From 8a38978a1fdd373e337440ab6478eeb092000895 Mon Sep 17 00:00:00 2001 From: Arthur FINDELAIR <41486077+ArthurFDLR@users.noreply.github.com> Date: Sun, 25 Jul 2021 16:19:53 -0500 Subject: [PATCH] =?UTF-8?q?=F0=9F=9A=80Add=20Google=20Colab=20link?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- examples/hand_pose_classification.ipynb | 1648 ++++++++++++----------- 1 file changed, 863 insertions(+), 785 deletions(-) diff --git a/examples/hand_pose_classification.ipynb b/examples/hand_pose_classification.ipynb index 8a93757..fb5c4f9 100644 --- a/examples/hand_pose_classification.ipynb +++ b/examples/hand_pose_classification.ipynb @@ -1,817 +1,895 @@ { - "cells": [ - { - "cell_type": "markdown", - "metadata": { - "id": "uMvrR9_jnDwO" - }, - "source": [ - "# 🤙 Pose Classification Kit: Hand pose classification model creation" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "ONjOBMTanXiN" - }, - "source": [ - "This Notebook can be used to create Neural Network classifiers running in the [Pose Classification Kit](https://github.com/ArthurFDLR/pose-classification-kit).\n", - "\n", - "First, we have to import several libraries to create and train a new model." - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": { + "nbformat": 4, + "nbformat_minor": 0, + "metadata": { + "accelerator": "GPU", "colab": { - "base_uri": "https://localhost:8080/" + "name": "OpenHand-Models.ipynb", + "provenance": [], + "toc_visible": true, + "include_colab_link": true }, - "id": "0Xvt8oHFae-I", - "outputId": "e4a71681-b3a0-4240-cdb7-0ac3710d36ae" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Available GPU:\n", - "GPU 0: GeForce RTX 2060 with Max-Q Design (UUID: GPU-63f90f3a-1a62-5290-ce48-ec73cce8a7ae)\n", - "\n", - "TensorFlow use GPU at: /device:GPU:0\n" - ] + "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": "3.7.9" } - ], - "source": [ - "import numpy as np\n", - "import matplotlib.pyplot as plt\n", - "import matplotlib.patches as mpatches\n", - "from matplotlib.lines import Line2D\n", - "import os\n", - "\n", - "try:\n", - " import google.colab\n", - " IN_COLAB = True\n", - "except:\n", - " IN_COLAB = False\n", - "\n", - "if IN_COLAB:\n", - " %tensorflow_version 2.x\n", - " !pip install pose-classification-kit\n", - " \n", - "import tensorflow\n", - "from tensorflow import keras\n", - "from pose_classification_kit.datasets import handDataset\n", - "\n", - "print('Available GPU:')\n", - "!nvidia-smi -L\n", - "print('\\nTensorFlow use GPU at: {}'.format(tensorflow.test.gpu_device_name()))" - ] }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Import dataset" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ + "cells": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Dataset loaded from https://raw.githubusercontent.com/ArthurFDLR/pose-classification-kit/master/pose_classification_kit/datasets/HandPose_Dataset.csv\n" - ] + "cell_type": "markdown", + "metadata": { + "id": "view-in-github", + "colab_type": "text" + }, + "source": [ + "\"Open" + ] }, { - "data": { - "text/plain": [ - "((4347, 42), (4347, 27))" + "cell_type": "markdown", + "metadata": { + "id": "uMvrR9_jnDwO" + }, + "source": [ + "# 🤙 Pose Classification Kit: Hand pose classification model creation" ] - }, - "execution_count": 8, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "dataset = handDataset(testSplit=.2, shuffle=True, handID=1)\n", - "x_train = dataset['x_train']\n", - "y_train = dataset['y_train_onehot']\n", - "labels = dataset['labels']\n", - "\n", - "x_train.shape, y_train.shape" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "xRWuVIAHae-P" - }, - "source": [ - "## Models exploration\n", - "\n", - "This section is optional. The following blocks can be used to compare different model architecture and training processes." - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": { - "id": "93-aeO7Sae-P" - }, - "outputs": [], - "source": [ - "model_train_history = {}\n", - "input_dim = x_train.shape[1]\n", - "validation_split = 0.20\n", - "epochs = 15" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" }, - "id": "PZ2p8pTDae-P", - "outputId": "1c58cf17-9f5d-486f-bfe2-6b2f8935e109" - }, - "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model: \"ANN-3x16\"\n", - "_________________________________________________________________\n", - "Layer (type) Output Shape Param # \n", - "=================================================================\n", - "dense_7 (Dense) (None, 16) 688 \n", - "_________________________________________________________________\n", - "dense_8 (Dense) (None, 16) 272 \n", - "_________________________________________________________________\n", - "dense_9 (Dense) (None, 16) 272 \n", - "_________________________________________________________________\n", - "dense_10 (Dense) (None, 27) 459 \n", - "=================================================================\n", - "Total params: 1,691\n", - "Trainable params: 1,691\n", - "Non-trainable params: 0\n", - "_________________________________________________________________\n", - "Epoch 1/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 1.9592 - accuracy: 0.4567 - val_loss: 0.7188 - val_accuracy: 0.8425\n", - "Epoch 2/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.4739 - accuracy: 0.8712 - val_loss: 0.3257 - val_accuracy: 0.9276\n", - "Epoch 3/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.2515 - accuracy: 0.9304 - val_loss: 0.2351 - val_accuracy: 0.9207\n", - "Epoch 4/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.1581 - accuracy: 0.9589 - val_loss: 0.1261 - val_accuracy: 0.9736\n", - "Epoch 5/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1052 - accuracy: 0.9767 - val_loss: 0.0765 - val_accuracy: 0.9851\n", - "Epoch 6/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0717 - accuracy: 0.9850 - val_loss: 0.0499 - val_accuracy: 0.9908\n", - "Epoch 7/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0503 - accuracy: 0.9894 - val_loss: 0.0261 - val_accuracy: 0.9966\n", - "Epoch 8/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0366 - accuracy: 0.9931 - val_loss: 0.0266 - val_accuracy: 0.9908\n", - "Epoch 9/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0300 - accuracy: 0.9928 - val_loss: 0.0162 - val_accuracy: 0.9966\n", - "Epoch 10/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0262 - accuracy: 0.9942 - val_loss: 0.0126 - val_accuracy: 0.9977\n", - "Epoch 11/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0278 - accuracy: 0.9919 - val_loss: 0.0103 - val_accuracy: 0.9989\n", - "Epoch 12/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0153 - accuracy: 0.9977 - val_loss: 0.0103 - val_accuracy: 1.0000\n", - "Epoch 13/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0190 - accuracy: 0.9954 - val_loss: 0.0090 - val_accuracy: 0.9977\n", - "Epoch 14/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0151 - accuracy: 0.9963 - val_loss: 0.0063 - val_accuracy: 0.9989\n", - "Epoch 15/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0186 - accuracy: 0.9937 - val_loss: 0.0135 - val_accuracy: 0.9977\n" - ] - } - ], - "source": [ - "model = keras.models.Sequential(name = 'ANN-3x16',\n", - " layers =\n", - " [\n", - " keras.layers.InputLayer(input_shape=input_dim),\n", - " keras.layers.Dense(16, activation=keras.activations.relu),\n", - " keras.layers.Dense(16, activation=keras.activations.relu),\n", - " keras.layers.Dense(16, activation=keras.activations.relu),\n", - " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", - " ]\n", - ")\n", - "\n", - "model.summary()\n", - "model.compile(\n", - " optimizer=keras.optimizers.Adam(),\n", - " loss='categorical_crossentropy',\n", - " metrics=['accuracy'],\n", - ")\n", - "\n", - "model_train_history[model] = model.fit(\n", - " x=x_train,\n", - " y=y_train,\n", - " epochs=epochs,\n", - " batch_size=4,\n", - " validation_split=validation_split,\n", - " shuffle=True,\n", - " verbose=1,\n", - ")" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" + "cell_type": "markdown", + "metadata": { + "id": "ONjOBMTanXiN" + }, + "source": [ + "This Notebook can be used to create Neural Network classifiers running in the [Pose Classification Kit](https://github.com/ArthurFDLR/pose-classification-kit).\n", + "\n", + "First, we have to import several libraries to create and train a new model." + ] }, - "id": "Ils0Rcv4ae-Q", - "outputId": "64c77f7e-c6d3-4d74-d961-1d0f979c73c5" - }, - "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model: \"ANN-3x64\"\n", - "_________________________________________________________________\n", - "Layer (type) Output Shape Param # \n", - "=================================================================\n", - "dense_11 (Dense) (None, 64) 2752 \n", - "_________________________________________________________________\n", - "dense_12 (Dense) (None, 64) 4160 \n", - "_________________________________________________________________\n", - "dense_13 (Dense) (None, 64) 4160 \n", - "_________________________________________________________________\n", - "dense_14 (Dense) (None, 27) 1755 \n", - "=================================================================\n", - "Total params: 12,827\n", - "Trainable params: 12,827\n", - "Non-trainable params: 0\n", - "_________________________________________________________________\n", - "Epoch 1/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.8996 - accuracy: 0.7662 - val_loss: 0.0885 - val_accuracy: 0.9862\n", - "Epoch 2/15\n", - "870/870 [==============================] - 3s 3ms/step - loss: 0.0773 - accuracy: 0.9779 - val_loss: 0.0456 - val_accuracy: 0.9828\n", - "Epoch 3/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0361 - accuracy: 0.9885 - val_loss: 0.1093 - val_accuracy: 0.9655\n", - "Epoch 4/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0294 - accuracy: 0.9925 - val_loss: 0.0247 - val_accuracy: 0.9908\n", - "Epoch 5/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0161 - accuracy: 0.9960 - val_loss: 0.0034 - val_accuracy: 1.0000\n", - "Epoch 6/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0175 - accuracy: 0.9960 - val_loss: 0.0184 - val_accuracy: 0.9931\n", - "Epoch 7/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0132 - accuracy: 0.9965 - val_loss: 0.0286 - val_accuracy: 0.9897\n", - "Epoch 8/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0096 - accuracy: 0.9971 - val_loss: 0.0052 - val_accuracy: 0.9989\n", - "Epoch 9/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0102 - accuracy: 0.9974 - val_loss: 0.0023 - val_accuracy: 1.0000\n", - "Epoch 10/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0109 - accuracy: 0.9977 - val_loss: 0.0013 - val_accuracy: 1.0000\n", - "Epoch 11/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0126 - accuracy: 0.9948 - val_loss: 0.0015 - val_accuracy: 1.0000\n", - "Epoch 12/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0047 - accuracy: 0.9988 - val_loss: 4.2082e-04 - val_accuracy: 1.0000\n", - "Epoch 13/15\n", - "870/870 [==============================] - 3s 3ms/step - loss: 0.0091 - accuracy: 0.9977 - val_loss: 2.6341e-04 - val_accuracy: 1.0000\n", - "Epoch 14/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0126 - accuracy: 0.9971 - val_loss: 1.9235e-04 - val_accuracy: 1.0000\n", - "Epoch 15/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0021 - accuracy: 0.9991 - val_loss: 2.9593e-04 - val_accuracy: 1.0000\n" - ] - } - ], - "source": [ - "model = keras.models.Sequential(name = 'ANN-3x64',\n", - " layers =\n", - " [\n", - " keras.layers.InputLayer(input_shape=input_dim),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", - " ]\n", - ")\n", - "\n", - "model.summary()\n", - "model.compile(\n", - " optimizer=keras.optimizers.Adam(),\n", - " loss='categorical_crossentropy',\n", - " metrics=['accuracy'],\n", - ")\n", - "\n", - "model_train_history[model] = model.fit(\n", - " x=x_train,\n", - " y=y_train,\n", - " epochs=epochs,\n", - " batch_size=4,\n", - " validation_split=validation_split,\n", - " shuffle=True,\n", - " verbose=1,\n", - ")" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "0Xvt8oHFae-I", + "outputId": "aadcc080-4d96-41f0-b09a-e1c506e68475" + }, + "source": [ + "import numpy as np\n", + "import matplotlib.pyplot as plt\n", + "import matplotlib.patches as mpatches\n", + "from matplotlib.lines import Line2D\n", + "import os\n", + "\n", + "try:\n", + " import google.colab\n", + " IN_COLAB = True\n", + "except:\n", + " IN_COLAB = False\n", + "\n", + "if IN_COLAB:\n", + " %tensorflow_version 2.x\n", + " !pip install pose-classification-kit\n", + " \n", + "import tensorflow\n", + "from tensorflow import keras\n", + "from pose_classification_kit.datasets import handDataset\n", + "\n", + "print('Available GPU:')\n", + "!nvidia-smi -L\n", + "print('\\nTensorFlow use GPU at: {}'.format(tensorflow.test.gpu_device_name()))" + ], + "execution_count": 1, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Collecting pose-classification-kit\n", + " Downloading pose_classification_kit-1.1.5-py3-none-any.whl (24.2 MB)\n", + "\u001b[K |████████████████████████████████| 24.2 MB 23 kB/s \n", + "\u001b[?25hRequirement already satisfied: pandas<2.0.0,>=1.1.5 in /usr/local/lib/python3.7/dist-packages (from pose-classification-kit) (1.1.5)\n", + "Requirement already satisfied: tensorflow in /usr/local/lib/python3.7/dist-packages (from pose-classification-kit) (2.5.0)\n", + "Requirement already satisfied: numpy<1.20.0,>=1.19.2 in /usr/local/lib/python3.7/dist-packages (from pose-classification-kit) (1.19.5)\n", + "Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas<2.0.0,>=1.1.5->pose-classification-kit) (2018.9)\n", + "Requirement already satisfied: python-dateutil>=2.7.3 in /usr/local/lib/python3.7/dist-packages (from pandas<2.0.0,>=1.1.5->pose-classification-kit) (2.8.1)\n", + "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.7.3->pandas<2.0.0,>=1.1.5->pose-classification-kit) (1.15.0)\n", + "Requirement already satisfied: h5py~=3.1.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (3.1.0)\n", + "Requirement already satisfied: flatbuffers~=1.12.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (1.12)\n", + "Requirement already satisfied: gast==0.4.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (0.4.0)\n", + "Requirement already satisfied: termcolor~=1.1.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (1.1.0)\n", + "Requirement already satisfied: keras-preprocessing~=1.1.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (1.1.2)\n", + "Requirement already satisfied: grpcio~=1.34.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (1.34.1)\n", + "Requirement already satisfied: wheel~=0.35 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (0.36.2)\n", + "Requirement already satisfied: absl-py~=0.10 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (0.12.0)\n", + "Requirement already satisfied: astunparse~=1.6.3 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (1.6.3)\n", + "Requirement already satisfied: google-pasta~=0.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (0.2.0)\n", + "Requirement already satisfied: keras-nightly~=2.5.0.dev in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (2.5.0.dev2021032900)\n", + "Requirement already satisfied: tensorflow-estimator<2.6.0,>=2.5.0rc0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (2.5.0)\n", + "Requirement already satisfied: tensorboard~=2.5 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (2.5.0)\n", + "Requirement already satisfied: wrapt~=1.12.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (1.12.1)\n", + "Requirement already satisfied: opt-einsum~=3.3.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (3.3.0)\n", + "Requirement already satisfied: protobuf>=3.9.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (3.17.3)\n", + "Requirement already satisfied: typing-extensions~=3.7.4 in /usr/local/lib/python3.7/dist-packages (from tensorflow->pose-classification-kit) (3.7.4.3)\n", + "Requirement already satisfied: cached-property in /usr/local/lib/python3.7/dist-packages (from h5py~=3.1.0->tensorflow->pose-classification-kit) (1.5.2)\n", + "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (1.0.1)\n", + "Requirement already satisfied: google-auth<2,>=1.6.3 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (1.32.1)\n", + "Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (0.6.1)\n", + "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (3.3.4)\n", + "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (0.4.4)\n", + "Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (1.8.0)\n", + "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (2.23.0)\n", + "Requirement already satisfied: setuptools>=41.0.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard~=2.5->tensorflow->pose-classification-kit) (57.2.0)\n", + "Requirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from google-auth<2,>=1.6.3->tensorboard~=2.5->tensorflow->pose-classification-kit) (4.2.2)\n", + "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from google-auth<2,>=1.6.3->tensorboard~=2.5->tensorflow->pose-classification-kit) (0.2.8)\n", + "Requirement already satisfied: rsa<5,>=3.1.4 in /usr/local/lib/python3.7/dist-packages (from google-auth<2,>=1.6.3->tensorboard~=2.5->tensorflow->pose-classification-kit) (4.7.2)\n", + "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.7/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.5->tensorflow->pose-classification-kit) (1.3.0)\n", + "Requirement already satisfied: importlib-metadata in /usr/local/lib/python3.7/dist-packages (from markdown>=2.6.8->tensorboard~=2.5->tensorflow->pose-classification-kit) (4.6.1)\n", + "Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /usr/local/lib/python3.7/dist-packages (from pyasn1-modules>=0.2.1->google-auth<2,>=1.6.3->tensorboard~=2.5->tensorflow->pose-classification-kit) (0.4.8)\n", + "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard~=2.5->tensorflow->pose-classification-kit) (3.0.4)\n", + "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard~=2.5->tensorflow->pose-classification-kit) (1.24.3)\n", + "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard~=2.5->tensorflow->pose-classification-kit) (2.10)\n", + "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard~=2.5->tensorflow->pose-classification-kit) (2021.5.30)\n", + "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.5->tensorflow->pose-classification-kit) (3.1.1)\n", + "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata->markdown>=2.6.8->tensorboard~=2.5->tensorflow->pose-classification-kit) (3.5.0)\n", + "Installing collected packages: pose-classification-kit\n", + "Successfully installed pose-classification-kit-1.1.5\n", + "1 Physical GPUs, 1 Logical GPUs\n", + "Available GPU:\n", + "GPU 0: Tesla T4 (UUID: GPU-65bbdbe4-42ed-f007-b0cb-f5b30cba9c3f)\n", + "\n", + "TensorFlow use GPU at: /device:GPU:0\n" + ], + "name": "stdout" + } + ] }, - "id": "lUaetVfOae-Q", - "outputId": "197b3a7f-8dd7-4e0b-c374-f4e2f2dbcbf6" - }, - "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model: \"ANN-3x64-Dropouts\"\n", - "_________________________________________________________________\n", - "Layer (type) Output Shape Param # \n", - "=================================================================\n", - "dense_15 (Dense) (None, 64) 2752 \n", - "_________________________________________________________________\n", - "dropout (Dropout) (None, 64) 0 \n", - "_________________________________________________________________\n", - "dense_16 (Dense) (None, 64) 4160 \n", - "_________________________________________________________________\n", - "dropout_1 (Dropout) (None, 64) 0 \n", - "_________________________________________________________________\n", - "dense_17 (Dense) (None, 64) 4160 \n", - "_________________________________________________________________\n", - "dropout_2 (Dropout) (None, 64) 0 \n", - "_________________________________________________________________\n", - "dense_18 (Dense) (None, 27) 1755 \n", - "=================================================================\n", - "Total params: 12,827\n", - "Trainable params: 12,827\n", - "Non-trainable params: 0\n", - "_________________________________________________________________\n", - "Epoch 1/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 1.9930 - accuracy: 0.3888 - val_loss: 0.5271 - val_accuracy: 0.8874\n", - "Epoch 2/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.7371 - accuracy: 0.7397 - val_loss: 0.1629 - val_accuracy: 0.9828\n", - "Epoch 3/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.4313 - accuracy: 0.8513 - val_loss: 0.0708 - val_accuracy: 0.9908\n", - "Epoch 4/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.3245 - accuracy: 0.8884 - val_loss: 0.0536 - val_accuracy: 0.9920\n", - "Epoch 5/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.2590 - accuracy: 0.9065 - val_loss: 0.0292 - val_accuracy: 0.9954\n", - "Epoch 6/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1947 - accuracy: 0.9298 - val_loss: 0.0196 - val_accuracy: 0.9966\n", - "Epoch 7/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1741 - accuracy: 0.9416 - val_loss: 0.0161 - val_accuracy: 0.9954\n", - "Epoch 8/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1483 - accuracy: 0.9497 - val_loss: 0.0100 - val_accuracy: 0.9977\n", - "Epoch 9/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1455 - accuracy: 0.9548 - val_loss: 0.0067 - val_accuracy: 1.0000\n", - "Epoch 10/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1100 - accuracy: 0.9681 - val_loss: 0.0063 - val_accuracy: 0.9977\n", - "Epoch 11/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1179 - accuracy: 0.9632 - val_loss: 0.0090 - val_accuracy: 0.9966\n", - "Epoch 12/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.1075 - accuracy: 0.9632 - val_loss: 0.0056 - val_accuracy: 0.9989\n", - "Epoch 13/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0918 - accuracy: 0.9721 - val_loss: 0.0035 - val_accuracy: 1.0000\n", - "Epoch 14/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0856 - accuracy: 0.9735 - val_loss: 0.0271 - val_accuracy: 0.9931\n", - "Epoch 15/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0780 - accuracy: 0.9756 - val_loss: 0.0095 - val_accuracy: 0.9966\n" - ] - } - ], - "source": [ - "model = keras.models.Sequential(name = 'ANN-3x64-Dropouts',\n", - " layers =\n", - " [\n", - " keras.layers.InputLayer(input_shape=input_dim),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dropout(0.3),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dropout(0.3),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dropout(0.3),\n", - " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", - " ]\n", - ")\n", - "\n", - "model.summary()\n", - "model.compile(\n", - " optimizer=keras.optimizers.Adam(),\n", - " loss='categorical_crossentropy',\n", - " metrics=['accuracy'],\n", - ")\n", - "\n", - "model_train_history[model] = model.fit(\n", - " x=x_train,\n", - " y=y_train,\n", - " epochs=epochs,\n", - " batch_size=4,\n", - " validation_split=validation_split,\n", - " shuffle=True,\n", - " verbose=1,\n", - ")" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" + "cell_type": "markdown", + "metadata": { + "id": "JSmWOsSRg8i5" + }, + "source": [ + "## Import dataset" + ] }, - "id": "FA1LSFoVae-R", - "outputId": "fb01face-dfb0-4bc5-f45d-7f96bf0e1b46" - }, - "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model: \"ANN-2x128\"\n", - "_________________________________________________________________\n", - "Layer (type) Output Shape Param # \n", - "=================================================================\n", - "dense_19 (Dense) (None, 128) 5504 \n", - "_________________________________________________________________\n", - "dense_20 (Dense) (None, 128) 16512 \n", - "_________________________________________________________________\n", - "dense_21 (Dense) (None, 27) 3483 \n", - "=================================================================\n", - "Total params: 25,499\n", - "Trainable params: 25,499\n", - "Non-trainable params: 0\n", - "_________________________________________________________________\n", - "Epoch 1/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.7411 - accuracy: 0.8277 - val_loss: 0.0596 - val_accuracy: 0.9931\n", - "Epoch 2/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0452 - accuracy: 0.9914 - val_loss: 0.0435 - val_accuracy: 0.9828\n", - "Epoch 3/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0196 - accuracy: 0.9960 - val_loss: 0.0077 - val_accuracy: 1.0000\n", - "Epoch 4/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0100 - accuracy: 0.9974 - val_loss: 0.0080 - val_accuracy: 1.0000\n", - "Epoch 5/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0085 - accuracy: 0.9991 - val_loss: 0.0021 - val_accuracy: 1.0000\n", - "Epoch 6/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0047 - accuracy: 0.9991 - val_loss: 0.0028 - val_accuracy: 0.9989\n", - "Epoch 7/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0028 - accuracy: 0.9997 - val_loss: 0.0012 - val_accuracy: 1.0000\n", - "Epoch 8/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0047 - accuracy: 0.9986 - val_loss: 0.0011 - val_accuracy: 1.0000\n", - "Epoch 9/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0037 - accuracy: 0.9994 - val_loss: 0.0019 - val_accuracy: 1.0000\n", - "Epoch 10/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0014 - accuracy: 0.9994 - val_loss: 0.0024 - val_accuracy: 1.0000\n", - "Epoch 11/15\n", - "870/870 [==============================] - 2s 3ms/step - loss: 0.0084 - accuracy: 0.9980 - val_loss: 0.4499 - val_accuracy: 0.9563\n", - "Epoch 12/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0198 - accuracy: 0.9963 - val_loss: 6.7393e-04 - val_accuracy: 1.0000\n", - "Epoch 13/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 5.4952e-04 - accuracy: 1.0000 - val_loss: 4.7648e-04 - val_accuracy: 1.0000\n", - "Epoch 14/15\n", - "870/870 [==============================] - 3s 3ms/step - loss: 0.0025 - accuracy: 0.9994 - val_loss: 0.0044 - val_accuracy: 0.9989\n", - "Epoch 15/15\n", - "870/870 [==============================] - 2s 2ms/step - loss: 0.0043 - accuracy: 0.9991 - val_loss: 7.2220e-04 - val_accuracy: 1.0000\n" - ] - } - ], - "source": [ - "model = keras.models.Sequential(name = 'ANN-2x128',\n", - " layers =\n", - " [\n", - " keras.layers.InputLayer(input_shape=input_dim),\n", - " keras.layers.Dense(128, activation=keras.activations.relu),\n", - " keras.layers.Dense(128, activation=keras.activations.relu),\n", - " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", - " ]\n", - ")\n", - "\n", - "model.summary()\n", - "model.compile(\n", - " optimizer=keras.optimizers.Adam(),\n", - " loss='categorical_crossentropy',\n", - " metrics=['accuracy'],\n", - ")\n", - "\n", - "model_train_history[model] = model.fit(\n", - " x=x_train,\n", - " y=y_train,\n", - " epochs=epochs,\n", - " batch_size=4,\n", - " validation_split=validation_split,\n", - " shuffle=True,\n", - " verbose=1,\n", - ")" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/", - "height": 355 + "cell_type": "code", + "metadata": { + "id": "_v5VYGMCg8i6", + "outputId": "15208251-c9aa-403c-ba55-3529a3e6dc7e", + "colab": { + "base_uri": "https://localhost:8080/" + } + }, + "source": [ + "dataset = handDataset(testSplit=.2, shuffle=True, handID=1)\n", + "x_train = dataset['x_train']\n", + "y_train = dataset['y_train_onehot']\n", + "labels = dataset['labels']\n", + "\n", + "x_train.shape, y_train.shape" + ], + "execution_count": 2, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "((4347, 42), (4347, 27))" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 2 + } + ] }, - "id": "4UhLcq1Jae-R", - "outputId": "91bd3127-5407-4d3c-82dc-37dad2977465" - }, - "outputs": [ { - "data": { - "text/plain": [ - "" + "cell_type": "markdown", + "metadata": { + "id": "xRWuVIAHae-P" + }, + "source": [ + "## Models exploration\n", + "\n", + "This section is optional. The following blocks can be used to compare different model architecture and training processes." ] - }, - "execution_count": 15, - "metadata": {}, - "output_type": "execute_result" }, { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6QAAAFBCAYAAABgqWLFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOzdd3hU15n48e+5906v6r0L0XsxxsbGuIEd95reHCebbPqW7P6y6dmSnqyTOI7tJE42TtziEndwwcYG04vooI56HWn63PP7Y4QAI4EEEhL4fJ6HBzRzy5lhNPe+p7yvkFKiKIqiKIqiKIqiKGebNt4NUBRFURRFURRFUd6bVECqKIqiKIqiKIqijAsVkCqKoiiKoiiKoijjQgWkiqIoiqIoiqIoyrhQAamiKIqiKIqiKIoyLlRAqiiKoiiKoiiKoowLFZAqiqIoE44Q4kEhRIsQYucQzwshxC+EEAeEENuFEPOOee4FIUSXEOLvZ6/FiqIoiqKcDhWQKoqiKBPR74EVJ3l+JTCp/8/dwK+Pee6HwIfHrGWKoiiKoowaFZAqiqIoE46Ucg3QcZJNbgAekknrAL8QIqd/39VA4Cw0U1EURVGUM6QCUkVRFOVclAfUHfNzff9jiqIoiqKcQ4zxbsBYSE9Pl8XFxePdDEVRlAlv06ZNbVLKjPFux3gQQtxNcrovLpdr/pQpU8a5RYqiKOeG9/K1Qxl951VAKoS4DriuvLycjRs3jndzFEVRJjwhRM14t+E0NQAFx/yc3//YsEkp7wPuA1iwYIFU1w1FUZThOYevHcoEdF5N2ZVSPiOlvNvn8413UxRFUZSx9TTwkf5su4uBbill43g3SlEURVGUkTmvRkgVRVGU84MQ4mFgGZAuhKgHvglYAKSU9wLPAdcAB4Ag8PFj9n0DmAK4+/f9pJTyxbP6AhRFURRFGRYVkCqKoigTjpTy/ad4XgKfG+K5pWPSKEVRFEVRRt15NWVXURRFURRFURRFOXeogFRRFEVRFEVRFEUZFyogVRRFURRFURRFUcaFCkgVRVEURVEURVGUcaECUkVRFEVRFEVRFGVcqIBUURRFURRFURRFGRcqIH0XUybGuwmKoiiKoiiKoijvCedVQCqEuE4IcV93d/dp7V/buYtHtn6fYCwwyi1TFEVRFEVRFEVR3u28CkillM9IKe/2+Xynt3/zVro6dvHSvgeJm7FRbp2iKIqiKIqiKIpyrPMqID1TBWU3Ml/PoLF9G2/VP4eUcrybpCiKoiiKoiiKct5SAekxNIuTKWW3Uh6V7G1+k11t74x3kxRFURRFURRFUc5bKiB9F3feRUx3lpIeDLD+8EscDlSNd5MURVEURVEURVHOSyogfRehGaRU3MKcmIYzFuXV2sfpiXSOd7MURVEURVEURVHOOyogHYQjfRrutBnMDfRhJmKsqv4rsUR0vJulKIqiKIqiKIpyXlEB6RD8k27GmYizQPgJRLtoDTaMd5MURVEURVEURVHOKyogHYLFnY07/2Lczbu5peTD5HpKxrtJiqIoiqIoiqIo5xVjvBswkXlLr8GROQe7Ow+Ag507MTQLRb7J49wyRVEURVEURVGUc58KSE9Ct7rRUysASCSiVLatx264KPRWIIQY59YpiqIoiqIoiqKc21RAOgw91asINm3gynmfw2q4VDCqKIqiKIqiKIoyCtQa0mGwuLKx+kqwaVZ0TSccD7Kx8RVMmRjvpimKoiiKoiiKopyz1AjpMDgyZuDImDHwc33gINta1mLKBItyrxzHlimKoiiKoiiKopy7VEA6ApHuKqI9tZQXXEprsIEdretIdWRRnjJrvJumKIqiKIqiKIpyzpnwAakQwgX8CogCr0kp/2+82tJ3eD19h9/GnjqFRTlX0hlu4c36v+OzpZPhzB2vZimKoiiKoiiKopyTxmUNqRDiQSFEixBi57seXyGE2CuEOCCE+Fr/wzcDj0kpPwVcf9Ybewxf2bUIzUrX/r+hazrLi27FabhZXf0IwVjveDZNURRFURRFURTlnDNeSY1+D6w49gEhhA78ElgJTAPeL4SYBuQDdf2bjWsWId3qwVtyNeG2nYTb92A3nFxefDuRRJhXah4lYcbHs3mKoiiKoiiKoijnlHEJSKWUa4COdz28CDggpTwkpYwCfwFuAOpJBqUwAbICewovxXCk07XvCaSZIM2RzSUF19PcV89bDc8jpRzvJiqKoiiKoiiKopwTxj3AO0YeR0dCIRmI5gFPALcIIX4NPDPUzkKIu4UQG4UQG1tbW8eskUKz4Cu/kVjfYfoOvwVAiX8aczIvZl/HVmp69o7ZuRVFURRFURRFUc4nEz6pkZSyD/j4MLa7D7gPYMGCBWM6TOnInI3NP4nug8/izJqPZnEyL3sZHpufQu+ksTy1oiiKoiiKoijKeWMijZA2AAXH/Jzf/9hZ09GSoPZA7JTbCSHwT74FM9ZHT9WLA49VpM5FEzrBWIDeaPdYN1dRFEVRFEVRFOWcNpEC0g3AJCFEiRDCCtwJPD2SAwghrhNC3NfdfXrBYEN1jJ0bwsRjpx5gtXryceUuJti0EZk4GsRKKXnh0P/xWu0Taj2poiiKoiiKoijKSYxX2ZeHgbeByUKIeiHEJ6WUceAfgReB3cAjUsrKkRxXSvmMlPJun893Wu3KKTRIJKC1cXjZcn3lN5B94f9D6JaBx4QQLM69miV51yCEOK12KIqiKIqiKIqivBeMyxpSKeX7h3j8OeC5s9ycAamZOjaboLE2Tk6h5ZTb61Y3AFKamNEAui0ZCOd6Sga2aQ81kebIHpsGK4qiKIqiKIqinMMm0pTdM3amU3Y1TZBdaNBcHycRH/502/Zt99O65ddIaR73+P6ObTy1/35qe/afVnsURVEURVEURVHOZ+dVQHqmU3bDjVVk6l3E49DamBj2fq68JXhLrgKOn6Jb4p9Gmj2L12v/poJSRVEURVEURVGUdzmvAtIzIRMxmh56iMgzT2PoUQ7Xnjrb7hGOjBk4s+adsGbU0CxcUXwHbquPl6v+wpq6p4jEQ6PddEVRFEVRFEVRlHPSeRWQnsmUXaFb8F+4HIJxXK2P01QTIpEY/rRdKSU91S8NlIE5wmX1cn35XczNWsrBzh08se9eanv2jbh9iqIoiqIoiqIo55vzKiA90ym73vmLsafPIKWvlkDDVpqrO4a9rxCCWG8TPYdeIB5qP+45XdOZl72M68o/id1w8nLVX3m99knC8eBptVNRFEVRFEVRFOV8cF4FpGdKc1mwFeWSnbUSu1ZFw9r/IhZsHfb+vvLrQEDX/qcGfT7dmTMwWnqoq5J1h18cdDtFUZT3OiHEg0KIFiHEziGeF0KIXwghDgghtgsh5h3z3EeFEPv7/3z07LVaURRFUZSRUgHpu1iLPWi9Pq64dTZe8Sa1z3+CWF/TsPY17Cl4iq8k1LKZSNfBQbc5Mlp6/aRPsiB7OQDBWECNliqKohzv98CKkzy/EpjU/+du4NcAQohU4JvABcAi4JtCiJQxbamiKIqiKKdtXOqQjhUhxHXAdeXl5ad9DEuxh9A7LRjhcvKuuJeal79CzfOfoGjFA1jcOafc31N0BX0Nb9O193EyF/3zCYmOjji2Numb9c/SHWnj5op/QNf00267oijK+UJKuUYIUXySTW4AHpJSSmCdEMIvhMgBlgEvSyk7AIQQL5MMbB8e4yafE4LrW0h0Rca7GcPS05HAWejCO9uP7j51bfDxEm8LE93fTaInOt5NOaVo2KS+rYlwNIJX5CEY/B7ldEhNEC5PR9om8K2lhES4A/1QHbawc6Ce/JkyiROMtyKCYexpZcSzvST8jlE59qhLmFja+rA0Bci+Lhd/wQRtp/KeMoG/NUZOSvkM8MyCBQs+dbrHMDIdCLtBtDrAhvAkPDn3Y7R8mJrnP07higewevJOur+mW/GVX0dH5UMEG9/BlXvBKc+5IPsyuiPt6JqOlJKoGcGm20/3JSiKorwX5AF1x/xc3//YUI+fQAhxN8nRVQoLC8emlRNIrDFI5327x7sZIxLs/2PkubBN8mGt8GGr8KGn2MalPVJK4k0hovu6iOzrJrqvm0THuRHgH5E+8K/6UT921e4YB93pp95wPEk71zdFMWQE6By1wyazlxhADSbQanPTZnVhmAnCumXgffHFQpgIIppBVNNhiIGL0aKbJqmxIOmRPtKjfaRGg+gkk3b2lLlVQKpMCOdVQHqmTGkiMbEUuonV9lJ6SSZOdyGeub+l9sW7qXn+4xSteACrt+Ckx3FmL6S3bg3dB57GkTkHzTj5hTPVkUWqIwuAA53b2dC4iiX511Dsmzpqr01RFEU5npTyPuA+gAULFgw/rfo5KrIzmagv87sLxi2gG65Vf+ujem8UdyjMzOwIOSJI8O1m+l47DICeae8PUP3JADXDPuSMpDMhTUmsppfwjnai+7qJVgeQoWSdcmHT0HxWfB8ow1bhJ7K7k3hXFO91Rcl9EyZCP3sroySSrnaTxpo4TbVxDtfG6e2JE0rbTqDoOSLuWpyksjD9OjL8ft48/BRNwWp8tgwuyrmeuZnLMLTTHImW0P61dSxeqHPFLeM3Q95MRAm37ybUso1g8xbCbTswY70AWFw5ODLn4PDOI/ZrH+LCDrqMB0lEu3FmzCZl6gdwZM8f9HPUG+2iNrCH6sAeart30xKqRSLRMZjbfQFTDkzF4fGwpWgTkYZuLoldQ26rSdbh5uQBNMGCzCCWST5i1T2AQOgCNIHmMhAeK5rHgua2JP9Os2NkOU/vPQjGiB3sIba/m9iBbuK1vWBKEGAUuLGU52Ip92Ep92H1T9yZB8p7iwpI+yXMOA9s/y7ZrkKuLr6F6P5u8jMkRqoBTKNoxf0DQWnhivux+YqHPJYQAn/FzbRs/AmBmlX4yq4ddjvSHNm4LF5WVz9GiX8aF+atwGG4zvwFKoqinF8agGN7B/P7H2sgOW332MdfO2utmsDClZ0Y2Q4suRP7mtLWFGfv7gQLLnERCTt57Z0IN328iLwig1hdL9F93UT2dRHe1k5wbfKGX/NbsVX4sE7yYavwY+Q6EdrIAtRER4RoTQ+ay0L0QA+hDa3EGvqgvwScsGoIl4GeZkfPsGOk29F9VlyX5iIMjURnBOGyoDmSt1Zd/7cfTDDynFjyXBh5rlGdemyakvamBA01cQ5XxzhcEyfUl2yrw22iT95Eu+9peuOHKUyUcZHjM5SaU5DVMcy+OJPzvkLt/MOsrn6E9r9X8lT6Rgoum80FeVcT/L/aU57fNsmH88JkZ3rnQ/vQXAb0RLBYgZiJsI79EiQzFiTUup1g8yaCTZsJtW5HJpIj1lZfKanly3BmzcOZNR+LO7lUKlrbSyubSJl7MdmzV9C193Had/6e5rc/jT19BmmzPkUsYypV3bs41FXJoa5K2kLJjhCrbqfIO5nLCm6l1D2d9C0e4k0BLNM8uK/OJ12bxo/e+Twt9i4+P/8HEJRE9ydH0iP7uwm+WAsmIEDPdGBkOBAaaPEEsilKrDcOSKxlXpyFyY6Nrv/bj7XCh3NhJjIh6V1Vj2bTEf1/ZCxBvCVMrKGPeF0v8eYQSMAQWEs8uFcWJDtvyr0Dn01FmWjUJ7OfrhmkO3J4o+5pps1cSOnd09FsOr09Jn09Jln5UyhccT+1L36K2uc/mQxK/SVDHs/mL8WZNZ9ooB4p5bB7blMdWbyv/BPsaH2brc1raOqt5sK8lZT4p43WS1UURTkfPA38oxDiLyQTGHVLKRuFEC8C/3lMIqOrgH8br0ZOFDJmEt3bhXPpqXMhjLd1q0NY7YKZZQnM2l7iIkjlg514ljswLALh0LHPTsM2KxWzKwpCEG8MEtnTReidZGZ84TKwFroRLgMj24meajsaoJqSRE8MMxAj0R3FPjeNRFOI0JY24g199M9mRE+zYS32YCl0JUeT8lwDo1jCOHHk0z4rbeDfUkpsk/3EGvqI7u0eGJ3WvNaB4NSS60TzWYd9f5CIS1oOx2mojnO4Ok5jbZxoJNlYr1+jaJKFrCJo966lYfN6jDrB0gPzKZS347WmJMvTGT0Y6UeD6WnpC5matoDqpg1sir3J0wceYHXNo9xo3EmRbwoWzTpke7Rjgmsjw47utZLojJBoj9D9lwNY8l1YS7xYSj3onqGPMxKJSA/Bli0EmzYRbN5EuG03yDgIDXvqZFIm34Yjez7OrLkY9tTBj9GZDFj1FBua4SBl2geJFCxh14G/sL/5LRr3/pC+g8nbY4fhosQ/ncV5V1Pqn06euxRdM0j0RAk8V0u8JYBjYSaOCzIRmsBLKrdM/ix/3Pk/rKp+hKtLP4BjbjqOucnpumYoTvRgz0CAGtnTCfHk/6GR58I6yYul0I21xAskP0dGthPdk3yvzViC6KEeEh0RzK4oZm8MGTX7/0NAcxoYWQ7sc9Lw3lCMlND10D6Y7ENzGCQCMYJvNBJJCIxiD/6Z/lH5f1GUMyWS+SDOD8ckNfrU/v37R7x/JBHmp+98kYRM8NVFv8BuONn8Zoi2xgRX3OJC0wSRroPUvHAXSEnRivuxpQydQEkmYqAZpz2NqCPUwpv1z9AaPKxGSxVFGRNCiE1SygXj3Y53E0I8THKkMx1oJpk51wIgpbxXJL9Y7yGZsCgIfFxKubF/308A/95/qO9LKX93qvMtWLBAbty4cbRfxoQR3tVJ+4+3k/aFGdhnp516h3Eg4yZNb7bzwotxZlztZ+4MQfdfDxIOSRqqYnj8Gpl5J/aju6/Kx1buI1oboPvxKmwVfhItIcI7O5IBKySnRrqN5JRIh4GMmSQC0eTzkuR0xnwX1lIvtukpyWnAoxRESVOSaA0TO9xH/HAfsYYgMhwHwD4/A9dF2cltOiLoabbj7hkO18SoPRCjsSZOY12cRHI30tKgMEuSMc1NbrGB3N5I3f5d/G3yX+iKtLFi13WUGFNIzyvAkuFA7w9CTxUAV3fvZlX1o+xp34jdcHFx/rUszb8el9V7ytfZcf8eovu6yPj6PMLb2olWBTD7E2jpafaB4NTIcgz7vigebCPYvLn/zyYiHfsAidAs2NNn4MxOjn46MucMO0FR32uH6frjfg58vptqvZJDXbvoi3UD4LH6KTBSSW07REZXA1nOPNJnfhJf2TWIY6YzRw/10PtSPe6r8rGWnvje/LnyJ2xtWcM/zv8hhd5JQ7ZFxkyiVUcD1Oj+HmQkOSVcz7RjneRDL4hBQiNRK4ju7xlYsyycBrZyL5ZiD5ZCN3q6DeISGTHRU6wYWU7MSILg2iZsk3zouS4aK/sI/L2WaF8CY1oq0z548rwoJzNRrx3Kuem8CkiPOJMbi+ruPfxy09e4xLqSS5qvpG9mDpveiXPhFQ7Ss5MXwkhXFbUv3IU04xSu+C321IqTHjPW20Skcz+u/ItHHJyaMsGO1nVsaXodi25jSd5Kin1Tx2StjKIo7z3qpiLpfA9Iux85SO+qBnL+9yI028TJ5i4TErMnip5iQ8ZNdnx9Bw24uPzbZVhtR69z61YF2fB6mGve76Zs2vADxURXJHmjv6+byL5u4vV9ySf05HTGZJIk/1mdziilJNEZId7Qh57hwJLtJN4covuvB3CvKMBW4cfsi1G9Nchzz8exyTi5/gQ53gSp1jhuM4roiwES510lrGt9iYPrNmDp0gjMj3NFye1U+OeincH61frAQVZXP8qO1rew6nYuzF3BpYU34rUNPuoI0P34IXpfrCf33qUDo9GJzgjRqgDRqh7iDUFAIhwG1hIPzouyj3vPpZTEeg8TbN5EqHkzwaZNRHtqABCGA2fmbBxZ83Fmz8eRPgPNGHnyx5ruPRx86B0mbSvgF3f+DDvp5FimUpYynXnls0l35CCEQJoJAjWradv+WyIde7G4c0md8Qnc3quw5vmB5GjnUJ+ZUKyXH73zBay6jS8v/BlWfXhrts1YnL7KSoLbq4gdDEOTDxFPriWV1gAiJ4Cl3I1rTjmOyRVo+ql/lyMhk5oDMWr2xwgHJU63RnGFhYIyy3G/YyOlrh3KaFJTdt+l2DeF5UW3sHHny0zvmk6eJxtdh8ba+EBAavOXULTyQWpeuIvaFz5J4dX3YU8bOgFRb/0bBFu24Mieh24Z2QinJnRmZ15EkbeCNXVPs63lTYp8kxFMnBsKRTkVaUoCz9biujQH3Ts6ow6KogxfuLIT6yTfhAhGpZTEm0NE9yQz1Qqbjv8jkzhcl+BtLYeF17hPuFFeeJmD6v0xXnmqj+wCA5dneMGW7rfhXJiJc2EmAGZvjHhrCEue66yscRyMEAIj1Y6RejSg0rwW3FfmY8lPjvKF93UTeKyOiyXkFhnoOhARaHYrRqaDRIqDnYkNrFr/A3plN5MqZnNZ8R2U+qePSod1vqeMj878Gk19tbxS/Rhr6p5mbcOzLMq5kmWFN5PqyDxhHz3FBgmJGYih+6wDjzlSbDjmpWOG4sRqeoke6iFW1wdWjUjXIQLrdxDp3EdQe5l4f913zerBmTUPf8UtOLPnYU+betwI5UhIKTnYtZNV1X/lQOd2VnS/j4jD5ENZvybankJHcwJ3WCNjVjLw27c9gjdFJ7vkKjzFV9Jbv4a2bffR/vdnCLR34brRTer8G9EcQycdcljc3Dn1i/xm63/w3ME/cGPF3YNuZyaihNsqkyPATZsItWw9moRpRh6Oy+fjMBYiRYxwZAPB5k3Egy20rwd9qx9H1rzkGtns+dhTJyM0Y+A1d7WZVO2L0lgTxzQhI0dn1iIrmbn6iNdXK8pYUwHpIK4suZPd7Rt5KPU3/FPaPWTlWWmsjTNjgRz4Jbb6iii65nfUPP9Jal64i8KrfoMjY8agx/NPvhVP0XJ0iwspTRKRbgz7yLLQ+e0ZvK/8Y4TjQTShE0mEqe7aRal/BhZd3eArE1ustpfAk9VoTgP35ac/RUhRlJFLdEWI1/fhvWXovAdnpR3dUSJ7uojs7UpO5dQ1rKUebFP8SCl5e3UIi9/CrAtOLEOh64KrbnHzl19388pTfbzvg+7TCrw0twXrBKxpqjkMbFOP3hfUxRxUWtJZPF/gneZMJlFKs9MnA6ype4q19c8SSYSYlr6Iy4tuo8g3eUzale0q5APTv8JVJXfyas0TrD/8EusOv8j87GUsL7qNDGfuwLZHMjcnOiMDAemxhE0gMxqJJTYRtG2i9ZHNJMKd2GquRVgcOJbMxjnzExid03CUlWFkOs8ouJZSsqdjM6urH6G6ezd2fJR2f4DSyCTc2YLymbkD2x1ZiytNSd3BGDmFkF1gYJqw79BC/KWLcRXsILb5KVp3PUXHod+QOuPDpEy5A93qGfT8k1Jnc3H+dbxZ/wzT0hdSkTr35EmY/GV4S1fizJ6PM2seFlf2u454a/8ocsPAGtpg0yZ6a18BQDOcOLLm4Myaj/As4K21pRhWnaIKC8UVVtzes5fxWVFGSgWkgzA0C3dO+xI/3/BVHt/1K64o+AqHa+N0tCZIyzr6llk9+RStfJDaF+6i9sW7Kbzq1zgyZ59wPCEEhiO5Zqen6kV6a18ldcZHcaRPH1G7NKHjtCS/+Ha3bWRT06vkecqw6Fb6oj3YDMfpp2xXlDEUbwoCkGgPj3NLFOW9J1yZrLVomzH0dMuxYoYTRPd3EdnTRbwxCAiMfBeO+elYy4+O2NbsT47kXPo+J4Zl8CAkNVNnyZVO3ng+yK7NUabPn9ila05XIiFZvzaGUeil5DYvQhN0R9p5rfrPrGt4gbgZY1bmRVxedBu5nrPTyZDuzOW2qf/IlSV38Frt31h3+CU2Nr7KnKyLWV50GznuoqMBaUcEij3IRIxQW+VA4HTc6J87D3f+0uT6z5vmYtjz0e0GZjhB56rdxDYeQnNbsBR7sJZ5sRS4hz2qZ0qTytZ1rKp5lIbAQfy2dBYYn8BRu5SZ8z24Nm5FTzk6W00Igc2ePLbQBMtvdGEml3HSu7ML460WdudlI7XJGM5/xeP4OEbgeXrefpLW7X8gfdqdpEz7IIbdf0JbVuTfxJ7mtfx52/e5o88KbXuOScI0hZQptyffg6y5wxqoEEJg9eRj9eTjn3QDALFgC6Gmzezb2kl9VTu5Df8LQGZsFhkpDtzabAjMx3TMRrOcXikZRRlrKiAdQq67hOvdHyH2ZCvNN6xF1xfSWBs/LiAFsHryKFr5O2pe+CS1L32Ggit/hTNr7pDHdWUvJNSyjbat9+ItXYm3ZOVp9QDOzryIIt9k3NZkKea1Dc/R3FdLsW8Kpf4Z5LiL0YTqDVMmhiMBabxNBaSKcrZFKjvRvBYs+WcnKZ6Mm8iERLPpxJuD9L16GD3FjnNJNtbJ/oGMoQPbS8m61SG8fu2UQebsxTaq9kR547k+8ksMfKnjPwV5tO3ZEqG7w+TaD7jpiDTzas0TbGhchcRkbtalXF50G5mu/HFpm9+ewY0Vd3N58W28XvsUbzc8z5bmNcxIX8xlvmsxgO4dL9PSvGrw0b93lWB5N82uk/LJKUSrA8SqepIj6js7EA4DW4Uf2xQ/eubgNWcTZoJtLW+wuuZRmvvqSHfkcNuUf2R+9mVEgzqBUpPMPJ3GzshJO2eEEGiapO+NRuJb2iguc1K+3EFXr6C9OUFHSxad8iMkLDeR6KjC+uomcrd8nPxZF+Mvv4FI16GBIDzSuZ8lVgvPZOfwhg1unPnRESdhGoqUko6WBKkZGXhLV+DuiSDSTcoXfpBw69b+UdTNtG9/gHZ5Hwgde9rUgSm+zqx56DbfGbVBUUbLeZXU6Eyz7L5bPBRl64+eZFfeTopKP0W0PYUrbnYN+kUY62um9oW7iAVbKLjyl7iyh17nbSaidO75K8HG9cmaV9M/csa9Vod7qzjQuZ3q7j3EElGcFjel/umU+mcMLNJXlPHSce8uQhtasRR7yPyPeePdHOUYKjFF0vma1EiakqYvv41tRgqpnxo618GonS9u0vm7vdim+HEtzUlmj20Po6cPHkQAHNwV5bmHe7niJhdT55161DPQleDP9/SQlq1z8yc8aOfRerhEXPLQz7rRUpuwLn6BLc2vIdBYmHMFlxXdTJpj8EBuNEgpCQclsWjyTzxG8t8xSTwqicUgHpP403UKyyzEQt28+GQV7SmPss+ygwjwhYe/xKGy9QRnb6IsZRbZOUtwZc8bsgTLKdsUN4nV9BLZ00W0qgdMiZ5ix72yACM9uQY3bsbY1PQqr9Q8RnuoiSxXIZcX3UZ2dDGtDZJZi49mLjaDcRo/vxZtSSParAa8xVdjTz9+3a0ZjBN4vpZ4Qx/2WWk4l+Yg9OM/Y9GwpKM1QXtznJa6DvKdvyfe+AjtffNp7L2CKVn34cuZRMS+nIRtJpW2jbzd+TgfmPIvzMu9+PT+g/rFY5L6qhjVe2MEuk0WLXOQlW8MWWIwEesj1LKNYPNmQk2bCLXtQCaS2afzr7gHT8Elp9UOde1QRtN5NUIqpXwGeGbBggWfGo3jGQ4r+eVTaKlvYuuUeykJ/QudrSapmSf2yFpcWRSufJDaFz5F3UufpeCK/8WVe8Ggx9V0K6nTPoTNW0znvsdpfueHpM26C6vn9NfW5bpLyHWXsCTvGup69nOoaye72zays3U9Plsqpf4ZlKfMwmsb2dpVRRkNsUY1ZVdRxkOsthezN4Z9DKbrSlMSbwoSPRTA7IvhuboAYWg45mdgZCbXgQpNYGScuCb0CNNMjo760zUmzx5ePgSPX+eSa52seqKPLWvDzF869PHPNWve2UtV3mMEszZgabFwUf77WFZ4Ez7b8Er19AXMgYDyuKAyKonHkn+7PBoVs5KB/5svBPGnacxYmAzuVj/Zx1DjFIIoIt5Bn2sn8Z1/JtKxD6Prdiq667iwwM8OezFBR5REPJ1X7FFeCW3EW3+I0t4tAx3kma78Ec3eEoaGtcyLtcybrOF5oIfI/m40j4VYIsLWd1azo24tu9J2kO8t56Mz/43p6RegCY0DlVF6uuLEY2Dp/2gdqUHqKp9DILiFtm33YnHn4y25GkfmHOLNIXqfrcUMJ5LlhKYMfs9ktQuyCwyyCwymL8gF/p1oz4ewVVZi653M9MtuQTOsbF8fpmZ/DINcPOlbeGTXr6hbW4TfmYrTpeFwCRwujZIpFjRNkIhLNJ1BA8veHpPqvVHqDsWTHQOpGnMutJOenbwnHarDR7e4cOctwZ23BAAzHiHctpNg8+YRLx1TlLFyXgWkY8E3KYeK2hms7X4Tl28Vwb7rSR0iw63FmUHRygeSQemqfyT/8p8PfAG8mxACd8FSLJ582rffT8uGH5Ey9QO4chaeUXsNzUKJfxol/mlEEmFqundzsHMnW1vewGY4mG5bRCwRJWZGBtajKspYkqYk0RICXWAGYpiRxITI9Kko7wWRyg4AbNNGpzPSjCSI1fZnSq3pTdbTFAJLgRuZkAhd4JiXPuzj7d8RpaMlwdW3u9D04Y90TpljpWpPlHWrQxRNsgxkwT9X1XTvZVXVI+wOb0DPsLMk8yZmu96HFvXSvE/SEI0MjFQaFsHsC5MB5IbXQiQSksWXJ2dZrX8lRF/APOH4AjAsAov1+MAlLUsfyFgshGDOhXY0HQwLiFgrse7txDs2E2l7h3jgUHK7sB3DOwfP3M9SmD0fR/pX0Qw7FUDTui1Y+jJZ0HIxEf9eEr79HOqqZGvLGwA4LR5KfNP6A9Tp5LpL0bXhXQ80h4F9ZipMtfN6w1OsqXuSKVsqKDMnseTym5mcOo9YU4hAh4kvTaNsmoWSKRb0/s+VlJK+A9sAsOXm4Cn7PsGmjfRUv0z79gfQO6di1M7CkpaB7/bSk3akDMbqLaT0wkJKj3ls5iIbFTOthPokpZ1f4k/1/8ShzPtZkvhXgr0mbc0SaUpKpyansW9/J0Jna4LlNySn11ftiRKLJkdjWxsTaFoy63JxhRV/unZ6ib0MW3LKbvb8Ee+rKGPl3P4GPwssxR789gwWxC9mvedhrJkXAEOv3TAcaRSuvJ/aF+6mfvUXyL/sp7gLlg65vc1fQtYF/0r7jt/RUfkHZDx80u1HwqbbqUidS0XqXIKxAHp/wqOq7krerH+Wmys+jd+eMeQ0D0UZDYmOCDJqYp3iJ7qni0R7GC337KxlU5T3unBlJ5YC96BZT4fL7I0ROdBNrCpArKEPTImwG1iK3FhLvVgK3afVyWQmJOtfCZGWpTNp+sjaJ4TgsutdNN7TzUuP9XHHZ7zoxsS/jkkpiYQldoeGlJKN+7bzZssjNMR2YMWN79BNlHM1liYPuwBIjujpOlisAsMijit5k56jI4+JP2cssCHlkW2P7mNYBh9BmzrXNtCuaHc1rtgmgnWb6G7edEIJFufkG5PlRdKmDFmCxZpug94Ayy6aReXGAoK7L2dOsU7O3C7qw7up6qrkUFcllW3rAbDpDop9UyjpD1ALPJOGrBwQjPXyZv3febPuGYLxAJNSZrPggzdQbJ+C7rQQCcSpvfcA0biAy1JxzkjByDoaVEY69xPY8xYaC9FTbAjNwJW7GEfmQrqeeofQ7iai3h3ES9qxRJfjMheddqmZI4QQ2J0CuxNSMoq5zvExntx/H+bkNVyatwJITsE98n+Tna/jTzv6/9tYG6e9JYHdKZg820pRuQWbQ+UHUc4/KiA9BT3Nhu62cLF5FVu09Ty866f8w9z/wWoM/dYZ9lQKV9xP7Yt3U//Kl8i77Md4CpcNfQ6bl4x5n6On6sVBs/SOhmNHQ3PcxSzMuRyfLdmLve7wiwRjPZT6Z1DgnaQy9Sqj6khCI/v0lGRA2hbGogJSRRlzR6Y4uq8aWQKcI3VCdY8FzWUh1tBHcE0jmt+GfXYa1hIvRq7zjGsZ7t4aHUjeczrHcrg0lt/o4u9/6mXdKyEuumr8M4iaCUkoKAn1mYT6kn8H+yShXnPg8YQpKbtyP6/WJ8uRWE0fVxd/lEOPXUiK38X8Sxw4XAKnW8PhFFisYsjR45LJxwdvmXnDu62TZiJZ+/NI+ZDmZAkWAN2RjjNrPs6Zn8CZNQ9bSjlimNNs9RQbiS3t5ObrZOS42L8zysFdUZobvEyZvZT5k5ej9WcNruraxaGuSqq6Knnh0J+A5CyvQm8Fpf7plPimU+SbTNyMnbLUTW+PyYZXwpCdySRPiPjeLnp2daD5bdgmJ5Mh2VMrcPmXExIBdP/R903oGhZHHrarSxGlTQSqX6Jz98NYXNnY/GXDet3DtST/Girb1vP0/geYlDKLdGfucVmlcwqPv/9acpUzOY1XY9RrhybMOLqmwgBlYlCfxFMQQmAp9iD3dXPzgk/zp70/4rH1f+UDF33wpPsZdj9FK+6n9qXPUP/KV/CWXE3K5FtxZM0btJdSaAa+smuB5IWio/Ih3AWXjPqXIYDHmsLMjAsHfrbpDqq7d1PdvReLbiXPXUq+p4w8T9lAFl9FOV1HAlLb9BR4vIp4e2ScW6Qo7w2RvV2QkNhnnHq6royZmOEEuseCGYjR88hBnBfn4JiXjrXEg//DFQNlPUZDIi7Z8GqIrHydkimn3wlaMtnKtPk2Nr8ZpmSyhdyiE48lpaQ12MChrp0c6qqkLdRInqdsYNrocNdnQnI0K9Rn4nBpGBZBR0uCqr0RsrRfY3O56BAfZ9+O6HH72B3JdYLeFIhkbWFn7AnWVB7Cb0vnxvK7uSDvSja/bhLtC7Hsw65hB5Wnq237/bRvf3CQEiz9GXC9hac9a0pPsUHMxOyLo7stTJljI7/Uws53wuzcGMGbopGWZeCzpTEnaylzspIzwvpiPVR37eZQ/wjqKzWPYcpH0ISGJnQSZnzIUjftzXE2vp7MT7DwlgxSM3XMSILogW4ie7roW1tD75tV2AozMA8baB4LwtCI1gTQ/TZ0nxX3yoL+15yFI2MW0Z5qbL7keboPPotmOPEUXXaa7/hRmtC4Y+oX+fE7n+fhXT/jc/P/C02cfHbBWIz8723fzGN7f8UnZn2dHHfxqB9fUUZKBaTDYC32EKnsYLqcR5n1IrZGHuOSwGLyPScPFnWbl8Krf0Pr5l/SfeBpeg49i9VfRsrkW/GVXYdu8w66XyLaQ7SnjniwdUwC0nebl30pc7KW0tRbw6GuSuoDB6nu3gOA355OvqeMEt+0cUszr5zb4o0hhNPAUuAGQ5BQpV8U5ayIVHYirBrWssE7FhO9MWJVAaJVPcTq+rAWe/BcW4juteJ5XxFGTnLEUVh1dOvorvveuTFCoNtk+Y2eM14ysnSlk/pDMV5+vI/3f86HYTVp7K0eCG4Ode2iL9YNJDtkM5y5bG56jbcbngcgzZE9kHSn0DUVazSTcFAS6pME+0xCvUdHPKPRZMafxZc7yMgxiEYlTXsr0eXfsRntuGbkMnvxyoERTrtTgDDZ1vJmfzmSWtIc2dxWlixHYmgWQkGTLW91UTrVMubBaPeBZ2jd9Avc+Zf0l2EZugTL6TjSaWF2RtDdyc4Bt1fjgssdxyWFbKyNkZqpY7MnR15dFi/TMy5gekYyGWQkHqKmZw+HuioJx0MsyVs56D1I7cEYO9aHcbo1Fl3mGJjOrNl07NNTsUyy0PT6n6EpBUvoCmI1vWh+KzKaoPfFeiyFbjwrCo77DAohBoJRKSWxQAOa9egsMzMeRjPsp/0e+e3p3FTxGf6868e8WvMElxffNuh2iVgf8d5GbCnlp32ud4ubMZ47+BBr6p4iy1U47JFvRRlrKiAdBkuBCzRBtLqXj17wWX60fjcP7/oJX1rw0yHXOhyhWz1kL/4amfO/QE/Vi3TufZTm9f9Dy8afJ0dNp9yGPX3mcV+Ghj2F7MX/htCTX+bhzv1YPYVoxtgVAdeERq6nhFxPCVJKuiNt1AcOUh84yO72jVg0K5mufOJmjD3tmyjxT8NlGTygVpRjxZuCGNkOhCbQ0+wq066inCXhnZ1Yp/gRlqM3nfG2MNGDPUQP9ZBoDQGgea3YZ6RiLTv6nW4tHbvv91hUsvH1EHnFBgVlQ9+GhIIm4eDgKV9t9mTAJ6WkpzvCpCtqeeWdrfz0tQP02PYRMZMzM3yWTErdcyl0TafQNY0Uaw7SBItD0kMNe1p2sr12Bzuj77ChcTUA1kQKvsgU/NEppCYmk24vwOnS8afrycyoboHHn3xPnbHVlOr/jK/8ehLhLnorv0nhVVm4ci4gbsbY2PQqr9Y8TluokSxXAR+Y9lVmZ158XCKfzW+GiUWTQe5YCrVso3Htt3BmLyT/8p+e8frIweipyfuUREck2QnZTwgxEIxGI5Ita8MUlFmYuWjwwM5mOAZyYAxGSsmerVEOVEbJyNaZf4kDi/X4jg1pxmnf/gBSbyfz+g9j8RbSsncDRrodYdXx3lA80N6hCCFIn3M30owDEOmuonXzPbjzL8FTtBzdenrJIedmXUJl23peqnqYySlzyNScxHobiPUexlt8FZrFSW/dGnoOPUfesh+hGTZ669/E4s457YGK1mADf6r8EQ2Bg1yYt5Lryz+BRR+7+0pFGYnzKiA9pg7p6B7XquO+ugAj04FusXJz2ef5/e5v88KhP3HdpE8M6xiaxYm/4ib8FTcRbt9N597H6Dn4LN0HnsKWOpmUybfiLb12oFDykWA0Ee2lbeu9GI500mbdhcWZMaqvbTBCCPz2DPz2DGZkLCZuxkj0fxm3BOtZf/hlUuwZuCxeuiPtdIZbyHGXYNNPv8dQOX/FmoLY+zN8Gml24mqEVFHGXLw1RKIlhPuKo+XEwrs66VvVAICR7cC5JBtLiQc91XZWE9vteCdMsFey8k7HoOeVUlK1J8buLRHMExPGkhBhHCXVkJ3M4FrVsRdTi0IpWPpy8fctJk1OwReZgt1MTsfthf4kQckgfPIsKxWzysmwlmLbdyWTZlnA38T+tp3sbamkIbKL/bG3AXAaHkr8ycywKf7pZLtL0TWNUNsuDr/xHzgy55C95BvIRJTqZz9M9atfpe2CT/NG86t0RdrI85QdV47kWH0Bk+3rwkyeaSUta+xuyWK9TdS98iUMVzZ5l/14TIJRAN3fH5B2Dr00w2oTLL3GidWW/L/v6UxgmuBPG94ovJSSzW+EOVwbp2iShRkLbCessZVS0rnnr0S6DpA242NYfUXJdvXEsE1LtvHYhEenIvrXWmqGE0f6DAI1q+itew1X3kV4ii7HsA8vi3Ui0kOst4Fo72GWmSkcQPDHjV/jTpmNgQBh4Micg81XjDN7ATZ/KULTkWacQM0q4qE2HOmz8JVfP+yRbSklG5te4W/7foMhLHxs5r8zI2PxsF+7opwN51VAOtp1SI9lm3R0ylPf7qkURq9gTd1TTEtfRFnKjBEdy542lZwl/0Hmwq/Qc/A5Ovc+StPb36d5w0/wlV6Df/JtONKnAaBb3aTN/CQdlX+g5Z0fkjr9IzgyRna+M2VoloFER7nuEm6f+nmcRjJwPti5gy3Nb6AJQYYzjzxPGfmectIc2SOqNaacn8xQHLMripGdnPqnp9mIbesd51YpyvkvvDOZoMY2PXmjHKvvpW9VPZYCN+6rC9Cc43P5j4Ylm9aEKZw0+HpPgO3rItQejJGdb1BYbiFs9tEQ2k19aBd1wV00hw9ihhOIao1cdwlz066iPHUGFRnTePZBQV+PyZW3u7APkY1U08DtSz5ns2ssveZIkrVCst2FLC2+BiklHeHmgaQ7784MW+guxX94CznuVBYu+wGabiUiE1RPv57Xah4nVPsoxd4KbpnyOaakDp47AmDjmhCJBCxaPnajo2YsSN3qLyDjEQpWPIBh94/ZuTSfFcTJA1IAj+9o8Ll3W5Tm+jiFk5JrTo8EqkMRQpCWrZOSkVx/PGjNzrrX6Dv8Nt6SFTizFwBghhPIYPyM1kJbXFmkzfw43tJrCFS/RG/dGnrr38SVuxhv0RUYzqMlj8x4iFDrDqy+EizODEKt22nbdt/A87rNz0rXLB7r3cTW9HzeV/YxDGfGQPBrcWYcNwiRtfjf6a19lZ7qlwmt+z6u3CX4yq5Btw2d6yMU7+PxPb9ia8sblPpn8IFpX8FvH35ZJkU5W86rgHQsSVMS3deF5rKQU2ijcP37CRTt5C+7f8ZXF/0CuzHy7H66xUXKlNvwT76VcNtOOvc+SvfBZ+na9zj29OmkTL4Nb8kKHOnTyFr0L7Rt+y1t2+7FW7ISb+nKcZv777H6B/49O3Mpue5SGgIHaeg9xOam19nc9Dp2w0HuQHKkUlXz9D0q3pwcjTiyFs1It2P2xJDRBGKU16QpinJUpLIDPc02MApk5LpwXpKDfUYqwhi/zsKtb4cJh+Sg01OPlCDLLTZwpsQ5aHmcNw5voLG3GolEFwYF3kksy7qJUv90in1TT7j2XnVrnL/e20Plhggr73Sf9sivEII0RzZpjmwW5lwOcDQzbOd29ta/zH5XcsnOcxs+R4FnEs19dQTjAUq9ZUw5tJ6yeB6F8+YM2YZAV4KdGyJMnWsb9ujgSElpcviNrxPp3EfBFf+LzV966p3OgNAFmt96yoD0WHOW2Nm3PULV3hiNtXGmzrVRUGac8L51dySIhCSZeck6nEMJte2ia98TODJm4y29duDxRFd/CZ1RSM5lcWWROv3DycC0ZhW9DW/T1/AW9rTJOLMW4Mq9ADMeoaPyIfyTb8PivBSrtwj/pFuwePKwuHPRrW5ygYa9v+KthheZVXgdZVrOkOfUdCvekqtx5S2hp+pFeuvfJNi0AU/RcjxFl6MZx/9O1XTv4U+VP6I70saK0g+xvOiWUyZQUpTxogLS4RIQfLsFI9dJ9qX5GOvtXGz7R/4e+AZP73+A26d+/vQPLQSOjJk4MmaStfCfkkHp3kdoXPstmt/5Eb6y9+GffCtZC79K556/0FP1PNGeGlJnfBTdMr7lM3RNJ9ud7FWez2WE4n0cDlRRHzhAQ+AQh7oqcVk83DH1iwghiJsxVVbmPeRIhl0jO3mh1NOT07rj7REsOeNfokFRzkcybhLZ3YXjgkwiOzqS03I9VhxzxndkJBw02bI2TOlUC1nHJO+RUrJ7SxRdh8mzbSS8h3m69oc09tVQnjKLq0reT4l/OkXeilOueUvPNlh8uYO3Xgqxd1uUKXNGb42cz5bG7MyLydjzLFPrDpF66X/R5klPjqJ276LUP53Lim6myDeFTv9jNL31HVo2/pSsRf806PHeeS25fGHRsrFb7tK29V4CNavIXPhPuPNHp8b5qegpthEFpBarYPoCO/llFna+E2HbujC1B3RmLrLhS00GUFJKKjdFiIYlGTn6kCVQYn1NdOz8PRZ3HqnTP3JcUGt2jl5AeoThSCNlyh14S1bQU7OKcNsuEv3Zi3Wbj+zFX8foH+XUbb5BM/W+r/wT7OvYOuwBDt3qIWXyrXgKltF98Bl6ql6gt/5NvKUrcedfgsTklZrHeanqz/hs6Xx23n9T7Jsyaq9ZUcaCCkiHSQiB99ZSNHey1y4tSyd0uIzLpt/MK7WPJbPDpS864/PoNi+p095PytQ7CbVspXPvo3Ttf4LOPX/BkTkHf8Wt+CbdRPeBZ2h++z/xlV+PM2fRWV3/czIOw0VZygzKUmYMTHnqiwUQQiCl5PG9v6LcP4v5OWeePl2Z+OJNQRBgZPQHpGnJG69EW1gFpIoyRqKHepDhBNYyL8G3mrH1xHBdPHqZVE/X5jfDRKMnjo4KIYiGJZoueav+eZ4+8AA23cEnZ3+TqWnzR3yeuRfZqdob4/W/B8krNvD4R29UqH3HA/Qceo6MeZ8nvfRasmAgM+yxUibfSqTzAB2VD2Hzl+GvuOm457vaE+zeEmHmItuotu9YPVUv0rb1XnyTbiR1+ofH5ByD0VNsxA8HR7yfL0VnyVUO6g/F2b05whvPBSmebGHSDCs2h8a8i+wghq7HKc04bdt+C0InffanTkgEmRiDgPQI3eYjpeIWqLhl4DEhxLDWedp0O++f9hV+uelrPLX/t9wx9YvDOqfhTCdt5sfxFC6n68BTRDr3Y2bO4M+VP+Fg1w7mZC7llimfxWGout/KxKcW+Y2A7jm6ViG3yKAvYHJhyu3kukt4dPf/0hvtHrVzCSFwZs0l75L/pPz2ZO9mItxJ45tfp/HNbyaTH2kGnXsfwYwGRu28o+nIlKdC7yQAEjJORcocslwFAPRGu1lV/QhVXbuIm7HxbKoyRuKNIfQMx0CWT+NIQKoy7SrKmAlXdoIGjrnp+O4ow7kka7ybRF/AZNu6MBXHJO+pPxSjpzMBwKT5UbY4f8YT+35NqX86X130i9MKRgE0TXDlzS6klKx6og9pDp6pd6QCNa/SuukXeEtXkjbrrlNun7Xon3DlXkjj298l2Lz5uOfeeSWErsOCS8Zm7WiorZLDb3wdR+Zcsi/8+lnttB7pCOmxhBAUlFlYdr2LogoLVXtjvPNaGNOU2J3akOuCob+ee/n1pM/+FIbjxNqyRwPSk1dHGA/FviksL7qFDY2r2dm6bkT7Wn1FZMz7PI1ZU/jx+i9Q17OPlbYKbi/6kApGlXOGCkhHqG9NI6GNrWTnGwigpV5w57QvJReO7/0VUo7Ohe9Yht1P2oyPUHrz0xSueAB33hK6Dz1HT9XzODPnoNu8ydT3h14gHu4c9fOPFkOzMDf7UvK9ySzIPZEOWoMNvFLzOA/v+ilr6p6iIXAIUw6SVlE5J8X6S74cofmtoAviKiBVlDEhpST0djNGpgPNaaCn2IYcUTqbNvUn77lguYNYVLJ5bYgtb4Wp2hvjYOdOfrLxS+xq28D7yj/OXbO/idc2vKylQ/Gl6sn6pFVxtq07veDoWOGOfTSs+Rr29BnkXPTtYQV4QjPIW/ZDrO486l/5MtFAMsNxe3OcvTuizLrAPlA3czTFgi3Ur/4iuj2V/OU/QTtFebrRpqfYkOEEZih+2sew2gQzF9m55BonxRUWTvV2x/qaAXBmzh5ynWyiI5qc5TZB8xdcWXInee5SHt3zSwLRrmHvF0tE+Nu+3/CHyh+Q4sjksxVfYrqWgm5LJp88UrJGUSYyFZCOUKIjQnhXJzaHRmqWTmNtnFx3CStKP8iO1rfZ3PzamJ1bCIErZyF5y35A+a3P40ibTvM7P6Tx7e8R66mlp+pFwm2VY3b+0ZbrKeGOqV9kRekHKfZNoaZ7Ly8c+j/+uvvnrD/8Eq3Bw2MS4CtnhzQl8eYQluyjU3OTtUhtJFTpF0UZdVJKAs/VkWiPYGQ7J8z3Z6A7wY7+5D1SwprnghyuijNppk5jxmPcu+XrWDQLn5//A5YV3jRqGdqnzbdRPNnCWy8H6WhJnPZx4qF26ld/Ad3qIf/yn6EZw1/zqdu85F/xv0gzTv3qL2DGgqx/JYTFCvOWjv7aUTMepn71l0hEAxRc8b+DjhSOhUQ0QLB5K/Fw18CU2NMdJT2WL1WnoGzwTLpH9DVtpGndfxLpOnTyNnZGxmS67mgxNAvvn/YVIokgj+355bB+fxt7a/j5xn/irYbnuLTgRj4//wfk515E1uJ/R7d6kFLSsvFndFT+cUIPWCiKWkM6QpZiN7E1jSS6IuQUGuzcECHQneDSwhupbHuHv+27jzL/DPz2sa0XanFlUbjit7Rs+gUdO39PuG0X2Rf+P+xpUwEINm0EBI6sodPNTwSa0MjzlJLnKeXCvJXU9eznUNdOdrdtZGfretIcWdww6VMT+jUog0t0RCBmDpR8OULVIlWU0SdNSe/L9YQ3tgDgXlkwYb43N/Qn78nM0Vn7YhC7UzB9WS/PNv+U6po9LMhezo0Vd59WtvqTEUKw/AYXf76nm5cf7+XWu73o+sjeEzMRpf7VrxAPtVN0ze+xODNH3A6br5i8ZT+i7uXPUvnsLzi4624WXWbH4RzdMQEpJY1rv0W4rZL85T/Dnloxqsd/97nifY2EWncSattBtLsakCAM7OZSIJ1ERwRL7thPGXWkT8dXshKrr/ik2yU6I2gTOCAFyHYXsqL0Q/z9wO/Y0LiaRblXDLqdlJK3G57n6QMPYtcdfGr2t5icNm/g+YHffRnH5i+jt24NwebNuAuW4S2+Es2icjgoE4saIR0ha3GyfEm0ppecAgNNg652E03o3Dn1S5gywV93/+KsTDsVmkHWwq+Qt/ynRLurqXvpM/Q1vAVA7+F1tO/8Ha0bf0a0p3bM2zIaDM1CiX8alxffzvunfZmlBe+j2Ddt4It1Td1TVHXtGudWKsM1kGE35/g1UnqanUT7mfecK4qSJBOS3hfqiO7tQjgMhNPAWuod72YByeQ9uzZFSM/Sqd4fI6fAwL9oCw8c+gqNvTV8YNpXuXPal0Y9GD3C5dG47HoXLYcTbHgtNKJ9pZQ0vf09Qs1byL34uzjSp592O9x5F5K16J/ZsWsWVkuEOUtGf3R0IOHS/M8Pms11tIQ79tG49ls0rftPug8+DdLEW3oNGfO+gCt3MeHgJmB0RkhPJh7uRCZiaIYDb+mKU5bCm+gjpEdcUnADpf4ZPLX/t3SEmk94vi/Wwx92/BdP7LuXMv8MvnrBL44LRo8lNAv+ipvIXpJcTxyoWUXjW98mUPMKUuXuUCYQFZCOkO63ofltxKoD2J0aV93qpqA0WcYk3ZnDdeWfYH/nNt5qeO6stclbdDkl1z2M4cyk7uXP0brlV6TP/jQpUz9ALNhC8zs/pKPyTyQiPWetTWfKZjioSJ3LnKyLgeQaiY5QC32xnoGf93dsoyfSQcI8/alYytiJNx4p+XL8jaaebsfsjiJjaq2wopwpGTcJPFdL9EA3jouyiTeHsE9LmRDrRgHWPJv8HnC4BNMWwcGM3/KXvT8k05nHVxb9nHnZl455G8qnW5kyx8rGNWGa6oa/nq5j15/o3v8k6bM/jbd0xRm3I+y8jfbQYgo9vyfc8PwZH+9YRxMuXUPazE+O6rETsT7atj9IsGUbALrdj8WdQ8qUO8ld+n2yFv0zvtKV2FMrSJ16J9mXJ8vcJDojSDNGx+6/EOtrGtU2mfEQrVt+Sdv2+4a1vYyZmIHYORGQakLjzqlfAuAvu39+3ADHgc4d/OSdL7K7fSPXlX+CT87+Bh7rqddbG4400mZ8hKwL/hWrt5Cu/U/QvvOPY/USFGXEzqspu0KI64DrysvLx/Q81iIP4Z0dyLiJxZqM6Y8U9F6cezU7W9fx7IHfU5Eyh0xX/pi2ZaBNviKK3/cnmt76Hm1b7yXUsp3cS/8LZ+YceqpeJFD3OsGWLXhLrsZTeBniHKsFatFt3FjxqYE1FfWBg6ypexoATQjcFj8emx+vNRWPLSX5tzUFny0VXTuvPubnjHhTCOE00DzHf9aOZNqNt4ePW1+qKMrIyLhJ4NlaYjUBXMty0VPtmF1RbDPOLCHQaOloSVCzP0ZGjs7UK1t4rObHtAUPc3nRbVxV8v6z+t18ybXJBEcvP97LnZ/1YbGePGDvrX+Dlg0/xlN0Belz/+GMzy+lZN0rIZwuQcWkGhrXPo7VW4AjY9YZH/v4hEvfOuOp2rG+ZkKtOxCagadwGZrhIB5swYz1AWBxZpIx5zND7m9xp6J5LSQ6o8QChwk2bcSZORuLKxspzVOOZJ6KlCbtO/9APNhKyuTbh7VPomvsSr6MhVRHJjdW3M1fd/+cNXVPsTT/el6ufpjV1Y+S7szl8wu+Tr6nbMTHtXryyZj7OcIde9GMscnwrCin47waIZVSPiOlvNvn843peSzFbkiYxOr7MBOSt18OcrAyOfVBCMHtUz+PoVl5ePfPzuronWY4yFn6PbKXfINg0waqnr6DSNfB5HSNC/8de8pkug88TdPb3yfYsm3CJLwYiSMX2mLfVK6f9EkuKbieWRkXke7MJZqIcKirkncOr2JV9SP8bd9vaA0eBuBwbxVv1D1NOJ7srQ/Hg0QSah3jWDqSYffdN0d6en/CC7WOVDkPSCmJ7O5EJs7+92nvS/XEanpxXZ6HfVYa4coOAGzTxzcgDXQnaKyNsW51EMMmSb/yVe7f8y9E4iHunvMdVpZ9+Kx3FNrsGlfc7KKr3WTtSyevkRnpOkTDa/+KLaWC3KXfO+MACqDuUJyGqjgLLnVQdMV/YTgyqFv9pWGPHEopkYMsBTo+4dLPR5RwaeDYZoJw53669v2NxrXfoent79J94EnCHXsAEEIje/HXcOctGfYxj5R+sfqKyL34O9hSpwDQvf8pWjbfQ7hz/4jbeUT3gacJt+0kZfKtw14nm+iYuCVfhrIgeznT0y/g+YN/5J5N/8Kq6kdYkLOcLy34yWkFo8eyp07G6i0cpZYqyplTQ0enwZLnAkMjVh3AWuzB7tKwHNPp5rOlccvkf+BPlT/k1ZrHuKLkjrPWNiFE8ks6bSoNr36V6uc+StaifyFlyh2kz7mbcPseuvY9TqD6pVHpmR0vQggynLlkOHNPeC4cDxKIdtIT6STFkUxA0RftoT5wkMW5yWlX21veYkfr29gNxzGjqin9o6pppDqyMM6xUeSJJt4UxD7txBtjPV3VIlXOH5HKTtp/uoOUT0/FuWjkCW/OhGNhBtYyL7bJ/oG2GDlOjNTRX584Enu2RGlqiHHwYAfxi37HC3VbmJ6+iNunfAGXdfzWthaUWphzoY2tb0dAwsUrnBiW4zvM4uEu6lZ9HmHYKLj856OS/EVKybpVQdw+jRkLbeiGnYIr/pfqv3+I+tVfpOia358wWiXNOF37nsCeNhVHxkxiPbU0b/gh6XM+iyN9Wn8A+SSBmlXEQ62kTf8IfYffRjMc/X/sCMOJ1VeEbnEhzVhyJpeWzFhrxoKE23cTattJuK0SMx5MJiRKnYS7cBmO9OlnlKFXT7ENJK879j3UHanEGt+hddPPsfnL8RRfhT1t6rBHdfsOryNQswp3/iW485cOuz1Ha5CeGyOkkLzPuXXK5/jx+s/TGmzgg9P/iblZl4x3sxRlTKiA9DQIQ8OS7yJaHcApJXMHSU4wJ2spO1vX8VL1X5iSvuCMe7NGypE+nZLr/0rD6/9G87r/JNSyjZwl/4E9bQpZF3wNM9aLEIJENEBP1Yt4S65Gt3rOahvHit1wYjecZDjzBh6blDqbSamzB34u9k3BYTjp6Q9cW/rqONRVOTBqLIQgxZ7B9eV3oWs6vdFubLoDy1mu53auMkNxzK4oRs6JN3O639Zfi1QlNlLOfbZpKRhZDnpfqMOxMGPMM9ua4QTR/d3YZ6ZiZDgwMpKBjIwmiOzrxrUsZ0zPP5RoRGKaErtDY+YFNnY+tZnGRfcitCA3Tfo0S/KumRBZf5dc5QQh2PpWmMPVca6+3UVaVvJWSJoxGl79KvG+JgpXPojFPTrvZfXeGM31CZbf4EQ3ku+BLaWc3GX/Q/2qL3D4jf8gb9kPB96feLiL9h0PEO2uQrO6cWTMRLN58JasxHAmM/hLKQk2rife14gz5wJifY1EuqtAHj8rK2PeF9BTKwi1bKd95+/IXvx1LO5seqpeIFD7CprFhSNjJvb0mdjTppzWCOtg9BQbkX3dJzzuKbgUd+4SehvWEqhZTdvWX2H1FuEtuRp7+syTfkYiXYfo2P0XbCmT8VfcPKL2JLqiA+06l3isfr608KdoQj/j2ryKMpGpgPQ02aamEG8OQUKCIUjEJeGQPK7I9c2TP8Ohrkoe3vUTvrjgJ1j1s/tFqNt8FFx5D+3b7qd1yy+JdOwhb/lPsPlK0G3Jac2Rzv30NazFnXfxeROQDkemK/+E9b0JM0FvrIuucBvtoUZC8T50LVlAe239swTjAW6q+DQAVV27sBsu0h3ZWM7y/+u5IN6czGb57oRG0F+LNEXVIlXOD0ITuFcU0PWHfUR2dw06K2A0hXe0E1rfgiXfddzNdWRfN8RM7OMwXbezNcGmN0O4PRoLllt4cvdDVGY+hZd8PrXwu+S4i896m4aiG4KlK50Ulhu8/Hgff723h6UrncxYaKNp/f8QbNpAztLv4cycfeqDDYM0JetWh/ClakyZe/y1wlNwKZkLvkzLxp/Qtq2cjDmfIdy5n/btDyLNKGkzP4kzay4Ahj0VX9m1A/sGG9cT6dxH+pxPkzH3c8lzSYlMRJGJMGYshBkPYXFlAWBx5+Ervx7dlhyhduVfjCNzDlZf8ahMSX43PcWGDMYxIwk0m37cc0K34Clchjv/Yvoa3yFQ/RJt2+7D4srFW3JVf7m649sUD3fQtu23GPYU0mZ9AjHCKd+JjgjCoaM5zr3bXr89fbyboChj7tz7zZwgbJN82CYdXav6zqsh4jHJ0muO1txyWjzcMfUL3L/t2/xhx3/y8VlfP+vTQIXQSJ9zN/aMmRx+/V+pfvr95Cz9Dt7iq5JtzJqHLaUC3eoGoGv/k1i9hTgy506I3uyzSdd0fLY0fLY0inyTj3tuZsaFxMxkD6uUkrX1zxJJhBFC4LWmkubIJt2ZQ7ojh1RHNjZ9fKfMjbeBki/ZgydN0NPtasquct5wXphFz5PV9L5YN+YBqWN+BtZizwkjPZHKTjAE1gr/mJ7/3RqqY2x9K4zdIciY2sE9m35CfeAA/pbL+PIN/4DLMTG/C4smWfnAPyaD0teeCXJgSxWl+gvkzP44/vLrR+08ByqjtDUluOpW16A1UFNnfJRI1wHatvwKMxYi2l2F4cwgfdYXsbizBz3mQMKl4itJn3M04ZIQAmHYwLANdDofYXFnH3c8izMTTqOm6nAd+XyaXRG0rMGnPQvNwJ23BFfOBQSbN9NT/SLtO39Plisbq+doh7EZj9C29T6QCdLnfBrdMvLapudKyRdFea9SAekZkAlJoiOMkeEgI9dg95YIwV4Tp/toz97ktHncOuVzPLrnHh7a+T98ZMa/jsvaRHfehZRc/wj1r32Vhlf/idD0D5O54EsIzTIQjJrxEKG2nQRqVmFx5+EtWYkjc/Z7LjAdTK6nZODfQghumfwPtIUaaQs10h5qorl/yu8RXlsK09IXMT19EQCRRPg9FaTGG4OgMTCd8N2MNPtAAhZFOdcJi4b78jx6nqgiVteLpcA9qsdPBGL0ra7HdXk+uscy6O9VeGcHtgrfCaNRY0VKyf4dUfZuj5KaoWNMXceDB34DUidjx+dZsWjphA1Gj3C6Na7/sJt3XtzLhrdSabP+kZXLT8xLcLrMRDKzbmqGzqSZgy/3EEKQuehf6WvcQEflH/BPvp2shV8eMgPqcQmXLv7umIxujgY9Nfl6E50RjCEC0iOEpuPKWYgzewHRroMDwWjX/qcwnJk4s+djcWXhK78ei2vwIP1UVECqKBObCkjPQHBdM+EtbaR+ehq5RcmAtLE2Ttm04y88F+ReRdyM87d99/Lnyh/zwen/PDAV9GyyuLMpXvl7mjf8mI7KPxJq3Unesh8MTOnRDAfZi/892VN56Hnad9yPxZ2Pr+yaU67teK9xWNwUWCZR4J008Fgo3kd7sJG2UBPtocMYItnxEIwFeHjXz7i08AbKU2bRGjzMKzWPogkdTejowkATGprQ+v+towmN6RkXkOsuoSfSwY7Wt5mevgi/PYOOUAvV3bvQhI5Fs5LvLcdnO/3kE2Mh3hRCT3cgLIPfLOlpNsyuZC3SobZRlHOJa1kOgWdrCbxQR+qnpo7acRM9UXqeqEKGE5h9MXTPiR2aiY4I8cNBnBed3s36iNsUl7z1dh17W3YSK9hHpXUPbQcOU+qbjnfbp4hFU5m+4Ny4+Y8FavF33MVFUxayo+Ub/O33fSy8JMGiyxxog4xojsTe7VG62kxW3ulGG6IubKyvmfbt92P1FSPjIXprXyVj9t2DBqTHJ1z6xagkXBorR4K/REd02PsIIbClJMv2STNBtH9NrKZbSZv58TNqT6IzkkxIqSjKhKQC0jNgm+rHkuMEkexp9adqgwakABflX0PCjPH0gQfQd/+M90/7Epo4+0Gp0C1kL/4ajsxZNK79FlVP30Hesh/gykmO5Amh4cpegDNzLsHmTfQcep62bfdh9RTiLb0Ge/p0FZgOwWG4yPeWk+89vg6uEBoLcpbjtyWTUVh1GznuYkxpYpoJTJkgIY/+HTOjmNIkbiYLuIfjQWp79lHmnwlAV6SVLc1vHD3B4ZdId2RTmjKDUt/0cc1ieUSsKYglZ+gaZwOZdjsiGFmqFppy7tNcFlyX5tC7qp74TSUY6Wc+OpjojNDztypkzMR7U/GQI03hXZ0AY7Z+VEpJa7CBQ107OdBRyd6WSkKiDVLAIVwUO6dxScH1ZPVexnNVIS67wX5C9tqJKBHpoW715xFCY/p1X2W61c+a5/rY8HqYukNxrr7NhTfl9K7TibjknVdDZOTqlE0belaUGevDjAXJWvBlAKqf+wh1r3yRohUPohlHg/qBhEvBZopWPDDkdN6JQvP3B6Sdp5e8Tmg6GfO/eEKSptMh4yZmd1SNkCrKBKYC0jNgpNrhmPT6OYUWdm+NEOozcbhOHPW5pPAG4maM5w49hKEZ3Dbl82jjNN3GV3oN9tTJ1L/yFWpfvJuMeZ8nbebHB6b/JKfQLMKZNZ9g0wa6Dz1P27Z7sfknkTH/CyooHQGH4WJ25kUDP/tsaVxScMOw98905fP+aV8e+LnUP50S3zRMaRKO91HVvWug/uo7h1eR7Sqk1D+ditS54zISL01JvDl00rV0R27W4+1hFZAq5w3XlXn0rm6g9+V6/O8vP/UOJ5HoiND9RBVIiffmkiGnvwNEdnag+awY+aMzAmTKBI291RzqSn63HOqqpC+WzJhqNX34o1NYnH8Dc0tnke0qQhMa0pT85dc9+FI1ps6d+Df+0ozT8Pq/EO2pp2jFfQPTRK+4yU1hWYRXnw7y8K96uOx6JxUzR/56dm2O0NNpct373CdcL6U0CbfvwZE+DZu/lJyLvoXQk0Fr7tL/pOHVr9D41rfJXfr9gX2PJFzKXfp9HKOUcGksaTYd4TJOOyCF/rrj4sxvUxM9UZDnXoZdRXkvUQHpGUp0RohWBbDPTSO70GD31uS03dKpg68XWV58KzEzysvVf8HQLNxc8Q/jFtzZ/GUUX/dnGt/8Fq2bfk6oZRu5S783kIUP+gPT3MU4sxfQ17gBMx5ECIGUkmh3FVZfiQpOx4EQAl3ouKxeZmQsZkbGYnoiHRzqquRg1062taxlStp8AFqDDfhs6Wcty3OiIwIxc9CSL0foaf0jpCrTrnIeMVLtOC7IJPhGI97ritDcp58vILSpFeIm3ttLT1pXVJqS8K5OHLPTTvu7OG7GqA8cGAhAq7t2EU4kE5Ol2DOZkjYv2RHmnc7h7akUz7XiTzu+s+vArhhtTQmuvGXw5D0TTcvGn9LX8BbZF30TZ/aC456rmGUjq8DgpUd7efGRPuoOxLjkWhcW6/BeVzwm2fBaiJxCg6JJJ34GeuvfoGvvo2Qt/GesvqKBYBTAW3wF0bn/SOuWe7D5y0mf9Qk6dv+Frj2PkDbzE/jKrzuzF34W6Sm2MwpIR0uio78Gaaoq26YoE5UKSM9QrKGP4JuNWArduNPteFO0kwakAFeVvJ+4GePV2scxhIXrJ901bkGdbnGRt+wHdO6aTfOGn1D1zJ3kLv3+QKr5I5LZ8C4c+DnSsZfWLfeQNvMTOLPmne1mK4Pw2lKZk7WU2ZkXE04kOw5MmeDlqr+S7S5iedEtACTMOPoIU+aPRLzx5Bl2ob+nWlMBqXL+8VydT+itZvpeO4znfUWnfRwZTaB5LCcNRgFi1QFkXxzbCKbrRhMRarr3DIx+1vbsHcginuksYE7WJZT6p1Pqn4bfnkHtgRgZfh2HUyNzyYnHM03J+tVBUjN0KmZN/Jv+rn1P0FH5R1KmfZCUilsG3caXonPzJ72882qIjWvCHK6Jc/XtbjJzT/3duWNDhL6A5KrbHMdd26WZQGg67rwl6FYfVt/gn4+02Z8i0nWA1k0/x4wFaN/xe9wFl5Ix7/On94LHyUQJSM3Oc7MGqaK8l6iA9AxZipO1O2PVAYx0OzmFBnu3RQkFTRzOwafjCiG4puwjxM0ob9Q/g6FZuabsI+MWlAohSJ3+IezpM2h4/V+oee6j+CtuJnPBl09IHX+ELaWMlKkfxJGRXNcYat2O0O3YUyvOZtOVQQghcBjJqXsCjSuKbx8IQHsiHTy5/7cUeSdT5p9Bjrtk1Kf1Hi35MvQIqdCTtUjjqvSLcp6x5LuxzUyld3UD7qvyEdbT+/2ScQnDGGkMV3aCYFgBaSQe4qGd/8P+zm2YMoFAI9dTwuK8FQNLAdzW47/zw0GTyo1hiiusTJ03+A39/h1ROk+RvGeiCDZtovHt7+HKW0LWwq+edFtdF1x4hZOCUgsvPdbLo/f1sOQqJ3MW2xBDvM5oRLJpTYiCMoP8kqMjn70NbxOoXU3Wgq+gWZw4s+YMeV4hBDkXf4dooI727Q9g85eTe+l/I8ZhCcaZ0FNsxGoC492MgaBY96uAVFEmKhWQniHdbUFPdxCtCeBYkEFukYW925J1xwpKh14fKoTg+kl3EZfJkVKLbuWqkvefxZafyJk1h7KbnqR1y6/o2PV/BGpeJXPRV/GVXXdCsCw0y3Ejpj1VLxHtqcaWMimZ/Chl0rsPr4wDIQSZrvzjHivxTaOmew8HOndgN5wU+6ZS6p9OtqvwtDtFYokooXgf4Xgfifo2hNNA81jY1PgqXlsqk1KTa55WVf0Vq2HHbfGR7jOIt3ShhVtxWbxYztKUYuXcIYRYAfwc0IH7pZT//a7ni4AHgQygA/iQlLK+/7n/Aa7t3/S7Usq/nq12e1YU0PbDbQTfasa17PTKiMiEROinzjEQqezAUuhG95x6ZHJ1zaPs7djMpQU3Up46m2LflIHOq3dLxCWaDnanxsUrnLi9Q7el9mAMh0ucNHnPRBAN1FP/ylewevLJu/QHiGHOFMkvtfD+z/lY/WQfbz4fpPZAjCtvdh1X4u2IbevChPokiy9PdshJM0bn3sfpa3gTW8pkpDSHdU7NsJN/+c9p2/ob0mZ+4rRqb443PcWK2RMb92zqic4IwqohXOqWV1Emqgn/2ymEKAX+H+CTUt463u0ZjLXYTWhTG2Ykgdurs/wGFy7Pqb98hRDcVPEZ4maMl6oexhAWlheP70vULE6yFv0TvvLraHrruzS+8XW69z9F9oVfx+YvGXK/zPlfpLdhLYHql2jd9HNsKZPxlV2LzV96FluvnIrXlsrSgutYkreS+sBBDnVVcqBzO3vaN+GyeCjxT6PUP4N0Rw5CCLoj7cTNGGmOZEbHbS1r6Yl0DASf4XgfoXiQuBkbOMdFdYtIzc5ACEF970EyzQiTmE3CTBCK99EaOkwo3scM2yRS61J4ce+9ANh0Oy6rl0kps5mRsRgpJQe7dpDpzMdrSx2X90sZP0IIHfglcCVQD2wQQjwtpdx1zGY/Ah6SUv5BCLEc+C/gw0KIa4F5wBzABrwmhHheStlzNtpunezDUuIh8GI9zktyhhxNOxn7VD9SnnwbMxgnerAH94qCUx6vNXiY12ufZH72ZVw36RMn3TbYa/LOqyGKJlkomWLF4z/5yFxbY4KMXGNC5xNIxPqoX/UFpExQcPn/HpcrYTgcLo1rP+BmxzsR3nwhyMO/7ObKW9wUlh8NwiMhky1vhimZbCG7wCAe7qR9+wNEe6rxFF2Jr+x9IxrltDgzyVnyHyNq50QyUPqlK3LSpFxjLdEZQUuxTejPp6K8141pQCqEeBB4H9AipZxxzOMn7fU+lpTyEPBJIcRjY9nWM2Ep9hDa2EqsthfbJN+wgtEjNKFx25R/JG7Gk9l3dcspM7BKKdm4JkwiJll8xdjUIbOnTqbo2ofo2vs4LZt+xqGnbiFt5idIn3UXmnHimiahW/AULsOddxG99W/SU/0SLRt/gi11Cr7Sa08azCpnn64ZFPkmU+SbTCwRpbZnH4e6drK7bSN72zfzgelfxRAW1jW8SCQR4vpJnwSgpnsPfbEeHIYbu+HEZ0vDYbiwG67+v51onS1YZiY/lzdMuuuYc+oDN8IJM0HHgf1E9zSzLPdGemUPfbEe+qLdGFryBi+cCPJ67VMszrua6bZFdIXbeO7gQ7itPlwWDy6LF5fVh8eaQpojG7fFp244zi+LgAP91wCEEH8BbgCODUinAV/p//erwJPHPL5GShkH4kKI7cAK4JGz0G6EEHhWFNDx612EN7fhWJAx4mPYpg5jCu6eLjDBPuPkHTZSSp7a/1sMzcK1ZR896bYdrQk2vhbCNMHtO/W1LBGXdLQmBk3eM1FIM8Hh179GpLuKwqt+PeTazVMRQjDrAju5RQYvPtrHU38IMPciOxde4UA3BFvWhomEJRdc7iDcsY/2Hb9DmlHSZt510im656uBgLRz/ANStX5UUSa2sR4h/T1wD/DQkQeG6vUmGZz+17v2/4SUsmWM23jGjGwnwqoTqw5gm+QjEZdsfTtMRq5BYdmpL9Ka0Llz6peImzGe3v8AhrCwJP+aIbff+HqYdatDABRPtpJdMDb/jUJopEy5DU/Rcprf+RHt2+6j59DzZF/4/3DnDZLZgv7AtOgyXHlL6G14k0D1Klo2/hh72lRSZ3zsnJx2dL6z6FbKUmZQljKDcDxIR7h5ICicl73suG2vK//ESYM+MxSnsefwSdePQjI4dWb6icpmiijHyDzxZsWm27l1ymex6skOEE3oFHor6I110xVppyFwiNgxI7N2w0GaI5s5WZeQ7SocWCOngtRzVh5Qd8zP9cAF79pmG3AzyQ7OmwCPECKt//FvCiF+DDiByzg+kB1z9nnp6Jl2Ai/UYZ+fPuLPodkbA0NDsw89ohap7EDYdKxlJx/t29W+gT3tm3hf+cdPOtugoTrG1rfCOFwai5Y5hhWQdrQmMBOQnjNx1ze2br6H3rrXyVr877hyF5/x8dKzDe74jJc3XwiyZW2YhqoYl1zrZOvbYcpnWLBFX6e18kkMZybps7+IxTWxa4aOFT31SEAaHdd2JDojWCv849oGRVFObkwDUinlGiFE8bseHrTXW0r5XyRHU885QhNYitxEa3qRUqIbgkhIEo+eYr7VMXRN54PTv8pDO+M8se9eDM3CotwrT9huxzvJYLRippX6qhhrXwxy8yc9Y3rTbTjSyLv0v/BPuoGmt79P3UufwVuygsxF/4zFOXjPv2bY8BZdjjvvYnrrXyfcvhfNSAYpsWArhmPkN2jK2LMbTnLdR0ezM5zHr3871f9ZvDnZUXKqgBTASO+/WWkLD9p7rgkdny1t4GevLYWLC45+RUgpiZoRusNttIebaQ810h5sRJBsY23Pft6oe5rryj+O355BINpFwozhtaWNW/1fZdT9E3CPEOJjwBqgAUhIKV8SQiwE3gJagbeBxLt3FkLcDdwNUFhYOKoNE5rAfVUB3X/aT3RfN7bJ/hHt3/14FUaWA88Q03GllIR3dmKb6kcYQ3+eY4koT++7n0xnAUvzBy8ZIqVk/44oe7dHScvUWXCpA6tteN/PbY3JtzUjZ2KuAOo++HfadzyAf/JtpEy5Y9SOa1gEy65zUVBmYfWTfTz22wBCwOS85+nevwZH5hxSp31o0BlF7xXHjpCOF2lKEl1R9JSJn/1ZUd7LxuMKMpxe7wH9vd3fB+YKIf6tP3AdbLsxu7EYDmuxh+j+7oGb6yVXjXwqraFZ+MiMf+V327/Ho3vuwdAsx41Q7dse4bW/BymZbOGKW1zs2hThtWeCVO+NUTJl7L9sXbmLKbnhMdp3/o727ffTW/8mGfO/QMrk24ZcF6MZNrzFV+EtvgoAMxak5Z0f4sq9EH/FTWPeZuXsGk7JlyOO1CKNt4U5nclUQghsup1MV/4JiZsA3BYvZf4ZuK1+AHa3bWRH69sYmoU0RxZpjhzSHNmkObLx2zJGPduwcsYagGOjsfz+xwZIKQ+THCFFCOEGbpFSdvU/932S1w6EEH8G9r37BFLK+4D7ABYsWDD8HsRhcl2UReCpagIv1I04IHUuyTrp6GiiJUSiLYz76hM/+8d6vfZvtIebuHvOdwYt95SIS7atC9NQHaeg1MLMC2wjqiPa2hjHYgVf6sTr5Am1bKNx7bdwZi8ke/HXxqQTtGyalaw8g1ef6SMtUyclU0e33oCn6Ir3fKer5jAQdn1cA1IzEIOEVFN2FWWCm5hdmseQUrYDnxnGdmN6Y3EqliIPwmFg9sSgf7RHSkksyrB7miEZlH505r/zwLbv8PCun6FrBrMzL6Zmf5SXH+8jt8hgxR1udF0wbb6NrW+FeeulZPIJ7SwUI9cMGxlzPoOvZCVN675P87r/pPvA02Rf+HUc6dNOub/QrfgrbsbiSd5nxvqaiXQdwJWzCKFN3DVIyvDEm4KgMegU3HfTU2wgIDFGpV/SnbmkHzPCOyVtHin2DNpDTbSHmtjfsZVd/VN+daGT6sgizZHNkrxrEEIQS0QxNMt7/qZyHG0AJgkhSkgGoncCHzh2AyFEOtAhk6lL/41kxt0jS0P8Usp2IcQsYBbw0tlsPICw6rguzyPwZDWx+j4s+cNfsmCbNHjJrSPClZ0A2KcPPQW3M9zK6ppHmZmxhIrUOSc8HwmZbHg9TGdbgqlzbJRNH/nnvbUpQVqWMeHKvcR6m6h75UsYzkzyLvvxmF5fLLKaq2+wYvXkA4PXNX2vGu9apAMlX1JVQKooE9l4BKSn7PUeL+3NcdxeDZtj5D29mtMg5a4px13M160KIQQjTjxk1W18YtbX+e22b/F/lT+mp11j56NTSM3Ued8H3RiW5Dl0XXDhlU6e/0svu7dGmT7/7H3hWn1FFFz1G3oOPUfzOz+k+u8fIHXqB0if97mTrhMVmnHcGp5g00Z6qp6n5+BzuIuW4867GM1QF45zVawxiJ7hOOkUwiOEoSVrkbadnZsVry01WYKGZAkaKSU90Q7ago39QWry7yO/w6trHiWWiAwkYlpb/xwxMzKQvCn5txuH4RxI6mSoTpVRI6WMCyH+EXiRZI6BB6WUlUKI7wAbpZRPA8uA/xJCSJJTdj/Xv7sFeKP//7KHZDmY+Nl+DQCuy3Lpfa6W3hfrSPnklGHvF2sKorks6J7BP1ORnZ3oGXaMrKE7f57Z/wCQXPs96DliyTqjCy6xk1M48s+ulJK2xgSTZ0+s6ZBmLEjd6i8g42EKVtyPYfeP2bmkmaCj8k/oNh8Z87+gOrDeZdwD0o7IQDsURZm4xiMgPWWv9+kSQlwHXFdeXj7ifRNxyd//rxdNg/d90ENKxsin7wkhkP15+oUQ+NN1Du6KEo3IEY2SAtgMB3fN/ia/fOc/eLrhRxTnfokb7rwYm/34G/2yaRayC3TWrw5SMdOKxXr2LoZCCHxl1+LOX0rLpp/Tsev/6Kl+iawLvoan6PJhXZi9pddg85fRU/0S3fv/RqD6Jdz5l+AuXKYSIJ2D4s0hLMOYrnuEnmYfsxHSUxFC4LOl4bOlUZYy44TnJ6XMRnK0ZmAoFqAz3Eoo3ntcMqVj5XvKuLo0+XX2Vv1zpDqymJI2H4Dq7t3YdOdAhuDBpk8qx5NSPgc8967HvnHMvx8DTsjALqUMk8y0O+50twXn0hz6XjuM96aSYY/U9Dx2CPvcdFwXnZgQR8ZNInu6cF6YOeT++zu2sb31La4u+QCpjuO36+5I4E3RcHs1LrvehW6c3nWjp8skGpGkZ0+c6e5Smhx+8z+IdOyl4Ip7sPnLxvBcEqHppM/5NJrFqYLRQegpNmK7+sbt/AMjpCogVZQJbazLvjxMsgc7XQhRD3xTSvnAYL3eo3E+KeUzwDMLFiz41Ej31Q3BFTe7eO7hXh65r4cVt7somjSyXt9EIEbP44dwXpiFbbKf3CKDA5VRmurix9UqG65Ijw3PW1+mc8r/UFf2CxqivoHRnSOEECy5yskTDwTY9naYBZee/dTqus1LzpL/wF9+PY1vf4+GV7+CO/8Sshb/G1ZP3kn3FUJgT5uCPW0Kke4qAlUv01P1PIHaV3DnXYS7cPmY9m4ro0eaknhzCPv0U5erOEJPtxPd2zV2jToD7w5Sryg5mhAlbsYGarGG4kHC8V5C8SBOwz2wTUe4BZuR/H1MmHFWVx8fNzkMJy6rD5fFi7u/hE2Ws4BMVz5SSiRSJV86T7ivyqfv1QZ6V9Xju/3UAZKUEkyJGGIZRvRgDzKSwDbEdN2EGedv++4j1Z7FssKbj3uusy3B2heCzLzARtEk62kHozAxExq1bf0NgeqXyVz4VdwFS8fkHFJKAtUvkYj04J98KxZX1pic53ygp1oxu6LIxNCf57GU6IyAIdDcavaKokxkY51l9/1DPH5Cr/dEkFds4Y7PePn7//XyzB97uehqB3OW2Ifd66m5DIwsB5oz+bZ6UzScbo3G2tiIA9K+gMmTvw8goi4+M+87/KXuGzy4/XvcPefblPiP7/jPK7ZQMtnCpjfCTF9gw+Ean5tYR+ZsSq57mI7dD9O6+R4O/e0m0ud8hrQZHx7W+h2brwTbnLuJ9R6mp3oVgdrXCNStIW36h3BmLzgLr0A5E4n2MMTMYWXYPcJItxFaF0HGzWFN850oDM2Cx+rH058waTDvK//YwL+F0Lh58mcIxgIEYwF6Y930RZO1V7sj7RzurSKWiDI78yIyXflEzQh/rvwRi/NWMjVtPsFYgJ2t604IYO26GpU5FxjpdhwLM+l7vRHP+4oGrhFDSvTPtNEH/50I7+wADWxT/YM+v7b+WVqCdXx85v/Doh/fsepP05g230Ze8ZnfoLc2xhEC0rImxghpT9VLtG39Nb7yG0id/pExOYeUkq59j9Nb9xrOrAUgTRAT4/VPRHqKDSSY3dFxWceZ6Iyg+22ICbbGWVGU402cbs1RcCZTdo/wpujc+ikvq57o5c0XQrQ1J7jsOtfAus2Tnl8TeFYezfArhCCn0ODQ7iiRsHnCdNuhREImT/0hQKjP5MaPecnONfh0+nf41eZ/5/5t3+buOd+hyDf5uH2WXOXkz/d0s/H1EEuvGb+prkIzSJv+YbxFV9K8/n9o3fQzeg4+Q/aFX8eZPX9Yx7C4c0mb8RF8ZdcSqFmF1ZcsQxLrPYyUZn/iCGWiiTf1l3zJGX5AqqfZQY5/4fSxpgmNFHsGKfbByyQBRBJh6J/yj5TMyryINHty5KU32s2u9g0kzOMrl+iajt+WwY0VI54Uopxl7qvzCa1voe+1w3iuOXkmeNkfkDLEiFKkshNrmRfNceIlvCfSyYtVf2Zy6jympS867jkzIdF0QenU0Vnz2daYICVdH9b1cayF2nZx+I2v48icQ/aS/xiTjhppxuio/CPB5s24Cy/DP+lm1SF0Crr/aOmXcQtIVckXRZnwzp0hiWGQUj4jpbzb5zt5dsJTsdoEK+9ws+gyO3u2RPnb73roC5in3rGf2RvDDCbzZ+SXGggB61eHiIRPfYxYVPLMn3rpbEtwzfs9ZBckbzg81hQ+M/d7uK0+frvtW9QHDh63X2qmztR5Nra/E6G784Rye2edxZ1N/uU/Jf/yX5CIBal5/uPUrfoC4c79wz6G4UgjZcodGI5kLcrug8/SuvmXSHNccpMopxBvGn7JlyOOlH5JtI9f0ouJwqbbB6b42gwH87MvGyhnk+nK56Mz/o0PTP8KN0y6iyuKb2dx3tVMS1tEjrtoPJutDJO1yINtWgq9qxqQsVNcC46MkA4ynTbREyVW0zvkdN3nDv6BuBnjxopPHRcshfpMXn68j6a60fv+bG1KkJ4z/qODsWAr9au/gG5PIX/5T9H00Q9AzHiY1i33EmzejK/8RhWMDtORIHS8EhslOqNq/aiinAPOq4B0NAlNcMFyJyvvcNPWlOCv93bT3HDqC7kZSdD5u72Et7cD4PXrLFzmoLfH5K2XQoT6hr4RSSQkz/+1l8a6OFfd6j5hmq/PlsZn5n4fh+Hivi3foOFdQekFyx1oWjK770ThKVxG2U1/I2P+Fwg2b6LqyVs5vObfiQbqR3ys1GkfJH32pxCagTQTtO/4PaHWnQOJpJTxFW8KIpzGiNbqGOlHa5EqJyeEwGG4SHfmUOSbzPT0RSzKvYILcq8a76Ypw+ReWYDZHSW4rvmk28lE/3VikBHSyK7+ci8zTlyrXd29m41Nr3BJwQ1kOI9fv1+9L0YsKvGmjM5lPxQ06e02yRjngNSMh6lf/UUS0QAFV/xioANzNCUiPbRs+jmRzv2kTvsQ3mJVY3S4tJTxC0illP0jpCogVZSJTgWkp1A+w8qtn/KiCcHj9/ewb/vJv1Q1m46R5SBaHRh4LDPX4ILlDsIhyeY3w4MGUNKUrHqij5p9MS673smkGYP38KbYM/jM3O9h0a38dMOX+eWmr7G2/lkC0S7cXo05F9rZtz1Ky+GJM4qoWZykz7qL8lufI23Gx+ipfpmDT1xP07r/Jh5qH9FxbP5SAOLhdiJdB2nbdi9Nb32XQN0azLgaZRtPscYglpyRrWnUU8e2FqmiTCS2qX4shW56X6xHmifpSIsPvYY0XNmJ5jawFHmOe9yUCf627z681lSuKL79+MPFJDX7YmQXGjjdo3PZb2tKzsRJH8eERlJKGtd+m3DbTvIu+S/sqZNPvdMIxYNttGz8KfG+JtJn331c2TLl1DSXARZtoPzK2WT2xSFmqoBUUc4B51VAKoS4TghxX3d396geNyPH4PbPeMnMM3jx0T7efjl40psJS7GHREtoYNouQFqWwYVXOJh1ge2EG3YpJWueC7Jve5QLr3QwY4H9pO1Jc2TzxQU/ZkXphwjGe/nbvt/wnTc/xn1bvoGsWIvFE+Stl4Jn9qLHgG7zkbnwy5Td8nf85TfSueevHHjsGlo230MiGjj1AY5hcWaSc9E3SZvxMTSLk669j9D45n/Qtf9J4uGOMXoFysnEm0Ijmq4L/bVI/TYSaoRUeQ8QQuBeUUC8MTgwi2YwQ60hlVISqezENjXlhCQt6w+/TEPgINeVf3xg6vcRdQdjxGKSslFaOwrJhEYAGeNY8qV9x4P0HHqWjHmfx1O0fNSPL6Wko/KPmPEgGfO/gCPjxPJQyskJIZK1SLvGISBVJV8U5ZxxXiU1OpOyL6fidGvc9DEPr/09yMY1YdqbE1x1qxur/cTRIGuxh9C6ZmI1AWxTj06r8qfpR9rJvu1RsvIN/Gk677waYvv6CHMvsjN/6cmD0SO8tmQv+BXFt9PYW8PW5jVsbX6DJw7+L9oCg4bWmbi3LWPpjAux6cM75tlicWWRc9E3SJ3xEdq2/Ir2bffRteevpM26i5Qpd6AZw2uv0Ayc2QtwZi8g0lVFoPZVAjWvEKh5BUfmHDyFl2Hzl4zxq1EAzFAcszs6ogy7R+hpNuJqhFR5j3AsyKDn8UP0Pl+HY0764Bv1T9l9d5mMeH0fZncU24zj14/2xXp4/uAfKfVPZ07WJcc9J03JoT0xUtL106qvPZS2xgQurxi3rO6Bmldp3fQLvKUrSZt115icQwhB6vQPI804FveJ9WCV4dFTrOMyZffIqOx4JFMaDbFYjPr6esLhiXl9fPnll2du27aterzboUxYJrAzHo/fNX/+/JZTbXxeBaRjTTcEy29wkp6t88bzQR79bQ/v+6AbX+rxF3k9w45wWohWHx+QHhGLQn1VHNOEmv0x3nk1zNR5Vi662nFa61Jy3EXkuD/MitIPURfYz+bGNayLvsEL7T9l9Ru/YkbGBczJXMrktHkYwyi/crbYfMXkLfsBqTM+RuvmX9Cy4cd0VP6J9LmfwV9+A0Ib/sfT5i/B5i8hHu6gt24NfQ1vEettIPvCryOESBYwV2t+xszRhEanEZCm24ke6BntJinKhCR0gfvqArr/fIDI/m5sk05Mwqd5rLguz0N/V+bpcGX/+tF31fp94dD/EYr3cWPF3Sd8zzXVxwn2mkybN7odk22NCTKyx+cWItyxj4Y1X8OePp2ci7496t/tweYthNv3kDL1TgznEJ0GyrDpKbZx+Y5PnOMjpPX19Xg8HoqLiyfk/UsikYjPmDGjbbzboUxMpmmK1tbWaU1NTfcD159q+/Nqyu6Z6n6iisCLdSQC0SG3EUIwe7GdGz7iIRgweeTeHuoOxU7YxlrkJlbbO+jUXqtNsHSlEyEka54NUjLFwvLrXWf8hSOEoNBbwY2T7+JjOb8ma8vXKLNcwt6OLfxux/f59psf4ZHdv2Bfx9YTykeMJ0f6NAqvupfCFQ9gcWXRtPbbHHryZnqqXxpxwiLDnop/0o3kXPxd0mfdhRACMx6hef1/E2rdPkavQDmdki9HGGl2Eh3ho9MUFeU857w4G+Ey6H2hbtDnNaeBfXoquuf4DsTIzg6MXOdxN9gNgYOsa3iRi/KvIdd94oyQQ7tjON0a2fmjFzzGY5KOtvHJsBsPdyQz6lo95F/+s2HPqBmJWF8Tsb5GZGLoewFl+PRUG4nOyMnXTY+BRGcENNB852bZl3A4TFpa2oQMRhXlVDRNkxkZGd3AsNY6nFcB6ZmsIZWmJHawh55HDtH01XV0/HoX4Z0dQ36BFpRZuP0zXpxujaf+EGD7+uOTFVmKPchIYmDk6N0aqmOs/luQnEIDmz3Ziz2aJs90UOicjrHuw/y/Rb/nk7O/ybT0RWxrWct9W7/Bd9d+jL/t+w1VXbsw5fBL2owlV85Ciq79I/nLfw5Cp+HVf6L6mffTd3jdiI+lGTYs7hwAzFgvmsWNZnEDyRuaWF/TqLb9vS7eGAQNjIyR3xzq6XYwx68sgKKcqVBwZN+hmk3HvTyP8NZ2Yo0nXiPMUJxYU/C48jBmJEFkfzf2Y6brSin52777cFo8XF3ygROO09maoKM1QekUywlrTs9Ee0sCaSbzK5xNMhGj/pUvEw+1k3/5z7E4M0fv2FIOJNnzlqwgc/7n0Yxzc2RtotFTbJCQmIHYqTceRYnOCLrPNqqf/bNNBaPKuUzTNMkwY83zKiA9kzqkQhOk//NsMr+7APfleUT2dNL+0x00f209PU9XE+84cQ6/L1Xntru9FE+y8Prfg7z6dJBEf3ZES6EbhCBW3XvCfodrYjz/l17Ss3VW3unG6dbZ/EaY2oOj92UtNMGSq5wEukx2bYwzNW0+75/2Zb518UN8dMbXKPVPZ/3hl/nl5q/xn299ir8f+D0NgYPjXkJFCIGn6DJKb3iMnIu/SzzcSe2Ld1Pzwl2EWnec1jENRxqZ8z8/kKE3UP0yTW9/j9YtvyLcvnvcX/P5INYURM9wIIyRf6Xoaf1lAdQ6UuUc1Fwf5/c/6mLL2tCIRoBcl+eCRaP3xRNHSWMNffQ8cvC4RDDRfV0Ql9iOma67ufk1qrt3c03Zh3H0d7gdq7UxjsUiKCgb3aUabUcSGp3FEVIpJY1vf49Q8xZyL/4ujvTpo3dsM0Hn7r/Q/M4PiIc7EUIgJtDylnOdPk6lXxKdEbSUc3N0VFHea86rgHQ0WHJd+O4oI/tHF5LymakYWU4CT9XQ/C/rafvZDkKbWpHxo73WVrvg2g+4WXCJncqNEZ78fYBQn5ks/5LrJFp7fPbY1sY4z/ypF49f4/qPeHB5NC5Y7iA9R2fb22Gq9ozeFKGCMguF5RY2vh4mEkq22aLbmJm5hI/M/Brfuvgh3j/ty2S7i1hT9xQ/3fBlfrD+s+xq2zBqbThdQtPxT7qBslueIeuCfyXSuZ/qv3+Q+le+TKTr0Bkd21t6Dd7Sa4kG6mnd8kua1n2f3oa1yMTZ7b09n8SbQlhOY/0oHK1FqgJS5VzkTdUoLLfw5gshnv5jL8He4Y2W6h4rrouyCb7dfEIGUkuOE8/1xejHTDUM7+wEi4atItnhGo4H+fuB31PgmcTCnCsGPUfFLBuX3eDEsIzuKEtrUwKLDbz+s3cL0bHrT3Tv/xvpsz+Nt3TFqB1XJmK073iAvsNrceddjG7zj9qxlaTxC0ij5+z60cHse/gydv9u1qj92ffwZcM675NPPokQgj179gw8Vl1dzcyZM53f//73B6YpfOQjHyn8xS9+kQZwyy23FGdmZs4KhUICoLGx0cjLy5s52PF/8IMfZFRUVEybMmXKtPnz50/etGnTSadabdmyxT5nzpwpVqt13je+8Y2sY59ra2vTV6xYUVpSUjK9tLR0+qpVq1yDHeM73/lO5j333JMG8OCDD6aUl5dP1zRt/po1a467kVm/fr1jzpw5U8rLy6dXVFRMCwaDAuC3v/1tSkVFxbTy8vLp//AP/zBQ9Pn55593T5s2baphGPN/97vfDfQeHj582Fi6dOmkk72u9zoVkA5BWDScCzNJ/+ossv57EZ73FRKr76XjV7to+ud1dD96kFj/dFyhCS680snVt7lobojz13t7aG2M416eh/emo2t6utoTPP1QAKtVcONHPQPZCQ2LYOEyB9kFBjs3Rti3IzJqo3ZLrnIQDks2vnHizb7dcDI/+zLumv0NvnnxH7h18mfRhcGD27/L3w/8joQ5/rVMNd1K6rQPUnbrc6TP/Sx9h9dx6MmbOfzmN4j1Np7WMXWrB1/pSnIv/jap0z6EEAadux/m8Jtfp/vAMyQio1s26HwnTUm8OTjiki9H6KnJa09clX5RzkEOp8Y173ez7DonDdUx/nxPNzX7h9ex6L4qHxKS3tUNxz2uuSxYiz0I69ERyEhlJ7YK38BjL1f/lUC0k5smfxpNnHgpj8eS1xCbffQv8/+fvfOOj6pK///n3Du9ZSa9994oSehFqgjSpFpwd/2qWFbBrr9VXHfddW2ouKC42FjWCDYQARGkFymhBUghvfc2vdx7fn8MCQQS0iYFnPfrlVcyd+4998xk5t7zOc9zPk9NOQd3b0GfpULqSg6i6sS7UAZNgvuQRx3WLm81oPr0ahir06GOnA+X8JnOFMleoD8jpLeSIOVMna/b7sj2UlNTMWbMGKSmprba7urqirVr13qaTKY2vzQsy9JVq1Z16Ar24IMP1mZnZ1/MzMy8+PTTT1csX7484Eb7e3p62j744IOipUuXVl773MMPPxwwderUpvz8/AsXL168OHjw4OsGFlarFRs2bHBfunRpLQAMHjzY+N133+UkJyfrrt1vyZIlIR999FFhTk7OhQMHDmSJRCJaUVHBrlixwn/fvn3ZOTk5FyorK4VbtmxRAkBoaKjl888/L5g5c2arN9fX19fm5eVl/eWXX9oUyE6cgrRTCDykUM0JgfdbI+C2LB6icBfodpWi6i8nUP2vMzAcqQBv5hCZKMa8B1XgeYpv/9OE/DICRmwfPOiaeGz5QgueB2b/UQml+hpnXpYgaawE/iECZJ21IOOUxSGi1MNHgKhEEc4eNUHb2L6RkVyowgi/aViW/C5G+k3DvqIf8NHp/4cGU3WP++AIWKEcHoMfQdj87XCNvQ9NuduQ+92dqDn3WbffJ8IIIfcdAa/hL8AjaRnELmFoKvgFZYdWdGvd6u8VrtYE2Gi3DI0A++QPoxaBq3GuIXVyc0IIQcIwCRY9ooJMzuDH9Toc3HFlCUd7CLykkCZ5QL+3DLzxygQg12CGJbepJRvHVmuCrdzQkq5bqS/GweIfMcxnCgJVkde1a7VQ/PqDHgXZjjfloTxFTYWtz+qPmhvyULrvBYg1EfAd+w+QNsR3d+DMjahK+wDmxgK4xf8RysDbHNKuk+thlEKAJX0qSHmjDdTE3VKCtD/Q6XQ4dOgQPv30U3z99detnlOr1XTMmDHa1atXu7V17NKlS6s++ugjL6v1xtlnrq6uLWklOp2ObZ4Ueu211zwXLFgQDADHjx+XRkRExGm1WsbPz882fvx4g1AobHWBra2tZY8dO6Zcvnx5DQBIJBLq7u5+3cB369atqoSEBINQaE/LHzp0qGnQoEHXfTi///57l5iYGOPIkSONAODt7c0JBAJkZWWJg4ODzb6+vjYAmDRpUtM333yjAYCoqCjL8OHDjQxz/XVqzpw5DevXr2/zvXJyiwnSnpgadap9hkCS6Aa3x+Pg/fYIqOaHgGuyoP7TLFQ8cxQNGy5BYzNi0SMucPdmseNrHc5sKEfDgQr8+KUWRgOPWfcr4dpOLTiGIRg8SoLgSCFyMyxIP+4YV7oRk6WgFDi2x9jhvkJWhHlRj+HeuGdRrivEyhPLkVFzssd9cBQCiQZew55F2LyfoAgYj+q091G67znw1rbNozoDIQQSTQTcBz8M71EroAy4DSKXYACAuSEXjbnbwNs6fu9+r/Sk5EszAneJsxapk5seNy8BFj6iQsJwMc4cMeGb/zShvubGjuaKaf6gRg76A1cyPiz5Wmi3FYJeFrTm5nIv8a6glGJz9icQsRLcEbakzTYpBfxDBdC4O140NtbzsFr6xtDIZmpA8e4nQFgRAiatAiPs/jXmaqz6SlSeWAmbsQYegx+BzDvJIe06aRvCELBqUUtd0L7gZi/5MlDYsmULpk2bhsjISLi5uSEtLa3V8y+//HL5hx9+6G2zXZ9RFxQUZElJSdGtWbOmQxH2xhtveAQEBMS/+uqr/qtXry663HZVfn6+eP369eoHHnggePXq1QVKpbLdNRFZWVkiV1dX24IFC4JjYmJiFy1aFNTU1HSdzjl48KBi6NChHQ4as7KyxIQQjBkzJiI2Njbm5Zdf9gKA2NhYc15eniQrK0tktVrx448/asrKyjpcrDx69Gj98ePHr1/s7wTALSZIe2Jq1FVYFxGUdwTC6x8pcH9+ECSD3aE/VIHqv52C/r3TuD1Uh9h4BqWndDi7pQ71NTbMuFcJL78b38QJIYhPESM8ToTCS1aUF/c8bValZjFohAQZpy2oqehce0O8xmF5ykq4iN3w6bm/YVvOlwMihbcZocIHfhPehWfycmgLd6Ng2xJYtCU9b1fmAXXkXAjl9iLo5voc6EoOghD7/81qqAIdQO/DQKCl5Es3U3YBgHWTONeQOrklEAgJbrtTjhn3KKCt57Hxo0ZknGp/GYYoRAVRtBr6XaVX/Akul0AiAnukwHS+DoxGBIGvDOerj+JS/VlMC7kXSpG67TbFBHHJkutqZDuC6nK7wO7tki+UUpQdeBE2fQX8J77f4preU6y6ClSdXAnKWeCZtBwSt2iHtOvkxrAacZ9GSJvFL+vqFKQ9ITU1FYsXLwYALF68+Lq03djYWMuQIUN0a9eudW3r+BUrVpSvWrXKm+dvvLb+pZdeqi4uLj7/17/+teTVV1/1AQCWZbF+/fr8Rx55JGTkyJHaqVOn6m/Uhs1mIxkZGbLHH3+8OiMj46JMJuNfeeUV72v3q6ioEHp4eHRoGmKz2ciJEycU33zzTf6xY8eyfvrpJ82WLVuUHh4e3HvvvVe4YMGC0JSUlOjAwEDzZTfZG+Lr62urqqpyumy1wy0lSPsDQgjEUWq4PhgNn5Uj4XJvOMAD2tQcxOxLR5S7GVUyFaYtUiIgtHOufYQQxAwRY8QkKXwCHTMLnTxOArGY4Miuzkf6PGR+eDLpbYzwvR17i77DR6f/ggbTwKmBTAiBW8IDCJiyGlZ9RbdLxNwIVcjt8Bn9KggrBKUUNac/QtnBV9BwaQushiqHnutmxVpuAJELwCi670rJuonB1fV9nTonTnqL0BgR7v6zC7z8BNj9gx47v9HDbGp7UKac5g+u3gzjcfs1paUmL0tAeQpzRgMksa6w8hb8mPMZfORBGOl3R5tt1VbaUF1u6zX38JpyGxgG7Wb6OApTzXnoS4/AI+lJyLwGO67d+iwADDxTnoZIdcOlak4cCKsRg2vou7quzghpz6mrq8OePXvw4IMPIjg4GG+//TY2bdp03bVlxYoVFe+//75PW9echIQEc2xsrOHLL79sMfh54okn/KKjo2Ojo6Njr93/oYceqtu1a5e6+XFGRoZEJpPxFRUVHQ4wgoODLV5eXpaJEyfqAWDRokX1Z8+evS6tQiKR8CaTqUP94+/vbxk+fLjWx8fHplQq+SlTpjSePHlSBgD33HNP47lz5zLPnDmTGRUVZQoPD+9wtsVgMBCxWDww6iwOQJyC1IEwMgEUE/3g+dckeKwYCtkob8jKGjCqLBfejXUwpXdtQbqHjwCEEGgbOaQdNLaYVHQHiYxB0jgJCrOtKMnvvJuskBVjfvTjuCf2GZTp8vHeieXIrE3r+MA+ROE3GiEzUyGQuqPol0dQe369QwdjjOBK5E8dOQ9ilxBoC39FxZG/oerkB9CXn/hdO/TaKgwQest6ZAYicJcAHL3ObdSJk5sZhYrB7D8qMWKyFDkXLPh6dRMq2sh6Ece7QuAnh3ZHMSil9kgpQ0AIgTVfC2qwQRyvwd7C71BvqsKcyKVgmesFIaUUF0+Zcf5E732Pqis4aDxYhzv3Xkt95kYwAhnUkfMc2q4yYDx8Rq+AUObh0Had3BhWIwZf7zjDxo5oEaRqZ0Cqu3z77bdYsmQJCgsLUVBQgOLiYoSEhODgwYOt9hsyZIgpIiLCuHv37jbTE1999dXy1atXt0QqP/zww9LMzMyLmZmZFwEgPT29ZdZg48aNLkFBQWbAvib0mWeeCdyzZ09mXV2d4GrX2rYIDAy0eXt7W86ePSsGgF9++UUVFRV1XepVTEyMKScnp8OZirlz5zZlZmZKtVotY7VacfjwYWVcXJwJAEpLSwUAUF1dza5bt87zscce69Bw5fz585LIyEjn+q92cArSXkIUpIR6SQS83x0JViOG9sdC6PeWwXSu6y5pTfU86qs5mE09u5APGiGBQsXg8E5Dl28KQ73HY3nySqhEGqw7+xq2564Hx994bVRfIlIFIujODVAGTkDViXdQduD/gbc5NgWUEAKpRzzcBz8M37F/h0vYLHDmetRd+BJlB/+C+qxvYdGWdtzQLYatwtij9aOAPWUXADin066TWwyGIUgZL8X8B1UAAb5d14QT+4zgr8oGIIRAOS0AtjIDzOl19pRd5kq6LghgCLFib9F3GOw5FmGa+DbPVVfFoaGWR2iMsNfcYmvKbb1ef9RmakBT/s9Qhd8JVuSYJVeUt8LcmA+g9SSjk76BdRWDWnhQfd8seeHqLWBUwm7Vxh6osBLH+uF01F5qairmzp3batu8efOuS9sFgFdeeaW8srKyTfWfnJxsiouLa3fN5sqVKz3Dw8PjoqOjYz/44AOvL774Ih8AHnnkkYAHH3ywKjEx0fzll18WvPrqq36lpaWCoqIigZeXV+Inn3zi9d577/l4eXkl1tXVMQDw4YcfFt17772hkZGRsefOnZO+/vrr15VjmDNnTuORI0eUzY/Xr1+v9vLySjxz5ox87ty5EWPGjIkAAA8PD+7Pf/5z5ZAhQ2JiY2PjEhMTDYsXL25s7ltYWFjciBEjop9++unyxMREMwDs379f5uXllbh9+3bNU089FRQeHt5SMHnXrl3KadOmOcs4tAPpq9mqviQ5OZmePDlwjHh0u0rQ+HUuZOO8wTdaIJ/kB0lcm+n27cLZKFgBAaUUNisgFHVvsJFxyozdP+gxbaEcEQldT2WxcmZszv4PjpX/ghCXWNwX/xxcxAPHNIxSHrVn16H69GpI3KIduvao7fNRmOsvQV96GIaqswC1wTP5aYjVob12zoEEb7Sh/M+HoZoXAuX0wG63Y60woOovJ6B5MBqykV4dH+DEYRBC0iilyb18ju8BfApgB6V0QKYs9cV9w2zise9HA7LTLfALEWDqfAUUKvuAmdp4VL50HKyHBNIUT1guNcL14RhU//M0KE+xbdYOXKo/i+eHr4Fa0nYlhRP7jKir4jD5LjlYgeMFqUHH49M3GzDmDhmGjLphqcAeUZv+BapOrkTI7G8hcb3eRbg7aIv2oSH7W3gNfxEipb9D2nTSeYwnq1H30UV4/jUJwoDe93WpeT8dfJMFnituXsOqjIwMxMTE9Hc32uX8+fOG+Pj4jP7uR3eYMmVK2MqVK0sSEhL6LC0rOTk5aseOHTkeHh4DJ5rTB5w9e9Z90KBBwR3td+tMHaH3XXa7i2ysD4hcAL7JCmGgAvpfy2DOauhSG82Di8wzFhz62QCjvntjuqjBIrh5sTi629hhSYK2ELJiLIj5M+6OfQqlujysPL4MWbWnutWX3oAQBu6DH4b/pA9gaSpG/tbF0Ff03iCTEAKJayTcEv4E33H/gCZ6EUQu9tqzjXnbUZfxdZ+lKPUHLQ673Sz50ozgsvGEsxbpLcsaAPcAuEQI+RchJKq/O9QfiCUMpi6QY/JdclSV2pD670bkZdjX1REBA8UUf1iyGmGrMoCwBLzBBkteE3ShJlyoOYbJQQvbFaO6Jh6VJTYERwl7RYwCQE2FfRzVmyVfKOVRn7UJUq+hDhOjACD3HQnX2PucYrSf6OtapLdaDVInjuWdd94pKSkp6b7xRRcpKysTLFu2rPL3Jka7wi0lSPvSZbcrMBIWiol+MJ2phWSoBwR+Muh+KYE5p+vC2ctPAJOR4sgvBugauy5KGYZg1FQpGut4nD/Z/RtDkvcELE9eCeXlFN4duRsGVAqvMvA2BN/5P7BiFxT9/DDqLqb2ujBkhXIo/Me2pMpRzgLKWVoea4sPwNyQe0u59NrKmx12eyZIiYgF4yJyOu3eolBKd1NK7wUwFEABgN2EkCOEkD8RQvpsUDAQaDatW/SoC5QaBtu+0mHfVj1sVgrZWG8QmQDmiw2AgIH5Yj1AgX3Sn+Eu9cW4wNnttpufaQFhgODI3ns7q8vt1y73XhSk+tIjsGpLoIle5LA2KaVgBGLIfUc4rE0nXaPPBWmd2emw66RdBg0aZL7jjjt0fXU+X19f25IlSxr66nw3I7eUIB3IyCf5gYgY6HeXQDUzCAIvKXQ/F8NSoO1SO66eLEZNkcJmA/Zu1WPvj3qkHzehrNAKs7FzAjUoQgi/EAFO7DPC0oN1qZ5yfzyZ/A5SfCbj18JNWHvmFTSau75GtrcQq0MQfOf/oPAfhcpjb6D88F/B2/rONEcdMQdu8fcDADiLDg1Z36Dq5Hso3fccqtI+RGPeDpjrc0D5m9cQyVZhABhA4NHz9D3WTexcQ3oLQwhxA/BHAA8COA3gA9gF6q5+7Fa/oXFnseAhFYaMliD9uBmbPm5CfRMgv80HthI9qImD6UI9OBGPTOU5zI58CAKmbbFpMVMU51rhHyKEWNp7t/Wacg5KFwYSWe+doz5zI1iJK1RBkx3SnkVbioqjr/8u1/cPJBgXEUD6RpDyZg7UYHNGSJ04uYlwCtI+glUKIRvrA8NvVeB0VihnB4N1k0C7rQjW4q5N0ri4shg3XYaYIWLIFAQleVakHTThl+/02LdVj6Z6e6SyvYggIQSjp8pg1FOcOtwzwy8RK8bCmCewOGY5ipsuYeXx5ciuO92jNh0JK1LCf9IquA9eisZLP6Dw5wdg1Vf2Qz8U8B33BtwSH4LcbzR4qx5NedtRlfY+Svc9j6q0VWjM237TlZKxVhgg8JA6xDhC4C6BzRkhvSUhhPwA4CAAGYCZlNJZlNKNlNInAPxuC4WzAoIx02SYdb8CRgOPjR83odjNvaXci/F8DQq9ChDjmYwYt/bXwhVmW8BxQGhM7wabqytsvVp/1KItha74ANSR80DYnr8WSikasr8Hb9WBlah73kEn3YawxJ4FU9f7pV/4hmaHXacgdeLkZsEpSPsQxVR/gFLofikBI2ahmhMMVi2Ctbxd87F2kcoZhMeJMHyiDLcvVGDM7TLEDBZDKieQyu3/1twLVuz/Sd+yVvRqV0cvfwHC40U4fdgEvbbnHiPJPhOxLOVdKEQq/OfMX/Fz3v/A04GRwksIA48hj8Nv4nuw1Ocif+vdMFT2vWhmRQrIPAdBEzUf3iNehO/4f8F90MOQ+48BbzWgKW8HbPoKAPaBWWPuNnCWrkXQ+xpHOOw2w7pJwNU6a5HeoqyilMZSSt+glLZyPextQ6WbgaAIEe5+3AV+IULs/dWGGi9XmE5Vg9ZZUehbgFkRD7Z7LMdR5GdZ4enLQqnuPbFotVA01PC96rDbkP0tQAg0UfMd0p6pJh3m+iyoQqeDFcod0qaT7sNqxH0SIW0p+eLqLPnixMnNglOQ9iECdwmkw71gOFAOXmcFIxXAZWEYZMM8AaDbA3GGIdB4sAiPtwvUZgdeqcK+vdng4uR+E/b/pMf5EyaUF1mRPE4CngOO73VMWSRveSCeTH4XSd4TsbtgI9aeXoEmc51D2nYEqqBJCL5zAxiBDIU//x/qs77t1/6wQjmkHonQRM6D94gX4Tf+TUhc7Y56lqYiNOXvbNnXUHkKjbnbYKrLBs/1XXHxG0F5ClulAQIfx5RQaK5FyjcOjNfnxKHEEkLUzQ8IIRpCyGP92J8Bh0zBYNZ9Coy5Q4YcsxzNXsQ+STFwk3q3fyAFQqNFCI/r3cF3bSUHSgF3b0GvtM9zFjRkfw9FwHiHOKNT3oqG7B8gkHtD4TfGAT100lP6TJBejsI6U3adOLl56J07i5N2UU4LgPFoJXR7y6CaGQQitM8J2KqN0O4ohvKOAAg8HDPA9wsWwi/4StqTuzeLqlKKohwr8rOsIAA0niwunDTDP0SAoEgRROKeuTOKWQkWxy5DmCYe32d9hJXHl+HeuGcR4Tqoh6/GMYg14QiZ+RVK97+AiiN/g6k2A97DX3RIelhPYYRXIo0Kv5GQeQ0BI7CvzTQ35kNXtA/I3wEQAUSqQEg0ERBrwiFU+IERKXut7mB7cLUmwEYdGiEF7E67zoHELcdDlNLVzQ8opfWEkIdgd991chnCEAwZJYF3hg215wywsUaMSZxzw2NYAUF4fO9Hgmoq7IZGvRUh1RbsAmeqhyZ6sWPaKz4Am7Ea7oMfA2F6t26qk87BasR2o65epln0MrdYyu6nb9bDoHNcBpFMQfB/L2g63G/z5s2YO3cuMjIyEB0dDQAoKChAQkKC7PXXX/f8y1/+UgUA999/f2BycrL+ySefrJ03b17wwYMHVYWFhelSqZSWl5cLkpOTY0pLS9Ovbf+tt97yWLdunQfDMJDL5dwnn3xSmJSUdMP1Oz/99JPy2WefDbDZbESj0dhOnDiR1fyczWZDQkJCrLe3t2Xv3r05bR3/wAMPBCxYsKD+jjvu0P3zn//0+Pjjj72Ki4vFZWVlZ318fGwA8NFHH7m+99573gAgl8v5NWvWFI4cOdIIAK+99prnf//7Xw9CCKKjow0bN24skMlkdNasWSHnzp2TC4VCOnjwYP2GDRsKxWIxfeWVV7y++eYbNwDgOI7k5eVJysrKzri4uPBjxoyJPHr0aJZQ2P/j0P7kloqQDtSyL1cj9JdDMsgV+t0l4M1XUlqJmAUjE7QI1N4gNEaEEZNlmLZQgdFTZYgcJIJ/iAAgwIHtBvzyjQ7nTzpmDV+KzyQsS14JmVCFT86swPbc9TDZup6a3BuwYhUCJv8bbgkPoCHrGxT+/CBshpr+7tZ1NItRANBEzoPfbW/BffBjUAZOACiPpoJdqD69GmUH/x/KDryImnOftuxv0ZaCMzf1av9aSr44SpC621+v02n3loQlV82YEEJYAM58unbwnBeIw7efwfejD6Cppv154/pqDqX51lbLMXqL6nIOYgmBUt0796j6zI0QqYIg9x3e47Y4ixZNeT9D4hYHqXusA3rnxBGwrmJQEwfe2LtO81y9GUQuACO+tSYiHClGu9JeamoqxowZg9TU1FbbXV1dsXbtWk+TydTmbDjLsnTVqlVt16m6igcffLA2Ozv7YmZm5sWnn366Yvny5QE32r+mpoZdtmxZ4NatW3NycnIubN68Offq519//XWv8PDwdlP/Kioq2LS0NHmzy+748eN1u3btyvb19W2VnhUeHm4+fPhwVnZ29sWXXnqpbOnSpUEAkJ+fL/zkk0+8zpw5c/HSpUsXOI4j69atcwWAe++9ty4vL+98VlbWBZPJRN5//313APj73/9emZmZeTEzM/Pia6+9VpKSkqL18vLiJBIJHT9+fFPz8b9nbilBOlDLvlyL4o5A8DobDIcqWraxKhFU80LAqsWglILX957zKsMSuHqyiEwQY9x0OVLGS2DQUXj4snDztF/AeY6C53p28fNWBGJZyrtI8p6APYXf4h9HHsSu/K9htPaZ03a7EIaFZ/Jy+I1/C6a6DORvXYzKSxexY6MOX7zTAIOu5+tqHQ0jkELqHgt1xGx4DXsWfre9BY8hf4Y6ch6knkMgkFyZ6axN/wz1WZtaHjdk/wBt0V6YajPBmRsdUgKnsyVfeKsBFm0JzA15Nyx502zR7xSktyQ/A9hICJlECJkEIPXyNidtwKrFsLg3gYMQRTnt3wuKcq24kGZGX5Q6rrlsaNQbmRim2kwYq85AHb0QhPR8WNKYuw2UM0MdOdcBvXPiKPqq9IuzBqnj0Ol0OHToED799FN8/fXXrZ5Tq9V0zJgx2tWrV7u1dezSpUurPvroIy+r9cbjWVdX15YBl06nY5uvMa+99prnggULggHg+PHj0oiIiDitVsusW7fOdcaMGfUREREWAPDz82sZWOTm5gp37tzp8tBDD7UbZdiwYYNm0qRJLTP2o0ePNkZFRV23VmjKlCn65rqhEyZM0FdUVLRMonIcR/R6PWO1WmE0Ghl/f38rACxatKiRYRgwDIPk5GR9SUnJdROvqamprgsWLGhZzzZ//vyGr7/++ncvSJ0pu/2AOMIFonAVdDuLIR/v0+JQ2vwlNByuhCW7Aar5oWBVvR9EGDpGigsnzSjOtWHYBHu68MVTZjTU8hg5RQqW7f4AxJ7Cuxyj/Gdgd8FG7Mz/CvuLt2Cs/0yMDZgFmbB/DTZVodPAiUJx8NuTKFrvBkZgBMexOHPUhFFTHBP5o7wNlLeCsBKHDuYYgQQSt2hI3KKve04TvaglDZnnLNCX/wbeqr/qWBkEcm8I5d4Qyn0gVHhDqPQHK1K2/RooD95qAG/Vt/wY8xoAKYW2cgf4EgMU/mMgUvrDVJuJugv/hcfQxyFU+EJfcRINl8WxSBUMt4Q/QSC9/v7FiFkwKiFsNX1XmqcvaTLX41z1YYz0vQPs7y+F8AUASwE8evnxLgDr+q87AxvThTrIa4UQC+yCNGls28s4EoeLYYgT9ega3Rl4nqKmgkN8cu8M8uszN4GwEqjD26+z2lks2lLoSw/b16LKb7D21kmfw2rs4xmu3gyhb++ZTDkFqePYsmULpk2bhsjISLi5uSEtLQ1JSVccv19++eXy6dOnRy5btuw6ARgUFGRJSUnRrVmzxm3hwoU3TF184403PNasWeNltVqZXbt2ZV1uu2r48OFR69evV7/11ls+q1evLlAqlXx2drbEarWSYcOGRen1eubRRx+t+vOf/1wLAI8//njAW2+9VdLY2NjuTfbIkSOK+fPndyl3/MMPP3SfMGFCIwCEhIRYH3/88YqQkJBEsVjMjx07tumuu+5qlZJmNpvJxo0b3VauXFl89XatVsscOHDAZd26dUXN21JSUoznzp373buuOQVpP6GYHoi6VedhPFEN2UivVs+Jo1xgPl+Hpu/zoZoXClbZu3nlQhHBsAlS7P3RgPxMK0JjRHD1ZCEQEocNdAJVEXgg8WWUaHOxO38jdhV8jYPFP2JMwJ0YGzALcqHKIefpClYLxdmjJqQd9IDVMg2BXscQKH4TeeY3cO5ICAJk/4GQNYJyFlDeAspb7X9z1pa/eb7149a/raC8Bc3uJBK3WHiPfBlSj/hef20S18iWvxlWBN9x/wJv0cKqr4BVXw6bvgJWXQWM1eegLzsCAFCF3AGXsBngrQbUnv8CyuApkGgiYKrNQPXpNQBah2FI4XhAzkBbsB+MUA6pRwKg9AcjUkLiHgtcrpkocYuBW8KD4K06NFzajMpjb8I17j5IPRKv6zfrJrlla5H+nLcBx8t3oclch+lh9/d3d/oUSikP4KPLP046wHC0CmqiQlOgDWWnbbBaaItZXTM8T8EwBHJl768db6jlYbMC7j6OHzJwFi0a836CKvQOsOKe3wd4mwEiZQBUoXc4oHdOHMmVCGnvGtdx9WYIg9ueXHXSNVJTU7Fs2TIAwOLFi5GamtpKkMbGxlqGDBmiW7t2bZsRvhUrVpTPmTMnfP78+TcUpC+99FL1Sy+9VP3xxx+7vvrqqz7ff/99AcuyWL9+fX5ycnLcvffeWz116lQ9ANhsNnLu3DnZwYMHs/V6PTNixIjocePG6S5cuCBxd3e3jR071vDTTz+1+wGorKwUenl5dTpvfOvWrcoNGza4HzlyJBMAqqur2W3btqlzcnLS3dzcuBkzZoSuWbPG9bHHHmuJev7hD38IHDFihG7atGmtUgK//vprl6SkJJ2Xl1fLmj2BQAChUEjr6+sZjUYz8NLz+ginIO0nJAmuEPjJod1RDOkIz1aRM4GHFMo5wdD+UICmH/LhMj8UjKx3/1WxQ8U4c8SEI78YEBwphG+QEL5B9ufqqzmUFVoRM0QMpocC1V8Zhj8m/j+UavOwu2ATdhdswsHirRjtPwPjA+ZALup9YcrzFJlnLPjtVwP0TRQh0UKMmiKDxn0Kqk5egPn0eyi1fYzzJywI9dgKwopAGCEIIwLDCgFGCIYVgbAisEIFiMT+HGGbf1/en72yHZSiIetbFPx0LzTRi+CR9ES70cjegBACVqwCK1a1EquAfUBo1VeAFdtT3TlzI3iLDpSzp9kIZJ5QhUwDI5S3/LBCOWq3lUAcr4Zm0l2tPr8ipR9cY+9teSyUeUAo8wAAiF2jUJv+GWrOfgJF4ASow2eDMFc+2wJ3CSxF/Z/S7WhMNgPOVB2EmJViT+G3CHKJRpz7sP7uVp9BCIkA8AaAWAAti6MppaH91qmBDMfDIrDAxUUBHQeUFlgRHHklW8Zk5HFgmwGJwyXwDuj923hNuX3s5u7t+Mh+Y85WUJsJmuhFDmlPoomAeNhzfW7y5qRjmuuC9mbKLrXy4JuszgipA6irq8OePXuQnp4OQgg4jgMhBG+//Xar/VasWFGxcOHCsBEjRlxXpy4hIcEcGxtr+PLLL1vWFD3xxBN+u3btcgGAzMzMi1fv/9BDD9U999xzgc2PMzIyJDKZjK+oqGiJzPj7+1vc3NxsKpWKV6lU/PDhw7UnT56UpaWlyXbt2qX28/NzMZvNjF6vZ2bPnh2yZcuW/KvPIZFIeKPR2Km1AceOHZM+9thjQdu2bbvk7e3NAcDWrVtVgYGBZl9fXxsAzJkzp+HIkSOKZkH6zDPP+NTU1Ah27tyZe217mzZtcl24cOF15SesViuRyWS/65p3nbqTEUKWAfgcgBb2NKshAF6klP7Si327pSEMgXJaAOo/zYT5XB0kg1qnMAq9ZVDODkLTZrsoVd0VAkbaewMPhiUYOUWG7ak6ZJ2zIGbIlYt5TYUNeZlW1FVzGDpGCrmy52t8/JSh+EPCiyjXFWB3wSbsLfwOh0p+wii/6RgfOAdKkbrH57gWSikKL1lxZKcRtVUcvPxZ3L5A1sqJ2GvYs/BM4VGxXoeyigcxZeGzEAgdM7BxjbsP1adWoz7zazQV7obXsOegCpnW7wMnVqRsJY6FCh94DX++5bFA6gaXsBmtjuENNvBNBRD6ds3dVyjzgFfK02i4tAW6or2wNOTBLeGBlhRe1k0M7nQNKE9BmFtnQHm26hAsnAmPDHkdW3M+x9cX38PylPduXM7j1uJzAK8CeA/ABAB/wi3mYeBIqI3CIjBB4+IFkxAoutRakBZmW2ExUShc+uYtrK7gwLCAq4djBSmlFPWZGyFxj++x+RDlbdCVHILCb/SAcE13cj1EyIBRCXtVkHINl2uQOgVpj/n222+xZMkSrF27tmXb+PHjcfDgQQQGtmhGDBkyxBQREWHcvXu3S0pKiv7adl599dXy2bNnRzQ//vDDD0sBlDY/Tk9PFyckJJgBYOPGjS5BQUFmAKitrWWfeeaZwD179mQ++uijgZ9//rnmT3/6U/38+fMbHn/88UCr1QqTycScPn1a8dxzz1U+8MAD9atXry4F7C687777rte1YhQAoqKiTNnZ2WLYNU27XLp0SbRgwYKwzz77LD8xMbHlQxscHGw5deqUQqvVMnK5nN+zZ48yKSnJAAArV65037Nnj8vBgwezWLb19bK2tpY9fvy48rvvvmvVp4qKClatVtvEYvHvWpB29m72AKW0CcBUABoASwD8q9d69TtBOswDrKsY2h1FbT4v9JVDNTMIXIMFTT8UtHLl7Q1CY4TQeDBIP946ZTIiQYzkcRLomygObDegrNBxhks+imAsiX8ezwz7EHHuw7C/6Af888hD2HrpMzSZHWcPX1Vqw+YvtNj6Xx1sNoppixRY8LCqlRhthhAGSWOlMOgoMk477sbJipTwHvEigu/8H4QyT5TtfwHFvzwCS1Pb//+BTE8cdgkjhCZqPtwSHoTNUI2qtFUtZkesmwSwUfBNt1Yt0mNlv8BLHoAwdQLuj38BALD+/JuwDpCasn2AlFL6KwBCKS2klP4VwIwOjvldQikF5XhYYIZUKIVvsBCFVxkbcTaKgiwrvPwFUKj6RpDWlNvg5nmlprWjMFScgKUx3yHRUVPNBTRkfwtzQ5tVHpwMEFiNGFxdLwrS+mZBeuuZeMsUjv3+ddReamoq5s5tbQw2b96869x2AeCVV14pr6ysbPNNT05ONsXFxbVbZmHlypWe4eHhcdHR0bEffPCB1xdffJEPAI888kjAgw8+WJWYmGj+8ssvC1599VW/0tJSwdChQ02TJ09ujI6Ojhs6dGjMkiVLqlNSUjq91mfWrFmN+/fvb5mFf/311z29vLwSKysrRYMGDYpdtGhREAC8/PLLPg0NDYInnngiKDo6OjY+Pj4GACZOnKifOXNmfWJiYkxUVFQcz/Pk6aefrgaA559/PqimpkaQnJwcEx0dHfvss8+2FFX+3//+px47dmyTSqVqlZa7Y8cO1eTJkwdueZA+gnTGbZMQco5SmkgI+QDAPkrpD4SQ05TSIb3fxa6TnJxMT5482d/d6BS6X0vR+FUO3F8cDHFE2+7AlgIttD8VQuAhhWpuMIio9wxRzh414cB2AxY/qoKHb+uIrEHHI+2gCQ21HIIjhYgdKnb4AKVKX4LdBZtwuvIAWEaAkX7TMCHwLqjE3TMga6zn8NsuI7LTLZDI7Gtl45M77jelFN980gSjnmLJMpcepypf1z7PoT5zE6pPrQLlrXBLfBBuCQ+AYW+Om6jhSCXqP82E5z9SIOxB2RebsRY2QzUkbtGglMJ0rhp1qzLg/tJgiMMHtlt2ZynXFeDd409iVsT/YVyA3bTlQs1xfH7udYzwvR3zox/v1/4RQtIopcm9fI4jAMYA+BbAHthnx/9FKY3qzfN2hYFy36AcRfWHZ7FZ9T9ETh4FdckdOLjDgD884wKVmkXhJQvOHTNj1BQp3Lx6P12XUopP32xAcJQQk+c61oSuZO8zMJQfR/jCXa3KXHUXi7YEIqW/A3rmpLeo/fA8bDUmeL3WO5ccw7Eq1H+SAc+/J/eqcVJfkJGRgZiYmP7uRrucP3/eEB8fn9Hf/egOSUlJUTt37sxxd3fv3UhPJ5g6dWrYO++8U3J1FPZW4uzZs+6DBg0K7mi/zk6vphFCfgEwHcBOQogSwO924a0jkY3xBqMQQLejuN19RMFKKKYFwFZlhG5vWa/2J3qwCAIhkH7i+u+FTMFg9FQpwmJEKMi24vBOA3RNjv0YeMr9cU/c03h+xGoM9hyDwyU/4Z9HH8Lm7E/QaK7tdDtGA4+DO/TY8EEj8jItSB4vwR+eUmPQCEmnRDQhBEljpWiq55FzwfFRLMKwcI29G6F3bYEycCJqTq9B/uZ50Jf95vBz9QbWCgPAEgjcezaIFEjdWlyCdcX70VBmn3nlam+d6/Kxsl1giQBJXhNatsW5D8PEoHn4rWwnTpbv6cfe9RnLAMgAPAkgCcB9AP7Q0UGEkGmEkCxCSA4h5MU2ng8ihPxKCDlHCNlHCPG/6rm3CCEXCCEZhJBVpL9z4zsLT8FTHhzLQcJKERhuz+IoyrGCUoq8DCvUrgxcPfvGqdmgozDqKTy8HSt+rYYqaAv3wCViTo/FKGexrzt3itGBD6sW927Kbr0zZddJx7z99tslubm5/R4BMJlMZNasWQ23qhjtCp0VpP8H4EUAKZRSAwAh7GuAnPQQRsxCPskPprO1sJZcl3rfgjjcBcrpgZCN8mp3H0cgljKITBQj66wZZtP1YpNhCWKTxEi5TQqjnuLgdr1DU3ib8ZD5YXHscjw/fA2Geo3HkdLt+OeRh/B91sdoMFW3e5zNSpF20Ij17zXi7FEzogeJsGS5GiMnyyCSdG08GhptT2E+ecDkkLqdbSGUecLvtrcQMPUjUMqjaOfDKN3/EmzGzovv/sBWYYDAXdJSssgRCKRuEPnY15LeKk67Vs6CtIq9SPAYeZ1h1+0h9yFMnYDvstagXFfQPx3sAwghLIBFlFIdpbSEUvonSuk8SukNZ18uH7cawB2wmyHdTQi5dqHhOwDWU0oTAfwNduMkEEJGARgNIBFAPIAUAOMd+bp6C2rjwVMOlPAQsVJoPBgoVAyKLllRWcpB18QjNFbUZ2vPq5sNjXwcK4Absr4DKAdN1MIetWPVlaH80CswVJ5xTMec9CqsqwhUb+u1JUhcvRlEwvaq54aTm5+JEyfqhw8fbuzvfkgkEtpcsub3TmdHkyMBZFFKGwgh9wF4GcDvPt/ZUcgn+oGIGWh/bj9KCgCiMBVYpQiUpzCdqwXlekckxaeIYbMCmWfajwx6+wswboYMKg0LphfNZ9xlvlgY8yReGPExkn0m4reynXjj6FJ8m7kGdcaqlv143r7e878fNOLIL0b4Bgpw9+MqTJqr6PY6K8LYo6S1lRwKLzledF+Nwm80Qud8B/dBS9FUsBO5389CfeYmUDowExFs5QYIfBxTp7UZqUcC3AbfB0YhhLWyAfXZ37U4/d6snK8+CqNNh+G+U697jmVY3Bf/LKQCBb5M/xeMtvYnpG5mKKUc7Om6XWUYgBxKaR6l1ALgawDXFqqMhT0FGAD2XvU8hd3NVwRADPskamU3+tDnUI6Cpxw4hoNYIAUhBIERQpTk2ZB70QypjMCnD5x1m6kptwsHRzrsUt6KhuzvIPcbDZEqoPvtUIr67O9BGCHErhEdH+Ck32mOXPINvRMQctYgdeLk5qSzI/WPABgIIYMAPAMgF8D6XutVNyGEzCSEfNLYeHNpZVYhhGycD4zHq2DrRGTIWqKHfl8ZrIU3NAjrNl5+Anj6sTh/3HzDyKBUzmDUVGlL2YGSPCt0jb0joNyk3lgQ/We8OGIthvlOxony3fjXb0vxyZlX8V1aKj794gR2/dAAmZxg7p+UmLlE6ZD1VZGJIihcGKQd6P2IHSOQwGPo4wid/R0krtGoOPo6CrYtgak2s9fP3RUoT2GrMkLgLe2V9ll3CayV9dAV7UXliXdg1betI6iN77XUL95mdMhkwG9lv8BV4oUwTUKbzytFGtwX/xzqTBXYlLGq1yLxA4DThJAfCSFLCCF3Nf90cIwfgKtn6Uoub7uaswCa25kLQEkIcaOUHoVdoJZf/tlJKb051jrZePCUB8/wELP271hgmABmE0VZoQ0h0SKHr2m/EdXlHFQaBmKJ47IhtEX7YDNU9djMyFRzAea6TKhCp4MV3tzrBX8vXKlF6hSkTpw4uUJn7zA2ah8pzQbwb0rpagADruowpXQrpfRhF5ebzwxFMdW+9kX3S0mH+4oCFVDdFQphiP1fQHnHD2ITUiSoq+ZQVnjj2sHNaWOczR6hzD7fu2nwrlJPzIt6DC+NXIskzXSUVFTjaGMqsoJfR+ltj6Fu+Ju4gG9wqe4sLFzP+8KyBENHS1BWaOuV1OS2EKtDEDhtHXzH/RNWbQnyty5G5fG3wVkHRgSNqzUBNtoth93OwLqJAZ0U7oMfBWduQOXxt6AvP9FqH8pT6HYWo3FTLnhjp+tb3xBzfQ4acn5E5bG3ULrveZQdeAm157+EvuJkt977GkMZchvSMcx3ChjS/qU2VB2H6WF/QHr1URwo3tL1fjde52o/EJEAqAUwEcDMyz93OqDdZwGMJ4Schj0ltxQARwgJBxADwB92ETuREDL22oMJIQ8TQk4SQk5WV7e/FKAvYZQi6GYxqHCrgOSyIA0Is68jNRtpy5rSvqKmwgYPB6fr1mduhEDuA4X/df+STkN5Gxqyv4dA5gWFf3cC8E76A6ZFkPaOw7hTkDpxcnPSWUGqJYS8BHu5l22EEAb2FCgnDkLgKoFsuCcMB8vBaTsWPkJ/OQghsNWY0LDhkt1kxoFEJIgglhCkH++cqGMFBGPvkCE+2W5OYTTwsFl7L9ojsrmhfus8BJz6BxaIP8X9sS9hVMAdMHFG7C7YhLVnXsErB+7Gv9NewPbc9ciqPQWTrXvvUWySGBIZQdrBvlvXSAiBS9idCLtrK9SR81B34b/I+342mgp29XsUzVbe/ZIvnUHgLoGt1gyJWyy8hr8IkcIfdRe+RN3F/4HnLKCUQv9rKSy5TZAMdoduZwnMWQ1dPo/NWAtD5amWx41526At/BWEFUEVMg0StziYajNQd/4LlO1/EZUnVkJf0XkX1uPlu0HAIMVnUof7jg+Yg3iPEdiW+wXyGy7ecF9T/SU0FVwpAa0t3DtgU7ububxu9NqfBzo4rBTA1fmc/riqdt3ldssopXdddnz/y+VtDbBHS3+7vG5VB2AH7EtPru3XJ5TSZEppsoeHR/dfoAMhLIFJZYZVaIWItV9PJTIGXv4shEICoajvoqMWM0VDHQ93BxoamRvyYCg/Dk30AhCm+0JXV3wQNmMV1JHzQBjnesGbhd6MkFKOgm+wgHXtd6+aW4rNmzeDEILMzCvZWgUFBUhISJD94x//8Gzedv/99weuWrXKDQDmzZsX7OnpmWg0GgkAlJeXC/z8/NpMFXrrrbc8IiMjY6Ojo2OTkpKi0tLSOnQ5++mnn5TR0dGx4eHhcSkpKa3c2m02G2JiYmInTJgQ3taxWVlZIolEMjQmJiY2NDQ0LiEhIaa53/3Fiy++2OWi5A888EDAjh07FADwz3/+0yMwMDCeEJJUXl7eckH86KOPXCMjI2MjIyNjhwwZEn306NGW1LbXXnvNMzw8PC4iIiJu5syZIQaDgQDArFmzQoKDg+MjIiLiFixYEGw2mwkAvPLKK17R0dGx0dHRsREREXEsyyZVVlayJpOJJCcnR1mtPQvadPYqvgjAPbDXI60ghAQCeLtHZ3ZyHYo7AmA4Ugn9nlKoZgd37iAGAE/R9F0+FJN8IY7WOKQvQhFB9BAR0o+bYdDxkCk6nruQyOz7UEpxcr8JnI0iaawESnX3Bh2UpzCbKcxGCpPx8m8DD7OJIj/TCpsVWPSICmIpgVTmjkRv+3jTaNOjoCEDuQ3nkddwHvuKvseewm/BEAZ+ijCEauIRqo5DiEssZMKOyxgIRQSDRkhwbI8RNRU2hw7OOoIVq+Az6hW4hM9CxZG/o3TvM5D7j4X3iJf6zVHSVmH3AXD0GtJmWDcJYOXBN1khcNHAI+lJNOVtR1PBLzA3FECumwdrphXS4V6QDvNA07f5QCeyBHirAaa6bEjd40BYIfSlR9BUsAsStxgwAik0MXeDFSnBCK6kIlPKw9JUCFPNRZhqLoAzN7a01XBpC5SBt0Go8LnuXBzP4UT5r4hxT4aLuOP7HCEEi2KW4YMTT+O/59/CU8Peg1KkAW8zw9yQC3P9JahCpoERiGGuy4a2aB+UARNAWCE00QsADGwDWULI57Cv62xFB6L0BIAIQkgI7EJ0Mez3oavbdQdQR+2K/CUAn11+qgjAQ4SQN2B/c8YDeL+HL6NP4HRWMBfMkJilEF/+LHI2e2T05H4TTEYeEmnf1B+trbQBFA6NkNZnbgJhhFBHzO1453bgLDo05e+AxC0WUvdrfa6cDGQYMQsiE/SKIOUbLQC9dR12/3rofugsDQ5rTyFS469jOl59l5qaijFjxiA1NRWvvfZay3ZXV1esXbvW85lnnqmWSCTXXd9ZlqWrVq1yf+GFF26YfvLggw/WPv/889UA8L///c9l+fLlAQcPHrzU3v41NTXssmXLAn/++edLERERltLS0laDstdff90rPDzcqNPp2r1wBQQEmDMyMi4CwMWLF0V33XVXOKUUy5Yta2UwZLVaIRT2fvxt1apVPv/6178qOrt/RUUFm5aWJv/ss8+KAWD8+PG6efPmNU6cOLGVOA8PDzcfPnw4y8PDg9u0aZNq6dKlQefOncvMz88XfvLJJ15ZWVnnFQoFnT59eui6detcn3zyydp77723bvPmzfkAMHv27JD333/f/YUXXqj++9//Xvn3v/+9EgC++uorl1WrVnl5eXlxl8/ftG7dOtdHH320rrvvQafuapTSCgD/A+BCCLkTgIlSOuDWkN7sCH3lkAx2g+7X0k470AlcJXBZFAaBtxS6X0qgP1zhsAhafIoEPAdcPNW1GwchBNGDRbCYKQ7uMLSUK2iG5+zCspmqMhvys66k75z9zYRd3+mwLVWHXd/pcWC7Acf3GnH2NxOyzllQnGNFcZ4V4fEiqN0ZHN1lwOkjV6KXElaGGPdk3Bn+RzyZ/A7+PjYVDw9+DROD5kPACHGoeCs+P/c6Xj14L1YeX4bN2f9BetUR6Cztrz1OHCGGUASc6sMo6dXIPAchZNbX8Ex5FoaKk8j74S7UnPu0X0x/rBUGMAoBWEXvXKTZy6VkuFr7e00YFi7hM+Ex5DHw2Uo0HToLEtwE6TAPeyR5QSjEMddPxFCeg7khH42521B54l2U7n8BtenrYG7IAQC7sB/1SosAFco8W4lRACCEgdglBC5hM+A1/HkoAyfa3wN9BQyVp8Db7OLc0lgIbeFeWPWVoJQis/YktJZ6DPeZ0unXLRXIsSTuWRisTfjyxEuoOL4SpftfQM2ZNdAW7YVVby/5pAyaCL/xb4Cw9vefFSn7zHG1B/wEYNvln18BqADobnQApdQG4M8AdgLIALCJUnqBEPI3Qsisy7vdBiCLEJINwAvAPy5v/xZ2r4N02NeZnqWUbnXoK+ol+HozpCcJ5EY5xKwUNivFr5v1EIkJKAVKch2Tot4Zqi8bGnn4OGYSjrca0JjzI5TBUyGQdj8g0ZS3DbzNBHVER8uQnQxEWFcxuDrHC9JbveSLI8VoZ9vT6XQ4dOgQPv30U3z99detnlOr1XTMmDHa1atXt/llXrp0adVHH33k1VHkzNXVtWVAqNPp2Ob72Wuvvea5YMGCYAA4fvy4NCIiIk6r1TLr1q1znTFjRn1ERIQFAPz8/Fouirm5ucKdO3e6PPTQQzUdvrjLxMbGWt56663ijz/+2AsAnn76ad85c+aEDB06NPquu+4KycrKEo0YMSIyMjIyduTIkZGXLl0SAfYo8D333BMYHx8fExwcHJ+amuoCAAaDgcyfPz84MjIyNiYmJnbr1q1KAFi1apXb/fffH9h83gkTJoT/9NNPyscee8zPbDYz0dHRsbNmzQppampibrvttvCoqKjYiIiIuP/85z/XDXA2bNigmTRpUlPz49GjRxujoqKuy4OfMmWK3sPDg7t8Pn1FRUVL+gDHcUSv1zNWqxVGo5Hx9/e3AsCiRYsaGYYBwzBITk7Wl5SUXJdykJqa6rpgwYIW8Tl//vyGr7/+2rWz73lbdOouQwhZCHtEdB/ss80fEkKeo5R+25OTO7kexR0BMJ2pheFAORRTOhcBY6QCqOaEQL+/DKa0anB1Zihv9wcR9WxW29WDhX+IABdOmDF0jKRLbroePnYX3tOHTDj7mwmFl1jwPIXZQGEx28Xp9HsUYBiCyhIbygpsCImyf+ZlCgaevgKIpQRiKYFESiCRMi2PT+43IjfDiuRxEoACMUPFEIntfTMZeBz+xQifQAH8ggV2Mw6BFJGuQxDpOgSAvRRHUVM28hrOI6/hAo6V7cShEvtY1UsegEjNYEwImgeV+Mp3SyJlEJ8iwZmjJgyfzMFF0zc1AK+GMAK4xd8PVfBUVB57E9VpH6Ax50eoI+ZCGTgBIpegPumHrcLQa+m6ACBwswtSW40JotArpVJosQfE9WNhDc6AINHcSoRRSgEKcOY6mGozYKrLhLkuG7zNAIBApAq6nIYbA5HKfj8QSNRd7lvzOcXqUPiNfwMg9s+BqS4TjblbgUvfQSD1wEFSCaVAiSjNoA7b5Cxa6MuOwlx3CbQhBxN4FXaZy3CIZzE5cCokrlEQqUPBsPbvx7Wi+WaAUvrd1Y8JIakADnXiuO0Atl+zbcVVf38Lu/i89jgOwNLu9rc/EfjJUTSrBnWltRCxEtjMgE+gAL7BApzYb0Jhjn0yri+oLucgkRHIVY6Z8GjM2w7equuRmZFVVwZdySEo/MdBqOhylpuTAQCr6Z1apLe6IO0PtmzZgmnTpiEyMhJubm5IS0tDUlJSy/Mvv/xy+fTp0yOXLVt2nQAMCgqypKSk6NasWeO2cOHCG7qNvvHGGx5r1qzxslqtzK5du7Iut101fPjwqPXr16vfeustn9WrVxcolUo+OztbYrVaybBhw6L0ej3z6KOPVjWXTnn88ccD3nrrrZLGxsYuDdJGjRplyM/Pb0kVvnTpkuTYsWOZCoWCTpw4Mfzee++tfeKJJ2rff/99t0cffTRg9+7duQBQXFwsPnv2bMbFixfFkydPjpo9e3b6m2++6UkIQXZ29sXTp09Lpk+fHpGbm3u+vXOvWbOm9IsvvvDMzMy8CABffPGF2tvb27pv374cAKitrb3utRw5ckQxf/78+q68xg8//NB9woQJjQAQEhJiffzxxytCQkISxWIxP3bs2Ka77rqr6er9zWYz2bhxo9vKlStblQDRarXMgQMHXNatW1fUvC0lJcV47ty5HjnLdXba8y+w1yCtAgBCiAeA3WhjIOCkZ4jDXSCKdIHulxLIJ/h2us4jYQnkE3whcJNAf6Acjd/kQXlnEFiXng1c4odJ8PNGHYpyrAiO7FpbEimDEZOkyLloRWWJDVIZA407gVhCIJHZZ/sBIGaIGHFJV24gETcYbFlMFGd/MyMkWtiSOusbdCVSZ7VSKF0Y5GVYkHvRArmSgW+QXZw2pw4LWRHCNPEI08QDAGy8FSXaXOTVn0de4wUcKd2B4+W7MTXkbozxvxPs5fVJg0dJcPY3E04fMuG2mf3n6ChUeMN/0nvQFu1D9ek1qDq5ElUnV0KkDoMycAKUgRMgcY8DuYGRTk+wVRghSejRRNgNYd0urzG6ynHadLEehgPlEEe6wXXaIhDG/tpMtRngGwQw7bJCOkGFmuJ/2dsQayD1HASJWwzErlG94sB59bo1VcjtkHknw1RzEVXVachtSEMSVaHiwP+D2DUKUvdYSNzjIJC4gvI26EoOQSj3gsQtBpSzoDHnRwjlvpD7jcY410g0Vu/BbxV7EKvxRaxbtMP7PgCIAODZ4V6/QwhDYBQYIBJKwBAGIgmQMMw+TgoIFaD4csZJX0TF7UsUWIeci1KK+sxN9u+DZ8cTNe1BBDLIfUdCFTa9x31y0j+wGnGvVAlwClLHk5qaimXLlgEAFi9ejNTU1FaCNDY21jJkyBDd2rVr2xwUrFixonzOnDnh8+fPv6Egfemll6pfeuml6o8//tj11Vdf9fn+++8LWJbF+vXr85OTk+Puvffe6qlTp+oBwGazkXPnzskOHjyYrdfrmREjRkSPGzdOd+HCBYm7u7tt7Nixhp9++qlLxqvXZhZOmzatQaFQUAA4ffq0fMeOHbkA8Oijj9a99tprLdGiefPm1bEsi4SEBHNAQID5zJkzkiNHjiieeOKJKgAYMmSIydfX15Kent7huthmhg4davzLX/4S8Oijj/rNnj27cdq0addlE1VWVgq9vLw6nS6zdetW5YYNG9yPHDmSCQDV1dXstm3b1Dk5Oelubm7cjBkzQtesWeP62GOPtUQ9//CHPwSOGDFCd+35v/76a5ekpCRdc7ouAAgEAgiFQlpfX89oNJpumVp0VpAyzWL0MrXovCGSky6inB6A2vfPw3CsCvLRnZ8BJoRAMsgNjEYM3fYiNG7MhXJ6IIT+3R+Mh0YLIVPYzY26KkgB++AqIl50Q5EpEHZ+sJN+3ASzkSJlfNtRIqULi2ETpLCYKSqKbSgtsCLnvAWXzlugVDPwCxLAN1gIufLKx1fACBHsEo1gl2hMxHxUG8qw5dI6bM35DMfKdmFO5EOIdB0MhYpB9GAxLp4yY9gEaafW1fYmysDboAy8DRZtKXRFe6Et2ova9M9Re24dBDJPKAJvgzJwAuTew1rSO3sKb7CBb7T0WskXwB7xJ3IBuFr74IKrN0P/aymEAQoobvcHYZvXKvOoz/4OQok/iDUZtF4ITcw9EKtDIZB59Xkaq0DqBkXAWPzGlYM2AGOjl0KmLYex5gJMNekQSD3gM/pVgLDQFu6G1CMBErcYCKRu8B33BljRlfvnXNdolOoKkHpxJZanvAc36c0dCSKEaNF6DWkFgBf6qTsDGlu1EaqzYshdlair4kAphaunXRQGhguRe9GK+hoerh69m6XBcxS1lRwSh3d6HHVDjNXnYK7LhPeoFT36bgokarjG3tPxjk4GLKxGBL7JCmrjOz3p3hm4ejOIiAGRO02uHEFdXR327NmD9PR0EELAcRwIIXj77dYWMitWrKhYuHBh2IgRI66bZUhISDDHxsYavvzyy5a00yeeeMJv165dLgDQHBVs5qGHHqp77rnnWtJaMzIyJDKZjK+oqGgZxPj7+1vc3NxsKpWKV6lU/PDhw7UnT56UpaWlyXbt2qX28/NzMZvNjF6vZ2bPnh2ybNmyqsceeywIAF555ZXS5ORk47X9PHr0qCw0NLRlu1wu75SouvZadqNrm0AgoDx/pVmz2dzmhz8xMdF86tSpi999953LK6+84rd79+6md955p/zqfSQSCW80Gjv15Tl27Jj0scceC9q2bdslb29vDgC2bt2qCgwMNPv6+toAYM6cOQ1HjhxRNAvSZ555xqempkawc+fO3Gvb27Rpk+vChQuvWytqtVqJTCbr9prBzl4JfiaE7CSE/JEQ8kfY1wFt7+AYJ91EHO8Kgb8cup+Lu1XSRRSogMuiMBAJC66xZ2kxrIAgNkmMgmwrmho6t661t7BZKU4fMSEgTAAv/xvfcERi++Bt5GQZpsyTIz5FDKGQIPOsBfu26lscgDnu+vfXQ+aLBwetwAOJr4CjVnxyZgW+TP8X6oxVSBorAccBZ472z1rSthAp/eAadx+C7vgUkXfvg+/Yf0DqkYDGnB9R/MujyE4dj9J9z6Mxbwc4S89mpW2X3Zx7y9CoGbvTrv09ZjViKG73h/LOoFaDF0IYeCY/BVXYRLDuEtjKDFD4jYJQ7t1vayp5yuN42S6EaxLh5zcWmuiF8Bn9V3iPeBmamLsv95vAe8T/a5W2eLUYBexR/PsTXgQArD//Jqxc75RI6CsopUpKqeqqn8hr03id2LFVm6DJVEBGFbh4yowzR80tUr655EtRTu+vHa+v4cDZAHdvxwjf+syNYIQKuIR2L7JJeRvqLmyAVVfmkP446T9anHYbHHtd4+rNYNTim2FN/U3Bt99+iyVLlqCwsBAFBQUoLi5GSEgIDh482Gq/IUOGmCIiIoy7d+9us+biq6++Wr569eqWWdUPP/ywNDMz82KzGE1PT28JaW/cuNElKCjIDNhTVZ955pnAPXv2ZNbV1Qk+//xzDWBfr/jbb78prFYrtFotc/r0aUVCQoJx9erVpZWVledKS0vTv/jii7wRI0Zot2zZkj9x4kR98/nuvffe6yK1WVlZohdffNF/6dKlVdc+d/n16detW6cBgLVr17omJye3RAy///57DcdxuHDhgri4uFg8aNAg0+jRo3UbNmxwBYBz586Jy8vLRYmJiaawsDDLhQsXZBzHIScnR3h1iqtAIKDNbrYFBQVCpVLJP/bYY3VPP/10xZkzZ64bcEVFRZmys7M7TAW4dOmSaMGCBWGfffZZfmJiYosgCA4Otpw6dUqh1WoZnuexZ88eZUxMjAkAVq5c6b5nzx6XzZs357Fs6+t/bW0te/z4ceU999zTcPX2iooKVq1W28Rice8KUkrpcwA+AZB4+ecTSqlzdruXIIRAeUcAbGUGmM7VdnxAG7AaMdT3hEMSZ8+isFUau12vND5ZDALgwonerTHaERfSzDDqKVJu61p0TixlEBIlwujbZZg8V46hYyQtUdmju4w42464jHVPwbPD/o1pofchs/Yk3j72GE40fYOQWCD9mBlm08ArtcGKXeASPhP+E99D5N0H4D/5QyiDp0Bffhxl+19Adup4FO18BPWZG2HVV3a5/RZB2otrSAG7066twtBSzkgcqQYRXn+5YoVyiFSBEPrIYK0w9EpN3q6QU38OdabKVmZGhBAIFd6QuEa2bGOEHb9/blJvLI5ZjlJtLrZc+k+7+3FWPSjfd0Y33YEQMpcQ4nLVYzUhZE4/dmngwlHwlAdDRKiv4RAaLQS5vH5fpWGhdmP6RJDWVDQbGvVckNqMtdDm74RL+KxOffbbwqqvgLEmHTZTl5ZNORmA9FbpF3sNUmfJF0eRmpqKuXNbu2HPmzcPqamp1+37yiuvlFdWVrb55icnJ5vi4uLarbu3cuVKz/Dw8Ljo6OjYDz74wOuLL77IB4BHHnkk4MEHH6xKTEw0f/nllwWvvvqqX2lpqWDo0KGmyZMnN0ZHR8cNHTo0ZsmSJdUpKSldihIUFxeLm8u+zJ8/P+yRRx6putZht5mPP/646L///a97ZGRkbGpqqtuaNWta1lT6+flZBg0aFDNjxoyI999/v1Amk9Hnn3++iud5EhkZGbto0aKwtWvXFkilUjplyhRdQECAOTw8PO7RRx8NjI2NbXlP7r333uqYmJjYWbNmhaSlpUkHDx4cEx0dHfuPf/zDd8WKFeXX9mnWrFmN+/fvb5nJfv311z29vLwSKysrRYMGDYpdtGhREAC8/PLLPg0NDYInnngiKDo6OjY+Pj4GACZOnKifOXNmfWJiYkxUVFQcz/Pk6aefrgaA559/PqimpkaQnJwcEx0dHfvss8+2lBL43//+px47dmyTSqVqNQDesWOHavLkyTdMy+4I0t81DXuD5ORkevJk5+sFDkQoR1H5/46DdRHB/aXBPZrx47RWNKzPhmSwW5dSgK/mpw1aVJTY8Kdn1WAFfT/7yNko1r/XCKWGwfwHVR0f0Akopcg5b4FExiAgTAiblSLtoBE+gUL4BApa1furN1Vja85nOFd1GC5CL4jSFmPSkJFIHt+7wsxRUJ6DsfoctEV7oSvaC0tTIQBA4h5vX3caNBEil9AOP2eN3+dD93MxfNeMcWiq1bU0fJ0D/a9lkN3mA/U94R32y5zVAN3OYrjcHQ6BR/+Z/mw4/zay687gldGfQ8g6ZmC0LedL7C36DotjliPZZ2Kr52ymehTvegxiTSR8x7zWTgs3hhCSRilNdkRfb3COM5TSwddsO325fuiAYKDcN4xnanBx614cGHEG0bUvYco8eatlDft/0uPiKTMe/n+aXr0WH/rZgHPHTHjkZQ0YtmfnqTn3KarTPkDo3M0Qq0O73Q5vM4KwEmcE7CbHWqpH1YqT0DwcA9lwxy0lr3jhGEThKrg+FOOwNvuTjIwMxMRceS39VfalPc6fP2+Ij4/PcFiHbjLmzZsXfOeddzb+6U9/6pdZsqSkpKidO3fmuLu792/6IoCpU6eGvfPOOyVXR2GbOXv2rPugQYOCO2rjhnmPbaz7aXkKAKWUOkYZOLkOwhIobvdH4/9yYLnUCHGkutttsUohFJN8IQzs0hrvVsQPEyM/y4rcDAsiE/reMCDzjBm6Jh4T5zhOABJCEHHVazHqeRh0FGd/M+HCSbtZUmCEEGo3BhqJB+6PfwGX6s5i86VPUJn4AbbW7oFf46PwcemfeqBdgTAsZF5DIPMaAs/kp2BpzIe2aA+0hXtQfepDVJ/6ECJV0OV1pxMh9Uhss2i9rdwAgYekV8UoALuo5Cnkt/l2avDZnEJs71//CFK9pQnp1Ucxyu+ONsUoz1NYLRRiSdfeu2mh96GoKRvfZa2BnzIUPopgAIDVUIWinUth1RbDffAjjngJvUlbL9q50KstLkdICRFDKifXrbEPjBDi3DEzygptCAjrvfp41eU2uHmxPRajlOfQkPUNZN4p3Raj5oY8iFyCb0qHaSfX0xsRUspTe4TU9dY1NOqJeHRy6/H222+X5Obmitzd3a9bE9uXmEwmMmvWrIa2xGhXuOGAgFLafQXjpMfIRntD+2MhtNuLeyRIAUAcbV9PTjkK3c/FECe4QhSo6PTxQeFCqDQMzp8w97kg5TmKtIMmePqyLWuoegOlmsVtM2VoqOFRlGNFaYEVRblWqNQMAiOE8A8RIsJ1EJ5O+QDbzv2Ig7aNeC/tCYwPnIPJwQtbitgPdAghEKtDIVaHwj3xQVgNVdAV7YO2aC/qLv4Pdee/BCNSQeadBLl3sn0g6RoJQphOl3wxm3g01dt/tA08GBadMkfhtFaYztZecUm0dG7ij1EKQWRCWMsNkCR2v75hT0ir3AuO2jDMt3XtUZ6nYBgCq4Vi7xYDxk6XtTLV6giWYXFv3LN4/8RT+DL9X1iW8i5YYz2Kdj4MzliHgClrIPcZ5uiX42hOEkJWAlh9+fHjANL6sT8DFspR8DwHBiK0ZZTtFywEw9rXkfaWIKWUoqaCQ2hMz6P8utJDsOrK4JnydLeOt+rKUXXyfbiEzYAq5PYe98dJ/0OkLIiYdagg5bVWgKNOh10nfcZ3331X0J/nnzhxor4/z9+MRCKhzWV3eoJzhnoAw4hZKCb7oemHAliLdRAGdF5Atgc12cA1mKHdUgD5OB9IBnVu8E4YgvhkMY7sMqKuioOrZ9/V4bx03oLGOh7T71b0eqoWIQQaDxYaDxaxSWKUFVpReMmK8yfMyDhlhm+QEInDxZg5aC6aPk1Cocsm7MV3OFW5D3eG/RGDvcbddOlkQpknNNELoYleCM6ig77kEHRlR2GoOAld0V4AsAtUzyQwlXMhiLQ73DaXldE18Wio5S4LUPtvs+lKYoVYQlqMUSxmCqOBb7OOK2+0Qbu5ALzOCtlYe2o5V2MGQjp+DYQQCH1ksFX0z0QhpRTHyn5BoCqyJYIJABfTzKiv5TB6qgyUt5clys+0ID6la86lKrEG98U/h49P/wVfn/sXRl86CmozIXDaJ5B6JDr41fQKTwB4BcBG2LNudsEuSp1ci42CAweWkYBpQ5CKxAQ+gQIU5Vgxupf0mb6JwmSgDlk/Wp+xEQKpB5SBE7p8LKUUDdnfgxFIIPcb3eO+OBkYEELAujq2Fqmz5IsTJzc3TkE6wJFP8IV2ezG0Pxc7ZF0EIxfCZUEotDtLoN9fBludGfJxPiCdSMuKTRLjtz1GnD9hwrgZfVOHk/IUJ/eb4OrJIjS696KjbSEUEQRFiBAUIUJDLYeiHCuMetqSwhYb7Yn6nf+HcXNux0n+c/zv4rs4WvYz5kQ+DF9FJ1TUAIQVKaAKnQZV6DQAgFVXAUPlSejLT8CQmwMpR1BW8CNOr/4FUTFVkHsn41LJeFRVqcGwBEoXBp6+AijVDFQa+8/VKaqnDxuhbeQxcZa8VSogb+bQtKUAXJMFqtnBLWlXzU67nUHgI4MltxG8wQZG1reXtsKmLFTqizE37HHkZVgQFCEEKyBQuNhfO89T+1rlUCGKcqyITBRDJO7axEWoOg5TvKdhZ/l2KCUM7hz7WSujpIEMpVQP4MX+7sfNAOV4WIkFAioFw7T9GQkMF+LoLiP0Wr5L0fbOUl1hN8nqqSC1NBVDX3oY7oOXgjBdv36bai7AVJcBdcQ8sKKeT8g6GTiwGpFTkDpx4qSFAV9LlBAyhxDyH0LIRkLI1P7uT1/DyIWQj/eB8XgVbNWOif4QEQvljEBIkjxgTq9F05YC8MaOXTqlcgbhsSJknLHAaukbM6y8TCvqqjkkj5O0OE32B2o3FonDJRg2wR7ZspgpygqtkCsJio8G4Ymhb2Fe1GOo0BfhveNPYXP2JzBYr6tlPOChlMJmpWis51CcZ8WlLA0y8ififPlz8Ij+AAAgiLodBnY8DLVFqDz+NkRFf0IAdz/iPV5BfPD3iI4qQGiMAB4+guvWS8YmiTF0jLSVGKU2HtqfCsFVm6CcHgChvxyMTAAiE4Cr6bwgFQYrIBvjY1/h3odQSnEwbycEkKDmyFBcSDOjutyeahwYLkRskrhFWITFCsFxQEF218sdGCpPISBtPYLNHI4rZShnBraz7tUQQnYRQtRXPdYQQnb2Y5cGLJTjYSM2sFTSZsouYF9CAfRe+ZfqMg4ggJtXzyZ26rO+AQgDdeS8Lh9LKUVDzmYIZF5QBIztUT+cDDxYjRh8vePKvjgFqRMnNze9GkYghHwG4E4AVZTS+Ku2TwPwAQAWwDpK6b/aa4NSuhnAZkKIBsA7AH7pzT4PRBRT/aH7tRS6X0qgvjfCIW0ShkA+2hsCNzF0u0vRuCkPypmBELjeOJUwfpgY2ekWZKdbEJfUuxd+SilO7jfCxZVBRPzAsHJvTscViQnGTZejIMuK/dsMuHjCBlPVWCwKTcFFZhMOl2zH6coDmB52P1J8JoNpb2TZy1CewmKhsJjtZjoiMYFRz6M4zwr/ECFkCgZVZTZknjbDbKawmCiuqtsMhgFUagae/gLYKu0O5ZF3DEOMUgRgTOsIasVJVJbssR8nUkLmlQSZTwrk3skQayJBGBZKlysRl7oqDmpXAt32IthKDVDc7g9RyBWfNIGbpGsRUldJh59fR2KzUpTkW5FzqRHp4kPwNo9CaJgSQRFCqNpISQbs65Q9fVkUZFoRFiPqtEuqrvQwSn59CkK5F/4w8gP8O+Nf+O/5t/DUsPehFKkd+Kp6DXdKaUPzA0ppPSHEcfaatxBWqwU8w4HlJWjDVwyAvTaoVE5QnGNFzBDHX4drKmxQuzJdjuJfDW8zofHSD1AGToBQ7tXl423Gatj0FVBHLQRhnMlctxqsRgyu0QzK0U5laHUEV28GWAJG2beZVE6cOHEMvX2V/wLAvwG0WIMRQljYjS2mACgBcIIQ8iPs4vSNa45/gFLaXKj2ZVwxxPhdwWrEkI30gv5gBZQzg8CqHCfOxNEaMGoxtFsL0bQpD4ppARAFt+9l5RskgKsni/PHTb0uSItyrKgq4zBxtqzHTo+9gUrDIi6FwalDJlw8ZUFAmBB5Z0SQMffhdr/bkC74HN9k/hu/le7E3KilCFT1LL2SUgpCCHiOorGOh1ROIJExMOp5FF6ywmyyC0/LVb+tFtpikz1ohASB4UKYjBRZZy1Qu7KQKRgIBPZ2VK4EIrH9RyonUGkYyJVMS3Svfr0JjEIAVnnl8ydUeMNFcSdcwu4EgOsEqq54H4DLAtVzCBiREgBFo84dFy6NRZypBC4GDogqhLGsCigDrhh7JwOFcpTs+fzKNgpceUX234xACrFLKETqUAjFIWCtbhAFqnv0Xt+IpnoOBdlWlOZbYbMB9e5HwTNmzBlxB0LdOhbEYbEiHN1tREm+FUERHX+Xmwp2o3T/8xCrwxA49WMIpG74Q/xLWJX2HP534R08PPg1MKTv1nR3E54QEkgpLQIAQkgw2nZw/91DRimwh/8Vd8Y8iBSfto3SCEMQGCZEUa4VlKcOzx6pLufg6duz4UFTwS/gzI3QRC/q1vHm2iwAgMQtukf9cDIwYTVigAf4JotDoppcnRmsRtyvmVROnDjpPr0qSCmlBy4PPK5mGIAcSmkeABBCvgYwm1L6BuzR1FYQe0jqXwB2UEpP9WZ/BzKKaQEwHK6A/tdSqOY6dn2i0FsGl8Xh0G4thOlMLYRBCoAC5ov1EAYpWgkQQggShomx/ycDKkts8PLvvY/Qif0mKFQMogcP3BQcliUYMkaCA9sMGDdDhsQRYhRdsqIkLwBhlpfhpj6KHOYrrDr5LIZ6jcfk4EXwlPvDqLc7z4olDDgbRWmBDRazXUBaLRRWsz2yab28zWIBQqKEiB4shtVKcWinAfEpYoREiWC1UuRcsLSISZGEQKlmIJaQVts07nbRonZlMONuRYvId/VkMcyzY4dgW3nHDrs3EqjG6nOgnD2tSsgTBDf4gavzRn3QcUglh4A6AISgOedWIPADq4uHuSHfvrk5F7fFNIoAhIC3aNGUt93ebsUICOoTwA/fC5Fr6GU34TCI1KEQqYLAdLM26NWD/vTjZjTWcfANEiIoQoj1BfvhzQchxDWqU225ebFwcWWQl2FFYLjwhiZYDZe2oPzwq5C6JyBgymqwYnsE2VcZgrsiH8GmzFXYmfcV7ghb0q3X1Yf8BcAhQsh+2P+RYwE83L9dGphYYIZVaIVUKLthBD0wQoiscxZUV/RcPF5Ns0N2bFLPJjnqMzdC5BICWTcdoE31WWAlrhBIPXrUDycDk6tLvzhEkDZYwGoGRiZVb1H+1BHwTY5L02dUQvi8N6rD/TZv3oy5c+ciIyMD0dH2CaKCggIkJCTIXn/9dc+//OUvVQBw//33ByYnJ+uffPLJ2nnz5gUfPHhQVVhYmC6VSml5ebkgOTk5prS0NP3a9v/61796/fe//3VnWZa6ubnZvvzyy4LIyMh287krKirY2bNnh6Wnp8vnz59fu379+iIA0Gq1zMyZM0MLCwvFLMti6tSpDWvWrCkFgEuXLonuu+++4KamJgHHcfj73/9eumjRosZr2y4sLBT+8Y9/DNq7d2/ODz/8oHr55Zf9rFYrEQqF9I033iiZNWuWFgCGDRsWVVVVJZRIJDwA/Prrr9l+fn4t62i++OIL9Z/+9Kew/fv3Z4wbN85gMpnIfffdF3Tu3DkZIQTvvvtu8Z133qkFgP/85z+at99+24fneTJlypSGjz76qBQA/vnPf3rIZDJ++fLlPXawvRnojzwYPwDFVz0uATD8Bvs/AWAyABdCSDil9OO2diKEPIzLA5zAwEAHdXXgIPSRQTLEHbo9ZVBMCwAjdey/jlXazY4ob4/C2WpN0O8phWKqP9hoEWx1JpjO1ELgKUWEnwRHhRTpJ0zw8u8do4nSAivKC20YN+PGg7KBQOxQMY7vNSLtgAkzlygRl8wiejBFRYkNhZfGQlk6BEXKLTiHX3C68gAGeY6BOGMGYkNCEZdsj6id/c2emkoIIBIRCC8LSYmMgVINCMV291/A/vywCVKo1PY0YKULgxn3dN6BmDCkW8ssbRVdL6lyrUBthjdzaDTlosBTjArXFIyZ/GyrdF4A0ClL0Jifi+Apm8B2kIbFWw0wNxbAVJwHS00xLPCDqTYD2oJdaAnEERYiVSDE6lCI1GFXxKoqGIyg/QFRRbEN6cdNGDdDBrGEQeIIcUv6c5k2H8XaS5gd8VDn339CEBYrwqlDJlSWcPAOaPu7XHcxFZXH3oDMZzgCJn0ARth6MmCY72QUNGbg18JvEOQSjVj3lE6dvz+glP5MCEmG/Rp9GsBmAP1aO22gYr7YgNCSUNSqBCgxWOEf2vZnv7nkS1GO1aGCtKbCvv65J4ZGxpqLMFWnw2v4i91yHaeUh7kuG1LPQTeda7mTzsE0C9I6M9C98rSt4OrNXSpldzPiSDHalfZSU1MxZswYpKam4rXXXmvZ7urqirVr13o+88wz1RKJ5LqMF5Zl6apVq9xfeOGF6hu1n5SUZHjmmWcylEol/+abb3o89dRT/tu2bctrb3+ZTEb/9re/lZ09e1Z6/vz5VrPpzzzzTOXMmTO1JpOJjB49OnLTpk2qhQsXNq1YscLnrrvuqn/hhReq09LSJLNmzYpYtGjRdeL4n//8p9f//d//1QCAp6enddu2bTnBwcHWEydOSGbMmBFZVVV1rnnf9evX540bN85wbRv19fXMv//9b6/ExMSWsizvvfeeOwBkZ2dfLC0tFUydOjXijjvuyKiurmZXrFjhn5aWluHr62u76667grds2aKcPXu29oknnqgdNmxYtFOQDhAopasArOrEfp8A+AQAkpOTb8lUMOUdATCdqoH+QDmUtwc4vH0iZFqECusmhvr+SJDLwpdvsMCS3Qjz+ToAwHgzj6r9LBpkakj8ZBB4ScG6ShyyFgQATuwzQionvZ4W7AiEIoLBIyX47VcjaipscPcWgBUQ+AUL4RcshK5JguLc+0FFc1Ag2YHDJdtgdjuIBjIcau1i+CpCMXmuHEIRAStAhwMwwhB4+V356vbFgI032MA3WSHw6bgGaXtYOQsEl502GTEL9eIwRFmA6p+NOLHPhDHTZK3WrLHudrHO1Zo6FKSMUAapeyyk7rGt+20zwdJYAHNDLswNebA05MHckAtt0T6AXq5xShgIFf4tAlWsDkWjOQoqr2Co3UWQq+yTAdzluc+rhfPx8l0QMEIked/WpffCJ1AAmZygtMB6nSCllKL23DpUn/oQisAJ8Bv/VruCeW7kwyjR5uBM5cEBLUgJIQ8CWAbAH8AZACMAHAUwsR+7NSDhCgzwq/JHk4cQtTauXUEqVzJw92ZRlGNF8jjH1UCuKW8WpN0fHtRnbgQRSOASPrNbx/MWHYQKX0jcYjve2clNydUR0p5CKQVfbwbbyTJ2TjqPTqfDoUOHsHfvXsycObOVIFWr1XT48OHa1atXuz3zzDM11x67dOnSqo8++sjr6aefvqEgnTlzprb57zFjxug2btzoBgDr169Xf/TRR56HDx/OLi4uFo4fPz7qwIEDmYGBgbbbb79dl5WV1erGqFQq+ea2JBIJTUxMNBQXF4sA+zipqamJBYD6+nrW09OzTTW+bds2zfvvv18KAKNHj26ZNE1KSjKZzWbGaDQSqVR6Q43xzDPP+D377LMVK1eu9G7edvHiRemECROaAMDPz8+mUqm4AwcOyBiGQXBwsNnX19cGAJMmTWr65ptvNLNnz9YqlUre39/fvHfvXtmECROuE763Gv3htFIK4Go15X95m5MOEIWqIIpWQ/dLCaiV7/iAHkAIAasWgxGzLefWLI2B+v5IKG4PgHqEO6yUQe3Jeuj3lKIxNQd1H12wz3YC4BrMsFUbQWnX5wYqS2wozrVhyGgJBMKbY3Y8YbgYQhGQdvB6Ex6FikHMEDFi4zwwPex+/GXUOkwJXoxC/Xm8d+IpfHbudVRzORAIyYCNBtgq7NdCgXfXBr6UUuTUn8OnZ/+Ol/bPxw9bPkDF5nRQGw8iYiFTsEgeJ4VRz+PUQSN4/srnhb28HrMrTrsAYCnUwpxRDwBgBBJI3KLhEjYDnklPwH/Sewi760dELTmGkDnfwe+2t+E+6GFI3KJg0RajJv2/OLHtCA7/mI20bz5E7fkvIZMakDxOCpmi9eXSypmRVrEXCR4jIRO2v+66LRiGYMRkGYaMbr3mlFKK6pPvo/rUh1CFzoD/hHduGL0VsmIsHfJ3LI5d3qXz9wPLAKQAKKSUTgAwBEBDv/ZogKKfABwcegDDRmswaMSN1yQHhgtRXmRP93cU1RU2SOUEMkX3rkWcuQlNeTvgEnonWFHXvhfNsGIVPJOXQ+Y1pFvHOxn4MAoBICCOEaR6G6iFbykX5sRxbNmyBdOmTUNkZCTc3NyQlpbW6vmXX365/MMPP/S22a53fQ8KCrKkpKTo1qxZ0+mZgrVr13pMnjy5EQDuv//+Bk9PT+u//vUvjz/+8Y9BL730UllgYGCn7OVramrYXbt2qe+4444mAHjjjTfKvvnmG1cvL6/Eu+66K2LVqlVF1x6TmZkpcnFxsbUlOL/88ktNXFyc4ernHnzwweDo6OjY5557zoe/7AZ56NAhWWlpqWjx4sWt0oEHDRpk+Omnn9RWqxWZmZmi8+fPywoLC0WxsbHmvLw8SVZWlshqteLHH3/UlJWVteSeDx06VL9v377uXUhvMvpDkJ4AEEEICSGEiAAsBvCjIxomhMwkhHzS2HhdWvgtg/KOAPANFhh+q+zzczeLVHGUGt4z/VAb64ejcn+4LImAYloAJIPdwbjYv0emc3Vo3JTXki1pqzOB2jonok/sN0IsJUgY1neOqT1FImUQnyLBpXQLGuu4G+4rEypxe+g9+MvI/2BayL0obMzEqpPP4j9nXkV+w8U+6nHXuCJIOxchtfFWnCzfg/dOLMfHp19GUVM2RvjejjpdBQ4XbcePuZ/BeLksjqsni4RhElRXcMg4dWVwIrgcIe2K0y4AmDMaYDhy4+8Hw4og0URAFXI7PIY8Bv8J7yJwxvdoCtgHvWo5wgd5ITTwIqpOvIucjVNQcexNWLQlrdpIrz4Ko02P4b7dq0bVbBhFL4twSnlUHH0dtec/hzpqAXzH/aNTtRvlQlW/uTh3AROl1AQAhBAxpTQTQOcW3f7OMNvsk/JituPJn8AIIXjOvsTBUdSUc/DwYbs9OdaYswWUM0ETvbDbfaB875SzcTJwIISAdRU7RJA6S770HqmpqVi8eDEAYPHixUhNTW31fGxsrGXIkCG6tWvXurZ1/IoVK8pXrVrlzfMdj//WrFnjevbsWdlrr71W0bxt3bp1Re+//76PSCSiS5curetMn61WK+66667Qhx9+uDI2NtYCAJ9//rnr3XffXVtZWXnu+++/v/THP/4xhONaj9WKi4uFrq6u1wnekydPSlasWOH3n//8p7B528aNG/Oys7MvHj16NPPIkSOKNWvWuHEch6effjpg1apVxde2sWzZshpfX19rQkJC7OOPPx4wdOhQHcuy8PDw4N57773CBQsWhKakpEQHBgaaGYZpEb2enp62srKy34V1dG+XfUkFcBsAd0JICYBXKaWfEkL+DGAn7M66n1FKLzjifJTSrQC2JicnP+SI9gYi4jgNhAEK6H4uhmy0d786yiUMk2D393pU1LPwj1RDfJWJrGSIG4TBShCG2OtMbi4AEbNQ3BFww9IcNRU25GdaMWyCtEclB/qDwaMkOPubCacPm3DbTHmH+0uFCkwOWYQxATNxpHQH9hdtxupTLyJck4jJwYsQpo4fMBFTa7kBYEmLSGwPvaUJR8t+xuGSbdBa6uElD8CC6D9jqNdtELIi6EIb8XPuBhwq3YpTVfsxPWwJUnwmIzBciKYGDnmZVqg0LALChPZapFK2yxFSgbcUluwGcFprh6m+zeiaeJzYZ4RBRzF4jAZBEV4AUmCsuYi6C/9FfcZG1GekQhkwAa7xSyD1HIJjZbvgJvVGqDq+w/bbo6bChtOHTRg5WYjGM39FU942uMX/CR7JywfM/95BlFyuQ7oZwC5CSD2Awhse8TtFkGZBaG0YLokZmH2sCGgnZRcAfAMFEAiBoktWhET13NCFs1HUVnEYPLJ7k4GU8qjP3ASp5+Buu+PynAVlB16CS/gsKAPGd6sNJzcHrMYpSAcydXV12LNnD9LT00EIAcdxIITg7bffbrXfihUrKhYuXBg2YsQI7bVtJCQkmGNjYw1ffvmlpnnbE0884bdr1y4XAMjMzLwIAJs3b1a+8847PgcPHsy6OgqZn58vYhgGNTU1Ao7jwLIdr22/5557gkNDQ00rVqxortKBDRs2uP/888/ZADB58mS92WxmKioqBFcbEclkMt5sNrea3c3NzRXOnz8//NNPP82Pi4tr+bCGhIRYAUCj0fCLFi2qO378uPzuu+9uuHTpkmTixIlRAFBTUyOcP39++Lfffpszbtw4w6efftoiVIcMGRIdGxtrutzfxnvuuacRAN555x33q1+jyWRipFJp76ZEDhB622X37na2bwewvTfPfatCCIFiegDq12bAeLwK4mg1qIUHtXBXfpt5UAsP/uptrX63tb/9MQigXhIBcaS6w75ExItwcIcB6cfN161zYpWiFndeImAgn+QH3S8laEzNhfw2H4hjNW0OuE8eMEEoAgaNuPluLM2puRdPmTFswvUpnu0hEcgwMWgeRvvPwLHSndhb9D0+Pv0XhLjEYkrIIkRoBve7OLFVGCHwkIAI2n5NVfoSHCj+EScr9sDGWxDlOgTjApYj0tXed8PxKnAKIRSxGsyPeRwj/Kdhc/ZafJP5bxwt/RlzIh9C7NBo2CyASnPlHAJ3Cbjarg1YhD72yQBbuR6sUt3h/lVlNpw6aAJhgBGTpHDzunJZlLrHwm/8G/BMXo76zK9Rn/kNtEW/wuQRg1yZEXeE3Nuj6KRCxUClpig99Dr4qm3wSHoS7okPdru9gQqldO7lP/9KCNkLwAXAz/3YpQELW8TDzeaG2mIBNB2MQ1gBgV+IEEU5joko1lVz4LnuGxoZyo/D0lQI38GPdLsPlLdC4T8OIuWtZ07opDWsRgxLTlOP23EK0t7h22+/xZIlS7B27dqWbePHj8fBgwdbmYcOGTLEFBERYdy9e7dLSkqK/tp2Xn311fLZs2dHND/+8MMPS3HVUr3Dhw9Ln3jiiaDt27dfulogWq1WPPDAA8Fffvll3ueff+722muvef3tb3+7YfrTk08+6dvU1MR+/fXXBVdv9/X1tWzfvl315JNP1p46dUpisViIj49Pq2hoQkKCubS0tGVmr6amhp0+fXrEa6+9VjJ16tSW12W1WlFTUyPw8fGxmc1msn37dpeJEydq3dzcuPr6+rPN+w0bNizqnXfeKR43bpxBq9UylFKoVCr+hx9+ULEsS5OSkkwAUFpaKvDz87NVV1ez69at89y0aVNucxvZ2dni0aNH6270mm8VBrypUVcghMwEMDM8PLy/u9KrSJM80OSRj/r/ZHbtQJaAiBgQEQsivvz78mNGJQIRMbDmaVH/eTa8/pYMIrzxQFsgJIgZIsa530zQa3nIle3vLwpSQn13OLS/lED/aymsxXrIJ/i2rFEFgIZaDjnnLRgyWgKJbMCnILbJ0DESXEwz48wRE0ZN7ZoBkJiVYFzgbIz0m4Zj5buwt/A7fHLmVQSqIjEleDGi3ZL6TZjaKq4v+UIpxaX6szhQ/CMya0/azX28JmBswCx4K67crHiDDcYT1RDHqIFY+ySpvzIMjw99E6crD+CnnM/x77QXMNTrNsxI+gNcxPblJpyNgnWTwNbFCCnrLgEEDGzlxhtOrFBKkZdhRcYpM5RqBim3tT+JIJR7wTNpGdwTH0Jj7lZsz9sAQgG3k/9BjVYPTdT8lpIsXUEkNMHb+jQMVcfgNfwluMa2OYd3S0Ep3d/ffRjI8DYOHMOBcGIwnbgMBoULcSDbisZ6Di6anpVqaXbYde+moVF9xkawYg2UwVO63QdWKIc6Yna3j3dy88BqxOAazD2upcvVmwGCliVDtyqMSujwsi83IjU1FS+88EKrbfPmzWtz+yuvvFI+evToNl3IkpOTTXFxcYYLFy60OSh67rnnAgwGA7tgwYIwwC4e9+zZk/PSSy/5jBgxQnv77bfrhg0bZhg6dGjMnDlzGocOHWry8/NL0Ol0rNVqJTt37lRv3749W61Wcx9++KFPSEiIKS4uLhYAHn744aqnn3665r333it+6KGHglevXu1FCMHHH39cwFxzgVWpVHxgYKD5/Pnz4vj4ePNbb73lWVRUJH7jjTd833jjDV/AXt5FqVTykydPjrBarYTneTJ27NimjoybysrKBLfffnskwzDU29vb+tVXX+U3P/fII48EXLx4UQYAL7zwQlliYmLLLPyJEycUb775ZtmN2r5VIN0xnRnoJCcn05MnT/Z3N3oVS5EOlsyGKwJTxFwnMlv9FjLtRreuxnShDrUr06GcEwzVzKAO92+o5fDf9xsxYpIUKbd1vOaJ8hTGtGoYj1aBcRFBOc0fAi/7NerXH3TIOmfBH59Rdzq6OBDZsVGHoktW/PFZF4gl3X8dNt6KE+W/Yk/ht6g3VcFfGY7JwQsR6z6sT9cMUp6i7NGDUEz2h8uCUNh4K05V7MfB4i0o1xdCIXTBKP8ZGOk3DUqR+rrjDUcqYTxZDfWSiDZnsM02I34t/Ab7izaDZQSYHLQQXvXTUFdOEF9fBuORSvj8e3SXxHjjd3mgVh7qxe1PTlWV2nBsrxHeAQIMGdV5Ay2Ot+HvRx6An9ANU+u1MJQfBxFIoA6fDU3svRC7BHeuHXMTinc9DmNNOtRJb4B1m+LQ8h2dhRCSRilN7vMTDzAGyn3j7HtbkSE9C7liGaISRYhMvHHUp76aw4ZVjZgwS4b4lJ6tuz+4XY/zJ81Y+rIGTBcFglVfgZxvpsEt/k/wTF7W7T6YG/MhUvp3av20k5sb3a+laPwqB97vjQSr6r6YrP8sC6aLdfB5Z6QDe9f/ZGRkICYmpt3necrDzBkhYWUghMBo1YOnHCQCOVimZ5NTneH8+fOG+Pj4jF4/UR+yfv169cmTJ2WrVq3qdxF4+PBh6dtvv+29efPm/I73HricPXvWfdCgQcEd7XdLRUh/T4gCFb1Sc0sS5wppsge024ogG+EJgceNRabajUVAmADnT5qRNE7S4SCGMASyFE8I/eTQ/VyMxm/yIBvtDUuwGplnLIgfJr6pxSgAJI+VIOe8BenHzT0qxyBghBjpNw3DfCYjrWIvfi38Bl+k/xM+imBMDl6IBI9RfSJMuRoTYKPgPIBf8r/G0dLt0Foa4CMPwsLoJzHEaxyEbNuDCWrlYUqvhShM2W46lVggxfSw+zHMZwq25nyK7XnroRb+giTN/WCZIFATB6q3gSg6P0AV+shgTKuxu/leMxFDqb3Wrocvi+RxEngHCLokdjNqT0JnacDo6CcQ5J4CU10W6i5sQEP296jP3ARFwDi4xi2BzDul3XZtxloU/fIIzA258LvtHWTmj0ZDpgmT5soHfN1dJ70L5XgwAvtgsjNRI7U7A6ULg8JL1h4L0uoKDm5ebJfFKAA0ZH0HUAp11Pxun58zN6HqxLtwCZsFVUj3zMKc3DxcXfqlJ4KUqzf/rtJ1m+9hVt4CnaURrJiFiJVAKuzYu8LJjbn//vsbampqBoQ2qqqqEr755pu/myokN/fI/xp+Dy67fYHLojAQBmhMzenU/vEpEugaeRRkdT6VROgrh8vd4RAGKWE4WI7zPzcCxJ7yerPj4StAYLgQZ4+aYLP2PAOBZQQY5jsFzw//CHfHPgUbb8V/z7+Fd449gRPlv6KkKQd1xipYuJ6bQ7RFdb7de2Z91bv4Jf8r+CnD8fDgv+HpYaswzHdyu2IUsEfcqZmDZKhHh+dxl/ngT4kv46FBf4VIKMCvxrewT78FQNeddgXeMoBS2CqNrbY31XPYv80AXSMPQgh8AoVdToM+VvYLVCJXRLkOBQBIXKPgO/bvCF+wE+6DHoax6hyKfn4Q+T8uQkPOj6Bc6++FVVeBwu1/hKWxEAGT/w1V8GSExghhNlOU5DvdRX/vUNsVQdqZlF1CCAIjhCjJs4Hjun+9oZRedtjt+liMclbUZ38HRcBYiJR+3e6DuT4bgP075eTWp7lMS0+NjbiGW1+QUkphthnRaKqFwWr3DhIxYqjF7hCxN/+4aSDx9NNPX1dTtT+YO3duU1RUlKW/+9FXDIhZAEfxe3DZ7QtYVzGUs4LR9E0ejGdqIB3sfsP9Q6KFkCsJzp8wIzSm87OcjFQA5Z2B0ObocPYrK6IHiyFjedjNl29uksZJ8MNnWmScNjusfA3LsEjynoAhXuNwruoIdhVsxMaMD1rtI2REkAtdIBepIBeqoBDaf8tFLvbfLY9VUAhdIBMqwJDr329KKbLrzuBA8WbITzIYj9sQHJGIuyNfhJc84Lr924LyFKbTtRD4yCH06fx62ii3oXhGswqHS7Yh7eQuJCEB+0/8hLF+syERdK4dweXz2coNEPpdmTUWighYFuD47g3cG0w1yKw9hYlB869LiRLI3OEx9HG4Jf4fGvO2oe7CBpQffBlVJ9+Ha8xiqKMWgDc3oXDnw+AtWgTe/jFkXnZR6+bFQu3GIPeiFYFhwn51z3bSv1COB8N2XpAC9nqkF06aUVlig29Q91JdtQ08zCbaLUMjbdFecMYaaKIXdevczZjqssAIZBCqOneNcXJz0xIhreuhIK03Qxyr6XjHmxAbb4WFM8FkM4CjHFjCQkTsYwpCyA0nhJ04uZm4pQSpE8ehmOwHw+EKNH6VA3GMppX50LWwLEFskhgn9pu6bKxBCMG5XBY8Z8WQSB4NX2RBeUcgRGFdN4gZSPgFC+AdwOLUIRPiksRgWMcJDIawGOw1Fomeo1GizUGTuQ56axP0lib7b2sTdJd/1xjKoLc2wcwZ22yLgEAqULQI2OafwqYsVOqLoBRpMI8uBlEIMGfI0i7103KpEbzWAvl4ny6/RpYRYFzgbCRIR8Gw7SKqKrPxxpFHMCP8fiT7TOwwVZmRCsCoxeC0VlBKUVpgg1+wAFI5gzHTZN02hzpRvhsUPIb5Tm7/3AIJNJHzoI64C/qyI6i78F9Un/o3as7+B4xAAoBB4LR1kLpf8X8ghCAsVoS0gyZUlnLwDnBemn+PUJ6C8jyIQABwQGcz8v1DBSAEKMqxdluQVpdfNjTy7vpnz1h9FoSVQO43ulvnBuyTYKbaTIhdI0EGfl1dJw6AUQoBlvQoQsobbaBG7paJkHK8DeW6AhQ1ZUNkdUO9qRoEgJARQy5UQcRK+t1134mT3sA56nHSJkTAQH1fBGreOgvdtiKo7gq54f7xyRKcPGDC+RNmjO6Cu6zRwOP8CRMiEkTQhEthqHWDwP/mXwdBCEHSWCm2faXDpfMWRA1y/M2SIQwCVZEd7wj7LGuzaNVZG1uEa7OI1Vkbobc0oc5YgaKmbLiI3bA4ZjkGe41F/eELgHfXIoqUUhjTasBqxBCGKLvz8gAAand3GMQsIuumIt/8LTZlrsLRsp8xJ+IhBLncOK1PfU84OEpwcr8JFSU2sKykWym6zfCUx/Hy3YjQDIKb1LvD/QkhUPiNhsJvNMwNuai7sAGm2gz4jvsnxOrQ6/b3DhBAJifIvWhxCtLfKxwFT3mwrBDg0Om1nBIpAy9/AYpyrBgxqXunrqmwgRB7tL6r2Iy1EEjdeiQkbYYqcOZ6SFyda0d/LxCGgHURgavvflbirVDyxWjTQ8xKwRAGJyv24Hz1MQgYIRLJZChFaogYMZg+MCly4qQ/cY56nLSLOEoN6QhPaHcWQzrKC0Lv9oWmwoVBSJQQF9PMGDFR2mljlrNHTbBagORxEhARC/lYezSN2nhofyyEZKg7RMHdFzT9SUiUEK4eLNIOmhCZKOrXWU0BI4SL2K2lpEpXsFUYIBnUteNspXpwNUbIJ/n16HUTQiDwkMBfKsbg6hXQex5FpikVH6Y9h2TviZgedj9UYtc2jzWagBP7DNA28IhLEvdY5OXUn0W9qQozwv7Q5WPF6jD4jH71hvswDEFojAjnT5pRV8XB1dM5APm9QXkKK2MBKxSCBQHbhY9sYLgQx/cZYTLw3SqbVV3OQe3GQCjq+ve1WZD2BHOdff2o2DW6R+04ublgXcU9ipBeEaQ3T+oqpZcnnhgWpdo87Mz/CjPC/gAveQCiXIfATxEKb0UQLmXldHqZykBAJpMNMRgMpwsKCoSPPPJIwM8//5x37T5X1+Zsr52//e1vnk899VSNUqnkAWD8+PHh3333Xb67uzvXm/130r/cUnkxTlMjx+OyMAxEyKDxqxx0VCIoYZgEJgNFzoXOzXZaTBTnfjMjNEYIN6/WIy/eyIE32qD9sQD6g+WgPTDr6C8IQzB0rAS1lRwKsm9OsxreYAPfZL2uBmlHCHzlUEwPhDhK3eM+CNwkYHVmJA6TQVYxGrPF72NC4DycrjyAN397FHsLv4eNb/3+1lbacHirFspzpUgJtyI0pucTAr+V/QKZUIl4jxE9audGBIQJIRQR5Gb8bnwMnFwFI2ax67bdMMSYMW2hAn7BnU+/DQwXABQozu3etaamonuGRgDAGWvB9lCQmuoywErcIJDe2LPAya0Fq+mpILW0tDOQ4XgOZdp8/Fa6E99k/hsXao4BANxlvhjiORZyoX2ZklriAX9VOAQ3cdmj4OBga1titLOsXbvWS6fTteiT/fv35zjF6K3PLSVIKaVbKaUPu7i49HdXbhlYFxFUc4JhvlAPU9qNjccCQgVwcWWQfqJzN5dzx00wmyhSxl9fGoVVCuGyKAziBDeYTteg6ds8cI033yA9MlEEpQuDtINdc4kdKNgq7JOYAu+ula8hDIE43KVTtW87gnUTw1ZjQmC4ECFRQhRnCZBA7sGzw/+NUHUctuV+gZXHlyOn/hwAoPCSBUd3GyGQsQjwZ6B26XlkWmdpxIXqY0j2ntirAwWBkCA4UojKYht0TXyvncfJwMXCmSBmu14uystPALGEoCin64LUZOChbeDh3g1DIwCwmWohkHRfkFKeg7k+BxLXKOf6uN8ZrEYMvt7c4YR3e7RESNUDT5CabAbk1KdjT+F3+Oriu9iRtwGZdWlQS9xbspXErARDvMdDIRp449YXX3wRq1evbnn817/+Fa+//jomTZqEoUOHYs6cOdINGzaorz0uKytLFBEREQcAOp2O3HnnnaGhoaFxU6ZMCTOZTC1f8HvvvTcwPj4+Jjw8PO6pp57yBYDXX3/ds6qqSjh+/PjI4cOHRwKAn59fQnl5ueByH7wiIiLiIiIi4v72t795Np8vNDQ0bvHixUHh4eFxo0ePjtDpdM4LyU2GM2XXSYfIJ/jBcKgSDV/nQBzvCkbS9qCFMATxKWIc3mlETYXthuYYVgvFmSMmBEYI4enX9n5EwEAxwRfCADn0u0vRmJoD+SQ/iCMG3oW7PViWYMgYCQ5sM6CssPuGI/2FtbxZkHY+QqrfXwZGKYJ0qGMiHay7BNTIgTfYEJskhq6RR/oxE0ZO9sL/DVqBCzXHsTn7E3x8+mWECEbDq+RuBPi4Y+gYKYSicIf0Ia1iLzhqwzCfKQ5p70aERAlhNtJOO6w6uXWwac2IT4+DRKDAyXIjIuJFcHHtnEhkWIKAMAGKcmwtdQo7S02FPfjQrZIvvA2cqaFHKbtWbQl4mwESN2e67u8NViMGtfCgBhuIvOv3R67eDEYlBBEOnAtmRm0a8hvOo1JfDJ5SSAVyBLvEIFAVCV9FyE3jjLto0SIsX74cjz/+OABg06ZN2LlzJ5588kmoVCocOHDAeN999/nfc889DUw7N6x33nnHUyqV8nl5eReOHTsmHT16dIub38qVK0u9vLw4m82GUaNGRR07dkz68ssvV3300Ude+/fvz/bx8bFd3dbBgwdlX331lVtaWloGpRRJSUkxkyZN0rq7u3NFRUWSDRs25I0aNapw+vTpoevXr9c89thjdb35/jhxLE5B6qRDCEvgcl84at44A+3WArgsCGt335ihYvz2qxHnT5hx28z2P14X08ww6ilSxndcEkUc7gKBpxS6HcXQ7SiCtdgV8nE+Dom+9QWxQ8U4sc+IvT8aMPdPSsgUN0e/gcsRUpZA4N650jWUUvA6K+DAsiUCN/u5bTUmiAIVGDpWikM/G1BXbV9nGec+DJGaQdj420acNW1BifcpeITdDUZwJwBBy8x7dyMvlFIcK9uFIFU0vBWBjnpZ7SKWMhg00llX7veI2WSERquByCqGTsuDs3V8zNUEhguRc8GK+mq+S2uQqyuaHXa7HiHlTA0AaI9SdoWqQHiPfAWs+OaZbHTiGFpKv9SbwXRHkNYNvBqk5boC/P/2/jw8jrNM9L+/T1X13q1dlrV4l/clduwkzmonZIfshEAIwzbwG2YYBg4wh3l53wPDNXOGOTDnDAxwfmSGnUAGQpJJIJA4+2YnsWM73vfd2nep16p63j+6pXjRrpa6Jd+f68oVqbq665bc6qq7nue574QTZ0X5lcwomE95cGy1FHr9/d//fRaiO9vXvjZwbYNVq1bR2NjI6dOnaWpqori4mOnTp/OFL3yBl19+mUQi4W9sbFQnT560Zs6c2e+n1auvvhr+3Oc+1whw2WWXxRYsWNC3dvRnP/tZyU9/+tMy27ZVU1OTZ/v27f7LLrus/5YAwIsvvhi+9dZb2wsKClyA9773vW0vl9lZbQAAWetJREFUvPBC5N57722vrq5OXHHFFbFM3NGjR4/m15tCDGlKJaRKqduA22prszMqIt7lqy0keNV0ujecInjF9LN6O54pEDSoXepl7/YEV9wYxOs7/0PYsTVbXo1RNdsa9oihWeCl4P1ziW5qIL6lCbsuSsE9cwccrc0nHq/ipg+E+f0vu3jsx13c9YnJk5Ta9TGsaYFhJ/9KKSLvnTXq6Vf9MTPJsNMSh5lhvD7FNbcGsTzvvrc8po/3r3qAVSfX83r0xzx56Mdsrn+Ou8o+RdFzHsI3z8A7Mzyq4x/t2ENj9AQfWPTXWfl5hqu9xSHW41I5c3KNqovRswscnr1sA/fM/0suH+AzdjAza9PvlWMHUyNKSJvrbEIRNarPJTveAoAV6L+42HAopfCEKkb9fDF5GZliRE5bAk/NyD+jnbbEsG+YjifbTdGZaKUkUMG6GXee16c6GwZLHsfLvffeyyOPPEJ9fT333XcfDz30EE1NTWzZsoV9+/bFb7rpJisWi434g2Pv3r3e733vexVbtmzZU15e7txzzz2z4/H4qC+MvF5v30WHaZp6NDGJ3JpS/2CyhnR8Fbx/Dspn0v7QgUETjuWX+kglYP87/a8l3bMtQU9n/2tHB6NMRejK6UTumI2nOoTyTZ6374y5Hm57IEJnm82jP+6kp2tyrA+066NYFcP7d3ITDk57+t88m+vAzMwIqdP87jrc3mS0qc7m2Ue7cV2NP2iwdMFM/vyir/HR5X9HzI7y4KGvc7JpP90nmkZ9/DdOb8BnBrho2lVj+0FGaP87SXa/ncxqci/yW9JJv8d95ugusCNFJsVlBscPjGwdaVOdQ9koCxrZsUxCOso1pK6doGXnz0l2Hh/V88XkZpZkRkhbR1cjwmnLjxHSvS1bePzAv9MebxqXZDRX7rvvPh5++GEeeeQR7r33Xjo6Opg2bRoej4dNmzYZp0+fHnT+8VVXXdX90EMPlQC89dZb/v379wcB2trazEAg4JaUlDgnTpywXnzxxb4L91Ao5HR0dJx3gXfttdd2P/XUU0VdXV1GZ2en8dRTTxVfe+21Xdn+mUVuTJ4repFzZsRL4T1zSO7rIPZG44D7TZ9hUVphsuPN8wsVuI5my8txplWbzJg3ugsg76wIofVVKJVuqN35X0dxOvOz4JHbnSL6ej3a1dTM9XD77A6qjtbx0r+eoOtUfhc60o7GboxhVQ5v/Wj8nRbaf3EApyu7FYWNsIXyGdgt5/++ujtdwoUG9hn//EoplpdfzpfXfp9r5t3JSesYr2x5lFdP/h7HHVmhvpjdw/bGV1lVcQ0+a+SFZs6kXY3dPPx/82WX+Ljm1qAUebmAJE53csW2K3FOe3nt6eioClvNnO/h1NEUdmp4NzIcW9PW5FA+ium6kK6wC4y6Oq4dayLesgc3NWAXCDGFmQVeUIyq0q5OOugeOy8S0vnFF3Fl9Xsp8pfnOpSsWrp0KV1dXVRXV1NZWcmHP/xhNm/ezPLly3niiSesOXPmDHpS+9KXvtTY09Njzp07d+lXv/rV6iVLlvQAXH755bFly5ZF582bt+wDH/jA3NWrV3f3PuejH/1o880339xX1KjXVVddFb3//vtbLr744sWrV69e/JGPfKTpyiuvHHCKr5hcptSUXTH+gtdU0vNqPR3/eQj/ilKM4PlvIaUUyy/18eKTURpOOmf1f9y/I0lnm8vVt4SzcqHttCdw2hIoMz8v2lOneohtbcE7rxCrIkBRbZCF3Unq97Rw+FutVC0PElxQgHd2BKsqlFc/h9McB1sPq6CRtl3i21rwzAxjRrI7xVQphVnqP2uEtNechV7mLOz/Bq3P9HPrvD+jbsUuTmzfyW/2Pcibpzdw98LPMLtweMVTtjW8QspNcGnVjWP6GQC6N5yk83dHmPaNNYP29O3VO31ysDWwWmvcrlT6ok5MeomeOOXt5RgJL61NDu4o2l3NrPWwfWOC08fsvim8g2lpdHBdxjxCOto1pN5IDVXX/E9AZgJciJRlYBR6R5WQ5lPLF58VYGHpqlyHMS527NjR93VZWRkbN24EYOfOnclly5bt6X0sGo1uBVi4cGHywIEDuwDC4bD+/e9/328LmN/97ndH+9v+1a9+tfGrX/1q36jHqVOn+gL4+te/3vD1r3+94cz9zzwewDe+8Y2zHheTg4yQihFRhqLogfm4XSk6Hz864H4LL/Lh8cKON99NIrSbHh0trTCZszA7SYt3TgFFf7YAI+RBa03PS6exW3M78ui0JUgcTPfC9S4opOjPFvRNew1dOZ3qv1rE9L9YyOFAMfsOuHRvbqHzsSO0/fseup46TupE92AvP2FG0vIlsbcdHbOzVln3XGaZH6d5dL3qimdVMTewhI/O+DI9qU6+t+Vv+c2e79KdHLpf8Runn6EyPJsZkbGtS9daE321HhxNz4ZTw35erMfl5T9EqT9xfr0IN2bT/ccTdPzqYLqQlJj0Usn0Z5eRmbKrRnGGrp7twTAZdvuX5rreCrujb/miTB+GZ+RrXnsppVCj+WHFlDDaXqR9LV9KcndDLprq4smDP6E5ejpnMQgxFcgZQIyYd3aE0Loqep4/NWDy5PUpFq30cWBnklg0Pe3s0J4UrU0Oa67xo7JYhbX3tdz2JIl9HXT86iCxt5rQ7sTecdeOJra5ifZfHST6Uh3adtOje/2MGFYvC7H2M9XsDFewwa3CWFeDd34hqbpo3/RjpztFdFMDTo6SjVQmIR1qNE9rTeztZsxpAaya0V+UDsYq9fc7ZXc4PJVBlIL5yaX87WU/YP3Mu9hc/wL/vOkzvH7qj7i6/2m8p7oOcbLrIJdV3Tjm0fzUiR7s01GMAg/R1+uH/W/qDyhsGw7tPnv/5JFO2h86SPJwJ4E15ah+ZipMBUqpm5VS+5RSB5VSX+nn8VlKqeeUUu8opV5UStVktl+rlNp2xn9xpdSdE/4DjFAqlf7bt4x0QmqM4nPS41VUzbKGnZA21dt4vFBYPLrLASfWghUoHdXfiJPopO61vyfWvHtUxxZTw1gTUiOHPUi3NrxCc/Q03lGu+xZCpE2phFQpdZtS6sGOjqFHPsTYFNw9GyPsof2XBwZM/JZd4sOxYc/W9FrSzS/FKCo1qF02PnczzWIfRQ/Mxzu3gOjG9LTikazZGwu7MUbHfx4i+no93tkRCu+bN2Rl2sqZHu78aAE9McUTzync1dMp/sRCfIuK069ZHyX2ZhMk0wl9qj5K4kAHbmJk6yBHy66PYUQ8GOHBR7NTh7tw2xMELi4bt/WOZpkfHbVxoyPsgwEYRV6U38Kui+KzAryv9uP8t0u/Q1V4Do/u+798d/OXOd554LznvXF6A5bh4eKK9WOOP7axAUxFyV8sQSddoi8N7266MhRzF3toa3ZobXTQWtP9/Cm6njyGETApvG9eOiHN4g2efKGUMoHvA7cAS4APKaWWnLPbt4Gfa61XAN8A/glAa/2C1nql1nolcB0QBZ6ZqNhHy06lL7DfTUhH9zozaz20NDjDWoPaVOdQNt0a9XvIjrVi+kdXYTfeth871oQ5htFVMfmNeYQ0R1N2OxIt7G99m4Ulqyjwjb7KtBBiiiWkUmV34hghDwXvn0vyYCfR1/ufrl823aJypsXONxMcO5Ciqc5h9TWBUd31H3ZcQYvIrTMJ3zITtztFx68PEt3UgB7FWqzh0Cm3b02tG7MJ3zqTyHtnDpnE9Zo+w+KOj0WIxzSP/aiLrna3bx2pr7aQ4k8t6iuLn9jVRvcfj9P24B46Hj1C7O3mUZ3Eh8uuiw5rum5sSxNGgRdv7fj93fX2InVGMUqqlMKqDJKqe7doyvTQTP5i1T/w4SVfpDPRyr9t/hKP7P0BPalOAFJOgrcbXmJF+RUEPaNrF9NLu5roG434l5fgW1iEb2kx3c+dRtvDK1gzY64Hr1dxaHcSpRRGyIN/dTmF983DKh9boaU8dylwUGt9WGudBB4G7jhnnyXA85mvX+jncYD3A3/UWud91ZzeEVLTTP+7jnYWa+/a0RNDjJJqV9Ncb1M2yum6kJ6ya41y/WiiZQ+GJ4SnYMaojy8mP7PYi445uLGR3XB02hKooIXhy01V27frX8RQFisrrsnJ8YWYSqZUQiomVvCKCry1BXT+9vCAa9iWXeKjo9Xlucd6iBQaLLxoYtZ6+OYXpkdLFxQSe7ORjocPYjdk93o0daKb9l8dJP52E74lxRQ9MB/fKJKy6TUWd34sQiKuefRHXXS0vTsCagSsvlHH0LVVFNwzF//FZeiYTfTVOtp/cYDEnras/UxnshuiQxY0Sp3uwa6P4l9VNq6jdGZZ+g74qKftTg/itifOuuBRSrFq+jr+du0PuHrG7bxZ9wz/vOkzvHH6GbY1vkrc7slKMaPE7jbcjiTBy9N9FsM31uB2JNOj38NgKk2t3Ubnnk66O12Cl00jdOX0YfeGncSqgRNnfH8ys+1M24G7M1/fBUSUUudmRx8Efj0uEWaZ3ZuQMvopuwBl002CYTXktN3OdpdUglFX2IV0UaPRJKRaa+Kt+/EVL5D1oxe43hHOkd5gdVoTfW1jJlpztI7D7btZVn7ZmG9aCiEkIRVj0FfgqCdF52NH+t2ndqkXf1AR7dZcfLUfcwKryBoBi8hNM4i8bxZu3KHjPw+TOt2TldeObW3u+5kL7ppD+D3VY7pLW1FtcefHIySTmaS09fxpucpQeKpDhK6cTtGH51P08YVYNSG6nztF6mR2CyG5PSncztSQCWlsSzPKb+FfUpzV45/LLOsdIR3diLBnToTgFdP7fcxvBbl9/if5/CX/h4rgDH6793v8du/3KAtUMq9o2ahj7hXd1IgKmPgvSl+0+5YWY1UF6X7m5LB7jBYlYwRicQ7vyc/2Rjn0JWCdUmorsA44BfT98SilKoHlwNP9PVkp9Wml1Gal1OamptH3qs0WJ5VKJ2eZPoajnbKrlGJmrYfjh1KDrqVvyhQ0Gm2FXe06OPFWzFH0ILWjjTiJNvwlC0d1bDF1jDohzWEP0s31z+O3AiwrvzwnxxdiqpGEVIyJZ0aY0Huq6XmpjuSRzvMetzyKi9b6KSg2WHJxbk4c3rkFFD0wn8Dl0/oSrNGuw+ydZumdHcG/upyiD9fimZGdu6PTqizu+ngEO5VOSttbBo/RjHiJvHcmZqGPrt8fx2nN3vRduz7d2muwKbtaa4yghX9VKcozvh8lRtiD8hr9tn4ZDqvMT2BNOUZg4AvvqvAc/vLif+JDi79Aoa+U9TPvHvOaWDfhEH+7icAl0/p+R0opwjfUkDrRTXJf/+vdtaOJbW1GJx2UZVDy4VrCV1Zw4lCKRGzkvSknqVPAmXM5azLb+mitT2ut79ZarwK+mtnWfsYuHwAe01r3O1SotX5Qa71Ga72mvDz3/QMdO4WpTHpzyNEmpJCethuPahrrBv4caa6zUQaUThtlD9JEO2h3VCOk8da9APhLhteCSUxdvaOcvW1chitXCenp7iOc6jrMRdOuwjeFixm1t7fzgx/8YMTPW7duXW1zc/OgHyqf//znqx5//PHIqIMTU44kpGLMCu6YjVHgpf2XB/u9G3/ptQH+7AuFWJ7cFV4xfCbBS6ahDIUbtWn/+X7i21uG/XytNV1/OkH30yeB9B3d8Zg2WV6ZSUptzaM/6qStefCk1PCZRO6YBaYi+mbjoPuORF+F3cqBR0iVUoTfU03wkmlZO+5gxzJL/WMqUuXG7CFHyJVSrK68lq9e8R+srb5p1MfqFd/ajE64BNee/TsKrp2GEfHQ/czJ857jtCbo+O0hoq/UkTiQTliVx2DuYi/ahaP7L5gWL28B85VSc5RSXtJTb584cwelVJl6d77n3wE/Puc1PsQkma4LkDDjxEIJPH6DSKEx6jWkADPmpdeRHj8w8Pulqc6hpMwc9Wdzbw/S0SSkiZZ9mP5SrOD4tIoSk4dZNPIRUm27uJ0pzOKJbfmitWZz3fOEPAUsKl09oceeaAMlpLY9+Frfl1566WBZWdmgFy//+q//evrOO+/sGluEYiqRhFSMmRG0KPzAXFJHu4i+XNfvPuNVfXU0lKXwzS/EmpGu7DjYlLbeKZVKKayKAFZFYNjTLEerbLrF3Z+I4Lrw2I87aW0aYqS0wEvhPXMJX3/u8rrRs+ujYCrMsv5HSN2o3Ze0ThSzzD+qoka9Ym820vn40QltBxTd2IBZ6sM7/+y1xcprElpfRXx7S9/vUWtNbFsz7b8+iNuZInzLTPxL363cGC4wqKixqD9pj/t7MB9orW3gs6Sn2+4BfqO13qWU+oZS6vbMbuuBfUqp/UAF8I+9z1dKzSY9wvrSRMY9Fidn17F93U5mL/Cy/rbQmD43g2GD8iqT44cGSUjHWNDIibcCYI4wIdWuQ7ztAP7SxaM+tpg6lMfAiHhGlJA67Zn11hM8Qtqd6qAn1cnq6euxjOz0U89XX/nKVzh06BArV67kkksu4eqrr+b2229nyZJ0sfPPfvazvqVLly6ura1d+u1vf7vvzlJ1dfXyuro6a9++fd65c+cu/eAHPzirtrZ26ZVXXjm/u7tbAdxzzz2zf/KTnxT37v+FL3yhasmSJYsXLFiwZOvWrX6A06dPW1dcccX82trapffdd9+sqqqq5XV1dVOzx5mQhFRkR+CyaXgXFdHxuyM4Xfm9zq03GbBK0lNtup85SfcLp9HJsxM/py1B56NHSBxMj1IFVpWlW2xMQHJdWpEeKe1LShuHSEpLfCjLwE04RN9qHHPCYtfHsKYF+ir+niv+TgudvzmM0zVxo3VW6dgSUt/yEgrumJ29gIbgdCRJ7GojsLai34JPoWurwFL0PHsKpytJ52NHib5ch2dGiKL7a/HNP79A1orLfFx9czCvbvCMJ631U1rrBVrreVrrf8xs+x9a6ycyXz+itZ6f2efPtdaJM557VGtdrbWeNHOcE04Mn5W9yskz53moP26TjJ//eRDrcenp1GOrsNs7QjrCNaTJzuNoJ4a/ZMGojy2mlpG2fslVy5eIt4h7F32WecXLJ/S4ufDNb36TefPmsW3bNr71rW/x9ttv853vfIf9+/cD8A//8A+JXbt27dm2bdvuH/7whxX19fXnfZgcP37c/7nPfa7x4MGDuwoLC52f//zn/RacKCsrs3fv3r3nE5/4RNM3v/nNCoCvfOUrVevWres6ePDgrnvvvbetrq5uYofDxYSaUncalFK3AbfV1tbmOpQLjlKKog/X0vj1LXQ+coTij0+OQhXa1Rghi/jWFlJHugi9pwpPTZj41maibzSmE7JxahkzlNIKi7s/UcBjP+nk0R93cvcnCigZYq1X8kAHsTca8c6MYFWM/sJ2qJYvgYvLsKYFMCMTd4fYLPXhdtu4MXvQtaAD6b0BMVFibzSC5rzpur3MQi/BtdPoeaUOpz0BhiJ0XTW+pcUDJpy+QPoeoutolGJK9h+9kJXvK6YkVsohf5LmepvLrhu8qNhQZs73sOWVOCePpJi7+Oxruab69E2u8umjvwwY7ZRdwxsmMusGfFLQSGSYxT6c1uHfcOxLSCewym57vImItyRnI6MvvvgiL710/oSPcDjMF7/4RV588UUA1q9fz7/8y7/Q3X1+scN169b1Pf7pT3+aSGT4yzgvvfRS5syZ0/f9L3/5S88nP/nJJQD19fWeXbt2+adPn37Wupjq6urEFVdcEQNYtWpV9OjRo/3+g91///1tmWNEn3jiiWKAN998M/z4448fBHj/+9/fWVBQMDFN2EVOTKmEVGv9JPDkmjVrPpXrWC5EnqoQ4Rtq6P7TCYJXTx9VC5ThSJ3sJrqpkdjbzYSuqCDyvlmjfi1lKEJXV+KtLaTn2ZN0/ddRjLAHtzuFd14hoXWVw+4pOh5Kppnc9YkCHs8kpXd9PEJpxcB/tr6lxXiqQ2O6a6wdjd0Yw79y4ItM5TXxzi0Y9TFG48xKu0bN6D66Uie6caM2voVFWYysf9FNDXhmhfFUhfp9XDsa5TXB1rg9NiV/sQSzcOgbwD1dLq9viLJ0tY+qWVN7ytiFxnZTWIYHwyArFckrZ1h4vHD84PkJaXNdeh3YmKbsxlpQhgfDO7LaJJ5gOUXz+2sZKy5UZrGP5KH+i7z1Z6JHSG03xR8P/5LpoZlcO+ueCTnmudavX8/69esHfbzXF7/4xUFfa6jH+xMKvXsue/HFF9m0aZO5efPmnZFIxL300ksXxmKx82Zder3evjv6pmnq/vYB8Pv9GsCyLG3bttxpvQDJlF2RVZHbZmEW++j45QF0FkcW7dY4XX88TsPXNtP4tS10P30CZSg6Hzvab2GYkfJUBin8UG16Sq7HJHzrTCLvnZnTZLRXSXk6KTUMePTHXTTXD1xQQCnVd4JO7G8neXzk7WCc5jg4ut+WL9rRdD5+hOTRia9FYGUS0tH2IgWIv9NKdGNDtkIaUOp0D6lj3X29R/ujTIVVGcSaEcJujGGEhpdkB0OKsukW/oB8fE81B2btp/GyDuYs8rJm3din7pqWonqOp99+pE11DuFCg0Bw9O8jO96C6S8Z0RRy104QbzuAdgcvjCIuLGaxF7fbPm/pzECctgTKZ45qtsxomMri6prbWFa+dkKOlw8ikQhdXf2f6zs6OigoKNCRSMTdunWrf/v27f3feR2DSy65pPsXv/hFCcCjjz5a0NnZOfq7ZyLvyRWNyCrDb1L4wXmkTvTQ88KpoZ8wCDdq0/NyHU3/axsNf/sGnY8cQXkNCu+vZfq/XM60b6zBv7qMjv88RM8r/RdTGgllGQSvmE7RR+aP2+juaBWXmdz9yQJMCx77SRdNdYNfzGlXE9vcTPdTx0dcmba3yE5/U3aTB9pJHe+GHBTVMUszI6RjqLRrVQZxO5O43eO79jW6sQEMCFx69nRdN+HQ/exJ7KZ0W53wddUU3jMXtz1J7K3h9cFUhmLVFf4hp2+LySfba0gh3f6lo9U9r7dxc51D+fSxvYfsWMuIp+sm2g/StOU7JNoPjenYIr9ordl46k8c69g3qucbxSNr/eK0JiZ0uq5SipqCWsqD2SsemO9KS0u58sorWbZsGV/+8pfPeuzmm2/GcRzmzp279Mtf/nL1RRddlJ0m72f45je/efr5558vmD9//tLf/OY3xWVlZamioiKZtjtFTakpuyI/+FeX4VtWTOfjRwmsKe8r6T4cOuUS35EexYq/0wK2xqwIELl9FsHLKs5bF1ny6cW0/NtO2n+2H+U3J6QFSa4UlZrc88kCHv1xF4//pIs7PhZhWlX/f8LKUERum0XHbw7R9cQxCj8wd9ijvXZvy5dzRki11sS2NGMW+/HMnvj2YUaBBzzGmAobWVXpnylVHx23mw7a1cQ2NeJbWnL+FFxXkzrena7YXJ5+L/uWFWNVBel+5iSBtdOGPdoU63FpqnOYWZv7UXwxdq52WbpzCVWnytnRFqen02Xt9WNbQwowa36m/cvBFMsvTSegdkrT1uwwb+nY3jt2rAUrOLL+rb6ieZRd9Gl8hXPHdGyRX5RSFHiL2Xjqj1SF5+AxR1Z/xix+t/XLcOofpHuQTkyNm631L2FrmzXTr7tgCsr1+tWvftXvdp/Pxw9/+MPEsmXL9pz72KlTp3YAVFZWcuDAgV2927/xjW/0TU/63e9+d/Tc/QGuueaa6JtvvrkPoKSkxHn55Zf3ezwenn322dC2bdtCgUBg6peYv0DJCKnIOqUURffXolMuHb89POT+2tUk9rXT9tN91P23jbR+fxfJgx2E1ldR/v9dRcU/XkLB7bP7PUkpy6DkL5firS2k7d/3ppPYKaywxOTuT0TweBWP/6SLxlMDj5SaEQ8Ft83Cjdt0/v4YOjV4sVGtNanTPSR2tWFEPOclsKnj3Tgtcfyry3JyUlZKYZX6xtSL1Crzg6Gw68avZU3yQAdOa6LfYkZGwKLoIwvwL393VEkpRfj6GlLHu0nuG/4aquMHU2zfFKerQ24YTwUpJ0EoFsKb8BCPahL9VMYdjcISg4Jig2Nn9CNtaXDQOt1iaiyceOvICxpZfgLlK1Cm3EiZSrTWzC9ZyY1z7h9xMgrvFicabqVdty05IetHe1KdvNP0Oj2pzgsuGc21gwcPelesWLFk4cKFSz7/+c/P/OEPf3g01zGJ8SMjpGJcWBVBIjfPoOv3x0lcXYlvUdF5+/QVJ3qjEac1gfIZ+FeVEby8At/i4gFbjpzL8JmUfm4Zzd/eTssPdlP2heUTUrQmVwpLTO7+ZIRHf9zFYz/t4vq7Qsxb0v8FgDUtQOSWGXQ9eZyuP50g8t6ZZ1VmddoSJPa0Ed/dTmJ3G25HerpUoJ+1j/EtzaiQJ6e/23Qv0uG3BjiXsgysikDfKPB4iG5sQPlM/Kv62rKRPNZFYlcboWur+l3zFLx8Gp2PHqF7w8l+/1b6M3uhh0O7kxzek+KitTJ9d7JLODEMbWB6LLQGI0u3i5VSzKz1sO+dBI6jMU1FU2YdennV6N83WrvYsdYRtXxxEp30nN5IsPJSLH+/3R/EJBSze3jywI+4ovpWagrSXQ62N75GWaCS6sjwRsJ7Z1INJyHVjsbpSExIQrqt4RW01qyevn7cjyXOtnz58sSePXt25zoOMTEkIRXjJnzrTKKbGml/6ADTvrYaZRnYrXFibzQS3dSIfbIHDPAtLaHg/XPxryzF8I3uAskIWpR+YQXN/7yNlu/upOxLK/DOmdgqsBOpoDidlD71q26e+nU3C1d4uea9Qfz9FCjxzikgeM10oi/X0fP8KcwyP4lMAto7UmiEPfgWF+FbUoxvcVHfdNJedmOM1MlugldOH/aNgvFglvpJHW0e02tY04PEt7egbRdlZXeSiE65xDY34V9d1vde1loTfa0BnXTSVXX7obwmoWur6Pr9MVL10fOmS/fH5zeomevhxKEUCy/ySpGjSS7hxDBcA9PypNv6ZLGlz8xaDzvfSlB/wqZ6tofmOgevT1FQNPr3jJPoBG1jBUqG/Zx46146Dj2Jv3QJSEI6ZRxp301XsoOgJ33Otd0Uh9t3sq3hZW6ccz+V4aEr4Rt+ExW0hpWQup1JcMe/wm5HooX9rVtZXLqGiFfer0KMJ0lIxbgxfCaFH5pH67/tou0n+3DaEiT3d4AGz9wIhffXErikHLMgO+tAzIiHsi+uoOmb22j+Pzso/9uVeGqyXvgtbxQUmdz7/xSw+aU4m1+KcfJIimvvCDFn4bu/T51ySR7uJHWki9TpKPHtrQAor4F3QWG6Pc/iYjw1oUEvgGNvN6M8Jr5lw7/4HA9WmR+3O4WbcEZ988KqDMLWZuymOJ7Ksa/RO1N8Wws65pxVXTe5vwOnOUb4xhmDJvOha6vo+uNxep49RdED84d1vLmLvRw/mKK53qFmjiSkk1nCjmG6JpblwXWzN0IKUDPXQhnpad7Vsz001TmUVZpjmoLoxNPLI8xA2RB7vivRug/DE8ITqRn1cUX+OdS2g9JABSWB9DIFy/Bw89wHeOrQz9lw5NfcNPfDVIRmDPk6ZrFvWAnpRLV82VL3AqayuGjaVeN6nMForWWqsJi0XNdVwODrxTIkIRXjKrCyDP/KUmKbGgctTpQtZrGPsi+toOmfttH8v9+h/CsrsaaNz7HygWkqLrsuwJxFHp79XQ+//0UXK+c7LCuN4xxoJ7m/A510QYFnTqTvBD7cnpcATkeS5IGOMY1gZ0tfpd2WOMYA/T2H0jv6aNdFs56QRjc1YBR5+6bdakcT3diAWRbAu3DwIkpmoZfgZRVEX6un4M7ZwypCFS4wuP7ukIyOTgEJJ54eIfWkE1Iri2dnn99geo3F8YMpLrtO09Jgs+TisV3M27F0QjrcNaRaa+Kt+/AVL5QL7CmkI9FCY/QUl1Zdf9b2gBXilkxS+vSRX3HL3AeGrFBrFnuHVWXXac0kpONYZbcpepojHXtYVXE1AU943I4zGL/fT0tLC6WlpfI3IyYd13VVU1NTIbBzOPtLQirGXfGnFuM0x7GqgxPyoWqVByj74gqa/9c2mv/lHcr/+8oJLQ8/0ezmOKFDbdzkaaOntQ3zZZsooEsDhK+anp6Gu7AII2ihUy5Oe2LYyShAYlcbKPCvHP5IyHgxSzPrjJrjeEaZkBphD0bEi9M++rWo/XG6UsR3tBK+vrpvtDm+sxW3M0nk9tnDeu+Hb6wm+lo9PS/XEbl15rCOK8no1NA7ZdfyeNOjItkcIiU9bfeNF2I0nLRJJaGscmyn/5EmpHZPA06iHX/pwjEdV+SXQ207UEoxr2jZeY8FPRFumfuRdFJ6+FfcPPcByoKVA76WWewjdWLo7iG9n93GOI6Qbq57Hr8VZFn55eN2jKHU1NRw8uRJmpqG1xJsotXX11uO4+T+wkDkKxfYadv2nw9n5ymVkCqlbgNuq62tzXUo4gyG38SY4KmznupQek3pt7bT/L/foey/X4QZmZgS8RMlvr2Fjv88hN2Q7mlpFHoJryohXlnIa3ss6jstlkV8XLkkiOFLJ0PKY/StD03sbcOqDmNGBh+JC6ydhmduZMj9JoJVlh4hHUulXYCiB+ajPNm94I+91QiO7puuq1MusTcbsapDeGYN7w67pyaMb0kR3c+dInxjTdbXuIr8lbCjGK6Jx+PFTWR3yi7AzPke3ng+xtuvpv92yivHNtvBySSk5jCLGsVb9wLgL1k0puOK/KG15lDbTqrCswl6+m8FFvIWcPO8B3jq0M94+shD3DL3I5QEzi+aB+mE1O1MDrm+32lLgMfACI3PJeyprsOc7j7C2qob8Zq5u5nt8XiYM2dOzo4/lCVLluzQWq/JdRxiaphSVzta6ye11p8uLByf/oJicvHOjlD6uWXYzXFa/vcO3OjALVImE51yaX/4IC3f3QmWQeEH5zHtG2uY/i9rKfnUYqreV8Vdny9n1ZV+dm5O8OvvdXDySOqs13CjNj0v1BF/e/ACQemRGjWsIjsTwSjwgqXGVGkXyHoyCunqulZNCM+MdPIZe7sZHbPThaBGMDMgfGMNbnuS2Ob8vCsuxkfCiWG6BpbXm/U1pADTqkz8AcXhvSkME0rKx5aQ2vEWUBamb3jF4xKt+7ECZSNuEyPyV1P0FJ3JNuYVLR90v4i3iFvmfgRTWfzp8C9pj/f/2WaW+ECD0z74tF0n0/JlvGZc1XUfJeItZFHp6nF5fSHE+aZUQirEuXwLiyj9yyWkTvXQ8t0duInJ3bPRbojS9E9b6dlwitB1VUz7/11M+IYaPNWhs07Olkdx1c1B7vlkBGXAYz/u4qU/9JBKpnsbGkGLgnvnErxq+oDH0rZLx68Okjgw/N6Y400ZCrPUj9MythFSN2bT9YdjJA93ZiUuuyFK6nBX3+ioG7WJbWnCO69gxMm8b2kJVmWQ7mdOorX0AL9QxO0YLYUteItDFJebFBRn9/RsGIqaeR7Q6WTUtMZ2MW/HWrACJSg1dJzatYm37ccno6NTysG2d7AMi1mFQ/+7FvhKuGXeRwh4wri6/xonvTUOhips5LQmMIvHb8bTmsrruGPBpzGNKTWJUIi8JgmpmPL8K0op+dRikgc7af3BLnRqWAW/8k709QYa//5t7OY4JZ9dStGHh552WjXLw4f+qpCL1vp4Z1OCX3+/g9PH0qOlVpkfZap08vR283nJjxtzMEKefvtm5pJV6h/zlF3lM3Hak+hkdm5QRDc1goLgpekqk719Tvvr5zpkbIYifEM1qWPd6arU4oKQdOO8vvI1wsunsfJyP/OXZ3+q4Kza9LT7sjFO14V0Qmr6h1d1O9l5HO3E8ZfI+tGpwnEdjnTsZmbBwmFPay30lXLn/E9TEqhIF7myz+4HPdxepE7b+PQgdVyHjkR6KrrP9Gf99YUQA5OEVFwQApeUU/TRBSR2ttH673vQzuQZeXJjNq3/sZe2H+3FMzPMtK+vJrBq+HUEPF7FNe8NcdcnImgXfvejLl79UxQ7lf4dJHa1EX21jvjWlrOeZ0Y8FNw5O+9a55hlYx8hVYai6MPz8S0ae285rTXRTQ34FhX1Fc/yzi2g+JOLsEpGd1ETvLwCI2zR/czJMccnJoeEHcNj+DDU+FWynjnfg2HC9Jqx32RyYi3Dnn7rJDowPBF8JQvGfFyRHwxlcP3s+7ho2pUjel7vTJ7tja/xXwf+nViqu++x3s/PwRJS7ep0Yb5xSEgPtG3j0X3/l9ZYQ9ZfWwgxuPwa+hBiHIWurkTHHToePkT7z/ZR9LGFWW0+Px6Sx7po/eEenMYYkdtnEXnfrEF7WQ6mZo6HD322kNeejrL1tThH9yW5/p4wFWvKsJtiRF+txyjw4KstxG6KofxmXhaCssp8uJ1j60Xaq3dUeCxrkZKHOnEa40Tem27+brfEsUr9Y4pNec10X9LfH8duiGJV5McaXjF+UrEE179xA4nKdjYd8FA502LBiuxedIcLDB74XCGRwrHfi7ZjLfiKh1dAMFixisC0ldK6YgpRSg2rt+hAaiLzSDhR/Na7NzxVwET5jEETUrc7BbYel4R0VuEibDdFsX9a1l9bCDE4GSEVF5TwDTVEbp9F9LUGOh4+lLdr9LTWdG84SdM/bkUnHcq+fBEFd8wedTLay+tTXHt7iDs+GiGVhEce7GTTszEC11VjVQTofvokqfooPc+fpuvxo3n5++nrRdo6tsJGdkOMtv/Yi30qOvTOg4htbER5DQKry7Cb43Q8dJD4jtYxvSZA6NpqMBXdz54a82uJ/Jd0YnQX9WAELUqmmQQj43N6LiwxMcb4OaK1xom3Yg2jwm42bvqI/JJw4rx+6o9901tHoyxYyWVVN6KUoivZTjTVhVIKs9g36Gd7b7I6HglpwAqxrHytvFeFyAEZIRUXnMjts3BjNj0bTmEETQruzK+y6k5XkrYf7yPxTiv+i0op+sRCzHB2W67MrPVw/18X8Oofo2x+Oc6RfSmuv7Uaz8vH6HrsKDrlEFpfNaITs9aaWI+ms92lq92lq91Jf92W/r6n2+X6u0LMWTS2Ude+hLQ5jqdy9COHRoEHHbOx66KjnpasbZfoW434V5ZiBCyUxyB4zXS8tcOrPDoYs9BL8LJpRF+tp+DO2Rih3LfdEeMnakapW9WMZ0aY5aMfeJoQbrIL7aYwhzFlN968i/b9v6Ns5V/gCY18TbXIPy2xOg60bmN+8Yoxv5arXTYceRiN5tZ5f5ZOSAcZIe1LSLPYWzxuR3np+OOsqbyO0sDAhf6EEONHElJxwVFKUXjfPHTcoevJ46iAReSm/LgCTOxto/XBvbg9KQrvryV03ciSwpHw+Q3ec1eYuUuSvPBfPfz2Z1EuW1PBnJOnUX4L35Kz11dqV9PTrelqd+hqd/sSz3TS6dDV4WKnzj2GIlJkUFBiYNuaF57ooWqWhS8w+tGf3l6kY11HagQsjCIfqfoogVG+RnxHK7rH7quuqyyDwMrs9QkP31BD9LUGel6qI3LrzKy9rsg/CTuGd5IUUrEzPUiHs4ZUWT484UqsYRZAEvmvKjyHDy35b3iMsS/pMJTB5dW38MyRX/Gnw7/k6qIrcffFBty/d/Q0myOk7zS+xqnuw1zCe7L2mkKIkZGEVFyQlFIU/dkCdNyh8zeHMfwmoXVVOYtHO5quJ47S9YfjWBUBSj+/HO/M8IQce85CL5WftXj5qSgb30xyvLScxRd5OfRqIpN0vpuAuucUpfUHFQVFBiXTTGYt8FJQbBApNNL/LzLw+d9NPBtP2fzmh5289kyM6+4YfaEko9ALphpzpV0AT2WQ5JGudL/VUST+0Y0NGBEP3sVFdD11HO/8Qnzzs9cH2TMjjG9xEd3PnSJ8Y82gzeLF5Ga1Ki5+cw2Jkk42vK5YdJGP2mX5t4YbMj1IGV5C6i+ej794/niHJCZI72flcCvrDkdleBbXz76PDUcf5gj7qWovRbu63xoPTlsCTIURyc6MkZ5kJ7tb3qK2aDklARnBFyJXJCEVFyxlKIr/fBFu3KH9FwdQfovgZRNfzMBuidP24B6SBzsJXllB4f3zMfzjV2mzP/6gwY3vDzNvSZIXnujh2ecdIEYwnB7hLKu0mLs4nWT2Jp2RIhOvb/hJ3LRqi5VX+Nn6WpyFK7xUzxndBUW2epECWJVBEnvacNuTI77j7kZt4ttbCK2rwj7RQ/JgB54Z2b+JEL6xhpbv7CS2uYngWrlgmqrsVArLtdAotAbyeBmbkxkhHarti2vH0U4S0zf2KewiP+xo2siRjt28d95HsYzsLSOojszl+lkfYO/2l6lyS4m3dRMojZy3n9OexCzyZq0g4dsNL4GGVdPXZeX1hBCjIwmpuKApy6DkM0to+dcdtP1oL8pvErhoeK0MsiG2pYm2n+4HV1P8qUU5TzjmLfEyY56Hnk6XSJGB5cnuVfFl1wU4tDvJ8//Vw4f+qnDUr2+V+rIyQmpl1qCm6qIjTkhjm5vA1gQum0b0tQaMIh++pWNvI3Mu37ISrMog3c+cJHDZNCm4MUXZqSSGMtEoQGPk8WD4cKfsxpreoXXXz5m+9qt4wpUTEZoYZ4fad2AZ3qwmo71qCmpx5/bAC+1s3PUkV1/5fjzm2bME3NbstXxpjzdxoG07S8suJeItysprCiFGJ49PeUJMDMNnUvq5ZXhmhGn9wS4Se9vG/Zg6mR6Vbf3BbqxpAaZ9bXXOk9FeXp+iuNzMejIK6Z6o194eor3F5a2XBl4nNJR0L9KxVdmFdGEM5TWx60deaTe6sQFregA3buO0xQleXjEubYSUoQhfX03qWDfJ/R1Zf32RH5xUCkMZ6MxZWeXx2dmOtYAyMH1Fg+4Xb92H4QljhaRQzFTQGmugNdZIbdHycTvG9Jp5AMSaO9hw9GFs9+zCBE5b9hLSLfUvYBkeVoywl6oQIvtkhFQI0gVuSr+wnOZ/3kbzt9/BLPZhTQtgTgtgTfNjlQewKgKY5YExT6dNne6h9Yd7sE/2EL6xhoJ75lxQawNn1npYtMrL26/Emb/MS9n0kX8MmaV+3I4kOuWiPKP/3SmlsKYHsetGlpDazXGS+zuI3D6L+BuNmNMCWamsO5DA5RV0PnaE7mdO4ltYNG7HEbmhtca1bQxl4ioDcPJ6hNSJt2D6i1HGwJ+FWmsSLXvxlyyQUf0p4mDbOxjKYE7RknE7Rm+yudRaw/M9T7O/dRtLyi4BMu2G2hL4szCLqbHnJEc79nHx9HUErNHXNBBCZEfeJ6RKqcXA3wBlwHNa6/+b45DEFGWGPZR96SJ6XjyN3RDDboqRers53Yj7DEaBJ52slmeS1WkBrPJ08mqErQEvvrTWRF+pp+PXB1E+k9LPL8O/fOKmB+eTq24Ocmx/iuce7+HeTxdgjHBksbfSrt0SxzN99K1fID1tN/ZGI27CwfAN72ZD7I0GIF1gKXWsi4Ibasb1otvwmYTWV9H1h+PYDTGsitHWBRb5KOnEUa7CUCZk/hZG+jcxkexYy5A9SO2eBpxkB76SRRMUlRhPrnY51L6Tmsg8/NbYPnMHY4QtsBTFiRJuq/04ZYF0scEXjz9GpZpFOOmiiiyePfKfQ77WjIIFLCxdhePavHDsd9SWrGB24WKiqS5eOv44ASvIsrK14/azCCGGb1wTUqXUj4H3AY1a62VnbL8Z+A5gAv+htf7mQK+htd4D/IVSygB+DkhCKsaNWeil4I7ZZ21zozZ2Uwy7MYbTGO/7Orm3ndjGs6eNqoB5RoL6brJqFHnpevwosbea8C0uovjPF2EWZb+x92QRCBpcc2uQp3/bwztvJFh5+cjaXZil6d+d0zz2hNQ7N5IuIOPqYe2vtSa6sRHPvAKS+zvwzIyMSzGjc4WuraLrjyfofvYkRR+WqqVTScKJYWgzM0Lam5DmOKhB2LGWIdePxlv3AuAvWTgRIYlxVt99jGiqm9qqsfceHYxSKtOLNEl5cF7f9u5kB3Y0vczDKPbSnRp6+ULCSe+v0XSnOkg66fO11hqP6ePSqhvOW6MqhMiN8R4h/SnwPdKJJABKKRP4PnADcBJ4Syn1BOnk9J/Oef4ntNaNSqnbgc8AvxjneIU4jxG08M6K4J11fsU/nXSwm+PnJ6vHu3G2NoNzRpJjQMHdcwjfMmNc1hpONvOXe9m7PcmmZ6PMXeShoHj4U6HNLPUiBdLTscuHP+KYOtaNXRcleOV03O4kwSsnZu2vWeQjeNk0oq/WU3DnbIxQ9ouKiNxIOHFMx8BUBr2fGPm8htSJt+IrnD3oPvHWvViB8mG1hhH572D7O3hNHzMKxv9mmFns6+s32ut9tR8jvqOFFnbiKQlyZ+2nh/16luHhzgXv7h/yFnDngk9lLV4hxNiNa0KqtX5ZKTX7nM2XAge11ocBlFIPA3dorf+J9Ghqf6/zBPCEUuoPwK/GMWQhRkR5TTxVITxV569B0Y7Gac0kq01xPHP6T2ovVEoprr0tyEP/1sELT0S5/c/Cw572ahb5staLFMBNOLjtCayKoUdbo5sawFIor4F3QdGIktmxCt9YQ/T1BnperiNyy8wJO64YX70jpEqZ6MzQaL5O2dVaY8daMAdJNLVrk2g7QHD6JRMYmRgvtpviaPse5hQtHZfquucyi30kD3Wet91pTaYfL7lwZxcJMVXl4h5sNXDijO9PZrb1Sym1Xin1XaXUD4GnBtnv00qpzUqpzU1NTdmLVohRUqbCKg/gX1pCaH2VJKP9iBSZXH5DkOMHU+x/Jzns5ylDYZb4slJpFyD6Wj2djx1F68Gn7WpHE3ujEf+KUgrvryV0bVVWjj9cnhlhfIuL6H7uFNp2J/TYYvwk7BjdgS6Y58XyG8yY6yEQzs+E1E31oJ3EoGtIk53H0U4Cv6wfnRKOdewj5aaYV7xs6J2zwCz24bQnzvs8dtoSoMAskGm2Qkw1eTwpKE1r/aLW+nNa6/9Ha/39QfZ7UGu9Rmu9pry8fCJDFEKMwfJLfVTUmLz8VJRYz/CTLKvUn5UpuwD+FaVE3jsThlhGmtjdhtuZIrCmDGWoYRdByqbwDTW4bcl0H1QxJSScGI2ljZjXlxAutlh5hZ/CEUxhn0jOMHqQxlv2AAp/yYIJikqMp1mFC7l+9r1UhmZPyPHMEh/Y+ryCgk5bAqPQe0FVpRfiQpGLv+pTwIwzvq/JbBNCXIAMQ/GeO0Mk45pX/jj89itmqT9rU3atMj+eGeEh1/ZGNzaAxyC+sw3tDK8IUrb5lpdgTQ/QveHUkCO6YnJI2On3vd/M/+rJdqwZYNApu5GZ11J+8WcxPONXjVVMHMvwMKtw0YS17+lt/XLuOlKnPXs9SIUQ+SUXCelbwHyl1ByllBf4IPBENl5YKXWbUurBjg5pHi/EZFJaYbH6Gj/7tic5dmB4U3fNMh9ue7oXaTakTnaTODjwZ4cbs4lvbSawqpTQlRUoMzdTKpWhCN9QQ+poF8kD8lk3FSScOEsOLcX9RTONp2x+/1AXbc1OrsPqlx0feoTU8ASluu4UcahtJ9saXsHVE/d+NIvTU3KdtnMS0lZJSIWYqsY1IVVK/RrYCCxUSp1USn1Sa20DnwWeBvYAv9Fa78rG8bTWT2qtP11YWJiNlxNCTKBL1gUoLjN48YkoqeTQI3+9vUid1uyMksa2thB9vWHAx+Nbm9FJl9B11fgWFWflmKMVuLwCFbLofuZkTuMQ2ZFwYrQUtuBfXkIwbDBviRd/ID/XkNqxVoAB15Am2o/QefQZXDs7f5citxqiJzjasSfdI3eC9I2Qtp19c9Jpk4RUiKlqXBNSrfWHtNaVWmuP1rpGa/2jzPantNYLtNbztNb/OJ4xCCEmB9NSXHdniM52l03PxYbevzSdkNpZKmzkqQzitidwY3a/j3e/cBoVsrCqcz8N0fCZhNdXEd/Wgt049O9K5LeEE6OhrJ6CK6oJFxosXuUjEMrPdXLpNaQK01/U7+OJtv10Hv4TysjPNbBiZK6ovoX31X58Qo9pFHjBVGeNkLoxGx1zpMKuEFNUfp7xRkmm7AoxuVXN8rDsEh/bN8ZpONl/YtirNyF1srWOdHo60bTrz1/HarfFSR3uwizxYXjz40I7dF0VGIruZ2UJ/mSXsGP43SDKVji2JhnXaDc/1wfbsRZMfxHK6L9rXMGcm6i6+h9QE9AeRIwvx01P052IVi9nUobCLPSelZA67ZmWLzJCKsSUNKUSUpmyK8Tkd8WNAYJhxXOP9+AMUjjILPaBkcWEtCIASmHXnZ+Qdv0h3akqckNN3lR4NIt8BC6bRvTVOtye1NBPEHkr4cRZvX8NHQ8f4tQxm6cf6SYWzdOENN4yaMsXQIoZTQFaax4/8CBv17+Yk+ObJT7cM4oa9SanvetLhRBTS35cWQkhRIbPb7D+thAtDQ5bXxs42VSmwizxY2ep9YvyGJjlflLnJKTa0cTebMQo8BC4vCIrx8qW8A016IRLz8t1uQ5lXCilblZK7VNKHVRKfaWfx2cppZ5TSr2jlHpRKVVzxmMzlVLPKKX2KKV2K6VmT2jwI5BwYnh1eppi78iokadnZyfWMmCF3WjD2zRt+39xU8Ovli3yU1P0FO3xZiLe3KyXN4t8Z4+QtvYmpDJCKsRUlKenPCHEhWzuYi/zlnh484UY7S0DV3c0S31Z60UK6XWkdkPsrJYuPS+fRvfYBC+vGLItzETzzgzjW1aM2zn1RkiVUibwfeAWYAnwIaXUknN2+zbwc631CuAbwD+d8djPgW9prRcDlwKN4x/16CScGB58KFPhZopGG3n2Xutlx1oGrLAba9pJsuMYysr/9jVicAfb3sEyLGYVLsrJ8Y3i9JTd3tZWfSOkRZKQCjEVTamEVNaQCjF1rHtfCNNSPP94z4Dr6awyP6mTPcTebs7KmjurMgi22zcNWKdcep4/DQpCN9cM8ezcKP2b5RTeNy/XYYyHS4GDWuvDWusk8DBwxzn7LAGez3z9Qu/jmcTV0lpvANBad2ut83bYLmHH8GgPWIrMsj1Unp6d7Xj/CanWmkTrXvwlCyasX6UYH47rcKRjNzMLFuI1c5MAmsU+dNJFR9O1BJy2BEbEg/Lk6R+GEGJMptRftqwhFWLqCEUMrropwKmjNrvf7r+SbmhdFUbQovX7u2j4yht0/ekEbvfoRwutyhAAqboeAKJbm7CbYngXFGIV5Oed+Xwbtc2iauDEGd+fzGw703bg7szXdwERpVQpsABoV0o9qpTaqpT6VmbENS+lR0i9KNNA942Q5jam/ripKNqOY/azhtTuqcNJduIryc2ImsieU92HiNsx5hUvz1kMvdV0e0dGpeWLEFNbHp7yhBAibclqH9VzLF57OkZPl3ve4955BVR88zJK/mopZnmAzt8epv7Lm2j7+X5SJ3tGfDwz4sEIeXDakrgxm+iLdZDShNZXZePHEdn3JWCdUmorsA44BTiABVydefwSYC7wsXOfrJT6tFJqs1Jqc1NT04QFfa6EE8PSHjAVbh6vIbVjLQD9jpDGW/YB4C9dOKExiew72LYDvxWkOjw3ZzG824s0nZC6kpAKMaXl4SlPCCHSlFJcd0cI29a8/If+E0xlKAIXl1H+5YuY9verCaytIPp6A41f20zTt7aPeDpv4QPzCV9bRepUD05zHOUzCKwcvKqoGBengBlnfF+T2dZHa31aa3231noV8NXMtnbSo6nbMtN9beBx4OJzD6C1flBrvUZrvaa8vHx8foph6J2y27uGVJGfI9/vJqQl5z0Wb92LFZiG5T//MTF5JJw4Jzr3M7doKWYOe8m+m5Am+/4vCakQU9eUSkhlDakQU09Rqcml6wMc3JXi8J7koPt6asIUf3QB07+9loL3z8FpjKWn8/7dm+npvMNoj2L40hdhnhlh3B6bwJppqDzpPXqBeQuYr5Sao5TyAh8EnjhzB6VUmVJ9qy3/DvjxGc8tUkr1ZpnXAbsnIOZRSTgxTG1lquxCDvOAQdnxdEJ67pRd7dok2g/K6OgUcKxjL7ZrM69oWU7jMAu9oNIjpDrl4nanMEuk5YsQU9WUSkhlDakQU9Oqq/yUVpi8+GQPifj5U3fPZYY9RG6ZmZ7O+5dLMEt96em8X8pM5z018HReN2rT/vBBujecRMcdgldMy+aPIoYpM7L5WeBpYA/wG631LqXUN5RSt2d2Ww/sU0rtByqAf8w81yE9Xfc5pdQO0oOO/z7BP8KwaK1JOnEsbaEsA9fVeVsUyBlgym6y4yjaScj60SngcPtOCn0llAfPXa49sZRlYBSkK+2+24NURkiFmKqsXAcghBBDMU3Fe+4M8dsHO9m4Icb620LDep4yFYHV5QRWl5M60U33c6eIvt5A9KU6fIuLCF1fjX9F6VnTI5XPxGmMkdzfgVnsw7ugaJx+KjEUrfVTwFPnbPsfZ3z9CPDIAM/dAKwY1wCzIOkm0GhiC1w8s8KU+S083vxMSAeashtvOwAo/MXzcxCVyKZ1M++iO9meFzdFzBKfJKRCXCCm1AipEGLqqqixuOhyHzveTHD62Mgr6XpmhCn+2EKmf2stBffMwa6P0fpvmem8T5/AzbQXUKai8P5aUie6Caydlpdr+cTUkbBjAKRWGPhqC6mosViwIj8vvO14C6avEGV4ztruCU0nPGMdhieYo8hEtgSsUM5HR3uZxT6c1ncTUkMSUiGmLElIhRCTxtr3BCkoMnj+8R4ce3R9R82Ih8itM6n458so+cwSzGIfnb85TP0XN9L+iwOkTvcQ394KLgTXVmT5JxDibAknnZD67QDadkklNcnE2Hvqjgcn1oIZKDtve7BiFcUL35+DiEQ2vXLiSY53Hsh1GH3M4swIaauMkAox1U2pKbtKqduA22pra3MdihBiHHi8ivW3B3ni59289VKMte8Z/YiMMhWBNeUE1pSTPNZFz3On6Xm1jp4XT6N8Bp4ZYTw1w5saLMRoJZwYylUU/9Yidk0ze1MR2ltcrrsj/957dqz1vCq6Wru4qSimN5yjqEQ2xO0oDT3HKQtU5jqUPmaRFx1zsOuiqKDVV3BOCDH1TKkRUilqJMTUN2u+l4UXednySpyWBjsrr+mdFaH4EwvT1Xnvno1Z5CN8Y35MWxNTW++UXX15AM+sMDPmeViwPD+ridqxlvMKGtk9DZx++StE67fkKCqRDX4ryD0L/5KFpatyHUofsyQ9Ipo80iWjo0JMcVMqIRVCXBiuviWI16d47vEe3BH0GB2KGfESee8sKv7npQSvmJ611xViIAknhjY01ooCPNODlFda1Mz1DP3EHHBizZjnJKSGJ0TR/LvxFs7OTVBizLTWOK6NUgpD5c8oZG8SatdFMYvz8yaNECI7JCEVQkw6gZDBNbcGaTjpsOONRK7DEWLUeqfsetst3IRDT5dLT9fQrY0mmmvHcO3oeSOkpq+AyKzrztsuJo+6nqP8evf/oSl6OtehnOXMUVEZIRViapOEVAgxKS1Y4WXmfA+v/DHKi0/2EIvm30W8EENJ2DECiQD8ro3koU7e2RRn2+vxXId1nr6WL/5zepB2nsCOt+UiJJElB9vewcWl2F+e61DOIgmpEBcOSUiFEJOSUoqbPxBi+WU+dm5O8It/7WD7pjiuk58VSoXoT8KJYbgmhjJRlsLVYOThmdnJJKTnTtlt2flT2vf9NhchiSyw3RRHO/Yyp3AxlpFfU8WVx8AIp2PqXU8qhJia8vC0N3pKqduUUg92dHTkOhQhxATw+Q3WvTfEh/6ygPJKk5f/EOXXP+jk+KGR9ykVIhcSThzTNTCUgTINtAsqD8/MdqwV4KypudpNYUeb8ITypzKrGJnjnftJOUnmFS/PdSj96h0ZlRFSIaa2PDztjZ5U2RXiwlRaYXHnxyLc+qEwdkrzXz/t4g+/6qKj1cl1aEIMKuHE8OJDKQWmwnU0hqFyHdZ57Pj5U3ZTPY2AixWWhHSyOti2g5CngMrQ7FyH0i+zJF3MSBJSIaa2KZWQCiEuXEop5i3x8uG/LuTy6wOcOJTil9/t4PUNUZIJmcYr8lPCjuFX6X66ylS4bn5O2bX7puy+24fU7qkHwBOSitSTUczu4VTXIeYVL0vfEMlDMkIqxIXBynUAQgiRTZZHsWZdgEWrfLz+TJQtL8fZuzXBFTcGWbjCi8rD0Sdx4Uo4MfwqkP7GMtBaY+RP540+TqwFwxvBMN9tv5HqqQcUnmBF7gITo3akfTeudplXtCzXoQzIt6wEpy2BCuThH4UQImskIRVCTEnhAoMb3x9m+aUpXn4qyobf9bDjzTjX3BqiokY++kR+SDgxfKQTUmUqXMfNy5smdrzlvNYuqZ46rOA0lJlfxXDE8Bxq20FJYBolgfy9oRBYVUZgVVmuwxBCjLM8nBgkhBDZUznTwwc+XcD1d4XobHP5zQ87efbR7rzs9SguPAk7hq9vhDS/p+ye2/Il1V2HJ5S/yYwYWFeyncboKWqL8rOYkRDiwiLDBEKIKU8ZisUX+5i3xMtbL8XYtjHOwd1JLl0f4KK1fkwr/0akRsJ187MQjhhaeoR0GpB+n+o8TUidWAu+koV93/dW2A1OW5m7oMSoRbxF3L3wL/BbwVyHIoQQkpAKIS4cXr/iypuCLF3j45U/Rnnt6Rg730pw9a1BZi/w5G1hj4E01dls3BCjvMrk8uvlwnIySjgx7FJNsKYSFbBYtMogFMm/96EdbyUUOL/Crkcq7E5axf7yXIcghBDAFEtIlVK3AbfV1tbmOhQhRB4rKjW57YEIxw4keeWpKL//ZTczaz1cfWuQkvL8L57R0erwxnMx9u1I4vMpZs2XNXyTVcKOQalJYHF6ndzM2vx7/7l2AjfZddaUXStQRtnKv8QbqclhZGI0mqKn2NX8JpdMfw8hb0GuwxFCiKmVkGqtnwSeXLNmzadyHYsQIv/Nmu+l5rMedryR4I0XYvz6ex2sWOvj0vUBfIH8mzcZ7XZ568UYOzcnMAxYfZWfi6/248/DWMXwJJ04QTuE3RzHLPXR2ebiD6i8ev858Vbg7JYvhuUjULYkVyGJMehMtFHffQyPKa1UhBD5YUolpEIIMVKmqVh5hZ8FK7xsei7Gto0J9m1PsvJyP4tW+QgX5D4xSMY1b78WY9vrcWwbllzs49JrA3kRmxg9rTUJJ07psUI6XjhA8WeW8vJTURau8LJgRf4kC709SM+sshtr3I7yBPEXz89VWGKU5hUvY27R0km3REEIMXVJQiqEEEAwbHDdHSGWX+LjtWeibHw2xqbnY8ye72HJah+zF3gwzIm9gHNszY43E7z1Uox4VFO71Mva6wMUl+XftE4xckk3gcbFnmUQXjgTTLhkXf7daLDj5yek7QefwBOulIR0EtnZtAlTWSwuWyPJqBAir0hCKoQQZyivsrjzYwW0tzjsfjvBnrcTHNmXIhRRLFrpY8lqH0Wl45sQuq5m3/YkbzwXo6vDZcY8i8tvCFJRLR/ZU0nCjgFgFvvw1RQCMH1GfiWjkK6wC+l1o70qLvkirpPIVUhihA60bueN0xuYU7iYRaWrJSEVQuQVuboRQoh+FJWaXHFDkLXXBTi6P8XuLQnefi3OllfiVM+xWHKxj9qlXixP9i7stNYc2Zti47MxWhsdplWZXHdXhJnzpGjRVJRw0glpoDtA6mQ3RmWI5jqHgmKDQCh/EtPeKbvmGUWNDE8QwyOVnSeDYx37ePXkk1RH5rBu5p2SjAoh8o4kpEIIMQjDVMxd7GXuYi/dnS57tybY/XaCDb/r4aU/RFl4kZelF/sorxrbx+npYylefyZG3XGbolKDm+8LU7t08rWiEcOXzCSk4X1euppOErx/AW++GGPFZT5mzffmOLp32bEWDE8Yw0qva020HyLespfIrOswrECOoxODqes+xgvHf0dpoJL3zPoApiGXfUKI/COfTEIIMUzhAoM16wKsvtrPqWM2uzcn2L0lwY43EpRXmSxd7WPBcu+IKqQ216d7iR7dn54WfO3tQRZf7MOc4PWqYuLFM1N2LW2hTIXrprcrI7/+7e1Yy1nrR+PNu+k8uoGCOTfmMCoxlJZYPc8e/U8i3iJunPMhPGb+3OQQQogzSUIqhBAjpAxFzRwPNXM8XBNz2bc9ye4tCV58Msqrf4oyb4mXpWt8VM2yBhzh7GjL9BJ9J91L9IobAqxY68fjza9kRIyf3im7lvbAGQlpnuWjOPGWs1q+pHrqsILlKEOmkuerjkQLTx9+CI/h4+Y5D+C3ZHq1ECJ/SUIqhBBj4A8YXLTWz4rLfDSddti1JcH+d5Ls256kqNRgycU+Fq3yEYqkR02j3S6bX4qx460ESsHFV/pZfY30Er0QvZuQpkdItdYAGHlWRNmOteArmtf3faq7Hk+kKocRicFEU108ffghNJqb536YkLcg1yEJIcSgplRCqpS6DbittrY216EIIS4wSimmVVtMq7a46uYgB3elR01f3xBj43Mx5iz0UFRqsuNN6SUq0pJOHABTW+kRUie93cizIVI71kKo8lIAtJPCjjURnL46x1GJgfSkOnG15sY591PkLxv6CUIIkWNTKiHVWj8JPLlmzZpP5ToWIcSFy+NVLF7lY/EqH21NmfYxWxMc3pOidqmHte8JUlyeZ8NgYsL1tX1xTZRlvLuGNI/yUe2kcJOdmJk1pKloA6DxhKbnNjBxHle7GMqgPFjNvYv+SgoYCSEmDfm0EkKIcVRcbnLlTUHWXh8gHtV9U3eFiGem7BquifIqdO8a0jy6V2HHMz1IMy1fUj31AHjClTmLSZzPcR2eO/ZbpodmsmLaFZKMCiEmFfnEEkKICWCailAkj4a+RM4lnBgew4tyyBQ1yqwhzaN7FnasFaCvyq7dXQcYWMFpOYxK9Mdr+vCavlyHIYQQIyYJqRBCCJEDCTuGzwyAo1GmkZdtX5xYeoS0b8pud2+FXbl8yAdaa2w3icf0sW7GndK3WAgxKeXRfVghhBDiwpF00glpaH0l/tVlFBYbXLIuvwpd9U3ZzSSkTqpbpuvmke2Nr/JfB35E3I5KMiqEmLTkFqcQQgiRAwknhs8K4JkR7ts2fUb+JKOQrrALYPnTfUgrLvlvaNfOZUgiY0/LFrbUv8j84hXpkXYhhJik8uvMJ4QQQlwg4pkpu8nj3TitCWI9Lg2nbOyUznVofZxYC8oKYHiCfdtkum7uHWnfzcZTf2RGwXyurHmfjI4KISY1SUiFEEKIHOgdIe36/THiO1tpbnB484UYiXj+JKR2vKVvum6saSfN7/wHTrI7x1Fd2E51Heal448zLVjDdbPuwcynssxCCDEKcptTCCGEyIGkE8dnTqfwnjmogEWF1+Kqm4L4g/kz2mXHWvpavrh2lFRPPYblz3FUF67GnpM8d/Q3FPpLuWHOB7EMT65DEkKIMZOEVAghRF5SSt0MfAcwgf/QWn/znMdnAT8GyoFW4AGt9cnMYw6wI7Prca317RMW+DAlnBg+049VkZ4OawJef36NdjmxFjwFswAIVV5KqPLSHEd04WqPN7Hh6MMErDA3zfkwPlNuDAghpgaZsiuEECLvKKVM4PvALcAS4ENKqSXn7PZt4Oda6xXAN4B/OuOxmNZ6Zea/vEtGIdP2RQWI72zFbo7T2eZw/GAK18mjKbux1r4puyJ3oqlu/nT4VygMbpp7P0FPeOgnCSHEJCEJqRBCiHx0KXBQa31Ya50EHgbuOGefJcDzma9f6OfxvKW1JuHECRCk5/lTpE5003jaYfumeF8/0lzTro2TaMcKlKCdFKdf/Ro9dW/mOqwLkt8KMqdoMTfNvZ8CX0muwxFCiKyShFQIIUQ+qgZOnPH9ycy2M20H7s58fRcQUUr1Duf5lVKblVKblFJ3jmuko5Byk2hcfKTbdShTod30yKiRJ2dmO94GaEx/KaloA068RSrsTrCUk6An1YmhDC6rupHSwPRchySEEFmntM6fqUHZopRqAo6N8ullQHMWwxlPEuv4kFjHz2SK90KJdZbWujybwWSDUur9wM1a6z/PfP8R4DKt9WfP2KcK+B4wB3gZuAdYprVuV0pVa61PKaXmkh5FfY/W+tA5x/g08OnMtwuBfWMI+UJ5v0w0iXV8SKzjYzLFClPw3CEmpyl5q3MsfyBKqc1a6zXZjGe8SKzjQ2IdP5MpXok1504BM874viazrY/W+jSZEVKlVBi4R2vdnnnsVOb/h5VSLwKrgEPnPP9B4MFsBDuZ/g0k1vEhsY4PiXX8TLZ4xdSVJxODhBBCiLO8BcxXSs1RSnmBDwJPnLmDUqpMKdV7Hvs70hV3UUoVK6V8vfsAVwK7JyxyIYQQQgybJKRCCCHyjtbaBj4LPA3sAX6jtd6llPqGUqq3au56YJ9Saj9QAfxjZvtiYLNSajvpYkff1FpLQiqEEELkoSk5ZXeMsjJ9a4JIrONDYh0/kyleiTXHtNZPAU+ds+1/nPH1I8Aj/TzvdWD5uAd4tsn0byCxjg+JdXxIrONnssUrpqgpWdRICCGEEEIIIUT+kym7QgghhBBCCCFyQhLSDKXUzUqpfUqpg0qpr+Q6nsEopWYopV5QSu1WSu1SSv1NrmMailLKVEptVUr9PtexDEYpVaSUekQptVcptUcpdXmuYxqIUuoLmX//nUqpXyul/LmOqZdS6sdKqUal1M4ztpUopTYopQ5k/l+cyxjPNEC838q8D95RSj2mlCrKYYh9+ov1jMe+qJTSmUI+YgJMlnOHnDfGl5w7smMynTvkvCFE9khCSvqkB3wfuAVYAnxIKbUkt1ENyga+qLVeAqwF/irP4wX4G9KFSfLdd4A/aa0XAReRpzErpaqBzwFrtNbLAJN0FdJ88VPg5nO2fQV4Tms9H3gu832++Cnnx7uBdE/LFcB+0lVc88FPOT9WlFIzgBuB4xMd0IVqkp075LwxvuTckR0/ZfKcO36KnDeEyApJSNMuBQ5qrQ9rrZPAw8AdOY5pQFrrOq3125mvu0if+KpzG9XAlFI1wHuB/8h1LINRShUC1wA/AtBaJ3t7GuYpCwgopSwgCJzOcTx9tNYvA63nbL4D+Fnm658Bd05kTIPpL16t9TOZSq8Am0j3wcy5AX63AP8H+FtACgNMnElz7pDzxviRc0f2TKZzh5w3hMgeSUjTqoETZ3x/kjw+UZ9JKTWbdMP3N3IcymD+lfQHnpvjOIYyB2gCfpKZJvYfSqlQroPqj9b6FPBt0nc164AOrfUzuY1qSBVa67rM1/Wk23RMFp8A/pjrIAailLoDOKW13p7rWC4wk/LcIeeNrJNzx/iarOcOOW8IMUySkE5iSqkw8Dvg81rrzlzH0x+l1PuARq31llzHMgwWcDHwf7XWq4Ae8mdq0Fkya2juIH0hVAWElFIP5Daq4dPp8t6T4o6sUuqrpKc7PpTrWPqjlAoC/x/gfwy1rxBy3hgXcu6YIJPl3CHnDSFGRhLStFPAjDO+r8lsy1tKKQ/pi4qHtNaP5jqeQVwJ3K6UOkp6Ott1Sqlf5jakAZ0ETmqte0cNHiF9kZGPrgeOaK2btNYp4FHgihzHNJQGpVQlQOb/jTmOZ0hKqY8B7wM+rPO3R9Y80heX2zN/ZzXA20qp6TmN6sIwqc4dct4YN3LuGF+T6twh5w0hRk4S0rS3gPlKqTlKKS/pBf5P5DimASmlFOm1Knu01v871/EMRmv9d1rrGq31bNK/1+e11nl5N1ZrXQ+cUEotzGx6D7A7hyEN5jiwVikVzLwf3kOeFtE4wxPARzNffxT4rxzGMiSl1M2kpwzerrWO5jqegWitd2itp2mtZ2f+zk4CF2fez2J8TZpzh5w3xo+cO8bdpDl3yHlDiNGRhBTILED/LPA06Q/m32itd+U2qkFdCXyE9F3jbZn/bs11UFPEXwMPKaXeAVYC/zO34fQvcyf+EeBtYAfpv+UHcxrUGZRSvwY2AguVUieVUp8EvgncoJQ6QPou/TdzGeOZBoj3e0AE2JD5G/t/cxpkxgCxihyYZOcOOW+MLzl3ZMFkOnfIeUOI7FH5O5tACCGEEEIIIcRUJiOkQgghhBBCCCFyQhJSIYQQQgghhBA5IQmpEEIIIYQQQoickIRUCCGEEEIIIUROSEIqhBBCCCGEECInJCEVYoSUUs4ZbRO2KaW+ksXXnq2U2pmt1xNCCJF7ct4QQoiBWbkOQIhJKKa1XpnrIIQQQkwact4QQogByAipEFmilDqqlPpfSqkdSqk3lVK1me2zlVLPK6XeUUo9p5SamdleoZR6TCm1PfPfFZmXMpVS/66U2qWUekYpFcjZDyWEEGLcyHlDCCEkIRViNALnTL2674zHOrTWy4HvAf+a2fZvwM+01iuAh4DvZrZ/F3hJa30RcDGwK7N9PvB9rfVSoB24Z1x/GiGEEONNzhtCCDEApbXOdQxCTCpKqW6tdbif7UeB67TWh5VSHqBea12qlGoGKrXWqcz2Oq11mVKqCajRWifOeI3ZwAat9fzM9/8d8Git/2ECfjQhhBDjQM4bQggxMBkhFSK79ABfj0TijK8dZK23EEJMZXLeEEJc0CQhFSK77jvj/xszX78OfDDz9YeBVzJfPwd8BkApZSqlCicqSCGEEHlDzhtCiAua3EETYuQCSqltZ3z/J611bwn/YqXUO6TvVn8os+2vgZ8opb4MNAEfz2z/G+BBpdQnSd/R/gxQN97BCyGEmHBy3hBCiAHIGlIhsiSzFmiN1ro517EIIYTIf3LeEEIImbIrhBBCCCGEECJHZIRUCCGEEEIIIUROyAipEEIIIYQQQoickIRUCCGEEEIIIUROSEIqhBBCCCGEECInJCEVQgghhBBCCJETkpAKIYQQQgghhMgJSUiFEEIIIYQQQuTE/x+2lCig+EFSTAAAAABJRU5ErkJggg==\n", - "text/plain": [ - "
" + "cell_type": "code", + "metadata": { + "id": "93-aeO7Sae-P" + }, + "source": [ + "model_train_history = {}\n", + "input_dim = x_train.shape[1]\n", + "validation_split = 0.20\n", + "epochs = 15" + ], + "execution_count": 3, + "outputs": [] + }, + { + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "PZ2p8pTDae-P", + "outputId": "7546402c-69b3-4ea1-9c30-49a4dce4129f" + }, + "source": [ + "model = keras.models.Sequential(name = 'ANN-3x16',\n", + " layers =\n", + " [\n", + " keras.layers.InputLayer(input_shape=input_dim),\n", + " keras.layers.Dense(16, activation=keras.activations.relu),\n", + " keras.layers.Dense(16, activation=keras.activations.relu),\n", + " keras.layers.Dense(16, activation=keras.activations.relu),\n", + " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", + " ]\n", + ")\n", + "\n", + "model.summary()\n", + "model.compile(\n", + " optimizer=keras.optimizers.Adam(),\n", + " loss='categorical_crossentropy',\n", + " metrics=['accuracy'],\n", + ")\n", + "\n", + "model_train_history[model] = model.fit(\n", + " x=x_train,\n", + " y=y_train,\n", + " epochs=epochs,\n", + " batch_size=4,\n", + " validation_split=validation_split,\n", + " shuffle=True,\n", + " verbose=1,\n", + ")" + ], + "execution_count": 4, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Model: \"ANN-3x16\"\n", + "_________________________________________________________________\n", + "Layer (type) Output Shape Param # \n", + "=================================================================\n", + "dense (Dense) (None, 16) 688 \n", + "_________________________________________________________________\n", + "dense_1 (Dense) (None, 16) 272 \n", + "_________________________________________________________________\n", + "dense_2 (Dense) (None, 16) 272 \n", + "_________________________________________________________________\n", + "dense_3 (Dense) (None, 27) 459 \n", + "=================================================================\n", + "Total params: 1,691\n", + "Trainable params: 1,691\n", + "Non-trainable params: 0\n", + "_________________________________________________________________\n", + "Epoch 1/15\n", + "870/870 [==============================] - 6s 3ms/step - loss: 2.0095 - accuracy: 0.4464 - val_loss: 0.7066 - val_accuracy: 0.7874\n", + "Epoch 2/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.3870 - accuracy: 0.9154 - val_loss: 0.1953 - val_accuracy: 0.9494\n", + "Epoch 3/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1556 - accuracy: 0.9655 - val_loss: 0.1074 - val_accuracy: 0.9805\n", + "Epoch 4/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0954 - accuracy: 0.9779 - val_loss: 0.0799 - val_accuracy: 0.9816\n", + "Epoch 5/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0665 - accuracy: 0.9836 - val_loss: 0.0762 - val_accuracy: 0.9805\n", + "Epoch 6/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0471 - accuracy: 0.9882 - val_loss: 0.0573 - val_accuracy: 0.9828\n", + "Epoch 7/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0391 - accuracy: 0.9891 - val_loss: 0.0868 - val_accuracy: 0.9701\n", + "Epoch 8/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0299 - accuracy: 0.9928 - val_loss: 0.0490 - val_accuracy: 0.9874\n", + "Epoch 9/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0236 - accuracy: 0.9945 - val_loss: 0.0406 - val_accuracy: 0.9851\n", + "Epoch 10/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0219 - accuracy: 0.9960 - val_loss: 0.0373 - val_accuracy: 0.9920\n", + "Epoch 11/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0225 - accuracy: 0.9934 - val_loss: 0.0324 - val_accuracy: 0.9920\n", + "Epoch 12/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0123 - accuracy: 0.9968 - val_loss: 0.0216 - val_accuracy: 0.9943\n", + "Epoch 13/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0157 - accuracy: 0.9954 - val_loss: 0.0146 - val_accuracy: 0.9943\n", + "Epoch 14/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0110 - accuracy: 0.9974 - val_loss: 0.0117 - val_accuracy: 0.9954\n", + "Epoch 15/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0099 - accuracy: 0.9974 - val_loss: 0.0154 - val_accuracy: 0.9943\n" + ], + "name": "stdout" + } ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "fig, axs = plt.subplots(1, 2, figsize=(14,5))\n", - "colors_graph = [\"#d69e2d\",\n", - " \"#927ced\",\n", - " \"#73bd4d\",\n", - " \"#e462c0\",\n", - " \"#eb5e52\"]\n", - "handles = []\n", - "\n", - "for (model, history), color in zip(model_train_history.items(), colors_graph):\n", - " label = '{} ({})'.format(model.name, model.count_params())\n", - " axs[0].plot(history.history['loss'], c=color, ls='-.', alpha=.7)\n", - " axs[1].plot(history.history['accuracy'], c=color, ls='-.', alpha=.7)\n", - " axs[0].plot(history.history['val_loss'], c=color)\n", - " axs[1].plot(history.history['val_accuracy'], c=color)\n", - " handles.append(mpatches.Patch(color=color, label=label))\n", - "\n", - "for ax in axs:\n", - " ax.set_xlabel('Epoch')\n", - "axs[0].set_ylabel('loss')\n", - "axs[0].set_yscale('log')\n", - "axs[1].set_ylabel('accuracy')\n", - "axs[1].set_ylim(0.95,1.01)\n", - "\n", - "handles.append(Line2D([0], [0], color='grey', lw=1, ls='-', label='validation'))\n", - "handles.append(Line2D([0], [0], color='grey', lw=1, ls='-.', label='training'))\n", - "\n", - "fig.subplots_adjust(right=0.85)\n", - "fig.legend(handles=handles,\n", - " loc=\"center right\",\n", - " borderaxespad=1)" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "JmJ_LECNae-R" - }, - "source": [ - "## Model export" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "km8bpVuSrwPE" - }, - "source": [ - "Once you have a good model, you can save it on your Google Drive. The model is saved using the [folder hierarchy of OpenHand](https://github.com/ArthurFDLR/OpenHand-App#pose-classifier-models)." - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" }, - "id": "MyD7zsZSfuMk", - "outputId": "14e58157-df9a-45d2-db51-fa9f06d507e0" - }, - "outputs": [], - "source": [ - "from pathlib import Path\n", - "import json\n", - "\n", - "model_name = 'ANN_RightHand_1'\n", - "\n", - "if IN_COLAB:\n", - " content_path = Path('/').absolute() / 'content'\n", - " drive_path = content_path / 'drive'\n", - " google.colab.drive.mount(str(drive_path))\n", - " save_path = drive_path / 'My Drive'\n", - " \n", - " for subfolder in ['Pose Classification Kit', 'Models', model_name]:\n", - " save_path /= subfolder\n", - " if not (save_path).is_dir():\n", - " %mkdir \"{save_path}\"\n", - "else:\n", - " save_path = Path('.').absolute() / model_name\n", - " %mkdir \"{save_path}\"\n", - "\n", - "model_path = save_path / '{name}.h5'.format(name = model_name)" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" + { + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "Ils0Rcv4ae-Q", + "outputId": "baca714f-39c3-407b-ecd4-0ccd3aa0b4f9" + }, + "source": [ + "model = keras.models.Sequential(name = 'ANN-3x64',\n", + " layers =\n", + " [\n", + " keras.layers.InputLayer(input_shape=input_dim),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", + " ]\n", + ")\n", + "\n", + "model.summary()\n", + "model.compile(\n", + " optimizer=keras.optimizers.Adam(),\n", + " loss='categorical_crossentropy',\n", + " metrics=['accuracy'],\n", + ")\n", + "\n", + "model_train_history[model] = model.fit(\n", + " x=x_train,\n", + " y=y_train,\n", + " epochs=epochs,\n", + " batch_size=4,\n", + " validation_split=validation_split,\n", + " shuffle=True,\n", + " verbose=1,\n", + ")" + ], + "execution_count": 5, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Model: \"ANN-3x64\"\n", + "_________________________________________________________________\n", + "Layer (type) Output Shape Param # \n", + "=================================================================\n", + "dense_4 (Dense) (None, 64) 2752 \n", + "_________________________________________________________________\n", + "dense_5 (Dense) (None, 64) 4160 \n", + "_________________________________________________________________\n", + "dense_6 (Dense) (None, 64) 4160 \n", + "_________________________________________________________________\n", + "dense_7 (Dense) (None, 27) 1755 \n", + "=================================================================\n", + "Total params: 12,827\n", + "Trainable params: 12,827\n", + "Non-trainable params: 0\n", + "_________________________________________________________________\n", + "Epoch 1/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.8609 - accuracy: 0.7831 - val_loss: 0.1198 - val_accuracy: 0.9701\n", + "Epoch 2/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0710 - accuracy: 0.9825 - val_loss: 0.0346 - val_accuracy: 0.9897\n", + "Epoch 3/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0328 - accuracy: 0.9905 - val_loss: 0.0480 - val_accuracy: 0.9839\n", + "Epoch 4/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0211 - accuracy: 0.9937 - val_loss: 0.1835 - val_accuracy: 0.9264\n", + "Epoch 5/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0115 - accuracy: 0.9974 - val_loss: 0.0108 - val_accuracy: 0.9977\n", + "Epoch 6/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0241 - accuracy: 0.9931 - val_loss: 0.0175 - val_accuracy: 0.9931\n", + "Epoch 7/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0076 - accuracy: 0.9980 - val_loss: 0.0061 - val_accuracy: 0.9989\n", + "Epoch 8/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0119 - accuracy: 0.9971 - val_loss: 0.0076 - val_accuracy: 0.9989\n", + "Epoch 9/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0042 - accuracy: 0.9986 - val_loss: 0.0158 - val_accuracy: 0.9977\n", + "Epoch 10/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0126 - accuracy: 0.9957 - val_loss: 0.0093 - val_accuracy: 0.9989\n", + "Epoch 11/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0123 - accuracy: 0.9965 - val_loss: 0.0097 - val_accuracy: 0.9989\n", + "Epoch 12/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0055 - accuracy: 0.9983 - val_loss: 0.0053 - val_accuracy: 0.9989\n", + "Epoch 13/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0081 - accuracy: 0.9980 - val_loss: 0.0082 - val_accuracy: 0.9989\n", + "Epoch 14/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0017 - accuracy: 0.9991 - val_loss: 0.0131 - val_accuracy: 0.9966\n", + "Epoch 15/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0084 - accuracy: 0.9983 - val_loss: 0.0025 - val_accuracy: 0.9989\n" + ], + "name": "stdout" + } + ] }, - "id": "5pc5o55gae-R", - "outputId": "528e1d83-80dc-4de9-e00c-4f0eff944506" - }, - "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model: \"27Class_3x64\"\n", - "_________________________________________________________________\n", - "Layer (type) Output Shape Param # \n", - "=================================================================\n", - "dense_22 (Dense) (None, 64) 2752 \n", - "_________________________________________________________________\n", - "dense_23 (Dense) (None, 64) 4160 \n", - "_________________________________________________________________\n", - "dense_24 (Dense) (None, 64) 4160 \n", - "_________________________________________________________________\n", - "dense_25 (Dense) (None, 27) 1755 \n", - "=================================================================\n", - "Total params: 12,827\n", - "Trainable params: 12,827\n", - "Non-trainable params: 0\n", - "_________________________________________________________________\n", - "Epoch 1/10\n", - "\n", - "Epoch 00001: val_loss improved from inf to 0.10254, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.8606 - accuracy: 0.7704 - val_loss: 0.1025 - val_accuracy: 0.9786\n", - "Epoch 2/10\n", - "\n", - "Epoch 00002: val_loss improved from 0.10254 to 0.04877, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.0633 - accuracy: 0.9816 - val_loss: 0.0488 - val_accuracy: 0.9755\n", - "Epoch 3/10\n", - "\n", - "Epoch 00003: val_loss did not improve from 0.04877\n", - "924/924 - 2s - loss: 0.0257 - accuracy: 0.9930 - val_loss: 0.0543 - val_accuracy: 0.9709\n", - "Epoch 4/10\n", - "\n", - "Epoch 00004: val_loss improved from 0.04877 to 0.00274, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.0152 - accuracy: 0.9962 - val_loss: 0.0027 - val_accuracy: 1.0000\n", - "Epoch 5/10\n", - "\n", - "Epoch 00005: val_loss did not improve from 0.00274\n", - "924/924 - 2s - loss: 0.0253 - accuracy: 0.9949 - val_loss: 0.0066 - val_accuracy: 1.0000\n", - "Epoch 6/10\n", - "\n", - "Epoch 00006: val_loss improved from 0.00274 to 0.00130, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.0141 - accuracy: 0.9957 - val_loss: 0.0013 - val_accuracy: 1.0000\n", - "Epoch 7/10\n", - "\n", - "Epoch 00007: val_loss improved from 0.00130 to 0.00106, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.0061 - accuracy: 0.9984 - val_loss: 0.0011 - val_accuracy: 1.0000\n", - "Epoch 8/10\n", - "\n", - "Epoch 00008: val_loss improved from 0.00106 to 0.00079, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.0127 - accuracy: 0.9959 - val_loss: 7.8661e-04 - val_accuracy: 1.0000\n", - "Epoch 9/10\n", - "\n", - "Epoch 00009: val_loss did not improve from 0.00079\n", - "924/924 - 2s - loss: 0.0144 - accuracy: 0.9970 - val_loss: 0.0040 - val_accuracy: 0.9985\n", - "Epoch 10/10\n", - "\n", - "Epoch 00010: val_loss improved from 0.00079 to 0.00064, saving model to C:\\Users\\ArthF\\Documents\\Projects\\pose-classification-kit\\examples\\ANN_RightHand_1\\ANN_RightHand_1.h5\n", - "924/924 - 2s - loss: 0.0019 - accuracy: 0.9997 - val_loss: 6.3679e-04 - val_accuracy: 1.0000\n" - ] - } - ], - "source": [ - "model = keras.models.Sequential(name = '27Class_3x64',\n", - " layers =\n", - " [\n", - " keras.layers.InputLayer(input_shape=input_dim),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dense(64, activation=keras.activations.relu),\n", - " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", - " ]\n", - ")\n", - "\n", - "model.summary()\n", - "model.compile(\n", - " optimizer=keras.optimizers.Adam(),\n", - " loss='categorical_crossentropy',\n", - " metrics=['accuracy'],\n", - ")\n", - "\n", - "model.fit(\n", - " x=x_train,\n", - " y=y_train,\n", - " epochs=10,\n", - " batch_size=4,\n", - " validation_split=0.15,\n", - " shuffle=True,\n", - " callbacks=[keras.callbacks.ModelCheckpoint(filepath=model_path, verbose=1, save_best_only=True)],\n", - " verbose = 2,\n", - ")\n", - "\n", - "with open(save_path / 'class.json', 'w') as f:\n", - " json.dump({'labels':labels}, f)" - ] - }, - { - "cell_type": "code", - "execution_count": 18, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/" + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "lUaetVfOae-Q", + "outputId": "17dcee7c-f45c-4cdc-b959-fe1c6a9ae83a" + }, + "source": [ + "model = keras.models.Sequential(name = 'ANN-3x64-Dropouts',\n", + " layers =\n", + " [\n", + " keras.layers.InputLayer(input_shape=input_dim),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dropout(0.3),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dropout(0.3),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dropout(0.3),\n", + " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", + " ]\n", + ")\n", + "\n", + "model.summary()\n", + "model.compile(\n", + " optimizer=keras.optimizers.Adam(),\n", + " loss='categorical_crossentropy',\n", + " metrics=['accuracy'],\n", + ")\n", + "\n", + "model_train_history[model] = model.fit(\n", + " x=x_train,\n", + " y=y_train,\n", + " epochs=epochs,\n", + " batch_size=4,\n", + " validation_split=validation_split,\n", + " shuffle=True,\n", + " verbose=1,\n", + ")" + ], + "execution_count": 6, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Model: \"ANN-3x64-Dropouts\"\n", + "_________________________________________________________________\n", + "Layer (type) Output Shape Param # \n", + "=================================================================\n", + "dense_8 (Dense) (None, 64) 2752 \n", + "_________________________________________________________________\n", + "dropout (Dropout) (None, 64) 0 \n", + "_________________________________________________________________\n", + "dense_9 (Dense) (None, 64) 4160 \n", + "_________________________________________________________________\n", + "dropout_1 (Dropout) (None, 64) 0 \n", + "_________________________________________________________________\n", + "dense_10 (Dense) (None, 64) 4160 \n", + "_________________________________________________________________\n", + "dropout_2 (Dropout) (None, 64) 0 \n", + "_________________________________________________________________\n", + "dense_11 (Dense) (None, 27) 1755 \n", + "=================================================================\n", + "Total params: 12,827\n", + "Trainable params: 12,827\n", + "Non-trainable params: 0\n", + "_________________________________________________________________\n", + "Epoch 1/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 1.9459 - accuracy: 0.3955 - val_loss: 0.6170 - val_accuracy: 0.8299\n", + "Epoch 2/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.7669 - accuracy: 0.7110 - val_loss: 0.2344 - val_accuracy: 0.9368\n", + "Epoch 3/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.4771 - accuracy: 0.8274 - val_loss: 0.1153 - val_accuracy: 0.9747\n", + "Epoch 4/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.3480 - accuracy: 0.8737 - val_loss: 0.0718 - val_accuracy: 0.9713\n", + "Epoch 5/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.2661 - accuracy: 0.9108 - val_loss: 0.0545 - val_accuracy: 0.9793\n", + "Epoch 6/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.2071 - accuracy: 0.9327 - val_loss: 0.0327 - val_accuracy: 0.9885\n", + "Epoch 7/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1903 - accuracy: 0.9387 - val_loss: 0.0537 - val_accuracy: 0.9793\n", + "Epoch 8/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1523 - accuracy: 0.9523 - val_loss: 0.0266 - val_accuracy: 0.9908\n", + "Epoch 9/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1417 - accuracy: 0.9491 - val_loss: 0.0278 - val_accuracy: 0.9931\n", + "Epoch 10/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1288 - accuracy: 0.9583 - val_loss: 0.0226 - val_accuracy: 0.9931\n", + "Epoch 11/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1063 - accuracy: 0.9620 - val_loss: 0.0256 - val_accuracy: 0.9920\n", + "Epoch 12/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1051 - accuracy: 0.9661 - val_loss: 0.0291 - val_accuracy: 0.9897\n", + "Epoch 13/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.1155 - accuracy: 0.9609 - val_loss: 0.0119 - val_accuracy: 0.9943\n", + "Epoch 14/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0904 - accuracy: 0.9712 - val_loss: 0.0210 - val_accuracy: 0.9931\n", + "Epoch 15/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0754 - accuracy: 0.9727 - val_loss: 0.0169 - val_accuracy: 0.9943\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "FA1LSFoVae-R", + "outputId": "6b2a5b4d-7632-420b-8db9-92431cde3fa1" + }, + "source": [ + "model = keras.models.Sequential(name = 'ANN-2x128',\n", + " layers =\n", + " [\n", + " keras.layers.InputLayer(input_shape=input_dim),\n", + " keras.layers.Dense(128, activation=keras.activations.relu),\n", + " keras.layers.Dense(128, activation=keras.activations.relu),\n", + " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", + " ]\n", + ")\n", + "\n", + "model.summary()\n", + "model.compile(\n", + " optimizer=keras.optimizers.Adam(),\n", + " loss='categorical_crossentropy',\n", + " metrics=['accuracy'],\n", + ")\n", + "\n", + "model_train_history[model] = model.fit(\n", + " x=x_train,\n", + " y=y_train,\n", + " epochs=epochs,\n", + " batch_size=4,\n", + " validation_split=validation_split,\n", + " shuffle=True,\n", + " verbose=1,\n", + ")" + ], + "execution_count": 7, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Model: \"ANN-2x128\"\n", + "_________________________________________________________________\n", + "Layer (type) Output Shape Param # \n", + "=================================================================\n", + "dense_12 (Dense) (None, 128) 5504 \n", + "_________________________________________________________________\n", + "dense_13 (Dense) (None, 128) 16512 \n", + "_________________________________________________________________\n", + "dense_14 (Dense) (None, 27) 3483 \n", + "=================================================================\n", + "Total params: 25,499\n", + "Trainable params: 25,499\n", + "Non-trainable params: 0\n", + "_________________________________________________________________\n", + "Epoch 1/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.7049 - accuracy: 0.8473 - val_loss: 0.0971 - val_accuracy: 0.9793\n", + "Epoch 2/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0516 - accuracy: 0.9882 - val_loss: 0.0384 - val_accuracy: 0.9920\n", + "Epoch 3/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0249 - accuracy: 0.9940 - val_loss: 0.0871 - val_accuracy: 0.9690\n", + "Epoch 4/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0144 - accuracy: 0.9954 - val_loss: 0.0092 - val_accuracy: 0.9989\n", + "Epoch 5/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0162 - accuracy: 0.9965 - val_loss: 0.0164 - val_accuracy: 0.9954\n", + "Epoch 6/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0108 - accuracy: 0.9974 - val_loss: 0.0071 - val_accuracy: 0.9989\n", + "Epoch 7/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0085 - accuracy: 0.9977 - val_loss: 0.0267 - val_accuracy: 0.9920\n", + "Epoch 8/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0047 - accuracy: 0.9983 - val_loss: 0.0180 - val_accuracy: 0.9943\n", + "Epoch 9/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0037 - accuracy: 0.9983 - val_loss: 0.0065 - val_accuracy: 0.9989\n", + "Epoch 10/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0031 - accuracy: 0.9991 - val_loss: 0.0060 - val_accuracy: 0.9989\n", + "Epoch 11/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0086 - accuracy: 0.9980 - val_loss: 0.0032 - val_accuracy: 0.9989\n", + "Epoch 12/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0048 - accuracy: 0.9991 - val_loss: 0.0063 - val_accuracy: 0.9989\n", + "Epoch 13/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0017 - accuracy: 0.9994 - val_loss: 0.0045 - val_accuracy: 0.9989\n", + "Epoch 14/15\n", + "870/870 [==============================] - 2s 3ms/step - loss: 0.0056 - accuracy: 0.9991 - val_loss: 0.0045 - val_accuracy: 0.9989\n", + "Epoch 15/15\n", + "870/870 [==============================] - 3s 3ms/step - loss: 0.0012 - accuracy: 0.9994 - val_loss: 0.0042 - val_accuracy: 0.9989\n" + ], + "name": "stdout" + } + ] }, - "id": "mPMnJf1Vae-S", - "outputId": "18ae8c05-5c6f-41bb-9e04-580b70e05d8e" - }, - "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "34/34 [==============================] - 0s 2ms/step - loss: 0.0069 - accuracy: 0.9981\n" - ] + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 356 + }, + "id": "4UhLcq1Jae-R", + "outputId": "0778730e-c2d6-4a26-f64e-f9f531d8e061" + }, + "source": [ + "fig, axs = plt.subplots(1, 2, figsize=(14,5))\n", + "colors_graph = [\"#d69e2d\",\n", + " \"#927ced\",\n", + " \"#73bd4d\",\n", + " \"#e462c0\",\n", + " \"#eb5e52\"]\n", + "handles = []\n", + "\n", + "for (model, history), color in zip(model_train_history.items(), colors_graph):\n", + " label = '{} ({})'.format(model.name, model.count_params())\n", + " axs[0].plot(history.history['loss'], c=color, ls='-.', alpha=.7)\n", + " axs[1].plot(history.history['accuracy'], c=color, ls='-.', alpha=.7)\n", + " axs[0].plot(history.history['val_loss'], c=color)\n", + " axs[1].plot(history.history['val_accuracy'], c=color)\n", + " handles.append(mpatches.Patch(color=color, label=label))\n", + "\n", + "for ax in axs:\n", + " ax.set_xlabel('Epoch')\n", + "axs[0].set_ylabel('loss')\n", + "axs[0].set_yscale('log')\n", + "axs[1].set_ylabel('accuracy')\n", + "axs[1].set_ylim(0.95,1.01)\n", + "\n", + "handles.append(Line2D([0], [0], color='grey', lw=1, ls='-', label='validation'))\n", + "handles.append(Line2D([0], [0], color='grey', lw=1, ls='-.', label='training'))\n", + "\n", + "fig.subplots_adjust(right=0.85)\n", + "fig.legend(handles=handles,\n", + " loc=\"center right\",\n", + " borderaxespad=1)" + ], + "execution_count": 8, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 8 + }, + { + "output_type": "display_data", + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "tags": [], + "needs_background": "light" + } + } + ] }, { - "data": { - "text/plain": [ - "[0.006858584005385637, 0.9981481432914734]" + "cell_type": "markdown", + "metadata": { + "id": "JmJ_LECNae-R" + }, + "source": [ + "## Model export" ] - }, - "execution_count": 18, - "metadata": {}, - "output_type": "execute_result" + }, + { + "cell_type": "markdown", + "metadata": { + "id": "km8bpVuSrwPE" + }, + "source": [ + "Once you have a good model, you can save it on your Google Drive. The model is saved using the [folder hierarchy of OpenHand](https://github.com/ArthurFDLR/OpenHand-App#pose-classifier-models)." + ] + }, + { + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "MyD7zsZSfuMk", + "outputId": "45b5fecf-ce4f-48eb-f852-b0873c8efd60" + }, + "source": [ + "from pathlib import Path\n", + "import json\n", + "\n", + "model_name = 'ANN_RightHand_1'\n", + "\n", + "if IN_COLAB:\n", + " content_path = Path('/').absolute() / 'content'\n", + " drive_path = content_path / 'drive'\n", + " google.colab.drive.mount(str(drive_path))\n", + " save_path = drive_path / 'My Drive'\n", + " \n", + " for subfolder in ['Pose Classification Kit', 'Models', model_name]:\n", + " save_path /= subfolder\n", + " if not (save_path).is_dir():\n", + " %mkdir \"{save_path}\"\n", + "else:\n", + " save_path = Path('.').absolute() / model_name\n", + " %mkdir \"{save_path}\"\n", + "\n", + "model_path = save_path / '{name}.h5'.format(name = model_name)" + ], + "execution_count": 9, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Mounted at /content/drive\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "5pc5o55gae-R", + "outputId": "5a4ca725-206e-4ee8-e4d1-8195807388bb" + }, + "source": [ + "model = keras.models.Sequential(name = '27Class_3x64',\n", + " layers =\n", + " [\n", + " keras.layers.InputLayer(input_shape=input_dim),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dense(64, activation=keras.activations.relu),\n", + " keras.layers.Dense(len(labels), activation=keras.activations.softmax),\n", + " ]\n", + ")\n", + "\n", + "model.summary()\n", + "model.compile(\n", + " optimizer=keras.optimizers.Adam(),\n", + " loss='categorical_crossentropy',\n", + " metrics=['accuracy'],\n", + ")\n", + "\n", + "model.fit(\n", + " x=x_train,\n", + " y=y_train,\n", + " epochs=10,\n", + " batch_size=4,\n", + " validation_split=0.15,\n", + " shuffle=True,\n", + " callbacks=[keras.callbacks.ModelCheckpoint(filepath=model_path, verbose=1, save_best_only=True)],\n", + " verbose = 2,\n", + ")\n", + "\n", + "with open(save_path / 'class.json', 'w') as f:\n", + " json.dump({'labels':labels}, f)" + ], + "execution_count": 10, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Model: \"27Class_3x64\"\n", + "_________________________________________________________________\n", + "Layer (type) Output Shape Param # \n", + "=================================================================\n", + "dense_15 (Dense) (None, 64) 2752 \n", + "_________________________________________________________________\n", + "dense_16 (Dense) (None, 64) 4160 \n", + "_________________________________________________________________\n", + "dense_17 (Dense) (None, 64) 4160 \n", + "_________________________________________________________________\n", + "dense_18 (Dense) (None, 27) 1755 \n", + "=================================================================\n", + "Total params: 12,827\n", + "Trainable params: 12,827\n", + "Non-trainable params: 0\n", + "_________________________________________________________________\n", + "Epoch 1/10\n", + "924/924 - 3s - loss: 0.8317 - accuracy: 0.7856 - val_loss: 0.1262 - val_accuracy: 0.9648\n", + "\n", + "Epoch 00001: val_loss improved from inf to 0.12617, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 2/10\n", + "924/924 - 2s - loss: 0.0701 - accuracy: 0.9827 - val_loss: 0.0483 - val_accuracy: 0.9816\n", + "\n", + "Epoch 00002: val_loss improved from 0.12617 to 0.04829, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 3/10\n", + "924/924 - 2s - loss: 0.0372 - accuracy: 0.9897 - val_loss: 0.0259 - val_accuracy: 0.9877\n", + "\n", + "Epoch 00003: val_loss improved from 0.04829 to 0.02588, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 4/10\n", + "924/924 - 2s - loss: 0.0262 - accuracy: 0.9946 - val_loss: 0.0203 - val_accuracy: 0.9893\n", + "\n", + "Epoch 00004: val_loss improved from 0.02588 to 0.02029, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 5/10\n", + "924/924 - 2s - loss: 0.0177 - accuracy: 0.9940 - val_loss: 0.0290 - val_accuracy: 0.9908\n", + "\n", + "Epoch 00005: val_loss did not improve from 0.02029\n", + "Epoch 6/10\n", + "924/924 - 2s - loss: 0.0143 - accuracy: 0.9968 - val_loss: 0.0099 - val_accuracy: 0.9969\n", + "\n", + "Epoch 00006: val_loss improved from 0.02029 to 0.00991, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 7/10\n", + "924/924 - 2s - loss: 0.0095 - accuracy: 0.9973 - val_loss: 0.0127 - val_accuracy: 0.9969\n", + "\n", + "Epoch 00007: val_loss did not improve from 0.00991\n", + "Epoch 8/10\n", + "924/924 - 2s - loss: 0.0094 - accuracy: 0.9986 - val_loss: 0.0065 - val_accuracy: 0.9985\n", + "\n", + "Epoch 00008: val_loss improved from 0.00991 to 0.00651, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 9/10\n", + "924/924 - 2s - loss: 0.0048 - accuracy: 0.9992 - val_loss: 0.0044 - val_accuracy: 0.9985\n", + "\n", + "Epoch 00009: val_loss improved from 0.00651 to 0.00438, saving model to /content/drive/My Drive/Pose Classification Kit/Models/ANN_RightHand_1/ANN_RightHand_1.h5\n", + "Epoch 10/10\n", + "924/924 - 2s - loss: 0.0173 - accuracy: 0.9968 - val_loss: 0.0232 - val_accuracy: 0.9954\n", + "\n", + "Epoch 00010: val_loss did not improve from 0.00438\n" + ], + "name": "stdout" + } + ] + }, + { + "cell_type": "code", + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "mPMnJf1Vae-S", + "outputId": "79d6066f-d31e-4d54-e88f-5cb1f405b8d8" + }, + "source": [ + "model = keras.models.load_model(model_path)\n", + "model.evaluate(x=dataset['x_test'], y=dataset['y_test_onehot'])" + ], + "execution_count": 11, + "outputs": [ + { + "output_type": "stream", + "text": [ + "34/34 [==============================] - 0s 2ms/step - loss: 0.0071 - accuracy: 0.9981\n" + ], + "name": "stdout" + }, + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "[0.007146174553781748, 0.9981481432914734]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 11 + } + ] + }, + { + "cell_type": "code", + "metadata": { + "id": "V3lsa7aKae-S" + }, + "source": [ + "" + ], + "execution_count": null, + "outputs": [] } - ], - "source": [ - "model = keras.models.load_model(model_path)\n", - "model.evaluate(x=dataset['x_test'], y=dataset['y_test_onehot'])" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "V3lsa7aKae-S" - }, - "outputs": [], - "source": [] - } - ], - "metadata": { - "accelerator": "GPU", - "colab": { - "include_colab_link": true, - "name": "OpenHand-Models.ipynb", - "provenance": [], - "toc_visible": true - }, - "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": "3.7.9" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} + ] +} \ No newline at end of file