From 0bcaf62e59606685433b1efa96dd84d96b0801ef Mon Sep 17 00:00:00 2001 From: Dario Coscia <93731561+dario-coscia@users.noreply.github.com> Date: Mon, 27 Feb 2023 10:59:18 +0100 Subject: [PATCH] Continuous Convolution (#69) * network handling update * adding tutorial * docs --- docs/source/_rst/code.rst | 1 + docs/source/_rst/convolution.rst | 12 + docs/source/_rst/tutorial4/tutorial.rst | 824 +++++++++++++ .../tutorial_files/tutorial_32_0.png | Bin 0 -> 62030 bytes .../tutorial_files/tutorial_40_0.png | Bin 0 -> 85799 bytes .../tutorial_files/tutorial_45_0.png | Bin 0 -> 84791 bytes .../tutorial_files/tutorial_49_0.png | Bin 0 -> 73330 bytes docs/source/index.rst | 1 + pina/model/layers/__init__.py | 7 + pina/model/layers/convolution.py | 154 +++ pina/model/layers/convolution_2d.py | 548 +++++++++ pina/model/layers/integral.py | 63 + pina/model/layers/stride.py | 82 ++ pina/model/layers/utils_convolution.py | 48 + tests/test_conv.py | 140 +++ tutorials/README.md | 1 + tutorials/tutorial4/tutorial.ipynb | 1031 +++++++++++++++++ tutorials/tutorial4/tutorial.py | 638 ++++++++++ 18 files changed, 3550 insertions(+) create mode 100644 docs/source/_rst/convolution.rst create mode 100644 docs/source/_rst/tutorial4/tutorial.rst create mode 100644 docs/source/_rst/tutorial4/tutorial_files/tutorial_32_0.png create mode 100644 docs/source/_rst/tutorial4/tutorial_files/tutorial_40_0.png create mode 100644 docs/source/_rst/tutorial4/tutorial_files/tutorial_45_0.png create mode 100644 docs/source/_rst/tutorial4/tutorial_files/tutorial_49_0.png create mode 100644 pina/model/layers/__init__.py create mode 100644 pina/model/layers/convolution.py create mode 100644 pina/model/layers/convolution_2d.py create mode 100644 pina/model/layers/integral.py create mode 100644 pina/model/layers/stride.py create mode 100644 pina/model/layers/utils_convolution.py create mode 100644 tests/test_conv.py create mode 100644 tutorials/tutorial4/tutorial.ipynb create mode 100644 tutorials/tutorial4/tutorial.py diff --git a/docs/source/_rst/code.rst b/docs/source/_rst/code.rst index c32039c..2e0b5da 100644 --- a/docs/source/_rst/code.rst +++ b/docs/source/_rst/code.rst @@ -8,3 +8,4 @@ Code Documentation FeedForward DeepONet PINN + ContinuousConv diff --git a/docs/source/_rst/convolution.rst b/docs/source/_rst/convolution.rst new file mode 100644 index 0000000..25f182d --- /dev/null +++ b/docs/source/_rst/convolution.rst @@ -0,0 +1,12 @@ +ContinuousConv +============== +.. currentmodule:: pina.model.layers.convolution_2d + +.. automodule:: pina.model.layers.convolution_2d + +.. autoclass:: ContinuousConv + :members: + :private-members: + :undoc-members: + :show-inheritance: + :noindex: diff --git a/docs/source/_rst/tutorial4/tutorial.rst b/docs/source/_rst/tutorial4/tutorial.rst new file mode 100644 index 0000000..bc9beeb --- /dev/null +++ b/docs/source/_rst/tutorial4/tutorial.rst @@ -0,0 +1,824 @@ +Tutorial 4: continuous convolutional filter +=========================================== + +In this tutorial we will show how to use the Continouous Convolutional +Filter, and how to build common Deep Learning architectures with it. The +implementation of the filter follows the original work `A Continuous +Convolutional Trainable Filter for Modelling Unstructured +Data `__ of Coscia Dario, Laura +Meneghetti, Nicola Demo, Giovanni Stabile, and Gianluigi Rozza. + +First of all we import the modules needed for the tutorial, which +include: + +- ``ContinuousConv`` class from ``pina.model.layers`` which implements + the continuous convolutional filter +- ``PyTorch`` and ``Matplotlib`` for tensorial operations and + visualization respectively + +.. code:: ipython3 + + import torch + import matplotlib.pyplot as plt + from pina.model.layers import ContinuousConv + import torchvision # for MNIST dataset + from pina.model import FeedForward # for building AE and MNIST classification + +The tutorial is structured as follow: + +* `Continuous filter background <#continuous-filter-background>`__: understand how the convolutional filter works and how to use it. + +* `Building a MNIST Classifier <#building-a-mnist-classifier>`__: show how to build a simple classifier using the MNIST dataset and how to combine a continuous convolutional layer with a feedforward neural network. + +* `Building a Continuous Convolutional Autoencoder <#building-a-continuous-convolutional-autoencoder>`__: show how to use the continuous filter to work with unstructured data for autoencoding and up-sampling. + +Continuous filter background +---------------------------- + +As reported by the authors in the original paper: in contrast to +discrete convolution, continuous convolution is mathematically defined +as: + +.. math:: + + + \mathcal{I}_{\rm{out}}(\mathbf{x}) = \int_{\mathcal{X}} \mathcal{I}(\mathbf{x} + \mathbf{\tau}) \cdot \mathcal{K}(\mathbf{\tau}) d\mathbf{\tau}, + +where :math:`\mathcal{K} : \mathcal{X} \rightarrow \mathbb{R}` is the +*continuous filter* function, and +:math:`\mathcal{I} : \Omega \subset \mathbb{R}^N \rightarrow \mathbb{R}` +is the input function. The continuous filter function is approximated +using a FeedForward Neural Network, thus trainable during the training +phase. The way in which the integral is approximated can be different, +currently on **PINA** we approximate it using a simple sum, as suggested +by the authors. Thus, given :math:`\{\mathbf{x}_i\}_{i=1}^{n}` points in +:math:`\mathbb{R}^N` of the input function mapped on the +:math:`\mathcal{X}` filter domain, we approximate the above equation as: + +.. math:: + + + \mathcal{I}_{\rm{out}}(\mathbf{\tilde{x}}_i) = \sum_{{\mathbf{x}_i}\in\mathcal{X}} \mathcal{I}(\mathbf{x}_i + \mathbf{\tau}) \cdot \mathcal{K}(\mathbf{x}_i), + +where :math:`\mathbf{\tau} \in \mathcal{S}`, with :math:`\mathcal{S}` +the set of available strides, corresponds to the current stride position +of the filter, and :math:`\mathbf{\tilde{x}}_i` points are obtained by +taking the centroid of the filter position mapped on the :math:`\Omega` +domain. + +We will now try to pratically see how to work with the filter. From the +above definition we see that what is needed is: 1. A domain and a +function defined on that domain (the input) 2. A stride, corresponding +to the positions where the filter needs to be :math:`\rightarrow` +``stride`` variable in ``ContinuousConv`` 3. The filter rectangular +domain :math:`\rightarrow` ``filter_dim`` variable in ``ContinuousConv`` + +Input function +~~~~~~~~~~~~~~ + +The input function for the continuous filter is defined as a tensor of +shape: + +.. math:: [B \times N_{in} \times N \times D] + +\ where :math:`B` is the batch_size, :math:`N_{in}` is the number of +input fields, :math:`N` the number of points in the mesh, :math:`D` the +dimension of the problem. In particular: + +* :math:`D` is the number of spatial variables + 1. The last column must contain the field value. For example for 2D problems :math:`D=3` and the tensor will be something like ``[first coordinate, second coordinate, field value]`` + +* :math:`N_{in}` represents the number of vectorial function presented. For example a vectorial function :math:`f = [f_1, f_2]` will have math:`N_{in}=2` + +Let’s see an example to clear the ideas. We will be verbose to explain +in details the input form. We wish to create the function: + +.. math:: + + + f(x, y) = [\sin(\pi x) \sin(\pi y), -\sin(\pi x) \sin(\pi y)] \quad (x,y)\in[0,1]\times[0,1] + +using a batch size of one. + +.. code:: ipython3 + + # batch size fixed to 1 + batch_size = 1 + + # points in the mesh fixed to 200 + N = 200 + + # vectorial 2 dimensional function, number_input_fileds=2 + number_input_fileds = 2 + + # 2 dimensional spatial variables, D = 2 + 1 = 3 + D = 3 + + # create the function f domain as random 2d points in [0, 1] + domain = torch.rand(size=(batch_size, number_input_fileds, N, D-1)) + print(f"Domain has shape: {domain.shape}") + + # create the functions + pi = torch.acos(torch.tensor([-1.])) # pi value + f1 = torch.sin(pi * domain[:, 0, :, 0]) * torch.sin(pi * domain[:, 0, :, 1]) + f2 = - torch.sin(pi * domain[:, 1, :, 0]) * torch.sin(pi * domain[:, 1, :, 1]) + + # stacking the input domain and field values + data = torch.empty(size=(batch_size, number_input_fileds, N, D)) + data[..., :-1] = domain # copy the domain + data[:, 0, :, -1] = f1 # copy first field value + data[:, 1, :, -1] = f1 # copy second field value + print(f"Filter input data has shape: {data.shape}") + + +.. parsed-literal:: + + Domain has shape: torch.Size([1, 2, 200, 2]) + Filter input data has shape: torch.Size([1, 2, 200, 3]) + + +Stride +~~~~~~ + +The stride is passed as a dictionary ``stride`` which tells the filter +where to go. Here is an example for the :math:`[0,1]\times[0,5]` domain: + +.. code:: python + + # stride definition + stride = {"domain": [1, 5], + "start": [0, 0], + "jump": [0.1, 0.3], + "direction": [1, 1], + } + +This tells the filter: + +1. ``domain``: square domain (the only implemented) :math:`[0,1]\times[0,5]`. The minimum value is always zero, while the maximum is specified by the user + +2. ``start``: start position of the filter, coordinate :math:`(0, 0)` + +3. ``jump``: the jumps of the centroid of the filter to the next position :math:`(0.1, 0.3)` + +4. ``direction``: the directions of the jump, with ``1 = right``, ``0 = no jump``,\ ``-1 = left`` with respect to the current position + +**Note** + +We are planning to release the possibility to directly pass a list of +possible strides! + +Filter definition +~~~~~~~~~~~~~~~~~ + +Having defined all the previous blocks we are able to construct the +continuous filter. + +Suppose we would like to get an ouput with only one field, and let us +fix the filter dimension to be :math:`[0.1, 0.1]`. + +.. code:: ipython3 + + # filter dim + filter_dim = [0.1, 0.1] + + # stride + stride = {"domain": [1, 1], + "start": [0, 0], + "jump": [0.08, 0.08], + "direction": [1, 1], + } + + # creating the filter + cConv = ContinuousConv(input_numb_field=number_input_fileds, + output_numb_field=1, + filter_dim=filter_dim, + stride=stride) + +That’s it! In just one line of code we have created the continuous +convolutional filter. By default the ``pina.model.FeedForward`` neural +network is intitialised, more on the +`documentation `__. In +case the mesh doesn’t change during training we can set the ``optimize`` +flag equals to ``True``, to exploit optimizations for finding the points +to convolve. + +.. code:: ipython3 + + # creating the filter + optimization + cConv = ContinuousConv(input_numb_field=number_input_fileds, + output_numb_field=1, + filter_dim=filter_dim, + stride=stride, + optimize=True) + + +Let’s try to do a forward pass + +.. code:: ipython3 + + print(f"Filter input data has shape: {data.shape}") + + #input to the filter + output = cConv(data) + + print(f"Filter output data has shape: {output.shape}") + + +.. parsed-literal:: + + Filter input data has shape: torch.Size([1, 2, 200, 3]) + Filter output data has shape: torch.Size([1, 1, 169, 3]) + + +If we don’t want to use the default ``FeedForward`` neural network, we +can pass a specified torch model in the ``model`` keyword as follow: + +.. code:: ipython3 + + class SimpleKernel(torch.nn.Module): + def __init__(self) -> None: + super().__init__() + self. model = torch.nn.Sequential( + torch.nn.Linear(2, 20), + torch.nn.ReLU(), + torch.nn.Linear(20, 20), + torch.nn.ReLU(), + torch.nn.Linear(20, 1)) + + def forward(self, x): + return self.model(x) + + + cConv = ContinuousConv(input_numb_field=number_input_fileds, + output_numb_field=1, + filter_dim=filter_dim, + stride=stride, + optimize=True, + model=SimpleKernel) + + +Notice that we pass the class and not an already built object! + +Building a MNIST Classifier +--------------------------- + +Let’s see how we can build a MNIST classifier using a continuous +convolutional filter. We will use the MNIST dataset from PyTorch. In +order to keep small training times we use only 6000 samples for training +and 1000 samples for testing. + +.. code:: ipython3 + + from torch.utils.data import DataLoader, SubsetRandomSampler + + numb_training = 6000 # get just 6000 images for training + numb_testing= 1000 # get just 1000 images for training + seed = 111 # for reproducibility + batch_size = 8 # setting batch size + + # setting the seed + torch.manual_seed(seed) + + # downloading the dataset + train_data = torchvision.datasets.MNIST('./data/', train=True, download=True, + transform=torchvision.transforms.Compose([ + torchvision.transforms.ToTensor(), + torchvision.transforms.Normalize( + (0.1307,), (0.3081,)) + ])) + subsample_train_indices = torch.randperm(len(train_data))[:numb_training] + train_loader = DataLoader(train_data, batch_size=batch_size, + sampler=SubsetRandomSampler(subsample_train_indices)) + + test_data = torchvision.datasets.MNIST('./data/', train=False, download=True, + transform=torchvision.transforms.Compose([ + torchvision.transforms.ToTensor(), + torchvision.transforms.Normalize( + (0.1307,), (0.3081,)) + ])) + subsample_test_indices = torch.randperm(len(train_data))[:numb_testing] + test_loader = DataLoader(train_data, batch_size=batch_size, + sampler=SubsetRandomSampler(subsample_train_indices)) + +Let’s now build a simple classifier. The MNIST dataset is composed by +vectors of shape ``[batch, 1, 28, 28]``, but we can image them as one +field functions where the pixels :math:`ij` are the coordinate +:math:`x=i, y=j` in a :math:`[0, 27]\times[0,27]` domain, and the pixels +value are the field values. We just need a function to transform the +regular tensor in a tensor compatible for the continuous filter: + +.. code:: ipython3 + + def transform_input(x): + batch_size = x.shape[0] + dim_grid = tuple(x.shape[:-3:-1]) + + # creating the n dimensional mesh grid for a single channel image + values_mesh = [torch.arange(0, dim).float() for dim in dim_grid] + mesh = torch.meshgrid(values_mesh) + coordinates_mesh = [x.reshape(-1, 1) for x in mesh] + coordinates = torch.cat(coordinates_mesh, dim=1).unsqueeze( + 0).repeat((batch_size, 1, 1)).unsqueeze(1) + + return torch.cat((coordinates, x.flatten(2).unsqueeze(-1)), dim=-1) + + + # let's try it out + image, s = next(iter(train_loader)) + print(f"Original MNIST image shape: {image.shape}") + + image_transformed = transform_input(image) + print(f"Transformed MNIST image shape: {image_transformed.shape}") + + + +.. parsed-literal:: + + Original MNIST image shape: torch.Size([8, 1, 28, 28]) + Transformed MNIST image shape: torch.Size([8, 1, 784, 3]) + + +We can now build a simple classifier! We will use just one convolutional +filter followed by a feedforward neural network + +.. code:: ipython3 + + # setting the seed + torch.manual_seed(seed) + + class ContinuousClassifier(torch.nn.Module): + def __init__(self): + super().__init__() + + # number of classes for classification + numb_class = 10 + + # convolutional block + self.convolution = ContinuousConv(input_numb_field=1, + output_numb_field=4, + stride={"domain": [27, 27], + "start": [0, 0], + "jumps": [4, 4], + "direction": [1, 1.], + }, + filter_dim=[4, 4], + optimize=True) + # feedforward net + self.nn = FeedForward(input_variables=196, + output_variables=numb_class, + layers=[120, 64], + func=torch.nn.ReLU) + + def forward(self, x): + # transform input + convolution + x = transform_input(x) + x = self.convolution(x) + # feed forward classification + return self.nn(x[..., -1].flatten(1)) + + + net = ContinuousClassifier() + +Let’s try to train it using a simple pytorch training loop. We train for +juts 1 epoch using Adam optimizer with a :math:`0.001` learning rate. + +.. code:: ipython3 + + # setting the seed + torch.manual_seed(seed) + + # optimizer and loss function + optimizer = torch.optim.Adam(net.parameters(), lr=0.001) + criterion = torch.nn.CrossEntropyLoss() + + for epoch in range(1): # loop over the dataset multiple times + + running_loss = 0.0 + for i, data in enumerate(train_loader, 0): + # get the inputs; data is a list of [inputs, labels] + inputs, labels = data + + # zero the parameter gradients + optimizer.zero_grad() + + # forward + backward + optimize + outputs = net(inputs) + loss = criterion(outputs, labels) + loss.backward() + optimizer.step() + + # print statistics + running_loss += loss.item() + if i % 50 == 49: + print( + f'epoch [{i + 1}/{numb_training//batch_size}] loss[{running_loss / 500:.3f}]') + running_loss = 0.0 + + + +.. parsed-literal:: + + epoch [50/750] loss[0.148] + epoch [100/750] loss[0.072] + epoch [150/750] loss[0.063] + epoch [200/750] loss[0.053] + epoch [250/750] loss[0.041] + epoch [300/750] loss[0.048] + epoch [350/750] loss[0.054] + epoch [400/750] loss[0.048] + epoch [450/750] loss[0.047] + epoch [500/750] loss[0.035] + epoch [550/750] loss[0.036] + epoch [600/750] loss[0.041] + epoch [650/750] loss[0.030] + epoch [700/750] loss[0.040] + epoch [750/750] loss[0.040] + + +Let’s see the performance on the train set! + +.. code:: ipython3 + + correct = 0 + total = 0 + with torch.no_grad(): + for data in test_loader: + images, labels = data + # calculate outputs by running images through the network + outputs = net(images) + # the class with the highest energy is what we choose as prediction + _, predicted = torch.max(outputs.data, 1) + total += labels.size(0) + correct += (predicted == labels).sum().item() + + print( + f'Accuracy of the network on the 1000 test images: {(correct / total):.3%}') + + + +.. parsed-literal:: + + Accuracy of the network on the 1000 test images: 93.017% + + +As we can see we have very good performance for having traing only for 1 +epoch! Nevertheless, we are still using structured data… Let’s see how +we can build an autoencoder for unstructured data now. + +Building a Continuous Convolutional Autoencoder +----------------------------------------------- + +Just as toy problem, we will now build an autoencoder for the following +function :math:`f(x,y)=\sin(\pi x)\sin(\pi y)` on the unit circle domain +centered in :math:`(0.5, 0.5)`. We will also see the ability to +up-sample (once trained) the results without retraining. Let’s first +create the input and visualize it, we will use firstly a mesh of +:math:`100` points. + +.. code:: ipython3 + + # create inputs + def circle_grid(N=100): + """Generate points withing a unit 2D circle centered in (0.5, 0.5) + + :param N: number of points + :type N: float + :return: [x, y] array of points + :rtype: torch.tensor + """ + + PI = torch.acos(torch.zeros(1)).item() * 2 + R = 0.5 + centerX = 0.5 + centerY = 0.5 + + r = R * torch.sqrt(torch.rand(N)) + theta = torch.rand(N) * 2 * PI + + x = centerX + r * torch.cos(theta) + y = centerY + r * torch.sin(theta) + + return torch.stack([x, y]).T + + # create the grid + grid = circle_grid(500) + + # create input + input_data = torch.empty(size=(1, 1, grid.shape[0], 3)) + input_data[0, 0, :, :-1] = grid + input_data[0, 0, :, -1] = torch.sin(pi * grid[:, 0]) * torch.sin(pi * grid[:, 1]) + + # visualize data + plt.title("Training sample with 500 points") + plt.scatter(grid[:, 0], grid[:, 1], c=input_data[0, 0, :, -1]) + plt.colorbar() + plt.show() + + + + +.. image:: tutorial_files/tutorial_32_0.png + + +Let’s now build a simple autoencoder using the continuous convolutional +filter. The data is clearly unstructured and a simple convolutional +filter might not work without projecting or interpolating first. Let’s +first build and ``Encoder`` and ``Decoder`` class, and then a +``Autoencoder`` class that contains both. + +.. code:: ipython3 + + class Encoder(torch.nn.Module): + def __init__(self, hidden_dimension): + super().__init__() + + # convolutional block + self.convolution = ContinuousConv(input_numb_field=1, + output_numb_field=2, + stride={"domain": [1, 1], + "start": [0, 0], + "jumps": [0.05, 0.05], + "direction": [1, 1.], + }, + filter_dim=[0.15, 0.15], + optimize=True) + # feedforward net + self.nn = FeedForward(input_variables=400, + output_variables=hidden_dimension, + layers=[240, 120]) + + def forward(self, x): + # convolution + x = self.convolution(x) + # feed forward pass + return self.nn(x[..., -1]) + + + class Decoder(torch.nn.Module): + def __init__(self, hidden_dimension): + super().__init__() + + # convolutional block + self.convolution = ContinuousConv(input_numb_field=2, + output_numb_field=1, + stride={"domain": [1, 1], + "start": [0, 0], + "jumps": [0.05, 0.05], + "direction": [1, 1.], + }, + filter_dim=[0.15, 0.15], + optimize=True) + # feedforward net + self.nn = FeedForward(input_variables=hidden_dimension, + output_variables=400, + layers=[120, 240]) + + def forward(self, weights, grid): + # feed forward pass + x = self.nn(weights) + # transpose convolution + return torch.sigmoid(self.convolution.transpose(x, grid)) + + +Very good! Notice that in the ``Decoder`` class in the ``forward`` pass +we have used the ``.transpose()`` method of the +``ContinuousConvolution`` class. This method accepts the ``weights`` for +upsampling and the ``grid`` on where to upsample. Let’s now build the +autoencoder! We set the hidden dimension in the ``hidden_dimension`` +variable. We apply the sigmoid on the output since the field value is +between :math:`[0, 1]`. + +.. code:: ipython3 + + class Autoencoder(torch.nn.Module): + def __init__(self, hidden_dimension=10): + super().__init__() + + self.encoder = Encoder(hidden_dimension) + self.decoder = Decoder(hidden_dimension) + + def forward(self, x): + # saving grid for later upsampling + grid = x.clone().detach() + # encoder + weights = self.encoder(x) + # decoder + out = self.decoder(weights, grid) + return out + + + net = Autoencoder() + +Let’s now train the autoencoder, minimizing the mean square error loss +and optimizing using Adam. + +.. code:: ipython3 + + # setting the seed + torch.manual_seed(seed) + + # optimizer and loss function + optimizer = torch.optim.Adam(net.parameters(), lr=0.001) + criterion = torch.nn.MSELoss() + max_epochs = 150 + + for epoch in range(max_epochs): # loop over the dataset multiple times + + # zero the parameter gradients + optimizer.zero_grad() + + # forward + backward + optimize + outputs = net(input_data) + loss = criterion(outputs[..., -1], input_data[..., -1]) + loss.backward() + optimizer.step() + + # print statistics + if epoch % 10 ==9: + print(f'epoch [{epoch + 1}/{max_epochs}] loss [{loss.item():.2}]') + + + +.. parsed-literal:: + + epoch [10/150] loss [0.013] + epoch [20/150] loss [0.0029] + epoch [30/150] loss [0.0019] + epoch [40/150] loss [0.0014] + epoch [50/150] loss [0.0011] + epoch [60/150] loss [0.00094] + epoch [70/150] loss [0.00082] + epoch [80/150] loss [0.00074] + epoch [90/150] loss [0.00068] + epoch [100/150] loss [0.00064] + epoch [110/150] loss [0.00061] + epoch [120/150] loss [0.00058] + epoch [130/150] loss [0.00057] + epoch [140/150] loss [0.00056] + epoch [150/150] loss [0.00054] + + +Let’s visualize the two solutions side by side! + +.. code:: ipython3 + + net.eval() + + # get output and detach from computational graph for plotting + output = net(input_data).detach() + + # visualize data + fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3)) + pic1 = axes[0].scatter(grid[:, 0], grid[:, 1], c=input_data[0, 0, :, -1]) + axes[0].set_title("Real") + fig.colorbar(pic1) + plt.subplot(1, 2, 2) + pic2 = axes[1].scatter(grid[:, 0], grid[:, 1], c=output[0, 0, :, -1]) + axes[1].set_title("Autoencoder") + fig.colorbar(pic2) + plt.tight_layout() + plt.show() + + + + +.. image:: tutorial_files/tutorial_40_0.png + + +As we can see the two are really similar! We can compute the :math:`l_2` +error quite easily as well: + +.. code:: ipython3 + + def l2_error(input_, target): + return torch.linalg.norm(input_-target, ord=2)/torch.linalg.norm(input_, ord=2) + + + print(f'l2 error: {l2_error(input_data[0, 0, :, -1], output[0, 0, :, -1]):.2%}') + + +.. parsed-literal:: + + l2 error: 4.10% + + +More or less :math:`4\%` in :math:`l_2` error, which is really low +considering the fact that we use just **one** convolutional layer and a +simple feedforward to decrease the dimension. Let’s see now some +peculiarity of the filter. + +Filter for upsampling +~~~~~~~~~~~~~~~~~~~~~ + +Suppose we have already the hidden dimension and we want to upsample on +a differen grid with more points. Let’s see how to do it: + +.. code:: ipython3 + + # setting the seed + torch.manual_seed(seed) + + grid2 = circle_grid(1500) # triple number of points + input_data2 = torch.zeros(size=(1, 1, grid2.shape[0], 3)) + input_data2[0, 0, :, :-1] = grid2 + input_data2[0, 0, :, -1] = torch.sin(pi * + grid2[:, 0]) * torch.sin(pi * grid2[:, 1]) + + # get the hidden dimension representation from original input + latent = net.encoder(input_data) + + # upsample on the second input_data2 + output = net.decoder(latent, input_data2).detach() + + # show the picture + fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3)) + pic1 = axes[0].scatter(grid2[:, 0], grid2[:, 1], c=input_data2[0, 0, :, -1]) + axes[0].set_title("Real") + fig.colorbar(pic1) + plt.subplot(1, 2, 2) + pic2 = axes[1].scatter(grid2[:, 0], grid2[:, 1], c=output[0, 0, :, -1]) + axes[1].set_title("Up-sampling") + fig.colorbar(pic2) + plt.tight_layout() + plt.show() + + + + +.. image:: tutorial_files/tutorial_45_0.png + + +As we can see we have a very good approximation of the original +function, even thought some noise is present. Let’s calculate the error +now: + +.. code:: ipython3 + + print(f'l2 error: {l2_error(input_data2[0, 0, :, -1], output[0, 0, :, -1]):.2%}') + + +.. parsed-literal:: + + l2 error: 8.44% + + +Autoencoding at different resolution +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +In the previous example we already had the hidden dimension (of original +input) and we used it to upsample. Sometimes however we have a more fine +mesh solution and we simply want to encode it. This can be done without +retraining! This procedure can be useful in case we have many points in +the mesh and just a smaller part of them are needed for training. Let’s +see the results of this: + +.. code:: ipython3 + + # setting the seed + torch.manual_seed(seed) + + grid2 = circle_grid(3500) # very fine mesh + input_data2 = torch.zeros(size=(1, 1, grid2.shape[0], 3)) + input_data2[0, 0, :, :-1] = grid2 + input_data2[0, 0, :, -1] = torch.sin(pi * + grid2[:, 0]) * torch.sin(pi * grid2[:, 1]) + + # get the hidden dimension representation from more fine mesh input + latent = net.encoder(input_data2) + + # upsample on the second input_data2 + output = net.decoder(latent, input_data2).detach() + + # show the picture + fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3)) + pic1 = axes[0].scatter(grid2[:, 0], grid2[:, 1], c=input_data2[0, 0, :, -1]) + axes[0].set_title("Real") + fig.colorbar(pic1) + plt.subplot(1, 2, 2) + pic2 = axes[1].scatter(grid2[:, 0], grid2[:, 1], c=output[0, 0, :, -1]) + axes[1].set_title("Autoencoder not re-trained") + fig.colorbar(pic2) + plt.tight_layout() + plt.show() + + # calculate l2 error + print( + f'l2 error: {l2_error(input_data2[0, 0, :, -1], output[0, 0, :, -1]):.2%}') + + + + +.. image:: tutorial_files/tutorial_49_0.png + + +.. parsed-literal:: + + l2 error: 8.45% + + +What’s next? +------------ + +We have shown the basic usage of a convolutional filter. In the next +tutorials we will show how to combine the PINA framework with the +convolutional filter to train in few lines and efficiently a Neural +Network! diff --git a/docs/source/_rst/tutorial4/tutorial_files/tutorial_32_0.png b/docs/source/_rst/tutorial4/tutorial_files/tutorial_32_0.png new file mode 100644 index 0000000000000000000000000000000000000000..31f3e55605cb3f5f13a98c05530f668b68589ddb GIT binary patch literal 62030 zcmX_n1yCH%6D}594tEId?jBqZ4Z%IQyIVM1gS#ZSySux)!{IIol0zQ9|9kbSc51h_ zW~ZiSr@O!IuP0JfSr!eM2pI|r3Qb;4N*xLc+U%ntBEo;XiMmbk`nU>Sr^DK3!Akx?gN~+!l(6%I7}q3fo@z+-0A2_x835T)TkBLd1H#T$xeu@E8|r zXqXYc)(A~bq*wbT73t!p7WV-4@qk66@RdOJzgxv;#+ct9FN$yhxWoVTnMaO52LIRF z`jb?sbMnViBh0V0w05?hmqa6Pm4Yyy}D1NJj^n}tAt-#;cEktrTl_$uIdYP3KAt`JBT z6X;}fG@Yc+b%P4DioZF%G}_(~3Xlv5G{Rf-f_&OBq60nAK0Su~t?ujG?J*ax^7!8M zD&1K;b-=l+$8adjV2OXpGGLg0&M=Az80q##FvOr30bi( z0+_0H%|PI;V>u}X;VA|){)$#Cd7N+M1tJIVJxfH`{>a-RAh|aRSpjeC-t~MeOvuO< zGl(ECVueZp5f4xsE2t!y4)cm!@Xg6qCe-T-cb0ly4FI4aS|j;VIBceV3{oCI!u04S@s z|D{34!;N$jJtd3^L;rz;M|fO0K0h6`I88`gId=W6>`cXVC_KNhq(_ zFfeE&6d>IfKHusy0fDx)*_LBy!JY;A!W)GpyC`=3U4d({j%f&*LE6+Khu%S?u#-R} zzF}d9wL;m1_M-=#ER65XJw(2ND{+gLr0-ZksJlg=dCZTmnPKjU4symJqWi zoD)Q065)43I5<`Rl1)!>`#FRcSONyniFP2n_?{<~Q(AI8wb$0$C;iRQr(<KN|57+mj1{af<8oBW<%Hy$fu=OZ=xqCd-> zm=c2@ijJmS#Du66p9BKfP;KBZzEBjoA-u=f9<^TvC;Ci4J&+kszC$-dbNwg|^F51w zSD+X?+r?=`JyH2$TWMm{)g#hZ$F2V@@TCcisJx^216-v-{25+b?Yea#WP*wVpMZFA zRdbiixDh}Z9fa4~KF5VoGPYu*&?cpzL+E$YiNXw}j1r*#iJE(r8oE??)G}7>y+ zg7Y$@rmN4K=YsYoC_<_(dA{^<8wfsbL3^SK&R7bd7^IVk2;0$2Hu_o0?m&j7xcM3x zbi^T{knv^f&(jx;odFthxeoys^_J`EQbJ%3Rn*6;doy~sBqDCbeXq{pip$H?$El4) z!ucN7o#CeN8NL{XxJMx(-Ug_T7w$4~oUq4ozuCiq#z}iW)bSH^fe#gXPjG(u)&=`? z0SC4FiUq9dPW47gP-+{^(>w?ua%L|?-wHHa>~TO&kl$@TgeJAMdl46eZVTssoBgC7 zKwU&4**i%aOow68qox~C$}TFrkA#@YiHT_Mvnxlgs?bKPjV??Nszu1rNFvrVLUaxT z{PLX@R#i6kv4O6G3kVAB`&&(sPe_hlfeY=dOa-zc7WP1( z7%W3rz1D}ph$TuI%!a=vR+rl)=^z`IN2Evu0ZNc5vLz6Ka1uoEboqUfy>Eg#h#b5p zwY~&Tj8D52@ZB(EcoKui6f)SgNIii;lIXY+*-S8)f`M>;y{5)U!3c9kA-o69aNXxB z=zEP1YS%u96h254%2*QoP;9{v{`nA=@fxYG1mt8hFsv}^+oop$m9m)=K|jB6+|Wb& zR@eM=_k|L}Oc!N?G5oM-sp=N6?`W(W2&kXgam7ePDQ~_oLWf~F-AV5|!YR>3haBbP zS#35WK^G9K7$i##JW{{`0rAi%KZ+B;ncif)Fe^<7y`H`ZP9Cuu$_j1JtrO!RF{0s#CsKR39(q zi!;Ef5Dzomy7ALn03sxIULwNE`FTh5!_o!=)yJu-Gsulmub~9g$02dc1Od^A6l8Gi ziVVcQPSVeT5?{IA&?rIt-w@}3hy{ogv{>ozfv(BnxHJmM1D`XcA2z5%f($rQf{tJD z7!J#S3l)C&V-j>VF6sa0m@T6Iud62E;gZ(X&72$`7t{@p$D$c`jKF33h9l2FtcYrZ zRgz1<6s( zj4*4PHgY{~nsif3Vc8Ut>EBfn(lN1l^$^rdNIs_bjxX0MNSSfRBKbcgarln*jcNKR z?tj?frRNLfoL*hVetv7NpGt;x9f7Cp>(o!W~Mf%t-^j zP5UDl2`D6oe1^teD_oh&XUM!2f@|06Tp2x1Ld5xfDxROzKhYQXTS|bsFuC)seN|NWW+?BcI_G6d2J?%DfV@4nz;T>BDjB{#8-9`XOR@d1`nY$W#CEJR z-i=|c@8G`ibDxL$_)99$_v9VATt>D`g8>VSF0_&!o@hs-+~A!G@Z%X&Wmei$l9uSm zRjfD8Vd^5Q(er{}wf6z#bk}`j?W=60Iyq8$6+gI@hVw=cux?EmE+dGENa27Z!Hh4# z%*M=Y7c43J%#zskory2DHU3_3d${e=aZik|JFH_~1M-Zo5*Xbg*hpD=#icE5PZRnV zjw5L15_|~}EokotCetGf*zBKM-Bc7yDl!-)siS{fwhKqKiPpbfWX3k?@_c9hv32kr znP%X%?-*~dOn99n9^0;zF_%TPRwZNgjF30$H-~XOerZEo%N*Zdpm1n9AxTpX(XA>IjdF6< zl_|yxw-C30H$S9z$i1xJLqF$oEtgSg|zRA;dY>@|WnLIHUIr_B`E%OTmc zA0k7YPw_L_eW+`KLzUA9r=aL-xQgcuC1>1KIaik%O5WNm8I(3SriEc+HG#oVEgh661+uSa)yKaY2}`mUf9Mp{jk zCv-%tJ>?3P=+bsw+ucP&(C7T@FZi@8*3va_e*f;h3W<05^D}74=hYJ!Ror^;IZtu78IM>kd0@YU zEp`9?WYotqXOEA8tsIVKk@0UYPtE$qFkJxjIX`s@Dh6ziy6(IS_(AlP&HHibYTcKVvDEiR z&sJW-<*5TQ+RMKIR!Z1gK>I&j=GSSOMxB3=80a=wkW^W|2$?~`Nc=Ff^`Ta|odj)& zr+v%)?Q|MS#vr`F=#uMa)I6;bZrEb@JlT@c{Op?Ts8wR_nB+fFxmov6zm>({En}!| zf}B?Sa2ZE=I@V7_^_`XB<1QCWe-I#|)KrKosN#svb-Q%N8m!RwAi>C!|B|oyE&i3@ zefngX4LiI7Gby+{h_ZJN>I-7<4t_Nn@9_2|WR9w`27-2^VzhGb+IYR53K8qfskt{Q zHt3l~Ea4hN-E~{ba9cZ}y{@52_jL9;uFDod;Un=dAF~!>AfPGTr`-(-^!KbNI1i#akZa|Ut^O>+CejG$B5Ati1M3_=u@P# z_So$xNSi1u`=6hnSYQWH{06?ateP%QDh|+x6z!MhMnHJudG_#c^3DE{n#uhnEsOAU zN@`=8R#h`nyUp~OV&JpiB9P+V8g>FXAO@+nb!l$(6@{6lF;7yvbR!>RmHL9J#^;P3vU6y->~=?P{fmnNGSewfDV05EVkS?u|EEKWAFc2P ze9w3HG@EbX8W?+a7&-IvHg9h&Dr;R8h&{zY7P3!3oGmzpfq?F)@>yD{TP;$viw41crjs9?PC6OT~yHOA_It&)HTuC)!KidA#|fHn9`h%SHatI~q$Tz}QaACc|w zkk9r~H!?N*z3yU5GkZ$)ksBbP*~d1}4}gsQ*Pw(SW@<92V}J{kJS=2EOSX?U{S{*1 z3d^h2#-bo!@~6bKcTe+%k>oWFTmy{j`3h@f5uai{-@sT7-)-7luYZtrXU83#Q898z z7{Xfs?Qa=4GR~nP&F_wrXr4;Q!H2V|ZYaUjyu;!?5Ng)8Q>o$X&G<7L^1xQGKQi+8 zjh9=tGI33hXt?FOr+`7b)OmHIOd-ybIhZ0zHftG@9x-pKom#klIZ#iHO||Ml+o!hM+s#EIW7JHRC{nE-8pl(CRHID&7g zc)rAl_EE&e;nFfxI;9DwR;1C1xkO_vn0Qx+O+x2dC=JFRFu}@xf_(p>DbFZ@Hj0n)- zR*f;zwlrtG(Jske=itz$56#Nk`O?5(h2_>Yt$gNWrSCd*+w_BAw*bG}a_=lB{hMlk zY5oYqDkIR8n|Q^5CpCSz{#6%R%Rn$nYbP7q1&stf*0!1eyMB8W?7L8Hz-biFmhAt0 zh)TK~$~7Yz_VhGtYoG0N1||CXT|NR^9Z~#|?39{TrfGj!ZZY`r9gcx!BJI+6dGarP zR6i&ZmXN!3L9OaRQ=6oXtj1bi2IG1nBak*5@#6Lf%HjOF@VtqfV<2-CLN3 zGliT@TCYD}-uR7)f%R0j|1dylVSg(gZ@>Bhll*#rooR&q_TbIr+~fM^^v+er=7GPF z%c10t`RRmXRzDB-QH$op-Gi}^G3(TB^zfFZk&*lT9D|Pd=AA8zG*=8 zm^QG@m=c_$oumLPA2Xwxa;Q+ zNtV!Ltby8DZW`U?h738lcDpXhn!j-ue1pZY-jlg0v23eZz;WaKK|ZE*w|~ZhmPMpa z5VWlOCyygRVM8&n%-=$da2#RL95NkyXOiJ~NNR=3?sa@XtivMGSBB3Of$&sf<9OHD@Y-vlp|s#>MQAV2LJ35THrms5wat(1_4^LSsaxyCD^T; z=-|;rJNqAihY<4Ro1?6IdZr3m4vw&j8lr)NG09&NZIM!>Uhf#%^;UZM1m3WPG&;@S zA<&_>2u7@XIWzcskHp$o0fzOTDBzS!&^~(>-^EJ^>YM^_M$ZsQfQ-))S(>{ z$FpNQY*yAXjQV5D^-i{$STyi0!S}?dSIfCQMt)e9Iot`3mNf%7m@Y+HV)qhSC|0*LE5hCS@{SAiw>t z>J2TTaB=WR4VX{K@XY~}__5h;nFOv{5?J?ZpOMeC-m}(0MSKuOrYg2x0i)wik)5c_ z$&$hjhQcV4JJ9=VXiciuuc-(jCWY%%a|UYV|Y<&k!TNXS;2pD`8($d?3+y zOp2b@$pGOfJWQWsx?cx%mMEznt*e8l8ezQOE0<#PwX`Oj!Gij#rgoHpW?f;Sy(vk} zju+Ww;=bO27mIw##)z2FdoT?#%uWSR^J$@o&Skx{JG|?J^Xz4(OIt9sycP<|XSOMHf9)d> z_>5BW+ZrfKJiYC<@EyD@7^}}m$;aMOh{|C`f_~`0me&2m@iE1rM69ja)kGuE4tv?% zvopl~0Vld|#{ko%)a-$6&kmVHll5j`X;o{B z)h#bs;GJG>fQR7P+8q2P8WR7_;{OomRO=^tmWtx*-Z;O1f62Am{P4>e`7N~Z^${Zb zNJXhpiew1*I3XT_+%gyvj&M za)*HjF&)OKw;b4ZbY+5Xcg`JE^9PIqD6ua!`GS$zTt$sAXFgH)pp~VsJ5k}A{TPyS zDvZ}F{PlZnHiZZmhV47wmKSBbC@&9n*T&v#>t2|%9E11C&l;U(KTGyZS@!#qzxn}d z^sGWrH1c{>Y6#oc5kS4W=c5hH&yGVXM35%qmVeN>;hIcdhew`XgJ374; zj=;rkB-C4}r-{QTyp1Pdp{VxNFcug~f}CzdvNkwNA`^c#=Kq*8MC>?8!B!LV=Z{o% z&y`o4SU2B9>UAaVAfUDd;s<{nSR154i<%GAJ?A?@yiiB^mH zd>ENleLNVSCR>A)x%NiIzS@3Ebcru&O<5(~*%P?n@I*s4992Fgi>l4>@hO4Al0hzS zH?eh+&tIMXl#<1eEm7G$zkJHF{p9SABW?10m!p))&wkPHUL(Z=-hn2o$vZZBD|5)C zhar-a49~LG_2V;R(c&@lv1%xb;V))|Tpn=53caF-#}s;z@Z#n5WSC_Ia9oRhULP)J zSQgA4@|O%s8u21?9QHVF1dqb!a79L8te#YWdA&YskN`hTiJ+g3ODlABmIW!4j4y9m zZz_3+`01o3ZN3hOt`B#!DHB(NtA>@j ze+iU?`Nrh2D;?^Lz1CR-r76-1f3gC+LmrQN2|^F%3VYZaBFEeN?ZC()^<`DMUG{#Q zvP6R%`TXv&uG`qC>T_7oYs>oeaRtx$n4G+EPYY32dH%|3@h5Hc+TXKWJSn1=8!o3O zeSfZ;tcD_Wu@>r|z#TjFUK@tOAKOV2w3{g}snhMix7P7kKvicKL^%sYSvWs_cErEA zwpBSEAY@u?=xJQ8+~ZSXSGw|j^(}sdgjvwIWV?(*b!W75R_d1~Ct?nCG#yyK4`a?X zEA}9mnvn$m5a8iHa&lcK;LZovaFtpl;+*gQ)td%8UI)&+McEW4yE!B&<(4G6IJUM_ z+GZW44co=AK%yyGJiALWdiy0e%R+*&jA4`8|l*;bDB~u6*gY|LUxE;Z? zZCU8ifM2-k3foLX=GLU3V#sAnt2(Z;Zn!;j&rDe;c77}~Lw8C(niHTw!W3u1L=PUB zwav4)AXycL8A4p`82NKDIbf;G&-lY@Zdyc`)1P-ES|P%7waxSia2L&#{<90sC?zf; zd4a;jR@~Wa#spl*u#5n8(C#g)@Pxrdm=ol0!Kx%i2O zV9Punj7Q5fuf`i+=K8(;yTyQJ!IOXd^YzVl?bJ;HXyZn#s8H>7?b>|BHcZZq)Y;?4 zgW`FyH*pTTUIh3tx!4|-c}%)No@M|J&BNDFHI!fBWq7gP*Ou`R>!e4DufD%4FY;?* zAVyl<*KFj3{p}G(!9hH3iTj5?QgK1s`P|;Byg#c#>t;9;H`%)jReol-n*)5ZQ+&A2 zC2eZ&0L>RDSo8p`d3t|8{9lust`J7`O+esuO*5Mo{tFrrdBdfAP42nUi-9WfbHb*&z zJgXaTx?O|M46V~5>f_fKjWZvF3Y4PO>RLRA9}x6Ihdm+~`@@CxcQe^HWhk3Su1xfqb zb_a?dk|{)TfvsPf%aB6^TS{h%?8}^db?R8~v&wh*l%|PAg6>C9p?0&e;p^SNWC6M@ zA{VRb{qjW4Apd?@uLUv3z`q z!ojz=`*2=4v$66)wpkxtZ%rNTOT5gZB585eIu);FxQ@9O4h_0oY8l;5Ty0Un=xE?5 zpKj;4_9QXnwcDX(snbUFC58~1-j9Odn9|od9(f7@E z?4GXi-~u(huFkkVf-+#!{k;8gqMp98xHXLdm%4mhg|AaZRDX}7NYYPFQTzx8>!7XW z#ypFrk(s*-kI?D z4^AhjA?RPfHVjf<4|Z^8$R#&p48|8_L+ioA58r0f692@1j49;dVEsvy>P6NjJ<y4TsRUdb~bG( zp2NV*Oh8n@{MuReP0agA_N^hcEs_8?GjTF1<4 zhgQ`{LKD_`F{Hie7~?ojWY^=9G_LT=di)2xT{09~+WwN#)AQ=ch`Y467HOnxFm~qd zd-Kojr9E;3T+%o>5E9=qEl4i=!(zFZK2aUL;?d{Q7~rXKVeQ2Kq zssot*0+==cOtS#{xrlk5SQ3+X5|vmIl^y}LUoTHa~%BTK@NH>6Pqxrs{JJ* zXSb2(Ivw4j&lUw>kNuaPtk3tt+x=fdF#*q-VZx6vflm)YcZ`fSE6)d+(Wg5fQBtkT zYS06rpWoCFW+>$6exrkkGyTudI*p~1EZsEm<5}+MtBo|VShA_pE*}$&^#Kiij_mC& zd}WqDLaLex(>~S<$i`{(yb%qucoAfh$0<#U=}%J|*?%t~e%L<#M^@9SMp zcUVb*YI4!Bd(+r5cMqD!T=&(sZP$O2yiO*QG9J^21-dQMWFB2Vf5i21kHz(}HG}7g zUYA%Rc<>MBLFb=$mSTpoYQGo8i1lS3>J6l-K%9Yt{0CdUQ4GvBZC-cF$|xzblb_B; zYAoQ>mW~B?xu6&hDSMGfY9|K8r*LG)wO{p&4{ED710PR2|2&X&~n{N>}K;?>x~laalJQkUQN4p z_{&uRE6>Om>5z;KOU{L~T|io4?0944Zh83d>xt;F=gD*qtn9)tu5a(9J} z&kIzI4jmla?jkqq3F&q?E0#*G7xI`g9l5=B^bGfyf)b^`*mP9?rs{n-lwt#am0s&Z zjUFw&-y^2pM3n7FeOqJ1C+`mL2M`yL$ir+L>C@dm#evk8*m&@!8G4nWoPU6BPs&B- ztJe#BDKuAvR%R!MB_Vzs!Ivyp5a&1Vmx4pJko02dCOToB^_W^#xa`n8zdDdIsfNJm zljz_Gxxm849IdKEgY=8=L=c%o1S=fqX6VZRu{j1inU^f$TzC*W`gQQP_~O-=_=viH zE&aDtzAi7LxlNZ5x<4)<(HULV2$|(h3O!_vj)SSh=e76Yk-wh@2Irb4JW^6QZ6OZ; z-}(>3-#$H3vEp(*D!8DzZ-Tf-M5h0Moo8F5$pV7bgm+*%mk%`C|5m|D`xmV}am)XaW;FPEt<9 zq3AMvRdqqnhU87}g_=X|VZXN=Ed56bot`;0`3VRcUjE8h6=PxUI0DAN&*qgjIgKMx zCg1dOn>Z0Scyb#auf5ulK$JnBMAi#@X%+Rm8=9|`kO?!n8{%Vf667}YH$ztu3BdIC z1iZrlHgX4@p(3IrQ+w`zuU$Ov4D; zNA}IDew5}o-nehy{@^fe@Ak~9{D+CI98-dUwpqW|?wsS(#j9RGn*U!I%?;h*PJ$T1 zx9-0Bz-5RKF~OEo2ke~s91hR$+51OP2p0Rd$288xQe5sIJ@D*ytF_BJVkMzt(;G8F zDBpIuB$_g0yK{zayV?`*Ja@~)zKjeG)+X>FT4`))=(73~8cUvvl9c07#!vusoAZr5 zKTPS!Y+FDd#Ey}zHxWxyP0n@jw(oW&;YhgAqSzCiWNQ4i&Fjp%U0D;HUHuvnZ_G6v z5j#*_YI0n_fNWi@ovSMl-@;ejYX%R}HFW>vIh7}qy*0i15N?6tfP9QW-ixoe5*|vq z{xuZrL!zyNDRHw4Owz8kpc7-=j3{T!7A+t~qoE=|F5IDu0|iL9hRN~mJ@6phJxfv|-9{B;JY7tR*`m|p zNRA(POXzRp7suh)HGTI(->109)HZ?4PD7KGoU>JWB*G;kq8_zA7IbzZXk9ZOdIo zcR&I%p4_b#b~!&4_FABhyVTN}{;Dwc%e}*EcZuBVrstKAvgOS zB@LZG4l*beu-;6{Kx3eh_`Fh9GD}fs4vv5n7cQ`HgaV5~3I5UwplcpxZUfH|U#D_; zo~l%*W)kpF9vnv;iB^{+v$BB~o6#CVe+wlZF@LYs@?f<=&$9C7@~Mc%3myAKkIdF3 zG)6}rHzmy!DLhia;!=J=!XYgy``*%2o>1~_W^X2Bgmj+3mj2!+z1bY4hzx^@lS`o4 zKU_E1bGkMln=>DQ$@37?Y=vO6%|Ci^&+DQpV^Y7qZ5hO2`}}QMuv;v!{Y45&St#EgKtK-|Z^dXLZD}s!kKF?XpR?VZlaAh;^r_D$R>~z z3F~v$x36ikpsuK(d_kJ4_V|)GWP*f2Hn2P3ZmZ)lFl80>V~0D^lUuC2sjO4zjq>Wn z9C_m5I3HQObE_|jO69k#`U0(J7_KOGP%qGw?1<$fl|Hx4O?PF7n|!Z`m>XQ6Va_>6kT0gJ;qnG)%UqjKfK+DxyL~u zwu;%m)T+s~?xg11I(<&7&s!RLGKdz}^v1*cKnnMiF3}vvqiXPlGwexf#_%aNLY?ug z$8z;fbT#p+JyBHo&CToaivtfBFQ%BEl3Sid0BiVc2l<0qI%8=oE$`nYBSqI<^wTh; zm{HD~d9m-1)E#HrNtv*}2JQ{$u^wg4NqXD9xhN_ALJS7n%y5k)5#evQx^2DrW15-p zNSioH3v?~4@6MvX;j(VFBw6xJnrN9LZDFrFJq9_WD%xhk@+=V!gk!T744xM&JSGvQ ztyccS0B`)Y(4B4vOzA?sCar_#`^~J;Ccf*8Ss)7+{ZSe%bd*>SiX=hRR69w+(nVP- zF5MKc;(j+2A*SkY!t3qqB-LNOPq!eSX=HYIz0%wC$+@4cI%i(*2-n#0PIGBNhv@1q);17`=S;OV5HRv-s|S>bSu{lJ%VTAR&=t0Ljb; z6WjHG?4l~zYN>e#%J*K;`x$Jcozj&y=!Kt_%MtVR9Ur5UgH%JoZ9)RE^`UW3KXBOD zL}RRt*wg7Td}nF+Z8;B(uz0BWMZyC=#bspdH^0BCqlM#oTybgke2wgH`{NF}lG>6^ zVY@bm%*J>}U89CPSHAlZ8l#mkcNV7g0iJu&{iQj44#uw2=`tKkREFtcAp^eQJOm|U z`8n3{SF+7w@)Da5t19d;wXc+LFloHo~n+!?8wN;m6J4saRlnVg`mZg@0o zu@Vt0abLQSNSI?JAH}xQx zEwI0WI9N3AZ})Yk7>{N_ZGGmn^_ZJ|(tKh6WlKDNoV0+it^F`SE}g$ZiyR9b3#>O_ zW@`~PxB}w^453?o8C=wGu$mV|5C1C5yra?@OYXv3Tg{zTeQTY)<_ti>oJ5JYe~OSW z@|LsE@fdOt-4uzo?Z{m>HM`4#H@Y+!e)O)WV`4n^AeruxR59)AHZWOHfT6Ptd(066 z_%clhoS#3IJ^bTfCY;kU9&@^G+i6^_btIQumyNqmQLndf6m{jhrd~EVZVc|C94=KM z*iO>w)&@9})1ldJw=EhI;W-M6C0%#x23zY6(zUjpR%-~{E*6J=Z-%Ef{z%Lk#lK{$ zTfDZ$UH2GB%ggh8Cg5a_?%58cl^^>lIU<_gj*~Zo$iedT5^Q6<8Vt{H$kcLBZ#BRw zgMnz3`spy`t5;{!7SWl8MuZ{Xqy^s`N%mYEzDnL?h^xgKn?<5`*iB~mR;HU4}+fs>V35^_zRANF9|J`L!dea76-$Bv{_7d zx4O_vZW_{L;04J2N*~?LXjILPcWt_Zua|#&vewBII)XzU{Et~dD zu57fo5B55zeEcss3_B1p^!u>6yUFy@%HEv&MY%2n?Bo3EXSiNl%FFAq*KMHBeed5b z_?OX#=F`P5<-s>Znd$isK3cji>{YjH$^?HaCXNjsn2zfaH3XN)gslx%*6IT%KHkPXb? zdiVdas=qRnAT9#;0n;)V#lFkBcuFV(wWBjk=HRffy*X$>3u+#_JAK7MgtBtPeY3OT zH6&I#4rM)1?j3Z^i$T(`%f!=1*|1DOt6j6M_yx}&z()pbhTF?w>!p1Fr4Ox3cHhR> z(Y=h<;S$@b?9cZHt17r01lq}uZI6j~8GERZ*ECMF!S zj$mO;Srk|YyZ+Uo6}bAsu*5hpBD8^jThoMKw+gVi*Ay3IDFprinl8ZnAmjNQWcvoo zZl9PPg_sfA>uJj?t(5Z(H&lA#QEHQR+1zq-_A1*+=}{7I5XzipDRw;^&rrTY^L9w& zHQU2NzJ6JU$*CzqtK8iwt=++H`{q7Sn36CP=R~h-vjK|WQlQ9}z4nm(v&s^SyBLS@ z@IUeEMo6nK7e~06ef}WVz-D)K1&!=v2eYN8=OYFKKcehbL%+A3^VY-3pC9p2Ibt?1 zS~w6)o|Z?l%zRf#;qNSUGW|hDEmQ&T!LnQJc{9ybhhr(O>Fr!Xg|X@z=}!r&b^pE( zh2kOOOWcG9(A8d4irp_1ZfJWf=e5>7(t`kY<)hNrdq2i3Pb=w>7C(8EHp?gihZ}<; zDZQkmx(JLwUl8SgZB4jJI@0!eSb0p#DP1|CDu6b&%gUtCV=>-M)^S4pR1Oe?dX2*4 z&7C$LTDfDjvtkLRWUiCB$%sTmzwSmGw*Z>1*q35gV`hlv@wGq@MTIfvE#2zr=nZg9 zj={2VLz#X6n4e@?^>Y2H*7VOL z&xy)u5QLMMK$|dD*j}037DLH9q+x1CNpo-iQMnotK25V`N|b?v%uKy-A84$ySZBz6-rK zZ|5iYY`TnpY33?VVUVOwuu04Q;OhH$cpke(;m%{@WJ5IKIUdzIgJ~e-(pLxjj1Zy9e-mffaU49B7VzECkRb{XL%ftEw&liI*VhWZXCszZupuGu;fdex zXY_fn9&z`~R%*$O5g{)@i#f)%2U9Y&V)moB|)v;C!dJ~ut*L4Zl5s`1ZVSWIN z?_+UYg>-|?M+u@p#oJW^N;S4~TmQxiAC4(dV9Z+1OD{VtZJD?$Vbq+M{ zO$sX=1Of>2-UwdlK4XzB;G<%2W~@38=+xcXRc%hrKmsj*Ym9D$08^m~&zrm{m7AGv zjHD}JC;a2@z8@61jN3dCb+lr)rRJK7wl?*KdRyJbR-q6^%^=v>dMm?uUc|0!a)g$< zeWZj6lrJj*--X5H6$qhJcv(V~JL~21UExObOe&2UYrD_gU%#Sr0k$~Dl9dGN3cfB# z%)HMz({U9jdXl}F=r~990=Of(1}oRJHZT?&A&hJGSMq5uD}+p~L+Uj5w=$NPo_6#@=SQniE@QOb$bQx1OKynGJe`?7S-{rh)kJdl}?M2*`m5 z&g(p^#C#wjTG?e$J(wQr`D1Vikvwi{DQ~2K+Vf4}1V1ec_5gAtaa4=I>ertWJ-mK> z(`T=Ua9#Tw^&3g?3F3}j0R0S};E9fynk}=)nnZ+U|H<}x19}OHIlKY?SGMm~3Xcg? zS*c^=^Y+Q7nuY83eyR7gWDP_NpB$L{Y^w6vX)nF1C)@w!AHT%A$JhOC-+TNVspZeV zs7gUvLvu44bgVX1^INu(b!`Isyttv8IgLo%P6HO{i}KR27((_4$YaWrU4bE^*`3R= zMn^t$W~9uZ!*zgTfT-%;Z^x4>qdh>wKpk_p-dfmwnVR;`8JTld@!e)t!_&X;B@;|( z^p4?=BR}HbMt&PrNFdfpD6BR9ppQ*hWilXpT&Dr*dJea~lRn&LPF%}y!AC^8J3UBK z@eBRHBZ&;y;?(Y9psP&GbV#4vsb{LRtyyur>ATR>_Z)XJC6O&g$OZ6FsKkm@SbS3f423qq54FvSMd-g6Dj8vxS|r{un^Dc9DHgXMU{ITz9pB@ zL-gVEw=!_%aPk=$M~WXLTHLO5eCd`fdgrKZh1Jox(NUt&a4WnWkIf3#H=Om?tqTPy zRjl9FA?R1Se=&r=_4N(*x zJ3$)B1A}Auxb~U{R&<8(-w~tz`YotO7e?64!eeBLVNB2YylZH4y(f1%qA16`ALb!kT~=M{?!Tlom;yBe*=09vU7?seb36#gv@58lVCD@tt$IUtf~F?Vgef2?GV#6@snRu zW!EeHWPT*g*T{<66gu%j{>68;=SeoN$H}baytQ?~d^P393N2E4rS-^tm;M2lljRzF zd*?6ZghD{UqSj>Ps~9KEqbskbot2vsPhkfZaakl^A&CuEfjNDYdd0@ zqF1$KugpGsj98@N9welsPlk#>lJE-ZN9uKmp=C4EW!i*6YmaWx!?Rj` zFtmJ*j&3KK(QJUNJZ7CrHgJw$#Y1(~g@NWC+45$g+xjbEid)Rse)ifdt7zd%T!~qhY9^>+Zf$+xaQ|&#H$wKV zRa}ISq_Cv9-e5~@K_*4Z+IBlUn9zc;5_bM)YMg@lGE)}Es~5Inb5qm$b~~hY?u7YZ zRy2#t_n!{VB5hTWt=RPf%>pO0$E2s3VMkjTg{c9{a9@0yZe(>57ClA5gQuweo#FVy zpM(vzTsLsRQu!UjqycqO6mb2;pK*g5e<0DzeQ;O3<4YZ5+sU#vHv0MHS?8na#{n$A z&^r+u8)2^<178xBm^@5$+>d{kBhzIshWChhDPPu42|YTW59FB;&;|?)T7RS{oNt!R z(&l(+xc}I`h2;p}KHJ?XGS%8L9>)`o=2*sAtK)}-C3}k~<=%-(lkqu7jue!y;-si3 z_e4NjyRT27tlfK=-}_xeg-K!62ne(pkTbrD%9L0->y~t{E`3qD*nB-t&+H)RF=sCZ zifylnrYQ`B;Nsw*e*&?qh5ks!0;;u@-XrO6xSOs%bx*y<2>c%a9YNy0h12m{zS?DW z#m-$lA0RU(D`eeI_Poc!;(4{}+7&w;U0K}S>t0l$^3b{q@kcaeNKR}so|D_%-lS{h zw|vaoKa=I93$Q76#-yFhu_>cb>ZKxE(&D{c|0LJF1ziTlV-pU##SP)YPJiS6bvILG zhPI-wscUPZ2FLH`UB`RhEn{CvC6;cX_+(wp3@Neuz@e%5 zd^%2t4NX&c_TQJ#txMa8`!r1pccqt@cqeLvWU0fUB^-EhJu~Fsm#%Emx(g=!C+fgE z6imYk6T}d>ITrr1C1lxSul|VJTbCf>^&9vNlb1gFg$HjxEo9v>{kkx%Ul%-n9lNF; zRtJV>S1u&Kx)MFu#>i)y_)L>O&iS67&%YLcxdq#Jeak}pmRZ}c&12{+EZWWw6K)Q% z=l*BY7yMMP8m|>Nmo1gKbcxTPjlCO-i?`rZ?3~ka1QRj_vao0arnDHH){m^X@M*2j z^fG(@549R-4Z5ZCL;ly?(Cw~}N&3=&hi*;%!OpYz`j zJ0}RtFcc(a@OJTAJ{A?tJD@Q3TRMucaM>E%W75KB9@re#QjF`~g>l_mrYGMU-z?(& zAC}h@pjW_hg60W!p3)`Kka&Y}`<`HS>x2KA%6qR}7jiF3DVg%#0zQ2HC%hg%cBhSt zZ#t3lZ#bdV6{?r5*n#)E5XONieB=lOrQQ>|+T&-(-r}%zEK9OrRpcIYzyX8m_m-8D zQ&vuowA2H}6o0pCbFCn*&z_1h3fSNf_71k_Ou=XS20p2CA^rJMj!U#U975s`dDT zQ5Y&Fc2Q$DH(6Eo2`L(s7_s=Q11Cl(2Ppgla9nC%KG^kB$TbIvD&Uxu!;5D4wekA& zP(fS|(+-+1AdDlie}4r9`zvUdUS~_5{`oTAd3riNPvF`Zy2*zx{)El0arV_me#X#r zT6dwK--R$%RfXe5_h9kDt(f&P@HCssi6{4~d#@^u76h_s&C`0_ew{nMn#R&Sd$B1B zUen~XK0WyN$%S5WK1_7I8rCw(~SY5h3;(8v_&$8kzjA_#^;^&Q# zWJ|_s#QE3U1dBF~EF>nest(dmphUbx6ISrrOHA23O;SpW={bKV-?J7&2AqO9)ywIX zcp7oG?l^@2e1W2Q4KBSRAVsTYpFaUB=&uf#DeAs&yW(K<0kvX3dEOoLPH?ct?WR&! zP=v&$SZIpHy=$N4tPUq~dFRtxohBHGyrK$}NV7a4kV;SpA1J!Afob#O7k{o5#5MRl zIv>A0tCb1k)~y-;rlTo*ULAk%>Hdemo<>HyBwR6eY&I2_%TDL4G;X-9F2TWJ*BCJ< zE8@O9x`97JtD44maQe?I*}WI9X;5ML@e7l0Hf>_;%dfMmv@~q}OM4H**`WERDg?*0 z?LcC%Q2fLWeQ-t)+7e0s zQo*M3DeS5Eh*Yfvr|=^}9EK7F3!4a3f@@>1K*bCSYz>>@L`B7}*cp&Gh~X)tA|Bgq zxcSGrhj?$-#}wiSvw#jiAW`u{Y45Y$4 z>>6#`C1SUS`$pQdi6^|8yAI`xU8~7+m*X*Y6a;ov!9%ONht!K6`r z`Ek+45L1Q2p>g*`CosH!XVQ}6`18G=v2FW4ygnU=!-gVZK#z7j^YnB^kLk&=$96?g z1Z~^OF(EwToF6lu-T6u&uPTvg%W?#o}Ls=^!8s3gnhZ{^QdzQ*g*{>bo= z-5EZz@$E}0-B>l%vVz6m5dx3j;J@GhLba~r^&2Qc@WYDD{Nuv$BqqCfXzI@tRaGN| zU}C=>n3ff>)-eQ{PR`k}lL9SN<(-M=v%Y8_>xv5K5SPH{_MJnVg0Y&NuTHw2 zuXnFucEMKC9I;&9?O6IHhn6XHs5VY%*PB`U7uL=govO*WHbY2vHlF@{AXK3;xx+=A z(f&eAX`(f;i$^3Bp)xY6t+ z5KTB(UGyl0o+Stb88PE<8vzqd{pV_pi$KA|LcuP~pp824L4847v%VrFim0Z5l@h$-Z6SDZwQ)5h1c*P@#iZIWX7%jG9Ac1T^J`2YR;5>~F+ zP4>n-3UW)C{ZlqS&s>jbN`C%n1A~Wj=HW-qqfPtB!q_xS_T`q6-r>;h$(ANiO*blx zTa2HB9S7=9ZB4_Xq_C1zi?%a#^no?Go}Js`a@lcv_0V2*F$xu4U$BEhpg50LH+cAq zpD^P@NKJqrH*CeJY3QmyVl1y=S?sHdoEa6Vcuajop&G8=)LB@(o)Uir zY4!x>7X|7K*Z*BEELp*&-ic&54?Sy22!Ylh-d4A&De-N^sV0+X5AQ+|3B&R*H}^8U zW+BP8z%`XJpqa>Eed+p`QfF)F`xJWS;B-onX{;HwrCiKXmmms!iUFlta2oOTz1B-y)iV*96XX$P)XkXtn!-Ku~F z*?NkUM0ORLl=@Z&9dvcuNKAw@un7gdfo(ZZnNQUIouGQ~TXrlFk3}^NJy22x?8Utj z#y0uMTh0)Z(-tuPjVU9oX}*K_$Xe;NPV@&!Iir^@dL6WErt@JT;@ z(_r4dEvzWs#g`}CL53^T4(l@wZlCpSNX0mhp)-5$4yNzg!f9Q4@y?2sJT-4Ve$&K~ zlKH!K^YNOsyno^OO^$`vRE0ObU&aSNEJsz8U|ng!|NeX~1G@*Dx2sq0LDLlc4P>s@ zt21j>HvNZnVdv%oj0Uw_biYC8?gy^g_iJ;M(khxn>oVvxfu8*{aYo1jQdEVmy$&t_ zHC5sD`!D9q69$kF=Oi)C$yq1%r+>G0b(zQYzA#fns`wkwEROHiB@hWwZyTqkF|u=0!3}RwVYu;_K6GhdNDEU6j0TPdfjD4qwx-}WtI;ht0D0AmSzY`pru3o9 zYE0>)#J7bxd4E0Ve#(3sFoTf~4T+?iD8i0SNoZsn3x^tusHZ!q!cLOC2dZ$@HvQBT zJ4yC9uI=_ZiMDoO%|L|^oY~=@*pxh}%YD`NlnbG<^kr-NfV5aeH;DJ$~+f`WuX3VWLQmJP14h zoeqt2uN{liWe=&_?Qq#Rm^6U*U;GK(Z(s&Z3O2h& zmmcZ#9n!Rp2l25^{(ALz{(5yF(&Fuz%lK@?`e3QaIuu0^r`f1R$E=r7V%ng>rZaoz z&S0T$6^Evh5g*GXg9mZh;P7NmoV?HkRz{f7}<&f{)LcEqn z$C&& z&aa~KAe3$zaTPqhLM4%>IjGD|P zSx@3nTuj?@3x2B-f}*g%vkDQ^EyH4W(eRmN2NcGg;?H1n*(;dRN6jNnpb=~9$cY_3 zXIJIdY%Bi=-Kr#2^@A+JG>Z|$1*i5kCIB0;whpw9J&6u+Co{b5p8)XayUDA38!1hO zB%ejBJ=8oEZ)?lNSuaqj7h+gmlI$I7A3TRbJ68tnTnCPXh}T@y7rai*pxSSz)>9mO z*w^nip$dhrap5-XW79Ju#;eC@HZJJdpIBiJv643($(xNE@4223?}vM z8L`k15K$1oS7`Wftbxh^&&s^I0#c3t9 zW;4ZaP8b6ur^WH-f1Q88b*kMyvNz=u7w4dNkBmkI_4~6{M!X?ys>;hZpF>tg3ZLF| z3ICn(3(I%qqB>Q|@P+iG1MunwKWy6C;=;I7Oggce6K_L<9g-MT4^~YBr>e8Zy%~rc zFT1+rJGNFX!jS&jAFXcrn40qhLzBl~Nk8@H1%WsLjleAudVM4oN%jM`XSb;djBWoO zs|z2b)VB^0q{ob9aGU#46pgNl7tl5F0xG=A$=-i8mh>V7BrDZaSx$UtsKQBy*wg5j zdRK@b&8>K!U8PSTWx#g6xBMkC6KyIzte*MJU|Y?<)Sj_Eika#E>Qei6%_qY;^)!B3V?tg!3}bA<3Q;wx3m{-(g5!$TEpV zfp-f<-V6*GDf>^LDKSXAL17K5a5A*bAJCM8SFVeZ z*ou%Os2)HNrzIo3Z?h@)eUf4((;?+S49m^V(kGDh&c%`@ITf#y7CVtxd-KX`wluf( zI2(z@j|Iz6P*8=!wOvo+?94HEOh1pWo5l}$>#(Fn`x9=c<#5ZQz@6I&GYALR~R@kK4yzZ+wTnJB!f$CN_skY^;-K z-o2i-8LhAMrsU|^gPX$dP8hG>n9Iv=&cWw5@C9X+cV9VwzVCBBe(T0C4f(_Y-Pn;| z9Mgh*a7;B2kAA4QEP2J9< zT(2h0k%*=Q9?~w=&clPx<^F+Z;xi56YzKd!B|4m3(t8m5s;e0P*_%}B{(z~&v{0=G z5C8De(ioP-j7{4j3S(7Kc>1n$Sh#j4zpUC$e2kOH$M>UiyU6vt>+}9AbJ)3cAG$yA z+c69ipGW7Ze}2usU%UFC+aC>Dv6bTYCyc-Tc2!6`Bn0e4Wfg9+H|Nu@cWAAJOU4am z%90HfS5)KE1L?93o5o+yABWQ(dU-s$&eRQ?=$ewkrouvOYQXlYTNcN3YR~raV!EfL zgk;OL6v_5@F3Fles?)*yyFZM$j$s+JaYUY$_lzIIs*(Y2!ngdrOi5)gY!`95Cns+y>3}k%PEKq}e4I`Yl62S~y?epW&bzHXa z*@>bE#tiPtn8A%_A_A;hy_?U!Tt-Q0H6urK=cj3_YXxx)mL*xgYByewKVsC=QAL70 zN~z9G;ddvD%gWsmjr58T?9DI5=G5riJ_Scm&?dw<`ShO4`TUo){IX#OZ4=|TbleaI zbqy8TZokgC?|npWc?BLLpp}}Yg~O>~X%-6#cCoTJm&fMM<;(LgB{Sj36dOM!<9HIZ z7+&4+YDmWoo3PP4p%-nPk%jTdHW#zCav^S`6kYmI(NKiO>FxiFrr4R>=`~gqz0S_c zStvrKd%{G9q+Z)%g>Nj&O>WU;`1EzbB?6n)nanny)3@zAY%aVNQ@Vqn8Q5iYnC35$ zxYQIPR_7Bo(qc~wTizP+ntqlPZ^maC9GlXkN%ch5748eI`X|70StuR_r>dbT3WjBo z=}KTvby@h#!Ls=8&-1x&;-ST$e?4Uduf8z{pD!TkckV1iR75uM1&-G`bXh>3=#RV=k_E5}$uHk~&+jU*F39(tBx_`ZB{Z7g6oc1|%Ne3i8YTg(URH zE)c{yhSDwN0k)Ss6ny9gviUkC-Vqj2QG;QY<1_PdYMH3Yk!_iEFjkaoB-0n#KT zW;`xic$&5)b#@frPhs^pD8i0Jr(OItbV~ZmQ7LulHhkQ(=AC*SyaT+n^*aV9ccEKc z<4H?b^*e^w)+`PGKtPdR&Cm|*NsWoYlCsh2mks?QO=a7T{SiBN_+1(?7T6Sl$L+`I zvSD}F`18Z(9MW-)G%Zz|;z-HFrf9~)fo)RC9l?Wq0-Ob8*IVZX;6*8ADS(u6f74E@ z<#9M{y!`a#{4jGJa~EvE?$lVkV{dKha}8dQ6w;oR3v`PU^Ga6<-6F`4B90$t^H0*vZ#1dL>sSe8T^p9g4=sG-qEC8Ts-aK z63DFzFITAufiIXgtq8%e&K+33J2y0eNeD)FYg%2Ah9oAqY7?v)9GHP8^XVKTkm~Na9MqN0zHKg3xp7hgDQL&LWmoM z5X*%S%L@w&n{enu!|Akf&g5fw_KAzR`ie2wZIPcjn~IA=8jeCBUijH5zl7D;y<+GP zToLvo0zG_cQG{T4`^<=1aj~z83+BGW7rU3U*ImNevK;PT{S_~7o*lOCt2;hsOXW`d zmX6!>qDzB?#VeSW_j6M&L8~LyIiA|-LoEWQ{a8f3khkjUH>s@r4^mbkkSGEYRFxFx zFjOU;rXr<9LDh#LVRHbM^pIQe>S3J8q2n|4ucYi25DgaX$#O+got3tyL2d%_Yy#&+q< zEyoSVp{m5$?Znz`bV!J&V|oIyb{hhSwb@8@#qjU(r$*fGhdUNgWq2`U-Elpp&Rg4l zp|W0=55J|exOi1q=|;b$^L_rDCZ2igA}#R&R3!yLAn`y57mA1{J^A0Y>qsf7s(cpN zAk0ll4^@@_ZPxKD>Bo|h?-j!eFaCcp+(rqzsuoe`-Ez=k_>D0cCDR#FpSdmNjYEjj%trgg>7C6rO4yUbi1-jGg!Y`#sN2`!ng8 z2akgUa5`+55`Sh1sr=F%N#KkyBOmVWeD7@}NjKhv2f);%1s1aN|cmSWTU60!^ zIK6jIx~8Nsd-o1D78TMtA&H6IdNi7Py0~aZZKKip*c6p56$K1U%>n|AKpXqHd(6o3 z!mS;f)`89$^Qo-<0 zq&pXLS(mXKlhP&P_VIQH-<^9E-|g7MqP)Fi#wBo3?}4<73!gE3vo@QzmaITi6t)x= z^W&y%Jaf)zoX{ikzAN2cp8j|aGna0`v@Avq$l@=Tk0&G8F+DXkj#pp2iamRaDX;L* zy=$Np?Qf2jao`;hnb;IzybS0egy<4H;Ke{gui!TXr2pGOh%^E!dgC2Fo5+jzeolEw zHCE7G<#qe<`wU+C$LBos?oCbJ&qWiDp;NnLKAN_SqS9(c49en)$-`N`BZmh*m=R*C zaN0GlnJ_G3VU(ikYA*ZybMh(!T_@gHxt!y(vUu+F)0xnvJ8O#axoqYqtSjC}vctuV z{f2W*pP{u{ONJ|v4Q07u_k(Vkv~kw`pklQc(j95!dH02^ryv-V)R*5XsuE99{56e! zE`)}vw#PKXE0$?mS4=Y>MI<7Wm?j;sOUgrRF1#hEAxl6(5e~AF?``s!w^hz&ec2RD zS!Z`H_3z=w+y^+f^9xPhPYsu4jbms^cYerQ&eqC&Ru=E2(q9A!wp8Y`tav+r?mL+? zGb0x?+Eluqb)^M39SS2lbR;7-JUV1eZUN8EU5L*x@zsmfzTuN^IId$R&t7mkDY13e z(6S^qJ@Fa4@=I!Kw9H$*g9U4LGPG-ZE|@TccBzSEwo9T@$L2kNS{1UzLfejrOgyMC zmQor*h}(f52^js}l2UFELOcvCmr_m*es+owBAb9m<}Xsp@HWiNWW}6~wSu??%d}WA zC%cg?8fWC7E{q%$sui4iOfTGCo#(!tPo>9))24CFgk!npU8SWM^`@!w zcJ1P=l`A=+dlnb}_%R+MkmFzAuH-)}=ds^Y#XUpE0dQ5f(ahh!6>q%`&^DpcFR4A9 zVnc88Tf1MzBiS!tSOz8*Hlg8A?OfjZtY)A4Z;AN$zfo59C!l_<1RKK2r)=L?z{24= zo7kj(A;f{5m&vXZ=$HNxIb|>4F}4$HA3(>%+ljIFZ1Q;4lztJGDlRc8^lha|FC^{| zYG3qD$fR7a=8HW`gQ0450>W$QJhSOLCbb_NmKgHC)eCre-D3QviA3_{_6?lfwJ(p4 z3N=T5v3h-@krT`JCQZlhgamG^0?}6_dEc<7@za?;Y@YQ0dsvttW_lRLOhsrsl`g8Wk zepGw?#5n8`HDZ8DpN}Pb_l5{!0Iy;2(V8`^tb$dY1$PV_MXb%vpyZA` zH0UfITlZaUor$5TS^WE$3&M`qC!st49ds|><2 zTNG6MRuyn?VO(WYGqYeGYs)v%#*xZt?M~w5t>0rb9q{jF;9e z!dI^*ipS9TW=A#`bRWPmZR`9;RX$&BjpK%xQnI6{gv|wo^vVc~eP>=VMucby)eB1S za5Y)2Uwprk^fn2cGjZ@i&%1TfG7pg>Aro752Hw~+k74eV6&UqWE;YN#*vZYSjc_o8 z5X89->;-OE7^pra9uMn^_XV|_&>w*+1UXgZ^hyrg_>(#eWJ0@s?5Qpx(e5JE(RdO< zJ6Bt-?|MEpL+^I>xs|UjFf&Hj^jy8 zdJIiFeCJYD+-TO8eu+hBRWnuCNU|Ties+ak$=}vKMwMQTU+M@13yYRh;u9#sg(<_M zBMfPhWD70IJ||}fkpzj?(3!DkYltwO*taLMx9-HPhj+w52I^7ltE{RO##!xB1D!nT zYtvE^sjHaRXYkQC%YM7Uxb-xv4x?2I4S-nl3u zmbUR%k`aF$U{dAz9*=K1Hmx%WG3PaKxV0>XZzeQo$`mes8m2Tzz=Nf)2_}eu&o_s{ z!qZ4kUx1<{HTn2nqmXT7uTtb$fQ22uRPY!TC_+JC!<0G#1y$G>m+@y5aqudA-{k&C zg*2-c2PJi;jWcux6CQjti&pxZ2D%Qk|8bVT4WgudLF^cturyzK5 z^Gg0Wa5OJ(nZ>8OmQrPS>66%=d-|WwpyZ}?Cpl8+W(nB^XJ7=IfI=y`<utwlB0 zVVHSHS&a}bN-CbBLz^#f+7BLvb!Z$sW6^5V=qFLAQG6_<^iW;>3Gwka*RJ9AEu*yR zBP>}-Ld;|mT&E*~%~yS9G4pfJL$@mMTNZw5I!!~O0zMf-jJ*%3&YlcNI)^l8cDt2YL0lgJI5ZQ}((##f=bhG$xmU+HAUg9 zuAxp{RZ)2D;p#x|WU#EF*-n*ZIiYicRs^JAn5J@2s z5ia+D2F131OGJyy9|l4Q2955?pwUftx*RD6w`<3wKE3!Vdqbc_lK`nm&?LT^`A>uA zgo)&{ZEL9%#Z1p#hd0s^TG8S9RdFa(2LRi}LAt1dNwMB`bu;K-ZDxO73FCeM%&ZW%$F`e3c zh!83}%iqVas<5Pij-C2nq$0p2+$3opPR#goqlc_AD)`UlzvDKlu&}5w%J3NRG#1V; zrNC)3143B}1|)Rje?xDLsB9!VUA!>vEN-7O1yuxurYS8R9C0#P3E|zBRYl>ffqfX= zwG*%YvXohycM$J%aP@J=aMqx@Qm>XJx%;K5RC|LNk|Ov&RUqPl3bWom1$Mi}tyiAV z=rw5dM1}F8vv|fjK6>T{_UDz-Yj9_-{@Y}_^liM|kH_n0#;gsj+mK6U`(!3h8boSJ z%N2*|otB2ft_8~(N^k_f-I@k^5tMeHY4CmSny?sSzh&^&wqN+ykYCU2;xh^Oktkn@z%joZJ1O3#lZ#GFR4XE9bVWxKK8Yq@{h0mnbT za3)^U6H-mjrFqeH7xlYDO9_TlYL^KCajK20x*oW7*vTEcv+SbVnUlK%!?HM`Ll;t< z2M+T~i;LwSlTPFxlOh)g+L~KPr8~mIPe4|;Hr#Q>Xr}zQhCR6@^y!hoRp*bSWBYL7 z|C=FdrZ_Zy_;3NQKJXPjPvA$g?B@-vUAUE}r`#PTko|sx*|W2G@wHia4HKVkVza4y z`1um9xNIb6PZ~l(+~L=8IJZw<{Pt1aa_xHNWC|-?vRs7?IwA4c@9SVQfjs z=8EPeGc*S&C57Gfdi>dPBra-DYEZDuLn4SAec;qia3^(+F4<6%}+T zK?vS@cRufYwS>E`A5Z5lX*~7G95(LWM|_NvOHUrc)sv1zRho9QrMhDH@SF?y+a-`lmpk39zJ!g+$thl9cicO3>M; zD6tW{gMfmf1X6^%Bwo#a&o5Ybu}L#1-0MzW6;AplUw6QIMjMx+d5TK+<&=PyAbIVEE_XLcBa(gN4D#lgBnFJ%d?ST+6sm-(YuT zS-rst@F^%7oZGW6!_zx)W>z1ZnvF9u2l0K*+FF}2fL+nJuKU;~9~Z!s24C!bimjC^ zP=tz^!X-{6~H)>Bd4yqXU}2yPg7EUpH1C~S(#*o>|`HTn!L>^`uzdfPwx zPv)v_qe-y2P!RM_>cH#AT|?LS=9PY4So95BD_3AhKYq)DF1-|b_cA5-xsdgCRZguH z#PvpviGtO@ZYGpuG81kl7C{UKPGKU1z$7q{7E;3n#1e33)gYyX5aAt%n}U=QtNIhH z{bR8T`d}4KLi!Hu7*w>*(8%%X?iCZuD#7smQn#IO(7DY=s7f-5h(i(a2%MxRJ%H0* zSH84E{OOGEG?y_QK4oIo49@AgfJvRc;jAt{(<|x9gGTXWI6H7%m%B-}rJ@NN3L4dh z3k!+Q@?&A)v;3?pU(dp##Z9?P&7#cX4XZ*}!2>0?=Rru5fAZ~%Ue>0$A-6%myyMHTE=(^xC5BYnPA_ zHDia%&M6li$IMR`OwabKsP-+w zvfRWvN1`g>Rm>D&BhFse3(loA?_t?HVIU9md!BN=lzi`gp5FWddV|n5pXF!fzIm)F z&EuECwP=dUq;|(~QK#|Po7o$$J_$}oz}yxgK`Aq_?p(Lz6$Wg;^<|>~OAYqVYw#mc z6hYhMgO@K|Hg^+m{pWjp^^~JEOSbdGA3q|oP2|i`Q#VOS{S~{fv|`P;)e2*Pn;tq3 zfEk}IMhLLkZCrcrnH+aYU7V}SWy7qWITX}hBw(q5N4;eaEYqaG5aF?gx@port@-5z z6(M-y_*41AF{9W}ypMKq@$^fL9Hus9?+TvWFs;^1@yq_rEG^!_J0owTXEW@^y485t zmtpC(!njl1B({`)fYm^7s|ajL3{n~(1gWkQXdizrLZ~FgUPNit2h<7OK#E>eQ za0KIW1wtf{l=^0q$6X&OO{jXIP7v4PMOHo2s4y17i7WMMeC2<^@-Ib73mN<`#0P~$ zwGE80y^T0$%bErvk5iCjPocygUJ(!i_utBrA`Dqn8-x7Go}XD?zJn(R-Ev6B+*nk| zWAo?WC0sFaY8rnWUDwtrInEVUhXTNKfZZ=bQmGXoh>vkHV!*+(8NYaaPKY1|F#INK z7H{XSht6c(FPreyi`-CDg-*RY(4ohn`AwQlH0k5e(QGQW{p&)myZ>w|OR7mpPe9c| zZGY3hTZ^LC^Zg5nr3nxMRTbz?0SH`98~uA`vNR_zq*k%ht}$suZxWlS71t-wg`siH z!IT!yZkirqFO~#~jlZ~gCjUFO>0Kv!#vjYtvbk6dGSyP-?TFFp#?UAES_<3?DEDti z*873lg+^w=wK&vR5*i**#dEuR^+KNI+4N(CQ_om}W86-r1B@y@^j> zha%!hjJt-^gu82#K8^%gv7NCgHoR8oxf+%eQxcL4;9q03v&Zns_bVta zcOxxH_fBp2*R7{F$-!N`KiGw&zP2L-noT8ZKxeLe;C$YF;4?J4hG|%I=$^sD-~6Rn z=X$tAqbUv(r^}ABL%4x|_s&9o{B|`^XX}y@6kVb?1(PNY#&8P!ev^|%_u{xAU0A&} zmw$gen=Ls-#5(OZp;r<+H?m@8+M|GWn5RKIP z5D6e0AS1M5QnQ+L9fWXVcz+0A4q%lA3&86XmHve)_cs{z=H!+vr>yE#3^SjMlxLcB zyhA||3U~Ing{LCg;;9B*sf z$<@}`Jo8ArErr{AyvVB3pV{NiCe5D7u+)hp+QMUC_f~#`z=y5DX-cV6;$2OeD-b?b zRPiiDl`n>L*f06E8dgA)WXSw1dPC9}c|}dh*QR zC;8vj*U7HfhKY%V0rk?YYfywjTW9lYf2`eCfZ5RYCcvR->?|*1NCaV=mJrKFcVEix z!cxkrJoISah8Twpziwb!l6Gm$uMRlzq7h8{U}4yDRS~rjIkAZ`3?3KW^YPcBHEYJL zHkzU-fbLM_lJ%g4l1GJn_@KT0sn8Sdwa=pO@w@Ilz|TI8_^G zbsUOAv4t(ap?l=ar~~gWmXT>^a#5#y7@u)@L_rKR+I9#8Qr3hR85CE(64vSn1O^&5 ztvjiS{~|ddbUJC7C3q`eqPpY`{3X|btfn@NS-O#CE_&5Vjh=&uy$<641R_|+S?vUg z|3x^OZqwBqkwVrKRL2;=qO$q}B;KGh6rtphu%jD0n|!=OMYnV|Rcyx$YQqlIpw6>V zGirrITP(*UH$UyadrB%wgh&s+X_A?6V4>Q%O)`VJv`5zs{`%5Xj=%i{#@+D(mp%Le zoAxxXukgzGTd>)~^*bp8&b?NF6is^_1%A`Q_?Tf`usa%2?v=o;{?*qjc=x05r}e?u zTUs8mD2*v4i+6{&@;MlP>T?>CGl$?%G@Oc!IL*mjy(e-~M$m?_ZUsMMjVoYJ*B zZDI~?W{^^H^HZNQd)0Py)55SMoA&JIh9^Izs4Q|l=SURhm7+(8Y%)!gl6~P_t$%fz z8rBvKZEHTKojHWBzgR{|aaFDMD+$C+*364^!{pPimvO}K3X+%>yVVFBZ5~b^K5bK(RrXKlHrt!A;u-~_)NRsV3 zC$t&G6nn2d`JV^nDZsKwKiH~o3#6(Jr zi{rayF>lANkizmVo5qbJBe!L03NF>&tVWKjN|*3`{^$7X3ClHQ%TR^Jg% z5g9K~=vl}<_b=F$L^{TwNsO&7HdqLSHVJ=aU-=^;rV1gPv`zeb*gmB+F#XG*cJeQR zWjSPcRD-?c>Q>Yp-^BDxLHB$HfX#Iw_SnlKMs}no|BL;_*Mjt-1|zQ}foZdWNq+UW zL(8H=(%(sniQLuuaOs=SgILXl&+iMV`JpK)XS6>lL=bPT$mfam-?6fIHx^(a zsL)N`-M)xicNzaV_S}fa7~Qcmv#z)nkFH}^)h4I+158_-9T-*phoT6CxNX?Aud&U- z7&)Oo?Oyts+?|E^eRV0;y4S~(x4g|?p1YD>L$aE@zav2-4dGh_ziBfCKvHrn_x|&2 zF2CUfq)Wxp6lAc(pbW-k#l+Z&jSYDAH}2ZU=d)HbchxqGM*4)+@b5DxGNx;1?5awv z-A0Vf#-GNG;lv(Y54c@R@?*i25J3!}OPv*^i|{p}Q?w8YZDYnTAnot;O1U*e5JMn~ z&SvXJT+9`yYA1wn;j|5;Q`&pPIwM7yC^WRwTEX`7vfh6n&|&Z!<0DJ8zrTkuu< z1Ixb%%fAR;#Xs;=7Gx1A1ZmHm{Ej(@EGj_1E_)f~) zGg@@KH7G*i{=R=C!`T*xV#lRAaj8yj?RGVpF?IDKc2w@?ss(SbvSc@ifH^}|C5mF< zH4SFwZDgP4z|(z~&DJbI46whn5mST^B*q@N15SN3o63Jby^Zs47>B9|uq@CGq}NaG zjza!^(F>Fo9%XYy)J$>Yx$25BbjVEMiKo89FsxcZ>~LsYf7S6=mc`xAO=ZQVy;xGB zsR}U;JFox!A~M@1Mcikc!_F%goJCPpHAQYWT~d;9Xor8vDs&H;P>9hEexq0DWf7MP z1a7k$huSo4UpK2+SM*QzyJsV@NU{&4Z^|FpS#lS;S&TqnNk8fF*U~BZ?~R%cgix@# z&OrA}19U7TYA{o;4h_XN5Va9A$Dv~BD=>ZY2#LUuZcP7NEPW+vgg$r<`(WCpy@hF2 zu`zcb5(6ucEFZF-lpb=+o**galolOx4eea%JTv$~ayr!@r5c8hG8<} z)5TnL`;ocE2Sdo#9Zm7O5yk)~pV*f+Y4N=B_8fNZDxyuBc&@o@3@07mn-70n&hkxr z@dX_We%-|5Gq~sFX?*a&Z9idECF|D|S{D0x&$aD~}95?|^;q zy2kZnZQ1g$l9qPGPNJ=8b!9E7GdJfn{8n9a+EV{Y*6+IjEfAPnyL^Auhr~GslNuYI z(p`fq@ewRz8>X=x378nD$Ut1NG*Rrw;!JLS8~cM{`W6wf@VxY4`WHnM#&tQu(8(lF|GIN*%eTEcf80PmpR*<;8A<>dsR@kg-=*36 z+n-w!F@j0=8RYLQim=76NHhcQ*JIFtj=cPI`1JUjdFyKhG1S-&B|CD9C@ig_P4bb= zk1z36aoys#s5a_^$&%t7T({^=zB=KKM$0Fj(f&dt8%xLs-0V0h*r zts+$W2o&3p!z!62l*B-?NTV=3;XsJyA51ni14YDO(X?wvqF_lKMYLTrAzdst`U53y=ToqzbCFWJA`i)8V}Jxm?bav-{tlH7_)gn+ELxHes=pX^(K z(MVk0loo}aN>-HY=7f~+e&`HmCcp3f0AC*Xgq<~8Nw%kOTE^M*itFF%V;9zZAF@o; zdv-%{`!=wv@a({Pe$C&`X`f)%8czHPH2W}UFd7xw!5hug*W)R<1TaB*(S;4wJ_bkP z3x}M7thuJ(t9mfZHv-KyrP*zTP{>SpkleE0A}lwS6hZS6TzJJ8y7tRv@Yr6gUbv-RLb*5?ux?WB934lNqSQ%CmZ?K#WwSb@rn4c347-S0?F zjHaa4O?;Gt(LHm}RD~tGbD6nzBc>%8&@qFTS1hNnrUoGdS#h!ad)gFwr5+58s`PpJ zYR^_Gd>+Ph$e~Zt!Ha))bunHmObn*9$a9y|Tzp2FvjbOlxw+NntuaxyPQi(8fZ4>{ ztw2Me78ytwk$`1-$Sb*-F6lE{bv*w>+YNpViu#2HruRpT+V`flv}~CcciR z>?WwssHwGbOe?(D0k04EMqLC@EpSQ1JQ;e)H@nXmK_K z0-L5VdT?jH*|GzVt|Ng@*KR^863Ye%*jZA>lsDey>JdY@b>t|vloxZ=H#7XH*p`JN z1XH{8;n6W?_=~D0cH+zZYcU!*ExndOqP_V#W`|4H_$%2}`8vVxXo4DPCr*ffkdTQY zTp)GqA%0mCk8d4@S=_4BVQFkbTH#)|-5BmeH^u;}V?sH_T`i6?=Lg{{k`D=4j^s@6m2j3hLz>GSII z8a(sQIR_U6j_dL0Jp9KG=$e(l>L0d}-X)a_ZkoW5vB$Q&QjY&A6d>T z*u=Q>w#{0%1&5Zx@bv$%qTqI9{fJVbTjHNdj~YQi)m!+?d=i}FsHl1h!>SFl7g0Z2 zZkc9=A7KP?!mcJwbnlN?<~}srz#tViAyicRp}+3NJIA017p4sEry?}cV=ii9tJbli zY*|pHkAwSGSYNT0F)1yldkZ0mvmeZy>=ctgu_rY1QUXbigX^Wm#DqNT>*$)){KQ{L zMGg1-=LCZZ`8E!+||oDjog{kQAcq$-%Y$=>Y+Y+SLM8}6OX znZIgX^2kx89UJ3z8)sZo=|%~cG3Rs;PL9GfEQ$_P@X?H=+;PjBK>|XR)oy0ZThET2 zMOa4Y#af@W7r@e>zRX1r5lGdiy#C;mbzeB1Hh!s!@CQuI{fqoyMWZKqAk zL{pod1Zs%TdLl$PNl_Q7tK~}AmYGj^;mK5$T~AHLU#TcQm&%e~wc%mVJc_WUTf2G>`T0e>jwMFR)mY!rU!)HH4zGpgBR-0|e6tlP04pKem?(W$QWa{p^z zvvF5w!Q9+^CA|95ys&mn)Ox{^nEJtouGj73<%eb-OawWO)P=XT^f?~1+mUhi0Ug5& z{Al4~00Kyf?lak+TgH+_+c|M;PZliO#(huEMiBxjC8}ZKa0$GPdTs`;X)92NQZyLA zNtOCD6%rXEz-yYQ;YJBiKT=6|#s%ruHl$N@a#KAzY&Mm@&+3xrF$0r7wMGeR%ibZz z)`fi)PvSKTF=Z`^aI(GZ-wer%&=;i^WtVQ4#l+R3hI!w{iay-6Os7eA3+u-`| zg-~!L{s&L#6@ihk2SgN#ZBVOR6jf!nW0|=$tSq8?m*B4c4^gpq9Fg-9W$(ez%z4!6 zo3W&uXnTJY(R^m)nzA{}&HoPy3KkY$ANV&(j-HH7y@Z~zCm@7ibn=PJEBqF{k=}>I zVqn7HLu%`|%&t5#>MR~uF_&_m8%?NO-fJko8$3Q_-9JqjM~vOh8!K1g)(x`bI`VcF(xp%9 z9_aszv|D3rciOn+p$mBCp7+tc29_mpxNNv$9n_)(m13$VFp_q=eQepdpZRF5Dz&4nQ(9WxTxP*en#=@)DYC>sWu9kWjFB6tK1{$)0)A_+CLQJRf&yZNIY zk{mIdoYtF#gJw60tSfz`-VoPeN-xWc|3#cAKn8SvE$P9+!-kRv z7?t@Am9<|`>s^9f>qJ8IMIjmvrB(l-wBnx#VMk)2DXCQyZ(_$rQ%c{cgErpQt>A_p`?h!M_;CBZ^+nXx;>K>b;!+Q-ULe4|x_uTjf zd3#Ih4YCBJO|73qa5`w%mLWXMor+prE~p}1eQsZW@Ai;9Hw+W=yQ8u@OeDAZYFaMti!Y{EF>OF z=Y?(G^XCE6$&5-kawiWkq#He~{}()FC9zs)f2(~vF|1k=qR%5C`n<5))_N9FTKO0f zZ@ue4H{IkFUq(*)f`cPuB>`LXm2GWx)8CkR?A_m&_}JTsi@EWzVwT&==E{Z0Cj7Bd zy4h6m6M0Ezkrma0$r-0}LefaqSFXb*GzKONIV?k5hu@bRKI77B7|hwWjmq%m=(;Y4O(dG)@aQqH6L&ryoFVtbo7XevlNBuZYCW4bA3zr% zRRt-)vME$6k|&;@izCsA*9^V6C7|mTmLl1*r;t(oJ9EV)r}D!0OW9ml6!P#?6+y?; zcy`v7*NfLHn!xTof|3ldN?Vw3!^eR1{n;eYU7N|!OYHE z>#s8w_-aYX4le)c1->}t4qQ$6m0P0BSIEW+zs_>s_>m;q)6j$qyAp%f3XLGqnlWS| zOj-MelvlqNM*rBtFc091015U)#ICiG>|XUl^o_)Np($-CYmy;BN(hP6Gb?}+hp9~lcFP6wXCuMqRHl=9&w z);kaQ5I8|fx%`N=vJnoajp>(+V)`Yc$lX`UKmPV9J9iY~(WoCC8-|6ixE3q^;E1^) zQuF^VKFi{#^}G1{^=Hr}J&B8+dp9K7LQ@o8y!0HBVq6qdRg)PPM}^ma;de<)tPdwj zkBa4n{v(@g_llA|+`9B_Ost?LSyJLQe0+0Y1JkmHwE2AeR5XjP^53k#&c4q7FQ;W( z#)+v@7?gZHD~g{9a#g6pLATg3l-GWVDZS`Y#R_PyB|CazQ=8X5FdJp_)ge$Z!x;Xy z5|j88mBr`#i>Bg1N`Hciw|8F)z$UYK5BIxk74lVxpR5t+;c+KU%#h-t$D@uh8V!K3pATTMszfG zRb}j;u3T}(3B<<)kLXVv+k=$USn?{|Vb6x{Gg$n?c4Qd&S)a$pN6*dW-j}a$-N`t@ z2-EN}+U7q~ui?7eCgE%pP_9T& z+*ZBc3Ibtl8W@`YUrPm_fNTCWgOr$Pt~hZBPMd}z1d0$iZ5lU@A4yJf08eTaB{|j3{Aa{7)M{+(t(bOiHd|OYzhDSOT3)`-yPW+s-(pEGrgW2A^(FK3uJ%{14p5~R@!HPoSyJ*o1)eR`h3rY` z|5y^LOb2G5gi@|Mm-qI*aLE0ZQgX$Dx0tzi9UjxbkQNK`cX0lsPeO85AwA?rlh27`MG4JXSi>1Ac%UpXV^t0pJg;-fF4qUIe8BNvGq z5D4KUHU4iXO7IQr)pwF#{%2}DKcWZ+@zEEO9{)!arFkE!q9zj+`)kVGvoL%s4)#kz zja2+C4!BCHJoDI=Kb1b23s75;iMPMzJ9M)$Xwucfz-<&zP`ikXXul5g4@Ga_F{|p& zuMieU&=eBQ^HbNP7h3Tg~5U8B?4QCW=7YS=D` z+tm4F&sr|;e$d~#skn%@*RP@2T@%*e`r8}cV0KAwTF2wZH9L9#+f`g}%HSqTlf}e3 zak^~OR)?<;)6^JGvug}K={P2T`~pl(U2oO&Vh%#0|g= z0G&FutpaXKbm^Rm%W1ffZ&o^ev(gW# z9Z#?RiD%a=Mz>6)l)S!iC0F+w#y!Kw(J4BKb>+4GL&^rv0tpHBDArVLVp-`He3pSP zFp>4~zJ+ADk~u5Of66Usv94kxWxg`H#dP&gB!;`dTJeNGP&+Sd6o_tBQ)A>2V+*gR zpeWIFNPds}iia@tH8`}cB**;*mm^p;%`i*ZRWJohRss@B>XcXi7oWbBoV53wZ(M-H zxIa)>av3tu};ppVM2@N<*3q*C-4qC=|lL3Pmcq4 zwY093@YSy-~H`?#KpzIUP1GJ7**_9=HmZ!&YysM8lP(f0p@B z*I#@f4vtqrIOJFyLtJ{(F@p;(eG(LKT@ zcSs=F-f~KHiMN`kx6H+3>h-djx-=2OLQ#UBqAt}&pLqYDuWq#57Ov${7!HE;=x9t{x39E`82xwFIk~(=K9TvC}a z<)>4)zk=mk^FrJicALuf++xz2xE~S{qj~VjOStv2m;BCi(?ocU`p1*QW#g2y26Nlr z&OW9gh%Isi+!PJw(I(^f6f6@$%#>1YY_wjR=FwVXjF$mDgb-cu8{vzAOY6Rr{?fqp z>M=rydBFc2OG6A$R#t=C9ez|B7v<#2i~V`=-!0opZfmLqbc?VM6QNhK1J*bFCog7?Z@FiW&@Qa$=f4nj+TbpD>NF+r#^a1$0b^ z=lT;`9DoJUIWw8n+xLfL-|41Bhveo*u^+v;2vau@rt~+!rXfueRa5!vvsW-^WOz5p zapTBM&K;F`v=)uAl+qAF+zNb!f70It`jw`FE|Gvi7J@cakZ#NwoqO=mvUwrPRiSWZm*9zu##MfW5Zu}8T>iHHZG4u2 zM4}1>r()-Z?q{M~y8prs(?{LXNpWN{F3n%8@{7FZLeg=;KVsfiy^=u*IBf%PwlNmjXO^<2;CgEG9S8(2z#&vX65~9H?8FDql!R9AFG|bc#12mf z5;;UzqalRCsU2=-bjq(N^6n(gmPU%BbN%<}&MdC(a%Z#0snlz^Zt2q$dMYra83@sH zV+jRQ9z2J2*cFXOhn+{?ge;_#A?@myl17RvihX)T(52wgG%g=7xYhly-r!=}`HlL=yVf7v?vq38qfmAf?(GKF1W7?M?4$y`Pyc8is!vW=sQpy(t ze;=`Vw1%|G@<=IXODTIwDZ5E29}Fz}OE^O;r5t~(N5udxmz`6_^~P>*B$DE^amhvg z^Yp{&olVXVs}>CEmPu?>>kBY8COs;af1G$K4pqaY*>I^E4prs8krU|@o6zF+C#Lq} zsbRn7q}2X&i|N3W%#pk@@=jtk7c=t~Q{;A1=5^vV4;pft&={2*oYmw{!wrwp=qK+Q zHw9HVF{Of6+AvX2nixKmc-E2V9E&%s#)hCm--#s+q%>HQe_On~uHmyTK*yXxzkeGaQ zE9V?Ctq*N0uAAQqd|nK{!%*5XkJgZOpMlpF?!5gpJYIvvOSYqFDyC_1$%P}CJkejp zBb@LM_{SzLr=9z+J^irTa9)r8jLYuA+%`v&URnSDO1*{~ zmOV@629 z^_j(#dsYScw*CEalUwx$9byl!GRM(mLE#4UCeEv(K%6SLyYCe2ij5H|T}fz#wsp6} zB!0Z&dX^r@quk?WSVk6U(TA@4a6+F>eD?6QlvdW_a@dG+wk~Pq;%iT4kU${ zlc>VZ#H|1DQ^6x_shkB!d{V`Z8I&cc3Mbv;F9smm)`d#nhOq6`lsF8tEbMopL5Ovg zZxhQ7T*lb!ud%5K&7K3pszL}gtQ$%{0vUgev82VGnr-ZJ?<7`BpkMq@?8>3NF-5mf z0X)eO%~@H4+x$weDhi{sS}x_(T7(cJ$G6RZ4nCm^x8FOJ=l}U7Uax^=S&W(3hu{9W zojx3ooX_~E%%lCHjBy=_iP0n`hF(}MhYexLkRL=*1iAU;{%2Vey!gXXo?Ea4cOX@I zQ+^>&E%^zlP+#U|O@2P#?%2W8Q=6~H))*-zD@ykA#lCd_OzAj)Vac5kLa?lKbJ&rU zQ*n^)%nC8YbxcVA6@wB^VQbZLY>J&eaifXW!Y4!sAxLu#Vn@}t7}AYT+E4(EgV7nC zV$PsX!qotDOT2|u1$RGs&EEHaDYDoBE@w|v)i_njg>2Td(Tt&EKlGV3JMNx>i!41M7KR*;rvcx zd3584c&y;Mb|7#ncK*=s+;-_(ZH#GJY}lBG&u7rTUv_=A>%_ALG5+)c6ctnu8|NZA zCcKbQdmx}fe%KtXd9cg zuzO{FAz!rB*UN=?}TnrNnD!3~=}(2^?AmS#ejA zSNVE<#6w+1y4B!zu%w3?eOHTZ6GEkT$^&dDx(iGC01E{Nni@yvgd1DjzP0Rk@8-?j zk7G1=kXX{f!p%!NALai3FQB$j3uAKnU^Z9m2YoVoE=*?mB^Pv8}&O z8dFkR%dE{?DR;X$F{cy5vNGFj*o(`nF)i8bTjAy{2l(T?AK(j2@>!PTcXyq^DdYX2 zgsQ5Lk=b-sU%TQ6W#CCEX9KevE&EFo>yB^DK7TgG_~TdR^XB7o@CM}jRxaGi#^t+t z{KH#YC0K1)60b+cG#Z$L20u((bkk4tb$R*NRW-;MTaARx0yd0V%iyNZXLA39iA?O) za`@KPiUMZrUV*29x}4k8d4KnE&h0#eQ`1MXtYj0u2BTvYg)Y$@$c_q4E?8LfBd_gx z8;OM}ExyeEnz1RPxvKMJ&EBXsC5kgU{+sXi|Ct)25Q)UD#Bx&ly+MW;U`^?3=s~50 z1QsSg6+X&^(PI%pp=;v*5$71f){=Yin&nthB1Q1nL=`R)9D`asM*yi&lNi?FefCwn zfydZKqU%JmV=oPHaTu1Dm8BoDvEnNL`ovFWSki@Pt#C8UFPar(h{5k#krGVHV0-l@ zdc_WCV(ah-A-JvAG%oEtjtv$2NOr{1FRs&JkA@fR*~<+de}NIs4k@`0JyeW0m^QYAqT% zAf0xt$x$&tg3B2a+u86(mEb><0b7$_UCnK?XE9^_x)$3uKYts#6*{Y+TNeMfdD9cp>uq$Y2g0Z%KFi0C#f#WnwY9~zBssb>IqPZM zGKXS4jy#{8|L%Q=Z}VRbN@mHe`6*;F&EMu)s`ULpzy}$13T|Zt`DQ8}X+t(BlOoWt zsfnaVpV4Bw>WH=Xrf=#$7~0`&x+L5jVu+=be7x`XEG_vj75aWE^!+R=d6$p&-qoZQ zaYKr|O=lf~#IV8(;55hfsscXUx00=uMT|}DOaJ(s!#2daX>#jlvvBJM9>YN5uMGX; zo;`fAX-kXS`*efr9(kXw`wGzw6Q6FfJHLz@|2c!on&2dz#Y=Vshpz=NO^Z2Stv#%M z92Z(2O>xZmS&WQ}4wQ$@Y?D&5bzN?=+iu>T&)=T-0&fEmA;ko{7i@`kbi?G@vqv$! zZ_`>C{WH@^jfrMoSw+23Rt&J}<_7B?X$|f%w_)(`!iAjIw_md(6^E*!2?eu}v5-=- zx26OX6Q553Uo=sg!rgtRGQDGXf$nt`8&HLU-sn_$Egy?ZmeMD#WuNCq`yQlJFZOFP z1V-a4N@tPj>Peq?zh;A7iKT{6!zD4Xs|O=T@@jsh$g>Vp>XgZ7ViXT9bOd12x-l^A zF;wNq`55=ruAl|C<3J}-k)vbeI>a5_YVCR&ua-MqDK zHH-81(K$YW5gFN-VIJUa-Qc738=2m(`RXML*6hUN)k7vkb<^UjrJK3nl)?4OYHB=T zlS~Z5qN1{8Rl#E}CcXRVvU*=vLu^vv>a#{Ld30~S z_+cZr{qrN1ZwfEND1_jZ%g?1taw0C9MvTKwl+Df!Lx+>-bT&CN;5H13Yr|LCsN?jk zeqoneSV+;Z-GW*}^6}nPP3}t(;Ts|lC@qvb0x0(tP~_bkB95a=FH1`g&Pnu5ya-JM z8&L&~bmw544MwZEH48E7C*1_qmWv`Yjshc&UG6eI-FF8ydQpq*YAyNh%@|>lU3BRt z-?O>J?Z>8^fm3Z1S|IL#HyeS@cP@5;~WM<5ac#xwMWEA&=T;;qK1+Csvn2FCZt3L{HpQ*4Y$ zZoWD&fY+))5o*|r(PLJlTRt{cETPm_OPu2*$~^PY6dRT_Nwjz2aCEy(>aaqiMmGl5!*h%LsmO>XwONncTl(F0YKgus&7( zsii+r?)6{<45BF|rfH(6CjQ6l!F{-FHZB;@a%N-yE*WT=f^G(W7b($dL9t-(zEbk? z%J}%})ucr^DLPPzBK?zebrsj0P8+@ZWHNHZ!8wtQD|hqU{hzU8<9=ddot%C3NnCW( zgb=-$<4t?^8UHMdF~G0yI~$u_<;(Y$AT3FBoRb^woyO3U4kis)mgLUgzRljfCFp@t z;byrT9HC`|JYF52O$3<vlvpcWA zWw+5aE{R=b1dXzy3J1mQV45{kyOX8wPc=)pa+Sh24Izg$jZSUkzt{|4%F_ z=^T?x_t-T0#dl~QyiU8!jxQ6g^f;8 zgZS6xoA6m4bjycLv6E;Y%+$1VNU&#<=m_0^_xLGnsrV2xOvavH2Gd|)&C(Xz)mpm8 zp2UKJ=L5+u2k$yv8oagVzewJrdu%^0>3Aavc7Lj`&(f*T%ZazeV^{1X*i*Q;)6Iv~ zj;A+$6LxTASph++CJcfEemJm;p07Mkk~4}shK=F7-Mi`yaYHmk#b#3x2z1lLrYKD4 z)`c^Bx12;Xplb&@rX{d*e+h3r z%&O2hEAuu!%XPRd!;UGp!U{B31YK+Qy7lzOBQ0(2$rm9_>nK`b0ANlUxY^+>{ zCN%mdoXCjO-?OynQH%!B7hY+j?#bn9ROaY#uqiIi&v}9`_dh_n&p(&owV==!6R;&F zTdUS_|Hj|aHF`KUHJW9mE9&L`Pfef1dD-W*%B9mBduxlse<29SDm6{rTXKn~mIoGp zkK${(Epx89oF8`Yqr&51bf=CC?9g(UTve@y8y=ZKZecl=<&XMSn@_w%G!}qjP*#-o=3**>m8yM*c}=P$uS{@ z7-0AAVvI%^c0z(ysna<@`X>kvR8;f8ck?*;g_oGIw&|#f8l!3bWS8n-WO7$>qLb+ylR{mI zKc()93N*`O=~U`&{<+cb*NwBqGB)`{qSbU(m+hgz6Fws@9KGXDPmGdKTTp4xI7Wrm##y3Quf-Va?m zaZ5W^;Ih+%lYvPWx4LcNFr~qc>dow|-hwGjk{vl**5xIx>iQM~6K|qWPs1l0j`k&{ z)LB`+k{^o~;j?^rEgwG1$JYgOnSEf+VYLS!J1YFOr3i^#RXOO|Z=5a;Qa^ZIdrX5` z%NH`5vndKEcI-rUd;%8_8o-SsPiU1P26*EA`RqPWirZ&kS%Ja?&G?yx8Tz%)XYkIu zi|YTsYx4nQ7^wlxrc#jG`fKl~;}3Li-RC&)vl$t;hV1MlG))Z}8G}Fb?xM4Xan_W< z2baY;`15nK@icg%+clNZ13Qz@pa^4TVt5%ryQY$plE|0)wg-RFph!cl?jL#o?c47e z-Ki5D6GBIU$&M(J9nl+-~FtxeQ8XZvomLA9u)KOm`zVY*4=`+wcA)+GP_2yPex?_Bnk1zA*FeUTMLv9nmHQi3)j-?-ANx!p6 zQ%w{_VwwUWOyX@$bk(F*H?b_R3B@1YXbR994Kh@~N6@gTYhn_QPi}okU_-uMy5(SX zj{t@Sn$gJU2C{U75o%;DOR{}?Vg3I*_v%32p5hRLtouyTvJwudoks&fWMq8I7&g2M zDJd~QlLm?+a5`*Em=GM6G;U~jZanuy?3#)u1&S%jN{Qp{D<%f-vvU?{@iAzP)ZR2j z;ewF^7?jyyVyB6-X;~Jtw}g-X6hiRe31{L`ZPaP4*@QxjX6JW<4wk4Hmed)$()_bi zQ&ddLtb2QDjARqj5zwRcH(hguAUP|0yJ4 z0{rR3o2ypfP-B_a>1i(L{2Hfc{e#Q9eas0d*CK?#lt$Rl#pbBeD|mV5zi^v1cr7#q&EJq1Nzc2JWjZqS|!elPW%`Qe)b2o8i;P>!>zr+T51lc6oSs?f00Lf|LrD zRPY%ZUR}ksR7@-uPQ8ssMoi=H!%yS&Q!eJkaTk!SB_q9|l9~X*6bK!J;s11wNhHw~ z{Oa%34R#lmQSI@DY~PS@Vp&oJY4;1tyPP&$P8$M&<}t$#niWORt?R+$kV|f!5Tcgo zaN0QS{1b?cZ&~>0n9`1;DSlDLSXC9C`u7!#J*g*}rl2T-p~E`!>~mM)iVB@Y*Mx#< z;SZaWkY8BI%@59?)^pHxpeTYjZoGgI-8y1Z6>O?P*VH6lx#4`$;$ryYq;W*qY^cq| z6q!=uHJWzJjqA{jm&aa6|F|qNv;?Mh9Kh_;Zy+b;V2bJaIU|YI?5F{O%f|M?uBe>R zVF>FgcH^^jOerv>g2eyvSQZa${NRuc%>^BA$EC(%55xuA6(=e7OeSSq0uW-T>u6jQ zGv1y>hp0g~)L6DwZQy~8zvGToSMh&quI1C*cQ9K}j;mu;`JyHgIV5Y!mr&p-U{(2g z4#pk&KR&6V%cd{;4$cmTm6Y+`&XtsSYX~-~0x1=w6bwwvB1*G!M#nx}+I<)&cj(Hv z%x?T?s{GGd~@g2NIUzxv*u@5}Y1<(AKlOBA5zkKvfSV1g+ zv4gsxHX2wcs=|n2U6}FIwcLIEWFEX@DsMh=IVTM2g3#(J0!T0|gl?irm^OV#{q|k^ zWb(jk*U_zSCPF|=yo(EN7|%_AnSRLU@n|8WY}MyDwe9#Ec=+()a{2P3CwnZV-yhbL zesN8$mosmE9PsH(oG?OV13gm{9HU;xkN=VuGLcJ4;G=;m4({$lK4}nu?Gh zd8ei^)xQ+Flao=5(rLBA40!`942K9BF@&fQH#&=vZ3N9N_~Z7 zy1LRUcBnrT4%qMB$&1_n7$R!#4{hs=O?5IZ^%TaXjzC5QZJ2>PooFM*@KnSryO4+tDCXI+r=_vb5zoS*_ zu6XsX!#-YH`9=AR8zO{gHUNI(&U`dY!Q1Fw_4*7JuGkUQ7{@rn>l<`UPU4Yiel6su zfBcCTmoCR=7$5`=RpsJA1DiC&)jlt`eE)fnHU+@T>z8p#c2|a_9~45$a3%2fL4F^y zl#=xo``A-mM3Fm&C#9Xkg2IoHNCXOo)TuI@)L8yRh7WSzV{_#uuIPLfU*&%m;?~fmj}>K$ zI4Ap3qP6DbOnb)m=exo$@L9pH<*HEW5!;VlHMvbxMiJPAicL{*sZk`@(kSv4lj-Wf zIa$-`8QY`P=OY{)T=6J~`XZxsXhPwcQ5Q97hyf<&bZ2tT!9eXIhf2 z1;so%cOj~WkO76qF!*@+27Z6m$=Dl(R;DG!^2tNj@XfNVY~5E#x9k*78`0Zeq&Ezz z>Mw9=Qx$vxU$d^8=(@>+_kGOk@7)|^eA~MMU3>N!|6+}COT@)E!}`wzkQjI9ey-cb zoXCW3-S~XdCQK(UoTU*E%`xjB{EoVe>AI7H*ZP8~vG5sP2 zC7i_O%H`~<-N%n5tMCM>T}ufyhQSXdt65Zf7jbGWBFvqk35^n8G0|EqK1)XvYS{L5 zjqXLS*a2**T32t7?TUl`al_~o)r~}38t?3RCuBRDuyIxAE6H?akrUNAorrQ@E${AH z!lI&`WJM)#dDqeOO$h&+UeSF7bMiI@JzP};-C|Q1o8Gm>V>RUNlTYHx4?n`w(8VPK z5e`z|(G9dFCb0TzJ}$X0~jZ{u$}?&uBWf*RU)eS~#Eg)~&-r;xi?} zE|B(v?T~@>4uxO$IEyb2tPIl74M(l%Wk+>BJz|3m%TLAYd1&LC7?z2N#gdYZe7J8O zPYt~tm)gAgS&Ac*v8mH|efO*Nh8RG%lt7{fn9{^2!A@vYh#?IMJrz8*`L7)C6k<~} zCZ&(%vYd0V1s(=cN*0%_VP~~Ug)f^J&7gC1EGMN-U{Hd;#8#{=mK(d>$O}7Ps4q2T zN|P&cuHeL!F|?LqPc;|Je~wDsgU>QpQ@W422R8BG;Bz=NOe%Z-#11?(Y%2GyoQ;LV zuq=ATr}6CQ3)?(u9G;cMqo<$7U%&m1Dz6vav=F90e6kKpfD&f37Z>FqAu2S!x2&?3 zRa&XJ~rj*lYz{Afcz^x6pwW}?DIhGQ&40B-X*Q;IYhLiBDWWtdj}5twd)B;;DJ?vv53~M( z6Y589h+&!hbMrfR8yIqr#}inm%%@RfI=Q~v(4bND z(8QrUIrvG|maoOaVnF-=Vzjo^W_x_&*Od8c>xG;wBpy@e?`uD2LWkaHN@$4Ol#czG zl-Zm2wyk3Efqf*|W7uC+MY^kb*Go9g>eY+Wd-UM_)$91@*9)oj1wUXFhd}cQEKAl8 zDx5Zrd(IjkVnpBkVks|uvM6AH1rD3a(|4aw&rYdL9^d29!@B!~fUJ&*q^31rLAqTb zBW{XzOJ-_3@BQ;Cw(TjRvc^N7t{FtTT37$;P|;13|E^mb{OU@RY$i_4#?-F8 z8PlO@CApRul-PyOPyP#Q%6IWY;RZg-U52-jk2Y3w(Jdy}hq<$6e~5w{_>JPcqSdW7 z#$!`XW={TWj3%NZEEFmYC$XA=T@gg7(M(7`gVkjR&>M~DeU{GRl9gQPI}ZuW%=r!2eOb#mC+bXJcFSjm!j}>J*!+x(@CVOj2=^7hq$fT6q zIe!-4=kBD|&`}V4ynQ{l4jRp^gGaZyEvlk$-oSo5`Rz|Y@VU`!D$iXyjSrTrV_j}O z9h2g@al%MObql_BR&UAUl}{Jr)BTc}0JUz(O@DcZtayJh(rKp-Wa^}WL8IMgpP0k` zFo_!ur;U5>Kd;SwIqpP^ar;FGLC?g=#3^ky_U|3(v2a;_)z)t zs?S+fx(^Eo_m*x*lQ>&=5vU*c@2NM$;2%x9O@pV`{KR=Z`qLr0?IUD|rt#(#=kx3L zKBK(WgAf8u5&Y}iGdZz)4kvccX?FXMzgvU1q2E_ZRJX)u>FlcXE3`fH`aG7d-pxaI zpAGmY>c07MUA;h1LtB#$?@1s3^Tzl{@!k3z{QIjP$t$a*TUru#oOu!> zx`!5wjCVSTcRIsLYYy#@$=xT63t4`+Xo||igRf*m#a>pI?IO`0&G__zVT)yUib^BR znMiJJQIJt|Dt4x2wcghX&@;9dj}CmCLT>>}N>}hn-t2n!1Nhw%7j_I5lIa$mLtbq` zi29si8DzLpF(ItJu7W~}BcWC8vZRIE)QM8&^mtc1>s7mIGl8F+g>-C+za}j%cDne0v~bjGLgU#>0->Vp0=h z$Vy9ScHJ*lZsM;W&aD>{TDPx&n_vBuXRe;ks2-iDbbDBrS3pv`BcW0B`6|}Xf=wSF*zjJ zE9_MfOVeZ<>b{I>5z&dV7Zw2j;J z9W0wF{Za+>^#cU-iSNL_hFuV}O|{QQUUdZ>V&aIlIZ%aY;HN)$E~2#NGXPsc2tl6= z|C?NSjhlyNe8+};`D7->A}od6q6)faCUfJNqv+Q;gZV4A^Oxtp3?3B|hzAa{e`!h9 zZpx!)*L3F0THRzjy7xHr+I`239 zUO4fx`gNt0JhF5FuWwk1rYPu^$>f~wf$A;66O>X)CU$ID%Ct57uyGf+yzpuL|8q+M z+9n3%msYZD>pq_P^>m({vCwab15XvDPQtiu#NTMYs;aPQ^*+vjWg+Dy)qy9X40*jj zWpc~9$?cIyXxq<-xr?{(-`UH9COXz_+s}Q^&gSvopBu8iM>i;`4lks$Ex&~A^Gop> zCf-1L>dxX)ZvO9Qd~)+;oSf77sP*G;QH8?L#GaVa;@_L*@%D}-xJ{i5R~&yFa5@t+ zdmokKZm!%%jbQ=7VOc6BTdMXkEU7!G&Lo~3at{aGMbw%ea-uTBMm79<^SA7+E7H+Sz2PVA7wPX&7eBJuv?y)^M0qHTviLA<}Dl3QNupLI;t6>ExTc)!XkydllF~pq+zerCSl`Go;-s|v{8kX>XNHR@NNp!nZc29?o0W zx!DUryRKZWglJLhSc2A&<~Cx*)anPlxy&brdl@1pP*()Jt06LZ0Z3!$FGFF(@kud*(`jFCUT6LYrifn1zhJp7|2$P_Kp+BY<0o? zB1wyHe^q5-lBSJ2R*St>)?{KX-JN_02*G1m!G1;K)Oxz=by5UA9dIsaW2V}`lK zdyQ181XGP8D$rtMRB98nzisB_leJ0?r3c2EDerVh6oZD(konDNjlQ1Jkb}}>8aBc~67?{FvB4wwC=?6)o3#i68VwGXdNFcWE7iU*pvi=}=Edjp5F9r!G^n@_Fjnync6|vx+ir|13M0>CrRRG=V7Y; zZM*^%+;4#xhWTGEHgQA)&n@TSo;5|Enw0=CWyn~Ln7}s0S5jw`y)LqlATNEl%!f*S zhEa6%TEgHNE+d`qKbS%+ri2+n+n-cS+?MlI%U;ZHkvX#OKvdw;2MQtn*zQ5Hn7rwat@dy6rPi+j@ z+<)AFyamYd%Za-X1(~*i>-F^EJU0!#C6t$rT{y*)sx{HmvfB7DtkbM`@TQy@{OJ6` zq2(}+(TYbX%~jzmYf-&HX3~_EF{nd5Lba%(Pd;92D^x|hV$@qRk2Xc{&4}WA2 z1SgcDOqV0N#`L<-JbyIvg)8qh>T~q$H6-PbC^p<|e#*!F&Uhi7+YWzZ_I*grFN4Vs zY#0d&zu!8-3CtM8m<${`)h(Z1dn5Jz^E(=v406Zp{)p+d$0b_xb`SN25t9E^92z0- zn=|i$W4pQuTCll-XgA52&Aq-Xdj0X-yYtl5h-Y%#yirwkjJr zV!Z9spvUotil>-=)5C8sc7Yut#GEfhPK%2=k1v`s#vacDop3ueYWwAg{VY%X!4WIA!Tmw#dEGgcmHLXd8N7iQ9y^qA!{wbHu4@dq0+4l zGEv1!NxoH4+F44PU=U@T0-l{yBZ4jWuK z^)aH3(nuy-z8cp2$^Y_>-TC4*0qLr6lIPp1g86b@>(4k=2< z7d`W0j`cX#T-B!i-_X;~y{DI6Xs0^$h}~HAU};_zA)Rgb3?!@-5)cuAKI|XJ_)?KJ zGjVW+d=9ScF?&a>^TvSo_v)f?wt5h~+K3)ae|7Mc+DLT9%Ek2CR0)my{a-FG@Xk;m z`mcLNn1Za?P&BwGMh6DHPQ5E&bDmbofs8NRo*w4TO|t*lJekwKB73v*>|UsPU;BWe zujT@a@m$i-2odCpYA$CnBp^ul9yXU9i1%hAcN)Ik|+*Ftm8Uo}n5wyV%$C=N%i_1)`9Q;Ts^fwf03RtGY?s22xY*($QTKmN`8(X)Y_k@c3-2o%@!z%uYCJgi@6+)J{)N1j3Qny|6H_ zIN=Y@B@xfo4}^y7N&D9RAuBd_)`!)gbTjiS|LQ~*aWqj3I!4iD*n+04o}iYPo+J5vd}46?`2^$PEwT2^ z0t#p^ospx^w;7y^VKv!skl<_^INu#t=lB+=Oh1rg z(>7Fs`#aQguA|n52#3=ks0=dN2Xn64-LO>tX0p(P)+abXY}@>gd|O26$WtGJNmoQW zl(0D~T>c0f8vlT%L0+hSL02{v24wgP%87MYU8>4$_|dwP1YYp@i2o)WY^saXD2Zu?xp(!*{JTgwirMlAIT-8|^qxrP2>W|X# zf_M>=4P}YP$;1oB{+2+D57N8K=h%{4F&J|N5DY;OCjU&G+u!lKbIs(c?zdO$Tyq<8d-Y0O~1EKsDeXC(iQqI}nX~E!A}HDrR++!1kLj1h*N{XbZ>QS5u+& z1L~Ehaci;*!=>uGH;{K+`Hk8NOMX<=qWf9AVO-`jlIT&a@)6Or3{Q~Qh=sG!gT~jR zC!t}-Ee9{cF~{qXI@(q%F8KHmNP!~rLrVfazTw$WXCfuRq^5t^)&9O_%L!lLM;_wW zUi? z^N06}XKi~!%!GU-&9hw;+^;dhX0mtUpSQ6=ij0=jO%)w8eQrQ^+yY2n3UKu?tWRov znQ_LvSh>y`og>cui=)`Au6fpb!aJq$T3OwDQ}y6yuJ%Mse2SXp-nYY_69%4h!$t3# z4R2wV1DqeE)D?uu)7Gq(jL4ubyvkaRWSPqJ?H^AZTM7q=%LTT0fi`{_txk750ndLX;Uk(7~I3vZ-pf{Q`<1%8qcB`Oj&#mtC5{#H*03pc4h&oKn`s+9tXYr|I z-^dg0)aRu#Hnx>M7p?Q<=Ru%*H93)>YqJnil0k|4S#Rcvb|~ttj@q`Qtr|fQO%!fv zu#Y#Gmp6UBT`<>i>AU^{kf$+n$zIFMqK~1tY!*XozZOkup|*H7EBskY+QY>!X{Aft zR+uxet`vk7c#=bV?suvV^MCNhJWd}NtVHn&+;;))D5*dmL1CCXJtJ#miw-fG^OsfU z@lO^3of+HLWN&WA_qUJ0!^b~Skz9pNVm?%!gE>p91K#^oZIfBT_f8Lg7+t4p zE`Kp=?E}`Bo)TrMuQ^JqzPYb6in-tGDjiPy5%#rIaKwxr_kp8{FPEQwF=cE>et$PW zHINf!OwXTM?5qp#eT6yxcNe<;^5}H)Ml39j7Fj@Z@BUk%x5x0z+h=QS&56VH!u1E% zkj;GgB;gWOBMZm1C`0u>6_=NfPwX^ZX6kxwN_NX$VRWDo`2cUtzb=X+d4lBN1}-In zlC=vyLChp=34Q#caKz0!QZaaMe3tLbD=P1MR=>9H!;>&}6t5dF{UTlr>2&tYcNV}# zvj1!z$UY~$s5D+K`dH3$t@wcgA=mlc!?rHj?pIVA3xToDS6T>Q^4+gu=zN*Jc-(_$ zS-dEByY+kIPJ;MO-0aawKJExmy6qFAr?l0XsJw*-6wqTF6)E%WsDrS_?aOc5#}(p@ zCtjJWaFP5zu+eyluMH1a#5~mQW~%uHJR-XeAMdb;b=IvU2UNB@ZcdY>Nv#bRl5h9V zX$BoE)l>vm+j@WJj=yUz!$(L$eEc;g@3o{px9Hkqhv7C#&+SMO@@ z6z)L4%^RD%%52QeHW@~@YkWwE@V75BGmz&H<1#MXD+UK%80oe%^h{po-t1zg>8i*gIQrgh-psHm1`Z6brGm z662Y!PLM7OM<0Qou8yLxys(gznV6zB*XtR7lAL*dtk}?&bk$usmq3S4o z93Gl`#=eZ_JeX7ETM)%i>Xv&XkJb+3AI!vv?mKr3G^hUx?wMxt@BqUQ&mV7DJyIbb zOI7|3?{B%P6sj8_b{}Vc?*7^^KA}x)G&}-*+11jV;pC4!j~vcFV4zFNC++3lnDu@5 z?-30*Y)|RSuwn2x8@&D6xn!g{kdkQPvQ2Q9v|PtM9uvP)IXZE$uIsv_rx8O-sB*TJ z-JQ*yFy0#4ardHdt-vCbFDuJIA^J19<&XhuRrjNBeSKcC;6Y_i@t?!ufu1Z(s{zdMGYA*C?mro zVla*tQ`O7{X0;GpEcFm0%#i&)ys=Izxqgyn*AQIjV-{MLtCUgKN)1wp@=nZ$&cn2< z&)q84%Ipeh_=73Qgqm{=UNN3271&pb3#4Xfgc_lPhDh(FrJeJ5ZtwJ8HOW|y4$Scz zTs?*VgJ6g0voAHqj~YD(aXNOg6un5HhPnSW(FvalmsGq}k>7*K&~0!;M!Lz{Wg?LM zOz*W^Fp{ZM1?S!G*g?gxhf&Tym^}UNhi${Y8x`A$v*A}0S$j13DcVZZyz?DDavOuW8XIHg?)yaCEX+l#Bhq>9k_9d}fVQ6$eFu8krR4DsY zu7&N%c6>aooj2&`8kN_(1*E_E5paZ~8McS4L+~}!Yy2i*c>~nd%ec7b!V%2$z@Naj z)ER%Tl)MIn7C`qUaxa@7_P2S*a3#7{xV!?_ZK(VXRu~=#sQ1it7EbdvIIC;=a{U?U zqd2Bl@hi2YDhICJeVIpCOJrrxzW{`tzY5U@u*^xb>%;ZiqiKXn=fjNKh4Lij)g=HF zaQkbOIr&U1o$aEpn7eOndp^)Gzxx$1{2Rh$rswRsLE`Pw3C7)=RW>orilLZEADtkT zUx?dcRcZKS(hyC~5VQH~B6aRLg)&1y#*ghF#;R3huphLh)9M*0_eB%o{lKTIa=b0e zwH*!7zE?||EcjvF6i;m(ell*tB$~D5MQrcIaG-?`FMGc#5BKtG+iPN=bqXRS3~GB< z34}Ye!c}Kaf8kzm(YKE4qEqyts zPfLeCNo2o$qX*>%gRAnbSpqvBPT^|_wPGgIgmp2zk%Jaoe3=QOzghohGg|&sxHqVO9(%c8R7)fbDK4rsrd@sui*Z) zr2iHPWGN@HuHTm8Z%%agMi-T2u9862z)LIXs7E*Qsvk90FaFELLTNcq0vLsFq5Eb{GCHpx;sLA-%C1NS$R2JqG_R+nDw z>$L_wztn$NNxAbEA9V1WwGYlXQ|ny=x#L&KK&K#KP(~!!m+PTj`NQbzpk%q%Mf6v| zdkfT<+ftcAQX$c3kDVJ7MNHip1H`!50TPjIdeOxO_T1YBA&z!qLNLlfjl~*B!0s=d zuyB|&<6YmL2rN=!=og)FbiCHNbUz98$Mr`y-3Tp7!)e2K?StCX{p)!G?MClzHHC9u zZ|oZ_hoQ5(a3BAxaS>pyT$j)4dHA-T3aqzZ;e+&QbE*X$=(tA)s%5Y(%ogIt9X$7g zzKY$C<3zN5Aq}8Aud!A-reu@jGnaG(#D#>NBmJnKPytkoz@;u;0j=zL{V}7l9S4GC z^@{HDVW#rA0d~&KhchOEvBn3<1PA)HhJJ;cIC&O^}ZAwyh3a$tI zAcnl>d#L_L=pa^E31S`#x}x^Q{uo-Xqp0xG@U9u7>oR1DYp@qdl-tYrnQ%{piXw7K zKu835rsQ667Zr&3T`9bJe>%ecCzan0V^r{>9B3W}-aR~F!fvMH2~)#!{cM4%9^5Mo zznpI*zP_bb*XQehRjY#HXy}Ttrk+$|%i^K=TUCb4(a85p#QOJ9haE6E#}+jIMLy1B zng*0Yt|Ag_$qBlS%;2Y@#t=9K7A?t6gga6O#irBWMRj|_%0VQlKVm+?#R=k(V^V}(Gz<1!zb4nd{7Op;ICou3$VG&!gK3 z1mgWm(_xGX8T81TFrU)GoZarM*1DcUX_4H)KzkichSp784B}bbQSNnauhJ$nae?E-2Qa-HKkO=Pn+^$vmZs z45_MQS`sF2ggzS#4+Q$qm43}rRzF&x|KXt9n^Zr?|1~PQTuwC}yw~8m;W)87ywH6x z>PhW-}D(V?8DsF*7IcV9+djwk?IYicmFD(?2rF$4(4h0Kp<5|i1r3N^4H13 z_SG@K_fDZF740AwG+KArTLTs#zzIqFHChqoBrwNI@EZbIk1$A zV;&_bYY~9vQ*fm&Q==JO4_I?|c*lO4k(3zjG|#H^tq-DT##azhwOroJ4|NRY@vlGK%Ef5uKuNT*U&FgZES>lAb) z&5_|i0}#ieo+_5o;Dvx(2izSZKzb5*`KYvQl5N0eFO-_YTd*|!(Dj{Xzj_sqFWpQp z5x_5VIdvxKOQ@CP*AukywFCh=-cmz}y+YZ_J)!f71V~a48fBMa?mQ6^485H5V7~+?ZB;(AW%qA4yPV(P0PoCB zYw*9vhWyq(xEXIgc{vbtij#miF%L&hE9HtLe}|6L@WKQ#6~4$5k-|s0`@$7Fm298i zo5aM^MP7Y*KNwCN=@>O)qq?}4LjQR4JE@e*uLF87_35{KBi}L6@+j+C$5k)w2&+x& zU50wbDm-N^)T)e(r~cxKv|>4t+$=`4m_9^#!F`#(%>z~0$N_ycarLoUwbbfgbNbCZ zZ-pbH3e{VR^jZf+D!89Dhv>sw(`IX>Y_6b8P6^U*afr)#LVDi@hq#YV@Awubff``X z$7@+RxWZI0xWn(|If!4ZN;$ImA$zg0odk?J{`Vx&)ynx04|^tOjW7YNBbT%=Okm67 zXW}HtNcs`RQgJ+`?RvR!OFgNp>@95-+bRMe%6)Mj{F*c+AJwMnwmL-;69~!qo(K>D zKSpN8d+S?u2Lu@ASHtUrbc2Ol)IkbC0|2{2x|Mhgr9p9|Hfzi4VvfX>&b8 zSKOg5M@^Ak7mJ#IRTZ1m4e&E2*0cat<@1H$^};h~e_FQN82q$ab8^%cSzTMB8We>N zH`|szh<`U?j^X1AynPwC^`e@XV0S}j;bLb(8MHh+uyp6U*M*O#<@}Hz`Z$C6ETG|H zc9f~wDJtrG}bN&p(MgaBEKrr}^B^LH$m(c=d-OTnqFZ7fPiTrS$L;9-k!22u#aujqHxd=~w;iKFA@M!Oo9s0Z41*z}H327?t zjAO`#%)5DAz2zXe+JX@4Jr#7tGCA81?>Y-+#xLMdg!YG&yfQ>XLE)W$2+;e5B>gNg ztZ-|+Ro91SyVbPtp+Y+3vF89@Z3FJGXiieVjT@Ywkn?sZ7#j8S<@Wf96X91D1vq-r zaQwvar6-E0Ju=?YdHOY7x|hNzPBSXIGZfR_HiBd)my}tayPt(@4wP#YDEZw4a))|~ z9eevvsOgLi^F8BVt~a{df9AaJLmlN+<4yZ7Hxrt0zgu3k26cB`(vJ$90^}YU1-8Ie zm>a-z)%qBmU1EY9;tH5q$mUjV?h6u|=mww-@iIId2ns4=Dxi7z`T@FU>?i`} zwtg4XEbQ1>l~YQ$4ocr>FnZ&P9gFwl9y{3%3heRqis#!m;#OT5?KxD2vuQqqn-ZS1Xr`CKkb`-aeUudvP?B)vqMkG$P1dq+F3;ljewD;{YkpL=7C=XA5^nCV z_FZtZ#Lg}T0pG+>>@n~_ZRdV#4h!GZkS57YO+l+-aeRheG$PWO=MgX`IsDo4fFW=* zV~g>ff%IU^Fusgs_5S3CynHLl$d}~LoawJ>pu*6JP7DS%xi3h5d6yeZK?eD3EDn!V}Fg29vFXCEHv~e09%o!6%Y6a zK1OJ=K5SkcT4jv*UT|!mODJuPdlH9!^7ZO`Q@>xaS8O|EilbEbn_(zKNo@`QyTNA* zs=Gz`Yf=Sx9O*uYuKSLdJf+y}gGJn(aw>LTU;RCs&vcXi_m7Xg%$$S)qkE-4k-8}% zjRKr*iId+@l%Xo26K2jo(lX0lWcUF!!luB(q`OO3?0PeFyotZ>3#~~E9fdq#(~NR1 z9uB;;F4b3u00tg*I$fD7RX#v>szwPT96X3d*X}ZVp>mKxZZxoiE%399g@lJmr@n`e zJ>;ELzm0Cc+PP$ckeLS1DR=(b%n5PT(efj?atly#11D!)WiLCHDSxedg@_eNqaUqCrJc zLE@$vuBIl5c!B<+z2Da#@b)7tj|;Z`o%t`Cp`0KCnRtXHFR+<7qVeN{Fk2$~LYGw$ z`ONx%1<9RGiU0TXr=)pkrX^K&Y;`bvZf9Dz(~a@j^OqBtf5+=;@9~aKOOvR@z=NZG z_tPuv-IVLyAH2Hleu7e{GVvGVjxes!0gR^Wk?!u^v-^A|^jQv};Bv?Gzs;_KK#wDg zDJt)yjhKGm<0xEiNmTwk09nv;!}-m_IvfqO)?fGDqg_TRIlDZNpWzhL^<=`| z?m@VXb-TdFt~6`1v+Vy3MHkUYddVA-5Pb}!GA?5P&kXrU4{W7Kc!a`7ImHuz^0l$;SPj@y)X-`{qOnhuz7edqw&$4RU-zLyX5f|3 zd4t$s5i|CW7v<`7^yc#E_UUAOYYKAbzVpKX8O25M@p$%^P_#JoDhdsgWIFWU?jo+B zl=83ea2N!^HswNQme4K!L--i756)kUW%|OyB~Cy3pOk=fa$#j7zicM*TV{cNcD>Yj-HlT(cHF%pJ-RY*K8lR7-)5iN=+{&3UC8?kWW+Td&}vR2nb&vHGI=e zTo>mRvK=u7<+h?{)Zh68#%Z8Guje^lUcKWrb|3eMZBvJ}`x6 zm|0;>oKHy*SsdY15fSOm)0(b*?j&UB0ejhx?YM+cyyaZ{PD@{xM`&bkTx5q3hNgm2 z%A#q`ukmXEKhw!7V7jW4Cqgv_oitL^P++2MGbGO(aBafs7LPrk&inx|%94AzJELf8 z+5FRZIbfq_Mm zmTIj%ho%zNNJM6+(bROmj_$j+_A`ERz6e$Q#+RW}uQZ%N-IVu?IhZXFh5o)25F!j?6VBBWC}maj%w`t}XEncVYGcO>2x~A!8waiqxBvD;rFYtDx4aD9U)F7JiSt~{%*dHC#r$gF zhHB@6SZv~8(tjK!-hV|#-B|D4O!5t1qV3~EI*Kp00&t3{0)Qcwk2r~#6l`Fej%!k! z>|*+3uIlwjQSw?SE-_O4t8h4od@HSx3B4J9qHW-Fkbb)QQ_;qP*6=x$_)M6vD-b^P zAGdnExbCSPD@OP^giAk^k%KvmcpsiF?K)p$>d&ETs4i@M{5k8(*4)hT;Xl@-FP4}l z!&!}9NM9%g+GW&a#C9eqr<<^)A&>7qz{Zk5GOt#CCyOG4;9eli28RuLiCix9Y)~Uk zqSA!e&HN$G?@Y=_m$R`Lbsn)F*s2p``;7m|wJNlD%8rFmnQ}y4J-n&I1?fEC=uMZ` zp07ox2u`)^L6J4M1yWTnWx$6Q@C`aPn_vIGR>gettQAJG3f8DJEJ0>@En5mw5#&MS zuSDFBKZU@>Rl1)D-}<#6JKA3!>Y(=KsXy`8&Lzld5*2T7B^(JViCGNAIAtZ^SKQmi z#2+E&{9gvTa^^qa(rKy{Sy zG5Y_<9dr$F2PLoMC1}CdLC(VQBfcV*wTipV(wOMr*rtteVtY)F9!2r%0s7(yESF`; zE<_mnn?HrWAV**wDT-Z)$usgQdx=LaUa|I;WwL*NjqQ;z=c5M4O3nW##5 zoR#{g2;8keBvQQ8`2W85_|Ot1bOr!*zzO=`*k@j5Y;oFCKX40B7YX*u3CZZAp-Pdr zg)di%N28p-si&U0cIB9HScN~Jax*%0%ztM4E%xFRx1bf5$?bqn(+`a5a%U|LOg|HU zKg9W)vpYGUhqIhBT=A0>uT%&s-ZBE7DFyJesAoasUfkpPr^+kl(j^sWUhb z^SiwoD^Ji4Djw}8h*6PvTnyd+j5qFo1_%td$C-PuyC}n_@^Ca;Dk4@YqERZsF%(yL zs4$y-T1|rN|25cMwii;feWDF%;ks~^p{eM(A7pvpL(mIO-P66i^2So8!{)Wm}1nBXOqnr z)3ph@#Z>j6(w>cOhj@Q$WeLQS0Yr?Tfga%6&%qF7pI7BU{XOv6IL~&RT53+jZfxip z@uGjn%GxI*l{aCCCjgeIw5cfhsxXLtsC|5B$5l*b@Q(kkC>fMjRs8j1F_I%P42GM= z6!*z`n@%;_KS3>k^LGWSZwE=YP@r!ie>Yms3nY5nC(e9lSz$Tt8xs&HUR+IqR4tTnYYM3=rjR@QfzOfC}tZLv}Q5_ zphPW7s0qXWlcFikOCAh~1{1wNm+qFbGr8?KJ)SSD3x<&sNz8 zS4s&Hh76Eme8>Ygu&w@3*GdK&=ir(M26V5Fd5AQ**o82PHd0SXig!FhZX{QB~`^#4Hp?c?Dx{B+{j)C z-82sIUGi4>)M_K>%;r(ab@j=0ya5{ze9U{-JF}}6*5h|~%&;$jeJ>%@Av$}gCblx4 z|78(F2^m-Fw%0KxI?fu1UE~dBMf^fk7=+He7re7pYC3V1#@{dD+eQSpyUIlzg~h?& zpunZ?5NarO(A~t>d9%Z(Q(GkwYD6y6y*pO?m|mhN?>4E%rt z5if@r&TFzGCEu2W|I$_LSqQ?4*}r$s=^E*p zQ$1b%-hH>;eHHaX;X5i45fTIh1gebmH)RM2C^qot83Z`+>*ERQEBKAiRZ`Pc#lhUw z!^qhT0%+vwXzSo=Yh_H~ZszP_gdAH!eaM-f57bEY{5czLSPIY z1kq7i%LM`g8T-Em36YgU009XBA@fa4)idWj*TR!zaWQaPfMb23({}vNQ+Q_|_wfCe zyaS~u0wN3oBot!kqY~6F35xrzY_=7+^|q(2DpOoJbq5`bBYzR|-*eU)EwQM7>#yz* zx>T;tr*in7R4GfoO2rbe84pHc?n_I&l)z(9{h#kQCXoQP|LFjqp-RfY!u~%!ykDaJ zUptBx5h!U8DUtmDIe{-gCFK%|p^PwlmrnLQHa|TB>yAMZA(0{U8iDdO6k|v{aybkG zvFE-G{AX_NzJz8lgqGuo^P(B&XNY6$M^cmyf;#y0WHTX9V8bCjJ{kw2FsX*8qc92? z98oBOIQ&;wNNH{mCZ54V==`UvBXsoDVLBaX}D||Lfd2! z8nh6U7(!R5L)$L#30=W+fCirqEf=^P#6W}xSHsccaQh-lIT*ata}0}LhpO$&g23D( zh0q{@phOw^48DtBqaS`@7)`-vEl8B5+*(f^7<+t_gUe#=jj9NzD0GaX;CjO-mqaRn7M+QNOBD75pK?$YH z{lvf})*$+S?gC*Jlb3LXGW0KGHyjznTwo7WJ_03sNcI)90(4SfRp2KAW%wtkC=NUu zMmx?sxjWlY;*saj7ZfEq7@%VefLUCR_hg{Th&x+l3P0w($d)D~=GAac-g z7{+@x$)os=fhc`Fv?ptu!S}))6T%<<B_ zYyUNfc`RCq)~4KY0ZN!lbWwewa<)xBr;Ag@i&Oa1#JM=zK%51dvM+5Za||WbsmjcD z&Qov2Dd3|-Q+CFz&rKz0xj2QmIOU=D53p9p5J1Bt2Vf{p`Q0m_8LbLU;~1YZM_Y0a z`g`?NiuivUnVf7Uf;`kK6rra01*m|bLKZ;@mgFZ_r#OOBBZjfeOtVf{$%ZKiD3dK}0_NRjJ8*_l-fwF)K7*STg4O)kwJ`w1~loM4o zewpq8Ny3qUjZI<~sX8N3NO^28Tx3U!QDZs)LL{y67EV1!R^<|Q3!%fLR2I+!{SjnV zv59<7VJU_<_7gieKc-&yd%Phq@=?xldp~GC3EW)Ck`4`q~u;sG%D-nnl3&g-??Y z+D>I@kp6Kv(kYZG5yo`9sp!8T!AsICsfDd#Y&`gkYiRs2CjkcaVTZsL8XL6oK?F+u zp;OKdMOi-9F>gr0utmBVR9!%cu3H;C(SF&G%JFW@u{rD2{?H)ph|>`&E!*WpqOErG z&%F)5dxEu}_Rs>P&!TMp?$xVG>sX8;;ThF0cJ(fHWaNU6us$s~a~`JX%xxF)N$gl! zY$i+)R0;(0zRn;P!7+ka%b3PeuC$Eh23fV9)&AGyApO3c*awX;%1X{V4RyNJOh)(+G;E9U>yk+RFLK z_gj88t-C5QSs5=SYVL@i2eIEE)TdW_MW(%RdUygkslABrUM??QGQxw*n+~EYh(Q-W zMc8(`8hQ?{O}_nN8q0$2{#l%I7mA_x)HIsen>}}GQ2z}f@k^*BOUZ^$8yw;O^-j&x zO}D%p?I1)f|HKgW4_>)(mGa zBPG+xo$VNEDu~Z+3LU0`%OZ~ce><@ESt@rjEHDb-zxiTa~b+)P`e)B^EA2YZI`^sKHlY) z;1MfIDaIDl+2z`bl)eVk|0=#o`&+bB;xY8;-k9=bOZHYuAGZSu~?hmer zV2K58hbj{!bHR^(pY|IOQ1mCx@&#QV(jWOZV0F^`zzwtbd}Q_jJw3(pIbHk5H(7i~ zbPyKKon&y48JFGb7a^T~PA;DvhXg%rdEqlLa)nuNXl04HA9Xq9^OR zh4sz5cKI+lo$vZ{!?u_F<J+)(dJpr`pi0{JN@q3xYp358CN*w)MiU)&WAx9NbE3MEN-Ak`+Xy93$K;(nJSK1N+&w1(abWXD zritSfk5lA^B~2L=viQ@U$lrtgUJg(C5K~hcXV<=n&Ek-OPSN-u?G=r+q0Wx%_TD^> zLQ+UPMq%I-&$my{s;esF>sjR@NF9{16hS_CbzO}Kcg*C|?dd-ym5eW-(8F5jY_yYe zU1uSeL%?M~{b-9BkcD&}+%bZs?k3Y(7 zCQcH>#brsljTtgDpuSQEmO=#w8d!|xOxo&qMribO!)oXvbF>n`4`z{)zV!2fgCvWU zmdwcq9+YlKbZoN~L3HmiYQ32XrLaR_H?*@Ouk9}&PtC+n`+uy|Z%7QI|M0f_M|K5@6x+R9QhJU5mUK-*};*{$A19#)WYRSdK03c{TJNOMBE_^KpyUNd}3>y ze`O`;k)&nRB!wLQ@J5`DdHa8+WBV`g8EgI*!5N-!kXawX0k)bQRQ(YQYPj=e5*_Ch z`>Xruy53V9yn62UfqhCvoi0y)$A%u3rf$+=^9zxLk(*XUdSz+zSU&L=zyc3ix*AkO zn)!XI?y2!i4Q_hYi0DE((BX|-+mi=N5it!GMkUb^1`lw&Sm^ldcaHN%`GE7*_7|)m zOdDzPB0>ORRIFgF$hd_=bh9cQ6o(KY3HJt8p6?Aj=Qtn zNm~`y$yb6$Mn*Bx`*Lu>(|Pz_l5q+;A6JsH(^Yhop?~-8c7(Q;*LIoD)7K#!csjDD zMa*ahpl17WJ$#IMefTEVey|qRfH+S%gJ%X@Cz^B!?KAOlnm55WvSD8rQ(;TqXM&YN zK2!*6oN%&wE1)G}`z1^_7ne`}{ABJuy&)_i`c)dhfF5iDPX1tlC*J z+Knu6$8-;O#BnjV?6dr=qS9GKTb=mU4H^at?oR!9_aaNOK!@AAmC;qBc#l`w^?)M)SNU)$Ncrx0Y`bC1=*MT_#h zV*4$@H*-O$E+L#Z2amN3$F&)}5Od_ml)Jt8rWrSERa{lvU-WcB0yl%f@L=@06Elc0 z%D_oiRoP~=0XS#?ydK~hP4Hap;csfoX9uDy>m3}bCkB4~Hy>cWkL+}=MRl^q648_M zN#fswCeBIE%Cz%z=KHotq`>P(*Sh&m`JqY+nH~+GjB- z^`;1?(9qaNjE5Uw>V8Hnm)AOT7@K-9clruuX3D?|-70+X@}`o0o!=?kcZOWR0FRb& z`{rifc_vY5Blr_JEU+5J-uGGD`!Qbo4Y|IWsxZvuxyY3B0)3OU{iekOz)*(_kc16O z-EcmJ?|SW5^L?A__-ld`wA&b#LLk{793T?F&|T2}w(hXR7OprSqO7E-AwmEwWJSv}K2`V+7{m~MEP|!XtQ?&2pUvbs%*2cSBKi)! z-Y#4;cuHy)=5MulSlGd_*)#AyDjr_YSW@w4UO!9%j&p~M*rBRrmJ213Qf!HROIZrTdULlev7g7S zv8j*c&EPv8E>iRxA2FyNytoOWXFZR`wcV~4OZMcBn3vLyFm=b}@o9(WQNjr-3o~{{ zy(eBt+*c%vowr&H0G}K9${L#!IH(v`7!r1nGl&WD@x;h7w1p6{mc2zOLWx6Vz&O^K51uZ@y`MPc{ zAAAM;n0%hp*-R}hXyi{L-YEoBp3XLfPo2}}hZFj82^PwP6nWSnks;9{eg=rR@CkXF zLMG%Tic(-yARt0stWlFMc>0e^!$1=Hq{MSVL52*;x&Khs1zv4k5D2!-E6&9g6*lIr z-nC*_8i%p&aTQw^c%g=(oYE>6uBtM3SPMG%c9YDf7~zVUUW=5b++QqS1|GlP+0k;m z3HTjLGssf{k%90atV|r|uz{`~8npsf4G{6fqTbFsaDJ8E`XoWN_$Q2Sn_9sCs z1f3b9hKp0vuz$D!0F!h<2Z=q@tScncVd1pDt#GuF{SmR%C7p*Moi2MdwJOhFs#rCZ zotD=-luoL>`l&C}L?~R0T(xu*hJ%|8q^1%IlKKABXTK~Bl-s%B!TX`B4D9hIZM*fr zKG_mtaE*PuP&rwygC^#|E9Ck2+ql(6{tt!sZp{)F0)g$Sa%gGA{pBW_+Pj<1*H{A$ zvn>Omp&>QU%W$hx=Q}uDWtGkpeFwAQ^Zt4wrihTjKWZy>))`nPLuFP3Xt3W?Q=u*# zEkTEI^D8P!*uR-$d&m4iXAiVHmP5tg5&C*5a|wA5b~qBVoL9~EwdxIt^&KcUnak#f zQ#D9{x;Y^MJ+8RqT@8U(Wpxq|Bl*H9cPzx@r5JhcAJV0M&4Th#N~&Vs=3m|;NV05M z7^%6 zKOu`H+cfDJX}bz?@jB^a(fZAfuclVN&Sn2P9(BL27%M@-PWI2zhc}w-3Fqpx0Dx0m+8Oe;o97S zqh~4Y+%G*OmAHQ+DTwd&tB;=($F9;=C{w6=@%Wzes>SperMZmxV+U;StAMlBmZ}WV zX66|vEPk58ujKmFXQBtL(uYI%wFeK&$4%~f!vFpu>iW(S(7-TFE2QT#e6VJMTlq*LRPhMBX$j1%9?YF-Ce6hzT$))$f2u~50YjS|dBbSSXL%;nz0*tFv?D^E@Q_j!L zn=@as^}=dQWhv_o5qR(#J%K^J`dM;n6U$@!15ys}DpHdS-{QFMJG0@BWWu2U{nxlU z3y0pbdBhe%@x1uj(#F#Bs zg6hCdF(u9%z47%tHX9V6@*Aw_V+E>88abGuN~Ut%|5c!+{y4VEQ`fsFZCR3gdjwDv z6c?ry^U@;1Yh_e6H>Zu)Jv)4(F!b#TK^JMGto-b%p$ydol-RnJP1=WG9r)#=**d)+ z{z)fWxbe4Z=F(EpI6lmA$#1bP=cJ;d3PKloRylF8%T14|`y~JEwuNvOId4?-y;VNX z>sq`ST3(;IeHZuEWgJtqm)g;y;+M86s%5k&5A_Wstd0{!^(2I5W2vy*5T?U^G`5uE zrj-f$Wj}Eo6}9>GpQiR1B#g`Sw)qhd&ejlw{ zn!RsVj7Z5`?w!L?wCG2rItu(75+dg$)j2^efd2`}$n(Y`MdoMLXI=7d_CgiHv}~C-JCJYWPfaXj@R|Hqw{*GfHx_5 z|JnS;oGicttvfKzNR|fms!ybrA5((?c+O#A7io>e0~{?J65zGJ?K3kX1!OcpY*B?< zK3Q1iVs_iVJ`rEPLMc7nyh5FC=#8Dj>hKxOeO#Sk4a8jHSPDn+s73uSgt*_~JLL+2 z5mM=QrfU#F0uVcijapuLrIlA1AT|>{Kmg-<-i^6&kAG;z;f^ zuRc#7ma%o6#kcmhz)`}G@|W!1z_z`;)Ma)W!&iQBT;a0sKG0Mk+pb$2dEINmRcb0V z_ic(3`8*tq+;Zx{10#=>PSHzPrt@9_UX%4J`Z>X?hmwq15h73Y9%MaL9ZtT(Cnl&g zMY=PXMTrU1lFiX_{S%bYvkM>C_h1GK8XJbO4}TubTPcHusL`zg@&woizo zrej|dXkVC3%1JTyo9tAmwOL~g z$pZ$3>+X2pyXdm_9$+(CO?9fKYz8Zym46p>dW#^4>7@l{nhhkHe)kCEQR_>ed+w;O zpx3#5$~WK45$S$}HK^SP6*;M#17a(RmTln|bWKf?e>@Q`+Fyy0-i`(5{h4Rx;5&_J1XLClko67< zyp-ozo(2VYZo+nbDz{Hm`{ zdLst~n~i&rrZs#fN%tnm#sskWd!3$k`F^~}=J~ke#ee$nbE(-Ef92HiXt!;?O6bR( z>{xm_Py4-1p)HB|wY9JbIyEdBSblubkS27G?6x0$wlJD%CG~7zm<=x8{WMFK8SARs#vo!GVLnIV6ne+YCx+lbhbMTlbG9W z21S_~$J=lC%$eR+R%y3Dz*anh1e?HQ?c{+1s-q8i9?$9QCE0P+j#0NK8{MMr*V2XA zSmag?hY{;kL8sMVr2YL>{=hJ5m}f{DYuS9bJprcp`0>2g;UYN!SL2(akC6zQdcEGn z-kkdCpU#w?jc2@|KtiM3R4h)Q<0jXL^H_z%7h@fTT=vLT5&h`8F#LkaQs?2Q+5DGM zkkjdG^+t}Omzq#1i>PR_64PJczimK_T%15O`O*bawytHRc>Lo(SBHEux1JZT_6;5j zGyw(%87KaB`#}Wfg|AR=PRk*O>~+;o=N{PuVp?iKS*78O;y|rGzXQXFT?j=A7=DwH zlYi8Qd!4x_rV&h2kD!`-yBd}+J^SPgBmR7b^Y4kP&8CaS#RlL6Z7+Y#-Aq_m^v?Jk zP>&P%7;!tbX8Cwe0xWphdn@m`um^gVmAJ9jVvq(lp z4C)HE$PpIB5|!BtG<@vaU)$7w?sqheu$iekd^{}4FCF8BAWWrv&ERIu8vc#+!w#wH zF6QMbf9i1Q6KLY21a>LIdDY852Qc0((~-OpMp?SgNANhRiv{o&;w~QNM;jX3zm2_# z=+C(T^P7QRV2Zq556Z-4M``7FuD7q!m6g;cx-UVdLGzU z(hSDZk-2Y&H}UuDt>~h6KFRp=wYKd{k0wyT&OHLWobLJL5ITK94xnD-H<7$pWBz=a zNnelzd;|0*36vTx*ZL`ixFlogbv>U@3x4!7s@H5oYjXHS=t=O%+dpqM-1CV-L5eyD z$2@&rA^B_mP8Kl83!f1q6HQ#r>52c)YXm3!NA9kDn%LyAH(3d1ev`)Q`DicrrQzc_ zZ>AgUEi#pqjh!9-qn0Dc_uk<&j{uq+DAL)`pKb{9`;l$it;AuSS8w7VdZL7^f4t!a z@OeNvOlpg10foAE4lMqq3EB(3Su&#xlek}`jgHsp4UCX8nVlW|aXrhE0W)4jk+|j~ zPeI7rTiJj{n>Riy8?bv;4NkoJhm`b+#yBxSP`_E9Y_KsIx0llE`ut=2uVs{)I#_Xa zKBmkz=+STTOXmM-{uZBu!Yq74clmCJXCGin0#sqh-DcIy?n!SD61aP`!*6IU4EL_f zuU)lb()K^Pm?&wk6)x_DAjn7|c(6n`Gr~)oD~8$Z`r`RhA$xeTjKt-~a+3%s94x$i zm(7W>j!zHpDZ)zI3Ih^8}Ue6`(&T&foZmaEe(cf$|Ym(4*) zCB7*%rnDl*cg|2j*s9*GCVbR|#UKAcGATLPd!_awz3@MZs|jY|f5(~DueigF8l6Pc zOau;#Zw{dGDM|)e<+*K{`uA@*$_>6_p%r43l~$1ORPqQNx0uZ5B|+O8oZTp?UF{wy z`DHq*vN*>u$V9s&oX!1OTz2Wb z@$sV*$GEC*;;&0M+Xe;m*O9Bp+RE&dQ*!2o8)b5j*W4e_-mwap{HUP5QMWAN-v)-4Ibl z=bU=VejoG;+A2LuTFlEWBU6PwRFav`j=M3L=#Bih;(7o|EPxvM?rDe3?JHK0=ohw- z#_47`9P%{!qTqiJlJK3C>{f;zAKr5?{KaJ6lH-9!*C!r#FOG}e@&G4mB4I^{rdlOw zX|)}>G)uRFQ+26OE9eTX?fj9t{U5ld5aK0Bp}j1?yk}eC@8}m|5p--~m6eV!CExE( zZA5nLu)h9tw8Z4%i~IJqnXeCPps=x$Oadc`YqnHYTy%sEyX5Z$OjaqoYoOWUijLNR zh7EU71=0xWxs0~B>615r+JRYXQtmEYW!HUPi0WsB34EiCgI5u{qe{dD__9*;1Jmfoqa|CU~RoJog=(QPl8ZM#tUVg;tNwI(pK6`EEbnumNO_z{7ryVTvkG$V!wc(m&SsU!Ai#|0GK7vJRv_Jb zCMaQbl4M2Cm}%Q7=W#$qNbFCx;{8z4QdEdg{czkopb{p7 ztB!@d!w3OaSkXc%LegK001gKxmptIMAe4jGHz-SDFNwTLyZnSu5HeM&H|mdaE3^4iWeDAOlQgS+YEld>-{WYd%MfKo(pQ z++X&NbkV)8(z3~2f{oUY#HVuYW*fXqskq6w>Am&UkMK1M3toP@x7+CDH=Ytm+FrQ= zr!C$&qNv}{UF|c@B~1=npwb*BEmXJv$_`eRUfg5IIHD)wS%@lz4C2AU8aLXM5U2Z7&yA zU@AlGQ$YyF;DLWT)xW6A)zI)UeI}=exUi`Esg#K9zd?F`K_CsKWaEqyQP3W; zg4#iY;r;i{6V+scKsi$~3Oypf+ttm4FK8XNPQH`;oA{bb_YU%uK^9k4C6xY>+pyK@ z>k6mreMk9&E1un$q%0xjv+IEAP^r+$yq@4&f2wJjWY)k_>QcSw3$rbEYSGC37q*6n z2Nr)TG&lFJp+AC-1NVcV4BIjG0D(2CpB z9e}K&vhWS}cU(Qv5L)(dhJUVBGl0jMwUMQO!{xB0Tvp!8O{j|d*yWrsF!~FDo2tiU$Dt}Xv);!0 zy~;a;v!tyut*ddYKAl+;YJdKaxIZQ1r_WPZCO2_?6@BM8!@mUE_HVDv)23^a8b4Ik zpb?zW`fKhU;_P5duSrm+rLRBdV566jGVC;&?Fx8dAQw?MerjfCEM$ufO^0?K`;E+; zKi76Tj{UMh8W|c`1Y)>h)LA2j&v*GvA(8c3}+9^j5agvkSKcJBttf$8JwzXV@ z1fZ53n%oRK+_=6T@$_AKb{jYV0S2D$!bU<wif_p#aiN8Y%6sL4TgkoGlTR;KBVql2udrb>YekeTK94Aj}a1U>Ai8WM!?#RnH zoWJo1=0N}ePVjbi8QtDCnsp5`4Y*-^#2(I5bT+|C{~nt)H61jxBu}=*5&S*+>6r64 zbN0rF=92ehTz_w^z21)DKF6MTtz))B(VyIK8XyH4$xZ~y%An#4C+Wr` z+C+{TZpw0*ujFclM%OS?jpV)Aq$!QWO;AUm4%Mp+uipL`J)|$ zdkB)l4g{`p=;)ld;t9##*TA7MA$zD&%->iEdTReNQx){R)aVeVE38&H!|0o|uB*%g zi=uiH)z!x0TWoix8h+k!oDlJmW3jG)-)?#;l447{1nRYX7auXBs9) zU^%=_|1NTPW>F_GubTIHW&3!L&Qrl={zFr9lw*1dEZF8d_1?6=qJhhbzB(TT%1+dc zE(rQM9+z`qgVl&OB_Z^GD;FwQqY4U!;HBY;b0&ljE>9h!K@>(Xk9zG<<1dK@8~To( zn_Wk1B5WTwWGy3YaUP)XR(l@0N3|r>u^ai|f__K?V3%j99F0^7H;9s|2vJSaaU>Mc zmI<)Gk4l`-%lH1a?`$pDbPy`ke_8p-H%LCAemFk@kJu4=a!rk+EenM-h#VAJ6#5 zAMuRo)kXDBd@yO2^D(!^&2;nusa$!HeEG{|ccV`>nXFy#9LKh{4gFWx(^Cg)474-d zda{GBEur5Et_KMVfBDABBs|_f#I_+hHt`;nrVawc)wZ z2Ax+{UmK3k!)!Fv4b}0r)cL)a4P`Xa)#vBLzrRnNMfg8mt%$C2MN-Ss8I2aVzVK7i zx(t-hf?OV-&MR#h9Vkf6O#?{`zf>S=sxLU#W#5(dObROs3G0pgvNj!|z8(dnC!IDk z5{^7vjW%sFl@>Je7yHund%sjh2ba?Qr_U6sWNy?o@YMiS{$*rhxnHWSBNtA|b()v4 z7ua}ezook@@5UHI&bfuC=^zns)g!wc+sCz$yKb6ZqGuBdH40$UKHA+mafpOe^I7uE zXE~f}m{?!VFo)eu^F}L{)}Y~vyYShe5Y9v;4s?(?1^}p=TF>_`hN>!VJTvEdXw$)& z%EI{|jNkwRUnGq`_iW)-d#EQXx%6%dLaOb`a28)3YV{{Mj&r60@EJ|)cJ=pc-I;l< zF?h%YJ_3(>>@R5<=|_)hpoqMQRp*(Op88@OS>(aFWWtl=xVRAD1xn_4mVV|+(n=?j&LwSZfSqs4d>U+^gEpa%P7HF!v{Jv~5T2pawU$@;`1P|t3X=BR% zbQ?n)*i)vmW!H$?x!APqJRYrFM%KYVC0>VKh>o(7n*%^uXpN8M*p2SPY8ed5fU{54 zH#Gv+`C^Oc! z=FwztWvd3M*8AvfsH?%z%u>CnsyUct$@VB&L|?pXU@9}}66m*Y zlZ%h$%LUx6P37}WC2|p5Vas6UoH$?3Q7UiyqPB@B7KAEpNI zGqhxJT}dwu_8SZ*8HpRccFHqmS?SGaJO0gR+E4J10y>*RDX&!H?^V$S8XeS{!t3p9 zl_I*BtcNk5yR&YtR)t7n+Y9o-ukVtBt3>1R z$Fnq2wR+wQA{8keEfVbb%YkK%rHGU@3}xdCLn3}fA(3P!6Yqp9=N;EK1E?rb9(2Hk z6W@#+!Yry>xh{h4hHbV`uT$@tc%*+IHt)t_r$1LC4PbVtYT953JW_)*ka<;WG{@!D zvBP~eAd6k|0XXGzL^@`DCV>cHa~s{}ers7PtpB1?kC8K_Cui=cJ3?kO@-HuIj22&iklLUt zp5+(XE+=MbL9X?92%HQ! zsM68&P96>+2jWT$^$6RY{34pBomi1jJApQrvyd`KS$dlyW)0l1N`A!*ESHT8VMe|D zh_XzQS-0ff_v)v|j9DSgHcOv29*SqhI*pO5x?>T=se@ne(*osAA2}q2)h{v!6T9?( zmJOj!5psgfprUzOF!r=52azZTqvz%fYQ1-rBsv_Y?aJTv*4*moSc8z8NPwR(WOxGr6lZw<9 z+9pwawloEXfz$s}jcqyKlai8j*omCc>4op)POLSGxU%@ypIBedVyk6Sm)o8|U#M5> zs%Ly_ZEe^a5jArH=$_n*_y7(kRuFAL@iO%_UKce)HeNw0nF;O1J6;9h#v<@r4`IZ)u zQpyDV$%gwp1$qc|xS`|uz4`dZ>cMt);QN~ZdhT{{b;aFlMvFJHz*Fjcj|8_1_eHF9 zKAeqmV?%wY)1MAj5uS7rPQVp&v#n;63H`goNBwH>(gdiW26&$-MPh{v=;J2gRRIsH}{<)oR3a1O|w-N z0`5!sstfCN9uep1(rCo=1mA_omRq=eR-<-u#pu5msa( zztkVd>yVYQaNBKGfG$MJQ|Y)r{76kqLy@G**(xy@ye0izQZ5eL9~Lvcpk=W7Y@eXq zSY^-R))65rlj~TSmf&1Tl%MAe$UoocjTKs#b{llk+BTwOK?w6Og>wv}(&c6->s!k^eV|DvMysH#Ti2b3uL3EbCA zZe~Kk(6cz=P2=^jTGF|;mn4~A(;6Fh3UoAZ)o+vq#9!}6`!fxqrtv@g2un@RPb82* zAAm$GsGd0c#U6ew6?Hwunvw5i6Dg5zDzc-kRV}yi)8@ajX&?s%%~~&(_Y%me$KHB3 z+A$B`Avt@ZPcbZ`#lsomzSHF3xOY?g+`V`otAm%k=lbu2Nw}Ls&YOX=8%@6I8E0$1 z%5yeK6f*?cf>2X^sC0!O+$GIC?6jn#ZV3EamT(d*U1o*4+2HI|-{Ry&o{;pn>yt>; zaAW^Y?cPET^f!`Rk-I>hb^9l4_n^0uQ7AP1LWGSci?rriJFeZY+RWx_J^rxK&Qx|D zI?k`JE&h6+E8O8IoK?7UWosyl(NWNoI5tU3Ixt9?m`q)dlQ~-C&Eb`jI+LKG+DF^8 z6b=jW_JeG}Pvvly=UHN&lytyyDXK);AAYo5$1dfj`)?@pgx}XMl9FrXNmkI<~dW%?jbsUqpu$s0Fk-F6=?AYCzs7rT&&|x^x1wh@x%1> zo=d0gZh+7Tz1ThrgE9Zh!Zi}SJMeIiO#)I=Uo}zyBXjMz*(ztQ2;an+pRN5n`rIB- z$e4OMzT;t+$1D)K$BMPckb}iu|1a&xd)!x_w%}AtkL$-2v z**}Zr;(ZA;Jdb=aD*T$=9%bhHNx}nd#nbiW z+jSVh@FTgs;Kkyz{WO|x*3bElhr`krgN9OeZ`aSbQ)g??R^*wDti2m`K<<|mrV>8|`qc{+ZMVz(B#D}pbv)iVv&k;Ab+7$o>J;^b*-jtsl8*FA$&w6d97^7R zJwtoGfG?GW7&E9JpmPgGW;#&~B47Yn0E8q0&bO5qE!OVi}qr7b=7agP#M=x zv)KQVa>m&B^u#m$m1RGK18DNER$?{KDKLEN-ADrn*rT)KqH~_}wp}nfZ5TY=?qtuS zqqi6MI{bLZh+K>p<{xFIj9uKELc4CV;qwbhqQ=+c5&w0B3U2JyI zTGif6Qh8Z^PLs%48=}1(;9F)BazI|M8vB=<^dDXU;{PYkrVxPk#-!^XSLMtZ&{>&)t8f;NbIY z;(WZz0#@72SjbxQx1gPg^ptui?0`$nV^b?qVDn2!UAlgcMFs1%eAcQS;V4&Fr9)}z z3r>p{4ob&2Mp;R>W8UsZ>A%+d9FZ`B^{8=k-llmxKl9TQvkvwy;~J{+aR*6*`T50t=`fxELlNZ6xl1$G!E+1r+Hew>>$8}mWCF= zk#Zvg5xzUQ(V~*p%a@W5Bosu{r2C4R{8^!2LVn3;8(u9zl$jU*5|hSEvP;?AW?g-0 zMDI}`C>bRysnc){T@LT#Z{X#>G#3<@RrP&dExP3yG=?(7S*tb{r}%amLpv~O19O*0 zLWMWmjBW|old*o_231Cu`&&fl?T?MenU5WOFlskKX9G?8XaC43+JH6i@X`5Co}X7e zS`nMIyHCU@?FVP3AEF5W1nW?szbHTlDF zz_*!m1y_e+LCpH?sHBu69Si^Q<@FXRQ*|xk(180BesYJkziw1g;b|9D75gt&>8uXp z_E>so%;X|DE>b9>2x2+&c&c9pw&Al%6VkAztLi{`X5;RO**a2X3RbbC>l}<*brh|! z0UOtNsfTWKmDEiizRUkeJl1BGeAPq{|EpoWD89$j>@Xx&6ajIWQFEz3nNdE2_773W zSAPZwd+MJ^c?PT@ZRiIKBFjDK2!#;MpeB!OTwd5a)CrhP{dFOiIb$e)tBj zp<~TQqn#}LW4SwAm_P8<96n#QF%*RT$;yq)-n^49Z@Pk%I48#*)r&bRw}+)qI&3Ou zO&SD1$DA~#o;xb)b$Seol;qhbpAJNWeU_6zdpK z@cRP1^2~R9_2IHmdaPzsdHcU}xa*M%7=4W8HV6a^UVLYMSaZ<^G>=Iju~9~>-@TWm zTlVn9f(^J8(<;B;u+kuub?30D%)9$KwiTC=9OtBSdP>-FuNyUr|14Y>W~p>)8um0F zH!gdZhH(2qDIIxk^rg(o+ksb)v=Rf%%-cr0G{Nf1-L!Y5az@T5GM&ljrp`Ni-e*zS z3RFc!O37IrPUP&4r$i_n-4io-_2^5vZ_RYJRF~jTZA|StoWBk{j&ffGPi}vPbXQI2 zUaj!O;WKQM`l{GoS48i`oL2uX+iOa=e(76y4eNSrabA4BxL_t?lVX=hwtMopZRH1C z+I<{DQo|=@Yl`uo=QNFJ>%)R*Z{a@0pS2o=b zqD{r&q%F#tBTk7~zGw#z-tkdm+=2ulCEhw8tCsI(-t5gBd;9^SO2AHx$&=Af=>7dMdf>!!Hf| zO_ZCf=YGKSQ-0rKlQFbf{=DM9_=4T1XBMwvO7;+@b{@z58{fufHVULo)p`7F?Mu8o z^419960g~LVE7r_KkRf&Bx*2`V_MN%=9ezTuIi}5#FPRhSUt{X*eLhcL?2fRY^*L| zXWc#mO&V4@oI1L$VHk~?yHF&4wVsvb^=z!#%a{3UdGqM2(1MnkQcpc4?mFxHRwdG~ zGAu&K@FI8CL7}jwhMc4%_SMw!$;u630oaCVQtS2e!LqenH)%9?Tseip(rR`UmRs>V zhRKNjo%sEk2X4am`TNW8H7h6Y_vw5;eI0+k{Z!N_<;Z_>-l#)#x!x4zGs<-5Y_ zd*}fJ9eDVzkNM!czhSp)l$6(EL<`b{kd>tY!;~!Cu$wyF4@#t{L0S`^(}woLVbd7Y zDZHp!1HT(SoO-{XmshL^rB3Ixvs3Kci`itn>nihkaqAqMY67aFL{te^75JoVCcaW1 zUc-+~(Rep+9{(ABHOnd%vABFC0jU!R2JL=cG=nrpDwDDzi{%YV@5E=vU56n}lwekr zDNP>S{5W2-x}iZds*q$j>nL_7qo7a|h@L6g60dBZhrdY>u@DfKJjVbJ*zBu7woH;KTRSrKh*3 zz%D`&DLuX%!n<*(gQQ_g_YcC4)^3X86HX$q!V8#Y7m8h*Yj zUPDcwhJe{D@D!6mcR4%i3h9*?z3z$-0=0?A{3!n$Y>GFy4isUcDJGgQK)M*8F<`&f z*_=J~k?D+53S4mkOjBU!0u+g&GzR#3jR4zfO89>77Ea8u&gax@p`K3CCM_MZi8WRj zAkFYDv!(_W9yd8jNo*?G7d4B_r|T@+Rls$VMw1Zd;I(@$W=n1f`6X5K?36*b4z240 z<*xTeU9JZ+oqh==Mgoz&kz*VRdAs(pZdG{TWpg&I*~8FL-APSyVYJXYsJWa`gdio} zNp@x;Meax|EI?*LJhz;3^a00s&Y=EG9Xt@zghacIiL)NYY!;K^Gj%@CU(M@duI2eP zOVJ5m??|e{ZFobIkASHY0RFZ9JzTaT0?pdV`%FJy6wZjM#5ABb$#~ydyPa}hnH5hM zTqdyso)T>y+`9ds+|~fwYD*#lc~#9MA=Ql$P>@of+6)>>VM#1m^U#LroRl*Nn_|iE zc)J6es)kK2LpwD5zv-YQ)HZ1+v2@p-2sr~*h4QU?%g9KKXY}B%On>2eR&3aVVVaB> z(1m!{f#dgTs@&m|w;-k|sVIwHzvkzlEy|iBPKhbZZy}&9h#GY7Y}sAH(v7(h>m^{S zK$GHjHdW!gNkiz8k!qFdk20$o&>@SlJ-b9KZ=l-m<(7rhm|wKpy5FTq*W?u9Y%Rx0 z_zayV$DG5(bKa!N@4=Lo5x8GkBKvfAM39`%46viFIBF!mDijLbC2Xu&#)j(kq}Wn8 zq1{RJPV5u)+#bW488sSiMDV*XY3C9gof&<H>M2slrVO0n^rzeLk z>)SP(zFo7Uj#)~{yd~TD;LDX%*LoN?qC1zIGm1gOy0BvLj))?99lK=U{KaBoKsVc# zpZZ8=&#qE5O~oI@Mbv;2bX&jr@#3g5Ljt-h{CBZw3a5A~E+i4T_ z8>qghIG4XJnns1sL%=X_sv6_kbwM{o)XHB0)1Yf|D))^)j(;xv4lP(T%8(`<(game zUc?ptI`$W2&uQ() zU{?}m;SiCU$#(u2tvY{11usXQe z>xhg&iQAKVCc}!QQM(7 zT{gx{?Azk8ernpHtZ7@LX$9>%q_nus;k0qmX+x1xa@W&e;qmL3N=S(YASK0xk)(t^ ztBPRa&>mcK>S$8poJ{J|J)#DH%cgPdr06jc4O}<(Q|1@##%~$~Oaom?c2rfcv9dU7 zHDJGCAQdlldL0+^>c-gishriVH}6floXdNTLX8sWrU->jahcea$d{#Gdik{Q6}~N= zPOf`5YpPfC;hzFjU&iXBd>?y3^uYcZq@$(fL>>6D0J7uJMc zSe2q5y!vE1q$hFKr~x?bkkOSz8Ye=Rj2{-^ zzE^~xPx}mdX1CI^>-KTYjc@Sb=PSs|FK5%1y*%>Nx4i%P3Z`ClG!9qe_vdih=+G&h zV^8_zTA;K=S#!imxGVoS8HXeC!BFig$DKHcu}Al!sI;20s-T}$2}b>^mIbEL<3YE9 zX#+z8oC@bm8jLMy!Fl-nlNi^tGj>fS-eJdS)42ZFF&sajSBqoHtF2^pSy5O9Py>cx zxjh^~#YhbnGfGPE^3bODxn|ja_$dDeHrH=uS=Dm>z3m-Fr1gy2rB)LP*Y!M^WLrF% zXncqTl1yhc9@C938)uGw)6e_)@35$dRX&c|BZt~pPQ zxSVr3jU+*HU{^FIWcTKhZlka(8u6Nqmb!fVq?;R;zebtYQkc?RaeRHw6}*4?g*-O# zB)&WUcT79?a@xg5QBD+v0h!tKOtXYWepNO3-kNabRX|31Jl!)}*P*wj*2BA>Eei|O z)OC}ZIxovN?%}`hU(adhjim3ujvRmLP@Z}78U_t(*-H0wU?N)li*fLD>y=HP0Xf+0 zk^XyAUxccu%wD_|pI>kA$rAFLs8Hs1V>-dKgLDXzGhG}r$m)NTkrdB!*PKO0 zVjOXH8%~=>r}SiAx#4^q`*B&nwRNS{oxTNp9y@kn?XL+HLz--;%IE6EuX5A!cbQeP zp0)1nyk0PyPxdV5>`tR`G*dJb1bq@aaBYtZu`7`lp-R1mKodEmOPzY%&1<{gqTE*! z{yoNb*{^yLkO6$AkIxFfWO4bj7PkvKWi zW`Z@hYLGRijff7{`(xV7^;A~5!>le|Z-CFHt>TmmNAr)zFJjb$-t-^Vncv-XJWsrN z4fZJSzn_z~CTos3chb0teYy3XGx^`sv)ET$O@H-x68XMlEmEsP)eH=2Ac85C0AABik*A!Q#Vfgg@Fo0n-8-Rc(WSv(`kYChq>j8i z@@C%6o6W4^rFhLcQtb7(v;fhRgn;SimEHeGysB|=*W2kDZxzKmuhSX)ch?)uww5>) zJ!GLVFevsEP~^#{gX_S4YnN*0w*IGaTmRF-mjAxj1eTZQQtS8f^p-hP`rS0lSM92f zBCHspuL@RjRnKFsZ%zmXXSQQ-W@BLE;;XLZ)itYlZp9MZ0W0A7=uVw^{P>eY+q|=4 z4ThAMP8ubPfD(_JO~v~d&_25Rs43fa?Zal%@HaE(`u#dfSLbs1`D3~M){|P;`{zf> zwrAq~HE{n^m+<1_-}3#}YY7B&vOA=3;u%MA=7pn(Pq5M(g%Dx!Sq%@53BeBX0#q=N zRMmT#y?zHL4zpafBYJe?yL+x>M{zkeRiSHE>H$-YbL%Rj7L{v2LE)mVW7ttw%+j)* zfW%`o#I-bxP2x2IJiB!^Ur)Y?wUxOnD%%nY#!a@x@vlJ_)6SJjk}ZKx3Z}EMdJ85d zje#QJ<1{c?Rk@Bb-y(V>4B)aZw_;bFbc*Xpx;>3D|NdJ5_)I@v7Jkjx^!;1B8|a_X zfj5p0FPM2_zbUM%%%{*(#fHiP-r2o0R6D!@ziBXk-!>|I^`tqhf}?RZ8^?9;9@^%r zVZ%A2Phal%W+n@B_n-=arYhWg^f*rLV`Zcnrpaq7SK(Dns|{&GH3&uW>dIAIJbXy2 zu4DPi-IfyE44X}5%l2Z%kMG5Z@hwlFe{R~MtU2P87=Rvq+cRoXA98n0f9SXQ?Tx6!iB>Nj;}m#pEkZd3Sj;y+kYwuO4#O{X{~IdPURH`$)R^}WuZ$TyEt zpEdTDa`z=*NDuXf@ zRZ2{HYCL+BKusY8*=@ALY+Flv{59Zk+88mmClw_%n5N0`XAI+*;|HKLX`Xt*G5z>> z#%lcj$ly@DS&0(2FTl2#zLFYdKaCiT|@z=F~N0%*IkN~OzjqWHZC3SiQHpPxpwJfkZ>$WnxboPe5;M4H2}wHBeI5`fRK0tR$2ub*oHMlAhK!8_u>hS#!iG zF&ozI;m%v$BM{Ir49Vt=g?#ndN}l`Q@95AmjcvL6ck4)yl9r zHm55K%Qhh~pykABz_Mrn#Hn`N##ZK(PDc|OhBW9D*Ntm?+(NvT5SEFcU}A=%&>F{J z69#IqU4m}5o=0@)R_XPmFKVucIf3*yc zVj>&@%>`VE}za`Z*j1quFknRXK_l0 z6N45Q5!%O6io%a4kR)kVaJ?=Atgf8TIXTy1Q*ER>()rKONBHl~*VtIIu|?U}Gr_Wk z6u39=b>ZXG>E%cyom>OSjyr}g3qK+ttw~nCdv|ERUG-a;Upk9h`}_-s+9=>hW^|)4 ze^tbF2TUEGtOE`_p^3BGqFUnP?KaZlT=tlQ(RWV z_5XUG4cqsIvPXmzbj(WO$=lE2nv2HcZ01jN+B9xBt>tvV25y`)jW;%}rq=JnkdkG^ z1)MwWJql{8xq87ntgXpIwdsU>7y=Z{vNQ~Dm&D&Uyv&yci_oPHO$oI}jL33vDt6B5 zXcZ(~S-F_GrQf2<0AABWK>Eq6-@zNX&jBzhZ5*0t=$dK?66-dz{%;coDOzJ8I1_`# zW$zF$8&azUr{$c&izA-q$zjiuq@|!XnNrvl2NN=nC()M7NBNJjwRRbX)X}96k6zB_ z1yAGCYr^7igaA!QObq;{pEvXV7gp<_cT)Rcz;DwFmk1%*QJ0IcAAf&qm^ySI2@X42 zlVED6rg385USUCb(eU~L{O(7CGaN-Oa+h<;WzQR{gjp0@%HUi#7pL5C44|4VKkMilO^Dv@hef$(yilfrvUyOri$Cka!pVg?q0C+tCRxaAk zrZszbbJ}uV{CFW%wbt_`1xUNXsZ$2>;WJm0lARdpoDNpNXPm>%@k4s=_q@e*wVde`fS< zQ=jFck^kh*OI~2}s@y}mmV==!%9sFAqo|-VV!4!(mCJYY(YLE3;uv(pq@=8t zw4^wGcl;=9s>Ulbmr>*KlAV&s&1W9N#6dkMuC3$iO^(8#@!y@Kv-$k6=&Lu|;;fhfQ&P;ma>t*aHD6{S%(RdZQOW-Is!!&-M zPiMj69Te@YA~ngyjb}eYWob>QuE?8D&SdkNJ>2#5r9b*w4hGrAOuR=rtC#G=h*BfQ z>kY7c(RSWkw*_yrSO(oN*|4jacIipHdD8_v@X`0o-(;0s7~8uu|F~cZ4x7r7+GvD?zdyYhD!93(Y&)A*<#OL^*D!d@&(ox~ zt;w1rN{O-AR1`%Jt*%sjyo2p|`=VCpHBFN}MU`}JpT_0K9mQpnhoc*2C?jd+mhJrY z%Wu)8iC-E#w{$7}GBSAUqVq{|I@wiS5e(c7KXJe`SzEdncas^25TI%jRhUff)PwmI z3&X1GS{s3iV0dZ|rnDbQmNSu>fSZ9SUFqOzOwFs;qZc(&go;P6C)3%EJNx~M_0?-=@$MX8lKj1e5=(1^{2Z>$i2;pN_?ZOBp1|Z3n%y}Iz$01i!>fMTh7oP!} zs)}(68#cwjCv9QJRFK4}K1>OfMq><7W4$ODHtWi(-$REu59v;;(3js-aqHGG6=8DE z#^<Ba~o-K&ld<5MjdCG=*KcrBv-Iqo%4ptT39- z8(`U-&E)UgN2eYKc4HkHZO_Deq?3^3LbEC8&8%s5yN2CmkGxNUfaD{Yi8=D5$I^V3^$Y)wj5ffEAn-%of^MR?6|OyvEF{ zucBLWYSchFo1)M)IfZ&}b(0K$;FuJROh*D9!w*PoiWw^J?>DX4(A|Tkw%EK8pkLAu z7MIP%Y}P=`ChVj;tQVsyREDM;#n6Joum}Jt5r)KH=ch2gf}P8D;`2nl01Vw^`-Xy_T#1piyU+35ny%;^H3tnAkZ$%YL za`Rc1myb8NpXE-L{QLU_RC~Pq^|)ggn32Wuy#*2deEfzAbTVvp zI25Zk@XW$Lq6r7><3@5+#*HM}JK{I$(WRTErOzRy+fsjmu{KF+J(Y$FV?VE^c8+R_ zeJKQ*^o_AAb|z#tPFy}K{GN{sz5|*lLb4$mQD8`e7j}Hiu#~=JI;}XR?X|_^)s@jB zF^g^qKixBXP)};m6>rm2c5E+Z@1E$%54l_R9{RvqGaTBUtZDN~%!jYc>Iv-I!O9i?>M@l3h``d(>nung4pkxK%~r#E$*g<=MvE z(s?LB*Ny~S1aOKZG9CMm(Kxx?*{rQv9txZl2sB~i!p^^IQD6a#%e zKnivt>EP_hRo$)&t^clMPFVa*!?88YP!OVD2KcUIDW`Ur#M`^)@nPOlY{8m1ict7n z_sNXO>_v~nNFlScs)GNnTgK*!QU+(V(g~=eLhlMu3%@`NKP5wkGH;A z5kAY7R$NsmM$?oiels}dGeC-Pf+zxwy+z-pTs_fAG9EVI#?w$cgQyw78$DI2f082MbE>r=7idQOrhL ztS&v|*UzWey^3_#K#JX~t)6AUAl(Kz((wN{#ZRpnyQs<<0 zXVWKXAONN``Ebune5OC@zM+vx$%3-g^i1f^|81N`UR@cgP|&48pQQF&)nfv~QoE68 zbA?}-luX~fjdwS%!Dkqp-LpUE_Zfsen7vfv^&*5I(GevT7UY??M(gtNsn^t1H9zXBBI49hECM8OpkC7w0bJ`_i z=-Mrl%q%NY#pSZ$h_m5!Hwp{G4*oGw3qM}zKxsh~3!sD3 z&9Jn~>6JK?j&YsC))%-7q6%67wMj*X+JKv>3+|`JZ{41L8uxi>+t)m`)j|KH4m>^T zQj&rKX7#=zE}H%!enZEQl1=6N_-w~U-aqw1zFNP9_tvgMN=cmE#_ba)aNeMS5o5UG z)KM(nkVkG&x#h187P0EwE`{?i9K!T3*O0%xG;AM_CxB0cjCc{0?yZV?y@wNPM&Gss z%aO(p->j#qtS+R$0O)=lyHn%pzZ_4;t{LQ2*v#1JGF*Ic`{S)H&Vejp@^$I%1k797z zrNn9B>EPXxrckFBv8wzuamf-OJlI^-{;Tdi=CTs z2fKoAEZ@f$J2&&txD&Yjn;C2>v9c@&waezQQ%!Th8FGk9>-0%21sc z#ilT6LSH)ATzvKE%CPsIl;HOpDDe)|YEQVV2`Vr93M=Gu&gzhH0Ems50U1wUXl(~ zdJ2Z=#VU=O85>lu}KD?O7>58JQAJ%JOz{`#c8!`ffI< zE*Ho348MPz4jZrDeG#vGxrA50TpB7npI2PPGoLNw-rt|dBX@m{(d_z7ML>EyT6uW7 zrD{`2Opac<{qQ0%%(f+1jx;u}&co}8w5SOL!judd**$Fe@nib1eB&NGz9ugk0aCF7 z_c6hjN|*v8kyhJ`;5MR3hjvvcB)hDcd%yHdK0527s59$Z22SSMKI7R_SAl_tdp16g zugQc=37*FcDKU~CN>_7q=7Ec^^-kzXc2@@ig}mB%EZ+Ady0k2bTdIoKRr?)hcX<<+ z7XG5#QT+u3npFAx z1k7+b*MQG7k;3G~?K9|;l*7=}ZtSQk=d&G~!YuM0J-~*t63+SH9X$1V=zCbRx0nmw zeV-q$`#p|kSxm$FcjU9@uHl+{-lMRjnt(JhFqyr4JLxI$Tzu&`-gKDOECjCj=$4>^g_(`8V)c-sN}=Pgn~9LwZ?J_Ea#Uzmd8?E=EuQ zDME&7#n9yPox+bIS{5peS808DQV3F1H#VYMQve31oDo%l0hpXU3AUw2?BwFKoc4P(}~dvGnvl}@bR9-04&{`A9WkI9>6X2 zuq!mB#O?R<^_H!Xow9+#ebp3K*AWnshHkvNVYN5ekzW?|_g7Q}H6dKSs+z*UQ9VdX z{ppM2wMALerY$jvDRCTq!T=V|+8EYN*5R~q;kAfjSzohhF>w%zdV>)T_yfof4nj*MnweO5Imwn z*MyNwZg+i)-%XM&k$(-mi+A!qU}McD?23c&8DmJ&vMKRa&^xgU4{drMFwqnfz2S#y zMkugMfXDFTQf)XjTbLy#bQL~9L=8XzoAXN<`RMa>NKWFOlP7ab&#s~0qF`TD)bSUT z)-Y~hN7k>;538_gx2ud8+Jy}>HX;YI2ILAL$?io{OzHzkOv!F^H`fOC@@d|61WXTx^ixp3 zf{m4*fgogcX+TFolZ~-&COU=n%NUvY$FS{^?47Cd?`&a|lAg9w1SaKXWbZSAnMih_ zkfxO56h0(SZ6=Wy$f9HX80z#g%6)n0(hmp}6i&~%n`C>-#p+IIHx9dE=iR)oDf3m4 zZCb?F?G#!5=9E6M@A@&Ze8bZ z-^}3eGiGsYkM2Bl;_;*gt@~T^_C+;ndwc=RI3=nQqiG6*2X|uQf^FCxHcZo`UFURe zf9mHc{(L&ENj{5v$hh0RfrmiQIAX z$qX6Pi7xHa=+-?G+Zq)k5NvgLKArRQwIC!u3fls{DF($HK&6sqlFoim7ydA z*c6lg$&EdM3O$>`6c~Vj^iUrtjbd6A7$_)Bo+1+irbenFRWcPNjymrG8tMz%(aEt@F(~V4l3hp9Kjjp*R?kM4 zetIO1qrL0EjnA^3S=`d+&n>S1ykIHos&}FVv$>4mtgJy7wM>~Jc&sGS?3VI+ao@qb zvw5v$<_%iigapMeA`}>;)s)ZhO7bhK`O`;V@y-=f8Q38kAp~hjaa7bqhTWwl$8r0e zr*Y$--@;w*!|ylHGzFWYuxjo$>3E_u$S)rxKSC-6HiXv9BBf;Pn4a8r>nX&?JD7M>PyYSc58(<7 zfDqVgK+jSzq={@HXv$^Nc>KIm=-aM!RbHFY$(7AsPie$^Py(7_q6kQ|*U>AXv2Mwd zeJ|iOYG`*=)iTZ;K45AD;0ie%N~lMw4Q63N}(5z0m_}LRG`llzJ-6 zIAq8HAki=xoi>${x;HR`Ze{_bs`a#U1Q?m~Pb89Eb<4wpK(H`k5Ge^er>`eMXAS*j=)4iCsV0K2`r-4I~z%OoPd-*_EbLuK&rhX zV=^!2=&awjN`Yw&f9ib}F4c}E6kN6dE}MZ(k*GoiX~U2TDhf{3#yx}23O&5HkDS1i z&OLCbDw+t^nP}-N6ydQ7zJSh)3zmjH|L%;@IGeeLoOX@hpEZW;>}1|~_j>;Lr{g*O z%wY`b)egP99^J3wuJ@x$Si5=;-+Z>}uzVk_K_ZZ{<-ZuMqi6pPyzs?Mygm00-d}JR zkGy^j)pcI}b^m9SmDS<*>jVM@ieq>kl!UO#&Hy$|;k080 zaO&y9c=MeNIPJ8dI82GB+Kb2I#~U!HHzY5;Fo&wjXn9UQ4Q)}@w0TQR0}iJx zl!vWaUl;y6eMzX;ROmQ1jmd+0(5F)tg}z!oTE8*M z*(AC3l@!+2wAiDkB^06X#IQf{_?GusRK6Ynn_`e+_l7AG;s;k&wenBmc^Wd*-sWb`RYXI&eX!V@c2w_#k&Fi_3P9qEr~HbI+2+g&)N@*DXEJLGG&yHQ^)>HMutufr2I! ztF?&`jLQ5o?V=Q}1LzPxh7R##kTQU~_DMXof1s`(ts3va^5@>CG9$r5P;WXw^6C z(J6y}UV9=bi7s~U+{Zn)zE4$Uz15E^N_?m(C39zO;Oq-VANm*48rq<&X_XQqgg6O! z3M-TJe^Sbaqii!3xEJf4uvSVLsj?p?O7_)+J@67R90I=+K3=e%Z&z;Pcju1fl#zXz zzI+Sbpi9YRw{iUNUJU5cjz{Lq=c`Ryn@vEZm0_U4riy)>^U=G^yF4@wKbb5@D)<*SGi%;CT{9|I;Z7~3fpd-Lct=Lpr7dFA!_{4jqjfq+54uhX=C zaZ~Dlj`}c7kcVjEZ41wMTNp!g3JWX4g1@Bz-KpT$;CHvbMXzpITz~mQ7Hrr}t;dgF zH?XS;nyPZ|tNkT9~&F|N_?GJBKQ{6=2j1s!&_vuuY?>`IcaC4}#=5TYsx3dsJ z1)c+r2MU1YLWnP=l$)At(+jv07%!!)6hi!@B}#ku%noZWVAutw(~9nv5^CK(UifG} zDP5A$P17>j+ElJRX*5?(8G$J!@2p-M(doFsA1duu6iI!+&$m0bb7r4@p)CTY&dTyU z3~6#yde>0tZm{YOEUQ?{ywc^ULSb^&SZ?Tf9WU*A850wU#3pPwRR>c$Hol}2ZJC_f z;crYY`ZuCU5xdhn-i=cctgU#9+CU-MuE7jRyMiQp&Ov=&<<*ab6qp8F!V9?Yl87Pe zv7wV7bW|bnNr5Q@4g`K_LpCW#t_UZcQ*HvF+Pjo3`>v%ifzqI7`lDpTpL$T^YYn%r zd5vwgg;r~bpyKu9)-UK3pUIGvZlUcuC8gkNlKvl5M+6xb^YR5E*t%to9V`(v$f2wWrf7Cxhu*x8XNh&gKKF zcGasZS8>C*F=4KyVs{N2D~o9tmqh=R?68-1wZE1bCG*%;yPM8&IsCQn?|38kEw6cOA+r^&xB`88eN|SU)4^GH=knJ@y*j7CQgrINIsq{}h z6IE!$)4F&a!zjY%nGQ5jkOX9^f*R2tk6sMuI0e`M9q>r7Vd4WMDum+q>^QY9Bs<3- zrAcM=oz&NVfDm@TKvO%Bneid2dU$Vv+`1C(S^r*GtC-z&#oV~^Wj+{xTUe3Nt|_Uk z-CG>>I880x%?TI=d#bAGme!c{*1R$CvV}X)Bi>JuI9)b|9M$!J+Z|30QPv#xiArls z45NS@Qp#Py6TA`ZE%e;P_m%d%XiAVJzEio>N0@GyYAoYO&yQ?ZgpI4OU@|WfBVo0l;lqn?-4m*eA z+V@9F$rIaNVO8a(kgI5A6UmD z`St4%LZNfqV6@=OrzI0^Yw_z_BbEsP@e1_I`Wlb>6Y6Ro39XA4k~%3+aL^(7Mha`+ zAYfDyZ|_U@w7W=j48$<&*u3v*Oc`#P-cf!V3HHG_ZHKRQ{GR$!cGeZ6%jOO00(_7+ zFHDIUke-E8)u?aCR1zSaRy9kPYUA>rLwRSxS^~{xg*H_sH7>j|XLF>q8UQv$Pqd>| zH(1K7bVs{GcfzIW4)MM{u|mfqMFp7FLYh6v6-Zxr9mSVw)mi5|)!Zj?2;3SOrP z^^KOd;>uc{{A?c2{plR`)zw9f&1;S+B~=~|SqTY9DYv%Y#OGm7VN?|_@|np0RthNNBiqrWRd-+>TzWV3G5s?f=_wIeC}YpP05!!+_kFZmRp z)U=c1JQc5$RCs11P|1uxjc#doAcUZkT)jnOgK< zd)<(HykH$So^doClats}6&cUaG#)qJW}|P0H4CW=c=^-Pw{aWceuz6)eZrTMb13)k zV^!tGuu|7P)6e(&7I9L$30%{Jrs9CYuV#m=3L?@nn~(NmqI*=%H|C(v(DN6x!qJncH9MjY#>9g*6*acv5j%vD?Q@oo<(ReRE6T8Cye;g=B2#4WjUtnBzv~4YpLV91#j@oVdwMo z)>jDI(Ik(=mJYUbcx3Z)+&$*zwDeI|C@^|bVD#MUa$g6k+8-&iwcyau^x?!p_kmlJ zDZMCcV?4dJ!i1hZ@M~Ju0dc>gVlF&jEWi4277@!%epfUE1V4QJ-<=*niK|N_5=|Ev z0fo{;I$?J8+N|4jaCvLzQv619Mz4;~QmMq;)g~Y;kJNHrPz-P4&v!eO+#6A?})!fH%J$_HGob08L z0E6->u~W+7r3BkikO)>b*Yb-c&vMgG$-J?#Llzvo{xb)4@5RuPGJXKD zKu*7V{E^)7#=9M+N_s+xp;*WLiaSl zda24>AAh|2JcbPJD4YM!LhLP1T`#5V`U3A8_URq}p0Urzus2)dx-P%F>@k*qwt-m0 zLcwZlbD;l(nVHqPiV4#v?3mUYC^!yDhwn zxMO~+4)EUws@4<3e<)^bE~zOcgh4R(N?PhqhECOvLV%DBAmh-56XVM4oDU&Xqzprr zbXVAJ)UJ7ZlS>A(uwgy1L^{cJy^5TWZF9$}nKZ^D zoZfpfLI`wK_n?;An?A#^Kx5KZj^2ozC3ao%;5Rc=+g(0eIxIh3PvG z&=QMt+T+ji&Pp%!cK5R%Gk@7;Ov~}g^=*ea3pQ}`J#P^W2=Xd(>Cvl*V~!fhBM)5k zzv#dK(j(h-_1g#h$KLV%c7xt)ndsg4i=UhrP18RXLXP7@%X2AKY((7A%>zQ9hmkgTbmu_@9W>;^&K5@ z43qoxM%6n$qnfJli$kX}tzTc(HPo}Jp{|3FoA|x8eFG=n{v?fU5kk>8SKj|B)lF?g zO^c{$(;SNO*Joz)?-iSf^$1chu$TcuyK%`6j^yr#F8W_|VAvO1Q!k~Qkn($UZ=39d z)Pc&mz+*y)!ac39FIQtBFh~f|Pk5}VbAUHHE%QcVrwAcR@w}O;sq&32@>hPtRo5#*I|9wc@DK zdp|TU_h|>c{zkTRF6l!5s!a~Pc5cf z?*E{O4hKyUK??m-$@d?W^%&N+yqlp43V;yojAYkUX*OGVX6q|_)ijqe#laK>=8kLa zBH0~nVsYaZ0Dd_Kxo_$MD&^%aaaFDl|=H z>HJI%Sz@33<<`?V=fnw=6y=bg=jXTshL9UH(%$`mp>fgC6Uj?dn*3q@98AT<@ymqM zB5?({s&{iN*ClFO#B7`EKYW*i%le^ZNqZC+e!2$>IHqDMMyIHSCNvH%Il$9X`d@?) ze)0c)qF%AGl?bqdcwzZU|#w)3*o=ijiMZ}^XVVcXZ%oXJ2UW=-f zA(FO6Lx=#yMfYY+4>{6iTjl&gdgzXKLY zDc?-YJ4Ohx8ZS3-qm(im31g4&1&s{tmv#j6_N;z-(3pN4G^XPvaa@;|<}K&Nc}t1L zEsh*Bh%2T~BA6gl&D*w(h?V~FmWoXDGLdP|Zq%{(aN9;s9dH0McPvSUdIPG?L1hD( zP~w&4{$$|A+_&-RB!gW5KOcG(Io~i6Bc-G*zJRE;4!_oqeE(EHV%eJ!B8aYiy(%RD zkrFMEQew6J4g&?V6VcT`B>HALKm^-bjBPC;+G^y*SDrtWE6<;roOWeHD-XZ*ISbZqqbN7X z#m7uw`h+3L`O9mwSFlI`%|!%l&|I(;h${@s>CRl=CR(i+F~>#~0z=WbbK(goLU2aU zV z4q+Oj^NBmnbXCrsEd>tnLG=P2-uQA_RT5-x-r8Wc;Yyd+*Q~}))gcq!J&oBGv$yQo zw|1mt_VSH%FV4rb?WD0$W1u+cw#Kv^wpBHO zJbD8!0`ug)4pjEZt&>u|1H98|+HEQSErIJ3zi*80eaq<4zl?2bcc=BH@dph~y5#E# z9;u4LxrdJB+(XA^onMxdgRUr^o=6%}9GH}QwL98MpPb>8`}1hFB2@Ydxpd%Ej_y7p z`O%(SI*xHgBUshC0aYjr&mV}smoC*^!Er;ZtvQ#Ny$LBDgwXIQIRr$+Q@4~3hCYnS zlE=Q0GXo%_EuUW?@-@% zKY4*;@#{mdoEE;QzKk|=8xhwf>S{oM&hCK3p=VAH z1ZFN6A&Ka8N!D$u+2b|-kKtRCJ;HzP9p7J$;pnp_^UgD$67S^12m})j%OrU33EfNb z`Q1gwb~!)SFldQqNfGG|q97r)FF-0TrV(LdTP=t6>cN7#EjX^h0fqhf@rXkikY5Ub zrl{OF^a^%IYuO&zMWw%--Z|aB)-k7|Ir=GU>#o8}&pN;#=TEa~Z+> zUw&h4mU#44I`d)L;nLOPYaykDlwq10FQuZ}SLEehPGfW)RiStAY27HxnSrL{QXiSi z%DSImxe${shR7!_(^Z@lRQl&0lKea7M44az2`igdFsF7WRwufEB4k!!w=!Qoicr|y z)<2~bLh$AztjDkIdUWqFYMq(NF=VB z@<66Oy8=SC$vvxQVI@*8mTU7-^=hW~NUsF&sRjoWrHdzh6YOZdlc=?hcG-U+V;F7| zsT7B}wSua;GwIXq({E&4q_jXNp7@&(7@bly*&u{Kk>DVaGD@iJL1IprnwH-or31Ls zxBiWi+^g6buc6J{MAVhIC>X-VC#-~V@rb>O&^f%@&oLBl)brM^7noc39;O?oE}n}^ zdU!n7h|?So;8h>CAFrfOP8p*Ldt*yCDRZ6*1;ONQUjF=^(~je;Cto0JS{*N>1R=q% z>l7IQ+Dt2rSLGTI&lmVx(hXfj+1FEa{$3#WKLM)!9sP!N=gQxl#>2P0ikcwjs+!7g zpS~)i67t*P!qMZnf8m0ZM@@PSB6ICD-d6j6zqoS|IROg?iGtv>rq!Hlj$;6&Y1^#~ zlv5e_b~-7J8)I$V)oD@*rHYMauS}E=dCpfX@&t-nK_K_iZ?qq;G&*~9q$^v~d8u&p z4c523M#S2LDlCReHqh#GGqn-2lnjAORHCAMD!IQ7+=|gp`~S9J#ixU zEm(k?6593ELqO3vZOTYKU9lBXBQHW==Yir%v=qPB@6)L)FW86c^gZKS+M4fUpt9ff zPSSlk#TEJ3wpab(^BaslXi(ZmIu^Hh;IY}vIP+c(JO0mHaqZK5wQSqh+E0)Ce9jo@ z@na+s*YAQarS`8AB$BY%@e)QIi%)9Tv8JV~uGRaFs9j6cuA@5iZ$967G;_9&qBis# zseOwKku41BU^!cztfH28bT+Z$HG=uVw0HD#VB07mY>HBM;$_+l0sA=@Ck)n zRj1JJ08 z=Kq(-FAU<-6hTTzm&>8>HMaR8*8}!`e!-%m_KpcjJx^MU2oCWmR)_fD_lGA zApDw^xE&I732Ne(1x{j+0Nw{a=2+am;;nCVj%nLaPO? zcFN~W7cM>YMvBxB`AUqTd1E-g|I-9DPx|2f>UU|enlYt`BC^$bfuTC&8gUfi#flW+ z^20&La#W?Ky>nFm0rV>@!j~dY7|=8h>D`AB<(_kSZ2y6rGi)SzmPSsSLUFZ5eoSZl zz#d$9_yN2$cbVtlcqfTsNfZ-oe@DX^Aq0M(#?%AT(`VVQlVtT58<{_I5p`9KU*{6< zJN7eMvtLDGq?Fux{sU}ZSDoa>SaF-Xul*;LeM;#+q6g1E|2a=RH4n#5`0uJ<{f26O zdee*a9$3n$r;O*rr5mYl4sp%-4;;MbXU6Zk( zlqNw1;!dE``k>>Ws}7n_$k#inde*nC^!P-HZRKbtrqzC9+PCX%Bxu-xjV4qcn{*8U ztz-0j_TbTU&&}h(2VA<~7T=debIMO9tOqMO{ zs0ei&oA=*e$p;^-WXADhDCl3vM~gNP@add!#2Ah{Xjs;WsHP|!(XT&mZ(N_?+|l(Y zp^I7DLJhKHM;fHbET5=2j)a6pypczx}`A;AodkbWgRD)rl5X&-P5qPO>3yBaKjpi04nQ1x8 zX+)AqXjTN|sd2!h!dICT4+U7-ycj#-jHyE6Q*7dHCtCyPszZV56(bDK@6QGO59>H5 zMd770&*q6Gi;28v>{m|A=2cbf+FGC4*wM6^`NSOh9x{Y~J@aX@{8|8tlVJL5f(_M;{PDTj z$>(r;Z8PtDv4$6aa{(p!nU2i3Zk0^v-kk;2JFwabUycH5=R7$GMWSly(>0+|?CYBF zX1f-z?6{kap|4Pdh9hly28VEN?^{rX2nO_{!GVhYWIFj?Ver@;Y!I5Xa1XznW6_{+iY)Xrpb=c5{_~1U!J46DdPFDZsM1I zS_I01%>3&W=z6B%bZm!(bJwwROC7xjmJ_OOkDo0*kI%zXCE!Jn{;`{9o?VfJsUl ztR*1h1PlIvFL*X)>|LacVOuM3tre&O3@I^FER6sHs0ct80@pYk5!>3uUcAm|*&??9 zs_@Y*=ZLgv0X2PaRqTt<7!MM6wdDJ52uVz;_!J#o(fH}$b2GLb&~<)z;ADPy;N-OF z3wG?_rIjm^I%Mqx;FmV9;)Y|V^5$o&xc#}0lFyGGk)7*FFnk*49zUL|&N_r_XsT`D zvwweu6}LEW`dIF~>|u6qsY@fPzyGg!44KfMBhTKqi1A*rAKIGz%86;JYeiKvGhD7K zsomMcXU~4#(YGc60VuY_)xgSA(w+j_aS27@{PVp<{Pe6tGqx*);2$TR#*MSzP2$a}+PF zn6g03w0Y;V)m$`VB0b9sGu9bURLsmX&*$`+&(UIf&V-C7`Y;(T0`O@X6Ur)C6WT^R zC7tP0G*0U?J}cJ?yl&4Q?EDFFwL13krl&Kc4I~OpZVnZ)HBp7waV|BE zN)OfX25-VU&9uRjFmUZ{D8`HeQJr;T3>`ntwYv1;0ryV@BU4)_em7A@t;6 z$KyG)M-K!B!Qc$KY1M3RekjptDS0pi^#K$Ik6?T2Q|OL~<-NUtXKQ-po=lNHox6BM z`7u1R?Fr0;K&c|2)E~ii1+G-k6bB(S3}q+bPJd~cnediy|T)igQcsLcBqA+X2 zM(%okHfFo{p_Ex|P2941|I;;DotRHvU&!58J%$hh*LC^FuV2P-T-;8r3gQuynGetU z?wlCd?`+L}#fcd>x)+w2WfbDq8Gqj~hLUSkRDK^=2((+5fCC3&3EVnxd*gsp0v z3@PozG6*Dsurx5GiiU%Z#E}Umf5NNq10BVn0W^JNm&ei$+4g78fguhuRoD(zevB~iXu6EoNnZeeZfHeT31 zoAVd^j@{8tl}24)@$RZFduB-~nY(JUmqrrc_{r|Uct?bpMbGGg-@HyuRTI&09LIKv zModiGMHCfgEUT&tk=C!rt=J>nJ^cQQ3>fjUdCG zt&ve^xd~}+O)jSy6TE>GiBxssm!AGGam=sim46wzzLBURfIy&%0Gd+Bu(H3>zvw!~ zRXj=G{4?m4b0|G>j;3G!`An{Slu>23WE|gt#gmyo|~6H!R@P_+b2F-#IjgjTifNge#@>#wA8`+9yV#)nJ;i#V~D%2dJHpW5s#W! zaT_ac5$Ld0b)|C!L{^4j{5Z#o-)1Zg|d0H_(;bI8a^STk( zfFFr2+DAwVLih;gcsdbMnu$uGEQr1W8F}S6L|;yu;>*7UL=Y*JjAv6M!nZ;cVcR>g zttA+WA5B>3!l58=C?$E9dD~G-7+82E6Dsaua^+u`SpFOO<)4D4WY^B=n=^<%jCzRk z`&@&Xl17peLY9G@hzz+%qE4I^E6ktPJeU1kQ}E34CH&#jIXGEVmjk-SnWIMH(^3MH zis#5=wJ!oj4oFw%l~VHgtW`Lclm79OpyXzz95q9u@5t|{Nv6b$Aa?cpFETLW!l~pJ z=b{-M&zL`GaO#yuP+cF&LRw5obvR`GSRe!?o%nYx(HLv0YiW&nk&x;A26Nr`g9&IF zxw?U&s!Z+Kmxqr$jU3(JpPT0svfDW$-p`f}zgqTMm*--+k*vofaf!RFl={1h7R!Wq zq=Kg8p(Nfo0hm&u+ELkMYv~^e5x^HX14SwFNd7XG{lgxgL-vMIHZt0AE{a-)k#jkS z?DinBQ-sG3g^ewUVp02d=y){ z%-X$@D+V2&^<4Vpj$(b=Vmd#bmHz&84<1ZV>qe`Us^jCuI&7(;E5-B(9ziVnOmbd3 zv_$gMi;AcizJtjxyaVj*#9A&u+G_v}HE=3={%wdv`LdJ}vMJ_5f^63TMITI{^gZIO zf5eTwn`H%oRPxzTeK2+>#otKmV|F)QMUT>FvaGjvs6u00(L}lj-eO0z3RenjsStJ2 zbsxYRrLJq+PS}o-qh%LFjM)}c}kViFWvcKH`x$imIUdG;Vn8YEwQrjO>-lkWx6pETL-$Ote(4px5pM?(lfZopu| zFwZe{T$ew*H-}f2uEbDPOvmBOaU;3q*u%JL%z<1md@P$=nkWzEP#)|UWX;^Mgt(ht zdM%M`Y^$Lm9-_pT*(|jpFcv3EBt{bf`s5x)zTTTRw=YJf4M)6ZwcgEVP+&M37On9` z8QM?^lDHK3ghUYq_;W77pL;E4=zgrwBOpTvr2;+w7WBZ09V@!62?O03Hn{0rBVYc{ zQH`-gTW$t7+n`Y5#9pPoWgU@}$hMTN)HeSTT`MQYzwc?~R1_M89NX=9zG_;G>r1AVb=_d|-Sz9)-qMmqf6VLOcK7t@Oc~f0LswH9pJN9EMns~k zDwgdK@avQn<#NfXlQXvCx~}IQ22&hA9fae6W1|RxW4rhQ2E`Qxoc`nKU+0qS75kg5 z+3%bfAq3M-I)G^>b%>w2u4LAGD>9dbWQ~0=pmW6N0aV2!tlnOOuBy0F@{=>Ba^Ntp zB4}4jE7!a-lW0n=a$|i1|MRbxne(H|(G-ONg~b{B$zFg5NSW-{lMTmqyPWOeH3?3P zg^Gr%Xq?dF2MD21<{v`X+K?zg_K0a3F1nIS|H2z+ZT%S?0#O}F9Tl)p@%c}otoZLZ z=9ieQ_d|RpB02goaiXsyon81!{+YbCQXdD_;>?59Gre_1>x~vT4J$q`k$UQY04#em zPSSx%W%b92#^z($JJGdo-YNI00hjT+)qf(CY@GMaJpD~gjUB;q1zevunwCgP^MgEb z#%Ty4xZ>zZtlD0~qK!L~xn}YtY7`25J`Nt$k1J0-h;AjB+Au;0y7wt%=jP0&tHE3! z7vFe33*TEtdG8WVyy_VG4F7IB2zM6Y`!Dv6f|6V=`_-vj_N!Bq)8@`u)9$OJA@%R> zWd%$;WDxHx+kk1?xUQsoSpj$d_$1WCi1?v-U+~(pRhW*0nGi&KZP_YHbA$X~>STht z!N}4~K2(>s2D(gT1hitt7GK7)#wT%P9FSCq}{GX2GZS)+59tv$U*Zb2(-pQWX~@I z5x(?6pAZ5)a4cr%VVtZ^jE-Y%B%Gx~wv;8#PrGO&Dg>N__Tfz5{Z$e3{oH%oY#w>>SJl+ zPIiWy7+TPcf%zSLo8mwaD^Xf4wZ9+uUK0kE?W`2Z}_Ve&}_a^h9TWS#$=;J<03?*vUvVfOf`k}#lO89%-J zD6VynVAIY90zQpi73n0`r@vg1!M}=HHcx-Cgda@J99_3#MvpN(v-vYjw_{xa^vNlu z)Ss#8)MmA^u<1)8b^#-be@0t;A(k7WZ{Csg%R351Xt>e_AyKoOmn$tA!!IE5!U961 z440%t5q?T@PA9095ve;A+%_OvG&hFT_Sd8nBeb5#-~|wUHYLg%0AnopvO65AZ!(Ux zf=p#!E6K&nc4sWRYtG}o;-ns@@y4#XfQ_MKPsRIHgXxtMiP$Ee)o#X>k}2i=$xY!f zRp#YpRf0D(l>x;a1LWa-D)`_}SMu5Ft^E1fkEv~I#gr~4u&J(zC*NPl@zb)4rZ(2L zq5Jeq{-_Xa+fc*ZGru=cZ4djIt@%IU#OS(8&)&uC*p@|oI8l&fnKm!JwveA(a}*WD zc~lf2Eo_GCQmD&DQiwyA>coX?j?W2wN_(B zUk4dQLG3~dz(LyEpxuelheGK)$neu(tp;NxB6uM}&FsP~r9`?_XvPR^b0rC3J%Cx9hZvB2To&t?Rs{E8;Lk30aZs6DtC-KGh^9REh~9o(|t%Ov89c! z7>p?x%N6~9mF(XJpa_jJ|0rr>%Sm+u1PVQK4n{g}WIYxk1cOQ+pdfHK@usW6Z6rI- z#e?glx402`H-hgVWaLGVQAFS*@E?!J(xg(lAyl;zMd*n>YVG?8AcT%0TQk<_fI}$y z0XhDovNnX=7et*7$4m$)^hIfkccg3liot=!Lz!K*jURsY2C7g{g@TLZ_5)7gxSpQF zarUTDJhAwzG&>`J9NpmHKIvp=T~kq9sHsgqHz}c}zKxF-ZQ!_rGFxCsSEA`^)*_ge zh7eN9?^G8iedp|Vwr0N~_U;KkzJ7X|E=N1<&DE9UbzQM~C-atW=9zaFGH>bTbk$X0 zO5ffD^~{O{SGtVrS(!0C73cMv!i!UX#MOh2q zI+O>(Nz4k3krDXAct$i(PZD1z+`E!_I(+dTN%r`-DJ+njmJQ#7ZPmrWlw zm|lei_)-X!0bOIp(82UC*?Y90o4=4A!D2K;r6$&zl@GJ1>9eE~WIH619g!_Wo%D~G z5DF73?qz)W-Q?;0QH2dM%9_UCu%hOCqOL+*YDA<)#MJ;lMY$K_(+4HiH)y|w?G8+; zBjf#d4$3!3>6mSQ!L?R{jNnx>nFu7hZxn&T+tH2T6c_#x1wUFsl&Zb@JuU@tsQ@mH zY-4-lO<#Ne?Kq}l8aY}JH8FY<0`hc?(|S&$#OTJn+6}mg{C@jSRV6Xk;^nO!#=OAq z=h;)wZ@x}Bj z%C4!?8tS6Z0xwdN4XUbe%cGa0XHoFAEr&NA`vdbU zqpy)!itXUU=Oy|iL0;~8^yoGZ-8dLW3PMsPEVVSTDUgKXUv^#cy%*%_Is9q#EgW4j zjWS;eeS^KYy#IMz(C2jC-?fsclZ~$ft^+l#CU0(8l_XXIROaSkC@O!RJD;Dt`8G3G ztx1~?(7U1#wTsly$bOkot9JMU2G`tr5k>W?!k)}pTjXj2hy*k zh+~EhcbtD^`vT6G_gDVD{ypwr_a0}@zlRsLraMVu4-EpkvOtECgA%)k zTv4BN_`P@BPOjHJy<5?rvCxu^lIA~|J?QuXeiY*n?D!YRECk%H;F@2MmFXg*IFXmU z;}U-SC9mKwdK4)Pw1iBL(uo2nWrViKH;+P0DLs%Ihx~})d4nl6iaD_O0RAxYCVB?D z^JU%EjH7JdcO(XHZCS_arVbmcDhdNjN?2W8%TM2YhhKj%o7FX$r2m6P_3iR_PMX?v zBjHg84PyM^!?VuQG?kTKeDmn{p0GdKn*EX!6K;$0<*Zehaf|VX4x==MOHf!CoSn&^$(xfEsD#{9ejz9nB#9IG^BLpfou5=M10Jxx3 z;DrB%6Y+sG(fr3_6#hADNH-H*5vkKd&YISfK;Sr<&BY+I=ldlR+YOZNW`nyvxI@7x2ol7ZB78-dw$w zo8S8o%W-g|?oWRTQ1C$m0$kVR-S%j z4lPY#Mvv)*bR5Vcs#jQ=z1ik>g8fVi{QnBJt*>U);;j^y7>!XhzMX#PE z+;!L6?A+0a5YTT>1?!`&NXy0wbciCx9fxP$Tg06|?1)AT?N-L$Νp&5ez0X>KOi ztkb)&u*>7Ajy7}e+IKKrJ5BoK{`K#3aM@6L<#-J$2j&h+MtM@I=bI_Fn4yd@7g8Z9ttrfV=rUcR7!XqRG01a0{wCM`kV>yAMzoWNT zD(!J+T^DSmHoz4^LG?`~s+40p+q+yRZhpOiLfDQlukKT}gtyWs*qdqP2lJ~DKTBK2 zm0o1X?V>u&aa5MnRdf3L|Ku0r4(H+_6R;eIi(Y({mDROGEDJ>lUR|+@UmbNAXOHP9 z(LVj~F+BVJV#3kXbA_RODi}MsXQ%x*JpYe*d@^el!5kljnu4M#X>A6)4M0J0W-0!6 zhW*Ue>~~Jg;<;=1{bi4%C;}-Zj^lE|rH5hb8jW>r95{6dhfN>JO*b6R4c8t^JZ6)d z>to85e?vObb6`?lZb!lixBY|eJqkHu^Z?d0)-!AEMoh;gr|jwyY6 z^ZdGHys~Kpst}w#XbdOx8^W@t?HLcT={mf%b7`6rb6n+VtZ!RIUAzX>u8QJ8xLM&LxO*vANAplCx7vK1@(8cy_0z)Vyk zfgO7n5$c1Hcimp^gXTXLJNi25&Wu3O$0nSl4ip4Wrn==0Pg_BD9vy)zG#VoB&@C^s z_ejPWO36RAJj}A@CBzebU_Py%#|hm}=D5le(w05B*Ers;T1KpMG)91Jc^{*QZLy`b zk*C)#=8_>3x%K^7X;mW;%VI-)1Fx)D#rfmLB&XkY(J@pMeblur8L;$|uX9T8VWC>+&&7_)Y*!A_wNNC}SJp3aAuV{^w>vpKQPaOQ2=#)|4% zqKR^AS4v{G&F^N-VaCwGC%u9z9=G_P z-m3E>~FSazj9)t;W&S|{BhzbUa9oyy!P@U&*O9@ zUwpETNAG!$Q!YG&<4+k!QNrUqWK4fPUbcyh(4>Is+BUX_+xWD42c#(@N~8mpA+hW@ zm%sWt`Q3G*(jw+~AG-C;^~~C_p-YX1P^WZ~Ukz~M&_A%IbvZku+bA(g8B;hBUCADb z)kF}1mP{Q<1P%gKc5e14(MWxPA~du=Sa8FB4Hn??qJNZ`S-Tf z9FX5D>*zIV+q}JYU78c4Ybr;K8_34(b!^;GPl4aa=(W?*9O-aiz_TEZgAB(JikeK; zDgrL}*$l=XnptV^o#4M|qpyt*Us=TGAFjlVbvVG$aGb~P_;;dLkl=$?7BXPc0M5MP zFb+CsAgZb`WlTRz*TD&rLHw2y)>YN9zN(J57OhE}=Ziwjl{mSuc-Jn*KlmWJqGG!) zj+AU{XyB8rTRHQ<0fa1p?R2DzBoMI_c7~hNwmB?+ICqWy3yT`RB;rIEo$QR%9kw$q>@PjJ%3J??+kpVOtw< z+w;ybe?oOdW&Z9hyt8Iw(t$~z zudr}O6-Ny0pPbgKqL9NU45FbeOw6ubakWsEgVZ7kie zi@Sex5|c;v1>o;@znflME+tBH2s(v#b)U|lG3n)%-#PZP(b4{ncb@y4cr^1P)Bbo! zDRFGcx=%N7duyLiOgc39H)-Ph6}Om_1@r3TIA>doM739IMLUTvIW(5B#LoR)(SBp z1f_*H6SvnBiG7eUUr~xsmD03ntyUc?o97aDBMi)&K+nK1gb=K3S)GxJmPppLuFEJ> zE;NG7JmMM_)NNr~sDXLAw(@!-z|HO4$$PJ!EUhp)8HY}m4nGVIs)I~94ZxIbYgx3_(b0BQs`MycC#=pc(Xx}IzK%Q?r`6;8fh~2_7TdpHwH)F{d5_osl-{_Td5JL1u zLBZ|xrU*or(X8&lLeA(j5h*3lty#t9)`m3W5zuu0XYc`x>{dZxz@MS08PK(~C=9?~ zpP!R7D(#4yHg`Y&F|Yjb5~Eu^ z{T*gJGXp)MLC zsA!mOXGdI67t0)g>?@geYeo|;A=ki>I=T`E6vD2K+3B;O2!mq( zcw9G3OZa)hu@BMI3JP;CMK`8oKOqrB&mwBq*l-D!wGxs-nb6kuXEc2%{=94V`FzR> ze#MUX=LtJw0|FuZlo#ER=6tMfna{tg{)#IdTpT{De+f+(9A17gs!%g3Nr7Y~WM?Qs zVM_S`rj!q0WI-=J-nlgyv2TY2rd9T*m>fP_zmaGsZk$h3nb@;~+`YBBkvU&&Otv}! zuxy8hmN4&qv6hpkj^yW8J%!b&;j9Q}NvBB&wMJOIXbU3`9MHv;`}Wx1Y|VaEoizTi zVYqh6&63dN8=siGwY8O9wJk&wVPTP2Pm6Yx99#MAbhETo5qA0n(FI}2P*nQoR*+4l zz;qod1BH7%w`{Q7C~KQv#E~&Y#NKLMT$ix|=wg&v`E?UnOpM@+N?UPM3qx}(0p6?bGC-(bYVm! zV+zOab^nhZb~eFgztC&$q7sVWy8Fpfj} zdtvyU_08Pz(rngl%d8{-*tWX~$8~w~@wwQY`=g{bA$43ZEr^?7Syb=nDr)u}LGI!A zUu59$lMl$M>5~mo2x2jdnKPFVkK0(5$RZyvQJy#(L4{`e=0nw75P@R8l_+st8aLm`uLjpY65B z($I1T(fB;tBCoNt?sP)tSLnf0A+4Eh097B4>YJ8!47R-u%j#%X&<-i1G&TJM(_FUC z=VRyt=v^_3{G4+!w4vnqr_;0aWpV=@DJav4@>bP7*wVyJ*{3Z{X4gDHznqe^4%Psg z(3o6u;9kYzMwL`@{pe}<6qTT+lcO2rXaoog{661{;N}Y{10znTMn-44P81n+2}m@M zB+)@sB4P?MFEbjWDhfA`n!!b%KStC^FCUahZd>*m_f9x}uMfT}9oDukW^L<2f?6I! zat9;G_Q4j|(wi|#FbZKt6lSu3%DoS72CpAn(=#=N3jn8)gsxZjY`)6F2 zb#0%f85Qtuk0K>;H_SmLhx2*EVich!6@jlA_!InU*PU)Rwbm1|;tVUS;L}ro$o$<~ zsEdV}P}Y-v`Nb@%s^aq3UiV7Tg~Y;P#=yb+{LpE1%k!dWtF~8j`F*coT3!RbY$J}` zpbu41sIO~H8kOYWTGJ8_r{hIs+TlvJuiC+u01-j%zBN_!8@Bfn+pi6||B>zXcbs(5 zA81tsTbNg~fvFXPy52|v5xa@CE$`A~Zl{}n1l6&1*j+S#jZ2LH7D!Np zMByP+fqY;9^i`MlQ^Z{2tL`pg`lQ$SoZ<=RgK?% zIGdN2twdE6bX7%D6n=5oA)Gd91iI=yCwDzNo3}n&O|)}FFnMk&-AnV?v9&(YxHoZ# z?Ka0}K-|iBoNUwK)kkJ?$noR8AJyLd(AMme6C;F}4*U&oSpT?`^3E=nI~{lhm?))O z^i2-v#7m|zddeVXz5FHdsL7xM`ttN6b7^i0V_RMb(g|p|K6S4lTnULb4_z2;7Et3Hvf*D z(G8ex3@J|cfj$S*t?^rRou3@{Pj0zGt;m2`SG&uW5-W$q>Up@ zHnq(~prHvLb`}`~R|*=PTqIr-+B{WqaF6HHmI+1QN$NwTojVGqwTeJs26|vdm&XL4 z>qFQRp(5tS+0bw!d1f^gIY$%JyB6tmrOo2H+u0s^7eyGj(xgxBaU4)~Gm6mI6k5VN zyB;F$L^Ez#yH8-^aAuD~2xvXo8Qn#9e>Vo?W)8aA@ygaE+`jBR6ro^ChyHn`Jao{x zoZi2qWcuwj>v(0wD&kJZKt=$Mez};l5~bDKs+)Lo{;ISTQfFw2!fA(&rm?Ce`)4j- zKq!jL5|P=qu9_8}Z)Eh8y*s<#2>Y3>*{c&Hyhaxf07nBmfrUbdH>H%TJ1v(7`~>*o z+a6S(A(dQqYe#wA5vPo2$rqdX$3wH(Sk-`|D#(VB@{Oyk#i#XSK*cMlN>0XJ!&Vhv)ZB=ZV(dcJ4nEm+1=D-K&GXx?r`g<1$DK&IXOcv7 zM}?iSU2F=kWKzkLtY>Gr4s&bPvbn8}XErXPE)q>MjkhfRHv{rZ8D7|pjZFt_Q=)EmrzI6q zfE`d#9MGd+>STQC+Sl?PtNiayvvH9!j zSsqs}b8m#*IxhU;ceL3NhUE3($^j=aAUFME|55dwX#)jY#y~0vp{4063B2ysE=p3Q z(?O-hQD}~DqR`jDt5DTq$PzS=*eFT?0Ijk4)VBQt+o`7@a17;n7p2)KAp+zV{Fc_1 z+mM~o*Y&newHrj@P~CJd8=LQAaMAVjDYzi(bzj~1C_6&$<4O}ZAtA9n^gaQ#oczFH zytwU;*fQQFU&2kyd2`o;{A}pc^v}&+)m0m7wK4~TwV@Cj>KhnbR?5P)+q-7Hawq{=uM&4`?1ZOVkr%{N9n?-yV@0^+8gX{Vyb|Jw6X%hQZ&4F# z;?e=f@6E52n47%{9m3x)H!6X04zt`T;#l9waZ&(NS4ZyHQy75iLzW z!-{`|i;F2Gtx`$j;_re|rh>T$6XDWdk# zZNlk9(BgRMB1hVMQuiL?iZY#;hIohz=08f487AtONC~XTAL(FO@vHIZws&j3a(Ei z?|~Kr85Bv$cvniUKY1$WP8*kXf6uO8#=T4Ddv1gf-1_+ksKIEv5O*5#d@*wUacbk9 zgR-=FJ=ZU}kG~&qJ%*|?t9l_VW+RH+nz^!Y(S<=tU%hT1!Ld1%_3=o94*mx*fFCGeug8(Lo^)Vdp6nt;H? zOH-%9rDGT+k!%co#jyO$&f@RvT*=LgX99E*A8dE{B-_JHEUw+b&p-H(nsA6{t4%i5 zc14J!K2<@EKYO!S>9LW9bT8$AgwUN3f*=3#1g^W}@yx|FHS2)M@avRyFUm6KJHUQu zYxafM+XL8<^7qcfv?PQO2LQdKlyC28Lqdovg%FE`5R2;S_DwgWGe(X~A9V;oZ*$N> zE>gHy0f^^#Mn@!`I^jTWo8U)>AH+|OJ%sx&IEfd3aBfxy2H@)1Z_pTv5VA~Sj!o3D zscH@3#{C(~`qa!3e$2IaZRdP$TzVT%Zhe+nwR6}N%P#wGx-mX$T1AKhn9}V*PU-s+ zS{ErWTUxAddX`<`x!7(LM@DIOi*TgCl^(BDO2`RbjNceVUCUj#ZW|rlr@cJ(?#5qa zG)hy{QVNP6AWDFklsFQcG&-~gNt~ugW&{Et>eM2$4A&$sE#@}dc6oCF0*Nc3&GFG} z2dT62lA|N<8fH&9-Ldj5YGN%!95c!L76Npw<36}jvbm+62fzB#b3{e<=9ocUXU5>c ziDMF+Jb45@J(JrY90b#sU+!VK(}fUH2r=Q?A9M!x8(ULanun?hufnmta!CrHCe_Q$~amOL**hFkkpDGg4>D-o0Aj$K` z>HO_8U5k6yzsQvf@8tRIAM$qf=Qv$7L5#Z*Hb-{RH}51a8t@Vam0g69RTNHA6`s%5 zwhw5Ctwu^o$knKG3|vXV-UuuV%8MVvaogC{a#vb(ManpqTT5NrzcRM&^G%|l@Q;WD z8MxIovb>#ul+;9KWtl9o-0aH3mK&lzo~r08a1+{FE&>|`w^Lh`xRYsXTFfXnFP=%< zwTZiSry7%vqg92%gG&~$qqT*wWqE_?EN7z%g<<7h^UtZH`)4I@12~Rk@!Fj%T-|XF zgNAj-=hHLOI6fndCj*`xGkiKzrjO3L{qF?-uk}t69zo)-zzus_R!aG(lyZWUaza_z zp1Zkrg~?e;#{{6*l0a0Tr~=gi)dt1(rVA;tOvw?$25|8qexJC7SaHOr3QTvv#jYQ_XDOma2A~pkLMTJN}NINCKornsEv0I+J?#)&42?MUMs2Mh_m4ChQ8OSht~TTQ&NX;ntwF7T=$g`GaYQWZCAb)<`Ab?X-5u7xcf9H~-g z=4RL$DIw;n#GDYzn`iU%)*A_^5^j@$|kTOn7G)zq`k8kJ5S%1KE_gB zjUjwlx_1g7SmXzXq+w|&dkB3+H~Mg5slHZ<#} z-ZLmp(wM|tgI3$`r9VXbODj=DhZAH=lMkz(W^my}X$*evne~jR#KD$f92sR}crH=t z>B=~$*r=Fjgq$YY;;U%1Z5rJGEv}yiH%Qpkz3oaDMagblea2WS%4gG}49Ag5MkRs3 zM^P}lZcLBdqsY_yqKXcqq>2E&b7wF(|5yt3Zm204OhqVk^C{@UKuJ7`hA=49OE|Rr zcud#k+QpA>>+)yV7~X{|9ddjY>K@(}pQ2G2$Y)4y_q6*WHE<13bQs&c8}}YRgGWz2 zC9PkUPuIBjij(nM3Q8Q*7^r54+4K9-WBdh$K@3%R)p4#1wu6Y8D2{}jAXCnm%)=kw z!i%f!;>1f2?edtvHRN7?dq*;jBfai@?P;9OXiwvG-Fo-iVZX99T~&@qAqZCFW_|D5 zgEspD({;NXwV%{$3}rrVAGw;3O9o6uSF%??iB4vGNlP!njt^IAsKP)>u%$vQp~CB8 z6SkY!5?M*HZ%DQ?1ibW&En_sBn+Q1;?a5_{f-4n}akjO-MBLhjB%6kfKv%O1uL&W@ zFZd-!!TrRfM%Dq75R?ax%DVoz(qEtofAZV~3aaokuJlF@t2i&k#(K~(DAFs?MSC<| zpbH;|l$}AoUc%~@9h~yXpLllbJQ~b4RK+FeGwJ*SrQX|ss&nn|!w{*p^8#GIMEWGC zl8eTVv)o6bLqfU0&l0f~8y_}MqDPTCvF%kojONGj{PN<1LwtVkySTu>E3USLlizlZ+$#T5GdMD26~=2tb& z95|96kDY?$Iw(S*3Wc?;J6PJZk$8&Nt_c+vi)Pb{O88Wpd_7LHokPUQ;j+Fb@nOvq zkfqUMU3>?H{^8j8P-PrJdk!~yf9yCqMq0VDzyUTi{+hV-M4;NAt%%f7aLLj7Aw<{L zVpaXmXg0SXFenf)uP6ghgb$xSkcxtzWL-}YI@5amgDovDvm^8_suG}o{+aa5osJNK zv-{l5oSH{j*E$y$hXJ{hIK1K-!cL4&>p!5;tY=u>SSA)v!>9UrdHZ};x9-4gXXYd> zfZwnP#}i70fPx@j&!sp}%=DhaIDfzdcDAyh)&NheXBlUusMIwd5Eq^;7We*&g?-F|%1(gaRQs2i% zrIe4R{GBb0Gb!cov^0(o;ydS}B(PuEn#yj4?AVbhKcM(kB4*d4^zLN^`?UQTeTVSU zh7}pM#*r@9k2#D|{|qeG@fw{9fwQ)OxpiNq*&2b!5QdDqI_IJ^8s@U25>n_evlIK69%kz)fyIIq4 z6RtE-5jaU_OrQuKgNlCIWrb3Se-uace1gGygH9U9k!PeUqJN+G7R27~h4Z#oCq_zX3n6|4ypLD* z_=J>ll@Q{0z#=K-oBMta-yY+}_MxyaNIY((^|H}4g*H12dlT=K9MY$pisJO>!@_Ml znYnBYu5>wk$N+YRT3EV!H-{JZ=Z(lJ97Pg$9Di z+E=NKHz%r@926mGHwJM>qg6tY&tgFCAf}WZ&%FBQh-cw_^Igp(SPfaQuPXzv#NFt~~jmjAi=`u3+Pu-H8JekWh*4Gob80=D@&S$#(Tp%5NO4 zNJ-;d{7tv9FWBE~&3R`}X4&%X#N!>$pkb)=9$dkOmgcPMD2m{mW6~QihE0>1%U82- zTNQna3z^Wn2eUV9p(Pq)Vr~z6d!bn+W z`9%zwwe5g+OuL(Yjg&FU^RLKQq9*i4lAEUsi7Fjj1VR`LD*7qC3NFl8Cl&cde9y2kUKOjvOcol5JncHJ62WEMjGOokt89&&|FO08z>p7%GrG< z@XOKZJ)M3rWg1`ZuA(~BM%1zh=sLNE!5xQZcDzqT&~I?>^`~*ey>EEmG69a`a@*y{ z(xbdEZP_8Cd!~1*R#i|{`~icBQ?o_WzCU~mV(<4dSlTC@7%AmDz&o9$-PXmjhwaNY z_Z8JU8mO&mqW7R~6qe+oD1yJ;eJ;Ph&lfLUH$p4nDaU7QotE!N|;Njg-HI*@=dvW9S$MN?EXAz5egQi2K4CcHG4&wY< zp2@nM0X<7NY(htSiN2@HKN=D$~;8%08+<1yT5NL{z z5&6?F)Bw{e|HLP|enQ7X)G+{ShbB}^sZ$j%BINqfg-xa2OrGAHmCR3K3lfX!<~u3M zITc+^?`dQ?4M;koABynW#&}Wgudl9sFzZ^se7~&(383 zGAlxoV>q1GcN!=39L3;*%v4Q*-_LvJUc|@SHnX}Z99w`(lgokv9`9A4?kQ<^PA(n_1Y+hMUK(`5I5tj8S2GQ@h>HTRX18ktUW@(ee5xNePP3=$$)*O>OTImmT!NuvCaUI(;>X?KY6mzM#R}MzPVGez}uTx^QkD+5G!-2L_PR z=3m?YPXF8?-IhxKLM(|m%pL~+K9E^M_O-EJrE&h#Sozgfet-S*n2BCl zx}ou_znsgc@qIb+@ZlUbZ76m1ZRF+p$;t7txvIevz)Nj2k$5iia{_52|NB2#KwT)D zY*FDP9^m%(Fk)G(udZkMkU?BEIeSWAWz#M${N!n3u8k`tTiWWGTfKpbd>=y1jOKRp z7xTkGX8^FFbq5{cTJt>c)vZufs?vU{?kw(>wuRn9K_;Rx9gd7D z{PfP9!K}J@iFpD`YHW#@Q>c}b%Q|G1hOah|DM$}LA7#+dHivMw*BWbUGk{Cn<7 zTq&6`bp%HqI5=5)y)iY+PUY_ zEjRsRIcN8JHn!F!9To4$14yZ6nXxs{^yH zuAU2DcqQ3{tsxTO@1HK9Arj%H!!wDtOE*;UqkCS(vK<^(vSQP2UYWCuXKp>8p5=x7 z^IxB`dv`NdB5fIu+r;BG|9bgL{_)sldpwr!8{bBNYJW@P{3p@Y8s*j>K0_?h@eW19 zCcpaOGd%PD4dfLBQB{R*6&}xJ-MZb}{g-#}?~GyPC|G$~!e)4;JrYZ=_W5w_#>^gD zonX5z_s;o@nzj(XntpgjVZoitKOo|yIPyT;b=cLa)1#9dh>F6vB2PA8TX;9OuDGAL z>m8FVRchj8gxv`g`_i3(VMQ0QGyE}*jG-jnG$}z9IULvPpUkRxF--yiJi)jXQ6B?~ z#}YEXNcp%WIBdzk1&YWlN2DsPd#XfYx+e4LKF!$P(cSOH zCI>0KeoAewpDL@ES+(y{6dy@qXb)+8O6 zRP3y3q^T*pZpD8L-%`6lO8JhIa-#oNbfx6AWvlsO+s=%6 z%bTjQtO2&GbZ#4mU-fZLZ;y-d@{aehI+Y2VuEngH1+-b|&uf8i03%AS!|@u)Atabm zqt5X&zwSwPMAme1BEWU*T-vkQ%Qy5zM zCkB`NftVBK#U1xyOA||)IM`UyYdBq(-_u> znmVrSbSp8( z2RP_T$H!F_KJxVLY-*iJvy+F4L$Ml8)+q^~&DAjJT0bM-H=2><_hd}p*fu{YTGtL2 zhs~kSkW!|NB6|e;XDuQ1TF|Hr&p#kz9wbzGKYc7sw9AUA%x|toI_=dWf~aHhYCFIQjK}f#)8i_ zbK1Jc2*)xVA8%QY(7F7JO?>?35~P$IaqZJ9aei?mNqwGiMF)PG#3Zf*x`)b0szoFTJsdk^QYO|TZlYj702Z4=3-BugX`wyc((z0?(!rJyc3S~<$ewglTj;Gfro}|I z+x-|OMTSrxcomJ6d1q0XcUIc;b!|&Mu1crFx24IFrcdabldcmpBySY!+ty^|0e16u zPb(xy#RF^R@bspHqah*gc<1-#=5?4&9p%`o8*2H&9f{ z1s%{pYwc*^;7ZBrnoQ?v(W>of->Y`SO`D(m=~eob6cX#A`c2bSBB2Ja4%~em*+7cz~ zIyk}1W{?6e>?QHM=6`>(lJcTlE;`{rR`0BZn54adv|{3Qvi=&J)suRU;oj9Br%m%5 zFHae=&Fmn?+`DEjFKzpZ5d}Rspg6m3f*}lg=k&^$-fYz*x0{sZmJo4mK5txxAc!Q0 zUoq)!G@}ZMf=d7V>8x#j0r2{v)H!(=!a>2sbTz`#&lmL%(L49+mgZ%Df*omd&$^pA zp~ra)&KpHdtdlNC~E;k?7`0su8tZ zht-WWjC<}rd~sj81Cv-(>LBACMe3I0&^lLrp^)7!o54N{amv zDhwIYjTi2p#d|M)K|JOepE++W%4H^Wt0@uxT z>pzkER)3uJ&5K$-qAeYA31`oJfZ@eCOfDNlg}<1c(VC2bgyq^)_%l-+O{P~klG?M) zl@N6mq`eJQ$b{CKH<;3d!La;;P=rW!i>`JGsCRNHP@}j~(Ch|Kh0Er)j~QGreV;Ew zwn&6`Y?4KdvvH)&ah3mrUk$P|vW5P+gUHkJX|)e4btRi>8&L{0 z0u7yHKKwd|PaDEjN8d?XbCh`0ME7aD_ShWmcBI;jxb}}{ z@Y}1N#4>GMN22K}zJS3cKR=!)|M3Y8^&uR`MQGVax)N|xA}cY|=DGJ5bKY^|`Pp-C zW1EQ#Z-P4$a8*K@-LZ~L+$;%wOA2X=$0_i4v>os%8n&e()$|dTZ7T>=RA2jaiNs-d zw2A6i6Q9;=^vl=Kgp*7u2?V;LGp+34jPt1QXGSv|sbEVjv2E`Jqpm@XS;WxXTnc?YY4Y6rhW@!z*w{87nM$CO5OEFatS#KX;g3X}d|W_R z3>;}Oxp)fo<|bOLCW`bjrguAsk%i;4Za?anEN@N^sdp8XOU?&c?HD5B>EdJ{Ws+ah zIBn!`^8M*8B(_#Jb?nEHX%0*(?0|w1?s$t;p>Y10leqorr!iw4MkgLM`QYV+9DBxr z3>e;h&nNI*;J=ZWZ-k+vdNO{>AQpeNA?e8Y13Kd-51{|hN`8LYeQDf;EQLE>BuM2@zG(ayDGgLb2P8o?m+^U1Sl7A& z7s=qf5uDlcLh_BmtlN37YE>6|7f3rxSRU_wZSA7G9syF5Ns6Lh)3|0z<{)wFt_I@f z9)wqgz}3N3g(sYuz@S0hSTlDmO?4rhL|WU5+gNd%zufQ~kIw2+0QbAaer9X-OD6_k z+#!Rx=hbU@_3=6E++4@-@qIb%vO_8DR>1su>%37BAZ+Ok7Zj4wnD%Sz8tD&VTR6`0 z?bS)ngMj4FJ8C!x6Hfu#?Fg9(psKl*8BhL`H!u1hN`hWka&>bJK2@j9a?*&iAg~?q z;n1$)qKJ+_QOvQ~+FHPr@*JvSn{cIzC3QyUPec`&Eb|I~Kl>r#T6L7|$9lqQyp zCVHtT)R=h)Anf=P&Y8e=JhI+`##Q`az*USZ8oSqhS+3Wy$#n&eD-tbh_QCa+BJ1}B zNTvaE4TB5EkLB8f)Ae4Yl-&5ho7i^Znd)kHBGF4H1-h>C$p9pkTe>}aqrMbK!0{=h=1NS{TiWPCqs%Cw zF5Zw(>kcp~e`MAzOe{H<^=)&o+!!EnrASsIbw=;tKw{=xw!~uzrzaZ|#)BsO6m%h4 z*BQ&2mou~K-_)B8bo2Em>}*0x7Z)d?JnY3&RUxsZ0ysFX_wJc)96P*>-;a6_EsNwr zXGClhS4wPKLAnA(l_+XLT2z%t`|G8G6$HBQBrFD%m2lfJM=-o1J@USC!!BN*wJd#P z)z_xAmRPSiWYHN0HtEZ(gL_zVMHL$)l_Cs5<-#Re>{f1WZv%53Nr+P}; z0RiZ3Hg-@&>I#agp!rooINk+!WOAf4Y+yx#(3^#)p?VP!B`&?IlJZ=cjzeQ4!lPd- zOjfm&`*X2eFT&9|QlO+H4O80+v?J=+d|KOzuH<7$FEm>nSmTNQ+UVhCw*kQbQt8@X;slWLm5y3iP6k zZMCD#YT>T6HxYHhxVW_1(5U1xGOrICL(2#`K3ePm?a3NPW*R-yHF>xC{e5y`@-#ov z@pWnup`s|RM>_0O{+Yb?wg62cZAjHNRuKGT(qw)(VKO@#TiDUqLS1W^%EFFx(zfa* z8d^Gtj|HU_hswy0L!dQ+xVQkuT z!_cz34vELl=hKL3VEIJaBkwmf#*gg9&Z-T2-f>C+K|7+B%?nGHbHRkM^eHL?phvJU zp=jH6M5izip)i=mCJ2R{;Rq4OPdmru*DIgnX6H7LF>D=@lqcAxwCwC-_sMOGk0 z5Yvh!2T!S}k8P$vjV6yk6%s`{30chaqqL=upgt1QwJ?;f--mg1^ZDntXE77iRIOGk zekG3)1!b&j-9p4Mu%&`T;8QGewRnbYwOo_6ZOa&0kUi3yiXnNGM8ihf=RhbfzMzek zV_~&)sDgXwYMA6fR~2TRe<|gG99CAd=PXgvhjM8uw0i}8#g8Y?w?=6(1LQXRqTlOb4&%lqI3V`Gw`b_ijZCQ zQE{Y0TzXB(Bm!GX*0k;BwVj`4NEDjv zT%Kvdv(nxb?*~*qiNFSrfagWc`G~P#WdSF zi5z-rBqY=Hw_2^)FGMHIsoqH3Zr@VJPa+UPqG}zH6)$HfkkN$KsRF8`qe&V_TXI~# z!TjLqnVkB!e{sX}Z!_aBPxA8@-%aME9oJ2F92J49bvld0HnpgOupE@QjS{!fC5)cX zFYCDoMKEo~e@KW*?&0@ed?O4S-W`9IkA~l`^Sj@l&RcK)jIopZCsmr;A6L_M=+Ui^ z)5eS-kj2)tBa6SE*o@%wtsSY}^ZHLfMZ*ed#G8GXtvc8W{c=mVXuu?Np%T=y6G2Va zBI>k9N<7%o;f^&g6Sgy3V3qj#a8l3TbM@dixMaW+WNUV!3LT9wogXf*sjH;SHxhx0 zBQ-Wf2l8cWH||<@A3yr)Cwy3)S?LL03H1xxUrIVK;3dnWuEWaaCR*$Omei8AChqD) zoy?MHDJ5aME$coy*WbkkzhmrYw&wq&6LZeF2Qg-BAACLyzh5U1FsSTa#7#G!gs!W!wZ$NX7}%$*T9PeUr03y^AmE*L8UuM^v1RPYtFtQxxdJU{t|WbdmXD zyV7N4TOBLH{pnY59o1GLmb8hxF>GnEA+&}kw%nI>UmHWK)5^iyVY=BC;$B!KHs>F zxNQ@PnV7c2N2@mS*qko`=-<60qv4w40})Wj%k_~HFffFI(QKib4#JUMV^Ili+~$)H zS8~H$=iu|}1af=?a(wvwIzPMXtp7~afW(s%=;HTZd^4!3!tHmQMP8ntV9>xH&@l{^ zlTR7XlqrKy6hUhw;RvKWg07VO{hxD~I-n1yjTw$l(=b$(pso|pbxs~Nlz^^vY4}w| zp&-y-0wrDC@FxTwz|TsUYFx~kF~3y_UJY`Y!2xl~xvX2#rKbIbCl zX|b{oU<}32;T0FBHB(fDPL5hYsgXTx;Y!K-^(GOHC(H~QEEQWi#M~HhH_l5tUtwAE zij4JItd^vnO)5OTok^x`5yadqowIT2pPRkIbt+=E#g{di#~a6A!UyME z#NCG-hd+z@t}sCj3)G1B+fmobg^#^NQzS|_ZYF=%zwkEIEuLd_)a0S~e3{=FMR|!)|j!V+kxDpYuDehg$0aJ(I^XvG7o~`NLw~QP9a{gY=`8&aWXlwTCbW&H) zb(K5*bT*qd*05niH5J_o7&E3%GJieqt4&#nQ>v=4c=ZmBn>w72Z@Ge5D>u>-iE-e7 zUJUJC#+sTs?wB=~1y$QK3iSfI#w8QH@T-)P*VnJ+)9u?m(p$>=wWGC_Pj_wOPvg$v zo;B|>bLS!)X(#)l0E{l`$?}ZW1Rf-?E84(K%bwxEiB~f;-^(T2uEWQ*Uoo$4897>j z>D_LiF201#;guBW6-+HZnG&Otm93v<7)A*+Sqip{@$$}>FqHJObEM6t&^p$%tVL6N z^v&sso(TJv`ns_>v>{3M>j;8m53Pzs1w;EhO=twwAjfw<4uGg_^2)Zwe6VX3dAgr- z`%PkMMTfRV*RO$KSYah6_8y&##Jqme1>7*_9hTH3DnJ}iqDuPc(FH7-A%hjOY@T1R zjBAdZj1Yo5uRehv{qG5&0mj{lJQfAkdyZx8X3WVORT>q3?AH#mtOfXixzI7wJplHaeb&rIHs;ES+l7+ z>w3%9Rw0Dow&RZAya{9Ybn`ZHeLjvEI*?qS!I&P)cyRs~)W$LucNHNxe1O;WIAU4+ z;^X&-#jO+x2!Si%?uDOn=HSshGxZ{Vz4R?=q8lg-Mo@(p5mtmki{m3*46_{qN4k7o zzlQ4;Kf)7}ewvI%R>zunZO3P94plRvpby7&zme4~?-RD07??Yei6tkorR_r&)jx(<4=RL!yC$5*kVK1tK?UVV;LZhe zX){es$Ky{brtoY{;@}i3`*FIiYeTE*!SzTN_J;wzB{|~ z%=3M|&+~Z}K3He;&1dSoy>kJN^|~D0G`Mrs^K38A!DIRm2!1%Unk(B*;O36gL*Dnh zwtwN>-A_=Vm!eA@i6GCLh+nFJ2+7x}jxG)Uw&^**isW41?ox)N_s5h?*yBpUM4$?j zMm85Ic9W)#H1d5FA>m&OL7ZYE&E6E7VkbK#oADWANwy{7HT+Eb@iq387Y17itSH#U z743&`OPA3AoY%T13v;*Ot|1)dP&EECY+8s`PpjlqKAdzZWj-&<_vLWo=QGi3sc48q zh1VLpl#)Y*k@v2M*ih79Y>u!iho zQjKU^r#_jm+cky^?HKmF5a6(Dc)Z~grnqV${|-$W)1gMG?PUXcbJ>92e6wjQ|M%r= z6k(}AIW-%vTr?S{P2;nT8@X@ZT-=6^l!{eA;538M5T!mZ2P?|y-YAogMqP{FG$`;E zv94q*30fTe(z^57wmH1JXJLd-W!V8;8tnHJu&!hmJyP4UsdO*D|M_M7rhzUERu*mN z!#y@$9e68Q@l~IpYwCHF>$z+!{s^zBqaexmC9u<5yY^Hio65HH(E5L4NQ0Jf8Qk0T z22@3mYEPoTmmin}SrH_=V#lGxQ5v>HB(WL}i?3OKK$5GKOSVer%-vmI z;XgZHU|G>p45gxjLpaaT^9;&8%M7{x)06_b}_gPeHkHL#h<`5^0^# zlv_Lgo~O4yORgsuujxfWK@%F&T2EtW`p~fEjegVM&gCDmrSxFP^KQe>f42R^)HVZ2 zcP272qazbqbmNOX>jH_>7MA4+ZrbOB7 z=u|)m!8xOP@WuQ!c&n+4JM0?gp3$pb`#7^pM@F`9%gUVnI5d^s%`;IIg{=hz+&gb> zQ1>V}<992_1S{E3N=cHvYN^MjsAR;YlMz={b$DHeGx;oM2`;1d-bMg9?gDzGw&n4S zALFj3MD8E_JK`qqIcRA+`ak+T&C4e>2OF1 zepBEzEjzQbB8NvdyntOTp~zPl*mvdlBp0?gkG2VId1CX^0bxA@yJ9CvOXU8pca!c6 zudv)vmd_1~{)ea9L`at=6}p$l*UjPOA=c8%6GPA9yszJ)$mOeEs%Hcs!&=lR$al~sk-%ti>on5JFPT{d*Dir=H*Ewi)c;34k%aaP3lS8~^|VSIh% z6(k88iVsv5D2BkWht?I?RFwpWo&P&~1Y36JF>l%CpaY}$O!N{z`}Y;`;a4lUn_p;wp#W9>U~(vPoNE{lmt zj{6`5-qMib0RV^MAYO~3b8;vC-t!SAWR7QGT0bsnc@cl_`EW!BCdXCG#S2~yYXML> zn&tVsgMYV7Or>pdTG(fObj7!<$cvOOW6?Mvp4nGT=l7=#CeE*-Re-HLv?{WaIGc_B zZCWs>T}xhl??+0?J$U^_&<5+e$)cZk@Wuy=m~!)LtlqefkKbJ!vAqev>7%-mkm!i` zT7wXUw&tXa!~l#R+LaOgJFsT!ew=oV><-NjtNGy94Icb>9bC4syI_3DAAob6`eR0nmIDrQBjK92;BfA zLEFB!i5Up9hJW5XZ$aQQXbedoLXVWbcnm*fJ`XSM`3_S`#y0Opr(|oXZ%0`^KOWqU zA+2hyu;p7RNp)1sE?!u@6pv3*{paa54Cd_KMv2!$ilg>ileUc-@%UM1a@VKd;6+CQ zHWh46je#w*$Sp3XRihMsKYj>J5)-)U$#?NpmOM%;pQmzoQi8|pXWoySFpE7A6DI+B z^=WlnXGd1qW0XygP7xe-jVJCtmz;ye>^)FK>*i@>Ha&da#pZp7xciN-$So>E08LX9 zdHm|L>C(c=5h-xF!X`^#-6%Doa(Y&4{xSUUjR^tTCuZ`=u=`kEu!V=$y^BltU=*=?ig0lt8oy1c_P_hSWIdafKxs zOC)}&hqWc}nqCeB+O(uQ8go&L^T_j*<5V<$$lcBAf*i6EQaGnocam(DO?YGLqL2to zO)9qo*jrZ2+JfA$^XxHnURl4C{!MGwwow&@8-@&I#pb=N+Is*`pcRW<5wvcUMpA-< z3XhKwo!WBK*g^cVVIN=rxB)3-a5)g8%GUV&2KWsg{MVPr9VGT>-10Jdq-1m5ia$|v#$?^h?LhcKZ)GxGp|^s2SAN2>0;{)_rLbwqkH4|) zC#JU_!tc9|W^>6wG)2Ly3)BcXIf*tWi6WkB=X}b~2XiR(d9C5p@SikAq0r;1(}@8X z*R>-V$w|DvXay@b?xi%}gNe?PLTeUwV0R(QR_&%;VhZ~X7K5}DdNHxet5sPgG9@Lx zaxV+M-GHeGE2&L}n&Bo*j`LCy(U)v@s*%+!m8@o|04!OxiV5)1G%!D)i6?1R7N-J#r%V-gPoiK?83(6tkY`$ z>~UE<*5fhe9GJtx{DpW;Z*an;2tm9W$3J>L#;5xh@Zr8$Va|%*G_VO3qlT?)5_I4?~yiDJ%$Xle@c^uNJOH z&+&v5Jb*;1AdPBf&XAPvE<#iU?7Rrs92AcKQWDW;Yffq>rV@|W&sF#TkNlE~pwjP0 zbAMsow*9<#{}kL_KdL79bjA7_B5PI^%d#zdF#`8ZQ6>=ymIjC^B~DEvJwBd$h7Lbs z2c{As1Q)a%!UZiOhoCzrx8mA%6M1FFOoY@pSfNqjRoRbV+7EB?P@jo(N)GSIqzaY6 zjfXP0@lY1$FXMx~pOEh>q_Mpb7q&Q;VNC|((!IEJ8xGYQG%iwFl?HYt9Yqv|EGgL) z8&|eD7bzvzEq;sbWqBdVv{G8GlE;v|xor_=W_71!LZd)b$2w<~7p(x2ZB8D@p3M25 zzl&SquP=htkvsP0^T5+z z^1w}Ja(e&v=!VJMRa+zWQ*wL#tlxPM(+=Je9V1ns*_7b*6{l%j)wkafJ22Hrcckz@ z_e*Pi{q`>BlItmCb7>BKT_xY8pqmQMtzSu=3+~K51G^HwB%fwaV`}TEOl~om5B7e* z;)2EKrcSryZd~5_a;y5tOvRK6UUBPCs>((Ed+@@CKVs-LCW3(5tD_hDu=%MmPYhTa1*8dzZsI)Tsl5}l098|I z*`|If>&J{Jv^A%M6EkngW=bnOK?eq`@ZI{|xs3Vai+Fr~>^2pt=wO#&>mAjjjfz*0 zC{#MNZGzpdF{({lreyaZ$#HDgDJQoW#^|Q~xOm}fczh+nlIK;0`?zf38_XJe3+c|< zC8r0c4`gt9cpr}ve+51x4zD4I)BGqx;xSa*x}7w;P(VPJY&9_V#81KxcRZ`i&B7y>kjCKQrv z4zB4||Him4f4!JeSLDQL307;Hh|WpXdXZM@x7yv*ah&0!yVdb(0}^Gn=9F+^)^9%$ zOx&!*P(c?GZy=1hGUV|8( z6NH<>iMw(LA;^z;PBCOXd2!P`B*Y*@8+;*k=N9jvwtJ`p7J^; zOf}h6eu&jYd$Fq;PQ^*FN1?cySa6)G(x-7-`lgMH==|&~D~K3bsh+Mcu_D@?QkwC? zps75#>MQoUicy5Z#Fjm{zt;pd6db~?D0m}865R^cYN&xbOyduN)6v6cCN7j2x$BZbnHK&9V(>gN^r(MHSZH6Rk^hdIu z1kD46Gc+0_YX~gH*|ad-hlXHRc|pV_E3c`Cl`hxCqJulRY4Hr}o-QS-rjU|gDQkjd zJe{hItJ~F{IM~>pgvSUk)s`~g#0b!Y!uTv}_K;$aq^Uv*^>PRjvEcl)8&(6$D*Un$p-|IUa+X zw&RrPuszH#UF z&2V`RQIJ=AyM%rN+tDmD^@zuJEQmH+b5fRA*W|3>-LctJglR1g2!HsJQ3#clfhL$$ z5-WgpV|8d6lX@S%jQVhFJamZHR{qR~>(^204G%}wq*HRUh!qP?MWbh&Yge@p2L|AlK7(*-q5iZ})tK0>ONaxLUtU2@acNj@sYvu|-vWoNmZc{N zahgiM?yVR-xKo(grKTzj?c0vqZXSoC2&58z)6q1AK|?zd7hiXYYeOI)*ajieBSyn< z&4oj7*i?irtz4?wl&vfOQLPGDj}P6{Y125rUp9%(-?mikGYr1ju!YwbtzgN{z4f|& zvKwU*-nhk@Fttudi|~$czSy^lxw-4`nmTU7kIVF1eabve%%T)1Nnvc`zPvc_iaHf% zcb69uUaoEZJ6^Ss=7{I1LFa|M$I|C=gyipJCM0os^VT?OEJ-^xjT?JaEg$KI$)18@ zN?hT+rD`ICpj}4f^HKTDI~_Le|HE0>?cw&w>9BG0v=Ll>^@xyufYot0ZA`gh=wZjw z5JaP`sh1NYgg8?Ou~`VQO$c#Mgx7BuLTnI1tPw(dD}*?<3%5r|a)OgdgS%t%NEDw$ z^;P9L)Ik59agAo~B_;SoGT+zOMQHs;E+VJ;b z<4JSI;ZQXks>-^uXD~b|7GmI-r(omId$%HLfjB)|riP*sS9^JQA(g&#u+``RbcBJodMXNJ@=ItL6~d zY$~0*Hs{Viol~#Rcg&Hs_cDpL&~IIw=ynER_<#QWyk>6 zbsmaC4GlrZsWvWb(}zZm`c=q$uxlyh9s|Em4Nh>%3*yN?q*7J^C1vo*+U-31^Wr-9 z;}40RR5>#O!9#r~@ZH(Jr*&fe9IJnip2(SPJ7HH;9I8sw_ynFCHIZITEcM$jR&Hd( z!!L8eQ*Sfs(O0yEVCb>T=Hwr~d|Jvp9fqxqpVEZ<~63zGF@l+8S{C(aEX+1L$z&20ZXNnE+NE?z(^_Ol)CoYzkh$ZV#SKzI{tdGWye8oy64@Hgge6q zzGO9Y=400w)~`KtH(O=cmEuMJzIHOb+cd|N5}R7T#L~L>X4_Wooc#?hJ=Bn-I-M-L z_J%scryj%4vl|xh$?jEnO`T!Q+Vg+e`<;=X>#&*Aa~?}oJpaH<+J$6rPJIk)$x;(rxXDRbmZ^f=@Tr{)? zmkjHJrYcA&dG_@Md_H>>Hci2&n}~8fs1*a267R5~ow(=Dv)Q;|KSf0qsEQyfs}Zf+ z9$sGk*CK@YSxQ-NL5QT(7P8OqKh*mA4`1AUMEg4)94k8m4Qp%u{?shyezzXInzFOI zsfwAP2BWBo5G2Jr$Vg3M&!OT#9+{+b%ci_=^ZBT%f=yM9NWV@3q+n>hOorct_3Ugy&9+SR$6Mto1jRd=0$5-YE z>v`l*Rpw8vW zV{SZWFumF}2jE~{DR=q;*qaJKWYG)uiZB2c0Vg>@+q?9`Vgb?oolYyXdky0+Gw*3-d+R+^3Z)MKH4fy?` zYoV$il01UcmdO;9xyVdU=7z~b7}d8Uca~MKWXm3$HXB2`w83e&PSbC9Uo1Xczb?#y z0nkm8l?V1SFspWvC8uiR?jGZ~yT`aAKJGKCeqv)`KGhu-X;{<2ns&)&8qD6ag&VsK z;pyc+@%h$`NC}fVcICE#L+F*3$>`=CS+IX=P#x2us7!6wJLten|Kwv<<>un@`%#48 zqxI{#Z^Q^L>)9*f_*6xqX?&<$(`qp;LdYL zguOQVyA6CfcQt;$f!}W+{3c3`Bww#r=ldTvbJdl@$nH`5cXHg3AZm#$oe-Qd&J8xHHlbWP3Rt>IT3@p1j8!DX}G$8UsAj(ipNFb4)Kr|zem-E>Y(=kY~9 zu>MdUnTbiEj6!JuUF@l*S_Pqr>4<=P={KNbKHqSTT>q=CJWeG?eBd+?aD`R zH4srBUI=j=a2-IaR<$Q`9c}!6Bcj@^%@ABW_B75KlO6JW-6?)TerUUuXavUuAb zQsNwp>eikFN7V%rDX*oHNL4yVi+47y!W$-KUPBhL8a72?%4hGht*9XQzIWEGV)l+L z=w3;&*Mp%7aH^<6AbfT{S-zEt&3mx7v?S=jRDLdxSLdOx=kwnmSJET188@6hkd962 z);%~_QjTFr6t6W*saT2DKEq&NQOU197axA{b4cU9NL5MznoVWmwB_^{UyEZJk(z9(QJ=^~JMt-_s^5dqRC-B;_vok;L-xV=|)*^gR9$hEj<>JBlGlIRMO1)n0m@|vle_6r0ygW-^CSC!X zhBP#ME*npNx0shEPU4=qb13&!E%|u-I)0bII+CIUjbeL zE5v&=k6jm|hIZ!bZ`XxsGZ==+pnml)+pP~$N*?`e0iXW79;8H76dwNcTVB5EeEPJp z!i(p1>B@?noUmw&0QzP&KdR%Yj!TcUwkyDlpjs(me}N0XuOh_p@tFoWWn~CMAWQ+e z0$v3j&@pl2W%_5IFtB-xpaTPd0cs#mZdYM3dx}b!w`DtTUw#oio7X-Ajn{1GrFxan zoB{a3maH+bZFr+LDJ7eB9wfJ*j4o}Pk(nMjL|su+*Fi)wuV9*zjAkjdjxTCBhBu{X z<8XEcj-RbLt$$kh|C_OjJdoOI8zJx$Xj- zHjUIcC!Lz6v*AEq$i3O7Dx96&`KXSiIwf9Dn)n+;oe|LUn|YVAw)YAwt7 zx?#||Nu#jm-5x(HH|#}G1pT_V#2&#(s;KayY6{eyRs%qKQ{kViVo3}+&1g*6fh)c5^8Ck7F+NV%j?!)b?<(PHYV^d1Xecq5wR4*Y7 z(o89F`*j|h^Ih#b0$!c7EaY&N&Wz13IIBmfAucR&G41}hx$&Rx%lq?@aSuE zFbtDb>-REq&T3X~*oTyo(+9Rk)2e>{rWXEzG)vTG_uf0kuDvNN~L-mpF-M@Iy@NmsWuUR`HLej)QWhd0X&9{*qgdux$pIIn+q z;_abH&I^`q;j9~8;4lCFiU*$klC!RVk>%@lQ&Hh%-uw;Bn!Senf->5*ZCWqq)#1>X zetC3MAl1oHur)`U`b1)u1097BZLy5XR3JpRS_shxco{fTN*Q%AF(CwZ-87B~W3pMe zcneOu#@G?vXx-xIYrTB3d|kvFzuz$VW%qu(ew{}@ozH=yQtYZix7Hbq>DQ6-yLO{V zLbyZa(+%$bd=9fWY{h39v}lyV{S!tpqGMGA=|E{27r*sBrCu*?pPzWUjV1{R%(!AI zjpD16`yL!LhV#Gr9~G6lCzZy*6i{xnW=aBUwiKwy(o_*D(OpR$&bTQh{j;)IottaD zw`O%@HP-E~UEjv#^|LKEKjLeOX_cTJEGR=a44OAd;m9A$AQBp)gA#Syqy7Pnx4e}UcHw23pcTCXCCd^G-1q$uEfV3 z-BSE=kC&CZ_FKQ_YIt>n&sJ>UhVg@Wa_$ezSh*ITZjc)9WK_p?Oz7T;QJvaFEKpdV zbC7#Jna%dR0)!BZ>e7yfCY?d!#Hwnf4_B?@;o0+%(z?!{|6wUtoz{;(j;v~#l9`aq z)!hd0#)cJmD%EiVk`)nGuZ5)rM(mTTONyc8#m14`cMmC{g6CgWqavjs=HxVx*M15aJeKHdb!y8&b;kLWqZf z6;jHXz!N|s@L|B~+#{u&bPUIR#0Vki(Itx>UFxT&@@w(c(n4JTTMiuJsjq(wCXX4q z!HzuzbV+MMlm6ijOiob==RW@qci7fy~*nmGz!HtD~wxm0>YL19Qm^Ct%_=jPMJ*j^D9i znq1bi7x9{nXO=8R6$-uxjU)k*91gY}$fsFq5=n8PW$QN8Dr=9_(9N!CTyo!A>^)e7 z5Q0WYP6`TLVKYd*%+H~)Bx=dt1I4I5iK&591;PY}P2;WCuEFm&usby3oX1}SroQ~f zWV_>@RC}`68GtyaowLXHJf_DmrL6TDGz^mmX3k~yhONORn!*Y<3%2aw>LL9jIxuf7 zTFKvLEe!rGrR2L!JD8R?gRgGA5=9Z@msjxc?D=@B**M)ad1L9%jO)^YZW&hPp>CKA zOlV0-BQIa<+en$sYsop7fjDZ7z=TavIkRm?=5E;*dBPNQd_os|1u&>Z3;sH0EPIMe z*jZeJG))YT8Fp^$no5#pV^?k=t(rDMsnLZp!D&ZR6#O;phh0>YSSondJx&gs*YxEi5Hu#C=+n|4KPF!|*# z8zTxU-SEz5lm|2lDwk1^;Po3kzi=sFISfDU*d3NEteYmYHf;(1{nWf4IrpWvdF}fZ zCDK5LY4a?ejA2pYJjJo);Paz^lA9Zv_<| z>oeOlmC-#qV5@O~J8T+ysh=G=`S|<>UY|~GetCEZT*_c?q|kv1(5!&7s)Od0sD6o% zk{^EBN=i!n3FW{9BVDyU4o{?{+Pb{`I+JRDEipS&f0E;zbj(Z-y9#Zp!tn0xn6-XO zSdNL;uk-Q`%fj}xIrk7x%>E(lc|%I_O3PThV-EoHx9*5|-FpH{*#LzVE+#(nU+((g zYra^z5&J=vrUw&=$BrgcI;5tdiK^?_p=x9$Ch_pFGk9p^7}A_^*aOS!y_;p=taP|) z9M++#B-$N3IC3;8PA4-jy^MFxzmSKiB_c)fnsZ#jU&VWVq@j4+RWLa$DIwQ3W~ckILM^J7yLR7K$*Qz!7` zm)}SHo^7g1Zb=y_@l|>NJM)X!TNqw_BLHc_&Vs_=bG7|;JbvA}2=^V}ze|6qnn4Br z=gIT4=FZ=f=&)1b_3+;MHT<-14=s~YnBKDwof=z}Lm!=gDP=w{QbMxB!KVB}Jn_?y ztUR!v#&K~BZrO@M6%{1d>`ck-!N3+(xi?J`5}4Sz3llna;?G}y%NOf5qM=~s`oWMG zLgEfMAGyWlJTvo0K3KS#S8lxkr`^V%FB(HmVF}xF3sHnXH%&S+4iwQUD~)DN z!V}d0ddYaMe0~N#-N575t!3oo1kM@Im03Hs;;T`KD1q(yg<<>o=;w72bLaqkhRLq{ zB8GNo9r!y!q*=n|eTM*eWafNwib{j~(|snIt}rRJ3xB`pOaRs$I>>+5tt8h~&Zt)H zn9`*O2{t)9)~ z1TTc3cV;HNGc)PaG=rPp|AO4oG8DH7`GF;#YF!wA-kQVO=Rf1pQGK{%WFHda9Xx%- zL~eikOB5ljz_e-7#vezXt46+#*8~Sp&G`_O|EZ@VW9TM6zfPQb0!@~Hdf1v;IabTo zF(>8-nES&<9)HP76!iskXIeCG#9h~orC$%Lj8jz%q`PGD_~q+ADW`zjj&EEfUbeit*R)bJLAZ6xk6_v5>zSJ zlzWKJS8oi900F2%@Wa-fOvvs`io?kbeFt(w-^krtlk8PXJzdh9@cQKQ*qv9z$BWmp zb#p$NqR=8EffNy0$yieq{(8<>?i@9g=})}RZnGpjIaq)$CHo6XIrskmU^f(+rzi8^ zb!Q^%3VZWQ=+r!&Wvh4d?7D?$9_w>f?ncoq2d}bKgx#*8>L#Yo43%1!5+Z~|RaMZt zSHqR#9yJcH-H?A60jeWSsvQxHNwp^nUZ2i`PkzDDUv^@e62mZ26u}w8y7Gr>Mw6Um z<+*jpNN3i)SMk-#P3*}prbnv`&gjtruWsNEI2*#SmYt*~P=ugI^YA3~Qg`k0`ic;A z%CIIggIcw;a&RKd5%%saL`up0joZT1$DEm!jGnr+eS|{fG>1tN(Ie=+O26b%76*s>T@x7WCbcY#qWNZm`6t}nDrR$lQjZUm+ zZzaD?C+3(eTgRN3da`?80e^pCc1SXp1orMP;?BQ*%(H*FIGArEgkVsIRt)M8dPBT! z^gzB_x)oV{i8M(tZDeoa9iczyj?L0hB9#0|K+^=bjI63I$cj%Uufi3!VLTeeE)S-R zq>^J%mu5{}e79~p|M%wCcy&D(M*U&KPTpO(ifhIVzu&;7ZMnSl z?C*k05Xo^)E*aW8%6^~l9aqlUV_%etbzM^y`|k`rh#?)Wlu-Wv28 zNxPLdP@m7>)%Slq>^_#R-i_TWg83~1u(_eRBV0tuuBkkF?pT~QE#yY+)Kp%({QO`f z2H>pjorpJ76rZKVqPne;X2X#5Y>~lP*`0$WmetWslVKg&9Cq9_@f!yBzcV|qTvhD_ zX(cwlK6e?r^CFi_?$0l&t2#SmU%_!2CNsZZ8~nRly9|;Xj)?ma5>#Dc^GFWvE1_Mh zCcO8;HQYRXB<)+L<4{#pDd?7+#pRa`A-j7Edi89{?YEBSzI#sG6pP=|&qPY81C>d& z8-W>>Nwq?V%9Bz(sdkkRV&<{>ET=M3O1_%6CK%}reGT#p%2@i#&N}yDm?mrH?Lu=A zY|N*5A!QFVvWH99kLlW;-mRL4iNW=5k-`72nG!4!PESrC)1HW<0BmKTm4TxG9B#qj z&aDtaFsx&nu(kxMVpTf*8opnBJ-hPxnvMAUI;x_e zC<+fx979Unk@w};l5+?n(i(bTiej3Q@7C?$%F+G8HtF25DTXPn-XTip`yik~of9k1VD=bn6q^lu;b+NSLX`F!4LcJ3{pu(-nN%~M^# z(v*Dt-FhyY6nbM;6@^zWKaV*ZwlZto7830?`ZR7vip@!3VFd}9R+Lx>!TlGH;SaBV zj^VEQz1eIEx142_SpRwQ7^c4X9_3yiu7KP_a-4&Ee`O=fdB1GM?Tg&3QUaUy=27AC zkq}{NrdPX6b{r_E({2MH$^SM@6d;fdhdUr`caZ(WPwDH29E~0g28YN|J+&LaZwd+jr!JPRU zxncBa2qAcA;u&1nH=FsJw-aZxaaML`S~RlqaQ5UDkzZ6!$5u^9N;tez6KA*8DWk0@ zBCMjQlJ=R68Puf}OE&GXA~}_(L0Y|+{HB4>3HdB>4jX+sWwCM7e*XK;Pwd>4Pt%NK zY?>N6W0h9PxG=1WqccW#IqE8@(TI>T>dGEt{5ljx(78hfo3`dgeBEJJX_Xb;vH7#P zt9ka#?<{?iko+%#j*Qo*v-d#c=!B*!oYkW<16pKp^CL4@R%}%^8ir)#zz#fc;{;Si zVeRf*G)2J(X)7RDxPB*>4DU@wN)q4PaTRmcZDH%d0y<@+Gp>6_95xNV-(b^@Tr^Fk zbK7QNMG=vRciK^esC_v~P-{cV>Ej15Yx(A2wPUrnSkeuqiLFA0M09n-q_RmlG_=&6dXyMwhwO><0en|ztY|Wu>C#zbhM^*+svZ``E zO3K{a^{>y_ygLt@s^HTNu9!H0tIr!8y!lHh*}f+qkH=5vcFnNaRN6L6Bco9gyYq@d zb8TeR|Ht%-EGh-?#5L#e=G^6c{Nq|Yex2dj?HJd$6Mo&`lXKAsi8VVFe)T;iRLz~P_HSOeo^hJq$eGBO7Z$-0QX*^g68YwfRjgXLmCG+bowLsD8Rq2w`MIyi z$qQ{(xN!M4vOBEif^ofhXTi!4ofxo6x8In%oJ)rH1}qWENqs_v+ZL_d!9&k{jg%5o zN)qE8Jbur)bZT3bXzX_R*uMQBDXHCGpE!_zJvE!fOSc7W zjh0`E7H7i?5vz}yH%p~cyC_%nKAQYWTXU?mpkhu;J<&9kXCJ`BmOXbBz?2Pt6CPX+!--qp%@a3KxVs?XfCUz!0d$^SiNB%?>u)EHd|HZ{YT%f z4l`&HOJ;xD-h6Tn6-Q2B6OdO_9{xO5srbzg&tO+h0Xh_bS1V}OBAw9#JFz>b zkh}l>DfvYeD2kQKbMH;#89AsE&t5y3YySNa#TD+Lz@MrJR3W&3>KQakt-V>AO;fpc z!XU1l5b9(YhDnAqo^O_KX5s4Xc>M;W`gh>EbBA!pT~!WDaC?uA&99&YKD(%PsvUNl zC9riWAc0fmvz`KU>5##HpP9~^GZypXvaOVrdr=g@=>t1(#|>k{`v0unloMHNLX*&l9rsn!_&ssdR&0E8EHIqQ}{iww8TaH^s#)kWE1PR=h7%8 zj>{(wq*IF~-1MJCAr7JLH&KLvT~qKyFe*-GWGCvR)#*j`K; z+Y7LAKk5pJI_w%{wY;WPNQ^j%%`H&^*S2X&4je4ukte_8zT3~j5tw1P+>s53q6pkx zKmEJ3$f-LHRUM86K) z`p^stiYh`T9}hn_i?%Hr(>kj$-#mCV%eL%g_n{&xJU%?-K34s(lV=~B#eWhPaPfKl znR5QXdWDUD{&^>lJvtMg&(gL@OH1VGM=qdE+a>_)+!eX~fN7Z6R0X|;_*7iHjm*r{ z!yfl!LBt|4Cl$>zQn~lm3BUwQQz8VlE{(>=+0i2;0wX0t2Z@QQOEixej0StXI( zn8&Kwo0&fOIXZUE;Ko0Wqf58LtFaq~$-nnKgS2 zBEmWF`E(SNuu>SiU1j*-+M~`VBhhASV%15%fy3RLPwUkt{K-JxhRrQ;l$k*XCipV& z!;-DsanA=qZIiKsI^(oQ@HZuNY15Q*26v-za(qZ0iYf$@>HP7J?{eu?uds5}Zoum4 zara+82|7lVRZ@yhP&=i#)Wy&1_TcjP!=k6UZt(eBt5Qf+6b5%`&D3GN8PmBPGv8Xy zmQ4rn`wa?0KB5uF^-C&t%a&qKc zouniuDamoz1LsFk1P+JBUAIlBm-BRT5GWn5?GfXIMo~naxm`UvHLvx%*1RUpQUjaY z42e_%n5M}ifBJ~;=B@{zS>q%c)mX{~uxSb-2U<7kdoCOmR>dSF+uU4Ent_LqXdj*3Bb%^|>oQz1U-%@Rl7$-HVs&IO@ z)`3vJv>sBTx=j?l?i-BH>*uAXW(T9Sf4*@7PKOPfW-Vts9U6^O6Zze_gUH!az`Xw~ zqv?urny+!wBwykvH%Z>%Ln$}Vb%VuAw(+MweF(s3b5_+_2m#iR`gChaVPWJYZ&g#m z6tpWUyv+P~1>PD?kI$>~zc<5MCzKSs*tl*V#YGjo_1Z$d{A2}Y^=m>(yk4D|A6e2S zefzW~Dajc&0a8_kTW>mpXa72taU;9Zw`VJ^nRYs_KYb;M36TXpPfDWD*2I>Rjv<}e zH=}K{GQzkC}% zt=>U$oP)_Dd(fp_29G}SIg5VU63iD7#U}P5ABIB-k>B_E4Q710oYMw{mu5mBA|nLJ ziB6_ZIt^VnBIemR9X1B{Z5#IdmTdhmDzY>ird)IyrDYyIo%u70A}~$Kxs&>G$;E?E6v4DhhcNBZAz{Zq zRfx9EvN0N#%cl-u`H~%Y3qJZJ7x{c_%$+qoz?AVcq$5VT$ zxZSQXVMJG2XQeTCa7Xs+E(j6#127FqyY@{(o;!G;7)?{CnVKV|WZRa5!M_dN;pO%@aj2Y4yfP%bI zewx1#)0AYSC9-eFAxkI5UgbdPeuIprR^OMr+!7wTdq!{;B>^-M*6XOUKNLWZ9+s?t zq6n_PW+Yck8$y15IqB(%C#G!ou|Q%aipMx*>C`EM+x{?~XP%v7<;v-PtLmnjRE(^Q z)Cdq>1I26oW;9LV)u-q3?JP^svt^rfI(E%u{rrurG1l9kJ)5Z-q zj=|xuA*E!^>OE}Qcz~oNCq9o4;q@a`1p*NxA)s}e^x!f6_r~|k|7Ib$*lL61|ux43LkU7S`)JGUsmqIh?FaXRbpJ-U$)s)dUVfX z+qxXmeXze$2uld!oHlxQYr!LbIEOc$o6F1(mSN~7reDVFKmNFdCm#O_pWh6#HHrba{K4lPP8;dziF`Y2O;9d> z_l`sC-f<}SKEKc4mt{L6u1E~YTgl5ne7v1@1KaVto5s+)PiuT$oqO(jk1d;X@%wbtqH-v4p(2n{ zA{{o&MoHM?Y+O41^xzmipRqiwnz6Q4C#t4$^Bre~J>R%-?G=`%F41ReVmUDlMBCQs zcuPDHrSk0#jn*yFu-Q~>HkD2to6#aAA@m{?!us*}4TQ@JZZD=sx0*@TKnPH5RbQ~_s~@DJ}&x}%WzoC<6uJ{%P~qq?=`ANQWmnxA*^<%i4hdHfhYJ*bHw4T(3W zl+w+Ia5`)h=9lr@V_)F&>ejY_s8XdAw=ls`;AY~uUfh05RgaMTywZ@Uk${N03qrtY zLppNz|DAILXZRE+%98D393LDG8%@%au-mnWwkEw<5(x{A}Tx(Z~}x72#?=#LzClz0(_PGyYQ0Jn7d#jlJX2I6zz1{Fr`Ehf{vXtc;J!qPoe6f--wi8)FO)> z@5De-JvyN@5fT_5Vi=msLcB$Nc-&R94~Nsn^|y^f)fCbiB?Px^ z*D-^Yfyx}DL==~U+Z!}0*lb|4p%zupcT^XC{>7>~YT2qWcicIFZr!r@-4(+^j=yI| z4MKE$~Hn-e8Aw(MnVAYSCvDr2J zHN`1)ldWs_TFXFn{9H{_GXK*R{4#G1<;|+KbE&1OQc8)>QjRDqO|(}s4UQCdR8ea5JRD6=&&C#C_(YSow*-npLty}pojtMr_-i& z6K2dOheQRA~&ICni>mk&``W=)zZug8Eeosbf{*zLMxx8QHjU%^ZNoW-_p*F@gP@5hlEZ_P3S`80KL z#Z61iM{h$jC1A8`PfVKW)Y5s!_veN^rR?1EL!!ZZ{$^0V#x) ztXrvvlba8Ll4(zcmwE3kV%_4cyu9ckR87UQC@BNU9 z5;v2t8UaA7HtF2`;JJtW{4tJ%=(9Dk#NH^Td+$~}@VAQsYq$}RCI~@~fswUeYSOKD zD^Ba32|k}CZY51djPA_kH(G{8Q4}6{VhWEua~UHh^q@&b3Ytx|WTwk2z*UtjC(GQl z&rIZs8%J~5^`n@6>sZ>hObz=%>3)NP14S&IU5g_Yw)M;Pq>1-o}cAv}o6q|17wN>+hRH*S@Vt zNOoG$AHx85mC&Y;lElJNE}YbdORpNnl@DIP#OcF`PYidYyly{l|LYqJJ-Tu>PAa0% z*3>H!BZN2;cm}J@&TCT2zt?!(0lWqD0}6mCQp#ONcc`Z(Y0XklwaC7vqG|m3g)2kW zg6rbN+umf+{}y5dqAVsbfBSCEoY8JH6w!0$)Zx;TCzFpxb6xsJJ;`H$k~1BHO?5nf|JBqiP@T+U3XAZ%NnPIBGQ`> zV<1LK5R@P#bZ$Th2uM++i3(CAiXt5^7$iv00D?$+s2GY8B;XT~4r&Z|6$oOeLIMGz zw9uP?_n>#)n|Xh}GqZDcc4yDG-*>*>*`2TKVcR6n$Wv8N*ZQ3Lt%Y?wOG7sv1~Rbb za{eK6W^3oNxO!oD1qb2^y0aP`^?}o+jlkxFFrh<%?8^Q68FtWU1U_ckG z0C4v&attVb>&4CeGhlM*XHwiT=myiF`Ws#IV#AxjHJBMfyd*yaPtu^O7%CbqoC?nD zZN=QX?L5r}9Sp#x!Sx$xu8)Ul>t%@;Akor}ZKZlnui)m*I5o0oRP))0$g4)UzS+Gu zMzrl$uNW(g9+1zg=!u!xX7-T7RLP}<1E+5nFzzOgG$St$edOli;76)+^5LDGF?~eg zFuX71_^#q|hC2CH-G1Cqy5@Oh2!n`#S?!@Kv{yD~NRAn?jBsuhzGG;cx6hV@w5PT< zD{cFVZ9mea7}t{|!@u>EcCOJ>P<2anhwD~8oB(Vi(wXdoA%Ri-(XpLPCfmtB7eBsD zj!Vwo` zPd{uEgDy_7-oQ)0?%z94qo0Ta}FLQo^r%N+}T)baSn?e%@@O|ee&&Vat4fhURs zE3(-qFn3IxMGQNs&7$T{B}Q=8L=tuC6c`VzuPYn9K$P=!f)-!b@vqjxPk0gTwoynD z;k(4iQF4n9I@S%WpPWibG+JE$D)f!s(BW*J-9dQ|$rZG;K1*2J6^t5hCrGI4&!wB9 zrsWc8DB}fNN7`8Qd$2JHkc(!Slg3bNE=BR3BNUyL=Itu>4yy^PN0@RvVRVKux-?ld)PqJkI>FQYXkosnCSA_cEcvG#JU zjBd$V#ImNrUE6hoEvBuNU7)*Ca9X8}-HG-2L&|`ml1lL1HLYaI`%+PR2sH!FLae1;GQ*}P5!@zBF`(3wFXrTu}>?=yLR_0Er;n!8Gx(3^GTpoEa zw;x_EP_9u0w}HA1DgC)qC(p8t4jS`!U&5ySOU6|OMioK$|74q{B zq^HbE_tDQFh>R8IccT%d$oBaWEr1&&Ft<>&beJ!2mx2QgYKOYVsg2hgg`D;RV|OX& zrX2>ig=2l3-mG;H$7T4EqD@fCA);* z!ojnjq=@BzTyfAH%!8xOhy`PPBUO!*#0~(5_gud3GI!5q6StP|L9ZrHbjx6#n~p=G zArm{ZHYioZ1RtX0@H|Jl@I4Z8vZ+muCB;(0soBxOwdNbM{w-`_m$y%aooR8M#YD93 z_?fCBCf3{?M6iB3^V!wze8nO#o%EGjM_y8Q=NGJZ+nZjiqK9SNTJT^mb;e%yGu_c% zzUuWL^T#iA0znlj7Hqaj>6CZs1Oti{H>Oy)SCwQfmJ2GJY@&g}`Hc3!;sIexp>^?~ zh&bAcvbKZp3yJbKEbjjE-pOpt%^S!m8-U5pv~nY{3)tUObsH;E2J+Mo_3|tVOQP-h zH~HbD^O_>_N6&ADLhT%MpKEh5&ppQ1&;+QC0Wo2oB$=4-I!SWC#oFUU zuL6<)^^$oXDM;$sojQLHJtG%V=>Mw43-su2ZK@>toyZ3{2nPwnqxfqjau1a(L*FfY ziSo9!(4T)O>ASDjlk>^1ApFeZuvZefh?3XYJ$K7<2$#K!qkCc6IlifiJa#AevnVAZGw1d_N?SYyfGfTf~fDTr0t vZ(JiCF@P)d_lE-TO&pq}{sZ{`t7AuLG5b;>b7tWO4{+HYceJ9Q&)@nR)DeZa literal 0 HcmV?d00001 diff --git a/docs/source/_rst/tutorial4/tutorial_files/tutorial_45_0.png b/docs/source/_rst/tutorial4/tutorial_files/tutorial_45_0.png new file mode 100644 index 0000000000000000000000000000000000000000..37003cd682206cfb7a9609315d2b75833c637d2e GIT binary patch literal 84791 zcmb5VWmH^E&@PGw0t6T&xCfZQH8=zhgAeWw!QFzpLx92Eg9mqa3&Gvp-RzQ_ zwHUZ$x4Or(a(ArI;*z0bS*dxXV+jJ}M15WSk@sI>WE* zi|_9q)Qe!5|F5b_E%*OVl{`MUSSbNQlzHI&Uwh*^zydK)o!;f0JQR}UQ}{I9&mK~b zb;S?_v510{WJ3v6z}AZ3j7sRCTjA5pdtdkketRvBpUf z7TtFLsTyXkaBGS2`cnDj`N|8`teF4Lc=DoXF|4U~d0+qwaN02{rV3yIJ&%K#TIK_P3>e{~9=ysQiDx}Z8)kh9#_#U8CqE>Ya= zYauk1OD^lL8rZ9D!&@-3wn+nYQ;}*!3o599fIra}oGOC7R6x#0@vMo~{Ek29S;t|4 z1*Y)K@T%__>>P|9;-jxTjE@3XRtkOXLo3`q+$y4{Jm{visAKMZsrsjU4?2zYKii?7 zD-}$`ZV^A(>%+4kCV%k!0Gun=_?Qmo1g{6@T?$QAbUeN26z%-74!2IZuC6b%e$p93 zxV&Nu-yU~;BE0VEGw|X>dv|5Cu3eyPTYaXj&%m>rzCG2i8~wgG$Nf9tVt~GYD$fI) zs=hWo8p81U;hBIw{Xc0kqZMSPEHG}@gKxD@-xAUPz#=au z0?_y(bs+#~5)=DM&9-Bx&ts=O)>JCKokWQRj|cDQ%S=1>!j7qro61%Jh}9CyNAUd2 zEezt^`?xHfTuT%_Sl0=(^Mwq<>LPgxi(OJ)0?$inIe4vQe$*x;n{nwfQf71(clyn% zjPb933e#{^0C6IxN8-u>3v>&i&Frjr7={UHL*rt5OYa>N5J7YMw8VcL(H;p4ASiuk z|D0S5O-(~UjHD}&i1%89Q-$%7(-z61%gTrT41pYW5vDi*Zyby)$JP*G0Orf8mSGxt zucBfSei@ykE32&65#N=s1hQ)R5bZ3NtnXW*+Hn38JzwJ>XTfvp#Zkr4nQGA)Sd}huw)1{)DBw)w=y=$WlUjd#TfIoNGH}vl=?2EmbF!P@4J)V z_1tTny7Q!S0bI-ug+xQ>GB7ty;RSzDrksO|lVF*>`hPCJOZ)LhrVoNe(A0A1B)E9W zuf|WBDEcW5GOPy9@ZagtnJi%~n5Kt9o zy+sfRcGLJ@NaGLrm9KhV9y+=@mLvZ@s6BWhv292B3+MKaGI+5r&`I6Mdg{Xe1w=Ce zC$-7rCNa`i5OlHu-n)_B|9>XVF4d(>AN1}MbR)khhZ&!vCa z`Q;v#W;HTl{2+3~T4PFrXzbo&2TevsV~XasexhF&T^QNqc}DR+`0kJQ3j+z z>a&bH;r+K^_h4vJ%NEM9@}Q51UvY1~0?fhG8Tr|-u=NW1q;5r_%WuZ6wNU~gp>{_m zuJ(ORKDjKK>p`~su{CQxA5yd*Q%*IBJYZUQQ`P-<$b_|K`8`;cg}P1N97$Z?4ul^I z@6MZJTGR7%TZ+@$yRmp(-D@Q+-d&TGcC z2T9b{Fu?YmtWTHG;~j~j^8DEc^H+ke^c4>adcf!fnI%iog-$(hM^vHaqoaHCj)QyZ z2F9&sy;T{I-X>xtUuL|X7el~O@WWP<&t@!NBKuuuO6#ep8QWc_hWvw9J*yxBGVT1~ zK3VTWt;twm-wKz&Fms19BhlP(b`4+8kQwRUrMDL}UXN<$flqfa=j}#9V|H)I1|Y+o z%{u#WeSsr|};a4J@kL!D|egt3u4Chv`!3FO|#&t6eqj=am?a4>1)+`99Ei zJy(8ak2~}Hn&P4(m!%*7EkmY4CVs)R=lY?&s^lIlSHxh@(49$VF=e&8nl-88DC;E6 z(g{Jhk@)Mw`J2A}5oX~tA6?L7_|{);|F_n4(ZMHg9n z8vE))Zu3dU%j3@TYl|8-4xHGrJ!(wq_+iML(lAn>aNw4O_SzP`&pp0m>w0$4PG%OW z0(ga%a)5tgD~7u}>8O$C(ue6qzFWIyQF}S2 zvFWg3B@^6AdtN%w3yDUzOK+}^Frq$fKlN+2UgA34&b8GM2if8#U)skkSEWyM@f+v| zDG?KZAF$Tm)|r3M%6jgSBX2MXj#IwNhEimUje0kd+#9oQ!`r~aJbn9^Xah6S`dhob zb37b8W)SPz<36I*%D?e(AxBtSt#b;d*V`F{pg&SH+!Z|~)HgxDnP0lz#aIn<;&e!;~fuD?OR`8NoB+|u4=J7#J7t&<-01PGyhtimE)5P zU*#luXKZ8`CC{_)Lif7Kp2N~ATC*U6*`XCcW}ry|l$(Y@{Q32Lg@FWMM0#>N?euot z%?x?;BU*Bvd%w7Qai}F-Y!Vn-!9e^>C{b34!lYqjj{={SG_Y9Z&T6pF9 z`}lO>5v&So+aBOPO)7_pj`5cmS3}Cd^c5vSwq3E%ik+~M zf48vLyy)5Y1d5bWpG7foHr$w#&F~hhccuJ8S*C@+yhrhW3o!xSuL&dcC}=RzG+1uy zvjZb$3vbFVHL8Tf)WoDV4k~lq_9V9go=i2LWh$3WA8qPZEYA9^X%nn>PZO#(bDWod z&UhU!^Y=!271|A#MoToIuRZpPGw4;r8U-pZZq=O$1~quEQ#WR^+h|$W_KQI|<%6*t zsn3NArp49cR{ouOPFr&r8MBPL3VDI<_zm72656f|*l#?q4)5Nxh6wX)@|# zN+9I^##6t>zvcX!zB``QJ=LsDe?d>}id>j2skyIE@!YFPsX0=FVT)JH}!vUhy42%Xpzk+%KTQiln^j29U_RZur>vmN;aj%LB2@f>31|aeT4oTOzf~5HcNkzpOT5ayQ0_$yV#!1}k$ z*H_lZrvdIeBDJgifTR0=d%8mjNz0Y}Ml=U{mmlX)`I{oJB=3gHD z3iuD1VmM>0H%HE1HD)%k2FkRxvl>-8KOom~>-{)UhAg15@v1hA-RS5(@5bLB^(D4& z@g|!sJ$j$EzY4!TJ&spBUdZD7Y|(FyAtVjR5Cv9MOULyoQQ0(YyV$fH%k?5>wcHzN zzm$qEJNCp4`~O~4^ixJL`c`+YEadP;NZM}38A8%<<|aBLf%=YHHCogXVyo3@FL0jg zQb!Yx$ikTR8*eM!GfAsG=lZ0?QQv=CXA?| z7S@R;8A1tBF*Y=4CG<%UitT>O&sXzXS|0`G=;SCv&vH>=dp{6IrSK5wAS(*PG%;|e z&Lcmt9?iAC6c+HiEnV&=q(Ihxu(Y6$d)#NL@Yfp)bUdu3ska#tIqu5uX05(aqI<1A z)3W7{OU=(rtd4^$MpBE=OXp742Rd3)JkIn27F^j*X93G6RKGum`@{84^pcpt#`4#c6&f7{g+Fg*;biK)hL7Urn z#GOljc~VYKt&ZXslYdngW>npQde0U7q6m&Xc4)}Spw5Lqjvj}ADXY?dI5V#tXAF1g`5oB@iY@kJEdt&&Dg}zI$ z5`%7aEbzzIH^OrD1{aUEQk2jdP-JCS)tao@2Ts-QsQd*XT{KV>hn>9z%PP1c(n&4& z?Jo0*J{FU&A?AEO?vv-~lI`Yv{M}k#%D4`4c%#m*SDvW87_0U0BU1A4Pt5!mFuHEt zBHD8XQ@g?yuUDeaui+l0sHkhy=?hW$#Lsmv05QZk! zr^CXL`+FE*b)0(Uu;f*$(N>WK!9wxg*|~FkvW~f`xmuH^VHdpHZHoLK^LJD!u=LH2 zS?Kx9vRqs_u*Fud%cP?vAjW4y<$2t-vM8Au5S@4^4@nfK&#W7)y_V!7tnJSiaAFm! zEm6{^I{~sRa6}&otCqxTfWirhQxn3}I|Egu!|ZYTrJ;Dm0=mDnQ3&W%>j@+cAkmyi z10Mwb9)3^Y;b&xE(;XIt2p0JLT$L@xzYOx}@KyTD>wo9H&!Ng+AEfuX&uq2G(iy4I z)_N3Bbn&c|8vi7q#sW~2d}l#`2_k~e2N7;o51%^)Q5#m9{MVS1hd~33+iG~La`=5? zr&^NOvZtL(EWQ@F&ig#GB9jSyIAxqP$?2s^`W8#}G6C_zp#39}N2IiJ0T9_T;fSnz zLtYN8nu-;SHd4*FO!`3Ch%^pMG0~98FGF9Z)9xmKgg#oX@>H(E6I1EzEjEIRs|QY> z)dE{DkL`or>h&x(_fdl6g-=IEhNbP#K$dWxgJiFVPChvk2fYC|HP2Zi%Hh;^{@_ct zGD?%t#1`5oT<*=^?g)Lq96P*pW(rDi3?p8-^1FAy5o! ztVaF_8u~dC?efUYhuRoKC>Z@+#onGojDt5CK=ng%s}a4@c24?+3%#u7z_J?GG{cgpFpkvC>JktNo3Q`3T67sxl*wG}u{R?ddRKKeJT!c+(HC1}FF}sX znt^WU29vi@@5iBci_mccV@>v!$FexZt-7$h_n1PJVX2!|@S3QbydOWng(Lo_V{cT~ z^+!cUG^5@b7iSi^H1EZ&-`IH7<|}evD^_jxnvTee567vaMcBNhv%W$TNTY__!c>rC z7VCzCA%Ozsts8N;yl`|o=l!l3qh#KWNK}iO0f&f*`kqnP%A##~q-2UHBs0iSV(bx* zZ2MzWzb`47Cq~4ysK(JJ!qU0t^lsFXQTwJ|1 zZSa)#&MU8<*iD%6dE4DkdcQOvVFHND9}fF|>F8SU%K0S-)mi4K+TfNa^Vfv*U}TJy z&oR-^1wnw?yJsllI|=s8enmIYxkV>d!6V+$ss_Rcj{YTp5D)R(6u2%NxEji)LM~Q#>a-AzSWDDbYA& z-Y?=fC=|RWo7+b(%WyaJHJ|5O4c>V#HX|(xLq=_F!r6c>i(PVdd8 zqz%wKC?`sq*vJm+&Jjf&odU|~4v3=Q?VG(#j39*ZlP<*X4NAvmu6BjI9f-G{UyITW z6`r=(%HE{cS_Uac(Sx<4j%ipEj#^sx1UmGyeV|@<8Qnq|jwm5yf@Vp3{95m}DQ#j{ zLliCET9ct-UF&x^TKqMX(riHsNxHY}+WOcK31k@=%+L2@CazPlSHP$sS!?l?S(UQG zv|^#ZGtfs!BNeRN#20|LVH7o(q5{4o)l~x4wDr3TApa|#RL@QN1PtzMI=yOrDxqA~ zYxO~UXwFOt&meS}jACV9?maA5 zx7pE;*|-L_gks73WK8=)4>WT9}eKh7`9%EAmoA5Zn?8{YIQ> zrvaFv86dTya#sGZv6sxEI4)IMdL^-Q0`MmSe4EeEH+7Zilv$3#x7GUq~{7fH_jfLx@1Xq?HWZYh~T^!s(W*&C+7 zMwVr$j}ygKE3%?ZpWy~&l-rtsir;C?g7@c9LWCw|C6*luK!RT8Z%w$`id%38 zcG7Z*GasGx^XuVL<3se$^P|V!FAsF(`OMI2Q^Olm{~UF+hvOoj@yUX;f0{;-25&EB z=MT*7sVbJPc2ngsoEjE=HODW1mL!o_n~gEcLmexkR?VF$3h)UnNfH@4Lor6Bv6CAO z7oFjRxik%RWXn|wtfZyO=<*nO-vy5FM~uKED4 znO~yuw4E^o;;WDGhik@11?Bji1QnHi*mF52w$bDS#qd5U{_Qg^-`tA+;3jGyo|>xM zjaLHrTEMa!BGl@KEvJ2-ZpYe(%@AQqa4?FOIqI!c|88>VP{-*DHFoENhYBHkb< zrybc@S>fyaJ;+wIRQ-7P2>#dQ%HwE#LI7?o)+H(%TA?*Z9?%XrQJva#+4VR_=W zjb4FXFs-tMI?gp&D&@YtLPftg;BVN02S+cFG*0-1LwBt&Yd6yCmTT*CBoz<`UXI@d zmigy`L77HXc}22G`os)Sx_nUn&IcZGN*lc8$a)(sM&erFOe<-Y@efXr9>kzRl7#RSRa4Ztlh*C~NE zm(mUulGHhLMNF(lC2Uw_M?Pv1nbl|s&F-BBQezeg?`WOtYl4$^qjGT!8u}UQFu`^7 zSS2=;u#`qm0F$myogd3@hDeteJC^5b?bkz+7)0KcRSa1X|LVnf(RrF46J>bh&bYA- zn*UIv*uj@$esjJ<8Ne>T@2>7G`LPmG zGWJRUsZFKqaW?sI9gE#kS&CSBc?9!H3Y7~m@iwiT=M&)8TnQ9iOfW5hx4gaP3AN<_ zH6vveSmY_ydxRz$OIAi6wJnA(xzpS&mVvPZXYMNDz%5^+Ck=WTEl#zFc_rLcy&N1i zZpivA=Mm3YtE>GkeT@F@#MDImWWx1ce}hPUMV2O&22#8|=9+_TE;`cIyglAA_}FZ3 zjqUFzeh+;Ggw4n&b0fp8Qx8=Y1{T~ON1Gpe@Tn5@l1OV&WEL93VGXOU+BexM-lmx> zAnhaR5Ak6--vSLAXoUK@@)^T~ju~2ViYcz>a??u&!!HYnC^Vz!%{HN?PKg~6)zP0H z4s3MJg$erAdKXHZHP~$0s%LZM;d(&>5EtO<*t?6vSdF2S*C+GB6E%K1kmsSh_C0Q;jfZ2O(5z!~9xq_JIq{!HTAbGN0tcfXDCTtXeu6^o>yj*ZugNcX*XOa|{# z-C{<8_H$XD2pjci?&d0qh;%n6WR{8&2h~6)C3`rFk^4*dc3%qQJ9q(EgnB096Rj~{ zdHTWW@Ok1K*|;ytM?MY&)`HC7Ugs?hMchbPXW$NWr@;KMkQ7GK0Me{6`vX7iKq>W@ zm3OI7d?Zx!xQ5?s&dy(GOs#2QcLbB4A{wx2N1v2`HPmR(#Xn8m#qHDt~_AQsPwJ;HR2wh_&jcr{b0?I%a$L`)=j5mv)}`P+Tj_3SX* zH&&++9GRvreAq}EB{64QlXt$mo6MINqL$%}hWE1LO^E@mq@N8*oI596Qp1-==JUlL zq}!D{Fc|0uq~{+lY=cP3R@i)}8@s3#L#=+-7f!>+BENIJu}Jaxp807-q;PlKK(ibU)-!c2N~+e4o&m49FNvKH&P#hqK5P2mT_%cy(3>Jb6~s^sR*$|&Hs=WZt}y)Cj8G* zyJ@%*KZ+86%QuG&|1OnC8C4b#Doxd6{pXZp1?p(e8CYWc;PmP@!O^Vc5wbJQ;bFo! z>-nubQKRzuiu%^f&&9Mh&DI14#Mn(J5Y={^LM~&0wUzlV1BH*G%U}-G(WZADx@4>M z*T?HBY8=e8`c8O`yHlI~0FJA*d%72ROAxF*&T@Kg5!o6aOv*+^(}@3zBCRM43KcRC_dC{l#?#u%&%H-5wBXk9Ey-Wb5Gf zQN-?Q`?!fNpQe5576;kecmmo1|_EA<$@lC3Q|{1L$<5V(|!JmPj)@VROl$nseczsC52_p{KS zM5?SHam#mOIYqtA{e#^I;|L}L_8K=z9yXk%RpWYG%l-TtAF7lkI0Id{Qi-B5F{iG3 zVWroc1y|A>3tei=rpo5?))MsfrfB3I9QnmRhbC|$^gKamWt7OD;u*6LDdfZRf+xFpFayVqHFgZFbhmKn|I1#7PrrgsBT`si4t-p?MWPS zMYxH_Z87Es&Hh2HZM>b$VO_n>Ky~Yp_a1QPS~F=mrrUJdECHv+2v7`zHm&x=Efr4X;gybs5tUHk;-w0tBjYk z^A_elsh37sE@o9EL3on%#F7C;p=`~6u_lpyMuRPc48#kA zf|OwGC;k{ueZ4dnomZLiIfi2p(!o#rWG`9sY!TS4XbS{~a-yBs_K@)vz*mz7#*BfFL!GKzsdNh{P#10cJR&PrTb6R?N-}Td2Pl zgG{$7?4eo8!v!sTEopo!9;_3_V~(>K*Ynk~p66Q~HGOS_*sNig9)$L|!k}8_Q3pZ` z#qW!8GEqh9wT2`p!bQqKdy1CdlZDK%@O)k}VA5X_^!~UE{HuU^czh&-%c9d@cK;fg zv>gKDIgY{*uY$7$k;YR~nhZs0K$4LwQ;~`}3YOD0FqwO&O0sOk_2Y0uF;?+s9=|Ju zRiA5zyt%KHy6-VF$tHSnk|@rS%v`1}KK5sP2NYRl2A6y*Yr-SH?D|-;=8Ogu*Vvw^ zKl4Dhb=zSruV|dhYvr{tGTgKjYRN2pyG727D|Dwr*Cz)e0UA}L^P|h2e*PfKf~7TH zNxjK$L9Fa>KmAc~??c8PKc#1fndXY-*gxkZ_mk)*Y?Ck>hB+q6;aZbyiRxsVFe-rJ+ET(ICY+rWITqt0yYF-r;mX~6tw_+fBFZZoureZhA3i`?T{7M8far*!s= zxK##Ki8i!9%sg(}iEcrylvw47e0|ZArN3cs{kKIm2r(4107n<%E==F|@oB6yxjkmt zwXiLwW8h4w{cg~~MC_wYl`c|T{uYf&4I!i+(#W*;$AU!?Df%B*n`6?P*Y!k{w#!4? z^<9j^^93sRQ(4~N4o&-wgnkTS1msX=9cli{u@GLYEyr;d>NCo|RP+NhRJ%nI&nwYr zB@tHeqI^WX!5JAhszzCXPLUjItN;tvXwT%I!4w;`0grv1M!RKkmX}Ydbv-Z$#<*J0 z;B%$|<+HKyME(tGWAx^-kH@F@=4vZqg55RmpNRbp14o$8ibv4^SyCG^Ny}`n+lWs7b|yB<8fNB-al_gYN-;& z8#s^T8kx8)qzK34OfRyP^rWPsV2D=>kO{uP1(-g9w`=XYs~@F})n=mIH^ajJs62W_z#DvYU?nUc=GpH}*iIm0$t&qQR)AJ)L_J2cA zEXsj~H`%&T4W#fiF6!pbv{o$7akJ6h;>uYS7%aNF_klv}c5*{N)DKRb{lug2$5<*^r_NLb@_Fp~RRY1W!k8%yNFKS`f~l9?sCFQ@O`&K@E?AL6s014HF} zJbve222*Lhs;<)=l8LCEH!h!$yEWgC7`?i4%~RZpJt5_&ecYHXAb(x)Dh%+7!Xd&` z&cke7gWcWXO(af4elvf!4Km+J(X^QwW0fr~Vtk-rNSM2tc4rMenLT-e{rx{_`Ftwt za`+_vOO_}N2V9hkayBJPzX_C0C`B?M_MS{zjK%_GjKSYaQ-8rwzRvssE_O9Z5uh9( zm#Jk^llM=ClFt?DXWJG3JWkqEE{qj$%H&wJ4h(T}+}Nf?)X@()-XDS;_xox08w0x1 z=|P6f(4S>`e8&n184AO{EndS?F?j9aX}Yb;@c*=ldzFY#ta}LEE4|0yAp0kuEQK^Z zhd;Rrz+l7d%uJB}I*Eo?4IQKdeDsa*RjX^1>2}K|n!h%_d0Pn8B2yQyn#Rnz6NBD; z_?>(xCPiwsMc;4Q82Q%;KjnahV`(FJK+WMmM|z%(yF1YjP`yZX?BC6SpMjXiPQLFe z!i=CX5=NBq(V&5{Oi`{=E(v{Y%c#-|7N&qD*Qe~?b_p!9N7?c3AZ!$)si294u{){? z6=RgC_qjziz{8<6T9ar|xZmV}ni6ixm6ChytuDD;so(*6S&K+pne9vC5Lz6H%*7in zFSJAcYKQxHYz=u{rcb%8rODTKeMm|Dtz?3|9Mtr4!6ZRg1>_O|K`bvtE?(i*G+zmurm1c z9U=3=KthsMW55W))WbtdGM0X@(t5NgWOtYN2`aH^{{#fF_(?$tDXu@oRfR{n8S0$^ zlnK7f7v*gF$CY5IX*6?c!&kwTZDfWk_J&k+gx`^Q@S730{4&vMU{^)MM=de3)scWy zafHcv?zM0Jf;YXaW0#}danJX93Qg=Dj4P#=@`!sQ4sFSvFFp_MOH^dJ>5wGoN5<3h z5por3O>EC3w)I+=;D7t`@O+)brQ`skmP2KwX>awCD&D}fbz&<}wSzBD{9E9YM>^tk zu#GiLv@Q2;L}yk4gbdE1^T5NQ%Z#u7jF=)EyMQ~E(I6y%%djD#Xxiq6u&+@vu}1v} zxgSiS*YRA5pcn~3q4KAR6&FlMVngvy74%a24dF5bDfmd45@V2FiZp9z{NAVk~+*RC(*2Ci`F30a@%C9%ITMq|O zu7C(-Y5HSI!r_Tj=GP|;Vor|_Bnvolh7&_3kd|LB;Vks~p~FQcYC0ir_oq8ni-Mz9 z=&55%diD%nIvRi9Zpv>Wb{*`s0H2vP#v^py{&CM`0OzSRlr$LLBh`l^3N$0m*4os6 z%7Iwh&)AH9auIQ1sMb?YMbPENCqxshl zMs!?63Cv!A)}EB*WJO33?vz6Y!_unl9aVBPzSvSaHs@TQpYsDnyVH_vqh@^C=|v8_ zoUw6!uU+w=PU#%GGD9_pun*rdvwV#B=gSs-^3q(7Px?<3sy-|(M!mkc8!bBlS>j&j z?kUr$e2Zw(K8GKKlpub|Je=p(U?j#Aj~DdM%u9n2G3uRBK3=tPQ9LdPVma`_1O8hw z3~#pd{IEhd#s$-DoNGRGltM3j%vYb+Jpm8VkVAxZE^x7^`1V3Z<~3O_#nad~OIZ^` z_)U;MDecFx1dG!Gd*c9@KGIJmK?~Kl)4iHXk;g#aX&v#t0 zU~R}AV{=;8vIAbwA5Qzac6b%Biep_r0wj$&>JRAqnZ1)$*3eBIJ*}k91>mzmA6A76 zaU?)az)cT5gFRzIB=%EJ4}YUMwvj>Sx_wzvxN=xQ>DxOG#-Zq5Lg1`mAM!u)t$1gU`@mt-;i?2$JoqEHKO7G}f& zK_BUGYLRi(bs#j^4Ex}$hm1Re*>lIi@=)cbnqt&qGyTQS!kxKJHI8jUnwrn07z&;t zsVU`QB>b06^1X>k7$^2pdw1+$*Z&5?VdD|W{pB^p{;W_In)}CM_!>c_#J0&$WkT6; zq_Ijx|8VsX4u53lV;LpHg9)7*)K`i}+?$5Zo8Dt4Gcc1{O6nCEqvJ|qAZffrY}uHy zt*f9ANfk;#7NlMM3h?&sFLgg5V=^A3hx`JD(c~6%-}|_(4NS0`UB-@7yt7Bh$&uN( zjRp4o%Aj211Tm{80BzwZ3)?JM=_&wEn!=~)U}Ctb?u`h;4kv_kwVOt;iyWA%gPKS} zSzY?CEUGg-^{gh|y3UQuj>b21IR^@ckN30o@rfB2s1ctfZas5~=C%8J$E;&}R3@d8 z+zW~lf_(a(@HzN%mnOUaDQ*2BrzTc~L{9`BxrGaA5wYwKw9GB?3cF`}cK0LM4x_aS z!PTmB{J&aGt~GE|!{o}MwSZpDf;ZTE&8s!D^r0N3BDRL25D3miI}y`f<&jp43FdcO z2L)FAK#|<@<;!^sIX)*L-aWVH?bRiNU~(c3l%gHO@zC=oA#IF!VjImi6e<_Le(!%( za(4cNmQD|p8^MtQo7^GEnPx7 zJgRE7s7-BYUJ~Bjj7IcgN%qfGsQ0sTYrU1;9z;iTXU31UOwm5C-p}tLH%h>ugb@1M zGOTszP#xVuVxSmdg1U$ch0$;ywrdkm(HFi6I-YRm#pk2bp6ci6w-h}0$8GQPWm-Av zuHQ+T@W&jo?BO{yPD75s<_ZB&2uBigx`l9}k}qvzuKQ7!q}^Trr!rz<^4dO;GO6jg zNaNs#&KJCiq~S|ur1Z{G4BqsZWp~`%-u5$r0EC%;1FEXDRNa?;B%a6pNAtX2!3_T4 zBWzVS`^oC}ods$JhccCkMcC3jaHWW~FHw&7%;|7fkJ2($ z2_PcH0aQ(jvl6w6Bj+R_dP+!2dRLRQG zPY>u}l1^S>#A)40?s30SL~>jTuS1X} zA?2M{5e#-Xss99jA^qF;xczh8a8*UulWtg zx;b?Y>Jw{(pIgeaQ9fXEy)D-5F}S-4gplwK?Bb+lQ={H0%<@x56wn%mxr3k8<*7-v zn#0b#k2Gh?R-WoQW*2dh`Sgr@5-fhG)3j~=^@o~GY;)ty;3#~{|&^dk& zgcFC3HK>`()WWL$QY=#qS}D>l)lXK}=oseaOozG6O4&VK2h=;E#Si)E*?kQ5iq2$> zH@V=oS$yXHkki4@k9XD)JwwH3Pogt4;ma&Vt)tXbk!&g}k<;k7^#~5Utzc01&rayB zr96y<3M2n``G6_@&){zE<{Rjr)V8X^!Swyc=|_sIiybYVtx;R-f#4qWj4?@}pC8^G z6{HkLe zz|?k<`hFHkzcCo407nB)L%Q||oVRMEfoGGTR#Q27z0^MX{@}UaM{76Qtzktpak>&4fHaeTLt8 zPcBN-1Y>+&_wd%6u}3o6LaKK`8Vmve>RqWTBT&IAYAc@p>2guOQ@!&DX4dom84#yz ziP|ZUk4(EKBVuloH+4HO4vvU zd_{;aM+VNesU0zPQ|n@=y|9;uRJ(Hdwc$2GV*+PFydX3qyZZU3nS?`abzim!k32}k zq^Nrr8K)ztyQ1R>D-$yzw_J=qvS1+&5L;YXh1f6kP)d)#aUpY8IT&xjdT)n%2L~zY z1Uy?lfq@DEz0{7@J@s#84^N8g`2x*@r&Vf0sd{g(=-ziBEuh4*vIW7^z3_ZNMVb@) zRm6z<1+LY<&~|i~*ERXd-M>3WCGOFa;m=OuHcHdV%(A{LrF5#WTeJBh$X14udBwkE zx=j1zPM|3A6~y`Hxm33WWj15@OLQY$n6;P6t1=y%*YW->(yjTE(Ye z_XU%O-mf-sv2a+i*S_3iG-zZwGb%q&t5*7=N}Xi)c2h>{fEKEczZUpR8<0KDGt&#l zDce>!C==^Vv5s%~J7YMEr4G31KhcO_vlD*U&imDsovbc@Jd&A|tn57r0f6=aG&w6`|T${KdcL|n+01CLEEQ*A$r#F%s_(@^YcNtERITDR7z z_AjPo(dq}O|>K?Zm=Ck7%t7b z5FBTdLDqPOJk&QgjJ~)G(WoA6VDL3Y^hMiS3&8hg390F*>IQ)^t_o3y*ZnoV-Dsz8 zKGGr&B3Ds-k_#HvSja>4mzss6B8Ac{0Xya!o}qgrd=Y8kr4KJ9Ax-ux;IC0p3Zaw| z{=N`mY$(FVtaa%jyq|?nVoHX&Unq2jdIJqZsRo+or&wqb z|Izz6m8=OjSl|*bD$+Gboc6Wnhj1;XOZPnSlUDE6jGfZ#i*_PsMunc5L@s(#K!XPJFll{RZ+EsSOsqEi@330s3+sXZB`guKqx;QC2iOsfYKp9^x56jS`Z*T z8c6XX04UQ`k)`C&&BPl}Ty|Wukn@@7DqG_HcZm;`y$_aCyc|F8xaEVqxR_r6e$Z^N zGOrlV76EsVIv!AxFx+>q9|z3jXP-Q`FkoWVk-)Y5j4c7)s?AydmD1Mg*>BML>MM%R zPAzD+Ji&0XCLxGMfC^vTE@^9)Xhbu6A9wvPw7pY`C$Mk*iVbydfH1Lv&%vLAb1r;k2yykJ3sU1yUK@<$?K~R!o7$c zZ5bZ#0o!S7E^q@`jpt!63i1?Uud?YkE#n9?^YII;1D7EO~DQm1bLJ0ePwQo~oT4$x9by{yLk#ui) z5Y~rF@%vC05nI47bLv_g#r?Qc$QRN}mOPh&B6fhByzgF&ZBQuotn@;-B{Q{*{GI)5 zJ5u#?zVoxq$ru_sr}2z*1w9Y<*8B_4Rp$$j%^QyuKGEh@NY=1x;&aqIek8K4gyZAI zQ`JKi>+B6iob*HM_P-zo)W;_0l|}m|DHv{rbQ_`XZG+4*0n?Z_4s?L9vv)d&xb}>L ze_$ur^0Q-478wmn{Qj_}1P8|u&!6zir?nR!{QQ&XWZ|C_}cMoq8TGLxWO!#TU-FNJwZ*<5>E{|kDfQ5xFY zftHqQ5NC^eoV_!th@%jGZ+qh)qL)^J!4RK1)8P^KJa%R#*sf6VJTDjPe88l8M z?!DTigf8qHEAa0>yOMk{L=e1b9(sph#$t@Jrg$BsJ&PZ#zii{=EvUL$LA*%Y} z^?Io7aXs&H7sF{{!;Wnu{OOv--166@{N()iObA70qRE=wDlrp?^!J2t!buCc^plBjodLI38PlZMk0+@K4%vhG8|76mk&r1PL%tEHv;N-1F|9zFMj%0&U)oh_^D?u<&q71oLGY8AHm)wYlN4b zbRd7c@;(Y_UGEV{dgD2ozMd>EebR@gEP|MbI2r9|k-TG>~z&KWXMWaQx zm3$B~*%Dp7V;tRcAg`SDd_H-{Rg5)SZz?bJ@vOI?c2R9|teB!ZTi9j(&`LWBM1(bC z+w0`Ap9(4Dn}Y@-G5qMEJM5xTNYX19mXB0Ryo`nc-(Gw(Lt%lRFTIO=`TiWBGs0%O z90zrNzUJMfK1HcpNDV^QnriQM42&Tn1PQ%Cr5q&dpdvwu*ir=z8P}`VwEGm+B9ku! zT=d)9`S8cj9KWAW116%Zc|wwRf{^}p2@&BZ=ibOwSKJ5M*3_X~c{;IEiUYByC}kBR zJBC5S!z(sh!dwk)#JE*qqQnyJcVTmB!8Q@1aTFilarl=bj^+#ZUr(eW`>aOgWL8zx z-$_Mi4I?tfHY21HfLS35C$}NxtirM74mb@9VqFl#Q%Ny9tF?%+DaTbL z6^oA~;oCY=Rs+Kpy3OIG2 zec0{}v21KT$}yNoQyNLt93dzzq`Qi^j<9%SDMyuN^64W_<4gDbxuvC9X+cc$!Uw=h zjc;8JBA2nEM=l*C!?f$?wUs)Q%?1<;exlgaURyQwRXNghi9#H5-lKOCnr2=So=H+w zjAsf<_B82$R>-6V!5@IpS@r;T4AMjOw(+18;h}8sLDAH=(G-Q%aQVpa;C*Yx-T!o9 zBFdU)O~&5aZ+)C=u6lqVh^+#XPBiNpJEg+%TcrHZ7%LoDsoHYYh$T&;1H=4+>&#*bIsf=Y=+cpM7InWaz zv4Vn&igZP2psaEdIX3gD7DZ1w&0t)tkg`>P%ccQLER_0k4GIj@Wkah$ zSTv}#u1hZ<$|}byaYba5L!m`x6wy%FFG3d-PQl8fxWho%^&5B)TBX`|9OiU&^Xj7) z@TDb}<7bn^o)Y7vV!|;G6Kfnl#?OYt;$T;CaM9XZ`0#?a;5!as)SwvDEMSse0oGOR z`Z0PJZY|%5nB6r*haVsUrPyI0$R^irxL2|6r$%Tk^i3KiKjsqz9>yr7(JIh_M^}qd z9HEpUosMu^g9sD{(l#AM`Ic(kNQq^qwv>sZ%a&v-FPu@qDJZLeCi}upAY z{19<@<)G{f)bR{nHX?4pZPChb@%kI+PHA=&6LL+pclg@!e`2HYT!Zh1wDR;QG^!?O zy4x?3^8$OU0*x>=J4)J#h>T(=$R-g3c6PZ=TuH*@I}t(bq7-;8cy4GT43z99QB$*t zM=20x(3LzhK&&k<$Wc}q^g?9VMTH4(piE@+AoA>wslR18kR9;n1BIORJ@XFU2Qb?oAPpL$qN>BL$cJzFt2FvhZnH&hds z?IJw4)uxU}Q_L#DD7MKnYbCG&9mGtTlw+t+B#2^bSyd3sifpKLhhRFSzEveiZwR3x zjuf#M^SxJ|$@}j9rA_!&vzTyPgVSBAJHNKIq6t~le-DMYH5s_8qrlK`t z@3pFNYsclnv`$sd0H~5dC(l-GwQNK<{Lp7Ud-z0@H51FZwFf8^g38}2YIgC|q2V{s z3QjwC0Uv(tDOQ0|^`F>-$JTA*yKg*;GY(k@zRe#Hc8QIkNvtoGa~z;dLeqdJgigc- z%OB*|EAFx3&n@6OhHSb-Z+4Waog?&R^K_;POz#|KMScb70ko2e*ii_MGEB)0)0--B z%9Q=+PHEi4>g1qFYY&BD-`q6Lnt33;j+66ERe@<3i=K3mo^**>okL8@;Ic-{o)Aw0W_tXkDyxJ((RvhD?{iY#VNT;1~r&`_OuSPyq_BF7Y#R22ldMP?G zi3QJ|J6b3al%7Cj$5RLaBBRIswDII2C`Fbpl|UEAA?uDKfOH7|{jQgD_yIGjVbkT0 zhETyN`_1DcXPn5wIg^O{421Q`X1Tu!ejd zi_{W6?%27M&Ro7`o0T~|xjfUmhnU;#Ff-Fh*0p9%6=ZOXi=#BYb~tnPA)GjK8;*)` zRLrDw-pZ{8W0j+Y?o`OM>?qSRBTP!?nUNi)!w)KRx6)Uon4g=%6mJrxd=`Vk^&&dD z^K|u%($SsA^^$q01S&52D$ga5l={FySba^Ry zQyDy^$&8nVUn`=T+IqEN^mw|ez|{OOlF}F_3E7X65(pxwV#L~ITUuY|X~aa7HP0GL z%&9Lvn3XHGQYybtRP&l&NFI(xOM!xOUU3|Yw`^&x#4Bq$oIk0D&z=1uuK)Q88!b>y zLPx`e%H~KzgPJG|2c_26z$zu3sZxy*$qEob5lf7tVRCkiV&oDVMYmt1D^*N_8zN?Q z4)MPI-@{+GK7v+?LwkBymtRai9^vrrqq%g$Uzm~^p%l9)41IncUq=|xj7C-<<=g;` zP>5ZIqAa?G#?qTAFcjFRCIxi(0sD03`TW8EfavY6@cTOS)Rkw*F_pluh`t6)vj6v1_%2w&r^sLzH$fTH%#zu)-AnD6 zfX&^|R#u)4Oqte2Z(j#nlOnQCJ|^z)T64%+Y@5>C!3&DZ=Kayi{|7PZPbh6K+0alMKro-)KVdo4kUO&R=^ZMB`)=Ab6==O_fB?QK0 zB=YOrNCJ3HjF*tvCZ+QriosIGa`qKuy_h2=_3){K-obV2m-5+%F2?mrmHs=fA=_Cb z3>>DZ>FnDznGFd7Zs{@`zKpdegnt#Zs~L>#r2Y^ky=A^QC9-zRRv8 zA2wU17H`=K*pdm3RwykMIzsivERQL6!HJ~ava*ZBUy`y}k0V#(jRfdXtJHbkGw;Tk zh_Ys{Q(|^(9pe1&UC-S&KZcibS#acBHfoc`Tee zg&UWz9QV3WY?$1cBjdTmj#T~HwJ*}f>y&E_)W9T+^ZIAlx?wVY*kYe8eeOtHSwUnx z%R=r-l_&@92!>@N%h}quj(^+#Wh~l$J(q9%F_G9%;$_1R6XT$iBIgD4dSjJqX=Ui~ zi)aj5K_r@-3dy=dY%caxO0WBlVRCwmj1%B!!BIldzz>%HfDbPC07v%r@aV|Sy8deg zDJSM*2fd6_r#-Lb*!CC4n#p!8(38rbl;RU7oWvJyzLC7dG%1fjh#?YCX^mHGP-K#y zBkBN*kS>x~SQt0uPu;nR!h=#cj^^0o_MH&IWulpQpAESEzLos^!rR!ib%T=JoclZ2OKDd`OF86j>!p z>t-cJDWVQV)B*mGp>yO3JRs#8Nyba(?K>fz@p$R0TFcL;8xu{|>{Uw4=x~7#zy3!I z59EnsL!9#1e*4PG3~^meCYNH;oL<)K9IjeK%1{j^J!RboGbVQt8$(BCyjW`J>`c*- z^0|5KYHMAoeg0dm!cYxwMQOiV3jpn3?H8))#scJ`H$IfZvesVOwb%|F!f>42C|o zhF?7#rHN^6W!Y977^! z$3OA%miP5^LWBvoI@;TD<86=eUqAX2r658@hV_r^V1uc$)%>)_{s+(CfgL+5G~C_2 zqB1Fuc{BS6qKK^TkuNsqei8-8?{D+{UA=bc6YQl-XD zaVn{Ol)T0Ll!G@l#nGSnp(Weu>O?M2uGka$y%NWq`(tMIbrQxk?Ur3eHS@qxipiNA zCAY|6$;JpP*A_?wg@Y#L$3zn1D1%OJ!x)XI)+XTkO@gQ6mgL=HJi_Cnn{7l8;Hrp{ zaSN3u4*j2%I%#DGF1QNnj=4pa2B!}N4> z(Zr4Ue%^iecL+>~n-x#HGv7tEN=Mx{3Kxf+LU2iyn{pU_$#fq)O zMmX%S+4S_<)$+(to{N6`R~G&0Aw1vZWp6x=SD$k-?$czYnP_2;dox6YbAR=>N;wj5 zFhH4xaM@zWibpr`zW+Fb|G97x<><#s8`QrV6%chQe(>l$eDd}i@rqJCux4Qub_(CP ze-W2&e2j;q&AVkSW*nDerq1Hlt*cwyh{g$XIH^2Q&8n+a9kpGd!)vx0iNJ$H%M-Vd zM#g7>*oPho1#lJz@NDGr}8 zjlZl|wd+B06~BA`n;9G{aLU)tYu>LE#U5=*yNQ4oWs@blrTWa^3-+19$KQ84n^*2& z>((Kr&g$ieqvu!lKYHI9e)Ik7S+{%(a~4eFpFVm9ha9uduKV{SVxq~Ky-JCB@Q#&~ z#>$COTE84h#gHWrtmV63eiI-1%!PHUqu7JkR~3PiNxPV-9uID`ged{*CQ2+V_IM%A z0!kPNiVSQi+S!&^krfJw#C{hWes}zgIehQMFDGvTUis&rR}(8?h0g2tIfAq19>C9+ z-^!AqH581!r&8?TH6}4hNOiTVCs!Fd+-AX|5OOu!%PT8a#iVseN=GOqcq*bKtyeN- z3ylJ6fk05KW?^myv(kri<(7?=^Ati_hk!w&aJ-m&oT4#T4MZ&Ln9KV7MlRiO6)Q(q z)93f`@ZdHK5v~STEBua8wvF11l&N%tpNc@Nzj;%xK`H7eD-~WCA!y`qTza}>P_J@Drk^9#0g&&{OG{#Rhgz=*}+W7ks zMlt;x2FPs9qf0Tktt%jVp2NZOr*Yg7^SS$>`qcHYkIm98)}=aqelJTqhlvM<00>ex z!J;A|ruz_d8lqm?%0nYt0d4b__{~do1!P^H_aAyZA2{p;zO(oiezEj!3UQTWo_1V* z`NFrbWOyxqSa%=mM*ArzZU|$I*HDwKo)=@`DnrU`OXKoFa+GW<0Gq zWy*eh_Q=!m9hc*#Snc`Rjra4-M=rO;)QrZ{mYZTpZG!LEh$2TBGH!sI+(X4;lqfi9 z9gubENoY!0t4caiV6r;EabvcH-3=2>(24IkNfrsfByW_A4tZ?odIacodl`swWOT?> ze*{BSP>CpVPKmYoNrcHN$=5L*-WdP+@F&?)iV2M6==28z_zrYwMJx_M>>v#*E3Khc z9DL`|>v`pzqxi4;Z{p&WiwT?%#{TsJvLqslpj(ITIF&o7g9mCgB~9g0E31GleW)nl93 z_rRGguJGi+z7!^wVA-4T>+jz{dfiCX%x`RA(nivEu@JI!@p``Wp_lW25b4uL)%$4vYnE#R@!Z)fyy0~ zVUon4nMrwDz_zB~ZXhAQU41KWpZ5Y@@yA~?SS%8G23&~KG0uonshV9zZ>Czk`=D7< zd30m7l$TQwa+?{4b^tqv^St9k<`MoG^?D0=uwXv}hTB``c1T+p^=5 zaxoog*fhv_U%7(w{`|@2#-C(NL|L=fC^27p-}&`*eqv*vWKErRYXlx%xtaHW_}85F zqJx<{ql?WuM{HD5&9QQ8@V~nTAK@*xt>a?{y^QBiJ{XN+X4d01^Jg&{7I|LpQJg;Y z`D`f-^62m?{zCR)MJqIbd#j0E&t(x0pDO44#fF`kkX94(QxU5L>@*ARhVH zXH#*sqS#dWrNG685vH6NLE-6`&^Rsc!-%kG%StX<@t`f9S-nBSu9HWrEm2qVMvq>^@_81R+vUc>G8ti-qi9S%A$=#s%5D%PD} zdC|F6NIt$~!&9!rz~Ap8Xrnc@ZW`bZzq+-4DQyJRqL^$P8!q_S-?;GSf8%5Sc{ZQB z@W$o_ljkAOHQCb2&8DqI2Eh0-F1OB!9keTWS`qsp0rr6_!~qwsx{rs?7W81iW$5O7| zxQI|j>Gq41Vuu|?>rxpCG8lpGR1r{wM%7ti%Fsa8iP1F+qS^jVHU+&5#yN~A42t0} zO;jTb9Ab{?T8&#%wjia8ys&RMkB-e?YHF0m;dc}q(7lzZnNbSHv%DQ;BWIQjZD?r$ zszt6dkK=b*>r2yePPl8w3byZX*-^+>IWFZ_PiKQGEQTZIOvi|zmExOkJd@Y{;J1Xv zfC$-b(wg^@VEW{0J)}n-*}#{-_(yi^7;QSv7@^6M<3QY`=bucGkaL$Ws@D8La2!MIj`5B~KV_7`YGOMeFcItXJNe?0 z-!VPY!Sg52;GTg8C?zhX$48fONB_;daqii?boz1hrD85!{~*4yh1{~<5GK)v9RyK| zQj&1OpwLnujo>Pe`I$bp2W#p`ayd63>jXri7>&~ulj2S)abIP0i4LdKQkm%UN<;}E z&ew6Bl9ST09M|Hf}`;^oI5 z$c=w_lm{2Ds}mx|*QijMxDs{kXw8)AJ>%?sGBNQ?yn8Ed|I4GzO058~L7bYbA<1)4 zO!8Cx{2SM??+MdcwP|O|cAXL!ui+&G^`x0`9;E#`C-M424&i~pwcNbraiW^%zhol* zvS}GB|GI;1g^^17eS=_(aN&waxM%1t&X|52i?%OeEG`n55C_d|{SR{1j1zeE>=*E_ z`@M`+W4E)pXbG%oH>S(qNi0bgafHa&8d07Wf);^2p7xlXD%Qn3Xa%!V1I&~G2Er`M z3sVVn0bBQh4QJyCMWh(!6?icJsjc>6p&$ zf|c(rFfHB94a=5~TRFV8c~Nl1J&$tL-Ag%a=5)UB@)z@mkG!29Tz?z)FW-PMHAOfZ zx}%h0`@jg_{l$$u@1Qw+^Q%`8SX!jc=;GD-wJ zRzy4paW&DjjyfSxEQUOA?;4bV6G5!3gE~%AFEoTo$h1J!bStXrMru%aT%|aA#w>Qm zqj;+vn;8@E{Rr2!FZS)>ppE2dvAPGj2#ykB zG^259oT)FQbc`m?V4T8H&Cyh4(q5F9n-$8|V?sa~I-HWV_y8s2G8*|c=L&r(IAhuo zT)bvUo$}{vms6%LWKt@Jt2AM&)a(r_1n~^27pB0 zzir4WtuEBcP*IRCh1`Ar8t%Dojg^=wYt?tk$0SPJUY5qJXmwp234V#@JLGs8eRrj>kFszJLc--q3PP>t90l zBc33E;-O7j_@`fA!jlQkMPsSZ?4IcPTjL5B5d0+%B6pLkV~(75VC^u zTvEeP)p|2tk;rlvjPUyRoI39QCleD<*6bBZOtvGmiwYhY@@Qo!icQQ_wVv3}Dn~_$d&kh@7HoYJs($a33hCAf(N4o8SFtXb)I2a) z9#4-u23WIm-uU!-`Go#Ug*cqn5y};IhF;Pa!F9%XX!Ld_Wi$iD9KLc1CFZ0l`|-7- zUQ02G`O1BNX&x|uT0eEeZ&rvRe)hn<-2TWqmTuWbP_RLoWd$~_b zvZF>=IU#gxgt2LYc;mOz3fuEz+`nxjIVXb_yJRWncgm##Vx!0hKCR{=qR>jL(lU+) z6-vY<>;KG@bh8Vi9*Sb@<13G&dJp1L2mK4H$2Ou-%+Jl@+70i;5H<#DR^TWfBL)&# zGuaUlO z>)HW@cu?wcNe_nAXQWajvSAoRa7GH8`pP4B*ZPx#2`Fp!3MJ;=zb$28+mNja=sGnk z9R;3)C`FWOu8=H1oGzPHtX;r~L4|@O>7W(XH?6Q)T+7=kEsR&#R0$VsxQ9!&Si`72 zM$lT&DoG|q%}NAN%2Z5oo)cq4wVdqbKE%Wn;(W_c>RR|Z#e$9_IAi)h(c?{FN8v7p z3%5`xb&>HFR>*oO9g~ufKwL^m@u5DiheLW^%l-YoB$8^$Xr(}l!JyFy-MYRXWe!b5 z>nO3ZY137PPE{nYJ&GpPw9-rI*eXy|3rh;a^rRhnGa={9dlM&4KAgTJ@mJP`AD{AC zK5*+F(TO*EEDVsag{BFJBVepmZ7c|RVEZPNDDWUkfm@(SKI)|4TdQG@eT7>}R(8s| zFfE(2{>`TyP2ZHKr>2rH@%+?#D}M2v>(QzO|DYD*Y>u){(;>AqJ{2d6q!#VQEh3Q0 zFvPt6FTWz~x&*NTG?`Su+(`qBm3+1ik5l>@!E+-#$BY*TQSG|~yDFlL2!uv6$?xIJ z=_m8r*)Jh9V-&-?84A}CnS;nUQvjQ@Mgb`zLUg4F?fzY_VN?DdM#6O%s)|9?@Ff() zCrENz3AGEh7ULo`@$4iysc|&Q`Dt7o(d*|)JCS`qh(fC@9+Ua0jVbnQzok1Q&M5StM-&_Wv~+o8Z!kyEwMuY;@qVR-?PXGU z6r*FoxXDYXoNd0MAp@m|R4K$yHNXF!3QR;&kT{KUTziG#1kUkT(KkN-I6SGd|oB<7y2A7Mw~PGTvdu)LY6uBuGF4qZw-(IzRz zyn#8Xf5DsbKJMM|3qq5}(;2$l9^&XxloG~Fs^#SC(WC4zjRMq?`Wn%+GOBC`FL8}k z_6zt`*C@mIXb$#CAds>eb(PGn zD3*MTQS^_MT{_L@c}~o3R@IMf=(wsf>1-xVa+W;6St>LQm?|uA^ez-l5M9F|GrrH={l8(u*lh?xkH0Si!D{BXTNpDb zmdDeZ`dF*vl=n%pipa-)-LbpMChMx0>6zg=Yp_B>k zO?O)ZN1-A?hX&(W4xTumwho}jiZE70qnP!)1$6g4T}5aoo~+p`l$f1chpJzK2r{X} za-zV~h*GFn(1i$-wwdE%4}vb;a@BLzF$Zzo%H`}FD}WJ#^5iYQ_>xm%SCThceH((H z{8nxtMG$J7T&q;y)-;-SBV4DP`vn{oHzl6Sx#hx)f}^@o>}c@m3ciX+Ii0+1{ugnz zoye<(e#6%M-%!LjfTIH3ttKKEoNE&$C=&zKF-~sQ35;7`YMMP>$QzU`q$%F(g!!n`b+N(KU#9C(V>|e&6{p$!#gjO2F zVKi`=J#amAO_(3+ypjTwlfMjlx?)$J6SojaB{&wWGX_&;s@IDr`fz$fN~>9u&; z;3}!Ox}jQHPAh{i4>Q{T8cvyd9ZG)*!QkpNx37N{j#*Eq6RX=Ys2?jo_lt zN^$l(PvVw4SFmNvAd@F|vZ<7Bsld3dX7P@#?93O?S`mma+wWv+Jch)k;!+YNNUFlD zqOt-gs}q^l))9Z4dK_hV+9s;=oS1B)tT=?Y$`GS*Rg6*!g~3rmpSKZ*fP9=H6c z5XS4#%0-Cqv}qZQpq0xDroERrxs$kT-6t7|wh@ivf9vynM&tUi1Zk?DQifeA30g&r zMWuvOBnUBR#il~Ux0hbUNLb{I8Aq})*h!$Q0@L)fI|d>L(TM_D(NZlRrm<#{IEpAO zxP_XK#UyjD)A+Ls=wjG($59yb$KTw>pMLfiOk~*iklB3x7w@HST6fF)-BV07vfbW> zVj*DN%56kJROKH;rV{yxQlv8;eN(zP^yqmk`tu`%Va&9BCo${=E#E{lrgU-MU+zYA zxWvZbh$1VBG3p;8ai}E5v)3x`ASp82%Cf2O-55t3V$-VBhau&}cseHS*g32e>?kS% z<5K%V)d=%5J6M?A%+4sy=28!#ICQ&tW~PRim0-bKL zZp~pN=F$GIC(%M7J_XWB8q?+^SFV?^236 z|3xL^AauDI7Iq#(#+${9W_*~dH+&u?&5-~Ar%rt>mu$R}VqE3vYlNJm8H<}u+2sg^ zSgiT7>{kaWr0G%wKU;nyKYHjs`iprY)7oWa2SpI4>}wXAg8tF{tNW!}k4N57Tyv-LQUeb*aa9dMz6yRu~FPmq_XpY)satcFVE|05)4lk(l za&>wE1A&8xVv;w?WPg;|nL&EI0s>6(#_09()IRBB(MImr`rn8cLJ2D_qN)<3DT?Nf z!gOAeTi+BCt(1+T6oG;W{bSd&EqDpNxz{tf^PIY4_e*W#{=(cOqm`uyYlP#v4q<-o zaJ0%WFS{?xN0-uKHsWA-%bZtm(fVr`2z_EvxXRGy=WU*yajYRqSZh&*Sg|XK~pLi&?d)pM`U$^0u>%=B#i3 zir7>Ym?#5bVm*=M!8C?SewYTjhy6zhaj0>%sc|ON%)TEd!Y47e>Jp(iM5573=+H$n^s{)=%UL-2N9@~iIRdq!HJzkC;e?E_muBaL3B&1$6#mKEwU~2Ecbz+7J5QIgW&q0H7q#{rx zuK58KkQQRG0>c|hR!MG>Gcm0EgjPUlg=kyTq*AD@c3?%UCCAqo&p`)q^KV{$n1DDk zY+SvK)k`qkK7~JQ2U7`SoRha3jtsJw_j5jt1siBw4!%%CtZD627nvySw(8e zdVc%+yEy5j{rU8#&c^c`(kYK$e(9~;c=zL6fBRCRSa{Jf`$0Zt|yno|=9)Aym zp;u@5&FX-88J}YO0FzRAy1aZv)@U@(pY%fR*m)a~)MhoecRW|d$4hmB+v;aERLuQA)Y{hM@6yn;PU*3qpDK*!XLIBUR?(yBAm=E~< z_pg0Ql$ePqYbLA`bK}K#Ruvef!1oZxpFmsZ*wE|iFksqiq2OA;-RTqZeA#}Ikt6m-{d+&b;{)Hqh^>U%sS37iL2|qSVv4NT@jhJrI}Y!C z10&(Lh@@D{dqm-|Z1_5S3F&e=$QvI)lTsnRtY>a+3SHjO+&yp~mP9C|bV!FQ6e5?Q zFxzwweIEXs^g( z5qaw#U33EAzFS+fEyV~sw-2+e6cA^8(26KgbZ)-+F>b$oIUoJ#nVj*`!|*+qmz{Jl zFFWbr%K4u=|GK*OO<@YHBOQ11v+3Rz=06~aJyva<%=DfidNM`2(**>LuH28}`}QVbtwa_4&( zF6~btT!xVVkBHfRf4$`yD8=hpvgMr|GVMI}%gtb0X@%`+B5Pdb5tuR7j@?dHrx`FE zm7WR__YT~`Yi7NL8@67{V7N264{N>Y@yCE3#*60AC`!9iU;x6JI$H|rf;Sz-=zJKR zXRW(#LD5}ayCv_DQm=<-NDW4)*k<|_0dW;*2dAq886Mqb24H}}?Zdn8>py}gLDoFk zctQ*vrPP^9sbxy36-ud3jdRQ!l~Ru?r5;gAU9hJei;@bmXbss6xGqX-hzvopPn6A^ zN;yvqaLeBw}u2Ncr|SmN)Ixx`)!;w>uTn7z6nRA=tvZpa+F^=?~4$z zY}-E(7cXRP?nGRb#!)Vw&fq8?!6lZEqWJharpGPP>*mQiA);b76dvTb-hJtCigY+7 za&AyrAv>puX%Js0`nYcz`4N7fP;azD5VZ=IA$U6Y2YLg zIRyCkC;J|r2{{u6HLLR4qG4D?rmVm;A_!u>`ORzi&bimu)ElZ|aI_{%tCmVO@V^U< z2FS*~$1dypC$VD7WWXksR0**)=LU3oC2CV>Im#|zlmeTxcf8ArEhj=mie2+a8quvw zb;DR*=qaVp$_`~(K_+R_$Qb$H0S=yi1^Z1tkBol=5pJDTxz-Xp8Z73X4KL(`z88>j zy75#7M>+U9jZzvRrerijaSpAd@=*#R3HjT$Ynkd#!o}E+DhECo7Id~w7zp5KP1Y@w zqFV?M0YZfep1>>BCi$te*ejr8b)NK&5FLtjH5v%h4*6Lb3e(dRW@X49J`0me@49)d zIQ--Tci)eH5PKqPo(k-C#Gfevcl2q8+9Y=ARQVeAf3PnHrIPshrVcYaSm572@@rPE z-_8TeHc%*9LDii}?N%`C7K!!}m&z%f>_=-uCLL4`QrKKDeW@|BUWls=8MjOTEuUxa zC_+~f2$f88l!HRZs(@a%K&Km^Rdw|d@8T!4!qvLJ92C_buUAl|3odEP&R$95Pb6i$gjO;1zRz#4(fql@q7_JF~MV6H91` zx#J9cEFsrzy{1l~s`N?0J~Fmw5zV05P{NiZ!c>*{NSRo>-L?i$% z0$$Z{%zJ=yMP$HUPqy(S;+c{&N1m`BQ508ytELBV*XLhz<%3-Irw3TPbOS5bZm%db zzT@tZs1a<@Nt^%3ZmE|WNf!}Z$DozXitTB^)h6M>2+GoFE4Ndg_mMd4jIu2fmr0TX zN}A_mi|f~(msU}tuuLJH{byYNg(%#nsiqZVRe27j@thebrRmAOh{I-FL09Ui7VTF@ z(@2SpgJ0vEeSXI2Gd{+NQ{K(lv%f+l)>1ZF$z0hy`h#IVtMkj6?hT+XmB&*p^3I52 zPIe0WO>#;5tuqJ%wgBxDg@8X;-%A||slkxcV1$l^*w+M;{6zb-9X4r)!ZGtPUTY_c z<7zUw6zBZsYoB;u{!Z+Pta&OHDo@0nR1t8=1vWYQxUe5nj)u)^tq#z zj>b5+#FR}#0UBSbX5@0)DHW33%Ka{2%Oc9zzmaWT6%4O zgdhYF7;)KF8ev;$gmg;bx*>xj*=B!j`MZFQ6e;bZm14%s4z_F_uH^8j+<#5rh!a|O zV^Y@><0+g%b(|U@zKbcw=-3v1!wC2ulP2YO|5x9@NoTbZ&U+3KkX?s8PF70YTl4*= zBJ$Ik@3VmQHQzS?$2YXGKY&u|Hms5Siz4!;f4tXCct+%m=Q?zz9d-_f;MeCIQGf~! z;*bAwiWGeB=Qokd_{2t-I=Pc?ede`H>8UEK$|!WH*&5?0%RU(|5uJFzmu5en#AIvfQ$~9wS zQM;>Ej;kyrgg_vY$T=Bfz?fiz4cNvQV`Jbkwh0Cv8yk!X4j<-O=M z=bPH?lczsnJ%9S%KVaoy%l0z{51bBw*9@f&MdgV&9&?RS>RTdmTgN$_ips1~>V6;O zKzcKX!Sy=zj|PrbO8r79bye8X=@6?O{ByOQ&`2^q{J)p7W!ns)*MX{O=t^xY20q0gVCf;{;VP8x7P<;7 zQIO@Dk2B#0I%B-hz|&1FM$s?<3z_pvgr055UCqN`D_E9<%P8vxYs0-v`4z^zI!nAo z4i9&;BA6#s29IWEUA9h;XALPps{&JnbBR!tRl>N6Dd`MPySSN&;#0`dyP4VjD16&? z(R9ghN}&muQWlASO1<`37U$o^-d*Pq>M6A9X7Yon`+Xcc{xl+G2uTUZ@CE>CdCKRs z$+HKyCa`||6pk5NMmb0*=sKy@T}2=Thm>3hlfG+hl9!s(78?lG6Ra=1`kXV{vp?r+ zDs}Ntf}m3Ow`f!4J%Gz%D*dn2`j=m0<9@}1(t@3tc(`1%44q?WnWfi&X$<9BkVIH zR88pCk80;JHsjDjU1BU)Lobue2$(3I?F5*G4-?P59hYL2U*CK_fd&Di-U!p40E0k$ zzag%DhI;Lj%+LHGi}Qa)IXG_U`|5cE?Y5O-{rCkGm10ykD5wNIKyf-{;t@o8uscRc z#LD6%lTpH0)FAR>ES6N-?I;DMLW=4vchEFY#ZJ<*zAJB&Y0ELImh#x~wS+a*Ixkp+ z$-CWL=$uuJ<&n#vZ0cS90CQ2^2`L0LPRN&s)}Wlid*jdVA}U4w*F1#uAS^Kti^cb$ zLajL<1}7pPZ!*%QOP3C9>=)m<9+SB@EXMSWjuL@=i_i?T>iFLH3i0}1ew<`{{(IN* zyhof!AveV-z>lGndYry`6;C|wNWS!{7x9L(&+Za-@|(EP%16u#ExLnAXayr-Og}k6 z#+9cZJSRhy@syP>W@Q-XK3w4Kij3SC9N;9D}OO$3zDJJz8%cE0JBtSE!sD`yOnEF>x zk^q*70hu^}Rz;@E&u{%b+qjk+cKsnsgL$Tc3ai6e)`fdm9qh%=u`)E8XRUraXHA_- zq}z+H33Hx`&nGwjKfbW#Ye0gLZpV@cE(Kc@VVIFiwr){1OA&ZDAG8)4UP|3&{Nwk3 z>=2tRkhM1$Q6QeehPic3DQo~H02L^vSC0*Na@P@+8^=T`R15W0k!HfzKY8ik>mCvs z?xQ|HKlL_XO~>D>^UpeIEATB5NkwEMa2IaP|0d~x1*rq&HJ5$kI)k4O#v zY^~3cB7h0|*{2BX+OxnRBO?@}ptat0rWoC-m|P4v@5sY=#UoDR>o0pLtH#C#TlJc< zO;0|qQV{rtiBi2cyERPlkDe(A8IJD5M?b; z73cz|E_oGCTl1fk{bl&d0jm{N$QgwS=lYjW-FX^G<^Q5+5TNlV-g66`h5!b)Xvclms18&E3KL(QlUMsjMqfUs*s78>KiB zG1N?_6=u4OghAJ)Vlu-|FS>H@H4h2hFRz(yDi8VbnvRPfQbnan2jWgTPzLu)5lMlK zN~tD;yXFAe@24i#h={Ye0CWl!BB-$-!5+4yNmDNbvMsG^*D?)-8DHWAn0Ueg0UJA2RP+>F5Nn_uA)!V8q>Tqudp;= zNjR&@FL(bpUWeuosqWfK{ zn|)lj^R;BQPF+IiC_7M4hC`!8mXE#$MZkBq{WDs5td3GDb~SUUX=MTkmibhX<#_*| zZa;^ybePATls7UGQ;&U;g@`^GUEwTg1~9q#y*b9J?Rf$y7D9gVu0P?XySMSL|MO!O zDs__7uw%YTP*C^{+h*@M5S}u*#yo)URpth#ojxQaJlLSRYk;GaQitKz_KShu%%gAM z1cy&6rKWMr=H3T$(;oWG^BH<(daVJ)s2z|kK$76a32Dj0Ocv3xOS=~jw_y!<*_jpo z>_cBA8ZR(4T4KKH@H+})A-k70xNPSpuGoDa|9i)c96z;`duC_4f`i2;&uS6q1Jx8J zN`vQ8lRq>Q2L^sm!nHFOlS$iZ=_xbdd_`SC)|by=NBs(#N!KS#zLeu9UW8JL=db$$ zn<|$w)3}qP?B@t%0VHjea{*fQ%{oOP55XiqVc_;7HR9?=S?Z-O(Llf|MOFNcD8hiF zm%I;Oj{>k^{_|{H_$rcHeLAx?*F+0UN2@t*@>OhF{3;i1dJS6ntc^4ab|sZG=HA+P zp1IX+C044*@B0UHq?l5QQtI;ssxt*=vK>HaVhQug-wJ1~&5A$=sP3~Kb0Tki`ng>A z&1?Abx2{D3MLo+AZ4|U{E(VZjMW<0HOAY&W%ndyEA))){HG4(mj0bUtPDN!k@H3^< zdJ&o1-v#$azi|flsArwdh41?|wh5__8F%EX)J{(zzK;}qbk?!>J%u(5cLf2Y!oIx~ zN`-*`di`^G=@-9DVmpK-;0t%%gbk%VfjZIhm4GKK3nn;bYJ$7#w}G;Rp1ZN+i`H7z zLdg``@udAEcuJ_5y^M6g!M0N+yGbdNi@ zgujeh?H6>Kb6agQAkIjsb|CP;UOEd=)0z|^EYmS{aiW)}M)1v6?%aBebLojNs$iyb zwRW_oF+|k)#mqNcKwiF1C;Y9PI(`=4-@O%^GulQ+xdK0TzooWi+v-xDwGsl)k_?*! z3TUR#`=Ao=8=x}P_2_Xu<@|Sjob%VMqf%>-r0w~Uj(UiKu#!(uBj~I{oNA)JiXb8= z5sp3Mu)+O#sObJpxx+6N$&^xW04~ODpPv$uTa{Ap2Cfy6Z{^=TNhx(3P7HjDhzw8X z`1R?xB1oP0tkd|;|GPX7;1nmwG|irQ5l9>Pq|+tvG`82HI(U!6OQD{)4gLI+kKl+? zm-8=IU*59*rPfkQQ+{6C;(#`rOD_jz1Yjh?FAAqESwkg_xnp5Bt0(4}EDtUU+gu0L zETo_ZXBB0V_BnKivrCUht#}*0@@be^B0bK|`c+)9^KVf~bJ*z9SrQ$AZhRk4W*Nk5 zT>EOk-xdbAmYwpN5_Q1UF4GCIWK^w(LP*0WaLwNLlI{M-{F=0b2lU>{Fse3i@BEk8 z6W@)+xLQvdR(e}lHvSqmR9XL+fVp66*=*IF>dgKb6NaHy)Qo!}cTfWO#3 z9^bs`R<68b3!{51S=Qkqqp76HKtV>;Qx_A{!2qLD$jOg>_%f4#2g7R~7DeTLg_z$4 zgIjadN{%}1Fz&c+V|V(Zh)c7!770A?dU95bvwrxCmTYuz-6UmPYISBTAGz#Gk^yL| zrir@SRMRcA-dv7FvbSF2^QXLsS6+P^v^1Fxs{le~6`^Mt@f&z3Qmb3SPX0j7z@rQe zs~G99A{p@z>1RZqDl!(jpxMx0N+bRSd-uHz_s%mK@h_NPcn?JiDAnnYZrXx|GcBbQ zo)j_E`@ZvV7MMu2m!MJLVMr}BSR^$r)<(qfWuH(PJCK&sFQtyFu7APFIy6<3?2)eoo}!==}3K*v2d zP>QVV5mXFb+&w?3Cr*e-()N=pc*J?f^25*mjHwmloO=F=c>b@E$#5X_Ah8=n?zAu% zzOLkr5$YmJDN1S?SMB{B5^8D4Fs>Pm##Hx?^vh)J z92&595<69@!?dM@zVZiw~Hhx8{R%zT2% ztUKZi8Yz;bt!qRCJHN<(zv8hgpa zvcjvgHmDL{*=Hge1Y`n!e!Wr(Expc}-add&k&H_B=gJ8(ZvAw3JhmQHh9wsaF^T0= z&?c|l)=k#lhi2K(+{wr{B*u^psS4|dLRe~$%okcc7IK$$Z%B=XYy|9_pqr`l)+G}l z!=FF_QRvf1vfRaKi0<5hH&^SsrZQjWpI-h^g2=}j%VVB>27mdVuSDOUBq4yBgKTd5Gf!5m8D}H3gOgOPC5Y1f<*(FB`qt^%9XOngvMgR#O+YpPFD(lbJ`aDPF|M_+ax(g__5l{->&fga zjr0zt^j;1R<~cT8U`%HewPlrGXSH8(ZuE}(E9K+q6?lbb1N~OV5RVa+F-}AQA0JJb zfRY4#F*HpYmO8DmjHpHj>_j&ye3bMT?6u;oG@j1r=TGkMPzNk4lm@ojI@jtx%aKA2 zMWZl<_NCf1WnV>B=)X%8`n>p=XYiM=emsGv(Y1t;Uu+=QwK%|lr$7Xg85-4?q>*y@ zx31$ypC6p&{($H~@){BOo`@VRB1eeGd-BiTkqa=0qrVi9<3(h>h+B|55m}`D? z)46a*&SF{W zFEXy`jHyLN)Gh*Z8!mIm_M8csNSBpFsdl}k_Sj7`GVy-O<8NVf`U`lpv$O&{N-D%t z8ee-v+9OmRzSd~1+d*anXwW2Th)0ITQ*@+NyB;aP!{d;VClKl)x`WAyK+Oaw6kf9o zoD!zO8skBoQ9tAPYaYkn9P=8&A=-sfVAA348VLt5q>e5&84HkCwC-SQ`K3>N)bTvx z5r^~I*FOQ@_nfN}An6uHrB?6bOd`V0jl0=)_l}``{IyB;bNrr$(w4h+V27nw`XN)*N3Oh3JTN1hhnQZf7$bR)nR|n-yC-#t4qhUzEiZ5Ehlu1b>R%B z6*qBi>1Ixeu3${vPA{L)3KS`RhhpX|z?Lm(z`XVI`brudoqwsZxD$E&xyKT1+1FRjq~~F~{)}r0-~PlU`|ba)O$QmTIcRGqrAfzx^gQrg1vD{+ z#oY_Aa-wVAG{-L<@87%>;KhQCl#2}S5n&Y$Ez#uPdA`Ta`3j$Z+Kc$bj!k^%#-DTR z%nqyt9c1`{B|}Jc10}BhQJ_ti)g{!1qL%@R*aj5D6>M7)EK)bu@bzuCV8w7q@i;+Y2XZer_eDP~n^X-eSC8?#T z(Eyain{POaPfr}zAkv;9t`81`d?2KLZ*h4*lqLqVQ_f#^7%w{hL_TrzwS4ie8`zo7p~MmuQan${fW7HF zzOT7w&k_z_y044Z*j6nS!HVM(B^`72_&xk;?>i`X1*{kj9eqAqYqyhjWF~10rp%qMo4JPPz`31^y%H+{( zsm^oRo?CNlssWZ+wLg1&_k$`+8!)q%4{1d_t_c>bi|y7h8^`?FdoCjMJ(AR*j$1~Q zWrRD2mt*OrxIPH_ZP!f##Pd*b;&2_b-?9*hQkbH4 zwUqWlL+d3qB}om5F}&-NOW0JK<&*bZ-6}BX2N^-=TuPzetXAz9NNN?6QN=M4EnQ5e z`P-OZW5jn_Lt^vbxcY4zQdmo<>m)e{okAGb^*+X8Bu`fy(0>J_2vOg|d3gR&l*)gG zEPfbjpTT7jXnaX;0Y`L9Donq`R8Pr3s(1)B^kgpES8mpm6ZI5r>{AUi#A;g2qNv02qNGGmTYhs z0xD5moKp#Sd3~ip%A88Ot5_=|cnevx*s$CCip~rkRJwzrRiS`X`inz`#fU>sSUtFd zzb=Uy5XAg@IuJVS_(KSzfciqmd+S&WIyLw;iwPCBpj+STc>lPt!$^rDoQ2vNHvu=y z?c)XC`2s6Tqujo4kK;AUV`RF4_M|nZ0fN9f9Gb-umX<4doi^xnWVp2;oHMb3f=OH*1M-eKI#}?+PTE#tCkr^A%xsa8DvW`i`BM~28k8tX;e_=(*;mNEU zIh%DOXLDa|3-7(-pD^rybDxeJSnQKlp27$3_y)$bQPP|D)yh(yuCchg%&;`;b_XGf z6`oct>t7$YZb)AcxwYcrVE`hZ+9|8?ezpOxcK=F{A9|ekoJZX6c?V1fA+I?o-AUW; z*~RTYzm0g`95TOv%*|sfRm8SS%Il5-kJeMzNDs=`19G2n9@{X`BH5&lcT>+YK5^}} zEdd6EAZiD{_4y-l5Z!_2im)|<)!meL1=Ys=dS8-*k(4d6C?(LW%)+Iu8WRKQ0xaClkpFJl^{M{tVB1h=c78A`|cL3jN|czrtocl4}R(or+Ub zBf(FEV9_`%AQ8kLZZR+e;;33pgM)9#$%-)7XmH!U-7Q`d78A5L3}OKi+lHfWuD{L~ zKIzv8lo`Nl5-Q2hSQ}P2EZobvC6NA- zyskBh9P{b`#lE8QDdD5hC{-rFC(11*3d@K#oD@_z&R<|uH9GKIoa;H#8Dk!twB!w} zET7k>psg;hVJe*3Z+}7^@|@KUxU=8|6g-QkjDxZ#d`gx&FsDw?8xn;uxwmhr0o&deVwR@G`-x8@z_Wt93`X3C{Y`#CPyw zn)?}|aiCQitWV!)WF6*W-4nNxn*O14Y0=Or;s`M2*D3f}UQcN-#WtZ%L?VUuqU$TE zpjZ%}CK`>8;6t7@`k^HpjZwZq`Hs1$RJE06_&V~$+5gb%*->AdyX zkEIa#t+~*gC+(MzgPjEm>nm)aFkZf)3E0u16KHfU%=jK^asm~F;CtvO;62}YD`!0M zq=94bfXVLX_&p7zZ+-Go-v9bfvpBndAVQrNkW~Uf)ez1T8qi92eC`H`!&)=eEetwUMwUH5{{>mUtz7E+X|lcK5HLaX0{$ZINJ$}+7j z%e7%b4{lp2D5x5d*#tu^9D}!;Hm_;6r6_8jCEhfn!AV3N{H^nT%RR%fL4wzz$f?$m zkea@yd0t;gpHfTGy2QB7x_MZ3ey}@ND}X zXqY~HmSxB6VPefZOOBW!D5f2K$q(zGJ)U&x(R}0WuVD3bzarO_y*x-T$p6lkj&utx zoe%*z-q-RqQ+(_f@1iui|Ekx&HXUTV=Ad&YeeK`BM{Thmrvz)ACC1apXbGh?-eQa? z1`aa@Tx>&PF(ciYZx*$_Zr8+VVzA{iOaaQ!Lj3iV>d=^+yCP}$XixA1SE#X(dK3c3 ze0*&PJkztBj7q`O)QHoHy1Cm`y$`VO@DKZ4uio9D_&e>4e%%!U%urQ{`mCMdFI(C@TJRdWZS+4;ylyH6a^}75Om^F z{JUDv;&RAD+{RWeIPgScVX^AmpT3XseS8fc|C`VA;miJ!!b4T#q-hrVJsmJ*so^vK z@=aHAU>F5~duOwCY+}f!%57|_rqrt18*monR+u5N6Xa)>w`k_ZZEF!L1qIJ1wh6`2 zm_ji@DZxlHBCIW!h_g2csrq>S{ zM_QnpQ_8SU{65F?)`d?nU3vneQCBf^M1)blz&nnAH`^OKsAg5xmDf@5+Q6KNP59-$ z>!_Os>&K7dxoghlvm1XzY||dT+@_#l2nq&IJm#{&j;{n{u@U2;C6Og4Tcw~I zh{`Can6hq%%2|{Hf0JGLX6yRd3>(uxS(%|VnA&{Rr9S&j>FoX}JO^+5-PqbiNc|#| z*$CL~n}`tLVHH?4q<&C-|0Xt|8K8Ly8T)GA<0#rfOeNjN#XCO0*6IxajEAeK*imMZ zS(X>p@Z=RQVP)Y6-gEnZvZ1osjg?qReu+07^Hl!xuFJ5#5Gbg`ZLQd3U~`LE(GOTZ zxrS@zw)gcZHipYLZ|09rKb>M2aN%FSgs=YMW`1o9_>cejDJHgE$h<95 z%7N=OGFLN6DF}R*&8Et*K&w+}z0US~cJcjBU&?b{|K$67GY3rjb(8*{`e^5tJ=81v ziHRr_^1^yxys&l4+M-ea@;wxn}1U zN>huj9$Qnm5$m#O@H+Mfb@-!L*$&ALknP zpzsu^UO2Qho0q2vlrYH1Z6#)R_1j%S%;v>QxMt4>iOnoZ`DlGISIkEkv79w^2G3jl z47SyG@~?ONyTf?GkXgg2lj}Ke*@;}T>t+J&QAty(iHm~^eW$=SA7Lac;MtIDK$bw{ zdEB;p505%@EoU9Ome2mpOZeQ6uVdrR8BRKEC3ZgH+uyt%WfkdIKNgxBMxl7euKOMB zq>=KecYK-0zvvuB_XkVzYtq5SYYr+QW^T^{yEg4{E~7y&Z)&11=t%)yJtb@6$|#pL zVEmr*Ozu8{p-wby70X9QIBwZeo^{-@tY5L5@9ns?FZKlyvQ!a-u5OZ23NtjQOR!>z zt-+`aqY4!LB^22}NhjTNRB)hE=57}9Y@R?2CFQ{K1Q_+aJScC_ot#z_p_|bwfU}_Z z;t$V2NUjDof;aLiO!5oF+|zf81B?TI_eR$$S3~OAC+XuEmX{9oNn!3w-{*i#U2>9Z6c| zw#EVl&r&z3+bn!1czVj?-08#FylZeozgCLXV{O5#6otI({L^{a`KL3#Sm#A=`81|8 zkh9m!n@gJ`sl^7K)9ytan&x7y&n+0&6MpuUUp}-#%x}pb{kuw+fA_|2H1JN+x+TCg zNdXv(S4*%XJr=ZPJ4zi8gI5-&Y7yn}GUu&4j29erBHuscdcLsb@-C%ADHpzFrKwQ3 zFO4XIA#EG#Xkbs0F_*69tkLbP^zR|;sgI-)R_g6&sSxFpIYnc0zEMyRD2?i?C#3Lk z;E~Q2q4M^ixiE&?VhqGrqDEec9(^08{x#QTABT0$AwBvx==fLItRHT}iZE|24J5y# zPrrI0Slf0L?x{&4380SQ?2C`i5mAuAbWI*(14E?iggOA+?reBW;1=$U~-i*UKFrxb(8y zdH4T(lYqAOKxKkoGw4jPz9K7mEGvc(o9G}6qap!c|xLeJx6r<^kIk^ko#SF$kQW@w_? znn=rvFHkOreB?i0!S^n@h8wTiz?PdgV;1Y(0*tjSnDCK@P%~7WvACTlmV)Z(-z^1t<kj18~|sEo)z;#mCp7wN{jjp6ROjVvS%owt1#g|#7%n?9NC_b9%) z;Vu{B)MJhFmE!2>B^PT`!FJcb`$ zcsVy+ekYY`f`lDtol)rcOuc~!4}DC`LB?wiDj_BaeLPR0Jg>VZgi;PGWe!)WGZh%cmpaqBIoLNUSfI($7qVtgV$ zCC)qwzR9EI-M3{w&KZj$mViKF0v!@tkHG5|)<$%FfCi(;&}5M1PBq9xMwv>hYm5H z?!?%Bp;fC+kUTKrqI?EXJqz6wQJ@1?)*()vuDUm_aN*52(#Rahi;V;wca$MO5un*V zzsQkOOZeusxAFO(|B70i@`RI);=5PgMw0UYZ6HjPtI^eQe+ASkF@N&HkD@Y5p%_3T zb$BtI{%oxw3<(jQ`PwH8e3l19!tVoke@*fO4}#pD(JRP!9-XBsI!V`4OlrXQhzbF5 zqkA2CiF0R1rGRtSAI^`jyQ^QX5m-nQjHoQ50@3VgAo6d5f&8?uW*e!c4Jf)Lw;!mxc#Du~2L^`bl+ zae%^!xX7FIcIWcR2f&K3(+)zbvXqsj((}Gbq<9_0j~COeL;9=;bw=HKWRjr=y#%0= zM12BGlTvHBb9Rnv@17$z#>ER6=@Mj&2$A)ft=GvKhEILxmt1-CCZ@(poO$FSES7}E zgqxF11n6EqgTQzG_ey@aVHb!nF+B=drEeTXYbX_oiXQ9FeFzIe9)!H+pb}!ve99>< zz)7h#gQ4v&^{O?@=t$Q!0^n;@Y9SQnbw9$U0=x{)JNa;)`PdWrlP`Ru+flFcu{!rHGBkR8)0HwF4?qTNFkN%c4W6Wp{Ob7xb8XyT0?+ za^kX&QS?{SX#A8cyR+*7bl&gbA-qB+vw5bF%8jt?>lGk#uVxrlwlg0Kw*EsJDnv=O zWlkQNY^szK=z6}6{B|MJrP1Md%EL5uA3qPv2i&R9k;~uEIPB; z+{Jg(?wuWd0$Y&QV)SZ?N|y5H-~0|wTz(iozj-6|JfQc&A6$cqrRBo3N)SB+Mo+-{ z8ZWan7V4-h>lY@K0^ehD*(hf|`Q*W4_5kT%Gw~km^prn%B%AKq=~RIZUegNH)gVK{ z+`NSbX0k-m7@$Z9j3=5`{n5ExbJJ#r1n$wj1`PBj`AEZ~e%}bC!@6q;b&^srU8pi$ zsNrd$AZZQ?*9q;%Pmfl>^vH8rH}y^=c{+`mKc+X^sOd@?MphO<4Rd%Y zB-;j+e;(M6R#hhvZ1szzjjwj~ZP53D0d_-uLaDPhiNT6>j6y6{ELrt`u<;c`uQga|lG30;R{J7awWD2z0bpr- zGD_W_sRC=q-_rVfb?GFgqC@dIn4X9e_8J}8Q4>M6^5L-7;DW9*R+28eE{$ zx!qfo7{d=X+}#pjT>XQ7|CMT+lSF8Xp6+Zl#OZoW?8{ITBE+(&asJs|$&z0G}0tFRP@}gGSaW5?^ zRa)3cT0x-gPyxnQKF>bn`K%dx9jc>nqM%#W-QMLsQaGE#m;5tYg?Q1q6qo%B9X%5$ zlGn2DQW^%){_C>~?0Y)<1NF+pkdTtrKMT@mDP~G3f+6Fg)ChqJ7||)CIweq`Rf)0k z3tH!#yyP`_sym)36UDvtr8Jp=ZkjuK0^cf0#C&78Tc+RzY~0ijYN80Oh*|IAbjn2x8RytQ#r>@i~n7DX*i$z))1@7D7`2`|E)}fs!Tq?Jx zTt$?C{)wbiiSd!waQKoBQSjHH@G1D~IC9Cqv2^6=9p@=taOnFudGeWr+Et~l+kly@ zOy?Z1(ljYUY8^uIzFLvg4xYl1_E}aOW8XgOGKYGfO+Y1W=3+!h186okEI>0Y4`9(b zHiL>Yr?P1cQP4MEz^YA>?dVfIcXL+=MA2^`< zY@$H2W`bn(1nGDIQ_A1%0D)f=RBV=i-~L?Dp3OFZtkicYp!mYsZ{p);znZ@|_9+y- zeqPgw0n-_k(bEe6ju|{tWP~ppIMULxXfOGSt*k802{XzynTzUbM{A`?{W0WYY(VT z2TBJcuQ{mPNln_ZanHcd{m|nDFFBj{y!)F@itTh3D}^!^FHx>aXwM5yIuv##&y!3CMIk@o$LkAfY`0oPk*s_8N~Lq=7?WTD2MS~s+Wr!#bsZUU;H)qw zJmLtBKIPB>J$aCHknx&>Qixf1{A%`Wo$34B8sVg~kHC+7vRcye3pU@TUo#H)u8}ha z4N>4zEeqCn=c*J4^7&)j#VQh%*NzjhVOxL!*t9UqtAF@)KKbP5@|YDz^TFGHKw@10 zZ>TeZz>rymuiNo5)hr}29;F~bMB!`0R8&PPh;5yFsyFkYyWhkLzrwO`Q0lK@q;MWb zOuV-5o5k8!nO*!3MA}(1)#Mr)BrRia0gq7W9PF+MKP!1CNQVg4DLzXZxC()^2kFHn zTz30--6{;xI-)T8R_@&QW~{`3p<2I~`Nm}&IsI|Qi;wPlhF|XaZ*HFb3JS%1=F`Xp z7hi#|AQg{T+L}#f+h$;66>;Vvcsz1?6}QgMQa1^pDC&sa^O{=P=qLoKK;zJQZEk&0 zfDfvk*DL8@30VX|Lt%YaGpUPP2-rxGl)=x3->YZn+zr*%<`fe6sL?XAPyu5&@{~h) z%ST_unxj_sz0(Iv2OD<7?*gqlbh^9HY))D$jy!QS$35;CysPiVrq)?+n-YQou^nP| zF36~|Pklu9ZAL=w=2no6E3~gL;w+)IqadvS-?{TnuG_wyAHL!hJaXyLY~6fiOGrtL zX2a|fPF}XdAz^oHf_ob!QtOdgkCVpkV|lcIR!}#yTtD*#uAlh=M}{+udHq15N|jkP z^=?L@Q~SQ#zW1fX=^a1@;*_$Pfw;VMmcio5#>Kc}`VQ>N(&-bG3DD z^h#nTb&*hcUBGw!oh=fEkK`V69QQWv#7GQmWj_5hn=3!yiL3t?U$;T)D%mXm@7}jj z$!0M&!{8RQr-X?BCW1O!8xzFV@tWR#?X8Tkdtn5vH0Mko!8Q9f;%SW)L0g|)i!qr9 z_y$jfAU#S|vq3AAwu9fGX)Jf|yyN#Sfmm|oDwPiRk;1Das9pi4NllBQQvBgxJ(cIZ z?(stf$_GOS8?QO&+(`g0`@3fo6}wHAQ7PovfBHn`=4zy|X}g5nu8gvRj@{^L3w}cg zYQpOAG5nI(I@fDBD@^A^(}|*FJyZZTkS+nHNez?n-m8Cxu@>Lcc^0M+2F|2Y%_8cV zPi7UFRYYD!!OxsXp%g~Eys~qX+znz;4hLcYWwwTh?F%%US^P+s0OKfes4e1#riP&= zS=$#Yh8TAw`C)r~Z>oPnw3bgVN-d!DGJ^12iX*RKa@nP9UHlAIYP9Vpf|WX(XaAw= zx^30#xn=fi7)hwxlzNBdr5TzKDB~i0@)u=tGg&?JSx5>b)+e=!TNXA^^y_^1^gran zb6(Hxxqw<~+igX84JSiefNA#MxcSihv}N-5APO{N%ZnVjcDmhh0gNS@F@zNt>)>Sq zTKA*)=3_=NLQRYlE}7unUwRX3k6Fb-P6wfh_jiFNrbl_)vrZ=}bh`n4GG^emK8G2oCRr?G(|9)-pfBqNZ95Klld4#^n!3xX*KS>zP0EnV(OAmCuKnz)blnvhq99dn6r(p ztid)FQ2r_C=y3$&|B4;`C+=DNDyD-6T1lv9yV$h&ovwDjz3n5+CA&y$Jr`iwch|}# zWh@W7)4_EPE_mn8W!$@G2@8!PbM=TTwr%C?iDUTuWB!=GJ>h9&sqX|BfXHGa+pQF; z_Rr!TC%`D_-ddD~z|$PRZaF2-ZwVqg7E0Srv?Q z%*U@s_ZA}Qm|V@@7Mz0tSv8LGIS1YUt2`ljU_gqk`}0@5yt9nz+0eU zdVQvfMdQ+~Q`=J{P!&kcHom*@G`_p>bgtR?W@6(^%f{~Qt@GvX(W3<H8Pw8dX@RymzG`AF-f=ad3cH#R0kH}(3bOtu z)&l@rRSMxEn1)cRrW|wh3cmKwujFt4;C$ZtlJl9a`}iGk+>+;Zr<2A0#+vt0!XN$h zQ<+>kI`o_eO$S-9-R}zh#XrA*=fCcJMkh*mzUKJT*YVyjzMj>GFJtrOeZw#8S$kP# zdFczzCYey!08k1WDNLC>NYfqG1AmVo0qz*E5Nm15?R)m{_?7F4Jhu^Lrd<;(HUam{ zE#+6cS99Cm6~v}{vr4T;!}f1rd$LjHvLabe27{G^8+QL4KiYH>muxwg&GVmgad2iA z&|g=rlbSqkuf$jp8j|+$nsi94T`yULMe7!2o7V&=Z#n7s`?w_?;itR*h0C|S5G_d# z&F2a-f?%Y<=Ed)j7-!keY|1^0*CQ<$KX>%Lyil9`^QQ1*SZsvM*Nd$TM}XK`F5k70 zMrPTxFhe!#5DGi_OOIfR2;t)3OyyJCG;lJPXeBS=t$MLGio=&uow1l2k zlc9?P9Y&?G0)P6R7Y_aYA)$ke*Bq4YB!KmgI);y2{?5M7Z@O+%3kC7f_uDMJ! zi#8-9;1@ey$#{4yXRZ8C8m2;$PrckaPl!PwBogNLgqjIi$coG-1!5V5$l0cAX71!I z$Gxb_-DMjd80C&Z*Gjj>rwjyDccw*1GQ;<-yMx(koqvDzvjF(|=YPqz#cqvOl!Ahf zRdk!eGlN&DLL=tWfB!{pyZm-u|ACjXWW~gR4EwJ|zqhzNF!FqlKYY^#{NbA}==<)z z`}$qlZBkC_7AQnMuYA>cY?zmq%iRwuH61IGN}SWYF$~`?l?aOf;SuA z!M8U*ne|g|;IOgh(ZmezLMs+&i7f3twv{3Q7V5=TGC?OL#&G4X%^W_qgpwan%Th=c zYFCs$=nx*YpFzlsWzSrdfBfVR_|RWGn`b}cH158AJKy^B&j-3YDMm2h^TJ4hS|#C4 zPyYZf`tv98oL4{Y;c+J&WW45}w8ZR>L>5l?n{p1?Hd;VW?!Kzd7R6i#g2@-s5Te83N zRr{2@_U0F;jAE-8q;scac@K?1|AjbBe}<?>9i$PO@tAZyWvz{0DmVHypWR)ZMMlXA8Omo4rQ=IZ0;t z#qFE8^43iN?4IesW$1VW$seNWZ2-ZWorgw@$t)(b{PN-(dD9dBc`##ouq69AeoqgG z2Bla#Qn@ceQyhtS^V^@swnD-eFT0T}v-dydijH8-!KXjEL+E!adv#$K^Y>dt65@MFOTtQ7TQs=JOe`BU`Oat- zlxf&O*?A6)RtKjHd|YqX#y*4=#BfU*acx@Us_Na`J$GS_v@1EE>()Zs97hl`ti;?h z^PkK#ZpKq;pmn`jTGI+tO_&QY(rV9DsqTL>T3M!|B2Pc`cnV(NM5zV8Pa*0erLYH3 z>$Q<&eBcY0FjuLwW83UtKB1J$&ZRwXlu~%UN2BKUmxY-spMBqV`QpD{eE&CnAas!N zn%^lQ<|p5}o-h5^Pca<`o2KBor|4ifR)lX{c`NU~@DgsjXD1R{{F>XObUOj4iv*z@ zH4!{+pYR;km6sB92r`~B6#a~LtrK@xG^6mUcma z&dD-g{Yuv-rq71jyrykGv83#(T}IW+^_|!36S$&xdy`RDDZIl4rbe%+{QA94P}|-FF*24Ui=TAA+>h6vK0nM;G?5} z$36Zy!a!p>eAp(lROYH&`Q2X)Zs38EIC*#2@q7C9s8NggKd<{V)%*6flLv;Kt1WN; z`cJsyU!?=YkQ%*hti#M>$VuuE}MiBNRFS_)eHu!@r&RLJD5@oF0kd0UNDi zUoyg6Qo!ae-6mzCNyWno2$~sFqHcz{%=$M_K`~T)k_KWBNN@i22u>y(p*%ddlj@m6uj|{ebX{CAU+T$n( zA)kNzOF3!D3RFW67hrMTmaR-t&}FqLe!0q3fcf+kUl~ z|9Iy&X;c~xk0X}`Y}jAt#D75BhD0 z6LMl{qOY_mQkfGmhOCvQb`q!n7fqsCzUdfOKEBttogoe%MJ$O}D%Mg{D{u*zLW(8p z7^uW5D%NAsdejB1ZL~A{_OnEL#iz;e5Y&LFr-0eC%+|&fbsG-27D|58WZf41S4LA@ zzHJxJ_|eBW>zg0sOZVL10y29em79W$2Vv;(rZdi=tOI8r?JN-5r(FnXV%WHShIDd( zCaQpDwu-0mJcy8p=0%c=^)dN9w;3QyynAydlA=MdOK?GA}YzT ze+C7i0rS(EI5VW?z&DGv)8l;Or7tD03cm_|6}*O3sh}H&4w)6Vv7f*^6Nzuq+LERXNj0ldZ7EN zNm?n2cs;QT0nq=P04WXyycNSDHJSp(Fh0{i)dgc`+L@UH&fnIU;J(IWtLSA{?>x)# zD|g~+gQqN7Spv`S>?4lgwO4$eCtv&#o_EQAbI;<=`%O7sIXc4e{^VdWXTUm5Qp0sS zc9V<^csI>C=xJ2o5zkcEcJJ)tFz1(W!Evc&&6dIDCAHrKSFWnX(`8+M_Pu4-b1R@4Nkf zxo-AmcS&vuX_@?5|L4JZ*3XeQJN>daVaAluq035Y;yHL&ED9|{1Q{zt<>yi=oJVE; z?Z84;y9;87@G1L;=Pn>aYIYM#+JR`Bq3m9?9sy~meTz88Qt5aS*RIe^ybiI%J^GLc zG$iQyEECI!a_NX!M=>@l5Zi#7>DE{xFQe80Z>kiO!xUo`p_*V_=`en}btkp#E=;D8 zoS~?KLJ|SvfhR2pd8HuGJ}+B;3NJkAL_T)$Rc$SzY2@kR$KotB;^JhiC-}NY>kz?f zWa!-dn*y^*g7!4Vc5iOQk?0}MP*VK=!{YMW<~8@+v4e{~c^Q{~=LU!C0|eV=$;OM= zi4x-VJ4Vuy8fsL~z!Oe6ibk4o?S`#{zQ=RUK9N^G@eICs(;cW9G$VTz&MF$z6@PT{ zsRP?PVPX}ZKKGT}FuQ^G-u`)Z$F=-g>9WW;ON$~EpM^}*XrcACI~q%XwwArwIBTNC z92VjJNC=f76QA44{RY1P98knv)JN37dSB`&bGw3bZ^ znv1vJL&LQDE6So*S?sO?m~LT60i`hDqtAUN*IvJwyEgA7PO=t{8N|n%srUw=u{0uF z7!ZTBcOGS2^j}6E@;87yUo-X)7n(g7dCh@{;g0||h{(MFN~tdb&jN1iIq&}f?*-n% zLqwJN8gF>wzcRaLfg}i9rN#xY=$yNKE3K176?jhM^Igq|VI+HC~NQqU<4+qO*k%I8IgK8cgZPvIq( zewx_Y4gn_L6g23_V(ZHN(JTro5M`RL+;9tLuUNw~PCAyaU3055=@-D5b}D3R;8gc? zlFofAe=u#q1|BMPjx`$uATDsUY3lWSkFh1AJo2fh3>e1;LBfL$wo^)70Q_GZYWyce ztgxW2R1-Jd)2p@*nur#_#jx`!2ys zw?#(N#A#8C`Y}fIz|sVDiD+a+CX0_@THQcB+k&sh$m}9ApunsMWGL=QC-G+f3n9Z7 zpOZsPv8Nc6A3lQ8)PDi6YsZNQ{X3u72IcX0P+j~GVrNNF>^32Dt=#kGxVWd%co+-; zhJBD*OXFQ!$ejDzt&_h&Mz-b>x+Q#70QY`GQp4Y zsj)1*C(1E4;=S%ZK0j z4QJ`?nds?~kv4p5o)pm9Y=CIi&{*Pi1Kf$A6G&6bCD-1~zWFLk#>pi=fm>%WDfAJ)q{_;l;xzmHB-$6^ven}~%fe!+&xW98n?l)#UtQp#VIRR@R+ZCjG1)NZD@J3-`adDm+rWSi=X;xrVB0$XizX&v(=jM ziKT#N3lwy{&!nhmx{HL^RQRvEo=u?YjC)m9_+UiSP;Kl63q-2lA1^3swn*ruC>5af zI7_EK#ID89vw8Oapzsl~C?ZBwN?m+x%Pif7B{wM4et;sHT2Q+voN5K%DO{}1Tvnzg z-9koe8{D(-V-gv6ZZ~OsrKCElJjMtpc=*o9Q0bd@dwFs4*b~vnfHn<0PdZ^pM1*bslr_p}g&ruNgXa4<$WF z6Yp=$0`lw7!^LZMZk-`+B+Z++x>s@PHpbBJiiK%;nDd{-eHIfb(ox;gnpM_DbJj|+ zbMFF6#>?EjXCL$R0V!}D|AvAA&V*I?Vgl4Nw5E)LrF8SW?`t4+nGIJ9*&+ee!6O^nKtS-** zqaDYgaQHe+L1-=(c1ecP+Xw>)1IVnR+6WmIN~(N8I^|QA_?Komh%H z8)zet=5Mg~MMojjV+y5^XTI_=Jo@Q_-H;D8Js4i|a9Df~B(QfIu%_eh)%j;l8U;=S zE>%ix0?r1$^-u)djW5YlgGF3$-TWfNF)A}C2e;VuVpX0Iq%t9hg=k;aA9R%Oy2q?t z+PZFd)o0MDLWyEmy1<@BrAvSTw;|TDgwNd79_K-Y+nP5o2QhWqU^WdMOjD==TH?DKtj+e1{=3R(1uu{q8 zKE*y8QI}zLbDJgDntjZlTVT;d>@($@qxX-Y!UY2M#>D`+P-4lb6Xp!lT9{8Fw${d} zWDaPm2{t7Hai&>J`*RgNue)!p5fvcotRt&h_tr5;;!?*4QqN%M#BEHv4MyCqS}g*Z zafb+g5(!Yz2mAb z0j5cHlku&4Z)1C<9q;B3RhWdS@V;CU?+%0?L=MW*{YxZpMQA0WE+O-_L`@0|3dRS} zs!SR_nHedtB{STya1nRzdMefU8pmsJ>$@SE1*?e%)VXerr;a;-126+9VANZ{ zkhNa%Vj6JWzSaC>=dol~kyx+ej_x;=<8L(&Ukk->2>u2jX~+0w`#Jkg3%4XK*|5zC&q%v@T^OANHfI<<<1 z@z`1!AN-ya6bjubXRW@}x{Zx~UE^2tf{*(Ff@*3_B7#?kAnx`Bds;a+W@b=vhRQ4| zv*^Td;;D!7sc*fBbw@29-1x&l_mlRxS47U}I6nSswcWOW{MEzI^-lV&@S2IKGKU|% zqWeQnW8 zB3Hog??3kg9{bjh@{Ox*AruV;RN6jhW~-2IhytoUT!b znLSyV-RV3tw&XGrnl2`B*+dPykIqW7ZC4efXtEY)l)@-iFJ#7+s7ORiKutpS*%FCp zEIxG!nac`nH4`n+tmgHi%1jo|rRc3WyegbQuENhSc}TkAefdG z?X6Y$>&w5#JFmHzebqAawUE>}USq6c+MnmR(LJp>>AdbuVP^PFvn!iGiR3HW2|fyo zh>#liRVAY#K?_jRtz)4;g9F@eoIsJsw=@NZoP*@TjS%SZJT+Q+xS_j#)C-u3PNC?p zWM%2Gj0VTDFWHL9Ys{rqGoQ87$c=@WwA8YCZ>vsXY1^ zlh^#Hy4Jb2_tOkmkb;l$6s3`fF!DL`F~@M;Q%}b8b;nq>-Dltb+V#It0R|ovuX$K3 zz6TT&BO>1ezSr~FJBFP5*aN-!{m~CD{CO8xn7aX7`TZL?;qmKfMNswO#CV9O@e|0j zMZA2`tF8aBWx#kd!D-7!Psi8GOAbXz+F#bTEbMeFCBEfI>Q7#vG%ZW{jv@6j84-wM zzYVdSPYGjn>wSaEOzUnOOu@N|lyRDbk7Bx5B+6iJwuZ^GPqIMJp74*axq$P}IvTAN z_uaOGuU>dXU#HX5aOuT2bNEp`+Wo^tIRktU)Utr!gqIfsf6!`PWFBZ$!c)rpTy{sYpa30SX}^VNW)O!Q;-xN~XLjV_uC+JjVSRqh4dkgA67gL)wGm+g$q| zFd81szQ)a{Aq6Wzv3|*)Fdm)7m+$)&SM9q5I1*1;Ms%HdQ%2=0KU4{^1Q?vVsHjus zlG1*y!>!Ti*;<-HVlR}x24|(;kMNUB!2D5A+9m`QXXOvH=GY^aar-^HojX%077raE zTr7YjJjq8Iau~^|_AG)|*qFh{|Ljg=dBY{Rd$Vv*lwoA!xfS@njkN)Z_l%K5ez5)JB+4IM5HsC_L6;Ol#R)9?n{@jI~~j2O~1xgl4CCHOvkvZv79~W zh~t;}9m>HGtQ>g;is5m3RuaCl^)p;Eb49zni}=i%5iA;mhe1=70vXM^Yl^xJSX!(M zv`A^C81V{S2<)A_24Hb6LQKwk45afRW@`!6P~Bdp0WTpPgZX0sWT_xLQ4-)))+;3SjRCRr;Z%1K>6Lb6r0y7T<%m03^U^=N-pY7vI$Nk_24f zU6WzhTsSOfY^0rCkldBjS9}e;7_vy(DUm}0u38ou6xs+TfWY^Nm3EWZ2ufS^Ab8*J@L@6(t0#zxro*!V+s}N#XKJrW^i)Zl2RTuK(t*^lL44kJ5tQ~m@SMB{8 znRE-EO%rlrJQDGpTWEl)SnyKnrrV_y>VSVa>2H_}rg-7c|D9d6MdD1Mw3Ec@)rd6l za<@@)oE7*MjYnrRgAE{Pv>8%&C-e7hA+DvM71C3NKpK{xeg8I|@Qs@p8HuRuTpUq1i8k677hb+H;R1-(zVEW*8!IZ71FgveC#Yq&rh&XS12kP)bqf@)_twUq~$9o zDs)>g9H3q%1+-FFskK^AM0re%?Xze!tj~AnPobIvTPk3}t8xDLZTJcf8G9+E;7GKJ z*t+;ddR2cRqZFLP)%!j}+7}PlvEKWh!~GrlhCmw<+kTcnc^o~qicg&NdM?{_EAP1J zOU%|SS__^Q(%7T2r;Igz*C-(h6wB8%72_;hHNkI^9t5v>SS-E=wI${??|uR0@lxOC zGBH96_$N2(N)O=OKme-tx4)*U>VBx8TccO|xredMfLOOooAEKd4F@j5lt4$EGVxRb z6?QNR*EwGeOOjLrVZ#+dO#*E2Yz`fNEs?hzUyrk*{0z=m^>xMzk46!;tTNyjixXhl zmL5SdYeq=L=h*3g1YkTmo=2|u0G=wK^QXsQ-mHE(G@R$s3} zoe?O(QyO1+EQ_Z3>tkNa_F98$_HE*0k9ZX?J^YbOMg^?#NE-#R)a!_Q(X>jGi)zOy z(EfVe^2r{p8W#tZGnk+X+9)y&CUStk04f#2c|%f5$&6)gzD`!~Fa@uxIe>LXF6Zdu z58*dTc0b4OX}>f!QR0u^^dy2_U@)zaaQ|Ro;;DgJcdxMpAgscHx?P%vRD~;fvft3+ z4#?Kcl_BuWH8AK);yO-O`pSkwa4()`##3@#+FuFW=Q(82zc7G zG@YhEGMAO8Ns(=Pzrk74&*$9LKgCn+SAEioPhxD9Z*P1Cnmq){uxJJpcFV_M))ZMx zcQYQY#@Lk2^%?e<IJb~Wzpv(zYD0Xk0!z0fSphnY`` z>}ia$EE?x!>(1whk;C}2Uw)3*x?q;f8vgo(3%LA(zhhrq;mtq)7ME?^(s76GJ4#?t zrh5}Qe0-9fTNj9%k$KL;4Olm+i1(=Rltxhc@Uv$dKust8>@)8j(K*hp<4C5~H zwFVCodw>#Xhp=ADBCgzZ7>``OvE%*;IKR~}=)nRq& zpl%>Kr{>(3EI}#Z^Y?v(R~+_!j#zql*EJKRC$hbEJO91mZ#kj3jiQR#l1|eLOlR@f z7%ye-Ow8j}yU+F|yVwvP%S=4O(xA$k(n%b@Q|hr zp(|TOIUz`)@FIoMim_tIE}OQMee1L^>1t9O)`4-VrB4iHuWiYadqtzh01sEa!iWtX z5ER@=zDBe{67T|#^Pjx_fco=L(!o~Gb{POxrtA4&;( z>&yAU6aRt8^SZ1Tr%heJhJ|mjt8quGuP%9DaNk0e(P%<0-U=!B z*n*GtG)Eu1im9oQ;eC3T=wRbD2emt?NiTfe`8?_=ClQ1`<F^r8C2qPb_7(n15 zy%FQa`JkIFCjpEP*#LusRymMa9dx7plC%+CcGfAp?Y#5&(9@p6qDfo~O$*jQku`iI zYr17}f3UVBSwKArs3%^B0_OsGYgs@o3;3(+|Cj&1>&MI{-3)78`M5ldu|1O)P1M3- zahtxPVnP;ei4E}zZmF%|?#2pcvynWg*Nun8vM1ic^|L={wN=X&_`>FQP_rR584 zf&=Zdn#HWpW$6iU1w@+9(l9xVvXeCyLV zWX;s@lOHaU{T#oi10v7U{L6p8f~70QDV9RY<%nX#P>2Ew1vk(7oGl{wiv}HcfaUFn zf*P(uXafgUo1Z>bhp4@D)rIJQ{!jqAdAMX4(b<^f&*Y3NB zU(Macjf-ozqq>rcEfOlHjCKOaHOJLg9#vD}yW2lWV(VQ&$VO7WvF&4Awbzw&dF=8F zICgA(Qx)nP14ZWs=bp)Z)~VPE8)l1Ccs({GzD2D-uPEPRNr$3oDJnC!$uixxeUQh!C0R>>?{-zKLSh;e7cl^V%_kZHU zMF$zLIp{4h;0oux<%2Kf6>oVOTkhV)%0rj1>d9zA^X*@OJSb0Pf&}de>SdrS zg7G04@AVh7NyJ%TyxgxRd3v?@io=Q%{Nbsm^2OV3att8Li|BL+CQHL94qH~_SNj~q zw}?ladH70WY>e;41YXBtw3;~cP#8Y5@e(fFd>QXJ@dZ42)hSf7I^W!O8TVF}G8HZ| z8P;11zF~b5ae_?4hSW@e6-A(P3yrwcX%9oGLm|Ura=0&bxWAix-c!`U}$p&m6K}EWle0$!AQ^4}vZ&^&Z z{ChWX>gk6uy?m^%orjI&;M0zNSLx6rmhqYI{}p%LvW;3T=D7812qK?-jCIe_sQ7;KW+_8zqU`kMl6aMtkXQ7nheV1K6w12NT=X5Td z`FYF5+9XeDG8sGx6vF2EZvOt}&#|t&j1QgqYL*tpxpQGFU)fq^XZ;A)k8fqvk8>A# z^QoYo`P{v*jBWKXcGbseuFqT4N{$@c&B;?cx}Q;bY_8tOv15;7Hksw)8-GA5V+fk6 z73L}>%3eRFOw%2nVOSb2XG7)2fzKiWU)%Z_zO?x!%6b(wenodMM!z*|f8M*yDt66{ zbQPUNK~zf68tTRgi8oa@^4Oz};%6JS5qF?C7yW>jpLhySIrJEA-?f`H6XQH`-C8nZ z*|cxAEi~pSp~0dM%xip4ll5AVMJUW=ZFfztsj9`VES2qZ{PYJm@xnhm=QkS@a}e^H z-w7e6NlRBwbmhZu-!a$qqP<#1HkNDV9WOR5I{`-F(yX(n?Z?(illt8mMhaXSHSpD& zZ{=He+`-C;QLM>lQleWfrAT?ae*M`zX5FDY|0ka!wiykZQt|^P!xBN@u{SX!#(^n` zsFopBatEUk%e!y+GV@85AME}qdy_d54 z0c9`AE7b}WQ=sH`TXzEO;QIH*TiI5>r)Okc^=VHA0X?~(!ox<+<8zxoF_0lxf%!CI zAzNf~^>#p!<|&$q){F-+O`NAp6bDib}%} zgaQ9_))TpY?>4qq7kTXJ!})__PG;wPg<~eBId=I{Htn6|6(9c!8~4s&YZfI=$Zf`W zTOd)EKx>S(EsffX1ut%!c)OGhg#*0-R6PY_x%P^?dGpIY&ZoZnCW8AxXMLzhaquyg z-(^xtaZFFF!M*qF>Q80u*jNOo^w_W+g4=nv0*oib-|b%` zu^GZ7_cg}R2v1mc8zP$VFlJ|cj34bd)Ds!5j-M5*SpKvBiC(2y5brU5&;15nXnKM=#!rgoK^610Y zame@>pZ&#k{KrqOq^1n1XYm@s#GMvpU6z{F28p~rK}IJ%45lwsN%Hdy)!1<1Cx3>? zEPwFIhpEux06G|X&F_>D(@SIH#V#;BZG>Px!4y2o6D1m+P?Zd26=K`Fl}?)$8Mia} ziQ9|B%Y%UxSl{7?+_7gbR)kx&?{b>=gkx4smr@w|9J^#GfBM63F*5^#&zkW`UVP+< z+_NylB|C3JLdO_~S7-$zmXF;1eI^QxoUycNJ}t4B7AfeI@u*H4DVNsInh6;35_cbT zS6ZXS$J2+fH{Q$U>TS+1-l4TMQ@W!8FrgxF*vJ_urD<40DbHl?&1l${lt{&IF)S}H zJD1pyxkOX&okiySWl!b1+pnZ%k`COPmZE|HjdZY!2*4OkntD+8z$y%I!>*m&v3D=) zro$k>`Bkf7w^7J6EChFx8AkBWnfnYU{M=&NpIxh>ay^w1pfsXh&#nUio** zN=rzyMyM>-x$4q8c=Qub9Qqs&IsM*k^srK~)L*j;0D*OfPjmO4S+YFUx;t)$mUTli zD9k{8skD*{oD0DTGS}X_)qSSG^Mq`)+j9K8^(XN8n{VR3uD+gnR>ISYM;yMK^AA6Q zuibtJS?V(K1}WycV}jJO1|PWn>%=zgI>RFTaQBHQgmJINuDXNeA4Wzr_byI#2{DPv!f&FB;l^fUR>AeSL&B78tQ6+g1W`HRL|`s~ho@ z;t59{!I2Y7xO(GO-tg6L6KG9hEo(=|*?!*~X=cd+z=yGo7Tp2Ype;zlxH-_2pOiYF zfLYaOTWaD)%0(Am!HZt@s1~B{Z-fpqUh_LG#H>GQO`mq8Eci9Ue}22$ zkX#%pS=1^62soj|Mr{eRxwofzX1C2XZIkbK-l8I#$nR-DCyG~I{8cJxJD_6o+yd{r z;tF)pB3?#Od4xeP*j+zs_eZK>=Xx=#qWu zD2ctFG%8>wE)%FI7l?$daf#1t{v2Q4`c=k*Dd_1=W<3^dfyk?LD-Z&!i$}1mu#(tR zFcM%fc&a_DbrW#w>egbD zuZ1c?unzS`4J+0QN!>^730t6<0I4ygz<2Jrv+LXjfQ5})_Br%#6kZBq1h3Kmj$O8r zpp7eogSGTR(6lMHz{I9c&`FbwYPG@G04S1&g$_bq^E)lXD6Q~)kE}P-1jsVOpT6kP zPKarKm|I&~Kvh5wwpvSh@?w{<)J49-kPH@@t{|w5rUOf?3t(c9K3*pa`Xk^{|dA6$Ip|97-%^`x`X_Jm)k z&?8AQ9&z$IuDoGm>+@d9LYx3Iq?q3zQU+|<{caJi+PZaM+>4|N0s%D1LP$88wEL%o z(PE=d@#!!m4x#&>PzoV2qZIt(*;?61#WXFkF{w=%4<-n8NYWuNDJ-e@>~4%Q5;Q33 z6j6#sR^nN!UdoMgH}gODev(>N!cxE^bLqvR$*jlj+GtCFY4@~km0nCDZr!(%xgQX{S-swz9<& zisJ0x@l=40+ux;SO;GK~-E1mLNj7ke7m6X}k-_j74;LM5yykZzJ82j#nI2{7veCZJ zeP6TY@TI)sb)OjESM2y*F_7UXoDkFXb?%B%7F1qD5bLZb&9uD!hVl(i6xC|P;$p<& zVo0SDHZ6!P3boIvQ|pNIa6v@`DrqnfG*9PC*+Ki4m`gL}@(R(FBxErOsipz*NkD4X z6R1(FC>qA+-s&>u(-KuvU{5m2ZHtGnHyI_dV?2KO3wZAxALaIi4P;_yLXlTZfi$i+)PpX||3UVQ` ze;7~$uzPH;vXwU@Iu}+_w_)hS)iv>AH`qFrHm;9uP8~rR@Tw_ZZ15VHL(h5JRS~u5S3+rPX|DcJnIOeQiS*l@s#`NYqZw<{N{~Zebc7>UZ0iJz$MZP zId1|X9v2!*AnmE}F)VPmb16vX5ziJ#7DH$#M#DOx-|Z%jbinCTN99>MgLkBM+){3zS`UPQlv$wEmPE%DY7Wy zqPwzUUE5J1;N`D*>~FpRPr4)bj_s^tI2wG^HLFiE|Mf~cX9n@+m%1B$9 zZKhq{Uk)2>0Yd=tSQ*va$aHh#Q>eNk7=|3#TnE-|#98W-r2$D2vUlGob8{sMIwaIS zkDNZ5_n-Q5UVr5C8MJ^zoGErDE|FT@s#@KQe)p82ZUZVNLMxxCaFR+AFcX)knUGo* zvY19Jq#jr8yN~PVM%Y##XG?7=uP2+=CCck7&84F}d-b1C@RqQ#dLKqiUw>LLA?=YG z*zdTbVyv|=UvCEUs@6bXH?x^P{>A57*L~!XFJwF@qtS>_Jtq3$@2?N*MJWcl^;;%; z=cD$kLRi6BUOa^a3KOZO0m})nnZ>VW_>C05k)dOQ&Mcl02t2}Kz~i5OG9UcRtAFE+ zrDJdDKYmXSnvOem6_0!Du@s773&Ao7JeI5+<%YYqyINWM8~+1%4{1S&mpCh-g!FBQ z{xCCb&&MNUHCC_L$Fk)M6bdeLQ7c8!3mNeX{ME6~;a^UEImeBy>A$$yq=I7DjF}mv z`6)u}^Q^U}^||z|Kw>uY2vSt*xINx4V8t5S3)2J|?V)+QE=04g{5kLue4ULx#8OF}d zS2`}tk4ECNb8nHSPd}D_KKZ4T{D>#7d?b&YJi7%s(TNL_4npsu-bnVP0?wa2Xz>-3lA71ch{KfH)<y4^<# zP(t%x)0FE~2E(%t)DTo|g#obQm<3G>o6AV@Q*lAu3X~Q6hVx0Q#I})sX-}VMq`+VN z!?PG28(t*yVIevAw4?vOky47c{N+=4+gqQ;*wQi-G-k|W)BIwK*9;|rpeh5Id{YS# zgmj$l-K}RTXSRRk2!e#|mmGvClS`_7f`W)8NLKL9EDSI zbhx|w*HoZ9))trYldYTByilZ>&^H?w*|fNXt<`ZhEl#nuGS1#cnfW9dSd*x zR+wUMycatVAt!g=Dp*!tMEBNG8q%(n&cW=aw_W#**6+_hay?(Y;AK4L@Do^F?uJIN zqXr2(c+LG#Uuc1lSGaaZVMba)j6lmGEFMU{T?7j$K{d-Q(b8qI7ZHNM;|WhWi6=be z+bI-#Yx9&G@gQTTP2ZXL|iHtPX} z(C4IeD@pwhGegrV(-dfIz5(Xuy_MM*BSC+Fm_nmP(?$>PUr>!q`1Bpu^N-j3u=Uw3 zbGxV{0g3TQt=AR=@_rhtNvzLe60xH?&K)yLxp#gE+bg4F_6V+-x!0MBaEv9fUI%M( zMq(r0e#_B&c_r7irPZVz+L5t}TIv&LKK0DUg;ume@RmIll> z3Z%wu`Bq?AslxK|0xp5US$3PSEh~!StSwJ*>eO0JnplgsAO5zpK2JSsXJ?%>y^;@| z`wTw+q(A6h+Y0PthM7u{{aApTmVT8d1T;TG93p(M0RAGVOm*!iqG{KrQGf=2G2L(6 zNtzfg{o$>{Pk0#VVB|Hw(}BHB`o@>9ChjOz?$@0UKs02fz|~ypaVseJ|Jpn6csa`I z|G%Fp+ip#6dM^ah10jTj-m8$%Ls1X}6|o_TKT$-#ieeYRf=CkqX;K8~B?6%nLlQ_x z@5#;WTc1#ONSyeHR8x*)Uf)WI7##t^1+}JnSJTRgH-SUkT;qQydOxbVs4M@BINE-r)g4M2n zW4w@#NV?g5B|?&(Ckcdh&XlS&vfb{8SUkn6^H*WpJzR1YSh~S3Fyj)hMzHVrK^(RJ zSpN9PD;NoIZ4kY(3ne|Ks)(n}6U|mxE>_axud`-x^()V?B_2f=250Owk#A2si22Q% z`0m2LluZuMHIJz4@P~yj@#lk2C#0*ahQjflr;-t(@enn&DLfBSc24s{DS2z_=A2|K zb%J&)Od;tueG@EKN0h0Ya9F?T036)sAfDXxH0$Fnyw}{Hgl!^_cv8@>riZj>7mgtq$Gs(A&<)V*~g z26|U_eNG@4+<8CUm8NTU3nQkZEyq9*IghsoX3WNkXsRPEP-=z?fP^IBSp4*!XYo*; z8_k12Cs?&0!}$hG?!TTR`P{Ad^S~wN5;AohPf{Dmt|)$cz{#BZ%snL}&pr(S-C%WB z8;!v*XN{P_wF@_sv;~IlB9J(qhGi@AT*5LjbT6;Y3J8QIu{>2#*YPxhB^DyeQ=<9J zBR!&a07LW8b2_F-G%g%@R_2(~hn~QSt}g!Bwt(C&x31LuLRiu_VDj&6+zN8qkfZYU z?bR@d>!zO06>r^_-%k(lJi!m2n~SNbRt<)x@z>*Kv){lWe6XbzEtSnm5-CWtZ8IHV zb^q(Nl#9_w3N&AdXqGZKH7DB#DRbx2*=nXN?-JFNeWU@>(7?)X2q^^g?N`sx;eE=S zU^_yVaAQ~Ti6fr0yU8Bx88jEPPLFpMt>f)g8?hYq9O*XHu7xh2rjl#GfdAT&JoU~} zMqk&MzW@MAm`OxIREHaRY0f2dS}7u?iRmW;95Q4iqgD)NwY8?CoJq?>0`Ir0knNyw zZ-zzsvMjnGqluC!gJ>*FJfZZLB!y)wE=v{yFI{>Y*c`26NPRo5SKjj>fetH3eL&@0 zr@cL+YKAf_G6;a69^{${Kj4@re?`p74+U+B)gcjOt|6$?>N5rmC?n;{Nyjj+c?C;5 zx8#l15%kjPIrgF32xuC%=P|ASVE%g4DXfonqVhES>q}7Ov(k$h4?Uj08V*Peqg8oy zA-}O>&M^OlKxz^vBrsDYtN|f#Y7By1PB!FJ5K;<;Og?kkv=aAW2cQCV&2AA+I^dAK z(DX9Ki3B6&sIY>ho%fqJI( zDeN|yFNfti46SVh;DkX_IBD=yG@)ZT1_>)b%FeR?6Lx@vU2HUXITKtWNVT7t`>YSY-|I6)HTZvFrO z+jDt-?Q-t_U?CG526OG+r(g&jL+I$9iQN(*x+y@?cIir5M3WZn@g(Oz_5jOUTj_`= z3wSa9YZnY#FubvWOQua@&iF}8890DbCr{wMFQ1c7*emd)l|(l`Qo<)FN!`y-0fZ)q zL;^JSuHlEjnX_X{faju+){swesNxeqpI$XI)_?qZ9G-wwK;|_e6$pe2fp%5+t`r0U zVM%WJ@0+aJ+D1b#lrdlcx{@irHu5Blfbfk)P zTd{{HA6t_#n*FAkTL8;IG`|vEJhXZN&um!AKMuZ-1Nx3aL-Ba94C3t}RyC(cxei^H zMbdG2ZNplwnfEMDtXP^OG5N0r2onw&F@moiFq1<@j$ofb1NqL8hw#HQ@)v8}LI}KI z;Wg49RH8XviK#0I1PlT}gP8~K$?>zN_WXLEG%8Hj>=vGw6K753@3+h)9xvU&&y2gc zVO>oG-M4RZYQ-Q_=%^&e)gKh-0=u4W&A!JLF5xSO&E%!!t5h~X(B#S|pQJOH!jUdU zosJXs6_=-6hr)!AZZK=a1bT;*f)qmV-H9i0SikXH`ud%?eh-99Y}z*PJb_{47tnL1 zz>^vQO}SHT&mfxA2nQTemO*!|UyExc(fuzcP*#Qtj@ZC#} z=JIPF?AZY&c(D>yH>BkG5d^z@W4YAt8-DRf*W<+{D@iyGOExt#q_Ke~7A@i4w-&13 zl5i6ZD zvK84TyXEd5{Q13?`NLx6%~jDF!QT!(oBx}7GPf*#gIgB7lMQ93Q^gqfL;of}A{Z5|6H3z^0C%uRQXVqNd3_1878e zGz+$N5_eLl4w$gGeG})*y^ZIN`#J$l=a7D5Sryx=45|VoQU<0`tauHfam=9ceE!J? zXi3ELRNK8DyvvLMLs{M0nW;dER0d38hS6IaHeiJrhiiQ$4SuMm1O^e_AHid2cw#Wv|})1 z?~x22)Qi;{+O|D7O=Q!^tQ_oGk@t7raPW<7QPKcP&kG6M@Z3C-PybgP<5$nlJmWEg zmN!&bQwAkoe;aB%_G;+M)l-jQQtyFz*VTl^kM=!@lOMaCgotM9(dYpOp_LRBAz)h5 zP~wiw25TiD)5?B`JfltRNT1{Ae*5#-hDA8i#__9$-?!!s!bX5g_dJr3b$yCFN!e?P zl7@@`#?<#^ZLFO?zV{NA=jItfU-<7m%scs;-1q+CtO4WSuv}duk*w-yLqqsYQ;Okv zlJjo)CqF#)D86ya!!&ngsSLsd?3{2*IYUND%tWdE5Nt`JJN}BbK;?xaft_o>_^32p z)1z>b5aKu?#1bLIG9kowOT2%%5Mr?qVu29iSs}#mPhQmSanx~>=+&!q=8F(Q^{$Yr z%(I!41Y;gGDF}y*Z2yYFMMU~ffAy`W$g)&6bKm@48aePLTepJ|~2c**wr3 z!Ld!_xMJ@^G5oNH<9U9(S>cL_$MDw!KFdK(V~MqeNhMWEwdd)iQbt}Hw&y{#E67O$ z4yL`Uu7G%#-8@fZR$RI;h-kfV3RXTIlDK2>;zkt?s-Ygo&4$|2*G#5#K*w>lJXtgZ zUmbfmmIE!xc#%p-I5u~_Kc5%Zt}M9O#c(}IP}k8*Xj9D>J7v3k=k~|*3>YADFEvYI z#>uNAljjAM(r-;6@B*MkU5uEE5p~hyWk_}sYHG^#)Y*X$WS_xO+pdCLLC*kO`Nd

Et7ZsFrp#!VlyZCp>{Ntv;gzz;d){+ANyYXX?< zCcm9Iiy8fg`yxv0oNmlbp$Jc(d=WpIeiVItkFZ|UDkbyr07y9w=Zx8l#*oZU=_Y?m zOQ#Esmz$T9bnU$JC0(0;teoc?W0G1UKhv``10Z{2>dX+oopB2Hu6U>D+yI{Rn74T) z?WvqNotJOGr0<>Ox|rS7c#^Pq%@)r6{oQ#6Oj=Hzd+^f|e{# zIO)^Bp}br>2$iU7wrRu&A#~tI;CNs>aJCR)e8I4HfN4_7$-w=v3lyqsb491P<5BL_eZ+v7nAv0M`0x%O`c7dKyo|$xGFPRQ) zOFU{*Y&?+{0-N0FNrSOa%O9X&R2Bsq=a|=Z-jl&yP8f^<5ohzq93StuzF3 zw07CB6hu=gCiU;19VbDvQGHD$7;z|9&=djE^pA1!4H={=`)wrsNhl?nCDGHh4?><*=*h8GyaV4PoyIg&a2pL)YjP3T5gJ0XU9F&=njq zX%t$SwgVDyLX18^waA(^ml*d z-wWRdAYukMV#sK^?3D6SG@FRUHHdd>5UWDqxARErlXp8daOcYB3K~h4$AD08TCGwp z2uL^<3+O|)FKI#z6I*~bww`DdqE0%CuSfDo}ccZ)Z%~e-uMU^u;js%(6gef>6k0zZs)33pfj| zCy7T>oV4FKLIFe7l<`2gQWeK@W$93p%=v-KUvd;7P1q~mug)VqxO4vhIPITzGk?=2 zBw)EN4=;J2iynRmP+p7IH?LF1LP`f$Q&y9w;l&KBn2GD^NWU^}o0Tlz3L#el52p1R z$&mxc648t6+~`7Ma__;ci+12jPZiZm1&Bt&#N#1c*Fbt2c3TK1rg%D@s}a`19NBLy z$Nu9szVh!!iQ6TVoP@3Iapo>tomIZ&o<<4MhP4|z`0X*rqa{G^^1G4y-mr8kV|S2- z8l$_a_*k$@5{#-Y=9z*`eud0D!_QC8&v>$>YW%mynEuQ(-=GI|E! zvhxll7|4&t2LlGj9WoA5#d9xZIlQ-MJ%>&nga6 z!ZZ2es1w<%zI0Eqf#F61y1^S;*AOq^O+kVeGjTVERRdU5tCrcrC-dZzbl??|L-$|LS=H)cQh(Y*Yt!FGZ|4I9=&5ql0H1i$*jd0hXi|M2qMC3v1>{KO%A^U4#r;QAJim{Qq=zdeAkWuZC@+8m3E>nO zI+p7)f73=z{l{IjCzH6sE9%^wK8sB0ejz4*rQ~UJ_f$zK@nQzCREQ@xZ{X>L8);}X z2?VpTFg?iJTO5k{9W}jbx#Y-v+;fN6<`S#=^m~-EKbR?2WsYneXxdnCr z7i7Eg_L!Q6Q|qY|X`Ad*nKM>6(C#F;|Gx_ec8I*omnqnpBcnR6=Hz@AzSl*OkMnQU zfPM`mQ&AF$tj|{l3<8GEl<|WJhfEGSY%GT!IezDnd2=C^;k}FF6Vo%W{dLW_(F6F! zcV_eJTb^a@#&$vhgHw*3!o_DCz*CPcD0xgB4eBA&7d7&-w{=f|Ht zmVdpwn3q3TO~TG{#I|Xf_W+qtlcQk?4+@I(C2fbRAAOqZpL&_rP#jnHFrs>1_hx}m zbGg}sf!TTCc|kP2ylcTq=`^?2<2nWxtbdHUfJva<#1&aXC8U{rYus^cjCRo{SW8ed zNR?_o;0oo5ZVgg9z~z#02Qb3u%au<(rFeCYKnrWgfTvo-q{AGW0whu{q+9|u8djsH z7A%%f9@-odZE%Y~c!Jwrc$2VZP~++u{bIx^;(Sq``S<0SybG&In( zXFU`49K><^k7HmTRf77?Z_MG==iX%1mNvF@MA_1r)BZSJJfu6Me~b$V!MMsR;FzV& z4Q0q|QXZeWZf1={rB?wYHArb_4FWf4t8kcq^0ZsljhSF6B9KH8kXzeec%NSAwKlrY zIHvyu=6$e<8MpogAp`*V7c<5U=g2905HJl6 znKX)j&wrl>-du>|dHiq1#^Q`g-}6YJhuB}UNXk%dy73kRw8Rp$btLc%iPuMFbyn^v zq=1-8Opr+pngg_ZwKR@PA#`OpWW<&Qt9%5rgQn2Be-DvR8G9}@7crv1YS!JLLMUIs0`4p26`g9 z(9mN#;h0W+ZGhP36i%JQsgVc=%nl9R51F|VqOOx&{t{w;5+7s3EXx6YdEb@oSY0M&y#H2q<9;0QwOV8 zZROoJ*YMy2Z!vUWFD^Xm0H#eI&d<&}9)KU*^(2qHz8Gxqa%w;c4R9zfAMWM+tL5+4 z{31%ilL);8BR*U55Y=aTTl3kGx1vWDJ->tl1jp*`qF` zcPPSfcil#FJf0!OTCR(*G_*zyFD+?GQ=u&%Y2w`GSd^rlZ!8K=AX0i(R`tpA#v~FB zL0!jkT~#Z?Q(TwaNmJNXK=_JVik!ZXISN>$lmSyzM^s0-h|^3p|DTX)Fk_#Q>^)&9 z|Geibyg2uLVzCtaP94D>d-N}H<0}L`XlmLWgb>F8zemyBZW zN-0-=tc&Jjm(()=0W*Knf9;YZ`QgtWCRMOtiYIL1i4J9ltZZfRfQ`H|eT=tF=BO4P zrB+oE2_X-e(qL5tBqR}Rp5LEH|0t1uG0c#caXxR_+)%Qwkp$LnfhT5N&P(f7@s(#D zC6z-m7OsXNgXp!k%7%0VZfiCz)N(zx#8khSTtLN&= zj_0fsrt;>(HPqDvIef;R)YXJa2vH&Un7ZbZMo)~GH-Irhh><9zG6y&(cYqLLAK)** zaZ<`H#i!W$h=c>&^t;b-$K9{+^b3pW>`G=WV;=%=EEh{iVo95}jyPBT>|b2{jT2~S zjBxK0Z?j}Wvua`QDWX;y^|e9z531*b<`z874q@-r(LxYltHA+VvkPR z9qJo9dY6V*D~ncOrlnJ(2Z_3Fjv*w#GLV*z3_B3?D7CsOxEv?maR-*|;uY7MLVBRL zn250S(=hV~OdBv5fGPb45HxfWZuzX9QjHps$NibHrOjbB(GGm|Ny!Q4gk|{g0Z~v(ync}a2>gqy#=ES{Ae}+ne{CKOQ98&Gk zs4&hOLWoDCl#6qQy@Mjv#)S}HMv-dIDn8ps#%`f!0A@}f%`bj*I=9^UGK=0@hg0Bw z7x-CqQIA%sLNO1$w2-;~Tfv>*KbNN#EagA*mk=>cy4;isHT%;-(*%c39LXz7*WlVN zmh?T0>3QNP&u-qN$({G5CX*yDoNaa)NWZ~qVb$P(8-$i1NtsI&m}n)c>?fM_-Z5&|bA@frkLP$Og- zM3bs#NqQwA5a_mmn1>h18Zue}NI=r^FoFV07Ai~uDFc!cJt>e<;?`3<+x)|s25379%EVIxa4>Jvh6)=AU&%n_4#{JDiVuE(*5 zjOUR3$5dpvs5>ZU(#KjIWk|IELWne}HeBf}ZqnYW&g)u17~K$(-x%q}560Rr5`&4O7=ScJDp;zq5{^Bc9~Gc?)=P z@e1Bu>o*-OFh&i@Gm$J23(y(laE|qRc`XQr`VKAv*e-0Wp9DpkhJcw7;uE_~0Dh>> z3kx*#xWtO&xH!_!WV$J$SH@xEdxmVwV|-IzhV^S?S$hk6_Z`TW_nFR!hF9)Mx!%$z!^Y=dmqM+os|&fhor zoRKs>QBv(k#qOkM0N#0b4Ypl6nLt9(Vj*Qaq}pAqP;ppUXo90B@4>Za9|u71+6Z4f zdIn!SdIlY_1iyN8E>AA_fTU%kCBX|*%xNju_44IkB;`09_q*G0%EeuLcs>Evl@hJo zkC=qudeg7TC6IB68}jhXV()QbmQ%U&g-w&w#!X~2;L$&R ziRQL0Y9k>cA(QjJa5HPywdRdAO`XFJ9YNSfJq~07<_&I19<7ei>b8(7@dM( z3z$)XCIwn5=Nb|W8yce~TUU3|->Biq%Pv4KwS@Cmo267>B4zW~ON*$F1nJjUix7g( z96FJ4BL@%;s{3jh8o_|UHD5WQGBS%k3CL}1uvJPqEoZx_+^M#O98&EsKd#ujJ9;(N z5HQsZu6_5JGU(m7gIJ4iNBD@55n}0sR{| z@0eY@dshHZ&{< zZGu3TAkZaPv~VMHD{!#by5FNdN1VR<=?~o_Xm#etUa1@?jb}fB60xTyg#pOc>suQGI&zrIThb zZT!%p<9A(r>;hCDmQ;IkrI2bX8Alv47G2lURS5EQy%QGbj;dEA9Q7S(n&`GxOjk6? z&u@5wh3~It#p-6Y81vEG8Ov0f75EDz-*70L{Q{UeY!DHD!89mb%Otyz#+8zW6_UtC zFC!BuFO0Au*rIfoZmW5lscvsUudgN4Y@(r|B}Hk{tyFrZ zez`_uqrmvU<&kIJ#VuFjAqN*v+m91Q?TPOAOEz0I)C~H1CMX5{V?iQOoi*#*xaawI zIQu7ea`U~fpr^<;Wc+(ume%7*zWD7sS-N6#X4hVO4&=^TFX7muCo+6^AI6U#%oX1} zp4tki>ix;#qwAXORnT&`=$-Uo@$GLOM?99|jW<@Pw(C-=HUK$eG)tmsE{%0HXuSih zZ|+o?Iz`5eWYd;TF2D9cYQh0r*W==|4&>~URIBuTM-SrZSC=CE3XZOyJ)rv`5!X;k zLK<+vA=Btvr&=fc=Ha>A|H`|WEFa-Qr~|?SLyR2Q#G;LxNX7l8gnqJ-@C3mQGChKX zt5T@TPDBYSx3p)f4GXEder9>3!A+vJe!J$N9x zBXBf{8O@S)B`P%39Ph$)U4)}j&TXG)t7%|(Lfs!d$EecDD$wlG65)cKbb0u>1qlGTKM;)|Km@8_#Ab06`TdoQ^=3);@xdXwUva%h6vaF z=ybk&?Z0T3R@Q>iAvbl`NJ^YA96V_xbJwpe%IpIa!L&>A*fa0)JoM>Lt~Cl-i< zxd8@F@59DKj7~d+zSYY!PtqavOrio{Z==C?SHjAyL(G`Kt&uW7;1HBy#?3@C5Rb=RKBzEek_p{gjB%1ShnlCF(W>f?f;{Tnm= zjdHy=T29B~Si&M{IsEnBmwDyw6-X&bY)N6nB)X$KCT1K0QGZgoN*M&A5Nd~Ha}v*! z#8Wo+MOhf^kjZmO9ETF;=I4H8Gw#hlGoo~!-|b9Md!2~SEUsu$;&oM>o3>y5Kgv* zLuT_-qErffEBZs%*?zX!*0 z>5L``c7c&p^GLTK+$J%i>a`hznhj9DQZitL&YIPVyYbAt_eoh!o>vE;MmPA@%maAf z!gDxf(s;BIc9kM|=FDe{ZbJ&WZwV;~ZPW;_(+IEC2yM~`wrCign&RK&AZa<=^Wyx> z-^({_<@F_N=!_>RL^NR$>JS7v^Sw~Qzm-;*f~QW6z(_!(RT6GhZAdgn-A^=M_O!0C zvz2$>U9_J6cg+JVS-Od2(k2>9VcV`h^(3j3!(I2i%EnFYIol;U zN0#E=4esVjJ>uSdDm^i2T>9lBx%9HzNyW?e{%~E7jceQZA{^2s8ivb$PS5kX)Cm&i*s+J%$bF#UWGoDB0`~;H_or zIrytLq6?bl9r@>T{Ju^uxO&&8i&m%eR-v?PHpL1BtlrY5yhWw~NG>{d ze=a_Le_A?YJo4IN7O&on?oz&wT1i0!NQq}=J>&q9=Jb68G|=NEz60CCvbNpgW!~#6 zIQ8VIA3~_g#Wr-!Rw?DQkA00E6tqm^Q*FeI9MP9Q-f%vbU3xpoWd20X{hACzqpdB* z)Ttvl^~fo_{l7IN;ud;B%@hHu`fkF*4QWV2(9#~`?f2J}Ra%*_NDh*g!`)B4%^64R zT{49csj`spRC;3bhEB9Pa9wpS&4NG&8JRQysrodBHOhN>dx zH?-4iff-drpW;c2xBj;_6N>YMi=(?NS=-DNfBrXa+-Akv7LvA&Vd&@v`NaTz7r9;2 zlS~DrK=h=kOM9#$AD5kT z2*H5CZ4bR#^i}0P{jD8wB0-Y_$Btn7ph3Lxzt!EIS4v5HXPh?{ugMgqEIN7N$_Ygj ze)^^N*mFcb!XcAHf&TM6Ny`6mZtF;(qd`zrvMuRtkt)GJ7Di{&)`*lCT@q=48!6u; z%{3vAQj5ZLfj(UmbjFf&#FCj)rhG5;txvsyVQAd^@H||{BdL3;a`Lw9ib4p1r-AOT znsUDrT^CFoIe_KwZzg4BT~eW-ftT3!g)~hN3|4gw+o_=Cr!t(Bi;*Mx^3!Y2AQ&{V zWv&ax*p{2A);fL42u2O-gOP4^LH1t(=qV3F7tGvyBn`EpvL2O^)`BPLY>D&eQwunC z>IA}$!0htS6WOqUmZNJtOA<&5g1SybGcZfIP?5!98qE)DAyv32)CQ4FWn!J4#E2&hCz&#S2=(2OM@u{c zcEaJcdCMT~P-~b3q7G(ZdWJw;bQ!bswU81cD(SVvp=q&0c(YUT zhcMXU(q}=E#-&Mmy`Q9JorMu~39YvXuUD@Nk4to=xZ@X36Q18m?V>2X7bVa;U3$>4 zx=2Q?FuT)-K(M7F#-Ru9Nx;+zhfG32lh2(olY#vkdT?sV!3T`_kOEaEKBlh8aPNc= z0TlNxFD#wIy*s(5+`HY#JpAa`d+*`=<)#a{=bl%2=9xvfu7~F-5^pGEa^@M+sjm+M z%Budu)n{}1U%>J_bbKpQ zNQ5Cs)aJWj0cu>6Z=G-`w&wB6JD$%Ep-?_MK~F(=bJkm!p-L& z0a%vHga3Mq2mk&R+M;o4LIHX=))0@U_{TqAC6#pYh`;I7aUo#(exo>L*6!in1)vAn z8&&LF2qC!U(i3>`%@4T!q1X6eb#umhqG^Iqz~oD39+VkBW>kN^dePzh_NHfX%M@%p z;fRS08qmnuM^EKvcRZUvz$d#Gl2VPBi*8F6zO|Y~udc+4S_FN{Td)A{Zc(cBuv z8>{i~9bcifEk?gSbp(SZgZeda?XMooYh{xQ!_fHQcTc9iF1&3)b`^X?U9-IkT2_r1 zA47*W@$GM)$i){Q&OQHpm3i})(a;d)^waj^;Dh(d8ygIo^zU2Gnsu$dm%z;wpp!w3 zUw&scA1vNLuimxH96Ows7Ol)XKGWMI*DDcI-VD!^xS5WPS#OGPB%y>L6;eOuy6*vT z<0-bT?O@7;A>DZcxnC$4xBtP|3Wz&IEg>s1qe&3 zMFz@asz>djIUDS-acXotQ^Sb4XtoDZ(z~}w^RlfBtgXe+wF1srxlb%Ip={R_xM2+q zff)n4MrJ}rwQHTCa`xGUlfOWcszY?Dm&A#vciljV5V(X;i-SERgllbkV$_9$&}x@l zl&0ah9?!qNlsU&vMc1;S=pzmtM{9eG8~*%U(J`i>bH`sVV&H(rZOX9A;A85V?ZLeh zg{<@)g=y2K$v5A8bK3(e6puf(fZzS`8Qiu6!pZ)?rX304ND$AH)b*=nqp(P&f3(xL zTS!Q{ICUx$AziswLTUF2sL8Ae^kn{bmL8gD(s2X%2d2X(#pm+FKUV%~%{9TQ6-b|B zVNB|P`~Os9 zEw9+1l9FJ;#c(_VK@-PzIsM$3oPWt-2qB0@lU((^`&jzHMoh!NwjGW-dLrNa=JA=1 z!<7gj#9JSR9M=#TBxjDfsOyQWg_1vOeG zMUXjjNNQQafBuOib!%P1%^r5Wp3wvp zl_>(zthXjDt5dHt`NXlj@xeNhR*BRqAG#(u^`J=*b$NX5q9RXr=JJljOnQWaCQ?FO zZHS+Ie>P*s3yZNhjg1wvr0Yp2(+2HFf23QGBz0A{DH+?Sm$R-LX~pWCRnq9f4l z5{9sU(O{cPqPLFJLDw`iLu0?ahVaIsH8{C0z;wl8A&LwG-BziY5>?ZtqzO~TFfyYq z!45Z*&gg^$PK`#O-BX`cfR~KXR1;)VfMUX-|2i1Zx$yLvjN4-%7o0hhTmShAQWJRb zeD94Bv%&Ik0LM}MuLtjbl_A5LIO_Px)YgXh{q>(^&6*ZAZE9!KsDAY6Q`hZ1-Zhcs zOjWUaK?uRGe|S3Qe)~48q?Jc_lYpIav0U{dzNMv$-t`e;NsE-_;5r_fBMG(Tr2F*H zcCVKju|h@TKVo{yqi#i#&XFMmszd7YjFR#!FwrQz7GUi&xZHTS$m(l|;7U z>l0Ovswd(}B=yS^1UlWE49jdjlLXAD$I|=Xr*S}tO^xM0cQ|Nr{;~T}7YeXw*+w=t zcaTWg`P0dh7;$wo5(%ZhlS!M)zjrV9-|=N4;Q-U7j^LqtzRJS)6i;>HxWSm_Zc-6E z50$5Dsz%I5gb;#}gLVg*8!ntp~eK$LVDvnnE~IV(J?E?=gg$ zR*x@y?ypEGIq<+e`TVT?S-GKw2cKC$%CfO+7sHcyB^t_kp2yr*mU7&Y69Hw+?K62; z_aAguM+NGdsuA-s(Ac|%K)@iGC|_p=KsPiZL6hbct?0>gs!(=fYMw+-xOg=t0#*DC z4d0Lv8PA7IIZ(FWk2;52qodo3LlTWAxcl$((1BWiS(Oe~=_!dD(tLwQ zV25?23DVS%fRXf+!J+$PyO7DGF$0OM=wNlbU+B(L#tWWAGc=_vOLVL@;tqOy65&c3 zd)F}M^9Lhz!7VpDN6Jt6T(!KJKm7Dje*BxWIsK<+v1&~Vuf4ef$8xy)kI!S3a20p7 z#ftXr?x?iRvQ>Pt=v5z~y(_WJanX2+I^DpCxkcPGwUAI<mTYB%K@MpE;+b&g80F4Q#uQWsbaCPv)xGahq=R@^sT z*O)VB2KV3o5-G2sp)2J%yqXK4HV4`qSdn1TKEt@=(xds=FCXXK4Q&}x?h7xx$HIkc z`THGT;(}9V@bWt=Si7-}W$$m|`4`_WImdocEdi z2o>o#DNG(WIMW*~L#mazTfv;| z=C{i=%j>~(RbMmas`*}`nzWX5(4#3_6P$>SupIO)9q8+}qIJa3JdbODOYdOI=8n9qBF~d-+0w=T``*21n#Q5~jp4kLr}LSkCld}8r^OqF z&XkF}QxCQtpaOMG)rk2Rh|~o6#r5aWuYWzEkVz3!?JZ8+H=7V->R(DaeDo)LYB%19Z91qR*FrpS_TLP^; zi57Fv9S@_+!B`h%_1o(J7_;X9WdOzOqFOZ~?U_kK`oT zM1&l2Z=NCJxQMnW+U9nQEuDx|c>*mcofBH$NqAW+TDnTQCkd};r)Fs@;k8}Ve9%T{ zZ3o8Mt!SNb^n{JRxdX9m6CIm78NEk;>{7j@$gx*)^EF zPn|NHaeECUlv8^u6f&8B9Zj%SI-mwp3fVv zE~j7r2Bu9L!LPr2KURr~z+T9}3!50x6vFi&X`@M*rK4yXQm8y@FJvM#4cE}|LK-q; zkWRhM7y_P$){#W>jiU5*WI)Gl4Ci0l^!vCtl3+^=E#azcgY?NXt*EhM=1IxfZw7Sk z{LkfVZtmjiU;jH^OQLjQl?G02Fr!fPsD)O_axdrE@;oA%#?NoLh`lBa$?Se_;X3ZS z^A$F%Z)4Kl!#Lx-1L)U-0#ucZ51pM98XylC^4YTQKe7B*)$F9*MAyU;7Hc-P(ywel=<6n zT>ky!0-ktgA-XO&@z^OGd&C6vGKK$k1$?CJqzcqERZh$&03ie;M)l47efpdkoIYm; z0I$99KDu742n3*|lve{!LQJvtGNE9Jn)rsUGOJS#G8|Ap{1!;ud@f76UXA0a-WDlK zt&{*v9o@EZd)I7ga;yS5d`Y`*3l8SktLiinkeBFmJ%!%5z*=D$ma!&n>d84e|~1^ zYkpcFrQds^irpPG;Q-@C4FFV6rte>POy<2yufC^r;)n!vQPlIsaV3NUxPyBk*0-T0 zZKUlWk`}^okVy+dfs%%d))mj=(4@oE=K5Bge)V__>6X^Y>#B9cF}8Lg0s*8Dgx5zi zq47*kQMx`=`hU4Adm(t|r6v6P_W!ca_<^ihwUu|@SzRQXkP>M*2q_sbuz{_eNgSv6 zrf8l-^Las~5R2urs0~AB_KEv2dpD`_{gF|Dx~6KxY-fzxYd~IOwp?VUpJnI&okn^U zAq04evm&G-?P9s03qQj^6>f-a#T%-XcBKwFGSQL5RF4RWz#SbhF+u) zWYU62NbUDLM0+%IUg^3B%O=&<$zy+e9^KGYjhH^Y6fwDgyeAwFvo#L3jn@_Dn?F5+ z8?JjCE61f93>qACnpzpv&j`++JHSr3OE*9l= zQy(E6elax6kUEFvb5%6QLr*#AmP=Fb8pe$14YpI{K22vO;Cg6lw<5N*LM(ysJQc=D zr65_{)?US({0BCk%sX#}&msh_<6$Lj?39CO?a0cV-aJ2!YN`4Qsf5M4MeF&|Z_i=y zu%--m=!8?J@y)Ae@tvQX#_&PC2ucu18{tZVA(IQfel)f9yHWY>PYM-CwyR>tFERmSzeuK>eSQ(`NBnqc7MY@t*AgD{lMlg`#o2J3)rJK0qYe#d@ zSB}KCT|zmTMpO40!Jqzg0rx-f25Z;0Fmd7#PCIoPy?fX8SdLE@Do@u`jhGz(P1Bft z-T}-$?|`EBuKMLL)~{^lE1&rdse*blnxSDv0*o3vh)aJsi<;UHzr5-505oszz;h-2 z1~wps;N0=w#!1Ip^r zCrlW^Nhj~a%{RZmgAcx;nxm%SNh0wS1Pwm_-IEzNbp+$4j^L!zreUQVwybZZv8j&Q zhA<1}eL&lm&hiFIPmuP|7A~Zu%!Jl^0^QKK^!u}U_Q5w;wrE|p=Ufh4*Q5WC-fFyG zYo-t*M)cu|Zyo=kKgOpL6{u@^WW)#|jst#=>O6U~l=9~V?+1Z9fPH}$;4CTSs!vwb z-4}yL_T?wHUBdO>yN?a4w<0vbz6bBcK_^UM)#=5aJOj<>H)S z7XfWj$~}Y-bAaoBv$lu8yDKK{Hi}>BQ@8Q*_-UXmn-#Wf_-5f@YEjG-m_^6O&iy-bwsOp;S4_()|@S;OG z|AK=_BrIxbf+ZI;<*?`8!}7yXg+@nVy^vDIieaEuNfJ%h@~Z=tbepRJ0Xpz?H0Pkt$<1Iwh8D*%KL_W-ki zT*WdAxW<2dKX8K(Ldcy}gk~2;OVZuVpJwkEcp~nsusmtr;}=>Mzgcu-Vh{ zlwzbd#P|Mm0m*oZWZa^msrExD%_@2Tsj|g(15|a*4uY<0)Yg7@T{C2Ke`X#tiC3R~ zHyi%m7LO&@kkR$bq6Y~>M)qaPhIW!^4~l7&3@Pb`hNf${uE*$!!}!6?7twcc@4PV+ zr;p}uFMN-V))>K%Nw~0HO%)#>6{u@^V#Evv*5>@Z4%olk7%8PAgy;a8fUUWMgb;-}%;F`*$4iMEB&$fA2T9!Nge)d(S_XC_dZVFqNwf<0R=(g4~<+ zA#_D|`fl@#x6@~4C!IqkdiA7GoXe(3k+ZF|Re@ROMoyGn8`j7b+^ARL-MGEq+8%nm zw?O*L32Wt7TaNzTRglN<1k?EbLVb*+YZ`xA8~rFK;CzzmR>yq9OTnWmQkpd~a|NHa zlt7*hUu<1W4jjDAEB>ZZn>47$*#-o(9OBnjOF>d5mv4-9FF1!$#SNjl=O){C zxg9SRha`0T$j|OHsE(<7q2H!E+4o9o`Sz!a`}a8{d|myf;+D@NCUxPwFVOsRQ-pEI ziVADN>2&Vy*4cyacTEfRlIi_B2Oqkse{M?EwlCfb zZfeG-w(*hvlC44;gy=BakYiLzaCyB^nNHr8xNknp!~fK$)pMsFCXjT4(oZUx>YT9( zdix$h+v&{@B;i#q)IoOHZh!cw74TEA0ed;=m;Nl|iRk>hhf$k1AkITOP`q^G<*oC{ zfQ4NS=cNm(E0}f$N2UtZ-e!U}J`;`J3W&LBqMX7YZwpUSt=b+&f`Es&ezqiZYCf@;c=FHg0lskx^Q%Q#K`?u&!}!58fL5>oBX zK_A;z?YWBb7Tixfb_OjCu)wWW_V#~M%hufEaubpd5{J`wRJ*5^Ya#>yX&i4hH(Na9 zfqBdr`GgJw8^?2NW-uE|90|NWqY~sAqGVN{mMmukU&WO;S_X61kdM3W-ZDPezy$hK ztbySP56XY9nWr|D#EQo5G_P~Y4k)g9GMO#@fNYhF6`=D*AL9oUo1g>NkiZYLhWF=y zvHscDv_;9xUv8ap1se%|aANQY05e5$l|Soaj=Q}zRwv8Udmuy%vjWvQgouI=86rf@ zSjGzsfm@nXwE=MxMrp4adgtqi5OG#A{;`vOFD6RtK(}$Er;tn+^7Nf$jb&4kI6Su#Cm5XyFaaiS_2o=85;EMK_>OcI}eOklDF<7^R4PI|JEA z-II8Ytr^BLI6}omm=&+C@n3boXR-o*cI&bQ;Rej-Lj(zSCN@x#Mm{4|99&i16ai(^ z>LZc;bl6KLC0ZGQiad)H6HIW%Bp9|1+v{mRBBi%;$11-QIs^rL70|jVlo_-KbCjE# zMT6XAFJ{5Uzi=I$m>#5_>_uMs(^@-ZdC%6&h9&@shi7nD3o(-fVeLFxbQSON`m@Pm z_3SMJ&iey}?GJ)`ywEm!!_n*V(PG-Swu$ZfHcN5Y$Q`Sx-z*ce^=Y=~veWBIO+^Tz-CXh9DViTlmY6nH!~0LKx(^&-u~2d#$}!cde?f zUDv*+IWqwyGMpzv$?noT70cuy8O;bA3lK z18m3X*9DF{yp67#9w>v&Y<*2LgdALSArzK0sm!bl9ukv3Prxv|y4&pI$5C!q zyZyWSoZQ@WyTwg~!OEVGgL)p1TlK6<{%2%$^9YJ8UDJPqQ{N!pI{q)?DDw3G89F$B zu`vHXb4?>&QRJvC|GPFf(+h7n=Uskibmp`68Y~av`bb@pM+!wvfJQ4`9#Hb8AvF_OPvtsbXdS% zP!LBz7)bUQNN)QW?J6(nt2fovJ?h_okIeo^hiKh#5I=@Me%pmCTl&!o?|t2XPf!NXp2i+BJKLW6hd}a`_;KF-SwPh0!Z%*P zw>@&FzL-MfgpK%d@#b3sdJ^ZT-&)iJ=kjx}n3%tT7(iflIr5tmawqG}hlA?N8dB08T9S@&P5>E6UB@jpiE~aQpQ6|e2*A~Q zzxqDXE*C8cSNxd!_ZQ|UGZ*ssare>f9E3m0>zvaSz-%`pYV+(YYCs{TFnZ#zxGC-{3i4Y5$%7H{22zra`0<&|wJrg>z4Rx*9r=wM*-Oz^ z!2vDl>-UjI@n`~4lE=v7E?DfWPBDbKo;~6qd7A5NAR~nLz}cAJ9FVpMeLJKLXgy>f;VWMu5JTm)5AU7c@7q0up6+M8 zAjy@!0BT3LSKUJ39w1MP?KNO~@IFu}sO)8vfuln}#)#L)o71-eP($p8Mj&qxA!zpf z$I@Y+0U47KWDw|pAaFn03qLLwh_8DO;Ll$_2!B{l`bN&SH=h0Pksk?U_P@t`tY)yg z9TH5QcLKoyypE7WPT#yif{}jyERJZeiygOTK^b`m8{1&MU>#tVc~PtM=X>21{yxx- zIIlzN60Ey_Q4e1z$Cy8|JwH#Jrjs{$@3Hr%)+^dU`!QbfqAJ8NnAOcdEQk9o`HJ27_`}hO6(3g+o>7KuS z+yy}w;xnzCno`XT^;LZz*XHYfV6N&?N&N%%=FME)FA1U`DFXx(1T%7i98BWCt7(o9 z7?4@`)N zf=FE$KJ?ctuKZgvrZ`zSqsbAz)_eBqE|Ek$0$C+VeF$GaHG}+U>!k|RNNAD=JY+8i zwTpZbxz{ormnib!xAVAaZ#w?~ZdHmhpU9ZNoEpaJF7d<>MAXxZrSj8@cAyCBvmYni z2BjJQl7)g>#cseBG*H{}_`BQ(0~yG1SEL;PU_`q93b%?Aq`s-2o9?pL zgncZ&;)0B8$$lYFa-Eq3s;^AzSoYofNo3(^f@ zdx(qWYjmF~#!LjA8@{&)p?7_14z$Ni9Ep|6GYNtgRugz4-|3JzBMFHi#7wlz77JR0 zR`1Fb%70TXmQtdOHC%2X?>i%RQrfKk`6c)F^Z?4SDbm!!(j%;p9Vc?)%;i=-8up{2 zniEw{wx5VZi_Ogcq_<~Q6i1fg)F{~U8^F3B5!gWC`9Vl71uBS;Nd^3c6qK1qCO5k# zK!vYw9JLeXK>EtDU0m&6iP3Cn)3i+)n`ruLgh$uY1uB z2EEm($_p|A0Q$%qK)Y~!r*dDNkgpWMCRg|aHBbO{hxiWw4}hP6WGz?TB#Aj&_Af^6 zgGO-pB?B^d+ddXl{7ygC7kogv5aTWLY;nLI93MQFup!m=)voSYQT!CG@Bc>dUxz(I zD*wNtHYB~`G~nyw>ERD!KqEu{2ue2w;ro0E#t-cJ!(wMssDM&@Q*fH z=A_SNMd_?&P-LeM^B1w#^IA`ZO^g}(Kb-|$2N4B+Wf&Mo>A@m*N@j5<`-!pEo> z2zeh^5UgTnWVD(`k()ayUcg-`XtzwScaNZ)aif*Zk1j6!4>ai-xFMl?fsoaHoF9TG><;Vq z2CQsD0?--!u8i#a0SpS;UdgRl&gTyy+g`)SQ$~|h z`3!qcHha^o-+wUqCp>z33cnwLgRl`cO)PNGB_$6BzI?_-v1s&ZO!H zc=3MK3HUH>i?(Drjz4+)=%3M-ZqD!i#pQr>HvUcP5HR{<91BxpwWmB1=Qo51U-ml- zoJ-6eO|ez+V-fY zb~*&)hFD3!RN7`1@pDGq;f&G$B@XTb9Ei~WrQ&t4-5;;p2}TW>6ipP8-LB39<`NrY zYW}RmZ!O@se1W-oY9S|HPWNxg?JL;cck0>w)Run7^3)Z`m4vnn!@~$31zLTzqt2%t z1jz$G$%8QR+x|j{_iI;XiJs!@b3Q7Bn$7leb}v*FvG^tJDe?vCedT<#nm@3+R|6GqRa$e6$cQ>+9r@>l!>4df(f+s+*FhSD||qUv62D}<-*1s~22!a+b1`wNO)sRmvr--QcP z6!C(7#xaWeF03^^o3+>1vm9Jl+c^rM{itkK_av;rR}_1_aS>&u6EXMexXT3*u$zhxBi)(`PD145hcK64oT@GuRW5*CZ; znB(-MY-X#Y-PG>&dj~dd!Und~pV03=+y+=@OAq~~RP&1wU$)`5azJ;#+Z0uf$$eft z-pXzPs1L%aj#IS)6{o9)L7lZ<84Ns7DevrjmzK7>@qGTVy6&S2aUsGrcKetej@lhyxTEvA2jrV z;t)5^#j;b(O-P$M*{nbz7AOv^!-r7gm1SCWsI% zNfRaaN`cXC*&lVYApD6gPw= zdq`~-H+umpE`M8C>EtNC_C3#7!ngOmGw!`tHve+w*LNum~?bag%9MV{-vc zjaL{-PTCaarsqkFiz03sMA`e|tYficbtnuHq=^MW9u@0&KAw1iWa1si*-&wQx-xw4 z16)NSIGtPx4fTs74niMJ8zbn>4BOd}Fd0*dN1Cj0VMYeTB*ZRwa={2;2ChePihK7O z2A8t^ZHDk$*8iz&&u@0@^8g4OuREr38}|n|bPkU<^oDTuAEj)&UgP^-?yMa^{^srt z460H(jn@%y(=6n5^ERc=Ue_`wm7eFfar6f5S=|-G|8BrJUt~ip;jvZ}Q!?>b9#czX z6W_dk%f>?fbrC)gdhtb4I3)-;`FL$^0O%5Ydu0^dYs(E=f7?`M^{{{XtMT{N1 zy%yOGSH*S7XqQnft{#*lnAzU6FU%kawugNHvFDA>-4^;PWh@wy>t5U>)?BgXe0{!V zlDJG;>-o3ygOmx3ITy3d#&$n|*8qR*bidXJ_y1Md_3g(H{^|ctP(0)exO6zTg9j~8 z$rDYi!I4`jpj8!J=Di0tS3+~@GAuKK&S7poy*M^R-Ig$ z?)x;R{(w!>F2+N=t32&gHal=PRuB?``V^>H~ZqB{j`E%@O4GGWM%QW^S9n8 zbIo@tR*B#|IrrSCSYpgA`#k7rA$LUsw0$)Q;8=>TL3`}e%-kDH)DgL_Lw&N#v2g*y zjKapN{HF;@g_$L=E$7@q(0SWEd!U3-*r5-#^Xq2vG1J)8y_gjVE0a}Bi282s_6^P8 z@oLXB5wmhApC7*ufLk!O1_&rfVCHQSLNxe65-q02v&;a1gWv!m|2HAFM9L&PVMmf? z|3mM#|JP0>a%+}+Bsvcr^Fr!T3t00+Vx$;un{aoWbe#NeL1PhFT97LEa+1K2!mvFp zU@DXuMfc<&>}$~4>8>(OKfFEpQjSNpp8L@`lw}`ZV$2M{XR zxvjki=-vA^Mdv~BzkH8>3P$QU@lFf`ondYM+Vx%Z`yqed>1s|I73D&ivdu&L7nVkY~T|kX&}}@)=gNp3mJogl7eVZ zZT-kq6(3TFY!77yxfm4Uo`V0SSN`+IW!r5*={h6ze%*BbvmTka8la*En{hFNuCYvA zOZqkp>-l%SAn!{HPKw9dHc9ve$ay=$?2Drem^I_OaA7#QcIZw=rEK#YLS;Pd?t78# zyc>FI=^KwAaVM?QVqrdmqExhXTDk4ZYVSEPF-jhV*7iI()ln#%6|G=qwo)HxV56!F zgmfaf=b;uCyEv6|9^`Xj1sju6Ez^a~Mq;Is_EyzS!^$eG%3IkIQDVepU5l?D`UQT? z&cy37`%)#GJKpen9#l_hlvv6^%DI^^H=3Bm3uTts&^mcZ&m;G6?)w$rh>s6Xuxqm6 zb7b!>J>Zi{@EdpUm5CvYs6$WUS9-_yU676aF{l46JRWx!clIL;TIUh}Z9l=2%OY)I z>W2o^3s_yXTX0r}=Wm4qm6n#VVW>VM!+j9Hj#m&iM#C~tacD_JlC7_|Gk2ZGEe;GE zYBbxl$n&9--$qyzl-OXYAPV ze|q%jd*tzd;_OSR@ZTlHNSG8Q=uTL@s^876IDc{RzvBgjw)b8b3?sU^RX`Cq+RQZ0 zOH%D;FT9L`ePhhf3FOUW6k9ZLtzl!K7kST3t)zYibM)mM>Oh}(ha3d0g557+5u^Ls zo(^-usoM(wmbUg>#MGPI3_{P&gOiuz#nL5*H1A9jMwNm^BL>5ELOF)r%CuqHWNp?#8BPe15)2VT zTO()|y2$E5!E08(xyfJs<590j!=LT3jN%k;zs}(mD+kx_GZuz(fYbI^d#pI20V~|9 zE(V<5;!L$n=1C)1dwHh71*SvJaR=fe3U5uuyuPTa=&F3(Kg!|PlzsGOq4XrjRMZ=u zTC-BUd51>ub($w?D@R|7qlq?)?~uPj?)F%Qa!M&tCm}a!j&2iZLi(O_;L?0l5m za*xi+rqi@P#sH#>|1l;Y+QN6xS>Jaog0WvmUwEHdP}l^`EOG;in=N?3yY&oV1%*vq&b+Q| z1V;AyZ-%ji@{Cc3F#|S^UX#xJxwVdpKB^-c$IDcnilT87a`zMV9{dHIwsy8J6_+Lk zR;!byXK%Z{CxRu!Gqqx>!wlFhEn2t_mR&aQ;X|G=s90iy#`T#mme!_46l%QumON(% zd(Q`Z=b`eho}9g(Q0QogD8QK1K<0~?Zk#O6^cCu|zMS48WQJ`+|Nd|YbKB0THTw5A z3vMs<9<=KV%#Df04(d?Oon|vp=oxw9L=14 zc4y(rBBlbOu8jM*D7SII4ya&1Ji@_PYa$J1bD{@clctux3PC)=v&`OC(sk3l@q85Y zjeYrkeCl~+>`VCl`}yh@A6Ow&7(>OwTYR%?&!gVj{oL&b+>S!X8Q(|Lw)Z95+(!+k zl6Qh;SiNxXB2RtEGIaIwBGtsN;ObKXxpJ+_#*og5Egs5NEEVr&Rg3!Da-{PN$bz9L zO12*}V5j1}Iuf+DtBw~)?|WRMZA@8M&eW>9ten>FS85Rbo<@EWDS-cnz^+8QZX_jo zpi)#R&zUlOpR)Bi{b(@Gsq4Wj_z&Np%Vm7d=cs<4@I`EyX|ZK=SYW}y^qqYe3|)KV;$@W?e03qZydy)KSPlRj~96f|d|Zyzi@I&O4A)lYUQ$e|PgXhgIwxrY#d%K*kxL8hvDI({iQ5)ns$h z;+jI>7l3~`&VjBVtg1CI*r$hEqpMuxI-xA7C`V?Owg%6@8u5`UVtwH%9tY`5N^Gz& zrG#83Sq;rmSmkCX5J6G`*0rI@rX8{vP-;?m+v@)!;taS_=zE|Qj9`5CNElVH@ge|O zw`(<@9Jva7>G@FYIS+Fvc&*d+r7R7<W}x z(uS74h)whWTZ1{xag|LH!MN$ZA)r&4kWyN}-)>8|Q=n&$=#qn8yUe2 z8rjzq-v%`)uiC%MRN?k^P3iYN@$p*O^Gn419Bg;mFpK#ckn6A>_d|YKAgZ54_#c~X zdv8T$`I*@y+p{UWY1PT(ePn_n3zv+!yUK&Ia7I^29Kj@%Lu5dEtrbe&Is9qk6PKmZ zfZNGAJ9DUGIDyaYYDWJ&BGFwkjM&dUFb)Z!mehe*V@A#1^dtOpCaE;~9MeX=x0LnD z%*!!=fOjH;l9_EZVQ|z%U5|>DegKZ2?O`t&vCCTMFD0H*Zq*6Z2W$>>$xqmMYF_^} zg5PrwPk59O@PtdC;_ua+^dF$OBJphad~Yz|*DS$l{cW;Rntcw-Qjzy&Y8=~RBbrGP zZ6NPR<-V07Q5X6LE}JR6(s<|(02mi5hf7$ue1a)j3?fn_y2kXq+bU1uB_)_M`o6`Z45hT?;*(h*gs)|xmG0lx?@gsZ&&JMu9*DiiG&< zxHtLTioWtvB$*5tMoCYSq&pMcAT(r9n@qN~l(oC_IS$q!i*dmKaO|(I$w-9P)I>pu zV5D@YT$`Zx(g@yvdwu^+v$bcKFmcKsG{LZ!y|V8)D-hsUM(>J0ZE`x~;mv%a&KV%l zk`67#Mb)ARXF-;!K+`JPjaWL4Dgpp>skNGL8lK-(iv}j!dIUE6VEMn-;-}X|7XXM=0mFh~^nG5vab#12ILG zK8w4@+;1?)d7ifaaFbAybDh$sDz7JaWi}A&7d2i)OSV|d`9)3g7?4FJ@WeI8*+y8GdQ_B&bk~~+>Cj`bL*4A zP&x(K<3GNlh5OKLT;RiyYKW}ls`pz?%Kd7x3FY4Et`u&&2wAdTey>F4U`!C86PD() z_Q@v3x{<}M*Yf44Lyy8aY>UdvE`xQe`@@%<)BtTyg92OVq11TdJL7>*YH`qcunLkj zYQZE2f8icQjpk5XzBBMFzCj1dv6D(Oa|AB!ko_8w<6Kr@;I01};czejK+gx<&fcDv z{a*}-=lk!d{r=C~_eVCUmE3bWoZk^%(Leb28Gks{H9-S0VIz zp9&lJ8ggt((wCOqF>$h|{gRSvYAKe~caqD7*8aAsvOxUTBh*fn)% zL6|Rmv#Lo<5{Z&^qdMcU7zn638@if^bvT>tAsDY|QI|RixLVZnohLTC@FlZ2-wc)b zS|Qbm0PiYhdZIlSjvxFRku~RW4IZ!7{_%)ni-XLDjFh(tZr{rkrGOJ}gWol~?~111 z&F2R{VqxRoX3HNtyb6Awx%ob%w1Yb7TW8*`@DK~VAV{7?4$J2#mLtngRy)O$3xDv? zyaAJX@3TmN%uLQ%C)sf&=Z~8WIV!TqJQCS!7*+-c4(VdFG7BzwoG3t?E9r-J>NN>CbiH5J*14M@%( zel}?gn)nl0V{5_uk8J zz-(MQ9^scc#=4_=sILJI78Z5{JLyl%Qw%NTY|~jQ_}y-_G}d945yF8r2Qy5kj1 zt|!$#h#(TsQ+2@$rVC4EG0le*m4wggo+e=qv6Q`KWpSa+knKxf6q}99susH@O(Hh` zJn@XyRN%5xL__$xAIJ9v+mO3d^$6Nc54?iX;<*PYIIQ@!SR>h4h7#)Kr4LTEIdp1< z-LY?Tz>57=bX;|)E%Vo2u$U#4{uhybzO*v8w4S+K=b|UlP4uAU3t4@v? zpVydT!f4jQjG{n3k}1rWy~xR1<}MHAI6D2n?Y29e@Zi(T*zjc`JAxnev2O3fI38~& z^L2(>ul9GGg1tc2gLMK=z_US zMQt%HD47eoJ&i1Tu=-^IRPuq~8K|*p-G~g!yq4N|W?rEQL>%r4njBn2hepYrEwh)8 zO(7ooEOK3IHoDOg9h}L*);dX>RB~J#8RE&ruo$JCGP4Qo^}@g#T~pwaXzr_E?qEui zkVo{^9n-zUwRME-hQ~Sk`>7>>_g?3MKYNa`gqa1OT-ik{pkMwD-ASGp`+T~yf&kx_ ze{8y^*$+#@!CdlsGOJgPg1tHIJ5tlkk!e1xI?9((s=oCm+>~p>UfHw0_W*WOn1AG3aATty)C0|D;~dTqjkf|~5Q;<3|?NGcF< zu`y7qnK1S4q6%F|JeEZ^VB{n>HjVn!vty2`bk@M7&u6%Yt(-ge_$ZI=cA_ z&YAD3!*-&`tyvGWlYpvPMIDvHpW}x&b61!(yn;bM8skcU@@uWEe@_laV0uDeazFSH zWV{)=KO5#2{X%a;#MM>52(1juN`Wxm_3H|4;ozqi+ldnj7Brm*XGoO~p)gNPCm~Gj zIdvmE?e`s5@A=>_I2_A~-)*-M{Gajzqnao$z-@2PTK~NXF9B;3Oh0wUX_i-CA~uU9 zldE=cHNGlGNR}lSDpnSm#OErei>8$=2`do+MYA#rS*Ef0eVtQ^3f0&g3JUr+3 zds`PT^c7`yw+sl>Rg$#kcILf!bzRy~G#or%{t_xiAg5@LDMEuMckLN8VE?P}{#?el zOO#g-an(jql)&o)dGL) z-jtBe+lo23@;g#Vw>6|TMj6r8IRqtOc^cF|Sc66$v(V;%M4gL@cv^ z$XZS_t^2$|hxv_mHDUBWj#@z_*u-q$<_rm6SDeJR#_&1J_u|S+)IwIZ z2Y#oW>Y1~MD1-f$xh#`P&azQsLLqI!Rfz$E02U0$Ar^!yH(eYvO1MgmP4LG z?ogxBNZFr5lYg%}I@v};s!(`(eKYWak7Cz#p49ih&gk=Bvq0@{XgviKvLgPZSnO`Q z()$C3B|Yj{914H2A7$yl`k-McZdSGLmKADM>aa1t9l>6w7;@G_jYfuOCX!j2jetXH z%BL3Ni@-_JXtJiWTrSoBQulRf7R(MV7@iPPntM+V247RZxV z5)lSBok(rvV;@l0u{@<`z};bYVhRSP&eSs65O*F9d;A{)xSV^rDm7Qzz~mlVjj zC^oAtr!+y-S+<%|?P$Jx+G(B;9m~Vny;-!#;gQk(kw- zxT94}jUKi#u~kv2c%sX3ibG@uvbjfS@ML>mZQS;n4A7SjSNRo0q{`reOP4OSy>-B| zAs|Y(IX8MX(31POY1vz)!Pi2Ac%*{XHRjOqUci%sqO6aV2~{W7g2QQc!pTr{GhphA zN^wy(QK-6zgEFW=Lc$|lxgv^D)OzZT%v%s-**-dDcuUY`??J!6$aSBJZ;y+IeP$08 z2;nCD;C()Q_r2A7frZ5Mp?%qvC>dbX!Xw%D<3F4SBY0%a-s%eiz}I#?!Jx`z3pxxT zqQ>Q<01HvQUfT2=&ym+^Z}5tv*Y1eWy?b*VJ{kf>ZUMXP4_XPlJ?FKM@D{}Q*O*D(Q}Z=yPt}_* z!vl4{Z`*+tbF6>L89AdTzJYiK`uo@L{Ud+ak-fJT#xqq;#dH*^ZiuayK`VciCKi#l zvVHuJFOKr#g70*UPRT_>o{b)|Wjq+KbTVtZaVE3OeALsw;PYS){L7`aEgc7!ja|*u-DjVF`BV4Lmj0b#? zZj4mrclY6aS_jL4a7`lD1fMBEZ(*q}?R^x4q+y*MmDwL z{A9W5dR#`J6yQ<{Rk|v6C4@FKBw*F!WZ$l~~cZEq@F}*(hTb62|l(c+kb& z#c`k4cs}|i+X(>;=rTep2JD(@@r7&FI$aVYVSnZTj>__Zx>xYlH`9{ue7RSEZ0p zh(L9n?B)@9s#r}*30SOc z4h)78VU^J{Tp#NVnKy!!3?bs8Y! z<vxq{d?P9HMO{NEym5<{!zjh$s zZUq~=Q<19oZ=@?%>FgmHs{H!702<>SAae+7O92>36pe(L2Ay~M`@W^@z5cU|fc_e< zTXxPwEZKA|HdNQj6qwPe1;SAstSekh^<}a1TTC2Eyo)xK{Mn+ct8+^n9C#NYda~ZP z?5OvqgE>!ebg(_=L0`TzUGaqxQfqUT7X>^@VI5nBOB0-=LT_N!oNAH{&e93H<*Qj? zzFn-jO+a0g>97RfUq}va|1F%a7`_W=skUL@FUbtzuaLPS@V78>KxX1UTdB=6Ea*@H zbz5d1!QjHEc(y(>kQ}}9UA)gg8gGarpx6tl#zZ2vMf8=zjmhAu>-_0?#DHUS`~D-c z2*ynp#M!QICqk>4FArooFv@b^Q| z%68O2fn)?r^u_R<#c)R&KPBgMU8UHAa_n7daiQ@7f8jt?*(+_xFVUi^e?1L2(rBo# zalzKd+gVY(?{z?|v+y8kDr1VD5b>7?i81$zg0Rp=~ly$MuEXlWscD7QY}Y&aIU)*dHMX;M>@`wSjnaV zbJ<|(QT5=X<5^zCYu?_wr2lzD0ehzu|AeE3x0QQfbvnsI$N(-B&)82B&$#$8W)Nca zqe4htN3lxiSV&r5;F+c&Ue)QxLDfXNEcB8(n>EEt*r(j713hJztrOHpyXcv;t0R}v_Rry zkYas5#WjUXDC?e5BcxiIAynbn*$5CX0Z0oEGO495h0wMpwR(~hJlM0jXXq-DxLE#9 zs7SW65=o1G7R%#v#NE80qvR#mqnQ+cQRJJ%j*Ef(4iT|IfoK)Ki)!LwHb)1ooSw9% z5kAa1SH|o1a_|)ojaFFsuuhTk3=T+{gR`CWfaa@A-%RAZ>*NHjlBPPUZLmmWX;fF1ci>mz&ovcRn#C)Y^)K9#bYAK}ZU?h&c z15(>&fU-A6a`ad}axSFqs{6T+;^*VV}h2suK{Z#)+b zzd4Vd$o`L-m$}|KNYIzG)onu8{~8{;hxGk9I=)tq;2u2ABeinRROUrpj$B~s#zBOWLpGtvNi3DNms^j}bR_Egv~E!?=F(+Tps-`K ziA|?kbL5xw0df|t(c~UnWoV}I`L}FsU6o3H)s)r|L_MK}l?lor<5FmJUrgYb5$Te^ zK|)&xljc+(;tQ7;qnY4c&PXk0_w6YYx;J-H3h~24?GveTq`7vx=L;>w7Onfs$w@5W zES#0w#t&236`}gG2|%~%&y>n|HjCb6{k=U!3^HkY&Y zv<)0#_vppeeNRkj{2%kZ-nAZ9%pVGu_#`Gw8SEUo(*>NpP+=4o)-r-MqrMJpEKAR+ zVbSC*T-2Cq-d{o{))yEPj5MdRR5TqbjgtuEIzY%}Nf=}aSo78xSg??UUa`K3hqA+l zEkT8@jO!!@YR6IYVYNxB=`QhEj1BPtw5MdR#pQYhOS^2ktV~HWCOiAhj?7s*T_J03luJ4GXonA^U92dT zVXUb8}$V`Ls8o_!n?r6`#B{jy*>IF2#7wyDi( zyYJ-*k0yfm=-orLt_G~nHz-%$+NwNrv<6T9-o2yRTuwX)=P|MTBxre7ljf(>S&IZM zajXy6e>g;{61l_wCEJV#I4}niZY;>31T;3Kvo{1eMM=b?Y4?8I}K+C7)Uy_%{~7ryQY)77EW|=z9R+KX}v21sbJQVeZh=A=o$RZY2cH4e56Hh z3;aN#6jtidR7w>*oSff~4y(dNK_qFDQ7}8*vBKoey3yLmk+741oQ#bh9xt}rih2In zRH9K_QWgmpB-;brML_e)7Gi zini^al)O7Bl67aEU}{^D3CLz#q#*(iOruUtE<($hxv!}j&u`v(QCUQrjS*ShEH4@g z?p-vHj$bGKJdZabTojfP_(w8JgI9&p*X@^?o^if9MyY%)2;_EYTy78lB8G{(;noxtDqT zg>9+P6YC@#SbbiLbGqgot;g+d>vF)wm-t@_r#5+KbQ8Rx;QiXdra%zb_n`B7bEAk3 zU+`tSzceuQQ&`y`D{?4Yb8J*(3@#Cdw5oLWPO%TLltsh- zK5@Q1l#r5Q)=a)d*OWDG!97z;CciylRiBjBQCuo$yRU9p3cmAgrbFM?5e0Z-*sx7JVs7&6e>&&3MZ5&Y{a{#V8tm$ zDB0^_xro_GKec5DI6TpVq98V?p|sn3)L{kX*hd2?_cYT~H%gq6+yZY>4Yn%d zy`(Z%8l~}VYv@i;_Md$p+Zl*Q=^sejpz-vE> zXiR5Pcw9Fj*)GF7B2aV_hWndAO%NAS^Exc1IYJi)5G#}4WNQK4}P z1_yk+)PiG6iN16-x+Bo&<_OT+PFY9MWnQY;8B$+-glU zZxc^u9bb6(2AWAhVif0vVVQ8-);a(OI~#P1IQ=^2P)s zQ}qpCI7q8OoPm0&nb)V#b|J(&pQrYkHK9>PAxW7i%Q8B2lqy2TS=E951w#9XlAWH+ z)Qm=re*&%<)+nzA+jE$jioTw-u+iKdd1-eet=SzTb06^bH-4P{_Bc*ffU;%KY`l<) zC#J;K5phuyiiB0Rm9X&1sJuK7%iYKGOK_bKl4o*iwTr?rlcuIHfn1}O2xT4&6Ok~s zRg%PW(#+9ll!z#a2@IOkpwU?C3?kDJ?+;IA93CNe35vwTnBiO_oEQyg9weE`$`{O5wQE)d=66+cI#hkiH6%vQB4mZdq{B^Vn6OaBe zQbtw$w4*ejP4M0$NT|&;1T)YYQ!&=eXciLQnKB%QguEJyAQgoY!8dFu#uv?c+39jb zMP1=|$yIOXo4@_(WZ8o_mGnM`STAuF_@m$WhY6?6WazNN=%g8KBk5*|DSGrTYEQ1X zCNeVH9=3(y3ivddZzQ~Rf^XIWCPh5xx<* zS`XS=RoJ_yM{50yf@p1J$YdS^W&heH-E_$Iput8z&&eT zjQduLKxcXuhIVxP#Oc*fX&44UP(Oa2W3B91`_sQ5Wc-yp2vEaS=-fVQDiR3cyB)&6b`Iz2y^J^JS0cbHCtG3-0*=!Y-sW=RFhi&bNGg;=4~|#K6>< zWfY|JVPD0x5PV@qwxxJ*z824^-c12#Lq=F6MkgtF!U8)_8dsyx_KfPoPwowCtPRK6 zMhOayrlXOTm{6t~Ym{cq13v|v58=y+cuF<2l7eo6tWHDj(M$ozI|mVS%>7yVienffUfVHNgxg85u=! zT@T{Xh9JWQ?_Jd7n$*$W6C5{8rCnZ-I$?IYc`j&gk{BPN303G9DWACQB2eLym6Ia| zra~vp`GUhYF>9j*V zChtpz!vysv1SD+RXi_?ZX{1_2M&_P)GxOZ#qC>V%K;oDet7kDU>yIF28rKZi9AHWh z&a*Kn_~{?`GoJL+NATDuT*5P+^F%Itr32C(W)laPW8^?HoH9B5IqbZ{@>ZlR4WoZ{tx5tLZ>*AK4kM0E zGcYVcfKwVtIYUgW(NPnbDMvQs`eCb$0mu7o_BQ)7{w~$WX~H5Bh|ZzbX}lV?%8bpT ziBp;B14PC}tI<=E#M4f4RvSaS7j8YhK4KJuB~^fgEHW^a%S^LVKqol$t?3?)_xlX` z2Jg(sB&2SEqP%|J6~k$noAgeK$zxrUnX1>L-8vp1wTiQXt+M;pjrxU0wuT(v8t}ST zzK?(S+YfTd6$kjOmwpe;_D)3h+(LnJ2kqAq_ZQ;KSgYKd99^9w?zTFDerj~3(XyNU(|#sH!;IyoYWOiX^E zbY^5WUW7f(9;q46TTt4PbJ_l542LP%g7Zy1Y3AsP%B7u?Y?loRZ?NJxkI_Mh{w4Z--34VF=jq5Fu6f#sBKC4y8#@ZH=aXuAHAT}b=DR;F;tC7=b2Gfx>9xIMknvWueCi)=6bAx}% zJQ$SAR!(9Uc<9e-w?b+>&7@>+qsMxq7g)2};i|y`*03erbijf376&>#*0Mh9%>ju~ zvcz-g`YBc$1B`e+ckd;<=-+-TUvS_G5<5%xtM<3e2DUY%v$s!cPmi=a#CC@ZwIwss zP~ej&I;RE z-X93L8_dFOiTjR!ddIc*$p6Wwb-|^EvuMK1slKjF=sl{SIN&H(9r-~=?@`6LsG|UD zC3H$O@nP$fXd}>`q*0PI%a}l}=64}VGbErH|ehRNE_i^nw^4HcYsYw5;1(=XFrXrzVvIk zpw-3BF_Bft_WEXn#@;^N%T8k&V}^0sC1!-PDMqa!B64~tHXf12HDYvF&<-`s6zf%8 zOG!5uNLUd??zrkV{(<9nZ=6NP&K-6mtvRnL+{Zt9D@ht^jOcJwrbh-o9_imFU!Fkpi zM?1}FCIziDU&x?@24Qbwn}e+_R?-1>DxpbWaHg{^)lUx`K+EQ=CVd(v_{JnMp7cgw zJsr?WijjE>l+=_QZlC6|l@oNc5Z0Xi`bv^%Nw9PDoMh&?Cyv!{P2Puz1MzxLGU z@W|EGi2)uHE}>LE;wh6vNg5s_64l{4KCiMxFCr`^L?xAN4F%bzBaGr86oRY;`dKx-==Q|n6&vQr(Fc%ABPgkWl}a==uYot1P*+lFMD!AV(Z*n(~&XQeqD z`OLIh#SC+jH5{GQ0nfbh0-pD|U(A#CAL4b-{`MW;d?nIl1Fb{bV_G?*E&(QL(k*DL zl2PaI9=;(Qdvb%#%M!K^8nzBw%9fc`hXXiIa-xszmt)$qAo$LE<<)=j&pX`Z{e#^` zYtCzPQkuoefOS)Fsj-gGctDjXb*FO(i_|jg*5_hP3+hEfdHO0N><+X{MT4nm6kj*7 zH+cGYa}w*Z#?wj*RF$MwEG=2&RUg~f*Vtyg(H{jsG;PU7(Hs$mwY0~2GT>O1o^}dW zJQp`lk(!EF!oa0$y4f1^O-c?7xlb!Nw*IvanCqGW< zJg0nv5c7Fv81b0IjMJaHc*9;|ou|o3qfo0;6?SU^ya&r4w(}&-9OYm zqd)~QC()F(C0{EUTz;0J`&AUB&eOOmX``UgbaVHCV2$p`z<_e7vpLr2*38FMr$u1s z6B5biYdzFYrJPZ640%1`DE`;Uayo>0q#jP>?J1r@DZQD)MultLynU4Cy!H8<&Nt|} z8WA`H#3&|n_~HiCfu<`M`h?YFKq8L5&*-_7{n-|onIvv4TYxD2TCBK;vYz%B`V=Fc zt_|Vb_4C@c;G))PZttx#k6;*qe>rsxRur4Sep8|zP7THcYNz_HVnQ7hEzMYOm=i)M zYqj5$Fy{5J#A?hT6X;=FG%|ztIwC@v)}kkv|2*$+ z?6$DShZ?^8YoEe3@BIve{%l5Ymq7nulQI^Vm6$kj-V^Z&4E6-Gw(UiBpkfY;1_oyO zM2-I~TFhj0(8NiH)=zV4*kQK8$bif_r;dG%tr%4_W?1G4aGOPgw#`W?SxfpTuu--t zF<6nxN8+4_{o!%<7si&j=JEYl!y_2+Ahs?3mSJI4<#wzn&WZFmaxweis}K`CP>uA%riy#%wmhN?Wt-;fPecXk9$D# z$0&l^Wzx|@4Nv*EkK)U}?kT+PuisCRk5?$42%C@W1}Dua&B(NAvS`D`EX59P3&ZiTOkb(vc|``OH-QGeUF>&H`gX zqT?coO@~PwN#?N@lGM}43RYV~I*k%*0)sM+?wN%1W?VB&pk;F&wQ`)*EXTy#5`jjQ zKq8`P%aMAxAvt@pEvy8o$4E>dmE>e*+VP`~KrIc5#bfa7Nw-Oi;wRKmV+gP(>(REu zDBUS5$$*t4r=snNc-pqWh+;)ajnYa>+KrO6W^Z)e>b|{>iK;P5t6R|B+aqm7ty`kd zP_(J2)#x9y17m2RSXF#V(Abm}(-Zd>N;)8>AeKe$M3Xlidrj%k)S`_9agI-a_-6k4 zMQ_JFKpA~i1CzLDTjIV$MEKqxeKuG9)erKOU;oq*Q!&_=%@&u9p|%y@LHxUQiG+lk zm8z~%Q<%6pNKT~4=>&#mqFz{4B4A?6ff7utMsF+_tHJk0gu6FZ84S}o)iXk`RE4~> z6A8-qHhS!DZIANC)X8Evr`AfGqh0Cqs8-N`7-97HaCVA>`Lw3)XIJ51wh^_=JEyka z*X~81LB)#Wa}V4;!N63oM!9t5)QDbI5aJr(gCIIM617#LbPo1dxnz^|D^4P|M2*L< zdfb`|1u@N-rCS6|qUF2#KDEDdhtk-JeW|n>XW?%J;9Ci6-+{qzP-S(TT4>6XzlZ1@pD~*&MH=v z)l~5@0)VE6P8rJDujH6+fz3*qow13_h@+L26AXwoI+5C2p~ln*r4K>d2G4rdTi8WY zTlf>>a4M|W9{ZEcQQOm_keW&6r@Cq?C6_mja%J-_9@RWT#||T75|GHe^Qxf|R?{H| z8e1G}Y>q;|M^}xXD+bDp2x-V7yuY)_N@GAX%@O>F+c%;kA0ZKFv`f;p0a%Y@C8ile zv@;(~K{`WZs&B3tOoM7CgTMYNAEnvU%cSW{MwJc zVuwb~EmURUw#2!^MVIX52cGvIX|{q^u6%;z&)&YBrorzcA;_islSm z!r|u0h4g8~&eoVR;(2uQE*{&ygU7XR=kcA}u#_BZ1pga@?kH$k<2lgY;)3oLk6u4Y zYUcU<1T<>x2+CTc&;Ir{ooq;|p23|RTyC+AlI}&Dq@9@7!a(=%7Au!-1Z|lCEVvbi zrh&%HFf(-quNk!Ym6m_Rt`(kSP|3s5U8z1J@suXByvMr+PHbJF|YaV>xGdvD?BU8i{T<1gk1e)g+580~_Rm=GzlU@Vmw zwbF^l)DqmyJzf;oq9%1KzU6DFr-Q1|wmf1rGANETo%9>=YH;3=)N&~#;z^7P1a73i zfCNu!9POl}VZ(%K+QpO9pjcCoQc{qzYWu85yY!-a8$0t=%FdA3?of7`Ox(y}crEF( zDe(MnPy-Dgh7$ejUj~VjXn+Q zMhmT1VY^5%CJ=;CQmKf-+u(RRV-`=aacSWwjFC-jFw!~JjKm}+48!KY*p3x--kHsFcp zIfhS#eASTjf`-SGL}n*mE3-H&B!i$9*F>OSB8>zsJ>onzv0VT0+wXTKM%jJNcIPt& z{haUl%HQVr-Q%KrFMYwA`G>##ATNIXkMgn?zK!jTFfW=%hZJk+^b`Ze?}{eD+-Zm4 zn)4VQV$w|J=*H(iS4sRC8JH|7IkdJBM5OrI4ogevE$uYt!uIBZ^Cu@nHE*|FHlZ!9 zj&R;A?8k~7a%tl@8)XYcShamRW;ki8iZr`p@{#lK<2!fJG>%VgUqIibbdw=hw2zK_ z{#4;mbAwYwi=j>@yQ}&IH*M|#lx9+}r#Y$uJL>O7-el?T%1=;lBxN8Cw`OBzFXtPClmZY21tO!m2^n4p^Y ztH}U3!Ip1F^Ae0}D0z%PCXQ9p8-;N8T}ICbLX*3|#H8XVRmhA|YE7B}hUFwtsrt}J z9IApbp>9*<;HO^34_=5BVW$d0s>TuqCRDiY2EQje2<gTHA;f z=ErCxjx0LcoXnf-$p+)p4$-94cWKT4EnKr77y63HsE^ET!#(WFwiiZOc1%tytLW7H zkcMf~vno8Yb(Hn2heqvD-2wS%zCx}^)Eou6Ru|Y*g}Vk_nrTTp+n5_lUO2OWfeG!o zJflA5)6n4iThp9`fl-LrHOA-J9$=!Wus_wL2tE^Bs@T5Zy3N@uRs$dc6J<@pTD14z z%Jdoc4|XA~+4W4!&DY(@bHDbt7!DnD=I3ANZRcF`-p@d-u8k0)11hD=;2@CBk$?#5 z-lEOe=Vhu0vnEPCv~ILGjuy}}qsoXm?JE-Y1v5(`C03woEY@tYnhw~Lj{RXgJvNG9 z-|E;Q4Lcd`*)hYKBQ11wfOrjzib^O1NyrF%VRD+T834kjZ*ba8MhMM@UyiYs3>Xwy zJgVxQj+h=*IN01mW66TGq|eYN^nJpirg9>0b8Oh5ljazd+j?t}5elUY7DaA}fEXoh zxR9i1 z_*u*PzAle_@}&#M@mwMJDC{o6az5fkzwvrr`?7zY%fM84!{2=nYYk;N7G4?~1=%MG zd>Y=r%_~=f({2qyh*)Xf4gCw5Xv^P=)-7#F{w!t0}6eof;$;hu}#K^>J-_qa5quWPmnPIJdQt;>7Z~I)C zo}wisL$!P|-^1OmJHONQKE`XbBxl=CnsheH2Dyepx0M#By06GeAfjB@+~ibIP-;Rq z$r-wYQ$=&c>=X+N8HPsFtN>|6S};1CQ$Y~avE?e7wBwqgAu4wDCMBIJJgfCj zJA~$=AkLwYfl(_^j&oIg(+~dp1^04pu^Va4u4iJd`nkVlxEWPvtp=F+U2bHS?P4|# zt{QIRB&4FBL!63MEb1r=UZ4d<9!0|g^-i!!c+s5`tesZ#hB`aw#~W;Qct0Vq0dJ5)H0I{QpXk~;_3Me6=B8n*<-iKq$CC~0LrQv zfWn5(4~LxTE5SKWRq#5Q&tc4jj8(QqLilzfjzW{bn2Az)iy~ao4ym(;KH=D4ej+@t z;GDq#okl_7EW;v5Q3}{3R;pbsAvgG*MXGAp2vmg}ryjCq;0lmZkRhY6>e0O_lx@Lm zEnuxocn3^5P2u5J|MWW_pipbUoh389#C&X=>hYRC|0lewnAiEV!nw)!qk;go&{$A* z_B{JI*@?Eef?~@Z|~DNX3n&RE2Hdm=JVkghG^6(8*FZ#Q*@E>f_zRvc znheU-5kohIV#H$wl6Fa!onqr;n_+*PdSX(`rua@|Fsrc{`OsCGIROis7sJ|TbmWhx zrta(=j`1nj*;GV;BBi2LcI1sM$V+sk%j*6W{=?UQ{?3DW&anGv&F=EhnZn(-o#2|c z$12rH6|*=yIaM{-sI}t5Yx@X9BOQc2-o#CU3A$_&GyE>@hwcz>DhUluSx4>zt&W78F61Fz;%?30Q(DB^6J*=KViU@@g2zo}|YkvZEYG zw@8hvF)-CVK-UanIArMaPReRQNrQ_ZMm(t$tlMqYY!AC2p@PUb^fDFC{$x8UZDph- zZJ3e^+nZ#jU_wSwDBU9T$8NqJDQ#u!Bl;_Ff6hY+b68 z_@QhIMQedd841pLgoIVq&6vcpx;{Vu&ACRXByJXNOPm>h{(?8*-Olg50zIST$xg%q zwDh#Zb^AMiB;TnYJ)*x7mD<&*?>Kb z0e9rbiEcBz(Yw=d#IZ!iC@eFqn%y-*)%s@>Em zpW8SRg}HHm=k`#lVb0R)zk7 zjG0Jft@6oe5SEG(203^S8=E}lafcotVQF{Lnq94ldDRQwGJe0B4JPVH$k<%qIt@w_ z)v;-C)KxpSO{#%mSAz2;jP$8WEzwxcx~XTfB3S&OG*hU{x>a=h#YjnT+A20Aj}_Hy6s1xCFO{TnC7o0OR@k0H z`$Z@^koGH&q-c_II`*e2f-`Tpj^ds217G*M{OKEi9GfgsTc11ZHfOu@31#7U^Q*6! z{BdfMC_#mvw^FbM$t>xq0X{LYS~ZxVBkvlDJ$9mvGKs{-NbrlW(cCaaY+PrY9d{bi zsLf#kMA>t2lMDB5;1v={x&3s5)OvYQZ^*U(Xu5$AhMV*&B|>%wg<3oavqo7#!5V8xsRrMJ zJBmGX`%~8fZ_O#EVg%N*A*n6NUFA2ZbBpGR!0O%}Hg#;C>>x>jv})2$1{@lqcQ%n7 zv=mihP6RX^uVa%3LEbVnw%k~&t00x8*DJ|qZXe&VDj#^=hxm*C`x?IGd0)No+UF9x zjn?dHG41+~-V#347lG1+%{jz*RF!hQ2@-QYI>l#!i6IHKaYFe4nOrmDXIli$d9?VM z@xH*==xCI5;vb<>^&C`;HKMdLN7Fj?GzY9V2Q-o~G0IFqHyLty`#39UUSr5ZyYs>_ zm7rn`M98pW`dplzWYvt5J&n%`A3A1OJ6{%tds zl4AtgW=Iy*y_JO1H04}vjOsW;C~C(z!eUZ$D4YD7CJtBC zJsERrVQY`;LoI?-B_w>1mWw8o8554)Y66es(QtgJh}Th8KGQ28FN2g+I4B%PZ#~Zc z`j!{&cl^Iv_CjG< zLf)08%~{KO?9F-;CT)*IfnpjOuM%MJ>`%9O(#kDNX5I^ENXa9c$9YQU%(Z+xq0bj2 zo3x}rW{TD+mXZtYsgXHpU%EZY6DR_y@jN0s#{5c3j08e}a)GK{RcKmIW<9C#q&w8C z6M=SjNSXx(X5tZ7YsFf)W)PsK+EOLyhinj&zcV?MBB3~>0?s3a!wk#Ndj;9*A;VdF z)1Uv!tNGAd{&ivh&n0#nt=ZK~%oU$~arh7%B1WsTs-j*|@55UiUjPMM%S;dvA=ENS zx99uIj6&q9d0D-XCYX^l`pS&5@tXM8R>{s3llYo?bP~#}BX4!>mMvIK2dpH+nPaNl zF_Ud7Ey*WTv?_)~JO|QkE=!KGE}NuMU@2)bWYug>9uIR&Re{hYF&e6AO4dxDl#+%i zS(9xpw5K?1PtdM~K`xSj4kag%+UyL0L6cQZQ{(3Q1t_C5=|(~|>~<0gD(Z+cnY^Ac zig)v~2?Jntq?sE%rT9V0_#Tvy^g|R0Cj2d%v8p*BDimu5@qu|UrN{OicIdD}k6+19 zYeqd3K&|ED@BI|N^W4AS!|(Vcem3LSIYD4dcUJ)h{@tGIxU_Mc71QIf%_Drz!T0mS zhyH1T)(EqLm{-^gjzfsRBhsTBw3}p7kWp~ZZ19-m4lYlR(K2(LLq<55ZgNHA?r4;* zDrGAN)6I!UsZ1&}`SjpG9&6Ky=xAF65ooQ>Chr^FY|LGyNzG*F6`W#n=-fFzH2j@1 zI#ugf!?1lI!Q?Im4b-mDY2h#0KbLj;`TzXeT>ZyyA7YBV&b=ccM3~G)rFI+w<)pSHm>7Kht((-;XsB7z z=?>_&hjiORI;}cG5{jHH(tIQm)U@+=8=MLm1r|p_$*SqIGSaFjRYqSVOkpFiCmnEc zdWwteDK1DhS(gEVVsY#@+gy>H=CWjib$x|F%GelyxObB7p>$cB6yUltsATuRx zlhcx6xV9K{%G6Y$VM_Me4KgW7r9|*EVbqxvGZ}z2mV?NB(9%>ols(PI(rc+O>^Z>H`rrCvZ>sUUIPIzf?IFS_CnzwUi@}`@mqe6 z@BgA-VC!^m`iT1u4JzJ6+u1-wo-HERiOBUL^1l}x^9>^M2@(0Yh`juq-ot~4B(;3r z(;qedAsGItr@0jc$;)v)jA>@4hIs;01>=PF9gJV@>XJsOa<++pq>5x7G4n4YRL^{T z&Z0SBY9^(5oX$qsVzX@21dqVLjE}5lW4KOcO7^4!K4uNy}`~#=7Z|G3f{k z!N;X71uHh^LUWqSlGCi4;i!(%n#s8|9k9o2b1>;~zzo?hIeSeW^GMVsD$lAJ(lsUf z%z%sS1{Wn8Ji2}PT!%6vN})-}J0Xsi$w_7GGz=ks%p;md>Ds}FcqZc6EQ0E*e)Yt{ zg)5s_nP(bORF$kbI|if6`DhO6fLnp06Uq}z^nB5}V;F}$M&~=3Ek;w>57Erlev0~U zRr=dQZoB>nFZ{_@F1XNrgL+xGoe3ghf!_z713V7+CJ}ku)G?0+eggPHRee10!{_=Q z9xOU@!?kzv;rE2J?ISu8C#-0EPpqKr3}KN6N5BFF7; z!7G0Kb$t3GH=W_C_YU`k*4!UBBNOvr5x9@P?|N?jC)4I&@xhd!Rzp?U`9=>AXmm=9(J3MmxFx~UNrxc9R-Q$Fj&R|Y&Dq!3 zicX=zzz4_FR7yH#K-&y4c-kfhgq(zighje*dhEAbB+&uQ2(+Z2!ECa+3P}u|oFRtf zteTQmOj0KTnJ9;nAr~h-_Lw0S$C??EiXRn`3>upf>)yiwbDHUaZBTM~dXz)SDOOCM zeaRM=q{rymKCPJbcf}4!t>@C_3HGHMtfxIz(jnbs$e2KkJDgh2C99`vjEqugqf+C* zDkhy|ltRrD4V7p^_NgcyLa`2h2OId*qdxO*^?_9d=g_i*;kbrYnU}od?>=zmYu*dg zl>^h;S>W@5>s9pzRn37{0$)9K%y$F7uc{}*_4Mv@+5ZQVk^1xBeG|jMr1V}1Mit6V z#%Q(M0YWHQF->dgP!wX8dqsob)Bw%R**>q>j$xzU!uv`v81JcQ&}!PJRi$?r_0iL@ z1J;usZ8M-{hOF9<^lrQ$crpp8FzZ&dMOvYqS-T(v9kfbA7}2ecz(rO$XdG<`)fjUu z2BjsQ6s50*rq&Jc1T<1YcrAnD1NMNL* zt4(EQ%DVYXHq3f2jR_^{je5KY2_+XbHZkI9+njCRh%$&$c;QgGMbnfch0&@Z1~CU> zni)r`j(v7EGo}bMaWknc9Kn4O)1DpIZ<}75;RcL=-MHtT)4vCY@5s!+?~%#8rCR>S59-Jw-W8I%CJZ% zH4U|wItBT_lMj>X;e~RtP9_H?ek|IMZ7HIHPeK^>#4xF1^ZIv2ywP0tRzUhWESE=e%weN?4x^Cb$;AvBBTnQi|?;*S@Kd-9) zc<*~Xg9n_>0KD~8ALM;+`PUGmzXR&R64VsCsuT6vdI#)GqB~$BO+(&M1_v2gI!9Ua zx04WQU=<@(zV0#R&S6oUx1_efsSO2Fqt@iY=4l$X;B?s@9lL6Ktfu+s9w21ujp-VL zvXQVN&Ewgs^}XJ+lWtCi*(p?R{I#JP4U-YbRA~wfG^aJC^{I}(#P;Z`;h1l7%ymXt zg;V268J8nC?Yc8eRb_57c$`{N<9Wi~JGk!D0eWt#K2q#cDg|O<0!~$iE{il@*wYD>K+92R84~WQXRrM3KW9ov+Nf9}CrgqPUi8*HkPDH*3_#S{uFI}9;^soNk z-?M#s8%?sA2dw~&XajPyBC&W#PrH!Dxn(Fiq11rH7g4QHHphgcetM{0ER%1T5>nNx zk|mZR9g?(*z@#ixpD7JnlA02)LL#0Enp+qXa->*M){+6DjBs!ixHBl(zKy{5H_XUoz z--}hnN{K^YRQh1SSecTR4)JQpv82pT>Y^IzNTt!me`}D%qLAandRf!q)r7=%)h&)z zoce%;ccAD(dWzuSJ8?{K#+zVbP-GYjfO@K&cB9T!3=Gs>A0yD|z~&al3Qzs_PgpPx z_YG%|f{v-`Q|{v>65!FmGk}YMcZ$dpRQ1$-Y`4NWWu(rA))aZk@Bh?aqlJTPVJYa) zUg4ugZysJBYBGpv8K{}N#|=3Qr(Kgh$v79v&JM0hN}ED^2ogfDk*NtWeg90* z+nIC~0$DU;Ez&59K<1TWuFKdXZ64;c2o}!;*=YtoW8hP^U4zuRu#U_*KCy~59kAGqnWh`iygLG~?09%(p?-1<$IOHwJ`s1FCZ?NiqAlPQ;$=5TQ$A+W#L^T8! z$lWK$gEDjbN~6`_i@x@$3&!WZ;htymUv^4nA1iwEuI6v*U8=PZ1Sx&=xT1mm>-Q!@O&=4l9C`~hDPudUV2Br!=){%;% zW5z4#{FojPlnYy*!$s*T2hA22m_C_-HRH%?4B6CZhLC0^8JN(Xj0!(5A{BKLRXUYv zwE8`X`lfWGfl*RQnv#=I1cIU&nejAC$%-9Pkjyy!MjhCwDs8BAJsXiE0$HORtH6y; zzeeNR_kb5%gK;>uj`vjpp%RFP87k>eDcTm-v|+)L$o!(%)3Xl9m0yf>TCDC{;Rk;G zn-`4ceZgWI^;z1fZwD@^eZM&VR>y6?YgM&S)f<6N5$5~@iL;_JaCWrj=1<)fa`EIuZ}AuUelnNJV)Kxapj^~^B&*iaG(#>=H?RnqEb{MTLXeo3 z2o9p=_zbVBmB|`O^?p?dDJ6nG&%R`nL+L5@CYu~?oFo;`YBFGd8oq~s2xXeqSaG9q z&C8v&bQO~3f+#4ecLn-;8I?PkM_U!3@T4bmXeIc}(m&9o>{whXl%0g4w;7zw=c&m= zNE#_${GUFbCqDf#XVvd}hBMNd{UY+g+V(xQljjL4ho>%ZC2*yPyhlXdDWeAWz8Bkdap)mU|zLCgXv|R#cmYA9A(qF(rxkx$4@9 z{)p0%$!MpZPg#j!&h>t^$=MpdhgESC+H4VMibjXI$tgCgM7?(q30k9hze^;BJc2N{zY&<`wq8wO9R5^9_7jNB4fo}uQ+VL%%vvkdNrIrBPm`D# zJkl6|EJcmrt)Xm6$SYOl9ZLn*jPGMzb5Va3^B#*2k#RdhL8@e$HR6^Sk0My(Ik>XP z{*@4UGada6aH42&*c6;?&`71}snRDD7*hNM{Q&>z5H2xS?zUOyxl+uTO2q3l?Sm9V+i*iWXh z0ok@EIg=^@zyzPX#n6!8uCj2|=uwh5KKtT3c>LkJIIz0CqYza1Wa9D5!k0qHq@Ykk zsY0%nQ?AW1*QM{r(KrTAgNmMw+c@ph@dAsGV8}3|;OHiX1WU#oL4qhaGR`f!6NFGR zZa$Pe`UPdiPcU}#@^~2u#c7Q|GbvfghOERaz7u96bLNDZ05_C(Ek*Gbw2{sac&Xt-Z}~Vs{LEkFuA7gZvETO?_ny|= zFB;!7GBFPZf%~GbeJaoSZ_fx{iW0I0G@IyZ7rL$BnJt1_ZGtrzAM$;e!efRG--sm! zlAwL@TP0GC`DqhM?##efW~K!X!?a40v_3gviY0I9=jqI8*@A*S)g@AKqw)ft@fxd9 zd$DGE>`k{>F#|d=J8s7eIcT=nZ~GjuJq{&3R*a*%-4Ha(XByCR>ua{+C+pm}jYonKRq;iD-#8OE0Csy{#*i&qMt z8b|~w71wExOivRXaxO|Ro?$QNFMs1|hW#^%;`b0EyVjiTY))5wy{cZRs*hCFUx~l{ ztg5~?epXfeQB{4Msy;zg&uOgqgT+Yw%vIk>cdb1+`KDGrJjc2G53Xang0z@GVL# z7SEdL)0C2C46~lcH1%}W+!D;uh7^m5W~M~U%z{(TY&#u6IoV!7F*hZ}xQ(96X7*O+ z2f8!X7@y2roGdz%!U=VgR)~bMF`rlcI^*%5>`l^ewD$Jq?Hc#Rhhnrd2Wc!syrBQ% zFGBw@7o3r2$Ok!p`P=_+hNJEU?mex!Uj&u2s-2{&uLoW~_1n+ZJ}dB}al6kD5x)1= zzL_`u%?Idh1u!sSRqq zMOv7oUCx1`0Y_^1Kx~VBy zGeeS5n;IheYJHMOTOE;Ts8ZA5gR|2Y^EOHsBV^-oAgxhsgABPi^wQ9EaDWmZxws@$#{Pn zwS@8H%IVh#ci(!PORl_dZj1K{XR)a5c0AB@24L?6Yy9LNej7je&40L%`50cjhFViw z-4?zH!5Ms@xT%7nRaE0}9r_pO=#`!OUM0Gt^Q+&7lvd$P7^odF0o4eO7=ucq5dYTAO?7X^CKe&#B$Ld2 zBbunHJ{$2IPByupplcExr-= zlGdndDI%4iVu2S|f(r1#C@{SN_zzJ~ah!#qvTHN;8e{m8-}yFdmWGtzQ{013=Q~@C z$_~^ZBA7Br$kq6pGT0=&4H_7&=`y8XeXfpHsAbfght<}Qp|j-9hILD!@FrA5g;8i> zU~`0oIZ}-Xsc~GKo}ytZn`~uFk{}5Ye^#C6e>@(CXt*&VG{n#nLq~%4C7Cn3tQgvo zvLY!RN$HZ&lmsi56-ffKAt6#JGojs(%nDjY*=s$SaWo`uCWL%vxQSd7Eh#vVY*L>I zMuHy<=)#&xl}2VyV?Za#X&6Vlk;m}yn7hU(tyYe)9+Bvo8`VZy?C|hPNbY4KT~pS? zaVyYBz}lE6#)a}2&Z8S!qxUNx7VJH=e(x?g56Z4fKy^Oh$xnX_k9*4H3-`PA)&h^y z**8^yWH6=8slr|j6l+=$fm$myI)lPoVJW~E2L^86Sfv*WpQ=_#%PWIX$ex;OGB;qT z;@cnBxShu}?;tTD$4tXGu58@JUK9Lm3O&2}z>08l(Z&-J+ofViC1_w~#z%q>)gWkq zJCe~QBa?*8SXwo?x{W1`!j_0&L}-g4VO-|U5Np0kPFo7ns4Ysxb769tH8Y#7w;H8w zHz&d){?bVbE^419vtm5ZeRmbLPCp#Y=HtU5Q>Tb1>3eoNX4>G8f3_* z-N|c7)bhA7LxGZ0)g`H4BZN9;Q(@FW-Rx+a;mEX9hiZsS>bs20dOB&rYL>I7-J_EY z0ER$$ze#Oin(`tE$`KosAtqJG4LDg)m<_VXQ&kZI(-$_w$mmubiJKYtSz=VUP_VR- zL6(g*hk*J}uvCoDY%eBs_X@hhw!}glx#c)-f7J*1rXPA17hWC-{`5q@C%JthT?$$A zHEbogB-HO}-!3$d66la9S!zOUFB3M}C6XK5Dt}QEEG9$kD?q*Hx}H#vX)e(S7u?wlNjY-bDKx3o?yMv zqhZUDS;>gfNF0e#604-Kh~&o2tGK~@`f7&IIzr6OJb%Hrg^uXZj#Jkw$=#ddcXj0N zH~H?a?d|T=%Duxbqcvy8YCnMZ=sQ2b&wSOd0YP(8sE{1G0B^=g=Mcf`o@8BIbv2X>g>fAPTk@?9-YEYJG@I63BZR)ecdVDGffi*jBapR_!Yyo_>+gX;dZd z14Glc#b}Zx;znAS%7HFnItiVIs;&#nS(!M5kY+y-9hYCx8RZiVg2tta7#OK_-yWt4bMR3Imm6{ri&Qv+cfXlf`E#rDHH zX517RQjmUuJ-HqA-X=H`+6Lq~X0?kyegd4M9C+@!`6!1Ud5Cj`ozm7XF$-1Y_x{%_ zc=KPq7aO&@DrThR9EItKkRI72zj%EzOX?0eQG6y?M^J`GBwIq>QBFTbBjEuIWZGpuxyi3^cYVfvb6FJMRC_j(rC$r{;SXSoQ(@yD7KgGLWsX>c;upDxMh4q zewfCema14&8&t~XO;ZjZ+d>^(st$!LlHCbGr}!l(U8*pSeFpeIw7JeJHF)2HB$xXZ zye!^d;C=t=!)xx$6PJ^wYfcmrGoR-CUtjivOb=#E4`$2`=Tt{Cs#zXbG3z3ak?_YJ z%$wCtZ3v1Ffty{3O~^MEb&{G#paF$3bMTplpsHq^3X}rHVr65SYSr;zf(yeKD5|8p zc#F`k!80^f!{LyMXJ{Oy4}{km|0gI3u7@8boa>yKv2DsoSS8dds18zZE>&TSVIZc7 zeen_kb?eoR5y23UmUXY=x)k8!USA=>nzLY^%}le#oa>V^#tW&2zo$N{9~F0GyBP6{ z#f1tNW}7s@D>bAmkeY?6FRN~m$xf$j&>os;lYyqbE&wSDZ4oDz7FHd|5mhlIfh=*u z;#29oV|MPa2eYp67d4&6v|SmE08r=oSKt4$$K1)sDJmi6XWi#%S&?7Do-GL+&octl#hP z$_dmpCkl!A&HwN_y#uQ*GZ0=3ZSvzy>5*xP-x`qLGDg$zc7;NY;Q-YTOy`GTm9?xl zlxP+A1k(wEmNaOLYJe#FO+UqHh#LvVDu-pp?W3bri(QqOj!3B$iD#1;+ezN4<*~5A zM^c8BQ>TBmo1%_=2?Jw8EJ>R}efgS}0_|)ZDz1rZ_c>7vcKMH06$LmwRS66H31aB*utuwGxVJ)~>C}_gA`YL46qa_%Hv||Kz*g@vYe8 z+K9BSDo?wO9+$lH$A5+SbZG^@JSq5v-Wh%LBJRP2;_h=WNSo^C=FkNcS3tcoyCq$} zdLxILY5=u=YhY?xI|(1-3y%r?kKQGA2KTN-*o)`k4H|Jt~psq%p}Jz z1)Z3kon~d!epQ=iZ3Qoas9GF0(y^{a?5NoqW) zR5&#_m2QA^{M(9BbCsR7x3-G1EkvX7f)}I|>i?R~6Kb2R%??y0o%V3jh~P&SO$}~n zkmU(+xf3<)X2UyCm5%xKOtbdKe4dsSnNRbp@3bV6R~Q$P1nPSmR@90qHMo$Dx!;0d z3hCcRh_hSxcn5umZ2MoqGfMNdC*TrHh6kN0g8#X5q&rB1v83|p~p+`$_q+!rm zfm62ZHDG|X`lAlaHPRjLeL6`s^lTtKYixx*C)`(B9`gXzT8$+FTQ=v;Y?o8Xgd`S# zw6y^f<*vaMOe;@G)_kT_Mq(_ku86&yitk`@X|$#{fC&rgn!;kW!@;GrauEn5ftN}S zCuH|tK6bilB=+xv|NEoYviG&i3DY$v%7}U87knzdY7LGoLk3xZZ{x(g4D#tP!JQq2 zdJ;sXQYy|U<%Up=12Ti_eQko&FVC7CB=tzpB^>f40(-QcLyG!(I{3F=7JYEjg z_jP7c-%XHUz*J)>2TK65DpEnR#`TO4h!_(~b!vnkt#SpaY9LUn@84k1{QcNH=hrm+ zI7Q*f^2Gc>lh?la)5r$L1fDKsd#Eba9cN)VD^eAlBJnf0QgWQ6v$fVT$)IF>X3F-h z2V_~*{J!kgsIvkQVOC|#$|RQc=s`iCDF_*kttVySq3U!%4pf>joprst*0<;wE!$l{ zw2xWRRdqH}f*~OU{nj5r7#rBfkgvq(tct0fF&YA$CvL!z8!-35{a<*?L&X`=SU|V0 z#oCCOJ4^0jMkI==;9dN9m>&$8Ufwuf+1A~uvtipV>Dd}xA0Zo2#h=;W@bjL3?ZaA? z$-!zm{u7oHsB2D?5p(vo)BLgT{aOa&3^9Tk4q!fqvJ5Gzvsr|AH8Q#x%AL_-c^*ZG zX^s#fIT-G*Kd)LydX!6MEi!h9^@vf@)L{)+aSW4+QChJ%C`jz$tu5U`6(}{oz|tT^ z*))s%{!B`C>@;46P>_+6#7XS8 z`YglnJ(CcnQyVo#LUGHXR7~7}{aB@TCz&xa{S!m2kf%k)ytHJA#~P1TpiaW462t-< zJ;Z|=0=v^mXh#Q7WyD2RWRu8QztYucKnnQG;!cmzETDDjbz#8dEyN`TGup&3+_+wa z9|GFA?(?*)%2&Sq^SI@%vmr}NBxI5e2jMfwLWWtA1Q(zxq!!&8t-cH`*MI=oOi3o| z$C@c)c2aAtCO~2=!|j}O=t!)`8ps-GCn6-$_bv*TXJa;!^~ln#gRb(c+nH;|N6TAy ztlDAEjj_#mT^+Mq*Z3m&btprp$6Z8J;=-j*lBG0lBcWW1xdsCU6idh?nrp(uXH-}m zmbs?vx-mb0>E*op@NV|1F%y@q%|2BiFB2{vZgBo^%rwuKTpn@d{!{Efyv@ZAon!a@ z(=@)grFN1j?0gtDE(F(fG6N}SVrERPtl>tDZN-OX1O^F*pZ)BstMhm<;xB&AA0iz`AF6SfNI#t0Ac2#WNGL)@=F}Ty?5EylE2)p}%!oJgKkBD%`^KRpg=U&I` zBwhb`T2^JeIpiPym%qRl{n6J3!^Ozr1vCcIG}KbcG8hi=qiEeO%hCW8V6fNkYN-Y= zxec_k49***0E6&(cRibLdDV-tiQ;qVc12pq%NnV?CACV{fs;c<Z)u>=c1m4+Z*g9|Q_=rT%hB*z5Q22wnUD~J$NGLGg= zeU8-%)T@`DG$CzQw|bh?#{io4ZZ{ZG4XQv{4e0X<1-4^+DkW8q%gy2` zoNekc1wXb@)#*!w$37`ja+hL_XPT!JRS51C6Y7|l#8XVuCR$O44ygonAYK{+X8Gp% z0i9J*u^}NFrPL;D!RF!nvZyg|w4Oy%P|nC6sLwm9dO_`dfICBT_973M@R+E7-cLm z@uaUC&DGQ^VqqZtZn{3*vxv&+SE`C)#Za^CE~>9BWGn>(aoi$D%rvD^Vb^bP-tREhF1C!seCXgO_fJXKptpn2|>YGrZ04yMu5&J|gEf>^``{iZQYZ zs7VOojoYN@wDj)}A!Bzm3~2B2MT(FlhS$C2HGKZpe|G;3c%pK`bj^t~V&46qf0qjn zUPk4Z>9(vO#uk`Kf!`i>6pRAZum`3vP`@P-EnryGOsav<%YMe?i_`e#Sg?jlJYX55 z71jz)1ruMK8gUkZ)GiDXYeI_Sq#9Cs%b8@t&4YO?{ul;kNy4RSiw)Z!tkvc~Py~l? zq_7crXAF>P4L2`Z2)$j%x{QFehK+=QbT#8zMZkMXlr*;Ik;crbTQ1ZxW!U^80;lW@ zaWK`cTV<~ruT;0*PNrNbH>tJ3VL4z@4hW~0#(-(MAOEze$5k;{+JbI6m|N`r`hnP~ zFDHT~O0_BI42G9n2paDkpdtMq0@XCH?IEAnZ1 zobu0p_+7YbA+i%`YOl39IM#{UfCMWGMA#&uxSv~!`~1kIE3Iz{h@utLc7TzPg}0dA zR?fyoS~N)Fscgoi%6R3*y?oBjJ$3g{b&M!63VXV2$S7MfZ|p0P;M(=Tkp?{y#&%q7 z$NdbjC^!(>8IoQ#^JrOWG}kOmj*fOR$70(6xYIh$J3~RJv|aaoQ%XKm+{8pPv^J|m zFAD!LOTywYr!^|ZNY^ZBjdf`16?x1ou-p56+o4*X0a|OCT2*KWQfvS}v z9#raRAGd7bV4%fwM41|J|E7iE+jA`6O%zL3= zMab1sAxwPJ^G2P7bZL)|!O-j3VCwdI&Y`;YT~)O4VNZR<5lR^^Ck0Ykf!G>qNKI-M z!Il_-Z9AvEOmQBrRO8NmUWMIqv$;C2!lYcZSNZnwsjn`bA(#;Ui zKssfo%HY5^X8!`TkJnPxvsLTeOe``8!ReBvU<}@Q9{$K<{Otet`W9w8X%O#?AoqUo zk!#*J4(lMDfXroc0LqOxwfn`f$qY6hf@E^s0}+-3B*VhUlUXIRp2W0^5h{ssJac<* zp-}{OvYgj!-`9zussBEk9FbD7pyc$VIPOZX&?j?+^Bx{@!X+27ArpiQLngLy^VI$C zD>ZN}QfE)rAAn>{VuTUlR_l3b<~U_s;LM7*b?Q^5Ed!?R*W&-RC{v#W7u1q^R58qb z%G75(d~^!umr=MrL=G5q5p_WV-^*~VS7@mXW%Ovt#-kNU3F*-)4^jO|4jum7PB5@V zkl_H?+z6lH05OKDQhxjgex55A`dN#gsGKxibK;Dc+n#+h!|@&W^%KkKGl(}Z?yCEfpOiQ+ybzD=+l12^L)De-T^kHc6CtO`>L&mF8=_}oiH%Ln4P(9)=%ker?urw zuSH)8@i1jQF~ol@0bNO#f)L}McD`sps_@;`;RX^IH8lxt$-0z5J;_PD2`fDctbJ%% zQOrwz`KSI@Z=avAoUpXV8$oV;#@TD%SwnV6r-V-Zyq6xf8`O@KrR9o-x;e**-6$~G zf)RGv2r&{6-eFR4b~xpQr!Piwwy4VlHqE@z56fZ%?n*B6jN~#mn|CD1NBCvl8T6=H{gxvp_#HO+(vXMpV_=k=#Q z#Aok(gt6%lj6N&|O9oNh#_|aWi=`f!-4f#L+DG;2hwO?YDFf-90O?#AT&YN=9y|49 zdk#BsWV;o~+=Ke{u?_KeYYo5k%OB|9=ZVS*(={i~i21BHzl!nJ2)sk*c~ftuc~of{ z;4)RxQHmcVwDJZ)0x3{-=#hM(blfQVb=QCrvK*GudN(BV(SQjXlBB;R{lR=hYCOYa zwT{QkWfVP;TNcN;pt-W_Td9RYJ%@-WxZjG$Q9Fla8|m#ON@^$ zy4C);*)j>V~$0^A#wOEx+tz4vj-aVw*zk(l-J4d3}iWW(b;>I*Bv-?Y_oZFCKkj$I?H(8;&O zl&qh4fWaNzbl7F@RsFX?dG^j_Hqu3cK#lR1kppJ!RS9I`*feuoL`oO0Y6yyLe~7J( zg#JaCNK+8D!mZLf^ejf}@7ikPiHHE>82Bs_h(Y^BnfruG#b)DrGZMI_w(X41-1!)N z$pZs>4EIKJEugu9a`nN7cS1OLCd|&pM3QW22d|9IR}7APv)HM{cu1$7Z0a$UlFpR$ z%9O!J&Xep;qY1e)e`hU@YeYVM38trTjr799yg{_=F0>`1ZE zLfWr}D@2Ps$+Q_%f@dr_Ln%4b5xY8O*N@rrwd-W@nX94jLS5^+zekHLo|e%PKMqeB zm5$T(hD+yaR4L=0*ZR(iz@~KFMxia(btAl53SDK~6@ig0NMp^YS`wpVX2FXg0-K{8 z6Jhpx442acbxW_ff|(fX#9~T=CPA^dEKq!l0Y41Ez3FW68fYq%TZSrIQgy1B(qs2$ zNad+xU)UVO#t4jwaYGKO3P}?NqXA#?C*E+4ckdeIgi(Rt2y)YHr?E*A9{P%DQj7B1 zFMAoBjV{g}T@&iAg%3tB#qpd}>qW7n1v9tz^e&b}Ld@B2-gUdG!o$-|?w_3IzWJH8 zwvq^3sJ3`demCzgp27c`KZp0_w{zgrPFzgAHZ%ga8nD-3XlQf+b#x8Nk>&J6zH z@yfBC9vdM)D;(S{96bYC;eD7wv2%|-%1QR;&8K1Vtj`T3mySpt-3uAI?QA+@7JoM# zP`p<@{S7Z!f5V=LoG@K;;*6Mo`*(kV-~E*jqQ)+Kq=4o*{%8_XR?CW4z2PN%@cj=r z5I-zdL*-a-bzpD&hzi;M>iHBv6itF#KvCSVixX>OxP)Sl{SwYzsm1%n+AP6wHk&dq zj*681auk(I2-6MON8$TeLfL@@j4YHSjC+eh3VEb*%2^uZ@lQ)|582VzLG}hJJPxJ8 zQAET=Q3@*jG47*g)ubAa_pt|S?u5czyQ`9ssn}rN_~KFJvfJRukB?n78VP>?MI1no z)59r=agD(uq7B)dh~k~%tJT-r%nVX3rFX`U_<+cXtKCE`iAy&(424ATian}G_bbxf zeD!{az-WM)=;MzFq6pmn{M&frH@$xCH9rwKY1|DbUA*`F!nNiAJ7IP|!5)sRtv`Kh??%A_21PFyfKFO<$41KJ2jW!A)$dxb}* zTg=^niOcw@^RMQ?{PapV1;Cyg^H6m*RzNZ+hF!nGgXOvAQd+&jt2D)0+t*+JU4Q1@ zknQ|va5ZbrTuWwJFwq=pz2htTiaypEkr1xMDg0h(xzDA1u*}x#2nC9zJN>-$5@*e% z(KS{*4;E+eSW@xru(d{vC?nfSqKHjq)^Ydg%SdZ+wD?4n?l1|QR7?Gek7>m|RtRyH z5OGfT<0b z$kZyK1iDXyLDN`70wirqJg_sGrBkR1)`VhhVjz{^r`=5Q25A8}@*^hsfGs=YTzc3X zpF{^y-#ael6vA^5{-j&T?Uhn2uy9j zC?OjuukTTFA7G{-YoODHLMkCc!Ye#n?r^BXz|y-WWv^Bi9>aDrXX+Aip9W4CU}#JB zXG6RiR72XfiEPP|LX|nRG>hk`AVm)^PTbZ!4h7@Zv1f)rz0swrnRiU(`u9~r`oFr6 z>6N8WReb7eUxc;SasBWKgmQ0G`OteF;Ap>x$+R{wVnu6J_`rXBKgl?wIui|U69Y-x zwA5`;NM@dL(~`^=!^2xBwy3S{R|Mw%G%gyN?Uw!j$aVDLR#aZT&@)Uk~d}6rlH!0PS#p0{`e!x_N@8crMb167y zD=O-2A2{@mQkAh7hN4XNzLmg}t7xQ-Xpgso>e%9sQ(rPu!w1U&M{%5Aa4B~u1shAZ z%&T(Al`Hzpit@~Kk7u%n^YCvEKaF~)nNQhPRty;>WJNYlqh*_JW(Aok*`JJQV(TOn z|5XLu@?c8Lm1QoPK2Zkykjxjmrlk+YN^7t98ek9eU?9Ny+H8Fyg0;d6UwQYnA9F2o z0(H%aLt=_~(La;JV(>S3tCC!vBDqIXniz%S-Xl{f$pKIVrBZEx*bvtcQX>VE3nm9~ z5#gDpRMSPxXXg#h%VLuCAr4stw$nLiNIg^4l^Rs4M--i;#jf<#TWYqUD{(S4N)|z-=A*2lFm+c-Y~YVk z%ap)pp@ep_C3fHXn7L0#Vu*5PJe9XN4Xejd5}auot9`bEAwy;oFm|2(t(Ep_$#6Us zLTwP$eKl(;2@5pUkYlG*m0$gLzk&DHdD(Kpg6&Sa%%{c57-?iFnpOZaD=?Eh%zlPI zXt<`JS-oI#VYJYfQ!8md-tBAJRA_i^_dlEN7s~2 z!d&LVqn-MUsT;CRm0u`#_%FLJ;iJWE4Vn>LxBXJavAgGo(SQkq!eRJuHQKQ=>;vc;SP^EuHNXKcH0GaqCwq?3NqsS7Z8=gb_G5o-Pa+02RD5Xq~yV>F}pY zuFh8+nu$Pr)Fmxt(h`AC&Z6Cs$6-LY%Jb|0`F+=Zz_rQ=(={i~huKP!(iZBcgmm028cwA}jz**=Oqkvq%F0q! zmdY87Q3kd|MaW$mp+;fmQ_jz~RxT;HGK4%T6t}(tAx8Y#TSAP*nX09svfzomPo8ef zOzl4E?UgE(26a^e6%fcOg$-(k-SAPMR6~m6NHYpj(Ap%D=zdBmd{8a{P8Keg8|+mZ z6q+!pQg(~c3ap&90`F>rvWao?(XdH;=v=kNz*Pd%W(WEesx>IARNY?mWypmYr)o&6 zR5e(--gD*AU4HEse!IW%CnzT^*zTmui(dIme07ykQU_Uj4P=K0^O)gy{AybT%FPJl zWWxLu%+C`1yB0KMktu1x4MR+pAA&0q_8;D2YknqF1yf;OrcCmTNuE(u34$_6i8ve(8liw-E{B|+sSLT~s(joVkJ4|EpgHyQZ#(bza#|O(>_+WV}`HHha zc&IqTo*PqoL*XqKi%ov}@EOfLSYGjn!0FNSYL6YZ9BNu-%o<&jE&+h$PP%r{)cM&M z7-iRH$Eboa{KVh)ELOXKCLyA` zFJilhZlv0yQoEO$Y=YJsQ**|&-x$I0+|8SpIu$0gtzk}Ll~NOCF5$2oG+8xejA~dxlxkuT#SrXU%en{aQJh!i>X@r* z6p2&iKno^XG^k5Oi>6;!vz+p#LlW;oy@CyV4TU$%y`j{AxD=XDvr3dLh3CmP&^?e$Vz zx+T-48zoTm)pAoMJL=vEO)PF}(8_XB$V)hy#Y3VYGnqEzcmM6DuXBpE9`kuxR%AFH zaOT$S<9E{T-%>@wx#c-(X<({AvEA}F8)2|70RiqDxcIJF_owY*r)Nzc;=@jl=M0iX z{KR77U6p~!c-wRTJc*3B;I{cdejBz^TA=QN=O(+)uHWFuXRR(O8&YtaJ>+(K$Vkf0 z z!z_K5@0DcMqicZqxeS*GE;GjtTixCWv349YyrThw2t{6Uu(z&)?Guy}rfW`IOw74E z&M?{z~g}lla#m+@3ImrQ-e6`tZYwX z5o~8eUVGE?`2GF!s3;;zYAXgw6&YhgfLNM1j3~1zjT{c4DsvXg*KcK~c-84o=f6Jo z6YP~6A?1;h+$M}o!Iqh0#bZVA7;-XlpJDNAnIeKj3;(wgTs5LlQ`C7_@jR<46RO~; zjxXF4ht);74~LR#$wr``YL1=b8&tsqbJf;BU+m`<4t&Ou4ja@&qp{^;K}(0F)KCZk zfw>RJl_$i<2-=YRCE`PAg4o}^Vu))QYt{&fag=4!1lomj2BJH6lr}K10wmy!VoC|s zhrN&WRFc70@NogYO$O( zGqz08c+LjerH`ntGIO;V!Bv%ci#Ipdigu)*DwGHhxsC31ECv#dc@a-r$M!6Al~SX> zig4iCE~0*OPXxA;IhV`LX0NVR3tw*AIpl%eGdOP$6P}6pLcMyC%jsI<4FYrT=u+QQ zBg}3aGP`XU7=KmCE*ES*nlEll%f+c%;hJQYw=PXq_l-aAUpiv@RQH}fVL5@i=9-Kc z5qTr<_X+djKULKqT(jR>fPV{ox~jhWiFD|dZ+Iy-3G9uihK!CRfuagEBF0M?zX&#W z!Dvaq&f-g)fTRMxy;A6Z!K4CRSBKs}Y7C`!T+%)M!SDSB?b|lkuFw`mzSmIgbeL96 z(MCj=dyDuG=`Y1){>A-&JGPJo!Z!njOUN~6&u_3JQ_dulPN40+AF*i)S~2XM>ngFR zR#XJK@NLu00`MrybU3h5uM7nyIEFF$`3T>e%59`U+oKChIP@8p{6+_v39z%rYp5XW zVhl^ImsB;ll3HgdH3R~;Wm=|@owwVQd7cGjvMd7Rw!6Cm27IO=?#W9p3~J{}cGNmF z8Z&NaNDGe}Cz#v?=RArj9ClhP(jK=rLgn>&fqEZ{;KZhk@R~1v`RYlZjC_(HE+0=_ z^JRbZv-!w-A7VZ$(8NF%Ij5Y5eBlagiHKWGH1su7|NT_#hwdIF_cTr?z-O&_+Yui zD!E);q$QdtDy@awdb2@LN;OTjnfhG)=uG$5${8-_VE5&lJRU3%T0Hj?(NSLmJ_CH zuIf$_5exh%@Fw8Jz}Jb$i+lFl0=@(IjVIHgY>@K(@BDL&PmPhm0BSVGIoLgbgH>ou z<0um!0(Y$oD|5%UUV? zT{Zk9fhWcoLmpw$p)EQc@JQSy%;P>|TdeqC9SM(AI~~NK(`91d8$y9_;Dst02~L%D zmEPC2RT63+ zgmlz+md-Zq{ND3tog!HSJ@xe{q?6UTMhJ??3t$94u(+`$*_)G0iiI(x8s63E9DFp* zVl`xdv$vk&bzl2gEBEt>iol6Wdwe4D@zyn;^QKSbO<(sJ$S8wh7IcjawV7 zSS;R_L?o0f==|2YqA~{T_eo?p2s#ea6uDd{68lhU(kza;LAR}YqBKOLbr{tBho{-k zM^tJle8Pp|&w;Q(r|V z5WOWyW_jzM`a*`IC12%}kdvrut|cbs)xbwp^&V9%fS&@syl20^4*U?%p9}mrr&zO1Bgdx?BqQ<+{bG`&Okw|b6wH0dK6GRx! zFAHVlg=z&Zlcwc(M0eY&E!7nlAb#U&We|V&bgEdd9F_&c&`v7SOLN5eg`#O(%|1d|Y*ni14=W zeiLStbti!O?`$Y@IF*JaV`dX&)B*DW(Y$mZe}x z5ipzUn1X|tT-m-qmc*$q>$r>_5quWkqB^*|WxRAwzWi_e`%_;19RA36zVX<8J#M)U zy5@SzwHPsX01vjlKNNp!L_}T*JVRCgf5$Z>B5xOwcZ`8y`e|>oU#(e zD3`Qn2#LU{9Y~X_D&Vc)#u>$J zTY-&vO`1xvV><>;-_&VoqM{5hhZ|yui)$ZiB(1v?uVi~A zloga7N{?E**tJbHQeVK5-1p(fx*)*yl25i2YY{QP-vR#2)%I1@cc|*CRP|Li-E`Bj z``-4fHQN=`xv)w~Lhxy}TZH(89&uUQnm%5tE#d+`F#Jj>3+UK)TM!fC9K7!i^7?Nh z9&6yp4Vk-?N)t-8JUlzYr0TDmql8SJi65|6Z3OmTU2LsCS9Hu=lV;^m8wO6LSLxJ8 z7Z_%`qzqea)csctBzF5MR7Xit!6r>>z2Vv?5?D`QErp5W9%}JDiA_c>!OVK@E zgRy&pd#zH#mHBvupbOw&5@QBrd0cC2V69$oawzN37DJO>-NAM$&+w8bop?wLewV|r}2=^2> z<5W62z$5JXO@6<)gGL-RHL=gB!&=s z(ow}%k$@HjUD~?hrjKoLUbye*EEi{6RB?gTt{Vl4SD2S+>%O(lJ z(iy#1Cv|r^DoPnXI%E96A`U0m#w$whSh=o*nk{)Uy5^}se6Qsy9|oS$`hI8pt(Gm| zQ-J>_BKHBG0lf3c!p(2}&M)b=B8`FVjn1qkK(;$yVKh{ry~r@@UA)bt)qc?g3EUzI zTG{B64sV-{@b0w8sNAA}n)iY=4r4-Ea3XCtR2UkM&Uff?orN|fLINe0ii8Ktb6lx5 zW82c8G+;PzL$0_XdsW5-x5Z_@g#q$9-ek!rDWV!UP|L2j%?fC$aM?#x#S4et@=!G- zkHNQX_f%J6ila}tSZ;9HZ&GN+Ww*tF9|xs~#LT1ta~YR#o)`z&e2U`I`%knOdz8Tt+k55vdACMEnk*z=J~sHIG- z;QDxCL_%!Ek_?6k&wu5!)}Q-{h+gFb=W0|Ue%`*+!gWJHo}_%-buyc%Dl`%I!%;87LbU6T@-U$peBMB!+tqrC!1qLo4Ah}MKKaq5s7&A z{035aW}z-L0*qx&A`2h+fQq3hsp8Z;(UgHySXDei*$U(XL$1QT)sPJqgjsWieQzlU zsmkYFN`+xV3Q}41$9oT#+$IwpuDrV*L#~5hGbr3&+)kw-PAD_^3JO|yON!&MVgnoR zNA5$WNKqMb*F(IefxWa(L_i&ZP17i(8iIIpuycc{YbwQmwB34*q!pndO#=)rr7m1g zODFT28BBY8;?=6v{PypE>r=8qy2>i(lBz!8iN1w^eB%>s;~M1?qHDh9n?8r%{M8S2 z4HQ+})?kH|w_&OF+Ibm#{StLvrq%B*p|f;s&gca*75qrMOUu^rzOu{?M$J)!tgoh8 zt4|mno$v6B@m}5g<+C=w=bG_Ad6wJLy}%VLp@o6uWa4ooIE^Y&yDg+t)B(auNZMTuL;D$&F*3Gh{*%}H`~o@2wTWNQ%{>KrhX-+w?TTOO)^2ZPhqEE$SA{0 zdGC3E;G!oMWlHaOBT2l7e$~f|}@0Uey229(9G}DY!=Iz7TcSA0_P1*xqHD(Q{ArVIs??I`STs^A7#Ajrx zY}gV}CHIz##TFAkKvkH#S}7@3JVMx~_gM=~!N+A$4KW1Ga}h9yf+@mpZlK}0I`COA zZ0mb3O2gi(s^NPD>GYUohLM%I3$HMl1>`WoZrLc~^vyf#+j$c5B!lW64*~*kD**v{ z^OH3o+VY9gHAP-6X63dav&PnJ@Ywb}x2>nY{4828$Nb#VZmq+TT}xFt{I)Jmvhrdr zdN3ZY=ZL9g?o%F~o#KwsUi@V>T(|Us73JpiaD|Y#(2P`aj3`cpNApw7*gsqr5szMGifE%)ZtAx2rs5Idt2j|!;LZ;*5aUVZ`tnoOvib{(>x?SQD@I^3$bU(us5y3YB zjVZU`;90QwpawT~jp#O~SHvmFeo1zv2y<-m`yc(%f>O%xTuR>6?L2jeB35tL5?X0Buhk>6uol(i~Scr>*TD;-PaU@}F#qQ(RT;~eTe z&wKeZ``dXkqCDB4x=#oK(kGt?T{D~H*u>zLbc^H;9ZK(H(g9nPfYjCBm-cI_Gh1*Y zX(pnsuYl-Eu=uU2yUnr0T&Wo2IdkfetBY{&nc-py z6GN$n-_SpLUeX_$4 z_N<9Pr6+<$1O^#+k4BVNj9E|v)J9BIUH`(5wOn1-JYm%Li6XvFP2|sYlIOncuJA|u zO{-iWzH6a{O;pxAocdKVMYjcYhGIA4RcG#EXcGj_ zFfB=wSbL;MxI7&*tNH?C!}qyMd3d(Xw5he!-5%UWtCMGq?CiM_hi4kvYJvZWfwaKm@cOYDVLqCiLOh|*C zx&gb@2Is3CCQ*UZB^W9VDb=crf{jLCU@KBn=bbJjk&tTH7d%}$yM%CouY_un(zZzL ztYyG7w@JY2A{$y*6afXS4+xBk27VwFd>TXj+L4J~v3UqI*Np*q2YZJ|ZP3KDe@PPF z_(wm7v$vjFd7e*Jwsao6S*mURuVqKAVk38wj=^IPwI_Mp21 z`T}@O9)A-emryvpF=6lG+>oQYHaL3LHpSU7eqj4UO^kt!F>G&!#eI^{!~{fzY6ltL z{FlCbwS(6|o}{jMN@Ctlv=Q?y-}e^ML9&3%h&O98M<(-N%v26HN;<)n3Y43MYGgxX z`_xkt9d%Yo3btaw53}Ye2!d3y@|$;-U;F4KUUTX>jI%I+@+x6gW|5WFaA~^LW1!Ui z6eiV(i}@Ch%uiAFKtanDXKS~O$AmgfKIL+?*-;82i)yq5DSLj*o*y&Q0p&`{VR-!% zVXkSMp!-pWs~`_+hv&p32LE?1jPOj4Z|*_2D@S@Ky4#DKLJ43#$= z6&aN`F>pEbQnE*$q*&Y~ zp9Fpwmw;M9)njKR=|RzfcI9Z^d!}rV@Vc*hP0yC=BTtrqN>$&bs?S%|=cwxU$KU=+ zp9@D-U-wj?FrO&7^^ViL?oBUU8UNyx^wI<~DT1ykdolxB_cuH6NUn=n&MT$}2?e$y zP&izJ#+9j%Tud;B26JRc4ioanz&)zG?)kzUx9pLnj>Fl2OGlejKB068`?LOvrAjEK z`oQcQ@7aF_A3nUfGYJc^;Ny~?;Gi1vaCwSDKO&F8yvym$ig15*HW>cxb1K|do}+Ji zQ#D+uw%5i;3vk74Mja!ZFSl_Dj})iknUy+7S0wc2x2Jb=vD}CzPuT0oFG`^3T*%wiuEQU@<+$U?sNFqIO)Q)k3MO8vwpz@!;k*l@DE38Z5R?AvlvZKJ zmZH+y#_K9ChctZ)nNGn-$~C3JQpN0s!W+^UztZm<3KG&pVbhKP%Rn^0frR}$<$=Q; z6ieX}CPm8SgAER+L-H!6C@n>0afQVfc44S2En$*m{Tji>Nk*#h{1wF(K76kLt$8>@ z<#bI2aq(`rSWjx*(yV=+#FDI#r(H)mVcZQjntbQ?eI=XQ>{AW9*%-KfV@wv|xX2v%o%lc*O7gjwKAAAMrV?q7V9e`Z4oUvyivOR6+{YD;H1aO39nukTRaV#-|>kzk}8xvVcK zj6yAAQ&7LWADZ9J)DJ@3Aky5pF3+3rf!Uo*T*hfLML3=(#{jnskJwCxcyE}@Q|4uY z^OmFekp1b9!}*ZPCpZlTi+aQS(l9KTjxwrg(hPm|pn2LGxQi;u1U4QEg+O;c5(rio zi=U;75urGnQf{ZXogwA9jo@j3CR@^tYSz|O84P=}C9kJEL0$7yA-*TB0OYcK+UsA; zKmCD!AHKAMo#NXgHVJVuN#JNuVPU{TF)p|R=+5b(xoFTf)FwS_)`Flz-}BCd0aX>s z!S0apmP3*W%Z$9cvk^CqCp5L5V+M{j+HfzR25+MO*;Y;KbD!}@ewwrC5m|(M6;{}5 zJ%)>Jn_JC(xHU4M)RIK!C@}LGQ$Jv6icSpTvK;vl2G7_O4RPI`8*#;Luwil(X@dVe z1|*P}=NhVZ_ZdkQ7SW1xmea-^t#t@2fg914!WE}YJVW~(Pzdj#Y5mYU4rN7uwB_(1 z;ljPQbFek3TWzHBmc8kSQ{!1|DJ+u0Vku7i1f%8A**3Qg_oD%`#l#Kxox{7hd2q*@t^0eJ89E?&NIKe<{mQ!@AclYL-E@I15+Fjy1d+j_1Bz+*=a zHw{KYbu^$l%0kdwECiVf%Z~*V#x1_Z6;!zIT0|&z64DDLYJ}>{C>E^sXkx(#W;%z$ zbq>#l319fNukOD&*F&D9u6ZhA-cGy`bLVqzK}=Znc5?SL=Hgqy2uZ6bKi*a^dUQ$2>M9{}2^F1;)zBb+y!FGdWl>tsRH zLUP1X+O(iXDK`as zWqkn!wSwJW7@8d!4d7_jq(P2028a><%K!5xL)^tvBsWrxq?0ON`ZcfNAN-AoZntCZ zMDSUt`lB_Ui;3HpjFEKeDYjSLmqH90cM6Ft@PuV!MxcLIYkI@3onh$T?{Ge1T4Zcw z1!p%VA^c?3W>%=OtT@GSdMwH)wdC`a}dwOcJo9JZJOe{ND5; zmi<_SwypnGtu+PCmqE#XaDE5BKfjGc9F;C;{T~8GQpDPH@!B*bA-J^;gn%Uf=Hw+* zTC|okt9{5T!@D25t8-0PqW}^Ec33q=Onpuw)q|PJiUA{~viV4S_K#CEkwsAm5lDnK zlW=)%lPp$Y7i$;{Q(pM;yLijDyzcRz>~Y9R(={jBodl4imaqGcFA86L2#Qc^(a`|e z7(-DZrNa~+dsJbMsufsE2vv!JuB#zhl_p=f6O6ApLb@L*S$RPlhDIZ3X>qwF&-OINR; zx=m(YK;=_rRd6xYq;{n@%$;RgrOc}C%1ME|TE`CBlBt22h0u3n05oSE(oB#4x@-N@>1H9m=$Vyu3 z)7sC|a!vBauY4ujr^YlXt0H(K_)&spwn>#2jp$vEGB=eTtK^R(+yt)Sc?tgE5!J-v zUH`=EyrHOq3%8G)O(K->*!&dx<&aylLq7BLJ)F%J#dQN;lRvgYR#I*rK9kS8^^c=y zq=~5O*Ga3+QTK(Mxe)wcZ&RdYz=UgrkZ@q}&Bur*vyMd8TsYz9EqNjPvuAR7w%I}r zx&z{&!&58uk&ffsXQFA2Na@F0aCd@C#H@fB?>ym!*ic~&C8b@6t!Oz95 zcb?`4{`uSafuH==r!2L)-LI3RYfd~81Mo+`>rI@!b%&q$p;|OvH3XQ?=8F7ovbI!vcx8VlhQbyKS@m&J8Z{fz+str{f0c zJ%@{V$)}(DQ~b{EuhQ1G0v!4=yVV9tNNpAq2=jK`n6&cTlI*kV#;9~%Lr#T#H*Q+0 z>U0ZrAy+^=XdetMeB{XD^djMvvsAvKvY}s`etwiX19gE`_LGP z?XL@6^NHfNJj_y9>$;q(#$vm^=>&hV`9CAQPmN)j7VaHM^UlVEuK z_eWpAXWaB@>=p33TmKxt^ypt*F`7i-aIS-kuBIt=1Z|H@jgk( z$W)!*6!79nO)%7!5f)Y0osamH%SU)k5N;u}j!(P!AyVsV03TJjFxl$8!CgdVN!KX2 zX|ft}U_dj4Hrk?bu~fnAENniuW*Me|L{k&(-hes~G5&3G7xYTgQ9}8hD0e*T94~sR z(yBZCJ$bt3hA?6PzWD1ui~YSB|MJIviOP2$xB%o88jTw}^9;14*s!>v4a~iDP@QyN z{CZwA1=M9W>J@v3U0Xo%_*>dKSOBXvSOExSkx>>2X%M=qo*b`_$8nn+q92o{H{^AsYY^YCmt@V>N; zDSV1*8OpgS9OeV^%96I=<7E*Ei(!`=v`$?Yg)+twwBZwt=o&3mSB2+XcV7*GAExJA zCKBHD`L~UN8yF<-F^iox$EB&|!|#2F7ro-CMPY=Kr&w-O8E*{uOF!~fzWxnA#KqcL zsb6>#XSmjT(`4Hl@QF~3V(_U(#a?YejKeE-rwA@~I{Rs)Dw#}3pbxy=HP8Otns)2lOC#!l5uUZuj zq-qs+jMw5U@4s?ulPazVxf9-V@pfK(<}m~~Dl+arJj1*SHP6=Y$psrCH@HhfG>ksB*4(CJeI=j~k$Ov#y4w;k#ZXO=6 znMUX!p2TCgTx_tJV8aE@d=d?q2-&Jo`h*z?l{W#x z6=iG-2Btz#&ga`q{4m@Xag1!)iK$T)Diy52Q8i!?D=QvV8RzHQl$!P`i~VxMK^|ON zp~82d+iQm|rPeWHQFzVgNl+y9L)3t{XsXDpX5y93K+_bo3W!^~8$JoyNgcCR!5(T8 zN6^+gQqNdUJMYViB!T!YNiEZ(t6|C4NuG8aT{oE=&N)1sqo;KC(uepoiRQk`w^Go(xwcxU=hx+ozA%+zw3KhvkrgFWF328FiH~ z{vVs~@Lw;!fX_I658G+CEJ$KJh4-CU3_&plZXPbk_I~mF$0+@6ym0G0sSUhN56sRm zi{yC0eQxn8Ov+Q#qwcU8GPEUDJkCoD#D(+4Hd}VihAmoIQN(tH-F(cnvYZ*t@M<_e z-Q=TtH?vxOT9{TDzwziZ!&6i3*(ctQIEF+3CsdD&Wr5F)#EaI)|encb1xYg(+0Z0?P2_X>0yr=BRQza)%*W6G>%$?7= ziQoS9d)Ds|;Wlpr`>q?94y0@IXv8duqN_Dhh_QR|-3UA}iqE1VSyuyXt+c1g5TJG8 zV2hi#4?8bt<%Po6e^4bE;s2Z||8vjS>eRA6OZo&!waele9cF%G6u z21hjKSUFRrVHLiP!52_t5FD=Zy!e$*EeaD(czl1Jmd7Wf@qpyG`c^GN#slS+4GWwl zE1`?|#zS>cyRF{Il)%77z6fVS*wEK2c z1Y$N3*el2U#|xjr3pXF-CEJfRs7xJzznoU>H7r=_*0!@bA2_;=-`&5HSDbo~nGBfu z6bP|zlWMTk!s5vV0XWOkdj(_)6Ym&+u zI$zu~S4k*QB8o_G(c46E`)SvJX^xU;fYc9VsY#$Ozo9rK_*5C}E%c2n87IwJH(qRV zP{mlgW3Fvy6MHb_)t~p0WADLrlM|?GPJCc*TfX9teKvMY7`YH;ITa8Ceq%Kb!Ilwm zS)zXx1(#}ce?;GFv?CN^n}W>^>7gOpv(V5O#6UhzsCYdWbS`KSR{!}lVSbd7PcvNE zo`8i}nepJ2Q|!+&W<|nbo|3zS?x+yY1cb2(JD4d~E zVc%^8&Y9SDsWuug()onS)8X6))sUlVxq5U^2(|H}3c0r&R%0HUZE;v+?B=5;)9JX3 zrSpf|t;*a%oUDd?nF<>5=;IxRPB*Y^TdS9q|}%A(=ut5C0w~ z$vosn`0SWC7gbD}@F)NBmowTprhM&nl)4gb-{)z0e3GP=FaP?_2r*B6L?%QXVL(&C z4_2YrUb+y#)pKJ#Ik`plq)(LjO;BzJL&pym(QF#gfnI4%wq<3xa&fz-bTiINL}`V?V$k8B3f(t+X#W-x<9OEgCAQKzrw3E+ z+TP=l@>FEd4adw}&0o{jHH*G^gu`;--d>6L&TR0uKl{bUT>JHvlcsBKC?w|gXP)DmzUzzlr+@oD*1;o- zjn*I{80YYF(Cu_Zu?R5S^~}#0w3Q+$4zdGK3*{JGUDUKbD1Mr?xrgMy23M0Pn26A{ zR_`hbV~6^nDAWsOE_ijM*}~W1RoI_zgyZ5{u#pv{iKB`|3{&fvmV*$48&MXghEwU3 zTZTss?@ZHPIczcW0-lf(ZJ~%67F}bhUus^=_FlQO927wTT2Z#s8Ad|UORt8*YCz>p zGdPu?4DFKjaAi7VUPWIW2CsoQZZ_-VbqM#?RYG5p&UPHkIx%G=AFCg`&;(^Cv!%sU zT-bJ-3kBM26@N`yJGBjK=P?R*KIdlM_*C*WEXhgpQ{U+Fwr~Fe{>v}DpYw}l*fCzu;hu+HB zTXr5_Kc9r0G+lE;88HA~`Hi2$&;8V|aq0YF%Lot3MF1+rZ`noc%CaG&!N;h+WpJ>7 z^0En(3RNAX*7u%^LQ+qN5`38&TWZ%onOkJs&Bt?vB}wblN9~los82MnY==Ll^cI8X z%*ZpX5?owC?Uftsm19zIY$bEf4W~@XG#W7NPAhYtvSD&OhJz{)+C5$=T*xA-+jc7K z7Gs9CWMEz7kAS>N$P(AMn<}+bJ`nrmWy0QUOzD@%`XX6Wzl3`$^Cal(_M!E7b|85T zopqnPl#m`QJw^qbVyA%;-b;9%-moZ|xukEBNfl$PLIZAKNDj;33j@S0R_>qj>33ht zExms7G~#ky#ahGneBWF6bARz)Qx>kXVs6_GQpYhYiK;I6iga;cs74~uD5-Z|Bo(Yk z7eEM$ueMFlJ7faVaJV<-^i4;{G#F8(RW*rI;V>UZ7jnY?yzpF76Y85y8ybX~8)|WI z;S=sVIKy*JU2dJ>>f+R3)%C>AMTBVy-Vg~z8T({G6G4M2ubi*!H(;9f%l=X&N-Jxz zCyOkZ#Ke&04wa!SMX?#;@S0?>zMJnAn=jWii0ii%s~BGRDNkLAfnGUby5@#*CqYmy zKl0Dsj!A5oJWLOiYYnDcz0*n{wW|%D+Hk2?L$$uKbI5kVPU+=>a5Zg7ZLc)th}w!S z1eo?6zswHVNlQfIoG6{oQ8A=&wgUmzN0gdyrQBe*7_nC^jEwd^5aLZBWE$-i!(+*v zR3UdE+o|5~s2nugoUcm`ghH0nGHtHNBTR}x1!Yn}#eeJVmQZ@!d19Y1|8A|V>It-4AIq?70#Y-*E)EvE38q7KG$B-0Y>2rjJ3 zVJe5jU_E1mZ}>mou>R&=Z&99f?WCt=RbKJ(XYyrV@`_%_zy0=_xvoGi(jC+hZGmdo z>=56-m@(JGL$!P|MPWVL{K`RH&Z^bMn``JnHhn{Ox7b+vglWr7)NN(){is~%meMCz zC6NpJ^}A;)Z>u5%!AZne6tP*p-+Sc=l(oJCtkR)>J^YJKX-F6*8wx&ed2i=C|v znAKyAhIz@241OR~?aQriP~@B%TrQfqNR5O{w$|FXEie9*J2`jr_Ej$CI?4&uH8&g* z1EFBs9nU<+gZEzS-@yb0VRMg5O=GH6nc%Y36?W|sP@+LnS28Za?m@W`R!~w~-b>U8 zdYn=3WWnwU4cgFWfCCuQ|0XEH<>`pcVbQxCLdk;lJhP#Vd6jT}wuR!kb9~6aR!!z8 zvM8Q5cL`G$0`11OATgd&E%Pd2S}ub0QsdZ2axTxe7^IGkG{^biJj$yC6fPcZ#=SeZ zS4+b0Wnou%V#2-BKr-be(0@0CQ?2{z1Y4NUvXjNCH%=7&cO|yh1Oh8Qkx))RY50JW ziNnquMinU>rcUt}&`}EWDx^UI!;Or;{*T|r>2q7XXM8F|;KT*neX__Gzv*SX>p%Q1 zd0usXZ{O`??r|!(p;^4*rvHtr>dYe_7E2Hfp(99gnBE*?_cX*%=rVWaH5o9Wha%c! zyhyV!T53+#e7~mm=@?z3!uk0o(<&5HJU0$2iAL8fozG*Kx|ER(B|DrK?%6%dmFY%w zmn(Pe?6Q&N+`D_0JGKvy7ICNul%zH+@BKOQR%)3-2(pVJ9+C*lc|bQ4j0#$@ux#5-Wto;4)_6u)wW1@v7p~-EHj)x! z`V588&j+Wca0mv^14pOW8O*SvT$*i?yO2V-mCZRdSfpA`-2knWqOwfN#Y!yjyJYOo zgH<=J;IPQZY;czq&T=$Ae&wbbGLC#DxFQM4Anh7*>II*hg%P>jmV(Tpu~NH%YJ`mt zxNFLOqp^dOiQooG8XrkhKdI(uaK15kJN+A`2=K?h{SCb3kA8N4u-;QFPY3WmHS(Ni z-^{b_zKM^1>^xP~zrYY6$(1Az?hTg;)#fU)pa$G9yc=P{Tf|(1O2^9RV#rJq1iVt7 zdV(ESL9e77?48EVE!!{HCE08#zN^S&GC;Dv1x5HAS3$$znN&k2Rf?)mddtIy+l-PD z=MA$e1o~R>+%%prN`ryu)N)uF5^+4dcbY4+QHaZnXz)F|b^MF)p-V$DTVmo2<{TXZ zrrs~~R~y}36&7EZ#o}`LI%>|BO^4aLEWIP)staJaYXTK>Om?>n_MLNB>@Qjk9;1cvnkLuPg@m+IK%LCLari>5Y8gyXs zZKAS>$at@0!|W4cz@RjwS$&_U<;lt4{DH6KRiE}OjFGOVQdVoSD;gWo^Z*|{4t(mG6GHcvS?+GKLD zC__?j^D5+}UCCBzU+o|*xYUZB8zhcVT2ZNGT4jx!%Bki2WQ$pJ7kMADo-R%{xH214 z^SLbg*miFTg-aK1t7?6=p!Y(R8)kb$%4v$rBP>!HC``yItpsujzvTj{;{WsE(6d+FpMzOUL%kh);-|KwWdg zA~678@%5j{mwwHw`3FDzF8<4Zd@n_Oqowp{Y8uNm{1{25Bp51nHR6UQVzZs9aaJa=N zEg2*gl@B$UG%{k&$_&yX6!k{&DjXI=_UA(?mo($c8c$+*II$MRw^oZC*2}cJ#GkFNGTb8Ws_9Z>6BJT{%4bsG*>Q`XZm%QvLs#1L|a>7l| zJ>m!c#@BH1;sM|Ctv}9QF{d0@FpwVk4p);V;76;5d!rjvgZpv~SV~7_3E93)rCUrp3S69y*+@$&Z!hlLB+(@fm{lnzZRN~@>qY5z8&$wS@x!vQMD0XKXW`;Zp}}}5Xzp`52@6)t z;(3cx?7U09i7H7Rk~*4Vx0PpbrReCHb+WxPB1xX&QYOdA3DY$A%#kxAXC>S1<)!vs#qc61Pj7jETN~Me?IE)u0 z7MWB4W5Pt7#IgY;NX^D?K?ku8lQ$Hz3^5*?`ADsAsj37KPd@9i>Q$vO7?sS*j9HmA z2u~VaFo(sE-T8=taa7(=X(%A%ma>6rgd|PY7kN7@1`Lu##()vUh2WW3JZ&LGVJY&( z>ol8rhNChlF@>&z3BUPrIlxz224Aylr!JD210wiAu>6y$U|OrfWpt%k99hLy3-?nX z7mM#f=}BiwAR1PldB<5Ey6+Mc9#i$)8xgz}Y|Dtb?T%AC=LNU*-p8k0;>!MMd1~dR zo3{B!Kl+XQ%^&{v{MJL~L4{I2FV<3rIbI}0Bv9mXly zz7MPQ+V>dq_XUuu&Td)wQIt4YPGYtaG^nQW5-jmH+^g|bmiSUv22&ag!4E6WAcbHK zZOE!A1?C1VOc+>=n@FMoBb|?Wp(?vpc#&I*7!*65r&P|eHyu;BkQq=$_S>!qE}}kt zqAJX)lv$Yu;;Tr0r7Dz_aCkW6@LFzG4YxEu#?Ne#A9;L;83wu-E-mohc4rtV&}(%+cyPP zl(T0x__2TaC)e-$agdXiG4RPMx8HV(ANj$rTgwOu+hm+R5uML)dgSBScXqxsRgDu-t(rE}y5QNGpSN3NhvclfmzUubBZ~6My z@I_zpvNe7F_{d4qHP?j1h{zj(zfS<<|Ea3};L`T*0=|U+!d(QuSyk`9&V9YAoH?`E ztmJFMq!v>uUOfzLG;lx*#ifEDL`5Nx%)u8C7B+y<=5j`Bm!-@nY6j^4K6pxPKil$t#JW@D-CJp8iN1~Wg!^x%q9a&8c6oac_5R^ zMkPto-(VlAU8hikA$Nw{En;C@r5qhCDp{)v)shRQ-sU~bCK*0Y5bKe&E+;McqBg`d z1PmyVKSQCA@M$Xw4fom3SXCd_vM^9VG#D#rC_!Nhu*H({&`1d`vc5oMOodbtp8dRA z`N#kKPkdrxVWJ3}xL~^{A~%Yj8Dt5@2;LoY(_4^))W&-XZqS3Mr@S!a;Nk|VFgiP9>yATY zN!K)Ux0==N7F%1|BtvcXgWh;xKVUt{xhO@8x?U1tbT; z@N!sKJ}$yLU2aYo?s|HYPQ3=CSHTS|$<(bK#{&HAAA1`weZ^h zevP=SQ7KWcsYC@Y-EF3zSxR>O)2+X~lFa<#wm z!Zgd{lp9IU0PJiI+1;P@Z;Vm1wTr2SDcoT3H>|v(&G-7>WClgeNflT2#w}L7_0PtI z3_L#|7%a=vs0#Lcie66fHk~X7yGVV+Q)VimTuf}+ z50&DS3k#^0eon6l*{(t2VrQ)Z<7@bSb-OItikg5fxw%c@u=DQ4R6|QLbIpDYq62)% zSG@8QZomL1QrBDy67y={qpErjfQbAQ@Z~_GTvYX!Tl@by@D10qlh-P*|AH6u6Yu!t zK&`)KVzIemW41{(6#UTizL2_(SAj4gdK2R+>gQPo!zQF_*ATFh9uU$bd%F`?WM0Fd zNd-#KOhOid@vEuZ8X(XjPiTj@t4izp>MEtoQu+$f$zOmuej;u%BQR{J@j#+m1WP%whC! zf}MdMDET>$8iE0%vV!DyPUZuFW`oGLd_EmcDzA(mDRBc6V)+t+%Mq~iRrit4EJbEW zW>vV*t(4ckz`W5_*@BMcn^r7c1Y*z(H<{D_AOIS&y(f$TBZnLFyCp;n=V`&ut z3oZ;wpk7h68`5Skv@QrIgkZhOAl|@uX{N1!B7yC0McNvrSSQ{G_fCENWo{SBO+ZXA zgo>jSYRI=yT*CZt7z~bKI9DIbCoM$>pc8{EoX;Uss775%p-DHb!s=2DXd-9=wlu-z zH(y=ZtcVihoB$KD$Fz*T=VfK&J=vjl;N3tazUDk4jUhQ0B)sh9&*X3Z;MX44n~%RJ z$Z@WX8$+&}o`I)c*L=g5znb@c=n;P7J@>6{jH}hQf#FLBlC)rg_WrDY!K3SFVOZ0& zv5elB0^B-{-YoCITuJbE6k$B?8_d*@>{_bRU4$mNU@jS z7KfEp->oXik;cHfwnW55R#&>U+;ZyYiGVowMnj7>eTPYKu2_fkJDpHU6nq071h{brB z&NZrH=GyRr4fd4Txv+3G)e=tZJR6VsRadj8san2HUh}fM`R1?wOkVo@+ppUHPlTK_ zU2`o)%)>yN0OQX1TSr9X^MStve6FhI%iBMm^0JrT#Si|-Tls;%_VZlXodD52WLID6 zBv-UjEQ-#c+z6y;UVw}Mz0yKWH3Z$QUGelN7f~UZ5UlV)93VtKlLlxJToo0A2om`R z6fpZ1H&irho#WCV3~te2;8=%LCiHW>G)rsP&leH~{VZ4o(t=d9GU)yQQs3_`suiaf zC=8mm&QXQmK+wV>lMs5lQioC^>JfjeX}XsKo~4TSCfaGPhMGD7H}PKw!^E z`4A@{b$kNBAWD|QdFtjQtgNKyQk0``^DW!_tDpQo`wu)R@;G>o`s4%z)!uM=2H>xK z=a=!L|KeBp>398hFv309!uKlr(Ul1IU4i*2L)m8atpUpxw4xFX^JCmlwP{9cCHDWEQ*>jb}_L2l7m7~?A!7{6*lK)|mi(*9ExeU)IG)P@Ww z$y6B}wC)!$z68Zs@k2VUzv0EmV%oFN#rJ0-3Gf&G^qctN&wuGPbo>)2Pg2)hZ@CsD z<~M=oiO91EO69G<*SB{Nkyil!1bCyWUbybPd;;={SKiIP`MK}l8~@lp;em%P$IO@2 zSEpksJ4j|8KeQNCTqLopiK)UrvbO$ny?ScC0F;oH;D&$-HdovzPSPcqm_Dr04dUy7 z*1CW3i=9JNNBuV|hp-1OD4_TapZegws-+Uh6MVH`wG9KctLe1HY1dv{eV@#NAWESQ z3Y0=f4os(wj`7vH9?{kci0*dr+HvieiooyF?ps7^vO_v9BZ>bLjWuZE3M-kfB4;a? zSl{{(JJ ze}?-WH!urc8Q+@o;(M16ip{#d8b8o^PdnhuG4?@?c-q6X>!vzJ#0%mrlop+WPGSIU zohT5E1x=LdjN{_&&YFvF{a}f!lQlhv?s30Afc^FbN>>-HJN6pKgSS15>(Mw15L!H+ z4+hL~Zv_}k33ZSFJ8_}*+A-4Tm?hcS9^L>3%*jJyt{S4Lsul3i6FzF$QC znuxqP?)JmLCh%`Wqj!P55%E!(<>0B9Hc4UWv znK3-z%P*1E{g27V zM+vfLk%@u1fy#hy88FM)JDs*##>vXf3VpCgLb7i$69cngMByOWk3n&*|Li;(TN+$N z*Cg6CSVSXCD%=EbUpLzyKmwdx1GNB0~FGU>wFK-aq+Ll@m6nd z{?*0)opHjo#oRxxl;P!LE!1h4QV2SB=tf0(b4hjUA{*!lOT1=5urnB(7Y64IcG|Mc zuEARCW8>}o8aAh-{D(V2R`WQmbPLsa^FOEI~DBEpER zvD8r9$NkkrRG`H#BVPI^-(`NM<$kpW_;P>i6^hBhR;IZKO_lOg48DnjQsSy_1%=gq zG-fCX^W0%7NDpj-1eq8gBvo_-nJ|c#Q_4+2K!a*6Tk8fBjKXC%k(md#kJ=a?6Au-q z3*Hvb0*N|KqM>Q>z}BoOkT0mv!)=Ob*D8pC@h}wV+}axQ1AqOiKk+Hl*C1C3wmT;x z?{0m6M~oS%<+=vdK2f=m^^Az{L*Mt+eD`1ZX&!!L7k3T8k8K2Tc{tYJ{c!2W`0sIV zq*ScBa+l=^&Np?H5UB!_uw1Seh%md;QQqXAaN7WPjJJ4n?}))gA=_K@GdmEfEW{T1 zG;rt?r|8BRLU89zx#m7(2LXwx26Wwa8nYJ#OdgDwV{oIzVaq!-EQZo`!$>e(lEhF| zj?aJni+Ix)ed@J!`V%al)PcPxQAC7q`}Qy3>;K4SanC&$*x4TOH~!WyaPLExSN5&Z zN?`qp9Y|vEt~dg^Ad0uLI7Xl*O$(v;0gMj_ zVdkoCCZup`{fP=eG$J?i0NrX9l#uvf+-R%d8HC@`Y3!B@o7#k5gZ;?k1B=3rz1l?= z*W@FEa}jsV5Wx`Hn9Fh0~`u_|#8%2Bt4X;-tyd;B~m9s;{~pr@EG) z+CQjr%gx*T^MCZm`PjV|+1;ChD1YJmeulE_!QTa3tHx5Fn?fN5!?$GRIky%a5~u)% zSHNXZZia()UT5KOZWbtFgjjv&niwlsk zO{n5EeCcyuz>mECP5g;}^RvABBlmYwqR|k)CM6BH~7B)>npf)>5vz_ z;C61kb!YXMlPFi$HOIletYR`69_JW@5V z#SV^vD(8x+94-rng?La=sxi12yJ2Tx)liMP>Adx1UTnf@4xu19$dEtXzU0&Rcfb37_OHyztDLv~ zoqxq^UvM{xGq_H`OMQ;^`NHpum7tK$g2KZBWnH%@N=+)|qnpUBwna=o&yMI+6AuJ~ z_4BKbiqEms#^;|M`bbb1U4e;U3T3&>oH4>{KkNCQ*y3tWNS-tS)eRX`dlE%NcR7e!%vi7dgvp(={=9}{4-}UA}jHuD@?>*8C-%{OyUtAV&uy0XpHpq z4{T@Vy!Zr4*n@36}nnf-;qo!~wJ1_FxJ!kuEGLG(=fKKtS5_uU27YQf?n}(H=}L zBvYl@1hEpXqxP|d%O;sA|IdH^tq|7(IFzHZ;6J?Q!{IWuEzz~eUCkOG~kU_OKg8a~5{c3*j@4bs(`^}HxeV|$>whV~GRNF9! zsv)b38V|#Lg#&H8!3IGO6@NuJx;@kpvo4t9ePd{JSvedL(1}_i(s1fy8pc8o3{?M= z#07^ff^NunfO-F@54GaLC7CJ61l-NNyFz8ydw9zKDDG`S{0ck|UfJi7-2-FUssR()vZXTEQ%n!qpz*#n#|cnr(C?}wM}9$& z8W)+me0)6YS$E&W+rHtmkH5H+Do;+=JSBvde;|z*fH!~XEBKN(y^Q_+DfiuXi9h?@ z|L@8^>D1#hk0y2gqlSsES_}el;4%dh_DN>oOh5s2p!9{>i3oP4xIy=2XbQjCE_Jxp zE}+_H+cxp*uNoD|5^-hdqmb?^X$+1_X3G2~7LaRPp(N9mKY0m-Ru@o^K^}e5qs5`U z#z%CY*>bVn5?WW*wHLx15mT8maJOA)UzTw7jQ?HZbbXmo$X2Wf(jwn0@?Nou5^2W_dvI2i0IZV*tH zq=*Jlt;kv;5K=JfA9~)QMp>O|9i;U&fbolgB`SZ@Hb? z`zyiKq|xm+8l-&Hm%sAHHele%x|43`pxPTj#-oh!DC5KTU+j&42}r4;Nm%KoN4|;I zX!e1H`y0Rs`rK)W!?LrSL{sGGN;MPs_M=k)?Y|K(pKXE?--Kn1~!+ zu}>Pv?^g6uyYC_xD-5!PKl>eTy0HxyuBvOUOZ?sAiM{&+ZN${F_K~P!D#(s1s$qaF z#C35(altBgsRz@oxnb&Tqvdy2H86U5FfO`CqLJc;*Z6`jb=pLBTYC}8Sp8P2fDGAr zL^z{wThq{`BoCz*LIzNTw`KKy$WrV}{cm=pxD4D_IwRN4TKchd{nC?JRg0|@GZp+8 zToUeSI**LLLgNxzC)ttWl8V7$Kao^(EwN8$Pi=71nN7att3I92``i~DbHoiICyapV zX;~v{9+RjtI9PI5tT|i=Xs92svf$}y84@L(!r*Xm^5YU0)``;8dhFfiQbH`@vZ3Jw zv)d`Ny|$5L*1z!Roao8e09xyZrL$S0fgy zDG$S4$hSSMGThz%t|7$|!ivY1QGrFMyV=)!k_n`<(3f&7v`as9+ETDSphkTnbQhtk z$1B;9#}vMHNe8sWtZLmfwfxM#__qFrZa@jo_h}>MMvxah=T_eT(MP*K)^1^KP1?sy z*5weqmHxrd`x*#7UC`yPS_aDvNTXMl`><}2<%XbENspAlu0yRbKWkU_j|CL_^gvUF zWu^Ds-t}!Txo6{%g6z=YtZ?wWQUCo}R|;FgJ(8PJT5OBZ5932M=PYI; zpaJ1XX3EB6!N)$l?6EWLT&u3vc)@t8&r!z;#r_TU2*noTxzkBcJ6c;Dp|(tNQr z-s@P2P0-cfy^`%#nB0@heKQdlLMBIFQK>h){*wTCSr+ftf1Z{TBR}+=Z)T9Kt{vst z5_;m!t#S4^j%oiZs+eij=#hF&y`V$@03FjwL_t(6*;yQWZPLs5sfGbLUOul%FkrUt zsW|=nB^!@=HXnAJdvC$;Vjb15n5k#yW6G(IDmx!lHt+Ef+193=W2~#z30wcFV&;z1 zADA)Nja*i_GPqp!-yc;>p={naXXBwfTyM`leWkKZ?n$P~iY3(-98ExBzVHn%x%&0o zXz~Pg%~M5a`7{y}<$J&HE&R^!+{aJ<>)*h85YuT9sIe`G1hnHg7&iEfS@5_( za2?G($y{P)O!!?{LM`IpQW1uieMrGf!bEK65kgKvwt!9-blKt+J5lnpAv-6T1!Kt@ zs77YR4O8ohWDZrwYQjeq{{|2Mz&{s$Rm3BUFJ2Pk!^$l^7$)d}{e z=b(jhXkSbQ;I6R~~W=bXgGL6NFtn5p+So8NhfXHLqs ze!Jxa_fEq=PX1VV*JRVXyIdW|5oQqoPJlW+r%)=#M zQ1Js}vxa;EoUsOiLlL*hcbSbTEAy1BiKo7o+7Tx<;NDeyK^l-wszXgaB}Iv+mA%m? z=xr-5rIZ10yUq_VIPZSsYiLJAYupHLHFtn8DfhMHTL*C?2o|P?gwnzn+v?>x)O7cU z03meFJnjMN9Fe9zcvfCe=O&nIZ!f#q$D|$C&vVhX zw|ZD-2XxWRF~Uut`uLVpL$24(Xdj!)VLXsF2YbqEPWw?}qv~ z7M-p6onhb}+wB?6v**{;B7V++nzkTI4{0s^XeIr3ifb0doZopQd-M<@(aV`HEQZt- zNwBw<2y{{pLK^~=wStdm1kc0hcy>J zjN1k7^|<0Z392!nkNe#MF1}58X|GsAR;4uywABR!DCzm(U zjhs+07GkWhI#5PCS0B}YdC5zddoZq3QsIc2TtV}AJiN_+rX#)QG4q>5S6|QFQcrMG z#9e3D94gIcb7BUq)QH}eZHm*E%-VZiHR$_mA(ii#oJ6Fk)_rj)`Nf?Pw0iy04$0{6 z)p@Z}UM^fmPfCb?JHaQc242M%3=N#mDN`X^4xXS_HVI`;%b_XW&bGhS9C!5JrYOD= z0Ik>VMqKjuwJZUw&I&)(qZw2OCXk^=36Ro%{iHD!D%0N*I81Um#1*j8lyiZWrmlm! zQo4ITvgg&C?srytLB{1v9C+mPOhMh|737&~zOwNkEEjjBCK?SZ?$Z4+TvfxwJS2NlT2Rz;t?uBd$ig1{m!ZSgy1eoY=i6$J@vP)V?ZTVH~`em zVhmLo{yNJ`lN)E=7_w3%5bMYi7|E$6#eF@Q6)7B`UHkq#LQb06-W*Mt{o|9hB%!qq zU5;Ev+hy^CNXb>1fQ})VB%k7D;@Gw}!A$wB(%&}@XU;{I0#prZ^9~%jv8;jc4>~@I zhbOb&*!#S`-_z~3ld`uA1iPbbu+m-N$ci?_<2a2U1x@Qd52umj4oTpX?BkJBU~H#R zXcpx$Ni>;{#1|Q9i}hvm1me$`*jrk#f~%Z&$d_|6UmAbQOMlWrwp!YrBfV+iJoRq# zW?ME=y7Zx&pGS|{ukjnV(X#viNTJkBuwrQ!v1tZRkxh4SK4`t-k&_f?B)T8s-W`6~ z6uLC*c|iCnGiKd=wBq1URlxIX`7Z(v&R6O$|2*ma#{v?h3;h~@NvRvY^F7_?WAs+9 zP23=L|2udBE(sin5vk4*IBCEHr?&s273vnDlp1uK)W0&(Br{n>YkkDr&#ZoMI4FsD zBZ|d{HpebbLfb2r=y9fsn>G3!xi<3;j1~v1wtP1u(1TwZaQkJEMGWB`dM zbc)idM&T$Z;kaLEt2$h!b3^leaX+A<5_uL;I!iD`BIcosBXQ=7)Z)$1;&KC$zc9#L zW@S#O^i&?T1nNv)xvNjBFsoSGWju(K?(2YmS3j$PuUi6d{%mx4KEAEi2Xih}T{81( zfq-Eorpr_5PR}AyQ?XK10ZV?I7|Gz;wRvgWbW2wo1+>zFEJZ^M7jZ?gNCE|VVh`U6 z-)_>lRrsb_AAVJAw&^sBbdb%M+BcJEQmSs7A9TBQMQZjvE}vq=}KEoJep0aE3qL2;yQp|&t&owLJQy*QY~q- zGF|X5$<~XMu~Jr-qZrHo7W04MN{SqOn9k}uIws~R zsjEH@p$|^uwBExrCl1DZdQH}{(ssBE!0ivo4GJ+S&;u;OTKzQpJR47*1ABN;awG+X&Pxt%I zMkG;#w2HB3pQk(?IcTLtgYwl zp8ee72lv%$X@jb4Q~M#_4SG*MOp~Yf+zi|I0$9%gmw{DHv_Z9^e2TordQk@|cSHrN zW}7W%#JgJvicja31FV?JI(4h~Uf^Fq?8b+4#DTv|k*Dg(_Ek(0{zjgatR4yZ#eA~A z7N3?!9~rz0^UEhx+V}QI@u!Ys4{gw+1aVmgqI2mjm8tYvL{uL83x( zJX|Ur@9o0xR~<>*d|(UDl37VIueyd%c!+xI!*Bv0-4{1nf*nfi1xC6bjAL$^fcj8z z{+vE#23qp)cGi<8md+v3o}q%Hmbu2#+MA~BxXNMss%z7eRy~f_Rq>54RB?kC22$cTS<(7Lx7r)(u#CJ(LtO4%qnB1zkJVA;9&j=3KDtBo_DzeuAbJ{z zbU@}X>gpmd2FN}XUukjv2U}{T&7ME))pb#2+mo;E-v^P{x9>iP@lulQO?(47f;(v= zS=1KJ37`u4X_vI3$%>*eLq!4U?^$C67nAbd$kZ8#)y;{xrVPImKWfQm-~;~oEbEGW zVvAgd5z3+-)m-BIm-!J-Ha9s3k|)#+ma3Bo2 z_=JQv;))E9tLV~;G6XEO>?iJQc7`?!5h~bhitY7(8w(tjdhFjGqticY$WxftB@cZn zbz7!-w#2T^k+<8@Ll?b_&;raPHmfP#{}44W#IleFqjUQ|OAQjD@?b1$W=EwAyFU1{ z-R`!>iQmIi83^=;sQ9uzvF*V&Kb0oTJf`D^aj#4{Th_8~P6g`sV{P1uhzv7kyFJ-n zu>5qrH>fT?ZW2Lvpt;dt+{l@yz)IJxo zJL!ZUx2%#Ez%LZ|@%1p}yvE%(h< zo2Hu{6m6D=QA?Oo9b&+bA()|%YrHA!(3^T%mSD91GYd#6g9TYE>93)>_~|^B++!P6 z`;A!2D-4ff%SBV&z<`>47Na#$@+ST3?~LrvdVI}}b*|T_RYv|Y`t1r#o9~3 z!gbswp@hC0GG*ry-&UMP|0;7n8pJvpQcBbX*R1=C46BhN_X?dq4j|N!ABIi;8z$T( zD-qGQ$>Ku|f}dLt#~u_J8+Pik)i^V-NgnB9`u=?5i??3N=4d=Q?FCCXkG&CpoPTSg zfNtq_x4-35rC0UP%Ajqd3JxYKzZbHW$^UxV!=?gy(-f5-#jmv?233{7ap*`_kLn$; zRY-DB-`p{>nwH2Vs^l^oyYgJgqE7|B&Z#4fXa5C#OOh+n;VX62+uk&7wkBAv)4U|A z2x+mY((|V~?t$9ZA5!|yo)SF$>UI!2IOBjx#_sGZ;Oa(Mx`v-?4iyJn^JK?swG-Hu z9d|{8RgJ&R#x=Q9s%RP$SU~66*gG!S%^8Vz_^;q5rKCz-9Y~Dqxjj7r^#^rz9e7dd zp?^c8L`bVVitkGmeaz%NwON$#C&Z~&p7gQzL7xW$xr{Hde6?G9@8=e9UYB^eqFIER%63NCl%{#4} zTTkk5+&06{cX!h_7}DL~eOs@zJDT1=^(OX0_wyz$?TL9Gc9g^~ehjZb@1+hLRQKH8 zx4&gZFFfw}ycQmn$7pH@KU>MJfX96d)@aYI?#&x2@cfq*-NlI9^C9pKDg>0O&PPn89ipis;UHdcxstKaoW`Fh@jne*tyyM zqQu>f{zcDEiNMzTtl9>)gSWYQ)!%4Gnqi0P zcW{n0u?$4ZM*$s~a@^V1Ojk|3YRB1Ku^CHwb>&i6dX7ixaq$2=6idI%J~_$^;EbI4 zoX8(h*y+PMxGqnMxaxKj)`b6tLH!_GH<}w7VqPSJiOgYcLX=EH_7NFHOoFb{DMIy* znwq3_}M5U?k8HpqYI?ME>{G?(~UiR&hf%m!ncQU5B5# z4@|_Bp)1e6d4)5vcR6huNmbW&-Xa8DGyS{;306}x?5;4e}F zoui(i+I;>jC{QX}{LxRsjU;zrA3AGstPjl_*T31#OH#UTA{%cmtFGLt#%giC+JP>_ zSL_X98Q$KH`P%$;!^O+s-`4x%<0xdD0~#!fJR-N(uRUQGl!B2r71;~lo#;^UZ1v)7X%iZ#~$^UWjwX{i9hx!Wb0qd?nN8p$l|ET0uhJov1#x&2TGKl(%NoU z-R@4g1nWF?-?W(_NOy{kt%oy1+iEz9KD%Ri3~Y4>y@U-1HsUQ^M-Bn_sn}JU=T26EA6_J_v3~8l)ND0|jPF-OH;F<9U%;jvpavR;3>JKA`-|MT9QW;b>c?KyY&O&u)(}usf#1ajr|niw?SAD`JAp}9=Zc1B zTH*HM5=ho01-B<77kN<$<7`Mjzw&zrdRZ@;xb7XTXlGSye#ofcJU;K7oSba8oiFjM z*f8^2q62W1*{vjP+Spf3#hAh+;YIc3AaM$wTBU}EG@kOy5g)Ev4OuNC(}8C<;w{#5 z?J!N*QJ&n+2eXt4Nu);{=bp$;#72bl4_b|VW+O#mTq|eJ+$I*>VcFb1s`?^4758s1 zY-ZhFW-Y|9jCfn;{{1GaEF{AIEaK;&S6m;iIb1BQD&B=|8cQ$RDWHXJ{#IJ5|jd`kQZ(WCVt-bwe-|vwi zealZKflMH%ihL--kQeOiGnt6Wvr7xe?*E!-bA_q046lv|ezk3lFQ_U%-5aK=y?mvE z!AGCiXU0v68SyddwK*nQD7=9AzMpzvU2Fjf&5i{-_a4yt|yp zPEx?CB`d?2wvP_vntlRvu)C zr~`I-#m~acx=fAL2~v`&xaWUH3iq!g)GL*oM#KW}79?lM3&SbOX!j~!lW+6UF<{C} z0L`ouo{an47{|*hkSS4yunTf?s|FI>*#?7rLYVrub*`6)7Briw&Cs}Ifr8N^ZUC)A z+p_Ul_czYrkthA8W*2#ERkpO)*OGjle6iD*vCu5f>V4waX-drrq{MRsyf!2oG-B)m zv^Td*&cEQXr7|cznKzV=tK0Jc*ZPWPtb&s9AexFJW58tqBmmDl!EPkYQh+r2Rp>N| zDBKH(g=W&LkQ$hKAP-+J=MjqQ*2hKyFR~A@F$<`HmUzy5wLlOb(VSL%t`9__7Y%PE zc;oy|5tZO~t86 zX9p?djXz!hN*7V9h}YQ%_$2Oj_C!5cC`bxH`-|RmN-s6!I|MqSam2!EsA<6hyMGnX z03(JFLFRDRWgCLyf}x!#ezpBRi^wksuD4BtslBXbpalCyrkMxj?^1j)Zb?4Utmv_L z;eP|_{-W>E`%fRye%3Nziu|8qeTn^lUU-#{{tq$qsssQ literal 0 HcmV?d00001 diff --git a/docs/source/index.rst b/docs/source/index.rst index 1e841bb..52fa72b 100644 --- a/docs/source/index.rst +++ b/docs/source/index.rst @@ -38,6 +38,7 @@ solve problems in a continuous and nonlinear settings. Getting start with PINA <_rst/tutorial1/tutorial.rst> Poisson problem <_rst/tutorial2/tutorial.rst> Wave equation <_rst/tutorial3/tutorial.rst> + Continuous Convolutional Filter <_rst/tutorial4/tutorial.rst> .. ........................................................................................ diff --git a/pina/model/layers/__init__.py b/pina/model/layers/__init__.py new file mode 100644 index 0000000..6dee651 --- /dev/null +++ b/pina/model/layers/__init__.py @@ -0,0 +1,7 @@ +__all__ = [ + 'BaseContinuousConv', + 'ContinuousConv' +] + +from .convolution import BaseContinuousConv +from .convolution_2d import ContinuousConv diff --git a/pina/model/layers/convolution.py b/pina/model/layers/convolution.py new file mode 100644 index 0000000..62cda88 --- /dev/null +++ b/pina/model/layers/convolution.py @@ -0,0 +1,154 @@ +"""Module for Base Continuous Convolution class.""" +from abc import ABCMeta, abstractmethod +import torch +from .stride import Stride +from .utils_convolution import optimizing + + +class BaseContinuousConv(torch.nn.Module, metaclass=ABCMeta): + """ + Abstract class + """ + + def __init__(self, input_numb_field, output_numb_field, + filter_dim, stride, model=None, optimize=False, + no_overlap=False): + """Base Class for Continuous Convolution. + + The algorithm expects input to be in the form: + $$[B \times N_{in} \times N \times D]$$ + where $B$ is the batch_size, $N_{in}$ is the number of input + fields, $N$ the number of points in the mesh, $D$ the dimension + of the problem. In particular: + * $D$ is the number of spatial variables + 1. The last column must + contain the field value. For example for 2D problems $D=3$ and + the tensor will be something like `[first coordinate, second + coordinate, field value]`. + * $N_{in}$ represents the number of vectorial function presented. + For example a vectorial function $f = [f_1, f_2]$ will have + $N_{in}=2$. + + :Note + A 2-dimensional vectorial function $N_{in}=2$ of 3-dimensional + input $D=3+1=4$ with 100 points input mesh and batch size of 8 + is represented as a tensor `[8, 2, 100, 4]`, where the columns + `[:, 0, :, -1]` and `[:, 1, :, -1]` represent the first and + second filed value respectively + + The algorithm returns a tensor of shape: + $$[B \times N_{out} \times N' \times D]$$ + where $B$ is the batch_size, $N_{out}$ is the number of output + fields, $N'$ the number of points in the mesh, $D$ the dimension + of the problem. + + :param input_numb_field: number of fields in the input + :type input_numb_field: int + :param output_numb_field: number of fields in the output + :type output_numb_field: int + :param filter_dim: dimension of the filter + :type filter_dim: tuple/ list + :param stride: stride for the filter + :type stride: dict + :param model: neural network for inner parametrization, + defaults to None + :type model: torch.nn.Module, optional + :param optimize: flag for performing optimization on the continuous + filter, defaults to False. The flag `optimize=True` should be + used only when the scatter datapoints are fixed through the + training. If torch model is in `.eval()` mode, the flag is + automatically set to False always. + :type optimize: bool, optional + :param no_overlap: flag for performing optimization on the transpose + continuous filter, defaults to False. The flag set to `True` should + be used only when the filter positions do not overlap for different + strides. RuntimeError will raise in case of non-compatible strides. + :type no_overlap: bool, optional + """ + super().__init__() + + if isinstance(input_numb_field, int): + self._input_numb_field = input_numb_field + else: + raise ValueError('input_numb_field must be int.') + + if isinstance(output_numb_field, int): + self._output_numb_field = output_numb_field + else: + raise ValueError('input_numb_field must be int.') + + if isinstance(filter_dim, (tuple, list)): + vect = filter_dim + else: + raise ValueError('filter_dim must be tuple or list.') + vect = torch.tensor(vect) + self.register_buffer("_dim", vect, persistent=False) + + if isinstance(stride, dict): + self._stride = Stride(stride) + else: + raise ValueError('stride must be dictionary.') + + self._net = model + + if isinstance(optimize, bool): + self._optimize = optimize + else: + raise ValueError('optimize must be bool.') + + # choosing how to initialize based on optimization + if self._optimize: + # optimizing decorator ensure the function is called + # just once + self._choose_initialization = optimizing( + self._initialize_convolution) + else: + self._choose_initialization = self._initialize_convolution + + if not isinstance(no_overlap, bool): + raise ValueError('no_overlap must be bool.') + + if no_overlap: + raise NotImplementedError + self.transpose = self.transpose_no_overlap + else: + self.transpose = self.transpose_overlap + + @ property + def net(self): + return self._net + + @ property + def stride(self): + return self._stride + + @ property + def dim(self): + return self._dim + + @ property + def input_numb_field(self): + return self._input_numb_field + + @ property + def output_numb_field(self): + return self._output_numb_field + + @property + @abstractmethod + def forward(self, X): + pass + + @property + @abstractmethod + def transpose_overlap(self, X): + pass + + @property + @abstractmethod + def transpose_no_overlap(self, X): + pass + + @property + @abstractmethod + def _initialize_convolution(self, X, type): + pass diff --git a/pina/model/layers/convolution_2d.py b/pina/model/layers/convolution_2d.py new file mode 100644 index 0000000..02f21b5 --- /dev/null +++ b/pina/model/layers/convolution_2d.py @@ -0,0 +1,548 @@ +"""Module for Continuous Convolution class""" +from .convolution import BaseContinuousConv +from .utils_convolution import check_point, map_points_ +from .integral import Integral +from ..feed_forward import FeedForward +import torch + + +class ContinuousConv(BaseContinuousConv): + """ + Implementation of Continuous Convolutional operator. + + .. seealso:: + + **Original reference**: Coscia, D., Meneghetti, L., Demo, N., + Stabile, G., & Rozza, G.. (2022). A Continuous Convolutional Trainable + Filter for Modelling Unstructured Data. + DOI: `10.48550/arXiv.2210.13416 + `_. + + """ + + def __init__(self, input_numb_field, output_numb_field, + filter_dim, stride, model=None, optimize=False, + no_overlap=False): + """ + + :param input_numb_field: Number of fields N_in in the input. + :type input_numb_field: int + :param output_numb_field: Number of fields N_out in the output. + :type output_numb_field: int + :param filter_dim: Dimension of the filter. + :type filter_dim: tuple/ list + :param stride: Stride for the filter. + :type stride: dict + :param model: Neural network for inner parametrization, + defaults to None. If None, pina.FeedForward is used, more + on https://mathlab.github.io/PINA/_rst/fnn.html. + :type model: torch.nn.Module, optional + :param optimize: Flag for performing optimization on the continuous + filter, defaults to False. The flag `optimize=True` should be + used only when the scatter datapoints are fixed through the + training. If torch model is in `.eval()` mode, the flag is + automatically set to False always. + :type optimize: bool, optional + :param no_overlap: Flag for performing optimization on the transpose + continuous filter, defaults to False. The flag set to `True` should + be used only when the filter positions do not overlap for different + strides. RuntimeError will raise in case of non-compatible strides. + :type no_overlap: bool, optional + + .. note:: + Using `optimize=True` the filter can be use either in `forward` + or in `transpose` mode, not both. If `optimize=False` the same + filter can be used for both `transpose` and `forward` modes. + + .. warning:: + The algorithm expects input to be in the form: [B x N_in x N x D] + where B is the batch_size, N_in is the number of input + fields, N the number of points in the mesh, D the dimension + of the problem. In particular: + + * D is the number of spatial variables + 1. The last column must + contain the field value. For example for 2D problems D=3 and + the tensor will be something like `[first coordinate, second + coordinate, field value]`. + + * N_in represents the number of vectorial function presented. + For example a vectorial function f = [f_1, f_2] will have + N_in=2. + + The algorithm returns a tensor of shape: [B x N_out x N x D] + where B is the batch_size, N_out is the number of output + fields, N' the number of points in the mesh, D the dimension + of the problem (coordinates + field value). + + For example, a 2-dimensional vectorial function N_in=2 of + 3-dimensionalcinput D=3+1=4 with 100 points input mesh and batch + size of 8 is represented as a tensor `[8, 2, 100, 4]`, where the + columnsc`[:, 0, :, -1]` and `[:, 1, :, -1]` represent the first and + second filed value respectively. + + :Example: + >>> class MLP(torch.nn.Module): + def __init__(self) -> None: + super().__init__() + self. model = torch.nn.Sequential( + torch.nn.Linear(2, 8), + torch.nn.ReLU(), + torch.nn.Linear(8, 8), + torch.nn.ReLU(), + torch.nn.Linear(8, 1)) + def forward(self, x): + return self.model(x) + >>> dim = [3, 3] + >>> stride = {"domain": [10, 10], + "start": [0, 0], + "jumps": [3, 3], + "direction": [1, 1.]} + >>> conv = ContinuousConv2D(1, 2, dim, stride, MLP) + >>> conv + ContinuousConv2D( + (_net): ModuleList( + (0): MLP( + (model): Sequential( + (0): Linear(in_features=2, out_features=8, bias=True) + (1): ReLU() + (2): Linear(in_features=8, out_features=8, bias=True) + (3): ReLU() + (4): Linear(in_features=8, out_features=1, bias=True) + ) + ) + (1): MLP( + (model): Sequential( + (0): Linear(in_features=2, out_features=8, bias=True) + (1): ReLU() + (2): Linear(in_features=8, out_features=8, bias=True) + (3): ReLU() + (4): Linear(in_features=8, out_features=1, bias=True) + ) + ) + ) + ) + """ + + super().__init__(input_numb_field=input_numb_field, + output_numb_field=output_numb_field, + filter_dim=filter_dim, + stride=stride, + model=model, + optimize=optimize, + no_overlap=no_overlap) + + # integral routine + self._integral = Integral('discrete') + + # create the network + self._net = self._spawn_networks(model) + + # stride for continuous convolution overridden + self._stride = self._stride._stride_discrete + + def _spawn_networks(self, model): + """Private method to create a collection of kernels + + :param model: a torch.nn.Module model in form of Object class + :type model: torch.nn.Module + :return: list of torch.nn.Module models + :rtype: torch.nn.ModuleList + + """ + nets = [] + if self._net is None: + for _ in range(self._input_numb_field * self._output_numb_field): + tmp = FeedForward(len(self._dim), 1) + nets.append(tmp) + else: + if not isinstance(model, object): + raise ValueError("Expected a python class inheriting" + " from torch.nn.Module") + + for _ in range(self._input_numb_field * self._output_numb_field): + tmp = model() + if not isinstance(tmp, torch.nn.Module): + raise ValueError("The python class must be inherited from" + " torch.nn.Module. See the docstring for" + " an example.") + nets.append(tmp) + + return torch.nn.ModuleList(nets) + + def _extract_mapped_points(self, batch_idx, index, x): + """Priviate method to extract mapped points in the filter + + :param x: input tensor [channel x N x dim] + :type x: torch.tensor + :return: mapped points and indeces for each channel + :rtype: tuple(torch.tensor, list) + + """ + mapped_points = [] + indeces_channels = [] + + for stride_idx, current_stride in enumerate(self._stride): + + # indeces of points falling into filter range + indeces = index[stride_idx][batch_idx] + + # how many points for each channel fall into the filter? + numb_points_insiede = torch.sum(indeces, dim=-1).tolist() + + # extracting points for each channel + # shape: [sum(numb_points_insiede), filter_dim + 1] + point_stride = x[indeces] + + # mapping points in filter domain + map_points_(point_stride[..., :-1], current_stride) + + # extracting points for each channel + point_stride_channel = point_stride.split(numb_points_insiede) + + # appending in list for later use + mapped_points.append(point_stride_channel) + indeces_channels.append(numb_points_insiede) + + # stacking input for passing to neural net + mapping = map(torch.cat, zip(*mapped_points)) + stacked_input = tuple(mapping) + indeces_channels = tuple(zip(*indeces_channels)) + + return stacked_input, indeces_channels + + def _find_index(self, X): + """Private method to extract indeces for convolution. + + :param X: input tensor, as in ContinuousConv2D docstring + :type X: torch.tensor + + """ + # append the index for each stride + index = [] + for _, current_stride in enumerate(self._stride): + + tmp = check_point(X, current_stride, self._dim) + index.append(tmp) + + # storing the index + self._index = index + + def _make_grid_forward(self, X): + """Private method to create forward convolution grid. + + :param X: input tensor, as in ContinuousConv2D docstring + :type X: torch.tensor + + """ + # filter dimension + number of points in output grid + filter_dim = len(self._dim) + number_points = len(self._stride) + + # initialize the grid + grid = torch.zeros(size=(X.shape[0], + self._output_numb_field, + number_points, + filter_dim + 1), + device=X.device, + dtype=X.dtype) + grid[..., :-1] = (self._stride + self._dim * 0.5) + + # saving the grid + self._grid = grid.detach() + + def _make_grid_transpose(self, X): + """Private method to create transpose convolution grid. + + :param X: input tensor, as in ContinuousConv2D docstring + :type X: torch.tensor + + """ + # initialize to all zeros + tmp = torch.zeros_like(X) + tmp[..., :-1] = X[..., :-1] + + # save on tmp + self._grid_transpose = tmp + + def _make_grid(self, X, type): + """Private method to create convolution grid. + + :param X: input tensor, as in ContinuousConv2D docstring + :type X: torch.tensor + :param type: type of convolution, ['forward', 'inverse'] the + possibilities + :type type: string + + """ + # choose the type of convolution + if type == 'forward': + return self._make_grid_forward(X) + elif type == 'inverse': + self._make_grid_transpose(X) + else: + raise TypeError + + def _initialize_convolution(self, X, type='forward'): + """Private method to intialize the convolution. + The convolution is initialized by setting a grid and + calculate the index for finding the points inside the + filter. + + :param X: input tensor, as in ContinuousConv2D docstring + :type X: torch.tensor + :param type: type of convolution, ['forward', 'inverse'] the + possibilities + :type type: string + """ + + # variable for the convolution + self._make_grid(X, type) + + # calculate the index + self._find_index(X) + + def forward(self, X): + """Forward pass in the layer + + :param x: input data (input_numb_field x N x filter_dim) + :type x: torch.tensor + :return: feed forward convolution (output_numb_field x N x filter_dim) + :rtype: torch.tensor + """ + + # initialize convolution + if self.training: # we choose what to do based on optimization + self._choose_initialization(X, type='forward') + + else: # we always initialize on testing + self._initialize_convolution(X, 'forward') + + # create convolutional array + conv = self._grid.clone().detach() + + # total number of fields + tot_dim = self._output_numb_field * self._input_numb_field + + for batch_idx, x in enumerate(X): + + # extract mapped points + stacked_input, indeces_channels = self._extract_mapped_points( + batch_idx, self._index, x) + + # compute the convolution + + # storing intermidiate results for each channel convolution + res_tmp = [] + # for each field + for idx_conv in range(tot_dim): + # index for each input field + idx = idx_conv % self._input_numb_field + # extract input for each channel + single_channel_input = stacked_input[idx] + # extract filter + net = self._net[idx_conv] + # calculate filter value + staked_output = net(single_channel_input[..., :-1]) + # perform integral for all strides in one field + integral = self._integral(staked_output, + single_channel_input[..., -1], + indeces_channels[idx]) + res_tmp.append(integral) + + # stacking integral results + res_tmp = torch.stack(res_tmp) + + # sum filters (for each input fields) in groups + # for different ouput fields + conv[batch_idx, ..., -1] = res_tmp.reshape(self._output_numb_field, + self._input_numb_field, + -1).sum(1) + return conv + + def transpose_no_overlap(self, integrals, X): + """Transpose pass in the layer for no-overlapping filters + + :param integrals: Weights for the transpose convolution. Shape + [B x N_in x N] + where B is the batch_size, N_in is the number of input + fields, N the number of points in the mesh, D the dimension + of the problem. + :type integral: torch.tensor + :param X: Input data. Expect tensor of shape + [B x N_in x M x D] where B is the batch_size, + N_in is the number of input fields, M the number of points + in the mesh, D the dimension of the problem. Note, last column + :type X: torch.tensor + :return: Feed forward transpose convolution. Tensor of shape + [B x N_out x N] where B is the batch_size, + N_out is the number of output fields, N the number of points + in the mesh, D the dimension of the problem. + :rtype: torch.tensor + + .. note:: + This function is automatically called when `.transpose()` + method is used and `no_overlap=True` + """ + + # initialize convolution + if self.training: # we choose what to do based on optimization + self._choose_initialization(X, type='inverse') + + else: # we always initialize on testing + self._initialize_convolution(X, 'inverse') + + # initialize grid + X = self._grid_transpose.clone().detach() + conv_transposed = self._grid_transpose.clone().detach() + + # total number of dim + tot_dim = self._input_numb_field * self._output_numb_field + + for batch_idx, x in enumerate(X): + + # extract mapped points + stacked_input, indeces_channels = self._extract_mapped_points( + batch_idx, self._index, x) + + # compute the transpose convolution + + # total number of fields + res_tmp = [] + + # for each field + for idx_conv in range(tot_dim): + # index for each output field + idx = idx_conv % self._output_numb_field + # index for each input field + idx_in = idx_conv % self._input_numb_field + # extract input for each field + single_channel_input = stacked_input[idx] + rep_idx = torch.tensor(indeces_channels[idx]) + integral = integrals[batch_idx, + idx_in, :].repeat_interleave(rep_idx) + # extract filter + net = self._net[idx_conv] + # perform transpose convolution for all strides in one field + staked_output = net(single_channel_input[..., :-1]).flatten() + integral = staked_output * integral + res_tmp.append(integral) + + # stacking integral results and sum + # filters (for each input fields) in groups + # for different output fields + res_tmp = torch.stack(res_tmp).reshape(self._input_numb_field, + self._output_numb_field, + -1).sum(0) + conv_transposed[batch_idx, ..., -1] = res_tmp + + return conv_transposed + + def transpose_overlap(self, integrals, X): + """Transpose pass in the layer for overlapping filters + + :param integrals: Weights for the transpose convolution. Shape + [B x N_in x N] + where B is the batch_size, N_in is the number of input + fields, N the number of points in the mesh, D the dimension + of the problem. + :type integral: torch.tensor + :param X: Input data. Expect tensor of shape + [B x N_in x M x D] where B is the batch_size, + N_in is the number of input fields, M the number of points + in the mesh, D the dimension of the problem. Note, last column + :type X: torch.tensor + :return: Feed forward transpose convolution. Tensor of shape + [B x N_out x N] where B is the batch_size, + N_out is the number of output fields, N the number of points + in the mesh, D the dimension of the problem. + :rtype: torch.tensor + + .. note:: This function is automatically called when `.transpose()` + method is used and `no_overlap=False` + """ + + # initialize convolution + if self.training: # we choose what to do based on optimization + self._choose_initialization(X, type='inverse') + + else: # we always initialize on testing + self._initialize_convolution(X, 'inverse') + + # initialize grid + X = self._grid_transpose.clone().detach() + conv_transposed = self._grid_transpose.clone().detach() + + # list to iterate for calculating nn output + tmp = [i for i in range(self._output_numb_field)] + iterate_conv = [item for item in tmp for _ in range( + self._input_numb_field)] + + for batch_idx, x in enumerate(X): + + # accumulator for the convolution on different batches + accumulator_batch = torch.zeros( + size=(self._grid_transpose.shape[1], + self._grid_transpose.shape[2]), + requires_grad=True, + device=X.device, + dtype=X.dtype).clone() + + for stride_idx, current_stride in enumerate(self._stride): + # indeces of points falling into filter range + indeces = self._index[stride_idx][batch_idx] + + # number of points for each channel + numb_pts_channel = tuple(indeces.sum(dim=-1)) + + # extracting points for each channel + point_stride = x[indeces] + + # if no points to upsample we just skip + if point_stride.nelement() == 0: + continue + + # mapping points in filter domain + map_points_(point_stride[..., :-1], current_stride) + + # input points for kernels + # we split for extracting number of points for each channel + nn_input_pts = point_stride[..., :-1].split(numb_pts_channel) + + # accumulate partial convolution results for each field + res_tmp = [] + + # for each channel field compute transpose convolution + for idx_conv, idx_channel_out in enumerate(iterate_conv): + + # index for input channels + idx_channel_in = idx_conv % self._input_numb_field + + # extract filter + net = self._net[idx_conv] + + # calculate filter value + staked_output = net(nn_input_pts[idx_channel_out]) + + # perform integral for all strides in one field + integral = staked_output * integrals[batch_idx, + idx_channel_in, + stride_idx] + # append results + res_tmp.append(integral.flatten()) + + # computing channel sum + channel_sum = [] + start = 0 + for _ in range(self._output_numb_field): + tmp = res_tmp[start:start + self._input_numb_field] + tmp = torch.vstack(tmp).sum(dim=0) + channel_sum.append(tmp) + start += self._input_numb_field + + # accumulate the results + accumulator_batch[indeces] += torch.hstack(channel_sum) + + # save results of accumulation for each batch + conv_transposed[batch_idx, ..., -1] = accumulator_batch + + return conv_transposed diff --git a/pina/model/layers/integral.py b/pina/model/layers/integral.py new file mode 100644 index 0000000..3269134 --- /dev/null +++ b/pina/model/layers/integral.py @@ -0,0 +1,63 @@ +import torch + + +class Integral(object): + + def __init__(self, param): + """Integral class for continous convolution + + :param param: type of continuous convolution + :type param: string + """ + + if param == 'discrete': + self.make_integral = self.integral_param_disc + elif param == 'continuous': + self.make_integral = self.integral_param_cont + else: + raise TypeError + + def __call__(self, *args, **kwds): + return self.make_integral(*args, **kwds) + + def _prepend_zero(self, x): + """Create bins for performing integral + + :param x: input tensor + :type x: torch.tensor + :return: bins for integrals + :rtype: torch.tensor + """ + return torch.cat((torch.zeros(1, dtype=x.dtype, device=x.device), x)) + + def integral_param_disc(self, x, y, idx): + """Perform discretize integral + with discrete parameters + + :param x: input vector + :type x: torch.tensor + :param y: input vector + :type y: torch.tensor + :param idx: indeces for different strides + :type idx: list + :return: integral + :rtype: torch.tensor + """ + cs_idxes = self._prepend_zero(torch.cumsum(torch.tensor(idx), 0)) + cs = self._prepend_zero(torch.cumsum(x.flatten() * y.flatten(), 0)) + return cs[cs_idxes[1:]] - cs[cs_idxes[:-1]] + + def integral_param_cont(self, x, y, idx): + """Perform discretize integral for continuous convolution + with continuous parameters + + :param x: input vector + :type x: torch.tensor + :param y: input vector + :type y: torch.tensor + :param idx: indeces for different strides + :type idx: list + :return: integral + :rtype: torch.tensor + """ + raise NotImplementedError diff --git a/pina/model/layers/stride.py b/pina/model/layers/stride.py new file mode 100644 index 0000000..02143e8 --- /dev/null +++ b/pina/model/layers/stride.py @@ -0,0 +1,82 @@ +import torch + + +class Stride(object): + + def __init__(self, dict): + """Stride class for continous convolution + + :param param: type of continuous convolution + :type param: string + """ + + self._dict_stride = dict + self._stride_continuous = None + self._stride_discrete = self._create_stride_discrete(dict) + + def _create_stride_discrete(self, my_dict): + """Creating the list for applying the filter + + :param my_dict: Dictionary with the following arguments: + domain size, starting position of the filter, jump size + for the filter and direction of the filter + :type my_dict: dict + :raises IndexError: Values in the dict must have all same length + :raises ValueError: Domain values must be greater than 0 + :raises ValueError: Direction must be either equal to 1, -1 or 0 + :raises IndexError: Direction and jumps must have zero in the same + index + :return: list of positions for the filter + :rtype: list + :Example: + + + >>> stride = {"domain": [4, 4], + "start": [-4, 2], + "jump": [2, 2], + "direction": [1, 1], + } + >>> create_stride(stride) + [[-4.0, 2.0], [-4.0, 4.0], [-2.0, 2.0], [-2.0, 4.0]] + """ + + # we must check boundaries of the input as well + + domain, start, jumps, direction = my_dict.values() + + # checking + + if not all([len(s) == len(domain) for s in my_dict.values()]): + raise IndexError("values in the dict must have all same length") + + if not all(v >= 0 for v in domain): + raise ValueError("domain values must be greater than 0") + + if not all(v == 1 or v == -1 or v == 0 for v in direction): + raise ValueError("direction must be either equal to 1, -1 or 0") + + seq_jumps = [i for i, e in enumerate(jumps) if e == 0] + seq_direction = [i for i, e in enumerate(direction) if e == 0] + + if seq_direction != seq_jumps: + raise IndexError( + "direction and jumps must have zero in the same index") + + if seq_jumps: + for i in seq_jumps: + jumps[i] = domain[i] + direction[i] = 1 + + # creating the stride grid + values_mesh = [torch.arange(0, i, step).float() + for i, step in zip(domain, jumps)] + + values_mesh = [single * dim for single, + dim in zip(values_mesh, direction)] + + mesh = torch.meshgrid(values_mesh) + coordinates_mesh = [x.reshape(-1, 1) for x in mesh] + + stride = torch.cat(coordinates_mesh, dim=1) + torch.tensor(start) + + return stride diff --git a/pina/model/layers/utils_convolution.py b/pina/model/layers/utils_convolution.py new file mode 100644 index 0000000..c1e6a4e --- /dev/null +++ b/pina/model/layers/utils_convolution.py @@ -0,0 +1,48 @@ +import torch + + +def check_point(x, current_stride, dim): + max_stride = current_stride + dim + indeces = torch.logical_and(x[..., :-1] < max_stride, + x[..., :-1] >= current_stride).all(dim=-1) + return indeces + + +def map_points_(x, filter_position): + """Mapping function n dimensional case + + :param x: input data of two dimension + :type x: torch.tensor + :param filter_position: position of the filter + :type dim: list[numeric] + :return: data mapped inplace + :rtype: torch.tensor + """ + x.add_(-filter_position) + + return x + + +def optimizing(f): + """Decorator for calling a function just once + + :param f: python function + :type f: function + """ + + def wrapper(*args, **kwargs): + + if kwargs['type'] == 'forward': + if not wrapper.has_run_inverse: + wrapper.has_run_inverse = True + return f(*args, **kwargs) + + if kwargs['type'] == 'inverse': + if not wrapper.has_run: + wrapper.has_run = True + return f(*args, **kwargs) + + wrapper.has_run_inverse = False + wrapper.has_run = False + + return wrapper diff --git a/tests/test_conv.py b/tests/test_conv.py new file mode 100644 index 0000000..f83023b --- /dev/null +++ b/tests/test_conv.py @@ -0,0 +1,140 @@ +from pina.model.layers import ContinuousConv +import torch + + +def prod(iterable): + p = 1 + for n in iterable: + p *= n + return p + + +def make_grid(x): + def _transform_image(image): + + # extracting image info + channels, dimension = image.size()[0], image.size()[1:] + + # initializing transfomed image + coordinates = torch.zeros( + [channels, prod(dimension), len(dimension) + 1]).to(image.device) + + # creating the n dimensional mesh grid + values_mesh = [torch.arange(0, dim).float().to( + image.device) for dim in dimension] + mesh = torch.meshgrid(values_mesh) + coordinates_mesh = [x.reshape(-1, 1) for x in mesh] + coordinates_mesh.append(0) + + for count, channel in enumerate(image): + coordinates_mesh[-1] = channel.reshape(-1, 1) + coordinates[count] = torch.cat(coordinates_mesh, dim=1) + + return coordinates + + output = [_transform_image(current_image) for current_image in x] + return torch.stack(output).to(x.device) + + +class MLP(torch.nn.Module): + + def __init__(self) -> None: + super().__init__() + self. model = torch.nn.Sequential(torch.nn.Linear(2, 8), + torch.nn.ReLU(), + torch.nn.Linear(8, 8), + torch.nn.ReLU(), + torch.nn.Linear(8, 1)) + + def forward(self, x): + return self.model(x) + + +# INPUTS +channel_input = 2 +channel_output = 6 +batch = 2 +N = 10 +dim = [3, 3] +stride = {"domain": [10, 10], + "start": [0, 0], + "jumps": [3, 3], + "direction": [1, 1.]} +dim_filter = len(dim) +dim_input = (batch, channel_input, 10, dim_filter) +dim_output = (batch, channel_output, 4, dim_filter) +x = torch.rand(dim_input) +x = make_grid(x) + + +def test_constructor(): + model = MLP + + conv = ContinuousConv(channel_input, + channel_output, + dim, + stride, + model=model) + conv = ContinuousConv(channel_input, + channel_output, + dim, + stride, + model=None) + + +def test_forward(): + model = MLP + + # simple forward + conv = ContinuousConv(channel_input, + channel_output, + dim, + stride, + model=model) + conv(x) + + # simple forward with optimization + conv = ContinuousConv(channel_input, + channel_output, + dim, + stride, + model=model, + optimize=True) + conv(x) + + +def test_transpose(): + model = MLP + + # simple transpose + conv = ContinuousConv(channel_input, + channel_output, + dim, + stride, + model=model) + + conv2 = ContinuousConv(channel_output, + channel_input, + dim, + stride, + model=model) + + integrals = conv(x) + conv2.transpose(integrals[..., -1], x) + + stride_no_overlap = {"domain": [10, 10], + "start": [0, 0], + "jumps": dim, + "direction": [1, 1.]} + + # simple transpose with optimization + # conv = ContinuousConv(channel_input, + # channel_output, + # dim, + # stride_no_overlap, + # model=model, + # optimize=True, + # no_overlap=True) + + # integrals = conv(x) + # conv.transpose(integrals[..., -1], x) diff --git a/tutorials/README.md b/tutorials/README.md index 95c722e..370eed3 100644 --- a/tutorials/README.md +++ b/tutorials/README.md @@ -8,5 +8,6 @@ In this folder we collect useful tutorials in order to understand the principles | Tutorial1 [[.ipynb](tutorial1/tutorial.ipynb), [.py](tutorial1/tutorial.py), [.html](http://mathlab.github.io/PINA/_rst/tutorial1/tutorial.html)]| Introduction to PINA features | `SpatialProblem` | | Tutorial2 [[.ipynb](tutorial2/tutorial.ipynb), [.py](tutorial2/tutorial.py), [.html](http://mathlab.github.io/PINA/_rst/tutorial2/tutorial.html)]| Poisson problem on regular domain using extra features | `SpatialProblem` | | Tutorial3 [[.ipynb](tutorial3/tutorial.ipynb), [.py](tutorial3/tutorial.py), [.html](http://mathlab.github.io/PINA/_rst/tutorial3/tutorial.html)]| Wave problem on regular domain using custom pytorch networks. | `SpatialProblem`, `TimeDependentProblem` | +| Tutorial4 [[.ipynb](tutorial4/tutorial.ipynb), [.py](tutorial4/tutorial.py), [.html](http://mathlab.github.io/PINA/_rst/tutorial4/tutorial.html)]| Continuous Convolutional Filter usage. | `None` | diff --git a/tutorials/tutorial4/tutorial.ipynb b/tutorials/tutorial4/tutorial.ipynb new file mode 100644 index 0000000..9de6490 --- /dev/null +++ b/tutorials/tutorial4/tutorial.ipynb @@ -0,0 +1,1031 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "source": [ + "# Tutorial 4: continuous convolutional filter" + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "In this tutorial we will show how to use the Continouous Convolutional Filter, and how to build common Deep Learning architectures with it. The implementation of the filter follows the original work [**A Continuous Convolutional Trainable Filter for Modelling Unstructured Data**](https://arxiv.org/abs/2210.13416) of Coscia Dario, Laura Meneghetti, Nicola Demo, Giovanni Stabile, and Gianluigi Rozza." + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "First of all we import the modules needed for the tutorial, which include:\n", + "\n", + "* `ContinuousConv` class from `pina.model.layers` which implements the continuous convolutional filter\n", + "* `PyTorch` and `Matplotlib` for tensorial operations and visualization respectively" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 1, + "source": [ + "import torch \n", + "import matplotlib.pyplot as plt \n", + "from pina.model.layers import ContinuousConv \n", + "import torchvision # for MNIST dataset\n", + "from pina.model import FeedForward # for building AE and MNIST classification" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "The tutorial is structured as follow: \n", + "* [Continuous filter background](#continuous-filter-background): understand how the convolutional filter works and how to use it.\n", + "* [Building a MNIST Classifier](#building-a-mnist-classifier): show how to build a simple classifier using the MNIST dataset and how to combine a continuous convolutional layer with a feedforward neural network. \n", + "* [Building a Continuous Convolutional Autoencoder](#building-a-continuous-convolutional-autoencoder): show how to use the continuous filter to work with unstructured data for autoencoding and up-sampling." + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "## Continuous filter background" + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "As reported by the authors in the original paper: in contrast to discrete convolution, continuous convolution is mathematically defined as:\n", + "\n", + "$$\n", + " \\mathcal{I}_{\\rm{out}}(\\mathbf{x}) = \\int_{\\mathcal{X}} \\mathcal{I}(\\mathbf{x} + \\mathbf{\\tau}) \\cdot \\mathcal{K}(\\mathbf{\\tau}) d\\mathbf{\\tau},\n", + "$$\n", + "where $\\mathcal{K} : \\mathcal{X} \\rightarrow \\mathbb{R}$ is the *continuous filter* function, and $\\mathcal{I} : \\Omega \\subset \\mathbb{R}^N \\rightarrow \\mathbb{R}$ is the input function. The continuous filter function is approximated using a FeedForward Neural Network, thus trainable during the training phase. The way in which the integral is approximated can be different, currently on **PINA** we approximate it using a simple sum, as suggested by the authors. Thus, given $\\{\\mathbf{x}_i\\}_{i=1}^{n}$ points in $\\mathbb{R}^N$ of the input function mapped on the $\\mathcal{X}$ filter domain, we approximate the above equation as:\n", + "$$\n", + " \\mathcal{I}_{\\rm{out}}(\\mathbf{\\tilde{x}}_i) = \\sum_{{\\mathbf{x}_i}\\in\\mathcal{X}} \\mathcal{I}(\\mathbf{x}_i + \\mathbf{\\tau}) \\cdot \\mathcal{K}(\\mathbf{x}_i),\n", + "$$\n", + "where $\\mathbf{\\tau} \\in \\mathcal{S}$, with $\\mathcal{S}$ the set of available strides, corresponds to the current stride position of the filter, and $\\mathbf{\\tilde{x}}_i$ points are obtained by taking the centroid of the filter position mapped on the $\\Omega$ domain. " + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "We will now try to pratically see how to work with the filter. From the above definition we see that what is needed is:\n", + "1. A domain and a function defined on that domain (the input)\n", + "2. A stride, corresponding to the positions where the filter needs to be $\\rightarrow$ `stride` variable in `ContinuousConv`\n", + "3. The filter rectangular domain $\\rightarrow$ `filter_dim` variable in `ContinuousConv`" + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "### Input function\n", + "\n", + "The input function for the continuous filter is defined as a tensor of shape: $$[B \\times N_{in} \\times N \\times D]$$ where $B$ is the batch_size, $N_{in}$ is the number of input fields, $N$ the number of points in the mesh, $D$ the dimension of the problem. In particular:\n", + "* $D$ is the number of spatial variables + 1. The last column must contain the field value. For example for 2D problems $D=3$ and the tensor will be something like `[first coordinate, second coordinate, field value]`\n", + "* $N_{in}$ represents the number of vectorial function presented. For example a vectorial function $f = [f_1, f_2]$ will have $N_{in}=2$ \n", + "\n", + "Let's see an example to clear the ideas. We will be verbose to explain in details the input form. We wish to create the function:\n", + "$$\n", + "f(x, y) = [\\sin(\\pi x) \\sin(\\pi y), -\\sin(\\pi x) \\sin(\\pi y)] \\quad (x,y)\\in[0,1]\\times[0,1]\n", + "$$\n", + "\n", + "using a batch size of one." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 2, + "source": [ + "# batch size fixed to 1\n", + "batch_size = 1\n", + "\n", + "# points in the mesh fixed to 200\n", + "N = 200\n", + "\n", + "# vectorial 2 dimensional function, number_input_fileds=2\n", + "number_input_fileds = 2\n", + "\n", + "# 2 dimensional spatial variables, D = 2 + 1 = 3\n", + "D = 3\n", + "\n", + "# create the function f domain as random 2d points in [0, 1]\n", + "domain = torch.rand(size=(batch_size, number_input_fileds, N, D-1))\n", + "print(f\"Domain has shape: {domain.shape}\")\n", + "\n", + "# create the functions\n", + "pi = torch.acos(torch.tensor([-1.])) # pi value\n", + "f1 = torch.sin(pi * domain[:, 0, :, 0]) * torch.sin(pi * domain[:, 0, :, 1])\n", + "f2 = - torch.sin(pi * domain[:, 1, :, 0]) * torch.sin(pi * domain[:, 1, :, 1])\n", + "\n", + "# stacking the input domain and field values\n", + "data = torch.empty(size=(batch_size, number_input_fileds, N, D))\n", + "data[..., :-1] = domain # copy the domain\n", + "data[:, 0, :, -1] = f1 # copy first field value\n", + "data[:, 1, :, -1] = f1 # copy second field value\n", + "print(f\"Filter input data has shape: {data.shape}\")" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Domain has shape: torch.Size([1, 2, 200, 2])\n", + "Filter input data has shape: torch.Size([1, 2, 200, 3])\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "### Stride\n", + "\n", + "The stride is passed as a dictionary `stride` which tells the filter where to go. Here is an example for the $[0,1]\\times[0,5]$ domain:\n", + "\n", + "```python\n", + "# stride definition\n", + "stride = {\"domain\": [1, 5],\n", + " \"start\": [0, 0],\n", + " \"jump\": [0.1, 0.3],\n", + " \"direction\": [1, 1],\n", + " }\n", + "```\n", + "This tells the filter:\n", + "1. `domain`: square domain (the only implemented) $[0,1]\\times[0,5]$. The minimum value is always zero, while the maximum is specified by the user\n", + "2. `start`: start position of the filter, coordinate $(0, 0)$\n", + "3. `jump`: the jumps of the centroid of the filter to the next position $(0.1, 0.3)$\n", + "4. `direction`: the directions of the jump, with `1 = right`, `0 = no jump`,`-1 = left` with respect to the current position\n", + "\n", + "**Note**\n", + "\n", + "We are planning to release the possibility to directly pass a list of possible strides!" + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "### Filter definition\n", + "\n", + "Having defined all the previous blocks we are able to construct the continuous filter.\n", + "\n", + "Suppose we would like to get an ouput with only one field, and let us fix the filter dimension to be $[0.1, 0.1]$." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 4, + "source": [ + "# filter dim\n", + "filter_dim = [0.1, 0.1]\n", + "\n", + "# stride\n", + "stride = {\"domain\": [1, 1],\n", + " \"start\": [0, 0],\n", + " \"jump\": [0.08, 0.08],\n", + " \"direction\": [1, 1],\n", + " }\n", + "\n", + "# creating the filter \n", + "cConv = ContinuousConv(input_numb_field=number_input_fileds,\n", + " output_numb_field=1,\n", + " filter_dim=filter_dim,\n", + " stride=stride)" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "That's it! In just one line of code we have created the continuous convolutional filter. By default the `pina.model.FeedForward` neural network is intitialised, more on the [documentation](https://mathlab.github.io/PINA/_rst/fnn.html). In case the mesh doesn't change during training we can set the `optimize` flag equals to `True`, to exploit optimizations for finding the points to convolve." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 5, + "source": [ + "# creating the filter + optimization\n", + "cConv = ContinuousConv(input_numb_field=number_input_fileds,\n", + " output_numb_field=1,\n", + " filter_dim=filter_dim,\n", + " stride=stride,\n", + " optimize=True)\n" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's try to do a forward pass" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 6, + "source": [ + "print(f\"Filter input data has shape: {data.shape}\")\n", + "\n", + "#input to the filter\n", + "output = cConv(data)\n", + "\n", + "print(f\"Filter output data has shape: {output.shape}\")" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Filter input data has shape: torch.Size([1, 2, 200, 3])\n", + "Filter output data has shape: torch.Size([1, 1, 169, 3])\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "If we don't want to use the default `FeedForward` neural network, we can pass a specified torch model in the `model` keyword as follow: \n" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 7, + "source": [ + "class SimpleKernel(torch.nn.Module):\n", + " def __init__(self) -> None:\n", + " super().__init__()\n", + " self. model = torch.nn.Sequential(\n", + " torch.nn.Linear(2, 20),\n", + " torch.nn.ReLU(),\n", + " torch.nn.Linear(20, 20),\n", + " torch.nn.ReLU(),\n", + " torch.nn.Linear(20, 1))\n", + "\n", + " def forward(self, x):\n", + " return self.model(x)\n", + "\n", + "\n", + "cConv = ContinuousConv(input_numb_field=number_input_fileds,\n", + " output_numb_field=1,\n", + " filter_dim=filter_dim,\n", + " stride=stride,\n", + " optimize=True,\n", + " model=SimpleKernel)\n" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Notice that we pass the class and not an already built object!" + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "## Building a MNIST Classifier\n", + "\n", + "Let's see how we can build a MNIST classifier using a continuous convolutional filter. We will use the MNIST dataset from PyTorch. In order to keep small training times we use only 6000 samples for training and 1000 samples for testing." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 9, + "source": [ + "from torch.utils.data import DataLoader, SubsetRandomSampler\n", + "\n", + "numb_training = 6000 # get just 6000 images for training\n", + "numb_testing= 1000 # get just 1000 images for training\n", + "seed = 111 # for reproducibility\n", + "batch_size = 8 # setting batch size\n", + "\n", + "# setting the seed\n", + "torch.manual_seed(seed)\n", + "\n", + "# downloading the dataset\n", + "train_data = torchvision.datasets.MNIST('./data/', train=True, download=True,\n", + " transform=torchvision.transforms.Compose([\n", + " torchvision.transforms.ToTensor(),\n", + " torchvision.transforms.Normalize(\n", + " (0.1307,), (0.3081,))\n", + " ]))\n", + "subsample_train_indices = torch.randperm(len(train_data))[:numb_training]\n", + "train_loader = DataLoader(train_data, batch_size=batch_size,\n", + " sampler=SubsetRandomSampler(subsample_train_indices))\n", + "\n", + "test_data = torchvision.datasets.MNIST('./data/', train=False, download=True,\n", + " transform=torchvision.transforms.Compose([\n", + " torchvision.transforms.ToTensor(),\n", + " torchvision.transforms.Normalize(\n", + " (0.1307,), (0.3081,))\n", + " ]))\n", + "subsample_test_indices = torch.randperm(len(train_data))[:numb_testing]\n", + "test_loader = DataLoader(train_data, batch_size=batch_size,\n", + " sampler=SubsetRandomSampler(subsample_train_indices))" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's now build a simple classifier. The MNIST dataset is composed by vectors of shape `[batch, 1, 28, 28]`, but we can image them as one field functions where the pixels $ij$ are the coordinate $x=i, y=j$ in a $[0, 27]\\times[0,27]$ domain, and the pixels value are the field values. We just need a function to transform the regular tensor in a tensor compatible for the continuous filter:" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 10, + "source": [ + "def transform_input(x):\n", + " batch_size = x.shape[0]\n", + " dim_grid = tuple(x.shape[:-3:-1])\n", + "\n", + " # creating the n dimensional mesh grid for a single channel image\n", + " values_mesh = [torch.arange(0, dim).float() for dim in dim_grid]\n", + " mesh = torch.meshgrid(values_mesh)\n", + " coordinates_mesh = [x.reshape(-1, 1) for x in mesh]\n", + " coordinates = torch.cat(coordinates_mesh, dim=1).unsqueeze(\n", + " 0).repeat((batch_size, 1, 1)).unsqueeze(1)\n", + "\n", + " return torch.cat((coordinates, x.flatten(2).unsqueeze(-1)), dim=-1)\n", + "\n", + "\n", + "# let's try it out\n", + "image, s = next(iter(train_loader))\n", + "print(f\"Original MNIST image shape: {image.shape}\")\n", + "\n", + "image_transformed = transform_input(image)\n", + "print(f\"Transformed MNIST image shape: {image_transformed.shape}\")\n" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Original MNIST image shape: torch.Size([8, 1, 28, 28])\n", + "Transformed MNIST image shape: torch.Size([8, 1, 784, 3])\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "We can now build a simple classifier! We will use just one convolutional filter followed by a feedforward neural network" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 19, + "source": [ + "# setting the seed\n", + "torch.manual_seed(seed)\n", + "\n", + "class ContinuousClassifier(torch.nn.Module):\n", + " def __init__(self):\n", + " super().__init__()\n", + "\n", + " # number of classes for classification\n", + " numb_class = 10\n", + "\n", + " # convolutional block\n", + " self.convolution = ContinuousConv(input_numb_field=1,\n", + " output_numb_field=4,\n", + " stride={\"domain\": [27, 27],\n", + " \"start\": [0, 0],\n", + " \"jumps\": [4, 4],\n", + " \"direction\": [1, 1.],\n", + " },\n", + " filter_dim=[4, 4],\n", + " optimize=True)\n", + " # feedforward net\n", + " self.nn = FeedForward(input_variables=196,\n", + " output_variables=numb_class,\n", + " layers=[120, 64],\n", + " func=torch.nn.ReLU)\n", + "\n", + " def forward(self, x):\n", + " # transform input + convolution\n", + " x = transform_input(x)\n", + " x = self.convolution(x)\n", + " # feed forward classification\n", + " return self.nn(x[..., -1].flatten(1))\n", + "\n", + "\n", + "net = ContinuousClassifier()" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's try to train it using a simple pytorch training loop. We train for juts 1 epoch using Adam optimizer with a $0.001$ learning rate." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 20, + "source": [ + "# setting the seed\n", + "torch.manual_seed(seed)\n", + "\n", + "# optimizer and loss function\n", + "optimizer = torch.optim.Adam(net.parameters(), lr=0.001)\n", + "criterion = torch.nn.CrossEntropyLoss()\n", + "\n", + "for epoch in range(1): # loop over the dataset multiple times\n", + "\n", + " running_loss = 0.0\n", + " for i, data in enumerate(train_loader, 0):\n", + " # get the inputs; data is a list of [inputs, labels]\n", + " inputs, labels = data\n", + "\n", + " # zero the parameter gradients\n", + " optimizer.zero_grad()\n", + "\n", + " # forward + backward + optimize\n", + " outputs = net(inputs)\n", + " loss = criterion(outputs, labels)\n", + " loss.backward()\n", + " optimizer.step()\n", + "\n", + " # print statistics\n", + " running_loss += loss.item()\n", + " if i % 50 == 49: \n", + " print(\n", + " f'epoch [{i + 1}/{numb_training//batch_size}] loss[{running_loss / 500:.3f}]')\n", + " running_loss = 0.0\n" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "epoch [50/750] loss[0.148]\n", + "epoch [100/750] loss[0.072]\n", + "epoch [150/750] loss[0.063]\n", + "epoch [200/750] loss[0.053]\n", + "epoch [250/750] loss[0.041]\n", + "epoch [300/750] loss[0.048]\n", + "epoch [350/750] loss[0.054]\n", + "epoch [400/750] loss[0.048]\n", + "epoch [450/750] loss[0.047]\n", + "epoch [500/750] loss[0.035]\n", + "epoch [550/750] loss[0.036]\n", + "epoch [600/750] loss[0.041]\n", + "epoch [650/750] loss[0.030]\n", + "epoch [700/750] loss[0.040]\n", + "epoch [750/750] loss[0.040]\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's see the performance on the train set!" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 21, + "source": [ + "correct = 0\n", + "total = 0\n", + "with torch.no_grad():\n", + " for data in test_loader:\n", + " images, labels = data\n", + " # calculate outputs by running images through the network\n", + " outputs = net(images)\n", + " # the class with the highest energy is what we choose as prediction\n", + " _, predicted = torch.max(outputs.data, 1)\n", + " total += labels.size(0)\n", + " correct += (predicted == labels).sum().item()\n", + "\n", + "print(\n", + " f'Accuracy of the network on the 1000 test images: {(correct / total):.3%}')\n" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Accuracy of the network on the 1000 test images: 93.017%\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "As we can see we have very good performance for having traing only for 1 epoch! Nevertheless, we are still using structured data... Let's see how we can build an autoencoder for unstructured data now." + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "## Building a Continuous Convolutional Autoencoder\n", + "\n", + "Just as toy problem, we will now build an autoencoder for the following function $f(x,y)=\\sin(\\pi x)\\sin(\\pi y)$ on the unit circle domain centered in $(0.5, 0.5)$. We will also see the ability to up-sample (once trained) the results without retraining. Let's first create the input and visualize it, we will use firstly a mesh of $100$ points." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 22, + "source": [ + "# create inputs\n", + "def circle_grid(N=100):\n", + " \"\"\"Generate points withing a unit 2D circle centered in (0.5, 0.5)\n", + "\n", + " :param N: number of points\n", + " :type N: float\n", + " :return: [x, y] array of points\n", + " :rtype: torch.tensor\n", + " \"\"\"\n", + "\n", + " PI = torch.acos(torch.zeros(1)).item() * 2\n", + " R = 0.5\n", + " centerX = 0.5\n", + " centerY = 0.5\n", + "\n", + " r = R * torch.sqrt(torch.rand(N))\n", + " theta = torch.rand(N) * 2 * PI\n", + "\n", + " x = centerX + r * torch.cos(theta)\n", + " y = centerY + r * torch.sin(theta)\n", + "\n", + " return torch.stack([x, y]).T\n", + "\n", + "# create the grid\n", + "grid = circle_grid(500)\n", + "\n", + "# create input\n", + "input_data = torch.empty(size=(1, 1, grid.shape[0], 3))\n", + "input_data[0, 0, :, :-1] = grid\n", + "input_data[0, 0, :, -1] = torch.sin(pi * grid[:, 0]) * torch.sin(pi * grid[:, 1])\n", + "\n", + "# visualize data\n", + "plt.title(\"Training sample with 500 points\")\n", + "plt.scatter(grid[:, 0], grid[:, 1], c=input_data[0, 0, :, -1])\n", + "plt.colorbar()\n", + "plt.show()\n" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWMAAAEICAYAAACK8ZV4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAAsTAAALEwEAmpwYAADxu0lEQVR4nOydd3wUZfrAv+/MbEtvkAABQu9NEKQpiiio2HvvXe/0mufp2U+989Sz9947iIqgNOm9ht7TIL1tm/L+/thN2ewGEggS/e3Xz0p2dso7s7PPPO9ThZSSKFGiRIlydFGO9gCiRIkSJUpUGEeJEiVKqyAqjKNEiRKlFRAVxlGiRInSCogK4yhRokRpBUSFcZQoUaK0AqLC+AgihPhBCHFVS6/7W0QIIYUQ3Y/g/u8VQrxxgM+vFkLMP1LHPxoIIV4RQtx/tMcRpWWICuMGCCGq6r0sIYSn3vvLmrMvKeUkKeW7Lb1ulHCklP+SUl4PIITICgp/7VD3J4SYI4Tw1vvuNzf4/FIhxG4hRLUQ4hshREq9z1KEEF8HP9sthLj00M+scaSUN0spH2nKukKId4QQjx6JcURpGaLCuAFSyriaF7AHmFxv2Yc16x3ODz3Kb4bb6333vWoWCiH6Aa8CVwDpgBt4qd52LwL+4GeXAS8Ht4kSpVGiwriJCCHGCSFyhBB/E0IUAG8LIZKFENOEEIVCiNLg35n1tpkjhKjR1q4WQswXQjwVXHenEGLSIa7bRQgxTwhRKYT4SQjxohDig0bGnRYcV5kQokQI8YsQQgl+do8QYntwP9lCiHPqbXe1EGKBEOKZ4LY7hBCjgsv3CiH21zerBDWvV4QQM4P7myuE6NzImBzBc9sjhNgX3M7VyLq7hRBDg39fFtR4+wXfXyeE+Cb494P1rsG84L9lQa12ZL39RbymzeQy4Fsp5TwpZRVwP3CuECJeCBELnAfcL6WsklLOB6YSENyRzu9BIcQXQohPg9dtpRBiUL3P+wTvjTIhxAYhxJn1PqvVduvdn38Kfjf5Qohrgp/dGBzzX4PX49vg8r8JIXKDx90shBh/iNcjSgsQFcbNIwNIAToDNxK4fm8H33cCPMALB9h+BLAZSAP+DbwphBCHsO5HwFIgFXiQRn7oQf4E5ABtCGhq9wI1OfDbgbFAIvAQ8IEQol2DMawNHucj4BPgWKA7cDnwghAirt76lwGPBMe8GviQyDwB9AQGB/fVAfhnI+vOBcYF/z4B2AEcX+/93Ajb1HyeFNRqF9U7n6Zef4DHhRBFwYfSuHrL+wFrat5IKbcT0IR7Bl+GlHJLvfXXBLdpjLOAzwncWx8B3wghbEIIG/AtMANoC9wBfCiE6NXIfjIIfJcdgOuAF4UQyVLK1wh8F/8OXo/JwX3cDhwrpYwHTgV2HWCMUY4wUWHcPCzgASmlT0rpkVIWSym/lFK6pZSVwGMEBERj7JZSvi6lNIF3gXYEBGST1xVCdCIgEP8ppfTX07waQw9u21lKqUspf5HBgiRSys+llHlSSktK+SmwFRheb9udUsq3g2P4FOgIPBw8/xkEBFB9p9x3QW3RB/wDGCmE6Fh/MEHhdyNwl5SyJHjd/gVc3Mj451J3TccCj9d735gwbozmXP+/AV0JCLbXgG+FEN2Cn8UB5Q3WLwfig59VNPJZY6yQUn4hpdSBpwEncFzwFQc8EfyuZwHTgEsa2Y9O4PvRpZTfA1VAY4LbBBxAXyGETUq5K/hQiXKUiArj5lEopfTWvBFCxAghXg1OpSsITI+ThBBqI9sX1PwhpXQH/4xr5rrtgZJ6ywD2HmDM/wG2ATOCpoZ76o3/SiHE6uAUuAzoT0BrrGFfvb89wbE0XFZ//LXjCE7fS4LjrU8bIAZYUe+404PLIzEXGBvU2FXgM2C0ECKLgBa4utEzD6fJ119KuURKWRl88LwLLABOC35cBSQ02CQBqDzIZ41R/7pZBGYy7YOvvcFlNewm8ICIRLGU0qj33k3j57cN+COBmdV+IcQnQoiG31WUX5GoMG4eDUvc/YmA5jFCSplA3fT4QFPfwyUfSBFCxNRb1rGxlYMC5U9Syq7AmcDdQojxQXvu6wSmqqlSyiRgPYc39tpxBM0XKUBeg3WKCAjxflLKpOArMegwjTT+bQSEyh3APCllBQGheiMwv4Ggqt3sMM6hMSR112YDUN+u25WAlrkl+NKEED3qbTsouE1j1L9uCpBJ4LrlAR1rbPxBOgG5hzj+0AVSfiSlHEPAzCaBJw9hv1FaiKgwPjziCQiWMhEIbXrgSB9QSrkbWA48KISwB51TkxtbXwhxhhCie9A8UE5gemoBsQR+gIXB9a4hoBkfDqcJIcYIIewEbMeLpZQhWntQeL4OPCOEaBs8dgchxKkH2O9cAg+NGpPEnAbvG1JI4By7HspJCCGShBCnCiGcQghNBEIajyegwUPA/jpZCDE26LB7GPgq+OCrBr4CHhZCxAohRhOwCb9/gEMOFUKcKwIROn8EfMBiYAmBB9FfgzbkcQS+608O4bT2Ue96CCF6CSFOEkI4AC+B+zjSgy3Kr0RUGB8ezwIuAtreYup+rEeay4CRQDHwKAF7rq+RdXsAPxGYPi8CXpJSzpZSZgP/DS7bBwwgMBU/HD4i8EAqAYYScPJF4m8ETCeLg+adn2jctgkBoRtPXZREw/chBE0QjwELgqaQ45p5HjYC17WQwHd7B3B2jVNOSrkBuJmAUN4fHMut9ba/lcB9sR/4GLgluE1jTAEuAkoJOGPPDdp9/QSE76TgOF4CrpRSbmrm+QC8ScA+XBaMQHEQcKQWEZhptAX+fgj7jdJCiGhx+d8+QohPgU1SyiOumR9gDO8AOVLK+47WGH6LCCEeBLpLKRt7cEX5f0JUM/4NIoQ4VgjRTQihCCEmEpgGf3OUhxUlSpTDIJpF9tskg4BdMpWA5/0WKeWqozukKFGiHA5RM0WUKFGitAKiZoooUaJEaQUcNTNFWlqazMrKOlqHjxIlym+IFStWFEkpG0sMahKnnhgri0vMph1vre9HKeXEwzleczlqwjgrK4vly5cfrcNHiRLlN4QQYvfh7qOoxGTJj5kHXxGwtduedvC1WpaoAy9KlCj/T5CYERM2WwdRYRwlSpT/F0jAOiKZ8i1DVBhHiRLl/w1WK874jgrjKFGi/L9AItFbsZnioKFtQoi3gp0D1jfyuRBCPCeE2CaEWCuEOKblhxnl/yMVxZWsnZdN/s59B185SpSDIAET2aTX0aApmvE7BLpXvNfI55MIFKPpQaCTwsvBf6NEOSSklLz+t/eZ8sJ0bA4buk+n/9g+PPDFn5FSMu2VmSz8djlJbRM5545JDB4XbS8XpWn8pm3GUsp5wULejXEW8F6we8TiYPnBdlLK/JYaZJTfJ54qL5pdw2YPvQ2nvzWLb1+egd+r4/fqAKybl82TVz7H7o15FOWW1C5f+dM6rnnkIs6987Sw/UeJUh8JmK0447glMvA6ENppIodGOhEIIW4UQiwXQiwvLCxsgUNH+S2SvWQrN424j/M63sa57W/hPze+hrvSU/v5F09/i7c6tCKo7jNYPG0lhfUEMYDP7ePt+z6husJNlCgHw2ri62jwq6ZDSylfk1IOk1IOa9PmsJJpovxGyduxn3vP+i+7N+ZimRa632De18t46JLna9epLKmKuK2UEr/HH7ZctalsXbnziI05yu8D2UR78dGyGbeEMM4ltO1PJofWFibK/wO+eXkGul8PWab7DNYv2sJ/7nyXxTPWMWT8ABQlvPuTzW6LuE/LtEhIaayVYNPx+3R+/nwJ//vzR3z2wgzKig7Uti7Kbw0pQW/i62jQEqFtU4HbhRCfEHDclUftxUeXgj3FvPiPz1j1y2Y0TWXc2UO54YFzKC+uYu2ibcQnxXDsiX2xOyMLtyPJ7o15mEb4RNA0TH7+bAmLZmygQ+cUYhJi8Lp9GH4DoQjsThtXPHABHzz6FT53nXasqArpndvQZUCnwxpXVbmbP57+H4rzy/C6/dgdNj55djqPf/EHeg3ufFj7jtJaEJhHtD3l4XFQYSyE+BgYB6QJIXIItNWxAUgpXwG+J9A1t6Zx5DVHarBRDk5VuZs/nPEUlWVupCUxDYufv1rG4p824K7yoagCRVFQNYXHP7qV7gMa7WXaKHu372faR4vYn1fGMaN7MOHcYThj7E3ats/wbmQv3oruN0I/kICq4qn2sXdHERf84wKqCopZOzebDj3aceGfz6T7kC644ly8fs+HqJqKaVi069KWR7/9G4EWf4fOx8/+wL69JRjBcfl9OvjgP7e9w+vz/3nY+49y9JGA1Xr9d02KprjkIJ9L4LYWG1GUw+KnL5bi8+jIened4TcpL6oEJdQq9cA1r/P+0gdRlKZbq5bM3sjjf/gQQzcxTYtVC7fy9Tvzee6rO4hLcB10+zNvOplpb8zCMMyQMWLTEMFx+Lw6S37eyHPf/zl8+1tOZcIVJ7B11U4SUuLI6tf8h0kkfpm6qlYQ16cwt4TignLS2iW1yHGiHF1as2YcrWf8O2P7hhx8EZxcANg0cDpqG857qn1sWbOnyfs2TYun7/kMn1fHNAOmBp9Hp6ignK/eitgbNIyU9ESem/MAo04/BmeMI/CAcNjBFSrINbva6D5ccU4Gju3TYoIYCAuvq0FKUBRB0b5yvI1d1yi/CQJJH6JJr6NBVBj/zujWLxOHqxGTgaqAIsDhAEAIgd8Xrg02xt7t+9EjrK/7DeZPX9fk/bTv2pb7P7ydr/Nfpl2/zginM8QM4HTZmXTZ6CbvryWYePloHA1s6IqqkNa1LTed9wLXnv40F459jOce/gZdb/o1i9J6kIAulSa9jgZRYfw74+Tzh+Nw2hANoxEUJfASIiCQhQAkvYdkNXnfrlhHrUbckJh4Z7PHKoTgn29cR1xSDK44B3aHhsNlY8SE/ow/b1iz93c4nHPjSQwa0wuHy4ZqU0EITE0lr9hNZbkbv8/A7zP4eepqXnzs2191bFFaBonARGnS62gQFca/M+ISY3h22p845vjedQttGsQ4gwI4oCGIBBcdRvbkpf/NYOWyHbz8vxlcecEL3HrNG/z84zoi9UZM75BM5x4ZYWFnTpeds65snibr9xu88uJP3HnXR5SlxJM4uDOTbxnPM1Pu4p4Xr2qWHbsl0GwqD71/C6MvGgWJcVjJ8ciEWBqGnPp8Oj9/uxqP2xd5R1FaNZYUTXodDaJV236HtOucxqMf3MLHr8zik1dmh5giJGAluFBjHWzcmM+mTfl8981KFEVgBR1qzz75PVs3F3DznRPC9n3/i1dwz5WvU1pUgRAC3W8y4fxhjDtjcLPG+PjD37Bk8Xb8QafZ3twyppSs5/RLRh3yeR8um7JzmT1vM4bDhpBABIcegKoKyoqrccU4ft0BRjksamzGrZWoMP4dM/nSkUz/fBmlRVW1oWRqnAPhtNUK6BoF2KoX2eD16kz9ajkXXT6S5HrJFD6fTnGFh0feu56y/HJKCivpNagjbTKSmjWufQXlIYK4Bl03+fLzJdz+h1MP4WwPndLSav7+t0/ZvasIv6aAZkf4TRTdBCnDfr6KqtAmI/FXHWOUlkBgHiV7cFOICuPfMXEJLl746k6mvL+AJXM2kpQSR4lusmVzwUG3tds0tm4uYPjI7gB8+/1qXnptFkIRGIZFj25tefSBc0lOim32uHL2FmOzqWHC2DAstm1p2XKZ8xdt5Y135pG/r5zM9snceM0JjDi2a8g6jzz4NTu27w/Yw2tMOXYVaWoIT2i2oMNp4+o7J6DZVEzTYtHCrWzcmEe7dkmceFJfYmOj2nJrJdDpIyqMoxwl4hNdXH77yVx++8kAPPbPr5okjE3TIjUtHoDVa/fwwqs/46tn7ti0pYB7H/iSl/93ZbPHlNkxFV0P79KraQo9eqY3e3+NMXveJp54+vvacW/fWcg/H/uGB/5+JqNGBB4yxcVVZG/MDXdMCoHl0FB0C4EkNsZBevskLrnpRMac3I/qah9/uON9CgrK8Hh0nE4br78+h2f/dzldukTrrrRGpBT4ZeMhk0eb1vuYiNJiWJak2u3DsiQTI9h2G7rqVE0hs1Mq3XoEBOPnXy0LEcQQENY7dhWSk1vS7PGkZyRy3Kju2B2huoDNpnHehS1XCvuVN+eEjdvnM3jlzTm176uqvGhqIz8DIejWpx3/fvM6vlhwHy9+fjtjTg7UTv7g/QXk5JTgCWrOXq9OVaWXfz02tcXGH6XlsRBNeh0Noprx75wp01fzxgfzqXb7cDhsnH5yf5QYDdMdKqQkgYg3zabSp18m9z1ybu1nhUWVtQK7/m2qqQqlZW4yO6Q0e1x/v/9s3n5jDtOmrsLr8dO3fya3/+EUMloo0800Lfbtr4j4WW5eWe3fmR1SsNm0WqFag6YpnH3OMG657eSI+5g1Kzuidr93TzFlZW6SkmIOffBRjggBB17r1T+jwvh3zIw5G3jxrdl4g9qh4fYxZfoaDJcKAoQpQYBUAyK2V892PPLQeaSkBpx2Xr/OM+/OJru0BDNJBQtsbgvFCIhmw7To1rXtIY3NZlO58Zbx3HjL+BY403BUVSEpKYaysvA6x2mpdU5JVVO4+8+TePyxqei6iWVJ7HaN+HgnF186stH9R6oqBwGHaLSMRWsl6sCLcpR46+OFtYK4Bp/fQLMpaBJ0vc5O6nBoXHPN8bWCGOChF75n4eqdmFZQwqigxynYKk1iNBvXXTWWmMay/VoBV106ilffnBNyDRwOjWuuGBOy3tjje9OuXTJffrmUffnlDB3WhTPPPob4+MZrbZx66gA+/WRJiBNSCEH3HukkJka14tZI1IH3/wQpJcuX7WDWzPUoisKEiQMZNLjTUa32VVgcuR6vaUlGD+/OkqXb0VQFVVW4+YYTGTa0S+06+0sqWbh6J/6GU3EhSGqXwH03n8rwYV1pzZx9xhBM0+LdjxbirvYRH+/i2ivHMPHk/mHrdu+Rzt/umdzkfV9y6UhWrdzNtu37MHQTm03D6bLxj/vObMlTiNLCmEcpoaMpRIVxCyCl5L9PfsecWdl4vTpCwNzZGznjrGO4uRGb469Bpw4pbN8V3t4qLSWWR/55DhUVHsorPLTLSETTQr3M+YUVgfCzCHbR5DZxrV4QQ0BTPf/sYZx31lC8Ph2nw9ZiD0eHw8azz13O2jV72bw5n/T0BEaO6oG9kYJDLYWUEsOw0DQlWtazmUgEumy9Iq/1juw3xKaNecyelY0v2JtNymDixDcrmHT6YDpnpR1w+w2b8/h25jo8Xj/jRvWkU4cUPD6dHlltcRzGj/uWq0/g749+jW7UCVQpIKt/Blf962MM0+L0kX04P31g3edSIoQgq31KRAeVqggSklxMWbyBkb070zbp8DtsHGmEELicLW9OEUIwaHAnBg0+vML2TeX7H9bw5pvzKCurJjExhquvGsPkyUOiQrmJRB14/w9YsmhboBh5A6QlWbZk+wGF8QdfLuGdzxfh9xtYEn5esgUhBE6nDaTkLzdO4JSxfQ5pXJ06peGLU5CVJooJUgFvksqcjbtq19mZX8KMZVvo3bUtUxdn4/MbDO7Wnr9ffBJnnTSQb2evw1vPLmpKyfK8fJZ/VoBpWdx2xiiuOvnXLerz/5GZM9fz/PM/4QveZ2Vlbl5+JZCEM/mMIUd5dL8NJKJVmyla72PiN4Qrxh42zYdA2qzT1Xhro+LSat7+dCE+X0AQSzWgmVpS4vb4cXt1nnhlBlt27j+kcU1ftBHLDv5EFW+KijdJpWHMu0832LCrgC/nrcPj07GkZOW2XK5+6lMuPfNYbr54DOmp8cQ4beBQ8CYIPKaBx6/jN0xe/m4RG/e2bNYcBK5DTbGi4io3c7J3sG5PQcQCRq2NskoPu/JLMBqpcHcovP3OL7WCuAafz+Ddd+e32DH+P2ChNOl1NIhqxi3AiSf1491IxdWlZOwJvcOXB1mxdjeqqkB9c0CDKadumHz142ruufmUZo+r0u3FqN9vrhGlwLIkUpchd4NfN/ls3hruPHsMF00ayoyVW3jowxnIBgLBb5h8uySbPh1bJnNud1EZj3z9M0u270VVBF3TUthVWIpdVbGkpG1iHK/feB7tkuJb5HgtSZXbxz9f/4ElGwLfq01V+NOlJ3LaqL6Hve/Cwsgx0yUl1ViWbDTULkodUtKqQ9ta78h+Q7RNT+Cv95yBw6ERE2MnJsaO02njnw+fd8AwJ5fTXmfvO4CgLGqkdf3BGDkgC6ejaU1HG87edNNk4946jVw3Ijcwt6QMMWMcDuVuL5e++DFLtu/FkhLdsNiSX4TfMKny+XH7dfYUlXHH21Na5HgtzT0vT2Pxht34DROPT6fC7ePx935i1Zacw953u0aSYdq2TaituLdo8TYef/Jbnn52OtnZ0QbtDQk48NQmvY4GUc24BSgtq6bvwI58+vUfWLl8F6oqOGZYF1wHicEdPiQLpUYYNzL7djo0Rg89tMiFob07MrxvJ5Zm78Hj0xs9hiRgT66PXVPo07EuoeO43p0iTrtddhsnD+l5SONryNfLN+DRdQxVggJKhJLBlpTsKixlT1EZndKSWuS4LcG+kkpWbckJcZYCeP0G7/+wnCE9Mw9r/zfdeCKPPjY1JL3b4dC44YZxSCl56JGvWbZ8ZzCaRzDzp/VcftkoLjuKJUlbI1EH3u+ITVvy+ebbVZSUVjFwQEcWLNvO1h37EUKQmhLLfXefTv8+HZq0L4dd4z/3n8dfHv0SaUm8polu1Qk8u00lPS2BSeP6HdJYhRA8eftkZi3fyg+LNlJS5SF7334Mr1mnCguwXIAKfnvgPRI0Q3DxuMG1+0pNiOWus8fy7JT56IaJJSUuu40TBnRhZO+WiSbIzt2HBzMwXzvArFtVFCo83hY5ZktRVF6NTY0cCphfHNnE0BxGj+7JffedxRtvzCEvr4z09ESuv+54jj++N8uW76wVxBCwt/t8Bu9/sIBTJgygTVrrM+kcDSRHr3B8U4gK42bw3fS1PPfyTPz+gDBaumZ3iLKZX1DOn//5OR+8cn1Iyu2B6N+7PVPfvpWV6wLaq2FaTJu9nqpqL+OO68k5pw5usqkhEqqiMGF4LyYM78VTU+eysrgAYgXUyHwFFFVgBbVRAASYMbAmJ58JST1q93XJuCEM69GRb5dk4/HrjB/cnRG9mp7Y4tF1npmzkK/XZqObJuO6d+Gek48nIyEgLIQqAkK4ZrKgAka4XBYCerVrXZXRurRLwbDCZw5CQFG1m6e+nMsV44eSXi8UUDdMZq3cysrNOaSnxDN5dD/aHCBUcPSoHowe1SNs+fwFm2sFcX0URWH58p1Mmjgw7LP/r0Q1498BHq+f51/+qW6aKAIaSEOHm2lafDdjLVc1Y3pos6mMOKYu++3kMY07/Q6HjKR4HDYVn25CPbOYZUnMhlEWhslzPy1kQv/QH3+PDmncfe7xzT62lJLrP/6atXkF+MyA9jh901aW7slhxi3XEOewE+uyh0heqYEwg/UeAEUI7JrKA+eOxxYheuXXRkrJsh05LNqym8QYF5dMGMInM1eF2NAtCYUeN5/OW82UxRv46K+X0rFNEm6vn2sf/4TconI8Ph27pvL290t57g/nNNukEeOyh3RqqUFRBK4DRPP8f0MCVit24EWFcRPZvKUAVT24BujXTfIKyo78gA6BM4b24cUfFuGjbiotREDYRbpH88oip1MfCuvy97G+YF+tIIaA/bfa7+erNevJSElgv7cam6qg19imBVgOsKPQMzWVPu3bcunowfTucGjFiVoS07L447vfsnjbXjz+gDBVhOCaSUNZvm432bv3oSOxbIAi0E0L0+vn+W8X8O9rT+fDGSvYs7+01qzhN0wwTO57/Xum/fuGZiVynHLKAL6ZujKsXKiUkhHDu7Xkaf/GEdG2S78HYmMcgYI5NTTmcHPaGDSg468zqGaSFOvi9VvO42/vf8/+imqklHRuk0yep5Jyb7i3rGub5pfGbIwt+4siLnfrOv9ZvAAhwKsbqA1m+poqSE+M55M7L0X9lZuU1mAFWy8JIQLa8K4cXpu9lOWbc2pNE/6g4+795av55u4rOPOhd7AaOPMsKVm6eQ8A05duqhXEtQ5UAYVVbtbuyGdQt/ZNHl+XrDbcevN4Xnz5p5B490cfPu+gTuT/T0g4apESTSEqjJtI925tSU2JIy+/tHbaLCUh02qbppKaHMv4sS1nZpBSkl9WiaYqtE1omh262u/n5V+WMHXdRqSEyQN6c9vxxxFrt9O/UwbT7r2G3JIKNFUhIymeL5et57Fps/HqdZqV06Zx96ljDnCU5tE5JQkRQSsRDoHH0GufbWYc2DwCxQjYu8f06MzDZ044KoJ40c49PDJ9NtuLSoh32Lli+BDyiyuYkb0NX5WOiJTTIWFrQXFjz2oSYpwA2INCUxIwxwAgBAaS21/4mvf/dglZGU1/GE4+YwgnHN+blat2YbNpDBuaheMwfA2/R6QUUTPFb5lFy7bz0ttz2JtXSlJiDHEJLnSfjqoo+HWDY47pQt7+cnw+gxNG9+TKi0a22I9g7Z58/vrxDwEtFkn39FT+e9npdEpNanQbS0queO9ztu4vxh80Cby/dDULd+zhy+sD2qUQgszUuoaa5x3bH5dd44WfF1NQXknXNin8aeJYRnY/vCgJj67z8qqlfLUpGwnY4lW0MgOjdoYhMRvKZxX0OIlL01hx621oqsKbq1fwxpoVlHm9DEnP4L4x4+jfpuXaM0VibW4BN386pfYBVenz88b8ZWCAYVn1/YwhSCDGYWP8oO7MWrOtVmMGcNo1rjhpKADnnTCQZz+fh9us8UHU7c3t8/Pvz2bz0p3nNWvMCQkuxp1waKnz/19oqaQPIcRE4H8EvC9vSCmfaPB5J+BdICm4zj1Syu8PtM+oMD4AS1ft5J9PTsUXdMiUlFbjdGhMPn0wQwd2pl+fDiQmNF7z9nAornJz3etf4vbXeck35RVy1SufMeOe67CpkadbC3bsZmdxaa0gBvCbJrtLy/hl+y7G9Ygcs3zaoN6cNqjlNHpLSi6Z8jnZRfvwW2ZtlISWrKCWC4QpGNS+HStL87EaSXG2ayqPzJ/NRxvW4jEC38HivBwu/OpTvr3wcgo91XyxaT1+0+TMHn04qXPXurjtw+TFXxbj00NtsLpuIYJDlRpIf7hAtqkKgzq3p8+lbalwe1mxLRebquA3TM4bPYDzxwwA4JwTBrJ04x5+Wrc97NhSwrLNe1vkPKLUEahnfPj3hxBCBV4EJgA5wDIhxFQpZXa91e4DPpNSviyE6At8D2QdaL9RYXwAXnvvl1pBXIPXZzB9dja3XXfSEU1BnboiG7NBqJQlJdU+P79s3sVJfSM7Ztbn7QsxN9Tg9uusz9vXqDBuaebv3c3m4kL80gwJVzOwcKbZmXPJdaTFxHLtV1/xy+7dIQJZUxRO6d6dcp+XD9avCXH6AfhMg1umT2VvVRlew0ACP+3azomdu/LChDNapIrZ1sLGTQ0AKMHQu+Dp2TUVTVV47uoz0VQFTbXz0m3nkltcTkFpJV0zUkmOq3twq4rCv289k5F3PheIbmmA3da0n2a1x8/y7D2oisKwfh1x2g88K9u4JZ+Nm/NJS41j5LHdsNlarw215WmxTh/DgW1Syh0AQohPgLOA+sJYAgnBvxOBvIPtNCqMD0BOXmnE5R6vn2q3j/g45xE7dm5pBT4j/EdqmJJ95Y2nR3dISsBp00I0aoAYm40OSQmNbNXyrN1fgNfQIyZweA2Dqds2ce3AoTw6YQLnffQR1X4/1bpOjM1GisvFvePGsae8HJuqhgljU0q2lhQjlTpx6TZ0Zu/eweK8vYzscPhJKH3S25BXVhEikKUKouY5J0DaQVqBaI+/nnYCZxzTmzinI2Q/HVIT6VDPJNSQM47ry7eLskPMGXZNZfJxB69n8dOSzTzy+o+19nSJ5PE7JnPcgKywdQ3D5N6Hv2bNur1YloWmqTidGs//5zIy2ycf9Fi/BwKhbU1+UKcJIZbXe/+alPK14N8dgPpTlxygYSfdB4EZQog7gFjgoIXNW681uxXQoZF6AE6H7Yi3GxrapQMxEbQcIWBQp4xGtzuldw+cmhYyXReAQ1M5tU/LpC03hXZx8Y3GAhvSYl1hoNJb+/h45lx3HQ+PH8+tI0bwxCmnMPOaa0iLiaFDfDy6Gf5AEgCqBIcJLhOcJqgWbsPPz7vCp/2Hwu3HH4ezgXbqsmsM75qJXVNx2TRi7DZcDo3nrzyTi0cNChPETTrOWWNo3yZQX0JVA3HUg7q15w/njj3gdgVFFTz82o94/QbVXj/V3kCVv7/9byoV1eHZiV9OXcnqdXvw+nT8uonb46e0zMODj7fOOh9HgmbWpiiSUg6r93rtYPtvwCXAO1LKTOA04H0hxAHlbZOEsRBiohBisxBimxDingifdxJCzBZCrBJCrBVCnNbMgbdKrr98LI4G7eSdDo0rLzwuUG3tCDK+X3faJyfUet0hEOEwoltHemSk8eKixYx++VWGPPcCf/j2O/IqKmrX+eSai+nfPh1bsHLYgPYZfHzNxRGF+5HitG49cSiRhbFT1eibWpdB59A0zu7blz+NHs3pvXphD9rDU1wxnN69F04t9DtQVZDOemnTCmCXKDaIs0cWiB5D55W1S5j4zduc9e37fL51XaO2aoA+GW1587Jz6d8uHU1RSIuN4c4TRvL2Vecz865rue/0k3jozJOZ95cbOb5nl0b3cyAM0+KO96ayR6/E65J47RI9Fgb264DrIN/VjMWbsGSkjD/BnOVbw5ZP+3FNxDjk3XuKKSxquXjy1k4LldDMBerHr2YGl9XnOuAzACnlIsAJHLDLxEHNFEfKWP1bYOSwrtx/9+m89PYc8grKSU6M4cqLRnLu6Ue+mLddU/notot5Y/YyflizGZumcv7w/lw6ajB3f/cDs7fvwBt0av2weQuLdu/hx+uuJtnlonNKEp9fewnlHi8SSHIdujlFSsnn29fy1sblVOg+xnfozp0DR9PGFXvA7Vw2G1+ddymTv/wAr1knBATg1DQu6B3ehy4ST554CslOJx9tWIvPNMlKTCI5yc6KwtxQ84cAS7M4rVu49q9bJhd89xFby4prx7K5tIgFebt59oQzGj320I4d+PL6S8OWt42P49xjDq1eSH1+Wr+V7Lz9ePwGqAEziFeavDFnGReOGECbA4QyVnn8oeVRg5imRbUnPDXajLAuBJyFn3yxhE4dUznp+D7Exx8509vRJlBCs0X8PMuAHkKILgSE8MVAwxtlDzAeeEcI0YeAMA7vgVaPptiMj4ix+rfC8SN7cvzInrXtiFqC7fuLmb1pB5qqcEq/HrRvxJYb67Dzh4mj+cPE0bXLdpWWMmv79hB7siUl1brOx6vXcuvIOtNV4mEI4RoeXPYTn21fi8cI/MA/3rqaH/duYebk60l0HHj/PVLSWHbVLfx97gxm7NyGIS1Gtu/Io2MnkORsWhSKTVW5f8yJ/GP0OPymiVPTGPPFyxHjyhyqGjFLcsburewoLwl5KHgMnR92beHWgUX0TA4oLJaU+E0Dh6r9Kq2MZmVvx+MPF5w2RWHpjhxOH9x4dMvoQV34dMbKsO7fQghGDswKW3/8uD588sXSsEJGhmHyxTcrcDo0Xn1jDk89fhF9ezc94eS3RksUCpJSGkKI24EfCYStvSWl3CCEeBhYLqWcCvwJeF0IcRcB+Xi1PEhnhKYI4xYzVgshbgRuBOjU6dfpG9ZStNSP87mfFvL2/BWYloUiBP+buYD7J5/EuUObpiluLizCpqghKc0APsNgVX5+i4yxhv3uKj7eujoQmhbEkBblfi8fblnFrQNGHnQf8XYHz598BquK97K5Yh+dYlPonJjU7LEoQtSaKzonJJNTHaESmhC0dYVrkwvy9lBthAs9gOX7cuielMrzaxbyevZS3IZORkw89w87iUlZvZo9zuaQHOtCESLcXCIECQexPw/s0Z5xQ7szd8X2QHlUAr6Ms8cNIKt9eLLIJeePYMHibeTll+Hx6miqgmFaSCuQXVhjwnjg0W/47P1bfpd99QJV21rGvBiMGf6+wbJ/1vs7GxjdcLsD0VLRFDXG6v8KIUYSMFb3lzLUqBU0gr8GMGzYsNbfP6eFyc7bzzsLVuAzQrWZR76dxfG9upAWd+CpP0DHxMSwkDcIaFPdU1oufRlgQ+k+HKoWIowhEFq2oGB3k4Sxx9C5fsH7bCovwJISVVFIc8TxwfHXkOY8tGamdw4czcr9uXjqabpOVeOcrv2Ij2AzzoiNw66oYeehKQpprlieWjmPtzeuwGMGhFpedQV3zZ9GnN3O2PaHZg9uCucPH8AXS9eHhSLaVIXjehxYWRFC8OBNk1iwZic/LtyIqiqcPqYfw/pGTsWPcdl57X9XMn/xNtZtyGHuL5soLKoKm2BUVnnZtbuILlmtqypeSxBIh269MQtNGdkRMVb/f+SHdZtDQphqUIRgzqYdEbdZt28fF3/xGX1ffI4xb73O8oJceqalYmuQHmxTVS4fMrhFx5vuiotYFlJB0Dk+qUn7eH7jbNaX5eM2dbyWQbXhJ9ddxn0rpx7yuEZkdOSZsWeQEROHTVFwqhqX9BzEw8dNiLj+BT0GoDaICRcEBPiodp3qCWIJQoJi4ZV+Hln28yGPsSn0yEjjvrNPwmnTiHPaiXXYSYuP4Y0bzms0qSfkHIRgzOCuPHLr6Tx40ySO7VdXzlRKyaJ1u3jg1R948LXpLN2wB1VVGDemF3fcNJ6U5LhG0x9+j1pxgIBm3JTX0aApmvERMVb/3tFNk91FZSS6HLWOmEi1GQ7EluIiLv7i09rss7zKSp6c/wuXDhhI+8REftq2HSklXVOS+depp9AhsWXjiPumpNMtMYVNpYUY9SY5DlXlmt5N6wj9zZ41+K1Qzc+QFgv2b8dvGtjVQ5ucTezci1M79aRC9xGj2bA1ErkB0C42njdPPo8753yLx9CxkLSLiee1k8+lSvcjAxUiQAkUHampZLe1qpCnVs/hz4PHHdIYayjzepi2ZQulHg8jMjM5tn2HWoF3zrB+nDKgByt35RFj1xjcuX2L1OF47K2ZzFyyudaEMWv5Vs46vj9/uvxEAE4/dQB79hSF2ZwTE1x07pR62MdvrbREBt6R4qC/hCNlrP49893qTTw6ZRaGZWFYFsd0bs9/LzmdiQN68sHiVWHTUktKTuwdnlH3/JLFYQkPHsPgw3VrWXbDzWiKgt80iXc0P761qbw7/iLu/GUKy/bnoApBvN3BkyNPo0dS0yY+hhU+E4DATWIe5i0ihCDR3jQn5ej2nVl68a1sKSvCrqp0TUhBCIFumWiKEii2L+pKRNT8+8amZVzUfTAd45IOaYxLc3O4dsrXWFLiMwxcNhvHZXbkfxNPY/qWrczZvpO28bFcMmgg3VNbRghm7yhgxuJNIXWVvT6db+au4+xxA+iWmcbpEwexYPF21q7fi64b2G0aiqrw8P3n/G414xaMpjgiNEktORLG6t8KXr/O6i25qKrC4B4dDljUfFNeIc9On8/8bbtDWsov35XLbe9N4cNbLuaa0UN5a/5yTEuiCIEQ8M8zx5MaF964dN3+fRFjYTVFYW9FOb3T2uDQDv4VLty3k2c3zGFXZQndE9pw94BxDEtrmgM11RnDhxMuocTrpkr3kxmX2Kz6D+Pb9WZazroQzVoA/ZPa49LCY2kNy2RZyWYKPKX0TMikb0LTO4kcDFVR6JMSWgvZpqjcOWg0/149GyPCtRbAvLwdXNbzmGYfz7Qsbv3uW9x6nfPQress2ruHU995l3K3B49uoArBp2vW8dRpE5nY6/ATcxas3RlWVwMCYW+L1u6iW2Yamqby5CPns25DLus25JCSHMMJY3oRE3PkHuytgWjVtt8Ae/NLefnjX1iVvZfEeBeXnjGM+GQXD7wxHVNKdMtCCLj2tOFcf8ZxYVPJuRt3cPeH39XG/tafDRmmxab8QnYXlXLHyaM4fVBvZm/ajqaonNKvR6Nt57snp7CnvDxsuW6atI9vWl+zn3M388clX9eGdS0r2sM18z7i1dEXMSq96c6pFGcMKc7GO103xp8HTGBJ0S7K/R7cph+nasOuqDx2zJlh6+73lnHHihep1N0Y0kRBoXdCR54cfD0O9cglrNzQ91iW7t/Nz3nbwj5ThSDG1vxsy3Kfl/8tW0i5LzwbzmMY+PRKCMpoU0pMw+DvP87gpO7dapNeDpUYpx1NVcOao2qqgstZdx2FEAzsn8nA/k3vLKLrJp5qH3EJriNam+VIEO2Bd5SQUlJe5sZm14iNPfDTPr+wgmvv/QC3Rw9sV+nlmXdm4xMmvpomnQBC8Nq0Jazdkc/zd55bezNaluSBL3/CqxuBPp/1vu+aPzVVobCyms5pyXRtk9Kkwu23jziORTl7a23GEEiYOKtXHxIOEuNbw2NrZobE1wJ4TYMn1vzE1FNuaNI+DodURyzfT7iNH3I2sK40l67xaZzZcRAJEcwLj234iEJfeUhmWXbFbj7cPYtru57aYmPymwZf7V3G93mrsSsq53Yczr1DT2LBvl1h10oCJ2eG9507EHsryznr6/ep9voxLBnRV2DJcO+537TYsG8fQ9pHjvOVUrJmbz5zN+8k1mHntIG9IsaoTxjek1e+WBBxHycNa9651GAaFm88P5PvvlqBaVrEJ7i46a5TOfHUpoVktgYkYEQ141+P/JxSZv24ju+/XUVJSTUgGDIsi7/efyZJyZFDxz6YuhSfzwgxLfj8gWpg2IEGKeUrt+SyaMMuRg8IaJYF5ZVUeLy13RrqU1OIXjfMZjfRHJzRjhdPn8yDc2aRW1GBXVW5bOAg/jJqDFJKvtixlpc2LKTYW83A1PbcM+Qk+qfU1a0wpcXe6rKI+95a0Tz/ao67kGl5iynylTM8pTfj0gdjV5p2+zhVG+d0Hsw5nQc3us6Wyt1sqdqCXZEYUsGwAhfTbxn8kLesxYSxKS1uXvommyvy8VoB1XRTRR7j0/vz18En8MiKn5HI2q/xit6DiLc5sKSFQDTJZPLIwlmU+bzBDiERfvySiIXpfYaBR48cDy2l5O9f/MjMDVvx6gaaqvDiz4t5/PxTmDggNB66bUo8D900kQdfm147g7Ok5LFbTyc5ofmzG4CXn57OjG9X4fMGHlalxVU88+hUEpJcDB3x22ntFDVT/ApYluTph6cwZ8Z6dN2srbYlNYVVy3fy1zs/5NX3IvcWW7c5D8OMnC4qzPD+cH7DZO6a7bXCOM5pR5dWSKnIwMED752qyvXjhhN/CIVkxmV1Yc7V1+HRdeyqWvvjenH9fF7asKg2Nnbhvl1cNPN9vjr1KnolBeyiCoJEm5NyPXyqnOYMPJgsaTElZylf7l1Itenl+Db9uLrreJLtdTHAC4s28PD69zGkiSktFhRt4LO9c3l+6O241MO3Mf60bzGvbv8Um2IgBNiliWEpuA0bgbKHkZ2Ah8Iv+zextbKgVhADeEydGfnrSLXlEh/nw28EulSrqsVnexYxe382Oe5S4jQHl3cdwa29x6EeoObLvJxdAVu/AGm3wB9YVwT/L5FghUbXBB4AguyiQkZ17hy+zy27mJm9DU/QFhzoE2hx75czGNuzC7GOUFPKScf2ZET/LJZl70EIGN6vM65DbHrg9fj5ceoq/A0iL3xenQ9fn/vbEcaydZspWu9joplM+3I5837ORg+me9ZaFkwLw7DIzytl44aG4dEBOmYkNWzyXEukB6mqCOJj6qbZCS4nMU5bhJ7yAa/8XZPGcsv445p5RqG4bLZaQew19BBBXIPPNHhu3fy6wwvBDb1H4mpgb3WpNm7tE2ip9GT2V7ywZRo7q/ex31vONzmLuWbxc1QbAQFuWCaPZ3+Mz9Ixg+YDr+knx13I1JyFh3VOAG7Dw6vbP8WwfDhUHYeiY1MMNMVEUyw0oXJC25ZrNb+kaBtu0x+23JKQ5y7HEhLNZqFpgXP1WSY57kAp1SrDxzvbF/GvtQds2BBq81VBOi2kTYJNYtMCgh41IIBr/gMwbZKcigiZhcB3azZFTJ1WFYVF2/dE3CbWZWfc0O6ccEz3QxbEAGUl1Y3ahwvyyg55v782NcXlm/I6GvxuhPG3ny/F5w29WQUEm4wF6krsyw93hgFcftbwsGLedptKckoMRKjOpqkqk0eF1pvNaMQJ57TbGNGtZRuU5laXR4xosJCsLQlNib6x1yiu63UcMZoNp6oRq9m5ve9YLu56DPmeEmYUrArREg1pUaG7mZa7DIDtVXkRq4P5LJ2f96867HNZV74VVRg4VANVSFRFYldMYjQdl2qS7kzmmha0F6c64rFHiEkWiDCnbKTAO6+p89We1VRGmG3UcF7P/jjqC2QBNrtCn7ZtA/tUQNoCna+lDazg36iw3xO5VnXDpJWQz45wKFpq24SIM0ohoEfvdkf02C2NFdSOD/Y6GvxuhLHHHa7t1Mc0LHr0Cq8DnL29gMffnIEvmJigCIFNUznpuF58/t/ruOXMkWiqgqoInDYNh03jH5efHNYs8qwhfXFE6M4Q57DTrU3z40fdup/Ptq3lyZVzmLozO6SNUtuY+JBQsfpkxYUWChdC8Id+J7DsrD/z06TbWHbWn7ix9yiEEGyqyEUT4YLJZ+msKAnUBXaotkZLTca0gIkieEREvRhfhEAISaLdy3NDbybBdmh2zkhM7nAMSoTb3qmpRA6ND/9h2hSFfE/kBzvAX4ePZXDb9rg0jVibjRjNRu+UNB4aPT5U4ItgtxC17jDVjdiMzxrSF1eE+0tKyXHdWq7Oi9vjr51d1mCzqVx+4zgczlDt2u6wceXNJ7bYsY80NcXlW6sw/t3YjEeN6813Xy6PWFbQ7tA4bkwPMhtkFuUVlnPbY5/VZimhgmoT9O2ewT9vnwTA9Wccx1UTj2Xllhx8hsnQnpnEOsNDnS47bjAz1m9l2/5i3H4dh6aiKAr/vei0ZocA7a0s45zp7+E2dNyGjiIEcv5U4l0aF3UdzF0Dj+ecrP58s2t9iPffpWrcMSByR2e7opLuCtXe2zoTa6fI9dGEQoeYwLXqHJNOG2ciOe6ikHWdip2zOgRCyw3LYEre18zePwuf5SUrtguXdbqCrNiDh84l2WIamRQKLCTT839gXNvRaIpGW0e7w445Tncl8t+hl3Pv6k/RLROJJMkewzPHXMEv+7bxypY5teYfBUGkR55umSQf4AHh0mx8Ovli1hftY3NJIV2TUhjcJjD2c3v35euN2SF1NWpwahrHd8qKuM8RXTtywbED+HTpWiwp0RQFCTxzyRkHrX3cFNZvyuPJl35kb14pihCMG9WTP910MrHBuOPzLxtJSmocH781j+LCKnr2acd1d5xMt56NNzpobUhE0DHcOvndCOPLrj+BhXM3U1HmxufVa1Nak9vGc96lIznvooaF5uCzH1eGxmIK0A2L7B372JlbjMc0+O/Xc9mwZx/JsS6uOnkYYxuxvTlsGh/edBFzN+9kyY69pCfEceaQPk0q/tOQvy3+gRKfp1YjDfwrqfDqvLNlGatL8vjwxEtxaTY+3rYKw7JIc8XywNBTGN62E27Dh6ao2BWN/d4y5u1fhylNRrfpT2ZMXeZc34SOtHMms7u6ELOe2NGEynkdA0WAhBA8NvA67lr5Eh7TFwwPMpnYbhhZsfFsrdzMz/t/Zk3ZKnQZEGI7q3fwn81P8M++D5PuPHAX52R7EqqiRnTSCSSLS75nVdlPWFgk21K5sdvdpDsPr8TjcWndmXHSPWypzMcmNLrHpyOEoEdCBr0SM3h/+yJK/NUMSOrAN3vXNAh3k6iKybnznuLa7idyTbdxjR6nf1o6/dNCz//RcSczoUt3Hv5lFnvKy2uzEO2qSqorhgv7RQ4V0y2LCnR0e8D5hyK548SRjO2ZdVjXAiCvoIy7Hvwcb1ApMYE5C7dQWFzJ849eXLveSRMHcNLEAYd9vKNJa06HFkcra3nYsGFy+fLlB1+xGbirfcyYtpo1y3fRPjOFM84bRrvMxvt73fn4FyxdvztseazLzk0Xj+GZH+YHCn8Hcdo1Lhs3hDsmR9Y+WwK/adLn46caSRWWqDaJXVH58KTLOCYtE8MKtBuKtznYVJHLvzZ8wc7q/QgEveLbstuTj5BBb70QXJk1gcu7jK/dY4mvkgfWfcy6sl0oQiHRFsN9/S9kaEr3kCOblsnK0m2U6VWk2m18lvMmHrMaKQVuK3xqraAwJu14rsy6+qDn/I91j7GjuuH3IIlXfTjUUCEdryXwyIDnUcWvo0esLtnLE+ums74slxpBbFMDCUBO1cY/B5zHhHbNdzBKKZm2dTNvr1lJhc/HqV27c8Mxwxqt83zfdzOZun5TXVIR4LJpPHXWJCb06h5xm6by3Juz+PqH1WERRQ67xhtPXUFWx6Nfq0IIsUJK2bSCKI2Q2CtdHvfaJU1ad8a4/x328ZrL70YzBoiJdXD2RSM4O4IWHIl+3TJYvSknrJKabpjM37IrrGuv12/wwexVXHfKcGIcR6YHnhA1IU+NPyR1y2RT2X6OSctEUxQS7E72ecu4bflreGojBSRbq/aGRolIeH/XT4xu048ucYHpZYojnueH3UiZvxqv6SfdmRTRFKAqKsem9kK3dO5ZeyfVZsDRJGs8Ug00DguLDRWreW/nY6Ta23Fc2mmkOiJPaf/a+w7uWvUPPFadU8wuDOxKuLbst/xsrFhL/8TmpycfCoNTOvLGqCs4ZdZj6A3qbHhNnfd2zDskYSyEYHLP3kzu2XgB+Ro8us6U9RvDGtR6dIOX5i85bGG8c09RxNBOTVPI21fWKoRxS9DMhqS/Oq3XgPIrcP4pQ3A4Qrs6OOwaY47pxo7C0sh1IVSF3OLI4UctgU1RGds+CzVsOiVD5F3nBo66r/YuDinKo4jIwly3DObsXxO2PMkeS4Yr+aA22Q0VazBluL0zEl6jiM2VK1hc/APPb7mLHVXrI66XaEvg6cGPkOlMw6VqxKs2YlQZMdxQSkmF3rjz7EhQoXtQG/mpFPsa79TdUpR5vI1W/CuoPPzj9+vdHrst3JGr6yZdO/2+KuG2Zgfe70oYe9x+vp+6ileem8nM6Wvx+SJ7pmtITYrlrYcvY8yQrjgdGskJMVx++rE8fOtpdE2PnK6sGybpSYdWFL2pPHHcJDJia5xtwfKOgKIG/k12xDAyPStkm51V+9CbmBwRKVStqVQbVVj17MsBTb5ujLXLkThVT+B4mOjSx5d7X2gkYgGS7En8e9C/+Eefv3Bjt+u4uNNl2JXwaA2JRde45nXgKPYVMi3vE97f9SLLSn6h1FfMrP0/MiX3czZVbGh0TDWkuxIbqY0hyYxJatZYDoW2cbERC0IJYHD7w3egnTtpCHa7FiLubTaVE0b2JKNt4mHvv7UgEZiW0qTX0eB3Y6bIzyvlzhvexuvT8Xp0nC4bb786hxfeuJaU1FDhWVRaxZRZ69iZU8SAHu154JZJxDWoVnXDxBEs3bI3pNyl06YxaVhvEmIOrbec19D5dsdmlhXkkJWYzIU9+5MWobGnTVV4YPhJzM7Zzhc71mNi1iaQxGl2vj7l6rA44wFJnVlavA1f0H5rSRGxT5xN0RjXdlDYcikl8wpn8NO+aVQZlXSMyeKczMvoEhtay6BnfJ8w4aUgEagoig3d0rErklilGq2Bdl5plFBhlJBoizztFULQLa4rAKY0WFw8h33ePHQZML3YFQeDk4aT0cCBZ1g6K0tnsLpsFgKFocmnMDj5JBShsqliLW/u/C+mZWJisrpsCbo0kNKBLg1m7f+RbnE9ua373Y3aoVWhcGP3k/jPxm9rRkrNw2eHexclvgpSHC1bSzrk+IrC38aP5ZEZs2sz8ASBRKC7xh1+sUQhBMKmIBWBtIIJKEgG9uvAytW7yd6UR2pqXKCqm+vImOd+LaIOvAi0tAPvb3/8kNUrdmFZdeejqgonnNSHvz94Tu2yzTv3cevDn2EYJn7DxGnXiI1x8NZjl9E2JTT0a/6GnTzx+WzySyuwayrnjxnInWeOaVIXhoaUeT1MnvI+RR43bkPHoarYFJVPTr+IAWl12s0L2fN4ZfN87IqGRBKvOTm5XR8q/T5Gp2dxWqe+2NVATOzc/euYmrsYv2Uwpk0/3tk+l0rDgxUUFA4FhDBRhYKUEk1RuajzCVzbdWLY+L7N/Yw5hdPxW77aZXbFzl09HyAzJitk3U/2vMfC4nm169oVOx1dWdzd615UofLQugvxRzRlSE5sM56T293epGvmt3zM2z+T5aULsCl2xqSdzLEpo1HqpSJb0uK9Xf8k170FXQbGYxMOusUN4cJOf+P+dTdTZYa2oq+pa2uiBsfv4MKOlzMmbVyjY/lsz1xe3vodHiMwjVWFhV0zcak2bup+BudkHjmnbg1ztu3kpflLyCuvYHCHdvzhhJH0aHNoZgQpJd/MXcd73y1jX0kllm6h+GRdzQwpsfvAoan4fDpOhw1NU3j235fQrUvbiPs0TQt3pZeYeCdqhGSpw6ElHHhxPTPk4JeubNK6Cyb8J+rAOxRMwwoTxBC4ORb+siVk2WOv/ojbW5cg4vUb+A2Tlz76hQdvPy1k3TH9ujCtXxc8fh27ph5WB4ZnVy4kv7oSPdjGyGea+EyTP875np/PvxaAuQXbeH3LQvyWWduvzWPoLCzazg8Tbg6x5/5n0xf8VLAarxU4ly2VuXSJz6CNvRuLi7fgVOyc03EEEzIGsrBoA6a0GJ3Wj6y48Gmtz/Qye/8PtRpoDbql80P+19zQ7a6Q5Rd1vILeCX2ZVzgb3fJxbMooRqUejypUCr270IQff8PydUhswmRj+Qx6JYyhY+zgg14zu+Lg5IwzODnjjEbX2Vm9ljzP1lpBDBJTutlZtYSf898POycIzDAUJGbwdvFbPhYVzTugMPZbOoqwcNka3GPSwmce2BzWUozr3oVx3VumJ98b3yzi/R+W1xWgV8ByguYJdp7SJdKQeINx+55gduuD/5rKe69dF3IvSimZ8t4CPnz+J7weP3anjYtvOpHzbzih1RWql63Ygfe7EMaBKXzkCASl3hO62uNnx97isHUsS7JgVeQedMABg+r3lpWTU1ZOj7TUA8YUf7drc60grs+eijIK3dW0iYnl/W1Lw+pNWEjy3eVsrSikZ2JAI9ldvZ+ZBatqTRIAXsvPruoCrsg6iUcHXRayj06xkTWZGkr1koC22eDySSQ5nl1h6wshGJw0jMFJ4YpDmV5AjKqgSx1d1t1eKhbxqhdDCrLLZzZJGNfHkgbbK2eztWIGqrDTO/EMOsUex66q9fiDURgqJnZRYzc3WVX2CSkalBix+KwDJ0YoByj8A3Bcal8+2PVTyDUHiSrgmJTDi2b4tfH69FBBXA/TDpovIIwjsX9/BQX7K2iXXmdLnv75Mt55ejo+T+DaGLrJhy/+hM2ucfbVR37G0HRad6Gg34UwVlWF40b3YPHCrZj1MvA0m8qJJ/ere68qjRYEsmnN03rdfp07vprG0j17sSkKXtPktD49+PfkSRHrRjRWblIiawvLRKquBgGbZaVeZz5YXbo94noe08+y4i2MadMv4ueNkWRLxmrE+dfcBIs2jiwkJrGqDtJPoES8RBNW8NpLTNk8TVJKi+k5fyffsxZDBpyCOdXL6ZV4GnG27jiFAHyBonkNLr0iIEWrZp8/ASvorw6YKeq+b7viYPQBtWI/ed4d9E5wsMftplJXUITEplhoisW9a5/lrA4ncnnn01udJhiJvKIKlEizPCGQSlAIHyi6soFp8+MXf64VxDX4PDofvzyrlQnj1q0Z/6ajKXxeHX/w6f6Hv55GRrskXDF2bDYVl8tO56w0britLsHBYdcYMSgr4r48uoE/Qquaxnhw+s8s2b0Xr2lQYfnxC5NvNm1i7Muvs6u0NGz9S3sPxNmg+aYqBMe07UBisFD8xA59cEYQ2hJJ/+S6giyJttiIJRxtQg0pfdlUnKqL41LHYVdCnTM2YWdSxrnN2leSPYPu8SNRhQNVSByKiU2xaoWkQFLk/pkdFd+jm5V4jP0HjWbY615GgWc1GhW4hB+X8KNSwabyKeyp+IhUrQSn8CNCHIYSBQubCFSCa2OvQMUMCmwNm+JERcWu2BmQOIjhKaMiHtttuPnHuod4b9fHFOu5JNh8ZMRUE6uZCAGmNPFZfqbkzubbvLnNulZHi7bJcRhmhIevDKTCO+wabdslYreH+0batIknIz00wqKkMHKoZ0VpNVaE2eDRQkowLdGk19HgN6kZ791RyDP3f8WmdXsRQjBsTA/++NC5vPnhzaxYuoOcvSVkdW3DkKFZYZrKqCFdWbh6ByH3iAiUypy5eDNuYTB3/U7SEmK4aOwg+nQMpLNKKSnzeImx2RACvt+4BZ9pIDVCTKP7qqu45OPPmHfT9SGOvhsGHMuyghwWF+wFAtPiFKeL5048vXadi7sO5Ytdq8lzl+MxdRQEdlXjgcGTcNQT5CPT+qAKBVVY2JTAA8RvaSjCxqT2h+ZzOK/jFbhUF3MLZ+C3fKQ50jk/80q6xDW/M8QZHf7CkuLPWVb0OaascZ4FnYpCR8XNysIHWFMIQqg4lCSOafsg6TF1ZUalNAGBEAq7KuahEdhPzdepSIkNN+X6XoSQSCGgntajChOVuljlGEUn01mKYcUxOfPPlBkWVUYlPeN60zm2a6PnMi1/OkW+YoygQ9IkIMRcmhddd1Lz5fssP1/mzOTMDuPC9rG3upjtVfvoGJNCt/ijX8shLsbBpFF9+XHRxhBThcOhcfdFJzC8f2fSkuP4872fsW37Prw+HYfDhiJhQI92PPP0dEaP6cmxx3ZFUQRt2iVRsLck7Dh2py2yBn4Uac3RFL85YVxV4eHuy1+lqsIb1Kgky+dv5c9Xvsbr0/7I8JHdGT6y8e1LKt1YNdWygsXfEQKP3+C/3/9Cta7j9RsoQvDDis3848LxJCa7eGjqzxRVVSOEYGK/HpiWFbHWMQSaTs7ZsZMJPepsiXZV5Z2J57O+aB/rigroEJfA6PadQ5yCsZqdr066nq93r2F2wVbSHLFc3u1Y+iWHlil0qDZOadebWfsX1s4YnarJuDZDSHc2nv59IFShMrnDRZzR/kJMaaI1sYtHJBShMjLtYoYkncyS/Q+ws3IVFgK7oiMAhzAQUiIFSGngMfexuOAuTsz8EE04WF30KIWeJQgEGTHjcBsBE03952qNi0DBxELBLsygjTrwQX1BXLO+IiWKqKbEv4vRbULt6o2xtHh5rSCujyDgBKz/467Qq0PW0S2Df6z+hEVFW9CEiiEt+iZm8vTQK4nVmlbxrtBdzbc7NlHu8zI2M4uhbdu3iCnkb1ecRIzTxtez16KbFm2SYvnL5Scxdkhdofj//fsSVq7ZTfbGPApyS5n9UzY/z1yPaUp+/mkDg4d05uFHzqNN+8jC2DRM9uWWkt7h0O7JliZQTTcqjFuMn4MdB+pPbU3DoqSwklWLtjN09IE1uZ6d2+Jy2XF79dBedTEKFV5fsINCoDiPVzd4+IufMONESE2AHzdsxalp6Ko/YiyvYVnsq4qcGRWpeEx9XJqNS7sN49JuoRqu39RZWLyeEn8FafYE5hcF6g3X/10uLFnBpd6JZDgPPWtKCIHWIN5Wt6op8W3FqSaTaA/vQhEJw/IwN/cyvGYhybbANZUyoJnIYH3p+phSZ2vp+xR65uKzyiAYoFfgnoMQzqAJItIPKVhoRxh4sdXaqCOfG2goJNia3v6qoekm9Mih48mK7QDAkuKNfLBrBhvLC6nySywkPgL3z5qy3Zz68+NIqTE0NYu7+0wkKy7y9zV77w5u/mkKEonfNHl13TJO6tiVF06aXOuXWJOfz8dr1lLh9XFqzx6c1qtnk0IvNU3lrkvGcceFx+PzG1RUe1ievZdZy7YwalAXnHYbiiIYNiSL3j0yuOC852tNggBer87qVbtZuGAr3kbK1zpjHOzPaz3COOrAa2H27iwKKyIPgTC2vD3FBxXGowZ1IT01npx9ZehBZ59NUzGcCnqEsoZ+m4VhhP64fYaJTUrsmoK/pt1SPRRgcPuWK7q9syqfP69+AV2aGJZJ4OdtYVcaOKykZGnxOs7s0HI1ZrNLP2Z18SsoaFgYJNm7cmL7p3BpB65XkFs1Hd2qgAbZeoqURHYVmpT6VgcddHXbSEyENLAJBb2BbTlQmU/U/p2gevBbdiwRj0U1DT1QUgbsxb0Sxjb5/E9OP5EP93yK36oTOAKBKUWtMBaAXbFxfddz+TF/Gf/b8gU+S6fK7wgT2Ja08Ek/ftNg3r4trCjexZcn3EGGK9QO6zMNbp/1bUjFOI+hM3vvDqbv2sJpXXrxzoqVPDVvPn7TxJKS+bt28/Gatbx/4flNjoXXVIU3vl/GBz8sR1WUYF1pwf/+fC4Duwect6tX70HTFPwNZK7XqzNrVjb9h3Vh5+YCjAa1XHS/QefuB67a92tzlNIqmkTrMug0gZ79M3FGyAJSVIWuvQ4uADVV4fX7L+bMEwaQEOckMc7JWSf2p3+3dnWZx/W+MFNEqvgb0CyO79YlEAlRbxunpjGycyf6p7fMTSil5MENb1FhuPGYPnRpYEgLUwqMBo4GgThgb7am4Der0S03AHnVi1ld/Cqm9KHLakzpo8S3hTl5fz3ofsp8GzCDkQ8NiaSbKMKOTUloZBsLhxqLqKc7qMJOkr0HqnDVNv3UFDtJ9jiu7fY6ibZ2Yc1AhRCc2eGf2JXIldEiMa7tWIYlD8EmbDgUB07FSao9hTt73sKQpN6kOZIYmtyPJwb+kd4JXXhl+9Ta8LeD/e4lEp9p8MGO8PZVywoitwhzGzpfbFlPmcfDf+b+gtcwamuouHWd7H37+WHzlojbRmL5xj18NH0Fft3E49Nxe3WqPX7ufvobjGBhokiOPAg8AJ0OjXOvHYvTZQ+p2+1w2TjjspEkNNIE+GghpWjS62jwm9OMT5g0gPdf/Andb2AGTQo2u0aXHun0HdK0jgfxsU7+evV4/np1INLi7ZnL+GzF+mCpyToURRCn2ahGx6yXUGLaoMKmM3vnzmChbxWHppIRH8slgwdx2eDwdONDJcdTSLEvUmEcgSEVbITWYz4u9cDHLvHlkO/dQryWRseY/oig8C7z72Ze/qMU+7YgEKS7BiHwY8rQcDuJSal/G5V6LvG2Do0eJ87WFVU4w7YXQuBS2+O3Smo/EyjECAtN5iFqGnaGbGNjeJt7yHWvYXfVzyjCRveEyfRPvpIKPY8NZV9QoefS3nUMfZLOxqHGc16np/gu92GKfDsQCGyKi1Pb30Pn2OY5OBWhcEv3GzjbU8D2qh0k25Ppk9ALRSgclzokZN1iXwVesy4EUVWsYJ2D0B93QHYGlunSZHVpeA+7A5mFFaGwLCcXm6riNc263ctAp5AfNm/lzL59mnR+U+eujxxvbFqs3JzD8H6dGTIkK2KDBLvdxqTTB5OWnshzX93OO0//yJrF24lPjOHca8cy8YJjmzSGX4tANEXr1T9/c8LY4bTx3Ce38ubT01k0ayOqpjDhzCFccfvJh+TYWJi9i1d/WFxrK67ZgwAykuN56PJTuO2TqVT7/UgZcNhbjsAKNTWHDcvCJhUemnAyx3Vq2X53pjQbrdilCEmizYMEdMvOdV0uJdWRFHFdS5p8l/sUWysXIkSgJlyMmswlWU/iUFxM23MLfqsSgqKwwLMKEdQtw2J30fCZZQcUxp3iz2Bz2SuY0kfNI06gEWvryIkdPmdX5ddsK/8I3aokVfWDLMI0ixHEBaf2NSYAG7G2DrSPPYkOcSczvO2fAPDou8mrfBspTYYkn0WsPdQ8FW9ry8VZL1CpF2JYXpLsHWofPIdCO1cG7VwHjoSI11zUF7xOzaDab6fWUxx82NfvNqEKha5x4TbsY9MzI8arx2g2LujZn1hhD7QKq9lVvbhgp6bywdLV7CwuYUCHDCb17Rmx0BAQURDX7K+mhKzNpvLYvy7k7/d8GpgEWhLTsrjo4hEMHBi439t1TOXvz1za+MVpJbRmM8VvThgDJKfF8ed/nX/AdardPrI35RMX56B3j4xGBfX7s1ZGvCFtqsrLN59DVkYKn950Cf+d8QvLd+WiOlXK8IX0pAPw6gY/bNra4sK4U0w6sZoTb0ODHZIYVQ9WTQOnqrPHvRmIHC+7uuQ7tlYuwpD+2h+tbvmYmvMEQ5PHYEk/9ecFgTgBjcAtEnp9JBZJ9gO3Z7epCYxt/x6rCx+k1LcOEKTHjGVwm/tRFI2uiRfQNfECKrzLyd5/NZb0IgQkK9VUWU58aKjCSYfY0+iXeleIIM2teJ9dpU8gZcB2vrf8JTITbqBz8h/CxhHfDGfd4WJXbUzMGM6PBUvxWTqKkMTafViWnU4x7Sj3+8l3V1Hfam5TVK7sFl7sx66qvHLyWVw/4+vgw9bEpiic1qUnp3TuTqHbHe6vCArkn7K3M3v9Djy6QcyabJ6fs4jPr7+UlJhw88yEEb1Ysn53XeuxIIZpcUyvzNr3/Qdk8vmXd7JkyXbc1T6GDutC27YJSClZvXQn2zfn0y4zhRFje6JFKMfZWohGU/zKfD1tJS+/MQdNU7AsSVJSDE89cgGZHcLLYpZWR7Zr2jSVCk9gytm1TQovXnYWAN9mb+K+GT+FCWMhBHat5W9CRSjc1+8q/r72VQzLxJAmAokmTOJsdQJaIllUvITzMs8hyZ4Utp9Vpd9hSF/IMolFgXcLRd4MDBkp+0+gKTFIvEFhDapwMrTNH9CUg1eui7dnMbbDO5iWDyEUFBGeklzhW4JV79iKkCSoHkAQpyUjvW+zKfd9kmIm0yHlQQzLzc7SJ5D1zsWSJjkVr5MWO5FYe/PKa7Y0t/U4G4nFjwXLUFBQhcK1XU/j7MwxeEw/j6+bxg9567CkRfuYZO4fcCbd4yP7F0a378ziS27m+52bKff76ByfyHsbVtPjjaexWSqqIkLMZwAI8GCg6gGh4/br+A2Tp3+ez6OTJ4Qd46Rje/Dd/A2s2pyLx6ejKgJNVbnnqpOJbeCbcTptnHBCXTF8j9vHX298h707i9D9JjaHSmysk2fevZ62Ga2v9Kbk6NmDm8LvThiv35jLy2/Owec38AVllXdfOX++73M+fuvGMA35xAHd2JFfHNbtw5KSnh3Ctapx3bqEFSSCgCZzdr+m2emaS//Errw34j6m5y3m473TcCheXJoRZj7QFI1cT15EYaxHFLYBp1+SvQuacNWmGtegCI0xGQ9Q6ttIbvUCYrQ29Em+hHTXkIj7agw1Ql3iujGnoAgHVpjjTuI3dmETJhKTMvcUPPpG4mMvD9qVQ7Gkn6Lq6UddGGuKyh97XcBN3c6kXK8mzZGIpgQe0i7VzsODz+X+gWfiMw3ibAd/oCU6nFzSexD7qqsY/9mbVOv+gKnDtFCtCA//8NLSGJbFjE3bIgpjVVF4+q5zWLh2J7+s2k58jJPJx/ejc7vI9bzr8/4rc9i5dR+6P/DbMQwTn0fnqX9+zb9fu/qg2x8NWrGV4vcnjL/+dlVIPCQE7ERlFW6yN+fTr3dorYVLxg1mypINFFdU49MD6bIOm8Zfzx+H0x56eTy6ztQNm+iaksKmwkJsasD2qlsWx3Rqz/byYrqnpYRky7UUyfZ4LsmaQKrTwfu7P4y4jmEZtHFEnpb3jB/DqpKpmA1MDjFaEr2TJrOp/HOq9AKs4OeqsJPs6ErH2FF0ihvNoNTrD/scpJT49NX49E3Yta447cNJiz2dXaX/iri+Vm86L9HxGzvxGbsb2Xvr0nhcmgNXI4kdNkXD1sykmnc3rAxkfNYsUIN9DSOctxLBDFzt8+PRdVy28NmJogjGDO7KmMGNZyICrFq9m8+/XEZJaTUjR3Tnp2mrawVxDZYlWb9qN16PP2LU01FFgjxKqc5N4XcnjMvK3RGN9IoQ7M0vxebU6Nw+BUdQ0CbEOPn0nsv5/Je1/LJhB20T47h03BAGdQ0V2m6/zrnvfURueQUeI6CVmqaFw6lhaRYLCnaxsjiXx+bP5cvzLyUz4cgUGz8l43gStBhe3fFqSMcNm7DRO6EXbZ2RhfHINheztXIhbqMUXfpQ0VCExunt/4ym2Jnc6TVWFL3OzspZKEKle8JEhqRe22KFbyzLTW7RZfj0dcElApvWmcw2n9O37TtsLrwVU7qDnxnYZXWY5i+RuNQUIlbnEzbSYk8LW94a8Zp+lhRvxmcZHJvSo0n1RNYV7gut+ifAclkoHgWnqqEpCoZlkelKYE91WfjMwbL478/zuW/iocWgfz1lBa++MQefLyDpd+4sRFRENvEFjtc6ddDfvJlCCDER+B+gAm9IKZ+IsM6FwIMEfilrpJRHxbU6dlQP1mfn4PXVqQdSQIVq8vg7M9HUQB70HZeN4+zxgUaS8S4H155yLNee0ngozser15JTXlGbiScl6NJCd/sxYwKOlGpdx2sY/G3WdD48+8Ijdo7HpQ0j2ZHAOzvfo8C7LxhmNYIrOjd+yV1qPNd0e5nsstnsca8hydaOwSmTSLAFyms61ARGpf+JUel/OiJjLq54Ep9/NZI6W69f38r+0ntpl/oywzIXUe3PBiQe3wryy/6FJPTHLhDEOYbSLeUBtpc8FHzoBrL5OiXeERZRUYNhluMz83FqmajKkW2ZdTBWlGzjnjXv1Gq0hjS5rccZnNcxsuO1hn5pbVmQu7u2cQAAKlixFskuJ38fMY6RHTuiGyYnPPtG2PYW8PWa7EMSxh6PP0QQA/h1E7tDQ3iN2u4gEPCd9OzbnpjYpqV7/9q0VDTFkZCJBxXGQggVeBGYAOQAy4QQU6WU2fXW6QH8HRgtpSwVQhy4gO4RZNLJ/Zny3Wpy80vx+YIabJyK1AR+3cQfDNf53/uz6ZiRhGUTPDvlF3YUlNAuJZ7bTh/F+MHhP+oZW7aFpETXIJGBOz1ovjOlZEluDl5Dx6kduIZuU8hx57PPV0znmPakOerseL3ie/L4wEfxml40oTWploRdcTI4ZRKDUyYd9riaS0X1ZyGCOIBOled7pDQRQiXO0R8Aly2LfRXPYFleJBYBMWzHoXUj1nEscc7hJLvGUuSeAdIkJWY8Llt4mrYldXYU309h9Tcowo5Ep138NXRK+stRKXXpNnzcs+adeh28A7y09TsGJ3ehW1zjSUtX9TuG19YsDf9AQJFezdAO7UmNicGSstHqlw39Ik1l2479ETt3+F0aMRbYFIHH7cfpsmF32Pjzw+cc0nGONC1Vm+JIycSmaMbDgW1Syh3Bg3wCnAVk11vnBuBFKWUpgJRyf9NOq+VxOGy8/PTl/DBzHXMXbMHu1Fi8Y29t6nMNXr/B81/8wqbykto+dzsKSvjHe9Nx+3Qmj+gbsn6is2We9IXeSl7fOpcF+7eRbI/hym6jOaV9eP1ht+Hh8U2vsK1qF5pQ0S2D0WnDuLX75SFZdk710Prx/dpIItcvAJP6TzMpJZ7qD0lVPFjCREpJtbQTE3Me7ZPvrxWiDq0dHRKuOuAx95T+hyL3VCR+zGA0SH7lu9jVDNolNK39TkuyqHhTRBuvbhn8kLeC23s23tGkXVw87eISyK0KL1epCIEvmDatCMGwzh1Yvjs3NIFJCEZ3a1pdkYYkJrhqE6waHJiBJ/Vi0ri+bNuYR7vMFMZN7I8rpnVqxQFp3CIP4SMiE5sSBd8B2FvvfU5wWX16Aj2FEAuEEIuDKnwYQogbhRDLhRDLCwsLm3DoQ8PptHHO5GN49omLufqKMdhtkZ85m4qLQxqOQiBe+Nkpv4TV2L1i6BBcEfYjAj18alGEYHj7zIhacbGvigvmvsQXu5ez113C2rIc7l/9FS9vnh227qs7PmZL5U78lo7b9KJLg4XFK5iW93NTLkGrI9Y5ntrpQy0Cp/1YRL2Qt+rqN6ms/A9SViGQKAISVJUU5zBUJZ6mIqVFQdWHIWFzAJb0kFvx6mGcyaHjM/Ww7EIIlETymA1nDeGc06Mvjgg1J5IcLjonJNW+f+j0k4l3OnAEQy2dmkai08E/Th13SOPu1DGVzp1SUdVQQeZwaFxw/nCOn9CPa++cwKRzh7ZeQRxEyqa9gLQaWRV83VhvNy0mE+vTUrmBGtADGAdcArwuhEhquJKU8jUp5TAp5bA2bZoWjO+p9vHDBwt48d7P+P79+XiqD37T1qdrZipmhALXNk1BF5ELX5dVe8ISQcZ26cwtxw3HoarE2e3E2m10SIinY3oCsXYbAoi12Uh1xfDv8adG3O+HOxZRZXgxZN1xPabOW9t+obJelw+/pbO4eFVY6Ua/pfN9/pwmnnnrIi3xAVQlBUEg8UAIF4qIJz3537XrSCmpqnwW2SDMTUoPVZX/bdbxLOnHkpHvFcMKL/5/MNxGCbuqFlHo3XLQYviNMTy1B6YMv+dcqp0T2g446PY3DRpOZnwiMcEHvV1RidFsPHNiaIeRbmkp/Hj7NVwzciiDO7bjtP49+fbmK+mYfOixv/96+Hy6dU3H4dCIjbHjcGjcdP04jhlyaNr20UEgraa9gKIaWRV8vdbMgzVJJjbc4GDkAvXTyjKDy+qTAyyRUurATiHEluBAljVp2I2wP7eEP57+FJ5qH163H2eMnff/+z3PTvsz6ZkHj4MEcDps3HzRWF759Jdap56mKsTFOEhMcbK3KLzuQ4zDjiOCFnzrqBFcPHgAn69dj9swOL13T7qmpDB79w42FhXSOSGJid16NJp6urhoR22j0frYFJUtFQUMTc0CAo1AG/u5e8zI8cKtHZvWgc4Z86ms/gKvvga71ovE2ItQ1frfo4nViKA0zYJmHU9VnDi1TLxGeN2HOHvTa4dIKVlU+Crryr5GFTYsaZJga8fkjv8h9iCV6xqS5kjkuq6n8NaOmeiWgYXEodhI1pJ5btOPpDsXc1mXsRyT0iXi9vF2B9+dexXf7djEory9ZMYlcFHvgbSLC58xfLM2m7eWrEBTFLYUFvPjpm08c85pnNAj8r4r3V6e/ngOM5duwbQsRg3I4i+Xjycj2DE9NTWOV1+8ij17iymv8NC9a6AU7W+OlnHgHRGZ2BRhvAzoIYToEjzgxUBDr+A3BKT/20KINAIqeuMdPpvIi/d+TnlJNVbQXuV1+/H7DF76x+c89O5NTd7PRROPoVO7ZD6atoyi0mpGDMriisnDWb4jhwc+mBFiqnDaNa6ZMCxiYZSi6mou+/xz8ioDXSfeWLmc47OyeO7005nQpTtrivK5bOYnrCveR7LDxY39hnNNn6G1WkuGK5ENZblh94NumbRx1v2gYrUY0h1p5Hn3haynIBiU1JffKqoST1L8NY1+LoSGqrbHNMMrlmlaV0wzH0UkIpSYJh2vS8pDbC68JaghS0BBEQ6yku9t8pi3Vc5mfdkUTFlndy717+HH3Ac4t/MLTd5PDZdljWNoSnd+yFvO1sp8VhbvpdJfBAi2VOazrHg7f+47mTMzIxc02lS6n3WlBTgdCiM6dCQjNjw6ZE1uAc/OXYjPMPHVi9W+4ZNvOK5jJs9ecDopsXXX0LIkNz3xGbvyS2prtMxbvYMFa3YytGt7LpgwhHapiWS2S6JTx+Y9gFoVssVC246ITDyoMJZSGkKI24EfCRj93pJSbhBCPAwsl1JODX52ihAim4BH5i9SyvA2zM1kxZyNtYK4Bsu0WDEnu5EtGmfkoC6MHBSqFZx6TC/cXp3nvp1PpceH06ZxzYRjuebkyCFuf5o+nV2lpRj1pqnzdu3irZUrOaF7Fhf/+DEeI5DjX+Cu5D8r57HfU8U9Q8cBcFW30czfvxVvvQ7QNqHSL6k9nWJDb/Jbu1/GI9kvYFgGJhY2oeFQ7Vze+axmn/tvifj4v1FW9iegfq0EG8LKp2zfWMDC4Tqb2KR/IcSBnZfJrhPol/4ROWUv4DG2E2sfQMfE24mx92zyeNaUfhGWKi4xKfRtpUovJO4Qal/0TsikXK/miz1Lg6FqdQLCa+k8s+k7JrUfHJYY8r/VC3h53WL8polE8uX2DZzdtS+Pjwo1R36xen2jkRPL9+Rywwdf8+VNdZ1OVmzeS25hea0grsE0TVau3suaFXuxB+PyLzz9GG66fOxvovFqRFpAMz5SMrFJccZSyu+B7xss+2e9vyVwd/DVYqiqIEK994hhNofKOaP6c/bIflR7/cQ47BE1YoBKn48le/eGCGIAr2Hw0Zo1rK3Oxdcg9M1j6ryzcQV3DBxFrM3OoOSOPDToLB5b911tnYmhqVn8+5gLwo7XJ6E7/x10L9/lzybHU0Dv+K5MzDiBJPuRSSZpDVhmAWblkzjR8KNjAQqBMqGaDLQ0MpH4PFMwzd0oSgZCceGIuQSbfWjEfcY7BtMnPTzutqn4zcgdWxRU/FYVcHBhXGVUM7NgHmvLN9LGkcpp7U7i5a3fY1iRu5dICbuqC+kRXxfqtqeyjJfWLcJXryaKx9D5Zkc2F3QfwDFt6/xHVT5fbY3jhphSsqOohM0FhfTKCIx9Z14xRoRoCdULSvBwNVmtX3y/kvQ2CZwzcfBBz7t10jIPkSMhE1t1Bt7YM49h7pQVGPVSLjW7ytjJzauNcDCEEMS5DuwFblgYqD4lHg/rivcFtZzQH4EqBDlV5fRKDtz4kzoMZEK7fuypLiHR7iLV0XgSQjtXW67velHTT+Q3jrf8MaRVhCZMtOCt2VBcKUhMfBj+JcElAp97CkLric/YjJR+HI7jSUx6DE07fOdSVtxo1pZ+iUVoVTNV2EiyH7x+doVeyd/WPkaVXo1f6ggEC4uWU+ITKMKFKcOjIwxpkmQLLco+J2cHkQSJ19CZsWdriDCe2Lcns7fuxK2Hd8RBBupR7KusqhXGWe1SA5086mvTUhKpq5jXZ/DRN8t+u8K49TSrDqNVC+ObHzqPHetzKNhTjGlaqKpCesdUbnrowOUzjwSpMTF0TExkR2mog0ki8Vo61d7Au4Z3r9vyk+wMLV2oCoWu8b9eaceD4TPyyCn7NxXeuSgilrbxV5IRfx1gUuX5Ho9vOTatMwkx5zVwuLUshm8mNGjKFC5+GqY0SMCDpa8JJoiAzzeHwsLTSE9fgKIkNfn4XqOQbWWvUuRZgIUdv1TwGBUoQkFgw5Q6AgVV2BiX8WcUcfAqfd/k/kiFXokhzeBoJbrUibWB29Bw6+HF5wcldaaNM3QG5FBV1AimAVUouBqEUZ7csxtDO3Zg8a49gRTqmssVrLjpN0z6ZNTlIAzr3ZH0lAR27SsJpDFLEAeIGCmvbDwNulXTcnHGR4RWLYzjEmN4ceY9rF24lT1bC+jYPZ1Bo3seNXvVfyZO5NLPPqudKtbEjBpCUlVlIGKCMYr1hqcgeG/zSv44cDQvb5nFp7uWUmV46ZPYnnv6n86g5KZ1JzlS6GYJ2QVnYASbgEIpueXP4PatRTXXYZh5SFmNEC5KKp6iQ5svcNoPHoZ1aBw8Y9Fq1OgnUahRfCyk9FBd/Qnx8TfXruH2LaO06iMsWUlCzGQSXKcjgs1XfWYx83PPQ7cq8FsWfmo6TUOiAj4ZQ4y9D6nOngxIOoc0Z+R6zmX+Er7P/5yNFatxqi52u5VaQVwfATg0E9PS8Zl1553pSuFfgy8JW/+UTj15YMlPYcs1ReGsrqFOXVVReO3is5i2fjP3TZuJbprIYEMQl03jgqEDaBNfp3lv3rufgooqUGuq4QVKEEkhEREu94AGxbZ+S0SLyx8GQggGje7JoNFNd7ocKQa3a8ekPr34KntDoEWTAKlI0CS+mvlPg+eEieSzbWupFMVMz12HN9gfLbs8jxsXv8OHY25qtJ7tr0Fh1YeYVjUhTUClB7f3W5wCahxpUgY6ihSU3E5WxtwjMhZbzHno1R9CWNr0ISC9GLVFiaCw4gWKKp5FSi8gqfLOpcz+MZ3afIgQKjvL38ewqrCkgY6d+l+kIsAl3CRpHk7M+HOjh6wyKvjP5r/jNqqwsKgwyvAY8TQWzi+lwGkzA0JZCgYkdON/Q2+NqGwkO128MO4sbp8zBVUEhKYpLR4ccTJZCeHdl1VF4ayBfRjdtROv/rKUWZt3EO90cPVxx3DWoLpSr5YlufulqVR7gxmSIvg/VSDigQqz3j4FdrvGrVee0Og1aPVEhfGhs3jGOj565gf255XSa3Bnrv7bZLr0bbzdj8frZ8asDazbkEtm+2QmTxpEakrLFYfpnZaGw64FyxlKcAS/3ZoGehGUdp9p8H3umrAYY79l8Na2X/jXkF/P7OIzCimomoLP3EeycwSV3qURakaAHYNId65ubKes9J/Ex9+IqmWGfX44OOP/iulfjWVsCqowkoBgrhuHgoIZqWgvDc2BLjRboNaFYe6nqPzpkPOU0o3bv4JKz48kxJxGsWcxVjBlu7Hfa7l/2wHHP79wJl7TE1JNz6V50XUX9btESwmGVGqXCQFxqo2ruh64ddjJHbuz7KLbmZO7A8OyOKFDF1KcBw7zS4uL5R+TTuQfkyIXCNqaW0ilO8LDT0BWtzbcccZopvy4hj25JfTpkcFV54+kY/tw4f+bIWqmODR++HABrz7wFT5P4Eey7OcNrF24laen3B0mkP1+g9mzs3nxtVl4DROfbmKzqXz4+RKuuGwkZ08aQkLc4ddxOLdPX/63dBE+E9BqBHHwfzW/4nrft01RGJHegQ3eqjBhbEnJ5ormJTM0Byklhd4VlHqzcWnpxGjxbNh/K5a0kPjIr/qCGMVG5NZKjYkkC3f1u3g9H5Kc+i4Ox5gWG69QYohN+xpTX4Glb0LRumAaJfir/oU080HEYRMOhFWM0WC8Omq9cxAIYSc2JuD8rPYuBKFBw04n0k2l5wcSYk7DpbWj3L+BA6lOTjXtgOPfVpWNIUOdZk5FR2oO3KYdu2LHlCZSqrj9KjX2cYdio0d8JkOSI1edq0+83cHkLgduYjA1eyPPL1xMQVUVvduk8dcTxnJsZuQHp2laoaVKpUQYICzYvaeYnftLeOQvZ6K1QARTVVk1374ygyXfryKtfQrn3DGJfqN/3WYAkcwurYVWK4xNw+Stx6bWCmIIaBQ+j593nvyWh96tswVu2VLAX//yMW6PH9MIWBUVFXQpQQjeenc+73y1hLuvH0/nLmk89ckcNu7eR3yMk0tPHsLVk45FVZp2s6W6YvjwnAv444/fsdNbEqoJ6wrYrVoNOUazkexw8bdjTuTi+eGx0YoQ9E5ovFLXgSjy5ZDv2UGSPZ1MV7gd3bR8zMu/jVLfJizpRxUOLOkmQfGiBu9IU7pxWzZiwrQxiS4FCjJCM1JQhA5Sp7zkdtpkrDysRp8NEUKg2YeBPZD0oDnAETsZKf2Y3nn4yu9EExJVKrX6p0CgCbCkioVE2oaSlPIMStDZqCixEbuDgIKiBFKEuyRezX7PL1h4sUkzKNzrTl4VTvomX3fAsbdxtGN71aYQzVgIiNf8jE+fQIXuoW9CL0akDGNa/kKm5y9BIjklYzhndhiD0gLX8YNVq3lyzjw8wTDLVXn5XPP5V7x34fkc0yHU1rtkyx7en72iLulJShSdWj+032/y8hcLWLkph//84fDi2ytLq7j12Hso3VeO3xvo3bjkh5Xc8vRVnHbd+MPad5ORAqLF5ZtPyf4KdF94aI6UsHlVXbcHy5L8497PqKysC8wXACYIBaQaWMfw6Tz11k/4EzV8wZuvrMrDW98vpai8mr9delKTxzYoPYNZV1zLuE/fYHdlWb3BCfApKBqc27Mvw9M7cWZWH5yajTMyB/F9ztpamzGAXdG4rvvYJh8XwJQGX+x9iq2VK1CEipSSFEcGV2Y9QqxW54HfXPY+pb7sYHdmMKQbkFRaTpJUT7396Ui1E3bK0a0iQKIiUWp0Y1nXHVoA9eNCpKzCNLaj2Q6u0R0uQtixjGwI1q0QQtSWHZJSIoIPDhUNRRaiqnVhbbHO4wkvUgRCOEiODTjLkp2DGJD6ENklj6GgIywTnUDfPoFG35TryYo/sEA6oe0klpfOx2/V789no0y389O+OeiWwdKSJawoXckdPW7mnMzjD++iNMC0LJ6Zv6BWENfgNQz+M+8XPr6kLkzy419W8ey38/H6DYQGqhnQhhsGBHn9BovX72brnkJ6dGpaBJDPq5Ozo5Ck1FhS0wMPu29emE5JQXntb1pK8Ln9vPKn9xl/6Vgcv1ZqdSvWjFtOpWlhEpJjG71ubTvU2aw2bcrD4wkX2gIQDQLZPZrE37BKm9/gm1/WU1HdvJoPQghuGHgsrgZ1KGyKytiMrjw1+gwu7D6wtnrbP/pP5qpuo0mwuVAQ9E/qwOvHXUPX+OaVfl5Y9A1bK1dgSD9+y4MuvRR69zIl97mQ9XZVTq0VxPVGjYmC1cBuJpRkMhNuJhZBDAZOYSKEwCQwkRZSwwXEEnBm1SClhRDhHYebgpQSq/oTrP1jsQr6YBVOQvoO7BhUtM7QyPHq7hUTyyrB8P1St51w0LnNR6hKMoqIQxHxCJykJz2A015XvrRD/OmM7zSXMR0+ZWLWDM7puoBJnb7hrC4/0yvpyoNG8WQ4O3Bdl7tJtqWiCRsKGtWGCwuJz/JjYeGz/KwrX8/CosVNuUzNoszjDatCWMOWoqLav90+nWenzq8thiVVMFyAiJwSIYANO/KbNIap7y/g4hEP8ZfLXuaa8U9y37VvUF3pYfG0FRGVK0UV7FjTWCutI4Bs4uso0Go1Y4fLzikXHcfMTxfj8+ohyy+9q644ut9vhPW8qyXoU0MBhMBUI19nm6aSW1ROQmzzbMqX9hnE+qJ9fL1tA3ZFxZAW3ZJSeebE8PY/mqLSP6kDXeJSyPeU0cYZi1NtfvH55SU/YMjQ2sAWJturVuG3vNiDXZtlE6PbFeGifdxFxLsGU1jxeIMLJJBIFCw0UWfjDG6JpnU7ZCeedL8LVc/UarqY25Gld0DyywhHeOt6ANV5KlQ8EtwmGMQma4Kx6u/cwmpQ38LlGEzP9quo9i3CstzEOkeiKuFVzBRhI85e1wvOpTXvYdk7YSAP9Huecr2UfO9+nt78AnqD0Daf5WfmvnmsKs3jl8K1aIrKpHbHcVGn8dib2RuvPglOR8DcFiFBKTOx7ly35BUGbMD1ZaMASw32z2vwI1EUQdvkg5cvXTZ3E2/953t89ZSjNUu288QfPyKxTeTMUVM3SUj9FbuvtGLNuNUKY4CbHjoPgBmfLkYIsNltXHvvmRx3Sl2cq8tlw6yfORTUXiRgqYFHvWUPLNOkwBDhsYa6adI+rfnlBRUheOL4U/nDMaPYULyPdnHx9EuNHKY2Ze8KntwwrdZM8cv+zSwt3sE7I2+kZzPsxrrVeNiXYem1wrhT3CS2lH+A1UBwqwhsSgxSGiAEqa4TaB9/LkKoJMVeRGnVB9QJ3UDsrsTEkCqacATjcgVCJJCU+maTx10fKU2oeqFOENfiRVY906gwFsKJK/VrfOV/wfIvRUoLGfwvxAeFRLUNjLC9jThny5oGIo9TkGRPocQfXhEQAvffypIiDFlUG4P8yZ6fWV++gycH3XJIx9xYVMhj8+fgJVwxcWoafxxd19YpOc4VVocCwHSA3S9CUqkVIYiPcTJiwMGzGb94fU6IIAYw/CZrFm/nrofOZN0vm/DVi9xQVIVOvTvQoceh+U2aTTTp49DRbCq3/etCrrv/bCpLqklJT0DVQm1/079bi6LX8/0LiaUJEAJVFehOAQhcThv9e2ayNCcvpFax064xaURvEpupFdcn0IWhcc3BlBbPbpoeYi+WBIqNv7h5Jv87tuldJ3rGH8vasrlYDTLVUuztiNHqxtAn+VoK3Aso929H1vuB2oVGZsK1ONRYEh1DiXfUeebTEx/B456O3wpEeKhBYSwE6CKGlORnEbIKRU3H7hiDaEIGWkRkBchGzELGgYv9KVpHXKmfIKUXT9k9WJ6vQrpnSEBRktDqJaaYZjHV3tmUVb2JT1+HIuJJjLuG1IS7a5M+jgRZsZ2xKTa8Vui5GpYTUxKSDOK3dLLLd7G5Yg+9EpqXCLSrrJTzv/g4kP6sgFBBMQVCQJuYWO4ZdzwndqvT9ju3SaZ7u1Q25ezHrNe/zum08cezx/DJ9ysoLncjLUn3Tm34162nN8nBXbw/2IWkvrYjBJpNJWtAZ6584HzefeAzNLuGaVh06J7Bw9/8pVnnerhEoykOE6fLjrNDuIF/985Cfvx2NVBn65ISFF2iOAWTJg7GcAr8fpOTRvfiuCFdWL01l39/PJstOUXEOu1cPH4wN04e2eixF+zczWsLl5FfWclxnTty86jhtE9sXrGeEl9VSKW2GiSwvjynWfsan34F26pW4TPdgS7PwoYqVM7qcEfIepriYkDKTSzd92cMaSKCjjmJztbyDzml8yyUoCCSUlJU+TzFFS9hySpAoGER2thBx+k8HqWRhp7S3I/h/hxp7kGxD0d1nYEQjdT7EAkgHCAjtGJSs5p0HYRwYnh/CF8OYJUGC8DrFJX+hSr3Nxj15uSWLKes6lUMM5+MlGeadLxDQREKd/a4hac2/w9LWvgsAylVvKYWZrqAgEa/rSqn2cL45RVL64pUCZA2MDWJXVX4/porSXKF29mfu/4s7nxjCtvzi9FUBcO0+OPkMVw8dggXnTiYguJKbJpKWlJs2LaNMWRUD/J3F2OZ9SSelFiWRYesNnS5ezKnXTeerat2ktQmgax+HRvf2ZEiKoyPDN98sTysN1dN5QJDQkZ6ApdcHCpoh/TM5OMHrsCyZKMV2mr4bPU6Hps5B0/QKZJTVs73G7cw5brL6dAMgRxva9zJ1dbZPPNIvC2F23u8yKrSn9njzibNnsmw1Ikk2sJjYPdWfoPAh63BaUqpU+JdTZorED5WVPEsRZUvhnTYMBAIKVFFoCtHYtzNjQpi078Sf8mVIA3Ah+mdhlH1Io60bxBK+HUSQkXG3gJVz0NIB2gnIv6PTb8YshFfQXAmUFr+KG73VMza/nv1Ey88ASGd+A+0g8QPHw69E3ryzOAn+d/W18mu2IaJiYUB9dKta1CEQltn82t/rN1XgNnQ9ibAoWnsLC9jSARhnJYQy0d3X8ruwlLKqjz0aN+GGIcNv25gWpJ2ac2vDjjxohFM+2Bh2HJVEbURObGJMQweF97z8deiNWvGrTaaoink55UGCptEQsLrb87l5TdnR1znYILYb5o88dO8WkEMYFiSap+fl+YvOcCW4ThVG2d1HIpDsYUtv6F75MyoA+8vlpFpZ3JRp3sYn3F5REEMNNp2CARS1oQYGRRXvhLW6ggEBho2rTupif8iOSF0OmmZJXgrX6K65A78JdeArKY2jVm6kWYuetVLjZ6DiL0O4v8EIljHWe0Eif9FOJpu09UckULWFFTHGMCisvoDJN6gMhSh5JCwoxs7m3y8Q0W3dDZVbMcMasMO1aDOs1EzaoVEWyzHJDc/7b97SipKhEgPv2mSmXBgodq5TTKDurRH1w3++tJUjr/9Bcbd8QKXPfwBm/Y0r6/wyvlbsNnD9TspA869VoEUTXsdBX7TwviYYV2wOxpR7pXArf7V1yt496Pwp/XB2FtaHrEojSkli3eHt/I5GH/qcxqTM4fgUDScio04zcldvSdyUkbTOnfolp8CTy5uo7rJx+wQdzpqhFAwiUmyczAAllWJjGQuAISIpWPGLyTEXRwS1mXq26jcPxZv5TPo3m+QMpKjyo/l/a7RsQkhUGKvRElfhEjfjNLmJxTXhCafG4Az8RGEkkRt9LNwIUQirsTHkNJXe14NBV8NUvqxaVnNOuahsLlyO1o927QiIN7mQxUBx6MmVAYmdeOZIXeEdP5uKrcMHR7WqNShqpzStTttYg5uZpBScvNTnzN71XZ008KwJJv27OeGJz+lqCxyPedIFOaVokeIbDJNk5J94V2tf3WaGtYWDW1rPqedOYSvv1hGSXFVrfYrCcRN1syLdMPk06+XccUlI5uV0pkc44pYcBugbVzzQ3Fsisq9/c/krj4TKfO7SXPEY1Oa5gD7ed/3/JD/BSAwpcngpGO5tPMN2JQDB8q3jzuVnKrvKPGuxJRuBDaEUBjc5lE0JSDAFCURRYnBtMIFssMWWUtzl9+DlJXURFs0VpOjqbfXoVbhU7RM4trOw+/+CsvYgKL1xR5zbq1pRFM7YZg7QxKl63AQ65qEph75Uqbxtviw37emSJLtOie0GcUVWecTq4U6kDeVF/DYmh9YXZJDjGbnki7HclufEyLeM33btOX1M87m/jk/s6e8DJuqcmHf/tw7pmkFfVZuzmFrbnHAzl7zXUiJ12/w5dy13HTWqAPvIMiAY7sy88vleN0N7iUp6DX4KNiHI9GKzRS/aWEcF+/k5beu46XnZzL7pw2BNtuaQNbzPEkBhm7icfuJj296xERKjIvju2Yxb8eukMLyLpvGjSMjt2VqCi7V3qxGjqtKl/B9/uf46wnLNWXLUIXG5VkH7gOoCI0RGS9Q6FnIPvcv2JVEOsafSYytLjZYCIU2CX9lX/kjIaYKIZy0Tfx72D6lNDD9y6B28i+CZTlkeFSDveXqVtQd38T0TEH3fAaAzXUh9tjLI0Z2pCb/i/3F14L0ogYjryWBabOBSX71t+RXf49Dy6JD0t9Jionc1ftw6ZvQg1jNhc/vC6n5oSkqZ7Q/KUQQb67I54n137GyeDcWYCGo0E3e3b6IXHcZ/zn23IjHGN2xM7OuuBaPrmNX1San9wNMW5wdGFX9h6IIhLht2LWvsc3CGDmhP+1e+pncnYX4fXUp1rrXx5O3vM1fnruSXkOymry/I0EjDeFbBb9pMwVAUnIs9/7zbC6/4QSkS8XSFGSwxKAMxmXFxNiJjQ149jfu2sfns1Yzd9U2jEb6hNXw7zMnMqZrZ+yqSqzdRozNxp/GjeHEHl0PuF1L8mPBlBBBDKBLnRWli/A1oVO0EAptY8YwIO3v9Eq5NUQQQyCLzmbrR1zMFahqJ4Rw4bQNpFPae8Q6I0WZKDS001rB9vOy3n8g0GKvaM6pHhQpJb7SW/BV/APLvwjLvwhfxT/wld4ajJ4IxeUcR0abr4hxTsShdUdTM/FLO26p4K0tPG3hM3aws/hOytwzWnS8NShC4f6+fyTD2RaHYselOnGpTm7tdjWZMXX1InZU7ueqBW+wsmQ3iIA5Q1UlimLhNQ1m5m1kn+fA032XzdYsQQywv7QRU4QQJMYeuANOfTSbyn8/vY3zbxiHqgiwLNANpNdP7o5C/n7RC5QWHmVzRdRMceS56ooxdMhM4YmnvsMwrFpB7HBoXH/V8ViWxd3PTWXFpr1YUqKpCjEOO6///SIy2yZF3Gecw84rF5xFUVU1RW43XVKScWiHf8kqdA+aUIjRDn6jV+hlEZcLBG6zGofaNG1fN6tZV/RP9rtnI7FIsg+kb8of2VlyF7pZjEBgST/p8dfSKfnvjZoOhFCwuc5A90wDdCwkajBTrw4NxTYQtREzx6Fi6SswffMIicCQbkz/XCx9JWqEPngO+2Dapr2FaVWxMXcgEgMrkCwfsp6UHnLLniAp5pQWHXMN7VzpPDP4QXI8+XhML11jO6E1yLZ7descfA1CIIUAVZFYlsSmqOysKibd1bJ9EHtktmFJ9p6IMmjiiN7N2pcr1kHfwZ2wY+HxhSoRhmEy49PFXHT7kbnGB0PI1h1N8bsRxgAnn9iXNmnxvPL2HHbuKqJNm3iuuWwMJx3fmw9/XM7yTXvxBR0Mft3E4zO45+VpfPDA5Qfcb1pcLGlxTY+3bIxN5Xncv+YrdlYVAjA8tSuPDD7vgH3wusb1ZG3Z8rCSlnbFTqKtaXVlLakzN+c0/FZdc9pS/0rW77sALRCYW7t8f9WHxDkGkxobntJdgyvxESxjG6axHQmYGKjIQFyxNFHsx+JMbn4b+4Nh+hYRsfC89GH6F0UUxrXbWhUgFOQBpql+48jWSBBC0DGm8S4Z68tyDtDJJFD/unNsy7e9On/cQD6fsyYkGUoAGakJjOrfpdn7259bEjGCSfcZ5O8sirDFr0g0A+/XY9CAjrz8dPj0+Jt562sFMdTYDiU7cosoLK2iTfKRzY8v8VVx3eK3qDbqhMmS4h1ct+hNvjrhjkbLJ57R/kI2VaxDDxaaAbApds7NvLzJJRdzq74LEcQAChZqhPoVlnSzr/JdYm2d8PhXo6ntiXMeH5Kp9n/tnXd4HNXVh987M9vUu2RJttx7t3HBBhsXqgu9hd4DBAJ8CSQhIQklQEINhN4JxXSDjQEDLoB7771JtnqXtky53x+7krXaVXMVZN/n2cfy7JR7Z2fO3Dn3nN9RlHhiUmZh6iuxjJ0oWi9UW1+kuQch4hGBuF1p1WB45iCtIlT7CBTbkMMqmSWUxECySKMwPOFAiOYfTDY1A0XEYNG0a0eiYEkdRbRdM+RI0Ck6mdzasrDfqVLDZkRz6ucvk+aK4bYBY5neJTReV0rJ26vW8MKSZZTVuumbnsYfTjk5RD6zIdmpCTx+yzTue/UrqtxeLEvSq1Maj9x41iH9Xr0G5xDuXd8ZZaf/qPDlqo4ZkZHxoSOlZPPK3VSW1tB7WGfiD7FqR51/+KAv2b/ci2Txxj1MHdNyILqUktyKSoSArLi4Nl2on+xbgdFIXN6UFoWeSlaU7OaElPB+6AxnJr/v/QBfHviEnTVbSbancFrG2fSK69/qYxfWfB+yTNR7dkNx+zawq/DswHp+zd/OaZ9g1xpO/Ak0+zBoMBoV2sEbzdTX4ym5BDBBetGFDdU+BkfiC4ecgqy5puCrejDMNwLNNaXZbYVQyEx4kNyy3yIsN5YUQfNVUoIhBCW1X5IaPe2Q2ndwX5Kt1VtYWbocm6IxKnkM2VEtRxNc32M8K0v3BGVrSgnCUvG6bdQEUsir9VL+sPhLSjy1XNMneDL5qR8W8cqyFUF6xle8/xEzfnURfdObFj0a2TeH2Y9eT15xBU67RmpC6+6zPbuKeO3ZbynMr2TEmO5ceMUYuvbLZvDYXqxeuKVe5EuzqySnx3Py1KGt2u/RIuKmOETy9xbzx0uepayoCkUR6D6Di35zKr+64ww8bh9b1ucSFe2ke58OLRrGU0f25u05y/FYgcqMDdZ/+N3vSE6MpqTWTXyUg9E9c9BUhSW7c5m5diOmJRnQMYOXVyynsKYGIQRZcbH8e9oUeqUeDI2q9Hn5as9WKnwexnTIoU/SwYt/d3UxXitMDKa0yHOX01x8RpqzA1d2ubnV560xUbZOwYlugNnE3K1AQ6EWGUgYkYBlesgruZku6TNbdTz/RNsNfg2K+oUGpvdHjNr3sUX/6lC6gVAScCa+jqfsJqjPqrPjTHoBEUaBrTEJ0Wdh09LZU/pnqnyb0Br4LHQEljQod88/LGMspeSN3a+wvGwpPsuLgsL3hd9yTtb5TMpoPlpjaFIO/xh8Pg9vmEWJtwZVCM7uOJTNhdUsrNodtK7b1Hli7QIu7zW0PtzNretBhrgOr2Hw9I+Lef7c5vulKIKOTcyfhOPzD5fx70e/rOs427fm8/G7S3jto1u596Xr+PTl7/ny7Z/w+XROnjqUi287Dbvz+Lx1+NvYvqMp2q0xllJy3xUvULC3JMj/9MGzc6nyGHz52WpUTcEyJQlJ0Tzwn8vJzmk6rfWKM05g7rIt7C4JTVDw+gx+8+wnaHE2BAKbpjB6cGe+2bIDj64jgU/XbkIqEssGCNhZUsaUN97m6hFDuHXkKLaUF3PVNzP8qdiWhSIEU7v04dExZ/iLqiZ24sv9azEaOS29lkGivfk6ZodL1/gr2VX5OsHvaIIaaSdOiICQkIkiXCB17CF+WROPbx2GWdyq1GFpbENa4V633RjuQzfGAKpjJFHpy7H0NQAotkFtEiyKdgynQ8LdVBX9Bq+sIThIWsOmtk0yszFbqzbXG2IACwtL+vg47wNOSB5JvC2h2e0ndOjLKRl9qNI9uDQbNkXjhA+fDruuaVkUuWvIjPZP6B2orAqbiSeBjYWty6YrKK2iqtZD54wkNK3p8+rzGjzzzzkHFwSO6/HoPPzHD+ialcDqH7aS3iWVC26awJCxx7a8UpNERsZtZ+/W/LATAR7dZOaMZTRcnJ/n44bzniGhUxITT+3PpVeMxRUVHMsb7bLz+8sncuezM0MEuP2jP7/oNoDUYea6zUHv8PXzXFJQV3zUsiSvr1zF1zu3U26vocYIngmftXszEzt25/ScnpyS3ocH1ocfWc4r2MS49LbNWrcFh5bC8PT/sKLgdmT9iFKhW+KfSY8aQUHVW/jMfBKc46ioegrDqg2zF1E/Wm6ZZoYfYQRywq4mTXTvj1jmPjTbwCAVNiFUVPuhv+7GO8egCBemDO6nIjTSYy5qYqvWsaJsWVClj4P7VlhfsZYxKS2newshiLMfzJzsFJNAsSc081ICSY6D66XFxIQ87OvomtT8xF9JRQ33/OcLNu7OR1MVFEXhD5dP5NQmoikWLdwSNpwQw2T9t+vZpCqYusne7QVsXLGbG+6dzpm/al3yyFGlHRvjdhtnXFPpRgmXMed0EE6OwjQsig5U8NGMpdxxy5tBAkKV1W5mzVvPjp1FWEboxVqftRfA0qC+zEUjsXXF9P8byHTAsCT57ircRqgqW62h88G2tQAUeiuJUsMne6wqbXt6dWuRUrK5/EPm5/+DYsOOmxw6xf2GMzqvIif+Qip82/HgwukYR1L0VOKjpiMIbadNy0RTm54EaojQeiJEOElRF6qr5UrYlplPeeHJVJVdT03FX6koPpfKkl+14WHQPIqw0S/9vzi0bBQRhSpiUEUsPVKexGXr3Oy21UYpa8q+ZE3ZHGqM8pDvbYodEea28tfpO7RX9NsGjsWpBo+bXKrGJT0G11eSAX8o5kUDB+BsFH7p1DRuGzOq2WPc8eSnrNuxH59uUuvRqa71cv9rX7NxV2jB3Nx9Jbz60rywdk3UesGSmPrBh67X7ePlh2YeTAQ5jtSFt7X0OR6025Fx9wEdwwv8aOFiAAJIie4z2Z9bxpJF2zlxbE8WLNvGfU/PRhECKSWqYeJwCbyNem4GrmkrUIdSNBjASZX6MxUSGSPAa5qo4UYJUD9SSXfGBenXNqRTdHJTPTpsNpW/z+qS5zEDkz+1Rglry94hxtaFfRVP4jULMWUtinCyvezfDM/4DzbP1+jmAaSsReAXlM9MejpM0dNyPL5VqEoCDtvg+u+FUHAk/gdP6RWBkbAHRBSKbVCrXBRVZbdjmbk0rCyie5fgrv4PUbF3HJHzEmXvwdDMedTqm7Gkm2h7fxTRfGbkmrI5zM3/DwoKFhZfH/g3SY6ejE29hJ6xJyCEYHTyicwr/Ba9kd6HxGJgwuBDauu4zK48MuosHlgxl3KfG02oXN5rKL8bPD5k3T9NGEeM3c4bK1fh1g2y4+O4b9IpzUZT7MgrZteBkiBtYwCvbvLuNyu5/4aDYY7uWh+33fwGVRXh3p5A6KGDkjrydhbSpU/rHuj/i7RbY2x32rjloQt45p730b0GliVxuGy4Yh3UogSVYqonMJJ2u31s2bif/oM7cd/Ts4NC2gBUjyA5zYUhJB0zEthQVASW6RdsCigbBpkds07vIngEDf5lNqkiCTW0UZqNc7v5ox6SHDGcnNabhYVbgibynIqNa7odneoTUlqsK32l3hDXYUovy4v/RZwoQAZ0fi3pwcLDhpIHGJX5DZW1X1DrXYJd60RC9EUhGg6lVf+htOKfCGFHYqEpqWSlvotN81eEUO3DiUr7CcP9GZZZhOYYiWIf02IlacuqwvAthZDz6cFb8+4RM8bgdwdE25sve19Hha+AufnPYUo9qGXFns18uO+f9Ikfw7nZvyU7qhNnZ53HJ3kfogglUJXa4sZut+JSD61eIMC0Ln2Z2rkPVbqXKM2O1kSWnaoo3HnyGO446UR0y8KutuxPLy6vCavbIqUkvyQ4Y27e9xvx+Qx/bRVN+LVqAwhAc9gwG1Tz8GsUgO4ziE8+/Fj9w6YduynarTEGmHT+SHJ6duDz1xdQWlDJyMn9GTd9GPfc+AZ5e0vqDbIEpE2tn0RwOm2kZ8Tz44odYaUyTdOittyNcKoouqR3VirbCkqoDiMADwGNZAssB/VWWiLrQ+RsispfRk3gvqVz/ZEElkmUZmNMhxymdD7oc3tg8Hk8uG4mcw6sw7QsLCQeS+eFbd/z5wHTyYxKOHInDzCkGz2s/xc8ZiWxWmh/q307MaxaEqLPIyH6vLDb1noWUlr5GBJvvetAN93kFV9GTvqCgyNkJQFb9JVtbXXT35h57N3fl/jY24iLufGwYpbbyubKhcgm/LFSutlU8RN5SWeRFdWDyRmnc0LSSNZXrsMmNAYmDDksQ1yH35ccPuPSaxjM2baNzUVFdE9O5qyePXHamnaL1Hp9PPHpQmYt24SumwhPaN/sNo3R/TsHLdufV4aniQLAqqpw+W9P490n5vjLK+kGdVU/pLRY++NWxp9z6Louh00kmuLw6DGwE3c+Hpwh98Sb1zPnkxUs+Go9mzcfQBcgVQUpwLIpVCuSJ56fiyPRidcMvbkloOsWliLZsqeQUf07c8Ulp/LC90vZXFgU8vBUFUHH1Hh21pbXi9crmsDuVIm223ns1NM5Kacz4zt25bOdGyn3ejgpswujMjoGGQynauOMzMHMzlvrrzMW+GpJyQ4u/+kFvjjlDlxN+JVbosxXyMYKv85y3/iRJNrT0IQLuxKD1wqNINGaMWQiTFn7hpRXvxZG/9jCMA/g0zfgsLc+BroxipKIqnXFNLYELZdSYiCxrDLKKx9FSi8Jcbcf8nHaiiXDVOqsaxtgWD62V68kK6oHAAn2RMa2YrLuSFBUU8O5/32Hco+HWl0nymbjXwt/4KNLLyUzLtR3L6Xk189+zKZ9hfgC8feqJlEa2FibphAf4+S8UwYFbdujZwdcLjvuGg/CCK49aJoWb77xE6Mm9GXRZ8uDJvhMw+TJu94hvVMKfYa1PavviPFzHxkLIU4HnsKvEPOylPLhJtY7D/gQOEFKufyItbIRDqeN6ZeMYvolo8jdV8o//vYpO3cU4tGoHx0bhoWvuBYr/uCIuSFS8/8uPsvip/W7+Pv1Z9AlPYmLnns3JNpCUxRevPQckmJcFFRVkxkXR4XXf+F3SUysDydKj4rlhv4jm2z3ipI93Lr0bUysoCZZUuI2fHxzYAPTsoe0+Xz8VDSLr/LfrL/Ovs7/L6dlXMaJqVMYlHw9K4qfCXJVqMJBt9gxlNd+iSm96FLFCmiuJTt6YlOb1z4wG2Xz1SFQMcMY/rYSk/gUlcXn+4um4kFKf4KKt36U5aai6hniY2855AQSj1HMmuJ/cqB2PgKFrJhJDEy+C7saHK9ca1SwumwmO6qXhaSk11FhODFRWV2+mhNTzmlR2rQ5qnUvD635mi/2rcewLMakd+W+IWeQHZ3Q5Db3f/c9BdXV9dU+anUdj2Hw57lzeeXcc0LWX7+ngK15xfWGGMC0CzSbQkZ0NHZVZeygrlxxxgnExwSP6E8c24OU1FjyKtxhz4buM1j8w3ZkGPlZn8fHR8/N5d6Xr2/dyTjCCNp30keL0RTCH8T5LHAG0Be4RAgRoogu/NPntwNtK4NxmGR3TOLZl6/hutsm4WwUUK5IUNxWvd8KAmFsGhia/19LBZ+Aa/41g5SYaK44cQhOm4YiBIoQOG0aN44fSU5yArEOB91Tkomy2+gQG0u3pKSwcZ1N8dTGb9Gt8JN4taaPXVX+WNBibwUbK/ZQbTQefYZS6ivgq/y3MKTu92dKHUP6+Cr/LUp9BfRKOJ8RqXcRpaUjUIiz5XByxoMMTb2PGEdf3NKBjoqJioFKkW8f+bU/NnvMGNeZCEJflyUGTnvbHyaN0Wz9SUj7EVfsXfikhkdKahq4CKSEWtPDuty+rNnbhZ2FV+I1Wh+RYlpevs+7nP0132JJL6Z0k1v1JfPzrvZXrg5QpRfx2s7rWVLyPgWeLdSF7Enp/1gSqkwHZqCEUq47l/f2vnDI/ZZScsWCt5i5dx0e08CQFgvzd3D+t69QrTcdSTJ3586QskuWlCzcvTuo0nMdO/LDP0y9wmJAv2w+eOhqbr9oHIlxofHvmqby7+euQhGCpq58wxfe3SclFOaWNtmPY8LPXLVtBLBdSrkTQAjxHjAd2NhovfuBR4BjW+41QGFRJZ4woTOaV2JJC9OuoEWreDWJUacC2cCQ7s4v5d5XvuTZ28/l1P49mLNuG0LAGQN70btD2wTIN5YUMnPnJiSSKV16MyAlA4DtVUX+iY8wv7ZLtdElJpU/rXmNZaVbsAkNXRqc3/Fkru92RpP+0Y0ViwNKvcFIJBsqFnFS6tl0j59K9/ipIetE2QaDewsNJ8ss6WVl4V85I+erJifb4qMvo6LmHQwjF4kbEAjhJDX+byhK8A3s1XeSX/5najw/oggH8dEXkh7/RxSleR+qoiYTFftryt2zMHwrDvZLQo1U/C0OuEqqPPPYlj+VPpkLUFuRibe/5jt0szJo0tXCwG0UUFC7iIxovw7zD0Vv4DGr6s+vEBZI0BQnlQbUWHaMBjO6htRZW7GUGqOaaK116cTry3P5eO8KagwvXaPT2VFZjM9q2C6J29T5ZM9aLu9+0N+6qbiIt9evodhdE9bg+tsb3mB2SW9ax+PHNTt5/9tVXHDK4CZLk8XEOnG6bNRWh68Q09QYT7NrDD7pOCZ//AJU27KAfQ3+nwsEvYsLIYYCHaWUs4QQTRpjIcQNwA0AnTq1rQJuS3TpnIrLacMdLspCgmpI+ndMZ1tNBSUeT8hT0JAWy7fmUlHjoW9WOn2z0g+pHU+v/oln1yxGt0yklLy+cSVX9xvGPcPH0Sk6iXXltf5DygbPAgkJ9ijWVGxiWekWfJaBLzCR9fG+hWS5kjkry3/KLWmxrmIjZb4yusV09U8khrvAWvGEP1A7L2wUiG5VU2PkEWMLr6egKNF0SH6P/IqHMPQ1OLSuJMf9BpcjWDXNMEvZVTAFS1YCEkvqlFf/F5++lZy095tvXIDEuD9QWHIZMuBm8Yd/N77ZLSzpprT6Q1Ljrm1xnxW+rRgydGLTlD4q9Z1k4DfGu6qXhTzohAAhTISSg2GGjjBVoVJllLfKGL+140ee2fotPtPAQmITG5GqBDMQuhPAbepsLD9Q//+Pt2zkj/O+QTdNTCmxhXloaorCpG5dwz7EB3buQOf0RLbvL0avcycELsrKcg///nAh23OL+dOVTZfBOnlCX+Z8vjr4Ggu8gSoCcDrAc3A0r2oK0bFOzrlhQovn5ajSjifwDjvpQ/iHT48Dd7W0rpTyRSnlcCnl8NTUI1vuZvzJvVG14O7IBn85HTYuO2ckMx64CpfDVm+sRMOVpX+W+VDZXVnGM2sW4zENTOm/jT2mwWsbVrC5tIjf9DkFp2rDNBWswKsuErrHpvPqqGtYULQOXyP9Co+l897eeQAUeYu5ffU9PL3tBd7c8x5/2fAgS0t2hHWVCKGQ7Mjgu4J3WFj4IWW+0IoNNiVcYgYY0kONHhrsX0dp7XyW7z+V3Jr5HPCWsKd2GQW180LWK6t+OxBt0fCX8FLrW4HH1/jFKjwu5xjSkl/DpvUGbKAkIQiNEpDSjVtf36p9xtq7hq0NqAo7sQ2SPhxKU6FYki7RPcImd0gg2d5ySnWJt5qnt8zFY+r1spm6NBHCQlGCn6Iu1UafeP/blcfQuXfeXDyGUe+a0FULBNhVFZuiEG2zkRUXx98mTQp7bCEEL/7mfE4f1st/7UjAkqge//3g8RnM+mkjhWVVTbb/mlsmkpoWdzBlOtAW4QncPy4nWmIsHXtmkNEpmdN/NYZn5/6BxNQjq8XcVo5U0ocQ4nQhxBYhxHYhxD3NrHeeEEIKIYa3tM/WjIzzgIZDpOzAsjpigf7AvMBTOAOYKYSYdjQn8Rpjt2uojWIl60yUKgW/Om8Eo0/wq4qlxcWwt6YUpF/BTdRVBrEkGYnhDVRr+Gbv9rAporpl8vXebdw2+EQeHnouj67/igPuClw2J9d2H8N1PcZS5qtuciRbqftTYZ/a9jxlvvKgiaT1VTsZkzScfbUrsAL+TkWoZLly+CT3cXTpQ0FlftH7TMm8icGJE+u37RZ/CauKHghoU/iR0l/qZ3Hh/XRwDabCt51ERx96J15FjK0jplXLpqLfYDWIppAS8ipfJ9E1nnjnwdGxR1+HDCNZKVDxGttw2ltXjNXlHE9WxngAarzL2Vn4q/rq1vX7FC5cttbtLzt6MhtKnsY0vdQNlQQaDjWZjKgx9esNTTqb+YUvYzTI/FPQ6Bx9AmPTLmJj5Wp81sFSSnbFwekZ5zc7gVfirWJ7VT7bKgvRFAVf45GaAJsq8Vp1xxM4VY1zOg8EYHVBfqj7QIBpl3RMiOeCXv3plpTEuC6dm634EetycP/lp1O4v4LlW/NC3Bl2m8r23GLSmrgfEhKjefDJS7j1spf8znPTRHiN+seTw2XjlLMGc9vfzj6mIYgtcgTcFA3m0Sbj9xQsE0LMlFJubLRem+bRWmOMlwE9hBBd8Bvhi4FL676U/tLA9eoxQoh5wP8dS0MMUFZWg9sdflQbF+Xkyov9efHfLdtKbm5p/cUnLJCKBNU/Qiiv9pAYe2gxoTZFDXvhKUJgDyhrTc7sy+TMvuiWiSaU+vUT7THE26Mp9gZHIwgEgxK6UeYrJ7c2L2RG32f52FpTze97Pc6GisUAxNuSmbX/WXRZJ1ZjYEn4Yv/z9Iw9gSjNPzrJiZ3G6uInsBpEQEjwS8WbeeypPgBYlPu2s6/6K07JegXD2INAwZQCnbooDIlNeimo/jjIGDu0PpTJb5HoaA0iSCQmjgZym23BIg5VzUAa++oTVkBBEU4SYy5o1T5Uxcn47LdYXfQQBbX+yuEdoscxOOWPQaJDgxOnUOTdxYaKb1CFHUsapDm7ckbmXTjVWO7o+SCzD7zPrpotxNoSmJx+DoMTQtOOaw0vs/cv4709C8n3VGAXDjyGhdcMc60g6BiTxJ6KGgxpcmKaP5oi1uafMI2x28P7iAV0TIjn+hNaHIAF0TkzmdXb94dk3+mmRYfk5kexH725CFM3EKb0N8CmIU0TVVP567NXMGhkeDfJcePITc4dlXm0Fo2xlNIQQtwKfIV/2utVKeUGIcTfgeVSytbpKh5loqLs4X2nQFKi/3WzssbDfS/6laaCLpGAQRaCNlWQbswZnXvw0LJ5IctVoXBml+CJi8ZVfoUQ3NnrXP62/m18loFEogoFp2Lnhu5nolu6/8IO00fd0klxZDIuzV+s8ou850LSccE/Yt5WvZJBCePrj6mpGdRatShILAm6VHEoRuD81A3bTAzpZnXxv+iXcA6mNPEGIggAJAIfggrvtvpj1fi2s7PqM3yWRl3NvBjhxaWoOG0DcLYxFtlnlrKm4Hpq9B0IBC7AIVRAEuM8iezEB9CUhFbvL0rL4MQOTwcSOURYoyGEwqkdbmd0yq8o8u4kzpZGiqNz/ffpziyu7nJns8ep1Gu5bunTFHoqMKQZCNFzYwkFUx48h3XYFY1Hh51H34QspJQh7eqXkkaKK4p9ekXQpeDSNC4fMLjV/a/jkklDmfXTRswGWao2TaF3pzS6ZDafpr9h9R4ss0ErFIFUNOzRDuKTY9qXIQ7Qhgm8FCFEwwHli1LKFwN/H7F5tIa0yvJIKWdLKXtKKbtJKR8MLPtLOEMspRx/rEfFAC6XnVPG9cZubySS4rRxyUX+8/TD6p1NzhCrFgzqlklsVOsLMDYmPSqWf5x4Kg5Vw6XZcKkaDlXlvpET6RSb0OL2J6b246lhv+ak1P50i+nAlMyRvDLyTmI1FzPzFlPkdlLmdeI1Dxpym9AYlRyc1eRPww2P0ugn7xo3BVU4MaWCV9owUVCaGD6UeNaR6ByLN6zGhqBC34klDaQ0WJN/JV4zPzAYEUgEVdJBtakQG3Nri+eiMRuKfku1z68jYcpaqqWNMplAWuK/6Zb2No4WBH6aQjR4O2mKWFsKXWNGBBni1vLunvkUeSrrdUn8E4CgKhYO1Z9IEq3ZiVbtOBSN2/tMpm9CVmDdcA8IwWtTziU9OoZom50Ymx2HqnL94OGMz2l7MkXnDkk8cdvZZKbEYdNUbJrK2IFdefL2s1vcNq1DQtjlum6QmHx0K+ccMq0PbSuum98KfF4Mv8NQ2jKP1pB2n4HXFu68/TQ8XoPFS7ajaSqmaXHJhSOZPNFfxSOc8BD4xyYup537rz39sNtwXo/+jMvuwtx9O5BSMrFjN9KiWn9h9o7rxN8HXln//0q9hhuWPUqlXoMuJaBR6VOJ0nwkORSS7YlM7RAsWj4wYTyryr6td1PUYUmTHrHBEQ+9Ei5kf82PFHk2BpIsmsamxqCpcSCiIEw0gsRCN8uo0bdiyVoaD+OlFFShsqrwVmLtfemf+g9i7C0XLfWZxZR7VgX5tv39cZNb9QYdYg+vMsfRZEHhBvQmzquiQJwquLvvFOyKjREpXUlohbZ1t8QkfrziepYeyKPc42ZYRhZp0Yeu+3BCn0589vC1lFbW4nLYiHK2Lmnl4mtPZuOafUE6MXa7xsiTe5KQ1A50KMJwhNKhj8o82i/KGDscNv7257MpK6uhqLiKjtlJuFwHL6zRAzsHZR3VIYBHbp5CelLzk3c/7tjDW0tWUeZ2M7l3D84f2p+5O3bw0foNqEJwwYD+TO3TmxRXNBf3HHhE+vRJ7gIq9Vr0BqNRicBjOrm049mMSxuNTQmOLsiO6sXolOn8VPwpEhkY60rOy74Tpxp8k6jCzsSs//BTwSNsqZwFSHxSxY4ZlCWoCifd4y8GINrWhUrfhrDt1ZQ4DLMs7HeiQYh1lW8Tyw5cxtjsr7GpCc2eA8OqQgg1rBtKPwIZf42xpMWemm3sc+8kVkugX/xQ7MqhvTFFt1ABPMkRzVlZg9v8Sq8qCqOzWi7l1FqEECTHt82ADj6hC7ffO5Xn//klXq+BZVmMmdiH3/65nT4cj5zP+KjMo/2ijHEdiYnRJCaGXlguhw1hyZAfRFEEufnl/mdZE7z4w1KeW7AEdyBVelN+IU8t+QmpUF/mZk1+Pt/u2MHTU5uvx9YYKSUrSvbyxb51AEzpOIBhyZ0QQrCsdHPYkZVTcZDuzAoxxHVMSP8VgxJOYWvVcmyKnT5xo4nWwidECCFIcw1iR9W3GNKNT9pQAI2D0RmdYs6gT8LVAHRLvIVVBXdSbUq8UkMTFjEqdI69CFVxEO8cjiVD4739YagHg6stqbO/+jNy4q8MWbchLq0TqnAGRXCAPwIixXVKs9u2lY2VK3l7z1P4LE/gGAoKDm7q/ge6RLc9YeH8TmN5dOOHeKzgunYgcCp2/tDv3HbpW22MaVps2JiHz2fQv192fbbrxLMGMf70ARQVVBAb5yI6JryQUXsgoEJ+2BytebSflTGeN3MV7z47l9KCSnoO6sjVvz+L7v2yWr39hh35OO02quuiLgIjNcuUfLdsK+dPGhx2u3K3h2fmLw4aVXssE9MkKIi8Vtf5bsdO1ubnMzAjo9XtenjdV3ywe2V9IcqZ+9ZyYedh3DPwNFIc8WwJE+5pSJNEe/iRfKGniMUly/BZPoYmDqJrTMu+xM6x41hS9HTgQSXwSDtCSmyKg+kd3yTWfrAYaax9AKVWMj7pD7nTpcRraAxw+CMJHFoGWXFXklf5en3EQ11GutXgdrCkhxp9V4ttE0Kld8oDbCi6C0v6AAsFB5oaR+eEG1vcvrXsd+/mzd2PYzR4kFjSwsLNSzsf5f7+L6C2UQtjcvpgNlfs49O8xahCRbcMFKFwYko/ru46gW6xrb9OGlNSW8uszVspd7s5MacTw7Iy22zYi8urMS3Z7Fvh5i0H+OO9H+ALTPJZluR3d53BKeP9oYSqqpCR2XRWX7viCGXgSSlnA7MbLftLE+uOb80+fzbG+ONX5vPm43PwBuT7Vi7cysblu3n8w1tbJVhtWRKvR8c0GpiDBj9MXHTTT/R1efnYVTXIGFuNhM0CNoxay+Cmz2dy/4SJTOzWcvjW1ooCZuxegaeBupzb1Hl/93LO7TyEU9NHsqxkE74Go2MVhS7RHciOCk0uWFj0I6/tehuJhSktvsz/mrEpo7mq82XN3qg2JYozO/6b7/b/mRqjEInErsRyUvq9QYYYYE3pa+hBpYUEFiY/FT7KhdFjEUKha+L/EecYwtaSe/Ga5VhSIkXw2EQVLhIcg1o8RwCpUZMY1uF99lW8gcfYR6LrRLJjLw1xcRiWl5WlH7Kx4isA+sRPZljShWitcDPML5oVZIjB71qREkzLYHv1JnrFDmhi6/AIIbit1zQu7TyODRX7SHbE0i+u02GPhhft2csNH3+GlH75zJeWrmBMTieePXtKs/HFdewtKOMPL8xi5/4ShIDM5DgevOEsenUKvqZ8PoPf3/Me1dXB8w+P/ms2PbpnkJ3dfDmn9kZ7Todut2WXGqL7DN5+8ut6Q1yH16Pz1pNftbj9wiXbOfva5/jro59jlfrQai0aOiCddq3JUTFAgssZEtvZ8EeV4D+TAVuTX13NbbNnMWP9uhbbNj9/G7oVOqugWxb/t3wGv13+AVU+f9ieTWjYFZt/km/AdSHbVOnVvLbrbXSpB8KoJD7Lx4/Fi9hctbXFtiQ5ujM562ncModiPY48n53/7n2Ib/NfDUpmya35KWQyzd/mWqr0/UDABxl1Cm5lBPlmKh5pQzYokyLQsKmJpEef0WK76oi196Zv6j8Y2uFtuiTcHGKIpZR8vO/3LC35LxX6fir0/SwreZeP9t7VpBZxQ4q9TWcdgsSwwgvgNMf2qgI+3beSrZVFjE3tS//4nMM2xLppcutns3DrBh7DQOKvDP3jnj18sWlLi9v7dIPrHnmfrfsK0Q0Tn26yO7+MG/85g8qa4CSdxUt2BIevBTBNizlfrT2sfhwXWh9Nccz5WYyMiw6Uh81sk1KydV1us9tu3p7P3574Am8DESHFBLsXbIl2dMPk2rNHMbxv01oZ/TPTSY2JZl9ZRb1RVgyw6uYG6x5pDe4xt2Hw8MKFnNe3X7MjFYeqoQkFs5GxkFKyt6YEFBOfTwFfNFGawt19pzK9Y3iB7nUVG1CFgt7oVHktH4tLltInrmWf50f7HqRUL/ZrMgRG4ytKZ5Hp6kmfeL9mg12JpYZQwyUxsasHI0fyan6kyLMOU+qUyWhiFA8udARg07oxMO0xVOXI+Rj31a6i2LMDs0GMtSl9lHj3sLdmBTkxzQubd4/px77aHVhh9DosJN1jWpfhB2BYJvesmsHCwq0BhTNBvD2KV0dfi5SCj3etZVdlKZU+L1U+L4NTsriy13A6RIUmWpS53by1bjWL8/bROT6BERnZmGEe4G7d4KP1G5ner/nqJQvW7MTj1UMmRHXT4qulm7nglMH1y2pqPGGTTAzDoqKyZVXBdoVs3+LyP4uRcWJKLFYYfVSAjI7Nvya9++myel9XQ2xS4eqzRjB5dG+++GEjt/3zI5ZvCi/DKITglcvPpUtyIi6bRozDTpTNzg1Dh5PodDY5M+A2DIprw1faqOP0rH5hl1tYSNGw3YJaQ/Lx3lVN7ksR4RsiECgtlDsCKPcVUOTdGyKOo0svS0sPzkn0TbwITQQbUQWNdOdgnA1Gq7k1CzHqJ90E1ZaLIjOOAiOOXe4yPt13BzV6UYvtai357s0YYZJddOkm37O5xe3HppyBU20UWib9scgXdbwRh9r6B8c/1n/O3AMb8ZgGtYZOjeEj313O9T+9xalfPsczGxbyya71fJu3naVF+3htyzJOm/Ui2yqCz0dBdTWn/vd1/rN8CYty9zFj43rumfc1epgHBoSV7g4hv7QKvXFhXinRa3Q+mrWKp96Yx+5cvwjSgAEdscIYfpfTxuhR3Vt3MtoTkZHx4eGKdjDpvBP49uPlQTGNDpeNS28NVpaqrvLwwvPf8f13GzBNiZLiDBsSZbOpvPrZYrzSwjAtdh8oZfXWPH53xQSmnhQcVqGbJl+u2oK3UsdlaAzpksm9U06hQ2Icd50yltPeeJ1d5eUhxxBAgrP5GzjNFct9g6dw78rP6kcgEgLGM/SGK/RWUeKtxqnaQsKmBsb3xwojS2VTbIxJbr46MIDXqkERatiL0WNWo1s+NlYsodRXS5zjJMo889EUG5Y0SbR34eQOfwvaxq7EIQhfH1Bi4jVrWFHyJidntCk2vklibSlowo7eqOafTThxqgn4LB/2ZnQjYm3x3NnzYebkz2Bj5QqklORE9WJa1hWkOTu0uh0/FW3no70rQk6jacGWknIkAssIfnDqlolhmdy/Yi6vjL+wPlX+qaU/Ue71YAQMoiklpmmiqCb+x+zBfbhsNi4Y0HJmY/8uGWiagm4Gfhcp0WpBsWDvvlLy8sr4+OvVpEU5KTpQiSL9k3R1FdedThu9+2QycsTBOZGNK3fz3P2fsWPjfmJinZx99UlcdNOEEL2Y40179hn/LIwxwK/vOxubXeXL95YgLUl0nJMb7p3G0JMOJg1YluTO377N3r0l6IFS4Wa5GxyhZYQ8Xh2fTQSZCY/P4Il35nPG6D4H1aiAO96exaJte+orgPywcTeX587gszuvwGW30TUuiV1l5UGDUqemcWG//ji0lk/xgv27sAwbBnWz1XU3mYXdfrCFCoIKXw1nff9PLCnpHJNM99g0+sV3ZGr2MOJsUfy623U8t+MlBAJT+pUjTs+YTPfYlicTUxydUMKUXFKFjU5Rg/jX5pvwWR50y4dNsZNoH8zUtHNIsGeT6Ajdf/f4KWyp+BAzTMaeXwbTYF/N0hbb1Vq6x57EgoLnA8kuB+86XRq8tee/SN6je2w/Lul4Iwl2f6qvz9Kp0MuJt8VjV+wk2FO4uNPNh9WOpzd9E3ZwZVkBQaomDIIEfjiwi56vPE6SM4o7ho3hu9276g1xQzRFxWZTEQh008SmKEzo1oUze7ecRDOgawc6piWwc38Jhmmh6H5DXIdpSUzLJK+iGs2S/qKjgZ8wOtrBeecN5/LLxtQb2l2bD/DHK1+qHyhVVbiZ8cL3lBVXcct9oZVGjisRY3z4aDaVm/5yNtfcPYXaag9xiVEojXyxq1ftZv/+8npDDKC4TSy74s+bD/wQTocGThV3mFhY07TIK6ogp4Pf/bE9vzjIEAMYlkV5jZsvVm2iGp2lO/f5BYfqQn4lJNud9E1JpaSmluToprOq3IbOnNzN+CyLhl4jCRiGVm+MVQQmErfpQwiJKiS7qwvYXVPAD4WbeX3nPF4ffTPDk4byROyjLC9biW7pDEoYQIazddrMqtA4K/M2Pst7DFPqSCw04SBGS2BPzS5qjMp6F4bP8lDs3c+mql2cmTmufh/57t1srFyGJjT6J5zIiNTfs7ToUYwG0RdeaaPuyeVqVObocLApLs7PeYIv9z9Auc8/l+C1oNAXE7AlJlur1nH/xt/SNXo4NjWaZaUr6uX+T02fxHnZ57TKpdMce2uLw38hCMgqNU3dW3KJp5YHFn+P4mu6LXOuuZwfdu+jwuNhdKeODOzQcphcVa2HGx6ZQW5hOZblj/pWm0+8RDR4ltbUeJkxYymnjO9DTo4/r+Hd/3wb4gr0unW+/mAZV95xOjFxh1+M9UjRnkfG7esdohXYHRoJyTEhhhhg9+5ijEYZdkKCVqnTOT2RlKRouuakcNdNk+nYJbwIimlaxDdQbduQVxhWL9itG6zYlcdrS1fi1g1UU6B6QPH6JwgPVFTzwHfzOPmFl5ixtumoihqjaf1kl+pgZHJXcqKT6RyTQl2JP6Xuigo0y2PpVOpu/rnR79eNs8UyNvlEotV0VpftocATPiMuHL3jT+Tqro8xOPFUusYM5ZT0K7m88yPkuXeE+JJNabCmfH79/+fsf5Pnt9/DdwXvMzf/XZ7achtlhp0Lus4mzt4PXTpxSztW4LLThJOBSRe3um3hMKXBjurN7KzegilNkh05XNblJa7q+hYDE2+iUE/DJxuO9iWm1FlZvpIfixfhs3x4LR8+y8fXBXOZfWDOYbUHIDsq/LWlCIldVTno2m+cLg5SP3hde0yDWssXMpqzKQonduxEdnwCFw8awI0jT2iVIQZ44v357D5QittnYFmBwgTN+JmFGfq1z2fwzruL6v+/a/MBZBipAc2mUpDb+mvvqOMPdG/d5zjwsxkZt4bs7CQ0TQ0aGYM/8+6is4Zx1pTBB5fFOfjzc7PxNHii2zWV0QM7k9CgCGNmYngZQbumkpOSyJf7tgd/0aBIQ63uH3n/7dvvGd2pEx0TQkeByY4okh3RHHBXBi1XhOCk9K48M+o8AG5a8iq7agqpy2Fr/HyQSJaW+NuyuXIvv1v9Ipa0kFJiYnF+9slc3/2ssH1pTJqzM2dmHhTz8ZhNT0LW3YK5tdtYXDK7Xi3OxF+i6Iv9L9Mn7gTO6PgkX+X9iULPZlShYUqdgUkX0i320DPotlSt57VdT2JJC5Bowsa1Xe+kW0xvYmwpVBq11Bo+fJb/lcWmmGiB6XRDhrpjfJaPL/PnMCXzzENuE8A13U7ijqUfYUm/s0lTTRRFYpoCIUzibU501aRWN7EssAnFH95oCtAbDTJUwEfQsKlrYhKPT259SGBDvl665WB1jwCGBprRyOgGsnSEEWpkLUuyfXth/f879+pA3p7iEIOs+0zSshIOqZ1HA0FkZHzMGDa8CykpMUGTBooicDpsTJgQHJY0bmh3bjz3ROyaiiYFiinJSonj7ismBu+zcxZp8TFojdTeNEXhgpEDGJTZYGKnibNpWRazt4SP/xRC8OAJZ+JSbdRprdmEQoxm53eDDpaoGZbUGbvS/LNTU1RMy+QPa16hSndTbXipMnRqDZP39y5gWUnLEQXhcKpRZEV1o3FFNVVoDAyEu60r/ykkYQL86cSbq1bgUGOZ1ulpLuj8Gqdn/YPLu33MCSnXIoTAlEbY0MXmqNIreHnnv3CbNXgtN17LQ41ZxfM7HqHW8GcG7qstotp04JMqPqlSY9pxm+HTx+uoNmra3Jbgdnl4eO33mJaClAqWVPAZGl5dxbRUTCwWTr2NR0ZM4+/DT+P1Uy7mxXEXkK0lI72hkpoAwhQIHYTh/3SOSSTRFfzq79UNdheVUe3xhmzfkHBiWVIFafcLyrucNlxOGwKB5jHDDpoVRdCt68HkkEtunhCiluhw2jjt/BOIjW9Z+OiY0o6jKX5RxlhVFZ769xWMPaknmqagKIKhwzrz7HNX4YoKnUXXhIJq4Q+bk5BfWMnND86g1nPQdaAogjduvIAR3TpiUxVsqkLX1CReueF80uJi+OOkcUTZbKjNxBSZUuI1mnbMjevQjZfHXkj/uCwytASmdhzA7NNvICfmYIrp+TkjidbsqAGBy8b2wq6onNFhCOsqduOzdEwpMKWCxP/xWBYPbfzgkA3NBR1vx6XGYg/EBdsVJyn2TCZlXAIQYqjraFwUM96eRYeoQTjUWLZWbeCBjb/jjlVX8X9rruXT3HcwW1COq2NV+aImEjkka8qXUOApYFX5eg7GHfo/Pqmiy6Yv+w7ODoeVlPHezpWUeGsINqoCKRWkFAxMzCJKs3NGxz5c2n0YJ3foyrjMbvxp5AScaqOHrQQ8ItBygZD+z/ydu6jxHbxGX5m/jLEPPM8Fz/yXkx58gb989E1YQSyAsYO6oDYaWCiqwohhXZjx1LXcff1kHrxjKm8+cjknDOqC0OoqAkgwLYRuokqYNn1I/fZd+2TywGvX0a1vJkJATLyLC24Yz01/mX7I5/FoIaRs1ec4te34HHj48OFy+fKjJ3sspd8f1pR+cU2tlzNvfh6vHnzzO+waN14whkvPDK2YUO3x4jNMkmKCn/Y7S0p54celrMjbz96aipAgeaemMePSi+mbHr422rd7t3PzdzNRhAgkfwhuGHAC/zfspKD18t3l/Gfrt/xQuAWP5UEg0RR/6l/32HT+Pfwa1lfs4i/r3qDaMGg8yhIITk0fQr+EHCZlDCLO1rZRi8/ysqHiJ0p9hWS6utIrdqg/FA7Y797Ji9v/GCJq7w9t64xNdTIu9RTGpY5HEQq5tbt5fMvfgta3CTtDE0dxWeeW9Sa+PPAhc/I/ClmuoDIl80IsXHyw7wP0kNG6DLQLTFmXNunHrti5vcet9I8PH/vdGi5f8CZLivaE/c6lqrw3/hr6JIT3736zZzsPL53P3spyMqJiySuqRDR2WwCqEPSMSeaSIQOJVmw88Nl39QJW4L/ezhnej3unhxb/LCyr4soH3qHa7cPt1XE5NFwOO2/86VI6pIS65EzT4uVX5vHR24uxTL+UgM2mYrNr/PPpy+jZOzjkL5wg/pFACLFCStm2MiaNiE7pKPtMv6NV66549a7DPl5b+UX5jBsihGg2AH7jznw0TcHb6F71+gwWrtgR1hjHOMPrG3RNTuKRaX4t5Kd//IkXly7HZxggBA5V5eJBA5s0xNU+L7d8/3mQNgXAy+uXcUp2V4alHxRCynAl8PdBfh+ylJL1FfvYU11El5g0+sZnI4RgQEKXsOnV4Pcrz8lfybzC9Ty3bTaPD72WAQmdw64bDrviYEhieB9vpqsrJ6Wdw4LCj5FIBAqG1KkynNRY/kSGD3LfY3PlJq7ofBVf5X8W4tbQpY8VZYs4J/tSorXm5Ux7xvbnu8Iv8FnBr+WqUOkR249dNblNVEYR2ISKTVFxqVEk2NIo8haT6erAOVnT6Rnbo9XnIxwdXPEoiPoio3VoQuGZURc1aYgBJud0Z3LOwUSKC959j5X794esZ1qSrQXFPDp3IaqHkAGFxzD4ePkGfn/WydgbhVamJcbyyUPX8M2yrWzdV0i37BROG9HbX6Q3DKrqL3BqUxV8AV+zrpvousmD933M6+/dHGR827sCXXv2Gf9ijXFLxMU4w/rPBJAYd+h+rtvGnMiEbt34fNMmTCk5s1cvhmaFFzJy6zqfbNsY1lfkMQw+3r4hyBgHtVMIBiR0YkBCcBq3S3VwZocRfJIXWgOxbsBeJ+f4pzVv8+nJfwwbypVbW8gX+xdR7C1nRHIfxqcNbdFnPTH9IgYlnMTmyuXsq93HjyWr8ASqXUsJNYbBDyWrWVR6F3FqLSLMnaEJjRJfMdFaLD5LRxNq2PZ1je5Fr9gBbKlaV2+Q7YqD/nHD6BTVjURbGu/seSdkO5uwcVXnK8l0ZZET1TqdiHJfLe/uWsKiop1kRSVwedfR9E0I/5te0X0Ec/I2Bj1cVQTdYlMYm961xWM15O6TT+LKDz/C09DFJQMRDlLg0Q20JuQypJRUe3wkxYT+Zk6Hjalj+wGtewOY+9W6sFmsJUVVFORXkNFExY/2SHtOh/6fNcY9c9JITYohN788yK3gsGtceNqQZrZsmf4Z6fTPaDq217AsHlo4n/fWrcNC4jMlwi6Q9oPtkIH1DoVf9ziLLw+sDNLQDUet6WVHdT49YoMNy0/F63lo45sYlomJxZKSjXy4bx5PDb0dl9q8+lmKI5OxqdN4ffer9YYYQLeUQKIHmNLEJ8FOaFSIIQ0KPRU8tvmvHPAUYlM0JqWP5bKcc7EFHgZSShaXrGFntUqFL5sYzaBzdCwnpU5gUIK/xFasLZYbu93ICztfqC81ZUqTK3Ku4MSUMbSWYk81F8x/jkrdg88yWFO6j7kHNvHg4HM4NUwqe//EDjw0bCr3rZyNhcSwLHrFp/Hs6AvbPGocnpXFo6efxl+//Y4qrxfdtBCmP3yyDkuAEma0FxflICEqNL63rLKWVVtyiXY5GNanY6tqPjbl6pOBzLyfFZGRcftDCMGTd5/HHY98TGFpFYoiMEyLWy4+icG9s1vewWHwyA8LeH/9uqDRk/ApICykzX+1RGk2pnZrXvClKaI0J48NuY571ryOKS08pg+z/oET6kduiGGZ/HPzO3gbGHKP5SPPXczM3B+4KCc42qQpku3JaELDkIZ/7odg/6zbtGHXgieZ7IqdfnEn8MTW1/BZfl+yz9KZW/ADVXoNt/X0i9u/tWcms/bPxxNYp8ynUWM6ub7b0KBR9NDEoTw5+EnWlq/FwmJA/ABitLbVZnt849eU+WrrhZwsJB5T5+9rP2dCh95oSmiI3JSO/ZmU2Zv3tq1i8YF9JNlj2F9VRbqreddLOM7q1YvTe/Tg/dXr+OfXC6n1BT9gFYdA8wksS9YPKpw2jXvOGh9iRN/8YikvfbIoYIAFDpvK03efT89OqU0e3+PRw7r7hICsjomkpjVfQbpdISNuinZLZmo87/3zKrbtKaKyxkOfrhlEu1pX/+tQ0U2T/65bG/zqSWC2XFewbCZOzcbUrr0Z06FpJbmWGJTYhc9P/gtrynfxQ+EGPstdhreRjzZWc9EtJtiHubNmf4iCHPiN4ryi1a02xmNTTmZO/myMJqIjLBQqDSexmoFdEUSp0UxIP4PVZQfQG43ofZbO4pKVXOk7DwnMzPs+qPqJTxqU+Sr4tmARUzLHB23rUl2MTA4q3NsqqnUvty5+jxVlO8N+r1sme2tK6RobasiklNw5fxbf5+6k1tBREHy8fT23DhrNrYNHt7ktqqJw3sB+PPntT0DwudFsKs9dOo0PlqxjXW4+HZPiuXHCSEZ1C752Vm/J45VPF+PT/ZKZALVuyU1/eYeseH8x0mmnDmTqpIFBo91nn5jD/jCJGw6njXvvP6/NfTnuRIxx+0UIQc/O4SfXjgbVPl9Y+UMAh9C4auBwJuf0YEjq4YVYgT/ueFhSdwYldGG/u4zlpdsxpIlN0VCEwkODrgg5hkOxB5IoQolqwUXRkER7Irf3uJOXdj5PtVGNJ8wuTRQyXQP4U9//q18268DfkWHuGE3RKPKWUuKrxKZo6I0mPL2WzorSjSHG+FD586rPWVW6DwsI95ZuSJM4W3gRqB/376k3xOAfTbtNg6fX/MR5PfrTIbrtI2SHpvHmFefz6/c+o7zWgxACTRH869wzObF7Did2z2l2+4+/W4O3od9XSlS3xGfp7Kr0K7Q98/o8lq7ezUN3nw34J+q++XJdvUBQQ2JjnXTKSQlZ3p5p70kf//PG+Ejx3fod/PurH8krq6RbWhJ3nHkSI7qHFoyMdzqJdzrDSmsO7ZDJ3SeMC1l+uGiKysODr2RDxV5Wl+0i0R7N+PSBYYtldopKI82ZSG5tUZBRdCp2pmW13tcK0DO2F48OfJxCbwFLSlfwSd7sevcD+N0SF3YMFpLpFp1DXm1BiPqcYRlkuFIxIay+roJCiiOhVe2ypIXb1IlS7WEfeG5D59sDm9EtEyGEPzqkwWqaUBic1IkUZ3ijOmfP1npD3BBVKMzP23XIxWp7p6fy3W3XsrmgCJ9h0i8zPRDa2Dzzlm9j/ortQY84YYZOZnm8BktW72brzgJ6dk1n25YDYQ0xQHW1J+zy9o5ookJ8eyBijJugqsbDjtxiUhNjyEpLaHbdL1Zu4q8fzq0XE1q3r4CbX/2UZ66ezqgenah0e1iXV0BStIveGance/J47pn7db2rQuCPDb177EnNHOXwEELQPyGH/gnNj6CEENzf/zr+b/Wz1JoeJP6Jr9M7jOTk1MGHdNx0ZwbTMs8izZHGx3mfU+otIye6Ixd3PI8ejdTkpmVOZl7REqQ8OLknJSgijmg1ip4xOSTZ48n3FAcZbE1ROSsz9EFmSouFBdvYVHGA7KgEDtRW8ur2n6g1fcTbXPy270TOyxkatI3b9B0smyr9UgX19QME9InvwL+GXRC0TbXPR151BR1i4oi2OVCFaOCnP3guorTmMwBbQghBnwz/m1x5jZuHZnzHhn35ZCbGIiyF3KIKumcmc+OZo+jXOYP5K7bzl+e/DB4VE4gHDrN/aUnWbsqjZ9d05sxa3WQ7uvc49Np9x43jmF3XGiLGuBFSSl7+aBFvfbEUm6ZiGBb9umfwyB3TiQ1TJ09KyeOzFgapugF4dIPHZy1k8tAePDtvMTZVxbQkWQlxvHjFOTw/ZRpPLV5EbmUF/dPSuXP0GPqlHTt3SXNkRaXy9ui/sKZ8O+W+KvrHdyXFkYDXMnAo2iG7T0Yln8Co5OarbeyuLaVaj8Wm1GBXTCwpqDVsVCFZWbaDYUnd+Xv/W3lw04vkuQtQhYKCws3dL6FzdHAYYJXu4fIfXuFAbTlu04cmVHRp4h8cCUp9NTy07kt/RlzWQR3gRHsUqc5Y8mrL8WfP+Q2rgmBMejdeOPGy+nWllDyydAGvb1iJpijolsnknO6B6i3BE5RSSiZ2bFnKtDX8sHEXt7zwaf3/c0sC2iYG5JdWsmzrPp769XSenbHQb4gDTxWlTsK4bsTfaL+qqrBp9T7Of2Uh5SXVgSeiDHm9P2Vyy7rJ7ZFIaNvPiLmLt/Df2cuDJjrWbdvPfc/O5vHfnxuyvkc3KKkOL6SzLb+E7fNL8Rom3kB66s7iUm7+72d8cvNlnJzT+aj143BRhcLQxJ6Y0uLfm+bx9o4luE2dDFccfxxwOhMzex+V426u3EeNYQHBYVkqJluq9jEsqTupziSeHHIPB9xF1JoecqIyw0Y1/Hvzt+ytLkEPGEVdmv4RNwcHSB5T55lN3wcZYyEE9w+Zyi2L38NnGZhSYlc0XKqdvwyaEnSMV9Yv542NK/2RMQFDN3fvDk7skMNPBXsDIvH+47008RyibYc/Qby/pJJbXvw0/JcqWIZfm/vRGfPIL6rw98lspMAmAZtA6gcNshAgKnUWLdyKt9aHsOrWF4GT5j9rNpvKmJNbLuHVLomMjH8+/HfWcjyN0vJ0w2LZhr0UllWxYPUOvlyyGbtN5bxxgzhlSHeiHHaq3KECLRKJp5GCnCUlu0vK2FlUStfU9l9Z91/rv+G9XSvwmP5zsr+2gv9b/hEvnPgrRqR0PuLH6+BMwqnY68PW6rCrGunO4HLwHVxNh2QBzMlbX2+I66ir9uy/K/2m5oC7ImTb0WldmTH+el7fvohdVcUMT8nhsm4jSXUGh8a9uHYZbiM0A25Zfh6LL7qJHw/swa5qnJzVGddhuijAL/Rz1dPvN2y+nzCRizsOlNA1OZ7c/LIQKUyBPxLDYVewDImUkpSEaMp3luH1GQ0McTBCEUw7bzhJyW0LEWwvRCbwfkaUV4Uvsqiogjv//Rm780vrZTfX7zjAklF9uGzMEJ6buzh4Awl6E8KomqJQ4W7/EyBuQ+e9XctDUrU9psEzm+bx5klXHfFjTkgfxHPbZ+G1GtoXgUu1c1Lq0Xk17hITPiqge1wqDwyd1uy2ZZ7w10ut7iPO4WRq10OLFW+K1bvyqKj1NKtBXEesy8HNF4zhvv98iRmm9JWumwwZmM1vrzwFTVUoKajkz/d+hBEm2w4AIXC57Nx466TD7MVxorkyK+2An1n6zNFn1MAcVDX0Slc0wZ6CsiD9Y7fPYPaiTUQpKjahhJXhC6fmZklJnw7twz/cHMXe6ibV2PZUlx6RY1TpHj7YtYqXtvzEutL9RGlOnh12C91jM9GEiiZUesdl8+ywW1pMx25Mz5gOIffewf/7++VUNe7qF1xHsS30Twmfadk1PqlVkQ5tpaiyJny18foZx8CEsF3j8olDmTiiF9dMHxU2cUNRBOnJsXTOTia7QyKZWUnoehOGOEDHnOR2rz/RHMJq3ed4EBkZN+Kas0fz/bLt1Lq96IaFEGC3afTtlcmSMNWjhYDdB0pxKP7JvvrlADo4Y2z4LL/PWAh/vOgfzxiP09b+T326MzasMRZA7/jWlXJqjhXF+7juh3ewpES3TGyKyoTMnjw24hxeGXEH5b4ahIB4W3Sb961bJssKCrEUgdIwX1iCgyiEZtA5JoW7+k5idNqhT6r9ZfQELp31Ph7DqPccODWNv41pXXJMWxmQ0+FgIdGGBLroEAqKTXDumAFcc/oIAK4+eyRfz99IbkF5kB6LTVM57/SDqf+pqbGMGtWdxUt2oNf46v3rdTgcGldcc/LR6NYxIRJn/DMjLTmWdx65kndnr2D5hr1kpsbzqynD+XH9LlZuzQu5ERRFMKx7R2av3xpitpyaxlMXTmVTYSHztu4iLTaay0YNYXDH1lcaPp7YVY0beo3lhS0LcZsNqnKrGrf1OfQKHeAPObtl0YygslOGafHd/q3M3reBszr2o9qn+0fDh+Bq3VNVhiXB67WhKBaKIv1REaZCanQs3085vKKjdQxJy+Tjab/i6ZWL2FBSQI/EFG4bOppBqUfnN85MimPaiL7MWr4Jd4O3NIdN5bVbLsDltJOeGEu08+BEoRCCJ+89n7v/+Sl795ehqv5Jxd9fNzkk4ekPf5rGC89/x+xZq9HdBkKCUCAhIZobb53EiNHd+dkiZbt2U7TKGAshTgeewl8E5mUp5cONvr8TuA4wgCLgGilleFHXnwE11V6K8yqoyq+mzBBUlrmZOrY/b361nEbzcWiqyukn9MJUJQ99/D2WZWFaErumcv7oAYzu2YnRPTtxzdhjJ43q1nX+s2wJH23aiJSSab1685uRo4mxt30m/4aeY0m0R/HCloWU+mroHZ/B7/ufSr/E8KplrWVt6X68ZugrsdvUeXnLYh5ZsYBiTw2WlPROTOPZk84mOyah1ftPcLgwLH85e8tSaZj0mOo6spNPfZLTeG5yy0LqliVZtHMvWwuK6ZgYz7heXbCpoVEgLXHvBRMZ3DmTdxauotarM2lQD648ZRhxUeEzAgEyUuJ445Er2HegjOpaL91zUtF9BiUl1SQlRde7Hux2jd/cdiq33DoZ07RQFIHHoxMVFT5B5udGex4ZtyguL4RQga3AZCAXWAZcIqXc2GCdU4AlUspaIcSvgfFSyoua2+/RFpc/VPbmlXL9/72Fx6vXv9I5HBq3XTuBpIxY/vzyl/WiLHHRTp74zdn07Oif1c8tqeDr1VvxGgbj+3WjT/ax9wtbUnL+jHfZVFSENzCKt6sqXRMTmXnJZUfFj3koLCvawzXz38Nt6AhVBiV4WD5bUJadIgQZrlgWnP3r8P7SJrh2/vv8VLAbn3XwCepSbTwxejqTs1suaX8kqfH6uPyVGewpKcdnmjhUlXiXk3duuJj0uKMbmXCgoIItO/JJT42jd/cM3G4fjz7+JT8t3o4QEBfn4q7bT2fUiCMTA300OBLi8rEJ2XLIybe3at2Fn/++XYrLjwC2Syl3Aggh3gOmA/XGWEr5fYP1FwOX8TPllXd+CDLEAF6vwXNvzOfzN27hm8dvYuOeAuyaSq9OaUGjhezkeK6Z2HxSw9Hmp3172VpSUm+IAXymyb6KCubt3sWkrsf/hltXnM+v535OtUcgAz4I1WGg2i0Uy18HrqE4uyUlFT4PP+TvZlxm6zWBnxg9nVt//ITlRXuxqRqmZXFb/7FNGuKimhrm796NpihM6NqVOEfrtThabMs3P7CjsBRf4HcxTAuPbnDvJ1/z0pWh8etHAsuSPPzMHL5duAlNU5FSkpmeQILNzsbN++sL9xYXV/PXBz7lmScuo3u3w58LaM+055Fxa4xxFrCvwf9zgeZksK4Fvgz3hRDiBuAGgE6dDl2R7GiybnNeWNF5w7AoLK4iMyOBgd0O7xX9aLK+sCBsvb0aXWdtQf5xN8Ze0+Cyr2ZQ4asL7fM/zEyvhk0zSLBFUaCHJtFYUpJfWxW0LL+6iu/37kJTFCZ37kaCMzhRJM7u5M1TLmF/TSXFnmq6x6cQpYV31byxahUPL1yAqvjLwv5xruTfZ01hYte2CcI3xRdrt9Qb4jpM6Xdb+AwjpCLHkeDTOav5/sfNQQlMe/YWs9dthdRC9PlM3v9gKX+6Z+oRb0e7wZ/bf7xb0SRH9AoQQlwGDAfCqt1IKV8EXgS/m+JIHvtIkZocS1FJdchyy7KIjwsV625vZMXG4dQ0avTgxJUom42s2OOvPTs/d1dYiU4Ab42NQssbUJ0P/s6QFoNTDj4EX12zgkcWL0ARAoHgz/Pn8sSkMzmjW09My2LZ/jzK3G6GZ2bhVGws2L6Xpw4soVdKCpcNGkxm7EGRn20lJTzyw0L/20QDg/mbWV+w+PobiHM27YttLeHEjeo4WnNKH81aiccb/GA2DAstzAGllOTmhUpl/tL4uY+M84CG8mPZgWVBCCEmAX8Cxkkpm68X3o654vxR/PWxz4MuYoddY/yJPYmOOnKvrW1lT3E5C7fswq6pTOrXPaQoah2nduvO/Qvm4TaMegMg8PuNp/Q8/imsFT5PE4ZJIE2QJvXREw39yLpXkmT393lbaTGPLl4Y5IoBuGPubDpEx3HLrJlUer0IIfAaBqpQkFLiNU1+2LuXN9es5p3zL2Bgul/s5rPNm8KGiwkLrnr3Y3YVleKy2bhk2EBuGjPikCbdTu7ema9WbiEwp4hl88euD+mYieMohTm6Pb7QhU243DVNZeDAo1tUoV3QjqMpWjMbsgzoIYToIoSwAxcDMxuuIIQYArwATJNSFh75Zh47xozozs1XjSc6yo7LacNuUxk3uie/u/m049amZ75exDlPvMljsxfyyOfzmfTwK3yzblvYdR2axgcXXMzA9HRsioJNUeiXlsaMCy4i+hCiKY40ozt0Cj8ylvjl0RBQpYFXOWicPQqyVuXxJT8C8NnWTehWqPFUFMF1X3xCfnU1NbpOtc+Hblp4DKPecOuWSa2u88e5c+u38zZ4cDVsj15tsiGvgBqfTnFNLS/9tJw7P5nd5j7nl1WxeN0e0EGx/BoRmgeihMaD557a5v21ljEndA8tiyQEzlgHTsfBB4CiCFwuGxece3znO44FQrbuczxo8ZEspTSEELcCX+EPbXtVSrlBCPF3YLmUcibwTyAG+CAwobVXStl8Hmk75pwzhjBl0kAKiitJiIsiJvr4jYjX7cvntQXL64WG6rjn/TmM6t6JWFdo23ISEvj4oksp97iREhJd7ce9kh0Tz1V9hvHm5lW46zR/QzIXBXg1aPR+9c2uHfzjFNAtK+zo2rQsqkxvq7RgthQX4TF0nJqNU7v34J21a4M0JoTub0vDfXkMg3nbd7G7tIzOSYkh+2yK/3y1iGrPwc7UeWBclkZWQvOuIykln36/jtc+W0JJRQ1ds5O5/dJxDO/b8pzL1RefyA9Lt1NV7cHrM9BUBU1TeeBP55CfV86Mj5ZSVeVh6JDOXHfVyaQkt130/mfFL0FCU0o5G5jdaNlfGvz9M01WbxqbTSW7Q+tvuKPFF6s24TNCR4GqIliwZRdnDT6onialZGdhKZVuL32y0kImtFrLntJyymvd9EpPbTFTcP2BAp5c+BObC4rokpzIb8aOZkSn5l93/3DCeE7MzOGZ1YtYUZDnf3WvGxU3c7c4A5Ncp3XtwZvrV4UI9FhS4kBrUhOkIYoQqIF6ecMzM5nWuzefb9mCW9cRQqDK8C2xKSpbC0vaZIwXbd2LGWZS2O3TOVBWRXZyfJPbvj1rOS9/sqg+DX/rniLufOxT/n33eQzqmcXGrQd4f+ZyCoorGTagE6OGdSUnK4m4WBcJcVHccc1EZn69muLyGgb0zebi6cPJ6pAIA3M464xBre7DLwEBiCM0gXc0ci8iGXjtHMOSITPf4Hd9NSzflFdawS2vfUZeaQWqomBJyR/PPoWzh7euHDvA7pIybnrnM/aXV2LTFCRw96knc9Gw8JUpVubu56p3P6pPBS6ormF13ic8dfZZTOjRfBTCuKwujMvqwsOLF/Da+pUomt9AgiDFHsWeimAlNZemcVn/wQAMzcjk/N79+XDzBjyGjiIENlXl9mGjeXbJ0hb7aVdVzujeo973K4TgoUmTOadPX2Zv3YpdU6ms8PD5+i0hvmRTWnRMbNp4hiM+ykl+eVXIckvKoDcbw7QwTBOn3e80NwyTVz9bEqSHAuD1GTz/wY+ce9IAHnn2K3y6v+jrhi0HePPDJdgRjBnRndKCKnbuKsTt1nE6bRTvrWDapEOrMvJLQRwBn3Eg9+JZGuReCCFmNsy9AFYBwxvkXjwKNJt7ETHG7ZwzB/Xks+UbcDcScDEti5N6dQH8I+LrX/qY3NKKoNf3Bz75jh4ZKfTLbjl29P2la/nbF9/6M0bxhzqhwsNfzad7ajLDOmWFbPPQt/PDykfe/828Zo1xWa0bm6IQ43Rwz6iTubTvQBbm7iHGbmdSTjfcusGln80gt6oSAZiWZHynLlw3eFj9Pu4/eRLn9OzLlzu2YlNVpvfoTa/kVDKj4/jDt9/gM00sKXFqml/03TQDAv8W/dLS+PuEYO0IIQQjsrMZke0f1e+vqGTOpm1BxtiuqvROT6VPevPSnY256pRh/G3G3KACBDZV4cReOcRHOfH4dB7/7zxm/7gRw7TIyUjknqsmkZ2W0GS9xB25xTz+4tyQCh5IiU9KfvhxK8KQ9WGaHo/fJfTX+z/l7ddv+EVk07WZI+emOCq5FxFj3M4Z2jmL6cP68emKDXgNA00oKIrg3rMnkBjtd0Os31dAcVVNiB/VZ5i88+NqHryo+cnH1fsO8I8v59VPNNc7C0y/SPkbi1eGNcabC4vC7m9/ZSVew8DRKHZ2w/4C7vn0K3aX+EOoTsjJ5uFzTqNTXAK/6ptQv160zc5XF1/Jyvz95FZV0i81je6JyaHnJiOToRnBMd/Te/ehV0oK/127lsKaaiZ27cb0Xr3ZX1XFlpJicuIT6JPasjHNjI/jrcvO595Zc9laVIwiFE7t3Z2/nTGRzbmFvPX9SvJKKjihR0cuHDuQlLjoJg3cWUN7s7OglDfnrcCmqeimxcCcDB669HQA7v3PbBav310fC7xrfym3P/Yxr9x7ceBtIZT0+BgKysOEogUqc1i6FXYiqqS0mv0HysnKPP4uuGNPm7QpUoQQDVOEXwyE5sIRzL1oSMQYt3OEEPz5nAmcc0JfvtuwA6dN44xBveiYnFC/TmlNLUqYEsaWlBRV1bR4jHeWrA7xS9cZZAkUVIXGXQMkRUVxoDL09VsAtkZpy8XVNVzx+ofU+A6GWy3ZtY/LX/uAL2+9KqT9QgiGdchiWIfQh0BL9E5J5f5GI98uiYl0SWybARqQmcFn119GrU9HUxXsqsrHP63jHx9+j276EydW7dzPi3OW4FI1Lh43mFumnRgS+iaE4LYzx3Dl+GFszy8mLT6m/vfLL6kMMsR16LrJe9+s4pLTh/HOnOUhoZZXTB/Bw080cX+3YG/+J0fFAdoQKVF8JNKhW8q9aEj7ECr4H8Pt8bE3r5Rad5g40Cbon53BbaeN4YYJI4MMMcDATh3CTvI5bRrjendpcd8l1bVN3r82RWF8Ey6HCV27hN74ElRTsHzP/qDFH6/aEBDuOYgpJcXVNSzZvY/2TJTdhmlY/OY/n/C39+biM8wQP77bMHh/3moefOfbJvcTH+VkWNfsoN8vr7DCn30nJYou0dwSrVZCrcXydXspyiunR0YKUU4bAshKi+f+m89k8uje9OnRAa1x6Jr016uTTdzZqamxdMhom8/7F0WdcltLn+Zpa+7FtNbkXkRGxscQKSUvvvMDM75YgRDg001iY5yMHNKZy84ZSdeO4StOtERitIsbJozgle+X1fuWHZpKRnws54xouTrGxD7dWLl3f0hRVQEkxUTxqxHhZ93jVDvCANngKhKGvxz6urx8RnQ5GFWxq6QsJDwP/KP3/eWVrejlobG7sIz8sko6pyZS5fGRFOMiOa7t+sj/eP87Fm/eG1ruiIP/9+gGs5dtpqyomtVb8nDaNc49ZRDXThuJpoVPFOmcmYRPN1B0UIwGu7agOK+SOXsq/BOUNoVfnz+aKy8YXb/t/b+fxt0PfsyO3cX4dKM+VltIcETZ6JAYR2FhJT6fgcOuoWkqf7337P/dkbE8YtEU9bkX+I3wxcClDVdokHtxemtzLyLG+Bgy44sVfDBrRdCkS3mlm68XbGL+ku089qfzGNz30LKgbpw4Epsp+HD+WnTTZNKwXvzmrDFE2VsWAz5nSD/eW7qWfWUV9QZZCBjVrRNPXHhWk+nAWYnxxAiNWo9R79eoqzKRER8cszq0YyZzNmzD3ShNWwJ9j0LVkyq3l9tf+oz1ewtASryGiRbQnRjZqxOPXHkmMWFitMPh0w2+WrEF3bRafJfUDZOf1u9BWBK3V+ftL5ezc38xj9waPuw+OT6aicN7MPe7TUHLBYCUSAWkJfH5TN78YDFnTRxASpJf5S0xPooXH72MPbklbN6ez+IlO9i0LZ/0tDguu2AUwwfnsGbdPjZsyCM5OYZxJ/XC5Tr+iT/HlSNgi49W7kXEGB9D3vl0WYhWQF3RXY9X57GX5vLWE1e1eb9SSu57eQ7zVm7H7fWHen3xzTo6xsTyq9Nadnu57Dbeu/ESPlqxnq83biMxysWvRg5mZNeOzW535oBePPbND4i6URl+I+6waUzqHSxINGVAb55bsBS9ysQIRAg4NY0TcrKPSgmq+975mrW7D/gNaAAj8PfSrfu4580veebGs1u1L6/uz9Crn9hsPDqWDUr1SKCh4p9u8OOaXeQWlpOdlhB2/+eMG8CChVtD/MYNqqcCoCoKS1fv5swJwW87OdnJ5GQnc9r40DDGwQM7MXhg+xTlOh4cidA2ODq5FxGf8TGkojp88UoAJOzKLamXNWwLyzfvqzfE4H/19/oM/vPRjxSXh598a0yU3cblo4fw1rUX8vQlU1s0xACxTgdvXXMB3VOTsGsqdlWlT0Yab197UYjegstu48MbLuH8of1JjnbRIT6WG046gWcvPvKJmjUeHws27AwyxEC9AfUZJku27KWksuXJTYAYl4OMRP9IP8joNvooQiAMGeLFsGkq2/cVN7n/rPSE8F80MhxCAZfj8CtM/09zZHzGR4WIMT6GdM9pJqRKgN2mhmoJNEF5ZS0r1u1lb14p36/YVm+IG6KqCj+t232IrW0dnRITGJeaQ3yxSkyhoJeWSEwTJemToqP465SJ/Pi7m/j+juu4edwo7E34Ug+HWq+vRb+oTVUobe7h2AAhBH+6ZCJOu4YqBIoFNiFwqiq9MlJwKiqpcdEM7ZyJwwo9rmFadMxIaHL/qcmxjB7aBbst9FxoNSaKx/S7LCSMGtbyhGyEJqjTP2nN5zgQcVMcQ267ZgJ3/O2Dep9xw+evw64xbdLAsCFqDZFS8vzbC/lg1gpsNn8R1JhYJ4qExnagzmXQVgzT4r2vVvLRt6txew1OGtKVm84bQ3JC8MSXlJKbH/+IzXsL66M55i7fysotuXzy4NXHbRSXEhdNUoyL/MZvBTL4z5zUhFbvc1TvHN78v0t4Y+5ydheUMqhLJpdPGlY/YgY4UFzJxX96I6heoE1T6dclg25ZwZOz1TVeiourSE+Pw+W0c99vzuTJ17/ny3kb0HULLInmtlAChkHV4eH7z8Hl/B/3+R4GAnnE3BRHg4gxPoYM7J3Ffx68hFff/4mV6/fi8erYHRqWJRk7vBs3X95y5d2vF2zioy9XBQmGG+W12BSJNyrYGluWZOygtouj//X52SxctbM+DXfWwg38tGYX7z98FTENZETX7NjPtrzioLA605JUe3x8uWQT557cttRbn26weO1uKqs9DOmTTVYTPtaWEELwl4snc+crn+MzzJBkGKdd47fTTsLexgdVj6wUHrjy9Ca/75ASx3P3XMA/Xp/Ltn1FqIrC5JG9+N1lE+rXMQyTp56dy9ffrEPVVCzL4sLzR3D15WO5+8ZTqS6o5ocft4VMNNmEQmbq/3BI2pGiiYzG9kDEGB9jenVN55E/nANAWUUt+w6UkZkeT0pi6+qgvTdzGZ5GLgnTtFCkwK4qqIHXfp9uEOMVnH31f/D5TOJinVw0bTiXnjOi2dH3voIy5q/cETSZZFqS6lovny9YzyWnH0xJ3pFXElY3w+3V2bSnbUqq2/YUces/PkA3/IpslmVxzoSB/Pay8U26HAzDDFQ6Dv1+TJ/OvH3nJbw9byXb9hf7Fd3cXjKT4rh60gmM6dO5Te1rLX27ZPDW3y7D49PRVDUkDvilV+fzzbfr/ec3cI4/+HApyYnRTJ86lIL8yrAz/jabSlFJFRn/yzHCh0u9TGv7JGKMjyOJ8VEkxocXiW+KympP2OWWJfnzpZOoMHy8/PZCRKVJtTTqFdrLK9288cEiyitqufWaU5rc/5bdhWiqGjKz7/EZrNqSG2SMO6YlhE3Xddo1unYITV9uCsuS3PmvT6ho1LeZ89YzrG8nTh4WHJmxbkse/3xpLjv3FmO3aUydMICbLz8Zhz34cu6RmcLfLj16esHN4QwTUmiaFjNnrcbbKKLG4zV4d8YSpk8dyuBBndi1uxjDCM3I65JzaHHoEQ7Snt0UkQm8nxmjhjbhdpCSzz5fRfn+Kqzq0Fdz8N/0H89ZTXlFaI25OjJS4sKOdjVVIadDUtCy4b06kpEUGzT6EwLsNo0pJ/ZpsS9SSr6av5GLbnmZsv2VKB4rKCzM7dX55Ls1QdvsySvht/d/yI49xUjpVzD7/Nt1/P3ptom+79hVyE9LtlNcEprO3VZM0+L77zfy1/s+5tFHv2DD+tyw6+m6ia+xsE+Aigr/ZOJF548gymULentxOm1cdMEIYmIOv/zT/zztOJoiMjL+mXH1BaP57Ks1+INbD8ahKj7Jpi0H0LHQDTMg3ho6atV9Jhdf9Ty3XDeBqWcNDvm+X9cMstLi2b2/tD4uF/wTUedOCM7EUxTBi7+7kH+8PZf5a3ZiScmgbpnce8UknDYbBwoqSIh3NTnp9OaHi3nr4yX1sddCguqRmE4gYIwax2W/M3N5SPifVzf4aeVOCkuqSGtBIL2i0s3v//wBu/eWoKoCXTc549QB/PbXk1ucPA2HaVrcc/f7bNyYh8ejIwTM+34zV1wxhosvGR20rsOhkZ4Wx4H8ipD99OrpLwGVnBzDi89dzetv/cCKFbuJj3dx0QUjmXhKyw+3CC1x/Axta4gY458ZKUkxxCoa1T6fX39AgmL49QjsDo301Di27SwMJCfIMAZZ4qnVefaFb0lPj2PE8OCRthCCZ++5gL++8CXLNuxFCEFGSix/uf50OqSEVqVIjHXx6K+nYgSEc2yaykefreDmNxdiWX4JxzMm9+c3N00MSgl2e3xBhhgOihMpusRyCJwOjdNO7B10vB17i8MKtdtsKnn55S0a4wf++QXbdxYGPWi+mrueHl3TmHrG4Ga3Dcesz1exfvUev1tHEUgEXq/OG28s5NTTBpKUdDACRQjBbbdM5q8PfFrvqhBC4LBr/PqGg66j9LQ47r7rzDa3JUILSP53qkNHODZMP3MwH89cGZRWbbernDF5AJMn9GXpqt1+Ixdih2V9UoI34KdsbIwBEmJdPPl/51Lj9uHVDRJjXSGTZFJKVqzYzdIlO4iJcTL51P506JDA9ws289JrC4ImGefMXY+qKtz264NJSbkHylGUUC+ZALD8yQ09c1I56+TgrLI+3dLZtivYmILfBdApM9iN0pjKKjer1uwN2dbjNfjgsxVtNsavPP8dM/67CNOqS/SQSJsCikBVVVav2s2EicHtHzWiG489fDFv/vdH9uWW0qNHBlddNoYundumkRzh0GjPPuOIMf4Zcu3lJ5G7v4wly3di0zR0w2TIwE6cfepAfvh2EyM6ZbBqxwF8qsRrHnw1E7pEMQ5ejEXFzftLo112osNoGZimxX1/+YhVK/fg8ehomsp77y7i7j9M4c13F4VEe3i9BrPmrOWma8djD0yypSbHhExS1ZGaHMPt10/kpKHdQqIRLpl6AnPmbwwyqA67xqQxvUhObF4AyO3Wm3RF1NQ0LaplGCYLv9vEkh+2kZgcw5lnD6G0pJpPP1iGZTXKuNMtpF1BCIhqonZiv75ZPPLghc22NcJRImKMIxxJbDaVB+49h/0Hytmzr4SO2UlsWL6b2656GdO0sEwLh8PGCSO7ctPvTueaG1/F4w42kKoqGDIoWLPANC2EIGjEalmSyopaomOc2AIZYgsXbKk3xOA3VoYBjz48CxETPtFDSklNrbfeGCfERXHSyB78sHR70AjfYdf422+nMKgJwaTM9Hief+ASnnr9e9Zt2U+0y8H5Zw7m8rOb0/b2k5YaS1ysk6KS4GQQVVUYPaJb2G18XoP/u+kNdu8sxOPWUVWFLz5cRo8B2XjDZD0CYPn3OSySLde+aKQb0t6IGOOfMZkdEsjskEBlhZtnHp2Nr4H/1ePRWb10J/t3lXDDNeN48eX59SNWVRW4XHYuu+REAA7klvLkgzNZu3w3QghGj+/Nb/4whcULtvLKv7/B7fahKArTLjiBq2+ZyLdzN9Qb4oYoiqBDajxbq0NjjKOiHMTHBYfx/fHW03nsxbnMXehXLIuJdnDHdRPDGuKSoirKS6vJzkmhW04qT9/X8shSSsn2zQcwdJMefTLRbCp333EG997/CT7dxLIkdrtGTJSdq381Juw+vvpiNbt2FOIN9Nc0LUzTYuPafWEHWQJwRdl5+JGL6h9eEdoLkQm8CEeZlUt2hNW08Lh15n29nrv+Mp3MDgm88/4SiourGDI4h8svPZH0tDhqa7zcfuVLVFa4AyFtkkXzN7NpbS7VNd56IwQmMz9YBgK0ZozM1DMH8ezL8/D69Prr3uHQ+PV144NcBAdyS/nq05U4iz388YoJDBjRhdTk2BA3Qk2Vh4f++AFrlu9G0xR0j0GHzAR6D8hm2iWj6NkvfCWQHZsP8Nfb36Gq0o1Q/CI+PftmsnHNPmymRXp2AvFdUxgxshvTzxxMXGz4Strff7W+wTk4iKYqqIoICVWz2VTefu9m4tsYPx7hGBExxhGOJpqmhk7W4Q+kqDOcI0/oxsgTQl/Fv/tyLV6vHhRbbBoWpaXVIdet16Mzc8Yy/vDohSxdsiNkdKyqCmecPpC+/bJ49c0f2Lz1ABnp8Vxx6YmMbDBRuHj+Zv7x+xmYpsQwTBZ8tZ7O3dN49OWrsTfSs/jHnz5kzbJd6D4DvdbfoNzdxeTtKWbhNxu45Y9TOHX60KBtfF6de258napA7G5dR1Yt2Vm/TsneMmStzkX/uCjkmA1pSv9XkdB7QDabNuQF/OYKiqpwx91nRQxxe0UCjZX82hERY/wLYPjobsgwvjC7w8bks8JX6ahj786iEH8yBOyXJcGrI0wTFAXptGFZkt49MzhrymA+n7kK8Ls9EIL7HzwfTVPp1iWNB+87N+zxdN3gn3/6KCgLzeP2sWtbAXM+WcG0i0fVLy8rqWb1sl3+uOJA/+qeOVL6Hw7/eXgW404bgMN50KAuWbC1ycnBOkzToqrKw8JvNjBxyuAm15ty/jDWrdoT8uCJS4ji4ScvZc2qPSz+YRvRMU4mndafzOzmIzoiHE8kyIgxjnAUcbrs/PmRC/n772cghMCyLJBw/uUn0ndg87rE3Xp3wOmy42lUj0+RFrLS7Q+HAyQmwqtjc9mJS4ji5lsmMW36UFau2E10tIMTx/RoVRWJ7ZsOhH1T9Hp0vv9yXZAxriir8bsmfNRXEWmMIgTbN+2n35Cc+mWV5bWYrYgn9dT62L29oNl1Ro3tyZTzhvPZB0vRVP8biMOh8cATl6AoCkOGdWFIZKLu50PETRHhaDN8dHfemX0ni+ZvxuPROWF0dzKyWq6GPG5yP9547lt0n4EZeIXTFIFSUYNR6/FnwtlsiECEhcuy6v3T2dlJZLdxJGi3+1XqwuF0BrsLMjslH7TA9WU2gjFNiz2b9rNp0VY65KQy8tQBDBjWuVU3ndNlJ6db81VGhBDccPtkzr5oBGtX7SEu3sXQEV2brGkXoR0TiaaIcKyIiXUyuZlX7nA4XXb+/eaNvPD4HBYv2IIiJbKgFF9tYKRsAYYJTgeoKlVlNVSV1xKbcGh+0a69MohPjAoZiTtdNs664ISgZXa7xg2/PY3nH5uDt9aHlMExvYoisHSDl/7yAbpXx+60ERMXxWOf38UpZwxk/lfrDrpgGhlzRRFExTjoP6gj+XuKSO+U0qwgfVpGPJPOaJskaIR2SGRkHKE9k5wayx//cQEAT931X76ZEUb+0usDl1+oxuY49MtGCMHfnv4Vd1//Gj6fgQykTE+eOoQxE/uGrH/mucPJyEpkxhs/sHPTAarLa+v9wwrgyy+p9+e6q7143T6evOu/PPDOLQwd3Y3ZHy5H1w1Gj+vNlvW5LJq3GSklAwZ3pGZnPtePuBchBAmpcfz+hevoP7rHIfctws+AiDGO8HNh6bfrMY0wkxxSomoKw8f1wnmYFYY7d0/nv1//juU/baOyrJb+wzqT2bFpd8fQkd0YOtIfCVJZXsvmdbnEJ0Rx74VPYTSaWLNMyZoftqD7DMadNoBxpw1o1A2JaVpcO+xPFOWVYgVcM4X7Srj3/Cd5aen9pGZFJuF+kUgJZttrTB4rIhKaEYKIakamsVP3dO58+KLDPoZlWaz6bh1LP/iRfcu2YNaG12gOR1xCFCNO6kmvAdlhpT6Bg0VCwyCEYO0PW6gsra43xHWYhsmctxa2ui0Rfoa0YwnNiDGOEMT0607B0Wjkq6oKfUd05dkv7jxkX3EdlmXx1/Me4+8XPM7nz33NB499zk3D7ubrN+e3eV9jzxoSkoCiKIK+I7phdzYdO1xyoCysIdd9Bgd2FbW5HRF+RvzcjbEQ4nQhxBYhxHYhxD1hvncIId4PfL9ECNH5iLc0wjHhzCvGcsq5w7E5NKJinThcdroOyOa+125sseJya/jps+Ws+nYdnoAwj6mb+Nw+nrr5ZWoqmxa9D8c1f5pOWnYSroAgjzPKTmxSNHc8cVmz2/Ua2iVkVOzf3sHgk3uH2SLCLwPpj6Zozec40KLPWAihAs8Ck4FcYJkQYqaUcmOD1a4FyqSU3YUQFwOPAIf/PhvhmKMoCrf/61dceueZ7NyQS2pmIl37hRftORS+f//HekPcEM2msvq7DYw5+4QwW4UnLimGF+b/mUVz1rBjfS5ZXVM5aepQnFHh1dLq6NQrk1FnDGLJnLV4A1EdNrtGUkY8484d0bYORfj5IEH+zJM+RgDbpZQ7AYQQ7wHTgYbGeDrw18DfHwLPCCGEbNKpF6G9k5qZSGpmy3HKbcXhsjcsUBJEc66FptBsKidNHcpJU4e2vHID7n7pBr545XtmvToPr0fnpGnDuPius0JcNBF+YfzM06GzgH0N/p8LNNYrrF9HSmkIISqAZKC44UpCiBuAGwA6depEhP89Tr/6FBZ8uARvbfDoWAjBoFP6NbHVkUdVFabfMJHpN0w8ZseMcJyREqz2a4yP6QSelPJFKeVwKeXw1NRIZYP/RQae3Jfz7jgLu9OGI8qBK9aJK9bJ3z/9XbOCPREiHBHa8QRea0bGeUBDgYPswLJw6+QKITQgHig5Ii2M8Ivj6r9fxJnXTmDFN2uJinUxcspQXNGRyscRjj6yHY+MW2OMlwE9hBBd8Bvdi4FLG60zE7gSWAScD3wX8RdHaI70nFTOvC7iIohwLPmZi8sHfMC3Al8BKvCqlHKDEOLvwHIp5UzgFeAtIcR2oBS/wY4QIUKE9sMvQShISjkbmN1o2V8a/O0BLjiyTYsQIUKEI4cEZDtOh45oU0SIEOF/AxkRl48QIUKEdkG4ijjthYgxjhAhwv8O7XhkLI5X0IMQogjYE+arFBoli/xMifSj/fFL6csvpR/Q+r7kSCkPKzlBCDEncLzWUCylPP1wjtdWjpsxbgohxHIp5fDj3Y7DJdKP9scvpS+/lH7AL6svh0tEQjNChAgR2gERYxwhQoQI7YD2aIxfPN4NOEJE+tH++KX05ZfSD/hl9eWwaHc+4wgRIkT4X6Q9jowjRIgQ4X+OiDGOECFChHbAcTHGv6Saeq3oy51CiI1CiLVCiG+FEDnHo50t0VI/Gqx3nhBCCiHabThSa/oihLgw8LtsEEK8c6zb2BpacW11EkJ8L4RYFbi+zjwe7WwJIcSrQohCIcT6Jr4XQoinA/1cK4RoW9mWXwpSymP6wa/8tgPoCtiBNUDfRuvcDDwf+Pti4P1j3c4j2JdTgKjA379uj31pTT8C68UCC4DFwPDj3e7D+E16AKuAxMD/0453uw+xHy8Cvw783RfYfbzb3URfTgaGAuub+P5M4EtAAKOAJce7zcfjczxGxvU19aSUPqCupl5DpgNvBP7+EJgojkRp4iNPi32RUn4vpawre7wYvzh/e6M1vwnA/fiLzXqOZePaSGv6cj3wrJSyDEBKWXiM29gaWtMPCcQF/o4H9h/D9rUaKeUC/NK6TTEdeFP6WQwkCCE6HJvWtR+OhzEOV1Mvq6l1pJQGUFdTr73Rmr405Fr8I4D2Rov9CLw6dpRSzjqWDTsEWvOb9AR6CiF+FEIsFkIc07TXVtKafvwVuEwIkYtf4vY3x6ZpR5y23ke/SCJCQccIIcRlwHBg3PFuS1sRQijA48BVx7kpRwoNv6tiPP43lQVCiAFSyvLj2ahD4BLgdSnlY0KI0fgLPPSX7bkefYQmOR4j47bU1KOd19RrTV8QQkwC/gRMk1J6G3/fDmipH7FAf2CeEGI3fr/ezHY6idea3yQXmCml1KWUu4Ct+I1ze6I1/bgWmAEgpVwEOGm9EE57olX30S+d42GM62vqCSHs+CfoZjZap66mHrTvmnot9kUIMQR4Ab8hbo++SWihH1LKCillipSys5SyM37f9zQp5fLj09xmac319Sn+UTFCiBT8boudx7CNraE1/dgLTAQQQvTBb4yLjmkrjwwzgSsCURWjgAop5YHj3ahjzvGYNcQ/e7oV/2zxnwLL/o7/Bgf/RfUBsB1YCnQ93jOdh9GXuUABsDrwmXm823wo/Wi07jzaaTRFK38Tgd/tshFYB1x8vNt8iP3oC/yIP9JiNXDq8W5zE/14FzgA6PjfSq4FbgJuavB7PBvo57r2fG0dzU8kHTpChAgR2gGRDLwIESJEaAdEjHGECBEitAMixjhChAgR2gERYxwhQoQI7YCIMY4QIUKEdkDEGEeIECFCOyBijCNEiBChHfD/AzujvzpE/coAAAAASUVORK5CYII=", + "text/plain": [ + "

" + ] + }, + "metadata": { + "needs_background": "light" + } + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's now build a simple autoencoder using the continuous convolutional filter. The data is clearly unstructured and a simple convolutional filter might not work without projecting or interpolating first. Let's first build and `Encoder` and `Decoder` class, and then a `Autoencoder` class that contains both." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 23, + "source": [ + "class Encoder(torch.nn.Module):\n", + " def __init__(self, hidden_dimension):\n", + " super().__init__()\n", + "\n", + " # convolutional block\n", + " self.convolution = ContinuousConv(input_numb_field=1,\n", + " output_numb_field=2,\n", + " stride={\"domain\": [1, 1],\n", + " \"start\": [0, 0],\n", + " \"jumps\": [0.05, 0.05],\n", + " \"direction\": [1, 1.],\n", + " },\n", + " filter_dim=[0.15, 0.15],\n", + " optimize=True)\n", + " # feedforward net\n", + " self.nn = FeedForward(input_variables=400,\n", + " output_variables=hidden_dimension,\n", + " layers=[240, 120])\n", + "\n", + " def forward(self, x):\n", + " # convolution\n", + " x = self.convolution(x)\n", + " # feed forward pass\n", + " return self.nn(x[..., -1])\n", + "\n", + "\n", + "class Decoder(torch.nn.Module):\n", + " def __init__(self, hidden_dimension):\n", + " super().__init__()\n", + "\n", + " # convolutional block\n", + " self.convolution = ContinuousConv(input_numb_field=2,\n", + " output_numb_field=1,\n", + " stride={\"domain\": [1, 1],\n", + " \"start\": [0, 0],\n", + " \"jumps\": [0.05, 0.05],\n", + " \"direction\": [1, 1.],\n", + " },\n", + " filter_dim=[0.15, 0.15],\n", + " optimize=True)\n", + " # feedforward net\n", + " self.nn = FeedForward(input_variables=hidden_dimension,\n", + " output_variables=400,\n", + " layers=[120, 240])\n", + "\n", + " def forward(self, weights, grid):\n", + " # feed forward pass\n", + " x = self.nn(weights)\n", + " # transpose convolution\n", + " return torch.sigmoid(self.convolution.transpose(x, grid))\n" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Very good! Notice that in the `Decoder` class in the `forward` pass we have used the `.transpose()` method of the `ContinuousConvolution` class. This method accepts the `weights` for upsampling and the `grid` on where to upsample. Let's now build the autoencoder! We set the hidden dimension in the `hidden_dimension` variable. We apply the sigmoid on the output since the field value is between $[0, 1]$. " + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 28, + "source": [ + "class Autoencoder(torch.nn.Module):\n", + " def __init__(self, hidden_dimension=10):\n", + " super().__init__()\n", + "\n", + " self.encoder = Encoder(hidden_dimension)\n", + " self.decoder = Decoder(hidden_dimension)\n", + "\n", + " def forward(self, x):\n", + " # saving grid for later upsampling\n", + " grid = x.clone().detach()\n", + " # encoder\n", + " weights = self.encoder(x)\n", + " # decoder\n", + " out = self.decoder(weights, grid)\n", + " return out\n", + "\n", + "\n", + "net = Autoencoder()" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's now train the autoencoder, minimizing the mean square error loss and optimizing using Adam." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 29, + "source": [ + "# setting the seed\n", + "torch.manual_seed(seed)\n", + "\n", + "# optimizer and loss function\n", + "optimizer = torch.optim.Adam(net.parameters(), lr=0.001)\n", + "criterion = torch.nn.MSELoss()\n", + "max_epochs = 150\n", + "\n", + "for epoch in range(max_epochs): # loop over the dataset multiple times\n", + "\n", + " # zero the parameter gradients\n", + " optimizer.zero_grad()\n", + "\n", + " # forward + backward + optimize\n", + " outputs = net(input_data)\n", + " loss = criterion(outputs[..., -1], input_data[..., -1])\n", + " loss.backward()\n", + " optimizer.step()\n", + "\n", + " # print statistics\n", + " if epoch % 10 ==9:\n", + " print(f'epoch [{epoch + 1}/{max_epochs}] loss [{loss.item():.2}]')\n" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "epoch [10/150] loss [0.013]\n", + "epoch [20/150] loss [0.0029]\n", + "epoch [30/150] loss [0.0019]\n", + "epoch [40/150] loss [0.0014]\n", + "epoch [50/150] loss [0.0011]\n", + "epoch [60/150] loss [0.00094]\n", + "epoch [70/150] loss [0.00082]\n", + "epoch [80/150] loss [0.00074]\n", + "epoch [90/150] loss [0.00068]\n", + "epoch [100/150] loss [0.00064]\n", + "epoch [110/150] loss [0.00061]\n", + "epoch [120/150] loss [0.00058]\n", + "epoch [130/150] loss [0.00057]\n", + "epoch [140/150] loss [0.00056]\n", + "epoch [150/150] loss [0.00054]\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's visualize the two solutions side by side!" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 30, + "source": [ + "net.eval()\n", + "\n", + "# get output and detach from computational graph for plotting\n", + "output = net(input_data).detach()\n", + "\n", + "# visualize data\n", + "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3))\n", + "pic1 = axes[0].scatter(grid[:, 0], grid[:, 1], c=input_data[0, 0, :, -1])\n", + "axes[0].set_title(\"Real\")\n", + "fig.colorbar(pic1)\n", + "plt.subplot(1, 2, 2)\n", + "pic2 = axes[1].scatter(grid[:, 0], grid[:, 1], c=output[0, 0, :, -1])\n", + "axes[1].set_title(\"Autoencoder\")\n", + "fig.colorbar(pic2)\n", + "plt.tight_layout()\n", + "plt.show()\n" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + } + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "As we can see the two are really similar! We can compute the $l_2$ error quite easily as well:" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 32, + "source": [ + "def l2_error(input_, target):\n", + " return torch.linalg.norm(input_-target, ord=2)/torch.linalg.norm(input_, ord=2)\n", + "\n", + "\n", + "print(f'l2 error: {l2_error(input_data[0, 0, :, -1], output[0, 0, :, -1]):.2%}')" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "l2 error: 4.10%\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "More or less $4\\%$ in $l_2$ error, which is really low considering the fact that we use just **one** convolutional layer and a simple feedforward to decrease the dimension. Let's see now some peculiarity of the filter." + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "### Filter for upsampling\n", + "\n", + "Suppose we have already the hidden dimension and we want to upsample on a differen grid with more points. Let's see how to do it:" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 33, + "source": [ + "# setting the seed\n", + "torch.manual_seed(seed)\n", + "\n", + "grid2 = circle_grid(1500) # triple number of points\n", + "input_data2 = torch.zeros(size=(1, 1, grid2.shape[0], 3))\n", + "input_data2[0, 0, :, :-1] = grid2\n", + "input_data2[0, 0, :, -1] = torch.sin(pi *\n", + " grid2[:, 0]) * torch.sin(pi * grid2[:, 1])\n", + "\n", + "# get the hidden dimension representation from original input\n", + "latent = net.encoder(input_data)\n", + "\n", + "# upsample on the second input_data2\n", + "output = net.decoder(latent, input_data2).detach()\n", + "\n", + "# show the picture\n", + "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3))\n", + "pic1 = axes[0].scatter(grid2[:, 0], grid2[:, 1], c=input_data2[0, 0, :, -1])\n", + "axes[0].set_title(\"Real\")\n", + "fig.colorbar(pic1)\n", + "plt.subplot(1, 2, 2)\n", + "pic2 = axes[1].scatter(grid2[:, 0], grid2[:, 1], c=output[0, 0, :, -1])\n", + "axes[1].set_title(\"Up-sampling\")\n", + "fig.colorbar(pic2)\n", + "plt.tight_layout()\n", + "plt.show()\n" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + } + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "As we can see we have a very good approximation of the original function, even thought some noise is present. Let's calculate the error now:" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 34, + "source": [ + "print(f'l2 error: {l2_error(input_data2[0, 0, :, -1], output[0, 0, :, -1]):.2%}')" + ], + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "l2 error: 8.44%\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "### Autoencoding at different resolution\n", + "In the previous example we already had the hidden dimension (of original input) and we used it to upsample. Sometimes however we have a more fine mesh solution and we simply want to encode it. This can be done without retraining! This procedure can be useful in case we have many points in the mesh and just a smaller part of them are needed for training. Let's see the results of this:" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 36, + "source": [ + "# setting the seed\n", + "torch.manual_seed(seed)\n", + "\n", + "grid2 = circle_grid(3500) # very fine mesh\n", + "input_data2 = torch.zeros(size=(1, 1, grid2.shape[0], 3))\n", + "input_data2[0, 0, :, :-1] = grid2\n", + "input_data2[0, 0, :, -1] = torch.sin(pi *\n", + " grid2[:, 0]) * torch.sin(pi * grid2[:, 1])\n", + "\n", + "# get the hidden dimension representation from more fine mesh input\n", + "latent = net.encoder(input_data2)\n", + "\n", + "# upsample on the second input_data2\n", + "output = net.decoder(latent, input_data2).detach()\n", + "\n", + "# show the picture\n", + "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3))\n", + "pic1 = axes[0].scatter(grid2[:, 0], grid2[:, 1], c=input_data2[0, 0, :, -1])\n", + "axes[0].set_title(\"Real\")\n", + "fig.colorbar(pic1)\n", + "plt.subplot(1, 2, 2)\n", + "pic2 = axes[1].scatter(grid2[:, 0], grid2[:, 1], c=output[0, 0, :, -1])\n", + "axes[1].set_title(\"Autoencoder not re-trained\")\n", + "fig.colorbar(pic2)\n", + "plt.tight_layout()\n", + "plt.show()\n", + "\n", + "# calculate l2 error\n", + "print(\n", + " f'l2 error: {l2_error(input_data2[0, 0, :, -1], output[0, 0, :, -1]):.2%}')\n" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + } + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "l2 error: 8.45%\n" + ] + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "## What's next?\n", + "\n", + "We have shown the basic usage of a convolutional filter. In the next tutorials we will show how to combine the PINA framework with the convolutional filter to train in few lines and efficiently a Neural Network!" + ], + "metadata": {} + } + ], + "metadata": { + "kernelspec": { + "name": "python3", + "display_name": "Python 3.9.0 64-bit" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.0" + }, + "interpreter": { + "hash": "aee8b7b246df8f9039afb4144a1f6fd8d2ca17a180786b69acc140d282b71a49" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} \ No newline at end of file diff --git a/tutorials/tutorial4/tutorial.py b/tutorials/tutorial4/tutorial.py new file mode 100644 index 0000000..bd8a899 --- /dev/null +++ b/tutorials/tutorial4/tutorial.py @@ -0,0 +1,638 @@ +#!/usr/bin/env python +# coding: utf-8 + +# # Tutorial 4: continuous convolutional filter + +# In this tutorial we will show how to use the Continouous Convolutional Filter, and how to build common Deep Learning architectures with it. The implementation of the filter follows the original work [**A Continuous Convolutional Trainable Filter for Modelling Unstructured Data**](https://arxiv.org/abs/2210.13416) of Coscia Dario, Laura Meneghetti, Nicola Demo, Giovanni Stabile, and Gianluigi Rozza. + +# First of all we import the modules needed for the tutorial, which include: +# +# * `ContinuousConv` class from `pina.model.layers` which implements the continuous convolutional filter +# * `PyTorch` and `Matplotlib` for tensorial operations and visualization respectively + +# In[1]: + + +import torch +import matplotlib.pyplot as plt +from pina.model.layers import ContinuousConv +import torchvision # for MNIST dataset +from pina.model import FeedForward # for building AE and MNIST classification + + +# The tutorial is structured as follow: +# * [Continuous filter background](#continuous-filter-background): understand how the convolutional filter works and how to use it. +# * [Building a MNIST Classifier](#building-a-mnist-classifier): show how to build a simple classifier using the MNIST dataset and how to combine a continuous convolutional layer with a feedforward neural network. +# * [Building a Continuous Convolutional Autoencoder](#building-a-continuous-convolutional-autoencoder): show how to use the continuous filter to work with unstructured data for autoencoding and up-sampling. + +# ## Continuous filter background + +# As reported by the authors in the original paper: in contrast to discrete convolution, continuous convolution is mathematically defined as: +# +# $$ +# \mathcal{I}_{\rm{out}}(\mathbf{x}) = \int_{\mathcal{X}} \mathcal{I}(\mathbf{x} + \mathbf{\tau}) \cdot \mathcal{K}(\mathbf{\tau}) d\mathbf{\tau}, +# $$ +# where $\mathcal{K} : \mathcal{X} \rightarrow \mathbb{R}$ is the *continuous filter* function, and $\mathcal{I} : \Omega \subset \mathbb{R}^N \rightarrow \mathbb{R}$ is the input function. The continuous filter function is approximated using a FeedForward Neural Network, thus trainable during the training phase. The way in which the integral is approximated can be different, currently on **PINA** we approximate it using a simple sum, as suggested by the authors. Thus, given $\{\mathbf{x}_i\}_{i=1}^{n}$ points in $\mathbb{R}^N$ of the input function mapped on the $\mathcal{X}$ filter domain, we approximate the above equation as: +# $$ +# \mathcal{I}_{\rm{out}}(\mathbf{\tilde{x}}_i) = \sum_{{\mathbf{x}_i}\in\mathcal{X}} \mathcal{I}(\mathbf{x}_i + \mathbf{\tau}) \cdot \mathcal{K}(\mathbf{x}_i), +# $$ +# where $\mathbf{\tau} \in \mathcal{S}$, with $\mathcal{S}$ the set of available strides, corresponds to the current stride position of the filter, and $\mathbf{\tilde{x}}_i$ points are obtained by taking the centroid of the filter position mapped on the $\Omega$ domain. + +# We will now try to pratically see how to work with the filter. From the above definition we see that what is needed is: +# 1. A domain and a function defined on that domain (the input) +# 2. A stride, corresponding to the positions where the filter needs to be $\rightarrow$ `stride` variable in `ContinuousConv` +# 3. The filter rectangular domain $\rightarrow$ `filter_dim` variable in `ContinuousConv` + +# ### Input function +# +# The input function for the continuous filter is defined as a tensor of shape: $$[B \times N_{in} \times N \times D]$$ where $B$ is the batch_size, $N_{in}$ is the number of input fields, $N$ the number of points in the mesh, $D$ the dimension of the problem. In particular: +# * $D$ is the number of spatial variables + 1. The last column must contain the field value. For example for 2D problems $D=3$ and the tensor will be something like `[first coordinate, second coordinate, field value]` +# * $N_{in}$ represents the number of vectorial function presented. For example a vectorial function $f = [f_1, f_2]$ will have $N_{in}=2$ +# +# Let's see an example to clear the ideas. We will be verbose to explain in details the input form. We wish to create the function: +# $$ +# f(x, y) = [\sin(\pi x) \sin(\pi y), -\sin(\pi x) \sin(\pi y)] \quad (x,y)\in[0,1]\times[0,1] +# $$ +# +# using a batch size of one. + +# In[2]: + + +# batch size fixed to 1 +batch_size = 1 + +# points in the mesh fixed to 200 +N = 200 + +# vectorial 2 dimensional function, number_input_fileds=2 +number_input_fileds = 2 + +# 2 dimensional spatial variables, D = 2 + 1 = 3 +D = 3 + +# create the function f domain as random 2d points in [0, 1] +domain = torch.rand(size=(batch_size, number_input_fileds, N, D-1)) +print(f"Domain has shape: {domain.shape}") + +# create the functions +pi = torch.acos(torch.tensor([-1.])) # pi value +f1 = torch.sin(pi * domain[:, 0, :, 0]) * torch.sin(pi * domain[:, 0, :, 1]) +f2 = - torch.sin(pi * domain[:, 1, :, 0]) * torch.sin(pi * domain[:, 1, :, 1]) + +# stacking the input domain and field values +data = torch.empty(size=(batch_size, number_input_fileds, N, D)) +data[..., :-1] = domain # copy the domain +data[:, 0, :, -1] = f1 # copy first field value +data[:, 1, :, -1] = f1 # copy second field value +print(f"Filter input data has shape: {data.shape}") + + +# ### Stride +# +# The stride is passed as a dictionary `stride` which tells the filter where to go. Here is an example for the $[0,1]\times[0,5]$ domain: +# +# ```python +# # stride definition +# stride = {"domain": [1, 5], +# "start": [0, 0], +# "jump": [0.1, 0.3], +# "direction": [1, 1], +# } +# ``` +# This tells the filter: +# 1. `domain`: square domain (the only implemented) $[0,1]\times[0,5]$. The minimum value is always zero, while the maximum is specified by the user +# 2. `start`: start position of the filter, coordinate $(0, 0)$ +# 3. `jump`: the jumps of the centroid of the filter to the next position $(0.1, 0.3)$ +# 4. `direction`: the directions of the jump, with `1 = right`, `0 = no jump`,`-1 = left` with respect to the current position +# +# **Note** +# +# We are planning to release the possibility to directly pass a list of possible strides! + +# ### Filter definition +# +# Having defined all the previous blocks we are able to construct the continuous filter. +# +# Suppose we would like to get an ouput with only one field, and let us fix the filter dimension to be $[0.1, 0.1]$. + +# In[4]: + + +# filter dim +filter_dim = [0.1, 0.1] + +# stride +stride = {"domain": [1, 1], + "start": [0, 0], + "jump": [0.08, 0.08], + "direction": [1, 1], + } + +# creating the filter +cConv = ContinuousConv(input_numb_field=number_input_fileds, + output_numb_field=1, + filter_dim=filter_dim, + stride=stride) + + +# That's it! In just one line of code we have created the continuous convolutional filter. By default the `pina.model.FeedForward` neural network is intitialised, more on the [documentation](https://mathlab.github.io/PINA/_rst/fnn.html). In case the mesh doesn't change during training we can set the `optimize` flag equals to `True`, to exploit optimizations for finding the points to convolve. + +# In[5]: + + +# creating the filter + optimization +cConv = ContinuousConv(input_numb_field=number_input_fileds, + output_numb_field=1, + filter_dim=filter_dim, + stride=stride, + optimize=True) + + +# Let's try to do a forward pass + +# In[6]: + + +print(f"Filter input data has shape: {data.shape}") + +#input to the filter +output = cConv(data) + +print(f"Filter output data has shape: {output.shape}") + + +# If we don't want to use the default `FeedForward` neural network, we can pass a specified torch model in the `model` keyword as follow: +# + +# In[7]: + + +class SimpleKernel(torch.nn.Module): + def __init__(self) -> None: + super().__init__() + self. model = torch.nn.Sequential( + torch.nn.Linear(2, 20), + torch.nn.ReLU(), + torch.nn.Linear(20, 20), + torch.nn.ReLU(), + torch.nn.Linear(20, 1)) + + def forward(self, x): + return self.model(x) + + +cConv = ContinuousConv(input_numb_field=number_input_fileds, + output_numb_field=1, + filter_dim=filter_dim, + stride=stride, + optimize=True, + model=SimpleKernel) + + +# Notice that we pass the class and not an already built object! + +# ## Building a MNIST Classifier +# +# Let's see how we can build a MNIST classifier using a continuous convolutional filter. We will use the MNIST dataset from PyTorch. In order to keep small training times we use only 6000 samples for training and 1000 samples for testing. + +# In[9]: + + +from torch.utils.data import DataLoader, SubsetRandomSampler + +numb_training = 6000 # get just 6000 images for training +numb_testing= 1000 # get just 1000 images for training +seed = 111 # for reproducibility +batch_size = 8 # setting batch size + +# setting the seed +torch.manual_seed(seed) + +# downloading the dataset +train_data = torchvision.datasets.MNIST('./data/', train=True, download=True, + transform=torchvision.transforms.Compose([ + torchvision.transforms.ToTensor(), + torchvision.transforms.Normalize( + (0.1307,), (0.3081,)) + ])) +subsample_train_indices = torch.randperm(len(train_data))[:numb_training] +train_loader = DataLoader(train_data, batch_size=batch_size, + sampler=SubsetRandomSampler(subsample_train_indices)) + +test_data = torchvision.datasets.MNIST('./data/', train=False, download=True, + transform=torchvision.transforms.Compose([ + torchvision.transforms.ToTensor(), + torchvision.transforms.Normalize( + (0.1307,), (0.3081,)) + ])) +subsample_test_indices = torch.randperm(len(train_data))[:numb_testing] +test_loader = DataLoader(train_data, batch_size=batch_size, + sampler=SubsetRandomSampler(subsample_train_indices)) + + +# Let's now build a simple classifier. The MNIST dataset is composed by vectors of shape `[batch, 1, 28, 28]`, but we can image them as one field functions where the pixels $ij$ are the coordinate $x=i, y=j$ in a $[0, 27]\times[0,27]$ domain, and the pixels value are the field values. We just need a function to transform the regular tensor in a tensor compatible for the continuous filter: + +# In[10]: + + +def transform_input(x): + batch_size = x.shape[0] + dim_grid = tuple(x.shape[:-3:-1]) + + # creating the n dimensional mesh grid for a single channel image + values_mesh = [torch.arange(0, dim).float() for dim in dim_grid] + mesh = torch.meshgrid(values_mesh) + coordinates_mesh = [x.reshape(-1, 1) for x in mesh] + coordinates = torch.cat(coordinates_mesh, dim=1).unsqueeze( + 0).repeat((batch_size, 1, 1)).unsqueeze(1) + + return torch.cat((coordinates, x.flatten(2).unsqueeze(-1)), dim=-1) + + +# let's try it out +image, s = next(iter(train_loader)) +print(f"Original MNIST image shape: {image.shape}") + +image_transformed = transform_input(image) +print(f"Transformed MNIST image shape: {image_transformed.shape}") + + +# We can now build a simple classifier! We will use just one convolutional filter followed by a feedforward neural network + +# In[19]: + + +# setting the seed +torch.manual_seed(seed) + +class ContinuousClassifier(torch.nn.Module): + def __init__(self): + super().__init__() + + # number of classes for classification + numb_class = 10 + + # convolutional block + self.convolution = ContinuousConv(input_numb_field=1, + output_numb_field=4, + stride={"domain": [27, 27], + "start": [0, 0], + "jumps": [4, 4], + "direction": [1, 1.], + }, + filter_dim=[4, 4], + optimize=True) + # feedforward net + self.nn = FeedForward(input_variables=196, + output_variables=numb_class, + layers=[120, 64], + func=torch.nn.ReLU) + + def forward(self, x): + # transform input + convolution + x = transform_input(x) + x = self.convolution(x) + # feed forward classification + return self.nn(x[..., -1].flatten(1)) + + +net = ContinuousClassifier() + + +# Let's try to train it using a simple pytorch training loop. We train for juts 1 epoch using Adam optimizer with a $0.001$ learning rate. + +# In[20]: + + +# setting the seed +torch.manual_seed(seed) + +# optimizer and loss function +optimizer = torch.optim.Adam(net.parameters(), lr=0.001) +criterion = torch.nn.CrossEntropyLoss() + +for epoch in range(1): # loop over the dataset multiple times + + running_loss = 0.0 + for i, data in enumerate(train_loader, 0): + # get the inputs; data is a list of [inputs, labels] + inputs, labels = data + + # zero the parameter gradients + optimizer.zero_grad() + + # forward + backward + optimize + outputs = net(inputs) + loss = criterion(outputs, labels) + loss.backward() + optimizer.step() + + # print statistics + running_loss += loss.item() + if i % 50 == 49: + print( + f'epoch [{i + 1}/{numb_training//batch_size}] loss[{running_loss / 500:.3f}]') + running_loss = 0.0 + + +# Let's see the performance on the train set! + +# In[21]: + + +correct = 0 +total = 0 +with torch.no_grad(): + for data in test_loader: + images, labels = data + # calculate outputs by running images through the network + outputs = net(images) + # the class with the highest energy is what we choose as prediction + _, predicted = torch.max(outputs.data, 1) + total += labels.size(0) + correct += (predicted == labels).sum().item() + +print( + f'Accuracy of the network on the 1000 test images: {(correct / total):.3%}') + + +# As we can see we have very good performance for having traing only for 1 epoch! Nevertheless, we are still using structured data... Let's see how we can build an autoencoder for unstructured data now. + +# ## Building a Continuous Convolutional Autoencoder +# +# Just as toy problem, we will now build an autoencoder for the following function $f(x,y)=\sin(\pi x)\sin(\pi y)$ on the unit circle domain centered in $(0.5, 0.5)$. We will also see the ability to up-sample (once trained) the results without retraining. Let's first create the input and visualize it, we will use firstly a mesh of $100$ points. + +# In[22]: + + +# create inputs +def circle_grid(N=100): + """Generate points withing a unit 2D circle centered in (0.5, 0.5) + + :param N: number of points + :type N: float + :return: [x, y] array of points + :rtype: torch.tensor + """ + + PI = torch.acos(torch.zeros(1)).item() * 2 + R = 0.5 + centerX = 0.5 + centerY = 0.5 + + r = R * torch.sqrt(torch.rand(N)) + theta = torch.rand(N) * 2 * PI + + x = centerX + r * torch.cos(theta) + y = centerY + r * torch.sin(theta) + + return torch.stack([x, y]).T + +# create the grid +grid = circle_grid(500) + +# create input +input_data = torch.empty(size=(1, 1, grid.shape[0], 3)) +input_data[0, 0, :, :-1] = grid +input_data[0, 0, :, -1] = torch.sin(pi * grid[:, 0]) * torch.sin(pi * grid[:, 1]) + +# visualize data +plt.title("Training sample with 500 points") +plt.scatter(grid[:, 0], grid[:, 1], c=input_data[0, 0, :, -1]) +plt.colorbar() +plt.show() + + +# Let's now build a simple autoencoder using the continuous convolutional filter. The data is clearly unstructured and a simple convolutional filter might not work without projecting or interpolating first. Let's first build and `Encoder` and `Decoder` class, and then a `Autoencoder` class that contains both. + +# In[23]: + + +class Encoder(torch.nn.Module): + def __init__(self, hidden_dimension): + super().__init__() + + # convolutional block + self.convolution = ContinuousConv(input_numb_field=1, + output_numb_field=2, + stride={"domain": [1, 1], + "start": [0, 0], + "jumps": [0.05, 0.05], + "direction": [1, 1.], + }, + filter_dim=[0.15, 0.15], + optimize=True) + # feedforward net + self.nn = FeedForward(input_variables=400, + output_variables=hidden_dimension, + layers=[240, 120]) + + def forward(self, x): + # convolution + x = self.convolution(x) + # feed forward pass + return self.nn(x[..., -1]) + + +class Decoder(torch.nn.Module): + def __init__(self, hidden_dimension): + super().__init__() + + # convolutional block + self.convolution = ContinuousConv(input_numb_field=2, + output_numb_field=1, + stride={"domain": [1, 1], + "start": [0, 0], + "jumps": [0.05, 0.05], + "direction": [1, 1.], + }, + filter_dim=[0.15, 0.15], + optimize=True) + # feedforward net + self.nn = FeedForward(input_variables=hidden_dimension, + output_variables=400, + layers=[120, 240]) + + def forward(self, weights, grid): + # feed forward pass + x = self.nn(weights) + # transpose convolution + return torch.sigmoid(self.convolution.transpose(x, grid)) + + +# Very good! Notice that in the `Decoder` class in the `forward` pass we have used the `.transpose()` method of the `ContinuousConvolution` class. This method accepts the `weights` for upsampling and the `grid` on where to upsample. Let's now build the autoencoder! We set the hidden dimension in the `hidden_dimension` variable. We apply the sigmoid on the output since the field value is between $[0, 1]$. + +# In[28]: + + +class Autoencoder(torch.nn.Module): + def __init__(self, hidden_dimension=10): + super().__init__() + + self.encoder = Encoder(hidden_dimension) + self.decoder = Decoder(hidden_dimension) + + def forward(self, x): + # saving grid for later upsampling + grid = x.clone().detach() + # encoder + weights = self.encoder(x) + # decoder + out = self.decoder(weights, grid) + return out + + +net = Autoencoder() + + +# Let's now train the autoencoder, minimizing the mean square error loss and optimizing using Adam. + +# In[29]: + + +# setting the seed +torch.manual_seed(seed) + +# optimizer and loss function +optimizer = torch.optim.Adam(net.parameters(), lr=0.001) +criterion = torch.nn.MSELoss() +max_epochs = 150 + +for epoch in range(max_epochs): # loop over the dataset multiple times + + # zero the parameter gradients + optimizer.zero_grad() + + # forward + backward + optimize + outputs = net(input_data) + loss = criterion(outputs[..., -1], input_data[..., -1]) + loss.backward() + optimizer.step() + + # print statistics + if epoch % 10 ==9: + print(f'epoch [{epoch + 1}/{max_epochs}] loss [{loss.item():.2}]') + + +# Let's visualize the two solutions side by side! + +# In[30]: + + +net.eval() + +# get output and detach from computational graph for plotting +output = net(input_data).detach() + +# visualize data +fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3)) +pic1 = axes[0].scatter(grid[:, 0], grid[:, 1], c=input_data[0, 0, :, -1]) +axes[0].set_title("Real") +fig.colorbar(pic1) +plt.subplot(1, 2, 2) +pic2 = axes[1].scatter(grid[:, 0], grid[:, 1], c=output[0, 0, :, -1]) +axes[1].set_title("Autoencoder") +fig.colorbar(pic2) +plt.tight_layout() +plt.show() + + +# As we can see the two are really similar! We can compute the $l_2$ error quite easily as well: + +# In[32]: + + +def l2_error(input_, target): + return torch.linalg.norm(input_-target, ord=2)/torch.linalg.norm(input_, ord=2) + + +print(f'l2 error: {l2_error(input_data[0, 0, :, -1], output[0, 0, :, -1]):.2%}') + + +# More or less $4\%$ in $l_2$ error, which is really low considering the fact that we use just **one** convolutional layer and a simple feedforward to decrease the dimension. Let's see now some peculiarity of the filter. + +# ### Filter for upsampling +# +# Suppose we have already the hidden dimension and we want to upsample on a differen grid with more points. Let's see how to do it: + +# In[33]: + + +# setting the seed +torch.manual_seed(seed) + +grid2 = circle_grid(1500) # triple number of points +input_data2 = torch.zeros(size=(1, 1, grid2.shape[0], 3)) +input_data2[0, 0, :, :-1] = grid2 +input_data2[0, 0, :, -1] = torch.sin(pi * + grid2[:, 0]) * torch.sin(pi * grid2[:, 1]) + +# get the hidden dimension representation from original input +latent = net.encoder(input_data) + +# upsample on the second input_data2 +output = net.decoder(latent, input_data2).detach() + +# show the picture +fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3)) +pic1 = axes[0].scatter(grid2[:, 0], grid2[:, 1], c=input_data2[0, 0, :, -1]) +axes[0].set_title("Real") +fig.colorbar(pic1) +plt.subplot(1, 2, 2) +pic2 = axes[1].scatter(grid2[:, 0], grid2[:, 1], c=output[0, 0, :, -1]) +axes[1].set_title("Up-sampling") +fig.colorbar(pic2) +plt.tight_layout() +plt.show() + + +# As we can see we have a very good approximation of the original function, even thought some noise is present. Let's calculate the error now: + +# In[34]: + + +print(f'l2 error: {l2_error(input_data2[0, 0, :, -1], output[0, 0, :, -1]):.2%}') + + +# ### Autoencoding at different resolution +# In the previous example we already had the hidden dimension (of original input) and we used it to upsample. Sometimes however we have a more fine mesh solution and we simply want to encode it. This can be done without retraining! This procedure can be useful in case we have many points in the mesh and just a smaller part of them are needed for training. Let's see the results of this: + +# In[36]: + + +# setting the seed +torch.manual_seed(seed) + +grid2 = circle_grid(3500) # very fine mesh +input_data2 = torch.zeros(size=(1, 1, grid2.shape[0], 3)) +input_data2[0, 0, :, :-1] = grid2 +input_data2[0, 0, :, -1] = torch.sin(pi * + grid2[:, 0]) * torch.sin(pi * grid2[:, 1]) + +# get the hidden dimension representation from more fine mesh input +latent = net.encoder(input_data2) + +# upsample on the second input_data2 +output = net.decoder(latent, input_data2).detach() + +# show the picture +fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8, 3)) +pic1 = axes[0].scatter(grid2[:, 0], grid2[:, 1], c=input_data2[0, 0, :, -1]) +axes[0].set_title("Real") +fig.colorbar(pic1) +plt.subplot(1, 2, 2) +pic2 = axes[1].scatter(grid2[:, 0], grid2[:, 1], c=output[0, 0, :, -1]) +axes[1].set_title("Autoencoder not re-trained") +fig.colorbar(pic2) +plt.tight_layout() +plt.show() + +# calculate l2 error +print( + f'l2 error: {l2_error(input_data2[0, 0, :, -1], output[0, 0, :, -1]):.2%}') + + +# ## What's next? +# +# We have shown the basic usage of a convolutional filter. In the next tutorials we will show how to combine the PINA framework with the convolutional filter to train in few lines and efficiently a Neural Network!