epochs=1,\n",
    "#           batch_size=100,\n",
    "#           shuffle=False)\n",
    "#         train_result=mnist_classifier.evaluate(input_fn=eval_train_fn)\n",
    "#         #print(\"\\n \\n \\n Train 100\",train_result)\n",
    "#         print(\"\\n\\n\\nLABEL FUNCTION\",train_labels[:100])\n",
    "#         print(\"\\n\\n\\nTRAIN DATA FUNCTION\",train_data[0])\n",
    "        \n",
    "        # Train the model\n",
    "#         train_input_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "#           x={\"x\": train_data},\n",
    "#           y=train_labels,\n",
    "#           batch_size=100,\n",
    "#           num_epochs=None,\n",
    "#           shuffle=True)\n",
    "#         print(train_data,train_labels,eval_data,eval_labels)\n",
    "\n",
    "        \n",
    "        student_classifier.train(\n",
    "          input_fn=getdatap,\n",
    "          steps=5000,\n",
    "          hooks=None)\n",
    "        # print(\"eva\",type(eval_data))\n",
    "\n",
    "        # Evaluate the model and print results\n",
    "        eval_train_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "          x={\"x\": train_data},\n",
    "          y=train_labels,\n",
    "          shuffle=False)\n",
    "        train_result=student_classifier.evaluate(input_fn=eval_train_fn)\n",
    "        \n",
    "        out.write('%f;' % i)\n",
    "        out.write('%f;' % j)\n",
    "        out.write('%f;' % train_result['accuracy'])\n",
    "        out.write('%f;' % train_result['loss'])\n",
    "        \n",
    "        eval_input_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "          x={\"x\": eval_data},\n",
    "          y=eval_labels,\n",
    "          num_epochs=1,\n",
    "          shuffle=False)\n",
    "        eval_result=student_classifier.evaluate(input_fn=eval_input_fn)\n",
    "        # output=list(predictions)\n",
    "        \n",
    "        out.write('%f;' % eval_result['accuracy'])\n",
    "        out.write('%f;' % eval_result['loss'])\n",
    "#         acc=eval_result\n",
    "#         out.write(eval_result)\n",
    "        out.write('\\n')\n",
    "        \n",
    "out.close()\n",
    "outlog.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "eval_input_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "  x={\"x\": eval_data},\n",
    "  y=eval_labels,\n",
    "  shuffle=False)\n",
    "eval_result=student_classifier.evaluate(input_fn=eval_input_fn)\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# loss1 + lamba * loss2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    " \n",
    "def custom_loss(y_true, pred_logits, tar_soft_t, pred_logits_t):\n",
    "#    print(tf.losses.sparse_softmax_cross_entropy(labels = y_true, logits = pred_logits))\n",
    "#    print(l * cross_entropy2(tar_soft_t, pred_logits_t))\n",
    "   return tf.losses.sparse_softmax_cross_entropy(labels = y_true, logits = pred_logits) + lam * cross_entropy2(tar_soft_t, pred_logits_t)\n",
    "mnist_classifier = tf.estimator.Estimator(\n",
    "  model_fn=cnn_model_fn, model_dir=\"./models/mnist_convnet_model_test1\")\n",
    "\n",
    "mnist = tf.contrib.learn.datasets.load_dataset(\"mnist\")\n",
    "train_data = mnist.train.images  # Returns np.array\n",
    "train_labels = np.asarray(mnist.train.labels, dtype=np.int32)\n",
    "eval_data = mnist.test.images  # Returns np.array\n",
    "eval_labels = np.asarray(mnist.test.labels, dtype=np.int32)\n",
    "lam_list=[0,1,2,4,6,8,9,10]\n",
    "temprature=2\n",
    "out = open('out_lam.csv', 'w')\n",
    "for i in lam_list:\n",
    "    lam=i\n",
    "    # Create the Estimator\n",
    "    student_classifier = tf.estimator.Estimator(\n",
    "      model_fn=student_model_fn, model_dir=\"./models/mnist_convnet_student_t_\"+str(2)+\"_lam_\"+str(i))\n",
    "\n",
    "    # Train the model\n",
    "    train_input_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "      x={\"x\": train_data},\n",
    "      y=train_labels,\n",
    "      batch_size=100,\n",
    "      num_epochs=None,\n",
    "      shuffle=True)\n",
    "    print(train_data,train_labels,eval_data,eval_labels)\n",
    "\n",
    "    student_classifier.train(\n",
    "      input_fn=train_input_fn,\n",
    "      steps=20000,\n",
    "      hooks=None)\n",
    "    # print(\"eva\",type(eval_data))\n",
    "\n",
    "    # Evaluate the model and print results\n",
    "    eval_train_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "      x={\"x\": train_data},\n",
    "      y=train_labels,\n",
    "      num_epochs=1,\n",
    "      shuffle=False)\n",
    "    train_result=student_classifier.evaluate(input_fn=eval_train_fn)\n",
    "\n",
    "    out.write('%f;' % i)\n",
    "    out.write('%f;' % j)\n",
    "    out.write('%f;' % train_result['accuracy'])\n",
    "    out.write('%f;' % train_result['loss'])\n",
    "\n",
    "    eval_input_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "      x={\"x\": eval_data},\n",
    "      y=eval_labels,\n",
    "      num_epochs=1,\n",
    "      shuffle=False)\n",
    "    eval_result=student_classifier.evaluate(input_fn=eval_input_fn)\n",
    "    # output=list(predictions)\n",
    "\n",
    "    out.write('%f;' % eval_result['accuracy'])\n",
    "    out.write('%f;' % eval_result['loss'])\n",
    "#         acc=eval_result\n",
    "#         out.write(eval_result)\n",
    "    out.write('\\n')\n",
    "\n",
    "out.close()\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "eval_input_fn = tf.estimator.inputs.numpy_input_fn(\n",
    "  x={\"x\": train_data},\n",
    "  y=train_labels,\n",
    "  batch_size=100,\n",
    "  num_epochs=10,\n",
    "  shuffle=False)\n",
    "eval_result=mnist_classifier.evaluate(input_fn=eval_input_fn)\n",
    " "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}