From ea454fec11c938ba3f7608697a443e40868a7fdd Mon Sep 17 00:00:00 2001
From: huanxiaoling <3174348550@qq.com>
Date: Tue, 6 Dec 2022 18:51:27 +0800
Subject: [PATCH] update the en file
---
tutorials/source_en/advanced/compute_graph.md | 177 +++++++++++
.../source_en/advanced/compute_graph.rst | 8 -
.../advanced/compute_graph/combine.md | 299 ------------------
.../compute_graph/images/framework1.png | Bin 27920 -> 0 bytes
.../compute_graph/images/framework2.png | Bin 25348 -> 0 bytes
.../advanced/compute_graph/images/jit.png | Bin 72774 -> 0 bytes
.../source_en/advanced/compute_graph/mode.md | 274 ----------------
.../source_zh_cn/advanced/compute_graph.ipynb | 2 +-
8 files changed, 178 insertions(+), 582 deletions(-)
create mode 100644 tutorials/source_en/advanced/compute_graph.md
delete mode 100644 tutorials/source_en/advanced/compute_graph.rst
delete mode 100644 tutorials/source_en/advanced/compute_graph/combine.md
delete mode 100644 tutorials/source_en/advanced/compute_graph/images/framework1.png
delete mode 100644 tutorials/source_en/advanced/compute_graph/images/framework2.png
delete mode 100644 tutorials/source_en/advanced/compute_graph/images/jit.png
delete mode 100644 tutorials/source_en/advanced/compute_graph/mode.md
diff --git a/tutorials/source_en/advanced/compute_graph.md b/tutorials/source_en/advanced/compute_graph.md
new file mode 100644
index 0000000000..06d16ab624
--- /dev/null
+++ b/tutorials/source_en/advanced/compute_graph.md
@@ -0,0 +1,177 @@
+
+
+# Computational Graph
+
+
+
+A computational graph is a form of representing mathematical expressions by using directed graphs. As shown in the figure, the neural network structure can be regarded as a computational graph consisting of Tensor data and Tensor operations as nodes, so constructing a neural network and training it by using a deep learning framework is the process of constructing a computational graph and executing it. The current support for computational graphs in the industry framework is divided into two modes: dynamic graphs are executed by interpretation, with dynamic syntax affinity and flexible expression, and static graphs are executed by using JIT (just in time) compilation optimization. There are more restrictions on the syntax for using static syntax.
+
+MindSpore supports both computational graph modes with a unified API expression, using the same API in both modes and a unified automatic differentiation mechanism to achieve the unification of dynamic and static graphs. In the following, we introduce each of the two computational graph modes of MindSpore.
+
+## Dynamic Graphs
+
+Dynamic graphs are characterized by the fact that the construction and computation of a computational graph occur simultaneously (Define by run), which is consistent with Python interpreted execution. When a Tensor is defined in the computational graph, its value is calculated and determined, so it is easier to debug the model and get the value of intermediate results in real time. However, the need for all nodes to be saved makes it difficult to optimize the entire computational graph.
+
+In MindSpore, the dynamic graph pattern is also known as the PyNative pattern. Due to interpreted execution of dynamic graphs, it is recommended to use dynamic graph mode for debugging during script development and network process debugging.
+
+> Default computational graph mode in MindSpore is PyNative mode.
+
+If you need to manually control the framework to adopt PyNative mode, you can configure it with the following code:
+
+```python
+import mindspore as ms
+
+ms.set_context(mode=ms.PYNATIVE_MODE)
+```
+
+In PyNative mode, the underlying operaors corresponding to all computation nodes is executed in a single Kernel, so that printing and debugging of computation results can be done arbitrarily, e.g.
+
+```python
+import numpy as np
+from mindspore import nn
+from mindspore import ops
+from mindspore import Tensor, Parameter
+
+class Network(nn.Cell):
+ def __init__(self):
+ super().__init__()
+ self.w = Parameter(Tensor(np.random.randn(5, 3), ms.float32), name='w') # weight
+ self.b = Parameter(Tensor(np.random.randn(3,), ms.float32), name='b') # bias
+
+ def construct(self, x):
+ out = ops.matmul(x, self.w)
+ print('matmul: ', out)
+ out = out + self.b
+ print('add bias: ', out)
+ return out
+
+model = Network()
+```
+
+We simply define a Tensor with a shape of (5,) as input and observe the output. You can see that the `print` statement inserted in the `construct` method prints out the intermediate results in real time.
+
+```python
+x = ops.ones(5, ms.float32) # input tensor
+```
+
+```python
+out = model(x)
+print('out: ', out)
+```
+
+```text
+matmul: [-1.8809001 2.0400267 0.32370526]
+add bias: [-1.6770952 1.5087128 0.15726662]
+out: [-1.6770952 1.5087128 0.15726662]
+```
+
+## Static Graphs
+
+Compared to dynamic graphs, static graphs separate the construction of the computational graph from the actual computation (Define and run). In the build phase, the original computational graph is optimized and tuned according to the complete computational flow, and compiled to obtain a more memory-efficient and less computationally intensive computational graph. Since the structure of the graph does not change after compilation, it is called a "static graph". In the calculation phase, the results are obtained by executing the compiled calculation graph based on the input data. Compared with dynamic graphs, static graphs have a richer grasp of global information and can be optimized more, but their intermediate processes are black boxes for users, and they cannot get the intermediate calculation results in real time like dynamic graphs.
+
+In MindSpore, the static graph mode is also known as Graph mode. In Graph mode, based on graph optimization, computational graph whole graph sink and other techniques, the compiler can perform global optimization for the graph and obtain better performance, so it is more suitable for scenarios where the network is fixed and high performance is required.
+
+If you need to manually control the framework to adopt Graph mode, you can configure it with the following code:
+
+```python
+ms.set_context(mode=ms.GRAPH_MODE)
+```
+
+### Graph Compilation Based on Source Code Conversion
+
+In static graph mode, MindSpore converts Python source code into Intermediate Representation IR (IR) by means of source code conversion, and based on this, optimizes the IR graph, and finally executes the optimized graph on hardware devices. MindSpore uses a functional IR based on graph representation, called MindIR. For details, see [Intermediate Representation MindIR](https://www.mindspore.cn/docs/en/master/design/mindir.html).
+
+MindSpore static graph execution process actually consists of two steps, corresponding to the Define and Run phases of the static graph. However, in practice, it is not sensed when the instantiated Cell object is called. MindSpore encapsulates both phases in the `__call__` method of the Cell, so the actual calling process is as follows:
+
+> `model(inputs) = model.compile(inputs) + model.construct(inputs)`, where `model` is the instantiated Cell object.
+
+We call the `compile` method explicitly for the following example:
+
+```python
+model = Network()
+
+model.compile(x)
+out = model(x)
+print('out: ', out)
+```
+
+```text
+out: [-0.26551223 3.0243678 0.706525 ]
+```
+
+### Static Graph Syntax
+
+In Graph mode, Python code is not executed by the Python interpreter, but the code is compiled into a static computational graph, and then the static computational graph is executed. Therefore, the compiler cannot support the full amount of Python syntax. MindSpore static graph compiler maintains a subset of Python common syntax to support the construction and training of neural networks. For details, refer to [Static graph syntax support](https://www.mindspore.cn/docs/en/master/note/static_graph_syntax_support.html).
+
+### Static Graph Control Flow
+
+In PyNative mode, MindSpore fully supports flow control statements in Python native syntax. In Graph mode, MindSpore is compiled with performance optimizations, so there are some special constraints on the use of flow control statements when defining networks, but the rest remains consistent with the native Python syntax. For details, refer to [flow control statements](https://mindspore.cn/tutorials/experts/en/master/network/control_flow.html).
+
+## Just-in-time Compilation
+
+Usually, due to the flexibility of dynamic graphs, we choose to use the PyNative model for free neural network construction to achieve model innovation and optimization. But when performance acceleration is needed, we need to accelerate the neural network in part or as a whole. At this point, switching directly to Graph mode is an easy option, but the limitations of static graphs on syntax and control flow make it impossible to convert from dynamic to static graphs senselessly.
+
+For this purpose, MindSpore provides the `jit` decorator, which can make Python functions or Python-class member functions compiled into computational graphs, and improves the running speed by graph optimization and other techniques. At this point we can simply accelerate the graph compilation for the modules we want to optimize for performance, while the rest of the model, we still use the interpreted execution method, without losing the flexibility of dynamic graphs.
+
+### Cell Module Compilation
+
+When we need to speed up a part of the neural network, we can use the `jit` modifier directly on the `construct` method. The module is automatically compiled to a static graph when the instantiated object is called. The example is as follows:
+
+```python
+import mindspore as ms
+from mindspore import nn
+
+class Network(nn.Cell):
+ def __init__(self):
+ super().__init__()
+ self.fc = nn.Dense(10, 1)
+
+ @ms.jit
+ def construct(self, x):
+ return self.fc(x)
+```
+
+### Function Compilation
+
+Similar to the Cell module compilation, when you need to compile acceleration for certain operations of Tensor, you can use `jit` modifier on its defined function. The module is automatically compiled as a static graph when the function is called. An example is as follows:
+
+> Based on the functional auto-differentiation feature of MindSpore, it is recommended to use the function compilation method for JIT compilation acceleration of Tensor operations.
+
+```python
+@ms.jit
+def mul(x, y):
+ return x * y
+```
+
+### Whole-graph Compilation
+
+MindSpore supports compiling and optimizing the forward computation, back propagation, and gradient optimization update of neural network training into one computational graph, which is called whole graph compilation. At this point, it only needs to construct the neural network training logic as a function and use the `jit` modifier on the function to achieve whole-graph compilation.
+
+The following is an example by using a simple fully connected network:
+
+```python
+network = nn.Dense(10, 1)
+loss_fn = nn.BCELoss()
+optimizer = nn.Adam(network.trainable_params(), 0.01)
+
+def forward_fn(data, label):
+ logits = network(data)
+ loss = loss_fn(logits, label)
+ return loss
+
+grad_fn = ms.value_and_grad(forward_fn, None, optimizer.parameters)
+
+@ms.jit
+def train_step(data, label):
+ loss, grads = grad_fn(data, label)
+ optimizer(grads)
+ return loss
+```
+
+As shown in the above code, after encapsulating the neural network forward execution and loss function as `forward_fn`, the function transformation is executed to obtain the gradient calculation function. Then the gradient calculation function and optimizer calls are encapsulated as `train_step` function and modified with `jit`. When the `train_step` function is called, the static graph is compiled, the whole graph is obtained and executed.
+
+In addition to using modifiers, `jit` methods can also be called by using function transformations, as in the following example:
+
+```python
+train_step = ms.jit(train_step)
+```
diff --git a/tutorials/source_en/advanced/compute_graph.rst b/tutorials/source_en/advanced/compute_graph.rst
deleted file mode 100644
index f107eeb165..0000000000
--- a/tutorials/source_en/advanced/compute_graph.rst
+++ /dev/null
@@ -1,8 +0,0 @@
-Computation Graphs
-==========================
-
-.. toctree::
- :maxdepth: 1
-
- compute_graph/mode
- compute_graph/combine
diff --git a/tutorials/source_en/advanced/compute_graph/combine.md b/tutorials/source_en/advanced/compute_graph/combine.md
deleted file mode 100644
index 37559dbf7d..0000000000
--- a/tutorials/source_en/advanced/compute_graph/combine.md
+++ /dev/null
@@ -1,299 +0,0 @@
-# Combination of Dynamic and Static Graphs
-
-
-
-Currently, dynamic and static graphs are supported in the industry. Dynamic graphs are executed through explanation, with dynamic syntax affinity and flexible expression. Static graphs are executed through just in time (JIT) build, which focuses on static syntax and has many syntax constraints. The build process of the dynamic graph is different from that of the static graph. As a result, the syntax constraints are also different.
-
-For dynamic and static graph modes, MindSpore first unifies the API expression and uses the same APIs in the two modes. Then, it unifies the underlying differentiation mechanism of dynamic and static graphs.
-
-
-
-## Implementation Principle
-
-MindSpore allows you to use the `jit` modifier to modify objects that need to be executed using static graphs, achieving combination of dynamic and static graphs. The following uses a simple combination example to describe the implementation principle. The sample code is as follows:
-
-```python
-import numpy as np
-import mindspore.nn as nn
-import mindspore as ms
-
-class Add(nn.Cell):
- """Define a class to implement the x self addition."""
- def construct(self, x):
- x = x + x
- x = x + x
- return x
-
-class Mul(nn.Cell):
- """Define a class to implement the x self multiplication."""
- @ms.jit # Use jit to modify the function. This function is executed in static graph mode.
- def construct(self, x):
- x = x * x
- x = x * x
- return x
-
-class Test(nn.Cell):
- """Define a class to implement Add(x), Mul(x), and then Add(x)."""
- def __init__(self):
- super(Test, self).__init__()
- self.add = Add()
- self.mul = Mul()
-
- def construct(self, x):
- x = self.add(x)
- x = self.mul(x)
- x = self.add(x)
- return x
-
-ms.set_context(mode=ms.PYNATIVE_MODE)
-x = ms.Tensor(np.ones([3, 3], dtype=np.float32))
-print("init x:\n", x)
-net = Test()
-x = net(x)
-print("\nx:\n", x)
-```
-
-```text
- init x:
- [[1. 1. 1.]
- [1. 1. 1.]
- [1. 1. 1.]]
-
- x:
- [[1024. 1024. 1024.]
- [1024. 1024. 1024.]
- [1024. 1024. 1024.]]
-```
-
-According to the preceding information, after the test operation, the final value of x is a 3\*3 matrix whose each element is 8. The following figure shows the build method of this test case according to the execution sequence.
-
-
-
-Functions modified by `jit` are built and executed in static graph mode. If the network involves reverse derivation, the part modified by `jit` is also used to generate a backward graph in the form of an entire graph. The backward graph is connected to backward graphs of operators before and after the graph and then delivered for execution. The cache policy is the same as that of the static graph. When the input shape and type information of the same function object is the same, the built graph structure is cached.
-
-## `jit` Modifier
-
-To improve the execution speed of forward computing tasks in dynamic graph mode, MindSpore provides the `jit` modifier. You can modify Python functions or member functions of Python classes to build them into computational graphs. Technologies such as graph optimization are used to improve the running speed.
-
-### Usage
-
-MindSpore supports static build in dynamic graphs. You can use the `jit` modifier to modify the function objects that need to be executed using static graphs to implement mixed execution of dynamic and static graphs.
-
-#### 1. Modifying Independent Function
-
-When using the `jit` modifier, you can modify an independently defined function so that it can run in static graph mode. The following is an example:
-
-```python
-import numpy as np
-import mindspore.ops as ops
-import mindspore as ms
-
-# Set the running mode to dynamic graph mode.
-ms.set_context(mode=ms.PYNATIVE_MODE)
-
-# Use the modifier to specify the execution in static graph mode.
-@ms.jit
-def add_func(x, y):
- return ops.add(x, y)
-
-x = ms.Tensor(np.array([1.0, 2.0, 3.0]).astype(np.float32))
-y = ms.Tensor(np.array([4.0, 5.0, 6.0]).astype(np.float32))
-
-out = add_func(x, y)
-print(out)
-```
-
-```text
- [5. 7. 9.]
-```
-
-In the preceding sample code, although the running mode is set to dynamic graph mode at the beginning, the `add_func(x, y)` function is modified using the `jit` modifier. Therefore, the `add_func(x, y)` function still runs in static graph mode.
-
-#### 2. Modifying the Member Functions of a Class
-
-When using the `jit` modifier, you can modify the member methods of classes. The sample code is as follows:
-
-```python
-import numpy as np
-import mindspore.nn as nn
-import mindspore.ops as ops
-import mindspore as ms
-
-# Set the running mode to dynamic graph mode.
-ms.set_context(mode=ms.PYNATIVE_MODE)
-
-class Add(nn.Cell):
-
- @ms.jit # Use the modifier to specify the execution in static graph mode.
- def construct(self, x, y):
- out = x + y
- return out
-
-x = ms.Tensor(np.array([1.0, 2.0, 3.0]).astype(np.float32))
-y = ms.Tensor(np.array([4.0, 5.0, 6.0]).astype(np.float32))
-
-net = Add()
-grad_ops = ms.grad(net, grad_position=(0, 1)) # Define the derivation operation.
-grad_out = grad_ops(x, y)
-
-print("Infer result:\n", net(x, y))
-
-print("Gradient result:")
-print("Grad x Tensor1:\n", grad_out[0]) # Derivation of x
-print("Grad y Tensor2:\n", grad_out[1]) # Derivation of y
-```
-
-```text
- Infer result:
- [5. 7. 9.]
- Gradient result:
- Grad x Tensor1:
- [1. 1. 1.]
- Grad y Tensor2:
- [1. 1. 1.]
-```
-
-According to the preceding information, the sum of x and y is \[5, 7, 9\]. The derivation result of x is the same as that of y, that is, \[1, 1, 1\].
-
-### Precautions
-
-When using `jit` to modify functions to improve execution efficiency, pay attention to the following points:
-
-1. Functions modified by `jit` must be within the syntax scope supported by static graph build, including but not limited to data types.
-
-2. The control flow syntax supported by a function modified by `jit` is the same as that supported by the static graph. An acceleration effect is achieved only for a control flow structure with a fixed loop count or a branch condition.
-
-3. When the `@jit` decorated function is used in PyNative mode, the parts that are not modified by `jit` support breakpoint debugging, and the parts modified by `jit` do not support breakpoint debugging because they are built in static graph mode.
-
-4. Functions modified by `jit` are built and executed in static graph mode. Therefore, `jit` decorator does not support the Hook operator in the modified functions or the customized Bprop function.
-
-5. Functions modified by `jit` are affected by side effects of static graph functions. Side effects of a function refer to the additional effects on the main function in addition to the return value of the function, for example, modifying global variables (variables other than the function) and modifying function parameters.
-
- Scenario 1:
-
- ```python
- import numpy as np
- import mindspore as ms
-
- # pylint: disable=W0612
-
- value = 5
-
- @ms.jit
- def func(x, y):
- out = x + y
- value = 1
- return out
-
- ms.set_context(mode=ms.PYNATIVE_MODE)
- x = ms.Tensor(np.array([1.0, 2.0, 3.0]).astype(np.float32))
- y = ms.Tensor(np.array([1.0, 2.0, 3.0]).astype(np.float32))
- func(x, y)
- print(value)
- ```
-
- ```text
- 5
- ```
-
- In this scenario, `value` is a global variable and is modified in the `func` function. In this case, if `jit` decorator is used to modify the `func` function, the global variable `value` is not changed. The reason is that statements irrelevant to return values are optimized during static graph build.
-
- Scenario 2:
-
- ```python
- import numpy as np
- import mindspore.nn as nn
- import mindspore as ms
-
- class Func(nn.Cell):
- def __init__(self):
- super(Func, self).__init__()
- self.value = 5
-
- @ms.jit
- def construct(self, x):
- out = self.value + x
- return out
-
- ms.set_context(mode=ms.PYNATIVE_MODE)
- x = ms.Tensor(np.array([1.0, 2.0, 3.0]).astype(np.float32))
- func = Func()
- print("out1:", func(x))
- func.value = 1
- print("out2:", func(x))
- ```
-
- ```text
- out1: [6. 7. 8.]
- out2: [6. 7. 8.]
- ```
-
- According to the preceding information, after the member variable `value` of the `func` class is changed to 1, the `construct` operation of the member function is not affected. In this scenario, `jit` is used to modify the `construct` member function of the `func` object. When `construct` is executed, it is built and executed in static graph mode. The static graph caches the build result. Therefore, when `func` is called for the second time, the modification of `value` does not take effect.
-
-6. If a function with the `jit` modifier contains operators (such as `MatMul` and `Add`) that do not require parameter training, these operators can be directly called in the modified function. If the modified function contains operators (such as `Conv2D` and `BatchNorm` operators) that require parameter training, these operators must be instantiated outside the modified function. The following uses sample code to describe the two scenarios.
-
- Scenario 1: Directly call an operator (`mindspore.ops.Add` in the example) that does not require parameter training in the modified function. The sample code is as follows:
-
- ```python
- import numpy as np
- import mindspore as ms
- import mindspore.ops as ops
-
- ms.set_context(mode=ms.PYNATIVE_MODE)
-
- add = ops.Add()
-
- @ms.jit
- def add_fn(x, y):
- res = add(x, y)
- return res
-
- x = ms.Tensor(np.array([1.0, 2.0, 3.0]).astype(np.float32))
- y = ms.Tensor(np.array([4.0, 5.0, 6.0]).astype(np.float32))
- z = add_fn(x, y)
-
- print("x:", x.asnumpy(), "\ny:", y.asnumpy(), "\nz:", z.asnumpy())
- ```
-
- ```text
- x: [1. 2. 3.]
- y: [4. 5. 6.]
- z: [5. 7. 9.]
- ```
-
- Scenario 2: The operator (`mindspore.nn.Conv2d` in the example) that requires parameter training must be instantiated outside the modified function. The sample code is as follows:
-
- ```python
- import numpy as np
- import mindspore.nn as nn
- import mindspore as ms
-
- ms.set_context(mode=ms.PYNATIVE_MODE)
-
- # Instantiate the conv_obj operator in the conv_fn function.
- conv_obj = nn.Conv2d(in_channels=3, out_channels=4, kernel_size=3, stride=2, padding=0)
- conv_obj.init_parameters_data()
-
- @ms.jit
- def conv_fn(x):
- res = conv_obj(x)
- return res
-
- input_data = np.random.randn(1, 3, 3, 3).astype(np.float32)
- z = conv_fn(ms.Tensor(input_data))
- print(z.asnumpy())
- ```
-
- ```text
- [[[[ 0.00829158 -0.02994147]
- [-0.09116832 -0.00181637]]
-
- [[-0.00519348 -0.02172063]
- [-0.04015012 -0.02083161]]
-
- [[ 0.00608188 -0.01443425]
- [-0.01468289 0.01200477]]
-
- [[ 0.00845292 0.00044869]
- [-0.00361492 0.01993337]]]]
- ```
diff --git a/tutorials/source_en/advanced/compute_graph/images/framework1.png b/tutorials/source_en/advanced/compute_graph/images/framework1.png
deleted file mode 100644
index b49791d8eed61c164c8df9d087be7ae34e2c7f51..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001
literal 27920
zcmdqJbyQT*|28^+iV|WW-7PKMNH<6~(j5W=(kKEV(%qfIfW*+DBHhCb9U?Uh!bmr~
z2fy!c-B@?sweEX=_n$j!&DwkR+2`!u`}sVd=S-BAh9V&zIUWcEB2-qA(*}WXB0wN)
z-uw4}l*dD*-azocL&?Ys1j2uE`^ADU248|ev>;`?7F5(`L%E%PX
zGpj$~B8q-f-VZ}MEhkFJc(~nb@(GQuTeX&VTJO*9M}b}sIv6bO3%KqbsV6(I;(fJ^
z4s);4m+{I86PEUHD*ap$P(1F1=Og
zv<&Y6?E|uS{{JQ9J4isKHHDqe?*tYT+%|33}>!&OMhu0+~{#J8CxhIjKB8w-{5SPU=dOuM+Ps=(AKC3@DQMOsBrX+^tT7pU5#O^vzMwYy|4d9&mGFd
zF_$LiSDY*#Isg4?5oSNS!5k+)fS6yrPUqZQmRM;Y_B7v(nmI6pqzT&2bvMW?vrXp*
z<5e05SNdFl2dFz5p*ovPTWV#O
zEc2)YCA-Uy5u-m{y-e*Ub$3}#97BZTC_*v|Hye$_6!D*;lAp)56PX`j#h4mhfvu*t@NfHHbeS-Z8Z$Odd<;Dfv&k>$wxWV+v9Cc=(RtVCK0*z5G~>vN*!?w@uvTqN4Zf#0D8~R?-fI@yAsmj6wiI+9j&YsrsP&
z1=ruGfh@4gW}mXF6-tRUUI%Kv?=~LK{cMKtT9Ou%Is74VVBJkcr(dUCrz`H*VM>bc
zD74o+)_9KAJ(?4`7rG?*W~9D>(}l0)4{^X^s_VGs?+qK~cJm#a6vp$w5!;{&hM5rsAuBmZ9`D3&Fq6p=qo{8X9iG8a3X|jd#Nu_0O)&8YM
zbrC#7J}2dVBi^InJE3s-H}^qghF=@>;h&D?eH*0h1>Ol?2889tJ@8lSSYE97&X=eQ
zx%u0bjl8+I`7)C`eK%)r&v{4KzRT|gp&P_RvC?2Xt7p;TQ-=gQF94)g>
z>(YALuJk2pv&kl6yA==N6TbAe#33EI%?+G&Gtv5Pt@kMt+5vdyW}(64l0I>!V_
zS69wI%E9%0Ctdc3W-d~vdLe%fQ=@g&JLBgyBq#`3IzL(aeJ4_Ufy|%QoNUg1b`Yhw
z8FW$D7aj1|ys^%RduMB0mBJay$w0nvoE*rKls&gXl0?n2+G5OtTZrF#=>V;?K}FHX
zAMo@$h89-_4~I
z8g4d~2Z06HnEzFOD3wzW`7Vx=wMwt!gmY)%K)>N(_7`;`)pjPN1k@Fnz#onh1=H35m-o&<8Sq$UtP|yww6yP>~HDYwN-Q3`m
z;EfK23fZli%u$$EhNx%d8j|nqDN8nRbx!yCWD^9K70U8MUnUYuUnqE+Pto&pYDAte
zv7Kz^OEZryxPHpJo)7e*YDl=5
zP?oTWaD^@s6I?XfW_vNY(|Ws>s*BNaLYwmeCiqnT>TkIHx|pNJZje{2g?4yN)+V>j
zoGc+|S1t}Xo2+c1=5y^Vb`$sk?&%MSPza4U{Q*R?E9utLv+PwH3{v?J@9y@BB*OXW
zj_q({x;fkyo|5z))qP!Oy!8bsmb^vJ|6nlaoZoqKf-@;Hi|g}Pbit;@_IP{{*W%#A
zEAg#1iuy;(Lu5L*3JoDwI&*nzkOFG{xXcb1IRVjAiKpbFBIU-&fDm#cMn%t7izo8B
zLRg>*-~^Y}ev9?JZ~!72(a1uoQF-d;>U8xF5T9*&^0meeV$5m$=a<;A
zNVbmQ8O$j`aVZSrU~l7Q&H5<|9sD}HBFlhFE})u^s0!;_&TS8{xTvn>?J44XROZ%*
z3({H&p?y@Z)Q|Yd8Ct(%)Zxnc=p`@pd~`~eCG_Dkesd|^6P^-_kIZpa`w2D^NuTP^
zM;_J$KDd_0mb`$l>HCzol&CK$J<9$Kd#dg7MwB`pvw5s1o^x$U*Z45|M4|*5^geys
zIU9Z8QXN7z
ziOC9eft&~0?o*=#XjvDk_|Xe<1oQIRm5i(u&c1u}u}81T#-f#(Y}tj#$TUA<#1T!0
z8|ow~g;+xhPcwTNV{`I!Qo
zqRxg_1{{!wE^3W`)u5@aRZ}0{Dz7F!>D({SKU$KK!!ZPEH89PO97(4e$vV*6#LB3-*EHs(!rB2MTNOY{y&p@iU30P~Zs^aVRBaek62Cpd(tj-NVs
zI=kS7zgpr~iyGq&Z^Y~ePLa282u(x)88r8I_H_6j>f)YG37Mo@YwK@ssm-%N*+ojs
z#8m&bOeTzP^%ptl^H39j!V_AEC*PJ^@FqV4dD;}Zn|SMurcgO)Jbk9(cmIC(=M`ci
zd2WPrb4ZrO%31Ch$zLSCNDcnO?a;~94UQ1Ms;shya+~&37&_KFp=ysvvA-qXJv@%Z
z3Rl3*MgFp!_fU!WnsV<#{98Bo;ECF~x%o4;3v6kM<=>*%(hS+c+qVf#=TuCt;QmieL@?C!kn_kDdR5x}tDACr?9Ijt
za&`;1dC!^HfjCWojdClQzMncotBy)f%qNn$2sgCX`ExyrsStMh835p@`+$;#FzuBfc
zdXmvEA5<_(m-nwJgelf^&RY#^RPavhZhIf+fBdG;!dc@
z;FCGBt1S2GXW5NrecnDx1c}Ao&^9SPl_+W$f_Zpp-!xhfv#4|!fzMA3t8aSh7?8xk-u_ohJcwcsw
zWlF?UZj(Q_(w=;*!hNuiPr4@_DNe`3VYv}y=Cnhu?B6YP0xQ=GYW>4k+8`C5N)~^z
z{`@FTrLsaVhA}4AL04N{CvWNSG3}2Ql*9|&GzCRyq%pFxBB-sg!PYQOP3Qg7(l2UV
zX;2v|5;9)kQ81~Vhw4GC;w9h45mAf5z@rFVscp&0@7=qsG?L-BYAPR9m6t8eL^>l&hSF+K`|FlRjWy#f)ochBuNvdFSm!xKI5E1p|5KlDPo=ei
zw_mxeKw_r9>wKUNYA1=2%t_2gUN+6wmhzbj7UAKv)72aY4Qhl4tO4IIxpU7pW>aL1
z^u1*TWJ}9E0D5lHG&IpcH+l7*5!;-dRlHGkFGN>JJPTaaWQ2BqeQLBsaHndev=sZj
zn_tHwZ?`72qiU=X!V>S6PO|VlzRY#H_?$63LiMhH--PA7ez8|b%5TeFMJK6fT6#6jzz1z{d
zCYB1XMC?AKMg6Oa-#-qwGG1zaQBE8+`&1J)fObV6$Uh$Zv9c%vrjD+dKbU|iZnIQv|t(5b}&+|1hdW|aXgS-4L>zH6-P
z=+)2jC)YD*QpK@OdEU68xikChqFa~-N}z8JXlX#vZK@8pzxy_jyC
zZ(Mk1R{#t>6)vd>HcfthTjoV8vfd4bHjYCr%tRLKIBv%YB)gxyDc6vL8|AkyT9Z#2
zfnn7eGdR@b3dVV!GY+kX!{5w10ueQ@m5a<8MbitHfCK*-Ti=*hsm33-iQT7fh3J9F
zVRFLUd|i9wWTq*Nk<)`ZGS!`rMDAn)M2mrgnn++ps%1lU!SmS{5mgI2y#tM=2RwJTx%PBNveY8tzC-j~MJ)&;wZcdD={B-n}et15MzO#{_*mXog8
z^yhR@Cbk`4VkJEi^`rKMiiuLmd97{*Kaxh&dagf}>RZ#~{A&#=SDLAH>qae3vn%EF
z(Z?}I4;42!+$tIuFX)v`x|Jd+32xLcf}V*Kdb3bS%z$OtKLjr91nxy$g}ML2rkF
zMaaJ9V4R>Kdssjy_cS{Iy2>rjGo-O%oFGR6HBlTr`){u0
z>Vm&`iD{9@+`1vg;aOju=^kTPmm~Ct^G)BjktuCfYue|Er;3Vjg0ZpZM~#8RCkNRR
zjb`fm&b0acT37A*U^TM$mMyh}Y^42t*P)C9MVvb@?HjHm#xotSHx^|ItC9I?b8P%$
zw84%(-StMDLSni_^_heH(#aIg<`(Lajqhv-TOfKDschoocBhNV;0baCFn=%|ZnLZ)
znA^5BOjIb0j4WbsW%ZuwL@A#3voZ*qkTUzZ)YYMIk}
z<*NaErt{(tBwD-Pf4QNnfA%`w`Nglr%Ib+38<_*^Cs%Lj7JYRP^49g{j)@9wjT**-
z5|_^VtA*mlH31oTgbX9A-F0F9(eSiWztK$0a>&B5=;s~dY4n&TGc#<5E;nMFJ<|9>
zHpKANTY4T2Rq8YUtb=AL+HmvV;#Et`Tc+DiNKVuIaD0;T>S}qjP5p9T)ae>ATtSTC
zg0sqbqNVg9Ayqx~F;0M<6%8dWr>$fGRv7J2`8)H%zz|BnGwQbdP;RNnT^5D_GMeB?|b91
zaB}>N(l0h*rOIWdwmKM7lTF5qQus9SWU?b^Qp9!1RdraagK-q&0@J`-)yF421C&S!
z#Rp(#v&D4t7V)5v`>ICn(Uf3bd=W^)>)3gpyK81Mg?f@?pfWBFM~at+j6
zzpkh_d1s_f#MalU3xB_o(-uC8H8=sSslSjsddXWD{sNzIjbG9Q*c>Wdyit@C-Wnp#BWQ^x)>Kv^`=9@Fy
zy?bY%|3b)3vt#M?X1x0?My-QJTPzN|R-k25pLZkOZ*b+BmkP8oA&GviT%&T1q}n=%
z?zgeBX3zVU)YM(#KLX$?fSS|wI*R@`#2W0!qC_P4Ke4AMPjYco*B%XJ%K>1IX+&gb
zajzh9+6vASIRSJT4xmeUF0>U7YtTzVcICqMb#$kpi;5&qyO+WF-(?IMJ@T?$G+pp}
z{^4HQtE(f(0)hd}l-=ylS59)
z#AB_eQeAlHW%qmFS7vgaXI0TX#T?k(Urj91Hs!RE1<_2U6yS$Fv7(=?
z9ucU7v{50J-AF`Pq~Efh6%^l^M4^8GW9~Nmf9;vM^~iWS9!tQpGl(;T;&fl5hCv`t
zde*ItX?V>1tHn+e)Iza>*d5S%S@K+gOS1eMVf#}4n@+q%sc1V=(4gM`&e)N$XMVWz
z2$bAqXGl2xK(=XUi=9kvgk;muWLc2l9xta*gyXDc&RnY^vu{
z($)kZxBc-B;!Hh8C<*YGW=&q0Y>O_u>`>Y^+%EMHFc4qr(QPed^gNE5c!(8x@k~c7
zS(RxqTW?7RMF|4oS!yJxeaxrA2Kl~7^eyFq
zt|QUpDF}hmrJ*+6O6WLZ+z0DUga-*9EoO+bfTC*zh5A(dIHvJjU@Aaf#MK!KWQ(ps
z2Fn38Qo0HTSLpA^_N5s0C2@940MP6SOp$?{Dy1RtvyosMth!pKC~4o`Varpwh^-Sp
zzbfUBQkIqv%Wu2zl_g-)L_|b@fy$XtlTT^6M`k*SBBZtMgZh=Aa{-Y8HZUfE#Nx+b
zq^kjKwC;tk@6fRNJ*vIaYQvqvd4GH?r9;vm^251i=%!$|&lqBU0<&X@4fJWy$
z*Und`Cuq*LhA9Hn&u?#I!hNfOeuq($DOB<0n}KZlgGew6npc*R=yaeYXlCr)Uh6=}gmdaY5Pe_z8_B)6j+=6ZQv
z;IBJy`+9T+{hP|ZHUn#ww;?n;h#`q1(T)812?Akk=@uaEA;>s*3Y}A3=9=uyFaHk)>93rRQ(f5>qn0&GmjBgd$etZFL
z2om1Gk#Wh5D~BwnjsG+$@13fL)SzCH%d|&%PEub|G;6FD<}?1~|vW
zvgEO3E@HXW*zCdn1dLns7uzpFp${;px(LN1L1aYYj(dX8qpm#q3qr!9iVB4X@+a>9
ztlrLB46=R7t%-?e1NEnU3tc|1&C%*KBW`m)Z;bZAcawQ@f7+-M
zF?9*>4DY2Ex_&mDl`gT1Zp<^~xz~&5j_ku+Q~P*q0N$aS!J-Bk8cRJ9C8vz=c0kQ=
z!TiMc8+PDXE4?BU*T+E;SumzLO!|QWD|3vQe+k?7<3^}4jobaPAcqpF@z5iZ-1OYMfd&mvbsG+h+Y9NS>RYV{&_u3)6bn
zDQ1D939{2?Qgkw^NMNtAzW
zmU3Q8lRVgIkbpqJ+>e*I7fzh`JM#M3`)_;wro-v9bDdU;K)E^z<}=JJ?8h^3s-3w?
zuwpmO7rBkLFBwnRKjN^;Cs<~r0&}{Lh;tGe}uo^IDC}s;K%Ht81aYKP>8)C(W)0
zxt53@k2F-DJwdmN!B>zET;MKey>Hxcx>^dK)LyFfS)0i9)l7xs%h+~Rzs4uw(lSfs
z{%|kU$_^O4bb0P9az6j4K%4JL$`9;knEl9$*i##gPgcS!FkK1lamHTT%vXL4&Q|x`S}R(%Fd~1iw$>gdJ>7AN%4J^*bnZ`JchOq;gx)u
zFTR8+0?d|}L~!XD$DvG)mY~cHObwg9^W3mgk4$+7;W^%mN%EW`<
zo9=(zqm|exQYI6(GLpZlPNBGUN+Q?|2jT8Ug1MChU}-fUdZ;A~D?qTf+TkRh}
zu1N>7v2OL)|GD0QOeX&t;^*Y%4NP-#h@YtF+&X6dzU-+owRW
z)33Ic$wqNNtd9UpEM4&;`hOz)ivMo!M^;9h943#+(M?`5@m976GH4E&I;3Lfe!l{^
z6ug)DbN_i;-k^cg;Px_?A|O)v?+AtXXrhA8x5`=rPTbZ`&JH%WSL38&XCqzLYhAA|
z&l1-vU3;{4U92hQ@!Fs>yQ-75P}0F1R~Ks&%O>LaAV<+gV%Zj!x#M49L{7U4ErTV1
zUX9Ow^veT+?)fJ>eS(cCbB$I5xbri;f{p5%|Hr7}Wg{7%r_|qinZys@e{nNi(Qj7Z
z+_viTV$H3R>c69X|EvAxG?-r`zVwlJX>_a&53-^-V$FBM{H*=>x)?sCl1K
zo$Dmm{px|80*#$_ufr0onNY+F72}U4RbPZj4>Dk0X%_Hdfxa;Uve8gw=7?BU{+=3r
zZa*Q1{^^Q?#Cst$@xJO0ak7%kcZ4agnR>J)2G6W)&aXQw*?r{xKsv
z*Xp3oB`i46L1ROgB%^l2TwzQgdHK{Y?ZH_67-V`$g55DJ*z+rt;Da+rTJn}d_ABLxAfbPuE%eZ>Y&0*y`*hkrAc@8P;8NpiNzfF$s+m(Jwp0Y&7K8p0KwoB%2%sQ(
zSGf}V)lv7%^vEaQ*mT;2bsk)}&OPSy2Fq%Z#KI;{TMJ3{oLE0`zIvEi;G3
zdv0COX&rdR;~n8Z8_)cJfn$lX=;7=`r)|NIg_el@wh&uytSq_Bv!5ld6JACwukhhC
zgrHyA!@PGuKSf2}aMCkyUQ8J44U9jPs)RLuad7A#o_Mcz1>u~l24jrgq`)edXR6LqCdNiDj<ezc%NhP&LREvL-n%bw}iKY0Y)vP_yCYP&C~z8*bu3sep8HO5-c
z6qKmtpPLt&SuENRh}t)Tq#gf>dPuRTOGV+}UnCKx9j1FwNX(jt?22Xo9zSKbBaV)C
zTeiVBtYna5?|X$pUzaY00Rj0aU9srK9OT4+<$mo0|LCK
zE1^~s#>nu1u~F~Ayec6s4EA|ZkjC%Q8S~!CR>BEtv2F09B$rD^e%9n-c~uzWzG{8&
z6F$t?c%f3w3ZO`1QmLU6{A-LT&2#y3?Y{@t4KTDG0~}15q(aaTxUWk`^MeChLOv$e
zJ0Rg#fOLbFz7_SQ!6n0c-G-LjloOINYS{6%#rz{PwGVnfd!
z5KxVnaYRvnpOFWra*WDhyZPstj-GRSvq5Dx%+LDIGO
zIsDrLHV}tuN3EdKO;Zk;`f_!tlPQH@iI7{>*xi~XNwQB|4W2Lan|#F~PyzFt&wk(t
z{$@ZU6a+_nERNa(v{8lMT~T@zlfN#*9YV$bqy5q0dnaVg*nyQpE+oK99ZBy+AmAW((Zpn!Y-<}CoQvd{Cb(5DW{%#W0YqpT_O?2(
zcj3jTGmIgo$vn$a%3!MCq<~8jE!?RH8g+
zFX}4jp5tckjF(i%cilBpA!d_zrTRmH`p)0q>CfH;N&i#uhp=stw5S|TQoOnqtwYv8
zAU8fYg-Cm@m)f{~2|eJajlFYAo#9hFvCg+P9am>B4ULeSG+esZVe??2US8a-_
zPahU2`NKaz>&0HWs?~LSvgV%_3`C@O+>J8#G8HtNc$EzkWN*0D1iS
zZC`(T#$jT_?iilRVOvcD%0Il7$j4Uzx;6FVYDEx8<E|D4N3S5xCj_6xv`%`!(l-sz!bI9ae%U`a^mI&CbQ2sn84ft01P0wAeF
z+oHWATzN=zwq=ukhq8nSFtl&=uFAD{l>loSka&#jr@Wqy>+_YDHBlpr2zs{_DQbUZ
zSl?+*-5vjFXp}TwQ73NJ~p<
z2HH5h7_S#vp_Ht*Z##}r{twYjz?c6mib3B)ps?toaVWE*jI(oump7FRH5K)~wtth*
zBzsJc4`(q!B$VwWD_IIqvNU3-7TIf#{JM>maqOW)g?=?|GnInu*1UaCfWr0oOX^jn
zMHU`Xvs9T*!-%)^6U9T{
zT>O9Ql#7k-!Irkd!eGOWRh`h2>k+HQGw3v3$Z^5y=Nu=8EZ=EF*-#@_j}o~66}f1m
zQ6|6dm{pM5&oYuO6P>)+l0S87*Jf;OmoNz5+uXTGaFx1pECna`MeSUgZzV+P_+ZU1
zj`P4lbi~PGw+UXGV}&6;6OYsn}5`(zI8VJ?7=GCg9lR%Y)PX8!q
zwBINkc;@Yi21l}IIyx-IRZmId5<&ZP9Rw^opMny
zILBE(z&+1DK#bmRciDAkJHP5FZ>pP?PQW|261`v%6|q(4tX)bh5K|QJ=8DQkGHPbS
zdb$SGXK|;mg<&qw3wR0Q#8nOCf@{pT6Z%#YSYGn;;2!k7rlu>*Z)rDD3V=j`!h5{0LRtraJq
zW6yjBHs@WSQyG5bm5%zDS~I7EcWZktQ&}t0uHI^;*~#E1|LE0+IrWzPJmULDf+miU
z`@FD*G7s0uYz9z1|G)130QmT2yYDyy^i_r(>$rnM>iN}~*+=G=HLs035!5z@5YyNf
z1TYxe3{$J
zAnMC?wG?aX;aap0js>WQTNR`BqZtaaSdqEq!q)47Hbwx;bbUb^D@Sd1D%U|tjfMX)
ziu;)ns4*m&oKTCBHTw4kXo?@=*y3HioHt!m+R+FKtpiA95ECuWBSLH^CnswuN|^uG
z^Emcj2W!_Q>k2TZ;*VIlBF}@H-Xt8+&+fNhrkE)QXzQrOcxNjt&0e!Us{f`MR@D<(
z7ylh*o6ZVK7TN%|jfwg(9Ez!0>ptrn!f+DMp0jUZ(`4QS&%T!Xy&w29jMjv44Sdu(
z&Gj5Uof)pzzxMl{;6~{Gq6_!u?;e&uEIIUhoV2-$
z-B921U$vGW{WEJjf&)RG!yZsi#P~y-3TsgG{1H>tMBI4riq`O$+kFsiEcTRW=S!8x
zi=!s4@m4U9$U{jmOCnkeH=ZsJ$2XzJ=>Ack%DE+a6F!x<@Mj?IjbCtr*_`gFfdh{8#kZWQJXNMRRYp)Rhm`_#8r_e<6y|Ggmys+0X+
zB#R96xtsIeZf5RY>~}d&?P!+#nhqa*2s}Gf+u&`S+gN4|_ohTa8{aOcx7&WH3wAf|
z&AG+Vh~`bb-5vMmqU^WT7#yDF>ne!rtk)63K)Jm2_X{~95`zPN-@v=`{xAUc2?!KV
zaG$b}K^Q*C!sUcD|HY+i3z1d7!Br2TW*(h&)
z@%oKXg?(iwd>hvtJ^qz4Z@Wjz3|U~Qtoj0Z{HL7*7`=apSuL!_`~0_hPABQjc-s58
zu5qdBQ*PsHYaLw4QM1v{7A0RQIg&hFXNfaR%_57p*7B>@261T|556*vOpOWt#Cdg}$OO5%WQVu7>l^TMR-G7__7Sm|~6
zoId7vIeU}$uW{sf9E2wK$Nl%S=NY~g!6&`iSGzGUB$^l-jt;*^DZl?)x=F3JPX?mZ
z0al?FPukL%l0@T~H_(?d5+Qh3>o>JD;>GIA_+zzg)whFD#-uC+zU)^#*Gkl_Xl#tM
z+ITQKsf84W1%>23?<2I17AZZtu%@6!@`d=M=An+)yvpF~3@gEoLyr6H<*GEM)0i9Y
z#TRjR`UWj$OGf9{_P-E39!6y%9*x1r5s&@)KU3Fh$EgRpQF%26H_~0#JNpL-`wwdcr=n|RYx{)x
z138(4%FdM9Em}Q4=a7K6sC1+nQdAEO;a+~MZQg#-++)Kv0|w*)H>DbVDHtzv*_z<7
zfxT=>zPBwjJNk%Wn!y5`8HWmndp(F45ve_$z&%{RCxToM&YzSO{duCS
zm78IalqH&zAIPwoxbZG_lnI`7lplQD>kHm(T1K`ArB=>gXbq1Kec8{gzE*Lp2`h5m
zqRo*j+6;$2^W?SF9^A&{hpScts~Q9qhTmjott*$F%?&{ofo8^-?D}!kwiOg
zs%@Ln*JI-O*EjWJcIwTf)VuNXt;&0@>(3uVK*cu>gYSSQKLRf-hBvLf0{cb;IBz%(
z;7`khbTm;P=H^OiM$>d~u^p0J^>s>dyWL*^6B>u~gj|NdQesB_mlojf-)0qwjulG@
z$;%cEi}upxkjwGn+?vhb{!&ZPYnzr=)qD#FJetNq0uAvIK^G@>LxDF=<-X3pOSv3#
zpNowudP_m8;gH4^zc^cMx8(rV>XTCK8h@yzZ-VuD*`Fs-oMYSPn`)CWY|DPeg*8?F
zzy6$yh^BRo)*g3Mu7>26+2`hz^&GI5=`t%YE>~=0tdyfC&2SZ`;jcuKX2zI|r=K_J
zZP`C9__aP>z>PTWXwDVH-AtZQ4|0&VkBh`<{IZDHzOh0#|LN$fKdQe;ZWi*`EAo2w
zBCf_JxT>P#dAH@+b^y_qy=ryD;Ev`ru!1(&O26d2la2wooI-SM<_$vLa6jKNmi
zBxboLS@N_{cvle=PY1vk+x0yrFI_65I;r!6wHD>O!+k7GE<_zV#6ppd17ZHvD&!QM
znfoRcYmh&9Z`BiNymAGvd@887Gd>N|Nl09_2x?|p3+MQaUBSHR`S0*4|1#5>5*H6D
z2)Qg90~;H_PI4j-y`SR`6{i}VabI8K_+P$ZwFsjTe{sd;c6Um%@Z$N<)qPMlEzo)A
zon{q%OgX8#R|`eq6tV04G-mT%oxl~zI#1wxyzH=w
zk@3<-6Msz317}_mlfNrGIVVrtM@_5$bx)M<+aGv%X=`=99*xu{^jn4A{s0%q@uB66
zXuzJ(QXwX`t-#}_C~kAN3zU8GJRf%hk9RGtG|M9O!9XGf2!HkmbroE
zyOe&$=vbSOVHF7~KV^TxEt?nL!Ai72Z~oBcqij8rCZlv_d{@BlEHWZV1-=!Qr}#Xhf(SRX$5RSBlmK|6Kc8Q2$$*(^|4QCk)lQXU
z(_!142058o)<|NvJ^nol-UFGtT!r`tMjid~*()uzk4+u+Z^$3{`!SVh
zY-+VNWA?AMNzG5Xyoy?6VnkB$>-f4XnO0y!BND4N>OTFknHV^cz(#ctqq1{V<_t%R
z6hxjx)p++C*VgY{q>N6U?G%xvDZ#h@D%_iXsVDm2fPbv<)8OX$N?gs>9@EN!7Rq$4
zE@M;F$mWyNJ$w=-691^H?Ul=u;i;C!nG;+MM`DDBS-|KxB;(SePA)QuS=6U?Ar3^#
z2ka?I6`oWww;#_=z%xF{4VI>7
zFKQs7{DFFrTuaGSbxkxZQnra}qO3{I8T2_e60es75I$VT2k#6Q-Pa_gjV}LmPeG_<
zxO2wlPN-q#N3?9$N8~h$0kb(AYlnyTOft;dD{AB1Y)?Z5Uf$5Tno(T|T+jR)GPVE9
z*A7+D6nQra%cqw@fa5>(x98TX3H);KS%*@SOgBt>myarYZ8eoE$h4wo_OLQAP&JJ*
z*Lh`-%dZu4bjh>$lOVmS`SHP;EQtorsq+^!x0LP@|Ch%a}XHFmTc6
z&tdeyph(z$9HrQf$n9d2=DIps$wSo;-7sSM&ZZXd00A{UrZ3@;9r%wBR*@|lr+R`0
z&t7Fo&-iHqYGHf>pV7$?v++c9dPcVW?hxaah#jf8jk*5hL5THUJ<)bc*i+HuF=BbC
zYngz^lHmvnlf;zzNvW@ae-ntG8Dn0kwfO!URDs*SAPJ?+Dl=HVPuL@VyZAgie+_a{
zou-XbcJ*154~{vNB$-Z2-Q0r9)h<-z3Q{FaoWo+tya)?4326u*VBcl)1}q8@UXwV3
zl4X1*n`_x~?yewK%_svWolRdDn`Q;e;huVod@z6%45T@_z3!Lwb64RY
zK@2Cb&Ttwr|3iZ7wu{B0$hJf@!7!7xdv|E%Ny_DNW>!`}ZW;{Rzc0Y|hU0dIDHDT=
zz{-<&#`agl)a2Stmv42jP+wabjWrJkk$!vEdH?NI*cYe%0WC?8!ulAU~#p&~k
zf9FfwSb!`Wz>-BM|3n|2+z(^C_W_5ZwKQq+wUeufq@ZBrCty|rmVV9Q-MJG^O^m}KP`=Db!mTkbk~O2JW%x-28iMht
z$*K&ib$w-N84EC3GSf`Jlw8Jkelh*T!s>k#CBCU$LSKJ$1>Q6@Fo^
zYC#-(BkDac6rPJ&IB@?u5i>TrfL{e~2W}}Uv4
ze2Umw*LO(5rw=@mI(u`@1X(QwbI5{@IxRBa~VSVWe3V(xyobXfz{==w=$B6so^
z!1jxkVfTUVCNwsSS88Md5Anra7$Tlmi?Mj2&@~1-LrgZ%Vxk-O_E&8QV#R#E4EB&&
zhN$(qC0Jy`dSk0Vu_NEM#OUnNqKBu@!MXN;Hrp^x{m34zYEd-hvJ^1`5}MkaJ;^!G
z3>wpQ^*HX$A0o&0&6^BY)7?G33c4jBrB4LPkM`%sE#i~FZELGL<#$){Ww~*cbj)j#
z45;h8D}mJ_l5zN%j^ReMzu4F|-{$+VRcolag4YsOe5MquAJIOX7uQDDzIx!_=OrwZ
zqs}iqld{9|-uxh{rD&(u_W-x*hCA}*S*%O
zIY%E^1&=GBD&dRuCllKsbn1XY)2X!Bj
za&Wy@b)N~B$iLhdk8b8=VNkRa-R3UzA|T>S7HnI|lzn3nC&WJ}K*2kIR+=Zy#N)AY%hZK5_HJJ*+M3S#JtKXk(#r_&w&w8B|fNN*CEFT+>uJ)yD2pV=L?0L)5z?Z>+
zG(+cJ!iUOfja%H&MkE2Hmq53kBm*F3tz+Vs)r*Ms)#!GtAP1~oy}2)-L#n{B9r4US
zO#z*9_O%&2+WWm)d@B@-3DPaImcJ@_R(fumM$-NId?4GDP&rMaJv#fn2+FM|fyrB}
zF)eiL1%cS+BN;)+K+`DuCaX9tqb#uL0RCFQiskn9M~c^TzjIm%wc5k!N$BHMH`G2|
zc3(;zu4|Ml2qzqw&LjL=#eg?;3+?G0ltK!60#S_BDnM_uoK2L$&kq{o&ks%d%EpO6
zDjvfq7-SfnKird(w0(8L(SO?G$jl-0DG})2g2M2-{YBg72j9`#_CQ+Pj)&y2q*?c%
zpvkT-_KXB1lgaT%uXZQ;zi{^d^uxe}{hrcp%(g|4*>AfX$z}lV$$s`|4sn^AMRI&*WU
zO=@ayJHy)RDaPD}h}jk*8Ao*58lL!%H^=x29CY0sEK}s7!U}lqTM9Pl#xdZPlx3re
zHWR^czXVyHS$P
z8oJ>H5~f^Rm&ZK?_g3J81;?i%xRdcr&(?s~V&jvOT#Y}G<329>t*1k4@~h_sthmX|
zs?$zuE)#VlZlm+l5+0reUTs}}Z7I`ulOSiTyso=3+$_$>WKY4>+050bGfrupL}u?H
z-(|z+25E@aoMqfsk@ZfPm&aK4&_NyBieyn*v?J`0t8)f#j-9YhJJ?q@N{Y^-%Gipy
z_rvEbZOxbfv7U{m6(?9%L3j6}13jB{KXX@6^@EFvX$p7r
zC~|E10qvK507p{<+sJ}tkuOC_BN35#gxRz8v?D6OFbnR?>y?v&;{KYmTT3Ukx$sRM
z-~A>V-gmRW*5B%8QVHC|+iV|_Q|j~BTa@&h1hLefR}%6`>!+JPB_YrqI!vxR9d;{h
zOqo+TW&6vCwSv>zL->%qkMAkRUGi6}9^t)8P}j7fVMFQpKuc3~r+blRDFQy5n3J=h~XVO^5CtRGn4OAX;e(b6djc=GaWB
zOEG1hS#L#5#{^$?3YwvI@{xfc0Wjcr3Rnbpy71YdX$sC$H6|~$%eEp6`$`EFa~r&Tu-~a_Hd|H`
zQoxZcELA_Hh_+kYhMc3UdLMYW2IxtRlsIooE#iLbF*U38_ra&6@~&?QVJm&l5u0g`
zRiMgcw(AWA8K>ffWflq6{Zc=Y&Q5+ms|`q4R?u{}T99IrY={%n*+t=Kv
zuqgjEA?)F!Q~oXAiL2SPzk(*kJFX-=%1F@Tp!j@!3ReDDIn8LGahZ?fb)ZPYfjKTx
zdFMpag}#=Ceqj9DtD4t4YQvjKiDe}&yVy-#TTQK{?&_BiN!<#kW$gB0A)s`L4-f=SEK}Uc}CYfbys7T8aQQ2
z+8j$$U7@;f0$DiXxq*LW!
zzgljK)-SKOd3YN+i&EZ$EmrInl_|RhdQEwjx>SM-GhHqfbib2VcWR}uUN3mN=?XhP
z)T^CySVP8dF1&ZehLA1Dv${8hI97~0wu~;8$N7;V%@uFF$z3?)LRG2l8;^!BKpj_7
zH{>Z43>9W)O~g_e*d&j(hbnYmaUCu@dq*Mq4=0V1fc*o}@^}JeR^-oAK+{?MsqV3L
za?D}+(Z2JKK*?|Mt?tD`U$Wa;6vrR^Qb03k63p$LrM6^Om7HB>_eQckrWxYe_S$hx
z6*k#{36_%EZ!~a6x|<4u-c%Zrg>PZM$D)QX_tqbFr~Adso6Y3!39iNjNt~8|K?lz)
zt_d)671r_wAQ53!V*!Os8hUvR@WAf@$AkTXHw-N7pB$E7wTzat!L&f>x>S6VNOoUjOuzq92P9
zD{CufNe%Be?|m?E;Ju?J4IO2;Z@!@Aqw#K+yQi;&Jqzg)z!e1A?GrMoFw46M-mQig
z?VS`xJMOu;DzUl-$@D_3k^`aGl9kU(Llh;jpQT)x!nKXBxs7V86D&#k^|gf$%V%D(
zv`kpC;x(*;z8Vf;lol)m52|5|5sPw?Ly?@vKB^Pc>NOF0pyVrIj3ZT#$y<`{KKU67
zt&rV*?P8(NQ(#NrOZ!epg$|kiFduuAc$B-d;VI_6@L6qxx5a)Zu_R;TA(?yHECByH
z_%<;>L1@kQ7XA_{eDrF^uh)@Lh{<1Jo>RB?n$Fsbvs(4$U=Tw2>52-xIL~Qz19|#p
z@212&f|otYo^U?!Q~vhg1g@~n-xGHDS?yy3)&iw*@_;K5?YUsk8>6ADHau-AAgnli
zdUq4EJ5#px*=jgi#j~evFNS0T{;|px|*T`
zE5DQ9p(mO6kNgmcbajK55C^#;oW##s(}|zmhMY57;nToT9TD!A(!2o>Gtez44W2wz
z(poh~m#ql59?V@Ue8E2&KB3*>;=mIm`8`q6dM{x;(+l?K-D4aWN
zi!VqW;1EdnFh>31UWDxb3vs*u-~*N%Up7UdIx
za2CIll)V>e@Zk%<%AY7Igz+Vi1_cH-j4^(AjU^f%K|Oih4F4vs+7hjV!$g6tEiv&aL8NIa^Bl^!%eNF(}ye+><$naPUx$
z@cf|+&|dWUa8S;EIz7i3k~uq1e8xP82qZJE6)W^WnNXc%@ktnY6JSN@Fv1Jmxuz0f
zTQf`fA`q<;BK7Y@_AZ@7%~z^Jpdsr1f221h2v}j@G#Lxu44^c|uhjuh?f?p)A1FY2
z3%vff0m6TB#(y)u_#dj*MLpDOsIM;=92|V8_a_i9|LfN;4F!dV97a1eKrC`)C09sD
z=&ga^g;961Ov4ye{}zV-y8+JsV8Fi|AH8Q1jP&yIvYRY5A$$KExFUIOzDm{M%mFTT
zb%dR#XC`>J4gc=~%bIm_0Z7JSR+c6dBC;?I6(XOr#
ztpc5_2-#v#vQA~^p9+J$*>z~4u=%ao#%foL)LV{`cTK_)6b~du7Q3VA^=K`wj5H;4
z=%K0HN$pJ8I&*4%S=Is1`AOn8Dlh9Gd*452ymTM!gthD4S&|CwV(Dc;eM_
zW2P@>z_|c75EuUu;BfNCC`=11lILnLr@%t5Nckl@$<1@x$hyl_2%jir?({o=5w*?~
zH%CgJ4nAEAGa#x$$p1a8lDxDZ{m5$<3iHra&HfjHqK`#q|{ruQdy=JTE6cpS003p;)|
zOw3|~*3(@gQI?QvAF@_s7$dvmWs@$y9wDl!~
zJ%9Q539*%zlbbebQEQXOh{=;o8t=?>EFgUTz($Mq+wA6UvN-!C#
zi-QCrg0?8s-+J#W$)gCVkZUr}nPCcKv29=4Vm>g%d}*7qHxbj}lnQMp3vJ$m^d1LC
z?W^{a7WRPFdQe`Nn77$airqvq6P2}yCfwHA?P78#`hLmv
z#(ZN(Pq4?4e^TvR#NE>>b(FJqI+-WllyG?Ed9)T;8@joug
z^f^yp7`L;MXdD_JzMnMy3}naun(uk|0C#*dq_TJvM%2$}3hHsAbon$!mM6ThP&s|N|FVXyr
zWDHoc8ys`}^*ahfo2$URe<&w5u3Y&9>3zb$ygRAd1bX<|h;3;hc-&uUekVKDgCrvi
zDlhIMrG1#ByHuCyqOH*%-dB@#vW~bv+#D7$V9&UR8>+EZ$ZF8CX!-qJ=WtKB#u-xN
zBEA09l{W%!3ZCpyCmHQ^*$tshv-3%+SdQT~a=ZurX?~Ye!QulO)#O+82$Iai(Km3-
z+&4b(^Nz?oe-*vSf`X?QY8GG_|5V{h@Hffd*kdfx&ueq>9Bq4MS1h|0JH7u@3YlIS
zLUT85)D7jUN!75*pkQ8SToihGqIRGZ73V+f`%A~b3sGZ=afD96Aw6Uy
z8Rdc!w0aU_OAnKD0)Bp2Ame%}R5+h6ekS}Lcm4wWD
z#T)Zkr|gE+R~kR8B!qgNWa}PR-IQ71?fT#)oo;tnjmVipi86Qu;WliFe;VcI+~NiK
zEg$i#b#?E_KO3TIE$E;T$Jf(yH6WK|neT;nN>b#{
z;Jc1nQWaj0rfIWY(4U9vYcCI)c&IL$s)^O#R3*btXOPN_j}ps#Z1gpG&(~JrnV$DW
zDgq8`L$oR_Tmk7@Xvq;rB+%$PK@Dp12J5ZOP;X&!QE?+n8Y(*~AqJmb;
z%c~^T`JBlEGy3!FzM9pRBkasOjIiQ?didH>Gl%Ju^;AQ}7vxOwD$A2ub-qTghJF=}
zD|#Zi;1z@eU!jTnfzc?8>6nE
zUytcrXh{}HTdhWU_76~!*E;y5}v
ztaghx_z=tRfi<(jtbIr75#akqn{W57M|?&gQxFuhdFj9j6aI{yBshUsx0&V9a!2S7
zj$~fZz!$JF1pZ++q&tm4T7vhZGjsA`X{(dCAyObYw#`@1>zSt`te&EhZSfxXL!upR
z0(!86W=e1_cuGG(FyZd*OQQqUYmQEZA#5d}OXhNPRmC&i*>SyoxW9&pT;lwQs)TLD
za*a*>jF(-xM|*WvjCNO_*=6^>A%ZJswOO_!T0UssYm5nFO?u|g#IM|pvilaGWs&7bP3fGbR5$K;`furb|=idwlG-8UC&qk*fafbC-Xg;>jHWeh%UDX8<-pSbd@@<9Ce6r
z8vw4)lfi
zpdloTM#jbMx~a*R#o~m`xr^_&3ymVBEiU0qyqknNA{}i8i*AdJdto{g8K;)JC1R|L)7V6eQh(9BQ$f><3eg5k0
F{{R^Y5Ul_J
diff --git a/tutorials/source_en/advanced/compute_graph/images/framework2.png b/tutorials/source_en/advanced/compute_graph/images/framework2.png
deleted file mode 100644
index a961c2f661836bd7c96b99d2ac25c12488819948..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001
literal 25348
zcmbTd1yCK&_b+&H3+@g#xVyU(2oAyB_2O>9T|;np*8l;61-B3&cyJ5u5NwCRNXMsbLMm(?e5c`iBeILMMWk;27y4R@^VsYAP@`%aIK375Bx7i$pJKgkRWn8
zt{@Q0+t(kISBasECJGKV4t7>Hc2;%{W_E4?ZVmwseu{tp0|ULem|F^{Ny+@XFW^oXZ0+U-5nyHY
z^z>x$i1zx!Hf|
zAnq>q|BTbzjMc*4!okAP%@v?!|1VRBwUe8ZtF_br3+w+r{vRCxR(SvZzkU2~$Kv4d
z-!5F;q&)zM|E;_KEw!tLH^hQf&BE2m-Nnp8+5;HoTQJ3|1_UIX93UmDez{d73()Uj0HkRK1TT%`Ic8>o+
z3XI9z#LeXYli1u$z|zUZ!2}qPjf07m1uMkS3QX}|)dMQ9cX9zb21dvEzqZRuN~*Xx
zS=!hG4_wt`B`D;jC3)ESd3cyPSOAItv66tiqpO>VqnU-glrR`*VX?6>7cghz=jJeF
zGhsF}<>6!I;<03BHs#sqGW*Z_Qch;>uWs|7_s#!*z7N;|h;S48
z|6@L{asTRX0&+I4ZcZ-V|Bg&`3+I1d+1pV3<6;6PX8&>lR}&8lbMU_#ZT^qt{})qF
zYl~N}`k!9$znc1AWUfw@Zk{GC7UEWbuKtsK{&CaS;G*E*VBw&kdv9W9~P5wg;q^Q@+|K>*E=D*q1!VzHY0_09k
ze)1|1$l$HKl(>dh_HmATr-n3fP~LIZD%9tbljOt`$3(6(^#%;V25fc=Ts<+W;*Jgl
z`h(7u+SRUQ-^tsdBW5URd5hdNr-2S%kIC#s&eIIf-#^PMeaPfMOmb9MQBvR};Lpf4
zC63%I3}x8w#Q!!XA4*3)=lP8l1&twyTK@KzYp{wSBQYlA{umCgEU0*`9F
z!ypE}SJ%*FN*89Z4kqACMZyzIO%@$0>aig%=I@P2F>?KF_f&)kC34JYb0xNGfY;o=-
zr0wwK5iV5g*-+Ighq-P=CCv=7v>)nn*=zxl@XF{=a#qo
zzupd^A~c4T^fU1aQ6*Q&gsALti%lRRre%@iZ+LVxEZJLg5jE@D6sF15!WrhFr%cPq
zLC#7&F`#3IyON=geuuCB>K!)g}bkj
zZi_c&^ZLGiXSZORvZ#8nQqGFUjX(ZECKEUR&Po
z{@M)1T@hgyQ9F^Nnn%Mem=*MgsEy`;w|937D;abN4}s2I0ZvL~*saU{(^j7~#8?)j
z>)W&Au|`^CrOiu6TqMH}Ul~!1Q4hWqbMyiDgbB5$Y+%*KV?o;Wy}AquU#&08MTDv}
zKM<1>gOm6Mf1%S9Uk#k3FUlqM_0sZo^B2VgJ)WG)j3l@{z#237z((x^w0`4dp2irp
zEl@OP3pkROP;TMr%eh#LUTH^~oFgoJjmphxYrFCnS~Z`x5~j{D*0~$-a{Zf+M%v*O)bRO_f^b7F$iI95Uz-=?9N2zM&dj
zeEBkLfA!R{gn^Plw4S|(%PrR1b}5$*2=yf+dW}}62CiRpA~g{zENs+^=R#c^iK^yw
zU61j3>nhD|rA$b+V$$|5HIHC;Jzo!BlHGikiH3S
zKcY0;5cJFub39g1jzNSh!n$C;*;@En4^GeV#-N5$TkO;fp;zr0eue-LXew)PmpD*q)J{!l7{
z8XDwupKxT*VpY-_bpQ6FI|P}%0dldgL+1CWn#`)7uU?lMOLz-SsAJlSOTmDLb}T%x
zzegrrG+j7
z&s3QjF?*j&^axh#V5Xf1?p%HoRn9Va&oxrfAbZ#gkN@Dgo<6)*TstG4`;$gVz?Wyb
z5qB$Mepnf%&*zJrHzgazu~V_!j9QfE9UX14#h1d8%J7d{K1XVZJ;Fkb>`m31J&lf$
zaRx5&`aSjXSynGfRk)lNm%d!1o&2PhES(oq0S0wPzJHW6j
zY0{mc>F$yL#$gS%Cw*R|r#PX{$DdUe+`hv`NI7q7vhzsjwc;3mvL;!U+hx7fOtyzS
zg0C5Nhq%!Ff=J$d0|vho<&#=nb(BR0*V9%$Cx#$T}u%B^<4dB|Z7xdtbZ$Zi3rOEaY{n={*#39{rnnWjoz_9_>2q5VWay
z?@+Fip7&CE!$#ZFAAXmpO$aQwG^*l!@?YjZ%|f80=5;(2NuW@o@a9?#=x8Gi5yfuj9Ua4V+8oGxwdMlvDjM*Ti4
ze_g_*2jN^C&V^!v#Z=&8JE$^923>UnqZ0F^ltxTlP6akB)ELgt|B+HZUVl>|b+aA)
zTfW~Mk(aRJFXoIX+O1p}RliP=8%j&ZrFX^{uyBZ;)%L`pa1!OeP6H|>e9-8^R#-Iy
z?d7!zr1l4H{lwA`TwJ8OHE^$dB4aTEt#m1cK!sO{J0UpT?!T`r=Tx7H55@N8U6tF
zCR@gYm+)qA?Ht2n0AG}TRY-oKaVHr%S;Ea>_4N<%YK?y0gQC3Q{(8KwvuKni8}nDZ
zDRR$@Om_!XWC73bvEIq1D9jsB!63wcwp)!YU>#ZuVnFZR2NB6=}23b6tm-KH4
z_#=M~^j#?9;*j(aqL8d~pajhu`p2ol2)g32K?@&>1}_%x
z8%-6c`fJpwhOY$(4}3e;`*9dIYvY0tXPZjFjnC9xNDiz!kwe}_X{Fh(C@NVpY)qeX
zIV{E^JBw$J;AJrx*7ZI6@2P5r3Am4)Eq~jxH)svcye3ww(0bnoLpcm0_9WRGuImsB
zYn-#4uJq;>f`U+Pjp|!dzj}m!BJwb`iofGzi6CNZ3G+(N9(gN+n{Tzp|{Gx59DR<@Aak_5cS|T&DZSDj78`)M6Zy~NQZ%D
zeJ+#&_c}JQIgQSDUTiiMSD13cOgB5kMK?RTRei#7aA(2j~Q9$KMH)D
z!-WLn^JoJl`tclqIm6zdL+C~DlijbqA?m6~M$GsDXm-l8Ka8hm2X<3C$jvB&n2=&}Xpad!rul
zpfC$~SRH0scnRk|y2g54>HL1YbI?F-6IKDg5qZ8MHuOJw{z$Ah`kp;G&f`3uQ2E-mE4RByH6nQKx*<$N`BN??XUR2v3M
zv0(lc_)LQf=7o9H)k5U+_$m=w(K=#oBrafK-6N3+VgX#mrvWprE8cXpq;cO{f#Em^A1zq~@e%a~vHI*wLVLdGtW2w(Kd|epOz)~CJ
z;h;NqbIo}0!p)i8vQku4w7+N_k{o8?x(ib2me$Ms#3+6o9tWi@
zqq^!Mj)zNd6+Kvc^=C60%vh}dne?KG=V*R5K6(oCe
zx;tS-KJ+FPDQrsQUdZUaT%(9bQdU&n%bZJ{=3Bzl+81iUn9ANCB`+ADUrcuyGS26Z
zr%7eno%Xbg?JCSXvAjdErYkXKbOs!e<5$&>@8jdf@Uik#k}1*S3p}{QaE>ths{-oS1aX;vPLkgnD$KtQI
z!gucBCHEh0S0949vpo0NePqj^ZYpI9Sb_%KIIi_68~H0!|-zLmq^Cxd#!N5!dF!A8_(J
zKd|cLGvRRhCX)MkuuHZ|N%(H}=>wGYQmOFVM+(|QTS;^3SQ}KG+4Yixo-}2|lC4_7jWcriPaE2w2H)-%SGK@bX@RMQkCiTY~jY^wL
z`fYvC0zpAQ+E3j2XpZjpat;g1bl)HO`}B~`aqj=?F(BRcI%}tOd9GIb#t_rs;g7%KLW1B>1ZiI^U;@CN_&L3PtQP-l1H#0}T{o^(+OdH@OFM%*;UT
z&$g1|i4}?c<}fIlUYCT`!t)3uI$sxF2ucwv%QXOu1V=&3Ns3-?ag7Y}K_!z|OS@jk
zf28n=W>%wBF}g~lh<6=w+Z6(%v;E3J7OB^}vxJ;s099|8`C}K
z(IuG4k?)Cswh*TF!#7BKoHkD|W*L<>Ky;`*WeDVTRWXDF^QA-riCO6HcnTb9MPdx1
z{Y$`41bYFitvUG_>&13j_@booD~*7^o04CD(T^dJD^igaz|udDM&AkxWFlKGB*7c5
zZ_tv%gwE|=HLUTUKk%!iuz=Wh!)V!I1l7YdJ&$6Pp&%#QNH-^ZGE%ARYXh+v
zmyw0xuwpGs?u(t4xKUP0n@DeEl`$$-3y
z-c#3oY_(o4J3J{z%SM-;m?*}P!-Mz!C5f^hLEjDA{XjIvnsG9?k7Ealxp%+uGz!Dew
zyKW$KwQfM7?A+wMh$Z6xWHXt=H>js9Eb)YQTrpSg5-=_sg*+#RaHT7EDIuqCQ@FJy
zr>xYQn-wz$Muw7>k;XFmiYy;ZSL!7m7V5%{h=kf}+mP&b!sSk^`TY?0_?PXE$Tw2I
z#0>a~vqe!NXV=khPrXLi`8_WP77ddce)vc6@x1Of^^J<5U67jLIh_K}STs%F4&QW_
zi3|CuKOQi~!6*5bD!iB|dhJlFv1IRspwl*Z4)K6|&2VKWQ#SmQ-`9;2ZMNWC+x(efKgWAw4%$eLC&PC9Gr^1_UcKSN?Iw=bQuWzG7_
zgP|?rr1iY^;OSuTQd@&h+UmGy>TxQf`1!A9Oy|p{-Bw%CyLPbI9+VnsVg|B>DXXh#
z)HV+-KYQ5ogxyS|I_)o3y+Ky>fS#kE`%#COY0+xZx0>X4B`q^LfU)nWeX|O>&S-1(j_#r$vns6!b
z`eck+G37YsTv@7g5WPzi8rEt9bXZt87?n~9ixb}qi^e97GasS!am(mWOmRcUG!$%z
zcJXf{fMRoUR+wAwSJ=w=i32W#vLI~2O?M%fNz0%y)w3U?!=bvk(bKNO5Ri!fF`N>K
zh}^*iuFYHx{z$Eablc+zv??WhAx7|G*cH^Q1fdZ!BRor42wVQU2_FzsP6q+83O`<@
zY=0OD!q^@wTs|<IL=OxD(UI%iWVARO@`@deKJ3P-lV)DnbY}
z`+F;@CXRv&rd&CVu9!Q##CF_v__9q;F}BN0^@a`?W+%8ykH@TO5k#j|h(kS^iqLa>
zQSW%JHVQhQR>9I1uFD^`*&%YT*Bw?4D!B8r#vtaU$(Lx!1QlEDL8kHzJ7138Dta9t
z{r3Jtne9QW5RegE*%lcEeM-&~`C5TJE97MHa!2r3MSF3b|A
zKmH&y5j2%f8E+o#Q2A1Daow>_C4CMqU0b~cK^}VyPssSP{mYo`g|hp)p8h6hxR3>X
zdwX(otI+Od7U(X_@SV_Pb+bKNm>&8!IT2?#Wz8YGpS5y0`tiuR$g!7;6uGYZ335+k
zds71_SX}&NKN~Bm7*vC2FBb9Jzg(k>1~8_6R#;motiUz77|Z;ZMlU5xyp}@l=F>%l
zAap~&=-a^b{tR}EG5;bLPJj;e&d^YWPKx4$85m*;_Kq)dtAu;Ntv_O&@o&R?|e)?8)^%B5BbWK2t*
z;izTIZ~XU>a=Sb+?C{4a1UVx>EP3JjK*r%GirX!mkCUmJN0~y)(AhK!aqDFWRmP0*
z*VMf5oO^x~2hXyQ@BX00^{okanYS7yU;R{E8;Wb)P0SCr*?R30p^>uraUp-T0arxH
z12z&q2oRF(Jnf&n#v0TQu1F%S~?)i~@3$4Sl1iIm_zJcT>Us-u^%5n{72h
zxOz5*uLI{pz+Z#$ARWk+@LyMJW15vy2sv=>Hky<{EJOk~
ze}+13L|u;7#VuKu>%%^1p0eCZXW)ir1O+dz~(p
zLGANCl8uHDH9(Uo_X^#3t@H_V1v*x>37a|!9<@K4g_86bZLAoS`;-gMhf6R`b-y49
z@;HLuQYV$k-b}o#SaWUMIDIs?vcQ&P2ljgmFnX;$CS$v0Q?zF@d)O>rOWEtbNtVIm
zE(Y-MzASBGy}KBxQocNID?>(`-S+w>QE1ZKJ|Byxz6n_Uk;edU4J}UO@1RFk9*+2WFm>
zi{>3p7^o?uCn#u)tO#1G3O8@c(*OrsQxgdZP3E+G|IJ7v)!>$yKHZD0Qn9o+>SUSZLbeEJ&^C$o8RvK&T(jKlKgNagH3Jrao>4_Jj#t2=IZ!a9D$vjY@=+8gS7bD`8R1p;5L?7|Q!YaGjEPWbDfLoQ+|1U~
z(HL_7fnm(2VR#OXY{i@_5yONCq?z66PJdpsqa`JCd!Sr-WBzlh$y%ch+3IIcCQ8W`
z%z@_=onNLg2@eWUTw3F&DvtGMcE5|QR;h3V2qivOCEDj1C7`lvJyXNGF;Fyoso2z(
zKS#=KGh2Y3MnX*1nv$Q0Vaw~-J9DaJepOT?HwE|k@TMi`P9u^>a=qIeNcYuiatA|j~d=WTI&VhOt
z9&=jeEB(MO0I7nA5Np(V)qjr=|A?gK4-rwwo~aUxeEI9wb3g)Vk1@`lS`uo+e`g+POyiiVwes3W+rx~Z$Nt{8v!;DR#R*}#H()O}+Qc^c4u`QZFFUjp&F
zK>*A_+U9J9dMuf6W-_(H&q8UGOkO9Y3=WIh$_WixlGdODo>=q?dT@dC0fhbY6GFc!!Y!ezui#1sjn
z2q{+)Sa1ukOtA;@p+a8t(OMG)H%u^`aI!FU44w6ZpuiZX)RvOgi28h11qyMnN9{b{
z?kp%@4rk)9MVwFgNEbR2Q<-a3i!d8Gz;8C+CzgDiK%f*roEJ;RZe-r%D>59GD+(NE
z@
zQ^)c!57CeH95yZ3B5#^HO*xCt?)^tghTc3LOyJ6mpAD_uACgipJubp?x7J;pJ5e(+
z#ipkd#FOy%is_LRdQs4imQ%6Oh%CP3UiTQAAJ*R`J0M^Z
zeqJix=uS}gj~^+{b$%RR|;9e>#GxKmHtFSA&>ze%&WILbfiQ?+R+p?wCQXU_Mmxbe^qrn0K
zL}fE0f3FX9Iy{(56h4=l^oMiV{fdo#mPkp?>K|k6a!blu!8RM>*yu;s>GWbxp;g{}
zemEVnY{DeL2^}%tlu*NxO&}3>VpFNnal^NFI@X?XIKI8)Lh<&Y`m5Ez+0kS{>!c)1l}z3gBHDl$4&s&1b1
z??!|?oniK`8oil9Z`bI7q02vZb(qFH*u>cV!8U4?(El;LwP$^Fya@)oyW^$`4gC%RCO{232?#qiGhq
z)jIdJVo8{o+uBT5M~f*JJEPM;0?XubRE)x+zPnj=0=gJ9=NLKCorh|d_QqW}#=yiC
z&kG=B1KD`gYon~dQYAA|I(@hSgv%E61TaSsg-pMYkPv>~8z$he{y|dwbEmYpgamYD
zWpZZb=+(iDPQ5V{px6`^-BwC6UV=9QOD#^JP;w0twFaX@dSRWYt3|R2$vHW3y%0Og
zzR1|v?Ui=du#gYI(029^2pc6O<(=O*Kl~b04~K(7_mGi?_ZQC-%^z;Oe-|ygqDmAOZ=MQ&)={VMR2bY7B^Zu?V!TLu%D)axN)7X5lq)`X0g7hj9*ukB0LZ#|B!O`fNQAm|B8G!j0sEtxm~Z?^ko55)%r}WNrJqU`;3?2slP0x;1LUN7-5S_O`FH(U^0|5Tp=;It
z6{H1yEV?EpCcw!98kx}WWRBqd`4Bn1W+e`@-j`zaGNqrzPvo*yxF?13Y9haVtSS0H
zdc4&;?VpXPg@hgp--qeJ0Z;)I6$~I`K2O+R$>KAA9vc{pBYur$pe>cvD3`_Ht4epk
z6YPgrtO@&tHg=p&XVJP@Ni)7J7?(fV7<4MOjr-b^6
z>$>A-4$S99KgAMpBPYy{Z4Sgh5wN>Zfwa0?r0Zr(xT409w+#JpKaWXENx>fS+UHxM
zsxssQ7PMb%4D(#`O@M-dXLH_^0@Rk!lTUGTyqa}7$h%RL;AJ{aVCa2s0VEz-KmF`C
z_T7WSxIzC$+H8MS@oE~`I%o5oBoWgV0qi^Y+yCLt;U&O!t~RF*A}d2C8bwP%Y(;mz8Y%1qAa)2GLC2=2TWkU`910lsX!`_cGg`w
zE+2k=b+_Bx+*kyKGh98acBTCpup+zA#uv7gu;L2B(p`BnNpD_t}&
z;M?^zfR7>a6Cu~p(E%`FSlHN-z&ulo^6Nd#6gW3C8Cope=$TwJfFwvV0!{UTqs1U#
zxtlzw*K>6MSZBQ&hXMy92(ai(wT-&-8F}vKy{SA}PW57}8iNiVis%*r60}Oqy2(`f
z!bDeFU812!f1OTiOp}6w&qre$k`d2?nhLczuiz)%3?b!Gz=GhnRyn2$fzhvX6CpG^|Hpfx}dAS
zDNAny6&teV$-9&yhPHZ58;m(J&hGv0M{R&Dq|MJ_B%6u~FMkmUNy!Se5@bR`LN3Rj
z+e>Kiu>SzJ?umiO0+jy>rH~1kKcr@4MCn*n($^M66B;nMP1ef{HYsZ!Qpv&{@>|924r=zJg=3Ih|fcZfVtzc>yK1+U)s
z_T+Jm1N@qx9xm^Tt0=Cicp2$!lry3E?O9o)5nd`K&|RHFynR~G^NFbLsW4()cn?O
z@UI_O^SeHJNoZbwK>dcQg3FfveJq3VBu?+g*g+btlhp{I)7E7qWaL7bIHI2ee;9t)
z-jDT(zu>mJ6%BuWvZS2PLE~S0nE>1vN1F2O+mMwN{lcOmo5jWww}!6+>D|Uqnw47A
zjGQU4v2XBLb-x!%$AC<1h%fK9Kv##$pKSnBx*uxg`uw5)9r)r8$_;$>{vGh(0Al!1
zAu%I(0?K#F6Y&$V!LpjnhRO#tq)}%`S~Vdf2Z1Ds@U1$2goZhot-xOMza`xoN+@$q
zp}E@sowG%&bhfF$8uhGE9f`NUQi8&$RlOc@3Y4cq;Dv;IxRgG}Jjldt(spmoesw=Q
z)LK6F*sa9HNI)-?p!k{BMmvK3c;Oj(zT#YG^d2%I8(BnrM==GehU
zABtr-LjIShaV`LEhah1LT3E%vg|}IhfJ%u;?v9u_prWUb`d{`>gn_nFBWseX-v<@^
z+REVrD!PI?mpS7XV<)JH`Z&n+3BK
zZZ9^xEYzkEsWGjZa>FhG1avZ3V_-T3-G{*%u&A!i9-ps!w6hV%bv_RgGZ}nM-QG6$
zx!G4$8CKCmLPOgX?*4gsXO_@l{OtFu)?n-{y68Bz*<_kE(c}2L5H=7IG5RUgx#SWJA
z#3YCe%ub1{2P!9QRd^cpX1-;je8H)~+kCrTLTj{{b?(Hr|19f$JC@L}NJ)c{R=YHy
zY^1op|J+qvMuka+`+o7=+&<48S@@gNs0q*_|BDDU71tXz7Hu!@D*~p)e^40S`nqc1
zuJ8oYg_rQer2BcTsz6S0%0i-xB1!mqpI#=z`zyf{Z!fBYm-EAKWUvSbmn*JQ)8(pf
z6`$QAk^^Oo*|c3(ggn*=)Uc*72}oR~*&RP)=NL{c@QN1teKOrPUk#0aMxn=qbBgy(
z01lVKPeNob&Xw0pMxahLh198d6M%eq4$Sp4ZO&U+tGFZNkW8x$Q21*IYNhBZ)muD#wzJy_|HDw+@BOZ}pyB
zVv^wv?njVYt}xd900t}I{wGw8j;!1?3|UHA+Sw`w4_J`fdk7*8#5DO%k>&bb$m22V
zN6y(&^YNwMkP4igbeyU+NWlH%?oFWq*
zZds1`PA*@IN~Ea?K(Ih{&<{<$H9^nFgE~dQ97$&@oLlLSzpFsar_)&#mwSon2D1)g
zng#E;V*URR(gdE7foi^9nJ#X&+Y4nX%`Z_q3+4cDod(FMP4p+kZS_lLE8_$%t
z_3LG)K-C9bl=%@3_3WoQE605ZFYCi9%TvU~z&O46vQ*1uAhxP1XIzy<~N{!Cnt?2^0jpIt`OkQl_oBfI45%`~jd*=XU-3p{mMvK;jF~H`$o-7a{Kz7r&nyW!AF$
zhLw7obCgX2Ot!BxBp)4O-c-bZ?SzvcuwzY`j_g06qNavR8IfwZ^8!`#5R)ay{;K)<
z?!xwZTQeVl>4O8*G-ycz3V{)S+~RQ3$Y!OUXKw~YaqSg>wmKlxtxOs;ZBBzPM@Ra8
z9;!d%Vaxn0I+47hJ{wwuwxV4yQx#+;ngf{^t_nJ1v9C_F&VK-5mjVDqa<0Qk#gjQl
z`4lDbt!GgaUwldcOvAC+V8xy}ZeB`rIUvZ|o^&p#USgWUpU~
z+b3Gj!*O8SjbQU
zkUI)NJ+8vazDz80*aYMQLJSgE6R1QB{JjJu41W)(gLd#2i7M=Qd?NS)8t3Xti3SR1
zRCz7P!ho}hNg+YnT0FzdQ>D@3m
z0UBy1k@K2N>5Yh
zuxwWlArE8`WpngmS1&d28`MJ3ILX=9$4h#NVVo?zVWt{*)khZSaoyoYq+2?S1|rPo
zqtA=KK6;^sAQeiyNW{ZSdaVz*%)(kvp8KD9TdJfPZ~WP!32;#OUkqJEKR&dp$__K$Vrb>NvCk!5|=z%lKIWRj}Q*
zdZKV6?c&hJVq?4e{ys?yb!bnJjgjHIjmh7swUo?M?QLz^&5f&-htnC?eVEzG4}>7u
z+P*MOdM|(EmP%TP2zv~F)h`o{jPjW_sJXEr7fL-LH$Mg{shAtB+Qp%Evj{q^Qs`SZ
zdnv_YLQq^cXSd@TvKUZ>PJfy~PruT|3icYEDP6VA{8005b7W0N#EQo=fhc@#b8~nD
zC6$#_QpK9>63>z5?L8Aal7-D;C1#0fGf<(6F{^_IiA5;YU_9LHlhJL9Ksd68h?-Dp
zGBc<_hd16%f~kFPB|=vV*4c_d4hm*s|idtbd31U!CqSo=eK`>r$ta2dAm3D}RT535+)=T?
zGb0$dHruOC0qf%@l&8Kfj`COR4r_m|I;9b*W7wx#g@1w3a
z`z{0u5H^j-&5b%AiF(Bg>$PHGk-bl`M5QxeN^K(KsFk|79q|?O-2WN%@Ze6z@u3A#
zdPjmKTf~BzQ8?#fXA=<>#e8Y}6CD}twO(_Pxdg~WV8NFu?rrx|IJIEN*}fMMw}ae4
zI%v`|x=qNpHmiv?1qlr)U*ewNi-Yj^Y4-_btvSDLv!g_f7itt*daE@`#>gJMUJK1c
zmexQ)gksFTf?u(AkJE}rYDz?)WF49Bs^E~z#uuBkE+Qqwg89I)!xV@-J-fnX{NDmuqbpUu*fQ&K
z)oN?RTW-5-t1y%(3}E1rj>q!bIq&Ea@Kw_W2k#bcRR?cQCwfsvq;e5wR76pw(#z_f
z?`Jt8eUf{Q+b)r~d6rxo?2NTkDr*|{{VbrQnhHLr3EbnXT3d2%cinz+XkBi28F5aw
z8*^6`+kL!az7@SEaV{wwc^44`l7q4I<0YhPvk_+s4T(LwN@oXP&U)cOcd$mvL{nOJ
zX#PYW1|6Gz`k$@Z>vW~Y&Tc=tcAoW0kr4`JavrEKz)#XzvNihhqMfd4{YZf7XVFEq
zeXP1cu~NQVExcY?t>WJ3AtZ+-+E^V~tM~g2uAHk;U}JOo!`u0AW&MY{3%f>Zs`kCz
z#d0)@R&?DiQ@&2h5{W{%+Iiw98>^`aYIrU
zn7tY{)br9h0EKieFev)K?iz51mBSuH-ElkpMTQz}&uf3ak-jXRB;RmGaIxN7xQIM3
zja($)v*q#E#7{|s1zm!d%hlRVP-~M$03v!oC8%{!WUAVUcDmLoaD@e*IE@(ufC)au
zo-D^Jm$4cjt%jVnk=#4L!iI5d3}sv$&LmL@?!RH%ZP>fpNb!u;Pcd{deM~)gLoiD$
z2zj>0Q+`7}9)CR+L-L_f-{MpD9n%BvE)JS?nw?e&XzOiv3}rm570nfyCYIbyM#kyy{JcfN1-_-X1pR*c3ZZ#o
z4Fn*Z$#*R5AFhKY5S+!&4z?5SE^RfM=kdh0
zw<{uA_z$t+Nf)-Z7tog)Lyc|yoWBXiKd(+J3>9UEI%FFBS6y_QPlso5#}E<2V>J+9
z^kylk*^~CxUeamv{BT&=TyHqYPS#L^Sn=&Qznx6*r5>AaCl=UiUnnHG&f5KoXSE#9
z*N~v{SwB|pHt$m~vstMW2JpnMpQr?_|7ZbHq`03vZnS!wN_vMQZHVAX)YMk-yJH`}
zZa+cztqIS8Q~;r~9A{Sz1tsy3nMyJ@22u%9bE{c~hnar(Ye**LJj71My
zru&i1+y&+vo8{DwB_3*3%v!#9{Q~sU#YU*0!I~xP51p`f_*2xv2TYPnYa@Na5f!Qp
ze2b0*ze|K4zBq}O%}kf-8M(p3*x`*^&WG=;%zdMF{c=+Bt{kmZw9R~ZsW4`D6X~3!
zxi_EZy|^ZwRwLKMvggk2+D(1nJddIGoui^9OwkExA5!*w8Tue_|Iek>1c;qXVY!?eD2#I(3Jlfy||cCDdOm2
zr!5k2Z18LKuQ3~ecB?tr(9Nl-)ZUlzHUCH#aX1}%t%*F%Vxy9GpUnq?XM}t{D1$~d
zYn4LqY8IQkg@@2j_kzFWAnG?HnH*`UDuEL+(T^V`1GGD~0a7+G{3D%`C6sBuzLk#^pe2=TL
z$k>OiHMo4o6g-V;!z_1&-DE(NUWZIJyDg=^+)hD5%qIim%^G-0FS(=Khia96)I2Ok
zkuxTH2da|g`lo!bM5$DV=
zM+c<8krQC`F7S$se*mzzCvpR}+&UdzEPKj%>*thD^g0`e{zRI6hv*;Wx*)l;D0>z2
z)zmUYQ4V$=>g{H9zs}Le$rKf6?ip^ITQzzf!vA8)9Se1?yF4LgP5Kk!>+4~%H|S3A
zo<`UGAZ}{U;_2r-Q`R*MI=xcb9eC<-*A+nbk}A#os)_NdgOEgNeK@O|e0iDs!)>uTy~UKv
zA7*3MTEvSu^LG;U5D{VbnJ)#VW5DrIaBaeJr_V^6I}mTHY?9F>7bjnQ;-{isOH}C3
z!r=+vGXQUkE69>&WU=R%N9XkfiIQ(9c@z2%{88G+1-q@qP7gTOGXJrw>khVb=!Rvc
z$Gd}}@6a>f9y9%y!%VrK%ix%@ctZx7N-ysH8nV2Oegu4bGKcmND`WieMAT|K^~a2m
zNiVs@(z5(#|J#B%6+xp&!r?A0ajx}Y+!
z{j_~D`To-A*4?gmSjjWd?U~Ck-?l;JTRJy>5b5EEpI_~q{;$5yGN`RDYWKlCNP`wD
z?i7j?3lOweaWBP+yB7