nathanhubens/TTA Viz.ipynb Secret

## TTA Viz.ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai import *\n",
    "from fastai.vision import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[PosixPath('/home/jupyter/.fastai/data/cifar10/test'),\n",
       " PosixPath('/home/jupyter/.fastai/data/cifar10/labels.txt'),\n",
       " PosixPath('/home/jupyter/.fastai/data/cifar10/train'),\n",
       " PosixPath('/home/jupyter/.fastai/data/cifar10/models')]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = untar_data(URLs.CIFAR)\n",
    "path.ls()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = (ImageItemList.from_folder(path) \n",
    "        .random_split_by_pct()          \n",
    "        .label_from_folder()           \n",
    "        .add_test_folder()             \n",
    "        .transform(get_transforms(), size=32) \n",
    "        .databunch())    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "img = data.valid_ds[1][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<h4 id=\"_TTA\"><code>_TTA</code><a href=\"https://github.com/fastai/fastai/blob/master/fastai/vision/tta.py#L32\" class=\"source_link\">[source]</a></h4><blockquote><p><code>_TTA</code>(<code>learn</code>:<a href=\"https://docs.fast.ai/basic_train.html#Learner\"><code>Learner</code></a>, <code>beta</code>:<code>float</code>=<code>0.4</code>, <code>scale</code>:<code>float</code>=<code>1.35</code>, <code>ds_type</code>:<a href=\"https://docs.fast.ai/basic_data.html#DatasetType\"><code>DatasetType</code></a>=<code>&lt;DatasetType.Valid: 2&gt;</code>, <code>with_loss</code>:<code>bool</code>=<code>False</code>) → <code>Tensors</code></p>\n",
       "</blockquote>\n",
       "<p><a href=\"https://github.com/fastai/fastai/blob/master/fastai/vision/tta.py#L32\" target=\"_blank\" rel=\"noreferrer noopener\">Show in docs</a></p>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This is how we perform TTA\n",
    "doc(Learner.TTA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<h4 id=\"_tta_only\"><code>_tta_only</code><a href=\"https://github.com/fastai/fastai/blob/master/fastai/vision/tta.py#L10\" class=\"source_link\">[source]</a></h4><blockquote><p><code>_tta_only</code>(<code>learn</code>:<a href=\"https://docs.fast.ai/basic_train.html#Learner\"><code>Learner</code></a>, <code>ds_type</code>:<a href=\"https://docs.fast.ai/basic_data.html#DatasetType\"><code>DatasetType</code></a>=<code>&lt;DatasetType.Valid: 2&gt;</code>, <code>scale</code>:<code>float</code>=<code>1.35</code>) → <code>Iterator</code>[<code>List</code>[<code>Tensor</code>]]</p>\n",
       "</blockquote>\n",
       "<p>Computes the outputs for several augmented inputs for TTA</p>\n",
       "<p><a href=\"https://github.com/fastai/fastai/blob/master/fastai/vision/tta.py#L10\" target=\"_blank\" rel=\"noreferrer noopener\">Show in docs</a></p>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# TTA makes predictions using the function\n",
    "doc(Learner.tta_only)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _tta_only(learn:Learner, ds_type:DatasetType=DatasetType.Valid, scale:float=1.35) -> Iterator[List[Tensor]]:\n",
    "    \"Computes the outputs for several augmented inputs for TTA\"\n",
    "    dl = learn.dl(ds_type)\n",
    "    ds = dl.dataset\n",
    "    old = ds.tfms\n",
    "    augm_tfm = [o for o in learn.data.train_ds.tfms if o.tfm not in\n",
    "               (crop_pad, flip_lr, dihedral, zoom)]\n",
    "    try:\n",
    "        pbar = master_bar(range(8))\n",
    "        for i in pbar:\n",
    "            row = 1 if i&1 else 0\n",
    "            col = 1 if i&2 else 0\n",
    "            flip = i&4\n",
    "            d = {'row_pct':row, 'col_pct':col, 'is_random':False}\n",
    "            tfm = [*augm_tfm, zoom(scale=scale, **d), crop_pad(**d)]\n",
    "            if flip: tfm.append(flip_lr(p=1.))\n",
    "            ds.tfms = tfm\n",
    "            yield get_preds(learn.model, dl, pbar=pbar, activ=_loss_func2activ(learn.loss_func))[0]\n",
    "    finally: ds.tfms = old\n",
    "\n",
    "Learner.tta_only = _tta_only"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TIL what the `&` operator is: a \"bitwise and\" operator\n",
    "\n",
    "It works like so : \n",
    "\n",
    "    3 & 5\n",
    "    \n",
    "    Transformed in bit => 0011 & 0101\n",
    "    \n",
    "    We do the 'AND' operator (if both bits are equal to 1 -> 1, else 0) bit by bit => 0001\n",
    "\n",
    "So in the `_tta_only` case, the `row` value will be alternatively 0 and 1 (e.g.: 0101010),the `col`value will be alternatively 00 and 11 (e.g. 00110011) and the `flip` value will be alternatively 0000 and 1111 (e.g. 000011110000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = create_cnn(data, models.resnet34)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "augm_tfm = [o for o in learn.data.train_ds.tfms if o.tfm not in\n",
    "            (crop_pad, flip_lr, dihedral, zoom)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "row = 0 \n",
    "col = 1\n",
    "flip = 0\n",
    "scale = 1.35"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = {'row_pct':row, 'col_pct':col, 'is_random':False}\n",
    "tfm = [*augm_tfm, zoom(scale=scale, **d), crop_pad(**d)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plots_tfms(im, rows, cols, width, height, **kwargs):\n",
    "    [im.apply_tfms(tfm, **kwargs).show(ax=ax) for i,ax in enumerate(plt.subplots(\n",
    "        rows,cols,figsize=(width,height))[1].flatten())]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAo4AAAFpCAYAAAASp6lHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzsvVmvZEeWpWdn8Nn9zkPciCAjgkNwyGQysztLUqmr0ahu9UPpURCgHyBA0I/QHxGgR0HQD+inVgsoVXdJQiWqa8pMMkkmyZjjzoPPfiY9sPWw91qZYclMerCA9b3ZhvnxM5jZset3rb2TpmmCEEIIIYQQryJ93ScghBBCCCH+caCNoxBCCCGEiEIbRyGEEEIIEYU2jkIIIYQQIgptHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkShjaMQQgghhIgiX+eXffLv/w3UN2xq204qLIHY1BXE6uBiSQ19sqQNsf5g07Q7vSH0SXL7ubrCYzf+xEMIvnxjXcddC8bwc2VVQOz09MS0i2IJfUYjvL5Op2XaVYnHDi5WFivSBT9X1XhfYkjSxLTZvfsv/pv/PoGg+M74d//7/wIPYTqdmfZigWPu7p27ENvZ3jHtLMWlJ0nw71iIJRn2SVPXZn1w6MDnAvZJM3ZObqyWC+iznE8gNptfm3bF1gKyrvjzSgK5PndfyK0MbF2BS46sQOu7sfP+yb/+bzVf18yzf/iP8AT9HGLjnA4Y/znWJ8VY6uaany/fxPxXkXlGBmNd2vfQYjGGPqvVDGJsruH3sXOImVeM372UM6v+zPcZr/6umh/MNNlzefif/1dRc1a/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaJYq8axIlpBkCZGahwbr3Fk/+dPSjwHp8srM9TuJU5fF6MXYGfQEL0fi9WgY8Bjz2ZTiD179ti0JxPUVd29ewdiR0e3TLuq8P6m7Jr9WTL9hXtWVDDBNC+N08VEfL/4brm4uIDY9fWNac/nc+izvb0FsY3RyLTTNur0qMgn2DHGdDl+lDFtFPsLmfWL+RwIs8l51w1bs3z71WsIDdJJ9S3nC5x7rCyx+a1N8bpgDyLm4WAfnB+sD9O0/+6/R7G5SGe610ayXg15v0RFXh3j7zM2j393eS+7B+w4vl9Dr5c984hriUS/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaJYq8axJlo6r3FMSLolnsvIxrheCA/mcw+mERpHn5/xm2O/mphcjyGgxrEs8Zyev3gKsV9++olpn5+dQZ/pFHWPu3u7pk1S3IWytPrQYoXnVJWoIW2aCI0j1W3Yv2Hqb5kPUvzhePrsGcTOz89NuyBjYG9/H2Kj0YZpb+Ut6JMQrVDS2HGRMo2j0/gwqWRdk3xzbhazXHJ0BLv5mkRofemJkRON0wRHqJOixJKRfdgtcOfO1l+xfqoK52OWudc8S0ZIxrCHS2vJsdzaTSR4MKbp6CGfg/cCyffrv/+bL/CxWIWfO0+qsX7lx75TuDaSdnx1n0j0i6MQQgghhIhCG0chhBBCCBGFNo5CCCGEECIKbRyFEEIIIUQU6zXHlEQw7jWrTNdKkumGxoqAaWLtihhfvCKUfC5ttf2n8PupkN5CU7Eyo407qTzH5Mg7O9sQ29vbMe0XL9DM8Mmnn0Ls4MCaFw7296DPam4Lxa8WmOg5z/C+9Dr23mWkDzMvpK6YPHueYr08fYqGrMnUJ6LHh3l6dgqxnW07Vge9PvTJMjTMwDQjwwK7ENMJMd74T7IZzcTwTW3XnrIsoA8zjsFaw8wxxLiAiYmJucwnRmaGFgzFdYqof6B8/d8PVqsFxNrtrmmzeUYLWvgxRSYIKxThk1b74g7fBG2MzlkyzhtnsKXvCWKYQaNqrMnFXXRkvvO4o0cYhP6A+L1HrM+HoV8chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUWjjKIQQQggholirOaYh5pimsoJNX0UlBG5yaZwgvS6W0Kck1U5WiTVhrLo96JO3O6ad5nibUmJgSTJn8EiwT8nMP84BsH+AlTc6vUOI3b5z27S/+PUX0OfrL7+G2M9+9jPTfvutt6CPr6pRrvD+DnreRBTCoatKM+yjCYLJgL2gOqaSgfhuefb8OcQ2NkamfesWjstDMn57vY6LkApKrNqKh7kwXCihphNWxcnOu4ZUl6GmgcquPVWB5piaVVWqXu0EZAL5xKnYmag9Rufuj8O+kFb5YMYF/7mI7xffPcfHLyG2s2PNj4PBEPpwU5Z7qsyASirH+HHGx5Qb0/WrDV8h4Lxi1ejoOhJljmETy1VIYiY7coG+V8Qt+NaTiJtwIz73e0xa/eIohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIYr0JwIlOLviEnhXqhYoVJp9eTsa2PRtDn4okQ02dlqPd8tqrELKWTZCa5pgwNWnhrfMax1YftSR5bwSxVt/G+oMu9FkWqNdst63GMCdazLlL5B1CCL/4+c9N++z8EvrcvnVk2sMenlNZoI5r2Lea0UEfNaQ11ak5LYkyCr922NjZcxrWvX3UMx4eHEBsNLJjPM3wb9YaNIBMVvXqsZNE6K5CQF1VwzIck+IDtdNXF2RdK5nu0R2rIRpHmvTY66pSKpT2B8JjR0wppsWKmYm/TzJh8YeDFYFoufdEp4PvvDQjOv7EJ4wmc4gm2/eZw+mpvrIL+z6vaaQaR7KOxI1i8jmnjWYFAeixvWSfJu3/du84SL7Pe32rY8eiXxyFEEIIIUQU2jgKIYQQQogotHEUQgghhBBRaOMohBBCCCGiWKs5Zjm+gFi1nJp2uUBB/mqJ5piVS+7NBOp1heaN1O2VVxkxvvgEpikRtqZkz+1Co91b0GWrh4YZn5A1Jcc+IYldHz3+2rRPz06gT0OSoSapveaKiIlLFysKPM4qRcMOkTNDhNE4Ma+sMa+fskSDR29gzU67uzvQ5+g2JgXvtq25arXEuXkzR4Obn685EfH7BPppSpLz06zZNsbmXSCJiavSjntWaKAiCcBDiEhwThKVB19IgIjqG5/gmGVdjjIpvDpJOA1pwn4vGAyw4EKva+deSt5nrOCCNzGysZEwg9e3cEqxj3gjawg4rypmjiGFRkLqryWiCEVAAwtNAM7wh2LGIl+44Ns6zFiRgohT+n3QL45CCCGEECIKbRyFEEIIIUQU2jgKIYQQQogotHEUQgghhBBRrNUcc3OGWe1LV91lMUNzzJTElk6QzgwejDS1l8wy5nuYaDXLcc/dyq2IfbR7BH1Gm1sQ29zaNu2T42Po8+kv/gFix88fm3a5xPtUEXNMlttKAkwsvVpYQ1K7QaNESLECQd5Y8XJNqvc0xLzABL7i9cKMVdPpxLZnU+jjjU4hhNB11SqaCvvMyLFyJ5DvtHDsZG5+JkRUn3iDSQghODNMlpLqGaTCTeXE9wWp6lQRY54X2lMjClOwO2F/U7+6UgWrStEQw0yMHp89Ty+/1+z9fjAYDiA2HFnzZauNhtBihev7yhlOMzKv+C9PvrRJRKUwNqbJ58pi6do490pSfS5zVdV8lbdvghjyE4QahNjgdzeGGWxTOBapwkMO7avX8J1PnGHm26JfHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFGsVeN4fY7avfHcahaux6jTu5lMIDaZWQ0ey/nJZBO102lUpJPXWzANYKfThtjhwb5pb75JdBR5D0JZyyZohYS/IYS8hY/K92KatFWBeo86sRqXgiR6Llf2/tbwbSFkHVRXpIV9fqsp6UN0pWluz4np1MR6qWvU6c2dDvH84gz6nJ2dQ2zQtYmJ8xzHUytH7ZUvEDBf4lhN3LjPyPhi+sXgxhgbc0ye5XXDPlFyCDz/ttdFM91lGpNgmCQJ9/pFmkecChq9OJJ0kYDxHw29PiYA39pxunoyNsYkYf18audem+j6MybTbez8aIj/oHaJuxvy/b5PCCEUoHEk7zdygS1XgCBtoz4/zfGdnmZu8NfkvURugt9WNOQ3Oq9VRM0j11Q2Xh9KE4BjLPWFTaBHPHo7CyGEEEKIKLRxFEIIIYQQUWjjKIQQQgghotDGUQghhBBCRLFWc8zZ5TXEvj4dm/bL8yvocz0hCcALK6ZNMxTWB5JounSq1fmCJKiurbg2S3F/vTEcQiwZ7Jp2kaNQObTQHOONIYPRBvQ5OLwFsWdPH5l2TswFK+IaWlb2mm/GaD7a6Npjtft43u0Ej724uTDtYk6GGEu07BK0pixJuFgrdY0i86q04vTpDc7Xr7/6EmLDnp0L21s70GcwwPlyPbPHn16dQp/VzK4heYYi94ysD94wQ8ccMZR44XuS43jOOyi+D8HGEmK6q8lak3qxPzMkRAjfmTnGx2gfciw4vgw03wsuLtGYNp9bo2Ovi2t5t0uKObhxXRVz6LNc4LujmN2Ytn+ffnMwG2uIEaYh766y9IU/iDmGjMXcmfPafXx/d0gsbbt7Rd5ddYNztnaFA4h3N6TOVJOR40Tk549K0P9NxB3t9yi6oV8chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUWjjKIQQQggholirOeazYxTSn4ytuPVqhhnkZysIhdJncGf62waP5U01kzmaY3zxhowIxlcJiokPrG8gNNQcg59rnAh5c3cf+txePYDY46dPTLvd/wL6lKSExWJlb+g1qcxz58AafbZ20MwwIvp/L15uahQv+4odIYTQpK+u4iHWCzPHpE503VQ4x66vLiF2cWFjw8Em9NnZxjG2uLZi/xtSYWIytuaYjIwdXzUhBKzUwCo3MHl640wt3twWwm8Q3w9Gpu0F+yGQKlIBRe3sPFHmTq6FGm9cmwnmyb2TF+b7yaNHaEzb3bFr+dHRHejTI2Nx4Aygs2t8VxZLNK6OnYGNVY5JnIOFGWhYEaXaV20i1a2YWaSq7HupJkafanEDsXbHmmOyNt6npINmo8RVqgktXCPqxm2/2BrFjHAYisNN9t/DG6NfHIUQQgghRBzaOAohhBBCiCi0cRRCCCGEEFGsVeN4VeA+tTXctu2aJNNNiR7B6fRKoqNYET2UV9wlRHvQcsmCWyQxcO41DCGEydyKHBcr/P75AjV/M/e5VYYqhqyLmql7D39g2h/eTKHPyyuMPXr01LT9eYcQwnRp72/SwqTKaRvPs3Sa0SRWM1W5mDSOrx0iGw4tp8cdDgbQZ293F2LDkdX3tdookM1yHGNNYufejMyf8cSO8VaGY4fpHn1i/5zM86Zhd8FnAMfxzBITN06f1WMyRKZgcnOoych5us815HqpgjP1icOj0n3D8ZvfRzAl/mCcHz+H2Be/+oVpVwWOzbt370NstGF1yCnREy7HqGcu3LujWLIiG17jGDd+YjSONdFd+/dJnuMcardx/en17frW6uJ6lxGNo4+lrI/XMxP/A004DlrlVyf2Z0jjKIQQQgghvnO0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxVnPM5sFdiN19455pH5+eQp9nz55B7MXLl6Y9nl5Dn8oJaUPApMYZEX5nzhaQEPFpUaDg9+L8xJ7jCxQq37l7G2L9oRXO5i3yWDI08ewc2mO998MfQ5/za0zuXbrbcnOF967x30cSE1cp3juvcU5JYtecmBe8KD9C2yu+a8iY6w6tyWV3/xD6HN3Beb69s2fabZJwmBncli7R/8UUjVzPLmwC8PkUkxJ3iBi+17Fi+H4Xxem9Nn6uk9tzyoiZjabIbtzEI8nxmwFeX8sZ41JizGuCfVY1yRyckIzKXiDPLiUlRQT89ckc8/1gORtDbDq26/tsyt4J+K70ie2ZMaTOcCzeLO1YOH15Dn3GN/Y82fdz04dvk3HHrsW9q7ptXNtGxOi3sWnf870u3rtOB+9By5lhcnLvfCzvMOMNSTie2+9LiKmPeux8O9KQxNAvjkIIIYQQIgptHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFGs1RyTdzF7+nBry7T3bxPziKs6EUIIhTNdnF1cQB9vhAkhhMRJRJkYPHHmGCYiXS0xO/11acXuxycvsM/NFcQOSmswaMh2PiWVIDo9K6a99+At6NPtonB3NLBi+6dPnkCfnc0N075z/wH0SZYoFJ4UNlav0KjAdPSNl+7+HsJd8YfhyBnXQgjh7v13TPvWXexzeITmmP3DW6ZdNSjoHpPKR8GJw6+XOKefX9oxdnp8DH1yMtG9QH5AzDHDHqkm0bLn3m3jMjro4udGA3stI1ZRo0RzTO3WlayP62Ea7LEbYhpIyHKfBHstrOIEzM0QQkhkjvk+kpBKR72efQewd0KakbHhTBes+knaxbE4ddXfnp7je+LpU2scLUo0itXE4ZG5qkndDs4zdg981aaUjOleBw0zB3u2CtbtvQ3os7uBBpZOy11zggZUbzZqDzexz3AHYpkrOZWTijMZMQOmfs7KHCOEEEIIIb5rtHEUQgghhBBRaOMohBBCCCGiWKvGsSb71L5LKPyDjz6CPq02/g//7OLStH/1+efQpyEax9RpeFKSKNfLdaoK9YwVOXbbJeGdEQ1TRbRHmdNIZS28Tywhttdt5KTPrtOQhhDCTz629/itN1GT1s7tOR0d7EGf5QT1mmFhtRyLS5KEvVxBDE5dmqnXzns/+CHEfvTTf2rah3v70CfvEd3T0o6DhuiQZkscF4vCjoPLKWqhXlzZefb4GMdlRuZdx2sce7jObA5R19VzCcB9O4QQBl3UGO1v2vuy8pn4QwgVGfe90q4/HTJ/Wi6W9ofQh2nUgk/uneK1eG0UJaaP+M5JSKGG/si+A7b3DqDPYITavbRl9YNpjS+Y9mgbYlXLHut0hmPj0ydnpl2QdyzLO99q23PaHOE4b5ExnDSv1ji2W3gOq8zqrivyuWWBa9LOhtUrtjKSpHvp9P9zXA+yCZ5Td2RjG5v4ju/1UXfpNaPLJeqpY9EvjkIIIYQQIgptHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFGs1RwzX6BZJG9Zgfr+PortVysUg3/+61+bdn+IItnZGBOPdpwRhYlW5/O5aVcFClSzHG/dzo5NFvrGvTehz+4+mkx6fStoTojQvKmIkN4ZdNhfAT2SIPXOkU2yfkTuuc+XzBKtrgYoxC6n1phwQRKAL0kS9NSZJaS1f/380R/9ZxA7vHVk2p02josqwTk1W1oBeV3hAx5PMAH41fXYtM8vMZGuN5nkHRyXK5JgeLGy83pGzmleoUI/d4OzlaDRp0ucamdju/7dXuC6ckQMM/vO+FIWuB52XaxNzAYtsoYkLsF6aOHzbMjzDN5USIwMYv0MtvD9Mtqx63t/Aw0tgSQA90YtZtyqiYOlcm+igphqekNroMlqHK9FTYyVzpyywKkXQo6JvPPMGt9YovuEJO1vuvY8rwr8wuUFvuPGdgsR2hka73zS/CYjpp7WJcSyjjUW9YdojtkjBqijA7s/ef7yFPr8FCIc/eIohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIYq0aR5+8M4QQBk6b2CZ9NjYwoXCnYzUDwyH2SYlOzhcybyrULHiNY0KShG+QhKn3H9w37YfvvQ99Dg8PIdbt2cS8DdF7lCtM1okSEKKNJLqUttOVJi0cBv6SveYxhBAG5J4fOA1cPUNN2slkDLHG6cuSRH/TvG4OmR635/W4+JzKGsdc5TR41QrH+IuXzyH22WefmPazp4+hj5+vTH+ck6TAjdMBNmSeF0S8lzo92IposcoS15XVjdU4NhlqOjOiue67y2Hrg9dhF0QX3l2hzrMztOfe6qFWPGmhPitp3P1kFQrE2nnz4Q8gduvNt02bPeMV0fc2LrF1SbT+s9kcYn4+TqfoNfDeBqZ59u+EEEIo3ZwtiN6X+QH6LiH27gHq+h88eAtit4/s++z05CX0+eqLX0Hs0YsXpt3JUHfZchuUbhfXrYxpjlO7X2h18X16M8F70E7tmvSrXz+DPrHo7SyEEEIIIaLQxlEIIYQQQkShjaMQQgghhIhCG0chhBBCCBHFWs0xPSdQDSGEwWDwys/lRDDuDSubG2hW6XdQ1J27z82IUcN7TFiS4zfffANiP/2pTZ/58OG70GdrC5N1tkDMTxKBNii2b7wov8Y+KRH3+0hKsm2n7ib4dggh5CRh6WDDXt9wcwf6nGYoym2coaJRRuHXTreNgu7gxOkpMaIkLJGtG9O1N1eEEK7GNxB79sIaZl4SA03i1wdi1PAJ5mmMJCouSzQE+F4sOT4zuVTu+8Ylft+JzxwcQmi5ZOK3tnGt6xTWeNMucL4WJHlx7ZMQJ8RYRGKpe8bysn0/OLpzD2I+2TaO6BACMYnWLlaSJPqnZ8cQOzm2c/TsGOds48xqCTFxpoG9z2y/mhhhCnKsnttnPPwATUQ/+tGPILZc2nl1fIpJs6crPM9zN48nY/zcILefO9xB01KbFTOoXcLxjBUHwTVpZ2jXpEdPT6BPLJruQgghhBAiCm0chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUbx2c0zPVU0pChTgLpaYHd57NfZ2d6ELmk5CmI1tJRNm+kicuH5/H7PM/+QnP4aYF9fu7eHnul007GReaM4qwBCRPqtq4UlSFOmn7vqoOSbx5hgkI+VkOq4qwWCIZqCshYLf0omcWcUbsWbIM/Bj04+lEEJIcxxzTcuN8QznpjePhBDCeGarq8xJRZTMfR8zplDDjAvV5PuZIcAfqt1Fs8rhwQHEtjY3TbtHTHdnLx9B7HRizyFJsYrUZs/ezw5ZR7NiBrFJaZ9Lv0RD1GgDZ79fy1vMSCXWzgEZd75iW0LeGw15B/hFn5neKvKuWrm1fLpYQB9/LG+2CiGElFR7arm5zSregFkuhLDl9gc/JEaYh+9/ALFPPrGVqyakUs5kjtdXuFvctMialNvrW5L1ZzHDOTt3czvNce512pcQe3nacm007MSiXxyFEEIIIUQU2jgKIYQQQogotHEUQgghhBBRrFXj2O2its1rY5ZL1O9MJhP8XMt+7ujoCPr0yfedvLRagx7RHOZOM/DwXUzk/eMfo8bx4cOH9vtJcvPUC6sCJs9NiN6E6bZ88t6iQB0k+1yaOk0E/fPh1Vo2lvQ3c/eu1cN7kLd7ECsLp13zyc3F2vHzIATUJuVEz5gTvVuS2Vi3h4Pnzt07EDu8dcu0hxsj6DOZWB1kkrJk9Xiefm4sVqgLZIN8tGl1ux/+8CPo8+YbWCDA36vJDSY8f/4Ck+NfXNn17+wKNVXbfauFGnRwae/1cD3sDuyx+nPUWe2T27Lv7gvTk4v102mR51B5jeyrE7qHgNphpl3e2tmG2MHRLdcHi0Bc31ivQUUSkLfb+G4eDO38H0+n0Ccjn9s7ODTt94iece8APQnhU6txnM5R4zhb4Z6ldHpFUlsgjF1y8fkS9zklSdofEqsv7nbJGkze18/cmntyrgTgQgghhBDiO0YbRyGEEEIIEYU2jkIIIYQQIgptHIUQQgghRBRrVTR3iGg1cwmqZyThJYttbtlkuttbmGi618EEux0n8GUJuQ+cSPbem/egjxfthxBC3yXF9Ym9QwghYeYYnyy0QWPIeDKG2Je//rVpHx+/xHPqYtL1g0MrFPaJiUMIoe+F9Dn5G4MIcIMTXuctNMK0umiYWS2s6LgOJOm7WC/k8frk+K0WzrF2B00YPuF3RRL33rmD5pj7D+6b9ldffgl9fvXZZ7/1HEMIoUPOKXcGu6LGOcau5d6DB6b9T3/6U+jz5j1cM548ssm9X748hj43UzS+jOdWfF+RJOhXY2sSGHVJIu8hzsX22AryW5co0D+/wth4as/p9i1MPC3WT8IKJ9QuRsxj3BzjksoTU80dYgL7yCX8fvb0KfT51JlOLi8xYTWbx6ORNcd0iQE1a+E7ve8KjXSIgS8lRri2S56eEfMRsa+EwhXsqIjZc+6Sidclzus8w/Nsd5wpdobr1nJBDDvu+MdnuP7Eol8chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUWjjKIQQQggholirOWY0GkLMV5OpG5SasooHt49u2z6kSgzLor/tKk9sjLASxd7enmnv7O6Sc0KhuRfbMyMMM5T4EMuif319DbEvnVHgM2cSCCGEPqnc8vbbb5v2vXtvQp9bh1bs3m6hyaah5hhrPmJC5RapHJNkVoSckHEg1sv1DYquh24Oj0hFFl/VKYQQUmeOKYlAf9/NuxBCeM9VbWJC+ydPn9hjlyhEZ+tD15nZQobrRbuD4/cdd04f/uAH0OfWEZrnfv7zn5v28xcvoM81qZK1cBWhVksU0U9r22dZ4G8CTYIlYLKVvS/pnBjzFnis2lWv6JH7JNYPM5T491CW4vNklZVyZ3xLEuzDTKkffmjnw5xUd0ndOfzyl7+APjV5Dw6dGaY/2oA+OTG0dd372ptXQgihJgaW3BmEWm00A3oTUQgh+Fd/m5hLp860tKqxT0q+r6jtfZlNcG+QJ/i52rkdb6b4uVj0i6MQQgghhIhCG0chhBBCCBGFNo5CCCGEECKKtWocvT4qhBB6fas9KAvU4WxtbUNsZ2fHtAderxRC6BJ9QFPahKUsAXjbJSr3eowQQkiJJsQnUU1S7BMCart83tEKpRbh6vIKYo9cQuFf/uKXeE7kHF4e28SfFxcX0OfjH31k2g/uY0Jjen8bdzEkaWya4z1Pc6uLa2ocB2K9/OX//f9AbNfpfX/6R5j8+qiPGtbMJd5n8tish2P1gw/eN+0J0QD6RLpf/PoL6NMhGryu0/9ubO9An7JGndVdl6h8MMC1Z0rOczy2mtHrMWpI5yS5d+X0vg3RTmdOn5nk2GdJrqV0xRW8DiqEELoFWSM7N6Z9tIPPXKyfjCSMhkfK5h55T3jdI3ufsc+13Fr+8cc/hj4+IfcR0QR//jnO49HQ7iEOnNchBK57TNw7nBUVGQ5xf+Lf1p026id3tnDd2HT7kXKJ37dwyf7rBHWXpX+fhhCmE6sZXcww2ffWBs7Hxj3PhuxFYtEvjkIIIYQQIgptHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFFMixvmAAAgAElEQVSs1RzT65Mk3c6c0m6juLffZ0mz7Z63nRORLjG1ZC4peIuI5jMn7g3UHIMxf07sc03DBKmN64PGEDh2CKFY2X43JEn4kojt53NrJiBa6bC1uWnahwcH0KdNEj37gyUssSxLouoSgDcpnrdYL3/1Vz+D2P0H9037HjFNHR4dQQzMG0SYnSYY8wl/33v4EPqULpnvYIhJ78c3NxDrOHNM1sJxeXWDc2rLJT1OyAy6JGa2qUuEPFsuoE9FDCwrZxisSlwfVqWdLwWZPnOSyLtpfMJ+nNN1QPF9sbLnfn6F91esH2Z4qFxC/ISMsTYxffhk4hlZy4l3A/rt7++T72u5Pvh++fAHP4TY1eWlafcGaGgZbmJS8tnCjldWJIAlBffXwky4ezukQIhL+L2YohHu8aNnts8C5/WSmIXnM7uO9Dt4ThubaCjudW2/4RBNRLHoF0chhBBCCBGFNo5CCCGEECIKbRyFEEIIIUQU2jgKIYQQQogo1mqO6XaJmSJYoW6HGSe8WSWEkCYuqz0R/GZEKNxylUxYBRhv6EhzvE0+E/1/OikfgC5NTcwx/tyJEcYLW0PACjeBGG9WRIB/emrF7kz0vLe7Z9rvv/8+9BkSE4K/BUmO19Ii1XpyJ8pvVmsdmoLw4vlziG1uWkH1eDyFPqxaEZhjiBEm4BQOqTOesAoTnY5dM3xVqRBC+Oyzz/FzPTunCjJ/xqQCTLdrjQQlKfXEqjHNl3beLZfoYKHVHHzlmBq/b+HE/wtyMzOyruSpnYvdHjmnBp/xamVNCY+en0IfsX6ePHkCsfPzc9NeknfCQ7K+H92xVdY2NtFAw15nfgi3Ozjudtz7ZUiqvRySue6rpmTk3Zz792IIYbawc49Vg2PGXF9Zrt3G72sKvJ+LlZ2jz1+8gD6nF9bocznG45BtTeh17HPY3T6EPlsbaI7xe52d7T3oE4t+cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxViFZi2xTFzObGDPpYUJPn4g0hBCSzMUaFAPUDWqBKifKqFMUaWROqJewLKc0bTaLWZgkpHbaqpoIR4YbqAG5c/u2ad+9exf6PHr8CGJlae9Vu4WakK5LKkpkVaGsSBJnp61qSKL0br8HsbZLxF5OUVcl1ovX94QQwuaGTa67vYV6wnYb9bghtTpEIrejScG9lrgLOuIQbrWtFson6A4hhMNDTDBcuq+7usEkvddj1Dh6TWVN1p4Lp18KIYSV0zT2ujgPEnKs5dxqk8ZjTFTslwyS3zj0e6gfT1zi/QUpGLCYYxLi8aZdp8ezOfQR6+fv//7vIfb0yWPTnpJk1EyzP3AJore2MdE1e6HBDGV2APd9TKvY6+H82HLJvasK50tF3sODkTtR8qpuyNxrtZzfIeD8WM5wrl9f2Xv8y09RY311Y99xdYPa8GF/BLHDfatppAnWSWET74EYEV1pLPrFUQghhBBCRKGNoxBCCCGEiEIbRyGEEEIIEYU2jkIIIYQQIoq1mmM63tASQlhOrfg8JUaUpCHmGJdUlArrmWEmONU4EZH7LNYJMXhQcb8/BWLqITmGQeBbk6yf/T4aDm45c8z9B/ehz2w+g9h8YQW+m5toJtjatKaHJEHhrjfZfIO9oTm5Ue0eXkvHCaFnxAQh1ss777wDsQ8+/MC09w/QdNKQsVK5eZ3mJPE+Mar5fkmJSXrrwo7nLlG+3z7EZMIr93WjTUyau0vmT+USfvd6mBh5VaCI3hsCNkcofN8a4NyYjK9MezlHI0qeWzF8kpP5QxL9LwprfCkWmIS4laKp5solfn9+cozfJ9bOX/6H/wCx81P7bFotfO2/8+57EPMJ6hOylrOYfzmS12BI3fuaFfCoSzSBZb4feaE2bL/gxn5Z4vycz4hpqLaJw/d3NqHPLbIGPv7aFk84PsGCAO22LaCxsYlGw91tjG24daNNiqawPYS/V60WK8gSh35xFEIIIYQQUWjjKIQQQgghotDGUQghhBBCRKGNoxBCCCGEiGKt5pg2qxwzsYLUsqBuFYxk9tRbRGyPbhXMDt8QkSx8jmW+J8eOsXPUJNN9WVqB+oqIghuiMN51GePf++BD6NMbYCWe6dQK4IcDzCB/cMsab0KCQ2W+wIoSdW1jfVIuqNXBigA9V6XgpnUGfcR6+S//2Z9A7M7dN0y71UVjCKs+0smsEDsnhjNWhiJxlYjqBueGnz91hePSV4MKIYTMGXZGxKwyGOH8KdyaMR7fQJ/FEk01s4k1uSTkPDvEuNC4CjOrJX5u6WKsesZyuYTYysWSmpjZRjhfz2/sNV85A494PTx98jXESmfU2tvDSiMJGS+t3JouCmKG9KaTEEJIvLmUvBhrX2GMmFyqCud65eZeU+F+oSbvqsqVUpreXEOf6RgrwAzb9ljvvfUW9Lk8xbF/dvHvTXtraw/6bG1bU83mFlbm6fVxffVrJ6s0V7NSbw5mqolFvzgKIYQQQogotHEUQgghhBBRaOMohBBCCCGiWKvGcTFB3U9R21jVoF6obvA0U6+ZIolzk/TVCaprppHwuiqivcqIJiSpXYzoICuiEyl9El6mcSTft+ESFt9/8Db22cCkxuOJSyDcoE6l17d6r8mMJExdMn2oPfdkRJJ9tzoQ629YjWNniNoysV4evv9DiHVdsuuajMs5SSIdUjuHmwTnQUa0Oo3TAdYr1OmVCxtrSjKnSVLgymuhSJ+iwO+bz23Rgouzl3js1RRiq5ld2ybXqON9PsY18vzK6rGevTjF83S3kydmJiGn+R4OMMHxxhYmIZ7N7D1ISHEHsX46bdTE9d2cfUB0erduHUFsMLAJqldEu5wT44KXE5OaHiE4DV5Vknld4Dri1wM6zgmzqUtY//Qp9FkQna6/vJMTnHuff/Y1xPp9O48Obt3HY7sE4EmG+5y8hfPKaxPnc7xPXvMcAmpNpXEUQgghhBDfOdo4CiGEEEKIKLRxFEIIIYQQUWjjKIQQQggholirOebmEo0v04UV3M69yjuEkHUGEOv2rOliSMwxLKFwcGLwmiQAr705hYjtM3LrEp/UmJljCpLA2JljWBL0huzxe31rIMmcYSiEEIYbKGwfOwG+b4cQwnRixcqn55gwtapJAuOONdqkdw6gz/4GPqv+hhUTb+5hMlSxXl6e4XzddN6JVY3Gqi4Oi1A5hXynQhF/xsxspTVy1Us0nQSXKDgnIvOQohC8dmvBcorz4PriBGJLdw6rCc6NjR5+3962vXmnxyi0f/ESv+/4zCYmXpA1MiT2ObRJQQQmhh+6AgE72zjvNja2IFa6dbPV+vZCe/GH4+Mf/xHENrfsM753/z70Obp7F2Le67kkxjRmjstq+x5ixqmmsuOnXODcq1do+vCFN3I27ojJ7vSlnWuf/uIz6HN9hvOxWtnF7JiYY84v8Nxv33lg2uz97fObV8So22SvTozOkqezhO6Jc8esVmShjkS/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxVnPMF58/htjMCTQrkgm+1R1BbOAqm+ztEjMFMacELyRtUGi+dJUoFgUK8pcXGMs61vSRt9AA0FT4fVXp7oFXzYYQKlJNpvIVZwo0+ixJdZfrG3vu52eYMf/4+Ny0Ly8voE9BxNLdnh1SSYOC38692xA73LYGqOEWVrwR6+XP/+L/hdjhwaFpHxzuQ5+9XTRkdXs9087baOSqijnE+i07X7YHuGSN+q4SkfuuEHDahxDCZGznwYvnWAHm2eMv8ft69twHXbyWrR4a+jZG9r48eYpGmKtrXFdSZ+wZDrHykjendFp4ToMBVmMa9O15djt47xJy87odu7b1uni9Yv388T//U4gNh/a9tL2DZqcN73oLITTBGqwKYtosa3wHhOCqPZHxM3WGsmY5gT59UjVlNPTGSuxTrvA8v/zikWn//B8+hz4Xp+cQqwt77gtibi0anGuNL9NSEgOL2wqkxMzLKtst3f6g9C6mEEJDjLL+XpXlt6/2pF8chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUaxV4/jLT7+C2Mr9D7/Tw+TQne4GxPb2rNaqLlE72JBtsf+vfkq0T/OJ1RmdXqL+4vSG6PuGVifSH+B5p0RW0NT23EuiZ1wu8fvmc6sJm01RHzW+GUPs4sImFD49QW3Hi+dWf3V1jTrIFUnQ2unaIZWTZzDsoUZr1yWp7Q5RjyXWy1/91d9BrO/0gyzxPtVQjezz7PTxc9UKE+k+uGN1gR+/fx+/74FNXpymmPx6ucKJ/uirp6b983/4OfR5+RR12bd27fVtDFHL/JQk8v67X/zKtOsGz3N39xBiSWr1ShlJetxymsacJABPfYGCEELqdGwJuXc+UXoIIWQuyfrGBmrkxPoZjFBf3HHrbVHhM764JIn1r+37JUnwcwnxJFQuAfeYJNZ//sRqDvsZvvPevY9a+E33XshS1PLNJvi+fvr4hWmfHGPS/skUx7nPJc4SnkPhjxBC6u8LSYLu72ZS43asIMnM/Vmy5xLIc/FJwb3G8ndBvzgKIYQQQogotHEUQgghhBBRaOMohBBCCCGi0MZRCCGEEEJEkTQsM64QQgghhBAO/eIohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRhTaOQgghhBAiCm0chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUWjjKIQQQgghotDGUQghhBBCRKGNoxBCCCGEiEIbRyGEEEIIEYU2jkIIIYQQIgptHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkShjaMQQgghhIhCG0chhBBCCBGFNo5CCCGEECIKbRyFEEIIIUQU2jgKIYQQQogotHEUQgghhBBRaOMohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRhTaOQgghhBAiinydX/bv/rf/ufGxx89emPbV9TV8Ls9bEPvogw9Me9gfQp8sw8+FkJhWmuItSLPM9cH9dZqRWGKPnWV47CTUECtWc9O+Hl9An1W5hFhd22Ml7tq+gVyfP3X2sQCPKoqmsZ+r67jjwOcavE//6r/7H+iZiu+G45//DTy8JEldO/NdQkgxlrhBl5D5kyT4eP3cS8ixm9LOn8UM15DFHGNVtfJngOdE5o+/ZrI80DnVfNs55eZQQ+ZG4w8NAf652vUjj4AdCtYatka+/cd/pvm6Zv7q3/yv+LQSG8rTNnTZ2NiDWLtr36kN+Z2pql49puqqgj6Nj9Ul9qHj3H5uvpxBn19//RXEcrfebIxwv7AxGkCsWLn3LrmWovDrSAir5cK0mxo/B+9YOmcxlqVuWpH5WdV47/xkLCu85//1//g/Rc1Z/eIohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIYq0ax9kCdXrPX7407fML1Pf1uj2I3Tk6Mm2mg+x2iI4qeI0W0V+A9ABFBHVFhAVO55PSY6P2oHJag4boRgLRCnoxAtM40hh8kGmvXq2j4IqtV+u42P1EjeMrDyNeB03y25ohhN8gmYXjkM+xceg1eEQjjLplMu/Z9/l5Tgf5q7+vYVccqXvE74v4HLvpMfOOxWAu4rHZfMXlQRP2ewERqYL89Q+ov03I51KYo9incn2Ypt3rGUMIoXJayMViDn0eP34EMa/nO9jfhz4P33kbYqnTdK9AFx3CeDyB2GRqY22i6W63rFa6nRM9NX1XugcYKUz+Q85Q/eIohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIYq0axxfHxxB7+dLGbsY30Kff70PszGkhhwPMwZSTPIo+tyKRIQJM45MmRPvU2H14RTVTRLdRWv0FzXtF8kC9WvX4m3RjXiOBXSIOTT/m9RdMfhED03aINUO1bT53KPscGYeNbxMNTs0GmcthSMYzzA0ikGW6YYhRXRCZZcmrNY509EZNuz9M6sNvPXteLWf8ppu7V1QHKb6fsEdF8v4FP6/o3MNcgD6/sG9/Eyt/azuEECryzpsvbN7Gpy+eQ5+vHn0NscXC5lU8OzuHPkwb3e91THvljhNCCJfneKyVy/+4McQ9zKbLG9ke4h6mjjAX+Ny633wO8XpUn9vyd0G/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxVnPM148eQ+z49MS0+z1M9r2zsw2x7a0N026R5Jks2XYdIfj16tOkwf11Q1wftRP3N0QUzEwu5comFfVmmRD4tQDMiUJMPI0TAcfk/143r/nrRQihYRJrnzSbJaYnD692g4x5y7i5y88pYkTxCfRL4nij5hjXj82xJsOQmz8NWx/wSCGkPpn5H3CUwxd+O7MKNwORfpCwX+aY7wX0dWbHNUusXRYFxLLMJbtOcC74xNohoBmGvQdr9zlmhGGxLLWDcXNzA/o8uH8PYp99/oVpf/HrL6DPkiQT39zcNO0WKTSSEXNKWdr7yaZHp2X3LINeF/owww7O0bi5Bwa2mD3Fb0C/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxVnPMs2fPIDadTkybiV13iTlmz8W6nXbUOVReEM/06U7wnxCBqhft01j6alFwCChMppVjAitx40wuLLN/yrL9u+of7Fq+VSQEFOpKNP+PlbJYQizLvDicjAsy7sFgFllSCCrM0Plq50ZNzGU1M8eUTsRPzEDMOFb7ayHKdyo790F2LbRCk6vGRMu7+Go27OtfXRmHGmHIsaByDL9isWaYMcybYViBpnKJcz1xZpgkRXMMvE8DMU7RyjG+ugwepybvvMHAmmd7/VvQZzzF6nOn52em/eTxI+jzxRdomBlt2P3Izs4e9Nnb2YVY49agIsObnrp57U1+/38vOLa/n8ScA1V/Qgh15fYZpE8s+sVRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRxVo1jscvX0Kscf/673YxCeb2Nmocbx1arQHJaRqurscQWy5sUtNWhrcgc5pGloQzIboCrwFh2siqxESrhdM4Mt0IE1t5rVMTkcw8hBBQosWO7XVVBBb8lpJGdg7i9XJxcQqxra0d005B8xhCQkRUXkcbnfy6ebUGLzjND0tmXBVMW2zXAiYVCgF1XaF2HYmOOCGJ0SFPeEqSmZOYD9F7B5pDUqCAJPz1844tISk7Tzi4NI7fBxqm5XXPvSLPeJWgxtEn/E7Ju9Inuv7mHOxcq2ucez6594poM0sypvb37F6A5CQPrTauSYU7T/JqDg25MZU7r4KsIwuSODz3utIUP5cU9nPVgsyhDE+0rvxcR5h+sVzZZ+w1j78L+sVRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRhTaOQgghhBAiirWaYxg+OelytYA+q3IFMW9Y6ZEE4BeX1xCbz6c20EIhbe0U4lzI/+oYM9V4sW0IIRROpOqFwyGEkJIEopC4myUPZknBiXD/VcemAnnvbCIwz4tPOvxN7Le3xfrJyDP3Q64iAuuEjAswWzGjBhn3XmjPkuTWzuSyIAmAV0sUsPtx2Op28NjMHJPa88xSXEZZUv3GHaomc7oiYnj/HFJmzPP3nN1fNu8iEqw33gwUAphh5G37fsAMHt4UxQyaLAH4cmLNpex9xuZs3dg5WpL3d+UGTNrCudfqYzGQVm7P4fTyAvqcnBxDbDy+8mf5ynP6ppc7T7Imzie43vRz14+YCFcze3/nFa5RbD5WMPfiJl/tnjt7drHoF0chhBBCCBGFNo5CCCGEECIKbRyFEEIIIUQU2jgKIYQQQogoXr85xgk9JxOs9nJ2fg6xk7Mz035w9w70abfw8lIndp1P0UBTF9agk5L9NRWouxirLlMSc0zlzinJUJDfJuYff1o1ES8TLe9vKL/huvgPEoG8r1zzn4IGKtulBW6krv++sbmDFZtauR2H15eX+EFS+SjzY7whFSdIxRcv2i8XM+zjKjfMqTkGTXfeNNDq9qBPqzeEWLs3MO20jdWuGjL3a2d8qXOc53WOa1bjKlKxOZ36yjxkPn3rGUYq43jvTYzhTnz3rIh5LDgD23KK79hijvOqcfM4JWOqJOY4H2N98m7ftHcO8f29s70DsZuJPc+vv/oa+jx/8gRiofTmH7yWosR7N1vYdWMynUCfzR6+m3P3vs4TNOMUC2vUTStcD9irOqZGE6sc1bgKPsygGIt+cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxVo0jyxdduySUNUkWOr72yTtDePL0mWnvbm5Bn3Ybk262nK7o+gr1HpMrq59kGscsRT2C78d0kBVLzOv65R2SiLgaQCzvWJ1ImpMErRAJIaXJy/3n3LFSPG/2VwcmS8fvqqn+KibpulgnOwf7EKtLOw7GROO4mKMOqFna2Pz6BPpUC9Qh+ljJ+jgdZEHWkJJovzABONE4krnYblv9UruDGse0g/M17bpYG7VRTY5rVuJjJJlw4mIwf0MIgaxHpBM5J9Kttv3q6tvrpcQfjtIXuAghzF3y69k1Js32ersQQvC1I6hmn7zPfCJtVtCi7zYDeY5zYXNjE2LjmV1H2m2cnzvb+LnLCzu3WTLzxQrXjXlhr4+99wcd1GIORiPTHvVwEmWukEBFC52Q+QhtWvkDY81vbf5O6BdHIYQQQggRhTaOQgghhBAiCm0chRBCCCFEFNo4CiGEEEKIKNZqjilI8uvaJRltiJB2PkOx/fPnz0379uEt6HNrH2MbQ5vQ9+YMRaTzmU0ozGwazGDiY2xXXjV4fV7cmhPBb03E/V7KT1KEhxBQSF+7M6PJit3l5czZRAS4PkJzfZOo98skJOmwWC9nJy8gtn9w27SHmyPoUy5wvk4ubVLuq5Pn0Gc1Y0mIrTi9JmuIF/HXNc6VmiQhXhYuufgNFgNIyBhvt+yc6g/60KdFDDO5M8PkJHF43kNTTeZMcE0bTTwQa2GfJMd1JUky14YuPNG///5K8/X7wOzyGGMTO66XxEBTkeTXlXumsxXOvfEczWrzpTN5EEPJ7oGdQ3ukwERJ3qDdvn1/HxweQp9lMYfYxZU1CHVP0NRXXqBRdr6y15cQU83GCIsEZG5uNxkzFrmk5MRg1mL7DGeY4TMvwlz6e0xZ/eIohBBCCCGi0MZRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRxVrNMcsCM6MnzgzDMqX3eygi39zcMG2WQZ5WdHAZ6gtS8ODGmWNCTUSrRPDrq9LkREjbEHOMV6k2Ab8vQT1zSPyx+mgAaHVRuB/cvWoSHAYVFIDhdWI83tPCrqUhlWOgjyrHvHZ+/jc/g9gHH35s2qPNXeizd/QGxMqpFaeviND+9BwrRI2ndi7O57iGrJxhpiBC/6LEuVEUNuarWIUQQotUY+q5dWVzhIaWjQGuR5sD+7lOG+ddkqLFLWvZz+VkTneGtlpGMsDnknbRyNT2xybVtnilKVc5hpiWxPq5JKazy7E1q91M0YQ2XeC8mi3tPLqZLaHPdIGxVWHHQm+I464e7Nl2jmauFTFk+moy7R7Ol6SF8+pyYq95UuA7aIrLRnh5fGraM2/8CSGcXd1AbOrWt34P14PcOdGqFRqNMvKq9GZA9jZNot6x3x794iiEEEIIIaLQxlEIIYQQQkShjaMQQgghhIhizRpHFBG0nKaxRbSKo9EGxHZ3rYZnMMAknHkLNY6lS3jLtA5nY6s1GN9gYlCWFLed29vZJXqhdgv36m13D9pE+9RfoZakLpwmosQ+VbUJsbzx9wrvk08nXhOdU+XFFiEElKWQPiSW+OMrn/Br58vPfwWxA5cAfLR5AH06ZL4mvS3TvixQI/zJ03OIff3spWnPV0S/6JOEk7GaZfh9vvhASZLslxVq9zKnTdrZwLXn3gHOuzd3rTZxRHRPrYzosjM7r5uUJEpvWx1b0sM1qzPcgdhwaJ9Vp4das24XY7m7n8cn+Ox+AhHxXfPk5AxiL5x2+GpCknaXuOD6hN/nN5jYf0Y0jt4SsLeH77zNif3cZIFzbzLHY6ctfKd6ehuo733vIzsa8+EW9Mk6qFXefPzEtOekSEG/i/N47+jItHdGxGuwsseaVaifrMk7PXW+Ae4GIO9Y93L+fVwE+sVRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRhTaOQgghhBAiirWaYypieOi0rAmj3UWB6vbuHsSObluR/u7+Ph48weSgTWaFrCtyCy7nVhT8/BQTE9dESO8TgDNzTK+D39dx5phuG4X8oz6K5reGVpRfkaSf3QoTcLcrm/g4rzBBa+aTDDf4/UlGhMpOcZsQAw1X5bok6CT5q1gvO9soIO8680TWJgmryRxujaxgvepuQ5/nUxyrn7y4Nm2WqHjpDCxphn8PD4jpo+X6xf4VnWcuce8Cz3t1hhn7Tyc2dm8fRfz9FjGTVdbMsCTJtkufTJjMzZokF89bdp5vbeFae/vWIcQO96z554svn0IfsX7+7ktMAH5yZY1SVUMKU5DxsnTvjmlJClPkOKbaLlYn+D47u7Dv1ONTNPVs7aCZK3PJvdMM3xMNKbzRH9nx+u7D96HP4SGO8y8++9y0L87RBMb2Am+8cde0t4dojqmcOeYixfs7u8Lvq+bWpERLc9DXrr1XSgAuhBBCCCG+c7RxFEIIIYQQUWjjKIQQQgghotDGUQghhBBCRLFWcwyrSnB0x5pc3n7vA+xz9x7Eei7ze53gpZTEjVM408Wixr3z1cKKz5+e3UCfyhlMQgghc2L7Hskov0FMLr5yTCvD8x5OMdv/dGlFuSUxx2yWeJ690hoMOqRPxwl+U18OIISQtPF5Nq6iRJPg/U2Y78UJmhv8OrFm9m/fgdjuLRvrbaCBpiJzKh9YcXo6QOH7uMDPlYkV7dcZDow0tWOuIrLvJVWC28/lxFTjDW8hhDAcWTPZ4QEaShip0OEAABxKSURBVHY20XB28sJWofjbJ5fQpynw+nI3h5KAfbI8cW0875Is91Wwc3gwwOogJxdYLeOjt60Z8dePX0AfsX6u0DsWqtw+45r8XsSMq0XtqpYQ00lGYj5ULPHd9fK5NVN99tmn0GdIqq1sbNl51eqgqYedk4/12OdGOGd/8N67pl0V96GPN9mFEMLQvedZn9JVg8trNNlc4TQO1ye2X1WQh07WCG+OgWptvwP6xVEIIYQQQkShjaMQQgghhIhCG0chhBBCCBHFWjWO9+8/gNiDt+7b9rsPoc/+EWqt0pbVENzMltCnxjy5YTa3/SbTOfSZL6zmL81RD4FqhBCW7gurJfaqE9QTtpzUoCH6yYsUdUY3c9uvJEmzWykRrzgBYckSChf23LukT2eAx047NvlzmhIdBYn5v2AalsFUrJXtW29CLOvapPPLAsd4WRBtcWHHT5LinJqR5N65SzCeEx1vcNqgpsKxWpV4ngvXL89QUDQYohbzwdtvm/Z7D9+FPq0WXt/ppU1mfn6BusDJBOf5cmnvSytB/dKgY899NMDEzEWD17cs7frXmuB9mi3xfo469hxuxmPoI9bPHnlXTqZ2TF1eX2OfG3x+lZtXKZl7dJl2w3NFEmT7uXd+dYHnNMe5kLbtliUnxTLYK8dvBhoiom+3cDuUuSIbDdlUsF/f2rmN5uSkeh2/tmEC8jaZ6yv3PKc3+Dwbcs+Z3+Dbol8chRBCCCFEFNo4CiGEEEKIKLRxFEIIIYQQUWjjKIQQQggholirOebjj38EsYfvWTPM/sEB9Gl1MWn2ygl1mwINJVWBQtaTsxPTPj09gT7T2dS00wxvExO7ekFqQzJdlyTpZuaPT0SsFRH8zl2i5csZmgv6N1OI7TlBc4eaGawotyIZYhvyd0c7tUnP8xZJAA4RFDQrAfjrZ/f2fYg1mRV0L5Y45ooVzruLy3PTfvb0EfS5ujzDYy2teSMhAv3MJfwmHjEqhq98R5I4+C1nhAkhhD/+kz95ZZ9PP/kET8KZ7GYlntOMGHtWLlaRAgF5ba8lJ6alGXlWKyf2T5d481YrPM/tvr1X1xNcZ8T6ef8HP4TY5ZU1T3z99dfQ5+oai1z4qZZFFnOo3FxjxjSftD8jZrLBaICxoY2RPPehIYm0K/8+w49Rsty+m1OyZSK5vUOW+DbeKF9woD/chD4ZKbxxeW4LB8xnaPBlxUB8JKGVOOLQL45CCCGEECIKbRyFEEIIIUQU2jgKIYQQQogo1qpxfPfBPYjtbm2YdqeNWoeUaI+8ZKlqmDYINT3HJ1bT+PLlc+gzvrYaAibkYFqrxJ8USY5aluRYrt0jms5+D2Ndl0B0Qe7B8Q3qH7odWzy+InrCtLSa0XmFGqZphQKTfmWH1Gi0AX26nQ7E/E1gmjSxXtr9IcS8eqgmiWZ94uAQQli58XQ9QU1VQZLrLlZ2DidEW+xjGelTE+Gjl+0yndUb9+9D7AcffWTaB4e3oM/f/u3fQmw6X5j2fIVFCxqy1nkhV0Hu783CHmu6XECfJdEqNom95lYLF63FEs/z0bE9T1ZEQKyft97BZPQeNs4fPXkCsZUbn502JpVnybYXbiwWRLe7s2n1fIeHmPz6zh1MZt7v99z343itSQENiJD3S03mnn/PM10gOwd/JFoHw2kcs6SHnQriWxjad2rWwuTpVU00jl7PTDOlx6FfHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaJYqznm6GAPYi1nlEiIwSNtWIZNZ8xI0KiRNnh5M5dQeDKfQZ+lSybuRazffB8Tydp2TQw0FTETZE6U2x9g4tPbt29DbGdry7TB1BNCOD9+BrEvXth+G10U5XacIL/VQWF0Z47XMnRa3romCc+38Vm1XBL0mjl2xFpJiLvLC8FTluA9I8l8t0a2vTmCPhubaKS6Ho9NuyAmjHbbjp0uGc8JEb5PF9ZAMhiiGWh3D9eso9t3XtlntInJfNvO9Eb066EkRoLKidpLWuwAJh70SUgC5yy3sbpEI0yxQoPdk2N7Tn7+iteDH2MhhHDvnjWl+vdbCCH87K//GmKzycS0+11iaiTjbOUSzTMz5NtvvWXa7777DvR54+5diLWcsSf3mbZDCHVF1iR3nsUKTScNeV9712ZKk6ATw55fJyNMfcwEnLXxefYG3hyDfVKyjtSlu2aZY4QQQgghxHeNNo5CCCGEECIKbRyFEEIIIUQU2jgKIYQQQogo1qpobuX4db5CSNKgcYKJT3Nv3kjxc1lORPojK8pnWfR9kQkmWPffH0IIqTtPJn5voE5MCD0nyn/n4XvQ5+2334LYfGaNPdc3WI1jskTx8uXcit0vM6wy0W/b6+t20RyT32A1me6NPaebCYrt50u8n7f3t0379AqvRawZUqEkye04yMicbqcYu/PGG6b9x//sT6DPi5cvITZxY/zk9IycqJ1TbTKnhyM04/Sd0aZPzDFlifcg98Y8Iqrf3d3Fc9iwovaErCGhROOCX/6YpL0o7TxnxrwWMTL5NWvlBfQhhOUSY5fObDAkVYbE+mHvyo0NO/a3t7ehz9YWxvrOZNYmJtHVEo1TK1dpaIt83/vvv2/a77z9NvTx5x1CCC03ZzJijglkWtVuHrNKRzfk/Vm78lJtUj3HV3ALIYSu65cxg60zw5BHFwIxGnZ7dq51enHrVu33MTLHCCGEEEKI7xptHIUQQgghRBTaOAohhBBCiCjWqnHMW6gF8EqchGgW8hz3t+22/d9/Ro7dJsmn33jDJhVleo9nLmGp1xKGEEKSEu2B04TMl6jvCySB6N7+vmm//8EH0Ofhe+9C7G/+49+Y9mxB9IQFah3GM6tpbEhC2FHHCkW2NzEped7CxK7jpX2iixKHWJ5jwtLMJUx99OIc+og1Q3RyPpEt0y3TuejmxkOncQohhD/7sz+DWO60y//nn/9f0MfrefzaEEIIO1uYkLs7fLXemWmF5k4jzBL2b5AE4BtO49gjicrZX/J1ae9BSfRSZWH1SxVLZkyKJHil12pFEoCTogX+WKMMn7n4fuCTVndIQu7dnR2I+bdnU+F7YuoS9IcQQq9nx/WHH3wIff7JT35i2m89uA99uiSZuX99Mo1jVeLY9/r/X332OfR58uQJxFK3/hwcHECfg30sALDl5vqwj9eS5/Y50JIXZM62Ok572u1Dn/kc/QdJYddqshWJRr84CiGEEEKIKLRxFEIIIYQQUWjjKIQQQgghotDGUQghhBBCRLFWcwxLFgwyVpKUsk0SeXuBLzNqFBXKTd9805pjWGLt05MT03767Bn08YLjEELo951INcU+SYb34NbRbdP+4EM0xxzeOoTYnzujwMXVFfRZrDB5b+PucU3ueeESs69IYuIFSaLq8qWGVYWi+UHvGj9XWlH+189kjnndZD7RdcDk1zmZm9QE57Lqb2+jGP/jH/8TiI2cgWWHiPg/+9wK3a+vMZHvcIgGlgfv2KTDWRvXECbQv7y6tH16aHLpkGMN+vYc9nb3oc9iNoGYT8ReLDDp8mJu5/lsifPVG2hCCKFZ2X5LYuhLSMrxjjPotDp4n8T6qWp851WkgIVnjxg8+m7s1+wdQIyju7t2jn5IzJ5vP3hg2iOSfD+nSbPtWExZAvAG78HNjTXxsHf6P/zDz/FQzmQ2Gm1AH29uDQGv777bd4QQwt0j+05nCcAbUtgkd3OtTdafbEwKduT2+SXNq8fFb0K/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaLQxlEIIYQQQkSxVnNMQwwlvjJDRQwXYDoJIbRdlYe8jZeSkoIHPsv7w3exIsvzZ89N+4Zkxy9JtZWBqyDhM7yHEEJCDAc7O7Z6Tb+HQvPlYgGxlTO+LIiwvSDCaKgqQf58KJ1taUGud1Vghv46dc8hxXNaLPFaTgt7LZc3aHAQ66UiVVOatn3mKTGAMVONN8f4igwhhLBHqjL4aiu+HUIIX3/9lWm/fPEC+nhxfAgh3L1337R7zogTAq/+VLi5sCrQgDYkYv8tV71mbxeNPsUA14yeq5y1mGFViOPTC3ucMZoWyprMV2emaEi1rRYxO/VctYo8x+cp1s+KjNe5M1P58RtCCHfu3IHY5sjOhxStrKEilYYe3L9v2nfv4rG3trZMu0WqPaXETOvfVSlxlHR6+Lnp1M6Hs7Mz6PPVV19CrCzt/GBmuc3nuN7455CRdfLAmWrYFPJG1hBCSNu2Y7ePa0abmPMat19oKuLGiUS/OAohhBBCiCi0cRRCCCGEEFFo4yiEEEIIIaJYq8bx8dPnELtxWraaJO98/72HEBtuWP1Fq0U0EglqMnKnwfvRDz+EPleXNpF2UaDW6/j4GGI9p3/YIZqpFdFwbrikokw3Np9j0t/SHWtV4XmWJCFs6XSPFdFoLSsbm6NkKjQNnme7a/VQScDznhO95nxhdTfjCWrSxHp5+vQJxPpOx3v77pvQh2p73ZgmU5Mm1c8Hdr6+98EPoM/dN+w5XLkE3SGEcE2S4w/cvGv3UEs9neNYbbftGPfFCEIIYWMTtZi7u1bL3G3h9d6c4XnO3Lw+PsHk+GOn4VoscU7XROOYuHvO1tEueZ5e48jWLLF+aq9fDyEkTgc4GOA4f/stLISx5cbwkGjvW0y7d2B9BGx+pE4HzYqDMI1jQ5LRA2RxGQzttWxvob44J8U5lgubkP/qEl+El5cXEPOFEjZHm9Dn3XfeMe3hEOcZm1WNO3Z3gMUNeiRWu/f8aqYE4EIIIYQQ4jtGG0chhBBCCBGFNo5CCCGEECIKbRyFEEIIIUQUazXHfPLpZxA7O7eJOMsSk5N6I0wIIew5Ae5oIy4BbZpYs8gWSSj8wfvvmXZFkmh/+dVXEPMJuHf2MKHx9RgTW/vE4d68EgImMA0hhMoZX3wy328gYmkXY9dXruxzKEii1VaOoufECfmrKSYrns7QMHNxbc0w0zl+TqyX/+Pf/luI7bkE+v/iX/4r6NMfbUGs1XbCb2KOCUz47sxyLSK033Jmla2dXehTkMTIPhF+Tb5/k8wNT0KS9LY7aDLZ2LJJwVsZHns+xfVhObfnfnKG5pjZzJp4GmqSwN8JWrk9z34XRfXDIa6RrTYmBRevH5aIvd2xz2p7exv6MFPUoGfnbI8k6W6TKdvtWBNNRsZK2rJbj6xFtiIJXkvtCgk0FUlKXuMeYsOZU46OjqDP3btvQOzFi2em7c2830BMom13DzK8d/O5Nauw5OJNg2tE5t7FHZIAfED2TKUrvFEQk2os+sVRCCGEEEJEoY2jEEIIIYSIQhtHIYQQQggRhTaOQgghhBAiirWaY372138NsamrEMJEuu+++y7EFu9ZA0ue4+d4eQoby/qo7n37wT3T7jJB/hYaAF68fGnae4e3oM/L01OI9Z24dUFEq0yUW5bWiNJpo7g21CiuXTkxf7HE76u8USAlQ4Vk2l/57PQLNMJcjYcQOzl3gv84r5P4DvmLv/hLiO3t2YoLTFB+6/Y9iLW79pkT7wZUuPgGV2GCfBCMIMQk1u7gsb0prCSVVbIWCvtrN6eKAo035QrH/dbAriP33jiEPtPLM4j9+isr0J/Nyfe5a8nIfG2R9aHvDBB9Uj2HVf7wRhtuzBNrh7zzul37/IZDXH/bZJy7V2XISFW3NnvHOho2rd07qCEGM+6Vs0FmXatLPM/ByJpF7t+/D33ef/99PAV3Dhl552UZ3rvNTbs/GAxIFbmVPfvFAiu/FRVWgGq7d2Pewvk5IJVqSmcQnM++vQFVvzgKIYQQQogotHEUQgghhBBRaOMohBBCCCGiWKvG8fjlC4hVTqfnk2GHEEJVEf2M09ikJMFmRvQXXjYBWr4QwmBoz+H+PdRxjUaoE3nzTdsvIbqRIUk4nmVWtMB0nkuSwNhrwgZ91Cf12/iI5zP7ffMp6rFy9zGW5Lggz2W1sOfZVPi58Qw1lVfja9PuDVG3IdbL8ckJxPzTrFkCXqIxgr9RSTJqr3v6Bnd8orPymsaa9KkrTApcu7WnIefUEN1j5YoULGYT6DO7uYDY3qZdMz4imqpQ4Pf96nOrcazIPfeaxi5J5M10Vl2nceywZM0puS/+sbDnItZO0qBObjG347Pfx3dXlb5aVN4QvXxS4/clrsgGyUkektyOKTp8Enx3+XHH9gZFgbpAL1bc2EQN4NvvvgOxbs/qgs/PcV43Dc6PzU1bhGB3F/XMjbu+yQzPe0G00p3cXsuoi/uFFtEqj1zi9xXxNsSiXxyFEEIIIUQU2jgKIYQQQogotHEUQgghhBBRaOMohBBCCCGiWKs5ZkmSQbedEWTQR1H35gYKWTsdK/5sEhR1Jy1U5aaZVdemxIiSrKz4PSPJxfeJYHzojD1LIvjtODF6CCGsCvt9nTZ+HxP8585UMyLGok6OQuhJyz72589fQp9V4YwD0COEskDDQVVaAXW/i4L8RcESnbpjLSS2f90wq0rPJRPudnE8++TQIZAE0WT+xJxDQ0Zi09gxV9fECFMTwXzwYwznXVUxs8HMtC9PcP4sp1cQ67hT3xmhUS7UbM3yayTO6c3cPpdeH+ddu4OC+TSzzyEjj4UZ4/w89+uFeD3kCRpYJtfW0NGU2KfXJ8m923ZM5Qn2acj31YWNNSRNdx3seMkaTE6fZOT7Gjs/SnItS2KOKZzJLSdJ7Y/u3IVYf2Dn2tEYk2Yvljj2a5eofFHgtTx9bs2HvnhGCPzejQbuXu1vQ5+tHq5lo21bvKEihsFY9IujEEIIIYSIQhtHIYQQQggRhTaOQgghhBAiCm0chRBCCCFEFGs1x/RYNnNXgeWtt9+CPnduH0HMC3dXrJpCjvtinyA/SfFzqRPy++o2IYSQE3F/11VdSBoUlR8S809RWVHscoVVYkoiZJ3PxqY9vjqFPlNicbi6vjHt65sx9Fk6TS4zx/gKGiGE0GnZ5+LFxSGEkJJSAl6AP55iNQ6xXjY30WDx4YcfmPbWNgqzEzI3CjdWsgzHZULGauKqPzWkUkVd2vlSFVgRoalQeJ5CtQxSCWmJnzs7PTPtX/3iE+iTE4NO161HEzLvvvz1E4j5ZXo03IIe7Y41KaUpiuPp7wROxJ+1sE+bVJNZJfb6igJF/GL97A7xHTu7ujTt5RTnx+YOrvAjV10l65BKLmTO1M5QVpK5l7j3WUbe30lOjHDBztliifNstcBY6apZNWTrs7G5A7HBwBrYWBG72Qzf12en9p6/eInv5rOzc9O+vMSqNMxDOBraub5c4PO8c4jXcnffrhu9ITHnRaJfHIUQQgghRBTaOAohhBBCiCi0cRRCCCGEEFGsVeP48Y//CcQGfZvM8s0334Q+G1uo6amc6o4l/eyQBOCJl00QzVTl9EmrFSYuL5ck5nQaFdEZMa1gsbQahckYkwdXBX5fnjrBBbmW88sbiL04sRqt8RSPvXKH8snGQ8Dk7SGEsDGymrcR0VFAMugQQq/nkqevWMJmsU7+9E//FGIff/xj0x5tocaxqklS3qXVAZGhE1KS/NrnEvf6qRBCqNxcrFcz6JM2eE5p2ycdxuVwOUf90uOvnpr23//9p9BneoVzuHaFBa6vcW4+fnkGsTS157mxgTq24PTUZIqFqiaaMdexJHpuVnygccdKIxO6i++WgujtFiubtHqyQL3d9hTX29tuTHX2UTfnNcghhNC492dJksNXTpfXrIhWMkWtotcmMg3yaomav5VbfwryflkQreB8ZteSyQQTgF9cXEPs7MzqFV++eLXG8eoa14xAEoD3e9ZHMJ/hOTX1fYiNBlYb2WthEvRYNNuFEEIIIUQU2jgKIYQQQogotHEUQgghhBBRaOMohBBCCCGiWKs55ocf/Rhiw4EVfu/soti+1UER59IJzTMibE1IkuHaCX6bCsXExdSKTYsZiu2TCkWrXrjLTAKLORpRbq6sSHY2RbFtF30DYXvDGkpe5Og48Mm+Qwjh5NR+36rG+9TK7T0fkOTtGxtofBmNbMwnag8hhJIkDu97cwwRPYv18s//xb+G2MGtQ9MekGTUfo6FEMLKzVeWSDfxTpgQQuXMG7MJzo1qbmPdDA8+IomRwalGTCDLOY7VT3/xuWl/9qvH0OfiHJP5Ju7wS2IaWGIoNIEl83Z9wPRHChsQZx6uh9ipJE4b/31No98gvg+8eH4CsaVzOp5eYXGF6zGut1lm3827JNl/k786kX+1wjl07s7hfPIS+ixr3J70Bna9acg7tioxVjrz7GyG7+HxGBPy39zY9+clMb1dXeKadOlil8RAc+MKADBzTkNMfb5gR04KalSkYEjb9XvjFpqdYtFsF0IIIYQQUWjjKIQQQgghotDGUQghhBBCRKGNoxBCCCGEiGKt5pi7bz6A2O7Opmm3SLWXQCqwLFdWNNqQ6idE5x3ylt0rNyUaX0pfuaVE4TCrmtKk9twXpOrE/9fe/fy0dQQBAF4MGGOcUgmlJTlFtH9B7733P6566rVqKzWXXqqQnwKSNgRc7GAM2D30tDtz2BOn77u91frJb+33PLJmZ86bavGllPLxw0l1vLlKqvg361RKKdOvn1THz5/HDhZXs7guw2FdQX4y2gtzRqOmyvyo7bIR5/x/7nozzCDZ8JB1mdhtNt/cJh0seFhfPX0WB5tOEZdX8fu1vorf+9ACpsSk+qyzyeKmTvI+efsyzNm8qzezHT2JSd973zyLb2mn6baSPDCWi5hkfnZSd4H4dBE7N0yvkq4XzQaS5HLLOuvA0gxtbCQbEto5WY+qdMNMfa67dXz+rsPOm3iq9Uby3ObB/fb7n2Fs3dxrs0X8fTm4jL+D28P6mfzt0VGYszOO38XwzE92wl2eX1THx28/hDn/TOOzZTypN+i0nY9KyTvVLJf1M2k2j9c7T7rCzGb1Jp7ZPG4s+pxsnr3+XG90uUl2vd01m3hWyf2ZPZPum8/v9DSuXduNrpRSJnv155nGWp384wgAQBeBIwAAXQSOAAB0edAcx19+/SOMPX58UB3vfxlz+SaTR2Fsd1zn1+3sxBy8rHB4KU1BzbuY1zAZ1PlJj8ZJAd6tGHO3qUD/TmNB0dev3oSxv0/fVccHX8Scw3mSj3B8XL9uOo3XMh7HtRvv1XkiW8O4dsNm7baTIqNZrlWbu7ZK8qOynKnhsF7j/aS4OA/rx59+DmN7k0l1vL09DHM2kny3VZOLtEwKvM+ToryfLuoCvO/evIjvaau+p3/4/rsw5/DwMIyNmnypQfJ9znKFrq/rnKbFIuYv3Ze4Lqvm/Pclvm5dYs5Wm4c4SNY3FE8fJPdmcn0hrypNvIxj8VTZs4CHdvI+FqgeNLn3d0mh+1WJrzvZr4tyn3+Mc3afxuf0uMn/30zy3Nui2ccvYu7yX6/OwtioKQC+leQEZ78v903DjqzBxGIZx5bN2O1tnHOb5FS2uYnJWyql1J/LILmHBsl93F7x8iZ+nmfvYwOCl6/r9RyNkqYInfzjCABAF4EjAABdBI4AAHQROAIA0GUjSyQFAICWfxwBAOgicAQAoIvAEQCALgJHAAC6CBwBAOgicAQAoIvAEQCALgJHAAC6CBwBAOgicAQAoIvAEQCALgJHAAC6CBwBAOgicAQAoIvAEQCALgJHAAC6CBwBAOgicAQAoIvAEQCALgJHAAC6CBwBAOgicAQAoIvAEQCALv8B4EgJO2RVDkgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x432 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plots_tfms(img, 2, 3, 12, 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The default value of `scale` may be too high for such small images. Indeed, on some images, we can't see the head of the cat anymore "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "row = 0 \n",
    "col = 1\n",
    "flip = 0\n",
    "scale = 1.05"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = {'row_pct':row, 'col_pct':col, 'is_random':False}\n",
    "tfm = [*augm_tfm, zoom(scale=scale, **d), crop_pad(**d)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plots_tfms(img, 2, 3, 12, 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This looks better ! "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}