synthetic.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Experiments on synthetic data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Required imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from utils.train_synthetic import run_synthetic_experiments\n",
    "from utils.results import get_joint_coverages, get_interval_widths, plot_timeseries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "BASELINES = ['QRNN', 'DPRNN', 'CFRNN', 'AdaptiveCFRNN']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Static noise profile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 81.24044494628906\n",
      "Epoch: 50\tTrain loss: 9.250524425506592\n",
      "Epoch: 100\tTrain loss: 9.006788682937621\n",
      "Epoch: 150\tTrain loss: 8.388805389404297\n",
      "Epoch: 200\tTrain loss: 7.657482576370239\n",
      "Epoch: 250\tTrain loss: 7.34607081413269\n",
      "Epoch: 300\tTrain loss: 7.26207537651062\n",
      "Epoch: 350\tTrain loss: 6.29205093383789\n",
      "Epoch: 400\tTrain loss: 5.312899732589722\n",
      "Epoch: 450\tTrain loss: 5.130415296554565\n",
      "Epoch: 500\tTrain loss: 4.849606561660766\n",
      "Epoch: 550\tTrain loss: 5.5908119678497314\n",
      "Epoch: 600\tTrain loss: 4.363440227508545\n",
      "Epoch: 650\tTrain loss: 4.283811330795288\n",
      "Epoch: 700\tTrain loss: 6.369759798049927\n",
      "Epoch: 750\tTrain loss: 4.4529987335205075\n",
      "Epoch: 800\tTrain loss: 4.771352291107178\n",
      "Epoch: 850\tTrain loss: 4.807813453674316\n",
      "Epoch: 900\tTrain loss: 4.194506573677063\n",
      "Epoch: 950\tTrain loss: 7.1635761737823485\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 83.23585433959961\n",
      "Epoch: 50\tTrain loss: 8.845319700241088\n",
      "Epoch: 100\tTrain loss: 8.229940462112427\n",
      "Epoch: 150\tTrain loss: 7.3740818977355955\n",
      "Epoch: 200\tTrain loss: 7.2438915252685545\n",
      "Epoch: 250\tTrain loss: 5.4991649150848385\n",
      "Epoch: 300\tTrain loss: 5.715927219390869\n",
      "Epoch: 350\tTrain loss: 4.763548803329468\n",
      "Epoch: 400\tTrain loss: 5.029480528831482\n",
      "Epoch: 450\tTrain loss: 4.403912997245788\n",
      "Epoch: 500\tTrain loss: 4.256645345687867\n",
      "Epoch: 550\tTrain loss: 3.945555877685547\n",
      "Epoch: 600\tTrain loss: 3.6085495471954347\n",
      "Epoch: 650\tTrain loss: 5.1070164203643795\n",
      "Epoch: 700\tTrain loss: 3.9269121170043944\n",
      "Epoch: 750\tTrain loss: 3.9320720434188843\n",
      "Epoch: 800\tTrain loss: 3.428136467933655\n",
      "Epoch: 850\tTrain loss: 3.87942111492157\n",
      "Epoch: 900\tTrain loss: 3.573975157737732\n",
      "Epoch: 950\tTrain loss: 3.725117301940918\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 85.75051574707031\n",
      "Epoch: 50\tTrain loss: 9.734579181671142\n",
      "Epoch: 100\tTrain loss: 9.17660150527954\n",
      "Epoch: 150\tTrain loss: 8.517406940460205\n",
      "Epoch: 200\tTrain loss: 8.613284492492676\n",
      "Epoch: 250\tTrain loss: 7.679318428039551\n",
      "Epoch: 300\tTrain loss: 6.8898694038391115\n",
      "Epoch: 350\tTrain loss: 6.665403604507446\n",
      "Epoch: 400\tTrain loss: 6.631478118896484\n",
      "Epoch: 450\tTrain loss: 6.35113492012024\n",
      "Epoch: 500\tTrain loss: 5.60222373008728\n",
      "Epoch: 550\tTrain loss: 10.00226182937622\n",
      "Epoch: 600\tTrain loss: 7.562596654891967\n",
      "Epoch: 650\tTrain loss: 7.488201713562011\n",
      "Epoch: 700\tTrain loss: 6.790530395507813\n",
      "Epoch: 750\tTrain loss: 6.643708753585815\n",
      "Epoch: 800\tTrain loss: 6.280843305587768\n",
      "Epoch: 850\tTrain loss: 6.1307238101959225\n",
      "Epoch: 900\tTrain loss: 5.650083923339844\n",
      "Epoch: 950\tTrain loss: 5.401419925689697\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 88.28747100830078\n",
      "Epoch: 50\tTrain loss: 9.003084564208985\n",
      "Epoch: 100\tTrain loss: 8.12756052017212\n",
      "Epoch: 150\tTrain loss: 7.349674940109253\n",
      "Epoch: 200\tTrain loss: 6.691097545623779\n",
      "Epoch: 250\tTrain loss: 6.196534824371338\n",
      "Epoch: 300\tTrain loss: 6.301280069351196\n",
      "Epoch: 350\tTrain loss: 5.904611539840698\n",
      "Epoch: 400\tTrain loss: 5.389737892150879\n",
      "Epoch: 450\tTrain loss: 5.4791717529296875\n",
      "Epoch: 500\tTrain loss: 5.446946430206299\n",
      "Epoch: 550\tTrain loss: 4.8739673614501955\n",
      "Epoch: 600\tTrain loss: 6.075085163116455\n",
      "Epoch: 650\tTrain loss: 5.10304708480835\n",
      "Epoch: 700\tTrain loss: 5.171582269668579\n",
      "Epoch: 750\tTrain loss: 6.107748556137085\n",
      "Epoch: 800\tTrain loss: 6.196619939804077\n",
      "Epoch: 850\tTrain loss: 5.08017783164978\n",
      "Epoch: 900\tTrain loss: 5.1339037895202635\n",
      "Epoch: 950\tTrain loss: 4.8066953182220455\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 85.4440902709961\n",
      "Epoch: 50\tTrain loss: 9.046202087402344\n",
      "Epoch: 100\tTrain loss: 8.305813837051392\n",
      "Epoch: 150\tTrain loss: 8.108648204803467\n",
      "Epoch: 200\tTrain loss: 7.060485696792602\n",
      "Epoch: 250\tTrain loss: 6.682833814620972\n",
      "Epoch: 300\tTrain loss: 6.383335590362549\n",
      "Epoch: 350\tTrain loss: 6.074161577224731\n",
      "Epoch: 400\tTrain loss: 5.295213508605957\n",
      "Epoch: 450\tTrain loss: 6.144737434387207\n",
      "Epoch: 500\tTrain loss: 5.465481853485107\n",
      "Epoch: 550\tTrain loss: 5.601260662078857\n",
      "Epoch: 600\tTrain loss: 7.024316024780274\n",
      "Epoch: 650\tTrain loss: 6.543147897720337\n",
      "Epoch: 700\tTrain loss: 10.790575313568116\n",
      "Epoch: 750\tTrain loss: 6.6365419864654545\n",
      "Epoch: 800\tTrain loss: 7.557180643081665\n",
      "Epoch: 850\tTrain loss: 6.072384023666382\n",
      "Epoch: 900\tTrain loss: 5.844747829437256\n",
      "Epoch: 950\tTrain loss: 5.203190326690674\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 79.26380310058593\n",
      "Epoch: 50\tTrain loss: 9.366790294647217\n",
      "Epoch: 100\tTrain loss: 9.060098361968993\n",
      "Epoch: 150\tTrain loss: 8.648244190216065\n",
      "Epoch: 200\tTrain loss: 8.065692281723022\n",
      "Epoch: 250\tTrain loss: 7.5783524990081785\n",
      "Epoch: 300\tTrain loss: 7.770314788818359\n",
      "Epoch: 350\tTrain loss: 7.852092123031616\n",
      "Epoch: 400\tTrain loss: 7.736189699172973\n",
      "Epoch: 450\tTrain loss: 7.965700244903564\n",
      "Epoch: 500\tTrain loss: 6.852958059310913\n",
      "Epoch: 550\tTrain loss: 7.9879326820373535\n",
      "Epoch: 600\tTrain loss: 7.506460332870484\n",
      "Epoch: 650\tTrain loss: 7.431897354125977\n",
      "Epoch: 700\tTrain loss: 7.757143497467041\n",
      "Epoch: 750\tTrain loss: 8.601996040344238\n",
      "Epoch: 800\tTrain loss: 7.779906606674194\n",
      "Epoch: 850\tTrain loss: 8.174870157241822\n",
      "Epoch: 900\tTrain loss: 7.901781702041626\n",
      "Epoch: 950\tTrain loss: 7.803011465072632\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 90.41195373535156\n",
      "Epoch: 50\tTrain loss: 9.262263202667237\n",
      "Epoch: 100\tTrain loss: 8.385658025741577\n",
      "Epoch: 150\tTrain loss: 7.9419738292694095\n",
      "Epoch: 200\tTrain loss: 6.862919902801513\n",
      "Epoch: 250\tTrain loss: 6.2690589427948\n",
      "Epoch: 300\tTrain loss: 6.438243961334228\n",
      "Epoch: 350\tTrain loss: 6.06789984703064\n",
      "Epoch: 400\tTrain loss: 5.595947408676148\n",
      "Epoch: 450\tTrain loss: 4.687317419052124\n",
      "Epoch: 500\tTrain loss: 4.5200238704681395\n",
      "Epoch: 550\tTrain loss: 5.078364419937134\n",
      "Epoch: 600\tTrain loss: 4.3638090133667\n",
      "Epoch: 650\tTrain loss: 3.9496761083602907\n",
      "Epoch: 700\tTrain loss: 4.353389596939087\n",
      "Epoch: 750\tTrain loss: 4.251759195327759\n",
      "Epoch: 800\tTrain loss: 4.050025820732117\n",
      "Epoch: 850\tTrain loss: 4.046064734458923\n",
      "Epoch: 900\tTrain loss: 4.901954507827758\n",
      "Epoch: 950\tTrain loss: 4.576928567886353\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 79.53520736694335\n",
      "Epoch: 50\tTrain loss: 9.07018690109253\n",
      "Epoch: 100\tTrain loss: 8.838050746917725\n",
      "Epoch: 150\tTrain loss: 8.201774406433106\n",
      "Epoch: 200\tTrain loss: 7.2700563907623295\n",
      "Epoch: 250\tTrain loss: 6.924431562423706\n",
      "Epoch: 300\tTrain loss: 5.619455146789551\n",
      "Epoch: 350\tTrain loss: 5.421317672729492\n",
      "Epoch: 400\tTrain loss: 4.624437236785889\n",
      "Epoch: 450\tTrain loss: 4.403502249717713\n",
      "Epoch: 500\tTrain loss: 4.459345293045044\n",
      "Epoch: 550\tTrain loss: 4.363511657714843\n",
      "Epoch: 600\tTrain loss: 4.138219332695007\n",
      "Epoch: 650\tTrain loss: 3.562211608886719\n",
      "Epoch: 700\tTrain loss: 3.7108464002609254\n",
      "Epoch: 750\tTrain loss: 3.37767608165741\n",
      "Epoch: 800\tTrain loss: 3.660181927680969\n",
      "Epoch: 850\tTrain loss: 3.6641828775405885\n",
      "Epoch: 900\tTrain loss: 3.3653597831726074\n",
      "Epoch: 950\tTrain loss: 3.374153184890747\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 85.47801513671875\n",
      "Epoch: 50\tTrain loss: 9.277861499786377\n",
      "Epoch: 100\tTrain loss: 8.506840705871582\n",
      "Epoch: 150\tTrain loss: 8.000013780593871\n",
      "Epoch: 200\tTrain loss: 7.212153434753418\n",
      "Epoch: 250\tTrain loss: 6.407902669906616\n",
      "Epoch: 300\tTrain loss: 6.268220710754394\n",
      "Epoch: 350\tTrain loss: 6.768305921554566\n",
      "Epoch: 400\tTrain loss: 5.776602125167846\n",
      "Epoch: 450\tTrain loss: 5.844431638717651\n",
      "Epoch: 500\tTrain loss: 6.655422735214233\n",
      "Epoch: 550\tTrain loss: 5.886592531204224\n",
      "Epoch: 600\tTrain loss: 5.329868125915527\n",
      "Epoch: 650\tTrain loss: 5.514274406433105\n",
      "Epoch: 700\tTrain loss: 6.357551908493042\n",
      "Epoch: 750\tTrain loss: 6.033068990707397\n",
      "Epoch: 800\tTrain loss: 5.373098468780517\n",
      "Epoch: 850\tTrain loss: 6.364412450790406\n",
      "Epoch: 900\tTrain loss: 6.571986675262451\n",
      "Epoch: 950\tTrain loss: 5.813973093032837\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 80.12358703613282\n",
      "Epoch: 50\tTrain loss: 9.82124366760254\n",
      "Epoch: 100\tTrain loss: 9.423162460327148\n",
      "Epoch: 150\tTrain loss: 8.918308973312378\n",
      "Epoch: 200\tTrain loss: 8.36513295173645\n",
      "Epoch: 250\tTrain loss: 7.73427734375\n",
      "Epoch: 300\tTrain loss: 7.912996530532837\n",
      "Epoch: 350\tTrain loss: 9.187416934967041\n",
      "Epoch: 400\tTrain loss: 7.783732271194458\n",
      "Epoch: 450\tTrain loss: 7.942233467102051\n",
      "Epoch: 500\tTrain loss: 7.41982102394104\n",
      "Epoch: 550\tTrain loss: 7.37455472946167\n",
      "Epoch: 600\tTrain loss: 6.412312269210815\n",
      "Epoch: 650\tTrain loss: 6.13273777961731\n",
      "Epoch: 700\tTrain loss: 9.176149320602416\n",
      "Epoch: 750\tTrain loss: 6.533852672576904\n",
      "Epoch: 800\tTrain loss: 6.452950096130371\n",
      "Epoch: 850\tTrain loss: 5.615079116821289\n",
      "Epoch: 900\tTrain loss: 5.047862386703491\n",
      "Epoch: 950\tTrain loss: 6.719021940231324\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 78.00207862854003\n",
      "Epoch: 50\tTrain loss: 9.181750297546387\n",
      "Epoch: 100\tTrain loss: 8.757673931121825\n",
      "Epoch: 150\tTrain loss: 8.24648699760437\n",
      "Epoch: 200\tTrain loss: 7.542593097686767\n",
      "Epoch: 250\tTrain loss: 7.65652322769165\n",
      "Epoch: 300\tTrain loss: 6.627083587646484\n",
      "Epoch: 350\tTrain loss: 6.229741001129151\n",
      "Epoch: 400\tTrain loss: 6.047731113433838\n",
      "Epoch: 450\tTrain loss: 5.172323417663574\n",
      "Epoch: 500\tTrain loss: 5.439112305641174\n",
      "Epoch: 550\tTrain loss: 4.670319437980652\n",
      "Epoch: 600\tTrain loss: 4.446900010108948\n",
      "Epoch: 650\tTrain loss: 5.230783557891845\n",
      "Epoch: 700\tTrain loss: 6.057297897338867\n",
      "Epoch: 750\tTrain loss: 7.611701154708863\n",
      "Epoch: 800\tTrain loss: 4.356600570678711\n",
      "Epoch: 850\tTrain loss: 5.857761859893799\n",
      "Epoch: 900\tTrain loss: 6.104545831680298\n",
      "Epoch: 950\tTrain loss: 4.7130985975265505\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 84.74791412353515\n",
      "Epoch: 50\tTrain loss: 8.76320309638977\n",
      "Epoch: 100\tTrain loss: 7.986342525482177\n",
      "Epoch: 150\tTrain loss: 7.286170578002929\n",
      "Epoch: 200\tTrain loss: 6.665844202041626\n",
      "Epoch: 250\tTrain loss: 7.410980749130249\n",
      "Epoch: 300\tTrain loss: 8.643399143218994\n",
      "Epoch: 350\tTrain loss: 7.000440692901611\n",
      "Epoch: 400\tTrain loss: 6.796496248245239\n",
      "Epoch: 450\tTrain loss: 6.437708806991577\n",
      "Epoch: 500\tTrain loss: 7.503562593460083\n",
      "Epoch: 550\tTrain loss: 6.756772089004516\n",
      "Epoch: 600\tTrain loss: 6.496200609207153\n",
      "Epoch: 650\tTrain loss: 6.835354900360107\n",
      "Epoch: 700\tTrain loss: 6.350413179397583\n",
      "Epoch: 750\tTrain loss: 6.015933656692505\n",
      "Epoch: 800\tTrain loss: 6.2089704990386965\n",
      "Epoch: 850\tTrain loss: 6.256645917892456\n",
      "Epoch: 900\tTrain loss: 6.910832691192627\n",
      "Epoch: 950\tTrain loss: 5.8197380065917965\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 81.62955627441406\n",
      "Epoch: 50\tTrain loss: 9.022040557861327\n",
      "Epoch: 100\tTrain loss: 7.932729482650757\n",
      "Epoch: 150\tTrain loss: 6.574428415298462\n",
      "Epoch: 200\tTrain loss: 6.120427417755127\n",
      "Epoch: 250\tTrain loss: 6.117426204681396\n",
      "Epoch: 300\tTrain loss: 5.307480573654175\n",
      "Epoch: 350\tTrain loss: 5.121606111526489\n",
      "Epoch: 400\tTrain loss: 5.0967470645904545\n",
      "Epoch: 450\tTrain loss: 5.239321446418762\n",
      "Epoch: 500\tTrain loss: 5.1202305316925045\n",
      "Epoch: 550\tTrain loss: 4.450476670265198\n",
      "Epoch: 600\tTrain loss: 5.044766044616699\n",
      "Epoch: 650\tTrain loss: 4.225764465332031\n",
      "Epoch: 700\tTrain loss: 4.654297304153443\n",
      "Epoch: 750\tTrain loss: 5.183893060684204\n",
      "Epoch: 800\tTrain loss: 3.9976943731307983\n",
      "Epoch: 850\tTrain loss: 4.303521299362183\n",
      "Epoch: 900\tTrain loss: 4.120017194747925\n",
      "Epoch: 950\tTrain loss: 4.077371144294739\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 78.14715270996093\n",
      "Epoch: 50\tTrain loss: 9.435545206069946\n",
      "Epoch: 100\tTrain loss: 8.737615728378296\n",
      "Epoch: 150\tTrain loss: 8.107306432723998\n",
      "Epoch: 200\tTrain loss: 6.742183780670166\n",
      "Epoch: 250\tTrain loss: 6.060596990585327\n",
      "Epoch: 300\tTrain loss: 5.837732315063477\n",
      "Epoch: 350\tTrain loss: 5.1867540121078495\n",
      "Epoch: 400\tTrain loss: 5.088409423828125\n",
      "Epoch: 450\tTrain loss: 4.666734647750855\n",
      "Epoch: 500\tTrain loss: 5.069767427444458\n",
      "Epoch: 550\tTrain loss: 6.823103713989258\n",
      "Epoch: 600\tTrain loss: 4.701733875274658\n",
      "Epoch: 650\tTrain loss: 5.65894341468811\n",
      "Epoch: 700\tTrain loss: 4.36440007686615\n",
      "Epoch: 750\tTrain loss: 5.165375471115112\n",
      "Epoch: 800\tTrain loss: 4.355100846290588\n",
      "Epoch: 850\tTrain loss: 4.554683113098145\n",
      "Epoch: 900\tTrain loss: 3.873276400566101\n",
      "Epoch: 950\tTrain loss: 4.378058385848999\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 78.15289421081543\n",
      "Epoch: 50\tTrain loss: 9.590887355804444\n",
      "Epoch: 100\tTrain loss: 8.888565969467162\n",
      "Epoch: 150\tTrain loss: 7.881770277023316\n",
      "Epoch: 200\tTrain loss: 7.821522665023804\n",
      "Epoch: 250\tTrain loss: 6.36688928604126\n",
      "Epoch: 300\tTrain loss: 5.864798736572266\n",
      "Epoch: 350\tTrain loss: 6.350349998474121\n",
      "Epoch: 400\tTrain loss: 5.734854888916016\n",
      "Epoch: 450\tTrain loss: 5.581589412689209\n",
      "Epoch: 500\tTrain loss: 4.877874803543091\n",
      "Epoch: 550\tTrain loss: 4.269944429397583\n",
      "Epoch: 600\tTrain loss: 4.103828573226929\n",
      "Epoch: 650\tTrain loss: 3.965160536766052\n",
      "Epoch: 700\tTrain loss: 5.3052088737487795\n",
      "Epoch: 750\tTrain loss: 3.7051788091659548\n",
      "Epoch: 800\tTrain loss: 4.484750390052795\n",
      "Epoch: 850\tTrain loss: 4.269074511528015\n",
      "Epoch: 900\tTrain loss: 3.778046154975891\n",
      "Epoch: 950\tTrain loss: 3.5355605363845823\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 83.97033157348633\n",
      "Epoch: 50\tTrain loss: 8.898229312896728\n",
      "Epoch: 100\tTrain loss: 8.446148777008057\n",
      "Epoch: 150\tTrain loss: 7.6867447853088375\n",
      "Epoch: 200\tTrain loss: 6.972411394119263\n",
      "Epoch: 250\tTrain loss: 6.254526090621948\n",
      "Epoch: 300\tTrain loss: 5.848462629318237\n",
      "Epoch: 350\tTrain loss: 5.52229323387146\n",
      "Epoch: 400\tTrain loss: 5.255409145355225\n",
      "Epoch: 450\tTrain loss: 5.041772413253784\n",
      "Epoch: 500\tTrain loss: 4.890948152542114\n",
      "Epoch: 550\tTrain loss: 5.259255886077881\n",
      "Epoch: 600\tTrain loss: 4.7583984375\n",
      "Epoch: 650\tTrain loss: 5.050693416595459\n",
      "Epoch: 700\tTrain loss: 4.641634631156921\n",
      "Epoch: 750\tTrain loss: 4.070526671409607\n",
      "Epoch: 800\tTrain loss: 4.175567388534546\n",
      "Epoch: 850\tTrain loss: 5.273600292205811\n",
      "Epoch: 900\tTrain loss: 4.424501967430115\n",
      "Epoch: 950\tTrain loss: 4.095605278015137\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 86.7265869140625\n",
      "Epoch: 50\tTrain loss: 9.130821037292481\n",
      "Epoch: 100\tTrain loss: 7.964997720718384\n",
      "Epoch: 150\tTrain loss: 6.4195130348205565\n",
      "Epoch: 200\tTrain loss: 5.629604291915894\n",
      "Epoch: 250\tTrain loss: 5.220414733886718\n",
      "Epoch: 300\tTrain loss: 4.93767032623291\n",
      "Epoch: 350\tTrain loss: 4.579490232467651\n",
      "Epoch: 400\tTrain loss: 4.600406169891357\n",
      "Epoch: 450\tTrain loss: 5.207606077194214\n",
      "Epoch: 500\tTrain loss: 4.651117897033691\n",
      "Epoch: 550\tTrain loss: 3.7423829793930055\n",
      "Epoch: 600\tTrain loss: 3.5629870891571045\n",
      "Epoch: 650\tTrain loss: 3.6564330577850344\n",
      "Epoch: 700\tTrain loss: 3.9673351764678957\n",
      "Epoch: 750\tTrain loss: 3.161724805831909\n",
      "Epoch: 800\tTrain loss: 3.819044589996338\n",
      "Epoch: 850\tTrain loss: 3.115344429016113\n",
      "Epoch: 900\tTrain loss: 2.9387707233428957\n",
      "Epoch: 950\tTrain loss: 3.23336501121521\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 84.19077377319336\n",
      "Epoch: 50\tTrain loss: 9.547489881515503\n",
      "Epoch: 100\tTrain loss: 8.972766017913818\n",
      "Epoch: 150\tTrain loss: 8.088125610351563\n",
      "Epoch: 200\tTrain loss: 7.2892204284667965\n",
      "Epoch: 250\tTrain loss: 6.478799104690552\n",
      "Epoch: 300\tTrain loss: 6.198852014541626\n",
      "Epoch: 350\tTrain loss: 5.602673721313477\n",
      "Epoch: 400\tTrain loss: 6.328011894226075\n",
      "Epoch: 450\tTrain loss: 5.482089376449585\n",
      "Epoch: 500\tTrain loss: 5.302216625213623\n",
      "Epoch: 550\tTrain loss: 4.71892147064209\n",
      "Epoch: 600\tTrain loss: 4.339744281768799\n",
      "Epoch: 650\tTrain loss: 4.727737379074097\n",
      "Epoch: 700\tTrain loss: 5.737258815765381\n",
      "Epoch: 750\tTrain loss: 4.261880588531494\n",
      "Epoch: 800\tTrain loss: 4.529950189590454\n",
      "Epoch: 850\tTrain loss: 4.127873778343201\n",
      "Epoch: 900\tTrain loss: 4.776248264312744\n",
      "Epoch: 950\tTrain loss: 6.28344030380249\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 78.31633644104004\n",
      "Epoch: 50\tTrain loss: 9.365627574920655\n",
      "Epoch: 100\tTrain loss: 9.072612762451172\n",
      "Epoch: 150\tTrain loss: 8.722649765014648\n",
      "Epoch: 200\tTrain loss: 7.824538326263427\n",
      "Epoch: 250\tTrain loss: 7.222533893585205\n",
      "Epoch: 300\tTrain loss: 6.6318799495697025\n",
      "Epoch: 350\tTrain loss: 6.335978937149048\n",
      "Epoch: 400\tTrain loss: 6.1315289497375485\n",
      "Epoch: 450\tTrain loss: 5.848065280914307\n",
      "Epoch: 500\tTrain loss: 5.335760068893433\n",
      "Epoch: 550\tTrain loss: 5.381454372406006\n",
      "Epoch: 600\tTrain loss: 5.4262940883636475\n",
      "Epoch: 650\tTrain loss: 4.867246961593628\n",
      "Epoch: 700\tTrain loss: 5.029692792892456\n",
      "Epoch: 750\tTrain loss: 4.734965658187866\n",
      "Epoch: 800\tTrain loss: 5.883873748779297\n",
      "Epoch: 850\tTrain loss: 4.844973540306091\n",
      "Epoch: 900\tTrain loss: 4.8646210670471195\n",
      "Epoch: 950\tTrain loss: 4.595320057868958\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 82.15459976196288\n",
      "Epoch: 50\tTrain loss: 9.875979614257812\n",
      "Epoch: 100\tTrain loss: 9.046921825408935\n",
      "Epoch: 150\tTrain loss: 8.466633939743042\n",
      "Epoch: 200\tTrain loss: 7.738572454452514\n",
      "Epoch: 250\tTrain loss: 7.12861442565918\n",
      "Epoch: 300\tTrain loss: 7.1785619258880615\n",
      "Epoch: 350\tTrain loss: 6.489000988006592\n",
      "Epoch: 400\tTrain loss: 5.810995769500733\n",
      "Epoch: 450\tTrain loss: 5.977119731903076\n",
      "Epoch: 500\tTrain loss: 6.185702753067017\n",
      "Epoch: 550\tTrain loss: 5.488721656799316\n",
      "Epoch: 600\tTrain loss: 6.411993455886841\n",
      "Epoch: 650\tTrain loss: 5.230825328826905\n",
      "Epoch: 700\tTrain loss: 5.23015022277832\n",
      "Epoch: 750\tTrain loss: 10.509976577758788\n",
      "Epoch: 800\tTrain loss: 5.156805801391601\n",
      "Epoch: 850\tTrain loss: 8.092289638519286\n",
      "Epoch: 900\tTrain loss: 6.047313499450683\n",
      "Epoch: 950\tTrain loss: 5.696239566802978\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 81.08464088439942\n",
      "Epoch: 50\tTrain loss: 8.651427125930786\n",
      "Epoch: 100\tTrain loss: 8.038590288162231\n",
      "Epoch: 150\tTrain loss: 7.542240762710572\n",
      "Epoch: 200\tTrain loss: 6.409868907928467\n",
      "Epoch: 250\tTrain loss: 6.5087645053863525\n",
      "Epoch: 300\tTrain loss: 5.96718578338623\n",
      "Epoch: 350\tTrain loss: 7.14032301902771\n",
      "Epoch: 400\tTrain loss: 7.941912984848022\n",
      "Epoch: 450\tTrain loss: 6.651282501220703\n",
      "Epoch: 500\tTrain loss: 6.563988113403321\n",
      "Epoch: 550\tTrain loss: 5.857947635650635\n",
      "Epoch: 600\tTrain loss: 6.634229040145874\n",
      "Epoch: 650\tTrain loss: 5.729459953308106\n",
      "Epoch: 700\tTrain loss: 5.4363562107086185\n",
      "Epoch: 750\tTrain loss: 5.36856427192688\n",
      "Epoch: 800\tTrain loss: 5.894646787643433\n",
      "Epoch: 850\tTrain loss: 5.408110666275024\n",
      "Epoch: 900\tTrain loss: 5.1193562030792235\n",
      "Epoch: 950\tTrain loss: 5.874623346328735\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 84.38402404785157\n",
      "Epoch: 50\tTrain loss: 8.984945154190063\n",
      "Epoch: 100\tTrain loss: 7.931896829605103\n",
      "Epoch: 150\tTrain loss: 6.596808958053589\n",
      "Epoch: 200\tTrain loss: 5.878898286819458\n",
      "Epoch: 250\tTrain loss: 5.446655225753784\n",
      "Epoch: 300\tTrain loss: 4.981298232078553\n",
      "Epoch: 350\tTrain loss: 5.18609619140625\n",
      "Epoch: 400\tTrain loss: 4.180840039253235\n",
      "Epoch: 450\tTrain loss: 4.535655188560486\n",
      "Epoch: 500\tTrain loss: 4.157743287086487\n",
      "Epoch: 550\tTrain loss: 4.344356513023376\n",
      "Epoch: 600\tTrain loss: 4.735836243629455\n",
      "Epoch: 650\tTrain loss: 4.018746304512024\n",
      "Epoch: 700\tTrain loss: 3.877059078216553\n",
      "Epoch: 750\tTrain loss: 3.671091318130493\n",
      "Epoch: 800\tTrain loss: 3.5929489850997927\n",
      "Epoch: 850\tTrain loss: 4.751744723320007\n",
      "Epoch: 900\tTrain loss: 3.4221625566482543\n",
      "Epoch: 950\tTrain loss: 3.471364212036133\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 79.64511108398438\n",
      "Epoch: 50\tTrain loss: 8.700117254257203\n",
      "Epoch: 100\tTrain loss: 8.310606908798217\n",
      "Epoch: 150\tTrain loss: 7.499184894561767\n",
      "Epoch: 200\tTrain loss: 7.1141773700714115\n",
      "Epoch: 250\tTrain loss: 6.540563535690308\n",
      "Epoch: 300\tTrain loss: 6.853567266464234\n",
      "Epoch: 350\tTrain loss: 6.373655891418457\n",
      "Epoch: 400\tTrain loss: 6.084713935852051\n",
      "Epoch: 450\tTrain loss: 7.599636697769165\n",
      "Epoch: 500\tTrain loss: 6.772329425811767\n",
      "Epoch: 550\tTrain loss: 6.687769365310669\n",
      "Epoch: 600\tTrain loss: 5.9620274066925045\n",
      "Epoch: 650\tTrain loss: 5.9943968772888185\n",
      "Epoch: 700\tTrain loss: 5.850201845169067\n",
      "Epoch: 750\tTrain loss: 6.502186679840088\n",
      "Epoch: 800\tTrain loss: 5.706582689285279\n",
      "Epoch: 850\tTrain loss: 5.911169910430909\n",
      "Epoch: 900\tTrain loss: 5.843581628799439\n",
      "Epoch: 950\tTrain loss: 4.871858263015747\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 88.2604476928711\n",
      "Epoch: 50\tTrain loss: 9.447468090057374\n",
      "Epoch: 100\tTrain loss: 8.619844818115235\n",
      "Epoch: 150\tTrain loss: 8.22054944038391\n",
      "Epoch: 200\tTrain loss: 7.445224714279175\n",
      "Epoch: 250\tTrain loss: 7.075113534927368\n",
      "Epoch: 300\tTrain loss: 6.821350336074829\n",
      "Epoch: 350\tTrain loss: 6.114449739456177\n",
      "Epoch: 400\tTrain loss: 5.532419490814209\n",
      "Epoch: 450\tTrain loss: 6.2603926181793215\n",
      "Epoch: 500\tTrain loss: 5.4449488639831545\n",
      "Epoch: 550\tTrain loss: 5.973920202255249\n",
      "Epoch: 600\tTrain loss: 5.917870712280274\n",
      "Epoch: 650\tTrain loss: 5.950586795806885\n",
      "Epoch: 700\tTrain loss: 5.249895858764648\n",
      "Epoch: 750\tTrain loss: 5.106182670593261\n",
      "Epoch: 800\tTrain loss: 4.460706162452698\n",
      "Epoch: 850\tTrain loss: 4.726734924316406\n",
      "Epoch: 900\tTrain loss: 8.473365688323975\n",
      "Epoch: 950\tTrain loss: 9.578587436676026\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 81.02315750122071\n",
      "Epoch: 50\tTrain loss: 9.183327436447144\n",
      "Epoch: 100\tTrain loss: 8.420903825759888\n",
      "Epoch: 150\tTrain loss: 7.909148645401001\n",
      "Epoch: 200\tTrain loss: 7.060808753967285\n",
      "Epoch: 250\tTrain loss: 6.960767078399658\n",
      "Epoch: 300\tTrain loss: 6.499058866500855\n",
      "Epoch: 350\tTrain loss: 5.884964179992676\n",
      "Epoch: 400\tTrain loss: 5.895913600921631\n",
      "Epoch: 450\tTrain loss: 5.517432832717896\n",
      "Epoch: 500\tTrain loss: 6.158428192138672\n",
      "Epoch: 550\tTrain loss: 5.528194808959961\n",
      "Epoch: 600\tTrain loss: 5.368613147735596\n",
      "Epoch: 650\tTrain loss: 5.110677194595337\n",
      "Epoch: 700\tTrain loss: 7.554749250411987\n",
      "Epoch: 750\tTrain loss: 7.750849008560181\n",
      "Epoch: 800\tTrain loss: 6.809018898010254\n",
      "Epoch: 850\tTrain loss: 7.344015264511109\n",
      "Epoch: 900\tTrain loss: 8.070838832855225\n",
      "Epoch: 950\tTrain loss: 6.384436750411988\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 81.68156967163085\n",
      "Epoch: 50\tTrain loss: 9.24165678024292\n",
      "Epoch: 100\tTrain loss: 9.172164154052734\n",
      "Epoch: 150\tTrain loss: 8.476519298553466\n",
      "Epoch: 200\tTrain loss: 8.29114727973938\n",
      "Epoch: 250\tTrain loss: 7.881560611724853\n",
      "Epoch: 300\tTrain loss: 7.660659742355347\n",
      "Epoch: 350\tTrain loss: 7.57628288269043\n",
      "Epoch: 400\tTrain loss: 6.613582563400269\n",
      "Epoch: 450\tTrain loss: 6.758566904067993\n",
      "Epoch: 500\tTrain loss: 6.653642702102661\n",
      "Epoch: 550\tTrain loss: 6.410918951034546\n",
      "Epoch: 600\tTrain loss: 6.143301963806152\n",
      "Epoch: 650\tTrain loss: 5.5058962345123295\n",
      "Epoch: 700\tTrain loss: 5.630884885787964\n",
      "Epoch: 750\tTrain loss: 6.375167465209961\n",
      "Epoch: 800\tTrain loss: 5.387549257278442\n",
      "Epoch: 850\tTrain loss: 6.046385860443115\n",
      "Epoch: 900\tTrain loss: 5.517148065567016\n",
      "Epoch: 950\tTrain loss: 5.485147428512573\n",
      "Epoch: 0\tNormalisation loss: 1.3731305718421936\n",
      "Epoch: 100\tNormalisation loss: 1.1400811553001404\n",
      "Epoch: 200\tNormalisation loss: 1.0287028312683106\n",
      "Epoch: 300\tNormalisation loss: 0.9724959015846253\n",
      "Epoch: 400\tNormalisation loss: 0.8876942038536072\n",
      "Epoch: 500\tNormalisation loss: 0.8243596494197846\n",
      "Epoch: 600\tNormalisation loss: 0.7573744356632233\n",
      "Epoch: 700\tNormalisation loss: 0.709328681230545\n",
      "Epoch: 800\tNormalisation loss: 0.683774185180664\n",
      "Epoch: 900\tNormalisation loss: 0.6392963349819183\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 89.50358963012695\n",
      "Epoch: 50\tTrain loss: 9.040993595123291\n",
      "Epoch: 100\tTrain loss: 8.678611469268798\n",
      "Epoch: 150\tTrain loss: 8.202519035339355\n",
      "Epoch: 200\tTrain loss: 7.724748849868774\n",
      "Epoch: 250\tTrain loss: 7.0677141666412355\n",
      "Epoch: 300\tTrain loss: 6.579661130905151\n",
      "Epoch: 350\tTrain loss: 6.411355876922608\n",
      "Epoch: 400\tTrain loss: 5.955724096298217\n",
      "Epoch: 450\tTrain loss: 6.235448408126831\n",
      "Epoch: 500\tTrain loss: 6.443108701705933\n",
      "Epoch: 550\tTrain loss: 5.114755249023437\n",
      "Epoch: 600\tTrain loss: 5.378562259674072\n",
      "Epoch: 650\tTrain loss: 4.783126187324524\n",
      "Epoch: 700\tTrain loss: 4.869310426712036\n",
      "Epoch: 750\tTrain loss: 6.207082033157349\n",
      "Epoch: 800\tTrain loss: 6.052699136734009\n",
      "Epoch: 850\tTrain loss: 5.003683376312256\n",
      "Epoch: 900\tTrain loss: 5.9649382591247555\n",
      "Epoch: 950\tTrain loss: 5.791038084030151\n",
      "Epoch: 0\tNormalisation loss: 1.3760009407997131\n",
      "Epoch: 100\tNormalisation loss: 1.0837734699249268\n",
      "Epoch: 200\tNormalisation loss: 1.0261471331119538\n",
      "Epoch: 300\tNormalisation loss: 0.9671974778175354\n",
      "Epoch: 400\tNormalisation loss: 0.9111883342266083\n",
      "Epoch: 500\tNormalisation loss: 0.8628953158855438\n",
      "Epoch: 600\tNormalisation loss: 0.8176236391067505\n",
      "Epoch: 700\tNormalisation loss: 0.7863347589969635\n",
      "Epoch: 800\tNormalisation loss: 0.7564339458942413\n",
      "Epoch: 900\tNormalisation loss: 0.729747861623764\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 83.68523330688477\n",
      "Epoch: 50\tTrain loss: 9.581727123260498\n",
      "Epoch: 100\tTrain loss: 9.04995436668396\n",
      "Epoch: 150\tTrain loss: 8.275908136367798\n",
      "Epoch: 200\tTrain loss: 7.4115190505981445\n",
      "Epoch: 250\tTrain loss: 6.989978408813476\n",
      "Epoch: 300\tTrain loss: 6.185967063903808\n",
      "Epoch: 350\tTrain loss: 6.102560186386109\n",
      "Epoch: 400\tTrain loss: 7.514665031433106\n",
      "Epoch: 450\tTrain loss: 5.62588586807251\n",
      "Epoch: 500\tTrain loss: 6.197639846801758\n",
      "Epoch: 550\tTrain loss: 6.866394567489624\n",
      "Epoch: 600\tTrain loss: 5.820081949234009\n",
      "Epoch: 650\tTrain loss: 5.085397052764892\n",
      "Epoch: 700\tTrain loss: 6.3403984069824215\n",
      "Epoch: 750\tTrain loss: 4.978895282745361\n",
      "Epoch: 800\tTrain loss: 8.083449649810792\n",
      "Epoch: 850\tTrain loss: 5.948006057739258\n",
      "Epoch: 900\tTrain loss: 5.138296699523925\n",
      "Epoch: 950\tTrain loss: 4.698595380783081\n",
      "Epoch: 0\tNormalisation loss: 1.434386098384857\n",
      "Epoch: 100\tNormalisation loss: 0.9247571110725403\n",
      "Epoch: 200\tNormalisation loss: 0.8496770918369293\n",
      "Epoch: 300\tNormalisation loss: 0.7849303007125854\n",
      "Epoch: 400\tNormalisation loss: 1.122332400083542\n",
      "Epoch: 500\tNormalisation loss: 1.0650983095169066\n",
      "Epoch: 600\tNormalisation loss: 0.9129484176635743\n",
      "Epoch: 700\tNormalisation loss: 0.8400116920471191\n",
      "Epoch: 800\tNormalisation loss: 0.7905907213687897\n",
      "Epoch: 900\tNormalisation loss: 0.7447066187858582\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 85.4475082397461\n",
      "Epoch: 50\tTrain loss: 9.045368003845216\n",
      "Epoch: 100\tTrain loss: 8.872663402557373\n",
      "Epoch: 150\tTrain loss: 8.169634532928466\n",
      "Epoch: 200\tTrain loss: 7.212144374847412\n",
      "Epoch: 250\tTrain loss: 6.9273388385772705\n",
      "Epoch: 300\tTrain loss: 6.132639408111572\n",
      "Epoch: 350\tTrain loss: 5.7136670589447025\n",
      "Epoch: 400\tTrain loss: 5.7717691421508786\n",
      "Epoch: 450\tTrain loss: 5.710860729217529\n",
      "Epoch: 500\tTrain loss: 4.990969920158387\n",
      "Epoch: 550\tTrain loss: 5.033152103424072\n",
      "Epoch: 600\tTrain loss: 4.482332968711853\n",
      "Epoch: 650\tTrain loss: 5.340632486343384\n",
      "Epoch: 700\tTrain loss: 5.191479301452636\n",
      "Epoch: 750\tTrain loss: 4.995418643951416\n",
      "Epoch: 800\tTrain loss: 4.2622384309768675\n",
      "Epoch: 850\tTrain loss: 5.196611833572388\n",
      "Epoch: 900\tTrain loss: 5.100488233566284\n",
      "Epoch: 950\tTrain loss: 4.552758431434631\n",
      "Epoch: 0\tNormalisation loss: 1.4605520606040954\n",
      "Epoch: 100\tNormalisation loss: 0.9604336857795716\n",
      "Epoch: 200\tNormalisation loss: 0.8890071392059327\n",
      "Epoch: 300\tNormalisation loss: 0.8330494165420532\n",
      "Epoch: 400\tNormalisation loss: 0.7660817921161651\n",
      "Epoch: 500\tNormalisation loss: 0.9832048058509827\n",
      "Epoch: 600\tNormalisation loss: 0.851666384935379\n",
      "Epoch: 700\tNormalisation loss: 0.7850785195827484\n",
      "Epoch: 800\tNormalisation loss: 0.7471330225467682\n",
      "Epoch: 900\tNormalisation loss: 0.7128818929195404\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 83.23604736328124\n",
      "Epoch: 50\tTrain loss: 9.219482898712158\n",
      "Epoch: 100\tTrain loss: 8.232393741607666\n",
      "Epoch: 150\tTrain loss: 8.02063512802124\n",
      "Epoch: 200\tTrain loss: 6.723915529251099\n",
      "Epoch: 250\tTrain loss: 5.935933256149292\n",
      "Epoch: 300\tTrain loss: 5.295427703857422\n",
      "Epoch: 350\tTrain loss: 5.764938163757324\n",
      "Epoch: 400\tTrain loss: 5.073060655593872\n",
      "Epoch: 450\tTrain loss: 4.181943750381469\n",
      "Epoch: 500\tTrain loss: 4.372122025489807\n",
      "Epoch: 550\tTrain loss: 6.108986806869507\n",
      "Epoch: 600\tTrain loss: 6.8242506980896\n",
      "Epoch: 650\tTrain loss: 4.223234105110168\n",
      "Epoch: 700\tTrain loss: 5.307839965820312\n",
      "Epoch: 750\tTrain loss: 4.192227602005005\n",
      "Epoch: 800\tTrain loss: 3.812542128562927\n",
      "Epoch: 850\tTrain loss: 4.440666770935058\n",
      "Epoch: 900\tTrain loss: 4.036766624450683\n",
      "Epoch: 950\tTrain loss: 4.2451557397842405\n",
      "Epoch: 0\tNormalisation loss: 1.2336423516273498\n",
      "Epoch: 100\tNormalisation loss: 1.120639592409134\n",
      "Epoch: 200\tNormalisation loss: 1.015909057855606\n",
      "Epoch: 300\tNormalisation loss: 0.9521346271038056\n",
      "Epoch: 400\tNormalisation loss: 0.8790603458881379\n",
      "Epoch: 500\tNormalisation loss: 0.811534458398819\n",
      "Epoch: 600\tNormalisation loss: 0.7572154939174652\n",
      "Epoch: 700\tNormalisation loss: 0.9594051837921143\n",
      "Epoch: 800\tNormalisation loss: 0.9641362071037293\n",
      "Epoch: 900\tNormalisation loss: 0.8635981380939484\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 79.12278366088867\n",
      "Epoch: 50\tTrain loss: 9.417754936218262\n",
      "Epoch: 100\tTrain loss: 9.024152517318726\n",
      "Epoch: 150\tTrain loss: 8.833436346054077\n",
      "Epoch: 200\tTrain loss: 8.52507848739624\n",
      "Epoch: 250\tTrain loss: 8.124370861053468\n",
      "Epoch: 300\tTrain loss: 7.746031951904297\n",
      "Epoch: 350\tTrain loss: 6.959705495834351\n",
      "Epoch: 400\tTrain loss: 6.3075151443481445\n",
      "Epoch: 450\tTrain loss: 6.273167133331299\n",
      "Epoch: 500\tTrain loss: 6.7977344512939455\n",
      "Epoch: 550\tTrain loss: 5.795707130432129\n",
      "Epoch: 600\tTrain loss: 5.698840999603272\n",
      "Epoch: 650\tTrain loss: 6.213014888763428\n",
      "Epoch: 700\tTrain loss: 5.0264856815338135\n",
      "Epoch: 750\tTrain loss: 4.771403336524964\n",
      "Epoch: 800\tTrain loss: 4.87660846710205\n",
      "Epoch: 850\tTrain loss: 6.06127610206604\n",
      "Epoch: 900\tTrain loss: 5.990324211120606\n",
      "Epoch: 950\tTrain loss: 5.399152565002441\n",
      "Epoch: 0\tNormalisation loss: 1.3791279196739197\n",
      "Epoch: 100\tNormalisation loss: 1.100799411535263\n",
      "Epoch: 200\tNormalisation loss: 1.0412758708000183\n",
      "Epoch: 300\tNormalisation loss: 0.9722307324409485\n",
      "Epoch: 400\tNormalisation loss: 0.9025334119796753\n",
      "Epoch: 500\tNormalisation loss: 0.8270188570022583\n",
      "Epoch: 600\tNormalisation loss: 0.7708219707012176\n",
      "Epoch: 700\tNormalisation loss: 0.7264029443264007\n",
      "Epoch: 800\tNormalisation loss: 0.6910638988018036\n",
      "Epoch: 900\tNormalisation loss: 0.6556360483169555\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 84.95330200195312\n",
      "Epoch: 50\tTrain loss: 9.165211725234986\n",
      "Epoch: 100\tTrain loss: 8.423638772964477\n",
      "Epoch: 150\tTrain loss: 7.477045583724975\n",
      "Epoch: 200\tTrain loss: 6.6220738887786865\n",
      "Epoch: 250\tTrain loss: 5.790329170227051\n",
      "Epoch: 300\tTrain loss: 5.40159649848938\n",
      "Epoch: 350\tTrain loss: 5.001340007781982\n",
      "Epoch: 400\tTrain loss: 4.721873617172241\n",
      "Epoch: 450\tTrain loss: 4.276969766616821\n",
      "Epoch: 500\tTrain loss: 3.979682993888855\n",
      "Epoch: 550\tTrain loss: 3.6354345083236694\n",
      "Epoch: 600\tTrain loss: 3.447826623916626\n",
      "Epoch: 650\tTrain loss: 3.664746570587158\n",
      "Epoch: 700\tTrain loss: 3.156619906425476\n",
      "Epoch: 750\tTrain loss: 3.472463870048523\n",
      "Epoch: 800\tTrain loss: 3.4147178173065185\n",
      "Epoch: 850\tTrain loss: 3.674015831947327\n",
      "Epoch: 900\tTrain loss: 3.0181112766265867\n",
      "Epoch: 950\tTrain loss: 3.02276554107666\n",
      "Epoch: 0\tNormalisation loss: 1.434452223777771\n",
      "Epoch: 100\tNormalisation loss: 1.119236809015274\n",
      "Epoch: 200\tNormalisation loss: 0.9658170998096466\n",
      "Epoch: 300\tNormalisation loss: 0.9142118394374847\n",
      "Epoch: 400\tNormalisation loss: 0.8652804851531982\n",
      "Epoch: 500\tNormalisation loss: 0.8171998858451843\n",
      "Epoch: 600\tNormalisation loss: 0.7650171875953674\n",
      "Epoch: 700\tNormalisation loss: 0.7206730306148529\n",
      "Epoch: 800\tNormalisation loss: 1.1055502414703369\n",
      "Epoch: 900\tNormalisation loss: 0.9003395915031434\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 84.7526699066162\n",
      "Epoch: 50\tTrain loss: 8.742105531692506\n",
      "Epoch: 100\tTrain loss: 8.256730699539185\n",
      "Epoch: 150\tTrain loss: 7.62621169090271\n",
      "Epoch: 200\tTrain loss: 6.8438465118408205\n",
      "Epoch: 250\tTrain loss: 6.4427022457122805\n",
      "Epoch: 300\tTrain loss: 6.229275894165039\n",
      "Epoch: 350\tTrain loss: 5.7944746017456055\n",
      "Epoch: 400\tTrain loss: 6.087716913223266\n",
      "Epoch: 450\tTrain loss: 6.227401351928711\n",
      "Epoch: 500\tTrain loss: 6.921637535095215\n",
      "Epoch: 550\tTrain loss: 5.781329488754272\n",
      "Epoch: 600\tTrain loss: 5.805483055114746\n",
      "Epoch: 650\tTrain loss: 6.216368865966797\n",
      "Epoch: 700\tTrain loss: 5.722856855392456\n",
      "Epoch: 750\tTrain loss: 5.34336485862732\n",
      "Epoch: 800\tTrain loss: 4.805745029449463\n",
      "Epoch: 850\tTrain loss: 4.743149280548096\n",
      "Epoch: 900\tTrain loss: 4.398545837402343\n",
      "Epoch: 950\tTrain loss: 4.309070205688476\n",
      "Epoch: 0\tNormalisation loss: 1.380394732952118\n",
      "Epoch: 100\tNormalisation loss: 1.021468061208725\n",
      "Epoch: 200\tNormalisation loss: 0.9561675548553467\n",
      "Epoch: 300\tNormalisation loss: 0.9058136224746705\n",
      "Epoch: 400\tNormalisation loss: 0.8495520114898681\n",
      "Epoch: 500\tNormalisation loss: 0.7959563374519348\n",
      "Epoch: 600\tNormalisation loss: 0.7447491824626923\n",
      "Epoch: 700\tNormalisation loss: 0.9997459828853608\n",
      "Epoch: 800\tNormalisation loss: 0.8814162313938141\n",
      "Epoch: 900\tNormalisation loss: 0.8136478126049042\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 85.29511947631836\n",
      "Epoch: 50\tTrain loss: 9.227641153335572\n",
      "Epoch: 100\tTrain loss: 8.718681144714356\n",
      "Epoch: 150\tTrain loss: 7.704185581207275\n",
      "Epoch: 200\tTrain loss: 6.889781188964844\n",
      "Epoch: 250\tTrain loss: 6.267303943634033\n",
      "Epoch: 300\tTrain loss: 6.4499163150787355\n",
      "Epoch: 350\tTrain loss: 5.35318956375122\n",
      "Epoch: 400\tTrain loss: 4.5866858959198\n",
      "Epoch: 450\tTrain loss: 5.4930823802947994\n",
      "Epoch: 500\tTrain loss: 4.443050503730774\n",
      "Epoch: 550\tTrain loss: 3.968444013595581\n",
      "Epoch: 600\tTrain loss: 5.660119247436524\n",
      "Epoch: 650\tTrain loss: 3.624089503288269\n",
      "Epoch: 700\tTrain loss: 4.956357765197754\n",
      "Epoch: 750\tTrain loss: 3.415421891212463\n",
      "Epoch: 800\tTrain loss: 3.387102150917053\n",
      "Epoch: 850\tTrain loss: 5.112363290786743\n",
      "Epoch: 900\tTrain loss: 4.048612713813782\n",
      "Epoch: 950\tTrain loss: 7.503252124786377\n",
      "Epoch: 0\tNormalisation loss: 1.5889132380485536\n",
      "Epoch: 100\tNormalisation loss: 1.09980548620224\n",
      "Epoch: 200\tNormalisation loss: 1.0203112185001373\n",
      "Epoch: 300\tNormalisation loss: 0.958864688873291\n",
      "Epoch: 400\tNormalisation loss: 0.8853005111217499\n",
      "Epoch: 500\tNormalisation loss: 0.8283346354961395\n",
      "Epoch: 600\tNormalisation loss: 0.7806220471858978\n",
      "Epoch: 700\tNormalisation loss: 0.7508142650127411\n",
      "Epoch: 800\tNormalisation loss: 0.7034551560878753\n",
      "Epoch: 900\tNormalisation loss: 1.151155561208725\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 82.87845916748047\n",
      "Epoch: 50\tTrain loss: 9.961201572418213\n",
      "Epoch: 100\tTrain loss: 8.808184671401978\n",
      "Epoch: 150\tTrain loss: 6.927392292022705\n",
      "Epoch: 200\tTrain loss: 5.9501574516296385\n",
      "Epoch: 250\tTrain loss: 5.357558393478394\n",
      "Epoch: 300\tTrain loss: 5.418920803070068\n",
      "Epoch: 350\tTrain loss: 4.522383785247802\n",
      "Epoch: 400\tTrain loss: 4.654705762863159\n",
      "Epoch: 450\tTrain loss: 4.000681686401367\n",
      "Epoch: 500\tTrain loss: 4.415667867660522\n",
      "Epoch: 550\tTrain loss: 3.9705204248428343\n",
      "Epoch: 600\tTrain loss: 3.8241904973983765\n",
      "Epoch: 650\tTrain loss: 4.301312398910523\n",
      "Epoch: 700\tTrain loss: 3.4357852697372437\n",
      "Epoch: 750\tTrain loss: 4.386287188529968\n",
      "Epoch: 800\tTrain loss: 3.4412041902542114\n",
      "Epoch: 850\tTrain loss: 3.869562292098999\n",
      "Epoch: 900\tTrain loss: 3.4680067777633665\n",
      "Epoch: 950\tTrain loss: 5.985481071472168\n",
      "Epoch: 0\tNormalisation loss: 1.316525673866272\n",
      "Epoch: 100\tNormalisation loss: 1.0859399437904358\n",
      "Epoch: 200\tNormalisation loss: 1.014825689792633\n",
      "Epoch: 300\tNormalisation loss: 0.9419260919094086\n",
      "Epoch: 400\tNormalisation loss: 0.8668361306190491\n",
      "Epoch: 500\tNormalisation loss: 0.804548978805542\n",
      "Epoch: 600\tNormalisation loss: 0.7503767311573029\n",
      "Epoch: 700\tNormalisation loss: 0.7020752549171447\n",
      "Epoch: 800\tNormalisation loss: 0.6651908338069916\n",
      "Epoch: 900\tNormalisation loss: 0.6354887962341309\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 78.27247123718261\n",
      "Epoch: 50\tTrain loss: 9.284101676940917\n",
      "Epoch: 100\tTrain loss: 8.75220866203308\n",
      "Epoch: 150\tTrain loss: 8.202619886398315\n",
      "Epoch: 200\tTrain loss: 8.598189163208009\n",
      "Epoch: 250\tTrain loss: 7.701869773864746\n",
      "Epoch: 300\tTrain loss: 7.472338247299194\n",
      "Epoch: 350\tTrain loss: 7.172509431838989\n",
      "Epoch: 400\tTrain loss: 6.679927921295166\n",
      "Epoch: 450\tTrain loss: 7.024628925323486\n",
      "Epoch: 500\tTrain loss: 6.037258148193359\n",
      "Epoch: 550\tTrain loss: 8.512858629226685\n",
      "Epoch: 600\tTrain loss: 6.964078760147094\n",
      "Epoch: 650\tTrain loss: 6.502758121490478\n",
      "Epoch: 700\tTrain loss: 6.8464436531066895\n",
      "Epoch: 750\tTrain loss: 6.437352085113526\n",
      "Epoch: 800\tTrain loss: 6.418541669845581\n",
      "Epoch: 850\tTrain loss: 7.575465726852417\n",
      "Epoch: 900\tTrain loss: 7.520419883728027\n",
      "Epoch: 950\tTrain loss: 6.295539808273316\n",
      "Epoch: 0\tNormalisation loss: 1.245104694366455\n",
      "Epoch: 100\tNormalisation loss: 1.0667146265506744\n",
      "Epoch: 200\tNormalisation loss: 1.0037746965885161\n",
      "Epoch: 300\tNormalisation loss: 0.926888108253479\n",
      "Epoch: 400\tNormalisation loss: 0.8682201862335205\n",
      "Epoch: 500\tNormalisation loss: 0.8059861660003662\n",
      "Epoch: 600\tNormalisation loss: 0.7681635916233063\n",
      "Epoch: 700\tNormalisation loss: 0.7405408203601838\n",
      "Epoch: 800\tNormalisation loss: 0.7097642958164215\n",
      "Epoch: 900\tNormalisation loss: 0.6957737147808075\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 81.46749114990234\n",
      "Epoch: 50\tTrain loss: 8.937374544143676\n",
      "Epoch: 100\tTrain loss: 8.428857898712158\n",
      "Epoch: 150\tTrain loss: 7.606823492050171\n",
      "Epoch: 200\tTrain loss: 6.932793426513672\n",
      "Epoch: 250\tTrain loss: 6.566731929779053\n",
      "Epoch: 300\tTrain loss: 6.121100568771363\n",
      "Epoch: 350\tTrain loss: 5.879490709304809\n",
      "Epoch: 400\tTrain loss: 6.235382509231568\n",
      "Epoch: 450\tTrain loss: 5.155001258850097\n",
      "Epoch: 500\tTrain loss: 5.193544244766235\n",
      "Epoch: 550\tTrain loss: 5.25482292175293\n",
      "Epoch: 600\tTrain loss: 5.636888217926026\n",
      "Epoch: 650\tTrain loss: 5.024979162216186\n",
      "Epoch: 700\tTrain loss: 5.760802364349365\n",
      "Epoch: 750\tTrain loss: 6.3883544921875\n",
      "Epoch: 800\tTrain loss: 6.403538990020752\n",
      "Epoch: 850\tTrain loss: 5.775782346725464\n",
      "Epoch: 900\tTrain loss: 6.237210607528686\n",
      "Epoch: 950\tTrain loss: 5.3478758335113525\n",
      "Epoch: 0\tNormalisation loss: 1.4244652032852172\n",
      "Epoch: 100\tNormalisation loss: 1.0311861157417297\n",
      "Epoch: 200\tNormalisation loss: 0.9713379859924316\n",
      "Epoch: 300\tNormalisation loss: 0.9182882905006409\n",
      "Epoch: 400\tNormalisation loss: 0.8684650480747222\n",
      "Epoch: 500\tNormalisation loss: 1.1534475803375244\n",
      "Epoch: 600\tNormalisation loss: 0.9505264639854432\n",
      "Epoch: 700\tNormalisation loss: 0.8768818318843842\n",
      "Epoch: 800\tNormalisation loss: 0.827026879787445\n",
      "Epoch: 900\tNormalisation loss: 0.7800727725028992\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 81.89487533569336\n",
      "Epoch: 50\tTrain loss: 9.2174476146698\n",
      "Epoch: 100\tTrain loss: 8.104479503631591\n",
      "Epoch: 150\tTrain loss: 7.080185794830323\n",
      "Epoch: 200\tTrain loss: 6.305279684066773\n",
      "Epoch: 250\tTrain loss: 5.99536018371582\n",
      "Epoch: 300\tTrain loss: 5.094821929931641\n",
      "Epoch: 350\tTrain loss: 4.537659168243408\n",
      "Epoch: 400\tTrain loss: 4.532218265533447\n",
      "Epoch: 450\tTrain loss: 4.217866897583008\n",
      "Epoch: 500\tTrain loss: 4.448298382759094\n",
      "Epoch: 550\tTrain loss: 4.321450686454773\n",
      "Epoch: 600\tTrain loss: 4.087678265571594\n",
      "Epoch: 650\tTrain loss: 3.875309705734253\n",
      "Epoch: 700\tTrain loss: 3.7133603572845457\n",
      "Epoch: 750\tTrain loss: 3.859604096412659\n",
      "Epoch: 800\tTrain loss: 3.988318109512329\n",
      "Epoch: 850\tTrain loss: 4.205815982818604\n",
      "Epoch: 900\tTrain loss: 3.8626057624816896\n",
      "Epoch: 950\tTrain loss: 3.8163368463516236\n",
      "Epoch: 0\tNormalisation loss: 1.4025108337402343\n",
      "Epoch: 100\tNormalisation loss: 0.9946534931659698\n",
      "Epoch: 200\tNormalisation loss: 0.9315947651863098\n",
      "Epoch: 300\tNormalisation loss: 0.8807870507240295\n",
      "Epoch: 400\tNormalisation loss: 0.8215189218521118\n",
      "Epoch: 500\tNormalisation loss: 0.7619557917118073\n",
      "Epoch: 600\tNormalisation loss: 0.7097862243652344\n",
      "Epoch: 700\tNormalisation loss: 0.6613797307014465\n",
      "Epoch: 800\tNormalisation loss: 0.6198291063308716\n",
      "Epoch: 900\tNormalisation loss: 0.6786653637886048\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 83.10616607666016\n",
      "Epoch: 50\tTrain loss: 9.278478622436523\n",
      "Epoch: 100\tTrain loss: 8.327937173843384\n",
      "Epoch: 150\tTrain loss: 7.116293048858642\n",
      "Epoch: 200\tTrain loss: 6.363328075408935\n",
      "Epoch: 250\tTrain loss: 6.009942531585693\n",
      "Epoch: 300\tTrain loss: 5.130108165740967\n",
      "Epoch: 350\tTrain loss: 5.286893272399903\n",
      "Epoch: 400\tTrain loss: 4.313185667991638\n",
      "Epoch: 450\tTrain loss: 4.501169228553772\n",
      "Epoch: 500\tTrain loss: 4.324339580535889\n",
      "Epoch: 550\tTrain loss: 4.090535998344421\n",
      "Epoch: 600\tTrain loss: 3.914230728149414\n",
      "Epoch: 650\tTrain loss: 3.9665990352630613\n",
      "Epoch: 700\tTrain loss: 4.1453594207763675\n",
      "Epoch: 750\tTrain loss: 4.40505199432373\n",
      "Epoch: 800\tTrain loss: 4.047308254241943\n",
      "Epoch: 850\tTrain loss: 3.4530006408691407\n",
      "Epoch: 900\tTrain loss: 4.067157316207886\n",
      "Epoch: 950\tTrain loss: 3.7043118238449098\n",
      "Epoch: 0\tNormalisation loss: 1.343706524372101\n",
      "Epoch: 100\tNormalisation loss: 1.0520248532295227\n",
      "Epoch: 200\tNormalisation loss: 0.982921302318573\n",
      "Epoch: 300\tNormalisation loss: 0.9266715943813324\n",
      "Epoch: 400\tNormalisation loss: 0.8595946490764618\n",
      "Epoch: 500\tNormalisation loss: 0.7966128289699554\n",
      "Epoch: 600\tNormalisation loss: 0.7367247819900513\n",
      "Epoch: 700\tNormalisation loss: 0.6818281650543213\n",
      "Epoch: 800\tNormalisation loss: 1.1069519519805908\n",
      "Epoch: 900\tNormalisation loss: 0.8375420093536377\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 84.87225112915038\n",
      "Epoch: 50\tTrain loss: 9.61821756362915\n",
      "Epoch: 100\tTrain loss: 8.816016674041748\n",
      "Epoch: 150\tTrain loss: 8.249310064315797\n",
      "Epoch: 200\tTrain loss: 7.942151212692261\n",
      "Epoch: 250\tTrain loss: 7.8853801727294925\n",
      "Epoch: 300\tTrain loss: 6.93205623626709\n",
      "Epoch: 350\tTrain loss: 6.64282636642456\n",
      "Epoch: 400\tTrain loss: 7.218336868286133\n",
      "Epoch: 450\tTrain loss: 8.927383279800415\n",
      "Epoch: 500\tTrain loss: 8.119023323059082\n",
      "Epoch: 550\tTrain loss: 8.755059337615966\n",
      "Epoch: 600\tTrain loss: 7.196173524856567\n",
      "Epoch: 650\tTrain loss: 8.247705841064453\n",
      "Epoch: 700\tTrain loss: 7.302032852172852\n",
      "Epoch: 750\tTrain loss: 6.5116979598999025\n",
      "Epoch: 800\tTrain loss: 6.706056547164917\n",
      "Epoch: 850\tTrain loss: 6.439429521560669\n",
      "Epoch: 900\tTrain loss: 8.30726466178894\n",
      "Epoch: 950\tTrain loss: 6.328287839889526\n",
      "Epoch: 0\tNormalisation loss: 1.8933709621429444\n",
      "Epoch: 100\tNormalisation loss: 0.9907392740249634\n",
      "Epoch: 200\tNormalisation loss: 0.9227840542793274\n",
      "Epoch: 300\tNormalisation loss: 0.8448610305786133\n",
      "Epoch: 400\tNormalisation loss: 0.756724339723587\n",
      "Epoch: 500\tNormalisation loss: 0.6940891981124878\n",
      "Epoch: 600\tNormalisation loss: 0.9300257325172424\n",
      "Epoch: 700\tNormalisation loss: 0.8411890745162964\n",
      "Epoch: 800\tNormalisation loss: 0.7724728584289551\n",
      "Epoch: 900\tNormalisation loss: 0.7246203899383545\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 83.97528991699218\n",
      "Epoch: 50\tTrain loss: 9.014692306518555\n",
      "Epoch: 100\tTrain loss: 8.575937795639039\n",
      "Epoch: 150\tTrain loss: 7.888586521148682\n",
      "Epoch: 200\tTrain loss: 7.04653491973877\n",
      "Epoch: 250\tTrain loss: 5.8486473083496096\n",
      "Epoch: 300\tTrain loss: 5.236519527435303\n",
      "Epoch: 350\tTrain loss: 5.091056680679321\n",
      "Epoch: 400\tTrain loss: 4.093488502502441\n",
      "Epoch: 450\tTrain loss: 4.971089482307434\n",
      "Epoch: 500\tTrain loss: 3.7501224279403687\n",
      "Epoch: 550\tTrain loss: 3.635694408416748\n",
      "Epoch: 600\tTrain loss: 3.696952748298645\n",
      "Epoch: 650\tTrain loss: 4.307092094421387\n",
      "Epoch: 700\tTrain loss: 3.3233880519866945\n",
      "Epoch: 750\tTrain loss: 3.546467685699463\n",
      "Epoch: 800\tTrain loss: 3.236264133453369\n",
      "Epoch: 850\tTrain loss: 3.154902625083923\n",
      "Epoch: 900\tTrain loss: 3.076595902442932\n",
      "Epoch: 950\tTrain loss: 3.1797818422317503\n",
      "Epoch: 0\tNormalisation loss: 1.3061949491500855\n",
      "Epoch: 100\tNormalisation loss: 1.1301359057426452\n",
      "Epoch: 200\tNormalisation loss: 1.0480585098266602\n",
      "Epoch: 300\tNormalisation loss: 0.9867150008678436\n",
      "Epoch: 400\tNormalisation loss: 0.9272753179073334\n",
      "Epoch: 500\tNormalisation loss: 0.8649227023124695\n",
      "Epoch: 600\tNormalisation loss: 0.8057882726192475\n",
      "Epoch: 700\tNormalisation loss: 0.7474342226982117\n",
      "Epoch: 800\tNormalisation loss: 1.0984981417655946\n",
      "Epoch: 900\tNormalisation loss: 0.9366654038429261\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 90.52503128051758\n",
      "Epoch: 50\tTrain loss: 9.099216890335082\n",
      "Epoch: 100\tTrain loss: 7.867575550079346\n",
      "Epoch: 150\tTrain loss: 7.123359298706054\n",
      "Epoch: 200\tTrain loss: 6.432214307785034\n",
      "Epoch: 250\tTrain loss: 5.811590051651001\n",
      "Epoch: 300\tTrain loss: 5.682885456085205\n",
      "Epoch: 350\tTrain loss: 5.445123481750488\n",
      "Epoch: 400\tTrain loss: 5.527906608581543\n",
      "Epoch: 450\tTrain loss: 5.378407287597656\n",
      "Epoch: 500\tTrain loss: 5.412281942367554\n",
      "Epoch: 550\tTrain loss: 6.8679563999176025\n",
      "Epoch: 600\tTrain loss: 6.150000381469726\n",
      "Epoch: 650\tTrain loss: 5.634285926818848\n",
      "Epoch: 700\tTrain loss: 6.123028898239136\n",
      "Epoch: 750\tTrain loss: 5.101238203048706\n",
      "Epoch: 800\tTrain loss: 6.446169424057007\n",
      "Epoch: 850\tTrain loss: 5.015711975097656\n",
      "Epoch: 900\tTrain loss: 5.473747968673706\n",
      "Epoch: 950\tTrain loss: 4.334996271133423\n",
      "Epoch: 0\tNormalisation loss: 1.5064809441566467\n",
      "Epoch: 100\tNormalisation loss: 1.0619194865226746\n",
      "Epoch: 200\tNormalisation loss: 0.998764443397522\n",
      "Epoch: 300\tNormalisation loss: 0.9488204479217529\n",
      "Epoch: 400\tNormalisation loss: 0.8859874129295349\n",
      "Epoch: 500\tNormalisation loss: 0.8240801453590393\n",
      "Epoch: 600\tNormalisation loss: 0.7787476062774659\n",
      "Epoch: 700\tNormalisation loss: 0.7275568842887878\n",
      "Epoch: 800\tNormalisation loss: 0.6782892405986786\n",
      "Epoch: 900\tNormalisation loss: 0.6370946764945984\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 80.3324363708496\n",
      "Epoch: 50\tTrain loss: 9.41322374343872\n",
      "Epoch: 100\tTrain loss: 9.037164783477783\n",
      "Epoch: 150\tTrain loss: 8.36141266822815\n",
      "Epoch: 200\tTrain loss: 7.7356208801269535\n",
      "Epoch: 250\tTrain loss: 6.9603358745574955\n",
      "Epoch: 300\tTrain loss: 6.367994356155395\n",
      "Epoch: 350\tTrain loss: 5.461265230178833\n",
      "Epoch: 400\tTrain loss: 5.416676092147827\n",
      "Epoch: 450\tTrain loss: 4.537454080581665\n",
      "Epoch: 500\tTrain loss: 4.889493846893311\n",
      "Epoch: 550\tTrain loss: 4.9521918296813965\n",
      "Epoch: 600\tTrain loss: 4.247673416137696\n",
      "Epoch: 650\tTrain loss: 6.6988708019256595\n",
      "Epoch: 700\tTrain loss: 3.9175872564315797\n",
      "Epoch: 750\tTrain loss: 3.948887038230896\n",
      "Epoch: 800\tTrain loss: 3.8540413618087768\n",
      "Epoch: 850\tTrain loss: 3.8282999277114866\n",
      "Epoch: 900\tTrain loss: 3.4215386390686033\n",
      "Epoch: 950\tTrain loss: 6.619118738174438\n",
      "Epoch: 0\tNormalisation loss: 1.4203537106513977\n",
      "Epoch: 100\tNormalisation loss: 1.1912163734436034\n",
      "Epoch: 200\tNormalisation loss: 0.9725974142551422\n",
      "Epoch: 300\tNormalisation loss: 0.9120678901672363\n",
      "Epoch: 400\tNormalisation loss: 0.8609631538391114\n",
      "Epoch: 500\tNormalisation loss: 0.8193521976470948\n",
      "Epoch: 600\tNormalisation loss: 0.7770210981369019\n",
      "Epoch: 700\tNormalisation loss: 0.7381188571453094\n",
      "Epoch: 800\tNormalisation loss: 0.6950426995754242\n",
      "Epoch: 900\tNormalisation loss: 1.108687025308609\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 84.94687194824219\n",
      "Epoch: 50\tTrain loss: 9.115227603912354\n",
      "Epoch: 100\tTrain loss: 8.587158918380737\n",
      "Epoch: 150\tTrain loss: 7.7464268684387205\n",
      "Epoch: 200\tTrain loss: 7.613473701477051\n",
      "Epoch: 250\tTrain loss: 6.633837080001831\n",
      "Epoch: 300\tTrain loss: 7.451458978652954\n",
      "Epoch: 350\tTrain loss: 5.743749475479126\n",
      "Epoch: 400\tTrain loss: 7.186801385879517\n",
      "Epoch: 450\tTrain loss: 6.0996373176574705\n",
      "Epoch: 500\tTrain loss: 5.2463315486907955\n",
      "Epoch: 550\tTrain loss: 7.0406383037567135\n",
      "Epoch: 600\tTrain loss: 5.2012238025665285\n",
      "Epoch: 650\tTrain loss: 6.0742706775665285\n",
      "Epoch: 700\tTrain loss: 5.112344074249267\n",
      "Epoch: 750\tTrain loss: 4.7405620336532595\n",
      "Epoch: 800\tTrain loss: 5.336548089981079\n",
      "Epoch: 850\tTrain loss: 5.478794574737549\n",
      "Epoch: 900\tTrain loss: 4.887904405593872\n",
      "Epoch: 950\tTrain loss: 4.40457124710083\n",
      "Epoch: 0\tNormalisation loss: 1.354984748363495\n",
      "Epoch: 100\tNormalisation loss: 0.944582724571228\n",
      "Epoch: 200\tNormalisation loss: 0.8762294292449951\n",
      "Epoch: 300\tNormalisation loss: 0.8128012418746948\n",
      "Epoch: 400\tNormalisation loss: 1.1439494013786315\n",
      "Epoch: 500\tNormalisation loss: 0.9429359912872315\n",
      "Epoch: 600\tNormalisation loss: 0.8697427332401275\n",
      "Epoch: 700\tNormalisation loss: 0.818221652507782\n",
      "Epoch: 800\tNormalisation loss: 0.7789436638355255\n",
      "Epoch: 900\tNormalisation loss: 0.7514292657375335\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 87.35642013549804\n",
      "Epoch: 50\tTrain loss: 9.827840042114257\n",
      "Epoch: 100\tTrain loss: 9.415369367599487\n",
      "Epoch: 150\tTrain loss: 8.666235446929932\n",
      "Epoch: 200\tTrain loss: 8.639724254608154\n",
      "Epoch: 250\tTrain loss: 7.285659742355347\n",
      "Epoch: 300\tTrain loss: 7.151046180725098\n",
      "Epoch: 350\tTrain loss: 6.733164358139038\n",
      "Epoch: 400\tTrain loss: 6.093073511123658\n",
      "Epoch: 450\tTrain loss: 6.08828067779541\n",
      "Epoch: 500\tTrain loss: 5.530384731292725\n",
      "Epoch: 550\tTrain loss: 8.709383249282837\n",
      "Epoch: 600\tTrain loss: 6.816505479812622\n",
      "Epoch: 650\tTrain loss: 5.618836164474487\n",
      "Epoch: 700\tTrain loss: 6.190145540237427\n",
      "Epoch: 750\tTrain loss: 5.970053195953369\n",
      "Epoch: 800\tTrain loss: 7.306299829483033\n",
      "Epoch: 850\tTrain loss: 6.1652404308319095\n",
      "Epoch: 900\tTrain loss: 6.609395837783813\n",
      "Epoch: 950\tTrain loss: 7.825184059143067\n",
      "Epoch: 0\tNormalisation loss: 1.4926727533340454\n",
      "Epoch: 100\tNormalisation loss: 1.0159858226776124\n",
      "Epoch: 200\tNormalisation loss: 0.9682438373565674\n",
      "Epoch: 300\tNormalisation loss: 0.9241810202598572\n",
      "Epoch: 400\tNormalisation loss: 0.8742155373096466\n",
      "Epoch: 500\tNormalisation loss: 0.8080674111843109\n",
      "Epoch: 600\tNormalisation loss: 1.1036350548267364\n",
      "Epoch: 700\tNormalisation loss: 0.8855434477329254\n",
      "Epoch: 800\tNormalisation loss: 0.8363765776157379\n",
      "Epoch: 900\tNormalisation loss: 0.7966308534145355\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 80.80132751464843\n",
      "Epoch: 50\tTrain loss: 8.776431560516357\n",
      "Epoch: 100\tTrain loss: 8.470914602279663\n",
      "Epoch: 150\tTrain loss: 8.130402660369873\n",
      "Epoch: 200\tTrain loss: 7.597188997268677\n",
      "Epoch: 250\tTrain loss: 7.472413778305054\n",
      "Epoch: 300\tTrain loss: 6.639237642288208\n",
      "Epoch: 350\tTrain loss: 6.381952333450317\n",
      "Epoch: 400\tTrain loss: 5.5809389591217045\n",
      "Epoch: 450\tTrain loss: 5.979782438278198\n",
      "Epoch: 500\tTrain loss: 5.521850824356079\n",
      "Epoch: 550\tTrain loss: 5.030989074707032\n",
      "Epoch: 600\tTrain loss: 4.713957262039185\n",
      "Epoch: 650\tTrain loss: 4.32127058506012\n",
      "Epoch: 700\tTrain loss: 4.949831295013428\n",
      "Epoch: 750\tTrain loss: 4.510575532913208\n",
      "Epoch: 800\tTrain loss: 4.492763423919678\n",
      "Epoch: 850\tTrain loss: 4.28149733543396\n",
      "Epoch: 900\tTrain loss: 3.8202861070632936\n",
      "Epoch: 950\tTrain loss: 3.833922576904297\n",
      "Epoch: 0\tNormalisation loss: 1.304414975643158\n",
      "Epoch: 100\tNormalisation loss: 1.047117495536804\n",
      "Epoch: 200\tNormalisation loss: 0.9812219917774201\n",
      "Epoch: 300\tNormalisation loss: 0.922183507680893\n",
      "Epoch: 400\tNormalisation loss: 0.8739524602890014\n",
      "Epoch: 500\tNormalisation loss: 0.982919716835022\n",
      "Epoch: 600\tNormalisation loss: 0.9041475832462311\n",
      "Epoch: 700\tNormalisation loss: 0.8503785789012909\n",
      "Epoch: 800\tNormalisation loss: 0.8062837243080139\n",
      "Epoch: 900\tNormalisation loss: 0.7679564535617829\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 77.09097671508789\n",
      "Epoch: 50\tTrain loss: 9.118207311630249\n",
      "Epoch: 100\tTrain loss: 8.857242918014526\n",
      "Epoch: 150\tTrain loss: 8.201157379150391\n",
      "Epoch: 200\tTrain loss: 7.61168360710144\n",
      "Epoch: 250\tTrain loss: 6.81648473739624\n",
      "Epoch: 300\tTrain loss: 6.353013134002685\n",
      "Epoch: 350\tTrain loss: 5.69087495803833\n",
      "Epoch: 400\tTrain loss: 5.418682432174682\n",
      "Epoch: 450\tTrain loss: 4.848593521118164\n",
      "Epoch: 500\tTrain loss: 4.686382913589478\n",
      "Epoch: 550\tTrain loss: 5.32959361076355\n",
      "Epoch: 600\tTrain loss: 4.087034082412719\n",
      "Epoch: 650\tTrain loss: 4.881962823867798\n",
      "Epoch: 700\tTrain loss: 4.120310854911804\n",
      "Epoch: 750\tTrain loss: 3.912178611755371\n",
      "Epoch: 800\tTrain loss: 4.375130558013916\n",
      "Epoch: 850\tTrain loss: 3.668883728981018\n",
      "Epoch: 900\tTrain loss: 3.7680122375488283\n",
      "Epoch: 950\tTrain loss: 3.803770089149475\n",
      "Epoch: 0\tNormalisation loss: 1.4265458822250365\n",
      "Epoch: 100\tNormalisation loss: 1.017558366060257\n",
      "Epoch: 200\tNormalisation loss: 0.953939539194107\n",
      "Epoch: 300\tNormalisation loss: 0.8952550649642944\n",
      "Epoch: 400\tNormalisation loss: 0.8316254675388336\n",
      "Epoch: 500\tNormalisation loss: 0.7776279270648956\n",
      "Epoch: 600\tNormalisation loss: 0.7234362602233887\n",
      "Epoch: 700\tNormalisation loss: 1.2421189546585083\n",
      "Epoch: 800\tNormalisation loss: 0.9137240767478942\n",
      "Epoch: 900\tNormalisation loss: 0.8233350872993469\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 80.38934707641602\n",
      "Epoch: 50\tTrain loss: 8.431652545928955\n",
      "Epoch: 100\tTrain loss: 7.790385437011719\n",
      "Epoch: 150\tTrain loss: 7.340012311935425\n",
      "Epoch: 200\tTrain loss: 6.92984676361084\n",
      "Epoch: 250\tTrain loss: 6.5844886779785154\n",
      "Epoch: 300\tTrain loss: 6.556162548065186\n",
      "Epoch: 350\tTrain loss: 6.51979513168335\n",
      "Epoch: 400\tTrain loss: 7.029332447052002\n",
      "Epoch: 450\tTrain loss: 5.855820798873902\n",
      "Epoch: 500\tTrain loss: 5.812616062164307\n",
      "Epoch: 550\tTrain loss: 5.850678873062134\n",
      "Epoch: 600\tTrain loss: 6.5557080745697025\n",
      "Epoch: 650\tTrain loss: 5.379201221466064\n",
      "Epoch: 700\tTrain loss: 6.1025062084198\n",
      "Epoch: 750\tTrain loss: 4.978126764297485\n",
      "Epoch: 800\tTrain loss: 5.71851921081543\n",
      "Epoch: 850\tTrain loss: 5.001962280273437\n",
      "Epoch: 900\tTrain loss: 4.692365646362305\n",
      "Epoch: 950\tTrain loss: 4.912478923797607\n",
      "Epoch: 0\tNormalisation loss: 1.2427197873592377\n",
      "Epoch: 100\tNormalisation loss: 1.0414158940315246\n",
      "Epoch: 200\tNormalisation loss: 0.9766607165336609\n",
      "Epoch: 300\tNormalisation loss: 0.9200258731842041\n",
      "Epoch: 400\tNormalisation loss: 0.8544817447662354\n",
      "Epoch: 500\tNormalisation loss: 0.7949779212474823\n",
      "Epoch: 600\tNormalisation loss: 0.7493811786174774\n",
      "Epoch: 700\tNormalisation loss: 0.7099384725093841\n",
      "Epoch: 800\tNormalisation loss: 0.6697041869163514\n",
      "Epoch: 900\tNormalisation loss: 1.1870090961456299\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 83.71227188110352\n",
      "Epoch: 50\tTrain loss: 9.557567882537843\n",
      "Epoch: 100\tTrain loss: 8.604257202148437\n",
      "Epoch: 150\tTrain loss: 7.797598886489868\n",
      "Epoch: 200\tTrain loss: 7.397903680801392\n",
      "Epoch: 250\tTrain loss: 6.881170463562012\n",
      "Epoch: 300\tTrain loss: 6.520148611068725\n",
      "Epoch: 350\tTrain loss: 5.747840166091919\n",
      "Epoch: 400\tTrain loss: 5.309753322601319\n",
      "Epoch: 450\tTrain loss: 6.248370122909546\n",
      "Epoch: 500\tTrain loss: 5.227959299087525\n",
      "Epoch: 550\tTrain loss: 4.682461500167847\n",
      "Epoch: 600\tTrain loss: 5.294703149795533\n",
      "Epoch: 650\tTrain loss: 4.749373984336853\n",
      "Epoch: 700\tTrain loss: 4.799724912643432\n",
      "Epoch: 750\tTrain loss: 4.689652776718139\n",
      "Epoch: 800\tTrain loss: 5.208345508575439\n",
      "Epoch: 850\tTrain loss: 5.685738658905029\n",
      "Epoch: 900\tTrain loss: 4.940444231033325\n",
      "Epoch: 950\tTrain loss: 5.201069355010986\n",
      "Epoch: 0\tNormalisation loss: 1.3878847002983092\n",
      "Epoch: 100\tNormalisation loss: 1.0307590544223786\n",
      "Epoch: 200\tNormalisation loss: 0.9606484770774841\n",
      "Epoch: 300\tNormalisation loss: 0.8877663671970367\n",
      "Epoch: 400\tNormalisation loss: 0.8153517246246338\n",
      "Epoch: 500\tNormalisation loss: 0.7551772475242615\n",
      "Epoch: 600\tNormalisation loss: 0.7122112929821014\n",
      "Epoch: 700\tNormalisation loss: 1.03649400472641\n",
      "Epoch: 800\tNormalisation loss: 0.8811654090881348\n",
      "Epoch: 900\tNormalisation loss: 0.8165050625801087\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 78.06697692871094\n",
      "Epoch: 50\tTrain loss: 9.01449375152588\n",
      "Epoch: 100\tTrain loss: 8.026634168624877\n",
      "Epoch: 150\tTrain loss: 6.853920841217041\n",
      "Epoch: 200\tTrain loss: 7.166627645492554\n",
      "Epoch: 250\tTrain loss: 6.523769807815552\n",
      "Epoch: 300\tTrain loss: 5.497459506988525\n",
      "Epoch: 350\tTrain loss: 5.480183696746826\n",
      "Epoch: 400\tTrain loss: 5.621191120147705\n",
      "Epoch: 450\tTrain loss: 5.071364402770996\n",
      "Epoch: 500\tTrain loss: 5.428377246856689\n",
      "Epoch: 550\tTrain loss: 4.854258441925049\n",
      "Epoch: 600\tTrain loss: 4.487707304954529\n",
      "Epoch: 650\tTrain loss: 6.109057950973511\n",
      "Epoch: 700\tTrain loss: 5.855729770660401\n",
      "Epoch: 750\tTrain loss: 5.139348936080933\n",
      "Epoch: 800\tTrain loss: 6.56358323097229\n",
      "Epoch: 850\tTrain loss: 5.339845371246338\n",
      "Epoch: 900\tTrain loss: 5.107042694091797\n",
      "Epoch: 950\tTrain loss: 5.726099109649658\n",
      "Epoch: 0\tNormalisation loss: 1.3915269255638123\n",
      "Epoch: 100\tNormalisation loss: 1.010163027048111\n",
      "Epoch: 200\tNormalisation loss: 0.9485545456409454\n",
      "Epoch: 300\tNormalisation loss: 0.9038596868515014\n",
      "Epoch: 400\tNormalisation loss: 0.8580649852752685\n",
      "Epoch: 500\tNormalisation loss: 0.8124093055725098\n",
      "Epoch: 600\tNormalisation loss: 0.767306786775589\n",
      "Epoch: 700\tNormalisation loss: 1.0084548711776733\n",
      "Epoch: 800\tNormalisation loss: 0.896350908279419\n",
      "Epoch: 900\tNormalisation loss: 0.8300711274147033\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.6378\n",
      "Epoch:  1 | train loss: 0.7003\n",
      "Epoch:  2 | train loss: 0.7053\n",
      "Epoch:  3 | train loss: 0.6328\n",
      "Epoch:  4 | train loss: 0.6638\n",
      "Epoch:  5 | train loss: 0.6068\n",
      "Epoch:  6 | train loss: 0.5412\n",
      "Epoch:  7 | train loss: 0.5835\n",
      "Epoch:  8 | train loss: 0.6277\n",
      "Epoch:  9 | train loss: 0.7117\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.6799\n",
      "Epoch:  1 | train loss: 0.6562\n",
      "Epoch:  2 | train loss: 0.6077\n",
      "Epoch:  3 | train loss: 0.6772\n",
      "Epoch:  4 | train loss: 0.6599\n",
      "Epoch:  5 | train loss: 0.5625\n",
      "Epoch:  6 | train loss: 0.5857\n",
      "Epoch:  7 | train loss: 0.6989\n",
      "Epoch:  8 | train loss: 0.5939\n",
      "Epoch:  9 | train loss: 0.5825\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 0.6440\n",
      "Epoch:  1 | train loss: 0.6351\n",
      "Epoch:  2 | train loss: 0.6175\n",
      "Epoch:  3 | train loss: 0.5762\n",
      "Epoch:  4 | train loss: 0.5858\n",
      "Epoch:  5 | train loss: 0.6332\n",
      "Epoch:  6 | train loss: 0.5688\n",
      "Epoch:  7 | train loss: 0.6146\n",
      "Epoch:  8 | train loss: 0.6148\n",
      "Epoch:  9 | train loss: 0.5772\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 0.5896\n",
      "Epoch:  1 | train loss: 0.5609\n",
      "Epoch:  2 | train loss: 0.6474\n",
      "Epoch:  3 | train loss: 0.6328\n",
      "Epoch:  4 | train loss: 0.6484\n",
      "Epoch:  5 | train loss: 0.5713\n",
      "Epoch:  6 | train loss: 0.6666\n",
      "Epoch:  7 | train loss: 0.5915\n",
      "Epoch:  8 | train loss: 0.6329\n",
      "Epoch:  9 | train loss: 0.6507\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 0.5996\n",
      "Epoch:  1 | train loss: 0.6446\n",
      "Epoch:  2 | train loss: 0.6327\n",
      "Epoch:  3 | train loss: 0.6463\n",
      "Epoch:  4 | train loss: 0.6034\n",
      "Epoch:  5 | train loss: 0.6363\n",
      "Epoch:  6 | train loss: 0.6250\n",
      "Epoch:  7 | train loss: 0.6142\n",
      "Epoch:  8 | train loss: 0.6017\n",
      "Epoch:  9 | train loss: 0.6309\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.6328\n",
      "Epoch:  1 | train loss: 0.7552\n",
      "Epoch:  2 | train loss: 0.5763\n",
      "Epoch:  3 | train loss: 0.6081\n",
      "Epoch:  4 | train loss: 0.6161\n",
      "Epoch:  5 | train loss: 0.6868\n",
      "Epoch:  6 | train loss: 0.6189\n",
      "Epoch:  7 | train loss: 0.5744\n",
      "Epoch:  8 | train loss: 0.5790\n",
      "Epoch:  9 | train loss: 0.6319\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.6092\n",
      "Epoch:  1 | train loss: 0.6533\n",
      "Epoch:  2 | train loss: 0.5853\n",
      "Epoch:  3 | train loss: 0.5606\n",
      "Epoch:  4 | train loss: 0.5846\n",
      "Epoch:  5 | train loss: 0.5787\n",
      "Epoch:  6 | train loss: 0.5692\n",
      "Epoch:  7 | train loss: 0.5844\n",
      "Epoch:  8 | train loss: 0.5350\n",
      "Epoch:  9 | train loss: 0.5892\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 0.6160\n",
      "Epoch:  1 | train loss: 0.6530\n",
      "Epoch:  2 | train loss: 0.5900\n",
      "Epoch:  3 | train loss: 0.6299\n",
      "Epoch:  4 | train loss: 0.6483\n",
      "Epoch:  5 | train loss: 0.6118\n",
      "Epoch:  6 | train loss: 0.6280\n",
      "Epoch:  7 | train loss: 0.7032\n",
      "Epoch:  8 | train loss: 0.6220\n",
      "Epoch:  9 | train loss: 0.5850\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 0.6454\n",
      "Epoch:  1 | train loss: 0.6738\n",
      "Epoch:  2 | train loss: 0.6216\n",
      "Epoch:  3 | train loss: 0.6614\n",
      "Epoch:  4 | train loss: 0.6861\n",
      "Epoch:  5 | train loss: 0.6972\n",
      "Epoch:  6 | train loss: 0.6278\n",
      "Epoch:  7 | train loss: 0.6621\n",
      "Epoch:  8 | train loss: 0.5655\n",
      "Epoch:  9 | train loss: 0.5445\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 0.6243\n",
      "Epoch:  1 | train loss: 0.6825\n",
      "Epoch:  2 | train loss: 0.5760\n",
      "Epoch:  3 | train loss: 0.6248\n",
      "Epoch:  4 | train loss: 0.6143\n",
      "Epoch:  5 | train loss: 0.7299\n",
      "Epoch:  6 | train loss: 0.7216\n",
      "Epoch:  7 | train loss: 0.6313\n",
      "Epoch:  8 | train loss: 0.5731\n",
      "Epoch:  9 | train loss: 0.6072\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.6899\n",
      "Epoch:  1 | train loss: 0.5916\n",
      "Epoch:  2 | train loss: 0.6192\n",
      "Epoch:  3 | train loss: 0.6774\n",
      "Epoch:  4 | train loss: 0.6000\n",
      "Epoch:  5 | train loss: 0.5883\n",
      "Epoch:  6 | train loss: 0.6012\n",
      "Epoch:  7 | train loss: 0.6156\n",
      "Epoch:  8 | train loss: 0.5479\n",
      "Epoch:  9 | train loss: 0.5486\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.6286\n",
      "Epoch:  1 | train loss: 0.5944\n",
      "Epoch:  2 | train loss: 0.6218\n",
      "Epoch:  3 | train loss: 0.6122\n",
      "Epoch:  4 | train loss: 0.5694\n",
      "Epoch:  5 | train loss: 0.5941\n",
      "Epoch:  6 | train loss: 0.5579\n",
      "Epoch:  7 | train loss: 0.6666\n",
      "Epoch:  8 | train loss: 0.6048\n",
      "Epoch:  9 | train loss: 0.5749\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 0.5830\n",
      "Epoch:  1 | train loss: 0.6047\n",
      "Epoch:  2 | train loss: 0.5999\n",
      "Epoch:  3 | train loss: 0.6587\n",
      "Epoch:  4 | train loss: 0.5945\n",
      "Epoch:  5 | train loss: 0.5913\n",
      "Epoch:  6 | train loss: 0.6180\n",
      "Epoch:  7 | train loss: 0.5959\n",
      "Epoch:  8 | train loss: 0.6198\n",
      "Epoch:  9 | train loss: 0.5878\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 0.5848\n",
      "Epoch:  1 | train loss: 0.5824\n",
      "Epoch:  2 | train loss: 0.6677\n",
      "Epoch:  3 | train loss: 0.6390\n",
      "Epoch:  4 | train loss: 0.6361\n",
      "Epoch:  5 | train loss: 0.5826\n",
      "Epoch:  6 | train loss: 0.5718\n",
      "Epoch:  7 | train loss: 0.6132\n",
      "Epoch:  8 | train loss: 0.5921\n",
      "Epoch:  9 | train loss: 0.5159\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 0.6765\n",
      "Epoch:  1 | train loss: 0.5781\n",
      "Epoch:  2 | train loss: 0.6513\n",
      "Epoch:  3 | train loss: 0.6124\n",
      "Epoch:  4 | train loss: 0.5603\n",
      "Epoch:  5 | train loss: 0.6266\n",
      "Epoch:  6 | train loss: 0.5861\n",
      "Epoch:  7 | train loss: 0.6426\n",
      "Epoch:  8 | train loss: 0.6040\n",
      "Epoch:  9 | train loss: 0.6509\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.6054\n",
      "Epoch:  1 | train loss: 0.5920\n",
      "Epoch:  2 | train loss: 0.6003\n",
      "Epoch:  3 | train loss: 0.5934\n",
      "Epoch:  4 | train loss: 0.6015\n",
      "Epoch:  5 | train loss: 0.6167\n",
      "Epoch:  6 | train loss: 0.6032\n",
      "Epoch:  7 | train loss: 0.5985\n",
      "Epoch:  8 | train loss: 0.5789\n",
      "Epoch:  9 | train loss: 0.5317\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.6301\n",
      "Epoch:  1 | train loss: 0.6681\n",
      "Epoch:  2 | train loss: 0.5945\n",
      "Epoch:  3 | train loss: 0.5637\n",
      "Epoch:  4 | train loss: 0.5412\n",
      "Epoch:  5 | train loss: 0.6905\n",
      "Epoch:  6 | train loss: 0.5769\n",
      "Epoch:  7 | train loss: 0.6526\n",
      "Epoch:  8 | train loss: 0.5829\n",
      "Epoch:  9 | train loss: 0.5963\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 0.6394\n",
      "Epoch:  1 | train loss: 0.6556\n",
      "Epoch:  2 | train loss: 0.5854\n",
      "Epoch:  3 | train loss: 0.5989\n",
      "Epoch:  4 | train loss: 0.6005\n",
      "Epoch:  5 | train loss: 0.5743\n",
      "Epoch:  6 | train loss: 0.6625\n",
      "Epoch:  7 | train loss: 0.6048\n",
      "Epoch:  8 | train loss: 0.5612\n",
      "Epoch:  9 | train loss: 0.6611\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 0.6123\n",
      "Epoch:  1 | train loss: 0.6339\n",
      "Epoch:  2 | train loss: 0.6215\n",
      "Epoch:  3 | train loss: 0.6127\n",
      "Epoch:  4 | train loss: 0.5932\n",
      "Epoch:  5 | train loss: 0.5580\n",
      "Epoch:  6 | train loss: 0.5769\n",
      "Epoch:  7 | train loss: 0.5735\n",
      "Epoch:  8 | train loss: 0.5956\n",
      "Epoch:  9 | train loss: 0.5767\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 0.6422\n",
      "Epoch:  1 | train loss: 0.6721\n",
      "Epoch:  2 | train loss: 0.7459\n",
      "Epoch:  3 | train loss: 0.6595\n",
      "Epoch:  4 | train loss: 0.6919\n",
      "Epoch:  5 | train loss: 0.6234\n",
      "Epoch:  6 | train loss: 0.6652\n",
      "Epoch:  7 | train loss: 0.6183\n",
      "Epoch:  8 | train loss: 0.6518\n",
      "Epoch:  9 | train loss: 0.6001\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.6065\n",
      "Epoch:  1 | train loss: 0.5980\n",
      "Epoch:  2 | train loss: 0.6010\n",
      "Epoch:  3 | train loss: 0.5622\n",
      "Epoch:  4 | train loss: 0.6008\n",
      "Epoch:  5 | train loss: 0.6064\n",
      "Epoch:  6 | train loss: 0.5568\n",
      "Epoch:  7 | train loss: 0.6096\n",
      "Epoch:  8 | train loss: 0.5584\n",
      "Epoch:  9 | train loss: 0.5964\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.7397\n",
      "Epoch:  1 | train loss: 0.6211\n",
      "Epoch:  2 | train loss: 0.6648\n",
      "Epoch:  3 | train loss: 0.7041\n",
      "Epoch:  4 | train loss: 0.5930\n",
      "Epoch:  5 | train loss: 0.5763\n",
      "Epoch:  6 | train loss: 0.6125\n",
      "Epoch:  7 | train loss: 0.6298\n",
      "Epoch:  8 | train loss: 0.6190\n",
      "Epoch:  9 | train loss: 0.5969\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 0.6575\n",
      "Epoch:  1 | train loss: 0.5970\n",
      "Epoch:  2 | train loss: 0.5687\n",
      "Epoch:  3 | train loss: 0.5997\n",
      "Epoch:  4 | train loss: 0.6690\n",
      "Epoch:  5 | train loss: 0.6746\n",
      "Epoch:  6 | train loss: 0.6293\n",
      "Epoch:  7 | train loss: 0.6102\n",
      "Epoch:  8 | train loss: 0.5670\n",
      "Epoch:  9 | train loss: 0.6291\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 0.6397\n",
      "Epoch:  1 | train loss: 0.5925\n",
      "Epoch:  2 | train loss: 0.6380\n",
      "Epoch:  3 | train loss: 0.6614\n",
      "Epoch:  4 | train loss: 0.6153\n",
      "Epoch:  5 | train loss: 0.6425\n",
      "Epoch:  6 | train loss: 0.6251\n",
      "Epoch:  7 | train loss: 0.6119\n",
      "Epoch:  8 | train loss: 0.6728\n",
      "Epoch:  9 | train loss: 0.6134\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 0.6337\n",
      "Epoch:  1 | train loss: 0.6897\n",
      "Epoch:  2 | train loss: 0.6381\n",
      "Epoch:  3 | train loss: 0.6558\n",
      "Epoch:  4 | train loss: 0.7261\n",
      "Epoch:  5 | train loss: 0.6151\n",
      "Epoch:  6 | train loss: 0.5870\n",
      "Epoch:  7 | train loss: 0.5780\n",
      "Epoch:  8 | train loss: 0.6216\n",
      "Epoch:  9 | train loss: 0.6109\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 14.5015\n",
      "Epoch:  1 | train loss: 14.0653\n",
      "Epoch:  2 | train loss: 9.6645\n",
      "Epoch:  3 | train loss: 11.3552\n",
      "Epoch:  4 | train loss: 10.3648\n",
      "Epoch:  5 | train loss: 10.5229\n",
      "Epoch:  6 | train loss: 9.5320\n",
      "Epoch:  7 | train loss: 11.5685\n",
      "Epoch:  8 | train loss: 11.2285\n",
      "Epoch:  9 | train loss: 10.1474\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 9.8017\n",
      "Epoch:  1 | train loss: 12.1080\n",
      "Epoch:  2 | train loss: 11.6380\n",
      "Epoch:  3 | train loss: 10.1707\n",
      "Epoch:  4 | train loss: 9.7969\n",
      "Epoch:  5 | train loss: 9.2541\n",
      "Epoch:  6 | train loss: 9.4700\n",
      "Epoch:  7 | train loss: 9.9569\n",
      "Epoch:  8 | train loss: 8.5410\n",
      "Epoch:  9 | train loss: 8.9986\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 12.8512\n",
      "Epoch:  1 | train loss: 11.9474\n",
      "Epoch:  2 | train loss: 9.7650\n",
      "Epoch:  3 | train loss: 10.5613\n",
      "Epoch:  4 | train loss: 10.5916\n",
      "Epoch:  5 | train loss: 10.4781\n",
      "Epoch:  6 | train loss: 11.9312\n",
      "Epoch:  7 | train loss: 9.0732\n",
      "Epoch:  8 | train loss: 8.3488\n",
      "Epoch:  9 | train loss: 9.5027\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 14.3251\n",
      "Epoch:  1 | train loss: 13.3876\n",
      "Epoch:  2 | train loss: 11.2828\n",
      "Epoch:  3 | train loss: 8.7488\n",
      "Epoch:  4 | train loss: 10.7719\n",
      "Epoch:  5 | train loss: 9.1372\n",
      "Epoch:  6 | train loss: 8.6423\n",
      "Epoch:  7 | train loss: 10.0059\n",
      "Epoch:  8 | train loss: 8.3689\n",
      "Epoch:  9 | train loss: 11.4922\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 12.0839\n",
      "Epoch:  1 | train loss: 15.1081\n",
      "Epoch:  2 | train loss: 11.3273\n",
      "Epoch:  3 | train loss: 13.3903\n",
      "Epoch:  4 | train loss: 10.5116\n",
      "Epoch:  5 | train loss: 9.8417\n",
      "Epoch:  6 | train loss: 8.4485\n",
      "Epoch:  7 | train loss: 9.5747\n",
      "Epoch:  8 | train loss: 9.0019\n",
      "Epoch:  9 | train loss: 9.3969\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 11.8064\n",
      "Epoch:  1 | train loss: 11.7800\n",
      "Epoch:  2 | train loss: 13.2717\n",
      "Epoch:  3 | train loss: 11.6357\n",
      "Epoch:  4 | train loss: 11.3512\n",
      "Epoch:  5 | train loss: 10.3682\n",
      "Epoch:  6 | train loss: 9.0819\n",
      "Epoch:  7 | train loss: 10.6028\n",
      "Epoch:  8 | train loss: 12.2019\n",
      "Epoch:  9 | train loss: 9.9401\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 11.0520\n",
      "Epoch:  1 | train loss: 14.0819\n",
      "Epoch:  2 | train loss: 11.7455\n",
      "Epoch:  3 | train loss: 10.1004\n",
      "Epoch:  4 | train loss: 11.6918\n",
      "Epoch:  5 | train loss: 9.4857\n",
      "Epoch:  6 | train loss: 10.7926\n",
      "Epoch:  7 | train loss: 11.0436\n",
      "Epoch:  8 | train loss: 9.2804\n",
      "Epoch:  9 | train loss: 8.9803\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 12.2259\n",
      "Epoch:  1 | train loss: 12.4678\n",
      "Epoch:  2 | train loss: 8.8301\n",
      "Epoch:  3 | train loss: 11.9483\n",
      "Epoch:  4 | train loss: 10.1855\n",
      "Epoch:  5 | train loss: 11.5354\n",
      "Epoch:  6 | train loss: 9.3730\n",
      "Epoch:  7 | train loss: 10.8201\n",
      "Epoch:  8 | train loss: 8.7962\n",
      "Epoch:  9 | train loss: 9.9606\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 11.7645\n",
      "Epoch:  1 | train loss: 13.6994\n",
      "Epoch:  2 | train loss: 10.1586\n",
      "Epoch:  3 | train loss: 10.7661\n",
      "Epoch:  4 | train loss: 10.2923\n",
      "Epoch:  5 | train loss: 9.7596\n",
      "Epoch:  6 | train loss: 8.6951\n",
      "Epoch:  7 | train loss: 9.4563\n",
      "Epoch:  8 | train loss: 10.7522\n",
      "Epoch:  9 | train loss: 8.2922\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 10.6968\n",
      "Epoch:  1 | train loss: 11.9105\n",
      "Epoch:  2 | train loss: 11.6027\n",
      "Epoch:  3 | train loss: 10.6599\n",
      "Epoch:  4 | train loss: 10.1674\n",
      "Epoch:  5 | train loss: 10.0081\n",
      "Epoch:  6 | train loss: 10.0826\n",
      "Epoch:  7 | train loss: 11.1627\n",
      "Epoch:  8 | train loss: 11.0785\n",
      "Epoch:  9 | train loss: 9.8457\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 13.4648\n",
      "Epoch:  1 | train loss: 12.0859\n",
      "Epoch:  2 | train loss: 8.6265\n",
      "Epoch:  3 | train loss: 8.1011\n",
      "Epoch:  4 | train loss: 8.5747\n",
      "Epoch:  5 | train loss: 9.4550\n",
      "Epoch:  6 | train loss: 13.4587\n",
      "Epoch:  7 | train loss: 10.3778\n",
      "Epoch:  8 | train loss: 9.0000\n",
      "Epoch:  9 | train loss: 8.3825\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 10.9910\n",
      "Epoch:  1 | train loss: 11.4373\n",
      "Epoch:  2 | train loss: 8.6309\n",
      "Epoch:  3 | train loss: 9.6723\n",
      "Epoch:  4 | train loss: 9.9977\n",
      "Epoch:  5 | train loss: 11.0179\n",
      "Epoch:  6 | train loss: 8.8984\n",
      "Epoch:  7 | train loss: 8.5947\n",
      "Epoch:  8 | train loss: 8.8481\n",
      "Epoch:  9 | train loss: 7.5154\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 12.0799\n",
      "Epoch:  1 | train loss: 13.2179\n",
      "Epoch:  2 | train loss: 9.5531\n",
      "Epoch:  3 | train loss: 9.2297\n",
      "Epoch:  4 | train loss: 10.7250\n",
      "Epoch:  5 | train loss: 11.3142\n",
      "Epoch:  6 | train loss: 9.5693\n",
      "Epoch:  7 | train loss: 10.4375\n",
      "Epoch:  8 | train loss: 8.7030\n",
      "Epoch:  9 | train loss: 7.6978\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 11.9693\n",
      "Epoch:  1 | train loss: 12.6482\n",
      "Epoch:  2 | train loss: 13.2238\n",
      "Epoch:  3 | train loss: 10.3141\n",
      "Epoch:  4 | train loss: 11.2489\n",
      "Epoch:  5 | train loss: 11.1592\n",
      "Epoch:  6 | train loss: 9.7584\n",
      "Epoch:  7 | train loss: 10.6049\n",
      "Epoch:  8 | train loss: 7.8218\n",
      "Epoch:  9 | train loss: 9.7281\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 11.4072\n",
      "Epoch:  1 | train loss: 12.6996\n",
      "Epoch:  2 | train loss: 13.5436\n",
      "Epoch:  3 | train loss: 11.9376\n",
      "Epoch:  4 | train loss: 10.6142\n",
      "Epoch:  5 | train loss: 9.2570\n",
      "Epoch:  6 | train loss: 12.2037\n",
      "Epoch:  7 | train loss: 9.5302\n",
      "Epoch:  8 | train loss: 11.0267\n",
      "Epoch:  9 | train loss: 10.0763\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 14.8730\n",
      "Epoch:  1 | train loss: 12.4226\n",
      "Epoch:  2 | train loss: 11.0865\n",
      "Epoch:  3 | train loss: 10.1762\n",
      "Epoch:  4 | train loss: 9.9338\n",
      "Epoch:  5 | train loss: 8.6440\n",
      "Epoch:  6 | train loss: 9.1189\n",
      "Epoch:  7 | train loss: 8.2765\n",
      "Epoch:  8 | train loss: 8.6916\n",
      "Epoch:  9 | train loss: 10.0454\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 10.7542\n",
      "Epoch:  1 | train loss: 13.2868\n",
      "Epoch:  2 | train loss: 12.0993\n",
      "Epoch:  3 | train loss: 10.0969\n",
      "Epoch:  4 | train loss: 8.3206\n",
      "Epoch:  5 | train loss: 8.6757\n",
      "Epoch:  6 | train loss: 11.2432\n",
      "Epoch:  7 | train loss: 10.1198\n",
      "Epoch:  8 | train loss: 7.8976\n",
      "Epoch:  9 | train loss: 10.2651\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 14.9698\n",
      "Epoch:  1 | train loss: 10.5997\n",
      "Epoch:  2 | train loss: 11.4696\n",
      "Epoch:  3 | train loss: 11.2004\n",
      "Epoch:  4 | train loss: 10.2945\n",
      "Epoch:  5 | train loss: 11.1360\n",
      "Epoch:  6 | train loss: 8.9633\n",
      "Epoch:  7 | train loss: 10.9883\n",
      "Epoch:  8 | train loss: 8.7732\n",
      "Epoch:  9 | train loss: 9.0977\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 15.9013\n",
      "Epoch:  1 | train loss: 12.9066\n",
      "Epoch:  2 | train loss: 11.9802\n",
      "Epoch:  3 | train loss: 11.6879\n",
      "Epoch:  4 | train loss: 8.2739\n",
      "Epoch:  5 | train loss: 10.0606\n",
      "Epoch:  6 | train loss: 9.2597\n",
      "Epoch:  7 | train loss: 9.0068\n",
      "Epoch:  8 | train loss: 9.4615\n",
      "Epoch:  9 | train loss: 10.2397\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 10.8988\n",
      "Epoch:  1 | train loss: 14.9089\n",
      "Epoch:  2 | train loss: 12.7159\n",
      "Epoch:  3 | train loss: 10.3906\n",
      "Epoch:  4 | train loss: 11.1511\n",
      "Epoch:  5 | train loss: 12.0162\n",
      "Epoch:  6 | train loss: 11.7325\n",
      "Epoch:  7 | train loss: 9.4879\n",
      "Epoch:  8 | train loss: 12.0782\n",
      "Epoch:  9 | train loss: 9.8465\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 11.0226\n",
      "Epoch:  1 | train loss: 9.9029\n",
      "Epoch:  2 | train loss: 11.2459\n",
      "Epoch:  3 | train loss: 10.1631\n",
      "Epoch:  4 | train loss: 10.3697\n",
      "Epoch:  5 | train loss: 10.8067\n",
      "Epoch:  6 | train loss: 8.5892\n",
      "Epoch:  7 | train loss: 8.6314\n",
      "Epoch:  8 | train loss: 9.8594\n",
      "Epoch:  9 | train loss: 9.4547\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 14.2134\n",
      "Epoch:  1 | train loss: 10.3638\n",
      "Epoch:  2 | train loss: 12.4455\n",
      "Epoch:  3 | train loss: 10.6633\n",
      "Epoch:  4 | train loss: 8.3095\n",
      "Epoch:  5 | train loss: 9.9960\n",
      "Epoch:  6 | train loss: 9.4533\n",
      "Epoch:  7 | train loss: 8.6023\n",
      "Epoch:  8 | train loss: 10.6289\n",
      "Epoch:  9 | train loss: 9.2135\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 14.3409\n",
      "Epoch:  1 | train loss: 12.7143\n",
      "Epoch:  2 | train loss: 10.8423\n",
      "Epoch:  3 | train loss: 9.8451\n",
      "Epoch:  4 | train loss: 10.9226\n",
      "Epoch:  5 | train loss: 9.6697\n",
      "Epoch:  6 | train loss: 9.9423\n",
      "Epoch:  7 | train loss: 11.2368\n",
      "Epoch:  8 | train loss: 9.2704\n",
      "Epoch:  9 | train loss: 11.5728\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 14.3293\n",
      "Epoch:  1 | train loss: 12.4069\n",
      "Epoch:  2 | train loss: 10.8012\n",
      "Epoch:  3 | train loss: 11.7216\n",
      "Epoch:  4 | train loss: 10.3629\n",
      "Epoch:  5 | train loss: 12.1549\n",
      "Epoch:  6 | train loss: 9.7778\n",
      "Epoch:  7 | train loss: 9.3933\n",
      "Epoch:  8 | train loss: 12.0224\n",
      "Epoch:  9 | train loss: 9.2936\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 14.7780\n",
      "Epoch:  1 | train loss: 10.2938\n",
      "Epoch:  2 | train loss: 9.6722\n",
      "Epoch:  3 | train loss: 11.0995\n",
      "Epoch:  4 | train loss: 10.1820\n",
      "Epoch:  5 | train loss: 11.3043\n",
      "Epoch:  6 | train loss: 9.7212\n",
      "Epoch:  7 | train loss: 9.1697\n",
      "Epoch:  8 | train loss: 10.0358\n",
      "Epoch:  9 | train loss: 10.5364\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['CFRNN', 'AdaptiveCFRNN', 'QRNN', 'DPRNN']:\n",
    "    for seed in range(5):\n",
    "        run_synthetic_experiments(experiment='static', \n",
    "                                  baseline=baseline,\n",
    "                                  n_train = 2000,\n",
    "                                  retrain_auxiliary=True,\n",
    "                                  recompute_dataset=True,\n",
    "                                  save_model=True, \n",
    "                                  save_results=True,\n",
    "                                  rnn_mode='LSTM',\n",
    "                                  seed=seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Joint coverage** \n",
    "\n",
    "Prints mean joint coverage across the horizon (mean±std of 5 random seeds, with each row indicating a different dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CFRNN\n",
      "92.8 \\(\\pm\\) 0.8\\%\n",
      "94.0 \\(\\pm\\) 0.4\\%\n",
      "94.6 \\(\\pm\\) 1.6\\%\n",
      "94.3 \\(\\pm\\) 1.4\\%\n",
      "94.3 \\(\\pm\\) 1.4\\%\n",
      "\n",
      "AdaptiveCFRNN\n",
      "93.6 \\(\\pm\\) 0.4\\%\n",
      "94.7 \\(\\pm\\) 1.0\\%\n",
      "93.6 \\(\\pm\\) 0.3\\%\n",
      "94.2 \\(\\pm\\) 1.3\\%\n",
      "93.7 \\(\\pm\\) 1.1\\%\n",
      "\n",
      "QRNN\n",
      "65.0 \\(\\pm\\) 2.7\\%\n",
      "65.6 \\(\\pm\\) 3.4\\%\n",
      "66.4 \\(\\pm\\) 1.9\\%\n",
      "65.2 \\(\\pm\\) 4.4\\%\n",
      "67.2 \\(\\pm\\) 1.6\\%\n",
      "\n",
      "DPRNN\n",
      "5.4 \\(\\pm\\) 0.5\\%\n",
      "5.6 \\(\\pm\\) 1.0\\%\n",
      "5.0 \\(\\pm\\) 0.9\\%\n",
      "4.7 \\(\\pm\\) 1.0\\%\n",
      "4.2 \\(\\pm\\) 1.0\\%\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['CFRNN', 'AdaptiveCFRNN', 'QRNN', 'DPRNN']:\n",
    "    print(baseline)\n",
    "    coverages_mean, coverages_std = get_joint_coverages(baseline, 'static', seeds=range(5))\n",
    "    \n",
    "    for m, s in zip(coverages_mean, coverages_std):\n",
    "        print('{:.1f} \\\\(\\\\pm\\\\) {:.1f}\\\\%'.format(m, s))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CFRNN\n",
      "[16.44722347 16.96501667 17.11944893 17.34209549 16.96876431]\n",
      "[3.69449142 3.34230051 3.50004207 3.76909034 3.27203236]\n",
      "\n",
      "AdaptiveCFRNN\n",
      "[21.05475494 21.90251205 20.8085498  23.28737286 21.55519516]\n",
      "[4.96181789 5.67663224 4.44719657 5.12214817 5.67844598]\n",
      "\n",
      "DPRNN\n",
      "[2.82013162 2.94538281 2.77061952 2.87406446 2.8536792 ]\n",
      "[0.33138791 0.3700536  0.36804268 0.35106769 0.38050776]\n",
      "\n",
      "QRNN\n",
      "[9.47492119 9.63299967 9.72052855 9.71000792 9.8446781 ]\n",
      "[1.99090573 1.84661966 1.91620127 1.80339317 1.98716116]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['CFRNN', 'AdaptiveCFRNN', 'DPRNN', 'QRNN']:\n",
    "    print(baseline)\n",
    "    widths_mean, widths_std = get_interval_widths(baseline, 'static', seeds=range(5))\n",
    "    \n",
    "    print(widths_mean)\n",
    "    print(widths_std)\n",
    "\n",
    "    print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Time-dependent noise profile"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training procedure for CF-RNN/MQ-RNN/DP-RNN baselines (for 5 different seeds):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 84.4234619140625\n",
      "Epoch: 50\tTrain loss: 12.888644886016845\n",
      "Epoch: 100\tTrain loss: 12.360594844818115\n",
      "Epoch: 150\tTrain loss: 11.653214836120606\n",
      "Epoch: 200\tTrain loss: 10.932942295074463\n",
      "Epoch: 250\tTrain loss: 10.140926742553711\n",
      "Epoch: 300\tTrain loss: 9.618991374969482\n",
      "Epoch: 350\tTrain loss: 9.163879299163819\n",
      "Epoch: 400\tTrain loss: 9.026247692108154\n",
      "Epoch: 450\tTrain loss: 9.197303104400635\n",
      "Epoch: 500\tTrain loss: 8.01125636100769\n",
      "Epoch: 550\tTrain loss: 7.914232587814331\n",
      "Epoch: 600\tTrain loss: 11.86298589706421\n",
      "Epoch: 650\tTrain loss: 8.87921862602234\n",
      "Epoch: 700\tTrain loss: 8.349535036087037\n",
      "Epoch: 750\tTrain loss: 8.381713342666625\n",
      "Epoch: 800\tTrain loss: 8.599508285522461\n",
      "Epoch: 850\tTrain loss: 7.528670167922973\n",
      "Epoch: 900\tTrain loss: 7.006804084777832\n",
      "Epoch: 950\tTrain loss: 8.182713890075684\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 95.2811279296875\n",
      "Epoch: 50\tTrain loss: 22.429939651489256\n",
      "Epoch: 100\tTrain loss: 21.807165718078615\n",
      "Epoch: 150\tTrain loss: 20.974736404418945\n",
      "Epoch: 200\tTrain loss: 19.58411693572998\n",
      "Epoch: 250\tTrain loss: 18.638779067993163\n",
      "Epoch: 300\tTrain loss: 18.041443252563475\n",
      "Epoch: 350\tTrain loss: 17.28767967224121\n",
      "Epoch: 400\tTrain loss: 16.09400806427002\n",
      "Epoch: 450\tTrain loss: 16.311839389801026\n",
      "Epoch: 500\tTrain loss: 15.689071083068848\n",
      "Epoch: 550\tTrain loss: 15.66753273010254\n",
      "Epoch: 600\tTrain loss: 14.688533210754395\n",
      "Epoch: 650\tTrain loss: 16.028208827972414\n",
      "Epoch: 700\tTrain loss: 15.016997146606446\n",
      "Epoch: 750\tTrain loss: 22.742497634887695\n",
      "Epoch: 800\tTrain loss: 15.881043434143066\n",
      "Epoch: 850\tTrain loss: 14.07880744934082\n",
      "Epoch: 900\tTrain loss: 13.89788885116577\n",
      "Epoch: 950\tTrain loss: 15.45349931716919\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 113.49898300170898\n",
      "Epoch: 50\tTrain loss: 37.95982666015625\n",
      "Epoch: 100\tTrain loss: 35.393855285644534\n",
      "Epoch: 150\tTrain loss: 32.876318740844724\n",
      "Epoch: 200\tTrain loss: 36.9261474609375\n",
      "Epoch: 250\tTrain loss: 29.681139373779295\n",
      "Epoch: 300\tTrain loss: 28.002218627929686\n",
      "Epoch: 350\tTrain loss: 29.329817390441896\n",
      "Epoch: 400\tTrain loss: 28.867386245727538\n",
      "Epoch: 450\tTrain loss: 27.474138450622558\n",
      "Epoch: 500\tTrain loss: 25.368652725219725\n",
      "Epoch: 550\tTrain loss: 24.691938972473146\n",
      "Epoch: 600\tTrain loss: 28.45384979248047\n",
      "Epoch: 650\tTrain loss: 29.986781883239747\n",
      "Epoch: 700\tTrain loss: 25.768702507019043\n",
      "Epoch: 750\tTrain loss: 25.198582458496094\n",
      "Epoch: 800\tTrain loss: 24.621386337280274\n",
      "Epoch: 850\tTrain loss: 25.225585174560546\n",
      "Epoch: 900\tTrain loss: 22.607170486450194\n",
      "Epoch: 950\tTrain loss: 26.53703784942627\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 134.02331771850587\n",
      "Epoch: 50\tTrain loss: 54.98128547668457\n",
      "Epoch: 100\tTrain loss: 52.21415023803711\n",
      "Epoch: 150\tTrain loss: 49.607368850708006\n",
      "Epoch: 200\tTrain loss: 47.99926223754883\n",
      "Epoch: 250\tTrain loss: 45.56360359191895\n",
      "Epoch: 300\tTrain loss: 44.0585376739502\n",
      "Epoch: 350\tTrain loss: 43.29169960021973\n",
      "Epoch: 400\tTrain loss: 45.42622680664063\n",
      "Epoch: 450\tTrain loss: 40.100886535644534\n",
      "Epoch: 500\tTrain loss: 44.25032768249512\n",
      "Epoch: 550\tTrain loss: 42.4025764465332\n",
      "Epoch: 600\tTrain loss: 38.10000858306885\n",
      "Epoch: 650\tTrain loss: 38.00457572937012\n",
      "Epoch: 700\tTrain loss: 42.11572456359863\n",
      "Epoch: 750\tTrain loss: 39.49422264099121\n",
      "Epoch: 800\tTrain loss: 35.957356452941895\n",
      "Epoch: 850\tTrain loss: 36.22798900604248\n",
      "Epoch: 900\tTrain loss: 36.64000835418701\n",
      "Epoch: 950\tTrain loss: 39.7930305480957\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 160.71371002197264\n",
      "Epoch: 50\tTrain loss: 85.52231369018554\n",
      "Epoch: 100\tTrain loss: 81.54395980834961\n",
      "Epoch: 150\tTrain loss: 77.15659103393554\n",
      "Epoch: 200\tTrain loss: 79.71729278564453\n",
      "Epoch: 250\tTrain loss: 72.0809829711914\n",
      "Epoch: 300\tTrain loss: 72.3434928894043\n",
      "Epoch: 350\tTrain loss: 69.46300621032715\n",
      "Epoch: 400\tTrain loss: 65.8802433013916\n",
      "Epoch: 450\tTrain loss: 70.10502433776855\n",
      "Epoch: 500\tTrain loss: 68.27158164978027\n",
      "Epoch: 550\tTrain loss: 66.0713264465332\n",
      "Epoch: 600\tTrain loss: 62.865806579589844\n",
      "Epoch: 650\tTrain loss: 62.11319274902344\n",
      "Epoch: 700\tTrain loss: 64.40327110290528\n",
      "Epoch: 750\tTrain loss: 69.27809829711914\n",
      "Epoch: 800\tTrain loss: 67.48683280944825\n",
      "Epoch: 850\tTrain loss: 66.5970069885254\n",
      "Epoch: 900\tTrain loss: 65.30392417907714\n",
      "Epoch: 950\tTrain loss: 69.02027473449706\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 81.79501037597656\n",
      "Epoch: 50\tTrain loss: 12.973886489868164\n",
      "Epoch: 100\tTrain loss: 12.734036827087403\n",
      "Epoch: 150\tTrain loss: 12.44111499786377\n",
      "Epoch: 200\tTrain loss: 11.599647331237794\n",
      "Epoch: 250\tTrain loss: 10.567817211151123\n",
      "Epoch: 300\tTrain loss: 10.029623031616211\n",
      "Epoch: 350\tTrain loss: 8.607649230957032\n",
      "Epoch: 400\tTrain loss: 7.97518744468689\n",
      "Epoch: 450\tTrain loss: 8.134216928482056\n",
      "Epoch: 500\tTrain loss: 6.904873085021973\n",
      "Epoch: 550\tTrain loss: 6.72160382270813\n",
      "Epoch: 600\tTrain loss: 7.068408012390137\n",
      "Epoch: 650\tTrain loss: 6.254941940307617\n",
      "Epoch: 700\tTrain loss: 9.200609588623047\n",
      "Epoch: 750\tTrain loss: 7.1578575611114506\n",
      "Epoch: 800\tTrain loss: 6.731553030014038\n",
      "Epoch: 850\tTrain loss: 6.427328968048096\n",
      "Epoch: 900\tTrain loss: 8.460420894622803\n",
      "Epoch: 950\tTrain loss: 7.296691656112671\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 102.12119750976562\n",
      "Epoch: 50\tTrain loss: 21.74437255859375\n",
      "Epoch: 100\tTrain loss: 20.10674171447754\n",
      "Epoch: 150\tTrain loss: 18.874035263061522\n",
      "Epoch: 200\tTrain loss: 17.779083442687988\n",
      "Epoch: 250\tTrain loss: 16.33564329147339\n",
      "Epoch: 300\tTrain loss: 16.320201683044434\n",
      "Epoch: 350\tTrain loss: 16.299690628051756\n",
      "Epoch: 400\tTrain loss: 14.754706764221192\n",
      "Epoch: 450\tTrain loss: 15.235816860198975\n",
      "Epoch: 500\tTrain loss: 14.455899620056153\n",
      "Epoch: 550\tTrain loss: 13.280168724060058\n",
      "Epoch: 600\tTrain loss: 13.185545730590821\n",
      "Epoch: 650\tTrain loss: 14.415399551391602\n",
      "Epoch: 700\tTrain loss: 17.609828376770018\n",
      "Epoch: 750\tTrain loss: 13.627500057220459\n",
      "Epoch: 800\tTrain loss: 13.327171516418456\n",
      "Epoch: 850\tTrain loss: 14.628689289093018\n",
      "Epoch: 900\tTrain loss: 15.468371200561524\n",
      "Epoch: 950\tTrain loss: 12.67688159942627\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 104.87369308471679\n",
      "Epoch: 50\tTrain loss: 37.163404083251955\n",
      "Epoch: 100\tTrain loss: 35.94003181457519\n",
      "Epoch: 150\tTrain loss: 34.78913631439209\n",
      "Epoch: 200\tTrain loss: 33.58717517852783\n",
      "Epoch: 250\tTrain loss: 32.90770244598389\n",
      "Epoch: 300\tTrain loss: 31.411839103698732\n",
      "Epoch: 350\tTrain loss: 31.731608390808105\n",
      "Epoch: 400\tTrain loss: 29.83922290802002\n",
      "Epoch: 450\tTrain loss: 29.22393341064453\n",
      "Epoch: 500\tTrain loss: 30.724624633789062\n",
      "Epoch: 550\tTrain loss: 29.149965858459474\n",
      "Epoch: 600\tTrain loss: 31.30580711364746\n",
      "Epoch: 650\tTrain loss: 28.88075695037842\n",
      "Epoch: 700\tTrain loss: 28.91574649810791\n",
      "Epoch: 750\tTrain loss: 33.99705696105957\n",
      "Epoch: 800\tTrain loss: 31.328807830810547\n",
      "Epoch: 850\tTrain loss: 30.658202171325684\n",
      "Epoch: 900\tTrain loss: 31.120278739929198\n",
      "Epoch: 950\tTrain loss: 33.21406192779541\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 130.59509658813477\n",
      "Epoch: 50\tTrain loss: 58.92678527832031\n",
      "Epoch: 100\tTrain loss: 54.848337936401364\n",
      "Epoch: 150\tTrain loss: 53.215538024902344\n",
      "Epoch: 200\tTrain loss: 50.38517303466797\n",
      "Epoch: 250\tTrain loss: 49.70459251403808\n",
      "Epoch: 300\tTrain loss: 50.9646411895752\n",
      "Epoch: 350\tTrain loss: 50.10303802490235\n",
      "Epoch: 400\tTrain loss: 46.73117523193359\n",
      "Epoch: 450\tTrain loss: 46.79228973388672\n",
      "Epoch: 500\tTrain loss: 48.19785690307617\n",
      "Epoch: 550\tTrain loss: 44.646545791625975\n",
      "Epoch: 600\tTrain loss: 51.0705509185791\n",
      "Epoch: 650\tTrain loss: 47.81349563598633\n",
      "Epoch: 700\tTrain loss: 49.876451873779295\n",
      "Epoch: 750\tTrain loss: 47.902710342407225\n",
      "Epoch: 800\tTrain loss: 44.54184150695801\n",
      "Epoch: 850\tTrain loss: 46.28085517883301\n",
      "Epoch: 900\tTrain loss: 45.8169662475586\n",
      "Epoch: 950\tTrain loss: 44.84323959350586\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 152.42222900390624\n",
      "Epoch: 50\tTrain loss: 84.06143035888672\n",
      "Epoch: 100\tTrain loss: 79.28846435546875\n",
      "Epoch: 150\tTrain loss: 76.26418914794922\n",
      "Epoch: 200\tTrain loss: 73.38643875122071\n",
      "Epoch: 250\tTrain loss: 71.33121795654297\n",
      "Epoch: 300\tTrain loss: 66.05809593200684\n",
      "Epoch: 350\tTrain loss: 68.1793212890625\n",
      "Epoch: 400\tTrain loss: 64.11015319824219\n",
      "Epoch: 450\tTrain loss: 65.76095542907714\n",
      "Epoch: 500\tTrain loss: 62.83871421813965\n",
      "Epoch: 550\tTrain loss: 61.248997116088866\n",
      "Epoch: 600\tTrain loss: 61.65150337219238\n",
      "Epoch: 650\tTrain loss: 60.66765365600586\n",
      "Epoch: 700\tTrain loss: 60.34023590087891\n",
      "Epoch: 750\tTrain loss: 57.899349212646484\n",
      "Epoch: 800\tTrain loss: 60.36548080444336\n",
      "Epoch: 850\tTrain loss: 71.26205101013184\n",
      "Epoch: 900\tTrain loss: 59.0400089263916\n",
      "Epoch: 950\tTrain loss: 58.449110412597655\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 80.43602676391602\n",
      "Epoch: 50\tTrain loss: 12.846064376831055\n",
      "Epoch: 100\tTrain loss: 12.365988159179688\n",
      "Epoch: 150\tTrain loss: 11.626377201080322\n",
      "Epoch: 200\tTrain loss: 10.847915554046631\n",
      "Epoch: 250\tTrain loss: 10.232018280029298\n",
      "Epoch: 300\tTrain loss: 8.999821281433105\n",
      "Epoch: 350\tTrain loss: 8.857147121429444\n",
      "Epoch: 400\tTrain loss: 8.70275411605835\n",
      "Epoch: 450\tTrain loss: 8.032811832427978\n",
      "Epoch: 500\tTrain loss: 7.8896409034729\n",
      "Epoch: 550\tTrain loss: 8.402475023269654\n",
      "Epoch: 600\tTrain loss: 7.395649719238281\n",
      "Epoch: 650\tTrain loss: 9.290483951568604\n",
      "Epoch: 700\tTrain loss: 7.6579601764678955\n",
      "Epoch: 750\tTrain loss: 8.0304705619812\n",
      "Epoch: 800\tTrain loss: 9.448664999008178\n",
      "Epoch: 850\tTrain loss: 7.1548097133636475\n",
      "Epoch: 900\tTrain loss: 7.878600120544434\n",
      "Epoch: 950\tTrain loss: 7.998249340057373\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 95.12761154174805\n",
      "Epoch: 50\tTrain loss: 21.312012100219725\n",
      "Epoch: 100\tTrain loss: 20.592498779296875\n",
      "Epoch: 150\tTrain loss: 19.194359397888185\n",
      "Epoch: 200\tTrain loss: 19.11586685180664\n",
      "Epoch: 250\tTrain loss: 17.650954246520996\n",
      "Epoch: 300\tTrain loss: 16.74121561050415\n",
      "Epoch: 350\tTrain loss: 16.831148529052733\n",
      "Epoch: 400\tTrain loss: 16.317352867126466\n",
      "Epoch: 450\tTrain loss: 15.728972721099854\n",
      "Epoch: 500\tTrain loss: 15.261245346069336\n",
      "Epoch: 550\tTrain loss: 16.11915216445923\n",
      "Epoch: 600\tTrain loss: 14.772070693969727\n",
      "Epoch: 650\tTrain loss: 14.504510879516602\n",
      "Epoch: 700\tTrain loss: 14.586884212493896\n",
      "Epoch: 750\tTrain loss: 14.487107372283935\n",
      "Epoch: 800\tTrain loss: 16.262297439575196\n",
      "Epoch: 850\tTrain loss: 14.017327117919923\n",
      "Epoch: 900\tTrain loss: 18.53071403503418\n",
      "Epoch: 950\tTrain loss: 15.967198371887207\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 107.01494750976562\n",
      "Epoch: 50\tTrain loss: 37.60706329345703\n",
      "Epoch: 100\tTrain loss: 34.58369960784912\n",
      "Epoch: 150\tTrain loss: 34.17727108001709\n",
      "Epoch: 200\tTrain loss: 35.58249626159668\n",
      "Epoch: 250\tTrain loss: 34.677588844299315\n",
      "Epoch: 300\tTrain loss: 32.92224044799805\n",
      "Epoch: 350\tTrain loss: 29.244219398498537\n",
      "Epoch: 400\tTrain loss: 28.46578598022461\n",
      "Epoch: 450\tTrain loss: 31.2123233795166\n",
      "Epoch: 500\tTrain loss: 30.882509422302245\n",
      "Epoch: 550\tTrain loss: 27.498036003112794\n",
      "Epoch: 600\tTrain loss: 28.643499755859374\n",
      "Epoch: 650\tTrain loss: 27.25665340423584\n",
      "Epoch: 700\tTrain loss: 28.763136291503905\n",
      "Epoch: 750\tTrain loss: 25.621175956726074\n",
      "Epoch: 800\tTrain loss: 27.38588104248047\n",
      "Epoch: 850\tTrain loss: 30.261739349365236\n",
      "Epoch: 900\tTrain loss: 25.467670059204103\n",
      "Epoch: 950\tTrain loss: 38.94464797973633\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 126.28602828979493\n",
      "Epoch: 50\tTrain loss: 57.64399375915527\n",
      "Epoch: 100\tTrain loss: 53.33676109313965\n",
      "Epoch: 150\tTrain loss: 49.263886260986325\n",
      "Epoch: 200\tTrain loss: 46.257103729248044\n",
      "Epoch: 250\tTrain loss: 46.686213684082034\n",
      "Epoch: 300\tTrain loss: 43.79640693664551\n",
      "Epoch: 350\tTrain loss: 46.98242835998535\n",
      "Epoch: 400\tTrain loss: 39.682346725463866\n",
      "Epoch: 450\tTrain loss: 41.81349754333496\n",
      "Epoch: 500\tTrain loss: 39.341467666625974\n",
      "Epoch: 550\tTrain loss: 36.52610912322998\n",
      "Epoch: 600\tTrain loss: 38.15698165893555\n",
      "Epoch: 650\tTrain loss: 36.74868049621582\n",
      "Epoch: 700\tTrain loss: 36.510972023010254\n",
      "Epoch: 750\tTrain loss: 40.37860336303711\n",
      "Epoch: 800\tTrain loss: 35.14694061279297\n",
      "Epoch: 850\tTrain loss: 51.8945499420166\n",
      "Epoch: 900\tTrain loss: 34.710107421875\n",
      "Epoch: 950\tTrain loss: 38.86952629089355\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 150.29981155395507\n",
      "Epoch: 50\tTrain loss: 85.18797760009765\n",
      "Epoch: 100\tTrain loss: 81.62702941894531\n",
      "Epoch: 150\tTrain loss: 77.94078063964844\n",
      "Epoch: 200\tTrain loss: 73.22370529174805\n",
      "Epoch: 250\tTrain loss: 71.82431640625\n",
      "Epoch: 300\tTrain loss: 69.2010425567627\n",
      "Epoch: 350\tTrain loss: 65.98355140686036\n",
      "Epoch: 400\tTrain loss: 64.69097557067872\n",
      "Epoch: 450\tTrain loss: 67.55713157653808\n",
      "Epoch: 500\tTrain loss: 64.10452041625976\n",
      "Epoch: 550\tTrain loss: 62.295197296142575\n",
      "Epoch: 600\tTrain loss: 64.76053199768066\n",
      "Epoch: 650\tTrain loss: 59.9636646270752\n",
      "Epoch: 700\tTrain loss: 60.6743221282959\n",
      "Epoch: 750\tTrain loss: 60.08495750427246\n",
      "Epoch: 800\tTrain loss: 63.9869800567627\n",
      "Epoch: 850\tTrain loss: 69.44180145263672\n",
      "Epoch: 900\tTrain loss: 62.55828857421875\n",
      "Epoch: 950\tTrain loss: 64.76601600646973\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 86.173486328125\n",
      "Epoch: 50\tTrain loss: 12.269222831726074\n",
      "Epoch: 100\tTrain loss: 11.505187034606934\n",
      "Epoch: 150\tTrain loss: 10.981892585754395\n",
      "Epoch: 200\tTrain loss: 10.285229682922363\n",
      "Epoch: 250\tTrain loss: 9.283590841293336\n",
      "Epoch: 300\tTrain loss: 9.389542388916016\n",
      "Epoch: 350\tTrain loss: 8.694770812988281\n",
      "Epoch: 400\tTrain loss: 8.734269142150879\n",
      "Epoch: 450\tTrain loss: 7.881217956542969\n",
      "Epoch: 500\tTrain loss: 7.76573805809021\n",
      "Epoch: 550\tTrain loss: 7.977513599395752\n",
      "Epoch: 600\tTrain loss: 7.506833505630493\n",
      "Epoch: 650\tTrain loss: 6.883513355255127\n",
      "Epoch: 700\tTrain loss: 6.97816014289856\n",
      "Epoch: 750\tTrain loss: 7.819342231750488\n",
      "Epoch: 800\tTrain loss: 9.31094570159912\n",
      "Epoch: 850\tTrain loss: 6.31754789352417\n",
      "Epoch: 900\tTrain loss: 8.122107934951782\n",
      "Epoch: 950\tTrain loss: 6.534470891952514\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 97.9305046081543\n",
      "Epoch: 50\tTrain loss: 22.322773933410645\n",
      "Epoch: 100\tTrain loss: 21.496275329589842\n",
      "Epoch: 150\tTrain loss: 21.026865768432618\n",
      "Epoch: 200\tTrain loss: 20.200413513183594\n",
      "Epoch: 250\tTrain loss: 19.779405403137208\n",
      "Epoch: 300\tTrain loss: 18.66576232910156\n",
      "Epoch: 350\tTrain loss: 17.96215190887451\n",
      "Epoch: 400\tTrain loss: 17.764425468444824\n",
      "Epoch: 450\tTrain loss: 16.792273139953615\n",
      "Epoch: 500\tTrain loss: 19.330153274536134\n",
      "Epoch: 550\tTrain loss: 17.29500494003296\n",
      "Epoch: 600\tTrain loss: 16.00704698562622\n",
      "Epoch: 650\tTrain loss: 15.977263641357421\n",
      "Epoch: 700\tTrain loss: 15.616984176635743\n",
      "Epoch: 750\tTrain loss: 16.36334991455078\n",
      "Epoch: 800\tTrain loss: 16.11426000595093\n",
      "Epoch: 850\tTrain loss: 16.199789237976074\n",
      "Epoch: 900\tTrain loss: 17.19579620361328\n",
      "Epoch: 950\tTrain loss: 15.501418209075927\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 112.49331665039062\n",
      "Epoch: 50\tTrain loss: 36.94883003234863\n",
      "Epoch: 100\tTrain loss: 34.42451648712158\n",
      "Epoch: 150\tTrain loss: 31.164532279968263\n",
      "Epoch: 200\tTrain loss: 29.544715881347656\n",
      "Epoch: 250\tTrain loss: 29.009761619567872\n",
      "Epoch: 300\tTrain loss: 26.37386875152588\n",
      "Epoch: 350\tTrain loss: 26.07402629852295\n",
      "Epoch: 400\tTrain loss: 25.056727790832518\n",
      "Epoch: 450\tTrain loss: 26.284784698486327\n",
      "Epoch: 500\tTrain loss: 24.24788303375244\n",
      "Epoch: 550\tTrain loss: 23.7326171875\n",
      "Epoch: 600\tTrain loss: 28.731057739257814\n",
      "Epoch: 650\tTrain loss: 25.464000511169434\n",
      "Epoch: 700\tTrain loss: 24.83842945098877\n",
      "Epoch: 750\tTrain loss: 23.427857971191408\n",
      "Epoch: 800\tTrain loss: 24.352308464050292\n",
      "Epoch: 850\tTrain loss: 22.105141639709473\n",
      "Epoch: 900\tTrain loss: 23.726190757751464\n",
      "Epoch: 950\tTrain loss: 33.44319801330566\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 127.73859481811523\n",
      "Epoch: 50\tTrain loss: 57.86774940490723\n",
      "Epoch: 100\tTrain loss: 54.30822868347168\n",
      "Epoch: 150\tTrain loss: 52.72433433532715\n",
      "Epoch: 200\tTrain loss: 51.03073844909668\n",
      "Epoch: 250\tTrain loss: 50.99278450012207\n",
      "Epoch: 300\tTrain loss: 51.70597877502441\n",
      "Epoch: 350\tTrain loss: 47.409484481811525\n",
      "Epoch: 400\tTrain loss: 47.51112442016601\n",
      "Epoch: 450\tTrain loss: 46.14914894104004\n",
      "Epoch: 500\tTrain loss: 45.31324424743652\n",
      "Epoch: 550\tTrain loss: 45.4570125579834\n",
      "Epoch: 600\tTrain loss: 43.05292015075683\n",
      "Epoch: 650\tTrain loss: 47.32657737731934\n",
      "Epoch: 700\tTrain loss: 49.41196212768555\n",
      "Epoch: 750\tTrain loss: 45.09057121276855\n",
      "Epoch: 800\tTrain loss: 43.491957092285155\n",
      "Epoch: 850\tTrain loss: 44.41350326538086\n",
      "Epoch: 900\tTrain loss: 43.14528846740723\n",
      "Epoch: 950\tTrain loss: 43.09449920654297\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 157.5089080810547\n",
      "Epoch: 50\tTrain loss: 85.27504043579101\n",
      "Epoch: 100\tTrain loss: 80.03738327026367\n",
      "Epoch: 150\tTrain loss: 75.02909851074219\n",
      "Epoch: 200\tTrain loss: 71.5799560546875\n",
      "Epoch: 250\tTrain loss: 80.70703735351563\n",
      "Epoch: 300\tTrain loss: 65.28086433410644\n",
      "Epoch: 350\tTrain loss: 65.07732734680175\n",
      "Epoch: 400\tTrain loss: 61.66347694396973\n",
      "Epoch: 450\tTrain loss: 60.05321998596192\n",
      "Epoch: 500\tTrain loss: 62.95950469970703\n",
      "Epoch: 550\tTrain loss: 59.71402816772461\n",
      "Epoch: 600\tTrain loss: 57.32377815246582\n",
      "Epoch: 650\tTrain loss: 68.3594253540039\n",
      "Epoch: 700\tTrain loss: 54.2302864074707\n",
      "Epoch: 750\tTrain loss: 58.12089538574219\n",
      "Epoch: 800\tTrain loss: 53.966177749633786\n",
      "Epoch: 850\tTrain loss: 57.0029052734375\n",
      "Epoch: 900\tTrain loss: 56.334313201904294\n",
      "Epoch: 950\tTrain loss: 52.03715133666992\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 84.78298416137696\n",
      "Epoch: 50\tTrain loss: 12.342616558074951\n",
      "Epoch: 100\tTrain loss: 11.281159687042237\n",
      "Epoch: 150\tTrain loss: 10.534625053405762\n",
      "Epoch: 200\tTrain loss: 9.693261432647706\n",
      "Epoch: 250\tTrain loss: 9.243228244781495\n",
      "Epoch: 300\tTrain loss: 8.766589069366455\n",
      "Epoch: 350\tTrain loss: 8.088664340972901\n",
      "Epoch: 400\tTrain loss: 9.057122993469239\n",
      "Epoch: 450\tTrain loss: 8.350885772705078\n",
      "Epoch: 500\tTrain loss: 7.449875259399414\n",
      "Epoch: 550\tTrain loss: 7.7396680355072025\n",
      "Epoch: 600\tTrain loss: 7.406132078170776\n",
      "Epoch: 650\tTrain loss: 11.164901065826417\n",
      "Epoch: 700\tTrain loss: 7.76407036781311\n",
      "Epoch: 750\tTrain loss: 6.626983404159546\n",
      "Epoch: 800\tTrain loss: 7.607175827026367\n",
      "Epoch: 850\tTrain loss: 7.114658689498901\n",
      "Epoch: 900\tTrain loss: 6.44792799949646\n",
      "Epoch: 950\tTrain loss: 7.262828826904297\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 94.84661331176758\n",
      "Epoch: 50\tTrain loss: 21.809113121032716\n",
      "Epoch: 100\tTrain loss: 20.80772171020508\n",
      "Epoch: 150\tTrain loss: 19.565254783630373\n",
      "Epoch: 200\tTrain loss: 18.577206802368163\n",
      "Epoch: 250\tTrain loss: 17.704934883117676\n",
      "Epoch: 300\tTrain loss: 17.120372772216797\n",
      "Epoch: 350\tTrain loss: 17.311895179748536\n",
      "Epoch: 400\tTrain loss: 16.080688095092775\n",
      "Epoch: 450\tTrain loss: 15.364257526397704\n",
      "Epoch: 500\tTrain loss: 15.06804895401001\n",
      "Epoch: 550\tTrain loss: 15.268016910552978\n",
      "Epoch: 600\tTrain loss: 15.058371353149415\n",
      "Epoch: 650\tTrain loss: 15.495054054260255\n",
      "Epoch: 700\tTrain loss: 14.845393943786622\n",
      "Epoch: 750\tTrain loss: 14.289745807647705\n",
      "Epoch: 800\tTrain loss: 14.875080394744874\n",
      "Epoch: 850\tTrain loss: 15.021088981628418\n",
      "Epoch: 900\tTrain loss: 13.871181297302247\n",
      "Epoch: 950\tTrain loss: 13.520666313171386\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 105.15418701171875\n",
      "Epoch: 50\tTrain loss: 35.55476837158203\n",
      "Epoch: 100\tTrain loss: 32.67681522369385\n",
      "Epoch: 150\tTrain loss: 30.196973037719726\n",
      "Epoch: 200\tTrain loss: 28.674854278564453\n",
      "Epoch: 250\tTrain loss: 27.431928062438963\n",
      "Epoch: 300\tTrain loss: 26.412607955932618\n",
      "Epoch: 350\tTrain loss: 27.57172336578369\n",
      "Epoch: 400\tTrain loss: 26.080477142333983\n",
      "Epoch: 450\tTrain loss: 25.62244930267334\n",
      "Epoch: 500\tTrain loss: 25.132897758483885\n",
      "Epoch: 550\tTrain loss: 24.346748542785644\n",
      "Epoch: 600\tTrain loss: 24.3595760345459\n",
      "Epoch: 650\tTrain loss: 24.153165435791017\n",
      "Epoch: 700\tTrain loss: 25.583980751037597\n",
      "Epoch: 750\tTrain loss: 23.055416107177734\n",
      "Epoch: 800\tTrain loss: 23.73332061767578\n",
      "Epoch: 850\tTrain loss: 31.287969017028807\n",
      "Epoch: 900\tTrain loss: 23.107582092285156\n",
      "Epoch: 950\tTrain loss: 24.10624942779541\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 131.68669509887695\n",
      "Epoch: 50\tTrain loss: 57.00568618774414\n",
      "Epoch: 100\tTrain loss: 53.13820343017578\n",
      "Epoch: 150\tTrain loss: 50.69826965332031\n",
      "Epoch: 200\tTrain loss: 46.41366767883301\n",
      "Epoch: 250\tTrain loss: 45.33090782165527\n",
      "Epoch: 300\tTrain loss: 42.847934341430665\n",
      "Epoch: 350\tTrain loss: 40.7726936340332\n",
      "Epoch: 400\tTrain loss: 40.179351806640625\n",
      "Epoch: 450\tTrain loss: 40.16936721801758\n",
      "Epoch: 500\tTrain loss: 45.250087356567384\n",
      "Epoch: 550\tTrain loss: 44.208642196655276\n",
      "Epoch: 600\tTrain loss: 37.293785095214844\n",
      "Epoch: 650\tTrain loss: 39.0457145690918\n",
      "Epoch: 700\tTrain loss: 36.78661003112793\n",
      "Epoch: 750\tTrain loss: 35.6746732711792\n",
      "Epoch: 800\tTrain loss: 50.601179885864255\n",
      "Epoch: 850\tTrain loss: 41.17609405517578\n",
      "Epoch: 900\tTrain loss: 37.09422302246094\n",
      "Epoch: 950\tTrain loss: 44.19507102966308\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 152.83582916259766\n",
      "Epoch: 50\tTrain loss: 83.65989227294922\n",
      "Epoch: 100\tTrain loss: 80.13255615234375\n",
      "Epoch: 150\tTrain loss: 76.67098999023438\n",
      "Epoch: 200\tTrain loss: 74.3056510925293\n",
      "Epoch: 250\tTrain loss: 71.01979904174804\n",
      "Epoch: 300\tTrain loss: 74.41039123535157\n",
      "Epoch: 350\tTrain loss: 69.69670181274414\n",
      "Epoch: 400\tTrain loss: 65.48036308288575\n",
      "Epoch: 450\tTrain loss: 63.771248245239256\n",
      "Epoch: 500\tTrain loss: 64.88014144897461\n",
      "Epoch: 550\tTrain loss: 67.1743350982666\n",
      "Epoch: 600\tTrain loss: 62.335418701171875\n",
      "Epoch: 650\tTrain loss: 69.85305938720703\n",
      "Epoch: 700\tTrain loss: 61.67037734985352\n",
      "Epoch: 750\tTrain loss: 62.46879653930664\n",
      "Epoch: 800\tTrain loss: 66.61092834472656\n",
      "Epoch: 850\tTrain loss: 59.62632713317871\n",
      "Epoch: 900\tTrain loss: 57.42679214477539\n",
      "Epoch: 950\tTrain loss: 63.0329086303711\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 84.8447036743164\n",
      "Epoch: 50\tTrain loss: 12.792031383514404\n",
      "Epoch: 100\tTrain loss: 12.162845611572266\n",
      "Epoch: 150\tTrain loss: 11.536083984375\n",
      "Epoch: 200\tTrain loss: 10.36197738647461\n",
      "Epoch: 250\tTrain loss: 9.360900020599365\n",
      "Epoch: 300\tTrain loss: 8.524717569351196\n",
      "Epoch: 350\tTrain loss: 8.107394647598266\n",
      "Epoch: 400\tTrain loss: 8.155165100097657\n",
      "Epoch: 450\tTrain loss: 7.202952718734741\n",
      "Epoch: 500\tTrain loss: 7.473969507217407\n",
      "Epoch: 550\tTrain loss: 6.908097743988037\n",
      "Epoch: 600\tTrain loss: 7.491472291946411\n",
      "Epoch: 650\tTrain loss: 11.629153156280518\n",
      "Epoch: 700\tTrain loss: 6.664625215530395\n",
      "Epoch: 750\tTrain loss: 6.344875192642212\n",
      "Epoch: 800\tTrain loss: 6.293830251693725\n",
      "Epoch: 850\tTrain loss: 7.452976036071777\n",
      "Epoch: 900\tTrain loss: 6.16534686088562\n",
      "Epoch: 950\tTrain loss: 8.17918882369995\n",
      "Epoch: 0\tNormalisation loss: 1.3267759084701538\n",
      "Epoch: 100\tNormalisation loss: 1.143422108888626\n",
      "Epoch: 200\tNormalisation loss: 0.9582558929920196\n",
      "Epoch: 300\tNormalisation loss: 0.8961601912975311\n",
      "Epoch: 400\tNormalisation loss: 0.8456060588359833\n",
      "Epoch: 500\tNormalisation loss: 0.7918730616569519\n",
      "Epoch: 600\tNormalisation loss: 1.1918627738952636\n",
      "Epoch: 700\tNormalisation loss: 0.9014257788658142\n",
      "Epoch: 800\tNormalisation loss: 0.8393739938735962\n",
      "Epoch: 900\tNormalisation loss: 0.7944740891456604\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 101.34948348999023\n",
      "Epoch: 50\tTrain loss: 22.343124008178712\n",
      "Epoch: 100\tTrain loss: 21.765906143188477\n",
      "Epoch: 150\tTrain loss: 20.784116554260255\n",
      "Epoch: 200\tTrain loss: 21.19084014892578\n",
      "Epoch: 250\tTrain loss: 20.464911270141602\n",
      "Epoch: 300\tTrain loss: 18.571663188934327\n",
      "Epoch: 350\tTrain loss: 17.587533283233643\n",
      "Epoch: 400\tTrain loss: 17.077655506134032\n",
      "Epoch: 450\tTrain loss: 16.58853359222412\n",
      "Epoch: 500\tTrain loss: 16.363551521301268\n",
      "Epoch: 550\tTrain loss: 15.606403064727782\n",
      "Epoch: 600\tTrain loss: 14.670666408538818\n",
      "Epoch: 650\tTrain loss: 14.970314025878906\n",
      "Epoch: 700\tTrain loss: 14.129783058166504\n",
      "Epoch: 750\tTrain loss: 13.652624607086182\n",
      "Epoch: 800\tTrain loss: 13.745926189422608\n",
      "Epoch: 850\tTrain loss: 14.614752674102784\n",
      "Epoch: 900\tTrain loss: 14.608681774139404\n",
      "Epoch: 950\tTrain loss: 15.249412250518798\n",
      "Epoch: 0\tNormalisation loss: 1.6420991897583008\n",
      "Epoch: 100\tNormalisation loss: 0.9908526301383972\n",
      "Epoch: 200\tNormalisation loss: 0.9238927125930786\n",
      "Epoch: 300\tNormalisation loss: 0.8630164325237274\n",
      "Epoch: 400\tNormalisation loss: 0.802228432893753\n",
      "Epoch: 500\tNormalisation loss: 0.7335247457027435\n",
      "Epoch: 600\tNormalisation loss: 0.6856681108474731\n",
      "Epoch: 700\tNormalisation loss: 0.645414012670517\n",
      "Epoch: 800\tNormalisation loss: 1.039108669757843\n",
      "Epoch: 900\tNormalisation loss: 0.9409434735774994\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 111.54848403930664\n",
      "Epoch: 50\tTrain loss: 38.42831993103027\n",
      "Epoch: 100\tTrain loss: 36.721792984008786\n",
      "Epoch: 150\tTrain loss: 34.42462711334228\n",
      "Epoch: 200\tTrain loss: 33.851576232910155\n",
      "Epoch: 250\tTrain loss: 31.083294486999513\n",
      "Epoch: 300\tTrain loss: 30.254301643371583\n",
      "Epoch: 350\tTrain loss: 28.847883033752442\n",
      "Epoch: 400\tTrain loss: 27.742107009887697\n",
      "Epoch: 450\tTrain loss: 26.68309841156006\n",
      "Epoch: 500\tTrain loss: 28.057644081115722\n",
      "Epoch: 550\tTrain loss: 25.889114570617675\n",
      "Epoch: 600\tTrain loss: 34.064266967773435\n",
      "Epoch: 650\tTrain loss: 27.378387641906738\n",
      "Epoch: 700\tTrain loss: 27.608916091918946\n",
      "Epoch: 750\tTrain loss: 26.7553373336792\n",
      "Epoch: 800\tTrain loss: 27.220262718200683\n",
      "Epoch: 850\tTrain loss: 25.467073059082033\n",
      "Epoch: 900\tTrain loss: 25.17544651031494\n",
      "Epoch: 950\tTrain loss: 25.08575572967529\n",
      "Epoch: 0\tNormalisation loss: 1.9702653884887695\n",
      "Epoch: 100\tNormalisation loss: 1.01887686252594\n",
      "Epoch: 200\tNormalisation loss: 0.9477696061134339\n",
      "Epoch: 300\tNormalisation loss: 0.8928329169750213\n",
      "Epoch: 400\tNormalisation loss: 0.8411316037178039\n",
      "Epoch: 500\tNormalisation loss: 0.7960816740989685\n",
      "Epoch: 600\tNormalisation loss: 1.280767011642456\n",
      "Epoch: 700\tNormalisation loss: 0.9691754758358002\n",
      "Epoch: 800\tNormalisation loss: 0.9072991013526917\n",
      "Epoch: 900\tNormalisation loss: 0.8633798837661744\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 131.32340774536132\n",
      "Epoch: 50\tTrain loss: 55.198051071166994\n",
      "Epoch: 100\tTrain loss: 52.0201961517334\n",
      "Epoch: 150\tTrain loss: 49.54126396179199\n",
      "Epoch: 200\tTrain loss: 47.470816421508786\n",
      "Epoch: 250\tTrain loss: 45.64619331359863\n",
      "Epoch: 300\tTrain loss: 43.882241821289064\n",
      "Epoch: 350\tTrain loss: 43.84517402648926\n",
      "Epoch: 400\tTrain loss: 44.00627670288086\n",
      "Epoch: 450\tTrain loss: 40.65523071289063\n",
      "Epoch: 500\tTrain loss: 42.296831130981445\n",
      "Epoch: 550\tTrain loss: 41.52278861999512\n",
      "Epoch: 600\tTrain loss: 40.914622116088864\n",
      "Epoch: 650\tTrain loss: 39.8517822265625\n",
      "Epoch: 700\tTrain loss: 40.8651611328125\n",
      "Epoch: 750\tTrain loss: 54.03617706298828\n",
      "Epoch: 800\tTrain loss: 41.77549171447754\n",
      "Epoch: 850\tTrain loss: 42.24918746948242\n",
      "Epoch: 900\tTrain loss: 40.448251724243164\n",
      "Epoch: 950\tTrain loss: 36.853488731384275\n",
      "Epoch: 0\tNormalisation loss: 3.006563115119934\n",
      "Epoch: 100\tNormalisation loss: 1.0467230379581451\n",
      "Epoch: 200\tNormalisation loss: 0.9874194502830506\n",
      "Epoch: 300\tNormalisation loss: 0.943942254781723\n",
      "Epoch: 400\tNormalisation loss: 0.9027596771717071\n",
      "Epoch: 500\tNormalisation loss: 1.186933594942093\n",
      "Epoch: 600\tNormalisation loss: 1.0336638987064362\n",
      "Epoch: 700\tNormalisation loss: 0.9528011977672577\n",
      "Epoch: 800\tNormalisation loss: 0.896102774143219\n",
      "Epoch: 900\tNormalisation loss: 0.8538949429988861\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 158.8522750854492\n",
      "Epoch: 50\tTrain loss: 86.88177032470703\n",
      "Epoch: 100\tTrain loss: 84.20453720092773\n",
      "Epoch: 150\tTrain loss: 80.49743347167968\n",
      "Epoch: 200\tTrain loss: 78.1982292175293\n",
      "Epoch: 250\tTrain loss: 76.43697891235351\n",
      "Epoch: 300\tTrain loss: 73.91760025024413\n",
      "Epoch: 350\tTrain loss: 72.74342269897461\n",
      "Epoch: 400\tTrain loss: 78.41397705078126\n",
      "Epoch: 450\tTrain loss: 80.58158340454102\n",
      "Epoch: 500\tTrain loss: 71.21786651611328\n",
      "Epoch: 550\tTrain loss: 70.81235847473144\n",
      "Epoch: 600\tTrain loss: 71.3131103515625\n",
      "Epoch: 650\tTrain loss: 70.26103782653809\n",
      "Epoch: 700\tTrain loss: 72.0260627746582\n",
      "Epoch: 750\tTrain loss: 66.71834182739258\n",
      "Epoch: 800\tTrain loss: 71.60065307617188\n",
      "Epoch: 850\tTrain loss: 69.5692970275879\n",
      "Epoch: 900\tTrain loss: 67.16917533874512\n",
      "Epoch: 950\tTrain loss: 65.87801132202148\n",
      "Epoch: 0\tNormalisation loss: 3.374381947517395\n",
      "Epoch: 100\tNormalisation loss: 0.9826013565063476\n",
      "Epoch: 200\tNormalisation loss: 1.134387230873108\n",
      "Epoch: 300\tNormalisation loss: 1.0658822238445282\n",
      "Epoch: 400\tNormalisation loss: 1.0267245709896087\n",
      "Epoch: 500\tNormalisation loss: 0.9932338535785675\n",
      "Epoch: 600\tNormalisation loss: 0.9602195620536804\n",
      "Epoch: 700\tNormalisation loss: 0.920594084262848\n",
      "Epoch: 800\tNormalisation loss: 0.8813749253749847\n",
      "Epoch: 900\tNormalisation loss: 0.846530431509018\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 81.69780960083008\n",
      "Epoch: 50\tTrain loss: 12.80743055343628\n",
      "Epoch: 100\tTrain loss: 12.28528881072998\n",
      "Epoch: 150\tTrain loss: 11.617803859710694\n",
      "Epoch: 200\tTrain loss: 10.836398029327393\n",
      "Epoch: 250\tTrain loss: 10.021251106262207\n",
      "Epoch: 300\tTrain loss: 9.510063934326173\n",
      "Epoch: 350\tTrain loss: 9.001793432235718\n",
      "Epoch: 400\tTrain loss: 8.491766834259034\n",
      "Epoch: 450\tTrain loss: 8.14349422454834\n",
      "Epoch: 500\tTrain loss: 11.049880886077881\n",
      "Epoch: 550\tTrain loss: 8.213187408447265\n",
      "Epoch: 600\tTrain loss: 8.28749942779541\n",
      "Epoch: 650\tTrain loss: 8.358230304718017\n",
      "Epoch: 700\tTrain loss: 8.182042551040649\n",
      "Epoch: 750\tTrain loss: 7.456598949432373\n",
      "Epoch: 800\tTrain loss: 8.657089853286744\n",
      "Epoch: 850\tTrain loss: 7.222389173507691\n",
      "Epoch: 900\tTrain loss: 8.623604822158814\n",
      "Epoch: 950\tTrain loss: 8.204286003112793\n",
      "Epoch: 0\tNormalisation loss: 1.5291688919067383\n",
      "Epoch: 100\tNormalisation loss: 1.0038879990577698\n",
      "Epoch: 200\tNormalisation loss: 0.9328949868679046\n",
      "Epoch: 300\tNormalisation loss: 0.8822942078113556\n",
      "Epoch: 400\tNormalisation loss: 0.8344945430755615\n",
      "Epoch: 500\tNormalisation loss: 0.7888985753059388\n",
      "Epoch: 600\tNormalisation loss: 1.179774236679077\n",
      "Epoch: 700\tNormalisation loss: 0.9417408108711243\n",
      "Epoch: 800\tNormalisation loss: 0.8524836659431457\n",
      "Epoch: 900\tNormalisation loss: 0.7919414520263672\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 96.71445159912109\n",
      "Epoch: 50\tTrain loss: 22.04219341278076\n",
      "Epoch: 100\tTrain loss: 20.787553215026854\n",
      "Epoch: 150\tTrain loss: 19.23086795806885\n",
      "Epoch: 200\tTrain loss: 18.607622718811037\n",
      "Epoch: 250\tTrain loss: 18.449608612060548\n",
      "Epoch: 300\tTrain loss: 17.897996616363525\n",
      "Epoch: 350\tTrain loss: 17.284758377075196\n",
      "Epoch: 400\tTrain loss: 16.55690221786499\n",
      "Epoch: 450\tTrain loss: 16.183551216125487\n",
      "Epoch: 500\tTrain loss: 15.49037275314331\n",
      "Epoch: 550\tTrain loss: 15.619722461700439\n",
      "Epoch: 600\tTrain loss: 14.666920375823974\n",
      "Epoch: 650\tTrain loss: 14.588118839263917\n",
      "Epoch: 700\tTrain loss: 15.328056335449219\n",
      "Epoch: 750\tTrain loss: 14.586916828155518\n",
      "Epoch: 800\tTrain loss: 14.173543071746826\n",
      "Epoch: 850\tTrain loss: 15.506175327301026\n",
      "Epoch: 900\tTrain loss: 14.238847064971925\n",
      "Epoch: 950\tTrain loss: 14.817945957183838\n",
      "Epoch: 0\tNormalisation loss: 1.4255507826805114\n",
      "Epoch: 100\tNormalisation loss: 0.9534731388092041\n",
      "Epoch: 200\tNormalisation loss: 0.8941116571426392\n",
      "Epoch: 300\tNormalisation loss: 0.8483847856521607\n",
      "Epoch: 400\tNormalisation loss: 0.807311475276947\n",
      "Epoch: 500\tNormalisation loss: 1.221135926246643\n",
      "Epoch: 600\tNormalisation loss: 0.978366243839264\n",
      "Epoch: 700\tNormalisation loss: 0.9146193504333496\n",
      "Epoch: 800\tNormalisation loss: 0.864894562959671\n",
      "Epoch: 900\tNormalisation loss: 0.822322553396225\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 109.920654296875\n",
      "Epoch: 50\tTrain loss: 36.79039421081543\n",
      "Epoch: 100\tTrain loss: 34.90499572753906\n",
      "Epoch: 150\tTrain loss: 33.609731101989745\n",
      "Epoch: 200\tTrain loss: 32.4336877822876\n",
      "Epoch: 250\tTrain loss: 30.7672758102417\n",
      "Epoch: 300\tTrain loss: 32.71134243011475\n",
      "Epoch: 350\tTrain loss: 31.110104942321776\n",
      "Epoch: 400\tTrain loss: 30.2500638961792\n",
      "Epoch: 450\tTrain loss: 30.676999473571776\n",
      "Epoch: 500\tTrain loss: 29.454038429260255\n",
      "Epoch: 550\tTrain loss: 29.522528457641602\n",
      "Epoch: 600\tTrain loss: 29.349404335021973\n",
      "Epoch: 650\tTrain loss: 29.197266387939454\n",
      "Epoch: 700\tTrain loss: 27.040664863586425\n",
      "Epoch: 750\tTrain loss: 32.10186405181885\n",
      "Epoch: 800\tTrain loss: 27.449439239501952\n",
      "Epoch: 850\tTrain loss: 30.998060417175292\n",
      "Epoch: 900\tTrain loss: 28.739632415771485\n",
      "Epoch: 950\tTrain loss: 29.840588760375976\n",
      "Epoch: 0\tNormalisation loss: 2.08091858625412\n",
      "Epoch: 100\tNormalisation loss: 0.9521580278873444\n",
      "Epoch: 200\tNormalisation loss: 0.8881326377391815\n",
      "Epoch: 300\tNormalisation loss: 0.8386759757995605\n",
      "Epoch: 400\tNormalisation loss: 0.7968529343605042\n",
      "Epoch: 500\tNormalisation loss: 0.7596980631351471\n",
      "Epoch: 600\tNormalisation loss: 0.7227226495742798\n",
      "Epoch: 700\tNormalisation loss: 0.6854663729667664\n",
      "Epoch: 800\tNormalisation loss: 1.087339061498642\n",
      "Epoch: 900\tNormalisation loss: 0.8692574083805085\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 130.52799530029296\n",
      "Epoch: 50\tTrain loss: 58.60606994628906\n",
      "Epoch: 100\tTrain loss: 54.745849227905275\n",
      "Epoch: 150\tTrain loss: 51.6585578918457\n",
      "Epoch: 200\tTrain loss: 49.01286239624024\n",
      "Epoch: 250\tTrain loss: 46.45946426391602\n",
      "Epoch: 300\tTrain loss: 51.822593688964844\n",
      "Epoch: 350\tTrain loss: 49.173594284057614\n",
      "Epoch: 400\tTrain loss: 46.91705055236817\n",
      "Epoch: 450\tTrain loss: 44.41569213867187\n",
      "Epoch: 500\tTrain loss: 43.960120010375974\n",
      "Epoch: 550\tTrain loss: 44.148668670654295\n",
      "Epoch: 600\tTrain loss: 47.550000762939455\n",
      "Epoch: 650\tTrain loss: 47.44259071350098\n",
      "Epoch: 700\tTrain loss: 41.085443115234376\n",
      "Epoch: 750\tTrain loss: 42.69338188171387\n",
      "Epoch: 800\tTrain loss: 43.037728500366214\n",
      "Epoch: 850\tTrain loss: 46.286447525024414\n",
      "Epoch: 900\tTrain loss: 42.79905395507812\n",
      "Epoch: 950\tTrain loss: 43.82791175842285\n",
      "Epoch: 0\tNormalisation loss: 2.902526879310608\n",
      "Epoch: 100\tNormalisation loss: 0.9423612773418426\n",
      "Epoch: 200\tNormalisation loss: 0.8740255653858184\n",
      "Epoch: 300\tNormalisation loss: 1.194383442401886\n",
      "Epoch: 400\tNormalisation loss: 0.9872547805309295\n",
      "Epoch: 500\tNormalisation loss: 0.9396163344383239\n",
      "Epoch: 600\tNormalisation loss: 0.8950032711029052\n",
      "Epoch: 700\tNormalisation loss: 0.8508612275123596\n",
      "Epoch: 800\tNormalisation loss: 0.810306578874588\n",
      "Epoch: 900\tNormalisation loss: 0.7707054376602173\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 155.28406219482423\n",
      "Epoch: 50\tTrain loss: 85.09510650634766\n",
      "Epoch: 100\tTrain loss: 81.64310455322266\n",
      "Epoch: 150\tTrain loss: 75.05121459960938\n",
      "Epoch: 200\tTrain loss: 70.50174102783203\n",
      "Epoch: 250\tTrain loss: 70.4432918548584\n",
      "Epoch: 300\tTrain loss: 64.43350067138672\n",
      "Epoch: 350\tTrain loss: 66.69871101379394\n",
      "Epoch: 400\tTrain loss: 61.81775245666504\n",
      "Epoch: 450\tTrain loss: 59.49547843933105\n",
      "Epoch: 500\tTrain loss: 59.8012939453125\n",
      "Epoch: 550\tTrain loss: 57.697017669677734\n",
      "Epoch: 600\tTrain loss: 56.46034660339355\n",
      "Epoch: 650\tTrain loss: 56.667021942138675\n",
      "Epoch: 700\tTrain loss: 56.35287742614746\n",
      "Epoch: 750\tTrain loss: 54.62410202026367\n",
      "Epoch: 800\tTrain loss: 55.51843795776367\n",
      "Epoch: 850\tTrain loss: 56.21785659790039\n",
      "Epoch: 900\tTrain loss: 58.30414352416992\n",
      "Epoch: 950\tTrain loss: 60.67893486022949\n",
      "Epoch: 0\tNormalisation loss: 3.247756028175354\n",
      "Epoch: 100\tNormalisation loss: 0.9370633959770203\n",
      "Epoch: 200\tNormalisation loss: 1.1745215117931367\n",
      "Epoch: 300\tNormalisation loss: 0.9484146535396576\n",
      "Epoch: 400\tNormalisation loss: 0.8915863394737243\n",
      "Epoch: 500\tNormalisation loss: 0.8465322852134705\n",
      "Epoch: 600\tNormalisation loss: 0.8055080056190491\n",
      "Epoch: 700\tNormalisation loss: 0.9534104347229004\n",
      "Epoch: 800\tNormalisation loss: 0.8986787557601928\n",
      "Epoch: 900\tNormalisation loss: 0.8603480339050293\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 80.70360488891602\n",
      "Epoch: 50\tTrain loss: 12.808109760284424\n",
      "Epoch: 100\tTrain loss: 12.310794162750245\n",
      "Epoch: 150\tTrain loss: 11.807731437683106\n",
      "Epoch: 200\tTrain loss: 11.376515960693359\n",
      "Epoch: 250\tTrain loss: 11.028923320770264\n",
      "Epoch: 300\tTrain loss: 10.320969486236573\n",
      "Epoch: 350\tTrain loss: 9.70308780670166\n",
      "Epoch: 400\tTrain loss: 9.515857601165772\n",
      "Epoch: 450\tTrain loss: 9.028320121765137\n",
      "Epoch: 500\tTrain loss: 8.467005109786987\n",
      "Epoch: 550\tTrain loss: 7.982444381713867\n",
      "Epoch: 600\tTrain loss: 8.055335569381715\n",
      "Epoch: 650\tTrain loss: 8.884514999389648\n",
      "Epoch: 700\tTrain loss: 8.013199615478516\n",
      "Epoch: 750\tTrain loss: 8.256792497634887\n",
      "Epoch: 800\tTrain loss: 8.439514780044556\n",
      "Epoch: 850\tTrain loss: 8.008050966262818\n",
      "Epoch: 900\tTrain loss: 8.919013357162475\n",
      "Epoch: 950\tTrain loss: 7.846853590011596\n",
      "Epoch: 0\tNormalisation loss: 1.415935182571411\n",
      "Epoch: 100\tNormalisation loss: 1.0567435145378112\n",
      "Epoch: 200\tNormalisation loss: 0.9963124036788941\n",
      "Epoch: 300\tNormalisation loss: 0.9318171739578247\n",
      "Epoch: 400\tNormalisation loss: 0.8678408861160278\n",
      "Epoch: 500\tNormalisation loss: 0.8142000496387481\n",
      "Epoch: 600\tNormalisation loss: 0.7667224049568176\n",
      "Epoch: 700\tNormalisation loss: 0.7300226569175721\n",
      "Epoch: 800\tNormalisation loss: 1.061500746011734\n",
      "Epoch: 900\tNormalisation loss: 0.9327688217163086\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 91.98578414916992\n",
      "Epoch: 50\tTrain loss: 21.34808235168457\n",
      "Epoch: 100\tTrain loss: 20.314129257202147\n",
      "Epoch: 150\tTrain loss: 19.29815044403076\n",
      "Epoch: 200\tTrain loss: 18.18748607635498\n",
      "Epoch: 250\tTrain loss: 17.034809589385986\n",
      "Epoch: 300\tTrain loss: 16.95942783355713\n",
      "Epoch: 350\tTrain loss: 16.4012412071228\n",
      "Epoch: 400\tTrain loss: 16.411675548553468\n",
      "Epoch: 450\tTrain loss: 16.88783941268921\n",
      "Epoch: 500\tTrain loss: 15.327233505249023\n",
      "Epoch: 550\tTrain loss: 17.617895126342773\n",
      "Epoch: 600\tTrain loss: 15.764149570465088\n",
      "Epoch: 650\tTrain loss: 14.850158309936523\n",
      "Epoch: 700\tTrain loss: 15.39426212310791\n",
      "Epoch: 750\tTrain loss: 14.91820468902588\n",
      "Epoch: 800\tTrain loss: 13.662085151672363\n",
      "Epoch: 850\tTrain loss: 13.657723522186279\n",
      "Epoch: 900\tTrain loss: 16.350941276550294\n",
      "Epoch: 950\tTrain loss: 14.249890613555909\n",
      "Epoch: 0\tNormalisation loss: 1.7868404626846313\n",
      "Epoch: 100\tNormalisation loss: 0.9899668991565704\n",
      "Epoch: 200\tNormalisation loss: 0.9182874977588653\n",
      "Epoch: 300\tNormalisation loss: 0.8630453407764435\n",
      "Epoch: 400\tNormalisation loss: 0.8170792877674102\n",
      "Epoch: 500\tNormalisation loss: 0.7766317307949067\n",
      "Epoch: 600\tNormalisation loss: 0.7340017557144165\n",
      "Epoch: 700\tNormalisation loss: 1.1206874251365662\n",
      "Epoch: 800\tNormalisation loss: 0.9800539314746857\n",
      "Epoch: 900\tNormalisation loss: 0.9295535743236542\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 107.19964828491212\n",
      "Epoch: 50\tTrain loss: 37.893201446533205\n",
      "Epoch: 100\tTrain loss: 35.733880043029785\n",
      "Epoch: 150\tTrain loss: 34.30326843261719\n",
      "Epoch: 200\tTrain loss: 32.40213527679443\n",
      "Epoch: 250\tTrain loss: 30.603710174560547\n",
      "Epoch: 300\tTrain loss: 29.094486045837403\n",
      "Epoch: 350\tTrain loss: 27.979226112365723\n",
      "Epoch: 400\tTrain loss: 29.02595806121826\n",
      "Epoch: 450\tTrain loss: 30.843321800231934\n",
      "Epoch: 500\tTrain loss: 27.690510940551757\n",
      "Epoch: 550\tTrain loss: 31.056016540527345\n",
      "Epoch: 600\tTrain loss: 25.43119087219238\n",
      "Epoch: 650\tTrain loss: 30.95918560028076\n",
      "Epoch: 700\tTrain loss: 27.671293449401855\n",
      "Epoch: 750\tTrain loss: 27.125728797912597\n",
      "Epoch: 800\tTrain loss: 24.934615325927734\n",
      "Epoch: 850\tTrain loss: 29.089559745788574\n",
      "Epoch: 900\tTrain loss: 25.915099334716796\n",
      "Epoch: 950\tTrain loss: 28.427610206604005\n",
      "Epoch: 0\tNormalisation loss: 2.3614852905273436\n",
      "Epoch: 100\tNormalisation loss: 1.0366677165031433\n",
      "Epoch: 200\tNormalisation loss: 0.9811807096004486\n",
      "Epoch: 300\tNormalisation loss: 0.934874427318573\n",
      "Epoch: 400\tNormalisation loss: 0.8958104252815247\n",
      "Epoch: 500\tNormalisation loss: 0.8567554235458374\n",
      "Epoch: 600\tNormalisation loss: 1.0034342348575591\n",
      "Epoch: 700\tNormalisation loss: 0.9410132765769958\n",
      "Epoch: 800\tNormalisation loss: 0.8937056064605713\n",
      "Epoch: 900\tNormalisation loss: 0.8470113933086395\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 131.1715835571289\n",
      "Epoch: 50\tTrain loss: 57.363517379760744\n",
      "Epoch: 100\tTrain loss: 53.1100414276123\n",
      "Epoch: 150\tTrain loss: 51.37166328430176\n",
      "Epoch: 200\tTrain loss: 48.10048561096191\n",
      "Epoch: 250\tTrain loss: 44.92236480712891\n",
      "Epoch: 300\tTrain loss: 44.18153266906738\n",
      "Epoch: 350\tTrain loss: 41.00429725646973\n",
      "Epoch: 400\tTrain loss: 39.4634334564209\n",
      "Epoch: 450\tTrain loss: 39.5370059967041\n",
      "Epoch: 500\tTrain loss: 47.552005767822266\n",
      "Epoch: 550\tTrain loss: 39.86001434326172\n",
      "Epoch: 600\tTrain loss: 35.5752628326416\n",
      "Epoch: 650\tTrain loss: 36.12265739440918\n",
      "Epoch: 700\tTrain loss: 36.74264411926269\n",
      "Epoch: 750\tTrain loss: 39.1431396484375\n",
      "Epoch: 800\tTrain loss: 34.548382949829104\n",
      "Epoch: 850\tTrain loss: 34.917837142944336\n",
      "Epoch: 900\tTrain loss: 39.07706489562988\n",
      "Epoch: 950\tTrain loss: 39.23990364074707\n",
      "Epoch: 0\tNormalisation loss: 2.5642863273620606\n",
      "Epoch: 100\tNormalisation loss: 1.080855906009674\n",
      "Epoch: 200\tNormalisation loss: 1.0196050822734832\n",
      "Epoch: 300\tNormalisation loss: 0.9720881879329681\n",
      "Epoch: 400\tNormalisation loss: 0.9299122869968415\n",
      "Epoch: 500\tNormalisation loss: 0.8893210053443908\n",
      "Epoch: 600\tNormalisation loss: 0.8492451786994935\n",
      "Epoch: 700\tNormalisation loss: 0.8098423659801484\n",
      "Epoch: 800\tNormalisation loss: 0.772295081615448\n",
      "Epoch: 900\tNormalisation loss: 1.1406767606735229\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 156.9016586303711\n",
      "Epoch: 50\tTrain loss: 84.98082275390625\n",
      "Epoch: 100\tTrain loss: 80.95439682006835\n",
      "Epoch: 150\tTrain loss: 79.02888412475586\n",
      "Epoch: 200\tTrain loss: 77.44264221191406\n",
      "Epoch: 250\tTrain loss: 75.9509780883789\n",
      "Epoch: 300\tTrain loss: 76.24576034545899\n",
      "Epoch: 350\tTrain loss: 72.21396865844727\n",
      "Epoch: 400\tTrain loss: 69.40503158569337\n",
      "Epoch: 450\tTrain loss: 69.1283950805664\n",
      "Epoch: 500\tTrain loss: 64.21777801513672\n",
      "Epoch: 550\tTrain loss: 63.580627822875975\n",
      "Epoch: 600\tTrain loss: 65.16306686401367\n",
      "Epoch: 650\tTrain loss: 67.23400688171387\n",
      "Epoch: 700\tTrain loss: 63.27817001342773\n",
      "Epoch: 750\tTrain loss: 64.33463973999024\n",
      "Epoch: 800\tTrain loss: 64.34588317871093\n",
      "Epoch: 850\tTrain loss: 65.19399375915528\n",
      "Epoch: 900\tTrain loss: 64.67699317932129\n",
      "Epoch: 950\tTrain loss: 65.42491226196289\n",
      "Epoch: 0\tNormalisation loss: 4.033884000778198\n",
      "Epoch: 100\tNormalisation loss: 0.9841663658618927\n",
      "Epoch: 200\tNormalisation loss: 0.9204454898834229\n",
      "Epoch: 300\tNormalisation loss: 0.8728857636451721\n",
      "Epoch: 400\tNormalisation loss: 0.8332749783992768\n",
      "Epoch: 500\tNormalisation loss: nan\n",
      "Epoch: 600\tNormalisation loss: nan\n",
      "Epoch: 700\tNormalisation loss: nan\n",
      "Epoch: 800\tNormalisation loss: nan\n",
      "Epoch: 900\tNormalisation loss: nan\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 86.16807479858399\n",
      "Epoch: 50\tTrain loss: 12.259513473510742\n",
      "Epoch: 100\tTrain loss: 11.379058170318604\n",
      "Epoch: 150\tTrain loss: 10.297104930877685\n",
      "Epoch: 200\tTrain loss: 9.282791090011596\n",
      "Epoch: 250\tTrain loss: 8.532085657119751\n",
      "Epoch: 300\tTrain loss: 8.779542016983033\n",
      "Epoch: 350\tTrain loss: 7.445439624786377\n",
      "Epoch: 400\tTrain loss: 7.385059213638305\n",
      "Epoch: 450\tTrain loss: 6.744339179992676\n",
      "Epoch: 500\tTrain loss: 7.140688896179199\n",
      "Epoch: 550\tTrain loss: 7.048989009857178\n",
      "Epoch: 600\tTrain loss: 6.2710120677948\n",
      "Epoch: 650\tTrain loss: 6.261430263519287\n",
      "Epoch: 700\tTrain loss: 7.317537307739258\n",
      "Epoch: 750\tTrain loss: 6.546491241455078\n",
      "Epoch: 800\tTrain loss: 6.438292074203491\n",
      "Epoch: 850\tTrain loss: 6.087098789215088\n",
      "Epoch: 900\tTrain loss: 7.399912548065186\n",
      "Epoch: 950\tTrain loss: 6.634091329574585\n",
      "Epoch: 0\tNormalisation loss: 1.3432991981506348\n",
      "Epoch: 100\tNormalisation loss: 0.8763927102088929\n",
      "Epoch: 200\tNormalisation loss: 0.8005275428295135\n",
      "Epoch: 300\tNormalisation loss: 1.0666260898113251\n",
      "Epoch: 400\tNormalisation loss: 0.9600163221359252\n",
      "Epoch: 500\tNormalisation loss: 0.8911144733428955\n",
      "Epoch: 600\tNormalisation loss: 0.8344018995761872\n",
      "Epoch: 700\tNormalisation loss: 0.7866984367370605\n",
      "Epoch: 800\tNormalisation loss: 0.7435751676559448\n",
      "Epoch: 900\tNormalisation loss: 0.7083934068679809\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 101.55772705078125\n",
      "Epoch: 50\tTrain loss: 22.212650299072266\n",
      "Epoch: 100\tTrain loss: 20.989109420776366\n",
      "Epoch: 150\tTrain loss: 19.868995666503906\n",
      "Epoch: 200\tTrain loss: 18.98191318511963\n",
      "Epoch: 250\tTrain loss: 18.28988609313965\n",
      "Epoch: 300\tTrain loss: 18.373023223876952\n",
      "Epoch: 350\tTrain loss: 17.498756980895998\n",
      "Epoch: 400\tTrain loss: 18.233312034606932\n",
      "Epoch: 450\tTrain loss: 17.4847562789917\n",
      "Epoch: 500\tTrain loss: 16.942180824279784\n",
      "Epoch: 550\tTrain loss: 16.82923069000244\n",
      "Epoch: 600\tTrain loss: 16.719596004486085\n",
      "Epoch: 650\tTrain loss: 17.73786792755127\n",
      "Epoch: 700\tTrain loss: 17.731486797332764\n",
      "Epoch: 750\tTrain loss: 15.93314504623413\n",
      "Epoch: 800\tTrain loss: 20.686806678771973\n",
      "Epoch: 850\tTrain loss: 20.922110939025877\n",
      "Epoch: 900\tTrain loss: 18.725978660583497\n",
      "Epoch: 950\tTrain loss: 18.063598251342775\n",
      "Epoch: 0\tNormalisation loss: 2.3396992921829223\n",
      "Epoch: 100\tNormalisation loss: 0.9781871259212493\n",
      "Epoch: 200\tNormalisation loss: 0.9274061024188995\n",
      "Epoch: 300\tNormalisation loss: 0.88034006357193\n",
      "Epoch: 400\tNormalisation loss: 1.0282755732536315\n",
      "Epoch: 500\tNormalisation loss: 0.9421161055564881\n",
      "Epoch: 600\tNormalisation loss: 0.8918230593204498\n",
      "Epoch: 700\tNormalisation loss: 0.8418133914470672\n",
      "Epoch: 800\tNormalisation loss: 0.7937333941459656\n",
      "Epoch: 900\tNormalisation loss: 1.0432867348194121\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 108.68727569580078\n",
      "Epoch: 50\tTrain loss: 37.854291152954104\n",
      "Epoch: 100\tTrain loss: 35.60001335144043\n",
      "Epoch: 150\tTrain loss: 32.947666549682616\n",
      "Epoch: 200\tTrain loss: 32.14724617004394\n",
      "Epoch: 250\tTrain loss: 30.04031047821045\n",
      "Epoch: 300\tTrain loss: 28.551446151733398\n",
      "Epoch: 350\tTrain loss: 28.267166137695312\n",
      "Epoch: 400\tTrain loss: 30.301368713378906\n",
      "Epoch: 450\tTrain loss: 27.17408676147461\n",
      "Epoch: 500\tTrain loss: 26.68271408081055\n",
      "Epoch: 550\tTrain loss: 25.87200183868408\n",
      "Epoch: 600\tTrain loss: 41.30018424987793\n",
      "Epoch: 650\tTrain loss: 35.194609832763675\n",
      "Epoch: 700\tTrain loss: 32.65792942047119\n",
      "Epoch: 750\tTrain loss: 31.770879554748536\n",
      "Epoch: 800\tTrain loss: 32.56878128051758\n",
      "Epoch: 850\tTrain loss: 31.276265335083007\n",
      "Epoch: 900\tTrain loss: 33.11222839355469\n",
      "Epoch: 950\tTrain loss: 30.089984321594237\n",
      "Epoch: 0\tNormalisation loss: 2.69455144405365\n",
      "Epoch: 100\tNormalisation loss: 0.9088240325450897\n",
      "Epoch: 200\tNormalisation loss: 0.8531341195106507\n",
      "Epoch: 300\tNormalisation loss: 0.8059117555618286\n",
      "Epoch: 400\tNormalisation loss: 0.7626091480255127\n",
      "Epoch: 500\tNormalisation loss: 0.7179029643535614\n",
      "Epoch: 600\tNormalisation loss: 0.982455950975418\n",
      "Epoch: 700\tNormalisation loss: 0.8969188690185547\n",
      "Epoch: 800\tNormalisation loss: 0.8421795666217804\n",
      "Epoch: 900\tNormalisation loss: 0.795195072889328\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 134.51704330444335\n",
      "Epoch: 50\tTrain loss: 58.606471633911134\n",
      "Epoch: 100\tTrain loss: 55.466775512695314\n",
      "Epoch: 150\tTrain loss: 52.611981582641604\n",
      "Epoch: 200\tTrain loss: 49.35291786193848\n",
      "Epoch: 250\tTrain loss: 52.862179565429685\n",
      "Epoch: 300\tTrain loss: 47.535278701782225\n",
      "Epoch: 350\tTrain loss: 48.03348655700684\n",
      "Epoch: 400\tTrain loss: 47.66410865783691\n",
      "Epoch: 450\tTrain loss: 48.61828384399414\n",
      "Epoch: 500\tTrain loss: 47.41182250976563\n",
      "Epoch: 550\tTrain loss: 45.99660224914551\n",
      "Epoch: 600\tTrain loss: 44.6152458190918\n",
      "Epoch: 650\tTrain loss: 44.1289119720459\n",
      "Epoch: 700\tTrain loss: 48.502933502197266\n",
      "Epoch: 750\tTrain loss: 43.6343132019043\n",
      "Epoch: 800\tTrain loss: 42.52341232299805\n",
      "Epoch: 850\tTrain loss: 43.53097267150879\n",
      "Epoch: 900\tTrain loss: 44.05606536865234\n",
      "Epoch: 950\tTrain loss: 53.957688903808595\n",
      "Epoch: 0\tNormalisation loss: 2.926976728439331\n",
      "Epoch: 100\tNormalisation loss: 0.9332593023777008\n",
      "Epoch: 200\tNormalisation loss: 0.8717446625232697\n",
      "Epoch: 300\tNormalisation loss: 0.825575053691864\n",
      "Epoch: 400\tNormalisation loss: 0.787418794631958\n",
      "Epoch: 500\tNormalisation loss: 0.7501756489276886\n",
      "Epoch: 600\tNormalisation loss: 0.7150205433368683\n",
      "Epoch: 700\tNormalisation loss: 0.6777422189712524\n",
      "Epoch: 800\tNormalisation loss: 0.6422137141227722\n",
      "Epoch: 900\tNormalisation loss: 0.9765039145946502\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 162.24438018798827\n",
      "Epoch: 50\tTrain loss: 86.07393417358398\n",
      "Epoch: 100\tTrain loss: 82.3587745666504\n",
      "Epoch: 150\tTrain loss: 79.31134414672852\n",
      "Epoch: 200\tTrain loss: 76.86992874145508\n",
      "Epoch: 250\tTrain loss: 72.45954818725586\n",
      "Epoch: 300\tTrain loss: 74.34286575317383\n",
      "Epoch: 350\tTrain loss: 70.43694686889648\n",
      "Epoch: 400\tTrain loss: 66.12868270874023\n",
      "Epoch: 450\tTrain loss: 68.66575126647949\n",
      "Epoch: 500\tTrain loss: 70.35631866455078\n",
      "Epoch: 550\tTrain loss: 65.26971092224122\n",
      "Epoch: 600\tTrain loss: 65.65645980834961\n",
      "Epoch: 650\tTrain loss: 64.86261444091797\n",
      "Epoch: 700\tTrain loss: 64.32746238708496\n",
      "Epoch: 750\tTrain loss: 62.39595260620117\n",
      "Epoch: 800\tTrain loss: 62.12804641723633\n",
      "Epoch: 850\tTrain loss: 64.06581954956054\n",
      "Epoch: 900\tTrain loss: 72.34855575561524\n",
      "Epoch: 950\tTrain loss: 64.99554595947265\n",
      "Epoch: 0\tNormalisation loss: 2.9796576261520387\n",
      "Epoch: 100\tNormalisation loss: 0.93860724568367\n",
      "Epoch: 200\tNormalisation loss: 1.0608595550060271\n",
      "Epoch: 300\tNormalisation loss: 0.9917778074741364\n",
      "Epoch: 400\tNormalisation loss: 0.9450596094131469\n",
      "Epoch: 500\tNormalisation loss: 0.9061052620410919\n",
      "Epoch: 600\tNormalisation loss: 0.8678486526012421\n",
      "Epoch: 700\tNormalisation loss: 0.8276494145393372\n",
      "Epoch: 800\tNormalisation loss: 0.787420529127121\n",
      "Epoch: 900\tNormalisation loss: 0.990388298034668\n",
      "Training AdaptiveCFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 84.44567260742187\n",
      "Epoch: 50\tTrain loss: 12.388182258605957\n",
      "Epoch: 100\tTrain loss: 11.634936904907226\n",
      "Epoch: 150\tTrain loss: 10.754131603240968\n",
      "Epoch: 200\tTrain loss: 10.279436111450195\n",
      "Epoch: 250\tTrain loss: 9.64091091156006\n",
      "Epoch: 300\tTrain loss: 9.822110271453857\n",
      "Epoch: 350\tTrain loss: 9.135330772399902\n",
      "Epoch: 400\tTrain loss: 9.430213165283202\n",
      "Epoch: 450\tTrain loss: 8.282647895812989\n",
      "Epoch: 500\tTrain loss: 7.901952600479126\n",
      "Epoch: 550\tTrain loss: 8.682451343536377\n",
      "Epoch: 600\tTrain loss: 8.835342693328858\n",
      "Epoch: 650\tTrain loss: 9.166709995269775\n",
      "Epoch: 700\tTrain loss: 8.409164905548096\n",
      "Epoch: 750\tTrain loss: 8.611199378967285\n",
      "Epoch: 800\tTrain loss: 8.154553365707397\n",
      "Epoch: 850\tTrain loss: 8.1891188621521\n",
      "Epoch: 900\tTrain loss: 7.089335107803345\n",
      "Epoch: 950\tTrain loss: 7.454970788955689\n",
      "Epoch: 0\tNormalisation loss: 1.5241156458854674\n",
      "Epoch: 100\tNormalisation loss: 0.9893463909626007\n",
      "Epoch: 200\tNormalisation loss: 0.9124066174030304\n",
      "Epoch: 300\tNormalisation loss: 0.8561822175979614\n",
      "Epoch: 400\tNormalisation loss: 0.8097961485385895\n",
      "Epoch: 500\tNormalisation loss: 1.1265826284885407\n",
      "Epoch: 600\tNormalisation loss: 0.9416178405284882\n",
      "Epoch: 700\tNormalisation loss: 0.8489664494991302\n",
      "Epoch: 800\tNormalisation loss: 0.784853321313858\n",
      "Epoch: 900\tNormalisation loss: 0.740243935585022\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 87.58925247192383\n",
      "Epoch: 50\tTrain loss: 21.897487449645997\n",
      "Epoch: 100\tTrain loss: 21.020578384399414\n",
      "Epoch: 150\tTrain loss: 19.306179809570313\n",
      "Epoch: 200\tTrain loss: 17.994976043701172\n",
      "Epoch: 250\tTrain loss: 17.389811325073243\n",
      "Epoch: 300\tTrain loss: 15.987369060516357\n",
      "Epoch: 350\tTrain loss: 16.155328941345214\n",
      "Epoch: 400\tTrain loss: 15.861713027954101\n",
      "Epoch: 450\tTrain loss: 16.198466777801514\n",
      "Epoch: 500\tTrain loss: 14.668617534637452\n",
      "Epoch: 550\tTrain loss: 14.519305801391601\n",
      "Epoch: 600\tTrain loss: 13.826822853088379\n",
      "Epoch: 650\tTrain loss: 15.028116130828858\n",
      "Epoch: 700\tTrain loss: 13.489307880401611\n",
      "Epoch: 750\tTrain loss: 16.514671993255615\n",
      "Epoch: 800\tTrain loss: 16.0770544052124\n",
      "Epoch: 850\tTrain loss: 14.327129077911376\n",
      "Epoch: 900\tTrain loss: 14.372275924682617\n",
      "Epoch: 950\tTrain loss: 13.745365142822266\n",
      "Epoch: 0\tNormalisation loss: 1.8662189364433288\n",
      "Epoch: 100\tNormalisation loss: 1.00782852768898\n",
      "Epoch: 200\tNormalisation loss: 0.9439215123653412\n",
      "Epoch: 300\tNormalisation loss: 0.894990611076355\n",
      "Epoch: 400\tNormalisation loss: 0.848849618434906\n",
      "Epoch: 500\tNormalisation loss: 0.80697580575943\n",
      "Epoch: 600\tNormalisation loss: 0.7638927698135376\n",
      "Epoch: 700\tNormalisation loss: 0.7255899548530579\n",
      "Epoch: 800\tNormalisation loss: 0.6833096325397492\n",
      "Epoch: 900\tNormalisation loss: 1.1617249250411987\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 105.77265701293945\n",
      "Epoch: 50\tTrain loss: 34.92742347717285\n",
      "Epoch: 100\tTrain loss: 32.254839706420896\n",
      "Epoch: 150\tTrain loss: 30.455157470703124\n",
      "Epoch: 200\tTrain loss: 29.23030891418457\n",
      "Epoch: 250\tTrain loss: 26.96883068084717\n",
      "Epoch: 300\tTrain loss: 26.422343826293947\n",
      "Epoch: 350\tTrain loss: 25.17423152923584\n",
      "Epoch: 400\tTrain loss: 24.874321937561035\n",
      "Epoch: 450\tTrain loss: 23.522957801818848\n",
      "Epoch: 500\tTrain loss: 23.051746368408203\n",
      "Epoch: 550\tTrain loss: 27.535110664367675\n",
      "Epoch: 600\tTrain loss: 22.589092254638672\n",
      "Epoch: 650\tTrain loss: 23.597799682617186\n",
      "Epoch: 700\tTrain loss: 27.506947135925294\n",
      "Epoch: 750\tTrain loss: 22.233683013916014\n",
      "Epoch: 800\tTrain loss: 23.584763526916504\n",
      "Epoch: 850\tTrain loss: 24.648665809631346\n",
      "Epoch: 900\tTrain loss: 22.874370574951172\n",
      "Epoch: 950\tTrain loss: 20.963798332214356\n",
      "Epoch: 0\tNormalisation loss: 2.0536038279533386\n",
      "Epoch: 100\tNormalisation loss: nan\n",
      "Epoch: 200\tNormalisation loss: nan\n",
      "Epoch: 300\tNormalisation loss: nan\n",
      "Epoch: 400\tNormalisation loss: nan\n",
      "Epoch: 500\tNormalisation loss: nan\n",
      "Epoch: 600\tNormalisation loss: nan\n",
      "Epoch: 700\tNormalisation loss: nan\n",
      "Epoch: 800\tNormalisation loss: nan\n",
      "Epoch: 900\tNormalisation loss: nan\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 127.78708343505859\n",
      "Epoch: 50\tTrain loss: 56.882066345214845\n",
      "Epoch: 100\tTrain loss: 52.855929946899415\n",
      "Epoch: 150\tTrain loss: 50.81613121032715\n",
      "Epoch: 200\tTrain loss: 49.504431533813474\n",
      "Epoch: 250\tTrain loss: 44.817701721191405\n",
      "Epoch: 300\tTrain loss: 46.40031394958496\n",
      "Epoch: 350\tTrain loss: 43.3583740234375\n",
      "Epoch: 400\tTrain loss: 42.340241241455075\n",
      "Epoch: 450\tTrain loss: 40.31144142150879\n",
      "Epoch: 500\tTrain loss: 38.390482330322264\n",
      "Epoch: 550\tTrain loss: 38.90307960510254\n",
      "Epoch: 600\tTrain loss: 39.49632377624512\n",
      "Epoch: 650\tTrain loss: 38.59342880249024\n",
      "Epoch: 700\tTrain loss: 37.93435478210449\n",
      "Epoch: 750\tTrain loss: 48.28556671142578\n",
      "Epoch: 800\tTrain loss: 49.8742561340332\n",
      "Epoch: 850\tTrain loss: 46.03072319030762\n",
      "Epoch: 900\tTrain loss: 47.40200653076172\n",
      "Epoch: 950\tTrain loss: 47.065721893310545\n",
      "Epoch: 0\tNormalisation loss: 2.884469723701477\n",
      "Epoch: 100\tNormalisation loss: 0.9413354992866516\n",
      "Epoch: 200\tNormalisation loss: 1.2021246075630188\n",
      "Epoch: 300\tNormalisation loss: 0.9934822261333466\n",
      "Epoch: 400\tNormalisation loss: 0.943819522857666\n",
      "Epoch: 500\tNormalisation loss: 0.9021033525466919\n",
      "Epoch: 600\tNormalisation loss: 0.8624977946281434\n",
      "Epoch: 700\tNormalisation loss: 0.8218292653560638\n",
      "Epoch: 800\tNormalisation loss: 0.782255333662033\n",
      "Epoch: 900\tNormalisation loss: 0.7438328385353088\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 150.95835723876954\n",
      "Epoch: 50\tTrain loss: 82.0126724243164\n",
      "Epoch: 100\tTrain loss: 78.62155380249024\n",
      "Epoch: 150\tTrain loss: 73.59943008422852\n",
      "Epoch: 200\tTrain loss: 72.4131591796875\n",
      "Epoch: 250\tTrain loss: 71.41692428588867\n",
      "Epoch: 300\tTrain loss: 63.54440002441406\n",
      "Epoch: 350\tTrain loss: 72.56935348510743\n",
      "Epoch: 400\tTrain loss: 70.05299034118653\n",
      "Epoch: 450\tTrain loss: 69.99685821533203\n",
      "Epoch: 500\tTrain loss: 65.4482234954834\n",
      "Epoch: 550\tTrain loss: 64.80992622375489\n",
      "Epoch: 600\tTrain loss: 61.3451114654541\n",
      "Epoch: 650\tTrain loss: 61.94207649230957\n",
      "Epoch: 700\tTrain loss: 59.54591178894043\n",
      "Epoch: 750\tTrain loss: 61.62533683776856\n",
      "Epoch: 800\tTrain loss: 63.04804916381836\n",
      "Epoch: 850\tTrain loss: 60.10301551818848\n",
      "Epoch: 900\tTrain loss: 56.15347747802734\n",
      "Epoch: 950\tTrain loss: 54.85137367248535\n",
      "Epoch: 0\tNormalisation loss: 3.640133094787598\n",
      "Epoch: 100\tNormalisation loss: 0.9537804126739502\n",
      "Epoch: 200\tNormalisation loss: 0.8869083404541016\n",
      "Epoch: 300\tNormalisation loss: 0.8356479108333588\n",
      "Epoch: 400\tNormalisation loss: 0.7965259253978729\n",
      "Epoch: 500\tNormalisation loss: 0.7617185413837433\n",
      "Epoch: 600\tNormalisation loss: 1.118407380580902\n",
      "Epoch: 700\tNormalisation loss: 0.9107585072517395\n",
      "Epoch: 800\tNormalisation loss: 0.8647272408008575\n",
      "Epoch: 900\tNormalisation loss: 0.8245938777923584\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.7714\n",
      "Epoch:  1 | train loss: 0.7145\n",
      "Epoch:  2 | train loss: 0.7381\n",
      "Epoch:  3 | train loss: 0.7455\n",
      "Epoch:  4 | train loss: 0.7004\n",
      "Epoch:  5 | train loss: 0.6584\n",
      "Epoch:  6 | train loss: 0.7208\n",
      "Epoch:  7 | train loss: 0.6718\n",
      "Epoch:  8 | train loss: 0.6941\n",
      "Epoch:  9 | train loss: 0.6832\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.9635\n",
      "Epoch:  1 | train loss: 1.0153\n",
      "Epoch:  2 | train loss: 1.0040\n",
      "Epoch:  3 | train loss: 0.9431\n",
      "Epoch:  4 | train loss: 0.9830\n",
      "Epoch:  5 | train loss: 0.9013\n",
      "Epoch:  6 | train loss: 0.9163\n",
      "Epoch:  7 | train loss: 0.9068\n",
      "Epoch:  8 | train loss: 0.9275\n",
      "Epoch:  9 | train loss: 0.9080\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 1.2327\n",
      "Epoch:  1 | train loss: 1.2981\n",
      "Epoch:  2 | train loss: 1.3547\n",
      "Epoch:  3 | train loss: 1.2714\n",
      "Epoch:  4 | train loss: 1.2810\n",
      "Epoch:  5 | train loss: 1.2763\n",
      "Epoch:  6 | train loss: 1.2342\n",
      "Epoch:  7 | train loss: 1.2104\n",
      "Epoch:  8 | train loss: 1.2507\n",
      "Epoch:  9 | train loss: 1.1590\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 1.6975\n",
      "Epoch:  1 | train loss: 1.6096\n",
      "Epoch:  2 | train loss: 1.6022\n",
      "Epoch:  3 | train loss: 1.4757\n",
      "Epoch:  4 | train loss: 1.6286\n",
      "Epoch:  5 | train loss: 1.5001\n",
      "Epoch:  6 | train loss: 1.6095\n",
      "Epoch:  7 | train loss: 1.5320\n",
      "Epoch:  8 | train loss: 1.4337\n",
      "Epoch:  9 | train loss: 1.4857\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 2.0161\n",
      "Epoch:  1 | train loss: 2.0302\n",
      "Epoch:  2 | train loss: 1.8750\n",
      "Epoch:  3 | train loss: 2.0723\n",
      "Epoch:  4 | train loss: 1.8195\n",
      "Epoch:  5 | train loss: 1.8148\n",
      "Epoch:  6 | train loss: 1.9757\n",
      "Epoch:  7 | train loss: 1.8661\n",
      "Epoch:  8 | train loss: 2.0964\n",
      "Epoch:  9 | train loss: 1.9251\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.7073\n",
      "Epoch:  1 | train loss: 0.7554\n",
      "Epoch:  2 | train loss: 0.7440\n",
      "Epoch:  3 | train loss: 0.7360\n",
      "Epoch:  4 | train loss: 0.7243\n",
      "Epoch:  5 | train loss: 0.7030\n",
      "Epoch:  6 | train loss: 0.6718\n",
      "Epoch:  7 | train loss: 0.7332\n",
      "Epoch:  8 | train loss: 0.7458\n",
      "Epoch:  9 | train loss: 0.7659\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 1.0338\n",
      "Epoch:  1 | train loss: 0.9611\n",
      "Epoch:  2 | train loss: 0.9355\n",
      "Epoch:  3 | train loss: 0.9583\n",
      "Epoch:  4 | train loss: 1.0007\n",
      "Epoch:  5 | train loss: 0.9882\n",
      "Epoch:  6 | train loss: 0.9618\n",
      "Epoch:  7 | train loss: 0.9588\n",
      "Epoch:  8 | train loss: 1.0495\n",
      "Epoch:  9 | train loss: 0.9844\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 1.4031\n",
      "Epoch:  1 | train loss: 1.3649\n",
      "Epoch:  2 | train loss: 1.2993\n",
      "Epoch:  3 | train loss: 1.2054\n",
      "Epoch:  4 | train loss: 1.2535\n",
      "Epoch:  5 | train loss: 1.3182\n",
      "Epoch:  6 | train loss: 1.2852\n",
      "Epoch:  7 | train loss: 1.2401\n",
      "Epoch:  8 | train loss: 1.2809\n",
      "Epoch:  9 | train loss: 1.2622\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 1.5937\n",
      "Epoch:  1 | train loss: 1.6612\n",
      "Epoch:  2 | train loss: 1.6265\n",
      "Epoch:  3 | train loss: 1.6029\n",
      "Epoch:  4 | train loss: 1.5544\n",
      "Epoch:  5 | train loss: 1.6024\n",
      "Epoch:  6 | train loss: 1.5447\n",
      "Epoch:  7 | train loss: 1.5907\n",
      "Epoch:  8 | train loss: 1.5105\n",
      "Epoch:  9 | train loss: 1.5252\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 2.0468\n",
      "Epoch:  1 | train loss: 1.8783\n",
      "Epoch:  2 | train loss: 1.9055\n",
      "Epoch:  3 | train loss: 1.8602\n",
      "Epoch:  4 | train loss: 1.9738\n",
      "Epoch:  5 | train loss: 1.8803\n",
      "Epoch:  6 | train loss: 1.8099\n",
      "Epoch:  7 | train loss: 1.9891\n",
      "Epoch:  8 | train loss: 1.8827\n",
      "Epoch:  9 | train loss: 1.9087\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.7042\n",
      "Epoch:  1 | train loss: 0.8607\n",
      "Epoch:  2 | train loss: 0.7456\n",
      "Epoch:  3 | train loss: 0.7316\n",
      "Epoch:  4 | train loss: 0.6791\n",
      "Epoch:  5 | train loss: 0.7378\n",
      "Epoch:  6 | train loss: 0.6986\n",
      "Epoch:  7 | train loss: 0.7082\n",
      "Epoch:  8 | train loss: 0.6924\n",
      "Epoch:  9 | train loss: 0.7475\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.9833\n",
      "Epoch:  1 | train loss: 0.9004\n",
      "Epoch:  2 | train loss: 0.9842\n",
      "Epoch:  3 | train loss: 0.9779\n",
      "Epoch:  4 | train loss: 0.9364\n",
      "Epoch:  5 | train loss: 0.9041\n",
      "Epoch:  6 | train loss: 0.9647\n",
      "Epoch:  7 | train loss: 0.9480\n",
      "Epoch:  8 | train loss: 0.9118\n",
      "Epoch:  9 | train loss: 0.9583\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 1.3981\n",
      "Epoch:  1 | train loss: 1.3069\n",
      "Epoch:  2 | train loss: 1.2704\n",
      "Epoch:  3 | train loss: 1.3082\n",
      "Epoch:  4 | train loss: 1.2041\n",
      "Epoch:  5 | train loss: 1.1990\n",
      "Epoch:  6 | train loss: 1.2361\n",
      "Epoch:  7 | train loss: 1.2519\n",
      "Epoch:  8 | train loss: 1.1898\n",
      "Epoch:  9 | train loss: 1.2562\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 1.5277\n",
      "Epoch:  1 | train loss: 1.6441\n",
      "Epoch:  2 | train loss: 1.7891\n",
      "Epoch:  3 | train loss: 1.6469\n",
      "Epoch:  4 | train loss: 1.5690\n",
      "Epoch:  5 | train loss: 1.6532\n",
      "Epoch:  6 | train loss: 1.5629\n",
      "Epoch:  7 | train loss: 1.5687\n",
      "Epoch:  8 | train loss: 1.6555\n",
      "Epoch:  9 | train loss: 1.6380\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 2.0383\n",
      "Epoch:  1 | train loss: 1.8160\n",
      "Epoch:  2 | train loss: 1.9486\n",
      "Epoch:  3 | train loss: 2.1947\n",
      "Epoch:  4 | train loss: 1.8448\n",
      "Epoch:  5 | train loss: 1.9572\n",
      "Epoch:  6 | train loss: 1.8254\n",
      "Epoch:  7 | train loss: 1.8432\n",
      "Epoch:  8 | train loss: 2.0312\n",
      "Epoch:  9 | train loss: 1.9052\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.7985\n",
      "Epoch:  1 | train loss: 0.7078\n",
      "Epoch:  2 | train loss: 0.7213\n",
      "Epoch:  3 | train loss: 0.6984\n",
      "Epoch:  4 | train loss: 0.6942\n",
      "Epoch:  5 | train loss: 0.7513\n",
      "Epoch:  6 | train loss: 0.6977\n",
      "Epoch:  7 | train loss: 0.6780\n",
      "Epoch:  8 | train loss: 0.6677\n",
      "Epoch:  9 | train loss: 0.6711\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.9650\n",
      "Epoch:  1 | train loss: 0.9688\n",
      "Epoch:  2 | train loss: 1.0490\n",
      "Epoch:  3 | train loss: 0.9750\n",
      "Epoch:  4 | train loss: 1.0232\n",
      "Epoch:  5 | train loss: 0.9303\n",
      "Epoch:  6 | train loss: 0.9606\n",
      "Epoch:  7 | train loss: 0.9060\n",
      "Epoch:  8 | train loss: 0.9663\n",
      "Epoch:  9 | train loss: 0.9875\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 1.4105\n",
      "Epoch:  1 | train loss: 1.3144\n",
      "Epoch:  2 | train loss: 1.2250\n",
      "Epoch:  3 | train loss: 1.3585\n",
      "Epoch:  4 | train loss: 1.2950\n",
      "Epoch:  5 | train loss: 1.3313\n",
      "Epoch:  6 | train loss: 1.2580\n",
      "Epoch:  7 | train loss: 1.2707\n",
      "Epoch:  8 | train loss: 1.1648\n",
      "Epoch:  9 | train loss: 1.2881\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 1.6930\n",
      "Epoch:  1 | train loss: 1.5346\n",
      "Epoch:  2 | train loss: 1.5979\n",
      "Epoch:  3 | train loss: 1.5889\n",
      "Epoch:  4 | train loss: 1.6760\n",
      "Epoch:  5 | train loss: 1.4737\n",
      "Epoch:  6 | train loss: 1.5542\n",
      "Epoch:  7 | train loss: 1.5804\n",
      "Epoch:  8 | train loss: 1.5506\n",
      "Epoch:  9 | train loss: 1.5503\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 1.9239\n",
      "Epoch:  1 | train loss: 1.9434\n",
      "Epoch:  2 | train loss: 1.8894\n",
      "Epoch:  3 | train loss: 1.9379\n",
      "Epoch:  4 | train loss: 1.9420\n",
      "Epoch:  5 | train loss: 1.8439\n",
      "Epoch:  6 | train loss: 1.9518\n",
      "Epoch:  7 | train loss: 1.9793\n",
      "Epoch:  8 | train loss: 1.9613\n",
      "Epoch:  9 | train loss: 1.8847\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.7583\n",
      "Epoch:  1 | train loss: 0.7098\n",
      "Epoch:  2 | train loss: 0.7465\n",
      "Epoch:  3 | train loss: 0.7308\n",
      "Epoch:  4 | train loss: 0.8585\n",
      "Epoch:  5 | train loss: 0.6921\n",
      "Epoch:  6 | train loss: 0.6799\n",
      "Epoch:  7 | train loss: 0.7345\n",
      "Epoch:  8 | train loss: 0.7370\n",
      "Epoch:  9 | train loss: 0.6868\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.9582\n",
      "Epoch:  1 | train loss: 0.9979\n",
      "Epoch:  2 | train loss: 0.9996\n",
      "Epoch:  3 | train loss: 1.0044\n",
      "Epoch:  4 | train loss: 1.0010\n",
      "Epoch:  5 | train loss: 0.9756\n",
      "Epoch:  6 | train loss: 0.9756\n",
      "Epoch:  7 | train loss: 0.9874\n",
      "Epoch:  8 | train loss: 0.9982\n",
      "Epoch:  9 | train loss: 0.9425\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 1.3526\n",
      "Epoch:  1 | train loss: 1.2796\n",
      "Epoch:  2 | train loss: 1.2216\n",
      "Epoch:  3 | train loss: 1.3256\n",
      "Epoch:  4 | train loss: 1.2424\n",
      "Epoch:  5 | train loss: 1.2991\n",
      "Epoch:  6 | train loss: 1.3024\n",
      "Epoch:  7 | train loss: 1.1446\n",
      "Epoch:  8 | train loss: 1.2892\n",
      "Epoch:  9 | train loss: 1.2441\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 1.7330\n",
      "Epoch:  1 | train loss: 1.6226\n",
      "Epoch:  2 | train loss: 1.5622\n",
      "Epoch:  3 | train loss: 1.5615\n",
      "Epoch:  4 | train loss: 1.6125\n",
      "Epoch:  5 | train loss: 1.5226\n",
      "Epoch:  6 | train loss: 1.5053\n",
      "Epoch:  7 | train loss: 1.4805\n",
      "Epoch:  8 | train loss: 1.6036\n",
      "Epoch:  9 | train loss: 1.6299\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 1.9440\n",
      "Epoch:  1 | train loss: 1.9157\n",
      "Epoch:  2 | train loss: 1.8917\n",
      "Epoch:  3 | train loss: 1.9503\n",
      "Epoch:  4 | train loss: 1.7849\n",
      "Epoch:  5 | train loss: 1.8418\n",
      "Epoch:  6 | train loss: 1.9543\n",
      "Epoch:  7 | train loss: 1.8287\n",
      "Epoch:  8 | train loss: 1.8259\n",
      "Epoch:  9 | train loss: 2.0436\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 19.8410\n",
      "Epoch:  1 | train loss: 13.8487\n",
      "Epoch:  2 | train loss: 14.2551\n",
      "Epoch:  3 | train loss: 12.4155\n",
      "Epoch:  4 | train loss: 13.7917\n",
      "Epoch:  5 | train loss: 13.4203\n",
      "Epoch:  6 | train loss: 14.4878\n",
      "Epoch:  7 | train loss: 12.5787\n",
      "Epoch:  8 | train loss: 15.0455\n",
      "Epoch:  9 | train loss: 13.3065\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 27.5054\n",
      "Epoch:  1 | train loss: 24.1595\n",
      "Epoch:  2 | train loss: 24.0495\n",
      "Epoch:  3 | train loss: 23.5248\n",
      "Epoch:  4 | train loss: 22.6089\n",
      "Epoch:  5 | train loss: 21.1649\n",
      "Epoch:  6 | train loss: 23.5767\n",
      "Epoch:  7 | train loss: 22.8174\n",
      "Epoch:  8 | train loss: 23.6026\n",
      "Epoch:  9 | train loss: 22.6461\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 45.9242\n",
      "Epoch:  1 | train loss: 39.4033\n",
      "Epoch:  2 | train loss: 39.7444\n",
      "Epoch:  3 | train loss: 38.1223\n",
      "Epoch:  4 | train loss: 39.3488\n",
      "Epoch:  5 | train loss: 35.5735\n",
      "Epoch:  6 | train loss: 36.9361\n",
      "Epoch:  7 | train loss: 38.4428\n",
      "Epoch:  8 | train loss: 35.8947\n",
      "Epoch:  9 | train loss: 33.9382\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 62.8493\n",
      "Epoch:  1 | train loss: 69.2182\n",
      "Epoch:  2 | train loss: 68.6527\n",
      "Epoch:  3 | train loss: 61.3044\n",
      "Epoch:  4 | train loss: 61.5192\n",
      "Epoch:  5 | train loss: 57.6742\n",
      "Epoch:  6 | train loss: 52.6909\n",
      "Epoch:  7 | train loss: 53.3031\n",
      "Epoch:  8 | train loss: 54.1062\n",
      "Epoch:  9 | train loss: 57.7629\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 97.4288\n",
      "Epoch:  1 | train loss: 96.6656\n",
      "Epoch:  2 | train loss: 84.6354\n",
      "Epoch:  3 | train loss: 84.8731\n",
      "Epoch:  4 | train loss: 92.2382\n",
      "Epoch:  5 | train loss: 85.5816\n",
      "Epoch:  6 | train loss: 90.2330\n",
      "Epoch:  7 | train loss: 82.0403\n",
      "Epoch:  8 | train loss: 79.0288\n",
      "Epoch:  9 | train loss: 78.3094\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 16.7985\n",
      "Epoch:  1 | train loss: 13.1820\n",
      "Epoch:  2 | train loss: 14.7854\n",
      "Epoch:  3 | train loss: 17.6894\n",
      "Epoch:  4 | train loss: 14.2310\n",
      "Epoch:  5 | train loss: 15.0478\n",
      "Epoch:  6 | train loss: 13.0958\n",
      "Epoch:  7 | train loss: 12.6777\n",
      "Epoch:  8 | train loss: 13.8005\n",
      "Epoch:  9 | train loss: 14.5159\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 25.6921\n",
      "Epoch:  1 | train loss: 26.1172\n",
      "Epoch:  2 | train loss: 20.3668\n",
      "Epoch:  3 | train loss: 22.1903\n",
      "Epoch:  4 | train loss: 22.8681\n",
      "Epoch:  5 | train loss: 20.8569\n",
      "Epoch:  6 | train loss: 23.5216\n",
      "Epoch:  7 | train loss: 21.7132\n",
      "Epoch:  8 | train loss: 21.9215\n",
      "Epoch:  9 | train loss: 21.5253\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 43.4091\n",
      "Epoch:  1 | train loss: 42.5208\n",
      "Epoch:  2 | train loss: 42.5969\n",
      "Epoch:  3 | train loss: 40.4571\n",
      "Epoch:  4 | train loss: 36.3583\n",
      "Epoch:  5 | train loss: 38.2379\n",
      "Epoch:  6 | train loss: 40.0475\n",
      "Epoch:  7 | train loss: 35.1275\n",
      "Epoch:  8 | train loss: 34.8667\n",
      "Epoch:  9 | train loss: 34.2496\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 64.7152\n",
      "Epoch:  1 | train loss: 58.7251\n",
      "Epoch:  2 | train loss: 69.3915\n",
      "Epoch:  3 | train loss: 56.1418\n",
      "Epoch:  4 | train loss: 56.1422\n",
      "Epoch:  5 | train loss: 64.3975\n",
      "Epoch:  6 | train loss: 66.0493\n",
      "Epoch:  7 | train loss: 61.8715\n",
      "Epoch:  8 | train loss: 53.9934\n",
      "Epoch:  9 | train loss: 54.0710\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 91.9167\n",
      "Epoch:  1 | train loss: 82.2487\n",
      "Epoch:  2 | train loss: 88.9494\n",
      "Epoch:  3 | train loss: 90.6061\n",
      "Epoch:  4 | train loss: 87.4785\n",
      "Epoch:  5 | train loss: 74.9120\n",
      "Epoch:  6 | train loss: 85.9943\n",
      "Epoch:  7 | train loss: 86.5989\n",
      "Epoch:  8 | train loss: 95.6154\n",
      "Epoch:  9 | train loss: 85.7444\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 17.0786\n",
      "Epoch:  1 | train loss: 15.9889\n",
      "Epoch:  2 | train loss: 14.7045\n",
      "Epoch:  3 | train loss: 16.5012\n",
      "Epoch:  4 | train loss: 14.0669\n",
      "Epoch:  5 | train loss: 13.8178\n",
      "Epoch:  6 | train loss: 13.1897\n",
      "Epoch:  7 | train loss: 11.9335\n",
      "Epoch:  8 | train loss: 13.0415\n",
      "Epoch:  9 | train loss: 13.0481\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 21.6588\n",
      "Epoch:  1 | train loss: 24.6951\n",
      "Epoch:  2 | train loss: 26.2913\n",
      "Epoch:  3 | train loss: 20.4144\n",
      "Epoch:  4 | train loss: 27.7347\n",
      "Epoch:  5 | train loss: 23.3295\n",
      "Epoch:  6 | train loss: 20.2400\n",
      "Epoch:  7 | train loss: 23.1401\n",
      "Epoch:  8 | train loss: 21.4648\n",
      "Epoch:  9 | train loss: 19.2841\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 46.4965\n",
      "Epoch:  1 | train loss: 44.0066\n",
      "Epoch:  2 | train loss: 36.9494\n",
      "Epoch:  3 | train loss: 37.1196\n",
      "Epoch:  4 | train loss: 36.9034\n",
      "Epoch:  5 | train loss: 34.3771\n",
      "Epoch:  6 | train loss: 37.6179\n",
      "Epoch:  7 | train loss: 39.0974\n",
      "Epoch:  8 | train loss: 36.9342\n",
      "Epoch:  9 | train loss: 38.6440\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 62.1366\n",
      "Epoch:  1 | train loss: 57.5680\n",
      "Epoch:  2 | train loss: 59.4419\n",
      "Epoch:  3 | train loss: 68.4156\n",
      "Epoch:  4 | train loss: 64.0303\n",
      "Epoch:  5 | train loss: 58.0671\n",
      "Epoch:  6 | train loss: 61.4508\n",
      "Epoch:  7 | train loss: 51.5770\n",
      "Epoch:  8 | train loss: 54.9209\n",
      "Epoch:  9 | train loss: 60.7734\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 92.6576\n",
      "Epoch:  1 | train loss: 87.6012\n",
      "Epoch:  2 | train loss: 88.5219\n",
      "Epoch:  3 | train loss: 88.0347\n",
      "Epoch:  4 | train loss: 90.2284\n",
      "Epoch:  5 | train loss: 91.9129\n",
      "Epoch:  6 | train loss: 83.4153\n",
      "Epoch:  7 | train loss: 88.6168\n",
      "Epoch:  8 | train loss: 77.0283\n",
      "Epoch:  9 | train loss: 83.7331\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 14.1768\n",
      "Epoch:  1 | train loss: 18.2145\n",
      "Epoch:  2 | train loss: 13.1111\n",
      "Epoch:  3 | train loss: 12.4878\n",
      "Epoch:  4 | train loss: 12.5495\n",
      "Epoch:  5 | train loss: 12.9788\n",
      "Epoch:  6 | train loss: 13.9774\n",
      "Epoch:  7 | train loss: 11.2851\n",
      "Epoch:  8 | train loss: 11.2171\n",
      "Epoch:  9 | train loss: 12.6364\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 28.2103\n",
      "Epoch:  1 | train loss: 26.7635\n",
      "Epoch:  2 | train loss: 22.6914\n",
      "Epoch:  3 | train loss: 21.3248\n",
      "Epoch:  4 | train loss: 25.4711\n",
      "Epoch:  5 | train loss: 24.1195\n",
      "Epoch:  6 | train loss: 22.3918\n",
      "Epoch:  7 | train loss: 22.5836\n",
      "Epoch:  8 | train loss: 24.9613\n",
      "Epoch:  9 | train loss: 20.2221\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 45.4793\n",
      "Epoch:  1 | train loss: 42.9064\n",
      "Epoch:  2 | train loss: 39.1494\n",
      "Epoch:  3 | train loss: 42.4022\n",
      "Epoch:  4 | train loss: 44.3275\n",
      "Epoch:  5 | train loss: 32.1160\n",
      "Epoch:  6 | train loss: 41.9855\n",
      "Epoch:  7 | train loss: 41.5141\n",
      "Epoch:  8 | train loss: 41.7776\n",
      "Epoch:  9 | train loss: 37.0314\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 68.1928\n",
      "Epoch:  1 | train loss: 60.9250\n",
      "Epoch:  2 | train loss: 64.4884\n",
      "Epoch:  3 | train loss: 56.4683\n",
      "Epoch:  4 | train loss: 55.8278\n",
      "Epoch:  5 | train loss: 59.2180\n",
      "Epoch:  6 | train loss: 60.3783\n",
      "Epoch:  7 | train loss: 59.8257\n",
      "Epoch:  8 | train loss: 57.8234\n",
      "Epoch:  9 | train loss: 55.9034\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 93.1515\n",
      "Epoch:  1 | train loss: 93.2158\n",
      "Epoch:  2 | train loss: 91.8771\n",
      "Epoch:  3 | train loss: 93.0415\n",
      "Epoch:  4 | train loss: 88.3366\n",
      "Epoch:  5 | train loss: 82.7292\n",
      "Epoch:  6 | train loss: 83.9762\n",
      "Epoch:  7 | train loss: 83.2848\n",
      "Epoch:  8 | train loss: 87.4261\n",
      "Epoch:  9 | train loss: 88.0535\n",
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 14.9991\n",
      "Epoch:  1 | train loss: 14.0670\n",
      "Epoch:  2 | train loss: 16.4768\n",
      "Epoch:  3 | train loss: 13.6899\n",
      "Epoch:  4 | train loss: 11.7472\n",
      "Epoch:  5 | train loss: 15.1485\n",
      "Epoch:  6 | train loss: 14.9840\n",
      "Epoch:  7 | train loss: 16.2987\n",
      "Epoch:  8 | train loss: 13.1073\n",
      "Epoch:  9 | train loss: 10.7872\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 24.0135\n",
      "Epoch:  1 | train loss: 24.2200\n",
      "Epoch:  2 | train loss: 23.2405\n",
      "Epoch:  3 | train loss: 23.9958\n",
      "Epoch:  4 | train loss: 22.1946\n",
      "Epoch:  5 | train loss: 24.7127\n",
      "Epoch:  6 | train loss: 21.3280\n",
      "Epoch:  7 | train loss: 23.8518\n",
      "Epoch:  8 | train loss: 21.1239\n",
      "Epoch:  9 | train loss: 21.3904\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 38.7094\n",
      "Epoch:  1 | train loss: 37.3846\n",
      "Epoch:  2 | train loss: 38.8874\n",
      "Epoch:  3 | train loss: 37.4158\n",
      "Epoch:  4 | train loss: 37.6326\n",
      "Epoch:  5 | train loss: 39.5551\n",
      "Epoch:  6 | train loss: 36.3515\n",
      "Epoch:  7 | train loss: 38.9229\n",
      "Epoch:  8 | train loss: 34.4390\n",
      "Epoch:  9 | train loss: 37.8410\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 63.7210\n",
      "Epoch:  1 | train loss: 65.3039\n",
      "Epoch:  2 | train loss: 62.0838\n",
      "Epoch:  3 | train loss: 61.1064\n",
      "Epoch:  4 | train loss: 59.9271\n",
      "Epoch:  5 | train loss: 58.4027\n",
      "Epoch:  6 | train loss: 58.7041\n",
      "Epoch:  7 | train loss: 57.3659\n",
      "Epoch:  8 | train loss: 51.2027\n",
      "Epoch:  9 | train loss: 55.7590\n",
      "Training dataset 4\n",
      "Epoch:  0 | train loss: 86.6487\n",
      "Epoch:  1 | train loss: 89.7640\n",
      "Epoch:  2 | train loss: 85.0025\n",
      "Epoch:  3 | train loss: 90.8506\n",
      "Epoch:  4 | train loss: 91.4057\n",
      "Epoch:  5 | train loss: 81.7964\n",
      "Epoch:  6 | train loss: 82.8596\n",
      "Epoch:  7 | train loss: 85.8879\n",
      "Epoch:  8 | train loss: 83.3970\n",
      "Epoch:  9 | train loss: 85.6696\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['CFRNN', 'AdaptiveCFRNN', 'QRNN', 'DPRNN']:\n",
    "    for seed in range(5):\n",
    "        run_synthetic_experiments(experiment='time_dependent', \n",
    "                                  baseline=baseline,\n",
    "                                  n_train = 2000,\n",
    "                                  retrain_auxiliary=True,\n",
    "                                  save_model=True, \n",
    "                                  save_results=True,\n",
    "                                  rnn_mode='LSTM',\n",
    "                                  seed=seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Joint coverage**\n",
    "\n",
    "Prints mean joint coverage across the horizon (mean±std of 5 random seeds, with each row indicating a different dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CFRNN\n",
      "92.7 \\(\\pm\\) 1.5\\%\n",
      "92.6 \\(\\pm\\) 0.9\\%\n",
      "91.1 \\(\\pm\\) 1.4\\%\n",
      "90.6 \\(\\pm\\) 1.4\\%\n",
      "91.3 \\(\\pm\\) 0.7\\%\n",
      "\n",
      "AdaptiveCFRNN\n",
      "93.2 \\(\\pm\\) 1.4\\%\n",
      "92.5 \\(\\pm\\) 1.4\\%\n",
      "91.3 \\(\\pm\\) 1.1\\%\n",
      "91.4 \\(\\pm\\) 1.0\\%\n",
      "67.5 \\(\\pm\\) 39.0\\%\n",
      "\n",
      "QRNN\n",
      "63.0 \\(\\pm\\) 1.5\\%\n",
      "60.8 \\(\\pm\\) 2.1\\%\n",
      "57.3 \\(\\pm\\) 2.3\\%\n",
      "57.4 \\(\\pm\\) 4.1\\%\n",
      "58.2 \\(\\pm\\) 2.3\\%\n",
      "\n",
      "DPRNN\n",
      "2.8 \\(\\pm\\) 0.9\\%\n",
      "0.4 \\(\\pm\\) 0.3\\%\n",
      "0.3 \\(\\pm\\) 0.2\\%\n",
      "0.1 \\(\\pm\\) 0.1\\%\n",
      "0.1 \\(\\pm\\) 0.1\\%\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['CFRNN', 'AdaptiveCFRNN', 'QRNN', 'DPRNN']:\n",
    "    print(baseline)\n",
    "    coverages_mean, coverages_std = get_joint_coverages(baseline, 'time_dependent', seeds=[0,1,2,3])\n",
    "    \n",
    "    for m, s in zip(coverages_mean, coverages_std):\n",
    "        print('{:.1f} \\\\(\\\\pm\\\\) {:.1f}\\\\%'.format(m, s))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Interval widths**\n",
    "\n",
    "Prints the baseline; mean±std of interval width across horizons; rows denote the noise configuration and columns the random seed. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DPRNN\n",
      "[3.01251169 3.15986606 3.63665878 3.89435101 4.27283962]\n",
      "[0.35518938 0.36311286 0.34558058 0.48825614 0.5684701 ]\n",
      "\n",
      "AdaptiveCFRNN\n",
      "[25.17730494 30.97222786         nan         nan 53.91521645]\n",
      "[5.09629808 4.67399336        nan        nan 6.70683835]\n",
      "\n",
      "QRNN\n",
      "[11.52802386 15.51419711 20.01357651 24.92123652 30.62702904]\n",
      "[1.69918445 1.69984578 1.90383696 2.07403603 2.41681679]\n",
      "\n",
      "CFRNN\n",
      "[19.77865334 26.04067698 32.76004543 40.68767166 49.18859081]\n",
      "[3.57548383 3.33551577 3.97135237 3.8959511  5.60516026]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['DPRNN', 'AdaptiveCFRNN', 'QRNN', 'CFRNN']:\n",
    "    print(baseline)\n",
    "    widths_mean, widths_std = get_interval_widths(baseline, 'time_dependent', seeds=[0,1,3,4])\n",
    "    \n",
    "    print(widths_mean)\n",
    "    print(widths_std)\n",
    "    print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sample complexity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training DPRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 10.5556\n",
      "Epoch:  1 | train loss: 7.9259\n",
      "Epoch:  2 | train loss: 5.6957\n",
      "Epoch:  3 | train loss: 6.0963\n",
      "Epoch:  4 | train loss: 6.9604\n",
      "Epoch:  5 | train loss: 5.4680\n",
      "Epoch:  6 | train loss: 6.0883\n",
      "Epoch:  7 | train loss: 5.0417\n",
      "Epoch:  8 | train loss: 3.7989\n",
      "Epoch:  9 | train loss: 5.1126\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 11.6278\n",
      "Epoch:  1 | train loss: 11.1659\n",
      "Epoch:  2 | train loss: 7.1649\n",
      "Epoch:  3 | train loss: 9.0527\n",
      "Epoch:  4 | train loss: 8.2037\n",
      "Epoch:  5 | train loss: 6.6389\n",
      "Epoch:  6 | train loss: 6.1422\n",
      "Epoch:  7 | train loss: 8.5512\n",
      "Epoch:  8 | train loss: 7.0019\n",
      "Epoch:  9 | train loss: 6.2032\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 11.5287\n",
      "Epoch:  1 | train loss: 15.5701\n",
      "Epoch:  2 | train loss: 14.4515\n",
      "Epoch:  3 | train loss: 12.5441\n",
      "Epoch:  4 | train loss: 12.3198\n",
      "Epoch:  5 | train loss: 12.1947\n",
      "Epoch:  6 | train loss: 11.7064\n",
      "Epoch:  7 | train loss: 10.4232\n",
      "Epoch:  8 | train loss: 12.9222\n",
      "Epoch:  9 | train loss: 9.9828\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 16.8015\n",
      "Epoch:  1 | train loss: 12.7351\n",
      "Epoch:  2 | train loss: 13.7319\n",
      "Epoch:  3 | train loss: 12.2638\n",
      "Epoch:  4 | train loss: 12.5810\n",
      "Epoch:  5 | train loss: 11.0966\n",
      "Epoch:  6 | train loss: 14.0130\n",
      "Epoch:  7 | train loss: 12.6540\n",
      "Epoch:  8 | train loss: 11.9889\n",
      "Epoch:  9 | train loss: 13.9595\n",
      "Training QRNN\n",
      "Training dataset 0\n",
      "Epoch:  0 | train loss: 0.5452\n",
      "Epoch:  1 | train loss: 0.3362\n",
      "Epoch:  2 | train loss: 0.2907\n",
      "Epoch:  3 | train loss: 0.2442\n",
      "Epoch:  4 | train loss: 0.2057\n",
      "Epoch:  5 | train loss: 0.1860\n",
      "Epoch:  6 | train loss: 0.1265\n",
      "Epoch:  7 | train loss: 0.0869\n",
      "Epoch:  8 | train loss: 0.0793\n",
      "Epoch:  9 | train loss: 0.0658\n",
      "Training dataset 1\n",
      "Epoch:  0 | train loss: 0.7167\n",
      "Epoch:  1 | train loss: 0.5005\n",
      "Epoch:  2 | train loss: 0.4507\n",
      "Epoch:  3 | train loss: 0.3769\n",
      "Epoch:  4 | train loss: 0.5031\n",
      "Epoch:  5 | train loss: 0.3223\n",
      "Epoch:  6 | train loss: 0.2772\n",
      "Epoch:  7 | train loss: 0.2405\n",
      "Epoch:  8 | train loss: 0.2984\n",
      "Epoch:  9 | train loss: 0.2350\n",
      "Training dataset 2\n",
      "Epoch:  0 | train loss: 0.6972\n",
      "Epoch:  1 | train loss: 0.7374\n",
      "Epoch:  2 | train loss: 0.7153\n",
      "Epoch:  3 | train loss: 0.6798\n",
      "Epoch:  4 | train loss: 0.6865\n",
      "Epoch:  5 | train loss: 0.6880\n",
      "Epoch:  6 | train loss: 0.6798\n",
      "Epoch:  7 | train loss: 0.6386\n",
      "Epoch:  8 | train loss: 0.6155\n",
      "Epoch:  9 | train loss: 0.6527\n",
      "Training dataset 3\n",
      "Epoch:  0 | train loss: 0.7373\n",
      "Epoch:  1 | train loss: 0.8796\n",
      "Epoch:  2 | train loss: 0.7820\n",
      "Epoch:  3 | train loss: 0.8135\n",
      "Epoch:  4 | train loss: 0.7398\n",
      "Epoch:  5 | train loss: 0.6859\n",
      "Epoch:  6 | train loss: 0.7412\n",
      "Epoch:  7 | train loss: 0.7517\n",
      "Epoch:  8 | train loss: 0.7153\n",
      "Epoch:  9 | train loss: 0.7393\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 167.2482452392578\n",
      "Epoch: 50\tTrain loss: 43.24018478393555\n",
      "Epoch: 100\tTrain loss: 26.187057495117188\n",
      "Epoch: 150\tTrain loss: 4.217774868011475\n",
      "Epoch: 200\tTrain loss: 2.107243299484253\n",
      "Epoch: 250\tTrain loss: 1.6684846878051758\n",
      "Epoch: 300\tTrain loss: 1.3749470710754395\n",
      "Epoch: 350\tTrain loss: 1.1831697225570679\n",
      "Epoch: 400\tTrain loss: 1.0621347427368164\n",
      "Epoch: 450\tTrain loss: 0.9858493208885193\n",
      "Epoch: 500\tTrain loss: 0.934427797794342\n",
      "Epoch: 550\tTrain loss: 0.8968338966369629\n",
      "Epoch: 600\tTrain loss: 0.8662256598472595\n",
      "Epoch: 650\tTrain loss: 0.8391954898834229\n",
      "Epoch: 700\tTrain loss: 0.8123094439506531\n",
      "Epoch: 750\tTrain loss: 0.7679761648178101\n",
      "Epoch: 800\tTrain loss: 0.7160883545875549\n",
      "Epoch: 850\tTrain loss: 0.47938206791877747\n",
      "Epoch: 900\tTrain loss: 0.32071131467819214\n",
      "Epoch: 950\tTrain loss: 0.23054395616054535\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 87.77689361572266\n",
      "Epoch: 50\tTrain loss: 15.27635383605957\n",
      "Epoch: 100\tTrain loss: 9.153653144836426\n",
      "Epoch: 150\tTrain loss: 6.44749641418457\n",
      "Epoch: 200\tTrain loss: 5.440326690673828\n",
      "Epoch: 250\tTrain loss: 4.95735502243042\n",
      "Epoch: 300\tTrain loss: 4.1591410636901855\n",
      "Epoch: 350\tTrain loss: 3.588036060333252\n",
      "Epoch: 400\tTrain loss: 3.2279155254364014\n",
      "Epoch: 450\tTrain loss: 3.0506937503814697\n",
      "Epoch: 500\tTrain loss: 2.8720579147338867\n",
      "Epoch: 550\tTrain loss: 2.721926689147949\n",
      "Epoch: 600\tTrain loss: 2.79974102973938\n",
      "Epoch: 650\tTrain loss: 2.6097347736358643\n",
      "Epoch: 700\tTrain loss: 2.486483097076416\n",
      "Epoch: 750\tTrain loss: 2.378326177597046\n",
      "Epoch: 800\tTrain loss: 2.374810218811035\n",
      "Epoch: 850\tTrain loss: 2.1860921382904053\n",
      "Epoch: 900\tTrain loss: 2.0735599994659424\n",
      "Epoch: 950\tTrain loss: 1.9736638069152832\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 88.00083923339844\n",
      "Epoch: 50\tTrain loss: 12.52962646484375\n",
      "Epoch: 100\tTrain loss: 11.29830207824707\n",
      "Epoch: 150\tTrain loss: 10.244065284729004\n",
      "Epoch: 200\tTrain loss: 9.577332305908204\n",
      "Epoch: 250\tTrain loss: 9.529282379150391\n",
      "Epoch: 300\tTrain loss: 9.673064613342286\n",
      "Epoch: 350\tTrain loss: 7.579775047302246\n",
      "Epoch: 400\tTrain loss: 6.976821041107177\n",
      "Epoch: 450\tTrain loss: 6.926629638671875\n",
      "Epoch: 500\tTrain loss: 6.397948741912842\n",
      "Epoch: 550\tTrain loss: 6.3770242691040036\n",
      "Epoch: 600\tTrain loss: 6.413407039642334\n",
      "Epoch: 650\tTrain loss: 7.327940559387207\n",
      "Epoch: 700\tTrain loss: 5.744447898864746\n",
      "Epoch: 750\tTrain loss: 5.812213897705078\n",
      "Epoch: 800\tTrain loss: 4.992693996429443\n",
      "Epoch: 850\tTrain loss: 4.921022033691406\n",
      "Epoch: 900\tTrain loss: 5.208474445343017\n",
      "Epoch: 950\tTrain loss: 5.0707169532775875\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 49.70023750305176\n",
      "Epoch: 50\tTrain loss: 12.749510192871094\n",
      "Epoch: 100\tTrain loss: 12.309471912384033\n",
      "Epoch: 150\tTrain loss: 12.204575958251953\n",
      "Epoch: 200\tTrain loss: 12.495136280059814\n",
      "Epoch: 250\tTrain loss: 12.349474411010743\n",
      "Epoch: 300\tTrain loss: 11.94230266571045\n",
      "Epoch: 350\tTrain loss: 11.834405460357665\n",
      "Epoch: 400\tTrain loss: 12.02137622833252\n",
      "Epoch: 450\tTrain loss: 11.826626358032227\n",
      "Epoch: 500\tTrain loss: 11.844655418395996\n",
      "Epoch: 550\tTrain loss: 11.674178104400635\n",
      "Epoch: 600\tTrain loss: 11.805206241607666\n",
      "Epoch: 650\tTrain loss: 11.64682222366333\n",
      "Epoch: 700\tTrain loss: 11.621992225646972\n",
      "Epoch: 750\tTrain loss: 11.594175338745117\n",
      "Epoch: 800\tTrain loss: 11.741921520233154\n",
      "Epoch: 850\tTrain loss: 11.926607265472413\n",
      "Epoch: 900\tTrain loss: 11.934220485687256\n",
      "Epoch: 950\tTrain loss: 11.568297710418701\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['DPRNN', 'QRNN', 'CFRNN']:\n",
    "    for seed in range(5):\n",
    "        run_synthetic_experiments(experiment='sample_complexity', \n",
    "                                  baseline=baseline,\n",
    "                                  retrain_auxiliary=True,\n",
    "                                  save_model=True, \n",
    "                                  save_results=True,\n",
    "                                  rnn_mode='LSTM',\n",
    "                                  seed=seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DPRNN\n",
      "6.0 \\(\\pm\\) 8.0\\%\n",
      "5.8 \\(\\pm\\) 3.5\\%\n",
      "3.6 \\(\\pm\\) 0.6\\%\n",
      "1.1 \\(\\pm\\) 0.2\\%\n",
      "\n",
      "QRNN\n",
      "0.0 \\(\\pm\\) 0.0\\%\n",
      "4.2 \\(\\pm\\) 1.2\\%\n",
      "54.7 \\(\\pm\\) 2.4\\%\n",
      "67.6 \\(\\pm\\) 0.7\\%\n",
      "\n",
      "CFRNN\n",
      "60.0 \\(\\pm\\) 16.7\\%\n",
      "94.4 \\(\\pm\\) 4.3\\%\n",
      "93.2 \\(\\pm\\) 1.2\\%\n",
      "92.6 \\(\\pm\\) 0.4\\%\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['DPRNN', 'QRNN', 'CFRNN']:\n",
    "    print(baseline)\n",
    "    coverages_mean, coverages_std = get_joint_coverages(baseline, 'sample_complexity', seeds=range(5))\n",
    "    \n",
    "    for m, s in zip(coverages_mean, coverages_std):\n",
    "        print('{:.1f} \\\\(\\\\pm\\\\) {:.1f}\\\\%'.format(m, s))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DPRNN\n",
      "[4.9435079  4.91403866 3.66253467 2.25822444]\n",
      "[1.33484906 0.41153536 0.2735284  0.33487339]\n",
      "\n",
      "QRNN\n",
      "[ 0.78771722  4.74291252 10.8338064  11.90018562]\n",
      "[0.87484591 0.97409898 1.54388235 1.81286484]\n",
      "\n",
      "CFRNN\n",
      "[16.2657159  23.71135349 21.04365135 17.35642841]\n",
      "[5.37997023 4.26279014 3.54724386 2.62391159]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for baseline in ['DPRNN', 'QRNN', 'CFRNN']:\n",
    "    print(baseline)\n",
    "    widths_mean, widths_std = get_interval_widths(baseline, 'sample_complexity', seeds=range(5))\n",
    "    \n",
    "    print(widths_mean)\n",
    "    print(widths_std)\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 936x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "coverages_mean, coverages_std = {}, {}\n",
    "for baseline in ['QRNN', 'DPRNN', 'CFRNN']:\n",
    "    b = 'MQRNN' if baseline=='QRNN' else baseline\n",
    "    coverages_mean[b], coverages_std[b] = \\\n",
    "        get_joint_coverages(baseline, 'sample_complexity', seeds=range(5))\n",
    "\n",
    "\n",
    "widths_mean = {}\n",
    "for baseline in ['QRNN', 'DPRNN', 'CFRNN']:\n",
    "    b = 'MQRNN' if baseline=='QRNN' else baseline\n",
    "    widths_mean[b], _ = \\\n",
    "        get_interval_widths(baseline, 'sample_complexity', seeds=range(5))\n",
    "\n",
    "fig, (ax1, ax2) = plt.subplots(nrows=1, ncols=2, figsize=(13, 4))\n",
    "\n",
    "figa = sns.lineplot(ax=ax1, data=coverages_mean, legend=None)\n",
    "ax1.set_xticklabels([0,1,'',2,'',3,'',4])\n",
    "figa.axhline(90, linestyle=\"--\", color=\"black\")\n",
    "figa.set(xlabel=\"log(training dataset size)\", ylabel=\"Joint coverage, %\")\n",
    "    \n",
    "figb = sns.lineplot(ax=ax2, data=widths_mean)\n",
    "ax2.set_xticklabels([0,1,'',2,'',3,'',4])\n",
    "figb.set(xlabel=\"log(training dataset size)\", ylabel=\"Average interval width\")\n",
    "plt.show()\n",
    "fig.savefig('{}.png'.format('sample_complexity'), bbox_inches='tight', dpi=600)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2016x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_timeseries(experiment='static', baseline='CFRNN', index=range(6,7), figure_name='CFRNN_static')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2016x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_timeseries(experiment='time_dependent', baseline='CFRNN', index=range(6,7), figure_name='CFRNN_time_dependent')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2016x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_timeseries(experiment='time_dependent', baseline='AdaptiveCFRNN', index=range(6,7), figure_name='CFRNN_adaptive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2016x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_timeseries(experiment='time_dependent', baseline='DPRNN', index=range(6,7), figure_name='DPRNN')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABn0AAAEsCAYAAAAGkHaAAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAADOSUlEQVR4nOzdeZyN9fvH8dc9mBn7bqzZkiVrtihb9n3NTjGhJJHyJSolFEa2FlkS2TOETEhUSsqgkrXsO0liGMPc3z8+M8yMM5yZOWfOLO/n43EeZ+Y+93Lp+/tdc5/7+nyuj2XbNiIiIiIiIiIiIiIiIpK8eXk6ABEREREREREREREREUk4FX1ERERERERERERERERSABV9REREREREREREREREUgAVfURERERERERERERERFIAFX1ERERERERERERERERSABV9REREREREREREREREUgAVfURERERERERERERERFIAFX0kybIsy9eyrJWWZe21LGuXZVnrLMsq5um4RESSAuVIEZHYKUeKiDim/CgiEjvlSEkpVPSRpO5D27ZL27ZdEVgNzPJwPCIiSYlypIhI7JQjRUQcU34UEYldks+RlmV5WZaVycHLx9OxSdKgoo8kWbZtX7dte12UTT8Bbq+uW5b1gmVZf1mWFWJZ1nbLshrH4dillmVNj+WzLJZljbIsq7TrohWR1Cq55EjLstJalvWmZVlHLMu6EfH+mmVZVoz9lCNFxGWSUY7MY1nWAsuy/rYs6z/LsnZYltXDwX7KkSLiEsklP8Y4Nq1lWTsty1rj4DPlRxFxmWSUI2sB/zl4TYlxXuXIVEpFH0lOXgC+cOcFLMt6DpgMLAG6AruBLy3Lqu7EsbWAlvfYpQrwBpAu4ZGKiNwlqebI8cBwYB7QCVgFvAm8FGM/5UgRcaekmiMDgebAO0Av4E9gnmVZbWPspxwpIu6SVPNjVMOBirF8pvwoIu6UVHNkMeAQpvgT9TUhxn7KkamUZdu2p2MQuS/LsoYDrYD6tm2HuOka6YDjwFLbtgdG2b4dOGPbdotYjusBvA48GLHpfdu2BzjYbwgwBshk2/ZNV8cvIqlXUs2RlmVlAM4B023bHhZl+zygiW3beaJsU44UEbdIwjmyIrATaGfb9oqIbRawB9hp23bXKPsqR4qIyyXV/Bjj+IeBn4FLmNzYIsbnyo8i4hZJOUdalvU2UMG27XsNPleOTMU000c8xrKsjJZlvWtZ1p8RLX/sGK8hEfu9DLQHmroryUaoBPgBC2JsXwvUtywrbSzHHQZmY0YfXXK0g2VZe4GJgA8QFvHv+9wVQYtIypRCcuSDQEbg6xjbtwG5LcvKBcqRIhJ3KSRH5sDkwy2RG2wzIu8qkD5ym3KkiMRFCsmPAFiWlQaYA0wDDjr4XPlRROIkBeXIYsBfAJZleTvaQTkydYv1j6uIO0WMYgwEHsNUnLcDNTBTDo8Ai4C1lmW9BHQBGti2fcmJc6Zx4vK2bdu3HGwvH/G+O8b2PwBfoBCmwBPzZFuI+LJuWdazsVyzJ7A44lxjI7addiJWEUmFUlCOPAY0xIzOjKoKcA24GPG7cqSIOC2l5Ejbtr8BHo24vi+QGeiBaWE0PsquypEi4pSUkh+jGIIpkL8JBDn4XPlRRJyWwnJkMcDPsqyjwAOWZZ0D3gPete+09VKOTMVU9BFPeQ7zILCxbdsbIrZtiGhzUQt4DSgABGB6VG4yeZSbtm1XieWcdYBNTlz7W6Cug+05gBu2bV+Nsf2fiPcsTpw7Nr8CBYFptm3/lIDziEjqkCJyZMQNcrRZPpbpV9wT0/ItPGKzcqSIxEWKyJExvAO8GPHzV8DqKJ8pR4qIs1JMfrQs6yFgFNDctu1rEXHGpPwoInGRYnIkUBywgWHAKaA1MA7IgFmCApQjUzUVfcRTegEboiTZSPuAVhFV6ROAwzu7WAQDVZ3Y779Ytt9vUbNrcYglpocBb2BHAs4hIqlHisuRlmUVAj4AWmCmqr8a5WPlSBGJixSXI4EpwHrMA4chwEyge8RnypEi4qwUkR8jRs7PBj6zbfteD1OVH0UkLlJEjowwCNhq2/ahiN+/sizLB3jFsqx3IlrSKUemYir6SKKzLMsP09pnsIOP8wFn4nnqK8AuJ/azY9n+D+BtWZavbdvXo2zPGvF+IZ5xATwScd1dCTiHiKQCKTFHWpbVHfgQuAUMAD6IMuUclCNFxEkpMUcC2LZ9GNO6Y23EGhZDLMvqG/GFXTlSRO4rheXH54CSQBfLsjJFbEsDpIn4/VpEmyTlRxFxSgrLkdi2HXMNIDAzxXthWr/tRjkyVfPydACSKhWOeI/WRzLiC25T4It4nrcOEObEa2Msx++LeC8bY3sp4IRt2xeJv0rAX7ZtX07AOUQkdUhROdKyrAHAfOAboKRt2+/HKPiAcqSIOC/F5EjLskZblnXAwbn+wHxPi2znoRwpIs5IMfkRqAbkBo5jRsf/BzwONIn4uWXEfsqPIuKsFJMjLct60LKs7hEze6KKnDV0NuJdOTIV00wf8YRLEe+lYmz/H5AdmBG5wbIsL0wLoDEOHhLGlNAplT9GxNYZs5gblmWlBToCXzpx3nspA+xJ4DlEJHW4FPGe7HOkZVnZgQnACqD9PWJUjhQRZ12KeE/2ORIzs6eEZVkP27b9R5Tt9TBf1iO/sCtHiogzLkW8p4T8OAaYFWPbNCAE8++JzInKjyLirEsR7ykhR+bDDKy8ivmuHakrsM+27fMRvytHpmIq+ognHAR2YvpMngf+AlphpnC/YNv2b1H2LQd0sm377fud1Lbt/4hIkPFh2/Z1y7LGAO9YlvUf8DvQGyiKWRANy7LKAQVs2/4qjqe/BDxiWVZj4F/goG3bf8c3VhFJ0VJSjmwI+AJfA40dLMD7rW3b11COFBHnpaQcuRCzSPlay7ImYNqKNAWeAgZEechwCeVIEbm/FJMfbds+GPHvuc2yrH+BK7Ztb4my+RLKjyLinBSTI4GtwDbgE8uyHgSORvxbWgPto5z+EsqRqZZ1/4KliOtZlvUAMB14AtO+IhgYb9v26ij7lAE2YKYnngK+sm17mJvjsoCXgReAXJg/CC/Ztr0t4vO5wFO2bTtc1M2yrCPAGtu2B8TYXhazEGV5zAPQWjFuVkVEbkspOdKyrFeA8fc4ZVHbto8oR4pIXKSUHBmxrTQQANTCLLS7F3jPtu1Po+yjHCkiTklJ+dHBOTZjij4tomxTfhQRp6WkHGlZVh7MrMgWmJlKe4EJtm0vjLKPcmQqpqKPJGmWZU3D9J+c7OlYRESSGuVIEZHYKUeKiDim/CgiEjvlSEkJvDwdgMh9VCYB0yRFRFI45UgRkdgpR4qIOKb8KCISO+VISfY000eSLMuy0gD/APls277q6XhERJIS5UgRkdgpR4qIOKb8KCISO+VISSk000eSsnzAdSVZERGHlCNFRGKnHCki4pjyo4hI7JQjJUVQ0UeSstPAdsuy9luWFeDpYEREkhjlSBGR2ClHiog4pvwoIhI75UhJEdTeTUREREREREREREREJAXQTB8REREREREREREREZEUIK2nA3BWrly57CJFing6DJEU4Z9//gEge/bsHo4kZQkODr5g23buxL6u8qOIaylHuodypEjyp/zoHsqPIimDcqR7KEeKJH/Kj+4TW45MNkWfIkWKsH37dk+HISISK8uyjnriusqPIpIcKEeKiDim/CgiEjvlSBGR2MWWI9XeTSQVOn78OMePH/d0GCIiSZJypIiIY8qPIiKxU44UEXFM+THxJZuZPiLiOj169ABg8+bNng1ERCQJUo4UEXFM+VFEJHbKkSIijik/Jj7N9BEREREREREREREREUkBVPQRERERERERERERERFJAVT0ERERERERERERERERSQFU9BEREREREREREREREUkB0no6ABFJfEOGDPF0CCIiSZZypIiIY8qPIiKxU44UEXFM+THxqegjkgq1bNnS0yGIiCRZypEiIo4pP4qIxE45UkTEMeXHxKf2biKp0P79+9m/f7+nwxARSZKUI0VEHFN+FBGJnXKkiIhjyo+xuHwZZs2CZ5+FW7dcemrN9BFJhfr16wfA5s2bPRuIiEgSpBwpIuKY8qOISOyUI0VEHFN+jGH3bpg4EZYtg5AQeOABWL0a2rRx2SVU9BEREREREREREREREXGHEycgbVrw84N9++Dzz6FGDahdG8qVc/nlVPQRERERERERERERERFxldBQM4NnzhxYtw5694bWreH6dfjoI8iXD9KkgZs34coVl15aRR8RERERERERERERERFXGD4cZs6Ev/+GPHmgVSsoW9YUeQoWdPvlVfQRERERERERERERERGJi/BwOHkSDh40Ldx69oTLl00Lt5IloVYtqFYNsmUDL69EC0tFH5FUaOTIkZ4OQUQkyVKOFBFxTPlRRCR2ypEiIo4l+/xo23DqlCnsVK0KGTPC4sXw9tvw11+mXVukGzfM5507Q/bs4O3tkZBV9BFJhRo0aODpEEREkizlSBERx5QfRURipxwpIuJYssyPv/wC77xjCj1//gnXrpntixZB4cKwfz/4+kL9+mZtnjx5TNu2vHkhUybPxo6KPiKp0q5duwCoWLGiR+MQEUmKlCNFRBxTfhQRiZ1ypIiIY8kmPx45YmbtZMwIO3bAtm3g5wd165piTt685vNz56B6ddO6LRFbtsWFij4iqdCgQYMA2Lx5s0fjEBFJipQjRUQcU34UEYmdcqSIiGPJIj9u3Qpt2kDOnPDqq6bAM3Mm+Pgk2cLOvSS/iEVERERERERERERERBJq4UKoVw/SpYPevU27tqxZIX36ZFnwARV9REREREREREREREQkKbFteO45GDvWtFVztfBweP116NYNHnoIRo2CypUhTRrXXyuRqb2biIiIiIiIiIiIiIgkHSdPwkcfmZ8//hh++sm0XXOVmzdhwwZ44gno29es35NCqOgjIiIiIiIiIiIiIiJJx44d5r1DB/jnH9i3DzJkMC3XMmWK/3lPnzZr9YSEmGJPrlyQObNrYk4iVPQRSYXGjh3r6RBERJIs5UgREceUH0VEYqccKSLiWLzz45EjkDYttG0L+fPDxYvw6acwciT072/eM2aM2zl37YKWLaFECbN+T/78pgCUwri96GNZlhcwG6gP/AP0ByzgMyCyGd8btm0vcXcsImLUrFnT0yGIiCRZypEiIo4pP4qIxE45UkTEsXjnx4EDoXBh8PY2v+fIYdbhKV8e3nkH5s2DyZPNTCDLuv/5Vq2Crl1Noah+fciXL0Ws3+OIVyJcoy2QAygMdAE+BIoDb9m2XSripYKPSCL68ccf+fHHHz0dhohIkqQcKSLimPKjiEjslCNFRByLd368dQuuXwdf3zvbcuWC0aNh/HgzC6hjR1PAuXkz9vPYNkycCG3aQMGC8MYbULNmii34QOK0d8sLfGrbtg3ssSwrN/AA8FMiXFtEHHj11VcB2Lx5s2cDERFJgpQjRUQcU34UEYmdcqSIiGPxyo/nz0PnzlCrFvj53f151aowZw4sXQonT8Lu3VC6tPksZru2f/+FKVNMoee556BAgfj9Q5IRt8/0sW37fdu2AwEsy+oDHMTM+hlhWdZey7I+sywri6NjLcvqa1nWdsuytp8/f97doYqIJBvKjyIisVOOFBFxTPlRRCR2ypEiSciOHfDNN2a2T2zSpIEuXWDwYNi/H957z7SD+/RTM7vnn3/MDKBr1+DVV+Hll1NFwQcSp70blmVltCxrJvAi8DSwAxgJlAGOAW85Os627Y9t265i23aV3LlzJ0aoIiLJgvKjiEjslCNFRBxTfhQRiZ1ypEgSsmOHeS9R4v77pkljZgNlzQqZMsHTT0ONGmY20DPPmOJRkSKQLZsbA05a3F70sSwrA/A9cAWoYtv2IeAz27a3RLR8+wwo6+44REREREREREREREQkiduxA/LnN2v4OKt0afj4YxgwAPbuhb//NsUePz9In95toSZFibGmz7PAFtu2B0fZts2yrM62bf8KtAa2JUIcIiIiIiIiIiIiIiKSlO3YAQ88AL6+cTvOywvat4cmTcxaPnnzmm2pTGIUfSoDtS3LahRlWz9ggWVZaYDdgH8ixCEiESZPnuzpEEREkizlSBERx5QfY2Hb5qHCxYtQrJinoxERD1GOFBFxLM75MSwMChUys3TiW7DJmNG8Uim3F31s2+4Wy0dq6SbiIRUrVvR0CCIiSZZypIiIY6k2P966BadPw7FjcPQotGljWoTMmWMWDD56FP77zxR8fvoJtA6ESKqUanOkiMh9xDk/pksHCxfCd9+5JZ7UIDFm+ohIEvP1118D0KBBAw9HIiKS9ChHiog4liry45EjULAgpE0Ly5fDyy/DiRNw8+adfZYuNYWd/ftNy5HHHoOcOc227duhaVOPhS8inpMqcqSISDzEOT/atplBnSaNG6NK2VT0EUmF3n77bUA3oyIijihHiog4lqLz408/QUAABAbCxIlQtizs2wcFCkCFCmYR4Vy5TGEnPByuXIH69U2/eMsy5wgNNZ+JSKqUonNkpFOn4PffzRoZRYtCliyejkhEkoE458enn4a9e+GNN9wXVAqnoo+IiIiIiIikPuHhpsgzaRJs3QqZM0OLFmb2TlgYPPqomcUjIpLahYfDxx/DK6+YojfA2LHQr5+ZIdm3rymS58tnXvnzQ6NGULjwnZmSafUIUkSc9PPPkCmTuSeTeFHGFRERERERkdTj1i3TLsS2YdgwuHYNnnrKzNzJn1+tREREorJtaNsWVq2CihVNC8tr1yBDBtiwwax3FhYGu3bB5s1w+bI5bupUU0j/4QeTY6dMgeefvzM7UkTEkStXTAvd9u11T5YAKvqIiIiIiIhIynfsmHkIuXw5fPUV/PUXvPACFCoE2bPrQaSISFTh4SYv3roFjzxi2lu2anV3Szc/P6ha1ex/8yaEhMCFC+DjA9u2wdmz0KYN/Puv+TlvXo/8c0Qkmfj1V1NsLlrU05Ekayr6iIiIiIiISMr188+mhdvnn5vfa9SAjRtN26EKFTwbm4hIUnT4MPj7Q4cOJlc+9BA8/vi9R917eYG3t3lly3Zne968ULkynDljHuSKiNzLjh3mvVgxz8aRzKnoI5IKzZgxw9MhiIgkWcqRIiKOJWp+DAkxrYNsGxo3ht27IWtW88qWDSpVgnHjzL6zZpn9Iz/PmtW0aStZ0iw4Xr266QvfrJk5V4kS5qGkiIgLpYh7yPBw+PBD+N//zO+lSpnZkPnyeTYuEUnW4pQfS5SAli3NOmESbyr6iKRCJUuW9HQIIiJJlnKkiIhjbs+PN27Al1/CnDlmds7mzWa0eZ48poATEgJXr5q2Qdevm33TpYM334QTJ6Kfq25dmD8fLl2C/v1N4adAAfWGFxG3Sfb3kIcOQe/e8O23pp3b009DmTLKmyKSYHHKj40bm3V9MmRwX0CpgIo+IqnQ6tWrAWjZsqWHIxERSXqUI0VEHHNbfjxyxKy189lncP485MwJjz0GW7ea0eXPPBN9f9s2o9Fv3TKvqVPNw4H//jOvq1fNWhJbtoCvr2lPpPV6RMTNkv095K5dEBwMffpAixZ3r90jIhJPTufHsDBzXxgWpoJzAqnoI5IKBQQEAMn4ZlRExI2UI0VEHHNpfrx0yczWyZsXDh6EadOgShUzsrx6dcie3awP4YhlmQcBkQ8D0qc3+4uIeFCyvIf86y/Yts2MrPfxgffeM4un62GriLiQ0/lx1y6oVg2GDDHFZ4k3FX1ERERERETE/cLDYdMm074tMNDMwOnZE86dgw8+MIuFa60dERH3Cw+H6dNh+HBT7JkwAfz84MEHPR2ZiKRmO3aY98KFPRtHCqCij4iIiIiIiLiPbcPkyTBlChw9CpkzQ+3aZq2I0FAt1Csiklhu3IAVK8yMnm3boHJl6NULihTR7B4R8bwdO8x94gMPeDqSZE9FHxEREREREXGfGzdMf/YsWWDAAKhZE/Lk0QNGEZHEcvMmpE0Lp09D165mVk/fvqZ9UubMno5ORMTYscMUoTNm9HQkyZ6KPiIiIiIiIuI+Pj7Quzc8/jjkzu3paEREUofwcFi3zrTPvHnTzLbcswfeesvMtMyWzayRJiKSFISFwW+/QaNGpkgtCaL/giKp0Pz58z0dgohIkqUcKSLiWLzyY0gIfPstXL4MOXO6PigRkSQiydxDXrhg1k776CM4fBhy5IA6dSA42Pz82GOejlBEUhmn8mN4uClO//OP+wNKBVT0EUmFChUq5OkQRESSLOVIERHH4pUfV6yA7t3htdfgiSdcH5SISBLh0XtI2zYvLy/4+GMYMQLKljUtNSNnWXp5eS4+EUnVnMqPPj7QoAFs3+7+gFIBFX1EUqElS5YA0KlTJw9HIiKS9ChHiog4Fq/8OGcO5MsHlSq5KSoRkaTBI/eQ//0HCxaYFm4DBkCVKpA3L4wfb4o+6dMnXiwiIrFwKj/++CMcOADZsydSVCmbij4iqdCHH34I6IGmiIgjypEiIo7FOT8eOQLffAMdOpi1I0REUrBEv4dcuBCee860z3zwQfOwNEcOKFwYihVLnBhERJzgVH4cOtS0dnvvvUSKKmVT0UdERERERERcb+5cs0h43bpaLFxExJXeeAPeegsefhg6djSzKTNn9nRUIiLxc+sW7NoFtWqBt7eno0kRVPQRERERERER19uwASpUgOLFPR2JiEjKUq4ctGgBPXpAnjyejkZEJGEOHICrV6FoUU9HkmKo6CMiIiIiIiKuFxhoXr6+no5ERCT527wZdu6ERo3g+nXo319r9ohIyrBjh3lXa0qXUdFHREREREREXMu24dgxjUAXEUko24aAABg2DAoWNGukPfAApEnj6chEJJn6/fffyZEjBwUKFPB0KMaOHeDjo5k+LuT2oo9lWV7AbKA+8A/QHzgNfA7kAj63bfsld8chInd8/vnnng5BRCTJUo4UEXHM6fx46RJUrgzt20Pjxm6NSUQkqXDLPeTly9C7NyxfDjVqQL9+ULiw668jIqlK06ZNqVSpEqtXr06U6903P44cCfnza20yF0qMmT5tgRxAYaA0sBj4C3gDWAOssSyrnm3bmxIhFhEBcuXK5ekQRESSLOVIERHHnM6PixfDoUOQNatGootIquHye8iwMHj8cdizB7p1g44dIUsW115DRFKdU6dOcfLkSf755x9CQ0Px8fFx+zXvmx8zZjRFH29vt8eSWnglwjXyAp/axh4gN1AFWGPbtg0sAxolQhwiEmHu3LnMnTvX02GIiCRJypEiIo45nR8/+cS05yhf3u0xiYgkFS6/h7QsaNPGtHXr2VMFHxFxie3btwMQEhLCjz/+mCjXvGd+PHkSXn8dLlxIlFhSC7cXfWzbft+27UAAy7L6AJeAixEFH4BTmMLQXSzL6mtZ1nbLsrafP3/e3aGKpBp6oJn8KT+KuI9yZPKnHCniHk7lxz/+gJ9/hsceU4uOJEj5UcR9XHIPGRYGgwbBokXwzTdQpgw88YRGvycS5UhJDYKDg/Hy8iJt2rSsW7cuUa55z/y4dSu8+y7891+ixJJaJMZMHyzLymhZ1kzgReAp4GaUj23glqPjbNv+2LbtKrZtV8mdO3ciRCoikjwoP4qIxE45UsSDPvkE0qaF2rU9HYk4oPwokoSdOgX16sGUKfD553D1KuTNC16J8uhOUI6U1CE4OJhSpUpRs2ZN1q9f7+lwYMcO0w74wQc9HUmK4va/HJZlZQC+B65g2roFAzmj7JIfOObuOERERERERMTNmjeHHj200LiISFx89x088oh5+Nm/PzzzDGTP7umoRCQFCg4OpnLlyjRq1IidO3dy7tw5zwa0Ywc88IBaWLpYYgwXeBbYYtv2YNu2r9u2fQvYZVlWPcuy0gDdgS8TIQ4RERERERFxpzJloGFDtSISEXHWl1/eaeH2xhvQrh2kT+/pqEQkBTp16hRnzpyhcuXKNG7cGICvv/7acwHZ9p2ij4+P5+JIgdImwjUqA7Uty2oUZVtLYDFmxs9c27aDEyEOERERERERcZdPPgFfX/MSERHn1K1rWrplyABFioBleToiEUmhgoPNI/gqVapQqVIlcubMyfr16+natatnArpwwbSyLFpUuc/F3F70sW27WywfVXb3tUXEsbVr13o6BBGRJEs5UkTEsXvmxzNnoE8faNbMLEIuIpLKxPseMmNG8PMz7dz00FNE3Cg4OBgvLy8qVqxImjRpaNCgAevXr8e2bSw35p9Y82Pu3PDHH6bFpbiUVoMTSYUyZMhAhgwZPB2GiEiSpBwpIuLYPfPjZ5/BrVtQu7YWHReRVCle95Dh4fC//8G+fWYhcxERN9q+fTulSpUiY8aMADRq1IjTp0+ze/dut173nvnx4kXIlMmt10+NdDcukgp98MEHfPDBB54OQ0QkSVKOFBFxLNb8aNswZw6UKgUPP5z4gYmIJAHxuoc8fhzGj4ejR90TlIhIFMHBwVSufKf5VqNGZjWW9evXu/W6sebHoUNNe0utY+ZyKvqIpEJLly5l6dKlng5DRCRJUo4UEXEs1vy4bRvs3Qu1aulLu4ikWvG6h9yzx7wXKOD6gEREojh16hRnzpyJVvQpWLAgZcqUcXvRJ9b8uHAhHDqk9SDdQEUfERERERERib+9e81aFI895ulIRESSl8iiT+HCno1DRFK84OBggGhFHzCzfb777juuXbuWuAGdPQsnT0LRolrPzA1U9BEREREREZH469ED3n9fI9VFROLqjz8gRw6zmLmIiBsFBwdjWRYVK1aMtr1Ro0Zcv36d77//PnED2rnTvKvo7RYq+oiIiIiIiEj8XLkC586Zn9Om9WwsIiLJzbFjpmCu1kYi4mbBwcGULl2aTJkyRdteu3ZtvL293d7i7S47dpj3EiUS97qphIo+IiIiIiIiEj+tWkHXrpAxo6cjERFJftasgcGDVTQXEbcLDg6+q7UbQMaMGalVq1biF328vaFCBciVK3Gvm0ror4pIKrR582ZPhyAikmQpR4qIOHZXfjx0CDZtgiefhMyZPRKTiEhSEa97yJAQyJDB5bGIiER16tQpTp8+7bDoA6bF2//+9z9Onz5Nvnz5XH59h/nxpZegUCHNdHQTzfQRERERERFJRS5fvsyJEycSfqJPPzUL79apowV4RUTiautWeOYZOH/e05GISAoXHBwMcM+iD8CGDRsSJyDbNi2CbRu8VJ5wB/1XFUmFJk6cyMSJEz0dhohIkqQcKSIpXZ8+fShatCgjR47k+vXrTh8XLT+Gh8PcuVCxIhQv7pY4RUSSkzjfQ27dCitWaKaPiLhdcHAwlmVRsWJFh5+XL1+ePHnysG7dOrdc/678uGkTlCkDR4645Xqioo9IqrRmzRrWrFnj6TBERJKkFJsjQ0PNQ1qAUaMgd24oWhTKlYOaNaFJE7h82Xy+di2MHAnvvgsffADz58PKlXeOF5FkKyQkhNWrV5M3b17GjBlDpUqV2LJli1PHRsuP33xjFiB//HG15RARIR73kHv2QLZskCeP22ISEQFT9ClVqhSZMmVy+LmXlxcNGzZkw4YNhLvhO99d+XHHDjh50nwnFbfQmj4iIiIiKdVff8FXX5nXpk2mcJMrF/j4mNH516/DtWum2HPuHHz5JWTKBPPmwfLlZrp9JMuC/fvNgpv58pl3EUl21q1bx7Vr1/j0008JCwujX79+1KpVi+eff55x48aR2dm1eapXh9deM6M0RUQk7vbsgQIFVDgXEbcLDg6mQYMG99yncePGLFiwgF9//ZVKlSq5N6AdO0zBO29e914nFVPRR0RERCSl2bMHWreGP/80v+fPD489Br/9Bg88AJUrQ40a0Y+xbTOT59Yt6N0bnnrKFISuXjX9lq9dMzfn169Ds2YalSWSTAUGBpIjRw5q165N2rRp2b17NyNHjmTq1KmsWrWKjz76iGbNmt3/RD4+ULYs5Mzp/qBFRFIa2zb3a9WqQVo9mhMR9zl16hSnT5+OdT2fSJFFofXr1ydO0eeBB1T0diO1dxOR1CE8HP7+24xS37LFjHZfuPDO55MmmYeYVavCk0/CzZseC1VExGm2Db//DhMmQIMGMHo0nD5t8l327NCrF7z3HsyYAcOGwSOPmJk+jmbpWBakSWM+S5/ezPjJnRuKFDEPdqtWBT8/86BXRJKlGzdusHr1alq3bk3aiIeMmTJlYvLkyfz4449kzpyZ5s2b0717dy5cuBD7iZYvh7FjISzM5A0RkSTm66+/pkyZMpw7d87ToTh2+bKZ5VO4sKcjcbsNGzbQuXNnOnfpQtWqVVmwYIGnQxJJVYKDgwHuW/TJly8f5cuXZ/369e4N6MoVOHDAfM/UfaTbaDiBSCqUPn16T4fgeufPm5FShw/feZ04AevWmVHrzz4Ln34a/RgfH7Pw8PXr8MMPcPAgZMxo1r3Yts2MiheRVCdZ5EjbhoAAmD4djh4124oUMaOlvv3WFG7GjIF06TwapogkLZs2beLff/+lXbt2d3326KOPsmPHDsaNG8fYsWNZt24dU6ZMoUuXLliWBUTJj++8AxcvwuTJiRi9iIjzpk+fzt69e/nggw8YNWpUolwzTveQWbOa76pOrqmWXG3YsIGAgACuh4biB5w4eZK+ffsC0K1bN88GJ5JKBAcHY1kWFStWvO++jRo1YurUqVy9epWMGTO6LIZo+fHaNdNVokgRl51f7qaij0gqFBQU5OkQ4u7SpehFnSNHzPuiRaZQM348TJxo9rUsM5I9Vy747DMzXTR/fujZE7JkgcyZzSt7djh1yjwUfeYZM60+bVoICTGFIBFJlZJ0jjx8GIoWNbMRN2wwDwz69oUKFcxNc/r0JgeKiDgQGBhIpkyZYu3p7uPjw6hRo3jyySfx9/enW7duLFy4kA8//JBChQqZ/Pj771C+PPToYe6nRESSmH/++YegoCC8vLx4//33GTp0KBkyZHD7deN8D3npUoof5T579myuh4ZG2xYSEsKIESNU9BFJJMHBwZQqVYpMmTLdd9/GjRszceJEvv32W+fa/TopWn7MnRuGDjX3lOI2KvqISNIRHm6KObt3wx9/mPfXXzcj1z/9FAYNurNvjhymqLNwoVn4LX9+GD7c/JwvnykEeXvf6Y+cL58n/kUiIgl38yasWgVTp8J338EXX5jCdLdu5oY5OcxMEhGPu3XrFitXrqR58+b43qd/+sMPP8wPP/zAtGnTGDFiBA8//DDvvvsu/fr1w+uTT8z9Ve3aiRS5iEjcBAYGcuPGDQICAhgyZAjz5s3j2Wef9XRY0Q0bBlu3wiuveDoStzp79qzD7ceOHUvkSERSr+DgYOrXr+/Uvo8//ji+vr6sX7/epUWfaE6fNgOwE6EYn5qp6COSCo0ePRqA1157zTMB2DacPGmKOg89BMWKwfffQ5MmZpZNJD8/83nJkubhwiuv3CnwZM5sijqRrYvy5vXMv0VEUhyP58hIly/DRx/B++/DsWMmz3XuDBcuQKFCpuAjIuKkrVu3cu7cOdq2bevU/mnSpGHQoEG0bt2afv360b9/fya98w47z50jU8WKZlCOiEgStHDhQkqUKMHgwYNZtGgRkyZNom/fvnh5uXdZ6zjdQ27ZAv/8k+LXSvTz8+OMg8LPA/obIpIoTp8+zenTp6lSpYpT+/v6+lKnTh2Xr+sTLT82aWKe5b37rkuvIdG59y+eiCRJGzduZOPGjYl70YsXoV8/s05O9uzmgWXTpvDBB7BpE+zbB3XqmDZFb70F8+fDnDnw+OPmQefDD0OzZmYR8rx5zUwerVUhIm7gkRwZVWTx+8oVeO01yJYNBg826/f06WPau6XVuB0RiZvAwEC8vb3jPGqzaNGirFu3jrlz5+Jz4gSXrl8nfOdOM3hHRCSJOXXqFJs2baJr165YlsXLL7/MwYMHWb16tduv7fQ9pG2b1uX58qX477T+/v74xihsZciQgTFjxngoIpHUJTg4GIDKlSs7fUyjRo3Yu3cvx48fd1kct/Pj9eums0+hQim+vaWn6YmBiLiPbcOZM+ZmNiQEPv/c/FytmknwBQuaWT5Xr5r3oUM9HbGIiGfcugWrV5sWbiEh8PHHsH8/vPeeKfKohZuIJIBt2wQGBtKoUSMyx2MdHsuyeOqpp7BHjKDgyZMmZ73+OkyeDA8+6PqARUTiacmSJdi2TZcuXQBo3749hQsXZuLEibRu3drD0UU4e9bM8ilUyNORuF3Dhg0Bs7YPZ89SsEABBr37rtbzEUkkwcHBWJZFxYoVnT6mUaNGAKxfvx5/f3/XBvT77+Y+smhR155X7qKij4i4x5Ytpk/x8eMQcYPH7NmaoSMiEtXp0/DJJ6bIc/SoaWv5xBNm9GeuXGrhJiIusXPnTo4ePcobb7yRoPOsz5+fp06exAIzaGfYMDMLUW12RSSJWLRoEZUrV6ZkyZIApE2blsGDBzNo0CB++uknHn30UQ9HiBnlDqZteSrQsGFDU/w5c8Z099B6uyKJZvv27ZQqVYpMmTI5fczDDz9M/vz53VP02bHDvKvo43aJ1t7NsqzWlmW9E/Hz45ZlHbEsa1/Eq1NixSEibvb779CyJdSqZUap169vWrvlz29aFKngIyKp3a1bcOOG+fnzz2HECMiaFQYNMuv39OljHqCqhZuIuEhgYCBp0qShZcuWCTrPqQwZOJw+PUSui/H336bwc/myC6IUEUmYgwcP8ssvv9ye5ROpd+/eZMuWjYCAAA9FFoOPj2ljXqQIGzZsoHPnztSrV4/OnTuzYcMGt1/eE9cUSQ2++uor3nzzTTZu3MjVq1c9HQ5gZvrEpbUbmBnejRo14uuvv+bWrVuuDWjHDrNGt9b1cju3P02wLMsCJgFdgU8iNhcH3rJte467ry8id8uZM6d7TrxtG9SoYWbzdO4MLVqYUTxuXjBTRMSV3JYjT540a5XNmgUDBkDt2pApEwQEQOnSauEmIm4TGBhInTp1yJUrV4LOkzNnTvbny0exEycgPBxu3jSzFEeMgIkTU/yC5CKStC1atAjLsujUKfq44syZM9OvXz8mTJjAoUOHKFasmFuu7/Q95OOPw9ixbFqxgoAPPuB6aCgAZ86evV2YimyL5mobNmwgICAgUa8pkhrYts3zzz/PoUOHADPLsHLlytSuXZtatWrx+OOPkz179kSN6fTp05w+fTrORR8wLd7mzp3Ljh07qFq1aoJjuZ0fu3c394sZMyb4nHJvifUkdiOwMMrvDwCuWw1KROJk+fLlLF++3DUnO3cO1q83o9YzZIAePWDKFHjmGShQQAUfEUl2XJojbRvWrIFWrcxoptdfhxw5TB/3kyehcGF45BEVfETEbfbt28fevXtp27Ztgs+1fPlymo4cae77une/88Hu3TBmjJnJKCLiAbZts3DhQurUqUPBggXv+vyFF14gTZo0TJ482W0xOH0PefMmXLzIrAULbhdfIl0PDTXr37jJ7NmzE/2aIqnBwYMHOXToEO+88w5r167llVdeIW3atEyZMoVWrVqRM2dOKlSowIABA1i6dCmnT592e0zBwcEA8Sr6NGjQAMuyWLdunUtiuZ0fq1Qxg8XV1cLt3P5f2LZtG1hjWVYuoFTE5sJAfcuypgLBQH/btu/qCWBZVl+gL8ADmvYlkrT8958ZnR4QAGnSwAcfmAJPjx5K3olA+VEkifvnH8ie3RR93noL/vrLzH584gl46CEVedxMOVLkjhUrVgDQpk0b15ywdWvYsMEM/EmfHj780Gz//nuYNg1efBEsyzXXEpdTfpSUateuXezfv58hQ4Y4/LxAgQJ07dqV2bNnM2rUKHLkyJHIEUawbTM48tFHOXXunMNdzp4967bLx3Zud14zOVGOlPgKCgoCoGPHjhQtWpSmTZsCcO3aNX7++We+//57vvvuO+bOncv7778PwIMPPnh7JlDDhg0pUKCAS2MKDg7GsiwqVaoU52Nz587NI488wvr16xk5cqRrAjpzBn744U6rc3ErTw3B3wGMBMoAx4C3HO1k2/bHtm1XsW27Sm4tZCziMsOHD2f48OHxOzg01MzkKVYM3nwTypWDkSPNguNagyLRKD+KuE+8c+TNm7ByJTRrZr7Mb9tmZvn06AEffWQehFaooIJPIlCOFLkjMDCQ6tWrOxz5HlfDhw9n+IQJULGiGfTTsSM8+eSdHb74AhYujPV48TzlR0mpFi5cSLp06Wjfvn2s+wwZMoSQkBA++ugjt8Tg1D3k+fOmaJ41K35+fg53iW27K3jimsmJcqTEV1BQEKVKlaJo0aLRtqdPn546deowcuRI1q9fzz///MPPP//MxIkTefjhh1m5ciW9evWidOnSnDlzxqUxBQcHU6pUKTJlyhSv4xs1asTWrVu57IK1G4cPH84yf3/o0MGsCSlu51TRx7KsjJZlPWdZ1njLsiZFfcXzup/Ztr0lYhbQZ0DZeJ5HROJh69atbN26NX4H//abWWw8f34zev2118z0TG9vl8YoIuIp8c6Rr70GbdvCL79AkyYmX3p7m+J47twO211qIV0Rcadjx46xfft22rVr55Lz3c6P330HETOIePZZM4sRTCEogesGiYjEVXh4OIsWLaJJkyb3nMFTrlw5GjduzLRp0wiN0eLMFZy6h/zjD/NesCD+/v74xlgLzdfHB39/f5fHFskT1xRJ6UJCQti8efPt2T33ki5dOqpWrcqQIUNYuXIl58+f59tvv+W///5j3rx5Lo0rODg4Xq3dIjVq1IibN2+yefPmBMeydetWvH791QyAjFEYE/dwdqbPMuBVIB+QPcYrPrZZllUh4ufWwLZ4nkdEEluuXPD22+ZVq5ZGrIuIgGnVsXAhVKpkWh0NGAAlSoCvb6yHRC6ke+bsWWzuLKSrwo+IuEpkazdXrOcTzZYtpuhz7ZopaP/vf1CzJowdC40bu/ZaIiL38f3333Py5Em6du16332HDBnCmTNnWOipWYl79pj3woVp2LAhQ4YMIa+fHxaQ18+PIUOG0LBhQ7dd3hPXFEnpNm3aRGhoqFNFn5i8vLyoXbs2jz/+OHPmzMHMj0i406dPc+rUqQQVfWrWrEnGjBlZv369S2J66L//oEgRyJjRJeeTe3O2D1MdoIRt26dcdN1ngQWWZaUBdgMaUiCS1J07B2vXQqZMZmZPjNFBIiKp2rFjpl1H48ZOj3K/10K6+uItIq4QGBhIuXLlKFGihGtPXLWq6ce+b58pdnt7w5gxrr2GiIiTFi5cSMaMGWnZsuV9923QoAHly5dn4sSJPP3001iJvQbZnj3mO3W+fIApwiT2fZ8nrimSkgUFBZEhQwZq164d73P07t2b3r17s3XrVmrWrJngmIKDgwESVPTx9vamXr16rFu3LsHxeNk2D/73n1n39tatBJ9P7s/ZmT5/AmEJuZBt23Nt2x4W8fO3tm2XtW27tG3bT9q2nfDmgCLiXuPHg7+/ebCpgo+ISHSFC8OPP8Ljjzt9iBbSFRF3OnfuHFu2bHH9LB8wA4AA9u69935XrphFe0VE3OTGjRt8/vnntGnThoxOjB63LIuXX36ZPXv28NVXXyVChDHUqwetW99zNriIJB+2bbN27Vrq16+PTwKelT355JNkzJiROXPmuCSu4OBgLMuiUqVKCTpPo0aN+PPPPzl06FCCzuN3+TLpbdt8Z27WzKzt8+KL8O678NlnsGkTHDhg7h1dQG3UnZ/p8xEQaFnWG8CFqB/Ytv2by6MSEbeK80K+p0/D++9D7dpQVktwiUjKFu/Fzk+dMmv3OMnPz48zDgo8WkhXRFxh1apVhIeHu2w9H4iSHx98ELJkgXs9ADh/3rR9CwuDadMgWzaXxSEiEmn9+vVcvHiRLl26OH1Mp06dGD58OBMnToxXO6bYOHUP2bo13LypNXFFUogDBw5w+PBhXnnllQSdJ1OmTHTq1IklS5YwefJkMmXKlKDzBQcHU6pUqQSfp1GjRoApovTr1y9e5zh27Bh/HjjAP0RZJ+bvv83rNwdlhaxZoU0bePrpeF0vso16ZFeNyDbqQKqa5ejsTJ/3gceAr4FdUV473RGUiLjXZ599xmeffeb8Ae+8Y76wt26dqDenqsyLiCfEOUdeuACPPGJGLcVhnTMtpCsi7hQYGEixYsUoX768y855Oz9almnxdumSWdMspps34aWX4PBhOHEChg836/+IiLjYwoULyZkz5+0Hk87w9vbmxRdf5JtvvmHnTtc91rrvPeT162aGZHi4y64pIp4VFBQE4JICcu/evbly5Qqff/55gs8VHBycoNZukR566CEeeOCBeK/rs2/fPh577DHOpk1rCj6ZMkGaNPc+6N9/zb1mPN2rjXpq4lTRx7Ztr1he9/lfSUSSvRMnYMYMqFMHypRJtMtqgXMRSTbWroWdO+NU8AEtpCsi7vPvv//y9ddf065dO/etVxEUBG+9BTG+VAOQNi307XvnC/u+fTB6tHq4i4hLXblyhS+++IInn3ySdOnSxenYPn36kClTptujvxPFzz9DhQqOR7aLSLK0du1aSpcuTZEiRRJ8rpo1a1KyZMkEt3g7c+YMp06dcknRx7IsGjVqxMaNG7l582acjg0ODqZWrVqEhYWx+rvvoEgR0z1o3TpYsMAsI/Hii6bVW82apmV6ZC7Pn//uE+7f79S9pNqoG87O9MGyrEKWZQ2wLGu0ZVm9LMvK7M7ARMR9Bg0axKBBg5zb+fRpk5gTeZaPKvMi4ilxypEAa9ZAzpzxKow3bNiQxYsXs2nTJhYvXqyCj4i4xJdffklYWJhLW7tBjPyYLh3kyQMhIY53rlULXnjhzu9bt8Irr5i+7T/8YO4xNdpdRBJg1apVhISE0LVr1zgfmy1bNvr06cPixYs5fvy4S+K57z3knj3mvXBhl1wvsakTh0h0V69e5dtvv3VZm0jLsujduzfff/89Bw4ciPd5goODAVxS9AFo3Lgx//77Lz///LPTx2zevJl69eqRKVMmtmzZwieffMKvadLA0aNmUFD+/GbWeJs28PzzMGYMzJ0LX30FS5ZAjRp3TnbjBsycCf37w8KF9712bO3SU1sbdaeKPpZl1QJ+A9oAxYEBwBHLspxfrVhEkoxdu3axa9cu53auUMGM4ixVyq0xxaTKvIh4Spxy5I0bZqRShQqm97CISBIQGBhIvnz5qF69ukvPGy0/Xr1q2rZ9/XXsB7RtC1Efxu7cCbNnw8iRZnuLFmbke1Ly6adm1Ol775lCldrSiSRZixYtolChQjz22GPxOv7FF18EYMqUKS6J5773kH/8ARkyOB7BnsSpE4fI3TZt2sSNGzdo1qyZy87Zo0cP0qRJw9y5c+N9ju3bt2NZFpUqVXJJTE888QReXl5Ot3hbtWoVTZo0oVChQmzZsoUHH3yQXbt2sePmTThzxrRvi42XlxlUFHUtoq++MsWe8HBzn7Z//z2vrzbqhrMzfcYDz9u23cC27a62bVcGhgKT3RaZiHjel1+aqefh4Ym+0KQq8yKSLGzZApcvQ8WK5gZVRMTDrl27RlBQEG3atMHLnXkpQwbYtcs8xLyXZ54xxR1Hrl2DXLmibwsPNwWhl16CadPMbMrgYDNC/vBhM0Po0iWzNoaj9YQcuXnTtCzeuhWWLoWAABg0yLQWiencOXP/u2oVvPqqme0+dCgsXw4nTzp3PRFxu7///puvvvqKzp07xzvXFS5cmI4dO/Lxxx/z770eQrrKnj1QsGCcWwInBerEIXK3tWvXkjFjRh5/3HVzIvLly0ezZs349NNP49xOLVJwcDAlS5YkU9TCSQLkyJGDqlWrOlX0mTdvHu3ataNChQp89913FChQ4PZnv2bNCo0bw8WLcQugeXN4+GHz861bMG6c4/bCEdRG3Ujr5H5lgMAY2xYA01wbjogkGX/9Zb7kNmsWvTVHIvH39ycgICDajWVqrMyLSBKXIQM88YQp+oiIJAHr168nJCTE5a3d7mJZpi3H77+b4ktsawdZFgwZYu4r9+2DQ4fgyBHzfuUKFCoUff+zZ01h5/RpMzPoXjJmNEWhqP7806xHmT69eTBw/DicOuW4B7yjhyEFC0b/PSwMfvnFvKZPN59Xr25eFSok+sAoETE+//xzbt68Ga/WblENGTKERYsWMXPmTF5++WUXRReLPXugZMlkmTfUiUMkOtu2CQoKon79+vjEmFWSUL1792b16tWsW7eO5s2bx/n44OBgnnjiCZfG1KhRI8aMGcOlS5fIli2bw32mTJnCoEGDqF+/PitXrryr6LQre3Zzj/bDD3G7eJo0Znb5M8+YQT9Hj8KsWaYtXCwaNmyY6oo8MTk7HOIwELNsWRU44dpwRCTJGD3aLMLbrNmdhdQSkSrzIpIsVK9uCuOahSgiSURgYCDZs2enTp067r9Y1aqmoPLPP/ff98EHzYyfgQNh0iRYscLMnol5n3n4sPPXd/Tg9MIF2L4dvv8efvzRFH1iW/TX0Toe9eqZ1sadO0PRond/fuKEiXvoUDMrXkQ8YuHChZQuXZoKFSok6DyVK1embt26TJkyhbCwMBdF54BtwzvvmMFCyZA6cYhEt3//fo4cOeLS1m6RmjdvTp48eZgzZ06cjz1z5gynTp1y2Xo+kRo3bkx4eDjffPPNXZ/Zts0bb7zBoEGDaNeuHV9++WXss4wyZjQt0uOqQAF47rk7v3/++f0HByU224a//04yrYGdnenzGhBoWdYS4ABQCOgODHFXYCLiPg899NC9dzhwAObPh6ZNzRd0D1FlXkQ84b45MtLff5sR6deuQSyjnUREElNYWBirVq2idevWpHPDoJ278mPVqqYd2x9/QK1acTuZZTleC+3RR03btcgZQYcPm6LStWvmdf36nXdHLZJi+6KdK5eZVRT5KlgQHnjg7v3y5jWvWrWgXz+T57dtM68dO8x1I8VcM8m24b//nP5PICLxc/z4cb777jtGjx6NFdsswzh4+eWXadGiBUuXLqVbt27xPs897yEtCxo2hG+/jff5PUmdOESiW7t2LQBNmzZ1+bnTpUtHjx49mDJlCufPnyd37txOHxscHAzg8qJPtWrVyJIlC+vWrYs2mzw8PJxBgwYxbdo0evXqxccff0zatHeXG27nx54978zKjquWLc2Anm3bzO/vvGPWinRRG7t4u3XLDDZasMD827y8zLPUsmWhXTtTsPIAp4o+tm2vtiyrNvAMUBc4Czxp2/ZGN8YmIm7y8ccf33uHt94yIydbtTKzfUREPOTvv//m6tWrPODowZyb3DdHRvr0U9Oy6IMPIF8+9wYlIuKEzZs3c+nSJbe1drsrP1atal736KseZ15eZpHz/Pnhfj3yHc3gqVAB3n3XFH9s23zRTsgaGn5+5p64VSszMvXXX83DhhMn7l6M/eJFyJw5ftcREactWbIEgM6dO7vkfE2bNqV06dJMnDiRrl27xruQdM97yL17TeE4TZp4RulZkYMxZ8+ezdmzZ/Hz88Pf31+DNCXVCgoKokyZMm77ntqrVy8CAgL47LPPGDx4sNPHBQcHY1kWlSpVcmk86dKl44knnmDdunXYto1lWYSFhdG7d28+++wzhgwZwoQJE2LNn7fz48CB8M03ZhCNr2/cgrAseOUV6N3brKt77pxZ/3H48AT+6+Lp5k34+mtYuDD67PHwcDOY/sCBu9e1tG0zqKlwYbevCez001zbtncBA9wXiogkCTdvmhGNjRpBsWKejkZEUoHLly9z8ODB268DBw7cfv8nomVQ+fLl6dSpE506daJ48eIejjjC6tWm9U8iFqRERO4lMDCQjBkzJt5DOD8/2LABgoIS53oxOXp4miMHVKvmnut5e98pdDkSHq4BUyKJYOHChVSrVo0HXdSVwsvLiyFDhvDMM8/wzTffUL9+fZecN5qPPoKZM83DwWRKnThEjCtXrvDdd98xcOBAt13j4Ycfpnr16syePZtBgwY5XYwODg6mZMmSsbdXS4BGjRqxcuVK/vzzTwoWLEjHjh1Zs2YNY8aMYfjw4c7FWKGCKfgcOQKlSsU9iJw5YfBgePNN8/v69fDYY1C7dtzPFV+hoebed/Fi8/w0qnTpzHNV2zYzkAoXjv75qVOmaJUpEzz8sJkNVLasW7os3fOO1LKsc7Zt57Es6x/AdrSPbds5XB6ViLhV3759gVhGIqVNCwEBsHu3vrSKiMuEhYWxf//+aAWdyPeYC8AWKlSIEiVK0KlTJ0qUKAGYxXpHjBjBiBEjqFy5Mp06daJjx44UjnkT5QL3zJGRLl0yU7ibN4cMGVweg4hIXIWHh7Ny5UqaNm1K+vjOarkPh/kxUyZT7AgPd/uIxSTPUbs6EXGpvXv3snPnTiZPnuzS83br1o0RI0YwceLEeBd97nkPuWePmXmo+0aRZO+bb77hxo0bbmntFlXv3r3p168f27dvp2psA05i2L59O/Xq1XNLPI0aNQJg2bJlrFu3ju+//54PPviA56KutROL2/kx4p39++NX9AGoWxd++MHMsgGYPh1q1Ei89chHjzbXjypjRmjbFtq3N89S9+wx7Ylj3hv//rt5v3LlTvtgMAOZpkxxaZj3e6LbIeK9jUuvKiIedeDAAccfHD4MYWEmOWlBRhFxkcuXL/PEE0/c7i8MZtHXEiVK0Lx5c0qUKMFDDz1EiRIlKF68OBkcfBl+6aWXOHr0KMuWLWPJkiUMHTqUoUOH8uijj9KpUyeefPJJCrioV26sOTKqdetMW6FHHjHTzEVEPOynn37izJkzbmvtBrHkx3nz4NlnzXss948bNmxIHS2BvL1N8UtE3GbRokV4eXnRsWNHl57X19eXF154gZEjR7J7927Kli0b53Pc8x5yzx4zktvbOwFRikhSEBQURKZMmXj8fm1oE6hTp04MGjSIOXPmOFX0OXPmDKdOnaJKlSpuiad48eIUL16cESNGkDZtWhYuXOh0m83b+fHhh02B4+jRhAXz4oum5W62bPDqq4lX8AHTsi2y6JM1K3ToAG3aRF9bKLZZ5zduQPbspiAUlY+PWVPShe45FMu27e8ifqxr2/a3MV9ATZdGIyKeNWSIqY6HhyfbXsMikrSEhYXRoUMHdu3axfTp09m+fTv//vsvZ86c4fvvv2f27NkMGzaMdu3aUa5cOYcFn0iFCxfm5Zdf5pdffuHPP/9k3LhxXL9+ncGDB1OoUCFq167N9OnTOXPmjPv/YWvWmBu8cuXcfy0REScEBgbi7e1N8+bNE/fCBQuaL7B//OHw4w0bNhAQEMCZs2exgTNnzxIQEMCGDRvue+oNGzbQuXNn6tWrR+fOnZ06RkRSLtu2WbhwIU888QT53LCe4rPPPkv69OmZNGmSa0986ZJp6eOhxbxFxHVs22bt2rU0aNAAbzcXcbNmzUqHDh1YuHAhISEh990/cpBl5cqV3RZTmzZtSJ8+PatWrYrfumrp08Mbb5jiT0JkymS6FL3/PhQpkrBzxebvv2H+/LvXkKxeHWrWhOefNy3eunePXvC5l1atYPly+Owz+N//TOeQwoWhTBmXP4e9Z9HHsqyBlmUNBEZG/hzl9RYw0qXRiIjn7NgBK1ZAvXqa5SMiLmHbNn369GHDhg18/PHHPP/881SuXJksWbIk+NzFixdn2LBh7Ny5k3379vHmm29y8eJFXnjhBQoUKMATTzzBzJkzueVokW9XGDvWjC5SKx8RSQJs2yYwMJAGDRq4JMfGSeRo0oMHHX48e/ZsroeGRtt2PTSU2bNn3/O0CSkWiUjKtH37dv766y+6dOnilvPnzJnz9qLkp0+fdt2J9+wx7wULuu6cIuIRe/fu5dixY25v7Rapd+/eXL58mRUrVtx33+DgYCzLolKlSm6LZ+zYsZw4cSJh//6RI6F8+buLKXFVqJB7ZvicOQOTJ0OXLjBnDmzaFP1zy4IxY8wMH1/fuJ/fsswggCZN4OWXYe5c0zLOxe7XdLltxCtNlJ8jXzWAYS6PSEQ84403IEsWM01Rs3xExAVGjRrFp59+yhtvvEHv3r3ddp2SJUvy2muvsXv3bnbv3s2IESM4efIkffv2df1IzUi+vlC6tPKliCQJv/76K4cPH3Zra7dYZc8OxYrBoUMOP465btv9tkeKb7FIRFKuhQsX4u3t7dZcN3jwYMLCwpg5c6brTlqxIixcaO4dRSRZCwoKAki0ok+dOnUoXrw4c+bMue++wcHBlCxZkkzOzjqJB29vb3LkyJGwk9y4ARcumFmQrnbqVPyLSYcPm8Gd3brBF1+Y5S8AFixwf/teN6ypfr/2bvVs264HLI38OcqroW3b01wekYi4XcWKFalYseKdDdu2mVZFTZqYSrmISALNmjWLt956i169evHGG28k2nUffvhh3nrrLfbt20fDhg2ZOHGiU1Pho7orR8Y0b56ZRq6Cj4gkEYGBgXh5edGqVSu3XifW/Fi1Kvz1l8Mv2X6xzCCPbXuk+BaLRCRlunXrFosXL6Z58+Zky5bNbdcpXrw4devWZf78+di2HadjY82RGTKY9j0JfVAqIh63du1aypYtS6FEenZmWRa9evXim2++4VAsA2wiBQcHu7W1W0JEy4+bN0P//rB7t+suYNume1Hv3rB0adyO/eMPGDHCHLthQ/QCT8mS4O/vujgTkbNlJH/LssYC39q2vc6yLH/A17bt990Ym4i4yeTJk6Nv+OknyJnT9JLUQ0wRSaCgoCCeffZZGjduzIwZM7AsK9FjsCyL1157jdq1azNz5kxefPFFp4+9K0fGNGGCmZJdo0bCghQRcZEVK1ZQq1YtcufO7dbrxJofu3UzvcyvXLmr7aW/vz8BAQHRZu34+vjgf58v0H5+fpxxUOC5X7FIRFKmzZs3c+bMGbp27er2a/Xs2ZPevXuzbds2Hn30UaePizVHzptnFi2vXt01AYqIR/z33398//33DBo0KFGv+9RTT/Haa68xd+5c3nrrLYf7nDlzhpMnTybZok+0/BhZ/PnrL6hTxzUXWL8epk41P8+ZYwYkPfhg7PvbNvzyi5mF+euvd39esaK5v61c2Xz3T4bu194t0kygGvBXxO/7gT6WZU1wS1Qikrh69zYPMdVjWEQSKDg4mCeffJLy5cuzbNky0rmjx66TatWqRe3atRk/fjzXr193zUmPHDEjkipWBB8f15xTRCQBDhw4wO7duz3T2i1Sy5YweLBp1xFDw4YNGTJkCHn9/LCAvH5+DBkyhIYNG97zlP7+/vjGyLPOFItEJGVauHAhmTNnpnnz5m6/Vvv27UmfPj3z5s1zzQlHjDAPJD14XywiCffNN98QFhaWaK3dIhUsWJDGjRszd+7cWNesDQ4OBkiyRZ9o/PwgTx44dsx152zQAEqVMj/fvAnjxjm8L73tv//g9dfvKvice+ghRmbPTr1du+g8cSIbvv7adTEmMmeLPk2ADrZt/wlg2/YWoDnQw12BiYj7dO/ene7du5tf9u2DvXshc2bwcjYliIjc7ciRIzRv3pycOXPy5ZdfkjlzZk+HxGuvvcapU6f45JNPnD4mWo6Mac0a8/7IIy6ITkQk4SIX9m3btq3br3XP/GjbcPKkw48aNmzI4sWL2bRpE4sXL75vwSfymPgUi0Tk/s6fP8/KlSs5cOBArA8Qk5LQ0FCWL19O27ZtSZ8+vduvlyVLFtq0acPixYsJjbG22L04zJGXL8OJE2aAZTIdLS4ixtq1a8mcOTOPPfZYol+7d+/eHD9+nI0bNzr8PDg4GMuyqFSpUiJH5py78mOFCmZAZRzbaMYqTRoYPvzOwMxDh8yMn9hkyWK6HUUe27gxP/bpQ8+jR9nyzz/YwJmzZwkICGDDhg2uiTGROdve7T8g5nDWMJwvGolIEnLixAnzw+bNUK+eGZnZooVHYxKR5O3ixYs0adKE0NBQNm3aRL58+TwdEgD169fn0Ucf5Z133sHf3x9vb+/7HnM7RzqyZo350n6vqeIiIokoMDCQqlWrJkpv+Xvmx44dzRfte62JFkcNGzZUkUfEDV555RU+/fRTANKnT0/ZsmWpUKEC5cuXp0KFCpQrV47s2bN7OMo7goKC+PfffxOltVuknj17smjRItauXet0Ud1hjty717wXKODC6EQksdm2TVBQEA0aNHDqO6WrtWrVihw5cjBnzhwaNWp01+fBwcGULFkySQy8dOSu/FipEmzaBCEhkDGjay7ywAPQr9+dNm9Ll5r70iNHTAviZ56Jvn/Hjub9ySchb16mdu4crR0xwPXQUGbPnp0s70edLfrMBtZalvUOcBjIC/wPWOLshSzLag3UsG17mGVZxYDPgVzA57ZtvxS3sEUkwWzbTGXMnRuqVLl7lo9tw4ULcPiwSZIx/6iNGWMW682ZE3LlMu+Rr1y5zGKVIpIqXL9+ndatW3P48GE2bNhA6dKlPR3SbZFr+zRv3pz58+cnrC1QeLiZBl6hgutuTEVEEuD48eP8/PPPjBs3ztOhmLUqVqwwLTXSOvs1U0QSW1hYGF988QUtW7akXbt2/Prrr/z222+sWLGCWbNm3d6vUKFC0QpB5cuXp0SJEqTxwBqwCxcuJHfu3NSvXz/RrtmgQQPy5s3LvHnzEjaT8o8/zHsiLfouIu6xZ88ejh8/zmuvveaR6/v4+NC9e3c++ugjLl68SI4cOaJ9HhwcTN26dT0SW7w8/bQp0ly/7trv1q1bw48/wvbt5rnm8OFme7p00KaNeV4Zyc8PXnjh9q9nHawlea/tSZ2zd+NvA5eBN4CiwDlgEfDm/Q60zOrNk4CuQGRvlYCIc60B1liWVc+27U1xC11Erl+/zoULF8iTJ0+cRxpUvnQJfvsNevY008y3bTOLSx45Yl5Hj5qKO8DMmXePav/xxzufO5I+vUmmOXLAK69EH9l04wZs3Wr2iXxlyHDnZx8fTX0XSSbCw8Pp2bMnW7ZsYcmSJdSuXdvTId2ladOmVK5cmXHjxvHUU0+RNr4PI728YNkyM0tSOUpEkoCVK1cCidPa7b6qVIG5c01/9mLFPB2NiMRi8+bNXLp0iWeeeYZWrVrd3m7bNqdPn+a33367XQj67bffCAoKut0CztfXl6pVqzJ//nwKFy6cKPFevnyZ1atX88wzz8T/Hi4e0qZNS9euXZk2bRoXLlwgV9QHhXGxf78ZQKmij0iytnbtWoBEX88nqt69ezN16lQWLlzIgAEDbm8/e/YsJ0+eTB7r+UQqXRp8feHnn117Xi8vGDrUrF1+5cqd7WFhsGqV2R4LPz8/zjgo8Pj5+bk2xkTi1F9M27bDgckRr/i43XDQsqw0QGWgnW3btmVZy4BGgIo+InHUunVr1q9fD0COHDnImzcvfn5+5M2b9/Yr5u+5cuUC22b43r1mFObSpXC/BSqPHIle9Ll27d4Fn8h9jh83r5gLVl6+DKNGxX6sl1f0gtAnn5gem+5i2ybO3bvNq2BB6NTJfdcTSUFeeeUVli1bxsSJE+kYOT06ibEsi5EjR9K2bVsWL14c+3oU93PrlskVidBLXkQkUmhoKH///fddr4sXLzJv3jzKlClDyZIlPR0mVK1q3v/4Q0UfkSQsMDCQjBkz3tWqxrIs8ufPT/78+WnSpMnt7aGhoezdu/d2IWjGjBn069ePoKAgrEQYBLNy5UquX79Oly5d3H6tmHr27MmkSZNYsmQJzz//fPxOMm4clC2bumaJ37pluoacOwdnz955f/xxT0cmEm9BQUGUK1eOggULeiyGChUq8MgjjzBnzpxoRZ/g4GCA5FX0AVPw+e03cHXrtNy54aWX4O23TbeOAgWgS5f7Xsff35+AgIBoLd58fXwS1i3Eg+5Z9LEsa5lt209alrUCcLiykm3b7e51Dtu2bcxsnlxAKUxLt38itgOcAurEcv2+QF+ABx544F6XEUl1fvzxR9avX0+vXr0oWrQoZ8+e5cyZM5w5c4aff/6ZM2fOcPXq1buO8/LyorS3N9nCwsyGmzdjv0imTFCkiKm+R5U2LQQEwN9/m5u5yPeLF+/8fOPGnf1j9oO+X8EoPByuXjUvb++7Cz4XL8L06VCmjLmBfvDBuLURCQ01I67++ONOoefy5TufP/xwki/6KD9KUjB58mQmTZrEwIEDeemlpN2ptVWrVpQrV44xY8bQpUuXe7YmqVGjxt0bw8OhRAmoUeOeo4MkaVCOFE+5cuUKISEhXL9+nWvXrnH9+vW7fnb0e0hICBcvXnRY3HF0PxfJ19eXCRMmJNq/z2F+jFS+vLkf+/PPRItH4k75MXW7desWK1asoFmzZqR3chCLj48PFStWpGLEel1FihRh4MCBLFiwIP4DaZx048YN3nvvPYoUKXLv/OMmkW3t5s+f71TRx2GMtm2+03pgDRC3sG0zev7sWfjnnzsF/0jBwWaUfXj43ccmg5yjHCmOXL58mS1btjB48GBPh0Lv3r0ZMGAAO3fupFKlSoAp+liWdfv3pMhhfnz3XTMDxx3r5dSrZwZ0X70K5co5NZA8cjDE7NmzOXv2LH5+fvj7+yfL9Xzg/jN9fot4/5VYij7xYAM3Y/x+y+GOtv0x8DFAlSpVXHV9kRRh9OjR5MqVi2nTppExllFDV65ciVYMinyFHjqEtXDhnR0zZzbFnSJFoHDhOz/nyOG4hVG6dPDII7EHF3kjeOECXLp090wfb2+oVcvMBop8hYTceY8sSIHjEfV79pgF3zZFTBD08YGSJU2xJvKVLZvj2C5dMou03avYdfBg9BiSIOVH8bTly5fz0ksv0bZtWyZNmpQoIz0TwsvLi5EjR9KpUyeWL19+z1lJDtfG2LHDrHHWtKl7Zx6KSyhHiidMmzaNgQMHxutYy7LInj07OXPmJEeOHOTLl4+yZcuSM2fOe74yJPIaivdcO8jXF+bPhzNnEi+gJGTDhg3MmzWLC+fOsbdwYcaMGUO3bt08HdZdlB9Tt59++omzZ8/Srt09x+7eU//+/Vm4cCGDBg2icePG5M6d24URRvf222+za9cuAgMDPXav2bNnT15++WX2799/31mVd+XIK1egf3946CGzdkRyc+sWnDhhvh//+eed98gBk15esH599HvjHDkcF3zAPB9I4pQjxZGNGzcSFhZGs2bNPB0KXbt2ZciQIcyZM4dp06YBsH37dkqWLEnmzJk9HF3sHN5DVqwIy5eb528xnxu6QokScT6kYcOGybbIE9M9iz62bY+OeB/lwmv+DeSM8nt+4JgLzy+S4v3yyy989dVXjBs3LtaCD0CmTJnIlCkTxYsXv/vDGzfg889h7Fh49FHXrk9hWaaQFNsfnLx54a23Yj/+5k2zmFtISPQZQ5F2747+e2iomRL62293thUoYKr6/fpB0aJ3tmfLBjlzmpFJUWXJYmYNlS1rWpJ4ed3znyiSmv3www9069aNRx99lAULFnhkQd/4aN++PaVKleLtt9+mQ4cOeMXl/8/XrDF54V4FbxFJ1WbPnk2ZMmV4/vnn8fX1JX369Pj6+kZ7xbbN29s7yRfPndKpE6xYYe7lEnHtDU/bsGEDAQEBhIeGkgY4evQoffv2BUiShR9JvQIDA/H29k7Qg8s0adIwc+ZMHnnkEV566SXmz5/vwgjv2LZtG2PHjuWpp57y6LplXbt2ZejQocyfP5+33347bgfv3WuK4Ul8Rnw0ISEwY4Yp7hw6ZL6XxyY83HT6yJPnzrbIn3PkMD/7+Zn3PHmir/MrkowEBQWRJUsWatas6elQyJ49O+3atWPBggVMmDABX19fgoODqVu3rqdDi7tKlcxyDqdOmQHo4lL3a++2ifvM8LFt+4m4XNC27VuWZe2yLKse8B3QHRgWl3OIpHZvv/022bNnj3df4fbt25Phxg3m58oFs2ZBtWpJa+R62rSmtVymTI4/b9bM3Dzu3m1m/TgaUXrypHlVrRq96AOmsOPtbWYElS1rpnoWKnSn8HX1atL67yGShOzfv59WrVpRuHBhVq1a5XRrkKQgTZo0jBgxgh49erBq1SratGnjcL/27dsDZjbTbWvWmBmF+rIqIg4cPHiQX3/9lffee4/+/ft7Ohy3cZgfo7p4EbZvN6PaixRJvMA8bPbs2VwPDSVq86aQkBBGjBihoo8kGbZtExgYSMOGDcmSJUuCzlW2bFmGDRvG6NGj6datW7Q1gFwhJCSEnj17kj9/fqZMmeLSc8dVvnz5aNiwIZ999hlvvfXWPQcN3ZUj//jDvMd3DRDbNoMgQ0PN68YNU4SJ+nOePHevo7Z/v5lVY1l3vuNG/hz5e+S6tnXqmPUvIvn4mNk79yr2+PreKebE7JCRMSOsW+e4nV0qnQkqyZtt2wQFBdGgQQPSuWM2Sjz07t2bRYsW8cUXX1C3bl1OnjyZ5NfzcXgPWaGCed+/X0UfN7jf8Ku5Ee/VgZbAdOAEUAzoB0yN53VfARZjZvzMtW07OJ7nEUl1fv31V1atWsWbb74Z76mbf//9N38DTJoEPXuaGT9JfA2baB54wLwiR3xduGBuqPfsMe8HDty5+dy9GyL+uNw2fLiKOiLxcOPGDZo1a0batGkJCgoiV65cng4pzjp37syoUaMYPXo0rVu3djiy/u+//46+4dQp05+8c+e71zgTEQGWLVsGQIcOHTwciXvdlR9jOnHCLFo+YECqKvqcjTmDPMKxY2poIUnHrl27OHLkCK+99ppLzjdixAiWLVvGs88+y+7du8kU24C9eBg2bBgHDhxg48aNZM2a1WXnja+ePXvSrVs3vv/+e+rUcbgkNeAgR+7ZY1oWxfVh5kcfwcqVprhzPx06QMzBoEuXwjffOHetHDngiShjudOkgeLF7xSscuY0a+iWKHHnPV++e3cKSSnrF4kAu3fv5sSJE4waNcrTodz2xBNP8MADDzBnzpzbzwWTetHH4T1k+fLm/fDhWI/bsGFDilljJ7Hdr73bpwCWZb0CNLFt+4/IzyzLWg3MAsY7cyHbtudG+fkgkLT/r1EkiXr77bfJkiVLvHvGR9O9u5lufvmy6debXAshuXKZEUqRN+A3bvDzggVsX7mSXZs3c3nv3uh/GJLrv1PEw7777jsOHTrE559/TrGYIwqTibRp0/Lqq6/i7+9PUFCQc+1N0qaFwYPNjMAkxKkb4JAQM6oy8vXff/DUU3efzLZd2+ZTJJVZtmwZNWrUoGB8R3OnFA8/bIrjhw55OpJE5efnxxkHhR8tAi5JSWBgIF5eXrRq1col5/Px8WHmzJnUqlWL119/nUmTJrnkvF9//TXTpk3jxRdf5Ikn4tRYxm3atGlDpkyZmDdv3j2LPnfZs8fM8rlHS3aHbNu5gg8k/P7t4MHoRR8wA0Nt2xR4cuRI2PlFkrmgoCAAl89oTAgvLy969erFW2+9xQMPPIBlWVSqVMnTYcVdtmxmcOXevQ4/jmyfez0iH545e5aAgAAAFX6c4Gyj5WLcve7OYeDeq9iJiEvt2bOH5cuX8+qrr5ItW7aEn9CyzNTr336DffuS5+KSDmz49lsCliy5/YcB/WEQcYnVq1fj6+tL06ZNPR1KgnTv3p0333yT0aNH07Rp0/uvo5EnD7RqZQomSUTUG2BvoOjZsxwYP54SGzdSxMcHTp82a5dFLrQbycvLFPyjFr+vXYNevcz0+urVoUoVs86ZiDjl4MGD7Nq1y2UPPJO1tGlNLvnrL09Hkqj8/f1vr+kTKUOGDIwZM8aDUYlEFxgYSJ06dVw6U/vxxx/nueeeY8qUKXTu3Jlq1aol6HyXLl2iV69elCpVyvGi3x6SIUMGOnTowLJly5g+fbrz7Y2vXTOtgWOb+XLlCqxaBe3aRZ9N7uNz5+d06czvMV/e3uY9ZitzMMWaa9dM4caOWLEh8ueov+fKZYr1MSXwf0eRlCQoKIjy5ctTIIm1+X766ad58803mT17Ng899FC8OwF5XIUKpvgcHn7X2tqR7XOjuh4ayuzZs/VszwnOFn1+AD6yLOtF27YvWJaVG5gAbHFfaCIS05gxY8iQIQODBg1y3UktC0qXhiVLzCieKlVcd24P0R8GEdezbZvVq1dTv359MmTI4OlwEsTb25thw4bRv39/Nm7cSIMGDWLf+fp1+PJLU/BxxYjt69dNwSVmP+i5c02BJmqv9shXzN/HjYuW57yBcWAWTt+27d7XDw+H8+chb94723btMgWi9evNy8sLypQxBaBHHzUtPjQLSCRWqaW1m9OqVzeLgF+7Bslo3beEiLy/nDdrFhfOnaNw4cKMGTNG6/lIkrFv3z727NnDc8895/Jzjxs3ji+++IJnnnmG4ODgBK15MXDgQE6fPs3WrVuT3LqRPXv2ZO7cuXzxxRd07tzZuYNWrjT3kbHdR332mfke/sUXMGgQ1KhhtnfrBl26mMJOfLpUdO5sXiKSIJcvX2bLli28/PLLng7lLkWKFKF+/fps3LiRKsn5Od7OnSYX9u4N2bNH+yi29rmxbZfonC369ARmAucsy7oG+ALfA13dFZiIRHfw4EEWL17MkCFDEjw6q379+tE32DYsWGAeGM6de/fDyGRGfxhEXO+PP/7g8OHDDBs2zNOhuESvXr14++23efvtt+8q+kTLkZs2mV7p//vf/denOH3azJz891/45x+4dOnu1/XrMGYM1KwZ/digIDh3zrngr12Lls+uAP8BDsd2pUtnCjx+fuY9b97oo0fB3GhHFR5u1kPbvRtmzza93KtVMw9yk3ivaBFPWLZsGY8++iiFklgLSHe46x7SkapVYepUM9unbFn3B5VENGzYkIa1a5scmsxnxErKs2LFCsC0KXO1rFmz8sEHH9CmTRsmTpzI8OHD43We5cuXM3/+fEaNGkXVqlVdHGXC1alTh0KFCjFv3rxYiz535ciQkNhPePw4RC5ofu6cKZRHinmvJpJE2LbNzp07qVixIl4xZmWkRF9//TU3b95Msp0uevfuzcaNG5P8ej5wj3vI48dhzRp4/HHzfTOK2Nrn+qWQLkXu5lTRx7bt00ALy7LyAIWBM7ZtH3drZCISzbhx4/D29uall15K8LnuWrzTxwc++ABatIBPP4VnnknwNTxJfxhEXG/16tUAtGjRwsORuIavry9Dhw5l0KBBfP/999SqVev2Z9Fy5Jo1ZqR6xYqxnyw8HBYtgjlzzM/3c+nS3duyZnW+6BMaelee+wbwAq5mykTnF180xZ18+cxoqft9IXv2WahfH376ycwU2rfvTusPgL//NkWpoCCoVQv693cuTpFU4M8//2TXrl2328imdE4tAN+6tckl91iUV0QS14oVK6hevbrb1h1r3bo1HTp04M0336R9+/Y89NBDcTr+zJkz9OvXjypVqvDqq6+6JcaE8vLyonv37rz77rucOXOGvFFnTUeIliNXroSAALM+joN9+fBDM+gSTIG8Xj33BC7iImfPnsXf358vv/ySoUOH8u6773o6JLdbu3YtWbJkoUbkLLwkpn379uzZs4cuXbp4OpT7ivUeMvJ79p9/3lX0iWyfG7WTj6+PD/7+/m6KMmVxuixrWVZzYBLwLpDOsqwebotKRKI5cuQI8+fPp2/fvg5vLl2ieXOzZsWyZXDihHuukUj8/f3xjTE6Sn8YRBJm1apVVK5cmfz583s6FJfp06cPefLkYfTo0Y53sG1T9ClXzsx2iU1oKHz1lXMFn7RpzWyfmDp1guefhyFD4NVX4c034d134b33zEOBOXPMjMzly+GRR+7Kc+8B7/v4kHPgQGjQwDw8yJnz/gUfMPuULAlPPWUGAAQGmhjq1797bZ8YN+IiqZ1auzmQObNpCxm1eCwiHnPs2DF++eUX2rZt69brTJs2jfTp09O3b1/CnbknimDbNn379uXq1avMmzcvQe3h3K1Hjx6Eh4ezaNGi++/800/mFaNdEQA//wxbt5qfLQsGDFArXUnSVq1aRbly5di4cSO1atVi/PjxfPnll4l27fHjxyfKtaKybZuvvvqKRo0aJdm85OPjw9tvv+2+54SJoUgRc+949OhdHzVs2JAhQ4aQ188PC8jr58eQIUO0bIOTnJrpY1nWAGAY8DHQHggFXrcsK79t2ym/tCviYe+88w5eXl4MHTrUJeeLnJoaFBQU/YPp082Dv/feM6OSkqnIPwCzZ8/m7Nmz+Pn54e/vrz8MIvF07tw5tm3bxqhRozwdiktlyJCBl19+maFDh7Jt2zaqRxQ0bufI8ePh2DFo3PjebS/Tp+enRo2oPGcOh4C/MmTgoapVebBKFciWLforY0bHX+qdaZkUhVvzXLZs0LChed26ZWb+bNtmHlxUq2a2iQhwp7XbA65Y8ysZiPUeMqavvjLrVLhghrqIJMzKlSsB3F70yZs3LxMmTKBPnz7MmTOHZ5zsHjFnzhxWr17N5MmTKV26tFtjTKjSpUtTpUoV5s2bx+DBg+/6PFqO3LMHChSATJmi73TzJrz/ftSDzHdwkSToypUrvPTSS8ycOZOKFSuyYMECihUrRo0aNejZsyc7d+506z3QDz/8QIcOHQgLC6Ndu3Y8+OCDbrtWTL///jsnT55Msq3dkptY7yEtywyydFD0gYj2uXqWFy/OzvQZArS0bfstANu2TwLtAPX3EHGzEydO8Mknn9C7d28KFCjgknNeu3aNa1F7BkcqVAgmTjTte65cccm1PKVhw4YsXryYTZs2sXjxYv2REEmAL7/8Etu2admypadDcbnnnnuOnDlzRpvtcztHbthgNjzySPSDYhQ9NmzYwKgFC3gRc2M0PiSEAT/9xAYfH9ObuGxZKFjQfOl34SjORMlzadLAww+bhTU//hhy53b9NUSSqb/++oudO3fy5JNPejqURBPrPWRMu3aZlpDJ8H5yw4YNdO7cmXr16tG5c2c2RP4tEEmmAgMDKVu2bJxbrsWHv78/devW5eWXX+b06dP33f/w4cMMGjSIevXq8cILL7g9Plfo2bMnu3bt4vfff7/rs2g5cs8e02o35vo8K1eaQUUAGTKAulFIEvXTTz9RqVIlZs2axbBhw9i2bRtlypTB19eXpUuXEhYWRufOnQkLC3PL9Y8dO0a7du0oUKAAXl5efPLJJ265TmzWrl0LQJMmTRL1uinVPe8hH3nEdM+I0sZNEs7Zok8WYG/Ez5Hz9A8AOVwekYhEM378eMLDwxNv8fT+/WH4cLMQudpyiAhmSn3BggWpeK91bZKpTJkyMXjwYL788kt27NgR/cPBg+Gjj6Bw4TvbzpyBF1+Er7++vWn27NlcDw1lDxBZDroeGsrs2bPdHr+IeI5au91D1apmNPu+fZ6OJE42bNhAQEAAZ86exQbOnD1LQECACj+SbJ07d47vv/+edu3aJcr1LMvi448/5vr16/ct4ty6dYunnnoKLy8v5s6dm2wWhe/cuTNp06Zl/vz5se907RocOmRm+kQd8HPpEsyde+f3nj0hhx6rSdJy8+ZNRo0axeOPP86NGzfYvHnz7TWmI5UoUYJZs2axdetWt6zDFRISQps2bbh+/Tpr166lSZMmzJ07l5uR62AlgqCgICpWrJii2psnWZMnm1bmjtqgS7w5+1d1I/B6jG29gR9dG46IRHXmzBlmzpxJz549KRz1oaO7+fnBzp3mYWcU8R35qBGTIsnX9evXWb9+PS1btsRKob3GBwwYQLZs2Xj77bejf3DtmlnTJn168/uWLdCnD/zxB0yadHv9s7Nnzzo8b2zbRSRlWLp0KdWrV081rd3ipEoV87537733S2Iii/hRqYgvydmqVasIDw9PtKIPmIfBo0aNYvny5bdbyzkyefJkvv/+e6ZOnZqs8mju3Llp2rQpn332Gbdia3n7zz/w2GNmfbOoPvkErl41PxcsCIn4v4uIMw4ePMjjjz/Om2++SdeuXfntt9+oXbu2w307duxI//79mThxIqtXr3ZZDLZt06tXL3bt2sWiRYsoXbo0/v7+nDp1inXr1rnsOvfy77//8sMPP6i1W2JJk8Z0lNBMH5dytugzAKhnWdZ5wMeyrKPAC8BzbotMRJg4cSI3btxg+PDhiX/x8+fNguEHDgDxH/mY3EZMbtiwgV69elG/QQOKFCnCggULPB2SiEdt2rSJkJCQFNnaLVLWrFkZOHAgK1asuN2q4/ELF0y7jWvX4MYNmDoVXnvtTqui0FBT/AH8/Pwcnje27SKS/KXG1m5xUrSoGb1+6JCnI4kTFfElpQkMDKRYsWKUL18+Ua87ZMgQKlSowPPPP8+///571+e7d+/m1VdfpU2bNvTs2TNRY3OFnj17cvr0aTZu3Oh4h/z5zYyeSpXubDt0CNasufN7//73XjNSJBHZts2sWbOoVKkS+/fvZ/HixcybN4+sWbPe87iAgAAqVarEU089xdFY1mSJq7Fjx7J06VLeeecdmjVrBkCLFi3InTt3og3C2LBhA7du3VLRJ7HYtlkHMqKlnriGU0Uf27bP2bZdA2gCdAa6ARVs205ed/Eiycj58+f58MMP6dq1q8sXq2vRogUtWrS4904BAZA5sxnNbtvxHvmYnEZMRhaozp4/jw0cPXqUvn37qvAjqdqqVavImDEj9erV83QobvXiiy+SKVMmxowZQ4sWLeiRJo1Z0+faNXj+eVix4s7Ofn4wZQo0bgyY/vW+Mfq1+/r44K8e7SIpVmpt7ebUPSSYdkY1apiieTJqF6wivqQk//77L19//TXt2rVL9Nna6dKlY9asWZw5c+auNuU3btygZ8+eZM2alRkzZiTLmeQtWrQgW7Zsd7V4u50jbRv+/ht8fe98+MAD8NxzkDGjaYH56KOJHLWIY+fPn6dNmzb06dOHRx99lN9//51OnTo5dayvry/Lli3j1q1bdOrUiRs3biQoli+++IKRI0fSrVs3Xnnlldvbvb296dmzJ6tXr06UgRhBQUFkzZqVGjVquP1aqcU97yEtC/78E3bvTtygUrj7Fn0sy0pjWdbflmWlsW072LbtZbZtb7Ft2z0rdYkIAO+99x7Xrl1jxIgRLj/3yy+/zMsvv3zvnXLlgvHjYf9+WLky3iMfk9OISUcFqpCQELf8byCSHNi2zZo1a2jUqBG+Ub+0pkA5cuRgwIABLF26lJaNG1PyyBHIkwdeftncgEaqVQtmzoSyZW9vatiwIUOGDCGvnx8WkNfPjyFDhtCwYcNE/3eISOJYtmwZ1atXT9z2u0mAU/eQkVavhpEjk1V/dhXxJSX58ssvCQsLS9TWblFVqVKFQYMG8dFHH7Fly5bb20ePHs3OnTuZOXMmefLk8UhsCeXr60unTp0IDAzkv//+u739do7s1MmsARk1n6RNCx06wGefmRHtybDYJSnP2rVrKVeuHOvWrWPSpEmsX7+eggULxukcxYsXZ/bs2Wzbti1BXWp+//13unfvTtWqVZk5c+ZdBeHevXtz8+ZNPvvss3hfwxnh4eEEBQXRqFEj0qZN69ZrpSb3vYesVAmOHYPY2mZKnN236GPb9i0gEOjn/nBEBODixYtMnz6dJ598klKlSnkuEH9/qFYN5s6lcO7cDne538jH5DRiMrZC1LFjxxI5EpGkYdeuXZw4cSJFt3aLavDgwfj6+rLmxRfh8mWzFkXkw8p06WDgQHjzTTMLMoaGDRuyePFiNm3axOLFi1XwEUnBDh06xI4dO9Ta7X4sy8yMvHbN05E4TUV8SUkCAwPJly8f1atX91gMb731FkWKFKFPnz5cv36dn376ibFjx/L000/TunVrj8XlCj169CAkJITAwMC7P9y1y8z2cTRoKls2yJvX3eGJ3NO1a9fo378/zZs3J0+ePPzyyy8MHjwYLy9nVwGJrkOHDgwYMIBJkybxxRdfxPn4Cxcu0Lp1azJnzsyKFStIH7mmahRlypTh0UcfZfbs2dhunEW8YsUKTp8+rfu8xFaxollm4sIFT0eSYjj7/82lgWmWZR2xLGtH1Jc7gxNJiKtXr3Lz5k1PhxEvU6dO5b///nPbDJO6detSt27d++/o5QVz5sCECfTu1CleIx+T04jJ2ApRyWlhURFXWrVqFZZl0bx5c0+Hkijy5MnDs/360WnTpugfFCoEH3wAbdtqVKaIpNrWbhCHe0gwi5U/91yy68+uIr6kBCEhIQQFBdG2bdt4P8R1hYwZMzJjxgz27dvHyJEj6dmzJwULFmTy5Mkei8lVatasSbFixaK1eKtbty6NateGv/6CAgV03yhJ1nPPPceHH37Iyy+/zC+//EK5cuUSfM6JEydSuXJlnn76aY4cOeL0cWFhYTz55JOcOnWKFStWUKBAgVj39ff3Z+/evfz0008JjtcR27YZO3YsDz74oMdmSaZU972HrFDBvO/blyjxpAbO/vUPBHoDbwBTYrxEkpxDhw7x4IMP0rhx42RX+Ll8+TJTpkyhdevWib7gpkMPPwxPPUXtunUZ+vzzcR75mJxGTDoqUGXIkIExY8Z4KCIRz1q9ejWPPvposm29ER+vDB3Kz1E3NGoEM2aAi9dWE5Hka+nSpVSrVi3VtXaLs4wZ4fhx0ypYRBLV+vXrCQkJSRIPLRs1akSPHj0ICAjg4MGDzJ07976LwycHlmXRs2dPvvnmG44fP357e6GQEAgPh4IFzaj1YcOitwoW8bCgoCA+/fRTRowYwYQJE/CJ8Qwkvnx8fFi6dCm2bcdpfZ9BgwaxefNmZs6ced+ZiZ06dSJjxoxuWyN6/fr17Nixg2HDhpEmTRq3XENiUbEiPPIIhGk1GVe5Z9HHsqxHLMs6BkwEXgWCbdv+NOorUaIUiYNz587RuHFjLl++zDfffMPrr7/u6ZDi5P333+fSpUu89tprng7lDh8f+OUXnlixgsULFsR55GNyGTEZWaDyy50bCyhcuDAff/wx3bp183RoIonu5MmTBAcHp5rWbpHy5cvHv9myEQ7QqxcMHw4O2guISOqk1m5xVKUKHDpk2hyJSKIJDAwkR44c1K5d29OhADBp0iSKFi3Kq6++Sr169Twdjst0794d27ZZsGDB7W2FQ0LMDwUKmHUgt22Dfv1g0SIPRSlyx+XLl+nXrx9lypRxyzOnYsWKMWfOHH7++WeGDh163/1nzJjBBx98wCuvvEKPHj3uu3/mzJnp2LEjS5Ys4cqVK64IOZqxY8dSsGBBp2IRF8ubFzZuBBfMOhPjfjN9pgKfA+WBIGC62yMSSYArV67QokULTpw4wYYNG+jTpw/jxo3jyy+/9HRoTrl69SqTJk2iadOmVK5c2dPhRFelChw+DFFuaFOihg0b8sknn7Dx6685cuSICj6Saq1ZswYg1RV9AFaWLs11Ly/44w9PhyIiSUxqbu0WL9WqwalTcPGipyMRSTXCwsJYvXo1rVq1Il26dJ4OB4BcuXLx559/prgOCsWLF+exxx5j3rx5t9cYueDtDU2bwo0bsGGD2TE8HEqW9GCkIsawYcM4ceIEs2fPdtkMn5jatWvHwIEDmTJlCitWrIh1v++++44BAwbQtGlTxo0b5/T5/f39uXLlCkuXLnVFuLdt2bKF7777jpdffhlvb2+XnluclCkT3LypwUIucr+iTyVguG3bu4HXI34XSZIi+4AGBwezZMkSatasydSpU6lUqRI9evSIU09RT/noo4+4cOFC0prlE6l9e9PmaNEi06pDRFK01atXU7RoUR5++GFPh5Lo/vX2ZmnBgvDzz/Drr54OR0SSkGXLllGtWjWKFCni6VCShypVzPvvv3s2DpFUZPPmzVy6dClJtHaLypNrC7lTz5492bt3Lzt2mCWvf8uWDV54AaKs9UOtWqZtkYgHbd68mQ8//JDBgwfz6KOPuvVaEyZMoGrVqvTq1YtDhw7d9fmRI0do3749xYsXZ9GiRXFqpVazZk1KlizJnDlzXBky48aNI1euXDzzzDMuPa/Ewfvvw7PPQuSMSUmQ+/3VTWfbdiiAbduXgaQxTEQkBtu26dOnD1999RUfffQRrVq1AsDX15dly5YRHh5Ox44dCQ0N9XCksbt27RoTJkygfv361KhRw63X6tixIx07dozbQZYFH34IadOaRXl/+ME9wYmIx129epWNGzfSsmVLrFS4AG3Hjh0JGzgQsmeHWbM8HY6IJBGHDh0iODg4Vbd2i/M9ZNWqUL+++wISkbsEBgaSMWPGJNtSO6V58skn8fb2Zt68eXTs2JGnmjSB1avhwAGzQ7p05iGmiAeFhITwzDPPULx4cUaPHu3263l7e7NkyRIsy6JTp07RnsVduXKF1q1bExYWxqpVq+K8xpdlWfTu3ZsffviBffv2uSTeXbt2sXbtWl588UUyZszoknNKdE7dQ+bMCVeuaB00F0nr6QBEXGHEiBF8+umnjBo1ij59+kT7rHjx4sydO5e2bdsyZMgQpk9Pml0KZ82axdmzZ1myZInbr9W/f//4HVisGGzfbmb9XLkC165prQuRFOjrr7/m+vXrtwvoqc3tHJkuHQQFwaVLkC2bJ0MSkSTg888/B1J3a7c430PmzAmBgbB2rXsCEpFobt26xYoVK2jevDm+vr6eDidVyJ49O61atWLRokWcPHSIdFmzQtTWUB07Qv78ngtQBHj99df566+/2LRpExkyZEiUaxYtWpRPPvmEtm3b8sorrzB16lTCw8Pp2bMnu3fvZu3atTz00EPxOnfPnj159dVXmTNnDuPHj09wrO+88w6ZM2fm+eefT/C5xDGn7iErVjTvBw9ChQpujSc1uN9MHy/LsnpYltXTsqyeQJqov0dsE/Go6dOnM27cOPr06cPrr7/ucJ82bdowZMgQ3n//fRYvXpzIEd5faGgo7777LrVq1aJOnTpuv15ISAgh8Z0uWbIk/PYbDBxoplwuXgzHjrk2QBHxqNWrV5MlSxZq1arl6VA84naOHDgQZs+Gq1c9HZKIJAFLly6latWqqbq1W7zuITNlguvXzZoWIuJWP/30E2fPnk1yrd1Sup49e3L+/Hk2ffyxyXXXr5sPcuYErRErHrZt2zbee+89nn32WerWrZuo127Tpg2DBg1i2rRpLF++nLfeeosVK1YwceJEGjduHO/z5s2blxYtWvDpp58SFhaWoBgPHDjA0qVL6d+/P9mzZ0/QuSR2Tt1DlixpiuZHjyZOUCnc/Yo+J4C3gDcjXqdi/D4qvhe2LGu7ZVn7Il5J7ym8JAuff/45AwcOpFWrVnzwwQf3bEM0btw4HnvsMZ555hn27t2biFHe39y5czl58mSireXTrFkzmjVrFv8TeHlBvnzw2GOwdCn07WtGw2uxNZFkLzw8nDVr1tCkSZNUu4BltByZLx+cPw9bt3o2KBHxKLV2M+J1DzlrFvTuDWfOuCcoEbktMDAQb29vmjZt6ulQUpUmTZqQK1cuNsUchNq3rzpjiEeFhobSu3dvChQowLvvvuuRGN59912qVatGz549efPNN3n66acZNGhQgs/r7+/PuXPn+PLLLxN0nvHjx+Pj48PgwYMTHJPEzql7yHTpoEwZFX1c5J5FH9u2i9i2XfQer2LxuahlWWmAy7Ztl4p4dY5X9JKqffvtt3Tr1o0aNWqwaNEi0qa9d7fCdOnSsWTJEjJkyECHDh246uGR2zdv3uT7779n+PDhjBw5kurVq9OgQQOPxhRnefOaRc7Ll4fx42H0aC24JpLM/fLLL5w9ezbVtna7i22b9cymT4dbtzwdjYh4iFq7JcBDD5lc+scfno5EJEWzbZvAwEAaNmxIlixZPB1OqpIuXTq6dOlCz6jPGEqXhuT2/V5SnDFjxrBnzx5mzJjhsbzg7e3N0qVL8fX1pUaNGnz00UcuWTe2adOm5MuXjzlz5sT7HMePH2fevHn4+/vj5+eX4JjEBZ55BipXhps3PR1Jsne/mT7ukh847aFrSwrw+++/07p1a4oVK8bq1aud7klaoEABFi5cyN69e3n22WexE3lmyoULF5g/fz5dunQhT5481K5dm4kTJ1KuXDlmzJiRPBdML1TIjID/3//g22/NIpUJnF4rIp6zevVq0qRJoxGikby8YMwYOHUKvvjC09GIiIcsW7aMqlWrUrRoUU+Hkvw88ghY1p1FzUXELXbt2sWRI0fU2s1D+tSuTamoGwYMMPeRIh7y66+/Mm7cOHr06OHx73aFCxfmwIEDbN68GR8fH5ecM23atDz11FOsXbuW06fj94g5ICAA27Z55ZVXXBKTuMDzz8PTT99pkynx5qm/QIWBCpZl/WpZ1jbLsh5ztJNlWX0j2sBtP3/+fCKHKEnVsWPHaNKkCRkzZuSrr74iR44ccTq+QYMGjBo1is8++4yZM2e6KUrDtm127NjB22+/TY0aNciTJw89e/Zk06ZNtGnThs8//5wLFy7wzTffUCE5L1KWJg288w6sW2d6Fp87Zwo/avfmNsqPqdvhw4cZN24cFSpU4NFHH+W6C2+IVq9ezWOPPRbn3JqitW9vFpVcuBBCQz0djThBOVLWr19PkSJFmDBhQoLPdfjwYbZv357qW7vFW5YsUKIEHD7s6UgE5ceULDAwEC8vL83W9pCy7dszL00a/gVo2NC0KJJkJ6XkyJs3b9K7d29y5szJ5MmTPR0OADlz5nR5+/DevXtz69YtPv300zgfe/78eT7++GO6detG4cKFXRqXJFDatKbFuiSIp4o+V4FZQBXgWWChZVnpYu5k2/bHtm1XsW27Su7cuRM7RkmCLl68SJMmTbhy5QpfffVVvBPzyJEjady4MQMHDmTHjh0ujfG///5jxYoVPPPMMxQoUIDKlSvz+uuvEx4ezqhRo/jll184deoUc+bMoX379mTNmtWl1/eoBg1g7FioXt2s8TNsGPz7r6ejSpGUH1OfM2fOMHXqVGrUqEGxYsV49dVXSZcuHdu2bePNN990yTWOHj3Kb7/9pocFMVkWTJgAf/8NS5Z4OhpxgnJk6mXbNhMnTqRp06b8888/DB06lLfeeitBs7vV2s0FqlWDQ4fUJjMJUH5MuQIDA6lTpw65cuXydCipkmVZXMqbl6wALVt6OhyJp5SSIydOnMiOHTt4//33U/RgvhIlSlCrVi3mzJkT53u9KVOmcP36df73v/+5KTqJl3//hccfh/XrPR1JsnfvRVDcZx/wu23bN4GdlmWdBfyAEx6KR5KBa9eu0bJlS/766y/WrVtHuXLl4n0uLy8vPvvsMypVqkSHDh0IDg4me/bs8T5fWFgYgYGBzJo1i2+//ZawsDCyZMlC48aNad68OU2bNiVPnjzxPr+rPf300+47uWXBgw/CAw/AtGlm4d7XXjOj5EUkTi5dukRgYCCLFi3im2++ITw8nPLly/POO+/QqVMnihQpgr+/P+PHj6dt27ZUq1YtQddbvXo1AC1T+RdVhzmyQQNo1crMYAwPV7sOSbVCQkLw9fXFKwn+/8C1a9fo06cPCxYsoEOHDsyePZuBAwfyxhtvEBoayttvvx2vVrrLli2jSpUqau1GAu4hn3oKcuWCK1cgJQ14Ekki9u3bx549e3juuec8HUqq5vfKK4S/9BJe69dDAp5XiCTEvn37GDVqFO3bt6d9+/aeDsft/P39efrpp/n++++pXbu2U8dcvnyZ6dOn065dO0qXLu3mCAXicA+ZNat5nnj0qFvjSQ08VfR5EcgNDLEsqziQFTjloVgkEaxatYqpU6dSrFgxKlSoQPny5SlXrhzZsmVz6vibN2/SpUsXtm7dypIlS6hbt26CY8qVKxdLly6ldu3a9OrVixUrVsT5QcDFixeZOXMm06dP58SJExQrVoxBgwbRvHlzatasSbp0d01gSxLcWvSJNGgQ1KgBHTrASy9B9+7mC3+aNO6/tkgyFhISwurVq1m0aBFBQUHcuHGD4sWL8+qrr9KlSxfKxGgVMWnSJNavX8/TTz/Njh078PX1jfe1V69ezUMPPcRDDz2U0H9GshZrjvziC9i+3bQoSsYj/0Tia+3atXTq1In+/fvz7rvvejqcaI4fP07btm1vt9V99dVXsSyLOXPm4O3tzdixYwkNDWXChAlxut87fPgwv/zyS5L793pKvO8hGzSAAgXgt99U9BFxgxUrVgDQpk0bzwaSynV+8UXYuNGsdfvCC+DiVlYi93Pr1i38/f3JmDEj06dP93Q4iaJDhw688MILzJ492+miz4cffsi///7L8OHD3RydRIrTPWSFCrBjhxlwmRzXPk8iPDVEbzpQwrKsP4GlgL9t2+EeikXc7IsvvqB9+/bs37+f5cuXM2DAAGrXrk327NkpXLgwLVu2ZOTIkSxdupR9+/Zx8+bNaMfbts2AAQP44osvmDJlikv7qdeoUYMJEybwxRdfEBAQ4PRxe/fu5dlnn6VgwYIMGzaMhx56iFWrVnHw4EHGjx9PnTp1kmzBB+DChQtcuHDB/ReqXh3++APatIH58yE42IyQP38e/vnH/dcXSSZu3LjBmjVr6NatG3ny5KFz58788ssvPP/88/z8888cPHiQ0aNH31XwAciaNSuzZs1i7969jBo1Kt4xXL58mU2bNqm1G/fJkSVLwnffwcWLiRuUiId98MEHtGzZktDQUN5//33+SUJ/x7ds2UKVKlU4cOAAX3zxBSNGjLhd2PHy8mLGjBkMGDCAgIAABg4cSHi48187Ilu7aT0fI0H3kJkzw+jRsHat1n0UcbHAwECqV69OwYIFPR1KqnbhwgUud+oEly+b4o9IInv//ff58ccfmTx5Mnnz5vV0OIkiY8aMdOnShWXLlnH58uX77n/t2jUmTZpE48aNqVy5ciJEKBDHe8iKFeH0aS0XkUAemelj2/YVQE+VUoE1a9bw5JNPUrlyZdavX0/mzJk5deoUv/32G7/99hu//vorv/32G0FBQdyK6PHt6+vLww8/TPny5alQoQJHjx5lxowZDBs2jBdeeMHlMb744ots2bKFYcOGUb16dWrVquVwv/DwcNavX8/kyZNZt24dPj4+dO/enYEDB1K+fHmXx+VOkT3pN2/e7P6LZckCgYHw2WeQP78p+MyZY/pzFisGVapA5cpQvjwkYIaCSHIVEhJCtWrV+OOPP8iePTtdu3ala9eu1KpVizROzoxr3Lgx/v7+TJgwgbZt21K9evU4x7F+/XrCwsJSfWs3uE+OPH0aPvjAzPZ5+eXEDUzEA8LDwxk6dCgBAQG0aNGC4cOH89hjj/Hxxx8niR7oM2bM4IUXXqBIkSJs3rzZYYsOy7KYOnUqPj4+BAQEEBoaykcffeRUi7ply5ZRuXJltXaLkKB7SF9fyJDBrJH288/wyiuQMaNrAxRJhY4dO8b27ds1IzEJ6NChA2lsm40NG6oVsCS6w4cPM3z4cJo2bUr37t09HU6i8vf35+OPP2bx4sX07dv3nvvOmTOHc+fOaZZPIovTPWSFCmbA+IEDZl1IiRdPtXeTVOCrr76iffv2VKhQga+++oosWbIAUKBAAQoUKEDTpk1v7xsaGsrevXujFYPWrFnDJ598AkDPnj0ZO3asW+K0LIvZs2fz66+/0qlTJ3bu3Imfn9/tz69evcq8efOYOnUq+/btI2/evIwePZp+/fqRnBf2S3SRNx03b4KfH5QqZaa9BwbC0qWmVdL8+eDjA2fOmN/VCk5SgVGjRvHHH38wb948OnXqhHc820AEBASwbt06nn76aXbu3BnnNm+rV68me/bs1KxZM17XTzUeegi6dYPFi6FLF9OuSCSFCgkJoUePHgQGBjJgwAAmT55MmjRpaNCgAVOnTmXw4MHxzlkJdePGDQYOHMiMGTNo0qQJixYtumfbYMuymDBhAj4+PowdO5YbN24we/bsexbXjxw5otZurpQrF2zdatZ6fPdd2LcPXn8dHMxiFRHnRbZ2a9u2rYcjEYBblgVffQWrV8ONG2rxJonCtm369OlDmjRpmDFjRrzWMEzOqlatStmyZZk9e/Y9iz5hYWGMHz+emjVrOt0KTjzgscfgnXfUEjiBNPRA3OLrr7+mTZs2lClThnXr1t137R4fHx8qVqxIz549mThxIhs2bODcuXOcOXOG7du3M2fOHLf+0cqaNSuff/45//zzD127duXWrVscP36c//3vfxQsWJD+/fuTMWNG5s+fz9GjRxk5cqQKPvGVNi3UrQtTpsCuXabAs3AhvPgiXL8Op07BgAHQurV5KLBqldlHJAXasWMHAQEB9O3blx49eiTo4Wlkm7fIhTvj4tatW3z55Zc0b96ctGk1HuS+Ro8277NnezYOETc6e/Ys9erVY8WKFUyePJlp06bdLpAMGTKEU6dOsWTJEo/FVr9+fWbMmMH//vc/1qxZ49Q6kZZlMWbMGN566y0+/fRTevbseVdb4ajU2s0N0qSBsWPvtD0aN87c/4lIvAUGBlKuXDlKlCjh6VAkkpcX5M1r2puLJILZs2ezceNGJkyYQKFChTwdTqKzLAt/f39+/vlndu/eHet+Cxcu5NixY7fXfpQkKm9eeOYZFX0SSE92xOU2bdpEy5YtKVmyJF9//TU5cuSI97n8/PyizbpxpwoVKvD+++/j7+9PtWrV+PXXX7Ftm3bt2jFo0CBq1qypPwrukDOnGS0Ppr/7f/+ZtTLWrTOjQbdsMTfNzz0HEdNBRVKCsLAw/P398fPzc9ko8vi2edu6dSt///23Wrs5q3Bh6NfPtHn780948EFPRyTiUnv37qVZs2acPXuWFStW0Lp162ifN27cmIcffpiAgAC6d++eqPdH27dvp23btvz9998sXLiQLpH3EHHw2muv4e3tzbBhwwgNDWXhwoUOi+5q7eZGdevCnj1mkd7z500xKG1aSMD3BpHU6Ny5c2zZsoXXXnvN06FITMOGwd69Zp1btXoTNzp58iRDhgyhbt269OnTx9PheEz37t0ZOnQos2fP5r333rvr81u3bjFu3DjKly9Ps2bNPBChxMnZs2ageJMmno4k2dJfHnGp7777jhYtWlC8eHG+/vprcubM6emQ4qR379707duXv/76i8GDB/PXX3+xbNkyHnvsMRV8EoNlmTWABgww0+HPnjU93zt0MC2Uzp2Dq1chYv0nkeTsvffeY9euXUyfPt2pEerOCggIIH/+/Dz99NNcd3L09OrVq0mbNi2NGzd2WRwp3muvQcmScPy4pyMRcalNmzZRs2ZNrl27xnfffXdXwQfMaMqXXnqJX3/9lW+++SbRYluwYAG1atXCy8uLH374IV4Fn0j/+9//eO+991i+fDkdOnQgNDQ02udHjhzh559/1iwfd8qRAxo0gGbNTIvfXr1g2zZPRyWSrKxatYrw8HDatWvn6VAkpqefNt9nldfEjWzb5tlnnyUsLIxZs2Y5tV5hSpUrVy5at27N/Pnz77qvA1i5ciX79+/XLJ/k4v33Ydo0cPC/pThHM33EZX788UeaNWvGAw88wMaNG5Nt+7OPPvqIDz/8MEX/sXzuuec8HYJz0qSBqlVhyRLT+mPPHhgyBI4cMe3gKlZ0/TVv3fp/e3cep1Pd/3H89TUYBhMxFBNKcleI0i6EQcge3ea2FMZadrnJkq3CRIudn6XbGspWlsmS4r4jQlJaZMuSsi9jmfP748xMY1wzZsZ1zTnXzPv5eFwPM9dyzmfOzLyduT7n+/3azaWiRb2/bZFYP//8M4MGDaJBgwZe/yM9bpq3WrVqMWjQoBSNIlq6dClVqlThNg2fBlKYkQULwo4dsGIFXLwIOXP6vjARH5s5cyZt27alVKlSrFixgmLFiiX53PDwcPr160dkZCTVqlXzaV3Xrl2jb9++jB49mkqVKrFw4UKvnGd269aNwMBAOnXqRIMGDVi8eDE5Y3+XNbWbZz45h7ztNhg2DBo3tq+Mb9zYHk2ZLZv39yXip86dO8fhw4c5dOgQhw8fjv945cqVlChRgjJlyjhdopAoI5s0sS9m/OwzePJJ54oSvxATE8OlS5eIjo7m8uXLREdHp+i2Z88eli9fzjvvvEOJEiWc/jIc16ZNGxYuXMiyZctokmCmGMuyGDFiBPfee+9190v6SfU55EMPwYULcOAAaPrSNFHTR7zif//7H7Vq1aJIkSKsXbs23aZk8wVjTIbv+jdr1szpElIvRw54+GH7KtA+faB7d3txt86d4c47b337Bw7YJ+Rr1thv3s6YcevbFPHAsizat29P9uzZ+eCDD3yyj5o1a9K2bVtGjx5Nw4YNeeKJJ5J87k8//cQPP/zgP83gdJDijMyWDe6/H957D/TGsPgxy7IYPHgwQ4YMoXr16ixcuPCmTeDAwEC6dOnCgAED+P7773nggQd8UtuZM2d44YUXWL16NZ07d2bMmDFk82IzoGPHjmTPnp127drx/PPPs2TJEnLlysVHH33Eww8/zD333OO1fWUEPjuHfOgh2LULOnSAWbPs6TyGDIHChX2zPxGX2bt3L3v37r2hqRP38enTp294Tb58+ShSpAg9e/bM8H+/+ovrMjJXLnjxRTvTTp/W2hSSpOPHj/PEE0+wb9++NL2+SpUqvPrqq16uyj+FhYVx1113MW3atOuaO2vWrGHbtm1MnTo1fo1KSV+pPoeMu8j7hx/U9EkjNX3klm3dupUaNWpQsGBB1q5dy53eeANefOpg7HREfrnAX8uW9lVTgwfbb7S2agWvvQa3cpXxjh3QrZs913K5cvb2fTGKSASYMWMGa9euZeLEiRQpUsRn+4mMjGTVqlW89NJLbN++nRw5cnh83rJlywC0nk8CqcrI1avttX2KFIGnnvJxZSLeFx0dTdu2bfnPf/7Dyy+/zMSJE1PcVOnYsSMjRozgnXfeYerUqT6pr0ePHnz++edMmTKFtm3b+mQfbdq0IXv27LRu3ZratWszfvx4vv76a9566y2f7M+f+fQcMmdOmDnTnru9e3d7WqRCheyR3yIZ1Llz5+jduzcTJ06Mvy9LlizccccdhIaGUqpUKapVq0aRIkUIDQ2N/7dw4cIEBQU5WLl4ckNGtmsHU6fa69TWqeNgZeJmb7zxBgcOHGDo0KHkzp2bwMDAG27Zs2f3eH9gYCCFCxdWIyNWQEAArVu3ZtiwYRw8eDD+d3HEiBGEhobSokULhyvMvFJ9Dlm6tP0e3f79PqwqYzOWZTldQ4pUqFDB2rp1q9NlSCLbt2+natWq5M2blw0bNlBUU2L5hSpVqgCwfv16R+u4Zb/9Bj162A2fggUhMBBy505+ocyYGPvq0ZUr7Tdpa9e2p4778kt4+WX7StMk3hy/GWPMN5ZlVUjTi2+B8tF/HDt2jPvvv5/SpUuzfv16n08juXr1amrWrEnv3r0ZOXKkx+c8++yznDhxgl27dvm0Fn+Sqoy8dAnuuQeCgmDKFHttsozu6FE7d1M5vZYy0n3++usvGjZsyBdffMGIESPo27dvqq8W79SpE9OmTePAgQNeH+m9ceNGKlWqRJ8+fVI0VeWtmj9/PuHh4eTKlYszZ87wyy+/aKRPIul2DnnpEvz8s33OtmkTVK1qXzWfPbu7m0DR0fa55nPPpeplysfMafPmzbRo0YJff/2V7t2707RpU0JDQylUqBBZs+r6XH90Q0Zaln2B4fffa+Qi2OeQlSuneqaOjJyRP/74Iw8++CDt27dn3LhxPt1XZrFv3z7uuecehgwZwoABA/jqq6+oWLEiY8eOpWvXrk6Xl2ml6RzyvvvsUZKjRvmkJle5ehXOnYMGDVL90qQyUmcSkmY7d+6kevXqBAcHs27dOjV8JP0VLw6LF8P58/aUIG3a2GvydO1qT7mU0JEjdqNn1Sr7ytGgIPtqq/vug2LF4F//yhxv1oqjunbtyvnz55k8eXK6rBtWo0YN2rVrR2RkJI0aNbphmreTJ0+yceNG+vTp4/NaMqwcOewpiNq1g6goCAtzuqKbWrNmDdOmTePYsWMUKlSINm3aEOYHdcvfLMvim2++YdGiRezevZvg4GDy5ctH3rx542+ePg8ODo6/EvSXX36hdu3a/Pbbb8ydO5cXX3wxTbV0796diRMnMm7cOIYMGeK1r/Hy5ct07NiRYsWKMXDgQK9tNznNmjUje/bsNGvWjEceeUQNHyflyGFf4XnkCEyaBBMm/P1Yliz22j8VK8Lu3TB2rN0MirsFBkLbtlCqlD1975df2hf6hIbab7pqDTZx2OXLl3njjTd46623uOuuu1i3bh2VK1d2uizxBWPsGSSOHrXfzMud2+mKxGX69u1LUFAQgwYNcrqUDOPuu++matWq/N///R/9+/fnzTffpECBAj4bMS4+NHcubN1qv8/n5ot+XEpNH0mT7777jmrVqhEUFMTatWspXry40yVJZpYrFzz+OHTsCAMHQqdOUL263QS64w77OePHw1df2W8gRERA69ZQtCjoKjpJJ8uXL2f+/PkMHTqUf/zjH+m239GjR7Ny5Upat27N9u3b4xcpB/jss8+4du2apna7Va1bw9tv29MSVa3q6hPSNWvWEBkZyaXoaACOHjtGZGQkgBo/LhcTE8OmTZtYtGgRixcv5sCBAwQEBHD//fdz/vx5Tp06xalTp7jZKP7g4GDy5s3LyZMnyZ49O59//jkVK1ZMc10lS5akXr16jB8/Pv6NC29455132L17N8uWLSNXrlxe2WZKNGzYkK1bt6brPiUZYWHwxRewfLl9kc/Fi/ZomhIl7HXVcuSwP46OhsuX7dupU3DihH1l6P/+Z4/CTOj22+3Mvvde2LcPDh36uyEUGOjIlymZx+7du2nRogXbt2/npZdeYuzYsQQHBztdlviSZdmNa8uyZ6kQibVx40Y++eQThg0bRsGCBZ0uJ0Np06YN4eHhjB07lhUrVjB06FCd2/mjRx6BkyftEeD6/qWa3u2UVNuzZw/VqlUjW7ZsrF27lhIlSjhdkoh9FVWXLvaInX797D/wo6Lsqz8LFoQWLWDYMHj0UXuUj0g6Onv2LB07dqR06dLpPqomODiYqVOnUrNmTQYNGnTdNG/Lli2jYMGCPPbYY+laU4aTNSuMGGFnzMGD9ihEl5o2bVp8wyfOpehopk2bpqaPC129epUNGzawaNEiPv74Y44ePUr27NkJCwtj8ODB1KtXj/z588c/PyYmhnPnznHy5Mn4JlDcLfF9MTExDBgwgJJeWBi1Z8+eLFmyhJkzZ9KxY8db3t6+ffsYMmQIDRs2pG7dure8vdQqW7Zsuu9TkvH00/bNk+rV7fO/hCzLnmItJgYaNrRHBf3wg33bs8eeNi5rVvvK+xUrYNGiv19boIDdABo2TG8uiFfFxMQwZswY+vfvT3BwMJ988gn169d3uixJD8bYmbNmjX2RokYbCvao7V69elGkSBG6d+/udDkZTsOGDcmbNy+9evUiT548dO7c2emSJC1OnLDP1YoXt5dikFRR08cFVq9ezfbt2+nTp0+q51FPb3v37qVq1aoYY1i3bp1X3igQ8aq8ee1RPa+8Yv9brBg884x9VafLf7/EHb7//ntWr15Nhw4dyJHG9Z0S69evH4cPH+ajjz4ie/bsXtlmania5u3KlSt89tlnNGrUSAt/ekOTJvYon5UrYdkye0qi0qXhwQftq81d4tixY6m6X9JfdHQ0n3/+OYsWLWLJkiX8+eefBAUF8dxzz9G4cWPq1KmT5FXhWbJkITg4mODgYIoVK5ZuNVesWJFHH32UMWPG0L59+1uavtKyLLp06UJAQADvvfeeF6uUTMMYe8Rl3P9td91l3xI3ti9ftqeIa9v276bQL7/A77/D6dNw5ox9ZWnx4jqHlFvy22+/0bp1azZs2ED9+vWZPHmyrurPbNq1s6clX7vWnmJcMr0FCxbw9ddfM336dK+Nkpa/5cyZk/DwcMaNG0enTp3Ily+f0yVJWly+bF/I3aqVmj5poKaPw3777TeaNGnC2bNnyZs3L+3bt3e6JI+uXbvG+PHj6d+/Pzly5GD9+vWUKlXK6bIkjXr27Ol0Cb53//3w/vtOVyF+Zs+ePVSpUoU//viDqVOn8uGHH1K+fPlb2ubmzZsZN24cXbp0uWFNnfQ0evRoVq1aFT/N2+bNmzl9+jT16tVzrCa3SlNGGgP589vzts+ZY68fNneu/VhoqD0FZeKr0R1QqFAhjnpo8BQqVMiBaiTOxYsXWblyJYsWLWLZsmWcOXOG4OBg6tatS+PGjalVq5ar3xAwxtCzZ09efPFFli1bdktXry9evJhPP/2Ud955h9DQUC9WKd6Qoc4hs2e38zk0FGrVuv6x6GjYsAGefx7Kl4fXXgO9YSSpZFkWM2bMiF84fPr06bRq1cr1F3pK2iWZkWFh9tpia9ao6SNER0fz73//m7Jly9KiRQuny8mwunXrxoEDB+ihaRVdIU3nkHfeaf+NfeCA9wvyAbetnaumj4NiYmJo3bo1AJUqVaJr16489thjt/wGo7d9++23REREsGXLFmrWrMmECRO4++67nS5LboHW7xC50S+//EL16tXJkiULU6dOZcCAATz++OMMHjyYPn36kDUN6z9dvnyZtm3bEhoayvDhw31QdcrFTfNWo0YNBg4cyNWrVwkMDNSUXh7cUkY++KA9BP3PP+11KNavt9eUOHzYnkrIGHjzTXuayTJl7NFApUrZbz6mgzZt2ly3pg9AjsBA2rRpky77lxsdOXKEypUr89NPP3H77bfTpEkTGjVqRPXq1Qn0o/VFGjduTLFixYiMjExz0+fMmTO8+uqrlCtXjldeecXLFYo3ZJpzyMBAe+q4gQPhjTfstdt697ZHBkmGdOHCBTZt2sT69evZsGEDFy5coEyZMpQpU4ayZctSpkwZChUqlOKGzfHjx4mIiGDJkiVUrlyZGTNmaB3cTCDJjAwIgJdeguHD/15HTDKt8ePHs2/fPlatWqUZF3zo3nvvZenSpU6XIbHSdA5pjH1R5a+/2tP3uviiCTeunaumj4PGjBnDhg0bmD59OnXq1KF8+fI0bdqUb775xhWLOZ4/f57BgwczZswY8ufPz9y5c2nWrJmuTMoAfvzxRwCN1hKJdfDgQapVq0Z0dDTr16+ndOnSNGjQgE6dOtG/f3+WL1/OrFmzuPfee1O13bfeeovvv/+e5cuXkydPHh9Vn3JhYWFEREQQGRlJvnz5qFq1qha09MArGZk/v72WRMOG9udXrthTBf31l/2H/rZtsHmz/VjWrPbzOnW6xcpvLu6E001XIGVmJ06coHr16hw5coQlS5ZQu3btNDWY3SBr1qx07dqVHj16sGXLFh599NFUb2PgwIEcOXKEjz/+2G+PQ0aXqc4hs2SB/v2hdm148UUYMACeew569bIfE792/vx5Nm/ezPr161m/fj1ff/01V65cISAggAoVKlCgQAFWr17NzJkz418TEhJyXROobNmyPPDAAzeMxFyyZAnt2rXjzJkzREZG0q1bt1ua9lL8R7IZ2aaNfR544UI6VyVucvLkSYYOHUqNGjWoUaOG0+WIpJs0n0OWK2dfTHnxoqvX53bj2rn6a8ohu3btol+/fjRo0CB+iPe8efOoUqUKERERzJ0719HmyqeffkqnTp3Yv38/7dq14+2339YcmBlI3DSC69evd7YQERc4duwY1atX5+TJk6xdu5bSpUsDkD9/fubNm0f9+vXp3Lkz5cqVIzIykoiIiBTl8/fff8+wYcN48cUXqeOiaRxGjRrFypUrOXDggKZ2S4JPMjJbNrsRlD8/LF9uLzC+b589EmjDBnvNiCNH4NQpmDrVngbkmWd8spB4WFhYmk4844arc+wYe0ND6fnWW4SHh3u9vszizJkz1KpVi19//ZXPPvuMKlWqOF3SLWvTpg2DBw8mMjKSefPmpeq127Zt4/3336djx4489thjPqpQblWmPIcsXx527oSePeG33+w3bfPnd/XVpnKjmzV5evToQZUqVXj66aevu1DnxIkT7Nq1i507d8b/O3nyZC7EvnFvjOHee++NbwTt27ePmTNnUq5cuevOKyVzSDYjixeHkSNh6VLXX7EuvjNixAhOnTrFqFGjnC5FJF2l+RyyXDn7b+f9++1lHFzKjWvnqunjgOjoaFq0aEHevHmZPHly/JuHFStWZNiwYfz73/+mSpUqdOjQId1rO3LkCF27duWjjz7igQceYOPGjVTUNAYikkH9+eefVK9enUOHDrF69WoeeeSR6x43xtC8eXMqVarESy+9RIcOHVi6dCnTpk3jjjvuSHK7MTExtGvXjjx58vDuu+/6+stIleDgYGbOnEnXrl1pGDcKRdJflixQooR9i5ta7cwZWLbMXkT87bfhnXfgiSegRg147LF0mwLOk4TD1QsBBw8dIiIiAkCNnzS4cOECdevWZceOHXzyyScZouEDdr5EREQwZswY9u/fT7FixVL0umvXrtG+fXsKFizo+FSYkjHMnj2b/v37c+DAAYoWLcrw4cNvLasCA+GDD+wr9Ldvh6go2LvXnq5Jo9J8Kioqio4dO3Lt2jVy5cpFrly5CAoKiv84qfuCgoIIDAxk586dqWryJFagQAGeffZZnn322fj7rl27xq+//npdM2jHjh0sXrwYYwz9+vVj0KBBZHfw/21xqcBAe5qio0fthrJkKvv27eO9996jdevWlC1b1ulyRPxDkybw8MP2BTgu5sa1c3WG6oBBgwaxY8cOli1bRkhIyHWP9enThw0bNtCtWzcef/zxdFvfJyYmhsmTJ9O3b18uXbrE0KFD6dOnj05URSTDOn36NDVr1uSnn35ixYoVPP3000k+NzQ0lFWrVjFu3Dj69OlD6dKlmTRpEo0bN/b4/EmTJrFp0yamT59OwYIFffUlpFmVKlXYsWOH02VIYsHBEB4O//ynvdDvzJn2+kBffgmTJ8Pdd8PZs/bz0nmaGE/D1S9cuED//v3V9Eml6OhoGjVqxFdffcXcuXNdNRLQG1555RXGjBnDu+++yzvvvJOi10yYMIGtW7cyd+5c8ubN69sCJcObPXs2ERER8SMx9u/f770mdVAQPPUUzJ0Lc+bYa7YNHAhFi95q2eLBli1baNCgAaGhoTzxxBOcP3+eCxcucP78eQ4fPsz58+fjbxcuXODy5cs3bCO1TZ6UCAgIoGTJkpQsWZJGjRrF3x9XR+K/8UXiXbkCo0bZa0Cq6ZPp9O/fn4CAAIYOHep0KSL+I0cOCAmxR0i6mBvXztXEsols2bKF+vXrszlunn0v+/LLLxk5ciRt27albt26NzyeJUsWZs2aRYECBWjatClnzpzxSR0JfffddzzzzDN07NiRRx55hJ07d/L666+r4SMiGdb58+epU6cOO3bsYNGiRVSrVu2mr8mSJQuvvPIK27dvp3jx4jRp0oRWrVpx+vTp65536NAhXnvtNapVq0arVq189SVIRpYlC9Ssab+hePy43fh56CH4808YNAiaNoUJE+Dnn9Pt5DepYekHDhxIl/1nFFevXqV58+asWrWKKVOm0LRpU6dL8rqiRYvStGlTpk6dekM+evL777/Tr18/wsLCaNasWTpUKBld//794xs+ceKa1F5hjD3qZ9YsO6PbtYPFi13/ZoS/2bt3L7Vr1yYkJIR169bx4YcfsnjxYlauXMnGjRvZtm0bP/74I4cOHeLkyZNER0dz+fJlTp06xeHDh9m7dy/ffvstJ0+e5L///S9vvfUWtWrV8ukai7ly5VLDR5KXM6d9gc+WLfZUkZJpxF3c0qNHD4oUKeJ0OSL+5dtv4d//tv8edqmwsDB69uzJHYUKYYA7ChWiZ8+ejq6dq6ZPrJiYGEaPHs1TTz3F0qVLCQsLY926dV7dx9mzZ2nZsiV33313slc+hoSEMG/ePPbt20dERASWj/6AuHjxIv369aN8+fL8+OOPzJw5k6ioKO677z6f7E9ExA0uXbpEgwYN2Lx5M3PmzEn1Vfb/+Mc/2Lx5MwMHDmT27NmULVs2/v8Ly7Lo3LkzV69eZdKkSY6uzSYZRGCgvXD4o49Cw4bQuTP84x+waJH9RmOrVvbc8ADXrsGYMTBuHEyZYo8UmjPn76Hwly/D2rWwcaN9dfrRoykuI6lh6UV1dXuKxcTE0KZNGxYvXszYsWN5+eWXnS7JZ3r27MnZs2eZMmXKTZ/bvXt3Ll++zPjx45WZ4hVJNaO93qRu0QK+/96efvP99+0GvXjFkSNHqFmzJsYYVq9ezZ133pmi12XLlo3bbruNwoULU7JkSR566CGfNnlE0qRtW3vEjzIj07Asi169ehESEkKfPn2cLkfE/xQubK/ps3Ch05UkKywsjHnz5rFu3TrmzZvnaMMHNL0bAMePH6dly5asWrWKRo0aMWzYMJo0aULt2rX5+OOPqVWrllf20717d/bv388XX3xx05NPX6/vExUVRYcOHfjll19o3bo1o0aNokCBAl7dh7jX66+/7nQJIo64cuUKTZs2JSoqihkzZvDCCy+kaTvZsmXjjTfeoE6dOrRo0YKqVavSvXt3Hn74YZYuXcrIkSMpUaKEl6uX9OLajMyWDSIi7Nvvv8P06Xbz59gxOHIEoqNh/Xr7jYQrV+DqVft1DRtCwYJw6hQknE4iIMAeUdS6tT1kPhmehqsHBQVp/ZUUsiyLV199lVmzZjF06FC6du3qdEk+9cgjj1C5cmXeffddunbtSrZs2Tw+b+XKlSxYsIAhQ4Zw7733pnOVkhauzccEihYtyv79+z3e73WhofDFF/B//2fn6PHjoL+pbsnp06epVasWf/zxB+vXr6dkyZJOlySSYinKyPLloWxZ+0KcFi18X5Q4bvny5WzYsIFx48YRHBzsdDkijrilc8gHH4Rnn4XVq+31FDUzVYoYX40i8bYKFSpYW7du9fp2o6KiaNGiBSdPnmTs2LG0b98eYwwnTpwgLCyM3bt3M3/+/Fte7Hrp0qXUr1+ff//734wYMSJFr4mJiaFOnTqsW7eOzZs3e2V9nzNnztCrVy+mTJlCyZIlmTRp0nWLUopI2hljvrEsq0J679dX+ZjRXLt2jebNm7NgwQLGjx9Px44dvbLdCxcu0KdPH8aNGwdA+fLl+frrr8mqhZ0lvVy9CjExN96uXYOLF+0ph7Jnt0f6/PILXLpkL0a+cCHMnw/9+sEzz9x0MfI1a9Ywbdo0OHaMvaGh9HzrrVStj5GZM7Jfv368+eab9O7dm7fffjtTjGhZtmwZ9erVY/bs2TRv3vyGxy9evEjp0qXJli0bO3bsIDAw0IEqJSNKvKYP2E3qyZMn+3YNsitX4Lvv4MwZqFw5VS/NzPmY0KVLl6hVqxabNm1ixYoVjl8hK+Iz48bBsGEwejRkpqm+jh618zGFo/fi+HNGXr16lTJlyhATE8N3332X5IUwInITy5fD889Djx72vxnN1atw7hw0aJDqlyaVkY5M72ZsE4wx+40x/zXGhKZ3DVeuXKFfv37UqFGDfPnysWXLFjp06BD/R3iBAgVYu3YtDz/8MC+88AJz585N876OHz9O27ZtKVeuHIMHD07x6xKu7/PCCy/c8vo+a9eupUyZMkybNo3XXnuNnTt3quGTSX377bd8++23Tpchkm5iYmJo27YtCxYsYNSoUV5r+ID9RtIHH3zAypUrqVq1KjNmzFDDx8/5XUZmzWo3dXLksBcZz50bgoMhXz57KHyRIvYV6EWKQKVKUKOGfTL5n//YTaBmzex55d991x49lGgtjDhxw9XnzZ3Ltm3bfPvmaQby5ptv8uabb9KhQ4dM0/ABqFOnDqVKlSIyMtLjVMXDhw/n119/ZcKECWr4+BF/yMfw8HAmT55MsWLFMMZQrFgx3zd8wB6NWb48VKzo2/1kUNeuXSM8PJwNGzYwc+ZMNXzEL6U4I9u1g02b7HUcJUObOnUqP/zwAyNHjlTDRzK1Wz6HrF0b7r7776nN5aaceleqHhACFAeaAsOBdFvt+rfffqN58+Zs3ryZtm3bMnbsWHLlynXD8/Lly8eaNWuoW7cu4eHhXLx4MdXzr1uWRUREBKdPn2bt2rVkT+UQtLj1fapUqUK7du2YN29eqt8sOHfuHH379mXcuHHcd999fPnllzz55JOp2oZkLN26dQNg/fr1jtYhkh7iplWaMWMGgwcPplevXj7ZT82aNalZs6ZPti3pK1NlZNx0R8WL2w2fTz6Bjz+2Fxlu0MBeU0jSbNy4cfTr14/w8HDGjRuXaRo+YF+81L17dzp06MCGDRuoUqVK/GN79uxh5MiRtGzZUhcg+Rl/ycfw8HDnGtMBAc7s14/FrYkYt+bZP//5T6dLEkmTFGdk9uz2xTgxMfb0vDrfypDOnj3LoEGDeOaZZ6hXr57T5Yg46pbPIbNkgVGjYNs2e0RM7txeqy2jcuqygtrATMu+7O9joEp67XjhwoWUK1eO3bt3M2/ePKZMmeKx4RMnT548fPbZZ4SFhdGmTZv46XtSasaMGSxZsoQ333yT0qVLp6nmuPV9FixYwMSJE1P12o0bN/LQQw8xfvx4unfvzvbt29XwEZFMw7Ks+KZ37969GThwoNMlibhTUBB89hls2AAPPAATJ0Lz5vDNN05X5rdmzpxJly5dqF+/PtOnTydLJryat2XLlhQoUIDIyMj4+yzLomPHjuTOnZtRo0Y5WJ2IuMWQIUOYNGkSffv2zfBrnonEO3gQOneGlSudrkR8ZNSoURw/fpzRo0dnqgt/RHymcWPo0AHOnnW6Er/g1F+fRYFDAJZlXQYCjDE31GKMiTDGbDXGbP3jjz9uaYcXL16kQ4cOvPDCC5QqVYpvv/2WZs2apei1QUFBLF26lHr16tGlSxdGjx6dotft27ePV199lSpVqsR3NNOqT58+PPfcc3Tr1o3t27ff9PkXL16ke/fuVI6dT3r9+vW88847BAUF3VIdIuIO3szHjGzYsGGMHDmSTp06ZapplUTSrFIl+O9/YdkyexSQMXDypL1GRUyM09WlmNMZuWjRIl5++WWqV6/OvHnzMu10Hjlz5qRTp04sX76cH374AYBZs2axYcMGRo4cScGCBR2uUCTzcTofE5s4cSKDBw/mpZdeSvHatyIZwt13Q968sGaN05VIAt7KyMOHDzN69GiaNWvGY4895sUKRTK5wEBYsgR+/93pSlzPqaaPBVxN8PlVy7JueCfBsqzJlmVVsCyrQkhISJp3tnv3bh577DEmTZpEnz59+PLLL7n77rtTtY3AwEAWLlxI06ZN6d27N0OGDPE4P3mca9eu0apVK7JkycKMGTNu+erOuPV9QkJCbrq+z3//+1/KlSvH2LFj6dixIzt27KBSpUq3tH8RcRdv5WNqxMTEsGrVKqZNm8avv/6aLvu8GcuyOHXqFD/99BObNm3ik08+YcqUKQwfPpxWrVoxcOBAWrVqxfvvv6+Gj0hq1K0L27fDSy9B/vzwxhvQpg1s3gzJnP+4hRMZGWflypX885//5IknnuCTTz4hR44c6bp/t+ncuTOBgYGMGTOGP//8k169evH000+nespkEfEOJ/MxsUWLFtGpUyfq1q3L5MmTda4mmUuWLPDyy7B7N/z2m9PVSCxvZeTAgQO5evWqmtki3nbuHCxYAAsXOl2J6zm1ps9hoDCwyxiTDbjki51YlsXUqVPp2rUrefLkYdWqVdSoUSPN28uWLRtz5swhZ86cDBo0iAsXLvDmm296PDmNjIxk48aNzJw5k2LFit3KlxGvQIECya7vc+nSJQYNGsTo0aMJDQ0lKiqKatWqeWXfIpJ5Xbx4kQ8//JAxY8bEX6kNcPfddxMWFkb16tWpWrUq+fPn9+p+L1++zO7du9m5cyeHDx/mjz/+4Pjx4/zxxx/XfXzlyhWPr8+TJw9t2rRh4sSJmXJaJRGvyJfPHv0TEQFDhkC/ftCsGVSv7nRlrvTFF1/QqFEjHnzwQVasWJHsFMKZRcGCBWnRogWzZs3ixIkTnDp1igkTJiiXRTK59evX07x5c5544gnmz59P1qxOvTUh4qCXXoLBg+0R1q+84nQ14iW7du1i+vTpdO/enXvuucfpckQylnvugTp1ICoK2rWDnDmdrsi1nDqzWgE0B1bF/uv18aynTp0iIiKCjz76iLCwMGbNmsUdd9xxy9sNCAjg//7v/8iZMydvv/02Fy5cYOzYsdf94bpz504GDBhAo0aNaNGixS3vM6GKFSsyfPhw+vbtS5UqVejYsSMAW7dupVWrVnz//fe0bduWyMhIgoODvbpvyTh0tYmkxNGjRxk/fjwTJkzgxIkTlC9fng8//JDy5cuzbt06oqKimDdvXvyVmQ8//DDVq1enevXqPP300+RMxX++Z8+eZceOHWzfvj3+tnv37usaOrlz5yYkJISCBQsSGhrKww8/TEhISPx9Cf8NCQnJ9FfXS9opIz1o395+Y+Ldd+3pSDLpdGVJuXbtGpMmTeK1116jWLFirF69mrx58zpdlmv06NGDqVOnsnjxYvr06UOZMmWcLknSSPko3rBjxw7q169PiRIlWL58uaYglwwj1Rl5111QrRqsXw8dO4KanxlCnz59uO222+jfv7/TpYi4hlfPIXv0gOXL7YZ506be224GY5KbosxnO7WHp0wBqgIHgcaWZZ1I7jUVKlSwtm7dmuJ99O3bl8jISIYPH06vXr28fjWhZVn07NmTMWPG0LZtWyZOnEhAQADR0dE8+uijHD9+nO+++44CBQp4db9gT7FUt25dPv/8c7744guWL1/Om2++yR133MHUqVOpVauW1/cpIjdnjPnGsqwK6b3f1ObjzezatYsxY8Ywe/Zsrly5wvPPP0+PHj2oVKnSDSMbr169ypYtW4iKimLNmjVs3ryZq1evkiNHDipWrBjfBCpfvnx8Dh8/fvy65s727dv5+eef46fMDAkJoXz58vG3cuXKUbRo0VQ1kUTEfTJKRia2bds2OnTowJYtW6hWrRozZ86kSJEiPtufv2rYsCE7d+5k586dGgElkkhGzUdP9u3bx1NPPUXWrFnZtGkTd911V7ruX8R1vv4atmyBIkXsNX4ysqNHoXJluPPOVL3MnzIyKiqKsLAwRo8eTc+ePX1UmUgmZ1lQpgycPQszZtjr0Pq7q1ftqesaNEj1S5PKSEcuI7Dsd/ba+nIfAwYMoEmTJlSo4Jv/F4wxREZGkitXLoYNG8aFCxeYOXMmAwYMYNeuXSxfvtwnDR/4e32fcuXK8eSTT2JZFi1btuTdd9/VVaWSIps2bQLgqaeecrgScYu49XrGjBnDmjVryJkzJ23btqVr167cd999Sb4ua9asPPnkkzz55JMMGDCAc+fO8cUXX8Q3gfr27QtA/vz5KVu2LD/++CO/J1hwr3jx4pQvX54WLVrEN3kKFy6sOd3FUcpISYkzZ84wcOBA3n//fQoUKMDs2bP55z//qfxKwty5c7ly5YoaPn5O+Si34o8//qBmzZpER0cTFRWlho9kOGnKyMceg6JF7amK9H6OX7t27Rq9evWiePHidOnSxelyRFzFq+eQxkCvXjB5Mhw/DoUK3fo2M6AMO3Y0V65cPmv4xDHGMHToUHLmzEn//v05ePAgX375Je3bt6dOnTo+3XeBAgX46KOP6NmzJ3379qVevXo+3Z9kLP369QPsubQlc7t48SL/+c9/GDNmDHv27KFw4cKMGDGC9u3bc/vtt6d6e7lz56Z27drUrl0bsKeI+/zzz4mKiuK7776jatWq143gyZcvn7e/JJFbpoyU5FiWxaJFi+jatStHjhyhQ4cODB8+XHl2Ezly5NC0mxmA8lHS6uzZs9SuXZtDhw4RFRXFgw8+6HRJIl6X5ozMnx8iI+3mT6dOEBjo/eLE52bPns2OHTuYM2cOgfoeilzH6+eQrVtDjRr29JjiUYZt+qSnfv36ERQURPfu3SlRogSjR49Ol/0++eST8Z1SEZHUOHbsGOPHj2f8+PHXrdfTtGlTsmfP7rX93HHHHYSHhxMeHu61bYqIOGXfvn106dKFTz/9lHLlyrF48WIef/xxp8sSEXGlY8eO8emnn7J8+XJWr17NxYsX+fjjjzVSTMSTChVg6lTYvh0GDoR773W6IkmlunXrMmrUKJo1a+Z0KSKZQ6FCcPq0/XHhws7W4kJq+nhJt27dKFOmDPfccw+5c+d2uhwRkSSdO3eOkiVLcvbs2fj1eipXrqwpiUREknD58mUiIyMZMmQIWbNmZcyYMXTp0oWsWnBZRCReTEwM27dvZ8WKFSxfvpwtW7YAUKRIEcLDw/nXv/5FxYoVHa5SxIWyZYMpU6BuXXj5ZejY0f63WTPw8vrU4ju33347vXr1croMkczjzBno1g2qV4fevZ2uxnX0l6oXVatWzekSRERuKnfu3HzwwQc88cQTya7XIyIi8MUXX9ChQwf27NlDo0aNePfddwkNDXW6LBERVzh37hyff/45y5cvZ8WKFRw5cgRjDI8//jjDhg2jTp06PPTQQ7q4SCQl6teHPXsgPBwWLoRKlaBIEaerEhFxp3z5oEEDWLYMOnSAPHmcrshV1PQREcmEWrZs6XQJIiKuduLECfr06cP06dMpVqwYy5Yto27duk6XJSLiuH379sWP5lm3bh2XL18mODiYmjVrUrduXZ577jlCQkKcLlPEPxUsCKtXww8/wG+/wcGDcOgQPPmk05WJiLhPjx6wYAF8/DHofa7rqOkjkgmNHTvW6RJERFxLGZm5xcTEMGPGDHr37s2ZM2fo27cvr7/+Orly5XK6NBHHKR8zt5iYGB599FG2bdsGwH333UeXLl2oW7cuFStWJFu2bA5XKOIsr2WkMXD//VCyJAwYAG+9BdWqQffuoPMREfFDPjuHfPxxe020zz6zR0kGBPhmPylx9Sp8/jk89pg9CslhavqIZELlypVzugQREddSRmZuo0aNom/fvlSsWJGJEyfy4IMPOl2SiGsoHzO3LFmyUKtWLVq0aEGdOnUoWbKk0yWJuIrXMzJrVhgyBC5fhrFjYdcueP11KFPGu/sREfExn55D9uxpj/LZtQucOlc9fRoGD4Zvv4WQEJgxA4KCnKkllpo+IplQVFQUANWrV3e4EhER91FGZm7t2rWjUKFCtGzZkixaPFnkOspHGT58uNMliLiWTzIyWzaIjIS6deFf/4KuXaFLF2jUyHv7EBHxMZ+eQzZuDKVLw86d3t92Svz8sz0q888/7eZT/vxw8iQEBjo68khNH5FMaNiwYYD+YBcR8UQZmbndfvvttG7d2ukyRFxJ+SgikjSfZuSzz8L330NEBISG2m8oumD6IBGRlPBpPmbLBg88APv2wdmzkCeP9/eRFMuyG/PR0TBhgt30yZrVzus5c+DUKXjhhfSrJwE1fURERERERERERNzstttg/nz7Tc3//c9e6+eBB+xRQBqdLCKZmTEwciRkz26PuvG1mBh7DZ+YGGjXzl5X6Jln7DoAHnwQ9u6FhQth/357TbZ0HvWj/xVERERERERERET8QZ488NRT9hXkY8bY074tWGA3g0REMiNj4N574csv4a+/fLuvCxdg0CAYOtTO4RdfhEqV/m74xJk3z56Oc8UK6NULzp3zbV2JqOkjIiIiIiIiIiLiL4KCmN25M6/dfjs7jxyBCRO41qgRfPWV05WJiDije3e4fBkWLfLdPg4fhs6dYdMmuP9+qFULChf2/NyAAHj/ffu2axd07AinT/uutkTU9BEREREREREREfETs2fPJqJDB0b+9RcPAU8BS2Ni2LxvHxw5Aps3Q1QUXLnidKkiIumjbFl7xM2qVfbUa972zTd24+aPP2DUKLuZkzfvzV8XN9rn4Yft0T4xMd6vzQOt6SOSCU2aNMnpEkREXEsZKSLimfJRRCRp6ZmR/fv358KFC/GfbwYaxcRQYs0afh4wACZPhvXrYdw4e82fevUgJCTd6hMRSSjd8rFHD2jQwG781Knjve1evAjDhtlrq02YAM89l7o1emrWhOrVYedOO5tPnICwMO/V54GaPiKZUKlSpZwuQUTEtZSRIiKeKR9FRJKWnhl54MABj/f/evAgFCtmj/JZtAjeew9mz4Y5c6BhQ/uKcxGRdJZu+Vi3rr3WWZ483tnelSuQNStcuwa9e9sN9NKl07atgAAoVw7efhvmz4d9+6BduxvXAvISTe8mkgktW7aMZcuWOV2GiIgrKSNFRDxTPoqIJC09M7Jo0aLJ3x8QAE2b2oua794NL70Ed94Jv/8Ox4/DJ5/AmTO+L/TcOdizx/7Ysny/PxFxpXTLx4AA6NrVXmfn/Plb29aff0K3bjBtmt38ad8+7Q2fOMbArFnQrBnMnQsDB0J09K1tMwka6SOSCUVGRgLw/PPPO1yJiIj7KCNFRDxTPoqIJC09M3L48OFERERcN8VbUFAQw4cPv/HJ998PU6faV6ofPw4TJ8K779qjgEqVgieegEcftT9OzXRFSTl3zm42ffEFbN0KQUH2NHO5ctk3Ecl00vUc0hj48Uc7g159NW3b+OEHeP11O8+aNLGnZwsK8k592bPbDZ8HHoDBg+GVV2DECHtEkRdppI+IiIiIiIiIiIifCA8PZ/LkyRQrVgxjDMWKFWPy5MmEh4cn/aKAAHu0zxtv2GtKdOhgj76ZORM6d4affrIbQ4cPw19/pa2wRYvsaeTeftve3nPPwQcf2G+Y1qsHwcFp266ISGr88gssWwZHjtz4mGXZo4AOHIDt2+3pMH//3X5s7157dM+rr0KWLPYonwEDvNfwiWOMPcpnzhwIDITTp727fTTSR0RERERERERExK+Eh4cn3+RJTuXK9g3g4EF70fMHH4RDh2D8eNi0Ce65Bx5/HB57zH4sW7brt3H2rD2iZ8MGaNECChaE22+3F09v0sRu+OTLZ79xKiKSnrp2tUcYRkbC3XfbIxofeQR+/dVucl+6dP3zIyKgalW7+RIdDZUqwfDhdv75aM0dAF58ERo3hm++SXuzPQlq+oiIiIiIiIiIiGRGd90FbdvaH1+7Zq+FsXgxrFkDCxbY0xCVKGFPCxcQAGvXwurV9puU165BSAgcPWq/SVqtGuTNq0aPiDjr3nuhQQP4+GP47jt7pE6RInYDp1Yte9Rj4cJ2/hUtCsWL29kVGAjdu3t9qrVkZctmN9jPnvXqZtX0ERERERERERERyewCAuDpp+0bwLFjsGKF/e/p03Dxon3lfO7c8Pzz9mLkYWH2CB9fXg0vIpJaH31kT++WO7e9jk5goJ1xL73kdGU3Msbr01860vQxxmQF9gHnY+9aYVlWTydqEcmMPvzwQ6dLEBFxLWWkiIhnykcRkaRlyIwsVAheftn+OCbGbv4sXGhfla5Gj4ikkCP5GBAAoaHpv1+XcGqkTxFgnWVZLR3av0imdtdddzldgoiIaykjRUQ8Uz6KiCQtw2dkliz2lEh33ul0JSLiZzJ8PrqQU5NsFgUOOrRvkUxv/vz5zJ8/3+kyRERcSRkpIuKZ8lFEJGnKSBERz5SP6c+pkT7FgFrGmPrAH0Any7L2JH6SMSYCiAAoWrRo+lYokoFNmDABgGbNmjlciaSV8lHEd5SR/k8ZKeIbykf/p3wU8R1lpP9TRor4hvIx/Tk10uco8B5QFngbmOPpSZZlTbYsq4JlWRVCQkLSsz4REVdTPoqIJE0ZKSLimfJRRCRpykgRySh83vQxxgwwxnyX8AaUAj60LCvGsqyVQEFjtPqbiIiIiIiIiIiIiIhIWvm86WNZ1lDLskonvAFlgM4AxpingH2WZVm+rkVERERERERERERERCSjcmpNnyHAXGPMK9hr+rR1qA4REREREREREREREZEMwZGmj2VZvwOVndi3iMDChQudLkFExLWUkSIinikfRUSSpowUEfFM+Zj+nBrpIyIOKlCggNMliIi4ljJSRMQz5aOISNKUkSIinikf05/P1/QREfeZMWMGM2bMcLoMERFXUkaKiHimfBQRSZoyUkTEM+Vj0mbPnk3x4sXJkiULxYsXZ/bs2V7Zrkb6iGRCcUHbunVrR+sQEXEjZaSIiGfKRxGRpCkjRUQ8Uz56Nnv2bCIiIrhw4QIA+/fvJyIiAoDw8PBb2rZG+oiIiIiIiIiIiIiIiKST/v37xzd84ly4cIH+/fvf8rbV9BEREREREREREREREUknBw4cSNX9qaGmj4iIiIiIiIiIiIiISDopWrRoqu5PDTV9RERERERERERERERE0snw4cMJCgq67r6goCCGDx9+y9vOestbEBG/8+mnnzpdgoiIaykjRUQ8Uz6KiCRNGSki4pny0bPw8HDAXtvnwIEDFC1alOHDh8fffyvU9BHJhBJ3kUVE5G/KSBERz5SPIiJJU0aKiHimfExaeHi4V5o8iWl6N5FMaPz48YwfP97pMkREXEkZKSLimfJRRCRpykgREc+Uj+lPTR+RTGjBggUsWLDA6TJERFxJGSki4pnyUUQkacpIERHPlI/pT00fERERERERERERERGRDEBNHxERERERERERERERkQxATR8REREREREREREREZEMQE0fERERERERERERERGRDMBYluV0DSlijPkD2O90HbEKACecLiKWm2oB1XMzbqrHTbVAxqinmGVZIb4oJjnKx2SpnqS5qRZQPTfjpnrSWosy0l3fR3BXPW6qBVRPctxUC2SMepSPGeP76EtuqsdNtYDquRk31aNzyLRz0/cR3FWPm2oB1ZMcN9UCGacejxnpN00fNzHGbLUsq4LTdYC7agHVczNuqsdNtYDqySjcdtxUT9LcVAuonptxUz1uqsXfuO3YuakeN9UCqic5bqoFVE9G4bbjpnqS5qZaQPXcjJvqcVMt/sZtx85N9bipFlA9yXFTLZDx69H0biIiIiIiIiIiIiIiIhmAmj4iIiIiIiIiIiIiIiIZgJo+aTPZ6QIScFMtoHpuxk31uKkWUD0ZhduOm+pJmptqAdVzM26qx021+Bu3HTs31eOmWkD1JMdNtYDqySjcdtxUT9LcVAuonptxUz1uqsXfuO3YuakeN9UCqic5bqoFMng9WtNHREREREREREREREQkA9BIHxERERERERERERERkQxATR8REREREREREREREZEMQE2fJBhjshhjphtjDhhjdhhjnk70+GvGmJ+MMT/E3u5Oh5q2JtjfvESP1THG/GKM+dUY0zwdahmaoJYfjDFHjDGPJ3g83Y6PMaa+Meat2I/vMcZsi/2+vePhucYYM8EYs98Y819jTKiP67nfGPNt7LH42BiT28Pzk/y++qCeisaY3xLsr1mi5/r0+CSq5bsEdfxkjNni4flePzaefrfd8HPjb9yWkcrHZOtxTUYqH1NVjzLST7ktH2P3qYz0XItr8tFDPcrIpGtRPvoxt2Wk8jHZelyTkcrHVNWjjPRTbsvH2H0qIz3X4pp89FCPMjLpWjJXPlqWpZuHG9AYWAIY4AFgZ6LHpwP3pGM9AcDaJB7LDuwFQoG8wM9ArnSs7X5gJbFrRKXX8Yn93owBjgFvxd73MfB87GMrgGcTvaY+sDD28WbATB/XswxoFPvxm0DvlH5ffVRPK+DlZF7jk+PjqZZEjw8BXkqPY+Ppd9vJnxt/vbkpI5WPSe7bNRmpfEx9PYkeV0b60c1N+Xizn5XMmpFuysdk6lFGJlFLoseVj352c1NGKh+T3LdrMlL5mPp6Ej2ujPSjm5vy8WY/K5k1I92Uj8nUo4xMopZEj2f4fNRIn6TdgX1ALcuyvgdCjDEmweN3AYfTsZ7CwJEkHnsE2GVZ1iHLsk4BXwIV06sw4F3gNSv2pzJWeh2fz4E5AMaYAOxjsTy2lo+AGomeX5vY7yv2L1gVX9UTKwf2LzbAV8A/Ej0/ue+rL+opChxM5vm+PD6JawHsqyKA6pZlTU/0kK+OzQ2/20AFnP258UduykjlY9LclJHKx9TVAygj/ZSb8hGUkUlxUz5eV08sZWTStQDKRz/mpoxUPibNTRmpfExdPYAy0k+5KR9BGZkUN+XjdfXEUkYmXQuQefJRTZ8kWJY1zrKsxQDGmHbAT4nCpBDwqTHme2PMoHQoqRjwUOwwsP8lGuZZFDiU4PPfsX+gfM4YUxG4ZFnWjkQP+fz4xP6yLAfi9l0AOJng++TpOMQfK8uyLgMBxhiv/B54qAfLssIsy7pmjMkJdMUO3ISS+756vZ7Y/fU3xuwxxvzHGBOc6GU+OT5J1BKnHzDaw/0+OTYefrdPAX859XPjr1yWkcpHD9yUkcrHNNUTRxnpZ1yWj6CMvIGb8jGJepSRydcSR/noh1yWkcpHD9yUkcrHNNUTRxnpZ1yWj6CMvIGb8jGJepSRydcSJ1PkY6YN05QwxuQyxkzB/iVpnejhlUBL4FGgkjGmgY/LOQ9Mxe4EdgDmGGOyxT5mAVcTPNcCrvm4njjtsTvsiaX38YGUHYfEz7lqWVaML4syxpQHtgI/ATMSPZzc99UXtgGvYw8nPIA9nDGhdD0+xpg8wFPYXevEfHZsEv1ut8KFPzf+wEUZqXxMGddlpPIxecpI/+WifARlZEq4Lh9BGZkc5aN/c1FGKh9TxnUZqXxMnjLSf7koH0EZmRKuy0dQRiYnM+Wjmj5JMMYEARuBc0AFy7J+TfBYADDIsqzDlmWdBxYBpX1c0g/AB5ZlXbEsazv2nISFYh87jD0ELU5h7F8kn4r9RakMrE90vxPHB+BPIH+Czz0dh/hjFfuLe8mXBRljqgFLsefQ7OThFzS576sv/MeyrC9ju8n/4cbvS7oeH+x5KVclunoljk+OTeLfbeAbXPZz4w9clpHKx5RxVUYqH1NEGemHXJaPoIxMCVflY+w+lJHJUz76KZdlpPIxZVyVkcrHFFFG+iGX5SMoI1PCVfkYuw9lZPIyTT6q6ZO0DsCXlmV1tywr8YEtBOwyxgTHBksd4H8+rqcr8DaAMaYEcBv28C9i913eGFPAGFMQu6v9Xx/XA/Ak9jFK3I104vgQW8e3xphnY/f7L+zFsBJaATSP/bg5sMbHZb0DNLAs69MkHk/u++oL/zPGPBT7cX1u/L6k9/GpCnyWxGO+OjbX/W679OfGH7gpI5WPKeDCn3Xl480pI/2Tm/IRlJE35dKfc2Vk8pSP/stNGal8TAEX/qwrH29OGemf3JSPoIy8KZf+nCsjk5dp8jHrrVScwT2CPRww4UJKHwBYlvWBMWYs8C1wEVhsWZavfyg/wB5W9jNwGmgDdDLGxNXTB3uexgCgu2XP9+drT2J3JwEwxnQBx45PnN7APOyO6QzLsr6Jre1z7GGeS4C6xphfsRcSa+yrQmI7sQ8Cs83fa+99bFnWvxPUc8P31UMX3ps6xNYTAHyH/XPkyPGJ9STwSsI70uHYePrdfh6X/Nz4ETdlpPIx5VyRkcrHFFNG+ic35WPcvpWRN+eKfIzdpzLy5pSP/stNGal8TDlXZKTyMcWUkf7JTfkYt29l5M25Ih9j96mMvLlMk4/G82gmERERERERERERERER8Sea3k1ERERERERERERERCQDUNNHREREREREREREREQkA1DTR0REREREREREREREJANQ00dERERERERERERERCQDUNNHREREREREREREREQkA1DTR1zBGPO5MeZq7C0m9hb3+c/GmGKxH9/mUH0zjDFjndi3iIgyUkTEM+WjiEjSlJEiIp4pHyWjy+p0ASIAlmVVi/vYGDMj9r7WiZ6mn1cRyZSUkSIinikfRUSSpowUEfFM+SgZnUb6iF8wxhQ3xljGmLwJPm5vjDlsjDlpjOlsjBlljDlijNlnjGmS4LUVjDFfGWPOGWO2G2OqJrGPXMaYObHbO2qMGW+MyWGMeRtoCbxqjJkZ+9x7jTGrjDGnjTE/GGOaJdiOZYx51RjzU+y2phpjgnx9jEQk81JGioh4pnwUEUmaMlJExDPlo/g7NX3Enz0L3A90BT4AgoB7gMHAZABjTCFgNTAGKAiMBJYaY0I8bK8nEAIUAcoDTwLtLct6DZgFvGdZVitjTA4gClgO3Am8Akw2xjyUYFuNgaeAfwAPAEO89lWLiKSMMlJExDPlo4hI0pSRIiKeKR/Fb6jpI/5siGVZZ4AvYz8faVnWRWADkM8YkwdoAWywLGuhZVkXLMuaC2wBGnrYXgx2WD8LnAIqAlM9PK8ucNKyrPdjt7kGWAw0S/CctyzL+sOyrGPAWOCFW/xaRURSSxkpIuKZ8lFEJGnKSBERz5SP4jc0N6H4syOx/8bE/ns60eMB2OH5vDHmUoL7swAbPWzvLeAqMAiYD3wF9AB2J3rePUBZD9ucm+Dz/Qk+PgjckexXIiLifcpIERHPlI8iIklTRoqIeKZ8FL+hkT7iz6wUPOcIMNeyrBxxN6Ac9jDMxKrEPvcxoDBwFHs4pqdtbkq0zfuB1xM8p2iCj4sDv6egVhERb1JGioh4pnwUEUmaMlJExDPlo/gNNX0ko5sH1DLGPG2MyW6MqQv8F8jr4bnhwFhjTL7YzwOB47EfXwUKxH68ArjPGFM/dptPAf8DSiTYVg9jzO3GmOJAP+yOvYiI2ygjRUQ8Uz6KiCRNGSki4pnyUVxBTR/J0CzL+gn4FzAee9jlW0Bzy7L2enh6P+wpD3/FHhYZiL2oGsCnQANjzBTLsv4C6gCvYc+5OQvoY1nW+gTb2gbsBL7FnutzoDe/LhERb1BGioh4pnwUEUmaMlJExDPlo7iFsayUjEwTkZQyxlhAecuyvnW6FhERt1FGioh4pnwUEUmaMlJExDPlo3iikT4iIiIiIiIiIiIiIiIZgJo+IiIiIiIiIiIiIiIiGYCmdxMREREREREREREREckANNJHREREREREREREREQkA1DTR0REREREREREREREJANQ00dERERERERERERERCQDUNNHREREREREREREREQkA1DTR0REREREREREREREJAP4f+4MCr09DGK8AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 2016x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_timeseries(experiment='time_dependent', baseline='QRNN', index=range(6,7), figure_name='MQRNN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Long horizon analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training code for the three underlying RNN models:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 116.17800216674804\n",
      "Epoch: 50\tTrain loss: 105.08755569458008\n",
      "Epoch: 100\tTrain loss: 105.04140701293946\n",
      "Epoch: 150\tTrain loss: 105.00509033203124\n",
      "Epoch: 200\tTrain loss: 104.94557647705078\n",
      "Epoch: 250\tTrain loss: 105.05139083862305\n",
      "Epoch: 300\tTrain loss: 105.01086959838867\n",
      "Epoch: 350\tTrain loss: 104.9856964111328\n",
      "Epoch: 400\tTrain loss: 104.9906005859375\n",
      "Epoch: 450\tTrain loss: 105.03883361816406\n",
      "Epoch: 500\tTrain loss: 105.01057815551758\n",
      "Epoch: 550\tTrain loss: 105.00690383911133\n",
      "Epoch: 600\tTrain loss: 104.97871704101563\n",
      "Epoch: 650\tTrain loss: 104.98605880737304\n",
      "Epoch: 700\tTrain loss: 105.0848503112793\n",
      "Epoch: 750\tTrain loss: 105.0183120727539\n",
      "Epoch: 800\tTrain loss: 105.04076232910157\n",
      "Epoch: 850\tTrain loss: 105.01460037231445\n",
      "Epoch: 900\tTrain loss: 105.04418411254883\n",
      "Epoch: 950\tTrain loss: 105.0252555847168\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 115.84004592895508\n",
      "Epoch: 50\tTrain loss: 104.69863433837891\n",
      "Epoch: 100\tTrain loss: 104.61384963989258\n",
      "Epoch: 150\tTrain loss: 104.58743743896484\n",
      "Epoch: 200\tTrain loss: 104.5912986755371\n",
      "Epoch: 250\tTrain loss: 104.62083435058594\n",
      "Epoch: 300\tTrain loss: 104.56500701904297\n",
      "Epoch: 350\tTrain loss: 104.58524093627929\n",
      "Epoch: 400\tTrain loss: 104.5987564086914\n",
      "Epoch: 450\tTrain loss: 104.7249038696289\n",
      "Epoch: 500\tTrain loss: 104.68297958374023\n",
      "Epoch: 550\tTrain loss: 104.66183319091797\n",
      "Epoch: 600\tTrain loss: 104.64169158935547\n",
      "Epoch: 650\tTrain loss: 104.62773513793945\n",
      "Epoch: 700\tTrain loss: 104.58800888061523\n",
      "Epoch: 750\tTrain loss: 104.62368850708008\n",
      "Epoch: 800\tTrain loss: 104.63006973266602\n",
      "Epoch: 850\tTrain loss: 104.64454193115235\n",
      "Epoch: 900\tTrain loss: 104.6767677307129\n",
      "Epoch: 950\tTrain loss: 104.63366394042968\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 115.44023818969727\n",
      "Epoch: 50\tTrain loss: 105.30974578857422\n",
      "Epoch: 100\tTrain loss: 105.28869934082032\n",
      "Epoch: 150\tTrain loss: 105.27025985717773\n",
      "Epoch: 200\tTrain loss: 105.24124221801758\n",
      "Epoch: 250\tTrain loss: 105.21414642333984\n",
      "Epoch: 300\tTrain loss: 105.2301025390625\n",
      "Epoch: 350\tTrain loss: 105.27389373779297\n",
      "Epoch: 400\tTrain loss: 105.1761085510254\n",
      "Epoch: 450\tTrain loss: 105.15733261108399\n",
      "Epoch: 500\tTrain loss: 105.18980712890625\n",
      "Epoch: 550\tTrain loss: 105.14850463867188\n",
      "Epoch: 600\tTrain loss: 105.23087463378906\n",
      "Epoch: 650\tTrain loss: 105.1828125\n",
      "Epoch: 700\tTrain loss: 105.2318359375\n",
      "Epoch: 750\tTrain loss: 105.26486053466797\n",
      "Epoch: 800\tTrain loss: 105.24966583251953\n",
      "Epoch: 850\tTrain loss: 105.2502555847168\n",
      "Epoch: 900\tTrain loss: 105.25208892822266\n",
      "Epoch: 950\tTrain loss: 105.32386474609375\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 115.9845588684082\n",
      "Epoch: 50\tTrain loss: 104.91316223144531\n",
      "Epoch: 100\tTrain loss: 104.87761688232422\n",
      "Epoch: 150\tTrain loss: 104.83979721069336\n",
      "Epoch: 200\tTrain loss: 104.83731002807617\n",
      "Epoch: 250\tTrain loss: 104.84060211181641\n",
      "Epoch: 300\tTrain loss: 104.77823104858399\n",
      "Epoch: 350\tTrain loss: 104.7824493408203\n",
      "Epoch: 400\tTrain loss: 104.73952560424804\n",
      "Epoch: 450\tTrain loss: 104.71489944458008\n",
      "Epoch: 500\tTrain loss: 104.73666687011719\n",
      "Epoch: 550\tTrain loss: 104.73846130371093\n",
      "Epoch: 600\tTrain loss: 104.76000518798828\n",
      "Epoch: 650\tTrain loss: 104.73836517333984\n",
      "Epoch: 700\tTrain loss: 104.87026290893554\n",
      "Epoch: 750\tTrain loss: 104.9089584350586\n",
      "Epoch: 800\tTrain loss: 104.89728698730468\n",
      "Epoch: 850\tTrain loss: 104.86479110717774\n",
      "Epoch: 900\tTrain loss: 104.7954719543457\n",
      "Epoch: 950\tTrain loss: 104.8195411682129\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 112.97627029418945\n",
      "Epoch: 50\tTrain loss: 102.6340118408203\n",
      "Epoch: 100\tTrain loss: 102.61421813964844\n",
      "Epoch: 150\tTrain loss: 102.59782409667969\n",
      "Epoch: 200\tTrain loss: 102.60124130249024\n",
      "Epoch: 250\tTrain loss: 102.60901565551758\n",
      "Epoch: 300\tTrain loss: 102.66170654296874\n",
      "Epoch: 350\tTrain loss: 102.6140235900879\n",
      "Epoch: 400\tTrain loss: 102.6051040649414\n",
      "Epoch: 450\tTrain loss: 102.59608688354493\n",
      "Epoch: 500\tTrain loss: 102.5944725036621\n",
      "Epoch: 550\tTrain loss: 102.57596130371094\n",
      "Epoch: 600\tTrain loss: 102.61970672607421\n",
      "Epoch: 650\tTrain loss: 102.60094375610352\n",
      "Epoch: 700\tTrain loss: 102.56781539916992\n",
      "Epoch: 750\tTrain loss: 102.5620346069336\n",
      "Epoch: 800\tTrain loss: 102.54615936279296\n",
      "Epoch: 850\tTrain loss: 102.52875213623047\n",
      "Epoch: 900\tTrain loss: 102.72762603759766\n",
      "Epoch: 950\tTrain loss: 102.58450775146484\n",
      "Training CFRNN\n",
      "Training dataset 0\n",
      "Epoch: 0\tTrain loss: 115.83352279663086\n",
      "Epoch: 50\tTrain loss: 105.03586120605469\n",
      "Epoch: 100\tTrain loss: 104.94470520019532\n",
      "Epoch: 150\tTrain loss: 104.88064880371094\n",
      "Epoch: 200\tTrain loss: 104.76230850219727\n",
      "Epoch: 250\tTrain loss: 104.71804656982422\n",
      "Epoch: 300\tTrain loss: 104.87625427246094\n",
      "Epoch: 350\tTrain loss: 104.85363311767578\n",
      "Epoch: 400\tTrain loss: 104.72554016113281\n",
      "Epoch: 450\tTrain loss: 104.61922225952148\n",
      "Epoch: 500\tTrain loss: 104.78587036132812\n",
      "Epoch: 550\tTrain loss: 104.95186157226563\n",
      "Epoch: 600\tTrain loss: 104.85871658325195\n",
      "Epoch: 650\tTrain loss: 104.78477630615234\n",
      "Epoch: 700\tTrain loss: 105.01043853759765\n",
      "Epoch: 750\tTrain loss: 104.77209396362305\n",
      "Epoch: 800\tTrain loss: 104.83765487670898\n",
      "Epoch: 850\tTrain loss: 104.7858268737793\n",
      "Epoch: 900\tTrain loss: 104.69740371704101\n",
      "Epoch: 950\tTrain loss: 104.87532348632813\n",
      "Training dataset 1\n",
      "Epoch: 0\tTrain loss: 115.75688247680664\n",
      "Epoch: 50\tTrain loss: 104.66805038452148\n",
      "Epoch: 100\tTrain loss: 104.57697830200195\n",
      "Epoch: 150\tTrain loss: 104.56703414916993\n",
      "Epoch: 200\tTrain loss: 104.53413772583008\n",
      "Epoch: 250\tTrain loss: 104.349169921875\n",
      "Epoch: 300\tTrain loss: 104.36441192626953\n",
      "Epoch: 350\tTrain loss: 104.37173538208008\n",
      "Epoch: 400\tTrain loss: 104.3967529296875\n",
      "Epoch: 450\tTrain loss: 104.47179946899413\n",
      "Epoch: 500\tTrain loss: 104.33271636962891\n",
      "Epoch: 550\tTrain loss: 104.43269958496094\n",
      "Epoch: 600\tTrain loss: 104.25617599487305\n",
      "Epoch: 650\tTrain loss: 104.25851974487304\n",
      "Epoch: 700\tTrain loss: 104.31623458862305\n",
      "Epoch: 750\tTrain loss: 104.25134353637695\n",
      "Epoch: 800\tTrain loss: 104.44346466064454\n",
      "Epoch: 850\tTrain loss: 104.3064453125\n",
      "Epoch: 900\tTrain loss: 104.60414505004883\n",
      "Epoch: 950\tTrain loss: 104.7297492980957\n",
      "Training dataset 2\n",
      "Epoch: 0\tTrain loss: 116.2881591796875\n",
      "Epoch: 50\tTrain loss: 105.29717864990235\n",
      "Epoch: 100\tTrain loss: 105.24379653930664\n",
      "Epoch: 150\tTrain loss: 105.14953536987305\n",
      "Epoch: 200\tTrain loss: 105.02491836547851\n",
      "Epoch: 250\tTrain loss: 104.82961502075196\n",
      "Epoch: 300\tTrain loss: 105.04059982299805\n",
      "Epoch: 350\tTrain loss: 104.90226364135742\n",
      "Epoch: 400\tTrain loss: 104.83191452026367\n",
      "Epoch: 450\tTrain loss: 104.69211273193359\n",
      "Epoch: 500\tTrain loss: 104.64795532226563\n",
      "Epoch: 550\tTrain loss: 104.61201934814453\n",
      "Epoch: 600\tTrain loss: 104.58943099975586\n",
      "Epoch: 650\tTrain loss: 104.79262237548828\n",
      "Epoch: 700\tTrain loss: 104.74981079101562\n",
      "Epoch: 750\tTrain loss: 104.56293640136718\n",
      "Epoch: 800\tTrain loss: 104.73923950195312\n",
      "Epoch: 850\tTrain loss: 104.79753494262695\n",
      "Epoch: 900\tTrain loss: 104.66920394897461\n",
      "Epoch: 950\tTrain loss: 104.61814193725586\n",
      "Training dataset 3\n",
      "Epoch: 0\tTrain loss: 116.19977188110352\n",
      "Epoch: 50\tTrain loss: 104.91102600097656\n",
      "Epoch: 100\tTrain loss: 104.8573226928711\n",
      "Epoch: 150\tTrain loss: 104.81196746826171\n",
      "Epoch: 200\tTrain loss: 104.77023391723633\n",
      "Epoch: 250\tTrain loss: 104.73728866577149\n",
      "Epoch: 300\tTrain loss: 104.77311553955079\n",
      "Epoch: 350\tTrain loss: 104.66194763183594\n",
      "Epoch: 400\tTrain loss: 104.64828338623047\n",
      "Epoch: 450\tTrain loss: 104.59880142211914\n",
      "Epoch: 500\tTrain loss: 104.84854125976562\n",
      "Epoch: 550\tTrain loss: 104.79998168945312\n",
      "Epoch: 600\tTrain loss: 104.76977157592773\n",
      "Epoch: 650\tTrain loss: 104.72329483032226\n",
      "Epoch: 700\tTrain loss: 104.6841667175293\n",
      "Epoch: 750\tTrain loss: 104.63412094116211\n",
      "Epoch: 800\tTrain loss: 104.65337219238282\n",
      "Epoch: 850\tTrain loss: 104.60695037841796\n",
      "Epoch: 900\tTrain loss: 104.51512832641602\n",
      "Epoch: 950\tTrain loss: 104.73683013916016\n",
      "Training dataset 4\n",
      "Epoch: 0\tTrain loss: 113.11035995483398\n",
      "Epoch: 50\tTrain loss: 102.62748336791992\n",
      "Epoch: 100\tTrain loss: 102.56647262573242\n",
      "Epoch: 150\tTrain loss: 102.45281600952148\n",
      "Epoch: 200\tTrain loss: 102.26483993530273\n",
      "Epoch: 250\tTrain loss: 102.22038879394532\n",
      "Epoch: 300\tTrain loss: 102.26835861206055\n",
      "Epoch: 350\tTrain loss: 102.04121551513671\n",
      "Epoch: 400\tTrain loss: 102.16726608276367\n",
      "Epoch: 450\tTrain loss: 102.11195831298828\n",
      "Epoch: 500\tTrain loss: 101.98408966064453\n",
      "Epoch: 550\tTrain loss: 102.05206146240235\n",
      "Epoch: 600\tTrain loss: 101.97759475708008\n",
      "Epoch: 650\tTrain loss: 102.05753479003906\n",
      "Epoch: 700\tTrain loss: 101.9084342956543\n",
      "Epoch: 750\tTrain loss: 102.51231842041015\n",
      "Epoch: 800\tTrain loss: 101.9041015625\n",
      "Epoch: 850\tTrain loss: 102.19611206054688\n",
      "Epoch: 900\tTrain loss: 101.92498550415038\n",
      "Epoch: 950\tTrain loss: 101.86601867675782\n"
     ]
    }
   ],
   "source": [
    "for rnn_mode in ['LSTM', 'RNN', 'GRU']:\n",
    "    run_synthetic_experiments(experiment='static', \n",
    "                              baseline='CFRNN',\n",
    "                              horizon=100,\n",
    "                              retrain_auxiliary=True,\n",
    "                              recompute_dataset=True,\n",
    "                              save_model=True, \n",
    "                              save_results=False, \n",
    "                              rnn_mode=rnn_mode)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Analyzing interval widths:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "widths = {}\n",
    "n_train=2000\n",
    "for rnn_model in ['LSTM', 'RNN', 'GRU']:\n",
    "    model = torch.load('saved_models/static-CFRNN-{}-5-0-horizon100.pt'.format(rnn_model))\n",
    "    \n",
    "    ws = []\n",
    "    alphas = np.arange(100) * 0.01\n",
    "    for alpha in alphas:\n",
    "        corrected_alpha = alpha / model.horizon\n",
    "        corrected_critical_calibration_scores = torch.tensor([[\n",
    "                torch.quantile(\n",
    "                    position_calibration_scores,\n",
    "                    q=1 - corrected_alpha * n_train / (n_train + 1))\n",
    "                for position_calibration_scores in feature_calibration_scores]\n",
    "                for feature_calibration_scores in model.calibration_scores]).T\n",
    "        ws.append(corrected_critical_calibration_scores.flatten().tolist())\n",
    "    ws = np.array(ws)\n",
    "    ws = ws * 2\n",
    "    widths[rnn_model] = ws"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_valid_horizons(threshold_width):\n",
    "    valid_horizons = {}\n",
    "    for rnn_model in ['LSTM', 'RNN', 'GRU']:    \n",
    "        T = []\n",
    "        ws = widths[rnn_model]\n",
    "        for i, _ in enumerate(alphas):\n",
    "            t = 0\n",
    "            while t < ws.shape[1] and ws[i, t] < threshold_width:\n",
    "                t += 1\n",
    "            T.append(t)\n",
    "        valid_horizons[rnn_model] = np.array(T)\n",
    "    return valid_horizons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams.update({\n",
    "    \"text.usetex\": False,\n",
    "    \"font.family\": \"sans-serif\",\n",
    "    \"font.sans-serif\": [\"Helvetica\"]})\n",
    "plt.rcParams.update({'axes.titlesize': 16,\n",
    "                     'axes.labelsize': 13})\n",
    "for threshold_width in [28]:\n",
    "    valid_horizons = compute_valid_horizons(threshold_width)\n",
    "    for rnn_model in valid_horizons.keys():\n",
    "        plt.plot(1-alphas, valid_horizons[rnn_model], label='CFRNN ({})'.format(rnn_model))\n",
    "plt.xlabel('Coverage $1-\\\\alpha$')\n",
    "plt.ylabel('Horizon $H$')\n",
    "plt.legend()\n",
    "plt.savefig('{}.png'.format('tradeoff') , bbox_inches='tight', dpi=1000)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}