From e0a81dca263df428deeca10dee7783346d64411b Mon Sep 17 00:00:00 2001 From: Luo Tao Date: Tue, 3 Jan 2017 19:42:51 +0800 Subject: [PATCH 1/2] add translated chinese docs into catalog --- doc/howto/deep_model/rnn/index_cn.rst | 1 + doc/howto/deep_model/rnn/rnn_cn.md | 226 ------------------ .../deep_model/{ => rnn}/rnn_config_cn.rst | 59 +++-- doc/howto/index_cn.rst | 5 +- doc/howto/index_en.rst | 4 +- doc/howto/usage/cmd_parameter/index_cn.rst | 11 + doc/howto/usage/cmd_parameter/index_en.md | 8 - doc/howto/usage/cmd_parameter/index_en.rst | 11 + .../k8s-aws/README.md => k8s/k8s_aws_en.md} | 8 +- doc/howto/usage/{cluster => }/k8s/k8s_cn.md | 2 +- .../{cluster => }/k8s/k8s_distributed_cn.md | 4 +- doc/howto/usage/{cluster => }/k8s/k8s_en.md | 0 .../usage/{cluster/k8s => k8s/src}/Dockerfile | 0 .../src}/add_security_group.png | Bin .../k8s-aws => k8s/src}/create_efs.png | Bin .../k8s-aws => k8s/src}/efs_mount.png | Bin .../usage/{cluster/k8s => k8s/src}/job.yaml | 0 .../k8s => k8s/src}/k8s-paddle-arch.png | Bin .../k8s-aws => k8s/src}/managed_policy.png | Bin .../usage/{cluster/k8s => k8s/src}/start.sh | 0 .../{cluster/k8s => k8s/src}/start_paddle.py | 0 doc/tutorials/index_cn.md | 2 + doc/tutorials/index_en.md | 1 + 23 files changed, 67 insertions(+), 275 deletions(-) delete mode 100644 doc/howto/deep_model/rnn/rnn_cn.md rename doc/howto/deep_model/{ => rnn}/rnn_config_cn.rst (88%) create mode 100644 doc/howto/usage/cmd_parameter/index_cn.rst delete mode 100644 doc/howto/usage/cmd_parameter/index_en.md create mode 100644 doc/howto/usage/cmd_parameter/index_en.rst rename doc/howto/usage/{cluster/k8s-aws/README.md => k8s/k8s_aws_en.md} (99%) rename doc/howto/usage/{cluster => }/k8s/k8s_cn.md (99%) rename doc/howto/usage/{cluster => }/k8s/k8s_distributed_cn.md (99%) rename doc/howto/usage/{cluster => }/k8s/k8s_en.md (100%) rename doc/howto/usage/{cluster/k8s => k8s/src}/Dockerfile (100%) rename doc/howto/usage/{cluster/k8s-aws => k8s/src}/add_security_group.png (100%) rename doc/howto/usage/{cluster/k8s-aws => k8s/src}/create_efs.png (100%) rename doc/howto/usage/{cluster/k8s-aws => k8s/src}/efs_mount.png (100%) rename doc/howto/usage/{cluster/k8s => k8s/src}/job.yaml (100%) rename doc/howto/usage/{cluster/k8s => k8s/src}/k8s-paddle-arch.png (100%) rename doc/howto/usage/{cluster/k8s-aws => k8s/src}/managed_policy.png (100%) rename doc/howto/usage/{cluster/k8s => k8s/src}/start.sh (100%) rename doc/howto/usage/{cluster/k8s => k8s/src}/start_paddle.py (100%) diff --git a/doc/howto/deep_model/rnn/index_cn.rst b/doc/howto/deep_model/rnn/index_cn.rst index 9e805ca85191b..9ecab5594cff4 100644 --- a/doc/howto/deep_model/rnn/index_cn.rst +++ b/doc/howto/deep_model/rnn/index_cn.rst @@ -4,6 +4,7 @@ RNN相关模型 .. toctree:: :maxdepth: 1 + rnn_config_cn.rst recurrent_group_cn.md hierarchical_layer_cn.rst hrnn_rnn_api_compare_cn.rst diff --git a/doc/howto/deep_model/rnn/rnn_cn.md b/doc/howto/deep_model/rnn/rnn_cn.md deleted file mode 100644 index 5ec05b2cab9ba..0000000000000 --- a/doc/howto/deep_model/rnn/rnn_cn.md +++ /dev/null @@ -1,226 +0,0 @@ -RNN 配置 -================= - -本教程将指导你如何在 PaddlePaddle 中配置循环神经网络（RNN）。PaddlePaddle 高度支持灵活和高效的循环神经网络配置。 在本教程中，您将了解如何： - -- 准备用来学习循环神经网络的序列数据。 -- 配置循环神经网络架构。 -- 使用学习完成的循环神经网络模型生成序列。 - -我们将使用 vanilla 循环神经网络和 sequence to sequence 模型来指导你完成这些步骤。sequence to sequence 模型的代码可以在`demo / seqToseq`找到。 - -准备序列数据 ---------------------- - -PaddlePaddle 不需要对序列数据进行任何预处理，例如填充。唯一需要做的是将相应类型设置为输入。例如，以下代码段定义了三个输入。 它们都是序列，它们的大小是`src_dict`，`trg_dict`和`trg_dict`： - -``` sourceCode -settings.input_types = [ - integer_value_sequence(len(settings.src_dict)), - integer_value_sequence(len(settings.trg_dict)), - integer_value_sequence(len(settings.trg_dict))] -``` - -在`process`函数中，每个`yield`函数将返回三个整数列表。每个整数列表被视为一个整数序列： - -``` sourceCode -yield src_ids, trg_ids, trg_ids_next -``` - -有关如何编写数据提供程序的更多细节描述，请参考 [PyDataProvider2](../../ui/data_provider/index.html)。完整的数据提供文件在 `demo/seqToseq/dataprovider.py`。 - -配置循环神经网络架构 ------------------------------------------------ - -### 简单门控循环神经网络(Gated Recurrent Neural Network) - -循环神经网络在每个时间步骤顺序地处理序列。下面列出了 LSTM 的架构的示例。 - -![image](../../../tutorials/sentiment_analysis/bi_lstm.jpg) - -一般来说，循环网络从 *t* = 1 到 *t* = *T* 或者反向地从 *t* = *T* 到 *t* = 1 执行以下操作。 - -*x**t* + 1 = *f**x*(*x**t*),*y**t* = *f**y*(*x**t*) - -其中 *f**x*(.) 称为**单步函数**（即单时间步执行的函数，step function），而 *f**y*(.) 称为**输出函数**。在 vanilla 循环神经网络中，单步函数和输出函数都非常简单。然而，PaddlePaddle 可以通过修改这两个函数来实现复杂的网络配置。我们将使用 sequence to sequence 模型演示如何配置复杂的循环神经网络模型。在本节中，我们将使用简单的 vanilla 循环神经网络作为使用`recurrent_group`配置简单循环神经网络的例子。 注意，如果你只需要使用简单的RNN，GRU或LSTM，那么推荐使用`grumemory`和`lstmemory`，因为它们的计算效率比`recurrent_group`更高。 - -对于 vanilla RNN，在每个时间步长，**单步函数**为： - -*x**t* + 1 = *W**x**x**t* + *W**i**I**t* + *b* - -其中 *x**t* 是RNN状态，并且 *I**t* 是输入，*W**x* 和 *W**i* 分别是RNN状态和输入的变换矩阵。*b* 是偏差。它的**输出函数**只需要*x**t*作为输出。 - -`recurrent_group`是构建循环神经网络的最重要的工具。 它定义了**单步函数**，**输出函数**和循环神经网络的输入。注意，这个函数的`step`参数需要实现`step function`（单步函数）和`output function`（输出函数）： - - -``` sourceCode -def simple_rnn(input, - size=None, - name=None, - reverse=False, - rnn_bias_attr=None, - act=None, - rnn_layer_attr=None): - def __rnn_step__(ipt): - out_mem = memory(name=name, size=size) - rnn_out = mixed_layer(input = [full_matrix_projection(ipt), - full_matrix_projection(out_mem)], - name = name, - bias_attr = rnn_bias_attr, - act = act, - layer_attr = rnn_layer_attr, - size = size) - return rnn_out - return recurrent_group(name='%s_recurrent_group' % name, - step=__rnn_step__, - reverse=reverse, - input=input) -``` - -PaddlePaddle 使用“Memory”（记忆模块）实现单步函数。**Memory**是在PaddlePaddle中构造循环神经网络时最重要的概念。 Memory是在单步函数中循环使用的状态，例如*x**t* + 1 = *f**x*(*x**t*)。 一个Memory包含**输出**和**输入**。当前时间步处的Memory的输出作为下一时间步Memory的输入。Memory也可以具有**boot layer(引导层)**，其输出被用作Memory的初始值。 在我们的例子中，门控循环单元的输出被用作输出Memory。请注意，`rnn_out`层的名称与`out_mem`的名称相同。这意味着`rnn_out` (*x**t* + 1)的输出被用作`out_mem`Memory的**输出**。 - -Memory也可以是序列。在这种情况下，在每个时间步中，我们有一个序列作为循环神经网络的状态。这在构造非常复杂的循环神经网络时是有用的。 其他高级功能包括定义多个Memory，以及使用子序列来定义分级循环神经网络架构。 - -我们在函数的结尾返回`rnn_out`。 这意味着 `rnn_out` 层的输出被用作门控循环神经网络的**输出**函数。 - -### Sequence to Sequence Model with Attention - -我们将使用 sequence to sequence model with attention 作为例子演示如何配置复杂的循环神经网络模型。该模型的说明如下图所示。 - -![image](../../../tutorials/text_generation/encoder-decoder-attention-model.png) - -在这个模型中，源序列 *S* = {*s*1, …, *s**T*} 用双向门控循环神经网络编码。双向门控循环神经网络的隐藏状态 *H**S* = {*H*1, …, *H**T*} 被称为 *编码向量*。解码器是门控循环神经网络。当解读每一个*y**t*时, 这个门控循环神经网络生成一系列权重 *W**S**t* = {*W*1*t*, …, *W**T**t*}, 用于计算编码向量的加权和。加权和用来生成*y**t*。 - -模型的编码器部分如下所示。它叫做`grumemory`来表示门控循环神经网络。如果网络架构简单，那么推荐使用循环神经网络的方法，因为它比 `recurrent_group` 更快。我们已经实现了大多数常用的循环神经网络架构，可以参考 [Layers](../../ui/api/trainer_config_helpers/layers_index.html) 了解更多细节。 - -我们还将编码向量投射到 `decoder_size` 维空间。这通过获得反向循环网络的第一个实例，并将其投射到 `decoder_size` 维空间完成： - -``` sourceCode -# 定义源语句的数据层 -src_word_id = data_layer(name='source_language_word', size=source_dict_dim) -# 计算每个词的词向量 -src_embedding = embedding_layer( - input=src_word_id, - size=word_vector_dim, - param_attr=ParamAttr(name='_source_language_embedding')) -# 应用前向循环神经网络 -src_forward = grumemory(input=src_embedding, size=encoder_size) -# 应用反向递归神经网络（reverse=True表示反向循环神经网络） -src_backward = grumemory(input=src_embedding, - size=encoder_size, - reverse=True) -# 将循环神经网络的前向和反向部分混合在一起 -encoded_vector = concat_layer(input=[src_forward, src_backward]) - -# 投射编码向量到 decoder_size -encoder_proj = mixed_layer(input = [full_matrix_projection(encoded_vector)], - size = decoder_size) - -# 计算反向RNN的第一个实例 -backward_first = first_seq(input=src_backward) - -# 投射反向RNN的第一个实例到 decoder size -decoder_boot = mixed_layer(input=[full_matrix_projection(backward_first)], size=decoder_size, act=TanhActivation()) -``` - -解码器使用 `recurrent_group` 来定义循环神经网络。单步函数和输出函数在 `gru_decoder_with_attention` 中定义： - -``` sourceCode -group_inputs=[StaticInput(input=encoded_vector,is_seq=True), - StaticInput(input=encoded_proj,is_seq=True)] -trg_embedding = embedding_layer( - input=data_layer(name='target_language_word', - size=target_dict_dim), - size=word_vector_dim, - param_attr=ParamAttr(name='_target_language_embedding')) -group_inputs.append(trg_embedding) - -# 对于配备有注意力机制的解码器，在训练中， -# 目标向量（groudtruth）是数据输入， -# 而源序列的编码向量可以被无边界的memory访问 -# StaticInput 意味着不同时间步的输入都是相同的值， -# 否则它以一个序列输入，不同时间步的输入是不同的。 -# 所有输入序列应该有相同的长度。 -decoder = recurrent_group(name=decoder_group_name, - step=gru_decoder_with_attention, - input=group_inputs) -``` - -单步函数的实现如下所示。首先，它定义解码网络的**Memory**。然后定义 attention，门控循环单元单步函数和输出函数： - -``` sourceCode -def gru_decoder_with_attention(enc_vec, enc_proj, current_word): - # 定义解码器的Memory - # Memory的输出定义在 gru_step 内 - # 注意 gru_step 应该与它的Memory名字相同 - decoder_mem = memory(name='gru_decoder', - size=decoder_size, - boot_layer=decoder_boot) - # 计算 attention 加权编码向量 - context = simple_attention(encoded_sequence=enc_vec, - encoded_proj=enc_proj, - decoder_state=decoder_mem) - # 混合当前词向量和attention加权编码向量 - decoder_inputs = mixed_layer(inputs = [full_matrix_projection(context), - full_matrix_projection(current_word)], - size = decoder_size * 3) - # 定义门控循环单元循环神经网络单步函数 - gru_step = gru_step_layer(name='gru_decoder', - input=decoder_inputs, - output_mem=decoder_mem, - size=decoder_size) - # 定义输出函数 - out = mixed_layer(input=[full_matrix_projection(input=gru_step)], - size=target_dict_dim, - bias_attr=True, - act=SoftmaxActivation()) - return out -``` - -生成序列 ------------------ - -训练模型后，我们可以使用它来生成序列。通常的做法是使用**beam search** 生成序列。以下代码片段定义 beam search 算法。注意，`beam_search` 函数假设 `step` 的输出函数返回的是下一个时刻输出词的 softmax 归一化概率向量。我们对模型进行了以下更改。 - -- 使用 `GeneratedInput` 来表示 trg\_embedding。 `GeneratedInput` 将上一时间步所生成的词的向量来作为当前时间步的输入。 -- 使用 `beam_search` 函数。这个函数需要设置： - - `bos_id`: 开始标记。每个句子都以开始标记开头。 - - `eos_id`: 结束标记。每个句子都以结束标记结尾。 - - `beam_size`: beam search 算法中的beam大小。 - - `max_length`: 生成序列的最大长度。 -- 使用 `seqtext_printer_evaluator` 根据索引矩阵和字典打印文本。这个函数需要设置： - - `id_input`: 数据的整数ID，用于标识生成的文件中的相应输出。 - - `dict_file`: 用于将词ID转换为词的字典文件。 - - `result_file`: 生成结果文件的路径。 - -代码如下： - -``` sourceCode -group_inputs=[StaticInput(input=encoded_vector,is_seq=True), - StaticInput(input=encoded_proj,is_seq=True)] -# 在生成时，解码器基于编码源序列和最后生成的目标词预测下一目标词。 -# 编码源序列（编码器输出）必须由只读Memory的 StaticInput 指定。 -# 这里， GeneratedInputs 自动获取上一个生成的词，并在最开始初始化为起始词，如 。 -trg_embedding = GeneratedInput( - size=target_dict_dim, - embedding_name='_target_language_embedding', - embedding_size=word_vector_dim) -group_inputs.append(trg_embedding) -beam_gen = beam_search(name=decoder_group_name, - step=gru_decoder_with_attention, - input=group_inputs, - bos_id=0, # Beginnning token. - eos_id=1, # End of sentence token. - beam_size=beam_size, - max_length=max_length) - -seqtext_printer_evaluator(input=beam_gen, - id_input=data_layer(name="sent_id", size=1), - dict_file=trg_dict_path, - result_file=gen_trans_file) -outputs(beam_gen) -``` - -注意，这种生成技术只用于类似解码器的生成过程。如果你正在处理序列标记任务，请参阅 [Semantic Role Labeling Demo](../../demo/semantic_role_labeling/index.html) 了解更多详细信息。 - -完整的配置文件在`demo/seqToseq/seqToseq_net.py`。 diff --git a/doc/howto/deep_model/rnn_config_cn.rst b/doc/howto/deep_model/rnn/rnn_config_cn.rst similarity index 88% rename from doc/howto/deep_model/rnn_config_cn.rst rename to doc/howto/deep_model/rnn/rnn_config_cn.rst index e6d8c1133a5e8..8d65b3512d0d9 100644 --- a/doc/howto/deep_model/rnn_config_cn.rst +++ b/doc/howto/deep_model/rnn/rnn_config_cn.rst @@ -1,4 +1,4 @@ -RNN 配置 +RNN配置 ======== 本教程将指导你如何在 PaddlePaddle @@ -20,7 +20,7 @@ PaddlePaddle 不需要对序列数据进行任何预处理，例如填充。唯一需要做的是将相应类型设置为输入。例如，以下代码段定义了三个输入。 它们都是序列，它们的大小是\ ``src_dict``\ ，\ ``trg_dict``\ 和\ ``trg_dict``\ ： -.. code:: sourcecode +.. code:: python settings.input_types = [ integer_value_sequence(len(settings.src_dict)), @@ -29,7 +29,7 @@ PaddlePaddle 在\ ``process``\ 函数中，每个\ ``yield``\ 函数将返回三个整数列表。每个整数列表被视为一个整数序列： -.. code:: sourcecode +.. code:: python yield src_ids, trg_ids, trg_ids_next @@ -45,18 +45,17 @@ PaddlePaddle 循环神经网络在每个时间步骤顺序地处理序列。下面列出了 LSTM 的架构的示例。 -.. figure:: ../../../tutorials/sentiment_analysis/bi_lstm.jpg - :alt: image +.. image:: ../../../tutorials/sentiment_analysis/bi_lstm.jpg + :align: center - image +一般来说，循环网络从 :math:`t=1` 到 :math:`t=T` 或者反向地从 :math:`t=T` 到 :math:`t=1` 执行以下操作。 -一般来说，循环网络从 *t* = 1 到 *t* = *T* 或者反向地从 *t* = *T* 到 *t* -= 1 执行以下操作。 +.. math:: -*x*\ \ *t* + 1 = *f*\ \ *x*\ (*x*\ \ *t*\ ),\ *y*\ \ *t*\  = *f*\ \ *y*\ (*x*\ \ *t*\ ) + x_{t+1} = f_x(x_t), y_t = f_y(x_t) -其中 *f*\ \ *x*\ (.) 称为\ **单步函数**\ （即单时间步执行的函数，step -function），而 *f*\ \ *y*\ (.) 称为\ **输出函数**\ 。在 vanilla +其中 :math:`f_x(.)` 称为\ **单步函数**\ （即单时间步执行的函数，step +function），而 :math:`f_y(.)` 称为\ **输出函数**\ 。在 vanilla 循环神经网络中，单步函数和输出函数都非常简单。然而，PaddlePaddle 可以通过修改这两个函数来实现复杂的网络配置。我们将使用 sequence to sequence @@ -67,16 +66,17 @@ vanilla 对于 vanilla RNN，在每个时间步长，\ **单步函数**\ 为： -*x*\ \ *t* + 1 = *W*\ \ *x*\ \ *x*\ \ *t*\  + *W*\ \ *i*\ \ *I*\ \ *t*\  + *b* +.. math:: -其中 *x*\ \ *t*\ 是RNN状态，并且 *I*\ \ *t*\ 是输入，\ *W*\ \ *x*\ 和 -*W*\ \ *i*\ 分别是RNN状态和输入的变换矩阵。\ *b* -是偏差。它的\ **输出函数**\ 只需要\ *x*\ \ *t*\ 作为输出。 + x_{t+1} = W_x x_t + W_i I_t + b + +其中 :math:`x_t` 是RNN状态，并且 :math:`I_t` 是输入，:math:`W_x` 和 +:math:`W_i` 分别是RNN状态和输入的变换矩阵。:math:`b` 是偏差。它的\ **输出函数**\ 只需要 :math:`x_t` 作为输出。 ``recurrent_group``\ 是构建循环神经网络的最重要的工具。 它定义了\ **单步函数**\ ，\ **输出函数**\ 和循环神经网络的输入。注意，这个函数的\ ``step``\ 参数需要实现\ ``step function``\ （单步函数）和\ ``output function``\ （输出函数）： -.. code:: sourcecode +.. code:: python def simple_rnn(input, size=None, @@ -102,7 +102,7 @@ vanilla PaddlePaddle 使用“Memory”（记忆模块）实现单步函数。\ **Memory**\ 是在PaddlePaddle中构造循环神经网络时最重要的概念。 -Memory是在单步函数中循环使用的状态，例如\ *x*\ \ *t* + 1 = *f*\ \ *x*\ (*x*\ \ *t*\ )。 +Memory是在单步函数中循环使用的状态，例如 :math:`x_{t+1} = f_x(x_t)` 。 一个Memory包含\ **输出**\ 和\ **输入**\ 。当前时间步处的Memory的输出作为下一时间步Memory的输入。Memory也可以具有\ **boot layer(引导层)**\ ，其输出被用作Memory的初始值。 在我们的例子中，门控循环单元的输出被用作输出Memory。请注意，\ ``rnn_out``\ 层的名称与\ ``out_mem``\ 的名称相同。这意味着\ ``rnn_out`` @@ -120,18 +120,15 @@ Sequence to Sequence Model with Attention 我们将使用 sequence to sequence model with attention 作为例子演示如何配置复杂的循环神经网络模型。该模型的说明如下图所示。 -.. figure:: ../../../tutorials/text_generation/encoder-decoder-attention-model.png - :alt: image - - image +.. image:: ../../../tutorials/text_generation/encoder-decoder-attention-model.png + :align: center -在这个模型中，源序列 *S* = {*s*\ 1, …, \ *s*\ \ *T*\ } +在这个模型中，源序列 :math:`S = \{s_1, \dots, s_T\}` 用双向门控循环神经网络编码。双向门控循环神经网络的隐藏状态 -*H*\ \ *S*\  = {*H*\ 1, …, \ *H*\ \ *T*\ } 被称为 -*编码向量*\ 。解码器是门控循环神经网络。当解读每一个\ *y*\ \ *t*\ 时, -这个门控循环神经网络生成一系列权重 -*W*\ \ *S*\ \ *t*\  = {*W*\ 1\ *t*\ , …, \ *W*\ \ *T*\ \ *t*\ }, -用于计算编码向量的加权和。加权和用来生成\ *y*\ \ *t*\ 。 +:math:`H_S = \{H_1, \dots, H_T\}` 被称为 +*编码向量*\ 。解码器是门控循环神经网络。当解读每一个 :math:`y_t` 时, +这个门控循环神经网络生成一系列权重 :math:`W_S^t = \{W_1^t, \dots, W_T^t\}` , +用于计算编码向量的加权和。加权和用来生成 :math:`y_t` 。 模型的编码器部分如下所示。它叫做\ ``grumemory``\ 来表示门控循环神经网络。如果网络架构简单，那么推荐使用循环神经网络的方法，因为它比 ``recurrent_group`` @@ -143,7 +140,7 @@ Sequence to Sequence Model with Attention 维空间。这通过获得反向循环网络的第一个实例，并将其投射到 ``decoder_size`` 维空间完成： -.. code:: sourcecode +.. code:: python # 定义源语句的数据层 src_word_id = data_layer(name='source_language_word', size=source_dict_dim) @@ -174,7 +171,7 @@ Sequence to Sequence Model with Attention 解码器使用 ``recurrent_group`` 来定义循环神经网络。单步函数和输出函数在 ``gru_decoder_with_attention`` 中定义： -.. code:: sourcecode +.. code:: python group_inputs=[StaticInput(input=encoded_vector,is_seq=True), StaticInput(input=encoded_proj,is_seq=True)] @@ -198,7 +195,7 @@ Sequence to Sequence Model with Attention 单步函数的实现如下所示。首先，它定义解码网络的\ **Memory**\ 。然后定义 attention，门控循环单元单步函数和输出函数： -.. code:: sourcecode +.. code:: python def gru_decoder_with_attention(enc_vec, enc_proj, current_word): # 定义解码器的Memory @@ -253,7 +250,7 @@ attention，门控循环单元单步函数和输出函数： 代码如下： -.. code:: sourcecode +.. code:: python group_inputs=[StaticInput(input=encoded_vector,is_seq=True), StaticInput(input=encoded_proj,is_seq=True)] diff --git a/doc/howto/index_cn.rst b/doc/howto/index_cn.rst index 6a14ce8ae75c3..bd3d0ec292057 100644 --- a/doc/howto/index_cn.rst +++ b/doc/howto/index_cn.rst @@ -7,10 +7,11 @@ .. toctree:: :maxdepth: 1 + usage/cmd_parameter/index_cn.rst usage/concepts/use_concepts_cn.rst usage/cluster/cluster_train_cn.md - usage/cluster/k8s/k8s_cn.md - usage/cluster/k8s/k8s_distributed_cn.md + usage/k8s/k8s_cn.md + usage/k8s/k8s_distributed_cn.md 开发标准 -------- diff --git a/doc/howto/index_en.rst b/doc/howto/index_en.rst index 983dc743eb453..1fbfcd260b912 100644 --- a/doc/howto/index_en.rst +++ b/doc/howto/index_en.rst @@ -7,8 +7,10 @@ Usage .. toctree:: :maxdepth: 1 - usage/cmd_parameter/index_en.md + usage/cmd_parameter/index_en.rst usage/cluster/cluster_train_en.md + usage/k8s/k8s_en.md + usage/k8s/k8s_aws_en.md Development ------------ diff --git a/doc/howto/usage/cmd_parameter/index_cn.rst b/doc/howto/usage/cmd_parameter/index_cn.rst new file mode 100644 index 0000000000000..4c8729821110b --- /dev/null +++ b/doc/howto/usage/cmd_parameter/index_cn.rst @@ -0,0 +1,11 @@ +.. _cmd_line_index: + +设置命令行参数 +=============== + +.. toctree:: + :maxdepth: 1 + + use_case_cn.md + arguments_cn.md + detail_introduction_cn.md diff --git a/doc/howto/usage/cmd_parameter/index_en.md b/doc/howto/usage/cmd_parameter/index_en.md deleted file mode 100644 index 2a96e7e976c43..0000000000000 --- a/doc/howto/usage/cmd_parameter/index_en.md +++ /dev/null @@ -1,8 +0,0 @@ -```eval_rst -.. _cmd_line_index: -``` -# Set Command-line Parameters - -* [Use Case](use_case_en.md) -* [Arguments](arguments_en.md) -* [Detailed Descriptions](detail_introduction_en.md) diff --git a/doc/howto/usage/cmd_parameter/index_en.rst b/doc/howto/usage/cmd_parameter/index_en.rst new file mode 100644 index 0000000000000..0e3c72d27aca0 --- /dev/null +++ b/doc/howto/usage/cmd_parameter/index_en.rst @@ -0,0 +1,11 @@ +.. _cmd_line_index: + +Set Command-line Parameters +=========================== + +.. toctree:: + :maxdepth: 1 + + use_case_en.md + arguments_en.md + detail_introduction_en.md diff --git a/doc/howto/usage/cluster/k8s-aws/README.md b/doc/howto/usage/k8s/k8s_aws_en.md similarity index 99% rename from doc/howto/usage/cluster/k8s-aws/README.md rename to doc/howto/usage/k8s/k8s_aws_en.md index 593158428803c..201bcae48df29 100644 --- a/doc/howto/usage/cluster/k8s-aws/README.md +++ b/doc/howto/usage/k8s/k8s_aws_en.md @@ -1,4 +1,4 @@ -# PaddlePaddle on AWS with Kubernetes +# Kubernetes on AWS ## Create AWS Account and IAM Account @@ -331,15 +331,15 @@ For sharing the training data across all the Kubernetes nodes, we use EFS (Elast 1. Make sure you added AmazonElasticFileSystemFullAccess policy in your group. 1. Create the Elastic File System in AWS console, and attach the new VPC with it. - + 1. Modify the Kubernetes security group under ec2/Security Groups, add additional inbound policy "All TCP TCP 0 - 65535 0.0.0.0/0" for Kubernetes default VPC security group. - + 1. Follow the EC2 mount instruction to mount the disk onto all the Kubernetes nodes, we recommend to mount EFS disk onto ~/efs. - + Before starting the training, you should place your user config and divided training data onto EFS. When the training start, each task will copy related files from EFS into container, and it will also write the training results back onto EFS, we will show you how to place the data later in this article. diff --git a/doc/howto/usage/cluster/k8s/k8s_cn.md b/doc/howto/usage/k8s/k8s_cn.md similarity index 99% rename from doc/howto/usage/cluster/k8s/k8s_cn.md rename to doc/howto/usage/k8s/k8s_cn.md index 2575701053ca1..ab07cb9cd5b13 100644 --- a/doc/howto/usage/cluster/k8s/k8s_cn.md +++ b/doc/howto/usage/k8s/k8s_cn.md @@ -1,4 +1,4 @@ -# Kubernetes 单机训练 +# Kubernetes单机训练 在这篇文档里，我们介绍如何在 Kubernetes 集群上启动一个单机使用CPU的Paddle训练作业。在下一篇中，我们将介绍如何启动分布式训练作业。 diff --git a/doc/howto/usage/cluster/k8s/k8s_distributed_cn.md b/doc/howto/usage/k8s/k8s_distributed_cn.md similarity index 99% rename from doc/howto/usage/cluster/k8s/k8s_distributed_cn.md rename to doc/howto/usage/k8s/k8s_distributed_cn.md index 53d0b4676c6a3..b63b8437a0114 100644 --- a/doc/howto/usage/cluster/k8s/k8s_distributed_cn.md +++ b/doc/howto/usage/k8s/k8s_distributed_cn.md @@ -1,4 +1,4 @@ -# Kubernetes 分布式训练 +# Kubernetes分布式训练 前一篇文章介绍了如何在Kubernetes集群上启动一个单机PaddlePaddle训练作业 (Job)。在这篇文章里，我们介绍如何在Kubernetes集群上进行分布式PaddlePaddle训练作业。关于PaddlePaddle的分布式训练，文章 [Cluster Training](https://github.com/baidu/Paddle/blob/develop/doc/cluster/opensource/cluster_train.md)介绍了一种通过SSH远程分发任务，进行分布式训练的方法，与此不同的是，本文将介绍在Kubernetes容器管理平台上快速构建PaddlePaddle容器集群，进行分布式训练的方案。 @@ -22,7 +22,7 @@ 首先，我们需要拥有一个Kubernetes集群，在这个集群中所有node与pod都可以互相通信。关于Kubernetes集群搭建，可以参考[官方文档](http://kubernetes.io/docs/getting-started-guides/kubeadm/)，在以后的文章中我们也会介绍AWS上搭建的方案。本文假设大家能找到几台物理机，并且可以按照官方文档在上面部署Kubernetes。在本文的环境中，Kubernetes集群中所有node都挂载了一个[MFS](http://moosefs.org/)（Moose filesystem，一种分布式文件系统）共享目录，我们通过这个目录来存放训练文件与最终输出的模型。关于MFS的安装部署，可以参考[MooseFS documentation](https://moosefs.com/documentation.html)。在训练之前，用户将配置与训练数据切分好放在MFS目录中，训练时，程序从此目录拷贝文件到容器内进行训练，将结果保存到此目录里。整体的结构图如下： -![paddle on kubernetes结构图](k8s-paddle-arch.png) +![paddle on kubernetes结构图](src/k8s-paddle-arch.png) 上图描述了一个3节点的分布式训练场景，Kubernetes集群的每个node上都挂载了一个MFS目录，这个目录可以通过volume的形式挂载到容器中。Kubernetes为这次训练创建了3个pod并且调度到了3个node上运行，每个pod包含一个PaddlePaddle容器。在容器创建后，会启动pserver与trainer进程，读取volume中的数据进行这次分布式训练。 diff --git a/doc/howto/usage/cluster/k8s/k8s_en.md b/doc/howto/usage/k8s/k8s_en.md similarity index 100% rename from doc/howto/usage/cluster/k8s/k8s_en.md rename to doc/howto/usage/k8s/k8s_en.md diff --git a/doc/howto/usage/cluster/k8s/Dockerfile b/doc/howto/usage/k8s/src/Dockerfile similarity index 100% rename from doc/howto/usage/cluster/k8s/Dockerfile rename to doc/howto/usage/k8s/src/Dockerfile diff --git a/doc/howto/usage/cluster/k8s-aws/add_security_group.png b/doc/howto/usage/k8s/src/add_security_group.png similarity index 100% rename from doc/howto/usage/cluster/k8s-aws/add_security_group.png rename to doc/howto/usage/k8s/src/add_security_group.png diff --git a/doc/howto/usage/cluster/k8s-aws/create_efs.png b/doc/howto/usage/k8s/src/create_efs.png similarity index 100% rename from doc/howto/usage/cluster/k8s-aws/create_efs.png rename to doc/howto/usage/k8s/src/create_efs.png diff --git a/doc/howto/usage/cluster/k8s-aws/efs_mount.png b/doc/howto/usage/k8s/src/efs_mount.png similarity index 100% rename from doc/howto/usage/cluster/k8s-aws/efs_mount.png rename to doc/howto/usage/k8s/src/efs_mount.png diff --git a/doc/howto/usage/cluster/k8s/job.yaml b/doc/howto/usage/k8s/src/job.yaml similarity index 100% rename from doc/howto/usage/cluster/k8s/job.yaml rename to doc/howto/usage/k8s/src/job.yaml diff --git a/doc/howto/usage/cluster/k8s/k8s-paddle-arch.png b/doc/howto/usage/k8s/src/k8s-paddle-arch.png similarity index 100% rename from doc/howto/usage/cluster/k8s/k8s-paddle-arch.png rename to doc/howto/usage/k8s/src/k8s-paddle-arch.png diff --git a/doc/howto/usage/cluster/k8s-aws/managed_policy.png b/doc/howto/usage/k8s/src/managed_policy.png similarity index 100% rename from doc/howto/usage/cluster/k8s-aws/managed_policy.png rename to doc/howto/usage/k8s/src/managed_policy.png diff --git a/doc/howto/usage/cluster/k8s/start.sh b/doc/howto/usage/k8s/src/start.sh similarity index 100% rename from doc/howto/usage/cluster/k8s/start.sh rename to doc/howto/usage/k8s/src/start.sh diff --git a/doc/howto/usage/cluster/k8s/start_paddle.py b/doc/howto/usage/k8s/src/start_paddle.py similarity index 100% rename from doc/howto/usage/cluster/k8s/start_paddle.py rename to doc/howto/usage/k8s/src/start_paddle.py diff --git a/doc/tutorials/index_cn.md b/doc/tutorials/index_cn.md index 97014d537655d..6a27004d58d24 100644 --- a/doc/tutorials/index_cn.md +++ b/doc/tutorials/index_cn.md @@ -2,6 +2,7 @@ * [快速入门](quick_start/index_cn.rst) * [个性化推荐](rec/ml_regression_cn.rst) +* [图像分类](image_classification/index_cn.md) * [情感分析](sentiment_analysis/index_cn.md) * [语义角色标注](semantic_role_labeling/index_cn.md) * [机器翻译](text_generation/index_cn.md) @@ -9,3 +10,4 @@ ## 常用模型 * [ResNet模型](imagenet_model/resnet_model_cn.md) +* [词向量模型](embedding_model/index_cn.md) diff --git a/doc/tutorials/index_en.md b/doc/tutorials/index_en.md index cce9d3a176a5e..77331a703b6f0 100644 --- a/doc/tutorials/index_en.md +++ b/doc/tutorials/index_en.md @@ -7,6 +7,7 @@ There are several examples and demos here. * [Sentiment Analysis](sentiment_analysis/index_en.md) * [Semantic Role Labeling](semantic_role_labeling/index_en.md) * [Text Generation](text_generation/index_en.md) +* [Image Auto-Generation](gan/index_en.md) ## Model Zoo * [ImageNet: ResNet](imagenet_model/resnet_model_en.md) From e0a85db7f5609da5b881707505144de99613e49c Mon Sep 17 00:00:00 2001 From: Luo Tao Date: Wed, 4 Jan 2017 17:16:28 +0800 Subject: [PATCH 2/2] fix dead links, reduce image size --- doc/howto/deep_model/rnn/rnn_config_cn.rst | 12 +++--------- doc/howto/usage/k8s/k8s_aws_en.md | 6 +++--- doc/tutorials/gan/gan.png | Bin 33275 -> 17810 bytes doc/tutorials/gan/index_en.md | 12 +++--------- doc/tutorials/gan/uniform_sample.png | Bin 20609 -> 24880 bytes 5 files changed, 9 insertions(+), 21 deletions(-) diff --git a/doc/howto/deep_model/rnn/rnn_config_cn.rst b/doc/howto/deep_model/rnn/rnn_config_cn.rst index 8d65b3512d0d9..ac2bd0775f4ab 100644 --- a/doc/howto/deep_model/rnn/rnn_config_cn.rst +++ b/doc/howto/deep_model/rnn/rnn_config_cn.rst @@ -33,8 +33,7 @@ PaddlePaddle yield src_ids, trg_ids, trg_ids_next -有关如何编写数据提供程序的更多细节描述，请参考 -`PyDataProvider2 <../../ui/data_provider/index.html>`__\ 。完整的数据提供文件在 +有关如何编写数据提供程序的更多细节描述，请参考 :ref:`api_pydataprovider2` 。完整的数据提供文件在 ``demo/seqToseq/dataprovider.py``\ 。 配置循环神经网络架构 @@ -132,9 +131,7 @@ Sequence to Sequence Model with Attention 模型的编码器部分如下所示。它叫做\ ``grumemory``\ 来表示门控循环神经网络。如果网络架构简单，那么推荐使用循环神经网络的方法，因为它比 ``recurrent_group`` -更快。我们已经实现了大多数常用的循环神经网络架构，可以参考 -`Layers <../../ui/api/trainer_config_helpers/layers_index.html>`__ -了解更多细节。 +更快。我们已经实现了大多数常用的循环神经网络架构，可以参考 :ref:`api_trainer_config_helpers_layers` 了解更多细节。 我们还将编码向量投射到 ``decoder_size`` 维空间。这通过获得反向循环网络的第一个实例，并将其投射到 @@ -276,9 +273,6 @@ attention，门控循环单元单步函数和输出函数： result_file=gen_trans_file) outputs(beam_gen) -注意，这种生成技术只用于类似解码器的生成过程。如果你正在处理序列标记任务，请参阅 -`Semantic Role Labeling -Demo <../../demo/semantic_role_labeling/index.html>`__ -了解更多详细信息。 +注意，这种生成技术只用于类似解码器的生成过程。如果你正在处理序列标记任务，请参阅 :ref:`semantic_role_labeling` 了解更多详细信息。 完整的配置文件在\ ``demo/seqToseq/seqToseq_net.py``\ 。 diff --git a/doc/howto/usage/k8s/k8s_aws_en.md b/doc/howto/usage/k8s/k8s_aws_en.md index 201bcae48df29..422dc3bd811ae 100644 --- a/doc/howto/usage/k8s/k8s_aws_en.md +++ b/doc/howto/usage/k8s/k8s_aws_en.md @@ -331,15 +331,15 @@ For sharing the training data across all the Kubernetes nodes, we use EFS (Elast 1. Make sure you added AmazonElasticFileSystemFullAccess policy in your group. 1. Create the Elastic File System in AWS console, and attach the new VPC with it. - +
![](src/create_efs.png)
1. Modify the Kubernetes security group under ec2/Security Groups, add additional inbound policy "All TCP TCP 0 - 65535 0.0.0.0/0" for Kubernetes default VPC security group. - +
![](src/add_security_group.png)
1. Follow the EC2 mount instruction to mount the disk onto all the Kubernetes nodes, we recommend to mount EFS disk onto ~/efs. - +
![](src/efs_mount.png)
Before starting the training, you should place your user config and divided training data onto EFS. When the training start, each task will copy related files from EFS into container, and it will also write the training results back onto EFS, we will show you how to place the data later in this article. diff --git a/doc/tutorials/gan/gan.png b/doc/tutorials/gan/gan.png index 001ed6cc19e8911f9b10f63211c9658160b3a06e..0eafd7cb49b545f412f8e775804bcd0b22c42454 100644 GIT binary patch literal 17810 zcmeIaWl&sC^o9w9B!eWl1$TD~?!hHE!9BQJ2qAcIcPB`Ypuyb<8rk4`2Ae#=Rc{JH_ zi=jt*e9RIF88nfqGMRb0cx>3KzvCa~84lm!)hp3#5L$iPd2|4(0< zP;&J1&sCL!9$nqHEr|Hb-vnm;L(7RD73h+CQT-IhcSo&_RPrfx7t{JeC#Ja^7~hpe zprS)%D;b}5X|Hbg%I0N< z-&mErKdjEltEqL;@+;(>)~YRU8H_t;=EUrxxkuMkXrp8zDBhKnifvXyzQZ?6_Ueh**3 z(f>@Er}d3n!v-A|txp~rjD>COLr(Q`^gF~SD zq*0XLY^62^V^BAQGPk^?-qVL@Ad8ekjiE%Lq|U5^OEQz!l?Ohn#Huo z*%0^8hvtnKSppUQQ~6wuvxA7j7U&IIjt=kZ%e$TClvaDbk6e)E81mqKwdi)xTKj83 z|M3*ZZQiCf+0kT}OeT`QVOrNMo!sGB>EyK~BSr4hiXY0tq=x>%m&Y4(r^Xtq>EX#D zwE;%0%C4bwzG;`;iQm9PhCzfnU@+uJV+**K#bwx+50iH3p?4jyX=p`oQ8pE%^Q9_u`df^QQ|oSxnP~MmM`EIxF-pRPTghT@*aVNoTAMnr1Fzs21|9aRd? zz38l>O*HUwCwDH;$m?b^afprLO%21IeWI>w#f4a=*QJ)zI2-ovvxt3um%S+$LOxe# zAW^LkE=Nm?E{6*|>vQq-5KS3&oTHFk3h1?KQK9NooQ%Inv{CX8Gd;x{h^mm<$O&a~ z#GsP1-%fYSsMqH~v7NIY#M00)R5yN$vz{Pzwl$=pzSS2^uKV=(zzFlr*sl3Po!g`j z^=hxI)vP`~-kX{2yq?95&SkM&<)K(Z~^G~ichLgxv%!=c~PC=&5-Q|JGq^L#9Ne_O_M=lvQ2m&gu zo4Ci?yG8f)a9M&*OiUr>@yDn_m4dlA42ItVmouIDS-D8ly+fhZ4PXlRgitsVk2Q7O zMP{t4h8Do$&@fxf<#0NzDHH|6b|Xik5pO2Q{9~Y>?1PbSK;)T8t zhvib0se-X9N&%y+7rpogo(ZH9fDPz;@x@Dd1c`tA!{ua{n8#} zdC8P3yx~w$5jFrqQ)BuI0*To`fo&BHZQLy?W>eQ3t-?bq1e)^|qMFtQRrrRzQY#fM zw(wZB`~68LcvywO5ehWV79<-(Tg2oLsof~)Tn=`$u`kf^)s%)??~Qb0ZaVgb zMl(vf5jRp+*ia2(&;5#oG!p&q{$gz?XEM@t_I;#56dBxfdC5o3nin`wIiaKY^HZC9 z8z-Z*gMMo&k0U=>`%0|(pVFjJMMd5iRUxR#u%~jkF#V_hzEmW6T7POa!ZH3$Z66gZ z?U41@L_Jjhh2arXzQ&n}EbDP9{0ki{cr&&jv{eU>$P z)Yu@Ee7fwvH)mx5hk$*4ptkJoVd~AvKC9z6mGji*ICm<%rZs6o35cWp7Dd)$@T^Hs*)==c9Z*Zypk zZYVaRPS9(@1-sUVEh@!~eZjkZ_9`kKSD#zCt4enZR zw(z0;s&U!-X>szW_VfXeC(7OPzuzMp)PH`MT^?5yve`bGD2#==fI_bvr|n_e`|A^l zd3_j)e+xT@qTG76vXA~9a?lT8B`$^Tbsf5~dQ7HDwBGMc{gmBr+{sT>ZnABCaGiGp zw@e((c&TnWRB7A=6dqiyCr<9XB^ zT3P1*4(Q6ofyTN-Rvkjt%#^O%VxP9hXr$01u1E0zrXx~7rEOPjW1i*^7_U`fB(cHxe!j;A8aEqXEFU!<3DP?~ zTujfq*`>cDmy8G#yg%-QNQ5pp4lue7*E?;$TY?M!bh&R~`o2F7kKeuAEAy>`(Z4-)Vu~6y#&BpJ z_<fG%w7OQUo!#V7(_%8g7scC_Hs`}c@b0qBCh=;r7CtFh`Ux%IP zvP>KA^>A)$_Dt?2cVkZnDY9-`3p&&v?gD3)^$<(}pM3e$2B=r|y4@+@(%ZobX4L|r zalf_paKF`UvSEm8e$Zoc?Q+hF8UM^=AKkD7^#rjFegv48#mN$_@fru)4dCC(ova&n z@?Bdpzuf-r5ku=yj(+y~?TX@6z{wz~Sm4ZC%JpQBA22u|8)tm0MOEn?2Ch@kuH|Nn za(KExk@)Tr+&1|&?1A;asZ#HE8s?78fhz7|fxA^gwyjY;-7S&#D8G6o5lsAK+>}K@ zx|y7BC3)ZWjB%ukB%w>(8FoJ3QPY`DnO<|rAaT#!)8oq|L%wA{(Q{ihH!yg!9j8IV ztn0EPl3@8#HmGtaeTDzy7KX-VcD-nf;caD#SY_DGf0Z@rlY4hVnSfB{@fV2I>TYCvkS`Qg>{g6q5u;d+%||MZsYC;IWpVvP|G+1ci&$Gd!I zJXurxUZ>%@7|CRr7&MRpo204AfYEA$pz5q|%g^CbwhM&qWeYz{pd&TZ@0B*DWj22N z3C(dB>O~-RA_zpze?-i9KH(HnK5t7AcX-|&@&GKlH8LAynj*Rbq^yQAp=X9SKkK(r z^7*Odk_^_<@oj#+N0_swmI`)Ig!94mezaKM~=ra_dco>z#qA~IDSea`=UGzGOih>VTq86G|9j{UT+ueOd_pEB-O21URU*EuX)4~*%c4Z!CuEMbRSC(xQj5( zg#YbM4~o$JA77K#g4Zh>FAGhoG3wSr=og|$#n`6AWZ22cuz<-%_1S=L0r?tI!m477 z5d72!E5=&jIKr?LIP!m|rR|It;CNAe3*(?Kgc)_c*qe5`X++@4x$&^UDtV~? z2M2TGK3i|rC9oaoG?hQ1V76qmuVF!2R(rUasP78)MQY~_{9ZZcS@OJ(EG}*6_X0va zB?;VC>U8!o4=l$(eAd&Ki%AXJ;WVB@Xo9P4T_KTB7NuR#xE{f`&)opTZ(F_FF_#PM zk2uto1NXE~P&oQjkC@a*CTKVXr?!2$=%#;F+xjpBqPvP0N6?{s6W#eTG{|&9@|64ZVnLi8=d>8h z6?J-YJg-rcsfI*#+zD3(>BQpO*0tNz%$h}P%Os1wX&!^dr0C2j?I~NEvUhMT;}T1i zMLlmedh$FkcCiG2%c*~lH%Z3o!&tk(b}GZ+aPyh|uRbD~oxExD+}ZX>BuJ!7{xW;t z$o>KuvE-LW^9|j$Xm7M&J?{sX#w9R(I5+j#qk;IgxHsa6(V3o_IYj)^53+6>x_xL@ z?dI3P_x#;|{ZIsH-rFuUd5qR?XW(<%E%Bdb`1}RoJL1d3g$Dh4?*p085yvOse!hcG z=GOP)`SRa~xR?3g0C#AEbPBZyf^~GK!B}B_-qKtzX5-Mv-)T1X$0|Sb>0{O` z(-Ww&q^Ot7G9SsX-Jhwb*QlT?MT_kl|{_XxWW-S@mNa^{@IrhoYQ?6v*0p65?W znYKpX)H851R9373uh^R}2aNRV!-`&_s!DB}>Ua;tr<)P)e9I@O@rT9~LXhs-7NomY zTCi&NP#7E|PWPkT?8T?A3xd6tp+EtUn`|ra9dOJma*@Q;puFMetF~KS^4PReuhxj+ z;#k9~df2Y5U;4CHuc9b0`HEO@p=LE-(sHuMRE(Aaqs@=UX-jS!EARUkO^=BKXnebb zCw;!!`t)Fj?{pqAL+HNZ=>;-w5y*0l2DoE38u#HeNWv3!MFK5kdVk3=%?@#qd>C2j8==>(7@k$mYLa`D(R6FZq@CI~CpzCJas5F%o8P_Gd-U8|b8!M7t659OV~v(C;eGk_q?38y&S`neJf4 z6QV~7zfQJ-Oo->pC9i#`vEGq-JwowdF`i$a?zS{*MB3C*!KT;#A!-b$Ze|gjIwg9k z0}v3+a_sd#;2aauso{y&7FQeoLXH|-aK{wTL1AE-(|YEQ z2NG)l{X=~OnM8+cPTB)*LB0Fcj4=w|2*`LX>HX|&&YvO1AM7A0V_*$@$?roe!HskB zK+;(M^tfL&Qo^{?Y7ysVzw@G5+}X-SW-1mR<`E>oHK{Z=#IQ?`N!VydO>0%B@amK? zskJ&pOiiz3#Yu=ES-71YBVp`pSWy*1`8lM9MOf>Ra(_^~2f;4u0JF_!mRR`icl=1} zftv9&9K+r2)YGCK#-?O6J9TjHqANp?Mhi5X3GjQQ$FW@QuR=ATS)SWK0^?omSnlhv zcruaf7nGTW0`b*Cdomb#Z6p`ZQ23R!T!shEPA}hm=>%#{&%CEE6@O{WbeV8JCd;f< zRmTl4_ZdXp_$fJGu6x`Gy39=O2MmR`?KiCHYJ!OhQ9|(Uqm2)j^NcJn1I|I+fk$X7 z@r@`_$o9k4a_zg-8f0(M87dVkKjqW+KW#hw5vF?Wy93X%2~&jWW04nytQqmm!`CP4gvJWuXlB|; zLcgzLD6Sv|(rXF1#H=mfE+o3FK~Aj_Z)!oG`~Leo?DRo?DF(Tc)^Y=me(ptOq~!{N zXe8#QKaa-MF`@A-Jt2%ViWz%hl4tzu%uJ+X@jtYQRZV&WD*jw^f^Bf_D`2MQcoPPB zBoQiqaQgZdM6&CnPqVD~W$kfo_Y6sLBBbP|_32Is zDWzRTTq)~Igb9MqJrQmhvT6A~P`OBS+_IenrvMnYGfw~Y%2oE8_quqkZex5MdVg@&-@BHwOK(@|8q$qF^*KAP{?i2`NDh<;1zm49Fa2>0t zzcD09+dIASMPe!!xgZhuMKkQdCrM9s2$c(o|FlvAu`{2X^(oq$`vO-PUjIju9YM(C z5UKsex#CjPj^sJtsoz>)4l$egkRoF3U$HMU=R$E16r@{|L?&Hj6zrwzoAf`4f3G%* z7QjZ${z0-}kXqY4_bt*@>irM_hR(0Lz!To%D1%|fOl?pOy1S9DXdyO!Z!`DHp(!~C zSt$%{b=Pz-*lMLY%kKTlUBA=sM+YG9C*Sp)1=;NMyl#E<<=HuNMzTl}hRiaZ ztk7}Ob}eQP`HmCKRvjvbcpK_^vO_zgY&wr1(?slEekOIQ*T+0|xSJq}jb#Qh{dUKK zwkXbyWvNtIr-}EHT7d{fG6$)KWxhYSDPF3;j>gUWe;aJo%O0cZY?6?b^VztP=wR+TK zhGFUzN`l-z)N7mFDYa>=3K`S-JE0+wRnm5s>(M~19M0#+f0;iR`L z9*;B#8qJjRc9u;JUScV~bqHaRp1hk-F*h;#S&OZ*+)1JHv}HH{*CZ`y!ai3OG1i4z zAq_V>$IK33+R<>?w|SR*C1oY{0MfL1}FXuA1x%=~whtkvS|)_W@29BC1-?&@R2# z&oG2)ha)bg8mn>MWi#b1636j)r3ddX8ASxXCQRAk+M4{L*HW48JJF~e&dO|w`*JDO za745w{>nE-W5qG6QaWx+W+aN*0Oks-l;p{E_h^yW7bRUdKqy1N8^Q#q^E~MmnJTES zh@~CN4Hq5V_9c#w_c~`LSE;_vONxlzp?C0x4f;gf7^N^eGIcG-y#fZYKx4d~=H)}rarymmN@`4z@i@U04KVa!Gwr3=DxI#Z9hokxCEbe?4)dn@*mHxw?kfUh(2Z0oFdHKwsUL6& zk>YTV&pR`Mm+;NRkn8J|rMTwPW4Gz#>!~X^Je+#vV>4a{`?t4b7jkIj@{pa_y3?Pc zWW~u&OU}-1`XTkvN21&PqQKwMxnHe3eb%-`)X}VxvQl}5@{IoIWg)qcLM+`DU*)8q zS9^T7c?|K(6BKMGmG0+PKZ)H!?vq3;*%OF*Z!7SCA?BZ{*73+a-{At z3PX{9G@?BoXRgnCE~r{B1G6d*Pp9&Hv`#XML(iSoASJ^#lG~}~r7T8q_>mxLyMmYy zGLu}*`m5u0=bDko8!bo6vM||Z35vEJfn=t3EaIAVo7JV>=b@o{-QF+V^=7b3Sm%bW z3@lE~23fumm&v6gBH>uQsb%VLTevxQiRez36u2Q+`1bu>c;(TT&w?&Q%8M1@@QV%z zOh*Q*EKO_Ac+d+t*01oDzef+h+N+O(ACahG91>lN9k7u$vM6;vXYi|NQbW^0OFWk@ z(aeCqMt!xp_!CXQ(95y6)SuV|b)8^4@{IqLRYA+^qyFEi~-b5Bu#m4fmLhdKLzXheqxti2lL2blv z)@%V8L8>-8*(b>}Mx3KT<>lpp*W`7K@Ur%cr^CqXam1>)1)l>7^8(VF7?WOC0)luB4)ih0xeG8KGiMa*gFjFHQ*r!u^5uEnC1sL^0r&! z5==wn&@;9k%$|=jp5G;Yvr+ZoB4nQsh$Zo7kA?BDE=d!*cS=;S3-;fvO_91N)biu8 zi`Ms);msvHW*NwZROrq9KIP*#!SfS-n{npPO2hkF_KIaX*rRA7^@QqFAm=aTENR;> z#GaKplm)kZU4Pc=^R(2}kT452oR~z$TlA&Z{uRP{O-gUU7G6GMWEUoCi87Tl`FZA8 zN!Yod_6L0=W{13sBcDo_`0LD9Gd<*4+A?v8js5SNRPsVOr(qx`=ON z@%W4pX!5u184zP>Il}JPv0fk`<;psk6^vk?Y9+ntRE_@wD5r`%U{IY$^^(Sk1oV<}&Im2npwyS|P0XK20h9VEi zp8c#ki9A_?6GD?hkRAxew+h#!@t$Yl2yGLo4^95618*E|7D5LK`7nwra6OKlmJtkA zTgn;Fh4=|HKV<8l7)hYwIMLK!4LWqqArrxDt>(D%yzCE@-e@2AiiJ~|)m0quy!S#> zfAuPc=*PRvtfwx<;d30>TY^nd;tlo(D#6-B~+^=t1;I^*(zbQSGq9I*qI0-^Xc+nsB68FM@xU|v0!ysANj;i3LNbt$j&@*-hu}-o0 zA7XR;oqNCzvvaedbr=S~EV#GtZt46A2x3N=s59DPOy@Ozf7T`*{9>T7BqM9%#qjO( zD=KNnXK&H7gDZ`Z!?T|UVu2HISpHij9(}4D5b#caMbDAhmrv-6Q2C3UGK!^@x5J_6 z2%T6Il;4oB=@YD^I0H3cBgOe8sUNV9;2-F`e)w8t{leAAI~pY}r0-RF*@P_;ox@Sc z*Gd~9_Jetf-3zal9x;{TSrg3*Zmvz-BZ^{V(>1i+U-3BdU)E1d^WFw-5Yc=>CyuL4 zkyy#zcoBaWM%|9|kl51UwZ((0KZr*j^eFk1m1*=Bo-}}Pe&@`c(A~_b6>nt;+QFyZ3wN(~NK^{lSDxZ|xrt`i(3PMiNwu&bu&c%*8 zgA=oaN=jrVZU3spDV#(cPbEWWOXV!ngl&?tm_Q))pv`~R-RVb&Rba)BqL8FkgtyK+KSt%f@k4G?29qnd}@)8p6$pav&+)Gqi}!wO#`8Vpl&<3 z(cnkenInS(j{8m$nva3Q+V9Tq@9>&vG*dMjb-r_CGwev6l-MKpu1grd5c#Z0KYg88 z4{=eW{i^er)zRex4~E1N2e7pS#{2ySj2H6vEJI#%kW$7E6bOqEm63u zLCKO2!E$Ht8VCC~4jk(YVR^ErkuVS33LUm$?R_=x0_o82gUo5x z)9UH+=;Rrw;iyGZ8g+htu9yntu64JpHQErq>(p5?vi26NO)lgmH8l{dCCAwtM%S+U z;81T=AN?iruk8kuIsBO;-o-h(Ql~EcG9{yhWieJQMT0($mN8p4P1b{+>v@2z0lKYL zz9?&J##hcUwZOg;opqh9IE>7#elodDxPhPA?{b-h5OXF%bg#1w9f~5i6ctHE;be1J zz15X@;N7#kUJcApo;lz}QuL>Mv-F2(l?1Q$ckM0t_^-a@L==h87|7A!=f1&T(CSND}qjU-^`z`;rSwBhXViESj zG{fBcrDl2!3uVKV==T99@=Lggt3Ile+(v4plH0F61n|uN;NNCi#K= zG*X!C$ZyE0HJjS8t8yhTG(mDythy{ z4RB!u)hHoI`96L-0lzcFjl@^OXAG8=DQeZIXtFu4;p`y&t9hZ~@FG$~VqNsi2^0>b zut=~xnx9FF5(??yH{nS?4vhTdOtjFv4R=stkgj}hh{g7evL}x@N zH1R2lD&C4kAi)341fi*d5~{Q$M5)Nq5FBj#Jdb+)B5JH;w=rigSQrkjZuWj{cndCe z0|(a8#&|KHY5Yd*F22}CDW%SA9no!bSlvr9d!sXLveL3brrx}`%%Y(ZTHD;9f?erkpbA;N~n3t>aOC?dXc%Q5nFp#vjKIMxRt(aEf7l%&x`8c4NG`z3vH019LNQ1O+$qZpjSrdLIF%)B( zqK8|Uk>x&+DZPT{$k%%Hhki9iJSIl<>&+y5YgZy}iT{AwJZ*atV z`TxtWHYq{MzmQpBLkoccV-(e5@Q%=f4llh?lY@^zGl12{zBEl2hhmwVfE_H+Y2j^z zsPg(m#f0mmNN*;aXH>YLNK;b*%1a7V>DhVc<1u~GNXtKtl5vA-Sg|6Yg!jEa392|+ zeF6CL>nodva^x{k@Iu5b5ysa61f?lmTAN8}36e}m-`HS5V_pL`r;mo7NB3OoylrFr%AX2j z$vYzb`$X0e&(*Vk32|pYF@GV0#1Tg<(t9Xas(8A3kpy3W<6$LBUcG^W9B=~F>-ty%G z?C4)fIV>H@VlgN(rle7)i%*s~1Mn(7LK6qoXCFNY0KUN(Mx12rKR*GsuW>}^X~EZX z$?Hm2GMAxU*g*ugtY*%VGH$ia=h@;fi}8iQWcHzlQ1nH>^2k<8Rn3#7%fp`rfvJ03 zpr8EJhp-Pu>)lE>jy{8k5daG>2`RtHWeP6lST?%sZN1^PwO&t%rB-BeK%TNts@cB= z)tyDqj$-U10tNJ;wek{vCOtSm3ILsWgIiKn5!HvEjOF7WlHM7rV1#!O1UsKDc1_1c zqq5-NC)6&wrZ%3B$u1tRcIs{qr`sq{bzD=70T$yf(&aY?2AK^Wembv95k8l75@FEh z2LLeemBo;3yr+JzMjzj{`xxY6k*6m@4YcV8iw`Kj4(WUcj0TMZgA;WHjYIaDtFCc4 z=K^J-juEWj;QWH~xV()V`W3?+!tN|6CEI*L45?7e6m;d^)@}J*k1z1)msTtkIg3x` zb1StG`^`2IYe5H(M)>J(wk-fBO6>UnsDb$gR-T9U7hlmOCcIB&ewu$zbVX2@5a`xm zGkrHjobNvY`UFB)aMv5Zo0s4yjJAJC`O7~VXf7I1KH#^qq7CH{j@#9^Ss-OK-Q9FG@5YB}}RUVD%&ZUOKu#Sb^Oi-xt6$>nBRGr~#m!PB(`?Y{{Y|;YHuH zUXK}Vs%5$eLP;@h(|%0tPy7A!WgNw4U>Oa3S1&H=eh0;l4(sMpui$L~(C-Ec!>MYn zlWH0dg#rvRL$}G2#9qJdK=nyx-yC1dS87n<1xbH1kTLbZcqIk?6LB|ivJHT~Y3g#g zTJ~;b#0j(~n5Lf8Lr-Qk4%7|$T$j6pG*+fBq z_SmHM(e$&Z+uh;W_&dO(m3pDfH_danr0E+C_Mc3bXtCcREsK!~QCZB})~yCkIOx&N zab>~?y2xD}F3RRT0QOf?>k(8(H4uQJ%ueHhSRjK=Ld}3A8A-@DHEamj>tZMZ zCeNh9#MuV`glTUSNe1OWpBp~TTp*U*n0vTY(e^eQ zpx$*__=-{?T@X|bzc5_Z4rt@ZrAP74n##>q-!xD$t>H!JKL^096M$@<#>e!)kByYZ zC>`dG14Is8XoWUVb_U^+w9owSpX|Z|3h%3_LW$-CDiD}5ZOkaK7b+=|sXh2RKBbE+ zxv#$}M)3wQC+)BvvcUn4p4K8zlm@wv6b1;*;*a;=7eJ@dAX(?F*lO4F?#^G4PzExt z&}}v!yWsVv)IJeTOd|lAZq4FfwC9A&Dy8gLSsCR8QvMbtL7kQakRcTq zeFX{emuth_^RRZ;ow06)HG|RJmQdkRD6K!q_x)3_FfF`&YEGMN{U)|H!<*13gE#)0 z&Gt?}EEloEW$MGZ=K9q6gJqR04e>XVsL(OV=hrjWdtY?p)>J6VXpL4H?jm3xfs(30 zsQI14hr2NPw~#%*%B=!%_){NnFA6SDPQHg9gPb4CPs?ZU&;G2Q`tXk;0HMc z9fCkyn~E^dVb_%|MIN+&U5bIzBjk61hh6j<(JfYCtz zhJa_OT(xr;M3o1O9-U+1sJL4MCNl|(W!$LC|og< z-M+CrS)1(Ly10Dt@!I%{bXIAGP7rXkl;L!$p!9a(HZjM?D4X6lW)1zI@;ung514f) zR`Lnry=0ygBy0vv9u{61_#&vtV0BF8i5sXvP$!%~rxmnjKEhB9xSRPK3(9={_er2j zC|RNqouCVM4cyBc!~AD*D&-EOF1k9J-vIF5>I$WBZC;0}6lh0=$)Zb5w0L=jwlyDj zAT7?CXWH}v#qN6yGKSL6R{%L<-I8c)n0;_!~?U$L`? zFZgnAu_%2`v9ry>P))^z6_o@Xr0Oj8CX3Bglu+RuMUPQ3XK{#AyOb%93xJ&{9A7co zd=n>eucOHAer$lhC(d)ay>0opvt%$EXuQVffZ=45V}n2xg4CaW+x`7(=xKiOsE3Gf zf|=d|m6|gAqa=*!H!O{mj$BOjk4j0OBP9n;c?6PzSbOU_HgFfJ95!(ls%bPCmc}rt zdgJq1W)N@PNWDbavSq$ub83Bg2%?a{2@8Yg!^Dh#&_g;0=^dyo!$HmV(k`q>^&HBTwje=-OVKndi~@d!|3)N52bseH zXovluz8h_wTQJpb-z6ilUx8N$Ko4!QF}L_6n$Ih)nKmetTMGIad`UrThcW!G46}$u zR5faDNVyqhDBC*Ss{tc5wIR(W#K}VtEje#Joi_j(4pme z77`EwFsOsod^xwO^gd7^;9`Gv9CSoEOn!*esk?@*+|s%=M5xzw>q3*<5RF%TiT1-4 z#MQ|!NUAnyy{8LRsNH4^gEnR2i<*fvm!Ley-L}w^MX6CR1d|W;IUJTJiqc(p zTMlST0Qe+LK=KqZfHl`6gdW|w%?5EVFhLKIqOgq0qA)!Fzt9Z)wt~@nMA#k;gs>Pe zr+^Vg`Suk}6Ht_kmc=`Sl_iH+`{K9R-`a@b&*PL70V~-jl~q)*Xi^e|_)VA-x>Tr+ z8%qAI2jNHPMYek8zri8(!DBb8drAEH&%@*0k;Kd=PAzyxsL?xIK$;YXm!T9ku>&mr z0-kwVwu8n68rlyt=(T=+=InviC;k;n5l~-*Zk?~VP3VfCKykbrSgSFxd*=ZnE1>z( z9^BY1sREv^!7rkBOF=tI4IRQM@J{t#9&WW-K0JbA;)k4j5U`J20VO)|2t;;T;P&tL zO;M-`W&pYofB;^%vOV3W-`0^aArT~Hk>WAx&pnaTlq4;_{3mDjmZtm zP=+SKAR=+^EI(e>qGN)-AEH|ORnh4>2jh7_4m!CFFxJ)qOlw;@5@XKfiQP5L}W*G8k z#sLlx(HK{Agnon^3PxG7MC>uRh%`4 z`y}PBK*3O|8$m1EROtxb>-3VqUO3rQ6T!5X(9IbD22$}*=wHR<9EEo~zN}cSSiPup zf1~Mczbeubnm5Hz$~yajJR2QY{eMOk1`g_@kdTAq9#jv_SBRsbF+|U|J#1O>J z_8iUeJiz500@vdhRt;SlY+z*!4oXpqlSr%pfU8<+3IT&^<_QQ6NwpGt|H}dNfPl)7 zCtc@J3D?&>sL)ucq70a8S8&q@{VY*VN%MZZslglg#g8yeOK*jezz`~}n96maSYTLE z1xCp1yUs>$2{H5fjE#sd0QvfV9RsDAS^`z|~1B3OzVq zfy}$EI}}G-)hY_g@nWh1#7uX&#cL>pA|~EJtfUFrT&fD->A;!uD3A{K-WU*Qd|*4T zFz)3gR@z`{I_?PDaAjRXQhT#k7O9b z3X^7wu)>(aZ-aE6F)+UvSWQXp{M(*d9YCc zHNe($-@3G>a%^fBr!J0`_1}V%`=9??@C}Vb(h6W%42y*Af`erY`6v9o*|?n)txVD= zQViaMti3-XfePFwevQj-@P)*QewE{)ph*ImE2TO!wuoU&b|15!`XRxD(Q3A9h%hhs zZwMC%|2kNqA|~u3j382&^q-6PpP=(!2oVKg;QxOKj4bpaEDHBT{E60A0b}Po5BR4h OAQ@K>FBLWL{a*m;J>miY literal 33275 zcmeEtS5#ANw=TUClrCMVB2_w}2&jliQ9(d@??sRf37t<`q=S?Y5kWzE4?Xm*^csqQ z)C3R$B#?uC|K8_*Uz~BqIOl>v@~)M&)~xR{pE=h%eO*mjDmE$t0s>mCCl6l`5D<3| z5D;T(3Mx~ z=b-XkGvJt#B!(ss8n~HV%R@oLr*=)|(cim1du?_9b0p&!Ze@+VA<`rt^UQ`gAOMcRP3a-?o`(@N()q4!6yp}r6*@sD}o&+RM3cyOt zeSbo9=6jbz|JF2!tVp*XX}tW`X!kGE6?&@oc?Xgn{*c9TWgPmr5iT~%{O8zP5)K8H zIJu{Od@1!>c$uM8GGjU@TtJRH-?wVRy@jJAJzS{RQ%wN$>#0rT*%v( z=?Vc6x>${#>Vc#u19pFP@0?kW6s@$Zj?{-){av_gDQg(jos2b4Nuv6s3r zF?fOfOik;*s^!}l%AFeD+)LO`mFvnW?yM^+z~^MsCJo^N^-IQb+B>ov2sqav@NMeG zeae6{cx9gIqk07wp%!`49dr8Nglw>S&QsXYhje|#uOlc;gb9OHmlDr3<94bP$V`Hh z%b3CtFgpJeMZA(*v9eW+@N4^UPI%fvpaNvw$%D6oWzslVIaR1&*H(|{_ zEF>PcrIehva_RDIzhXvzfOfofkg2M=tPgvxVM8;GZT-@N;P^wyXg5qyc4ovFq zTyQE6f8n<;0?3X31XZ}A4b$(IJAUTtgI?WlU`I2Wn$zS^jf%>jm_~Z9<&&yDmn*&y z=F{Q*EHR0V_3m{u$vf3%u!p%Z!4EDgU^WIvK-Ar{`+H`O{Q=q_gtRD~|0RPkvWhAK za~znQE-29P$9z>=n?|J8c>1*oZU~ku-8jd|jniV`RQBrVW+GJj;;l6k2 z`KunVJY+V1j&NL!Az3f;BFsJKqzRBK#Mu$()kl$QBj)gGBR}Ra_#MRfspN>Y6STYH z`w#7c#};DcM?EC}YD3tRV>YXyO$whNlBvYIr83g>>J`@%&+%*zJ#abR)!O0FSzw?O zQpa`m^L;Nb0y;CY*wgUeM<&>$0B(6cTkNhD`h@cxbZgBtyfYB~co6t+@C)nrphkm{ z>0_E16GJ~2i#H9Q1Cz9H0sN}b+@=Vip#3-YWU_r|^dRj@Uu?~^_@28P|fnm&}#M?|M!;=^7r?gHCEpi?7Fc&2OKm}#v;r57u#?t*~C}EAw9gG#AAEn_j+*mpqrNZrSeqCq(|Uf@SJq5`Vz};e(MD zPI&);mypC+U>2B9={_L8QO*E-WhQ?rSdy)<-*ir7oMs}|My)h0K4+w-@E?n)+m4k? zCF4Pgv4G2!a{z2j62kn9Lk8Zpj_HUhN&lz{yjQza^~5O{{j&1BoHMj8&GbJmk+>B6 zFvBuP7<&rdXE#wP^6%`OSR;}T4lPl#*K)!}B`3SbUS-c_J>~j0T%#Y)+XdX;Q|{Dz z7Hf%$Tk_kdmp|U?4~_lA4=SV8>zriq_K-VH9QP}2S6}$n9sEBzSvp|$cnqaS0|2B;GH~3wbB1Q(+#Ov32+LM ztd*x*4&inX376^ktXAo#_)MI3W{F1Cf_2B|Yyh@jn`&ho-!I7yiCAFn=ys^AIjsJi*`@yc>;n8=CcH@d$&-&~umm4kQi!mL`w-iyj6u zs|7Wn2=$FM1toHLpVzMmIa^>WZRH(5IGZZsOh>|xFl3TR<(x$TOd*O1o7PjG8C&^B z4%WSV2r+DX`^%W1XP0Q`e*G>NY0d`(+7OM9KfBP(nxWpDig*lBD?Im_29Cc0mn3UM zTnpf~L#BEf<8Rb8quGH|E|@(Y_o9<86X)XMDS4j7u-P4c+b|*mw8h03kU+)vKeA+vZ0mnf%{3yq$j{kx`ZVd+7trW0Zal;BRCgF3@GoTQVqBtP z&{l6hUOCvlk)U%8^{6gbxq^9l&-3a%JNeoP9;Jp_{JEHQ4K`b1L)EF795+jo52s^; ze%LGdG%2x5-}qHEq?xIgHca`b32i9hpOk7+vh8Zk-RsTXSj^Piq(8O4k;x63D zK+@xR8N+4@k9(wohS853+@Q8ufBkKqzo@8ljlJMV;~MrwN5}s~O(`!TDF47lzPtSY zp7sA1{vXsw+D^R#=#zQwU(wULV$3O>AKh0OzAiz2oZ9~l(xS>4AU}}Exli(MRK1wY z$XTspNa2jGU1!wsBe-__r=09xJLeNCy~WbLQm6aH+B9@KFN)@bbA_J>cE;tv`oHV9 z$uU~yJRm^XUzHJZFweIJ3W;E;rty>;&1%`2u-TuBP~v+>43*J?Z1(D=dMM&+tJU$6{^z@CiX3rVS(nzufcW`C?M=Z>%h(mU zSucgMGAd_Fjq^?YHI4#*u~CH`R#Vjo{yeM85&a9*l16gICP7R)^NVNU*S=qyml(uD zk<=NkF8 zbCIMsbgP_Zdy;mRk(ofw$KrT?wF+%nM%sL{?Q8ix%oA|cJsT>VKo#K+B9K{oGI$yTF6dJS)MylNqX}gWOtlQQ4Q_cF(i(RK}^BLOs)#R^DD7gx* z$y_GkibnSN#7IhR2P7&(YNhSN`%|Z%S2B(d<{C6M)R+o(<8s%}j^yP}=JEzwL?tk$+Z3e?9ImigwuA5>Ld?gF9_T+x=ci&a+vOt}WK zRRM+nm5c&+E6gEde=?LaJg`{T?q~r=8BnFtf>AUJ5KIFpRbO@*u2yTWWYZPkHei>o z0_q!LCaFjSIC^^Q(Lt~k&thiBz0i-T$P7BQ9M*{v2$p^m%n+O)E@-a<+C>oxvV_lX zJw8yalp0nN#b!Bh-cUVv_X`m^CzUUHH7$#k4cFx?*rQ9v0+#{zitU+y6#SNf={2G+ ziz>p`W*$#NTy6l*mv(`(>+t^2QvXOrZbU9gzNE_b!t8+fgY}1>UuwU5dMf&7SKR+l z4`NnGKS{wv@{};lIUJ_(2-O9bjPwlILVUF*-!6#jQb&B-irtwRri;LvGtF&|btc4gLE&v1ei5gO|I^*9z16qI>xvq{#(_O*&swa#%)q*W(Vvis+UoHg6cJhRP;)jDU4 zzq~!8>s3qGN!XKUBUYe1vqvE}%2=7^Jng?E_+k735eoifozo`P>4=DxA?M`2tFzv+ zEr}9?NrRu@b$n1cTw%A4T}Fl?;yz3^rW&X2c

3TkgP49@aJ{wQn#Z7- zOt5WUc6R5Pr24IvKidR|SGQO+HtrU!pxl_NUr$M`Z*@N&-{2Oy;@4;SAs%r={;*!O zfmv23!c--E-e2u?YvmwfbynF583Qx7#OEKM^pC=!k23_@X~?C?Ud5g=xzpw?M|E0P zJyW0u1>N2{6b~X@zM|eqkN#c#fKF%%(7*oM@hd;t?jjhto&HTglU6mUyWA7}HQhVS*sWDQAC_~u@_Ttz$2$YP; z;+?YYLMZQ$=T_y%!y4>^<94kl-yP;=pd6>U`;+VR>{9tTV-#?|A=q}{;R4q=zcj$q z9{WAsn+sp%7(Q|y|9tRWh1#{EfB04TLi)YmGsBmf*OE8ZTEhupM`4v)7Gt-lpT&rJ z_Hnr)?<8;3iNZp1&zfg$n@(vy%o`he4+8g|ezXi9gf*`Z^+G7Ekgcl?s#5h}Rn{!} zcJ~=uhUS(-b6QiQ&grN6b+%5JBZWY& z4gDQPrxG)rJvJzIi~OWd?czpBJClP5Y1i`9f)$C`#j{D)CEoaLND6R{Z`>Kmukx(+ z*Y;>rTg zW29;08#4#b#|ZGpy-C2+f)e;Vm@bm|smAv!lZ8>?0${lz*`o|WP>v52*{hi`)LMrS zOJ>acSBp!-mW0I(wCP4Pkw>wyGEb~Ji!EXiC)o?oSJA#Qu*sI6`9_HY$s0yS>N%Wm z_bkVV?i;0*gMkL+?8U2@Xi#i~g2TIm@miW^+qus{aKhX)4d|71g_WX>;U+w^i0@xU zv_1!^2XMzMv!>$&N=&i+MR({s?$$1{C}XbgBv(|ydgv2V|KYl1Z<{~J=+pk^hqFsa!8n8r z65(UN(nGBQ#k9qu1iPx`{qOHykvQtnrem3-wJRbX5r1SW51V2^h+{M$YG@nH`qJg~ zJ44?rqENz!&yl<+$Rq=onIHH;+d+L3O!z1=0wj84ZOSpaRc zrzl!{LM#N!r}FCjZt}*0Sj5_DGIq!}24Am!Y@Lji z%cJGh3wI7iWH{2RJPTL2Vm`BbF@5Qi+0p;Qdcc$L>wVJYq@cR*|6mI|?avxRe_ok5 z{*;*Qli2^!Zw?Kmg?NV1ncBIxVT#emGs)OZok&Pqlz31_N(uZsgQ-S5YrK!vz<*4= z_T3ex!EO9LaLcEC&I@twh3bTjee*;_ zXoaJ4;PzpX zzgQT({f+k%?HH(R1aH4sx9ml0@P&Y>oLCovR0+7nOc#|N`{128%!Wy*hUUj(W?bD! ze66%`2(bak37b%n2l<=Ms1|iPG_p~p$`xcSN~&8b>)Hs5UG)u7h(-{~4*Jf|L7*dU z^Dp1P)JS=vu^_tqq-;!;Ut^)JT;@F80R6Z4HCO@cG0%f8RVXg(kst?J`8OsdAqO*0 zUW6aZJh@hmuvf8e#DY{qA6;w`(@LA{c5@J!kCU z-_A>YsgA34?~}7PQo~5P&9Rs@;)0#T@n7n*HH^|8R6VFCexK&2uKC{Y|8n~KDbK=~ zM*b2IDWB5*{-+j*dQrQ-ZA(L=_Q_cIrI{CkuFupGZ zwb|tDGl2A6OjI^;3THYM@xT1te9@x+OM}3mKj*Lhrw5gmI0SoGq#roKq~B~=UHHO! z<-ZQBF|AHo+**jOkUls3=dbsFG9Pu+kpXYR3XQq z)rHA^9-25ZBf}`(gV-P9p7?bpom0v`?Jr5WB(N(P$q!_^E0e%MOukR2$4!u#C_G!G^S8gq+6Dw9 z;iy1nG_eMMo1^EK!A=u5RYIwCl?qpS|LtCW6bUBw+a*QSM|!CKy;5jB@NGV^&pq{T zL_s}Se-+PIwB=%qBMO-yb$b7Y?wdLW7%Ko&u#Hx|J~#i+xp#-l01}lPE#rVx%ew>l zxwKF)5rd~l*5SWwZ*-H3q=+*uv4bFxUW*`;F1BdNdZh&i`IjlNpAV5|WC!pDb*fcd zKYheI`O`%HHxp~9+kf+!I`WK)jM{#TDIYn#857F0|9AU!H1hSMYewdi)6`hZBgC<) z?)L~Pr&AyF;hlg2P7!(G?99i)INcCZu^^_TD~7mlPyD3k=XCPPxp6f9GBG3HQaKZA zU4^R%J@72eDl|ISYr554?;bSrF+>nZweNBzN}h8X?rjg7HyTKC*fRJEC~M=vum1!R zS{&b8RJh3nr#)q-h0un#4k0|C8sDo1+Mc~wt@aZY$)|yigfqM*QtgU`cHR{f0N&qE zU~Z*sMMZ|cT>XVvjSMb044QS9`UK9{`}N6OW1Kp}g6ySmhmYt=ebxB7^{-%NgOjoB zPb78iln%>ebbx50puBHKLY9>1c7OGnca<0oi8uG9k=(P~jS4lLumKh2FX%7id(?|p zQZ&s>&3k|unb_E;>dlsHch$fugjh4;Vsm2rbi@*z$Xx24Ho;OVC#*L97ho#P0fNx# z*c0#y^)0Etc@@BR+NLx|_*8f=Lpd`U_tmT4hp6SJbcq~AkX)E%*d0#+tih)|?PNbM zMX!}tPV={9l?X3WIMe*7CZ_HA+VzIRON|IbQ#g~;6#bx7sD$u(Z7Zm))jZFvt#Nf> zib2*TF0h(czp;hC!+xc5eYKV!eJ1q)VF!(=J^l(M1Gkg-x^}cRJgC5?awb#NoG`$& zv>Vr-G&alX1FvTY?mDS9vKe1UOSf@MF9G;bI)3`}Xc8QjDFJkAb2ug)rIsgGFk&Zk z)AeL=a6IJs9<2Ut1X4HZR-V<58ssh#BYSvT9Jf58{tnG%b(Nh~WnLvb zY4*OVbORhpO;(wLc2{HTroNb=cQX%NTNXXZ4HDPpjiQE6#%(y>OhNtU^BB{scq6&%>qZS= zSR)Ti)M);u0BhX=ZY1STaW2mShmCE%rBX*`aV4XThm8y2o;^!4*yhx)KCks%|~l z@LWw4K3sSPXMK$$M}poT2bO=OoQgw;OWAhQpE?HmYC9=VE?@C)NiOUI+GXj|jg%-+ zbi+-SBGj?!smK)=NBP38pnAHj1b?{C6`W0Jl5Vs~%J0(S+Y$5bsL;&If~{7-E?4-v zrIz2XALabtq!)d~l~ity`Wrg+NF1=P+;cL$^T%qNE7Ve;q>8kv4p?oWdkFU_Ueuuc zNI7JV>cb@w`&G=BF@zSTW3961rY23)hj{CT?PS{}469O=X3645$6mb7PAFkbLq{HW zM-;czj)gnzeskA3st>Z?daSXqzZ^~t*T{>s-hiN2-=H_;U>6!Yh|PL^KH)htWzL;; zt|rvckyqD&{#pR*CCOsrU4xk~_i#pxU)}DM7h2x2HGIJnDsAR-XK$J7YUh1Z6;X=No3gUMJ6-&XkAS7?(jrZ7sY8$g+DPE_8EU?`{KoQzfug{iTt|`lgeuc{{`Vd)50ZJZ@K%*CpUyM>0=GxE9$%hXX;m4VDX(vRxK7{is0y9q3$h!s zKB*UPC!;Yv=iHiu@z5^jDn(}!1CYFHP8>_;q4moI6ie99orO;_ zVZjIgEXuuodV09jMxnTP6MoQL0#C9PdnZy6OmSdNFWe3!ea)HTo2v-vG!>5-L@5#O zwvXH{=UzQhdv76#^%EWHoD)}{i?yBMW~pP{=n851;e9OSH5I4dG{f^tQps<@Pg5Ol z13OxfLqs?$aNG6<>+|+lKFr-E30P$=2{x)mWccauOF@fFwivHn-jTcH#HcMc*`UM4 zFHAxFdXnm{xIhxMyWABJ1Yh5ctryhfuUC@`sCF;Gbw4^ZauD;U;9xy1J1nK{pU#H# zRcL#XkVo-<#mmN(;n&yukKXjGL#A%;Y<8w`Oba&`(UX1*m}Hrt9v#FQgWGQ~t$*BG zfZnP*?eNd3@x zCiN3IVr-71M(a6ideh|$wO#4e)&uC-iA+9S{QB?yW^V8rz9#*?_R82x`gSqu^4I`R zBSkEc(A@pQs`@!^?lM-|d!>jU2nY&+PmA2HNr80psG68w@QBz6zjaSLS_+VSjoq}~X7B|v;Vu?mc7c0&7)pXphY$pfSfK$5Y}BPM& zm5GrRV~@6|kNx-Hr*~{!{68metaMCtvAN~sn}yx$m{F9jZ9QS7h|!pb&0)^{t)1ZZ zax0!`k$J*|`2iF-1#yD|N#41e&W^ti11mj@d;K;dSAI2I11N*D)&^BKt`Ak_7%O?Z zkgMKQKBin!Ikg-|jSWae*jlE>z$5)S4(Ue5!Cb=Jn{aUY;e$l|^Vk~Ln`%8Od7eF! zB|~2&dNyeC3=Fwsz^h(jh<5EhkI*nR)y?|A9~y*`_L6HlyakYx=hN8^R5$ z{^Vvk<{i2xH0V!{y?TeoV@6(v)!C!ZBhQds9>@-tVOD$rix4Kg4 zx}2e%EA|6PLWp7Z)qYC!UX9o9If;`69OSO^40?nVwqI?3J1Yw-$weEiy4PZYo;`kW zxQ87MrH;VYBLk-_0N8yNVW^KH?%ek@R2EMK=FV$hX9g=&EBQI?`MEN8!K8e+P-htR zg~%GM7useNtqA6>olsU>=(D~0Qh^6sK#h$%!Cfl;8>p6-(Ue{sIjHdMSo;t@tvOao zWK65uVQKk;+W~$e70i#6Eshr~|2%5~@13ES)CgQR$u}(d=b5pyC&B@J6;$s2Ei^w& z)#pvEq3_0P`G32RA=4s6RUJ6j`#)ELXKRt{nH{(LI^WMr_*0h2b8h-aRVA)`p&ei0 zDDe>xQE|cMFPI`9W?&SUJzOYWjz#o%B2qM0kglsa0$v)vLVg6fP14}LXFxh#02~`M z^9H@KUU|hm8zNRV&(JncJ~$orP3BtI0%zF1rAibgwyXD|@|@hbp6CiNBzF*Z@dI`! zhrX~eZlRJ@AsvE-b%qu^S2bol?de;2{@zdS>0i%ElKw=)k#nAj$hgOI`NdB2dL-Gv zRQ^Q#tibT`VJ$c;@9Z$*gAp-Z)Zl(T31DQwnDmSDq(Q^0r7+7>PD7jJZ!}w1sa2dJ zq<-3d4}X+zWVFJ-BmFbKyd3OuzGOq?!hS$HeiBv&pSg?`BN`rAVH7Hxb$iR9s^NtM zUmtQ-{!9Wk2*5iFn#WI&<884H4k4`Xx2%2M9;>0>7@wcHhP&D}w8BJ`S`@F0MUx1) z0O@ddta?poFNB&*13zrO;U8IrQO~LD86N%ew4MvZt(Z0snz1oj=UkDuu*kwo^4&FT zsUx|Ni)7J@AOxNte<>>MLT+^y#v(9GtK)PjoX=wvzgZ6p2Utns6zO+ReAjK&_ML&G z2lKv{f$Rp#s>0_Zo%3S_9^hGNWp8kgs6>N6?unilQGh)by8Z><(a*r(X?bzlz0m9P zfF*!GJpEJb@35rxYR{MZPFeZ!{%j3xw4Xn#&5v2At0w#Nf+b2O2;Em2XB} zSrt}~^|EG(KYT`TU3X|SV8$LsGOjc}&`8RO6_lm*lwHZmoyjj0*v#1K4IBT3Mbem% zn@ygBH61m{*bkoXz6Z*SI8ibPAZpjG9?6t^r*)BzsZWHd!y}q>sRcVrRIPRSv%snb z?aVHhWZ$5XimGZ=92WF#n~$>QZwYXd9&^XMU@F)oaZ)DPX(MoETJw$4+K`f+0L3~4 z+q;Q_NI`St0w+-~APM>>VQh0D!HsPcZl;z6 zYyA3BK1B!x|E`ijiEo@~J2Z({jUn!WWdvi%)bT3D0_{(ig%y!*VrDyul_kFY?eSG# zCRJoQ1mQWgifXXhrg8$wOaGz@t|$B9eSpUpRaV*e`VIIOtkPC3xGS-I9*M0g?ual2Gc}dEyUH(hYGcS1k84v&^dOS5zMy zb$J(*xIm&rg$|u@_h7xMUQWJFl7fXGn^^;ROojRYkXM7E!ZTT6DcDzsRe}2o#}%P9 zZo?}b!1)26f5NzlqAtXCOm2ly>NhBJco9oyk#VqNYa_L-f9Sxxj~96M6D!F~t!*4| z+YK5u!31RO_!_i+NieQh@<;hjpdSzlw96NspzPBJ`gHi>H#0XG!bMDSK z@2a1*vL{ejY3Yb&0|T30$+twMCpM5LTA zeDvg1MrUa;5xXkUcrfFsIL|RE@G*nYLeaP^%a)h9BTpw2CY~c-(zP*EYQ?~cQlCZ$ z>Vpi14kJj(cCN_e2;@WvriqFW!czFQ@+-RQ*%9Ml=`iJk=$deU2)Aq{`w`+`jbL;SrBlBTyta31;KXCA>DZdO}7VQ0|HdXi2S7R>`{2M12ajb5H# zM#9)C9*P4t1`>8q!XEbbxdLY4bWZWkq~{W)8rnOaM@W(%t76o`MpCM|dC8^|Bt`)U z25!xkn{klf>-y)xZ|it|Nh@L7(miP++Mo8RTpxB4nSp=yi6vU1&J#tS#1hW;5%3f$~XR`c2~Icp!o zl(CQ$?vixgZjU`xM*B{`(M7R+RmBgol7HYLsgN5xM^!Xma~`{;>wUli=NoAk{8&ld z3(-l?RCi@NQl)=Z?z7Dvh&v?zdHX7+|I)Va5$SW0YS2Ws+PMD(cwhMjdq240w#_<+>C|JuVrsiJsvyefTppp5mpG7)18FC^xaPdIRb~fILr2ot zeN#<;6<%hmd>nd!--`_bmRA}_j^L#_2+&4$i7rMy)q=JUi8J{ zlIon}ZT@QyiC960pK}-M^L6OuCC{p+VD|ETh(ff^yHjT^p5wiygOoif(HiFTDE~90 zTII>sqGe%~CO?RVBEm487dyt448NMt>u1vlAfi$88n4dn`_f=}19xI&)8|%wo1z|< zcG9>?iB_X#P>t0>cDH4`18rc(Nxi-eTwlW_}iyo)@7THWR{3Lav%Ki#q1A z1}7b-5MVq~&!9`Y_|bXrP=%V;rPR9o=O3bi%{~_^TS3mr~=~ z5q%D=WKCLOw<(gYMX3_YyM@Q}ORB9&A~~*teFD}lKQgR*yQZZ^C_|E4_m3}{s%#;; zVPI}%aDp{{6E&7Xi7vNy)C!YRzL8EKh{h+==I~-e%=+Hda5| z`rw`Zh}GQz^skulE1<)c(z2ne`hp=QDww4!(-ZLzVr2va`^gErm=TXv-HILz9qe?4 z*oAbtFR)x`0fh{}eq-Z)cUPuu(1FCPfr>P*X@f&JzT&Ov&JL!`EIt(%xK0AcxPY8( z;2R(XMwL05R(zCA_PPc3ewsPQT-&Ic*Mr%##B#?@_~mNxXY`1xm)fLswolubLYbwo zJ#KOLQNSe^s~=qpmUNHvba+73wm@fA1~qImIpJR%pp6t z8)hUPIwi60IPV?cazDg^xXqb>4#YBdFh?dKFd8g?I=fqI-cF`;L>}^z4b;qJ;dW!L zUJrb>l{(x7L(jh-7c@J&nu+P4e*Urz)lPp=7QZ-CK^QUzPf|rxLA^Wa^SdmYB z{qDVH@BPE%`m*Mw8{SO&xq4iW$pP)$*>?e661)^*m!cG(m+7%trG1Z+-b}TcEa1=< zb~&P%TOneZms}+4+N5e|_vvOG(TD<^CG0RsRmy+HdD|%Ya5#svcvVJC!uR_l?WYssiEaxgqdc<)!me<8aT2x5 zmvCUB>^}90c1U^Qy(1QKP{%d)d z<8*zGX-AE<|Agi;jatzjBzRd`USL|?a@*?!RB1F@*LHD}LhQ~P?&xJhO0UZR`JL3V zk-?;h3V4)xX|x13CpR{eUb3FND|~|H`hqtQE2kCe8**IlmQ-80 zm6Ql~@|fXmN3Z3vd-~@PuJiZ?%gJC7OCV=lud!P5TC{}H4p<^`c_Lft4<>puaiwk!ZF)N}@tULT14PAse5DJ5F`F5aHRDM#KWjiEt>%}K-IwR6 zjGEsUzc6^gurfB5B*J~Jg{wc|3Y#bW{bQfm+EE#{nWQk&YyI-st9F7;&v%$_i;C4D zLwvJnq@Rk;wWh{Att}tBms9uyVs?0`4=^m) zt{QMuoY82wH`AIN>cTUL*TJ>@F>K#-0a3DuyPUek1qiO)*g36SsA>AeU1WYsUWw(N zlK|tIzrNxqeBJeNE8;M2`~9i)Uj4~Oi$~R)<+vVSA5x)uc~Iy=B^dw8l1hz>;hs;IF*7IAeJnOT-TI`sGDNs6hV(<& z4Nut_!8D3&7el5(>sl*$csN)WRWpysLEImNy786iJ(w+>n;#~6-R)tz7S?}sp>`g>`_GrT8KYmJgVMB zE0whnz4DE<_ebmituD!gq8!`SaClfN7qRzQUt2Hr10qWF@XOUI z|JC!8O8!>@hV3h3_RG=e?)$&M3}uTIE07a)aF#b^0>JcFA%(fS(#B=*)-{TCr z^V?M*f(aDKsA!gt2$<18Bnwu{Fb`JQJ5gMGD=ZDBo~@^_@Ic0L^~IHlC+{pKhhl3( zrIuOXe#Xm5M(cOau0(%TiqxVqRlHa`>d+k?UIzgsH`hne)g{0%-xwmtjmyPdx7ax; z!%;H&T<2=j(g;E%?!`%?NZ=5A(c%J<=JI!RdM}Y9Jb|H-l z*F0?W%cIW^z_h0E(Qz{LXVZ;?_H#3ym!x8FdNvTn7-OU;tq+$J5)=_U^kAQ5^}E{h z>&ZajDOOMu%W}fu;=+5uf~DvBt|Yjl*Lzp(b8rYrM--=vvmwqI4H0p$a0IKYi_4Qs73;pqoIh;FT?}b0Av=gX}4l^#G#EhPT83h zrdR1YzJ$VhR`rVr)8^=&0wR`>qFKAu%I+@bcm{Htc{Z8W;>m`Nx_61@AtX2a{P%jc z7$BuD;=31~4j9x1*#&A4NpaTe4&mRp>cupK3%YiO@vbzN!2f~Bgcmz>;j$|iEN)@r zh6QXoKPGY$%=Uvt*x&n`A?7_&FYb5Eyrq zXROGt+wau~CwdaL<)3J_-`MvJ?aeR5#5?!FIP9f!Ts|46USyJ|-Qx3Y9||RQSs5JI zs6+A*%x_at{M`j);OEu&Z=&jFiE|R%6$udCqXz8DlDQhkvdf+)})++*}0dWHS`lat-!>Uxv zFq3`i&8+V`&%zxoVh*2|egsKezi-idTSCdL>T9^8y+FHUnMb#-br9eli*CYef}+kg zzOuRT659>iRqij9!uU6Kf!*tK4anbwZej&L<;;@bcy+rO8<2aL-I&LpyfDr?fAp%y zBRRB9Rrv5Q-9NIBnW5z{=lb=Tj!*4Wn^}g1S8bP9`?=T3iM5olZkyV)AY`?-(0@>grk@Rm47(JJWAWQk7X8ePi zC-08;PrFMnOs|GF6=pO?9pEm$lxB>lfHAG?(y!%b_p8m1HxBz)t5ZXKu6zMw4600Z zAKjKRwTx>NlQ&qt3N9PXn=rx|UdN8$11}ivFek;!+5*RDw4k&N$_!1OkMIdl{-+?7fIs1*I7?mnV}+@}3q?z^p+s)aj@;?g+OG z6{J4pMGXIj(J7&rW(rlTVVpVgk9>4I8?|B-*0mb+Cb;N8HF5`PaPntMq0dheqj z1L(E)dwugMeV94n|JB`l2SxQo`<_jftmNDVKqL!DPQnk7j7U;KgMegc$wbjEcOrE+AKWz?r;q;l@{GE;(^C&UwMH=?K8*A^`1N`%ja>rnz5 zxR*n34Bj`+9Kc&F4oE6jC5$1|1y01`%3DRQKdvN**0jS1*)4Ep!g=Jghh)b+>m`f~ z0eBzG!*N}07}fUe^U6S4CuR~>la7qJ#d*qol}X+yc9Co?K`GYFAhzisLp7$KMB)?? zTFAEdu9m}Ly3Qoc9G`FR1QIl3Zt=e$VO>L1MWY7$pcbRAAl?CKm~XCeX|H^@-6Yp# z&aCkGXUY0swq?v%`l#01a_46gYau}yXNsB9F$%HB$gBdu{AW91Bhz@_v9x7zc(T7HH0%bNJ=b6Q_MkQo`7i&%tv1+VbXxx z#x=Xq{E$bq$rZr@u83(B;l&AEvn|8nd)xhS<_jH4(VQUq!+x5Uv?5W(yz$X^e?z`2Kq+KzSkBsBFZ)&4{ZOTSJukCDaG(usTkwoT0T zWzWL95hUWvYTgU(BA**>W`2tYiTl@O+=B(+w)T`WWi_`1ufSMwS~R8Khw&}dj;+NHwYVr3)n*5cX{z@%Lrb+^1&03LMB~paq(< z8!o%HBy0dbRu+MZh*9&~1_?mLaAjQ%IryQn|Bugdk*XgWLV+AqN7E;8?(oCpSzCTM zFnQl&%S41NVkko5r*S4C?foFOFH{3Q)d*kGX6gOn1yo91jovk^eyeSJ0TQ*mZkOSGhEM`V$X;mNSF{~ zdjfTvE5fqVxX70~<4oFgk`6uSpR&61!P3#{?%vHPHAlhcP_k)dd8GXkaKh80&J*v+pF2y+2?h4!py)D7MD^K|PIAWsiw60766t(V` z7?Yx5D46rf5dWPfUWQ8YihmQ6=6tqHCe=1)h1`WR{Whc5oP;@fC&wVh=e&y9m|}EN z9^qQgs6LP99vd~Ceg8L`fn6q)oUMki{H>)7mRY0JA(*&KC^Y>ow+w8Bj3X>arlTjtVzH!hJFqB_UaX$wJ3CE`Yo9Y8=hdgbL+p?4PrO{JzKZBVlOxpvg(uV2{4WB`T;UQ9M?uoeTca_>f$N`#b_G(D*%p6CR=ctkwhw{%460 z>(#xZtJ^EIa(Rb82%;`C5J#%%7DWMs1S0O6RS&p)G8eEQ{ z<+c@-+3lf54KC+AiQ#;EK{9;ToqOqA6%mF}u2|e(t@R$c`gkttj_%1!SnOKQ(TALs zv9bk{mNq2EOq4sPlfuq#%w={Xb|#eL(J$gKKk1=rYYhQ!{JU5ou`|@K$0!`XJ7>P* zhF+!)pUUziM0e9$y)fk{pTFU1Q~vJws^N!h4y#x;J~*2~i+j!9N995DU-bt2X4v7d zr@<>{@HwkXY$Lg}6M4qRHnIcd0_DL%i<2hZ*OpkX?0j3<&k1vkP;ZH-2Kt;MzhPgz z{XlVNczCDi;%_!suR?6tGLnS5+x6%Wbk-;xqY*YYy?*(UV4$9yWfj_T|4~Fin>Sdg zTku^>@iu3`A9;dWZ{1p1&r02r+VGVmQy@Ophu(kxW(F(T1)VYhs!)l2V>~L;D0E3D z&r5wzAs87HvYa=4I%v6i*$ED>mAN8!6$-axM?sf@NuL}$fO|vN`$yFQ`y?w_)m%wH2j_Kzq_CnogBFod7RaFL*k>Lrvb79o z;QT#+`uBZ&KBbdxqfnQLc1-G4!|YqR*;&>RH=lLv@cUuQ84Nt?wXDF@NzH$&>aWebz2BCT9q}gf3&k0Zv=#^0lL#b#OM~NBD}?AcP!l^ z&y}N>@h=fc?1o?FT1LxaQq%3Cr1iz6V)jR7K0BFHbqa;r*hcZKyZzm^N7y$rjk!Bt zGu>m4&dd+C)a6J|>!_39*jwVKT3D8tcj~1w0nG57_-7>?S2%*cC{vw~a&yASIU@h$ znJm5e?`MqMvkbh9O%i)n^j=${LCIOL?Wg)oW9O`EHN97oFtTxMOYRUY;HguuW!TDj z3o-uWD{+KvTC9>mjj=$LYg2g&Y-TO|`dl!Bp;4cBUhMG9aWeC&OKUS81!WVLhSXT)B zZ4yj^;DrwY+eNTuv-7ufvW-`!#P&M23;A#L-I_hzmq_$ZpmyqO za+`atjr_m-706DCRxa=R^M*ux6W+m&C!u4MNlAq$Jyjv+u)3z=2 znRZUhyO8nE7l(NP`VUQ84@(>Wq`P?_u*HlUsa`ZhZ9d#5f*MHk-!Kt#*k~WVakk2C z=l9m`Dn}z{<|y7OQ;$HpE}R&nwCb39OsZSTIf{EYWt-Zf(jQEz`0mVk{KNgFWY+`J z7|3SyF#3&a#{);2y_r%1)UumzGG^j-P59&Ii)UKU&EL8gbEa{(A4FCRMFo{8o^E~u z@fEys%!Rm1{w6f8VjeN%mbv?&hR8GbsA5YAM3+wPq)Fu)<_v~pta#(iClJ!6niv)ju>mXv&kCuq;yzb?CJj8$X0*DOUY>xa zxq4>9Y|^Xx?Y7g|(wZru*R!gMfgu5%5=HO?%gYSnby@2Cp6b*QwAi|&OvAdBb`+> zc*O}%WgzO6&;7yAml=RNT~JB(mXSXAHp4*fEL2Zq7^j7vDu``vA-1Qaphn^5P-znu z(MUdZUo_Hr*v_^7Ggk)A=Cb=&d6_l)r&^g=6N)Tk+0HTaBaYBzmx zS0aIiLd;XU3+mr6aqNW1wI$yPYw2B1V-p3hITLut$b9Z$3xrc_<*G!2`sFq)Z}v}u z%f7{erL}t&Z?~)=Ntuw{Y&@#xr!HrE%JoOib`L@5CYra@AuZtWIr*s1^L;6IOL0_} zTcnU?koO?LCPMYJy_e{_`e(iF`5lWzp<#?~kqju7gCOOSNUJq$u21n|y*E}@>V-eQ z{yFc|GoJeS27r$YFTJ;+N;un3*~Y9K`jz>Hn@NORC3dnCs4VH zU}UKHrHN08!0f?I|4Jt9CJOPEkK09wm8P0W&uMHfrRg1sz2zxz#EQRM0qNQIfPiC@ zuJtQt494NA$14s?xH@Cwlp=d$SDt*=Uo@|I#nFy!zFw;e$?L5hKqX&K$goST`ow6S z`30LNIUQW#wA=0{8?j)dXvh5ae%tAOTP^vi+8Rfghz`1sC-4u?dJ~W0}rZ;Ry zPp_sQ`}BO&gL_t~D{B2n>XX+eK;|t@!aELxPWB}U^inC#n4D`^jEyM*Ai1jLFLLod zGu`tMj-M~V?tW6A-hi6}O|XGU=HaC94jcBxyxm^;<+Fk;%2R`=nMUgc7c6AB*yfd0 z*GQhos3ZOkjX!bys6-oSZ>x0gw@tkwABy37LCbW;e`NsM4lVds1sja)>_MSPBpnlV z?I@Q0xe4JyaVJ@@!yZ*xv@$X!=huwxSCjEedt{qq+n2n>RhNAg`cHloiBpu)a-j9N zFss&x<5Sdt+VIk3Okii;eDE_bAcE4V$W$yn=8N91p4#88`0L#+CIP& za-%S&F^#d!vZDPR5Q=@2d-0s;IJ$-Qr7(*FB>4{mZ$GdMkn92wgfAVxTVPOS(oZ#a zudH@N)T@x^NwL7qh6-KYL6^T1Cyyrwk?|WT8mDN~`l1ZYX z{!5SQAk4YE4n#&@FKw57rZ-Ech|aohYCJxO@GjPDIr;Yk9&Ln^D5ad zjF^uj)xZT8pxt`8&weZeaxv2vUj7SDw%iP&wk?f}4kUif4M=`#%-ktFiirl-83%m+ zXyEk{3l)N0E22(^01&a>=EWhOkDq9BpmF<$-;MwbSREe{#z`Nzz)qqXNTeVbsmyyV zS9K+K?0{X_Fla|HlmUUo4y=B6w=q)wZeWkr8dR4++SMJP2J+32w;&y?)QDUI3#h8( zNW?cU`>}Q96Kj$?pgX=<(47@3T>RK+WM*~pzWU~K=vKNl>b}lmDzkfT+|2OE~ z_Kek+ca0xOE>{;A0Fp+|ulrB`AkS{XIKQO*$9yAr_RZB8FHV18a<`tXwFHo)ByoO) z?*Ab({0H1HA3H_pq~-;bS{=CVzN=GBqPNQ=M}@A}ezm^T>wR5cwJlQHix9@i^+)}( z+B;1oj0FT7chvNjGJxwN+IIbSDlaUIf8OF%Ao#+lG`%59#n5(f)m|KJq49#}IZfX- z-fxrYJ#jC+(OBEOzOj(!%cl!a^3pWwbZN!ZUERj}bO8zt69%yFj(_+xGVxO;i)sBi z8%W)QTSeadi8!{Apol_}j?Hpir`NBnv??gP2f{{dmA@M0Oo!-Jb0xYGs^HgK$nr}H zIKtt49anGLbJ>+Kmv|$B(b#PRt|!?0?O{m8q}RGt)cOIG4kw{xTsf1sC9UKKnkA-; zUckPW9LfO%FRsVt^8~V3Gs$MZp&bs)jdD3IE17w};2AKvf`8`Zn{LCXadMH`I{v87 zNi%vUlK1H*uTxT1hKQ5>)Nn@Q-qRDyH^zg5F7V4q1+--pC3dgFt%Ljxz$Wyx; zA_c9C4jhV4c!f5GqBb5F;h2D5zP63{+t46iHO+B;g(Q?*Z8LAQlF9#Mw^+@w%BI)qvJ*b2X+ekF%iDP_PRKw!pmIr0uIj%7I46& z)12K1{^~73rn`Z?~FJL~UV=z$Ytp*!K^AN=&`>`C?z>^E2q< z{Jp><5#`)I(R0apj>;NNr)m7LIPx9y^C>Uc?rIB1*2!kDeuLC^zw?zHA58$`ddLh6 zda`eW&-TXUnz<^4G4do3ifr@bqe+vvV1@F~+&Dzkm<3Ix|IFO#&apzc1ugdEHiaYM zo@ZFEZO4)~f+gQMds!Vucs*&;JETwilguZ9Y|tsAlB#1Ky{Vl}N-_6l{qRHz+65Q} zcimNy?A}glur^=-=648eK2lUMM>c3Q=!CpK*aDkTP9O&w;bS~@%}ksA`1H0N%_fVh zXd-u(3OLpJkAGWyt?Xtwyx^p~l~;EW*$dZ=k(yE%7|v;a0}E#?o?Ool;;79(L=A9-^=7{Z1ViHk9PTV zzmN{!b({&B=PBa5ucl{sP3hOO94FHzH#y$yTo2JYNOzqFM<#uMD15fE!)&zbI(P!d zYOUdxMsx!YV!qy-E|_L;vfGI+sCjYDLQPlKNDG!<`u73SrU22o2v5_6 zu|entYW)^;WN<33<_b7fw5;nGosK^-sqMfd=NhN@BxTl1+|mL?~E!dR2D(%KSaQ zL?*BxlTU0D-Ipa%iB_x|uyrq4hf35=m&C6kk`Z5PP@9m&AILk-bix$cEkaduAr(q9 zj;P{d=ppXor7)ZGA)neb9smG_R>cFxkMTw>DW?lh|wHLkse?33*M{l&nlKO@(bo&A$A7ZnbV~E)9ZfK_Umtb+w`qwN7 zxxTMsU*s7u^l(S?=98GvFO?;1XRr>KNB)3QT zVHY#Ly=PUNy*tZ*3`uRia#Ex2E_cp9<6ix$h`yT219Bdj8^!v;pM!o#t_6M+w~)C~ zdJ5C~(=uuw7yhhjS%0h+PPGT^e{3{ZBF$gm(Es`edVR(-42G0LXQg~e+jc{UCp@*ur{^e&5B<#o9H(uhiCIp~ioY#g#KK}PrE~aV<7aUXC4we0>;>ABt`S38K zVR*LfeZPBS^P;y8T6%4#9y*h$OI)d~k30i>w{GaN2n}5i& zytiHFfh~l#mBmtbZ>%v)FZw@9ZwwE6Wq;nskQvX?8uCXroIbuXxi033MDg!m@(=RL^)NvH`WxU7k2$% znN-j9j_KUWw>mrl0*RO4#B=he`!(|IxXMPkQg%v}iV(f(T#rXNK2+@4gV-0hB=7NO zbyhYD(b-ed2L&%N5)4{Vf7{eHenkKNdvKMiQfe5@4WF4eLLX*ZUK5v2;;TG)9FeA5 z#an6A4%r**w>;rAeniTv6*?k%8y7a|{EMaIn~=4Fc@wZ8@Lh(95P@qs9}xdB(D2(S z&MqYC7iG@2Y6dc|@FCVt3^+PLFyZcv6BQ5ynGqWiEXc7=)oL1?RnfB5QtMKD0%c=~8S8?H7T?U_ z4UR0vCqb1mD4^J}N+WF&)OzclG_o~0IusP>Nzaf3Kl&AS8zVCwSRn*u$M{+mam_XOx(Yw*f}j#H|N9Z zVNSdto9=EB+?%5Ot59jRK}Rvuk!H)}3;TdCvHCW9LlC>Jw17UxzHo-GGTO=vcVBjd zqY)jdlQqz~b3)RNxL37HtrILGM_Kk7VMT4DV3@Lw%Dk()9a`!+_X{;QuyX{DN}2BF zSRNBXHPgDan*ZW^jQH}-Y{t1D0HoCz*~b_CSZV?AvCD*CBaHNW=GyjK;|IiD=}~4owH_H-y+{@ z2J-1d-Sbl%47xibe@Gs)4Z}TzH^_>dmq9t%#Ob}79#+jor6{h8p}>dvd-~hA%1^QUagf#!>W(OUS}%NyHk59?Ne%4gq?>@UfiSbBam;P zJi1DW%{}3`1pu7XIBV(og zG(F=ZKEO=HrxTnDN5CO%3(A1R%z;MBRk=xZBtKyy3I!=CJF zEOa$&F{-O_6h{ zt2i1~zh861Cnjfd+q)Oh32xv#&IfzfBqTi=PHCdIyBdJQmUpipA!no8R}#gRqmA7R zQV4x(>1}c4%dlxq!>g(jLToHT<|T7`%b36j&|Al2#3_Q8}}u+Qt!`MSPaKR=?i zirD0dSaNKk17cN-=;Awzu$^U>S6TMrf^u$8a%ne< zFvAd;$$-PTobDp}Y{Gj?%nnY~70CtPJC;j;?rmkZpO!257Ic=@(k{3@0Ea$OlzfAkH!|FS*`D-?ARb+7+< z+*9mJ&97`PT>My^8u-zL9tlOuKV-V^KxZFS7Ae_H?AJj1gx+-nMuy~Hb4L|X_l#WV zqy@Y=9*&#okXUFd_(j9&aPD*F#cpaq{4xOaTkZ}{f#*Vm4!K%g^yuRLym-^s_6CE+ zvhA%B6|I)6b0X4odBG~v0}GxDC#+6AyuX{{sJjrgo}#>}(A1Vo+FJ1f_7~a5yYf=@ z#;+#Dl5g>t?wC-k*pQ_#R6{io=fvq`Wm0eRP=A_SIw>;H(Nf9i2JxC)!Qlz5t4TAc z@FF=QBOo9?`))rUF{eEGbFFWjqfLh6x!dV5+LpS*9Ly=6#c7ZEoBv5~h$RrTtPuJT z++o(T7+qw9HI)VlDv%ar5>hIV?m&)c*=DJ`3D!{=CJdHuHCxh<=mI@vp18mS#V|-UK+nNvs5Y~A3EJXAr)VYNY+2`-k($dYl zB0}3k9TcvPM3_&sr&god{BjN@_WK)hYr9RUROx$h4_{`0aG&Q7{47X$4jIJGn9u=) zUNRi-p#j_90%I;b8}?{8)zsFmvzYAslS~!tecQ@rnd-$>A!Wo;Q$*jy z{EdpChzB*w7~3zSXDolC;W9enj<+IBaxkv%+dLHx_KJNXz?i=g6BGyD2` zjoSgDrhzs97qzgh)HJ3#(A>J=hmqpeim8m69`QBnQ2$>800!kxMb-cM_ww;VF0!Pu z@!s~WBs#`SyydH4za`l#lO3D0*RDPr5J`Tx2FD%<+)TaKLb7V5=~}*z=kLMh@54`r zv^s1Wv@vRnc?{iUovLfvk^B@0&X-HADaTayxGkW2n_OV|6#q}-;N9tz2z&BLh-b8g z79)=YYaf=Feh&}b*KjqiU1=@PUjgcLiJEgzI)APBO*cK~#cThBYV?93V@06fclMx1 z+w|z|?TeC}9|>(d#;q*L>_jSp;dAQBQ{WKbTxy&R)iYpKhQ9{O8)UV#e`frMt=1g$ zMO?X)Aa~oQ4ZxI`4KrQNf?4=OxEEIZ`PKerM9`yI;lk5ad2?-x9RhB@YYNQ&2((l8 zuOZs`+=0c^!@np4aPZ#)+H%035_p0d>e{n8eef3&_`e5!`QQJR9mSv>JmrePNsdeI z7)ZkQzt=e+{?FF?cDg5Au2s}*By)CDXsybnrswd-8z(ayp!~V-9ssoCRQrm!juhBY zWR+2Rx2L>Hd6IKh`hUGu_aGPQ_6BEd#!8PJ5#@(KZ0@Y)dHn=9WG?sa0TDyHQ*5 z%yVDOigIwTvT7y4yCkEUr`me=%7T zq#~RkE42Ee-O)(-BKex~F+FyjaHTeSeqHH`UfA)TK2CP(r=9J$jW$#7;UuTx?Q9LV-<&}tm{9!qzz*#?p$>hKj!W3GtF{;2G<}9KoqR^>HqUj@ zYquDE(RZ=46YojaY-`oenR69E9e5b)W+7mM9eRyH)eGVKB{+4<&Pcb>vg+PP&K-yU z#=i7D_jMl1#Ic3@?a5!5XTCtf=2*F8?gD=6!0SEt*N>#~VO`lu)!8EQ%9y_wDE}-^ zB0(y+t=R_s>)~cUagE#6yK}UCFLy_~TlovgpWfK?XJ_S8W37#s>@YNH;8b~@!(ii7|Lru4$F}k!&iuFJW_Mxc5YQVZa z5jorTJbTN8Z6#wshY^?{=DXV?_S$wyu`hwj()`as`F-oDM%cSAx#wxKr*k5N?>Xps zA?~+i(f;UwwHDGEBvlez3LFbWiJ4!?c{C8UvM>k5h^M2PJc^jC%=L!}%H>He*c05b ze%r*f9HC2B{WCQfA@<@3T#4)|$WDaU+v+=p{K2sq_fD!AE=;k+-mO_-}gxESro)@Rbfx3goXJZwA$hiV5tm0 zL>WjRns4o|=0T@F#$3##0IB(?L-=f?-s{YhnZn?X=d-@d3e;d=drfep5S65Cxf`~> zA{y^pVa(_kqpA~ZgedR=kKzS3;xWjBG|~~H^UybwN@f5oKF?5TU-j};3a7o_HN*X@ z;;Jp#vAC*hM7xSeEwvN*y`Xnn`*s?aHgV+(sck#J4zG_GKK+ zu(MT>Y@V1{PYkaAs{KI7)=7euG65Y2KAP6kear4E{`gVSYz(?`59 z1)4u&9&vo_jM=Y)S^*C~ap0jve`y;@2NFl+9uj2&>E%9hST!E|xi20|=N_s)UtH<( zg>#^Q12Dbz>l2Ny!wv3^c7CHzjiGMaUP9T7(}Ghs)8LG}3M{p^e+AyN0@|h!9mfwp zRkHIN{4|V)?Y4b8uo}`iB;Uspl)<1ky|g|@h`%fPad_Fb?gdWNe~RBk&-*=p~Rgc1KE{VXXXVzz3ER0c2MoAsGxh)`Sv^y z$UD$c5LYKeV4(g39DNtsKdJlXwUF_F6yWFPZ6yZ$0~`6Sq4%ei_6F!bRs=AV@8bVf z4WGtNGoRhn->p&w^eQ3o%al?#Bdrq%qXuVfW~167m}ds-+DI|&>#5$MB$^?3iW85a z%Bqvzb)o0n=Y?zWtIDSjK5h9NtKSv~42XXP8fjMKYkxAm#g1Dl-wSuw_G>tO5Ch45X7 zE(v(sa>8Yf4gp|Lg&?r3lj*-WE`PvC`OYvPZHGS?-OIwTY zR+Bxjd#I@k?yuT=UQD}hQPjL4VCQRtml*>L+H1onU+L?nu;lA(-SeOCb9ZV6QFE0> z#B&Tu{^dIRH7f6Telilz-B4(`;ap%_20ydl!HM|4QJwFCxILn9WEP)4YyLp1F%@<+ z@9kL8<91%{w)?zP+uGNxHv{t&_WhnO;GfmEn|p>|+d_Oc$FaZUr@O1nBra2B)>K54 zq7HcGBl|=#tA}ky)^lUKLHn%xuF}U9WT!$gr3#DYJ{aANK>b~+lb`mAfZysad#;L( z#VoqDN{WvmX$7xAYd1R)K~tEr3x3bqTj=PL25Y1wzx^m4C(-+H-8mqfA+ocgV&H;m0W0$Esq zTsD|e9Pi-2{M2k_``?nVY=WgE3FRMV6~1aOSz#f5ryN_JFS_xt0Ur&ZG@Tj@eR_2C z)%wkMs79q~oDz-p8wrm(G}oNUerSq5()yyy=S((O_2a?UNss#2jv}ovw~fat{P`c= zud?eFxLqhu?nSvt^-;IN-LobfWn_>?I6360f7SwIy=C1m6@K@{OYg4-YJ&62eo{We$kMGSWFiWBl?r$^FW8~oD z+0eID*7rE{YUKUqKOTSebNGSf?p$_uQk-CBRT$5nW%q9E?GddlO)|jp^Ih{;u z;#?p90G8+ded0C72MT^f+6EG<*)f4xA&XD$U)GLi?LUQCI%azx=^foeCOzumieu$x|Qf&R@bo$T=vFn|D;(ZxH|OqU^3hc8 z^LP;YsFw->t4`tep3A`6FIe}5uk`)wov+|Bq14c3iRUjPchR6yk2Os__UB{vdke6x zS5l>VvB}m$5xoM`_hm-s_GZ?pYCfI`f7gl0xhqBl~;zBY&BeZZVXuG5}uYKFg zoPYPum0O`n#ursHgv%6-x^KdvzqL1cZ)!(^HOsrFC%*+EQlF5b0GF2Y&EI{b%!qY< z9_YvFIvUpd9Ik^tyS(T&^9vI3%+8ZHAC!}G)0I2c@8W6d?YSxRKet;D(w(HkgpKP-~7g}6Lmcj-9{ zou%ftw-Q92P`Ac~i@Z-Tq(yb^b~qJxR#`}Y%8eXtN%y(BrPliE)86l|RtcxctqZ&E zMk1A07v0FOO&-3;XKGtSTg=#n6>v*Jx2q(lhaFdL zzf(1AA--bX4WMR5xk=mBIm>jo``23+T0FGf#!$sZeFfYgf?+$s5aOKQgYu;efeF?z zii9A43VdOZR@rjrN_bwB(RC~dZ`cf&#he)WskF0vxKb}X4!0#>jM~MG zcZvbzqbK16$teHHYoPHmLUj)N2d1AV|@Y;NS6?g>3^p9|KEy)UlHhc*WF?r z)c^za!{wo3>|t%`VIytjW&`{Ki3mRw7Z8>d5E0dXC?+i;CH+X0Pxzs2?O*%;zZa;%?1}*eKx!)5|CB3Pg#I5$9aPK! diff --git a/doc/tutorials/gan/index_en.md b/doc/tutorials/gan/index_en.md index 99c8d730117a4..ac9ed37b22647 100644 --- a/doc/tutorials/gan/index_en.md +++ b/doc/tutorials/gan/index_en.md @@ -4,9 +4,7 @@ This demo implements GAN training described in the original [GAN paper](https:// The high-level structure of GAN is shown in Figure. 1 below. It is composed of two major parts: a generator and a discriminator, both of which are based on neural networks. The generator takes in some kind of noise with a known distribution and transforms it into an image. The discriminator takes in an image and determines whether it is artificially generated by the generator or a real image. So the generator and the discriminator are in a competitive game in which generator is trying to generate image to look as real as possible to fool the discriminator, while the discriminator is trying to distinguish between real and fake images. -

- -

+
![](./gan.png)

Figure 1. GAN-Model-Structure figure credit @@ -111,9 +109,7 @@ $python gan_trainer.py -d uniform --useGpu 1 ``` The generated samples can be found in ./uniform_samples/ and one example is shown below as Figure 2. One can see that it roughly recovers the 2D uniform distribution. -

- -

+
![](./uniform_sample.png)

Figure 2. Uniform Sample

@@ -135,9 +131,7 @@ To train the GAN model on mnist data, one can use the following command: $python gan_trainer.py -d mnist --useGpu 1 ``` The generated sample images can be found at ./mnist_samples/ and one example is shown below as Figure 3. -

- -

+
![](./mnist_sample.png)

Figure 3. MNIST Sample

diff --git a/doc/tutorials/gan/uniform_sample.png b/doc/tutorials/gan/uniform_sample.png index 4a96c45cae82673f5a1df986f2643a8026da7937..e716c48e782019a757bed0cb443f2ed97386cbe2 100644 GIT binary patch literal 24880 zcmdRW1ydZ+7G@v=4DRmk?(Xgq++6~}CAhl=cXyWrcL^jo!QEYh2HEDlw^jQWcB-bR zX1b?u-`jo8`Q%Kbijp)U93I?<4<8U^WhB%-e1Iqh-XdUV;6FDR-~!+e$VE(60}Q-; zz~&LaHiV0swAhE*34){d53NKMML&FKh=+eRfdaN+9c6S}K72rPeSbr&TM~PI_+Z>3 zD*XaiByDXe<88fTDJLiAN=-#3ii89Vh71ZI*|Y?0 zw|e>D8ls?pKoFuLC@QEZp%lnL|NEQ?1tqjN+6IN{KcAt1BuGcu|950QJg{|SiWdS4 z9QA%MbB&?}yv|3M~|BA)G3eqcV+e@{#TmJk(b z606jxKoj(RN_JT9lwGJbASWiypvBP#dTTZc^ua=dLBT!u-sXf_;EDRzB#=Vum>pHC7+Ec|*h35UfXQzRW<8ZY!D^Zy(X8z4TB;2Dmg9$6jJ zI(~C7Nh=UdgC4%j-4f-aGXh~WIhTjZi zs-B)a4HhvkP_K5Z(TXzR>8I7IGVoXo`Ai1wQ^0A8c!Zkm7UJB`S7rCcQi>815*9t@ zUp5}0Mh?uYi+z246^oJ|D1X?@Qax=3BP$dod)B&NtSi30zH<3{{e2?Ibjn@SIUfB+ z?Yc9JNT=UgPn(R^Q(b4+RoPA4Xz@$RQbkTc`5DQm!4zt(-7PwaO3`*@A*Ab+K$XpE zf`B|@C>om{I6D67!uRQ~s^w(v8JNpe+-$yD2Ts$ueQQKD27!=oPPN!o+zMFI_N^h+ zm~^LHBEhf43_6X2swhW66I%n}nA#DpebM9@G-|~WkWhB)vg1}V!2A{0bw5Sb8ucg% zdf#biWtA!A6h1%RRAnX#p5SqPma$)ME~rp1n^|ly`_yEx+ppJ&&1|?>D(kkM=@Hvm z$gbzL;#Ht(VUe*oXk2PK?6rOIZD|({3?`|yr0W#lUdt66Va~)vTX9t{Rj8oL9E?P- zx)XQTHKrV*)M+&5u$`lk%i(KjU+eaN>ptp!o)%PCWR-FX~TYvDzku zh}T^7D=mJe%_XZsDs~tH?Rs^`3H1XALXeaMPFA^zw37!hP`xZJz}(1q%o#D?lJJ=eL|02mmM3eS$bjeuF&7QcY9s^!!(D+&2TAETN4B`W+oWLvb@v zBZCDSC=r>|r{R&V+l(V?zx+F2KF!>N0{Q0(>cicNgGPtDY$mH&kjNexupDh*Yuw=% zaJ~3IHt&h`a`Rp?tC{6JI|acZR$!?j{$A4}(XE7lg$VuR6-D8SFftccn0=d3^MeA{nknpeo zt_W9&M8gySL)=Hw!HxPZ{`v_lcD_1jzD6(prs$K3LR*E)!R}ny?!7Tvo9hM|C-EY zuk-Z!uaAhlqXP;vboU%L;R*_f>RRC6Is8OvrFsEnoaGCLCv{faR`s=>9Q5pk3BUQ8 z$kKozFNXGJ^Y#=FOaI;6u6eDZi85#|{$>xXXCoqW+4t9%`=f{-nsH%BXM1x_F%mv6 zPbqX%8$LHspC~mZd%JtWn?@D2C9N{&m5n%<*%=!zKGbwj5N|vEf71lck$B9olSN&o!B9R29lpbpoIJsSGZgdH(iL zw3#F^9^=NpyR#xlC^!<{OyXE#;q_0AxVG8rUEylsNRc$^RKk@iCRAc6j^?&Clrdj# zO>YL==G3CwvCO6u8BWg&{LB)ds>4D@=!?FtCQ1d&@0u_Q#Xi zEaWQHO5jT5)8ag(In9O<{_%2oDSx@}beH$s{Rxr&h^DsHAKbtrBZ5r4J|LD)e^?@I zACfZ+K3?~YSo^#7hgB)N6^Xdr?}bjqk |6?(2(hcJAWnD<)m>0t*BjdFSGC!8od zdA<%QE03tCD66UKzdL)Mqo&SR@7L$YBOnc_E>p|ozjO}%Mz@IloWuV^s>!9?PW67q zD(}xQtZ}nZf95ogh9E+mdXUb=PT*TPgSW3 zZG_H-p_t6bI*Tgl~3i)Td#ZGZu`89C3TxDD|RB<1J!_#c2pFU8^8fBHs2%uypIo1IzbO}9-d&fN zov&(>H|wk<)2b7>IHi35Y!h@-gn#*&VVgzuWUe9-b4p#e^-`-+wMe)|?}vcYGVr7o z!=iT=b$S2b|Lh};4ve2WcU08%;WD0O!`_+ZaVT2LRGKe-rfd=d$}j%)UO4)sT|6As zlu5#PB<}VSg-EJQ>1C&1HI^5HU+HB5&a#DqrNV};#cn~<@pT|D7h@4ArNL~3?!8id z`$CJ)?V!PC`MbPWF-tQLE0&Gux9eE3>^F`p(wv;|-d}0TtRro#Wdu{vCtpi^|2<73 zwL0zKJquzUgXcKLjQkm0h|KuP2+*$84hZ19=SePqF&AAi97wz<`IaIRty(E^$J?V5 z5L9C@Xsh&bl4P-7ZWC$RF;jD%u61w&H6;AKBOi|qh_nhbu+s~k1oGNAVJktZ#d@ZA zEWIvQ^=J1#2luCcU))hP7mc=lHfO{l5j-rISEV2yF@sytAcW9=r?XPe>U8-qxLPZ9)$(_93VNH%s#~GtUc1PNv!!lA6 zm-Q+J_wm`>caR5vQzh}wIcrJL4(ejF4W(l4v*22*)1RwBMVZ7&&}^xNbb&AGjQ!Bw zEdf=s1~J(vk}UC!F%A*iJw93 z_+;~?yQ9>ziR2w9N^*v^4dajNT^;2L8SzsL__g%{W$4s#+fza|pVtKLE*izJ7b`rv zHXDW({Silk2*m;Y`fT2KV48VDPuq6P=&iQ?7_`y&@3Nhu5@Itpy{woEYg z_4#c4IxB$ihNIx|Zo;(Dg{(DKfzQj?UD*E<67f26Q#diD zl73jFR$Uac$6r-J0o=ND^Nk;0_+Rc;XsXBJt%IlyI|q|B4Cag?@|& zmo=ABw{TMUOAY!&a|0@~p;)JW zWe|8lvb?52Hx=V7QMJqMgwJ9;6TU5XtXHvn_k=~N^+&j6>R{B0BT<^cin+vJrPP?< zvu--?`|A>8d8xTt*R6iDZOM@N>#guxlTrgP8v$1Z_TD6M5Kq+GaC1zg-|Mra^loL?(L~1jd;W z3;%ej^B!GW7NtS#M&%!Ba2+RtKu)Q6_%Ug&8GlO8=r^nFjeFJTO@dN*g`BwfNUaJq zX+A3#JZU-H5Eg0NNNiMH`Rm<^gpvmu-5iZsbeZJ zd9(ZDb!6Q#R|x+ZR&F#FBaG6pgd=-+qm}YrLQDf6YP@dqVFdQELUv^50a_}6T(^^) znvn~q_I~+L|3p)&BFy2vp)=BceGWA?^^Rmwu}f@!A)RSmOd~(B7x?b5mRPD_6NbSS zdi0B#ndEei>x@iNMJA2u<3t&idOVT%&joIG$}~DmBjio_KV{@J&q`46*BNOT%A&v> zrDWM|)N@g8Gm9|9_nqPA^%6s&cFb+VnrrTMxC_@ zKK$K~XRs(-xbv0Eei_S(u}KnFQiN{or_*F=B2DDzgiI4SuOnA7xvg7KYc!3K7k8R7t= z$4TUWs$vF;N-vGJ#y!(g@(OrHfNpWbf-FqP4y~TT1n>y3T@MQ&0br z0a#>3O7B`EX}nO&2#(gkrs9-u#e&Ns4L49N0*4Bud*KGb&LNO5- zgrNbubdi(rQ*=!xj6P(sSx^anbxoeg6OJTlq*S2FiFilMy6KL ztFthqsOaSCWQe1&E^xDFxZ6syi1s>x2?W1JIPZfC$d4Zn=NX~Y^h`hc z#r@s&7DRwTUBCn;%o%EoEsVt2806#oO|9!|D5)D$Ezx%V)MYo=g^eQa9<@>`BXytq|e>+$5i<-lJc@jEROq&}H za;@E0UcHn-`hhyGnN$+RQ2-b?L41Wgjl)tS)}tRE04BcoR=N>ls8grK&7fL2K;B#1upi?Uq+bz_p_V~RVHcbjO zg$mo?e}we?hg6iwS(os`gNWuIQM2{5z_>P}(JQs}UdisL%WMd2=SLR2_D=_NGN+Sgz2TvQ}O008&eWW#jEJXUWOJf~rWb#nbh#i=#o%{C{sRK6`3{TNCAW)y$6 zmtp{mh#1pwF_scN#YJojwq>5HE@xI9p;!5=8GAGuD<)?6q3Seo|7+v*I5(jHE?VMR zAb68TnpUG6rF)=&OZ`u?`1SR5|4H;0$2YysuC9#-{~r!(*1u92wTFyKfR!_o)Z~vs z+ZXzL2T)mww#(kn4-G4rEoH0;LcU0HS*epBMha^sqawUC3d9i8dJ{No;9^_^O{@PU zvlL(}Bwm>jcppr(JggXN5YMs95v#VnU@w+<4}fl`D1GqVP>Ay$=!PYb$s}5pWWP~y zSMvPb9%9$+`U;W5>oQl*hLdxZkbu&^AF_xyeDQR7>1g_5Y&Tt;fJUtp6TX!axeE}J z#}@!a=NlXjfC7~YwUW8XjBCbIK| ze3g1*eaLbIyb@-LWiLKdHheo=Z4>@2HfYItke_877E3|uZQq|j~CwU35 zu$5?IS=@DM0Ohp@PzBR;0BT_{LNkN7fOCi^4I!8M8!GUT3}Z%oXkh`_IW$9zs|YS|Em5uBShNEp2^;?G+}9L% z?EZAf5%Re@a?`gH2>U*rx*klJUuq(@u#IOe#AI^*TeavE@_+qcSznC9T>EP{IglW_ zw=B<3!et);dYX-H3y2w0UEb8ABh-OrFwU$kR+G6&J^rlbBudeV@w2<58(VO4TG%e6 z1Cmk!li!|SUXBi@i+JGuX3I#piaByfhotmhYY77nCW%FHuL1?>$(eEve4Y z{|vyOx%H;(XkXsYOE_)wP^m@kw4%v4h-M;(B_l}X~o4dO}m=mh6 zcb}2nSJTQJP5k2%4M3boKrTXH(RyaXW!Z!|)#TcAP{B-d(r>a%G8Rgp5cnbuq)q|SFOnoC z{StHE?Ujq`9c(rvxv@9(@eer*5Xjw^=Z$n%~Upu39=amgr*5(Hf5>BH$%7R~+CCCr(~_k{02 zuybW5OTQ`gU}61@Or_lwad0mi96$7Hu4+Ml&{5AqP1b-um4;$r`>b&qpnIa4l8ckc zhQeC~(xZrk=F&R8{|cC5QSofuVLxy-4`eliHu?0zsDGGey{XjR4j)LZd`_~f3yomJ z6cp<2db()n0X&5O8<}Ct_l8;O`>(XMts8~-IoK5ey+S^1pDJaJ*VOmzMWyxIft1+j zw(E22&J12M!;!M!*AUZBgvia&fnGnP&z+G197>;dM?dnkJ_FE-XI(6TL5FYb8#Mu1 zV@2FI?biVLk=r$f2Hr62Q1dvi7_y=9EUl|0Y1F(05voOa;51jJHTsV*Y^2O(D?gZL z3&e}cKkLB9Ee``c=9k^mtB;8nDllZPx~R)t-bGx6GyHInWy*OX)!i{aX#}trAR1>j z{J*&H$b}&9Wrq+k6qRGh?T zSgq9n`DloqM(0Sa9H(y-KlcXEmVsf2CvgQBzI6bZ@O&n4rM-a(lvEK*(=7z1RSu z9m*UcThJ~`>DH>5?EAO27@dbvG zfxuG<<9H+X{!j&_!!pzC-bN)@(s-eO`XF_x{gB zd?dligwhgWC@8^#BuWD_wSP}1cL~X6!j0_Hh)>5=L`a`$YU|%E|r#hW}LQJVt?95 ziG5aT-beee&L)%1b7idQo3=e9zssg+I4-R+=*b+Iao6UEd&7q-gEBiR0mgrl&X@g` zih<@#=)YNY`3IlF&)H+(8h1Usw94SzL~VdUfQ~rg>wlP zrvnKhLd036l1nqS`#d^_9BNf4%6qqD~g-dP{zFH(O1*E4r>O z@w)62;P`2Q;N!1u__pb;CGBa00Mz6K4T2CSVR1VKrN4bu@A1)$h-Z@YTUv%O&rCj&_wzTLHY<jNBCz;ldr|_v}>p%PRo_3c>@BMn?^P{%ZX~qEb&{G`Rb#lmIqM z-QrYqE;xuuf)+vchRV2er~H)k2iBX%OB+mh7c4@jvU2WcRpuosrPB-#9etiFw&b{X z;9N9%X|DPFd`;%QDv@4(t=1RR;*{PqJml!9-tJ1vT1we|i}xe*7hnrGF@3b)5$%0> zpO-fRFbpaf5YQ1U+oOL@eNCd%sy%krFICQqn&L;)aM7vNpS3wufqmJ&mQ8>V!~=3V zk>CrNk2xp`i!so83Hk79XMB*zKRZP6-DlE@1nf{dpeq#x5eSg*7`U zGo$*jfC~j580#+dl*Lc~cb*tAh4zNkd4!?_4pVnFi4q<}1aKOcDj*K};*T&$DwDZ_ zHfOrO)*{Pg);qn>w;o_-^F@OX0qvd$&P}5sF?4RH0f#SHaCxHWPNFN$C2KuwfP&=` zij3YhUsb#R;k~ z98DAz0N(-|VZl)@PpxCK-idpcz#^IVJHtf`X~J{Ng3)3u`H!P>3X=g= z!vWhYv42d7^Ij&9IlU37gIO5Bhma7AkBjnYOf}W|k91!IFk+K9?dEBA<&tR%PFLG* zY6kuu|B=t&ut~zwMEK)zdn}Au=v)qWEIF4Pw!M+~Tgb=LNH&`XlZB?()0gB>}523HZq zY1Hil5Y0aAPCMf$H!uM2I({z%fM1AW^f+l1rD-}-Vq#(g8fz^Urz^4xv?@u2p+4%_ zso63Q1J>sAVVOiSO!J=jIS~Hl7FNI^Xn2{FQZW;{V-=?NkF);~cku*M5r1|l=>Z9g z^X`wz3inr^$rk5l>?L4clhV@C3hZoe zQW0BoY~bRL#hDJUowK&q)(LSsK|Ne-K%78g?$h`O&uPa6A!pNmo1-|-#=?S9dMs|c z)?|h3G_UQQoJ18mn3<4*mR4WS(gJvtIOk-?GdcfoS^P>Z0Wc{}rSVdPM*0;nqG`a` zG>4a?#yfAo~?R!sLpOOcyr$363~@Gt~e^VbVt$S4$u6kGpt^B63vX^#4Ev zhlGbuResW&G9$7zxjon(n*E$_TwD5l?*DLE@btQ$_vSmIM_)q~s?q7`wA}rYqBRYu zNg8{}Olw&lJTqm=%0M3Yv{KKDd^bds`K^U^B2(ia9Iz%;?F%=h0XCfJ1kxUo3_axQ zt4KR{Cqc1p9;X=L_^mCIpFS)2^so0eJ5te2kSpfRRe%=j`>+*Wk20qcZ*{6Fmx(b0Ff0myx{r8!I2~;Jr&7{vB<>UIJ|VLAf>VaD zzfg<)N@~0NStVW8bdj_)KXP?4z=Yh9)tik#7o!qG%wEWP%gP9os1!-Lovs>fd-LNl zr`1`XXqr)hU5Cs8^An_I-VxwdK(bwb#zn-r45knJ^Mdot^g|lA!8j8GkbgsMR=(S7 z6w9CjdL81QDlKKG6JSimfWBGPqgAO9>3+5>*_pM?<&#SR&}9^m!fyg_X|_H2N;zV* zU)JURP6~$;%C4JPLqQ7_AmZ1GtQ)DksQOFD-RTgK_tXS6icf6#Uj*(QP>kpN3$Ix# z@#T2u$rPBq?@s61-RwTm>c|1YH7(#Is(Z_$R?eldH|GLDNR#B#kaB#+DU0O^=lkHs zdk(sF0@=hDZJlHx-1qi>LNH3XYBhVP9oaIj>J<`#{d`7tW+npo+|bN{&`!3whnWU& zaFAFG9fgFBgjQN#Y41qtop#{#_i!MYRfN;Gy+}tZ=1RbjweL)!onxwMc0_M7EdlAS z;GEb`6OPF4mpf_dI7^2auj}1#m_yQb-+ver-b=33z$6d?ct=0HC1#fbG?rEuBR*&4 z$Zb%!ub#wEwH%1Wp6reg8gvE(_(nqZqXKo1TCp*={(UkaQ%?QPFe|{g=m-zgmJ;;5 z{F2FK&yhqNH4x53Et9SWu%pE{Fk8d1O&Ayys(rTtih2I)kRB{yvsQU)Auh_Q?fyna zCi$AR`U;ft^L;r#J^)5EHh}>KleoVzm6!XGz$s^rs>4G=)rO)<97Q%j4Kvr9-CD>w z&c{oif%Pe^Q1|#4B3NFu3Aw7cMtfB8_hN(a^m4);d@PyW;5LQs-5*VI`JQI@d~pEV zM>q}6DdR9zYF0$zaO6pl+gQy(<52C4glKuH!8ED_{b)74$V}-TPar>;=d#k#EIxfe zO>HlRW%fHoTh?zUd{w5rq+kWhGY0W_r_U9e)3N4-ytVeA#m-V`>`-!N;W@m&E|s7u?ENHdbiE`R>&4L3cZYXW@z$zVHyW2NIxF@&Q(vxCDSSN*r-x;NEYMEQi#-U=Kw{+?sqk$b$RWr0A%bkMm={@$nI{n zirwECFwdbRmwR|15kDoHnM2{mr$1F`6=ox0b3eX#e|>%^FeI|s8W?@g&uqPSNy-Vm z*!Q9nWUFeve4dVw6SVs)@?-Wd?e6wr4ITxrv0u}>ANp5Kijt^&+3V|90^&_JlWEzH zVoR(ay&f)!Fd4LisFWQPi=;3m>i5LLHuciyfI^1i42LfkvaVV9>CkKtg2VD5btInG z$zUDqb8QTB_0vWhcIy!J-ES0&?;l}B!>*VoG>yVyztIhSWQ`ZRuw1@&IFa-F-2ozg z$)sQ_4&1|B&!5i;0pr)ckklEAJHtWLmcama`6w!)U#Y4jE5inWyrt`Ruomr&!k0YI zobN8ZlDg%)eI)xfXTVf-TL-9~S=YEZ7)^k=YBZF6wQN#r_e2k8>)3_7Eo~3}Knz-6 zAi3v_-*s=yM)!L)f3EA*Zh0de^NuLLOEK^5-T!D}j>dWK=a@cs*eb&f*sx;iIo)emlOu%3~5xVfNa zrfx2u8|Y<;X`+P07CO6Pby`B54<53C)baTN6~LyP?q$3Qu66O1F^hJL_R9##tUDve z@Ok9ChYdKgxP(0Qm;lv`&C5JvO2MU|K&^E8hrGIdUlt=Ny0pywY@cP}rCO+3!br73 zEi!>!Zo^QADM%{Ji40I!{*gHk%5^d@&r;_#Yf9q@buzDY+;jTvU_R>{ueSA|^Ijs% z(5e-O{#DSnnV*Cx1!FxvJ2 z=O!xR0d|avdwz_b{guYpw?fQeBIaN0e^JUW6#G-f(eXo}$sj3l-k_BE2C2X=+$3>h z46>jAE0P33ZzRu+oWfA}XfnbfT~#Th>^sk^q_53vGjJv8r_E&L!LMmdIc0d9QRdx{_ObZ)BEiC;p3rq)9`-`L1(O;AL=A368;|Qwu%?~XG-KpraJ{}S?7w&?YG`0AH8C<_9t>6 zmFt~2Y}8-~P$nfcjuVH``vVBREHy^GKFKq=;5i;Eyc^|$l`HayVRBa^zk)bFxba~! z>JBHUDQ5J;MWon#nlp&#x{D^=V?mp3vC{$o|68Q(AKkbPIjESy_Dfr_cSX!tZ8A?- z{7+>|L8C?Vbhj@*>U}i&KisaKvHYGNydvbfz_cqjWP80l z+2o_Yrq9Fsqv9gbnic=Sybofg!1xq4HHVNJpDj2N-TNub}hn>y7kADxR{rCW+ zM=6&)h(tFtqzeM*{#b%f#D!kabNX}jCTesB?TsyKg!8o?7f6>Vw?*e`D||7fC~Q2 z2tpE-#@&CLg=QdXgT(Ivbu2Ph8o;G^9skZUJXe{Ow}yroiOY$Lo?><^;HQU7S}z*? z=I@c9&Gv=_Ia6W|SWGb-aDQWh6|%Sz)k?-lEK;I}8FNb4SuyOt=8@GsW+p zeYf~xHrt+-Tm~C=q5VN$Te_JwUkEEH8a?2gn;DM9uK*}r4v+Jdo`-kGAW-(t0OoZw z3pu>^a?A1M3CeciT<`0{B>`LEc_EeJvbs!!fJ@B5Ksfhl=pU9BCIddjvlZm9vVBRw z(T2Z$>kGY3*%GEorSeS-aIB&F-Yr?p)fovt9hK#k`@f~5`}#iarL|hIQ8OZrw+FrX!3OcnZf&0}Lao7ctBqkw}^r;%dVU?u{U+?e`YA`+ci(cfL1@SiBnd1gX!@0cAy8A0$J1soUCXRq?rq*_9uim=EpUKs@Y5`05``=r zpVu{s+aXK%*Sl5uGUZ>-_%@B4w8c|XD(N3Ojeec%?MX5No(l2TR7O3ruQwF8?3z+n z0KUP>1Hq!`MYzZm85NZT zIQ}A4Eo!#w}*WR)J0;+KpH zcrbzd9_LfNfgg{uzI68m)i49?8*@x0yP*crYh29$E+K#YB<{)CqK>HucuXV4Jr$BX z3=>Tt*by+hld` z#C6NW+(foZ^l`{F{_8!_xhl`0`XY8Qrn`)K znbOdmKhB?buQYPy^hHlk``mCGe22%yu!*xDpgeNgFV+WYR&NGNMHR*{D0KTg&fxi# z#FA`cfp}0V(iY5O#$i$07|H~^c(5@|4v53}If@$iJ}Vzg;Ep68V380C05Xv{3p4KK zPnAqO&dMX>W3dS$kmtw#c^2;|s!+g`Nc9Q+8tDLNM)*0RuDQ1b%oB?ZMs)uqFz>~2&NwDRB?i5JqE3sN+=wrBfZehNrQYc z4QkCctzl=YfxDz$XwqF!06R*#dKqbGP6D5W^$?s;-fMEHf~+{usu8{uxjQp+ zq@gpEO2y1yB~@RDnkRY%wSq**r#-Ek=}87j$k!aQ(@PqW3u7~(+;6jwbKX3(fQi9u z1TE6n`4Z~+?+#I)j5)#%ZK&^`QnnXE=%W(SjvHrUe+(69v7&EdBQK6f@K~3bUr}G) zO0-uF6cm&zf@?2r#}&;uB-e^IVPlxHbyzW28k4^8_lSKO=z?&f+=fX%&y8Z6nFYZy z1FiuDcZWPk6A~`a_OL~h=I%2W1Q?+CeiA)irv!8~cDs8vj0cxJ`@X$0y6*nm$R(?- z-Ngtkt~x@oEEFeEA}}I&T^Pt~%ug-x*bodJI7y$h0ZEGX8iFc6(TiVn-Uz2>f8A!N zkLofn&dUKZTO&FT<63{eVW{4WdYi`u6BDmTIR6bY(7z<;4;V|;R}?Zi(5y&myM8|J z#P1yS1$VW2@+G>T3xHNQ%F+_Zz&uESfC@VGbbIRBpE&I?J5zFhbo@!yV2dZjtc>El z^?_iXjDjM9PRlR8(Q-U~(4=#FLE7v2fs~02y9DB+v&u*uHp9HwfC~P>%=7kZ_G&fHYg0~1(zUBD!X1~8u_!Jh|{5$i?mrXDwC+dKO^m|D# zXr821^4$4@!5Ia>p$7a+1&jGe@8eM_zkb<8d=>xxx$Wv=K7tGR@`f#G!(v{jRRsm) z8q(Wnk%cnki)7I8J$rv#uaW*v22?S0MfLZqYdWQ#*C$~E07`)&qsmne#5iS)=G&m z4Y%W5sTM&`y}sxj4Itx_idb=D$G_ff1qChU@WJAF4K1?pJkT-~U-#8jTexz1^~aFp zqk{lroDx+(9@o>~M`9+o^Qjy*)AgN^1DcPl^yCic%tZ7Is#5d6TbkdUJ}W2*+|l-) zvUULd8UYb_bMPjLff!eAa^7e!%~bSF=>=bW025GGwI&M zT8vyWPE-}*X^G{8^Tj%%=XOib&VVSNizUAAUsdweQEy8nlQohQ8Vc?JhY3J zDY_wDx7Cq}O#5F2ww1@(vMs}YsNOi()9b3uY!^xvM}YPA))vrJ7#j61rWouZ(Jn;4 zbLKapr;zhi7h*9k&ph6o^3qGvLq2`AZI&Z{F^KCunjI;Z{Wbw$N((;s&!j2LZ{6?Y z{93d~5Qw(i?WEC|OxPCc>)Q)|gz0*3mx3f-m)(MUtF+cf8~HFYB?D48SlI7bKSE%? zZO6?x2^=tr1lKw0mkN4I^=XW1kri7@0VdL%R` za8V`|D>VCCy3IMo`euK0k(RTJNN~bLv`Lv=ku*Rr0WPW|Facx#JQre;P*|&00FW>p zKJ-?tq|B9$V*rFy=mDgu+Gw2tzx4gZhE#6Ry#if<-(Zpq;!_6v)}S%Kvfh0C%^zg< zMJ?GeSeE2eH&`T!5OmlNrltzu{*?EHx9-!sop4DRmPG?tiExwBEvwOd)Io?;lB|pb!0m>*-%ok1 zo63$%2Fp8T0r2TKrkyLMfss`9yE)dmR-YGJ&`)Z-QI%$5lt;?FN27U^-DUv2w5@beX3 z2^1T|$6Iek)|@cZfFO`X@~x+xHJYe;#CX%Rb!=C-sL4_>1}~g_d)JQu69BWb`(2i> zKeQUTa`8}f*>nJeN5!WqOLsMAhaYi--ejakO2v*7N*CS#;O$p`Zm9umZ;*&NNV8@+ z?pJToR>dR!-nbr^znGxB&1Nt{6f&6*9I^6)FyH0~uwYT@wk9A?z*m?9Ef#?)6#ud9 zU^Ob8;A3Ga09|s~Qbu5S$R{b1m^k8%cnx&PE5g{NV(+(CM7eKa%uq(+7V3sntr^6H z8Wn8EdIHpr)D9a;AO7f2yu5gMzxje5X_LvNb=T;~E!C0}n{q(~RURNFqwa_`B{YdV z91!yFVlY7`qf{!hK2K;5X7mHU5jjne{C3)3ro!_F^lv%QWyOFw+1T!`6an`cCEp3U zM)x)?nm1wV-93p1db+g>xBOh`hY4S2x1hx8PH)f-w>LKAkqBLM7yfuoR=9Dqtscer zo?)RFVOnzd+^~MTmf4KN_0USfMf&fNXotp-Y^nf?j8PELf#7yha!~Nbd9~x^ChF7G zhzcJ2Fib%R5`QwbGHIzSDHJp`+38nU&s6h~Ef1+x%{aWhHpi|0-GhU&`Rl!318zJ& z)$4t$vVnPnf&l{a6f5M0zcAwJ6;m0nEi_H0(fXMEOz~*B)uhSaKU(4_)|`=Xl(Wos z0H+Svz7p<>=?uDkGW;I4VE)rlBg1DghM|wuQ9S=UgMIeLhF_HtM+p0Sz{d)N$DRPl zvVjD|dA~2eeE+#mg}`o1uXin@@3om`0&Smw=BN)an3+HyGj_QgCejSuP{uolPOn)l z5@Pke2@PQF;{hX?^km-6_@2Lf2HRU$5wqTRhP6(w*^Fzjy=dClSaV zAFJ(J^*otJ^hr}wd42BMT1t|Mw>Q>eaQ zitg0?k#H_UU#SdvoWYPlV?zm#)2TY+-EEZd%E}5pGS$D;H$81yNv(Ik8Q@k6I*^^F z@xQyhEde0e3_yT6;Wl9lOltCia7ecWK;P~4PP zH$N>TlZv;i8gH5-e8;_obUVR%=^y7@?c|l)|DEkPZEDFM*1-uIkZ5m}pTYJR2<=_v zq*G_unrcb}0_=t2fJ|EarTa1L>F!KkIZvoXnbt{&!do8b>Q31iiLY9_*yyR-jG@t0 zi^bzqd#3?*SX9K|Fp|1Hdn)t*O?mVX5FTqx7n^-A2~7nbj#Y*hxNDsEf6M;3*#$!b zHm466GeypE^=ai5$sN2! zY|O$MfGdt(S%gPKEWYS53b^8}y!udmTsF1X@X>7d;PFPuDZH)`zH`KXuOYX#wiE*&VnIY&l&w@erZu#=jF?f6 z@h8Ee5?c;T?9x!mB&^b%4-9<#)09iV3#a!-=(|*v7IS#g7Pci*v3-QRdf6)^b9K@~ z59JLVsf~^((5*A8r6Fn`ZwGPYBvhXF$zOp@=<8pU+R*_gscms1pJ`-%wG);&ME1uH zl1*B$3y)4Gb zMNl#-d9LRz9_Y=yFN`tuM%ldp>?x>I=8Y~a%%zmrSXfaV2$)-VMwx6DW05%nHd*FC zBQAtOo>c;5AupjIA7<(FM$IX4oZe3M>>=ARlfIIJ_|CB8S_#l-hrZuJ;K^8YnXJ>~S2PfwRk)`)N6*7gP;eK8%}hFDThF%5&=jpA9E^5m0VQek-dT|w^MBeq zuc)TFwp}ZNASwb`O06zRQpP(XU`0s_*c2asZrD!upeN)?eB z2p~QD6TkNxtmgGt}9bvQqv+2qXf)D>61V`Fu1+&%t6wQ?JFovFyw#r!(`W3Pv4k(&F>(SRW+y-N)j+M zWXNvi*AtgV!-xI#F ztuM7^e-g>aT)ZNZ@@u5&vH97NVxfl<2}Xn3<+0_Pl-UNI@7lq)XdI%PNvexSrmJX+ zybjAt*B?dVQb~@9!AY`uYqYMg^D3j({JLr*wM;eXND9NN$Zd zxZ%nrS!X!r7t$49z&l>Jdt;y?Ct6C(xI;fpsFe6fPxbV0u;4{riJOb+hbO5@py;JV zW1lw_+zh?@`TJ(~S7pW&oohBM4@wEH#xpV5&azquG)$3`3P^JrIL3f{( zj`|mb$?t(zBg=Wm&>`Zxj;Pk+!1&F&u5@TzowC34E>l3jV`R1KV`u|y6iaT2_V8WU zyj9)&gIg0uiOW919`1DiM)AsF+ z7dSa4Cja{`F;Byqh6*-jDH4=#d&4@2~qg`F>C53^{4@a6LcV zGV^-b4|vIzt=rc)R;b!&zBiHsW+@8rr^qg{oS4n)W8L(#)fVqJ@C|mYGqU3!&5o{S zIcY;7RQLdclIf27@!d~^PXfAuabU#wTvecmdn0~LE^J2xAIi!d4zZeeXvxlKb!j3k zG*MLT=zu|Scm^^87#?=LFV=REZ{Kj8*Vde#AFcqi1GE19Z%>Uo>`|f+Pq~H7>0HCe zs%FPErGn>93C@sMvN~Y2r_vcM(yte*D?Wt2=t3(Evap%~UzG`}AotCnw89PYcQAOp z;R~8)%E_qCw1e-RorG?&+|Aa9WG~=!GVlSiYHA3X1pNAV${vG`Wb3Tn$w|I!HKF1g z47*r{T!D|*w@m|L7cNVO^`Y#hs$f3OA12)|hA>U#1qNTa+tdp(jaKYgmjO2r`X#(0 zP129qaf7bqN0J}u3e2+Xfz?zYqLF4~^bvg2nXG@${};VggQca@;kK#Dd0v$jsrmdj zXn)4+)x-S!(X(TNLKmxE+it<-UiM$I;6910QZhJ(3MvaZvvxIYTI{pq!BjYUCI zu})~&)#PLz)}xX=byo~gPRu{yO=WP*xYZSI8}3}W@Uv_>RTdX{s)F6wdj8H9Shy$A zoFa>Ta5ZggfAPq{h87*4nY}#|C5Of>XDADS7qfdXr3;Y(^O84yYg?OrnGShs*nDgyByku@fuOkZm1vIJWqz*A!^2rm02EYGo;W;n0 zBlHxW4BOwKZnvcRdkY^#GRc7aB z4K{`g%B;5fB!<+4f@FlKm%+DbkFJ-tg}5#5k;j^ws53}@v#BXr?!5T)P<6B5;q%Ow zsS-Y@@pI1G045Hvgyoi9DW^%~e-kzPjKwje;5p{)s{jRTpIY#(wVcp=7EwkCw4P42 z`R72VFT!1bxhG|n4zyM}*{@H`Q@*foS9>*lw35sipniR9fvCii-D(nNp^3Ek>G^+< zcXy&DQ%=L#AD*r*0DR^sjw*Ejh1DmfThZtE#=mLq-_J?lOue>TE$j2#Vyfcv%*dqh zn@^R(kgb{VP=`&e9y6%r`_Il%^su9btsFV0U7Z;s-RPZBy?WpCR@!PIJ4Dpr2K@dY zvfgneArck*(I+L+hk})9fxDxh+F~;+F_z_2UYXG0N@;$UsEhp22Wn~I0#KP9VfV4g zesj?LSb$EbMAE%G>?9~o)aCxlL@BLh@A;|QLX|a{cV_3HyKV@o!Owo@ZUptQ$m!Em zVY|Tb^So>;vZ0)qc9YedjKZqm?v%&EKKo=|W14a{iWs!htpG(@QSZ7$s2}|IUDep; z8$Z7+qu;ijiREriACGJy$(B{xviS=vqrd4l`GVHc~3EawkO$z za#Yp|KuCqf7Xf%GW7=b{kegpPUSv6rUw4!~D6iu)H20@dLNQK&my1vGgFJE zLp)wC+I(OT2@?}8E|k8u(O)wW#j@q>?{DSuW3|}6e`1(8XEG%IL^TX|2dP54oQ8Mwl zWNCJ2reH=q+ls+X|NF;Hn(`6uH8;^N8)Kxt!U&pdjoFuKfM5_CfxcpiYo~Ya) z_FPkIc-mJUB8%Pr* zFz$eJ<^SwSl##}3nKS?i0tLYJqDCW7Sd87b@UMAjW9xQ-6HGr)*S82W}8)PoGb7^T}I%k$^l66H8oKG+{}DV zeZ$Ejqz`JPtyv0)S=#?Kl}Cj8gI@m8uJOC)5Pzak#$}w8{FL8wS?Q(LpqoTI@fm3X z7J)E=>6m+4*$%_x1ON4+2ZPPVK&;lgrHeq|wR*0c#3}7vbY1;ksZJH!wtAnm`?=5i zc)=_6$e*?Wqh>EbG4RruWq4^briZyMU=aqIXCTb(Gh032Fws{D)7PKOvLu6?s3DMU zbNg0u#lMk)3!Gq+hZ4frg9A;EfM(a_Sm_zI> z_86WY}FHZU-z?7Qp1Yp-a9RuoHHf4cDa^P^R;+CacBR!`^$ z5ldr(0BAn(tfiKInO{UbHN1Irbhj_X}}z*)oLfOP11`Hl31R!E_M>m+%5%OoS`l@gYEh zmT|EDuU&H)%Y{?rJ&+44l(<&%QNJukx7Jn#K&~u*z7!jL6^f)4R+WpcwOd?VT*m;X zY7SUZ$!qEPOo&|8#~`}za$#D09U4T*|DG!q{*ieHgK7IYD`T7({NE2rWTddnBDa*7 zAA&IjK3m|%2X@Q)a0?o(%j5x2Q1u2wmv8<$xs2j?$9{Syt4oWNmBg25HMu-`im6)vv1~0*M679-Tc4fjF%BSGkwAZ4`Tmb z$l%NOPv!!Zl*=T!|9{Z`7nc|CNWhU@S&~`6#h}8PM6lj{-IxFc>WPaf;!*fKyL- zY;4{VNRS}z<$7{{IBx;q9(DI3y;|-_2JtlP>frNvzYB2yo4p5nCusA z-4E2#OF`Q&26FY^9{w{^WA%wWjM=>=Ns)g5poTo5iE*htV2N1+9&|Hu;Db)1AWvTg zfoR!$F(OW8VW4meLM=Jvg9kFjivzPjaX@&+UJ{8!ss#QnbppH)yT!;SPh(|%Nq{&_ zJd?4**Jhh*OfLfAnMbGX^H6?D{vHZpTYZpT%ErUPlU~qs6WVq!IT%cEz;(^mn9T_! z$vE|v|J7?61V9SOC}oO?5D*ZkXFmG94@B~}^0@)n@XsFDAS z1~eZlb%+{W7H)`mA60yI*BRzBGh+uZQZ6tRtA7upR54y!st0~dMt~M-oRg;-6_|i@ zO}E?W&C^&lVAOu8&3NST)%U#Nq|dZFE~vG&_0~n2A!|Ek2V`E2fK)9Ih-V_yz5go6 z`50@uR22gdCp23RkCxEa8q0B|bC2C6Ii+N7dXOTO4l?t=l41~+bQRa?2VgVMlt9WI z!)Db}Ny>fucAgqu_3`!|Bkbb@ajKC>S2LKMQC|BmP5%Bwn$^%maxMPURvn1LbO8!lTBlKW+WNU=sFK5Cv_v+q_S*sN424V5ZV^laRbP#dZMr68IMzZ+*jl+;4LMbwURi z{mx=N54Tl72jh`M^YY--tu7ebe`nG_e}c<)LukcuuuTw9S`USMv?aVw_91yniApZV zX7KA@peVkj9^yXhkS?sk->Hm=$XrG$KAHYb|Ce>^X;kAY`J_Q@8iz$0pI1_7r=B#y z;+ct@h+qBb1q?BJ2Km29%XqyJy;)q zNuRTim`W2cA7XMTn_7hA0|lo0cec-meVNK@z0v! zAR~a`tHP$~YE)o2ig=q10$?QoJQKuI21SB=1P-vmYv7Jo8T0JszY!;VfN4iM^nn!N zs*oJKX!4b0q_28~a>NJ~k95s#^c~7)nJ==R5*T!o((~I(sIO0!m(z(lD+DZu1@fHH z_J8!JU}pi2*K=ZSxnk^MpW38j%bNVkT_iC5pGoHC~r7gK&<5#>3pbOeBTm`Tpp};6i0#v(LCD=ZqN1v(ojq#*j0)z`Y*5HN!`F zhP76E+GU&m^jRYy9Ai=@AGkbPRN0Adp2aQm(7Y-l?k!s#`6zvjAbD7m%3+TGXq%>v@@ybe$B+t6bc=9~AUp7!c%0MlU~wRT<414Dt=Fri zoVwNQx(Gm=Wn0fwE1uIzrq?Z75T&}y1Y}s3pj*lF4dJH~u4yW+m6(C{9l< z@A3#txDGL_^n;`yu9SNiP1&g|UH+iRdv`prD@x#qTrL%AdkBWiXQQ$y8_Ib)gWh&C zC^|)xfh*r>Gih!iwQa=v{9sbReaCQVHNMSft=|6K4WTVP;?8c9H$`1}z~^M?e2<;` zG6my=9w@l%%vp0G_Z8`MevVh{92mNBK&|+E?`|r!@0SOIloz=9`c~B>(5K6GUH(ep z>4skH{Y?db@GI3XgyuLBvY-Y-dhZf&>DQ?pyxW>A&q5R^tc06}MZo+DaJ1(^vpcN` zZHbrUYsE5?#2~My5#i%rZ>(F)hTrOFA)h1qC4vH_>aORI_Vl&k@itiqRhYichSMT- zhTPT4J-p|^rk(r9V4Y%Ykgi1g*vsx1=C_{AQ!m_u+Mu;vL3SK_%=LLYSF21ZKRwdn zMDN^W!Y2E>IiYpKOFnVh9Lsgu7%5NWvfJ&nC)av}*g(V{pq0; zK}0lSk_R8>D~I@Jr-s;-Axz(0LL`uw+IQHAf>Io5{p-yZz$eS)3`0`EJdC@Bgf3Q? zdXr_j?L5I~V5ILi+T+uOV^=Vg7XuS4xztB^qyueZg%p*P1(1hADwRe4l%aV`AF(wd zFM+RYoL4}9>SjOr>%jhp=m)hF-v-;6-QC^tYJo9}{$x2?5>XW-V=qWaNZmW}q5md2 zaJdz=8lQWQW-sgdCq)GOgXJpmRNB+$4`04H)yBenQ8BMel$S--|(~<-d z-x0rMJ0!8E>EfDT^>Y&Z8L!DY5zqcea6txU?nl5waIizdCgJihe3I&7Ak*rMVfC$) zMBUFemp^ACFI&P@yOVjc)~`Ns-PBGL#6>g?aS~M4CIE%N?oN7paiu*|$GuAf1D7~k z4J1^`xXHaq)7`Wzm%WfzxMO5(J;(2v!e3XwF~wEr6yyxyP}Vkr+8ac-G7Lhz?APiS z@=6y7$_DQGh2Ty$coGd81)Fmj*5N2Q8)EYC}9Uk%S?x_2mcLqWbUL;E=EgP$M z!W4ZNEAEo@{uzK_*wvan$(f0H&gNbhaN_2gl|Jpj0&cq26Av7W%~!T*(u*FYnj9be zDLXA!vsc$THcd4QP_wU*vuVAkbr;8CD(*vwB0hfGAe&w--TQR>xt@~P}9{b0@F1F>Eo4j0;O_v!rA{zyX9!z&<9k^J{&@}PLXo+VrckGk(ok&4)b3VdE z;b`&W+)&ntiBl2s!JAo=sG@Kccm1h|KCJe)M`DtkZa;;hnnzLv?)eGn>0k?XVg7EV zu`1f$Y#rE($!FWozHR39!(EDOPTf2?^@n5FLfvo zN0-tn=d!ILM)j|VYq$)7{r~TO2!Oj|GF;DiLzjCN{Hwtg NMLAX3vZscD{{un^Hgf;~ literal 20609 zcmd_S1yok;x-L8cu_zS@X_b%$32CrF5Ebc0KpN>T6+}crQW}(y?oL7JZX~6}mMV>9o#U`1ctzaV1LxA<#nqV0;lxGC&X-MB>g(S-Z%^ zVOJ~Jt(V8geLAl&-sjxB^ZKnQ%~xe3M#JXwscgRZm!w^ba7wVw2&=pkHmSr*NT3;H z%BYk!&p*)5j(IJ2t~1K;rct50j^JG}<{;jxf(2#Mw`VV2NwT@ZTF>see|S}=*2#D}rTWfcmtJ@xBebsS^;ruw4D>Hd?rQMMj#!HXensEy$3_r0 zz4JD3Y)t9DJrrW1wKLoCS-Hq8E+r+!>MJ(Mm)KabRE13Gxqct=8-~-(VPt|fX|2LL zsyx{$@UMUt3y5_O=$Rmy2h%9gdsWD^+Fo z8}|F_M4sQq+h~jAx6HF!8&_4R>Ef&FouE+h-Y^nw&Cchm>~+{^xgiHvtT!2$4$mU@ z{%$oBWec7C^hcgSAX~}oClS9n(vsC|KWC|lI={k_clHML{+An>-Y zwT;YOmb7~OpVmEm?6@;*oA0<|`h8;U&cG>Zr`VJfru>2eL+G}W^+sk(#oO=uO?5<9Ph>)e`uTO%54)#)r)cFohxW!D}A4Z17l)3KxQ-Q zvcDv@DcR@dIT)OJmm7%+g({E^bS;GpURH*|+=D zY@@=%!$}qKVRneg&+tHhU_q;LJkaYRZC4gCVZr5|hP?w>WvOO^Uqash-#C=Y(MNfoaf^way)RxOu&1{-WFZ9u zF}~A}JtonW*mI3+m_Av$Q$)i0-MvCaii}qC76h?*Mv8^BapKHr$hcYHI52$=mDuMm-`yT9xqvyN9_%xNk_S5yaj-3#TgZ$_D{# z%-{e+TAu*f&tkDMAH>ysjZVhC)*}6prS&>IBEnLQ$c~h-8?lp-knkcS>rE{#R&v~4 z)QVE8y7?^oW8%Y9ey$IBY8<38v@%Hn0Ra@fxaxPHxsRBcnV-tZ5#1y)I~Q`&JxYH7 zcFh7U1F?Re5ZFyM$>GCcbeOAnS@2GmqPwY^i6YrKo2%#yP|4<|S^Rc_2To*0zy13} zn>Sit5)wROahf>a!%Xg{C-yezjBh}9&s%Z+jZIBF2Aw*v5l)pDu3kf@P8J^dgDO*8 zZJ0y_nzzqPVzQ$5BR0PFQk4e1OVjsfoWk4vJFN%QaH0sVw%b)s$v#0lHVpA4ZRZazd@nZ3b(Gm_Te>#JSZgc=@BjoG`a@!eYwKDLqwO>quc zmf_(=JC?tW=S~XT-HO*?bC&CPUP!)BijvbkyCmy^7K!iANu`oETKcd{xaVEu-z`3& zS-HM$f&KdO?qR;+%F+OPvNYPp*YA5g@=uqVyn7zHd82z}$9ND8t|qK45;o-9=Is`!J2wsy1yNX@hL}xuV>qL*9 zY)rRwVUwuFCPh0uC>*wluvzTyNN&1l+GjpBIy96f?6^H?Wzs8K8?~-A%*7xaintx< zs`OKGK64*&-cf^Sr~2^W!;+tG$Yn&f^+fl>3QUJntQUIcGD?>NXdgxi$ibRGqw)N_ zKI3|LPR@0e*H3K8scHQLqekXN-U~kT9B$KP>Yv3#);&#XNKfLD#{R7F+*#;HFEZ}U zE`35xf{fSr<8sBmX*onH0qc3$f)p>IuE7)BwNVa!!%Q%P-gls&>c@9CalP) zKY0>IAo5G>jn-tW{OYR;`dH58Zk0%(pY3=zDUhLhBCW#we7S^#1hq!NMtmwEecRwM zjASV>?d>dEKIiJVz}I|8QJ9Ig4VU?1KPM3pku@VVpINL_S9i8{oyBA3A~gZR_DH_v zi*x4d?lrzNBUDy`wVsduv5%A;A+a&4=WR~mk;M}-NLw+Ax*QZaiNu`5BkNvRvz%`3 z@V|UJpt1UAR`pLa`GqsKTA0c%=Ia|vlQj6qz(ub+@P9xfKDCJPoOa2T#gLGY6j*AK z_V)+gp13O}RhJ0c6?teJBM+Pe`QKuR z@mlxw7irhwtJw})hX^M!*Dm_PED*F=oRz6C8RW&I5>kW_;{QTaR+UWH{xMAlFG*xH z?EnMmHCOF-{(I~UVti^f+kUzH*SNP=4mz!j`Z9W8$d&L&IcP}aJWlahV=D79FlPt# z+>hLIl5E8rcN_lIe5+tU;Jw!F`jo@*;g0oK6^2A7yUQBR(rBedfUJs2DumNVaGMzU zO!agKv-dyW^9kstvI%8T`s8}p=h_2NHBPJM&0KO)w2Tl{-)jO9{h1iV&!Y20p7BHZ zci*3d^QKQXuNacxzPaet&3|QRDj+a=1%Hvx)2U96IPIBHzw@{kL*9!~5s3f@uMm6V zL)9H50%V^){fypQB<2w|b^~53Ha0f830E90y;wLU`s<}HrC{Q_cW2Yp%fs2R?4%3m zndu7gN3R%of9g>7sOd4(SB&{ib~PfB^c#lYY8ZCvufO816|SABq2aSvWU4hBxIUDE z@(rw|0vpFX*Kt3$aRm8{Vjwm@xtJ@JqWUOqJSVBQ9r&0Q^k9Jl;lqwmLU$gRW(>@9 z?7BUdB$d7vfBYC!`K$R=uxfTIoBxaJQE8KtJi0G)IXTZbJ^%UuLLt7qynG_8cNgdO z+#sT%D-n`QU(4Yeau5Q9-X)}sog7d$Bv?kf;#NPlP)24#T7Lj(aZq}~wuRTuW1H*) zsvI9JT(Rh9lFzI{QX8~Pqd$IgJZ*NWn$16nl;I5H&P&$egzd<@XILwjt0+&Gic!sz zGFQY|vd?+NgYu>4sn69R~*-De2PLIfG>3 zB=_~!G$(>J{?j@nuq%(}^^#{c)`omDE0DzuRAF<7ByOZL8olj=kE!(BLGkI+K*k1R zL&HXEIsaF~rM3|PQpv4B&C;SqQWA+;q{SJZlD<9TC#iQY-E1{CohsIgs!js*VL|Fkn_CD4&KS|XU zUvxcGtQs!u9QbniLh@5S)%4oh@Na7`#p<7w%JW+@D>09t%^JAh@1ad=n=HLz<(*q1 zTmA-nsGK3HYBI!|iotW;xWe-n#fcohESx$s!v(xZPs&t>zL=`XO>e3~_jM`sv8f5W z$F3!(6?z?fXW8nuv&Ve)h?&>!4BBZO5E~mpd&xH+t+Q;Y(3(8qR!%i5Qik5OR29@S zwP-iQ#CJ zaZJ3x-Oqy;@fBu@o^8(i@i$QZg99!p9# zg}Dw}^s*K{2-sd6_zlhKi2arM zSOB0S^fC*FMnXyH(q$54?9y*M6ifvEde6ta&@t~OHrH(LUt|3NjqYk2EhKxHb0fnqlht1DZ``a z?t|S89*aqFG!6s@2fw@bnURv*=N<|!)a-tGpjjNGHeX5vLcDLTpw~kBiEZ?nKgQ8X zhI}b0QEm!heX6Au2Z&JJ^>9nwZ~r^HOPnUhm#0b6bEC(Hqg(yjeg>;!U&A@I--NKK z89rV5c~4qZe1rqnq(AE>^jd)^FM{mW0}EiV_1bu}LJeBsu-?GHqt{9y(dk9VVAaCY zH~Yz(tAR8$zI-r2AO}5_tjve$fTlMrImai%8qA*7EtkrW9bZwV@ zW^NH|!{ndbY%Qj(!h2HcoJ&}z|P4l=ql1!<$iV_@q z-%Sk#+-_kFD3}dac-}Ze9wdum4vGx<+ckTOhd0e=L-0=cKP<>wqyp*yx=XF`6h6hr zqH>0rNK0QdQ2@wxxS1r{^T=hLPz|O@xQKH}!HZE-rKp$2Yn>_Q2Gm>A_vKf}X`}oN zjvlM)pSi!M$a_wVcc3#z5QvE=KJ!nKyK575gwRn&8vEh!-3&@W8`{Q1>8KPx*Mo_0 zVaEa(pzy=9SW4a5Om|=(v>TG$C)s|EW?-i$pUb^HdVs&&B{xRp86`dnc|Hp1nKopP^+K{>F4LWdpIFZ6YxsP9m_s>KO2Bg>%4sG6 z=T6e<_&W}>ByrkL;Hh|`hsJrvN7FHgbSXQ}VOphKP?ZmI&r*`gUxps}?_+UoV z)=|~SD0OaQw;3CU*s6FbRpT&4>G5MtDp8lRx;hg>!|QZmfWE^HF~mYh-)Qyr^+f>U zEnzw&_Ws_i$)WS$!NG~wm$`_&w-$CFbLm>4Hhf?tWiT3cOdKXMDoPu?`#aJFO;w(h z4qob0s@R`(mbaZ2maB6->#GGz!8&S~A=>WN=)v?2y8M+dHt15U#$oS%;j%$4(Rp;Z zzxDMLwLT=_0r6c)kBs`Ww1A)SM)L~J?1TOfSMOVG^(ZVHY_#IPyY7*D6#?=}QzQTIO z)Qxb~8d4@ECT^HdtcqD;p7@mLR-R|r!?ZM9s=%y}AvKhln0Qa6z_??sD~V15AYHhi zZN_pXP*QdiKIk4Kam(V@U1G4v)D&k3#6*aIdLikC#{rxBMntD&*Xhm_A;SyXyR#b>$pH)pYxZ748{cfciWS(K zDCFQ9`*}N6JCtmD-I>35TJ}N82kr+lKA0~b^!;u-Bvo$4>pw`j-{@?<=NuThyHn!q zKu%6zN6A`hM#sMNfIdpt)co(kwwUnW$7vmjI)@S$*YqD6`}=e~4UlnY`fpjKKmH9u znOa%b&uWomOeU~PWa2l}UnD_xUoKlC^4g?Ws!hzL0rn>1_js*T5!y$p;Tf<(wY7bK z^w?OAv?Cf`K+^8@dl=bINd}mEtC?f?yDC1%cawwMTDZ_v!UnhD?L8wRzlaCw#Pw>o zujhCGfhAVJ@tnjAsb!_#==^Amh?zo^<#SVbqoWh3P^*K5SOAyB_jb@zM~4u&WH)1= zAQetLim+H+ z5-zUrHPM z=3SC|_lh+@4!Rz3GP>k@z$B=={rt#XEtasPu|4e1*I#w9&6jj#dlYGrretP@0_4(H z%mijgRRIsb&Jb6Ofl^JkLqC1o{OJ5x{-XMsWtPAp!}+M1x>`r95Gvc zyQrzkoEz)64{XXdFLT@F-ry$}Yz_`!H}~JwAuq>X6{u!2BZm84?$I=`CCLh8tB>T1 zYw)tRxXzA0I?m%po9M~I!xI+JoS2xXs(|e#33Lx2?XZ9ge*x`8gYfC);nKFCzOaCS zIXxf725#J%oRUkwH!hVuQsLbNjpH8+_6^t!y$!6dzWUGY5D0IS4C}$3Rq6WW67Cxd zwfwwQGBbwM4Km=C+Py23_bdmnRsSSFk+^DRn6cO$!1{LiRk&Ef+>qrnz__^zSu zjJ9-S`{Xx)Vu`xDYL=F{01hV_O*-xIVi^E8ik-Y3#xdP%?(x@2PuLuBu75W|gfGLf zH=RqEHZ_<><*=6F@@KZ;VWZF4V9?y2%NYB)4PR0{YWA{zRc;3tgSk^5AY0-VRm7~~ z%4dEnkZ|FI;~JTmEdXMK_kpbfZ<58hd!lMjgKyM^KN&*>+XqMZPS%G$`4?v6z3DbR zjKH(~$P4FUM)Gh@-hOvyzTcg`F7j8&on;Q-;f%i(-C8l*?>C~+fF@Kuf6r_sn)$x` z9FA|Xng3Z2Z7i3Ah2m1U=w&fhd8^sn^6c$>o@7doKf6Wb<^)-Te}Qbtx8cSn z)?TSEnErmGxi~Z?>T25E3xP!+A7j%{11TAw{#jCpEL{%x*z&u!e`f2@AYC;3pHpI> z2gEAvvLz!WRFN@Z7`Ryyb{j9p{O@GQ`XwzAv3q5`bW~p5K=sONYok&4baBaF#}I+F zXWl%>Bo{olRgDpAFXEpGJZI-K(l^S=%KB?3kRr@fPDI>G!Il_sHLP2)2W4fs(5a=Q zwMgtE^--l}Q(SBeWWT>R$8mj9q9sOD0?^|`~>*V;sn%Yc`p2QZ5A zdxZ2-?Vuf{rE;(U!^dmfA3Hrg4Qqo@Tvv~C(|DsK9GS@m5-S$2koe^DJpoziKC+3LGw}uF^oNs2|tE~3%u@Z&Hc9FaIUstvglIo#gX#CAgk&87dCrJ z+7GVd)zn)PVbCz7wHg8P%TOzo_weut=P{E1l=kI|IG0{)cW=6iqe+}vj*iy@Nl7^s z6=n$mhs`fjTOK?Dwa&E@P2D2P8#4KR`>H9mwGMB%`BZ_@22NbYC;mx*Aq29VT*Y^! zXEBjtEy@1je4|i6l5(o5k+MMUeE@Jte#5jI@5WGUf@J7Cib+6ANjiV^o<1;5A-$Id z>O#v)truCI|BYXTA))P)v&4S$3K{r5L=fK?(nG4Oq9UiP{3D?(FHbhnU>*i@m}=)} ze`TS3XQT@B3dr75YV%UeQ17 z|GVE|%CNG2kfqYK0Q2R!nTz62Sctpw40!b9-x~a(sm#}WOaa|G{*7JK9N%81069-) zA7?*IFS-LHGH%_*i-+r>mEAoG8tM0`?7x*fxVj)vv3Pi-od2l(KPhOZftr2jC>>v~ zXz@^1)>24E(8Q@Eh+5P|$lUQ!okXV_2G(ZV@j;uH1;|Z~%f}Yy(r%UJyeR^+?1Cja*7K zXfHgWy&%CP==3Wq!OuXlA5xbkn=iKjwbB}VdwYJE>W0}AJ$8>ab>$^@8g(zVh z(B0eHtJkkqW|v*-M?D&oa<%DK&MA0;O$8=PQM(uXDz-j8z)-eZ_o?X(X#cqWUTdhx z{5tMcPL{Zu0+YeXja_cjp(0big`y>P*IjSGbT#pV0`g`>tJOH^$=ob*f6$0X5Tb(} z7`%=}DBM8}7@zTJ;%p6pX!koJ<_o%L?wcZ)Dho?jlCAhoKpyCpOC#libNy)dtl6oy zfkgai5i-ai49`{V27~l+RENE7ExQ(5Jw7;1l3uhQ^A6DfR_G$JrIz`I!bg%fqOUh!UTN`#TGv zl?MrwJ*WI{{}Em;v(L(}>m>Xut)uv|m>Nd;e*{Qx@Df1RXO^wMn{J6nLT4{X<778X zBSCT?I(=d>MFfe+E<3kJhGk6rA6|fJorHsLusE6uvK(Qa>vw#*tDL9*#BaxbDp;X} zl^Uq{XyKnPIT8 zJEZ#F@uoy*$YH;x@F5mN#OUE~np`#>`=@6x-Fg&#ATD0`_n39ioDJy*#j_7YKXP7U zO>K-K>Gce=ecB#Ivf)ItB7r)dTxz(G8emO4oJoyXR7`D z#~DFtQtM4V;_=ci#S-GlBtFyjTs!xrx#`a-)dHnv9=;*5`g!s|Xl)Y@x1d)z7fW|d zE?ZFwRLn2g*&94pxLW)n#TNMVyO%@zDd>e&V^N;bLw7dAD%w}V1?F`{g*yC<(-Lm-{Ynhi z@x$rf?5KK`%Xt$oM2k4vlCy!J*nfysw7iSes%ghvPG6=@V0`urYiZVr;b5dS#I#U| z)rwk6c)Ws>iUQDC^#v;h%vous;dR`Xw@UOiZ@xGUr{=!Juwn={pU7*!tj3)cd*l6p z2c19(^XCp_05v~E&HgAQZc5+8Di#ce_{bS!MM(*KM}VVj;_xOFE76a;?V2S;?r^(_ zjirX=88Y|D1OE_szrW#%jw;M*wzRtI3W&`2bJsRW@L13k02#Xm_ax-b8Z!n^=k~e% z)+w5Z)1~#2f0i`+ld7-!aajI%X|--;)z7vNWRj$sZT+lwkIHKk;8vyY_TL@R6l_@h zoIrt$S-GoC;t{8|i6}Q}F*uw5nwc(_aPRF4XMJX|9N&EtHBn}pLS`C z_4kw8=byT4AMEw7yH}MUsx2$5$`IVDVU+G7)wbnN^SV>*;Zfhbz3;GIh_)QGS`<&K`Ap~wlcD{6lPVj)}p zYR9v?RRvjs+O@c>N;!9WdU})#jKj|2;Gi<=-a=;5fuOq6+Fg^uf`3nOn0o2A zeE+?RVTx$vY5ge(T1$h43mzou?K>4a!pqg|mDn{2O{f;YPiOFyUWhFTTzxGG`h*)2-eNn8PiF!wAfv9V%+$qg4gIU`cidZ8 zkpMKF5rj*5x$|z4d?qF`@$AJ5P7X;bvjT0{BTrWaRH~OjV=XcH0Br+E6kP^viUYW2 z@|^c=n&wYcIa`$T8rUiZiSkgY&HQTs>9fz^TNB3Cc@0DnNUi@Vz1jh`iWG5H&edz{ z1+>lWa$qO9a2h%M-iSH(d4a-B&Ddbze}gWCz$=0Mgos%lHG~BODSG`}aXl{GULH8EDLFmQ%x8LAOFyVOCob{|RIA%r63MPZGO+Vz_g{D>iH00qOSfB59Sz(vC{cF@LkLxEHCnTjNiLE|ZblF80SQ zWR$k%)aB`t)_{Jy`O}ly@cTO=ube=JM}`1XVA6Nfg{odj79Ls#@M9hW_zzV_MUQve z0$o7wo@@x*+K4$;y)k0^AEyei1=7zlbv9aD2eC z`Af!?BDhnk$q^5c^3P_XNibM|qn`WRr((_L#g~i>X&69x5P`v1#-mm3{>>RA^|NR9 zGC-n*x}JxP>O<4{VEz>q+sC$)U;rP2St5nxZaKqTPy0aRrStXXHMJ4~Kaf^LXWm z(FEDd&bC?T)oY+9-0#QM#DTn>-O4KO!u`+r@9G}r7tWf7jEbtfMAW)wsY-h;Yk+0P zjrHT>tq&vgISKDF3=nZVU`$v-mvjro`nW;519)bUwq-zBl86y&3z1!uzjyBhP5#ch zaevm!sWnYz*k+rudl~#k!FELdFgy(wCDX;T$z^4#f%FgMRa8_mQr`#xC4jaCW7qpk zNN zJmciX_yu26QAEA!gxn;V7D-6sOjU1(H3{g;j(x)khASm1kRI--s}ytb6C;uCa37v( zHkQDZm#l#qf+dHD-5EmJ50dU`iu9vcD)ef$R{I#M0lsHa{suP6L8AkSaS}sb+f6qB zYYU zNEQP#1w)MOEFoA5_|b3H*~~0OnsE*f_s(?^1=x7C38E;yh;-sC#)1_|4(7|ovVmn) zi`ws2b<&aG3)OoN?VtCwli}M8rOa1NlB(}AmlQs-vpyCpnA*3Qp(N>Ax@o%?);R|g z5JcH8#l-`?8|VG%+71?$b4`D4;Fo{aB{AcKJNN!Er^$_AGbTOkoY=rW1U5+Zy}8H0 z*9XL63eq^CP9t%Zr%T!<>C4agx- z+TXm{;?JP;cLa|Wy2ai1#%+b3?nC&8=)Qui*yLSvM(z_5&Y6-M?))K1q%wYgsUMf;?{VmSAz{+4CD? z($f6{c*|cus?m5V;?WdF3hbT~Yx}7B&g#6)ONyUZgZv+zu!>FvN`7F2ZGwM1FP#sg z3h&QxEoS;>&h7-ph&VNRyqeAU>z%ZWUsWgFj^<1(7=4)dfM_bvoHpK3#6ytIKF;E9 z4&0u;z9tUb!x`6%Cw!_8Kc6kpd_~4mcPA}6n+$YtjhTi7Sc1>3$*q9iZsN2oc#(Dz zp^5qJlnimvSnc}qY6OU~5bZ0(NUY#Mc-z-`-dQ|YTL1aM&Ch2YZo>@y_!b+PP1Bye zDO4e~ben)pZomC?0!;5mMm*SLa_!``i`B$Z-(b@ZS=y!er@lVt!$h-Kt&ScP*CP-H zyclwXCmj0(TR7WzsF+ zGQe!Hj>BTP~9!#G74LF##@EvxVgaJBm@oJ-a1eqsynP6mc^ z>G{$ogG|(?YUzXK#eM3I>2=zD=_*MorV>lV3ke1`~BLfw5jNsNHRBj^!7i?hT{TIH<;IQJZ zH!<2dPZOmYV$XBh@CJeB_G?)_f!6?a{@DCZ(pzo4d zc;(Sn=5r3x>vVK3pdeiG{tg>|+nImEZyF?LW{Peo59mF3o)qJpNSj+xA<`$8_x*m` zOgNY8KAB|9qf%-DGr>Q~W33N4Lk))&5N7e~0OY~UZ2+`XIn`)q=O^$Yqh>{U(Ed*n zP#No5|GcF>Yq~61VS9h@yQDT-pvon|rcCJIJ`3F?#VRF$Ob_LUF& zor-lVa95LndJ7*^jfMOuIX(R$y7GE6H8Rf-($g+aZ0yQD_8;m$KwYT;o4+RfYBzyo07z_E3nJv36j5u>^2r;;pjv}4Yx?o9<4CC2_LM! z1{2_IKkKF>;#K1~o{LCXO~@HTH=@OlP80EG%=8{P48UFzzsKDAV`DT^pJQm==rE!I20f2px?d0MLUI9-Tf zuDivEFatrvPqCKPm#7-MXx_E8OdLLzkM^`$sJ*g-fN;{pmeVL17sX}DYTveoEmTBn?WD%R}hwk1;n0VaQ?F)FTBD{CKo6>V;g5By1V6h>Yq z!7$i+iW03u^HqOhtIwI-Y%AhJ8JD)!)T3oltK*n>c8LH;pr^vwx_$Fi|8DdNp?sZc zmr#gZJ~1v-dEVy3&SJ&<%3j<)&#u$!$zHbZKhB*F0ZfHY-h!fL=Fy`7b{5O$R}SJz zpM5_ok)#_gdQt6*+u`hZE4yms41>*Mte}V`xu(WO0qYOjLmrKu*EQ22EBa}c51QNs z-Iro-*-Wdhf*F&+=0#@skcR*eX3Mj;bQd1l-DG34!9b~#$;bCe4F5P`9KDzSIAQn- zad>{~i(~+r7#dA0cIBN97c_d3M79|3_e6;K$__YZ<5W=r?y_k1$#pMP3Ej-^vzd!_ ztoBuU_{s;->OfwYV7KfE!s;^8 zxAkoKV|?7ZI+Dqj?-c&%@VrxaEtVL^872QI8m{(S@~i~YQ;k#!#Y3SYa}9 z8E0y>W;JGmocL@**A-_KnU^-pLt&{Y7_B=0ea~>2D8?nO1Oe0%bMkeZ zszSQP5pXKua5O0B7sQL_EX0z#wxD_KH)mTX>imc1HESWoLd*r8m8)~+37Aj!@)!>* zMcR@n+iLFnWluM2r)4h|x7Kd>;~m9J7d>O17<3>K@aAL8(Pnya z#ua~JP_bcp)ui8yEo7fIcCX#r+vA$i9aWmqjbUs^%gkKaE^v#U#}rotl;#qvMTjT(N@g!?@>cI>H8DiD_sw51jQm z+!bD`yZBM2K`H$KU7~FJg7)B{5Lkr_n4<&PK1Irsl*SWkMp--wz(b|WRFO8f5yhxJ z(p>Md52@ynhdq7$O~wza=M8W(&~>A^L#|>xQl_-j^!`w+p~;Um&}y%FX*hS^KFcB4 z+gONTfNWtfX6E{!b(Ttsc2If2t-2!fiPnQA${0N{X{8Hsl&?b6B3oL`Bg4cKJlG=A z?dEKDGZnKx^~onB#kiF8)mb)e_`j2iKE+(p)8))IJd_{PT5PM*Ce2i~DRRq2oAma{ z)1`b7Q4@SzALdCc3n~|7me$+2^-HJh#pPUMPBEAEJpJ=EX}7KMa|D^?e*TbHT%uY0 z#31iC(JY&y(Oi6x6T!EJ-!-%E8Sgf;arm0we4LW;7(YprAbGxRW6XVZ%3VO`%J80J zK3zBS%GF8sYE!tYS*}&ZN6R(c`>8Stb3z2tv+}mLD6a>%?z{J~j7iDQ)%sNkxWarr&XVuR%bpf1WLg2B_eF>7iY7i7p2GvNz1rHpG@dw6&cg_q7 zV-4JgYNE-vgvu*Xk5A<-DB=_e&)VQJhyBB?V{hu_0!Pn{52J+VVua zAL<6?a~rg=KQRRP6yk=ZlQYca7L6vkW~VcyNKZt%b3!uAcY^Tb%SE*AXp93s%<=p` zxQPF~N7Zq4?B+0(C!hvZD5cokZF5ye-HhLPpaRxyaNz%FQWXRp+O%|uw;V;a%eIxN zo>UIvg0{=uCWXU^1qHkjhTxAyZsh3BW&a2L8u&;gncs3+JOFH^s3TQ>q|Dx7DZkG_ z$OZ&w7^;pHmmNVLYEkt7usGAIS+?+?F4nF3tA71!3~;q758xO9y`}>x%wyD>c1H3E##Pf{0kl5UE0$S8yUw=ULi{9l*9*HT+-sMs4f3A0GZB zaBbh}(-@s9fO3WWyu4Wu`35pM;GRNXL$U_{ckkzR5%A7Xm9lA<7Ek2|g<=R?yUv&L z>60f1%aw-NVfLshxrD|c_u9JxKIAKgXkgxk}g@@aP zl|8`<>I1r+Dz?Pz>XCWjXhq{W$|WE?5k$2yB#T)BCP2sG&S+*yi2&5~%>6LR z?D=O-Gk`q8+uJ+ee&cxykj#D}dtanuMA9K1xOVSU9y_a(9T@9Vz&!p0ts1;` zG3WE%>|NRm4c0_LP!jxCRrT)@>jhI#9byg_^?sCCr$#$2^E>a9j`k>2tP_=k%Cf$H z2!(pvX8Jn_at;T5FMw8NU9@E@zr*2H-&Vxv!95&emUhrQwkD{L72#gL>*CEP?g<{9 z*mNU67XGCGdl;w(bWc-rTKirCGQiM;P46Ku1a&mGcccDCu zu8VtXI^u=0j2?7EQ4H(};O8e3b#V$^hmFm7_3J6HfGmv;{P@9?EX<8A`r&N4#e#x@ zl)St=v^4JP>r-7vQGTLFMQ?B~eF8@Q{lY0Y>jjh44gwm~`eZ}UvWl!Mt%Lw@{Ntlh zs4oRPKmFO-20cHs>!`XQ!{7Q$eJDN?-_w@3_-ksc&YAAN@_+Y*i>XnUes}TfJyxu>0qlsBUscw zbgu08WV$4ot-;(cKdr`5QNdG80jXASS z=6Flv_-;D{8nFiu1cgBkg2dx_`C2VG%Kri7m55fK?r$%3!5;4jdowYHTdc zc{@J}{!giy70RF}IxJ?_mEhWw3OiU8F0B5JK<;YE&h!(ghyD=UOnsnI$U^NWYFdxk z>rM(-w*5||0B?Yx$@WTB7pe~5(T@GkiD9Qy^>^4~1Q_|D(V?&mM=XRZsc4w?XCPeoVAjQHtv|jO= z{LoSJP`egrN%3D)@stR78!fWk_oM2m_P~K+k^~&KA~Com$cxsEbvrjKC%vUwS6~_N zSWZi!Wm;%c?)g!rE9E|)gO$6aq06FFe&ZCxB9jpiD5t@6r-$0Frk_3%)>Q}4Grc<# z^I~}*H>pn>2JYFOlgI`zve&5{7dbaj;}f6(sjiis2!J1XbL?imV796gf;y3AsI!Q8 zBzl;~uT_2GcR2*q7(}T130h3PRg#x~4(j>$X~IWBT5L)=-zdN}#c$Ys4JsQ3WF2j1 zSNN&Gmb42#txi=aHQsIJYWq0)tM*3or%zWl=2J6y{isDvpDqo@Rn8RULS1nf?90z@ z-h8>cmFYTM{iEa&{fzC;Y7g`yd$4_dWR_3k2xZqWb(vz;$Tq%dUpT+Q?*bFD<<%?9 z4cMZ1(Ssse-A|_hFg}N8v}_2VVN-ttw0f=>`{_#?f}^{>_ArnUP@FRzciXcw!&|RYaQ zo_@jFI#sIGd}d$6{5#R~RX-tgIIB$=!XiiE_TeyZraKcP(=!rNbuD~>Hk7smZgd)? z>kGQnF2CY_TTcPjh4c)V0KXY_TDtQ1a0QwdYC+mW`)8`b!v2w93>tA_+YHxAO%*6ri5k^@m*9dX^%J{}W)1 z#*$CQP_0q|%Wexiql(HeZZu3%j9N6Ysx=CzZ|rUgEJ@PO6x9HOs1u$z-5O=2E2qX7 zxKRrB+~AN9`ZpdF!VcZ*dnHfkXMVkK-H!leQC9I+2Ys?jr3OSd?He7^|EOS>x_-Z+ ze)3=h-t?l@C<5`^2(anL;SaOXS>z!nqQ7k!aE7>`%=vH z*+%JAn}WR)EWM>cwn0(la^_6v4e>5rzXKb}rM@;+2SL9?gKax1?d*?WO!=@iQ{@f& zn_~{ROo)hy4I6_P2U0#Q=sNgTZl%94fRrL!z&b@v&MN;XdUR*hRRdyCw7TOGkML?0 z*52`8aixFah$#Npie(EUfic?%hg|>OOK~RJsvtbD8*+n!EjZ z_wygM=#1K9h;lA0)N&}PeSFXIslD3l=;mc6tv5fWqend|eOAI*ZMTyScgr+ML}rAN z+BVkA_7?4h9yvP;Q;9geU0gHCvxbo8lzCcl@uhp8&jVj?QSxGT^wF9~Evd`lUVfS-Yc5-EZmwi& zmAj|sYtAEH^bHMfJ$gg{_T%$JL{qXFuG822LMB}2Pl-Jq)x64GCs}N}LiJ;C z(ERAYK0ZFai1nW9%mZJIGCK^Pd#{gQSzthi6xK7W3)YCKdsU)kjFmAtCz`{rgT zZeHFWEQ6=G6qbL2v_K;)ObMak@wJPJii%)A_YeJH_$xR8#1el$2XHZ(@*^qnP&I-jn6(1e)dvXAs4PIu`aEnAuofUS6iY{m1a_4*ffhj)JXGf_hz4WMuDP*9O#D zrO$nZg{ArM=Be|xJrv)vv$Hwtr!2&2F9d{WZ!7TMw9Aq;MPL&Hnubo@rlTl9f(g@q(sTwJ{D_cph-UQY1U>*Ubw zn^Sng+?cfzDAmr?sC*;Kl<(MaMe$5a_V?9Qj_svEPpD;^7|7KFelZ$n*s6SSG%&HYA;={a~&@DQHnld7?Y(G=Ns7gdF@0m zbJxK}^twT(y1M$+Fz4rv)42nIG1(~Wv^6hRyCDP>6CwE~HZ~y= z2{IvJVRO4?|xCUU#amU%zg=J-|bXYYh9oG1TaeTyl#s&m&lzoTP=RsY4&w z@MBu4YCA0s30z&j3x*Ka>~&=|14dNKv11tFIM zQoXk5yM~5LK@aGx1SlOG9I~JY8u3;c)X!Irxb!M4?2?Y2UX3P4Y(W8WTwI*Ksp&~= zTr4as2%AflWC3mG!?nYt-oU_o&d9)r^2zMt=P>EO>9){g{_A=izxvs>4L;{RZYcAZ z_Gv$S)qc?0(xL@X`q#lOS4wK?hoGPn!4EtNUyOcAO~tV&KG48xm^s~I{`4s!85tQ( zV8df|b^O2v_rby!+WPvZAwweC4i4c^ zEBxwvZEa3{y?B9M8`bu73wh~AKww}P%=jar>({I8qK_w%a&kH&YJSa3&&(*PsZmOb zi)&u_@D@lXA8T9N!N|avn6_KDZY}OfOG#yQcFGhyZy|#le+TAn>3N6ODurOlBY1?R zmdn3{6qYAO*M!=j&B8<}8Wwbw8kH`(kioFf)4%b$b0s7q!l*B_C}aZord&v>a%X4t ziO-+^YGG*^cJK3={H)7We*Rs*Hrb|mrMq{ran78{R(Ysm+H$_xeY)>SB@~|sP~NX< zUc7MY^kw&R>!o9kSl;K*zn{QkvN;1E-4d_*^eptxM=bxZe6&Zs+)d)Z|EskdQX@p- MuJoOZTbi%_7b3Xb=>Px#