pipeline/getting_started_with_pipeline.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "41c3f8f6",
   "metadata": {},
   "source": [
    "# Getting Started with Pipeline"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "db9d09ef",
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "source": [
    "## What is `Pipeline` in Towhee\n",
    "\n",
    "`Pipeline` is a pythonic computation framework for unstructured data in machine learning and data science. It provides a [method-chaining style API](https://towhee.readthedocs.io/en/main/index.html) that makes it easier for a data scientist or researcher to assemble a data processing pipeline.\n",
    "\n",
    "- Pythonic Method-Chaining Style API: Designed to behave as a python list or iterator, `Pipeline` is easy to understand for python users and is compatible with most popular data science toolkits. Function/Operator invocations can be chained one after another, making your code clean and fluent.\n",
    "\n",
    "- Feature-Rich Operator Repository: There are various pre-defined operators On the towhee hub, which cover the most popular deep learning models in computer vision, NLP, and voice processing. Using these operators in the data processing pipeline can significantly accelerate your work."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f7ac0f05",
   "metadata": {},
   "source": [
    "## Preparation \n",
    "\n",
    "Before getting started, we need to install the dependencies and download the test data.\n",
    "\n",
    "### Install dependencies\n",
    "\n",
    "First we need to install dependencies such as towhee, pandas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "97b26871",
   "metadata": {},
   "outputs": [],
   "source": [
    "! python -m pip install -q towhee"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6fc95353",
   "metadata": {},
   "source": [
    "## Create a `Pipeline`"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "8441de76",
   "metadata": {},
   "source": [
    "This is the simplest example of `Pipeline`, it recieves the input and outputs the data immediately. We start with this `Pipeline` to illustrate the structure of a `Pipeline`, a `Pieline` should at least contain a `input` and a `output` node."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "bfa67b0c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from towhee import pipe\n",
    "from towhee.datacollection import DataCollection\n",
    "\n",
    "\n",
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.output('num')\n",
    ")\n",
    "\n",
    "res = p(1)\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "498c90b3",
   "metadata": {},
   "source": [
    "## Nodes"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6652d0a0",
   "metadata": {},
   "source": [
    "### Map\n",
    "\n",
    "Usage:\n",
    "\t\n",
    "\tmap(input_schema, output_schema, fn, config=None)\n",
    "\n",
    "Map node get the specified data by input_schema from each element as input, apply it to a function, then specify the schema of output as output_schema.\n",
    "\n",
    "<img src=\"map_node.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "85ef02c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.map('num', 'num', lambda x: x * 2)\n",
    "\t\t.output('num')\n",
    ")\n",
    "\n",
    "res = p(1)\n",
    "DataCollection(res).show()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c7ce9e50",
   "metadata": {},
   "source": [
    "### Flat Map\n",
    "\n",
    "Usage:\n",
    "\n",
    "\tflat_map(input_schema, output_schema, fn, config=None)\n",
    "\n",
    "Flat Map get the specified data by input_schema from each element as input, apply it to a function, then flatten the results, and specify the schema of output as output_schema.\n",
    "\n",
    "<img src=\"flat_map_node.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e2c23d89",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">4</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">6</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.flat_map('num', 'num', lambda x: [i * 2 for i in x])\n",
    "\t\t.output('num')\n",
    ")\n",
    "\n",
    "res = p([1, 2, 3])\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "fbba34b9",
   "metadata": {},
   "source": [
    "### Filter\n",
    "\n",
    "Usage:\n",
    "\n",
    "\tfilter(input_schema, output_schema, filter_columns, fn, config=None)\n",
    "\n",
    "Filter only select elements that satisfy the predicate, it will check the function with filter_columns as input. And the input_schema and out_schema are the data before and after the filter node, so their lengths are required to be equal.\n",
    "\n",
    "<img src=\"filter_node.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "dbc5a04b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">4</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">6</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.flat_map('num', 'num', lambda x: [i * 2 for i in x])\n",
    "\t\t.filter('num', 'num', 'num', lambda x: x > 2)\n",
    "\t\t.output('num')\n",
    ")\n",
    "\n",
    "res = p([1, 2, 3])\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b9249e20",
   "metadata": {},
   "source": [
    "### Window\n",
    "\n",
    "Usage:\n",
    "\n",
    "\twindow(input_schema, output_schema, size, step, fn, config=None)\n",
    "\n",
    "Assign all the elements into windows of finite size, over which to apply computations(fn). The size of each window is configured by the size parameter, the step parameter controls how frequently a window is started. \n",
    "\n",
    "<img src=\"window_node.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "83d42f4b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th> <th style=\"text-align: center; font-size: 130%; border: none;\">nums</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4,8] len=2</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">4</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[8,12] len=2</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">6</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[12,16] len=2</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">8</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[16] len=1</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.flat_map('num', 'num', lambda x: [i * 2 for i in x])\n",
    "\t\t.window('num', 'nums', 2, 1, lambda x: [i * 2 for i in x])\n",
    "\t\t.output('num', 'nums')\n",
    ")\n",
    "\n",
    "res = p([1, 2, 3, 4])\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "13f0d796",
   "metadata": {},
   "source": [
    "### Time Window\n",
    "\n",
    "Usage:\n",
    "\n",
    "\ttime_window(input_schema, output_schema, timestamp_col, size, step, fn, config=None)\n",
    "\n",
    "Assign all the elements into time windows of finite size based on the timestamp_col column, over which we can apply computations(fn). The time interval of each time window is configured by the size parameter, and the step parameter controls the time frequency at which a time window starts. \n",
    "\n",
    "<img src=\"time_window_node.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th> <th style=\"text-align: center; font-size: 130%; border: none;\">timestamp</th> <th style=\"text-align: center; font-size: 130%; border: none;\">nums</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">0</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[1,2] len=2</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1000</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[2,3] len=2</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">3</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2000</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[3] len=1</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">4</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">4000</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4] len=1</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">5</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">5000</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4,5] len=2</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('num', 'timestamp')\n",
    "\t\t.flat_map(('num', 'timestamp'), ('num', 'timestamp'), lambda x,y: [(i, j) for i, j in zip(x, y)])\n",
    "\t\t.time_window('num', 'nums', 'timestamp', 2, 1, lambda x: x)\n",
    "\t\t.output('num', 'timestamp', 'nums')\n",
    ")\n",
    "\n",
    "res = p([1, 2, 3, 4, 5], [0, 1000, 2000, 4000, 5000])\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "94871afd",
   "metadata": {},
   "source": [
    "### Window All\n",
    "\n",
    "Usage:\n",
    "\n",
    "\twindow_all(input_schema, output_schema, fn, config=None)\n",
    "\n",
    "Assign all the elements into one window, over which we can apply computations(fn).\n",
    "\n",
    "<img src=\"window_all.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "220d16b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th> <th style=\"text-align: center; font-size: 130%; border: none;\">nums</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4,8,12,16] len=4</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">4</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4,8,12,16] len=4</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">6</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4,8,12,16] len=4</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">8</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[4,8,12,16] len=4</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.flat_map('num', 'num', lambda x: [i * 2 for i in x])\n",
    "\t\t.window_all('num', 'nums', lambda x: [i * 2 for i in x])\n",
    "\t\t.output('num', 'nums')\n",
    ")\n",
    "\n",
    "res = p([1, 2, 3, 4])\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b1bd3245",
   "metadata": {},
   "source": [
    "### Concat\n",
    "\n",
    "Usage:\n",
    "\n",
    "\tconcat(*pipeline)\n",
    "\n",
    "Concat one or more pipelines to the existing pipeline and update all data from each pipeline.\n",
    "\n",
    "<img src=\"concat_node.png\" width=\"600px\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "af875770",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">a</th> <th style=\"text-align: center; font-size: 130%; border: none;\">b</th> <th style=\"text-align: center; font-size: 130%; border: none;\">c</th> <th style=\"text-align: center; font-size: 130%; border: none;\">d</th> <th style=\"text-align: center; font-size: 130%; border: none;\">e</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">3</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">-1</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from towhee import pipe\n",
    "pipe0 = (\n",
    "\tpipe.input('a', 'b', 'c')\n",
    ")\n",
    "pipe1 = (\n",
    "\tpipe0.map('a', 'd', lambda x: x+1)\n",
    ")\n",
    "\n",
    "pipe2 = (\n",
    "\tpipe0.map(('b', 'c'), 'e', lambda x, y: x - y)\n",
    ")\n",
    "\n",
    "pipe3 = pipe2.concat(pipe1).output('a', 'b', 'c', 'd', 'e')\n",
    "\n",
    "res = pipe3(1, 2, 3)\n",
    "DataCollection(res).show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c3bf17bd",
   "metadata": {},
   "source": [
    "## Towhee Operator\n",
    "\n",
    "Operators are the basic units of computation that can be applied to the elements within a DataCollection. There are many predefined Operators on the [Towhee hub](https://towhee.io/operators?limit=30&page=1), including popular deep learning models ranging from computer vision to natural language processing.\n",
    "\n",
    "We can load an Operator from the Towhee hub as following:\n",
    "\n",
    "\tops.namespace.operator_name(**kws, *args)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "8e38fcd1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from towhee import ops\n",
    "op = ops.image_decode.cv2()\n",
    "img = op('1.jpg')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "121a949c",
   "metadata": {},
   "source": [
    "### Custom Operator\n",
    "\n",
    "It is also easy to define custom operators with standard Python functions with `towhee.register`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "fab1f328",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from towhee import register\n",
    "\n",
    "@register\n",
    "def add_1(x):\n",
    "    return x+1\n",
    "\n",
    "ops.add_1()(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e699c40",
   "metadata": {},
   "source": [
    "If the operator needs additional initializations arguments, it can be defined as a class:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4e9c4179",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "@register\n",
    "class add_x:\n",
    "    def __init__(self, x):\n",
    "        self._x = x\n",
    "    def __call__(self, y):\n",
    "        return self._x + y\n",
    "\n",
    "ops.add_x(x=1)(2)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2ffc3114",
   "metadata": {},
   "source": [
    "### Run Operator with `Pipeline`\n",
    "\n",
    "When an operator is uploaded to the Towhee hub or registered with @register, we can call the operato directly on a Pipeline:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "268722a5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">url</th> <th style=\"text-align: center; font-size: 130%; border: none;\">image</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1.jpg</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \"><img src=\"data:image/png;base64,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\" 128 = \"128px\" 128 = \"128px\" style = \"float:left; padding:2px\"></td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">num</th> <th style=\"text-align: center; font-size: 130%; border: none;\">processed_num</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">3</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (\n",
    "\tpipe.input('url')\n",
    "\t\t.map('url', 'image', ops.image_decode.cv2('rgb'))\n",
    "\t\t.output('url', 'image')\n",
    ")\n",
    "DataCollection(p('1.jpg')).show()\n",
    "\n",
    "p = (\n",
    "\tpipe.input('num')\n",
    "\t\t.map('num', 'processed_num', ops.add_x(x=2))\n",
    "\t\t.output('num', 'processed_num')\n",
    ")\n",
    "DataCollection(p(1)).show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1c86d9ef",
   "metadata": {},
   "source": [
    "## Complete Example"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c36f9e06",
   "metadata": {},
   "source": [
    "Here we show a complete example of towhee `Pipeline` that implements a image embedding task"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "2b707c3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">urls</th> <th style=\"text-align: center; font-size: 130%; border: none;\">images</th> <th style=\"text-align: center; font-size: 130%; border: none;\">embeddings</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">1.jpg</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \"><img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAFwAAABcCAYAAADj79JYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAABcqUlEQVR4nEz9d5htyVXfD3+qdt4nn9Onc/fN986dHDXSzCggJCEhiSCSyckYC+OEjcHGJhhj+GHzyhhsww8bkUSQiCIKZY00I400eebm0Ln7dJ98dg5V7x8tv897nuc8e/9Xu1atWrV3re/6lNg6CnQuDN700CnsepdsHHF61eGVgwHnVZttr6BbdwkOx8yfWed1D5xntLfH04MriP0qu3HKfEOws3fEueY6sumyuNqm7jWZqzUZFGOuvHCJ6VFAa64KpYPvmhRexPCoxKvnVJo2ic4o922Crk01TckLg2/6mn/DJA3wZEKETVNXUL7JUXCEdzCgN7yN4btEmYHq9zFrDlcHO5xYOkPTreO6HidWzlDYDnluk5opprT4wlNfxMlC7IpBbzdlrtPk8rWnsN05tEx547u+HlcLptMhFa/O9Zc/SZRFGHbA/+fn/4jv+Gffw1d/1T9ApoJXX/04165v8NjXvJsgAt/2aC8s8rfv/w+cfHuFWx8LqLo5O9OCx++7B1MIC9uQJEVG1y3YSyYcDOdY8Fv0CFnqLsNhgDAs5uU8+TDkVryNNh2KbMpXfWOd7qxJ2l1n+5Lmju5FPvLi51iub5LYFo5jMZUJy/edYra9y1zTZiwTrLJCYQ4pSk2yZ9GvW5w7tch0ex8pBA++7p0sPXKRo2euQaWLoGQ6KcjHI3SSk1c7dOYWqTmSQEnUKclwNqFbOcurt68yHjzPqm7DVy/Satio7JC6W8Wo+JxZOc/u3hdIm+u4w6u8dCXCTsa0Vu6gGPdwbAMjyilVlX4QURgFeSMhnuZc33iVN77xjUzKKeF4wtZBH6KMv/rfv8/Xftu3MowCOkFAFhlkGxojloSxZq3hE5spZmaWyLJksD+iWhZUGzWknnGwP0JqzfzcAlvZEXqS8fzu85Siyc5WQXVZMTALBpcMirrPYKtg/+AI4h2MJKVXM1iquBylS9TDEYiMIhckUmK1HLqFRqgq6WGKsgrWcoOo34cyJBcOb//6d/GNr72Tv6nX+bq7zlAaJUqBlAZaZQjDRAMSDUpSihKlJUprJkqjBPTjElNKBllOL0oY9EvG4Qw9+yzpFqh4wMbWderLj1IMlqjqChv72zTiklzCnNphmCmm+z101UBmFn/+gd/lNV/9Zl5+5QbR0YDOwir9VFNJ9/j4X/4+ad3kSqJQccbu0ynBJKTTbaN0iRgoxOXbQ12gyKIc03ewhEYYNo4sEJaJoSWllGilKYoIR1pILRG+gVUamIYmBQxdYiBBGighQCsQBpYsMFSJsGzKUmCIkkzl2Iak1BpTGkhtgtRkgA0UKFRpYoucIItpWA7KkAgpOegdUKs2qFSqKKmwtAQNSihAIgUIDVppNAphGggNShcYWJQq5et+4F+QjDSNTp1gmtGuNXnhlaeoScGrN64QHt6iMCRlWWIBz22N+PxTTzOLJhS5wvNdpHIIyEiUgzOMubXxPN3FNh/44B8gtaYoFbZU+IaDSiMS06TlWIhXtidaCzBTC23llKaBVCWmNDAMjUAihSBTCmEIjEKjMZEyA2UgTIGWEqk0AFqAqSWlUWKgEcrAEJrCsMiyhLIAwxZ4WlJIA0NCrgsc6ZDpHBOD/MuG1ECRltiWAgxKNEiBKUy0wZfb1AgUWQlSCITWIA0sDampMUuBkgopDJIy57/+7PsZJSNuXH+Wo/0x5XTEa776a/Fa67zwyT+l2Vjmt/7wv1ITHoVOMKSLKhUoUFIT5xrXBIXGMEAbFnahUapECINSKAwpkQo0mhKB+LITGApMB4kSoGoSq7SRAkopKYVAorFKAy0UhrQwSlCmxtQKQ3oIQ1FKhYVBaSlkKSikQgLyywYypKAwJIZSONIA18BEog2B0AVaCAxlkVNgCQno46sqKYsS27EQwkCXClMcD4I2NaYWSAO0Pm5LCIDjgVFaU6CPn1eCKQykEMzXTF7cuMmibbM3K1iQOdbaOa5f3uDCgx4nFu5nFpWcqBn0Q5DCRQiBaZlorZGioGmaaAFaC0ChFGhDIo3jZxdliS41liOpSpNMCYZKY+YZkWsjcczjzmgNpsDQYAuBIyRaCHIDtC0wpDqewkpjGiZCKLQAQwuUUEgNpdBYWiARaDRSSoQhsUpAgBYCrUuULkCVWEikBmkKHGliGvK4YxqEkLi2g8pLtCpREoQEw5DY5ZdjNyDEccelIRCGREow/v/+tjSQQoOAu9ZOceH0EgeEOFXY6W9ydetZdjc+wtUvfgGjXTCWE37wO/45hgStFEJotC6RKCxMjuOXwpQaQ0gMAQb6y4MNYCKE4K/+8kO87Vu/j6//+m/gG975bj7ywd9ElgVmnke4dg1lCGRWohyBLDUlJbZpoDVYCLRpQKERhkab+jhEa3G8iOnjBqXSKFMiAaElkmNPNaQBqgTDoCgLlJCYGrTQCOP/Do8+jkdCUKIRQGGANg0MQyCERgqJ0AIMELpEcOwUBlBqzf/9CQ0SiZagdQlK0PAFj7/tXUz2RxSbV/ByReDP4aVgNU+STo842ih44IHXcm2jjyo00jjuP+LYzkIrJAIhj9vIhEYIQZwVlGiicMgv/uhPcWV7g8TtYJklOxt9TrUr/Kf3/U8+/IlnETcHUz0OIjynjmNI4jLH0hppSQxMoiCgUnXJSomNJDNyPCy0LimFiaCAErQt8YRLkeVgFCAthBBIVf7/YruUkkwIrOLYawC0ZR0bSJVow0QCSgp0mhJkEaqAmuNSOBJbQVGWSAMsw0ZpMKQAVZIrjSElwgCtBHAc06WEQpX8y/f+GDf7Je3FBuNL17l+9AJBfETVXaJiuxjzTda8Ok73FPfOr3L+7jm+4Vu/EzSkKsc3TYTSFEJjoikFyFIQqpJGTfJf/+2P8PmXh4g852h/F2suw3Vctg4UcTwjiI7wAwPxYz/+bfqJN76L3/4f76fQCTXXIS9n2Lak05yjUqnjeR6iqjE8kK7ASpqU1oQ8M5ERyHBCrbuO3RBoW5MdGdQbBrXaHJavMVUVo9IhTEZ4HiBM/MoclqxiGIIiTvGbTcpshpIezSJBuR6EMUnzLNValbgsMQ2BZUgMKUiTHM/zwDAxpUB9edEuhUBpfbygSnDIUMLj4p0P8fhXfytmNuDg9gYYDkE4o7e7QWdtia3dXRaFIvUMlqtn6Z5Y5/zrHqMTXeP6dsgv/PJ/RFIitYFWxy6eFRmf/vgf83u/9zcMBkNs4Nat29yzusS6ZTPKCu5a9ljorlJrFdg0MB98ww9iNOqsWnM8H21QL+qklFSHLvu5YpEA0+ny3OCLHBwNWFg06O9b3LX4MIeTAUm+gRtbbH38VRYWT+IiCI0ZD60sExS3eL68hnerTrfV5drWZfzVjESbFDsTKq02junQbayj24LcOmRlchcnXn6Jt28F1NIA/fxTHGUpslAIaWBrSSw0VpERZZpMZdiOgZQWhgRTmljG8ZuM69g8/+nnWDnhkkUTitEh2nMx59fxjQrB6JBlr4pZTLjr/HkOD6+ybK0SNQre+477+fD1Hi9efZl/ntXw73gDP/Xd7+DCyRMEuqSepvzF334Co9Jga7NHIeFg2mfRU7x+rkuj4xFLA1dGyPyQnX0T2+xjnjt5N5gFm9OAk+fO0pEmtjOhGRgsz9d4cesQ86VXaK9ViZtTtsch5TjksDlFjwcstB9lPLpC9aRH+OoVpi2f+WqDp17YZM2XLCw30bbHJDrgsTMX+MPnPk3NsXjsvgc58vbYSXN2X7rNQq1KMpwxblzi3s0e21YHGR1xIk+pWnNc+eOfRvRGSK+CKQvmv/e9GMyx9+IXaC3NI+wGpuUhHQvTsJGUGDMXq7fDlRsz8jxBJCWGzBn1e8ytnGC50UJNJbm02E33aTWXOEg2OZmf5zv+yY+x9vBpDrZG/GtiPvTE21moe/h6wif/+gukbp0ytZiODrCLCaqAr1s+w/0n2nh+RDCOKZGYfsFev0dkt1mUJaZZNaAU2CfaBNuHqDmHpK9YvLjK6KDPYrXJURGxPT3AcT1OxCfQTkBLF7ja4jIvEKsh9524i6o9R9o2ydf2WNlpsf/cIZWyw7ve8SC/+zefZGjP+MY3PsLGYMxgcYzb7fDogSR7tMHBZsT8nENn1cWTE6r/6b/TbC4zvvkcs0ufJxvO0FLip0Miq8H2r78Pu1bh9omH2duHrcF1JnsHjDauEgRbHBwl7I4O+MPvfg/rpx9hfu4E2oYpgrZrMckCql6FqPQRsUNdlARijBF4tM/ey31xj5vXN6hU63zp/u/mlytjXvnsF/hYmlFaNkX/JtPtGaUXcWdznifOdfAbcyTDAeNgxCiImWHS3wtZXVumlUUkpcCU0gBKWqJLfi4nHI3ozlXoD7axS4OKKLleG+AJSdNq4VHS7CzTc2ZQerQNjVr0Ca7FTBoJ7Xyd9GYFr1mlcr/Aq8BfXHmGxrKPuy655k0pRgV6W+ComGjoc7JxAmspx4im1CstjGDG8sd/h2uux2g2ws8F2hKIdMaktPBlQOJJxnt95kafgEab7/v+f4a257l85TpFMuTinQ9g7GzyyS99ibO9S/w/X/cYH/jiS8QLp7HygJVTZ3jnW97Mk89vcONLX6CZNTBTyU/9zD+hefYOfvo/fJ7wpRAHn/dWdvnEX3+Ehy7eQ0hJMOhTJhH3dBo8cuI0ShpkScIo2KRiCvLS5eLbvpeP/eYvEBkGyfZNFtsdssBCXNoaa2XC+3/hv/Ny70NUqnVqZQcjntE3c5Y7Hv1MsqAVN+MBS605ekZGMypJZYXDyZDVlTb0PazFnLmKTd8KGd+Aheoc0WgfvDoXlu8gGx+xMd3m+f09WpWc/Rs5Z5a6nDt9niyIGcYxBAd8/ff/J1rP/iVZnJCWPqgRR0nMsvCIrQq6CInShEprnrrpUjhLlE7B9PbLtM6c4Pw7v5Fzd1/kxb/4BEHicOn2Z7A2D/jw3z/FL//dh4gjmF9soVWBbfn8l1/9M77wmT/GXzrD8tEel0ZDxtER80oRTVP06VUetes8GeXc3ZXcVavg6CnSaRLnDpbRpxx77KUD7n7rt/L4e76Gem2JH/+qs1S9JQ6iXd7z5ncThhESU6AUvO5NX40uq4y2coSooFsN2rUKQVRFWgZ7ZoZpm6RWynnLYRCFFOOSlUaHYJYi/YxFY56NqyFi5tCxbLY3tslxsUrN85svcFgmLMyt8h0rj7PUeIS3PfFa7rh4Dl21kF7B/Jxg6ezdzL/0aZLYQCmJCgOyOEYmNmEco6OYqChwdEZdBGRWQSy2KPOQuRNnGfUzLr//dzGUyUu3DPae+iAtY55eGOB2a7zxK7+Wat0gz3MQoMj55z/0bvqXXiW+9BT3P/wa/vcf/R1naz6duQUunD3H8GCXM2fu4S1LVd7cbtFsKTJlM5wNMcuC6VCwNx7y+Hf9R9oPPE6YFkRHV+nUG2TZjHe++S3EdoQnPExTmmidcc/r7qT9EYf1+SWujgvOWg435QQdSIwsprRCVOIQuAkSG7PMqdcLJlbMeqNDGisCGdI9UScbx6Rhwd13nOcgnVErCoqKw6ASMxdbTBdKzmsHRAM/tqm6DmZ9iaPZgDdXFgjjKWkxxlAS083RI4ErcxzDYGoIZBGTSgNZWyWIB3RwKWWCFjZLtQpHkyHTMOe7f/gr+MCvuOx+9Oe5cOcdGG4b034RPe2T+12E7yFliWnZJEXINz7wDubrMZPBAZFc57w5IbrzIm9bP81Hn/0Uj52cY6AEuq/wFYSFZjDeYS/NiaKEx157N4f5lN//tZ+BvQBlSO69734GhwP8xOQztz+KqbUGaSI9j/LAQF2w6BCRV+rkuy3mlmF0a8p996xz5TDmhGyx48Y80D3PrcNDqkKwP5qyPF+HaEKUFtSbFgeppty4STE3jy8yHGFTDjZ5Ml7kRE0xN7fIxqd79MMDDN3gvoWIb3jsCfJ0zEQHFEIgiowyKxnJERY+tm0SihwdJCysrVGUI+pCk8YCaopqu0sUFyxGEX/5k+8lrlT5jp/+H1T/2RtQwiCXBT/6pm/gW7/5+/ntP/sLKCMsp4mDxff+5C/xwp/8Nv/mm9/I5z/0IT78qT/lP3zDW9l+8Wnufc/3cVpEXNUed2e7VHDYTDPS2ZjQksRpyLTUmLUKJ+oNws9+jiCJ6aeSeu0me1M4s1ij9ujbkcIwcKVJW5aIjmZ7sk1pFQyCiDOtnHpgU1/Q9GcRNWGSeQW+Krk+DQgTC9vxMH2H3f19dmb7WB7MEkErzKiu1lmspshSEeUlTXOZxzorNOotrn8k5uboEgYmLTfh+971TajRIYmWtE0P0hylIY+nuMpBpYq9oynxYESr2SSdjun3dyGZYJmaIusznYzQ8QGBGGGXHt6h4Fe/8x08+em/Q6MwtOTined43V33sfnKk0SzlDyLUEbJZz/zGZadnK2rfeYeuxctSo7OX2DdqlHvzmO50Ex2yavr7FkZmaUZWS6kFjVRZ6Fw0FLxr9/9NjZL6BtznLrrQab1eU7OuZQnHqW9dwtpcbztmBoGpvRIIg9QlN6MaSHwHZuWc45w7OI6JiUuwVTjhxK/1DStBi3Lo91Yp9k8g0o86qLGnafPM6eaGFGd+soiTuGA1eCoMuXKJ2ZsHWxSKVoEWcbHPvtppr1rCFehJlMmkzFRGjEaDUgKm7iwyHNBnGXUIpckKUlnM6LBkETalNUqs7xK241R0wl4dTRTLCuh3Z7jyh/+FlJIpNTMz83hzdX5nz///5IUioP9LR7/irfxa//tZ+gFAbNgyPMf/G1kmVE++SXsusOtT34EU9icaNT545dfoakETaeCnzgou4JnlLzn3/4iOC5M+9SESWAE5FHIV33TtzBYvpvWxgucqVWQJQoTi1oBuc7pdCw6dRMdGGTeEMevcLJzirlTVRJDI0qB2zJYOn2WmlVFeQFDESC9goYMUZUZQSdnS13C8TVhY0IqJzirBvFCwEt/F1CzfeIgJoymeKdtnvq1X6CCZICHU9oUeU6DktzQxL5BrscciRjDbxI4CXW/wly7hVv18bSHKjSLds44CElUQjAdI6UNRYnnWKRVhw/97m9BWRAbJucX5njgocf5l69/HV/1ju/lh/75z2FbFv/oDz/OdLpNY6FBKlxYfRgrVeiDDYpZCtR4e7vOXx9OqOoORqMkkhley+Mt3/gWXAfs0kToKo8sPsxONOATH3uW5ZtXmOQjgrKHlNJAyZyJBWreITNy8tKgfWaBerNL5h2wa98mLmcsrjTRiyHCHDNLd6mu1cgjSa3ZAb9O4rWQTg1fuLS6ZymrCa4jEFmMJQzcm/PY5NzYuI2cM8lEwIOnVxkNtpmRY+c5kewT64KsyHBKjcxipqFkLjfJrZihsNgZbZKVJnOWRjgS180ITUUUarRpIcqIsFBIsySeTbCDkptPPkWK4NSbHsWSVeYade69+yE++vG/5s1P3IEuYN7LmGzs8blXLnMUlig/Rno17ut0EHVJzS6RFYfq9iGDqsQMXJYrHo6wiZOM8azEtzOSdMb1ycvsXtpiee+LiDxFOha57CDll7c1FdDRNo1OFbtTper55L6gL0sKA6qWQzgbMstKZLtNxXTpdpcQVcm8CSKfUdU2zayAWYiFIihzoEKn0aDtWnzx6Zvs7u2QlYLB/pCD6zd595kTWMpHlSVhOWQ8K5CTCWEWk+eCvEyY1za2lyFLSTvJ8QOJLsZMRQvHb2DOEryggickbpohvC4tV5FjYwsDw5Dc9aa3oXLB7rPPoYwq8/MtLp67QMM/TsuVosC2XRZdl8qpJ/ARuL3bmA+9C2lkyGBCapScueMsZ9oug1u3yU4sM8pSSqskThT//Zsfp7QbxLpkuDsgkTnT/oiROh70MldIDDCAtFR4ZoNKBod729w63OBsOU+18AhmIfEMAjnBNFLaNIl9zTjaYcGpEqQedVVHdlxkzWOh0iIuSqjWOVU/i2c8zuhzA27HEe3lGmF6QLh3iyCXtIPjhMY4jqlmDmE2IU5jEG1wTHzRIG4kzERCOo0YRzGW2iMJIhaXFohLSWr42CZo4YEdYUc9RlmMMCYkroFpJ5x/6HHiIieaZUyzBMMW1BeW+bVf/U0a7SqzIODGtVf40q1tarUSy5Z05yyuvvyniKpClSXdrESYCVFZMN9yGWzdJM5NorzNxsFtZmnBfjgh1WCKIe/5ofcQMSMKHdK0wNMeUusS0zIxtaJajYl9RaNbY6HqMNQT3JZCyhy/VnC6c5quVUHaKe2Kj4HBUSmwspCqV0CSUDfqeLUac5ZgTkhUPSATX+Lpl3YwD2/T2zsgSxXCbdA4vcxEQVTMsJUgm87wMslUQ5GNUbOYfDqhMhbYSRXfKLDQjEoPWW2Rjwt838CU4HYriOkRaeSQhTllAeVUQSaIk4LNmy9hGB7T6QSPKf54RiAVrdOP8p53fSNFnrO3dZtdXeGDf/B3RKnGpc4jnksYaKKiwbaKGR8MWT+/wujWLc50WgyGIxIhefEDv4xnWpRZQrWIiJgn7LZ4cRfyZI9YxtwOe0jLsLGlIk1LDrZDapxADXxmY5MyNEmHGWWeEKuC/VlBnFVQWQTDhMKtQ6mZ63TpT0qy2YygVPT0gFJqzNoceRETZT6iPYdwXPS4wBAmdtXgb37iH+LaGUk4JAlGbBYDIgqaSpOVkgBF4Somhk0SR4SZiYinFARk2YzISpgMYmYiZzCZUNqKMA+JkwKimJmWoFKUgOef+jhZkZOGAfHsiM1hn2A24OOf/Ft6+4c8+ZEPsrd7yNL8ef7myY/xyY9+jN3SoW5kWLak5kOJib84jxRtFk+uU+oxp5bb9I4us7eTMjNqJCJllma0Xv86rPGYtcffTFL47EaaptfE3NvfpT7fQeQW/+YnPoDAwDJN8jJDaIXWglyU1AyTUpkoeZxNUUJhCYnQOY5tAYpFz+bWXozfEvh5QlGv4tuSH/zH7yXYH+NLn5ExYCI19cQiHeXEk33yw5RESbzQQ8opIz1FFzX81MGQDsaCQbkpyGXAYaxYma9geQ2M1AIvJTRMZJISJDkOHpkOkDgYdoI30xQSepsfo2b/AoYI8ZTNdDThRvsi/d4fMTVCfvKXf4P1OOU3ntylVq3wa//t33FX4pI8dALHVjh1jTE2mG33iJII6SiM0ZiObEFnmfDwBQZejSiMKYTJSumw2DE4s2zy558SdNyAJ37szzCb9XkKZZDlEaYhkKZGao2QGhNJqaFqekihiOISKQR5keNbHqUq0NJAF8e5vhujCHfOJc0hF6AjRRoaXHl5E73okV4fINwqK57k3scfpz++RTzJKJOYTGiUAXkYYusa2ipIsNE6It1NyXSM1IKqiEkSF206JJaiLDTWcEySGlhEhNlxRlQZAqlMxnUDIylYqi8QRwWv+5Ff5lM//69xTFhYWyD9lImKm1SKkn/3/g/RaFp84m//D3fYOcIGtwlObKCiAkWCpy2mVkGtXiMoL5KVOW6WsT+1cQ92uKoq/J/f+kUefPxriWcJtm/zPd9l8QPvfhtGcoQUtkLkMbZjYzsuUgCmwDEtFBLTsLCM4zyjaSosQ1BxPExbYVoS25Q4lolhSjzHQRbHGW0tBTYGs8EIRzokW7uUtseCv8jy6btZXJpnetRHSgftOxRFiSpjVFkD38SRNm4RksXHGXclPHK7yqTuI8mIMTFtgUwHjEYjCivClAZhERG5OSYpAgs7UFgyx7ZtPvgbP8fZZZNJ+16+uHCGP//jP0VmCulNeNvXvIdOt06SpvzVr/0WhVNhtVUn6GlcHLQjyDJJLjW+cIm2d8nKKW3HBSUgzzC0x3mxyNkHv453vPXHef///gUy4XA0M3jk7f8IW0yQKjcQto0jDEpVYJoWNgYyy8l1hmEeazykEFjSwBQCSYktDMqiRAhNWeRQlERpfKwHEYJSHMsWPvyh93MU7tCZO0UabTFLpjzy6AM88NpHMRyDcjQmj0dgSHRs4YmMqtkkV4qgVSdKAkwnpeM52KXGzLbQdpOmdAh3Nrh84wZpEjIJIoZ5hOXb2EoR5RnCANc0SaWFsiO2PvEsk8Tmp3/2H3N+4RS35RGGWSAHM177hkepGy7f96Y7ObVa5WJ9Hl3klCJiOJ7SyZbpdpskVQNDFoTVOVqGz97elGwWkuSSgVvyy0//LUUJf/zn/4Uf+OH/wI/+8K/x/W97nMXGNv/nd/4XpmFLdKnIxbHOwlQCLUtKx6aigVIjJZS6pAAsXSKlSa4VlmlQFCW2kBiGgStcrC/fe/pYsvZ//vTPCIMxs+CQOG8iBvuUhsXpu87R/6xmlI8xhCArBaks0K6Dk/TJE4toMqW0bYgVqZVBaJGHK5hzMeF0QJzMWLBKyoqmZYPOq0T5DOl00ZaBYSpGxQg3rBE7Ar9W8ovf9haq5x/jxuwQIwwZFxOa7Qa+zvjH33YnS3WH+Uxiz9uEBw5xf4jVbLJb7CFiC7uQlEabRTdhb+MSteYC6CrW9AA7uYuqX0dq0IamLFx+7pf/Cbbzg7zlrru4Z+0hzDTOsWyNlC62lCSFQBYZoNCGxikyCtumojxyNUULSWaBLHIwJDYOhZGjRElZKrSWuEV+LJ3QNkoY2E4Xx7UwwoA8VqxfvJsH7ryLv8wzHF0nSQfEtoVn5IhckBY2GDG+neNql8wCMzcJRQiOom7bJI5Bsj+hqbpQWBRJifIEDeESGxZeGZCNjhCGJHQtRJbT9Sy0W+Go0eb2Zz9HPVfoUDGnNb/zX/49lVxS8Uw20m0WpgsIHVC4JUGUU7FzbLtKjsTzIxLlML94lv54jPYMygw++MUPkilNgcDMJRqJKRRFVPKpFy7xmSe/hPnLP/zNDIND/tGvf5jf+v1vIYsTupU1QiNj2TAIDwu8Rg2chGIacHJhhdJxaVMna0qcskatYmIbPkkOfqtGHZd6Z50sGOMFMTPXwpFNymBCabU5ZZY4MiNJJCoJsC0T3wQjqyMoiYuUvdymKXySakxNVZk4CdbeGKPt4TbmYSdglgcYtRUmoz2mxhyrXkmWmxiOw8gBI3ORwqIlBYHVIC13aLl1rjgFRXqAGSeMZkOi5hwntMF+w8ZkyqKuUvEVtwOPVpoglM+0d0R3yafa9PByh5mRUG1VGKR9qtrE8iAVIKXGkQJXmsRSYRWazDBRObzusfsQf/PRL+gk1SyfXeSTf/mfKUWDXtKjTCKqrQWsEoroiKrfwC3qPDV7nnm3TqWzwILyyGUFZzbDNDRTx2JgHSADi3bT5dKfHvGxrVdxc0mYpjSBlJInHjnHa7/+IVoffxqtDUapRU6MrTRF7qNsmIUpZBLDL6gqC4OU6/u36XbXqFdd0vEEQ0sm2RG+0wHTw7FdJCY0SrwiI6m2cWcZmWFRw6Rvhnz02QGcv5PLn/0Eyne40/Ypk4DltoclZ+SpxXLDYm1xnbgsSC0bK01wpcKdX6FhV4jjkjhJECLlSOdEpcnJhy7SnFtm+fQqs+GEhfVljGNpKkLalKbAyquYq+snyKVJXiSEFY2WExI3ZyGuUniCZgVU0GFa0UjT4e7sDmQesZJV2LKBfIo1bzAuYmqlSVBW6TQtOo7Jp7IRvjAZDXo0W036ZcGaVeeoH3D7szt0pUFogJOXMLHBA60laVZSyoyMnEriMrA1DcMgTAUtIZlEJa5b5/b2dbZHgrm5IXWzTq2ZYlddRFxFaoc87eOziEvGVIOaSXaLFupLX+T7/+V/5/f/24+xYCrGhqYlMwrpkMocrerkjJHKRSQJloJANjDKgCzyMZ0AZbRQuYsX9BCmy3/+lT/n5IUz1Je6BL1dvvnd54g9FzvWDKMBcenhDDSmrFVxtaCMUqqGS9Y0WB2EFLrAJGMYhWS5h5/WcB3YpkpVmvR0TJHl1BurZGVJbIKOj2gEJpOuSSX1mBSahmjhnl1mGI8oe1vsGwXnig5WyyHZszF0yb62aeV9BpWMOK1TEzUKcmqGTcV1GKkcp1rF0/MIU7B+agGz3iDO9qh5iktDl2k+pevWqCpBt94ijcGzHeJ0RFFvUM628bXN5vZL/Off+D1+6l+8l8dPnWLW32PNrDPrOjSmCr8codwB0/Q0HddCiBbB0S5GdY/BYYXQDfECFy0S0tIiL0pcx0C6HkfDKXOqzvq9HW5lm5SZy1xRZ2s6pNudZzhKkZYoMNE4jsNYCSpTQaxqWO0mVlnDUB6i4uMIA4yShXTMom0Qu8d6UsNIMLMRLbek6SzRcwrMJGaYTOgqh4PkgNl0mzVpsmh5VLCQZ+vc8Zq7ETIhm0a0ZUIgSuozG0cFlE6CITRGpU4sYcmRuEHJSMecX13CDAr8QDCna2SGxx3dnLXTTapxnTSaMSsPOMgP6U8FM8MgCkIct8OfbhS85a3/gJ/+rm/lTfc+RhSPMSwDS2tOxh7KycninN62j5cEaLeJjHvQMZkMTQ429+ntDIjjkCBOKewcT9eYyIKHHrtAEBxwaXAN34zwg3kqYR2DDitOB1Hz8esepi4NtKkp85LwqMd+f8bKygINx0BXNWEBUdIjjFIwV6hKm/Eoot1wSWdHpHmER4OoSHAMlzvXTyJzlxs399jq79GqLBAbCbe3tmjWaghTk8RjzNmAvAC0wWwyYXeS4VQ9OsomjPuIxCC0QypFgdVsMxAp07jP//yrz7J+8gS18jI1lSBLh6rfZSZnOPWQw0Cg9xWZFyJNiTMVeKrCyDXYzh3Onb2POx6+jlNW6DYaxFsDBvmExfYifr9FXCuIZiW9vM9CNk/F62DnIyxZwaqM6U9naM/izPIcoqwyUkfMmTXOnSlon7qT+bk6cRIhbUlpliiGlK06TcugrIRIw5BYElzTpduW3H3/IywuruL68wgj5zA8wPIXmO+sEqU2jnRxnBa6lDTNk8w3z1NYY+zEwkigyAySJGFv2Ed0LJQ0Odn2CU0LvCpVS7O6tMTRrSPiuCDW22RxielAPh2TZHsQlqQ1wbL0j5WtqYFTpnhVxYY2CKcBR8Me2ItMUpPMLnCkwsWh4whyETCc2ah+SoRgGAhG8Sbthdcz7ff44Ic/yXB2g3tsC2GWCOlREQVe3cWyahyk+3TdLuN4DMqkIkrqXRu3rOI5Jm3LYK+MSZ2AesVELy3iJx2WrS6yqGDkFnbs0cg8iqCgGhwxPdqklItIIUoKYZIWCYboYBQxmSqIoz5xCKfM+3nAu5uziw/yyPp9NCsVKlbGye4ccS0kTGJOte5lqb2MqtuQz9jZGiImAoIaVR3wynNXcM2Mhl/HX2jw8MX7sRctDie7JGWFSRKh8ilVLKbGPMNUwbRkmIXgSPZnh9ycDfj0jiY9Snh+a48DWgzSQ779P/4rRumMsKwxSlOKvEFnrsV83WBiSgb7Y2ICvjgqueO8wzN/+3He8dBJVNDjtoo5VIrcSjiKc6L8EMvMcZxlculTyS1G+YxQlqRGzmrH446VBawsY26WY6CJHYO8JRjSgyxhsXRYNFawEpe2atEQHUYDqBuLOGaOmWqNVeSY2iAPQrQfkRsCO5fEag9ZOUVUrXI4LXCVyTDYJsZldnuD5U6dWbTDXhpjVapkrkSmQ86d7bJzY4+FE22ORjb1lkYlPTIv4tyJM+TeBnHuU6+c5iDYowhnGGaDvpkxX5pYcw1qXkFLeOQZ9PKCvVmKnWnmF2zqzS5qNuSVKykfef8fovoeLAyIyw5Vc8pgqjBsh7X5NlXvHm5cfg6Vdji49TIPvOdr2PnC3+HFB5ipZlFF2O4yBgZtdwGxuoAlN9FuiRfZBKZgOi7w/RahtYukIChL8nEPxzBoq2Mx6D3LDxGJDYw4Is8VjVWHLEtAjTldO4NXiXAUSLdQ2NpE2ZKV9gkc16eTL9GstRGyzXTSY3ewR0nAJD6iamnMmqLuLXFza4+DqE8xGTGnq9SCkro+RW8w5faNgPHGFuHubeJ2wjgMKLOQkw+fJfOa+G6due/4SopJykGSkdoKNxFMQkkWHSCkS1zmbB/tEU57fPLmLs21Lu2lNapGTG55yE6Dz3z8RUJ3wv4AavYMoRzMGdiFRTE9wCJBN1v4Fy7w8KMP8JPf827e8C0/xMUf/WUQFlEmGTmKIDoCw2ew+wKUgjiOsCsBanIA2mQ6vsY0U6RZiTFXZzMpKIySOA6orsHQ3+D6fkZsubTussh9gW4YBEaVQTyid3SLjVEfKUwTJUqEElyLnqc/DuinNxgMDpgcHFIWBfZkihlNyPMew6OAeWsZUVWsnrqLtYUHKRyTK/0bOJbBLFWcPnkWt2pzlGWMlMY+TJHaYKFeRcy2UUlBnZSKaWDaOWazjVlqCkMR6gLXbFK455gMhlzaPOCFrZz7ls5zwrRxdURiOtTClEnYQy/5vPk//kPO3P0Ek9Qkr9Zo+YKiDNibCDZvXT2e+sPbXL6yzfyZFZrOLtt//j6iCxWs9hJpkjEZTUmCferVZcrqiH5Y0E9K/ExhDgcUWUqlmJIog1OVLveeWceNYZgmkGmqTps7z52k1a4xPoiJBkOkAYuVLr5/iOMv0G4tIKWwSIoUwzDgaIZvKVrNGl7V5uyd51lZX6Ey10TZBUVR4NQ6lAqUMUGWBrMoYL5r4lZNap05tDfhS0/fIC9TbMel1VhFOA2aCwv0jwJUEZEUfXIvpG9s0m2epWn7zKSmsH1aXk6iNHe8+UHu/bbvwfUE8w1F3SmhUlLmgsSMkZUK0rc46TVhNuX0O09irN9Br3+V63nI9lAzDTNu6Ziv/dlfZ61iMrn8NKPLn+Xjv/1+Nm7eZvLcNrlOiDDQvkdmDjCTQxpmjZrpUhYpo1qNsS+I0QxTQbPQ6GqHqlOQhFMGMkc5mgYdSl2wnx5Ss07SaZ2AouCwGDFvnMcTVQrpIYWhcVwfpTV+x6bl5YRJTpxMyJIRRm5QErBUO0nT61DruFS0QgWCIp9hhDn7w5u4E0lvco25WoVXb24zHJbUCwe3SPGTEqfpEyQxy2fPs+I3GU5dskgwi/rksxBSjUaS6QrlYMoX/uLDXHJe5V0/9yMUwmMyHVANMuYdk5O6ipP6LHpzFFkKdUgtgzu/qcSdfxhChW9P8OQUMwz4n//i27l54xaZyPjE3/8KjRWPda+Nl8UMPJNkOiJxDHS4QGJosCIalVNkukElnrHudqjLjMI2cYRNp2YzN7fEbt6j1eiSDXaJ1SGBvM2a26HZLEmNhKrZpe26pPaIwhOE6QEyLUtsJTC0ph9HHBkxRqrIpYNVLrO5NWB4INmNJ6TOHrMiZGoFWLYiEQc4NUFNrmIsNUiEw954iBGadDsOB+GEKDokMwvavofXrUM55mgyYK6WEzx1QJhOMFTOVDiYIsRxFANTcfEHluguSRxzF8s0qRkWjumTGgUVz8c76dNwK9idOh4aV4zx8nt47IfO8tafew/jYgUhZ/hI/LDPqmfSKsf88f95hUXDYLK7zyDM0KMEt9GAIaT5lP1+zv60JMuO6IcjJqokcjNMt8lCmtCY7zAYDfGVw960yrnvOAGGj8ojaswxyhXbswP2964ynPTQscZ0F0mCAGYm0jLF8YcPEI7alEEDag0uzN2JVbVYXWvTrnW5cfU66dadmKMO6lqO2bPZv3nIqKcIZxOmt27RLDOskYHfqhMpsHyPUTaj3jGYhjGnjDp+3sRKTjIbphxcHSJrDUTX4XyrgWXVIU/pFROKoEplWkWICt/9s+8gK1IKPUIZFTLHoBGHtKRDHEyZTQaURkmzOqEqKlREle/66bfwtb/8E9zxzV+PkBKR9KmkCk+XbH3+MkqkKCIKUXDKrfJyNGYWjxF5jm/adNZ8mtUK9jQh66dYZp0r04Ld2Q5pGCNlnR/4wa9gsCsR5gqeUyKyGU4x4qx1mgfn30DbX6BVGrQQnGi1OXm2g2nZNkausA2TB089RFlKOqnFJJxgmgqhO6w4i3TXulhuRmoVaL9DVblU/DWGkwOWa/cwDSYI6yyLLYs8/DD5bJf15hIvZ3MYokLdtWlQI7855fRKC7t+B+nJIcMdF1fDnNeimW3y8kHCz/3SrzONU/TkAKc1j0/Ct3/XCr/zV79EE0WWJ3j1BpN0gi0XuLP6WmbRcfn4CM00ehkhaswN4eTdOWv3fBNNt8GGjKhPFvjon/wR25/dozO/QnjrBtTXaHYa3N6P8UTEHW7Oq8/sYXkmVGo4/SE3JynrizV6RyV5PEA1a6z3O7zrwUfZs8ZsbqdUWnWqpSDTIxJVZd5vMMHgYLbDJD7CjKqYFhpllhipZrl9jnG4j+tVMWcZ7fkTHE738ao2nuniSo9ZMGCch/SyLRqNJo06SMsmShX3uy63jg6YTiNiq0F4dESrENy1doL+cJ9Kt8G5ey+QiAy3aVJNc3ashNNuFW32OdofM3/yNK9sfI6Lp17DZOKQpgZJUnDmNWdp/GWHpl1iC4FbsahmJm41xLQEhhVRlh5rnoEyH2RSjKhY83jpCT7/yifoO9A/OEK0HL73W74L/zsU2HX2B/uMximW1eLZj73A5z73GfauDIgtxaKA82s+ZqtGLTneajUdgzSN6G+8ylL9zQRmhhpB065ihRV8v0LL7DKapWzMjmhZXVp6gbnWKXYmryKu7vR1p+YQpgVxnqEENGwPLSSxkphKUhoFVmngmjmmJdFRjPIrlEpQJiXSsjAdgVXk7O4P+c1f/QWeub3BrZs9yskBH/j9P+Iv/v5veeDB+/n2d72TUkve+vBDPH5hnURkqDChXimZjGb8zB8/hVbHG/lSaXIpKbIUURZ89zu+mjNrEtWo04xhe3rIotHhu37mNzkc3cSTBsKyyIIAx5NYvk02SYmDHkGU0WxZ9A4OqfgebqWkUllClwWlKtndm7C6bJGkBu/71T9iZ/M2jTQhchX32C5lu0bLC6l4FVzbIuhPqC9WeOCBN7Pf8FlvL2FagkargyoVWW4SZgG2Y1MOA1KhmJ/vYN5+9TLR3XdipBohXUwDpomEXCE8QS40dmliaMGstKBUFMLGzjSWKcE6LufOEoO0LLh+/Qa7G0N6WwPsWcjUNkhSg4XGOp/48Od57A1fhSUNqtUuuyrEzuv4ashBv+QN/+zniE0TSoFlm6CP692lV4Gi5A+ef5Z/8eYHmatXEf6MWlAlLkrOPXSa9WKJLC2OCRRCkJYcAxBsj7TIsKWH72Q8ZBigMtJCUmhJlmRERcYZ2yPXJVlZ8EuPfRv//pu+gtloTNcFsogLbVicu5+D2QGH0xRRq7C3H3HP6z2+4ft+BiVMXAMO+2PqtRqkmkrFJM4UthCUpcI0JWZ7dYUKLpGZYFkSledYCCJLY325hj4G8hKkUKBMpDwu81dKfLnmXCDKArQkSGI2tgbUfIupbHHKdonSiDtefzdKelRsh5qT0bRj3JlJFu0QlzlaaN76hrcwy2LM0iTNS1zXQkqBzgvSJGXUHyM8E5EUhAJ0Q5BrF2ULxhOBkhLHsJGGwJeCEoUQJa6QCJmTl4ISMLRNYSgoBb7vYWuH3ACrNLGVzQ9//7dSlgVNryAxDW6mHlY/JYyvYXc8KsWA/bJO269y/dlneOL7NbMsIRMGhuWRJIpclGTJscQjOr58ucTdMEGUSGEglSYuC0qtsZTA1gJLCUytwDzOxliGxjCOa80pC6QEoTSWYeBbDkaiycwA21e86bHX0ayfILJC0mTK+vwasRa866vejko9Dqcx4yhmFufs2U3SbIYhHBzLxHbMY1xHqRBSoiQ0WnM4dg0tUjzTZ7VsUjEiwtERwjSRQhBlCdPBLpPZAJXHROGEMBqSZROK4liUWWooFWh5DFKwhMTn2HlKNPXZNvOtBvVuB0PUOCMEsRGxIx8jnpRElQW8VCPMiEFvxEf+63sxtUGm9TF4wQBPGmgh0fKY4aLRoDRmGcWYjTqWcwxhcUwX4Zk4GAglMBRoYWKIkhxFriW+kihRoqTEFvJYKFTkKLNkLAtcYeELF3u5Q30yQ+YNpBmytFYn771KmSoKMWNnMKVFTlDV/OZffRqhc0wBhpBoKSkBy7IAje/7kGvKWYjpazzLJuj60I/BrqDSFNM0QJfQ6JCnBaVwcGoeTgGRBF0qUAJpSgyOHSXTJalSWMJEKYOKK8F2EIaCw4iVSo1eZvEv//0HWH3tI/ziNz6IqVJwNUfFPEu1glkvwNCaQmlMyyIX5THQBoUojnFQxv9dl/x2jVgYSGVhCANT5BgZ5GWOljlC5lhKIQCExJAFuc5Q6ph7UuYleVGQ5IoyVthaMb92gqNgyN71LR7+iiewbbh1dZuFEyf40X/6EwjL5Fo4RCcZ//Q//mdmM4XWGkMfzyKlBZSKUinKsqTQx+QfWSiu7RVot8aEBJVrpLTQFR+JhUoVaAslSjAkSThGF4pJEFPGOVmmieKcOCuJk4wwSUjjBJUURGlKmE35kz/6LU535sHIiP0apSrBSFl64A6QmtW3/SyDIGJ9ZRUvzzBNl5v7hzh1E1NClsXItPgy0oTjLROOWV1CCszfft+P8sLLO3z1930X/jhEqW1Mx6AfpTQ9RcVood2MtcoFbGwKOaPueOSuSziZ4lKQNmuMjkK688vcfOUT7PV3ENUGs8FNbn52SOfBe9h7+RrRhZPMZiGN0iIUVX7tr97PK5/5az74zOdxxQxLS1LruIyxxESWCul4SHHMrdIO3PnI6yiTV6lKg2EyQY3G5JkmVaCLFNOpkIYF+STErzcpw5RSQapSFCamNjCLGC0kprAJdIahM6LDKU5nlU//7v/iZKNOLgSVsmBaxLz27n+McAwKnfFt/+gr+Tcfb/P5K1d54/lH+NzGJc5cfB2/9L3/kB/+H7+OlC4CTWkILF2i9LGzlrJElxLzsXf+cx55+xTTMOjMlzzz/AF3zC9y/11nGPS2ubox4MKJu5n2DpAEmE6TBcNkd7+HNE2GZsDkYJfXLjzErdEWxfgA08ixdIVHLpzhL/7mM9z96L3ocEQ63GLz9i3On1vlvf/0FzCV5PalW0yuvMpePEDkJgUJoqyS54fkWpKGAQurTSp2DWnWeNNr7udTn9yk2i5wY83Q87n1iT/DMzThNKQUNhUxITcr7M9K6g0XV+ZMk5hcGTjS4sADmeU4XpOjKMIzLSbTHXBrZGaNiJgyyvBsg97REee+6w5u3/x7hOhQEy4//4Hf4wfe8BhP3X4aq3uByWCL2F1AGTElAlXmOIYL5vGblkSjlIkJmOdPrDJVKSJO6B1d57H7H2R7tM1suEci4NTJFteuXMOwR8dTaE3xav95kjjnpLXEIJmQOJKPjZ5EZzmxpxFZhEhhUAZ0mh4yTrnn7kepnGjiL7S553VfxZvfcA83r77MV777TWxtX6ZTb5IgQFpQTnD9Jp5jE5oewbAkNkc4dsT9r1/lkx+PEKXPQRJTw6FVMRnOAmp1m7TUJKZAmAa1qiYvNDYVjI4PRcHe5BLVqEZs5ohiStVpEeuAemeBshpSYpHHirnWPDsHOzw9gu8a3SYLPVwzRHt1Xn3hcxxMXZy6wer0iGd6KRfvWWDzMx+mWm0yHfQZRj2iUU5n7QS1Vo2sP8H1PKR2JC3Xw3Gr2JnBtYNPsB/f4nr0BfqjS+Rlj7VFePy+1/LIPe/gQuV+Hpl/M2fMO4ntiMXOQ9zVvB+ZV1lquTxx9mGyJGcvHDDY2mc4PmD+pMtW/CJRXPCD3/5veO8Pfg+9vR1GO3vMLa5Tb3aITMEo3mGajtg86jGJQxzTodasYNkloU4pLSjjiN7OhDKH/aOYrBC4nUVac/PESuFbLpWiTrY/ZXQ4Ihz1Mb2SbDohzfp49iJBNkE7BkuVkwTpmNLwuDG4yuZzNzF0gWFHFLlmq1/w8//v96DTnJIIz6kSlAkPzL2Ov/7E79DLLdLC5cT8MkVe8qu/+SS744BqZ5Xl5WXueegRJsUUEc8oioydnS1M27ApFLg2VO0Kw2mKKTtgVVmsL+C1K4hkgecvP4ee7XD3/d/Mx57+PaS/QKQzls2P03aXCEroVAwc0+Q1Z5d55nbK9uYE29DMkiHNeps//LM/5l/9q/+FXxcE1/aYOTsc7Hnc2r1GpdFgeaGLkEPOrN3H5uHLSNvHLOa4mV1mdGjwxANrtGtLCNOjFxiYssHmNOUL1/+cC503YFcUiSjRJayeW2V7PGMUvcQL4Q4tWWGv3CelTqtVpZ0UTMobuF6Ho/6n6dQe5tf/15/w+lXJNOry0mjGq5MR8/Y5DrOCqBgSl9s4ac4r6T6W6VH6S7w02oTcpez3OXPXA0TOjN1eid+ImM36FIOCqRozd7KD2EqQQmhcQyBLRWpV6acDDCdg2a3je+tcvXGb/eFlFhcXuOOO1zOa3uBr3vZ9XDhzmgfOPMw9F96JsdzlnjOnSMs6ayfvZunek9jVgMOoR627yssvXMGo1PjRH/1fHG5f4XNPfpI/+Z3fpcNpxukznLzzNHWnYHaUcmljzOXdyyzZJ0hmkl5+maVWndfcfwepXWLWJIdZxHQ0wbEgK6bsPNtnc3yJln+Ra1vX+eirX+TV60dcObzJpesTlpMHqFnrNI11yrFk49ZtPnv7OpeuFURhDq0lDDVhsWlQFhVOnlwliVJ+89/+GFYYced8hXOdcyRRAWaL+dZ5Vjqn+Y1f/Ulmk5T6yjJri+cQbp2//+Dn2cyexQsrZEVBMQ+ptukfQGl1MY9DuuYjf/ybPPj2r6Kq58i1RTjJUePrVDt11mQdCpNXDzfxhcbPFc2iw9z8HEkSclK1wOvgNXwOBxu8+d4neOozXyIyNN/z1W/Db9dZW1wgHG5iOhUsZfBPfui7ORokpOU6Vp6yNP8Q0hGcyDtkZUaeaUgFy5WLPH34GSb9W2Q7h6y0VglmYzYtTTMZ0nKWuL4/4c3NZXqHt7k4t86C06bWcjhn3sULjsnUvEG3eA0PLZ/jNcspk3CPZ27ewrUF0tPEUYlTd2m5NrcS+MJL2wzHMY89cRdPXX+V1eYyvWjGor+EJyY4CCZ5xNF4h1/7wM/yL374N5BzCXoiCVsNvvLsu9kfbtG0a+wevYxl3U/Vt48xeiQzcrfCm7/h+9i79SKt+Xm2klc5d+K1ePEepZoS+23MMGZlcY6WthhPQoQaEAxi7NRgPw9pEFMkHmE2IxuMaWce/vI6C6dPMZ3exvPmkU4dz5aM9m9xrn2GmmNheys0pIvvKNIyIss0aSpwpMSvu4TWmDctLmPK+xicmDGJb3LfY49y2OtRCSIOdUo+SHCsknRxgKGXOLU6h4oKbu1d4VS3i3IS4vw2h1FMHNQx7YxarWT74BLr9deTxiNGZYVpJFjtNNlNjpjcvsYrkw2W1xZ45toznD29BgWEeYZpe0TBEXbT5mg44+Lpe7l1/WXGg12ktChtiVuv4tou9/BaqtJkLALMtsBEGph5xO72Bns7mxwlG9S8VW6Pnsc1PXLVxxdtYr2HnAlUbZGesY9RyYnGfRq2S6XbZTgeM9E7SJUxm83QJ6qcP7eK8me0aku8cvgCT5x+I0mW0V4wudy/RGj0WBDrTGwPeRRgmTUClSHSCcLsUC1yZGJxrTdgrvN5lut1mt4SP/LD38CP/8ivUK10maI4sXAX4cTCkC2GYkS12mAW9mm2T+FZms3ZkHmzS7gfsnxqnVe3n6blrdBedshEyOOn7id3Tf5eK5556nksPeGnfuPf4aO5ffQM95x/BM9s0Us2qVfq6HyIYQCpZrle48f/3bv5nve+jFE1GG5vce0Ln+T0Q19NMushagsUMsOIFDIxMQtMqrbJiZMX+cRL/xPkHFm4iWPOo50mR4MRS8k2iSGZxIfsDvdoNeZYjFbZ0M9xNIJmmCIrgEwwbZ+lR+7ic5+7Rn/2Kre2FMx1ectDP4g0LbLJlM34FXx3HZF2GPgZRqzBLqgbOXP1DjvTAL/cZSLnESIhdGc0TZetLGB0dB13zuShtz7K3/3Zh8mk5Of+08+xMfpr8kywO9qj196gNneC3u0bOE0HlMbwS/YqUw62vwQtG8c/lkb3gz2GYYWOXgbbwHILZCFYWW9yS93GkS024x2K8dM4jUUS0aIsduhnfSpYGGFK3DTJFLSnMVEx41/94u/xZx+4h1rdIBrvMCMniMe4zWWkY5tEhktASsO4F6Oc0vBOc7p1nhuXP48jTEZRnzwacs5fp2IsYkUJMz3jnLuG4a0T6RAzt3Ein3nZ5tKlz2D2tqnadaR3BwudJfxqF9NzSOIZZXuNQu2hK3VGOqVZazJI+hwUEzb3X6GSQm/YwvZhexCSuYLtcI/DeI/aXJOqXOLcm5a55+G3ssA8Vq1Gw21Qa9Q4u7RMt7mCWxrMajsokdNsrRHoGBGPCBuCvYOX2dp/jmEvoIrFSmeO9/+PD5AOFEkac2s3ofQDmmIOt9Ll8s51Bpjszfa4tv0qqqzQLAx6wRBtePT2L/Hz7/sBHMti5pi0V9Z4bvQir4RH3FA7jNKILxabvCpvIrNSI5IRvutRDFNapYfIMq5fus7a/D1U/ISat4DnN5imMVUjIE777O68TC94FT3YQ8qMXI3BT9ncvUUlWuK1X/8OtrYTLHfC5sFzpNEI8pLewSHjg2eJswpaRQyOrpHYIxpelSIWtJwlEsdlYd6gtz9kpWVSF0ssdc5iKoPYEgRhTDQd8+3f8jZ+8lf+M/M1h3NnvpMsianYLQ5Hh1zpHVIVK6RBTnh0nYPNLW7c3uBgdwNTXGAw8dgf7/LirasMRwXDzQxVmTCOUn7xb76DF2/cJkxjZsmExZpPXPrs9SI8UUXboA2fTuc0wnGw6g1G6S7f+OPvxJlplhyTheo6e/2rXN59kiO5z4KssPnybeTtm1fxqnWEFgyTCb3GESrLoDmHEDaz1GYy3WFra4/9wWVuX7vM1tVdFtceIt5bInP7TKcl40gziavUGicodcphGnDy4hqDwR797SMatkMwC+gNNjD1jEpT0VUpVu5xuHNEEHjs7V3nmd1XyYcx8STk1PJFKOsYIufpl/6CzMy59dLL7BzM6DpLXB++wsLiAmGeYxkFvcMtbh/tU8xgEF5BBJpUDYhSj0rb4+6zr2GufheBMKgaksVui2a3w2rlLvpBwNEw462Pv47Ln/4s9VLQ3xszGW3y8tVrtNI6KsjRyYT9+CWCuM9ob4ci9dmeXkHlHosNhTRgkiuKbMjVrVssdM7R79+gNxhgE2MuVbq8sPEi661VkskB/uWSg+aERiVgds2lszLhSr/k4vKDONYAzy1gwSCZ7GF3LMJihYbO8cwGcRERZz0aay1mmx7z3QJvVHDm4rsZTKYE4ZAb0w9hRy2OyiscjCtsh4dMqMDci4xvHXB6/QS3wxApLnPpyvN0Tl8giiLuP/mNXL/5NH77ceJswG68w43ty7z5we9kcrgPJxsc9DXzxpSFpTWMMEd0u3CrwjQY0GlBq1Il7N3ElH3GpNRq99EMS86cfYD3/8Zf8nd/+9+xujHnTj7KIBgzGG7jJvPUnFVmB32sYoiSa1z93NNMpor6nGLDPWR90aQ32+TK7hY6SelaFk9/8YgTldNceXWX+9oneTXa5dTqCcy/feZXkGWPK9F5HnvNo7yy/wzZeEBNrmIvNTCyKuu1Y8JxnIyxai1SMSYsGvTGtwl2TYyLTfpbE0b9XWzh8MATJ1lsDumaS/TEDd702q9EyozDS3u0Wvdhd3KacpGBilhvrFK3FE7yMN2vWECqkr3webruw2z3Mtw85czaAwThjFOnXofv+uyPb1AMS+paU29aGLRwLMU/ePdP8pd/9JO82LvKxn6CsfgsWaOgCBV4LWRjkSQ2GUcZ9hK8eusyj6++gyzMyLJtOvMlW6S8vHOZ2aEmocAuLlNtrvHF219gfukeXvirv2P97BkcY0LdqxL3XGRrHVNf51TjHPe9ZcgLT+4w/LjirV+3wD2r97J9q8/5u5ZICoWcTmZkjgOVCU9v/gHDUcId51cp8xoq2KN7ZpX1TgNH+QwmAbW5RSb7Q3Zu71Fza4iZphsuc/T8bbS3RNWQPPmRT9Mf7jAJE8qZxhIRaZxCbjLpTdi4cciNF3dp1ASnlizuWHqQtDLh1UtPc23zaW5c2mKvt4NTG7N9kPHpJz/Ebv8aW7df4mjzFeTAZq93RHelzu//6S9huVXMXOFLg0999En2NwO6Z+vc/dr70cEiVqvgYH+Lz135KIoe2pgy29Z0Qp/7HngEw6uwt7XJSxubXP3Ms7SCB0iDAyY7+1hGh80rB9z92J3UFzOq5jlsmXJyfoXRYcnSvYogvI0MK+wmB7ztm+7GbbeoGC7FZIHdl26xdXSNjcs9xHaJTLlF7+ptnj94iVg5nHUX2N064uDGkIFIudV7iaTQ2O0WFavLjedvEactTp6+wOyyxdkn7mEsA+549AxLCzmnH17jwYcuYKQrZGJI9cxFUuHSOzigHE8pkOR6RuUuxeHuiK1XUibPHnCwfUitssLZxQdZOXuSSVInnEDH6tFdvYejaY6VrDAYHnAwNKl3h8Rhg42rr9Ko1Ll16zqWtHjH1309K3es4dohf/25P2H17BLryw8w110n3xHceiWGos3utQ1+4B/9EjqvUsYJL9z6eyaXE6zIJ+9+lLUVybVnDlleu8DdjbPsXx+w+9xnuPetdeJxn37f40znNNefuoG2HEqzzonKBQ5upqTxmM1xn9LIOHnPfbzr7W/jdOMEI7WLWSQWZ+5+gPb+NWz3JNoaMtorWVxI6cUJp9qrvPzqi3jx5xnHgjMnLqL9AiYZZx45ibR7lAuSKDDwo4zEsBl5MzpdwXg444ff89MofcwGvLL5JI7hYZnLqKzKWEnWujaHpuacvotbr1xGJhPiapXFZpVSTNB2nZoBq/U1xuOESy+XPPjmKoPJA0z3t2l22sR5zp0XLxBlBXfd+3W8dPvPOQiXWG52yPQrbG1PWHMfYGVtQFq4DLMRTk1SRhF1T/Dyq18gyQoe+aY34LdukvVsCrPgh//VgxRDwexeC7k5ZOY2qfodli/cT57ZREFKc7FLOJ4hmzu40/sokz52YmMYki/8zassfnuHl5+5TN1o0V08hRxLh73b1xB2E6Wm+Fad7mKdclqhWV1idzOh0mkyZ5zhvvnXcRj0cZqaIIhQOkP2FHsbQ+zAYnd3xNZzVwh2IwbjiNITHPZ2mBzskaiQXf0KuTCYTGGh3kENA9zUxh+D68c88pWvpXqig+/H3B4/jZ5mQIn0C0Zpn+q6x9KjHchjgiyjubpAnpR8/vN/yGwWE8/6nD65gikHCON5KnMJUrU5eeZOKksuStS5drnH6cYCv/6+z9Geq5OSc2vvGc6fN9jtbUKquHG0y8sbO1zbvkG8tM149yr5NcXyyVXi5AY2sHSqSW6kYPiYpqJq1xlNboM3x71fdwd33HE3J06tECf7nDl/lupqm7AokOvRSdLQpuPW8PUaX3jpMmF5hO/VqboLbAwukQc73NzZZ0dt4NbXeOXjPfr9G6jokNhapuPUOdyI6W85jGYhecWn061gFQHFbMDWtee5/szTXP70IWa8QWNFcnAzYK5RIS18xkXOC5de4tmnn6OMJixXl6nps2ROTm3xYQ52DwnDmO2t5zB7hwxnY+xYYtsKd07wyc/9Ka32AkmaYRuCOKrRSuYwpYcpHQ4Pr7O1cQPHMXjb1y0w6WfMwpAoTnjlmS+wxTWev7JNp93k45/ZwCh85i/O01qa4/a1AaLq07iriVup0ovh0sFneeHpv6HuNfHihGb9HJPREf3CRlke73rDQ5RxCm6LK9diWuYCxaZPvWtjri12STChaDEbbTLnrXDCW+XmQY+1isuifZrdvQHdSkmxbdFZT5GNJoezCcOxyWtX5/nSK7ucXluh1XQ4df4C1195HoRNS97H/mjE5z7/LCcWF/mKdz5KsGEwvz6HmRUERpOFOYNbNw5xqzYt32c0Cdjev81ge4p7tkV1cMjR7pThXsR8t4qcL1m0fdzWgOjAZXpkUMo2ty8/Q3vpBMFsyrd900/wc7/0L5m+8nl8d5kksrh494wosrhyZZv3fs/7UGVOFCfMrBx1ZHF0ZYydX+Lc8ioLJ1coezlpaHN2dY1ePMFKt9m/OcUSR4TXFojFmKyrCD1J/wsfxb/bYq4eMTqacWlsIFTIzn5AL5rQPvEKFy5e4PLLL2HaVoWGtDEtk9a50wiryu7mAXfd8QhXnnmaxbWTrDcvYFVSmu0lrr7wCma95Fz1PrSZEScZd953kTyr0KpWCKOSk3NnSX0fVcSsrF7k4UcEX3r5Y0zCCWfPrJBOpojmKq889RT2xbMo0aCrK2xe2qN73zrkBq5fwdr22fP2ETWf5fYyp0+eYFrsMOkfIscJtiNZaKzQXvEIFBjhjMneNqXpsHH5ecg9HvvaewgPUp575hZn76ww2VpivLdJVEuJZxEvXf4wHos89NA7GWzdZBYlVGew1b9B9UyVqF/HQlJpLLEw3OJo0+Sxt97F2doZnv/idTzL5M2v/wZ+6U8/xL33QRE3mOg+H37lSVqqzolT54hHCePWHq35uzHrtRabg1ssuKuYToXZZEKj4hAcbvLgo3eiDYPZIELiMtq/xalTZ/AtwVGYEe1tUVlV9AYBou4g/BoNlVGv2uzmB7zta3+CLNijaimqhctCpUu0m9KYP0k5C3jiiYdwC8VcdYVkNEIrh+XGOjeGn+Psa+4l741Ic4806dJcaKONhHXV5TCvMHfXKrc2LiGmKc7E4jOf+TCnuis0qwvMr2Z81du+i45M2Rheout2eOfjT9D2HXoPFiyu3s1kMuT2pWc5s/Rakp0rVKpVDLeF0WzTnoPzJ17Ps9k1OJoReQ53u6d4atrn8cdez0E5YetwQn1BMwoKnGqLr7r/KwgmB0y2htw04HXz93Bt5xbJJMTKTxH0avhaYo5HIxqVJsRTEmUzmBxS0z6d7jJVWWL6XRyRYpBieXOk0xRhSZardYaeQVzYGK6FKHtYxhqTLEeYPumoisp6pNOUwbSHMC2yJMMzLFw5Y6oKikFCqCxkdMRis01YpjTnl3APlqhkkOPg1WoIo6DTqePYVWaTI1ZX1tje2OX8+h0oUdJpLKE3XmTa26dz8jzjwx2+85t/kD/8wPtoOGv4boXKSp1Lz7+KPvKw73YRpsULN77I0skO7dYpGvUW7cYSpWUyKLfJ05BTswX2dI87uxeIppKzK2e4eunT3PHoV1LLbIpWFQ4afPKF5+gutmjUz7G8mtA8OsJ54j6+Wb6VT332GkeHioWFjEpnGbPhFNTqS5SmIstyKt1zCCWoe5ooEEgz4mh3l7mlBdxS4DtQGBauNlhaWeDy1du03QZW7QzDgxELvkRZkk6rxUuf+TyN9XWS0ZD77j59LAXLJDrVqPSI+fkTmL6gLEu0AjOFeOMW85VlLNNBtGKUTmnVXYppyNWNS7h2neVQsKi65KMxw/EU1YzozK0iTzd59eMf4XVvfhupDCh2Jc2TLTzLIt+bcOrUHSgbDM8g3o6598y9LJ5YYDrKqDdqyDQhDXPGmUezs8TKcgOnt0HT1YwKSat0aD3wMIP+JtOkSaXj0zZrOFUXQ7nE2Rin6nNycY390ZjFM2vo+CVSVzPaDWk4M8y15XPIhsXg4IBCCGyV0+0uICW4HYdgsEerZVB1qmAm2FaL8WhMYhW0ixpn1ubp1FZAprRadW6+fImzrTrT/gR3+RRffOaThHuvcvbU13CQHiOd5hfmWGCOYRQQTiMqlWMB5vkT6+RlgRQlDaNFJj2qtTkmcR8jS7lw8gKTYIpZcchnIUnkE8YFXm2ZesXm1b1bzJ88Sa4Nxrduce+b1gnLmGQ3Y1KGLJUur3vb15JlGXNnT/DycxGXr94m3J2R2wAzas4q48EtjPummPokXRxM26U0ImRaxzcC1rrrmJnN7dE2Hd8n3e2jZxb1SoWFZpf9nR2mW0d88rBHJCR2XmA3LmBWFOJTf/EHehiOqdsO0pDEaYlfqRHlKb4pyaQHecSNV77A6QsP4lUrSApAkBc2FCE1w2IcJ9iehUg02DbXrz1Hgs3m5Sv0rZzv/ZbvJEoCPO1RSM1okFBpgshsPO/43LU4ifA7bfT0WPoWpT08r4nlNEinI2rVGs+/8iqduUWkmBDGAUtzawzihPlajTBPMKw60snZ3Zlx74XThMMYjJyD4IDHnvgarj37JHGcsHbH3dy8dZ2m62Apg1u9PYLokLnaAq7rYDlVqqYiKzWhVuQx+DJCmRaGXUPoCCEt4lGA5RvoosA1XKJco2SBi8P2QZ+/+czz3P/IEzj1Bmc6AvFXv/MbOiGkLqr00xkqy7G0jVu1mM1GtNsVBsOQ0azk7HwTLT0CDvH0MQXZdarIeo2atDgKp1QNk/3tI1priySjI5rNeY5Ge5S5xGj4eFQYHm5z+vSdfPbvnuSuh87huB6yYTEb7FFfXILIAFNR73Qp44zB4T5KaKxUkpRT3PUOZiqRqqSMBFZVESUuFhNsd44rr7zM2bvXkcxR6JJwtI9Rc7n9xWs88LoHODwc4DeaHGwfcXKpQ27MuP2lK5x/4xsY9HrYukTbOXMLi8yihNs3btNxF2gvV0kKA5SmyEDKlPEsY35hDteSaNeiHCZ4lZwoc9BC8Qd/+DT3vOXtVAqBs+Igfv99v6jrTQ9hSAIEDa0YjcbEswHCr7CyukQwLvEqFQbTARUDDGlgd5oc7W4x5yyzs7tFdaGKWWnRrliMiwBDO+R5hJ9mzKRF22shVMJgY5+Fu06gRIVpkmCoCaZRw8hL3G6bdDJDuAIz0yRxSJEnuPNVytnxARmZEkyTKVZpM5seYBgeZhTTOrVKpiwa9TpmkqA8TbKfoWshSagpvBgzBRV5NBYbhIMBotlGIBne2mHhbAchHabhhLou2Ts4xK60MeWE3mHJp/70aX70//mn7BzexKrUCK5ssnD+PKZRZZhOWdCKcdUlHo1YmVtgmmriIKJRlUQHip1ozHi/j/jjX3uf9hoVRv0epmVQdWooF8LeCFmzqTkNJkgaTZt44wi5VMVLM0ZFwoo/x1AXWMKhmIV4tQ4GGbkRY+sWRTpCSYvw8JD28jxlLsiyjHE0YHVlCSV9wskEr1ph1N9hrjJPYlk4maY612LvYJd6pYpleCjXIB8f4fguu3t9TGljiYT60hoqDojSGSjJYDDErzUQccryhfPYyuPlS09z6q6Lxx9bs5Qwzmi4HqZjU1QyVO7QdH3Gkx4YPtODfUxc2itdjm7eprt2itRKGR8O6VQdqNXRYUaiYXC0x8riEtI5BmRqlTDcuMbiybNow6ZRb4MSRIVCFDPEb//cj+tc1wnLIXM1B2+ujUgE4/GUtbUVrt64Ts2p4zpTrMYqahKgrRFKr2PYM5rtLvkwYzfp4xsVSsug4bvksxS3UqE6Z1PEitloijANXKmIo4TBYY9Td93FdJxQb2dEE4NCWBxdvUle0RCVnLz3BBYLeHVJnkXIwmay3UM1bSgdah0bmZSopkd6MKBUBcKT3Hr+Es21JepzXSajkHmnxGqeRJZD8tJibmGBwcGIiu2h6w6FnhEeDCiEQdVtYBohMs7IPItUm/imIEgMWu059jeex3UWqXmK/jBloEdUYofu4iqBOUAMBK2VVYpkRhprjJpNNp4w16oyzk3Er/7I92rPr9BZP4n0W+h8xjQNMWYJWbugaS0RD1I6TQ+tLdyKBRWXNJ6S9ffwW/Nksosx6xNIk8WqR2ibFP09pmFE1W9w49ZznL34EFW7zuGkT81yGIxusnzmAtJehPCAg96UWrVCfbFJqRoM9m7QWF4gTEPcNCM3apTJBHSGKQW5FrQXVwmnMYvzDYLhHtqqI4AsFuhMUcgRZRzjNHye+djLPPaGB4j18dG7K2uLJKVidDjG9SRGXiA6Daoa9nohvq7gtgtc0+KoPyBNCtBtzHzA9v6EB9/yWrZefA6rUcWvNLm9dYPO8jryYJeD3R3mV09RXVknH+7jNqsUYYohXcRf/tr79FE+o92Zp3+wQb15gpc++3nO37nI3Noa8TTHMgSWYzPNUwY7O6ysLuEbHbb2blGtuJh1h7nGAkooTCDLBP3eLTq1eQq7QZTustBZJpj1ieOIqt0isiAZhLiuouav0ux6hJmi6tVJ0gRTBCA9JmHCXKfFeDii2Wlw0OtRq3aRFAgD0qM+yrI4HG2wdvpOomlIKSQVoZB2Dc9rkidHBEFIbsKc1SEWGXkQICseFAXVxhxxMUDNJMKyOBpcpds6Rx72icOEhcVltGcSxinZrIcQNrpi4mYS07GJDkO81S7p6BCz6mMI+5gkHczIC02oZvh2C1u6mLNsyGpnjqP9MaYxT3/7Ng+85l5uXv88kyTF9gTl0RFr99zLS599krnaCtemL+NYPnmsGKuAaSk4tXoOJSUFA5YWT+PYDtev3GT5zAkqzRZHG7tcu36Vux54gCSbcTRMaLYWsWouMskZHUVYLZ/NK8/jrS3RkDZhGBP0e5SzCZ3uEqODTczSIA8HCBNyoIhztFuysLDAaGeIXZGkmeLFz3+cE3fcTXOhRZl6CHdGTbbIPE0xGzC+vUtQEVhGlVUXstJhMtrFaVYZZw7NsqTS7mJ7EaVTISomyCIniQXtlSo3Xuqx0K3SqXW5vnubRtCjs9LAHNmkQhGO9rAXlzAFuH4XcxqjWgrx13/y2zrobSOVgWX7TFVM3W4yzQOq7QqD0ZTFZgWpLWaly/b1TRq1CsIvqLs2g/0+K9UOkQUHt26zfMc5ymmA7NYxkXiGRKic/jCm1a5hCckkichVRtNy8XKTtGJhO3B46wbN04sUpYUf2eQNRXB4yFKzxbioUq+aaAVKVQj29vEXWpQyZnzQp7WyTt7rEVjQrjXILBdLjSAx8BwTUWkTjgPyfIpb61BwgJVVKYWDiCc49QqOXSVghi41KtO0fJ/heEY2G7M32aZVc3Erp1BBTu4q4tmUhZUzZMNDKguLiCJEK4PBNKC72GU6neBoE7tiEc8ywkwhb7z0PGatja64TKXG9SymR9tsv/wK+ShjUfoMDkf0d7ZoioJOLaVTOTZ21fU40WxQ7R4fcnr24lmqpaLZabHUnMNG0HDq2I0WC40GRVKA5dD1GrRsE79RRXeq+DonPhxR7yxiizpOUWP35k30MGVt/R5mZYX/b0Fwm6M0EABg+G07nX5CCyvgqok/vIF39AKewQMZ/5ioiTGyEMCy3Zm2DDPt1OeZh4Hm6UCrety9xRYRiXT0zZXleo25NIR5QpEkhFmKbRtc65lCSXcT9M0F252IZYAzd7ACPSgQEy5b8vPXnuPlhOhD3IsjCRYoZUiTJZ0Leb/9iBzfEriYqBL8+P3E8c+RNAiZEBitwQW4TpFYQz801HHBbC321tOoPZPtCb58/jQXi5QRyV31PKxrUimw08DheCFNBdbFxFKwWm7I6xLdNsxBxLn9xyqvMOaZ+uGR8K5QzpCLJflizWxeGIKIOBNMBiLv8POIxTErTb3b8qINcZVTiQWq/cscpeTlikm3JLJC3RrKassUeuZxxneaMsvQgyIrJPrukV4gSzC3gaR8xe25I1nneNMziYzx2iCXr3FDz+GwZ7OpmeOEqirR5xPhHEEu8D6gLDJG59FGkWcVMogoohjdWc6HI4sqZPXuDXfviSgQRcy3r9/58LjGjo5YSHweMF4d9W6HMxopMjrb4dzIfzZuAi0x4m1DAAAAAElFTkSuQmCC\" 128 = \"128px\" 128 = \"128px\" style = \"float:left; padding:2px\"></td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[0.010864214, 0.0, 0.2484118, ...] shape=(2048,)</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">2.jpg</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \"><img src=\"data:image/png;base64,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\" 128 = \"128px\" 128 = \"128px\" style = \"float:left; padding:2px\"></td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[0.0, 0.0, 0.0, ...] shape=(2048,)</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">3.jpg</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \"><img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAFwAAABcCAYAAADj79JYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAABjcklEQVR4nJX9Zbhd5dXGi/+mL/ftHncSIhA8uENx10LRFimU0uJOoS1W3F2LUyBAICQhCcQ92ZLtvnyt6fN82Lw95/yv6/zPe57rWl/WWp/GHPN+7nE/9xiPcM2zU7zxFT4kez/0TJRx4/7Dt190odXFWTj3bAbWfUV/aD1+Oc6uzWfQOP1Hvv9sLTVTT0MSlnHk/H34ao1BuLoXKV/iwIPnsHbJUhKT2ih2igRLDWT8k5k4ZSdCTmLf2Xtzy1NLmd9yNeMbxzHUu5tc/j2WDWwhs2uEP1xyOD1FeG3F9xijIR65+S4+/Oge5uxzHFZ+B9t3DdHbv5FH/trFJ+88yFZ9Fy3hHEO7Exx3wBW8supEZgVm4VXqLP1uHTfccjUPPrWBtjXf459Yz0mHxrD1vfh88U+880Y71eNLXPPHOqytTcxrOJb7tt/MHskJ/NLZQ926MsrE8Rx9eoyXHviF+Uf42TyisGB8A3tPqOHhRxZzy2V7cf1DQa66JUP7jxOpSuUR3SbOP2cuq9aJ3PDPi1AKPnZ1+wgGMgiX3qN61dGpdGyvIDBtIwM780yLLODK82/g0WUnESyHkLR9eOPxLygGbfadHWCvQ09hY/vHlHpn8e8Pf+DT11/h8w0XMTh8EIfP6GRi05945T/30b6jHWUkRq5hhDnOPAJ1hzE+2kQgUomHRKo6SSGb4Z0fL+HkvQ7FCvrIulv47ovtNE2fQ6bcQdvATvYdN5+XXt/G3kdZXHz4xeheC8/c/y/mHRkgny6w/37n88pH73DmYWVe/ddBnHpJPXPmNvDJ4nuQM9BbLiHUleheZXPJaVewZP2bjI5UUd2SYeP3Iyw6ZyaJwRPY1PoNsyaEuf3hXzjjhk6mB/fh06+Wc8ZRf6N1cAWUMwTyfs694kFuvHU+f7z0af7x6LuMltez73697DOth9n7tHD+iQ04cjXB2t2o2pFcf95FPLb4aOpDNyG8/vnhnjM8DSv0FC3xaxnMdrIt/Qv1bpiNyjom1no887TLFVc2oBRljpv/Lu9/dBOnHn8ndz15Emqkhs3fd3Lcb49iZHQpDk10DW6i4A4Tt5vIjO7m4lM30bdzPaLmI6SFmDB1AvssmIUkSTz7wht4nkWvdjXffO/SMm0Af9BB0/dlbdsygt40Kl2XikiSN39ZRiIS4/tH7uKOD59mn4bTeHv1c5TpRCzOZ6B1NROmxjlpUYh3Phmieb9K5vkO4/n3XueGC1/mpQ8uorNH58FnPuOSi09lyrTJnFlTx9c9a2kYtx/e6A90jNRz0uzTKMpBEtOew8scRqppf154/4+06AXWds3gt2cl2fxVmtHiJ2TiMzj04NdZNGUP/vjAcaxc9xVNFXEeemQDkYpKXn32Wj556UNOOecijj/2fIT3nj3Q68ivoE/S6B7ycfx+e7N58HvccoHa+JGMbJPYPbKU8689jo4fXH4p/sQ+M+ew9pcUWe9HDtvzJKbVL+LtT15DUbaycdcKYpWV7Hd0lAXhjWxZ/x4IDXhenvqJzZx08CIKRhHDMNA0Dct0ePmN98h7/fTyFrKwmp5Wg3ff0nngTZl171dw5CkXcMJpL1ATM7hw+sn85GykofYI1nc+yB4Lp7Bw3BF8sfVBOpfC0QsOZMai/di48x2G+3ZT11LBSK/Oqk/KLDxsPHnJY2igxMSuQxlo+IiZ+8fo2dFHUp/J0r71HDKtig8Xt/K7627ggAqNP310Nw8ceSb/fORjMv467rjoKVItQW67+XNmH9zNaUf+iZOOm8alNz7Cuddfx1OX78/XI2k2fjpEMtJJRewANpprCAwex7adHyDc+4+Et3ZjgN/sV8FHq3uZ1eSyedTC31KHNLyDi475mtHip/RlPifsLqR15N9kyqMkohXcePogb3x/FEsXb+OYw09iTc+b6PoELj7+Y1o3bqBQKOB5Hv5EmItPP5VSMY9uGmiaim15CIJA0Kfwz8efJxauZEPmCgadHUyoiPH4o2W6fygSPASuuaCSz16TcDSb9UuHOOWCODXj6/ju803UjIfJtdPJZ8dx/Mklvv1SZM2qr7nkhusY+H4lxFJMmZrkky1foegBjtvrYd5cdRHssom3RGjbvZtjZ/+NR//9Z27+7Rw+WJ/hnH00Hn+ih4MPtukp5XF7xnHEUWdwyaX3cslt8+ht/5mnbn6Re174K61tEfaZuZW3n4HTL7uLDRvfZNZ+e9C3ZoQrLn0avednTr38ZlasamX595sRm6YfiEMFcw+7hXJ0gHWDExkpNtIc6+WpP+1ALT1A286HyGYrWfXZ87jpIyl1X03c3YvnP3uIuuAFrNweYsH0qzl+j285rPYWVi9bRkkv07Z7F5dfdA4nHnUIhXwW13UAD9M0sW0bURQxHJvzLziTomFRqT5BqRTjuyUBUjUGD3x0Ils+OItKp4aDjrSZXbEHf/vkeJZ9ksbpMVm5K8D0SA0/L20nOd3DP3QINXUJTjjrCXasWUyr3ko6s4u+oQ+p8E1mXCLFZTedzBR5CnvNmU68cgKZ3gR9vr9ywPxZfLKtneYak5y4H0oyz1fL92J3m8aNf93EdY/dT7JqNj5vHlec8wx3Pv4S5m6RoN7I5Oq7Wbw0wdNvPcC/3urG3rKS/RdNJVXXQsWeJ/LgA2dxwbVTmHfADwh//WeFd/zRVzHY10/3tm4GfKu58bxW7nvydKoqC8yfcBVXXncKt9/6CocfeCINB07mqxc/YUvrWWz9ZZB4499prJyIWMpRsspEwnE++/Qj3nr9RYxiEVeQ8bABAdd1EJBRVAlcD1VVsBwHgJXLf6atfZQ7XzuA0Q4Y8Y0jpLUzLEdwesrM2yvABYefSyTwC+/+MMKlZx7Bo/94DdWOctYZ17Ot7zqGuho5bOLRlGoqsUf6+GjlU8ydVcmcSX+lN/ML69qWcsSCq3hn+Quk/B4HzLqeZUt/wE2+gNR9HSVnN821Io89vIyTL83y5X9cmpwjaNN/5rUn9ubia97n87e+5ItvbmH1GoMzLjiKdZvvJW8czCFzz6RnIMwpx5/BwZdCeafG4CaFzz++hskL7sawdLZt+QHhnk9l7zfT7kMsxunIfMckeQ0Nk6ez8pc1bGrzuOPWPm5/IMIRB3zJ39/5K1XWbu5+aDMHHNDIySfeQ2VVEssxsUyXxx/7O6tXLEfVZCzLREDGcS1EUcR1QPUpKKKEiIPlSjiOgyKAoMok4inuf+BhsqUMlRWfcNcjq2jr7ef310RpcMbx2UCRnC6SDG1leNmR1Oyznduuup1nX7sH005x2EHnsm3r++wx9WCmO7U8u+YPLAzFWanPpjffyi9LW5lYE6NqxkQW1lzHzta1ZMTneP+Nbg7cv4LfnPQXLrjwBibPrqGwW+VPV+3HZ598yUDR4eCDp/Psk1u44dqruPR3f+X8i05EbNyAL6czKZzmiAt3sOjQOQi5AvmyQUUDDG0Ncuc//sjd193Pph3rSDSPZ+bMJMILH8/3NiyxEcatJZk4k2++e4eqtMb8BR6D1QJvPlTikFnwS9rPuk9LXPynY5k79ViaGyuwC35KpsnFF5yNY+soioLrgONaeJ4LngSCiyAICEgIgoBj24iKiF/S8DwPVZVxRHBdlwcefIhYNMXS1gc5YOI4vtq4nKzmcOFhL3LP3/7AWcfczRFHVPHDymdpntHMW2++iybn+c1ZF7Prh6XMnnkWu9OL0YVxbF3/IpvXB2hYFOeqva/i+2VDfNp2N1W6RbnnVCZPqkS2m2lu2sXra//OAcfN4cm/ruW2m6/lwX+8wsLxLjs6fUzZu8SE+nMpMsj4aICjD42xact0GmafyPvP/ZMX332YCy44lBgPMPOIt3j32SUsXryUI/b+HVMOe558/mqefvhddqzvZtHBIKR3v+ZF66v57Icii5fcx47tHpGmKHOmN1BhR5kyzmF5z0z2rZ/B9t270RSRYDhC2dA54+QTCIV9eLaI5er8zxIEAc/zwBORZAHbtv8bcFHwcPEQbJdgKEDZGnsDFEnCFwjwyOPPIbgOGeUaqiseJ16/jvaN79O/LcP6bC8vXGmybOBPeGqQvCxQLea56v6nWJBMcdDhfnZmm+nfaqFHv0d0PE7a84/854uXOOSQO3h2+aXsP3U8u1YeQtSRmDdrJuVilmHb4cdtz3LcEefz9yfuI+K3WPldhIP2T/Cf17/Hv8c8vn71GM687CU++NeXPPrBlRw/dYBrn2vmoINPYEHFozzxLlx2/dXU+0W+/uArXv1yJadcNAG9VaeiMsLXX22hfYeG8PWyl7za2ld47YMS/7x+JX+8bS6bulvpbE0jGSE++LSN9ctX4XgqKAKyqHD0EQcTCvkwTRNBELAdC0kagw1BEBBED/B+zWiQZJAkGdcdy3bbcpFVBcH1kAWwhbH/Igh4tsMzz79ERE3z+sCdCL2DVDfEOfmYa1n50zvUVNlkwidT7FvC/PA0ftzaRSTWyfIOiQm+dpb2p7jp9HE8+elyzph9Crd//hp7R/ZjcrOO1tvDtpEYtfm9sZJp9tnzGKqrNXZ0dpIbGOaKOx/h5Rfnc/Pdqzh777N48L03CCTght8di9K/guNPu49FV1yKuSvM5Q9aPH65zrTjJ9M+uJ3X73iXLvl6tqycTVd7B6OZLWheEKlssWbARBiRWHRYNWK63Ma02DfMmXwh1XtGmXHENgSxnmXfedx3+3us/H4llqKRqEhwxqnH85sTDsbnF7EsCwDP85BEGQBRAg8HWZbH4MSTkGQBzxWxbRtJkvA8D01VUUQBT3DxZBHBA0WS0GSFSDCEKIpISpxtP4gsmnMb+884mx+3Ps83rZsIBybz2TP3MTm+gXee/4HfnHwx/r5ZLJozmyGpiasPKvLcvzrZb1IdVq1EkCh7HarR9rXEPgc8zWZfkY7d31OpGqg+iYKuUl/fTOuOnTxwxTEY5vmU+msYSNVx8KJaLj9tCrfceAMT9ryOWftdSqw/TH2Dy/jYPby/+G6mpVxaaOGvD9zIPec08Nk733DYMTdz/Tk/8fHrQ5x37snUeTaBUJx5h9yNsPibu7zX3l/Dt2s/Y8kLrxOsPpGfln6JpPiwHJtxLc3ss/c8BM/Ftm08nP/Ched5YwGXBVwHPE9CFF0kSfo/oQXnvxAjCAKyLGObDqIoAuD+zwOC/36XSRd47YP30XMlwtFPufk/b1NRgocvPoA3OjqIWdPw1U7i3Y8e5fdHnE+/1cpP30q0LPSI+Hfy/XcFZh8ssOVrWPqOw8+/fMGqje/w/ZIPiZZm4aYtKupqmDBjDzS/j1hEwSNMRXw8G9d+ysc/PcUJC+7jmpvuYp8ze+jZIRBTLmTj2nf5+oc+zr8mxqYlKv/4V5C9jzqQHz7K8vW3Hjtbl9CxrcyEORquT8UeDHHg4SKvPtXD7LlB2rYqCIvO8Hnnn3A9ulSFT4xQEa3CdRTGTRzPofsvoFzU0S0TARfH9kBwcV0XSVQQRRFRFHFcC9d1f4WRMWiRFRHXdfE8D5/Ph+cK2I6J4zhIooLr2SiK8t+3RBRkBNFDkiQM3eLnNb+wc20rP355Ktm5BpedcQsvvPEuyzZtY/vncPx5zbT2d+I3guw7v5GTj72HZ768jLktR7GudTN7TTiI/3zfye2XXMnznx3GQTOO56UHVjA6kOa443/Dhs3Q1fUtDz72LKefejSXnHs9hlPgl+XfkEXh5j/eSIfQyUcrlnHN+X/jivP246TDT6F1Wx2TjnqUadFr6G1dTPMBV1EY+QuPv+Nyyzn38/mHP/H2B+/SXNeLIwjkhj0GM7VE1D6UkILwxdJ7vEz3EjYUN9Ps/YtkhcS0GVPZZ/4eOPZYMDwcHGcsKwXbRVIVDMvk/7Y8EUUdo3oAsidgeg54Yw9FFDxEDyzXA0FAEsGnqBiujWML+HwqjuP8iv+gKj6eee55RtryCHOvQcmFWbwiixZq4v2X13PZldMIGYexcfBTDFFjasUEtvWv5ch95zLFP5dti1ehJGI4rsBPPy6juSZF49TxlN0E8doE993+JicedzjFUgeiUsnTL7zEq0/fzqbtw/jDMRLNcQZ39LHowDlEYiex73EtXPIXH0OrKzjvgke57obzKAp9jGuGFT9EiNTmaN8A1ZUWQX+EjvQADSnIDCvU1st0tglIioj497/9hWB1hM7+DLtHz8Fz+2nfvpOAz48gjsGA5wqIgoqA9CuMjGW45wr/5++iiG1beJ6HIsl4goKqKvgVGVEQcD0BGwFBGst+D5GiYeK5ArICjuPgejayPAZHhllm2+YNhKpkflyisnS9zIn7nML772zl8APjLGsboV99jbIRheE4RtBjsH0UL7eQvz/zJqPICMEWjNESdRUtPP+frdz9xDd89OHnxMM1fLz4dQYLQ0ybP58zLzyF3555Jc8+/y5r1q5mzxmT2b2xi08+fo4PP/g3G356ghcfvJY3X9zNq/9ZSdP0+xg3pYa2rUWWLS0gCN20rs5zyKETcGSBzvQACxpj9LQqqCGL7sEyir+E4RQQ9jsP77g96lmwTwvX3LuUl+9eSX9flpJpcfml51IumWMluFnCc0VECSRJxLY8ZFnGcRws20CWVDwcREHGdkz8mg/dNFBlmbFSx0MVJUxvLOCGYYzBkeMgICFJEq4Lnucgyb8yFtfl+RdeQZZF7nz9Eny6yNR5LkctbMFmNsvbX6Rv6XgmHu4xssHi6MNP58jDL8fVTW6+dk/iHE88IrFq+UrOPP98nnriFboKJWZNbOD8352DFg4x3DPIl199zKL9D0cvjPDdsi9pajmMtvZViF6cllof+fwA8/aeiirGee+7T/lsxXb8bi21zQU2bSkSD0YpjmaRRB/+uEVlKEhXb5ZwCPIlF1sAnwyhuB/x7NnHcdP13WxvreCS427kosv3QrcdFFGgVB7bJHWjhG07iKIMnozjjG2MgiD8F8//B4c9z0MUBUTXQxNlQMBibMM1XRsBB8caw3bTcJElFUH0xgoiwUESf8V0JERB5vRTT0NA4ZRZVzP/sFlcftzZFHcdiij2UufWcOkfzySiO9xz9R088eQKDj98P06+9kwe+VeRRSftSXe2k8kL9yVjSfQU0ujlQUaGd/HQfQ+x9qe1hKRKJk+ZQCbTT/+gzhmnnkdH91aGOzK0d29h5/YyE+vHs32jQd9ABwdMWcjTf7mIVHKYwd0qATtMTMsTjTrodpGgX8MwXSZMTTI06iEoGq4DkgJ60UScMncqf7rmHt5Y/CE5fRMofv7xxM2YVonnnn0BXAEPGZ8WQJFFZPnXzVAeE6FMy0RWRGRZ/i8bkUQV0wMLF9O20MsmjuciiAqm5VLWdSRJwh9Qxsp7ScUTBWAMakzDxvFcRFEkFg9h2zmmTpuB39nKh99+yar0W5jCb1g9MIEf3vw7c8cdyn6nXkKf/SMJr0h62040MUQ2K3D9nU8yfd5U2rvXcvTsueTKAlm9TOfwABMbFzB77xj1ySl89NZShoe2UjtOYXLzIKaSYeG+s1nxy2Z+WN/LsqWvc8EF5/HFh1vZsmYd4fx4Lj1nFrNnpzBkF09OcejhTRy4zwzmzz4KJaoyeU4Is2igeUECCmRHHMSvVmSx1Pe449LfUXaH6d4qsW35Wkq6i6aqPPrkU2iCh2VZiCLIeMiigl408GRQFAnXdbGs/2Eq4NgmCO7YRiqpBPwasqTiejaSLODzqziWjWs7/6WPqjQGT1F/kGgwQET1k4jHWfLdj4iChu06NPn+xtknvcnRex7L0h+fZ2F8Fh+uKjKqOzx99TscNPFyNvfJ3HnlW2xev4I958zjzSf+SdByicVrWbJ1Pcmkxu4uCcvyCCW3UhjoZ9eu1Zi+nYTCSc477l6KpQrOPOMANqzYyvSpAX5a/QPLN8EVV12Ar3o3f396DdffupAqfwt7Vh7IjaedxI3nj6dndYEptXO49Px9ueGcm3nq3lt458MHuO76Qxjo15g8qRnxjVdfwvVlGSxnWbJsNXc+cRe/eeBSzj3nNHwBhXA4SigWR9M0TNPEQUASXII+FdcYo3iCIPyXrIgSeAJIkoQii4CLp0j4RHmsuFFUZFEioI3xfEmEYknHcSwQBAqmjidKXPTbC7nz7jtoa99NSTfIWBqHn3Y6qeo9iMZuprtrHZMXOCQ0cH0nMflIkSk1JpffOJER8Xg2bXuA3ev7mT19Eh2dP2HlBgknKxgZ0mmsjzCuoYXfnf8Un37+DbHKHA3V09i2YZDf33gl6ex2Vn39E+t27GTVpl72mObn3288SVavoLXfxyP3nsn1t7zKEy99zvQ5VTz13hK+WVzNpsER8uW1vPL8R/zjoefZ3Zbkw5e/wtLnc8efHySelBH22h/vp+/SQIxp84NkSip9rRn2W3ggf7n1HsxSmsYJ4zhg4Txsy0VRx4oT0fVwkbBsG0F0EYUxbu04NpoawLIs/D4fruvi2mUURUFwlTGpVlCRxDGamSvkcZEIaCo2CnquwFPPP0kknsB1XQLhBBMOOJjCUJaQ6iMZktB1HUI2NWqAT/59F3d8dD+HN0xkR8cIzz66kidfeJht/Z9y61nPElYzrFixmva2HXz61XpEbDIlFZ+nkzNtzju9hsbxTez8vpdlOzYyUk4yo2k2s2fqbOzZyY4dGkkxgB3o4+6/3MBhi24hVKWgaPXMm9LIEcdE+H55L6eddxiDfTa33/5P7rvzY97/5BmaGzU+eelL7KDKkSdMQbVTCKedfLD3zdfLGc6OArt48aPlvPL0ozz96HOccML5/OPvT2AaBieccBx+nzrGoWUBy7LwXAlRZEwR/FWc+p+KUhQEwrKMhUfZsfArCrZnIzFG+2zbRhZUXNelZI8VRO+88wGGVUJEAFkh0DCbx+6/l2sffJj7b76Xgc6NHHry6fztjxegOhauN0ZLRUni/Xe/pnPHPVQtPIwH/vxvfn/FXE5ddBZdbd28+vaT/LJqO5YdZrSg4/ML5HWVhlpY0LwHRx+xBwUpz1vPfcywVWDDzjSTKmdRWWXy9Y9bWLionn1mT+DHFatZMOcQerJ9fPL2TsrWCKnENFoHtzCtOsLWnixHHLgX/bk0lt5FR7/E/KaZjJ8Ux5JKtDQFkTuHVzDi6ASCUY478XA+/89SGvfUmLPvPtRXNlPIW4QiEq+88gqXXHQekur7lZfLODhInogsqZjuGI8Wfg0oooshg4pEQBRx7TF+bhgGoiSjiBqeLBIJhVj+xTds27EVxzFRJB9lLUX/6DAzpRCvvP8cV9/4Tz57+0Fuf/p1nvvrH1A9iVjI46KTjqauKgHAKacehuvtxw8rl3PcSYN8+PkSYsHpLJzWgFkUqKqppmiIZMp9uMTwqQ6WrVJSOnn5jTU0toxDCrrsN3EPioXNpKVdjK+ZxsEHTGawaydv91i0rqvgp01fMK1lAaY2TC4PPnsHtREf27qzHLTfDKqqKhF9Cp8u3sH0yS1IgTZ+2lLN7X8+lwfvugvhtt+f5dUcU81+E6axaeBfnHvIFpITdYxBlerqWkZHunjmiY+wrTLPvvgKX3/xKYqi/Ff/8MnKmGooS5iWhSB6yJKKKIzpLCCiyQ5BUSHvgWfZ+BQVRIU//vlPTJg0Hr3sIogypuTHC1ay55w98Cdg7sTDOeXMY/n432+x56GHslddM4JfYnRgFNUPvzn8UNa0dpMMSJx23OGAgedpCAJgwzn3+nntz528/+oL/POhZ5BUl5KXIx5KkMlmqW2uJjPYz8w9QviMGIOjArKso/iiWG6ZgaxLcXAUvSRhiWkGeg0mTq1laKSbTDZCoVCgZHhMmuXQ1aqQyetcdfEUNmwLUW6zOPGMaTzx5Ndce/XF/LLmC37Y1o9w3BkzPSFQy7b167jj7gtZ/PkuXnj8PabNGcfmDW2IikDE8nj1o0+RBJmzzz4d2zHH9BJJQsLDk8cKGVmWMQ0bWRHHhCpPIKio5G0T8JAEAREP0xV54onHCIQCCK6CpGqI9XtQGMkQr4owMZFkatMMbvnXffQMmjzyx7O59ZEXePKxR/nT3XdiZwcI+pMMp0dRFJlbrjqb5Wu28fWXn3HXrTf/tz4QRRHTKpEd3sJ7b7xF3471/LKhA39IIGvY5LImzck4E+cnGWkfZcmyjey710IMY4T0oEvbqI4mjTCpYW/6smvIZlxsK86QUQBbRPGy+EMVDKdLOG6JyePm45RH2N6xjXAihCj4yWY8pqfCbMuWkbUc4mBbmR8Xf0nf0AA1ykmMH1+muSlI1XyPcEWQ446bR9WkIPnCKLZr8dPPvxAJBJCkMd3EEcEvyciChmX+qmu7Ho5lIyMwWCgRVmQCkoJRNnjg74/zyhsvI6shZMWHWTGFn3ek8WkCW7tbueL8qznmsr9y4oW/4e1H/s6hLTHueO0DAjUJbrnzHm4+6xhsJ8qR8ydwxw2/ZfuuXXyztocD957932D/X5VHVQmQqp7L5dc/hCvXM7Ehhk+MEBTCeE6RRDjO0JZhfFqIEw6dT2f3NsolEzlZIDfkglhDTZWEbicoSjKTFhapDMn86bYUBS9OMZPn/r8swiiZHLfoGDbv3klldRUNmk1TxRB1lWX8TSZ1lWWq/SnEltkKTBR5+OHf8tJHN3Hz9Z/T3lmksK4WnySwomsXN9/2J/7ylz/geR7bN28hnc2hSTIIAnrZxNBtdKuIrIxxalGUcfDIWToyDq4W4ssl3/HW+x8QiYQQPB9FwceQWEdlTS1/uv4cnnvgX6iSw4HHz4W25dx23e/Y8+Tf8vqSH5hTP57R1q2M5Id44uUXefLhB5kwaw53P/wyW9f9wpELmnBt/m/09P+6/uf7ux95kb+99QPVqXFMq4emeIy68VNIVkcIBAIU8hayEqC3mGHOxCYuumhPwjGFzq7tdA4Octy+Ucr9Lfz23EN491WdiOpwycX7cdVN/6GuooIPPn2OCU0NXHruLAbLFsXyBFy5zFF7nkdbd5a0WUIQtaB31c31vP+fDoY6DMwBhdVrX+Cc425k4aG11IwfR8yX5Nbrn6K1u5uf16xGkhTOO+s8XM/Esu2x11cCvWwS1HxomkLRslAklRdfeA5JEbBMEWQJ1R/BDFTSNLEFQRdp27mR2y4/GXPyQiaFghxy+GFU+gNc9vurOOXKawmHAjxx+WXc+szznHzEvsyYMoU7H3uJloZKTj96ET6f7/8x0P+/y/M8XNfm1ltvp6oqSmN9AaTp6KVhVn3xOMnaaj7+ZBMH7juTVDKEqoRobGjhzr+/Rm1iPL7AVmZMP5r3PnifRccewLtPfsaMA2ayZm0Xs/YIomUaSI0f5ovFBXzBPkLCOE4/9TT2nhbn7W9W0NHTgXD4EfO9FbvX0ZAMECzVsHrtbs6/uIHvlg6h1jTRsXUdviz4GkGQZV575CN0XefsM87D9WxEaazS1BQfgucgSRK6blJXV8d9f7t/zH8iKyhiELdmCkOtm5i6YD8euO8uVn2zHEep5fq7z+KjJ1/jszee5K7nX8MXSXHUQYs4cd589r34GhbNThAWU/SaBp5VYP6Eeo455jBcd0z6/R8q+v8ecXj0qWf41zN/pRCsYqJTZMuuPp557WPKnd8wOJznx682kqxySTUOsmxxHkPvJxnbl4svPZrH7t/FFXf2Mdrbwo7+8eTSH/DvN1qpqBKQRQnJDbPwmP159623KZc1GmMK1175O55//XFaR/1kh7oR/3x1iFv/ZlExp8CNz/eBavHdd0Pcd/ddjGxuZZ+FdZQCgBrB6Aty5513Iskqz7/8PLIyJloJooftmITDYXLFHH9/9BFuveuvIAqoioIebmbHkEMsqLHomKM5Yv/9mHbAidiSiptvZ/u3i7nm+mu49u4HScUqqY76WPHt95x32585Y98m2nrydJslIm6emy+/kCOPPByQ/4vT//+C7Xljms9bb75OQ0MLD93/AFdc80+qvDJrt7QjqAInHXsUdz38LuOn7cucBdXEqkpsWFcknhAZKSQ5+KiJiF6JQDjE1+/48HwKbrYbSe/nrnvmMHtKlIXzmtjWNsjzz7/I3guqeO/lu2ia0sdA/yAXnnc6h85q4cJzj0F4ZdlxntNTR7Iiym8WPY4nhaioz2OVREKBCCW3H7PsQxdLOKPgSvCvR56hob4aUdI499zTsMsGik/jg39/RG9fF35/ANsRyLk+5HAl8+fuiV81uOf2h9m+YwuiVMVV15xLb/sQ05qquODMU9nzNxfiiMN88o+HOem6v/D+I/fw2pfL2NneRipRwdVnHInn/Wq5+F9k8/+wFLBpbmqiq3cA1/YIhsc2fM3nMnFmjLWr+qmoTZHpyxGNBXj0nqv59OOvwcgxlC/RMrGKUr6SULCN44/7DRt/8fh6x53skbqXnzfdS66tgdl7Hcj6Nd9w2Y3nc9fdL/PAvQtIj0ylZfwszj7xDCor4mTEEicccRDCdZce4KXZxPsfOxx+6BTee2MlogCiJOHYDqFgECkoMv+ERXSv2sLOnbtQHPjwwy+wHJvrfn8db7z1Ju0dO3/VtRXwx5CiDSQifkSfypRxKWpj40iNn8qt11/L6m0/8NiNd3LZ3fcTTFZz4W9OYK9pMzj4wrOZE/WxayDNtDl7UrAMvnrrLdav/npMrv01o/83yzAMvvvuO+65+1G2bl1PfVMlmzZtIhgMcu99D3PPPTcwmk4jqOAPKWBYVMVq0TST8U02+++7Nx2tNs3jGxHEIB0bezn7zD8xWtpCoWDzzrs/cfBRXUwfdwjXX/86IyPr6R3y4UoCNYkYp53gp7K2j9beffn4/Vaw0+x9UBzhr/+Y4L39xgg7N6eJ+mpYuN9kvvv6R5SgRHHYQAvKqEGFaDiCIHiU9Cx+OcaXX/3I9s2bUJSxQDiWi6SIVO2xL38493Jue+pukqEIf7ntNtSGaax76zO6t3zHS489TVb0M5Ddwfhp+9HTvoOwEqBvoJeW+mrW9Q3jDwSJovDn689HFqX/QsP/W2aPFVrg2g777r8fuzq6uPqKy+nt6+btt9+mkDd49NFH6di5kWdfeQbH0QknfZTSHtEUeGYQRTNpjEeZ1RRhbZtDbTLB8UefzMb1qznmlL0YGe1l1y6YNFviy3fzDPWtZ0fbGi657DjaujKURgJUJ0fJlTPsM/8mlq17nc8+7Oe04w/myzUvItz41EJv/8RJdPfFuP3hSxjuA8eDSFIGQ6NYLCKKItEYmKaMJVi4JQE54fLC818Rch1KlOgdldnvsIMpZktMH1fBxVf9hf3n7cFzL32BZ+pcd+WZ1FQlOffyc5l34Iko4Rif//M2rn/4UW6+9Hd8veoXVq5cSTiUYNEe4zho0X7/a/YBgGVhKzJ/u+NOnn/tLebuvSfffb2YUqFIsVgkGgty4WV/wDZ1ikMdfP71EpSwTiFtIEgB9pkSYM0vcUraLgolm/de+Cc/L/+B3btXM61hNoFkkJ6e1czb7wDc8iy+/OYfZEdTjJswjQV7JVn21WqWb9jEvPl5fPmDidQkKUlb+fjLXi4652B2bY2i2z0Iy9f/xquN38UT/7ia2qlzWLXpRb56L81ov0Z721b22HMm2WwRX8CPFlIIBl162wq88vJj7L3gFHqHenEqakn6NVw8rr7qTjYsf5c5Rx6Nzx1hwfhjaJgR4Mf1O6hNtODzB7lgrwnc+v5nPHXLjUyYtT/nXnMeezbPoCEWYP6eUxEVGUlW/9exdl2XX5Yv4dIrr6NrYAhF8qioqqO7Yye266CpQU45+xxiwSidHZv48MP30fw+aiuipIsjDPQZKAo0tTTQvbuPytoI5b4iDz50Ov+4+1MOP3QKc+dfxmDuUzxXxcyqxCsmoYxbQr1wGQedGOXjV17jztuWctJx+7Oj70c2LO/iz7dczMdfvMSX3xUJhacQTpQQWnf+3Zt14HW8/8krXHTjeUxGps/vcszBQdav87H4lQSNzTFefX+QO26cQF1DgjlziqQqC3StW0Ks9kou+/29CASwbAvFU9iys42F+x3KMedciGTYhAWPaHMjcxprGLJgnxnjuPmufzJ90QIevup3vPnZT5x2/N4I7v9+UwSwXQcRgeaaOGq8lmxuBMf28ByHcDiMrusMDQzyx1vupbB7LSgW7332JZocRNAyROQYtijS1jZIJKlgFzRSLQVUt5JwwGDmxEWMaxIJRSZQF21k++7XaVtbz+Lvv+HC30v4tQUkgjNY/8M3BGLT+GrJSm68eio/rm7j5ItP4/PP/0MqMJ6G6XHWfr+G4a2zEP7x4p7eG9+WiJptPHpnLx923E549Wa0BUewc+vt7DunmgMnfMfUg8fx5Us/0D7we8Txd1LY9Snf/+cjuktTyPT28ePX2zBdB9UTQRQouwYfv/UVq9s78EeDfPLa29x+zSVEJkxhZCDNqUfvzyPPvMXGX9by9FN3j222/x+Wi8fSxV9y6R/+SHlkmILrIXkunudguw6SILLo8GPwhVNM8Q/y2fIlDFs6biGHFpYpZAU8TyMU9KPrJiPpUZJJjWJBQvAkLLuEYwY59qgFBEJrOfmMe/A7Nq1bF2OZLjt3ZxnY2cw+B68mN+BjyfIMp556GF3bNxMKq4SaxxMOhlj/zRa2Dizj66UF9p0+DWHFuxd5r/d9yS3HP0fVxKM54YhGftywm2nzfdiVDpOqLJY+4ufzFSs4/uy9yJVELry2nq6fh1i2I8MZF0rce+4mLrvyBdLD/bz+zstICAi/vurg8tkXy4jU1nDofoswcj089MxLjPTnuP+vl2J5oKq+/0/BBjj/rFP4ZsVGnF/bV/5nw0RwEUU49sSLaaiUWfbZq6zrHsFxTcIJkd52m1QcTBOC0SCOZeILioiiQTbto5zVUfwKsmIxflIV9cVRaJzN7NmzMUcDHHnkFF554WkOWFRJyU4wOhAkFBtFCwg0xuYwbcZsClmZj795nWyHSCBs88H3P6FbIgGtH+HWvyreuy9P4PxrTiTZ+AGHz/yREy+vRhfH40mDDH1bpnEfAykkUGyPMBjwKPXm+OfjfyXhS/HUc9dw7Z1X8Y8rM9RMqKeyOsrKt9/hx42r0P0yPn5lD4LALQ89TUUyyVXnnjJ2SvT/geZZloWsKNiGycGHHUBnVz+GYWCaJuACIrZtks/pnHPRXxha8y11Uzy+Wr6OYh48tUBjbQ25cpHCcBHHc4gmQpR0C9Ny8El+Muk8Dc0Ku1shEIVxqQDdo0VMAy465VKOPGoPFBl++O4lyqUI41tmEQ54PPfq89z6l+solbeSHW5CMP1Y/n76en5k1aoe9t/vz6z47kn2PuB4BPx+76AF8MrrS2kaNw9MP/96X8KfNqieJPP96v05aM/TefPTz5k57XOM8HxWfZ0mldrI+iUprvjd1Tx0551sbxM5/6pjWbN2gHEtc0jEo/zj1j/iT0THaKPnIv0KzbZtI/8vN8X/yVzTNLny0ov5dulPGJaOYRj/PbT2PAdBkGhonszsOfsyOd7B31/5GMdU8LCIxlQiqsLggIAr66gBBV/Qh+sNEzQkQoLH1gGH6oYIuC7dvQb1VUGKVhlP0CkXFJKJCfzrvtuYMq2CtZteYPf6PCP5ITDr+WXjCh791+N0d2xgR+cA377xAfMPHsemDTZlF7oGykTNVvY7+kjEgxYu4LBzwzQ2zOOaO87koj//luvOsdi2Zgq/PdnEp9icdsTFHLBXmHc+ryVhH09l7SAdK+bjq1L45O1HcJMh4oEgHX3f09G2jlXLPqN711YefO4tzrzockq2jiSIwNjnfxtsAMMyERyBAw85lMU/rKJQyqPrOo4zZiq1LAtBEDjroqs59KgTGNr6Pg889RlGWUEOeSSSfmTZh2HlkEJFLMshHNAYHBpAz2vkkOkVRJIVASyjTHU0QKM/wHB2FJ/fwrUVkskg3d1b+fTL9Xz11Q8sfi9AvPo3zNl7FtVVUQJqlO0by2xat57OTUN4oV72PPRIKsKQCkUpDnVy4hlH8tWH3yAqke/5ZMkwf7x9DnKxlQV1Arf97TD0miZOvdvhsgvOJx+FTT8NIwz3s3rzN2x/eYTvl29h9ozDGXWznHrI2Vx6W4BP33uPW2+4jIpJ8N3KJTz//JNIsRSzGpu57da7xvTz/1Vaj6GE49mccuwJTJk+nq7d7RRLmf86dgUPXERUWeO4868kIWgMbvyCnzYWiMc0irkipm6QHsnioGBYKooTQsKgXB5EdCVM26C6JkFNyMVyFTRN4+etA0wa5xAMxrEyQQIE8aQcobDGp989x0/rd3DyFQswRvvpWD1MuLoFwlXkCr3YYowpswzqquoZaOth1NyT6poE/qo0eAr7nTgd+YzjH+KmR//Ij68uZdG8A3hkx5O8+YLEn37Xz9aBt6mUz2NosJI/33Qa0bqF1NdKaIfO4p4LjuRvL97L1rYgleO/JhLSmDL5QvK53RTKIfY/wqK/YzffvPIsr7z9GocdciieK7Lyx5XM23tPFEn+f6Z/AuDY7LfoIPq7e7CtMqb1q3sXcD2XsmEhCyKnXXoDNWqOl198FNOWmD4lxta+HjQflLMOngCBko2i+SjpJfxRCXCIBBzKZUh391O2BQzXIKR6yIqfFQM5wkqUkVyBeCDEUJdAqlIhPVrgow8+IxBsIBnZSmPzJIaKRbqHW0EIUlVTgeP08fPPPvKFEVJJix9XbEPPpVi1qpcDj5iLMPOoiLf/lNl09WtsWrWck04+iIfu/pRDjm5m587d1DQGEJt0+jcEOeioCXz+ci89vf2ceWCKjnycI484kP5gO6nCMpb94vLY429z1PHnM9Be4m8PXMqDz7/Om/d/wEMvPcU+Cw/kkgtO5aOPF3PBRWeg2ALer8AuCMJYK4rnsmDOTAaLFrZexDCsX3XsMbx2XRdcj3zZ5A833Yqb3soLz77J+Kn1dHVkUQMurucxvjlOe3qU0rBFNKISDai0tw1TUxEkW3JJ1crs2logGlVR/VAqWvgCLroO4aCCX1ZJ5wxcT8V2StiGBKKH6PnQ1GZOOW8h5Pw43jK6OgIce9Q8bLGVUn+KlcuXkWjKs2SxwB5TJtHV18nMWRG6dhcRW7/Icf3VU1AiBj88OZ6Hn/mMUy7Zm4GuIWTg5zUlkqEA5593DyveHeL0U13+/PB+/On119kldDD+6C7mtpxLato5tGWaOf/ys1m48DAWnugw2vcyj/9+fzYXetm0fjVPPfhXbr7z7xy+/9489MSr7Dl52n/bUDzPw9V19tprX3ozNnouh66b4HrY5v/ZbeF5Hv54imtvvp3Wn5fw2lvvEEz56e7oIFGpEY0YBAOQbuunxnQR0TGRSBeKVFZBX06modGHVZDxqx6ep6B6cQTPhyYlEC0YGVRo7S4TjoaQXBEZEGyX6RPiOJ5NQd/C2y+/wY6Bn6msOpjenjZeev1HJGcchj7EvINbsK0kB+5fSV/fVpCL7LvXibTu7kT+x5sncec9H9GfHmDcSXD9rafwz5uW8MebTiYmruWxDzex4fsklx/1OV9VhFi7qhptRx+ff3Eipxx1Itf85hP2OryNf9z/KL5yDVWpJOt/7mHbh+O4+61WXnzqIhqFN9hr7lSGhv/D1Ak72dK9lWP2n01Y+jMtE6cQFECWRdJWGb1kIHjemH3516wWpDE/o6wJnHDiZVRUV7FlxVvs7F1DqSwQVjVcJc9QOktTcxAZh/60RsoXRMp7hDwBz2egZySK2SyZLKSHoT4ZZChfRFHBSJcYHNQJBgLEPA/Lp1HIOyAquFaIphY/A6M5amoTdO0eRdUClIb66Muspa65kUSqBU2rZv9D5vDCC3dTLgus29QGTowF+zaw4udP0YIywgFn7uO1Ld5MWigxcaZGKJ9ia8cIh+2ziPb2YbZs24rgz9LUEOXCs9PYvnv54bt7eejZVzjljGs46apq9greydEHSfz7kxz/+eFRbrxUxAs/x5233MX6rWtID++gUBL4z+K/kTVf4Yh5q+gaKrN1/fdEko0ccfhBqIDj8mtLiwOMWZkRRRxDJ5JIcsqFVzDU38Xnbz8FgozsF6mocGjf7hKOCriOwpGHpvj2hwy2U8axBWJRB89QGcy5yJKG6xZIRiUkJ0k0pdM7EKBYHiQckSkXxlpnbNch7PeRKRaQRBnDtJFRUGWLGVNb+GXTKDUVITzJpLGultb2Xs4++xhGhwrMGjeLsr6NxT+uIDfoB3eAyvES6YEwJX0EcWaLn+TsImbG4vqzHmHOlCauvt+jMHs551y5C1GL0jShmROvDnLpn9p58pubOfGsU5k99VTm7VXDJ6+08cb7L/DAa49x2hkn89oHS9lUvoeJLePZ3vUFyz56iwOOC3DeOVfTpv+bA/f8kiVLvuSJp/emvmEua35awScffYXgD47Z4lwXzxP+CzWubRNLNHPRxb8josE3X7yNqCiEowK2qSB4QcJRj0gkRlWDx7crurFti0hUwx90wdQYSRtjkOQUUD0fRVugaA4yOhRhdLSfgC9IOmNSKhuUDQNNUVFVmbrKGgRPQBQgGPIjqH7auvqY0OijaGSwHYfO7gFiqTCrfhpi5rwpVE7qYXCgRDIapKgXQdFIxSciqINUVtQhHLD3Pl5tPMMeh52E6dvKx8tWYmzv5rjf/IG/PfYCdck8PQNxWvYpc/j0BEuX5fnk28VceeLvaJw8hVfefgs7JyNJNp4Erg7fbjuPsw9/n/SgydPv3ERzfJRLbniaCbUxhjNN7No6THZwN/FwLc88+RrLfvmJ8eNb+OQ/3/Dlv99A8/uQPAXdLnHd3X9HGNnGzl2dfPfDF4SCGoZlkk3baL4xp244GiQ9WiCXdZk8uZr2jn7qWgJ0by/TNE7ALFdhGMPEE2MSwvBAASUaI6LA8KgBrs1wxkSRoSrpo2SMWaYtFw6ZXENr326KUpBsHnyKC56GI3noRhnHcgiHooxvqaJ/aADFjXHDdYewaVsXG9e3U1Wtsrs9Q6pKZHRYQLz6/C2MNMI3P37GPo0auxf3ctDk2aQzEc4++xD07QGm/CbN7w48hor4vpj5POeefjkH759n9Y8rOOu8Km6+bSbFdJBxcyfx2tOPsk/Tazz20hWYZojbrn2EQMWlFFtjbF0bRBA2c8XlxxEIphge6ueM847k3feeY/vOVmbPmsK9f38Wp1hCkFP84aob6Fj5JZ+9/DjLVizHcTzyBZ1spowii5QLDi4Otq5RWROkZUKIQsElVakRDhmkqgV6drt09/eRiqsMj3qYrkcu56FoJkMjIrphIQhQ3wiqqjI8aqKqMj7Fj2uY5Ev9yKFqclmLUEimpqaOYFikMhhEcgPEfUEkLHZsbwNbIp0dZCjnozpciWv6sEaSxKskAmqUoZyFuGo0wM8rtnHi/iU+evFt7njlVKZOGODJ+wus+XIV/WaRdy//DreU4tW3v+DSS6/i5EUzefj1NsL1MUQrytqNwyz7ZRla/w5e//hr8MKccsjfOPGkNIZYQ3VQ4ZDz5jGU7Wbl1zq33v44bkFAC7skqiPYpQLPPvoIa5auZ9fGHznhpOM45qgKSiM/kIo5tGV9IGXGmgE8G0HwMEyXcMwPHnieiCZJZIclBLGAbbkkAhWUygKxlIqiaXR22iQaCli6QjQaoJj3EVLy+E0L02+SGRAxDQdJUhkazhFJBKmurWLrcIRMOkuxaGIYFt3dnSDKjDgmyVSQUDTMzMkTsV0wPI/aljq++nw1TgBOO31vqptLpNwwo0MZgqKF2Lqzl389/BLfbd7O9z+FkNLNDMfHcdHZbzDnqJkccHo9P3QvZt3ASqycn0OPX0T9PJdLr/WYcrjJK0/uYOLsaTz4xNmsXNfLV19+x/KVf0UKhBnoPopQIM+UKQfz73e/JJ6SCAXBzUOo0k9F/SQ818KyRggHTb7+/k0+/+LfCOIoRkHjn88u5vk3v6KgZynkLHJpHREBnypgGGUcN0tVRQWum6enL4Pi1ynlRUo5i83b0xQzDtgqfsUg53gMd4hEfVAomTjGKEMlEzERxMlJmLZHJOLDsA0qq2K0t/ViWRZNjSlcUaG2oQLdsKipTZHJlsj0DROJRRFkm62d/dTW1xCNBBgZGmV712YUeQFt/XmSsfHokoUnGZSKQwiahPfsKzdy5smnc+hlc1m7BvzAUYeGUJP7U+4bZN6M7Uyb+x1HH7kXwojLbXc+wM1/uXbMfV8WidYFiTZWoBoK2aEhdM/DrwUYNydBd1s32XQOVYhQPa2RXas2EVA1EgkFHTAKRUb6BRpqKkkX0ojyGCd3TAs14KIqsbEuC4pYhomhi0wYL9LV4RAK+xjJGDhFEc/nocke4ZifRNSlL11EyEsQ9uP3QBJLZAsh8uUiqqigyDrFfBxRyxPSRKLJBG1t/dTVVtM12E9F0IcoSdgeiFIU0c1SKJkIokK5UKKiupZysUQgLGJZLpUV9aRSCdK53eSzNolIEyecsD/5chvDHaNkR3to7zUQPvzsIW9BQ5L7n+rkgDPXMdRxOA/862ZOO3gRb//0C2/eEmTT5iKX/mE3owP9OJ6HbnlYjHFkUVCJB1UiWoimOQ1IXohSuhcBFS9iM7i7QFhTideo9O3OE4sHcSUHt+jiSCaKEkGVPTKjeRIpjULBA1dA86sosodlGWhqhKLukqzK47kyuREBzxAoWhZNVTDYLzKSN5AkEGUBUZZQBIlUWMHQwbIUTCFN2RSoSIQolC2icZv0wJi7NxIJYIpFVL9GZsBGVXy4FJClIIoqYBQVQlFxjM2kM8ye3cTGTX3k8iNUVSdxbQ3HcwkFJARBwiopxCMyx514OYefcBiPPXQ18ZDOmuUdCAdPDXjNCxew+qfVXHzbIVRqB7F+yzcM9SzltXcKyHk/w4U0udH0f0tw61elzvM8BGmsoao6kWL8hEnIQRvbCDA4vI3KRJxSuYxtiHhSkZFBl7rGALolUkwbyKpNMBhiaCBPMulnZLiMoIhEAgFExURQTXxykJ4BmYp4kUBAIpMxkLwIqmRSMBw0xYcmCsiqSK6YwR8MIToig+kCAVlG1kIIgkGxXALXQ5J9lMs6oVAIWRawTQfNp2A7HoIE4ahIbqhAvuSQqghTLgp4QolizsLxYNz4Ogq5PK6tofg8mpsaGBroprpuAp2dHSiajF4sUbDy3HTmXVTPTTBSGuKFRx6goaoZWW6q5Nsff6Rtm4USFJCEb4hX11AV86hviVOhxOns6cQvB1BlcBwXy7H/azFzLWFMKgWOPeYY3nn7Ffr6d2HaDrapsN9RC2jduJFYKIWTyOA6MsGQiuC4lAoiiuwQrgBF/rWbwnGxHZ1YxI/tOeRzBnXVRVw7Rqa/gEuQVCpHz5BEfV2IoeE8oqySy3vUVUdoa81hWWMCmFIpodhlgiGBkbxHQFJJhELkBAvJdbAsAUlyyefzBIJhLF1nuORSM95PbquB4Jq4qBRGLfxBP5FYGEnUUDUbXTBIJavoGxxCQKOtrZ1gSKa/uwtZrSNpTmH/c+bxzaefIfnrmDpzFkbeQezo0vjsyw1ceGcIR4fLrjoFubSLbMnHsQeM56clOwkrIcClbI3NqvJ+Lb0ty8KyDHS9RDaT4aTfXsPH36/Hs0xUz2HYzHPtjXcQiEpks1mi0SilXBnPMikXbOIJleEBKAwKmEWDYERBkASKpkOlmkcSA7iuR3cH2F4Gy7VJpFRGMn5sy2F4JEc04eHKJrGohCzYKDLE41EiEYW6VAjNJ5MM+RBN8GsKI+lRPNGPLYpIkoxpeUiyhqnbY93SjkfrxhINEyMITpCgXyaSECmWy2iKQi5fIBAKkojGGBnNUl1VR019BYGQzdDAMGUL6msdvnppNju3bcUo9ZEZzuL3VRCLVyDu2Lqdu+68nxrrXBYcPpGPf/iOshNi3LhmXnp2I7hg+n7tt3ctDNvAsb2xtm3bwzDyWJbHhg0b2PzD/WTaV1M/uZYp+09k77nTSUkKDTVhtIBBf+8QoWSYUl4mkpDRzTK1DQESyShF3U8y6VJVE0ATNboKIIsGruiQTPqwDHACETq6TfJFHdGLUC55ZIdFosrYgXDfoIPrgk81aKyKsGXXECXXZVt7CUEVKOkWAb+GgIQs2ziOQ0U8Qiiogmhj6CCpQfZd2MzQrgxOSEHzBSiXfNTWptASUWRJxHNtIlE/NdVJHLMDI5ehq3sYzwkQjaZoHxlGqAohYtHesY3B7kGSMchmBhGmH9TklYsimaFePDFI3K+iiAa93VlkJUZDuIJvfllGJjOK6zp4joNju1ieSTgQJhmr4sgj5mPJU3nuib9yzdU3sWnXSiQ3xryFft59cwNzDmog2wXIBkbJRC+7xBM+PNeHaacxyz5sS0c3XGJxDzwFWbOQ3CClvI7hOlTUaGTTFgG/SzQapKuzyKRpEXLDJZSQilGUGR7JEY/GGOgrkIopSDKk8w6xgJ9iWcdFwHIsUsk4juOQyeYJhQKU8wUUfwDXhlhEI1GRomyUSI+W8EybtGSRUoL4fRKlskWqIoLtGFTHptA9/BOuF6ac10mkkuTyI0yonkbrSpfzDtyIs/B0hnrbMDxo7xhELvX1cu29D3D/n/7I6PAoyGHm7TWbkcxqKhur0K1+fvz5Y0486EJsCdKZAppgYPsFnv7L+SQTMUx/I+mRIuPHTWLVL0uJxBX00jDvvjrClZeex8juLMW8TTxVgW2ZRKI+/FoQ0yrjDwWQohKDwwZaQEKWfUSTHgJBynkP03WIJlTSQyaaohKLhigUDTRNome3iSRp+FyDgX6dWDJIf1+GVDLK8HAWf2DMhjdaKKDIY4zKJ40Np0xVJDDNElgmWihCuVQEz8FyfEQ1k5ASZ2pTBd8sayPpU9BUPwIePs3FMotEbJn+oW+JRSfT0z3I0EgWx3PJZvMI4jZq90ixfaSKaO9ulny3noFMCVnWEGpbol4oWMYVoxSUME7vAMkJcTq78/hcDw+TodZdHH3eP8j0t+MURnjy9vNxcpuwJIW85eOelzYz0N3G1KmT2bJ+OcXRDH0jGXa17mDBvPmoQYNAwIdedMnkDayySTAik0z6GBoqUD/eT1+bRMN4CdtUcGzo780TjfjJFQqEAwEK+tgxVzQg0rnbwK/IqKqPMnlUKcRItoDsSehlh2BIRlUUCvky8WiIsmGhySrZUo7G6rGTe8dxQBJxDJf62io2bW1j/PgahrNFokIAXR/FX91ALCSRTg+jajKe40dVfAT9JcRwmZ2rXQIhl2E9Q1WigqyTpiLuI9efYsaMAN9+v4VgUmVWbQ1NzY3s2LwZsaouynBOxMrIBA2NUtGmfXUvpd1ZGmsrSYSj+OR6xtVV4NOSTJi9H2t3dJG1dMxSHgrbIJOlWB4Tgp555Q3+dO9j/Pujt1i41574QmDqEvmcgS/kIUs2ycoQkqaSSRcQPIHCiEtFZYLhPouRoVE629K/jm1yEC2QBYhEQuQyBoOjNqGoD8kXJF3M45gKIkGqK/2Ai6YK2LZLvqiTTMUoWzaCJJItFaiMxSnbOqoqk6qIoskKtuswnMmiaQqJaAy/JFPGwgtHmNxYhyiKaJKMJIWpipqMdu0mFC7RvtUmWaVS39JE1BfCF3cJBiWGO2VscZSNm3ZQVxMhmfCz/IeH2HPBLPbabzpi7Zwkpq5QNbGPeCqPoBsYjksiGmZ3Wx/9HSNUNlRyzgkH4CkOu9q6SNY0IQouohhCiTQxY2YVguCxfddW9HQ/yWSQP1/7R1yvjOWaNNYFqEjEyaV1XEtCUWQiahi/30c4GmJ0xEHU05T1EsGgn7pUDcGoTnlARwmpBDQQDBc5DIGsRckqkU5nCQY0BNkkECthmRAK+lE1EHCoS0QpF/LopTLRcByfKKD6FAYG08iySFRwGc2UscwxtuV5AkNDaVyxRDBsUledpG+4F9tyKZsGfknCsMqEqiNs3upQLOiYlkJjLIhgizhlgXJaRfGbxKpq8fmaEH1lku5EJlU+wOWX3c6ifY5BvPCwG5g1S2GkNYxuBHEkgUTch1EyqKpIEAiqCKJJTcrG8xRiGmzY2o6rTsEONSDkWtm2dQNqUCNVleD7L/9NS8LEFiyiYT9xv8bQ0Ai5zDCyKKBpHpJk0Dcwip7xGO4v4/cLOJqMVYbBHpNhvRvHThGdEmJ4oERXv81Aj4OnBwjWxnAGwTBdhoZ08BQGessoikDM7wdPIxwI0D+aRZR9aJqP0dEhpGAA0xwb+xTy+WkdGkUFRM/Gdh1cTOSAS11lNWZZZ3h4+NcGMZFAIECmuIOiUCJfKiJKLrV1FUSSflZu2U1eLxBySqiWR0yOMdi2jeHhdsrZIGV1EKl5B73to8yYPRHx8qtvYvPPo+R18EmdNO3hoQVUqqdpDPX34DoishiioaoKxzTRkfngk28JRnxIVhonMpczDtsTxXFRvShbtvVyznmXYNomfi2E6dmEQ1ECQYVoMEAsCJkBh8oaH7pnUFcjIwph+nt0BHHMNVvKgujalEZzNNUn8PlEfKpJyDMYSedINYaoTAUIB32UChbV8RgDPUV2dg8jSSJFY2yUk14qEQ6HcDwPRZAYSmeJxaLs6hxkyviJHHDIAYSjUVzTIRpTKIzqdPb1EAj7KY6MDVwo5LMUcn1ksuA5MRbMmYGkeIxmRsiVOrBdk3Mu+A27hwWqKvzUjK/h6ENmMXGPCq6/9iK2bRrBVYOkqoapn7gXolDqJFAt0LJgIb/8WMYfnoqjOhiDNkoAIvEASAWmTtmTM04+koDfT2XjXNb8/AWW6UfL/ELjpEaCgRSZXDcbtnbhOAZBn07ZMjFLEpgaPh/EwxEkSUBWHPIZg1AoQE7XKRujKD4DvyxRyBeJxnwkkybhkEBv/yjFooMn+yjrLp4A/QMFRkZL2J6Oqvrp7hkm4PMxd1IKTZHwbJFEXEZQAzjmWEFT0ssENQ3PsqlNBcnm86xdvYrR0Sz+iIyARDCq4rkCJTNHVX0LZUNncGSYaKIRWQuRz1is+nkbjmuguzr5UQ83UGLx199R0SRRPzFKenQ73YXtzJhWwb23/539D9iLajnMu48sH5tNcOMdt3DWyUcxqzpBNKGQUl1qZ4+jYXwjsuKhqQFq6mL4ptQgugOEwjGCQYGla2Q8oYCh1AEBZLGIZ7l4kosg19BcK1OdklBlAcMr4pkunjgKQE1lGMscmzTk1+J4tkwgEEB3IBSBUlmnu7fIEQv2IpqSiQUC2IKJrfjwqWOt3Yoi45d9mIaOogUIBgW6hkoUMzn2nZlAEKNjE4ZUhXg0wORJ4yjpBormI5vPMzCco2iYNE5IEPK5qKpHJjuMqgSxjCD5XBuSbBOKKESjEQRLQxQilIwcZtlDtFVkxYdqC2SzFkFXo63VQ5ST7NrsURhuxHWDlHOd3PrI63y+8jlsV0B89vHX+OjfW8ip2xH9dQxno9SUFcqlAlXJCTQ2RvCLNdREFa66+GBE12Z3Rw/NMyeBmUNXBezSLvbZexqi5DChWUKSc/y0qUAuq6IgMX9qAd2M49pF6qstRNXFNB0MA8qOw6R6jXTWIp4IgiCiaeA6Iq99/DPpIZtQPIQ/oOE6FkOZMtF4EAkfezSpyHgYtkG5/Kvyl0yyudugpFvEInEEx0KKqPT1DpNIJZFVGdNWMawMkugyPDhCJufgIeM4MqGIgmEYILsEFT9xfyWtO3fjYTLYP0RFbCLBYARNlnAsl8yITioZpb/XJZfLMXfBZMIqbN/4Nc0T/cg+k/uuP5BXHn8Mvb+EeOIZUzn2hGqWrFzLtNmV4OugdvIe1DZUkIxDTb1BMAySleS0s+5gR1s7yXCAxR+8j1voJFNUSI8O8O933sBf7KM0uoMpVQrNNSnKZoGWWof1OwXGVbog+klEmtHEALYLNTURFLtMS3Mt0ZCFrNqIsouhg6ZpFAs2oZRM38gonucRDESJxyNoapGyXaB9yEINBHGsscPnfKaAZ5Wora0l4VexDR3Ttij1mmOe9VyBYjlHNCYTjSSwbRtNDiLJKq431uLYNzSIJGo0T5zGtrYB0tluhDKoWohpU2J4Qp5yWSAUSeKZBk01VRQLFvFYFWdefDZLFn+FqATYd9Gx+DQDyxjkkMMPJKp3YfevRty6cxv77304ldZkJk+axpm/+Qthy6Q6MR7PVUkmW7jitycz2Jlnjz3CrPr4Sp585Gr+fM+1/NQ3Dn9qD977z04ifj8tM8KUpfGk7SClEkyY5LKrRyWRmEDO85MdsdnWnkWRivhFEdczGB61+G79LgZ6XUYGLVKpCD7Vj/TrWNXCsA1OCdeBfD6LZZcYyggE/H5ypoLngd8voQYkREVA8QdwPBvUAAge4ViUaFWAfD5HwSgTCEmUyjqWZSKrCqLiIxIQcE2BsuES8ofR5CAbV2wiFowhuTaJGoNiLsP2XWmGR/qZt2eccDzAnFl7IGh+bKmAEnR47pF/UkgHKNh+Xnzxc5LJmRS7Nf78+4soyRv4z/PXIAbkfbj+6keIBlyaqxIoRj+X//YybDLcctMNBIoRwr44//rXTZx0TAMrVn6Nm/uezWu3srttNR+8fRuBVJaGCQ49/QKeN0wyaVBbV0kpX4koS4yWRjHLaeRAgkhMIp11CEbClHUHQQQjJxOKioiSRc9unXithmGYhMMq4WgQ0fFQfQ6SoJDNOsTiUCiU8SsS6q9T4BxbJBSLM5TOs7O1jVLZIJMv0dnVSzadp1DQqa6tpJjRUUUJ2zIIBoNkMyOUDQvFp4AEPkUhUZnHES1MKw2+GiKRRvaaO5GayjDJaJjSUJmR3l2Ma5yHhszCic0M9fQxubmJ6uYK9tivmVDQJT1awHQa+OW7TXRu+Q+mqyOe87vtnHHWTBTVo7WjjVDE4tmXHuWS8y7hrdeeIjmuiYIukh4toUrj0EtBBHeURfs0cuGF53P+eZcSCE6hbEfRtCp86jiGszXoJYvqyhiJmjg14RQ5U0T2Kwiyh6jY1NYpFDM2murDw/qvjzNRIWMZMobuEQz5KRd1bATyORMEi8pYgqCWJBENUzYsFi1ahO25hCsUtJhDpDJAZTRIbV01rmdTkaqiUMjT0pLEcw0sx0bWlLFh8rrNhPFNhGNJRodzaLLGSDpH+y6bytokFbURBgf62LhxF509w2SLGRqam1i7YxRZsFm86iP8dQLtxWEOPmQS4WCCQrqP/u294ESIxhV2rb+dhtA6Fr+7HccVEeYe5vceffB9Nnd8wJf/3sHfbroWJJmuvs00VE0klw9TGc7xzfdrOPa4Waxc24frpXG9CAvmzuW7xd8gOkXc1Gzat23k59Ub6eroxNQFGqYq9HbYFEp5nFIeQVSRgiUwbJJRmTVrTWQ/uJ6NZTpoARlJEhE9UKQQulnGMCx8Ph+uayNIoEohHNvGci0S8ThFfZhyVscVJERJxTJNYtEQHiKFfBktaBP0JTH0EpI8ZvCpTVWiaDLdPZ2ceMrRvPDCm9iGQG1Vkq7BQRBcIqEKwn7IFQ2iqRiKl6ZYCBIJyviDIrJax87tuzj/7ANYuu57ij0absxkXMhBDs0jGLORCzZ/u3J/Nq1/js0/5SnnbcSjDjyYffY4iktOPIehzkF27uoj3TuAUwwBBYb617NqHdS1TGfDFonJ4+oJ2Eni4Yn8vHonM+fMZcacY6jQLA5ZMBt/0CQYsaiqM+nttEDSMMrQ2FKJ67oUc0UUVUL1JAgaFErG2G0ngQCWDvFoEsuS0M0y5ZJLIqVgmEUUzUUSx7JeUkQkTyeTLmAYNoKqoIVUFP/YuSOSjO1AJO5D8iIoqkAoEqSmrhZFltGtIr0DvZiWyOKvV2LbMsavM7yiIT9WAaKxAJIaYNqUBKIuUBoxOen4g0ATGc0ZuOlhUlGRj7/4mOyQjBBRqQiGWLFDxfC6+OSzVfz95mn4tVG6fsry40aT/6wxka+/7iYuOc3Hm19atCwIMnn8eLK5PJm+PjZszCJjY9smhm1Rzo8QjDUQaW4k5Utwz92/5dQzz2P5km8Zzo7gODambSE4JWxvHKX0TprGNSA4PkKeRSwWIxYNM5rpIleWESyZWFxCEkQc18S2PPp6h3+dCKdgOhajoxAKKWTTFk3NNZSzOcp6keqmGE4R8vkAgmMjCiqOVsafipDenUaURWxBRfMplIoWAb/McP8gpuXiOQbZnAmCQ74gIojQ2FBLSS+j+kPM3aOeUDRIZ1c/K5ancTwbn1/kjXcXM2/2QWQKa5gx6SiWfvNvfBWz2L6pkxOPjdO+YRRFtDhwwUTav4+iOhP48I27WbLF4963X2Xrig8RnnrlFu+151/lkJNPoXvNah645QY6Oodo79rGYYcey89rFxMOz8ERPdavX8/WX75i2fJd1DVJjGQsqqrCFI1BJDuCjUdVIMlAdpSRdIFQyIfgjBJINBIIBOnq2k4x6wdBRVBsdncWcDwXxxbxXIFAUEPXDSzLGpt7qCj4Q1AoeiQrDSwTYpWQ7hWoTFVR39DMyEiZbZu2gTQ24daWLGKBFIZhEgwGME0LRRERRI9IwE9PTw8eASzHpKqiAkXTcIwiuzv7qKquJZoUEQWNXK5Af9cA0WQARVCxPJNIOIrg+KiujjGU70GQXDI9GhNaZGJVCVqmhRnqHseWX5ax7JUL6Ov6lM9f3Mrx9z7GUI9NQuxF3vH5PwnLKZ688TEuuPRgcqJE49RmGiZOoXtgiOaWA+npHWLzphV88tm/yQwOoykeQz1xXFkk4cFwWiNfHCWejDNaGsQvlEimKinl+jBK1ZSdNMP0MbFuKrlghuGiTT7jkopF6O7PEVRl8pb9X3ty0O9HEgSqGxx6umx8qNgWSEKQ/p06M2bPRHYFCsUyXd1tKJKEqATQrQKJUArDsNB8Y+7bSFTDNEAQTbq6ukAJYJkmtu0ymh5EUTS0SJTaaJBSqciIIzJrRopicYQJk2cg21mGikM010yjfWAbuf5hxIBJS9MUAnKETHI9/Z0BMsZa0qWpuLrHpOpFqF6aFe90smb0ZK5W8tRWh8hG5iD/8y0JX6Kd+qoQW1evxhwu0G320NGTZerkeQQCfu6982IEz0WOWvj9flIpjVx6AFOQ2bbbIVEhEVYryTlFQqJHyZXxCUOMmkEiviEiqUpG9RLdPa14WCiejD8u4+YEEFzSuk3cJ+I4Lrgu8USUTCbDYI8EkoksOxRGNcJhP4ViEVnyMFw/0aCErSvkSyVUzcHn0wCJWCyI69qUSgaSqKHraRzLo1B0UVWD5nHN5NIDaIEYpuUiFm1020JU48yeNpXVPy+lfkozu3fuRiirxGJhgokC0f56TjltEht3b6S6uhrPLjKYns/eBwzxy/J9yWey4C7lmb/N4OPFH9GZHebP9x5B0etGECIkxx2FnIyX2PfoFBMTIu0bA/SWDJqqEjjtLqt//pY3nnmBcFRk92AeO12gpaKJtvYBmqp8KKaIpTgElQCm4iA6MukBhSkTbIayRSzHoLFRxTJHyBkR0HR8vggyJo1V08iM9rFtZ5agIqAFZIq6i2k7ZLNpUsEgqC5DQwaZjElFdTWlYg6f38fqn3ZSKpcIBFT8viCJZBhFUQj4griiSy6Xoa62knA4zM4dHaRqUhTzWWpaEug5m7w1Su9AjlRKYHAgi6aKVFdUU1Gr0dW7EU3TSHdkCQfDiL4SyNC6qYvxUyahVYps/XKYIL1s6+gimXDJZvbEF17HyLDJNaf/huGiTPeqESQpzKSJIfSygImDPbwMYcrEZq+6upKgl2VALPL0HfcjhHz88MO3vPraB8QVA9MFwdNJxlNoAQvLCYDloJfzOILL0JDDgllRMkaakf4Yo3qRSTVBRMVkqMej7OrEYjKCplEum9iFIEGfQK6cYfnGPIqmoYpgI6AoKsViEX9w7DBZk4OYpoGiSti6g+7oxEJRCrpOKpUil8uhKAqGUUJWFSoTlRh2Adt2KZZKOK6Orlt4gkIyFcL1VPLDOUKhIKaVJ5t1SUZlJjSNRwuUyWQ1qmvr2LBmDWXHJhSSMPRRQv4YRV1D9ixUuUzJNPGHmognwKdFyRse9VVJ7j7vNHZ3fMTmbxdTEhTixl5c8fSpyGIQx5QQAz6NxsZm+nWFkDOR1tESG1auI905RNAdJW3oYErIfj9KYJR0TiCTHxybUKZZaHKY/fdU2NGRZWQwjG7lqIwIFIdyDOklXJ+J5hu7/8EwCxQzFrZcoi/fgytKVMYDSJJIJmeg+VVs08QwHAoFd+wiD9GgZDiYZglBscfGa0sSfl8QXS+jaTL+gEg8nkRT/Pj8KpZlMjgwjCHYiLJGseARDQUpZm0KowVEGZyAh+oFaGpoRgsmWLZ6I6o/QfdAF109uzEwCPh8jA4KCEqIPfbYi1DQT+PkShYccCxnnXMlLg5GNkCwcgJTprXwyh8vQUi/TvuPK/H7QmMD16q24LoBZFHBU6OIwewOtq1Zze6uHH1D65jZVM+2no1MnD2NQChJbU0LWtJBQkXPVSIrRTR9bFKD40YoFDOsadURNZGyO4wk5jFNGVH1o+gBRF8UWRGQvShmKcr0Wc2Uinkak2H8Yo68ZaKoApGIn1JJRzcNwvEQfrUCQbawXYdAwiQajVNRWUtFdQLXdfAFxkyfll3GNMZ6OhVFobezj1xapyJZSUCSqa+uZkJLPUHVJqCp+PwKoulA2gDNo1gYYTQ9RGWtnyU/rmKvBXPZtm0n4UAtoajGCadMoJRRkeNlIgmJkWGd5Uu3MNK7mKHBPkbSo+jZXbTt6qOr0MXmn9dSLA7TXyiA4eLmQwTDFoYdQfNriK99/T3n/+4ErvvDeUQCZTK5doRsgXdefJ5iwU9vaxsuoGoy/oBKTaIaW3JQixaK5WKLUJsUKRUVamINuFKSQFBlRDcwXQPPTGOUS8h+nVTcpX1XF4FgmLQTwHJ8lEtjF3oYloVj2Wh+mVLRpGwO4ff7Mco6GH56uofIZvOkRwtomkIhr+Pzi4SCAVLJMD4JjMIomWIGfyhEMpZg+rhGBgeyjI4OY2cU/KqCKHiE4lFqG6sIigqlcoa6ujoy6TKVlRrf/PAdlakqbGcU0fPx6ms/MX7SeNb91E+5LOBaLo2NBt+sbuWWv1zEHf88hG0bB7hs7+ORBj+lr09AConIjoQteCAWeeXFMp6k4tg64qVnHcA+TdO46c930dUN2ZLJjvZOTM2molpHURxEw0bP58mMlnHsHIGAyjlXXkcsatGYDFMsQUBVMC2PqqSKKo1x2WLeJRRUqW2qIpKIEAxAySogSi51CdBdcF0Py4TKqujYNTOeQjKpEgoEKRc8bARcR0f1KeTzBRRFplAoEQr7sCyHYqmMDJSKOpFIjJqKFKVCli07t7Bs9TYEo0xD83hcxUEJRIiGEoSjCWzdwTZgzvTpFLIZAv4w8UANPimK4Onkizk2bd5OZbyCigqJgFakv62Vsj3Csh93Mrpb5/slG/jjFZ8Qj1pkyyMsWzKKYeiIuo9ISCAsO+zULZYuvo9dW7djWiDPG3cS+USS+Y3NfPz2A/RkXSpDQdJmGsX2c/Ipp/HTysUM9omYviwDQzphOcC3X/yN0ayPVFIjXxBwXUjGiriCH8ln0VKVoGOwjCKFKBYMEGG0L4fi0zCMDD9vk+jsGZtRa5omAU1nVDAxTAHXkwipLuNTYaxginTeYWQ4PXbLlaZRKpXwkAkoPsqWxNZdXbiuSygcoL8ni2vrhCKQHnEJVAfYvGYze+2zgJ6+LgKqQjadwSzbSJLH9rZekqkYgSCYhRyNqRQVtUlUPUO2MUWeIYxyjs5dfehCLYzqTJ4osXtQYnevSCgok/QFyO/6J72lJHVJH7Zl4JgWw2WH760WjtxbY/x4B0+wEP/85FPEUbnjH4chjbqsXP8ebX1t5IoFcnqBJYs/oqDHqWtQ0FSVyoqpFMw4kegMkqEUg306nqAwaZyf3t0meklnoFtluNhBXU09nb0l+ruKSI6DoqoYBVA8lYroKIoiIcsyiqzRPwiyGMfvCyGJoFsi29JZ1mzoRC/bBAIhNN/YZU0tDY1EQgq6naVncJCgJCK7DqMDaWZPG4c/HMY0A0yb2kIgXIlfhQPmTKOns48995hBNBqnvjFK2SoxMjoMbpq+jgEiFfVImkopl6c147K58xf23i9MZjRHMDGJg/aZRjjsYHl1TJ+wJyF/hpMvrGKvJo+MpFEVHqCvv4hpypiezOruEOND81j7c5h3Xg0TDEYQy4Ut/Pz1+zx641uUo03MXLAQzacg6yECIT+mqKHn+xgcsGiuaaKzqxdLLLJuwyY6B4YIxctEtDDZvEdVrUI0qpKotIEQotTNpGabWJPOaG4YTyqj+QR8ahxXFKhMOUSjkbH721SFWNhEEmxMw8UX0AgG4tSNC2GURvDMIrJhoaki5XIRr+yRHiogSgJFx8KWJVxZob1nCMmv4Q+EOOLEk9i5fRMnn3k0z771JUEJPvpiGYVimZBSydTx05F8AWrqZzN/wUJ0XSdWEWX1z1vJ5zqQ7Qo++7iVvj6Jg46u4+e1a5i15540tTTQ0/UzwZBK+rNWMoMZ1BGBdEElFFAYGSkglS2KkyeRs3MUzG4+W3UL6BnE0w45kx+3rueMh55k9sxDSPrraWiowXNtnJJDS301FSEflVUqrR1t+DQZSTeZ2tJCIhrBsX34RBAFj9G8weBImpqUCpKPUFhjJFOk0JvAr4SJhUNUpDRqGmLYWQXPdNGLoyiagK2bgI9MdswM5NOKNDVGcMsakVgVtiji+UPYpk4+n6d3eJQFc2cRCynEEhoNdfXU1iex7AK5bBpNMnnh0cdIVtaxePHPHHXQHISgwrxZjVRUVEAAuvozHHXY8bS1dtDesY2ugR4KhRL7778XFdHpqFqZ5uqJDOQ6cPIwf0E9G9avY/uu1Tz74X0YWT/dgkpYDmEZOgVdJqfrBMN+XuuIctQeAfTMbvKDfrbvHmVooIx4/nUXMP+oFsqLv+euCyZy542P0717GNEfJhpJ0dbaRdHxo+cKVARl/D4Zhwq6dg8TDmpglZFDVQhuGVVQEN0ApikSRGVk0MFxkjQ2ydhOnnyuTDov0dXZiymN3e2j+QOEtSCOa5HLlZjSXINrg21XsLu9RNDvI53JoCo+PAE81SKdz2E6Jba2tyHKEYySn3yuhF4wcPHTXF9DSZCJREKoKiSTKmvXbmHe/IW0tnfT1drOpHoF1Sexdt1y9miM0dU7QkUygGMZbN2xlRNPWITueOhmmurIJFrbB5m19xwKBY0Fc/fn6w/HU00X80IWZjHNUN5Fd018IuTzZU4+83e8+HIvBAYIRhzCMY9N6/MINx0/ydtn35lMnD2PZ1/5lB272qmtURgeNUlWBBjoyeAL+RFUEdWnkc46JKMysixhZQcZHZTGGotkP1XJJIXhNKYi4nk26axFOKQQ8rmU8kUkySLvWfiI0zrUgyzA1naRUNBPQNUwHRHF58enyViOSamY/7XX3sMseogBj2gkxWDvEIIjUdeSYLAvjeBBJBrCMIqULQOPAImARkdnH01NTRSLHhMnVdLR3klQ8yOIMiPlUaZOnUPbtm3EYgkyuTQN1SH0dI7m0ET65RK2U2TczDi7W7PMmDqNt19Zwp5zGlBrC+xYa3J0ZYGyNUxUgZGshCZDMqDwnVTJtFgDG/u30b7O4NRj9+eQAyLs3JFC8PJrvMP2PZk3vl3Cj5+/yW//cRuP/f63vP/GhwiqAGUfiWSYnt4h/P4kWjCHZBbpGbCQAxqVqQgeIWwjQ2ZEJxgV0ESZslfCL/oYHMkTDwfQtCxDaZeQT6FsmBRMnaGhFIOFAoIUAKtMJBrAMAUsV0dWRRxTRvNLqJJKsVQiny8hCC6qJqAKAfLZDOMmTWbzth2o0lhGO8JYh3IhYxCOxAlFFWxdQVMlRkZGCIRCBP0hjHIGgQBN45pYt24NjfVVtPYN4JMUmpqqGOzPE/TF8eRKWuo76OqJsbt1PamaiVTEiiRtg1SkD3cwiO551KdMMC06M0lGJs/hklP2Y23XBpZ93E4xk2fCvCQdm/P8HxABSTPbXmOGAAAAAElFTkSuQmCC\" 128 = \"128px\" 128 = \"128px\" style = \"float:left; padding:2px\"></td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[0.0002684289, 0.0, 0.17076984, ...] shape=(2048,)</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "emb_pipe = (\n",
    "\tpipe.input('urls')\n",
    "\t\t.flat_map('urls', 'urls', lambda x: x)\n",
    "\t\t.map('urls', 'images', ops.image_decode.cv2('rgb'))\n",
    "\t\t.map('images', 'embeddings', ops.image_embedding.timm(model_name='resnet50'))\n",
    "\t\t.output('urls', 'images', 'embeddings')\n",
    ")\n",
    "\n",
    "res = emb_pipe(['1.jpg', '2.jpg', '3.jpg'])\n",
    "DataCollection(res).show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.8.5"
  },
  "vscode": {
   "interpreter": {
    "hash": "f7dd10cdbe9a9c71f7e71741efd428241b5f9fa0fecdd29ae07a5706cd5ff8a2"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}