Predicting TEXT LSTM .ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers.core import TimeDistributedDense, Activation\n",
    "from keras.layers.recurrent import LSTM\n",
    "from keras.optimizers import RMSprop\n",
    "import numpy as np\n",
    "import random\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['a', 'little', 'baby', 'plays', 'croquet']\n"
     ]
    }
   ],
   "source": [
    "captions = [\"a little baby plays croquet\".split()]\n",
    "print captions[0]\n",
    "wordList = []\n",
    "for word in captions[0]:\n",
    "  wordList.append(word)\n",
    "#list(set()) is slow for large datasets, use bloom filters for those\n",
    "wordList = list(set(wordList))\n",
    "#insert system start token into dictionary\n",
    "wordList.insert(0, '#START#')\n",
    "wordList.append(\"#END#\")\n",
    "\n",
    "#convert words into one-hot tokens\n",
    "wordtoix = {}\n",
    "ixtoword = {}\n",
    "for i, word in enumerate(wordList):\n",
    "  wordtoix[word] = i\n",
    "  ixtoword[i] = word"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Building Training Data\n",
      "6\n",
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "5\n"
     ]
    }
   ],
   "source": [
    "print 'Building Training Data'\n",
    "#array has shape of num_seq X num_time_steps X dimensionality of features\n",
    "ins = np.zeros( (1, len(captions[0])+1, len(wordList)) )\n",
    "gts = np.zeros_like(ins)\n",
    "\n",
    "print len(captions[0])+1\n",
    "#encode the poem into the training sequences\n",
    "#ins begins with start\n",
    "ins[0, 0, wordtoix['#START#']] = 1\n",
    "print 0\n",
    "for t, word in enumerate(captions[0]):\n",
    "  #the ground truth at time t is the next word\n",
    "  gts[0, t, wordtoix[word]] = 1\n",
    "  #the input at time t+1 is the previous ground truth\n",
    "  ins[0, t+1, wordtoix[word]] = 1\n",
    "  print t+1\n",
    "#ground truth ends with the end token (or reuse start token)\n",
    "print t+1\n",
    "gts[:,t+1,wordtoix['#END#']] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Building Model\n"
     ]
    }
   ],
   "source": [
    "INP_DIM = len(wordList)\n",
    "print(\"Building Model\")\n",
    "model = Sequential()\n",
    "model.add(TimeDistributedDense(input_dim = INP_DIM, output_dim=64))\n",
    "model.add(LSTM(input_dim=64, output_dim=64, forget_bias_init='one',return_sequences=True))\n",
    "model.add(TimeDistributedDense(input_dim = INP_DIM, output_dim=INP_DIM))\n",
    "model.add(Activation('softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "rms = RMSprop(lr=0.03)\n",
    "model.compile(loss='categorical_crossentropy', optimizer=rms)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/150\n",
      "1/1 [==============================] - 0s - loss: 1.9490\n",
      "Epoch 2/150\n",
      "1/1 [==============================] - 0s - loss: 0.9513\n",
      "Epoch 3/150\n",
      "1/1 [==============================] - 0s - loss: 1.6461\n",
      "Epoch 4/150\n",
      "1/1 [==============================] - 0s - loss: 0.2363\n",
      "Epoch 5/150\n",
      "1/1 [==============================] - 0s - loss: 0.0273\n",
      "Epoch 6/150\n",
      "1/1 [==============================] - 0s - loss: 0.0168\n",
      "Epoch 7/150\n",
      "1/1 [==============================] - 0s - loss: 0.0126\n",
      "Epoch 8/150\n",
      "1/1 [==============================] - 0s - loss: 0.0101\n",
      "Epoch 9/150\n",
      "1/1 [==============================] - 0s - loss: 0.0084\n",
      "Epoch 10/150\n",
      "1/1 [==============================] - 0s - loss: 0.0072\n",
      "Epoch 11/150\n",
      "1/1 [==============================] - 0s - loss: 0.0063\n",
      "Epoch 12/150\n",
      "1/1 [==============================] - 0s - loss: 0.0055\n",
      "Epoch 13/150\n",
      "1/1 [==============================] - 0s - loss: 0.0049\n",
      "Epoch 14/150\n",
      "1/1 [==============================] - 0s - loss: 0.0044\n",
      "Epoch 15/150\n",
      "1/1 [==============================] - 0s - loss: 0.0039\n",
      "Epoch 16/150\n",
      "1/1 [==============================] - 0s - loss: 0.0035\n",
      "Epoch 17/150\n",
      "1/1 [==============================] - 0s - loss: 0.0032\n",
      "Epoch 18/150\n",
      "1/1 [==============================] - 0s - loss: 0.0029\n",
      "Epoch 19/150\n",
      "1/1 [==============================] - 0s - loss: 0.0027\n",
      "Epoch 20/150\n",
      "1/1 [==============================] - 0s - loss: 0.0025\n",
      "Epoch 21/150\n",
      "1/1 [==============================] - 0s - loss: 0.0023\n",
      "Epoch 22/150\n",
      "1/1 [==============================] - 0s - loss: 0.0021\n",
      "Epoch 23/150\n",
      "1/1 [==============================] - 0s - loss: 0.0020\n",
      "Epoch 24/150\n",
      "1/1 [==============================] - 0s - loss: 0.0018\n",
      "Epoch 25/150\n",
      "1/1 [==============================] - 0s - loss: 0.0017\n",
      "Epoch 26/150\n",
      "1/1 [==============================] - 0s - loss: 0.0016\n",
      "Epoch 27/150\n",
      "1/1 [==============================] - 0s - loss: 0.0015\n",
      "Epoch 28/150\n",
      "1/1 [==============================] - 0s - loss: 0.0014\n",
      "Epoch 29/150\n",
      "1/1 [==============================] - 0s - loss: 0.0013\n",
      "Epoch 30/150\n",
      "1/1 [==============================] - 0s - loss: 0.0012\n",
      "Epoch 31/150\n",
      "1/1 [==============================] - 0s - loss: 0.0011\n",
      "Epoch 32/150\n",
      "1/1 [==============================] - 0s - loss: 0.0011\n",
      "Epoch 33/150\n",
      "1/1 [==============================] - 0s - loss: 0.0010\n",
      "Epoch 34/150\n",
      "1/1 [==============================] - 0s - loss: 9.4188e-04\n",
      "Epoch 35/150\n",
      "1/1 [==============================] - 0s - loss: 8.8520e-04\n",
      "Epoch 36/150\n",
      "1/1 [==============================] - 0s - loss: 8.3192e-04\n",
      "Epoch 37/150\n",
      "1/1 [==============================] - 0s - loss: 7.8176e-04\n",
      "Epoch 38/150\n",
      "1/1 [==============================] - 0s - loss: 7.3456e-04\n",
      "Epoch 39/150\n",
      "1/1 [==============================] - 0s - loss: 6.9018e-04\n",
      "Epoch 40/150\n",
      "1/1 [==============================] - 0s - loss: 6.4839e-04\n",
      "Epoch 41/150\n",
      "1/1 [==============================] - 0s - loss: 6.0977e-04\n",
      "Epoch 42/150\n",
      "1/1 [==============================] - 0s - loss: 5.7352e-04\n",
      "Epoch 43/150\n",
      "1/1 [==============================] - 0s - loss: 5.3943e-04\n",
      "Epoch 44/150\n",
      "1/1 [==============================] - 0s - loss: 5.0741e-04\n",
      "Epoch 45/150\n",
      "1/1 [==============================] - 0s - loss: 4.7794e-04\n",
      "Epoch 46/150\n",
      "1/1 [==============================] - 0s - loss: 4.5009e-04\n",
      "Epoch 47/150\n",
      "1/1 [==============================] - 0s - loss: 4.2411e-04\n",
      "Epoch 48/150\n",
      "1/1 [==============================] - 0s - loss: 3.9985e-04\n",
      "Epoch 49/150\n",
      "1/1 [==============================] - 0s - loss: 3.7680e-04\n",
      "Epoch 50/150\n",
      "1/1 [==============================] - 0s - loss: 3.5534e-04\n",
      "Epoch 51/150\n",
      "1/1 [==============================] - 0s - loss: 3.3496e-04\n",
      "Epoch 52/150\n",
      "1/1 [==============================] - 0s - loss: 3.1578e-04\n",
      "Epoch 53/150\n",
      "1/1 [==============================] - 0s - loss: 2.9755e-04\n",
      "Epoch 54/150\n",
      "1/1 [==============================] - 0s - loss: 2.8027e-04\n",
      "Epoch 55/150\n",
      "1/1 [==============================] - 0s - loss: 2.6391e-04\n",
      "Epoch 56/150\n",
      "1/1 [==============================] - 0s - loss: 2.4879e-04\n",
      "Epoch 57/150\n",
      "1/1 [==============================] - 0s - loss: 2.3461e-04\n",
      "Epoch 58/150\n",
      "1/1 [==============================] - 0s - loss: 2.2121e-04\n",
      "Epoch 59/150\n",
      "1/1 [==============================] - 0s - loss: 2.0877e-04\n",
      "Epoch 60/150\n",
      "1/1 [==============================] - 0s - loss: 1.9706e-04\n",
      "Epoch 61/150\n",
      "1/1 [==============================] - 0s - loss: 1.8584e-04\n",
      "Epoch 62/150\n",
      "1/1 [==============================] - 0s - loss: 1.7530e-04\n",
      "Epoch 63/150\n",
      "1/1 [==============================] - 0s - loss: 1.6538e-04\n",
      "Epoch 64/150\n",
      "1/1 [==============================] - 0s - loss: 1.5594e-04\n",
      "Epoch 65/150\n",
      "1/1 [==============================] - 0s - loss: 1.4706e-04\n",
      "Epoch 66/150\n",
      "1/1 [==============================] - 0s - loss: 1.3874e-04\n",
      "Epoch 67/150\n",
      "1/1 [==============================] - 0s - loss: 1.3093e-04\n",
      "Epoch 68/150\n",
      "1/1 [==============================] - 0s - loss: 1.2385e-04\n",
      "Epoch 69/150\n",
      "1/1 [==============================] - 0s - loss: 1.1715e-04\n",
      "Epoch 70/150\n",
      "1/1 [==============================] - 0s - loss: 1.1081e-04\n",
      "Epoch 71/150\n",
      "1/1 [==============================] - 0s - loss: 1.0479e-04\n",
      "Epoch 72/150\n",
      "1/1 [==============================] - 0s - loss: 9.9108e-05\n",
      "Epoch 73/150\n",
      "1/1 [==============================] - 0s - loss: 9.3683e-05\n",
      "Epoch 74/150\n",
      "1/1 [==============================] - 0s - loss: 8.8537e-05\n",
      "Epoch 75/150\n",
      "1/1 [==============================] - 0s - loss: 8.3728e-05\n",
      "Epoch 76/150\n",
      "1/1 [==============================] - 0s - loss: 7.9178e-05\n",
      "Epoch 77/150\n",
      "1/1 [==============================] - 0s - loss: 7.4906e-05\n",
      "Epoch 78/150\n",
      "1/1 [==============================] - 0s - loss: 7.0833e-05\n",
      "Epoch 79/150\n",
      "1/1 [==============================] - 0s - loss: 6.6998e-05\n",
      "Epoch 80/150\n",
      "1/1 [==============================] - 0s - loss: 6.3362e-05\n",
      "Epoch 81/150\n",
      "1/1 [==============================] - 0s - loss: 5.9944e-05\n",
      "Epoch 82/150\n",
      "1/1 [==============================] - 0s - loss: 5.6725e-05\n",
      "Epoch 83/150\n",
      "1/1 [==============================] - 0s - loss: 5.3745e-05\n",
      "Epoch 84/150\n",
      "1/1 [==============================] - 0s - loss: 5.0884e-05\n",
      "Epoch 85/150\n",
      "1/1 [==============================] - 0s - loss: 4.8142e-05\n",
      "Epoch 86/150\n",
      "1/1 [==============================] - 0s - loss: 4.5599e-05\n",
      "Epoch 87/150\n",
      "1/1 [==============================] - 0s - loss: 4.3194e-05\n",
      "Epoch 88/150\n",
      "1/1 [==============================] - 0s - loss: 4.0890e-05\n",
      "Epoch 89/150\n",
      "1/1 [==============================] - 0s - loss: 3.8704e-05\n",
      "Epoch 90/150\n",
      "1/1 [==============================] - 0s - loss: 3.6638e-05\n",
      "Epoch 91/150\n",
      "1/1 [==============================] - 0s - loss: 3.4671e-05\n",
      "Epoch 92/150\n",
      "1/1 [==============================] - 0s - loss: 3.2823e-05\n",
      "Epoch 93/150\n",
      "1/1 [==============================] - 0s - loss: 3.1055e-05\n",
      "Epoch 94/150\n",
      "1/1 [==============================] - 0s - loss: 2.9386e-05\n",
      "Epoch 95/150\n",
      "1/1 [==============================] - 0s - loss: 2.7796e-05\n",
      "Epoch 96/150\n",
      "1/1 [==============================] - 0s - loss: 2.6306e-05\n",
      "Epoch 97/150\n",
      "1/1 [==============================] - 0s - loss: 2.4915e-05\n",
      "Epoch 98/150\n",
      "1/1 [==============================] - 0s - loss: 2.3544e-05\n",
      "Epoch 99/150\n",
      "1/1 [==============================] - 0s - loss: 2.2273e-05\n",
      "Epoch 100/150\n",
      "1/1 [==============================] - 0s - loss: 2.1061e-05\n",
      "Epoch 101/150\n",
      "1/1 [==============================] - 0s - loss: 1.9908e-05\n",
      "Epoch 102/150\n",
      "1/1 [==============================] - 0s - loss: 1.8835e-05\n",
      "Epoch 103/150\n",
      "1/1 [==============================] - 0s - loss: 1.7822e-05\n",
      "Epoch 104/150\n",
      "1/1 [==============================] - 0s - loss: 1.6829e-05\n",
      "Epoch 105/150\n",
      "1/1 [==============================] - 0s - loss: 1.5915e-05\n",
      "Epoch 106/150\n",
      "1/1 [==============================] - 0s - loss: 1.5040e-05\n",
      "Epoch 107/150\n",
      "1/1 [==============================] - 0s - loss: 1.4265e-05\n",
      "Epoch 108/150\n",
      "1/1 [==============================] - 0s - loss: 1.3491e-05\n",
      "Epoch 109/150\n",
      "1/1 [==============================] - 0s - loss: 1.2755e-05\n",
      "Epoch 110/150\n",
      "1/1 [==============================] - 0s - loss: 1.2080e-05\n",
      "Epoch 111/150\n",
      "1/1 [==============================] - 0s - loss: 1.1424e-05\n",
      "Epoch 112/150\n",
      "1/1 [==============================] - 0s - loss: 1.0808e-05\n",
      "Epoch 113/150\n",
      "1/1 [==============================] - 0s - loss: 1.0212e-05\n",
      "Epoch 114/150\n",
      "1/1 [==============================] - 0s - loss: 9.6759e-06\n",
      "Epoch 115/150\n",
      "1/1 [==============================] - 0s - loss: 9.1394e-06\n",
      "Epoch 116/150\n",
      "1/1 [==============================] - 0s - loss: 8.6228e-06\n",
      "Epoch 117/150\n",
      "1/1 [==============================] - 0s - loss: 8.1857e-06\n",
      "Epoch 118/150\n",
      "1/1 [==============================] - 0s - loss: 7.7685e-06\n",
      "Epoch 119/150\n",
      "1/1 [==============================] - 0s - loss: 7.3314e-06\n",
      "Epoch 120/150\n",
      "1/1 [==============================] - 0s - loss: 6.9340e-06\n",
      "Epoch 121/150\n",
      "1/1 [==============================] - 0s - loss: 6.5565e-06\n",
      "Epoch 122/150\n",
      "1/1 [==============================] - 0s - loss: 6.2188e-06\n",
      "Epoch 123/150\n",
      "1/1 [==============================] - 0s - loss: 5.9009e-06\n",
      "Epoch 124/150\n",
      "1/1 [==============================] - 0s - loss: 5.5830e-06\n",
      "Epoch 125/150\n",
      "1/1 [==============================] - 0s - loss: 5.3048e-06\n",
      "Epoch 126/150\n",
      "1/1 [==============================] - 0s - loss: 5.0068e-06\n",
      "Epoch 127/150\n",
      "1/1 [==============================] - 0s - loss: 4.7684e-06\n",
      "Epoch 128/150\n",
      "1/1 [==============================] - 0s - loss: 4.5101e-06\n",
      "Epoch 129/150\n",
      "1/1 [==============================] - 0s - loss: 4.2518e-06\n",
      "Epoch 130/150\n",
      "1/1 [==============================] - 0s - loss: 4.0531e-06\n",
      "Epoch 131/150\n",
      "1/1 [==============================] - 0s - loss: 3.8743e-06\n",
      "Epoch 132/150\n",
      "1/1 [==============================] - 0s - loss: 3.6558e-06\n",
      "Epoch 133/150\n",
      "1/1 [==============================] - 0s - loss: 3.4769e-06\n",
      "Epoch 134/150\n",
      "1/1 [==============================] - 0s - loss: 3.2981e-06\n",
      "Epoch 135/150\n",
      "1/1 [==============================] - 0s - loss: 3.1392e-06\n",
      "Epoch 136/150\n",
      "1/1 [==============================] - 0s - loss: 2.9604e-06\n",
      "Epoch 137/150\n",
      "1/1 [==============================] - 0s - loss: 2.8412e-06\n",
      "Epoch 138/150\n",
      "1/1 [==============================] - 0s - loss: 2.7021e-06\n",
      "Epoch 139/150\n",
      "1/1 [==============================] - 0s - loss: 2.5233e-06\n",
      "Epoch 140/150\n",
      "1/1 [==============================] - 0s - loss: 2.3842e-06\n",
      "Epoch 141/150\n",
      "1/1 [==============================] - 0s - loss: 2.2848e-06\n",
      "Epoch 142/150\n",
      "1/1 [==============================] - 0s - loss: 2.1458e-06\n",
      "Epoch 143/150\n",
      "1/1 [==============================] - 0s - loss: 2.0464e-06\n",
      "Epoch 144/150\n",
      "1/1 [==============================] - 0s - loss: 1.9471e-06\n",
      "Epoch 145/150\n",
      "1/1 [==============================] - 0s - loss: 1.8279e-06\n",
      "Epoch 146/150\n",
      "1/1 [==============================] - 0s - loss: 1.7285e-06\n",
      "Epoch 147/150\n",
      "1/1 [==============================] - 0s - loss: 1.6689e-06\n",
      "Epoch 148/150\n",
      "1/1 [==============================] - 0s - loss: 1.5696e-06\n",
      "Epoch 149/150\n",
      "1/1 [==============================] - 0s - loss: 1.4901e-06\n",
      "Epoch 150/150\n",
      "1/1 [==============================] - 0s - loss: 1.4106e-06\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x252ec3a50>"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(ins, gts, batch_size=1, nb_epoch=150, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INPUT\n",
      "(7,)\n",
      "#START#\n",
      "a\n",
      "little\n",
      "baby\n",
      "plays\n",
      "croquet\n"
     ]
    }
   ],
   "source": [
    "print(\"INPUT\")\n",
    "print ins[0][1].shape\n",
    "\n",
    "for x in xrange(ins[0].shape[1]-1):\n",
    "    #print ins[0, x]\n",
    "    word = ixtoword[np.argmax(ins[0,x])]\n",
    "    print word\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s\n",
      "a\n",
      "1/1 [==============================] - 0s\n",
      "little\n",
      "1/1 [==============================] - 0s\n",
      "baby\n",
      "1/1 [==============================] - 0s\n",
      "plays\n",
      "1/1 [==============================] - 0s\n",
      "croquet\n",
      "1/1 [==============================] - 0s\n",
      "#END#\n"
     ]
    }
   ],
   "source": [
    "ins = np.zeros( (1, len(captions[0])+1, len(wordList)) )\n",
    "ins[:, 0, wordtoix['#START#']] = 1\n",
    "for i in xrange((ins[0].shape[1])-1):\n",
    "    predictions = model.predict(ins, batch_size=1, verbose=1)[0]\n",
    "    index = np.argmax(predictions[i])\n",
    "    print(ixtoword[index])\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{u'sentids': [0, 1, 2, 3, 4], 'cnn features': array([  8.41589998e-09,   3.18888098e-08,   5.94773581e-08,\n",
      "         1.34680349e-07,   5.59009152e-07,   5.02435995e-08,\n",
      "         2.87913409e-08,   3.16422643e-07,   4.71011759e-07,\n",
      "         1.63704699e-05,   5.04545845e-08,   3.01402210e-08,\n",
      "         1.73545772e-07,   2.87365314e-08,   3.70973687e-08,\n",
      "         3.02698652e-07,   3.40937760e-07,   7.39699644e-08,\n",
      "         8.50642834e-07,   1.02383908e-07,   8.51808082e-08,\n",
      "         4.70065800e-07,   3.04309833e-06,   1.10937754e-06,\n",
      "         1.28562112e-06,   6.56475976e-08,   3.00554461e-07,\n",
      "         5.61098602e-07,   2.35904185e-08,   1.75602523e-07,\n",
      "         9.52802814e-09,   3.71192392e-08,   8.47269490e-08,\n",
      "         1.25647404e-07,   1.49475881e-07,   5.19190202e-09,\n",
      "         4.43526780e-08,   6.13872331e-09,   3.86867532e-06,\n",
      "         5.82997984e-07,   2.10374651e-07,   3.08866106e-07,\n",
      "         3.79565961e-07,   2.09374690e-07,   4.47989748e-07,\n",
      "         2.64900677e-07,   1.89438595e-07,   5.67727483e-08,\n",
      "         1.38458446e-07,   6.04028219e-06,   6.86134172e-07,\n",
      "         4.24259596e-08,   7.19297333e-08,   2.04784385e-08,\n",
      "         2.83503397e-07,   7.41681339e-09,   9.44634166e-08,\n",
      "         6.53844001e-09,   1.66423845e-07,   8.98615937e-08,\n",
      "         7.06459730e-07,   8.19186852e-08,   1.87076651e-07,\n",
      "         4.52499989e-07,   1.13534398e-07,   8.73306476e-08,\n",
      "         1.84952785e-06,   3.67330529e-07,   1.75248783e-06,\n",
      "         1.09223759e-08,   6.54612933e-08,   1.94263819e-07,\n",
      "         6.23355145e-09,   5.59112550e-07,   3.81516436e-08,\n",
      "         2.47892178e-08,   1.13358431e-06,   3.12189684e-07,\n",
      "         1.35374293e-07,   2.36198090e-07,   2.13333783e-06,\n",
      "         7.98608255e-07,   3.01858176e-06,   3.05543404e-06,\n",
      "         6.20071091e-07,   2.66241591e-06,   2.63429297e-06,\n",
      "         1.28779121e-07,   6.69094504e-08,   1.71157630e-08,\n",
      "         1.93548271e-08,   3.09328044e-07,   7.02132468e-08,\n",
      "         3.08299036e-07,   2.61834202e-07,   5.73617953e-09,\n",
      "         1.04262547e-07,   1.06352251e-08,   4.05570404e-07,\n",
      "         1.13366866e-06,   3.59215392e-07,   3.78844874e-07,\n",
      "         6.97164580e-07,   1.10646988e-06,   2.07822413e-05,\n",
      "         4.73115506e-07,   6.07526226e-06,   9.58590363e-09,\n",
      "         4.80364122e-08,   5.13476328e-09,   8.73929622e-08,\n",
      "         3.32907547e-07,   6.58178294e-07,   9.63332081e-09,\n",
      "         4.47975808e-08,   3.60055523e-08,   1.22103188e-07,\n",
      "         1.65276037e-09,   5.93618736e-08,   2.17162395e-07,\n",
      "         7.09825230e-08,   1.13007945e-06,   1.47625627e-08,\n",
      "         2.49534970e-08,   1.83561948e-08,   3.11787574e-08,\n",
      "         5.22272671e-07,   4.52871689e-07,   1.19557274e-06,\n",
      "         7.70366171e-08,   6.05334876e-08,   1.33436899e-07,\n",
      "         1.47625144e-07,   1.78572790e-07,   1.30385695e-06,\n",
      "         2.17481372e-07,   7.51617648e-08,   2.88747259e-07,\n",
      "         2.33415017e-06,   8.74054308e-07,   8.42001953e-07,\n",
      "         2.09084874e-06,   5.89173624e-07,   1.07043479e-06,\n",
      "         2.08497767e-07,   9.15949499e-07,   1.79400999e-07,\n",
      "         2.57604188e-06,   2.64504229e-06,   5.25246113e-09,\n",
      "         9.82751885e-07,   6.29095257e-06,   1.43006819e-05,\n",
      "         2.97159590e-06,   6.74249975e-07,   8.90938645e-06,\n",
      "         2.64945902e-05,   4.47900675e-06,   2.73474052e-05,\n",
      "         2.29411216e-05,   1.59261934e-03,   2.43434170e-06,\n",
      "         7.81403401e-07,   6.64690242e-06,   6.59774596e-05,\n",
      "         2.32120192e-06,   3.69765221e-05,   7.81656054e-05,\n",
      "         1.19033939e-04,   5.17956913e-03,   1.83769628e-01,\n",
      "         6.79825052e-06,   8.16244938e-05,   1.79826852e-03,\n",
      "         1.23597065e-05,   3.53580923e-03,   1.05012232e-03,\n",
      "         6.43258035e-01,   1.00594934e-05,   3.56646274e-06,\n",
      "         8.65407401e-07,   9.55633586e-04,   3.70333763e-03,\n",
      "         9.21081461e-04,   4.93787788e-02,   1.33183226e-03,\n",
      "         1.02438265e-03,   2.26834727e-05,   7.49448640e-03,\n",
      "         7.77365221e-03,   8.03834773e-05,   1.86260277e-03,\n",
      "         7.32741086e-04,   1.90441860e-04,   4.30757122e-04,\n",
      "         2.36850447e-05,   4.45019279e-04,   2.14122306e-03,\n",
      "         1.04301202e-03,   5.24507377e-05,   6.34799217e-05,\n",
      "         2.55871309e-05,   2.94591766e-03,   3.08123883e-04,\n",
      "         6.75803085e-06,   1.36988150e-04,   3.78652388e-04,\n",
      "         2.52307829e-04,   3.35354489e-05,   1.51535933e-04,\n",
      "         2.98679504e-03,   5.09060395e-04,   4.44510300e-03,\n",
      "         1.70834106e-03,   2.60235771e-04,   4.09048720e-04,\n",
      "         1.62900160e-05,   9.92566493e-05,   1.52575230e-04,\n",
      "         3.57110366e-05,   6.83773542e-05,   8.60223605e-04,\n",
      "         3.24875553e-04,   3.47038767e-05,   7.80885690e-04,\n",
      "         5.08010380e-05,   5.17836132e-04,   8.67020863e-05,\n",
      "         1.30615328e-04,   1.07148480e-04,   5.88962976e-05,\n",
      "         2.06628873e-04,   4.36717819e-04,   2.99684983e-03,\n",
      "         4.88210435e-06,   5.01365539e-05,   3.87837063e-05,\n",
      "         1.44071873e-06,   2.23251118e-05,   7.04956910e-05,\n",
      "         7.14824782e-05,   3.82620310e-05,   1.28973070e-05,\n",
      "         3.27292628e-06,   3.72306531e-05,   4.19208118e-06,\n",
      "         3.31708688e-05,   1.32289148e-04,   5.65981172e-05,\n",
      "         3.44346954e-05,   1.25074934e-04,   3.50736664e-05,\n",
      "         5.93019358e-04,   1.40753991e-05,   8.51637196e-07,\n",
      "         1.90208797e-04,   3.59969883e-04,   1.05997868e-04,\n",
      "         6.86901549e-05,   1.29729506e-05,   2.73978912e-05,\n",
      "         2.37249442e-05,   1.91876767e-04,   1.17671107e-05,\n",
      "         7.76741945e-05,   7.21329052e-05,   4.24298669e-05,\n",
      "         3.12306016e-04,   2.04070384e-05,   1.20834576e-03,\n",
      "         1.50897587e-03,   1.51580884e-04,   1.04081910e-02,\n",
      "         2.20129988e-03,   4.88445396e-04,   6.57426863e-05,\n",
      "         1.45936934e-02,   2.39809044e-03,   1.07794802e-03,\n",
      "         1.83754112e-03,   2.98952241e-03,   2.47271128e-05,\n",
      "         1.75820369e-05,   5.01614886e-06,   1.97140216e-05,\n",
      "         5.59595355e-05,   2.35083367e-04,   2.40657874e-03,\n",
      "         1.87809928e-05,   1.21840649e-05,   9.48403340e-07,\n",
      "         4.98044901e-05,   7.42897964e-06,   7.44979945e-04,\n",
      "         5.27423690e-04,   3.09508068e-05,   2.44482188e-04,\n",
      "         2.02134129e-06,   2.50334211e-04,   5.58734173e-05,\n",
      "         1.70963173e-08,   6.25895495e-08,   1.73377117e-08,\n",
      "         2.08288782e-08,   7.14956938e-09,   1.37520626e-08,\n",
      "         5.59776012e-08,   7.71441719e-08,   5.15827558e-08,\n",
      "         1.47241437e-07,   1.01770763e-06,   1.56359377e-07,\n",
      "         1.20653709e-07,   1.03022302e-07,   1.14457279e-07,\n",
      "         1.47181922e-07,   1.71056787e-08,   9.42088629e-09,\n",
      "         1.07596372e-06,   8.99065995e-08,   8.70356054e-09,\n",
      "         1.16631739e-07,   1.16235270e-08,   2.19923812e-08,\n",
      "         2.66519748e-08,   2.68241980e-08,   8.91461749e-09,\n",
      "         1.06611617e-06,   6.85434713e-08,   6.28125576e-07,\n",
      "         1.37751067e-05,   1.06462678e-04,   1.39697789e-07,\n",
      "         1.26196869e-06,   1.25474498e-05,   4.48229512e-05,\n",
      "         7.05740604e-05,   7.87748741e-06,   3.22897478e-07,\n",
      "         5.80860842e-07,   2.35775565e-06,   6.56412203e-06,\n",
      "         2.64235605e-05,   5.72219005e-06,   1.55528198e-07,\n",
      "         2.78232278e-06,   3.68568493e-07,   1.00191273e-05,\n",
      "         5.24310344e-05,   4.10808898e-05,   2.00746854e-05,\n",
      "         9.44281624e-07,   1.38186635e-06,   9.65234358e-06,\n",
      "         2.12567465e-05,   3.70949530e-03,   4.58857697e-03,\n",
      "         7.52207940e-04,   3.92485854e-05,   1.47607483e-04,\n",
      "         6.99159427e-05,   6.39788996e-05,   1.62402168e-04,\n",
      "         4.16155103e-07,   3.36715800e-07,   1.00041561e-07,\n",
      "         6.60540422e-08,   1.52023787e-07,   2.06997493e-06,\n",
      "         6.96480129e-07,   1.17590730e-06,   1.46852515e-04,\n",
      "         3.28362512e-04,   5.56626437e-05,   4.47538196e-06,\n",
      "         9.00530381e-07,   1.74368452e-07,   1.67885730e-06,\n",
      "         6.09108884e-06,   3.43178186e-07,   1.87193909e-05,\n",
      "         1.97115673e-06,   4.91982007e-07,   5.19084097e-06,\n",
      "         5.17354329e-06,   1.42266671e-07,   6.49489380e-07,\n",
      "         5.58422289e-06,   2.01768501e-07,   3.02741228e-08,\n",
      "         1.67427075e-07,   3.52612062e-07,   4.87624554e-08,\n",
      "         3.04703729e-09,   1.81568510e-07,   3.20703521e-07,\n",
      "         1.40385517e-08,   2.63617768e-08,   1.78524857e-08,\n",
      "         5.63581580e-07,   5.33533075e-08,   6.00583803e-07,\n",
      "         1.14444617e-07,   3.10663438e-07,   1.38267112e-07,\n",
      "         3.89648818e-07,   3.29079285e-07,   1.91012805e-06,\n",
      "         4.32585097e-07,   2.76170545e-06,   1.98448083e-06,\n",
      "         1.31574907e-07,   1.82388903e-05,   3.29677823e-06,\n",
      "         1.60830034e-08,   5.44719612e-08,   2.37905823e-07,\n",
      "         1.21814040e-07,   1.46215740e-07,   6.01743579e-07,\n",
      "         4.37912661e-07,   1.18722326e-07,   1.98476666e-08,\n",
      "         6.37969961e-08,   1.05081298e-07,   2.92747086e-06,\n",
      "         1.52249766e-06,   6.03459569e-07,   4.68283906e-06,\n",
      "         2.04675942e-07,   2.69168481e-07,   4.49019694e-07,\n",
      "         1.77977043e-07,   7.79816531e-08,   7.68060943e-07,\n",
      "         1.92184780e-06,   2.21067424e-07,   4.99142106e-06,\n",
      "         6.02107548e-07,   9.43237097e-08,   4.18178075e-07,\n",
      "         2.50357942e-07,   1.53065923e-06,   1.93722130e-07,\n",
      "         6.62393518e-07,   7.82976883e-08,   8.34146192e-07,\n",
      "         3.92444662e-07,   1.63121513e-07,   3.93939700e-07,\n",
      "         1.87814408e-07,   6.42507771e-08,   2.47294327e-08,\n",
      "         9.67220444e-07,   4.08946086e-08,   3.84579089e-08,\n",
      "         2.05411780e-07,   1.15959523e-07,   5.98772601e-07,\n",
      "         8.06470712e-09,   3.76605885e-06,   5.06494473e-08,\n",
      "         1.40780639e-06,   3.47298146e-06,   2.66475922e-07,\n",
      "         3.87264549e-07,   1.16228618e-08,   1.18715874e-07,\n",
      "         5.31392516e-07,   1.83573157e-08,   3.48557876e-07,\n",
      "         2.94123321e-08,   8.66354583e-07,   2.75358559e-07,\n",
      "         3.51285756e-08,   1.35633783e-07,   3.90570527e-08,\n",
      "         1.74554145e-07,   1.39715303e-06,   3.18180156e-07,\n",
      "         1.37928976e-06,   1.68124217e-07,   9.05526534e-08,\n",
      "         7.12529345e-06,   5.16849582e-07,   2.29573374e-07,\n",
      "         3.19519330e-07,   3.00180787e-08,   1.89145517e-08,\n",
      "         5.86798876e-07,   9.69057368e-08,   1.07433129e-06,\n",
      "         1.32719401e-06,   5.07258801e-06,   4.18551345e-08,\n",
      "         5.95965105e-07,   2.22292115e-07,   6.04314266e-07,\n",
      "         2.63461573e-07,   1.37635766e-07,   2.03719696e-06,\n",
      "         1.02251738e-06,   6.71026896e-08,   7.65382140e-08,\n",
      "         3.16504725e-07,   3.57531519e-07,   1.31346800e-08,\n",
      "         5.07939824e-08,   2.32599390e-07,   1.16453185e-08,\n",
      "         2.50104654e-06,   3.35661952e-07,   2.67076871e-07,\n",
      "         9.82752368e-08,   6.84791814e-08,   7.02228888e-08,\n",
      "         3.11915016e-07,   1.54926425e-07,   6.00050498e-08,\n",
      "         1.50830590e-06,   8.01716737e-08,   1.20128846e-06,\n",
      "         8.33946885e-08,   7.40667076e-07,   1.17309913e-08,\n",
      "         5.75625336e-07,   9.48309321e-07,   4.28480888e-07,\n",
      "         2.41305571e-08,   8.51153192e-09,   2.94692745e-06,\n",
      "         8.49399271e-07,   8.00739031e-07,   5.09481424e-07,\n",
      "         8.74436155e-05,   1.36895153e-08,   2.84124042e-07,\n",
      "         8.47319243e-05,   1.47372518e-07,   5.95545316e-05,\n",
      "         1.26809118e-06,   3.83275633e-08,   1.27259794e-08,\n",
      "         1.13929858e-07,   1.02148960e-07,   6.77910634e-08,\n",
      "         3.04563997e-07,   2.87398780e-08,   1.39215740e-06,\n",
      "         3.22066353e-06,   2.95978907e-07,   4.23622652e-07,\n",
      "         1.53078474e-06,   1.12235500e-06,   1.43436097e-07,\n",
      "         2.05120685e-07,   1.42597503e-06,   4.94455662e-07,\n",
      "         7.67043815e-08,   4.37483948e-07,   6.96361511e-08,\n",
      "         6.67902356e-08,   1.44829329e-07,   4.17369108e-08,\n",
      "         7.81120590e-08,   1.84716669e-07,   2.86475483e-08,\n",
      "         6.55442562e-08,   2.35073253e-06,   3.59297644e-07,\n",
      "         3.00867491e-08,   7.16749753e-07,   2.20964012e-08,\n",
      "         9.09514668e-08,   5.90048259e-08,   1.11550804e-07,\n",
      "         1.95314314e-07,   2.30581065e-08,   2.03475125e-07,\n",
      "         4.29331692e-07,   2.80990271e-08,   1.38060701e-08,\n",
      "         4.40241479e-08,   6.75935496e-07,   6.17455385e-08,\n",
      "         1.01904732e-07,   6.96565792e-07,   2.11481915e-06,\n",
      "         8.91030822e-07,   3.90137018e-07,   1.22803755e-07,\n",
      "         1.03041472e-07,   3.42961108e-08,   3.65284976e-08,\n",
      "         1.40919781e-06,   1.56832940e-07,   6.52353265e-06,\n",
      "         5.10303408e-08,   2.66249089e-07,   6.61256763e-07,\n",
      "         5.85479086e-07,   6.43476596e-08,   2.56659831e-08,\n",
      "         9.21159938e-07,   1.99199501e-08,   5.41099432e-08,\n",
      "         5.91029618e-07,   9.91523379e-08,   3.10224486e-08,\n",
      "         2.34342451e-06,   9.49433172e-07,   4.32440538e-06,\n",
      "         3.98151556e-08,   6.47125944e-08,   4.34974723e-08,\n",
      "         2.19850804e-08,   3.36366135e-08,   1.65049698e-07,\n",
      "         7.30164942e-08,   1.20810412e-07,   1.15907721e-06,\n",
      "         2.05269203e-07,   1.54242827e-07,   5.53728171e-07,\n",
      "         1.08126997e-07,   2.44486478e-06,   3.37486085e-07,\n",
      "         1.21018388e-06,   2.93555757e-07,   7.76053710e-08,\n",
      "         6.08636768e-08,   3.40229690e-07,   2.38327516e-07,\n",
      "         7.15864701e-08,   6.10313080e-08,   4.40375807e-08,\n",
      "         9.99974006e-08,   1.61454284e-06,   1.16619503e-07,\n",
      "         2.68246225e-07,   1.81463960e-07,   1.22370963e-07,\n",
      "         7.77040157e-08,   2.07753921e-08,   2.34707667e-08,\n",
      "         5.58467477e-08,   1.59853539e-06,   2.01438297e-06,\n",
      "         9.38479332e-07,   1.31303227e-06,   6.04048864e-08,\n",
      "         1.20339882e-05,   1.83108746e-08,   2.09470659e-06,\n",
      "         1.78682581e-06,   6.89435140e-08,   1.04795629e-06,\n",
      "         6.59977530e-08,   1.24983401e-07,   4.73822368e-08,\n",
      "         4.95000108e-07,   2.05677381e-07,   3.32745685e-07,\n",
      "         4.39716541e-06,   8.02994180e-07,   1.02988758e-06,\n",
      "         3.45996618e-08,   1.84267648e-07,   1.09521082e-07,\n",
      "         1.88759177e-08,   2.87184918e-07,   1.08130030e-06,\n",
      "         2.87627722e-06,   6.61597070e-08,   1.15686917e-05,\n",
      "         1.85432634e-06,   1.61604712e-05,   2.00405111e-06,\n",
      "         1.83371313e-07,   4.39507941e-09,   1.60268587e-06,\n",
      "         8.80399796e-07,   3.55996377e-07,   1.28542425e-07,\n",
      "         4.24464410e-08,   1.99166070e-08,   6.53419150e-08,\n",
      "         1.76453966e-07,   5.27852364e-07,   6.72440095e-08,\n",
      "         2.60233520e-07,   4.07028651e-08,   3.86546475e-07,\n",
      "         6.77448838e-07,   4.11389323e-08,   1.07242066e-07,\n",
      "         4.15634702e-07,   1.61208334e-07,   9.20756804e-07,\n",
      "         4.57381788e-07,   3.74197543e-06,   2.81124335e-06,\n",
      "         4.53265088e-08,   4.11449355e-06,   1.39841219e-07,\n",
      "         6.94117901e-08,   3.30324099e-07,   8.49317843e-08,\n",
      "         1.76358753e-07,   6.96536944e-08,   2.20997117e-06,\n",
      "         8.39234175e-08,   6.17740454e-08,   2.62558103e-07,\n",
      "         2.23136709e-08,   2.01411563e-08,   1.18440914e-06,\n",
      "         5.59137379e-06,   6.60495971e-07,   6.97520353e-09,\n",
      "         3.16089199e-08,   2.38097158e-07,   1.20893048e-07,\n",
      "         3.27055467e-08,   1.79098421e-07,   1.31696552e-07,\n",
      "         1.21194523e-06,   9.00304968e-08,   1.84292355e-06,\n",
      "         1.86389244e-07,   2.95953214e-07,   1.06935317e-06,\n",
      "         5.26297370e-08,   8.57924590e-07,   7.14928319e-07,\n",
      "         1.33746866e-07,   6.88299565e-07,   5.21401518e-08,\n",
      "         3.93049561e-07,   1.21823220e-07,   4.08520606e-07,\n",
      "         8.58775195e-07,   2.69024645e-06,   1.71661199e-07,\n",
      "         3.79105956e-08,   9.49129841e-09,   3.79441843e-07,\n",
      "         3.56289291e-08,   9.38748315e-08,   3.63054617e-08,\n",
      "         5.70535519e-07,   6.43432713e-07,   2.32087984e-07,\n",
      "         9.15540795e-07,   3.15308966e-07,   1.01030594e-06,\n",
      "         2.44586403e-07,   1.17636159e-06,   2.32685025e-07,\n",
      "         5.66548408e-08,   6.16455827e-06,   2.99054278e-07,\n",
      "         3.32324845e-07,   4.42912835e-07,   1.43854402e-06,\n",
      "         3.68274073e-07,   9.76128902e-08,   1.34958071e-07,\n",
      "         5.64578954e-07,   1.93557867e-06,   1.19488362e-07,\n",
      "         3.52244030e-07,   7.51449249e-08,   6.44465388e-08,\n",
      "         5.12021323e-08,   5.93825966e-08,   8.55815529e-07,\n",
      "         5.31729597e-07,   4.17675494e-07,   2.88856626e-07,\n",
      "         5.32595948e-07,   1.53877352e-06,   2.46374634e-06,\n",
      "         1.13586381e-07,   3.28943713e-07,   2.83385404e-08,\n",
      "         2.44700391e-06,   7.93366226e-08,   7.35148831e-08,\n",
      "         1.70028315e-07,   5.65709399e-08,   8.08705806e-07,\n",
      "         2.85453621e-06,   1.67622161e-07,   5.83836481e-07,\n",
      "         8.40320899e-06,   8.66827747e-08,   1.17420279e-06,\n",
      "         3.76426868e-07,   1.84572889e-06,   9.81451365e-08,\n",
      "         1.97913479e-07,   2.60744423e-06,   3.08470339e-06,\n",
      "         5.91429398e-07,   2.94501410e-06,   1.53213492e-07,\n",
      "         2.34612003e-07,   6.60263311e-08,   3.10190771e-08,\n",
      "         2.86521811e-07,   6.15869649e-05,   1.10605569e-07,\n",
      "         7.98394240e-07,   1.58476269e-07,   3.41048505e-08,\n",
      "         8.29397795e-09,   6.40261746e-08,   1.23391189e-07,\n",
      "         3.89633215e-07,   2.34352022e-08,   9.72838095e-08,\n",
      "         2.88676620e-08,   2.24796288e-08,   7.32545743e-07,\n",
      "         5.89201454e-06,   7.53643192e-08,   4.62837488e-06,\n",
      "         1.67894768e-07,   2.52987519e-07,   1.08217421e-06,\n",
      "         2.01697208e-06,   5.36081224e-08,   4.03058323e-07,\n",
      "         2.57662037e-07,   1.42321497e-07,   3.04320139e-07,\n",
      "         5.11227711e-07,   1.02655270e-06,   8.26697089e-07,\n",
      "         3.24995483e-07,   1.21995058e-07,   5.66677016e-08,\n",
      "         2.14317382e-07,   3.85841588e-07,   3.32231991e-07,\n",
      "         1.71364320e-07,   8.59406413e-08,   1.57774608e-07,\n",
      "         2.46794389e-08,   2.92363609e-07,   9.88827378e-06,\n",
      "         3.54006852e-06,   2.74126705e-06,   2.29737637e-07,\n",
      "         2.59840192e-07,   3.46858741e-07,   4.58612192e-07,\n",
      "         1.66152756e-06,   3.86087095e-07,   2.95032862e-07,\n",
      "         6.52049437e-07,   1.06612582e-07,   2.62583519e-08,\n",
      "         9.43001908e-07,   1.07321355e-07,   2.43793664e-07,\n",
      "         4.04923469e-07,   8.40675568e-07,   2.47264978e-07,\n",
      "         5.65494986e-07,   2.57618296e-07,   7.64995093e-07,\n",
      "         6.05014066e-06,   4.62846295e-08,   5.21273989e-07,\n",
      "         9.65458526e-07,   9.73121104e-08,   7.69533486e-08,\n",
      "         2.32474989e-07,   2.95987525e-06,   3.72256295e-06,\n",
      "         4.07482901e-07,   4.22909565e-08,   7.33164313e-07,\n",
      "         3.28352421e-08,   1.78373469e-07,   2.69284811e-07,\n",
      "         6.39241307e-07,   2.54285226e-07,   2.00588744e-07,\n",
      "         1.11265955e-07,   1.83471820e-06,   7.55673852e-07,\n",
      "         1.08630502e-06,   1.76793733e-07,   9.24205381e-07,\n",
      "         2.30930073e-06,   6.47323816e-07,   9.68654035e-07,\n",
      "         1.47438584e-07,   2.40291826e-07,   4.15612504e-07,\n",
      "         3.09656372e-08,   1.86306897e-06,   5.46019635e-07,\n",
      "         1.17348016e-07,   1.86116068e-07,   1.15847335e-07,\n",
      "         3.48428415e-08,   2.81128735e-07,   1.55914009e-07,\n",
      "         1.89463981e-06,   4.36741357e-06,   6.79047162e-07,\n",
      "         1.24940198e-06,   9.56550775e-06,   2.08576623e-08,\n",
      "         1.59942985e-07,   2.52077780e-06,   1.92841412e-07,\n",
      "         1.52423524e-06,   1.51593525e-07,   4.68828567e-08,\n",
      "         2.09163034e-08,   3.42687763e-08,   1.16572290e-08,\n",
      "         9.16321170e-08,   2.38912694e-08,   1.38086051e-07,\n",
      "         2.76502057e-08,   6.11460749e-08,   8.81325093e-08,\n",
      "         1.16119594e-08,   9.07845674e-08,   1.68745284e-08,\n",
      "         3.60196317e-08,   1.35766044e-07,   2.73536767e-08,\n",
      "         3.74870091e-08,   2.90941866e-08,   8.66364491e-09,\n",
      "         4.18528181e-08,   6.29129389e-08,   4.43892167e-09,\n",
      "         1.19177734e-08,   1.05328564e-08,   1.57016338e-08,\n",
      "         6.64163622e-08,   3.58170524e-08,   7.66861703e-08,\n",
      "         9.91837812e-09,   1.28141018e-08,   8.47962589e-09,\n",
      "         3.08904262e-08,   7.87273624e-09,   4.67559369e-09,\n",
      "         9.86844739e-09,   5.38734241e-07,   4.79142521e-08,\n",
      "         1.87804730e-07,   2.02872830e-08,   1.76352358e-07,\n",
      "         1.85942724e-07,   3.46882061e-08,   4.94499268e-08,\n",
      "         8.73525536e-08,   2.53816275e-08,   3.43038465e-07,\n",
      "         1.04763700e-07,   1.25577208e-05,   1.26446650e-07,\n",
      "         2.99924013e-05,   4.70625814e-08,   4.95580935e-06,\n",
      "         4.62725848e-06,   1.43686657e-05,   2.37719869e-05,\n",
      "         1.01009455e-05,   1.79579627e-06,   1.06105983e-06,\n",
      "         2.28260447e-07,   9.73303926e-08,   2.30288340e-08,\n",
      "         4.81443067e-07,   3.55609693e-08,   1.42960372e-08,\n",
      "         9.78636763e-07,   5.98692367e-08,   8.85203306e-08,\n",
      "         3.06405070e-08,   1.42716203e-07,   5.13964515e-08,\n",
      "         1.71135106e-08,   1.52060107e-08,   5.96984222e-08,\n",
      "         7.55639235e-07,   1.19618395e-08,   4.45140643e-07,\n",
      "         2.11517431e-06], dtype=float32), u'filename': u'2513260012_03d33305cf.jpg', u'imgid': 0, 'raw image': array([[[  90.,   76.,   59.],\n",
      "        [  74.,   60.,   51.],\n",
      "        [  70.,   57.,   51.],\n",
      "        ..., \n",
      "        [  72.,   62.,   59.],\n",
      "        [  28.,   25.,   19.],\n",
      "        [  43.,   36.,   31.]],\n",
      "\n",
      "       [[  34.,   27.,   18.],\n",
      "        [  63.,   59.,   51.],\n",
      "        [  53.,   42.,   35.],\n",
      "        ..., \n",
      "        [  82.,   59.,   56.],\n",
      "        [  62.,   46.,   42.],\n",
      "        [  87.,   82.,   65.]],\n",
      "\n",
      "       [[  42.,   34.,   27.],\n",
      "        [  62.,   49.,   35.],\n",
      "        [  73.,   56.,   46.],\n",
      "        ..., \n",
      "        [  43.,   41.,   22.],\n",
      "        [  39.,   29.,   21.],\n",
      "        [  56.,   46.,   34.]],\n",
      "\n",
      "       ..., \n",
      "       [[ 237.,  239.,  237.],\n",
      "        [ 232.,  234.,  233.],\n",
      "        [ 242.,  243.,  244.],\n",
      "        ..., \n",
      "        [ 205.,  213.,  216.],\n",
      "        [ 212.,  221.,  219.],\n",
      "        [ 212.,  217.,  224.]],\n",
      "\n",
      "       [[ 237.,  237.,  235.],\n",
      "        [ 237.,  237.,  236.],\n",
      "        [ 233.,  231.,  232.],\n",
      "        ..., \n",
      "        [ 210.,  216.,  223.],\n",
      "        [ 221.,  224.,  225.],\n",
      "        [ 212.,  217.,  220.]],\n",
      "\n",
      "       [[ 215.,  215.,  213.],\n",
      "        [ 224.,  224.,  222.],\n",
      "        [ 228.,  228.,  227.],\n",
      "        ..., \n",
      "        [ 212.,  220.,  224.],\n",
      "        [ 218.,  221.,  226.],\n",
      "        [ 218.,  219.,  226.]]], dtype=float32), u'split': u'train', u'sentences': [{u'tokens': [u'a', u'black', u'dog', u'is', u'running', u'after', u'a', u'white', u'dog', u'in', u'the', u'snow'], u'raw': u'A black dog is running after a white dog in the snow .', u'imgid': 0, u'sentid': 0}, {u'tokens': [u'black', u'dog', u'chasing', u'brown', u'dog', u'through', u'snow'], u'raw': u'Black dog chasing brown dog through snow', u'imgid': 0, u'sentid': 1}, {u'tokens': [u'two', u'dogs', u'chase', u'each', u'other', u'across', u'the', u'snowy', u'ground'], u'raw': u'Two dogs chase each other across the snowy ground .', u'imgid': 0, u'sentid': 2}, {u'tokens': [u'two', u'dogs', u'play', u'together', u'in', u'the', u'snow'], u'raw': u'Two dogs play together in the snow .', u'imgid': 0, u'sentid': 3}, {u'tokens': [u'two', u'dogs', u'running', u'through', u'a', u'low', u'lying', u'body', u'of', u'water'], u'raw': u'Two dogs running through a low lying body of water .', u'imgid': 0, u'sentid': 4}]}\n"
     ]
    }
   ],
   "source": [
    "import pickle\n",
    "\n",
    "dataset = pickle.load(open('flickr8k_Entire_dataset_with_cnn_features.pkl'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "35\n",
      "[u'a', u'black', u'dog', u'is', u'running', u'after', u'a', u'white', u'dog', u'in', u'the', u'snow']\n",
      "[u'a', u'little', u'baby', u'plays', u'croquet']\n",
      "[u'a', u'brown', u'dog', u'in', u'the', u'snow', u'has', u'something', u'hot', u'pink', u'in', u'its', u'mouth']\n",
      "[u'a', u'brown', u'dog', u'is', u'running', u'along', u'a', u'beach']\n",
      "[u'a', u'black', u'and', u'white', u'dog', u'with', u'a', u'red', u'frisbee', u'standing', u'on', u'a', u'sandy', u'beach']\n",
      "[u'a', u'cyclist', u'wearing', u'a', u'red', u'helmet', u'is', u'riding', u'on', u'the', u'pavement']\n",
      "[u'a', u'man', u'dressed', u'in', u'a', u'purple', u'shirt', u'and', u'red', u'bandanna', u'smiles', u'at', u'the', u'people', u'watching', u'him']\n",
      "[u'a', u'boy', u'wearing', u'a', u'red', u't', u'shirt', u'is', u'running', u'through', u'woodland']\n",
      "[u'a', u'girl', u'in', u'a', u'white', u'dress']\n",
      "[u'a', u'skier', u'in', u'a', u'yellow', u'jacket', u'is', u'airborne', u'above', u'the', u'mountains']\n",
      "[u'a', u'photographer', u'looks', u'over', u'the', u'hills']\n",
      "[u'a', u'bunch', u'of', u'girls', u'in', u'cheerleader', u'outfits']\n",
      "[u'a', u'blue', u'boat', u'with', u'a', u'yellow', u'canopy', u'is', u'floating', u'on', u'calm', u'waters']\n",
      "[u'a', u'dog', u'catches', u'a', u'frisbee', u'in', u'midair']\n",
      "[u'a', u'little', u'old', u'lady', u'sitting', u'next', u'to', u'an', u'advertisement']\n",
      "[u'a', u'blond', u'girl', u'wearing', u'a', u'green', u'jacket', u'walks', u'on', u'a', u'trail', u'along', u'side', u'a', u'metal', u'fence']\n",
      "[u'a', u'family', u'of', u'nine', u'people', u'including', u'four', u'children', u'pose', u'in', u'front', u'of', u'a', u'brick', u'fireplace', u'with', u'a', u'white', u'mantle']\n",
      "[u'man', u'in', u'a', u'sweater', u'pointing', u'at', u'the', u'camera']\n",
      "[u'a', u'dog', u'looks', u'warily', u'at', u'the', u'brown', u'dog', u'investigating', u'his', u'area']\n",
      "[u'three', u'children', u'in', u'a', u'field', u'with', u'white', u'flowers']\n",
      "[u'a', u'boy', u'in', u'a', u'red', u'shirt', u'in', u'front', u'of', u'a', u'long', u'blue', u'wall', u'raises', u'his', u'eyebrow', u'at', u'the', u'camera']\n",
      "[u'a', u'brown', u'dog', u'is', u'carrying', u'a', u'wet', u'stick', u'on', u'the', u'shore', u'of', u'the', u'ocean']\n",
      "[u'two', u'dogs', u'are', u'racing', u'on', u'a', u'ractrack']\n",
      "[u'a', u'baseball', u'pitcher', u'with', u'shelton', u'on', u'his', u'shirt', u'throws', u'the', u'ball']\n",
      "[u'a', u'brown', u'dog', u'wades', u'through', u'water', u'and', u'walks', u'toward', u'a', u'rock']\n",
      "[u'a', u'group', u'of', u'5', u'girls', u'in', u'school', u'uniforms', u'are', u'standing', u'together', u'and', u'waiting', u'to', u'cross', u'the', u'street', u'while', u'one', u'girl', u'takes', u'a', u'picture']\n",
      "[u'a', u'boy', u'wearing', u'a', u'jean', u'jacket', u'with', u'his', u'hand', u'on', u'his', u'chest', u'smiling', u'at', u'the', u'camera']\n",
      "[u'a', u'group', u'of', u'people', u'in', u'front', u'of', u'a', u'house']\n",
      "[u'a', u'woman', u'stands', u'in', u'front', u'of', u'a', u'display', u'case', u'and', u'another', u'woman', u'looks', u'at', u'a', u'digital', u'camera']\n",
      "[u'a', u'snowboarder', u'jumps', u'high', u'into', u'the', u'air', u'off', u'of', u'a', u'snowy', u'ramp']\n",
      "[u'a', u'group', u'of', u'mountain', u'climbers']\n",
      "[u'a', u'black', u'and', u'white', u'dog', u'is', u'jumping', u'through', u'a', u'black', u'and', u'white', u'hoop']\n",
      "[u'a', u'girl', u'using', u'a', u'laptop', u'another', u'girl', u'and', u'a', u'boy', u'looking', u'at', u'his', u'cellphone']\n",
      "[u'a', u'dog', u'races', u'while', u'wearing', u'number', u'6']\n",
      "[u'a', u'group', u'of', u'people', u'stand', u'at', u'the', u'edge', u'of', u'a', u'cliff', u'over', u'the', u'ocean']\n",
      "[u'a', u'couple', u'holding', u'hands', u'and', u'another', u'woman', u'are', u'walking', u'down', u'a', u'wooded', u'trail']\n",
      "[u'a', u'car', u'sponsored', u'by', u'riwal', u'is', u'smoking', u'its', u'tires', u'on', u'a', u'wet', u'road']\n",
      "[u'a', u'bearded', u'man', u'wearing', u'a', u'bandanna', u'is', u'pointing', u'his', u'finger', u'whilst', u'riding', u'a', u'bicycle']\n",
      "[u'a', u'crowd', u'of', u'people', u'waving', u'red', u'flags', u'are', u'passing', u'under', u'a', u'bridge']\n",
      "[u'a', u'man', u'in', u'sunglasses', u'gets', u'ready', u'to', u'pull', u'a', u'chair', u'out', u'to', u'sit', u'down', u'for', u'his', u'meal']\n",
      "[u'a', u'boy', u'holding', u'a', u'racket', u'is', u'standing', u'in', u'the', u'grass']\n",
      "[u'two', u'blond', u'girls', u'are', u'playing', u'together']\n",
      "[u'four', u'bicyclists', u'ride', u'along', u'a', u'dirt', u'road', u'between', u'two', u'wire', u'fences']\n",
      "[u'a', u'dog', u'climbing', u'a', u'steep', u'dirt', u'hill']\n",
      "[u'a', u'child', u'spins', u'on', u'a', u'merry', u'go', u'round']\n",
      "[u'a', u'brown', u'dog', u'is', u'jumping', u'up', u'at', u'a', u'woman', u'who', u'is', u'wearing', u'a', u'brown', u'skirt']\n",
      "[u'a', u'man', u'in', u'the', u'dooorway', u'of', u'a', u'bus', u'looking', u'onward']\n",
      "[u'a', u'little', u'girl', u'is', u'peaking', u'out', u'from', u'behind', u'a', u'tree']\n",
      "[u'a', u'group', u'of', u'people', u'hold', u'drinks', u'and', u'point', u'at', u'the', u'camera']\n",
      "[u'a', u'border', u'collie', u'jumps', u'over', u'a', u'bed', u'with', u'a', u'tennis', u'ball', u'in', u'its', u'mouth']\n",
      "[u'a', u'climber', u'walks', u'along', u'a', u'snowy', u'peak']\n",
      "[u'a', u'white', u'dog', u'chasing', u'a', u'black', u'dog', u'across', u'the', u'grass']\n",
      "[u'a', u'black', u'race', u'car', u'spins', u'its', u'tires', u'while', u'smoke', u'comes', u'out']\n",
      "[u'a', u'person', u'behind', u'some', u'bags', u'inside', u'an', u'igloo']\n",
      "[u'a', u'brown', u'dog', u'drinks', u'from', u'a', u'body', u'of', u'water']\n",
      "[u'a', u'man', u'wearing', u'a', u'black', u'hat', u'is', u'walking', u'through', u'a', u'busy', u'city', u'street']\n",
      "[u'a', u'brown', u'dog', u'chases', u'something', u'a', u'man', u'behind', u'him', u'threw', u'on', u'the', u'beach']\n",
      "[u'a', u'woman', u'taking', u'a', u'photo', u'standing', u'next', u'to', u'a', u'tall', u'man', u'and', u'in', u'front', u'of', u'another', u'man', u'who', u'is', u'also', u'taking', u'a', u'photo']\n",
      "[u'a', u'man', u'is', u'skiing', u'down', u'a', u'mountain', u'with', u'another', u'large', u'mountain', u'behind', u'him', u'and', u'a', u'river', u'between', u'the', u'mountains']\n",
      "[u'a', u'boy', u'climbing', u'a', u'rock', u'face']\n",
      "[u'a', u'brown', u'dog', u'is', u'standing', u'in', u'a', u'lake']\n",
      "[u'a', u'dog', u'plays', u'with', u'a', u'toy', u'in', u'the', u'grass']\n",
      "[u'a', u'elderly', u'man', u'in', u'a', u'straw', u'hat', u'is', u'sitting', u'on', u'a', u'bench']\n",
      "[u'a', u'dark', u'haired', u'man', u'in', u'a', u'red', u't', u'shirt', u'with', u'a', u'beard', u'playing', u'the', u'guitar', u'in', u'a', u'park']\n",
      "[u'two', u'black', u'dogs', u'are', u'jumping', u'and', u'playing', u'with', u'a', u'yellow', u'tennis', u'ball', u'in', u'the', u'snow']\n",
      "[u'an', u'attractive', u'woman', u'holds', u'a', u'halo', u'above', u'her', u'head']\n",
      "[u'an', u'elderly', u'person', u'with', u'a', u'white', u'hat', u'is', u'reading', u'a', u'paperback', u'book']\n",
      "[u'a', u'looker', u'on', u'catches', u'a', u'motocross', u'bike', u'on', u'its', u'side', u'after', u'a', u'spill', u'while', u'another', u'races', u'by']\n",
      "[u'the', u'men', u'are', u'climbing']\n",
      "[u'a', u'skateboarder', u'jumps', u'off', u'his', u'skateboard', u'in', u'midair', u'in', u'hopes', u'of', u'landing', u'back', u'on', u'it']\n",
      "[u'a', u'dark', u'colored', u'man', u'is', u'holding', u'a', u'protest', u'banner', u'in', u'a', u'busy', u'city', u'street', u'with', u'people', u'staring', u'from', u'behind']\n",
      "[u'a', u'man', u'is', u'wearing', u'a', u'blue', u'shirt', u'with', u'the', u'word', u'mambo', u'on', u'it']\n",
      "[u'a', u'man', u'in', u'black', u'shirt', u'sits', u'on', u'a', u'stool', u'while', u'trying', u'to', u'sell', u'stuffed', u'animals']\n",
      "[u'a', u'guy', u'in', u'navy', u'blue', u'and', u'yellow', u'shorts', u'jumping', u'into', u'a', u'pool']\n",
      "[u'a', u'boy', u'jumps', u'from', u'the', u'top', u'of', u'a', u'blue', u'plastic', u'slide']\n",
      "[u'a', u'man', u'is', u'playing', u'with', u'a', u'fire', u'baton', u'in', u'the', u'day', u'light']\n",
      "[u'a', u'brown', u'dog', u'playing', u'with', u'bubbles']\n",
      "[u'a', u'guy', u'with', u'a', u'scarf', u'and', u'woven', u'toboggan']\n",
      "[u'a', u'boy', u'sitting', u'down', u'smiling', u'for', u'the', u'picture']\n",
      "[u'a', u'black', u'dog', u'and', u'a', u'woman', u'in', u'a', u'red', u'shirt', u'playing', u'tug', u'of', u'war']\n",
      "[u'a', u'group', u'of', u'people', u'surf', u'on', u'a', u'very', u'large', u'wave']\n",
      "[u'a', u'man', u'rides', u'a', u'bike', u'on', u'a', u'dirt', u'path']\n",
      "[u'there', u'are', u'two', u'women', u'crossing', u'the', u'street', u'in', u'front', u'of', u'two', u'buses']\n",
      "[u'a', u'group', u'of', u'people', u'walk', u'down', u'a', u'city', u'street', u'at', u'night']\n",
      "[u'a', u'skateboarder', u'is', u'jumping', u'almost', u'as', u'high', u'as', u'a', u'building']\n",
      "[u'a', u'man', u'in', u'a', u'red', u'soccer', u'uniform', u'getting', u'ready', u'to', u'kick', u'the', u'ball']\n",
      "[u'a', u'boy', u'cheerleader', u'has', u'a', u'girl', u'cheerleader', u'on', u'his', u'knee']\n",
      "[u'a', u'driver', u'is', u'at', u'the', u'wheel', u'of', u'a', u'race', u'car']\n",
      "[u'a', u'girl', u'wearing', u'white', u'polka', u'dots', u'and', u'black', u'boots', u'walks', u'and', u'a', u'woman', u'in', u'a', u'blue', u'outfit', u'hurridly', u'walks', u'behind', u'her']\n",
      "[u'a', u'man', u'and', u'woman', u'walk', u'away', u'from', u'the', u'camera', u'as', u'a', u'redheaded', u'man', u'walks', u'in', u'the', u'other', u'direction']\n",
      "[u'a', u'blonde', u'woman', u'in', u'blue', u'shorts', u'walks', u'on', u'the', u'beach', u'carrying', u'her', u'shoes']\n",
      "[u'a', u'cute', u'puppy', u'fetches', u'a', u'yellow', u'ring', u'chew', u'toy', u'in', u'the', u'yard']\n",
      "[u'a', u'boy', u'in', u'blue', u'shirt', u'splashing', u'in', u'water', u'under', u'a', u'dock']\n",
      "[u'a', u'little', u'girl', u'dances', u'in', u'a', u'room', u'with', u'balloons', u'on', u'the', u'floor']\n",
      "[u'a', u'lady', u'and', u'her', u'little', u'girl', u'are', u'holding', u'hands', u'and', u'walking', u'up', u'the', u'sidewalk']\n",
      "[u'a', u'boy', u'in', u'a', u'blue', u'shirt', u'is', u'walking', u'along', u'a', u'tan', u'and', u'blue', u'wall']\n",
      "[u'a', u'tent', u'is', u'being', u'set', u'up', u'on', u'the', u'ice']\n",
      "[u'a', u'few', u'girls', u'with', u'green', u'foam', u'statue', u'of', u'liberty', u'hats', u'on', u'a', u'train']\n",
      "[u'two', u'brown', u'dogs', u'are', u'playing', u'with', u'a', u'red', u'ball']\n",
      "[u'a', u'boy', u'baseball', u'player', u'in', u'a', u'green', u'and', u'yellow', u'jersey', u'leaps', u'for', u'the', u'ball']\n"
     ]
    }
   ],
   "source": [
    "from collections import Counter\n",
    "def word_processing(dataset):\n",
    "    allwords = Counter()\n",
    "    for item in dataset:\n",
    "        for sentence in [item['sentences'][0]]:\n",
    "            allwords.update(sentence['tokens'])\n",
    "            \n",
    "    vocab = [k for k, v in allwords.items()]\n",
    "    vocab.insert(0, '#START#')\n",
    "    vocab.append('#END#')\n",
    "    vocab.append('#NULL#')\n",
    "#     vocab.append('#UNK#')\n",
    "\n",
    "    word_to_index = {w: i for i, w in enumerate(vocab)}\n",
    "    index_to_word = {i: w for i, w in enumerate(vocab)}\n",
    "    return vocab, word_to_index, index_to_word\n",
    "\n",
    "\n",
    "N_Samples = 100\n",
    "data_set = [dataset[1]]#dataset[:N_Samples]\n",
    "data_set = dataset[:N_Samples]\n",
    "MAX_SENTENCE_LENGTH = 35\n",
    "print(MAX_SENTENCE_LENGTH)\n",
    "vocab, word_to_index, index_to_word = word_processing(data_set)\n",
    "#print len(vocab)\n",
    "ins = np.zeros( (N_Samples, MAX_SENTENCE_LENGTH-1, len(vocab)) )\n",
    "gts = np.zeros_like(ins)\n",
    "for ind,dataset_val in enumerate(data_set):\n",
    "    print((dataset_val['sentences'][0]['tokens']))\n",
    "    assert len(dataset_val['sentences'][0]['tokens']) <= MAX_SENTENCE_LENGTH - 2\n",
    "    ins[ind, 0, wordtoix['#START#']] = 1\n",
    "    #print 0\n",
    "    for t, word in enumerate(dataset_val['sentences'][0]['tokens']):\n",
    "      #print(t, word_to_index[word])\n",
    "      #the ground truth at time t is the next word\n",
    "      gts[ind, t, word_to_index[word]] = 1\n",
    "      #the input at time t+1 is the previous ground truth\n",
    "      ins[ind, t+1, word_to_index[word]] = 1\n",
    "      #print t+1\n",
    "    #ground truth ends with the end token (or reuse start token)\n",
    "    #print t+1, len(gts[ind])\n",
    "    #print gts[ind,-1,wordtoix['#END#']]\n",
    "    gts[ind,t+1,wordtoix['#END#']] = 1\n",
    "    #print gts[ind,-1,wordtoix['#END#']]\n",
    "    #print t+2\n",
    "    flag = True\n",
    "    #print len(gts[ind]),t+1\n",
    "    # JUST HAVE NULL CHARACTERS AFTER END\n",
    "    \n",
    "    for time_step in xrange(t+2, MAX_SENTENCE_LENGTH-1):\n",
    "        #print \"Came in \"\n",
    "        word = \"#NULL#\"\n",
    "        if(flag):\n",
    "            word = \"#END#\"\n",
    "            flag= False\n",
    "        ins[ind, time_step, word_to_index[word]] = 1\n",
    "        gts[ind, time_step, word_to_index[\"#NULL#\"]] = 1\n",
    "        #print time_step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INPUT\n",
      "#START# a black dog is running after a white dog in the snow #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a little baby plays croquet #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog in the snow has something hot pink in its mouth #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog is running along a beach #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a black and white dog with a red frisbee standing on a sandy beach #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a cyclist wearing a red helmet is riding on the pavement #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man dressed in a purple shirt and red bandanna smiles at the people watching him #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy wearing a red t shirt is running through woodland #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a girl in a white dress #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a skier in a yellow jacket is airborne above the mountains #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a photographer looks over the hills #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a bunch of girls in cheerleader outfits #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a blue boat with a yellow canopy is floating on calm waters #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dog catches a frisbee in midair #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a little old lady sitting next to an advertisement #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a blond girl wearing a green jacket walks on a trail along side a metal fence #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a family of nine people including four children pose in front of a brick fireplace with a white mantle #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# man in a sweater pointing at the camera #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dog looks warily at the brown dog investigating his area #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# three children in a field with white flowers #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy in a red shirt in front of a long blue wall raises his eyebrow at the camera #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog is carrying a wet stick on the shore of the ocean #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# two dogs are racing on a ractrack #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a baseball pitcher with shelton on his shirt throws the ball #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog wades through water and walks toward a rock #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of 5 girls in school uniforms are standing together and waiting to cross the street while one girl takes a picture #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy wearing a jean jacket with his hand on his chest smiling at the camera #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of people in front of a house #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a woman stands in front of a display case and another woman looks at a digital camera #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a snowboarder jumps high into the air off of a snowy ramp #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of mountain climbers #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a black and white dog is jumping through a black and white hoop #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a girl using a laptop another girl and a boy looking at his cellphone #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dog races while wearing number 6 #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of people stand at the edge of a cliff over the ocean #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a couple holding hands and another woman are walking down a wooded trail #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a car sponsored by riwal is smoking its tires on a wet road #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a bearded man wearing a bandanna is pointing his finger whilst riding a bicycle #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a crowd of people waving red flags are passing under a bridge #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man in sunglasses gets ready to pull a chair out to sit down for his meal #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy holding a racket is standing in the grass #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# two blond girls are playing together #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# four bicyclists ride along a dirt road between two wire fences #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dog climbing a steep dirt hill #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a child spins on a merry go round #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog is jumping up at a woman who is wearing a brown skirt #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man in the dooorway of a bus looking onward #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a little girl is peaking out from behind a tree #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of people hold drinks and point at the camera #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a border collie jumps over a bed with a tennis ball in its mouth #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a climber walks along a snowy peak #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a white dog chasing a black dog across the grass #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a black race car spins its tires while smoke comes out #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a person behind some bags inside an igloo #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog drinks from a body of water #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man wearing a black hat is walking through a busy city street #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog chases something a man behind him threw on the beach #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a woman taking a photo standing next to a tall man and in front of another man who is also taking a photo #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man is skiing down a mountain with another large mountain behind him and a river between the mountains #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy climbing a rock face #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog is standing in a lake #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dog plays with a toy in the grass #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a elderly man in a straw hat is sitting on a bench #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dark haired man in a red t shirt with a beard playing the guitar in a park #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# two black dogs are jumping and playing with a yellow tennis ball in the snow #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# an attractive woman holds a halo above her head #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# an elderly person with a white hat is reading a paperback book #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a looker on catches a motocross bike on its side after a spill while another races by #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# the men are climbing #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a skateboarder jumps off his skateboard in midair in hopes of landing back on it #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a dark colored man is holding a protest banner in a busy city street with people staring from behind #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man is wearing a blue shirt with the word mambo on it #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man in black shirt sits on a stool while trying to sell stuffed animals #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a guy in navy blue and yellow shorts jumping into a pool #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy jumps from the top of a blue plastic slide #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man is playing with a fire baton in the day light #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a brown dog playing with bubbles #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a guy with a scarf and woven toboggan #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy sitting down smiling for the picture #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a black dog and a woman in a red shirt playing tug of war #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of people surf on a very large wave #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man rides a bike on a dirt path #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# there are two women crossing the street in front of two buses #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a group of people walk down a city street at night #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a skateboarder is jumping almost as high as a building #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man in a red soccer uniform getting ready to kick the ball #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy cheerleader has a girl cheerleader on his knee #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a driver is at the wheel of a race car #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a girl wearing white polka dots and black boots walks and a woman in a blue outfit hurridly walks behind her #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a man and woman walk away from the camera as a redheaded man walks in the other direction #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a blonde woman in blue shorts walks on the beach carrying her shoes #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a cute puppy fetches a yellow ring chew toy in the yard #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy in blue shirt splashing in water under a dock #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a little girl dances in a room with balloons on the floor #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a lady and her little girl are holding hands and walking up the sidewalk #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy in a blue shirt is walking along a tan and blue wall #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a tent is being set up on the ice #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a few girls with green foam statue of liberty hats on a train #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# two brown dogs are playing with a red ball #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n",
      "#START# a boy baseball player in a green and yellow jersey leaps for the ball #END# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL# #NULL#\n"
     ]
    }
   ],
   "source": [
    "print(\"INPUT\")\n",
    "for ind, val in enumerate(ins):\n",
    "    s = []\n",
    "    for x in range(MAX_SENTENCE_LENGTH-1):\n",
    "        #print ins[ind,x]\n",
    "        #print np.argmax(ins[ind,x])\n",
    "        word = index_to_word[np.argmax(ins[ind,x])]\n",
    "        #if(word!=\"#NULL#\"):\n",
    "        s.append(word)\n",
    "    print(\" \".join(s))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Building Model\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-309-d90fa284a366>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0mrms\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mRMSprop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlr\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.03\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'categorical_crossentropy'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mrms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 10\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgts\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnb_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m500\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/Users/danielsampetethiyagu/anaconda/lib/python2.7/site-packages/keras/models.pyc\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, nb_epoch, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, **kwargs)\u001b[0m\n\u001b[1;32m    625\u001b[0m                               \u001b[0mshuffle\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mshuffle\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    626\u001b[0m                               \u001b[0mclass_weight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mclass_weight\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 627\u001b[0;31m                               sample_weight=sample_weight)\n\u001b[0m\u001b[1;32m    628\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    629\u001b[0m     def evaluate(self, x, y, batch_size=32, verbose=1,\n",
      "\u001b[0;32m/Users/danielsampetethiyagu/anaconda/lib/python2.7/site-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, nb_epoch, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight)\u001b[0m\n\u001b[1;32m   1122\u001b[0m                               \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mverbose\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcallbacks\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcallbacks\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1123\u001b[0m                               \u001b[0mval_f\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mval_f\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_ins\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mval_ins\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshuffle\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mshuffle\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1124\u001b[0;31m                               callback_metrics=callback_metrics)\n\u001b[0m\u001b[1;32m   1125\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1126\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mevaluate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/danielsampetethiyagu/anaconda/lib/python2.7/site-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36m_fit_loop\u001b[0;34m(self, f, ins, out_labels, batch_size, nb_epoch, verbose, callbacks, val_f, val_ins, shuffle, callback_metrics)\u001b[0m\n\u001b[1;32m    840\u001b[0m                 \u001b[0mbatch_logs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'size'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_ids\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    841\u001b[0m                 \u001b[0mcallbacks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_batch_begin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_index\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_logs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 842\u001b[0;31m                 \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    843\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    844\u001b[0m                     \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/danielsampetethiyagu/anaconda/lib/python2.7/site-packages/keras/backend/theano_backend.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m    790\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    791\u001b[0m         \u001b[0;32massert\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m{\u001b[0m\u001b[0mlist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 792\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfunction\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    793\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    794\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/danielsampetethiyagu/anaconda/lib/python2.7/site-packages/theano/compile/function_module.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    871\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    872\u001b[0m             \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 873\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0moutput_subset\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0mNone\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    874\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput_subset\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0moutput_subset\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    875\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/danielsampetethiyagu/anaconda/lib/python2.7/site-packages/theano/gof/op.pyc\u001b[0m in \u001b[0;36mrval\u001b[0;34m(p, i, o, n)\u001b[0m\n\u001b[1;32m    858\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mparams\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0mgraph\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mNoParams\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    859\u001b[0m             \u001b[0;31m# default arguments are stored in the closure of `rval`\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 860\u001b[0;31m             \u001b[0;32mdef\u001b[0m \u001b[0mrval\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mp\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mp\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnode_input_storage\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mo\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnode_output_storage\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnode\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    861\u001b[0m                 \u001b[0mr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mp\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mo\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    862\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mo\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mnode\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "INP_DIM = len(vocab)\n",
    "print(\"Building Model\")\n",
    "model = Sequential()\n",
    "model.add(TimeDistributedDense(input_dim = INP_DIM, output_dim=64))\n",
    "model.add(LSTM(input_dim=64, output_dim=64, forget_bias_init='one',return_sequences=True))\n",
    "model.add(TimeDistributedDense(input_dim = INP_DIM, output_dim=INP_DIM))\n",
    "model.add(Activation('softmax'))\n",
    "rms = RMSprop(lr=0.03)\n",
    "model.compile(loss='categorical_crossentropy', optimizer=rms)\n",
    "model.fit(ins, gts, batch_size=1, nb_epoch=500, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "testins = np.zeros( (1, MAX_SENTENCE_LENGTH-1, len(vocab)) )\n",
    "testins[:, 0, wordtoix['#START#']] = 1\n",
    "# print (ins[0].shape[0])-1\n",
    "sent = []\n",
    "for i in xrange((testins[0].shape[0])):\n",
    "#     print i\n",
    "    \n",
    "    predictions = model.predict(testins, batch_size=1, verbose=0)[0]\n",
    "    index = np.argmax(predictions[i])\n",
    "    sent.append(index_to_word[index])\n",
    "#     print(index_to_word[index])\n",
    "    if(index_to_word[index] == \"#END#\"):\n",
    "        break;\n",
    "print \" \".join(sent)\n",
    "        \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['#START#', u'motocross', u'pointing', u'splashing', u'yellow', u'four', u'protest', u'skiing', u'go', u'chair', u'children', u'to', u'bike', u'under', u'fences', u'hats', u'brown', u'woman', u'sitting', u'very', u'baton', u'animals', u'bubbles', u'fireplace', u'school', u'large', u'race', u'guy', u'woven', u'round', u'chew', u'paperback', u'bicycle', u'fence', u'dooorway', u'boots', u'tires', u'street', u'drinks', u'blue', u'plays', u'stands', u'stool', u'liberty', u'uniform', u'waiting', u'above', u'racing', u'body', u'men', u'water', u'busy', u'baseball', u'path', u'along', u'cyclist', u'boy', u'colored', u'smoke', u'climber', u'family', u'buses', u'straw', u'cheerleader', u'standing', u'from', u'takes', u'sweater', u'two', u'next', u'few', u'camera', u'calm', u'6', u'wades', u'trail', u'carrying', u'stick', u'baby', u'hold', u'women', u'word', u'room', u'car', u'ride', u'nine', u'shelton', u'high', u'something', u'tan', u'dress', u'pink', u'sit', u'skirt', u'holds', u'hot', u'beach', u'helmet', u'waters', u'after', u'jumping', u'wave', u'man', u'a', u'chasing', u'light', u'shore', u'green', u'tall', u'playing', u'cute', u'shoes', u'over', u'whilst', u'through', u'looks', u'smiling', u'its', u'bunch', u'group', u'redheaded', u'croquet', u'soccer', u'wooded', u'dock', u'hands', u'front', u'day', u'halo', u'slide', u'rock', u'crossing', u'side', u'house', u'navy', u'girl', u'sandy', u'out', u'haired', u'looking', u'hill', u'red', u'dirt', u'sidewalk', u'yard', u'spill', u'skateboard', u'outfits', u'throws', u'scarf', u'tennis', u'statue', u'hoop', u'number', u'one', u'fetches', u'another', u'ring', u'city', u'little', u'toy', u'top', u'girls', u'airborne', u'plastic', u'stuffed', u'white', u'chases', u'dogs', u'beard', u'shorts', u'park', u'flowers', u'tree', u'bed', u'toward', u'steep', u'and', u'bridge', u'uniforms', u'blond', u'border', u'sell', u'mountain', u'snow', u'also', u'chest', u'bicyclists', u'rides', u'surf', u'who', u'sponsored', u'knee', u'the', u'cellphone', u'picture', u'looker', u'metal', u'dog', u'face', u'walking', u'ractrack', u'black', u'riding', u'wearing', u'his', u'sunglasses', u'watching', u'riwal', u'banner', u'him', u'catches', u'eyebrow', u'including', u'river', u'set', u'are', u'racket', u'wire', u'away', u'behind', u'finger', u'between', u'reading', u'across', u'tent', u'outfit', u'guitar', u'merry', u'tug', u'taking', u'ramp', u'passing', u'comes', u'point', u'wall', u'walk', u'dances', u'climbers', u'blonde', u'boat', u'polka', u'three', u'brick', u'staring', u'threw', u'peaking', u'direction', u'fire', u'races', u'toboggan', u'child', u'smoking', u'case', u'balloons', u'air', u'while', u'investigating', u'almost', u'is', u'it', u'player', u'in', u'ready', u'sits', u'bags', u'shirt', u'woodland', u'cross', u'gets', u'wheel', u'bearded', u'ball', u'foam', u'driver', u'snowboarder', u'hand', u'running', u'leaps', u'advertisement', u'climbing', u'canopy', u'off', u'ocean', u'person', u'edge', u'flags', u'jersey', u'jean', u'bandanna', u'being', u'photo', u'laptop', u'hills', u'photographer', u'onward', u'lake', u'bench', u'other', u'attractive', u'wet', u'has', u'hat', u'kick', u'skier', u'couple', u'dark', u'using', u'lady', u'pavement', u't', u'night', u'spins', u'warily', u'old', u'crowd', u'people', u'some', u'back', u'elderly', u'for', u'purple', u'ice', u'skateboarder', u'jumps', u'pose', u'by', u'on', u'getting', u'of', u'meal', u'stand', u'mouth', u'road', u'mantle', u'into', u'down', u'raises', u'mambo', u'her', u'area', u'there', u'long', u'war', u'mountains', u'head', u'snowy', u'bus', u'pitcher', u'landing', u'hopes', u'trying', u'with', u'pull', u'inside', u'up', u'grass', u'display', u'dots', u'an', u'as', u'at', u'walks', u'floating', u'collie', u'hurridly', u'floor', u'dressed', u'field', u'book', u'5', u'holding', u'digital', u'puppy', u'smiles', u'frisbee', u'midair', u'waving', u'peak', u'pool', u'cliff', u'building', u'together', u'jacket', u'train', u'igloo', '#END#', '#NULL#']\n"
     ]
    }
   ],
   "source": [
    "print vocab"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "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": 0
}