diff --git a/Exercise4/exercise4.ipynb b/Exercise4/exercise4.ipynb index 5e218ec..f0e9f0c 100755 --- a/Exercise4/exercise4.ipynb +++ b/Exercise4/exercise4.ipynb @@ -18,7 +18,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 8, "metadata": {}, "outputs": [], "source": [ @@ -87,7 +87,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 9, "metadata": {}, "outputs": [], "source": [ @@ -128,9 +128,22 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 10, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjoAAAItCAYAAAAwm9DjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydZbwVZdv2D55H0ttAbEwsxO7E7u7uQrHztrHjthMTW1GwMLALOxC7EwNEEIv0ed73w/se5zqGPcOOtWbN2ovj/+n8nXuvteaa65prZs5s83/+z/+BMcYYY0w98l9FH4AxxhhjTF74QccYY4wxdYsfdIwxxhhTt/hBxxhjjDF1ix90jDHGGFO3TDe1P06aNKkuU7LatWvXBvD4WiseX+tmWhnf5MmT63J8bdu2nSbGV+/rs97Hp9iiY4wxxpi6ZaoWnUrSpk2bBrLW8JnW6/n813+Vnjl5LlrTOWlsfpXWNC5jjDH5kvf9zxYdY4wxxtQtVbPoTJgwIeS///4bANCpU6fQTT/99CHzSe5///d/q3R01YUWj//5n/8JXZ8+fUL+7rvvAAD9+vULnT7x1gr//d//HbLOL8fVsWPHBjoAaNu2LYDk/NrKUzuodS6NWpkrvSYaO+a0v2dZlFvLvqPj12uRY/nnn3+qfkymfDivumZ1/6wXdHzjxo0LmWu5Xbt2oSt3z6m9u6cxxhhjTIXwg44xxhhj6pZcXVdqWj366KNDvueeewAA6623Xui22WabkDfffHMAwKyzzhq61mJObgrTTff/TvtDDz0UuosvvjjkQYMGJf4PqK3x0+Q4atSo0J199tkhf/LJJwCAbbfdNnQjRowIeauttgIALLfccrkeZ7k05g7JolZcO4pei2luUF1flCdPnpz6GTUpp30+z/Gri2bMmDEh//bbbyGnjW/ixIkhc17pQgWAWWaZJeSZZpoJQG1dcwrHN3bs2NB98MEHIc8xxxwAgIUWWih0LV3L9UKWm7NWXEJ6TKNHjwYA/PLLL6FbbLHFqn5MeZGWrHLKKaeEvOuuuwIAVltttdDpXtQSbNExxhhjTN3iBx1jjDHG1C1tpmZmLrdyopoL1Qw1dOjQBn9XEyJNV1dddVXoZpxxxpDLNSkXURlSTe6vvfYaAGDLLbcM3WGHHRbyaaedVtZv5TE+Na1yzRx11FGhu+WWWxp8RjPp/vjjj5DpJrjhhhtCR3cl0Lg5OY/x6fyQSZMmhZx2nehnVOa6znLnNObayWN8w4YNC/mxxx5rcBya9UD5m2++Cd3MM88c8qKLLgog6XpceeWVQ1aXcxotGR/P5QsvvBC6888/P+Rff/015LS51PGlubZ0/dEN2759+9A1xx2XR2VkHdPIkSMBAMcff3zoBg8eHHLXrl0BAM8991zoOnfuHHK5+2drqYysrn917eletOqqqwJIzm8RlZH1WK+88koAwLPPPhu6Bx98MOTWeP9TOFaGOADAfvvtFzJDWxZYYIHQNSeD0JWRjTHGGDNNUbU6OhdeeGHIN954I4DSmwmQfHu89957AQCzzTZb6udbS2CdvoV98cUXIe+www4AgA033DB0J598csgcnz65FzHmrKC9vn37AgDuvPPO0HXo0CHk8847D0DJMgck3zhPOukkAMD9998fOrVuVStAUOfn999/D/nll18GAJx77rmh+/nnn0PmeVl77bVDxzdDAFh99dUBAN26dQudWgfynEu1Vvz0008hH3HEESEPHz4cALDggguGToN5eawawPvGG2+EzCB6Hcc666wTcv/+/UNmrayWvIXqWLg/6Jr6888/Q9bAYq4ffQvUN2Yek/6dew4A9O7dG0AymLeImjR6fvX8cf/U+WECA1A6R3p+NNg6zeJYT/C61vHrtazngp6GohMI9PdfffVVAMlrsrXc87LQ42cdvWuvvTZ0u+++e8i05FTyPmCLjjHGGGPqFj/oGGOMMaZuydV1pabRnj17NpBpwgKADz/8MOR99tkHAPDkk0+G7owzzgiZQa5FmxuzoJlO2yKo62DuuecGAFx99dWhU9M7z1vR5kr9fTUD0zWlpvNDDz005H333RdAcky77LJLyHSDMCgbSNaM6NKlS8iVMq+nnUt1lx533HEh082m9WL08zT9DxgwIHQMoANKrh+6KIFknSHOP1B5N51eExpAfP3114fMeZlzzjlDp/PL49fvuummm0K+/PLLASSDOhdZZJEG3z/ldzQX/SzdDTvttFPqMev/zjXXXACAFVZYIXQajH3dddcBSNbm2HrrrUOeZ555ABRfY0XXnB7/kCFDAAC33npr6HQsvGay6nCxxQzHCRTTYibtmgKA119/HUByT9Bg8TT087yv6DWtoQMMxi8aHb/OH13Ou+22W9WPKS90T6Brm+sYAE444YSQ3dTTGGOMMaYZ+EHHGGOMMXVL1bKulLQS7J9//nnINOPp34t24zQHHqvWAVI3zTPPPAMgWW+kFjsNqzlYa8rQZaVZS9rCgy4fHZN+F+dVM5k0w0Cz7cpxXemaoRn/q6++Ct2BBx4YstZnoctnhhlmCJ1mldGkThcrAPzrX/8Kmabxu+66K3RaU+iiiy4KmeeiUmZa/R7tHr/UUks1+F89t5qJwvnVTKT77rsvZGaobbLJJqHTFi9pbtiWkOaGu+yyyxr9bv5+1v7BtcxWCUDStUrXXS1dk3Q3AaV2AHTRAUk37Pzzzw8g6fpR1yXXAv8PKCYMQPcE3f95XWmdrrQ6Xlmun3//+98AknWEtN2OznuR7smsDMl3330XQDKrtTWi9wdmegKlMBTdM+add96Q87jubNExxhhjTN3iBx1jjDHG1C25uq7UtKjmKLpuWHgOKBVpA0rmRM3kaWk59mqhZrq3334bAHDJJZeE7rbbbguZ5fLL7ciaB2pO1awaLbjF4z7mmGNCpy0AGjMHp5meK+WazPqe9957D0Cyo7rOmWZoLL300gCA7t27h+7xxx8Pmdl0p59+eug0Q2v//fcHAKy33nqh00yPM888M2S6ZPJY0/qdev3xHOm50qwUmpbVXTXffPOFzE7DOv/qmsujEB3Hou4oRedyxIgRAJItRpgpBpTmaquttgrdKqusEnLR2VZpqGvx+eefB5B0B6ib5tNPPwWQ3D81W22LLbbI7TgbQzPB2AoISB4fi8dpWwCF61bnSdth0E2l11yPHj1CrpX51evvqaeeCpkuSQ1taE3FHdP2YL2/MzRh5513Dl3e47NFxxhjjDF1S9UsOtrugcFmWmcmrRHfsssuG3JaifeiyQqGY2lrbUSqY2EwpFpPirZSpT2FawAfa48AwMYbbwwAOPHEE0PX2Pzo93Ousyw65ZwLXUfvv/9+yKzDoZZBrUOib/T8jtGjR4du9tlnD3nPPfcEkB10u/DCCwMoWT6AUqO+KY+xWuj5ZbDxww8/HDq1MrE+DYM6gZKVCijVX8lqUZAnujb0+tEGiBwLrXhA+lp79NFHQ8dGpQDQq1evBt9fxPWpv6n7B2sWsZUKALz55pshM9j2iiuuCJ02YOW48h5T2jnXpAzW2wKSTVdpccqqY8Xj1hYC2vSSNa30nNVSYDnRe4a2wznooIMAJK2ktWj9bwxttKsNeLmvqJXSFh1jjDHGmBbiBx1jjDHG1C1Vq6OjdVjGjBkDIFmiXs3EdH1o2fw111wzZJbQLzpAS82pGmx1++23A0h2h9bAVZokzznnnNDpuSjCTM6x/PXXX6FTc/BMM80UMseidVrS3FU6p/q93377LYBkbRcNdm3JvKYFKGow4tixYwEATz/9dOjUXaW/ye9QF5MGzrGmTtZx8nPagoAds4HqldtPc1cBpfokd999d+h0zdHlpmXZNYiU56eI6y/LXaxuHAa5qptSzeS8/n744YfQnXrqqSHTNbThhhuGrmjXle4P2223HQBg7733Dt32228f8qWXXgogWZtEr4s8x5I1Pw888AAA4JBDDgmdun61tg1bp2htIw1M/vLLLwEk3a0aYE3Xsf5+LdVh4/X//fffh07bHbHFTtHhDM0h7fyq61sTOzi/un/kPVZbdIwxxhhTt/hBxxhjjDF1S66uqyzTK11SK664YujUNcJy4Go6107RWhq8SDSSX4+VZlSNNGfZdqDUekBdczQ3A8WWY9ey8S+++GLIWidnpZVWApB9nDRjqotGswoGDhwIINm9XTMsWpIhwd9UFw3riQBAz549AQDLL7986BoznWoLCKWpLht1nYwaNSpkdeNqZkVzSTsOPedpbQ/0WDp37hw6tnUASvOzwQYbhE6v1SJN6llZV9opni5jdU1y/oHSXqOtOOhuBkrZWjr+IshyPX700UcAkq5V1n4CSq79IjKN1MWpGVYHHHAAgKQLuHfv3iGra5FhAIMHDw7dZpttFjJdjksssUTodH/i+tY6NLXkuuKxaB2hTTfdNOQll1wSQG1miilZWbOvvvoqgGRbHb2+6LKs5vhs0THGGGNM3VI1i06XLl1CZv0VfSNTmUGqDMoCknUyWNNDGylWKzBS36LYfA1IBtaxCnJWsDWb6elTrlp0ikSby+lbvr4xk8be7D/++OOQWW0XKL2d6dtdubWRGqu2rI1Cm/udTSGtzoc2Lf37779DTqtS3NTv189q5e21114bALDaaquFTq8JtRxx3amVp3///iEfccQRAIDTTjstdFpzp9KNSJtCWjVnHZ9aeXmO1KKm1y2vRa0NpA1MaVHQ79TPV2vcOtbx48eH/NZbbwFIBvAOGjQoZNYpY70joHq1x3ROtOnoTTfdBCDZ/FfXpJ7ThRZaCEBpTQNJ6xqDlNUip9bJatUJailcS3rN6Vzy+qr12jm6PrUpM68bDZDXfamIOni26BhjjDGmbvGDjjHGGGPqlqq1gPj6669DZpCnltVXMyNrrmgAngaGVqsOSRr621o7R4+VwY5qotN2F6wDoQHKRZBW80JN+OpuXHfddRt8Lq2tAwB89913AJKN9lh7BgCuueYaAMkA9DzMmTpXavovl7Txq8xgSa0ZoqZ1Dbxuqnmd36/jeOSRR0KmOzDLFabngq1J9Lu4JvV/9ZiLrpnDBrMMxAWANdZYI2QNgqWc1dSUf9cAcR2furxqBXUzsp2Ojp/uLKA0Lq2jUy30Ol5ggQVC7tatG4DkeVbXjK5PtoPQFjPqBqPLVoONq1UnqKXoWua9QFvUaDuLouvDNRWdM3XD8fi1jk5anbxqYouOMcYYY+oWP+gYY4wxpm6pmuvqyCOPDJmuKc260aylZ555BgDw448/ho61IYCSabloE6VmDajrisel7hw1LTODhfVKagmdM3U3qZuR/6OmZ61Zw9Lsr7/+euiYiQaUzNCVNNGmHdM333wT8meffQYgmUmmrrnG3KG61ijTnQIk57Jfv34AgLXWWit0ffr0CVmzBctZw+wyDpRqVmg9laxMIbqptI6TdvJm1qNmXak7p4isI9Z3Yud4ILl/aAsAdurWY1bX1ldffQUgmbWm1y8/X0SmVRZ6Lc4555wAkq5z6oBSTZqij1l/P81dofOrf2cG6pAhQ0L3+OOPh8zx1XqdGUXHyq7eur6WWWaZkGvZdaXj0EzShx56KORevXoBSO6vRY/JFh1jjDHG1C1Va+qpb0ysnDhs2LDQ6dM//1cbeWpNiyLrJKjFQN8u1aLBKs7aNFDf6Pk5rdZa9NsX0TcrbcT5xBNPhMzzf+6554buiy++CJnNEPUz2uCSY63kmPld+ha/3nrrhUzr0r777tvgOAGgR48eIbOmBYN2AWD48OEhs2K0vsWMGDEi5I033hhAMsBQj6slgZP8P32z1yrPV155JYCk5UJrB3344Ychs6aJvpEdfPDBIbOpp46/iABCPTesCbP++uuHTuuo3HjjjSGzDotaedQifNdddwEA3n777dBpsCstcWm1kaqJ/qauH1apfu6550LHarpAqX5XrewpWahFQ603rLml86uWyiLWYrmoxZgWHa1zpE1/a3HeePxqRdY9QwPD99hjDwDpVvCisEXHGGOMMXWLH3SMMcYYU7dUzXV1zjnnhMzS6xosqtDlcNVVV4Wua9euIRcd2ES0Hoo28GRrCK19ocGqdFmpObqIMaU1SFxkkUVCp8G2Wi6fpmM9Zp1flqDXYLS8Awc5FnXdaAA86wBpgPSpp54aMmt3AKW1pp/XEu0M5tXP012ish5LuePn+HTNqemYril18er8duzYMWS6dNhcFkg2wKQZvWgXQdrxX3DBBQ10AHD33XeHzNYjWqdE1ypdJqztAgDXX399yKuvvnqDzxRBluuKx6doTR3WHCv6+NNQF47WMdL9ka4PDTYv2vVRLml1ttTdXWRtuKbAa+b7778P3X333RfyQQcdFHIRLWIao7bPrjHGGGNMGfhBxxhjjDF1S5upmZcmTZpUMduTmu4Yua2ZWApN51qnppJm2Hbt2rUBKjs+hS6LrHL8dGPomJraxboptGR8/H11t6nramqfAUq1LYCSmTMv10dTx6fmYM6JZs1pJpW6OehyW3zxxUOn7Sp4jrIyJTjulppupza+rEwgZsipaVlR1wdbA6jrR78rb5dHOdefzqnuH6yzAwBvvvkmAODWW28N3ZgxY0LmvJ500kmhW3jhhUMud91yfJMnT67Y/qLj/umnnwAkj1/l7t27A8hvHtu2bdvi8en61b1Gr0u6PlrSKqUScHyVvD+oG5v1j7hOAeDwww8PuZL3gjTKuT/oPI0dOzZkHR/3yrzHkQXHp9iiY4wxxpi6xQ86xhhjjKlbqua6UpoaYZ6X6TVv11XRlDM+NTc2JxNA5ypvM3O1xtfY+strnC0ZX3PmKo+Cjc2hUtdfVvf4lpCHa7ySriuFY81yY+Y9r+W4rpSs+St6febhulLS5q+aGXJ57J/VdH03hl1XxhhjjJmmqFodHaXoJz6TTWON+Fo79Tq+afGaqqUS89WkaItHpfD8tb4xt9b90xYdY4wxxtQtftAxxhhjTN0y1WBkY4wxxpjWjC06xhhjjKlbphqMPHHixLo097Rv336aSC8fP358XY6vY8eObQBgwoQJdTm+Dh065JqeXDR5p+8WDa+/et8/63189b4+631/UWzRMcYYY0zdUkh6ufl/1GrBrKbSnIJt02oqqTFFkXVN+vorlsYKPqbNWz3un+UWbG3Wb1Xsm4wxxhhjaoyatuhklZhu7U+0fGLXgkvjx48PmV2l2QUcqK0x8/jHjRsXur///nuqn5l55plD7tChA4BSF3djTOXR7u7adXr66acv4nAAJPcx3d+1+zX/R/eHxiwetY4e84QJEwAAf/75Z+hmmGGGBn9Xi0bnzp1D5n2hNRXsU3j8f/zxR+g4Zv273jPKxRYdY4wxxtQtftAxxhhjTN1SlusqK5iqUowaNSrkNNeHmvZq3Yyn54rHeskll4Tu7bffDrlv374AgNlmm63BZ4pCj/+1114DAJxzzjmhe/3116f6mV69eoW8yy67AAC6d+8eOjVdm+qjbgTKaWsWmDb7arUmOH+XXnpp6B555JGQH330UQBAly5dQletOdU1pa7vV199NWS67ldZZZXQtW3bNuTWuP503HRZHXPMMaHTMfFcqOvxkEMOCfnggw8GAMw+++ypn68VdE/R/YPjO+KII0L30Ucfhcz5v+eee0K38cYbp35Xk4+l2Z8wxhhjjGkl+EHHGGOMMXVLs/0FajZS0xoj+ZvjwsrKo+d3HHXUUaHTrIG55poLALDDDjuEbtVVVw05LYK/aNR0+csvvwAArr/++tAttNBCITdWX6Ba6DE///zzIdN0Onr06NCpaZlrRM2pN9xwQ8j3338/gKS5epZZZgm5VuasXuH60mv5gw8+CJnz8+2334aOcw4AK664YuJ7apXWmLXZWG2tKeWpoa6N9957L+SRI0cCSLrGq+X60EzSd999N+Rtt902ZGZb7b///qE7/vjjQ55//vkB1PY8TomOmxlGY8aMCd0rr7wScqdOnQAk73n9+vULmdfqNddcE7pacmPxuvvrr79Cd/bZZ4d8xx13AADGjh0bOs0EZAbyAw88ELpNN900ZLuujDHGGGOEZlt09C1J3+Kb83TN/2VQKwDMOuusIXfr1g0AsNhii4VOg3XbtWsHADjssMNCt/vuu4esQV58ki766V+tTHyTUYvIfPPNF3KRT+T65vHpp5+GfPjhh4fM4+Y8AKUAMgCYY445ACStVJ999lnI7du3B1Bdi0BaNdLm1ONorFqpwvkres0pOq9ff/01gGQw4Icffhjy6quvDgD44YcfQqfBgLQOLLjggqErOlhe4VzpG6Vef1x/aZ9pCrpuOdeVuma1dsywYcNC/v3330Nef/31ATS+vrp27Zqq/+677wAAPXr0aPFxthQ9Zt3zrrrqqpC/+uorAMCDDz4YukGDBoV87733AiitU6B4K0Yaes198803IfP+NGTIkNDNPffcIdN6o+fk8ccfD/mpp54CkLSSn3baaZU67Bah18+PP/4IADj00END9+STTzb4zK677hrywgsvHPJll10GIPl8US626BhjjDGmbvGDjjHGGGPqlma7rtRE1RzTkpp7GYR0wgknhE5dH/379wcAnHrqqaFTky7N5AyqA4Azzjgj5H333TdkBtwV4UbQMWuLhGeffRZAsu3DEkssETJrBlXTHMt5/fXXX0OndTgYQA2U5l3n/+abbw6ZLkcta/7bb781+M1//etfIVdqftRcrPD71QVAE35T0PNCl52OT+s8MVheXSQ6l9Vai+quUdfvjjvuCCB5zBr4N8888wBIlmXXYEm6mWvJXaBrcfjw4QCA4447LnTLL798yNTrnqLruzE3JV0rALDkkksCSJbobwlctxoAvvnmm4e84YYbhrzBBhs0OKY0dH50rJ9//jkAYJNNNmn5AbcQdXHOO++8Ie+3334N/ldDE/bYY4+QmaSiCRIM4AWKdRnrnq97xj777BPyO++8A6BUDw4AzjrrrJDpktPzo2EAdN1pgPK6667b4PNAvudCx6r7Kl3iL7zwQug02YSJDccee2zo9H9ZZ6iSx26LjjHGGGPqFj/oGGOMMaZuKbTuvroZtIUAs0JWW2210GnO/bXXXgsgWdZ8q622CjkPl0hLUBO4ug5uueUWAMAiiywSun//+98h0zVSRHdvmkUB4L777gtZzaxEzdCaobPssssCAGacccbQqZskrc5OudCMqqXEBw4cGDKzinSd6dpQNw+zxtR0rHUeWLpe50ddczwXp59+eug0KzDP1in63W+++WbIWqdkm222AZBs4aGuF7qstE7W1ltvHTLdzEW7rnTO9PyfeeaZAICHH344dFpHhhmAumZ/+umnqf6WzpO6l1gfRM3wLYHzpteUursr2SKlyPpHWWtfx83jowsVAHbaaaeQ6RoZMWJE6NS1U8Sez3FNmjQpdOedd17Iuv9zLpldBAA777xzyDx+zUrTa3WBBRYAAJx00kmhe+ihh0JeYYUVQqb7vFLnROdP94cTTzwx5MGDByd+G0i65hhaou5mvRa5r1ZyTm3RMcYYY0zd4gcdY4wxxtQtVXNdqemJ2Sq77bZb6NS0zBL0mukxYMCAkD/++GMAwJFHHhk6LWinGVxFmNfTzNDMJANK2Vaa9UBzJFCMy4q/+cknn4SuMReLjk+LzzGr7JRTTgndTDPNFLKWnq8UPFa6ZYBkVhVdC+piUNOwuqZock0rLJeFlmt/+umnAQC9e/cOnRZnW3nllVM/Vw4092v2g14fPXv2DJnZdOqO1OuEBck0U+S6664LmWb2Iq8tIGnuPvroo0Pm8euc6rrgd7AUPZAsfpn2WyqrS5NFE/X6aE7xwSnRNavnV4v/pX2/6hgSkJWByLGqC0v3nHKOv6Xob3INq+tOi5fy/qF/L5q0rDnd8/WexLWq7jiFaylrTTFDTbt7azuhgw46KGSGR1TKdaVrRtvGqJuYY1V3lrru01rQ6D7I9VnJgpC26BhjjDGmbqmaRUefBClrbQgNVrryyisBJOuUbLTRRiGztLQGcKWVZa8maW9/GsyrQb4MdtV6P0UHdvKY9S1Y33LT3njTPg8AL730EgDg/fffD12XLl1CZhCe1g6q1BtH3759Q1YrH8+1/qYGDio8lubMiQaLbrbZZgCA7t27h+6tt94KWRvQVho9Zh2fHh/fpGgZBUpzBpRqUulbvrZjKfL60mO68MILQ3700UdD5r6hY1aLzuKLLw4AWGuttUKnJfqn/B6gVC8HSDZQ5LrW42pqfTHds3itaCNjHicAHHLIISHzt3QetEEkg61ffvnl0Kn17q677gKQDFZX6/J2220Xcp7WHR3/0KFDQ+ZY1WLzxRdfhNynTx8AwFJLLRW6ovdP7hlq5dAAed0LDjzwQABJi3FzWqhw/tVil2eCg/6W1ptSiw1r3wAlT43WQdL1x+PTFi30AgCN32tagi06xhhjjKlb/KBjjDHGmLolV9eVmtO+//77kB977DEAye60rE0ClEy/GkypdWbSzKlFmy71mDiWW2+9NXRqxmRNBDWHF93pmmZkDdbdc889Q04752ltPYDSvKo7RE3TvXr1AlAKGgXKL+HO+Vd3qJr7WQdGy8bThaifrwQMrBs1alTotLZTHvD4tXaRdg9WN56aiYmec5auX3PNNUPXrVu3Br9VBGqu13ohTGAASsGs+r/q5qabXNsKKDwXuuZ1reu5am6nev0ebSXBY1EXjbrrWTsMKLmktON8mrlf15yeC7bIWG655UKnAfJZY600aQkqQOm4tU6OugvZ+mCNNdYInbohq7WX6vqgm03bjqjrVPclujub467S32KtMLoogaRrqDlJFE2Fx6r3NHWNMigfKN3fspKCeF60hQVr7wClmniVvD/aomOMMcaYusUPOsYYY4ypW3J1XWkdHLorgJJ5Vts2aG0K1pLQPHo1vTbH5JcnWXUozj//fADAq6++GjrtLsty/GqOLNr1RtiZGsiud5NmRtSx0KT8888/h46ZdABw0003AUi2a1DTeUvMlGm1J/baa6+QhwwZAqDUXgQA5pxzzmb/ThY6/pEjRwIAvvzyy9Ats8wyIee5fnVNqutRXU88F9rRW+snnQXnPgAAACAASURBVHDCCQCSdYDUDVLE9cfzq+PT+dUMpddeew1AMkNE64w8+eSTAIAddtghdGlZIWkuqnKPX8+d1vFhnRg9DnUxqptg4403BpB0bWmdJrZO0HYnmiHDOiuqK2Iv0vOr7XCYrTp69OjQaVYPXdLHHHNM6J544omQuYfl7cLSc8b50zpWK664Ysha26bc3+K9UM+Jhnlohms55yDNzXrDDTeETrML999//5DZqVzXkX4Xw1i0bYWOj1lbmgHsOjrGGGOMMRlUzKLDJzZ9SmMFRyAZJMnKrFoZWd8eafHRt7Sig3XT0GNS6wVr5miA4KmnnhoyA+tqxTKVRVZlTspZc0LrllrptCnkjTfeCCBZ2ZONQIHkm0Jz513Pqa45vj2rZbCSb656fhjsufTSS4du4YUXzuV3pyRrztZee+2QNXCTqMVtnXXWAQBsuummoSvC4qhvgQy8VYuHBkBqA+BVVlkFQDIBglZEoJjrjvOiY9JgWu5/2vxWLY6cE6BU0VuvEw18pawJAgqtDzqnRVRDVnROaJFQ67KOj1V2taml1uxi4G/eFeZ1L2HiBaveA8mmpLoXpq2FtOtLv1//l5ZinTOtc6WB3eWsdf1+elk0wUJ/U2vCpVXk1uO44oorACStlLo/0tJcyTVpi44xxhhj6hY/6BhjjDGmbinLdaWmpQkTJgAALr/88tAttNBCIdNdAQArrbQSgGQdiXfeeSdk1u/QYKRaQU2IWudBa5b8+OOPAJIl6tV1UCuBx2loc9UXXnghZDWd0/WWVcKcpkst8f3MM8+ETJMyg9aA8txVTYHHqt9ddm0GWQt//PFHyGeeeSaApLtS13IRTVvT5kfrsLz99tshX3311QAqGwzYEtQEfvHFFwMAHnjggdBpAKY29aTLkkHXADBx4sSQmeSg67darnH9nfXXXz/kDTbYAEDTzPVpwdI6v9RnNeqkS0/nNKsBaLVIC4bW8em1xiQJrdOS9/6RdpzaYoX3L3VdabC4Hiu/g/WqAKBz584hcy70/vL555+HzCByDabfcccdQ67Utapj5b2A93kguykyj19r4912220hDxo0qMFn2NYDKLn5KulitkXHGGOMMXWLH3SMMcYYU7dUzHXFrCOtx7H33nuHrFk1NPOfe+65oWOmClCKytbo8SLM/Wmoife5554LWdsZrLfeegCAffbZJ/U7ajGDjOZOLcutNRO0DgtdV5ohoq4BZpvpnGpNE/6vmnP1vFZqrivpplK47vWY2REaAL799lsASddE0Rl2eq3y/J999tmh05ofrG9V9DrVc8bS+mruv+SSS0JWN1XXrl0BJDvGq2uArvE81lxLSXNH5YW6gWoZPU51E7E+m9Z20gygamU16vHRdaY63f+0tcftt98OoFTvCUhmaPE39F6q+yf3lfPOOy90mpVWqfHrWJkJpZnEzMQCSnseUGpNoRnYAwcObPC9munJ2nJT/m6laB0r3hhjjDGmBZRl0dEnLwYWzTTTTKHTwFatAsrAJtabAZLWj549ewKoraBdvv198803oWOAJJAcN4NQ9Y2jlsZC9C2fjShZYwNIPr2z2i+QfLtIg2NVK4+eiy222AIAsM0224SuaItHc0gLJmRtKKBkadDzV/T41HrBit0aIK5VeNlgtehj1mtmqaWWAgD0798/dGp91CQIVgTW49f9hc0Cix5f3qiVKitwuUiyaj7ROqG6AQMGhMzEFjaPBPJr0Ds19Jpigo3e0/r27RuyNmUdM2YMgOT8aM0gNlvVavFa5ZzXgtaUymNO9TwyGUU9M++++27I9GIApYBsTUbRas1HHXUUgOSYNJg8j/mzRccYY4wxdYsfdIwxxhhTt1TMdcXaBhpgdPrpp4d82GGHhcz6KRqsfNppp4XMIORadPdoozmtA6Tl3FdYYQUA+QXDVgo9JppBr7nmmtDdeeedIWv9EtYJyiphTjPyJptsEjptOkiTpZpea3GuFTWjc9xaJ4fuEKDUOkHHVPT8q2mbQfTaYqB79+4h1+Jc8PjVBXrEEUeEzDo0QKkmkNY0UdM6zeS14sLJC22how16eS0WHZSsv6/1WZ5++mkAyToyel9hYLk2Si4CXT8777wzgGTbDQ32V9cUWx+pa0pbmLDBaVbtLe4lea9f3bOYeKL3Bw2GHjx4cMjc/7UFlLZ7YmucPAKos7BFxxhjjDF1ix90jDHGGFO3tJmaSX3ixInNtrerif/vv//W7wqZ0erM7gCSZqy8ad++fRsAmDRpUpPHx3GpuVDL/mvUeNGtK9q1a9cGAMaPH1/W/Kk5UcfKDK0sOJdaB0kzFBrrft4YHTt2bAMAEyZMyNUfpOdCj/XBBx8EkDTj3nfffSHTTdBS03KHDh3aAMDkyZPLGp+6Bv7888+Q6WZjvSogWccib5N427Ztm339paHzo+srbV2pLm8zOa+/luyfefH777+HzNYX6jpuDtw/yx1fWgshoOSmGjZsWOi0+zezdjTTqpKdrltyf+D602tHszL1/sb7g55/PX5+R17ubq7PluwvWXOm93reC/X+ntYuKK/xcX9RbNExxhhjTN3iBx1jjDHG1C0Vd10pjUX1F5WV1BLTJEnLvgGqaxpvjHJcV0qWObgxM3FjroNyqZbrSt0hWiSRGWh77LFH6DQbqFzXT6VcV2ltHwDgnnvuAQBsvvnmoWNbDyD/9Vsp11UWaeuzmvtLLbqu0vaqlp6TSrmuskibvzRdXuu0UveHrH2ymu0+0ijHdaU0Ntai7u92XRljjDFmmiJXi06tUs4Te2ugUhadWqWIYGR9e+Tbcdbfy6VSFp0s0urIVNMKmbdFp2hq0aJTSfK26BTNtHJ/yGt/KRpbdIwxxhgzTeEHHWOMMcbULVN1XRljjDHGtGamWqWv3n2w9e6jrHcfc94xOkXBGJ1//vmnLsc33XTTTRPXX72Pr96vv3HjxtXl+Dp16jRN3B8Uu66MMcYYU7dUr+/CNE7RtT3yprGaCkXXFiqXxmp7tPbxpZFVB6sex1pPcF1mtTBpjftO2lps7WOaVilifdqiY4wxxpi6xRadKvHXX3810GnTs0o2pasW+pY1fvz4kNlAUsenlYNr3SLAudDxaQM7NjVV3Ywzzhgy69S09rdMbQSqVaI7duwIoHWNT68vNljU4//nn3+qfkyVJM26eP3114fu+++/D/nMM88EALRr1y50tTiXOqazzjor5IcffhgAcNxxx4Vut912CznvprRNJW3NNQXORZbFI++mmJUiq4sA909tBKqNsLnXVHJ8tugYY4wxpm7xg44xxhhj6pZCXVdZpi3Kaea6KfW1CMelprkNNtgg5J49ewIALrjgggafaQ3wWLXRZe/evUOmaXndddcNXf/+/UOeZZZZANSWCyvNTfX++++H7qabbgr5s88+AwAMGzYsdCeffHLIxx57bIPvrHXUtM5x77zzzqHbZJNNQr7wwgsBJN1ZtXhN6jU1adKkkF944QUAQJcuXUK3xBJLpH6utaBrbeDAgQCS+8u9994bcprrrhZ57733Qr722mtDZhjAZZddFroNN9wwZM5rEePTtaOu7SFDhoTM49f/pTsYKB3/jz/+GLr55psv5KWWWgpA8vormrRkFF2Tn376ach0qb744ouhGzx4cMizzTYbALuujDHGGGOahB90jDHGGFO3VM11leam+uOPP0J3++23h/zss88CANZZZ53QHXjggSF36NABQO2aXjnWN998M3RqhlU3VmtBzZB0WR166KGhGzp0aMi77747gJILCwAuvfTSkM844wwAxbs+dE1qVtz5558PAOjbt2/o1A1JNGtFx7fFFlsAAJZccsnQ1UomiJJlZr/hhhsAJM3NOhb+r2bS1eK1qMfUr1+/kA8//HAAwHLLLRc6NaO3lqwyvX4++eSTkI8++mgAwAknnBC61VdfPeRaXIu6v/C+0KdPn9Dp+uR19+uvv4bup59+CnnWWWcFUPyeomtqr732Cplj0f9t3759yMzgHD16dOjUzbrjjjsCAM4+++zQFeEm19/Uc80MP3VHqeuR+8pKK60UOmZi5YUtOsYYY4ypW3K16GRVjuWT7jnnnJP6OdZf0b/rEy2fjvN+CmwpHPfw4cMb6FoTesx6rmnxeOSRR0J3yy23hLz11lsDAOacc87Q6RM9g5TXX3/90FXr7Uvfgt94442QL7roopDfeustAEDXrl1Dp2NZdtllASStHKxNAgAPPfQQAGDxxRev1GHngr6RfffddyEPGjQIQLK2xTbbbBMy3zhrvfaMrqnPP/+8wd+19lMtWjmy4BrW/WX//fcPmddVr169QldLgf9E95dx48aFzGtJA3g1WJ4Wt/POOy90DNAFihkrryW13NOyBiSPv3PnzgCSVuKJEyeGTIu5np/ffvstZFrPdX1n1dypNLpn6PV/1VVXhXzNNdcASFrcNBlgtdVWA5D04sw111whuzKyMcYYY0wz8IOOMcYYY+qWXF1Xag6++OKLQ77zzjsBJN0FWnOFgVnHHHNM6NKCQWudrDoHtR7kSLKCVR9//HEAwKqrrho6BsgBpRYIRx55ZOg0WHnkyJGVP9gmoudeXTM77LBDyDQ508QMAPPOO2/IM888M4CkOVZr7tDlqnVoFlpoodRjKBKd359//jnkH374AQCwwAILhG6VVVYJuVaOvzF0fDPMMMNU/17rqMuAbh4G9QNJ1wjry2htFnXncNxFN8XUMX3zzTchM3Cc+wiQdDPuuuuuAIDtt98+dNVy3WTB39RrRt29d9xxR8icP21hsfTSS4ecFpKhexVd5tVM5uD51T2PLiogGRid5tJmOANQcjnOP//8odP1adeVMcYYY0wz8IOOMcYYY+qWiruu1Jz2wQcfhPzAAw+EfNdddwEoZa8A6WbUXXbZJXRsGwDUZoZEmpvn6aefDp2a8ziWWm8RoGNibSOgVHPmkksuCZ2amWmGZD0LIFli/+233waQnN9qoSZSzZrq0aNHyEcddRSAZB2Zyy+/PGTWj9F1qDLPm9Y2WWSRRVL/twh4fGoiZ6aV/p3nAQC6desWctHH31R0rr/88suQed1p7ZJad2Pp8bE+1ZNPPhk6upOB0nWn8ztixIiQOX9zzDFH6PT6zRMdx++//x6ytnPg/6TtmQCw3377AUi6czS0oYjWCLxnqbt77733DpmZmPq/uv917949ZM5PVlZj2vnJG55TbUuhdcZ0T9hpp50AlOqJAcBGG20UMrOq83ZXKbV9pzXGGGOMKQM/6BhjjDGmbqmY6yrNnEYXFQAccMABIbP0epbpjaZlzZR4/fXXQ6YZvZbMzXosLPj04Ycfhk7NqXTZqbm4Voofqjtt1KhRIWvxRmZbLbbYYqFLMz3qd9FcCZQyLNR0WYQbT82tDz74YMjsrqvHxCKCQMkNp2PW+eO8LrjggqGrpUwlrtWXXnopdGz7AJQyxLbddtvqHliOpGWoaCsWzVCqlbnS9ffRRx+FzGydgw46KHRaMI/Hr8Xr1ltvvZDpetYib9rOJU+XiF7zV1xxRcjaXZ0uRXUD3XzzzSEzW+f+++8P3TPPPBMyXUIsTDfl7+aJzpm2QPj2229D3nfffQEA88wzT+i0BQ0/Rxc/kFyfLAipLRTyGJ/e07hXPvroo6HT+4Peqzm+tddeO/X4CinoWPVfNMYYY4ypEhW36Pz555+he+edd0LWOitpb0z6xsXAMq0zoDVb+L+1WNYcKJ0LfbpnI1KgtpsGaj0OfWP64osvQmZNJLYCAJJvgRy/zo82cKXFowiLnK4zfWPSN1oenwZrptWRyaqjw9+opRYJOm423Tv55JNDp8GcDKKce+65Q1eLa7WlcH632mqr0Om6r5Vgaz3nWnOMQbja9iFt/9QWCfPNN1/IDLzXBAMG+ALJZrWVmnfuhVqvSZv+qnWb180ee+wROrUOMAnipJNOCp2Onxbju+++O3QauJxLnZb/Pz5te6AtZnT/ZzsHtfjoXvv88883+C6dk9tuuw1AMhhYG2Dnad3RRr76O1pnjUkMajFlo2cAmH322QEkA8x1fPzeSo7DFh1jjDHG1C1+0DHGGGNM3ZJrHR01B2udjhVXXBFAso6F1lQ48cQTASQDzNSMWYuuK3XD0PSq7gA1HbOdQC0ev87DCy+8ELKajhkE15iJX03EY8eODZntPnSt5H0uOD7tkqy1O3TcNEMfdthhoVt00UVDTnNdaX0Jfp6tIvQzRaHrk4GtGuCq42M7DJ2fWnLDNQbHqu5STQzYdNNNAZSSIoDauRb1nL/66qshP/XUUyHfdNNNAJJtSTQY/qyzzmrwXU888UTIdP2oO0ddK3m67oYNGxYyW40AyfXJOkDqzlDXGl1eM800U+rn2fVcQx8uvfTSkPMMPNd97quvvkr9Tc6rJgNoMDLHr/e8n376KeRXXnkFQMmFBQArr7xy6m+VMz79LPc0dfdqbSB1w33++ecAgM8++yx0t9xyS8hMTNEWHnvuuWfIrCmkzwflrklbdIwxxhhTt/hBxxhjjDF1S8VcVzT9qjlUTYcHHnhgg8906dIl5BdffDHk0aNHA0h2N9WaJLViZlbUdEo3xvDhw0PH2gcAMNtsswEo3p2h8Pi1XpGWlR8wYEDIaS0Q0r5Lza1qhmeGUzWzrmh6VXO5mo6VzTbbDECy+7jC47799ttDpx3Z2VpC62QU3R1aXXY0eWtH6I033jjkhRdeGEDtZB81F45b3VXajoYZoGoaL/papJuJNbgA4PDDDw9ZM1hYE2fixImh0zpIrAmlrmd1IzAbSOuc5T1+uj51T9FMnbSsG63ZosfP+W1s/9BMSHVNa02vPF1X6o7SexavO90fDjnkkJDXWGMNAMm2CXremC2ne46eyzzGx+PXrCt1B6qbnq1JsrJuOX5ds7qXct9lCAsAdO3aNeSW7Eu26BhjjDGmbql4MLKibyE33nhjyOeeey6AZE0FrWPCzx199NGha001PfhUr2/MWuWTwWK1aJlijQcgGWw755xzhtzU83/PPfeETCsJUApmrub4ecxaD0gre+r8HHvssQCSb1z6FsH6FqygPOXfOb4i6gTp3Ojva02rxx57DEDJsggk61zwjbk1WXTSkgE0AFItzXxjVotXrVyLaoXSYFats8L947rrrgvdhRdeGPIdd9wBIJkAcuutt4bMBrVqMc97/JwTvf50zlSmJUSb4rak6agGWFer0ecCCywQ8lprrRWyeiwYZKxWHLWIMIlHrVA6f1zLm2yySeh0/8pjLrmv6Dzp+rnmmmtCZhKOJuOcffbZITNYWdeC3itp6dE6Qvr8oNdyU8dqi44xxhhj6hY/6BhjjDGmbsnVdaVm9A033DBkBtPp39XMyGCmueaaK3RaM6JWzMyK1hn58ssvG/xdx1qLrjeeUzUnZp1/omZMNQ0z8LN///6h03LlNENXszYLj5U1KoDkMTMAFyiZnNV1oy49Bomy1DwALL744iGzNL+6RvKec45Pj1mDIa+88sqQGbjIVg9AqbbVlN/RWtBzzToq2qiVjXSBUguEWoLXn7oY11xzzZA1GJPXl7qrttlmm5C5LnXO9X+LcB0z8Pvaa68N3T777BOyBg7zf1virgJK15qeP203kMe4+Z3qgtK2HRqMS5dxVjA817ImcLz88sshc99S17qu/zyvXz1O/R0dC4Oh2eoBSDaQHTp0KACgd+/eofvuu+9CpmtWW2jo35nsAdh1ZYwxxhjjBx1jjDHG1C+5uq4UNTGldffWEuaMUE8z5035XbWCHqt2cCd6/LUIj3/EiBGh06wrrdlB06n+Xes8sCu2ZgVoifIi2glwfNrqgNk3APDWW2+FfPzxxwNI1ozQmiTseq5zvvzyy4e8xBJLVOiomw6PRY9JM3jUDE6Td69evUJXLdN3JVHXqdYJuvnmmwEk1xkzPYGS66CW9hGef5a/B4Dzzz8/5Isvvjhkuua0jopef6wPpS0Cim53wXXZrVu30GnWkdYM4l6jtXUay2DUOjKrrbYagGQmr7qp83Qj67mdccYZQ1aXVtr517XK+dP51/W93XbbAUiGgxQxpzonev+m+zWrhcm3334LIJlVlfa955xzTug0tMB1dIwxxhhjhKpZdBpDnwi///57AMlqyLWOviWwfoC+JWtgXC3WKeEbz5Zbbhk6rZxKKw1QeqO47777QvfII4+EzCDXPn36hE4bvBbx9pEWLKjVurUi9BVXXNHg8zq/fPvSAG19e+SbaDXHmfbGq8HgWqX6iCOOAJAMoK4l60ZT0etr4MCBIbN+E5t3AsAqq6wSci0mA/CYVl999dBpZW4d36effgoAWGyxxUKnTS932WUXAMnA+1qZXz0OHZ9aLPr16wcgWQ1ZA+vTPAKLLLJIyKzIP8ccc6T+brVQK41WZKd1Qy1WWjOJTVvVM8AAcqBk6VCLURH3Et1z9PwyCP7uu+8OXdr+r9ehNmhlM2VWMM/6fHOwRccYY4wxdYsfdIwxxhhTt9SM6yqNWqx3kYUGu9G1o40wtYFpLUIzYs+ePUOndTi0wRpdIuqOu/POO0NmkFxao76i0eNQN53WGfnPf/4DINlgUV1eO+ywA4BkHRCt01JkYKCay9W0rzWR9txzTwBJc3AtuVGnhprLNQBVg3F5LWaZvmvRdcU1o/VetPaIBm7SdaXuGq2zw+9qLXMKJGs6bbHFFgCAjz/+OHRjxowJmWtA53TppZcOmS7lIq5DveZ0/zjooINCZpCu1gnSz1Hea6+9Qqeucd5Lip7frHYz3P/1+PRaZZ0yda1qGADdzJUMd7BFxxhjjDF1ix90jDHGGFO3tJmaGXfixIm52njV3KU1WVjfg/VMgGRtknLNWO3bt28DAJMnT851fBo1ryWy1aWTB23btm0DAJMmTWr2+HROdG1oCXO6R1jqe0o5raZLJWnXrl0bAJgwYUJZP5A1VrZ7UNOruiaZ7aDzWEkzcocOHdoAwD///FPW+PSaUtOxZmsUwXTTTdfi60/nTLsjM9MIAOabbz4Aydozuj7zdmnw+it3f8nq7s21qOOophuD4yv3+suCrht15zRWR6eS54LX37hx45o9Pj1mrXN05JFHhsxaZVozafvtt2/wXVrnS2t6lVuHrFOnTi2+PzSFNNepzg9bPGTVNuLnWnr/4P1BsUXHGGOMMXWLH3SMMcYYU7fUjOtKzXHvvvsugKS7Ks010lKq5boqqm1FOa4rJct0nkY1x1cp11UWjbXryHuslXJdZc1f0Rlw5biuskibs6LGWSnXVa2St+uqaMpxXWXRkhZAep+rZBhA3q4r7jVZ94y0sVRyfHZdGWOMMWaaotA6OvoUp4FJzKPXN7JarH3RGEW/OZdLXm8UtU5rnzcyLc1fvcyZqU+mpfWZdzJKS7BFxxhjjDF1ix90jDHGGFO3TDUY2RhjjDGmNWOLjjHGGGPqlqkGI+edXl4UTC/PK72uaJhe5/G1Tji+ek9Prvf9Zfz48XU5vo4dO04T11+9r896Lw+g2KJjjDHGmLol1/Ty5hSca4zWmCrbkiJRWbT28bfG4zfGVB/uGy0tcun9pfo0dn8vek5s0THGGGNM3ZKrRWfy5Mkhv/nmmyE//PDDAIDXXnstdNrdm91Ld99999BtueWWIc8111yJ/6s1+EYyfvz40Om50LFODX0Kbtu2bcjTTVeatlo5B/pEz2MaMGBA6NZff/2QZ511VgDTVhEtUztwrWqR0jTrQa1cW9MCav3lvqnd6Tt37pz6vyTNolC0FUHR40vrzl5UJ/rm0pS2Mmnd1fWex3mp5v5vi44xxhhj6hY/6BhjjDGmbqm460rNwexCDgC77rpryL///vtUv4MmLf38Z599FvJZZ50FAOjYsWPoijZT6rh/+OEHAMARRxwRup9++inkZZddFkDSRJl2/Pr3BRZYIOTddtst5IUXXhhAbbmBJkyYAAC46667Qrf22msXdTgtRk2zOr+NmV6LXItZx9zYWmvJbxR9zTWGngt19/7xxx8AgC+++CJ0Y8aMCXmRRRYBAHTt2jXvQ6wYlUz2qBbquvntt99CPvzwwwEAjz32WOiuueaakIcPH97gu9ZZZ52QV1111QZ/L8IdpOObNGlSyJ9//jkA4Msvvwzd8ssvH/J8880HoPg9XY+f60vdUnr93H///SHzXqefX3nllUPmXM0zzzyhy3v92aJjjDHGmLrFDzrGGGOMqVsq7rpSc1uPHj1CPvnkk0MePHgwAGCllVYK3fTTTx8yzWS33HJL6FReaqmlAAB77bVXpQ67RWRFoD/55JMAgGeffTZ0M800U8g///wzAGD22WcPnWYV0OWgrrlXXnkl5KWXXjpkuq6KRl0DzKZTd10lawrlQdpc0sUBANdff33IdCNuscUWodN1/69//auBLm/TLI953LhxoVNzf69evULmWmyOaVxN/3/++ScAYMYZZwxdLc0vrx+6UAHg5ptvDpnX55AhQ0Kn522xxRYDAPTv3z90ep0VnRXTmJuK86rzm1bTSl2bKue5btOyjwDgzjvvDHnQoEEAStm1ADBs2LCQr776agBJN8riiy8eMl376623XuhWWWWVkNOygiqFjkn3v3PPPTdkjm/UqFGhW3TRRUPeb7/9AAC9e/cOnWbd5rmX6PyMHDky5I8++ggAMHDgwNC99NJLIasbjhlW6q674YYbQt5mm20AANddd13o9P6Xh8uudnYnY4wxxpgKM9Xu5eX2+tCnQ/0dPrFlvQXy7eK9994L3YYbbhjy0UcfDQA44YQTWnRclep1pW9BH3/8cch8o9AAx/POOy9kBubyzR8AOnToEHJanQ+lsYrT1eoFlfVmyWDxu+++O3Qvv/xyyLPNNhuAlj+5V2p8uv70LY9BkGeffXboNBiec6VrUt/O+Pa24oorhi7rWkijJb2uOBZ949pll11CphURaPr51/X3ThgWZwAAIABJREFUzTffhHzYYYcBSAabq3WnsfHl3euKc3HMMceETms6zT///ACAPffcM3Q6vw899BCApEVn0003DTktMF3Jo9dVWpD5t99+G7oHH3wwZO5FagWhlQoAfv31VwDA6quvHrqddtop5IUWWihkrisdZ0t6XfF7tDbOvffeG7JaPLp06QIgef45ZwDw/fffAwBOOumk0D3//PMh0xK79dZbh049AmopT7sGWtLriuPTdaQeB71+qF9zzTVDpxYPBvkOHTo0dDPMMMNUj7k5pPW64vHffvvt8X99+vQJmfOm86dWKK2TtuSSSwJIrj9dq08//TSA5P507bXXhtyuXTsALbdcudeVMcYYY6Yp/KBjjDHGmLol1xYQWaanpgYujh07NmQ112nga7VRs7Wa8S699NKQGYSmtYN23nnnkBmsVYnaJkXWMtF5ZG0IALjnnnsAFDtPWej8MagWAPr27RvyxRdfDCAZTLfCCiuEPGLECADA448/HjpdnzTJHn/88aE74IADQtbAwkoF3vF7NEDw4IMPDnnmmWcOubE1QzeJti1hAChQci2oC6AI1J2jx0qTu7qrdP4Y+Nq9e/fQaTuaN954A0DSdalrRU3ueZJVYv+RRx4BAFx11VWh05pjDKxm0gaQXPfvv/8+gGRbHg1c1zXEmkLl7jP8/R9//DF0rJcz5ffT9a+1w3Sv6datG4Cku0PruNB1ru6sp556KmR1aVWqJhTn59FHHw2dzskpp5wS8umnnw4guQ/oWOlGZD02AFhiiSUyf7OSaO0iTaCZY445ACTvadtvv33IrP2j6JrTMA62dnrggQdCp3Oy7bbbAkhe0+Vii44xxhhj6hY/6BhjjDGmbqk934JAFwKQ7AReRGlsmuEmTpwYuiuvvDJk1uYASlk5b731Vui0jhDdWFoPh5HmQPF1OpqKmnvVdMwMH5qYgfJL1JcLTd86JzQhA0nXBc206o5ca621QmZ9lldffTV0amZmNolmBWpNFzXZc121xHSu55Sm4Ycffjh0N910U8ia1ddUk7C67l588cWQWT+oWrU9stDxqxuZcznvvPOGTjMAmcHD2iBA0nXJFjWaSadZI7r/lDN/aaTNKZDMSrrvvvsSvz3l71PWOdc9hS1ojj322NAddNBBIav7hRmuLUHdTZwfrZ2i60fnivVjslqYEGYPAkk3Cl3ndEECSTeJZggx87Xc+ePnWSMOANZdd92Q9ZpPa6eg+wNdwrrONMOMoRH/+c9/QjfrrLM2OJbmwN/q169f6PSc85i13p3OT1ptItWpG4zrTvcU3ZdZZ6c5maqNYYuOMcYYY+oWP+gYY4wxpm6pSdcVi8t98MEHoVMzKE2ylTRtpaG/SXP2RRddFDot8pRmJtay2Nrp9dZbbwWQzMQ58cQTQ2aGTJ6lysuBJkstcX7FFVc0+LtShGsjraAji90ByfW13XbbhcxsGy1SpmuN5m7NFNC1QtO4toi44IILQtYMILrEWnJ+9DeZQfPXX3+FTkvoVyprBgA6depU1nflQZo7aZZZZgndc889F/Lrr78OoFQYEEi682g61xL8yyyzTIPvByrvstLj4HECSZfIlltuCQA49dRTQ6cZVFzruv7SMnjmnnvu0DHTCShv38nak+lS1cJ9imay0XXcWIiCHqcW1FtjjTUAJLOetKDiEUccEbIW9WwuOtbRo0cDSLpD1TXOTEVF988jjzwyZLauyMoKpfsojxAHdU2loXPanN/X/2U7Di2YSHcsUMrK0rYe5Y7VFh1jjDHG1C1Vs+ikNXPTN1IN9jvuuOMAAL/99lvoNt5445A18KzS6DHxKR0oWVz0zUrriHTt2jVkBqEtt9xyodM6M2zqphahX375JeR///vfALIbdhZhHUmzbqkVSoPF+b9FBI2nBUACJUucBpiqxUODxVkCX9+uGzvn+sbB0uiHHHJI6PTtW4Mk2Q6kJeg1xetHA9y1lH/a9ddYgHhaI0igFASat0W1MfT31crE8896M0DSksfAyD322CN0Bx54YMgMjNUAykrUvJoatBKpRUAtSgsuuGDIrON0xx13hE7bkWyyySYAknWs1CLFwFdNIND9Vfet5o5V50QDbFlzSvcEDUZeddVVQ+a5aOwtXo9NkznYwFYtdmwbAVSuAa2OlRZ7tcKo5U/ngteqBijr/LIO0GWXXRY6TQyg9U2bQldqTea1Z0/RQgRAcsxau4ntMtSiUy626BhjjDGmbvGDjjHGGGPqllxdV2oiVDMm60NosNj1118fMt08ao4+9NBDQ2a5bK1pUynUxKguBnbS1XoPWttigw02CJkBXTp+NXPuuOOOAJLBalq74sMPPwSQrNPDAC6gmDo7aSZpDbZOQ83pzWlBUA46fxqASjO/uj21hLwGrpZbepzHoPU6tA4GS/SXi55HBsMzkBZIuhO1eznnQgOX9bsYOKmuP4XB1Gqar5abUn+TLlQAOO2000KmyyptTEDJZaPuEoWfyzsZQK8prgnd50aOHBmyuvGJutGfeOKJkDk/6hpS1yzXt14L+r9KOdeqfpZzpXu27pk6Fy35zbQgWV2TaXWGykW/h6ELM844Y+i0O/umm27a4Pi0bYK6KXn/007ibBsEAPvttx+AxusM5Y2u3yw5Tcf9UcM99Fxy3VfyPmGLjjHGGGPqFj/oGGOMMaZuqZjrKq0L7PDhw0PWcvp0KajrQ+vQEDVzaglvduXVstd5mO7U9MlI8csvvzx06ppJy0DRz+vf2RFYayNoBgV/Q2v2aOn0vMfdGMwsUNdIWp0ZmliBYlpcqOuBv3/MMceETs+j/m9jptc0k2raZ3R962c026BS8Pt1fWpHaj3nzMbSOh7qpmLWjWYdfvrppyHTJJ/WUTkv6AbW7Bl18zz77LMh77nnngCSLRS0e/WwYcMAACuvvHLoqpU1lpXpxtL7dFsDybYjulZZZ0zn7O233w6ZLncds7qO066/vMdPd4WOX91l6nJuCRomwNYs6vpT8nBdzTPPPACSWZSa9bXDDjuEvNVWWwEAevToETrd659++mkAyQwuPVesGVREWx09z7pnajYrO9Sra03vFe+//z6AZG05nX9e43r/1N9tiZvcFh1jjDHG1C1lPUbrUxbfEjSo9swzzwz5u+++C5lPelqFUWtWrL766onvBJLNCvl2ecopp4ROn6TLCTzT/9fjY82btKAyoPGnzLRgOX3L0poracFo+namVUzTLGmVIsuKwWDPr7/+OnQawMumpd27dw9dETV10mqfaIC7rt+0BnU6/rTAd/28Vlnmm5zWKdH6NqzcqsfVEvScskrzkksumfrdWjOINTn0LVHrdPBzWk2XtZ2A0pto3nV00gLgWTUVKL0ZAsDNN98cMt+ehw4dGrrHHnss5M8++wxAcn3kUe24MXT+WBuHNcSAZAVfhTW3tE6QNmP86quvACSrlbORJ1C5RpYtQdfcK6+8ErJWdu7ZsyeAxtdX1ls+z6XqaHEBkveaSle2Puecc0KnXgqtE8N7iQYYa2Aurc5aZ02tUxxXNeeP49M9/4wzzghZrzUG1mdZf8aOHdvg71qTjmtZz5/WwdL/beo5sEXHGGOMMXWLH3SMMcYYU7c023Wl5kQNlmKjSjZEnJK0YFR1I1x11VUhs4S5mqXeeeedkI8++mgAwF577RU6raOx9957h8wgp6aauPT/tCYCf1+DBRlUDCTHl+amSft91alJly4fPdcazFUt9PdZlhsolSbXADL9XzZrK6LOQ1ptC6C01jToW2uraAsS1iTRv2uDRY5Lx6w1e1gzhEGHQLJmj5rOy3Hp6VhZByiteWBzoUlZa0ZpCwG6FtJc15VEv5/Xn7oItUT+rrvuGjLnRV1venxsGpnV4qII2NRSy96nNccFSvvS8ccfHzqtScMWNNpO4o8//giZwazVHDPdGToPuqddeOGFITOIXK8f3V953PpdDEAGSi5/nd/zzz8/ZG2tU+lzoG1Xrr766pC15hPnQudX9yquX3VHagID94+850/3N7qetJExg46B9DpMWcHSdD1l/Z2uLXWNaSPkjTbaKOSm1jyzRccYY4wxdYsfdIwxxhhTtzTbdaXuiiFDhoR87LHHAkiW0lezfFodD3Vzqek1zZyvNS/oflAzmmZgqWmLZbabHJ0t5k41zbGEv2ZCaQbW5ptvHjJbVNBEPKVMtKy7uuZoslRzoJ7XaqGmRa1TktYaYP/99w+Z57+ITCuN7ldzJ1uMqItTM9m0HQKzWtQ1qZ2C6RJZb731Qse2HgCw2WabAUh2gW6sjkm5cH1Xwpyd5hpgpg6Q7VKpNGmuXdWpC1CPietT2yLo39UlXW10f9E6YzTTs/M2kMz6VNcH1yezMwHg448/DpnmfN0f55577pDzvC6z3PHMRurTp0/oNBPpxRdfDJlZtXp/Sds/dR8aNGhQyMy60mtS7wlZx1sJ9NxqaIbOpc5FGtyLdK3QnQOUMiTzrqOj54bXDzNqgaTrWPeKrHYiU6LnSrNamW2oc6Z7aUtas9iiY4wxxpi6pdkWHX0K0wApVvHU2ghau+DII48MmZVLmxOUqU9x/F21CO2zzz4h65uA1t1oCvpkqk+RDHbU8d14440h33PPPSEz/59Bj0Ay2IywngcA/PrrryHz6ZlWMqAUoA1kN6urBPqWoIFen3zySch8u9T/1QZ1fPsvtzlmS9DzodZH1lbRar5aW0XrjND6osG4alHjb+gbl5Jm8SiignVL4bzqOlOLV7XQuaQVWIM9tWaJ1ili4Lw2VdQqtGx8Wm611Zagv6Nv9mywedZZZ4WOlgkgWXOlc+fOAJL7q1q8Dz/8cADJRsDVIs0KAJSOTyt377HHHiGr9ZwVu3V/1fOmDaKJBivze9XioLVXqjXXzannltYUU8+fVh5mYHc1KyPzt7Selt6/NdmClsqspr88F1onR6vo03uiVuTm1KxLwxYdY4wxxtQtftAxxhhjTN3SZmomtYkTJ07V3qamKQbIaYDZYostFrKajsttW8DPqztLS4iry4Gy/lb79u3bAMCkSZOaPD7WDNI6OhoA99Zbb4XM/8kyjacFji666KIhM2BPm/o1VqdHadeuXZPGl4aaQ7UtgDYNVDMq0TordC3mZSIuZ3w6J43VUWlpK5Fy3Ykc3+TJkwsp7sLzoq4TbWD7zDPPAADmnHPO0DXHNde2bds2QOP7S9oxaVsNraPFBACgtO402F/dGKxZk1cdEu4v48ePn+oP6LXG8em1xT0VSJr56TLQRp+NNRWu5Fg7duzY7OuPx6fubG26qvWRGB6giR/qJtl3333x/38/dBpgTjdgS5sf8/przvosl7QkmHXWWSd0ui7YtFVrLjUnQJfrc8KECWXdH3RNaWA85y8raYFzoa5bvdcxcLul948OHTo08OnZomOMMcaYusUPOsYYY4ypW8pyXSk0vWmmi5qeWpL73hhqRtPfbSxCu6muK4XjUxOjmk41E4ByY1Hxeu7VNK3R5qQ5ZrxyXDtK1vlNG5ee87wzjCo1vlqlVlxXo0ePDp26idgpXDui5+264prT6+ySSy4JecCAASEvtdRSAJLdv7WmEsdXtOsqDd1fsvaPNNd3Nds5tMR1RdLcdUDy+LmWtG1FWh2krDGnZQ02hyJcV2mtlbQ2mYYGsPVHWluMplCO60rJmsuWoMdfbsiDXVfGGGOMmabwg44xxhhj6paKua5aEy1xXaWRVuRpSrmppJmhW2qOnlZcO/U+vqJcV1y/Wpb93XffDZlmdC3C1py12hLX1ZTHNqXcGNV085TjumoNlOO6yiJtLrNcW43NX6WyHou+/2W5g8p17VTKdVWr2HVljDHGmGmKZreAMCWKCgY0Jk+4lrXtwBprrBEyg0WLWPO+5uqTtLlsTW1T8qCIpsj1ii06xhhjjKlb/KBjjDHGmLplqsHIxhhjjDGtmanG6NR7VHa9Z+0UnTWQF8waKCorKW+YleT12Trh+hw3blxdjq9Tp07TxPqs9/2l3sen2HVljDHGmLrFWVemSaSV+86rO7IxlaCxEvVev6ZScP00pc4T111L6q3VE421OKkktugYY4wxpm4pxKLDJzlt1KZPd3yi0zoKfuOqPvoWrA0ezz//fADACSecELo55pgj5NZY/yJt/en4da02Vt+Cf/eaLYa0pp2//vpryJyfzp07h07nt7XMW5bFKu1NWddsPdZnSbOkqC5v613aOZ88eXLq72szzmmZv//+O2Rdv6y4Xsl5skXHGGOMMXWLH3SMMcYYU7fk6rpSc56ahv/66y8AwDfffBO6H3/8MeQFFlgAALDwwgunfldrMS23dvSc33HHHSE/8sgjAIBjjz226sdULlkBcBMmTAiZptNRo0aFbujQoSHPN998AIBOnTqFThtgzjvvvACADh06hK7W12yWm46oO7IWXR96/JzLW2+9NXSnnHJKyIsvvjiA0joGkm6sWpyrNDeVuka+++67kH/55ZcGf+/Ro0fIM88881R/q1rj1zlTuaktRvSc/PPPPyH/8ccfDb5zhhlmCDkPNyWPRV38Z555ZsizzDJLA71eU9NSYDLP/wUXXBA67pkA0Lt3bwDJ9VsutugYY4wxpm7xg44xxhhj6paKu67UBKfm/LvvvjvkO++8EwDwww8/hE6zImha3W+//UKnpudq0VgdjpbSGrNyBg8eHPJ00/2/ZdOajp/nXNfk008/HfLtt98eMtff999/H7r3338/5O7duwMAunbtGrqRI0eGfOqppwIA1ltvvdDV0rniWlZz/5tvvhkyXT5qWlc35WKLLQag+DFlZdVcd911AICzzz47dJdccknIm266KYCkuyota6dWx/fMM88AAG677bbQqWuVLldd67vvvnvIF110EYCk6zXrdyt9DtRt9Mknn4T88ssvh7zvvvsCSO65zXFjXXvttQCAJ598MnTHHXdcyFtttVWDz5U7To7r22+/DV3fvn1D3mSTTUJm6Mb0008fOv39tJo8afcf/Yyuj1p0g+nx07X63nvvha5bt275/n6u326MMcYYUyAVs+ik1S7o06dPyFdccUXIBx98MICklUbfrl577TUAwH/+85/QLb/88iFvscUWAJJvpJUi6y0iLcCvsSdn/TytIAAw66yzNvitWgzw1OMfMWJEyHPNNReA7DfCWkHP71dffQUA2GWXXUKnAZwaDLfUUksBSFpkNHButtlmA5AMMHzooYdCfumllwAAa6+9duiKfsvSN+mvv/4aAHDppZeGrn///iHzWlSLgFp8Bg0aBKCUNAAUUztJx6TB8ldddRUA4IYbbgjdDjvsEDLnYvjw4aFjADpQsugVnQChv//cc8+FvNtuuwEAunTpErptttkmZAbj6vWp4+dY9txzz9CtuOKKIR9++OEh8xyXO35ei7qmaHkDgAEDBoS85pprAigFjQPp60uv7z///DNkrtU33ngjdG+99VbIvH8A6YH3LYH3BbUSq5Xi6quvDpnzknVO06xMaYk73KeAZLB1LaJzReu5Wsmvv/76kPPYS2zRMcYYY0zd4gcdY4wxxtQtFXNdpZnhNABuyy23DJmmySxz/qqrrgogGSD6xRdfhJyHG4DfOX78+NCp64zuiN9//z10WW6utHMx44wzhrzOOusAAA477LDQ1UodDx2TuqvGjh0bMl0yai5V1xu/o7GmbXm76/Q80h2htZnWXXfdkA899NCQGWTcvn371O/l+LTOzuOPPx4ygwz194twXalZXoNVGeSv7lStObP00ksDSLp21l9//ZA///xzAEnXVbXQ9akB4DfddFPIDDzdfvvtQ6dz8eKLLwIADjzwwNBpCxO68TTYvOi2Juo6ZDDrNddcE7pdd901ZB5rVgLFgw8+CCDpbs2qSVOpxAMei7qLH3jggZAnTZoUMpNUllhiidCl1ZzR47zllltCZuiDrk89P23btk393uai54xrUe9Z88wzT8h0d+vnGnNdaZgEA8gBoF+/fgCAk046KXQaJlIrYRC6/+heybnaaaedQqfnKpeQlIp/ozHGGGNMjeAHHWOMMcbULRWvo6PmuGWXXTbkNNNoVp0AuoloIgeS0fp5uq7GjRsXOq2DQHOnuq7UxKZuDmZVaSaAmkiZATBmzJjQaVZakWZyPbe//fZbyOrSSyvRrp9jTSTW+wCS5lRmMzF7S7+zkujx0TWoJm41Yauctr7UtP7OO+8AAE477bTQffrppyHT9aGuoWqZk9VcrDU91EzMbBZmJwHAggsuGDKvRa0jpMfPDLYiTOQ6vqeeeipkbSGz3XbbAUjOqdbsOvroowGUXOQAcM8994TM9auZeEW7rth2BCjNz4cffhg6dXPw73rMrF0GACeffDKApGtOM60q1bpEryNeP7qnau20ZZZZJmS6rLLOOdcAs8uApBuMezjr8QDAkksuGXI56zYrk/bee+8FULo2gFK9JiDZsbwl51Tnl/umuuOKrvnUGM8//3zIzPrUNZf3XmKLjjHGGGPqFj/oGGOMMaZuybV7eWPR05rVo2bos846C0DS3Ljzzjs3+XtbAk1/M800U+ieeOKJkOmyGjJkSOj+/vvvkNW0zEJOappT0zpNw+oGq0XUnahm2rnnnhtA0jWg42Mhsrfffjt0arplIS3NVNLiZ3mYMTkWLQynqOmX2UZa/EuzXliuXtevFr+kS6Sa5uS0TLCjjjoqZHUNsDgXXaxA8pqia+DVV19t8P1AadxFmMt1bayyyiqp+meffRZAMutTCwrqWiVrrbVWyHTp1Er2ClByxwEl16hmXWnBQGbNaQaQtkCgm/LGG28MnRZkbSyDtKmkubPvu+++0GnWpmYVcX/J6u7NfVdbfAwbNizklVdeGUB226BKjUnDKb788ksAyexaLUyo+19jblD+ht4ftEUG1/UiiywSulp0Xam7TQuuci3POeecocv7+G3RMcYYY0zdkqtFR0kLktQS5Fqum9x1110hL7TQQiFrYGil4BOlHqcGUxN981PSgq21kZu+PfF/1Uqiv8s3yayg6yLeNPWJm28y+sR+zDHHhPzBBx8AAAYOHBg6tRgwSFADE6tVByLrzUH1tHjoW4hanFifg80tgaSlb4MNNgBQerMG8g9m5VxobRi1OGnNFFpy9JjSWkTomtU6ILSu6meqFayrv6M1kbbeeuuQzz33XADJt2Bdizxutciqxetf//oXgOLfkvU60AaQvFY0GPXEE08MmRZTDdDdaKONQj7zzDMBlJrTTvlblRq3WgHZYFPPudYu0mslrQ6XyrTOaYsPPeZevXoBSFoss/aU5u41Oia1njLBRK04eq9ozvXBseia1SBnniu9f+j3p81lNet48RzpMavFbf/99weQTODJe/+wRccYY4wxdYsfdIwxxhhTt1TNdaWmRZphtY5Fz549Q2b9Dq1tocFyWaXNK32caSbcLHOq1py55JJLAJRcOEDSzM/v1To7WvNk9tlnB5B092SZsSttXtfv0zobGjjITrrq2tE6CWydsfHGG4dO65jwu7T2TNHdvfX36cbo3bt36DbffPOQdd0SdvQGgMsuuwxA0p2gbpZKocfM9aEBxHRRAOntDDSY/Oeffw6ZdWa0js6xxx4bMsdSS8G6uhYZEKoBrupGpl6vI3XtvPDCCw0+U3Qnc93zWC5fXT9sewCUXAbnn39+6Pbaa6+QOe95uKvUnfLRRx+FfN555wFInketzaVByrzW9PhGjx4dMusHqbtDg31Z52qFFVYIne5fbAcDlBJDmjp+/T+tNzTLLLMASNYG0vFpkktaHTKdX7b4UNcc7wlAKVhb22no/qpuQD3GasE51r2IxwwAG264YeL/qoEtOsYYY4ypW/ygY4wxxpi6pWquKzVDMoPl4osvDp2a7ujS0do5zz33XMg0KedRT6cxssyN6obiuNRcqq4rfu6VV14JHdsiACXTHk2YQLLcuZZzp/m9Um4E/R6tCaFZR48++iiAZB0c7aRLM7men19++SVkZl6wi/2Uv1s0rEmy7bbbhk7dPGnHqq4BunauvPLK0NGdBSTXQjnoWuTxHXDAAaFTd4a6Eele+OSTT0J39dVXh8zWHWou12uR2RJFXH9K2viB0nFrzSR1bbH+iLoWlltuuZB53tSdVS2y2uJoBssZZ5wBoORCBpLngpl3zG6ZkjyvNT3+sWPHhsyaMLr2td2Ounm1NQnRvZBjVdej/i472WvWmbrUttpqq5AZZtDUrB89z+oCY00nPXatXaQ1nZjNy+w+ILm/Dx48GEDy/OhaYP0kvX9o6Edahm813a38Tb1na1YjM6iruX/YomOMMcaYuqVqFh19IuUbRdYTHZ/U9YlYmzGus846AJJvcUVXaWUwGlBqmnf//feHToM9edxaWVMbfDIITcenNSEmTJgQsr7VVBo9pzo+HrcG2OkbC9+edH717YqolaCWLDppFpe0Nz5d01rbiTVDNDAxb3jMBx98cOh22223kFnbR/9Xx6RVSrmmdM3VUoNLom/xGth+3XXXAUgGUGsdLq5LnTNNBuD36prMMwFC0TF9/PHHIe+3334h87hp+QWSdZLSAlCrtT/q8eueRouOWrm1DpfOBYPgs4KlaVHMqlJMS94hhxwSOk0W0fptzT0v+v96fPytd999N3SaTKP3AjblZKNhILlX8F6R1QiUlhytfK7Xvdanqda8q8WMweLa1Fnv5dx/bNExxhhjjKkAftAxxhhjTN2Sq+sqrREgkAxyTIPmtqy2DzSdZ5n2qkVabSAAOP300xv8rwam8vhZqh1IBsjRpKfBbptttlnIqq/WuNloFUgP0lRzKU2T2hZBg0FpZlXXV9Hl9htD1zJN5hqAzgBIALj55psBlFyYU34+D7imtF7O3XffHbK2WKHLYPHFFw+dukO5Fvfee+/QqZm9VtyMek55zhUNxk2rY6UBruomYh2Xarmr9LfvN8PPAAAgAElEQVQ0aF/dVeqmYc2ZL774InTquqKbRl2M1RqLrg02DwVKrk914WttNG2dwNYquia1ThWTINS1r8G4bBq5xx57hC7rXtHctZxWuwoo1TTq169f6LQpsF5/dO0MHTo09fgZ2KzJHuuuu27IrFOm7uSsOk9FtIBgkLzWPtKms0XsH7boGGOMMaZu8YOOMcYYY+qWiruu1ESsdTpYGwAodbpuLKtBaxJoBgjdJLXk7tBjYQS6ZhgodE2p60DdXWlZLVnm1jzPgf6OlhVnzRh2+QaSri265FijAgBWXHHFkDn/aZl4lSStXYeawzVTQGWidSzY0RsAfvzxRwBJM7VmW5x66qkAsl0P1UJN2/PPP/9Uj4XmcKA0F2purhV0zWimitbsYDaKlp1X1zczdLRFhLr8WD8py02RBxzXyJEjQ/fee++FrBksSyyxBADgscceC51mMHEsae66vNG9S/c31t5S1xXrGQHZbhiin3vwwQcBJF3/WgeKLjM9ljwyBfU4+f2anbrLLruEvP3224dMl7e6djTrkfuOuv7pzgNK48tqEZTVpihP9FjYyV3dcfPNN1/Idl0ZY4wxxlSQilt09I1L30KeeuqpkFnfQ4Nq9U2bNXO0KRirQQLpdUBqCb496VOuvnHx7wsssEDodCx84i3aYqW/r8F8tD6xwjEA9O3bN2TW8dAAazb1A0pvb3mPT9cUgzw7deoUOq38rFVcv/zySwDAwIEDQ6d1MDiXWq9EG4Dus88+ABqvppw3+ptpb3xq5dCx7rvvvgCSVo6i12Iaes3oXDOwWCtD0woAAI888giA5F6ldZ6YBFErQddAcix8+x8wYEDo9I15/fXXB1BdixTRdaK/T4uEWiaUtJoqqtNq8Nz/eZ0BycDntKaZeZNWe0nHrxZjWn202rySdvx6/0i77xVhMdbf1L2E+6fuj6yjBJQSG6o5P7boGGOMMaZu8YOOMcYYY+qWNlMzH02YMKHZtiU1B2sAJ2sbACXTlbpuPvvss5BHjRoFIGlO1qZg5ZaO7tChQxsAmDRpUi62M54DrXPRp0+fkNlgcccddwydNiAs16TXrl27NgAwceLEXMZHk6U2RWQAGlBy2aywwgqhUzNmuS7H9u3btwGAyZMnNxifmlO1UR7bAeiaXHTRRUPWFgAffPABgOx1xiBzdadqOf5yG+m1bds21/VJ0//LL78cOq3TdO+99wIo1ZMBkqbzcqnU+lR3gDZN5bWmweQ9evQIma1HNEBUXSqVWp/jxo1r8vi4Z2htJm1hoa5TBpZrU0+t2cRgbD0/lXQTdOrUqcXrM6veS9r/6PXHRphA6RrWth9ZYQAtgeszbX+pBnTD6vrU/emOO+4AkNxfmzO/3F/KHZ/e67Um1dprrw0gGVivDUi5fvNyp3J8ii06xhhjjKlb/KBjjDHGmLql4q6r/8veW0ZdUbbv/wdriaDYiq2o2IiBHY+iIqBYjy0WdhcKdgd2i1jYqNjdnajY2CIGditKPN+1/i9+/+Pcx3jPcNeePXNvj88bznVy7z1zzXXNtWfOVLSOg9aEuOCCCwAkM3nYkRyomL5YVrva5O26ImraUzePuqlINc14ebuuiI4vreO3mp6raTqfmutKUTM5zajaKuCJJ55I/RzXpbY9UDPyWmutBQDo3Llz6PRatHasebuu6NK44oorQqednpkhqR3Py+i60vnVtUaXpd5T6jpluxZ17VSzk3JLXFdE15FmNTITFQBuueUWAMk6Ldq9muPKyzXQGtdVU0jLYHr22WdDZgaPrk9twVMt139Rriuev7pe1TVE17lmkDaHarmuFL3mPFfV1bLdj11XxhhjjPlX4QcdY4wxxtQtubqulLTuz1nFpajPy/RaK9eVkubayMuEVyvXVVE01XWlcH1Vs7BW3lkDebsGPv/889A98sgjIW+zzTYA8utYnsf61HlNayeT1tE5r/uvNa4rpTkdx9PGlxd5u67SSHON59UKp2jXFclq5ZD2+9icfS0P15WStm5rWXzTritjjDHG/KuomUWnTBRh0akltui0bfK26JCsYHIG5tri2DKqZdEpK0VYdGpJWSw6eZG3RadobNExxhhjzL8KP+gYY4wxpm6ZquvKGGOMMaYtY4uOMcYYY+qWaab2n/UerFTvwXQeX9uE46v3ZIB6D9at9/mr998Hz1/bxMHIxhhjjPlXMVWLjjFTo7GCVo7/MkXSnIJ7adSyyJkxJj9s0THGGGNM3VI1iw7f3vUtSrsD/9///V8DuZrl+MuIFmFLa4GRBd8kq9lRuZrw/CdPnhw6dowGKp1q27dvH7qyWHd0Teob+7/97T2txH7WmuW6LNM1S7unfvzxx5BvvPHGkF9++WUAwC+//BK67t27h7z33nsDABZbbLHQlWmsjZE2l7r/GlMLsgqSci3W8p6yRccYY4wxdYsfdIwxxhhTt1TNdUUz1Z9//hm6559/PmQ1DS+44IIAst0ZjQWzlsUN0ljH5O+//z7kcePGhTx+/PgGn1HT8tJLLw0AWGKJJUJXy+7Eaei5/vbbbwCA448/PnRXXHFFyEceeSQA4Jhjjgmdmi6LOH8e86OPPgrd3HPPHfIMM8wQcmMmVY5Fx1F210baWlXdl19+GTLH9e2334bu559/DnmllVYCAMw222yhK2L8ev56/9x///0AgHPOOSd06mbt27cvAGDttdcO3TvvvBPywQcfDAC46qqrQjfPPPOEXKv1m+Xipj4rGYD7y99//x26hRdeOOSy7J9K2liyzjPtupT9/muMtPFnBdNzrEX/JjTGN998E/LHH38c8rLLLgugEuIA5H/+tugYY4wxpm7xg44xxhhj6paq19G5/fbbQ95zzz1D7tevX8ibbbYZgKS5asqUKSEvv/zyACouLgCYaaaZQu7UqROApLmvVqZLNSdqVsdrr70W8pNPPtlA9/777zf4XFrtGQDo2rUrgKTpvXfv3iHXKptJz09N/yeddBIA4Lrrrgtdx44dQx46dCiASvYKAMw777wh1yqbLG19nHrqqaFT19Upp5wSsmZmEXWNvPTSSwCA2WefPXSLL754Fc64umTdH++99x6AZCbSo48+2uBzP/zwQ+jUddWjRw8AwK233ho6vZZFuBF0Td18880Akqbxyy+/PGSuxemnnz50c801V8jDhg0DkDS36/rN457jvpLlzlb5119/BQB88sknoeOeAwD33HMPAGDdddcN3Zlnnpn6XbWC1yzL3a97Kd3g7777buh0f6HrQ12PSy65ZMic17K6dtJcU3/99VfIvD+vvPLK0M0///whr7POOgCSY55jjjlCLjq04dlnnwUAHHDAAaHT0I1evXoBAK655prQzTLLLCHnsT5t0THGGGNM3TLV7uXN6YXBp1QNarzoootC1sCkr7/+GkCy9orSoUMHAMB3330Xuummmy7k/v37J/4FktafxmhJrys+sU6cODF0++yzT8h8itVz0XNec8019fgAkpYDHSvfSNVyM2LEiJBXW221kNPenqvVC0oDiEeOHBnybrvt1uD89G8nTZoEALj44otD99///jdkWuda+ubf1PGlWTR23HHH0OlbyPDhw/X7G3xe37h22WUXAMCrr74aOp3/Ll26AGj5m0m1el3pnDz00EMh77zzzgCACRMmhI73HADMOOOMAJL3Z+fOnUNmMDKDdoHkmmxs3Hn0utK5YsC5jmnRRRdtcH5qZdWx8L685ZZbQqfB6o29MTe115WuP+4rek4PPPBAyFdffXXIvL90TIssskjIW2yxBYCkFVivRWvf+JvaK0mPkxZA/MILL4S87777hjx27FgASSuy7jWUeZ8CycQNzuWmm24aOt2LdX2mWZry6HWVthfp+NWiPGbMGABJi83mm28eMi3KvI8BYKONNgq5qfdfa3td6frVvZAeG73+W265ZchMVtG9eNdddw2ZltiWrlP3ujLGGGPMvwo/6BhjjDGmbql6C4gFFlggdGeffXbIGizI0utacyetZoDW8fjqq69CZjAsTXxAuutBz6u1pLW40GDbQYMGhczARQ2w1mDHtLGqGZN1QNR0q6bzPFtn6Hera4MBmkB64KSeK/XqDrjttttCZmCoBtjlHSDJ+dPaIqusskrIumb++Rkgafqma+rpp58Onbo0y4jWOaLrY6mllgrdJZdcEvKoUaMAJOskac0cuqQ1QaDoOiY6V3Rj6P2nNb2eeeYZAEl3kI6PQaB04QHVG5/eMxpsO2TIEADAiy++GLqVV1455COOOCJktqZg0gYAzDrrrCHT9ZbmoqklabVhHn744dAdcsghIX/22WchL7TQQgCSySz6//feey+ASlA2ALzxxhshDxgwAACw/fbbh+7AAw8MWWu61Wov1blgksl5550XOv19YP2m9ddfP3S6v3D8RaPndPTRR4fM664B8HfffXfIn3/+OYBkMosmKzEwuZpr1hYdY4wxxtQtftAxxhhjTN1S9To6WeYmzQCZc845ASTNdfo5/i3ryQDJmhE0KasJUjOY8jDT8js1+n+ttdZKPSbN3I3V+VHXx1577RUyXXbMbgKSEfh5ugl0HFoCX2sCpbUQUNMjI/A1k0zrfGy33XYAklktedcpoetUx9GzZ8+Qs2oapf0/3Rzq7soq114Wfvrpp5BpclZ3lWYFshaWmqZ1frVmBynadaVwrpi9CCTbkdAlq7WP1A2nbthqwfVBtyEAnHDCCSGz9gszaoBknSZ1nfK7NBxAXSO1qlPVGHrP0GVMFx1QcWEAwDbbbBMyXabaPV7nRN1QRMMcTjvtNADA9ddfHzp1Ez722GMh87ekWnuOjpkhGkBy3HSNaibVhRdeGDIzHLV21Q477BAy3XSsJwfUzjWpv7PqDtY6SMyqevPNN0O3++67h8z9X2ua6VrPYy8p9+5sjDHGGNMK/KBjjDHGmLql6q6rppBmmlLXFguVaVaEZnCx9LeWmFbXQZ5mvKy2FWlklXOn6VSLZOn/s3iUZnXl3f2b56qZDFqwTOeMxce0yNPJJ58cMgtCPvfcc6FTkz3/dvDgwaHTcuB5trjQ79P5SXN96pj1czTD65g0g7CMbLXVViG//PLLAJJuEm2HwNYQWoRO1yrN80W0Emgp6m5joc+DDjoodJoBcuKJJwJIZgpWq+Ceril1bbCgmhZmVHcNO5Lr57TtRprpP8sFUEQGFt2Imh2lWZeagcRxqQtO3cSaLUiWXnrpkFdccUUAyRYYb731VshffPFFyMssswyA6l0T3VO0+KOur0022QRAMhNS7z+Gaai7S7NyWZxTQzfynlPe8+rO1tCD9dZbL2SORX+/Vl999ZA5Lr0n83a32qJjjDHGmLqlZhadtDdpfePQJ27m36tFQNs9sGaN1pYpOhgyrVmdNkXUOjR8ktc6LmrdYP2MWjYt5fxoAJ++8WvTNVraNJhXrTALL7wwgOxgQjaQ1O/Xtyz9XLWsBpwTrfP09ttvh3zuueeGTOuMlijXYGm2g/jtt99Cx7YDQOWNsmiLh95zbFsBVNpVXHDBBaG78847Q2awrgZLMoEAKE+waxa8V7TtiNYkYRKEvkXzzR4ATj/9dADJOWUjSaB6b88aYEwr7h133BG6rKaPtJiqRWeNNdZocK5s/ggkg61pHc6yWFYL/U5asfWeOPbYY0NWS1ba+mpOg06OVS3+++23X8jqHVBLS7VRK7zu5Qwi1sQI3Xdp9fr0009Dp2M+7rjjACSteHnfkzx/rUP2wQcfhKxjveyyywAk20Fpg1LOdS33EVt0jDHGGFO3+EHHGGOMMXVLzVxXatqi6VW7m6vpnMFaaR3BgYoZrUzuKjUtMnCMrSqAZHdXBmHRRA4kg+kY5Jz3+PT86VrSejcabK11HPr06QMgvXaQ6tVdpX/LwN88y6//Ex5TXRhnnXVWyFrun3+rrh2WpQcqAdtptU2Kgtcya0408I81ZXifAcA777wTMt2QWq+kVsGOWUkFzbkX+Dl1bev8Eg2sVDM7P68tUKqFmv5Hjx4dMmuG6Zz07ds3ZG1RwSB4rS2mbmC6lrVtRK9evUJmnRPdU6vVLkLnT133vL7aqmKllVYKWX8fuO+0dH/g5zRAVt08WlOn2ui102BrbdfBYP977rkndLpXamIEYYA8AKy66qoAausa5/2nLlDdH7XdD+sg6ZhY2wioJAGouzXvOni26BhjjDGmbvGDjjHGGGPqlkJcV++99x6ApOlLs1p69+4NoJK9A7TcjJ0nalrVEvt77LEHgGTHda0D8ccffwBIuk7OOOOMkOlmqGW9C2Yavf7666FTc6J2EiaNzYOasdVNwNLm2uJjvvnma/L3NpW0TL/DDz88dFqPQ8+lY8eOAIBvvvkmdFo2npkaOj7NOqsVuv64ptRdo/ecujnoEtY6QGmdltU1lwdp94+WjVfTP9s1NNaqQ8mqmcR2AVq75fHHHw+Z41Y3S2vhuWjZfq0dxUxLbYuT5U7iWuvRo0foVGb3bp3z888/P+QNN9wQQNKdoG4y3auaey/qddaO48zQ0ey2eeaZJ/U4rXVpp9V8UTeK/tZUGx2Hrl+tCcf7j9lJAPDhhx+GTNedti1RNzKvcRH1kPSYyy23XMg33HBDyMx21LYyOj7WjDrqqKNCp1lx3H/dvdwYY4wxpgnUzKKjga2sbMwKrQBw6aWXhswgOq28qG9frOmS9xOtvhGTrLcsfVNjxWB9O1OLz8UXXwwAuPfee0O38847h8z6NLW0XHEsOk/6ZtWcysxpweIjRowImXUidtppp9DpW2Qe88pz0bd0veZK2hsl34KBSi0erQOiNXWKgFYItcJp5VUNjOca1rnWN3rWZ9Hx6fxVC73ODPw9+uijQ8cK6UDljU+rGWdZd9LmT++14cOHA0gmO6glgpWT1crX2jXJz+s633jjjUNOm5PGyApGZX2gFVZYIXTaoJcyLT9AMnBZrZ7NtR7otddGj7w/tHYR39z/CY/VHMuOzh8tlWoxV+uOBs7mie5/6p3gPaqNPGeeeeaQed9pAkitKv83Bx2f7qscq96rat1iYo6uP12rTBipZrC1LTrGGGOMqVv8oGOMMcaYuqWQFhA03WltATWzjxkzBkCyzow2FWQ7BW1LUC03j5oIWe8HqJg+tZ6KjknNsDQDq+lVg3xpstSaBGr6K8I0SZO6noe6Dr7//vuQ2ewxK4CQ8/vUU0+FTgPraKaliwBIusbydNnpdzfnOOpy6NKlC4CkaVXrsFDfnMDZlpAWoKq1X9Q1o65TrrsTTjghdBoM+PHHHwNIujPyqHmk15/XdNtttw2dutu4P6iJWwNw9f6jy+SVV14JnTadHTduHIDkva7B9lyrGmCeR4B8c9xUzSGtqafeX2y2yKBsADj11FND1mvB+jNNdSPoMbt16xYyA/81XEFrp6kbuzF3Gdeirkl1TTHJ5fbbbw+drnVt9pnnXqPrS914hx12GIBki5G77747ZNaXybuRczVJ21d1fjSxgUHKSyyxROgefvjhkHV+qoUtOsYYY4ypW/ygY4wxxpi6pWauqzTTd1amRPfu3QEkXVuap89OvtWsc5F2HtoOgbVTRo4cGToty65jofuAXboBYMiQISHTjcCOxUCypkutsq30nJkNoS0S1J2o3dVZ/4PuBgD45ZdfQuZ10zoJ6vrhtVx00UVDV5baSFmkuYl0rWjNEmZ9dOjQIfXzeZBWO4StKoCk64HZZgMHDgwd6ygBwNNPPw0gaW5X1xwzKKo5Z7w+u+++e+jUdM/7f/vttw/dWmutFbLei3QJaHdlraPC7sm6PlnbA6i4wcq+JpuDjoWuZR2zhgm0xqWmx9Hu6qwTpO4adWfr+qP7Mq1tB1DZX7VtidYEYlaPZkoyExZIrqs0N3NryOo4v//++4f83HPPAQDOPPPM0DHTFqhcw7K7q7Jghp3+JqTVtNL9MS83bhwz1283xhhjjCmQVll0Wtqo8X//+x+AZOAkAwSBSgOzt956K3S0AgCVipp5NzVbZZVVQmajMn2j1AA3Dfa86667AAD3339/6HSsDAbUoD+tf1LEmyStFIMGDQqdvhGzUSlQqbmiwWRpVVDZfA4ALrnkkpBpqWtLbyxpgbMaDP/++++HzCrFWXVC8iSrgnhas0WdEw2Wp/VEK5tqNdk8A6u1nsghhxwS8lZbbQUguT6ff/75kNX6xOuuday22WabkHnfaVPLlgaptxV0X+b+q9ZpNuoFkkkSzZ1rvXb6xk6L9hdffBG6UaNGhazzypppSy65ZOjUCkNLjiaLaJVlNotWi1VWtedqWXL4Pby2QLLy/3333RcyA+PVepnWFLktoYH7vId1f1HvAK03muyi3pu0BsWtxRYdY4wxxtQtftAxxhhjTN3SbmrmoSlTpjT4z6w6M6zZoaZAldU0xwZfGkw2evTokNdbbz0AydoeWjOjtSat9u3btwOAyZMnN/iirNonrCmi5jg1h6rJkm6olVZaKXTHHXdcyHSJaTn0aprLp5122szxNYaOWd0B55xzTsi33nprg//XwFC6rLROTjVdH60ZX0tR0zkDKrXeAxuVApUgwwMOOCD1uxobP8c3ceLEJo8vre2GBhtffvnlIdPMrK5flbfYYgsAyXYMTBD45zFaQseOHdsBwF9//dXk8XHf+fvvv0On+4c20GU7Fg2G1WaSnMu8XFTTTz99s+cvD9LcVUClfo2W6Fc3lq5r/Rzh/KX9PjR2LjpPN998c8iabELXqroz1A1M14i6VjfYYIOQ2aKlpXVo+PvQkvtP204w3AFIBnuzBUVRdXJaMn+Noc8FY8eOBQD069cvdOuss07IbJasNen0+vB3saXXhPOXOL8WfZMxxhhjTBvADzrGGGOMqVta5bqiiRGoRFVr2f8sWF9Azck77rhjyHR9aNR+Nc3MU3NdKTrW8ePHA0hmgmhbBK2DQDcbXXBAMkOHboK8TOfVcu2oaZW1YYBKBpmatdO6EufV1qEI15VCMzWzA4FkCXe6fDTTpzlm2Ja4rohm77377rshqxuLWWGa1aIZKr169QKQdBekuTBaSktcV0TdMbq+0rJnsjKp8nYT1Mp1pWP+7rvvQmZIAecZSK7Pl156CQCw7LLLhk4zhBpzqbfG9aFzprVTNAOHe41mf+lYuS/r+tT/5/7aWEf7LJrqutLvpEtOwxV4HwHA8OHDQ+bvWlHZfXm4rhReF/5mAsCVV14ZMttBqOu0mmEcdl0ZY4wx5l+FH3SMMcYYU7c023WVZQ5sjmmwsb/lMfIyMTfVdaXQXNqccWYVb8ubPFw7jc112lzlNX9lcV1lXROOu6Vz3hrXlaKu15as27zWbGtcV22BsmRdZZG2lzXHtdca10dTfj+aWjAu79+H5riu2Pbg3nvvDZ0WVNXis0UXBMzbdUV0/0lDr0M1r4ldV8YYY4z5V9HsFhBZb4bNeSIr+om2JdRjWfjmkNfTd1skb4tjtfi3r1mTTpHroim/H2W/r4ieJ2v7aPNQvc55tysqI2Xaf2zRMcYYY0zd4gcdY4wxxtQtUw1GNsYYY4xpy0w1RqesWQOthVHnRWXt5A2zdv7++++6HN90001Xk6yBomhJVmBbgutz0qRJdTm+Dh06/Cvmr97HV/Tvn2YtaY83FoTs0qVL6JrTN4u/f0Xff1kFP1tbnJT3n2LXlTHGGGPqlmZnXRnTGGl1ZtpiVoXSWB2hMmUYGGPaLmn754033hgyO70/8sgjoVPrSNmhpWr06NGhmzhxYsirrbZa9Y9Z9W80xhhjjCkJtujkiD5lpz1xa22Ftl5nQf3JrBLK5q0AMPvss4fcvn17AOW37OicaQPbIUOGAAAmT54cunPPPTdkbaxZRnh+WZVLaZ3KaorZFtE54bgbqwZcVGVz8++G+877778futNOOy3k/fffH0CyqWnZ4Z4PAJ988gkAYIMNNgjdZpttFjKbeuv+1NrfClt0jDHGGFO3+EHHGGOMMXVLuW3sJSQrKFXNbHRDMQ0QAF599dWQJ0yYAABYaaWVQrf44os3+HxZ4bh1zE899VTI++23HwDgu+++C92xxx4b8sEHH5z3KTYbNY3SdPztt9+Gbu+99w75/vvvBwCsueaaodM5K4vrKqup5zvvvAMA+Oijj0KnrpmuXbsCAOaff/7QzTHHHCG3lRYYOv4xY8aE/O677wIAlllmmdDNMMMMIfNadOrUKXTqei37uNNIcwOUfRyN7bV6/mXfMxtD5+f3338HkHRXqZtq++23b/CZMo5fz+/LL78M+eijjwYAzD333KHr3r17vueS67cbY4wxxhSIH3SMMcYYU7eUw8behlAT4U8//RSy1jQYNWoUAOC5554LHSPNAWCxxRYDkDS9MpMHADbeeOMGxyoaNR1PmjQJQDLT6NZbbw351FNPBZB0x911110h//nnnwCAGWecMXRFm9E1w4rzesghh4TuoYceCnmnnXYCAJx44omh69ChQ85n2DR0HD/++GPIV111VQNZzcnTTTddyLPNNhsAYJ555gndVlttFfI+++wDIGlOL1NWEk3mr7/+euj69+8f8tixYwEAiyyySOjUTcX1udxyy4XuoosuCllN7q0Zd9r9neX2bMn9oZ/54IMPQmanbZ3foucvLQNQqwG/+OKLIT/++OMAkpWBd9lll5C5LoveUxojK/Rh5MiRAIB77703dNxTAWC++eYDUPycZcFxaVYq9wwA+OWXXwAAjz32WOhmnXXWBp+v5vzZomOMMcaYuqU0Fh19E+WToOr0Tac5T3p86m3t0yE/f9BBB4XummuuCVnfztZaay0AybeMrbfeusF3/uc//wn55ZdfDnmTTTZp1blWi6z+Iwwm0wDk2267LeSllloKQLKOjlp3tKZCGXnyyScBAHfeeWfo1l133ZBpydIA3db2Z2ktfCNUK87AgQND5lsiAPTr1w8AcNZZZ4WOa2ToYQIAACAASURBVBaovDHrW+Q555zT4Jj77rtvyNNOO23IRbxJ617x8ccfAwAGDBgQum+++SbkWWaZBUAy2FznsmfPngCS10wtCttss03I+tY6NXgvTZkyJXTHHXdcyLz/jz/++NCtvfbaITen8i2PpXvSsGHDQn7mmWcAJK2sCy20UMi1shTomH7++eeQr732WgDAJZdcErqll146ZN6LJ510Uuj0/mOyg17rMqLj12SVk08+GQCw3nrrhW6HHXYImfNbdouOWkE1MYX7qiY75F2zyxYdY4wxxtQtftAxxhhjTN1SGteV1vS4+uqrASTrXGjg6nvvvQegEtT0T9TNteWWWwJI1qxpDVq2Wr9zwQUXDJnnrbU3HnjggZDPPvtsAEDnzp1Dp3VaymKSVNeVuunefPNNAEnXjgZ20mSuZn0NDGUw9vLLL1/lM245avpmszkNENx5551DZrBu0e4qnR9e89NPPz109913X8h0VwEVlwCDGvXzQCVwV2tb6PgZOK+mdQ3crVUQvc7PH3/8ETJN/xpsrcHilHVPUTcVx0IXDwB8/vnnIVfLJcJWKUDFdbXrrruGjvWagOS90pJ1p2uFdYTOOOOM0KmbQV3L1XZD6nkwaQNI7n8MJlbXla41uhzV3bPhhhuGXKYkjjToslI38xFHHBEy91KdEw3WLcvvg5LmOr7ssstCp67TBRZYAEBt909bdIwxxhhTt/hBxxhjjDF1SyGuK5q53njjjdBtuummIdOkp5kQ008/fciseaFR2+uvv37IauZkNkFrzX00uW6++eahU3OdmmRpsjvzzDND98UXX4RM07jWSdCxFOkSUXcAXYQAcMUVV4Q8fPhwAOnuKqDiOtTuu5dffnnIK6ywQuJfoJhMHZ0zll0HKtlk6jrVTLgy1ufg+n722WdDRxMxAFxwwQUhzzvvvACy1xnXwJJLLhk6vae4FvQ71cye1k6hWuj6nDhxYsi6vu644w4ASXeVugZYQl9dW+oa4jmntc0Aki5ZZpu1ZE2oi4jfo+6s77//PmQ9l6aSVaeFstbW0fWf1u6jtfCYuqf07ds3ZM1QpftVaxvpWmWYwwEHHBC6hRdeOOQyunZ0Lrh+zjvvvNB9+OGHITN0Q+/fso9Jax4NHjwYQPZvchGuRVt0jDHGGFO3+EHHGGOMMXVLzVxXauaiGU7L6mtBoSOPPBJAJWMKSJYrZ8GvphSeq3an3qyOuZqJwa7lev5qerzlllsAJIvslaWjt5q41fXEjutA5frrtdDPvf322wCSZb+1e7lelyLRNaluAropevfuHTptkVBG11Uamqmhcto9odeCLTCGDh0aOs0A4rV49NFHQ6fdwddYY42Qq2VyTyu4d/7554esnZ65V+g9pcUNuW/QhQckrwXPWY+lbtrWFLxUt5e6w/557kAyk7OxNafzR9exZqJpuxr+bVrbhbzgNVUX5xJLLBGyuvnTWji89tprIR9zzDEAkmPingNUwhyKvk/1+uq8s+CqZiWpGzitBZB+F69l0dllek6a4clsRS0oq5nQPO/G3LFuAWGMMcYY0wRqZtHRtyO+KWpTSH3i49P7gQceGDqtOcMnwqIDtPSJUwOTDz/8cADJAGqFzehOOOGE0Glg7jrrrAOgmCd2PabWDNKaG3zjuOeee0KngZ/bbrstgGSAub5dc66LfuNS0gI3syweeTSdywNtJKvXv1u3bgCSwaj6xsmml5oswEaQQGWuNQBR5TzgnLz00kuh07YUGgDNwOPddtutweeB9PtK/z+tjpLWadG9qqmJA1wz2hZF68gwGHnVVVcNXVYwKted6rSmGJMFnn/++dCpxUPbddQKnutXX30VOrX86TkxMFqtPw8++GDInGu9/7TBbJH3pe4Teh66V950000AknOtljwGKauXQ+e6T58+AJIthJRajV/X/vXXXx/yjjvuCCDZtkPPn/eV3odqfaTFVK3orcUWHWOMMcbULX7QMcYYY0zdkqvrSs14p5xySshsgaDlzrWmAgMLtcXAXnvtldt5VgMda5ppVdl///0BAC+88ELotJMxzZwM+gVq56bTc55ppplCHjFiRMhsEaDmfHVt0KSq3YVb2n2+CGhapQsHSLoZVlxxRQDl6sJO1+lOO+0UuiuvvDJknb+0YEY1E7POk7p+tGM3u54//fTToWtJnZfGUNfRuHHjACRbBWgLh6OOOirkPffcE0BynaWtuawWEuzartdkzjnnnOp3NQY/o+5sbSHDwGQN4Nxjjz1CVjcFXZJPPvlk6P7888+Q2ak9rd4PUNsg5H8ek25TILm/q+vmkUceAZAMANfAd/4+aIByWrBrEWgIw3PPPRey1nHi/ae1c7SOkM4l0b2G7YQ0mUevX577a1qICZCsj8T7J2t//OyzzwBUgsqBZIukpZZaCgDw+OOPh07vm5aMzxYdY4wxxtQtftAxxhhjTN1SddeVZlepmUuzpm6//XYASXeVmvxoktOsg7ZEY24mjpX1FABgtdVWC5klzlVXRIaZHlNrXlx77bUAgF69eoWOrgUAuPTSSwEkTf/V6vhcC2hm1wwRtg0AgK233hpA0jWnptW0FgJ5kJYVtt9++4VO21a89dZbIU+aNClxnkDy/mQGoLprNKuK97W6CPIwl+v333DDDQAqZm8AGDhwYMi77757g3Np7JzUhaMZLsxGW3PNNUOn16I196JmJ6rrim4AdVuoOf+uu+4KmXOt+6ue0+yzzw4geX/qtdD6XbWC53zYYYeFTudXMwDZAqJHjx6h098HzpXW9iraHc61pFl12jbnhx9+CJluRM4TkGxnsfLKKwNItgPROjusb6aZduq6qhUPP/xwyHTnA8nWMUTPlS5ZuliBZNYxu9LrnpOVwdxUbNExxhhjTN3SKouOvhExiO7NN98M3aBBg0LWOh5pQUoa2MRgNA3gqif49rHooouGTp/IWb9klVVWqe2JTQV9+3riiScAAF27dg2drgVWftZGifpGVvTbVxr6Rsz6EPoWoW/abCCp6/jkk08OWcdaK3hN9dgazLnYYos1+Nu0z6usutdffz1kWofUytHaYF2i56/Wm5EjRwIA5ptvvtBtt912IetcNBaMyrWqb4xaZfnXX38FkEygUEtMSyw6vCZ6nkxKACoWDd0H1TqgVZx5jTTAePXVVw+Zli4GdQLJZq+8lrW0EnP8uk4Y1P7Pc+E10nWkTUdZE0gt3kWQVu2fDTmBZLCwWkxp0dDK8VoHilx33XUhf/rppyEzCUQbmea9p/Ke0X1Q1+pGG20UMi35eh9qsgAtylyHAHDjjTeGzMbZmozTWmzRMcYYY0zd4gcdY4wxxtQtVXNdMTDp5ZdfDt0cc8wR8lxzzRUyTVpqOlYzZr9+/QAA888/f+iKdnfQXJxVSr45ZmB+lwaIalNJBimmNXKrJWqa1WAyNnC75JJLQqemddaE0ABP/f+yoGtK6wDR5fPll1+GTgM/KbOUOwDssMMOIXfv3h1AMfU8tDaTuivUTD7bbLMByF5THJ+ay7V+DZspqmt58cUXD7la49ax0LWjx1TXaGOtGPRe4t/q+lU3AQNDdf+p1v2n36Oua5ru1fWvLThY9h+ouAY02F9dekTXbN6B8U0l6zqqyzKt6aO6/BjEq+6gItDzYzKGtsrRNamJN3T5a20grUnF+jH6+8i2OkClQbLOf61+HzSpRIOlNTCcc6nBxkxAAipj0dpxGsDM+0/XRGvHZ4uOMcYYY+oWP+gYY4wxpm5pletKzUl0V2jZ/MGDB4ec1t1YS7hvueWWIbM+SRFtA9QcmVbHQ8t6awsL7cTLc9Xv0rGMGTMGADBgwIDQaYl9ZksU3Z1dz19dazST00UDJDOUWBNBS5yrab1oNyTR89CsN9b6OOigg0KnZui0rB29PrVyE6StVZbHB5LmcDWd0+Sf1pEdqNQPOvzww0OnGVA9e/YEkHRnZdV0aSo8PjMygGSncprpe/fuHbqs7La0rDHNFmGGFbtEA8n6QyeeeCKApLsrjzWr14nuxA022CB06q5qrGZR2jXPOuei95U0Gru+aXupukuLQM+Jbhrtoq7utldeeSVkZmNNnDgxdPo53lf6m7jgggs2+NsiXOM6Jq5ZADjnnHNC5nmpG1bdXDfffDMA4MADDwyd7jW876q5Tm3RMcYYY0zdUjWLDq0QfFoDkk3l0p4+9Y1Mn2j5pFzEm7++xal1ioFTGgCmdTB0LDx/fWJnbSCg8sSudRAuvPDCkGkdKeObFwD8+OOPAIAvvviigQ4AHnvsMQBJi1VbYosttgAAfP3116EbOnRoyByrVs7Vt8si5o3rltYWIGkRYbVRAFhooYUAAD///HPotGkpLR5sHglUrglQsRrNOuusoavW26VahtQKyCBItaiqFU6vOedHx3THHXeEzAaE+kbJAE+gUp+miDozjQVVNwedk2WXXTZkWp/VIqk1ecoSuKzoWNiAVS3GRaDntOqqqwKoWOv/SVoVcyXNYpUVjFuEJYfHV8s9LZ9ApVE3AIwePRpAckzaoJa/f7om02oSVRNbdIwxxhhTt/hBxxhjjDF1S9WaetI0p66fNHdU2meaqq8FajbT2i8MgtTmcywVDyTrB40fPx5A0lyuNRP69+8PIBmApSXAy+Ky0nnQa0GXpJZgV5M7XQLagqAsAchZ6DXnXGgLEw3m/fbbbwEAyyyzTOiy3Ch5oteU991uu+0Wuueffz5krVnBwHqOA0i66dhMV9taaNPMTp06AaiuCT2tRYK2QKGeDR+BZM2StMBsXZPdunULediwYQCSwZ5F16zKAx2HltNnEoW6HjTkQF3OdGkVcf/qnOrx6cbUANei3W1cP7qOWks13ZjVQteU3lPcU5oC70/9rtxbWOT67cYYY4wxBeIHHWOMMcbULe2mZjKaOHFiuf0NLaRjx47tAGDy5MlTHZ+aIVnievjw4aHTFg6Mugcq2QBaZ0DNwezg25yOy81h2mmnbQcAf//9dy7zx2uhHYU1W4bjzqsO0nTTTdcOAKZMmZLL+Bqrg0R9S1uANEb79u2btD7T0EwNbeGg9XW4PnV8zMQCKm4u7XhezawIrs9JkyY1GJ8eR033F198MQDgzjvvDF2Wm6BHjx4AgJ133jl0Opa0TtHVXJ8dOnRo8fzlhV5XdkW/7LLLQqcZploThVmmen04f3mPL6umGWu2afdrdjQHgKWXXrrBZ5oDx1fvv39p9189wPtPsUXHGGOMMXWLH3SMMcYYU7fYddVEaEbNiu5PKwiVlTWQVqK+muTtumrqtchrfHm7roqmNa4rRV07zclKSZu/as7l1FxXSlaLitaSd1ZVGV1XaWS5/hq7PrVyXSnqkmUhyAcffDB02q6FhSxbumbtumrb2HVljDHGmH8VVaujU+80x0pR9poxrSVvi42pDm29HkxeFiXz/2hL60MD05n4oY2UWVsH8FoxDbFFxxhjjDF1ix90jDHGGFO3TDUY2RhjjDGmLWOLjjHGGGPqlqkGI9cyfTCtQaGmP+ZRObje0+vKnt7aUv4t6Z/1nj7/119/1eX4pp9++n/F/ef9s21SRHmAWsLxKbboGGOMMaZuKTS9XK04WhDq+++/B5DsX7LuuuuGXM3iYSYbvc5pxcWy0n8d92WMMaYs2KJjjDHGmLqlUIuOWgm04NOFF14IAHjppZdC17Nnz5Bt0akNkyZNCvmXX34JmYXGOnXqFDrtCN2hQwcAySJfbdHKk9WCgNbHLCsWr09bHHM9kbVPpM1fNWMATdNIu6f+qSeNtdApe/FDHdM000zTQJeGjkn3UtN8bNExxhhjTN3iBx1jjDHG1C2Fuq7UdKedaG+99VYAwMknnxw6mvuAcpop09wcjXWPVnN50cG8PL+//vordHQhAsBll10W8vLLLw8gOSfqutprr70AJHvR6LUoo0tHz4+ynqdel1deeQUAMPPMM4du7rnnDnmOOeYAkDTHl3HMbZ3Gupv//fffIU+ePDnkd999F0DS9brccsuFXMa5SnPzNMeFr3tNEftn2p74559/hvzRRx+F/O233wKouMCB5FzNMsssAJL33IwzztjgWEXPY1ZoBkMymHTzz7/lXHXp0iV0K6ywQshF/j7UgjzGZ4uOMcYYY+oWP+gYY4wxpm4pxHVF0+sbb7wRun322SfkzTffHACw1VZbha5oM2Qaam7UDKWvvvoKAPDoo4+G7tVXXw2Zro299947dIssskhu59kc1B210047hbzLLruEPPvsswNIZmLpWHfddVcAwI477hi6448/PuQympY/++yzkEeNGgUA+PHHH0P3wAMPhPzyyy8DqJjQAaBz584hX3TRRQCA1VZbLfW4eYyb7ojmZA9lmaN5fnp9dF1QX0Smkp7Td999F/ITTzwR8osvvggAGDt2bOg0a2X06NEAgDnnnDN0Dz30UMgLLbQQgOJd5DpWXYvXXXcdAOC9994LnZ4r53XWWWcNHd3JALDkkksCyH/+9PzpplEX+LXXXhuyzuVss80GILnmpp122pC57+j/9+jRI+Thw4cDAOaaa67Q1WoudczffPNNyLfddlvIQ4YMAZCc07SaZf379w8d5xyo3J86pjzGl+Ya1vtI/z8tA66xrNW02mz6+Wruk7boGGOMMaZuydWi01jl4wMPPDB03bt3D/mUU04BkHyKL/rtiug4fv3115APOeSQkBlYrW/5EyZMCPn3338HANxxxx2hu+aaa0JmFWgNYMsbzpVe8wUXXDBkHffXX38NoDIOANhtt91C7tOnDwDgv//9b+h0LCeeeCKA5BtZ0QF2w4YNC/nyyy8HkLwWWsfp/PPPBwBMnDgxdAMHDgx5/PjxDb4/j/Hpdx533HEAgKuuuip0ev5p56LB1MrPP/8MAFhvvfVCt+2224Y8zzzzAACWXXbZ1GPlAc///vvvD91JJ50Uslo3FlhgAQCVoHmgYsUAKuPS79puu+1CpqVBP1Or/UfvM64jIGldZTC83p8rr7xyyF988QUA4MYbbwwd5wwAunXrBiAfi46ePy2fAPCf//wHQGVugGSyid5faokhev257z777LOh0/2X61fHnPf80TrBYwPAnnvuGbJeC1pF1lxzzdDpvcS1qFb+n376KWQe45NPPgnd0ksvHbJe4+bOse4TGsD/2GOPAUhaltQiqudCWS02q6yySsj8rdff/Pnnnz9kjqWavw+26BhjjDGmbvGDjjHGGGPqllxdV1l1BE477TQAyWBBraPDYN2yuKuA9GCss88+O+S777475GOOOQZAMphax89xDxo0KHTqxqJJs4jaM1n1fDRYcL/99gMAHHrooaFTE+m8884LoFIPCaiYy4HK+Pr165f6+TxR06zOCc39QMVkevDBB4fusMMOC3mmmWYCAHz88cehO/zww1OPUW30u9V1xgBbDRBvzJ2kf6trjffdvffeGzoN1uX/v/baa6FbbLHFGvx/a9FzeuuttwAk75k01x1QSWZYeOGFQ9exY8eQ6V7p3bt36OhuBYCrr74aAHDEEUeEjgH4/zxutUjbXzRYV9vhbLTRRgCA8847L3Rq+qdLecMNNwyduiHyPP+0fR6ouAvvvPPO0KnrVMedFoyq6571c7beeuvQbbzxxiGntaDJG67VP/74I3TqTtX9jfOndcq4p+jnbrjhhtDRdQRU7tszzjgj9fOtmV91F+ma23nnnRt8t9YuUnn66acHkJwzuluByl6i10TPn2EcvXr1Sj3HlozPFh1jjDHG1C1+0DHGGGNM3VJ111VWppWa4WiSZb0RIBl1TtNlmjm9lqTVNtDaP5rhojVx6ObQ8SvM5tAS59tvv33IrFmz4oorhk7HX0SG0kEHHRQy63No1oCeE03GrEcCVDKVgEq23frrrx86dbPkOT6dU82Eo2sEqMw1s9+AZIsLfu6WW25JPYaWq682em3UHUNzuNbuyMqqSvsurQPF71XX2Kefftrge3X95jFn6nrgXsH2AEByT9lss81Cpkk8qzs57yUtq7/vvvuGfNNNNwFIui5rhbp+9Jrr+uP+oq45vVasQ7PooouG7p133gn5t99+a/Cdrd1fea3bt28fOs3UO/LIIwEk68mou1Bdg1x/ek5p7XLUzUJ3yT8/Vyt4TLYXAZL3FO9PoJLVqb8P6nqly1/DBVZaaaWQhw4dCgBYffXVQ5fm+msJ+ll1hzKrdKmllko9J63ZlIbuJdyjRo4cGTqOCQCeeuopAMmsz9aGA9iiY4wxxpi6xQ86xhhjjKlbqu66yio4xEwGoJIVoaZNha4DjVq/6667QmaGx5Zbbpl63DygufSDDz4InRaH0gwiumGyov45vummmy50mgHDgkvquqoV6i7UTCTNwHnyyScb/G3aWNUMqnPNbDV1s6ibK0/XVZbrp2/fviGzhDwLVwLAiBEjQj7zzDMBJLMm1HW5zjrrNDhW3u5Guhb1PNScn3Z81alpmS6NMWPGhG6ttdYKmV3BP//889Bp8bo8x6ruMm3boecy33zzAcjeE6h//vnnQ6ctTOga4vcA+c9fWtaOdvReZpllQua+kLW/cHw6pw8//HDIdNPqnFYLdTGpO57nr+5AbUuhbuwddtgBQLKFirrhiBYM1P9nwcG850zXF8etRSi1O7u65rhur7jiitDdfPPNDb6XrSKAStYTUHETqZuzWmPV+dPCg+paS6Ox42tWFQs5fvnll6HT1iB5ZODaomOMMcaYuqXqFh19y1eLzA8//BAy61PoG7W+/fNJmPVaAOCFF14ImW9atAwBycC0aqFPqRzXqquuGjotYa3tLGhpUouMBunxuzSYWa9b3tapqaEBctooUd/YGTje2JO3Xj+1XvGJ/plnngmdBlbmSZZFR8vRM+D1kUceCZ2uxaeffrqBTtsRcKxFNL1Ui0dzjq/XgpaC66+/PnQauM06SV27dg1dHm/Pek+zUSwbrgLACSecELK+EdKiNmDAgNBpnR8GG6tFTusg8VhFoKX0NYB4ueWWC5l1crSprFrPaZ3Sz+u6yGqmmCdsx6H1mPSNXvcarrsLLrggdGkNTmn5AZKBsbVqOpuWeKOtOu67776Q1WJD74QGK+vvA+8/vec0saBWTZGzgvlbC8+fzXf/+f1c69WsI2eLjjHGGGPqFj/oGGOMMaZuqZq/J622gQZmqWuK+fFZ5ijWJ9GOreoGYnfTWppgef4a9KYtDhigClTaQWhbB4V1LDSYedNNNw157bXXBlDbYNY0Pvzww5A1iJBujuaYM3Vd0Iz+9ttvh07nslYuH12TahpmmwEtW64tPhhkqSXuNfC3lqXn84DBxnr/6vyx5ox2h85jfer8sGaTujgeeOCBkDWYmCZxnTMNBqVr8qijjgrdHnvsETLHWst6LLx+2oWa+wBQqS0CVIKINcCT7iygUn9F3dBap0xdjrWC97S6SLU7/BJLLBEyg5Q1QUNDBujmYT0aIDnWwYMHA0i6y2s1lz169Ah5xx13DFmTObjv6znrnkE35CWXXBK6r7/+OmSOTxM4ytQuKQ3d3xnG8vjjj4dO908GrjeW7NKs47fq08YYY4wxJcYPOsYYY4ypW6rmumIktUaSawlrliUH0jOktDsr6wto2wHNAGGtEI1Ur1WEvR5HMzk06+P7778HkLwW6oZjhoea444++uiQmdWidRLyJq32hrqW1IzeEvR76TrYZJNNQleEay4rq4Dzqm4EzRCh6b+lGU5lIasTOrtia20adTOwa7R+vlYZIHPOOWfodt9995A1A4euAc3K0tYDNPOr60cpsoWAulC1jop2qmY2oNahWmONNULeYostAFS6QAPJmiisk1T0PZfljqBL57rrrgudhi6wdYC2bdltt91CZsjAgw8+GDodfx73KudPXXOaialZpaxJo9dCu6+ztYNmnelcsp1ErTJVW0paCyWg4oYdO3Zs6LhmgUp4SDXvQ1t0jDHGGFO3VN2io/UcNNhW67AwSOyvv/4KHWvrAJUnbg081AZfbMZXdACWvhloYBnr/KjlSt+O+SamTSO7desWchHBrJw/BqICwLhx40JuSZ0ivSa0cgHA+PHjASQrWxc9l/rGwfP7+OOPQ6dvJ6xemzX/bQUds9b8oPVArTyLL754yLSqFG0R0Ouv1rU55pijwf+r9YaBjyeeeGLoZpxxxpAZbF6ElU7vA62GrFV0GZiq+4Q2YKQlQ4NZ1frM8Rd9zyl6/7Ax5sUXXxw6raPGe1Gbsr700kshs4GoNh3WKsoaxFvtOdZrqsHQOj+sE6d1urRBJyvra7KL7sVtZa/RPVN/61mlWz0eWqWb96I+S7QWW3SMMcYYU7f4QccYY4wxdUvV6+h06tQpdCuvvHLIxx9/fMgM3NWmdWqaY+Cyuqs02IymOzUTFtk2AUia1HleGkys7Q6++uorAMkALHUNFWEy5/lrMJ2aeLWmRWNwLnR+tNw+a2bQxaDHryW6ZlR+/fXXG/ytBk4zMH6fffYJnQYGlj0wmWPVYFatA8XzX2WVVUI3cODAkGmSL5PrQ+fvjTfeAJB0DWiLFgYuayNJXZ90I3Tp0iV0RYxV15HelwyG1zGrG4tuHrpAgEpbDKDisix6/tKaYgLAWWedBQDo1atX6LTBatr9pWNl02ANfdAGvVp/h9ci7zpQOj9pYR46vzwXdVEV/fvWEvSctbUJg5E1wUHvRTf1NMYYY4xpBn7QMcYYY0zdUvWW32qi0xLYWofk1FNPBVDpQgskTXe77LILgHR3FVAxCZbdnKemy1dffTVkjkXLmhcNzaWaKdC7d++Qhw4dGjLrqGRlYnFe1HR8ww03hPzkk08CSGbKlKltAl2Offv2DV3//v1DZoaArumy17RQaBo+/fTTQ6edrueee24AwLBhw0Kn3bPL6JpT1wNbU7AeFVCpHQNU6nTo/qIZWM899xyApOuqaNLawWjWnHa6ZoanuoY1QXzl4gAAIABJREFU66pol1Ua6sb54IMPACTXZ9r+r6iO+5KGPtCdCSTXb63aCOn8cd879thjQ6ftVLh/ajiHuuZakgFbS3j+uqer65DtSrStkN6reaxPW3SMMcYYU7dU/dEwq1GiVvk87LDDAAAXXXRR6LTOCp/Esyqvlt2SQ/TNQRuY8e1k1llnrfk5NYbOnwZLawNL1rfYbrvtQsfaMkClGeRNN90Uuoceeihk1gcp2oqj60jfiHn+Ws1b55Kf09odra0cnQdZwaoMPB4xYkTo1LrGpopaGbqMVoAsuFfomLUmEPcatcipxUer6JYZnV8dC5vRqhVH6wgVEfjfGLq+WGdGreAamKzWDaJjYtNINocGkt4Frahfq2uhHgtaZLROV58+fRp8Rhu1atNnBsuXaR7TEjs0GUA7H9AiPmDAgNTP26JjjDHGGNMM/KBjjDHGmLql6q6rpjT669y5M4BksJlC05WasNqi60rPkwGeAPDFF18AKGdQp56T1tG55557Qt5ss80AVJqTAsmx0jSprhHWzgGKd1k1Bsei7sZPP/00ZAZsa9nyMqJzooGNV199NYCKiwAANt9885DZNFeDHsvuutL9gS5hbVuha5GuyVGjRoVOG4SuttpqDb6zjOj5aZ2SX3/9FUAywFVdk2UZl56HJkFceumlAIBtt902dBos37179wbfpXW+ONfqej700END1gDkPPdgHR/XFFBJbLjxxhtDp6130uqQKbUKoG4Oek50HbIeEpAM02CLjllmmSV0ee8v5btixhhjjDFVwg86xhhjjKlbCknIT8uKaIy24q5SslxX7ORaRhOkovOjnYCZLaCZVprJQJNxWevkEDUts6MzUDFzMzsQSJq4WV9G6yCV3bWj7VbY+kFr42ina85f2cek6Fyy+/ERRxwROm1B8+233wJItvAYPHhwyFy3ZR9/WqYRUKlJo20TdK8pi+tK0XPq2bMngEo9IyCZ9ckMnvfffz90uj8xa1DrtGh3+lqFDOiY1E3DkA39f207w720W7duoRs0aFDIzDoreh6zMqVuu+02AMnaRZrBy9ZQabWh8qLcv7TGGGOMMa3ADzrGGGOMqVvKXUu6jaPmYkaaA8AKK6wAIFmQrWgzZGOouZdm2KyChzRjljGrTNFrrufKDIkXX3xxqp/PygosI2o679SpE4Cka0fbBZTdZdMYPP/ll18+dPfdd99UP6Pz1xbH369fv5DPPfdcAMm2JW0JXn/Nmrv++utDbuxeS8taKvr+1HNhhtmFF1441c9kZTDzu4oek6L759ixYwFUQjSAZLunn376CUAynCNvbNExxhhjTN3SbmpPhZMnTy7PI2MVmXbaadsBwKRJk2o2Pm1KR1kDdKv5FtmhQ4d2QP3P38SJE+tyfB07dmwHAFOmTKna+HStsY6TNq3Umjl5vym2b9++HQD89ddfdTl/008/fc3vP7Uecy7z2l+K2D9ryb9l/2zt+LIShBgYr22ddP2xxYp6BPJYn4otOsYYY4ypW/ygY4wxxpi6ZaquK2OMMcaYtsxUs67q3UdZzRiIMsEYiLznT320aTEeeRUJrJaPuaz8W9ZnvcdY1fv6rPcYnf/97391Ob5pppmmHQBMmDChLsfXqVMnx+gYY4wx5t9Dzero6Ns/5aw6AZTtVisnzPCYOHFi6NhxGKh0TWaXc/1MW6Cp69MYkx9Z91+9k7b/tMXaTmWi7fz6GGOMMcY0k1wtOvoWr0+nrJioFgFWawXSm+q1xSd6HbPW0Sljg8vG0LmcMGECAOC///1v6LQp3cknn9zg82V/O9PzY5XPSZMmhU7nj031yj4mY9oi3Pf//vvv0GnT3bbY4LkxdEwct/7+6e+j95rmY4uOMcYYY+oWP+gYY4wxpm6puutKXRzqmnrkkUdCZoO2UaNGha5v374hb7XVVgCAXr16hU5dB2U33dEMSRcPALz//vshs6lnWwrQ1et/zTXXAAA+++yz0LGtAFAxs2qjtzLOmY7pySefDJkNIN95553QaYPIgQMHAgDmmmuu1O8qY9M9Y8pGmrsYAE499VQAwA033BC6s846K2R1mdcLun+cf/75AICnnnoqdHfddVfIM8wwA4By7S9pYSqNuRizGunmMa6280trjDHGGNNM/KBjjDHGmLqlaq4rmq6YUQUAZ555Zsh33313yMyqUm699daQb7/9dgBJd9dqq60Wspo5ywivxf333x+6E044IWSaJOebb77QlXFManr8888/Q3744YcBADfffHPoaE4Fyp1VpmbRsWPHhnzLLbeETNfcuuuuG7pXXnkl5D333BMAMO+884Zu9913D3nFFVcEkKwWXfY6GDrXrXWpZpmky0ya6T2PezKtnhiQ/3XS8ZVlTnT8uv/TTXX88ceHTveaLl26AABWXnnl0JVx/2wMHb/umez+raEdH3zwQcirrrpqg8/Uiqx7+7XXXguZIQ26Z/7xxx8NvmOTTTYJ3TrrrBMyu5pnrdOWrF9bdIwxxhhTt1TNosOnNH1z6NOnT8iHHXZYyHwi0wCzYcOGhczALA1mbkvwWqhF6tdff20gzz///LU9sWaiczlu3LiQ27dvDyAZoFv2NyqO5ZdffgndgQceGPISSywR8tChQwFUxgkAs8wyS8jvvvsugORbjH4Xg+iPO+640HXs2DHkIoII9Zga+EjUEss3su+//36q3zX77LOHTtfKdNNNF/KCCy4IIGndqhVZ1pM0i83bb78dMsfXvXv3qp+TvoVrssLMM8+cOHZr4Pgeeuih0GngPJMhikDnYcSIESHvscceIQ8fPhwAsOOOO4bupptuCvnoo48GANx7772hUy9BmYJ0p4beM5qsQutV586dQ6fzV4RFjnuGWsGZNARUrFBApQ6Qrm9Wywcqe8HIkSNDd8ABB4R80EEHAUjuvzqnvFeAplufbdExxhhjTN3iBx1jjDHG1C1Vd12pCXH11VcPWc38bAD5wAMPhE5NuqxTsvbaa4euLAF0TYFmPnVX/PTTTyHTJVD2UuZqLnziiSdC3nLLLQEk53rKlCm1O7FWoO4Kdb3svPPOIatLjuj63HjjjQEkXatvvvlmyGxmOtNMM4Vu0KBBIau5NU8ze1YLks8//xwAcOedd4ZOAwdZU+jHH3+c6vcvsMACIdNFBSQTBw4//HAAwBxzzNGsc/8nTa3Noaib8uuvvw6Za3n06NGh02B0ut7UHJ/m7msOnPOff/45dNtuu23IF198MQBg6aWXDl1L9zweS+dvyJAhIXPfVRdA3q5nnpMGFe+3334ha+IKr0tWI13ed23FRZWFnv9HH30UMn8r1J2ubqwix/3bb7+F/PHHH4esrv1dd90VQDLAWPcKtvPgmgeS64LJLurumnPOOUO+4IILQp5xxhkBNH5NbNExxhhjTN3iBx1jjDHG1C1VT4VQE+9jjz0W8i677BIyzbfsAg0AW2yxRcgDBgwAkJ01UfYMH5qce/ToETo13bH+StndcWoO1BYPG2ywAYDyz0Man376acjPP/98yMsss0zIdB9oJpK6objGNStA3bQsUa/mWK5pIGmSrfY11Ptv8uTJIaub6pRTTgGQvBazzTZbyKwf1K1bt9CtueaaIdOMrPevfl47TTPDoqnmdnXrqcxMDs1efPrpp0N+9NFHQ6ZLSq+F1vHgOWntp2222abB9+qx1PXWEtcB9y+955999tmQ6TrVTKmWHpPH0KwYzQDlutxrr72a/J2thRk2hxxySOgOPfTQkPfff/+QOe+ataMZjMzQ0XXWFvci/U3TPYEhD5tvvnnoNJOxyN8NPQ91V+le2L9/fwDJPVHnh67xl156KXS///57yKzZpqER6vqnuwpoxr7SpL8yxhhjjGmD+EHHGGOMMXVLrt3Lx4wZE/J3330Xclr3VXVzsUXCvvvuG7q99947ZJrJy26unHvuuUNWM39rMziKgJk6QOPXnSbZMmZFLLfcciGrafWqq64K+cMPPwSQdC3o+mNWlRbBe+utt0J+7733ACTnX83suXTn/f/vO50nuqiAZFYRS6xrwc5NN900ZBaybM46zRpTc8eqWUnqbmHBuC+//DJ0momxyCKLhMysj969e4dOM4zoElLTuLoRdtppJwDARRddFLrTTjst5JaU3k/LSmWXaqCSCaWF43r27BmyuivSMs/S3PzqbtWsK2Y7rbTSSqFTuTX7qp4H3Y0AcO655wJIuji0IGDaXGgm4Pjx40NmIceyZ602hq4jFkkEKlllmnVV9Fi5/vQ+0y7yev5c1xquoWuB+5Jmmmk7CO6vDJEAkm6ylqxPW3SMMcYYU7dU3aKjT6mstwIAiy66aMgMbL3nnntCp4GRrP+gjTAZwAQAl19+OQCga9euqcctC1o7R+sg8I26jBaPLLQmEMlqBJk2F0U0uNRz4luABq1qozwNtr7xxhsBJMd8zDHHhMy3F84jULECAZUg1r59+6aeS7XQ73zxxRcBJC1Pn3zySci6/liHYqONNgqdBhlyXdYy6JFj0aBDDVakdZc1jID0AE6g6W+/GqytFgV+lwbDthZeSw2k1ABcltbXVgds3ggk23HQ6qN1gnT/5Juy1iHT+iesaaZv4WrpVEtec/co/Xu1Yh5xxBEAkq0ett5665DXWmutkHn99Z7SsbDOldbuSru/yrq/cn3q+Ws7BFqatR5V0WPh8dUip8k2t912W8j33XcfgKQVTu/JV199FQBw8MEHh27w4MEh0xKpvyOt9d7YomOMMcaYusUPOsYYY4ypW6ruulJzN+vFAMlO3fwbNWNqnYuXX34ZQDIAkAHKQKXmxIMPPhg6rYlRtJmPZtRRo0al/r8GJrcV1E2QZiZW0ytN8urauf/++0NefPHFAeTjGlETqZqGr7zySgDAiSeeGDrWiwGSpnO6Eb755pvQaTAcZf3+lVdeOWTWhNA6LzTXAsB6663XxNFMHZ0HunbVXaX3gbo56AZaeOGFQ6d1ZHbffXcAyQDeWrmx9JyXWmqpkLlXZJ1T2j2vpm8NhmSwuLp71I3CdaGl5vX6aZBva/Yadecy8PTss88O3VFHHRWyuvnZzkLnX/faFVdcEUDSDaZB3uyErTV72DEaSF731oQE6PwwcLVXr16h0/1b3ZTsSq6dsrWmEGte6frVxBWGTKjrrIw1yzTYXWs2bb/99gCSYy5L4o1eR22bo+tk6NChAJJtcfQ+2W677QAkwwHybidki44xxhhj6hY/6BhjjDGmbmk3NdPr5MmTc/EBNbUOhNbeYRdkALj77rsBVLqcAskS9Y2Z+aaddtp2ADBlypRcx6emPXWD0DSblulSDdq3b98OaP386Zxod2HWN9Au35r1wgwJzeTQmgl0c7XUtMz50/HxXHXuL7zwwpAvu+wyAMk5oYsGSJr+2fph3LhxodNsmbSsHP08a71suOGGoTv11FNDVjcR0flvyfqku+yFF14I3TvvvBOyuoaZgaP3j3b3ZrbYySefHDptkcFr3NLaHlyfEydObDB/b7/9dvydXjOev9bGURewXj9mDWl3ZZW7dOkCADj22GNDp/PHWkSaVaKu82WXXTbktL2mY8eOTbr/1PXEDBU9prrLdK9gzSNtm7PCCiuEzDABnXNdq7zXdH2q60oz99LuS67PSZMmNXt/ycrU1Ewvunx32GGH0GlWD685QxyAirsEqOxLvOcBYPbZZw+5sb22Q4cO7QDgf//7X9U2ZR0r50Vrx6jLmWtN6+hUM6t4mmmmaQcAEyZMqNr4dP7effddAMnQFN3/99xzTwCVTDwgGRrQ2t/CTp06NdiYbNExxhhjTN1S9WDkppD2xJb2Rqa1P+abb76QyxKYpeibCitbqkVqscUWC5mB02Ws/aNkBfYysFbfOPWNn4GxWptGa0LQYqIWnWqdq9ZGYT0HANhnn30AAAMHDgydvmXp+uvUqROASnPPf5IWjK1vNLQEsR4UkKyDosdqTcVT/R4G6fbr1y90+saub+aUtVHipZdeGvJ5550HIBlMrm/HGkRbLTgWtZbccMMNIfNa6pg1gFMD/ydNmgQgGWCuzSRZsVoDIHVPYbCuWknyuFd17hl4y70DSDZVPf3000Nef/31AaTXtgIqY9H506arvIZaeVnrS2kDZl6jalmc9Xuy9nHWt1Lr1iqrrBIy77811lgjdAxwBSoWqQMPPDB0WjNIr1utEld0rllFXa2XWtOKQdZl/J1rChyrNurUPeP6668HkPx9GDFiRMi5VI6v+jcaY4wxxpQEP+gYY4wxpm7J1XXVmFk+qyw+g7WuuOKK0A0bNixkBg6qO6jo2jk6Vprs1LWjZti2go5JAx/ZNJHuICDpBqDrQIMFtXS4/m214PyrWXrAgAEhs86Nnoe6I3T9NLaWaFJWc6wGe95xxx0AKi4wPT6QdHNVq4Eizznr3NMCPzUAVwMD33jjDQDJ2i1ap4RjycO0ruev60TrcBFtGshGj805RlYAPK+PzpMGdufRAHP06NENdBdffHHI6maiG7kxd1pja4H1doBkzZ0012oe+6uOVV3OdENqsoMen+PWz2toA9236o6luxzILwnkn2S5/m+//XYAyT1jt912C5l7VFt1XbEZ67fffhs6vWe4v2gCRN7YomOMMcaYusUPOsYYY4ypW6ruulJznZZdVzMc/0ZNW2omZ2lwLQuuGQQ0Sc4555yp318E6hpgho2OT90cRbvZmoqa9tm2AajUf9DaFjS9A5XS5WpaP+uss0JmtlUeZdn1O9NqFy2yyCKhy+p4zbnMyo7iMbQ2hNYZohtgs802C51mcOU5/zr+rDol/BttC6BZVTQ9q7uImUpA7dZvc9yJ1YRzTRfsP+XWZMopumf1798fQLLjs95zWW7WlsDz16wlnf8sN0+e6LXg/qktHnT9pn3mgQceCJmdsLUOj9bBKqIdBDPJgMpexOw+IFkHroztKtLQ+0BrNjHbVVvdqGuZvxVZWa15YIuOMcYYY+qWqll0+MStbySbb755yFrzIk3Haq1AJQhR34j322+/kPkmUqYnXz2XBRdcMPEvkGzwuOOOOwIov5VHz0nPlc3Yzj333NCp9YqWgCOPPDJ0+vaYx1j5nRpsrHVU+Hanb36s8Awkx8eKzlrThdVygUqdoMcffzx02mySNT30+mgjyDybmWptCq18rGN9/fXXASSb6unnWL9F69hosGeZ7rs84LXUBII8xqz3AWuGnXHGGQ3Oo9rH53FZIRoAunbtGrLW8skTHb9ajlg5XIOJ+/TpEzItsWzU+0950KBBAJL7j+4Lee61ankaM2ZMyJrMwd/IIUOGhC6ryneZ0fWpdayeeOIJAJV6XEB6g8/mJBC0Flt0jDHGGFO3+EHHGGOMMXVL1YORtfaFymoap3lvoYUWCp3WPGE5bG0boO0CGIRWJhOfnouaSYkGZnP8ZTr/xkg7V629ctRRRzX4Ww0WrJW7Q4+j7jIGTl977bWhYylyIFmzg3+z7bbbhm7RRRcNmW6oq6++OnTzzDNPyHSDabBzrcavQcPqRt5yyy1DZosKdVccf/zxIdO1qq6NtrRWWwuDuNX1rHVQ8oDXt1qBzlODa1H3VA3SL2Ku9f5gyIK6i7WOE908es5s9AxUmgrr708R7la9/9V1zXYcGpqhiQNFJ9a0BK0JxN8/dR3quuYepaEFeWOLjjHGGGPqFj/oGGOMMaZuaTc1M+XkyZObbcNUE5XWY1CZf6OR5mpGpT6t43I1mHbaadsBwJQpU3Kx0XJ8WsJdzZhsnZBX9/L27du3A1o2f20Bzl9j49O1mNbdXN2JuhY5L5pBkdZpPa0Fg8otdQG0ZH3yWGoC17H+8MMPIf/5558AknWoNGuM487L3M/1OXHixFKuT87rd999Fzp1XWnribRr1LFjxzZx/+n61ZpKWnOGbghdy1yfkyZNqtr49Pu5hvXaqmuEf6uZWmluKv18c1yCHTp0aAcA//vf/1o1Ph2T/v5xf9HrnNUOKQ+mmWaadgAwYcKEVo0vre0RAOyxxx4AknXwdPxXXXUVAGCrrbYKXTV/Czt16tRgsm3RMcYYY0zd4gcdY4wxxtQtVXddKWqOSzMdZpV4zzvqP2/XFckyR+adAWDXVTZp7qymkGYGz2ud5rE+09ZiUS0Wyu66Ilnro7Fr1VZcV0qWGzaNPFxXjdHYvVrN9Vst15WSdv8VVXizWq4rRcfH4r9aJFeLULJ4Yl4Fc+26MsYYY8y/iqrX0VHqvVR8Y/zbx19GqmHFaIs1ZbwWm09bnOeWUvaxlv38GqPe7z8dH2sGaQsotcgx8LiWc2qLjjHGGGPqFj/oGGOMMaZumWowsjHGGGNMW8YWHWOMMcbULVMNRq5l+mAtYfpg2dNbWwrTW/NOny+KtpKe3FL+LfNX7/tLvY+v3tenx9c24fgUW3SMMcYYU7fkml5uDEkr1FdUwTrz/9A5aaygmefHmPzQ+6/eU9GLwBYdY4wxxtQtuVp09ClVO0H/3//9HwA/ubZ1ssqyc96zysr/8ccfACqdkYFk9/B6sR4U1QKkqfA+BICffvopZF7/WWedNXRarr0s52/qB73n9beC6FptTuuWMqP3EfdEAJh++ukBpF8H0zJs0THGGGNM3eIHHWOMMcbULVV3Xam5/scffwz51VdfDXnVVVcFAHTu3Dl0zTGH08xZ9mBWNbGq6T8N9v8AyjkWhXOscz1p0qSQv/vuOwDJcXzzzTchH3744QCAG264IXRdunQJuezjbwzO+/jx40On12qeeeYJuVZuIF2LPOY111wTuv322y9kuhQHDx4cOpVpWi+7C0uvucppwfA6lrKPi2QFk1POGlMZ7y9103z99dcN/l/vmTKef3PgvP3888+h69OnT8gXXXQRAGDttdcOnbru2gotdTHmMb+26BhjjDGmbvGDjjHGGGPqlqq7rtREOmzYsJBPPvnkkHv27AkAWHDBBUM3wwwzhEwz5oQJE0LXq1evkBdeeGEASdfX7LPPHjJN60AxZk6ajj/77LPQ6fh///13AMAss8wSuuOPPz5kXpcymWjVtEyX5JAhQ0L3wQcfhDxu3DgASXfd999/H/LkyZMBVK7DP8nK1mor0I03YMCA0B1xxBEhzz///CHXyk2i15Hrs2vXrqE7+OCDQ+Z9e8YZZ4ROz/Ooo44CAEw33XSp/18Euma4b3z00Uehe+CBB0IeO3YsAGCxxRYL3cYbbxxy9+7dARSz9tLqTQHJ68v5+/XXX0M3atSokF9//XUASXdIt27dQu7QoUOD7ywCHd/EiRND3nPPPQEA++yzT+g23XTTkNuiG0fhupp55plDt/LKK4d8xx13AABWW2210OleWsY9MW3PnjJlSpM/r78vadlmrV2rtugYY4wxpm6pmkWHT2GvvfZa6PSNsGPHjiE/99xzAJJPafpEyDdFfXK9+eabQ+Ybjb6FnX322SF36tQp5Fo9/aYFYe+www6hU4sG37T0LUytIxdffDGA4p/idUyffvppyAMHDgQAvP3226Hr0aNHyGuuuSYA4L777gvdJptsEvJmm20GAHjkkUdCt8wyy1TrtFtEWs2brLfrNPT/WRNDA6z17UwDz4uAa2m99dYL3VprrRUy6+dccMEFoWOAJAD88MMPDXS1WqtZ9VbUurHXXnsBAJ5++unQ6fzOMcccAIBnn302dNdff33II0eOBAAsu+yyocvb+sH1o2tDrRxq8ebf6PU/99xzQ+Zc6PztvffeIR977LEAkrWrirDu6D3z7rvvNjiXFVdcMXSNram2aAWmZQ1IWnRefPFFAOUfh15zXbdnnnkmgKRHR9fX33//DQCYaaaZQtevX7+Q+bu+/vrrh06vVUvWqi06xhhjjKlb/KBjjDHGmLqlaq4rmpOeeeaZ0KlptH///iGvtNJKAJLBgBpglhYsp4HJ/FsNOlbTbhEmPz0mTXZq2nv++edDZrAxXXgAcPXVV4dclhLneh6jR48OmS4rmvgBYLnllgv5wQcfBAB8/PHHoVPTOk2WWidCr1+e49fvZlA0UHHHAJW1pnWgtOYFgwj/+uuv0OnfbrTRRgCAU045JXS6PosOAiV6HnqvsmbO3HPPHToNVr7xxhsBVO5jANhtt91yO09FXVC6J5x44okhP/XUUwCSAax77LFHyAwG/+qrr0K39dZbh8zxqTs5jzWZ5m5RF/wvv/wSsq6lW2+9FUDSNaXBxkxs0ASICy+8MGReQ02AKML1o+tPa2odeuihAIAFFlggdHqv8vz1ntPED7qByu760d+8d955J+TGaq6VBb0XP//885C5hnV+F1988ZC33357AMkxa2gK77+DDjoodEcffXTImgTR1Dm2RccYY4wxdYsfdIwxxhhTt7TKRqbmzj///BMAcOmll4ZuttlmC1mj/plfr20DPvnkk5CZlaN1ctLaPahOzYBFZFqp64NuKJrggGTtFGZg6bXaddddQ6bpsiwuDgBYaqmlQqZJXds6aFT8WWedBQC49tprQ6euG85Vljsnj/lLy2rheQLppn1mBwDJmhA0ner6XX755UOm62q++eYLXdGZVgqzlbLcFcyQ3GmnnUKnGUq33HILAOCyyy4LnWZAqsur2jVP1Kyv62v48OEhn3POOQCAXXbZJXRq7ub8aiao/j/3srzR68+sRmZcAslMHF2LzCbT7vLq+lliiSUAJN2t+++/f8j33nsvAOCYY44JXa3cJbpnqmubtY2ASgan7glpdVY4z0Ay03aVVVYBUH7XlZ7f+++/H/IiiyxSxOk0G723uaaASrag7gm6ruedd14AyT1VQzeYFah7smbtrrHGGiE3tVaPLTrGGGOMqVv8oGOMMcaYuqVqritGULNYGpA0B7OIFwCMGTOmwec1qp6F5NjlGkhm9aS5doo2U7700ksh/+c//wFQMaECyXNld2/9zIEHHhhyWbKu9JqqG4amQ3W3qbuCUfWpPT22AAAMoUlEQVRZJtg012Pe8JqOGDEidNddd13ILDsPJN2MhB29gUq23F133RW6I488MmS6bIsuVZ9WBBEAvvjiCwDJInvqJua1UneBFs9jiXrNWtKO09ppOk9+++23kLWcfu/evQFku0bZekQztX766aeQF1poIQD534dprit1UalrTV3jzOBkdhIALLrooiHTTbrtttuG7uWXXw6Z2ZLaIkOztvK8L3VNMjsOSLrGmZWpLkRdq7yvNNNKi8vRnZHWSqBM6P6gru2if8umRla4xuWXXx4yfwtYWBZI/n5w3evn77///pCZTbnddtuFTrO2WrKv2qJjjDHGmLqlVRYdfbrj07U+bal1R5/eGGymAWgaVMQGfGrx0BLRbLq3zTbbhE4bZBbxRPzmm2+GzCBBtQLoG+V7770HAPjyyy9Dx0Z8QKVdQNEWAb2OLJsPACeccAKAZLAZrVRAxeKjAY5FB1bzjemJJ54InbYo0bWUZtFQ6wED79SiteGGGzY4lr6FatPZWr0xq5VF67DwvtI6LTq/aZYMtXhwLg844IDQ8Z4Eardu0xrNApWaG2ql0HuRQbC333576JZccsmQaQnRa1mt9avXVq/To48+CiC5D+rxhw4dGjLrcDFpA0hP1tAEgXXWWSdk1vlSi4peqzxr6qjlnvsgkGy6yn1RrVvffvttyLxGtEwCyQDVslty0iwa9HIAwOqrrw4g2yJbFrKsUFtssQUAYNVVVw2djoUNoHfffffQqcWR9dXYSgJIBt67BYQxxhhjjOAHHWOMMcbULa1yXamZlQHEGkCm5tIrrrgi5LnmmgtAsqO3mvkfe+wxAMk6GVongiZNNfdpzn0R5czVTMeu5dqiQk2vDGLVjt3qOilTzRWic03XR5aLgiZH7WiursUiXHI0nWorCg2WTwts1zo5gwYNCpn1SbTjtbZQ4P/rmjzttNNCrtX8arD0lVdeGTKDXHVOtOYVXc56fVibBQA22GADAJWgc6B2bkpdO1pb48MPPwyZbmCayIGk643ByOqu0ppX7Dqft7tV9ya6GdXtou10NNiaweA8TyC9jpjugzqX1GuygP5tnuPWMdMFByTDHG677TYAyTWprkfuRRqsTXcPULmGRbv+s+BepK47/d3s2bMngKTrsan1YvImK5xB3aBcqzpn6oZkTR12aQeSLVpOOukkAJXnBKD1c2mLjjHGGGPqFj/oGGOMMaZuabbrSk2rmjXErCOWdwaAU089NeSFF144ZJpG1fSqMmvmbLXVVqHTcvMsl61l6TWDglHfQL61MNSMp246trvQjuVqJqYZT2uvtDaqPA80Ul5dAxdddBGAZNaSmsZZP+H0008Pna4FujmKyI5Tc6uSlrWiridmxQCV7uxqWtc543XT2kJFjLV///4hP/TQQyHTNazzq1lTO++8M4Cku0rN0HTTqTm5VmtWj7n00kuHzEwioJLNovuTmsZZop7ZgwDQtWvXkGs1Ft2bmJWnLgo9D+3ETpdbVtsbfq+6XtUNxjpH2rakVutTfz+0LRAzbYCKS0PPT2sinXfeeQCA8ePHh07/tqwuK8L5eeGFF0KX1p29jOg60bnUOjlE1/JNN93UQN5nn31Cp+08+FtSzXks7xU1xhhjjGkl7ab2JD9p0qQG/5lVzZiBw1pBlfVggJa9MWiAp76dsemXNhLTwF+tScM3JX076tChQzsAmDhxYi6vMTyWBp2OGjUqZNYEGj16dOj07bm1b5QdO3ZsBwBTpkxp9vh0fvWNcMsttwyZb1JqsdJgM1aE1mBBHSuf2Fv6Ftm+ffuqz58G0959990AKpYNALj11ltD7tu3L4DsNw7On94fei0aY2rzl2WhTLuWWXVmWIX0kksuCZ3WgaL1QN+y1CLZ2jctzl/a/tIcst4ued+xng6QrEMzePBgAMmmltW0/HJ/aWx8ekzWoTrqqKNC16tXr5C33nrrkLkvZu0TvBZqMVDrOC06r7zySujUitDYfcnxtWR/UXT8aTWL9J7Umk+sH3TYYYeFTmt6tXb/5Pps7fiy4PxogoPO1T333AMgaX2uppUxj/Gl3T+arHHIIYeEzN+6++67L3RzzjlnyNXaXxRbdIwxxhhTt/hBxxhjjDF1S7ODkdWsqQGKGlhGWmtuU9O/BnbS5Pf444+HTk2bRdehoWlSzbFnnXVWyGlNL4tu5JbW9kDdba+99lrIDNLV2h5aB4Pzvu6664ZOa0KUBR2rmvEZTH3EEUeErk+fPiE3ZlrlvGuAdrXmV9e2rq+0AEa9/9QMPmDAAADJsvlqWqZJWdtesLYJUAmcLzpoPqudAoMdtWmruobZukNdI3pda9VUV9cE9zetHZb1t2nXPa12GIN2geRcbr755i074SqjY0q7p3ScWoeMySia4FJ2dK8ZN24cgEo9JCDZgJVunKJ/xxpD15zKrKlHFzEAdOrUKWQmFmmdnLzHaouOMcYYY+oWP+gYY4wxpm5pVQsIpTFzaktqA+jn1bRFlw+ze4BKJgmQbC2h5rFao92j1fXDrAGtM6Bm9CJRc7LWKdJsOmaA6JxonZYZZ5wRQLJOkI6vyDoXug617QjdOUDFpaPunJa4nqrpjuR5az2izp07h5zmjsk6PvXaAmK99dYLmXV23nrrrdBpp2itH1Qkuj+wNg4AjBw5EkAyE1Dr0Mw///wAkuuwVu6qLFriBsxyHXD+NJNHO50PGTIEQNKdUrQbsjG0Zg5rrnEegeJd/42ha40ZnN98803o2FYFKP9YiO6lH330Ucj77rsvgGTWrtbBY82jWv4O2KJjjDHGmLolVzOCBhPrG6EGJhENlvv/2jtbnViCIAqfJ+AtEEgIkmAQBIFBoXAgAMMDIJAoEhQIHAmGEAwJBkJCgiXhDUh4BdBXVe3Z3Fl+Zrd3Zme/T1Wan0xv9/R2V1XXCdE9/72bm5u0Q0DMa3/4icWT1JrYHcdJKZKyJOnr6yvttbU1SWWSVesSu3OvV3R1dZW2C/A9PDxIks7Pz7MtTpFSL4nXq2U2fWKME6970bxmiT9f1FwZJPrZBDE/fJ6HOJ7USzx1L4/X7vFk/ThpRe0WqX9+xv/Y3NzMNh//pudq4F4M9z5FYrnX5vBkz/g7H9OmPTp1GFS5fGdnR1K/l8svi0RtMV+f24h7jN27HDW9vFpyW+ak43PK150Qm/XLKIuLi2k3vdZ8xyBlBK/CHp6qw8PDbKtKgB/nmOHRAQAAgM7CRgcAAAA6y8hDV+5O9WSrlZWVtEOMb2ZmJtu8jkmI8g1KRo4kZC/hHsmYUr9Lswk3YCRZPT09Zdvc3FzaXl+mbfhn7qGbx8fHtKN+UYivStLFxUXa6+vrktqV7Bjz8vX1Ndv8maPsutRLpm6TOGC4eefn57PN66SE0OjHx0e2Rb0RqT8pf2FhQVK/xIr3Nd6vpaWlymdpeiwDfw6vmfP5+SmpF8KRqiVWJjFcJfWe28fMy+3H+rm/v59tLsfjYZQ246E3n9eRkN+m9aWKKqFVqZeM+/b2lm3eV//+aguxfnqC8fHxcdq+robE0fb29n9/LzUzVnh0AAAAoLOw0QEAAIDO8mf18r/gLtLLy8u0o5y31zGpUhz3GyZeZyBCKn4r6y+usRLq5e5GjRsQy8vL2Ra1BSTp6OhIUrmy13XUy6vc+BECkPprkgQe2opwj9Rz2ZZyUdZRL4/5cX9/n213d3dpn5ycpB1j2dRNjt+On8/5qptE/v55e8hx/BS68f6P8rMYRr3cZWdOT0/TPjg4SHtjY0NSvzp7KSXoKn6rXl6XmJ8vLy/Z5utj9NV/7jVnhl13RqVe/hO+5lxfX6cdN3x8zR0lpdXL4730cNWg77IS1OlfrBX+Pb67u5u21+SKmnYeZh+nnAXq5QAAADBVsNEBAACAzlI0dOWM0h0Xz1zXnV4idOX9e39/lyTd3t5mm98KizBPqdBIndBVMKisfFWYY1Boo3TIp07oKhjUpzbd2hjV+NVlXONXZ33xcIWHZp6fn9Pe2tqS1FMEl8Y7vqVDVzHGXmT17Ows7ZBrWV1dzTZfn4Yd33GFrpwqdfZSlA5dRV+aWn9+27+q59vb28s2v7XqYeJI02hKfZ3QFQAAAEwVY/PotIkSHh0nTk9++vTdbekTyTAegUlgGI/OJDAt41dnffF3xwVM3WMR71pTEg+lPTqB97lKzLXUmtOER2eclPboNM0wycguJROXiiRpdnY27ZiLTV3mwKMDAAAAUwUbHQAAAOgs34auAAAAACYZPDoAAADQWdjoAAAAQGdhowMAAACdhY0OAAAAdBY2OgAAANBZ2OgAAABAZ/kH6DhjGXYrYM0AAAAASUVORK5CYII=", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "# Randomly select 100 data points to display\n", "rand_indices = np.random.choice(m, 100, replace=False)\n", @@ -157,7 +170,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 11, "metadata": {}, "outputs": [], "source": [ @@ -215,7 +228,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 12, "metadata": {}, "outputs": [], "source": [ @@ -223,7 +236,8 @@ " input_layer_size,\n", " hidden_layer_size,\n", " num_labels,\n", - " X, y, lambda_=0.0):\n", + " X, y,\n", + " lambda_=0.0):\n", " \"\"\"\n", " Implements the neural network cost function and gradient for a two layer neural \n", " network which performs classification. \n", @@ -306,24 +320,67 @@ "\n", " Theta2 = np.reshape(nn_params[(hidden_layer_size * (input_layer_size + 1)):],\n", " (num_labels, (hidden_layer_size + 1)))\n", + " \n", "\n", " # Setup some useful variables\n", " m = y.size\n", " \n", " # You need to return the following variables correctly \n", - " J = 0\n", - " Theta1_grad = np.zeros(Theta1.shape)\n", - " Theta2_grad = np.zeros(Theta2.shape)\n", + " # J = 0\n", + " # Theta1_grad = np.zeros(Theta1.shape)\n", + " # Theta2_grad = np.zeros(Theta2.shape)\n", + "\n", + " # print(\"Input dimensions:\")\n", + " # print(f\"s_inp={input_layer_size}, s_hidd={hidden_layer_size}, s_out={Theta2.shape[0]}\")\n", + " # print(f\"m={m}, Theta1={Theta1.shape}, Theta2={Theta2.shape}\")\n", + " # print(\"--------\")\n", "\n", " # ====================== YOUR CODE HERE ======================\n", "\n", + " #### Forward\n", + " # input layer\n", + " a1 = np.concatenate([np.ones((m, 1)), X], axis=1)\n", + "\n", + " # hidden layer\n", + " z2 = np.tensordot(Theta1, a1, axes=([1], [1])).T\n", + " a2 = utils.sigmoid(z2)\n", + " a2 = np.concatenate([np.ones((m, 1)), a2], axis=1)\n", " \n", + " # output layer\n", + " z3 = np.tensordot(Theta2, a2, axes=([1], [1])).T\n", + " a3 = utils.sigmoid(z3)\n", + "\n", + " # vectorized classifications\n", + " tmp = y.copy()\n", + " y = np.zeros(shape=(m, num_labels))\n", + " y[np.arange(m), tmp] = 1\n", + "\n", + " # calculate cost\n", + " hyp = a3\n", + " J = -1/m * ( y*np.log(hyp) + (1-y)*np.log(1-hyp) ).sum()\n", + "\n", + " # regularize cost\n", + " J += lambda_ / (2*m) * np.power(Theta1[...,1:], 2).sum()\n", + " J += lambda_ / (2*m) * np.power(Theta2[...,1:], 2).sum()\n", " \n", + "\n", + " #### Backward propagation\n", + " delta3 = a3 - y\n", + " delta2 = np.tensordot(Theta2.T[1:, ...], delta3, axes=([1], [1])).T * sigmoidGradient(z2)\n", + "\n", + " Delta2 = np.tensordot(delta3, a2.T, axes=([0], [1]))\n", + " Delta1 = np.tensordot(delta2, a1.T, axes=([0], [1]))\n", + "\n", + " Theta2_grad = 1/m * Delta2\n", + " Theta1_grad = 1/m * Delta1\n", + "\n", + " # Regularization\n", + " Theta2_grad[..., 1:] += lambda_/m * Theta2[..., 1:]\n", + " Theta1_grad[..., 1:] += lambda_/m * Theta1[..., 1:]\n", " # ================================================================\n", " # Unroll gradients\n", " # grad = np.concatenate([Theta1_grad.ravel(order=order), Theta2_grad.ravel(order=order)])\n", " grad = np.concatenate([Theta1_grad.ravel(), Theta2_grad.ravel()])\n", - "\n", " return J, grad" ] }, @@ -351,9 +408,18 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 13, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Cost at parameters (loaded from ex4weights): 0.287629 \n", + "The cost should be about : 0.287629.\n" + ] + } + ], "source": [ "lambda_ = 0\n", "J, _ = nnCostFunction(nn_params, input_layer_size, hidden_layer_size,\n", @@ -371,9 +437,29 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 14, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise neural-network-learning\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Regularized Gradient | 30 / 30 | Nice work!\n", + " Feedforward and Cost Function | 0 / 15 | Your answer is incorrect.\n", + " Regularized Cost Function | 0 / 5 | Your answer is incorrect.\n", + " Sigmoid Gradient | 0 / 40 | Your answer is incorrect.\n", + " Neural Network Gradient (Backpropagation) | 0 / 10 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 30 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader = utils.Grader()\n", "grader[1] = nnCostFunction\n", @@ -407,9 +493,18 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 15, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Cost at parameters (loaded from ex4weights): 0.383770\n", + "This value should be about : 0.383770.\n" + ] + } + ], "source": [ "# Weight regularization parameter (we set this to 1 here).\n", "lambda_ = 1\n", @@ -429,9 +524,29 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 16, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise neural-network-learning\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Regularized Gradient | 30 / 30 | Nice work!\n", + " Feedforward and Cost Function | 15 / 15 | Nice work!\n", + " Regularized Cost Function | 0 / 5 | Your answer is incorrect.\n", + " Sigmoid Gradient | 0 / 40 | Your answer is incorrect.\n", + " Neural Network Gradient (Backpropagation) | 0 / 10 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 45 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[2] = nnCostFunction\n", "grader.grade()" @@ -470,7 +585,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 17, "metadata": {}, "outputs": [], "source": [ @@ -502,11 +617,13 @@ " in `utils.py` file accompanying this assignment.\n", " \"\"\"\n", "\n", - " g = np.zeros(z.shape)\n", + " # g = np.zeros(z.shape)\n", "\n", " # ====================== YOUR CODE HERE ======================\n", "\n", - "\n", + " def sigmoid(z):\n", + " return 1 / (1 + np.exp(-z))\n", + " return sigmoid(z)*(1-sigmoid(z))\n", "\n", " # =============================================================\n", " return g" @@ -521,13 +638,24 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 18, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Sigmoid gradient evaluated at [-100 -0.5 0 0.5 100]:\n", + " \n", + "[3.72007598e-44 2.35003712e-01 2.50000000e-01 2.35003712e-01\n", + " 0.00000000e+00]\n" + ] + } + ], "source": [ - "z = np.array([-1, -0.5, 0, 0.5, 1])\n", + "z = np.array([-100, -0.5, 0, 0.5, 100])\n", "g = sigmoidGradient(z)\n", - "print('Sigmoid gradient evaluated at [-1 -0.5 0 0.5 1]:\\n ')\n", + "print('Sigmoid gradient evaluated at [-100 -0.5 0 0.5 100]:\\n ')\n", "print(g)" ] }, @@ -540,9 +668,29 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 19, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise neural-network-learning\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Regularized Gradient | 30 / 30 | Nice work!\n", + " Feedforward and Cost Function | 15 / 15 | Nice work!\n", + " Regularized Cost Function | 5 / 5 | Nice work!\n", + " Sigmoid Gradient | 0 / 40 | Your answer is incorrect.\n", + " Neural Network Gradient (Backpropagation) | 0 / 10 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 50 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[3] = sigmoidGradient\n", "grader.grade()" @@ -557,7 +705,13 @@ "When training neural networks, it is important to randomly initialize the parameters for symmetry breaking. One effective strategy for random initialization is to randomly select values for $\\Theta^{(l)}$ uniformly in the range $[-\\epsilon_{init}, \\epsilon_{init}]$. You should use $\\epsilon_{init} = 0.12$. This range of values ensures that the parameters are kept small and makes the learning more efficient.\n", "\n", "
\n", - "One effective strategy for choosing $\\epsilon_{init}$ is to base it on the number of units in the network. A good choice of $\\epsilon_{init}$ is $\\epsilon_{init} = \\frac{\\sqrt{6}}{\\sqrt{L_{in} + L_{out}}}$ where $L_{in} = s_l$ and $L_{out} = s_{l+1}$ are the number of units in the layers adjacent to $\\Theta^{l}$.\n", + "One effective strategy for choosing $\\epsilon_{init}$ is to base it on the number of units in the network. A good choice of $\\epsilon_{init}$ is\n", + "\n", + "\n", + "$\\epsilon_{init} = \\frac{\\sqrt{6}}{\\sqrt{L_{in} + L_{out}}}$ where $L_{in} = s_l$ and $L_{out} = s_{l+1}$\n", + "\n", + "\n", + "are the number of units in the layers adjacent to $\\Theta^{l}$.\n", "
\n", "\n", "Your job is to complete the function `randInitializeWeights` to initialize the weights for $\\Theta$. Modify the function by filling in the following code:\n", @@ -571,7 +725,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 20, "metadata": {}, "outputs": [], "source": [ @@ -606,11 +760,11 @@ " \"\"\"\n", "\n", " # You need to return the following variables correctly \n", - " W = np.zeros((L_out, 1 + L_in))\n", + " # W = np.zeros((L_out, 1 + L_in))\n", "\n", " # ====================== YOUR CODE HERE ======================\n", "\n", - "\n", + " W = np.random.rand(L_out, 1 + L_in) * 2 * epsilon_init - epsilon_init\n", "\n", " # ============================================================\n", " return W" @@ -627,9 +781,17 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 21, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Initializing Neural Network Parameters ...\n" + ] + } + ], "source": [ "print('Initializing Neural Network Parameters ...')\n", "\n", @@ -720,18 +882,57 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 22, "metadata": {}, "outputs": [ { - "ename": "NameError", - "evalue": "name 'utils' is not defined", - "output_type": "error", - "traceback": [ - "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", - "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", - "\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mutils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcheckNNGradients\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnnCostFunction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m", - "\u001b[0;31mNameError\u001b[0m: name 'utils' is not defined" + "name": "stdout", + "output_type": "stream", + "text": [ + "[[-9.27825235e-03 -9.27825236e-03]\n", + " [-3.04978709e-06 -3.04978914e-06]\n", + " [-1.75060084e-04 -1.75060082e-04]\n", + " [-9.62660640e-05 -9.62660620e-05]\n", + " [ 8.89911959e-03 8.89911960e-03]\n", + " [ 1.42869450e-05 1.42869443e-05]\n", + " [ 2.33146358e-04 2.33146357e-04]\n", + " [ 1.17982666e-04 1.17982666e-04]\n", + " [-8.36010761e-03 -8.36010762e-03]\n", + " [-2.59383093e-05 -2.59383100e-05]\n", + " [-2.87468729e-04 -2.87468729e-04]\n", + " [-1.37149709e-04 -1.37149706e-04]\n", + " [ 7.62813550e-03 7.62813551e-03]\n", + " [ 3.69883257e-05 3.69883234e-05]\n", + " [ 3.35320351e-04 3.35320347e-04]\n", + " [ 1.53247082e-04 1.53247082e-04]\n", + " [-6.74798369e-03 -6.74798370e-03]\n", + " [-4.68759742e-05 -4.68759769e-05]\n", + " [-3.76215583e-04 -3.76215587e-04]\n", + " [-1.66560294e-04 -1.66560294e-04]\n", + " [ 3.14544970e-01 3.14544970e-01]\n", + " [ 1.64090819e-01 1.64090819e-01]\n", + " [ 1.64567932e-01 1.64567932e-01]\n", + " [ 1.58339334e-01 1.58339334e-01]\n", + " [ 1.51127527e-01 1.51127527e-01]\n", + " [ 1.49568335e-01 1.49568335e-01]\n", + " [ 1.11056588e-01 1.11056588e-01]\n", + " [ 5.75736494e-02 5.75736493e-02]\n", + " [ 5.77867378e-02 5.77867378e-02]\n", + " [ 5.59235296e-02 5.59235296e-02]\n", + " [ 5.36967009e-02 5.36967009e-02]\n", + " [ 5.31542052e-02 5.31542052e-02]\n", + " [ 9.74006970e-02 9.74006970e-02]\n", + " [ 5.04575855e-02 5.04575855e-02]\n", + " [ 5.07530173e-02 5.07530173e-02]\n", + " [ 4.91620841e-02 4.91620841e-02]\n", + " [ 4.71456249e-02 4.71456249e-02]\n", + " [ 4.65597186e-02 4.65597186e-02]]\n", + "The above two columns you get should be very similar.\n", + "(Left-Your Numerical Gradient, Right-Analytical Gradient)\n", + "\n", + "If your backpropagation implementation is correct, then \n", + "the relative difference will be small (less than 1e-9). \n", + "Relative Difference: 2.41759e-11\n" ] } ], @@ -748,9 +949,29 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 23, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise neural-network-learning\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Regularized Gradient | 30 / 30 | Nice work!\n", + " Feedforward and Cost Function | 15 / 15 | Nice work!\n", + " Regularized Cost Function | 5 / 5 | Nice work!\n", + " Sigmoid Gradient | 40 / 40 | Nice work!\n", + " Neural Network Gradient (Backpropagation) | 0 / 10 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 90 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[4] = nnCostFunction\n", "grader.grade()" @@ -790,9 +1011,64 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 24, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "[[-9.27825235e-03 -9.27825236e-03]\n", + " [-1.67679797e-02 -1.67679797e-02]\n", + " [-6.01744725e-02 -6.01744725e-02]\n", + " [-1.73704651e-02 -1.73704651e-02]\n", + " [ 8.89911959e-03 8.89911960e-03]\n", + " [ 3.94334829e-02 3.94334829e-02]\n", + " [-3.19612287e-02 -3.19612287e-02]\n", + " [-5.75658668e-02 -5.75658668e-02]\n", + " [-8.36010761e-03 -8.36010762e-03]\n", + " [ 5.93355565e-02 5.93355565e-02]\n", + " [ 2.49225535e-02 2.49225535e-02]\n", + " [-4.51963845e-02 -4.51963845e-02]\n", + " [ 7.62813550e-03 7.62813551e-03]\n", + " [ 2.47640974e-02 2.47640974e-02]\n", + " [ 5.97717617e-02 5.97717617e-02]\n", + " [ 9.14587966e-03 9.14587966e-03]\n", + " [-6.74798369e-03 -6.74798370e-03]\n", + " [-3.26881426e-02 -3.26881426e-02]\n", + " [ 3.86410548e-02 3.86410548e-02]\n", + " [ 5.46101547e-02 5.46101547e-02]\n", + " [ 3.14544970e-01 3.14544970e-01]\n", + " [ 1.18682669e-01 1.18682669e-01]\n", + " [ 2.03987128e-01 2.03987128e-01]\n", + " [ 1.25698067e-01 1.25698067e-01]\n", + " [ 1.76337550e-01 1.76337550e-01]\n", + " [ 1.32294136e-01 1.32294136e-01]\n", + " [ 1.11056588e-01 1.11056588e-01]\n", + " [ 3.81928711e-05 3.81928696e-05]\n", + " [ 1.17148233e-01 1.17148233e-01]\n", + " [-4.07588280e-03 -4.07588279e-03]\n", + " [ 1.13133142e-01 1.13133142e-01]\n", + " [-4.52964427e-03 -4.52964427e-03]\n", + " [ 9.74006970e-02 9.74006970e-02]\n", + " [ 3.36926556e-02 3.36926556e-02]\n", + " [ 7.54801264e-02 7.54801264e-02]\n", + " [ 1.69677090e-02 1.69677090e-02]\n", + " [ 8.61628953e-02 8.61628953e-02]\n", + " [ 1.50048382e-03 1.50048382e-03]]\n", + "The above two columns you get should be very similar.\n", + "(Left-Your Numerical Gradient, Right-Analytical Gradient)\n", + "\n", + "If your backpropagation implementation is correct, then \n", + "the relative difference will be small (less than 1e-9). \n", + "Relative Difference: 2.3399e-11\n", + "\n", + "\n", + "Cost at (fixed) debugging parameters (w/ lambda = 3.000000): 0.576051 \n", + "(for lambda = 3, this value should be about 0.576051)\n" + ] + } + ], "source": [ "# Check gradients by running checkNNGradients\n", "lambda_ = 3\n", @@ -808,9 +1084,29 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 25, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise neural-network-learning\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Regularized Gradient | 30 / 30 | Nice work!\n", + " Feedforward and Cost Function | 15 / 15 | Nice work!\n", + " Regularized Cost Function | 5 / 5 | Nice work!\n", + " Sigmoid Gradient | 40 / 40 | Nice work!\n", + " Neural Network Gradient (Backpropagation) | 10 / 10 | Nice work!\n", + " --------------------------------\n", + " | 100 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[5] = nnCostFunction\n", "grader.grade()" @@ -828,7 +1124,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 26, "metadata": {}, "outputs": [], "source": [ @@ -874,9 +1170,17 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 27, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Training Set Accuracy: 94.620000\n" + ] + } + ], "source": [ "pred = utils.predict(Theta1, Theta2, X)\n", "print('Training Set Accuracy: %f' % (np.mean(pred == y) * 100))" @@ -899,9 +1203,22 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 28, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "utils.displayData(Theta1[:, 1:])" ] @@ -932,7 +1249,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.6.6" + "version": "3.7.6" } }, "nbformat": 4,