diff --git a/kgcnn/backend/_tensorflow.py b/kgcnn/backend/_tensorflow.py index e2b0689b..11ec554b 100644 --- a/kgcnn/backend/_tensorflow.py +++ b/kgcnn/backend/_tensorflow.py @@ -54,4 +54,4 @@ def decompose_ragged_tensor(x, batch_dtype="int64"): def norm(x, ord='fro', axis=None, keepdims=False): - return tf.norm(x, ord=ord, dim=axis, keepdims=keepdims) \ No newline at end of file + return tf.norm(x, ord=ord, axis=axis, keepdims=keepdims) \ No newline at end of file diff --git a/kgcnn/layers/casting.py b/kgcnn/layers/casting.py index f7d59887..2c4f4eeb 100644 --- a/kgcnn/layers/casting.py +++ b/kgcnn/layers/casting.py @@ -407,7 +407,7 @@ def call(self, inputs: list, **kwargs): if self.static_output_shape is not None: target_shape = (ops.shape(attr_len)[0], self.static_output_shape[0]) else: - target_shape = (ops.shape(attr_len)[0], ops.amax(attr_len)) + target_shape = (ops.shape(attr_len)[0], ops.cast(ops.amax(attr_len), dtype="int32")) if not self.padded_disjoint: if attr_id is None: diff --git a/kgcnn/literature/GNNExplain/__init__.py b/kgcnn/literature/GNNExplain/__init__.py index e69de29b..05389503 100644 --- a/kgcnn/literature/GNNExplain/__init__.py +++ b/kgcnn/literature/GNNExplain/__init__.py @@ -0,0 +1,7 @@ +from ._model import GNNExplainerOptimizer, GNNInterface, GNNExplainer + +__all__ = [ + "GNNExplainerOptimizer", + "GNNInterface", + "GNNExplainer" +] diff --git a/kgcnn/literature/GNNExplain/_model.py b/kgcnn/literature/GNNExplain/_model.py index 90db3561..d135e6d6 100644 --- a/kgcnn/literature/GNNExplain/_model.py +++ b/kgcnn/literature/GNNExplain/_model.py @@ -194,10 +194,13 @@ def explain(self, graph_instance, output_to_explain=None, inspection=False, **kw gnnx_optimizer = GNNExplainerOptimizer( self.gnn, graph_instance, **self.gnnexplaineroptimizer_options) self.gnnx_optimizer = gnnx_optimizer + if output_to_explain is not None: gnnx_optimizer.output_to_explain = output_to_explain + gnnx_optimizer.compile(**self.compile_options) - gnnx_optimizer.fit(graph_instance, **fit_options) + + gnnx_optimizer.fit(x=graph_instance, y=gnnx_optimizer.output_to_explain, **fit_options) # Read out information from inspection_callback if inspection: @@ -265,21 +268,20 @@ def __init__(self, graph_instance): self.node_mask_loss = [] def on_epoch_begin(self, epoch, logs=None): - masked = self.model.call(self.graph_instance).numpy()[0] + masked = ops.convert_to_numpy(self.model.call(self.graph_instance))[0] self.predictions.append(masked) def on_epoch_end(self, epoch, logs=None): """After epoch.""" - index = 0 - losses_list = [loss_iter.numpy() for loss_iter in self.model.losses] if self.model.edge_mask_loss_weight > 0: - self.edge_mask_loss.append(losses_list[index]) - index = index + 1 + self.edge_mask_loss.append(ops.convert_to_numpy(self.model._metric_edge_tracker.result())) + self.model._metric_edge_tracker.reset_state() if self.model.feature_mask_loss_weight > 0: - self.feature_mask_loss.append(losses_list[index]) - index = index + 1 + self.feature_mask_loss.append(ops.convert_to_numpy(self.model._metric_feature_tracker.result())) + self.model._metric_feature_tracker.reset_state() if self.model.node_mask_loss_weight > 0: - self.node_mask_loss.append(losses_list[index]) + self.node_mask_loss.append(ops.convert_to_numpy(self.model._metric_node_tracker.result())) + self.model._metric_node_tracker.reset_state() self.total_loss.append(logs['loss']) @@ -320,6 +322,9 @@ def __init__(self, gnn_model, graph_instance, """ super(GNNExplainerOptimizer, self).__init__(**kwargs) self.gnn_model = gnn_model + self._metric_node_tracker = ks.metrics.Mean(name="mask_loss") + self._metric_edge_tracker = ks.metrics.Mean(name="mask_loss") + self._metric_feature_tracker = ks.metrics.Mean(name="mask_loss") self._edge_mask_dim = self.gnn_model.get_number_of_edges( graph_instance) self._feature_mask_dim = self.gnn_model.get_number_of_node_features( @@ -368,7 +373,7 @@ def call(self, graph_input, training: bool = False, **kwargs): training (bool): If training mode. Default is False. Returns: - tf.tensor: Masked prediction of GNN model. + Tensor: Masked prediction of GNN model. """ edge_mask = self.get_mask("edge") feature_mask = self.get_mask("feature") @@ -377,16 +382,19 @@ def call(self, graph_input, training: bool = False, **kwargs): # edge_mask loss if self.edge_mask_loss_weight > 0: - self.add_loss(lambda: norm(ops.sigmoid( - self.edge_mask), ord=self.edge_mask_norm_ord) * self.edge_mask_loss_weight) + loss = norm(ops.sigmoid(self.edge_mask), ord=self.edge_mask_norm_ord) * self.edge_mask_loss_weight + self.add_loss(loss) + self._metric_edge_tracker.update_state([loss]) # feature_mask loss if self.feature_mask_loss_weight > 0: - self.add_loss(lambda: norm(ops.sigmoid( - self.feature_mask), ord=self.feature_mask_norm_ord) * self.feature_mask_loss_weight) + loss = norm(ops.sigmoid(self.feature_mask), ord=self.feature_mask_norm_ord) * self.feature_mask_loss_weight + self.add_loss(loss) + self._metric_feature_tracker.update_state([loss]) # node_mask loss if self.node_mask_loss_weight > 0: - self.add_loss(lambda: norm(ops.sigmoid( - self.node_mask), ord=self.node_mask_norm_ord) * self.node_mask_loss_weight) + loss = norm(ops.sigmoid(self.node_mask), ord=self.node_mask_norm_ord) * self.node_mask_loss_weight + self.add_loss(loss) + self._metric_node_tracker.update_state([loss]) return y_pred diff --git a/notebooks/graph_explanation/explain_GNNExplain_cora.ipynb b/notebooks/graph_explanation/explain_GNNExplain_cora.ipynb index 209965ca..83bd8332 100644 --- a/notebooks/graph_explanation/explain_GNNExplain_cora.ipynb +++ b/notebooks/graph_explanation/explain_GNNExplain_cora.ipynb @@ -5,22 +5,26 @@ "execution_count": 1, "id": "ebf591c1", "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Using TensorFlow backend\n" + ] + } + ], "source": [ "import time\n", - "\n", + "import keras_core as ks\n", + "from keras_core import ops\n", "import matplotlib.pyplot as plt\n", "import networkx as nx\n", "import numpy as np\n", - "import tensorflow as tf\n", "from sklearn.model_selection import train_test_split\n", - "\n", "from kgcnn.data.datasets.CoraLuDataset import CoraLuDataset\n", "from kgcnn.literature.GCN import make_model\n", "from kgcnn.literature.GNNExplain import GNNExplainer, GNNInterface\n", - "from kgcnn.graph.adj import precompute_adjacency_scaled, sort_edge_indices, make_adjacency_from_edge_indices, \\\n", - " make_adjacency_undirected_logical_or, convert_scaled_adjacency_to_list\n", - "from kgcnn.data.utils import ragged_tensor_from_nested_numpy\n", "from kgcnn.training.scheduler import LinearLearningRateScheduler" ] }, @@ -53,23 +57,17 @@ ], "source": [ "dataset = CoraLuDataset()\n", - "nodes, edge_index, labels = dataset.obtain_property(\"node_attributes\"), dataset.obtain_property(\"edge_indices\"), dataset.obtain_property(\"node_labels\")\n", + "dataset.map_list(**{\"method\": \"make_undirected_edges\"})\n", + "dataset.map_list(**{\"method\": \"add_edge_self_loops\"})\n", + "dataset.map_list(**{\"method\": \"normalize_edge_weights_sym\"})\n", + "dataset.map_list(**{\"method\": \"count_nodes_and_edges\"})\n", + "dataset[0][\"node_attributes\"] = dataset[0][\"node_attributes\"][:, 1:] # remove ids\n", "class_label_mapping = dataset.class_label_mapping\n", - "labels = labels[0]\n", - "nodes = nodes[0][:, 1:] # Remove IDs\n", - "edge_index = sort_edge_indices(edge_index[0])\n", - "adj_matrix = make_adjacency_from_edge_indices(edge_index)\n", - "adj_matrix = precompute_adjacency_scaled(make_adjacency_undirected_logical_or(adj_matrix))\n", - "edge_index, edge_weight = convert_scaled_adjacency_to_list(adj_matrix)\n", - "edge_weight = np.expand_dims(edge_weight, axis=-1)\n", - "# labels = np.expand_dims(labels, axis=-1)\n", - "# labels = np.array(labels == np.arange(7), dtype=np.float32)\n", "\n", "# Find a color to visualize a label\n", "def get_label_color(label):\n", " return plt.get_cmap('Set1')(label / 7)\n", "\n", - "\n", "# Map label to class\n", "def get_label_name(label):\n", " return [\"Case_Based\",\n", @@ -92,6 +90,22 @@ { "cell_type": "code", "execution_count": 3, + "id": "d85f8fff-214b-499b-a17a-26291a2c795e", + "metadata": {}, + "outputs": [], + "source": [ + "model_inputs = [\n", + " {'shape': (None, 1432), 'name': \"node_attributes\", 'dtype': 'float32'},\n", + " {'shape': (None, 1), 'name': \"edge_attributes\", 'dtype': 'float32'},\n", + " {'shape': (None, 2), 'name': \"edge_indices\", 'dtype': 'int64'},\n", + " {\"shape\": (), \"name\": \"total_nodes\", \"dtype\": \"int64\"},\n", + " {\"shape\": (), \"name\": \"total_edges\", \"dtype\": \"int64\"}\n", + " ]" + ] + }, + { + "cell_type": "code", + "execution_count": 4, "id": "e561f2c8", "metadata": {}, "outputs": [ @@ -99,7 +113,7 @@ "name": "stdout", "output_type": "stream", "text": [ - "[TensorShape([1, None, 1432]), TensorShape([1, None, 1]), TensorShape([1, None, 2])]\n", + "[(1, 2708, 1432), (1, 13264, 1), (1, 13264, 2), (1,), (1,)]\n", "(1, 2708, 7)\n" ] } @@ -108,6 +122,7 @@ "# Make test/train split\n", "# Since only one graph in the dataset\n", "# Use a mask to hide test nodes labels\n", + "labels = dataset.get(\"node_labels\")[0]\n", "inds = np.arange(len(labels))\n", "ind_train, ind_val = train_test_split(inds, test_size=0.10, random_state=0)\n", "val_mask = np.zeros_like(inds)\n", @@ -117,12 +132,8 @@ "val_mask = np.expand_dims(val_mask, axis=0) # One graph in batch\n", "train_mask = np.expand_dims(train_mask, axis=0) # One graph in batch\n", "\n", - "# Make ragged graph tensors with 1 graph in batch\n", - "nodes, edges, edge_indices = ragged_tensor_from_nested_numpy([nodes]), ragged_tensor_from_nested_numpy(\n", - " [edge_weight]), ragged_tensor_from_nested_numpy([edge_index]) # One graph in batch\n", - "\n", "# Set training data. But requires mask and batch-dimension of 1\n", - "xtrain = nodes, edges, edge_indices\n", + "xtrain = dataset.tensor(model_inputs)\n", "ytrain = np.expand_dims(labels, axis=0) # One graph in batch\n", "print([x.shape for x in xtrain])\n", "print(ytrain.shape)" @@ -138,7 +149,7 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 5, "id": "1d611691", "metadata": {}, "outputs": [ @@ -146,780 +157,211 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:kgcnn.model.utils:Updated model kwargs:\n", - "INFO:kgcnn.model.utils:{'name': 'GCN', 'inputs': [{'shape': (None, 1432), 'name': 'node_attributes', 'dtype': 'float32', 'ragged': True}, {'shape': (None, 1), 'name': 'edge_attributes', 'dtype': 'float32', 'ragged': True}, {'shape': (None, 2), 'name': 'edge_indices', 'dtype': 'int64', 'ragged': True}], 'input_embedding': {'node': {'input_dim': 95, 'output_dim': 64}, 'edge': {'input_dim': 10, 'output_dim': 64}}, 'gcn_args': {'units': 124, 'use_bias': True, 'activation': 'relu', 'pooling_method': 'sum'}, 'depth': 3, 'verbose': 10, 'output_embedding': 'node', 'output_to_tensor': True, 'output_mlp': {'use_bias': [True, True, False], 'units': [64, 16, 7], 'activation': ['relu', 'relu', 'softmax']}}\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model: \"model\"\n", - "__________________________________________________________________________________________________\n", - " Layer (type) Output Shape Param # Connected to \n", - "==================================================================================================\n", - " node_attributes (InputLayer) [(None, None, 1432) 0 [] \n", - " ] \n", - " \n", - " optional_input_embedding (Opti (None, None, 1432) 0 ['node_attributes[0][0]'] \n", - " onalInputEmbedding) \n", - " \n", - " edge_attributes (InputLayer) [(None, None, 1)] 0 [] \n", - " \n", - " dense_embedding (DenseEmbeddin (None, None, 124) 177692 ['optional_input_embedding[0][0]'\n", - " g) ] \n", - " \n", - " optional_input_embedding_1 (Op (None, None, 1) 0 ['edge_attributes[0][0]'] \n", - " tionalInputEmbedding) \n", - " \n", - " edge_indices (InputLayer) [(None, None, 2)] 0 [] \n", - " \n", - " gcn (GCN) (None, None, 124) 15500 ['dense_embedding[0][0]', \n", - " 'optional_input_embedding_1[0][0\n", - " ]', \n", - " 'edge_indices[0][0]'] \n", - " \n", - " gcn_1 (GCN) (None, None, 124) 15500 ['gcn[0][0]', \n", - " 'optional_input_embedding_1[0][0\n", - " ]', \n", - " 'edge_indices[0][0]'] \n", - " \n", - " gcn_2 (GCN) (None, None, 124) 15500 ['gcn_1[0][0]', \n", - " 'optional_input_embedding_1[0][0\n", - " ]', \n", - " 'edge_indices[0][0]'] \n", - " \n", - " mlp (MLP) (None, None, 7) 9152 ['gcn_2[0][0]'] \n", - " \n", - " change_tensor_type (ChangeTens (None, None, 7) 0 ['mlp[0][0]'] \n", - " orType) \n", - " \n", - "==================================================================================================\n", - "Total params: 233,344\n", - "Trainable params: 233,344\n", - "Non-trainable params: 0\n", - "__________________________________________________________________________________________________\n", - "None\n", - "Epoch 1/10\n" + "INFO:kgcnn.models.utils:Updated model kwargs: '{'name': 'GCN', 'inputs': [{'shape': (None, 1432), 'name': 'node_attributes', 'dtype': 'float32'}, {'shape': (None, 1), 'name': 'edge_attributes', 'dtype': 'float32'}, {'shape': (None, 2), 'name': 'edge_indices', 'dtype': 'int64'}, {'shape': (), 'name': 'total_nodes', 'dtype': 'int64'}, {'shape': (), 'name': 'total_edges', 'dtype': 'int64'}], 'input_tensor_type': 'padded', 'input_embedding': None, 'cast_disjoint_kwargs': {}, 'input_node_embedding': {'input_dim': 95, 'output_dim': 64}, 'input_edge_embedding': {'input_dim': 25, 'output_dim': 1}, 'gcn_args': {'units': 124, 'use_bias': True, 'activation': 'relu', 'pooling_method': 'sum'}, 'depth': 3, 'verbose': 10, 'node_pooling_args': {'pooling_method': 'scatter_sum'}, 'output_embedding': 'node', 'output_to_tensor': None, 'output_tensor_type': 'padded', 'output_mlp': {'use_bias': [True, True, False], 'units': [64, 16, 7], 'activation': ['relu', 'relu', 'softmax']}, 'output_scaling': None}'.\n" ] }, { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\keras\\optimizers\\optimizer_v2\\adam.py:110: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n", - " super(Adam, self).__init__(name, **kwargs)\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/change_tensor_type/RaggedToTensor/boolean_mask_1/GatherV2:0\", shape=(None,), dtype=int32), values=Tensor(\"gradient_tape/model/change_tensor_type/RaggedToTensor/boolean_mask/GatherV2:0\", shape=(None, 7), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/change_tensor_type/RaggedToTensor/Shape:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/gcn_2/pooling_weighted_local_edges_2/Reshape_1:0\", shape=(None,), dtype=int32), values=Tensor(\"gradient_tape/model/gcn_2/pooling_weighted_local_edges_2/Reshape:0\", shape=(None, 124), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/gcn_2/pooling_weighted_local_edges_2/Cast:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/gcn_2/gather_nodes_outgoing_2/Reshape_1:0\", shape=(None,), dtype=int64), values=Tensor(\"gradient_tape/model/gcn_2/gather_nodes_outgoing_2/Reshape:0\", shape=(None, 124), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/gcn_2/gather_nodes_outgoing_2/Cast:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/gcn_1/pooling_weighted_local_edges_1/Reshape_1:0\", shape=(None,), dtype=int32), values=Tensor(\"gradient_tape/model/gcn_1/pooling_weighted_local_edges_1/Reshape:0\", shape=(None, 124), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/gcn_1/pooling_weighted_local_edges_1/Cast:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/gcn_1/gather_nodes_outgoing_1/Reshape_1:0\", shape=(None,), dtype=int64), values=Tensor(\"gradient_tape/model/gcn_1/gather_nodes_outgoing_1/Reshape:0\", shape=(None, 124), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/gcn_1/gather_nodes_outgoing_1/Cast:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/gcn/pooling_weighted_local_edges/Reshape_1:0\", shape=(None,), dtype=int32), values=Tensor(\"gradient_tape/model/gcn/pooling_weighted_local_edges/Reshape:0\", shape=(None, 124), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/gcn/pooling_weighted_local_edges/Cast:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n", - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\tensorflow\\python\\framework\\indexed_slices.py:444: UserWarning: Converting sparse IndexedSlices(IndexedSlices(indices=Tensor(\"gradient_tape/model/gcn/gather_nodes_outgoing/Reshape_1:0\", shape=(None,), dtype=int64), values=Tensor(\"gradient_tape/model/gcn/gather_nodes_outgoing/Reshape:0\", shape=(None, 124), dtype=float32), dense_shape=Tensor(\"gradient_tape/model/gcn/gather_nodes_outgoing/Cast:0\", shape=(2,), dtype=int32))) to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", - " warnings.warn(\n" - ] + "data": { + "text/html": [ + "
Model: \"GCN\"\n",
+       "
\n" + ], + "text/plain": [ + "\u001b[1mModel: \"GCN\"\u001b[0m\n" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/1 - 2s - loss: 1.7499 - categorical_accuracy: 0.2368 - lr: 0.0010 - 2s/epoch - 2s/step\n", - "Epoch 2/10\n", - "1/1 - 0s - loss: 1.7431 - categorical_accuracy: 0.3032 - lr: 0.0010 - 32ms/epoch - 32ms/step\n", - "Epoch 3/10\n", - "1/1 - 0s - loss: 1.7344 - categorical_accuracy: 0.3028 - lr: 0.0010 - 26ms/epoch - 26ms/step\n", - "Epoch 4/10\n", - "1/1 - 0s - loss: 1.7224 - categorical_accuracy: 0.3028 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 5/10\n", - "1/1 - 0s - loss: 1.7070 - categorical_accuracy: 0.3028 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 6/10\n", - "1/1 - 0s - loss: 1.6877 - categorical_accuracy: 0.3028 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 7/10\n", - "1/1 - 0s - loss: 1.6646 - categorical_accuracy: 0.3028 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 8/10\n", - "1/1 - 0s - loss: 1.6373 - categorical_accuracy: 0.3028 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 9/10\n", - "1/1 - 0s - loss: 1.6063 - categorical_accuracy: 0.3028 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 10/10\n", - "1/1 - 0s - loss: 1.5722 - categorical_accuracy: 0.3028 - lr: 0.0010 - 22ms/epoch - 22ms/step\n", - "1/1 [==============================] - 0s 466ms/step - loss: 0.1714 - categorical_accuracy: 0.2952\n", - "Epoch 11/20\n", - "1/1 - 0s - loss: 1.5349 - categorical_accuracy: 0.3028 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 12/20\n", - "1/1 - 0s - loss: 1.4929 - categorical_accuracy: 0.3028 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 13/20\n", - "1/1 - 0s - loss: 1.4469 - categorical_accuracy: 0.3041 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 14/20\n", - "1/1 - 0s - loss: 1.3987 - categorical_accuracy: 0.3410 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 15/20\n", - "1/1 - 0s - loss: 1.3499 - categorical_accuracy: 0.3849 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 16/20\n", - "1/1 - 0s - loss: 1.3010 - categorical_accuracy: 0.4116 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 17/20\n", - "1/1 - 0s - loss: 1.2534 - categorical_accuracy: 0.4407 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 18/20\n", - "1/1 - 0s - loss: 1.2086 - categorical_accuracy: 0.4694 - lr: 0.0010 - 27ms/epoch - 27ms/step\n", - "Epoch 19/20\n", - "1/1 - 0s - loss: 1.1671 - categorical_accuracy: 0.4797 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 20/20\n", - "1/1 - 0s - loss: 1.1275 - categorical_accuracy: 0.4846 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "1/1 [==============================] - 0s 27ms/step - loss: 0.1271 - categorical_accuracy: 0.4686\n", - "Epoch 21/30\n", - "1/1 - 0s - loss: 1.0871 - categorical_accuracy: 0.4928 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 22/30\n", - "1/1 - 0s - loss: 1.0448 - categorical_accuracy: 0.5150 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 23/30\n", - "1/1 - 0s - loss: 1.0004 - categorical_accuracy: 0.5650 - lr: 0.0010 - 22ms/epoch - 22ms/step\n", - "Epoch 24/30\n", - "1/1 - 0s - loss: 0.9543 - categorical_accuracy: 0.6085 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 25/30\n", - "1/1 - 0s - loss: 0.9067 - categorical_accuracy: 0.6512 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 26/30\n", - "1/1 - 0s - loss: 0.8584 - categorical_accuracy: 0.7124 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 27/30\n", - "1/1 - 0s - loss: 0.8097 - categorical_accuracy: 0.7394 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 28/30\n", - "1/1 - 0s - loss: 0.7604 - categorical_accuracy: 0.7714 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 29/30\n", - "1/1 - 0s - loss: 0.7094 - categorical_accuracy: 0.7850 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 30/30\n", - "1/1 - 0s - loss: 0.6531 - categorical_accuracy: 0.7969 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "1/1 [==============================] - 0s 29ms/step - loss: 0.0793 - categorical_accuracy: 0.8081\n", - "Epoch 31/40\n", - "1/1 - 0s - loss: 0.5970 - categorical_accuracy: 0.8338 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 32/40\n", - "1/1 - 0s - loss: 0.5481 - categorical_accuracy: 0.8621 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 33/40\n", - "1/1 - 0s - loss: 0.5113 - categorical_accuracy: 0.8461 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 34/40\n", - "1/1 - 0s - loss: 0.4809 - categorical_accuracy: 0.8363 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 35/40\n", - "1/1 - 0s - loss: 0.4544 - categorical_accuracy: 0.8396 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 36/40\n", - "1/1 - 0s - loss: 0.4266 - categorical_accuracy: 0.8515 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 37/40\n", - "1/1 - 0s - loss: 0.3987 - categorical_accuracy: 0.8839 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 38/40\n", - "1/1 - 0s - loss: 0.3832 - categorical_accuracy: 0.8945 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 39/40\n", - "1/1 - 0s - loss: 0.3640 - categorical_accuracy: 0.9019 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 40/40\n", - "1/1 - 0s - loss: 0.3439 - categorical_accuracy: 0.8978 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "1/1 [==============================] - 0s 26ms/step - loss: 0.0580 - categorical_accuracy: 0.8672\n", - "Epoch 41/50\n", - "1/1 - 0s - loss: 0.3288 - categorical_accuracy: 0.9007 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 42/50\n", - "1/1 - 0s - loss: 0.3211 - categorical_accuracy: 0.8958 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 43/50\n", - "1/1 - 0s - loss: 0.3227 - categorical_accuracy: 0.8991 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 44/50\n", - "1/1 - 0s - loss: 0.2959 - categorical_accuracy: 0.9052 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 45/50\n", - "1/1 - 0s - loss: 0.2838 - categorical_accuracy: 0.9097 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 46/50\n", - "1/1 - 0s - loss: 0.2830 - categorical_accuracy: 0.9126 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 47/50\n", - "1/1 - 0s - loss: 0.2589 - categorical_accuracy: 0.9151 - lr: 0.0010 - 26ms/epoch - 26ms/step\n", - "Epoch 48/50\n", - "1/1 - 0s - loss: 0.2670 - categorical_accuracy: 0.9114 - lr: 0.0010 - 26ms/epoch - 26ms/step\n", - "Epoch 49/50\n", - "1/1 - 0s - loss: 0.2408 - categorical_accuracy: 0.9224 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 50/50\n", - "1/1 - 0s - loss: 0.2497 - categorical_accuracy: 0.9175 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "1/1 [==============================] - 0s 26ms/step - loss: 0.0543 - categorical_accuracy: 0.8782\n", - "Epoch 51/60\n", - "1/1 - 0s - loss: 0.2291 - categorical_accuracy: 0.9216 - lr: 0.0010 - 36ms/epoch - 36ms/step\n", - "Epoch 52/60\n", - "1/1 - 0s - loss: 0.2361 - categorical_accuracy: 0.9188 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 53/60\n", - "1/1 - 0s - loss: 0.2178 - categorical_accuracy: 0.9261 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 54/60\n", - "1/1 - 0s - loss: 0.2229 - categorical_accuracy: 0.9233 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 55/60\n", - "1/1 - 0s - loss: 0.2079 - categorical_accuracy: 0.9311 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 56/60\n", - "1/1 - 0s - loss: 0.2121 - categorical_accuracy: 0.9233 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 57/60\n", - "1/1 - 0s - loss: 0.2000 - categorical_accuracy: 0.9307 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 58/60\n", - "1/1 - 0s - loss: 0.2029 - categorical_accuracy: 0.9265 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 59/60\n", - "1/1 - 0s - loss: 0.1914 - categorical_accuracy: 0.9335 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 60/60\n", - "1/1 - 0s - loss: 0.1929 - categorical_accuracy: 0.9319 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "1/1 [==============================] - 0s 34ms/step - loss: 0.0555 - categorical_accuracy: 0.8745\n", - "Epoch 61/70\n", - "1/1 - 0s - loss: 0.1835 - categorical_accuracy: 0.9380 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 62/70\n", - "1/1 - 0s - loss: 0.1848 - categorical_accuracy: 0.9323 - lr: 0.0010 - 43ms/epoch - 43ms/step\n", - "Epoch 63/70\n", - "1/1 - 0s - loss: 0.1768 - categorical_accuracy: 0.9389 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 64/70\n", - "1/1 - 0s - loss: 0.1771 - categorical_accuracy: 0.9360 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 65/70\n", - "1/1 - 0s - loss: 0.1701 - categorical_accuracy: 0.9397 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 66/70\n", - "1/1 - 0s - loss: 0.1703 - categorical_accuracy: 0.9348 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 67/70\n", - "1/1 - 0s - loss: 0.1643 - categorical_accuracy: 0.9421 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 68/70\n", - "1/1 - 0s - loss: 0.1642 - categorical_accuracy: 0.9389 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 69/70\n", - "1/1 - 0s - loss: 0.1588 - categorical_accuracy: 0.9434 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 70/70\n", - "1/1 - 0s - loss: 0.1579 - categorical_accuracy: 0.9430 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "1/1 [==============================] - 0s 34ms/step - loss: 0.0594 - categorical_accuracy: 0.8745\n", - "Epoch 71/80\n", - "1/1 - 0s - loss: 0.1531 - categorical_accuracy: 0.9450 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 72/80\n", - "1/1 - 0s - loss: 0.1522 - categorical_accuracy: 0.9446 - lr: 0.0010 - 37ms/epoch - 37ms/step\n" - ] + "data": { + "text/html": [ + "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
+       "┃ Layer (type)                  ┃ Output Shape              ┃     Param # ┃ Connected to                   ┃\n",
+       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
+       "│ node_attributes (InputLayer)  │ (None, None, 1432)        │           0 │ -                              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ total_nodes (InputLayer)      │ (None)                    │           0 │ -                              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ cast_batched_attributes_to_d… │ [(None, 1432), (None),    │           0 │ node_attributes[0][0],         │\n",
+       "│ (CastBatchedAttributesToDisj… │ (None), (None)]           │             │ total_nodes[0][0]              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ edge_attributes (InputLayer)  │ (None, None, 1)           │           0 │ -                              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ total_edges (InputLayer)      │ (None)                    │           0 │ -                              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ edge_indices (InputLayer)     │ (None, None, 2)           │           0 │ -                              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ dense (Dense)                 │ (None, 124)               │     177,692 │ cast_batched_attributes_to_di… │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ cast_batched_attributes_to_d… │ [(None, 1), (None),       │           0 │ edge_attributes[0][0],         │\n",
+       "│ (CastBatchedAttributesToDisj… │ (None), (None)]           │             │ total_edges[0][0]              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ cast_batched_indices_to_disj… │ [(None, 1432), (2, None), │           0 │ node_attributes[0][0],         │\n",
+       "│ (CastBatchedIndicesToDisjoin… │ (None), (None), (None),   │             │ edge_indices[0][0],            │\n",
+       "│                               │ (None), (None), (None)]   │             │ total_nodes[0][0],             │\n",
+       "│                               │                           │             │ total_edges[0][0]              │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ gcn (GCN)                     │ (None, 124)               │      15,500 │ dense[0][0],                   │\n",
+       "│                               │                           │             │ cast_batched_attributes_to_di… │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ gcn_1 (GCN)                   │ (None, 124)               │      15,500 │ gcn[0][0],                     │\n",
+       "│                               │                           │             │ cast_batched_attributes_to_di… │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ gcn_2 (GCN)                   │ (None, 124)               │      15,500 │ gcn_1[0][0],                   │\n",
+       "│                               │                           │             │ cast_batched_attributes_to_di… │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ mlp (MLP)                     │ (None, 7)                 │       9,152 │ gcn_2[0][0],                   │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ cast_disjoint_to_batched_att… │ (None, None, 7)           │           0 │ mlp[0][0],                     │\n",
+       "│ (CastDisjointToBatchedAttrib… │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "│                               │                           │             │ cast_batched_indices_to_disjo… │\n",
+       "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n",
+       "│ cast_disjoint_to_batched_gra… │ (None, None, 7)           │           0 │ cast_disjoint_to_batched_attr… │\n",
+       "│ (CastDisjointToBatchedGraphS… │                           │             │                                │\n",
+       "└───────────────────────────────┴───────────────────────────┴─────────────┴────────────────────────────────┘\n",
+       "
\n" + ], + "text/plain": [ + "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n", + "┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mConnected to \u001b[0m\u001b[1m \u001b[0m┃\n", + "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n", + "│ node_attributes (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1432\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ total_nodes (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ cast_batched_attributes_to_d… │ [(\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1432\u001b[0m), (\u001b[38;5;45mNone\u001b[0m), │ \u001b[38;5;34m0\u001b[0m │ node_attributes[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ (\u001b[38;5;33mCastBatchedAttributesToDisj…\u001b[0m │ (\u001b[38;5;45mNone\u001b[0m), (\u001b[38;5;45mNone\u001b[0m)] │ │ total_nodes[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ edge_attributes (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ total_edges (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ edge_indices (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m124\u001b[0m) │ \u001b[38;5;34m177,692\u001b[0m │ cast_batched_attributes_to_di… │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ cast_batched_attributes_to_d… │ [(\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m), (\u001b[38;5;45mNone\u001b[0m), │ \u001b[38;5;34m0\u001b[0m │ edge_attributes[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ (\u001b[38;5;33mCastBatchedAttributesToDisj…\u001b[0m │ (\u001b[38;5;45mNone\u001b[0m), (\u001b[38;5;45mNone\u001b[0m)] │ │ total_edges[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ cast_batched_indices_to_disj… │ [(\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1432\u001b[0m), (\u001b[38;5;34m2\u001b[0m, \u001b[38;5;45mNone\u001b[0m), │ \u001b[38;5;34m0\u001b[0m │ node_attributes[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ (\u001b[38;5;33mCastBatchedIndicesToDisjoin…\u001b[0m │ (\u001b[38;5;45mNone\u001b[0m), (\u001b[38;5;45mNone\u001b[0m), (\u001b[38;5;45mNone\u001b[0m), │ │ edge_indices[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ │ (\u001b[38;5;45mNone\u001b[0m), (\u001b[38;5;45mNone\u001b[0m), (\u001b[38;5;45mNone\u001b[0m)] │ │ total_nodes[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ │ │ │ total_edges[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ gcn (\u001b[38;5;33mGCN\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m124\u001b[0m) │ \u001b[38;5;34m15,500\u001b[0m │ dense[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ │ │ │ cast_batched_attributes_to_di… │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ gcn_1 (\u001b[38;5;33mGCN\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m124\u001b[0m) │ \u001b[38;5;34m15,500\u001b[0m │ gcn[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ │ │ │ cast_batched_attributes_to_di… │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ gcn_2 (\u001b[38;5;33mGCN\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m124\u001b[0m) │ \u001b[38;5;34m15,500\u001b[0m │ gcn_1[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ │ │ │ cast_batched_attributes_to_di… │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ mlp (\u001b[38;5;33mMLP\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m) │ \u001b[38;5;34m9,152\u001b[0m │ gcn_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ cast_disjoint_to_batched_att… │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ mlp[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", + "│ (\u001b[38;5;33mCastDisjointToBatchedAttrib…\u001b[0m │ │ │ cast_batched_indices_to_disjo… │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "│ │ │ │ cast_batched_indices_to_disjo… │\n", + "├───────────────────────────────┼───────────────────────────┼─────────────┼────────────────────────────────┤\n", + "│ cast_disjoint_to_batched_gra… │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ cast_disjoint_to_batched_attr… │\n", + "│ (\u001b[38;5;33mCastDisjointToBatchedGraphS…\u001b[0m │ │ │ │\n", + "└───────────────────────────────┴───────────────────────────┴─────────────┴────────────────────────────────┘\n" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 73/80\n", - "1/1 - 0s - loss: 0.1478 - categorical_accuracy: 0.9483 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 74/80\n", - "1/1 - 0s - loss: 0.1468 - categorical_accuracy: 0.9479 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 75/80\n", - "1/1 - 0s - loss: 0.1428 - categorical_accuracy: 0.9503 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 76/80\n", - "1/1 - 0s - loss: 0.1415 - categorical_accuracy: 0.9454 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 77/80\n", - "1/1 - 0s - loss: 0.1381 - categorical_accuracy: 0.9508 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 78/80\n", - "1/1 - 0s - loss: 0.1365 - categorical_accuracy: 0.9528 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 79/80\n", - "1/1 - 0s - loss: 0.1337 - categorical_accuracy: 0.9495 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 80/80\n", - "1/1 - 0s - loss: 0.1316 - categorical_accuracy: 0.9516 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "1/1 [==============================] - 0s 34ms/step - loss: 0.0640 - categorical_accuracy: 0.8598\n", - "Epoch 81/90\n", - "1/1 - 0s - loss: 0.1297 - categorical_accuracy: 0.9553 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 82/90\n", - "1/1 - 0s - loss: 0.1271 - categorical_accuracy: 0.9557 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 83/90\n", - "1/1 - 0s - loss: 0.1257 - categorical_accuracy: 0.9528 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 84/90\n", - "1/1 - 0s - loss: 0.1231 - categorical_accuracy: 0.9586 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 85/90\n", - "1/1 - 0s - loss: 0.1218 - categorical_accuracy: 0.9590 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 86/90\n", - "1/1 - 0s - loss: 0.1203 - categorical_accuracy: 0.9561 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 87/90\n", - "1/1 - 0s - loss: 0.1192 - categorical_accuracy: 0.9594 - lr: 0.0010 - 46ms/epoch - 46ms/step\n", - "Epoch 88/90\n", - "1/1 - 0s - loss: 0.1205 - categorical_accuracy: 0.9553 - lr: 0.0010 - 44ms/epoch - 44ms/step\n", - "Epoch 89/90\n", - "1/1 - 0s - loss: 0.1225 - categorical_accuracy: 0.9569 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 90/90\n", - "1/1 - 0s - loss: 0.1267 - categorical_accuracy: 0.9516 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "1/1 [==============================] - 0s 34ms/step - loss: 0.0698 - categorical_accuracy: 0.8487\n", - "Epoch 91/100\n", - "1/1 - 0s - loss: 0.1169 - categorical_accuracy: 0.9581 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 92/100\n", - "1/1 - 0s - loss: 0.1092 - categorical_accuracy: 0.9639 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 93/100\n", - "1/1 - 0s - loss: 0.1109 - categorical_accuracy: 0.9586 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 94/100\n", - "1/1 - 0s - loss: 0.1113 - categorical_accuracy: 0.9610 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 95/100\n", - "1/1 - 0s - loss: 0.1066 - categorical_accuracy: 0.9639 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 96/100\n", - "1/1 - 0s - loss: 0.1033 - categorical_accuracy: 0.9622 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 97/100\n", - "1/1 - 0s - loss: 0.1054 - categorical_accuracy: 0.9622 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 98/100\n", - "1/1 - 0s - loss: 0.1044 - categorical_accuracy: 0.9618 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 99/100\n", - "1/1 - 0s - loss: 0.0993 - categorical_accuracy: 0.9639 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 100/100\n", - "1/1 - 0s - loss: 0.0992 - categorical_accuracy: 0.9647 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "1/1 [==============================] - 0s 38ms/step - loss: 0.0704 - categorical_accuracy: 0.8708\n", - "Epoch 101/110\n", - "1/1 - 0s - loss: 0.1005 - categorical_accuracy: 0.9639 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 102/110\n", - "1/1 - 0s - loss: 0.0969 - categorical_accuracy: 0.9655 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 103/110\n", - "1/1 - 0s - loss: 0.0940 - categorical_accuracy: 0.9664 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 104/110\n", - "1/1 - 0s - loss: 0.0944 - categorical_accuracy: 0.9655 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 105/110\n", - "1/1 - 0s - loss: 0.0939 - categorical_accuracy: 0.9659 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 106/110\n", - "1/1 - 0s - loss: 0.0914 - categorical_accuracy: 0.9651 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 107/110\n", - "1/1 - 0s - loss: 0.0893 - categorical_accuracy: 0.9672 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 108/110\n", - "1/1 - 0s - loss: 0.0893 - categorical_accuracy: 0.9680 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 109/110\n", - "1/1 - 0s - loss: 0.0892 - categorical_accuracy: 0.9668 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 110/110\n", - "1/1 - 0s - loss: 0.0871 - categorical_accuracy: 0.9688 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "1/1 [==============================] - 0s 35ms/step - loss: 0.0739 - categorical_accuracy: 0.8598\n", - "Epoch 111/120\n", - "1/1 - 0s - loss: 0.0849 - categorical_accuracy: 0.9696 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 112/120\n", - "1/1 - 0s - loss: 0.0841 - categorical_accuracy: 0.9684 - lr: 0.0010 - 43ms/epoch - 43ms/step\n", - "Epoch 113/120\n", - "1/1 - 0s - loss: 0.0839 - categorical_accuracy: 0.9705 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 114/120\n", - "1/1 - 0s - loss: 0.0833 - categorical_accuracy: 0.9705 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 115/120\n", - "1/1 - 0s - loss: 0.0816 - categorical_accuracy: 0.9713 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 116/120\n", - "1/1 - 0s - loss: 0.0797 - categorical_accuracy: 0.9713 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 117/120\n", - "1/1 - 0s - loss: 0.0786 - categorical_accuracy: 0.9721 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 118/120\n", - "1/1 - 0s - loss: 0.0781 - categorical_accuracy: 0.9725 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 119/120\n", - "1/1 - 0s - loss: 0.0778 - categorical_accuracy: 0.9721 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 120/120\n", - "1/1 - 0s - loss: 0.0771 - categorical_accuracy: 0.9737 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "1/1 [==============================] - 0s 39ms/step - loss: 0.0767 - categorical_accuracy: 0.8635\n", - "Epoch 121/130\n", - "1/1 - 0s - loss: 0.0763 - categorical_accuracy: 0.9733 - lr: 0.0010 - 45ms/epoch - 45ms/step\n", - "Epoch 122/130\n", - "1/1 - 0s - loss: 0.0748 - categorical_accuracy: 0.9746 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 123/130\n", - "1/1 - 0s - loss: 0.0734 - categorical_accuracy: 0.9733 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 124/130\n", - "1/1 - 0s - loss: 0.0720 - categorical_accuracy: 0.9754 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 125/130\n", - "1/1 - 0s - loss: 0.0709 - categorical_accuracy: 0.9750 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 126/130\n", - "1/1 - 0s - loss: 0.0700 - categorical_accuracy: 0.9754 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 127/130\n", - "1/1 - 0s - loss: 0.0692 - categorical_accuracy: 0.9762 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 128/130\n", - "1/1 - 0s - loss: 0.0686 - categorical_accuracy: 0.9750 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 129/130\n", - "1/1 - 0s - loss: 0.0685 - categorical_accuracy: 0.9758 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 130/130\n", - "1/1 - 0s - loss: 0.0693 - categorical_accuracy: 0.9737 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "1/1 [==============================] - 0s 34ms/step - loss: 0.0829 - categorical_accuracy: 0.8450\n", - "Epoch 131/140\n", - "1/1 - 0s - loss: 0.0718 - categorical_accuracy: 0.9746 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 132/140\n", - "1/1 - 0s - loss: 0.0738 - categorical_accuracy: 0.9729 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 133/140\n", - "1/1 - 0s - loss: 0.0740 - categorical_accuracy: 0.9725 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 134/140\n", - "1/1 - 0s - loss: 0.0687 - categorical_accuracy: 0.9754 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 135/140\n", - "1/1 - 0s - loss: 0.0632 - categorical_accuracy: 0.9770 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 136/140\n", - "1/1 - 0s - loss: 0.0624 - categorical_accuracy: 0.9774 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 137/140\n", - "1/1 - 0s - loss: 0.0646 - categorical_accuracy: 0.9750 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 138/140\n", - "1/1 - 0s - loss: 0.0653 - categorical_accuracy: 0.9774 - lr: 0.0010 - 44ms/epoch - 44ms/step\n", - "Epoch 139/140\n", - "1/1 - 0s - loss: 0.0619 - categorical_accuracy: 0.9754 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 140/140\n", - "1/1 - 0s - loss: 0.0591 - categorical_accuracy: 0.9787 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "1/1 [==============================] - 0s 38ms/step - loss: 0.0857 - categorical_accuracy: 0.8561\n", - "Epoch 141/150\n", - "1/1 - 0s - loss: 0.0591 - categorical_accuracy: 0.9807 - lr: 0.0010 - 43ms/epoch - 43ms/step\n", - "Epoch 142/150\n", - "1/1 - 0s - loss: 0.0603 - categorical_accuracy: 0.9750 - lr: 0.0010 - 44ms/epoch - 44ms/step\n", - "Epoch 143/150\n", - "1/1 - 0s - loss: 0.0605 - categorical_accuracy: 0.9807 - lr: 0.0010 - 41ms/epoch - 41ms/step\n" - ] + "data": { + "text/html": [ + "
 Total params: 233,344 (911.50 KB)\n",
+       "
\n" + ], + "text/plain": [ + "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m233,344\u001b[0m (911.50 KB)\n" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 144/150\n", - "1/1 - 0s - loss: 0.0582 - categorical_accuracy: 0.9783 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 145/150\n", - "1/1 - 0s - loss: 0.0560 - categorical_accuracy: 0.9815 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 146/150\n", - "1/1 - 0s - loss: 0.0549 - categorical_accuracy: 0.9819 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 147/150\n", - "1/1 - 0s - loss: 0.0551 - categorical_accuracy: 0.9778 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 148/150\n", - "1/1 - 0s - loss: 0.0558 - categorical_accuracy: 0.9811 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 149/150\n", - "1/1 - 0s - loss: 0.0551 - categorical_accuracy: 0.9783 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 150/150\n", - "1/1 - 0s - loss: 0.0540 - categorical_accuracy: 0.9815 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "1/1 [==============================] - 0s 37ms/step - loss: 0.0891 - categorical_accuracy: 0.8598\n", - "Epoch 151/160\n", - "1/1 - 0s - loss: 0.0519 - categorical_accuracy: 0.9803 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 152/160\n", - "1/1 - 0s - loss: 0.0509 - categorical_accuracy: 0.9828 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 153/160\n", - "1/1 - 0s - loss: 0.0507 - categorical_accuracy: 0.9840 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 154/160\n", - "1/1 - 0s - loss: 0.0508 - categorical_accuracy: 0.9803 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 155/160\n", - "1/1 - 0s - loss: 0.0511 - categorical_accuracy: 0.9828 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 156/160\n", - "1/1 - 0s - loss: 0.0499 - categorical_accuracy: 0.9799 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 157/160\n", - "1/1 - 0s - loss: 0.0488 - categorical_accuracy: 0.9848 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 158/160\n", - "1/1 - 0s - loss: 0.0472 - categorical_accuracy: 0.9824 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 159/160\n", - "1/1 - 0s - loss: 0.0463 - categorical_accuracy: 0.9840 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 160/160\n", - "1/1 - 0s - loss: 0.0459 - categorical_accuracy: 0.9848 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "1/1 [==============================] - 0s 36ms/step - loss: 0.0933 - categorical_accuracy: 0.8635\n", - "Epoch 161/170\n", - "1/1 - 0s - loss: 0.0457 - categorical_accuracy: 0.9836 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 162/170\n", - "1/1 - 0s - loss: 0.0457 - categorical_accuracy: 0.9865 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 163/170\n", - "1/1 - 0s - loss: 0.0453 - categorical_accuracy: 0.9848 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 164/170\n", - "1/1 - 0s - loss: 0.0452 - categorical_accuracy: 0.9865 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 165/170\n", - "1/1 - 0s - loss: 0.0444 - categorical_accuracy: 0.9848 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 166/170\n", - "1/1 - 0s - loss: 0.0439 - categorical_accuracy: 0.9869 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 167/170\n", - "1/1 - 0s - loss: 0.0433 - categorical_accuracy: 0.9848 - lr: 0.0010 - 44ms/epoch - 44ms/step\n", - "Epoch 168/170\n", - "1/1 - 0s - loss: 0.0432 - categorical_accuracy: 0.9860 - lr: 0.0010 - 45ms/epoch - 45ms/step\n", - "Epoch 169/170\n", - "1/1 - 0s - loss: 0.0439 - categorical_accuracy: 0.9856 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 170/170\n", - "1/1 - 0s - loss: 0.0455 - categorical_accuracy: 0.9824 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "1/1 [==============================] - 0s 35ms/step - loss: 0.0975 - categorical_accuracy: 0.8524\n", - "Epoch 171/180\n", - "1/1 - 0s - loss: 0.0488 - categorical_accuracy: 0.9844 - lr: 0.0010 - 36ms/epoch - 36ms/step\n", - "Epoch 172/180\n", - "1/1 - 0s - loss: 0.0492 - categorical_accuracy: 0.9791 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 173/180\n", - "1/1 - 0s - loss: 0.0480 - categorical_accuracy: 0.9819 - lr: 0.0010 - 44ms/epoch - 44ms/step\n", - "Epoch 174/180\n", - "1/1 - 0s - loss: 0.0413 - categorical_accuracy: 0.9848 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 175/180\n", - "1/1 - 0s - loss: 0.0395 - categorical_accuracy: 0.9877 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 176/180\n", - "1/1 - 0s - loss: 0.0421 - categorical_accuracy: 0.9844 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 177/180\n", - "1/1 - 0s - loss: 0.0414 - categorical_accuracy: 0.9848 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 178/180\n", - "1/1 - 0s - loss: 0.0379 - categorical_accuracy: 0.9869 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 179/180\n", - "1/1 - 0s - loss: 0.0370 - categorical_accuracy: 0.9873 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 180/180\n", - "1/1 - 0s - loss: 0.0392 - categorical_accuracy: 0.9852 - lr: 0.0010 - 45ms/epoch - 45ms/step\n", - "1/1 [==============================] - 0s 39ms/step - loss: 0.1019 - categorical_accuracy: 0.8598\n", - "Epoch 181/190\n", - "1/1 - 0s - loss: 0.0388 - categorical_accuracy: 0.9869 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 182/190\n", - "1/1 - 0s - loss: 0.0360 - categorical_accuracy: 0.9877 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 183/190\n", - "1/1 - 0s - loss: 0.0362 - categorical_accuracy: 0.9881 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 184/190\n", - "1/1 - 0s - loss: 0.0378 - categorical_accuracy: 0.9885 - lr: 0.0010 - 45ms/epoch - 45ms/step\n", - "Epoch 185/190\n", - "1/1 - 0s - loss: 0.0358 - categorical_accuracy: 0.9881 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 186/190\n", - "1/1 - 0s - loss: 0.0340 - categorical_accuracy: 0.9885 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 187/190\n", - "1/1 - 0s - loss: 0.0344 - categorical_accuracy: 0.9885 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 188/190\n", - "1/1 - 0s - loss: 0.0346 - categorical_accuracy: 0.9881 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 189/190\n", - "1/1 - 0s - loss: 0.0333 - categorical_accuracy: 0.9885 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 190/190\n", - "1/1 - 0s - loss: 0.0325 - categorical_accuracy: 0.9885 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "1/1 [==============================] - 0s 36ms/step - loss: 0.1085 - categorical_accuracy: 0.8450\n", - "Epoch 191/200\n", - "1/1 - 0s - loss: 0.0331 - categorical_accuracy: 0.9897 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 192/200\n", - "1/1 - 0s - loss: 0.0331 - categorical_accuracy: 0.9889 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 193/200\n", - "1/1 - 0s - loss: 0.0319 - categorical_accuracy: 0.9893 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 194/200\n", - "1/1 - 0s - loss: 0.0316 - categorical_accuracy: 0.9897 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 195/200\n", - "1/1 - 0s - loss: 0.0322 - categorical_accuracy: 0.9906 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 196/200\n", - "1/1 - 0s - loss: 0.0317 - categorical_accuracy: 0.9897 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 197/200\n", - "1/1 - 0s - loss: 0.0312 - categorical_accuracy: 0.9906 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 198/200\n", - "1/1 - 0s - loss: 0.0308 - categorical_accuracy: 0.9889 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 199/200\n", - "1/1 - 0s - loss: 0.0313 - categorical_accuracy: 0.9897 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 200/200\n", - "1/1 - 0s - loss: 0.0315 - categorical_accuracy: 0.9897 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "1/1 [==============================] - 0s 37ms/step - loss: 0.1132 - categorical_accuracy: 0.8487\n", - "Epoch 201/210\n", - "1/1 - 0s - loss: 0.0310 - categorical_accuracy: 0.9906 - lr: 0.0010 - 40ms/epoch - 40ms/step\n", - "Epoch 202/210\n", - "1/1 - 0s - loss: 0.0301 - categorical_accuracy: 0.9893 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 203/210\n", - "1/1 - 0s - loss: 0.0299 - categorical_accuracy: 0.9906 - lr: 0.0010 - 42ms/epoch - 42ms/step\n", - "Epoch 204/210\n", - "1/1 - 0s - loss: 0.0292 - categorical_accuracy: 0.9910 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 205/210\n", - "1/1 - 0s - loss: 0.0285 - categorical_accuracy: 0.9906 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 206/210\n", - "1/1 - 0s - loss: 0.0276 - categorical_accuracy: 0.9902 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 207/210\n", - "1/1 - 0s - loss: 0.0274 - categorical_accuracy: 0.9910 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 208/210\n", - "1/1 - 0s - loss: 0.0278 - categorical_accuracy: 0.9910 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 209/210\n", - "1/1 - 0s - loss: 0.0277 - categorical_accuracy: 0.9914 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 210/210\n", - "1/1 - 0s - loss: 0.0277 - categorical_accuracy: 0.9910 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "1/1 [==============================] - 0s 35ms/step - loss: 0.1143 - categorical_accuracy: 0.8487\n", - "Epoch 211/220\n", - "1/1 - 0s - loss: 0.0273 - categorical_accuracy: 0.9910 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 212/220\n", - "1/1 - 0s - loss: 0.0274 - categorical_accuracy: 0.9934 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 213/220\n", - "1/1 - 0s - loss: 0.0271 - categorical_accuracy: 0.9914 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 214/220\n", - "1/1 - 0s - loss: 0.0266 - categorical_accuracy: 0.9930 - lr: 0.0010 - 39ms/epoch - 39ms/step\n" - ] + "data": { + "text/html": [ + "
 Trainable params: 233,344 (911.50 KB)\n",
+       "
\n" + ], + "text/plain": [ + "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m233,344\u001b[0m (911.50 KB)\n" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 215/220\n", - "1/1 - 0s - loss: 0.0257 - categorical_accuracy: 0.9922 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 216/220\n", - "1/1 - 0s - loss: 0.0251 - categorical_accuracy: 0.9922 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 217/220\n", - "1/1 - 0s - loss: 0.0249 - categorical_accuracy: 0.9918 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 218/220\n", - "1/1 - 0s - loss: 0.0248 - categorical_accuracy: 0.9930 - lr: 0.0010 - 36ms/epoch - 36ms/step\n", - "Epoch 219/220\n", - "1/1 - 0s - loss: 0.0246 - categorical_accuracy: 0.9926 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 220/220\n", - "1/1 - 0s - loss: 0.0245 - categorical_accuracy: 0.9918 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "1/1 [==============================] - 0s 34ms/step - loss: 0.1198 - categorical_accuracy: 0.8524\n", - "Epoch 221/230\n", - "1/1 - 0s - loss: 0.0249 - categorical_accuracy: 0.9934 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "Epoch 222/230\n", - "1/1 - 0s - loss: 0.0251 - categorical_accuracy: 0.9914 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 223/230\n", - "1/1 - 0s - loss: 0.0264 - categorical_accuracy: 0.9918 - lr: 0.0010 - 48ms/epoch - 48ms/step\n", - "Epoch 224/230\n", - "1/1 - 0s - loss: 0.0269 - categorical_accuracy: 0.9906 - lr: 0.0010 - 49ms/epoch - 49ms/step\n", - "Epoch 225/230\n", - "1/1 - 0s - loss: 0.0297 - categorical_accuracy: 0.9902 - lr: 0.0010 - 44ms/epoch - 44ms/step\n", - "Epoch 226/230\n", - "1/1 - 0s - loss: 0.0309 - categorical_accuracy: 0.9877 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 227/230\n", - "1/1 - 0s - loss: 0.0348 - categorical_accuracy: 0.9869 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 228/230\n", - "1/1 - 0s - loss: 0.0353 - categorical_accuracy: 0.9844 - lr: 0.0010 - 36ms/epoch - 36ms/step\n", - "Epoch 229/230\n", - "1/1 - 0s - loss: 0.0346 - categorical_accuracy: 0.9873 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 230/230\n", - "1/1 - 0s - loss: 0.0283 - categorical_accuracy: 0.9902 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "1/1 [==============================] - 0s 35ms/step - loss: 0.1227 - categorical_accuracy: 0.8487\n", - "Epoch 231/240\n", - "1/1 - 0s - loss: 0.0248 - categorical_accuracy: 0.9918 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 232/240\n", - "1/1 - 0s - loss: 0.0294 - categorical_accuracy: 0.9902 - lr: 0.0010 - 38ms/epoch - 38ms/step\n", - "Epoch 233/240\n", - "1/1 - 0s - loss: 0.0258 - categorical_accuracy: 0.9906 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 234/240\n", - "1/1 - 0s - loss: 0.0223 - categorical_accuracy: 0.9926 - lr: 0.0010 - 36ms/epoch - 36ms/step\n", - "Epoch 235/240\n", - "1/1 - 0s - loss: 0.0258 - categorical_accuracy: 0.9922 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 236/240\n", - "1/1 - 0s - loss: 0.0240 - categorical_accuracy: 0.9918 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 237/240\n", - "1/1 - 0s - loss: 0.0212 - categorical_accuracy: 0.9926 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 238/240\n", - "1/1 - 0s - loss: 0.0235 - categorical_accuracy: 0.9934 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 239/240\n", - "1/1 - 0s - loss: 0.0223 - categorical_accuracy: 0.9926 - lr: 0.0010 - 37ms/epoch - 37ms/step\n", - "Epoch 240/240\n", - "1/1 - 0s - loss: 0.0206 - categorical_accuracy: 0.9934 - lr: 0.0010 - 39ms/epoch - 39ms/step\n", - "1/1 [==============================] - 0s 67ms/step - loss: 0.1274 - categorical_accuracy: 0.8487\n", - "Epoch 241/250\n", - "1/1 - 0s - loss: 0.0220 - categorical_accuracy: 0.9938 - lr: 0.0010 - 49ms/epoch - 49ms/step\n", - "Epoch 242/250\n", - "1/1 - 0s - loss: 0.0209 - categorical_accuracy: 0.9930 - lr: 0.0010 - 46ms/epoch - 46ms/step\n", - "Epoch 243/250\n", - "1/1 - 0s - loss: 0.0201 - categorical_accuracy: 0.9938 - lr: 0.0010 - 47ms/epoch - 47ms/step\n", - "Epoch 244/250\n", - "1/1 - 0s - loss: 0.0208 - categorical_accuracy: 0.9951 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 245/250\n", - "1/1 - 0s - loss: 0.0199 - categorical_accuracy: 0.9938 - lr: 0.0010 - 41ms/epoch - 41ms/step\n", - "Epoch 246/250\n", - "1/1 - 0s - loss: 0.0196 - categorical_accuracy: 0.9943 - lr: 0.0010 - 45ms/epoch - 45ms/step\n", - "Epoch 247/250\n", - "1/1 - 0s - loss: 0.0200 - categorical_accuracy: 0.9955 - lr: 0.0010 - 46ms/epoch - 46ms/step\n", - "Epoch 248/250\n", - "1/1 - 0s - loss: 0.0192 - categorical_accuracy: 0.9947 - lr: 0.0010 - 50ms/epoch - 50ms/step\n", - "Epoch 249/250\n", - "1/1 - 0s - loss: 0.0190 - categorical_accuracy: 0.9955 - lr: 0.0010 - 49ms/epoch - 49ms/step\n", - "Epoch 250/250\n", - "1/1 - 0s - loss: 0.0193 - categorical_accuracy: 0.9947 - lr: 0.0010 - 30ms/epoch - 30ms/step\n", - "1/1 [==============================] - 0s 35ms/step - loss: 0.1292 - categorical_accuracy: 0.8561\n", - "Epoch 251/260\n", - "1/1 - 0s - loss: 0.0187 - categorical_accuracy: 0.9951 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 252/260\n", - "1/1 - 0s - loss: 0.0185 - categorical_accuracy: 0.9951 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 253/260\n", - "1/1 - 0s - loss: 0.0187 - categorical_accuracy: 0.9943 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 254/260\n", - "1/1 - 0s - loss: 0.0182 - categorical_accuracy: 0.9943 - lr: 0.0010 - 23ms/epoch - 23ms/step\n", - "Epoch 255/260\n", - "1/1 - 0s - loss: 0.0180 - categorical_accuracy: 0.9947 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 256/260\n", - "1/1 - 0s - loss: 0.0181 - categorical_accuracy: 0.9947 - lr: 0.0010 - 24ms/epoch - 24ms/step\n", - "Epoch 257/260\n", - "1/1 - 0s - loss: 0.0178 - categorical_accuracy: 0.9947 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 258/260\n", - "1/1 - 0s - loss: 0.0175 - categorical_accuracy: 0.9951 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "Epoch 259/260\n", - "1/1 - 0s - loss: 0.0174 - categorical_accuracy: 0.9947 - lr: 0.0010 - 27ms/epoch - 27ms/step\n", - "Epoch 260/260\n", - "1/1 - 0s - loss: 0.0174 - categorical_accuracy: 0.9951 - lr: 0.0010 - 25ms/epoch - 25ms/step\n", - "1/1 [==============================] - 0s 28ms/step - loss: 0.1321 - categorical_accuracy: 0.8487\n", - "Epoch 261/270\n", - "1/1 - 0s - loss: 0.0171 - categorical_accuracy: 0.9951 - lr: 0.0010 - 22ms/epoch - 22ms/step\n", - "Epoch 262/270\n", - "1/1 - 0s - loss: 0.0169 - categorical_accuracy: 0.9955 - lr: 9.7750e-04 - 21ms/epoch - 21ms/step\n", - "Epoch 263/270\n", - "1/1 - 0s - loss: 0.0169 - categorical_accuracy: 0.9951 - lr: 9.5500e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 264/270\n", - "1/1 - 0s - loss: 0.0168 - categorical_accuracy: 0.9951 - lr: 9.3250e-04 - 25ms/epoch - 25ms/step\n", - "Epoch 265/270\n", - "1/1 - 0s - loss: 0.0166 - categorical_accuracy: 0.9963 - lr: 9.1000e-04 - 24ms/epoch - 24ms/step\n", - "Epoch 266/270\n", - "1/1 - 0s - loss: 0.0165 - categorical_accuracy: 0.9959 - lr: 8.8750e-04 - 24ms/epoch - 24ms/step\n", - "Epoch 267/270\n", - "1/1 - 0s - loss: 0.0164 - categorical_accuracy: 0.9955 - lr: 8.6500e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 268/270\n", - "1/1 - 0s - loss: 0.0162 - categorical_accuracy: 0.9967 - lr: 8.4250e-04 - 26ms/epoch - 26ms/step\n", - "Epoch 269/270\n", - "1/1 - 0s - loss: 0.0162 - categorical_accuracy: 0.9959 - lr: 8.2000e-04 - 25ms/epoch - 25ms/step\n", - "Epoch 270/270\n", - "1/1 - 0s - loss: 0.0160 - categorical_accuracy: 0.9959 - lr: 7.9750e-04 - 23ms/epoch - 23ms/step\n", - "1/1 [==============================] - 0s 26ms/step - loss: 0.1348 - categorical_accuracy: 0.8487\n", - "Epoch 271/280\n", - "1/1 - 0s - loss: 0.0159 - categorical_accuracy: 0.9967 - lr: 7.7500e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 272/280\n", - "1/1 - 0s - loss: 0.0159 - categorical_accuracy: 0.9963 - lr: 7.5250e-04 - 21ms/epoch - 21ms/step\n", - "Epoch 273/280\n", - "1/1 - 0s - loss: 0.0157 - categorical_accuracy: 0.9963 - lr: 7.3000e-04 - 22ms/epoch - 22ms/step\n", - "Epoch 274/280\n", - "1/1 - 0s - loss: 0.0157 - categorical_accuracy: 0.9967 - lr: 7.0750e-04 - 22ms/epoch - 22ms/step\n", - "Epoch 275/280\n", - "1/1 - 0s - loss: 0.0156 - categorical_accuracy: 0.9963 - lr: 6.8500e-04 - 25ms/epoch - 25ms/step\n", - "Epoch 276/280\n", - "1/1 - 0s - loss: 0.0155 - categorical_accuracy: 0.9967 - lr: 6.6250e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 277/280\n", - "1/1 - 0s - loss: 0.0154 - categorical_accuracy: 0.9955 - lr: 6.4000e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 278/280\n", - "1/1 - 0s - loss: 0.0153 - categorical_accuracy: 0.9963 - lr: 6.1750e-04 - 25ms/epoch - 25ms/step\n", - "Epoch 279/280\n", - "1/1 - 0s - loss: 0.0153 - categorical_accuracy: 0.9963 - lr: 5.9500e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 280/280\n", - "1/1 - 0s - loss: 0.0152 - categorical_accuracy: 0.9963 - lr: 5.7250e-04 - 23ms/epoch - 23ms/step\n", - "1/1 [==============================] - 0s 26ms/step - loss: 0.1367 - categorical_accuracy: 0.8450\n", - "Epoch 281/290\n", - "1/1 - 0s - loss: 0.0151 - categorical_accuracy: 0.9963 - lr: 5.5000e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 282/290\n", - "1/1 - 0s - loss: 0.0151 - categorical_accuracy: 0.9963 - lr: 5.2750e-04 - 22ms/epoch - 22ms/step\n", - "Epoch 283/290\n", - "1/1 - 0s - loss: 0.0150 - categorical_accuracy: 0.9967 - lr: 5.0500e-04 - 22ms/epoch - 22ms/step\n", - "Epoch 284/290\n", - "1/1 - 0s - loss: 0.0150 - categorical_accuracy: 0.9967 - lr: 4.8250e-04 - 23ms/epoch - 23ms/step\n" - ] + "data": { + "text/html": [ + "
 Non-trainable params: 0 (0.00 B)\n",
+       "
\n" + ], + "text/plain": [ + "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ - "Epoch 285/290\n", - "1/1 - 0s - loss: 0.0149 - categorical_accuracy: 0.9967 - lr: 4.6000e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 286/290\n", - "1/1 - 0s - loss: 0.0149 - categorical_accuracy: 0.9967 - lr: 4.3750e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 287/290\n", - "1/1 - 0s - loss: 0.0148 - categorical_accuracy: 0.9971 - lr: 4.1500e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 288/290\n", - "1/1 - 0s - loss: 0.0148 - categorical_accuracy: 0.9967 - lr: 3.9250e-04 - 26ms/epoch - 26ms/step\n", - "Epoch 289/290\n", - "1/1 - 0s - loss: 0.0147 - categorical_accuracy: 0.9967 - lr: 3.7000e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 290/290\n", - "1/1 - 0s - loss: 0.0147 - categorical_accuracy: 0.9971 - lr: 3.4750e-04 - 23ms/epoch - 23ms/step\n", - "1/1 [==============================] - 0s 26ms/step - loss: 0.1381 - categorical_accuracy: 0.8450\n", - "Epoch 291/300\n", - "1/1 - 0s - loss: 0.0146 - categorical_accuracy: 0.9971 - lr: 3.2500e-04 - 24ms/epoch - 24ms/step\n", - "Epoch 292/300\n", - "1/1 - 0s - loss: 0.0146 - categorical_accuracy: 0.9971 - lr: 3.0250e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 293/300\n", - "1/1 - 0s - loss: 0.0146 - categorical_accuracy: 0.9971 - lr: 2.8000e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 294/300\n", - "1/1 - 0s - loss: 0.0145 - categorical_accuracy: 0.9971 - lr: 2.5750e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 295/300\n", - "1/1 - 0s - loss: 0.0145 - categorical_accuracy: 0.9971 - lr: 2.3500e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 296/300\n", - "1/1 - 0s - loss: 0.0145 - categorical_accuracy: 0.9971 - lr: 2.1250e-04 - 23ms/epoch - 23ms/step\n", - "Epoch 297/300\n", - "1/1 - 0s - loss: 0.0145 - categorical_accuracy: 0.9971 - lr: 1.9000e-04 - 24ms/epoch - 24ms/step\n", - "Epoch 298/300\n", - "1/1 - 0s - loss: 0.0145 - categorical_accuracy: 0.9971 - lr: 1.6750e-04 - 22ms/epoch - 22ms/step\n", - "Epoch 299/300\n", - "1/1 - 0s - loss: 0.0144 - categorical_accuracy: 0.9971 - lr: 1.4500e-04 - 22ms/epoch - 22ms/step\n", - "Epoch 300/300\n", - "1/1 - 0s - loss: 0.0144 - categorical_accuracy: 0.9971 - lr: 1.2250e-04 - 24ms/epoch - 24ms/step\n", - "1/1 [==============================] - 0s 32ms/step - loss: 0.1388 - categorical_accuracy: 0.8450\n", - "Print Time for taining: 23.296875\n" + "None\n", + "Print Time for taining: 172.5625\n" ] }, { "data": { - "image/png": "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\n", + "image/png": "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", "text/plain": [ - "
" + "
" ] }, - "metadata": { - "needs_background": "light" - }, + "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ - "1/1 [==============================] - 0s 28ms/step - loss: 0.1388 - categorical_accuracy: 0.8450\n" + "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 120ms/step - categorical_accuracy: 0.8229 - loss: 0.1132\n" ] }, { "data": { "text/plain": [ - "[0.1388462632894516, 0.8450184464454651]" + "[0.11321771144866943, 0.8228783011436462]" ] }, - "execution_count": 4, + "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model = make_model(\n", - " inputs= [{'shape': (None, 1432), 'name': \"node_attributes\", 'dtype': 'float32', 'ragged': True},\n", - " {'shape': (None, 1), 'name': \"edge_attributes\", 'dtype': 'float32', 'ragged': True},\n", - " {'shape': (None, 2), 'name': \"edge_indices\", 'dtype': 'int64', 'ragged': True}],\n", + " inputs=model_inputs,\n", " gcn_args = {\"units\": 124, \"use_bias\": True, \"activation\": 'relu', \"pooling_method\": 'sum'},\n", " depth = 3, \n", " verbose = 10,\n", @@ -935,7 +377,7 @@ "epostep = 10\n", "\n", "# Compile model with optimizer and loss\n", - "optimizer = tf.keras.optimizers.Adam(lr=learning_rate_start)\n", + "optimizer = ks.optimizers.Adam(learning_rate=learning_rate_start)\n", "cbks = LinearLearningRateScheduler(learning_rate_start, learning_rate_stop, epomin, epo)\n", "model.compile(loss='categorical_crossentropy',\n", " optimizer=optimizer,\n", @@ -952,12 +394,12 @@ " initial_epoch=iepoch,\n", " batch_size=1,\n", " callbacks=[cbks],\n", - " verbose=2,\n", + " verbose=0,\n", " sample_weight=train_mask # Important!!!\n", " )\n", "\n", " trainlossall.append(hist.history)\n", - " testlossall.append(model.evaluate(xtrain, ytrain, sample_weight=val_mask))\n", + " testlossall.append(model.evaluate(xtrain, ytrain, sample_weight=val_mask, verbose=0))\n", "stop = time.process_time()\n", "print(\"Print Time for taining: \", stop - start)\n", "\n", @@ -993,7 +435,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 6, "id": "ff18a377", "metadata": {}, "outputs": [], @@ -1006,30 +448,31 @@ " self.node_index = node_index\n", "\n", " def predict(self, gnn_input, masking_info=None):\n", - " return tf.expand_dims(self.gnn_model(gnn_input, training=False)[0][self.node_index], 0)\n", + " return ops.expand_dims(self.gnn_model(gnn_input, training=False)[0][self.node_index], 0)\n", "\n", " def masked_predict(self, gnn_input, edge_mask, feature_mask, node_mask, training=False):\n", - " node_input, edge_input, edge_index_input = gnn_input\n", + " node_input, edge_input, edge_index_input, node_len, edge_len = gnn_input\n", "\n", - " masked_edge_input = tf.ragged.map_flat_values(tf.math.multiply, tf.dtypes.cast(edge_input, tf.float32),\n", - " edge_mask)\n", - " masked_feature_input = tf.ragged.map_flat_values(tf.math.multiply, tf.dtypes.cast(node_input, tf.float32),\n", - " tf.transpose(feature_mask))\n", - " masked_pred = tf.expand_dims(\n", - " self.gnn_model([masked_feature_input, masked_edge_input, edge_index_input], training=training)[0][\n", + " node_len = ops.convert_to_tensor(node_len)\n", + " edge_len = ops.convert_to_tensor(edge_len)\n", + " edge_index_input = ops.convert_to_tensor(edge_index_input)\n", + " masked_edge_input = ops.convert_to_tensor(edge_input) * ops.cast(edge_mask, dtype=\"float32\")\n", + " masked_feature_input = ops.convert_to_tensor(node_input) * ops.cast(ops.transpose(feature_mask), dtype=\"float32\")\n", + " masked_pred = ops.expand_dims(\n", + " self.gnn_model([masked_feature_input, masked_edge_input, edge_index_input, node_len, edge_len], training=training)[0][\n", " self.node_index], 0)\n", " return masked_pred\n", "\n", " def get_number_of_nodes(self, gnn_input):\n", - " node_input, _, _ = gnn_input\n", + " node_input, _, _, _, _ = gnn_input\n", " return node_input[0].shape[0]\n", "\n", " def get_number_of_node_features(self, gnn_input):\n", - " node_input, _, _ = gnn_input\n", + " node_input, _, _, _ ,_ = gnn_input\n", " return node_input[0].shape[1]\n", "\n", " def get_number_of_edges(self, gnn_input):\n", - " _, edge_input, _ = gnn_input\n", + " _, edge_input, _, _, _ = gnn_input\n", " return edge_input[0].shape[0]\n", "\n", " def get_explanation(self, gnn_input, edge_mask, feature_mask, node_mask, node_labels=None):\n", @@ -1068,7 +511,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 7, "id": "1223fb98", "metadata": {}, "outputs": [], @@ -1096,7 +539,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "id": "eb3ec32d", "metadata": {}, "outputs": [], @@ -1116,7 +559,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 9, "id": "25de373b", "metadata": {}, "outputs": [ @@ -1148,19 +591,10 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 10, "id": "50c43786", "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\patri\\anaconda3\\envs\\gcnn_keras_test\\lib\\site-packages\\keras\\optimizers\\optimizer_v2\\adam.py:110: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n", - " super(Adam, self).__init__(name, **kwargs)\n" - ] - } - ], + "outputs": [], "source": [ "gnnexplaineroptimizer_options = {'edge_mask_loss_weight': 0.001,\n", " 'edge_mask_norm_ord': 2,\n", @@ -1168,8 +602,8 @@ " 'feature_mask_norm_ord': 2,\n", " 'node_mask_loss_weight': 0,\n", " 'node_mask_norm_ord': 1}\n", - "compile_options = {'loss': 'categorical_crossentropy', 'optimizer': tf.keras.optimizers.Adam(lr=1)}\n", - "fit_options = {'epochs': 80, 'verbose': 0}\n", + "compile_options = {'loss': 'categorical_crossentropy', 'optimizer': ks.optimizers.Adam(learning_rate=1.0)}\n", + "fit_options = {'epochs': 80, 'verbose': 2}\n", "\n", "explainer = GNNExplainer(explainable_gcn,\n", " compile_options=compile_options,\n", @@ -1187,7 +621,7 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 11, "id": "7ccdf57d", "metadata": {}, "outputs": [], @@ -1198,15 +632,15 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": 12, "id": "b2799fe7", "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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", "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -1221,20 +655,18 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": 13, "id": "10fe1fa2", "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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", "text/plain": [ - "
" + "
" ] }, - "metadata": { - "needs_background": "light" - }, + "metadata": {}, "output_type": "display_data" } ], @@ -1248,20 +680,18 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 14, "id": "37e0d9bc", "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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", "text/plain": [ - "
" + "
" ] }, - "metadata": { - "needs_background": "light" - }, + "metadata": {}, "output_type": "display_data" } ], @@ -1274,15 +704,15 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 15, "id": "58ae4d21", "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAApQAAAHzCAYAAACe1o1DAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd1zV1f/A8dfnXvYGGYIsRRAF3NscpZk5SsuZiqNc6ddyVLbLXJmWDUstt7Y1y71n7n1BxInIEpA9L9z7+f5BXEVAxr24OM/fg8fv6/18zvmcz+2O9z3jfSRZlmUEQRAEQRAEoZIUD7sBgiAIgiAIwuNNBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJSCIAiCIAiCXkRAKQiCIAiCIOhFBJTCE+3SpUtMmTIFW1tbJEmiRYsWzJkzh/j4+IfdNEEQBEF4YoiAUngiaTQa3njjDerVq8fKlSvp0aMHrq6u+Pj48Mknn+Dh4cHSpUsfdjMFQRAE4Ylg9LAbIAiGplarGTRoEBs2bODLL79k3LhxmJmZ6Y4nJSXx7rvv8tprrxEdHc1HH330EFsrCIIgCI8/EVAKT5xPPvmEjRs38tdff/HCCy8UO+7g4MCiRYtwd3fno48+IigoiD59+jyElgqCIAjCk0GSZVl+2I0Qqp/8/HyMjAz/e+b48eO0adOG6dOn8/7779/3XFmW6d27N0ePHiU0NBRHR0eDt0cQBEEQqgMxh1J4IGRZZv/+/QwaNAhJkjA2NkaSJPr27cvu3bsx1O+ayZMn07hxY955550yz5UkicWLF6NWq5k1a5ZBri8IgiAI1ZEIKIUql5eXx2uvvUanTp04c+YMM2bMYPTo0cyaNYuwsDC6dOnC0KFDUavVel1HpVLx77//8t5775W797NmzZqMHj2aFStWkJ2drdf1BUEQBKG6EkPeQpVSq9X06dOHHTt2sGTJEoYPH44kSbrjsizz66+/Mnz4cNq3b8+mTZuKLKCpiPHjx7N+/XoiIyMxNjYud7mrV69St25dVqxYwbBhwyp1bUEQBEGozkQPpVClZs2axY4dO9i8eTMjRowoEkxCwbDzoEGD2LFjBwcPHqz0iuuMjAxWr17Na6+9VqFgEsDHx4euXbuyaNGiSl1bEARBEKo7EVAKVebMmTPMnDmT999/n65du9733I4dOzJ9+nTmzZvH4cOHK3ytf/75h/T0dEaNGlWpto4ePZqjR49y5cqVSpUXBEEQhOpMDHkLVaZnz55ERERw+vRpTExMyjxfo9HQqlUrrK2t2bt3b4Wu9dZbb/Hnn39y/fr1SrU1ISEBZ2dnfv/9d/r161epOgRBEAShuhI9lEKViIiIYMuWLUyaNKlcwSSAUqlkypQp7Nu3j4sXL1boeiqViqCgoMo0FQAnJydcXFxQqVSVrkMQBEEQqisRUApV4scff8Ta2pqBAwdWqNxLL72Eo6MjixcvrlA5fQNKgKCgIEJCQvSqQxAEQRCqIxFQCganVqv56aefCA4OxtLSskJlTU1NGTlyZIXS+CQlJRETE2OQgFL0UAqCIAhCxYmAUjC4gwcPEh8fz6uvvlqp8q+++iopKSns3LmzXOcXBoGGCCivXr1KZmamXvUIgiAIQnUjAkrB4M6dO4e5uXmlAzxfX1/s7e05f/58uc4PCQnB2NgYPz+/Sl2vUFBQELIsExYWplc9giAIglDdiIBSMDiVSkVAQABKpbJS5SVJqtB8xps3b+Lh4VHh/JP3ql27NgA3btzQqx5BEARBqG5EQCkYnCEWyAQGBlZoPqNCof9LubIBsCAIgiBUdyKgFAxKo9EQGhpqkPmM4eHh5ObmGqhlgiAIgiBUFRFQCgZ17do1cnJyDBJQajSaCuejFARBEAThwRMBpWBQhcPUgYGBetVTWF6k8REEQRCER58IKAWDio+PR6lU4uLiolc9tra2WFtbc+vWLQO1TBAEQRCEqiICSsHgJElCkiSD1CMIgiAIwqNPBJTCY8/U1JSMjAy96ymsw8zMTO+6BEEQBKE6EQGl8Njz9/cnLi6O5ORkveoJDQ0FoH79+oZoliAIgiBUGyKgrIYiIyP57bffWL16NVevXn3YzdFb4YpyfRfwqFQqLC0t8fb2NkCrBEEQBKH6MHrYDRAenD179vDFF1+wfft2ZFnWPd6pUycmT55Mr169HmLrKq9evXoYGxujUqno0KEDAHmaPA7H/ktIoorwpHBuZcWh0WowVprgae2Bn309mro0p5FTIxRSwe8qlUpFYGCgQZKkC4IgCEJ1IgLKauK7775j4sSJNG3alKVLl9KxY0cAjh07xvfff88LL7zAJ598wkcfffTYLYYxNjbG398flUpFbn4Of17+g03XNpKZl4lSUqKRNbpz1Vo1YUlhXEq+xN9XN+Bi4UJfv/4869UVlUpF06ZNH+KdCIIgCMLjSQSU1cC8efN46623mDJlCnPnzi3SA1enTh0GDhzI7Nmzef/998nKyuLzzz+v9LUsLS3Jz88nPT0da2vrSteTm5tLZmYmVlZW5To/KCiIsKQLTNjzOvFZ8cgU9MDeHUzerfDxW1m3WHj2W/ZG7uF6wjWGBQ2rdJsFQRAEoboSAeUT7uzZs7z77ru89dZbzJ07t8RzJEnivffew8TEhLfeeovnn3+eTp06Vep6AQEBQMECl9atW1e22Vy8eBGNRlPuBOmubWuSVjOZhKwEXTBZoeslhdFqdktcXJwrXFYQBEEQqjsxWewJptFoGD58OA0aNGDGjBllnj958mQ6dOjAiBEjyM7OrtQ169evj0KhKHWBjCzLpKnTuJ2dSFpuapG5nHcLCQkByrfjzom441yuFQ4SaNFWqt1atBiZK9mQu57ojOhK1SEIgiAI1ZXooXyCbd26lXPnznH48GFMTEzKPF+hUPDjjz9Sr149fvvtN4YPH17ha5qbm+Pr61skoEzLTWV35C5O3TrFlZTLZOVn3TnfyJy6dr40cW5CF6+u2JnaAQULZDw8PLC1tb3v9ZJzkvny5DwAJIV+cz8VSgU5mhzmnZjLvI5folQo9apPEARBEKoLSS6ti0h47PXs2ZO4uDhOnjxZoXLdunUjJSWFo0ePVuq6/fr1IzExkY3bN7LywnJ23diJVi7oOSxtOFqiYHedpz06MyJwJIP6DALgk08+YeXKlezdu5eIiAhq1KhB69atmTFjBn5+fsw+NpNjcUe5vj2CqP0xZERnkJ+Zj6mDKY6BDvgN8MXC2aLItTb22VJiG/yH1MP3ZR8AhgWM4GXfvkWOP/vss+zatYvx48fz3XffVeq5EQRBEIQnkeihfELduHGDLVu2sGTJkgqXHTt2LH369OHMmTM0adKkwuWDgoJYtXsl43aNJj0vXRdM3o+MjCzL7L25m+Nxx4iUb9CjYU8+//xz/v33X/r160fDhg2Ji4vju+++o2nTpmzY/RdHYg8DkHo9DQsXc2q2dMbY0pis+Gwid0Zy62QCHb96CjOHorvfODZyxKNTrSKP2dax0f3vPy/9Tq86L2CiLOjZXb9+PUeOHKnwcyEIgiAI1YEIKJ9QP/74I9bW1gwaNKjCZXv27EmtWrX44YcfKhWQmgeaUj+wHmnqtAovkNHKWjLU6XiP8sQ814zJL0zm559/LjJkP2DAAIKCgnh/+ge4jnZGK2tpOKb4XMuarVw4OPVfbu6N1vU8FrJys8T9noDybpl5mfwbc4inPZ4hJyeHKVOm8M477/DRRx9V6H4EQRAEoToQi3KeUNu2baN3795YWlpWuKyRkREDBgxg+/btFS57Iu44hxQHQSp9eLssMjKSJHHK7AR4ycXmf/r6+hIQEMCVS5fv2/tp4WQOQH5mXonHNbkaNOqS0wopUHA45l8A5s6di1arZerUqZW5nQrTarWcPXuWFStWcOLECdRq9QO5riAIgiBUluihfAJpNBpCQ0MZPHhwpeto3LgxX375JWlpadjY2JRdgILFNwtOfQlgsOToX59ZgH+N+jiYOegek2WZ2LhYlM7FF82o09TIWpnsxGwu/X4FAMeGNYqdd3NvFBHbboAMVu5W+Pbzwb3DnR5LLVouJV0iMjKSOXPmsGzZMszNzQ1yT6XJzs7m22+/ZdGiRVy/fl33uLOzM6+++ipTpkyhRo3i9yIIgiAID5sIKJ9AV69eJScnR7fHdWUUlg0JCaFt27YAnDhx4r4LZH5S/UhmfmaRnkltvpb9kw6REZVBg2H++PSuozuWFZ/F7jH7Srx+08mNqdXejVxNDovOfc+lH66wcuXKoifFwJ4J+3nmu466h3a+tgdtXkGvpbG1MYGvNcCpsVORYvb+9ri1rYmFiwU5STlEbL3Bma/OkZ+Vj3c3L915yblJvDHpDZo0acLAgQMr/iRWQGJiIi+88AKnT59mwIABLF26FCcnJzIyMvj5559ZuHAhf/75J1u3bsXHx6fsCgVBEAThARIB5ROoMIejPgFl/fr1USqVqFQqXUB5vwUyW/dtZX/UvmLD3Nc3R5CdeP+clrXau+LctGhCcft69kDBnMqjsUcwz7fC1NSU6dOn88knn1DDpQaOvRwwtjYuUq7Vhy3QqDVkRGUQdSCG/JziQ9pPzW5T5N+enT04MPUQF9eE4/G0O0rTgp7PRNVtjv51nGPHjpXjGau81NRU2rdvz+3btzl48CAtWrQocrx169a8+eabdOvWjbZt23L06FFq165dpW0SBEEQhIoQAeUTSKVS4ejoiLNz5Xd9MTU1xc/Pr0g+ycmTS18gM+3TaTi95lAkUXluSi6Xfr9C3T51CP/lcqnXsq1je98FMgpJQVT6TRQKBYsXL8bR0ZHv1y9kyY1Fxc51DCoYEnZp5kzNli7se/MgRuZKanf3Lr1+YwXe3b1RLQoh5WoqNRo4oNVoCfkplAGv9C8W4BnalClTiIqK4tSpU/j5+ZV4Tp06dTh8+DDNmzdn5MiR7N69u8gWmoIgCILwMIlvpCeQSqUiKChI73mMQUFBRQLKtm3blrpA5nJ48QUyYavDsapliXvH0oPFQvk5+bqh6ntpZS2xabHk5OQQFxdHo0aNmPPJnDLrtHS1xLa2DdH7Y8o819yxIK1QXkbBAp6ovdFkxGQybuzrRERE6P4A0tPTiYiIICsrq7Tqym3Hjh0sXbqUL7/8stRgspCjoyPLli1j3759LFpUPJgWBEEQhIdFBJRPoNDQ0HLvgX0/QUFBhIaG3vccWZa5desW0j2LyZMvpXBzXxQBIxtAGXFt+G+X2TpoB5sHbOPAW/8SfzahyHGNWkP0mRhkWSYrK4tNmzZxYttJVItDyc/Ov2/dGrWWvKz7nwOQdasgODSxKQiYsxOzkfNlOrbvSO3atXV/AKtWraJ27drs2LGjzHrLMmvWLNq2bctrr71WrvOfeeYZgoODmTNnDhpNySvUBUEQBOFBEwHlAxYeHs68efNo1qwZixcvJiam7N6zisrIyMDe3l7veuzt7cnIyLjvOWvXriU6OhrXdjV1j8myTMhPodRq54qD/33aIUk4NXakwTB/WrzXjIAR9VGn5nLssxPcOhlfUJdG5tS8M6jT1HR8tgO//PILv/zyC4MGDSJi2w2OTT9BvjofdUbx1EDJl1JIv5GOXd072zfmpuYWOy8/O59rGyMwsTHBzqfgXPf27oxfMI6//vqryB9A9+7d+euvv2jVqtV9n5uyXLhwgf379zNx4sQK9SZPnDiRmzdvsnXrVr2uLwiCIBSVkJDA559/TmBgIJIk8dxzz7FkyZIyvwsFMYfygTl9+jRvvfUWe/bs0T02duxYxo8fz4svvshXX32Fp6fnQ2xhxV28eJHx48fTpEUTaj19J6C8uSeKtBvpNH+76X3LWziZ0/rjlkUec+9Ui30TDxC6IgyX5s6Ergjj1ol4XFo483yf58nPL+ht7NKlC9cTrnNgywGi98cQuiwMt3auWHtYoTQzIv1GOjf3RGFkYYRvv7q6+iO23iDu2C1cWrhg7mhGbnIukbujyE7MpskbjVAYF/zGsqxlwdvPTMPTpvh/k9q1a9O7d+/KPm06ixcvxsnJiT59+lSoXLNmzWjevDk//PADPXv21LsdgiAI1Z0sy8ydO5ePPvoISZJo164doaGhaLVaxo0bx1tvvcWPP/5I//79H3ZTH1kioHwAtm7dSr9+/fDx8WHNmjX07dsXU1NTUlNTWbNmDXPnzqVNmzZs3bqVhg0bPuzmlktcXBw9evTA1taWRat/YEbYdADysvK4uCacur3rYO5Y8byNJtYmeDzjzpX118hOzCbtehoAt07EM+3Eu8XOlxQSyRdT8Oziwe2Q28QeiUOj1mBmb0at9q749qtbZC9vB397ki6mELnrJup0NUamSux87Wg8IQjHho5AwSKg+g4NSgwmDSUrK4uVK1cybty4YvNSy2Ps2LGMGjWKiIgIvL29Dd9AQRCEakKr1TJx4kQWLlzIW2+9xTvvvFMk529kZCTvvPMOAwYMICYmhjfffPPhNfYRJgLKKnbw4EF69epF9+7d+fXXX7GwuBPc2NraMn78eF5++WV69OhBhw4dOHPmzCOfEiY1NZXnn3+elJQUDh48iHcdbwgrOHZ1w3W0+TJuT7mSFV8wLzE7MQcAdUYeWfFZmNmb6XoCS1IYiOZl5NF2Rmvd42ue/xkbU9si59o72pOXmUfj/5UvEHdq7FQsL2VJxjQcV+Ljd69i18eRI0dITU1lyJAhlSo/cOBAxowZw/bt2xkzZoxB2iQIglAdLViwgO+//54lS5YwatSoYsc9PT1Zu3YtHh4eTJo0CT8/P7p37/4QWvpoEwFlFcrMzGT48OG0bt2a9evXY2RU8tNds2ZN9uzZQ8OGDXn11VfZtWvXI5sSJicnh169enHp0iV27dpFgwYNAKhpWZO4zDiyE7PJy8hj38SDxcpeWXeVK+uu0uHLp7CtXfruO5n3LJABcDBzKBZMpqenk5qUSt2aTVFIivtuw1gRr/gPwdvW2yB1lUalUmFmZoa/v3+lyltaWhZL6yQIgiBUzMWLF3n//fd54403SgwmCykUCj7//HPOnz/PqFGjCAkJMchahSfJoxm1PCE++eQTYmNjWb58eanBZCFbW1uWLVvG3r17WbZs2QNqYcVoNBoGDBjAkSNH+OOPP2jT5k6C8MZOTVFISmr38Kb5tKZF/hqOK1hx7vFMLZpPa4qFc0EPZEkLZLJv53BzdxQ23taYORSk8pHzZOpZFA+8PvvsM2RZZsrgqXjb1EYh6f9yftrjGfr69dO7nrKoVCoaNGiAUll8+8jyCgoK0iWxFwRBECpuypQpuLu7M3PmzDLPlSSJn376iYyMjHKdX92IHsoqkpGRweLFi5k0aRK+vr7lKtO5c2defvllFixYwKuvvqpXHklDDM3eW8eUKVP4559/6NWrF0lJSaxZs0Z3LCczB62LBjsfW91K6UKFQ9/WHta4trqzeCds1UUy47JwbOiImb0pWfHZ3NgRiSZHQ8CrDXTnZSdnM2fUXG6+Eq3r0du+fTtbtmyhW7du9H+5P700vZh9bCbnE89V+D4lJGRketV5gZFBrxkkMC1LSEiIXjsZQUFAuWvXLmRZNtje6YIgCNXF9evX2bp1Kz/++GOR6Wj34+7uzqhRo1i+fDmfffYZ5uYVXyvwpBIBZRX55ZdfyMzMrPD8tnHjxtGlSxcOHTpE+/btK3VtFxcXoqOjK1X2blFRUUV22zl79iwAGzduZOPGjcXOf3v/FMKTw8s99OzU2InMbZFEbLlBXmYexpbG1GjggG+/urqgVCEpqOvqi2svd3bu3MnKlSvRaDTUrVuXWbNmMXXqVBQKBZYKS6a3m8G2iK0sU/1Enjav2DaQ9yoMJO1M7Xmz2SSaON9/VbqhaLVaQkND6ddPv57QwMBAkpKSiI2Nxc3NzUCtEwRBqB6WLFmCjY0NAwcOrFC5MWPGMH/+fP744w+Cg4OrqHWPH0k21CoDQUeWZZo1a0atWrVKDLzuR6vV4u/vT4sWLVi7dm2lrj98+HAuXLjA8ePHK1W+UI8ePQDYvHlzuc6PTIvkjb0T0MiGS7itkBTM77gAHzufcpdJzU1lV+RONl/bSGJ2IgBajRYjpRHyf/8HUMe2Dr3qvMBT7h0wVZoarM1luXr1KnXr1mXbtm0899xzetezfft2unbtasAWCoIgPNnUajXu7u4MHDiQb775psLln332WTIzMzl8+HAVtO7xJOZQVoFTp05x5swZxo4dW+GyCoWCMWPG8Oeff3L79u1KXb9whxuttnhPYXZKAjFnDnBt/wau7fuL6FN7ybodV+IQeeEWjuXlaeNJcINhlWpzaQbWG1ShYBLA1tSWl337srTrCpZ2XU7/GgMJWx1OJ9tnGBn4Kp+0mc6a539mwdPf0tnr2QcaTAK63Yf03c2odu3aWFpainmUgiAIFbRt2zYSEhIqnSVjzJgxHDlyhEuXLhm4ZY8vMeRdBfbv34+FhQXdunWrVPmXX36ZqVOncuzYsUqlJggKCiIrK4vr16/j4+NDZmIMV3f9ztW968lJSSixjKm1Pd4dXsS360Csa3qRkpLCzZs3KzzPr3fdl4jPSmDz9Yr1zJbkWa+uDKg3qNLlJUnCycIZl5yaXPv7Or0X9Hkkcjamp6cDFMlzVhkKhQJ7e3vS0tIM0SxBEIRq49SpUzg7OxMQEFCp8s888wxQsGmJn5+fIZv22BIBZRUICQkhICCg0it4vby8sLa2RqVSVSqgLOz5On/mNGmH13Fx8wokSUIuoceyUG56Mpe2riZ88wrqPNOPXN+C+ZsVDSglSWJ0wzHYm9mxNmwNkiRVKJ2PQlIgyzL9/AbwSv3BBllskpSUBICDg4PedQmCIAiPv4qOwN3LwcEBNzc3VCpVhedgPqlEQFkFVCqVXjveSJJEYGBgpXMMurq60sDDibQNc7mYnwmyXK5V37K2YO7j9X3r0BzajJ+TZaXyJEqSRP96A2nm0pyvTn1JZPqNMvNEFh53tXTlzaaTqedQufyMJUlKSsLIyAhra2uD1SkIgiA8vkJCQnTrBCpLpG4rSgSUBqbRaAgNDWXw4MF61RMUFMSxY8cqVTb15mXeauWIkToDKtHBJ2u1SOos3u/oQeq1EJz8K7f62ceuLt8+s5DziefYfG0TZ+LPkKvJKXaeqdKUhk6N6FG7J42dmxg8bU9SUhIODg4itY4gCIJAVlYWV65c0Xsee2BgIOvXrzdQqx5/IqA0sKtXr5KTk6N3jsHAwEBWrFhBXl4exsbG5S6Xm57Mns9GYKKoVCypIwFGCtg3exTd5/2DpVOtytUjSTRyakwjp8ZoZS1xmbFEZ0Sj1qiZM3MOpMFvq/+s0tyPhQGlIAiCIFy4cAFZlg2SC3j+/Pmkp6eLETDEKm+DKxymNsQLVa1Wc/nyZd1jJ06cYMKECQQEBGBpaYmnpyf9+/cvssrs5LLPUGekEpOWw5wDNxmxPpxRGy7x/bEY0nLyi1wjOTuPhUdjmLL1GiPXX+K1vy7xwa4IDkSkFiTLBjTqXI4t+qDYkPmzzz6LJElMmDCh3PekkBS4WdWiRc2WtKv1FB54Eqm6WeWJxEVAKQiCIBRSqVRIklTpBTmFCr/nCzN3VHcioDSwixcv4uDggIuLi171FL7Qw8LCdI99/vnnrFu3js6dO/P1118zevRoDhw4QNOmTQkJCSEu5CiRh7eSmJHD9L2R3MpQMyDQiR5+DpyJzWD2gZvka+4Ehum5GpKy82jpbs3gRk70C3TC3syIRcdj+U1VkL9R1mq4FXKUyCNbdeXWr1/PkSNH9Lo/AE9PT27evKl3PWV5VAPKqtjNSBAEQbi/iIgIXF1dsbS01KueunXrAgU77ghiyNvg1Gq13i9SQFdHbu6d/a4nT57Mzz//jImJie6xAQMGEBQUxJw5cxjdyAZJoeTvsDhy87XM7OKNo2XBcLlPDXNm77/J/ohUOvvYAeBpZ8aHT3sVue5zvvZ8cfAm268k0T/QEYVCQpIUhG9ZhVfb7uTk5DBlyhTeeecdPvroI73u0cPDg+TkZDIyMrCystKrrvtJSkqiTp06VVZ/RTk6OgIQExODj0/FcmzeLS8vj/j4+CK7GQmCIAj3J8typbOw3M3IyEhXnyB6KB8rbdu2LRJMAvj6+hIQEEBoiIroU/uQtRqOR6XTxNVKF0wCBLlY4mptwtGbZecsdLI0Rp0vk68teJPIspbbl8+RevMyc+fORavVMnXqVL3vx9PTE6DKeylv3779SPVQFg6TVHYVf6Hw8HDy8vL0nl4hCIIgCPoSAeVjTpZlbt26hY2ZEcgySVl5pOVqqONgVuxcHwczbqQUX2WtzteSlptPQqaaAxGp7I9IxbeGOSZGd788JM7u38acOXP4/PPPMTc317vtHh4eQNUHlI/akLerqysODg4lBpRaWUtU+k2Oxhxhf9Q+Dkf/y/XUa+Rr84udW1he35WKgiAIgqAvMeT9mFu7di3R0dG81r0dUsYFUnKyAbAzK/6f1s7MiAy1ljyNFmPlnWBx2+VkflXd2UEnwNmCsS1di5SVlEqmf/U9TZo0MVgS11q1aiFJEpGRkQaprySyLD9yAaUkSUXyl8myfFdqpdPkanKLlTFSGNHAIYDudXrQqmZrlAolISEh1KpVC3t7+wd9C4IgCIJQhAgoH2MXL15k/PjxtGnThmf8XYk5HoL6v0U3xsriSYMKH1NrZIzvmj7SxtOG2g5mpOdqOB2TQWpOvq6eQiGxqew9fZNjx48brP3Gxsa4urpWaQ9lZmYmeXl5j1RACQW9inv27OFaylUWnP6SiLQIlJISjawp8fx8bT4hiSrOJ57D0dyRN5pO0nunB0EQBEEwFDHk/ZiKi4ujR48e2Nra8ueffyLJWmRZxuS/oDFPU3yScOFjJvcEm06WxgS5WNLW04YJrd1wtjJm1v5I1PkFO9totDKrzsTTOcibFi1aGPQ+PDw8qrSH8lHddjEwKJD8+mom73uTyPSC+y8tmCykpeC/R1J2Eh/++z6JvgkEBonhbkEQBOHhEwFlFdBo7h8YVKSOknZ3SU1N5fnnnyclJYVt27bh5uaGkak5kkKhG+pOySk+5y4lJx8rE0WR4e6StHK34XZWPmGJWQAcjEglJj2XF9oEEhERofsDSE9PJyIigqysrErdZ1WnDnoUA0pZlkmqnYj/0Hpo0VZor3O4E1jWeMqejBZpaLT6v94EQRCqCxMTE7Kzs/Wup/B7797FstWVCCgNzNPTk9jY2EoHWIWuXbsGgJdX0bQ+OTk59OrVi0uXLrFp0yYaNGgAgK1HXUDGwcIYG1Ml15KKL765mpSDl13xxTr3UmsKApbsvIL/n5iVh0YLE5dsonbt2ro/gFWrVlG7dm127NhRqfv08PCodgHl1uubOZqlfx5PSZKINo5iRegyA7RKEASheqhXrx6JiYkkJCSUffJ9XLhwAQB/f39DNOuxJ+ZQGlhgYCCyLBMWFkazZs0qXU/hCt67M/lrNBoGDBjAkSNH+Pvvv2nTpo3umEOdQGRtQQDYwt2agxGp3M7Ko4ZFQeqgkFuZxKared73zgKOtJz8gtXh99h3PRUJ8P4v+GzjaYOXvRn+PYbjXL+57rw+ffrQvXt3Ro0aRatWrSp1n56enkRGRhbszFMFe20/6IAyIyODL774gmPHjnH8+HGSk5NZvnw5w4cPByA2M5alIT8BIGtlbuyI5Mb2SDJiMlGaKLHxtiFgZH1sa9sUqTczNpPwXy6RcO42+Tn5mNcww7WtK/WH1OPvqxto5dqGrGtZrFixgmPHjnH+/Hny8/NFfjRBEIR73J267Zlnnql0PSqVCmNjY+rVq2eopj3WREBpYIUBoEqlKhJQ5mnzuJEWwZWUK8RlxpKvzcdEaYq7lTt17epSy9odpXRnpYxKpcLT0xNbW1vdY1OmTOGff/6hV69eJCUlsWbNGt0xTX4eFuaW5Gdn0rt+DY7dTGfGvki6+dqTky+zKfw2HramdKx9p74NYbe5lJhNw5qWOFoYk6HWcDw6nWtJOTxX156a1gXd+LVsTHG3t6TP+CmY2hRdUVy7dm169+5d6efLw8ODnJwcbt++rUv4bUhJSUlIklTkeaxKiYmJTJ8+HU9PTxo1asS+ffuKHF907nvdXMmz350n+kAM7p1q4d3dG01OPqnX01CnFl3lnXo9jcMfHMWshhk+L9bG2NqE7IRscm4XDNkoUPDNmQW4HqrFTz/9RMOGDalTp06RLTkFQRCEAnXr1sXU1JSQkBBdQKmVtagSVagSznE55TLR6VHky/mYKEzwsvGmrr0vzV2a42NXV1dPSEgI/v7+GBsbl3apakUElAZmZWVFnTp1dD2MtzJvseX6ZnZEbCMzPxOgSOBYGFzUMKtB9zo96er1HLamtoSEhBRbwXv27FkANm7cyMaNG4td+9TKOVzaupoaFvDR056sPhvPr+cTUCokmrhaMaSxc5H5k01crbiVoWZ/RCppufmYKBR42JkypkVNOnjfCcAkhRKvtt2LBZOGUJiLMjIyssoCSnt7exSKBzO7w9XVldjYWGrWrMnJkyeLLGKKSo/iTPxpAGL+jSVqbzTN32mKa+uapdYna2XOLDiLlbsVbae3QmlafHcHLVriMmPp268fqdNSMTc3Z8KECSKgFARBKIGRkRH169dHpVKhkTVsv76Vv678xa2sOJSSEq2sRebO6M6trFucuHWctWGr8bH14WW/frRzewqVSiXyAN9FBJRVICgoCFWIij8u/c7PYWuQkYssvChpNe/tnNusubCKPy/9zpiG41CpVLzyyitFzrm3t+teGfHRXN7+M7JWg7utKe929Lh/O2taElSz7G0iZVnGv+eIEh/X19275TRt2lTv+u71oHNQmpqaUrNmyQHijhvbUEgKtLKWq/9cx87XFtfWNZG1Mhq1BqMSph8knE0kPTKDlh80R2mqJD9Xg9JIgXTPSn2FpOB4xjG6mj9XJfclCILwJAkKCuJCdChv7Z/ClZTLusdL+n6WkXXfd9dSrzH3xBxa1mxF+M1wevTo8cDa/KgTAWUVqN+4Psdsj7D6wsoKlZORyc7PZsHpL3Ee6EiDoAYVKm/lXIvGQ6ZyesWsCpW7L0ki4KUx2HtXzaRjJycnTExMqix10KOU1Pxcwjm0spa8rDxSLqfg3c2LsDXhXN8cgSZHg4WLOfWH+uPW7k5S+YTziQAojRUcmHqI1KtpKIwU1GztQtDoAEz+m5ZQMFxzvsrmogqCIDxJPFq6c9s5nmupVytUrrDn8mTcCZp80gh3R/eqaN5jSazyNrB0dTpRjW5g4anf1oSurWtyzukMedq8CpXze24wro2eQpL0/08rKZQ41Akk4KWxetdVGoVCUaUrvR+VgDJPm0dk2g0AsuKyQIboQzHc3H2TBsP8aTKpESY2Jpyaf4b403dWHmbGFEyTODnvDFa1rGj2dhN8+tQh9kgcx2edKtJLnJ2fza2suAd7Y4IgCI+ZsNsXuOAegsJIUeG0bYW0aDGxMWabtJm4TPG5CyKgNChZlvnixOekK9JRlJHrsSySQuJK1mVWhCyvYDkFT035Bqf6zUGPoFJSKLDzqsfT7/2I0qhqc2xVh4AyKTtJN5SSn1Pw//PS82jxbnO8u3nh3qEWbT5thYm1CZf+uKIrp/nvXLu6djSd1Bi3Nq74v+JHvUF+JF9MJvH87SLXER9sgiAIpcvKy+LzE7PRIiMp9BvNUSgVZGoymX9ybpkbU1QHIqCsgBMnTjBhwgQCAgKwtLTE09OT/v376xY/7IrcydmEM2hlLek3Mzg6/ThbBm1n29CdnF5wltzU4ns0Z8ZmcnLuabYN2cHmAds49O4RElUFQYKMzMZrf6OKP8+KFSt44YUX8PDwwNLSksDAQGbMmEFOTvF8k0am5nR670fqdR8KSEgVWZDyXxBau2MfOn+8ChOrql8dXZg6qCo8KgGlRr6TaF5pUvAcW7iYY+9np3vcyNwIl+bOpFxJQftfLlDFf+fWal90b/VaHdwASL6YfM91xIeaIAhCaZaHLiUlJwWZ4j2Tl/64wsY+W9g38UCxY0kXkzn07hE2D9jGjhG7CPkplPzsfLSylvDkcDZf2/Qgmv9IEwFlBXz++eesW7eOzp078/XXXzN69GgOHDhA06ZNOX32FEtVPwKQnZjN4Q+Okhmbhf/gevi8WJv4Uwkc/eQ42rw7L+LsxGwOTTtCUlgSPr3rUH9IPTQ5+Rz99Di3QwvyJypQ8N2J7xgxYgQJCQmMHTuWBQsW0LJlSz7++GOef/75EhfHKI1NaBo8jS6frsGhTsEqNElZfIVwIS0Fv9Rs3X3o9O6PtBo7A2PzshfsGEJ16KE0VZrq/reZQ0F+T1Nb0+Ln2Zog58u6nknduXamxc4DUGcWnRJx93UEobqRZZlbt25x5coVMjMzH3ZzhEdMQlYCOyK263Ybu1t2YjZX1l1FaVb8ezL1ehpHPj6GRq0hYER9PLt4cGPHTU5+cUZ3zq8XfyZPU7Epak8asSinAiZPnszPP/9cZJulAQMGEBQUxNufvo3FsIIv/8vrrpKfk0/7ee2wcCqYS2nna8fRT45zc28UXl0LVjZfWXeVvMw8On3dHqtaVgB4dvVk74T9hC67QIf5T6FFS3TWTX7e+jODug3SXXfUqFF4e3vz8ccfs3v3brp06VJim538m9J15m8kR4Rx/cDfJF46S3JEGNo8NQAKI2NsPf3Q2Ljy5ryfWPjL17g2fsrwT959eHh4EB0dTX5+PkZGhn1JPioBpYNZDcyUZuRocjBzMMPU3pTsEnYzyknORWGiwMi84Hmw9bGFnTfJuV303Jykgt5uU5ui0xE8rD2r6A4E4dGVkpLCTz/9xOLFi7lypWDKiFKppHfv3rz++ut6Ja8WnhzbI7YhSVKJnTAXVl7E3s8OWSujTlMXOXZxTTjGlsa0/awVxv9tFmLubMH571XEn03AubETGXkZHI75l44enR7ErTySRA9lBbRt27bYnp2+vr4EBAQQciEE6b9evtgjcbg0d9YFkwBOjRyxdLMk5t9Y3WO3w5KxrW2jCyYBjEyV1GzhQuq1NDL+W5BhbGLMbZfiW0T16dMHgLCwsDLbbu9dn6bB0+g641f6rzpDv5Wn6LvyJP1Xn6Xb7D/pPu0blE5erF69ugLPiGF4enqi1WqJjY0t++QKyM7OJjs7+5EIKCVJKpIQ162dKzmJOSScvfPfNTdNTdzxWzgG1dDN7anZ0gWFsYKbe6KQtXc+BCN3FfToOjW6k7vT3tQBW9MHk8BdEB4VV65coXnz5rz//vu0atWKP/74g59++on58+dz8eJFOnfuzLRp09BqK7f4QnhyHIjaV+IinNuhScQejiPg1frFjuVl5ZFwLhH3jrV0wSSAR6daKM2UxP73na5AwcHo4kPl1YnoodRT4RCL7CQXpP25nYM6VY1d3eJf7Ha+tsSfuhNAaPO0GFsWz7BfmLw69WoqVm6WaGQNobdDi50XF1ewAKOiCcElhQIjM4uij0kSw4YNY8aMGXz//fdYW1tXqE593J3cvPB/G0JycsH8wgcdUH733XekpKQQExMDFCSij4qK4sbtm+Q3z8fI0gjfl32I+TeWk3PPUOcFb4wsjLmxPRJtvhb/wXe28TKzN8W3rw/hv1zm6PQTuLZyITUijcidN6nV3hU7XzsAFJKSurIfM2bMAODkyZMAun97eXkxdOjQB/gsCELVu3DhAp06dcLBwYGwsDDq1KlT5PjEiRNZsGABkydPJiEhgZ9++kmk1aqmsvKyiCshC4askQn5KRTPLu7YeNkUO55+Ix1ZIxf7TlcYK7CtbUPqtTSgYNX3peTwqmn8Y0IElHpau3Yt0dHRNOpdsKtNbnLB0KSpvVmxc83szcjLyEOTp0FprMSqliVJF5LIz87XDXECJIUVzJ/MuWtINCYjmpz8HMyM7tQ7d+5cbGxseP755w1yL0OGDOH9999n3bp1ur2nH4S7k5sb0oPex7vQvHnzuHHjhu7f69evZ/369QB0XdwFI8uCOZHtZrXhwoowrm2MQJuvxaGePU3fbFRsH2/ffnUxtjLm+uYbhCy7gKmdKb596+LX/06Pp1bW4Kv1ZdCHg4qU/fDDDwHo2LGjCCiFJ0peXh5DhgzBycmJAwcOUKNGjWLnSJLEpEmTcHBwYPjw4Tz99NMMGTLkIbRWeNgi0iJKfnz7DbLis2n9ScsSj+ck/ze9yL6EOe/2piSF3VkYmZKbQpo6DRuT4oFpdSACSj1cvHiR8ePH07B5EB5PFyQ31agLutOVxsVnEyj+e0yr1qI0VuL1nCe3TsRzat4Z/Af7oTRVErEtkpSrqUXqgoIV32nqNF1AOWvWLHbt2sX333+PnZ2dQe7Hw8ODZ555hpUrVz7QgNLGxgYbG5snJqCMiIgo9djic4vYen0zWrRY1rSgxbRmpZ5bSJIkanf3pnZ37xKPKyQFjZwaM7DtQAbKAyvZakF4vMyePZvz589z7NixEoPJuw0bNowdO3bwv//9j2eeeQY3N7cH1ErhUZGZl1HsMXWamvBfLuPXv26JiyThzvewwqj4d7rSRIkmt2hmjay8zGobUIo5lJUUFxdHjx49sLW1Zc6Pc3Rb4RWmhNHkFZ+nUbjCuzAVjEszZwJHNeD2hSQOTPmXvRMOEH8qHv9XCoY8je5Zbab9LyXMb7/9xgcffMCrr77KuHHjDHpfwcHB7Nu3r0gP24Pg4eFh8NRBhQFlWV82D9LQBsHYm9nr5tsagrHCmAmN/2ew+gThUZeens4XX3zB5MmTadas7B9lAN9++y0AX3/9dVU2TXhEKUrIy3zx50uYWBuX+mMd7nyna/OLf6dr1BrdFLU71yk9m8qTTgSUlZCamsrzzz9PSkoK27Ztw8vDW3escKi7cOj7bjnJORhbGaM0vvOCq93dm67LO9Nudhvaz2vH0992xNiyoOPY0q1o2h4LY0t27txJcHAwPXr0YNGiRQa/t5deeglLS8sHvjjH09OzynooDdWDawgWxhZMbf52iR9ulfV64wk4WTgbrD5BeNStXbuWrKwsJk6cWO4yhcPey5YtIze3eE5g4cnmYFZ0pCojJpMbOyOp3cObnOQcsuKzyIrPQpOnRauRyYrPQp2uxuy/oe7c5OKvmdzkXN1xAAkJ22raOwkioKywnJwcevXqxaVLl9i0aRMNGjTA28Zbd9y8hhkmNiakXEktVjblcmqx+XEARmZGOPjbY+dji6SUSDiXiMJEgYO/ve4cW1M7ws6E0adPH5o3b87vv/9u8BQ7AFZWVrz88susWrWqxNQKVaWqeihtbW2r5HnSR4BjINNavodSUqKo5FuwsIdzdNAYnvYQKVGE6kOWZX744Qd69eqFu3vF9lEeM2YMiYmJrFu3ropaJzyqPKw9MZLufBfkJOWAFkJ+usDuMft0fymXUsiMyWT3mH1c+v0K1p7WSEqp2He6Nk9L6vU0bO76TnezqoWpUfH1E9WFCCgrQKPRMGDAAI4cOcIff/xBmzZtgIJep5oWNXXnubapya2T8WQnZuseSzifSGZMJq5taxar925JF5OJO3oLzy4euhXgChTYpzrQo0cPvL292bRpE+bm+u0Vfj/Dhg3j8uXLHD16tMquca+q6qF8FFIGlaSVa2tmPjWHGuY1Kjz8rZAUWBpb8m7L9+np80IVtVAQHk1Hjx7l/PnzjB07tsJl/f39efrpp6tkdEd4tBkpjPCxq6v7vLX2tKb5tKbF/qw9rDB3MqP5tKZ4dnHH2NIYx4aORO2PJj/7zo5nUfuj0eRocGtbsIuZUlISUCPgodzbo+LR6rp5xE2ZMoV//vmHXr16kZSUxJo1a3THlNHGKOop0KLFt68PMYdjOfzhMer09CY/J5+rG65j7WWNR+c7v6iz4rM5Ne80Li1cMLM3JT0ynYjtkVh7W1N/yJ3UMepsNb9N/Z3k5GTeeustNm/eXKRdPj4+uuDWEDp16oSHhwerVq0yaL334+HhQWJiItnZ2QYLlh/lgBKgQY0GLOy8iK/2z+dQ4kGUJsqCpLuU0DMsFyzOkSSJju6dGBn4msg5KVRLGzZswNXVla5du1aq/IgRIwgODiYxMbHCKdeEx9tz3t0IT74IFGwK4dqqeAfP9Y0RAEWO+Q/24993j/DvB0fxetaDnNs5XP3nOk6NHXFu6gQUbHvbxatyr8knhQgoK+Ds2bNAQV7BjRs3Fjv+wl89ADB3NKfdjNaELg8jbHU4CiMJ52bOBIyoX2T+pJGFEab2ZkRsvUFeeh5mNUyp08Mb3351i6QRMso2Jj4mHoBp06YVu+6wYcMMGvgpFAqGDh3K999/z1dffYWZWdV34d+dOsjPz88gdd6+ffuRDigBzIzMUC0N5eLpyyzY9BVnE89wKfkSqbkpd07KBc0tLSO7jaSz17PYmdo9rOYKwkN3/vx5mjVrhkJRuQG25s2bA6BSqXj66acN2TThEde+Vnt+Ui0hKz+rQuXsfGxp/UlLwlaHE7o8DCMzIzw7e1B/aEHHjwIFHjae1LOvV0ZNTzZJfpAT5Z5w35/9rtR9QvXxWtBoXvB50aB1liU8PBx/f3/++OMP+vbtW+XXu3r1KnXr1mXXrl107tzZIHV27twZJycnfv31V4PUVxUiIyOpU6cO8+fP54033tA9npWXhVqrxkgy4qu5X/Hl/C9JSkoSSZmFas/d3Z3g4GBmzZpVqfL5+flYWloyb948/vc/kR2hutkRsZ3vzn5j8HpnP/U5AY6BBq/3cSLmUBrQ8ICR2JnZVXqhxb0UkoJ69v70rNPLIPVVRL169WjVqhUrV658INcrnFxvyIU5ZQ15h4eHM3nyZNzc3JAkCTMzM8aNG8e5c+cM1oayfPPNN1hZWTFy5Mgij1sYW2BnaoeViRUtmrcgJSWFa9euPbB2CcKjKDk5mejoaIKCgipdh5GREfXr10elUhmwZcLjop7kT87VHLQaw/SlSUj0qvNitQ8mQQSUBmVhbFGweleh1DvPoEJSYG1szZTmbxk0xUxFBAcHs3XrVuLj46v8Wqampri4uBh0YU5pAaUsy3z88cf4+/uzatUq+vTpQ+vWrRk1ahR///03jRs35n//+x8ajaaEWg0nLS2NH3/8kTFjxtx3q8vCPHuF2ykKQnUVEhICQGCgfl/eQUFBIqCsho4cOUKrVq24tiwCFzNnvb9bJSQaOzdheOAIA7Xw8SYCSgPzd6jPx20+xVhpXOkXq6yRsTa2YVb7OdS0vP+q8Ko0cOBAFAoFP//88wO5nqFTB5UUUGq1WkaNGsX06dOZPn06UVFRLFy4kCNHjvDtt99y48YNvv32W77//nv69euHWq02WHvu9dNPP5GVlVXmsJuTkxNeXl4ioBSqPZVKhZGREfXq6TdXLSgoiJCQkAeaGk14uFatWkWnTp3w8/PjyL6jfNXla3xs65Zd8D5au7Xh/VYfYqwwNlArH28ioKwCDZ0asaDTN/jY+lSoXGGvZvzpBJz+dcbD2rMqmlduDg4O9OrVi1WrVj2Q6xkydZBarSYjI6NYQLlo0SKWLl3KypUr+fDDD4stODI2NmbChAn8/fffbNy4sdLztMqSn5/P119/zcCBA8uVS6958+YioBSqvZCQEPz9/TExMdGrnsDAQDIyMh74jmDCg6fRaHj77bcZNmwYwcHB7Nq1CycnJ2xMbZnbYR5DGwwryAlczg4gCQlzI3MmNZ3CtBbvYaLU77X4JBEBZRVxt/Zgbsf5jG04Dpf/clQqpeJD4QoUuhdyHds6vNPiXV6268eC2V+zd+/eB97uew0bNowzZ848kOEhDw8PgwWUycnJQNF9vK9du8Zbb73F66+/TnBw8H3L9+zZk/fff5+ZM2dy+vRpg7Tpbn/++SeRkZFMmTKlXOc3a9aM06dPo9UadsGXIDxObt++jYuLi9711KxZU1ef8ORKS0vjxRdfZP78+SxYsIAlS5YU+TGiVCjp59efRV1+pHfdPlga3dmdTikpUUpGKO/aSrGGWQ2GNghmybNLedrzGbFI8h5ilfcDIMsyqsTznE84z+WUS0RnRJGvzcdEaYKXjTe+dr40c2mOj11B97tGo6FLly5cuXKF8+fPY29vX8YVqo5araZWrVoMHz6cL774okqvNX/+fD7++GPS09P1fqOGhYXRoEEDDh48yFNPPQXA4MGDOXToEKGhoVhZWZVZh1qtpkWLFjg7O7Nz50692nM3WZZp2bIltra27Nq1q1xldu7cSdeuXQkPDzdYWiVBeNwMGDCApKQkvd+PZ86coWnTppw8ebLce4ELj5dr167Rq1cvoqOj+e2333juuefKLJOnzSMy7QZXUq4Qmxnz3/e0KR5WHvjY1cXd2v2hrWl4HIg8lA+AJEk0dGpEQ6dG5TpfqVSycuVKGjZsyPjx4x/YHMaSmJiY8Morr7BmzRpmz55dpdsYenp6kpmZSUpKit5BdOE+3oU9lAkJCfz555/MmjWrXMEkFNz71KlTCQ4O5vLly/j6+urVpkIHDx7k5MmTbNmypdxl7l6YIwJKQRCE0u3bt4++fftib2/P0aNH8ff3L1c5Y4UxPnZ1dZ07QsWIUPsR5enpyQ8//MAvv/zyUANKKFjtHRcXV+7etMry8PAADJM66N6Acvny5UiSxPDhwytUT79+/XBwcGDJkiV6t6nQ/PnzadCgAd26dSt3GQcHB+rUqVPqPMorV66wbt06Tpw4QXZ2donnCIIgPOmWLFnCs88+S5MmTTh27Fi5g0lBfyKgfIQNGjSIV155hddff/2hTh5v2rQpAQEBVZ6T8u7dcvRVGFDa29uj1WpZvHgx/fv3p0aNGhWqx8zMjBEjRrB8+XJycnL0btelS5fYuHEjkydPrvCw/r0Lc7RaLcuXL6dFixb4+vrSt29fWrZsiZubG5MmTTL43uiCIAiPqvz8fCZOnMiYMWMYN24cW7dufeR3SnvSiIDyEbdw4UJsbGwYNmxYledFLI0kSQQHB7NhwwZSU1Or7DouLi4YGRkZLKC0tLTE1NSU48ePc+3aNUaNGlWpukaNGsXt27fZsWOH3u366quvcHJyYvDgwRUu26xZM86cOYNGoyE3N5dXXnmFkSNH4uTkxPr16zl+/DjHjh1jzJgxrFmzhubNm3Pq1Cm92ywIgvAoS05O5vnnn+eHH35g0aJFfPPNN1U6PUsomQgoH3F2dnasWrWKAwcOMH/+/IfWjiFDhqBWq/njjz+q7BpKpRJ3d3eDDXkX/jo9e/YsSqWSli1bVqouPz8/atSoodvLvbISEhJYsWIFEyZMqNT+6M2bNycjI4OwsDB69erFhg0b+PPPP9myZQt9+vShRYsWtGzZkjlz5nDhwgVq165Nx44dOXbsmF7tFgRBeFSFh4fTqlUrTp8+zc6dOxkzZszDblK1JQLKx0CnTp146623+OCDDzhz5sxDaYObmxtdunSp8pyUhkoddHdAqVKpqFevHqamppWqS5IkXSJkffzwww9IksS4ceMqVb5p06YAzJw5k127drFp0yZefvnlEs91cnJiz549BAQEEBwcTFZWVqXbLQiPCjs7OxISEvSup3D3Lzs7O73rEh6e7du306pVK4yNjTl+/DidOnV62E2q1kRA+ZiYPn06DRo0YMiQIQ9t0cWwYcM4ePBgle4pbajdcu4NKPXZ+xcKEiHrk4szJyeHhQsXMmzYMBwdHStVh52dHV5eXvzxxx9MnjyZLl263Pd8CwsLVq5cyY0bN/joo48qdU1BeJQEBgYSFhZGXl6eXvWEhIRgYWFB7dq1DdQy4UGSZZmvv/6a7t2789RTT3HkyBF8fCq2kYhgeCKgfEyYmpqydu1arl69yrRp0x5KG3r37o21tTWrV6+usmsYarecwoBSlmWDBJRBQUFcvny50gtz1qxZQ0JCApMmTdKrHUZGRiiVSj777LNyne/v78+7777LwoULdcneBeFxFRQUhFqt5vLlyyUeT8tN5WrKFS4lh3MzPZJ8bX6J56lUKgICAlAoxFfg40atVjN69GjefPNNpkyZwt9//42Njc3DbpaAyEP5WAkICGDu3Lm88cYbdO/evVyJWg3JwsKCfv36sWrVKj766KMq2SXAw8ODqKgoNBoNSqWy7AKlSEpKwsvLi5iYGFJSUggMDNSrXUFBQWg0GsLCwmjSpEmFymq1Wr788kt69eqlVw7JtLQ0bt68iSzLGBuXf+/YsWPHMnPmTFatWsUbb7xR6esLwsNW+D4OCQmhQYMGuk0jdtzYjipBRXJuUpHzlZIRXjZetHVrR1fv57AztQMKAsrGjRs/4NYL+kpISODll1/m2LFjrFixgmHDhj3sJgl3ET/PHjMTJkyga9euDB8+nMTExAd+/eDgYK5du8a///5bJfV7enqSn5/PrVu39KqnsIeycJha3x7KgIAAgErNo9y2bRthYWHl3maxNGvWrCE/P5+8vDwuXrxY7nIuLi689NJLLFq0CLExlvA4c3R0pGbNmqhUKkITQxi/eywf/Pseh6IPFQsmATRyPtdSr7I2bDXDtw1l0bnvychN58KFC3p/JggPlkqlomXLloSHh7Nv3z4RTD6CREBZgoyMDOLj48nPL3m45GFSKBQsX74ctVrNmDFjHniA0L59e7y9vassJ2VhcnN9h70LA8qYmBgAvLy89KrPxsYGBwcHoqKiKlx2/vz5tGjRgvbt21f6+rIss2jRIrp3744kSaUmOC/N2LFjuXjxIvv27at0GwThURDUKIhzZmd599A7xGQUvL+18v1TqsnIaGUt265vZdzOMZh7m1ergFKr1bJr1y5ef/11OnfuzJo1a8jIyHjYzSq3f/75h7Zt22JnZ8eJEydo06bNw26SUAIRUP4nMzOTRYsW0bhxY6ytrXFxccHKyoqhQ4dy9OjRh928Itzc3Pjxxx9Zv349K1aseKDXVigUDB06lN9//71KFgcVJjfXZ2GORqMhJSWlSFJbQ8yVqkwdZ8+eZc+ePUyZMkWvKQJRUVGoVCpGjBhBvXr1KhxQduzYEQ8Pjwpt9ygIj5p8bT41+tkjNygIILVoK1Rei5bUvFTaTm8F7lXRwkfPqlWr8PPz49lnn+WHH35gz549BAcH6zZAeJR31pJlmTlz5tC7d2+6du3KoUOHdN8RwqNHBJQU9Ia1bNmS8ePH4+3tzbJly5gzZw4ff/wxhw8fpk2bNnz66adV2ht49uxZxo4diyRJSJJEixYt+OKLL0od1n7ppZcYMWIEEydO5OrVq1XWrpIEBweTlpbG33//bfC67ezssLS01KuHMiUlBaDCu+JUhfnz5+Pl5VVqep/yKhy6b9q0Kc2aNatwwnJJkmjcuLFeK9UF4WFbePZbMmtk6Dd/WwJJKfF9+LdcTXmwn50PkizLvP/++wwbNowmTZpw6NAhtFotsiwTERHBxIkTWbx4MV27dtXtLPYoycnJYejQobz77rt8+OGH/PHHH1haWj7sZgn3Ue0DyitXrtCmTRsyMzNRqVRs2LCBESNG8M477/Duu+9y+fJlZs6cySeffML48eMNHlSq1WrdG37jxo0MHToUX19ffHx8+OCDD/Dw8Ch1L++vv/4aZ2dnhg4d+kCH5+vWrUvbtm2rJCelJEl6pw66dx/vhyUqKopff/2VN954Q+9dG1QqFdbW1nh5edG8eXPOnj1b4dQphsilKQgPy9HYI+yO3GWQuiSFhEbW8OWpL8jT6peC6FE1ffp0Zs2axbx58/j9999p166dLhD39PRkxowZ7Nmzh7CwMJ577jm9UzEZUmxsLJ06dWLdunX8+uuvfPrpp2JF/mOgWq/y1mg0DB06FDMzMw4ePIirq2uxcxQKBe+99x6Ojo6MGTOGjh07MmDAAINcPyMjg969e3Pw4EGWLl1KcHBwkcAjMTGRyZMnM3jwYOLi4pg8eXKR8oUpfNq3b8/s2bP58MMPDdKu8hg2bBjjxo0jNja2xOdNH/qmDro7oHzQvbd3+/bbb7GwsODVV1/Vuy6VSkVgYCCSJNG8eXNycnK4cOECjRo1KncdgYGBREdHk5ycjL29vd5tEoQHJTc/h+/OfMPlP65w8edLWHtY0embDrrjslbmxo5IbmyPJDMuC6WpEts6tvj1r4uD/53Xen52Plc3XCP5cgopl1PJy8hDniXx/bs/PIzbqjKnTp3is88+4+OPP77vYsDWrVuzbds2WrduzZw5cx7od0hpTp06xYsvvogsyxw8eJDmzZs/7CYJ5VStQ/4vv/ySY8eOsXLlyjKDotGjR9OvXz9ef/11vVcgF5oyZQpHjx5lx44djBw5slgvlqOjIytXruTtt99mypQp7N69u1gdbdu25f333+fTTz/l+PHjBmlXefTv3x9jY+NSe0/1oe9uOY9CD2V6ejqLFy9m9OjRBsmRFhISokuZ0rhxYxQKRYXnURYuQhDD3sLjZn/Ufm7F3OLyuqsozYqnE7uw8iKqxaFYe1kTMKI+Pi/WJjMmk8MfHCX5UoruPHW6mku/XyEjKgMbb2sAzsSffqJ6KbVaLSNHjqRhw4a8//77ZZ7fvHlz3n33XaZPn054ePgDaGHp/vjjD9q3b4+bmxsnTpwQweRjptoGlFlZWcyaNYvx48fTrl27cpVZuHAhGo2GBQsW6H39HTt2sGTJEubNm0fHjh1LPU+SJGbPns3TTz/NyJEjSUtLK3bOhx9+SNOmTRk8ePADW7lnZ2fHiy++yMqVKw0+DcDT0/OxH/JeunQpmZmZTJw4Ue+68vLyCAsL0wWEVlZW1K9fv8IBpZ+fH8bGxmLYW3jsbLr2DxdWXsTezw47H9six7QaLRHbbuDapiZN32yMV1dP6vbxoc2nLZE1MtEHYnTnmtqb8uyyznRZ8gwNhvkDkKPJ4Vjso7XwUh979+7l/PnzfPXVV+XOV/vBBx/g4ODAwoULq7h1JdNqtXzyySf079+fPn36sH//ftzc3B5KW4TKq7YB5W+//UZqamqxYeT7cXJyYtiwYSxduhS1Wl3pa8uyzJtvvsnTTz9dro3sFQoFy5YtIzExkS+//LLYcWNjY9asWUNMTIzeuQ4rYtiwYahUKs6dO2fQej08PLh16xa5ubmVKn/79m3MzMwwNzfH3NwcoMRAvCI0Gg0ZGRm6+u4nPz+fBQsW0L9/f10aJH1cuXIFtVpdJM1JZRbmmJiYUK9ePRFQCo+VdHU6p46cJvZwHAGv1i92XM6X0aq1mNqZFnncxNYEFKA0ufM1pzRWYmZf9DwJBWfjz1RN4x+CRYsW0aBBAzp06FD2yf8xNTVl5MiRrFq1iszMzCpsXXGZmZkMGDCATz/9lFmzZrFmzZpyfc4Kj55qG1AuWrSIbt26VXgv17Fjx5KQkMBff/1V6WsfOHCAsLAwPvjgg3KvVvT29mbIkCH8+OOPJS7A8fPz48svv2TJkiVs3Lix0m2riK5du+Li4mLwnJSFQVhlcj5C0X2869cv+AIKDQ3Vq03Xrl0jJydHl+D8ftavX8+NGzcMFtynp6cDRVetN2/enHPnzlX4h42joyOpqakGaZcgPAjht8MJ+SkUzy7u2HgVnz6iNFVi52fHzb1RRO2PJishm7SINM5+cx5jS2M8u94/zYyMlvDk8m8U8CiLjY1lw4YNuowhFTF69GjS0tL49ddfq6h1xd28eZP27duzdetWNmzYwLvvvlslO7AJD0a1DChPnz7N8ePHGTt2bIXL1q9fn44dO/LDD5WfxL1o0SL8/Px4+umnK1Ru7NixxMTEsGnTphKPjx49mp49e/Lqq68abJ7n/RgZGTF48GB+/vlng64QLMwzVtl5lPcGlEql8r7zBmVZRqPV3HfovrBXr6xkyLIsM3/+fDp16kTTpk0r0fryad68OWq1WvQ2Ck+8pUuWkhWfTb1XSt+2tOmbjbBys+TMgnPsHr2X/ZMOkXotjadmt8GypkWZ1yhMkP64W758OSYmJgQHB1e4bO3atenWrRuLFy+ugpYVd+TIEVq0aEFSUhKHDx/mxRdffCDXFapOtQwo161bh7OzM927d69U+ZEjR7J///5KbX1469Yt1q1bV6lfkE2aNKFVq1alBrOSJLF06VIkSeLVV199ILvoBAcHEx8fz/bt2w1Wp7t7Qcbhys6jvDugNDU1xc/Pr0hAqdao2XdzL1+d+pKxO0fR++8X6PPPC/T+uxejdoxk3sm57IjYTk5+jq6MSqWiRo0auLi43Pfa//77L8ePH6/yqQeNGjVCqVRWeB6lIDxObt++zZoFa6jX3xdTW9NSzzMyN8La0xrv571o/k5TgsYEIGu1nJh9ity0snvx87X332nncXHgwAGeeeYZbG1tyz65BC+99BInT56s8mTnq1evplOnTvj6+nLixAkaNmxYpdcTHoxqGVCeP3+epk2bVjo3YIsWLYDKrZb9888/kSSp0vuQjh49mh07dhAfH1/icWdnZ5YuXcrmzZsfyC/NRo0a0ahRI4PmpLSwsMDR0dEgPZRQ0KuoUqnI0+Tx68VfGLZtCF+emseBqH3EZMYg/7fbhozMraxbHIo+yHdnvyF462CWhywjJz8HlUpFUFBQmT8C5s+fT7169Sr9Y6W8LCwsCAgIqPA8SkF4nHzwwQdY21pTu7t3qedoNVqOfHwMIwsjgkYH4Nq6Jt7dvGj9SSsyb2VxdcO1Mq9jpCi+cvxxpFKp9ArOgoKCkGWZCxcuGLBVd2g0Gt555x2Cg4MZMmQIu3fvxsnJqUquJTx41TKgLAwOKsvX1xdTU9NKDTeeOXOGwMDASq9ALtwP+n4LYXr27MnYsWOZPHnyA0kDERwczN9//01ycrLB6tQnddC9AWVgYCDX064xce94frm4lsy8gknnmlL2/9XKBQFmjiaHDVfWM373WK5kXCnzNXP58mX+/vtvJk+e/ECS8DZr1kz0UApPrMuXL7NkyRKGjh5KdnI2WfFZZMVnocnTotXIZMVnoU5XkxSaRHpkBjVbOBcpb+VmibW7FUlhZX8uuVk9/iuKk5KSiImJ0eu7rXCOeFWkFktLS6N3797MmzePr776ip9++gkTExODX0d4eKpdQJmWlsaNGzf0etMZGRlRv379Sr3p9A1m69Spg7m5eZnXnjdvHh4eHgwePLjKd0B45ZVX0Gg0/P777warU5/UQfcGlA5B9gS9G0BsRiwyFZsGICNzO/s27q+54tDs/snAFyxYgKOjI0OHDq1UuyuqefPmqFQqcnJyyj5ZEB4z0dHRaLVa5r4/l91j9un+Ui6lkBmTye4x+7j0+xVyUwuGtOUStvXW5svI2vu/5yUU1LP3r4pbeKAKOzgK89VWhpWVFXXq1DF4QHnt2jXatm3LgQMH2Lx5M2+++aZYfPMEqnYBZeFqX32CusLyFX3TabVaQkND9XrDK5VKAgICyry2paUla9eu5dy5c3z66aeVvl551KxZk+eee86gq70N1UN5MSmMXdJOFEYKtJTwjVMOWrQgwRnbU5yMO1HiObdv32b58uWMHz++ylJe3Dsntnnz5uTl5VXodfgg5tUKgiEEBgby119/8ddff9H90260mNaM5tOaYu1hhbmTGc2nNcWzizuWbgX7O0cfKrqwJuVqKhkxGdjWvv/GAjJaGjk1rqrbeGBUKhXGxsbUq1dPr3oq8912P/v376dly5bk5uZy7NgxunXrZrC6HwXp6elERkY+8HRLj6JqF1CqVCqUSiX+/vr9Ig0MDCQkJKRCX9ARERFkZmYaJJgtz3B78+bN+eSTT5g9ezaHDh3S65plGTZsGEeOHOHy5csGqa+y+3lrtVqSk5NxcHAgJz+HL058jixrkRT6/RouLP/Vqfmk5RZPu7No0SJkWeb111/X6zolKUwXFBsbW+Txhg0bYmRkVKFh75iYGDFnSXgsODo60rt3b3r37s3E4Deo2coF11Y1MbExwcjMCNdWNbHxssHOxxbHRo5E7Y3mxJxTRGy7Qfgvlzj6yXGUJkrq9PIuUu/1LRFc+uMKkbsL0pLdPpXEnuV7mTFjxmOdUkulUuHv71/uZOalKe/3S3ksWbKELl260LhxY44dO6b39+6jIi8vj7Vr19K+fXtsbGzw8vLCysqKLl26sG7dOrTaynVePO6qZUDp6+uLmZmZXvUEBQWRkZHBjRs3KnTtwrL6CAwMJDQ0FI2m7JWJ06ZNo02bNgwdOlTv5N7388ILL2Bra2uwxTmenp6kpaXdt80nTpxgwoQJBAQEYGlpiaenJy+//DJarRYHBwfWhK0iMTuR25eSOL84hANTDrGp71Y29tlSap0b+2wp8e/yuqvIyGTmZ7Lk/J3FTr/99hutW7fmww8/RKvV8uKLL7Jnzx6DPAeFateujaWlZbFeAzMzM4KCgsq9MCcrK4srV8qeCyoIj5qO7p2wNrFBouQfhi3fbUa9Qb5kxGQSujyMa5sicPC3p92sNljVsipy7tUN1wn/+RI3thX8YL15OIpPPvqEDz/80KDzwB+0ixcv6vLu6qNBgwbExsbqFVzn5+czceJExowZw5gxY9i6detD3bnMkNLS0ujevTtDhgzBxMSEZcuW8fXXX7N48WKys7Pp27cv/fr1q/KV8o+iyi1zfozdvn27zH27y6NwW6iEhAS8vb3LVUalUmFvb6/39YOCgsjOzubatWv4+vre91ylUsnq1atp1KgREydOZMWKFXpduzRmZmYMGDCA1atX8+mnn+q9KKUwufnNmzdLTSb++eef8++//9KvXz8aNmxIXFwcX3/9NQDxKfHsvrYDGZn40/FE7rqJjZc1Fi4WZMbcf2jCsZEjHp1qFXnMtk7BsJlW1nIgej9DGgSz6ItFTJ8+nWbNmiHLMm+//TYJCQlER0frde/3UigUpU5zaNqsKWcunub0rVPk5OegUChwMnfGy8YLI0XRt3dYWBiyLIuAUnjsmBmZMaHx/5h9fCZtZ7QudlxpqsSvvy9+/e//eQjQZUlB/l+FpMDN0o2vn/4OY6V+vXqPArVajaWlpd71FNZR2Z3KkpOTGTBgAHv37uWHH36oVL7nR1VqaiqdOnXi+vXr7Nmzp1gu6dGjR/P3338zaNAgunbtys6dO/XuvHqcVLuAEjDIZODK1HHz5k1q166t9/Xr1KkDFORpLCughIIerm+//Zbhw4fTo0cP+vXrp9f1SxMcHMySJUs4cOAAnTp10quuwuTmkZGRpQaUkydP5ueffy6yUrBBgwa8/PLLrPh9BW7jC3JGenfzom4fH5SmSlRLQssMKK3cLHG/J6C8m0JSsPDvb5k5fTbz5s1j2bJl9OzZk88++6yit1luQUFBnD59Wvfv66nX2Hp9C8nP36ZWD1c+OfJRkfOVkhI/ez+61e5OO7enMFGa6ALSBg0aVLodt27dYsWKFSxcuJD4+Hjeeecdhg8fXuEdpwShotq4taVjrU7si9qHvh/hEhIKScHk5lOfiGDyUREeHs4LL7xAYmIiO3bsqPDmHY+6SZMmcfXqVf79999Sf5i/+OKL7N69m06dOvHRRx8xd+7cB9zKh6faDXk/bIZIJ1OZOoKDg+nbty9jxowxeA9aobZt2+Lj42OQYW9XV1cUCsV9F+a0bdu2WNoJK6uC4a3IiBu6Fd2mdqYoTSuWZ06Tq0GjLj2t0M9LfqFmzZr4+/sTGhpaJXMn7xYYGMiFCxe4lXGLTw5/xBt7/8fOGzvIU5SctFkjawhPCuerU/MZsX0Yh6IPolKpqFOnju45qojc3FzGjh2Lu7s7n3zyCTdv3iQ3N5cFCxbg4+ND3759H+v5Z8KjLycnhz3T95F4OkGvegqCSSXvt/qQunZl/yAXymfHjh20atUKpVLJ8ePHn7hgcsuWLSxfvpyvvvqqzFGeNm3aMH36dObNm8eRI0ceUAsfPhFQVhOSJLFo0SLMzc0ZMWJElUwaliSJ4OBg/vjjD71XvBkZGeHm5lbhhTm3b98u+B9Wle/CuLk3ii2DtrNlwHb2/u8AUQeKB+BRZ6Jp2qwpkyZNQqlU0r17d1xdXfnuu+8qfd37CQoKwqGZPeN3j+Vswhmg9DyahQpXtWeo05l7Yg4X3S4Q2LjiGQZSU1N5/vnnWbFiBbNnzyYmJgZZlpFlmZiYGJYsWcLu3bvp0KFDlf1YEaq3nJwcXnrpJfbs2sPH7afTq84LACgq+BWmQIG9mT0z2s2kmUvzqmhqtSPLMt988w3PP/887dq148iRI/j4+DzsZhncRx99xNNPP83IkSPLdf7UqVNp3LhxlWdZeZSIgLIaqVGjBitWrGDnzp18++23VXKNoUOHkpGRwYYNG/Suy9PTs8Kpgwq3gHRtd/8tEktj72+P/yt+tJjWjKAxAUgKOPPVOSK23Vl8pc7IQ52m5sDBA1y6dIkBAwbw22+/0bhxY/73v/9VyQ5FqS4pNJvaBLVWrUu8Xl6FPbWK2hLmfUyLbClZZllZ5pVXXuHMmTPs3LmTqVOnYm9/Jx+npaUlr732GocOHSI5OZnu3bujVpe91Z0glFdhMLl37142btxIt67dGNVwDLOemkNNq4L56Erp/iMQhUPcz3l34/vOiwlwrHzqNuEOtVrNmDFjeOONN5g8eTL//PNPpbd9fJSdPHmSU6dOMWXKlHJPWVMqlbzxxhts376da9fK3q3pSSACymrm2Wef5Y033uCdd94xWGqIu9WuXZsOHToYJCdlRVMHXbx4kV9//RVTCxM8nnav1DWfmt2GOr1qU7OlC97dvOgw7ymsPa24uCYcTW5Bj6AmJx+A9NR0HBwcWLFiBf3792fz5s00aNCAGTNmVOrapTkTf4YVV5YV/EOPuWOSQiLbKuu/VErlS3e1fPlytmzZwpo1a3S7NJUkICCAv//+mwsXLhj8/oXqKycnhz59+uiCyS5duuiOBToG8UPnxXzWbhZt3NpiZ2pXrLxSUlLbtg6v1B/C8udWMq7xeCyMLR7gHTy5EhMTefbZZ1m5ciXLly/niy++QKl8MrawvNcPP/yAp6dnhXNo9u/fHzs7uweyDfKjoFouyqnuZs+eza5duxg8eDDHjx/H1NTUoPUPGzaM1157jejoaGrVKn1xS1k8PT3LnWMxLi6OHj16YGJiQq36tZCUhtmFQWGswLu7N6pFIaRcTaVGAwcUJv99aErwzjvv6PK+KRQKBgwYwMcff0xkZKRuYdH9ZGRk8MUXX3Ds2DGOHz9OcnIyy5cvZ/jw4QBk5mWy4NR8Infc5Ob+aDKiM8jPzMfUwRTHQAf8Bvhi4Vz0C7K0tEj+Q+rh+7IPJ24dZ+/NPTzj2Zldu3Yxc+ZMVCoV+fn5+Pn58b///Y+hQ4eSmprKpEmTdIu5ytKkSRM++OADPvvsM4KDg6lbt26ZZQShNIXB5L59+4oFk4UkSaKRUyMaOTUCIDknmcTsRDSyBnMjM9ysamGsqB6LbiRJKlcqubIU1nG/nriQkBB69epFVlYWe/bsoV27dnpf91GVkpLCL7/8wvvvv1/hgNnc3Jzhw4ezbNkypk+fbvDv2kdNteuhtLa2JjExUe96Cut4HLv3zc3NWbt2LRcvXuSDDz4weP19+/bFzMyMNWvW6FWPh4cHUVFRZc73LJzjl5KSwjPPPIODuWHznZk7FqR9yMso2MLSxMoYhYkCSSExZsyYIuc6OxfsJ1zefHaJiYlMnz6dsLAwGjVqVOz4LxfXkpqbSsr1VCxczKnbpw5BYwJw71iL+NMJHHzrMDlJxYewHRs50uSNRkX+7t7rePH5H/hj/R907doVtVrNJ598wsyZMzE3Nyc4OJivvvqK1atXk5mZycyZM8v3RFEQYNva2labX+RC1cjJyaF3797s37+fTZs2lRhMlsTezB5fe1/8HfzxsvGuNsEkFPwAN8TQ6tWrV7G0tNRtqHCvjRs30qZNG2xtbTl+/PgTHUxCwTSq7OxsRowYUanyI0eOJDExkYMHDxq4ZY+eahdQBgYGcvHiRb33t1apVJiZmVVo8rGZmRkZGRl6XRfQ1aFPfqtGjRoxc+ZM5s+fb/BE3DY2NvTp04dVq1bptdWfp6cnubm5JCSUvqozJyeHXr16cenSJTZt2lSQ1NzaAVdL/XONFsq6lQWAiU3BinJJIWHtaQ1aim2zGBNTsP1beXejcXV1JTY2lhs3bvDFF18UvW5eFtsjtqFFS8MxgTSZ2AifF+vg2cUD/1f8aPlBC9Rpam7uLb4QpjD10d1/1p7WuuPZ+dnM/moWrq6u7NmzhwkTJjB+/Hh2796Nj48PK1asYNGiRfTu3VuXc7U8zMzMGDFiBMuXLxd7jAuVUhhMHjhwgI0bN9K5c+eH3aTHQuGWifpur6pSqQgICCiWTUSWZT7//HNefPFFnn32WQ4dOoSXl5de13ocnD9/Hjc3twp9Dt6tQYMGmJubG3x/9EdRtQsog4KCUKvVpW4RmJ2fTVR6FJFpN4jPii91AYRKpaJBgwYV6gKvX78+ly9frnTC2EKFcx/1yScIBXkcO3XqxLBhwwy+Q0RwcDAXLlwo9y4uJbk7uXlJNBoNAwYM4MiRI/zxxx+0adNGt493Q6dGZU7Uv1duavH/LvnZ+VzbGIGJjQl2Pv/1RssSNl7WyLJcZK5oTk4Oa9eupUGDBuX+8DE1NaVmzZolHjsYfYBcTemvFQungmA2P7PkH0f3S30kIRGdGIO9vX2RYRgjIyMcHR3Jz88nNDS0UkmJR48eze3bt/nzzz8rXFao3rKzs3nxxRc5cOAAmzZtEsFkBQQFBZGamlos00KGOp2z8WdYd+kPfjy/hMXnfmDNhdX8G32IW5m3itWjUqmKpcXJyckhODiYadOm8cEHH/Dnn39WKv3Y46ik56MilEolDRo0qJI1C4+aajeHsvCFURgQAlxOvsT2iG2oEs8Tm1l0v2RTpSl17erS2rUtnT07Y2VirSsfGFixlYJBQUFoNBrCw8Np2LBhpe9BpVJRq1atIqttK0OhULBy5UqCgoIYN24cv/zyi0GSvgN06dIFV1dXVq1aRfPmlUvPURhQRkZGlljHlClT+Oeff+jVqxdJSUmsWbOG69evY2VlRc6xXDQuBcFUVnw2UfsLPmRTrhbkSrz0xxUAzJ3MdbviRGy9QdyxW7i0cMHc0Yzc5Fwid0eRnZhNkzcaoTD+7/eXJNO8fgts4mwZP348ly5dwtPTk9WrV3Pjxg02btxYqfu9V0iiCoWkKPKjRp2mRtbKZCdmc+n3gntwbFh8aOrm3qiClekyWLlb4dvPB/cOd+azysjY1Lci9K9QPvzwQ4YNG4YkSfz888+cPHmS9u3bo1areeaZZyrcbj8/Pzp37sxPP/3EkCFDKnHnQnWUnZ1N7969OXjwIJs2barUa686K/w+UqlUuLu7E5IYwqZr/3A09ghaWYtCUvy3daUEyLq0Y3VsfehVpxft3TuikBVcuHCBYcOG6eqNi4ujd+/enDt3jl9++YWBAwc+hLt7eFQqFX379tWrjsLe4yddtQsoa9SogaurKyEhITR7rinfn/2OKylXUEhKtCXk9cvV5BJ6O5QLty+w6sIKetbpxUC/QYSGhlZ4x5m73/CFAWVOfg6Hog8SkqgiPDmchOx4tFotJkoTPG288LP3o2XNVgQ5NtQFe/r+Yrqbh4cHixYtYtCgQfTq1YvBgwcbpF6lUsmQIUNYtmwZ8+bNK5aAvDwcHR0xMzMrtYfy7NmzQMGcnruDuJiYGLZv387kvW9yNfUKWfFZhP98qUjZwn/XCHDQBZQO/vYkXUwhctdN1OlqjEyV2Pna0XhCEI4NHQGQZImMW5nMCf4Yrw+9efvtt1m2bBmZmZk0btyYzZs389xzz1X4XktyMelisR7yna/tQZtX8JixtTGBrzXAqXHR4XV7f3vc2tbEwsWCnKQcIrbe4MxX58jPyse7250hKt/+damt8WHmzJm6ldkWFhasW7eOadOm0a1bt0on4n/++ef56KOP0Gq1BknmLzzZCnsmDx06JILJSvLy8sLKyopTF05x1uE0R2IPo5SUus+Q0kbbrqde4+szC/jt0m+87NiP3Nxc3ffL6dOnefHFF9FqtRw4cIAWLVo8sPt5FKSnpxMREaH3921QUBC///77E/95WO0CSoDAoEDOKc8yZd8ZXZBWUjB5NxmZPG0eG678xf4b+zFyNarwi8zW1hZPT09UKhWZeZn8evFntkVsI1eTg1JSFklUnZ+fz8WkMC4nX+Kfq3/jaulK/3oDecajMyEhIQbdPnHgwIFs2rSJ119/naeeespg82KCg4P54osv2Lp1Ky+++GKFy0uSdN/UQfv27Svyb1mWMTU1ZcGCBbz++utcS7nK5P1v4hhYg15/dS/zek6NnYoFZ/eSJRn5kESLsS0BqmxvdICE7Phij7X6sAUatYaMqAyiDsSQn1P8dfvU7DZF/u3Z2YMDUw9xcU04Hk+763YNUhgrqOFpR9++fXnppZfQaDQsWbKEIUOGkJWVxaRJkyrd9qCgILKysrh+/foTmeRYMJy7g8nNmzc/cTusPCgKhYLGXRtzwuMoUlzB91pZmx/AnTy18Vm3WBjxDT696xAUFMQff/zBsGHDCAwMZMOGDZWeQ/g4Cw0NBajwaOS9AgMDq8Xn4ZMbKpdCI2twfNkBuZEGGblSSaKT1Um0m9kaI8+Kx+OBgYGEJKt4fdcYNl77h1xNjq5dpbUXIDYzlq9Pf8UHB98jLj1O7xf4vb777jvs7OwIDg42SOoJKLjXpk2b6pWTsiLJzTMzM8nLy8PBoWCVdx07HwbUG1Tpa5fk+tYbvNn/TYPWWRJZLvm16RhUA5dmzvi8WIfmU5tw6ffLXN8Scd+6ClMf5WXm64b8AUKWXODY7mP8+uuvDBw4kMGDB7Nr1y4cHBzQarV6/Sq/uzdeEEqTnZ3NCy+8wKFDh9iyZYsIJvVwMSkM+1ds0RppK/y9Bv/1YErQYJg/H/75Pv3799ettK+OwSQUfH4pFArq16+vVz13T7V7klW7gPKn8z+S4Zyu11xBWZJRGCn44fJ3RKRGVKhsrc5uSN1kUnJTKvWmD0kKocO8drjUq9xOMKWxs7Nj1apVHDx4kHnz5hms3mHDhrFp06Y7WyJWkIeHR7kDyqSkJABdQAkwoN5AnvbQf/hMQkKKUqA+kEfPnj31rq/M60kSRor7/2CxdLXEtrYN0ftjyqzv3tRH2jwtkbtv0urplkWGYIyNjfH39wfA17fy+xy7urri4ODwxH+ACpWXlZXFCy+8wOHDh9myZQudOnV62E16bKXkpjD9yCdIyoIsFPqKcY1myteTWbt2bbFMFtVJUlISdnZ2ej8HhQsvDZGy8FFWrQLKM/Gn2XzdMAsmJIVEvpzP/FNfkK/NL1eZvTf3EFcnBkmSdMMMFSWjxcTahNW3VxCfVXxIVB8dO3bk7bff5sMPP+TMmTMGqXPQoEHIssyvv/5aqfIV2S2npIBSISl4o+kkXvTpXfDvCr7kpf+2pmlu14JNk7cw6Y1JD2wOTC2rsnf70ai15GWV/fq7N/WROl2NrJGxNCq+UjM+vuB1ZWlpWZHmFiFJEkFBQdViZaNQcSKY1N/w4cORJAlJkrA3s+eXnr/zT5/NbOyzhY19tpB9+07aLm2elst/XmHPhP1s7r+N7cN3cWzGCbITs3Xn5GfnE/7LJY5OP862oTuZ/8aXLF4q8skaYqGqoRa7PuqemIDyxIkTTJgwgYCAACwtLfH09KR///5culSw+EKtUbPg9FdISMhamYhtN9g/6SCbB2xj29CdHP7wGKnX04rUeemPKxyfdZLtw3exsc8Wwn8turBDK2u5kRbBhivr2bVrF08//TSOjo7Y2dnRsmVLVq9erTs3LjOWhWcMs3+2pJTIyMvgq1PzK9XLeT/Tp08nICCAwYMHk52dXXaBMjg5OdG9e/dKD3t7enoSGxtbrryhhb2gdweUUBBUvho0ihntZumSnpcVWBYGktYm1rzf6kPi1yVib2NPcHBwZW6jUurZ1yuYVK/Ros4ofv/Jl1JIv5GOXd07yfXLm/rI1NYUY0sjDmw7UGTv7YyMDC5evIiVlZXev8oDAwNFD6VQTGEweeTIEbZs2ULHjh0fdpMeS2PGjGH16tXM/n6WbvOCxm80QmmqxMrDCvMaBaMS2nwtx2ae4PKfV3Fu4kTQ6ADq9qmD0lRJ/l0/RtXpai79foWMqAxsvAuymRyPO/ZQ7k14PD0xi3I+//xz/v33X/r160fDhg2Ji4vju+++o2nTphw9epQEm3iScwp6sM5+d57oAzG4d6qFd3dvNDn5pF5PQ33Pl3H4z5cwtTfFto4NCWdK76r+bs137P/sIO7u7jg7O3Pjxg1OnDhBcHAwiYmJvPnmm3x9egF5mjxu7o0i9mgcqdfSyMvIw8LZHLf2bvi8WBulSfG8ibkpuVz85RLxJ+NRp+dhameKY8MaNJ7QkNDbIeyI2E632s/z66+/MnfuXC5cuIC1tTUvvPACn3/+OY6OjhV6Hk1MTFi7di3NmjXjnXfe4ZtvvqlQ+ZIEBwfTt29fwsLCKjwXxcPDA1mWiY6Oxtvb+77nltRDebeGTo1Y/OxPHIs9ytoza4hS39QND0nc6TWWZRkPS09eqvcST7l3IDM1k2XLljFlyhQsLAy/D/B3331HSkqKLin6xo0biYqKIjItkpxGBb0Mu0btwa2dK9YeVijNjEi/kc7NPVEYWRjh2+/OFoflTX2kVCppM6gNB346SOvWrXVzZ5cuXUpOTg7du5e9iKks/v7+YsccoYisrCx69erF0aNHRTCppzZt2tCmTRvmnpiDl5snGlnD7QtJaHI1uHe4M+fx2sbr3A5Not3MNtj72ZVan6m9Kc8u64yZvSkpV1I4+NZhQhJV5GpyMVU+2VsGCobxxASUkydP5ueffy6SnmbAgAEEBQUxZ84cXF5zKkjm/G8MUXujaf5OU1xbl5xQulDnxZ2wcLYgN03NjmG7Sj0v9J8w7B3tiYqKwtPTk5YtW7Jv3z6cnJxYsWIF3YKfI/R2CPk5+Zz99jz2fnZ4PeeJqa0JyeEphP96icTzibSZ3qpI13h2YjaH3j0CgNdznpg5mJGTnEvK5RTdOb+G/8zVrdeYMH4CnTt35ssvvyQqKoqvv/6akydPcuzYsQrvqNOgQQPmzp3LxIkT6d69O926datQ+Xv17NkTe3t7Vq1axezZsytUtnA/7Js3b5YroFQoFNjY2JR6jrHCmKdqtWfvsn389u0fHAw9QExWDHkaNUYKYyzyLOjU8GlWL1tNZ69nAfhy8ZdoNBrGjx9fobaX17x587hx44bu3+vXr2f9+vUA9Fn+AmpLNZ5dPLgdcpvYI3Fo1BrM7M2o1d4V3351i+zlXZ7URwBatMz6ZBY3no7k66+/5tNPPyU3N5eGDRvi7Oxc4jaQFWViYmKwBV7C468wmDx27Bhbt26lQ4cOD7tJj73MvEwOx/yrG6mKPhgDEtT6L6CUtTLXN0Xg2qom9n52aDVatPkyRqbFOy+UxkqU9kUfz9PmcSz2KB3cReAvlO2JCSjbtm1b7DFfX18CAgIIvRBKWkrBG+zqP9ex87XFtXVNZK2MRq3ByKzkp+HuL+r70WTnY2pjSqwqlpo1a3Ly5ElatGiBtbU15ubmbLm+uWDo0khLu9ltcPC/k5Dcq6snFs7mhP96mcTzt3FqdOdL//wPISiUEu3nttPNfbtXYnoi7773Lh06dGDnzp26gLRt27b06tWLH3/8kf/973/luo+7jR8/ns2bNzNixAhUKlWFezrvZmpqysCBA1mzZg0zZsyo0O5CZe2Wc7ekpCTs7e3LNcdx9+7ddGjbgQCnQAIoumLe19uXAwcOMGjQINRqNd9++y1DhgzBxcWwC6EKRURElHrs7yt/sTTkJwJfLd+uSOVJfaRAgZOFE61cW9Pulad45ZVXihx/ktNaCA9HVlYWPXv25Pjx42zZskUEkwZyNeXKnTyT+Vpi/o3Fvp697rsr/WYGOUm5WHtZc+57FVF7o9Hma7H2sibw1QY4BpW8X3chCQXhSeEioBTK5YmZQ1kSWZa5desWFrYFc8HysvJIuZyCXV07wtaEs3XwDrYO2sHusXuJ+Te2jNpKVyPAgbhrt1i4cCFXrlwhKioKgOvXrzP1rakciT2MRtagMFYUCSYL1WxdEKhkRN3Z5zs9KoP40wn49K6DiY0JGrUGbX7x+ZKZNzNJT01nwIABRXo3e/bsiZWVVaUXwygUCpYtW0ZeXh6jRo3Se3/YYcOGERUVxd69eytUzsrKCnt7+3ItzCncdrEsWVlZHDlypNTkyR06dODgwYMA/PLLL8TGxjJ58uQKtdtQevq8QF27uhVeTHQ/WrRMaja1zFXkgmAImZmZumBS9Ewa1tWUK7rPhoQzCeSl5+He8c5wd2ZsJvDfsHfIbRqOC6Tx/xqiVWs4Nv0EaRFpJdZbSEbLpeTwqrsB4YnyRAeUa9euJTo6mpbdCpJQZ8VlgQzRh2K4ufsmDYb502RSI0xsTDg1/wzxpxMqdR3f/nWp1c6NmTNn4uvrS58+fQCYMGECrbu2Qq1R37d8bnLB3M27eyETzxfM2TS1NeXIR8fYMmA7WwZs5+j0E2TFZ+nOy1MXTKouaQGFubk5Z86cQaut3MIdNzc3lixZwoYNG1i+fHml6ijUsmVL/Pz8WLVqVYXLljd1UHkDysOHD6NWq0vdJ7hDhw6EhoaSkJDA/Pnz6d69u977pleWUlIy0HUweTn5yFr9gvpC/f0G0KDGw7kfoXq5N5hs3779w27SEyU5N1nXkRB9MAbJSMKtravueOHGB5psDW2mt8LjGXc8nnGn9aetkJG58te1Mq+RlFO5lG9PAhsbG1JTU4ssXKyMwnRBdnZ2BmjVo+uJDSgvXrzI+PHjadOmDe17t0eBQvfmykvPo8W7zfHu5oV7h1q0+bQVJtYmuv2dK0phrMDCzQIPDw+eeeYZWrduDcD333/PuzPeK7P8lQ3XMLIwwrnpnaHKzJiCoPHcDyokIwXNpjah/tB6JIUlceTj4+TnFtyLpZslSHDo0KEidYaHh5OQkEB2djbJycmVui+Al156iZEjRzJx4kSuXr1a6XokSWLYsGGsW7eO9PT0CpUtb+qg8gaUe/bswdnZudQgsbAHZeHChahUKqZMmVKh9hpSWFgYL3V+iRuLIzE1MkMh6feW7V67B4PrDzVQ6wShdIXB5IkTJ9i2bZsIJquALMtISORn5xN3PB7nxk5FOiaUJgWfF/b+9pg73ul0sHAyx8HfgeTwsr8bDJ1J5HESGBhIfn4+4eH69dIWZrsw9IYkj5onMqCMi4ujR48e2Nra8ueff2JmYoYWre7NZeFiXmS1m5G5ES7NnUm5koJWU4lk40tCuXXiFl5eXqSnp+vy+Gm1Wv5Y8sd9y17+8wqJ525Tf2g9jC2NdY/n5xT0PJrZmdLqg+a4tXPFp3cdGr0eRFZcFtEHClYEm9qY4NbOlVWrVjF//nyuXbvGwYMHGTBgAMbGBfXpm/5nwYIFuLi4MGTIEPLzy5dzsyRDhgwhOztbt+CkvMq7W055A8rdu3fzzDPPlJobzMPDAy8vL1asWEHjxo0f2u4dp06dokOHDtjb27Pz51182ekrPKw8KlyPQlJiJBnxauBrjGk4rtrkRBMensJg8uTJk2zbto2nnnrqYTfpiWRhbImMTNyxW2hyNbrFOIXMHAoWZJraFZ+Db2pnotvo4H6yUrL466+/uHTpUrVbZFfWjl+5+TlEpt3gasoVbqZHkqcp+flUqVSYmppSt27dEo8/KZ64SVSpqak8//zzpKSkcPDgQdzc3LgVfwu4681lWzwFgqmtCXK+jCZHg8Ky/HF2wY4jUTTu34gDPx8A0C3K6dy5M3v27EGbp9Wla7lb9KEYLv58Cc8u7nh3K7p/dmEKIbd2rkV2PnBr68qZr8+RfDEZr2cLgouGYwMxsjdl6tSpTJ06FSgI3nx8fFi/fj1WVsWTV1eEtbU1a9as4amnnmLWrFl89NFHlarH09OTp59+mpUrVzJs2LByl6tID2VZq5NTU1M5efIkr7322n3PCwoKYtOmTcyYMeOhBGAHDhygZ8+e1K9fn61bt+oC5a+e/oZ1l//kr8vryMrPKpLu6F4KSYFW1tLQsSGjG47F3brsROlQkFJI3yEeALVaXaEFWEL5xcXFsW/fPk6ePEm3bt3o0KFDkQwXD1NmZiY9evTg1KlTbN26VQSTVcjbxhuNrCHqQAxKMyUuLYsuHLTxskYykshJyilWNicpp9TFnoVkrUzEiRu89OVLQMECy/r16xMQEFDkr3bt2g9sw4cHyd7eHnd39yIBZVT6TbZFbOX0rVNEZ0QX+fxVSArcrTxoUbMFz3l3o6ZlwfQDlUpF/fr1MTJ64kKuIp6ou8vJyaFXr15cunSJXbt26YY0fewKVq2aOZhham9KdklvruRcFCYKjMwr9pQU7jhib1J8sY1Go0Gr1ZY49y3hbAJnvz6PSzNngsYW7wY3cygIek3siga/klLCxNqYvMw7v4Ts7exZ+/ev3Lx5k4iICLy8vPDy8qJt27Y4OTkZZN5GmzZt+OCDD5g+fTrPPfccrVq1qlQ9wcHBDB8+nBs3buDl5VV2AQoC0eTkZDIzM++7e0t5eigPHDiAVqstdf5kocJpAobIx1hRW7Zs4eWXX6Zt27Zs2LABa2tr3TEjhRED6g2kd90+HIo6wNHYo1xKDic5987QlUatwd+lPkGOQXT1eg43q1oVur6Pj4/eQzxQMO1Cn+0bheJCQ0P59NNP+euvv3SjBfPnz6dmzZqMGjWKd99996FulZeZmUn37t05ffo027Zto127dg+tLU+qzMxMdu/ezebNm9l5eCe+U+uQeD6RWu3diqUDMjI3wrmpM/En40mPysDavaBzIf1mBskXU/Dqev8RD0kh8eHrH9FsanNCQ0OL/G3cuJG0tIJFPRYWFiUGmp6eno99oFm441dcZhw/nP2OMwlndD/W76WVtUSm3yAq4ybrLv9Ja9c2jGk4jpCQEN1+3k+yJyag1Gg0DBgwgCNHjvD333/Tpk0b3TEbExs8rb24mR6JWztXrm+KIOFsgi69Sm6amrjjt3AMqlHhfVALdxy5dPAyarW6SC/B2bNn8avnh/KeN3nypRROfH4a27q2NJvaBIWy+BvO9r8dTXJuFw1+tXla1Gl5ul+WEhK+dn5IkoSnp6cub2NKSgqnTp3i5ZdfrtD93M8HH3zAtm3bGDJkCGfOnKlUz+fLL7/M66+/zpo1a3j//ffLVebu1EGF+0yXpDwB5e7du/Hy8qJ27dqlnhMbG8vx48eBgh2YnnvuuXK10xB+/fVXhg4dSo8ePfj1119LzSFqqjSls9ezulyZmXmZ5ORns23rdoYOGsqKuNU4Od0/fVBpgoKC+P333yt9D4VUKtUTP2foQdq9ezd9+vTBxcWF+fPn07dvX2xsbLhy5Qo//fQT8+bNY/fu3fzzzz/UqHH/dDBVISMjgx49eohgsgpcvXqVLVu2sHnzZvbt20dubi5169blhR4vcPzwMWSNXGy4u1D9IX4knk/kyEfHqN3DG4DrmyMwtjKmbt+iQ7DXt0SQl5mv69G8dSKew6uOcNb4HP/73//o0qWL7tzCDSfuDTTXr19PRkZBxhJLS8tiQWZAQADu7u6PzdSbwMBAdkXuZMLuceTLBT/iyppXWnj8eNwxziWcJd7yFi8FvVTlbX3YnpiAcsqUKfzzzz/06tWLpKQk1qxZU+R4z3a9+P7cd/i+7EPMv7GcnHuGOi94Y2RhzI3tkWjztfgPrlekzM190WQnZKP5bwHM7dBk3cId9461sHA2R1JK+PTy4eKv4bodRwrzCiYnJ/Ptt99yyOoAMf91jaffzODYjBNYOJnT8v3mxYLNQjUCHTCxNSH6QAy+fX10Q+A390Yha2VdvkpJkmjgGFCs/Lvvvkt+fj6TJk2q/JN6D2NjY9asWUPjxo2ZMmVKpXZBsbKy4uWXX2bVqlW899575fpQKQwoIyMjSw0os7OzycnJKTOg3LNnz33nT0LBzjWmpqbY2tpy4MCBBxZQLl68mHHjxjFkyBCWLVtWoeERS2NLLI0tqe9dv2CYKiKi0gFlYGAgX3zxBWlpabok8VpZy8WkMMKTwrmaeoXknIIeURsTG2rb1sHX3o8gxyBdKiJZllGpVLz55puVaoNQ1N69e3n++ed55pln+PPPP4v8mGvcuDHfffcdwcHB9OjRg44dO3L8+PEq2dWpNBkZGXTv3p0zZ86wffv2EvMCC+WnVqs5dOgQmzdvZvPmzYSHh2NsbEzHjh2ZM2cOPXr00PX+BzQLwMTWBKeGJecKtvawpu2M1oStvsjlP68gSRI1gmrQYJi/bnvGQlc3XCc74c6c+9ijccw+WrAZxZAhQ7C1vbPNqyRJuLu74+7uXuQzUpZlIiMjiwSZKpWK3377jaysgsWmNjY2JQaarq6uj1ygqWgm4d7WDbW24tOAtLKW7PxsAibURzZ68hc3SbK+CQYfEZ06dWL//v2lHs/Ky+LV7cPJzMskIy6TCyvCSFTdRpuvxaGePfWH1sPO165ImcMfHOV2aFKJ9bX5rBWOgTVAhut/3+Dq79cwMTEhJyeH/Px8NBoNrVq1omfPnlxMCiOpWSIoYO/EA+Qk5VB/cD3dnM5CFjUtiuSpvLk3irPfnMeuri3unWqRnZDNtc0R2PvZ0XZ6aySlhIREQ1UTrodfp1WrVhgZGbFhwwZ27NjBjBkzyt0LWBE//vgjo0eP5u+//+aFF16ocPndu3fTpUsXjhw5olsRfz9qtRozMzN+/PFHXn311RLPiY6Oxt3dnc2bN5c6TB0fH4+LiwurV69myJAhJZ6TmZmJh4cHwcHBREdHExcXp8tJWZU+//xzpk2bxoQJE/j6668rPUyUnJyMg4MDv/32G/37969UHWfOnKFp06YcPnyYJi2asPX6FjZd20hCdjwSCiTpzi9wBQr47982JrZ0r92dnj4vkHk7Ezc3N/766y969+5dqXYIBdLS0ggKCqJ27drs3LlTt9iuJBcuXKBZs2aMGTOGBQsWPJD2FQaTZ8+eZdu2bSKYrKTY2Fi2bt1aMJS9cyfp6em4ubnRvXt3evToQefOnYtMfymUp81j4p4JxGbGGHRFtkJS8FWnr6ltW8cg9Wm1Wm7cuEFISEiRYDMsLIycnIIeUXt7+xIDTWdn54cSaG69voUfzi00WH1vt5jGU7We3GwHT0xAWR6Hog8y98Qcg9VXuOPIh4GfsGDeAr766qtSz+32Y1fytGp2j9lX6jnuT9eiycSii0qiD8ZwZf1VMqIzMbY0wrWtK/WH1MPI3AgFClq6tqJRYhOmT59OWFgYGo2Ghg0bMnnyZPr162eoWy1ClmV69+7N4cOHUalU1Kx5/y0s76XRaPD29qZXr158//335Srj5ubGqFGj+PTTT0s8rlKpaNiw4X2D1N9//50BAwYQHR2Nm1vJw0MLFy5k4sSJXLlyhU2bNjF16lRSU1MrvH1lecmyzHvvvcecOXP48MMP+fTTT/X64JRlGTs7O9577z3eeeedStWRk5ODlZUVn/34GVfdLpGQnVDqwp97KVBgaWJFezowrsfrXL58+Ylf2VjVxo8fz8qVK1GpVPedqlHoq6++YvLkyRw6dKjKh53vDia3b99eZKqRcH9arZYTJ07oeiFPnz6NJEm0bt2aHj160L17dxo3blyuz4PLyZeZun9Sud+n5THIfzCD/F8p+0Q9aTQarl+/XizQvHjxom5xYI0aNQgMDCwWaOqzg1tZYjKi+d+e8eRpy14JXx4SEmZG5vzQZTEOZmVnI3kcVauAEmDeibkcjD5gkDeeQlIw+6nPqf9fkuiYmBjmzJnD4sWLsbGx4e2332b8+PFYWFiw5domFp3/Qe9rFpJlGTlPZozT6/R6upfB6i2v+Ph4goKCaN68OZs2bapwEPTee++xaNEiYmNjMTUtvur+Xq1bt6ZBgwYsW7asxOP79++nU6dOhIeH4+fnV+I5Y8aM4cCBA4SFhZV4XKPRUK9ePZo2bcrvv//OuXPnaNy4Mfv376+S3T00Gg0TJkxg0aJFfPnllwabntC4cWPatGnDDz9U/vXW+tVWOPd0LJh8TsV6PQpXnt/YFMnpxWfFSm89JCYm4u7uzkcffcR775Wd0xYKXleNGjXCz8+vwim6KiI9PZ3u3btz7tw5EUyWU0pKCtu3b2fz5s1s27aNhIQE7O3t6datG927d6dbt26VDpL+ufo3P6mW6N1GBQoCHYP4pO30h7qbVn5+PlevXi0WaIaHh+sWpDk7OxMQEFAs2LS3L75Itjz27dtXapq4p+a0wb5eQb2X/7xC3Il4suKyyM/Ox9zRDOdmzvj29SmSRSYrPqvUTqRffvmFgQMHVqqdj6onZg5leU1s+iap6lTOJ5yrdFApISFJEm83n6YLJqGgF+2bb77hrbfeYtasWbz//vvMnz+fadOmMWr0KA5FH+JCUqhBhiUkSSL7QC4vfv8i77zzDp9++ukDTRvi7OzMsmXL6NmzJ4sWLWLcuHEVKh8cHMzs2bPZtGlTuRYOlbVbTlJSwdSE+82h3LNnD127di31+D///MPVq1dZu3YtUDCX0M7OjgMHDhg8oMzLy2PYsGH89ttvLF26lJEjRxqsbm9vb65fv17p8puu/oNzr4IvtYoGk4DufeXV05OVYcsZGXj/FE1C6VauXIksy4waNarcZZRKJePGjeONN94gOjqaWrUqtsq/PEQwWT6yLBMaGsrmzZvZsmUL//77r24U6dVXX6VHjx60bt3aIOlkXvB5kXxtPitCl903nVhZgpwa8n6rDx/61qxGRkbUq1ePevXqFfmOyMvL4/Lly4SGhuqCzZ07d/L999/r8mS6uroWCzQbNGhQZA7o/QwbM4zzZmeLPGbheifDSMrVVGy9ran1lCtG5kakR2UQufMm8afi6fDlUxiZFX3uarV3xbmps+7frwaNeiLfM9WuhxIgT5PHovM/sPPG9gq/8RSSAktjK6Y2f4smzk3ve25ERAQzZsxgxYoVuLi4MPX9qVzyvUhslv5zXV706c2w+iOYN28eH374IQ0bNmTt2rX3XQVdFV5//XVWrFjB6dOnK3ztVq1a4eLiwj///FPmuVOmTGHTpk2lprNZunQpr732Gnl5eSV+OEdGRuLl5cW6det46aWSV9sV5su7e9ehnj17olar2bFjR3luqVyys7Pp168fO3bs4Oeff6Zv374Gqxtg0qRJbN26lYsXL1a4bHjSRd4+MNWgQ2fTWrxH21pixW9FabVa6tWrR8uWLXU/csorLS0NNzc33nrrLT7++GODtis9PZ3nn38elUrF9u3byzUPujrJyspi7969uqHsyMhILCws6Ny5s24ou3ChYVU4deskX5/+ijR1Wrm/ZxSSEpAZWO8V+vr1e+jBZGXk5uZy6dIlXU9mYbB59epV3fbD7u7uJQaahYvcCnsoR84dwW2/RLRy+RO5xxyJ5dTcMzSd3Jha7QumVBX2UDYY5o9P74K5qApJwUu+fQluUP5czI+Lx+9VYwDGSmP+12Qi7dzasfDstyRkJ5SaV6pQ4fH2tTowquEYbExsyryOt7c3P/30E9OmTeOzzz5j6v+m4unrSYfpT5FsUvJin/spbMPAeq8wyP8VJEli2rRpPPvsswwePJimTZsyf/58xo4d+8AmMM+bN489e/YwePBgjhw5UqFe0uDgYN58803i4+Nxdna+77mFPZSyLJd4b0lJSdja2pb6S3/v3r1IkkTHjh1LPH7s2DH+/fdf/vrrryKPd+jQgenTp5OXl3ffxRDllZaWRq9evTh58iQbN26skhXk3t7eREREoNVqiyzuOXHiBCtXrmTv3r1ERERQo0YNWrduzYwZM/Dz80OtUfPVqflIkkRaZDqhyy+QFJaMwkiBczMnAkbUL/dwTuGHqoTEd2e/IcAxkNU/rmbhwoVcu3YNR0dHBgwYwGeffXbf3KLV2Z49e7hy5Uqp0zzux8bGhldeeYUff/yRDz74wGDTDkQwWbKIiAhdALl3715ycnKoU6cOL774om7lfVXNw75XM5fmfN95MeuvrGPb9S1k5GWglJRo7gmOlJISraxFkiTaurWln98AatuWPUf3UWVqakpQUFCxfI85OTlcvHixyLD5X3/9xZdffklhf5qXlxcBAQG6rBaXbl/COssShYmixLR+JbFwLsiqcHeO6Lvl5+QX1GVcEPQ/iQFlteyhvJtG1nD61im2Xd9K6O0QsvKzip3jaulKW7d2dPPujoulSwm1lE94eDiffvopv/7+K00HN8G9txuyJJfZG6SgYB5bTYuaTG4+FX+H+sXOycrKYurUqfzwww/06NGDpUuX4uJS+bZWxMmTJ2nTpg1vv/02M2fOLHe527dv4+rqyhdffMEbb7xx33PXrVtH3759SUhIKHGO0bvvvstvv/2fvfOOiur62vAzQ+8gihRBQBEb9t4V7L3G3mPXGI2JvUWNLXZj11hj712wtyiKCoqIqIiK0nsbZu73Bx+jI20oivF3n7VcCfeee+65Q5l39tn73ft48eJFptf3798fHx8f7t+/n+n5Hj164OXlxdOnT1XefG/fvk3dunX5999/qVWrltrPlhlhYWG0atWK58+fc/r06S9WDXv8+HE6duzIu3fvsLKyUh7v1q0bN27coHv37lSqVIn379+zZs0a4uLiuH37Nm8Mg/jrwRoSwxK5OvEGmvqaOLS1R56USsCxl+gV1aXh4vrKrk/pgvLz7RyAIuWLoG+RZrAtlUiRHVdweutpunXrhqurK0+ePGHdunU0a9aMc+fOfZHX4b/OlClT2LFjB2/evMnTB8SLFy/i6urK48ePs+xbnxtiY2Np1aoVPj4+nD9/Ps/NDb4HZDIZN27cUIpIX19fNDU1adSokTIK6ezsXOgWODK5jLvv7+AX+ZRnkX6EJ4ajEBQYaBtS2rQ0pU2dqGNVB7PvtEgkOxISEvD19VURmp6ennz48AENXQ3kSXIkUglFyptRfkBZTEubqlwvCAIpsTIEuUB8cDy+O/2IehZF45UNlQby6X8j0+dDkuYxXaFPWS7PvIaWNP9Bim+J/3lB+SmCIBCSEEJoYghyQY6uhi52xiXR0yzYzhM+Pj7MmTOH4xeOU/WHKti1tCVJmub9JUVKqjw1LbL0/3+LyhUpT/tSHahtVSfHH8BTp04xePBgBEFgy5YttG//dQp2FixYwPTp07ly5QoNG6pvi9ClSxcCAwO5d+9etuPu3r1LrVq1uH//PlWrVs1wfvjw4dy7dw9PT88M5wRBwNbWlp49e7J06dIM51++fEnp0qVZvXo1o0aNUjmXkpKCmZkZc+bMUba1zAtv3ryhRYsWhIeHc+7cOapUqZLnuXLi0aNHVK5cmRs3bqiI1ps3b1KjRg2VKLK/vz8uLi5069YNk0FGvIkN4uEGb4IuvqHpmsboF0v72Q99GMbt2XeoNLIiJVukmedntp2TGUkRSbgPu0TvXr3ZtfOjP+yaNWsYO3as0j9WRJV27dqhUCg4ffp0nq4PDw+naNGi7N27lx9++CFfa4mJiaF169b/02Lyw4cPnDlzhtOnT3P+/Hmio6OxtLSkTZs2tGnThubNmysjXCL/TW7evMnM+TMJcwhB21ib2KA4Ao69QJ4sp8EfdTFx/JiDmRSZzIXBHsqvdc11qTCoHNb1P36ITwhN5NFf3ljWLo6uuS4J7xN4ceIlieFJbN+3nf7d+n/V5/vS/Ld7IhUwEomE4gbFqVjUhcrFquBcpGyBi0lIK/Y4cOAAd67cwfKDFQe6HuLlwkAayZowsMIgXh54hVO4M/Pr/8E/bfezqNESGtg0VOvTTNu2bfH29qZ27dp06NCBESNGEB8fX+DP8Dm//fYb9erVo1+/fkRHR6t93YABA7h//z4+Pj7ZjvvU3DwzIiIisuwO4u/vz9u3b7Nst7hy5UpMTU0ZOHBghnPa2trUrVs3X16Uz58/p0GDBsTHx3Pt2rUvKiYBpbVMusF+OvXq1cuQkuDk5ESFChV46POQoNjXCAgE33pP8RoWSjEJUKxyUQysDXh3IzjTe6YmpaKQZZ4yEukXhSAXqNGqusrx9ArHvXv35ur5/lfw9vbOV7s2c3NzrKysVPoQ54WYmBhlZPLChQv/M2Iy3dZnzpw51KpVC0tLSwYPHszr16+ZMGECnp6evH37li1bttC1a1dRTH4H1KtXj2V//4mdmy2WtYrj1LUUDRelfSj33aWav69tqEWd2bWoNbU6zr2c0DbWJjUxVWWMfjE96syqhX2rkljWLI5jewca/dkAHWNt5kzN3ALvv4woKAuRypUrc/ToUTw9PbEzL8mkHr8yo8ssAk8G8fjgE46sO4rPfR9yG0S2sLDg+PHjrF+/nh07dlCtWrVMI3cFiYaGBjt37iQiIoJx48apfV3r1q0xNzdnx44d2Y6zsLBAW1tbWektCAIRERG8efOGlJSUbNsuenh4oKmpqSy6+ZTIyEg2b97MyJEjs+ws0rBhQ65du6ZM7M4Njx49okGDBujq6nL9+vUsLY0KEiMjI8zNzdWq9BYEgQ8fPqBjnJYbmRieREp0CqalM1ZDmjqZEP0iJsNxv33+nOl1nlM/nOXqpBuEPAhVOa9ITXvdQmWqx9Nf75yi0/+LREdH8/r163z3/3VxccmXoEwXk0+ePOHChQv5Tvv41omOjubgwYMMGjQIa2tratWqxfLly3FwcGD79u28f/+e27dvM3PmTKpXr/6f71MtkpHPAzcGVgZY1ipOuHcEgvzje7FUS0qxykUpXrM4ZXo44TKsAg/XevPh7ods59c20sa2WQle+L/gzZs3X+QZCgvxt+EboHr16pw6dYoFCxbw+vVr4uPjOXfuHLNmzaJOnTq4uLiwbdu2XAlLiUTC8OHD8fLywsjIiLp167JgwQKlrcKXwMHBgTVr1rBjxw4OHDig1jXa2tr07t2bXbt2Kb3FMkMqlVKiRAn8/f1ZsmQJTk5OmJubY2tri4mJCV5eXshkmSdDX7x4kVq1amXaZWLjxo3IZDLGjBmT5b0bNWpEZGQkjx8/VuuZ0rl16xaNGzfG2tqaa9eufdHKzs9JL8zJid27d/P27VsquJZHQ6JBcmRaxwods4wFBLpmusjiZMhl//8zJJFQrEpRyg8oS82p1akwqBwp0cn8+/tdPniGKK8zsE4rurl987bKfOlR37dv3+blEb9r0iP2hSko/xfEpCAI+Pr6snTpUpo2bUrRokXp3r07d+/eZcCAAVy5coXQ0FD27dtH//79cyweFPnvY22Y0WZLr6geilQFqclZv0cVKWuGjpkOb66+y/Ee+kXTPkyn2919L4iC8htAEATmzJnD1KlTqV27Nubm5krvuDJlymBiYsLgwYMZO3ZsrgWhs7MzN2/e5Ndff2X69Ok0adJELaGRV/r160f37t0ZPny42kKhf//+BAcH4+Hhke04c3NzNm/ezPTp06lbty67du1i9erVzJgxg5iYGA4dOsTSpUtVhLdCoeDSpUs0a9Ysw3wpKSmsWrWKPn36ZNvtp3bt2mhpaXH16lW1ngfgwoULuLm5UbFiRS5dupTnvtp5xcHBIccI5dOnTxk9ejR169alQqvyCAjIU9KiiRpaGf80pBfjKP5/THbbOY///mgeb1rKBNMyplzbdZ1t27bx6tUrzpw5w/Dhw9HS0iIxMTHDvf7X8fHxQUNDI982YC4uLrx8+ZK4uLhcXRcTE0PLli2VYrJmzZr5Wse3RGJiImfOnGHMmDGUKlWK8uXLM2PGDAwMDFi1ahWvXr3Cx8eHRYsW0ahRowJxdxD572CiY4KZjqoxesKHBKTa0gz+kp+jkClITchadKYjjUwr/Pza7wtfGlFQfgPMnDmT2bNns2DBAs6dO4e1tTWdO3fm4sWLFCtWjJs3b+Lo6Mi6desYPnx4rufX1tZm/vz5XLlyhaCgICpXrsyuXbtyvZWuDhKJhPXr16Onp8fAgQPV2iauXr065cuXZ/v27VmO8fLy4uHDh0BaTuTOnTvp06cPY8aMYerUqejq6uLm5sakSZOYMmWK8jpvb2/Cw8MzzZ/ct28f7969Y8KECdmuT19fn5o1a6otKA8fPky7du1o3Lgx586dU9tMtyDJKUL5/v172rZti4mJCQcPHlRWtmtop/1JkGeSD5meIynVzvrPRvp2TvzbeBLDPgrFGr9Wo1ipYgwePBgHBwfat29Pjx49qFq1qtIDTuQjjx8/xsnJSa0uUtlRsWJFgCy7Q2VGdHQ0LVu2xNfX97sRk69fv2bdunW0b98ec3Nz2rRpw6lTp5T/jYiI4OTJk4wcOZKSJUsW9nJFConQ0LS0nGrFq/+/NydEv4zh/d0PFKtcFIlUQmpSKqnJGQM7724FI4uTYVLq49/75OjkDOOSI5LxP+dPpUqVVFw4vgf+J30ovyVu3brFggULmDdvnlII6enpkZSURNOmTbl27RoXLlxgxowZvHjxgi1btuDo6Kh2G7ZPadiwIQ8fPmTMmDH069ePkydPsm7dujy3qcqKIkWKsH37dpo3b86qVasYP358tuMlEgn9+/dn9uzZxMTEZEhuT0pKonfv3piZmaGpqYmdnZ3K+ZSUFOLj4+nbty8tW7Zk0qRJNG7cmNatW+Ph4YGurm4GvzxBEPjzzz9p2bKl8k03Oxo1aqTsWpKdFcjff//NkCFD6N69Ozt27Piq3Ys+xcHBgdevXyOXyzN4EEZHR9O6dWuioqK4du0a1tbWGIeZIEGi3OpO3/r+lKTIJLQMtdDQyt7TUK9oWjGPLE6m/H99cz1GbBpGT/PevH//HicnJywtLbG2tv4qeaX/NRISEgrkg0j6HOoW5qWLyadPn+Lu7k6NGjXyvYbCIDU1lZs3b3L69GlOnTqljPg2aNCAOXPm0LZtW8qVK1fotj4i3xY//PADenp6OFctw8uIl2kdcM4HoaGtQbl+absF8cEJ3Jr1Lzb1rTAsYQgSiA6I5s2Vd+hZ6OHYzl45n++Op8S/T6BopaLomumQEJJI4PnXSJIlrFy5spCe8sshRigLEZlMxqBBg6hZsya//fab8riurq5yG1AikdCiRQtu377NiRMnMDY2Ztq0aTRp0oSbN2/m+p4mJibs3LmTf/75h3PnzlGpUiUuXrxYYM+UjpubGz///DOTJ09WK4erb9++JCcnZ5p7OXPmTF68eMHw4cN5//59hm3/T9suTpw4kZYtWzJ06FCioqK4ePEi9evXz2AqfPHiRR4+fMjEiRPVep6GDRsSHBxMQEBAlmOWLVvGoEGD6Nu3Lzt37iw0MQlpEUqZTMa7d6r5PElJSbRv355nz55x8uRJpT+ho6kjckGOnrku2sbaRD3PWKkf5R+NiUPOlazxH9K8XLWNPz6/RCLBydQJJycnGjZsiKWlJU+ePCE4OBg3N7f8PKpIAZEuJv38/P6TYjI0NJSdO3fSs2dPihUrRuPGjdm2bRvVq1dn//79hIWFcfnyZSZNmkT58uVFMSmSgU6dOhEWFsb2dTvw2fiEd9eDsaxTnEZL62Nkm7aTomuui1VdS8K8w/Hd5ceTbU+J8I3Evk1JGi6pr/J3r1iVYoCEV6cD8d74mNcXgihdrRS3bt2iSZMmhfOQXxAxQlmInDhxAj8/P+7fv6/S4SU9QvkpEomEdu3a4e3tjYODA/7+/tSvX5+WLVsyd+7cXCfM9+zZk/r16zNgwADc3NyYOHEi8+bNy/cW26csWLCACxcu0KdPH+7cuZNtpwgbGxvc3NzYsWMHQ4YMUR4PDQ1l5cqVTJs2jRo1aiCXywkODqZEiRLKMZ8KSolEwubNm3FycuKvv/7iypUrKlvg6fz5559UqlRJbTFTv359JBIJV69epXTp0irnHj58yODBg5Wm6Tt27ODy5csMGzaMYcOGFUqeTLp10MuXL5XFQHK5nB9++IFbt25x7NgxlV6yjoYffSSt6loSdOkNiWGJyghj6KMw4t/F49jeXjkuOTpZpXMOpFWJB3m8wdjeCN0iH7/fCkGBk9nHSKRCoeDXX39FX1+fESNGFNyDi+SJz8Vk9erVc76okBEEAS8vL6W5+J07dxAEgRo1avDTTz/Rtm1bsRJbJFeMGzdO6VLyOuY1P10ak6HDkI6xNpVHqlcsZ9PQWtmGEUBHQ5e/XNdTTP/7yp1MR/xNK0TWrVtH3bp1Mxh16+rqZhCU6djZ2dGxY0fMzc3Zt28fr1+/pnbt2rRv3x4vL69c3d/W1hZ3d3cWL17MypUrqV27dq4rmbNDV1eXXbt24efnx/Tp03McP2DAAK5evapSTPL3338jkUgYNWqUcqv7cy/KTwUlpPVr7dGjB2vXriUuLi5D/uSTJ084c+YMEyZMUDtKYWJiQpUqVTLkUa5bt46qVaty//59XF1dWbt2LTt37sTNzY358+dTqVIlHjx4oNY9CpL0PLBP8ygnTpzI8ePHad26NREREezatYtFixbRpk0bqjhUJfplDIJCwKlbKTR0NLg5419ennqF/6Hn3FvihVFJI2xdPwp53x1PuTHtFn77/Ak8/xrfXX5c+fka8iQ5FYaodmZ5tu0562dsYN26daxatYp69epx+vRp1q9fnyGFQeTrEhUVRYsWLf4TYjI2NpbDhw8zZMgQbGxsqF69OkuXLsXW1patW7cSHBzM3bt3mT17NjVr1hTFpEiesTO2Y2CFQQU658jKo75bMQmioCw0/P39cXd3Z+TIkRnOfbrlnRkjR47E29ubEiVK4O3tze7du3n27BnVqlWjS5cuubIJkUql/PLLL9y5cweZTEaNGjVYvXp1gRXsVK5cmQULFvDnn3/mWMXdqVMnDA0N2blzJ5AWxdqwYQPdu3enaNGiykhbuhdlOp8LSoARI0bw7t079PT0MrxBLlu2DGtra3r16pWrZ2nUqJGKoJw/fz6jRo1CEARWr16Nu7s7o0aNom/fvmzZsoUXL15QokQJGjVqxI0bN3J1r/yir69P8eLFVcR5urA9ceIE/fr1o1+/fkyePJkzZ84QExPDoFqDkUgl6BXVo/68OhhY6uO704+AIy+wqFaMurNrqeRPZradY16+CPUX1qVoxY8m81KkNK7dmLt37jJp0iSmT5+OoaEhHh4e9OvX72u9JCKZEBUVRcuWLZV/j741MSkIAn5+fixbtgw3NzfMzc3p2rUrt27donfv3ly8eJGwsDAOHDjAwIEDs3VrEBHJLR1KdaJz6a4FMteA8gNpZpd5c43vBXHLu5DYuHEjRYoUoVu3bhnOZbbl/Smurq6UKlWKdevWUa9ePXr37k2PHj3YvXs3c+fOpVKlSvTo0YNZs2ap3cO3SpUqeHp68ttvvzFu3DhOnTrFtm3bCqQK7eeff+bUqVMMGDAAb2/vLIuADAwMlMUsM2bMwN3dnYCAAKXpuYmJCUZGRllGKD+dt06dOhgaGmJgYKCSTvDhwwd27tzJnDlzcp3j2KhRI1auXMmbN2949uwZ06dPRyKRKCvOP8fS0pJLly7RqlUrevfujbe391ftpvFppbcgCMybN48tW7awf/9+EhMTadmyJYMHD6ZDhw7o6OggU8jwvHSXN3FBGNkZUWdW9mkUn2/nZIYECUY6xvz5y58YTs3oAypSeKRHJp8/f467uzvVqlUr7CUBaXm+V69eVW5lBwQEoKOjQ9OmTVm2bBlt2rTB0dERQRBITU0VbX1EvhgSiYSBFQZRRNeMvx9vQ0BAIajf4EIqkaIp0WR45ZE0L9niC67020CMUBYSHh4edOzYET29jK0ds9vyhrSoYo8ePVQifpqamgwYMICnT5+yadMmbt++TcWKFenbty/+/v5qrUlPT49Vq1Zx5swZHj58iIuLC0eOHMn9w2Wy3u3btxMfH8+IESOyjX7279+fgIAAbt68yeHDhylTpoxKrp+dnV2mEUpDQ0MVgZicnExiYiKhoaEqr+XatWvR0tLKk/1SeqedkydP0q5dOyQSCUeOHMlUTKaTHnENDw9n0qRJub5nfnBwcMDPz49FixZRtmxZGjZsyOXLl5k8eTKBgYGcOXOG7t27K/NmtaRaTKiuXpGSuggIjKv6E4baopjMLQWxS5DVHN+amHzz5g0bN25UpvO0bNmSI0eO0Lx5c06cOEFERARnzpxh5MiR+Pj40Lp1a6RSKdra2hgbGzN48GDu3r1bqM8g8n0ikUjoWLozq5qtoYyZMwAakuydLtLPVypambWu6/8nxCSIgrJQSE1N5cmTJ1SuXDnT83p6ejmaPVeuXJng4GDCw8NVjmtpaTF06FD8/f3566+/uHz5MuXKlWPQoEG8ePFCrfW1atWKR48e0bBhQ7p06cLQoUNzbYz8Oba2tqxfv579+/eze/fuLMc1atSIkiVLsn37dh49ekTNmjVV8hxtbW0zjVB+3nbx5s2byOVyFAoFT58+BdKsWP766y8GDx6cJ6skCwsLnJyc+PXXX0lMTGT79u107Ngxx+scHByYP38+mzZtIjAwMNf3zS0ymYxjx47h6enJrVu3mDVrFjVq1MDDw4OAgABmzJiRZdceR9NSjK4ytsDW0tO5FzUtv78OK18aKyurLPvW54b0nzdr64+R5KioKJo3b16oYjI1NZUbN24wdepUKleujK2tLaNGjSIyMpKZM2fy6NEjAgMDWbduHe3atUNfX5+4uDjatm1Lx44dCQ8PZ/HixYwcOZIJEyYou2H99ttveWqRKiKSE7ZGdixutJRlTVbQzM4Vc92imY6z0LegpX1r1jT7i7n151HcoPhXXmkhIoh8dXx9fQVA8PDwyPT8lClTBAcHh2znePz4sQAIly9fznZcYmKisGrVKsHS0lLQ1NQUfvzxR+HVq1dqrVOhUAibN28WDAwMhFKlSgm3b99W67rs6Nu3r2BsbJztGqZPny4YGxsLRkZGwsKFC1XODRs2TKhWrZrKsZEjRwpVqlRROTZt2jShSJEiAiDs3LlTEARBWLdunSCVSoWAgIA8rT04OFgwMzMTAKFx48a5ujY2NlYwNjYWpk2blqd7q4Ovr6/w66+/CsWLFxcAwc7OTpBIJMKHDx9yPde5l2eFDkfaCh2PthPaH2mTq38dj6Rds8d3t6BQKL7Ak37/7Nu3TwCE0NDQfM2zbNkyQU9PT0hNTRUEQRAiIiKEGjVqCEWKFBHu379fEEtVm7CwMGHXrl1Cr169lL+bRYsWFfr16yf8888/QkRERJbXRkRECNWrVxeMjIyEM2fOZDifmpoqLF26VACE/v37iz93Il+FmOQY4Wm4r+AT6i08i/AT4lLiCntJhYoYoSwEcurTm1NRDoCTkxNaWlo5FuDo6uoyduxYAgICWLhwIUePHsXJyYnRo0fn2BpRIpEwZMgQvLy8MDc3p379+sydOzfbnts5sWbNGkxNTenfv3+WbST79+9PTEwMsbGxGV4jW1vbTLe8P49QXrx4EVdXV0qWLIm3tzcKhYJly5bRuXNnHB0dyS2vXr2iQYMGymfPrJgqOwwNDenXrx+bN2/Osud4XoiLi2Pr1q00aNCAcuXKsWnTJnr06IGXlxebNm1CEIQ8RZdb2LdkaePlWBukRbYkqFcNL0GCqa4pc+vNo1fZ3qLXXx5J/7nP7Pc7IimCu+/vcPLFCY49P8q5V2fxDX9CcmrGNBlvb2/Kly+PhoYGkZGRNG/enBcvXuDu7p7BXaKgEQSBBw8eMH/+fOrXr4+FhQV9+/bFz8+PUaNGcevWLd6/f8+OHTvo2bNntrsGY8eO5fnz51y7do1WrVplOK+hocHEiRPZvXs3O3bsYMOGDV/y0UREADDSNsK5SFkqFK2Ik1kZDLQMCntJhYpYlFMIeHt7Y2FhkaU/YU5FOZC2tV2uXDm1K7r19fWZOHEiw4cPZ+3atSxevJgtW7YwYsQIJk+enG11pJOTE9evX2fevHnMmTOHs2fPsnPnTkqVKqXWvT8l3Vi9SZMmLF26VMXQ/dP7lS1blqdPn2boYmNnZ0doaCiJiYnK/NPPBWVsbCx37tyhf//+JCQk4O3tzYkTJ/D398+2vWNWPHnyhObNm6Onp0f9+vU5e/ZsngoBRowYwdq1azl27FimxVjqIggCt27dYuvWrezbt4/4+Hjc3NzYu3cvHTt2VPp9pufOvnr1Kk8i2snMiZVN1+Dx2p0TL44TFJu2Bash0VB6s2lINFAICgQEiukVo51je1rat0ZfSz/PzyeCsu2it7c3TZs2JSY5GvfX7px+cZKQxBAgTbxLkKBAofy6UrHKtHVsR83itdCQauDt7Y2Li4tSTL58+RIPDw+qVKnyRdYdFxeHu7s7p0+f5vTp07x9+xZDQ0OaN2/Oxo0bad26tcr2uzocPXqU3bt3s3379izThNLp3bs3V65c4ZdffqFly5ZKP1YREZEvj0QQvkBDZ5Fs6dKlCzExMbi7u2d6fvXq1co8vezo27cvL1++zJMlTUxMDKtWreLPP/8kOTmZ0aNH8+uvv+Zown3r1i369u1LSEgIq1atYuDAgXmKQk2ZMoU///yT27dvZ5rD1blzZ44ePcq7d+9UKs0vXbpEs2bNePbsGU5OTgDUqFGD6tWrK6MSp06dol27djx79oytW7eye/du7O3tle3YcoOnpyetWrXC2tqaM2fOUK5cOaRSKYMHD2bZsmW5fu5atWpRqlQp/vnnn1xfm16hvnXrVnx9fSlZsiSDBg1i4MCBmfYfTk5ORk9Pj02bNqmYxecFQRB4Gf2CZ1HPeBEVQFRyFIIgYKRthKNpKUqbOlHGrAxSibjpUVBUrVqV6jWq02lyB7Y93kqqIhWB7P9cSyVSFIKCEoYlGF91AlVsqzJ16lSOHj36xcTk8+fPlRXZV65cISUlhTJlytC2bVvatm1Lw4YN89w1ShAEypcvj4ODA6dOnVLrb01sbCxly5aldevWbN68OU/3FRERyT1ihLIQePLkSabbNumkV3kLOfSNrlChAidPnszTGoyNjZk+fTpjxoxh+fLlLF++nHXr1jFu3DgmTpyIubl5ptfVrVuXBw8e8NNPPzF48GBOnjzJxo0bsxyfFXPmzOHcuXP06dOHe/fuoa+vGtHS1NREIpHwzz//MGHCBOXx9GKS169fKwXl5xHKixcvUqJECUqXLo2LiwtBQUEEBQVx8ODBXK3xypUrtG/fngoVKnDq1Cmio6OJjY2lWbNmGQzO1aVmzZpcuXJF7fGpqamcPXuWLVu2cPLkSaRSKV26dGHVqlU0a9YsW+NmHR0drK2tVczN84pEIsHRtBSOprmPSovkjfJVy/PG5TUbvdXfvk23NHkX/45J1yZi3cqK7du3ExERwcWLF3OM8KlDSkqKiq2Pv78/2traNGnShCVLltCmTZsM3aTyypUrV3j69Cnr1q1T+4OrkZERI0eO5I8//mDp0qWYmpoWyFpERESyRwwnFAKJiYnZ+hGmb+UmJydnO4+JiQkJCQn5WoupqSlz5szh5cuXjBs3jlWrVuHg4MDMmTOJiorK9BojIyO2bt3KgQMHuHz5Mi4uLly4cCFX99XW1mb37t28evUq023v2NhYrK2tM2xRp7dc/DSPMjNB2axZMyQSiTK6aWtrS6dOndRe38mTJ2nVqhW1atXiwoULFClSRJle0LJlS7y8vIiJiVF7vnQqVqyIn58fKSkp2Y7z9/dnypQp2NnZ0b59e169esXy5csJDg7mn3/+wc3NTa0uIPb29irm5iL/DeJSYpG3lKFpk709SVakpyGU718Wg0b6+RaT7969Y/PmzXTu3Blzc3OaN2/OgQMHaNq0KUePHiU8PJxz584xbty4AhOTAOvXr8fZ2ZnGjRvn6rohQ4aQnJysbJIgIiLy5REF5TdIeg5cTnmUBYm5uTkLFizg5cuXDB8+nKVLl2Jvb8/vv/+epXDq1q0bjx49okKFCrRo0YKff/45V2suV64cS5YsYc2aNZw5cybDeXt7ex49esTDhw+Vx3R1dbGwsFBaqqSmphIdHa0UlOHh4Tx48IBmzZoBadvEAAMHDkRDQ7035z179tC5c2dat27NqVOnMDQ0BNKKqUxMTOjYsSMKhSLX2+eQVmyRmpqqtDL6lPj4eLZv306jRo0oU6YM69evp0uXLty7dw8vLy/GjBmTofgoJxwcHAokQiny9VAICv64M58knUSkGvn/E23XvgRhpiG5ukYul3Pr1i2mT59OtWrVsLGxYfjw4YSGhjJlyhQePHhAUFAQGzZsoGPHjsrfkYLkw4cPHD58mBEjRuQ6rcbKyopOnTqxfv36Auv6JSIikj2ioPwGSY9Q5pRD+SUoVqwYS5Ys4cWLFwwcOJD58+fj4ODAwoULM60WtrGx4dy5c8ot85o1a/Lo0SO17zd69GhatWrF4MGDCQ0NVTlXvHhxLCwsMkQpPzU3T4+ipgutS5cuASgF5aFDhwDULoJZt24dffv2pU+fPuzfv19p+g0oCxzKlClD8eLF87TtnV5klB7tFASBf//9l2HDhmFlZcXAgQOV0dt3796xZs2afPkEOjg4iBHK/xhnX53BO8xbWWzzOVEB0dyee4czvc9zutc5bs2+Q/TL7KPlGx6tJywxLNsxERER7N27l379+mFpaUm9evVYt24d5cqVY/fu3YSEhHD9+nWld+SXruA/duwYCoWCAQMG5On6oUOH8uTJE548eVLAKxMREckMUVB+gxRGhPJzLC0tWbFiBQEBAfTs2ZOZM2fi6OjIn3/+mWGbXSqVMn78eO7evYtEIqFmzZosW7ZMLYNhiUTC1q1bkclkDBs2TCWaIJVK6d27N7t371axKvrUOujzPt4XL17EyckJW1tboqOjOXXqFJDW1jEnFi5cyKhRoxg7dixbt25VadkIHwWlRCKhYcOGXLt2Lcc5P8fU1JQSJUpw584dli1bhouLC3Xq1OHs2bOMHz9eaenSu3fvTLso5RZ7e3vevXuXY/qEyLdBXEocW322ZHk+KiCaG1NvkfA+gTI/lKZMDyfig+O5Of02cW+ztodKUaSw7bN5BUHA29ubhQsX0rBhQ4oVK0avXr3w9vZm2LBh3Lhxg5CQEHbv3k3v3r1znSedXx4+fIizs3OemhAA1K5dGyBXH3BFRETyjigov0G+BUGZjo2NDWvXruX58+d06tSJyZMnU6pUKVatWpVhfS4uLty5c4fRo0czceJEWrRowZs3b3K8h5WVFZs2beLo0aNs3bpV5dyAAQMICQnh3LlzymOfdsvJTFCmRyc3bdqUY64ipL2x/vbbb0yZMoVZs2axYsWKDPmJKSkp+Pn5Kf0BGzVqxJ07d3IVRZbL5Zw+fRqZTMbq1auZMmUKFSpU4OzZs7x8+ZK5c+cWuM1J+nxfo0OPSP65GOSBTJ71z6zfP8/Q0NagwcJ6lOroSOnOjjT4oy6CQsB3l1+W1ykEBTfeXedtxFtOnDjBiBEjKFmyJJUqVWLevHmYm5uzfv16goKClN6R9erVUztN5EuQ/gEur5iZmWFjY6O2tZqIiEj+EAXlN0hhbnlnhZ2dHRs3bsTPz49WrVrx888/U7p0adatW6cS/dLV1WXZsmVcuHABX19fKlWqxIEDB3Kcv3PnzgwZMoSffvqJ58+fK49XrlwZFxcXduzYobKWoKAgBEFQEZRv377Fz88PV1dXZDIZK1euxM3NLdv7yuVyRowYweLFi1mxYgWzZ8/OdCsvNjaW1NRUpV9no0aNSElJ4c6dOzk+W0BAANOmTaNkyZK0bduWlJQU7OzsePv2Lfv27aNly5Zf7I3b3t4eQMyj/I9w5uWpbK2BIp5EUrSSOdrGH214dIvoYl6hCCGeoaQmZt10QC6X4zqiGR06dMDDw4MuXbpw7tw5wsPDOXr0KD/++KOy6K2wEQQBHx+ffAlKSPuQKwpKEZGvgygoC4nsEsXVjVAWRs9aR0dHtm3bhq+vL02aNGH06NGUKVMmQwcYNzc3Hj16RLNmzejRowcDBw7MsSp6xYoVWFpa0rdvX+XrI5FIGDBgAMeOHSMyMhJIi1DGxcURFRWlIigvXrwIQJMmTThw4ABv3ryhe/fuWd4vJSWFPn36sHnzZrZt28ZPP/2U4/Oni82KFStiamqaZR5lQkICO3fupGnTppQuXZo1a9bQvn177t69S+fOnbG0tKRo0cx7wRYktra2aGhoiHmU/wHiUmJ5G5d99yqFTIGGdsYPHxo6GihSFcS+js32+jqdauPn54e/vz8rVqygRYsWKnnC3wrv3r0jMjIyQ2OD3CIKShGRr4coKAsBKyurbLcg1RWUgYGB2NjYFOja1KVMmTLs2rWLx48fU6dOHX788UfKli3L9u3blfmO5ubmHDhwgL///ptDhw5RpUqVbE3YDQ0N2bVrF56enrx584awsLQigt69eyOTyZSRTjs7OyDNOigiIgJdXV309PS4ePEilSpVomjRovz55580b95cac9kYmKicq+EhAQ6derEkSNHOHDgAAMHDszV82toaNCgQQMVQSkIAnfv3mXkyJFYWVnRv39/AHbu3ElwcDDr1q2jRo0aX7UdoaamJiVKlBAjlP8BAqIDchxjYGNA5LMoBPnHD6QKmYIo/ygAEiOy/pshkUpINZVRpkyZfK/1S5MuAgsiQhkYGJgniy8REZHcIQrKQiCnT83qbnkXxJZQfilXrhz79u3j0aNHVK5cmYEDB1KhQgX27NmDXC5XRhgfPnyIlZUVjRo1YsaMGVn2s65Tpw7Tp0/H19cXLy8vBEHAysqKli1bKqu9083N0wWlubk5giAo8yevXLnC/fv3mThxIt7e3hmigdHR0bRq1YorV65w8uRJunTpkqdnb9iwIbdu3SI4OJiVK1dSuXJlatWqxYkTJ5S9hy9dukTfvn0zGLd/TcRK7/8GoQmhOY6xb2VH/Lt4Hqx9RGxQLDGBsXitekhSZFraiSI5+12LmJQYUhVZb4t/K/j4+GBgYKBM2cgr6X8ffXx8CmBVIiIi2SEKykLAxcUFX19flcplALlCTmDMK+5Fe+LYwQHP5Lu4B17gRVRApm8C3t7e+d4SKihcXFw4fPgw9+7dw9nZmT59+ijzJxUKBY6Ojly5coXZs2fzxx9/UL9+fZ49e5bpXNOmTaN06dJER0cTEJAWtenfvz83b97k+fPnWFpaomeqy+N3PgSnBGNRqhj+z/15/fo1rq6uLF26lIoVK9KiRYsMr1FoaChNmzbF29sbd3d3mjdvnqfnlcvlaGlpER8fj52dHZMmTaJMmTKcPn2awMBA5s2bl6de518Ce3t7ZYTy0aNH/PTTT0gkEiQSCZMmTcLPL+tiDpGCQxAEYmJiePXqFffv38fDw4MDBw6wYcMGFi5cyP4D+3Kcw75VSUp3LcXba++4PO4aV8ZfI/59AqU7pfVq19DLORc3vQ/7t8yHDx+wtrZWy7w/O9JzQkNCcufDKSIiknvE1ouFgIuLC8nJyTx//pyyZcvyMvolZ16e5lKQB8nytEhDuf7O3Jd4cs/rLgCaUk0a2jSijUNbypg5ExUVxdu3bws9Qvk51apV4/jx49y5c4dZs2bRo0cPXFxcmDNnDp06dWLGjBm0bNmSPn36ULVqVZYvX86PP/6osg2spaXFX3/9RfPmzRk3bhynT5+mQ4cO2NWyZe7V2Wg8l+K2rRnXuQrFoEQta37z+YX6C+oSaPCKc5fOsXHNRiQSCT4+PrRv3x5Ii2i2aNGCyMhIrly5QqVKlXL9fB8+fGDmzJn8/fffBAUFIZFIaN26NVu2bMmxD3ph4eDgwMmTJ2nRogUXLlxQFhbp6emxbds2li5dSufOndm+fTtGRkaFvNpvn5SUFCIjI4mIiCAiIiLH///0v3J5RjEnlUoxNTXFvllJbEpbZXJHVcr1daZUJ0diX8eiZaCJcUljZYW3oXX29lhSiRQtqVbeHvwrUxCpIV8zvURE5H8dUVAWAukRM09vT87Gn+ZikAdSiQaKTyIHUg2pSrVnqiKVq2+ucCnoIrUsa1M9qQaQ/xyjL0WtWrU4c+YMN27cYNasWXTp0oWqVasyd+5c2rZti5eXFxMmTGD48OGcOnWKzZs3qwiypk2boq2tzZkzZ1h3dB0PDb2oPMWFSHkEkqSMbxJyiRyzsqacizpDy7/doKxAVGwUAQEBuLi44O/vj5ubGxKJhGvXrin7gKtDYmKisg/4iBEjMDIyolevXgwZMoSpU6cCfLNiEtI8OMPCwnjw4AEHDhygY8eOaGmliYqkpCT279/P2LFjadSoEadPn1a2q/yeEQSB2NhYtUXhp/8fHx+f6Zz6+vqYmZlRpEgRihQpgpmZGRUqVMhw7PP/NzY2RiqVEhjzirEXR6u1fm1DLczLf+yaFPYwDF1zXQxtsu9YU8LQFqlE3JgSEREpeERBWQgUK1YMx9oOHOYAijdpOU8KNbah0reqPD/c5b78HpbVi3/zCfb169fH3d2dK1euMGPGDNq3b0+tWrWYO3cuGzZsoE2bNgwdOhQXFxe2bdtG69atgbSil0pVKqFZX4MznEQSlyYiJRpZRxzSoxFSLSlHAw5zJfAyhnaGaGtr06BBA4oUKcKFCxfUskYRBAEvLy+2bNnCnj17lB15Bg0axOrVq5VG6Y0aNWLFihUoFAq1t+dCQ0MzFAl9KZKSklizZg0AO3bsoFWrVirndXV16d+/P1WrVqVVq1Z07dqVa9euFar/YG74NFqY26hhdtHCT0WflZVVjsLQzMxMWUyXV0oY2qIl1UKmyDy/OCveXn9H1PNoyg8si0Sa9e+HhkQD5yLO+VqjiIiISFb8TwvKhw8fcuTIERYvXsy4cePo0qULNWvW/OLbJE8jfCk/sSwyDRkSIff3UggKFBIFNaZU41HEQ2pY1vwCqyxYGjduzJUrV/Dw8GDGjBm0atWK+vXrM3fuXLy9vRk8eDBt2rRhzJgxLF68GA1tDeyGlCClaFoKQHbefJkhIBAhi6DBwrpMXvwbJUqU4OzZszlGEiMiIti9ezdbtmxRFhKNHDmSQYMG0a5dO4yMjFS67jRs2JBZs2bx+PFjtaPFPj4+dO7cOVfPk1dmz57Nu3fvADIVUOm4uLiwf/9+GjZsyPLly/nll1++yvpANVqYG1GYXbRQT08vg/jLTbSwMNCQalDLsja3g29lmecY/jiCZ/v9KValKNpG2kT6RRF08Q3FqhbDoZ19tvPLBTm1Let8gZWLiIiI/I8KygcPHjBq1Chu3bqFrq4uSUlJLF68mEWLFlGtWjVWrVpF/fr1v8i9I5IimH1zJhJNCeRHt0rSbED+uDOflU1XU8LItsDW+KWQSCS4ubnh6urK2bNnmTlzJq6urjRp0oQ5c+bQtm1bJk2ahIeHBx1XtEdWLCV/4l4ioKGlQZXJlfiz4fIsxaRCocDDw4MtW7Zw5MgRFAoF7du3Z968ebRq1UrZgtHFxSVDtWjt2rXR0tLiytUrFC9VnHhZPFKJBDMdM4x1MkYh4+LiePHixVcppvL392fJkiXMnTuXOXPm5GgdVL9+fSZMmMD06dPp168fxYsXz9X9chMt/PxYZmJXIpEohV76f62srChfvny2orAgooWFRVvHdtx4dz3L87pFdJBIJQQcfUlqYir6Fno49y5DqQ4OSDWyF8LmuuZUt6xR0EsWERERAf4HBaW7uztdunTB0dGRQ4cO0b59e7S0tJDL5Zw7d47ff/8dNzc39uzZU2BRJH9/f2bMmMH169cJCfuATlEdbBpa49jJEU0dDRJCEvAYfjnL6+3cbKk8WjX6FRUQzbN9/kT4RnJcdgrn0s4MGzaMcePGFciavyTphSytWrXixIkTzJw5k8aNG9O8eXO2bt3KmtOreZzgkz/BnX4vDQlSTSmbnm1kkeUSlfyxwMBAtm3bxrZt23j9+jVly5Zl/vz5WYqpihUr8tdffym/TkxN5GrIFZotbcJ5izOcP39GZbyZThEqFK1A85ItqFysClKJlMePHwNfJ/d1w4YNmJmZMXHiRLZt26aWddDUqVNZu3Ytf/75J7169SrQaOGn4q9cuXLZisLCjhYWFhXMK+Js5ox/1PNM02AMrAyoM6tWnubu4dwTDcl/I5VBRETkv4dEyK5ly3eGl5cXtWvXxs3Njf3792NomDGBPTk5mf79+3Pw4EE8PDxo0qRJvu4ZFBREpUqVMDExoV2fdlyLvEKkXyRBF99SvKYFtabWIDUplfe3P2S4NsQrlLdX31H9l6pY1/9YKBHyIJS78+9h7GiMdX0rNHU1KCeUx86oJIsXL87XegsDhULB0aNHmTVrFn4vn9JyS3P8jwfgt+cZRraGNFnVSDn25vTbhD+OyDBHsapFqTPz4xtt7OtY/Pb6E/0imqTIZDR0NChTtgxzfpuDTCZjy5YteHh4YGBgQM+ePRk8eDB16tTJNiJ66NAhunXrxpt3b7gRc40Dzw6QLE8CgSzFr1QiRSEoKK5fnBGVR/Hg5EN+/PFH4uPjlX6jX4Lo6GgcHByULS1Hjx6NlpYWffr0yVEgqhstzC5C+Omx/2q0sLB4G/eWsRdHFZhfpFQipVyRcsxvsPA/U5AzZ84c1q5dm2+7n+fPn+Pk5MSFCxdybMMqIiKSP/5nIpTJyckMGDCA8uXLc/ToUbS1tTMdp6Ojw549ewgODmbQoEF4e3tnKjzVZefOnURFRXH9+nX2Rf2DQ5g9JVvYISjgzeW3pMTJ0DbUokSTjB1vgi69QVNfk+I1LZTHZAkyHqx8iEX1YtT4tRoSqQQJEooZmbOo2aI8r7MwkUqldOnShU6dOjH38ByuR17l+aEANHQzj6bomutSrq9qcYFuEdX2cQmhiaQmpVKiaQl0zXSQJ8t5fzOYbt26AdCgQQO2bNlC9+7d1f7+uri4YGBtwIy7UwlThH08kU0kVSGkFV2FJIQw59YsDFOMcCrnpJaYTM8tzG1eYWRkJHFxcQBs3bqVrVu3Kuf08fHJMVoYHh7OjBkzWLFiBe3bt/+fjRYWFjaGNoyoNIo1D1bley6pRIqhliHjq038z4hJgAoVKhAaGkpISAgWFhY5X5AF6Q0kKlSoUFBLExERyYL/GUG5bNkyfH198fT0zFJMpqOhocG2bduoVKkSc+bMYcmSJXm+b3rLLw0jDR74eymP65rpgBSkmpmrkaSIJMJ8wrFtYqPSu/ft1XckR6VQto9zWiu1pFQ0tDUIjA0kIPo5pU3Vt8P51pBIJLw1DsJ3pR9mZUwRFAIpMSkZxmnpa2YqwD+leHULildXfSNyaGPPrTF3MNYz5tq1a7len0YxKQ0X1SNMHpbr7fj0oqLY4jE4jrRn285txETE5CgQs4sWfioKixcvrhSGZmZmbNq0CR0dHVavXk2RIkXYtWsXGzduJDw8POe1CgKHDx/m2rVravU3Fyl4Wti3JEmexGbvjUiQ5LooDUAq0cBQy4AFDRZS3CB3+bCFTXpKiLe3N66urhnOKwQFMSkxyBWpaGvoYKhlmOnugre3N+bm5krvVRERkS/H/4SgTE1NZe3atQwePJjKlSurdU2pUqUYN24c69atY/bs2SqVvbmhSZMmLFq0iKFDh6LRWoK2kRYRTyN5de41Dm3t0dTN/Fvw9nowKMCmkapwCnsUjqa+JkkRSdxdeI/4d/Fo6GpQorEND8s8/E8LypCEEJ7c9SX45nsaLauPz6YnWY5VyBUoUhRo6qn/I6ypqUkJ5xK8f/o+12sLSwhlxo1paOpr5iu3UyKVoGmlwYZH63i01AczU9Xt4U9FYVbbyiYmJtlGCxUKBbNmzWLGjBk0bNgQSHuDjoiIICYmRtnfPMs1/n/x1P79+/P+oCL5pkOpjhTXL84qr5XEy+KU0e6cEBQCEqmECuYVGF/tZ4rp5z3CV1iULl0aHR0dfHx8lIIyNCGUC4HneRjqxYvoF8omEABGWkY4mZWhpmUtmto2Q18rrdWpt7c3Li4uosG5iMhX4H9CUJ48eZK3b98yYsSIXF03bNgwFi1axL59+xg8eHCe7t2qVSt+//13fp//OykXPkbbnLqVomyfrD3h3l59i46ZDkVdzFWOxwfHI8gF7v5xDzu3Epj3dSb8cTgvTwUy/+f5dD3XLU/r/BZ4Fu6Hz+bH2LmVwLhk1qInLjieMz3Po0hVoGOqjV1zO8r0KI1UM6PISk1KRZ6iIDVexvu7Ifhef0rnjp15+vRprraQS410wLxSkRwradVBoiHBomoxDvocoI1D23zP9zmvX78mLi5OpfDHwcEBgFevXqnVIcjFxYUlS5aoJUBFvhy1reqw3rwCu3134h54gRRFijIv93M0JBrIBTnaMh0e/P2QvzZt+E+KSUjbJSpfvjze3t6EJYax6dEGbgffQiKRZPrssbJYvELucz/kHtt8ttDWsR29yvbB29ubFi1aFMITiIj87/E/ISjXr19P7dq1qVq1aq6uc3BwoFWrVqxfvz7PghLSeinbV7FHr6oOWkZafLgXgv+hAHTMdHBoY59hfNzbOKIDYnBsb5/BqDg1UY48WU7JlnZUHJqWF2RV1xKFTIHX+QcsWLCA4sWLK3s15+Yf8FWv+/za5XuWkxCSSJ3ZWVex6lvqY+5ijrGdEanJcoJvBuN/4Dnx7+Kp/kvG7++Tbb4Eng9K+0IKxatZcPjwYQ4fPqwyTkNDA4lEglwuR1tbG3Nzc+rUqUP9+vVRlJTz3Dat73hsUByPtz0hwjcSqaYUi+rFqDCoHDomqjmc8cHx+O70I+xRGHKZAhNHE8r2LqPyAWGr92bqWNXF/bg7y5Yt4+nTp2hoaFCxYkV+/fVX2rbNm9hMzxv71JooXVC+fPlSbUEJaTmX9erVy9M6RAoGI20jRlQeRb/yA7jy5jJPwh/jF/GUsMQw5IIcHQ0dbI3sKGPmTC3LWjjoOFJ6VGnmzZvH+vXrC3v5ecbFxQV/4Rmj3IeTokhBQCC7GtL0tIAURQpHnx/h2ptrhEpCv9luYiIi3xvfvaAMDAzk3Llz/P3333m6fuTIkXTo0IEHDx5QpUqVXF+/d+9ehg0bxk/7xvJE/hgBAau6liAI+O7ww6aBNdrGqjmdb6+mGVHbNM6YJ6ihnRYhs2lorXLcppENgeeDmDZtWq7X+K0g1Zbi3NMpgzj7lCpjVMWQbRMbHv7lzesLQTi2t8fM2UzlvEN7B6zqWZEUkcS7G8FINTT4+++/cXR0VG4hjx49mtu3b9O9e3cqVarE+/fvWbNmDefOnePWrVtsCt2AJEZCQlgCN6ffRlNfk7J9nJEnpRJw7CWxgbE0XFwfqVba9yYxLJHrk28hkUKpTo5o6GgQdPENt+fcoe6c2phXSGuZl6pIZcK8n9k+fwdt27Zl4cKFJCUl8ffff9OuXTsOHTpEly5dcv06ent7Y2JiotIRqHjx4ujq6uboRZlO2bJl0dDQEAXlN4SBlgFtHNqqFdWeOnUqv/32GxMnTsxVm9FvCeNGRpiYG5EkT8r1tQIC4Ulh1P29FrpFRZcBEZGvwXcvKG/evAlA+/bt83R9q1at0NLS4saNG3kSlH/99RdVq1bF0saKp0G+yg4YxWsWJ+jiW6JfxlCsclGVa95ee4eBjQGmpTIaY+sU0SU2KA4dU1URqm2S9vXy5csZN26c8pO8IAj5/lcQ8+Q0x9y5c7l44yKl25XK9WtcqqMDry8EEfooPIOgNCphiFGJtCpu26Yl+HfWXebOncuOHTsoWrQoFhYWTJo0iRo1aqgUa/3www+4uLgw/fdpCL3Tvmf+hwJITUql4dL66BdLq9I2dTLl9uw7BF16Q8kWdgA8PxSALF5Gk5UNlb2V7VrYcWnMFR5vfUKjPxsAoEDB4W1HqFGzBidOnFBGbAcPHoyNjQ3bt2/Pk6D08fHJkDcmkUiwt7dXy4sS0loyOjk5KaOdIv8tRo0axYoVK5gxYwZ79+4t7OXkmtMvThJort7PalYIpOWSHo4+QM2IGpQtUq6AViciIpIZ372g9Pb2xsbGhiJFiuTpei0tLcqWLZvnN9YPHz5gZmaGvXFJldwfQS6o/DedyGdRxAcn4Nwr86iCaSljwh6GkRSepBQrALLItP6/FhYW/zl7F39/fw4ePMjY2WPxiXyoPC6XKVDIBRJCEtDU00TbKPPqfN2iaeJOFpuxIvxTBEHAtLQJzw4/p0GDNFFnaGiIo6MjpUqVyvCvfPnyeD/2oZKkAgpBQfCt9xSvYaEUkwDFKhfFwNqAdzeClYIy3DcSEwdjle+Ppo4GljWL8+pMIHHv4jG0TivySklIxqiIkYr4MzY2xtDQMM8+ld7e3srn+xR7e3u1I5SQtuUoCsr/Jnp6esyePZuhQ4fy22+/5TrdpzB5ExvEJu+NBTKXRCpBEASWei5hbbO/0NEUo5UiIl+K715Qpkdr8kN+3ljLlCnD+fPn0QjXUrH+eHvtHUjB2N5IZbxyu/uzLe10rOtb8fzwC157vKFopY+RzVcXXqGhqZFvI/avTWJiIsePH0ehULBy5spMx3gMv4xDO3sqDimf6fmE9wnAxyhtVkgECY7FSvGM5+zatQt9fX0CAgKU/w4fPkxgYKCKVY9ZKVPkcjlJkcmkRKdgWjpj1NjUyYSQe6HKrxUyBVoGWhnGaeik2T9FB0QrBaV5RXOuuF9h9erVtG/fnqSkJFavXk10dHSeLXtevHjBoEGDMhx3cHDgxo0bas9TpkwZbt++nac1iBQ+AwYMYMmSJUydOpUzZ87kfMFXIi4ujiVLlvDvv/9y584dIiMj2bZtGwMHDkQQBFbcX57BJkmRquDKz9eJexNH+QFlKdXJUXkuu05j1SZUwaahNaEJIex5uodBFQejUCjYsGEDGzZswM/PD319fSpXrszy5cvVdgERERHJyHcvKL29venevXu+5nBxceHkyZMIgpBr+4lJkyZx5swZhnQYQvEWFqAv8MEzhJD7odi52aJb5OMnZkEu8O7GO8zKmGJglblNkYmjCbauJQjyeIMgFzCvUIQwn3CCb75n4q8TsbbOXIh+K8jlch48eIC7uzsXLlzg+vXrJCcnY2pqSqXKlZDXTiFZmhZp9Nv9jNSkVCoMKY+BpT6yBBlSLSkaWh99OQVBwP/gcwCKVfnYqzs5KhkdU9VcTIVCgfsODwA8PDz44YcfGDdunMpWt0wm4/Xr12zcuJHFixfj4JpWGJUcmZbHpWOWMcKha6aLLE6GXCZHQ0sDQxsDIp5EkJqYqmJrFOGb1uEnKeJjTljloS6Yppoybtw4ZdvMokWL4uHhQd26dfPwCqc9p5ZWRkFrb2/P7t271f451tLSQqFQz6pG5NtDU1OTefPm0b17dy5fvvzNfNgMCwtj7ty52NnZUblyZS5fvqw85xvhy7NIvwzXvDz1isSwxGzntWlohUU11ar29BQYAYFTL0/wg3NPRv04it27d9O/f3/GjBlDfHw8Xl5e+e7KIyLyv853LShjY2N59epVgUQoY2JiCAoKws7OLlfXNmrUiJs3bzJ79mxunr1JTFQM+hb6lO1ThlKdHVXGhj4KIzkqBadupbOds9KIiugV0yPI4w3B/75Hv5genSd2Yumipbl+tq/BixcvcHd3x93dHQ8PDyIiIjAwMKBx48YsXLgQNzc3KlSogEQi4dCzA+x4sh0BgZcnXgFgVTvNlDjMJ5z7yx5g08AKAysD5Clygm9/IPJpJHYtbFVyTh+t9yE1IZUiFYqgW0SX5Khk3l/7QGpiKk2aNOHKlSts27YNU1NT2rdvT9euXWnRogV6enrIZDLWr19P3bp1cergSKQsEnlKmrDS0MqYTpBejKNIUaChpUHJlnZ8uBvCvaVelO1TBg0dDV6dfU1UQDSAci4ATR1NLEpaUNO5Fu3atSM2Npbly5fTpUsXrl27RunS2f8s5AYHBwdiYtLM1POaAiLy36Jr165Ur16dKVOmcPPmzW/Cj9HKyorg4GAsLS3x9PSkZs2aynOnX55U2h+lkxyVzLP9zynd2RG/f/yznNfE0STbhgcp8hR+Xz+H7du3c/jwYTp37lwwDyQiIgJ854Ly8ePHgKp9Sl5Iv97b2zvXghKgVq1anD59msikCEa4DyMxNfNP2hZVi9H+SJsc55NqSnH+wQnnH9LyLDUkGqxqlv82bQVFeHg4Fy9eVIrIFy9eoKGhQa1atRg9ejRubm7UqVMn045F9ikOKGQKyKTron4xPYqUMyP43w8kRyUjkUgwLGGAy4iKlGxhqzLWur4Vrz2CCDwbSEqsDE09TcpXKs/cNXPp0KEDgiDw6NEjDh8+zKFDh9i5cycGBgY0a9aMf//9F2NjYw4ePMjsxzNA9rG6Xi7LGLFT/P8x6f+PKV7dgoo/lsd3px9XJ6ZtMRtY6VO2tzO+O56i+UlLyX+X3MXKyIpLn0RpOnbsiJOTE9OmTWPfvn25e/Gzwd7eHkjzohQF5f8GEomEhQsX0rx5c44fP07Hjh0Le0no6Ohk2rlGEATuvr+jIiYBfHf6YWhjQInGNtkKSkjznZVqSJUf8j5FgoRdG3ZTq1YtOnfujEKhIDExMc9NK0RERFT5rgWln1/a1km5cvmr7rOzs8PAwAA/P788ewMCmOkWYVilEay8vzxf6/mcvuX6UcLINueBX4ikpCRu3Lih3Ma+f/8+giDg7OxMmzZtcHNzo0mTJpiYZMw/TCchIYH58+ezZMkSKnVzwbqHJfXm1VEZo19cnxqTqqm1JpuG1so8VKlEir2xA382Xo6GNE3MSSQSKleuTOXKlZkzZw5+fn7s2bOHJUuWkJiYiJaWFsOHD8eqf3EkWhLlVnf61rfK80cmoWWopbIV79DGHttmJYh5FYtUS4qJvTGvPdL8MA3+P38y/n0CIfdD6bWgl8p8RYoUoUGDBrnKd1SHT70oq1VT73UU+e/j5uaGq6srU6dOpV27dmhoZPJp7RvgQ8L7DB+2I59FEXT5DfXn182xQ5XfPn+ebH8KEjApZULZPmWw+CQNJiUhhTdP3tJxVCemTp3K6tWriYuLw8HBgYULF9KjR48v8VgiIv8zfNeCUiZLq3zW1c1fZZ9EIkFXV1c5X35oZuuK14f7XHt7NU/9eT9FihSXYpXoVDr31jL5QaFQKPMg3d3duXbtGklJSVhYWODm5saYMWNwdXXF1lY9kXvy5EnGjh1LcHAw06ZN49dff2Xxg4V4fbiPgvzl8EmQoCnRZEL1X5RiMjNKlizJpUuX0rbdDx3i9evXHDp0iGcH/XDqUQo9c120jbWJeh6d4doo/2hMHDJ2k9HU1aRI2Y82RqEPw5BqS5XHkqPSWseZ6xTNcK1MJiM1NTXXz5sd5ubmGBgYqG0dJPL98Mcff1CrVi127drFgAEDCns5mRIUG6TytSAI+Gx+jE19K4qUNSMhJCHzCyUSilUpimXt4uia65LwPoEXJ17y7+93qTWlBsVrpOVVJrxPAAH+2fsPWppaLF68GBMTE1auXEnPnj0xNjamVatWX/oxRUS+W75rQfktIpFIGF99AjKFjFvBN/M+DxIqFK3AtNozshVKBcWrV6+4cOGCMg8yPDwcfX19GjduzPz582nevDkVK1bMVY7W69ev+emnnzh69CgtWrTg/PnzShPmX2tMps+OnsjN5Bm6BamLIBfQ1NRkRt3Z2Blnnaogl8v54YcfuHXrFseOHaNNm7S0g/Hjx3Pd/xqLHy8E0joSBV16Q2JYInr/b1UU+iiM+HfxOLa3z3YtEU8jeX/7AyVb2SkrwA2s9JFIJXic8GD6z9OVr92bN2+4du1aptY/+UEikeDg4JAr6yCR74OaNWvStWtXZs6cSc+ePdHRybp5QGGRIle1/Qq6+IaYwFhq/Jp9NF2/mB51Zql21yrRxIbL467y+G9fpaBMTUrbSo8Ij+D27dvUrl0bgA4dOuDg4MC8efNEQSkikg9EQVkIaEo1+bXWZA77H2KP7y4EhEz702aGVCJFEAQ6O3WhT9l+aGlkrOYtCCIiIrh06ZJSRAYEBCCVSqlZsyYjR45U5kHm5Y1JJpOxfPly5syZg6mpKfv376dbt24qYnTR/EUc/+MkI3eNIFD6CgmSXEV0JUiQxch4fzAUp+bZdwqZOHEix48fp3379kRERLBr1y7lOUEQKGFly9u4Nzh1K8W7m8HcnPEvju3sSU1KJeDoS4xKGmHr+rErTUJIIveW3qd4zeLomukQ+zqWV+deY2RvRLm+H/u365nq0ahzQ64cuoKrqytdunQhNjaWv/76i8TERKZMmaL2836KpqYmSUmZdxfJjbl5UlISmprin4jvhXnz5lGhQgU2bNigdBT4lvj0g7EsQcbTXX6U7uSo/PCWG7SNtLFtVoLnh18oPwCm50GXtC+pFJOQ5kXbvn17du3aRWpqqvgzLyKSR8TfnEJCQ6JB9zI9qGVZix1PtuP5/q6yl/TnETkJab2uFYKCCuYV6V9+AM5FyhboepKSkrh586ZyG9vT0xNBEChTpgwtW7akefPmNGnSBFNT03zd5+rVq4wcORI/Pz/GjRvH7NmzMTZW3S7esmULc+bMYcGCBUzpPoV/g2+z8s5y4oS4HIWloEg717ZUO6o71KTR6EaMHDmSv//+O8vo6YMHDwA4ceIEJ06cyHD+wqvzrPJagV5RPerPq8Pjbb747vRDqinBoroFFQaVU8mf1NTXRMdMl1dnApHFytA118GxrT1O3Uur2AhJkfLP1n841OQwW7ZsUQrImjVrsmPHDho1aqTei/oZZcqU4enTp5mec3BwwMPDQ615fH19KVOmTJ7WIPLtUbZsWQYNGsS8efMYNGgQRkZGOV/0FbHUt1L+f8DRlyhSBawbWCm3uhPD0j4kpcTJSAhJQNdMN9Pim3TShagsToZeUT2lRZuVpVWGsRYWFshkMuLj47PN9RYREckaUVB+RR4/fsyxY8fYvXs3o0aNokOHDpS0tWdGnVmEJoRw8fVFlu9cRomqNgjaaS0JDbQMKG3mRBkzZxqVaIyNYda2GLlBoVDw6NEjZQTy2rVrJCYmUqxYMdzc3BgxYgRubm55qmrPjJCQECZNmsSOHTuoW7cu9+7dy9RE+OzZswwfPpwRI0YwefJkAGpb1SFlt5y3smA6TGrPv6/S+ml/ilQixdbIFvPoYvwxaCGz3OdSo0YNNm3aRN++falbty4jRozIdG2f+uBlhkJQ4PH6Ar4RvhjZGWXYXvscbUMtak2pnu0YgH7lB2BlbM2YMWMYM2ZMjuPVxcXFBR8fn0zPpXfLUceL0tvbm06dOhXYukQKn1mzZrFr1y6WL1/OzJkzC3s5Ktga2aIl1UKmkJEYlogsTsblcdcyjHt+KIDnhwJotKxBprnL6cR/+P+GB8ZpbhK6RXQxKGLA27dvM4x99+4durq635zIFhH5LyEKyq/A3bt3mThxIteuffzjmG5k3b59e1asWIG9vT0uikr8u+Auiy4voXHjxgW+jsDAQGUltoeHB2FhYejp6dGoUSN+//133NzccHFxKdDWjXK5nE2bNjFlyhSkUimbNm1i8ODBmd7j/v37dO/endatW7N69Wql4AkKCuLo4aOsXLmS9iXaM7v+XHYf201g6Cv+XP4n1V2qc2LXSbQ1tJHL5Ww328HKlSvZuXMnffr04datW4wbN46qVauqbHWpi1Qi5adqE/jp0hiSU5PzXSgklUhxNnOmQ+kvY+FSsWJFDh48iEKhUL7OSalJeH64S5hDCJUnV2TU+eFoamliqmNGaVMnnM2cqVa8OtoaaW++8fHxvHjxIt8eriLfFra2towePZqlS5cycuRIihUrlvNFXwkNqQblzcvjHeqNQ1t7LGsXVzmfEp3Co3U+2DazoXit4uhbpEUgk6OT0TFRTb1JDE8iyOMNxvZGysikFCkN2zbk7M6zXLhwgebNmwNpRuvHjh2jWbNm/7m2tSIi3xLftaBMFyRyuTzfVhmpqal5MgU+ceIEP/zwA+XKlWPfvn106tQJbW1tYmNj2bNnDwsXLqRu3bqcPn2aJ0+eABRY+6/IyEguXbqkFJHPnz9HKpVSo0YNhg0bRvPmzalbt+4XS9C/f/8+I0eO5M6dOwwePJhFixZRtGjGimZIK/pp27Yt5cqVY+/evSp5TOvXr0dfX5/+/ftz4cIFAJrUaMLJkyeJeBJJpH6UUghpaGgwbtw4fvvtNxYtWoS1tTXLli3j3r17dOvWjfv37+fpTdTSwJI59eYx88Y0UuQpeRaVglxAGielrlYDUlNSle0YCxIXFxfi4+N5+fIlptamHPY/xIXAcyTJk5BKpJhXNOdt4ltIhEAC8QnzRi7IMdAyoJV9azo7deXp47Qt8/x6uIp8e0yZMoVNmzbxxx9/sGzZskJZw5o1a4iKiuLdu7RWsydOnODNmzd8iAkluXIypqVMVBoVAMqtbyNbI2WzAwDfHU+Jf59A0UpF0TXTISEkkcDzr5EnyanwSbtWAYE/5/7JQ/eHdO3alQkTJmBiYsL69euRyWQsWLDgKzy5iMj3i0QQhPx513zDXLx4EVdXV54+fYqzs3POF2RBaGgoFhYW7N+/P1dtHC9fvoyrqyudOnVi165d6OllTC4PCQmhbdu2+Pn58cMPP+Du7p5nW5fk5GRu3bql3Mb29PREoVDg5OSEm5sbbm5uNG3aFDMzs5wnywfR0dHMmDGDtWvXUqFCBdatW0f9+vWzHB8REUGDBg2U67ew+Ng+LSkpCVtbW3r37s3KlSuZNm0a27Zt4927d2zatIlhw4bh5OTEs2fPVO5fokQJxo8fz++//w6kVU5Xq1aNypUrc/bs2Tx/wHgRFcDCu3/wIf59nmyfjCKMubfcCz8fP4yNjWnfvj1dunShVatW6Ovr52lNn/Pu3TtsbGxYcmQxdzX/JVmenKuiLwNNA8pGlGfWgNnExcUV2LpEvh1+//135s2bh7+/f4GlteQGe3t7AgMDMz3X9e8uyEySM/x+pffs/ryX99tr73h19jVxb+KQxcvQMtCiSDkznLqXVopSqURKXat6/FZrCi9evOCXX37Bw8MDmUxG3bp1WbhwoUrHHhERkdzzXQvKdCF44MABunXrlud50oWpr68vZcuqVwwTGxtLpUqVKFmyJB4eHtkKmJiYGCpVqkR0dDRNmjThyJEjat1DoVDg7e2tjEBevXqVxMREihYtqhSQbm5ulCxZUq358osgCOzdu5cJEyYQFxfHnDlzGDt2bKZ9pdNJSkqiRYsWPHnyhJs3b2YoAvn7778ZNGgQz549w8nJiZYtW6Kjo8Px48fZsmULQ4cOpWjRooSGhqpc99NPP7Fnzx6CgoKUPqQXL16kefPmTJ48mfnz5+f5OZPlyezx3c3xgGMoBHm2wlKKFAUKTLRNGFZpBA1sGgJp+bSHDh3i8OHDPHr0CD09PVq3bk3Xrl1p27ZtvgoD5Ao5dX6uhVXTjN1I1CG98CnsYgQ3Vn4b7fpECpa4uDgcHR1p164dW7duLezlqOAV4sWsm9MLdE49TT3+ct2AuZ55gc4rIiLyke9aUAJYWloyfPhw5syZozyWnJqEX6QfAVHPeRP3hhR5ClpSLawNrSn1//lk+lofozIrV67kt99+Iy4uTm1LiQkTJrBx40YePXqEo6NjjuPTRWv79u05fvx4luNev36trMR2d3cnNDQUXV1dGjVqRPPmzXFzc6NSpUpfPRfIz8+P0aNH4+HhQbdu3Vi+fDklSpTI9hqFQkGvXr04fvw4Hh4e1KtXT+W8IAhUr14dS0tLTp8+jSAIFC1alHHjxjFr1iy2bt3KkCFD0NTUJCUlRUX4PH/+nDJlyrB582YGDx6sPL5o0SImT57MsWPH6NChQ76eOSY5GvfX7lx5c5nXMYEZWsbpaerhbFaWFvYtqWNVF01p5j87/v7+HDlyhEOHDnHnzh20tbVxc3OjS5cudOzYMcs0gcwQBIH1D9dx+uUpCkIH/uDckz7l+uV/IpFvjtWrVzN+/Hh8fHzy3U2soFn3YC1nX53Jd/OHdCZU/4Umtk0LZC4REZHM+e4FpZubG8bGxhw+fJh3cW859eIk5wPPkyxPQoJE6esokUiUfpBaUi2a2bnSzrE9JY3tGTp0KPfv3+f+/ftq3TM2NhZra2t+/vln5s6dq9Y1QUFB2NnZYWtrS2BgoFIcRUVFKfMg3d3defbsGRKJhBo1aigjkPXq1ct3N6C8kpiYyPz581m8eDG2trasWbOG1q1bq3XtL7/8wrJlyzh06BCdO3fOcP7GjRs0aNCAM2fO0KpVK169eoWDgwMnT56kbdu2yuglpEVcPu/J27FjR16+fMnDhw+Vr6cgCHTp0oWLFy9y7949Spcunc9XIA2ZQsauE7v4ZepEjh87gaOVI5b6lrmO7gUFBXHkyBEOHz6sLOJq3LgxXbt2pVOnTtjYZF/lf+XNZf70XJLn58iMOfV+p6qF2KrxeyM5OZmyZctSrVo1Dh06VNjLUSFVkcrCOwu4+/5OvkVlv/ID6F5GbKsoIvKl+e4F5c8//8ypM6eYd2wue57uBlArnyxdaHZx6sa6YetxLu3M9u3b1brn+vXrGT16NIGBgTlG6dI5ceKEMmK2atUqQkJCuHDhAnfv3kWhUFCqVCllBLJp06YUKVJErXm/JKdOnWLs2LG8ffuWyZMnM3ny5EzzRDNj1apV/PTTT6xatYqxY8dmOqZnz57cv3+fp0+fIpVKOXjwIN27dyc4OBhLS0t27NihbCP35s2bDGLr0qVLNGvWDA8PD5o1a6Y8Hh0dTc2aNdHT0+PWrVsFliO4b98+evbsSUxMTIHYj4SEhHDs2DEOHTqEh4cHqamp1K1bly5dutClSxccHR25fPkyTZtmHnlpsLAuZs5p+bKKVAX+hwJ4c+kNSeHJ6JrrYOtqS+kujkg1PkazUxNTCTj6gkj/KKL8o5HFyaj3c10uLHJXidqLfB+k/w79+++/1KqVvR3W1yZVkcqmRxs58+qUMnVEXaQSKVKJlB9dhtPaoc0XXKWIiEg6372gXL9tPbtDdijfWHOLBAmxQXG4ytyYNiHnvB5BEKhatSr29vYcPXpUrfHe3t78+uuveHh4IJfLEQQBc3NzXF1dad68Oa6urjg4OORp/V+C169fM378eI4cOULz5s1Zu3atsmWiOhw+fJhu3boxYcIEli5dmumYt2/fUrJkSZYvX64UnJMnT2bXrl28efMGgJ07d9K/f38AfHx8qFChgsocgiBQpUoVSpYsmSGNwMfHh9q1a9O1a1e2b99eIHmC6RFTmUxW4N02IiMjOXnyJIcPH+bs2bMkJSVRpUoVqlWrxtatWxk3bhxJVol4frhL+q90sWrF0Pl/D757S714dzMYO9cSmJQyIfJZFG8uvcWuuS2VR320BkovfNArpot+cX3CfSKoMrYSM8fOpLNT1wJ9JpHCRy6XU7lyZSwsLPDw8Pgm82UfhHixymsFYYlhSCXSbAMCGhIN5IKcskXK8VO1nwvMt1dERCRnvmvboARZAveL38XEOO8FDgIC+tZ6PNXyJSo5ClMd02zH37lzh4cPH7Jw4cIsxwQFBankQYaEhCCVSjE1NaVu3bqcO3eOJ0+eqFQ7fwvIZDJWrFjB7NmzMTExYd++fXTv3j1Xb0I3b96kT58+dO/encWLF2c5bv369ejp6SkjkAD37t2jRo0ayq8/zRONiorKMIdEImH8+PEMGTIEf39/FdFbsWJFNm3aRJ8+fahbty4jR45U+xmyIjExEU1NzS/Sus3MzIx+/frRr18/4uLiOHv2LIcOHWLPnj0AHDp8kBrLq2NT1jrDtVH+Uby7EYxT99KU7Z1W9GTfqiTaxtq8OP4ShzYlMbZPM4jWMdOh+VZXdM10iHoexbVJaf3mT744QcfSnZFKRJ++7wkNDQ0WLFhAx44dcXd3V3ozfktUsajKpuZbufv+DqdenuRJ+GNkClmGcfqaBtSyrEUbx3Y4mzl/k+JYROR75rt+d1jjtYoQWYjKll5ekGpIiRViWHp3MTkFdK9du4aBgYHKH+bo6GiOHTvGmDFjKFu2LHZ2dgwZMgQ/Pz8GDx6Mu7s7tra2DB48mDVr1pCamoqnp2e+1lzQXL16lapVqzJ58mSGDRvG06dP6dGjR67+aD979owOHTpQq1Yttm/fnmXhUFJSEhs2bGDgwIHKtoyCIHDv3j2qV//YgebTe2cmKAF69epF0aJFWb16dYZzvXv3ZsyYMfz000/cvn1b7efIioSEBLW3/PODoaEh3bp1459//uHYsWMAOFZ3JEWWjEKeMXoT/iQCAJuGqi3nbBpYgQBvrwcrj2loaaBrltGXNDQxFP/IZxmOi/z3ad++PXXr1mXKlCkoFPkz7f9SaEg1qGNdl9/rz2d/+0OsafYXj5b5UOGDC7PqzmFzi23803YfE2r8QtkiZUUxKSJSCPxnI5RxcXEsWbKEf//9lzt37hAZGcm2bdsYOHAgADff3uD6u2uc6Hw6yzmKVjan7uyPnVOeHXhOlH8Ukc+iSIlOocwPpXHumRbRUQgKHoU95HzgOeLvJ7Bv3z7u3r3L+/fvsbW1pV27dsyYMQNvb2/Kly/PjRs3lBHIO3fuIJfLcXR0xM3NjXnz5tG0aVPMzdMsLKKioggMDFRuzxoZGeHt7U2bNoWf+xMaGsqkSZPYvn07derU4d69e1SpUiXX84SEhNC6dWssLCw4cuRItkVE+/fvJzQ0VKUd4cuXL4mMjMwyQhkZGZnpXLq6uowcOZI///yTuXPnZuhF/ueff3Lv3j26d++eZ9PzdBITE7+KoPwUbe20Le27FzxJOpaERCqhSHkzyg8oi2lpUyAtfxJAQ1vVuirdVD36RXSO95Eg4XnU8wLvIS9S+EgkEhYuXEjjxo05dOhQrrx2CwMNiQYlDG0JvPYax36lqF68Rs4XiYiIfHH+s4IyLCyMuXPnYmdnR+XKlVX6McsFOZu8NyBBQtWfMnadiQqI5uXJVxSrrCoe/PY8Q8dMBxNHY0K9wjK97zafLbgPu4S1tTV9+/bFzs6OR48esXr1anbt2kViYiKJiYk0btyYIkWK4OrqysCBA3Fzc8vSPujhw4cAVKlSBYlEQsWKFfH29s7jK1MwKBQKZctEgI0bNzJkyJA82RHFx8fTrl07EhISuHjxYrYFRYIgsGrVKlq2bKliRp8esf00Qpm+Fi0trSwjlAAjR47kjz/+YOvWrUyYMEHlnLa2NgcOHKBatWr06tWLc+fO5dn0vLAEZdeuXVE4y3krvCEmKJaAYy+4Me02Df6oi4mjCYbWhgBE+EaiX/xjYU34kzQRnhSelON9pBIpL6NffJmHECl0GjVqROvWrZk+fTqdO3f+ImkbBUl8fDyA2HtbROQb4j+75W1lZUVwcDCBgYEsWaJqk3L/wz3Ck8IRECjRxCbDP3lSKkgybgG6bmhCi62uVB1fJcv7JqQmMGXtFE6fPk2pUqW4fPky+/fvRyaTERISQlxcHK1atcLT05PQ0FD279/PsGHDsvWi9PLyQldXVymgXFxc8PHxyfuLk0+8vLyoW7cuI0aMoFOnTvj5+fHjjz/mSUympqbSs2dPfH19OX36dI4m67dv3+bevXuMGzdO5binpye2trYqeaXp21qmpqbZCkpLS0t69uzJqlWrSE1NzXDexsaGvXv3cunSJWbMmJGLp1OlMARlvXr1OHjwIGVaOlG8lgVOXUvRcFGan6fvLj8ALKoXQ6+YHk+2+xJ86z0JIYm8uxHM091+SDQkyFNy3uZUCAriZHFf9FlECpcFCxbw7Nkz/v7778JeSo7ExaX9LIqCUkTk2+E/Kyh1dHSwtMy8E4h74IUsiwfkMjnBt95jXqEIekVV3/z1LdSwRRHgzLtT2NraMmjQIHx9fRkwYIAyl00QBH7++WeqV6+utgB78OABLi4uyqhAxYoV8fX1RSbLmHj+JYmOjmbcuHHUqFGDhIQErl27xtatW/O8DSwIAmPGjOHMmTMcPHiQqlWr5njN6tWrKV26NK1atVI5/nlBDnyMUJqYmGQrKCGtc05gYGCWpvFNmzbljz/+4I8//lB+L3NLYmJiobUp/DRnzMDKAMtaxQn3jkCQC2hoa1B7eg20jLTxXHwfj+GX8Fr5kDI9nNAy1EJTt+D7iYv896hSpQq9evVi9uzZJCYmFvZysiU2NhZIyycWERH5NvjPCsrs8I14kqW1RMi9UGTxqdg0yqOdhAT07fTZu28vISEh3Lt3j0WLFql0mnBxcclmgox4eXmpiC0XFxdSUlLw9/fP2xpzSXrLxLJly7J161YWL17M/fv3adCgQb7mXbhwIRs2bGDjxo20bNkyx/Hv3r3jwIEDjB49WkWMpxfkfC4o00WUOoKyevXqNGzYkBUrVmQ5ZtKkSXTu3Jn+/fvn6bUvjAhlOkV0iyD95NdZr6geilQFqclpEVkjOyOarGxIk5UNqTe/Ds23NKNkc1tSYlMwsDbIalolUokUU50v2wNepPCZO3cuHz58YO3atYW9lGxJF5RihFJE5NvhuxOU0cnRRCVHZXn+7dV3SLWkWNfLW59jADQE6reup9ISb9GiRUrrn9xE9JKTk3ny5IlKoUv58uUBePr0ad7XqCbPnj2jRYsW9OrVi/r16+Pr68vEiROz7b+tDrt27WLq1KnMmjVLpfVhdmzYsAEdHR1l95t0AgICiI6OVsmfhI8RSmNj4yyLcj5l/PjxXLt2jXv37mV6XiKR8Pfff2NpaUnXrl2VeVrqUpiCspRpaZWOIgkfEpBqS9HU/ZgLJ5FIMLIzwrx8EbSNtAnzCQcFFKucc2tHuSCntGnBdBUS+XYpXbo0Q4cOZcGCBTl+SCtMxC1vEZFvj+9OUEYlZy0sZAkyPtwLwaJaMbQM8ieYopI/Vsbu2bOHLVu2ULt2baXNjbo8fvyY1NRUlQhlegvB5OTkfK0xOxITE5kxYwYuLi68ePGC06dPc/DgQWxtbfM998WLFxk8eDCDBg1i1qxZal2TnJzM+vXrGTBgACYmqr6hmRXkgKqgVOfNr2PHjtjb27Ny5cosx6S36QwICGDEiBE52kR9SmEIytDQUACczZyVgjL6ZQzv736gWOWiSKSZ26fIk+XKIjSbhhm9KzOjjJlzzoNE/vPMmDGDpKSkLJsOfAuIW94iIt8e352gzO79P/jWexQpCmwaq/cGmsOdgDTfySFDhtCyZUuV9n7q8uDBAyQSSa63yfPD6dOnqVChAosXL+a3337Dx8dH7f7bOeHt7U3nzp1p2rQpGzZsUNsP7sCBA4SEhKhYBaXj6elJyZIlVSLCkLstb0gzcR47dix79+4lODg4y3EVKlRg8+bN7Nq1i7/++kut9UPhCMoffviBtm3bcmT9USIuR+Gz9Qk3ptxCQ1uDcv0+Wvx4LrmPz5YnBJ5/TcDRF1z95QaxQXFUHV8ZTT3Vit6Xp1/x7MBzXnukdST6cDeEsOMR7Fi1g+jonC2GRP7bWFtb89NPP7F8+XLev39f2MvJFHHLW0Tk2+O7E5RG2ll/Yn179R2a+poUr5H/DjQGWoY8fPiQDh06ULFiRQ4ePJinKugHDx7g7OysjEp+SYKCgujSpQtt27alVKlSeHt7M3fu3AITQW/evKF169Y4Ojpy8ODBXG2br169mubNm6vkoqaTWf4k5D5CCTBkyBB0dHRYv359tuN69erFuHHj+Pnnn9U2PS8MQdmpUyfCwsJYvnw5t9f+y7vrwVjWKU6jpfUxsv34u2Ba2oRQr1B8tjzB/1AABlb6NFhUj2KVMm53Bxx9id+eZwSefQ1A8O333Np2mxkzZqiVWiDy3+fXX39FW1ubefPmFfZSMiUuLg6JRFJoRXAiIiIZ+bbNxvJAEV1zDLQMiJep5r8lRSQR5hOObdMSaGjlr6pVQ6JJSkgKrVq1wsLCgtOnT+d568XLyytPRuG5Ib1l4pw5czA2Nmbv3r257nKTE9HR0bRp0wYNDQ1OnTqVq8hBujl9ZhXYCoWCe/fuMXXq1Azn0tefG0FpYmLCoEGDWLduHVOmTMnWYH3JkiV4enrSrVs37t+/n2MrzMIQlOPGjVNaLCWnJjHm4ihCE0JRoFqUVrpzKUp3LqXWnG4bmyr/XyqRUtLInmVNVqAhFavB/1cwMzNj8uTJTJ8+nQkTJmRre1YYxMbGYmBgkKcP8SIiIl+G7+63USKRUMbMGQmqYunt9WBQQIlG+d/uLpZalDat2iCVSjl37lyebXUUCgUPHz78ooLy+vXrVKtWjcmTJzN06FCePn3KDz/8UKBiMiUlha5duxIUFMSZM2ewts7da7x69WocHR0z7Qz0/PlzYmNjM+RPwscIpZGREVFRUWq3jRs7dixhYWH8888/2Y7T1tZWeoz27NkzUw/LT/larRezQkdTl5+r/5JBTOaXCTUmimLyf5CxY8dSrFgxtfOgvyaxsbHidreIyDfGf1pQrlmzhnnz5rF161YATpw4wbx58wg6/JaU+BSVsW+vvkW3iA7mFc2znC/o8lueHXjOi+MvAQh/HMmzA895duA5CSFpvmwSJLjPuMiLFy/o27cv169fZ9euXezatYtHjx7lyr/txYsXxMbGquXPmFtCQ0MZNGgQDRs2xMDAAE9PT1asWJHroqGcEASBoUOHcu3aNY4ePaqsUFeX9+/fs3//fkaPHp1ph5qsCnJAdctboVAoKz9zwsnJiXbt2rFixYoci25sbGzYt28fV69eZfr06dmOLcwq73TKm5dnRKWRBTbfz9UnUtLYvsDmE/nvoK+vz6xZs9i9ezePHj0q7OWoEBcXJwpKEZFvjP/0lvfSpUsJDAxUfn348GEOHz4MQLvNbeD/0xLj3sYRHRCDYweHLKteAYLcgwh/HKH8OtwnnHCfcACKlDND30IPTakmL31fAbB48eIMc6T3VlaHBw8eAGSIUMrlcoA8RREVCgWbN29m8uTJQJoVz9ChQ7/Y1tCMGTPYuXMn//zzD40bN8719Rs2bEBLSytLayFPT08cHBwybdeY/vqkv7FERUWpLZjHjx+Pq6srly9fpmnTptmObdKkCQsXLmTSpEnUqVOHTp06ZTruWxCUAG0c2yGRSFn/8C8kEkmWnqxZIZVIkSDl5+oTaFQi999Tke+HwYMHs3TpUqZNm8aJEycKezlKYmNjxQpvEZFvjP+0oHz16lWW504GHGej9wYADG0MaX8k43bq59SbVyfHMV2dunNIOJrpuS1btvDjjz+qLSy8vLywtrbOkJsXEBAAkGObwszmGzlyJP/++y8DBw5k8eLFed6OV4eNGzcyf/58lixZQs+ePXN9fUpKCuvXr6d///6YmppmOiarghxQ3fKGNEFpZ2en1r2bNm1KxYoVWbFiRY6CEmDixIncvn2bAQMG4OnpiZOTU4Yxhdkp53NaO7ShjFkZlt1bSlBsEBIkKj6VmSFFigIFjiaO/Fx9IrZG6r2WIt8vWlpa/P777/Tq1Yvr16/nu9lBQSFueYuIfHv8p7e8s6ONYzvKFSmPVFIwuV+CQiAmMJbH/zxRRhA/p2LFigiCwJMnT9Sa88GDB5nmT3p7eyvnU4eYmBjGjx9PjRo1iIuL4+rVq2zbtu2LismTJ08ycuRIxowZw8SJE/M0x8GDB3n//j1jx47N9LxCoeD+/ftZCsrMIpTqIpFIGD9+PCdOnOD58+dqjd+6dStWVlZ06dIlU9PzbyVCmU4p09KsaLqan6qOx9Hkk6IKBWktGSUaaHzy++FcxJlfavzKkkbLRDEpoqRHjx5UqVKFyZMn58qX9UsSFxcnRihFRL4xvltBKZVI+aXGr5jomGTZ1zs3cxnqGNJQ1oiF8xfi5uaWqY9hhQoVgI+CMB2ZXIZfxFNOvzjJFu9NrH+4jm0+W3itG0jZes7IBVWB6uPjg729fY6fwAVBYN++fZQtW5bNmzezaNEivLy8aNiwYb6eNyfu3r3LDz/8QIcOHVixYkWeC3xWr16Nq6trlnmXz549Iy4uLtP8SfgYoUx/Y8mtpU3v3r0xNzdn9erVao03Njbm0KFDvHjxgmHDhqm8uaamppKamvpNCUoALakWriWbs7zpKta7bWJijUngI4EACc3sXOlUugu/1pzMpuZbWdRoKY1KNBYLcERUkEql/PHHH9y4cYPTp08X9nIAMUIpIvIt8p/e8s6JYvrFWNhwMVOvTSYyOSLXuWTw/2JSy5B59f/Avq09rjXc6NWrF1WrVmXPnj0qZuaGhoY4OjoqBeW7uHeceXmK84HnSExNRIJERdza97XDX+LH4LMDaOPYjhYlW2Kma4a3t3eO0clnz54xevRo3N3d6dKlCytWrCiQLjc58eLFC9q1a0elSpXYs2dPpoU06nD37l1u377N0aNHsxyTXpBTrVq1TM9/Lihz2ypOT0+PESNGsGLFCubOnZuhQ09mVKhQgS1bttCrVy/q1q3Ljz/+yLNnz3jx4oVyzm8Va0NrrA2tmXJ2KmXKlGHsrz8V9pJE/iO0bNmSxo0bM3XqVFq3bl3odj2xsbGULi22AhUR+Zb4biOU6VgZWLG62RplcYG60Urp/780NYvXYnWzv7A3sQfSCjQePHhAxYoVad68Ob///rvKFriLiwvej7355+keRnkM58SL4ySmplV+CwjIBbnyX3pkLzI5kj2+uxh+YSjnX53D29s7y845iYmJzJw5ExcXFwICAjh16hSHDh36KmIyPDyc1q1bY2xszPHjx/MlnlavXo29vT3t2rXLcoynpyelS5fGzMws0/Ppr5+mpib6+vp56j08cuRIkpOTlU4B6tCzZ0+GDBnCTz/9hKWlJZUqVVIW6vz+++/s37//m9kazIzw8PAMXYdERLJDIpHwxx9/8OjRI/bu3VvYyxG3vEVEvkG+e0EJYKhtxITqvzCn3u+4FK0EpNn/aHyWX6kh0VD6VzoXKcvU2tOZWns6ZrqqgqZ48eKcO3eOmTNnMmvWLFq3bk1ISAgAZauWRd5Wxj9Pd6MQFGpHRQUEkuRJrHmwCut+xSnnkrFjzJkzZ6hYsSILFy7k119/5fHjx5l6N34JEhMT6dChA5GRkZw9ezZf+ZkfPnxg3759WVoFpXPv3r0st7vhY4RSEARMTU3zJCitra354YcfWLVqVZa5sZ9z//59Tp48iVQqRSaTcfToUfbs2QOktXf84Ycf+PHHH3P0rSwswsLCMDfP2j5LRCQz6tatS8eOHZkxYwYpKSk5X/AFEbe8RUS+Pb7rLe/PqWpRjaoW1XgX946HoQ8IiHpOYEwgKfJktDS0sTWypbSpEy5FK2FnnH1RgoaGBrNmzaJ+/fr06dOHqlWrsmX3Ft5UCkRPkXX3FXUoXrM4Xob36CnviZaGFkFBQfz8888cOnQIV1dXTp8+jbOzc77ukRvkcjl9+/bFy8uLy5cvU6qUeh1XsmLjxo1oamoyZMiQbO95//59OnbsmOWYdEGpUCgwNTXNc1vA8ePHs2vXLo4fP07nzp2zHXvnzh1cXV0pV64c58+fp0WLFqxYsUKZh7l27VoCAgIYOnQo4eHhHD58uEBN5POLXC4nMjJSjFCK5Il58+ZRqVIlNm/ezKhRowptHaKgFBH59vifEpTppOeSFQRubm54eXnRu3dv5l6bTdGK5kg18hf4lUglBCQ8Z4v3JuLcE5g9ezZGRkb8888/Bd7lRh0mTpzI0aNHOXLkCLVq1crXXDKZjHXr1tG3b98st7IBnj59SkJCQrYRyvTXIV1Q5iVCCWmm6Q0aNGDFihXZCsqEhAT69u1LuXLluHTpEgYGBuzbtw9XV1eWLFkCpOVQDhgwAFNTUzp16sSGDRsYMWJEntb1JYiMjEQQBDFCKZInKlasSL9+/Zg7dy4DBgzAwMDgq69BEARxy1tE5Bvkf2LLuyCIi4tj1qxZtGrViiJFiiCRSPj777+BtG3TSVt+oaiLOUggNiiO23PvcLrXOc72u8D9FQ9Ijk5WmS/2TRxPtj/lys/XON3rHOcHe/DvvLtEPY8C0rbAT786xaLtC6lQoQLv37+nV69eSKVSJBIJEokk2z7UBcXy5ctZuXIla9asoUOHDvme79ChQwQHBzNmzJhsx927dw/IuiAHVLe8zczM8iwoIS1KefXqVe7fv5/lmBkzZhAUFMTOnTuVb6SNGzdm0aJF7NixA/hYlNOxY0eGDx/OL7/8kq1f6tcmPDzNqF+MUIrklTlz5hAREcHKlSsL5f6JiYkoFAoxQiki8o0hCko1CQsLY+7cufj6+lK5cmWVc0mpSezw/RuAxLBEbk6/TXxwAmX7OFOqowMh90K5PfsOCtnHfMrX7kG8vvAa09ImVBhUDsf2DsS9jef6b7cIfRgGpHlfdlzYnlatWgGwbt06du7cqfy3bdu2L/rMBw4cYMKECUyePJmRIwumnd/q1atp2rRplkVH6Xh6elKmTJlsK68LKkIJaQKwZMmSWb5JhoeHs3btWqZOnZoh3WDChAlKq6ZP7aSWLFmCvr4+K1asyPO6CpqwsLSfLTFCKZJX7O3tGTlyJIsWLVJ+QPmaxMbGAoiCUkTkG0MUlGpiZWVFcHAwgYGByu3NdK6+uaKs5PY/FEBqUip159bGsZ09Tt1KU/2XqsS8iiXo0hvlNTYNrXHb1IzKoytRsoUdpTs70nBRPbQMtfDb6w+kbX2HyEIIT0oTAd26daNv377Kf7169fpiz3vt2jX69etH7969mT9/foHMee/ePW7evJmlkfnnY7Pb7oaCKcpJR1NTk7Fjx7J3717ev3+f4fz27dsRBCHT7WuJRKKMuI4aNUppem5kZMTgwYPZvn07CQkJeV5bQSJGKEUKgmnTpiGXy1m0aNFXv3dcXByAuOUtIvKNIQpKNdHR0cHS0jLTc+cDzyqrw4Nvvad4DQv0i3201ClWuSgG1ga8u/ExemVaygRNPdUUVm1jbYqUNyPuTZzymFSiwYuoNI9DQRCIiYn54pY0vr6+dOzYkXr16rF169YC85xbvXo1JUuWpH379tmOS01NxcvLK8sOOekUVFFOOkOGDEFLS4v169erHBcEgfXr19OtW7ccq9vfvHnDjz/+qPweDRs2jOjoaPbt25evtRUU6RHKzHqji4ioi4WFBRMnTmT16tW8ffv2q95bjFCKiHybiIIyn8gUMgKiAhAQSAxPIiU6BdPSGbdpTZ1MiH4Rk+N8yVEpaBtrK79WCHJCEtMsiRwdHTExMcHIyIi+ffvy4cOHgnuQ/+f9+/e0bt0aa2trDh8+jI6OToHMGxISwj///MOoUaPQ1My+FszX15fExMQcI5SfbnnnN4cSwNTUlEGDBrFu3TqSkpKUxy9evIi/v3+2xTWJiWkR6k2bNvHPP/+wZs0aIO171rJlbK7umwAAY+hJREFUS9atW5evtRUU4eHhmJiYoKWlVdhLEfmPM3HiRAwMDJg7d+5XvW96hFIUlCIi3xaioMwnQTGvla0TkyPTRIiOWcZiGV0zXWRxMuSyrL0Ow59EEOkXiXV9K5XjyVpJjBw1gg0bNnDw4EGGDh3Kvn37aNiwITExOYtUdYmLi6Nt27bIZDLOnDmDqalpgc29adMmpFJptlZB6dy7dw+JRELVqlWzHff5lndMTIzaXpJZMXbsWEJCQlTMm0+cOIG9vT0NGjTI8rrExEQ0NDTo3bs348ePZ8KECdy8eROA/v37c/fuXaVXaWESFhYmbneLFAjGxsZMmzaNLVu28OzZs6923/QIpbjlLSLybSEKynwSmfxxm1WeklZ0o6GV8WWV/v8xRUrmRufJUcncX/YAfQt9Snd2VDnn0N6eOUvm0Lt3b7p27cqKFSvYvn07/v7+/PXXXwXyHDKZjO7du+Pv78+ZM2cKtPPOp1ZB6hSDeHp64uzsjLGxcbbjPi/KAfItsMuUKUPbtm1ZsWKFctvax8eHatWqZWvXlJiYqKzwXrx4MXXq1KF79+58+PBBWan+eY/3wiA8PFwsyBEpMEaOHIm1tTUzZsz4avcUt7xFRL5NREGZTz7thKOhnfZyymUZRWN6hbdUO+NLnpqUyp35nqQmplJzSvUMuZUA8s867vTu3RtLS0vc3d3ztX5Ii/CNHDkSd3d3Dh8+TKVKlfI956ccOXKEt2/fqlWMA+oV5EDGCCXkvp93ZowfP56HDx9y5coVALV6q38qKLW0tNi/fz8KhYKePXtib2+Pjo4OPj4++V5bfhEjlCIFia6uLrNnz2b//v1Kq68vTfqWd2F4YIqIiGSNKCjziYHWx22X9K3u9K3vT0mKTELLUAsNLdVWgwqZAs9F94kJjKXmlOoYl8z8U7eBVsY/nra2tkRERORn+UBa94stW7awZcsW3Nzc8j3f56xevZrGjRurJVRlMhkPHjzIsSAHMhblAPkuzAFwdXWlQoUKrFixgpCQEEJCQnK0OfpUUEKaK8D+/fu5du0aM2bMoHz58mKEUuS7pH///pQtW5apU6d+lfvFxsaip6eXYy62iIjI10UUlPnE3the+f965rpoG2sT9Tw6w7go/2hMHFS3cAWFgNeqh4Q9Cqfaz1UoWjHzN3ozHTOMtFWFpiAIvHr1Kl89tQH+/vtvZs6cye+//07//v3zNVdmeHl5cf36dbWjk0+ePCEpKUktQfl5UQ4UTIRSIpEwfvx4jh8/zvnz5wFyLSgBGjZsyOLFi1myZAnGxsbfhKAUI5QiBY2mpibz58/n/PnzXLx48YvfT2y7KCLybSIKynyir6WPpf5HOyGrupZ88AwhMSxReSz0URjx7+KxqqdqO+Sz6THvrgfjMrwCVnUztySSIsVGUiLD8XXr1hEaGqo0Pc8L58+f58cff+THH39k2rRpeZ4nO1avXo2trW22Pbk/Jb0gp0qVKjmO/VJb3gB9+vShSJEibNq0CR0dnRz7lyckJGQQlAA///wz3bp14/bt2/j4+KBQZJ5D+7UQI5QiX4LOnTtTs2ZNpkyZ8sVtzcS2iyIi3ybinkEuWLNmDVFRUbx79w5Iq/598+YNkaHRyGvJ0TDQwKlbKd7dDObmjH9xbGdPalIqAUdfYlTSCFvXj8LwxYmXvDr7GjNnUzS0NXhzWdXLzbJOcTR1NVGgYFnnFQT3/ICLiwu6urpcv36dvXv3UqVKFYYPH56nZ3nw4AHdunWjRYsW/PXXX1+kP3hYWBh79uxh9uzZam9PeXp6Uq5cObXeMD6NUKYX8BSUoNTT02PEiBEsXryYsmXL5rj+zCKU6WvcunUr5cuX582bNzx+/DjHaOeXQqFQEB4eLkYoRQociUTCwoULcXV15ejRo3Tu3PmL3UuMUIqIfJuIgjIXLF26lMDAQOXXhw8f5vDhwwC4bWiKnoEeekX1qD+vDo+3+eK70w+ppgSL6hZUGFROJX8y+mVaNXKkXxSRflEZ7uVavgmaupoYaRvRp08fbt+6zaFDh0hKSqJkyZL8+uuvTJs2DX19/Vw/x+vXr2nTpg1lypRh3759XywXadOmTQAMHTpU7WvULcgB1QilpqYmRkZGBSYoIa3rzYIFC0hOTs5xbFaCEtKqUVesWEG3bt2YNGkSZ86c+SIC/lMEQeDGjRv8888//PXXX9SvX5+uXbuiUCjECKXIF6FZs2Y0b96cadOm0aFDBzQ0NHK+KA+IglJE5NtEFJS54NWrV1meW/dgLedenUWBAiM7I+rMqpXtXFXHVabquMrZjgHoVbYP7TZn31kmN0RGRtK6dWt0dHQ4efLkF9s6Sk1NZd26dfTp00ftiFhKSgoPHz6kT58+ao3/tCgHKJBuOZ9ibW2Nra0tr1+/Ri6XZ/sGmZiYmK24d3BwAODcuXOsXr2acePGFdg6P8fHx4c+ffrw6NEj5X39/f355ZdfgDSj9q5du35xUSvyv8eCBQuoWbMmO3fuZODAgV/kHuKWt4jIt4mYQ1lADKgwCDNdM6QF9JJKJVLKFylPG4e2BTIfQHJyMp07d+b9+/ecPXs2y1aSBcGxY8cICgpSuxgH4PHjxyQnJ6tVkAOqW95AgXTL+ZwyZcqQlJTEiRMnsh2XXYTyU/r06cPEiRO5ceNGQS1RhStXrtCgQQMkEgkXLlzg+fPnCILAhw8fOHToEJCWfzt27Nh8m8CLiHxOjRo16NatG7NmzVLpNlWQiBFKEZFvE1FQFhD6WvpMrjUVDamGsq93XpFKpJhomzCxxiSkkoL5FikUCgYOHMjt27c5fvw4zs7OBTJvVqxatYqGDRuqVVyTzr1795BKpWpf8+mWN6RFKAtaUJqZmWFqasqKFSuyHaeuoBw3bhx169ale/fuvH//voBWmUZAQABt2rShRo0aXL16FTc3N5U+7OmpDYsXL2bdunUsWLCgQO8vIgJpNmRv375l/fr1X2R+UVCKiHybiIKyAHEuUpY59X5HW0Mb8ljoKJVIMdExZUHDRRTTtyiwtU2ZMoV9+/axe/du6tevX2DzZsbDhw+5evVqrqKTkFaQU758ebXzQj+PUH4JQQlgZ2fHlStXePDgQZZj1BWUWlpa7Nu3D0EQ6NmzJ6mpqQWyRoVCwaBBg7CwsODo0aOZdhkKDw8H4KeffmLq1KnMnTs322cSEckLzs7ODBo0iPnz5xdoa9jU1FTkcrm45S0i8o0iCsoCpmJRF2zu2hLxLDJXojI9qlnHqi6rmq7BxtCmwNa0du1aFi9ezLJly+jatWuBzZsVq1evxsbGhk6dOuXqOk9PT7W3u+HrRCgBLCwssLOzY+XKlVmOUVdQwkfT8+vXrzNlypQCWePOnTu5du0a27Zty/LNNiwsDCMjI7S1tZkxYwblypVj1KhRBXJ/EZFPmTVrFrGxsSxbtixf87x79465c+dib2+PlpYWmpqa+Pn58eTJkwLNlxYREck/oqAsYK5cucLK31fRUaMzo6qOwfr/haFUIs24FS6AIE8TQ2XMnJlaezqTa03FRMekwNZz7Ngxxo0bx88//8z48eMLbN6sCA8PZ/fu3YwaNQotLS21r0tOTubRo0dqV3jDly/K+ZQxY8awZ88ePnz4kOn5nATl5958DRs2ZMmSJSxdupSDBw/me31r1qyhdevWNGnSJMsxn3pQamtrM3fuXG7dusX9+/fzfX8RkU8pUaIEY8eO5c8//yQkJCRPc5w/fx5nZ2cWLVqkrB6fP38+EomES5cuUb58eby8vAp45SIiInlFFJQFSFJSEoMGDaJhw4ZMnDCRVvatWee6gT8aLKKncy9qWNYk8UMiOjIdShiWoL5NA5JuJOO/5AW/155PHau6Bbqef//9l169etGlSxeWLl1aoHNnxZYtWxAEgR9//DFX1/n4+CCTyXIVofwaRTnm5ua8f/+eoUOHoqmpmWVeWE6CMjg4WDlfOuPHj6dHjx4MGjSIp0+f5nmNnp6eeHp6MnLkyGzHfd4lp127dtjY2HyxXDeR/20mT56MVCrNU67u/v37adu2LY0aNeLt27ds2rSJefPmKds7zpkzhxIlStCoUSOuXr1a0EsXERHJC4JIgbFz504BEB4/fpzpeYVCIWhqagpr165VHnv8+LEACLt27SrQtfj7+wtFixYV6tevLyQmJhbo3Fkhk8kEOzs7YeDAgbm+dsOGDYKGhoaQkJCg9jUfPnwQAOH48eOCIAjC8uXLBQMDg1zfOzvWrFkjaGpqCsnJycKoUaMECwsLISkpKcM4Y2NjYcmSJVnOs2DBAsHY2FhQKBQqx2NjY4Vy5coJ5cqVE2JjY/O0xqFDhwq2trZCampqtuO6dOkitGzZUuXY7NmzBQMDAyEqKipP9xYRyY558+YJ2trawqtXr9S+5vnz54K+vr7Qq1cvQSaTqZxLSkoSAGH79u1CXFyc0KRJE8Ha2lqIiIgo6KWLiIjkEjFCWYCsX7+eZs2aUb58+UzPJycnk5qaqlIwUb58eZo2bVqgUaL0lozm5uYcO3YMXV3dAps7O06cOMHr169zXYwDaVG2ChUqqJ2HCJkX5cTHxyOTyXJ9/6xwcXEhNTWVZ8+eMW7cOEJCQti7dy8AckHO65jX3P9wD8OyBiSaJBCbEpvpPD4+PlSsWDGD96OhoSGHDx8mKCiIoUOH5rptXWJiInv27OHHH3/M0Ug6sz7eQ4cOJSkpif379+fqviIi6vDTTz9hamrK7Nmz1Rr/aXHZxo0bMzRdiI1N+/0yMjLCwMCAnTt3Eh8f/0V9XUVERNRDFJQFhLe3Nzdu3Mh22zG94vFzy4uRI0dy/fp1vL29872OhIQE2rdvT1xcHGfOnPmqXVFWrVpFvXr1qFatWq6vzW1BDmRelAMQHR2d6/tnRXqbRG9vb5ydnWnTrg0bzq9n2vUp/HCiG2MujmT2rZlUn1KVu8X+pc/pngw6O4At3pt4G/exnaa3t3eWLRfLli3Ltm3b2LdvX7aFP5nx+PFjEhIS1OrpnlkfbxsbG1xcXLhz506u7isiog6GhobMmDGDHTt28Pjx4xzHnzlzhmvXrrFp06ZMi8vi4uKU80JarubSpUvZtWuXWvOLiIh8OURBWUCsX78eS0tLOnbsmOWY9E/Xn1u6dOzYkeLFi7Nhw4Z8rUEul9O7d298fHw4efKkskvK18Db25vLly/nKVKQlJSEj49PrgpyIGNRjpmZGUCBFuaYmZlhY2ODj48Pt97dxHCIHkW7F8EnzJsURUqm14QnhXHixXFGug9j0Z0/CIsL5enTp9n28O7WrRsTJ05k0qRJXL9+Xe31eXt7I5FIsoyKf0pmEUpIE80+Pj5q31NEJDcMGzYMOzs7pk+fnuPY9evXU7VqVVxdXTM9/2mEMp3+/ftjYWGR77+fIiIi+UMUlAWEu7s73bp1y7ayObM/hpBWcdutWzfc3d3zfH9BEBg3bhwnT55k//79uY725ZfVq1djbW1Nly5dcn2tt7d3rgtyIPMtb6DAC3MqVqmId5EH/HFnPkmkdf8QcvCEUghpa7oVfJNRF0diVsk0W0EJsHDhQurVq0ePHj3UNj339vbG0dERAwODbMcJgpBphBL+r737DovqzB44/r0zMPQmXWyIgAiILXYTE9PUuNFo1JjeTDWmbHp2k3VNd9M00RSjJppmqiaWRGNJ7FhBFBEFRUBEep12f39MGEU6DDj8cj7Ps8/K3Pe+vJdQzpzzFoiJiSExMdH6dRTClnQ6Hf/973/58ccf2b59e53t0tPT+eWXX3jggQfqPBa06nfo+dlLnU7H3XffzZIlSygtLbXt4IUQjSYBpQ2Ul5dz9OhR4uLqP5u7rpI3QFxcHCkpKZSXlzdrDG+++SYffPAB8+fPZ8yYMc3qo7ny8vJYunQpDzzwQJO2Cqqye/duHBwc6N27d5Puq6vkbcuAstJYgddkd9RQy+doKJC8kFk1U2EqZ+BzAygNKKm3rYODA19//TUAU6ZMadRc0PpK6ecrKirCaDTWmaEsKSkhPT29wX6EaI6bbrqJ2NhYnnnmmTrnCVeVuW+66aY6+6kqeV/4O/Tee++luLjYOr9ZCNH2JKC0gaSkJMxmc4N/2OsqeYPlj7rZbObQoUNN/vxffvklTz/9NC+88EKTt+uxhU8//RSTycT06dObdX98fDwxMTFNXjzU2hlKVVV5M/4NKjwrUDQtOE5Tsfzvw+T5JJ2tf55XUFAQ33zzDVu3buWZZ55psOuqxT4NqTolp7YM5fnzRIVoDVqtlldeeYVNmzbx66+/1tpm+fLlTJ06td5TcOqq8oSGhjJq1ChZXCbERSQBpQ1UzT9raB5bfRnK6Ojoan011saNG7njjju47bbbmDVrVpPutQWTycT777/P1KlTCQho3lGRzVmQAzUzlB4eHiiK0qiActeuXTz88MNER0fj5uZGly5dmDx5MkeOHLG22XDyd3Zm76DoZDHbZ+1k1U1rWXPrb+x5Zx+VhZU1+izNKiX+jT2sueVXfpmyhj+f3UZugiWQUxQFVHhr9/+oNFYwb948oqKicHJyIiQkhMcff9xarhs+fDhz5szhrbfeYvny5XU+Q25uLtnZ2Y3KUObm5gLUmqEMCQnB29tbAkrRqsaOHcuwYcN47rnnakyvKCsrIyUlhcGDB9fbR20l7yqDBw+W72EhLiIJKG0gISGB0NDQWgPF8xUXF6MoSq3z3Tw8POjWrVuTfiEePHiQ8ePHc+mll/Lxxx/XOe+oNf3888+kpaU1a6sgsEwXOHjwYJMX5EDNRTkajabRp+W8/vrrfPfdd4waNYp3332X6dOns3nzZvr160diYiKlhlI+PDCf8txytr6wndKsMnreHEnY9aHk7D7D9pd2Yjac+6NYnlvOn89sI+9QHmHjuxN1SySmCiPb/7OTswfzLOPEzJmyHCbeN5EZM2YQExPDu+++y8SJE5k7d261+aePPPIIU6ZM4a677qoza52amgpAREREg89bX4ZSURQiIyM5evRog/0I0VyKovDaa6+xZ8+eGqdDJSUloapqg9n2kpISdDodOp2uxrWYmBiysrKs3+tCiLbl0HAT0ZCmzGOryqLVJjY2ttEBZWZmJqNHj6Zr16589913tf6CbQvvvfcegwcPbvYioAMHDmA0Gpt1/4Ulb2j8ed6PP/44X3zxRbWv25QpU4iNjeW1115j6n8mU2GsIOW7VIwVRkbMGYarv2WPTO9wb7a/tJOTGzLoenUXAI5+l4qh1MDId0fgHmLJnnS5ugsbHt7EwU+TuPR/wwEozytn3ZINTLtlGss+X2b93BEREcyYMYOVK1cybtw4FEXhk08+YdCgQUycOJEdO3bUeMNiMpkAGvXfvipDWdc2UjqdThbliFY3fPhwxowZwwsvvMCECROsc66rdiuoqtTUpbi4uM437udP3ajvCFIhROuQDKUNJCUlNWoeW3Fxca3zJ6vExMQ0ai+1oqIixowZg6qqrFq1qt4+W9PBgwf5/fffW7Sp8O7du3F0dGxUQH6hC0ve0PiAcujQoTUCsfDwcKKjozl06BA/H1sJQNa2bAIHBFiDSQD/OD/cOrqRuSXL+trZQ/l4hXpag0kAByctQZcEUnisiJJMSzk7P7kA1aQSM6r698vUqVMBqi0qcHd357vvvuPkyZPcfffdTd70/Hxnz57F1dW1SRvHC9EaXnnlFVJSUli0aJH1tcbuVlBcXFznHMvw8HB0Op2UvYW4SCSgtIHS0lK8vLwabFeVoaxL1Ukv9TEYDEyaNIm0tDRWr15NSEhIk8drK/PmzSMoKIiJEyc2u4/4+HhiY2NxcnJq8r0tyVDWRlVVTp8+jVcHLzJLMyk7W46+UI93j5r/bb3DvSg8VmT92Gwwo9HVPKlG62R5rTDVstm62WgZ66nKjGrtXF1dAUuAfb6qTc+XL1/OO++806zngrr3oBSircXFxTFt2jT+85//UFZWBjS+ylNSUlLn71BHR0eioqIkoBTiIpGAsg3VV65pDFVVuffee9m4cSM//PBDo7KirSU/P5/PPvuMBx54oEXl9uYuyIG6M5TN3dh82bJlnDp1iuFjhgFQmW/Zc9LJp+bqc2cfZwwlBkwGS9nZPcSN4vQijOXGau3yDlnmT1bkWfpy62jJwOzavqtauz/++AOAU6dOcaFJkybxz3/+kyeffNLarqnq2oNSiIth1qxZ5OTkMG/ePKDxAWVDv0ObMm1ICGFbElC2oYZK3g156aWXWLJkCYsXL+byyy+34ciabtGiRRgMhmZvFQSWlZ1JSUnNWpADNRflgOVkm+ZkKA8fPsxDDz3EkCFD6De2HwoKJr2lX61jzR8TzV+vmf9q0/WaLhhKjeyes5fCY4WUnCohcWESBX9lJqv68g7zwjvCmz1f7WHRokXWTPN9992Ho6NjnfuQvvrqqwwfPpzJkyeTlZVVa5v6SIZS2JOwsDCmT5/Oq6++Sn5+Pjk5OY2qttRX8gbLUYynT5+25VCFEI0kAWUbaqjkXZ9PPvmEWbNm8dprrzFt2jQbj6xpTCYT8+bNY/LkyQQFBTW7n/3792MymZqdobRVyTs7O5uxY8fi5eXFt99+i6qYURQFrc7y42Ey1FysUrXCW/NXm8D+AcTc24uzSXlsfmILGx7eTM7uHHpOiwTAwflcOXzAU/3wDPXkrrvuIjQ0lHHjxjF58mT69u1b5x9LBwcHvvrqKxRFafSm5+eTDKWwNy+88AJ6vZ4333wToFG7VNRX8m5sH0KI1iGrvNtQcXExgYGBTb5v9erV3H///TzwwAM89dRTrTCyplm1ahXHjx9v8akUu3fvRqfTtah0ryhKsxblVCksLGT06NEUFBTwxx9/0LFjR3THnFBV1Vrqrip9n68ivwJHd0e0jucCxdAx3eh8RSeK0orROGrw6ubJifUngXOlbgAXX2eueH0kr/R8nezsbMLDwwkKCqJjx471bgEUFBTE8uXLGTlyJE8//TRTpkwBQK+v/Uzx8+Xm5ta7grayshKttuYcUCFaS3BwMI8++miT5gYXFxe36E2sEKL1SIayDTUnQ7l7925uvPFGxo4dy9y5c+3iHfjcuXMZOHAgAwcObFE/8fHx9O7du0VzMBVFaXaGsqKignHjxnHkyBF+/vln68b0nTw6o6Li4uuMzlNHwdHCGvcWpBTiFVpz+oKDswMdevrgHeaFolU4sz8XjU5Dh54+1dp1cu9EeHg4I0aMICgoiKSkJLKysrjyyivrHfOwYcP43//+x9tvv23dBD85ObnBZ60vQ6mqKsnJyYSHhzfYjxC29OSTT+Lk5NToHQxKSkrqLXkLIS4eCShtpDG/EBuaQ3lhH2lpaYwdO5bo6Gi+/PJLu8ggHTp0iN9++61FWwVVacmCnCoajaba183Hx4eKigoqKmpmFc9nMpmYMmUK27ZtY/ny5QwZMsR6LcwrzPrv4CFBnI7PoTz33NzGMwdyKc0sJXho/ZmSvMP5ZG8/TZcrO+Podu6Mc62iJaJDpPVjs9nMU089haurK/fff3+DzzxjxgxuuukmZs6cib+/f4OLEFRVrXcOZUZGBoWFhc3aukmIlvD29rYeMXrmzJkG27d0YaMQovVIydsGgoODOXHiRIPtGspQpqen07FjRwDy8vIYPXo07u7urFy50rqtzMU2b948AgMDufHGG1vUT2lpKYcOHeLRRx9tUT8ajaZGhhIs53nXVxp74oknWLFiBePGjSMvL4+lS5dWu96jUzipBUcJnxRG5tYstv5rB92v64axwkjqj8fx6OpB51GdrO3LcsrZPWcPgZcE4uzjRPGJYtLWnsCjmwdRt0RW63v/xwngo4GhGgwGA1988QU7d+5kyZIldOnSpcFnVhSFjz76iP3795Oens7evXurXS/WF7MvZy9HC1JIK0qjpLKE/v/qS0pQMj8fW0m0bzShXt2t7asC0ou5a4D4+3r44Yd5+umnWbFiBc8//3y9bRtalCOEuHgkoLSBxm5V0VCGsmrrjIqKCq6//npyc3PZunVrs8/ItrXCwkKWLFnCE0880eKTefbt24fZbG5xhrK2kjc0HFDu27cPgJUrV7Jy5coa139PX8/be/6Hi58Lw2YP5uCiQxz6PBmNg0JA/wCi74yqNn/SwdUBJx9n0lanYyg24OzrRPex3Qi/sQcOLtV/zDpGBnN8fRpP/vAkGo2GgQMHsn79+iat3Hd3d+f7778nNjaWTZs2oaoqJ4tP8H3K92zO2IhRNaJVtJhUy7ZGfjG+pJHGxwc+REUlzCuMf4SN57LOI0lISMDDw4OuXbs2+vMLYSuurq4oisLOnTvZt28fffr0qbNtQ4tyhBAXjwSUNhAbG8uvv/6KqqrV5jiaVTOZJZkcK0wlryyPkCs7ku+TR0ZxBh3dO6JRzs04UFWVxMRErr32Wm677Tbi4+PZsGGDXc1rW7RoEZWVlY0qyzYkPj4eJyenBo9aa8iFJe/zA8r6bNy4sd7repOeJUmLyK/Ix6OLB4NfrH++qM7dkYHPNm77o6cefJp//O/6RrWtT2RkJNOnT+eDBR/w8CcPciogA1Aw/xVEVgWTVVTOBd7HCo/x9p7/sTZtNSeOZRATE2MX83PF35O7uzs6nY7nn3+eX375pdY2RqORioqKegPKvLw8CTiFuEgkoLSBmJgY8vPzyczMJCQkhPSiNFYfX8XvJ9ZTYbLM5VNQiLmnFzs029ixfhvOWmeu6DKK0aFj6OrZjVOnTlFQUMCuXbtYsWIF33//PYMHD77IT3aO2Wxm3rx53HjjjQQHB7e4v927dxMXF2c9y7e56stQtoROq2Nmv8d5cesLLernfBpFQ5hXGGO7X2ezPqfdfhN7O+7mhF86itr4gFDFEoQfzk/GOMpARAf7eeMi/n5iYmJwcHBg1apVbN68mUsvvbRGm+LiYoB6S96JiYkydUOIi0QW5dhA1WKGXQd28dbuOcz4/SHWpK2xBpNg+QOuaM79wa8wVbAmbTUzfn+It3f/j/gDlpNTfvrpJ+bOncv48ePb9Bkasnr1alJTU22yGAdssyAHal+UAzT7tJzz9Q3oy7ju/2hxPwAaNOg0Oh7r/wRaxTaLqyqNFXxX8i0dIn2anV00qyYUB4WCgXkk5ibaZFxCNFVMTAxFRUX069ePZ599tsYCxQpjBSfzTuLcwQkX99rPo6+q8khAKcTFIRlKG+jWrRvBccF8XroYU4alzGi+oNxYG7NqyaxtytjIxvIN+ER6c/e4e3jooYdadbzNMXfuXAYMGMCgQYMabGsymcjPz8fZ2Rk3N7cawU5JSQmHDx/mySefbPG4LlyU4+rqioODQ4szlFXuir2HEkMJG07+3uw+NIolmJw1bDadPDrbZFwAiw8u4mhhCoq2ZaVqRaOgovLqzpeZf+WHeOqaf5qTEM0RGxvLkiVL+OGHHxg7diwrf15J8CVBbM7YxOG8Q5wus5x+c9XCUXxcuYBf1q2kl28vrug8il6+0SiKwsmTJ2W3AiEuIgkobeBQXhL9X+iDQWOAxm2nVo1ZNWN2VBk6azC3X36b7QfYQsnJyaxdu5bPPvus3kzY/v37+eCDD1i2bBmlpaUA9OzZk/vuu48777wTLy8vAPbu3Yuqqs0+cvF8F5a8FUVp1mk5ddEqWmb2e4xgt458lfwFcO6NQKPGh0KIewhPDniGbl7dWjSWjRs31rlwZ/hrQ/CJ9KnxuqHUwO8PbkJfpKf/k33pOLT6dIWC1EIOL0sm/3ABqqqyK2433yz4pt6FEULYWmxsLHq9ni5dunDlfaNYcPZ9HLY7oFE0NX7eVFROlWSQXZrFb+m/0sm9E/f2vp/MhExrX0KIticBZQudKcvhP9tetGSJWpAo0mgV0CjM2v4S749agL+rv+0G2ULz5s0jICCAyZMn19lm7ty5zJw5k44dO/LEE0/g5uaGr68vv/76K08++STvvfcea9asISIigvj4eJydna0bibfEhSVvaN7xi/V+DkXD1J43MSh4EPP3f8DhvEPVVlBfSMGS8XPSOjOhxw3cGDkZR03L5oqe75FHHiHF/QjZpefO9HYNdqu1bfKXRzDpax9nQWohW57bhouvMxFTeqCaIW1NOpdedim7du4iMjKy1vuEsLXY2Fh0Ho68nzwXl2udUM2Wn+n63rxV/fydKjnFi1tfIKAgEG8/70ZtvSWEsD2ZQ9mAkpISXnzxRa699lo6dOiAoigsXrwYsMzZeXfPO+jN+mrBpNloZsOMzaycsIrUH4/V23/GplOsnLCKVTetBQX0Zj3v7X3HGiTNmzePqKgonJycCAkJ4fHHH7dm/9pCUVERixcvZvr06Tg5OdXa5sUXX+SRRx7hscceIy0tjf/85z889dRT3H333Xz99dccOXIEZ2dnhg4dyv79+9m9ezd9+vTBwaHl72cuzFCC7QPKKqFe3Xnj0jm8e/m8vxZTda22Uh/A3dGdPgF9ebDPw3w+ehnTom62aTAJENEvHIdLNHQaGWL9n5NnzW2citKLSVtzgh4TutfSiyXY1Oq0DH9tKGHXd6fHhO5c+tow9EY9zz33nE3HLES9XOGyOSM4pckAqDbfvCFVC8xyPE8z+KWBlBvLG7hDCNEaJEPZgNzcXGbNmkWXLl2Ii4urtt3MtsytHMjdX+Oe47+kVTtZpS7GciOHPjuM1vncIg2zamb/mX1sy9zKT++t4I033mDSpEnMnDmTpKQk5s6dy8GDB1m7dq1Nnq8hixcvpqKios6tgtatW8esWbOYPXt2nZsSh4aG8ueff3LFFVcwbdo0DAYD11xzjU3GV1uG0sfHxyaLcuoS6hXK9N6Wr4fepKegsgCTasJZ64y3k3erb7+TdDYJs5MKjioabd3vCQ8uTCJoUCAdenWo9XpeUj7+ff3QnReM6nx0dOjlw88//yzH3Ik2UW4s519bnsPZz7lFVR404NjRgZd3zGLWsJdttvhNCNE4ElA2IDg4mKysLIKCgoiPj+eSSy6xXvsp9Uc0aDCft79fZUElR745So8J3Un+MqXevlOWH0Xr4oBvjC/ZO09bX9eg4Yudy/jwrY+59dZb+eyzz6zXIiIimDFjBitXrmTcuHE2fNKaqrYKmjhxIiEhITWuFxcXc/fdd3PFFVfw7LPP1ttXhw4dWLp0Kf3790ev1zd4IkZjXbgoB1ovQ1kbnVZHgGvbbjz/yQufoC83oGgUOvTyodftPfHu4V2tTeaWLPKS87l87qWU5dT+5sZsMKPV1fJHVwd6vZ7ExES72rpK/P/02cHFnCo5ZZt6mQIJuQmsTF3B+B4TbNChEKKxpOTdACcnp1pPXDlVcopDeUnVgkmAQ58n4x7iRqfLagZg5yvJLOXYyjSi74yqsUrXjJkd23dgNBqZOnVqtWtVH3/11VfNeZwmWbt2LSkpKXVuFfTxxx+TnZ3NwoUL0Wga/laKiYnh5ptvBqB799rLsE3VliXvi02n0zFx4kR63xPLJc/2J3JaBEXpxWx5fjuFxwqt7UyVJpKWHKL7uFBcA+o+stMtxI38IwWopnMZXrPBTEFKAQCnTp1qtWcRf18HDx7kxhtvpHv37ri4ujBjxEz+fH4r2btO12hbfLKE7bN2suqmtay59Tf2vLOPysLKGu0q8irY/0EC6+7bwC9T1nDL8Fu4b8Z9nD17ti0eSQiBBJTNlnT2YI3X8o8UcHJjBtF39WqwdHPw0yR8YzsQ2L/27JbZaAmSXFyq77lWdab37t27mzHqppk7dy79+vVjyJAhNcdnNrNgwQImTZpEt27dGt1nVdutW7faZIx1LcrJz88nMTGRZcuWsWvXLiora/4Ram+GDh3KN8u/IfiKIIIGBhI+MYwRrw8F4NDSZGu7o9+nYjaqhE8Mq7e/btd2oTSzlH3vH6D4ZDFF6cXsfW8/FfmWr1V5ucxFE7aXnp5OcXExt99+O9fMvJrIyZZN9Xe9spv0X09Y25XnlrP1he2UZpXR8+ZIwq4PJWf3Gba/tBOz4dybSGO5kT+f2Ub2jtN0HhlCzD3RBPQPYOGChVx55ZU13nAKIVqHlLybKbXgaLWVvqqqkvjJQUKGBdOhpw9lOWV13ns6Pocz+3K57O3hdbbxDLEcH7Zly5ZqW8X88ccfQOtnj44cOcLq1atZtGhRrXMCf//9d1JSUli4cGGT+j18+DD+/v58/PHHPPnkk43KbNbnwgylXq8nKSmJ48ePV9s+xM/Pj7vuuosnnnjCbs5GtwW3YDeCBgaSvf00qkml/Gw5R388Ruz06BpniF+o27VdKc+tIPWnY2RssHw/efXwosf47qR8myrzJ0WrGDNmDGPGjCG3PJe7195BKN3oNrorm//5J6krjtP1assq7ZTvUjFWGBkxZxiu/pY31t7h3mx/aScnN2RY22XvOk35mXIGPj+AwAHnfrZdPF3Y99U+9u/fT9++fdv+QYX4m5GAspnOlOVU2zbm5O8ZFKUXM+CpfvXeZzaYOfhpEl2v6YJH57rPnPXo7kGn6BBef/11QkJCuPzyyzl06BAPPPAAjo6OrZ49ev/99/Hz86tRcq+yYMECoqOjGT687qC4NvHx8Vx22WV8++23rF+/nquuuqpF4zw/Q1lUVMQNN9zAhg0bUFWVtWvXEhQUREVFBV999RUfffQRX3/9NWvWrKFnz54t+rwXi0bR4OrgRpnx3Ep/Fz8XzEYzxkojyV+m4OLrjF+Mr/VNTeVfGUd9kZ6ynDJc/Fysq2ijbokkbHx3ik8U4+jmgGdXT2u2MyIioo2fTvyd7MvZa12hrWgVXPxcKDhaYL2etS2bwAEB1mASwD/OD7eObmRuybIGlMYyIwBO3tV3OtB6Wd6sXljlEUK0Dil5N5PxvGDSUGbg8NJkeozvjotf/b+8jq08jr7YQOTUhs9OnvTyROLi4rjrrrsIDQ1l3LhxTJ48mb59+7Zq9qi4uJhFixYxffp0nJ2da1zX6/WsWLGCO++8s0krmgsLC0lJSWHcuHFERkby7bfftnisVYtyysrKGDlyJPHx8Tz//POoqsqwYcPo3bs3AwcO5K233uLAgQO4u7szbNgwDh8+3OLPfbH08O5R7eOy02VodBocnB0oP1NOaVYZ6+/fyPr7LP/b89Y+ABI+PMj6+zZiLDdWu1/n7ohvrw54drWckJO7P5eOIR3bbdAt2oeDWQcxFhspzSoldcVxcvacwa+3HwDlZyvQF+rx7uFV4z7vcC8KjxVZP/aN7gAaSFx4iPzkfMpzyzm9O4eUb1MZeOVA+T4Woo1IhrKZXBxcrBtYp/54HLNRpePwYGtWqDzXco63vsRAWU4Zzj7OmPQmjiw/SrfRXTGWG61/2E0VJlChLKcMrU6Lk7cTCgqBwYH8+eefpKSkkJ2dTXh4OEFBQXTs2LFVs0dLliyhrKyMBx54oNbrycnJGAwGBg4c2KR+9+zZA8CAAQO45JJLOHDgQIvHWlXyfuGFFzh06BDbt28nM9NyYkZBQQFubuc2/O7cuTN//PEHgwcP5vbbb2fLli022QuzrZw5cwZ/f396dogi8WwCZtVM4fEisnedJqCvP4pGIXJaBPpifbX7ik+UkPzFEcImdMcn0hutU93bqZz6M5OCo4U8/+YLLZ6OIER9Fr+6mH0r/tp2TQPBg4KIuTcagMp8y+9PJ5+ab2idfZwxlBgwGUxoHbV4dPYg7oFYkhYf4s9ntlnbdb68E1NervswBiGEbbWfv6Z2pqtnN7ZnbUNVVcpzyzGUGNj4yB812h39LpWj36Vy6VvDcXRzwFRhIvWHY6T+UHPD8/X3bSRwYCADn+2Poih08+wGQHh4OOHhloxmUlISWVlZ3HHHHa3yXFVbBd1www106tSp1jYJCQmAZdV2U8THx+Pm5kZkZCSxsbH8+OOPqKraon0bNRoNJ0+e5PPPP2fOnDnExcVRVmYJ6gsKCmpsd+Tj48OSJUsYNmwYb731Fk899VSzP3dbmzJlCi4uLkT3j+Z4bhrFGSWc+PUkWp2WqFstWRjfWvacdHSzrHT17uFF8KBzOxacPZjHkW9S8O/jh85DR35yASd/zyBmWAyPPvpomzyT+PvqfUMszv2cqMirIHNLFqpZRf1rMaJJb/l/rWPNNzWav14z681oHS1vjpw7OOMd7k1AP39cAlzIS8rn+C9p/PTeTzy+9Ik2eiIh/t4koGymHt49rMeChY7tRtCgwGrX9YV6DsxPpPMVIQQODMQ1wAXFQcOAZ2rOsTz+czr5R/Lp93gfnP96R25WzYRdUNo0m8089dRTuLq61rnReEv99ttvJCcn8/HHH9fZJjExkU6dOuHjU/Ps6Prs3r2bfv36odVqiYmJoaSkhPT09CatEr+Qoihs2bKF2NhYZs6cCWAdV11bBw0ePJj77ruPOXPmMHPmzDpPALI348ePZ9myZSx8fyEFhQU4ejoSNDiQyCnhuNVx9GJ9nDs4oWgUUn88jrHciGuAC5HTIlj57sp2lbkV7VNgaCD5XnmAJZu47aWd7Hw5nuFvDEWrswSNJkPNFdpVK7w1f7XJO5Rnue/1Idb9WIMHBeHo6sgfX/xJ0nNJNjnmVQhRP/mr0Qjz5s2joKDAWkpduXIlx08c59iRNDqPDsE7zAvvsOpzfapK3x6dPaplhc7/d5XsHacpOKpUu+akdWbxy0v4RL+QPn36YDAY+OKLL9i5cydLlixptfNq586dS58+fepdbJOQkNDk7CRYMpRVm7FXrcBOSEhoUUBpMplIT0/no48+Qqu1ZCu8vb0B6j0tZ8aMGcyfP58ffvihzoVH9uaRRx6x7gl6tCCFJzY+Zl3UUB+/GF/G/TCmxutuwW4MfvHctAUFhWu7jaZbh242G7MQdQlxDyEl/4h1cWPHoUEcmJ9IaWaptdRdVfo+X0V+BY7ujtbsZPqvJ3Hy1tXY3D9oYCDJXx1h69atElAK0QYkoGyEOXPmkJ6ebv34+++/5/vvvweg46VBOLrZ9qxmjaLh6q5X49jfiXfeeYdly5ah0WgYOHAg69evr7aNUJVjx47xxx9/oNfrufTSS4mMjGzy5z169CirVq3ik08+qbcMnZCQwOTJTZublJ+fT2pqKgMGDACgU6dOeHl5kZiY2KITf4qLi9HpdNx0003W16oCyvo2N4+KiuKyyy5j/vz57SagPF8P73BujJjM8iPfNCqobIhG0eDj1IE7Yu6yweiEaFiYdw9+P7He+rGp0hJYGkqNuIe4o/PUUXC0sMZ9BSmFeIV6Wj+uLKhENdf8GTAZLP0ZjcYa14QQtiez7hshLS0NVVVr/C+rJBOPoNq3/nENcGXcD2MIG9/wiTB9H4ljzJfnzrbWKlrGhf2DO+64g3379lFSUkJRURHr1q2rEUyuXbuWUaNGERYWxh133MH06dPp2bMnw4cP54cffmjSc77//vt06NChWnB2oaKiItLT06vt8dgYVQty+vfvD1hK1TExMdb5mM1hMBgoLCwkKiqq2qp3Z2dnnJycGjwt54EHHmDz5s0kJSU1ewwX09Se0+gX2B+lRQcgW4JJnUbHvwb/GxcH2WJFtK6cnBwAYv1irW+GzEYzGRtPodFp8Ohs+VkOHhLE6fgcynPPbZF25kAupZmlBA89V81x6+hGZYGe3MTqp+Jk/ZkFIHtQCtFGJKBsgSC3YG7rdYfN+709+k6C3IIbbPe///2Pa6+9lrKyMj7//HNOnjzJyZMnWb58OTqdjhtuuIF//etfNU6SqU1JSQmffvop9957b737tuXlWeY8BQc3PL7zxcfH4+7uXm11eseOHTlz5kyT+jnf3r17MZlMREVF1bjWmOMXJ0yYgJubG6tXr272GC4mB40Dzw58noHBg5rdh0bR4ObgxisjXqO7d/0n6whhC/fddx+jRo1i8dtLqNyiJ2V5Kpse/YPCY0X0nBZh3ZA/fFIYWictW/+1g+O/pJHy3VF2v7kXj64edB51bsFg6JiuaJ217Hw5nkNLk0lbe4I9b+0jdeVxrrrqKgYNav7PhxCi8aTk3ULXhY1jb84e9ubsRaVlR3wpaOgb0JfrujdcAn7llVd4/vnnefbZZ3n55ZerlagnTZrExIkTefPNN3n66acpLy9nzpw59fb32WefUVpaWudWQTXG2sSV2bt376Z///7VtqJpyepuOLfa3NfXt8Y1Hx+fBgNKnU5HdHR0i7KkF5tOq+O5gS+wNn0NnyR8hNFstC4Wq49G0WBWzQwKGswDfR7C28m79QcrBJbdChYuXMj8+fPJPZuLxlmDV5gnUbf1JGjgucWNLn4uDJs9mIOLDnHo82Q0DgoB/QOIvjPKOn8SwD3EnUvnDOPwsiNkbDpFZUElzj7O3PXwXcx9fe7FeEQh/pYkoGwhraLl2YHPMXvHfzlwZn+z57MpKPT2782zA59Do9SfON69ezf//ve/ef7555k9e3bt/SkKTz31FA4ODjzxxBOMGTOGK664ota2qqoyb948xo8f32qLfeLj45kwYYJN+0xMTESn09W6IrnqPO+GxMTEsH//fpuOq60pimUxzYDAS/jl2ErWpK2m1FCKgoKiKJatmf76d9UCiH4B/RkX9g/6BtR/spMQtjZ16lTrvGVVVXlp27/Zf2ZfrW+EPLp4VFs4Vhf3EHfrKWWWxWVjeKDPg7YduBCiXhJQ2oCTgzP/HvwSXxxeyvcp36GgYG5ktlKDBhWVG8InMa3nzThq61/gYzQaueOOO+jduzcvvvhig/0/+uijrFy5krvuuoukpCRcXV1rtFm/fj2HDh1i/vz5jRpzU509e5bjx49bF+TYSkJCAs7OztXO8q7SmJI3WFabf/HFF5hMJusq8fbKz8WP26Pv5KaeN5Ocd5ijBUc5UZxOubEcraIlwDWAMO8eRHWIwtfF72IPVwgURWFG35nMWP8g5cbyRv/erItG0eDn4s8dMXfaaIRCiMaSgNJGHLWO3B59J4ODh7Dk4GISzyZYy4oXqsoWmVUzvXyjuSPmTiJ8Grcq+5dffiExMZEdO3bg6Njw6nKNRsNHH31EZGQkX3/9NXfeWfMX7XvvvUfv3r259NJLGzWGpjr/hBxbSkhIwMXFpc6Asmryf31iY2OpqKggNTX1/83Z1Tqtjlj/3sT6977YQxGiQX4ufswaNpvn/3wWvUnf7KBSo2jw0nkxe9jLsrhMiItAFuXYWGSHnrwy4jU+GLWASeGT6ePfF3fHcyuQ3R3difPvw6TwyXwwagGvjHit0cEkwIIFCxg4cGCTjj0MDw/nmmuuYcGCBTWuHTt2jJ9//pkZM2a0eE5jXeLj4/H09CQszHaLPnJycsjJycHFxaXWRUeNmUMJ5077ac/zKIVo78J9InhtxBv4ufo3e9eCbp7dePOytxq1oFEIYXuSoWwlnTw6c0uvW60fVwU9LQnajh07xtq1a1m4cGGT773//vsZP348e/bsoV+/c/Pm3n//fXx8fJg2bVqzx9WQ2hbktNTBgwcBcHV1rTND2Zg5lIGBgfj7+5OYmMjEiRNtNj4hRNN09w7j/Ss+4PNDn7EydUWjpg4pKGg1Wm7qeTM39JiIVtO+p60I0Z5JhrKNKIrS4gzgxx9/jJeXF1OmTGnyvWPHjqVTp07VspQlJSUsXLiQe+65p9a5lbXx8LDsu3n27NkGWp4THx9v3X/yfLm5uXh5edVyR8OKiooAcHR0rDVD2dg5lGBZJV7VnxDi4nFycOae2Ol8es0Spva8iUDXwFrbKSh08ejCXTF3s+TapdwYMVmCSSEuMslQtiNr165l/PjxjQ7+zufg4MCUKVP49ttvra8tXbqU4uJiHnyw8ashfX19CQoKavRpObm5uaSnp9c6fzIxMbHR2xTVRaPR1JmhLCwsxGw22zQzKoRofb4uvkztOY2pPadRaigl+cxhxk4Yy8xHZzL+mvF08wzF2cH5Yg9TCHEe+UvbThiNRpKSkujTp0+z+4iLiyM9PZ2ioiLrVkHXX389Xbt2bVI/sbGxdc45VFWV3PJcThSdIKM4g23x24CaC3Kq5kA250zw8ymKUmdAaTabKSkpaVH/QoiLy83RjXCPCLJ3nqabEkrPDlESTAphhyRD2U4cPXqUysrKJh95eL6qexMTE6moqODgwYPMndv0jX9jY2P56aefrB+XG8vZlLGRPzP+4GhBCmXGsmrtr/zoClYWrODqs1cT1aEXiqJYA9KWPA9YMpR1LcoBy3nenp6eNa4LIdoPvV4PWA4jEELYJwko2wlbBGBRUVFotVoSExNZtWoVMTExjBw5ssn9xMTE8Pbbb1NYXMjqU7/wXcp3VJoqUFBq3djdxd+ZjRkb+P3kOrp4dOGBuIdITEzEycmJHj16NPt5oP4MJUB+fn6rbdYuhGgbBoMBoFFbpQkhLg4JKNuJhIQE64rk5nJyciIiIoItW7awcuVK5s+f36yFQrGxsbiFuPH4ppmcMZ47i7u+U4LMf53QklGcwbN/Po2LyZVeMb1avJl4XRnKqoCysQtzhBD2qyqglAylEPZL5lC2EwkJCS0uD4MlGPz999/x9PTk5ptvblYfzp2cGP76EHINuU2+t2obkLJupXS7pwsGk6FZY6jSUIayMQFlbQGpEMJ+VJW8JUMphP2SgLKdSEpKIjo6usX9REZGcurUKe655x7c3NyafH92aTYvx/8XB2cHVKX5gZiiUTAGGnhnz1vNCuiqMrUGg6FFAaWqqmRmZhIQENDkMQgh2oaUvIWwfxJQthOlpaXWhSYtkZKSgqqqTdoqqIpZNfPunrcpN5WjaGxwqo4Cf5zazB+nNjf51qrV4WVlZbUGpI6Ojri5uTUYUJ44cYLi4mKbZH+FEK1DSt5C2D8JKP9GVFVl06ZNaDQaQkNDa1wvKSnhxRdf5Nprr6VDhw4oisLixYut19el/8bBs4mcTc7jwIeJbH7iT36etJqVE1bV+vnKc8tJ/jqFP57cwppbfmXNbb+x9YXtnNl/rlSuoPDBvnmU6Iur3XvvvfeiKArXXXddrX17enrSpUsXSktLa2QoVVUlqzSLrsO7cExNZVvmVo4XHsNoNtbop2qxU0u3LxJCtB4peQth/2RRzt/Ipk2byMrKqvNdfm5uLrNmzaJLly7ExcWxceNG6zVVVfkuZTkAOXtyOLHuJJ5dPXANdKU0s7TW/rJ3nubo96kEDwqk0+UhqCaVjI2n2P7STuIejqXLqM6oqJQby1l/Yj3X9xgPWE7WWbx4Mc7O9e81Fxsby/bt21FVFVVVOZR3iFXHf2ZX9k7KjeWE3R/KcVJ5defLADgoDkT5RjE6dCyDg4fgoHEgMTERLy8vOnfu3MSvphCirUjJWwj7JwHl38h7771HUFBQnWXg4OBgsrKyCAoKIj4+nksuucR6LTE3gazSLAC6XduVHhPC0DppSfjoYJ0BpV+ML1d+fAVOnucC2K7XdmHzY3+S/GUKXUZZgjgVlZ+PreQfYdcD8Mgjj3Dbbbexfv36ep8nNjaWdevWYXQ38uTmJziSn4xW0WL6a0X5hYyqkYO5B0nITcDHuQOP9J1JQkICMTExLT4WUwjReqTkLYT9k5L330R6ejo//fQTl112WZ1tnJycCAoKqvXanpw9aBXLFj9O3k5onRre7seji0e1YBJA66gloH8AFWcrMJafK0GfLssmpyyHzz//nMTERF5++eUG+4+NjSXo8kD0o8s5WpACUGcwWaVqlXlBRT7/2fYimaEZxPSWcrcQ9kxK3kLYPwko/ybmz5+Ph4dHtaxjU6TkH2kwWGusyoJKtE5atLrqQemBjP08/fTTPPfcc3UGtuc72ymX3vfFgNayYKgpqvbMdO/nhv7SCgzmlm1fJIRoPVLyFsL+SUDZjjR3v8Ty8nI+/vhj7r777maXjE4Un2jWfRcqzSola3s2wUOCULTnysxaRct7b8zFxcWFxx57rMF+Npz4nd/y1rZ4PIpG4axzLgv2z29xX0KI1iFHLwph/2QOZTsREBBARkZGs+794osvyM/P56GHHmLBggWNyv5dyGDWN+tzn89YaSL+zb1odVqibo2sdq3kVCkbPtvE1199jcFg4JVXXuH06dNkZmaiKAqLFi3ijjvuAOBseS7z97/P3vf2k7HhVI3P4xbixhXzapb2S7NKSf7yCGf2n8VYYcTF15ngocFE3RLJb+lrGdpxKP0DB2A2m/nwww/58MMPSU5OxtXVlbi4ON5++23i4uJa/HUQQjSNZCiFsH8SULYTsbGx1i1umkJVVd577z2uu+46unfv3uwTdxyUln2rqCaVPf/bS8nJEgb9awDOHaqv4D6wMIGIvhFMnDiRtLQ0Zs2ahVarxcvLi7y8vGptP074CP1fAa7GUUPcg9Wfx8Gt5lgLjxex9YXtOPs6E3Z9KI4eOsrPlFNxthywbF/03t53WXj1Iu656x6WLVvGbbfdxsMPP0xpaSl79+4lJyenRV8DIUTzSEAphP2TgLKdiI2N5bvvvsNsNqPRnJupoKoqueW5pBYepaiyCBUVD50H3b3CCHQN5I8//uDAgQPMmTMHsOy7ePvttzf583fy6ETh2cJmj3//Bwmcjs+h32N98OvtV+1a7oFccvacIWpKFK+++io9e/Zky5YtTJ48meDgYPLy8igvtwR+ueW5bMvcap0DqWgVOo0Mqfdzq2aVve/sw72TO0NnDap1QZGKSn5FHq989ApLlizh+++/Z8KECc1+XiGE7UjJWwj7JwFlOxEbG0tpaSnHjx8nLCyM7NIs1qStZl36bxTpi2q9x83RjYqESmJHxHLllVeSn5/PqVOnmpWhDPeJ4HDe4WYtzElafIiTv2cQfVcUISM61rhenlsBwKavN7Pp6+qn5pw6ZSlpr1ixgnvvvZd16b+ioFgDSrBkP42VRhxda89enNmXS/GJEga+MACtkxZjpQmtg6baHE4ADRo+nLuAgQMHMmHCBMxmM+Xl5c06olIIYTuSoRTC/klA2U5UBYF7EvawvvQ3fj62Eo2iqXd1c6mhFHO4mW6Pd+a9ve8QVWQ5C7w5p8L09ovjx6M/NPm+oz8cI/Wn4/SYGEb3cTVP5wHwjfVl5AuXMqPfTAx6A8ePHyc1NZWlS5diNpvR6/WsWbMGDw8PLnt9BLrOjvBXLGiqNLH65l8xVZpwdHckZHgwUbf1xMHl3Lf2mQOWk3m0jho2//NPClOL0DhoCBocSOz0aHQelqxHZVklmYeyGP/ABJ577jnmzp1LSUkJoaGhvPbaa0yePLnJzy+EaDmDwYBWq61WnRFC2BcJKNuJ4OBgOvXuxNf6LzAcs7xbb8xWORqt5RfwhpO/84dxM/69/IiMjKyz/bx58ygoKCAzMxOAlStXkpGRgVk149HLk2JtEWU55WRssmQOC1ItZfAjy48C4OLvQue/StBZ27M59Nlh3IJd8ejkTsbG6gto/Pv44eTthJu/Kzdfeis3RN5Q7fpvv/1G586d+fPPP3n22WfxD/BnQ8d11mDS2ceJHuO74xXmhWpWydl7hrQ1JyhMK2bo7EHWZ6/aeD1+zl4C+vrTY2IYRceLOfp9KuW5FQx7ZTCKolCWXQYqfPnVl+gcdbzxxht4eXnx7rvvMnXqVDw9Pbn22msb/JoLIWxLr9dLdlIIOycBZTtxrPAYfZ6PpdJBT3POdDGrZvSKnoEvXcKRwmSi/WrPUs6ZM4f09HTrx99//z3ff/89AB9uWMDPBSsoyykj+Ysj1e6r+tg3uoM1oCxKs5TiS7PK2Pvu/hqfa8h/B+Hk7YSjxpGrul5d7/gjIiK4YdoNbFi1zvpa1K09q7UJGdER945uHF52hKyt2dbyuqnCUqb37uFNv8f6ANBxSDBaJy2HlyaTe+As/nF+GP9ql5+Xz/bt2xk0aBAA//jHPwgNDWX27NkSUApxERgMBgkohbBzElC2A4WVhfx76wtoHDWoSvP2ogRAAxpF4T/bXmTeqPkEuAbUaJKWllbn7SaziYMbE9HEahj3w5gGP13k1Agip0Y02O6umLvxcfapdSznHwFpMhtrtLlQ93GhHP7yCLkHcq0BpUZnyVSGjAiu1jbk0o4cXppM/uF8/OP80P7VLrhLsDWYBHB3d2fcuHEsXboUo9GIg4P82AjRlgwGgyzIEcLOyV9GO7Rx40Yuv/zyWq8Nf20IPpGW4Ctn3xky/8yi4EgBxadKcPF14cqPar/PSoHUDccJ/Ecgbm5ulJSUNHpcWo2Wxwc8weMbH0VV1WoLY5pDo2iI9Yvl2tCGg1MAndap4TE6adG569AXnzv5pmqLIifv6vc7eVn+QOlLDdXa+fr71ug3ICAAg8FAaWkpXl5ejRqvEMI2pOQthP2TgNKOPfLII/hH+vP9kW+tr7kGn1txfGpzJplbsvDq7oWzj3NtXdRgLDeStCQJrbMWczNWbHf17MYLg1/kv9tfwmw2W8/GbioNGrp7hfHswBfQKI2baO/m6IaXzotCfd3bFxnLjeiL9ei8zmUzvMK84LeTVJytqNa2Iq8SwHreuHMHZ5x8nMg7XX3fS4DMzEycnZ3x8PBo1FiFELYjJW8h7J8smbNjI0aMwNTXQJfLO9NpZAidRoZYgx+AqFsiGb3saoa/OgTPbo0LdFKWH0Xr4kDwwCCM5uadzd03oC+zh72Kt7M3miZ+Cyl/zQAd0nEoLw9/FVdH1ybdH9EhEgUFk96EsbxmCfzIN0dBhYC+/tbXggYGonHUcPL3DFTzuazqiXUnAfCPO7cvZpcRncnMyOS3336zvpabm8tPP/3EFVdcIatMhbgIpOQthP2TDKUdK9IXsSNtOxqdxrpi+XwXnjbTkJLMUo6tTGPAM/3I3JKFGTPpRWl09ezW5LH18u3FB6M+ZPHBT1mbtgYFpd5spQYNZsx46jy5P+5BhoUMr7f/ulabH81PRd9Pj6HUwObH/6TjiI64h7gDcGbfGXJ2n8G/rz9BAwOtfTn7OBE+KYzkL1PYPmsXwYMCKUwr4sRvJwkZEYx3uDdgOU/8pgen8vGuhUycOJHHH38cLy8vFixYYD0OUgjR9qTkLYT9k4DSjj08/WHKS8tRNAodevnQ6/aeePfwbnZ/Bz9Nwje2A4H9A8jckmV57ezBZgWUAK6OrjzY52GmRE7l1/Rf2ZyxicySUzXmVjppnenZIZJruo1mcPAQHDQNf9vVt9p89MdX4+jmSMCAAHL355Kx4RSqWcUtyJWet0QQdn13FE31tfDhN/bA0d2R47+kk/hpEk7eToRP6kHE5B7WNibVxNQBN3H7n3fyz3/+k7fffhuDwcCQIUNYunSpnOMtxEUiJW8h7J8ElHZIp9MxceJEPOM8SKo4SOGJIlJ/OsaW57cz/NUheHVv+qKQ0/E5nNmXy2VvV88MphYcbfF4fV38uKnnNG7qOY1KYwUnik9QZixDgwZfF1+C3IIbPU+ySn2rzedsfJPN+Rvp92ifRvenKAqhY7oROqZbrdc1ioZw73B6dohC8VWswasQ4uKTkrcQ9k8mhNmhoUOH8u2339Lz2kgCLwkgfGIYI14fCsChpclN7s9sMHPw0yS6XtMFj87V51pml2bbZMxVnBycCfeJIM6/D7H+venoHtLkYLI+v/76K/+9YTb6PD0tXGRejYKGmf0eR1Gas8unEKI1SclbCPsnAaUdM5qN1vKxW7AbQQMDOZuQh2pqWiR1bOVx9MUGIqeG1/I5DLXcYZ/Wr1/P2LFjGT5kOK+PnoOj1tFmQeW9sffSyaOTbToTQtiUlLyFsH8SUNoxFweXatk9Fz8XzEYzxsqGN/iuYig1cGT5Ubpc1RljuZGynDLKcsosp8eoUHlGT05OTmsM36YKCwu58847GTFiBD/++CNxHeN4YfC/wUy1ldvNcUvUbYzpfp2NRiqEsDW9Xi8lbyHsnMyhtGNdPLpUy8CVnS5Do9Pg4Nz4/2yGUgOmChOpPxwj9YdjNa6/O2kuadef4Mcff7TBiFvPM888Q35+PosWLbJmKvoG9MNvZwDZkZk4dnBs0kbrGkWDk9aJ++Me5PLOV7TWsIUQNiAZSiHsnwSUdujMmTP4+/sT5t3DuhVP4fEisnedJqCvf40VzPXReTkx4Jl+NV4//nM6+Ufy+e8H/2VUzCibjb01nDlzhk8//ZRZs2bRtWvXateci104/WEuMxc9wsrUFVSaKlBQ6gwuNWhQURkSPJR7Yqfj61LzVBwhhH2RgFII+ycBpR2aMmUKLi4uDBw8kMycbPLS8zjx60m0Oi1Rt/a0titKKyJ7l6VcXZpdhrHMUt4G8OzmQdAlgTg4aQkeFFTjc2TvOE3BUYUZt87AQ2ffp78sWrQIRVG4++67a1xTFAWz3sxtvW7nxojJbDq5kfjTu0jOT6awssDazsXBhTCvHvT2781VXa/G18WvRl9CCPuk1+vllCoh7JwElHZo/PjxLFu2jPfeeY/CokIcPBwIGhxI5JRw3M47erHwWBHJXxypdm/Vx50uDyHokkDqoqCgVbR2H0yazWY+/PBDbrzxRvz8agaBGo0GVbVkI10cXLg2dDTXho4GIPtsFlGxvZjx0MP857lZsoJbiHZKMpRC2D9FrfprLOxSTlkOD6ybjsHGq7E1ioZ3L5/b7E3N28qvv/7KNddcw5YtWxg6dGiN6zNnzmTDhg0cOHCg1vunT5/O6tWrSUtLQ6vVtvZwhRCtYMSIEXTv3p0lS5Zc7KEIIeogq7ztXIBrAHfG1Cz1ttTUyJvsPpgEWLduHZ07d2bIkCG1XlcUBbO57iMfp06dSkZGBkeOHKmzjRDCvsk+lELYPwko24ExoWPpHzgAxQb/uTSKhp4dopgUMdkGI2t9iYmJxMXF1VmuPr/kXZvevXsDkJCQ0CrjE0K0Pil5C2H/JKBsBzSKhmcueZbe/r1RaP48QA0awrx68OKQ/zTqPG17kJCQQExMTJ3XG8pQ+vn5ERQUJAGlEO2YHL0ohP2TgLKdcHJw5t9DXmJC+A0oKE06zlDz13/ma7qN5uXhr+Lm6NbAHfYhPz+fjIwMYmNj62zTUIYSIDY2VgJKIdoxKXkLYf/aR5pKAOCoceSO6LsYHDyUTxM/4XDeIbSKFpNqqrV91bVQr1DuiLmbOP+4Nh5xyyQmJgI0GFDWl6Gsuv+nn36y6diEEG1HSt5C2D8JKNuhnh168salc0gvSmPDyd85nHeY1IJUKk0VAOg0OkK9utOzQ08u6zySHt41z/BuDxISEnBwcCAyMrLONg2VvMESUL799tuUlpbi5tY+srNCiHOk5C2E/ZOAsh3r6tmNO6LvAkBVVYyqEVRw0Dj8v9hzMSkpicjIyHr/kDSm5B0TE4Oqqhw6dIgBAwbYephCiFYmJW8h7J8ElP9PKIqCo/L/6xdueXk5np6e9bZpTIbSy8sLgLKyMpuNTQjRdqTkLYT9k0U5ol1rTIZSCNG+SclbCPsnAaVo1xqzKEcI0b5JyVsI+ycBpWjXGlPyFkK0X6qqSslbiHZAAkrRrknJW4j/30wmE6qqSslbCDsnAaVo1yRDKcT/bwaDAUAylELYOQkohV1rKPvYmAylBJxCtF8SUArRPkhAKexWcHAw6enp9bZpzKKcqj5CQkJsNjYhRNvQ6/UAUvIWws5JQCnsVmxsLFlZWZw9e7bONo0peScmJuLq6kpoaKithyiEaGWSoRSifZCAUtitmJgY4NyZ3lVUVSW3PJftWdvI7pBJ8FWBrDm+mqSzSVQYK2r0k5CQQHR0NBqNfLsL0d5UBZSSoRTCvslJOcJuRURE4OjoSEJCApdddhmFlYX8lr6WVcd/Ibc819LIH7pN6soH++cBoKAQ69eb67qP45KggWg1WhISEoiLi7uITyKEaK6qkrdkKIWwbxJQCrvl6OhIVFQUBxIO8NPRH/ksaTFGsxGV8xbhKKBoz51brqKSeDaBA7n7CXHvxKN9H+PgwYPcfPPNF+EJhBAtJSVvIdoHCSiFXYvuH83J2DQWJn7c6HvMqmVOZVZpJk/98U86jeloLZ8LIdoXKXkL0T7IpDJht4r0RZiuNaAN1jbrfrNqRkUl6taepHoftfHohBBtQUreQrQPElAKu2RWzby642UqdRVotC3/Nl116mfWp/9mg5EJIdqSlLyFaB+k5C3sxh133MGSJUvqvH7lJ1fg4usMQN7hfJKWHKbwWCGOrg50HBZMz5sjcXCp+S1dkFrIka9TGH1oLA4mB7p378706dN55JFHWu1ZhBC2ISVvIdoHCSiF3bjvvvu48sorKTeW8/GBD/9agAMJCxJxCXCxBpOFx4vY9uIO3Du5E31nFBVnK0j96TglmWUM/vcl1frM2XeGXS/vxrO7J5GTw+kZ1JMwUzgZGRkX4QmFEE0lJW8h2gcJKIXdGDJkCEOGDOHHoz8Q7B2EisrZpDxMlSY6XdrR2u7w0mQc3RwZ+t9BOLpa/si4BLhy4IMEcvadIaCPPwCGMgP73t1PQH9/BjzVD0WjYFD0PH3NU/g4d7gozyiEaBopeQvRPsgcSmF31hxfZd0a6NQfmaBAyF8BpaHMwJn9uXS6LMQaTAJ0HhmC1llL1pYs62unNmdSWaCn582RKBoFY4URs8nM7yd/b9sHEkI0m5S8hWgfJEMp7EqRvojM0kwAzEYzmVuy8In0wTXAFYDi9GJUk4p3D69q92kcNXiFelJ4rMj6Wu6Bszi4OlCRV8Gu13ZTmlmK1llL9ugcxn5xHc7Ozm33YEKIZpGStxDtg2QohV1JLUi1/vvM3jMYig10uuxcubsivxIAJx+nGvc6+ThZrwOUZpWimlR2vbqbgL5+DHiqH11GdWLXD/HceeedrfgUQghbkZK3EO2DBJTCrpwpy7H++9QfmSgOCh2HBltfM+ktm5ZrHGp+62p1WkyVJuvHxnKTZf7lyBBi7okmeEgQMfdE0/Xqznz11VekpKS04pMIIWyhKkMpJW8h7JsElMKuVJ1yYyw3kr0zh4A+/ug8z/0h0eos37Jmo7nGvSa9Ca2TtkbbkBEdq7ULuTQEgG3bttl28EIIm5MMpRDtgwSUwq64OLgAkL3jNKZKk3UxThXnv0rdleeVtqtU5ldarwM4dbDMkXTyrp7ZcPG2vJ6fn2+7gQshWoXBYECr1aIoysUeihCiHhJQCrvSxbMrABmbM9E6awkcGFjtukcXDxStQsHRwmqvmw1mCo8X4RnqaX3NO8zy74qzFdX7qLC87u/vb/PxCyFsS6/XS7lbiHZAAkphV7p4dMFUbCL3QC7Bg4NwcKp+jrejmyN+vf3I2HQKY7nR+nrGplOYKkzV5lt2HGb594n15zYx1ypaTq7LwMHBgZEjR7buwwghWsxgMEi5W4h2QLYNEnZFq9Gi3e+IalJrlLur9Lw5gi3PbmPLC9vpelVny0k5K47j38ePgH7nso5e3b3oPKoTJ9dnoJpUfKM7kJt4lqyt2Tz77LN07Fh7/0II+2EwGCRDKUQ7oKiqql7sQQhxvrgBcRw+epirF45C0dY+b+psUh6HPk+m8FghDs6Ws7yjbq15lrfZaCblu1ROrs+gIr8C9wB3XvrnSzz+2ONt8ShCiGZKT09n9erVvPPOO5w+fZrjx4/j7e19sYclhKiDBJTC7qiqylObnyCl4Chm1dTwDU3wUJ8ZXNPtWpv2KYSwnYSEBJ577jlWrVqF2XxuNwcXFxemTZvGq6++KvOfhbBDModS2B1FUXi0/xNosN2qTo2iJdavN1d3vcZmfQohbGv9+vUMGzaMo0ePsmDBAoqKinjqqafo1q0bzz//PCtWrGDo0KGkpqY23JkQok1JQCnsUoh7CPfHPWiTvjSKBg9Hdx7t95hsPSKEndqyZQujR49m2LBh7Nq1i3vvvRcPDw/0ej0uLi48//zz7NixA41Gw7Bhwzh9+vTFHrIQ4jwSUAq7dXW3a7gndjoASjOzlRpFg4fOk5eHv4a/a4AthyeEsJHS0lJuv/12LrnkElasWIG7u7v12vmrvENDQ9m8eTNms5n7778fmbElhP2QgFLYtX+EXc8Lg/6Nh84DjdL0b9dYv968PfJdunh2aYXRCSFs4cUXXyQzM5NFixbV2CLowlXegYGBzJ8/nx9//JFvv/22rYcqhKiDLMoR7UKxvphlh5ayLv1X9GY9GkVjPabxfFpFi0k1EegaxNTIqVzR5Uopcwthx4qLi+nYsSMzZ85k9uzZNa7feeedJCcns3Xr1mqvX3XVVRQXF7N9+/a2GqoQoh6yD6VoFzx0Htwf9wC39bqdzRmbOHg2keT8ZHLLz2Aym3HS6ujq2ZUIn0guCRpInH8fCSSFaAe++OILysrKuO+++2q9XtfG5g8++CA33HADe/fupW/fvq09TCFEAyRDKYQQ4qJQVZW+ffvStWtXfvrpp1rbTJkyhby8PH777bdqrxuNRrp27cp1113Hhx9+2BbDFULUQ+ZQCiGEuCh27tzJ/v37uf/+++tso9fra81QOjg4cO+997Js2TKKi4tbc5hCiEaQgFIIIcRFsXHjRjw8PLj66qvrbFPfWd433ngjpaWl7N69u7WGKIRoJAkohRBCXBQJCQnExMSg1WrrbKPX6+s8yzsiIgKdTkdCQkJrDVEI0UgSUAohhLgoEhISiI2NrbdNfRlKR0dHoqKiJKAUwg5IQCmEEKLNGQwGDh061KKAEiA2NlYCSiHsgASUQggh2tyRI0cwGAwNBpT1lbzBElAmJiZiNtfcl1YI0XYkoBRCCNHmEhMTAYiJiam3XWMylCUlJZw4ccKm4xNCNI0ElEIIIdpcUlISgYGB+Pr61tvuwqMXLxQdHW3tTwhx8UhAKYQQos1VVFTg4eHRYLu69qGsUtVHeXm5zcYmhGg6CSiFEELYrYZK3kII+yABpRBCCLvVUMlbCGEfJKAUQghhtxoqeQsh7IMElEIIIeyWlLyFaB8koBRCCGG3pOQtRPsgAaUQQoiLojGbkTdU8pYNzYWwDxJQCiGEaHOdOnUiIyMDg8FQZxtVVRsseR8/fhyAzp0723yMQojGk4BSCCFEm4uNjUWv15OSklJnG6PRCFBvyTshIQFFUawbnAshLg4JKIUQQrS5qjO8ExISqr1uVs2cKjnFppMb+Tr5KyJvCueo6xG2Z23jbHkuqqpWa5+QkED37t1xc3Nrq6ELIWrhcLEHIIQQ4u/H19eX4OBgEhISmDJlCnkVefyatoZVx3+hoLIAAI2ioccNYSRoD3Bgx34AOrl3ZlzYPxjZ+XJcHFxISEiwBqdCiItHAkohhBAXRWxsLAcSD/DT0R9YkrQYs9mMmXOLbMyqGY2DBpVzWclTJRnM3/8+yw59ziN9HyUhIYHp06dfjOELIc4jJW8hhBAXRVSfKEovLWZh4icYzcZqwWRdqoLLYn0xs3fMImCCP71ierX2UIUQDZAMpRBCiDZXoi/h7MAcXHBu1v1VgWXXqzuT5JWISTWhVbS2HKIQogkkQymEEKJNqarKnPg3KNYWo9G27M+QoijsK9rLN8lf22h0QojmkAylEEIIm9q1axdLlixhw4YNpKWl4evry+DBg5k9ezYRERGsP7GOPTm7ASg+WcLBRUnkHcpH46AhoL8/0XdG4eTlZO2vOKOEk+szOLPvDKXZZTi4OODV3ZPIqeF49/AG4OvkLxkUNIju3mGsW7eOl19+mYSEBIxGIxEREcyYMYNbb731Ynw5hPhbUNQL92AQQgghWmDSpEls2bKFG2+8kd69e5Odnc28efMoKSlh85+beCPjNcqMZZTnlrP5iS04uDoQOrYbpgojqT8dx8XPmRFvDEPjaMleHlx8iJPrThI8JAjvcG8MpUbSfz1BeU45g/59Cf5xfmgUDZE+kQwruJTx48czZMgQbrrpJhRF4ZtvvmHz5s289dZbPPbYYxf5qyPE/08SUAohhLCprVu3MmDAgGobkqekpBAbG8uwa4fidocLAAc+TOTk7xlcPu8yXP0tr53Zn8v2l3bS+4EYul7dBYCC1ELcO7rh4HKuqKYv0rNhxmbcOrox/NUh1tfz3yviWPIxjh07hpOTJctpNBrp2bMnbm5u7N+/v9WfX4i/I5lDKYQQwqaGDh1a43Sb8PBwoqOjSUw6iIICQNa2bAIHBFiDSQD/OD/cOrqRuSXL+pp3mFe1YBJA56mjQy8fSjJKrK9pFS2nck/h4+NjDSYBHBwc8PPzw8XFBSFE65A5lEIIIVqdqqqcPn0a1V9FRaX8bAX6Qj3ePbxqtPUO9yJn95kG+6ws0KPzPBe4mlQTvtE+xH+1h3/961/cfvvtKIrCF198QXx8PN98841Nn0kIcY4ElEIIIVrdsmXLOHXqFHHjLafaVOZXAODkU3PbIGcfZwwlBkwGE1rH2rcCOpuUR35yPuGTelR7PWh8EDeabuTll19m9uzZALi6uvLdd99x/fXX2/KRhBDnkZK3EEKIVnX48GEeeughYvrH0PnyTgCY9JZNzLWONf8MVS3GMetr3+i8sqCSPW/twzXAlR4Tule7pjqY6RLWhUmTJvHll1+ydOlSBgwYwC233ML27dtt+VhCiPNIhlIIIUSryc7OZuzYsXh5eTH7o9ksPPERAFqdJWg0GWoGjea/XtPoagabxgojO1+Ox1huZNgrQ2rMrUz86CBpWSc5sO8AGo3l/smTJxMdHc3MmTPZsWOHTZ9PCGEhGUohhBCtorCwkNGjR1NQUMCaNWvo2qmL9VpVqbuq9H2+ivwKHN0da5S7zQYz8a/voSi9mEue7Y9nV48a10+sz2DMmDHWYBLA0dGR0aNHEx8fj16vt+UjCiH+IhlKIYQQNldRUcG4ceM4cuQI69ato1evXhRUFlivu/g6o/PUUXC0sMa9BSmFeIV6VntNNavsfW8/uQfO0v+fffGL8a1xn75Yj2pS0dSSKzEYDJjNZkwmU8sfTghRg2QohRBC2JTJZGLKlCls27aN5cuXM2SIZZ9IbydvfJx8rO2ChwRxOj6H8txy62tnDuRSmllK8NCgan0mfnyQzD+ziL0vmuAh1a9VcfFywdnDmR9++KFaJrKkpISVK1fSs2dP2TpIiFYiGUohhBA29cQTT7BixQrGjRtHXl4eS5cutV5zPKlD00uDWTUTPimMzK1ZbP3XDrpf1w1jhZHUH4/j0dWDzqM6We85tvI4aWtO4BPpjVanJWPjqWqfL2hwIA7ODqhalVvuv5lP3lzI4MGDue222zCZTCxcuJCMjIxq4xBC2JaclCOEEMKmRo4cyaZNm+q8Pu6HMdZ/F58o5uCiQ3+d5a0Q0D/Acpa397mNyfe+t5+MDadq6wqAUR+OxDXAFXdHD5aM/pzlXy3n3Xff5ciRI1RWVtK7d2+efPJJJk6caJsHFELUIAGlEEKINvXGrtfYmrkFs1r7tkDNdXfMPVzfY4JN+xRCNI7MoRRCCNGm7uv9AK4OrtYjGFtKo2iI8InkurB/2KQ/IUTTSUAphBCiTXk5efHPAU+hKEqLg0qNosHN0Z0nBjyJVqn9VB0hROuTgFIIIUSb6xfYn6cveRaNokGjNO9PkUbR4O7ozivDXyXYLdjGIxRCNIXMoRRCCHHRpOSn8L/db5JVkolK4/4cKSioqPTx78sj/R7Fz8WvlUcphGiIBJRCCCEuKoPJwA9Hv2dl6k8U6gvRKlpMas0NyKte7+TeiRsjpjCy8+Uoim3mYQohWkYCSiGEEHbBaDayK3snibkJHMlPJrM0C5PZiE6ro6tnN8J9Iugf0J9evtESSAphZySgFEIIIYQQLSKLcoQQQgghRItIQCmEEEIIIVpEAkohhBBCCNEiElAKIYQQQogWkYBSCCGEEEK0iASUQgghhBCiRSSgFEIIIYQQLSIBpRBCCCGEaBEJKIUQQgghRItIQCmEEEIIIVpEAkohhBBCCNEiElAKIYQQQogWkYBSCCGEEEK0iASUQgghhBCiRSSgFEIIIYQQLSIBpRBCCCGEaBEJKIUQQgghRItIQCmEEEIIIVpEAkohhBBCCNEiElAKIYQQQogWkYBSCCGEEEK0iASUQgghhBCiRSSgFEIIIYQQLSIBpRBCCCGEaBEJKIUQQgghRItIQCmEEEIIIVpEAkohhBBCCNEiElAKIYQQQogWkYBSCCGEEEK0iASUQgghhBCiRSSgFEIIIYQQLSIBpRBCCCGEaBEJKIUQQgghRItIQCmEEEIIIVpEAkohhBBCCNEiElAKIYQQQogWkYBSCCGEEEK0yP8BLqCGA6wmZUkAAAAASUVORK5CYII=", "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -1294,7 +724,7 @@ "plt.figure()\n", "cora_graph = nx.Graph()\n", "cora_graph.add_nodes_from([(i, {\"label\": labels[i]}) for i in inds])\n", - "cora_graph.add_edges_from(edge_index)\n", + "cora_graph.add_edges_from(dataset.get(\"edge_indices\")[0])\n", "hops = 2\n", "khopgraph = nx.generators.ego.ego_graph(cora_graph, node_index, radius=hops)\n", "for n in khopgraph.nodes:\n", @@ -1305,6 +735,14 @@ "explainer.present_explanation(khopgraph)\n", "plt.show()" ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "dd7cdc5b-5bac-44f3-a627-3d22aa17e60c", + "metadata": {}, + "outputs": [], + "source": [] } ], "metadata": { diff --git a/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_1.ipynb b/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_1.ipynb index 79b59f5d..a65188ec 100644 --- a/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_1.ipynb +++ b/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_1.ipynb @@ -901,7 +901,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.9.7" + "version": "3.10.13" } }, "nbformat": 4, diff --git a/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_2.ipynb b/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_2.ipynb index 994f4ac0..eb4981db 100644 --- a/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_2.ipynb +++ b/notebooks/graph_explanation/explain_GNNExplain_mutagenicity_2.ipynb @@ -960,7 +960,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.9.7" + "version": "3.10.13" } }, "nbformat": 4, diff --git a/setup.py b/setup.py index 6f558590..8fc63000 100644 --- a/setup.py +++ b/setup.py @@ -14,7 +14,7 @@ long_description_content_type="text/markdown", url="https://github.com/aimat-lab/gcnn_keras", install_requires=[ - "dm-tree", + # "dm-tree", "keras-core", "tensorflow>=2.13", "torch>=2.0.0",