{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "2b1ea32f",
   "metadata": {},
   "source": [
    "# nb02 · 공식 코드로 보는 FNO1d — Bouc–Wen 비선형 SDOF 지진응답\n",
    "\n",
    "**깊게 파기 — Neural Operator · W1**\n",
    "\n",
    "### 이 노트북에서 배우는 것\n",
    "- nb01의 \"주파수영역에서 $H(\\omega)$ 를 곱한다\"가 FNO에서 **학습된 복소 가중치 곱 + 저모드 절단**으로 바뀌는 걸 **공식 코드**로 확인한다.\n",
    "- 공식 구현의 핵심 클래스 `SpectralConv1d` / `FNO1d` 를 노트북에 **그대로 옮겨**(아래 출처), 선형 $H(\\omega)$ 로는 못 푸는 **Bouc–Wen 이력(비선형)** SDOF 지진응답 $a_g\\!\\to\\!u$ 에 적용한다.\n",
    "- **훈련이 목적이 아니다.** batch 10, epoch 1 **smoke run** — 텐서 shape가 맞물려 돌아가고 loss가 한 번 내려가는지만 확인한다.\n",
    "\n",
    "### 출처 (2026-10-29 실제 clone·확인)\n",
    "- 공식 repo: **github.com/neuraloperator/neuraloperator**, commit `a6648eeeb60fe0837f2d7d021517a945a159a0c8`.\n",
    "- 스펙트럴 컨볼루션(1D forward): `neuralop/layers/legacy_spectral_convolution.py` → `class SpectralConv1d.forward` (L573–607): `rfft → 0버퍼(width//2+1) → 앞쪽 half_n_modes개 모드만 contract → irfft(n=width) → +bias`.\n",
    "- 모드 곱(dense contraction): 같은 파일 `_contract_dense` (L33–58). 1D einsum은 `\"abc,bdc->adc\"` = (batch,in,x)·(in,out,x)→(batch,out,x), 즉 아래 `compl_mul1d`의 `\"bix,iox->box\"` 와 동일.\n",
    "- 전체 아키텍처: `neuralop/models/fno.py` → `class FNO` (L25~). `lifting(ChannelMLP) → FNOBlocks → projection(ChannelMLP)`, 기본 비선형 `F.gelu`(L75,186), 기본 skip `fno_skip=\"linear\"`·`channel_mlp_skip=\"soft-gating\"`.\n",
    "- 아래 코드는 이 구조를 **교육용 1D 최소형**으로 전사한 것: `lifting/projection`은 `ChannelMLP` 대신 등가의 `nn.Linear`, skip은 `nn.Conv1d(·,·,1)`(pointwise) 사용. 모드 절단·복소곱·`rfft/irfft` 순서는 공식과 동일."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "028adea3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:17.048320Z",
     "iopub.status.busy": "2026-10-29T09:18:17.048142Z",
     "iopub.status.idle": "2026-10-29T09:18:18.378082Z",
     "shell.execute_reply": "2026-10-29T09:18:18.377671Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch 2.8.0 | device: cpu\n"
     ]
    }
   ],
   "source": [
    "# [공통] 라이브러리 + 시드\n",
    "import numpy as np, torch, torch.nn as nn, torch.nn.functional as F\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "torch.manual_seed(0); np.random.seed(0)\n",
    "rng = torch.Generator().manual_seed(0)\n",
    "print(\"torch\", torch.__version__, \"| device:\", \"cpu\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b99595be",
   "metadata": {},
   "source": [
    "### 1. Bouc–Wen 이력 SDOF — 왜 선형 $H(\\omega)$ 로 안 되나\n",
    "\n",
    "복원력이 변위에 선형이 아니라 **이력(hysteresis)** 을 그린다:\n",
    "$$ m\\ddot u + c\\dot u + \\big[\\alpha k\\,u + (1-\\alpha)k\\,z\\big] = -m\\,a_g(t), $$\n",
    "$$ \\dot z = A\\dot u - \\beta|\\dot u|\\,|z|^{n-1}z - \\gamma\\,\\dot u\\,|z|^{n}. $$\n",
    "- $z$ = 이력변위(상태변수), $\\alpha$ = 항복후 강성비, $\\beta,\\gamma,A,n$ = 이력 형상.\n",
    "- **비선형**이라 전달함수 $H(\\omega)$ 하나로 입력→출력을 못 쓴다. 입력 지진동마다 응답 모양이 달라진다 → \"연산자 $a_g\\mapsto u$ 를 **데이터로 학습**\"할 동기.\n",
    "- 상태 $[u,\\dot u,z]$ 를 고정스텝 RK4로 적분(빠르고 결정적)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e19e54ba",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:18.379907Z",
     "iopub.status.busy": "2026-10-29T09:18:18.379673Z",
     "iopub.status.idle": "2026-10-29T09:18:18.621307Z",
     "shell.execute_reply": "2026-10-29T09:18:18.620927Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "지진동 AG:"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " (60, 256)  응답 UU: (60, 256) (N, L)\n",
      "응답 최대변위 범위: 0.0122 ~ 0.0736 m\n"
     ]
    }
   ],
   "source": [
    "# [1] Bouc-Wen SDOF 데이터 생성 (RK4, 작은 데이터셋)\n",
    "L   = 256                       # 시퀀스 길이(시간 격자점)\n",
    "dt  = 0.02                      # 50 Hz, 총 5.12 s\n",
    "m, zeta = 1.0, 0.03\n",
    "wn  = 2*np.pi*2.0               # 2 Hz\n",
    "k   = m*wn**2; c = 2*m*zeta*wn\n",
    "alpha, A, beta, gamma, nexp = 0.1, 1.0, 2.0, 2.0, 1.0   # 이력 파라미터(강한 비선형)\n",
    "\n",
    "def jennings_env(t, t_rise=0.6, t_level=2.0, decay=0.6):\n",
    "    e=np.ones_like(t); e[t<t_rise]=(t[t<t_rise]/t_rise)**2\n",
    "    tail=t>(t_rise+t_level); e[tail]=np.exp(-decay*(t[tail]-(t_rise+t_level))); return e\n",
    "\n",
    "def ground_motion(seed):\n",
    "    g=np.random.default_rng(seed); w=g.standard_normal(L)\n",
    "    W=np.fft.rfft(w); f=np.fft.rfftfreq(L,dt)\n",
    "    W[(f<0.5)|(f>6.0)]=0.0                       # 0.5~6 Hz 대역\n",
    "    ag=np.fft.irfft(W,n=L)*jennings_env(np.arange(L)*dt)\n",
    "    return (ag/ (np.max(np.abs(ag))+1e-9) * g.uniform(2.0,5.0)).astype(np.float64)  # 세기 랜덤\n",
    "\n",
    "def bw_deriv(s, ag_t):\n",
    "    u,v,z = s\n",
    "    zdot = A*v - beta*abs(v)*abs(z)**(nexp-1)*z - gamma*v*abs(z)**nexp\n",
    "    vdot = (-c*v - alpha*k*u - (1-alpha)*k*z - m*ag_t)/m\n",
    "    return np.array([v, vdot, zdot])\n",
    "\n",
    "def simulate(ag):\n",
    "    s=np.zeros(3); U=np.zeros(L)\n",
    "    for i in range(L-1):\n",
    "        a0,a1=ag[i],ag[i+1]; am=0.5*(a0+a1)\n",
    "        k1=bw_deriv(s,a0); k2=bw_deriv(s+0.5*dt*k1,am)\n",
    "        k3=bw_deriv(s+0.5*dt*k2,am); k4=bw_deriv(s+dt*k3,a1)\n",
    "        s=s+dt/6*(k1+2*k2+2*k3+k4); U[i+1]=s[0]\n",
    "    return U\n",
    "\n",
    "N=60                                            # 샘플 수(작게)\n",
    "AG=np.stack([ground_motion(s) for s in range(N)])\n",
    "UU=np.stack([simulate(AG[i]) for i in range(N)])\n",
    "print(\"지진동 AG:\", AG.shape, \" 응답 UU:\", UU.shape, \"(N, L)\")\n",
    "print(\"응답 최대변위 범위:\", f\"{np.abs(UU).max(1).min():.4f} ~ {np.abs(UU).max(1).max():.4f} m\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f4aa461d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:18.623104Z",
     "iopub.status.busy": "2026-10-29T09:18:18.622966Z",
     "iopub.status.idle": "2026-10-29T09:18:18.819281Z",
     "shell.execute_reply": "2026-10-29T09:18:18.818855Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABEEAAAD5CAYAAADItyvfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8ekN5oAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOxdB3gUVRe90iFAIARC7733qnRBwIKiIhYQG/auv1ix995BBRuCqKACokiR3nvvHQKhhN75v3N37uzsZGZ3ZnbS3/m+gWSzZba9d995555z0YULFy6QgoKCgoKCgoKCgoKCgoKCQhZHjvQ+AQUFBQUFBQUFBQUFBQUFBYW0gCJBFBQUFBQUFBQUFBQUFBQUsgUUCaKgoKCgoKCgoKCgoKCgoJAtoEgQBQUFBQUFBQUFBQUFBQWFbAFFgigoKCgoKCgoKCgoKCgoKGQLKBJEQUFBQUFBQUFBQUFBQUEhW0CRIAoKCgoKCgoKCgoKCgoKCtkCigRRUFBQUFBQUFBQUFBQUFDIFlAkiIKCgoKCgoKCgoKCgoKCQraAIkEUImLYsGF00UUX0ZYtW9L7VDId7r33Xrr00ksdXfeGG26g66+/PtXPSUFBQUEhdbF9+3bKly8fzZw5M+J1V61aRbly5aIVK1akybkpKCgopAewlhg0aFCGrMHNaN++PdWtW5cyIvbv308xMTE0fvz49D6VTA1FgihkOqBgxCCa0UmZzZs301dffUVPP/20ftmuXbv43JcsWZLi+v/73//o119/paVLl6bxmSooKCgo+ImXXnqJWrRoQW3atNEvGz58OH3wwQcprlu7dm3q0aMHPf/882l8lgoKCgpZs+63qsGzCooVK0Z33HEHPffcc+l9KpkaigRRiIhbbrmFTpw4QRUqVKCMMhi++OKLGZ4E+fDDD6lSpUrUoUOHEBIE525FgjRq1IiaNm1K7777bhqfqYKCgoKCX9i3bx99++23dPfdd4dcbkeCALju6NGjaePGjWl0lgoKCgqZA17qfqsaPCsBc8aiRYto8uTJ6X0qmRaKBFGIiJw5c7KsFzI2BWc4c+YM/fjjj67bW3D93377jY4ePZpq56agoKBw7Nix9D6FLIsffviB21uuuOIKx7fp3LkzFS1alMkTBQUFBYW0r8EzE2rVqsXtOrAsUPAGRYIoePIEqVixIl1++eU0Y8YMat68OZMklStXpu+++87yttOmTaMBAwawhKtw4cLUt29fOnjwoKNeQTzWrbfeqt/fddddxz+D3cVtcEydOtX2/JctW8a3x/nhPEuWLEm33XYb99S5BW4DZQyeQ5EiRahfv37cvoJzMA5EeF2SkpK4sBXgHJs1a8Y/9+/fXz934+3Qu4jFycSJE12fm4KCgoIVMK5irMFu2o033siL7Ysvvjhk0d6kSRPKnz8/xcXFsT8RPC2MWL9+PfXq1YvHT4yjZcuW5eslJyfr18Fj3H///Vx81qhRg6+H+8X4b8bixYupW7duPJYWLFiQOnXqRHPmzLGcP+Cr8eijj1Lx4sW5D/rqq69mtYURCxYsoK5du1J8fDw/D+wAYpw34vz586zEqFOnDp9bQkICz0vmuciuPxyHGZhbMEcZMWbMGG6FwfMy3n7cuHG0detWfew33i537tx8nd9//z3iuSgoKGQ+WI0VxvE5HDCuYjw5fvx4ir/16dOHx+Vz5845Hgut4KQGd4pTp07RI488wmN2oUKF6Morr6QdO3ZYXjfSXOCl7reqwQHcBrcdOXIkt8ngdcOcgvMzz3kCzJt43AIFClCZMmXorbfeCvn76dOnuZURc11sbCzf3yWXXEJTpkxJcV8jRozg6+E1wfOtV68eK1aMOHToED388MNUrlw5yps3L1WtWpXefPNNnr/MwJrhzz//pAsXLti+Fgr2yJXeJ6CQebFhwwa69tpr6fbbb2cy4JtvvuGBEl9wFJnmARykAQbYtWvX0ueff87FoAxITtG2bVt68MEH6aOPPuIBDEwoIP9bAYTCpk2bmHjAgLdy5UoaPHgw/4+B1unjYwDCzt68efPonnvuoZo1a3LBiuduxqxZs/h+0eIiwDmiTxyD5V133cWDJNC6deuQ3nBMWij6UegrKCgo+AUUktWqVaPXXntNL5peffVV7ivGjhl6jEEufPzxxzzWojjFuI0iD0U1CtsHHniAx9GdO3fS2LFjuWBD4Sf477//uMDEOI0C7rPPPqPLLruMx00xmcPYi/EPReCTTz7JBMCXX37JJABuDwLBCDwmiJsXXniByXgQGZhT8DjA3r17qUuXLlxwP/XUU3zOuB5UdUaA8EBBjbkA54ee8U8++YSfJ8ZcnIcfO5Dz58/nOcKIZ555hgkjLATef/99vsxIkgCYOzGnHD58mF8bBQUFBaB379706aefMpEqhAAAUgSLYNTeUG07HQtTG5hLQK6DdEeNi5YN+B6Z4WQu8FL3W9XgRmDew9/hxYfXDHMKCBO0qqMGF4Agx/x1zTXX8Bz5yy+/8G1AXoC4ATBew3sEZNSdd95JR44coa+//prnTMx7DRs21NciuA5IHpAawOrVq3nueeihh/T3s127djy/Yr4qX748P5eBAwfS7t27U7RTYs7AfILXMaOauGZoXFBQiIChQ4eiWr6wefNm/bIKFSrwZdOmTdMv27t374W8efNeeOyxx1LctkmTJhdOnz6tX/7WW2/x5b///rt+GX5/4YUXUjw+Hqtfv37676NGjeLrTpkyxdH5Hz9+PMVlP/30U4rzj4Rff/2Vb/PBBx/ol507d+5Cx44d+XI8V8HNN998oVixYinuY/78+Smua0b16tUvdOvWzfF5KSgoKIQDxlWMO3369Am5fMuWLRdy5sx54dVXXw25fPny5Rdy5cqlX7548WK+PcbecMB1cCxYsEC/bOvWrRfy5ct34eqrr9Yv69mz54U8efJc2Lhxo37Zrl27LhQqVOhC27ZtU8wfnTt3vnD+/Hn98kceeYTP+9ChQ/z76NGj+XoYX+0wffp0vs6PP/4YcvmECRMsLzejXbt2fJiBuQlzlGDDhg18fx9//HGK6/bo0SPkumYMHz6cbzt37tyw56KgoJD5YB4rzONzOGD8K1OmzIVevXqFXP7zzz+H1LJOxkI7OK3BI2HJkiV8X/fee2/I5TfeeGOKx3A6F7it++1qcNwe94PX8vDhwylexw8//FC/DOM9Lvvuu+/0y06dOnWhZMmSIe/D2bNn+XIjDh48eCEhIeHCbbfdpl/20EMPXShcuDBf3w4vv/zyhZiYmAvr1q0Lufypp57iOW/btm0hl8+aNYvPceTIkQ5eFQUzVDuMgmdAtSBqBgDMMyTQUF2YAeWDcZcNu2TomU6LeCcjq3vy5EmWyLVs2ZJ/h6mQU0yYMIGfA5heQY4cOei+++6zbJvBzqUX4HY4RwUFBQU/YTbqxO4gFG7Y4cKYIweUHlCMiJxXlB5///23pRzbiFatWvHulAA7WVdddRXfFnJtHP/88w/17NmTWxQFpUqV4l1DyJixs2aeP4yKPcw7uB+oCQHsdgJQpkCJYYVRo0bx84B82Phcca5QZFhJl71A2iy9jP9yGzX+KygoGIHxDwoQ1MxGzzio4dCiIe2NTsbC1IbU9VBvGIEWDyO8zAVOEakGR0s+WlIEULXjcc1rEswNN998s/57njx52ALAuM6BAgeXA5hPDxw4QGfPnuWgA+MaA+9NpHZ3zFOY32QdIAdUKni9zK2las6IDooEUfAMFLdm4Atp1V+Ngto8sGDASYuEFwxIkJqh/xuECMga9EgCxn72SEDBjXNGX6AR6NezgtcePdxOmdAqKCj4DRn3jD4fGG8wPmNcNB6Q6UImLLeDJwckv+gzh8wX0myr8dM81gPVq1dn8gStNjjwMwhzMyBvRhFp7s02zzVS+MlcA/kw/EqQHoDzA+kydOhQbt8xPlecb4kSJVI8Vywq5Ln6BS/jv9xGjf8KCgpWLTFIavzjjz/4d4xbWLSDHJExw8lYmNpArYwNwipVqoRcbh7zvcwFfo3B5nkKrx9qefOaBN5X5vHYap0DQ+v69euz1xS8DzGvoHXJOEfee++9PBeijQb3C58WbK4agXkKl5nnKPE2Mc9Tas6IDsoTRMEzwH5awW+DHjF78grscqKn7oknnuDePBAwGFzR52dlNOQHMAg6MduzAm5ntZBQUFBQ8EsVB2D8Q/H0119/WY7nRs8KRHej7xyeFdi9wy7f66+/zr5KKOjSc67Bc0CvNs4F/fFQnaDAxDnjMhnzQYDAtNUKKDTDAY9hNbeZ5yeM/YCX8V9ug8WLgoJC1oLdQtVpjQsFM0xKf/75Z1ZKYKwDKQJyxPgYkcbCtK7B0wPR1OBu1znwPsHcCEUL1hmYZ3A7zI/GyHNcDs8RvCeYc3GAoIIqRVLBME9BrQh/FCuARDFCzRnRQZEgCmkCsJvGrG4w2DD56d69ewi7CpM9I2DIh+sZ4YbxxAAxadIkZsVhSGo8H7eoUKECS6bBXBvVIDCINQOmqSi2wQIbTQMjnTskdGC+4VStoKCgkJrATh2KOSg9zMWVFWAGh+PZZ59lYrlNmzb0xRdf0CuvvBJ2bF23bh2PmUI04GcYZJuxZs0a3kGEK74XYJGAA6Z3w4cPp5tuuond+GHSh+f677//8jmbySAnwPxk1eopLTlG1QruH6arZkQa/3EbPH8n74WCgkLmglWNazWGRNrUQ5oI2kTQCgNSRNq7nY6Fbs7PqgZ3UitjMQ8CwKjyMI/5mA+czgVulQ52NbjdPIV5ELU81BxuAdIJ7TxoLzWeJ4y8zUDbDAIWcOA1gjoERrAwJ4cSBfMU1kfmVBs7yDwTziRWwR6qHUYhTYA0FmN/ItJhsOAXd2UAX35zvxtuZ2ahET8FWE0mdiyueQfP7LDsBJCA4zkMGTJEvwyDGGThVn3xeMyFCxe6OndEccG3xJgYo6CgoJAagOM9xkiQxOYxEr+LvwUKbozXRoAMQZFqllnPnj07pA8apC7UI0gswGPhwM+4zCg9TkxM5GIdve1uk1FAdpvPXxz55fyweMBc8vLLL6e4PZ5bpPkE8xMKc2M0L+LR4exvBHyj0AuOmEozMP6Ha8HEfIFkNauiXUFBIXMDYwi+/8uWLdMvA8EwevRox/cB1QfGNCgH0DaBcc3tWBju/JzU4JEgdT3SXMLV3W7mAjd1f7gaXPDdd99xiouRyMB7YVyTOIXVOmPu3Lk8Fxoh86kA86eQLsZ5CreDWsQMPHfzPIznh/nCnMip4AxKCaKQJgCbjFgofMHB+iI2EQOcUfEAhhrGfehnhBwMBSYGArPMCwM6Bh1ETGFCQQxjx44dWWpmBgZQxGsh1xsEBgykIOW22qWLBEjdYIj02GOPMWMMphm9mfAcAYwMMJ4b5HjYecS5GScZmCNh9xSmTBjYEQEmvfowTAIzjuevoKCgkJrAeAQVB+L3UIRijMO4hPERhTkMSR9//HGON0QkLXrPoVJAIfb999/zOIzx2gjE9IEwNkbkAiBaBHhMjHUYJ7ETBpNs7IahEMRY7RZYEOBxECuO54TiFmQ1xn9RG6JXHpGDkChDkoziG4QFdgRhRofdVZjj2QGS8vfee4+fG2Lh0ZuNcRzFp9m8D334iMQ1R93ChBW7t/BXadasGUvTsSMIYH5CJCReDwUFhayHG264geNVMU5hfISqGBuCGFOdmvQ3btyYFQMYXzBeGlthnI6FdnBag0cCanREweI8UKNjUw+KbCvVtNO5wE3dH64GF8TFxfF1EJcO0gUEDV5XY/CBU1x++eWsAsFrjhhgzJ+YGxAeYTSxxeuL9QLOBy2kUAAhjh7PTZQcaKfBugL3iRYbzBkwU12+fDkTNZinje8HXjvMIcoTxCNS5MUoKDiMyEXcX6QYQbntf//9d+Guu+66ULRo0QsFCxa8cNNNN13Yv39/yG0RN/u///3vQnx8/IUCBQpc6Nq1K8cNWsVzDRky5ELlypU5MipSbNaOHTs4nrFIkSIXYmNjL1x33XUcwWUXBxYO+/bt45gvxHfhvm699dYLM2fO5PsaMWJEyHUffPDBC1WrVk1xH4gFrl27NkdQmuNyW7RowdFeCgoKCn5BIhgxftnFf1988cUczYejZs2aF+67774La9eu5b9v2rSJo/6qVKnCcbdxcXEXOnTocOHff/8NuR88Bm73ww8/XKhWrRpHpjdq1MhyfF60aBGP8ZgPMN7j/hD3Z4TMH+a4R4k5lPvFfSH+t3z58vyYJUqUuHD55ZeHRPUKBg8ezJHt+fPn53G8Xr16F5588kmeEyIBzwvzDiIdGzZseOHvv/+2jL1MTEzk8f37778Pufzo0aM8f2Auwvkbb/fXX3/xZevXr494HgoKCpkT//zzz4W6devyGFKjRg0eU5xE5BrxzDPP8PWt6ks3Y6EZbmrwSDhx4gTXwIipxZxyxRVXXNi+fbtl3e1kLnBb99vV4DJ3/PTTTxcGDhzIrw/mAqxnEOduBNYyderUSXG/5jEf8cWvvfYaXyZz3tixY1Nc75dffrnQpUsXfky8/3iPBgwYcGH37t0h93/kyBE+N5w7rof3o3Xr1hfeeeedC6dPn9avt3r1an4u5nlYwTkuwj9eCRQFhUgYNmwYM63z589niXBWxJgxY5gBRpwX+s0F6B+HWgTmR1DBRAJ2J8HyY0dA5IsKCgoKmQXYjUJk+CeffELZHVCLwAtl+vTpjq4PFQ5ePzfSeAUFBQUFa1jV4FOnTmV/Qqj/win/MgMQOYz2JbTEKCWINyhPEAUFF4ATtxHolYScDVJDEBhGwCgJhfAbb7zh6L5xPQzKigBRUFBQyNyAKR7If7NniBUQRzx27FhLvxIFBQUFBfdwW4NnJsBfBJH1aCdSBIh3KE8QhWwPEBlGszsroHcbxwMPPMBECEyX0LOIPkCkJLz22muWiQPo93QKOHcrKCgoKGR+ICUGJtdOgH5ws+GdgoKCQkbDnj17wv4ddXBGMnZ2U4NnJsDvxOg3ouANigRRyPZAeoEYk4bb1Rs0aBAbGiFvHbt2KHBhpAQlCEwDFRQUFBQUFBQUFLIiSpUqFfbv/fr14zZ4BYXMAOUJopDtATIDfh6RZHU4FBQUFBQUFBQUFLIbkLYSDqVLl+ZUFAWFzABFgigoKCgoKCgoKCgoKCgoKGQLZOh2mPPnz9OuXbuoUKFCyvhFQUEhzQGO+MiRI7y7kSOH8pH2E2p8V1BQSC+osT31oMZ2BQWFzDC+Z2gSBINouXLl0vs0FBQUsjngG1O2bNn0Po0sBTW+KygopDfU2O4/1NiuoKCQGcb3DE2CgEWWJ4EIUgUFBYW0xOHDh7mYk7FIwT+o8V1BQSG9oMb21IMa2xUUFDLD+J6hSRCR0WEQVQOpgoJCekFJev2HGt8VFBTSG2ps9x9qbFdQUMgM47tqhFRQUFBQUFBQUFBQUFBQUMgWUCSIgoKCgoKCgoKCgoKCgoJCtoAiQRQUFBQyMT799FOqWLEi5cuXj1q0aEHz5s0Le/1Ro0ZRzZo1+fr16tWj8ePHh/z91ltvZQmh8bjssstCrnPgwAG66aabWOpcpEgRuv322+no0aOp8vwUFBQUFBQUFBQU/IQiQTIgdh46QWfOnU/v01BQUMjgGDlyJD366KP0wgsv0KJFi6hBgwbUtWtX2rt3r+X1Z82aRX369GHSYvHixdSzZ08+VqxYEXI9kB67d+/Wj59++ink7yBAVq5cSRMnTqSxY8fStGnT6K677krV56qgkBGBKL6N+47y/woKCgoKQZw6e47Hx0XbDtKGvUdo7+GTdPLMOTVeKmQIXHQhA38S4e4aGxtLycnJ2cZcafzy3XTvj4vogY5V6bEuNdL7dBQUsjUy+hgE5UezZs3ok08+4d/Pnz/PjtgPPPAAPfXUUymu37t3bzp27BgTF4KWLVtSw4YN6YsvvtCVIIcOHaIxY8ZYPubq1aupdu3aNH/+fGratClfNmHCBOrevTvt2LGDc9mtcOrUKT7M7t0Z9bVVUHCCD/5dRx/8u56evKwG3du+anqfjkIWGdszM9Rrm32AJeSh42do64HjtA3H/mP8/9b9x2n7geO0+/BJslpl5s55ERXOl5sK589NhfPl0v7H77moUL6UlwWvm5sK5ctFBfLkVKbGClGPQRk6HSY7AgQI8PHkDYoEUVBQsMXp06dp4cKFNHDgQP2yHDlyUOfOnWn27NmWt8HlUI4YAeWImfCYOnUqlShRgooWLUodO3akV155hYoVK6bfB1pghAAB8Jh47Llz59LVV19t+divv/46vfjii1E9ZwWFjIRDx08zAQK8NWEtXdekHBUvlDdVHgu7qIP+WEWJh0/SD3e0oITC+VLlcRQUFBSMOHvuPO06dDJAbhwIkBzb9gvpcZyOnDob9vYgLIoWyEPHTp+lwyfO0PkLRGfOXaD9x07z4QU5c1xkSZSAILEmV0IJloJ5cykSRUGRIBkJkIgpKCgoOEFSUhKdO3eOEhISQi7H72vWrLG8zZ49eyyvj8uNrTDXXHMNVapUiTZu3EhPP/00devWjcmPnDlz8nVBkBiRK1cuiouLC7kfM0DWGAkYUYIoKGRWDJ+3LeT3H+ZspUcurZ4qj/XwyCW0Yudh/vn3JTvprrZVUuVxFBQUsh+OnDwTQm5A2bFdU3SgRf8cmIswSCicl8rHFaDycTH8f4ViBaic9n+xmDw64QDlyLHT55gMOXzyDB0+cVb/+cjJs6GXnwz+jPM7rP397PkLfD4Hj5/hwwtyXEQUXzAvNa8UR22qxlObKvFUvlgBT/elkHmhSJAMhJkbkkJYTrCvuXIq2xYFBYW0ww033KD/DOPU+vXrU5UqVVgd0qlTJ8/3mzdvXj4UFNICKLb/W7ePapQsRKVi86fKY8zbfID/r1qiIG3Ye5RW7ExOlcc5euosrdwVIECA6euTFAmioKDgGOfPX6A9hwNqDiuy40AERUaeXDmoXNH8VKFYgOSQAyRH2aIFKH+enI7OA2QIVBg4SlN+T+P6iTPngiRJCgLFgkg5of395BlKPnGGVSjgdPYeOUVjl+3mAyhbND9dXDWeWuOoUoxJEoWsDUWCZCAYixywnIlHTlGZIqlTvCkoKGRuxMfHszIjMTEx5HL8XrJkScvb4HI31wcqV67Mj7VhwwYmQXBds/Hq2bNnOTEm3P0oKKQlfpq3nZ4evZyaVChKv97TOlUeY92eI/z/1Y3K0Nt/r6W1iYHf/cbyHckhffVzNx+gE6fPOV54KCgoZA+cPnue5mzaT+v3Hg36cxw4TjsOnKDTEQIXoNgQ9UYo0RFDJQrlpRyQT6QzQKIUyANPkFxUMjafJxLl1NnzTIxsTjpGszbu5w3oJdsP0Y6DJ2jE/O18ADVLFqLWVeLp4mrFqHmlYkzcKGQtqHc0A2H/0aBpILDz4AlFgigoKFgiT5481KRJE5o0aRInvIgxKn6///77LW/TqlUr/vvDDz+sX4aEF1xuB5id7t+/n0qVKqXfB4xT4UeCxwcmT57Mjw2jVgWF9Aak0yBAgIVbD7Kc2++5FDuKu5JP8s9X1C/NJAiKaKg2/C6Wl+04xP93q1uSi/XdySdpwdYDdEm14r4+joKCQuZUeSzcdpDGLN5J45bvZqNSK+TKcRGVKZo/RMUhLSzl4vKzX0ZWB0iUfLlz8lGicD5qUbkYtzBi3J6/+QATIjM37qfVuw/Tmj1H+Phm5mZ+7RqUK0JtqhRjpUij8kUoby5FQmd2KBIkAyHpaKgcbcfB49yvpqCgoGAFeGz069ePTUqbN29OH3zwAae/9O/fn//et29fKlOmDJuSAg899BC1a9eO3n33XerRoweNGDGCFixYQIMHD+a/Hz16lM1Le/XqxaoOeII8+eSTVLVqVTZQBWrVqsW+IXfeeScnypw5c4ZJF7TR2CXDKCikJSavCVUq/bNyD/VvU8nXx1ivqT5KxebjXnLslEJejcsblS/q62Mt2xFos6lftgidv3CBSZCNe48qEkRBIRtjXeIRJj5+X7KLiV4BzJmbVSwa4s+B/zFWqRZ7a4C47lCzBB9A0tFTNHvjfpq1MYlmbtjPihoQ6jg+mryB8ufOSc3gJ1KlGHuK1C5VOEMoZRTcQZEgGQj40gH5cuegk2fOsxJEwRqQAuN1Uu7OCtkZiLzdt28fPf/882xKiqhbxNWK+em2bds4tUXQunVrGj58OD377LNseFqtWjVOhqlbty7/He01y5Yto2+//ZbVHiA1unTpQi+//HKIn8ePP/7IxAfaY3D/IE0++uijdHgFFBRSYuO+YyG/T1yV6DsJgh1CAJ4j8j9IkHWpQIJgVxKoVyaWDh4PbJZA4q6goJC9sCf5JP2xdCeNWbyLVmnjgiziu9YpST0bleYWDvgKpifgaYjkF3M8rrlkT3GWFqd9kenCSPdhXhfgtYGniRvAD+SKBqX5AOCbAkJkxob9NHtjEm9aT1u3jw+gSIHc1KpygBDBUbFYAbU+yQRQJEgGgkRFNShbhHt+jcyuQhATVuyhh0YsphualaMXrwos3hQUsitARti1v8DM1IzrrruODyvkz5+f/v7774iPiSQYkCkKChkRW5ICJMjl9Uux6R364/0GyA6gRkKABKmeUIgNS9cl+vtY6GGXWoB3dPcHEgxgbKigoJD1gda7CSt2M/ExZ/N+nVhAi0b7GsWpZ6My1LlWArd4pDWOnTpLm/Ydo437jrI5NP7HsSXpeEQPkrQCCKEKcQWoSomCVK1EQTayxlGleEGKcdi6CK+U3nHlqXez8jwmw/8JCpFZG5LYgwUtSH+t2MMHUDo2H7fNsNFqlWLceqOQ8aBIkAyoBKlTOpZJkMTDgX5jhSA27TtKd/+wkH/+dvZWRYIoKCgoKIQAhndAu+rFmQTZd+SU70ai2BkEYBoIIDkB2OXz5gWKaxj5ASUK52UiBFBKEAWFrItTZ8/RlDV7mfiYvHYvG54K0OpyVcMy1KNeKSoakyfVzwWL/n1HT9HGvcdoA0gOITv2HtV9kawAMYpRkWJWhVyweJwUj53iOu7PH0ETm5KO8QFVoBHwihJSpKqBJClSwP51hcKjZsnCfNx+cSU6c+48+zaBFIGnyKJtB/l1+WXhDj4A3G8bjRBpWaUYFc4G/iuZAYoEySDAl0jMjCrFB4qcY6fOpfNZZTzAsMiIk2fQFqPMiRQUFBQUAoW0kCANyxWhwvlycWzi9oPHWa3hF+DLAZQqEtjhk6QCxFD6iV3JAVIlvmAeNuKrEBcgXdCjDkNE1YeuoJA1gO8zNkB/XxIwOEWsq6B6QkEmPq5sUJpVCanVwrL94Ald0aErO/Ye5TE0XKpMFU1ZEVBYxPDPIBjSenwyEyn4FW2KeC4b9h5hVaA8L7S0QGWHA3HqRmC81cmR4gWpWkIh/hneT+Y2l9w5c1CTCnF8PNipGh0/fZbmbznIKpGZG5M4+ROPi2PYrC1MDsHfqU3VYtSmSjw1rlBUrWPSCYoEySCQjG58OZC5DcCtWCEU5l02yPBqly6cbuejoKCgkJaA9Hbgb8vpgY5V6ZrGZSkzAy0lp86cp3plY327TxS2mDtRp8KwFMeKnYe5fSQ1SJDSsQEFSIImd04MszPq1QMAKKU9Tuki+VgGj53hxCMn9csVFBQyH7BoX737CBMffyzdpY8rQMnC+eiqhqWZ/KhVqpBvHhNWLSz4f8v+Y3TmnLXUAmsTkC9mogNHWqhRnML8GuFXENQ4Lq4WH/K3g8dOs7JlfWLg+ePnDYlHWMWBeSTp6AGas+lAyG0K5ctlIEaEJClEZYsGCR/E90KFiEMeB/M2CBGoRUDSI+ULx6dTNlLeXDmoZ8MydH/HqqlGcClYQ5EgGawVJi4mL3/JgGOnFQkSiQTBoKVIEAUFheyAvUdO0g2D5/DPj49aysVxehvgecXibQf5uSDtZPJj7X0r/kQFgqIUygmkIjAJ4mP7CHb60KdvpQTBrqOfCg1ZFMn9I90BMZdb9x/nQ5EgCgqZD0h/RKoLyA+jjxDq/+51S9FVjUpTi0rFPI/v0sISIDmOOW5hQeBA5figZwb/XyKGKhaLyXJqBZA3zWLiqFnFuBQk0UYhRzSCCMfW/cdYnbN42yE+rF43Jka01w0/o10Sj9OtXik+ZB2DtplZGwPtM5gzRi7YTr8u2kG9GpdVZEgaQpEgGQT7tXhcSLDEqAdfRAVrEqRQ3lx0BANVKhjeKSgoKGRE/DR3u/7z+QtE/63bSx1rBpKAMhPQo33vj4t0r4sf526jp7rV9OW+heyQthEpJv0kQYSYQOqA9HYXL5iXd0vPnr9AScdOUYlC+XxWggTvr1zRAkyA7FAJcgoKmQaHjp/mNpffF++ieVuCCoM8OXNQx5olONmlfY0SnskGkK8Lth6k8ct3c4BAuNa8jNTCktGANRjaVXCYfVpAsgspgvYWrEGgqkGiJ9J6jIk9AFR7eF071y7BSTMw0i5dJD9d17QcHyCrJHYXSTNGMuS+DlVZyaiQelAkSAZTgiCWCYUVYOwHVAhg16HAoN62enGeTMDSKigoKGQHwHBNimY478M0LzOSIGv3HAmRff+8YDs9eml11zGGVtivzaXo3QagBDEamfqB3YdC1Rmi0MD8jV29xGT/SBCzEkQMUgEYviooKGRcwLfu39WJPFaDtJZ2E7RptKgUx20Q3eqWotgCuaMiPsYt28XJJBh/MlsLS2YBlIViiGrnpbJ+75GA+kYjSo6dPsdJMjjQ+gKDVJAhSC6rXLwgt+80rRhH393WnMmQDyetV2RIGkKRIBlMCVKsYB6dBMEuGb5cKK4UAgONMNtwWQYJAnmagoKCQlYHdoyW7ghIcO9pX4WLJYlpTQ1gbIVZXJ/m5dn4zU8s1MgcOOVj5wyeWHgudcvE+hY1j7lUVBMAjFH9Nis1qjOEqMAiBPNUPfLH52S3xWMJwYL2KAUFhYyndJu1MYmJj79X7gnx96tVqjD1bFiarmxY2nMrG+5/wZYDrPgwEx9op7m0dgInx6BOzmotLBkRWKNVio/hA6+9cc4GiQ2SauzSXTR17T5Wj7w3cR0fdUoX1gkReEE2qVDUkgz5hcmQMnR/h2qKDPEZigTJIDhwPFC4FS0QbIeRhJjYAooEARKPnOLBP3fOi6hmqYDB3cFjgb5sBQUFhawMtD8gQQxqiSsalOIiCZF/GBP99gUBIXHlJzNY4gv0bVXR1/tftDVAgjSvFMftI/M2H+AdND9IEFFVFiuYN8Sw1E/VhChBxBRVEHisZF8TYqQdRp4HUFxTuRgXPwoKCumf8jh42ib6dtaWkO8mWkxgcNqzURnP5sxCfIzTiI99FsQHFtMgPqBYUEh/QOWB1pcrcTQoTYdPnqF/VibSn0t30YwNgdQYHG/8tYYaly9Cl9cvTT3ql7IkQ35esIN+XbRTkSE+Q5EgGQRisgYSBEWuyJ2Pnj7rWSaXVf1AwJ6j/9qYqqOgoKCQlSEqEOweVYovGJgjzp7ncdFvE7VXx63WCZCxy3b7ToKguAMaly/KxTxIEEiHfVVVanJvUYQcPH7GN2XlnsMnQkxRBQlam4qfCTFC6hjba6TVR7XDKChkDKzYmUxP/LKMVmueEEUK5Kbu9UrR1Y3KUJPyRT35bID4mG9QfJiJjy61S1KP+iUV8ZFJAP+oa5uU5QNrF/i2gBCZs3k/Ldp2iI+Xx63iNikoRNAmpciQ1IUiQTIIko8HSJDY/IG3pGC+XPwlOap8QSxIkHx6P+OJM+foxOlzlD+PmgAUFBSyLsRwrX6ZWFZ+QHqLPmP4IvlJgkDCK94jAHYf4bMhygo/5joxKW1Yvgi78ANw4vcD+48F/bVkYwHrDxjJYk4tYVBUeMXew6dSqDMk0tLPNhXsLB/WaoA4Qw+/IkEUFDKO58dHk9bTl9M2MWkB8uOZ7rU4ucuLx5EQH+OW7aYJK0OJj8Ks+CipKz788FBSSB9gPL+xRXk+9h4+yQofECIgQhDLi+P531fy+3xF/VL0cZ9GPFd++O96blNVZIg/UCRIBsGhE4HdqyIFAoVOTN6cdOAYhfQSZnfIDh+kwDF5cupqmYPHT1P+PComUEFBIeti2/4AcVAxPpB6gthCkCAwYOtQo4SvbTcw5UaBXTk+htbsOUJT1u7j3Ss/sEXzccJCHjtjMOwD/DK5NvprASCMED0PRQUiI/0gQZI0BaIQLQI8jp8KRbQ/iYlibP7cKdthfGy7UVBQcAfs0D/5y1KOoAXgwzHoyjr699MN8QE1nCg+RP0lxEeXOiV1j4+sSnyAfIeJKFTxh7UDPyNCPW/unJQvV07e7EQUrf5zrpyUN3cOypsrB7eeZFZgTurfphIfiE4GAfbnsl0c7Q71B45nRq+gdjWKU68mZemutpW57cpIhlzTqAxH6yKSV8E5FAmSQSDFjrS+xORRMbl2LUMoBjHgFY3JTYmHT3HBib47N3h9/GoeQEbe1Uq1GykoKGR4iHpC0k4qxwfIA/iC+IllO5P5/9qlClODsrFMgvjVqgJs1Z5HRa1Yq1aikE6+oL0nmiIfhXSQBAkuRBA9j4VFkva3aJGk7c4K0SIQtYZfJAgIfqBI/twhvi9C5GDRgBrB6COmoKCQujh++iy98/c6GjprM124ECBDX+lZhy6rW8o18TFu+S6asCIxBfHRtU5J6g7FR5XMQ3wgQvbwibMBIuNkkMiAmk2IDVweIDpCr4fr4DXxAvAfIERAkOQHYaIfOfh/uSxvyN+DP4Nkya8dMLeuUKwAt1OmB7ECg9QB7arwgTheGKr+sXQXG6pOXJXIB869U60EGtCuMi3dHlCOjFq4g35brMgQt1AzZwYjQVDsAJIQo5QgKUkQyA1F5gwSRApFN/JFSBeB35fu9L3fXUFBQcFPYHEvShAhQdAOA/idkLVc8x6pXzZWV51s8ZFo2ardl8h34aOB+Q5zHYgeUYZ4wZFTZ1kdaPQECSo2jujxudGAiRZpudGUH2YSBP4jfkDIFMx1RuD1KpAnJx0/fY7l8n6TIDsPnaCCeXKpDQIFBROQ+vLUr8t1UvqaxmXo+ctr6yrucMAif+7m/az4MBMf2NzrUjshQxIfICoQaw6/k037jtGh46dDyA0hM8RHKhog+ACvRWEc+QLkL2r2wHFe/xmt8MKZgIji1vgz5+gg+TP2YozFXAtCBHMVSPsKcYGf4Uvotxm5FTDHP9CpGh94/dEuA4UINgygFpHzxPVAmODzpcgQd1AkSIZb4OfRPUEARYJYvEb580S167Zgy8EUzvsKCgoKGZkkxwIfEP8P8aMQfwq/ALd6AEktYkAtLSz+KkECzwO7bfB5wk5X4uGTUZEgogJBYWiMhoQSBDAuOrwCxIMU+ymVILn9VYIICWIgdASQ3KMYRgqFkFXRAsaxfYbMoflbDrK/ydQn2quITQUFUKgnz9Drf62h4XO38e+lY/PRq9fUc9SKCI+l3xbtYCNMoxoNi/2udRLYQLV1BiA+sIgGqb5GIzxW7w78D1LUKSCeKJQ3F5MYTGbkA6GRS/9ZJzisLsuXm1UOThQYIKPPnLtAJ8+eo5PamIyf4RHIRMnZ8/wz1ClMmuDyswESBWTJqTOBv/Pt+bLzdPzUWfYe3H34JK+94MMlXlxGoBW/bFx+JkVAMoAoCRwxVLZo/lQxqa1RshDVKFmDHutSndtkQIZAJbIrOXCuRigyxDkUCZIBAPMz+RCLEkR2dlQ7TBBgnwHpjZbC0G3BiWgqAQZ7BQUFhYwMIQ6gmpBFqSSR+B2TKjuc8AMRohmLbRSdfsiDRblS3lCUgdAREiQaiNLDaCJq9O7wox1GiBYU61BjGCGKDRD2mNdzR5lEc+C4tRJEPFUCJIh/RD5M+UCAAIj5nbp2H11Wt6Rv96+QPvj000/p7bffpj179lCDBg3o448/pubNm1ted9iwYdS/f/+Qy/LmzUsnT2bfDaMpa/fS078tp93aptlNLcrTU91qUqF84ZVSm/Yd5fjTf1YlpiA+etQvTa2rFIt6jPAKjFFrdh/mGnjNHiz2j9C6PUeYILACSJ9apQpTtYRCTCoLuVHYRGSAAPGShOMWmIvy5MKRgx/fT4AU2XHwBM9VGGMxJ2IjAGrM7QePs9oQihgcRPtM5xWIToeKpGJ8ASofF0qSiMo/muddr2wsH09dVpMJNihExi0P9ZIxkyFIKbq/Q1XfCPOsglQlQV5//XX67bffaM2aNZQ/f35q3bo1vfnmm1SjRo3UfNhMB8jJBBhEAEhRAUWCBHFIPEE0iW6cVhjKbpkbOaNA4swUFBQUMosfiDEyFcUsijY/duxRNIk6rkzR/FQsJi/LflEYo+3CD1NRFJVGJQhQQqJlo1S1CMlhVmiIP4h4eUQDmKsKsWImhaDkxEWQZ0O949Yg0QyZ20RhYkXs+BkT/9+6vSG/Q7avSJDMjZEjR9Kjjz5KX3zxBbVo0YI++OAD6tq1K61du5ZKlLBWMRQuXJj/LsjMppPRbry9NHYV/bZopz7+vtmrPrWqUizs7fCdRGLMD3O20tnzFzidqmejMpwYk9bEB8Z0LODXaKoOEB5QeNipO0Du1kgoxIRHzZLyf+Fs1RqHuRSKRCtVIl5PqEWMxAjmNP75wHFWCuK1xTF70/4Ut0ebJgiRysULUvNKcWx2W8alp6EAZFPTinF8PH9FHZq7aT8rRGCuKxYLcs6/LNxBoxUZkrYkyH///Uf33XcfNWvWjM6ePUtPP/00denShVatWkUxMeoNMC/uYYYkfWbSDiMSaIVQY9QQJYhLTxAwvAIw+5jonPRzKigoKKQHtuskSHDexC4cXPFPnT3PBIUfMblQYqBoz5XjIiZZMB+hQAsUfMejJkFgKCjKlQqG5yKtPdEqQYQQMPqBhLTD+EAYiNrEKjIYrxfUnPAEgVdV1CSIVshatcN4VUKGA8zCgVtaVqDv52ylf1cnRm1Wq5C+eO+99+jOO+/U1R0gQ8aNG0fffPMNPfXUU5a3AelRsqRz8uvUqVN8CA4fzvybSxNW7KZnx6zk3XVwQLe1qcStCAW0DUorgIweNmsLfTp5g167d6xZggZ2q8kKitQG4sdX7znMCg8QHSA8kCBm59WBsb1WqUJMcjDZUaoQe1+khd9FZgVeG8y1OEBgGAG1JIh4UZBAwWlUk2Cs3q8dUN2BmABAiqAdCgQZCDZz6pjT82pdNZ4PJBQNnbmFPp60ns2z7ciQp7vXSqGazG5IVRJkwoQJKWR2YJ4XLlxIbdu2zRYDqStTVMNCXLXDWA/wRmPUOO3/g8ecGyFhkBIyRSJ2YTjUonJ4Zl9BQUEhvSC7dmWK5AtZqEBBsf3ACSYP/CBB5HHgkC+FMAo0JkGSjvHOVTQQOTkkwcadxQSJfI2ytQPmfECs5hsliC/knxIEBSzfp03xCHIC5IUf5ISuBLEg6b0qIcM9FnrNgQc6VqUxS3ZyVDIM99CPrpD5cPr0aa63Bw4cqF+WI0cO6ty5M82ePdv2dkePHqUKFSrQ+fPnqXHjxvTaa69RnTp1wqq+X3zxRcoKAKE86I+VNG55wHgSagCoP5pUKGp7m/PnL/AO/FsT1upjKNK1nulRK8VC2U/AwwdqA5hkIkYV/hC26o6Sham2gfDAd9oYu60QPTAng/jGAXWG1fwkypGVu5L5vVu2IzlAmOzfRj/NC/jNQIEDMgTESIvKca7bfeBJcne7KuwH8tbfa3WyxUyGzNt8gL65tVlUPlyZHWnqCZKcHIjei4uLy/IDqRskn9Bi8AxFYcG8AWnzsVPW/XnZDUbyQoxRveyEwXtFYrjqlClMi7cd4t5nBQUFhYyKRK24TYgNVWIkFMrHJIhfviA7NZWcUZ6LyD5gV7Jzc7yIz0Nrf0mpBDnlS2spVDJGiDJEUl38UYJYkyB4LPSK+0GC6J4gFoSLbgzuUxINPFkAGPtB8VOtREHerVyXeESRIJkUSUlJdO7cOUpISAi5HL+jTd0KaFeHSqR+/fpcs7/zzjvcyr5y5UoqW7as5W1AsqDlxriBWa5cOcpsNebvS3bRoD9X8sYkSOB72lWhBzpVDWt0iYXkq+NW0dIdgfUNDIUf71qDF6Cp4Y0h0bpjlyFad49Oygrw/Q0QHcGWlgpppO4AGYR2Qaitdxw8zoQQfsa8slubP9AKhAPqMmxEIg0GP4deFviZ/zf9PXBZ8Ge5PS7DRnLl4jEZ1swZZAYMx3H0qF9KJ0bmbz5Aszbup5kbkjSPlsABNQfetnpli7BKBKlBIOPym7yo7IBx/J3rGrCyD59rrHeM2HbgOF392Uz6/KYmdHG11CPrMjLSjAQBo/zwww9TmzZtqG7dull2IPWjzcOoBFHpMAFA0gWZdkg7jOyEuWiH0VUguXJQuaIFeFAA86+goKCQUZGoKSRQYBsR9NI46a/ipGj+EANOwA+iRQhnKE2MkDabaJ+HKEHMO2dGw9JoEfQdsZYsy2P5qQSxMkYNpqP5M39tTjoaEr1cPaEQkyDrE4+k2qJz6rp9nAyERZtCxkCrVq34EIAAqVWrFn355Zf08ssvW94Gxqk4Mivgg/TM6OU0ac1eXcXx1rX1ebEazvT0zQlr6O+VAdPTmDw56Z72Vej2iys7XqS6IRdggDl22W726TGOxRgH4NvTrW5JalCuiO8moSk8ow6fZFKDSQ7+P+B/gd93HTqpR5RnFEDl0LBcESaECuULRIvjyJ/b8DP/H/gdLaZp6YGD96tTrQQ+hGSfs+kAzdyYRLM37mcl3tLth/j4fOpGJnsaVyiit8/gPY/kMYPr/Hp3a1b3wajX+Pk5cvIs9Rs6j17pWZf6NC9P2Q1pRoLAG2TFihU0Y8YM2+tk9oE02nYYIwkiDsKqHSY0GQbkBaR9RhNZfIm9EE5+FvcKCgoKqYU9yadCFBNmc1S/xjDxSyprUIKIr4UfUbxCgpifhzHpJpoUmuQTZ0PmBoG03nCMYpQmsuLAb/YdMZMTB31UglgZowaVkP4oQQJJB0RVigek0SKRFoWI3/hh7jZ6bswK/hmLx/9dVjNVHic7Iz4+nnLmzEmJicF0EgC/O/X8yJ07NzVq1Ig2bNhAWQ0Ya0bO306vjlvNHh5YYD7YqSoNaFfFdmFpZXp6Q/Py9HDnavp47Ne5QV2CGFS05kgrofgHgvi4vH5pbpvwy2gViVYghLZbEBz4GX+TzUg74PUoFZufiXTMI1Cm4OfSRfJTzosuYpIEPkOIt8XjIXp456GTATNR7XH8rMk37D3Kh1Pg/PPnBjESSpKAiK5TujDVL1uEGpSN9cUk3Aog16ESEaUITFihEkGgw6wN+3kOBUmC472JxOfHBqtV4vmzAALPSoGEy65pXJa61ilJn07ZQF9N36wTViC3Bv62nIm9p7rVylaeMGlCgtx///00duxYmjZtmq2cLjvDigSRIg0Fm0IoeSEFshBFGERd+4qABJHCW7XDKCgoZFCgUJQ2jpQKirRTgkgqSjSQ5BmzokWIFhTHGOu9GlXr7TCasbgAsY0o7FDs4f6jIUFEeWjXDiPkhFmmHl1tkPqeIJuSjqVQggBoh/EbUF++9VewHWPItE10a+uKKcgxheiQJ08eatKkCU2aNIl69uypq7LxO+pyJ0A7zfLly6l79+6U1cymn/ptGc3cEEjwgFrg7Wvr2xqYohb/dtYW+mTKBn3jrUON4mwu6ZfpKYiPlbsOs+Jj3PJd3OooQL3bpXYCXd6gFF1ctbgvZsUnTp/jxfXkNXtp+vokJjsicBzcfgJCo4x2oF2SCQ+QHUXy8xxlJGUw5sJ3b+G2g/yao94GycHH4ZN02MUmZloAzx/Kc6OhqACvkwBzGGJqQYiAGKlfNjZVAhbwWl/bpCwf+HxAGQJSBCoRvHfwn0KUOQ6xVWhVGX4ixahr3ZIpiDl0GTx5WU3q3awck3/G+OYh0zfTxn3H6OM+jfRuhKyOVH2WeMMeeOABGj16NE2dOpUqVaqUmg+XaWHlaB8kQTKWtMyuSEeaQGpKyIzkhUAKXbQMOd09NJIpxX0s7hUUFBRSAwF1RKD4NBtkSoHjV0vfHq1vGzt5+mNoC9N9PhAtOgliInPQc1+0QCBVBTtdnkkQaYcxKUEwN2DMx1wLYiGaxbYYcdudo05OuEwts5LASzusmdQBimrqEKhFolHPCLALCKCnHqiWEFCCIBXo1NlzYX0R3OKPpbt45x07q6h1Fm49SD/O3UaPXlrdt8dQCAAt5v369aOmTZtS8+bNOSL32LFjelpM3759qUyZMuzJB7z00kvUsmVLqlq1Kh06dIjefvtt2rp1K91xxx2UFYDv1bezt7CJKaK/oSx+vEsN6t+mkuUOOL5bfy7bTW9NWKMr5dBa8Uz3Wr74KOD+keAydimIj928yBVglx9tEpfXL0Xtqhf3xesCRMeUNXu59QcLaaSLGYF2EJAaQnCUNRAc+B11czilABLA4HExf8tBmrQmkY0/Mzsw/jYsX5TXaDA0hbIE89SeVSdpooFEQIQyyJAGGimCdio/yQSM8YjWxXFzSxgXX2DvEFaJbNzPEbmY3xCRiwPxzlAL9WtdkUk+I+ATM7hvU5qxPoleGruS1iUe1YmeOi/8TbMHdgypA7IqcqV2C8zw4cPp999/p0KFCtGePXv48tjYWMqfP+u/uG5JEGNUEeRYmUEJgjaVbh9O58Lpxztaptrj6KaoRvNYrTAEc4tsbieDTWg7TD7fZN4KCgoKqQFReWC8MstcRY3gV0yq+F2IwgQwksXRLrbluZiVIDL/RZuqElSCpGwfKaKTING9VnJ7K58Oo6JTzsUrjp85x+QXUMji+Ui9APWM0/kvXMoETPIAFNjyHmERhvtGr78oRPzA9PWBXcueDcswIQYSZMzinYoESQX07t2b9u3bR88//zzX4A0bNuTkRjFL3bZtGyfGCA4ePMiRurhu0aJFWUkya9Ysql27NmV2bNx3lP73yzJasPUg/96iUhwnv1S0+WzP33KAXhm3mv0YjKaniBeNtmUAC2mYm0L1YWzXAAmBWF0sXvF/tP4i+G7j+zV57V4mP2SxKwC50aFmcX6semWKcJy4mzEeao6/VyXSJ5PXR2VsDWULVIeYb4L/59N/R2zsoRNnaO2ew0z6Q40DIvXoybNMFsv/IMIDG6PkG6BWQfqOAMRZ3dKx/DhYU2Csh3IEYygOvKcAXsaqxQsGWmjKxVK9MrFMoPll3Ip6oHbpwnzccUll3pAG6TR7YxJNXL2XP7eIw8UBEqR/m4rUrW6pEBXRxdXiafyDlzAJ/cIfK/XLW70+mb64uTFdVjfQlpNVkaokyOeff87/t2/fPuTyoUOH0q233pqaD52pIFLnohYkCJjqjAx8udCriAP+JaklocLgZ24ZwmskEuejDh/beD/KE0RBQSGjwy5Rxe+WCBRQQkCg4BQU135GDzd2maySSqI1RpVFPaS40gLiBSKtNqfDGH1BZA7wChA1AJQrVpDHjlbmLW2emOPEB8sIzH9YMGEnF+9bNHMv2qDw/uL+SmkEFRZCkGJjgYa+dL9IEGzszNkUaEG4pHo8L8KwnsTiASSZaonxH2h9sWt/gUrbiPfff5+PrAaQbE/+uoxJQ7SWPNWtJt3YvLylfwLUGG/+tYYmrAxs3IIMRFIMFprRkBJb9x/jBfKfS3fxDr4AXiRtqxenKxqUYuWHtHp7BcaD/9btpclr9tF/a/eGjEUYT5qUL0odapZg4qN6QkHHpAeUBxv2HaVJq/eyr4TT4IYqxWN4zDeSGhIlG/g5H6stcB4YH/D6w6MIpBXGCvyP30HIZgRAoS9EWjiAiIGnEo5fFwViaqHoRNoWt9CUCbTS4Hc/fDjQhoT0GBz3d6zGJAhauPCZW7L9ED00Ygm9Umg13dSiPN3Yory+EZwrZw5Wi1zRoDQ9+vMSvbXm7h8WcbrQb/e2ZuPYrIhUb4dRcC6vLRYTLDzz58mRKUiQBVuCAwEGKfTIpQbA8pp3+DBgYrIAE4uC0UnxpCtBCgTbYXCZ33JfBYW0wqeffsqSZezcNWjQgD7++GOWPdth1KhR9Nxzz9GWLVuoWrVq9Oabb+r93mfOnKFnn32Wxo8fT5s2bWLVXufOnemNN96g0qVL6/dRsWJFlkgbATn1U089lYrPNHsiEnFgNNCMBkKAoBgzqhywayStKiCMvZIg2JGUth0rJUgRH9pIIilBjK2VXoACXeZku3YYeexolSAy5yHRwGqRgsvw/mMDAu9dubhAlLEXiOkiSA/johDJLSBBxCvGD2BXGosILH5qJBTi54F0mFW7D/POO3bAFRT8BNo+Hh+1lE090Vby2jX1QmLABSCTP5q8nr6fHTQ97d2sPD1yaXSmp/DEeHnsKpqxIUm/DC3k2IXH5/3S2gkhG3xe1lqrdx+hKWv30qTVibzgNXp7QEHdvnpxJj7w/J22G2K8w0L6v3X76LOpGyNeHyqSu9tVoaYV47i+xu9WdTXOF6pDtODN23yQSQ450HJkt3TEa1a+WAGKj8lLuXNdFBKna4zNDY3cTRmxm9cQtZtbv+1FdOzUOZ5vsfHA/6PlBUfyKd0QO1qAbF6x8zAfww3jLNpbbmhWzjZ1zAuQCvNe74Y0sHst+mneNjbzxRz+wb/rmcTqUa8U3dqmkt4qg/lkWP/mPA5f98VsvgxkXe3n/6YPb2hIVzYonabJOWmBrEntZDLst2iHEbkUjIsyKrBzaJSIYQBLLRIEsjdjC4wgSII4Y6SN7TA4MChiZwDFOfodFRQyE0aOHMk931988QW1aNGC+727du1Ka9eupRIlSqS4PmTNffr0YcLi8ssv53ZFGOYtWrSIo8uPHz/OP4MkAaECafRDDz1EV155JS1YsCDkvtA7Dtm0AC2PCv5DJMZWRbgQElhUYq6IZpdSCArMQ+ZdKRS0ARLkJO9aeZ3nUJjjvq0KvWhVLSjYpb/d7AkCFPEhJldUKngOVj4dxscWfxKvkN3bcLvCIKuYBImSBLMzrJWF4u5D/pmHL90RaC1oUbmYXlAj3QAkCDZVFAmi4Ce2JB2je35cyKQGdro/uqFhioUcNsGwY/7x5FDTUywexSDYCzDWvD9xHX0/ZysrljGsItoUHh9I6YhGVQfvDZi6wsNh6tq9IekxAHbwO9UKqD0alivqSGkApcfsTfvprxW76Yc52yJev0KxAnRf+6r8OFZjOtYITG7sBcERUHbI7+GUchhbq5QoyElVgSOGf4fnhl9JOG6B54I5EqQIiJrfFu3QFRPRAB9FvHdv/72WPvx3PRvfwiQaChG/gPn7wU7VmKCCugmfdW5BXLKLD5Al/VtXpO71Aq0yzSrG0YZXu1GrNybrdQFUJENnbuEo3XDR0ZkNigRJZ2DQsXKbFxIERR2uYyXZS2/ANV7ICcBNDJXXXTFzQYhdMsAxCWJI4sFECKk3drnAjioSRCGz4b333mMiQkzuQIaMGzeOvvnmG0tVxocffkiXXXYZPfHEE/z7yy+/TBMnTqRPPvmEbwvlB343An+DsgS94+XLlw8hPZzGLAKnTp3iQ3D48GFPzzm7QYoQo0+HICZPTpbXYncJ80j+PPmjfhxpfzECBAx6yaPxT5KdNCuSBSiiGX1Ku4lbyByAohJpMGbITuuhE94JA5mroSqx2xGTx4mGbAFEam7lByKQmiHadig7tRGUIQDaYfwCdsVlkSZoWrEoDZu1hXcgFRT8Ar6Dt387n8nLBlr6i/l7i+vc+e0Cmqd99vC5fLZH7ahMT1Gz/7xgO73191pdYXdZnZL0TI9aUSm2sBAfvWgnG6iCrMAGngAtcxdXjWe1R4caJfTvrtNa/pnRy9nMNBwQv3r7xZXokmrxthGxu5NPcLsMFCmzLIxXBXgbYLgaJDqCZAcMSDOa4gDkC15THI3LF2VVhBBo09clsb8LTFKtUmXCwah6QWTtb4t28gF1BsiQbvVK+qZSB8GB88axbMchHnNhyAu1z8Mjl9Cr41dzixjaZfD+zn+mM33530Z6XUvygsLo8o9nsGJlYLdaeotpZoYiQdIZ2C0CQwwYJcjiCQJgEInWHCk1ICZ6aUKCnDpjqQQREsRpb6LZYBUMKZMgyhxVIZPh9OnTtHDhQho4cKB+GQzu0L4ye3ZAymgGLodyxAgoR8aMGWP7OMnJyVyQFCkSujOBFhmQKCBGbrzxRnrkkUcoVy77KQXqkxdffNHFM1QwkgdGnw4B3hfMGyBxUWy7KXzNkJQsaRM0wo8krf1HU6agGSHzn9cFvSgvQJRbbRrImB+N54hOgoQp/qQdBuqcaNosxRPEitAxq1uiNcYVJYi5pVQnQbTUID+wZndKEqR+mcDYsj7xKLdNoUddQSEa4HN0//BFrEBAu8GQW5qkMKREu0O/b+ax5B/fs+euqE29GpeNyp9h0baDNOiPlXoqStUSBemFK2rTJdWKe75PrBHgI/LexHW6gTEAEgFKDxwtKxdzZbiJlJO7f1gYEsNrRvOKcXR9s3LUqkoxy/YhIXyW70xm0uPf1XtZ0WUE1jJVSsSEkh0lYqhisRjfDELTExjfO9dO4ANqRKhDQIiACDLaGQTSdfIz6SGEWzgs0YiJQX/mpr4tK9BNLSv46pcEpcl71zfkiOef5m6jH+ZuZdXph5PW02dTN7AqBCTMgHZVWA3Vf9h8/bYj5m/n8/vhjhaWdUlmgiJBMkgrDBbzRsde4+CAL1JGJEH2mwpiyNxSC0JypFSC5A5RirhphwHEHFXF5CpkNiQlJdG5c+d0l38Bfl+zJsDcmwHfEKvrS3KXGSdPnqT//e9/3EJTuHBh/fIHH3yQGjduTHFxcdxiAyJm9+7drEyxA65jJGCgBClXrpzj55vdSRArhYYoK0CCRBvJKkoQq6JGWjWjURyIAbhR8RjyGFF6gogHh11vvXiCRGOMKgSKXTJM0MMjsMMHdUregjmj9gSxgzynaP1HhATBYtGI0kUCv/vlCYKda6kTapYKjidYHGAnG8QRFnmSUKOg4BWIB52+PokX4UP6Nk2hXIAfRd9v5nFrA0je725rzskdXoFWwTf/WqsbYIJUeahzNTac9NrCAe8MEAvv/L2WY3QB+GwgzrdL7QQmWNwoJkBYDJ21hf1JrADy5/nLa1P7GsW59cTuvtGKg2jVyVrUrjGiHTeBUgItMp1qJlC1EgV1UhokAcYaKMuW70jmNhC0KcG3I7fJ3wOeH2bfjqCvR6gfiFwvJm/OdCVQsW67rG5JPvAaTVkTIETwOmEMlXEU7xveP7zGaEsZtTDwmbGbcz6avIEPtB8hzQipRn4pZeIL5qUH0CrTvgpNWLGH1SE4p9+X7OIDCqpbW1egP++/mEkzeQ4gDnt/OZuG39kyU5tZKxIkA8bjymAkfhUZ1RxVdvbgRIwvzdYDx6OOUIwkdTaTIPK70/5rkULrJIgmMd+nyYEVFBRIN0m9/vrr+TstSV8CI5lRv359ypMnDw0YMIDVHnnzWi/Wcbnd3xS8KUGMC/Jo1QB6O4yFEkQ3YI2GBNGVINbPQ9QVBzwqNZLDmKKGeIL4ogSxJ0FQ8GNewpwFcsLrTtkRJySIT4k30g6TQgkSG2yH8WNu35R0lBc9eE6lDYQLXjMsDGAWiCQFRYIoRIPvZm+h72YHjLs/uKFhCg8DLMBvHTqPNyErFitA39/ewnObClpU4LEAs0nZrLu2SVl68rIaUZmpztqQxO002HEH8J2BpwN2590mQW0/cJwG/rY8xJjVqFB469r61LpK0KPHbZsLxju0yDQqX4RVAyAzsVhGSxDGDhw7D530zVzUDiBCKheP4bGkWolCnHxTLaEgVSgWk+ZeIkhT6VG/FB/4XOB1Q0rLf2v3sWpelPPXNC5Dswd25HkcKqLhc7fpEbtmbN1/nG4YPId/fqBjVbqvQ1Xf1DS5c+Zgzxwc+H6ADIH6CK0yj4w8xPNYlzoJnNaDEAwAKishQqJRoaYnFAmSzpDC0EyCAGCwmQTJoOaoSdrOHphekCA4Vwx+qaFasVeCuGuHCXqLaO0wBQOTlIrJVchsiI+Pp5w5c1JiYmLI5fjdzqsDlzu5vhAgSICZPHlyiArECjBlPXv2LCfO1KhRw/NzUki5cydzRHwhGwWFDyqNEMWJBQkii22vfh1G1aOtEkR7Hoe8KkHCxOOGRuSe9kEJEr4XGkQMSIxofEGkHcbcAmrpcxLF+yJtAVaeIPI75nUQYNEmF4gfiKTCGIFFC0gQLA661onqYRSyMaav30cv/hlQOoCIgAGpEVAwDPh+AXs31C1TmNMwvBKVeCy0vmAxCDQoG0uDrqxDjcoX9Xz+ID2g/BDCAgopKD8GtK3sONUFOHbqLI1fvpue+GWZ5d9hcAnvBzviI1Kbi3l+gNrhrxXWilIj8HxAvLC/Rmx+3uwFkQQ/DHhbnT57jv/ny87K5efpzNnAZSBf+Hf9+oHrAPgfCoVA/PDukFQZRHyDEME4I/9XjC+QJqmQWLdc1bAMH9iwnbgykf5ctotbZ+D/gfdpQNsqNKBdZTbO/eTGgGoGRtEgI/5dHVqzATDxxQHCZ+itzaLymjEDARfvXt+ABnavSSPmbWNjX7TKgKCBEgkKelkzbdl/nJNkRtzV0tdzSCsoEiSdoZuiWpAgGCzQiosvg1dA9oZ4q1/vaR1VBJcVpDjHBx+DDHZ4UGBGY84Xkbwwp8O4NEZFMSe57yFKEEWCKGQyQH3RpEkTmjRpEie8AOfPn+ff77//fsvbtGrViv/+8MMP65fBCBWXmwmQ9evX05QpU6hYsWIRz2XJkiXsR2KVSKPgHVhEY1wNp6AoGhOdgiJlO0weW7WJV4LC2D5pt+AoEq0niEY42BmJSuuIRNJ7gZxbpFQHJMRgJzRcAkIkHHFgjKr7qETxvsBvQApaczsMdhlBToEAwXWiJUE2JwUWi/AFMAMLE2C9JvtXUHALEGj3/riIP9PYYb+nXZWQv6M14ZGRS3jx3KZqMfri5iZhv1/hlBWvjFtFf69M1Ot3EC7XNSnnOcQABOG7/6ylf1YF7hMtH32al6f7O1S1NSE1A2qteZsP0MgF23lxbQbqXqwF7Np+4KOChTkW3eY2l3BAS5EAC2QQHAGiI59uJlpGO0CY+K0Wx/PGewoyd/3eI2ziDX+hDXuPsLLs+Olz/D8Ooj0hinsogYQYEQUJ1CSp5VcCgrxXk7J8QGWBzxEMaeHFMWL+Nnqia026plEZfnyY84pBLzbD527eTw/+tDhkXsFn/pK3pvDPX9zcmEk/P1tl7u9YjX1B0Crz1t9rLD1kMNdd8ckM+u2e1plOxadIkAzaDmM0R/VKgmBg+HrGZv75l4U72NU5Nc4dRTMIFuz0oWgvpUlo/YTulK8pOARiGufEEwSvh7QWyQAnffZKCaKQGYG2lH79+lHTpk05wQURuceOHdPTYvr27UtlypThNhUAcbft2rWjd999l3r06EEjRozg6NvBgwfrBMi1117LMbljx45lzxHxC4H/B4gXmKvOnTuXOnTowAkx+B2mqDfffDMVLep9B0zBXp0hcd5WiDZa1qzUsPIeCbaqRN8OYzXXGS9HgefFHDOYppJ60bWihAlnjArEamqUaLw67FpAjZDziEZxgs8YFo1YEFgRVFjUYK7HgqhWKYoKYupYvljKHUMsPgAsYBQU3ALjH5Jg8L1pWqEovX5NvZDFIFpWBv25kr160KLw3vUNXKsAsBD9/L+NnJgBRQK+M31bVaCHO1f3vMm4bf9x+uDfdTR6yU4+N3AoVzcqSw93ruZ4Z33v4ZP007ztvIg2R+UCPeqVomcvr2Vbm8PAGaTJ51M3hhivWgHja4OyRdjHx0xwJMTmTRNlhRl4n/PkuohfLxwdayaEKFpg7AwCZAPIEY0kAXmAOSMQ3XuMJqwM3h8IKCgy0P5xae2EqNqawgF+Gz8PaMUKmtf/Ws0Ew+OjltKwWZs5oQhmtwIo7NvXKEHLBnVlleCr41azQakRd/+wiP9/9WqofCr4dp65tVYZeMVAZYX1pBlQI3Z89z+a+EhbqhZFrHRaQ5Eg6QwMXoDVDoss1L16ghj7t80mpn5A7hM7lDoJEuVupFsliLD4R7T0mHAw9jBCZWNUgsDUSkEhs6F37960b98+ev7555msaNiwIU2YMEE3P0WsLRQagtatW9Pw4cPp2WefpaeffpqqVavGyTB169blv+/cuZP++OMP/hn3ZQRUIe3bt2dfD5AngwYN4sjbSpUqMQliTp1RiB5i2GylzkiREBKlMaqQKHEWjxVsVYlisS3tMDYkSCC2PGAoCo8LtxJ1yL+BmDw2JIiYaJ866zl2PllrpQlnjGp8rGgIF5nzCjvxBInifZFkNHzGrFIx0B4FebkfGwXYQQdgCGgG4jGBLfuPpZq3mELWBFoiYNoIzwQszr+8pYm+GMdn6f2J69hYEgBp8cIVdVwlwOA+sFDFwlOMIVtVLsatLzUMKUdua3+0M4C4gIpBYnQf61Ld8SISY96X0zbRR5PWW/79wU7VqH/rirbKNWywosUBJrJ2wMuEuFak27StXpxbfjJTehPG+bJFC/CB6GDjewovJChGWCWSGFCNIC4YRBoU9DieHbOCjV671klglQX8RfwExjkksSDhB0TdJ5M3cFsgvD/wmIiirRgfk2Ld80av+tzSBPLqyV9DW56eGb2CDyhDLq1dMqq0I/PjvnNdAz5XeMxYke+Xvj+Nxj94CdUu7d1kOC2hSJB0xnZNRlauaMqiQLw1vHqCoJiw+tnviFz0eAf7rf0nQVCwHj0d3hjVSTuMUVEjBJMwvHgushumoJCZgNYXu/aXqVOnprjsuuuu48MKFStW5OIgHJAKM2dOwJxLIXUhY2w4QsAPTxCMsdJSIcoSu3YYrwSCTprbPBeMvSAPUFjhubglQUQJYmcaKAoRfLwxn9gZqIaDkA3SWmMHUZ1E5QliEwtvRGz+6NuUIhnv6vHIPpAg28KQIGWK5mcSDNJ1P/xHFLIHMF89N2YFzd18gOvBr/s10z87qOmwiP1p3jb+/ZHO1enBTlVdEWxYFMP3A2agABQPz/SoRd3qems7wHcVahIseKU9G6aij3epwcoAJ8DzGrVgO73wx8qQzT0Z5x7sWI36tChvqyJDMs6A7xdq7SEpASIJhEfbavHUqkq87630GQF476CMwYHnavw8IcEKrU7/rNxDS3cks+chjtfGr2E/I5ATXeqUpDqlC/tG1mJNgrYTmOq+/+86JqdwDvBa6deqIie4mN8HkFGIMMaxcOsB6v3lHL191qgMeahTIKXIToXpFiBtYIIL1crMDYHvhRHXs1lqC47hzehQJEg6I1xRkD9KJcjmpKCsLWAU5B8wUOiRh5oSJNqizw7Hz5zjwtVK6iy/OyFB5HWE1E2cokHgYAw7py0CMkLmtdoFU1BQAJLEp8PCrNRPNQAUC1I7WZnvyWPgOhhrhfT22j5pBxiOYg7xQqaLEqRg3py2RaYkruE5eCFBzBHrEZUgJ7x7gugRuaYWUKv3BS1EXkl8IUGKpTIJgk0ImOvZ1TvYuU8olI93Z1EXKRJEwQnQ8g0PDHz0P+7TSFdm4PP20IjFvJDE317u6b5FAG0vSGjBdwtjB9JZ4DPixfwf49M3MzbT4GmbdL+fxuWLsAdEqyqRfbcEUCfA18Sc1FUuLj/d064qe6FY+Vkg3QUqgx/nBgghM0AEdK5VghUf8MnIrjUonnfVEoX4QPoKkm0mrkqkf1btoTmbDnBUMQ4oi0AWdaldkkmRphXjfNlExbj3Ss961LdVRVYe4f3+asZmjl5G29WNLcpbJt00qRBHG17rzmo78QgRwG8EB9pZHru0BhufRotSsfnp+9ta0DczN9NbE9bq5rSyIXHTkLk07LbmnB6akaFIkHRe7Io8FAOYGfmi9ATZopmQyc+4H7/MfrBbIyw2K0GEBEmFdhgpBmG+mtfUFx80Ro38uKKoyWfoWcRggp1PtPJAFpyeJAjen6d/W05T1+2jkXe1zFR9dQoKCv5DT2wJMy75QUBLQY2dQyvvESxQY/Lk5EQFtN24JUGOnz7LcwYQbnGrtzd6aCPBuQHh4iPRWgJ1Dbw6sKPrOYY3Agki70k07TC6J4iDdBh+rBNnIhq2hvOCsSOngr5Z0bWM7jh4XPfxsvNUATkCEgQK2WgSNhSyB5Bc8ur41fzz091rUYeaJfTv3Z3fLmB1CGJTP+rTkC6r687QBkqN1/9awz93qZ1Az11e23P6xerdh6n/0Pl6FHXNkoXoia41uK3AKdmwZs9hXhRPXx8acwsTT+z0w/fDqk0FqobXx6/mhBczYIT87nUNqFmluDSPkM0sgOcJVBQ4oOJBTPDfK/fQtPX72BAWJAAOqCxAIqFlpk3V+KjXWoga/va25jR17V5+36HagfIH8c9QIqG1x+qzg8/oljd60LR1+6jvN/NC/gbTWxyYy4f2b07NK8VFdY5QhN5xSSDR5uGRi0P8nED03fL1XPrm1mYh3iYZDYoESUdgZwVSNpCHVhnLQWPUULmbU2w2tMBgBw9GQOa89GiLZnhrwHG6SCoqQY4aZMHmL73I/aTADgd5HfOZWHzsdDEJcuQk1ab062MDAfLb4oCj95Dpm+itaxuk27koKCikP8RM1M5Hw6jciGbslVYYSZqxe5xjp0/wdStRjKfnARIbBZgdJN7Wi4JC9wQJS4LkZhLEaZqYZyWIH8aoEYxeASxcMAdi5w3vixcSRFcb2ZBTkkwRrRJEVK8o0u0WfmXj8tO8LUHvEAWFcKQAkjKgEu7TvJxu/I86rt8385l4wHdjcN8mvEhzgzGLd/KCEwDB8Mil1T2fJ9JaxLAVm51oe7mifmnHLYXwDnlv4roUJpjw0AHxY0ekLN+RTC/8sYIWbTuU4m8vX1WHbmxRQbV/uwTmQEl2waYqiBAQIiBGsCb6ecEOPjDHXdmwNN3bvmrUsbEwQ724ajz9NH87e9vAxPW2YQu4hQpkSM2ShW2VPQue7cxpSfgMmjcM0LJya+uKfB/REmDw//jj/ovpjb/WcKSvAOuyW4fOo6/6NtNTbjIaFAmSjpCiAASI1YcwWmNUoxLE7xhY2T1CKwwG4NRshwnnkl8gj3MSRF5HIZcEeP3RLiT+LOkB7HyOXRbMNf9j6S56pnttT7JzBQWFrAF9nA2jnhACGgvhM+fOeypoDmixsVZ+IAIQJDAF9OI/EZwv0H5oX3hL64cnJYjeDmNf1gih4IWcgFJP+u8jjcvSDhOVJ4j2GkRq24GqAu+9Vz8u43sTTgkiJr1eIdGKVqpXgXijKRJEIZJC7vZhC3gx17JyHL14ZV0eV1DzYvcbtTVIvWH9m7ne+Ju8JpEeG7WUf8YiESktXvHvqkS6b/giHjeaVSxKX/Vr5thfA+o5tM7gMNe3z3SvRXdcUinFWAp1+eyN++np0ctpy/6U36EhfZtSp5olwhIwGOfgg7Jy12FauSuZOtVM0BU2CkGgJQqqDxyYd+dvPsCECCKOkdCDtJ5RC3ZwexLaaqIxVIXC55aWFeiqhqXp08kbaOjMLawI6v7hdG6PgUrJKpUH34Hhd7Sgd/5ZR1/8tzHF34fN2kIj52+nCQ9fErXhK9arMApGK9WgP1eFbD73HzaPBt/SNEN+jhQJko7Qd0YsTFGB/HlyRGWMisxsADItsJTRSHPtCjQpKkUenBrGqEfDFLcxWv83JoxIkLYiSYYRVI6PocmaWVR6YcrafdxTh3PJlfMilpVNXpvIUWkKCgrZE7pZaRiFhrE1A4t7L14KYqoaTkkg5qhCmPhpippCQeFBqXH0lIN2mPzO08TMEOIE64eCNgk05sfx8jwAFNWiXAxH6ggJAlm211ZUx8aomp+HVyCmErBSvQpk13S71jqjoGBVx8HUE4QsFlyf39SEW/hW7EzmXWcovSoUK0Df3dbc9cIOO+b3/LCIPUCublSGnr+8tmdvDBiXPvXbcr4vEA+f3NjYkZcIrv/rwh30zj9rUyQyod6G7wnUAUbArBqL71fHr9LJRgFIF9wGygHzcwFJu0ojOwL/H6YN+47yOQhy5ciRIRevGQnYeGhdNZ4PEAH4HH0yZQMTFVCG/LpoJ/VsWIbu71iVKpmSXtwAhPjA7rXY2+aNCatp/PI99MOcbZyKhEQk2RQ2EyhPdavJCT8wMpX1lHFzuN3bUzn1pVfjMlF7wdzaphITcEZFCNKP+g+bT1/c3IQuq1uSMhIUCZKOkMHKyiQstB3GGwkiBRjMe0CC+KnSEHWGkCB+SLIjGsRZyIIL5M6lf8lgeGfVzy44aaMEqVy8IP+/2aScSUv8vWIP/48BAmQVSBBjf52CgkL2g7QdxsXYkweQNGNsxJh8yCMJIvG6YZUghoQY74qW8O0a4gnihbCPZIwauH/v7TZC8IPgiCRll93eIx7nQ5nzInmCAEUkIUaL7/XccmXz3kiMPNpzsCHjxRQS2JN8UvchsEO5ogGCxLyQU1AQpQPahpHUge8ylBUgbmdtSKK7vl/IC7zapQqzl4KQd04BEuX2YfNZtQHS4q1r63tKwQIGT9vISSJAr8Zl6Y1e9Rwp9Kav38f+D1ZBBvD+gJKjilavAqh5f1+yk3f50SYRcv34GHr9mnrUonIxft1AqIDsWLkzQHas3J1s+z3DximST9Dm0LZaMDlFITJAIuA1x4HP6ceT17MPB4xNRy/eQVdpZIjxfXSL8sUK0Gc3NWHT1Ht+WMhkS79v5tHXtzazVQ5ibVE9oSBHSVutLR4ftZQmrNhD717fIOo0oIHda3Kr0CbTZxKPDULuigalKaNAkSDpiAVbA31a1RKsvwzRtMOcOnuOB0gABnDLdiRH1Z9srwQJfFmCxqjeo/psHyuMEsRYkKFAC0eCyOuY10SCCDNr/sKmFcDiz9gQMLvqXDuBlm0P9HButIkvU1BQyG4kSOSWCCZBPKoBnClBcoeoU7x5m0RQgujGqKnnCRK4/zOp5gcSqmjx9n7I84dqMdLiSY+nTyUlCIxM4eWCxSFaalGAe8HuQwESpGSsvRIEMbl83eQTnqOYFbIuPpu6kX3TQPx+dlNjqlqiIM3ckMSmo1DSojVmcN+mrpOfoALGIhK1JswiP72psae2QpAN8EX4ctom/v2utpVpYLeaEXfXMbY8MWopqznkO4fvmyRuINXjoz6N9OcF5fOIedvpq+mbaJdGLgowPg3t34waly/KbWVY3MJcU+LWzcD6AIRHndKxgf/LFKaShfNl23QYP4FklGH9m9OS7Yfoo0nrOe529OKdTFyBCHigY1VOofGKdtWL0/e3t2AF1PwtBzmRBQoou3kcG76j721DA39bzi33Zvy7OpEavPgP/TygVVSmqWjNef3qetR78JwUf3vgp8W8NoWvSkaAIkHSCZCuoncP6FQrwfI60aTDGHeSRH7qVZrrxKcjNT1B5LkUtJjYQHrA/RuTxbHT4aMbRV5sVoLAYErc60EeWfXWpSbW7T3CrxsMZuuVidWLebh6KygoZE+gJULG03BKEBl/t9MJz0R3kGwJ5wkSfTtMuHjcaD07ROYbE6ZVRb9/D3OhtJs4IkEMEbleIs+lXUc2GcJBPGG8kCAgGoLRxdafMZw7/ob2g6RjUZAgh7V2mDBKkBKF8unKTpBtKiZXQTBhxW56+++1/POgK2pzlCvqtWdGL+f6DzGlH97QyHUqBwi3W76ex2o1kABf9WvqKdnj7Lnz3P7yy8Id/DvIjwHtqkS83db9MLqcz0oOJCCiDWdT0jFWEQAD2lWmJ7vWZOIHY8nPC7Yz0XLQ4vuOhJg3r63Pa4ZBf6ykH+du5e8SAD4R6gMj4QGlh1UkuoK/QCsKUlJgVouoWpANvy/ZxUQE3rMHO1XjNBivRMtPd7ZkL5zlO5Op9+DZ9MPtLXRDazOwSfDhDQ2pUfkirDo6a2h9EsA09cGOVfm8rBKHnABKmN5Ny3F8NVRJl9ZJoC//C5CD8Nw5e/489W5WntIbigRJJ8DvAR8+yJPsesRksX7CQzqMkBRwKZYdvHA9wyi2/1m5h7rVKxWxB9m6HSY102HCG94VyJuTTh8/H9EXxM4YFbJJcdhHb53XwcgrxLkZgxl2H0Qmh3PxanSooKCQuSGLWqyfIy28o22JkIJaWl5Sqx0mHMlinE/cKkGwABHT0oyhBAlcB4szkO9uW0j0+dXBXBzN3AulihTB4d6b+EIaCeLRXB1kS2Jy4LYlw5Ag2NQAUYZda8SJKhJEQVpVHhkZMCvt16oC3dKqoh5jC/8B1HDvXt/QNXkBAvDmr+byZxsLNbTRuFWRACAd7h++mBe3IBve6FWfrm9aLuLt5mzazy0CGOvRJvZsj9r0/r/rOMkR34U3e9XTfeGgdH52zApuqzAD5AmSYq5vVo7VIUOmbdIjw+EFAmPOBmWLeG5lU/AH9crGMsmGzzPaZP5emciBCDi61ytJD3SsRrVKuU+ohPnvzwNa0k1fzeVWF5AYP9zRgsra+E2C2O7fphJvuiI9xuw9A3w0eQNNWrOXPXe8Et9oi5m0JpFJPfho9W9TkU1dgf/9upxTm6JNz4kWanWVDsCACyYXCJddLgOWF2NUnTjIl8tg0mZfJKGn8IlfllGvz2ZxweL0/lO0w5w4w2y1nwg+lg0Jok18xzRjPDucPG1tjIoBAf2WwIItB+mnedvoxT9Xejaa80qCNKsYkJ9hMoQqBMUpiBAFBYWMC4x309bt48LVT8gOPciHSFGG0bZEODFgLRJFO0ySY2NUb54gUvAbzbL99gQRksFoRGsHbD7Ie+alJSacD5YdARbN+4L5O1wrabxGkAiZ5cVzBoQQCD1Re9ghQdvBFGN3BYXXxq/mTSws6JGEIZ/djydt4J+f7FrD0eadua5EGwEUGKi5vru9ua0aKtK40PfreUyAoG3sy1uaOiJAfp6/nW75ei6P2Q3KxtITXWtwqgvmkYTCeWnUgFY6AYLUm6s/m8kECIYVLJTxWACui5YIXN7urSn0wb/reTysXzaWfryjBf+tZeViigDJQABpgc/J+AcvYfIDgMlptw+n0+t/rebNT7dAW82oAa05fQvE4PVfzI4Y9tC0YhyNffDikNYXmA0L4B3T9u0p9NuiHZ7WdVAZyff14ykb6Mbm5alx+SL630XZlZ5QJEg6APK9xMOnWAUi2eZ+G6NK4QWSQgiKcMXYlDV7+f+1iUd0OZ+bdBiZgMCfeI309a4EcRaTqxujWkwG4raNSQj9cmArB09PGSnlNzCwzNkUIEFkIAIpI2oQvxdWCgoK/gFjCnbyIEW99P3/6L2J61KBBIm86I62HVH3BAmjBBGlwEEP7TAHHBujelOCSAth7pwXhW1n9EqyuFWCYAwvHEVrj6t2mCgIMD0aOYJCRxaHXpUgYoqK+wlHtgDwIwjcJro0GoWsAZAdUEwAr11dT5fnv/vPWvbwwGIf5qNux+07v13AXnkYX7+/vbntrnk47D18knp/OZvmbTnAqi34MVxa27q9XYDkFZA6T/66jFtVLq9firrWLcneHRhj0Drx5/0XU4NygcUiYlev+HgGm6ViTKlWohCt3n2YlW+tKhejOy+pTE/8spRjSUFSQln+6Y2N6ff72lCbqvGun5NC2gHtSDA4/fvhttwWA6Bl5IbBc3iz3C2g2AARghZ/eMVc/+UcWrPncNjblCiUj8kyRO3KegvtMkY8+vNSemjEEk/z5pUNSjN5CR+Q539fySlJArQDpWcqJ6BIkDTGom0H2SkYu0SICwpXUIliwQupoPto5M2ly/vCFcgiJQbmbwksyt20w0C5IG3P5gimaGF8Lna7boDTdhgrySQy1/EcjMD7lNqAdA2TPN5r9OgJEPEGeBkIFRQU0gZfz9jMklYAGyWQuMKB3w/oxEEEP5BofSH4sXQlSOR2mGiMUeMdG6Oe8d0UNTR9xrsSxKlzfjSES6Q5zwjp6fcST++UaBPyyqsSZJc2j4VLhhEkaNdBO4yCAkgAbK5Bui/SebQTjJi/nX9GjK0bA120zj3402KavWk/145ogfFiTgkvj2u/mM3kBMi9kQNasQ9COKA2vuu7BTRYM069t30VrkffmrCWnyPInBF3tWQ/B5wnVAGIAwbZAwNTjF/YqATqlinM399Xxq3miOwShfLSq1fXpX8eaUs96pdSxqaZCDVKFmIzXpj9gkyDH0yPj6bTpNWB2sIN0G6IzyJSkrC26P3lHFqqhS3YIXfOHDToijpsNIxWRJCOs57qqBNxQlh0/3C6640WfA5f7VmP1zj4zs3auJ89SwQd3/2P0hOKBEljoIcRuKZRGT2a1Q7RGKMaSQrdqd5GAozBFi7SAifFh1mdgQ96jGZIF6ktJZrWHitINrZREu2WBEHx/cmNjahzrRL6FxRSsL1HUrcQk1SY5pWKhexgimQYOw0KCgoZDygwPpsSkGO/37sBu72DCHnHJ4nngWOBnfCiEZJhAF3t52EhjPFfCptw6TBGxYEbaSyuu197LpGUIEZDUb9NUQP3L0oTD/4ZbkkQB5sPto/lph1G9/xyT1CIv0s4BZBRCbJPa59xC6kpHJEg2tyXaEq9UMie+Gv5Hv6/u7ZTjvHkpbGreKzFmAtJv1Og1RvmpUhhgSJpSL+mVL9scKHnFCC6e30+m7YdOM4bVr/d05p39cMBxvvXfj6LfRbQyoKoUJjfQ3kNDufZHrXonevqc32KuhP+DmIkWSOhEJOpxk2xFTsPMyGCRTNaaaY+0Z5ualFBechlYuAzjvYUEH6YZ2//dgG9Om6V6/YYjNcwS8XGKuYffJbmamoqO+TJlYOVVsBP87Yzcf3L3a3oDkO3Asi258as8KRQeahTdf4ZzwffFWkDAhAdnF5Q35Y0BjwnADhAR0LQGNW7JwiKqEjtMBhYjQ7BsmvjrEjLnaIXW3bl/ILeH23XDqN7p4R/XLt0GEHHmgmcO39xtXgehIBp6wIkRWoB8W7AxVVDdxDQ5wlYGRYpBACX9NrPT+CsdAWFtMbI+duZeMVYcVWDMvTopYFJfuq6fRwlGi2CrQoOlCBCUHhYcAd8nLT7CbPAF5UIvB0itR6a5wpJKIjUdiGEPeY8N4WfEO8FnSpBovAEcUqCBImps74T/5YqIC9KECFBIrXDFIquHUb8PcTvIxxKxgYeSylBFKB0wO4xIIumv1bsYR817Cw/1a2m4/sCefLq+NVMOkCJ/UmfRmzM6BZIk0G7AkhweHOMurtVROPIhVsPUM9PZ7JqBCau2Kk/f+ECqwjRwje0f3O645LKvJmI59bjoxk0d/MBrm3xGEgQNLcIYtF65yWVaNqTHdj4VDYDFTI3KhSLoV/uaUW3tg6Y/w6Zvpmu+2I2k2huAJ8wbOiiZQrzSb+h8yLWys0rxdH1TQOtZc+MDpAdz15em7sWjIoQRPy6BdT2NUsWYhP2V8atok/6BNtiYHqcGqEaTqBIEAfABwgM7rNjlkdl+oldfRAOUKnVN8iM7CDeFWLo6Qa6Z0fe3IbdtTOWpqebk47x/9Kruzv5ZMTnedTkCWKUIrvt544ESAHDKkG0x41ojGqTDmMFJLUA6zXpYWoARb70upp7N0voJIgqBK0AQu/JX5bxYgzMsoJCWgLj46gFATn2La0qsBwbvdggRDB0wiTPLyVIsQgLVCBW0mE8qAGkvQWL9nBxeBg3ZY6QNgo38bggKCKlNxhJDDfziK4ECWOKaiRZUjsdxvhYXtphgp5bDlRAhnQYJ6bmRkj7VKR2mGiNUfceDnwGINmPBGWMqiBAYiE8NBDnisUhajj4aQB3ta3CLSJO8c3MLdy+CLzZqz51qRPciXaDjyZt4LEJ7SgjB7SMaPSLXe4+g+dymwGexx/3t6GyRfNzhC2ARJB21YvznDJ42kbqM2QOk+jwDLysbkn2/8CcYtwEhMfClMfb0zM9akckMBUyH6AKH3RlHSYfoF5csv0Qt6Lg++AGWJMN7d+MOtYswZvAd3w7n6Omw2Fgt1q8WQGV0ZDpASUSPodQKgngD+KWlIFC6fVr6vH697dFO3ntA6WJoNsH03wP1XACRYI4wN8r9tCCrQfphznbmIX2isVaX1b1EoUc9fpGowQxEgfSm4z66JiFWkISSOAgDWBhGWn3SgpU4/OQn31XgmgmcTHReoLYpMNYQSZXmAulFjCw4bXGgFOrZKiUUiZWGOgqpMT3s7fqP29JOu6pZUxBwSswH8CBHTt1YmgmxYL0sUeLA7JAdUSCeG+9cGqOiV1KWSy78R5xaooKgIQRZZ8bosKtJwg8sNyOGa5JEMPmg9/qRyPkfFA/ut2AEENc8RWJqATx2A4jbTTYBY8EidDNaiTItGnT6OzZlO8PLsPfFFJi3PLdIa0wIDEgyYd57t3tKrvacPpUa118untNuraJOyNVow+IkN8vXFEnbJwuFnQwb8UuN9RzXesksGqkVGyAAMGOOHbG72lfhUnce35YRK+NX8OkD5Ti8Gj4fckuvi94NRzV6lsYTX7dr5krAkghcwL1xLgHL2FvDigq7/p+ISdXwmTUKbDxADIFPjFQZN43fHFYs9SiMXnome4BwuOjSet1qwRE6hp9Cx8duZQ/q27QqHxRuqVlBf75mTErOCUHn21Za4lXTlpCkSAOgJxjPyJ9Fm8LkCDGD1I45IuGBDH0FKP/MI+2w2dVJG/ZH1CC1EgoqBe5u5JPONp5M04COgkSgYzwvx3GmSfIybP2niBmlCoSKMR2p6Ix6Yz1gVaY1lWKpTD2kh0z5QlijcXbAm1lAAoMaStSUEgLyI5M1zolQxbeXbRkAJh/ndLGm7RQgui+EKlojunVHBU7oE6fR6g5qvN5RAjwmAiScMwh4hfoljDwaozq5T0xG49H2jUU4sitaW0wGjn8eyPvHcgvL/GNQSWIg3YYTQmCRWJWIrc7dOhABw6kNJ1PTk7mvymkJOgwjgLd6pZkUkyIDLTBuGn/+G/tPh7n4JVwWxv7RMZIQPwsWseh3GgWwYtkzJKd9PHkwPne16EKfX5TEz5nEORjl+3mlpy3r21AuXJcRA+PWEwTVu7hOv3lnnXppavqsHeJLDKREgiSEwvID3o3jJiwpJB1ADNgxCWLNweSK6/9YpYrJQY+Lx/d0IgVIfhMvfdP+BS7axqX4TYaqEee+30FE3r4vL5/fTA1BolIXkgL+Neg3R8dCPg+f92vqf631/9aE9G7xG+ob1IEoJA1+kLgjfM6MUtqgFMjpqAx6nnXMiGjuzzH9YXpT5YiFTJUsNTS9xjOSE96wo0tKrIY8DMdBs87sjGqeIJEMEbV/u4kLz34OpxMAz+QlH2pcAcHwP5mpULQL8CQDCivucVLW5GCQlpg0upApHjnWqFxiNjVwIIROzVIMIgGotBwogTxalpqXAhHMsc0XsfNYjtoihpZBWBc+LtRUBzVWiEjKUFANhfM460lJi09QYIkiLPHkvfFrS8IiIbA7XNHvH8UwW5boQTibeVECYLXDRs3fLsspITE99IqsWP//v0UExOTLueUkTFxVSIv2KCWQIgAElRQdzYuX4Sualja1X39phkv9mxYOmzLXzisSzzCxAbweJcaYa+L+hlxoMAjnavTE11r8tiTfPwMPasZS97VtjLVKxtLP87dRv+u3ssEyE93tWCi44U/Vuo1juDBjlWZHHGThKOQNQASA94cX/VtyuMjop37fTPPVaslxm+ooPDxgTHw8h329QnGqVeursufSaRkjtfMiSvGx9CLV9bRr/fmhDWu6xzMaXIfX/y3kWulPs3L6X+//6fFabr5q0iQCMAHBYvw+IJ5dCWCeXByCmk7QYazExgX68YIWyeQAk921cL1Jxsd4ktrCohdh+w/hEaSw7hTlRrtMHjeYqpn10JUwKEhq5AJ+QwpLHaQ1wHmbG4lX07fH2mPsspyRx+gFIJ+mCz6DZwTWFwYfqVHMSnfQdl5F18bBYXUBnLtNyUd4x28S6rHpygexE9ovmaCnTaeIIFxHruUbkxLA4/jzBwzcJ3cIW0UbuJxHStBdCNv5/OIjP0FI3iCeL1/zB0yB4sHR8TH0eZGT0oQF8aofE56RPLpVCHAsPAStYjb+Qjzp3yWnXiC4DskLTFZwRz1mmuu4QPP69Zbb9V/x3HVVVdR165dqXXr1ul9mhkO41cEW2HQOvzroh16G4qb+FcQD/+uCpDW1zT21gYDoLUF/DJUKSAvwtUn8CsDkdmwXBFWgQhgCInvT+XiMfRQp2q0Ye8Rvgx48rIa1KRCHJtOwjPBCMQAP9qlhoq9zeboXDuBxj14Madsbdx3jB4YvtjV+gRR0Fc1DIRyvDcxfFdDleIFuVULGPTnSn3tCJKuZeWgCgpx05E2oM2AghYbSFjb4b7hiyPA9+OBnxbzZntaQJEgEbBpX2BxVbNkYaoQH9h13uJhwQW2S+RLYNOcIJ9B8ub2Q2ZWT8iOkpUEWC+EYnI7UoLIfcBbwxjHJaZ0sivnB4yES4yN/FEuj1T86+kwDpQgkO2COcUAkxokBBzAcd+IV4PczQxMdmKOmtF6ozEY3vL1XG4NQ0zcN5rZWFoB/eV4L8Fot6kWWIRiUaqgkBaYvCZQULeoHGfZEy4yaUkC8wIU0gddKEHYtFQbi12rAY45a4kwLpbFr8SNMaoTT5AQJYiLXa6gMWoux/fvRgkiRAbGHFGSRH4cbc7VPK28GaM6eyyv7VDy3jv5jBXzaI6K9x91Ol47p2ogicnNCiRIbGwsH/hOFypUSP8dR8mSJemuu+6iH374Ib1PM0MBxIUoZZEK89KfK3WZPvwR3GDs8l3cMgtFSaQYWzss23GIk1zAQUgKmB1+mLOVpq9P4vr43esb6MqTaev20aiFO/g+3upVn3JcdBEbTKKWgdEp2nTgv/D4qKX6faEGfe/6BnSbIaZUIXujbNECNKRvU/58Ie3ljb8CRsFOAfINn6spa/fRwq3haxSQIDB8xxroHc0KIofWxiVA7S1mxU6B9c0zmtHqnE0HuE1N0nAAJCO9E6Flxy8oEiQCxC+jYnwBqhAX41kJglQYFAL44DrZDQEweEph69YXxNxTXDCMWkJM7mCOJtGsSUdOOzBFDV0A6O0wPqbDGNt67GSA0g4TyYtEXkMnniAYJBK09ymSP4oXzNAmeCsViED6pzNaTC4GQ0S9CT6ctN53M9xwEKMmEHY1Egrpl6UVc6yQ+YEduGdGL2cyD4omNy1n0grTqWZoK4ygacWAEgQqKa9u51jUo3AH4hy0qRhbHt2qAQ54aIdx8xiyaC7mIOrXqyeIU2NU4/27aVMRcgGvsVM5ejCJxntErhNjVHM7lFMgSUael9w+HIp7jMmV+QsEiLTUREKCmKOmYjtqWmHo0KF8vPDCC/T111/rv+P48ssvaeDAgRQf7z6qNStj4upE3iVGQsrKXYdp0bZDXOf97zLnkbgCUVX0ikIFIguyqxuWoWpazWEFbJDC3BTAuWI3Xb7PA39bzj/3a1WRmlaMo3cnruXnhla0d65rwJG5d363QFc+Q2n45c1NolKvKGRNwFAUnxmJ0BWzXifAJnyvxgE1yPsTwxMNWCu92rMu//z9nK2syAKwcfvq1YHL5W9TtM0hx+dRrACv67AZjHX2vR2qhIRWIJnGS9qdW6hgaackSLEYKpTvdEhbi9f7cSNpw4cCxbB3EiRQ3MTksffr0EmQ/Ll1l/hwPd9BU9TQj48UbH4uiHVFS5hisIBDJUiQBHHG/ZUqkp8di3ejNag8pZkfiFtzVDj2PzFqKb9Wd15S2XP0mxNgUPpZG3C/va05u5yjFWXE/O10exrtVhj9QGCih7YhSNVBNCJGT0EhHMYu28W7baIMw64dPGUCuyvhCVIsGudvCbSAdapVwvI62G1EAQu/BXwmsXPjFqICgcLDiXJNFrIYC7yqAeK0VpdwEMWAeEm4aodxqwRx8TyEAHeSuuYlJtetH4hx7nWbDgPizK0nSDAi2fljQWkjSmonBFhQCeKOBBElZXGHKhCgpLYZkxWUIAKQIMC+ffto7drArmqNGjWoePHi6XxmGQ9/aakwHWqUoDf+CpAK97avoscnOwVICex2g3tz6yNiVO1CxcEGpp3tVSBYzD02ainXmTCVBNkheGvCGm0uyM/GkLM2JOmmkm/0qs/PCwtS4+bSj3e0oBZaYqMChajiZ21M4pQgqKRBsvJx+CSPNRe02hmvqfyPzd0yRfNT80rFHM0RmQGX1y9N6/YcoY8mb6BnRq/gFiu0UzkB2k9ADs7YkMRGpOE+Z62rxtM1jcrQb4t30tO/LeeIZ2zQ39i8PE1YsYfrJwCf/X8eacuqDifAOrhKiYK0dPshNv5F21u/1hXpy/826d8nKELQOpOaUEqQCED8ppAXYK6ArR6UIFs1uT7aH9wgaI7qsR1GU4DIF99MgqBNRy5DIRTc6bMvpqR4NPcr60oQH9NhdNVJGFmweIJEisiV11CihyMBfXeRWoO8AIP1usSjLIvEZGkHWTREMqIbu3QXS9vgQfD+v+t9PVczhs/bxovH2qUKU9tq8dS/TWCiH68VLWmBbftP6CQIdmUh1wNUS4xCJExYsZt7WPEZBgEJaTN2GDGRfzw58ncHxTB8N2CAake4Ia2jurZbuGKnfRRdOMhC00mLigAktkjJvZljRn6sOC+eINpzcVochWvdjNYY1Xj/btpt5DV1R4J4U4Lgsyl93k49QXQlyAnn74vMK9i8MLa12kHePzFSd4q9RwJEhrR3OoEsdrMSCXL8+HG67bbbqFSpUtS2bVs+SpcuTbfffjv/TYH076UsrLDQhTk94mDvuMR5JK5g9OKACuTiasV1s3m3hOTbfwdImOublaPyYep3kBogXPB9euf6BrpiDCTKd7O38s9vXFOf05Ue/Xkp+4v0aV6eF3kg1qGoFcAAUxEgQeA1m7J2Lz0ycgk1eWUi3f7tAjaP/WzqRvpl4Q6el0EgQXWIcQ0//6e1H30yZQM99/tKum3YAmr80kTq+808+m72FialMjtAyiF6GRvlA75f6Pg5QcmBzzPw7sR1ERWraF3BHLNq92EaNmuLTmK82au+3q2A1/2pX5e5Ur9W0+Jx1yce5f/vblslhKRKi8ADRYKEAd7MYDtMDJWXdhjtMjfYsj9IpriB7AK6JUGEEBCVRIyNSkOKJizIIfWVYiqcEsQuvs/uMVJbCRIjSpAIXiQ6CeJwZ1UKMb89QWZuDEzwdUvHhu3FjtPk4yJXt8PaxODuwfrEI1FHc4b7Poi0FP17GAQ7aekYiKx1u/jyiu2at44UJEKCbNb8exQUrDB9/T568KclvPt9XZOyrGR6sFM17rkGhkzbHNHvadLqQFx6p5rWKhBB3TKB3vNVWiKYWziNLjVCFuhulSBujFFFKegmIUSUIE6fixelxnE3xqgeyIlolCBuSRB53piTY5yqgPRWqDOuya8iDhRAgPh5QG3kLR7XhRJE24DIShHxjzzyCP3333/0559/0qFDh/j4/fff+bLHHnssvU8vwwBjLBZ12NnGQhZ4unstR23MKeoVLRVG5P9ugcfH5hLSOR40mDeasWbPYb214PkrajNpIzXn/35dxj/f0KwctalajNtiQO5Vjo+h5y6vxWPLwyOW6PcFzxM7lWF2A7xYnhuzglq8Non6D53PpBYU31DUXFanJPVtVYGVNW9dW5+G9W9G4x+8hA/8jMse71KdjTxBFGADGp8rECZI7mnzxmTq9uF0boeNtIGaUQGi7b3rG7LfDcjpO79d4Pi53N+hKhMYIOlmbtgfcewf2C3QivbexHW6T2HpIvnpFa1dBkDK0e9Ldjk+f2wmARv2HdVrEKOifFaE8/IDigSJYMCILxwI3XJx+fmLJ8kpbnu9t2rEiVu5vqgWTpx27ncAdYf0FYpfhuwomZUgUjShuEO/bpAECaME0fuVQ4un1EiHOXoqskGcPMdw7TDwi5DXxEk6jLFodxMH6QQz1u+P6AfCj6+9F5EWHEYJJXaphVX1G6t3H2HZGgqCy+oFJGqY7MHmYmE5fUOgYEltCCklRbUYDQthmd3w6aefUsWKFSlfvnzUokULmjdvXtjrjxo1imrWrMnXr1evHo0fPz7k7xjbnn/+ed6xzJ8/P3Xu3JnWrw9VSRw4cIBuuukmKly4MBUpUoR3M48eTZ3PnR+AP8dd3y3kIghGe5Agiz8BduLaVi/Of0MfargxZKpWlAv5Z4c6pQPpASt2eVSCuCQOjKklqWmMKv4kTnt1ORnkuNt2GPdKDZnXhPR3dP8uXiejJ4hTCNmCz5WbTQw9GUaLt3cCL14t+vvuQAEEICHPixIEdZTTeFwB2hyzmhLk119/ZU+Qbt268biJo3v37jRkyBD65Zdf0vv0MgzGLQvEcbaoFMffawzTXeqEH2+tsGDrQdp+4AQTiV1qu5fUYx58V/MC6duygk7MWdXbj4xcyt9zJF5c2yTo4fH1jM3cLox2jKd71GJlwl8r9nBrzYc3NOLx6tVxq0J28B/uXC3bp8BA+YH41Ss/mcl+E6iBMf5g8+3Xe1rT9Cc70Be3NKGXrqpL93WoStc3LUfta5TgVlQc+BmX3d+xGr3csy59eUtTmvp4e/r30Xb0VLea1KxiUf5crd59mA3+O77zH/2xdJdnD6/0BDafv+rXlNsVodRAq6+T5wEC48YWgT5/+NNEus11TcqxAhzrrBmaUosvb1qWOtQItvRNXevcG6Sq5pmDdYXg9ksqhWzwulGdeoEiQRy0wuDDAomzyEEx2Fl5a4TD9oNBCb8bCPvtxhPEmCQjRaEdQSEkiBRRxmLK7ksRsR3Gz3QYgzGqHZwYo540RAy76bEHDmj9+X4Ar6kTPxAgTvu8yYLI7v7QF2gsMld63H2OhN+X7tR3wY2pGO21ARAMe1pASCFZVOlKkGzYDjNy5Eh69NFHud980aJF1KBBA45c3LvXeiKaNWsW9enTh0mLxYsXU8+ePflYsWKFfp233nqLPvroI/riiy9o7ty5FBMTw/d58mRwQQICZOXKlTRx4kQaO3YsTZs2jVMOMiL+XZVIfb+ex2MoyI73ezcMMWhEwXlPu0AUHJROdkoKmPNhvMS40Lh8EUdKkBU7k6P7jLtqh3HvC4ECXhbdTh5LiJJI6jRBYB4hV4ttL0oNGfsdGaPmTxslSEwekBjkmtCR87JKHvKTABNyX9Q9kRDv1RhVV4I4b0cQFWZi8qlMuTCxAlpeEhJSLuZLlCih2mEMteW09YE6olG5onpd46Rdy4zftEhdeA04rfmM+HvlHlq+M5lJFIkKtcInk9fzYhpj4+vX1AshMGSxCA8G1LLwUAMe7VKdY3ZBEo9fHiB9AETqtq+evT1ikKLZ+8vZ9PnUjfz75fVL0Xe3Nac5AzvRoCvrcAS9F5IIt4Hy4O52VWjU3a1p/jOd6c1e9XhzG2Qr2mR7fzkn1ern1AR8x0AK5c55EX+ephlIinC4t30V9tRbvO0QTV27L6LqpI6WrmS0CMDr+qTBsHiti01YUYJs2ndUbwHFvGdMYIIvSGpCkSAOzUwBDKSizHAjB8YkLvG4oiZxrQRxQYIcPxMoosA2Y9ceEFntMRNBESyEcoeQIFAU2BE9QkyY1RnhEmi8Qgr0cMWt/A0MpV3BZCSG8KV3Are7nk6QePgUD7hYgzWuEH4hVcyBEgU9s8dOn+PBD5NFND4EkZIE/tRkblc2KJ3COAmArM4O+Cx9P3uLLz1+8t2TdiHISo1x1mbA0dpvX5eMgvfee4/uvPNO6t+/P9WuXZuJiwIFCtA333xjef0PP/yQLrvsMnriiSeoVq1a9PLLL1Pjxo3pk08+4b/j+/PBBx/Qs88+S1dddRXVr1+fvvvuO9q1axeNGTOGr7N69WqaMGECffXVV6w8ufjii+njjz+mESNG8PVSCyC5hs7czP28iCH8ad42+nn+du4JHr14BxuegvBA2wt6rMct200Dvl9Ad3y3gL8jkCJ/cXNjJrTNQO49koYwzto5rU9aE2iFQZEqsYd2QKQ6AMM2LzsZemJLKrfDyNgCUsjJolvmCfhWOIltl2QYJCBEes1SpLe4IA5kXiuYWp4gHkgQFI1yPq78TWzmV7/9YIIbIM6eU7w23ro2Rj3qvh1GCH1sOLlt78qoaNWqFZPVRjL5xIkT9OKLL/LfFALx4yBmMaeL6qpkrLuaGYDyauyygE+Zl3QVLMhEBYJ4WrtoZ8yXw+cF5gss0I1qJ/xNFtQgN/5dncg1av2ysTSgbRV9B9xYZz9yafVsrQIB8dT9w+m84YDx77ObGtMnNzbmzQun84dT4D3t3aw8q0Meu7Q6BybM23KArvh4BvvAZDbytVnFOPaYMRKAkVCicD5uKZI2l0jPGYIAYCfCIgyQFCQAhKDTpEZ4k+TRgg1kjWz0CkkLX5CsYZObSpD+cMTjCsD2QrqG4s5pawuui6IRY1upIu7MmSTJxI2cVtpCRCERqtIwK0E0EkSbcED0SNoGiiQrd3rdE8RUcEpkrluVTDhIjGG4Al2eJyYunLdV76i8fng9nU4yev+7jyQIdhYAGCdGkm4LIRWOcFun+YFgEMJEK4OQ34C0FEk5eM87mPwQAsx8wPcGvYJmB/f9R0+xrBHfG3y2Jj7SLqzBWDhgkJaedCGJpB0GUcZ4n43v/+9LdtJDI5Yw6fR41xp0b/uqlFVw+vRpWrhwIUcsCnLkyMHtK7Nnz7a8DS6HcsQIqDyE4Ni8eTPt2bOH70MQGxvLZAdue8MNN/D/aIFp2rSpfh1cH48N5cjVV19t+dinTp3iQ3D4sLvPKVQVL/65itwCn83+rSuxDFZI4ZTXuYidyZ8evZzlt7e1qZQiCnWyRONGaIWR8RatYvjM4zvq1uTugId2GCEokj2YY2Ih7CT6FYt6EK5oLQSBkj9P+AWK/l11kQzixVBU5pwYR54gQoI4v//DHkgQeSw8DzfPRVdaukgxkLnKjRLELdEWX0hLhzl6mklxp1HBYozqph0GYzheaxAg2DhwqlbJyAABjbG2bNmyrNgDli5dym2Jf//9d3qfXoaAGKx3q1dS9x0o5cHQFIQDvnMYg9FW4xbwOFu/9yiPReEMWVETYYzDhmOX2gkpNqgwxmC8RL2H1higY80SuhJx0baD+vWhLoTZfHbFNzM200tjV+m+KJ/0acSL5NQGxpoHOlWjXk3K0mvjVzN59umUjaysvLOtezPe9MTVjcqwCS/IJMyJTuaQu9tVoR/nbuO1yT+rEsOmsZTW1q+7TAasqKsQHCLelwgPMRIjdsD3AIQnWvpBCMqa2thlkNokiFKChMFWCzNTPbHDRV8sBkMgEOeZM9WNUcUg1LjItmuHsUoGiJQQI0WamSCRAtRPEkQeSyTMVjA+T7vdSbfJMMYFiBt5eSQs3xHI2a5XJuAZEA7yWcN7hKIz3GcUA458TsU41E/8obXCdK1bMgXJhEIfvYJ2ahAMsNLzCpJq0J8BSagXgODDfRjfH5AhKFZAYhvjq/HZeWXcav4ZL997/6yjbR7irTMqkpKS6Ny5cykk1vgdRIYVcHm468v/ka4DCbcRuXLlori4ONvHBV5//XUmVOQoVy7gTu4UmIB71C9F3eqW5IKzc60SXFC2q16cLqkWz8UuiieYhKFFCjtu6CGe8FBbNquzI0AEPRuV5lYMfIamrgttJ8LnBkUxil3sSjlB9YRAEbDO0O/qXu3kngRxM14FSRBnjwOyyAk5G423iewAOyUO0D+O3WPnShD3EbxelCDGx3Jj8mpnPO7svQ8QFE4gGyBO33t5D6ESdaqiAWntpR0GgI+CkUTJ7Khbty57K2EcbNiwIR9vvPEGX1anTh3K7kBtKpJ8tLAgFQaw8+IIBzFwx6LQKVlnhIyhlYsXDPudX7Y9UM/VKFkoRV0EfwagWolCPPeAWAEalQ+0+QDwBxFkZxUI2iHemBBI4cEGxKgBrdKEADGrHKA6ebZHLf791fGr0zT10A9gIxS1DzbdEV/rBMUK5tUVJBNXBdSukZQgVsrqaloingQ0uDZHNdRJRuU/CBI3nRduoZQgYSAeA0YSRO+JdvGmeG2FCfEEcSA9TpEMkzeyEsSqLxgFFVo27Now5D5StsPkMhiznvfUx2mG7NZZKVKMbKKoV9AbbrWzJe1EbhzGi8aEFpZeJlMzlmlKEPSDRnx87T2BwgVFuNXzSjTEDyIHXST4eA8iLfqcAu8lWguAqxqGtsIYpXgrdx3mNoQrDO0yuO2PcwPxcEji+GjSepa8WilGnEC+d3i/RQGEwgFs8tIdyfydRUECwKUaJqpgqFFIzdkUiKF7V0sDUUhbQLFiVKFACeKGCGlSIY6P1ALI1N7NytGQ6Ztp2Kyt1LFmkAiasHK3/jl3uhDG7h+iq90UBOY2kmIeyAM37QNeyBZcl9t8HCjkoAIzmmo6gZGkwCI60sLASOw78wRxn9rilQTRVSeaotGVMaobTxDtvMB/4PZOzvOg5nXlVAmCDRyQhJiTsfvtRJ2BcxHS2o0SREgTRMkLiZIVgFZFtC8qpATqAnxWkOKBTZUh0wIm1aVckiCY8yVV5mqPqTBSL+bXlNh2WKJtaoF8NwP1EAAfBYyzskvesGzwukYftUgecVkVGOOhwETNis0MJOakJxmEdJLtB47Tt7O30sMjlzAZm5p1h5/A6wbiD60taBE2mvSGQwVNmR0pWaZUbDAcxAxsxE6kAImyds9RuiwYGuOaBDF3GczbvJ8uqxto9/cbigQJ88WURBdjO0wxvS/WvRIE5jVu4ckTJEw7jNk8NNmiLzhSTK4QE3bGqPw4p876ImHVlSARdsXw2KfOnrZNiAEz6lYJIkaDKCyx8+XH81m+I9mxEgQkBhYEKNYhXbYqVvcZdtmwYEK7D54rmFq3SUR2gLkX1ChYyLSykfVjFx754WYlCG4LOTNue1+HKlycLN1+iOWqN7UI9CK6gXzvYFJsnCgrGkgQY1EFIA+9dZV46vnpTDalxUDvJEUioyM+Pp5y5sxJiYmh7D1+L1nSWtKIy8NdX/7HZUiHMV4HO5dyHbPx6tmzZzkxxu5xgbx58/KRkdG3VUX6asZmLk4xKWOCxlwwYn6g79tI8EWC7Ix4SWsym/+mVkxqNIqTcAli5vuXedMNcQDFAcaySKaGQspjvHRCvOskiwt1xqE0VYJETkQzA+S+jP2Y052cp94O49ATRMZdzP/7jpymqg5SPIXAQFHr1pxSPERAuGVW/PHHH46ve+WVV1J2xl8rdusqEMztXpUgSPnAxhF2xZ1I8sOTIOE/s8u2B+q5BhabWhKPDhJkyfaACqRK8RjdxFgIYuCjPo2yrQrk5wXbeYMK49erPUONZZ0QXqg5MZZfpM25aCuK5rXEbZ+/og6rlxH5ilS5aU92cESwZwQICTJr435eBwhxEQ75cjnbbJd2GMy5eM2NNgVCZhjb9J0ASiljTK7V2nL2RkWCpDlg5nVMj8c1kCBaUWocwFJTCRIVCZI7+PZKQWU2RtV3uAyFUKR2mKBxW2jxhAIUhSgYXXxJ/CANRLIcTgkihM+BY/YxuV6UIExC5M3Fu1ko5qN9PlCUyCJejBMjAYsTJkGOnaYqFip8KRBRMGLwRg/sxn3HmHizI0EwQMmE4QTw1QAur1/a1pyqmdZ3i0grPE95rf5ZFZDkdatbincS0cYAEgSyOy8kyAHNlM+8aKscHxiA1+89og/mksLTqWYCq0M+v6kxR6d5cYrPiMiTJw81adKEJk2axAkvwPnz5/n3+++/3/I2MODD3x9++GH9MiS8iDFfpUqVmMjAdYT0gGIDXh/33HOPfh+HDh1iPxI8PjB58mR+bHiHZGZgrEfMIT6fH09ezzGGmIBhugtz6SttlFDh2mHkMxmN+W9qGaN6VYIATgxfkzyQORjLoe7DQgaFVqTvq4z5Tj00pHDDHOVU4Rd9O4wHY1SXRTfmbSwcD504TeWpgO/tMEKCbEo65tgcVfcD0Vpb3KB4FmiHkXFZgDnabD4oCza0NmZnbNbSGBtr7SK6J4hLY1TxEEDbpFfo7dNhxh6MHeLxFlYJUiZWV3wYW2Gw8Bd09RABnBWAddSrWssyEkGceMXh+4P0kw/+XcepJmagBn6sS3VPhrgCzD8gpi59bxqTITM2JIX1yshoNQwigOdvOchqaHh+REI+3XYhvKEpNg9BmmMDBL4ghUsG50MQfF5IEF0JknhUV36aCSfjd8VvKE8QG4i3gMTjRtMOg6zyaNthTkX4cBpxLFw7jKkYk90wY3EnC1g7JciRU/bGbVK4mckWr9DjAsN4ghhVL8dt/EiE4RSjWacoEuN81zMSRA4JbxinC3H5vNnF5O41tMMAZTS10U5NfWQGUjW6fjCNLv94Rkg2vR3wusEsCQi3AERxXLl4DPtyLNhyUC8SwKQDl2qmYWIeNmvDflctXoIkG48BcyTprI1JLK3FhCiL0W4eo/IyMtBeMmTIEPr22285tQVExbFjxzgtBujbt2+IcepDDz3EyS7vvvsurVmzhgYNGkQLFizQSRNMQCBIXnnlFd7FXL58Od9H6dKl9YIeqTJImIGse968eTRz5ky+PUxTcb3Mjoc6VWMzVRQQ/6zcQy9osYaQVrsxq5TJHZ9ZN/MFCnBRN7hTaASui9uiDS21SBB3niDujVHxGXSjoHBjigrIfWOsOhpB/ht1O4zeepO6niDGc3OqBLLyA3Nqjuo0Jhc7tW6TYQTiIZKZ22FADMvxzz//MLH8119/MYmMAz8jnQtjcmri008/pYoVK7IJK4hqjNvhMGrUKKpZsyZfv169ejR+/HhKbZTRdpn3JJ/gxZAoQdy2w4gizukYGL5etB9TNiUFkl2wUVnVpDjB2IjzxzxSq1RhfbHeyBCtjtZhQa4c2XMpBg8YKMuQBgIvECfvywM/LaZ+38zTX1O8vqgv29cozhsVqGsf/XkpJ8lFAyz44Tsm55mZcHWjAAE0etFORyk3+V1stgdbYkLXD1UMShD4p50666y+R6cF9iGw2SybujEmpTY2V90ID9wg1b95bgff1AIieyANchp7ZOUHErIodVHUSltN+Tj37QmyaHOzYDxh0Q5TUPtQnTaYyNmlr4g8NpISxKpFJeg94o+ZqJA0keIbpb0B6h0ryBfS7SJYYnLtdj3dxGjJ50D675wgUkyuDBrir4FFP7DDguBAQYoFHU4ZBIH03IaDxLqVi8tPjSx2O4wQF3bEjAGLtx/kxwQx1lJro8HCEOeIz+Hcze5dn4Py+tDCXdqL0MKAVqxJWitMp1olsrTMtHfv3vTOO+/Q888/zwX2kiVLuKAWY9Nt27bR7t1Bc6/WrVvT8OHDafDgwZxQ8Msvv3AyDEz7BE8++SQ98MADdNddd1GzZs3o6NGjfJ8YwwU//vgjF8mdOnWi7t27c0wu7jMroG6ZWLq+ScCr5K7vF/KEjs/bI52D2fVOgDEJ3xu3OyPyGUeqQKQ2QCOM13Vq+hlsiXBPgjjzBNHa11yQLEYCINmBl4Z4gsQ4bHHDwkb8kpwoNEBKyZxpVEw6QbD1xn3STST1o605qoP3HvOWzGnifeUEbtuBhQQp7tIUNSsao4JcloSYwoUL84GfEXP+4IMPptrjjhw5kslyxPMuWrSIx308rrmlUTBr1izq06cP3X777bR48WImv3GsWLGCUhNSH287cJxrT/GSkQ0epxDSREiU1GqHWaK1wqD2MCtkJRoX64cCuXOy+hVoVK6oJQkiaTHZDUgRAZpWjIsYgQtSq/+weZzegrnxzksq0bxnOtFfD11CQ/o2pWH9m9PC5y7lywEkyXnx4zKivZaEOHXt3kwVmduDW8oC5IGMweGQP7fzAA4xRzX7gmCNZvR9MramhwNEBqJaF18QfB9AaBk3rudahC5keBLE7eCbGsDC8+Wxq6jJK/9Sq9cnU4d3puq+DJFuZ/YDMS6+nO7s4Yu7XduVh2uvZ2NUT54gwaLQuEtmNJKTHS6j0iLYDpPyOWIgCO5U5Q5DgpxLs4jcwONqShCbnT0hhtx4gkRSxcDRutmr/9Jd3y1wVKht0eSebkiQcLuuIHaEqJKdNlEbGTO3BVBHQGIu+GnetohJPtgNB65sUDoimSBEB4gTfE7ktp1rJ+iLDtxH2+oBA7Bp6wLtKm5gt3ONvHMUzXh6kKFKnCnSQ7I6oMLYunUrx8+ibcXYkjJ16lQaNmxYyPWvu+46Wrt2LV8fhS1IDCPwHr300kuc9HLy5En6999/qXr1UAIASTAgU44cOULJycn0zTffUMGC3vqvMyJeuLI2da9XUlc5fXBDQ1dqBnO/q5tizJjY4obAQxEpSjynLTGScuamXUW8iZyo4/Z7UJoAhfI6V1DoJEhe94SRE7JIxlgUZm5bVGSOdOM/Itd1ozoyelglOyCnMO7Dc8VLO4wx+jgSjO2anpUgmdgTxIiNGzdytLgZSMrasmVLqj0uSBao9qAOrF27Nn3xxRds0Iox2wogaqD0e+KJJ1j19/LLL7Na5ZNPPqHUhNRFUGELgQEvMbeJiiW1neo9qUyCLNNNUWNtW2Fqly5MG/cd5V1ubEqKKhUqWblOdgYMSIHyDpJgBk/bxG0RWBz/eEdLeqZH7RSJU1gvPd29lp7gJipmr4AHHkz48XmESXNmQa6caLsL/FzAwTySTyManJAgotgyK0EAKHoEbl6vcAkxssGZWlG5qUqCuB18UZSj/9x4uMVDIxbTFR/PoC7v/0fNX/2X2r09lfO5pShES8LdPyyMWCTKgtVOCeKUBMGXHAtPDKays5H6niBaO4xB9YACGV9mwLjwtW6HsW8BQc+YFE9m8xq+TCMjzFG8XgACSZ535HaYwN8jeYLkdUmC6P3vFoUlkh8gdcdA+/yYlS6UIM7JsDjdgybl44tMGASDvH9Cgli1w6AFBQBTjuthp2WBYTfCDJBg/2lRoVc1jOyyDsIBgyn8ExZtO8SMfeC2oS0Sl1QLTFDT1ruXGMrrYLUgrVcmUGCOnL+d043w3RFiRkHBDTCefHpjY5r2RAea+3Qn/TPrFtV0X5Cj7pNhPJAusS7UAMZxzc1COE5aBFOpHcatl4YQ7u5IEOcJMfpGQb5crlVlnjxBbNLX/IxIlutgjHTjkyVkmbQlRoL4OnipfXRj1MOnMtUurB2gqsOmoNGYGj+DbGjevHmqPObp06fZu6lz5876ZTly5ODfZ8+ebXkbXG68PoDNS7vr+1W7ix8ElCB7Dp/wHI/rhxLk5OnIymFRd9Q3pL1YJcNI24ZRMbJZqwUBqBqyOwkiislwGDpzi672hueFHTBGX6q1sYgvnFdgbGygvb9LNdIrM0BIaozvoqjwa7O9lK4ESbnGMBoRr9vj3hfE6J8m60tRiLlR02YIEsTL4Iv8dLDicriJTzSyTzAr4mi1I6e41wixS0P7N6N5T3fiLxt6xj6fujHs/WwRJYhpwRqUgzrbnZD7AcvtRZafP49zhi6cEsS4sySeISgsZCcstB3GXgkifiB4KlZfLrsoXi8wFo6RdsWE8LEjX7ykw0QihNbuCU70S7QJ0ctnKhyCyqOUnzfZIStuSEoJKkFSDlAzNwYmhNZV43VyIJzEDDnjZ85doJolC3Hcp5Ndz0trB3bPe30+i4lC7OSYo9/aVImnXDkuYtZXdsixa/PmhDX01oQ1IWoVM+R1sIoObajtyPy6aAf/j++9mwJfQcEIfKdQmEcjVa6uKUHctcPYf8adjleS+pU66TDONgLQQiJtIG4ico0KCifkgYz5QsA7u3/nShCvfiChZIt7TxCrTQYnBJib1B43yTCelCAaUe8lDl3aIFCY+1FPpDew+YfWxPLly1PVqlX5wM87d+6kr7/+OlUeMykpiQ1XpT1SgN+h9LMCLndzfb9q9wpxQRJEpPYlC7v30RPiBBshXuHESF8U3sbgBMEWQzu9FdFs9LnJzjWK1KlOzG9BfGAufqtX/YhrqdPnLngikq2wWyPkynlI90wv6F5MhUNTFP2wXSgtJIgFyWg0R0UrjlOIp45VTC7a6gP3XTBzpcOEG3xhyGcFGPiBKReATXY7mCJjGoQBJHRYkMOs0bjAf6Z7Lbr7h0X0y8Id7CBsFakHckAfxEztMNI/i0U1HifSACaO115aYdxEF0VSgkhRhV1GKRrBqMp6UwzcIi389QItr/WumB7F6wsJEnh8kC2R+gWdKkHckiDhPEHWGuRemHAjfR7EbNdNO4ykQxyweC/2mUxRgTJFCujnAx8ced3A2mLCwSTSvGIcG9vhOzA3jMRM2lmcqEAENzYvT38uDdwOuK9D1RTvHYp1pLSgbea3xTvZEAtGrVJYI8nF7jHDyesRhfvOP+v03x/sVM3xeSsopAaqe4jJFbWT2xYStwkx7Atx3EM6TBiS3GqhDcIzUjujXbuKE28p3RjVRey1blh6KnVJkCDZctZDLHxuT+0wSIeJBHnf3SaeCZnllARJlDnKgycI5nTUGXh/Qfi79UjJaADpsWzZMk7jkhoY7SbYGMzsvlV+1O5liubnzTXUcGJw7tYU1XgbjD9OanQrnHCwaVa7VGFODUFbDOJ4zQpAbMZic6yT1pK7YOsBPf2ijmbk7lYlltWATTvUqou3HaQmFezVHcDHfRqxwlGI2HBYpSlxYJoaDaCUl2ALvN+ZBcYNUifIL54gBr9IO5TWvl9WSpCq2qaP54SYvcdSrCWhLDcnK/mJDGVJnDdvXt0wSg63aF0lnjrWTKA2VeN5YDIXEp1qJfBEjkl8smaeaAakniAIMCCbWV5MyijqnBrDBckUbySI9HPZGX6GV4KEDuAxWpEo8mEp7vIYWmUi+WAETVGtC5JgOkz0A7sUjk6KH1GliMrFDFHSuE+HsX4t0NNp7vO38uEQgMSSBbzVzkEk6Xk4JYix3xo/Q14JNUWioZdaJgUwrhhcRAmybEeypY8KlBlzNOPSKxo4z+duVaUYR4thUXVzy/J0a+uKltfr1ThAcnw3awvd9NWckKL6o0nrbdUg+gLRYmcZhfaHNzRk9RdiwWBwqaCQnpDJHd99py2UXn00QhbCDuYm9KlD6eX2sfSW0AiPId9pXN9JDK0RooJwowRx0w7jhpyQ1zLWQ0S6rmhxYRQuc6xbT5CiLlRAusG0S4WOLEDs0srMkB1vt+aWApnbpK0mswML4C5durARKo5LL700BQGCNJbt27f78njx8fGUM2fOkBYcAL8jCt0KuNzN9f2q3bFxWVpTBIhC1Us7DMhKqfO8fm6swgXMaFk5YAQ/1yK+s0ONoKEm4nPRsox1BeKlpX7OrmaoRkir9JglOyNeF5tpTggQBGD8vTKgWpJWFq9YsztQ48PM360pdnoiaEjtbNzNp5EgUG+GU2IblSD4bpmva1xfndVqCyeQ+0TNIN898/xnJhozPAniZfBNC0D5cXWjwAJs/PJgaoJV2wIGZLMpEyYsN7234pBbyUULRLQeGxJPm6IdxhQ7qLfC5M8dMhFLMYUCFGoCOyWIFaQQRYEdLfQdsQh+IEayyE4xczJqJUjoew1VBcgmTG41tN1eUXpYQYxTMUi4SXzQlSAWRaf0ZRsnBiw2ZEDZofVbAqt3B426hIjBjgn8XZZqLudGjF22i42VmlYoSmVdygBhorrw2c70Ss96tjtcHWuV4F0AkHtoXcPr+Os9rZhE27jvGO+iuEmHEUBBsuSFLvS/y2q4OmcFhdQAZKaS2AQjZVdmpV6UIC48QeRxUOi72S0VY1SoIcMpFIUE8eJt4sZLQ0h/N6SBmzaV6Nph3HmCoKiUDQ/PniBO3nuD+a4bxGtFNV5zOxNyAWoWUel4aYcxkidOEg6yCmCSeuaMP+l6efLkoSZNmtCkSZP0yxDXi99btWpleRtcbrw+APWK3fX9hBhkSt1c0sPnBjWHtFd49QXR68UwJEgLvaV4fwrPmrbVivNmDGobkNqSrGckTBob4nKzK3rUL82byit2Ho46yQXA+/DcmJU87uD1bacZpHrFKi3lJ1pFSUYnQfIb5v9I1gsgpvHZxgaKWRFojBI2t8GHgyQlQSUvRIpxPsdcWNmjkCDdSBAvg29aQXwL/lu3z5L1kr4ktNJYQYqhSEoQfCFXaYtPMchzCy/tJSfOWLfDmM3ggiRIaLFlLPbMBZWVkWq052sHeSwnSpCgJ8i58LnvLiNyhRAyv9fS7wZlhbQ6hSNB9B69QvlcSV91TxCLz5qdd4AsuuB9I1it+ZfUKhWUqzXW5GWLtgV67oz4Q2tpMZuaOkWk5why8ac7W3IPIXx6frqzBTWpEEeNNUmkFQmC91DamsLtXONzntnlxQpZBzKPICXAlRLE5S6923YYGVPcKk6guhMzv3BzYHDM80KCOE9VORqNEsQBOSHzZKwDMt627ebkWUfmnkbvC9eeIC5UQF7aoOS9F9VoJDWIKBVxG7eqlhQJMQYPBQV3QJvKkCFD6Ntvv6XVq1fTPffcQ8eOHePAAqBv377cziJ46KGHOBL93Xff5badQYMG0YIFCziFLLVhbhUWM3m3EPLEa0KME0+Q+mVjecGGzSjz2A4yWuorqEGMhImgRaWgabsbz7+sBIw/7WsEiIqPJm+IqEKIhO9mb+U2a8xPb/Sq71qBaATG6/Er9oRsHmY6EsThBkReQydAJHNUxBrjbYKSyfz9wGsv6OAimXGiluJzaa0EvXY3zn9QgUTzXqZbO0ykwTe9AIYQxSKUHEs00xUj1mqutrLDb4bsnkSSnUItgB0XMJ1emcQYvYXlbNTtMEJ2SFF32Ka1BbIz2cEyF1R2xElK5co5H+Nxczluh7GNyJVJzWXcWjAOMvR1gHkXUKl4jD5xy2V+MLNWu67m52aXlCIkiNEcdbUm6zN+DhtpOxHiXi7AjjXaZDDIda/nvBXGLaBG+eeRdvTf4x2YAAHqar2yKyxirMWMGKoRr0W1gkJaQwy9pLc1EoLkpnvyoEh+Fy0RHr1HUKSEi+5O0ZfsiQTR5j1X7TDOx3aZ89wYo0qrkZfngeLezq/KCFGmYIxzGw2qm+I6UoKc8fzei/JwXwRfkGAyjDcVSEhCjIMIegVr9O7dm9555x16/vnnqWHDhrRkyRImOcSvb9u2bWzYKmjdujXHnw8ePJgaNGhAv/zyC40ZM4bq1q2b6ucqCTHScj56yU5PyUDRJsTIplnOMJsp+H6Kj8Vsq5YYbRE4Zc0+alkp2Dojz+eWVhX06347K/UikjM6+repxMoCeMk9PHIJp0K6Bdo4nh2znF74Y6XuRefEzD8cRszfTvM2H2Ci67omZSmzAJ9dSV502nq/QSPx8D5g7gmH3xcHWpfaVosP2QyHHcAaQyLMxdWcKUHwfos1RZc6wS4R46aGKKkyHQkSafBNL2CRLznSVr4ga7Rdcxg0WiGccagRWEgCNUsV8uwALR8ExJmaW1PscNymHUYv/LRCK5zMVxbf5pYfua2dJ4if6TDulCDhjVH1dBjXShAhQc6ETMZ7kk/oJkHGaDc7WPl3OAHInTw2O2923gHSviIxuVgkSItXKAkSmMBhSmV8bqICgZzNi5TdDUC0GBleyQRfockQLUmfmDxK6aGQaVDFpRLEq1+DcSx31BLhUQ1gvI0TJYg3EsR5OowXY1Q37TaHomiHgcxYev/dPBc3LZMCY5tupIWjEG0yz3syRz3ijATx6gdivK3MnwreABXH1q1bOcp27ty51KJFC/1vU6dOpWHDhoVc/7rrrqO1a9fy9VesWEHdu3dPk/OsoMVhYgGLjQ6YUi7YmnKj0nFCjFanuUV1Tbk9QfOWsIOoOeZYGMyLwmHWxiSqUyaW1QkwAZU60UgOvv7XmiwRA+0F8G/8uE9jfn1AhAz4fqErZQzm1Zu/nks/zNnGPo5PdK1BD0Vpig8F0WvjVvPPj3ep4crHL70xZPomJv+wGXpZXWfWE0Nnbub/u9YpGdaQG5/R37X1QU/NVkJgXEsjmdHpRiVaYbAWRU1hNMddaPjeN0zF1rEc6Tn4pic6aAMUWFrzmyxKkJolrdUbYhwayYUdrtF2OeJOYdzdcqquOC7tMKadMZHmisLC6AliRtAc9Yy1OsOmICzoqzGqc08QeZ3slCBePUGksMROnlE6LfFQJWPzpyAd/FwQYLEfjMk1kSB6z72pHaZoaDsM2FnMryBgjP4hUF3AFBdkihh24fP/x5LoWmGigZiZwln61NlzUcd5KihkNiVINOkw7tQA2uN4MPx0shEgSgFP7TC6t9SZVDFG1efCVPYEwfgdJFwiP1Ykz61wEKUKfJ4ibUKIx5WX9143R41g9KubonpIhhHIbbOKMapCeIiqFsofWcCN1nae01IJ0lczdEctFC4JSQzmjQoPAdJEMPZhY275jmTdpFNMX4HL6weVtmjPz67oUb8UDe7blNsysJhu9/YUevHPlbwQtiKHUM//tmgHXf/FbOr07n+s2MCYOeSWpqwCiWaTDI8HVQl8DdGGAaVKZgHGyc+nbuSfn+pW09Hm+4Fjp+m3RYHv2G0Xh3+uSDtC2z+6DC6tHSpm+Hf13hTGwE7wz8pAK0zHmiX0DQMEOeA9FURrcJtp0mHSEjDMwfcEnh3GvkEsVlHY4b2w8/EQyXEkY1R8YIAGZb2nVEByJ/3Xdskn9koQsyeIyRhVV3Xksn2O5p2+cLfxmwSRAttJoZZf2wW09QTxmA6DQUReR2NrkHxmMNkKSRGuKHTbo+ckjUEezyybL6eRIKL+EFNUc0uWUc45TZuAV+46zIQIJiOjNC2tAPYaCyyYLpljRaNJzVBQSC9U1kgQ9NJCthsOIP7EVNqTMaq2EHaSXCax31EpQZyMeVG0wzhLh3FvjBr0HDnrnATxmA4QVGA6UIJo1/ESB4u5TVSDkWqTaFRAxRwqQaSFJUEpQRQcQlS18Nnooi2yxi3bnWJDJBKwOQVAeeEFkN+jbj997jz9NHeb7fUalIvlWglECQzdjcBCXNQgU9gXJGWajHEh+fWMwG58dgUWzt/d1pzHpMTDp2jozC3U6/NZdPGbU+iOb+fTjUPmUM9PZ1KX9/+jpq/8S4/+vJTmbTnAazXEEI+5rzV1Ni3MvWDsst28oMe6661r62eqFJ93/1nLax20uhsJtnD4ad427jSAChtBCOHwu7ZBiu+msdMApLusIYTQcEo4iR+IfN/N7WEgRlNTkZ5tG+vxooJdAlEB46Ibmpfny6WnqWKxGFsWLdgqYl8AQpEgXgvid+AV2OFCUeOUWNA9QXLbtcOIEuSsfTuMLq218wTx1g4zcv42mrlhP71+Tb2IO3dSqDmR7EbyBAlG5LpvS0JLzPHTAX+XClrKTwgJUjC48JAceLuC0Is0WCdBDO0waI2SQtesBJFeSHiCgPCyI0EATNKzN+3nXQgw3r9rUWWdayWki+8GXju4QC/aFmCcjTG3dkawCgoZGVgEYnxC6se2A8eoaolCEXfoUXiFk6XaoagWqR2OnDCTil5aIpx4gkRD/Hpqh3HlCaKRLE48QY57V4KEmrBGfiy5jpexN+DVkpsXECBuyoW5bjSqOqdKEJxH9J4ggdvuywLGqC+99FLYv6NtHPjyyy/TvWU8vYAxD5sgqG2QcgeDUxAZUGw7lfb7oQTBdwn1EDwqvp+zlQa0q2LplYCNJBigooZCS4xEohsX9j8v2MFrjOevqEOfTtmop8ngMZpVDK4Npq9P4lotsyWR+AkYyM4e2JGmrUuicct28QIZimajyb8A6YK9m5aja5uW1dOAogUW8k//tpx/vr9Dtah9RdISK3Ym06iFO/jn5y6v7UgNc/rsefpudoBwuO3iimFvgzUHUiMlhdGI6QYCBPV7RYdJLvAqxHsLAv+SasX12sUYmZxa0biU3UkQYatAgkB+JSSIRPUYF19epMAYEMEiY2dbesK9IiZPgARx4rOBtg1RPZiLQrMxarKHdphwtzEWb1bnCuOc//26XHfWvuOSyr7tVAorKfGCZpzw2A4jCwt8UWV3Fa+x7DBg8JWFB7fMnDhruWMo0nAvu6I6CWIoOoUgwphljjnEokaKB7R2BUmQlAN6+xoluB919sb9vJvxqyaLuzIdWmEEIJqYBDlwzFr5kso+JQoKvhN7xQty4hF2C8ORIGL+i++0Fzd0GSvgY4HxKNwuloyvxTKkJ4hzY1QhvlNdCRIlCeKG0HEbj2tsiQH5EE4Jgs+FbG7I3OUGbo1RS0RDgmibBlBH4X02+5wJ7DYfMhJGjx4d8jticDdv3ky5cuWiKlWq6CTIjTfeSNkZFeIK0KHjyVwvXtWoNH353yYavXiHKxJEPEFQ02ChF8ns0QowhX91/GpWIf21YneKhZ+xJQYkyKTViXRzy6DZKdCmWjybvGLcX7fnCLcfY3Pql4U76Lqm5ZjoQToevE9EDfLOdQ0oOwPEEhQyOLB5CXIIcwnWMzF5cnGbP8biWiUL+5YYgvED6oOXxq7i5JMWleLonvZVKLMA5//quNXc9n5lg9J6MlEk/LViN88XmKN71Atf88/YkMQKLcz9ZtPTkFYYF6kw/6wKeO5cXLW47tcIQ1ooU9KKBMm27TDGvqWZG5J0ud2sjQGDozZVg/FVdr234Rz4/9PyktvVQNtNdF/UYItJZEmgsWXGrLQwG6NKISStL0bIwjqFEsQmUcb8mFbn+vGkDSGSs0iQRb95kW8FaVkRV28zTmnGqF6VIMZdWkysssDA4IFBW94jWcTY9UcXL5jPl3YY42tjtdARwgPtLaJuqmMR8wUDMEzCGHQgMQyoXQqwvDC9UF4zodpmihz2mmahoJBZzFHley3mk17HKhRDkaJS3SjtzBDiRIgOM7BglQW9l0WwxONhIyGSSZ7MNe48QZypM1BcRk+C5HbvCeKRBBECPpxfGTZBJInSydxqhigPxZPKDtLCkuCBBDN6w4jKM9yu/p3fLaSL35xM/2rS6oyIxYsXhxwwHEUqS6dOneiRRx5J79PLMCivqW2hBL1aM1/ERqWT6GcBWqhBOGAc9JosBOLk5hYBUgOtGXYAOYMabMraffpOuQB18v0dq/LPr/21mqpoShGkmCCFD2heMbjWgBJ3r/K/0YF6HWTIjS3KMwmFdpfWVeKpTulY3wgQpJM8O2YFDfozQIBc26QsfXd7c0/EWXoBJASIOJzzk5fVcHQbzG1fay1YfVtWiPh8xSsQbTa5cwavi7XQ5DXBcdfN2sHcCgO1yQ9ztoZcR5EgqQgsCrGIhXpg/uaDXLQt1Xw88EWzg7SKhNsFE5Oj9loKTTQQRYcTJYi0zIB9NmY/WxmjhvNYCMqqz7gyK5VzBRljNjQyZkhDgbNZM+O0g7y+jpQgYR43RAniMh0mNCEmcD67NGkeijshIKzUGsZBQl5rT+0wBVK2wxiTUqxQU5NUIukF7VH4LKDFywwQdP+7rGbIZQ90rMYJSukFu8jhoAeKIkEUMqcvSCRz1GhMUQEUJ7JYD9eqYvy7l++TEBt2Xg1CjkB5J4tYNyiYJxer3CIpKLDLC6LEqzEqbh+OZMHYCaNRowLULWTDwIkSRIgSL61QTv3KhPxCS5CxmHUKaW/CrmBqK0EwP5XSIt93H7JfHG7df4x32PO69PxKbxQuXJhefPFFeu6559L7VDKUEkQ8lBBOgPYQeISNWx5540yABXJCbF59I8grsPgGmYJ6FSl6VkCK5H2aauCZ0StCPAYBJJXc3LI8EzJrteRJjCsP/LSYN1+7GRQueJ7fau0JCqkPEGv9vplHP84NJMsM7FaT3r62vut48vTE9gPHaZAWDXzHxZX0oIZIWLTtICeYgvzA5zwcsLn8t5aUZFZELdl+UO8YQFtLU0OLVzhAXY/vJnu61CqhrxGNrU/47tW22Lz1E5lrxvAZGCiFpIAsaMb6JC54sDMeLhIpNkI7zJakY7Rl/3EmIlpXdZaVHA5BdYVzkzgoI8wKFLMSJFwUo10CTqSI3EJ5A5djwDe2pkB6LYtYtMIAC7TWIyuAyBACxpknSC79cY1SKoEoRPJ5GNzMBIdMciK5NF7Hqk8a5AmIELwdnkwIC6a870gmodJXKjFTiKyyIzZ61CtF1zQqw61bj15anX9OT4gSBDtBfsnrFRQyQkJMJCWIH+a/ToyajaSqFyWIJL7Y7bIav6telJCYm0GERFJQGOdEN2SLU5JF4nFhkuellTLEE8SB/0jQGNVjO4yNl5efKVvSjhguNQMbNuJN5iUdyMrfYZdN3ClqBRAggNMFQEZCcnIyHwqh5qiiBJV6ZLTWqusUXWsHyIW3Jqzh3X4vwPh1eYOAweQwg1mjGQ90qsbGklCNPfnrspCNOIx/L11Zl9sURIEFYAH49oS1rG4wJvENmbbZ1ttOwT+s3JVMV382i9X/mDsG39KUvV8yeludERv2HqXrvpjNxAE2D9208HwzI/B57tmwdMQW84mrE3k9h7VxY1Nc7cRVoakwThU0EzVSBeEM8vjyHcNaBAABktqEVLb2BAEgt4OZDJhAHMBlEVIxRBmQfMLaCFNUIE0rFvXFXFJvh3EwMEpRaPW4ot5AcYJJIVgM5Q1jjGodkWsnDWaH+pw5eHcORZ+chxT/+HCjXw0M5Pq99gsCfOFkh89ROoyhQMVrYGx7wXt0Umt3ypcnh/dFhbZoEFmu0YwpnBJEWmHwPLzsvAUjck9FjMcVNKtYlFUqIF+AXo3L2t4/Pr/v9W5IGa0I2p18IqSfVzeXjSJyUUEhPVBZa4eBEiScf4Ef5r8Yi5DwFE4JAlJYUmi8kIryHcTYZvV8ZIyEN5FXgAjAOYZTQMrfoHRzo15jkiVvLiZAQLLYvQZGU1SvxXEwle1sqkbkhmxehFOCRGGIa2zVwmOgjrCa00QFwu0sUdZApSXpw6YdBs9HlJ6li2TcueGjjz4K+R3fG7TDfP/999StW7d0O6+MqgQRJSj8yV7/azUt2HqQiRGpDyIBxATideHHAb+HSB50dujfuhJHiCKl5unutSyNfvEdeL93Q+rx0XQ214Sk/5ZWgZhdGW/evb4BjzVomxF8NWMz+yu8dFVdTo2Bjxvq3gd/WkKf39zYU72oEB7rE4/QB5PW8/spa5Kv+jXNdIa0IHH6fj2PNzuqlShIP9zRwnGqGPx2sPEPOIkA/l2Lqb6qQZkU8yC8cLz4gYBYMSYkrdlzmOZsOsDrFtRLIHZSuxUGyPbfMCg1sEsuQPFxd7sqjnZbIF2zMuIUEqRddX98FSIlrliRIFaFh7Gwwi6O3J/VjpD4nhhbfrAglWLDTgmCL4hIjY2FGBhLAD2REj28LjHgVRHOtA+kipMWFkwyQoTIDpQAyhAh5r3s5smkl6gtwrE4N+5QRSJBojFFDdx33hTKo0jJCyBoMGHL57WjJjfLDMBzwvsE/kakcefPX9Dl10oJopDZUCk+hpUH2CkMp9AIR0w7RThVmnn8wPiKhapbSFsfxlYrc1F9jIxiUeokIUY2BryQBk6ia6P1A3HtCaIbo3p7PDlPUbCENRz34AciRIu049sRbcFWmOjHavkMyWfKDFGBIIUpI8vY33///ZADpMjUqVOpX79+nAijEIAk8GHuB8mG+quNpqg2pkY4+S6IP8KH/6639S+KhHplA9GhUIn/aPIrMALJME91C7QWw1DVrPoDofHZTU14g8qIx0ct5bYYkCgCtAU8PGIJeyQo+AN4sDw8YjF1+WCaToD0qF+KxtzXJtMRIFB43zB4Ds/xdcsUppEDWrlK4fp+9laur1tXKRbxuWOMlzVtz0ah5qkgJY2b2RIJHQnYXJCo6Es1xZbE4kJkABNWSalMbWR7EgR44Yo61KBcESodm48XjpGkQVigye60WXYKdl9aEC72oRUm1BjVTVxgyqIQO2UiGRY/Dsh8ZafKLgFHpH3GIi6ccZsuyTW00ugkSPEYqqalI6xPPBp5MeCiUBNzVDMJYuz59mKMKoWcRP7pu5wGEsSsFvGzjSNO82cxGtHJuYTrt76tTUX6oHdDGnprswxdHFoRacGWmGN6H7u0FHk1jVRQSC9g3BGJZzhfEN0TJIrPuKjDjB5CZuw7ejKqdhU8H2nZ2GfRErNL82+IJrpQ5hgn7TBe1AaFHNy/HySIKDDdeIJ4NUYNGpqfSbWWK+zUCUln1xIj6kc/VHuiBJHPlBmZpRUGSTDGY+PGjTRnzhx67bXXqFChzBPFmdpA+xRqUsz3Py/Yzpf11HwIoOyw8nyzw3VNynH7NchFtMV4heyWQy0ezkOoX6uKXPefPHOeHh25JEUbDjb0vurXjGobFp7Y3Hns56WcSDKgXVCtAg+UJ35Zpqt5FbwBC3UQTZ3f+4/GLNnFG6Jd6yTQXw9dQp/e2DjTbaohyOOWr+fyfAJCbfidLV2N5Zgzf5oX6Hq4zYEKBJ9DEIDw0DQn2/00P3A/QIOysY7H+ylr9/J9QsGCDSKso/HdljkF79F1TcqmSKFJDSgSRGNwf7+vDc0a2CmiQQyAotHOgAyLUxROKBSqlwzNDPcK3WzUQTqMEAB2O2Oi0tiSdFwvmqyKYJHKGtUfsmOG4jFc9KK8NsYe6A0aK47XGokkwvTbqVu8SHbFHNXcT4kJCYBHixd5oTCs4totslzEm6VUgqQsCqWNI1olCF5/mVSd3Cfe156NylAjh3FZGbIvWJPESlENsik9TVsVFFLTF0SMRqPxUQg3Fvm5SNV9QbT7MkJ27aNpT9C9NMKQB0c9JMOkUIJoLZ5WQMurX0qQSEk0vkTkahsQct7RRs/bQYhoO2NcmZ+gzogWstlgrwQJzBHlinon3BQyDqDqfbhzdf759fFr+H1HAgs2H7F5t3RHsqv7wiYngLZ3GJx6QZc6Caz8BYEYLtkQj/f2dfWZxMF5fjolmIgowFjy7W3NefEnQAzsF9M2siebkSDBwvCZ0ctZCavgHHi9Fm49QE/9uow6vjuVI4nxEiK5ZOwDF9OXt2S+9hcA6Vf9h83ndR46GPA5cmOiDUINnyfMqfAQ6eigfeUPTX0lRKRg+vp99PnUjfrvHWsmuE6FkVaYkfO36+s0AN4jL1wZ+N6mNtRqwiPsdlzQ1wRggPNr991NO4xcR1QRZsgXRnbY7QohKEagEjG2YciuWKQvnS7JNbw2svuJhQDktLJ4R3+eX+ZtBXIHPU8sk2E8GtsJCYIJEKSQlRIknARdlCBeFxx4PYVzkvak4H1mLhbbtUO8Zo4mLUXxEVRaCgoZ3xckDAmiEa1upK12pGnYdhhp0Yvi+6T7glgshP1Qgjhphzmue2DljEKhkbpKELmt3Fc4yHP10qLkNh3GqyeIcQPALrFlT3JkpaLzx9JIkEyuBFFwjn6tK7IfAOrZ58as4HoWu/fA6EU7XN0XjBevaVxGj6b1Qihg8+yWVhKXuzmsGgVj3ss96/LPH0/eoCdOGoH69/vbm+vqQOCtCWvppT9X0TvXNQhJdhwxfzu9+OdKVwqY7Ai0FEFhMPC3ZdT8tUnU6/PZ/NpBcdC2enEafW9r+vrWZlS3TCCYIbMB8ckDfljIaxB8F+BjUkAzD3dKgDwxaimrYbAhPOiKOhFjhkEwz99ykBXYVzQItsJgI/jBnxaHkBa3tg564ER6n6au3auTIDiv72YH28xwSu9f39AXP00nUCSIRwQTYkILTfG5qJHgn7zRTTtMOGNUY+EHs6hwxpqsdpFoWK2YDsbjRiBB9HaYwPXxIZdYWYk/hQzKeB7RxOOmiMk1vU6SDJPXIwkCk1ghhNDrLP3ORk8QeR2tYpOjbYeB6kbeCyGHgjvGGdcIzl8lSPRxiwoKGUMJYj3moTiX73U0O+hBI+XTkZUgUTyOnKNVQoyVb5JbiBpCElPctn/6odAQ4kLGXy+QTYNkV+kwHj1BTHNvailBhJiQed0Mudy4yPMKIdLQ0mBFWIkSpKxSgmQZoOZ569r6XHf9u3ovqy+gagWwq7/KZeztU5fV5JoYhMSvLkkUQZ9m5ZmcQKoLjFLDAUkwl9cvxbXvIz8v0WtQI0DaTXj4Et23RNptLv94eop0ym9nb6XX/1qjiBATMB78uXQX3T98ETV5+V/qP3Q+/TRvO7dUYP5A6s4vd7ei725rnikV0QBIj48mraeHRy7hzxPSktDG42aTHbXF/35dRr8t3snfrY/7NHJkYvrV9M38f8tKxfRNX6jR8XobPQo/v6mJPvdEwuyN+9lLExu4DcoWSRGLC09OpzG7fkCRIB5R1KbYWLNHI0FK+keCxOTxxxjVWFAs33koogGfOSFGb1GJ8GEXU1W5HYpksLFgH2XRXlGTAooixQzZxRTFjROI+kWUHwJJhsnvIRlGCCE57xU7k/m5YCAxEhDyOlr14cvCprgfEvejATWKvBd+GM9lRIgniMTk+dEmoKCQkZUg2KE/q/ve+NEOE9kYNSoliN4meCrFTo+YGBtbBt1C1BCp5QniJLVFVz/6oAQJ13aTQgniuR1G5t5Acp1ffltmyPtqF1u745B/xATeWyH2hBQ3Yqt2mVKCZC1UTyhE93Woyj8P+mMl7+C3rBzHC6j+w+aFLJycjFUPdgrc15sT1jhqTTMDyqnbLw54KCAGd4KWrmFXM77Ssy4TxVBBv/HXasvrgey8t31Vmj2wo34ZhCrioWfE4Gmb6P2J6yg7AwTujPVJ9OV/G+nWofOY+Hjgp8VMkmF9hPrw5pblmfRY+Oyl9OENjdJ0Qe035m7aT90/mk7vTVzHPhl4blAKuWkJBwHy1G/LmDzEuuWjGxpRt3qB2Odw+H72Fj2y9tY2QZXH23+v5aQm/fdr6ztW12BOEj8SREOj1oGySgCTV2mFSytk+4hcr9AX+seslSAYwP2CFHhGhQMGAyz2sdNm9PSQtBq7olCUGGKsGS6KUS+otP5iMUGLVKCb5b87NbkqmETxEqmkOYCLQasZeyxaTiJBpGFm75STp6Nrh5HzwKS7ZEeAPMJga/RF0Y1Rj6WMTU7ysc8f9y/vA0ilaArZzOAQj6IXr2dWb/9RyPqoqilB8JkGUWDeyRGFWbGYvFFFI4oqzc60MiRuOioliHg1nLQcu7FrGokwd2ZcGi4dRvPAciELDt5/Su8qM4TI96MdBvO1MfLbDBguSiy8V2NUaYdBch3aQq3qAD/aYUThYacE8btFBcpAzH1oj6xTOlhwY1dSiHIhGRWyDkAQ/LV8D61NPEKvjVvNXg7XfTGL1iUepX7fzKNf727teAf61taVaMS87Rwf/tG/6+nZy2u7Pp/Hu9TgDRksKLH4HnxLTtsdddTPb13bgM8TSg6oO0CiWHnwYXNy+pMdqN/QeWGNsz+avIHHj/s7VqOsDNR8iAteufMwK29W7Epm9Y8V8VU5Poa61CnJLSJQFkRq8cgMAFH9+vjV7GMjHkzP9qjNyhY3RuYgQJ4evZx+XrCD20wQlIBEnEiYsGIPPf/HSv4ZPjVd6wQSXP5ZuYfJOEGf5uXpuqblHJ8PPHL+XpnIa6frm5ZjUg+xuAKcn938mFpQJIhHFIlJqQTBF3ezNoDBANQvSAuLPBZSQtq9PZWZTxj9oM8txc6YjSeILC4F4SSxUlCJ7Mlp1KvZnG2nhTRWyJgtNkoQuY0bOXUwHcbUDqMpQ7wkw5il39LfaSZn5HVEZKS5+Iy2HcZIsqDdxqgsyQoDvhXwWcFTw3uHz50fr6GCQnoCn11IsjFuY+FWzUSUi6IiWjNJo4cRFolWhEq0sd1GI8rtWjuC2Q8EagEvyTMpPEFOnYnsgRWVJ0hkJYjMhV7JHLwM2MnD/dm95nIeuK4XUkfmQLQQgARBvWBFghw8dsaHdhj7xBYod4Q8KuNTi0rFYjG0eNuhFPXC9gPHWT2FDY6SqlUyywELojd61aNrPp/FUv4rG5amYf2b0zWfzWK1xJ3fLaDvbm/uqLbDfT1/RW26deh83uG+oXm5FGkXkYB6681e9ZmwhPoAHg1I35MIXzPaVS/O8v4v/ttIr4xbzZ/X56+oYxksAJJk3AOXsHElnqsd3vlnHZMqn/RpxDvwXlRwGQFYL8GgE5ukIFNR86O1DWp6EB52nlZYOyCppH7ZItS5VgnX72FGf01AsL02frW+7gLRgHYup2SfkQB5ZswK9kVhn43eDUN8Pewwf8sBenDEYp6v8NgPdAwoqFCzgPgzpsEMutI5kTh22S7+7AIvXlmHJq/Zy6SeYNAVtdPlvcyc354MAGnTMHpAYOKXnSk/+1OlqN2nFcn4kErxN2nNXi46pGCM1CMt5IMgXBGsm79qg5EsRCMpQfSIXO1LbEWCiDP21qTATr+5WA6mC+R3rQRJGZF7PmoSRFpf5m4+YHleKD6x8wkSBCyuvP7oBUUvs19KEKhKdH+MLEwIoGDB7gg+O0gykr7vaDwGFBTSExjjEBGO1AAkxJhJkEQfTFEBqMNkIYwx22oM9SMdxuzbI5AWx2jnQCdKkGg8NJx4gohKJBolCBZOaO1BwR+OBJHzwHW9ktvi5YX3HS0xZk8OKJCkRvCDBME8jWLbeL4y30MF5Je5ndQtovoQyK456omsuiGQ3QEvh/6tK9E3MzfTM6NX0N+PtKVhtzWj6z6fTfO2HOB4WXgcOHn/29cowQtn+Iy8+OcqbptwS9SCwMCCErUeUi7u+HYBm5zatV3877Ia/F2ApwfIi52HTtJHfRpamloiQvfd6xtQo/JF6KWxq3gMtwK+370Hz+Gfa5YsxAqIBuVwxLIKPRoloR8A+Y76HxuhB46d4XEC44KR8ACBGq7FH68z1JN1yhRm9ReIj9qlC7tKQ8lMAKkHAkzWGPCVfO2autSkgvt2Hqypnv9jBbee4Gvx3vUN6SpTuosV1ice4c8zFIudayXQy1fV4e8HSL97hy/kz7zMUZ/d3MSxL8mibQfp0Z+X8s9QQ2EjFz4nAiTd9G3lzFjVbygSxCNkZyjZYA4jkz9IgmgW3GZIUYzFNJQeS7aHRoSt33uUGmumP5GMUc0kCHKmI6ldhJGUPm+Jx7NDYZt2GOOuEFhvzD14TmB8jcQKiiqrGFqnSpBjNkqQaNphKmqvm7RZNypXJOTvGCig1tiVfJKfjxhbCXGUL3eOqApCkRWj/1l68YtnUVNUAXx18J1avjOZ5bCByzJfrFlq4cCBA/TAAw/Qn3/+STly5KBevXrRhx9+SAUL2qvQTp48SY899hiNGDGCTp06RV27dqXPPvuMEhICzvtLly6lN954g2bMmEFJSUlUsWJFuvvuu+mhhx7S72Pq1KnUoUOHFPe9e/duKlkyIJtUsEbl4gU1EiSlAi7RJyUIFgMgN/DdAbFiHkPPnjuvt8pEQ6SW08akQLF7RicKIDc3GsF6hYyXYUkQbb4Tfw9PEblh7l/Ul0LsewV28YQEsYNT43EntQnmHWNtYlaBYIHh5TUTJECFeFGg7QafJaNh9Y4D/qe1QAkCmJUgm5IC3gmqFSZr4/Gu1emfVXu4zeqdv9fSoCvr0Jd9m3Crybjlu7lGhsrDCZ67vDZNW5fEsbT/rErUpf5uAJLhkxsb0Z3fLaRp6/axuuTHO1owEWEGasMB7arw9wEmqTCCvGHwHE73sCKhcf1bWlWk2qVj6d4fF+rzgh2gnMAxcsF2vdbE9wXjCL7jGOdAKIP0hfqtkPa78fI8OXNwK94Zw3H67IWQ3/Fdl5+PaGMZNoFl/JefcTjxLxSgbsYcBcK2VJF8rKKvWzqW6z8/11EZFSAY0CYCtRBeY7x/8MUAWeCFzAIBghSkH+Zs4zUWPETEVDgc9iSf5O8T3svG5YswsSjeI6+MW0UrdgbNiD+7ubFj02uon+76ToiVEtylYCRAMLfiHNOLxFYkiEcUsVCCSB+sXxJQYzGIDw5UJmDQzJFb6/YcCZIgETxBzEZ44QrVYAyw5gkiSpBI7TCmmD7pGzYW4xjcSus7/cdCSBAUVRgM8J1AseUU0gJkduIOtsN4Z8eRgT3oz1X671byx7iCARLkwLHgpLXvaIDMwe5fNNJwUc7AQ0Vex5KxWVcJAtQrE8uSOcjooObB+yeGqQpEN910ExMPEydOpDNnzlD//v3prrvuouHDh9ve5pFHHqFx48bRqFGjKDY2lu6//3665ppraObMmfz3hQsXUokSJeiHH36gcuXK0axZs/g+c+bMydc1Yu3atVS4cJCUwu0UwgNKEABKEDMSxafDB3ITXh9CgpiBXmuY76HwjcaAFXMMCHGQ4yh0YjVzNNmdj3ZhGozItScO5G9eCGZRmoTzBNHNwKNQTQAgiLbTibCPJWRMtDuduhLT4rGC5uZ5opqPUByj/QTzHT5nRhJE2qP8VMOK6kgi0wX6Z02bHxWyJqCaeOOa+nTz13Pp29lb6IoGpah1lXh69/qGHNUJlQgSi+64pHLE+0JL+J1tK9GnUzbSy2NXccuKl8U2dsG/vLkJm7TC16DvN/NoxF0tqVYp640aeDGgZsNO+7IdydzSM6x/M9sWAET7/vnAxXT/8MU0T1MHCEC24Nu7xLQWwJgGgkICGtITGF4wlmE8wliBdREWzjiY9Cian9cAUL9kV0xfv4+eHbNCH9c61ChOL11VN0U6kBsChBVOs7fy6/9Wr/p0TeOyEW+XfOIMG81iPMe8/XW/Zvr7gmheECqCx7tUp0uqFXd0PlA33v7tfK4RapcqzJtAxhYY4LWr60Wtfo0GigTxsdAQJUhZH2LhzMCHBDtsu7WdceDiqvE0Y0OSvkse6pZvPbAYCx8MRuEKISFMpDjX+8gjtsPkCVWC2MTlYWGPv6GQMUoJ8UWU5+zGBTm/jTHqKR+UIFKECawikCUhZr8hIcYP2blx8YTXSvxB/IxhzoioXzawqEIvuDxfq17a7IjVq1fThAkTaP78+dS0aVO+7OOPP6bu3bvTO++8Q6VLp+z9TE5Opq+//ppJko4dA270Q4cOpVq1atGcOXOoZcuWdNttt4XcpnLlyjR79mz67bffUpAgID2KFEm582UHKE9wCA4fdhdzmBWAIgCwVIIYxr1okaCNN1a7iOLjgF23aHdfUKwJCSIO8frufHzBVG+Hkb95MRIVxYXd/cuOJxCtAbWTmNygEiS6sizWlM5mHT0fvaQcO9uYq9EOZYyfDJqi5vddCQITXuycyqI1SLj558GmkDFxcbV4uq5JWTaL/N+vy2ncgxdzFO2e5BP02vg17LmBsdOJ7wEMV39duJM/q0OmbaIHOnkzGsVC8at+zajv13Np0bZDdPNXc2nkgFa2noBobfjt3jbUf+g82rL/OBMhMHttVaWY5fVRN0Jh8sZfa+jrGcEUDWyEgiR5tkct+nLaJl1x3KpyMbq5ZQW6oBHEGL/wP5Kp5HeQrYdD/naGo1dz58rBygOQ42inxM+ov/NoP/OB6+S4iAlwrIGw4Yl6Hz+DWI3V/sflGF9VvZYSeK2hBvp21haatXG/rv584Yo61K1uSc/kNAgQtFBJosub19R3ZFp66uw5GvD9AibOsFn7bf/mOum/Ye8RemjEEv26UHLgu+MEUJyCwIOJMcfhlosNMVUFEPfb3UFSTWpCkSAeEVRJnEmTvHrs7IEEwZcGUjO0foBZBgmyPvGoq8jAu9pW5gH1gxsahn1MeR6YKNCiIrtIkcz0RBaN84QEyk4hg4mCz39vKGstyhG3/g9C/Jw4Y2qHkXSYKBlnmAShx655xTjLxUOcRowZoymdEkdOSBiMjXhN8ZoBNW12HLKSEsSImqoVRgeICRAQQoAAnTt35raYuXPn0tVXX53iNlB5QDGC6wlq1qxJ5cuX5/sDCWIFkCdxcSn7Uhs2bMikRt26dWnQoEHUpk2bsOf8+uuv04svvkjZGaK8Q0yu2QtJxkmQE9FCjJutlCA7tfhS7MJFCyizQFKKLwgKH/FtqBSlEkSIg3DSavlbVEoQG6WJEAYY6qNtUdFjcsP5j2h/80sJYlSpCqQNKho/EAF2DOHJYCb0JPGtvMmEPRrAUwEHWnORwAdTRNQkq/cEiNRqCYoEyQ5AQsaUtfvYP+HTyRvo0S416M5LKjOxi8Uf/EGgbrMjFQSojwd2r8kLvE+nbqBeTcp6jvPG2DO0f3O66as53DKA/38e0CpFCIFx8w9ECExdF249SH2/mUtvXVufrm5kvWMP8gEtPFB//O+XZbqyGbfFpigiT7GoxvNHe8+09fvovvZV6c62lbNFO0lmART1I+dvp+/nbNXneswt8MJ4rEt1T75WRmIFaTJDZwYIkDeuqUfXN4tMgJw/f4G9OqBkwucYyiRRoSBgYsD3C0PWg1BeOdk4EUUKWsWw+YzP7k/zAu1aAmyKD7qqDqU30tc9JxMjaP55mj9Idt4XfkF2B2HEBGDXDYZIgFEJclRTQcSEcZd/smsNWvhsZ2oWIT9bvgwgJZKOneIvmpMCCn2IUtuv2XOYjUrBCJtbGaRwgaeJJQniclISpUeKiNyzgd+dmvjY4YUrajPzjl5QK4gSRCIIAb9STXDuQkqJWZafMcwZEZBYG/0RapbK2s/XDfbs2ZOi/SRXrlxMVuBvdrfJkydPCvUG/EDsboN2mJEjR3JLjKBUqVL0xRdf0K+//soH2mbat29PixYtCnvOAwcOZEJFju3bQyfF7AB4MrEX0smzOkEqRYMQCbLrHQ0k+hatL3ZKED/mKRnT4VUEbD94gtM60LpWKkpFi5AUmD9ArlhBlBpeCkgjySJzuJV/BnY6o93R1GPjLdQZAuzW+uUJYmxjtU4gyucboWdu7QJJAVT3MSEPZKEojUQJu3n/MX7/8VnL6nOhQgBQGrykLZw+m7qRVu8+zJ8NkATYRYevxV3fL6C1DtpBoCLBhhZabV8dvzq688qfm76/rQWrVaG+u3HIXMs4VwFqaCg8etQrxfXcIyOX0seT1vM8EO58x9zXRm+NBqDE6jNkDjUsV4SVMc0rBZ7PuxPX0WUfTKMpa/dG9bwUogfWQAN/W0YtX5/E5rggQLB+RGrQtCc7sL9NNATI7I376YqPZ9BXmlII7SU3NC/v6Lavjl9N45btZuXPl7c00ePH8Tl8dvSKEIJ78C1NHRuEg5AD2YNaB5vdsm6FegnA5e9d3yBDmNwqEsQjAj21xL3VsuNi1/bhByT+TQgPxBNV0YoMLLQlFlb+D7czBombtKyEA7ej5AikDCCvG8BuTCSzHty/LNinrNmnF8vm20nhYlSyALpyxOXraExkMeLE6fO+KEHAqqPn1Nj/bEQxzTD2gGU7TPT+HUZ5OV6baBILMgtua1OJyS24UXeokfU9J5566iku6sIda9asSZNzWbFiBV111VX0wgsvUJcuXfTLa9SoQQMGDKAmTZpQ69at6ZtvvuH/33///bD3lzdvXvYQMR7ZDRhDKmjEwbo9wXEPYzh2+LDW9mP+kHYYGX+sxlevO59GiOwbixEAO7RC5ETbamNscbFTg8jlQpi4gYyfWHdYKTSC/hnRj7M6CXIi9ZUgQriLkXlqJBABVUpo/jaGTQzUH/L5Mqcf+aUMXL4jQIIs2xFok0Thnt5pGAppB5AdXeskMNn6v1+XMUEqiS0w+gcxBn8DSRi0A+bSF66szWMuFoKvjltlSYY6BVoIvr+jOfvTYC1w05A5epKf3VwA88kBbQM+JiAu8HzQhmcHmIX+fn8burR2wMhcgOjSn+fvoJ/ubEkf3tCQ60202/QfOp8VJ2hXVEg74DM5YcVuumHwbLrsg+msggA5Bb+YN3vVozkDO9FT3WpGZR6NFDa0sYAEW7X7MNfIb19bn25s4YwAQRvY1xpxAmNSo88hYnWNMc0gLJDM4wST1ySy146sQ4W07lI7gRWw0o3QonJ4tVZaQc0cUUR4iqkc+lTBnKWWMSpgXnhDXsTuztrCH7t7aD2ReNho+4oBTCxSKC/WDJicGulV0hbsk9YEGEAjey2ophXQmDCMRS6+0IDb3Z3UTIdxAlHIWLbD+ECCVDdIfjEZZgfAVX3li11p8fOXUsVsYH6H5Bb4fYQ74NOBFJa9e0N3ec6ePcuJMXYJLbj89OnTdOhQqJlaYmJiitusWrWKOnXqxAqQZ599NuJ5N2/enDZsCDW8UrCGGOet2h1M+RIlBcZbzC2p2Q4jSrsyPrTdYAcSAEmO3uIFWwMGfohsjBZY2IqZtZ1vh0TkemmHwesst5MENGv/jOhbR8yJaanpCSIEh9V7D2N1v0h5UYKg/UVUohv3BnYO4Vvlx+tm5RElRfVSLSXPj8+aQuYByAuYR4L4hMGotACAVBjStykTs6jJQQCEaz8TAm1gt1r885Dpm+neHxel2ERzA/bwuLMFbwKChLjpq7m036D4MwNE8cDutejlnnWZjPl5wQ66bdj8sGbQqPthyIqFq7H1HuawVZ4ez61Akx5rR3dcXIlreOzCd37vP3pv4rqIxJBCdDh47DSnvLR7eyrd/cMibjPBe9C9XklukRr/4MXUu1n5qNqU8Jl+bfxqfk//XpnIn5ubW5anqU+0d+QBImanon56unvNkPhckMwDf1uu/35LywqOzFVlM+SB4YtZHGCct0vH5mNVEuZZjOOPXlqdMgoUCRIFhCDAIh4LXxQ4UIf4IWc2wxybKBO/9I9jcJPFN9QbfsmMZJCdqflQOG1REbd2TFJ2JAjUKEIOIJ8a4D7fXQESBLngbiC560IEmY1Ro0mHcQIp+hCR63c7jBACkG+KKW52AQoFNwa5mRnFixdnn45wB1paWrVqxWQGfD4EkydPpvPnz1OLFi0s7xvKjdy5c9OkSZNCEl62bdvG9ydYuXIlR+D269ePXn31VUfnvWTJEm6TUYgMuKQDq3cHJdvio2GOMI92IQzCwyyxDioWo38sKPww7kGGvmrXYT3FALJsPyAyYavFDEgXPK5XY1SgqGYQaiSurZJU0kYJ4k86jLRCWZEgcpmdmtENsIsJIunU2fM6sSatMKnh0VFPq3nQ6gBz1MXbDvLvMNxTyF7A+PZM9wB58e7EtbwrLjUlfA1A8sHoccB3C3mcCAd4Z0A9AUPQCSv30A1D5uh1mxeUis1Pw+9oyepttHrf8vU8XhyHAxaaiMzFRh6ie6/7Yrb+nbKria5tUpamPt6ed/+NaP5qoO3i0S7V6a+HLqGWleP4O4pY0lavT6brv5hN38/eovsDKUQPLP6f+jXQ8gITW8yxUBDe274KTX+yA312UxOeE6NJ5IK65Ic5W6nD21PZYBQK/UuqxdNfD7WlV3rWo2ION6ixlnt81FJdaX2nIVHpj6W76IpPZoRstj97eeB7FglQPd0+bL6eUGoE2rZwvmj/Gn5ny6itCfxE9lhZpBJkJw2DlfROQcqcGmZExihbxBIKOSFEDM5hvxbNClmeX5nL5TS5FkyYgPoms0o7mEkPu7hESTgR9Qfi9Y6cOssTkp3DdiQliLQEpbUSBO8LYJxc9voYewkVzsgBLVlK179NxajvTyHzAokul112Gd155500b948jrhFessNN9ygJ8Ps3LmTSRP8HUAk7u23306PPvooTZkyhQkUxOqCABFTVLTAgABB+wuuB68QHPv2BdragA8++IB+//13Vn7g+g8//DATMPfdd186vRqZCyIrBWlgVoKUj/OHQC8Xl593iFCQmL1HgpHl/2/vPMCjqtav/9I7CSUkgRBCDS30Lh0EBEGEa0GxX7uoXPtfsd5rL4ANy716/VBRVPQigvRepHdCJ5RAaAm9z/esd2YPZ5KZkHImmZms3/McQiYzyZxkZp+9117venM/JmFS19zlBkFotylVsE8EKerh+LBidYeUzSQDKzOMwOFtkWJuy3snSDFbBDBvpVB2OkGww2k2O4z4YfK96vpo+5kbsJuIhSXKIGDjXr0nTV/jbWsGhq2a5C03ta6u3VBQZoCdayP2Qpz76q7W6vJatP2wPDl+zRXLXLATPvbvbTWrAZ1Xrv9kgXtjLqdB9nCEYE6Iue3VH8yRX1bsyTTzo3v9SHULRLgEnN4j58p/5u/ItDwGm0PY/d/6r2vk9vY13Ld/tyRJGr74p8zalKLtTpFjZzbQEGY84rf10uZf0+W2fy+RH5ftznRcIhmBcx0OmxG/rpMu78ySa0bN0xISiE3YvIUwtei5HvJ0n/q2lJ2ilW7f0fO0nS42WbGe+urO1vLN3W2y5QrHnANhpypINInWjENcw1FF8PL/1mu7aQNyQsYMbZElwQLuKZRdmc6e3q4VEC3xOsyJa9OfUATJBVB8Aax3ptbJX63aYKFGkNKdHeLkrcFN3IqieQ4ohzGtWU0LVTtI3+nG2J+zK4J4c4JYg3JgGwPrXQuDelFls13n6xZB0gWjGhGkhJ9FECN0YKKJix0uvKYu2w4nCMDfHVb33CjKJDT49ttvVeRA2Qpa43bs2FE+//xz99fRCQZOj1OnLtcDI7fj2muvlcGDB0vnzp21DAbtbw0//fSTCh5jx45VZ4c5Wrdu7b4PSmpQtpOQkCBdunSR1atXy/Tp0/V5kKyXwyBQEjvaIMm1k2mXEwQTFzP52nno8t8fYxOccpo9YlPZZvNY5zXhnT8TdYGKxapdHdJMuae3chhrKUxORX+3COIlRNQEXJt2gXaElWYtE8SechhsJphucYYUGzNBgAkrXbrTuUmyavdRv4VY45p3Q6sY92sNdKkX4S79IgULvB7eHJygDl8IsFjMW8tcPh3aQl3RE1fvk9v+s+SKuRgQbn95sIPEVSqtpe2DPl0oC10O6JxuXGLXGwtWzAPRheOmzxe7BUNf7yeEn6JkAGMeWp72HTXP7cTOTAxBidDCZ7t73A5HSKe3Z+n5QBha9Fx3XfgiUxC6EFwnT/+0Rlr/c7r8/b/LtEwi/ZhBnA71dXvT5ONZW+WmzxZJ81en6qIf4Z+7Dp/S1xlEhfEPtJffh3VUYcqOzXBkbKE8Cm4itJqFo/Dl/g3lz8c7S7f6VbK1DkCAKrJyIODAHYScj8KFC2kVwYCP5rtb64LaEWVk4rCO7vXllTvMrFJR2hsQAsfe01YdV4G4bgksSSaIy2GMwosXj79AcI01vCa9G8UIDSag084Ju6FJFq2nVtEDimIDH+1NO9SuJKNmbJFF2w7p79DU+zaKzr7F1SiMJ85d8Gg/eSaPnCBmYglVFbVvlxwOd620nX8TQgA6wXz33Xc+vx4XF5dh56lkyZLy8ccf6+ENtLrFkRlPP/20HiRnoPU3dhzRXh07fhCWt7pEdEzA7QJjMCa/Ow+fdDszTAg1SjbtsqT2b1pV683NZismg3ZNdjJzUOSmPa7BuDy8iSBuJ4gN5TCZ/Ry7nSD4fZQpXkRdQBC9aprrIkQRl1XZDicIwE78T8v3yOLth9WBaRyjV9X2T7kmuh5gIWJea6ivJwUXtKF94up4zTf456SN0jy2gjtLrlPdCHn/pmby1PjVsmDrYen1wVx5qne83NEhzme3J2xion3tfd8sk2XavvYveXNwEy09yQl4LlMe6yxfzNsuH87couWCEDXu6VhTHu1R1x3mbwVu8gkPXaWiDsQ+uKuQLdKnUZQ836+Bu2ujrzXJptf6qDNmgivYEmV9KNFACQUCKVF6g4B/lBBBIJq4OlkbLqDNLg7Mk3s0qKLjOkTGgtpmF45uODDmbj6kH9MHTWPDonPdCOlcL0JzWOx0OKCz18jpW7T8BRsLEFlua19DHutRN0tNLaxA5PjXpI3y+5pkt/v+s9ta6fV/wdZD+tqygtcm3idZ/bu/OzVRJq/b73OD5JNbW2RJTMkvKILkAqsAYcKUrGUreYHbCZJ22l2GUcnVqtUOutf37MiR1bIODORmov+v6xN87qY1iw1XJR8DDJROpHTn1E5t6sKx7sNup7tbzHlXdxg/D+aoj4YLB3a1/Wln5MKlS+4JJ9PrCSEAAkGrGhVk+sYU3Z1BQLTJB0ENrl1gkjZvC5wgl9vcbUlx/pzslhpm/nPKyBO94nXCjgkW/m8XmWVpGHdITvNArK3uvQWjHnHdZosTxPyck+c9BHp/ZIIYQX77oZOaAWI2JEweiIokNk3Y29V2lqJg82LmphS1WePab5ejKT343uiqMDvxoO6ap++SQQoeKA3+fc0+3Ym+7qMF6g4xQY9oK4sSbnRdWbLjiDorJq1NVje1rzEQgiVKY576aY2KBMhPgFNv+NX1ciTuYl74cLc6cl2zqvLqxA0ydcMB+Wzuds1fePHahtKncVSG7wuRZkibWOnbOFo+mL5ZHQfIK0G4JFqr4vDV7RCLV+zy4/2BAE2zEedNDHmke109kLOD3yGeE5wNWDDjgBOvV6Mo6d80WjdgQ3kei81LiLhzVfg46HalGyAst69dWbrUq6zCB657doPSJwgfEEDMNa9H/Sryf/0aZHttiSycL+ftkI9mbnV3nhvaroaKhvi7jpq+RV9bVuDY6Fg36wL2Z3O2aatqb6A864V+DW0JevcnFEFscIIkp55x7xr5yr7Ii+dggtzsTGWHXQpvjLu+/ktuaxeXrcehb3rq6XOqyPsCamTruIpqyxv+wypJOnJKB5trErx3uMgMiBx4o2PMh6XPTPTOuASqvFC0MfmECIIJpynDscsaTggJDTrUrqwiyMJth9SajIkqFnh27piYgG44QQymha2dIgh4sEttFUBa1Khg6zibuQhyPsftcQ0VM8kEwW6c3scVnpobzDUZQa5Wgd4f3WFMOKoRQTKGotq3SYLXLMJxcd02YXsI6/On7fm+zrX1IMSUgvznztby2LhVMn/rIf24FLkX1zbU+SW6yqF17Hd/Jckbf2zUhS7yFbCrDjHA28IeY9iom5pJbMVS8vGsbTJ65lZ9jb/1tyY5dtAhq+Tz21vJjI0H5KX/rVeX3oPfrlC3xSsDGnntfhdWupi8PKCRCiLIbEDGCZzTcF/BFYJ2wd7ea7gNO/roSoJFMM7dmEKNGAJHFTY5ezWMki7xESpeo2sHBE2IPxBBUOr/84o9emCchbOlTkRZbY+NawgW5jgvX86aQAXXE4SR49oIR8yq3am6IZE+1LNxtfJut0eL2Ap+W9DjWoOMkU/nbJPtrnxJXE8RSprZ+skXyIJ5ZeJ67VAE0Dr6lQGNNY8s7dR56TVyrnsuAHo2iJR3b2iSZZcJXH+d3prl0QTCUKJoYXn9+gQZnEP3VF5DESQXmAnrfstEI7ttXe2YhJiSHOMEMQGddgFlcPmIq7MdPpfgaml3JWDNgwhiwlFhwzOdXrIDBn5MLrFDiHpo42E540oHL1Xc/4ok6pNxHnhNmJr13PQCJ4SEHqasEZN1k7Nk8pHswjgAdlgyQdzBlTZ374Do3dMPu/LGQeGvcpjwTMpUTMaWHd1hINBjcojgPCxC0osg2LXD1+woh/EVjmpEkEgbQrqtXN+8mi7MEFAJrklglyiSt6Azxn/vbqOvQ5SdjF2cpC2UYcVH+QjGJ+yCI0fh/35ZK3M2H1Tn2h9rk+XtvzXRDJH04DFP9a6vIt/zE9bJr6v2afDj57e1zHZJgpUeDSJ1/P9k1lYZM2e7PhcsSiEkP9i1tlcRGeGX393bVssOUNqA+T7a+aKc/KX+jXyGY2KNAif2/Z1ry+iZTvHEgHnyb6v26YFGBB3qVFJBpGfDKvJ8v4baOnh50lEVRPB7glsbApIpeTNAGEBAcm2XKOIUR8ro//OrlAZuOzzfpCMnNRMLweMQOyAKwNXjzfln1k4QHTrXqywd60TYluXnq0xl6voD8uf6/epScpfOlymunX1ualU9210RcY6v/b5BN1iMC/3/+jZQFxLWR8g1ufbDy91fwMibmrm/npXfK0pfP5y51evX8V5BFo+391OgQhEkF+ANgxeZSVzHjh46eOQlaJGLejFMoIx9q6KN5TAGu1ruegNK+KAW1eSXFXtVeX2sZ90cfy9MiDG4W8OdTuexEwSgHMbsJNIJQgixUi+yrF4rIFybCUWrOH+JICfUZosdz61+7N7hDzJzghgRxB4nyPkMkz0725tjggkxBeI4SkSrp6v2xO4cwIZqTjvdeL0OWTZoYHM3E1U7gdV/dmKKliP8vWNNvZ4TktfAjQAnQ4vYcHn8h1XqaMCCD6UhEB7MpiFa6CIv45WJG3TOjBIalJcM61HHq8sDuTNoJ/7g2OWa6THok4XqPPHm3MgqmIv+o1e8DGxeTV0h2ASEgPPrqr3q/OgW71mGbsaQvgnR+rUxc7bpgUBYuFpQ2jK8Zz11jvjqVvPuDU1VZEGpBYQNK3CoocQMx/O/OhsgQBBBuRkCV1G2g3J1ZFdtS7n8EW4zlJEg2wqH5/N1/r6NY8R8RKmccY7gX7P4NktwfFrIfFbIetvl34P5Eq4LTjcHRA6n0GE+emvVagXjOjK4UNaCklQIU2hfb1dnTW+ggcaf6w9oaRO6EFmpH1VOW8jecVVcttdbWON8Mnurllnh74E1IZxAw3rUdW8S4OtvT3EGSgOUxEwZ3tm9kX4llu08In8bs8jn17vFR8jIm5r7fA0GKhRBcgHejGjTZSaxqJ3LazCxxWQXO3xGBAnGEM53/tZUe2pjkMyNldYdjmrpJmDKUvJCBEELPyOCGGcOnSCEECsY4/7eqabakgFcAunzl+woh4GIgIkidoCQFQUXAiZIeZ1dlVPCSzmvZUZQ9poJkgsnSAV3Jojn98fvDAsDYNduILJFIEqYrjNWzG3YYbZjEm6Ed1j4DbB/mwWRnWAneNx97WX9vjTb3UyEZJeu8VVk0qOd5OFvV2iZwz3/XaZzSwgk2FnH2DuoRYw6nF/6bb26Kz6atVV35FHugrKH9OC+Pz3YQTt1YOGPzjFf3N5SWtbIXStwBLGizekfa/fLq7+v1wX8XV8t1RDUF/s39NpeFVkgyCdBWCtcIVhQo7MH8jwQaHljq+o+y1Mw7n84pLn+Pj6YtlnzSaxEli8hB46dlZVJqXq8NWWTlvhDDIEocm1CtMf4BPfC3qOnZevB47It5aSK7Oh6BpEEYi9KfnBAXMlrsIyoGubMJ4LQ4RQ8nP+HEGxXLlJmQEzHugyvrSnr9rudmOb54bXWu1Gk9G4UlaOMEXx/qzvIlCPCHWRKXtEYouNbs9zrEYC/P8qfimThWgMH4d/GLJTdR5zf32DyD3EeKC17tHtdvwpI/oIiSC5BrZ4RQfIrpAvWZuuby84WuXkF3ox1bNidNAOb2SUEZ/MoGBVEhTknzJjsQggBdrWiJISEDtipQRgdJknv39jMdrEUExLkLSHxH2U3ZmzExMtXqF6gkVl3mMsiSM53nkzoaXoRxLhAICLZ1UXHLbh4qaM2ThRzH/tcQJfzYEw2jMmKsRO8nlrF5W5BSIhdYHf7x/vbazAoBAKEN65IOiqjhzR3h/vj46dDW8rktcky4rf1Ooce/OlCufuqmvJkr/gMYyRKTiY81EFFFbhMhnyxRF0m1zapmqvnClEGHbWQyzFq+mb5z4KdKmwgoBOLy7s71vSaW4IynzG3tZT5Ww7JyxPXqwCBrjDfLUmSlwc0zFSgQddH5JPAjYDyBpTkmHEIOSIo64cIgswqZFR8Nme7HnAvXt3QmSOCjijYWISoiqN7fc/FOQR3pyhiEUdSTmgTh3RN63IMBH38HiBuxLnEjbjKpSW2YhmpXrGUbWN3doAwBNcEHB8QP4w4YZ4vfm8QPXo1jJQquWhVvuXAcf27o/ORec0jBweiitlI3rT/mPQZOc/jcZMe7ZilcpXT5y7K+9MS5Yt5Ozxuh0CH+QQEELRzH3Vzcy0zC1YoguQSKLVjhrbQCVle54EYnNbm/R71kQUVY402Ighs4GZHr3QeTPyjXDkxUE/3HHXuvLEchhCSHkxsEdiH7IncWKszo21NpwgCG7eZSGcn/T2/MZkgqV5EECOMmJKZnGDyPlAjbu3aYkpc7Wolqz8rs3a87hDW4raKIHB/YFKOTQbjCvFX5xZCAgk4lFBagjLDZ35aI4u3H5F+o+fLR0OaS9tazq5GJsMGC1N0jkFJ9r/n79CQSnSZQYC1FSxaf7i/nYav4j6PfLdS31fI88htGDAcbcjiQKDkiF/XydKdR+WNyZs0x+O1gY2lneU5W8F4PvmxTvLNol0yctpmFWgGf7pIBjWvJs9cU99dGucNdCNDlgoWte/+majZFHCloNMTWgn/c2BjWb0nVbMrELYJN8H3f+3WA/PprvERuvkLsR2uC+MEwO8C6xAc1t91ejDmOj+6Prfe5r4P/u95PwOuaYEQyopMp4VbD6vogdeFNTAU3S+71qsivRtHSvf4yFyXiyAQfPSMLfLVgp3aPhevc5Rz4TVoFe7em5rokd0B4Qsi3pUc8ZcuOeS31Xtl+A/OoGtruQ4CVFFWg/gSfL/Phra03VmY11AEsYE+jfM3CMwqvhQrUsjWiVuwUcZVT20yQazZIHlhf6vhqre21kdmteaOEFKwKFeymB7+orWr1bgJSgs2EcSdCeIlyC7ttCkhyY0TxPlYCAWwbxuhIuW408VnZzBeZp1oTGe33AQuWsGCBJNj1IfvSz2t54WgQBDsk1ZCsgOcGvWjystD3y7XXItbvlyiZSP3darlXrTjfQc3HkL5EZwKYeOWL5bILW1j5blr6nuM0QjtHzO0pZYg/GfBDs1ZgNgIocKOFrJ4rnCx/Lxir3azgUPl5s8Xy8BmVbWJQKOq5TMILvi5cBYi4PKdKYny4/Ld8svKveooQd7eTa1iNW/Pl1ADEWPcfe3UVfDu1EQtI4Lz49vFSepE+ef1jaVk0SKyZMdhFUSw0IfbGYIJDuO0RulM+gwQODN8OTLc+R4eTyv/RY0rCR4oDUHoKsqXViSlqkBkdZ/DIQHBoHfjKO0uY4fzEuIEsmzenLLJ7VSEADWiX0OPMR3PL/6FKR6PRR4MyqeuxPJdR+TR71d5uFfAezc01c0UlI0BCGwI3A0WR2lmUAQJAaxJ/7e0ic2TxX6gUtbtBLnoYZlGzX1e9DjHLlvVsJKaIm525PIrIZsQUrBJqBamOzYbXZ23YBFuUi14ktut3WGsTg1zm/U+OQGTc5SgwAly4PgZtwhy0B9OEHf+yHnf7XhtEkGwuEMNPBZ9yDCo7GonDKeJP0POCQlEsCD/9eGr5IUJ61QcQBbTsp1HdXFn3ZlH6OjU4Z01CwMdZlBaggUuWn5aLf9wHyCzA/M9tCIdt3S3LhzRacMOJzbGOSxar24QKe9M3STfLknS7jQ40Dp1cMtqMrBZtQzlFChXQa4JxBsErkLMwHngwHXgplYxGsbqTWzFz4RAflWdSjIrMUXe/XOzdjqE6+C/C3fK0HaxWgaDdr6vXtdIHScQQyCwb005rtl7KO002YQG6EwoUzGiiLWLTG5cfP4EbgsIHBDDUEYIkct87qucB9eKXo0ipU+jaGlbq6Kt6w1keuHvaTrzYF3xUv+Gmn9jBV9HSZeVpc/3vKKYv+foKXUdTVqT7HE7Qq5RkvrMz2vUdYJN9hf7N5KhbWP92gY9Lym4q+UQAm8IuA1Q+vFYz3pSkHEHo551TvpOnst9eF52wMCAFls/LNutn/dvwnaBhJD8AZP1D25qKgM/XiBFChWSz4a2ynbbvUAIRsUEDGn/1nEczg2Q24k07OIqghw7K/Vd2eamtaydThAjsHgNRnVlgoS7nCl2zQsgguw4eEJOnHEullgKQwoqcHC8d2NTza55+X/rdWf72o/myae3tpTGFmEYro9/DkxQB8mzP6/Rtqp3fb1UW0GjS4p5HwOUjGDuPez7ldrhpdPbs+T29nFyb6eatoghEGjwXG5oWV2+mLddg0wTDxyX1//YpEJOZ+2sGKP5EtbNNpS5/PJgB1m0/bD8sHS3OkIghL88cYO8PnmTZlKgBSta7KYPs8Qctnv9SC3hwOMQoAo3yseztumBxX6PBlWkR/1IeahrHQ3YxNpj95FTGTJA0EHm+NkL+jvEYXUkGtGmThVnO12rSIKNRH8usr210EXrXGcr3VNuZ54vUAoEYQfjKZ43fl/Nq4fbGgx66twFdeZMXpesDhAIL/i5w7ojJybO7a7BuUDsQmivVWDv2aCKfHF7K5+/RxPaOmGlswQsPTg35NIYwQevlaf71NfOQaEERZAQAIrjlMc7yaVLzkGzIGMmySddThBTDpOX7hj0WzciyHXNq+XZzyWEEG/26plPdFU3XLDlRaGeuniRwprrBOeHNxEktyUkEEFQvnjA5d4DB11J+iZE0Q5M/khqZpkgNjlBQM3KcIge0EUJOjSYum5CCipYEMIlAYfcQ98t17IGdHp5uX8jGdKmuseCERkckx/rrOGQWCRisTh9wwEtLxnarobUdZWh92wYKeMfaC/P/rJG1u09pq1rv1m001YxBKLGR7e00DEQu/U/r9iju/6mrS2y8K5tEi2DW8RohyacBxbkaPuKA2POb6v2qWMFYgha5OJAXh26ycB1kr4TDR6PlrwQTLAQx8+du/mg5iWZXBCMzx3rVNYWxD3qV8nQIdO0GrcGoxqRBOU0yBjBgbwWK/i+KK8pUriwhokWcR34f2HzsVAhKVrE9bVCrq8Xcd3u8ZjC7sdAWIDIAcHmSi100WACZSYocY+tVEY/mtDVymWL+0WkwfNCHgsOCFgoZzQMQKlW3wYSFea8Jh0+cVbzYuDgSA9yXry1KjfCx6S1zr9nkqV7WHpMS3WU9TzUrbbXrkmhAEWQEMGfdeXBhBE7TBmMHW0UswsuCLAe1ot0KtuEEJKfeGu1GAxgoglhHxNpTOSt+UrucphcOkHcbc2PXRZB/OkEQRCuLxHEusucW5rGOHe3F2w95N4l9hWuSEhBIiEmTH5/pJM8MX61OkL+b8Ja7eiB7As4RgzIPEBYaT+XKwRi6X8X7dKjTc2KKoagWwacJBMf6aiL15HTt2ipiD/EELjeIOLgQOenX1bs0TBXlOIYYQJlcHCHwLmCzilGKIZr5fb2NVSo+WFZkooiEEfRHeaD6Zs1u+Km1tV10Ys8IQOEBLhicCBvAoLFjI0HVBBC2TfcHcbhgTEH8198jwbR5XT8RskOjg51PLOokKGxzUMccX7E4vvM+Ut65F0LXae44fzo/Dwv1lQXLl5SMWtmYorM3Jji0eUTQKSCuDSgWVXt9oP743f//V9JGVw1oE1cRfnPXa091jtG+PgDwsfaZLe4kRmFC4nm4zzYtbZuooQyFEFISGaCXA5GvZjnIgh+FtK6CSGE5H7iDxHE2iYX9msTRJebTBAQ6dpZQ0evjE4Q+0SQCNciCDuf6TFhqcYtYgeo70cNNzJBDOiCQQhxuqa/uL2lfDZ3u7zzZ6Jmhazbl6Ztc9NvXqEE4I9HO8mCbYdk7OJdugBFxy0ccAXATQFhAgJA9/pVvIoht7WvoWGsdrnxUO6GUpThPetpVxe4Q7DQRdkJhA0c6A6GbjNwdGBeClECAlBCTII837ehTFmfrOUyEDbQJhcHcoMgoEAQSd/xEiUYcBjgQDbIxuTjTkFk4wFZvSfNfeBno6RFBZGGkdKuVsUM4ah4PnC44LCCsR1hznBBoAzyouvA/y85HHLhous2Bz5e0s/19nT3vZj+sa5MixpopVuptAoM+dFCF2P97M1wexyUOYkpcsy1UWsEJzh5IHzgdYRSG/zNIBCh/Gnc0iS3A9LK/V1qySPd6riFGwgfyHOB28O8JqwCB7q7eKN4kcL6enmgSy39PRUEKIKQkKJsCeegZibIRgwx4gghhJDgIdxLhxgjiGA3L7c7dpHlS3iIIJhAYhJuFUjswLhK0D4RO3rWbBaTE1LRxkwQ/F7g/EBWAcCE2s7yHkKCHSww0V4UIgdyPZChM+DD+fLm4Ca6E24FJRXIe8OxP+2M7sZjUYosoU9mb5NP52yT7vFV1B2CwMr0Ygi6rXyzcJfc3sFeMQTPC+Kmtvm9rpFMWbdfBZGF2w6rOILjxd/WqWMFC1y0/cViGy6X65vDMRIjOw+dlPHLd2t5Bc4H5T848Hu5uXV1ubZp1QwbifjdNaxaXo9hPepKyrEzer4QROZvPaQukf+3eJceZYoX0d8bBJFu8RGZnjvK+0NpAY7rCRxEpsxlZdJRDxECgdl4vSB4t0vdCHekAdYwPy7bLT8uc5Y+eeP5vg3k1nax6l5yOj7SnI6PNZ7CB0phEfqLUhp83bSAN6D06Na2sdp9yJTbFBS4MiQhRdkSlwcQgFAmUJA75hBCSLBiMj+sgaJmNwydTjCht6McBpN/4wI5de6i7phht9AuUGOO54qdSYTyWSebR13BqHY6QUC/hGi3CHJDFlokElIQgVg46dGO8uj3K9UVAUEEC89nr6nvtbsf3rvDr64nj3Svo04IdF/Bwn/GphQ9MG7AGQKHyP8euUq7rUAMWbPHIoa0ryH3dq6l4aB2gcUwSmFwQMhFjgkEke0HT7q7y0SHldQOMcgPgTAK4iqXkad611dXCcIwx/21WxfsCNzE8ervG3QsgTvEZI6kByUvN7eJ1eP0uYuycNshFURmbEzRRTdCVnHgociXQMkMwjuN2yGUwPkv2n5Izx3dhUy3SAOymRAuC6GsWfUK7msYhIwl2w+r8AGxAh13vF1H8Lob0iZWxQ24cUypC0qk0gsffRpHqbCE18HXC3d6fK/yJYvKnVfVlDs7xKkDqCDClSEJKcr4coK4bieEEBI8wG4ODh2/LIKknT5nSymMCUa1ZoKYmunoMHvt0tixxblAbEk5fsYtgsD2ba5XdosgWLSgxh3ZACYfgBCSEbikxt7TVks54OzAgvHXVXu1iwrcHd7eP1hc9mkcrcf2gye0lS3cFMjaeHtKonZWuaZxtD7+t4fTiSFzt8s3i/wjhpgcqIe71ZGHutZWIQOL4ImrkyU57Yx8CufK7G1aijK4RTUNT61ZqYy609DpBAdKEJE5gpB/iCjjl+/Ro3ZEGbmhVXUNRI2PKue1FSxcJhqW2iBSy1BQZqTZIRsOaJkGBCYcaEWMshk4P/B8q4aX1HEXH/E5BJtgyTtELguEI4geyGE6awk1RdDrVbUrS/cGVVSYSJ/RlZx2WrNdxi/b7eHgsII8rAe61lYxG2LHRzO3qvixPZ3w0TU+QjNs8Pf8Y12yPD9hbYYQWJS9PNGrngp1wfL7DSoRZOfOnfLaa6/JzJkzZf/+/VK1alUZOnSoPP/881K8eMFUm0jeUM7lBDHih5lc5mUmCCGEEHuo7CVLw90ZJpehqFYRBN8f9ehGBEEnAH8stFQEcblOrKGo2Awsb8P5WMEOa/pAQlJwOHLkiAwbNkwmTpwohQsXlsGDB8uoUaOkbFnfge1du3aVOXPmeNx2//33y5gxYyTUgQiANqBwO7z0v/UqZkCs+Hzeds1pQLgoFrPeWqHWiigrI65tKE/1jtfOK2OXJMnq3anyv9X79IiPLKelC2P/3lZDWNOLIZoZ4gcxBGNA89gKeuD5wZ0AcWNW4kF9fjgAOsyga06TmHANOG1SPVyfDw4IFsgO+X1NsnZ3QT6FWXQjFBZlMxBUmsWES/WKpTycHfhd4Xvi+MfV9ZxigeaIpMiibYfVJZHeKWEFzwsBptEuYQSiiVMgcYolEJPtzvaAIwMu8tST53V8hgsRwdxHXZ87j/N6G9qbI+fDGqxtRAs4PXCgVCm9owghs9M3pGjJy7wtB33mdEDEhpDVqGqYTN2wX/qOmuchfCDEFiVGyH2BowkizH8X7vRZQvPawMYqpHhzOBVE/LIy3LRpk1y6dEk+++wzqVOnjqxbt07uvfdeOXnypLz77rv++JGEeDpBzniKICyHIYSQ4HWCHD6ZUQSxQzSAvRiTeezcoUXhrsPOCSa6BdiNCVq11mQjX8D5tZK5FeTWLAAAJNtJREFULu0hxMqtt94qycnJMm3aNDl//rzcddddct9998l3332X6eMwX3/11Vfdn5cuXbBcRHAwIKdBF5SLdmpJmemCUiuijNzerobma3jbRcfiEk4JHOv2pmmQKjqxJB44Li/+tl4FBJSjvDEoQcXQkdM3a5jo53O3y//zoxgCIBZgsYwDoi+e1+S1yZpZgk6KyBHBYR17IV5AHLkmIUrLMPB1OBDgLsFjjKvDmnGhgohLGGkaE+5RagFx4Lb2cXpgsxK/o31pp2Vf6hl1ROAjSnngWEH2E35G4pnj+vvzBX5X6h4xYokKJJf/D6EAwoVbtLD8P9UiakDMMOIGwlSzA4ZuiGfI9uhRP1K7Q6Yv8zGBpeOX7VGXkbeQ0/SlmgjAHT1ji4dQhPPpWg+Oj2gVWfD3+2nZHnnm5zVeu+qg281jPepqzo03505Bxi8rwz59+uhhqFWrliQmJsqnn36aqQhy9uxZPQzHjh3zx9MjIYyZFEPFRfjc5XIYiiCEEBJsVHYJB9ZymFTTHteG8hHsVMLWjV1ZBNi5nSCV/OAEcYWwohzGgMm+3SGshGzcuFGmTJkiS5culVatWultH374ofTt21fn4XBo+wKiR1RUVJZ/VijO3SFIIsgTB1q3QsxAqQtKQ16euEE7yiB7A+UsddN1UTHAJYGQ1ef6NlD3Bcpl8L2+W5KkR4vYcHWXlC5eRMtT8koMAfi+93SsqQfmygiFXbMnVZ8DPibuP67ZRSbQ0ypiNIkJ0/apyGRCZxa0ul21J0027jumIsLsxIN6WN0MbrdI9TB1NUAswuZk20zadmP+DmFkLwSS1NNO14gKJKclOfWMukogXkPQwYEx3E4QGApRB9cZCDkov0TJIlqZV0j3f5yjuR6hxNHZ8vekbD90Ql8z+BwfrV3OrgTcJcZhAuEDXXmudQkfEFBQ4tR39DzZfcQZ5J0e/J0e6lpHejWM9OpeInmYCZKWliYVK1bM9D5vvPGGvPLKK3n1lEgIAns0xFeHwzlRNo4QiiCEEBIa5TBprhISO8phQIOo8k4RJPmY2wmC3TO7iXB1Z7E6QUxXmmhXWQ4hdrBo0SIJDw93CyCgZ8+eWhazZMkSuf76630+9ttvv5WxY8eqENK/f38ZMWJEpm6QUJ+7I7zz5QGN5Mne8TJhxR7576JdKmaY7icdaldSMQMlM9auT9Y233e5AigRvDp2yS75c91+WZGUqgcW0XCOwEWAMNF1e4+pGKKtddtBDKnt7i7lD/CcTaeXm9s4bztz/qK6FtbsTtWxcfWeVC3DgPCAY/K6/e7Hwx3TpFqY9HWFcKKsECLKKjzm4ElJOnJKD5QF6c9zCc+XhZFwbUuc3gkHkaROlXJ6eAPOCoguEEaMe0T/n+YSTVJPy4HjZzWMulyJom7BIlNRQz93/j+zkhH8bHT6wvmt33dMS6CM2LH76Gn9mTkF6xU4SeKjymuIKn5XENSKFCqkbY3v/WaZvo58gZbEyINBbkuohc7aTZ6sDLdu3aoK9JVKYZ577jn5xz/+4aEmV69ePQ+eIQkVMJjjgpPqsraxHIYQQoK/HAZdWwymU4wdwaigQbRzkr1mb5rWvJuOCXZjFjLWTBDjBClorQmJf0EeX5UqVTxuK1q0qG5G4mu+uOWWW6RGjRrqFFmzZo0888wz6uT+5ZdfpKDP3bE4RRkHgk6RZ4FSmWkbDrjLSOCSQO4HwlS9tYHFgtS0s4Ub7Melu9URgkU7RA+AdrIGlDZ8MW+HCi03t46VqxtGaslFXuQ54GegiwsOw/Ez51WggVPECCPITcHi33SgARAz6kWWkzZxFbVDDoADYsuB47Jqd5oK2hAOcMAdY847IQb5IhXULQJxBLkfmYHfJ8QMHBAJvAGXC+SInJaBILsj6fApvS4YNwfcHXC/HHNtstrhOEHuR/1op+iB7kJGvIDYAqHs8XGrVCDzRd0qZaWzq0TG+jcjmZOtleGzzz4rb7311hUtePXr13d/vnfvXi2NueGGG7TOMDNKlCihByG5oWLp4iqCHLGIIGVLUgQhhJBgdYKgNhy7k5icm3a2aMtoB5h8AmPhhvBS18fuox2ZIFZBxzhBKIIQO+fhOQWZIYaEhASJjo6WHj16yLZt26R27dpeH1PQ5u4mcBgHXBHfLt4l45bu1v+jKwxCT/s3qSp3dKihmRreQAbQI93ryoNd62j2yPjlu3V331u5BMQQdKvBARcFfm7nupWlU90Ir9kT/gIZKEbEMRxGGcreNFmzO81dTgORY2PyMT2soKQD7j2EnWI8t4IOJjj/9A4HiCOlSxRVoQBdVvCxRLEiHp+XtBzu24t7uc3jvp6fY72ADj9wu0DgwEd87qtbS27Bzx55czMtVfH298Nr6YUJazXA1hd9GkVJl/gIFT8gwJHsk62V4RNPPCF33nlnpvdB/odh37590q1bN+nQoYN8/vnnOXh6hGQfqMIYwDCoMROEEEKCFzj7ihUpJOcvOu3HmOzZXUKCchgrqLn2R0gpQvrA3qOnMwSjIgSPELvm4ShlSUm5nOUALly4oB1jspP30bZtW7ej25cIUpDBeISOMo/2qKvdU9CZA0GVyGvAgVIPiCEII/XWxcSaPYJ2ssglWrz9sB5/7TySITwTgZ1zNx/UQ2SjznfRdrVzvcra6tZfGSK+gOMFPx+HcS7A3WbcIs4jVV0TyMqwlgJmBYgj6Vu8BjIQqTILVW1Ts6KGlKKEKr34gda3L/62ToN4fQHHCEJ7m8eGM+TUBrK1MoyIiNAjK8ABAgGkZcuW8tVXX2kdIiF5Aer7jGX6xFnn4FmmOEUQQggJNjBRrFSmhAbEYdcRiw67S0jCShdTqzms7aBng0jxBzVcbXexUwprOXZWTfAdnSDEznl4+/btJTU1VZYvX67zcDBz5kzt3GiEjaywatUq/QhHCPEN3AR/axkjg1tU084paHv7+5p9+v9VP6TKP3/fKEPaxGq5jK8yD4RXmmyOuzvWVFEEXVEgiCyBS2LH4QyiCDb7jOBiyiK6N6ginetG5FnpTPrxWlvZhpeSPo2j3cLIqXMXXV1ZTJtZV2cWS7cW922nzqs4fO5ixk4n+UXxIoWldIkiupZAF8rSro8Vy5RQt4rJIdl15JRY+93C/dK+ViVtY4vMlxqVLpdZwtm4aPthDcX9a4fvjI/Xrmukv0t/5sIUVPyyMoQAgl7jqCtEDsjBg5ftPNlRoAnJaTkMcGaCOC8YsN8REkpgR2/YsGEyceJEFZkHDx4so0aNkrJly/p8zJkzZ3Qncdy4cZrm37t3b/nkk08kMvLyos+bNfP777+Xm2++2f357NmztQZ8/fr1Wvv9wgsvXHF3kpCcUrlccRULIB4gdM+EpEba6J4YdXMzGfbdShXPYS/2B+imgJa8cLSgC02jquXpBCF+oUGDBlqKjjL0MWPGaIvcRx55RMdx0xkGc3WUunzzzTfSpk0bLXlB+1x0kKlUqZJmggwfPlw6d+4sTZo0ye9TCgpw/WweW0GP/+vbQH5YmiRjFyfp+PXRrK3y6Zxt0rtRpNzePk7a1qyYaSkLRJEG0eX1QLAqRJHNKcedgsj2w+oGSe+S2JJyQo/P5jgzRrAA7wFRpF6ECiT5EZSJn4lcPhwx2YirQJ4HGhwYYQRz+gwiSjrxBB/RGEF/nquUxv1/l3BhPkdJSpHChQWGCggMCEM1z9P9WNfjIGYY4GhZtuuIljGhc47JkTJUDSupggcchSgdwuMB/n5T1++Xj2dt1bKhzBh5UzMZ0LQqu7r4Gb+sDNGTHNY5HDExMR5fgyJIiD+p6ArSg+3O9MxmMCoJNW699VZJTk7W8RYT3LvuukvruTGJ9QUmtJMmTZLx48dLWFiYTooHDRokCxYs8Lgf3HvWNufoMmDYsWOH9OvXTx544AHtIjBjxgz5+9//rjuFEFUIsRtj8Uag6MHjZ3WSixIZCAp2gYnqv+9sLf4GgasQQWB9rl6htJw+71zE0AlC7AbjM8Z4CB1GKB89erT767huIPT01Cln7kHx4sVl+vTpMnLkSDl58qQK3HgMRG6SfbCwRu7H/V1qq8sMpTJLdhyRP9bu1yM+spzc3qGGXNesWpZKtrEgrq8dQ8prNxosqrcePOEun8H3TA+cBjhkkjMnBgtriCL5UTqTk0YHeI7ZeZ5YY/pD6Ek5dkYzo2Ylpmi5iskbNCVNcN1A9EBZEHJacI1CWf7PK/bKZ3O2aYBsZuApQ/jolxDttcMQ8Q9+WRliR5C7giS/nSAmlAk2NrtaKRISCCD4bsqUKbJ06VJ3C0R04MIOHtx3ZqcvfZvyf//73yqSdO/e3S12YMdw8eLF0q5dOw/Rw5drD7uKNWvWlPfee08/x+Pnz58vH3zwAUUQ4hdiKzrLSGA1NuUjCBYMxl2yuEplZPmuo7Lz0EmpVqGUe7GU17Z1EvqgE0xmonhcXJzHxiREjzlz5uTRsys4ILsBmSA4Nu0/pqUyE1bs1VKX5yeskxG/rtMgZnRHaRoTJgkx4dol5EpjQmFXFxYccJbgb4nWvU5R5IhMWpuc4TFoU2ta1QK40W5uXV0GNK2mZYHBjl0CCFrcovvNbLg9ElO0K44VhGd3qVdFg0nhsFmZlCqT1u6TNydvyvLPuL9zLRnUIkZqR5Sh8JFPcHuchGwmCMKpQHR4cE6WCfHFokWLVKgwAgjo2bOn7vYtWbJErr/++gyPQW04dv5wPwM6ecXGxur3s4ogDz/8sLo7ELAHxwdcJmZygftavweA+PH4449n+pxRfoPD2kaRkKxQ09WudsfBk3LAVT4SWT6wdzF9UdOVC4KuA+FlnO+Bhq7uNISQ0AYujtevT5Bn+tSXn5bv0c4ycAxAEMGB2wCcbvFR5SShWrhLGAlTsSOzMExco+tGltMDrXw/cji0rSsEkTmbD7ozj6ygTe2I39brYahVuYyWz6BVq8n3QBYTFv75UU6TV6CUZu6WQyp8zN58UPNW0uMsoSkkh054ZrFcCTh9Hu5WR4a2i9UsKBIYUAQhIYexSJtSGLaOIqHG/v37pUoVZxq7oWjRorrzh6/5egzsztbSFoA8EOtjXn31VXWKlC5dWqZOnSoPPfSQnDhxQh599FH397FmiJjvAVHj9OnTUqqU9/fbG2+8Ia+88kqOz5kUXNwiyKGTbieIr3DBQAflMGDn4ZNSvGgh924sIaRgdb26p2NNPdDtaq2ri4q2m92TpgtwuA9wfP+X8zElihbW0NSmMeGSUC1MmsSESa2Isj47WUGwqFOlnB5D29VQpwgEl0XbDsu8LQflz/UZRRGg7WEPeeZcWIFzrXn1cG0tHgOBxCWWRIeVDCpHG34f6MYzfcMB+XaJM7flSpjyxcyoU6Wsun66xkdIzUpl3BuzJPCgCEJCjvQDDkUQEiw8++yz8tZbb12xFMafjBgxwv3/5s2ba234O++84xZBcspzzz2nYaoGiCawXxNyJWpVdob97jh8UvalnrY9FDUvge0dbNh3TDvEgEZVw/L5WRFC8guMZZENS2qbXLM435t62qPFLJzNx89c0LILHAYEdzaqdrmMBh9RPujNsYHbakeU1QOiiAE/b9muo/L9X0nyy4q9V3y+yGWauuGAHr4IL11MhZIWsRXUmYJ5eNXwktrS1/kzRRyun+1wf+7Qj5Lu8/T3k0y+pv9z3+78HJ1pkMmxeneqLNx2SJ0xdoDsD4gd6CyGfBeWtAQfFEFIyGaCGEzdNSGBDjq3XClPCSUqyOtISUnxuP3ChQvaMcZXlgduP3funLZNtLpBDhw4kGnXLrRTfO2117SUpUSJEnpfPMYKPi9fvrxPFwjAY3EQkl0whsMejlT+WYnObnOxFYNzXMfEGQsCLHI2Hziht9EJQgixihUxFUrrgUU2QAgqMpHULeISRuASQXcYtFe1tliFywQuEeMWaRITri4NX6UsuL11XEU93r+xmbZu3ZB8TMNADxw7q04VI8rAjZcV0MUFY7UZr4MZ5LO0q1VJGldz/k6Z4RE6UAQhIUdkWAmt2zO2NTpBSLAQERGhx5Vo3769ihnI+WjZsqXeNnPmTLl06ZKKFt7A/YoVK6bdXJD4D9AZICkpSb+fL1atWiUVKlRwCxi47x9//OFxH3Soyex7EJIbYPfG7iZaESL4D7SpWUmCESw4+jSOkn/P3+FesJjgV0II8QZy7VAWiAPdZEx4J8ZDtzCyN0027jsmaafPawcTHAbkeUAMwSK+aXUs5sO1rMUbKGmBgyMzIJSgW9eB4xBKnGLJ5v3HZeXuo25xN5hAi/KyJYvqJipKjih4FAwogpCQo0TRIlqLN3mdM+cAajohoQQ6sqCF7b333qvdWhB4ilaIN998s7szzN69e7U14jfffCNt2rTRlrj33HOPlqQgOwTOjWHDhql4YUJRJ06cqK4OfF6yZEkVN15//XV58skn3T8bQakfffSRPP3003L33Xer+PLjjz9q611C/AUs3BBBQIXSxXR3Lljpm3BZBHn1ukYM7iaE5EgcRngqjhtaOUtL4ZbbfOC42y2CjwhcRZDnTHQ62XTZQQoRpHzJoip6IG8EH63/d38sVljn1SW9fCxZtIgKuSjpaR1XQe7pVFNvw2Pw8aLDoYGjKcfPyu4jp7Sl70G0Oj9xVkWUlONn5OgpZ1mg3dmA+L0gTBbXiujwUlK2RBFthY6Q0tLFi0iZEkX1HEM57JVkDkUQEpL0bhRlEUHoBCGhx7fffqvCB4QOdIWBu2P06NHur0MYgdPj1KlT7tvQxtbcF+Ut6OryySefuL8Op8jHH38sw4cP1zrbOnXqyPvvv69iiwHtcSF44D6jRo2SmJgY+fLLL9kel/iVIW1i3TXomNgGs3DQskZFeXtwEy3zuapO5fx+OoSQEKF40cLqYsBxS9tYvc2Ut6zZfTl4FV1jkO2BIy+A2GAVVSCQwGHhK9TVG2hFW6NSGZ3TOw+UDJWS6hVKS/lSRSlmkGxTyGFtEh5gIDgPu5dpaWm6a0lIVoEdsMVr09QCOP+Z7pm2FSPEFxyD/Ad/tyS7PPr9Svnf6n3y/o1NZVCLmPx+OiSI4fjjP/i7DXxOnL0g21JOaGjomQsX5ez5S3LW9RGfQzgx//f8eMn5NcvHs5bPrV+7lI3VJdwZVmGDIgfJizGIThASksCet+CZ7lK4MPqtUwAhhJBgB+LHvZ1qSeNqXFgRQkhOgejQtPrlgHS7wf76hUsOlzDiFFjSf0TpTuWyJShykHyDIggJWaLCgrOFIiGEkIzAPp0Qw3ayhBASyEDQQEcvbEKW41ScBCjcIieEEEIIIYQQQkiBgCIIIYQQQgghhBBCCgQUQQghhBBCCCGEEFIgoAhCCCGEEEIIIYSQAkFAB6Oa7r1odUMIIXmNGXsCuJN40MLxnRCSX3Bs9x8c2wkhwTC+B7QIcvz4cf1YvXr1/H4qhJACDMYi9Bwn9sHxnRCS33Bstx+O7YSQYBjfCzkCWAa/dOmS7Nu3T8qVK5fl/tFQfzDw7t69W8qXLy/BTiidTyidS6idTyidi53ng+ERg2jVqlWlcGFWD9pJQR/fQ+lcQu18QulcQu18OLaH5tgeaPA9E3jwPAKLYwF8Hlkd3wPaCYInHhMTk6PH4g8SaH+U3BBK5xNK5xJq5xNK52LX+XCX0D9wfA+9cwm18wmlcwm18+HYHppje6DB90zgwfMILMoH6HlkZXyn/E0IIYQQQgghhJACAUUQQgghhBBCCCGEFAhCTgQpUaKEvPTSS/oxFAil8wmlcwm18wmlcwnF8yGh93cNpXMJtfMJpXMJtfMJpXMhgUsovc5C5Vx4HoFFiRA4j4AORiWEEEIIIYQQQgixi5BzghBCCCGEEEIIIYR4gyIIIYQQQgghhBBCCgQUQQghhBBCCCGEEFIgoAhCCCGEEEIIIYSQAkHIiSAff/yxxMXFScmSJaVt27by119/STAyd+5c6d+/v1StWlUKFSokv/76qwQrb7zxhrRu3VrKlSsnVapUkYEDB0piYqIEI59++qk0adJEypcvr0f79u1l8uTJEgq8+eab+lp7/PHHJRh5+eWX9flbj/r16+f30yI2wbE98AilsR1wfA9cOL4Tuzly5Ijceuut+l4PDw+Xe+65R06cOJHpY86cOSMPP/ywVKpUScqWLSuDBw+WAwcOeL3v4cOHJSYmRl+rqampQXUeq1evliFDhkj16tWlVKlS0qBBAxk1alS+XtPHjx+v73ncPyEhQf744w+Pr6PPx4svvijR0dH6nHv27ClbtmwRf2PneZw/f16eeeYZvb1MmTI6T7j99ttl3759QXUe6XnggQf0fTBy5EgJJEJKBPnhhx/kH//4h7bsWbFihTRt2lR69+4tKSkpEmycPHlSnz9elMHOnDlzdLBdvHixTJs2Td/kvXr10nMMNnBBw2Ry+fLlsmzZMunevbtcd911sn79eglmli5dKp999pkuAIKZRo0aSXJysvuYP39+fj8lYgMc2wOTUBrbAcf3wIbjO7ETCAd4b2Ps+v3331Wgvu+++zJ9zPDhw2XixIm6AMT4h8XpoEGDvN4XYkRevOf8cR4YAyFsjx07Vr/3888/L88995x89NFH+XJNX7hwoYoy+J2uXLlSBXcc69atc9/n7bffltGjR8uYMWNkyZIlKiLge0Lw8Rd2n8epU6f0+4wYMUI//vLLL7qxMGDAAL+dg7/+HoYJEyboHAGCTsDhCCHatGnjePjhh92fX7x40VG1alXHG2+84Qhm8GeaMGGCI1RISUnRc5ozZ44jFKhQoYLjyy+/dAQrx48fd9StW9cxbdo0R5cuXRyPPfaYIxh56aWXHE2bNs3vp0H8AMf24CDUxnbA8T0w4PhO7GTDhg06Vi1dutR92+TJkx2FChVy7N271+tjUlNTHcWKFXOMHz/efdvGjRv1+yxatMjjvp988om+32bMmKFfP3r0aFCeh5WHHnrI0a1bt3y5pt94442Ofv36edzWtm1bx/3336//v3TpkiMqKsrxzjvveJxniRIlHN9//70tzzkvzsMbf/31l/5tdu3a5Qi289izZ4+jWrVqjnXr1jlq1Kjh+OCDDxyBRMg4Qc6dO6fKJexPhsKFC+vnixYtytfnRjxJS0vTjxUrVpRg5uLFizJu3Djd9YRtOljBTm6/fv083jvBCqyPUJtr1aqluyNJSUn5/ZRILuHYHjyEytgOOL4HHhzfiV3g2oHSkVatWrlvw3sE1xa4CLyB6xDcbtb3EsoBYmNjPa5FGzZskFdffVW++eYb/X7Beh7exnc7xvacXNNxe/oxDE4Fc/8dO3bI/v37Pe4TFhamZR3+mif44zx8/d5RSoK/czCdx6VLl+S2226Tp556Sl18gUhRCREOHTqkk5bIyEiP2/H5pk2b8u15EU/wpkBN8lVXXSWNGzeWYGTt2rU6KYbFDrWUsHo1bNhQghFM8mF9g1062MHF7uuvv5b4+Hi1Sr/yyivSqVMntechs4AEJxzbg4NQGNsBx/fAhOM7sRMsmFHuYaVo0aK6yMfXfD2mePHiGRajuBaZx5w9e1bLBN555x0VFbZv3x6U5+Gt/AElE5MmTcqXazqel7f7m+drPmZ2H7vxx3mkB9chZITgNYXMl2A6j7feektfi48++qgEKiEjgpDg2ZXCpCWYa3kxCVu1apWqsz/99JPccccdWlMZbBPl3bt3y2OPPaZ1pAg2CnauueYa9/9Rh4tJc40aNeTHH3/UukVCiP8IhbEdcHwPTDi+k6zw7LPP6uIrMzZu3Oi3n4/cDISIDh06NKjPwwrGdWQjIS8CmU8kb4Bb58Ybb9TAV4R2BxPLly/XIF2I8HCxBCohI4JUrlxZihQpkiGlGZ9HRUXl2/Mil3nkkUfcoU0IoAtWoKDXqVNH/9+yZUvdZcObHcFzwTZIIfSoRYsW7tugBuPvg/Ar7GjgPRWsYJejXr16snXr1vx+KiQXcGwPfEJlbAcc34MDju/EG0888YTceeedmd4H5VS4dqQPfbxw4YJ2WvF1XcHtKB1Apxeri8J6LZo5c6a6ySCgAmf0k/M6hnBROJiC4TyspT09evTQoNUXXnhB8uuajtszu7/5iNvQHcZ6n2bNmok/8Md5pBdAdu3apa8pf7lA/HUe8+bN09cl3FDW6w9e1+gQs3PnTgkEQiYTBBMXTFhmzJjhYc/F58FczxsK4CKASTJsxXgz16xZU0IJvM4woQw2cGHDxRq7nuZAXSlqrfH/YJ4gA7SI27Ztm8cFkQQfHNsDl1Af2wHH98CE4zvxRkREhOZbZHbgmoJrB0QAiIUGjGF4v8Nl5A1ch4oVK+ZxLULnDmTTmGvRzz//rO1lzXvuyy+/dC8K4ZYLlvMA6ArTrVs3dcP961//kvy8puN26/0BXG7m/rj2YAFuvc+xY8c0F8Vf8wR/nIdVAEEG0vTp07WNsT8p7ofzQBbImjVrPK4/yHNCPsiff/4pAYMjhBg3bpwmAX/99deamHzfffc5wsPDHfv373cEY6L7ypUr9cCf6f3339f/+zMd2F88+OCDjrCwMMfs2bMdycnJ7uPUqVOOYOPZZ5/Vzgc7duxwrFmzRj9HCvfUqVMdoUAwdw944okn9DWGv82CBQscPXv2dFSuXFk7VpDghmN7YBJKYzvg+B64cHwndtOnTx9H8+bNHUuWLHHMnz9fuygNGTLEo7NFfHy8ft3wwAMPOGJjYx0zZ850LFu2zNG+fXs9fDFr1iy/dofx13msXbvWERER4Rg6dKjH2G7X++1K1/TbbrtNx18D3vNFixZ1vPvuu9rJBt2i0OEGz9Pw5ptv6vf47bffdPy+7rrrHDVr1nScPn3aluecF+dx7tw5x4ABAxwxMTGOVatWefzuz549GzTn4Y1A7A4TUiII+PDDD/WNXbx4cW35s3jxYkcwYgbO9Mcdd9zhCDa8nQeOr776yhFs3H333fpGxusLF4gePXqEzAQ52CfJN910kyM6Olr/NmjJhc+3bt2a30+L2ATH9sAjlMZ2wPE9cOH4Tuzm8OHDKhaULVvWUb58ecddd92lIrUBghvGM4zZBiyo0SoWrbNLly7tuP7663WBmp8iiD/OA4tab2M7xse8uKZjrEp/Tfzxxx8d9erV0/s3atTIMWnSJI+vo03uiBEjHJGRkbqgx/idmJho2/PNi/Mwfytvh/XvF+jnESwiSCH8k99uFEIIIYQQQgghhBB/EzKZIIQQQgghhBBCCCGZQRGEEEIIIYQQQgghBQKKIIQQQgghhBBCCCkQUAQhhBBCCCGEEEJIgYAiCCGEEEIIIYQQQgoEFEEIIYQQQgghhBBSIKAIQgghhBBCCCGEkAIBRRBCCCGEEEIIIYQUCCiCkJBg9uzZUqhQIT0GDhyYpce8/PLL7seMHDnS78+REEJI9uH4TgghhBA7oQhCgpKuXbvK448/nuH2xMRE+frrr7P0PZ588klJTk6WmJgYPzxDQgghOYHjOyGEEEL8SVG/fndC8pgqVapIeHh4lu5btmxZPYoUKeL350UIISR3cHwnhBBCiB3QCUKCjjvvvFPmzJkjo0aNctudd+7c6fW+P/30kyQkJEipUqWkUqVK0rNnTzl58mSeP2dCCCFXhuM7IYQUPOLi4jKULjZr1kxLGwnxBxRBSNCByXH79u3l3nvvVbszjurVq2e4H24fMmSI3H333bJx40atKx80aJA4HI58ed6EEEIyh+M7IYQQQvwNy2FI0BEWFibFixeX0qVLS1RUlN62adMmr5PkCxcu6MS4Ro0aeht2DQkhhAQmHN8JIYQQ4m/oBCEhS9OmTaVHjx46Mb7hhhvkiy++kKNHj+b30yKEEJJLOL4TQgghJKdQBCEhCwLxpk2bJpMnT5aGDRvKhx9+KPHx8bJjx478fmqEEEJyAcd3QggJHQoXLpyhnPH8+fP59nxI6EMRhAQlsEtfvHjxivdDqN5VV10lr7zyiqxcuVIfN2HChDx5joQQQrIPx3dCCClYREREaJmj4dixYxS1iV9hJggJ2hTpJUuWaNcAtEG8dOlShvvg6zNmzJBevXppa0V8fvDgQWnQoEG+PGdCCCFXhuM7IYQULLp37y5ff/219O/fX1uhv/jii2xxTvwKnSAkKHnyySd1cIQNGupxUlJShvuUL19e5s6dK3379pV69erJCy+8IO+9955cc801+fKcCSGEXBmO74QQUrB47rnnpEuXLnLttddKv379ZODAgVK7du38flokhKEThAQlmPQuWrTI/TnaI6YHO4JTpkzJ42dGCCEkN3B8J4SQggWE7XHjxnncdscdd+Tb8yGhD50gJKSIiYmRIUOGZOm+r7/+ulqtve0yEkIICSw4vhNCCCHEDgo50kfxEhKEnD59Wvbu3av/x8Q3Kirqio85cuSIHgCW67CwML8/T0IIIdmD4zshhBBC7IQiCCGEEEIIIYQQQgoELIchhBBCCCGEEEJIgYAiCCGEEEIIIYQQQgoEFEEIIYQQQgghhBBSIKAIQgghhBBCCCGEkAIBRRBCCCGEEEIIIYQUCCiCEEIIIYQQQgghpEBAEYQQQgghhBBCCCEFAooghBBCCCGEEEIIkYLA/wdXWYQVktrubgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1100x260 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# [1b] 한 샘플 확인: 입력 지진동 / 응답 / 이력루프\n",
    "i=0; t=np.arange(L)*dt\n",
    "fig,ax=plt.subplots(1,3,figsize=(11,2.6))\n",
    "ax[0].plot(t,AG[i]); ax[0].set_title(\"input a_g(t)\"); ax[0].set_xlabel(\"t[s]\")\n",
    "ax[1].plot(t,UU[i]); ax[1].set_title(\"response u(t)\"); ax[1].set_xlabel(\"t[s]\")\n",
    "# 복원력 근사(이력): 변위 vs (alpha k u + (1-alpha) k z) 대신 u vs u로 비선형 흔적만 간단 표시\n",
    "ax[2].plot(UU[i], np.gradient(UU[i],dt)); ax[2].set_title(\"u vs u_dot (phase)\"); ax[2].set_xlabel(\"u\"); ax[2].set_ylabel(\"u_dot\")\n",
    "plt.tight_layout(); plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "732628a7",
   "metadata": {},
   "source": [
    "### 2. 공식 `SpectralConv1d` 전사\n",
    "\n",
    "nb01에서 본 \"모드별 복소수 곱\"이 여기 그대로다. 차이는 $H(\\omega)$ 대신 **학습 파라미터 `weights`**(복소), 그리고 **앞쪽 `modes`개 저주파만** 곱하고 나머지는 0(고주파 절단 = 저역통과).\n",
    "- `rfft` → 길이 `width//2+1` 의 0 버퍼 → 앞 `modes`개만 `compl_mul1d` → `irfft(n=width)`. (공식 `legacy_spectral_convolution.py` `SpectralConv1d.forward` L573–607 과 동일한 순서.)\n",
    "- `compl_mul1d` 의 einsum `\"bix,iox->box\"` 는 공식 `_contract_dense`(L33–58)의 1D 축약과 동일."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "239a17eb",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:18.872471Z",
     "iopub.status.busy": "2026-10-29T09:18:18.872214Z",
     "iopub.status.idle": "2026-10-29T09:18:18.880318Z",
     "shell.execute_reply": "2026-10-29T09:18:18.879930Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SpectralConv1d in : (10, 4, 256) -> out: (10, 4, 256) | 복소가중치: (4, 4, 16) torch.complex64\n"
     ]
    }
   ],
   "source": [
    "# [2] SpectralConv1d (공식 구조 전사: rfft -> 저모드 복소곱 -> irfft)\n",
    "def compl_mul1d(inp, weights):\n",
    "    # (batch, in_ch, x_ft) x (in_ch, out_ch, x_ft) -> (batch, out_ch, x_ft)\n",
    "    # 공식 _contract_dense 의 1D einsum \"abc,bdc->adc\" 와 동일\n",
    "    return torch.einsum(\"bix,iox->box\", inp, weights)\n",
    "\n",
    "class SpectralConv1d(nn.Module):\n",
    "    def __init__(self, in_ch, out_ch, modes):\n",
    "        super().__init__()\n",
    "        self.in_ch, self.out_ch, self.modes = in_ch, out_ch, modes   # modes = 유지할 저주파 개수\n",
    "        scale = 1.0/(in_ch*out_ch)\n",
    "        # 학습 가능한 '복소 전달함수' R_theta : 공식은 FactorizedTensor, 여기선 dense 복소 파라미터\n",
    "        self.weights = nn.Parameter(scale*torch.rand(in_ch, out_ch, modes, dtype=torch.cfloat))\n",
    "\n",
    "    def forward(self, x):                       # x: (batch, in_ch, L)\n",
    "        B, _, Lx = x.shape\n",
    "        x_ft = torch.fft.rfft(x)                # 실수 FFT (주기성 가정) -> (B, in_ch, L//2+1) 복소\n",
    "        out_ft = torch.zeros(B, self.out_ch, Lx//2+1, dtype=torch.cfloat, device=x.device)\n",
    "        out_ft[:, :, :self.modes] = compl_mul1d(x_ft[:, :, :self.modes], self.weights)  # 저모드만 곱, 나머지 0\n",
    "        return torch.fft.irfft(out_ft, n=Lx)    # 공간영역 복귀 (B, out_ch, L)\n",
    "\n",
    "# 모양 점검\n",
    "sc = SpectralConv1d(4, 4, modes=16)\n",
    "xt = torch.randn(10, 4, L)\n",
    "print(\"SpectralConv1d in :\", tuple(xt.shape), \"-> out:\", tuple(sc(xt).shape),\n",
    "      \"| 복소가중치:\", tuple(sc.weights.shape), sc.weights.dtype)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67f9a86e",
   "metadata": {},
   "source": [
    "### 3. 공식 `FNO1d` 전사\n",
    "\n",
    "공식 `FNO`(fno.py)의 뼈대: **lifting**(채널을 width로 올림) → **Fourier blocks**(각 블록 = `SpectralConv1d` + pointwise skip, 그 합에 `F.gelu`) → **projection**(채널을 출력으로 내림).\n",
    "- 입력 채널 = `[a_g(x), grid]` 2채널 (공식 예제도 좌표 격자를 채널로 붙인다 → 위치 정보 제공)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "99d6450a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:18.881917Z",
     "iopub.status.busy": "2026-10-29T09:18:18.881782Z",
     "iopub.status.idle": "2026-10-29T09:18:18.889186Z",
     "shell.execute_reply": "2026-10-29T09:18:18.888851Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "FNO1d(\n",
      "  (fc0): Linear(in_features=2, out_features=32, bias=True)\n",
      "  (convs): ModuleList(\n",
      "    (0-3): 4 x SpectralConv1d()\n",
      "  )\n",
      "  (ws): ModuleList(\n",
      "    (0-3): 4 x Conv1d(32, 32, kernel_size=(1,), stride=(1,))\n",
      "  )\n",
      "  (fc1): Linear(in_features=32, out_features=128, bias=True)\n",
      "  (fc2): Linear(in_features=128, out_features=1, bias=True)\n",
      ")\n",
      "\n",
      "학습 파라미터 수: 74,209\n"
     ]
    }
   ],
   "source": [
    "# [3] FNO1d (공식 fno.py 구조 전사: lifting -> N x (SpectralConv + skip, gelu) -> projection)\n",
    "class FNO1d(nn.Module):\n",
    "    def __init__(self, modes=16, width=32, in_ch=2, out_ch=1, n_layers=4):\n",
    "        super().__init__()\n",
    "        self.fc0 = nn.Linear(in_ch, width)                                   # lifting (ChannelMLP 등가)\n",
    "        self.convs = nn.ModuleList([SpectralConv1d(width, width, modes) for _ in range(n_layers)])\n",
    "        self.ws    = nn.ModuleList([nn.Conv1d(width, width, 1)         for _ in range(n_layers)])  # pointwise skip\n",
    "        self.fc1 = nn.Linear(width, 128)                                     # projection 1\n",
    "        self.fc2 = nn.Linear(128, out_ch)                                    # projection 2\n",
    "    def forward(self, x):                        # x: (B, L, in_ch)\n",
    "        x = self.fc0(x).permute(0, 2, 1)         # -> (B, width, L)\n",
    "        for conv, w in zip(self.convs, self.ws):\n",
    "            x = F.gelu(conv(x) + w(x))           # Fourier block: 스펙트럴 경로 + 국소 경로, 비선형\n",
    "        x = x.permute(0, 2, 1)                   # -> (B, L, width)\n",
    "        x = F.gelu(self.fc1(x))\n",
    "        return self.fc2(x)                        # (B, L, out_ch)\n",
    "\n",
    "model = FNO1d(modes=16, width=32, in_ch=2, out_ch=1, n_layers=4)\n",
    "nparam = sum(p.numel() for p in model.parameters() if p.requires_grad)\n",
    "print(model); print(\"\\n학습 파라미터 수:\", f\"{nparam:,}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fc0d880",
   "metadata": {},
   "source": [
    "### 4. 데이터 텐서 만들기 — 좌표 격자 채널 붙이기 + 정규화\n",
    "\n",
    "입력 $a_g$ 를 표준화하고, 공식 예제처럼 **좌표 격자** `grid∈[0,1]` 를 두 번째 채널로 붙인다. 출력 $u$ 도 표준화(스케일만, smoke라 역변환은 생략)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "51cc4a59",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:18.890654Z",
     "iopub.status.busy": "2026-10-29T09:18:18.890542Z",
     "iopub.status.idle": "2026-10-29T09:18:18.896930Z",
     "shell.execute_reply": "2026-10-29T09:18:18.896596Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X: (60, 256, 2)  Y: (60, 256, 1)\n",
      "한 배치 X: (10, 256, 2)  Y: (10, 256, 1)  (batch=10)\n"
     ]
    }
   ],
   "source": [
    "# [4] 텐서화 + grid 채널 + 표준화\n",
    "AG_t = torch.tensor(AG, dtype=torch.float32)\n",
    "UU_t = torch.tensor(UU, dtype=torch.float32)\n",
    "ag_mu, ag_sd = AG_t.mean(), AG_t.std()\n",
    "u_mu,  u_sd  = UU_t.mean(), UU_t.std()\n",
    "AGn = (AG_t-ag_mu)/ag_sd\n",
    "UUn = (UU_t-u_mu)/u_sd\n",
    "\n",
    "grid = torch.linspace(0,1,L).repeat(N,1)                 # (N, L) 좌표 격자\n",
    "X = torch.stack([AGn, grid], dim=-1)                     # (N, L, 2)  채널=[a_g, grid]\n",
    "Y = UUn.unsqueeze(-1)                                     # (N, L, 1)\n",
    "print(\"X:\", tuple(X.shape), \" Y:\", tuple(Y.shape))\n",
    "\n",
    "ntr=50\n",
    "ds = torch.utils.data.TensorDataset(X[:ntr], Y[:ntr])\n",
    "dl = torch.utils.data.DataLoader(ds, batch_size=10, shuffle=True, generator=rng)   # batch 10\n",
    "xb, yb = next(iter(dl))\n",
    "print(\"한 배치 X:\", tuple(xb.shape), \" Y:\", tuple(yb.shape), \" (batch=10)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24f38bea",
   "metadata": {},
   "source": [
    "### 5. Smoke run — epoch 1 (훈련 아님, '돌아가는지'만)\n",
    "\n",
    "목적: (1) 순전파 출력 shape가 입력과 맞물리는가, (2) loss가 유한하고 한 번 역전파·갱신되는가. **batch 10, epoch 1**, 5개 배치만 돈다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "92828d24",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:18.898392Z",
     "iopub.status.busy": "2026-10-29T09:18:18.898263Z",
     "iopub.status.idle": "2026-10-29T09:18:19.653628Z",
     "shell.execute_reply": "2026-10-29T09:18:19.653183Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "forward: in (10, 256, 2) -> out (10, 256, 1) | target (10, 256, 1)\n",
      "  batch 1/5  MSE(normalized) = 1.14237\n",
      "  batch 2/5  MSE(normalized) = 1.22880\n",
      "  batch 3/5  MSE(normalized) = 0.66972\n",
      "  batch 4/5  MSE(normalized) = 0.67098\n",
      "  batch 5/5  MSE(normalized) = 1.17183\n",
      "\n",
      "epoch 1 완료 | 평균 loss 0.97674 | 첫 배치 1.14237 -> 끝 배치 1.17183\n",
      "핵심: 에러 없이 shape가 맞물려 돌고 loss가 유한 -> FNO 파이프라인 OK (성능은 W2에서 제대로 훈련)\n"
     ]
    }
   ],
   "source": [
    "# [5] 1 epoch smoke run\n",
    "opt = torch.optim.Adam(model.parameters(), lr=1e-3)\n",
    "lossf = nn.MSELoss()\n",
    "\n",
    "# (a) 순전파 shape 점검\n",
    "with torch.no_grad():\n",
    "    yp = model(xb)\n",
    "print(\"forward: in\", tuple(xb.shape), \"-> out\", tuple(yp.shape), \"| target\", tuple(yb.shape))\n",
    "assert yp.shape == yb.shape, \"shape 불일치!\"\n",
    "\n",
    "# (b) 1 epoch\n",
    "model.train(); losses=[]\n",
    "for bi,(xb,yb) in enumerate(dl):\n",
    "    opt.zero_grad(); out=model(xb); loss=lossf(out,yb)\n",
    "    loss.backward(); opt.step(); losses.append(loss.item())\n",
    "    print(f\"  batch {bi+1}/{len(dl)}  MSE(normalized) = {loss.item():.5f}\")\n",
    "print(f\"\\nepoch 1 완료 | 평균 loss {np.mean(losses):.5f} | 첫 배치 {losses[0]:.5f} -> 끝 배치 {losses[-1]:.5f}\")\n",
    "print(\"핵심: 에러 없이 shape가 맞물려 돌고 loss가 유한 -> FNO 파이프라인 OK (성능은 W2에서 제대로 훈련)\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "02384b9e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-29T09:18:19.655442Z",
     "iopub.status.busy": "2026-10-29T09:18:19.655238Z",
     "iopub.status.idle": "2026-10-29T09:18:19.736571Z",
     "shell.execute_reply": "2026-10-29T09:18:19.736193Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxYAAAD5CAYAAAC6RKNtAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8ekN5oAAAACXBIWXMAAA9hAAAPYQGoP6dpAACVJUlEQVR4nO2dBXhT59vGn7pRWmgpBVoo7u7uMDY2ZK7ANrZ9U+ZjytyFjfl/7hszxjaGu7u2WAsFCrSlVKnnu+43501P06RNGjsnfX67MtLkJDn6nvex+/ExGAwGYhiGYRiGYRiGcQBfRz7MMAzDMAzDMAwD2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGC/iyy+/JB8fH9q6daunV6VOMGLECPGQpKSkiP2P4+AsEhISaPr06eQpXnvtNerQoQOVl5eTHpDH4I033vD0qjBmvP7669SqVSvy8/OjHj16aOL8ZiqDa2fOnDmmvz/66CNq3rw5FRUVeXS9GP3AhgXDMIyHWb9+vbiZnz9/nrRETk4Ovfrqq/Too4+Sr2/duV28+OKLdNlll1Hjxo2rTLS0gqfPmX/++ceu/bJ48WJ65JFHaPDgwfTFF1/QSy+9ZHG5/fv3i++FgchY5vvvv6d33nnHLb8Fo6+4uJg+/vhjt/weo3/qzp2CYRjGxbRo0YIuXLhAN954o92TxGeffdbiJDEpKYk+/fRT8gSff/45lZaW0rXXXkt1iSeffJK2bNlCPXv2JK1S3TnjLsMCv28ry5cvF8bpZ599RjfddBNdfPHFFs9vGBb4XjYstGFYBAcH07Rp0+itt94ig8Hglt9k9A0bFgzDOAwmn/Bq6YX8/HyXfC+827gRI9XDWQQFBVFAQAB5AniW4bnHNtUlkpOTKS0tjb799ltPr4rXcPbsWQoJCaHAwEDNnN+MbVx11VV07NgxWrFihadXhdEBbFgwjIPk5ubSrFmzRK4wbpIxMTE0duxY2r59u2kZ5OF36dKFdu/eTcOHD6fQ0FBq06YNzZ8/X7y/atUq6t+/v7jxtm/fnpYuXVrld3bs2EETJkyg+vXrU7169Wj06NG0cePGGtcvKyuL+vXrR3FxccI7CJAv+8wzz4h1wDrHx8eLNAVb8mjVOezwmrVu3Vp8BzyNssbD3Nu4cuVK8Tr+Nd8n+NzIkSPFPmnWrJnI6bcFfN/dd99N3333ndhnmPz27t2bVq9eXWk5pFVgWfzOddddRw0aNKAhQ4aY3sfkEZ/Dvm/YsCFdc801lJqaWuX3PvnkE7GtWA77c82aNVb3jXmNRWJiorg5N2rUyHSMn3jiCdP6Pfzww+J5y5YtxefV+9BSDvrRo0fpyiuvFOuL/TZgwAD6+++/Le7zn3/+WaT24PhjH+G8OXz4sE2Ta5yvY8aMqfIe6i1w7Dt37iy+EylDt99+uzjX1GDdJ06cKNJgkFOPZTt16kS//fZble+0ZZtAYWGh2Gft2rUT39ekSROaOnUqHTlyxOoxw/nZt29fEYWwBay3M1FfM7asE7z7Q4cOpbCwMIqMjKRJkybRgQMHTO/XdM6Yg2sdk/f09PQq7912223iN7BfJf/++6/p98PDw+mSSy6hffv2md7H+fj++++L5/K38bAG3oORCoNeLiuvEfX5jddwDgCMCXJZ9bjhKCdPnqSbb75ZnLM4BjiHEZmTIOKImiI88Fxy7tw5ca4NGjSIysrKTPsBYzHO3fHjx4v91bRpU3ruueeqePdtvWbk/sd9Avse4z3OE0Qp5LiJ6wITfbl/1OerrWM7/r7//vvFmITfgQPhxIkTFvcZxkdcl3/++Wet9ztThzAwDOMQ1113nSEwMNDwwAMPGP73v/8ZXn31VcOll15q+Pbbb03LDB8+3NC0aVNDfHy84eGHHza89957hk6dOhn8/PwMP/74oyE2NtYwZ84cwzvvvGNo1qyZISIiwpCTk2P6/N69ew1hYWGGJk2aGJ5//nnDK6+8YmjZsqUhKCjIsHHjRtNyX3zxBe5mhi1btoi/09PTDT169DA0b97ccPjwYfFaWVmZYdy4cYbQ0FDDrFmzDB9//LHh7rvvNvj7+xsmTZpU4/YmJyeL38D6t2rVSqzL22+/bTh27Jjp97GMmhUrVojX8a+lfXLfffcZPvjgA8OoUaPEcv/880+N64HlunTpYoiOjjY899xzYr+3aNHCEBISYtizZ49puWeeeca0vtg+/M77778v3nvhhRcMPj4+hquvvlq8/uyzz4rvS0hIMGRlZZm+A8cV3zFo0CDDu+++K/ZbZGSk2H5sh/m+wX6Q7Nq1y1C/fn1DVFSUYfbs2WJ/P/LII4auXbua3r/22mvF57Afv/nmG/HIy8sT72Obpk2bZvq+06dPGxo3bmwIDw83PPHEE4a33nrL0L17d4Ovr6/ht99+q7LPe/bsaejdu7f4bpxjOO79+vWrcf/i/MXnd+/eXeW9W2+9VZwvM2fONHz00UeGRx99VJyfffv2NRQXF5uWw7q3a9dO7KvHHntMrCu2G+u6ePFiu7eptLTUMHr0aLFe11xzjWHevHmGl19+WZw3f/zxR6VjgO1u06aNOC9ee+01cVzj4uIqrV9N4PrBd+EccgR71mnJkiVi32K/YRl5TjZo0MB0XdV0zphz6NAhsSzGHTVFRUXie2+++WbTa19//bW4Ji666CKxPNYV1wOOofz99evXG8aOHSu+U/42HtbAe0OHDhXjlVz2yJEjVc5vvHbvvfeK73388cdNy+L8cAb4HuxvjDkYMz788EPDZZddZtqPEoypGJvvv/9+02s43zC2JCUlmV7DegcHBxvatm1ruPHGG8X5OHHiRPF9Tz31VK2uGYwd2P8Y21588UUxVuGz+H6A6wZjOs4JuX9+//13u8f2G264Qawn7l9Y76lTpxq6detm9XwfM2aMGEcYpibYsGAYB4ERcNddd1W7DCafGLC///5702uJiYniNUye1MbBf//9V2VyOnnyZGG8yJsxOHXqlJiIDRs2zKJhkZaWZujcubOY/KakpJiWwY0Iv7lmzZpK64ibHT67bt06myZJmCyfPXu20nv2GhZ4DRMZ9UQHRtbll19uqAl8Fo+tW7eaXoNxgxv9lClTqhgWmIipwT7B5AE3bzUwSnAjlq/jph8TEyNu5lg/ySeffCK+tybDAscHxwnrpqa8vNz0/PXXX7e43ywZFpgwYFn18cvNzRWGJiaAmFyo93nHjh0rrffcuXPF62rjyxJPPvmkWA7frQa/i9e/++67Sq8vWrSoyutYd7z266+/ml7Lzs4WBjIm2fZu0+effy6Wg+Fhjtyf8hjAkDt37pzp/T///FO8/tdffxk8ZVjYsk44z3C+ZWZmml6DIYFr9qabbrLpnLHEwIEDDf3796/0Gow29XWJfQ4DApNf8wk5xjn16xjz7PFN4hzGRLqm8/uXX36pMlY4i1tuuUWcexkZGZVeh9GA7SsoKDC9BicA9vnq1atN6wTHj/k24fV77rmn0nl4ySWXiPEa548918z58+fFWIHjdOHCBavjBb4f+80cW8f2nTt3ir/vvPPOSsvByLB2vt92223CsGKYmuBUKIZxEKQRbNq0iU6dOlXtcgiZI81GgnQYfLZjx44iDUoinyO8DhB2RyrJ5MmThVSjBGF5pPasXbtWqPeoQUgbofSSkhKRGoSiYskvv/wifhOh/oyMDNNj1KhR4n1b82gvv/xyEUZ3BOyTG264wfQ38q+RZiS3vSYGDhwowvQSyCIibeS///4zpStI7rjjjkp/Ix0H6QlIUVLvh9jYWGrbtq1pP0C6F/nh+Lw6PxxpEBEREdWuH1JPsP+ReoF1U1Nd6khNRbPYR+p0LuxHpLQgFQYpX2pmzJhRab2R4gJq2seZmZnk7+8vvlsNzh9sN9L91PsNxwHLmp8/SA2ZMmWK6W+kdqB4F6l9p0+ftmubfv31V4qOjqZ77rmnyvqa78+rr75apL3Zu92upKZ1Ql3Hzp07xbmF1BNJt27dxP7Gfqot2OcYp9QpY0gjRKoMxgqwZMkSUQyOYn31sUXNEMYlPefYwxeB8+fSSy8Vz9XbhzSm7OzsSumrSDdD2hIKl++8806xj+69916L342UTPMUTdScyZRWW68Z7H+k1j722GNV6ppsGS9sHdvleWS+PUjptQbOW6SGFRQU1LgeTN3G39MrwDB6BzUBuPngBo0bBdROcBNXGwEAOe7mNwfcbPA589eAzL3F5BSDOQwRc3ATweQYNQG4CUqgSoRJIfKyMVFWc+jQIfG6NaMAk2iZU6wuyEZtgHoijdxuR7G0T3ADQ26/LcAAMAe599hf2G/qbTdfX+wHTDAsfQeQBaXIZbb0W3jf/BibIyeMqCVxFlgftSGqPhfk++rfMzdo5MTWUm63LWC/YRKGWqLqzh8Jcr3NjzGOEYDRgGNk6zZhUozrAOd2TTh7u51BTeskzzVr1zoMZtQpIJe/NkYNJo4wJp5++mlxDBcuXCjy7OXxwbEFciJqDoxCLYGJLrZDjfl4J8F4AKMJNS541HTuwhhH7QXqGzDJR42Ipck9lK7MxwH1+W3PNSONvtqOF7aO7TjPsN6o9VFj6byTyJqR2jpEmLoDGxYM4yDweMPz+Pvvv4vIAppAQfsfHnEUW0usKQVZe90RaT8Us3799dc0d+5cevnllyu9B0Oka9euQj7QEtLQwXegqFwC40ldlAxDwxxrNx3z6IErt90a5uuL/YD1RaGkpfUw99Trldru46ioKKH2BQ8qijvV+w0TJExQLeFoFMtZuPPc0sM6wYhBIb00LCAcgQJedcRQNkH85ptvLE7QbTHo3MlPP/0kInK27Eu5bdhejGWWQGRIDQw5gMJ2TNpr60xx1zVj69heG2D8QlTB0rjPMGq0NUowjE5BWhLC5XjAK9SrVy+hxKM2LGoLbjoY0KWik7naEDxP5jcMpIrAU4wJBKIMCK1L4KXatWuXUAeqzvv05ptvVvLuIqWlJqQH1lxbX3pinY30sKo5ePCg2F813ayxHzAJwWRBehgtIdPI8FtqTy7SzKCc1L17d6uflZ7MvXv3Vrsu9ngBsT7WzgX1+joK0ikAtlE94cJ+Q4oHGp3ZMsmAAhX2s3obcYyAVLOxdZvw20jnwb73RolSuZ3W9gXSwGS0ojaeY0RSkSoIJSpMctGnQx3plB5sTIItqYGpcZXn2p7vRQoT0odsQaofwclR07YBRE2h7gTDBelpt956K+3Zs6dK+iMm84hMqscQ8/Pb1mtG7n+MFxi/7d1Hto7tOM+w3jICKLF03kkwDsgIIsNUB9dYMIwD4CZlHorHTRmTcFukW231co4bN05I/anlJM+cOSMkCJGXbilF4amnnqKHHnqIZs+eTR9++GGlCAskFy01XUNqgezxgLQu3IDlAzKhNSFvjGrJV+wja6kHjrJhw4ZKedFICcN+wv6qqZcEIjJYBs24zL2c+Bs1BqBPnz5iUvLRRx9VSg1D9Kam5mT43LBhw0RKxfHjx6v8hkROFm1pdoZUu82bN4ttl+CYYR9jImPLcbK1fkXWmKjB+YNj+vzzz1f5DCIc5tuA2iNE8ySoB0I0DfKz0itu6zahrgc54/Pmzavy297QvAsOCuyXr776qtJ+xEQT0VDZVM7ec0YCRweME0RUEY1URyvkRB1jCbpiw3gzRy1XW5vftwV7vhf7Sz1GVWcw4FrH+YM6C0uGvnrbsO2oc8E4jqgvrnWMt0gbs4T6fMR5iL9h+GKCb881g3ELxg+izGr5X/m96n1kft+xZ2yXDq9333230jLVNd3DOAupXYapCY5YMIwDIE0EdQJXXHGF8FwjfQaeKXgE4fF3Fi+88ILwzMGIQFQEKQkff/yxMF6q6/uAtCzcgO666y5xw8JEAvUX6G2AYmQU88GLhpsePKJ4HeF/TKZrA7yf6D8AYwY1GihA/fHHH8XN0xUgFxmTIRQhQrP9gw8+EK/b0hEYRhD2K9YVBhuK47GP4JnDRBiFwzDMMEHActCcR8QCuepYBjnXNdVYyJs3jhuiWPhOREjwe9CihycUyAJ09LZAgT9+E0WmlnLpEX364YcfxOQA2419jIko1gmTJkSwnAG2DfsX5zOKzyUoYsW+wOQH64/JENYXER0Uj2IihutBAk/uLbfcIq4JaPfDyMIkDfvP3m2Cxx1GyQMPPCAMEaQgYrKEdcR1AW+8M0AqEKJsslAVhjLOAYDrR0YW0F8B/RbQNwDFvs4A1yz2Aww77DdMCN977z3hKVf/hj3njATLYFlMfDHRNu+oDqMCTghsI85XLAvjGEYxzleMFXISLX8fxwvXIL5PLU5RW2BY4btg/GDswnWN685afYI9vPLKK2LMQz3PzJkzhcGKcQqTZpxDeA5wrHFuL1u2TIwJiNgh+ouO7Di31QYe6i8WLVok0qvwvUitxL56/PHHTVFTW68Z7P+3335bREdQ2yH77iAKgXMR14Tc90gDw3WA5XDfwbG3dWzHPsaxx3iJfQyDAdtqrb/Ntm3bxL5x1vXFeDk16kYxDGMVyHiiLwU09yETCDlFPEdPBDWQJIX0qzmQDIR0oDm4NM0lbLdv324YP368oV69ekKnfOTIkUJPXo15HwsAqU5IrUJCVWr9Q0IV+vRYJ2jLQ8seGuXQzIccqC3SmZC7tAQkcaF5ju9FbwLo0UOb35LcrKV9AglHS1KK1vYR+i1ARx6/BwlTc5lKKTcrpR/NgRTqkCFDxLHDo0OHDuJ71Xr1AMdU9g7p06ePkKHENtQkNyv7kEACF1KekMNt3759FZ179CdBDxPIRaplRM3lOOU+vuKKK0zfh74UCxcurLSMlJuFVKYaa+toCci64nxTy3Cq5XZxzkCCEuc++lOgPwdkkM3Pb0goQyMf+w7713ydbN0mgHVBrwsci4CAACFPjM9JKebqzk9bpWOlFLKlh/r8gkwsXoOcZ3XYu05Lly41DB48WOxbyDqjL87+/furfNbaOVMdmzdvFsui34E1sI0YayDBimPRunVrw/Tp0ytJO6OnCGRWGzVqJPou1DSdsFVuFnz66adCJhty0M6Wnj1z5oy4vtHLQp4/6I2C8xls27ZNjJVqCVm5veg5gd47sseN3Cace7J/BMY8HE8pkWzvNQMWLFggeubI449r4YcffjC9j34lkIbFtYL9ox4vbR3bIWeLniGQQMY24BxLTU21eD6i5wZ6IaklbxnGGj74n6eNG4ZhGHtBDjEiMZbSYhjnAG8mIheIisF7bi9IY0LUA+pD3gg6GiPSAk8vPOt6AN5veKwR+YGHm6k9SJdCEXxeXh55K4iK4zpGVPG+++7z9OowOoBrLBiGYRiLIP0Gk2ek50hVHaYCpJuglkkvRgVA/j1SZ1BjxDA1gZRFpG2Z9wFiGGtwjQXDMAxjlUcffVQ8mKqgbkQv/PXXX6LRIAri0cCtNr0wmLoHDAo2Khh7YMOCYRiGYbwcSFCjaB6Fx7aIGzAMw9QGrrFgGIZhGIZhGMZhuMaCYRiGYRiGYRiH4VQoJ4CiRjSBgt61q7qRMgzDMAzDMIy7QXIT+nahaWRNvZLYsHACMCri4+M9vRoMwzAMwzAM4xJSU1NFU+DqYMPCCSBSIXc4OmcyDMMwDMMwjDeQk5MjHOhyvlsdbFg4AZn+BKOCDQuGYRiGYRjG27Al3Z+LtxmGYRiGYRiGcRg2LBiGYRiGYRiGcRg2LBiGYRiGYRiGcRiusXATZWVlVFJS4unVYBjGwwQEBJCfn5+nV4NhGA1yPLOADGSg+Aah5OvL8vWM/mDDwg3av6dPn6bz5897elUYhtEIkZGRFBsby31vGIYxse5wBt342SYqNxAFB/hS34SG9NZVPahReJCnV41hbIYNCxcjjYqYmBgKDQ3liQTD1HFHQ0FBAZ09e1b83aRJE0+vEsMwGqCguJQe/XW3MCowTSgsKac1hzLolX8T6c2runt69RjGZtiwcHH6kzQqoqKiPL06DMNogJCQEPEvjAuMDZwWxTDMm4sP0omsC9Q0Ipj+nTWMdp84Tzd+tpl+23GCZgxOoC7NIjy9igxjE1y87UJkTQUiFQzDMBI5JnDdFcMw249n0efrksXzF6d2pYiQABrathFd1r0pGQxEL/1zQEQ7GUYPsGHhBjj9iWEYNTwmMEzNFJeW07N/7aPZv+2m7Avea4S/8V+SMCCm9mxGI9vHmF5/eHx7CvTzpfVHMmllUrpH15FhbIUNC4ZhGIZhNEVJWTnd+8MO+mJdCv2wOZUmzVtLSadzydvIyi+mTcnnxPP7x7ar9F58w1CRBgXmLjvkkfVjGHthw4JhGIZhGM1QWlZOs37cSYv2nRYe+yYRwZSSWUBTPlhHW1KMk3BvYUXSWSorN1CH2HBhSJhzy9CWoph7Z+p5OpNT6JF1ZBh7YMOC0T2fffYZjRs3rtplMjIyRKHsiRMn3LZeDMMwjP18tjaZ/t6TRgF+PvTxjb3p73uH0sBWUVRQXEZvLzlI3sTSA2fEv2M7Nbb4fkx4MHWPixTPlx0wqskxjJZhw4KxyIgRI2jWrFmkdQoLC+mpp56iZ555xvTa9OnTafLkyZWWi46OpptuuqnScgzDMIy2QJHyj1tSxfOnJ3aikR1iqGFYIL1+ZTfhuUe9AZrIeQNFpWW0SqmdGNPRsmGhNjqWKUYIw2gZNiwYh24ApaWlHl2H+fPnU/369Wnw4ME1Ljtjxgz67rvv6Nw57wqlMwzDeAvbj5+n5Ix8Cgnwoym94kyvxzUIpSFtosXzX7YZDQ+9s+FIJuUXl1Hj+kHUtRo52dEdjQXdaw9n0IXiMjeuIcPYDxsWnmiQVVzqkYetcnXw+K9atYrmzp0r1GvwSElJoZUrV4rn//77L/Xu3ZuCgoJo7dq1FiMEiHYg6iEpLy+nl19+mVq2bCl0/Lt37y6MgurAb/3xxx9VOhZ/+eWXpr9//PFHuvTSS01/z5kzh7766iv6888/TeuO9QadO3empk2b0u+//27TfmAYhmHcy6/bjemqE7rGUr2gyq22ru4bL/79ZesJUZegd5bsN0YgRndsTL6+1pXi2jcOp2aRIVRUWi6MC4bRMtwgz81cKCmjTk//55Hf3v/ceAoNrPmQw6A4ePAgdenShZ577jnxWqNGjYRxAR577DF64403qFWrVtSgQQObfhtGxbfffksfffQRtW3bllavXk033HCD+N7hw4fXeptg2Nx4442mvx966CE6cOAA5eTk0BdffCFea9iwoen9fv360Zo1a+iWW26p9W8yDMMwzqewpIz+2nVKPL9CFa1QpwQ1CA2g0zmFtPpgukiT0itw9NVUXyGBgwzLfLk+hZbuP1Pj8gzjSdiwYKoQERFBgYGBoolXbGxslfdhbIwdO9bm7ysqKqKXXnqJli5dSgMHDhSvwSiBUfDxxx/X2rBAV/Ps7GwRhZDUq1dPRETwm5bWHcvu2LGDvM1YPZaRTwH+vtSoXhCFB/tznwSGYXTpwc8tLBXe+QGtoqq8H+TvR5N7NhMStD9tSdW1YbE/LYfO5BRRaKCfKEyvCaRDwbBYlniWyssN1UY4GMaTsGHhZpA3isiBp37bGfTp08eu5Q8fPkwFBQVVjJHi4mLq2bNnrdfjwoUL4t/g4GCbPwOjA+viLZSWl9OxzHwqLisXj/yiUnGcW0aHkb8fZzoyDKO/NKipvZpZnTgjHQqGBbz9uYUlFB4cQHpkx/Hz4t/eLRpQsA335v4to0RqWEZeEe0+mU094o1KUQyjNdiwcDPwJNuSjqRlwsLCKv3t6+tbpX6jpKSiS2peXp749++//6ZmzZpVWg51GtXtq+q+NyoqSiyTlZVl87qjcBvpV94A9s2JcxdEd1povUeEBNC5/GIRwUjPK6ImESGeXkWGYRibgJGw5pCxfmCqhTQoSYfY+hTfMIRSz10Qhd7D2+lzPN9zIlv8K6VkayLQ31dEcWBQbU05x4YFo1nYpclYBKlQZWW2qU9gop6WllbptZ07d5qed+rUSRgQx48fpzZt2lR6xMfH2/y9hw4dqhRtwDriu/fv32/zuu/du9ehKImWgPGQU1gijKsWUaHUJDLE1GApM69YNJliGIbRA1uPZYmCbIxliLhWR98EY93cFqVjtR5B1AF0jbOuBmVOz+ZGY2JHqjHawTBahA0LxiIJCQm0adMmUbCN5nJQdbLGqFGjaOvWrfT111+LyT96RWACLwkPDxdF1ffff79QbDpy5Aht376d3nvvPfF3dd87b948UROB77/jjjsoIKBy2Hv8+PGiVsN83Xfv3k1JSUli3WWUA0bJtm3bamympwfKDQbKyC0Wz5tGBlOIEgVDfQXC6ng/M9/4PsMwjNbZdNRoJPRTjIbqkMts1mkXbhSpHzyTK553s8ewUKIUO5U0KobRImxYMBaBIeDn5yciAogcINpgDUzu0aTukUceob59+1Jubq5oRqfm+eefF8tAHapjx4500UUXidQoyM9a48033xQRjaFDh9J1110n1gkF5Wqg7vTPP/+IIm7JzJkzqX379qIWBOu+bt068TokaJs3by6+T++glgL1Ff6+vtQwNND0OqIXMeFBpqgFivwYhmG0zubkTPFvfxsKmfsohsXO1POiyZweC7cRnYmuF0Sx9W2vEewWHymaBJ48f4HO5hS6dB0Zprb4GGxtbsBYBdKmUFLC5BbN2tRdoZOTk8Xk2Z4CY8Y+rrzySurVqxfNnj272uUGDBhA9957rzBS9E7quQLKKiimqHpBQkFFDS7ppDO5ovaiaWSIuHkx2oLHBoapAH2Wus1ZTKXlBlrzyEhTSqc1MMb1eWGpiMrOv2OgydDQC1+tT6FnFuyjUR1i6PPpfe367Pi3V4vx/eMbe9P4zlWVDxnGnfNcS3DEgtE9r7/+upCZrQ6kRE2dOpWuvfZa0juIQmRfMKZ3RYZUVURB1AKyswDF3AzjbaRk5NONn22iaZ9vpiw+x3XP9mPnhVHRNCKY4hrULDqBMa5PQgPdpkPtVgq3q+u2bQ1ZtI1oDcNoETYsGN2Dmop77rmn2mWio6NFqpY39HdAwTZqKKAEBQ10S0AhykfJ5UXkgmG8AXiqf9x8nC5+d41QEFp1MJ2u/XSjkOBkvCMNytYxWs8F3HtOnre7vqJKAfdx29UQGcadsGHBMDpDRisiQgOs3oTRw0IWdEPGkWG8gW82HqPHfttDBcVl1K9lQ2oUHkSJp3Pp6o830NlczjnXKxsV4wDH1FbkslJNSk/1cYfP5tU+YqEYFpCr1dN22wM7w/QNGxYMoyNQsJ1TWCqeR4ZUFG1bAgpRAJ1sGcYbJhsfrDgint85ojX9OHMA/Xz7QJE+cyQ9n95ectDTq8jUAkRVZVpPfzsMi05N6lNYoJ8Y36TCkl4Kt2EPoGg7xo7CbUnbmHCx3fnFZXTorH62uyZwDsz+bQ+NfGMltXvyX3p/xWFPrxJTS9iwYBgdkV9YKtJBgvz9KDjA1ybDIq+oVKROMYyeWbDrFJ3OKRSqZ/eNaSs6M6PfwRtXdRfv/7UrjS4U608hqK6zK/W8MBoRfaqpf4V5VLZXC2OdxRYd1VmY6itqkQYF/Hx9qJvSVE9279Y7xzLz6aqPNtAPm49Tcka+eO31/5Lo792V+2Mx+oANC4bREfBSSaOhplzkkAA/IUcLo6KgiKMWjH6BMf3JamO0YsbglsKwlgxoGSU6McOAXrSPJyJ6A6lMMrXJ3hq4Xs0bVJqs64E9J5T6ilqkQZnXWXhLP4uX/0mk4rJyUZj+v5v60PRBCeL1B3/ZSXuVRoKMfmDDgmF0BCZPAKHwmsBN2pQOxYYFo2NWJqXTwTN54ry/rn/zSu8hcnFFr3jx/JetJzy0hkxtQa2AuvmbPXRsYpS9TDydQ3oBNUGgU9PqJTuro7uyr3YpRoqe2XAkkxbtOy0iMa9d0Y3GdGpMT17SkYa3a0SFJeV0+zfbRLocox90Z1i8//77QgUI2u/9+/enzZs3V7v8L7/8Qh06dBDLd+3aVTRTk6Aj86OPPipeDwsLo6ZNm4rGbqdOnXLDljCMfZSWlZsG2NAgo8FQEybD4gIbFox++ViJVsCogOKZOVN7NRP/rj+SSSeyCty+fkzt2aN4pDs3td+D37FJuPgXRifGR62DYmuZ6tMmpnqJ9JrqS8CR9Dwq0cF2V7c/nl+4Xzy/rl9zatc43JTm9t51PUUdCpoB/rfvtIfXlPFaw+Knn36iBx54gJ555hnavn07de/eXXR9Pnv2rMXl169fL/oWoDvzjh07aPLkyeKxd+9e8X5BQYH4HnSExr+//fYbJSUl0WWXXebmLWOYmoESDkAaSICfbZduvSB/o+xsKWRn2evD6A9MLDYePUe+PsY0KEugodqg1saOzb9uO+nmNWRqC3qQ4PjW1oMf3yBUSG6jRkNO2LXMqfMXqKi0XEiFxzWovglgdaApKsb2krIKQ0WPLNh1UhSzwwF2/9h2ld6rHxxA1/YzRie/33TcQ2vIeL1h8dZbb9HMmTNpxowZ1KlTJ/roo48oNDSUPv/8c4vLz507ly666CJ6+OGHqWPHjvT888+LDs3z5s0T76OL4JIlS+iqq66i9u3bi87MeG/btm10/Lj1E7moqEh0IVQ/GP2zcuVKkT50/nz14eVly5aJ86mszL0T9fxiJQ0qyHIaFNb9jz/+sCo7m19UtwwLRDbfeecdl//OiBEjaNasWaa/MY78+uuvTv8dTJ5+3pJa53KOlx04I/7t3aKB6CRvjSv7xIl/529PFTUZjPbZd8p472wRFWoxElUTSINrH2v0ch9QUoy0zOF0o8wsitSR+lNbsN3tGhsjHgfS9Dv/+H2HMTvk1iGtqGFYVZXDq/rGCYfCpuRzJoleRvvoxrAoLi4WE/4xY8aYXvP19RV/b9iwweJn8Lp6eYAIh7XlAdqVY4IWGWk93/Pll18WRol8xMcb83u9ienTp4v9YP44fPhwpfdfeeWVSp/DxNa8AA8T8LfffluknCElrUGDBjRhwgRat24d6RE02nvyySfJz6/mOgdnIg2DMBvToCSyHkMaJmoQpRs3bhxFRRkbU+3cuZM8hfkE3VG2bNlCt912G7kbnBuPPfYYlZc7L0UB6T1XfbyBHvl1N039YD39saPueOWXHjBGpEd3bFztchd1bkKB/r6Ueu4CHdWxF7cusfeU0UjuUos0qCp1FjqYYB9RJsetY2xXv7JG+1jjdifpwKCy1uh1w5EM8fySbk0sLtMkIoRGdTBe92iMyegD3RgWGRkZYoLauHHlmwv+Pn3acv4dXrdn+cLCQlFzgfSp+vWth2Vnz54tDBD5SE1NJW8E0Z60tLRKj5YtK1IRYCS8+uqrlJVlvQMoPIfXXHMNPffcc3TffffRgQMHRGQAxhgmkuYedkfB75WWuq6eYO3atXTkyBG6/PLLyd25qFJKM0yJQNhKqBLhkKlUavLz82nIkCHiOOoBe45vo0aNRETT3cBozs3NpX///ddpxY2XvLtW6LzDywn1lFk/7aR3lx3yes88xAo2HjF2ZR7TMabaZUMC/YSqjF67MVcHaqteXZRIL/1zQKSF6KlYuTpk9K1zs9oXMneUEQs9GBbpRoO3daPa11dIOijbrVfDYkXiWZHK1bpRWLX1Jtf1Nzpu528/wUXcOkE3hoWrQSE3UqJwo/7www+rXTYoKEgYHuqH3RTnW3+UFNqx7AXblq0F2M7Y2NhKD7WXHtEgvIYIjjV+/vlnmj9/Pn399dd06623CsMEtTGffPKJqGXBa5jcWiIlJUV40X/88UcaNGiQMGS6dOlCq1atqpK+hElc7969xTpj8g9vMdYLvxcSEiJ+E+uhBoX87dq1E++PHDlS/F5NYF3Gjh0r1kXNn3/+KdLs8HqrVq3o2WefrTQBxjrivMKkE7+HZczXZ8+ePTRq1CjxPiII8Lbn5Rk9XAXFpWQgA/3183fUs3tXsZ1NmjShu+++u4oBPmXKFDGhbtu2LS1YsIBCFUOkqKSMysy86DfeeCM9/fTTVSJ79vDll19WifCZR67mzJlDPXr0oG+++UakKCHSB4MTE3AZAcNxRfqijI7heFg7vjDuJk2aJBwF9erVo759+9LSpUurTYXC9/zvf/+rsn/UoP4Kxwjfie/G/sE+leBchcAD3sf+f/PNN6vsD1wjF198sThXnGFQPjx/l+i2DiWYFQ+OoNuGtRLvvbXkIC3Zb0wT8lbWHkoXhhRSZWyZjPVLMDZY2+xFhgXuSQ/9sos+XHmEPll9lB7/fQ9dPHcNLfWCYy9ToWrTgVrSwaQMpf0JNoqtnWVYyBQwPWy3JRbvM56/4zvHVrvc8HYxognm+YISLuLWCboxLKKjo8UN+8yZyoMp/sbk1hJ43ZblpVFx7NgxUXNRK0PBXl5qav3x842Vl329jfVlv72i8rLvdLW8nAvA8XjppZfovffeoxMnLMs8fv/992Lyfumll1Z578EHH6TMzEyxz6sDNTJYFgX4AwcOFN+Fz6lB6gnSshAR6datmzAqYMygDmffvn10//330w033GAyShBlmjp1qvgupP/AwMF31MSaNWuoT58+VV7DZBMRmf3799PHH38sJtsvvvhipeUgEoBIx65du+j6668XE2usr5ywIk0PaWJI4YGaGSbK0nBAGtTPX39Gzz/xkDA4YIRgUtymTZtKvwGDBufy7t27xeQWv5ObfV4UCxqsRC3cBYwBGBwLFy4UDxwLmUoHgwLHFjVUMjqmTjE0P74wuLB9qHfBeYHoGo5ldbVR1vbPuXPGSShqa2DY9ezZk7Zu3UqLFi0S4wWWV5+LWG8YkosXLxaGD4QfzOnXr584L5zh1TuRdYEiQwNEp+nmUaH0+MUd6ZYhxsjh/9YmU51Ig+rQ2KYeB32Vzs2bddQwrSbeW36YFu5OI39fH6GKhagMOjc/t3A/FelYkCG3sMRUeFwbRSjzCXZadiGdLygmLXPUiYaFjFig+B1pRXoCkYeVSWdtMiwQpb2ij/Fe8M8e7lOjB3RjWAQGBgqPJSYSEnil8TcmJJbA6+rlASax6uWlUXHo0CExkYOnmDGCyR88s/Jx5ZVXVlkG3l94oqHUZYmDBw+KQmdLyNexTHVgco0JOZaH1x/e7s8++6zSMki1QiShdevWQjoYBg+K+jFZR3QAHnEYFpj0A3wPloXHGYX7mGBimZqA8QlZYvPJKia+06ZNE7+F9YBQgPwtCfYfDBgYWngfBgqMMmmAIRUPxhCiMpjgQkgAHn5MblEf8cm7b9Jd99wnDBh8B7z05jUJ2Aak8sHgwD7ABBySzDJq4UnDAtcrDC5s39ChQ0U0QF6fOKa4xhFJsBQdUx/fhg0bigjU7bffLr4LkQfsT7xnHoEwx9r+AdjfMCrwOiSq8Rzn0IoVK8Q5imVx3r3xxhs0evRoUTP01VdfWUzNwjkC49XROouvNhijaFf3iRepPpKZQ1uJGy488/uUPHVvA9EaGFa2pEFJUOCNYk8YY2nZZtFcHbJo72kRmQIvTO5CL03pSt/d2l90Hz9+roA+X1tzlFWr7FeiFVA4slS4aytQD4prEKJ57z2Mnow8o+HTqpHjNRaRoYFCjhUc1PB2W2Ld4QzR7LVJRDB1s6EDubz+1x3OFCIWjLaxL1nbw0BqFpM3TMjgEUSaAzy9UIkC8Bo3a9bMlJqDCdjw4cPF5PGSSy4RqQnwRCINRxoVV1xxhfA4YhKNGg5Zf4HJCyY6LuPxanpl+JgVBT98uJplzWzDWXvIWSA9SJ0Whgm7JZCfj4nwQw89ZPF9R/PA1Yagv7+/OP7S0y9RRxFQYA4pYUxEzQUAMFkE+Dz6oFj7HWtcuHChShoUIhAoRFdHKHAuwVDAesg8f/Pvx9+yWBrrg8myeh8PHjxYTEwTExOJIptR+pk0Gju2+pQlePMl+C5E3yDH3CfIj6Dq6EnDAqlJ4eFGLxtAKpE1qWhzzKNEmOQjvervv/8W0Q1M7nFsjian1Gr/yOMIIwJGtKVoC74f55D6vME4AcPUHKSz4dhBQQ7Pa5s2seZQBsFRf8OAFpXei40IpgldYoUn+8t1KfT6ld1r9RtaBjUlmfnFQopSRiJqAhKc8H6jNwKMrkk9jP0t9AgmUHMW7BPPZwxOoGsU6U2INzw2oQM98PMumrf8EF3eqxnFKBNMPbFXMSw6O9AoTtIhtr4wJlHAPaBVlKbrKzCZtleAo7pozemcQmFQ9VHSAPWATGka18m2SCSK+6PCAsV4sP14lmaPMaNDw+Lqq6+m9PR0kRMOAwCecqQryAJtpEFAKUqCvHx4gqHS8vjjjwvPJlIx4OUEJ0+eNHk48V1qMMFAcbHLCAzz/LI1gImXeaqNJYYNGyYiAyhqN/f6w7NubgRI5OtYxhnrKpF1CZh0wtBUgxx9R1PyzIvV8XuIWiC1yhxzI6Q2oAFSmLLeSGmqjoCAypKNGLQxwYXWu6lWw2CwaTC3FVxz5sYjjHZb180WzI1aGLGIPr786mvUIDaeig3+9OAd0+hsdr5QXompb/k4V7cOOI5Ip7JUyA4jSCqi2QLSq7DOtTUqwDcbjol/R3eIEX0azEFPBxgWf+46JSaaUfUcO7e1hkyVQAdeW/u2gL4JDYVhsSVF34bF33tOiUljdL0gcXzVTO7RjL7ecEwYX4hovHJ5hcGsF/YphdtdHKivUDfKW3rgDB1Iy60T9RXqdKhVB9N1VcBdXm6gZQdsS4NSy+sOa9eIft9xklYmpbNhoXF0kwqlTotBOgo8gZs2barkPUS+M1ItzNNP0PQOy6MwE3nVag8qJkSWHi41KrwQ5L//9ddfVaR8UUeANDO8Zw4iSUg9M48smLNx40bTc3imITtsLb0KoMcJDAgYmjCM1A+Zt4/Pm3dtV/+ONRDxQB2FGhRt4xwz/y081Iau+ffjb7kd+Bcec3UhO6Ig+Hx8yzYUVi+c4uJb0PLly6k2BAf4ka+Pj0gvQYMmZwL1JRRhq9e9NrK1iBDa2hsE++aGm26izoPGUOOEdhQVE0OnThjV2ZA2lpJRQPYGynAcUY+DccH8OMJIQKoVDBOMOxIYmZZS+TDWyOhYbcgvKqVftxnrlm4amGB5fZtHUve4COHZ/sELpRg3HjXWUQ1tG23X5/q11H8BN+5Bn6xONkUr0BTTfKIljY2/d6fpMj3EJDXrgCJUFclZDatlVRgWznP+yfoSPRkW2A+IPAQH+NociQQj2jeq5HBgtIvuDAtGmyDfHHUK7777bhXDAnUYSGFDfjqUflA4i/x4RIug0mMtxUry/vvv0++//y5Sgu666y4xmbv55putLo90G3i0UbCNHHiksSDdDfUM+BvccccdwuBBMS6MAkS2zI1SSyAyA1UiNYigoTYCUQtMTBGJQdodImVqUJCNnH1MRFGTAsNGFmdj3yG6gf2ESSkiZvfcc4+oQwhvYJxYPTj7CWGMYR9j3eU22QKMCpmjr06HgmcdRoA0lrAv8Lc1SWZLwLhHuheigtjXtu5LczChx6Qd5wiUmKqLZrRp25Z+mf8b7duzm1IO7qfnH/w/IkM5NQwNFDnXUNAqLTdQuR3WBc4t7A/UYKCAHtvy33//iVRLGDxIkbrlllvEOQMDD8cJETq18ShB4Tb6g9SW1QfTKbeoVKghDWljeWKNaMv0wUaj4xfFCPEWIK0Mbzyw1zvZN6GB+PfgmTzR2VmPIJcc8qmINF7f35gCZUkBK7peoDhPEJ3RW/GubHjmSOF2FenVM7nCeaJFjpxVpGarkVatvTJUjm6kp7ceM0b8IUJgTyRyaNtGIi0UaV+ns82UM72EwpIyr5DUZcOCcRoosDWfDGLyA8lZTDrRJA/56CjcRdQJEabJkyfbFA3BAzUImNTDIEFKUnWgmBcqTKi3QTQAqkFIjZJ9OJo3by66IyM1Dt8L9SgU7dYEDAAYD5iAq40N1OhAJQgF1ei8jG1t0aJyXjwMDxgcyPOHIfLDDz+I6ArAxByTWExs8R2o/UGBMAqKZf+Km26aJuqKPvjgA+rcuTNNnDhRGBi2ok6HkmBfwrOOGiRpCOJv7A8JJs/VRfBQZ/Dtt98K+V4YmNgu1D/YC4xBFGxjnyAKUp3C02NzXqJ69SNo2uTxdPf0a+jiCReJiAM8ufENQ0SjNCIDZReU2HzDRcE1IiEwImAUYFtQHA8pXWk8vP766+L8RcoUJHrRAwSiEmqQYrl+/XpT7VdtWKc0jhrZPkZskzXGdoqlAD8fOpZZYFLY8QaQRw2Ne+SjN7eQBlYdSAmTXmE5idEbn6w5Kv69qk+8KNK1BM4LnB9AppboBRgVmP83CA0QheiO0iIqTFzzhSXldArFZF6uCCVB/weIOOQUlgpVLD2wTbkm+7SwryYEBf7d4iJNjhdvorzcQJ+uPkrdn11MHZ9eRINeXka3f7NVtwIUPga9mLkaJicnR6jaoFmeWqoWxbvJycliMuuMXPu6BjzX2HeQEzWvgfEk8FjjmJurPlUHDCxEXWwxpNTg8oTWOzzv7RqHi5Sm2oJeCMcy88V34LtsBQIIKOSvjbHgCnIulFBKpnESnRAVRvVDKtdNyFSio+n5InKBiam1yZkrQJNNRNWkSIQlahobRr2xUnSP/uTG3jSuhjzk6z7dSOuPZNLTEzvRzYoMrd55a3ESvbv8ME3p2Yzevtr+a3/2b7vph82pdMfw1lXqE7TOoTO5NPbt1ULdatXDIy3W10gW7U2jO77dLiJbKx8a4dTaKVfy2/YTovi8f8uG9NPtNYtm2ML4t1eLiMUXM/qaDC6tgFQ1TBgRTdk4e7QQX3AWY99aRYfO5mlyuy0x8o2VwgnyxfS+NLKDfev79pKDNHfZIbq4ayx9cH1lh45eOZ1dSPf/tJM2KKmfanBd/zBzADWNrH2tnqvnuZbgiAXD2MkTTzwhohGOSonaAuohYFQglSlIeOFrT4hilBSVlAsPiS1gEEFKkDXFL3cDQwsFrQBFrZaMCgDVlUaKJ/RsbpFb0wRiYmJExKy2wEsFowITy/42pAHJycQKL8o93njUmNqDiWdtkAXB+3XQjdkcFKiCUR0aV2tUgCFtG5kiVjhn9IKsCZApTM5ASrjCoaA1jp/LF0ZFWKAfNbYiLFFbZNdqiFZoncy8IlNktVdzY8qiPcg6C6jllZbpr67IksE5/YvNwqjA/Rly0lufHEM/3jZARN5xXV/zyUbNRuGswYYFw9gJUmOQ2mUpt97ZyDQoDDqOeiMxAfGHghMZbM7jhIcCzQ8tSbB6AoT8se5+Pj41plAg/xwGGZZHBMNdoJmjVKqrDesPGz1XXeMiKcKK4aRmZAfjzXbT0XOV0tzqYn2FeTEv6hT0BAzgv3Ybpcgn92xqk7yu3EfLdZQOhcgCaOcSw0J7E2yISYCE6DCnR5VkapWUs9VDGlTbmHoUEVrz2GYOUqEgP51bWKppBTBbeX/FYVEzAindv+8dIhpgwmGGa/rH2waKaDv61dz+zTabnYFagA0LRrNI1S4tpUHVFmyHvWlQ4IJiAKibo9UW3NDk98jv1ds+PKtEK5BH719D4R/eb6A03pKNqfQA0prAoNZRNk8s0CCsuKzcZJTomR3Hs8S2oPkXUgFqAzzhmL+l5xaJh17YkXqeUs9dEJ5tdBu3BcgRg2WJZ6guRywqJtjaMywwOQS1PZ9tilhocLut1lcoAgv2gnqSPi2Mn92sM8ECSw0iYViAZyd1plZmtTdoHPnDbQMoPMhfyGcv2FVN7zONwYYFw2gYdcTCGcjvkd+rJ+ClgkGEKASiEbYQrRgWOYUlulDbgPG0XincHtw62maD0ZvSoaTM7IBWDWvt3UWn+ZZRYbqLWizYaZw8jO3U2GZnAlKmwJaULCFWoHVQ6yULjdvaUetVE3Jips1UKKNhUVNqm0MG1Vn9GBa97SzcViMlarfoWE66tKycHp6/SygXXtQ5li7p2sTicjAu7hjRWjx//b8kXdzDABsWbsAdufiM94FJpjMjFurv0VvEQkQrcmW0IrDGaIUkKMBPyM/K/F6tjwnIP8akC40QeyueOVuQ6VBoHqV3PQ5TfYWDTbBkOpRe6iyQg//3njTx/NLuNadBSZpHhQqvNT5vqQBUaxxU0qAwaZLXpjNToVBTlVuoLQMrVTEs7FU4s2e70RtCy/LKRaVltFtpimjP2GZJZhlAYlmvY93C3WlClAWprs9N7lytA+XmwS1F9Pbk+Qv01foU0gO66rytN9DwC3n4p06dEvKZ+Fsvqh2M5ykqKaOykiJxzhhKi6mwzPFzx6esTHzXhVIfKijwr1bKVEtcKC6l/IILYl+E+wcKVSVbCQ8wUHZeMWXmlFBkkI8Ip3sK3AiLi4spPT1djA0YEyylQfVqEWmXMTmwVbSQ28TNBwox9qh+aQl45GR9RW0LtyWdmtYXE3W9RCwQqUHaFiYb0Oy3B6SHQMJ114nzdFEX27oZezoNql1j59ZtwUiBYAP2IaIW3eON0qRaili4wrCAUAVkmeGQOJqRR73DHLtuXMXek9miWBn1BAkOpIR1jYsQYx0MKQgWOFO+1133gM/XGZtfzhzakmLCq1cIw33gwXHt6OH5u2neisNCglqm+GoVNixcCCYOkJNMS0sTxgXD2AMKcc/llwg1qJQC5ymJZJ6/QGUGIkNukNLvQfucLyimvKIy0YsjtcD+QTUrp1D0RSg5HyDSZDwN+pagl4q5AMAGU32Ffd2mcfPBRBxqKfgOvRoWu1LPi/oKFDC2jHasQ3HHJsZ9oBfDQqZBQUrT3usSk+gft6TS7hNGo0wPhkX7WMc7bpvTKjrMaFhk5GnGsEDRrSsNC4DJNQwLNOFzJM3IlWw/Zjw3e7Vo4JCDFV3o0Vxvc/I5kQ6lN8Ni+/Es2n0iW9zXr+1nufmlOVN7xdFna5NFoff8bSdo5rBWpGU8f4f1cuCRxASitLRUNN5iGFv5ZPVR+mnLSZEWMaub8/oTfDh/l8h1nTWmLV3avRnpIXJz38cbhLLTa5d3p5a1KPxbuTaZvt10THj2X5jSljwJmgD6+/tbvLlKb31fJdxvb4oADAsUNU4bZOzIrTdkB+l+LR2bfKhToaCWg0iIIz1gXA3SmJYeMBZfT+xmexqUpFucUV53d2q2mMhqORIpFaHaxzp/Qoiu1puSz5m6XGuB9LwiIRuOQ+KqfgRoCLn2cIamC7j3njKmQcEocBSMdTAsMNZdY+PkXCt8vtaYzjS5RzMhQmILiLI/e1lnkeY3sZvlegwtwYaFG8ANMiAgQDwYxla2ncijk7llFBcd4dQGi82iI2jB3gzafrKAruyv/caN/yWepIMZRSIne2C72FpNmkZ1iaNXlx6lBXvT6anJfjbJuLqbc/nFIpUJdG5W36GiRoTb9Zh2iQLk2hpW5iAvGZ2dswpK6NCZPJFCoVVgUCK1A1Ka/WqRAmZsnulLuWgMmZFvUgrSGjgvTRGLxq6JWABELLSCjFbAqAiwsTbMXuTxRjqcllWQQCfF4HcEMdatqHBE6IUTWQX0715jHdWMIfY5fxytOXMn+siDYBgbPH5/7DhJt329leYs2Cc6u+pJZtISiadzKqV0OIuuSvMw5LzqAYR+weW942rtiW0fGy6005Fms2S/NmU5ISkIkAJUm6JWeALRqwReLTmZ0ds1vP2Y8wwLGFZ66WexTIlWDG+Hhnf235bxmS5NlaiFhtOhcG5CFQoeWFl07OyIhdaUoY5nujYNSutSu1KFUK5b56aOGxa9mkeKCBCkmdG5Wi98s/EYoR0FpMQ7uCAVUCuwYcHonpVJZ2nC3NU066edtHj/GfpyfQo98PMuGvXGSk3fZGvyXp/JKXJJLrKcgMBziGI6LQMPPkL84MrecQ59l0wxWag0INMa0tCTXaPtBak+aCAFkCagNzD5h8cdTd+kQeAoelGGWqY0txvTsfaNFeWxR52KVkGOOEDxritS01pHK4ZFhrHTtbf3sDA3qPBbUF/SopMMhwO1UzH1HY+ShwcHmK5tvfSzKFecn0Cvqaq2woYFo2t+3XaCZny5hQ6eyaP6wf5098g2NH1QgvCGYZJy0+ebTaF3PUYr4OXCRMuZxDeEzKO/8N5L6UetgoEYioIDW0U5rAE/sbsxN3XtoQxNyjLuOWE0LLrWIg1KIj39eksRAFuVdUZxp7OUuzrpwLCAFCnqDrDNI9rbpwalpnu80SDdqZxHWuSgqTGea7y1zRqEiMJ3OExOKWmFWpGadUUPC0lMeJC4T2DyfkyJkGgJef1Bqc1ZyLFum07Gum3Hs4SzEOmOjlzneoANC0a3YNL50PxdYuJ5VZ84WvPoKHpofHuac1lnWnD3EJEacr6ghG74bJMpHK0XEtOc35lWnSLSWYlaaHnCBf7bd1r8e1kP+wtaLaULYKKJpkTye7WYClXbiIUselbXKugJuc79atmV1xLqVCitat7LNCho+0eG1l5GUhbFHjiVo9lIpIxYuEq1DMaZbIyolbQgVytCyTEdBdxabZSHng3OSoOSyF4Y249rN0Kn5u/daabml1C28mbYsGB0yaK9p+mBn3cKo+K6/s3plandKhXkwnvz1Yx+YmKBWos5f+0jPSFzwjs4KSXEaoqIMuBrEXgcIcuHGmRHUkTUQMoTSAUerZClKtx2xLCA1CT2FxrtyYaCegCTfpnS4Iz6CnVRKyab6Np+Okeb+2NZokyDMnZPry2YuEaGBohIpIx4ag0ZIUXNk6uQtRtQA6srhoU6HUorBpWrDYuezRVDOi1H1HBombJyA/2jNL/Ug6qTo7BhweiOfaey6f6fdoqwLyIVL0zqYrGoNyI0gD68vpeYWCxPPKur9BDp2evoohuwDElruahVTv57N28gGl85g5EdjJO3dYczhQSp1qIVyD13pBsxjGuZZrIlWT9RC6RvwAGAjuPO7D+AtJgWyoROi4o56BCNxnhgtIPGM7zWpjoLDaZDYXLlTsMiJcPzhgUmvChYd4thoRRwa+08L4WhK1OhnOgog0ogUsAQgZbjp5bTPM/mGtOghrTx7jQowIYFoysw+Zj51Va6UFJGQ9tG00tTularFJQQHSY6VYLXFiVqNh3CfCCWN2BnFbFW1zxMq/tk8T6jYTGus3OiFfLGBhlSnD9yQuctaVASmUqkJ0NariskYZ1d1KtlKU50WkfjRhiUzmj01UOR1NViAbexsLhcyOK6cpLdQkmFSsnM14S8KMCE0tUS1xXKUJ7fbjWInuK4o7lpgnJsnAEMaRm12HFc206Uf5RoxbhO9je/1CPev4WM1wAP8x3fbqNT2YVCr3zetb3I3wZpxvtGtxVdLpHDvTIpnbROSqbxBhwS4OeyG3DbmHAhTZpTWGpKwdES2QUVntyxnYzpS866GY3sYPQYrVBSULSkCCWlgB1B9rPQkzLUxqPOT4PSg2Gx5lC6SWbWGXRVIhZalJJOUtKzMPY4qzjfEgkaMizUaVCu7isjIzWYyGvJWSTr+OAkc3bjxl7NZZ1FlrbToPaerjNpUIANC0Y3Um1If0LHaHh/Pp3WR6Q62UJsRLBJ3u21/5I0NehaQuZHI13AVR104TWRHq4DSqG4lliRdFaEuNs1rif6OjiTke2N6VDLk85q5lzY40TDQhY14jxCt3I9II3Iga2j6phhYZRSHtLWOYaFFHtAnn1JmbYKuJNO57k8DQog+gNOZl3weBG7u+orpJwtbhd5RaWa6uHkivoKCRTkZAG3VsZyc2D04HhAiXFwm2iqC7BhweiCF/4+QP/uPS1ysD+5sY/daQP/N7y1CMUi9UfrKhKy7sHZjfGs1VlosYB78f7TptCxs8HgjvMIzZW0UOiIwu0TWbLjtuOGRZOIEJF/jBokLabEWJLjRNTM39eH+igThbpgWECpDrUl2O4BrZwTqYlrECKEK5BepaUmcSDpjOIwcZEilAT1WBjrcf7LVKS6YFhAaSiuQajm0qFQE+kqwwKOGFw/mLhrMfIOVilZEiPax9SJNChgk0D+u+++a/cXz5gxg8LDXTuAMN4PvBBvLj5In69LFn+/cVX3Wnk1G4QF0oQuTejX7SdEJ2fp1dW21KxrO3Oi3uA3Oqm5Am54GeVgDGk+ZxMW5E/9WzUU3mIU9beJ8ew4Jfc/Jh/OysOGJw83WkT4BmncS7bhiDFagaJtHBtnI50QmfnFwojDWKAF1hxON6VzoOGXM0C6DaJ8cJ4gYuXq6IA9yH5Crl4n7APUWeC6guHWygm1K472sIhzg2Eh06FgzCAdyhXRv9rcv6XjqlMTx50m5qAeCw4yqAfinJeGlZZYddC56Y56wKZRfNasWRQXF0d+frYV1aWmptLEiRPZsGAcnmA+9ttu+m27sVvlk5d0pMu6176fweW9mwnDYuGuU/TMpZ1c0vnVqYpQLircrqLxrzFpyp2p5ym/uIwahgU6JTXIEqM6xJgMi9uGtSYtHG9n9iyB5/+vXado6zHt5h6bp0E5y2tvDowVRHBgaB1Oz6O+Ya75HXtBo0YwpK1zDb/2sfXFJEtLzS9RH4faMeAOYwfpUDAsPF1nISOR8Q1C3PJ7SBtFHeFRDURiQXpeEWUVlIgUrbaNXWPg9YyPFIYFCrgdmR+4goy8IlOa69B22nbwOBOb3UNbt26lmBjbdLbZoGAcBWkLj/++RxSgotDvpSld6Oq+zR36zgEto0wTDDRIm9SjGWmN7AslppCuq2/A0rCAVw+yl87ymjrKusPGCdeg1lEuqzGBYfHsX/tpa0qWyEl2dndze5ATQGcaFhXNo7JEfZKr9qMzPJobZH1Fq2iXavwLw+JsnksKxGuj/CbPc6jbORN5HskIgRZAyiGKWNFnAxKhrsakDOVByVmc26jzkClq7kBGZ45qQGoXHFTqaqDO6CpHHqKzX204pskU5zWKOAPSwGLCg6muYFPC1zPPPEP16tlubT7++OPUsKHnB29Gf2TmFdGrixJpwtzVwqhAruz/pvVx2KgAmFxd3jtOPEc6lBaRkwEYQK6WJ0REANKr6t/VAuuPGCdcrix0w8QDqUcoEN+c7FnZ2STFsGjnRMMCk0uoiqExHLz0WgVGbVp2oVAoc2V6YhtlwnXojDb2xe6T2UKRDQWdsveEs5AOCRkJ0wJJqo7brlZHUhdwyyiJJ8i5UEq5inhCs0j3pOi0VoQukAqlBeTY5sq6mp7xxnFj/6lsTfUmAqsPZtS5NCi7DIvQUNsvjNmzZ1NkpHMHS8Z7gScLeYh3freNBry8jD5ceUQUH8KrvOi+YSYVH2dweS9jlGLt4QxKy76g3Y7bbsqNNhVwa6TOAipGOxTP0+DWrg0dS8Nl7aFMj3o1D552/s0XMsw9lEZzqLPQehoU1jUk0HWpiTINQytGlkyDwjnobOlVeR4hDQfROC1NMN01rsFDDo55MBXqxHmjURMVFujSc1tNS0VyFnUWnlbEAgdVBqWriG8YQtH1AsWcQRaKa4HycgOtroP1FaBulKgzmuRsbiG9s/QgDXttBU37fDP9s+e0GByQV//pTX3o8+l9qbnieXKmp7pfy4YEZbo/dpwirUrNdnCxIpSlRnlaYHPKORFFQOqAs4+9OUMUw0KmpHgCpOegngQeezkZchYyAqBlw6IiDcq1haZSGeqIRpSh4NhwRX0FQHG6TDfSSiRSHbFwB7KXBYwrT8nunnBzGhRABBpRfjjrpCKVJqKxLjzuxkZ5xrFOOqW0IrObmV8s0mylLG5dwW7DIjMzk+666y7q1KkTRUdHi5Qn9YNhagJ5zo/M30VDXllB7yw9JCZXSAmYPiiB/rl3KP11zxCXqAFJLlUKvJYeMHZ21hKyp4SrC7cl8ne0Ijm7XplwuTpaAaRqCm5+MHI9OeGCclGADc0ea1VnoVHDQtRXKIpQA1xtWCipUMKQ87AXH+kaO5UJ0CAXnecyHUorhoX0XLsrYgHDCh2+4aQ45SEZUllf0cyNhgUm2S01kg4Fj/0hmQoV61plLtmBW0uN8lYdNDZgHdwmyulju9axu2LxxhtvpMOHD9Mtt9xCjRs3dku+JOMdoJvy20sP0jcbjwmPCujVPFI0rxvfOdZtKk2jO8TQU8oghJqOqHquLya0dSCWEwFXS82qJWfl5BrHxJUdcW1h3WHjRHNQG9dLJaLGBEV18CxhguuJYn5TDrILJlzyZotCTi2d5xKk353NLRITQFd79ODFR0oKPIgoJHZ2XYM9YNwpLiunxvWDTLUAzgYTeKieyW7XnhakOJVtNNzbuiligXq6Fg3DxPWFCbYs5nYnUoTD3RKoMCwwphmVoVznoLM1GoueQa7e/700GLFYraQ7DqtjaVC1MizWrFlDa9eupe7du7tmjRivA57JP3aepOf+2i+k58CYjjH0fyPaeKSfRNPIEDGhxsRmRVI6XaEUdHsahK4vlJRRkL+vyyYc5mDAR5Evfhc3YJky4gkw+ZW1Hq7y5FpKh8JNGDnvnjAsXJmDHBkaKI4nIoSQnYXxriWWHzB69Ia0aeQWpwL2RWbyObE/PGlYbDp6zhSlcZVjDpKzWingll7rphHBLhekMO9EDcMCAgGeQDbngxCHO5HKUJ6OWEi1O/TWcLXHvltchHCKQQgCtZNoEqqVqORgN93LtITdR7tDhw504YL2il4ZbYIUk5lfb6P7f9oljAo0b/r2lv70v2l9PdqkDoYNWJ6onXQoWeeASSaKb90BBmNZz+HpAm6Zbw9vK7rnuoPBqjoLGMDuJklRKXKVaoqUVt2SbJzMaomliWcrXYuuRisduGXBev+WrovKmSRnz+R65LxWI40bZ6qe2YKsWfJULwsZsXC3YdFaKeD2dOd1V0ZjzQkN9Ded89uPeT5qsV2JSqLmBQZuXcPu2csHH3xATzzxBK1atUrUW+Tk5FR6MIwpSrHjJI19a7WoZUBx6oNj29Hf9w51ScGivYzq2NgkB6cF9QxwwM15yFUa5XnYsJCTX1fn25tPvBGqR6qGu6UpUVQqi4lddfOVTedQFK8l0nOLaFeqcQIA9Td3IDtwe3LCBU/mDmW7XdUQUBpRyGo8X1Ai0s204Ll2dxdwWcDtuYiFkgrV0M0Ri2jZyyLP6xWhLKV+olGep9loqh1rWCfLBew2LCAjCwNi1KhRomFegwYNxAOv419X8/7771NCQgIFBwdT//79afPmzdUu/8svv4goC5bv2rUr/fPPP1UmwE8//TQ1adKEQkJCaMyYMXTo0CEXb4V3A0/N7d9so1k/7RT5tchjX3D3ELpndFvNFDF1axYhvOKQY9zk4T4GkkRlYu+uwm2tFXBvU24IfRLcF8mCDGSvFpGVlHrcBaQw4dUKC/RzmVdTRiz2nszWjPQoWKFEK5DCEKP0UnE1aJIHUGPhya7ycGRg7JFFtq4AqWXSY+/pAm75+67sZVB9Lwv3G5K41mDUeSJikRBt3O6MvGJx//UUro7GWquz0EIB90ZVumNdxO5Z3vXXX08BAQH0/fff07Jly2j58uXisWLFCvGvK/npp5/ogQceEH01tm/fLuo8xo8fT2fPGm9S5qxfv56uvfZaUWi+Y8cOmjx5snjs3bvXtMxrr71G7777Ln300Ue0adMmCgsLE99ZWOgZlRg9k1NYIprbjXxjJS3ef4b8fX3ogbHt6I+7Brt9smxLcd8opT/GMiXX29PIlAF3Sc2aF3B7MmJRUFxqUsSSNwh3YZKdVYrt3EWS0pUWBa2u6oyNeiLovEMrYauGohbLlBREd0Ur1CkimGii87Un06BcWV8haRdjHEcOeTD1C447d6bEqGmhGFap5wrcfrylIhRqSsKD3VdXAvB7Um7YU3UWpW6Ixpoj7xt7T+VQUannGuVdKC4TDoS6bFjYXbyNSTkm6e3btyd389Zbb9HMmTNpxowZ4m8YA3///Td9/vnn9Nhjj1VZfu7cuXTRRRfRww8/LP5+/vnnacmSJTRv3jzxWQx677zzDj355JM0adIksczXX38t1K7++OMPuuaaa0jroEYgI7eYyIfI18cH/xDuV+KB/5R7l4/6PeV141xGPvcR6ixhQf5CdxkPPA8P8q920gOPCC4ipD39uzeNCkuMA3j/lg3pmUs7mxqwaZHRHWPop62pIlXrmUs7eTRkCQ+X1B13lyKUBKlX2HSkTGTkFVG0B9SDdqVmC1WqJhHBYjLsTlBn8cbig6LjtzuVsdzRlRb0S4ii1HMnRCf7EU5sNllbcNOHYhEYo6QkuoOmESFijMMYlZp1waURg5oNC9dLs6Mp4KJ9FcXTnkp5g+ce15RMRXMXTeoHU6C/r4gQoag3vqH7ct1PKs3x3B2tkKBgGuM5lKFko0x3cgwN+srKhTCIu/YBahmg9Hcuv1hE32VvC3ezo47XV9TKsOjTpw+lpqa63bAoLi6mbdu2ia7eEl9fX5G6tGHDBoufweuIcKhBNAJGA0hOTqbTp0+L75BERESIFCt81pphUVRUJB4ST9aWvL/iCPU98RW1902t8h5K9srIhx4sudP02i1+/1AX32TT3z5kMBocZKBCMtD0kruojIwKLTP8/qV+vomiPiLQ10f8K8RbDAZhlM32mUWHlUaX0/0W0ce+Oymsnh+1xAUeEkA+S7AGBrE8XfklUZhSW7Hlf0T7FyhWjq/xQfK5D9ElbxFFKAo9e+YTHVhQ+X3185FPEDVoYVz24GKixL8qv6/+/gF3EDVsZVz22AYamfo3PR6QQqW5PpSzcDlFhAZVfKb7NRXLntpBlFg5hc60h7Ft3a4iatS+YtldPxKVlxEZyokMyr/lyvM+NxPF9zMue2Ir0dq3xfulBYX0RUAmBfv7UMNfPzUuO/BuonbjK5Zd9JjynWrvm2JFDriTqOsVxpfO7CNaeH/l903/wrVzk3H7wLmjFPbXffRr6HkqKC4jv2/fRyVcxbJdLifqeYPxeU4a0V/3KV+rfKf6Oda19zTjawXniBbcY+20JWo1gqjfTOPz4nyK+fcW+iggj2L9g4l+/Lzysi0GEQ28y/i8rJRo/nTr39u0F9FQ1TX/ywyicispQI27EI14VDRkhAH9dNk8yv3mS4q0pFwT1YZo9NMVf/81i6ggs+J4yAeOe4MEoolvVSw7/2ai7BPGc0XFpIx8ahcQTmdiP6p48Y+7xDERmAxd5diFNCC6+puKZf95hCj9QMUy6s8E1iO65jvxtH+rhtRq9xs0aPtJovTIyt8LfP2Jrvux4ntXv050YpvZOqiWv+orIj9lH62fR3RsvXItYB+ornn8e+VXREH1KpY99B8VFBTT55RNgSE+1Pm/CON7WP6a74nCFA/fhg+I9v1Wdbvkvpj6KVFkvPGlbV8R7f656vrK83Li20RRrYWDZHr9bTQw5z8Kn/8hkTCgzc7hsc8RxXQwvoZrfofc36rrRy4/7BGiJt2Mfx9dSbT1i8rfpX4+8E4qjOkh5DD7+CTSZYd+Ijruqzp35P4rJxp0D1HCEOPnU7cQrXnDbExTjYG9pxO1Hmlc9sx+43iijHlXZl2gOP8sijoURLQglqjzlIplcY6tf69i3xufVGxf+0uI2o0zPse5u+pV68u2HU/U6TLj87yzREvnmJYtzymkNwPSxfUVvHABUetRxvFSjhEY0yTmRebYB3I8Kc6vPJ6YL9t8AFH/2yvGiF9vIV8y0KfB6cJpE/zHt0T1Q4z7pmnPivFEXsvY7/JeoX5EtyHqe2vla64kXxnP1eN7GVF0W6JRT5rqK94JmEdtCkuIvq6nLGeoWBb3likfVnzvj9cT5Zysev2AiHiia3+oWPa324nOHTG7byrnRmiUuD5bRtcT6TjNt7xItD+94r6mPo8Cw4imflLxvWveNJ5D6nunOI8NRD5+RJPfr1gW13LazqrrqxyXgx1eFP9CrMV304dExzFGqFFdS1M+Mq4L2PIZUfIq68viWg5VjPKd3xMdWlKxlI8PfRh8js4UF1L9v78mumEuUbiihLf3Vyv3cAWM7XIecWAh0X7jHNEiwx8znhfg0FKiPcrYoxCZlkNvBeRS8+AQ8jnbhKhxJ+MbyauJdhjHZNVKVzzH+YtzE6RurhhP1Ay5n6hRO/I6w+Kee+6h++67T0QBULOAtCg13bopA62TycjIoLKyMhFNUIO/ExMTLX4GRoOl5fG6fF++Zm0ZS7z88sv07LPPkhbo06IBTcw+RF0KlYmAGWXkS782f8o4ppGBxmcepX5F5hd5BW+FPUDni30ov6iMuvseoQl+W4xvlCsPVcrm2UKEOusJj8QloVnU99xuIszhLAktlaoKCDMOWxg8VIx7UfUj+4n2/2l9WXGDUAaE07uJtn9tfdkuUyuMhbSdFLDxPbpNqlya7774/qpldxOtfs369zbtUWFYYNs2qSaK5uDGKg2L3NNEiQvFU/iURgqjDZMUZdnOUys+V5RDdEI5FpbIU+30olyi1E3VrMNI1bJ5YsDrhef4ffPTvkmPiuclBWJiaBU50ZPHW9k2i8ibAygvpdYZK6g1fh+Re/PLGcaGCQPRARiPVsBNWw2WLbeSZ4x9iq/386UBraNo7JGtFJFspdAzTjlmkoOLiHLTrBssak7tNE4EzGiNOb1PY4pURyzSdhGd2WP5e+uZycXixm7tOAcpE3YlenjMJ5l6Fu8hqroaRsNCzckdRAf/JauoJ3VYh6S/rS+r3vcZSeJcgx9xAOYs+JrjqmXLVGPE+ePVn+/q8eT8MaJja60vi0mpQueQczQ8f3fV81yiNkqzUoiSqpmM9J5RednqJiOdp9Du4pZUVFpO7UNzKPxwNWMajHlJ3mnjuWbLtYxlVZOc5njg0CKrdzsRNepQsTwMgK1mBryaiLgKwwIGQHXjar3GFYYFxp6dFZMnnLGX47rGfWEXij8iKwyL0kKi3T9Z/96A4ArDoqzEODm0hq9fhWGBE0s5FsPxPz+z86zkQmXDAsajNedDq5GVDQs4jYoUb5qlMUIxLJAKNd53PzW+cL5iPLdyTpqcQVkVDr/Ky5qNSWf2Gh+WCG9SKe2vYdZOorT9lpcNrhgjTBNfGMiW8A2obFgcW1fttXEo8tGKwu2T26ofsy9TDFw5nlR3v7/oFdWyuxXnQwX95fE+Y7aPT+8l2jvf+vcOVhxmAM6aPb9YX7YvHGKKYZFxsMo53AkPrANuL3l3VhgWmUeIdqscOOZ0mFhhWGA82fV91WV6XOedhsXVV18t/r355psrWYrwYONfTP69HURN1JEQRCzi4+M9sy4XdyRKuM/oVbKAH/nQ9wMHVLyQeA9R1kSzpSq8YCv6jibyM54WxQcDKffMIbpQUm58lJaLNAIcZ38/P/q84zBq2zTaqE2eWo8o8xIz77jKywdvqwTe8rg+VT110utZT9VQpv3FxsHS5BFRe0XLicKNXbRNHq5RT1VdVn5/fdWyTboLz+DWlEzaeTyL2jQKpRFoZGNaVtXTADdkMZhYAF4deKgl8HYOfdDo4cHNTu39wt+xKsM7tqsxOuPrR7/tSKP1yVk0vH0sXdojrsKzZlq2m9Gja4rAKMhtxTpKotsRXf2dRU+ScdmOlY2Byz+jRXvT6J89adS3RSTdOKBFxbIxqmURcZr0fmVvmvq5+nuD6xu9S9bAOspf8Quil3xmiojJHcNbU7x5p1pp4Mn9jX1mjUhMp1Rc/JpZhEeF6txBncWLSddTx0ZBNGOQ6nhK6pmlEOE8K72gOr6qY60+1+U6YDIjj5vBIELl9/ywnQoMQfSWOgd53PNGg0d9vOTzSgYWEY2YbYyaSNSfURkLzRuG0htBU+iPgsH0fyNaUVsl9960vHkKYL9bidpfVNU7Lf/GeSzpeSNRi8Fm0UTVvwGqVIBe06mg6WB64s99VFJmoAfGtVc091URGdOyN1Z47S2da+GNK0/EhTGnPs9V66syeM/Hj6H7TwVS/5YN6Jq+8VW/V32uIap26btVrx/5XDoTpCNiwutmy6jWO6Yjbd1jrHEJiOtB1OmVyp5m9TjRrHfl6x4TLzmOVRrXDMbflTRsTTTuBdOypeXl9PbiRHFvvnNEK6qnXhaGw4jHK/42j2K1kPse+zrWeL6rUZ8zaqMbx3BMhdNt4Z402n0im4a2bURDoQaIcVcSVN+4vhVfWvn71WNPQEjlSaX5slGtVX/7El2MKA+J9NwNRzJoUKsouqhzjHHfNWhZeVtGPq6KQKj2sYwsqBn+CFFZceUxXT6X3nFELM5foOdKbqIre0TTiPaxynI4zsry5pP6y941jhGVrh9lO7HtarAf5BhR6ZwoN40RSIUCX/hdSc9PwrViqLq8X2Dl7+1zC1HbcVWXE3nWZn1mesrrUx3Jq1j3pEOq+oqmNxC1GEhWUW9f1ysr3yPNI1NBqrGy46VEDVtWWu7YuXz6Yl2KSOV+MLRhxVmCaLr5GK5GPTeAMaket6q7xyCarnKEQunv9cVJIqX2rpFtqCEi3RLMecY+b/171ec79gGip+ao5xoaxsdgp8j1sWPHqn2/RQs5KXF+KlRoaCjNnz9fFGBLpk2bRufPn6c//6xq5TZv3lwYALNmzTK9hsJvpELt2rWLjh49Sq1btxY1Iz16VHhmhw8fLv5GjYYtwLBAClV2djbVr6/dmgLGcs7zNZ9sFHUFW54Y7bE6i8s/XE/bjmXR3Gt6eKRR27IDZ+iWr7aKfP//7h/m1t9GHvCoN1eJxoB75owXedHuBr0Nxry1Svz27mfGubxhG1SaJr63VuQEb3tyjMvPu3t/2EELdp2iWWPa0qwxnvN4fbU+hZ5ZsE+cZ4tmDXX79fbXrlN0zw87RA+dX/9vkFt/+9avttDSA2fpyUs60q1DzSasLmLUGytF5/VvbuknJvfu5rJ5a4Vh8dENveiiLkZvujv5ZuMxeuqPvaJXCnonuYtJ768Tcsof3dCbLuri/saUKNqGiApqivY/e5HLxCGsMfrNlXQkPZ++vrmfWztPQ86525zFwnGz+uGR1NzNNQ7oh3T9/zaJWsH1j43yKqlZe+a5dt/BYThU93AVgYGB1Lt3b6FEJSkvLxd/Dxxo2RrG6+rlAYq35fItW7ak2NjYSstg50Edytp3Mt4FtK8xoUXRsqdkKMvLDSZJRk+pZ8nfxT7A4OxOYFBJ6VFPGBUydaBx/SBR6CnXx5XI440cZHfcfPq1NKaebVD01T0BfFg/bDbmpFzbL94jN11ZQAxD0p2N4/Bb8rxyZ2NQFHCDQ4r0pzuB17aih4VnxrUKydkCj6hCxZlHX90Eor6oiUSGQVqOexUucf+Q+9vdSmBwCHWNM0aDtnhABc+dqm9axqa7+IIFC6ikxHY9ZPSKcEV3bkQfPv30U/rqq6/owIED9H//93+Un59vUom66aabKhV3oxZk0aJF9Oabb4o6jDlz5tDWrVvp7rvvFu/jwCOa8cILL4ht3LNnj/iOpk2bVoqKMN5LkL+fqW/Ceg9NulDohwJDNGrzhFINgIclMjSASssNbu9MvP34eY/IzKrBWCC7cLujn4W7FKEkIg1FMeIgC+0Jdp3IFpLKMOSn9IzzyDrI6wtqdlCPcRfw3mYVlIht79zULA3GhbT1oOQsZF4xsYXXHOl4nkA2yTueWSAMHXdNrOGo8qRhgboxuc8REXYnaECJfY0UaSl7607k/XzrMfcbFtJxM7COyszaZVhMmTJFpBvZCtSU0tKsFDY6AOo73njjDdHQDqlKO3fuFIaDLL4+fvx4pd8dNGiQ6LfxySefiJ4XSKNCGlSXLhXFlY888ogoSL/tttuob9++lJeXJ74TDfWYuoEcBDzlzT1wOsfkXfRUA0FMrDsqXsX9bu5nsV3x5PZyoye32n4W7jAsZMTCTR69FlFhIioDw3HNQff265D8sMkYrbi4axOKCHWvtr+6IaKUv0SKkLvP8e7xkW6NylVELHI91pcHxo27JJwtOUzguUdqzGk3ee7RIBag8aWoP/QQxvol9/eykFEqd0Vjzenbwhid3ZKS5fZeTLtO1O3+FXYVbyOMO336dAoKss36dGVzOUQbZMTBnJUrqyoaXHnlleJhDZz4zz33nHgwdZOBrTGhPEgbjmaKtCR356PKxnTu7l9hKR0K+8CdHbjzi0rp4NlcU1qaJ5ERiz0nsym7oMSlk19580UPEXcxumNjOpJ+VDSnu6Sbe/PdMdn6c9dJ8dxUNO0h0IEb64MGXrIzuauR3lN3pkGZRyykwIq7SFQcJkIZyEPAc4/+FfCiH8vId0tPBUSgQVyDUI+mw7RSonPYdncio7GeOu7yGkPkHVFJ1LG5g+3HzgtRiqYRwaIpaV3GJtcJCqRjYmJE4YYtD3Tn5iJmRi8gtz800E80cpJeNneSqHSc7ujmjtvmyGaG7uzAjd9CqjvqG2LCPRslbFw/WHj1sT4bk10XvUIaDhp2ya7b7mKk0hxvZVK629JCJC/+vV+kxfRLaGiq9/AUUorTnTVVW5WIBeTB3QnUgeAnwTmHZnXuZO9J4zjSpZln5wIyHSo5M9+t9RXNPJQGJZHKUO6uHTyo3EPdXV8haRAWSG1jjNEad9TLSbi+ws6IxRdfWGjUwTBeAtKPMNnBhAvdl93dLVx69jxVuC2Rhg1Sodzl3YQ6EkCTOi0wqHW0yIdHWtz4zq5Rc5FpKfBs1Q8OcGvucXiwv/DiIWTvrpqWtYcy6J89p0U6zLOTOnv8pisLuHGc3QH2t/QauztigWJWpMEhHQZRi5j67jPe958yXtvurCmxhOx+fMxNBdyy67an6is8nQrl6YgF6JPQUJzvW1PO0dhOlfuUuQpE+8GAOp4GBTyT0M0wGkN6Ud3p4ZCpQMfOFbg9LcZa2gTykXMLS015wq5mj+LV9PTkQzKotfGmAAPTVSS6ub5CbUAPV6Qflx8465bfhMrWMwuMzbzQH8XTxnNlwyLPrfUVbWLqUSS62rsZ/K46/c5dxtQpJSrnbkeNtYhFipsm2DIVyh1pV7YIFWAsd5fSH+5ncvs9aVj0VQq43aUMJeorUo31FQOVe0hdhg0LhhEpCg1NKQvulKGEd0f0zwoPoqh67lfQUIOiUjnpOqCkZ7mafYpXs4tGIhbGMDYmYXl0Ntc1tWImCU4P3HhHdTCmQy1PdL1hgXSrJ//YIyIDUWGBdP9YbXSMbR0TplItKnNbGlRvD6meoYjW3cpQ8rrG5BbNyupUxEIjqVC45uoH+4v7S4qb0sDkOdYoPMhttQ2WkLVTqJfDpN/VwCEJYQwYk3EePu5agA0LhlHqLOCtRx5y6jn3eOsrFW5rwJMLOinr4Y4Cbkzq5I1IK6lQyM+V+8BVKmFJHsxBHtE+RhhOSHc75cKoVGlZOT3w8076eesJkeP/4pQuHlXIUdOoXpBICUOZiTsmm9uPK4aF4kV1N7KAW+a+u7O+orOHoxVqzz26MkOcw53F254EKYemdCg3pf2Z6is8GK0AmNxjko9ianeoQ8n6iv6tGno81VMLsGHBMEousvSau1P/2lS47eE0KE8UcOM34NWOrhcoire1gkyHcoVhgWhYhRyj+485vIjSm/fL1hMu+Y09J7Jp+hdb6M+dp8jf14feu9YzXZetgRu/u9KhYGBhf4BeHlI9k+eZMTrqnmjsXo3UVwBMMHEeQjzgjIuikOrUP/kbnk6FUhdwu0taWQv1FfIaN6W1ukE+XPbA4voKI2xYMIyCVGyRqQt1qXBbItfDHb0s9ipREUw+tOTlQQG3qxomIiKGRmnw4svcd3dzw4AW4t/vNh2jkrJyp3wnVIf+3HmSpn+xmS6dt1Y0GUTDx49u6O12aVtbMBkWLk4PwkTrQkmZiJC0iq7nsdQvFM6jdkqqkbkaGfH0tCKUlJyV6SkpGa6NUKVlXxCpR2iECIeJp3G35Ky6h4WnkfLh61xYLwdyC0tot+I8kMZMXcfu5Mea+j2geR3D6JHeLRrSp2uSaZubGuvAeygjFh08LDVrblgcP1cgBsxwF6oW7VMUobQw+VDTt2VDMRHDPkAePnTwnV24jYJSRMk8wUWdY0UO9NncIvpv32ma2K1prVPZFu8/Q79uOyGaCiLHGMBomtSjGd09qo1pAq81ZJ2FqyMWO5WCzu5xkW7vjyMJ8vcTE0ykHSINr6mLPekYN6QSkRYiFgDKWCmZBXQsM9+lxbXq+gotOEtkKtRhNwkVuLvxZ3XISf6+Uzl0vqDYZcIJm5PPicg7ank8nf6mW8Pi999/r/R3SUkJJScnk7+/P7Vu3ZoNC0a3SClINGyDB9bVOeHIxc0tKhW1HVqZgCFVJrZ+sOhSi5sEZPtcnS6hlfoKCYpNu8dF0Pbj50U6lDMNCxkJ6ujB3HMU6V/XrznNXXaIvl5/zG7DAgbFF+tS6MOVhymnsKIwEtrxYzo1pqv6xJvy2rWKvN5cnSKy8/h5TTR/RD0PDAsYtiOVAn5XIYUfIKfsyQJeNQlRobQKEQsX19Ropb5CIvs5IDLnaglxKIHBWaGFVCgAaWVsP857jOMTuromcioj2xytcMCw2LFjR5XXcnJyRGfuKVOm2Pt1DKMZ4MWF1wEFnSi4lA3FXIWsY2gjZF61k5WIfhYwLDAJdpVhUVRaZvJuacWraR5Gh2EB2dmrnNgpWqaIyAJxT3Fd/+b0/orDtDnlnFgnWyVBl+w/Q8/+ta+SpOblvZrR5J7NTN5RPWBqkufiCZeMWPSI96xhASnrhbvTKElJvXRHb5rOGnIYIGIBELFwJScUQQQt1FeAhOgwUV+SV2RMg3NltEqm9TZvGOpxJTD1OA7DAqmZrjIsEK1Vp9AyTqqxQJftZ599lp566ilnfB3DeDxq4Y50KJkWo5XCbXcWcB86kycUOxAV0qI8n0yXgDfKmQWvMmLhaW1/dBm/qIuxAeBHq47UuPyF4jJ6/Pc9NPPrrcKoQLH9m1d2pzWPjKQHxrXXlVEBmjc01h3kF5fRmRzXdKTOKSwxpaB42rBoH1u/0pjjnsJt7aQ4ygiaq5vFyVQorYxpcFjJbXd1HxOZ1uupjtvV9yVyjcJfZl6R6Zri/hUVOM1Nmp2dLR4M4x39LM7VucLtKgXcLpSc3auqr9BCLrI56EqNlCGE9p3VoRkpREeViaanIxbg1qGthPTsgl2n6Kctx60uh/Ng4ntr6PtNx8Xytw9rRSseGkGX947zWN2Ao+DYtlBS3FxVZ7E7NVsU8sY3DPF4jxrZfBNFvM4q2K+xcFtDkUh1LwtXKmOdyNJG1201Mi3psIuFCkz3Mw0ZFv1bRYmaLxiUrpDXlt22cX1Fe/ga1xJ2x6vefffdSn/jIk1LS6NvvvmGJkyY4Mx1Yxi300fRmkcKA27ArkxROqCxwm2JnPRC0QZFafDsOhs0LtJSYzxzUFgNlTB4ujYcyXCKghNSv1DfjMZVaIjoaeBFf2hce3r9vyR66s991KlJBHWNi6g0tqOW4pV/E6m4rFys89tX9zCpregdRFlQYwHDwhXbtDPVGPXsGe+Z/hVqkJoTFugnIjToQN3WRTnw6Lwse9No6dpGzQOGMSh0QZkN+feuAF2utZQK5c7O6zK1VSs9mQAi4l3jIkVXbKQsXdnHeWmtYN1hWV/hHWOixwyLt99+u9Lfvr6+1KhRI5o2bRrNnj3baSvGMJ6gTaN6QhoS0owYKF11c0Q3UNkNVWsRC+QjhwT4iZswPD2ukEXdq0GvpqUwOgwL3DxuHJjg1DQorURp/m94a9px/DwtPXCGbv16C00f1JKGtYsWr/2yNZV2KTKKYzrG0GtXdNdMMa6zlKGWHnCd5Cz2oRbSoAAiS1DqwTohdcNVhsWuE+eFMwKF27ERrpm81zZCBaUmND9FAbcrDAv0LJFyvlop3gZt3dB5Hcdc9rCQ0TGtMLRNtDAsViSddbphAacT4MJtBw0LKEAxjLeCGzAmAmsOZdCO41kuMywOnkHRKInwqdZCqIhQIE8WURtMhp1tWCASJOs3tOTVNGcQvNiLD4pwNzr2Opr2I1NEtGRIYpvevKo7TXl/nfDev7ookV5dVPE+9PifvKSj6H2hFWPIWVQ0yXN+3j2iPabCbQ8rQkk6KIYFHCaXdieXGlM9lVo1LQGJZ6NhkU/9WjpflOJMbpGYYEPlTwsRySqpUGdcJ1SAong0IAwO8DUVymsFKNXNW3GYViWlC9EQyC87KzoFIxX3S3TcZirQjhQNw2govx5AFchVyIk1FJi0iCsLuJF6gg614UH+pjx3LdKtWYRQN4H0sDMaBsp9qYX6CvN0gd/vGkzPT+5CQ9tGi4kRJqEwKNY8OlJEa7zNqACu7L6NAvfM/GKxL7VyvNsrE0xXFnBvV5qLyjFUS8CwAEgFcwUnzhnrK6C8pKXaI2w3lKEgbQ61P5f2r2gc7pLUWUfHcRh6SAOE7KyzWJ54VvwLR6Qr+z3pEW1ogjGMhpCa84hYuIpEk2GhjUmHOwu49yjpNTBetHQDttSxF55N3EBwQ3IkuoKIh8mw0JBajtq4uHFAC/GoK0jJWaSvQI7TmRKZO5RoRaemER5rhGhNGSrpjGtEGeANl9vdSyNRGmsF3HWlvkKmgUF2FsXbUONrEuH89TtwWptpUAD3mNEdG9MPm4+LlM8RTpKRX37gjPh3dEfXytLrEY5YMIwZstgSYU7Iybm0cFuDAzGQXlZXRCzQCVXraVBV5QqNubS1BV284THDTR5dkBnPg0680fWMNSPJTk6Hkk6Jnhqor5BIGVCkA8GQcjYYL9EkDee4FnvTmCIWLuploTWpWUuN8lxVwC0dZR0U41VrjOvUWPy7dP9Zp6iCoUZynRL9GNPR+N1MBWxYMIwZEaEBJm+mzJN2JhjYDmhUalYCgwfZL5BbzXCycaWWmtU6Upt8U/I5kb5VW2QqFdJREAlhtEErF6VDaaUxnhoU3qMJqDp1xRVpUF2bRQjjQmskRLtWcraiaWSoZg0LV0nOarVwWz2Ohwb6iVSwvScdd5ZB0AP3AxiRct8yFWjv6mcYTdVZZLkkZA7VKeS9yjxvrRGmqn9wZtQCxY0yYoEJiNbpGFtfeLULissc6m2i1fqKuo68/g6ddd5EGxMOeY7LtEqt0EVJw5PGvTORY6UW06CkUhOcJYjWoP7F2Zw4b0yxgvqU1pAqYK6IWEBiWKaXaak5nhqkIw5r20g8X7L/tMPftzxRSYPqEOOV9WeOwoYFw1igp2JYSJUTZyInHVBb0qJnz5UF3MkZeULGFnK2LaO1aVSZ5+cOa2e8Ia06mF7r75ETOa0W69dVpIfVmR58XC8wLhAhaK4xcYJucZEmWVhnI8UutFi4LSeXTZX6AlcUcMvJtazl0BJqyVlnR2tktAIF0p5uBFkdY5V0qMX7jUaBI/Vyyw4YC7dRu8FURbuzGobxIL1aKDfgVKMuuzORk0yte+zhrXd2AbcMQ8No0Zp6iDWGS8MiKb3WN6JtUi1HgzKcdRlpWMiaJ2fWV3SPi9CcN7Ob0gBxtyKg4CwQBUhS0ju1fI7LdCjUgzhbQlt2dtai0l3L6DAx3iJSfibHuamticq1o9VohWRUhxiRJQBVNEccCXAMIkUYDSdZZtYybFgwjAXaxoQLlRgU3Do7fCy7Tqu7HGs7YpFb54wqNUPbNhIpFLghnVYaYNkDvIQ5haUix5dTobSFLDZFemJOYYmT6yu0N8GWEQvUlDizgBsOGPhfoIjU2EVdrZ2B7LGAvgvOBEYFth99X2Qdi5ZA7wYZSUlUDEBn38+0WLCvpkFYoEnBCc0/awuUpeR9wVk9MbwNNiwYxgLw7nSPj3B6OhTC0BXFy9oeiGVh+eH0PCosKXPyTUg/E2yktHRXJmSra5EOtSXlnKmQlwu3tSfU0ETpEH3QSelQ0rDQWn0FwKQXXbGRDePMOovNyec0H60ACVGuiVjINCikvmktSiWRE3+Ziuss5Hkko2Fa5iql8/bvO07WSowD9++/dp0Sz1lm1jp8l2OYGmRnnVnADVWKjLxiYbho3XuNCVdkaIBIBYP+uaMgJUimVWndu2U1HaoWhsVWxbDok8Bhc02nQznBsIDcqpy0dteQIpSlqMVuJ9ZZrD1slGMerKio1bWIBeSkjd+vvTQo88J9Z6a2opO1jIDoIQqNcRzGNYr3ZYM7e0BK69GMfBF9ntC1iUvW0Rtgw4JhaqizcGajPNkcDhJ1WmmcZQ143rooBsDuk45PQqAfj+6vSBdopxQT6oXh7Y2GxZpD6VRaZp+na6tSX9E3Qdve3LpKB8XAl1r8jqYEgVaNwkTTQS3STYnE7nJSnQVSyGSUZkjbaNIyspdFcka+U4uYUxXDIl6D9RUS6czZe8p5kSrUKpSUGahBaIAm+3eYg4jx5b3iap0O9bPymUu6NnFqQ01vgw0LhrGCzJE+kp5P5wucI0+olzQoiQxvS4PIGWlQqN3QW0oQUqEQvUGthD29TdKyLwh9e9SpS6UxxnuVoXZosH+FOd2aOTdiga70iGqiQBiSrloGEQVkKqGIGZFjV6RCaRWZfop1dVY9kRQBwP1Mqylg5lzZx2hYrEg6S2dzCu2S1V24O008v6qvMaWKsYy+7u4M4+bcetwsndkoz1S4rTPDwhneTWmc6GXb1SB1rUIH3Xa5wq0pWaZ6FfZwabuWCMX5jnqxtdhx2xwpGoEO3FlO6Oew9pAxDWpIG21HKwCixPGK8ePMZnF6SIVC8TKK652ZDqWn+gp175o+LRqIYvtvNx23+XN/704T/YwwJ8DnGeuwYcEw1SAnCFKj3REwadmjyK1qXRHKPB8bylgXih0r4N6tM6PKnAldYsW/8FrZOgGVMrN9ub5Cs2CiEODnI1SSZPfk2gCvvRR66N1Cu8cbKVrSYSKvSWfUVwzVeBqUBP2DpCiFM8BYIFOhtByxUEctnFW4LyMWXZUomF6YMbil+PfztcmiLsqeNChEPPQSnfEUbFgwTDX0VDwTzqizgH54Rl6RLgq31QXc0fWCxKRpvwM56Cjc3mfybunrJiQZ2SFGaJdDmlSmvNiqCNWH6ys0S4CfL7WJCTdFLWoLilhhnIQH+Wte09/Uz8LBSOyJrAJRr4AxbYDGC7clrRsZjaojTopYnC8oEbVjQOupYDIF1xkRCygFSil2PUUspJMIRhau1w9XHrapdgq1cjjPr1BqNBjrsGHBMDZELJAKhcmxM9Kg9FC4LYFnBo2+HM3JhpIGeoIEB/iabux6A8dMdm+VkoPVkVtYYupa3kfDHmwGzSBlnUXtJ1wy7Q3OCK03f5RRQ0dTPGUaFGpK6gdrs1jdWsQCvTycwTElWhFbP1jz47opYuGEAm6MbaXlBoquF2iSbNYLvr4+9PD49uL5VxuOiVo4a+C+/8yCfeL5pO5NKUbDfVq0AhsWDFNDYWdIgJ8o9nP0RrRHZ4Xbkq5O6NYrQ+9QJtFb4baaid2amvJta+rIvjIpXeTxIu86Vmc33rpGhyaOS86a1L90kH/dr2VDU/8JdI2uLWsO66e+okoqlJMiFsd1kgalvvdg2x1NbVULkegxNQjSs/0SGop+Fu8sOWR1uV+3nxAGOKLVj07o4NZ11Cv6vcMzjBvAJFiGeR3tZyE9/nqrMZDN4RyJWFTk4upr280Z2i6a6gf709ncIlOakzX+2HFS/HupYoww2u/A7Yjk7DblfOitg7Q3GPiQCEUKj5TItRdMyGTEQi/1FbJ4F6RlFzql+7gepGYlMeFBIsIAh4ejHbjlmN5Np2M6jKFHLjJGLX7ammqqoVAD9axXFyWJ5/eMbqvprvJaQjeGxblz5+j666+n+vXrU2RkJN1yyy2Ul1e9x6GwsJDuuusuioqKonr16tHll19OZ85UKLrs2rWLrr32WoqPj6eQkBDq2LEjzZ071w1bw+gJ2U3WkQ7cCKduVzyavXQmOyojFkhnQnpPbdij9MHQu2ER5O9H4zvLIm7r6VCZeUWmZnqTe7JhoRdlKNQL1GayibqbU9mFIgVKy1KzEqznYCXKsFoxDuxl3eEMyr5QImqw9CSlHBkaKCbX4KgT0qFksz0tK0KpJ9MV/SxynKNwqNOaOdm09P9GtBbPZ/+2h5YdqJgfQjHtvh92iLrIVtFhdLNS8M14kWEBo2Lfvn20ZMkSWrhwIa1evZpuu+22aj9z//33019//UW//PILrVq1ik6dOkVTp041vb9t2zaKiYmhb7/9Vnz3E088QbNnz6Z58+a5YYsY/SlD1T5icehsnuiBgI6dHZW0C72AiQNkCiGEtFdRtbIHpAztO6UvNazquLR7U5M6lLVJ6N970kT+MQwpWRjMaBd040WDL3hya+PBl93V0d04NFAfssJSPhlNH2vDAqXOaGK3JpqvKbEWtXBGOpSeUqHUdRb7HaizgEGZpBRuyxo8vfLI+PaiaR7uU3d9v50e+mUXfbr6KE2Yu4ZWJKULxbgXJnehQH/dTJc9ji5GwAMHDtCiRYtoy5Yt1KdPH/Hae++9RxdffDG98cYb1LRpVY9gdnY2ffbZZ/T999/TqFGjxGtffPGFiEps3LiRBgwYQDfffHOlz7Rq1Yo2bNhAv/32G919991u2jpGLxGLg2fyhBcDeuD2svWYceLRs3mkLmsMkA4GryzSoQbaqf4CryD0v1GrIm/oemZQ6ygh1wnv9pfrkunuUW2rLPO7kgY1uWczD6whUxt6t2gg5GYhESy9+fYWbmtZZtYc2SUbhlR2QQlFhAbYpQi0eN9p8fzS7k1Ib6DOYlPyOacUcKMfiF5SoYAptfdY7SPwW5LPCUcTOszrvZgZUZxXLu8qmuAuSzxL87edML2HSMW71/bUXV2kp9HFDAeTfaQ/SaMCjBkzhnx9fWnTpk0WP4NoRElJiVhO0qFDB2revLn4PmvAIGnYsPqbQ1FREeXk5FR6MN4LPPZQcgK4GdUGPU481EiJ2F21qLOo6M5aX3eeTUvAMJw1xmhMfLz6qJiUqUnJyBdpc9jWy5ToBqMPw0JdhF2rwm0d1FdImkaGiAk2ojTrjtiXDrUi8axQeUMkU2+pnc6MWBSVltEpRVFID6lQ6p46iDjY2sPBnI1HM8W//VvqQ2LYFsnpT27qQ1/O6Et3j2xDw9o1oluHtKS/7hnCRoW3GhanT58WKUtq/P39hQGA96x9JjAwUBgkaho3bmz1M+vXr6effvqpxhSrl19+mSIiIkwP1Ggw3s2AVlGVBtTaRiz02rFTTro2J2fZ3Z1YTrpkEbg3gILs9o3DhVrYp2uOVnrvhy3HTUo5SLFh9IGcIO84lmWXtDTqjqRMrR4Kt9XIomt706FMaVDdm+hSEchZylCIcGE4RIprVC0i2Z4gql4QtWts3P7NybW7n21UPjeglT4dZZaAI2hE+xh6aHx7+vrmfvTkxE4UFqSLpB7N4VHD4rHHHhODUnWPxMREt6zL3r17adKkSfTMM8/QuHHjql0WdRiIbMhHampVNQHGu5DpPxuO2D8Qn8kpFOFyOOuRCqVHusdHUJC/ryhkO5JuLFa0FXnz6q8YZ94AdNAfGNdOPP98XbJJGebPnSfp41VGQ+OqPuxw0Ju0NCaIUEpCTZStYEyAHZIQFUox4fpKC5F1FqsPZtjsMIAhtTzxrK4Vz6RhcSyzwCG5XdlkDylBejKwKhxl52pVXyFr5uT3MIwaj5pjDz74IE2fPr3aZVD3EBsbS2fPGgcySWlpqVCKwnuWwOvFxcV0/vz5SlELqEKZf2b//v00evRoEal48skna1zvoKAg8WDqDnIARfgYij/w+tibBgVJy3CdNJGypIYEj+6Go5kiaiNvzDUhDRHcc/WUJmIL4zo1FoWLu05k09i3V9HkHs1M+bnTBragi7taHpsY7aa4QdFp/ZFMUWdha/fslYr6F3Tx9Ub/Vg1FcSrqp+C9b9u45m3+d89pKiotF5NpWQisN9DQDUYkar9gXNg6npkjDdA2OqsdQwrT1xuO1SoCD6ECUV8RHcbyq4z2IhaNGjUSdQ/VPZDONHDgQGEgoG5Csnz5ciovL6f+/ftb/O7evXtTQEAALVu2zPRaUlISHT9+XHyfBGpQI0eOpGnTptGLL77o4i1m9ErDsEDh0ayNl0f2O9D7xBqTEHvrTNCACyBtCDKP3gQ8lHOv6SmOa2FJOf24JVUoQU3q0ZSeubSzrjyYTOWUPxgWtgAv/6oko2GBNAq9AQUraRB9t8mYwlcdUM75aNURU0ROr+c41tsZdRYyYmGLQabFsTzxdK4QJKlVfQVHKxg911hAyemiiy6imTNn0ubNm2ndunVCtemaa64xKUKdPHlSGCJ4H6D2Ab0uHnjgAVqxYoUwSmbMmCGMCihCyfQnGBVIfcJyqL3AIz29dvJ7TN2IWmw4al+ho5yk9FaK5ryhzsTWtAlpWPRXOv16GwnRYfTz7QPpoxt6C2nZKT2b0RtXdhepUox+FeBslZbGpBTefkhR6jUtZNqgBPHvL1tTa+xT88+eNNHPJjI0gG4Y0IL0TEWdRe27rZsiFrWMeOhRkEQ61rypvoKpg4YF+O6774ThgJQlyMwOGTKEPvnkE9P7UIBCRKKgwJjrDN5++22aOHGiaIw3bNgwkQIFKVnJ/PnzhRGBPhZNmjQxPfr27ev27WN0ZFjYUWeRX1RK+5Vuvnot3JYgTQQTqPTcIiG1agvyptXPS9RDrHk/L+oSKxRE3r66h1AYYfRJr3jjNYrzGymPNbFSiVZgbAgJ9CM9ApEBTIyh8vTL1gqpTXNQ0D5v+WHxHM3C6um8sFVGoOX4bC/YHzLaISfp+oxAZ9pZX2FU+dOrIc24Ht3cAaEAhZ4Uubm5omD6888/F920JQkJCcKLOmLECNNrwcHB9P7774tajPz8fGFUqOsr5syZIz5j/khJSXH79jHaBx4aRP5RM3A2t9Cmz8AIQfoAZBkh76hnggP8TM0CbUkHgwxroqKW089LIxaMd4FeDnKSuP14zdLKKw8aa/9G6LC+Qm0YT1eiFl9tSLGqiLX0wBlRYxYe5G+KcugZ2YF6fy07UCNSdaGkjAL9fHXTHM/RAm70r8DpwfUVjFcYFgzjaVAj0DG2vl1RiyX7z4h/x3TUX/61o7K7qC2RTZRYdpXRC30VI1gqH1UXjdySbEyZGtFev4YFmNqrGdUP9heFzCuSqm53QXEpvbn4oHh+06AWFBGiTxEKNbLwPCWzoMYUMEvIaAWaZeqx6ansQQHnD5rD2cJ/SlNEextIMnUL/V0NDONBhrYzDqiL9xkNhupApGJZonG5sZ28QyFIHT6vqc5ic4p311cw3snFXYydpP/dm0bFpdalSKEeVVxWTvENQ8TkUs+giPuafs3F8+cX7hcS2epx7N4fdohoBXo13DKkFXkDDcICqWmE0et+IC231oZFG6UnhN6Aswf9LDCMSwdYTd3WF+2V3db1KTPMuAc2LBjGDi7papx0wGCAx7I6dhzPooy8YgoP9jdNyPUOJGcR+j+TAxnZ6tVUNnlZd1am7vSsQXHr+YISWnvYupDHcsVpMKJdjG7VkdTMHNqK4hqECA/+tZ9uFOmeiFTA0Fh64Kyor0J3YijkeQudlKiFrBuwh0NK0bce6ysklykGwm/bT9rUbR09XmCM6b1ekHEtbFgwjB1A+adFVKiQF11WQ6qE9AKN6hDjNQW9qLMY1MZoKPy509h911r+8e6TXOTH6LMD78RuRgfCAivnOOqH/thhfG+Cl/QrgQf7h5kDRD3Y0fR8GvbaCur09H/05XpjzeHbV/UwyfF6C52UOgvZ8K0uKEKpmdIrTvyL/kQnsiqEbywhx/tLezRl1TumWrxjtsMwbgKeSTnpWLjL+sQaaUKLFcNibKfG5E1MVW5G8HJZK/T8ffsJEWIf2CqKYpV0A4bRC5f1MHpycQ1fKC6r8v73m4+Lwl0oC+Ec9xbiG4bS9zP7iwZycJ4ARChemNyFLlHGPW9C1lnYW8CN8f3wGakIpa8eFmpgRMrztzpHEdSgliu1N5O6N3Pb+jH6hA0LhrGTid2amqQmc6wU/SFNCJKVSBvSY0femjpOQxkGUQlZR2F+0/1VCa1f3ttohDCMnoD6GWon0JkZakhqSsrK6SvFi3/LkJZekQalpkVUGP1731D6867BtOOpsbT9qbG671lhjU5N6pvSmqqrpzEHqaBIC0J0KyFaf4pQ5oX74NdtcAYZrBZtY/8g7atjE/0aUox7YMOCYewEXsrWjcJE4eYSK0Xc/ymvI187PFj/Cirm6VDSe4mbkTloLgajKjTQjyZ08Y40EaZuAWNB5p//ufNklSZxp3MKRR2GjGx4owJe9/hIUeDszaCmBApXJWUGOngm1+76CqTFBvnrs3+JZELXJhQS4CcaH+5MtSyx/LviKJrUo6nXGdKM82HDgmHsBAOrVMX4a3fV8DFSJ77ZcEw8R+M0b0SmQ2GSZZ4qMn+b8SY0oUsTCtN5Ey2m7jKph9GTi8Lln7emmhSSPlubLJ7fOKCF7ieVdR2M5TJqYU86lEkRqpF+6yskaHQo71PyvqXm3z1pogYD0Rl5TTBMdbBhwTC1QBoWSIcy7+nw6ZqjwqOJ/NUpPb1zIO6b0ECkiqBb7+L9RglCKUkoa08u7+2d287UDdo1Dqc7R7QWz2f/toe+2ZBCV3y0nnafyBYKSdcPMMqzMt5RZ2GPMtRBWV+hU6lZc25QzuXfdpykBaraQfS3eOrPfeL5/w1vLWpwGKYm2LBgmFrQulE9urZfvHj+6K+7TV57SDR+tOqI8fUJHUTakLd6+qb2NEYtXv4nkVIy8kV+7ttLD4rcYxhVA1hmltE5D49vL3LQEanABGvH8fOivujNK7uLVCjGeyRn96fZHrHYfcKYMtRBaZiqd3q3aEh3jTQa0Y/9utsUkXlu4X7KyCsSylf3jG7j4bVk9ALnKTBMLZl9cUcRsUC32tf/S6K7R7WhV/5NFAWfPeIj6VIvVFFRM2NwgkiFguwidO/7JjQ0ebvuHNmaJQkZrzCgX728G2XlF9OKpHQhxPDy1K7UNDLE06vGOInOKsnZ0rLyGrto5xWV0gHFCOmT4D3yu/ePaUfbjmXRxqPn6PIP15vUoFBS8doV3Tjtj7EZjlgwTC2pHxwgJhng83XJ1Ov5JaZGQ09N7Oj1RW4o8Px+5gBRyJ6WXSiMCtgSL03pStf3904VGabugR40/5vWl5Y9OJy+nNGXjQovA954FHDDISR779TU+BQq24jKNonwnnMBBtW71/akmPAgYVDggfEcUTs0RmUYW+GIBcM4wIj2MXRd/+b0/abj4m90Jb1xYIIILdcFZFOtGz7bRCezLogb0+iO3tW3g2FQuIr0R8Y7jy16OSzad5rWH86ocRK9NSXLVGfmbcSEB9OCu4fQgdM51DQihJo1CBHF3QxjD3zGMIyDvDCpiyhswyTbW2sqqiOmfjD9e98wKioto9BAHlIYhtEXg9sYDYt1hzPp7lFtq10W6UKgd4J3Oo/Q0JSbmjKOwKlQDOMgqCWAWkZdNCrUXj82KhiG0SOD2kSbjAZLndYlqMFAnx5vjVgwjDNgw4JhGIZhmDpLq+gwiq0fLJqebj12zupyiadzRS1GeLA/tYvhDtQMYwk2LBiGYRiGqbNAaGNQG6M8NtKhrLElxWh09G7RgFXvGMYKbFgwDMMwDFOnGdzamA61/kiG1WW2KvUVfVpwGhTDWIMNC4ZhGIZh6jSDlTqLPSezKbugpMr7aAC6VYlY9PHSwm2GcQZsWDAMwzAMU6eBElKrRmFkMBBtOFo1anEkPY/O5BSRv68PdY+L9Mg6MoweYMOCYRiGYZg6z8j2MeLfbzYeq/LeZ2tTxL8j2jeikMC6qwDIMDXBhgXDMAzDMHWeGYMTREQCBdzbVOpQ6blF9Ov2E+L5bcNae3ANGUb7sGHBMAzDMEydJ65BKF3RO048f3fZYdPr32xIoeLScuoRH8n9KximBtiwYBiGYRiGIaI7R7QRDT9XHUynnannqaC4lL5WUqNuG9ZKSNMyDGMdbpXLMAzDMAxDRM2jQmlyj2Yi9Wnm11spLNCPzheUUIuoUBrfOdbTq8cwmocjFgzDMAzDMAp3jWxNgX6+orYiJbNAvPZ/w1uLSAbDMNXDEQuGYRiGYRiFVo3q0T/3DaXUrALy8/Gh8GB/UV/BMEzNsGHBMAzDMAyjok1MPfFgGMY+OBWKYRiGYRiGYRiHYcOCYRiGYRiGYRiHYcOCYRiGYRiGYRiHYcOCYRiGYRiGYRiH4eJtJ2AwGMS/OTk5nl4VhmEYhmEYhnEacn4r57vVwYaFE8jNzRX/xsfHe3pVGIZhGIZhGMYl892IiIhql/Ex2GJ+MNVSXl5Op06dovDwcPLx8XG7FQmDJjU1lerXr+/W32Y8Ax/zugUf77oHH/O6BR/vukeOzo45TAUYFU2bNiVf3+qrKDhi4QSwk+Pi4jy6Djgx9XByMs6Dj3ndgo933YOPed2Cj3fdo76OjnlNkQoJF28zDMMwDMMwDOMwbFgwDMMwDMMwDOMwbFjonKCgIHrmmWfEv0zdgI953YKPd92Dj3ndgo933SPIi485F28zDMMwDMMwDOMwHLFgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LBgGIZhGIZhGMZh2LDQOe+//z4lJCRQcHAw9e/fnzZv3uzpVWJcxOrVq+nSSy8VnS/R4f2PP/7w9CoxLuTll1+mvn37Unh4OMXExNDkyZMpKSnJ06vFuIgPP/yQunXrZmqYNXDgQPr33389vVqMG3nllVfE2D5r1ixPrwrjAubMmSOOr/rRoUMH8jbYsNAxP/30Ez3wwANCsmz79u3UvXt3Gj9+PJ09e9bTq8a4gPz8fHGMYUwy3s+qVavorrvuoo0bN9KSJUuopKSExo0bJ84DxvuIi4sTE8tt27bR1q1badSoUTRp0iTat2+fp1eNcQNbtmyhjz/+WBiXjPfSuXNnSktLMz3Wrl1L3gbLzeoYRCjg0Zw3b574u7y8nOLj4+mee+6hxx57zNOrx7gQeDp+//134cVm6gbp6ekicgGDY9iwYZ5eHcYNNGzYkF5//XW65ZZbPL0qjAvJy8ujXr160QcffEAvvPAC9ejRg9555x1PrxbjgojFH3/8QTt37iRvhiMWOqW4uFh4tsaMGWN6zdfXV/y9YcMGj64bwzDOJzs72zTZZLybsrIy+vHHH0V0CilRjHeDyOQll1xS6X7OeCeHDh0S6cytWrWi66+/no4fP07ehr+nV4CpHRkZGeLm07hx40qv4+/ExESPrRfDMM4H0UjkXQ8ePJi6dOni6dVhXMSePXuEIVFYWEj16tUTUclOnTp5erUYFwIDEqnMSIVivD/L5Msvv6T27duLNKhnn32Whg4dSnv37hW1dN4CGxYMwzA68Gji5uON+bhMBZhwIE0C0an58+fTtGnTROobGxfeSWpqKt13332ihgoCLIx3M2HCBNNz1NLA0GjRogX9/PPPXpXuyIaFTomOjiY/Pz86c+ZMpdfxd2xsrMfWi2EY53L33XfTwoULhSoYCnwZ7yUwMJDatGkjnvfu3Vt4sefOnSuKehnvA+nMEFtBfYUEmQi41lE7WVRUJO7zjHcSGRlJ7dq1o8OHD5M3wTUWOr4B4cazbNmySukS+JtzchlG/0BXA0YF0mGWL19OLVu29PQqMW4GYzoml4x3Mnr0aJH+hiiVfPTp00fk3uM5GxXeX7R/5MgRatKkCXkTHLHQMZCaRagcA1G/fv2EigSK/WbMmOHpVWNcNAipPRvJycni5oNi3ubNm3t03RjXpD99//339Oeff4r829OnT4vXIyIiKCQkxNOrxziZ2bNni1QJXMu5ubni2K9cuZL+++8/T68a4yJwXZvXTIWFhVFUVBTXUnkhDz30kOhFhfSnU6dOiVYBMB6vvfZa8ibYsNAxV199tZCgfPrpp8WkAxJ1ixYtqlLQzXgH0LYfOXJkJcMSwLhEQRjjfQ3TwIgRIyq9/sUXX9D06dM9tFaMq0BKzE033SSKOmE8IgcbRsXYsWM9vWoMwziBEydOCCMiMzOTGjVqREOGDBF9ivDcm+A+FgzDMAzDMAzDOAzXWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzAMwzAM4zBsWDAMwzCaZOXKleTj4yMekydPtukzc+bMMX3mnXfecfk6MgzDMBWwYcEwDMNoghEjRtCsWbOqvJ6UlERffvmlTd/x0EMPUVpaGsXFxblgDRmGYZjq8Pf0CjAMwzBMdcTExFBkZKRNy9arV088/Pz8XL5eDMMwTGU4YsEwDMN4nOnTp9OqVato7ty5plSmlJQUi8vOnz+funbtSiEhIRQVFUVjxoyh/Px8t68zwzAMUxk2LBiGYRiPA4Ni4MCBNHPmTJHKhEd8fHyV5fD6tddeSzfffDMdOHBA1GFMnTqVDAaDR9abYRiGqYBToRiGYRiPExERQYGBgRQaGkqxsbHitcTERIuGRWlpqTAmWrRoIV5D9IJhGIbxPByxYBiGYXRD9+7dafTo0cKYuPLKK+nTTz+lrKwsT68WwzAMw4YFwzAMoydQlL1kyRL6999/qVOnTvTee+9R+/btKTk52dOrxjAMU+dhw4JhGIbRBEiFKisrq3E5FHYPHjyYnn32WdqxY4f43O+//+6WdWQYhmGswzUWDMMwjCZISEigTZs2CTUoSMaWl5dXWQbvL1u2jMaNGydkaPF3eno6dezY0SPrzDAMw1TAEQuGYRhGE6C5HVKdkOLUqFEjOn78eJVl6tevT6tXr6aLL76Y2rVrR08++SS9+eabNGHCBI+sM8MwDFMBRywYhmEYTQBDYcOGDaa/ISVrDiITixYtcvOaMQzDMLbAEQuGYRhG08TFxYneFbbw0ksviTQqS9EOhmEYxrX4GLirEMMwDKNBLly4QCdPnhTPYSzI/hbVce7cOfEASKdCfwyGYRjGPbBhwTAMwzAMwzCMw3AqFMMwDMMwDMMwDsOGBcMwDMMwDMMwDsOGBcMwDMMwDMMwDsOGBcMwDMMwDMMwDsOGBcMwDMMwDMMwDsOGBcMwDMMwDMMwDsOGBcMwDMMwDMMwDsOGBcMwDMMwDMMw5Cj/DyJzCjG6g8MQAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 800x260 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE(1 epoch): 0.0165 m  (훈련 전이라 큼 - 당연)\n"
     ]
    }
   ],
   "source": [
    "# [5b] 학습 전 모델의 예측 한 샘플 눈으로 (epoch 1이라 아직 안 맞는 게 정상)\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    pred = model(X[ntr:ntr+1]).squeeze().numpy()*float(u_sd)+float(u_mu)\n",
    "true = UU[ntr]\n",
    "plt.figure(figsize=(8,2.6))\n",
    "plt.plot(np.arange(L)*dt, true, label=\"true u(t)\")\n",
    "plt.plot(np.arange(L)*dt, pred, \"--\", label=\"FNO pred (epoch 1, untrained)\")\n",
    "plt.legend(); plt.xlabel(\"t[s]\"); plt.ylabel(\"u [m]\"); plt.title(\"smoke-run prediction (epoch 1, not yet fit - expected)\")\n",
    "plt.tight_layout(); plt.show()\n",
    "print(\"RMSE(1 epoch):\", f\"{np.sqrt(np.mean((pred-true)**2)):.4f} m  (훈련 전이라 큼 - 당연)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1badfa0e",
   "metadata": {},
   "source": [
    "### 요약\n",
    "- 공식 `SpectralConv1d`/`FNO1d` 의 핵심(= **저주파 모드만 학습된 복소수로 곱하고 `irfft`**, lifting→Fourier blocks→projection, `F.gelu`)을 출처·라인과 함께 전사해 돌렸다.\n",
    "- nb01의 선형 $H(\\omega)$ 가 못 다루는 **Bouc–Wen 비선형** 응답에 적용 — 입력마다 응답이 다른 연산자 $a_g\\mapsto u$ 를 데이터로 학습하는 틀을 세웠다.\n",
    "- batch 10 · epoch 1 smoke run: shape 맞물림·유한 loss·1회 갱신 확인(에러 셀 0). 성능은 아직 아님 — **W2에서 제대로 훈련 + 해상도 불변성** 실험으로 간다.\n",
    "- 비교 포인트(다음): 모드 수 `modes` 와 `width` 를 바꾸면 무엇이 달라지나, 격자 해상도를 바꿔도 같은 가중치로 되나(discretization invariance)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
