DQN-tensorflow-Movie2Flights.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DQN\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training of DQN in Tensorflow\n",
    "--------------------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "code_folding": [],
    "init_cell": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.5/dist-packages/matplotlib/__init__.py:1401: UserWarning:  This call to matplotlib.use() has no effect\n",
      "because the backend has already been chosen;\n",
      "matplotlib.use() must be called *before* pylab, matplotlib.pyplot,\n",
      "or matplotlib.backends is imported for the first time.\n",
      "\n",
      "  warnings.warn(_use_error_msg)\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "from util import *\n",
    "from dlg_manager import *\n",
    "from alg import *\n",
    "from agent import *\n",
    "from user_sim import *\n",
    "from state_tracker import *\n",
    "import random\n",
    "from config import *\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from nlg import *\n",
    "from six.moves import cPickle as pickle\n",
    "import IPython\n",
    "import copy, argparse, json\n",
    "\n",
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Data\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Action Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "code_folding": [],
    "init_cell": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keys = 11\n",
      "Sample of dict:\n",
      "- closing: 5\n",
      "- inform: 1\n",
      "- greeting: 4\n",
      "- thanks: 7\n",
      "- deny: 9\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "act_set_path = './data/dia_acts.txt'\n",
    "act_set = text_to_dict(act_set_path)\n",
    "sample_dict(act_set)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### slot set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "code_folding": [],
    "init_cell": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keys = 7\n",
      "Sample of dict:\n",
      "- flightDate2: 1\n",
      "- origin1: 3\n",
      "- destination1: 0\n",
      "- taskcomplete: 6\n",
      "- ticket: 5\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "slots_set_path = \"./parser/slot_set.txt\"\n",
    "slot_set = text_to_dict(slots_set_path)\n",
    "sample_dict(slot_set)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### flight dic: info about flights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "code_folding": [],
    "init_cell": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keys = 1000\n",
      "Sample of dict:\n",
      "- 291: {'origin1': 'FCO', 'flightDate2': '6', 'flightDate1': '0', 'destination1': 'SHA', 'travelers': '3'}\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "flight_kb_path = \"./parser/fkb.json.p\"\n",
    "flight_kb = pickle.load(open(flight_kb_path, 'rb'), encoding=\"latin\")\n",
    "sample_dict(flight_kb, sample_size=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Language Generator (pretrained)\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "code_folding": [],
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#\n",
    "nlg_model_path ='data/trained_model/nlg/lstm_tanh_relu_[1468202263.38]_2_0.610.p'\n",
    "nlg_model = Nlg()\n",
    "nlg_model.load_nlg_model(nlg_model_path)\n",
    "diaact_nl_pairs_path = \"./parser/flight.nl.pairs.json\"\n",
    "nlg_model.load_predefine_act_nl_pairs(diaact_nl_pairs_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "code_folding": [],
    "init_cell": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pretrained_model_path: None\n",
      "max_epochs: 200\n",
      "slot_set: data/slot_set.txt\n",
      "cv_fold: 6\n",
      "decay_rate: 0.999\n",
      "write_model_dir: .\\checkpoints\\template\\07102016\\\n",
      "data_path: .\\data\\movieMultiLine.Annot.Corrected.Final.v3.csv\n",
      "split_method: 1\n",
      "momentum: 0.1\n",
      "activation_func: relu\n",
      "sdgtype: rmsprop\n",
      "grad_clip: -0.0001\n",
      "save_check_point: 20\n",
      "model: lstm_tanh\n",
      "trained_model_path: None\n",
      "valid_test: 0\n",
      "dia_slot_val: 2\n",
      "smooth_eps: 1e-08\n",
      "act_set: data/dia_acts.txt\n",
      "learning_rate: 0.001\n",
      "batch_size: 16\n",
      "eva_metric: 2\n",
      "slot_rep: 1\n",
      "reg_cost: 0.001\n",
      "check_point: 20\n",
      "feed_recurrence: 0\n",
      "init_rnn: 0\n",
      "hidden_size: 100\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "model_params = pickle.load(open(nlg_model_path, 'rb'), encoding='latin1')\n",
    "params = model_params['params']\n",
    "params['batch_size'] = 16\n",
    "batch_size = 16\n",
    "save_check_point = 20\n",
    "params['trained_model_path'] = None\n",
    "for k in params:\n",
    "    print(\"{}: {}\".format(k, params[k]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# User Simulator\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### goal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "code_folding": [],
    "init_cell": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "goals length: 128\n",
      "Sample the first goal: \n",
      "{'request_slots': {}, 'inform_slots': {'origin1': 'MUC', 'flightDate2': '3', 'flightDate1': '2', 'destination1': 'MCO', 'travelers': '4'}, 'diaact': 'request'}\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "goal_file_path = './parser/fg.json.p'\n",
    "all_goal_set = pickle.load(open(goal_file_path, 'rb'), encoding=\"latin\")\n",
    "print(\"goals length: {}\".format(len(all_goal_set)))\n",
    "print(\"Sample the first goal: \\n{}\".format(all_goal_set[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split goal set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "102\n",
      "0\n",
      "26\n",
      "128\n"
     ]
    }
   ],
   "source": [
    "# split goal set\n",
    "split_fold = params.get('split_fold', 5)\n",
    "goal_set = {'train':[], 'valid':[], 'test':[], 'all':[]}\n",
    "for u_goal_id, u_goal in enumerate(all_goal_set):\n",
    "    if u_goal_id % split_fold == 1: goal_set['test'].append(u_goal)\n",
    "    else: goal_set['train'].append(u_goal)\n",
    "    goal_set['all'].append(u_goal)\n",
    "print(len(goal_set['train']))\n",
    "print(len(goal_set['valid']))\n",
    "print(len(goal_set['test']))\n",
    "print(len(goal_set['all']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### user simulator param"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "code_folding": [],
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#\n",
    "usersim_params = {}\n",
    "usersim_params['max_turn'] = 40\n",
    "usersim_params['slot_err_prob'] = 0.00\n",
    "# slot_err_mode: 0 for slot_val only; 1 for three errs\n",
    "usersim_params['slot_err_mode'] = 0\n",
    "usersim_params['intent_err_prob'] = 0\n",
    "# run_mode: 0 for default NL; 1 for dia_act; 2 for both\n",
    "usersim_params['run_mode'] = 0\n",
    "# 0 for dia_act level; 1 for NL level\n",
    "usersim_params['act_level'] = 0\n",
    "# train/test/all; default is all\n",
    "usersim_params['learn_phase'] = 'all'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### a flights dictionary for user simulator - slot:possible values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keys = 20\n",
      "Sample of dict:\n",
      "- distanceconstraints: ['closest', 'visalia', 'near the space needle', 'area', 'nearest', 'near 98119', 'closest time', 'space needle', 'east side', 'downtown', 'close to 95833', 'near space needle', 'south beach', 'near', 'south barrington', 'near me', 'south side', 'local theater', 'closest theater to you', '12 miles', ' seattle area', 'northern part of the city', 'near here', 'near my location', 'safeco field', 'near you:', 'north side', 'vicinity', 'around the city', 'far away from disney', 'near you', 'near safeco field', 'in your area', 'your area', 'ave', 'general']\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "flight_dict_path = './parser/dicts.v3.p'\n",
    "flight_dictionary = pickle.load(open(flight_dict_path, 'rb'), encoding=\"latin\")\n",
    "samples = sample_dict(flight_dictionary, sample_size=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Create a User"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "user = RuleSimulator(flight_dictionary, act_set, slot_set, goal_set, usersim_params)\n",
    "# user = AlternateSimulator(flight_dictionary, act_set, slot_set, goal_set, usersim_params)\n",
    "user.set_nlg_model(nlg_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Agent\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## param"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pretrained model path = None\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "agent_params = {}\n",
    "# maximum length of each dialog (default=20, 0=no maximum length)\n",
    "agent_params['max_turn'] = 40\n",
    "# Epsilon to determine stochasticity of epsilon-greedy agent policies\n",
    "agent_params['epsilon'] = 0\n",
    "# run_mode: 0 for default NL; 1 for dia_act; 2 for both\n",
    "agent_params['agent_run_mode'] = 3\n",
    "# 0 for dia_act level; 1 for NL level\n",
    "agent_params['agent_act_level'] = 0\n",
    "\n",
    "############### DQN #################\n",
    "# the size for experience replay\n",
    "agent_params['experience_replay_pool_size'] = 10000\n",
    "# # the hidden size for DQN\n",
    "agent_params['dqn_hidden_size'] = 60\n",
    "agent_params['batch_size'] = 16\n",
    "# # gamma for DQN\n",
    "agent_params['gamma'] = 0.9\n",
    "# # predict model for DQN\n",
    "agent_params['predict_mode'] = True\n",
    "agent_params['trained_model_path'] = params['pretrained_model_path']\n",
    "#####################################\n",
    "print(\"pretrained model path = {}\".format(agent_params['trained_model_path']))\n",
    "# 0: no warm start; 1: warm start for training\n",
    "agent_params['warm_start'] = 1\n",
    "# run_mode: 0 for NL; 1 for dia_act\n",
    "agent_params['cmd_input_mode'] = 0\n",
    "\n",
    "success_rate_threshold = 0.3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### create an agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "code_folding": [],
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "act_cardinality  11\n",
      "feasible_actions 16\n",
      "80\n",
      "INFO:tensorflow:Restoring parameters from trained_model/flight_200/model.ckpt\n"
     ]
    }
   ],
   "source": [
    "# agent = RequestBasicsAgent(movie_kb, act_set, slot_set, agent_params)\n",
    "# agent = AgentDQN(movie_kb, act_set, slot_set, agent_params)\n",
    "# agt = 9\n",
    "agt = 10\n",
    "agent_params['batch_size']  = batch_size\n",
    "if agt == 9:\n",
    "    agent = AgentDQN(flight_kb, act_set, slot_set, agent_params)\n",
    "else:\n",
    "    agent = DQNAgentTF(flight_kb, act_set, slot_set, agent_params, transfer=True, path=\"trained_model/flight_200/model.ckpt\")\n",
    "\n",
    "agent.set_nlg_model(nlg_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dialog Manager\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dlg_manager = DlgManager(agent, user, act_set, slot_set, flight_kb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Running Episodes\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Param"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "code_folding": [],
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#\n",
    "status = {'successes': 0, 'count': 0, 'cumulative_reward': 0}\n",
    "# the size of validation set\n",
    "simulation_epoch_size = 100\n",
    "# the number of epochs for warm start \n",
    "warm_start_epochs = 20\n",
    "# num_episodes = 60"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\"\"\" Warm_Start Simulation (by Rule Policy) \"\"\"\n",
    "def warm_start_simulation():\n",
    "    successes = 0\n",
    "    cumulative_reward = 0\n",
    "    cumulative_turns = 0\n",
    "    \n",
    "    res = {}\n",
    "    for episode in range(warm_start_epochs):\n",
    "        dlg_manager.init_episode()\n",
    "        episode_over = False\n",
    "        while(not episode_over):\n",
    "            episode_over, reward = dlg_manager.step()\n",
    "            cumulative_reward += reward\n",
    "            if episode_over:\n",
    "                if reward > 0: \n",
    "                    successes += 1\n",
    "#                     print (\"warm_start simulation episode %s: Success\" % (episode))\n",
    "#                 else: print (\"warm_start simulation episode %s: Fail\" % (episode))\n",
    "                cumulative_turns += dlg_manager.state_tracker.turn_count\n",
    "        \n",
    "        if len(agent.experience_replay_pool) >= agent.experience_replay_pool_size:\n",
    "            break\n",
    "    \n",
    "    agent.warm_start = 2\n",
    "    res['success_rate'] = float(successes)/simulation_epoch_size\n",
    "    res['ave_reward'] = float(cumulative_reward)/simulation_epoch_size\n",
    "    res['ave_turns'] = float(cumulative_turns)/simulation_epoch_size\n",
    "    print (\"Warm_Start %s epochs, success rate %s, ave reward %s, ave turns %s\" % (episode+1, res['success_rate'], res['ave_reward'], res['ave_turns']))\n",
    "    print (\"Current experience replay buffer size %s\" % (len(agent.experience_replay_pool)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def simulation_epoch(simulation_epoch_size):\n",
    "    successes = 0\n",
    "    cumulative_reward = 0\n",
    "    cumulative_turns = 0\n",
    "    \n",
    "    res = {}\n",
    "    for episode in range(simulation_epoch_size):\n",
    "        dlg_manager.init_episode()\n",
    "        episode_over = False\n",
    "        while(not episode_over):\n",
    "            episode_over, reward = dlg_manager.step()\n",
    "            cumulative_reward += reward\n",
    "            if episode_over:\n",
    "                if reward > 0: \n",
    "                    successes += 1\n",
    "#                     print (\"simulation episode %s: Success\" % (episode))\n",
    "#                 else: print (\"simulation episode %s: Fail\" % (episode))\n",
    "                cumulative_turns += dlg_manager.state_tracker.turn_count\n",
    "    \n",
    "    res['success_rate'] = float(successes)/simulation_epoch_size\n",
    "    res['ave_reward'] = float(cumulative_reward)/simulation_epoch_size\n",
    "    res['ave_turns'] = float(cumulative_turns)/simulation_epoch_size\n",
    "    print(\"simulation success rate %s, ave reward %s, ave turns %s\" % (res['success_rate'], res['ave_reward'], res['ave_turns']))\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "code_folding": [],
    "collapsed": true,
    "init_cell": true
   },
   "outputs": [],
   "source": [
    "def run_episodes(count, status):\n",
    "    successes = 0\n",
    "    cumulative_reward = 0\n",
    "    cumulative_turns = 0\n",
    "    \n",
    "    \n",
    "    if agt >= 9 and params['trained_model_path'] == None and agent.warm_start == 1:\n",
    "        print ('warm_start starting ...')\n",
    "        warm_start_simulation()\n",
    "        print ('warm_start finished, start RL training ...')\n",
    "    \n",
    "    for episode in range(count):\n",
    "        print (\"----------------- Episode: %s ----------------- \" % (episode))\n",
    "        dlg_manager.init_episode()\n",
    "        episode_over = False\n",
    "        \n",
    "        while(not episode_over):\n",
    "            episode_over, reward = dlg_manager.step()\n",
    "            cumulative_reward += reward\n",
    "                \n",
    "            if episode_over:\n",
    "                if reward > 0:\n",
    "                    print (\"Successful Dialog!\")\n",
    "                    successes += 1\n",
    "#                 else: print (\"Failed Dialog!\")\n",
    "                \n",
    "                cumulative_turns += dlg_manager.state_tracker.turn_count\n",
    "        \n",
    "        # simulation\n",
    "        if agt >= 9 and params['trained_model_path'] == None:\n",
    "            agent.predict_mode = True\n",
    "            simulation_res = simulation_epoch(simulation_epoch_size)\n",
    "            \n",
    "            performance_records['success_rate'][episode] = simulation_res['success_rate']\n",
    "            performance_records['ave_turns'][episode] = simulation_res['ave_turns']\n",
    "            performance_records['ave_reward'][episode] = simulation_res['ave_reward']\n",
    "            \n",
    "            if simulation_res['success_rate'] >= best_res['success_rate']:\n",
    "                if simulation_res['success_rate'] >= success_rate_threshold: # threshold = 0.30\n",
    "                    agent.experience_replay_pool = [] \n",
    "                    simulation_epoch(simulation_epoch_size)\n",
    "                \n",
    "#             if simulation_res['success_rate'] > best_res['success_rate']:\n",
    "#                 best_model['model'] = copy.deepcopy(agent)\n",
    "#                 best_res['success_rate'] = simulation_res['success_rate']\n",
    "#                 best_res['ave_reward'] = simulation_res['ave_reward']\n",
    "#                 best_res['ave_turns'] = simulation_res['ave_turns']\n",
    "#                 best_res['epoch'] = episode\n",
    "                \n",
    "            loss = agent.train(batch_size, 1)\n",
    "            if agt == 10: \n",
    "                agent.model.update_target_params()\n",
    "            else: \n",
    "                agent.clone_dqn = copy.deepcopy(agent.dqn)\n",
    "                \n",
    "            agent.predict_mode = False\n",
    "            \n",
    "            print (\"Simulation success rate %s, Ave reward %s, Ave turns %s, Best success rate %s\" % (performance_records['success_rate'][episode], performance_records['ave_reward'][episode], performance_records['ave_turns'][episode], best_res['success_rate']))\n",
    "#             if episode % save_check_point == 0 and params['trained_model_path'] == None: # save the model every 10 episodes\n",
    "#                 save_model(params['write_model_dir'], agt, best_res['success_rate'], best_model['model'], best_res['epoch'], episode)\n",
    "#                 save_performance_records(params['write_model_dir'], agt, performance_records)\n",
    "        curve.append(successes/(episode+1))\n",
    "        losses.append(loss)\n",
    "        print(\"Progress: %s / %s, Success rate: %s / %s Avg reward: %.2f Avg turns: %.2f\" % (episode+1, count, successes, episode+1, float(cumulative_reward)/(episode+1), float(cumulative_turns)/(episode+1)))\n",
    "    print(\"Success rate: %s / %s Avg reward: %.2f Avg turns: %.2f\" % (successes, count, float(cumulative_reward)/count, float(cumulative_turns)/count))\n",
    "    status['successes'] += successes\n",
    "    status['count'] += count\n",
    "    \n",
    "#     if agt == 9 and params['traained_model_path'] == None:\n",
    "#         save_model(params['write_model_dir'], agt, float(successes)/count, best_model['model'], best_res['epoch'], count)\n",
    "#         save_performance_records(params['write_model_dir'], agt, performance_records)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run a Warm Start"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train & Eval\n",
    "-------------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "warm_start starting ...\n",
      "Warm_Start 20 epochs, success rate 0.2, ave reward 14.8, ave turns 2.8\n",
      "Current experience replay buffer size 140\n",
      "warm_start finished, start RL training ...\n",
      "----------------- Episode: 0 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 2261\n",
      "- cur bellman err 7.2320, experience replay pool 2261\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 1 / 200, Success rate: 0 / 1 Avg reward: -60.00 Avg turns: 42.00\n",
      "----------------- Episode: 1 ----------------- \n",
      "simulation success rate 0.0, ave reward -40.0, ave turns 2.0\n",
      "Train on : 2361\n",
      "- cur bellman err 8.3681, experience replay pool 2361\n",
      "Simulation success rate 0.0, Ave reward -40.0, Ave turns 2.0, Best success rate 0\n",
      "Progress: 2 / 200, Success rate: 0 / 2 Avg reward: -50.00 Avg turns: 22.00\n",
      "----------------- Episode: 2 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 4461\n",
      "- cur bellman err 4.5332, experience replay pool 4461\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 3 / 200, Success rate: 0 / 3 Avg reward: -53.33 Avg turns: 28.67\n",
      "----------------- Episode: 3 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 6561\n",
      "- cur bellman err 1.9567, experience replay pool 6561\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 4 / 200, Success rate: 0 / 4 Avg reward: -55.00 Avg turns: 32.00\n",
      "----------------- Episode: 4 ----------------- \n",
      "simulation success rate 0.0, ave reward -57.9, ave turns 37.8\n",
      "Train on : 8451\n",
      "- cur bellman err 1.2116, experience replay pool 8451\n",
      "Simulation success rate 0.0, Ave reward -57.9, Ave turns 37.8, Best success rate 0\n",
      "Progress: 5 / 200, Success rate: 0 / 5 Avg reward: -53.20 Avg turns: 28.40\n",
      "----------------- Episode: 5 ----------------- \n",
      "simulation success rate 0.0, ave reward -50.99, ave turns 23.98\n",
      "Train on : 9650\n",
      "- cur bellman err 1.4055, experience replay pool 9650\n",
      "Simulation success rate 0.0, Ave reward -50.99, Ave turns 23.98, Best success rate 0\n",
      "Progress: 6 / 200, Success rate: 0 / 6 Avg reward: -51.50 Avg turns: 25.00\n",
      "----------------- Episode: 6 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 11750\n",
      "- cur bellman err 0.6705, experience replay pool 11750\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 7 / 200, Success rate: 0 / 7 Avg reward: -52.71 Avg turns: 27.43\n",
      "----------------- Episode: 7 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 13850\n",
      "- cur bellman err 0.4185, experience replay pool 13850\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 8 / 200, Success rate: 0 / 8 Avg reward: -53.62 Avg turns: 29.25\n",
      "----------------- Episode: 8 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 15950\n",
      "- cur bellman err 0.2084, experience replay pool 15950\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 9 / 200, Success rate: 0 / 9 Avg reward: -54.33 Avg turns: 30.67\n",
      "----------------- Episode: 9 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 18050\n",
      "- cur bellman err 0.1550, experience replay pool 18050\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 10 / 200, Success rate: 0 / 10 Avg reward: -54.90 Avg turns: 31.80\n",
      "----------------- Episode: 10 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 20150\n",
      "- cur bellman err 0.1847, experience replay pool 20150\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 11 / 200, Success rate: 0 / 11 Avg reward: -55.36 Avg turns: 32.73\n",
      "----------------- Episode: 11 ----------------- \n",
      "simulation success rate 0.65, ave reward 27.25, ave turns 23.5\n",
      "simulation success rate 0.56, ave reward 15.2, ave turns 26.0\n",
      "Train on : 1300\n",
      "- cur bellman err 3.5087, experience replay pool 1300\n",
      "Simulation success rate 0.65, Ave reward 27.25, Ave turns 23.5, Best success rate 0\n",
      "Progress: 12 / 200, Success rate: 0 / 12 Avg reward: -55.75 Avg turns: 33.50\n",
      "----------------- Episode: 12 ----------------- \n",
      "simulation success rate 0.02, ave reward -57.58, ave turns 41.96\n",
      "Train on : 3398\n",
      "- cur bellman err 1.6884, experience replay pool 3398\n",
      "Simulation success rate 0.02, Ave reward -57.58, Ave turns 41.96, Best success rate 0\n",
      "Progress: 13 / 200, Success rate: 0 / 13 Avg reward: -56.08 Avg turns: 34.15\n",
      "----------------- Episode: 13 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 5498\n",
      "- cur bellman err 1.2560, experience replay pool 5498\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 14 / 200, Success rate: 0 / 14 Avg reward: -56.36 Avg turns: 34.71\n",
      "----------------- Episode: 14 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 7598\n",
      "- cur bellman err 1.4780, experience replay pool 7598\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 15 / 200, Success rate: 0 / 15 Avg reward: -56.60 Avg turns: 35.20\n",
      "----------------- Episode: 15 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 9698\n",
      "- cur bellman err 0.9768, experience replay pool 9698\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 16 / 200, Success rate: 0 / 16 Avg reward: -56.81 Avg turns: 35.62\n",
      "----------------- Episode: 16 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 0.7, ave reward 31.0, ave turns 28.0\n",
      "simulation success rate 0.65, ave reward 25.23, ave turns 27.54\n",
      "Train on : 1377\n",
      "- cur bellman err 7.3198, experience replay pool 1377\n",
      "Simulation success rate 0.7, Ave reward 31.0, Ave turns 28.0, Best success rate 0\n",
      "Progress: 17 / 200, Success rate: 1 / 17 Avg reward: -49.06 Avg turns: 34.24\n",
      "----------------- Episode: 17 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 3477\n",
      "- cur bellman err 2.5931, experience replay pool 3477\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 18 / 200, Success rate: 1 / 18 Avg reward: -49.67 Avg turns: 34.67\n",
      "----------------- Episode: 18 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 5577\n",
      "- cur bellman err 2.1357, experience replay pool 5577\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 19 / 200, Success rate: 1 / 19 Avg reward: -50.21 Avg turns: 35.05\n",
      "----------------- Episode: 19 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 7677\n",
      "- cur bellman err 1.1695, experience replay pool 7677\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 20 / 200, Success rate: 1 / 20 Avg reward: -50.70 Avg turns: 35.40\n",
      "----------------- Episode: 20 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 63.18, ave turns 35.64\n",
      "simulation success rate 1.0, ave reward 62.7, ave turns 36.6\n",
      "Train on : 1830\n",
      "- cur bellman err 1.7254, experience replay pool 1830\n",
      "Simulation success rate 1.0, Ave reward 63.18, Ave turns 35.64, Best success rate 0\n",
      "Progress: 21 / 200, Success rate: 2 / 21 Avg reward: -45.33 Avg turns: 35.52\n",
      "----------------- Episode: 21 ----------------- \n",
      "simulation success rate 0.0, ave reward -57.48, ave turns 36.96\n",
      "Train on : 3678\n",
      "- cur bellman err 4.2667, experience replay pool 3678\n",
      "Simulation success rate 0.0, Ave reward -57.48, Ave turns 36.96, Best success rate 0\n",
      "Progress: 22 / 200, Success rate: 2 / 22 Avg reward: -46.00 Avg turns: 35.82\n",
      "----------------- Episode: 22 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 5778\n",
      "- cur bellman err 1.9987, experience replay pool 5778\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 23 / 200, Success rate: 2 / 23 Avg reward: -46.61 Avg turns: 36.09\n",
      "----------------- Episode: 23 ----------------- \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 7878\n",
      "- cur bellman err 1.1439, experience replay pool 7878\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 24 / 200, Success rate: 2 / 24 Avg reward: -47.17 Avg turns: 36.33\n",
      "----------------- Episode: 24 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 9978\n",
      "- cur bellman err 0.8574, experience replay pool 9978\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 25 / 200, Success rate: 2 / 25 Avg reward: -47.68 Avg turns: 36.56\n",
      "----------------- Episode: 25 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 12078\n",
      "- cur bellman err 0.7118, experience replay pool 12078\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 26 / 200, Success rate: 2 / 26 Avg reward: -48.15 Avg turns: 36.77\n",
      "----------------- Episode: 26 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 68.0, ave turns 26.0\n",
      "simulation success rate 1.0, ave reward 68.0, ave turns 26.0\n",
      "Train on : 1300\n",
      "- cur bellman err 0.7604, experience replay pool 1300\n",
      "Simulation success rate 1.0, Ave reward 68.0, Ave turns 26.0, Best success rate 0\n",
      "Progress: 27 / 200, Success rate: 3 / 27 Avg reward: -43.85 Avg turns: 36.37\n",
      "----------------- Episode: 27 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 0.15, ave reward -39.9, ave turns 37.8\n",
      "Train on : 3190\n",
      "- cur bellman err 1.0409, experience replay pool 3190\n",
      "Simulation success rate 0.15, Ave reward -39.9, Ave turns 37.8, Best success rate 0\n",
      "Progress: 28 / 200, Success rate: 4 / 28 Avg reward: -39.64 Avg turns: 35.57\n",
      "----------------- Episode: 28 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 0.81, ave reward 38.82, ave turns 38.76\n",
      "simulation success rate 0.83, ave reward 41.26, ave turns 38.68\n",
      "Train on : 1934\n",
      "- cur bellman err 0.9444, experience replay pool 1934\n",
      "Simulation success rate 0.81, Ave reward 38.82, Ave turns 38.76, Best success rate 0\n",
      "Progress: 29 / 200, Success rate: 5 / 29 Avg reward: -36.14 Avg turns: 35.66\n",
      "----------------- Episode: 29 ----------------- \n",
      "simulation success rate 0.12, ave reward -44.64, ave turns 40.08\n",
      "Train on : 3938\n",
      "- cur bellman err 1.3427, experience replay pool 3938\n",
      "Simulation success rate 0.12, Ave reward -44.64, Ave turns 40.08, Best success rate 0\n",
      "Progress: 30 / 200, Success rate: 5 / 30 Avg reward: -36.93 Avg turns: 35.87\n",
      "----------------- Episode: 30 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 6038\n",
      "- cur bellman err 1.5528, experience replay pool 6038\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 31 / 200, Success rate: 5 / 31 Avg reward: -37.68 Avg turns: 36.06\n",
      "----------------- Episode: 31 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 8138\n",
      "- cur bellman err 1.4315, experience replay pool 8138\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 32 / 200, Success rate: 5 / 32 Avg reward: -38.38 Avg turns: 36.25\n",
      "----------------- Episode: 32 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.84, ave turns 12.32\n",
      "simulation success rate 1.0, ave reward 74.79, ave turns 12.42\n",
      "Train on : 621\n",
      "- cur bellman err 0.7393, experience replay pool 621\n",
      "Simulation success rate 1.0, Ave reward 74.84, Ave turns 12.32, Best success rate 0\n",
      "Progress: 33 / 200, Success rate: 6 / 33 Avg reward: -34.94 Avg turns: 35.52\n",
      "----------------- Episode: 33 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 2721\n",
      "- cur bellman err 0.8864, experience replay pool 2721\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 34 / 200, Success rate: 6 / 34 Avg reward: -35.68 Avg turns: 35.71\n",
      "----------------- Episode: 34 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 61.74, ave turns 38.52\n",
      "simulation success rate 1.0, ave reward 61.8, ave turns 38.4\n",
      "Train on : 1920\n",
      "- cur bellman err 1.5676, experience replay pool 1920\n",
      "Simulation success rate 1.0, Ave reward 61.74, Ave turns 38.52, Best success rate 0\n",
      "Progress: 35 / 200, Success rate: 7 / 35 Avg reward: -32.89 Avg turns: 35.77\n",
      "----------------- Episode: 35 ----------------- \n",
      "simulation success rate 0.21, ave reward -32.28, ave turns 36.96\n",
      "Train on : 3768\n",
      "- cur bellman err 1.5134, experience replay pool 3768\n",
      "Simulation success rate 0.21, Ave reward -32.28, Ave turns 36.96, Best success rate 0\n",
      "Progress: 36 / 200, Success rate: 7 / 36 Avg reward: -33.64 Avg turns: 35.94\n",
      "----------------- Episode: 36 ----------------- \n",
      "simulation success rate 0.24, ave reward -28.48, ave turns 36.56\n",
      "Train on : 5596\n",
      "- cur bellman err 1.5568, experience replay pool 5596\n",
      "Simulation success rate 0.24, Ave reward -28.48, Ave turns 36.56, Best success rate 0\n",
      "Progress: 37 / 200, Success rate: 7 / 37 Avg reward: -34.35 Avg turns: 36.11\n",
      "----------------- Episode: 37 ----------------- \n",
      "simulation success rate 0.2, ave reward -35.36, ave turns 40.72\n",
      "Train on : 7632\n",
      "- cur bellman err 2.1706, experience replay pool 7632\n",
      "Simulation success rate 0.2, Ave reward -35.36, Ave turns 40.72, Best success rate 0\n",
      "Progress: 38 / 200, Success rate: 7 / 38 Avg reward: -35.03 Avg turns: 36.26\n",
      "----------------- Episode: 38 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 0.22, ave reward -30.3, ave turns 35.4\n",
      "Train on : 9402\n",
      "- cur bellman err 2.6276, experience replay pool 9402\n",
      "Simulation success rate 0.22, Ave reward -30.3, Ave turns 35.4, Best success rate 0\n",
      "Progress: 39 / 200, Success rate: 8 / 39 Avg reward: -32.21 Avg turns: 35.64\n",
      "----------------- Episode: 39 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.0, ave turns 14.0\n",
      "simulation success rate 1.0, ave reward 74.0, ave turns 14.0\n",
      "Train on : 700\n",
      "- cur bellman err 1.0655, experience replay pool 700\n",
      "Simulation success rate 1.0, Ave reward 74.0, Ave turns 14.0, Best success rate 0\n",
      "Progress: 40 / 200, Success rate: 9 / 40 Avg reward: -29.55 Avg turns: 35.10\n",
      "----------------- Episode: 40 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 62.0, ave turns 38.0\n",
      "simulation success rate 1.0, ave reward 62.0, ave turns 38.0\n",
      "Train on : 1900\n",
      "- cur bellman err 1.4783, experience replay pool 1900\n",
      "Simulation success rate 1.0, Ave reward 62.0, Ave turns 38.0, Best success rate 0\n",
      "Progress: 41 / 200, Success rate: 10 / 41 Avg reward: -27.32 Avg turns: 35.17\n",
      "----------------- Episode: 41 ----------------- \n",
      "simulation success rate 0.82, ave reward 44.96, ave turns 28.88\n",
      "simulation success rate 0.84, ave reward 47.52, ave turns 28.56\n",
      "Train on : 1428\n",
      "- cur bellman err 2.0531, experience replay pool 1428\n",
      "Simulation success rate 0.82, Ave reward 44.96, Ave turns 28.88, Best success rate 0\n",
      "Progress: 42 / 200, Success rate: 10 / 42 Avg reward: -28.10 Avg turns: 35.33\n",
      "----------------- Episode: 42 ----------------- \n",
      "simulation success rate 0.0, ave reward -58.85, ave turns 39.7\n",
      "Train on : 3413\n",
      "- cur bellman err 1.4034, experience replay pool 3413\n",
      "Simulation success rate 0.0, Ave reward -58.85, Ave turns 39.7, Best success rate 0\n",
      "Progress: 43 / 200, Success rate: 10 / 43 Avg reward: -28.84 Avg turns: 35.49\n",
      "----------------- Episode: 43 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 5513\n",
      "- cur bellman err 0.7681, experience replay pool 5513\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 44 / 200, Success rate: 10 / 44 Avg reward: -29.55 Avg turns: 35.64\n",
      "----------------- Episode: 44 ----------------- \n",
      "simulation success rate 0.0, ave reward -60.0, ave turns 42.0\n",
      "Train on : 7613\n",
      "- cur bellman err 0.5849, experience replay pool 7613\n",
      "Simulation success rate 0.0, Ave reward -60.0, Ave turns 42.0, Best success rate 0\n",
      "Progress: 45 / 200, Success rate: 10 / 45 Avg reward: -30.22 Avg turns: 35.78\n",
      "----------------- Episode: 45 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 68.21, ave turns 25.58\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 68.28, ave turns 25.44\n",
      "Train on : 1272\n",
      "- cur bellman err 0.7769, experience replay pool 1272\n",
      "Simulation success rate 1.0, Ave reward 68.21, Ave turns 25.58, Best success rate 0\n",
      "Progress: 46 / 200, Success rate: 11 / 46 Avg reward: -28.09 Avg turns: 35.57\n",
      "----------------- Episode: 46 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 68.19, ave turns 25.62\n",
      "simulation success rate 1.0, ave reward 68.18, ave turns 25.64\n",
      "Train on : 1282\n",
      "- cur bellman err 0.6654, experience replay pool 1282\n",
      "Simulation success rate 1.0, Ave reward 68.19, Ave turns 25.62, Best success rate 0\n",
      "Progress: 47 / 200, Success rate: 12 / 47 Avg reward: -26.04 Avg turns: 35.36\n",
      "----------------- Episode: 47 ----------------- \n",
      "simulation success rate 0.14, ave reward -41.1, ave turns 37.8\n",
      "Train on : 3172\n",
      "- cur bellman err 0.7239, experience replay pool 3172\n",
      "Simulation success rate 0.14, Ave reward -41.1, Ave turns 37.8, Best success rate 0\n",
      "Progress: 48 / 200, Success rate: 12 / 48 Avg reward: -26.75 Avg turns: 35.50\n",
      "----------------- Episode: 48 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.81, ave turns 12.38\n",
      "simulation success rate 1.0, ave reward 74.78, ave turns 12.44\n",
      "Train on : 622\n",
      "- cur bellman err 0.3089, experience replay pool 622\n",
      "Simulation success rate 1.0, Ave reward 74.81, Ave turns 12.38, Best success rate 0\n",
      "Progress: 49 / 200, Success rate: 13 / 49 Avg reward: -24.67 Avg turns: 35.02\n",
      "----------------- Episode: 49 ----------------- \n",
      "simulation success rate 0.82, ave reward 49.36, ave turns 20.08\n",
      "simulation success rate 0.93, ave reward 64.02, ave turns 17.16\n",
      "Train on : 858\n",
      "- cur bellman err 0.9587, experience replay pool 858\n",
      "Simulation success rate 0.82, Ave reward 49.36, Ave turns 20.08, Best success rate 0\n",
      "Progress: 50 / 200, Success rate: 13 / 50 Avg reward: -25.38 Avg turns: 35.16\n",
      "----------------- Episode: 50 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.82, ave turns 12.36\n",
      "simulation success rate 1.0, ave reward 74.79, ave turns 12.42\n",
      "Train on : 621\n",
      "- cur bellman err 0.2286, experience replay pool 621\n",
      "Simulation success rate 1.0, Ave reward 74.82, Ave turns 12.36, Best success rate 0\n",
      "Progress: 51 / 200, Success rate: 14 / 51 Avg reward: -23.41 Avg turns: 34.71\n",
      "----------------- Episode: 51 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.2581, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 52 / 200, Success rate: 15 / 52 Avg reward: -21.52 Avg turns: 34.27\n",
      "----------------- Episode: 52 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.56, ave turns 12.88\n",
      "simulation success rate 1.0, ave reward 74.58, ave turns 12.84\n",
      "Train on : 642\n",
      "- cur bellman err 0.1831, experience replay pool 642\n",
      "Simulation success rate 1.0, Ave reward 74.56, Ave turns 12.88, Best success rate 0\n",
      "Progress: 53 / 200, Success rate: 16 / 53 Avg reward: -19.72 Avg turns: 33.89\n",
      "----------------- Episode: 53 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 0.77, ave reward 43.72, ave turns 19.36\n",
      "simulation success rate 0.75, ave reward 41.02, ave turns 19.96\n",
      "Train on : 998\n",
      "- cur bellman err 0.9647, experience replay pool 998\n",
      "Simulation success rate 0.77, Ave reward 43.72, Ave turns 19.36, Best success rate 0\n",
      "Progress: 54 / 200, Success rate: 17 / 54 Avg reward: -17.96 Avg turns: 33.48\n",
      "----------------- Episode: 54 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.81, ave turns 12.38\n",
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "Train on : 620\n",
      "- cur bellman err 0.1640, experience replay pool 620\n",
      "Simulation success rate 1.0, Ave reward 74.81, Ave turns 12.38, Best success rate 0\n",
      "Progress: 55 / 200, Success rate: 18 / 55 Avg reward: -16.27 Avg turns: 33.09\n",
      "----------------- Episode: 55 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.96, ave turns 12.08\n",
      "simulation success rate 1.0, ave reward 74.96, ave turns 12.08\n",
      "Train on : 604\n",
      "- cur bellman err 0.2773, experience replay pool 604\n",
      "Simulation success rate 1.0, Ave reward 74.96, Ave turns 12.08, Best success rate 0\n",
      "Progress: 56 / 200, Success rate: 19 / 56 Avg reward: -14.64 Avg turns: 32.71\n",
      "----------------- Episode: 56 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.82, ave turns 12.36\n",
      "simulation success rate 1.0, ave reward 74.79, ave turns 12.42\n",
      "Train on : 621\n",
      "- cur bellman err 0.1958, experience replay pool 621\n",
      "Simulation success rate 1.0, Ave reward 74.82, Ave turns 12.36, Best success rate 0\n",
      "Progress: 57 / 200, Success rate: 20 / 57 Avg reward: -13.07 Avg turns: 32.35\n",
      "----------------- Episode: 57 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 0.74, ave reward 39.6, ave turns 20.4\n",
      "simulation success rate 0.76, ave reward 42.41, ave turns 19.58\n",
      "Train on : 979\n",
      "- cur bellman err 0.8390, experience replay pool 979\n",
      "Simulation success rate 0.74, Ave reward 39.6, Ave turns 20.4, Best success rate 0\n",
      "Progress: 58 / 200, Success rate: 21 / 58 Avg reward: -11.55 Avg turns: 32.00\n",
      "----------------- Episode: 58 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.75, ave turns 12.5\n",
      "simulation success rate 1.0, ave reward 74.84, ave turns 12.32\n",
      "Train on : 616\n",
      "- cur bellman err 0.0778, experience replay pool 616\n",
      "Simulation success rate 1.0, Ave reward 74.75, Ave turns 12.5, Best success rate 0\n",
      "Progress: 59 / 200, Success rate: 22 / 59 Avg reward: -10.10 Avg turns: 31.69\n",
      "----------------- Episode: 59 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.82, ave turns 12.36\n",
      "simulation success rate 1.0, ave reward 74.83, ave turns 12.34\n",
      "Train on : 617\n",
      "- cur bellman err 0.0423, experience replay pool 617\n",
      "Simulation success rate 1.0, Ave reward 74.82, Ave turns 12.36, Best success rate 0\n",
      "Progress: 60 / 200, Success rate: 23 / 60 Avg reward: -8.68 Avg turns: 31.37\n",
      "----------------- Episode: 60 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.79, ave turns 12.42\n",
      "simulation success rate 1.0, ave reward 74.84, ave turns 12.32\n",
      "Train on : 616\n",
      "- cur bellman err 0.0384, experience replay pool 616\n",
      "Simulation success rate 1.0, Ave reward 74.79, Ave turns 12.42, Best success rate 0\n",
      "Progress: 61 / 200, Success rate: 24 / 61 Avg reward: -7.31 Avg turns: 31.05\n",
      "----------------- Episode: 61 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.86, ave turns 12.28\n",
      "simulation success rate 1.0, ave reward 74.76, ave turns 12.48\n",
      "Train on : 624\n",
      "- cur bellman err 0.0344, experience replay pool 624\n",
      "Simulation success rate 1.0, Ave reward 74.86, Ave turns 12.28, Best success rate 0\n",
      "Progress: 62 / 200, Success rate: 25 / 62 Avg reward: -5.98 Avg turns: 30.74\n",
      "----------------- Episode: 62 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.77, ave turns 12.46\n",
      "simulation success rate 1.0, ave reward 74.76, ave turns 12.48\n",
      "Train on : 624\n",
      "- cur bellman err 0.0271, experience replay pool 624\n",
      "Simulation success rate 1.0, Ave reward 74.77, Ave turns 12.46, Best success rate 0\n",
      "Progress: 63 / 200, Success rate: 26 / 63 Avg reward: -4.70 Avg turns: 30.44\n",
      "----------------- Episode: 63 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.78, ave turns 12.44\n",
      "simulation success rate 1.0, ave reward 74.85, ave turns 12.3\n",
      "Train on : 615\n",
      "- cur bellman err 0.0232, experience replay pool 615\n",
      "Simulation success rate 1.0, Ave reward 74.78, Ave turns 12.44, Best success rate 0\n",
      "Progress: 64 / 200, Success rate: 27 / 64 Avg reward: -3.45 Avg turns: 30.16\n",
      "----------------- Episode: 64 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.72, ave turns 12.56\n",
      "simulation success rate 1.0, ave reward 74.81, ave turns 12.38\n",
      "Train on : 619\n",
      "- cur bellman err 0.0287, experience replay pool 619\n",
      "Simulation success rate 1.0, Ave reward 74.72, Ave turns 12.56, Best success rate 0\n",
      "Progress: 65 / 200, Success rate: 28 / 65 Avg reward: -2.25 Avg turns: 29.88\n",
      "----------------- Episode: 65 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "simulation success rate 1.0, ave reward 74.83, ave turns 12.34\n",
      "Train on : 617\n",
      "- cur bellman err 0.0260, experience replay pool 617\n",
      "Simulation success rate 1.0, Ave reward 74.8, Ave turns 12.4, Best success rate 0\n",
      "Progress: 66 / 200, Success rate: 29 / 66 Avg reward: -1.09 Avg turns: 29.64\n",
      "----------------- Episode: 66 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.9, ave turns 12.2\n",
      "simulation success rate 1.0, ave reward 74.81, ave turns 12.38\n",
      "Train on : 619\n",
      "- cur bellman err 0.0436, experience replay pool 619\n",
      "Simulation success rate 1.0, Ave reward 74.9, Ave turns 12.2, Best success rate 0\n",
      "Progress: 67 / 200, Success rate: 30 / 67 Avg reward: 0.04 Avg turns: 29.37\n",
      "----------------- Episode: 67 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.81, ave turns 12.38\n",
      "simulation success rate 1.0, ave reward 74.72, ave turns 12.56\n",
      "Train on : 628\n",
      "- cur bellman err 0.0298, experience replay pool 628\n",
      "Simulation success rate 1.0, Ave reward 74.81, Ave turns 12.38, Best success rate 0\n",
      "Progress: 68 / 200, Success rate: 31 / 68 Avg reward: 1.15 Avg turns: 29.12\n",
      "----------------- Episode: 68 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.82, ave turns 12.36\n",
      "simulation success rate 1.0, ave reward 74.78, ave turns 12.44\n",
      "Train on : 622\n",
      "- cur bellman err 0.0275, experience replay pool 622\n",
      "Simulation success rate 1.0, Ave reward 74.82, Ave turns 12.36, Best success rate 0\n",
      "Progress: 69 / 200, Success rate: 32 / 69 Avg reward: 2.22 Avg turns: 28.87\n",
      "----------------- Episode: 69 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "Train on : 620\n",
      "- cur bellman err 0.0230, experience replay pool 620\n",
      "Simulation success rate 1.0, Ave reward 74.8, Ave turns 12.4, Best success rate 0\n",
      "Progress: 70 / 200, Success rate: 33 / 70 Avg reward: 3.26 Avg turns: 28.63\n",
      "----------------- Episode: 70 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "simulation success rate 1.0, ave reward 74.74, ave turns 12.52\n",
      "Train on : 626\n",
      "- cur bellman err 0.0248, experience replay pool 626\n",
      "Simulation success rate 1.0, Ave reward 74.8, Ave turns 12.4, Best success rate 0\n",
      "Progress: 71 / 200, Success rate: 34 / 71 Avg reward: 4.27 Avg turns: 28.39\n",
      "----------------- Episode: 71 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "simulation success rate 1.0, ave reward 74.85, ave turns 12.3\n",
      "Train on : 615\n",
      "- cur bellman err 0.0399, experience replay pool 615\n",
      "Simulation success rate 1.0, Ave reward 74.8, Ave turns 12.4, Best success rate 0\n",
      "Progress: 72 / 200, Success rate: 35 / 72 Avg reward: 5.25 Avg turns: 28.17\n",
      "----------------- Episode: 72 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.8, ave turns 12.4\n",
      "simulation success rate 1.0, ave reward 74.79, ave turns 12.42\n",
      "Train on : 621\n",
      "- cur bellman err 0.0344, experience replay pool 621\n",
      "Simulation success rate 1.0, Ave reward 74.8, Ave turns 12.4, Best success rate 0\n",
      "Progress: 73 / 200, Success rate: 36 / 73 Avg reward: 6.21 Avg turns: 27.95\n",
      "----------------- Episode: 73 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.76, ave turns 12.48\n",
      "simulation success rate 1.0, ave reward 74.86, ave turns 12.28\n",
      "Train on : 614\n",
      "- cur bellman err 0.0240, experience replay pool 614\n",
      "Simulation success rate 1.0, Ave reward 74.76, Ave turns 12.48, Best success rate 0\n",
      "Progress: 74 / 200, Success rate: 37 / 74 Avg reward: 7.14 Avg turns: 27.73\n",
      "----------------- Episode: 74 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.9, ave turns 12.2\n",
      "simulation success rate 1.0, ave reward 74.84, ave turns 12.32\n",
      "Train on : 616\n",
      "- cur bellman err 0.0886, experience replay pool 616\n",
      "Simulation success rate 1.0, Ave reward 74.9, Ave turns 12.2, Best success rate 0\n",
      "Progress: 75 / 200, Success rate: 38 / 75 Avg reward: 8.03 Avg turns: 27.55\n",
      "----------------- Episode: 75 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.1059, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 76 / 200, Success rate: 39 / 76 Avg reward: 8.91 Avg turns: 27.34\n",
      "----------------- Episode: 76 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 74.98, ave turns 12.04\n",
      "simulation success rate 1.0, ave reward 74.97, ave turns 12.06\n",
      "Train on : 603\n",
      "- cur bellman err 0.0821, experience replay pool 603\n",
      "Simulation success rate 1.0, Ave reward 74.98, Ave turns 12.04, Best success rate 0\n",
      "Progress: 77 / 200, Success rate: 40 / 77 Avg reward: 9.77 Avg turns: 27.14\n",
      "----------------- Episode: 77 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0707, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 78 / 200, Success rate: 41 / 78 Avg reward: 10.60 Avg turns: 26.95\n",
      "----------------- Episode: 78 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0683, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 79 / 200, Success rate: 42 / 79 Avg reward: 11.42 Avg turns: 26.76\n",
      "----------------- Episode: 79 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0459, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 80 / 200, Success rate: 43 / 80 Avg reward: 12.21 Avg turns: 26.57\n",
      "----------------- Episode: 80 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0598, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 81 / 200, Success rate: 44 / 81 Avg reward: 12.99 Avg turns: 26.40\n",
      "----------------- Episode: 81 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0502, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 82 / 200, Success rate: 45 / 82 Avg reward: 13.74 Avg turns: 26.22\n",
      "----------------- Episode: 82 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0391, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 83 / 200, Success rate: 46 / 83 Avg reward: 14.48 Avg turns: 26.05\n",
      "----------------- Episode: 83 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0354, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 84 / 200, Success rate: 47 / 84 Avg reward: 15.20 Avg turns: 25.88\n",
      "----------------- Episode: 84 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0359, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 85 / 200, Success rate: 48 / 85 Avg reward: 15.91 Avg turns: 25.72\n",
      "----------------- Episode: 85 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0321, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 86 / 200, Success rate: 49 / 86 Avg reward: 16.59 Avg turns: 25.56\n",
      "----------------- Episode: 86 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0292, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 87 / 200, Success rate: 50 / 87 Avg reward: 17.26 Avg turns: 25.40\n",
      "----------------- Episode: 87 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0341, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 88 / 200, Success rate: 51 / 88 Avg reward: 17.92 Avg turns: 25.25\n",
      "----------------- Episode: 88 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0342, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 89 / 200, Success rate: 52 / 89 Avg reward: 18.56 Avg turns: 25.10\n",
      "----------------- Episode: 89 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0251, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 90 / 200, Success rate: 53 / 90 Avg reward: 19.19 Avg turns: 24.96\n",
      "----------------- Episode: 90 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0304, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 91 / 200, Success rate: 54 / 91 Avg reward: 19.80 Avg turns: 24.81\n",
      "----------------- Episode: 91 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0238, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 92 / 200, Success rate: 55 / 92 Avg reward: 20.40 Avg turns: 24.67\n",
      "----------------- Episode: 92 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0257, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 93 / 200, Success rate: 56 / 93 Avg reward: 20.99 Avg turns: 24.54\n",
      "----------------- Episode: 93 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0233, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 94 / 200, Success rate: 57 / 94 Avg reward: 21.56 Avg turns: 24.40\n",
      "----------------- Episode: 94 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0214, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 95 / 200, Success rate: 58 / 95 Avg reward: 22.13 Avg turns: 24.27\n",
      "----------------- Episode: 95 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0247, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 96 / 200, Success rate: 59 / 96 Avg reward: 22.68 Avg turns: 24.15\n",
      "----------------- Episode: 96 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0225, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 97 / 200, Success rate: 60 / 97 Avg reward: 23.22 Avg turns: 24.02\n",
      "----------------- Episode: 97 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0217, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 98 / 200, Success rate: 61 / 98 Avg reward: 23.74 Avg turns: 23.90\n",
      "----------------- Episode: 98 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0201, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 99 / 200, Success rate: 62 / 99 Avg reward: 24.26 Avg turns: 23.78\n",
      "----------------- Episode: 99 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0220, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 100 / 200, Success rate: 63 / 100 Avg reward: 24.77 Avg turns: 23.66\n",
      "----------------- Episode: 100 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0250, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 101 / 200, Success rate: 64 / 101 Avg reward: 25.27 Avg turns: 23.54\n",
      "----------------- Episode: 101 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0196, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 102 / 200, Success rate: 65 / 102 Avg reward: 25.75 Avg turns: 23.43\n",
      "----------------- Episode: 102 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0241, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 103 / 200, Success rate: 66 / 103 Avg reward: 26.23 Avg turns: 23.32\n",
      "----------------- Episode: 103 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0177, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 104 / 200, Success rate: 67 / 104 Avg reward: 26.70 Avg turns: 23.21\n",
      "----------------- Episode: 104 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0209, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 105 / 200, Success rate: 68 / 105 Avg reward: 27.16 Avg turns: 23.10\n",
      "----------------- Episode: 105 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0197, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 106 / 200, Success rate: 69 / 106 Avg reward: 27.61 Avg turns: 23.00\n",
      "----------------- Episode: 106 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0232, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 107 / 200, Success rate: 70 / 107 Avg reward: 28.06 Avg turns: 22.90\n",
      "----------------- Episode: 107 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0197, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 108 / 200, Success rate: 71 / 108 Avg reward: 28.49 Avg turns: 22.80\n",
      "----------------- Episode: 108 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0159, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 109 / 200, Success rate: 72 / 109 Avg reward: 28.92 Avg turns: 22.70\n",
      "----------------- Episode: 109 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0187, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 110 / 200, Success rate: 73 / 110 Avg reward: 29.34 Avg turns: 22.60\n",
      "----------------- Episode: 110 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0183, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 111 / 200, Success rate: 74 / 111 Avg reward: 29.75 Avg turns: 22.50\n",
      "----------------- Episode: 111 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0198, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 112 / 200, Success rate: 75 / 112 Avg reward: 30.15 Avg turns: 22.41\n",
      "----------------- Episode: 112 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0160, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 113 / 200, Success rate: 76 / 113 Avg reward: 30.55 Avg turns: 22.32\n",
      "----------------- Episode: 113 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0165, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 114 / 200, Success rate: 77 / 114 Avg reward: 30.94 Avg turns: 22.23\n",
      "----------------- Episode: 114 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0184, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 115 / 200, Success rate: 78 / 115 Avg reward: 31.32 Avg turns: 22.14\n",
      "----------------- Episode: 115 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0228, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 116 / 200, Success rate: 79 / 116 Avg reward: 31.70 Avg turns: 22.05\n",
      "----------------- Episode: 116 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0205, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 117 / 200, Success rate: 80 / 117 Avg reward: 32.07 Avg turns: 21.97\n",
      "----------------- Episode: 117 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0174, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 118 / 200, Success rate: 81 / 118 Avg reward: 32.43 Avg turns: 21.88\n",
      "----------------- Episode: 118 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0167, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 119 / 200, Success rate: 82 / 119 Avg reward: 32.79 Avg turns: 21.80\n",
      "----------------- Episode: 119 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0171, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 120 / 200, Success rate: 83 / 120 Avg reward: 33.14 Avg turns: 21.72\n",
      "----------------- Episode: 120 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0213, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 121 / 200, Success rate: 84 / 121 Avg reward: 33.49 Avg turns: 21.64\n",
      "----------------- Episode: 121 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0219, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 122 / 200, Success rate: 85 / 122 Avg reward: 33.83 Avg turns: 21.56\n",
      "----------------- Episode: 122 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0255, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 123 / 200, Success rate: 86 / 123 Avg reward: 34.16 Avg turns: 21.48\n",
      "----------------- Episode: 123 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0189, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 124 / 200, Success rate: 87 / 124 Avg reward: 34.49 Avg turns: 21.40\n",
      "----------------- Episode: 124 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0189, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 125 / 200, Success rate: 88 / 125 Avg reward: 34.82 Avg turns: 21.33\n",
      "----------------- Episode: 125 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0171, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 126 / 200, Success rate: 89 / 126 Avg reward: 35.13 Avg turns: 21.25\n",
      "----------------- Episode: 126 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0171, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 127 / 200, Success rate: 90 / 127 Avg reward: 35.45 Avg turns: 21.18\n",
      "----------------- Episode: 127 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0162, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 128 / 200, Success rate: 91 / 128 Avg reward: 35.76 Avg turns: 21.11\n",
      "----------------- Episode: 128 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0209, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 129 / 200, Success rate: 92 / 129 Avg reward: 36.06 Avg turns: 21.04\n",
      "----------------- Episode: 129 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0182, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 130 / 200, Success rate: 93 / 130 Avg reward: 36.36 Avg turns: 20.97\n",
      "----------------- Episode: 130 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0162, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 131 / 200, Success rate: 94 / 131 Avg reward: 36.66 Avg turns: 20.90\n",
      "----------------- Episode: 131 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0180, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 132 / 200, Success rate: 95 / 132 Avg reward: 36.95 Avg turns: 20.83\n",
      "----------------- Episode: 132 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0174, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 133 / 200, Success rate: 96 / 133 Avg reward: 37.23 Avg turns: 20.77\n",
      "----------------- Episode: 133 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0206, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 134 / 200, Success rate: 97 / 134 Avg reward: 37.51 Avg turns: 20.70\n",
      "----------------- Episode: 134 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0263, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 135 / 200, Success rate: 98 / 135 Avg reward: 37.79 Avg turns: 20.64\n",
      "----------------- Episode: 135 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0216, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 136 / 200, Success rate: 99 / 136 Avg reward: 38.07 Avg turns: 20.57\n",
      "----------------- Episode: 136 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0206, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 137 / 200, Success rate: 100 / 137 Avg reward: 38.34 Avg turns: 20.51\n",
      "----------------- Episode: 137 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0257, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 138 / 200, Success rate: 101 / 138 Avg reward: 38.60 Avg turns: 20.45\n",
      "----------------- Episode: 138 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0189, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 139 / 200, Success rate: 102 / 139 Avg reward: 38.86 Avg turns: 20.39\n",
      "----------------- Episode: 139 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0185, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 140 / 200, Success rate: 103 / 140 Avg reward: 39.12 Avg turns: 20.33\n",
      "----------------- Episode: 140 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0205, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 141 / 200, Success rate: 104 / 141 Avg reward: 39.38 Avg turns: 20.27\n",
      "----------------- Episode: 141 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0215, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 142 / 200, Success rate: 105 / 142 Avg reward: 39.63 Avg turns: 20.21\n",
      "----------------- Episode: 142 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0162, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 143 / 200, Success rate: 106 / 143 Avg reward: 39.87 Avg turns: 20.15\n",
      "----------------- Episode: 143 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0193, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 144 / 200, Success rate: 107 / 144 Avg reward: 40.12 Avg turns: 20.10\n",
      "----------------- Episode: 144 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0218, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 145 / 200, Success rate: 108 / 145 Avg reward: 40.36 Avg turns: 20.04\n",
      "----------------- Episode: 145 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0219, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 146 / 200, Success rate: 109 / 146 Avg reward: 40.60 Avg turns: 19.99\n",
      "----------------- Episode: 146 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0191, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 147 / 200, Success rate: 110 / 147 Avg reward: 40.83 Avg turns: 19.93\n",
      "----------------- Episode: 147 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0169, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 148 / 200, Success rate: 111 / 148 Avg reward: 41.06 Avg turns: 19.88\n",
      "----------------- Episode: 148 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0207, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 149 / 200, Success rate: 112 / 149 Avg reward: 41.29 Avg turns: 19.83\n",
      "----------------- Episode: 149 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0202, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 150 / 200, Success rate: 113 / 150 Avg reward: 41.51 Avg turns: 19.77\n",
      "----------------- Episode: 150 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0191, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 151 / 200, Success rate: 114 / 151 Avg reward: 41.74 Avg turns: 19.72\n",
      "----------------- Episode: 151 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0211, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 152 / 200, Success rate: 115 / 152 Avg reward: 41.95 Avg turns: 19.67\n",
      "----------------- Episode: 152 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0303, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 153 / 200, Success rate: 116 / 153 Avg reward: 42.17 Avg turns: 19.62\n",
      "----------------- Episode: 153 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0265, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 154 / 200, Success rate: 117 / 154 Avg reward: 42.38 Avg turns: 19.57\n",
      "----------------- Episode: 154 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0233, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 155 / 200, Success rate: 118 / 155 Avg reward: 42.59 Avg turns: 19.52\n",
      "----------------- Episode: 155 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0225, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 156 / 200, Success rate: 119 / 156 Avg reward: 42.80 Avg turns: 19.47\n",
      "----------------- Episode: 156 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0234, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 157 / 200, Success rate: 120 / 157 Avg reward: 43.01 Avg turns: 19.43\n",
      "----------------- Episode: 157 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0212, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 158 / 200, Success rate: 121 / 158 Avg reward: 43.21 Avg turns: 19.38\n",
      "----------------- Episode: 158 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0196, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 159 / 200, Success rate: 122 / 159 Avg reward: 43.41 Avg turns: 19.33\n",
      "----------------- Episode: 159 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0196, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 160 / 200, Success rate: 123 / 160 Avg reward: 43.61 Avg turns: 19.29\n",
      "----------------- Episode: 160 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0235, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 161 / 200, Success rate: 124 / 161 Avg reward: 43.80 Avg turns: 19.24\n",
      "----------------- Episode: 161 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0212, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 162 / 200, Success rate: 125 / 162 Avg reward: 43.99 Avg turns: 19.20\n",
      "----------------- Episode: 162 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0145, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 163 / 200, Success rate: 126 / 163 Avg reward: 44.18 Avg turns: 19.15\n",
      "----------------- Episode: 163 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0185, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 164 / 200, Success rate: 127 / 164 Avg reward: 44.37 Avg turns: 19.11\n",
      "----------------- Episode: 164 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0190, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 165 / 200, Success rate: 128 / 165 Avg reward: 44.56 Avg turns: 19.07\n",
      "----------------- Episode: 165 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0203, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 166 / 200, Success rate: 129 / 166 Avg reward: 44.74 Avg turns: 19.02\n",
      "----------------- Episode: 166 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0167, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 167 / 200, Success rate: 130 / 167 Avg reward: 44.92 Avg turns: 18.98\n",
      "----------------- Episode: 167 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0171, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 168 / 200, Success rate: 131 / 168 Avg reward: 45.10 Avg turns: 18.94\n",
      "----------------- Episode: 168 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0148, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 169 / 200, Success rate: 132 / 169 Avg reward: 45.28 Avg turns: 18.90\n",
      "----------------- Episode: 169 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0172, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 170 / 200, Success rate: 133 / 170 Avg reward: 45.45 Avg turns: 18.86\n",
      "----------------- Episode: 170 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0190, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 171 / 200, Success rate: 134 / 171 Avg reward: 45.63 Avg turns: 18.82\n",
      "----------------- Episode: 171 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0176, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 172 / 200, Success rate: 135 / 172 Avg reward: 45.80 Avg turns: 18.78\n",
      "----------------- Episode: 172 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0239, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 173 / 200, Success rate: 136 / 173 Avg reward: 45.97 Avg turns: 18.74\n",
      "----------------- Episode: 173 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0195, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 174 / 200, Success rate: 137 / 174 Avg reward: 46.13 Avg turns: 18.70\n",
      "----------------- Episode: 174 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0224, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 175 / 200, Success rate: 138 / 175 Avg reward: 46.30 Avg turns: 18.66\n",
      "----------------- Episode: 175 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0189, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 176 / 200, Success rate: 139 / 176 Avg reward: 46.46 Avg turns: 18.62\n",
      "----------------- Episode: 176 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0188, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 177 / 200, Success rate: 140 / 177 Avg reward: 46.62 Avg turns: 18.59\n",
      "----------------- Episode: 177 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0201, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 178 / 200, Success rate: 141 / 178 Avg reward: 46.78 Avg turns: 18.55\n",
      "----------------- Episode: 178 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0198, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 179 / 200, Success rate: 142 / 179 Avg reward: 46.94 Avg turns: 18.51\n",
      "----------------- Episode: 179 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0189, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 180 / 200, Success rate: 143 / 180 Avg reward: 47.09 Avg turns: 18.48\n",
      "----------------- Episode: 180 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0218, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 181 / 200, Success rate: 144 / 181 Avg reward: 47.25 Avg turns: 18.44\n",
      "----------------- Episode: 181 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0166, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 182 / 200, Success rate: 145 / 182 Avg reward: 47.40 Avg turns: 18.41\n",
      "----------------- Episode: 182 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0145, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 183 / 200, Success rate: 146 / 183 Avg reward: 47.55 Avg turns: 18.37\n",
      "----------------- Episode: 183 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0164, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 184 / 200, Success rate: 147 / 184 Avg reward: 47.70 Avg turns: 18.34\n",
      "----------------- Episode: 184 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0268, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 185 / 200, Success rate: 148 / 185 Avg reward: 47.85 Avg turns: 18.30\n",
      "----------------- Episode: 185 ----------------- \n",
      "Successful Dialog!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0245, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 186 / 200, Success rate: 149 / 186 Avg reward: 47.99 Avg turns: 18.27\n",
      "----------------- Episode: 186 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0230, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 187 / 200, Success rate: 150 / 187 Avg reward: 48.14 Avg turns: 18.24\n",
      "----------------- Episode: 187 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0255, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 188 / 200, Success rate: 151 / 188 Avg reward: 48.28 Avg turns: 18.20\n",
      "----------------- Episode: 188 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0237, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 189 / 200, Success rate: 152 / 189 Avg reward: 48.42 Avg turns: 18.17\n",
      "----------------- Episode: 189 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0283, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 190 / 200, Success rate: 153 / 190 Avg reward: 48.56 Avg turns: 18.14\n",
      "----------------- Episode: 190 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0314, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 191 / 200, Success rate: 154 / 191 Avg reward: 48.70 Avg turns: 18.10\n",
      "----------------- Episode: 191 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0241, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 192 / 200, Success rate: 155 / 192 Avg reward: 48.84 Avg turns: 18.07\n",
      "----------------- Episode: 192 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0209, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 193 / 200, Success rate: 156 / 193 Avg reward: 48.97 Avg turns: 18.04\n",
      "----------------- Episode: 193 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0176, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 194 / 200, Success rate: 157 / 194 Avg reward: 49.11 Avg turns: 18.01\n",
      "----------------- Episode: 194 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0147, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 195 / 200, Success rate: 158 / 195 Avg reward: 49.24 Avg turns: 17.98\n",
      "----------------- Episode: 195 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0169, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 196 / 200, Success rate: 159 / 196 Avg reward: 49.37 Avg turns: 17.95\n",
      "----------------- Episode: 196 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0171, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 197 / 200, Success rate: 160 / 197 Avg reward: 49.50 Avg turns: 17.92\n",
      "----------------- Episode: 197 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0164, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 198 / 200, Success rate: 161 / 198 Avg reward: 49.63 Avg turns: 17.89\n",
      "----------------- Episode: 198 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0162, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 199 / 200, Success rate: 162 / 199 Avg reward: 49.76 Avg turns: 17.86\n",
      "----------------- Episode: 199 ----------------- \n",
      "Successful Dialog!\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "simulation success rate 1.0, ave reward 75.0, ave turns 12.0\n",
      "Train on : 600\n",
      "- cur bellman err 0.0181, experience replay pool 600\n",
      "Simulation success rate 1.0, Ave reward 75.0, Ave turns 12.0, Best success rate 0\n",
      "Progress: 200 / 200, Success rate: 163 / 200 Avg reward: 49.88 Avg turns: 17.83\n",
      "Success rate: 163 / 200 Avg reward: 49.88 Avg turns: 17.83\n"
     ]
    }
   ],
   "source": [
    "performance_records = {}\n",
    "performance_records['success_rate'] = {}\n",
    "performance_records['ave_turns'] = {}\n",
    "performance_records['ave_reward'] = {}\n",
    "\n",
    "best_model = {}\n",
    "best_res = {'success_rate': 0, 'ave_reward':float('-inf'), 'ave_turns': float('inf'), 'epoch':0}\n",
    "\n",
    "curve = []\n",
    "losses = []\n",
    "agent.warm_start = 1\n",
    "run_episodes(200, status)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9UAAAJBCAYAAABIypb+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3XeY3GW5//H3nZBQQoDQpSShGXqNFCFCQJoKoQjSQYoo\notj4HUEEpBzLUbBxlHIApQkCUqRIjTQDCUrvSSCAlIQkBEJIfX5/PDNuLzM7O7Oz+35d11zfnedb\n5t5kr8BnnxYpJSRJkiRJUun61boASZIkSZLqlaFakiRJkqQyGaolSZIkSSqToVqSJEmSpDIZqiVJ\nkiRJKpOhWpIkSZKkMhmqJUnqRSLi1YhIEbFTrWuRJKkvMFRLkupaRFxeCJFja12Lui4iRkbELyPi\niYiYFhHzI+K9iBgXET+LiI1rXaMkSY0tVusCJElSRU0EPgY+qnUhpYiIwcDFwJcaNS8E3geWBbYp\nvE6OiGtTSgdVv0pJklqyp1qSpF4kpbRLSmn9lNJjta6lswqB+kFyoF4A/A74FDAwpbQCMBDYFDib\nHLK/1MajJEmqOnuqJUlSrV0EbEbuYR+TUrqr8cmU0iLgaeDpiPgVcEX1S5QkqXX2VEuS+qyIWDoi\nTo2I8RHxfkR8HBEvR8SvI2LNNu4ZHBFHRcR1EfFMRMyMiDkR8UpEXBQR67XzeanwGh4RG0TEHyLi\n9cK84ZsK1wwvXld4v3FE/Cki3i7U90JE/DAiBrbxGa0uVFao+T9zzyNir4i4v1D/h4U5ywd38Oe1\nRkT8X0S8WahlUkScHxFDmj+/syJiK6A4lPu05oG6uZTSe8Dnmz1jbOGzj2rnc9r6czmz0H55RPSL\niBMj4rHCn0uKiM0j4qXC1yd28L38rXDd+a2cG1h49oMRMT0i5kbEaxFxaURs0N5zJUk9m6FaktQn\nFYLMM8C5wEhgKfLQ43WBbwBPRsT2rdx6JHAZcACwPnnebz9gHeA44F8R8dkOPn4UMAE4gjxfeEEb\nNe4GPEYe7rwEMAAYAZwFXNfJb7W15/4QuAX4TKFpEHm+8tUR8a027tkUeBI4GlgNmA+sCnwLGA8s\nX2Y5xxeO04ELOnNDSimV+VntCeBG4DfAlkDjz7imcDykzZsjVgZ2Kby9utm5T5D/Hn8D7ED+O58L\nDAW+DPwzIvbr+rcgSaoFQ7Ukqc+JiGWB24FhwJ/JQ4+XSCktTQ7HVwNDgBsiYrlmt08jB/GtgaUK\nc36XADYAriIH1KsjYlA7JfwvOYhuklJahhzov9vKddcCtwJrpZSWA5YBTiEHvjER8blSv3dgc+AM\n4IfACoXnrgpcXzj/44hoEpAjYnHyn9PywMvADimlwcDS5F7jQYXnlWN04Xh3SunjMp9RCfsBewAn\nAMuklIYAqwCTaAjJ20XE8DbuPwDoD7ycUhpfbIyIAcDN5J+xe4FPk3/WliH/cuKX5J+fKyJinQp/\nT5KkKjBUS5L6opOB4cA1KaUDU0pPpZQWAqSUJqWUDgXuJIeqYxvfmFL6U0rptJTS+JTSvEJbSim9\nABwO3AOsBHyxnc9/F9gzpfRMo/sntnLdeOCglNKrhetmp5R+AtxWON/eZ7RlWeCMlNI5KaWZhee+\nQ+41n0oOeF9ods8hwCfJc573SCk9XLhvUUrpdmCfwnNLUgic6xbePlnG91JJSwPfTCn9LqX0EUBK\n6d2U0qyU0ovAPwvXtTVEvth+TbP2I8mLrj1I/jv/R0ppfuH5b6WUvg1cSP7Fyrcr9+1IkqrFUC1J\n6ouOLBx/0c41xd7JXTv70MKw5GLgbW3oeNFvU0pzOvHIn7Qx1PmmwrGcPZs/JveONlGo529tPLc4\nNPn6lNKkVu59FBhbRi2Ne8Snl3F/Jb0HXNrO+eLPQ4tQHRFDyT3Qja8rKv6s/aoYpltxVeHY6Z81\nSVLP4erfkqQ+pbAA2RqFt7cXFwRrRXEhsBYLlkXEGuR5158lDxcfTMtfVK/WThn/6GS549tof7Nw\nHNLJ5zT2XEppdonP3aJwfKid5z5Iw1DuejQhpdTq3PaCPwE/AzaJiI1SSs82OncweU72Pwu92gBE\nxGLkaQIAF0ZEW3PG+xeOrS6OJ0nq2QzVkqS+5hONvl65E9cv1fhNROwI/JU8XLjofXIPMMCS5LnP\n7c2pntqJzyWl9EEbp4qfNaAzz2mmrWe299wVC8e32rn332XU0rh3utyFziql3b+TlNKbEfEAsBN5\nOPwPGp0u9l4376VenoZfzqzQiRqW7MQ1kqQexuHfkqS+pvF/+4aklKKD1/DixYU5wFeSA/U95NWz\nl0wpLZdSWjWltCrwneLl7dSwsLLfUn0qDId+pfB2s1rWQuf+TloMAS+sIr8ZsIjcm91Y45+1LTrx\ns9bez4wkqYcyVEuS+pp3Gn09tMR7tyMPHZ8OjEkpPdjKitWrdKW4Hmpa4fiJdq5p71x77i8cd42I\nJcp8RnHYdnv3l7yQWiuuB+YBa0XEtoW2YsB+IKX0ZrPr36MhrJf6syZJqhOGaklSn5JSmkxDsN6z\nxNuLc7FfKq4Q3YqO9qiuR/8qHHdo55pRZT77osJxefJ2Vh2KiOY9ujMLxzWaX1u4fl2g+dZoJUsp\nzSCvCg8Ne1a3NfS72BM/ofC21J81SVKdMFRLkvqiywvH70XE6m1dFFnjMPZ+4bhea72qEbEb9b1Y\nV1v+Ujju39o+zRHxKcr8vlNKE4DrCm/PjYh2V8COiBVoWGG96OnCce82bvt+ObW1oRieDyz0Vq9L\n7r2+vo3rLy8cj4qIdoe4R0Q5C89JkmrMUC1J6i0GRMSKHbyKC3D9BJhEXoDrkYg4MCL+s0hURAyN\niK+Q9ybep9FnPAx8RF506o8R8YnC9UtGxNHADeQhv73N1eS5z0sCd0bEdvCfXzrsQd7i6/127u/I\nceRgvAR5RfYLImLLiOjX6HM2jogfARNp2et7PZDIK3P/qviLkIhYOSJ+Td4/vK2RBaW6BfiQPMy/\nuJr3nYVe7Nb8HzCu8L3dFxHHRcQyxZMRsWpEHBoRfwdOqlCNkqQqMlRLknqLT5NXcG7vtT1ASmkm\nsDvwPHmu67XABxExLSI+Al4DLgQ2J4c1Gt13SuHtAcC/I2ImMIscnl4BftSt32UNFOaNH0AeZj2C\n/IuID4DZwB3kkHl24fK5ZTx/Fnlo+fXknUlOAB4H5kXEe+Se4KeB08krq/+h2f3P0rD39jeBGREx\nA3i78Kzj6eSK652odQ4N+4RvWTi2GPrd6Pr5wBjyL2SWJw93nxER70XEh+QV1a8kL3rX1vZukqQe\nzFAtSeqTUkqvkPdfPoG8WNYM8mJWC4CnyOHn8+TA0/i+XwP70dBrvRjwAnAGOdi3t2VV3UopPUFe\n5foyclgdUDieR96LeVHh0pmtPqDj589KKR1QeNZvyCF6FjlEzyL39v4E2CCldFQrj/gu+e/ySfLW\nYAn4G7BzSunycmpqR+MQ/SG597pNKaV3gR2BQ4HbyQF/cOH0C8AfgQPJ358kqc5ESv5SVJIkdU1E\nXAEcBvwopXRmjcuRJKlq7KmWJEldEhFrA/sX3t5dy1okSao2Q7UkSepQRIyJiP+OiI2KC75FxOIR\nMQa4j7yI2biU0sM1LVSSpCpz+LckSepQRBwLXFx4u4g8d3oZ8pxyyIu77ZJSmliD8iRJqhlDtSRJ\n6lBhf+pjgZ2BYeTtyD4mr3h+C/CrwurokiT1KYZqSZIkSZLK5JxqSZIkSZLKZKiWJEmSJKlMhmpJ\nkiRJkspkqJYkSZIkqUyGakmSJEmSymSoliRJkiSpTIZqSZIkSZLKZKiWJEmSJKlMhmpJkiRJkspk\nqJYkSZIkqUyGakmSJEmSymSoliRJkiSpTIZqSZIkSZLKZKiWJEmSJKlMhmpJkiRJkspkqJYkSZIk\nqUyGakmSJEmSymSoliRJkiSpTIZqSZIkSZLKZKiWJEmSJKlMhmpJkiRJkspkqJYkSZIkqUyGakmS\nJEmSymSoliRJkiSpTIZqSZIkSZLKZKiWJEmSJKlMhmpJkiRJkspkqJYkSZIkqUyGakmSJEmSymSo\nliRJkiSpTIZqSZIkSZLKZKiWJEmSJKlMhmpJkiRJkspkqJYkSZIkqUyGakmSJEmSymSoliRJkiSp\nTIZqSZIkSZLKZKiWJEmSJKlMhmpJkiRJkspkqJYkSZIkqUyGakmSJEmSymSoliRJkiSpTIZqSZIk\nSZLKtFitC6hXK664Yho+fHity5AkSZIkdYPHH398WkpppY6uM1SXafjw4UyYMKHWZUiSJEmSukFE\nvNaZ6xz+LUmSJElSmQzVkiRJkiSVyVAtSZIkSVKZDNWSJEmSJJXJUC1JkiRJUpkM1ZIkSZIklclQ\nLUmSJElSmQzVkiRJkiSVyVAtSZIkSVKZDNWSJEmSJJXJUC1JkiRJUpkM1ZIkSZIklclQLUmSJElS\nmQzVkiRJkiSVyVAtSZIkSVKZFqt1AZIkSZKkPuD99+GNN+D11/Nr0CA45JBaV9VlhmpJkiRJUtd8\n+GFDWG4cnBu//+CDpvdstZWhWpIkSZLUy82bB2++Ca+9BlOm5Ffz0DxzZunPff31ytdaA4ZqSZIk\nSerLZs5sCMvF4Nw4QP/735BS5T/33Xdh7lxYfPHKP7uKDNWSJEmS1FstXAhvvdU0JDcPzrNmVaeW\nxReHNdaANddsOM6bZ6iupYhYEzgf2BUI4B7gWymlKZ24dyhwNjAaWAl4HbgO+HFKaXa3FS1JkiRJ\nlfLRR633Lhe/fuMNWLCg++sYMKBlYG7+9YorQkT311JldRuqI2Ip4D5gLnAkkIBzgPsjYtP2gnFE\nDCIH8AHAD4EpwKeAHwHrAV/q3uolSZIkqRPmzs0BefJkePXVlsd3361OHZ/4BAwdCsOG5WMxKBeD\n88orQ7++uWNz3YZq4DhgbWBESukVgIh4CngZOB44r517tyeH591TSncV2u6PiOWB70XEUimlj7qv\ndEmSJEkC5s/PC3a1FpgnT87zmbvbEkvkoFx8FYNz8bjGGnU/RLs71XOo3hsYVwzUACmlyRHxMDCG\n9kP1wMKx+eSBmUA/8lBySZIkSeqahQtzMG6rp/n112HRou6tYcUVWw/Lxa9XWqlXDsuulnoO1RsB\nN7fS/ixwQAf33kPu0f5pRHyNPPx7a+Ak4PfOqZYkSZLUabNn55A8cWLT16RJOTjPn999n92/fx6C\n3TwwNx6mPWhQ932+6jpULw/MaKV9OjCkvRtTSh9HxA7ADeQQXnQJcGJb90XEV4CvAAwdOrTUeiVJ\nkiTVo5Rg6tSWgbn49dtvd99nR8Dqq8Naa8Hw4fnY+OvVV4fF6jnW1b8++acfEUsA1wIrA4fT0FN9\nOrAA+Fpr96WULgIuAhg5cmQ3bNQmSZIkqSYWLMgLgjUOy40D9Icfdt9nr7JKy7BcPA4dCgMHdvQE\n1VA9h+oZtN4j3VYPdmPHADsB66aUJhbaHoiI94GLIuL3KaUnK1apJEmSpNqbNy8P03755fx66aWG\n4Pzaa3n+c3dYYYWWYbn49bBhsNRS3fO5qop6DtXPkudVN7ch8FwH924CzGgUqIseKxw3AAzVkiRJ\nUr1ZuDAH5GJoLgbol1/O85u7Izj365fD8dprwzrrNLyK75dZpvKfqR6jnkP1LcDPI2LtlNIkgIgY\nTt4u6/sd3Ps2MCQi1m28ejiwTeH4ZoVrlSRJklQpixbBG280DczFED1pUvcsDDZoUNuhedgwGDCg\n8p+puhAp1efU4IgYRO5NngOcBiTgbGAwsGlK6cPCdcOAicBZKaWzCm3DgafI4fpc8pzqkcAPgZeA\nrVNK7a5rP3LkyDRhwoSKf1+SJEmSyIuDvfNO097m4tevvAIff1z5z1xlldZD8zrrwMoru+1UHxMR\nj6eURnZ0Xd32VKeUZkfEzsD5wBXkvaXvBb5VDNQFAfQn7z9dvPfViNgWOBM4B1gReJ28CNm5HQVq\nSZIkSRUyd24OyS+8AC++mI/Fr2fNqvznrbYafPKTsN56+bXuuvm11lqw9NKV/zz1enUbqgFSSlOA\n/Tu45lVysG7e/hxwYPdUJkmSJOk/UoJp01oPzpMm5eHclbTyyg2huXmAds9mVVhdh2pJkiRJPcj8\n+Tkktxaep0+v7GcNGdI0MBcD9LrrwrLLVvazpHYYqiVJkiSV5uOP8/zm555r+nr55bzfc6UsvXRD\ncG4eoFdYoXKfI3WBoVqSJElS6z76KPcyF0Pzs8/m48SJlR2yPXQojBgB66/f8BoxIs9/dnEw9XCG\nakmSJKmv+/DDPEy7GJqLr8mT83zoSlhyyRyUG4fnESNyD7TznFXHDNWSJElSXzFnDjz/PDz9dH4V\nw/Nrr1XuM1ZbrWVwXn99WHNN6Nev4/ulOmOoliRJknqbhQvzgmHF8Fx8vfJK5YZtDxsGG27Y8Npo\noxyeXSRMfYyhWpIkSapXKcE77zQNzs88k4dxz5nT9edH5P2bGwfnDTfM4dk9nSXAUC1JkiTVhw8/\nzIH5mWeahuhp07r+7H79YJ11WobnESNgqaW6/nypFzNUS5IkST1JSvDGG/Dkk/DEEw2viRMr8/y1\n1oKNN4ZNNsnHjTbKi4UtsURlni/1MYZqSZIkqVbmz88LhzUP0NOnd/3ZK6yQg3Pj10YbweDBXX+2\npP8wVEuSJEnVMHNmDs+NA/Szz8K8eV177hJL5LDcPECvsop7PEtVYKiWJEmSKiklmDKlac/zE0/A\nq6927bkRsN56DcO2i+F5nXWgf/+KlC6pdIZqSZIkqVwp5T2eH3+86eu997r23EGDYNNNYfPNG14b\nb+yiYVIPZKiWJEmSOqO7AvRqqzUNz5ttZu+zVEcM1ZIkSVJzjQP0hAn5+M9/di1A9++f93duHJ43\n2wxWXrlydUuqOkO1JEmS+rZigC6G5+KrKytwDxqUg/MWWzQE6I02giWXrFzdknoEQ7UkSZL6lvfe\ng/Hj4bHH4NFH83HatPKfN2hQDs9bbZVfI0fmfZ8dvi31CYZqSZIk9V4ff5xX3i6G58ceg1deKf95\nSy/dNEBvtZUBWurjDNWSJEnqHRYtghdfbNoD/eSTsGBBec9rHqBHjsxbWhmgJTViqJYkSVJ9eu89\nGDcOHnkkh+jx42HWrPKeNWgQbLllDs6Ne6D79atszZJ6HUO1JEmSer5Fi+C55+Af/8gh+h//yL3S\n5ejfHzbZBLbeOr+22QY22MAeaEllMVRLkiSp53n//dz7XAzRjz6a28oxfHhDeN566zyke9CgipYr\nqe8yVEuSJKm2UoKXX27aC/3MM7m9VEOGNPRAF1/uAy2pGxmqJUmSVF3z5+cVuR98ML8eeqi8La36\n9YNNN4XttsuvbbeFddeFiMrXLEltMFRLkiSpe330UV5QrBiix42D2bNLf86QITk8f/rT+fipT8Hg\nwZWvV5JKYKiWJElSZb33Hjz8cEOIfvzx8ra12mijpiHa1bgl9UCGakmSJHXNlCkNAfrBB/Mq3aUa\nPDgP3y6G6G22geWWq3ytklRhhmpJkiSV5o034P7782vsWJg8ufRnrLoqjBoFO+yQj5tu6pZWkuqS\noVqSJEnt+/e/c3guhuhXXin9Geuum8Nz8bXOOi4oJqlXMFRLkiSpqXfeaQjR998PL71U2v0RsNlm\nDQF6hx3gE5/ollIlqdYM1ZIkSX3d1Knw9783hOjnny/t/gED8hzoYoj+9Kdh2WW7p1ZJ6mEM1ZIk\nSX3NRx/BAw/A3XfDPffAU0+Vdv9ii+UQvdNOMHp0XlxsqaW6pVRJ6ukM1ZIkSb3dwoXwz3/mEH33\n3fDIIzBvXufv798/7wk9enQO0ttvD4MGdVu5klRPDNWSJEm90aRJDT3R994LM2Z0/t5+/WCrrXKI\nHj06h+jBg7uvVkmqY4ZqSZKk3mD6dLjvvhyi7747h+rOioAtt2wYzj1qFCyzTLeVKkm9SV2H6ohY\nEzgf2BUI4B7gWymlKR3cdyZwRhun56aUlqhknZIkSRW3YAGMGwd33gl33QUTJkBKnb9/7bVh113z\na/RoWH757qtVknqxug3VEbEUcB8wFzgSSMA5wP0RsWlKaXY7t18C3NmsbVCh7ZZuKFeSJKnr3n47\nh+g77shBeubMzt87ZAjssgt89rM5SK+9dvfVKUl9SN2GauA4YG1gRErpFYCIeAp4GTgeOK+tG1NK\nbwBvNG6LiMPJfx5/6K6CJUmSSrJgATz6aA7Rd9yRFxvrrAED8lzoYm/0llvmBcckSRVVz6F6b2Bc\nMVADpJQmR8TDwBjaCdVtOBJ4B/hb5UqUJEkq0TvvNO2NLmWBsU02aQjRo0a5QrckVUE9h+qNgJtb\naX8WOKCUBxXmZo8GfplSWlCB2iRJkjpn4cKmvdGPP975e1dYAXbfHfbcMw/rXnXV7qtTktSqeg7V\nywOt/ep2OjCkxGcdBvSjg6HfEfEV4CsAQ4cOLfEjJEmSCt5/H/72N7jllhykp0/v3H0RMHJkDtF7\n7pn3jnZItyTVVD2H6ko6AvhXSump9i5KKV0EXAQwcuTIEpbXlCRJfd6rr8Ktt+YgPXZsni/dGcsv\n39AbvfvusPLK3VmlJKlE9RyqZ9B6j3RbPditioitgfWBb1WoLkmSJFi0CMaPbwjSTz/d+Xsb90Zv\nvbW90ZLUg9VzqH6WPK+6uQ2B50p4zpHAfODqShQlSZL6sDlz4O67c5C+9da86FhnDBnStDd6lVW6\nt05JUsXUc6i+Bfh5RKydUpoEEBHDge2B73fmARExEDgIuCOlNLWb6pQkSb3ZzJlw221w44151e6P\nPurcfSNGwN57w157wXbbwWL1/L9lktR31fO/3hcDJwI3R8RpQALOBl4HLixeFBHDgInAWSmls5o9\n4wvk4eLuTS1Jkjrv7bfh5pvhL3+B++6D+fM7vqdfP9hhh4Yg/clPdn+dkqRuV7ehOqU0OyJ2Bs4H\nrgACuBf4Vkrpw0aXBtCfvLp3c0eSVwv/azeXK0mS6t3kyTlE33gjPPIIpE6sWTp4MOyxRw7Se+6Z\nt8CSJPUqdRuqAVJKU4D9O7jmVXKwbu3cmG4oS5Ik9RbPPptD9I03whNPdO6eNdeEMWNykN5xRxg4\nsHtrlCTVVF2HakmSpIp7/nm47rr8eq6Ta5+OGAH77ZdfW22V95OWJPUJhmpJkqQXX2wI0s8807l7\nttyyIUhvsEH31idJ6rEM1ZIkqW96+eWGIP3UUx1fH5EXGttvP9hnHxg+vNtLlCT1fIZqSZLUd0yc\n2BCkOzNHun9/2GUX+OIX8xxp94+WJDVjqJYkSb3bO+/AtdfCVVfBY491fH3//rDzznDggblHesUV\nu79GSVLdMlRLkqTeZ9YsuOmmHKTvuQcWLWr/+n79YPToHKT33RdWWqk6dUqS6p6hWpIk9Q7z5sEd\nd8DVV8Mtt8DHH7d/fb9+ecurAw/M86RXXrk6dUqSehVDtSRJql+LFsGDD+Ye6euvhxkzOr5n1Cg4\n6KAcpFddtftrlCT1aoZqSZJUf158Ef7wB7jySnj99Y6v33RTOPRQOPhgWHPN7q9PktRnGKolSVJ9\nmDkzLzh2+eUwblzH1w8bBocckl8bb9zt5UmS+iZDtSRJ6rkWLoS778690n/5C8yd2/71K6yQ50gf\neihst12eNy1JUjcyVEuSpJ7n+edzkL7iCvj3v9u/dqmlYMyYHKR32w0GDKhOjZIkYaiWJEk9xcyZ\ncM01OUw/+mjH1++4Ixx1FOy/Pwwe3O3lSZLUGkO1JEmqnZTgoYfg4ovhz3/ueBustdaCI4+EI47I\nX0uSVGOGakmSVH3TpsEf/5jD9AsvtH/toEF5nvSRR+btsJwnLUnqQQzVkiSpOhYtgvvvz0H6L3+B\nefPav3706Dy8e7/9YOmlq1KiJEmlMlRLkqTu9fbbeRusSy6BiRPbv3bYMDj66Dy8e/jwalQnSVKX\nGKolSVLlLVqUt8K68EK49VZYsKDtaxdbLK/efdxxsOuuDu+WJNUVQ7UkSaqcGTPgssvgd7+DV15p\n/9p114Vjj81DvFdZpSrlSZJUaYZqSZLUdU88ARdcAFddBXPmtH3dwIF5C6zjjstbYtkrLUmqc4Zq\nSZJUnnnz4IYbcph++OH2r91ggxykDz8cVlyxOvVJklQFhmpJklSaN9/Mc6Uvugjeeaft6wYMyFth\nffWrsP32EFG9GiVJqhJDtSRJ6lhK8Pe/w29/CzfdBAsXtn3tGmvkIH3ssc6VliT1eoZqSZLUtnnz\n4Npr4bzz8rzp9uyyC3z967DXXnlFb0mS+gD/iydJklp67708xPu3v4W33mr7usGD4cgj4YQT8rxp\nSZL6GEO1JElq8NJL8MtfwuWXt7+K94YbwoknwmGH5WAtSVIfZaiWJKmvSwnGjs1DvP/617avi4Ax\nY+Ckk/J2WC48JkmSoVqSpD6rs/OlBw2CY46Bb34T1lmnevVJklQHDNWSJPU1s2bB73+fh3m3N196\njTVykD7uOFhuuerVJ0lSHTFUS5LUV0ydCr/6FVxwAcyc2fZ1I0fCd74DX/xi3mtakiS1yVAtSVJv\n99pr8ItfwCWXtL34WATss08O09tv73xpSZI6yVAtSVJv9dxz8NOfwtVXw4IFrV8zaBAcfXRefMz5\n0pIklcxQLUlSb/PYY/DjH8NNN7V9zfLL5/nSJ54IK6xQvdokSeplDNWSJPUGKcG998JPfpKPbVl9\ndfjud/PiY0svXb36JEnqpQzVkiTVs5TgzjvhRz+CRx9t+7r11oP/+i847DBYfPHq1SdJUi9nqJYk\nqR6lBHfckcP0Y4+1fd0WW8App8B++0H//tWrT5KkPqJfrQvoiohYMyKuj4j3I2JWRNwYEUNLuH+D\niPhzREyLiDkR8WJEnNSdNUuS1CUpwW23wTbbwOc/33ag3nHH3IP9+ONwwAEGakmSuknd9lRHxFLA\nfcBc4EggAecA90fEpiml2R3cP7Jw/1jgWOB9YD3ACWaSpJ4nJbj9djjzTJgwoe3rvvAFOPVU2G67\nqpUmSVJfVrehGjgOWBsYkVJ6BSAingJeBo4HzmvrxojoB/wRuDeltG+jU/d3X7mSJJUhJbjrLvjh\nD2H8+LZJX7PjAAAgAElEQVSv23tvOP102Gqr6tUmSZLqevj33sC4YqAGSClNBh4GxnRw707ABrQT\nvCVJqrmHHoKddoI99mg7UI8Zk4d433yzgVqSpBqo51C9EfBMK+3PAht2cO8OheMSETEuIuZHxLsR\n8euIWLKiVUqSVKrHH4c994RRo+CBB1q/Zt994V//yntRb7lldeuTJEn/Uc+henlgRivt04EhHdy7\nWuF4LXAXsCvwM/Lc6qvbuikivhIREyJiwtSpU0uvWJKk9jz7LOy/P4wcmRcZa82++8ITT8CNN8Lm\nm1e3PkmS1EI9z6nuiuIvE65MKZ1e+HpsRPQHfhIRG6SUnm9+U0rpIuAigJEjR6bqlCpJ6vVefTXP\nh77yyjyHujV77AHnnOMQb0mSeph67qmeQes90m31YDf2XuF4d7P2uwrHLbpQlyRJnfPee/Dd78KI\nEXDFFa0H6uIQ8DvuMFBLktQD1XNP9bPkedXNbQg814l727OorIokSeqMOXPg17+GH/8Y3n+/9Wu2\n2grOPRd22w0iqlufJEnqtHruqb4F2DYi1i42RMRwYPvCufbcQd7fevdm7XsUju1sACpJUpkWLoTL\nLoNPfhK+//3WA/VGG+X50uPHw+67G6glSerh6jlUXwy8CtwcEWMiYm/gZuB14MLiRRExLCIWRERx\n7jQppfeAHwNfjYj/jojPRsT3gdOBPzTepkuSpC5LCW6/PS8sdvTR8MYbLa9ZY40cuJ98Mi9GZpiW\nJKku1O3w75TS7IjYGTgfuAII4F7gWymlDxtdGkB/Wv4C4SzgA+AE4HvAW8D/AGd3c+mSpL7kmWfg\nO9+Bu5sv41Gw7LJw6qnwjW/Aku7qKElSvanbUA2QUpoC7N/BNa+Sg3Xz9gScV3hJklRZU6fmFb0v\nuggWtbJUx8CBOUifeiosv3z165MkSRVR16FakqQeZ+5c+M1v4OyzYdaslucj4LDD8vlhw6pfnyRJ\nqihDtSRJlZAS3HQTnHwyTJzY+jU77wy/+EWeWy1JknoFQ7UkSV311FNw0kkwdmzr59dbD37+c9hr\nLxcgkySpl6nn1b8lSaqtmTPhm9+ELbZoPVAvuyycd15erGzvvQ3UkiT1QvZUS5JUqkWL4PLL817T\nU6e2PN+/P3z1q3DmmbDiitWuTpIkVZGhWpKkUkyYACeeCI8+2vr53XfPvdMbbljduiRJUk04/FuS\npM6YNg2OPx623rr1QL322nDrrXDnnQZqSZL6EHuqJUlqz6JFcMklcMopMH16y/NLLJH3mj755Py1\nJEnqUwzVkiS15Zlncu/0I4+0fn7fffNQ7+HDq1qWJEnqORz+LUlScx99lHumt9ii9UD9yU/mYd43\n3migliSpj7OnWpKkxu68E044ASZPbnluqaXg9NPh29+GgQOrX5skSepxDNWSJAG89VYOy9de2/r5\nvfaC3/wGhg2rbl2SJKlHc/i3JKlvW7QILrwQNtig9UC9+up5mPfNNxuoJUlSC/ZUS5L6rokT4dhj\nYezYluf69YNvfAPOPhsGD656aZIkqT4YqiVJfc/ChfDrX8MPfgBz5rQ8v+WWufd65Mjq1yZJkuqK\noVqS1Lc8/zwcfTSMG9fy3KBBcM45cOKJsJj/iZQkSR3z/xgkSX3D/PnwP/8DP/oRzJvX8vyuu8JF\nF7lFliRJKomhWpLU+z3xRO6d/te/Wp5bdlk4/3w46iiIqHppkiSpvrn6tySp91qwAM49Fz71qdYD\n9d57w3PPwZe/bKCWJEllsadaktQ7vfQSHHEEPPpoy3Mrrpj3nP7SlwzTkiSpS+ypliT1LosWwQUX\nwOabtx6oDzoo904fdJCBWpIkdZk91ZKk3uONN/Lc6bvvbnluxRXzNln77Vf9uiRJUq9lT7Ukqf6l\nBFddBRtv3Hqg3ntveOYZA7UkSao4Q7Ukqb5Nn57nRh92GLz/ftNzgwfDZZfBTTfBKqvUpj5JktSr\nOfxbklS/HngADj00D/tubqed4PLLYdiwalclSZL6EHuqJUn1Z8ECOP10GD26ZaBefPG87/S99xqo\nJUlSt7OnWpJUX157DQ45BB55pOW5LbeEK66ADTesfl2SJKlPsqdaklQ/rrsONtus9UB98snwj38Y\nqCVJUlWVFKojYnBEnB4RD0XEyxGxXaF9xUL7+t1TpiSpT5s9G449Ni9I1nwxslVXhbvugp/9DAYO\nrE19kiSpz+r08O+IWAl4CFgbeKVwXBIgpTQtIo4ElgO+0w11SpL6qqeeggMPhBdfbHnuc5/Lq3uv\nvHL165IkSaK0nupzgFWBbYBRQDQ7fzOwS4XqkiQJLr0UttmmZaAeOBB++Uv4618N1JIkqaZKCdVf\nAP43pfRPILVyfhKwZkWqkiT1bR99BF/+MhxzDHz8cdNzI0bAo4/CSSdBNP/9riRJUnWVEqpXJA/7\nbssiYImulSNJ6vNeegm23TbvMd3cMcfA44/D5ptXvSxJkqTWlBKq3wbWaef8FsCUrpUjSerTrrsO\nttoKnn66aftSS8Ef/wiXXAKDBtWmNkmSpFaUEqpvB46JiE80PxER2wBHkOdVS5JUmnnz4JvfzKt7\nf/hh03Prrw+PPQaHH16b2iRJktpRSqj+EbAA+BfwY/K86iMj4hrgAeDfwE8rXmE7ImLNiLg+It6P\niFkRcWNEDO3kvamNl2MKJamapkyBUaPgN79pee6gg2D8eNhoo+rXJUmS1Amd3lIrpfR2RGwL/BY4\nmrz69+HkcH078LWU0vRuqbIVEbEUcB8wFziyUMc5wP0RsWlKaXYnHnM5cGGztpcqWackqR1jx8IB\nB8C0aU3bBw6E88+Hr33NxcgkSVKP1ulQDZBSeh0YExHLACPIwfqVaobpRo4j75U9IqX0CkBEPAW8\nDBwPnNeJZ7yZUhrXfSVKklqVEvz2t/Dtb8PChU3PDR8Of/4zjBxZk9IkSZJK0enh3xFxREQMB0gp\nzUopjU8pPVYM1BExPCKO6J4yW7U3MK4YqAt1TQYeBsZUsQ5JUik+/hiOPjrPoW4eqL/wBfjnPw3U\nkiSpbpQyp/oy4NPtnN+mcE21bAQ800r7s8CGnXzG1yJibkR8FBH3RcSoypUnSWrhjTfgM59pfbus\nM8+Em2+GIUOqXZUkSVLZShn+3dGktgHkvaqrZXlgRivt04HO/B/ZlcBfyQusDQNOBu6LiF1TSmNb\nuyEivgJ8BWDo0E6thyZJKnroIdh/f3j33abtgwfDFVfAGAcZSZKk+lNKTzXkxcBaiIjlgM8Db3W5\noipJKR2eUro2pfRgSulKYAdywD6nnXsuSimNTCmNXGmllapWqyTVtZTg97+H0aNbBur11oNHHzVQ\nS5KkutVuqI6IMyJiYUQsJAfqK4vvG7+A94ADgT9VoeaiGbTeI91WD3a7UkofALcBn+piXZKkovnz\n8wreX/saLFjQ9NznP5/3n95gg9rUJkmSVAEdDf9+Avgjeej3EcCDwKRm1yTgQ2AccE2lC2zHs+R5\n1c1tCDzXhee22hsvSSrRzJl5u6x77ml57gc/gLPOgn6lDpiSJEnqWdoN1Smlm4GbASJiGHBOSune\nahTWCbcAP4+ItVNKkyCvQA5sD3y/1IcVtgn7AvBYBWuUpL5p0qS8kvfzzzdtHzQoL1L2xS/WpCxJ\nkqRK6/RCZSml0d1ZSBkuBk4Ebo6I08g9zGcDrwMXFi8q/DJgInBWSumsQtv3yPts30/DQmXfA1YF\nDq3i9yBJvc/DD8M++8C0aU3bhw2DW2+FTTapTV2SJEndoJTVv/8jIpYGlqOVOdkppSldLaozUkqz\nI2Jn4HzgCvIQ9XuBb6WUPmxcLtC/Wa0vAvsWXssCs8j7Wx+TUrKnWpLKddVVeQ/qefOatm+7Ldx0\nE6yySm3qkiRJ6iYlheqIOAg4DWhvVZn+XaqoBIUAv38H17xKs+3AUkq3Ard2X2WS1MekBD/6UX41\n96UvwWWXwZJLVr8uSZKkbtbpFWIiYh/ganIQv5AcVK8B/gzMBx4HzuqGGiVJPdnHH8Ohh7YeqH/4\nQ7j6agO1JEnqtUrpqf4e8DywFbA08FXg0pTSfRGxMXn49LmVL1GS1GNNnw57753nUTc2cCD83//B\nYYfVpi5JkqQqKWUvk02BP6SUPgYWFdr6A6SUngEuAk6pbHmSpB5ryhTYYYeWgXqFFeDeew3UkiSp\nTyglVPcH3it8PadwXLbR+ReBjStRlCSph3v6adhuu5ZbZq2/Pjz6aA7bkiRJfUApofoN8tZTpJTm\nAO+Sh4IXjQBmV640SVKP9MADMGoU/PvfTds/8xl45BFYZ53a1CVJklQDpcypfgT4LHB64f0twLci\nYg45nH8dV9SWpN7thhvyomRz5zZt339/uPJKWGKJ2tQlSZJUI6WE6v8F9o2IJQs91T8AtgbOLJx/\nlryYmSSpN/rf/4UTT8zbZzX29a/Dr34F/au2o6IkSVKP0elQnVIaD4xv9H4qsHlEbAosBJ5PKS1q\n635JUp1KKW+NdW4rGzycey6ccgpEVL8uSZKkHqCUnupWpZSeKn4dEZFS8y4MSVLdWrAAjj8eLr20\naXv//nDxxfDlL9emLkmSpB6ilIXK2hTZYcALlXieJKkHmDcPDj64ZaBeckm4+WYDtSRJEp3sqY6I\nLYD1gOnA31NK8xudOwg4g7z69wfdUaQkqcrmzIEvfhFuv71p+worwG23wTbb1KYuSZKkHqbdUB0R\nSwJ/AXZt1Px6RHwWmAv8CdiWHKbPBc7vpjolSdXywQcwZgzcf3/T9qFD4a67YMSI2tQlSZLUA3XU\nU/1fwG7ABGAssC6wD/B7YE1gdeBs4PyU0vvdV6YkqSpmzoQ994Rx45q2r7ce3HNPDtaSJEn6j45C\n9ReBh4Cdiit7R8Tp5G20pgCbppQmdWuFkqTqmDoVdtsNnniiafvGG8Pdd8Oqq9amLkmSpB6so4XK\n1gZuaLZV1rWF488M1JLUS7z5Juy4Y8tAPXIkjB1roJYkSWpDR6F6CWBqs7ZphePLlS9HklR1r74K\nn/kMPP980/YddoB7782Lk0mSJKlVXdlSa2HFqpAk1cYrr8CoUTCp2cCjXXeFO++EZZapTV2SJEl1\nojNbah0REds2er8EkIATI2KfZtemlNJJFatOktR9Jk6EnXbKQ78bGzMGrr0WFl+8JmVJkiTVk86E\n6t0Kr+aaB2rIYdtQLUk93aRJMHp0y0B9yCFw+eUwYEBNypIkSao3HYXqtapShSSpel59NQfq119v\n2n7UUXDJJdC/fy2qkiRJqkvthuqU0mvVKkSSVAWvvZaHfE+Z0rT9iCMM1JIkSWXoykJlkqR6MmVK\n7qF+rdnvSw87DC691EAtSZJUBkO1JPUFr7+eA/XkyU3bDz44z6E2UEuSJJXFUC1Jvd2bb+ZA3Xzb\nrC99Cf74RwO1JElSFxiqJak3e+utHKgnTmzafsABcOWVsFhnNoGQJElSWwzVktRbTZ8Ou+0GL7/c\ntH3//eGqqwzUkiRJFWColqTe6MMP4fOfh2eeadq+775wzTXuQy1JklQhXeqmiIjFgDHA8sCtKaW3\nK1KVJKl8c+fCfvvBuHFN27/wBfjTnwzUkiRJFdTpnuqI+FlEjG/0PoB7gOuAC4GnI2KdypcoSeq0\nBQvg0EPh7rubtu+4I1x3HQwcWJu6JEmSeqlShn/vATzY6P1ewGeA/wEOKbR9v0J1SZJKlRIcfzzc\ncEPT9i23hFtugSWXrE1dkiRJvVgpw7/XBBqvdrMXMDml9H2AiNgIOLSCtUmSOislOPlkuPTSpu3r\nrw933gnLLFObuiRJknq5UnqqBwILGr0fTR7+XTQJ+EQlipIklegnP4Ff/KJp29ChcNddsNJKtalJ\nkiSpDyglVL8ObAf/6ZVeG/h7o/MrAx9WrjRJUqf8/vdw6qlN21ZaKc+rXnPN2tQkSZLUR5Qy/PtP\nwA8jYmVgI2AWcHuj81sAEytYmySpI9ddByec0LRtmWXgb3+DT36yNjVJkiT1IaX0VP8YuJzcW52A\nI1JKMwEiYllgb+DeShcoSWrDgw/C4Yfn+dRFSywBf/0rbLFF7eqSJEnqQzrdU51SmgscU3g19wF5\nPvVHFapLktSeF16AMWNg3ryGtsUWyyt/jxpVu7okSZL6mFJ6qtszIKX0fkppfoWe1ykRsWZEXB8R\n70fErIi4MSKGlvGc70dEioiHuqNOSaqod96BPfeEGTOatl96KXzuc7WpSZIkqY/qdKiOiD0j4sxm\nbSdExCxgdkRcHREDKl1gO/UsBdwHrA8cCRwOrAfcHxGDSnjO2sBpwLvdUackVdTs2bDXXvDqq03b\nzz47DwWXJElSVZWyUNnJNAqeEbEB8Cvy4mSTgS8BjwG/rGSB7TiOvAL5iJTSK4WaniLvpX08cF4n\nn/M74CpgBKX9eUhSdS1cCAcfDOPHN20/5hj4wQ9qU5MkSVIfV8rw7w2ACY3efwmYA2ydUtoTuJbc\nY1wtewPjioEaIKU0GXgYGNOZB0TEIcCWwCndUqEkVUpKcNJJcOutTdt33x1+9zuIqE1dkiRJfVwp\noXoIMK3R+88C96WUZhXejwXWqlBdnbER8Ewr7c8CG3Z0c0QMAc4H/l9KaXqFa5OkyjrvPLjggqZt\nm20Gf/4zDKjazBtJkiQ1U0qongYMA4iIwcCngAcbnR8A9K9caR1aHpjRSvt08i8AOvI/wEvkbcI6\nJSK+EhETImLC1KlTO3ubJHXNn/8M3/te07Y11oDbboPBg2tTkyRJkoDS5hD/A/hqRDwL7Fm4945G\n59cF3qpgbd0mIkYBRwBbptR4g9f2pZQuAi4CGDlyZKfvk6SyPfxwywXIllkGbr8dVl+9NjVJkiTp\nP0oJ1WcA9wPXFd7/IaX0HEBEBLBv4Xy1zKD1Hum2erAbuxD4P+CNiFiu0LYY0L/wfk5hX25Jqp3X\nXoN99oG5jf45Ku5FvckmtatLkiRJ/9HpUJ1Seq6w4vf2wPsppQcanV6OPD95bGXLa9ez5HnVzW0I\nPNfBvRsUXl9t5dwM4NtUbxVzSWpp9mzYe2+YNq1p+yWXwGc/W5uaJEmS1EJJW0gVFvS6tZX2GeTt\ntarpFuDnEbF2SmkSQEQMJ4f+73dw7+hW2n5JnhP+DeCVVs5LUnWkBEcdBU891bT99NPhyGpusiBJ\nkqSOlLwvc0R8BtgNWAX4RUrphYhYmrw11VMppZkVrrEtFwMnAjdHxGlAAs4GXicP7y7WO4y8l/ZZ\nKaWzAFJKY5s/LCJmAou1dk6Squqcc+D665u27b8/nHFGbeqRJElSmzq9+ndE9I+Ia8nzpk8FjgZW\nK5xeANwEnFDxCtuQUpoN7ExewfsK4CpgMrBzSunDxqWTe6BLWelckmrjL3/JPdKNbbopXH459POf\nMUmSpJ6mlJ7q/wL2B74D3Ak8XzyRUvo4Iv4CfA7474pW2I6U0pRCTe1d8yo5WHf0rJ0qU5Uklenp\np1uu9L3CCnDTTbD00rWpSZIkSe0qpdvjCOCPKaVfkfesbu55YJ2KVCVJfc20aTBmTF6grGixxfIw\n8LXWql1dkiRJalcpoXo4ea/qtsyk9S2uJEntmT8fDjgAJk9u2v7rX8NOO9WkJEmSJHVOKaH6A/Ie\n0G1ZF5jatXIkqQ/69rdh7NimbccfD1/7Wk3KkSRJUueVEqofAg6LiBbzkyNiCHnhsvsrVZgk9QkX\nXwwXXNC0bdSo3EstSZKkHq+UUH0usB5wH/CFQttmEXE88E9gEPCTypYnSb3YY4/B17/etG3YMLjh\nBhg4sDY1SZIkqSSdXv07pTQhIvYHLgEuKzT/nLyy9rvAviml5ypfoiT1Qu+9l+dRz5/f0LbUUnDz\nzbDSSrWrS5IkSSUpZUstUkq3RcRwYFdgA3Kgfhn4W0rpo4pXJ0m90aJFcMQRMGVK0/bLLoPNNqtN\nTZIkSSpLSaEaIKU0F/hr4SVJKtVPfwq339607aST4MADa1OPJEmSytbpOdURsUVEfL2d81+PiM0r\nU5Yk9VL33w+nnda0bdtt4Wc/q009kiRJ6pJSFio7A/h8O+f3BE7vWjmS1Iu99RYcfHAe/l20wgpw\n7bUuTCZJklSnSgnVnwL+3s75vwNbd60cSeqlFiyAgw6Cd95paIuAK6+EoUNrV5ckSZK6pJRQvSIw\nvZ3zMwvXSJKaO+00eOCBlm177FGbeiRJklQRpYTqd4GN2jm/Me2Hbknqm269NS9O1tguu8AZZ9Sm\nHkmSJFVMKaH6HuDYiGgRrCNiQ+CYwjWSpKLJk/P2WY2tthpcfTX071+bmiRJklQxpWypdQ6wHzA+\nIi4Fnii0bw4cDcwDzq5seZJUx+bOhQMOgJkzG9r694frroOVV65dXZIkSaqYTofqlNLEiNgFuBw4\nodnpZ4Evp5RermBtklTfTjkFHn+8adtPfwrbb1+beiRJklRxpfRUk1KaAGxc2I96vULzSymlJyte\nmSTVszvugPPPb9q2zz7wne/Uph5JkiR1i5JCdVFK6Qkahn9Lkhp7+2048simbUOHwqWX5m20JEmS\n1Gt0eqGyiNglIn7czvkfR8ToypQlSXVq0aK8MNnUqQ1t/fvDNdfAkCG1q0uSJEndopTVv/8LWLed\n82sVrpGkvusXv4C7727aduaZ8OlP16QcSZIkda9SQvVmwLh2zj9auEaS+qbHH4dTT23atuOOecEy\nSZIk9UqlhOplgdntnJ8DOLZRUt/00Udw6KGwYEFD2/LLw5VXuh+1JElSL1ZKqH4T2Kqd81sBb3et\nHEmqU//v/8GLLzZtu+QSWGON2tQjSZKkqiglVN8GHBkRn21+orB/9ZHA7ZUqTJLqxp13wgUXNG07\n9ljYd9/a1CNJkqSqKWVLrXOB/YG/RcQdNGyptTmwJ7mX+uzKlidJPdy0afDlLzdtW2edlntUS5Ik\nqVfqdKhOKb3z/9u783i7prvx45+vJOYGIS2PCmLMYPgRbVQVoUVrLC0tqrREPYZWtTU/YnhU1VBD\nf8byIKqoCm0NTQ1VpcSUiimpRBB+QgiCIFm/P/a5zz3n3HPvPffknHuG+3m/Xud1s9dee9/vPjv7\nnv09a+21IuILwP8lS6K/2rYKuB04LKX0avVDlKQGlRKMHZvNS91mscXgmmtg2WXrF5ckSZJ6TU9a\nqkkpvQh8NSJWoH16rWkppbeqHpkkNbprroGbby4sO/542Hzz+sQjSZKkXtejpLpNLol+pMqxSFLz\nePllOOKIwrLNNoMTT6xPPJIkSaqLspPqiBhSTr2U0szKw5GkJpBSNhDZ3LntZUsumbVcDxhQv7gk\nSZLU63rSUj2D7Pnp7jghq6TWdvnlcOedhWVnnAHrrVefeCRJklQ3PUmqT6FjUt0fWAvYFfgX2YBl\nktS6ZsyAo44qLPvSlzp2BZckSVKf0JPRv0/ubF1EDAUeBCZVISZJakwLF8L3vgfvvddetswycOWV\n2ajfkiRJ6nOqcheYUnoBuAQYV439SVJDuuQSuPvuwrKzzoKhQ+sTjyRJkuqumk0rrwDDq7g/SWoc\nM2fCT39aWLbtttk81ZIkSeqzqplU7wY4X7Wk1pMSHHJIYbfvZZeFK66w27ckSVIf15MptU7qZNUg\nYAwwEvhFNYIqV0SsBpwLfBkIYCLww+6m9YqI1YHzgY2BTwPzgCnAmSmlP9c0aEnNZ/x4uL1oHMYz\nz4TVV69PPJIkSWoYPRn9++Qu1r0GnACcuUjR9EBELA3cDcwH9icbmfw04J6I2DClNK+LzZcF3iCL\n+WVgIHAQ8KeI2COldHNNg5fUPF5/HY48srBsyy2zlmtJkiT1eT1JqtcsUZaAOSml90qsq7WDgKHA\neimlaQARMRmYCowFzulsw5TSFOB7+WUR8SdgOnAAYFItKXP44TBnTvvyEktk81Tb7VuSJEn0bEqt\nF2sZSAV2AR5qS6gBUkrTI+IBsnmzO02qS0kpfRIRc4FPqhumpKY1YQLccENh2bhxsO669YlHkiRJ\nDafippaI6B8RX4iIb0TEiGoGVaYRwFMlyqdQ5ijkEbFY7jhWzj0zvi5wYRVjlNSs3n0XDjussGyT\nTeDHP65PPJIkSWpIXSbVEbF1RJwfEZ8uKl8TeBS4H7gemBwRv6ldmCUNovRo43OAFcrcxy+Aj4FX\ngZ8Ae6eU/lqd8CQ1tRNPhJdfbl/u1y8b7bt/T56akSRJUqvrrqX6u8D2KaXXi8qvAjYA/kE2+vbT\nwP4RsX+1A6yx84DNgJ2B24HrImKnzipHxMERMSkiJs2ePbu3YpTU2yZNggsuKCw7+mjYeOP6xCNJ\nkqSG1V1S/TngrvyCiFgf2BL4W0ppy5TS0bl6U4Hv1CTK0t6idIt0Zy3YHaSUXk4pTUop/TGl9E3g\nIeCXXdS/NKU0KqU0avDgwRUFLanBffIJjB0LCxe2l625JpzU2ayCkiRJ6su6S6pXJkuW821NNur3\n5W0FKaUPgOuADasZXDemkD1XXWw4Wct5JSYBa1cckaTmd8EF8NhjhWW//jUsvXR94pEkSVJD6y6p\nXgL4oKhss9zP+4rKXwKWq0ZQZboVGB0RQ9sKImINYIvcuh6JiMWALwL/rlJ8kprNzJnZs9T59t4b\ndtihPvFIkiSp4XU34s5MOrYGfxF4PaX0UlH50sDb1QqsDJcBhwETIuIEstbzU8mS+0vaKkXE6mSJ\n8ikppVNyZSeTdRN/AHiNrEX+e2Td2L/de4cgqaEceSTMm9e+vPzycO659YtHkiRJDa+7lur7ge9E\nxEiAiNgdWIdsUK9iGwCvVDe8zqWU5gFjgOeBa4DxwHRgTErpvbyqAfSj8FgfA0YCF5A9M/4L4ENg\ny5TS9bWPXlLD+fOf4ZZbCsvOPBNWXrk+8UiSJKkpREqp85XZ1FlPAUsCbwIrkk1BtWlKaUpevX5k\nLcS/TykdXtOIG8SoUaPSpEmT6h2GpGr48EMYORL+nff0x+jR8MADsFh33z1KkiSpFUXEoymlUd3V\n60bm0vUAACAASURBVPJuMaU0HdgK+DNZUn07sHV+Qp2zTW79hMrClaQ6+sUvChPqxRaDiy4yoZYk\nSVK3unummpTSJLJ5nLuqM5Gs+7ckNZcXXoAzzigs+8EPYJNN6hOPJEmSmorNMJL6th/+MOv+3Wbw\nYDjttPrFI0mSpKZiUi2p77rttuyV76yzslG/JUmSpDKYVEvqm+bPz1qp822xBey3X33ikSRJUlMy\nqZbUN517bvY8dRsHJ5MkSVIFvHuU1PfMmtXxuekf/AA22qg+8UiSJKlpdTv6tyQ1rblzs2ek58+H\ngw6CddfNyo89FubNa6+3wgowblx9YpQkSVJTM6mW1JpSgq9/He6+O1u+4AI44QTYaiu4+urCuqec\nAiuu2PsxSpIkqemZVEtqTX/4Q3tCDVlr9YkndnxmesQIOOSQ3o1NkiRJLcNnqiW1no8/hmOOKb1u\n4cLC5V/9Cvr7/aIkSZIqY1ItqfVceilMndp9vd13h223rX08kiRJalkm1ZKa18MPw9e+Bt/6Fjzx\nRFb2zjsdBx3bd1848MDCsiWWgF/+snfilCRJUsuyz6Ok5jRnDuy4Y/YT4MYb4aijYMECmD27vd4y\ny2QjgK+8Muy3XzaV1pw5cOqpMHRofWKXJElSyzCpltSczjqrPaGGLJk+66yO9X7ykyyhBth66+wl\nSZIkVYndvyU1n9deg/PP777eyivDj39c+3gkSZLUZ5lUS2o+Z5wB77/ffb1x42DZZWsfjyRJkvos\nk2pJtTNvHjzySDbFVbXMnAkXX1xY9t//DUceCRHtZRts0HFwMkmSJKnKfKZaUm38+98wejS88Qas\nsgpcdlk2UveiOvVU+Oij9uUhQ7IBypZYAvbZBy64IJt3+tRTnX9akiRJNecdp6TaOOaYLKEGePVV\n2GknOOAAOPdcWG65yvY5bRpceWVh2UknZQk1wGabwdVXVx6zJEmS1EN2/5ZUfdOnw803dyy/8koY\nORLuuquy/Z58cjbKd5u114b9969sX5IkSVIVmFRLqr7zz4eFC0uve/ll2H57GDsW3n23/H0+8wxc\nd11h2bhxdvGWJElSXZlUS6quuXPh8ssLy1ZcsWO9Sy+FDTeEe+4pb7+nngoptS+PHAl77115nJIk\nSVIVmFRLqq7LL4f33mtf/sxn4IUX4Gc/g8WK/uTMmAFjxsBhh8E773S+z2eegeuvLyw7+eSO+5Mk\nSZJ6mXekkqrnk0+yrt/5DjsMBg6En/8c/v53WHfdjttddBEMGwY33FDYGt3mtNMKyzfYAHbfvbqx\nS5IkSRUwqZZUPTffnM0j3WbJJeGQQ9qXN98cnngCfvSjwjmlAWbNgr32gh13zKbjavPss/Db3xbW\nPekkW6klSZLUELwrlVQdKcHZZxeW7b8/rLRSYdlSS8E558B998Faa3Xcz513wogR2TPU8+d3bKUe\nORK+/vXqxy9JkiRVwKRaUnU8+CA8/HBh2Q9/2Hn9LbeEyZPhuONgwIDCdfPnZ63RI0faSi1JkqSG\n5p2ppOo477zC5a99DdZfv+ttll4aTj8dnnwSttqq4/pp0wqn5hoxAvbYY9FjlSRJkqrEpFrSonvl\nlex56nw/+lH52w8blk2tdfXVMHhw5/VOPNFWakmSJDUU704lLbpLLoEFC9qXhw/PpsrqiQjYbz94\n7jkYO7bjQGbDh8Oeey56rJIkSVIVmVRLWjTz52dJdb7DDuuYFJdrhRXg4ovhH/+AjTfOypZeGi67\nDPr1W7RYJUmSpCrrX+8AJDW5m26C119vXx44MGtxXlSjR8Njj8HUqfAf/wHLLrvo+5QkSZKqzKRa\n0qK58MLC5QMOqF4CHAHrrludfUmSJEk1YPdvSZWbNAkeeqiw7NBD6xOLJEmSVAcm1ZIqV9xKvf32\ntixLkiSpT2nqpDoiVouImyJibkS8ExE3R8SQMrYbFRGXRsSzEfF+RMyMiPERsWZvxC21hNmz4frr\nC8sOO6w+sUiSJEl10rRJdUQsDdwNrA/sD+wHrAPcExHLdLP53sAI4HxgR+AYYBNgUkSsVrOgpVZy\n+eXZyN9thg6FHXesXzySJElSHTTzQGUHAUOB9VJK0wAiYjIwFRgLnNPFtmemlGbnF0TEA8D03H5P\nqknEUqtYuLDjNFqHHuqUV5IkSepzmralGtgFeKgtoQZIKU0HHgB27WrD4oQ6V/YiMBtYtcpxSo3p\njjtgvfVgo43g2mshpfK3vesuePHF9uUll8xG/ZYkSZL6mGZOqkcAT5UonwIM7+nOImIY8GngmUWM\nS2p8778P++wDzz8Pkydn80rvsgu88kp52xe3Un/zmzBoUPXjlCRJkhpcMyfVg4C3SpTPAVboyY4i\noj9wMVlL9RVd1Ds4IiZFxKTZszs0dkvNY8IEmDOnsOyPf4QRI+DKK7tutZ41C267rbBs7NjqxyhJ\nkiQ1gWZOqqvpQuALwL4ppVKJOgAppUtTSqNSSqMGDx7ce9FJ1TZ+fOnyuXPhwAPhK1+BadNK17ni\nCliwoH155EjYfPPqxyhJkiQ1gWZOqt+idIt0Zy3YJUXEz4GDgQNTSndVKTapcc2eDXfe2XWdiROz\nZPn00+Gjj9rLFyyAyy4rrHvwwRBR/TglSZKkJtDMSfUUsueqiw0Hni5nBxFxPPAz4IiU0jVVjE1q\nXDfcAJ980r48bBicdx4stVRhvfnz4YQTYOON4b77srI77oCXXmqvs9RS2fPYkiRJUh/VzEn1rcDo\niBjaVhARawBb5NZ1KSKOAE4Djk8pXVijGKXGU9z1e9994cgjswHLxozpWP+ZZ2DrreFb34Kzzy5c\nt9desPzyNQtVkiRJanSRejKNTgOJiGWAJ4EPgBOABJwKfArYMKX0Xq7e6sC/gVNSSqfkyvYGrgPu\nBMYV7fqdlFK3Ld2jRo1KkyZNqtLRSL3k3/+GtdcuLJs+HdZYI/t3Stn0WkcdBW+80f3+HnwQRo+u\nepiSJElSvUXEoymlUd3Va9qW6pTSPGAM8DxwDTAemA6MaUuocwLoR+Gx7pAr3wF4sOj165oHL/WW\nN98s7Op93XWF67/4xfaEGrJno/fbD559NhuwrCsbbgif/3zVQpUkSZKaUdMm1QAppZkppT1SSgNT\nSp9KKe2WUppRVGdGSilSSifnlX03V1bqtXUvH4ZUG0cdBYMHw0orwS9/mSXX115bWGfffUtvu+KK\n2Sjff/979kx1KWPHOkCZJEmS+rym7f5db3b/VkN77DHYdNPCsnXXheefb18eMABeew0GDep6X20j\nfh9/fPvc1quuCk8/DQMHVjduSZIkqUGU2/27f28EI6mX3XBDx7L8hBrgq1/tPqEG6NcPDjkEvvEN\nuPRSmDULfvADE2pJkiQJk2qp9aQEN93Ufb3Oun53ZsUV4dhjK4tJkiRJalFN/Uy11NLuvReOOw7u\nvrtn2z3xRDbKd5tSzz0PHAg77bRI4UmSJEkyqZYa0z//CdtuC2eckf389rfbn2fuTnEr9U47wZ/+\nBKut1l52wgmw5JLVi1eSJEnqo0yqpUb061/DwoXty7/9LWywAdx+e9fbpQQ33lhYtuee2fPTzzwD\nt9wC998PRx9d/ZglSZKkPsikWmo0CxfCHXd0LJ81K0uODz4Y5s4tve1TT8HUqe3LAwbAzjtn/15m\nGdh112xuaqfCkiRJkqrCpFpqNI8/Dq+/3vn6yy6D4cNhwoSO64q7fm+3HaywQnXjkyRJkvS/TKql\nRlPcxfuzn4XFii7VWbNgt92yaa5efbW9vLjr9ze+UZsYJUmSJAEm1VLjKe76fdJJ8I9/wHrrdax7\n000wbBhccAFMnpw9N92mf/+su7ckSZKkmjGplhrJW2/Bgw8Wlu24I3z+89lUWccdlyXL+ebOhSOO\nyJ6VzjdmDAwaVNt4JUmSpD7OpFpqJH/5S+Go3yNHZt2/IZsC6/TT4dFHYbPNOm777ruFy3vuWbs4\nJUmSJAEm1VJjKX6eescdO9bZcMOsNfu88+BTnyq9n379YPfdqx+fJEmSpAIm1VKjKDWVVqmkGrKk\n+cgj4bnnYJ99Oq7femtYaaWqhyhJkiSpkEm11CgmT4bXXmtfXnZZ2GKLrrdZZRW49lq4996sqzhk\nrddnnVWzMCVJkiS16999FUm9orjr97bbwuKLl7ftVltlSfm0abDqqrD00tWPT5IkSVIHJtVSoyjn\neequRMA661QvHkmSJEndsvu31AjefjubizpfT5NqSZIkSb3OpFpqBBMnwoIF7cvDh8OQIfWLR5Ik\nSVJZTKqlRnDnnYXLO+xQnzgkSZIk9YhJtdQIJk4sXN5++/rEIUmSJKlHTKqlenvhBZgxo315iSVg\nyy3rFo4kSZKk8plUS/VW3Eq9xRaw1FL1iUWSJElSj5hUS/VWnFRvt1194pAkSZLUYybVUj0tXAh/\n/WthmUm1JEmS1DRMqqV6euIJmDOnfXn55WGTTeoXjyRJkqQeMamW6qm4lXqbbaBfv/rEIkmSJKnH\nTKqlevJ5akmSJKmpmVRL9fLhh3D//YVlJtWSJElSUzGplurlwQfhgw/al1dbDdZZp37xSJIkSeox\nk2qpXkp1/Y6oTyySJEmSKmJSLdVLcVK97bb1iUOSJElSxUyqpXp4+22YNKmwzKRakiRJajom1VI9\n3HsvLFzYvjxyJKy8ct3CkSRJklQZk2qpHpxKS5IkSWoJTZ1UR8RqEXFTRMyNiHci4uaIGFLmtv8d\nEXdFxJsRkSLiuzUOV2p3772Fy3b9liRJkppS0ybVEbE0cDewPrA/sB+wDnBPRCxTxi4OB5YC/liz\nIKVS3ngDpkxpX15sMdhyy/rFI0mSJKli/esdwCI4CBgKrJdSmgYQEZOBqcBY4Jxutl8upbQwItYG\nvlPTSKV8f/tb4fLGG8Nyy9UnFkmSJEmLpGlbqoFdgIfaEmqAlNJ04AFg1+42Tikt7K6OVBP33Ve4\nvNVW9YlDkiRJ0iJr5qR6BPBUifIpwPBejkUqn0m1JEmS1DKaOakeBLxVonwOsEIvx6JWNGcOfPxx\ndff51lsweXL7coTPU0uSJElNrJmT6l4XEQdHxKSImDR79ux6h6NaGjcOPv1pGDgQjjgCXn21Ovu9\n/35IqX15gw1g0KDq7FuSJElSr2vmpPotSrdId9aCvchSSpemlEallEYNHjy4Fr9CjWDWLDj1VFiw\nAD78EC64ANZaC37602zk7kVh129JkiSppTRzUj2F7LnqYsOBp3s5FrWSRx7JEup8H3wAZ50Fa66Z\nJdezZlW2b5NqSZIkqaU0c1J9KzA6Ioa2FUTEGsAWuXVSZZ58svN1773XnlwffDBMnVr+fufOhccf\nLyz70pcqi1GSJElSQ2jmpPoyYAYwISJ2jYhdgAnAS8AlbZUiYvWI+CQiTsrfOCK2iog9gR1yRaMi\nYs9cmfqyJ54oXB4woGOdjz6Cyy6D9daD3XaDe+4pfFa6lAcegIV5M7kNGwY+RiBJkiQ1taZNqlNK\n84AxwPPANcB4YDowJqX0Xl7VAPrR8VjHATcCF+SW/zO3fGMNw1YzKE6qJ06Es8/OBi4rlhJMmABj\nxsCGG2aJ9rx5pfdr129JkiSp5UTqrnVNJY0aNSpNmjSp3mGo2ubOheWXb19ebLGsy/dSS2XPVV95\nZdb9e8aMzvcxcCB8+9tw0EGwySbt5aNHwz//2b7829/C3ntX/RAkSZIkLbqIeDSlNKq7ek3bUi3V\nRP4c0gDrr58l1JD9PPTQ7Dnq8eNho41K7+Odd+Dii2HTTWHUKLjoInjxRSj+EsaWakmSJKnpmVRL\n+YoHKSuVOPfvn7VEP/541qV7jz2yFu1SHn0UDjssG9gsf0TxddaBVVapXtySJEmS6sKkWspX/Dz1\nxht3XjciG737pptg+nQ45hj4zGdK1y1+zMJWakmSJKklmFRL+YqT6s66eBcbMgTOOANeegl+/3vY\nYYcs6e6MSbUkSZLUEkyqpTaffAJPPVVY1lVLdSkDBsDXvw63354NZnb66dm0W/mWXx523HGRQpUk\nSZLUGEyqpTbPPQfz57cvf+YznXfnLseQIXDccfDMM/Dww/DjH8N++8Edd8CKKy56vJIkSZLqrn+9\nA5AaRvEgZT1tpe5MBGy2WfaSJEmS1FJsqZba9GSQMkmSJEnCpFpqV+kgZZIkSZL6LJNqCbIpr2yp\nliRJktRDJtVqTdOmwYUXwsSJsGBB9/Vfew1mz25fXnJJWGed2sUnSZIkqSU4UJlaz4wZMHJk+0je\nQ4bAAQdkr9VXL71N8SBlG2wA/b08JEmSJHXNlmq1nuuuK5waa+ZMGDcO1lwTvvxl+M1v4O23C7fx\neWpJkiRJFTCpVut57rnS5Sll3cG/971s/unddoPf/hbmzvV5akmSJEkVsX+rWs+0ad3X+egjmDAh\new0YAP36Fa63pVqSJElSGWypVuuZOrVwedw42HTTzut//DF8+GFh2YYbVj8uSZIkSS3HpFqtZe7c\nwlG8BwyA446DSZNgyhQ48URYe+2u97HWWjBwYG3jlCRJktQSTKrVWoq7fg8d2j6K9/DhcMop8Pzz\n8Mgj8JOflJ42a+edax+nJEmSpJbgM9VqLcVdv0u1SkfAqFHZ68wz4dln4ZZb4KGHsiT85JN7JVRJ\nkiRJzc+kWq2luKW6VEt0vggYNix7SZIkSVIP2f1braWclmpJkiRJqhKTarWWnrZUS5IkSdIiMKlW\nayluqTapliRJklRDJtVqHaWm01pttfrFI0mSJKnlmVSrdXQ1nZYkSZIk1YBJtVqHg5RJkiRJ6mUm\n1WodDlImSZIkqZeZVKt1OEiZJEmSpF5mUq3WYfdvSZIkSb3MpFqtw+7fkiRJknqZSbVag9NpSZIk\nSaoDk2q1BqfTkiRJklQHJtVqDQ5SJkmSJKkOTKrVGopbqh2kTJIkSVIvMKlWa7ClWpIkSVIdmFSr\nNTidliRJkqQ6aOqkOiJWi4ibImJuRLwTETdHxJAyt10yIs6KiFcj4oOIeDAivlTrmFUjTqclSZIk\nqQ6aNqmOiKWBu4H1gf2B/YB1gHsiYpkydnEFcBBwErAT8CpwZ0RsXJuIVTOlptMaUtZ3K5IkSZK0\nSJp5zqGDgKHAeimlaQARMRmYCowFzulsw4jYCPg2cGBK6cpc2X3AFOAUYJfahq6qKjWdVr9+9YlF\nkiRJUp/SzEn1LsBDbQk1QEppekQ8AOxKF0l1btuPgd/lbftJRFwPHBMRS6SU5tco7traeWf4y1/q\nHUXvWrCgcNmu35IkSZJ6STMn1SOACSXKpwDfKGPb6Sml90tsuziwdu7fzefjj2F+c34fUDUOUiZJ\nkiSplzTtM9XAIOCtEuVzgBUWYdu29R1ExMERMSkiJs3Of4ZXjWXUqHpHIEmSJKmPaOakutellC5N\nKY1KKY0aPHhwvcNRsf794ZvfhD32qHckkiRJkvqIZu7+/RalW6Q7a4Uu3nb1TraF9hbr5nPrrbBw\nYb2jqI9+/bKRvyVJkiSplzRzUj2F7NnoYsOBp8vYdveIWLrouerhwEfAtNKbNYHFF693BJIkSZLU\nZzRz9+9bgdERMbStICLWALbIrevKbcAA8gY0i4j+wF7AXU078rckSZIkqVc1c1J9GTADmBARu0bE\nLmSjgb8EXNJWKSJWj4hPIuKktrKU0uNk02mdFxHfj4htgeuBNYH/6sVjkCRJkiQ1saZNqlNK84Ax\nwPPANcB4YDowJqX0Xl7VAPrR8VgPAK4ETgP+BKwG7JBSeqzGoUuSJEmSWkQzP1NNSmkm0OVQzyml\nGWSJdXH5B8BRuZckSZIkST3WtC3VkiRJkiTVm0m1JEmSJEkVMqmWJEmSJKlCJtWSJEmSJFXIpFqS\nJEmSpAqZVEuSJEmSVCGTakmSJEmSKmRSLUmSJElShUyqJUmSJEmqkEm1JEmSJEkVMqmWJEmSJKlC\nJtWSJEmSJFUoUkr1jqEpRcRs4MV6x9GFlYA36h2EOvC8NCbPS2PyvDQmz0tj8rw0Js9LY/K8NKZG\nPC+rp5QGd1fJpLpFRcSklNKoesehQp6XxuR5aUyel8bkeWlMnpfG5HlpTJ6XxtTM58Xu35IkSZIk\nVcikWpIkSZKkCplUt65L6x2ASvK8NCbPS2PyvDQmz0tj8rw0Js9LY/K8NKamPS8+Uy1JkiRJUoVs\nqZYkSZIkqUIm1ZIkSZIkVcikuoVExGoRcVNEzI2IdyLi5ogYUu+4+oqI2DMifh8RL0bEBxHxXESc\nERGfyquzRkSkTl7L1zP+VhURW3fyfr9dVG+FiLg8It6IiHkRMTEiNqhX3K0uIu7t4lq4I1fH66XG\nIuKzEXFBRDwYEe/n3ts1StRbMiLOiohXc3/fHoyIL5Wot1hEHBsRMyLiw4h4MiL26I1jaSXlnJeI\nGBURl0bEs7k6MyNifESsWWJ/Mzq5jnbrrWNqBT24Xjr7u7VxUT2vlyoo83o5uYvz8mFRXa+XRVTO\nPXGuXln3XuV+BtVT/3oHoOqIiKWBu4H5wP5AAk4D7omIDVNK8+oZXx9xNDATOA54Gfg/wMnANhHx\nhZTSwry6ZwC3Fm3/bm8E2YcdATySt/xJ2z8iIoDbgDWAw4G3gGPJrp+NU0ov92KcfcWhwMCiss2B\nc+h4bXi91M7awDeBR4H7ga90Uu8K4GvAT4AXgP8E7oyIzVNKT+TVO5Xsb+HxuX3uDdwYETullP5c\nm0NoSeWcl72BEcD5wBRgVeBEYFLu79ZLRfXvJPtMyvdcFWPuC8q9XgCuAi4pKnu+aNnrpTrKOS+X\nA3cUlS2TKyv+fAGvl0XV7T1xD++9yv0Mqp+Ukq8WeAFHAguAtfPK1iRLHI6qd3x94QUMLlH2HbIv\nOMbkltfILX+/3vH2lRewde49366LOrvm6myTV7YcMAc4v97H0FdeZB+a84FBuWWvl9q/54vl/fv7\nufd7jaI6G+XKD8gr6092g3lrXtmnc+dvXNH2fwUm1/tYm+lV5nkp9ZmzOrAQOKWofAZwbb2Pq9lf\n5ZyX3LoEnNbNvrxeevm8lNhuv1zdrxWVe70s+jkp5564rHuvcj+D6v2y+3fr2AV4KKU0ra0gpTQd\neIDsP61qLKU0u0RxW8voqr0Zi3psF2BWSumetoKU0lyyb1C9fnpBrrfNN4DbUkpz6h1PX5EKe9B0\nZhfgY+B3edt9AlwPbB8RS+SKtwcWB64t2v5aYINS3ZJVWjnnpdRnTkrpRWA2fubURJnXS7m8Xqpk\nEc7L/sD/I2uVVhWVeU9c7r1XuZ9BdWVS3TpGAE+VKJ8CDO/lWNRuq9zPZ4rKz4iITyJ7/v3WUs+P\nqOrGR8SCiHgzIq6LwvEGurp+hkTEsr0TYp+2O/Ap4H9KrPN6qa8RwPSU0vtF5VPIkoK18+rNB6aV\nqAd+FtVcRAwjawEt/swB2Dn3vOn8iHjI50Nr7ge59/r9iLg7IrYsWu/1UkcRsRqwDTA+l6AV83qp\nvuJ74nLvvcr9DKork+rWMYjsWYRic4AVejkWARGxKnAKMDGlNClXPJ/sGauxZH/MjwY2AP6RuxlS\n9c0FzibrEjaG7Bm27YAHI+LTuTpdXT/gNdQbvgO8DtyeV+b10hi6uz4G5f18O+X65nVRTzUQEf2B\ni8laqq8oWn0b2TOL2wP7AB8Cf4iIfXs1yL7jWrJxI7YDDgZWBO6OiK3z6ni91Ne+ZHlQqS9yvV6q\nrJN74nLvvcr9DKorByqTaiD37doEsmfaD2grTym9ChySV/X+yEY6nkI2UIl/sKsspfQ48Hhe0X0R\n8TfgYbLBy06oS2D6XxHxH2Q3n7/KbzHwepF65ELgC2TPhxbcgKaUDs9fjog/AA+RDQJY3P1Yiyil\ntF/e4v0RMYGsRe404Iv1iUpFvgM8nlKaXLzC66W6OrsnbjW2VLeOtyjdmtbZtzuqkYhYiuxbzqHA\n9qmbkaNTNkLr34HNeiE8ASmlx8hGYW17z7u6ftrWq3a6ajEo4PVSF91dH3Py6i2fG9G1q3qqsoj4\nOVmL6IEppbu6q59SWgDcCHw2IlapdXx9XUrpXeBPFP7d8nqpk4j4HLA+ZXzmgNfLoujmnrjce69y\nP4PqyqS6dUwhe+ag2HDg6V6Opc+KiAHATcAo4KsppX/1YPPiLmCqvbb3vKvrZ2ZK6b3eC6lP2h94\nMqX0ZA+28XrpPVOANXODyeUbDnxE+zOhU4AlgLVK1AM/i2oiIo4HfgYckVK6poJdeC31nvz32uul\nfvYnG/jqugq29XopUxn3xOXee5X7GVRXJtWt41ZgdEQMbSvITXy/BaXn31OVRcRiwHiy53Z3Syk9\nVOZ2Q8i6gz1cw/CUJyJGAevR/p7fCqwaEVvl1RkI7IzXT03lzsVwymwx8Hqpi9uAAWSjswP/+/zu\nXsBdKaX5ueI7yG5U9ynafl/gqdyMFKqiiDiCrEvx8SmlC3uwXdv5m5lSeq1W8SmT+zzZicK/W14v\ndRARi5PNB357JyNUl9rG66WHyrwnLvfeq9zPoLrymerWcRlwGDAhIk4g+ybtVOAlsoF+VHsXkV3w\npwPzImJ03rqXU0ovR8TZZF9mPUg2mMx6ZBPdL8xtpyqLiPHAdOAx4G3g/5C9568A5+eq3Up2Tq6N\niJ+QdTU6FgjgF70dcx/zHbLnrMYXr/B66R0RsWfun5vmfu4YEbOB2Sml+1JKj0fE74Dzci0P04Ef\nAGuSlxCklF6PiHOAYyPiXbJrbi+ym6pdeulwWkZ35yUi9gbOI0vO7i76zHknpfR0bj/fIpue5s9k\n9wSfAf4T2AT4Vu2PpLWUcV6OJvtbdQ8wi2zu8KOBlfF6qZnuzkte1Z3Iug2X/CLX66Vqur0npsx7\nr3I/g+qu3hNl+6reCxgC/B54B3gXuAVYo95x9ZUXMIPsy4xSr5NzdQ4km6fvLbJvqF8j6360Xr3j\nb9UX2R/oyWSjgH9M9iF5KbBKUb1BwG/Ins15H/grsFG942/lF9k3z7PJ5qYutd7rpXfOQ2d/wZAa\ngwAACKlJREFUt+7Nq7MUcE7uHHwI/BPYusS++pEN/vci2ejtk4E9632Mzfjq7rwAV5V57kYDd5PN\nx/sx2ZeLE8meb6z7cTbbq4zzsjPwAPBG7v1+kyx5+FyJfXm99NJ5yas3IXdOFu9kP14v1TkfM7o4\nJyfn1Svr3qvcz6B6viIXqCRJkiRJ6iGfqZYkSZIkqUIm1ZIkSZIkVcikWpIkSZKkCplUS5IkSZJU\nIZNqSZIkSZIqZFItSZIkSVKFTKolSZIkSaqQSbUkSZIkSRUyqZYkSZIkqUIm1ZIkSZIkVcikWpIk\nSZKkCplUS5IkSZJUIZNqSZIkSZIqZFItSZIkSVKFTKolSZIkSaqQSbUkSZIkSRUyqZYkSZIkqUIm\n1ZIkSZIkVcikWpIkSZKkCplUS5IkSZJUIZNqSZIkSZIqZFItSZIkSVKFTKolSepGRMyIiHtbOY6I\nuCoiUi323aoa5f+FJKm+TKolSX1SRAyNiEsj4tmIeD8i3oqIZyLifyJim3rHVwsR8d2I+GG94+hM\nLr7UxWtavWOUJKlY/3oHIElSb4uIUcB9wMfA1cAUYClgHeArwLvAPXmbrAe0Qivud4E1gPNKrDsI\nOKQ3g+nC+cAjJcrf7e1AJEnqjkm1JKkv+i9gaWDjlNKTxSsjYuX85ZTS/N4KrF5SSh+TfcnQCO5P\nKd1U7yAkSSqH3b8lSX3ROsCbpRJqgJTSa/nLpZ6dbSuLiI0iYmJEvBcRr0fE2RHRPyKWjIhfRsQr\nEfFhRPwtIoYV7ePkXLfmNYpjKPd53Yj4SkT8LiJeiIgPIuLtiLgrIrYq3h+wFbB6UZfqrXPrSz5T\nHREbRsQfIuLN3HE8HRE/jYh+RfWuyu1vuYj4v7n34sOIeCAiPt/dcfRURGyd+33fjYjDI+L53O97\nPiIO72SbL0XEXyJibu69eiwivtdJ3bUj4sqIeDkiPoqIWRExISI2LVF3/Yj4U0S8m9v3TcVfzEiS\nWpct1ZKkvujfwHoR8fWU0s2LsJ/PAn8BfgfcRNZ1/CjgE2AEWZfynwMrAUcDt0TEsJTSwkUJvsh3\ngUFk3dhfBlYFvg/8NSK2SSndn6v3Q+CMXCw/ytv+mc52XNRN/iLgNWBn4ExgI2CfEpvdCcwGTgFW\nJHs//hQRa6aUyu2+/amIWKlE+QcppXlFZYcDKwOXkHUP/xZwfkQMSimNyzuWnYE/5I7h7FzdvYHL\nI2JoSun4ouP+KzAAuAJ4iuw93gr4AvBo3u9fFbg3t++fkL0vY4GBZP8fJEktzqRaktQXnQZ8Gfh9\nREwF/k72DO+9KaVOk8wS1gK+mVK6Mbd8cUQ8SpZc3QZsl1JKABHxJvCr3O+9szqHAcBBxYlmRFxM\n9pz4scD9ACmlW3KDlC2VUrq2zH3/ClgC2DylNDm37wvJvkT4dkT8JqX016JtHkspHZoXy9PADcC3\nyRLfcvymk/KLgMOKytYFhqWUXs79vovIzucJEXFFSunlXKv6hcB7wOdSSrPy6t4DHBMRV6WUpkZE\nAFfljvtzbcedc0ZEFPfyWxvYK6V0Q94xLwQOjYj1UkrPlXnMkqQmZfdvSVKfk1J6ENgU+B9gOeAA\n4NfA07lu2kPL3NUreQl1m78DAVzQllDntLUYr1N55B3lJ9QRsWxErAgsAP4JVNztOiI+TdYqe2t+\nYpk7ptNzi7uX2PTcouW7cz97ctynkH35UPy6oETd8W0JdS6+j3Ix9CdrVYfsXA8BftOWUOfV/QXZ\n/dCuueKNyXoZXFmUULdtU9zLYFZ+Qp1TyTFLkpqULdWSpD4ppfQvsq7TRMTqZF17vw9sCUyIiE1z\nSVdXppcoe6uTdW3lK1YUcCciYi2yJHd7YPmi1YsyYvmauZ9TSqx7BlgIlPry4YWCAFJ6M2v87dFx\n/yulNLHMuqV6Fjyd+9kWX1fHMqWoblsi/HiZv/+FEmVv5n5W9VxLkhqTLdWSpD4vpfRiSulqssT6\nAWAk8LkyNl1QwbrI/9VdbN/tF98RsSzwN2AHsq7ae5Il118may2NzreujZRSOcfdSrr6P9CqxyxJ\nymNLtSRJOSmlFBH/BLYgG4Cq1ubkfg4CZrQVRsSSwCrAtG623xb4D+DAlNKV+Ssi4rQS9XvSct3W\n0j6ixLr1yb6YL9VK29uGlSgbnvv5QtHPUsdSXPf53M+NFz00SVJfYEu1JKnPiYgvR0SHL5YjYina\nR2x+unh9DbQlcNsVlf+I8j6j21pJC1pEI+IrlH6e+j1ghdxgXF1KKb0O/APYOSJG5u07yAZAg2zE\n63rbJyI+27YQEYuTvX8LgD/mih8DZgIH5E91FREDyAaVS8CEXPGTZF3CD4yIDkl4Oe+dJKlvsaVa\nktQXnQusGBG3Av8C3gdWIxuhel3g6twz17U2EXgOOCU3wNh04IvAaOCNMrb/O7kponJzXb9M1sK6\nH9lxbVBU/yFgJ+DCiPgHWeJ5dy6BLuVIsim17s+NlP1abvvtgetKjPxdLVvmWutLGV80ANzzwD9z\nI56/S3YONwNOTSm9BFmX9Ig4jOxLgEci4tJc3b3I3uv/TilNzdVNEXEA2ZRaD0dE25Ray5M9HnAH\npQdMkyT1USbVkqS+6Ciy0Z6/COxBljDNBSaTzcF8VW8EkUv2dgHOJ5tv+SPgLtqf7e5u+7cjYnuy\nEawPJ/tcfxT4KvA9OibV55INyLUncAhZa/g2QMmkOqU0KSK+AIwDDgWWIesm/TOyuZ5r5Ygu1l1P\nNg94mwvI5oQ+nGyE75nAD1NKv8rfKKV0W0RsC5xA1jq9ONkgZ99PKV1RVPeRiNgMOBH4Jtl79Qbw\nMGWcF0lS3xKFX/ZKkiQ1vojYmmyO6QNSSlfVNxpJUl/mM9WSJEmSJFXIpFqSJEmSpAqZVEuSJEmS\nVCGfqZYkSZIkqUK2VEuSJEmSVCGTakmSJEmSKmRSLUmSJElShUyqJUmSJEmqkEm1JEmSJEkV+v8K\nXybRLBH6nQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12b4238f28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_learning_curve(curve)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA+4AAAJBCAYAAADY2e8xAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSaowA3dJkiRJkirs/wPiko2glAOU7gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f129821a5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_loss_curve(losses)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9AAAAJBCAYAAACu4126AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12981a0908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_loss_curve(losses[-100:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "saver = tf.train.Saver()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'trained_model/flight_200/model.ckpt'"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#saver.save(agent.model.sess, \"trained_model/flight_200/model.ckpt\")  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_curve = [0.0,\n",
    " 0.0,\n",
    " 0.0,\n",
    " 0.0,\n",
    " 0.0,\n",
    " 0.0,\n",
    " 0.0,\n",
    " 0.0,\n",
    " 0.1111111111111111,\n",
    " 0.1,\n",
    " 0.09090909090909091,\n",
    " 0.08333333333333333,\n",
    " 0.15384615384615385,\n",
    " 0.21428571428571427,\n",
    " 0.26666666666666666,\n",
    " 0.3125,\n",
    " 0.35294117647058826,\n",
    " 0.3333333333333333,\n",
    " 0.3157894736842105,\n",
    " 0.35,\n",
    " 0.38095238095238093,\n",
    " 0.4090909090909091,\n",
    " 0.43478260869565216,\n",
    " 0.4583333333333333,\n",
    " 0.48,\n",
    " 0.5,\n",
    " 0.5185185185185185,\n",
    " 0.5357142857142857,\n",
    " 0.5517241379310345,\n",
    " 0.5666666666666667,\n",
    " 0.5806451612903226,\n",
    " 0.59375,\n",
    " 0.6060606060606061,\n",
    " 0.6176470588235294,\n",
    " 0.6285714285714286,\n",
    " 0.6388888888888888,\n",
    " 0.6486486486486487,\n",
    " 0.6578947368421053,\n",
    " 0.6666666666666666,\n",
    " 0.675,\n",
    " 0.6829268292682927,\n",
    " 0.6904761904761905,\n",
    " 0.6976744186046512,\n",
    " 0.7045454545454546,\n",
    " 0.7111111111111111,\n",
    " 0.717391304347826,\n",
    " 0.723404255319149,\n",
    " 0.7291666666666666,\n",
    " 0.7346938775510204,\n",
    " 0.74,\n",
    " 0.7450980392156863,\n",
    " 0.75,\n",
    " 0.7547169811320755,\n",
    " 0.7592592592592593,\n",
    " 0.7636363636363637,\n",
    " 0.7678571428571429,\n",
    " 0.7719298245614035,\n",
    " 0.7758620689655172,\n",
    " 0.7796610169491526,\n",
    " 0.7833333333333333,\n",
    " 0.7868852459016393,\n",
    " 0.7903225806451613,\n",
    " 0.7936507936507936,\n",
    " 0.796875,\n",
    " 0.8,\n",
    " 0.803030303030303,\n",
    " 0.8059701492537313,\n",
    " 0.8088235294117647,\n",
    " 0.8115942028985508,\n",
    " 0.8142857142857143,\n",
    " 0.8169014084507042,\n",
    " 0.8194444444444444,\n",
    " 0.821917808219178,\n",
    " 0.8243243243243243,\n",
    " 0.8266666666666667,\n",
    " 0.8289473684210527,\n",
    " 0.8311688311688312,\n",
    " 0.8333333333333334,\n",
    " 0.8354430379746836,\n",
    " 0.8375,\n",
    " 0.8395061728395061,\n",
    " 0.8414634146341463,\n",
    " 0.8433734939759037,\n",
    " 0.8452380952380952,\n",
    " 0.8470588235294118,\n",
    " 0.8488372093023255,\n",
    " 0.8505747126436781,\n",
    " 0.8522727272727273,\n",
    " 0.8539325842696629,\n",
    " 0.8555555555555555,\n",
    " 0.8571428571428571,\n",
    " 0.8586956521739131,\n",
    " 0.8602150537634409,\n",
    " 0.8617021276595744,\n",
    " 0.8631578947368421,\n",
    " 0.8645833333333334,\n",
    " 0.865979381443299,\n",
    " 0.8673469387755102,\n",
    " 0.8686868686868687,\n",
    " 0.87,\n",
    " 0.8712871287128713,\n",
    " 0.8725490196078431,\n",
    " 0.8737864077669902,\n",
    " 0.875,\n",
    " 0.8761904761904762,\n",
    " 0.8773584905660378,\n",
    " 0.8785046728971962,\n",
    " 0.8796296296296297,\n",
    " 0.8807339449541285,\n",
    " 0.8818181818181818,\n",
    " 0.8828828828828829,\n",
    " 0.8839285714285714,\n",
    " 0.8849557522123894,\n",
    " 0.8859649122807017,\n",
    " 0.8869565217391304,\n",
    " 0.8879310344827587,\n",
    " 0.8888888888888888,\n",
    " 0.8898305084745762,\n",
    " 0.8907563025210085,\n",
    " 0.8916666666666667,\n",
    " 0.8925619834710744,\n",
    " 0.8934426229508197,\n",
    " 0.8943089430894309,\n",
    " 0.8951612903225806,\n",
    " 0.896,\n",
    " 0.8968253968253969,\n",
    " 0.8976377952755905,\n",
    " 0.8984375,\n",
    " 0.8992248062015504,\n",
    " 0.9,\n",
    " 0.9007633587786259,\n",
    " 0.9015151515151515,\n",
    " 0.9022556390977443,\n",
    " 0.9029850746268657,\n",
    " 0.9037037037037037,\n",
    " 0.9044117647058824,\n",
    " 0.9051094890510949,\n",
    " 0.9057971014492754,\n",
    " 0.9064748201438849,\n",
    " 0.9071428571428571,\n",
    " 0.9078014184397163,\n",
    " 0.9084507042253521,\n",
    " 0.9090909090909091,\n",
    " 0.9097222222222222,\n",
    " 0.9103448275862069,\n",
    " 0.910958904109589,\n",
    " 0.9115646258503401,\n",
    " 0.9121621621621622,\n",
    " 0.912751677852349,\n",
    " 0.9133333333333333,\n",
    " 0.9139072847682119,\n",
    " 0.9144736842105263,\n",
    " 0.9150326797385621,\n",
    " 0.9155844155844156,\n",
    " 0.9161290322580645,\n",
    " 0.9166666666666666,\n",
    " 0.9171974522292994,\n",
    " 0.9177215189873418,\n",
    " 0.9182389937106918,\n",
    " 0.91875,\n",
    " 0.9192546583850931,\n",
    " 0.9197530864197531,\n",
    " 0.9202453987730062,\n",
    " 0.9207317073170732,\n",
    " 0.9212121212121213,\n",
    " 0.9216867469879518,\n",
    " 0.9221556886227545,\n",
    " 0.9226190476190477,\n",
    " 0.9230769230769231,\n",
    " 0.9235294117647059,\n",
    " 0.9239766081871345,\n",
    " 0.9244186046511628,\n",
    " 0.9248554913294798,\n",
    " 0.9252873563218391,\n",
    " 0.9257142857142857,\n",
    " 0.9261363636363636,\n",
    " 0.9265536723163842,\n",
    " 0.9269662921348315,\n",
    " 0.9273743016759777,\n",
    " 0.9277777777777778,\n",
    " 0.9281767955801105,\n",
    " 0.9285714285714286,\n",
    " 0.9289617486338798,\n",
    " 0.9293478260869565,\n",
    " 0.9297297297297298,\n",
    " 0.9301075268817204,\n",
    " 0.93048128342246,\n",
    " 0.9308510638297872,\n",
    " 0.9312169312169312,\n",
    " 0.9315789473684211,\n",
    " 0.9319371727748691,\n",
    " 0.9322916666666666,\n",
    " 0.9326424870466321,\n",
    " 0.9329896907216495,\n",
    " 0.9333333333333333,\n",
    " 0.9336734693877551,\n",
    " 0.934010152284264,\n",
    " 0.9343434343434344,\n",
    " 0.9346733668341709,\n",
    " 0.935]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9UAAAJdCAYAAAA83lQeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xe4JGWZ9/HvPcOEMzkPcZghDjkNQTAAguKKIiAqKgoq\nRlZQcd119QUT6i6LCcVFdEFERUURURRBBiRLTpIZBgYGZpjI5HC/f9xP0XXqVPfp06dP/n2uq67q\nrvhU6Oq66wll7o6IiIiIiIiIdNygnk6AiIiIiIiISF+loFpERERERESkQQqqRURERERERBqkoFpE\nRERERESkQQqqRURERERERBqkoFpEeoSZTTczN7N++woCM7swbeOZPZ0Wkc4ys4PM7CozW2RmG9O5\nfXpPp0vaMrPf6Pg0h5mdkvbllT2dFmmOrjimZrZrWubLzVqm9C0KqkVyzGyEmX3MzP5gZnPNbKWZ\nrTCzp9JNynvNrKWn0ylSi5kdbGZnmtnbejotfY2ZnZg97GmgO7Gn099VzGx34G/AEcBYYCHwArCi\nJ9OVyd0k64ZWeoVckFXs1prZfDP7s5m938z6/L24mZ1d2MYd2pn+oML0evgjfd4mPZ0Akd7CzN4C\nnA9smhu8AtgITE/dscA3zewEd/9bd6exn1kHPNLTieinDgbOAC4CLu/ZpPQ5q4hgsWgwMCl9Xgys\nrTJvf/UxYChwNfB2d1/ew+mR2uYR19eXejohAsRxWJ8+jwSmAm9M3XvN7Eh3X9NTiesC7wO+UGP8\n+7srISLdpc8/HRNphpTDdDkRUD8CnABMcvdR7j4GGAe8HZgNbA68tmdS2n+4+zx3n+nuM3s6LSIZ\nd7/U3TctdsC+ucmOKZvG3S/tqXR3g11S/xcKqHs/dz81XV//r6fTIgAckbtOjAa2An6cxh0GfK7n\nktZUc1P/vWZmZROY2XDgOOIhw3PdlTCRrqagWgY8M9sD+CHxe/gTsJe7/8zdX3nC7+5L3f0ydz8E\neBegm0oRGUiyai8qXi3SSe7+LHAycFMa9L4eTE4zPQbcDmxN9cyHtxIZFX8BFnRTukS6nIJqEfgq\nMIwoLvdud69ZhDPlRp1THG5mw8zs02Z2m5ktNbNVZvaImZ1jZpuWLCpff3N2+n68md1sZsvMbIGZ\n/c7MdspNv5mZfc/M5pjZajN73Mz+3cwGlyy7VUNgqQ7TlWm5K83snlQPsfQ6YGaTzOzjZvZ7M3vY\nzJan+uUPpW3avMp8xfUekOqjP29mG8zs22XTlSznKDP7k5m9YGbrLBpHesTMfmFm76wyTzOOwVvM\n7DozW2JmL5vZrWZ2fNm8HWFmw83sS2lfrjKzF9O2tFf3bGg6Tn9P+2CNmT1tZj/Jnxtp2ulpf56R\nBr2/pD7fdDPbKn1eb2ZjStZ5fxq/vMq59Xwaf3DJuFFm9nkz+0c6BqvN7DEz+66ZbdXOtk5P5/cj\n6RxdbmZ3mtnnzGxklXny2zXNzH5kZs+m/fSURV2/NtvYlaxSv/BcM9vEzD6VtmNpGr5dmq7F4jd/\niZndl47vaovf90UW9ZirrWNhWtYsM5uS9tvTabufMbPvm9mkKvMONrOTzeyGtM516Xx8wMzON7ND\ni+sBZqVBv87t8wdKlr23mf00pWW1mS02s+vN7INVzqVWjfuY2WstrjnzLa4XX+3Y3u+4tD8+YGZ/\nS9u7Np1Dl5jZXlXm2cTM3mpmF5jZ3Wn/rTGzeWb2azN7dY313ZG2+e0W19lz0m9klZk9m5uuM8e4\ntKGykv29t5ldltK/2uL6/jkzq1o90OI3fpbF/89qM3vOzC42sx2Ky+8IM9vFzL6czpfsP26xmd1k\nZv9qZkOrzNeq0SkzO9biernU4hpyk5kd3c66tzaz/7O4tmX/rd8ws1Ed3Y56ubsD16Sv25Ztn5mN\nNbMPpWP0UNqmlWb2aDr+M0rmGWVxbXcz27pk/HVW3/X/7Q1u2k9Tv9qDgvcVpqsq/c4+amY3pnNh\ntZk9YWY/KNu2wrydOqZmdmj6Hc1Lv7mFZvYXMzumnvllAHJ3deoGbAdsQdSZduDfOrGcycBdaTkO\nrAaW5b4vAg4ome/ENH428M30eV1h3oXADsD2wDNp2DKi6FQ2zfdLlj09N/7YtFwn6oOuy437HbBJ\nyfxn56ZZR6VOWDbsRWD3dtb7zty6lhD1UL9dnK5kGV/LLSPb3lW57/O76Bh8MX3ekNKbT8NpDZwX\nF6Z5vw7ckj6vAZbmlrsCeG2V+TcD7slNu6GwTauIosjZ9FsB84ncxGz8/EK3VZr2yTTNmwrrnEjl\nN+HAvoXxO+T27/DCuJ2AOYXz5mVaH4ODqmzrMYVjvCKdL9n3+4CpJfNl448iztHsfMmf4/8AhnTy\nWjE9t7yD25k2++2cR9RBzvbF4vR5uzTduwrHdlHar9mwNcCxVdaxME3zDuKBoKd9vSY3/8PAqMJ8\nRlR1yZ/b2W8z+35NbvoH03mTv35k59LswrI/Vzh3lqXtyr5fBQwtzLNrLu0n5abPrlNfrfP4nJIt\np4PHdTzw91waN9L697kO+EDJfAcU5lmSztn8sE9WWecdaZpTqVzTVxEloJ7t7DFO8/4mjT+9xv4+\nKresJYVj97MqaZ8IPJCbLn+dXUpUnerwcUjLfji33JXE7yF/nv4daKlx7K8EvkH1a/jJVda7F5Xf\npqfjkF2L7gf+I1t+B7dn19wyZ1WZ5ku5acq27d9z49cT17j8b3UpJddUIrfYgRMKw4fR+jpb6/o/\npQPbml3zrknLWJvS1lKYbgrxm1oCDKfy/3Z6yTJHAzfk0lr873wZeEOzjymR4fi9wrmzlNa/jx8D\nVu231dFzX13/6Ho8AerU9WQHvCd3kZzZieVcRSVoOA4YnIbPIoIBJ25CJxXmO5HWN7WnAiPSuN2o\n3GT8FrgNuBnYI40fAfwnlRu4XQvLnp7btiUpjTPSuJHAZ6ncwH6+ZJs+mf54diMF3URjTfsAf07z\nPVDyx5Jf73Li5m56GrdJ7vMr05XMn6XrrPw+IwLnY4Efd9ExWE80rjIujZsK/JrKTe+EDp4XF+aW\nvYK44RySxu0J3JlL1/jCvEOo3BhdA7wqN+9mwLeoBJ/bFuY9M427sI60faMw/GgqAVHZTfnJafgN\nheFjgafSuF8Bu+eOwTbAJbltHVeYd1/i/F9HlBzZIne+vYoIih34S8l2ZOfaYuBa0u+AuHn8AJUg\n9eOdvFZMz63r4HamzW4wl6fj8wHSA4h07Eamz29K0x6YG29pf52XOw5tbm6pBFyLiWvDrNx5c1xa\ntwP/rzDfv6Tha4GP5tIyiHjI+CHgayXrywLBt1fZ5uxauoi4jk3IHYe35M6N/ynMl92Irid+YxcD\nW+a2Zes6j0+Hg+q0r7OHHrcAhwDD0rhJwJeJa9E6YO/CvHsC/wscSi6oTcf3K8Q1eT2wW419uRx4\nPC3D0rjtOnuM0zTtBdXr0vwXUXnQNprWQV6bh33AZVSuacdR+W/Yi3iomQUyjQTVFxANWG2ZG9ZC\nPHDLzp9v1jj22YOY04ExadzmRLCd/ZaKD5lagCfS+H+SHrwS155jifM5C86bGlSn8++mNP75Kss4\nkbie703l+j8oLTs7xnNp+7AquwZdUBj+mty+qHX9f6iD2/pKUJ2+Zw/uji9Md1oa/qP0vVZQ/TMq\n/3EnZttItO+QPaReCkxr5jGlcl/1TDofR6fhI4j/8AVp/L9WOd4Kqgdo1+MJUKeuJzviBt6JG29r\ncBnZn5QDbywZP5XKE/cvF8admJv3jHaWvYhCMJKmuZbym+fpuXkfIN0sFqY5M/fHNKID2zyMyMFy\n4HU11nsjMKjKMl6ZrjD8HdmfYQ8cg/8smbeFyJV34H0dPDcuzC37PSXjJ1G5cf5CYdyH0vAbqJLL\nSrQF4MC5VY7rhTXSdlKa5pbC8G+n4VlpgT8Uxmc3Ol+p8lv6eY11Zg8+ijf6N6bhH6ky3wSiQZs2\nN6d1nONZjsPfOnLsap2v1B9UO1GlpNF1XpqW8ZmScdl58zTppq8w/ow0/r7C8C+n4b/uYFqqBtXE\n9WA+EYC+psr8uxA5TavIPUCideDxFxq/DjcSVL8tzXMPVa5/ufP6lx1MT1by6Hs19uUqCg/EmnGM\n07j2gmoHfltlvdel8d8tDN8tN+9bS+abRKW0SFMDC2BnKoHzJoVxp+TSdWrJvGOo5HAeUxj3MSqB\n29Yl8x6RW3bTgmpgS+IhQjb+vxvYJ4OJB+1OoUQL8SDLgccKw79A6+t7tev/eR1MSzGoPiZ9/1Nh\nuqw02WvS99KgOne8q/13jqFSyuMHzTqmxL3CauKBU2lGC9GwnBP/SYNywxVUD/BOdaploJuY+os9\nXRUbkNU7usPd/1Ic6e4vEMEPRMBYZi0l9bSJp9ir0+fz3H1JyTTXpv6uNdL4P17+uo5z0vLHAG+o\nMX8raVl/TV8Pame9G+tdbrIs9cea2Yg652nGMVhN3FAU511F3OxD7X1cy9PAz0uWvZDI7YLKNmTe\nn/rfcfd1VZZ7Seof3kCabkj9Wda6vvLrUv9c4mn+q611vfts/PVV0vs/NdaZ7YNX0mtm2xLn0BIq\nreG24u6LiIC81bwF51Q5x7NXijV67DrjWeAXnZj/D6lf6zf2fS9vjTvb7p0Lxy/7fW1qVt46bwOO\nIG5Gb3f3v5dN4O4PEjfQw4Fq9Y3P7sR1uBHZOXueu6+sMk2jv7F6jt1v3f2JOpbV0WNcr29UGV7t\nN5PVJX3Y3a8ozpSuZ6W/4c5y94eI3MdxRMBVZilRwqM47zLiHevQdpuy6+7P3f3pknn/TASCnfVn\ni3YC5pvZciIg/GAadzPxsKtD3H0D0bgqtD3PbiRKS2xnrds/ya7flxBVgKpd/4vX9466kniQ/QYz\nmwpRZ54o0fBUSl8tx6b+U+5+SXFkOqbfSl+PK4zuzDF9F/GQ8Ep3f7hsAne/hnjQvhmVtyKIKKgW\naYK9U/+6GtNkf+g7WHmDS3PKbppSQLowfW3TKFCSvVN3fI31zy4bmP6Y7k5f9y6ON7OZFo0t3WfR\neNpGqzQudmqarLTBsuSWGuOquY34M94MuMXMPlzWGEtBM47BQ+6+osq881K/1j6u5foawUJ287Jr\n1lBNaiRovzT8f3M3Y606oloARF3qDkk38/OIIvkHpvWOI4ptP+zuzxN1GMcBe6Tx2xA5LOuIG0HS\n8K3ScIA/1Ujvd0rSe2DqjwKerTHvO0vmzftHleGdPXadcVt7QaKZTbZowO42i0bDNuR+YxenyWr9\nxtrb7sFEsd7M1USO8quBv1o0lDa1/U2pKTuGe1Q7fukY7pGmq3YMG7ledEaW7rNqpHl2mmZC8SFf\nahDqsxaNYi2waPAtO3bZw4VmXB87eozrsZGoflJrucXfTNZoW62AqPShSr3M7M0WDb09lRrk8tw+\n3TZNVm2f3uvuZe+Ph+rblP131AoiOxtgQjzAn5q6fENZ3yNKe1V9o4iZzTCz/7FoEG9p4RrxlTRZ\nq33i7ouJusOQAmUzG0Kc8wvSQ4rrKb/+Qye3OR2HS4lz8z1pcPYQ6+I6Hp5lx2V2jWmy//RJZjat\nZN5Gjml2TTiynWvZhDRdh/97pf+q2rqjyACRvTZrvJlZg7kkk1N/Xo1pshZdjSgiVwzenq8x74Z2\npsnGD6mxjFppy8ZNzg80s3cRrXNmy80a8MlyA0cRdbNLW2VOOvy6DHdfbGYnEMWMdyfl5KY/squB\nn7h78Q+xGceg1mvSstICtfZxLfXs/8HEDd8LxB921hLsxLKZClran6TU9cC7iZuuvxLF6AdRuZG5\nnihG+Dri4UuWi3FnIWdvs9znKXWsNx+cZPNuQtxwdmTevGrHLzt2PfF/V/P8t2hZ+q+0PsbLqNQD\nH0Y81Kj1G2tvuyF33rr7fWZ2GlFc8/Wpw8yeIEoD/K+7V3uAV012DFuo71wsO4ar3b3bXtdl0RJ5\ndt2YUGvanBFEA1pYtDw8m6gWkHk5jXfifJtIc66PHTrGdVqVcjprLbe4zKyl8Vr/Vw2/d9jMfkJU\nS8msJR6wZiV1JhL7tdo+7dA1PD28HJe+1kp3ret3vfZ19zvSejclrqv/DXwCuJWSkkxp2iOIh6fZ\n78qpXCMgzsnRlO+T64kHWa8jSszMStP/KTf+JNpe/x9PD1U766dEUewTLN768Z7c8PZ05D89m35u\nE45pdi0bReuHH9XUW5pOBgDlVMtA98/UHwbs2MllDe/k/L2GmU0GfkTcgFxK/BkPd/fx7r6pu29K\npehV1SKkNW7aanL3PwEzgA8TjV49B2xKvIpjtpmdX2XW/nIM8tfmvdzd2usaXE9WBPx1hf71hX61\n8WXpHV9HeqeXzHtvPdvp7ic2uK09ob3z/2IiULiVqKc30t3HuvvU9Bs7MU3XrGLaALj7uURjaJ8h\nimkuJnIBTwHuTUF3R2TH8JI6j+HZJcto6FrRCYOo7NfD60z3wtz8PyAC6oeJVrTHufvo3LF7fZqu\n1rHr7m3utczsOCLAW0c0kDmD+M+ZmPvPyR72NPX30N3cfb67/4ioijQIuMDM2tx/pJIRPyUC6j8S\nuagt7j4ut0++mE1esqpmXd8b4u63Eu+t3pNooGxz4OY6qzxkuvs/PbuWfa3Oa8Jvujl90ospqJaB\n7nriyS/AWxtcRpbbMK3GNFnRWKdSnLs71SqCmI3L55q8iXhK+xDR0NKd3rZeb2eLjNbk7kvd/Ufu\n/k5334Kou/SjNPpkM3tzbvLefgzq2f8biOAGogRFdsNda5s6K7t52s/MWqjcVM1O/buJ3J/XpPq3\n1W66Xsh97mh6s3kHVDE6M9uZSuNdR7r7td62Xm+X/cbc/Tl3P8fd30IE9gcSOViDgP+y9C7tOmXH\nsCvP1aZK17NF6WuH0m1mo4E3pq/HufsV7r60MFmXXh97SHbd3KzGNLXG1ZLVi/2uu3/D3eeUlBxr\n6j519/VEWw5Q3zW6qdz9auKVli1EyZGig4kc2OeIBtZu8bbtRtTaJ1lQPTNV8Wh1fXf3OUTL4cXr\nf1OC6iSrwvKNwvf2dOQ//ZXpm3BM+9y1THoPBdUyoLn7s1SKQv2rmY2pZ74UYGSyBi9eVxied2jq\nP+rV6+12pdeVDUw3h1n9o3zDHdmf1X1e0tBY2s5Di8O7krs/5O4fJnL1oPU29fZjULr/C+MeSPXQ\nshv+O9LwNzWwvuyY1czR8WiI5UWiqPkbiDqTj7r7/DR+A9FY3kTiVUzTiWD/psJynqJyM9LR9Gb1\nSieY2f4dnLcvy35jz7j7S1WmOaw7EuLhFuJ1Oi8RJVQOrD1XK9kx3M/M6qmu0Ftk6e7oObsZUV1j\nfY2i8t1y7LpZ1v5GtYbmIIoQNyL7PdxdNjI9hGo0YK8l++94bY1pal2/O+urqX+kmR1QGJftk4e8\nel3xqueZuy+gUhrvUKIxs4XEmzsyN9D6+g/ND6qduKasIUq+1SM7LgeluuBlsv/0he4+t2TeRo5p\ndk04LBUlF6mbgmqReMXEGuIP7OdmVrO4kZm9A/h0blBW/GcXohhgcfqpxPtgIYoy94TPWGoEq+A0\nonjVMqK+cibLddm1SpB6MpVGY5qqSjrzVqX+sNyw3n4MppvZ8cWBZjaBKOIO8T7svAtT/0Qz24Ma\nzKzY+E7WwvO44rQlstyM/yQChdmF8dkN1hmpf7dHA3dFF6b+6Wa2RY20WmoQB3glsM8elPxXjRso\nzKzFzIZVG9/HZL+xaWY2tjjSzPYjgtymauf3tY5KCYmO7Oc/ETlFw4Cvt7P+nmgwrpoLU/9oM6t1\nA15Md3bsNjGzmSXTTgc+0oT09Ta/S/2ZhZJCAKQHKh8sDq9Ttk93K1muAWc1uNz2ZNfdd1s0uFhc\n9+HAPl20btz9Lipv8PhCYXS2T3YqC/DM7Bhg33ZWkV2/P0XUvb6hUAIgG59d/+cUAtROSbnhnyPe\nCvEpjwbU6pH9p29BpYGzV6QMkE+lr8X/zs4c018Q94ObpXRX1cuuZdILKKiWAc/d7yEaC3HgzcDd\nZvbeFPAAYGZjzewYM7uOeNI6Ojf/34E/p68/MbO3p0ZwMLN9iGA1a4Aqa/24u00Dfpdu9jCzEWb2\nGeJ9xgDfLBQ9vYbYH7sC382CIDMbY2afBb5PpZG3ZvuYmf3FzN5tZq/kTJjZODP7PFEkDiqvueoL\nx2Ap8CMze092c2RmuxPbMJnILf5BYZ4fE8HmcOBvZnZyviSFmW2alnc9lZbYM1lOxKvNbPt20pYF\n1dnNWTGX4vp2xme+QbyiZRJws5m9IxUpzNI7zcw+TOQivK0w7yeJG5nXAtea2Suv8TKzwWa2m5n9\nv7T8rsit6gl3E7lGQ4Ff5n6bQ1MjgVdRualupu+Y2S/M7C35hxvpQcj5RENz66nc6LcrXTuyetgn\nm9kvzeyVVxelbTrAzL5FVCnpUmY2qZ0uOy8vI64bg4lW608tXPcnpWvJn8m98sjjFX1Zy8oXW7wq\nKDtX30S8haDaa/D6LHe/j8rrti42s2Nz19k9iOvZ4AYXn72i8bTCdXIG8Z97JF3ze7iIuK6MJF57\ntX9a72AzOxr4ZRetN++/U//NFo0XZq4nGmvbAvixmU1JaRtpZh8jXovV3v9ws67vDXP3/3b30929\nzevOaszzT6KxUohr1vuzB67p9/ZnIiNkGW1fDdfwMXX3ecCX0tevWrz9ZJtsfLp3OsTMLiDuk0Qq\nvBe8LFudut7QETf6LxDBZNYtJy7a+WFzgNcW5p1M3CRn06wqzLcIeFXJOk9M42fXSNecNM3BVcaX\nLoMoqput/1jiJs+JervrcuMuBzYpWe45he1eTORiOfGH9tX0+cJq621nf5dOR9yc59f7clp3ftj/\nliyvK4/BmWXbWsc5dWGa7+tEgOxEq61Lc+laUTyfcvNPIV5fk027gbiJermwP84ozDcEeDyN20gE\n7XNSt2Vh2t0Ly9q8ZFkrcuPfUmN7tyOCpmza9UTguLKwjveXzPsmoi5cNs3qNO/awrxbF+bLhk/v\nyHnWwPVhem5dpb/F3LRnp+nObWe696Tjky13aW57HyMabnKiakBx3oVp3Kwqyx6VW+6k3PALCvtz\nKXGdy75vBD5Zsrw70vi319ieT9D62rIina/rc8OWF+bZNQ1/uZPH55TCdtXqTs/NN4Z4gJHf/kWF\nfeLA9wrrO5h4EPTKdhHXHCf+R46ttl317MvOHOM07jfFba13fxMBbLXzbjKtf+OrqFzPFgPvTZ9f\n6uDxG0Hr6/c6Ktf9jUTpsNL9ljv2VzbymySKPef/Y5bljuUDRMNpNZdfZZ275pZZegxz096Tprus\nMPzzhfNwMZXf0y3AZ2uljag7nJ9/95JpnsuNP6nB31+2f6/p4HzZdp9eMm4M8VAgS9saWv9HrADe\nUGW5DR9TotrUWYX9tiwtL3+9vr+jvy11/btTTrVI4u6XEy3ifoIozvgs8fqOTYhg5DfE64d2dPcb\nCvMuAF4FnE788a8jcqAeA74N7OJRZ7FHuPtlwCFEC6IbiD/le4F/JRpAWV8yz6eJosl3E39mg9Pn\n04gc/TbzNMnPieLllxL1wdYRN4/PA1cAb3X3NkUre/kxWEPchH8ZeDqlawHxxHzv4vmUcfcXibpf\n76FSxDYrJfEw0TLsOyg8qfeok/16oj7bPCKXfuvUFYsR3k+lwabH3b3Va0jSsrJ3Um+kxjtq3f1x\n4mbm40Ru3WJgLHGu3EfkhL6ZSg5Eft6rgB2IhzV3EftsHHEzc3Paxn3c/elq6+9r3P0S4AhiX71M\nPMB4krStNPBKujqcRRSbvJL4bQwizseniZyvV7n7dxtZsLt/H9gZOJf47Tpxvi4g3in7BSrvqu4V\nPKoy/AtwDPB7YD5xvRlE7J9fAidQKArq7rOJusVXEkHlEOIadS7R2vEj3bIB3SxdZ/cHvkmcq4OI\n4OZiIrdzTpp0Sdn8NZa7krjWnUOci1lg/Sfgje5+ThOSX23ddxPH7CLi+A8lrpv/Rfyn1HpVV7Nk\nudVHZyUfUtrOAo4HbiOCwk2Ikkj/QeyvVdSQruePp6+LqJSwyMvnTpf+F/WE9Ns8lPg/uZl40Doc\neAr4IfGffnWVeRs+ph4+T1yDfwI8Qez3EWkZVxEPeQ7v9EZKv2IeT1dEpJ9JxUmfAvDGX7kkIiJS\nFzM7lXiI+Ud3P7Kn0yMi0l2UUy0iIiIinWLxXuVPpK9/rTWtiEh/o6BaRERERNplZjPN7Hwze1UK\nojGzQWZ2INFw0/ZEcf+f9mQ6RUS6m97BJiIiIiL1GE60eXEygJktJuqaZq9gexk43ut/dZKISL+g\nnGoRERERqceTRKNt1wJziSB7A9Ew27lE69J1v45NRKS/UENlIiIiIiIiIg1STrWIiIiIiIhIgxRU\ni4iIiIiIiDRIQbWIiIiIiIhIgxRUi4j0UmZ2oZm5mZ3Z02kZKMzs4LTP55SM65LjYWZz0nIPbuZy\nO5iG2T2dBhERkb5KQbWISE4KLLLuijqm/11+nu5Io9SncCxrdeN6Oq2NMrO3mdmZCoYHpt5y/NPD\nqDPN7G09mY6OMLNpZnaamf3BzOaa2RozW25m95rZN8xsszqWMdTM/s3M7jGzl81siZndYmYfNjPr\nju0Qkd5B76kWEanuCDOb7O4Lykaa2UTgzV24/ueJV9Us7MJ1DATLgFU1xm+sczm98Xi8DXh/+jy7\nB9MhPaO3HP+DgTOAi4DLezAddTGzrYA5QD7wXQaMBHZP3YfN7Fh3v67KMsYAfwP2SYNWAi3AAal7\ni5kd7e7ru2QjRKRXUU61iEi5ucAQ4Pga0xyfpnm6KxLg7v/h7jPd/dyuWP4Acqq7b1qjW1bPQnQ8\nRPqNwan/R+A4YIK7jwVGAP8CPAWMBy43s02rLONHREC9CHgLMCrNfyKwGjgS+FIXpV9EehkF1SIi\n5X6W+ifUmOZ9hWmlBjPb38x27Ol0iMiAtxjYy92PdPffuPtiAHdf6+5XEYH1amAM8JHizGa2F/CO\n9PUkd7/SwwZ3vwj49zTuU2Y2pcu3RkR6nIJqEZFyNxK5FbPMbKfiSDObCexL5FJf397CzGxbM/tf\nM3vSzFZPf4BmAAAgAElEQVSb2WIzu8HMPmRmg6vM06ZhLDN7Txo2v9p8abpXpenWmdmkkvG7mtlP\nzOyplJ4lZnaTmX3UzIa0tz0NehPwcFrPh1LxyT6jvYbKzGy8mX0rNTy2xsyeMbMLzGyrWg2gFZYx\nwczOScdljZnNM7MfFet3ZsujUvT3jGJd8WZsc2eY2QkpLf8oGTfJzDam8VeVjN8xjVttZsNzw0eb\n2Ylm9iszeyCdt6vM7HEzO9/Mtq+RnmzfTDeznczsonSM1pnZ5Wma6fn9Z2b7mdnvzWxBqm97s5n9\nS26ZQ83scyktK83shfQ7n9DgPpthZueZ2aNpu1aa2dMWDcn9R/Zb7ujxN7O9LeoJ32iV+sMvpeXW\nugadmZZ3oZkNMrNTzOz2tN/dzPZM6zojzfL+YjrMbHoj+6IruftSd7+3xviHgVvT131KJnl36j/i\n7mVtb5wPLCWKgx/TmbSKSN+goFpEpJxTyYF+X8n4bNgladqqzOxI4AHgw8AMIgdkJPAaogjhn81s\nZJ3pupyouzcVeH2N6bJi61e7e6s6wGZ2CnAvcBIwHVhHFF08EDgPuNrMRtSZno54Nq3rQGK755vZ\nxWZ2qFnfbtTHzLYE7gJOA7YGNgDjgA8CdwLb1bGYbBmfAqYQ59XmwIeAm81sfG7atcALxLkEsCJ9\nz3c9LXvYtJeZjS6Mey2V+qwHlgR1r0v929x9dW74+4H/I4rsziT28yBgW+Bk4G4zO6yddL0GuIP4\nDY8FSuu8mtlRxMO1txDVPEYBrwL+YGbHpWD/L8A30vohjtuHgWvMbGg76Siub2/id/lRYHuiiPIa\nYBqxP84CZqXJO3r8rwY+BxwETCSuIRPScn8EXGFmtdrZMeC3wPeAvalc80anda1I31eXpGNDfXug\n13kp9cseOByS+leXzejuq4C/p6+HNjldItILKagWEanup6n/HjN75XqZAsD3FKYpZWbbAr8EhhNB\nxkx3H0fcjH6EuGk+DPhOPQly9xVAljNSWt87BShZ0cSfF8a9jbgxXgH8GzDZ3UcTdQGPAB4jGh36\nVj3p6Qh3vwDYggg87yZycd4LXAs8lXLFZjR7vd3kZ8QDiheIupSj0n49iKhz+d91LON7RLHUA919\nJBHEHQUsScv+j2xCd7/Z3TcFLk2Dzi7WFW/KVnWCu88lSnIMJvZDXhY0LyeK2O5VZXyxFMhC4GvA\nfsAId59I/LZ2Ih5wjQR+3s5Dqh8A/wB2c/cxxLn/mZLpLiJ+35ul3+wU4PfEvdO3gLOJwP5I4liN\nJo7X8rQ9H6qRhjJnp2XcBuzt7kPdfXzapn2BbxO5n40c/6uJ68Vm7j4yLXcUUb1lPlHc+VM10nYM\ncX34ODAmzT8VuDet6+w03aUlbRY808H90OPSA4bsnH2gMM6I4w7wYI3FPJT6Ozc3dSLSGymoFhGp\nwt0fB24BtiICzcwhRO7R7e7+SDuL+TxxU/wE8C/Z9O6+xt3PBz6ZpvuAmdWTmwmVQPnofNHYQvqm\nErlRr7TEm4Ltb6evx7n7f2e52Kku4V+IItorU3rafaVMR7n7Anf/jrvvTbSwew4RiG5NFCF9wsyu\nM7P3dSD3vj3fsSguX9Yd0dmFm9khRBDowLHu/kd33wgR/BDByLA6FrUGOMzdb0nzrk9FS7+axr+9\ns2ntAVlQ/LrC8Oz7ue2MbxVUu/sv3f0L7v4Pd1+bhnkqrnsCcA0wmdr76kXgTe7+QG7+J0qmu8vd\nP+TuL6TpFhAP05YRD4c+AbwzHe8NqbuCygOUjh6vA1L/VHe/O7fNK939Dnf/VHZudJS7vzvtu/m5\nYSvc/WdUHsB9vMYiRgGfdPfz3H1lmv/Fehv5q0e+6H0D3ZxmpSP5BLAp8WaAiwrjxhDXdIDnaiwj\nG9f066iI9D4KqkVEastyovNFwN9XGFcq5Wgcm75+K7sZLbgAmEcUr6z3JvzPRO7nWCKHqSjLwb4i\n5WxnDiaC1wdSAN1GCi5uJV65eHCd6WmIu9/v7p8hij0fCfyGKNZ6MHEj+7xFneRXd3JVY4iHDGVd\n2UOJjsrqTN7k7jcVR7r7HKK0QnvOd/eXSoZnD0ZmNPFBQ3e5IfVfCZpTMfbdgH8Cvy4Zvy0RtK4j\nHmrVxd2daM0Z2uaM552biue25xsl61hBpa7tze5+Q3EaouQFwK51rCMvC1C7NQhz97+TSkOY2eZV\nJnsJ+EkXJ2UDbYuO19uVvvawEWa2O/D19PVcd3+oMEn+N1jrPMqu96OalTYR6b0UVIuI1HYpkYN4\nrJmNSHWNjyWCv/YCpW2IwBeg9F2nKUdzdvq6dz0Jcvd1RAAKlQZzADCzYVSCvFZFv4m6zADb18i5\nnZ+bbqt60tNZKUf2j+5+HBFQfBy4nSgK+0Hg72b2H7WW0Y6T3N2qdM14p25WdPnGGtP8vca4TJsG\nvZJ5uc/j6kpR75HlNM/K1dN/DXH/cT1Rh3gp8JpcFYsswP5H2YMoM9vSzL5pZnemBrM25Brnyqot\nVAsOof5A/f4qw19M/QeqjM/qM4+vMr6aP6X+T1OjYgdYExsNTPXAL08Nla0qNGqWnVfV9tsd3sXv\nW3b3Z0qKjtfb7duMNKTSOZcTVVPuJOqhi4i0q1ajFCIiA567LzazK4lA+mgiR3kUcHmVXMW8ybnP\n86pOFQ14Fadvz8+JBpHebGaj3X15Gv4m4gZ5EZGjnZflgA0jcmnbU1djZWZ2OnB6yahnOnqzm/b3\nZUQaNydysaG+4tM9JWtd/fka09QqJppZXjbQ3Vfn2nHrqpbZu4S7P25mzxHH8kCieHYWNM92941m\n9neipMLuwD258W1ygc3sdcCVtM79W0qlwa4WWhfPLVNXrqa7VzueWcNb7Y3v6D3WZ4Edif30udSt\nNrNbiBz9C+vMYW8l1Q/+FXH9yqwh6qdnaZ1MPOiott+alhPcW1m02H410ZjkY8CbvXUjeZl86Z+W\nGovMrp8vNyeFItKbKadaRKR9+SLgdRX9LtGMYsZ5NxDB+HBav7IlK/r9m5SjnZdd839fI+c2351Z\nZ1pGUV60uu6HBGY2LOWk/YF4APEtIqB+kah33dH9Lb1HsQh4sb50sd51aX3qlGv7M+J8u4ZoQbzF\n3cd5pXGuT2eT10hPr2yNOj2kezVwOPBdojG/oUQbCT8AHkitzHfUyURAvZJow2Erdx/u7pNz+y17\n6FNtv/XKfdYsZjaWaMl9V2Au0bZBtRb0l1EJrGuViMjG1XrYJiL9hHKqRUTadxWRq5O9qmcRlbqb\nteRzd6ZR/TVH2Y1y3blB7u5m9ksih/h44CIzG0W8/gfaFv0mt/5p9a6nzrScCZzZyLxmdgDxmqR3\nUikuu45oZfn/gD92dbHTJlhI5DDWqgs7kBsruh54F/A6i3eT70m833d+bjxp/OVUXklWrJ/+KuK3\nsgg4qkobBfWUwOi1Ur3wa1KX1T8/jqjjuw3xsOm4Di42m/4r7v694sjUgGGbd9l3NzPbiupVINrT\n4VIxufWOJIrezyJaQj/Mo+X6Uuna+880/S41Fp21+l2sky0i/ZCCahGRdrj7uhTAnpIGXZq1PNyO\nJ4kGgMYRuU1tbhhTPdKD09e7Opi0nxNB9evNbAqRw9VC5GCXNaCU1SXd3cy2cPdaRdK7jJlNI1pq\nfh+wQ27UA0Qg/TN3f7Fs3l7qbqJhrFoNqr2mi9a9MfV783u+s3NxP+IcHUzrXOi7iCKyryX3W8hV\nachkD58erRJQQ+XBV7/g7ouB81Ojhz+kbSvp9Rz/bL/dXWX8QXS+JE0zzsPBNP5QpKyYdrvMrAX4\nA1Hk/iUioH6sjlmvI4Lqw6ssdziV3/y1ZdOISP+i4t8iIvX5PvA/qTu3nWmBV3Kdfpu+npprqCnv\nQ0RLx06lJeS6pNfuPEw8ID2OSqNlv0zrLroWeIa4ea353uSUQ9ZUZnaQmf0NmEO8JmoH4r3MPwD2\ndffd3P2cPhZQA/wu9Q8ys1cVR6aHCO/qonVnLUb32gbMUuvJC4h68VnDT7Nz47Nc6YnEq4yg7fup\nIb2jmWhor00QaGZvIB5e9TlmNijVfa4mq0tdbFugnuOf7bfdSta7CZVXtnVGp89Dd59TZ7WUsm56\nR9dnZkOJ6/MhxMPPN7h7rfdO5/0i9Wea2ZEl408mGqlcReX6ICL9mIJqEZE6uPvD7n566jpSnO8s\nov7d5sAfzWxHeKUO8clE3UmAH3v5u3LbkxXz/giVXJOyot9Zq+GnEAH88akl4D2z8WY2xMxmmdl/\nAU81kJb2HE7cwDrRINDxwObu/gl3v6ML1tddriNa9zbgMjN7U8pZzIq3/5loLb4rZEHAEdYF7xVv\noqz186yIbjFovr6d8RCB90oi+P5ptr1m1mJmHwAuI3Ib+6IxwONm9p9mtlsqkp0F268HvpamK74K\nr57j/9fU/6KZHZVb9kwil3Y/Wje+1YgsHa82s+07uawul/bBz4l3yC8n3lted0mh9EDzV+nrhWb2\nL9lyzex9wDfTuG/1wYeEItIABdUiIl0oBcrHE8UTDwYeNrPFxI3c+UTO07XAaQ2uIgugdyNahn44\n3fBVS88VxGuq1gJHAXeb2Uoze4nIVfkH0Qrx2GrL6IS5wBeArd39je7+yyqt6/YpqVTAe4nt24yo\nn7nCzJYTRe4nUGkdfU2TV/87oo7xDsCzZva8mc0xszlNXk9n5YPkx9y92Bp6fvxGSl5P5u5LgOzV\nascBz5nZEiKX9MfA48CXmpbi7rc1kWt8H7Aq/SbXEvWrtySqk3y6ME89x/9s4AkicL88LXsp8Z7w\nw4GPEu0CdMbstI4JwCNm9mKWjgYbV+tqBxFvdIC4bl5e4zWD1ep5n0y8dmsi8cB0BfFw4iKiGs6V\nwBlduxki0lsoqBYR6WLu/gci6P0RUfR5BJHjdiPxWqw3untDOUUpaL89N6g0l7owz/8RDWt9m8hh\n2kDccL9E3ByfkcY3lbv/xN2/5u7Ptj9135IaNtqbKHkwlyhiv4Q45vtQyUFd0uT1LiRy/39LFLGe\nTARnWzdzPU1wfZXPmX8QvwmA+1IA3Ya7f5do7T7Ltd6EqAJxBlEvtvS1ZH3AMuK1Yt8mfs8LiPe0\nryD2zX8CexZ/O/Ucf3dfBBwAnEfl9X2riAD7de5+YWcTn0rBvB64mGi9f3wuHb2x/Z78/e9wyt9e\nUPMtBu6+jDjn/p1437oTD81uJUoOvbUPNLIoIk1i5dXuREREpFnM7CtELv1F7n5iDyenDTObTTSC\ndYi7z+7Z1IiIiPQtyqkWERHpQmY2gShyD5X6rSIiItJPKKgWERHpJDPb38y+lxp6G56GbWJmhxIN\nmW1GFP2/rAeTKSIiIl2gN9ZzERER6WtGEy2rnwKQGqMbCQxN4xcB7+wPDbOJiIhIa8qpFhER6bx7\niDrT1xPvAh8BrCMagvsfYFd3v7367CIiItJXqaEyERGRAU4NlYmIiDROQbWIiIiIiIhIg1T8W0RE\nRERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFRERERERE\nGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpF\nREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERE\nRKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSg\nWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFRERE\nREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRB\nCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRE\nRERERKRBCqpFREREREREGqSgWkRERERERKRBCqpFREREREREGqSgWkRERERERKRBm/R0AvqqSZMm\n+fTp03s6GSIiIiIiItIF7rzzzoXuPrm96RRUN2j69OnccccdPZ0MERERERER6QJm9nQ906n4t4iI\niIiIiEiDFFSLiIiIiIiINEhBtYiIiIiIiEiDFFSLiIiIiIiINEhBtYiIiIiIiEiDFFSLiIiIiIiI\nNEhBtYiIiIiIiEiDFFSLiIiIiIiINEhBtYiIiIiIiEiDFFSLiIiIiIiINEhBtYiIiIiIiEiDFFSL\niIiIiIiINEhBtYiIiIiIiEiDFFSLiIiIiIiINEhBtYiIiIiIiEiDNunpBIiIiIiIiEj/sW4drFwJ\nq1aVd6tXR3/oUHjb23o6tZ2noFpERERERKQf27ChbWBbDHqb+X3DhvrSNW2agmoRERERERFp0Nq1\nEYyuWFG7nwWsjQa6a9f29JaWW726p1PQHAqqRURERERECtwjIG0v4K0VCLc3zfr1Pb2VPWvVqp5O\nQXMoqBYRERERkT7JPXI7X365dbdiRdth+eH1BsXSmMGDYcQIaGkp74YPj/7o0T2d0uZQUC0iIiIi\nIl1u7dryYLfeYWXDV6yAjRt7esv6hnxQWwx4899rjat32iFDenpru5eCahERERERaWPNGli2DJYv\nr3TtfV++vHpQvG5dT29R7zN4MIwcGcHoiBGVz8V+Fqw2GgwPHw5mPb21/ZeCahERERGRfsA9gtl6\nA+D2visIjmC0vYC3M/0hQxTs9gcKqkVEREREelDWINbSpZVu2bLyz/nvy5ZVAuFlyyI32L2nt6b7\nDRsGo0a17kaOrP595Mj6At6WlshJFmmPgmoRERERkQZt2BBBba3gt55xA6EV6MGDo2GqYsBbFgTX\nO2zkSNhEEY30MJ2CIiIiIjJguUdQvGRJdIsXt+6XDcv6S5fGvP3VkCERBI8ZE/18VxyWfR81qtIv\nBsFDh6qos/RPCqpFREREpE9bs6b+QLhsWH9qPbqlpXrA29Hvw4b19NaI9A0KqkVERESkx7lHneBF\ni+Cll6Kfdfnv2efFiytB8apVPZ36zhs2DMaOjW7MmMrnWt/HjKl0We6wikKLdD/97ERERESkabLg\nuFZgXC1Q7qv1irPc3XoC4bLvY8YoV1ikL1NQLSIiIiKlsvrGCxdWugULyr/35eB4xAgYNw7Gjy/v\nVxs3dmwE1GohWmRgU1AtIiIiMkCsWROBb7XAuCxoXru2p1PdvsGDGwuKs3FDh/b0FohIX6agWkRE\nRKSPWrMGXnwRXnihdX/BgvJAube3VN3SAhMmwMSJ0c+6su/jx1e6kSPVqrSI9BwF1SIiIiK9hHu8\npqksUM76+c9Ll/Z0isu1tNQOjKsFyS0tPZ1yEZGOU1AtIiIi0oXco5Xq55+H+fOjXytQ7m3FrVta\nYPJkmDSp0hW/Z92ECQqORWTgUVAtIiIi0oD16yMIfv751gFzvps/P7o1a3o6tWHw4MgVbi9Izn8f\nMaKnUy0i0rspqBYRERHJWbmybYBc9v3FFyMXuicNHhwB8NSpMGVKpT9lSuvAOPs8diwMGtSzaRYR\n6W8UVIuIiMiA4B6Ndc2bB88+G/2sy39fsqRn0zlyZOsAuVZ//HgFySIiPU1BtYiIiPR5a9dGDnI+\nOC4GzvPm9Vx95VGjYLPNott00+jKAuUpUyKoFhGRvkNBtYiIiPRqGzdGcetnnoG5c9t2zz4bRbF7\nwqRJESBnAXO+yw8fNapn0iciIl1PQbWIiIj0qOXLIziuFTSvW9d96dlkk8g5bi9QnjoVhg7tvnSJ\niEjvpKBaREREuow7LFgATz0V3dNPtw2au7MO85gxsMUWlW7LLVt/32KLKIKtesoiIlIvBdUiIiLS\nKUuWwJw5lcA567JhK1d2fRrMIue4GCQXv48e3fVpERGRgUVBtYiIiNS0cmUlQC4Lnrsjp3nCBJg2\nrbzbcssolj1kSNenQ0REpEhBtYiIiLBoETz+eNvuySfhhRe6dt1Dh8JWW5UHzFttFZ0a+hIRkd5K\nQbWIiMgAkNVtLgucH38cFi/uunWPGgUzZkQ3fXp0+SBadZhFRKQvU1AtIiLSjyxaBI88Et2jj7YO\nnJcv75p1Dh0agXIWOOcD6BkzYOLEqPMsIiLSHymoFhER6WPWrYti2VnwnHUPPwwLFzZ/fYMHR85y\nWeA8Y0bUZ1ZOs4iIDFQKqkVERHqhrLh2MXB+5JEIqNevb+76hg+HbbeF7bZr3W2zTQTUagRMRESk\nnIJqERGRHuQeDYE99BA8+GCl/+CDUZS7mUaObBs0Z93mmyu3WUREpBEKqkVERLpBFjwXA+eHHmpu\n8Dx0KGy/Pey4Y3Tbbx/ddtvFe5xVt1lERKS5FFSLiIg02Usvwb33wgMPtM6BbmbwvPnmlcA53229\nddSBFhERke6hoFpERKRB69dHC9v33gv33Rf9e++F555rzvJbWmCHHWDmzNaB8w47wOjRzVmHiIiI\ndI6CahERkTosWtQ2eH7wQVizpvPLbmmBnXeObpddott552htW/WcRUREejcF1SIiIjkbN8ITT8Bd\nd1WC5/vug2ef7fyyR4yAnXaqBM0KnkVERPo+BdUiIjJgbdwIjz0Gd94Z3V13RbdsWeeWO2RIBMu7\n717Jed5ll6jvrOBZRESkf1FQLSIiA8KGDfGO57vuqgTRd98NL7/cueVOnQp77BHd7rtHf+ZMvddZ\nRERkoFBQLSIi/Y47PPMM3H473HZbdHfdBStWNL7MIUOi6HY+eN599wiqRUREZOBSUC0iIn3e0qVw\nxx2VAPr222H+/MaXN24c7L037LVXJRd65sx4B7SIiIhInoJqERHpUzZsgPvvh1tuqeREP/xw5E43\nYvx42GefSrf33rDNNmDW3HSLiIhI/6SgWkREerXly+HWW+Gmm6K79dbG60FPmlQJnLMgeuutFUCL\niIhI4xRUi4hIrzJ3biWAvummeJ3Vxo0dX05LSwTN++0H++8ffQXQIiIi0mwKqkVEpMe4w0MPwezZ\ncOONEUQ/80zHl2MWjYjtv38lgN51V7XALSIiIl1PQbWIiHSbjRsrQfTs2XD99bBwYceXM2kSHHgg\nHHBABNGzZsGYMc1OrYiIiEj7FFSLiEiXaVYQPXMmHHRQpdt+exXjFhERkd5BQbWIiDSNe7TEfc01\nlSD6pZc6toxhw2DffSsB9IEHwsSJXZJcERERkU5TUC0iIp2yYAFcey1cfTX89a/w7LMdm3/MGHjt\na6N79aujZe5hw7omrSIiIiLNpqBaREQ6ZPXqaFDsr3+NQPruuzs2fxZEH3xwdHvuCYMHd0VKRURE\npEe5w7Jl8Nxz8PzzMH9+6/6mm8LZZ/d0KjtNQbWIiNTkDg88UAmib7gBVq2qf34F0SIiIv2MOyxf\nHsFyFjBnn4vfa900zJypoFpERPqnlSvhb3+DK6+EP/6xY0W6W1oiiD7sMDjkEAXRIiIifUq9wfLK\nlZ1f1/z5nV9GL9Cng2oz2wr4FnA4YMA1wGnuPreOeacBXwEOASYDzwC/Ar7u7iu6LNEiIr3U3LkR\nQF95ZQTUq1fXP+9ee8Eb3gCHHx6Niw0f3nXpFBERkQasXg3z5sWT8mqB8nPPwYpuDIWWLImc7JaW\n7ltnF+izQbWZjQD+BqwB3g848FXgOjPbvVZgbGYjiQB8CPBFYC6wL/AlYHvgnV2behGRnrdhA9x2\nWwTRV14J999f/7xbbFEJog87DCZP7rp0ioiISDuWLYtg+dlnK4Fz8XNHX8fRLC0tsPnmUX96s82i\nyz5vumm/KM7WZ4Nq4GRgG2BHd38cwMzuAx4DPgKcU2Peg4jg+Y3ufnUadp2ZTQBON7MR7t6E8gwi\nIr3Lyy/DVVfBFVdEv97/15aWKMr9xjdGID1zpt4TLSIi0uU2boSFCyvBcbWA+eWXuz9tw4dHYLz5\n5pUu/z37PHZsv79p6MtB9VuBW7OAGsDdnzKzm4CjqB1UD039ZYXhS4BBRFFyEZF+4aWX4A9/gN/+\nNhoaW7OmvvmmTYMjj4zu4IP7fMksERGR3mXjxqhTPHdu9VzmefNg7druTdewYeXBcfH7uHH9Pliu\nV18OqncBfl8y/EHguHbmvYbI0f6mmX2MKP69H3Aq8EPVqRaRvm7ePLj8cvjd72D27Cjq3Z5Bg+DA\nA+HNb45Aepdd9F8pIiLSsGXLImCeOxeeeabyOeuefRbWr+++9AwaFAHxlltGPa5qOczjx+sGoIP6\nclA9AVhcMnwRML7WjO6+2sxeDVxGBOGZC4BTmpZCEZFuNG8e/PrX8KtfwS231DfPuHFwxBERRB9x\nBEyc2LVpFBER6RfWrYtGvYqBcj54Xrq0+9IzbFgEyltuWQmai5+nToVN+nL413sNyL1qZsOBS4Ep\nwAlUcqr/H7Ae+FiV+T4MfBhg2rRp3ZJWEZFaXngBfvMbuPRSuPHGeG1ke6ZPh2OOgaOOipxp/b+K\niIjkuMOiRbVzmZ9/Popvd4cxY9oPmCdOVO5yD+rLt1KLKc+RrpaDnfdB4GBgO3d/Ig27wcyWAueb\n2Q/d/d7iTO5+PnA+wKxZs+q4dRURab6FC6N+9KWXRtHuev7Td9klAuljjoE99tD/roiIDGDusGAB\nzJlT3j39dHPewVyPCRNgq62iKwbKW2wR3Zgx3ZMWaVhfDqofJOpVF+0MPNTOvLsBi3MBdeb21N8J\naBNUi4j0lFWrorGxiy+GP/+5vipY++0XQfTRR8MOO3R9GkVERHoF93gCPWcOPPVUeeC8alXXp2Po\n0AiWp01r3WXDttoKRo3q+nRIl+vLQfUVwNlmto27PwlgZtOJ12X9ezvzzgfGm9l2+dbDgf1Tf16T\n0yoi0mEbN8INN0Qg/ZvfRHsn7Zk1C975TjjuONh6665Po4iISLfLB83Vuu7IaZ46tXWQXOwmT47G\nwRwxBsEAACAASURBVKTfM6+nAl4vZGYjidzkVcAXAAe+AowGdnf3l9N0WwNPAF929y+nYdOB+4jg\n+mtEnepZwBeBR4H93L1mgcpZs2b5HXfc0fTtEhH55z8jkL7kkqi21Z499ohA+h3vgG237fr0iYiI\ndLlVqyI4fvJJeOKJ1v3uCJpHjCgPlLMAesst4z3N0q+Z2Z3uPqu96fpsTrW7rzCzQ4FvARcT75a+\nFjgtC6gTAwYT75/O5p1jZgcAZwJfBSYBzxD1pb/WXkAtItJsy5ZFHekf/xhuu6396WfOhOOPj2B6\nxx27Pn0iIiJNldVrzgfL+c/zurjg6KhR0XLn9OkwY0bl8/TpUdRrwgQ1QCJ167NBNYC7zwWObWea\nOURgXRz+EPCOrkmZiEj73OGmmyKQ/tWv2n/oPnVqBNInnAB77aX/ehER6eXWro1Gv6oFzi+/3P4y\nGjVyZNtgOd8paJYm6tNBtYhIXzR/Pvz0p/CTn8Ajj9SetqUlGho74QQ47DC9/kpERHqZlSvh8cfh\nsceiyxfVfuaZrnvt1MiR1XOaFTRLN9PtmYhIN3CH666DH/wALr8cNmyoPf2hh8L73x8B9ejR3ZNG\nERGRUqtXR5CcBc75rquKaZtF/eVttokGQ7L+jBnxWe9lll5EQbWISBdaujRypX/wA3j44drTbrkl\nnHRSdDNmdE/6REREgCiq/eST5YHzM8/E0+FmGzmyddCc/7z11jBsWPPXKdIFFFSLiHSBe++F886D\nn/0MVqyoPt2QIXDUUfDBD8Lhh8Pgwd2XRhERGWDWrYv3NheD5scfj7rPXVFUe/PNy4PmbbaBKVOU\n2yz9goJqEZEmWb8eLrsMvve9aICslp12gg99KOpKT57cPekTEZEB4qWXonjUI4+07j/5ZPxZNdOg\nQZGrvP32lS4LnGfMiMZBRPo5BdUiIp20dClccAF897u13yu9ySZwzDHw8Y/Da1+rh/MiItIJ69dH\nkFwMnB9+OILqZjKLdzNvt13r4Hn77SN4Hjq0uesT6WMUVIuINOippyKQvuCC2m8F2Xxz+MhHImd6\n8827L30iItIPLF7cNnB+5JEosr1uXXPXteWWbYPmLHAePry56xLpRxRUi4h00C23wDnnwG9/W7v6\n2aGHRq70W98adadFRERKucOLL8JDD1W6Bx+Ef/4zhjfTZpuVB87bbgsjRjR3XSIDhIJqEZE6uMOf\n/gRnnQU331x9uuHD41VYp54a9aZFRERe4Q7z51eC5nwQ3cwi28OGwQ47wI47wsyZ0e24YwwbM6Z5\n6xERQEG1iEhNGzZE42NnnRUtelczdSqccgp89KMwaVL3pU9ERHohd3juudaBc/Z5yZLmrWfq1NZB\nc9bfemu9TkKkGymoFhEpsW4dXHIJfOMbUXWtml13hU9/Go4/XtXNREQGHHd44QW4//7o8kH0smXN\nWceQIVE8u5jrvOOOMG5cc9YhIp2ioFpEJGf1avjJT+C//ite2VnNG98In/kMHHaYWvEWERkQVqyI\noPn+++G++yqB9MKFzVn+8OFRb2jnnWGXXaK/007RSNgmumUX6c30CxURAdasiVa8v/Y1eP758mnM\n4Nhj4fOfh7326t70iYhIN1m/PlrWzoLmrHvyyciZ7qwRIyJgzrosgFaRbZE+S0G1iAxo69fDT38K\nX/5y9ZzpwYPhve+Fz31OjY+JiPQbWaNhxeD5wQfjSWtnjR7dOnjOAuittoJBgzq/fBHpNRRUi8iA\ntHEj/OpXcMYZ8Oij5dMMGwYf+AD827/B9OndmjwREWmm9eujgYx77oG7747+vfc2p+h2S0sEy7vt\nFg1tZDnPW26p+kEiA4SCahEZUNzhiivgi1+MDIkyLS3xfunPfCZe5ykiIn3IihVR5zkfQN9/fzSa\n0RlmsN12ETznu223VbFtkQFOQbWIDBg33ACf/Szcfnv5+KFD4SMfiTrTm27avWkTEZEGvPBCBM35\nAPrRRztf93nKlLbB8847w8iRzUm3iPQrCqpFpN977LGoD/2735WPHzwYTjopcq+nTevetImISB02\nboQnnmgbQFdrWbJe+aLb+W7q1OakW0QGBAXVItJvLVoEX/kKfP/78d7pIrN4v/SZZ8YrQEVEpBfI\nAug774Q77ojurrtg+fLOLXfKlHh1w157wZ57Rrfddiq6LSKdpqBaRPqdtWvhvPPgS1+CxYvLpzn6\n6Gjxe9dduzdtIiKS4x6vqioG0EuXdm65228fQXMWQO+1l+r1iEiXUVAtIv2GO/z+99Fa92OPlU+z\n//5wzjlw4IHdmzYRkQHPHebMaRtAV3v6WY9hw+LpaD73effd43VWIiLdREG1iPQLjz4Kp5wCf/1r\n+fhp0+Cb34R3vlNvOBER6RYvvhgtQ952W/TvuCPq5TRq/PjWwfNee8GOO8KQIc1Ls4hIAxRUi0if\ntmoVfP3rETCvXdt2/OjR0Zr3qadGezQiItIFVq6MxsOyAPq22yJXulFjx8KsWbDPPpX+jBl6Kioi\nvZKCahHps666KnKnn3yy7bhBg+DDH4561VOmdH/aRET6rY0b4eGHWwfQ990HGzY0trwxYyJozgLo\nWbNgm20UQItIn6GgWkT6nGeegdNOg9/+tnz8IYfAd7+rRshERJrihRfg1lsjeL7tNvjHPxpviXvU\nqLYB9LbbxpNQEZE+SkG1iPQZ69bBd74Tr8BasaLt+KlToxGy449XBoeISEM2bIAHH4Sbb650TzzR\n2LKGDIE99ogWIvffH/bdF3bYQQG0iPQ7CqpFpE+45x446aToFw0aBJ/4RLyTeuzY7k+biEiftWxZ\n5D5nAfStt8awRmy7bQTP++0X/T33hOHDm5teEZFeSEG1iPRqa9fCWWfB174G69e3Hb/vvvFO6n32\n6f60iYj0Ke7w1FNw002VIPr++2N4R02YUAmes1zoSZOan2YRkT5AQbWI9Fr33AMnngj33tt23Lhx\n0er3ySfD4MHdnjQRkd5v/fpokfuGGyqB9AsvdHw5m2wSr6864IBKEL3ttqpnIyKSKKgWkV5n7drI\nmT7rrPLc6Xe9K+pWq1VvEZGc1aujEbEbboju5pvh5Zc7vpyJE+HAAyvdrFkwYkTz0ysi0k8oqBaR\nXuXuuyN3+r772o6bOjWKeh99dLcnS0Sk91m+HG65pRJE33ZbPJXsqJ13bh1E77CDcqFFRDpAQbWI\n9Arr10fu9Fe/Wp47/e53x2uyJk7s/rSJiPQKL70EN94YAfTf/w533dXxd0OPGBHFtw88EA46KIp0\njx/fNekVERkgFFSLSI+bOxfe8564VyyaOhV++EN429u6P10iIj3qpZdg9v9n777DrKquh49/N3VQ\nQMCGWMAWrIg/sRessURFxa4gVixRMdHYGxbEYO8t9mDBAsaCFWKP6Bt7wYqJHZQiHc77xx4yc2aG\nmbly5557Z76f57nPzF7nnOsygXHW3XuvPQaefz4W0u+9l/t7LLMMbLUVbLklbLEF9OgR90hLkvLG\nn6qSMvXII3D44fDzz9WvHXRQnJ3u1KnweUlSwU2ZEmegn38+vmo6Q7AuXbtWFNFbbeVSbkkqAItq\nSZmYMQP+/Oe4R7qqZZeFm26CPn0Kn5ckFcyMGbGZ2IIi+o03cl/O3b17LJ4XFNJduzZMrpKkhbKo\nllRw778fO3jXtJJxxx3hzjtjYS1Jjcrs2bFwXlBEv/JKbo3FQoD11qsooLfc0h+WklQELKolFUyS\nwC23wKBBcYKmspYt47nTJ50EzZplk58k5dX8+fEog2eeiUX0iy/Cr7/m9h49e8K228I228Q90R06\nNEyukqTfzKJaUkFMmQJHHAEPPlj92qqrwn33xaNQJamkff89PP10xeuHH3J7fo01YhG97baw9dYe\neSBJJcCiWlKD++AD2Gsv+Pjj6tcOOgiuvx7aty98XpK0yGbNikcXPP00jB4Nb7+d2/Mrr5wuort0\naZA0JUkNx6JaUoN64AE47LDqKx4XXzwW0/37Z5OXJP0mSRI/IRw9Or7GjoXp0+v//HLLVRTR22wT\ni2pJUkmzqJbUIObOhVNPhcsvr36tZ0+4//540oskFb1Jk+C55ypmo7/+uv7PtmsH220H228fv3bv\n7hFXktTIWFRLyruff4b99ou9eaoaMCDOULdpU/C0JKl+kiSeEf344/H1r3/FpmP1EQJsuCH8/vfx\nOIONN46dGCVJjZZFtaS8+ugj2G03+PTTdLxlS7jmGjjqKCdpJBWhadPg2WdjEf3EE/DNN/V/dvnl\nYwH9+9/HGWmbi0lSk2JRLSlvnnwynj89ZUo6vvzy8NBDccJGkorGp59WzEaPHVv/M6PLyqB371hI\n77gjrLmmnxZKUhNmUS0pL66+Op4xXXWF5KabwiOPwLLLZpOXJP3PnDnwz39WFNKffFL/Z9ddt2I2\nesstY2EtSRIW1ZIW0bx58Kc/xaK6qgED4MYboXXrgqclSdGUKXEZzciRcVn35Mn1e27xxeNS7j/8\nAXbZJS65kSSpBhbVkn6zX3+FAw+EUaPS8WbNYNgwGDTIFZGSMvCf/8QfTCNHwgsvxBnq+lhttYoi\nundvPxGUJNWLRbWk3+S772DXXeHNN9Pxtm3j2dQ775xNXpKaoCSBd9+NRfTIkdV/MC1My5aw1Vax\nkP7DHzznT5L0m1hUS8rZZ5/FbYWff56OL7983Ka43nrZ5CWpCZk3D15+GR5+OBbSX35Zv+eWXjp+\nIrjrrrDDDvEcaUmSFoFFtaSc/L//BzvtBD/8kI6vtx784x+wwgrZ5CWpCZg7F8aMiccJPPIIfP99\n/Z773e+gT5/42mQTaN68QdOUJDUtFtWS6u2FF+LvpFOnpuM77RSXfDvhIynvZs+G556DESPijPTE\niXU/E0I8w69PH9hjD1hjjYbPU5LUZFlUS6qXRx+F/farfoxrv35w221xa6Ik5cWMGfD003FGetSo\n+nXsbt06duvu0wd22w06d274PCVJwqJaUj38/e/Qv3/cwljZn/8Ml14au31L0iKZOTMeefXgg3Ev\nybRpdT/Trl0soPfaK54h3bZtw+cpSVIVFtWSanXbbXDkkbG5bmWXXgqnnJJNTpIaiTlz4Nln4b77\n4h7pqntLatKxY5yN7ts3zkyXlTV8npIk1cKiWtJCXX01nHhiOtasGdx6Kxx6aDY5SSpx8+bBiy/G\nQnrEiPrtkV566bg3eu+9YZtt3G8iSSoqFtWSajRsWPWZ6BYt4J574t5qSaq3JIF//QuGD49dDb/9\ntu5nllsuLuvee2/YYov4A0iSpCLkf6EkVXP55dUL6lat4lbH3XfPJidJJejdd2NThvvuq9850p07\nw777xtemm9qwQZJUEiyqJaVceWVsQFZZmzax+/fvf59NTpJKyLffxkL67rvh7bfrvr9jxzgbvf/+\n0Lu3Z0hLkkqORbWk/7nmGjjppHSsTRt4/PG4jVGSajR9evzk7a674JlnYP782u9v2zbukd5/f9hh\nh7gURpKkEmVRLQmAG2+EE05Ix8rK4sk2FtSSqpk/H8aMiTPSI0bUfQRWWRn84Q9wwAGwyy7xEztJ\nkhoBi2pJ/P3vcOyx6VhZGTz2GGy7bTY5SSpSH3wQC+l77oH//Kf2e5s1izPRBx0Uj8Fq374wOUqS\nVEAW1VIT99hj0L9/+hzq1q1h5Mh4BKwk8csvsXP3bbfBm2/Wff9668UfLAccELt4S5LUiFlUS03Y\nmDGwzz7x2NgFWrSAhx+2KZnU5CUJjB0bC+kRI2DmzNrvX265OCPdrx/06FGYHCVJKgIW1VITNW5c\nPB5r1qyKWAhxRecuu2SXl6SMffMN3HEH/O1v8Nlntd+72GLxLOl+/WC77ezcLUlqkiyqpSbo009j\n4Tx1ajp+002w337Z5CQpQ3PmxDb/t90GTzxRe/fuEGIB3a9fLKjbti1cnpIkFSGLaqmJ+fFH2Hnn\n+LWyoUPhyCOzyUlSRj7+OBbSd90F339f+71du8Jhh8GAAbDSSgVJT5KkUmBRLTUh06fDbrvFmerK\nTj0V/vKXbHKSVGCzZ8czpW+4ITZWqE2rVrDnnnDEEfEogGbNCpKiJEmlxKJaaiLmzo2NeF9/PR3v\n1w+GDMkmJ0kF9PXXcPPNcOut8N13td/bowccfnhsPLbkkoXJT5KkEmVRLTUBSQInngijRqXj220X\nf78OIZu8JDWw+fPh2Wfh+uvj+Xm17ZVu3x4OPDAW0xts4A8GSZLqyaJaagKuuy7+Tl1Zjx7w0ENx\ndaekRmbixNjB+4Yb6u7gvcUWsaHC3nvHbt6SJCknFtVSI/fMMzBoUDq2wgqxwe8SS2STk6QGkCTw\nr3/FQvq++9Ln5VXVrl3c+3HMMbDOOoXLUZKkRsiiWmrEPv4Y9tkH5s2riLVtG0/OWX757PKSlEez\nZ8OIEXDllfDGG7Xfu+66cOyxca90u3aFyU+SpEaupNt4hhBWDCGMCCFMDiFMCSE8HEKo9zkfIYQ1\nQwgPhhB+CiHMCCF8HEI4sSFzlgpl0qTY6Xvy5IpYCPD3v8el35JK3MSJscvgyivHInlhBXWrVnGv\n9Esvwdtvw9FHW1BLkpRHJTtTHUJYDHgemAUcAiTAhcALIYQeSZL8WsfzvcqfHwMcAUwGVgfaNmDa\nUkHMnQv77gvjx6fjQ4fGQltSCfvgA7jqKrj7bpgxY+H3desGAwfGs6WXWaZg6UmS1NSUbFENHAms\nAnRPkuRTgBDCO8B4YCBw+cIeDCE0A+4CnkuSZM9Kl15ouHSlwjnzTHjuuXTskEPg5JOzyUfSIkoS\nGD06LvEePbr2e3fZJS7x3mknaN68MPlJktSElXJRvTvw2oKCGiBJki9CCC8DfailqAa2BtYkFt9S\no/LQQ3DppenY5pvDTTd5Qo5UcqZPjzPSV10FH3648PsWWwwGDIhn5/3udwVLT5Ik5binOoTQLoRw\nTgjhpRDC+BDCpuXxpcrjazRMmjVaG3ivhvj7wFp1PLtF+deyEMJrIYQ5IYQfQghXhxDa5DVLqYA+\n+ij+Xl1Zly6x0G7dOpOUJP0W338fl5ysuGLcA72wgnrFFeOnaP/5Tzw7z4JakqSCq/dMdQhhaeAl\n4pLrT8u/tgFIkuSnEMIhQAfgTw2QZ006AT/XEJ8EdKzj2S7lX+8HrgVOA3oBg4EVgT1reiiEcBRw\nFMBKK9W7H5pUEFOnwl57wbRpFbGWLWNT4GWXzS4vSTkYPx6GDYM776z9SKxNN41n5e21F7Qo5UVn\nkiSVvlz+S3wh0BnYGJgA/FDl+khguzzl1dAWzNDfkyTJOeXfjwkhNAcuCSGsmSRJtWmBJEluBm4G\n6NWrV1KYVKW6JQkcfnj1yawrr4y/e0sqcm+8ETsJPvxw/Atdk+bN4xl5gwbBxhsXNj9JkrRQuSz/\n3hW4PkmSt4idtqv6nDjLWyg/U/OM9MJmsCubWP71mSrxp8u/rr8IeUkFd8MN8OCD6Vi/fnDMMdnk\nI6keFjQf23Zb2GijuE+jpoK6Y0c47TT48ksYPtyCWpKkIpPLTPVSxGXfCzMfKFu0dHLyPnFfdVVr\nAR/U49nazP9NGUkZ+Pe/4U9VNl2stx7ceKONyaSiNG9enJG++OL4F3hhunaFP/85Hom1+OKFy0+S\nJOUkl5nq74BVa7m+PnFZeKGMAjYJIayyIBBC6AZsXn6tNk8Sz7fesUp8p/Kv4/KTotSwpk2D/fdP\nb71s1y5OeC22WHZ5SarB3Lmxk/c668SD5BdWUK+3Htx7b9xfffzxFtSSJBW5XIrqJ4DDQwjLVb0Q\nQtgY6E/cV10otwBfAiNDCH1CCLuX//O/Bm6qlFvXEMLcEMKCvdMkSTIRGAIcHUK4OISwfQjhNOAc\n4M7Kx3RJxez44+Hjj9Oxm2+GVWv7+EtSYc2aBbfcAt27Q//+sU1/TbbZBp56Cv7f/4MDD4ydBiVJ\nUtHLZfn3+cSzof8fcSY4AQ4JIRwJ7AV8AwzNe4YLkSTJryGEbYErgLuBADwHDEqSpFL/YwLQnOof\nIAwGpgLHAicD3wJ/BS5o4NSlvLjnHrjjjnTsiCPizLWkIjBjBtx6a8WRVzUJIXbwPvVU2HDDwuYn\nSZLyIiQL6zJa080hrEg8guoPVBSpCXEW+5gkSRbyW0Pj06tXr2TcOFeJKxuffQY9e6aPz1pzTRg3\nzmXfUuZmzIhNDYYOjedN16R5czjoIDj9dFhjjcLmJ0mS6iWE8GaSJL3qui+nwy2TJPka6BNCaA90\nJ84Cf5okyaTflqakXM2bF1eQVi6oy8rggQcsqKVMzZwZ918MGQLffVfzPS1bwqGHxpnpVVap+R5J\nklRS6l1UhxD6A/9MkuTLJEmmAG9Uud4N2CpJkrvymqGklL/+FV55JR274orY+0hSBmbNgttui928\n//vfmu8pK4OjjoJTToEVVihsfpIkqUHl0qjsdmCzWq5vXH6PpAby73/DOeekY7vuCgMHZpOP1KTN\nng033QSrrw7HHVdzQb344rGQ/vJLuOoqC2pJkhqhXJZ/13XibUs831lqMLNmQb9+MGdORWzJJWNT\nYc+jlgpo3rx4NNb558diuSZt2sAf/xgL6qWXLmh6kiSpsHLaU01sSlZNCKEDsXnZt4uckaQanX02\nvPdeOnbzzdC5czb5SE1OksDIkXDmmfDBBzXfU1YGxxwT90wvu2xh85MkSZmodfl3COHcEMK8EMI8\nYkF9z4Jx5RcwEdgXuK8AOUtNzssvw7Bh6Vj//vEkHkkFMGYMbLop7LlnzQV169Zwwgnw+edw+eUW\n1JIkNSF1zVT/G7iLuPS7P/Ai8HmVexJgGvAaMDzfCUpN3cyZ8fzpyqffrbgiXH11djlJTcZbb8EZ\nZ8Do0TVfb9kyNiA7/XRYfvnC5iZJkopCrUV1kiQjgZEAIYSuwIVJkjxXiMQkRRddBB99lI7dfjss\nsUQ2+UhNwqefwllnwf3313w9hLhc5LzzoFu3QmYmSZKKTL33VCdJsk1DJiKpunfegUsuSceOPBK2\n2y6bfKRGb+JEuOACuP76dFfAyvr0gQsv9Bw7SZIE5N6oDIAQQlugAzXsyU6SZMKiJiUJ5s6Fww+P\nXxdYbjm49NLscpIarVmz4NprY7H8yy8137PVVvFTrk03LWxukiSpqOVUVIcQ9gfOAtas5bbmi5SR\nJCAeaTtuXDp2/fXQoUM2+UiNUpLAgw/CaafBF1/UfM/668PFF8OOO3p+nSRJqqbW7t+VhRD2AP5O\nLMRvIjYvGw48CMwB3gQGN0COUpPz2WfxCK3K9t4b9tgjm3ykRunVV2HzzWG//WouqLt2hXvvjZ9u\n7bSTBbUkSapRvYtq4GTgQ6AncE557G9JkuwP9AK6E7uFS1oESQLHHw8zZlTEOnaEa67JLiepUfn6\na9h/f9hss1hYV9W+PQwdGjsEHnggNMvlP5WSJKmpyeU3hR7AnUmSzATml8eaAyRJ8h5wM3B6ftOT\nmp5Ro+DJJ9Oxyy6Dzp2zyUdqNGbMiE3Iunevuat38+Zw3HGx8/df/gJlZYXPUZIklZxc9lQ3ByaW\nf79gDq3yoT4fA8fkIympqZo+HU48MR3bcksYMCCTdKTGIUngkUfgz3+GL7+s+Z7dd49dALt3L2hq\nkiSp9OUyU/0foCtAkiQzgB+ADSpd7w78mr/UpKZnyBD46quKcfPmcN11buWUfrP334cddoC+fWsu\nqNdfH154AUaOtKCWJEm/SS4z1a8A21Oxn3oUMCiEMINYnB8HPJbf9KSmY/z46sdlnXACrLtuNvlI\nJe2XX+C88+IxWfPmVb++1FLxU6xDD42fXkmSJP1GuRTV1wN7hhDalM9UnwlsBJxXfv19YjMzSTla\n0Jxs9uyKWOfOsSaQlIMkgbvvhpNPhh9/rH69eXP44x/h3HNjB0BJkqRFVO+iOkmSN4A3Ko1/BHqG\nEHoA84APkySZv7DnJS3co4/C6NHp2GWXxSbEkurpww/h2GNhzJiar2+3XTwAfu21C5qWJElq3HKZ\nqa5RkiTvLPg+hBCSJEkW9T2lpmT2bDjllHRs663hgAMySUcqPdOnw4UXwrBhMGdO9evdusHll8eD\n3m1QIEmS8iwvh2+G6GDgo3y8n9SUXHcdfPZZxbh587gN1N/9pXp4/PE48zxkSPWCunVrOP98+OAD\n2HNP/1JJkqQGUa+Z6hDC+sDqwCRgbJIkcypd2x84l9j9e2pDJCk1VhMnwuDB6djAga5Oler09dcw\naBA8/HDN13faKX46teqqhc1LkiQ1ObUW1SGENsAjwA6Vwl+HELYHZgH3AZsQi+mLgCsaKE+pURo8\nODYpXqB9e5uTSbWaPx+uvx5OPx2mTat+vUsXuPJK2HtvZ6YlSVJB1DVTfSrwe2AcMAZYDdgDuBFY\nEVgeuAC4IkmSyQ2XptT4fPJJrA0qO/NMWHrpbPKRit6HH8IRR8Arr1S/1qxZbKE/eLAd/iRJUkHV\nVVTvDbwEbL2gs3cI4RziMVoTgB5JknzeoBlKjdRf/gJz51aMu3aN51JLqmLOHBg6FC64IH3u3AIb\nbQQ33gjrr1/43CRJUpNXV6OyVYCHqhyVdX/510stqKXfZuxYGDkyHRs6FMrKsslHKlrjxkGvXnD2\n2dUL6vbt43KPV16xoJYkSZmpq6guA36sEvup/Ov4/KcjNX5JAqeemo5tsgnsu282+UhFafr0eNbc\nxhvDO+9Uv77bbvD++3DMMbFlviRJUkYW5ZzqeXnLQmpCHnsMXn89HbvsMnsqSf/zz3/CYYelz5pb\nYOml4Zpr4qdQ/qWRJElFoD5Fdf8QwiaVxmVAAvwxhLBHlXuTJElOzFt2UiMzfz6cdVY61qcPbLZZ\nNvlIRWXmzNit74or4pKOqg4+OF5baqnC5yZJkrQQ9Smqf1/+qqpqQQ2x2Laolhbi/vvh3XcrxiHE\n3ktSkzduHPTvHzt8V7XiinDTTbDzzoXPS5IkqQ51FdUrFyQLqQmYOxfOPTcd239/WHfdbPKRZfZ9\nCgAAIABJREFUisKcOXDhhXDRRTCvhl1Fxx0HQ4ZAu3aFz02SJKkeai2qkyT5qlCJSI3dnXfC+Ert\n/Zo3h/PPzy4fKXPvvx9np996q/q1lVaC22+HbbctfF6SJEk5qKv7t6Q8mDWregF96KGw+urZ5CNl\nat48GDYMNtig5oL60ENjx28LakmSVAIWpfu3pHq66Sb4+uuKcatW8dhdqcn5z3+gXz8YM6b6tWWX\nhVtuicdlSZIklQhnqqUGNmsWDB2ajh1zTFzdKjUpDz8MPXrUXFDvsw+8954FtSRJKjkW1VIDu+MO\n+OabinGbNnD66ZmlIxXe9OkwcCD07Qs//5y+1rEjDB8ODzzgUVmSJKkkufxbakBz5sAll6RjAwfG\nVa5Sk/Dvf8MBB8BHH1W/tv32sYNfly6Fz0uSJClPnKmWGtDw4fDllxXjVq3g5JMzS0cqnPnz4cor\nYeONqxfULVvCX/8Ko0dbUEuSpJK3SDPVIYQWQB+gE/BYkiTf5SUrqRGYNw8uvjgdO/RQWH75bPKR\nCub772HAAHjqqerXVl89ftq0wQYFT0uSJKkh1HumOoRwaQjhjUrjADwLPADcBLwbQlg1/ylKpenh\nh+HjjyvGzZvDqadml49UEGPHQs+eNRfUhx0Wj9CyoJYkSY1ILsu/dwJerDTeDdgK+CtwYHnstDzl\nJZW0JIGLLkrHDjoIVl45m3ykBjd/fmwgsO228F2VRUtLLAH33w+33QZt22aTnyRJUgPJZfn3isD4\nSuPdgC+SJDkNIISwNnBQHnOTStbjj8Pbb1eMQ7DjtxqxSZOgf//4B7+qLbaAe+6Brl0Ln5ckSVIB\n5DJT3QqYW2m8DXH59wKfA8vlIymp1A0Zkh7vsw+ssUY2uUgN6l//gvXXr7mgPvNMeOEFC2pJktSo\n5VJUfw1sCv+blV4FGFvp+jLAtPylJpWm116DV15Jx844I5tcpAaTJHDNNXEmesKE9LVOneCJJ+DC\nC6GFJzdKkqTGLZffdu4Dzg4hLAOsDUwBnqh0fX3gszzmJpWkK65Ij3fcEdZbL5tcpAYxZQoccQQ8\n+GD1a5tsEvdPr7RS4fOSJEnKQC4z1UOAO4iz1QnQP0mSXwBCCEsAuwPP5TtBqZR8+SWMGJGO/elP\nmaQiNYyPPoKNNqq5oB40KHb/tqCWJElNSL1nqpMkmQUcXv6qaipxP/X0POUllaRrrolNkBdYe23Y\nYYfs8pHy6rHHYhv7qVPT8fbt4W9/g759s8lLkiQpQ7nMVNemZZIkk5MkmZOn95NKzpQpcMst6dif\n/hQ7f0slbf58uOAC2H336gX1euvBm29aUEuSpCar3kV1CGHnEMJ5VWLHhhCmAL+GEP4eQmiZ7wSl\nUvG3v6XrjWWWgQMPXPj9UkmYOjW2rz/nnOrX+veHV1+F1VYrfF6SJElFIpdGZacAPywYhBDWBK4i\nNif7AtgP+BdwZT4TlErB3Llw1VXp2LHHQllZNvlIefHZZ9CnD7z/fjrevDlcfjkcf7xLMSRJUpOX\ny/LvNYFxlcb7ATOAjZIk2Rm4Hzgkj7lJJePRR2OTsgVat4ZjjsksHWnRPf009OpVvaBeckl45hk4\n4QQLakmSJHIrqjsCP1Uabw88nyTJlPLxGGDlPOUllZSqx2j16xeXf0slJ0lg2DDYeWf45Zf0tZ49\nYdw42GabbHKTJEkqQrkU1T8BXQFCCO2ADYEXK11vCTTPX2pSaXj7bXjllXRs0KBscpEWyezZcOSR\ncMop6Tb2APvvDy+/DN26ZZKaJElSscplT/WrwNEhhPeBncuffbLS9dWAb/OYm1QSbrghPd5uu3iU\nllRSfv4Z9t4bnn8+HQ8BLrkkFtou95YkSaoml6L6XOAF4IHy8Z1JknwAEEIIwJ7l16UmY8oUuOee\ndMy91Co5n38Of/gDfPRROt6hAwwfDjvtlE1ekiRJJaDeRXWSJB+Ud/zeHJicJMk/K13uAFxB3Fct\nNRn33AO//loxXm65eJSvVDJeeSV2+P7pp3R8lVXg8cdhjTWyyUuSJKlE5DJTTZIkk4DHaoj/TDxe\nS2oykgRuvDEdO/JIaOlp7SoVw4fDoYfCrFnp+GabxZb2Sy+dTV6SJEklJJdGZQCEELYKIVwYQrgl\nhLBGeaxtebxD/lOUitMrr8C771aMmzePRbVU9JIELrwQDjywekF9wAHw3HMW1JIkSfVU76I6hNA8\nhHA/cd/0GcBhQJfyy3OBR4Fj856hVKSqNijbbTdYYYVscpHqbfZsGDAAzj67+rVzzoF774WysoKn\nJUmSVKpymak+FegL/AlYE/hfG9gkSWYCjwC75DU7qUj9+CM8+GA6ZoMyFb2pU2HXXeGuu9Lxli3h\nzjvh/PPt8C1JkpSjXPZU9wfuSpLkqhDCkjVc/xCLajURt98eJ/wWWHVV2H777PKR6vTjj7DLLjBu\nXDreqRM88ghstVU2eUmSJJW4XGaquxHPql6YX4COi5SNVALmz4ebb07HBg6EZjl3KJAK5MsvYfPN\nqxfUq60Gr71mQS1JkrQIcikDpgKdarm+GvDjoqUjFb+XXoLPPqsYt2oVGyhLRemdd2I37/Hj0/EN\nNoCXX4bVV88mL0mSpEYil6L6JeDgEKpvuAshdCQ2LnshX4lJxeqOO9LjvfaCpZbKJBWpdv/8Z5yF\n/vbbdHz77eGFF2CZZbLJS5IkqRHJpai+CFgdeB7YtTy2XghhIPAWsDhwSX7Tk4rLtGnwwAPp2IAB\nmaQi1e7RR+H3v4fJk9Px/feHxx+Hdu2yyUuSJKmRqXdRnSTJOGL37zWA28vDw4AbgDbAnkmSfJD3\nDKUi8vDD8OuvFeMuXWxQpiJ0yy3Qt2/1M6iPPz4emdWqVTZ5SZIkNUK5dP8mSZLHQwjdgB2oOFZr\nPDA6SZLpec9OKjJVl3737w/Nm2eSilSzSy+FU0+tHr/4YjjtNI/MkiRJyrOcimqAJElmAf8of0lN\nxpdfxm2olR1ySCapSNUlCQweDOedl443axbb1R9+eCZpSZIkNXb1Xv4dQlg/hHBcLdePCyH0zE9a\nUvG56670eJNNYI01sslFSkkSOOOM6gV1WVk8g9qCWpIkqcHk0qjsXOAPtVzfGThn0dKRitP8+dWX\nftugTEUhSeCkk+CSKn0iF18cnnoKdt89m7wkSZKaiFyK6g2BsbVcHwtstGjpSMXpxRfhiy8qxmVl\nsN9+2eUjAfHTnmOOgauuSsfbt4enn4bevbPJS5IkqQnJpaheCphUy/Vfyu8pmBDCiiGEESGEySGE\nKSGEh0MIK/2G9zkthJCEEF5qiDxV+qrOUu+5J3TokEkqUjRvXlzWfdNN6XjHjvDcc7DZZtnkJUmS\n1MTk0qjsB2DtWq6vQ+1Fd16FEBYjnpk9CzgESIALgRdCCD2SJPm1tucrvc8qwFnEfz+pmunT4cEH\n0zGXfitTc+bE1vP33ZeOL700PPss9OiRTV6SJElNUC4z1c8CR4QQqhXWIYS1gMPL7ymUI4FVgD2S\nJHk0SZKRwO5AV2BgDu9zA3Av8GH+U1Rj8I9/pM+mXn552G677PJREzd7dtx7ULWgXm45GDvWglqS\nJKnAcimqLwTmAW+EEK4NIRxR/roWGAfMBS5oiCQXYnfgtSRJPl0QSJLkC+BloE993iCEcCDwf8Dp\nDZKhGoX770+P99/fs6mVkdmzYZ99YkfvylZcEf75T1hzzWzykiRJasLqvfw7SZLPQgjbAXcAx1a5\n/D5waJIk4/OYW13WBkbWEH8f2Keuh0MIHYErgL8kSTIphJDn9NQYTJkCjz+ejtmgTJmYOxcOOghG\njUrHV14Znn8eunXLJC1JkqSmLpc91SRJMg5Yp/w86tXLw58kSfJ23jOrWyfg5xrik4CO9Xj+r8An\nxA8JpBqNHAmzZlWMV1kFevXKLh81UfPmwSGHwIgR6fjqq8eCeoUVsslLkiRJuRXVCyRJ8m/g33nO\npWBCCFsC/YH/S5IkyeG5o4CjAFZaKecm4ypBVbet7r8/uKhBBTV/Phx5JPz97+n4qqvCmDHQpUsm\naUmSJCmq957qEMJ2IYQhtVwfEkLYJj9p1cvP1DwjvbAZ7MpuAm4D/hNC6BBC6ED8gKF5+bh1TQ8l\nSXJzkiS9kiTptfTSSy9K7ioBEyfGo34rc+m3CipJ4Ljj4Pbb0/Fu3eIMtQW1JElS5nJpVHYqsFot\n11cuv6dQ3qfmI77WAj6o49k1gaOJxfeC1+bAJuXfH5O/NFWqHnkkbmNdYM01Yd11s8tHTUySwEkn\nwY03puMrrBALalfLSJIkFYVciur1gNdquf56+T2FMgrYpPycaQBCCN2IxfGohTyzwDY1vN4G3iv/\nfsTCH1VT4dJvZSZJ4PTT4aqr0vHOnWNBvfLK2eQlSZKkanLZU70E8Gst12dQvwZh+XIL8EdgZAjh\nLCAhHun1NXF5NwAhhK7AZ8DgJEkGAyRJMqbqm4UQfgFa1HRNTc9338ELL6RjLv1WwZx/Pgwdmo4t\nvTQ891xsTiZJkqSikctM9X+BDWq5vgHw3aKlU39JkvwKbEvs4H03cC/wBbBtkiTTKt0agObk9u+q\nJm7EiNgfaoGePaF79+zyURMydGgsqivr1AmefRbWWiubnCRJkrRQucxUPw4cHUK4P0mSZytfKD+/\n+hDg1nwmV5ckSSYAfeu450tiYV3Xe22dn6zUGNS09FtqcLfdBqedlo4tsUTsmNejRzY5SZIkqVa5\nFNUXEQvY0SGEJ6k4UqsnsDNxlvqC/KYnFd5//wsvv5yOufRbDW7UKDjqqHSsbVt46inYoLZFQpIk\nScpSvYvqJEm+DyFsBtxALKJ3WXAJeBL4Y5Ik3+Y/RamwRo5MjzfaKJ5gJDWYl16Kn9xU3nPQujX8\n4x+wySbZ5SVJkqQ65TJTTZIkXwG7hBA6UnG81qdJktR1LrRUMh59ND3ea69s8lAT8d57sNtuMHNm\nRaxZs7gHoXfv7PKSJElSveRUVC9QXkS/kedcpMz98kv1rt977JFNLmoCvvoKdtwx/sGr7MYb/YMn\nSZJUIupdVIcQVqrPfeXNw6SS9OSTMHduxXiNNez6rQby00+xoP7mm3T8ggvgyCOzyUmSJEk5y2Wm\n+kvi/um6NP9tqUjZq7r0u0+fbPJQI/frr7DrrvDxx+n4ccfBmWdmk5MkSZJ+k1yK6sFUL6pbAKsC\nfYB3iQ3LpJI0axY88UQ65gpc5d2cObDPPvD66+n4PvvAVVdBqPMEQEmSJBWRXLp/n7ewayGEVYBX\ngXF5yEnKxPPPw7RpFePOnWPnbylvkgSOOCLuM6hs223h7ruhuQt9JEmSSk2zfLxJkiSfAzcB5+fj\n/aQs1LT0u1le/oZI5S64AO66Kx3r2RMeeSQeoSVJkqSSk8+S4b/AWnl8P6lg5s+HUaPSMfdTK6/u\nuw/OPTcdW2WVOGvdvn02OUmSJGmR5bOo3gPwvGoVhSSJpxXNn1+/+//1L/juu4px27ZxRa6UF6+9\nBgMGpGOdOsHo0XGfgSRJkkpWLkdqnbOQS52AbYF1gEvzkZS0KKZPh622gjffjMdhjR0Lyy5b+zNV\nl37vsourcZUnX34Zlz3MmlURa9kyLvlebbXM0pIkSVJ+5NL9+7xarn0HnAUMXaRspDy49dZYUEM8\nsej00+Fvf6v9mapFtV2/lRdTpsBuu8EPP6Tjt94aP/mRJElSyculqF65hlgCTEqSZFoN16RMPP98\nenzXXXD22bByTX+CiYV35eOCW7aMM9XSIpk7F/bbD957Lx0/4wzo3z+bnCRJkpR39d5TnSTJVzW8\nJlhQq5jMnw8vvZSOzZsHQ4Ys/Jmqpxv17g1LLJH/3NTEnHQSPPVUOrb33rEDuCRJkhqN39yoLITQ\nIoSwWQhhnxDC2vlMSvqtPvoIJk6sHr/jjti4rCZV6x5nqbXIrr02virbcEO4807PaZMkSWpkav3t\nLoSwdQjh6hDCMlXiKwNvAi8C9wHvhBDq2LUqNbwXX6w5PmcODK1hx/+MGbGRWWU775z/vNSEPPUU\nnHhiOrbiijByJCy2WDY5SZIkqcHUNWUyANgxSZIqXXa4A1gXeAW4AvgAOCSEcEi+E5RysbCiGuC2\n2+A//0nHxo6FmTMrxl27xo7h0m/y4Yew777ps9zatoXHHoPllssuL0mSJDWYuorqjYCnKwdCCGsA\nWwL/TJJkyyRJTi6/bzxg9x1lqmpR3aZNxfezZ8OlVQ59q7r0e6edIISGyU2N3C+/xKOzpk6tiDVr\nBsOHw3rrZZeXJEmSGlRdRXVnYrFc2dbErt+3LggkSTID+DvQI5/JSbn46iuYMKFi3Lp19SL65pvh\n228rxjUV1VLO5s+Hgw+G8VV+XA4bBrvumk1OkiRJKoi6iurWwIwqsQ3Lv1bZicrXgD2TlZmqs9Qb\nbQRHHRW3sy4waxZcdFH8/osv0kdptWgB227b8HmqETr3XHj88XRswAAYNCiTdCRJklQ4dRXVE4Cq\nnb23AH5IkuTrKvHFgF/ylZiUq6pF9ZZbQqtWcPrp6fhNN8Enn8Do0en45ptD+/YNm6MaoYcfhgsv\nTMc23BBuuMG9BJIkSU1AXUX1i0D/EMI6ACGEPYHVgSdruHdd4L/5TU+qv5qKaoDDDoOVV66Iz50b\nC+2qS7/t+q2cffAB9K/SSmLZZWOhXVaWTU6SJEkqqLqK6iHEJeBvhxB+AEYAs4HLKt8UQmgO7A68\n1BBJSnX56afYeHmBZs1gs83i961bw5Ah6fsffhieeCIdcz+1cjJ1Kuy1F/z6a0WsRQsYMQJWWCG7\nvCRJklRQtRbVSZJ8AfQGngAmEmeot06S5P0qt25Tfn1kQyQp1eWlKh/n9OyZXsq9775xRW5lc+ZU\nfN+5M/SwzZ7qK0ng8MPTm/IBrr4attgim5wkSZKUiRZ13ZAkyThgtzrueZa4/FvKxMKWfi8QQmzE\n3Lt3zc97lJZyctVV8OCD6diAAXD00ZmkI0mSpOzUtfxbKgl1FdUAW20Fu+9e8/Mu/Va9vfQSnHJK\nOrbeenDddX4yI0mS1ARZVKvkTZsGb72Vji1sBe7QodC8eTrWrBnssEPD5KZG5vvv416CuXMrYkss\nAQ89BIstll1ekiRJyoxFtUrea6/BvHkV49/9LjZgrskaa8ARR6RjG28MnTo1XH5qJObNg4MPhm+/\nTcfvugtWXTWbnCRJkpQ5i2qVvKpNympa+l3Z+eenmzMPGpT/nNQIDRkCzz6bjp1++sL3FEiSJKlJ\nqLNRmVTsXn89Pd5889rvX3ZZeOUVGDUK1lkn7rWWajV2LJx7bjrWuzcMHpxNPpIkSSoaFtUqaUkC\nb7yRjm28cd3PrbgiHHdcw+SkRubHH+HAA2H+/IrY0kvD3/8ez6WWJElSk+byb5W0L76AiRMrxu3a\nQffu2eWjRmb+fOjXD775Jh2/+27o0iWbnCRJklRULKpV0v71r/S4V6/q3b2l32zYMBg9Oh07/XTY\nccds8pEkSVLRsahWSataVG+0UTZ5qBEaNw7OPDMd22IL91FLkiQpxaJaJc2iWg1i2jQ44ID0edSd\nOsHw4e6jliRJUopFtUrWnDnw1lvp2IYbZpOLGpkTToBPP03HbrstfRabJEmShEW1Stj778OMGRXj\nzp2teZQHDzwAt9+ejh19NOyxRzb5SJIkqahZVKtk1bT0O4RsclEjMWECHHVUOrbmmnDZZdnkI0mS\npKJnUa2S5X5q5dX8+XDIITB5ckWsVat4HvVii2WXlyRJkoqaRbVK1htvpMcW1VokV14JY8akY0OH\nQs+emaQjSZKk0mBRrZL066/w3nvpWK9e2eSiRuC99+L505XtsENsWCZJkiTVwqJaJemtt+Jq3QV+\n9zvo2DG7fFTCZs+Ggw+OXxfo0CE2K2vmj0hJkiTVzt8YVZKq7qf2KC39ZuedB2+/nY5dfz0sv3wm\n6UiSJKm0WFSrJNmkTHnxyitx33Rl++0HBxyQTT6SJEkqORbVKkkW1Vpk06fDgAHpfQRdusRZakmS\nJKmeLKpVcn74Ab78smLcooUNmvUbnH02jB+fjv3tb9CpUzb5SJIkqSRZVKvkVD1Ka731oKwsm1xU\nAqZPh6lT07FXXoErrkjHBg6EHXcsXF6SJElqFCyqVXJc+q16u+++OPPcsSP06wcTJsCMGXDooZAk\nFfd17Qp//Wt2eUqSJKlkWVSr5Lz1Vnps52/V6Isv4LDDYNYsmDcP7rknnr227bbwySfpe2+7Ddq1\nyyZPSZIklTSLapWcqqcfuZ9a1SQJHHNMnJWubNYseO21dGzgQNhuu8LlJkmSpEbFololZdIk+Prr\ninGLFrDWWtnloyJ1330wenTd9620Elx6acPnI0mSpEbLolol5Z130uM114TWrbPJRUXgiSegRw/Y\nYgsYMSLOUE+aBCeemL5v3XXjcVlV3XYbtG9fmFwlSZLUKLXIOgEpF1WXfq+3XjZ5qAh88QX07Qsz\nZ8bxyy/DppvCUkvBjz9W3NeqFTz4IKywAlx5JVx+OUybBoMHw/bbZ5O7JEmSGg2LapUUi2r9z0kn\nVRTUC7z6avX7zjoLuneP3595Jpx2GsyZ4zlskiRJyguLapUUi2oB8OSTMHJk3fetuSacemo61rx5\nfEmSJEl54J5qlYw5c+C999Ixi+oSULUD96KaNQtOOCEd69Ch5ntvvjku/5YkSZIaiEW1SsbHH8Ps\n2RXj5ZaDZZbJLh/VYeJE2GQTWHxx6N0b3ngjP+972WXw6acV4xDguedgzBjo1asifvrpsYGZJEmS\n1IBc/q2S4dLvEnPKKfD66/H7f/4TNt4YDjsMLr74t38aMmECXHhhOjZwIPzf/8XvX38d3norLu/2\nAHNJkiQVgDPVKhkW1SVk/Hi46650LEniEVa/+x1cdVVcz5+rP/85vZx8ySXhoosqxs2axdnq9deP\nM9iSJElSA7OoVsmwqC4hgwfDvHk1X5s8GQYNioXv88/X/z3HjIlnUVd28cXQqdNvTlOSJElaVBbV\nKhkW1SXiww/h3nvTsWWXrX7f++/DdtvB3nvDV1/V/p7z5sUjtCrbYAM4/PBFy1WSJElaRBbVKgnf\nfx9fC7RuHVcRqwidd15c6r3A2mvHovnyy6F9++r3P/RQPEf6lFNg0qSa3/OOO+Df/07HrrnGo7Ek\nSZKUOYtqlYSqs9TrrAMtbLNXfN59Fx54IB07//z4KchJJ8UW7gMGVH9u1iwYNgxWWQWGDIHp0yuu\nTZkCZ56Zvv+AA2DTTfOeviRJkpQri2qVhKqTlDZ2LlLnnpse9+wJe+5ZMe7cGW6/HV57DTbcsPrz\nkyfDGWfAaqvBTTfFZmZDhqSXKZSVwSWXNEz+kiRJUo4sqlUS3E9dAt56Cx55JB07//zYkbuqjTeO\nhfVtt8Hyy1e//u23cPTRcen45Zenr51yCqy0Uv7yliRJkhaBRbVKgkV1CRg8OD3u1Qt2223h9zdr\nFs+tHj8ehg6FDh2q3zN+PMyeXTHu0gX+8pf85CtJkiTlgUW1it7MmfDRR+lYjx7Z5KKFeOcdGDky\nHRs8uH5nRbdpEwvlzz+PX8vKFn7vkCHQtu2i5SpJkiTlkUW1it4HH6SPPO7ateZJTWXooovS4w03\nhJ12yu09OnaMM9bjx8MRR1RfNt6rFxx88KLlKUmSJOWZRbWKnku/i9xHH8GDD6ZjZ51Vv1nqmqyw\nAtxySzzHum9faNUqHrl1770178+WJEmSMuShRCp6FtVFbsiQ9LnUPXrUvpe6vtZYA0aMiB3AW7T4\n7UW6JEmS1IAsqlX03n03PbaoLiKffx5nkCtblFnqmrRsmb/3kiRJkvKspNdShhBWDCGMCCFMDiFM\nCSE8HEKo86ydEEKvEMLNIYSPQgjTQwgTQgj3hhBWLkTeys2HH6bHa6+dTR6qwdCh6Q3va6wBe+2V\nXT6SJElSgZVsUR1CWAx4HlgDOAToB6wOvBBCWLyOx/cH1gauBnYGTgP+DxgXQlixwZJWziZPjkcW\nL9CiBay6anb5qJKvv4bbb0/HzjgDmjfPJh9JkiQpA6W8/PtIYBWge5IknwKEEN4BxgMDgctreXZo\nkiQ/Vg6EEF4Gvih/33MaJGPlrOpRWquv7mrgojFsWNzvvMAqq8ABB2SXjyRJkpSBkp2pBnYHXltQ\nUAMkSfIF8DLQp7YHqxbU5bGvgB+B5fOcpxZB1aXfa66ZTR6N0pQpcOONcbZ56tTcnp04EW69NR07\n/fS4lECSJElqQkq5qF4beK+G+PvAWrm+WQhhTWAZ4MO67lXhVC2q11gjmzwanSSBffeFY46Bww6D\n1VaDm2+GuXPr9/z118P06RXj5ZeH/v0bJldJkiSpiJVyUd0J+LmG+CSgYy5vFEJoAdxInKm+rZb7\njgohjAshjPvxx2qT3WoAzlQ3kFdfhdGjK8Y//AADB0LPnul4TWbMgKuvTsdOOimeJy1JkiQ1MaVc\nVOfTtcBmwMFJktRUqAOQJMnNSZL0SpKk19JLL1247Jowi+oGcsMNNcfffx922im+3qtpIQhxufhP\nP1WMl1gCjjwy/zlKkiRJJaCUi+qfqXlGemEz2DUKIVwCHAUcliTJ03nKTXkwc2Y8Brkyl3/nwU8/\nwQMP1H7P6NHxQPCBA+H77yvic+fCZZel7z3mGGjfPv95SpIkSSWglIvq94n7qqtaC/igPm8QQjgT\nOBU4IUmSu/OYm/Jg/HiYP79ivNJKsHhdh6WpbrffDrNnV4y7dYNDDql+3/z5cZ/1aqvB+efH880e\nfjj9SUerVnDCCQ2esiRJklSsSrmoHgVsEkJYZUEghNAN2Lz8Wq1CCCcAFwJnJklybQPlqEVQ9Tgt\nl37nwfz5cNNN6dgxx8Add8Cbb8LWW1d/Zto0OO+8eGTWKaekr/XvD8st10DJSpIkScWvlIvqW4Av\ngZEhhD4hhN2BkcDXwP+qhhBC1xDC3BDCOZVi+wNXAk8Bz4cQNqn0yrlzuBqG+6kbwLNfm0yoAAAg\nAElEQVTPwmefVYxbtYJDD43f/9//wfPPw6OPxgPBq5o0CSZMqBiHACef3LD5SpIkSUWuZIvqJEl+\nBbYFPgHuBu4FvgC2TZJkWqVbA9Cc9L/rTuXxnYBXq7yub/DkVS8ep7WInngizjwfdBC89VaMVW1Q\nts8+ULnpXgjQp09sUnbVVbDkkgt//z59oHv3vKctSZIklZKQJEnWOZSkXr16JePGjcs6jUatZ094\n++2K8dixsNVW2eVTUv77X/jd79JnSe+7Lzz0EMybVxF76SXYfPOFv8+UKXDllbE52ZQp6Wsvvwyb\nbZbfvCVJkqQiEUJ4M0mSXnXdV7Iz1Wrc5s2Djz9Ox1z+nYO77koX1BA7flcuqNddt+6iuH17OOcc\n+OILOP10aNcOmjWDv/zFglqSJEkCWmSdgFSTr76KR2otsOSS6VXKqkWSxA7fdTn66Ljcuz46dYKL\nL4bBg2Pjsg4dFi1HSZIkqZFwplpFqck3KUsSOPts6NUrNgObOLH+z776ajyPrDaLLw4HH5x7Xi1a\nWFBLkiRJlVhUqyg1+eO07r4bLrwwHnN12WWxIdjNN6eXby9M1VnqPfaAW29NH311zjlxabckSZKk\nRWJRraLU5Geq77gjPZ44EQYOhE02gddeW/hz06fD/fenY4cfHl/jx8OoUbE5WdXzpiVJkiT9JhbV\nKkpN+jitH3+Mrc5rMm4cbLopHHggfPll9esPPwxTp1aMO3eGnXaK3y++OOy2W+z2Xd+91JIkSZJq\nZVGtopMkTXym+tFHYf782u8ZPjx+0nDqqTB5ckW86tLvfv3iPmhJkiRJDcKiWkXnhx/g558rxost\nBiutlF0+BTdiRHp87LHxjOmqZs2CSy+FVVeFYcPiRvTnn0/fM2BAg6UpSZIkyaJaRajqLHX37vFo\n5CZh0qTqhfHxx8d90s8/Dz17Vn9m4sS4R7pHj3R8o41grbUaLldJkiRJFtUqPk166feoUTB3bsV4\nrbUqNpRvs03cU3377dClS/Vn58xJj52lliRJkhqcRbWKTpM+Tqvq0u+9906PmzePxfInn8DgwdCu\nXc3v07o17L9/g6QoSZIkqYJFtYpOk+38PXkyPPNMOta3b833Lr44nH02fPFFXPpdVpa+vs8+0LFj\nw+QpSZIk6X8sqlV0Pv00Pe7ePZs8Cu4f/4DZsyvGq60G665b+zNLLhmblX32WWxo1rUr7LgjXH55\nw+YqSZIkCQDP2lFRmTMHJkxIx1ZdNZtcCu6hh9Ljvfeu/3nSXbrAddfFlyRJkqSCcaZaReWrr2De\nvIrxcsvFI7UavWnT4Mkn07GFLf2WJEmSVDQsqlVUqi79bjKz1E88ATNnVoy7doUNNsguH0mSJEn1\nYlGtovLZZ+lxkymqH300Pe7bt/5LvyVJkiRlxqJaRaVJFtVz5sSZ6sr23DObXCRJkiTlxKJaRaVJ\nFtUvvhiP01pgqaVg002zy0eSJElSvVlUq6g0yaJ61Kj0eNddoXnzbHKRJEmSlBOLahWNJIHPP0/H\nGn1RnSTVi+rdd88mF0mSJEk5s6hW0fj2W5gxo2Lcvj0suWR2+RTEBx/AF19UjFu3hh12yC4fSZIk\nSTmxqFbRqGnpd6NvgF11lnrbbaFt22xykSRJkpSzFlknIC3gfmpc+i1JTdysWbOYNGkSU6dOZd68\neVmnI0mNRvPmzWnXrh2dOnWidevWeX1vi2oVjSZXVH//Pbz+ejq2667Z5CJJytysWbOYMGECHTt2\npFu3brRs2ZLQ6JdsSVLDS5KEOXPmMGXKFCZMmMBKK62U18La5d8qGk2uqH788diobIENNoAVVsgu\nH0lSpiZNmkTHjh1ZaqmlaNWqlQW1JOVJCIFWrVqx1FJL0bFjRyZNmpTX97eoVtFockV11aXfu+2W\nTR6SpKIwdepU2rdvn3UaktSotW/fnqlTp+b1PS2qVTSaVFE9YwY8/XQ65n5qSWrS5s2bR8uWLbNO\nQ5IatZYtW+a9Z4VFtfLuySdhnXVg003hrbfq98zkyTBxYsW4ZctGvhL6uefS54etsAL07JldPpKk\nouCSb0lqWA3xc9ZGZcqrOXPgkEPgxx/jePvt4dVXoXv32p+rOku98srQvHnD5FgUHnssPd599yZw\nfpgkSZLU+DhTrbyaMKGioAb4+WfYZRf44Yfan2tSS7+TJE7nV+Z+akmSJKkkWVQrr77/vnrs88/j\nROz06Qt/rkkV1R9+CF9/XTEuK4PevbPLR5IkFdRjjz3GxhtvzBJLLEEIgYMPPjjrlBqlvffemxAC\nP/30U9apqJGzqFZeLWxG+vXX4eCDYWE9AZpUUf3UU+nxNttAmzbZ5CJJUpEIIeT0uuOOO7JO+Tf5\n8MMP6du3L9988w1HHnkk5557LnvttVfB8zj55JMJITBs2LCC/7MVvfTSSwwYMIDVVluNtm3bUlZW\nRrdu3dhvv/145JFHSCofvaqi5p5q5VVty7wfeQSOPx6uvRaaVfk4p0kX1TvtlE0ekiQVkXPPPbda\n7Morr2Ty5MmceOKJdOjQIXWtZ4k2+Bw9ejRz5szhuuuuY3dP/mhQV199NRdeeCEdO3bMOpWUmTNn\ncvTRR3PnnXfSsmVLtt56a/r06UOrVq2YMGECzz77LA888ACHHHJIyX541NRYVCuv6to7fcMNcRn4\nrbdCi0p/+oqyqP78c2jfHpZaKn/v+euvMHZsOmZRLUkS5513XrXYHXfcweTJkxk0aBDdunUreE4N\n4ZtvvgGgS5cuGWfS+HXp0qUo/3c+7LDDGD58OL169eL+++9nlVVWSV2fO3cud955J2PGjMkmQeXM\n5d/Kq6p7qgcNguWWS8fuvBP22w9mzYrjWbPSW4whdv/O1KBBsbJfZpm4IXzs2NhgbFGNGQOzZ1eM\nV14ZVl990d9XkqQmqlevXrRt25YZM2Zw1llnsdpqq9GqVSv++Mc/AjBx4kQuueQSevfuTZcuXWjV\nqhXLLrssffv25c0336z2ftOmTSOEwK677sp3333HgAEDWGaZZSgrK6NHjx4MHz682jPz58/nlltu\nYeONN2appZaiTZs2rLTSSuyyyy48+uijAPzjH/8ghMBf//pXADbccMP/LWUfN27c/97rxx9/5OST\nT6Z79+6UlZXRsWNHdtxxxxoLrGuvvZYQAiNGjGDkyJFsueWWtG/fnnbt2uXjf9qU2bNnc9VVV7Hh\nhhvSrl07FltsMXr16sUtt9xS7d4kSbj55pvp06cPK6+8Mm3atKFDhw707t2bBx98sMb3r+v/x8r/\nrk899RRbbLEFbdu2ZYkllmCPPfbg008/rfaeNe2pfu+99wgh8Mc//pFPPvmEvn370qlTJ9q0acMm\nm2zCM888U2N+kyZN4thjj6VLly6UlZWx9tprc91116Xerz5Gjx7N8OHD6dy5M0899VS1ghqgRYsW\nHH744dx6663/i1X+96+q8p/ZyhYs8R83bhy33347vXr1YvHFF2edddbh2WefJYRAv379aswzSRK6\ndu1K27ZtmTZtWuraqFGj+P3vf0+nTp1o3bo1q6++OmeccUa1+5oSZ6qVV/+/vfsOr6LK/zj+/qbT\nu/QuAmJDQKSoNAVBsaAiulKFdXexLLpW+AHKqmtZRXRXV7pYEJHqotIFFxQQRYM0ARVEaaF3cn5/\nzL0xtyTchCQ3hM/reeZJ7pkzc8/MZHLne08Lrqlu1Aj69IGrr4Zff/09/cMPvVj1ww9hy5bAeLVy\n5Sh3Md68GYYP9353zpv+asYM72AefBC6dIGEhOztO7jp97XXaiotERGR05Samsp1113H2rVrad++\nPWXKlKF69eoArFy5ksGDB6c1sS1RogSbNm1i+vTpzJw5k9mzZ3PllVeG7HPHjh1cfvnllCpVim7d\nunHw4EEmTpzIHXfcQUJCAl26dEnL+8ADDzBixAjq1KlDt27dKFq0KL/88gtffPEFU6dO5cYbb+S8\n885j8ODBfPrppyxZsoS+ffum1aL6f65bt442bdqwdetWWrduTadOndi3bx/Tp0+nbdu2vPXWW9xx\nxx0hZR03bhyzZs3iuuuu409/+hPbtm3L0fN7+PBhOnTowGeffUaDBg246667iI+PZ86cOfTr148V\nK1bw+uuvp+U/efIkf/zjH7nsssto3bo15cuXZ8eOHcycOZPbbruNZ599lkceeSTkfTK7jn4TJ05k\n6tSpacf6zTffMG3aNJYvX87q1aspXrx4RMe0bt06mjZtSoMGDejZsyfbt2/n/fffp2PHjixevJim\nTZum5d2/fz9XXnklycnJNGnShO7du7Nr1y4ef/xx2rRpk6Vz+Z///AeA/v37U6ZMmUzzJiYmZmnf\nGRk6dCjz5s3j+uuvp127dhw5coQ2bdpQtWpVpkyZwoEDByhatGjANvPnz+enn36iR48eAesefvhh\nnn/+ec455xw6d+5MuXLlWLFiBc888wyffPIJixYtonDhwjlS7jOKc05LNpZGjRo5CdWqlXNeJOot\nn37qpa9f71z16oHrwLlGjZz7978D0668MqqH4Nzbb4cWNP1Svrxzjz/u3ObNWd/3uecG7mv69Jwv\nv4iInJFWr14dfkVmn0n5cclB1atXd4DbtGlThnkaNWrkANekSROXkpISsn7Xrl1u9+7dIekbNmxw\nZcqUcY0bNw5I379/vwMc4O677z538uTJtHXLli1zZuaaNGmSlnby5EmXlJTkateu7Y4cORLyPjt2\n7Ah4/eCDDzrALVu2LOyxxMbGumnTpgWk79y509WtW9cVK1Ys4BhHjBjhABcbG+sWLFgQsr/M+Mvx\n/PPPR5z3kUceCTgfx48fd7fffrsD3Jw5c9LSU1NT3Q8//BCyn0OHDrlmzZq5pKQkt3PnzoB1p7qO\n/mNNSEhwn3/+ecC6/v37O8C99tprAeldunRxQMA1+Pbbb9Oub/Cxf/DBBw5wt956a0D6ww8/7ADX\np0+fgPQNGza44sWLO8D95S9/CSlzOOecc44D3JIlSyLK7+c//kmTJoWs8//NdurUKSDdf91KlCjh\nkpOTQ7Z7/PHHHeDGjBkTsq579+4OcPPmzUtLmzFjhgNcmzZt3L59+8KWb+DAgVk6rmjJ8P9tEGC5\niyA2VPNvyVHBNdXnnOP9PPdcWLQI6tYNXL9iBfzpT4FpUe9P/eWXma//7Td4+mmoVcubX3rKlMAm\n3RnZsMFb/OLjvZG/RURE5LQ988wzIYOZAZQuXTrsQFW1a9emc+fOLF++nN27d4esL1WqFM8++ywx\n6UZXbdy4MQ0bNmTlypWcOHEiLd3MSEhIIDY2NmQ/ZSMcm+Xzzz9nxYoV3HXXXSEDmJUpU4ZBgwax\nf/9+pk+fHrJtt27duCqXpuc8duwYr7/+OrVq1eLpp58OOB9xcXFpo4e//fbbaelmFrZZc6FChbjn\nnns4cuQIC4PHmPHJ6Dr69erVi+bNmwek9evXD4AvT/UMl079+vV58MEHA9K6dOlCmTJlQvYzbtw4\nEhISGDZsWEB67dq1+fOf/xzxe548eZIdO3YAUKVKlYi3O1333nsv559/fkh6jx49AO/40jtw4ACT\nJ0+mevXqtGrVKi19uK8l56hRo0K6GPTv359zzz034O/gbKLm35KjgvtU+4NqgKpV4bPPoH17+Prr\njPcR9aD6iy8CX190EaxaFZovNRVmzvSWsmXhzjuhRw+45JLwTbqDm35fcQUENbURERGR7Lnssssy\nXDd//nxGjBjBl19+yfbt2zl+/HjA+q1bt1K6dOmAtPPPP59CYfqjVa1ala+++or9+/dTqlQpYmJi\nuP322xkzZgwXXHABt956K1dccQXNmjXLUt/mJUuWAF6z83CDtm3duhXwpuQKltmxn65Vq1Zx8OBB\nChUqxJNPPhk2T1xcXEi5fvjhB5577jnmz5/Pli1bOHz4cMB6//EEO9WxNG7cOCStatWqAKSkpGS6\nbXqXXnopFuZ5rUqVKmzatCnt9bZt2/jtt9+oX78+FSpUCMnfsmXLiN8zWjI6p+eddx7NmjVj4cKF\n/Pjjj2lN7SdPnszBgwcZMGBAwDlasmQJRYoUyXRE8k2bNnH06NEca7p+plBQLTnmxAnYtSswLfjL\n2XPO8cbq6tEDpk0Lv5+oBtXHjsHKlYFps2bB/v0wYgSMH+/9HmznTq8f9vDhcN55cOutcNttcOGF\nvwfY4fpTi4iIyGkrXLhwhgHshAkT6N69O0WLFuXqq6+mZs2aFClSBDNL69981D96ajoZ1ZbG+aYv\nOXnyZFraG2+8Qb169Rg3blxabWZ8fDydO3fmxRdfDOkXHM4u30PURx99xEcffZRhvnCDQYUL9nKK\nv1zJyckkJydHVK7Vq1fTvHlzDhw4QKtWrbj22mspXrw4sbGxrFu3jnfffTfsOc/sOvqFuy7hrsmp\nZHZ90+9n7969AJQvXz5s/ozSw4mNjaVcuXJs376drVu35lltdWZ/Hz179mTJkiWMHz+eQYMGAb/X\nXHfv3j0t39GjRzl48CDg9dHOzIEDB866oFrNvyXHpBtYEYAyZbwWzsFKlPAGKPv738NX6EY1qF61\n6vdhyQGqVIFKlbx266++Cr/8Am+84dVGZ2TdOu/gLr4Y6tWDgQNh2TKYPz8wn6bSEhGRSES/l3TW\nligIV+PoN3DgQIoVK8bXX3/N5MmTeeGFFxg6dChDhgyhZg5NNxIfH8/DDz9McnIy27Zt4/333+fa\na69l8uTJdOrUKaJgr0SJEoDXtDazvpsjRowI2Taz4z9d/nLdddddmZbr22+/TdvmueeeY+/evbz3\n3nvMmTOH4cOH89RTTzFkyBBaZ9L1LTePI7v8A5/9Ftwc0yej9Iz4a7bnzp2bpe38ze7Tdzvw27Nn\nT6bbZnZeu3btSlJSEuPHjwfgxx9/ZMGCBbRo0YJzzz03LV9iYiKJiYlUrVr1lP2LTzUAW0GkoFpy\nTEb9qcOJiYHHH/cqgdO3tipb1qvcjZrgpt/BzWWKFoV+/eCrr7y+13/+M2TS7yctwL7sMm+Cbr/K\nlaFBg5wrt4iIiIQ4ceIEP/74I5dccklIH9/jx4+nNbnOSRUqVODWW29l2rRpXHbZZSQnJ4ed7inY\n5ZdfDsCiRYtyvEyn4+KLL6ZQoUJ8/vnnpKamRrTNhg0biImJ4cYbbwxZl1Ff6vyqUqVKnHPOOfzw\nww/8mn4qG5/FixdnaX/+/t+vvvpq2L786aWvzfePC/Bz8Dy0EDAlW1aln5Ls888/56233sI5l9bf\nOr3LL7+cn3/+mc2bN2f7/QoqBdWSY7ISVPu1b+8NVtarlzfm14wZkJSUO+WLSPAAF+mmUwhgBk2a\nwGuvwbZtMHEidOwIcRH2qOjQQVNpiYiI5LK4uDgqV65McnJywFzFqampPPbYYwF9Z7PrwIEDYQfI\nOnr0aFrT4XB9s4NdddVVXHrppUyYMCHsXNgAX331VZb6DecE/+BiGzdu5KGHHuJYmMFZf/75Z9au\nXZv2ukaNGqSmpoZ8QTBlyhTeeeedXC9zTuvevTvHjh1j4MCBAekbN27kX//6V5b21b59e7p168a2\nbdvo2LFj2AD15MmTjBs3jr59+6alNWnSBIC33norINjevn07jz/+eJbKEKxnz56A1+z7rbfeolCh\nQnTt2jUk34ABAwDo3bs324Mf/PGmHsvKYHEFifpUS47JbJCyzNSoAaNH53hxsie4pjqjoDq9pCSv\n//Rtt0FKitdZfNIkmD0bggZCSaP+1CIiInnir3/9Kw899BAXXXQRN998MzExMSxcuJDNmzdz7bXX\nMmvWrNPa/549e2jatCn16tWjYcOGVKtWjUOHDvHxxx+zfv167rjjDqpVq3bK/ZgZkyZNom3bttxx\nxx28+OKLNGnShGLFirFlyxZWrlzJmjVr+Pbbb8OOZp5dEydO5Lvvvgu7rnPnztx88808/fTTfPfd\nd7z00kt8+OGHtGrViooVK/Lrr7+ybt06lixZwiuvvEJd3zQv9957LxMnTqRTp07ceuutnHPOOXzz\nzTfMmTOHW2+9lffffz/Hyp8XBg0axMyZMxk1ahTffvstbdq0YdeuXbz//vu0adOGqVOnBoyKfiqj\nR48mPj6e8ePHc95559G6dWsuvPBC4uPj2bJlC3PnzmXbtm1pwS7Aueeey0033cSUKVNo2LAhHTp0\nICUlhZkzZ9KmTZuwA9hFql27dlSqVImxY8dy/PhxunXrFna+786dOzNw4ECGDRvGueeeS4cOHahZ\nsyb79u1j8+bNLFy4kI4dO/LBBx9kuyxnKgXVkmOyU1Odr+zZA+m+ZSUmBho1yto+SpWCnj29JSUF\npk/3AuxPP/09wK5Z06vVFhERkVw3YMAAihYtyquvvsro0aMpUqQIrVq14v333+fNN9887aC6TJky\nPP3008yfP59FixaxY8cOihcvTp06dXjkkUfCNqPNSK1atVi5ciXDhw9nypQpjB8/HuccFStWpEGD\nBvztb38L6OeaE5YvX55h8+EqVapw8803k5SUxKxZsxg3bhzjx49n2rRpHDx4kHPOOYfatWvzzDPP\n0KVLl7TtmjZtyuzZs/m///s/pk2bRmpqKg0bNuSjjz7i5MmTZ1xQXbx4cRYtWsSgQYOYMmUKL730\nErVq1WLYsGE0aNCAqVOnhg1CM5KUlJRWEz1y5EgWL17M4sWLOXHiBBUqVKBFixbccccdIc3nJ0yY\nwGOPPcakSZN47bXXqFGjBo899hh33333aZ3T2NhY7rrrLv7xj38AZPo3+9RTT9GmTRtGjBjBZ599\nxtSpUylZsiRVqlThL3/5C3feeWe2y3EmMxelASXOdI0bN3an03+hIHrsMXj22d9fP/kk+AYRPDPM\nng3XXPP76wsvDD+VVnbs2eN1IN+1C7p2hXLlcma/IiJSYHz//ffUr18/2sUQkSx46aWXGDBgABMm\nTDhrA8ozUaT/b81shXMudB63IKqplhxzxtdUB/cByck5H0uWhG7dcm5/IiIiIpJnfvnlFypVqhSQ\n9sMPP/Dss8+SlJREB83qclZTUC05Jrt9qvON7PSnFhEREZECr3379hQpUoRLLrmEEiVKsHHjRmbO\nnMmRI0d45ZVXzspppOR3Cqolx5zRNdXOKagWERERkbB69erFe++9x6RJk9i3bx/FihXjiiuu4P77\n76dTp07RLp5EmYJqyTHBQXX58tEpB87BP/8JU6dC7dreqNzt2kFCQsbb/PRT4AEULgznn5/7ZRUR\nERGRfG/AgAFpU0qJBFNQLTnCuXxUUz1nDjz0kPf74sUwbpw3KvdNN8Ett0Dr1qGTYQfXUjduHPmc\n0yIiIiIictaKfEI1kUwcOACHD//+OjERihWLUmE+/TQ0LSXFmwy7Y0coWxa6dIGxY3//JiA3BykT\nEREREZECS1VxkiPC1VKbRacsrF6d+fqDB+HDD70FoGFD2LEjMI/6U4uIiIiISAQUVEuOyDf9qSE0\nqC5WDPbvzzj/ypWhaaqpFhERERGRCKj5t+SIfNOf+uBB2Lz599cxMbB1K3zyCfTrB0HzC4ZVoQJU\nrZprRRQRERERkYJDNdWSI/LNHNVr1gS+rlXLq6m+5hpvcQ6++gqmT4ePPvJ+dy5wm2uvjWLbdRER\nEREROZMoqJYckW9qqoObfgdPi2UGjRp5y9ChsHMnzJ3rDW62YoU3Bdff/5535RURERERkTOagmrJ\nEWdMUB2sbFno2tVbREREREREskh9qiVH5JuByrIaVIuIiIiIiJwGBdWSI/JNn2oF1SIiIiIBFixY\ngJkxZMiQ09rP2LFjMTPGjh2bI+UqCIYMGYKZsWDBgmgXRaJIQbXkiHzR/PvwYdi4MTCtXr0oFERE\nRESyy8wCltjYWEqXLk2rVq0YO3YsLniA0XzIX/aYmBh++OGHDPO1bt06LW9BDFT9QXjPnj2jXZSz\n1po1a7j33nu54IILKFGiBAkJCVSqVIlOnToxatQojh49Gu0iFgjqUy05Il8E1evWQWrq769r1IAi\nRaJQEBERETldgwcPBuD48eNs2LCBKVOmsHDhQpYvX86rr74a5dKdWlxcHCdOnGDUqFE8/fTTIevX\nr1/PggUL0vLJmal///7cfvvtVKtWLdpFCfHkk08ydOhQUlNTadasGT169KBo0aL89ttvLFiwgLvv\nvpt///vfLF++PNpFPeMpqJbTduIE7NoVmFauXBQKoqbfIiIiBUZwU+XPP/+cK6+8kn/96188+OCD\n1KxZMzoFi1D58uWpWLEiY8aM4cknnyQuLvCxe+TIkQBcf/31TJkyJRpFlBxQtmxZypYtG+1ihHj6\n6acZPHgwVatWZdKkSTRt2jQkz8yZM3nxxRejULqCR82/5bTt2hU41XPp0hAfH4WCKKgWEREpsFq0\naEG9evVwzrFixYqAdceOHePVV1+lY8eOVK9encTEREqXLk27du2YNWtWyL6aNWtGQkICBw8eDEi/\n6qqrMDP69OkTkP79999jZnTv3j1LZe7bty+//vorM2fODEg/fvw4Y8eOpXnz5pyfyfPK+vXr6d69\nO5UrV05rttu9e3fWr18fNv9vv/1Gnz59KF++PIUKFeKSSy5h3LhxmZZx9+7dPPbYY9SvX59ChQpR\nokQJ2rZty6effpqlY80p7777Lq1bt6ZkyZIkJSVRv359hg0bFraZ8tSpU/nDH/7AeeedR5EiRShS\npAiNGjXilVdeITV960Wfnj17YmZs3LiRESNGcNFFF1GoUCFatWoFBPY9//rrr+nUqRMlS5akcOHC\nXHXVVfzvf/8L2WdGfarNjFatWrFz50769etHxYoVSUxMpEGDBowZMybssR89epQhQ4ZQq1YtEhMT\nqVmzJgMHDuTo0aNp+4vE5s2bGTJkCPHx8fz3v/8NG1ADXHfddXz88cdpr0/V975GjRrUqFEjIC19\nP/uPP/6YVq1aUaJECcyMrVu3EhsbS8OGDTMs67XXXouZ8d133wWkf/HFF9xyywYYWhAAACAASURB\nVC1UqFCBhIQEqlatyh//+Ed++eWXiM5BXlNNtZw2DVImIiIieSk+6Nv73bt3c//999O8eXOuvvpq\nypUrx7Zt25gxYwYdO3bkzTff5O67707L37ZtW5YuXcqiRYvo0KEDAIcOHWLp0qUAzJ07N2D//tdt\n27bNUjm7devGgAEDGDlyJDfeeGNa+vTp09m+fTv/+Mc/2LBhQ9htly1bRrt27di/fz+dO3fm/PPP\nZ82aNUyYMIFp06YxZ84cmjRpkpZ/586dNG/enI0bN9KyZUtatmzJtm3buOeee7jmmmvCvsePP/5I\nq1at2Lx5M1dccQUdOnTg4MGDzJw5kw4dOvDGG2/Qt2/fLB3z6ejduzdjxoyhSpUqdOnShZIlS7J0\n6VIGDRrE3LlzmT17dkCN/6OPPkpMTAxNmzalcuXK7N27l3nz5nH//fezbNky3nrrrbDvc//997No\n0SI6depEx44diY2NDVi/fPlynnvuOZo1a8bdd9/NTz/9xOTJk2nbti1ff/01devWjeh49uzZQ4sW\nLUhISOCWW27h6NGjTJo0id69exMTE0OPHj3S8jrn6NKlCx999BF16tShf//+aV++JCcnZ+k8jhkz\nhuPHj3P77bdzwQUXZJo3MTExS/vOyAcffMDHH3/Mtddeyz333MOPP/5I5cqVadeuHZ9++inffvst\nF154YcA227ZtY/bs2TRq1CignKNHj6Zfv34kJibSuXNnqlatyvr16xk5ciQzZsxg6dKl+a+5vXNO\nSzaWRo0aOfHMnu2cV1ftLVdeGaWC1KsXWJClS6NUEBERkaxbvXp12PT0H21nwnK6AEeYHS1cuNDF\nxMS4hIQE98svvwSsO3LkiPv5559DttmzZ49r0KCBK1WqlDt06FBa+ty5cx3gHnroobS0jz/+2AHu\n6quvdoDbsGFD2robb7zRAe6nn36K+BgqV67snHOuT58+LjY2NqB87du3d8WLF3cHDx50TzzxhAPc\nmDFj0tanpqa6evXqOcBNmDAhYN/vvfeeA1zdunXdyZMn09L79u3rAPfAAw8E5F+2bJmLi4tzgBs8\neHDAuquuusqZmXv33XcD0lNSUtzFF1/skpKS3K+//pqWPmbMmJCyZsafv0ePHhHnvemmmwKulXPO\nDR482AHu5ZdfDkhPf438Tp486bp37+4AtzToWbBHjx4OcJUqVXIbN24M2Xb+/Plpf3/Bx/j66687\nwP3pT38KW7b58+cHpPv306dPH3fixIm09OTkZBcbG+vq168fkH/8+PEOcFdccYU7evRoWnpKSoqr\nW7euA9xVV10VUuZw2rRp4wD35ptvRpTfz3/8wX8nftWrV3fVq1cPSPNfNzNzs2bNCtnmnXfecYB7\n8MEHQ9Y999xzDnCvvPJKWtratWtdfHy8q127ttuyZUtA/jlz5riYmBh34403Zum4wsno/20wYLmL\nIDZU8285bflijupjxyC4KVT9+lEoiIiIiOSEIUOGMGTIEJ544gm6du1Ku3btcM7xwgsvULFixYC8\niYmJVKlSJWQfJUqUoHfv3qSkpLBs2bK09ObNm5OUlBRQIz137lzi4uIYOnRo2muA1NRUFixYQJ06\ndahatWqWj6Nv376cPHmS0aNHA17t8OzZs7nzzjspXLhw2G3+97//sWbNGpo1a8add94ZsK5r1660\nbNmStWvXsnjxYsBrTv72229TrFixkKa7jRs3DtkHwDfffMPChQvp0qULt99+e8C6kiVLMnToUI4c\nOcLkyZOzfMzZMXz4cOLi4hg9ejSFChUKWDdo0CDKlCnD22+/HZBeu3btkP3ExMRw//33A/DJJ5+E\nfa+HH3440z75LVq0CBmxvHfv3sTFxfHll19GcjgAFC5cmH/+858BNeHnn38+LVq04Pvvv+fAgQNp\n6f5m+sOGDSMhISEtvWTJkgwaNCji9wSvBhgIe0/klhtuuCGt1Ud6N954IyVKlODtt9/m5MmTAevG\njRtHfHw83bp1S0v797//zfHjxxk+fDiVK1cOyN+2bVs6d+7MjBkz2L9/f+4cSDap+bectnwx8vf6\n9ZD+Rq1SBYoXj0JBREREJCf4g1s/M2PUqFH06tUrbP7k5GSef/55PvvsM7Zt28aRI0cC1m/dujXt\n96SkJJo3b878+fPZtWsXZcqUYd68eTRp0oRmzZpRvnx55s6dS79+/fjqq6/Ys2cPXbt2zdZxNG3a\nlAsvvJDRo0czcOBARo4cSWpqaqbNqr/66isA2rRpE3Z9mzZtWLx4MStXruTKK69kzZo1HDp0iCuu\nuIISJUqE5G/VqlVI3+olS5YAsHfv3rB9aHfs2AF4/clz26FDh/jmm28oW7YsL7/8ctg8iYmJIWXZ\ntWsXzz//PP/973/ZuHFjSB/59Nc8vcsuuyzT8jRu3DgkLT4+nvLly5OSkpLptunVqVOH4mGeR/1f\nzqSkpFC0aFEAVq5cSUxMDM2bNw/J37Jly4jfM1oyOqeFChXitttu48033+STTz6hY8eOAKxYsYLk\n5GRuuummgIHe/H+XCxcuDPgizG/79u2cPHmSdevW0ahRo1w4kuxRUC2nLV/0qVZ/ahERkQLFa3kJ\nBw8eZMmSJfTp04d77rmH6tWrhwSbS5cupU2bNpw4cSKtNqt48eLExMTw9ddfM23atJCBrtq2bcu8\nefOYP38+bdu2ZeXKlTz++OOAF7TOmTMH51y2+1On17dvX+677z5mzZrFmDFjaNSoUaaDN+3duxcg\npEbez5++Z8+egPzlM2guWKFChZC0Xb6pW2bPns3s2bMzLEv62tTckpKSgnOOHTt2hHyZkpE9e/bQ\npEkTNm3axGWXXUb37t0pXbo0cXFx7Nmzh+HDh2c4B3O485FeyZIlw6bHxcWF1LZmdz9AwL727t2b\nVv5gGV3XjFSsWJHvv/8+wy8VckNm57Rnz568+eabjBs3Li2o9n/Jk75fOfz+d/n8889n+n558XeZ\nFQqq5bTli5rq4KC6QYMoFEJERCTnOXfqPAVZkSJFaNeuHTNmzODSSy+lR48erF27NqDp9LBhwzh8\n+DDz588PGSH5mWeeYdq0aSH79Qfmc+bMISYmhtTU1LTAuU2bNrz77rt88803zJ07FzOjdevW2T6G\nu+66i0ceeYR77rmHrVu38n//93+Z5vfXNv/6669h1/ub9/rz+X/+FlzT4RNuP/5thg8fzn333RfB\nUeQef1kaNmyYVkt/KiNHjmTTpk0MHjw4pKZ9yZIlDB8+PMNtzSzbZc0txYsXZ/fu3Zw4cSIksM7o\numakZcuWzJs3j7lz54aMZJ+ZmBivZ3BG86bv2bMnwy8KMjunzZs3p06dOkyfPp09e/ZQpEgR3n33\nXcqWLZsWZPv5/xb27t0btpY/v1Kfajlt+aJPtWqqRURECrSLLrqIvn37smXLFl566aWAdRs2bKB0\n6dJhpxxauHBh2P01adKE4sWLM3fuXObNm0ehQoVo1qwZ8Hut9H//+18+//xzLrrootOai7hkyZLc\ncsstbNmyhSJFigT0IQ3HX4sdPE2T3/z58wG49NJLAahXrx6FCxfm66+/Tqu1Ti/cfi6//HIAFi1a\nFOlh5JqiRYvSoEEDkpOT2b17d0Tb+EdN79KlS8i6jK55ftawYUNSU1PDTtvl7zsfqV69ehEfH8/k\nyZNZHfyMHCR9bX6pUqUA+Pnnn0PybdiwIezfVqR69OjBkSNHmDhxIh999BE7d+7kjjvuCBnJPz/9\nXWbFGR1Um1lVM/vAzPaa2T4z+9DMIhpf3cySzOx5M9tmZofNbImZXZnbZS6I8mVNtYJqERGRAmfg\nwIEkJibywgsvBPRtrVGjBrt372bVqlUB+UeNGpXhYFWxsbFceeWVbNiwgUmTJtGyZcu06YVq1qxJ\njRo1GD58OIcOHcqwb3NWDBs2jClTpvDJJ59QrFixTPO2aNGCunXrsnjxYj744IOAdR988AGLFi3i\nvPPOS+trGx8fz5133sn+/ftDam2XL18eMsAXeP2Gr7jiCj788MO0QdSCffvtt2wPftDLJQMGDODY\nsWP07t07rVl7eikpKQG12P75koO/MFi5ciXPPPNMbhY1V/jnQB84cCDHjh1LS9+7dy9PPfVUlvZV\no0YNhgwZwrFjx+jUqRPLly8Pm88/BZZfvXr1KF68ONOmTQu47ocPHz7t1gzdu3cnJiaG8ePHM378\neICQweAA+vfvT3x8PH/9619Zt25dyPpjx47ly4D7jG3+bWaFgXnAUaAH3rD1w4D5ZnaRc+5gZtsD\no4BOwN+AjcBfgE/MrJlz7uvcK3nBE/U+1SdOwNq1gWka+VtERKTAqVy5Mvfccw/Dhw/nueeeSwue\nHnjgAT755BNatmzJbbfdRokSJVi+fDmLFy/mlltuCQlM/dq2bcvMmTPZvn17SJ/ptm3bMmrUqLTf\nT1e1atUinlvXzBg3bhxXX301Xbt25YYbbqBevXqsXbuWqVOnUqxYMcaPH5/WXBfg6aefZu7cubz8\n8sssX748bZ7qiRMn0rFjR6ZPnx7yPu+88w5t2rShT58+vPLKKzRt2pSSJUuyZcsWVq1axXfffceS\nJUs45zQf7hYvXhw2gAKvtv2+++6jd+/erFixgn/961/Url2b9u3bU61aNXbv3s2mTZv47LPP6NWr\nF6+//jrgBWnPP/88DzzwAPPnz6dOnTqsX7+emTNncvPNNzNx4sTTKnNe6969O++99x4ff/wxF1xw\nAZ07d+b48eNMnjyZJk2asHbt2oDrfSqPP/44J06cYOjQoTRp0oTmzZvTuHFjihYtym+//cZnn33G\n+vXrAwZli4+P5/777+epp56iYcOG3HTTTZw4cYLZs2dTqVIlKlWqlO3jq1q1Kq1bt04bZf/CCy8M\nO65AvXr1GD16NL1796ZBgwZ06NCB8847j+PHj/PTTz+xaNEiypUrx5o1a7JdllwRybxb+XEB7gdO\nAuemS6sJnAAGnGLbi/GC8F7p0uKAtcD0SN5f81R7UlOdK1QocH7KPXvyuBBr1gQWoGLFPC6AiIjI\n6Yt03tSCjgzmqfb79ddfXeHChV3hwoUD5lCeMWOGa9q0qStatKgrUaKEu/rqq93ChQsznVt51apV\nae/35ZdfBqzzz68bFxfn9u3bl+Vj8M9TfSrh5qn2W7NmjfvDH/7gKlSo4OLi4lyFChXcnXfe6das\nWRN2X9u2bXO9evVyZcuWdUlJSe7iiy92Y8aMyXT+4X379rm///3v7tJLL3VFihRxSUlJrkaNGq5j\nx47ujTfecAcOHEjLm915qjNbbrjhhoBtZsyY4Tp16uTKlSvn4uPjXfny5V2TJk3cE0884b7//vuA\nvMnJye7666935cqVc4ULF3aXXnqpe/PNN92mTZvCzo/tn6d606ZNYcubnXmaM5unOqN5pTMqx+HD\nh92gQYNcjRo1XEJCgqtevbp7/PHH3ZYtW8Keq0isXr3a9e/f3zVo0MAVK1bMxcfHuwoVKrgOHTq4\nkSNHuiNHjgTkT01Ndc8884yrVauWi4+Pd1WrVnV/+9vf3MGDBzOdpzqSv4m33nor7bq/8MILmeZd\ntWqV69Gjh6tWrZpLSEhwpUqVcg0aNHD9+vVzc+fOzeppCJHT81SbO0NHvzCzuUCSc65FUPpCAOfc\nVZlsOwgYBJR0zh1Klz4UeBQo7pwLP1ygT+PGjV1GTSmi6fryXzJ7+0V59n4O4xiJaa8TOMqRhBLk\n6fgPqalw/Pjvr9u2hTlz8rAAIiIip+/777+nvlpaiUiQ2bNnc8011/Doo4+ekU3b86NI/9+a2Qrn\nXOgca0HO2ObfQAMgdChHSAZujWDbTekD6nTbJgDn+n4/4xw/GcNRkqL2/uX5DTuW6fcRuU/9qUVE\nRETkDPPLL7+ENLHetWsXjz76KAA33XRTNIolETiTg+rSQLjZ13cDpU5jW//6EGbWD+gHRNwn5mxz\nPpmPMJgnMph8XkREREQkvxowYADffPMNzZs3p1y5cmzZsoVZs2axe/du/vjHP3KZnnHzrTM5qM5z\nzrn/AP8Br/l3lIuT75zHWp7l0egVIC4ObroJbrklemUQEREREcmGm2++md9++40ZM2awZ88ekpKS\naNCgAX369MnSfNOS987koDqF8DXSGdVCB29bPYNt4fca6zPO9M0XkXriSFTeOympOrA0Ku8NQGws\nBM11JyIiIiJyJrjtttu47bbbol0MyYYzOahOxusbHex8OGUb5GTgJjMrHNSv+nzgGLAhZ4qY9xKK\nJkS7CCIiIiIiImeNyCc7y3+mA5ebWS1/gpnVAFr41mVmBhBPugHNzCwO6Ap8eqqRv0VERERERETg\nzA6q3wQ2A9PM7AYz64w3GvjPwBv+TGZW3cxOmNn/+dOccyuBicDLZna3mbUF3sOb53pwHh6DiIiI\niIiInMHO2KDaOXcQaAOsA94C3gY2AW2ccwfSZTUgltBj7QWMAYYBHwFVgQ7Oua9yuegiIiIiYTmn\ncVBFRHJTbvyfPZP7VOOc+wnocoo8m/EC6+D0w8AA3yIiIiISVbGxsRw/fpyEBI2PIiKSW44fP05s\nbGyO7vOMrakWERERKUiKFSvGvn37ol0MEZECbd++fRQrVixH96mgWkRERCQfKF26NCkpKezcuZNj\nx46pKbiISA5xznHs2DF27txJSkoKpUuXPvVGWXBGN/8WERERKSgSExOpVq0au3fvZvPmzZw8eTLa\nRRIRKTBiY2MpVqwY1apVIzExMUf3raBaREREJJ9ITEykYsWKVKxYMdpFERGRCKn5t4iIiIiIiEg2\nKagWERERERERySYF1SIiIiIiIiLZpKBaREREREREJJsUVIuIiIiIiIhkk4JqERERERERkWxSUC0i\nIiIiIiKSTQqqRURERERERLLJnHPRLsMZycx2AD9GuxyZKAvsjHYhJISuS/6k65I/6brkT7ou+ZOu\nS/6k65I/6brkT/nxulR3zpU7VSYF1QWUmS13zjWOdjkkkK5L/qTrkj/puuRPui75k65L/qTrkj/p\nuuRPZ/J1UfNvERERERERkWxSUC0iIiIiIiKSTQqqC67/RLsAEpauS/6k65I/6brkT7ou+ZOuS/6k\n65I/6brkT2fsdVGfahEREREREZFsUk21iIiIiIiISDYpqC5AzKyqmX1gZnvNbJ+ZfWhm1aJdrrOF\nmd1iZpPN7EczO2xma83sGTMrli5PDTNzGSwlo1n+gsrMWmVwvvcE5StlZiPNbKeZHTSzOWZ2YbTK\nXdCZ2YJM7oWPfXl0v+QyM6tiZiPMbImZHfKd2xph8iWZ2fNmts33/22JmV0ZJl+MmT1mZpvN7IiZ\nfWNmXfLiWAqSSK6LmTU2s/+Y2Rpfnp/M7G0zqxlmf5szuI9uzKtjKgiycL9k9H/rkqB8ul9yQIT3\ny5BMrsuRoLy6X05TJM/EvnwRPXtF+hkUTXHRLoDkDDMrDMwDjgI9AAcMA+ab2UXOuYPRLN9Z4iHg\nJ+BxYAvQEBgCtDaz5s651HR5nwGmB22/Py8KeRa7D1iW7vUJ/y9mZsAMoAZwL5ACPIZ3/1zinNuS\nh+U8W/wZKB6U1gz4J6H3hu6X3HMucBuwAlgEXJNBvlFAJ+BvwEbgL8AnZtbMOfd1unxP4f0vfMK3\nz9uBSWZ2nXPuv7lzCAVSJNfldqAB8AqQDFQGBgHLff+3fg7K/wneZ1J6a3OwzGeDSO8XgLHAG0Fp\n64Je637JGZFcl5HAx0FpRXxpwZ8voPvldJ3ymTiLz16RfgZFj3NOSwFYgPuBk8C56dJq4gUOA6Jd\nvrNhAcqFSeuO9wVHG9/rGr7Xd0e7vGfLArTynfN2meS5wZendbq0EsBu4JVoH8PZsuB9aB4FSvte\n637J/XMek+73u33nu0ZQnot96b3SpcXhPWBOT5d2ju/6DQ3afi6wKtrHeiYtEV6XcJ851YFU4Mmg\n9M3AhGgf15m+RHJdfOscMOwU+9L9ksfXJcx2d/nydgpK1/1y+tckkmfiiJ69Iv0Mivai5t8FR2dg\nqXNugz/BObcJ+Bzvj1ZymXNuR5hkf81o5bwsi2RZZ+AX59x8f4Jzbi/eN6i6f/KAr7XNrcAM59zu\naJfnbOECW9BkpDNwHJiYbrsTwHtAezNL9CW3BxKACUHbTwAuDNcsWcKL5LqE+8xxzv0I7ECfObki\nwvslUrpfcshpXJcewG94tdKSgyJ8Jo702SvSz6CoUlBdcDQAvguTngycn8dlkd9d5fv5fVD6M2Z2\nwrz+79PD9R+RHPe2mZ00s11m9o4FjjeQ2f1TzcyK5k0Rz2o3AcWAcWHW6X6JrgbAJufcoaD0ZLyg\n4Nx0+Y4CG8LkA30W5Tozq49XAxr8mQNwva+/6VEzW6r+obnuT75zfcjM5pnZFUHrdb9EkZlVBVoD\nb/sCtGC6X3Je8DNxpM9ekX4GRZWC6oKjNF5fhGC7gVJ5XBYBzKwy8CQwxzm33Jd8FK+P1R/x/pk/\nBFwI/M/3MCQ5by/wIl6TsDZ4fdjaAUvM7BxfnszuH9A9lBe6A9uBWenSdL/kD6e6P0qn+7nH+drm\nZZJPcoGZxQGv49VUjwpaPQOvz2J74E7gCDDFzP6Qp4U8e0zAGzeiHdAPKAPMM7NW6fLofomuP+DF\nQeG+yNX9ksMyeCaO9Nkr0s+gqNJAZSK5wPft2jS8Pu29/OnOuW3APemyLjJvpONkvIFK9A87hznn\nVgIr0yUtNLPPgC/xBi8bGJWCSRozq4T38Dk8fY2B7heRLHkVaI7XPzTgAdQ5d2/612Y2BViKNwhg\ncPNjOU3OubvSvVxkZtPwauSGAS2jUyoJ0h1Y6ZxbFbxC90vOyuiZuKBRTXXBkUL42rSMvt2RXGJm\nhfC+5awFtHenGDnaeSO0Lgaa5EHxBHDOfYU3Cqv/nGd2//jXS+7JrMYggO6XqDjV/bE7Xb6SvhFd\nM8snOczMnsWrEe3tnPv0VPmdcyeBSUAVM6uY2+U72znn9gMfEfh/S/dLlJjZZUA9IvjMAd0vp+MU\nz8SRPntF+hkUVQqqC45kvD4Hwc4HVudxWc5aZhYPfAA0Bjo6577NwubBTcAk9/nPeWb3z0/OuQN5\nV6SzUg/gG+fcN1nYRvdL3kkGavoGk0vvfOAYv/cJTQYSgdph8oE+i3KFmT0BPALc55x7Kxu70L2U\nd9Kfa90v0dMDb+Crd7Kxre6XCEXwTBzps1ekn0FRpaC64JgOXG5mtfwJvonvWxB+/j3JYWYWA7yN\n12/3Rufc0gi3q4bXHOzLXCyepGNmjYG6/H7OpwOVzeyqdHmKA9ej+ydX+a7F+URYY6D7JSpmAPF4\no7MDaf13uwKfOueO+pI/xntQvTNo+z8A3/lmpJAcZGb34TUpfsI592oWtvNfv5+cc7/mVvnE4/s8\nuY7A/1u6X6LAzBLw5gOflcEI1eG20f2SRRE+E0f67BXpZ1BUqU91wfEm0B+YZmYD8b5Jewr4GW+g\nH8l9r+Hd8H8HDprZ5enWbXHObTGzF/G+zFqCN5hMXbyJ7lN920kOM7O3gU3AV8AeoCHeOd8KvOLL\nNh3vmkwws7/hNTV6DDDgubwu81mmO14/q7eDV+h+yRtmdovv10a+n9ea2Q5gh3NuoXNupZlNBF72\n1TxsAv4E1CRdQOCc225m/wQeM7P9ePdcV7yHqs55dDgFxqmui5ndDryMF5zNC/rM2eecW+3bTze8\n6Wn+i/dMUB74C3Ap0C33j6RgieC6PIT3v2o+8Ave3OEPARXQ/ZJrTnVd0mW9Dq/ZcNgvcnW/5JhT\nPhMT4bNXpJ9BURftibK15NwCVAMmA/uA/cBUoEa0y3W2LMBmvC8zwi1DfHl6483Tl4L3DfWveM2P\n6ka7/AV1wfsHvQpvFPDjeB+S/wEqBuUrDYzG65tzCJgLXBzt8hfkBe+b5x14c1OHW6/7JW+uQ0b/\ntxaky1MI+KfvGhwBvgBahdlXLN7gfz/ijd6+Crgl2sd4Ji6nui7A2Aiv3eXAPLz5eI/jfbk4B69/\nY9SP80xbIrgu1wOfAzt953sXXvBwWZh96X7Jo+uSLt803zVJyGA/ul9y5npszuSaDEmXL6Jnr0g/\ng6K5mK+gIiIiIiIiIpJF6lMtIiIiIiIikk0KqkVERERERESySUG1iIiIiIiISDYpqBYRERERERHJ\nJgXVIiIiIiIiItmkoFpEREREREQkmxRUi4iIiIiIiGSTgmoRERERERGRbFJQLSIiIiIiIpJNCqpF\nREREREREsklBtYiIiIiIiEg2KagWERERERERySYF1SIiIiIiIiLZpKBaREREREREJJsUVIuIiIiI\niIhkk4JqERERERERkWxSUC0iIiIiIiKSTQqqRURERERERLJJQbWIiIiIiIhINimoFhEREREREckm\nBdUiIiIiIiIi2aSgWkRERERERCSbFFSLiIicgpltNrMFBbkcZjbWzFxu7Lugyi9/FyIiEl0KqkVE\n5KxkZrXM7D9mtsbMDplZipl9b2bjzKx1tMuXG8ysp5k9EO1yZMRXPpfJsiHaZRQREQkWF+0CiIiI\n5DUzawwsBI4D44FkoBBQB7gG2A/MT7dJXaAg1OL2BGoAL4dZ1xe4Jy8Lk4lXgGVh0vfndUFERERO\nRUG1iIicjQYDhYFLnHPfBK80swrpXzvnjuZVwaLFOXcc70uG/GCRc+6DaBdCREQkEmr+LSIiZ6M6\nwK5wATWAc+7X9K/D9Z31p5nZxWY2x8wOmNl2M3vRzOLMLMnMXjCzrWZ2xMw+M7P6QfsY4mvWXCO4\nDJH21zWza8xsopltNLPDZrbHzD41s6uC9wdcBVQPalLdyrc+bJ9qM7vIzKaY2S7fcaw2s4fNLDYo\n31jf/kqY2b995+KImX1uZk1PdRxZZWatfO/X08zuNbN1vvdbZ2b3ZrDNlWY228z2+s7VV2bWJ4O8\n55rZGDPbYmbHzOwXM5tmZo3C5K1nZh+Z2X7fvj8I/mJGREQKLtVUi4jIIrHtUgAABdhJREFU2egH\noK6Z3eyc+/A09lMFmA1MBD7Aazo+ADgBNMBrUv4sUBZ4CJhqZvWdc6mnU/ggPYHSeM3YtwCVgbuB\nuWbW2jm3yJfvAeAZX1n+mm777zPacVAz+deAX4HrgX8AFwN3htnsE2AH8CRQBu98fGRmNZ1zkTbf\nLmZmZcOkH3bOHQxKuxeoALyB1zy8G/CKmZV2zg1NdyzXA1N8x/CiL+/twEgzq+WceyLouOcC8cAo\n4Du8c3wV0BxYke79KwMLfPv+G955+SNQHO/vQURECjgF1SIicjYaBlwNTDaz9cBivD68C5xzGQaZ\nYdQGbnPOTfK9ft3MVuAFVzOAds45B2Bmu4Dhvvf9JGcOA4C+wYGmmb2O10/8MWARgHNuqm+QskLO\nuQkR7ns4kAg0c86t8u37VbwvEe4ws9HOublB23zlnPtzurKsBt4H7sALfCMxOoP014D+QWnnAfWd\nc1t87/ca3vUcaGajnHNbfLXqrwIHgMucc7+kyzsfeNTMxjrn1puZAWN9x32Z/7h9njGz4FZ+5wJd\nnXPvpzvmVODPZlbXObc2wmMWEZEzlJp/i4jIWcc5twRoBIwDSgC9gH8Bq33NtGtFuKut6QJqv8WA\nASP8AbWPv8a4TvZLHip9QG1mRc2sDHAS+ALIdrNrMzsHr1Z2evrA0ndMf/e9vCnMpi8FvZ7n+5mV\n434S78uH4GVEmLxv+wNqX/mO+coQh1erDt61rgaM9gfU6fI+h/c8dIMv+RK8VgZjggJq/zbBrQx+\nSR9Q+2TnmEVE5AylmmoRETkrOee+xWs6jZlVx2vaezdwBTDNzBr5gq7MbAqTlpLBOn96mWwVOANm\nVhsvyG0PlAxafTojltf0/UwOs+57IBUI9+XDxoACOLfLq/zN0nF/65ybE2HecC0LVvt++suX2bEk\nB+X1B8IrI3z/jWHSdvl+5ui1FhGR/Ek11SIictZzzv3onBuPF1h/DlwAXBbBpiezsc7Sv3Um25/y\ni28zKwp8BnTAa6p9C15wfTVeballvHXucM5FctwFSWZ/AwX1mEVEJB3VVIuIiPg455yZfQG0wBuA\nKrft9v0sDWz2J5pZElAR2HCK7dsClYDezrkx6VeY2bAw+bNSc+2vaW8QZl09vC/mw9XS5rX6YdLO\n9/3cGPQz3LEE513n+3nJ6RdNRETOBqqpFhGRs46ZXW1mIV8sm1khfh+xeXXw+lzgD+DaBaX/lcg+\no/21pAE1omZ2DeH7Ux8ASvkG48qUc2478D/gejO7IN2+DW8ANPBGvI62O82siv+FmSXgnb+TwExf\n8lfAT0Cv9FNdmVk83qByDpjmS/4Gr0l4bzMLCcIjOXciInJ2UU21iIicjV4CypjZdOBb4BBQFW+E\n6vOA8b4+17ltDrAWeNI3wNgmoCVwObAzgu0X45siyjfX9Ra8Gta78I7rwqD8S4HrgFfN7H94gec8\nXwAdzv14U2ot8o2U/atv+/bAO2FG/s4pV/hq68N5O2gAuHXAF74Rz/fjXcMmwFPOuZ/Ba5JuZv3x\nvgRYZmb/8eXtineun3bOrffldWbWC29KrS/NzD+lVkm87gEfE37ANBEROUspqBYRkbPRALzRnlsC\nXfACpr3AKrw5mMfmRSF8wV5n4BW8+ZaPAZ/ye9/uU22/x8za441gfS/e5/oKoCPQh9Cg+iW8Ablu\nAe7Bqw1vDYQNqp1zy82sOTAU+DNQBK+Z9CN4cz3nlvsyWfce3jzgfiPw5oS+F2+E75+AB5xzw9Nv\n5JybYWZtgYF4tdMJeIOc3e2cGxWUd5mZNQEGAbfhnaudwJdEcF1EROTsYoFf9oqIiIjkf2bWCm+O\n6V7OubHRLY2IiJzN1KdaREREREREJJsUVIuIiIiIiIhkk4JqERERERERkWxSn2oRERERERGRbFJN\ntYiIiIiIiEg2KagWERERERERySYF1SIiIiIiIiLZpKBaREREREREJJsUVIuIiIiIiIhkk4JqERER\nERERkWz6f+wrpYwMebl2AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12983b20f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_transfer_curve(curve, raw_curve)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Initialization Cell",
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "nbTranslate": {
   "displayLangs": [
    "*"
   ],
   "hotkey": "alt-t",
   "langInMainMenu": true,
   "sourceLang": "en",
   "targetLang": "fr",
   "useGoogleTranslate": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}