From 99a7b6d6210c8186e712b467fc504fc5298af1af Mon Sep 17 00:00:00 2001
From: Cassie Breviu <46505951+cassiebreviu@users.noreply.github.com>
Date: Fri, 3 Mar 2023 10:46:05 -0600
Subject: [PATCH] Add stable diff csharp tutorial (#14892)
### Description
https://cassiebreviu.github.io/onnxruntime/docs/tutorials/csharp/stable-diffusion-csharp.html
### Motivation and Context
---
.../csharp/stable-diffusion-csharp.md | 334 ++++++++++++++++++
images/latents-noise-example.png | Bin 0 -> 6009 bytes
images/sample-output-stablediff.png | Bin 0 -> 563091 bytes
images/stablediff-example-image.png | Bin 0 -> 589295 bytes
4 files changed, 334 insertions(+)
create mode 100644 docs/tutorials/csharp/stable-diffusion-csharp.md
create mode 100644 images/latents-noise-example.png
create mode 100644 images/sample-output-stablediff.png
create mode 100644 images/stablediff-example-image.png
diff --git a/docs/tutorials/csharp/stable-diffusion-csharp.md b/docs/tutorials/csharp/stable-diffusion-csharp.md
new file mode 100644
index 0000000000..8a4e048c8d
--- /dev/null
+++ b/docs/tutorials/csharp/stable-diffusion-csharp.md
@@ -0,0 +1,334 @@
+---
+title: Stable Diffusion with C#
+description: In this tutorial we will learn how to do inferencing for the popular Stable Diffusion deep learning model in C#.
+parent: Inference with C#
+grand_parent: Tutorials
+has_children: false
+nav_order: 1
+---
+
+# Inference Stable Diffusion with C# and ONNX Runtime
+{: .no_toc }
+
+In this tutorial we will learn how to do inferencing for the popular Stable Diffusion deep learning model in C#. Stable Diffusion models take a text prompt and create an image that represents the text. See the example below:
+
+```text
+"make a picture of green tree with flowers around it and a red sky"
+```
+
+|----------------------|----------------------|
+| 
| 
|
+
+
+## Contents
+{: .no_toc }
+
+* TOC placeholder
+{:toc}
+
+## Prerequisites
+This tutorial can be run locally or in the cloud by leveraging Azure Machine Learning compute.
+
+- [Download the Source Code from GitHub](https://github.com/cassiebreviu/StableDiffusion)
+
+To run locally:
+
+- [Visual Studio](https://visualstudio.microsoft.com/downloads/) or [VS Code](https://code.visualstudio.com/Download)
+- A GPU enabled machine with CUDA EP Configured. This was built on a GTX 3070 and it has not been tested on anything smaller. Follow [this tutorial to configure CUDA and cuDNN for GPU with ONNX Runtime and C# on Windows 11](https://onnxruntime.ai/docs/tutorials/csharp/csharp-gpu.html)
+
+To run in the cloud with Azure Machine Learning:
+
+- [Azure Subscription](https://azure.microsoft.com/free/)
+- [Azure Machine Learning Resource](https://azure.microsoft.com/services/machine-learning/)
+
+## Use Hugging Face to download the Stable Diffusion models
+
+The Hugging Face site has a great library of open source models. We will leverage and download the [ONNX Stable Diffusion models from Hugging Face](https://huggingface.co/models?sort=downloads&search=Stable+Diffusion).
+
+ - [Stable Diffusion Models v1.4](https://huggingface.co/CompVis/stable-diffusion-v1-4/tree/onnx)
+ - [Stable Diffusion Models v1.5](https://huggingface.co/runwayml/stable-diffusion-v1-5/tree/onnx)
+
+
+Once you have selected a model version repo, click `Files and Versions`, then select the `ONNX` branch. If there isn't an ONNX model branch available, use the `main` branch and convert it to ONNX. See the [ONNX conversion tutorial for PyTorch](https://learn.microsoft.com/windows/ai/windows-ml/tutorials/pytorch-convert-model) for more information.
+
+- Clone the repo:
+```text
+git lfs install
+git clone https://huggingface.co/CompVis/stable-diffusion-v1-4 -b onnx
+```
+
+- Copy the folders with the ONNX files to the C# project folder `\StableDiffusion\StableDiffusion`. The folders to copy are: `unet`, `vae_decoder`, `text_encoder`, `safety_checker`.
+
+## Understanding the model in Python with Diffusers from Hugging Face
+
+When taking a prebuilt model and operationalizing it, its useful to take a moment and understand the models in this pipeline. This code is based on the Hugging Face Diffusers Library and Blog. If you want to learn more about how it works [check out this amazing blog post](https://huggingface.co/blog/stable_diffusion) for more details!
+
+## Inference with C#
+Now lets start to breakdown how to inference in C#! The `unet` model takes the text embedding of the user prompt created by the [CLIP model](https://huggingface.co/docs/transformers/model_doc/clip) that connects text and image. The latent noisy image is created as a starting point. The scheduler algorithm and the `unet` model work together to denoise the image to create an image that represents the text prompt. Lets look at the code.
+
+## Main Function
+The main function sets the prompt, number of inference steps, and the guidance scale. It then calls the `UNet.Inference` function to run the inference.
+
+The properties that need to be set are:
+ - `prompt` - The text prompt to use for the image
+ - `num_inference_steps` - The number of steps to run inference for. The more steps the longer it will take to run the inference loop but the image quality should improve.
+ - `guidance_scale` - The scale for the classifier-free guidance. The higher the number the more it will try to look like the prompt but the image quality may suffer.
+ - `batch_size` - The number of images to create
+ - `height` - The height of the image. Default is 512 and must be a multiple of 8.
+ - `width` - The width of the image. Default is 512 and must be a multiple of 8.
+
+* NOTE: Check out the [Hugging Face Blog](https://huggingface.co/blog/stable_diffusion) for more details.
+
+```csharp
+//Default args
+var prompt = "make a picture of green tree with flowers around it and a red sky";
+// Number of steps
+var num_inference_steps = 10;
+
+// Scale for classifier-free guidance
+var guidance_scale = 7.5;
+//num of images requested
+var batch_size = 1;
+// Load the tokenizer and text encoder to tokenize and encodethe text.
+var textTokenized = TextProcessing.TokenizeText(prompt);
+var textPromptEmbeddings = TextProcessing.TextEncode(textTokenized).ToArray();
+// Create uncond_input of blank tokens
+var uncondInputTokens = TextProcessing.CreateUncondInput();
+var uncondEmbedding = TextProcessing.TextEncode(uncondInputTokens).ToArray();
+// Concat textEmeddings and uncondEmbedding
+DenseTensor textEmbeddings = new DenseTensor(ne[] { 2, 77, 768 });
+for (var i = 0; i < textPromptEmbeddings.Length; i++)
+{
+ textEmbeddings[0, i / 768, i % 768] = uncondEmbedding[i];
+ textEmbeddings[1, i / 768, i % 768] = textPromptEmbeddings[i];
+}
+var height = 512;
+var width = 512;
+// Inference Stable Diff
+var image = UNet.Inference(num_inference_steps, textEmbeddings,guidance_scale, batch_size, height, width);
+// If image failed or was unsafe it will return null.
+if( image == null )
+{
+ Console.WriteLine("Unable to create image, please try again.");
+}
+```
+
+## Tokenization with ONNX Runtime Extensions
+
+The `TextProcessing` class has the functions to tokenize the text prompt and encoded it with the [CLIP model](https://huggingface.co/docs/transformers/model_doc/clip) text encoder.
+
+Instead of reimplementing the CLIP tokenizer in C#, we can leverage the cross-platform CLIP tokenizer implementation in [ONNX Runtime Extensions](https://github.com/microsoft/onnxruntime-extensions). The ONNX Runtime Extensions has a `custom_op_cliptok.onnx` file tokenizer that is used to tokenize the text prompt. The tokenizer is a simple tokenizer that splits the text into words and then converts the words into tokens.
+
+- Text Prompt: a sentence or phrase that represents the image you want to create.
+```text
+make a picture of green tree with flowers aroundit and a red sky
+```
+- Text Tokenization: The text prompt is tokenized into a list of tokens. Each token id is a number that represents a word in the sentence, then its filled with a blank token to create the `maxLength` of 77 tokens. The token ids are then converted to a tensor of shape (1,77).
+
+- Below is the code to tokenize the text prompt with ONNX Runtime Extensions.
+
+```csharp
+public static int[] TokenizeText(string text)
+{
+ // Create Tokenizer and tokenize the sentence.
+ var tokenizerOnnxPath = Directory.GetCurrentDirectory().ToString() + ("\\text_tokenizer\\custom_op_cliptok.onnx");
+
+ // Create session options for custom op of extensions
+ var sessionOptions = new SessionOptions();
+ var customOp = "ortextensions.dll";
+ sessionOptions.RegisterCustomOpLibraryV2(customOp, out var libraryHandle);
+
+ // Create an InferenceSession from the onnx clip tokenizer.
+ var tokenizeSession = new InferenceSession(tokenizerOnnxPath, sessionOptions);
+ var inputTensor = new DenseTensor(new string[] { text }, new int[] { 1 });
+ var inputString = new List { NamedOnnxValue.CreateFromTensor("string_input", inputTensor) };
+ // Run session and send the input data in to get inference output.
+ var tokens = tokenizeSession.Run(inputString);
+ var inputIds = (tokens.ToList().First().Value as IEnumerable).ToArray();
+ Console.WriteLine(String.Join(" ", inputIds));
+ // Cast inputIds to Int32
+ var InputIdsInt = inputIds.Select(x => (int)x).ToArray();
+ var modelMaxLength = 77;
+ // Pad array with 49407 until length is modelMaxLength
+ if (InputIdsInt.Length < modelMaxLength)
+ {
+ var pad = Enumerable.Repeat(49407, 77 - InputIdsInt.Length).ToArray();
+ InputIdsInt = InputIdsInt.Concat(pad).ToArray();
+ }
+ return InputIdsInt;
+}
+```
+
+```text
+tensor([[49406, 1078, 320, 1674, 539, 1901, 2677, 593, 4023, 1630,
+ 585, 537, 320, 736, 2390, 49407, 49407, 49407, 49407, 49407,
+ 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407,
+ 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407,
+ 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407,
+ 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407,
+ 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407, 49407,
+ 49407, 49407, 49407, 49407, 49407, 49407, 49407]])
+```
+
+
+## Text embedding with the CLIP text encoder model
+
+The tokens are sent to the text encoder model and converted into a tensor of shape (1, 77, 768) where the first dimension is the batch size, the second dimension is the number of tokens and the third dimension is the embedding size. The text encoder is a [OpenAI CLIP](https://openai.com/blog/clip/) model that connects text to images.
+
+The text encoder creates the text embedding which is trained to encode the text prompt into a vector that is used to guide the image generation. The text embedding is then concatenated with the uncond embedding to create the text embeddings that is sent to the unet model for inferencing.
+
+- Text Embedding: A vector of numbers that represents the text prompt created from the tokenization result. The text embedding is created by the `text_encoder` model.
+
+```csharp
+public static DenseTensor TextEncoder(int[] tokenizedInput)
+{
+ // Create input tensor.
+ var input_ids = TensorHelper.CreateTensor(tokenizedInput, new[] { 1, tokenizedInput.Count() });
+
+ var input = new List { NamedOnnxValue.CreateFromTensor("input_ids", input_ids) };
+
+ var textEncoderOnnxPath = Directory.GetCurrentDirectory().ToString() + ("\\text_encoder\\model.onnx");
+
+ var encodeSession = new InferenceSession(textEncoderOnnxPath);
+ // Run inference.
+ var encoded = encodeSession.Run(input);
+
+ var lastHiddenState = (encoded.ToList().First().Value as IEnumerable).ToArray();
+ var lastHiddenStateTensor = TensorHelper.CreateTensor(lastHiddenState.ToArray(), new[] { 1, 77, 768 });
+
+ return lastHiddenStateTensor;
+
+}
+```
+```text
+torch.Size([1, 77, 768])
+tensor([[[-0.3884, 0.0229, -0.0522, ..., -0.4899, -0.3066, 0.0675],
+ [ 0.0520, -0.6046, 1.9268, ..., -0.3985, 0.9645, -0.4424],
+ [-0.8027, -0.4533, 1.7525, ..., -1.0365, 0.6296, 1.0712],
+ ...,
+ [-0.6833, 0.3571, -1.1353, ..., -1.4067, 0.0142, 0.3566],
+ [-0.7049, 0.3517, -1.1524, ..., -1.4381, 0.0090, 0.3777],
+ [-0.6155, 0.4283, -1.1282, ..., -1.4256, -0.0285, 0.3206]]],
+```
+
+
+## The Inference Loop: UNet model, Timesteps and LMS Scheduler
+
+
+### Scheduler
+
+The scheduler algorithm and the `unet` model work together to denoise the image to create an image that represents the text prompt. There are different scheduler algorithms that can be used, to learn more about them [check out this blog from Hugging Face](https://huggingface.co/docs/diffusers/using-diffusers/schedulers). In this example we will use the [`LMSDiscreteScheduler`](https://github.com/cassiebreviu/StableDiffusion/blob/main/StableDiffusion/LMSDiscreteScheduler.cs) which was created based on the Hugging Face [scheduling_lms_discrete.py](https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_lms_discrete.py).
+
+### Timesteps
+The inference loop is the main loop that runs the scheduler algorithm and the `unet` model. The loop runs for the number of `timesteps` which are calculated by the scheduler algorithm based on the number of inference steps and other parameters.
+
+For this example we have 10 inference steps which calculated the following timesteps:
+
+```csharp
+// Get path to model to create inference session.
+var modelPath = Directory.GetCurrentDirectory().ToString() + ("\\unet\\model.onnx");
+var scheduler = new LMSDiscreteScheduler();
+var timesteps = scheduler.SetTimesteps(numInferenceSteps);
+```
+
+```text
+tensor([999., 888., 777., 666., 555., 444., 333., 222., 111., 0.])
+```
+
+### Latents
+
+The `latents` is the noisy image tensor that is used in the model input. It is created using the `GenerateLatentSample` function to create a random tensor of shape (1,4,64,64). The `seed` can be set to a random number or a fixed number. If the `seed` is set to a fixed number the same latent tensor will be used each time. This is useful for debugging or if you want to create the same image each time.
+
+```csharp
+var seed = new Random().Next();
+var latents = GenerateLatentSample(batchSize, height, width,seed, scheduler.InitNoiseSigma);
+```
+
+
+
+### Inference Loop
+
+For each inference step the latent image is duplicated to create the tensor shape of (2,4,64,64), it is then scaled and inferenced with the unet model. The output tensors (2,4,64,64) are split and guidance is applied. The resulting tensor is then sent into the `LMSDiscreteScheduler` step as part of the denoising process and the resulting tensor from the scheduler step is returned and the loop completes again until the `num_inference_steps` is reached.
+
+```csharp
+// Create Inference Session
+var unetSession = new InferenceSession(modelPath, options);
+var input = new List();
+
+for (int t = 0; t < timesteps.Length; t++)
+{
+ // torch.cat([latents] * 2)
+ var latentModelInput = TensorHelper.Duplicate(latents.ToArray(), new[] { 2, 4, height / 8, width / 8 });
+
+ // Scale the input
+ latentModelInput = scheduler.ScaleInput(latentModelInput, timesteps[t]);
+
+ // Create model input of text embeddings, scaled latent image and timestep
+ input = CreateUnetModelInput(textEmbeddings, latentModelInput, timesteps[t]);
+
+ // Run Inference
+ var output = unetSession.Run(input);
+ var outputTensor = (output.ToList().First().Value as DenseTensor);
+
+ // Split tensors from 2,4,64,64 to 1,4,64,64
+ var splitTensors = TensorHelper.SplitTensor(outputTensor, new[] { 1, 4, height / 8, width / 8 });
+ var noisePred = splitTensors.Item1;
+ var noisePredText = splitTensors.Item2;
+
+ // Perform guidance
+ noisePred = performGuidance(noisePred, noisePredText, guidanceScale);
+
+ // LMS Scheduler Step
+ latents = scheduler.Step(noisePred, timesteps[t], latents);
+}
+```
+## Postprocess the `output` with the VAEDecoder
+After the inference loop is complete, the resulting tensor is scaled and then sent to the `vae_decoder` model to decode the image. Lastly the decoded image tensor is converted to an image and saved to disc.
+
+```csharp
+public static Tensor Decoder(List input)
+{
+ // Load the model which will be used to decode the latents into image space.
+ var vaeDecoderModelPath = Directory.GetCurrentDirectory().ToString() + ("\\vae_decoder\\model.onnx");
+
+ // Create an InferenceSession from the Model Path.
+ var vaeDecodeSession = new InferenceSession(vaeDecoderModelPath);
+
+ // Run session and send the input data in to get inference output.
+ var output = vaeDecodeSession.Run(input);
+ var result = (output.ToList().First().Value as Tensor);
+ return result;
+}
+
+public static Image ConvertToImage(Tensor output, int width = 512, int height = 512, string imageName = "sample")
+{
+ var result = new Image(width, height);
+ for (var y = 0; y < height; y++)
+ {
+ for (var x = 0; x < width; x++)
+ {
+ result[x, y] = new Rgba32(
+ (byte)(Math.Round(Math.Clamp((output[0, 0, y, x] / 2 + 0.5), 0, 1) * 255)),
+ (byte)(Math.Round(Math.Clamp((output[0, 1, y, x] / 2 + 0.5), 0, 1) * 255)),
+ (byte)(Math.Round(Math.Clamp((output[0, 2, y, x] / 2 + 0.5), 0, 1) * 255))
+ );
+ }
+ }
+ result.Save($@"C:/code/StableDiffusion/{imageName}.png");
+ return result;
+}
+```
+
+The result image:
+
+
+
+## Conclusion
+
+This is a high level overview of how to run Stable Diffusion in C#. It covered the main concepts and provided examples on how to implement it. To get the full code, check out the [Stable Diffusion C# Sample](https://github.com/cassiebreviu/StableDiffusion).
+
+## Resources
+- [Stable Diffusion C# Sample Source Code](https://github.com/cassiebreviu/StableDiffusion)
+- [C# API Doc](https://onnxruntime.ai/docs/api/csharp/api)
+- [Get Started with C# in ONNX Runtime](https://onnxruntime.ai/docs/get-started/with-csharp.html)
+- [Hugging Face Stable Diffusion Blog](https://huggingface.co/blog/stable_diffusion)
diff --git a/images/latents-noise-example.png b/images/latents-noise-example.png
new file mode 100644
index 0000000000000000000000000000000000000000..5cc8dc71c2322f415d2b8e60e37020fb099779ba
GIT binary patch
literal 6009
zcmV-<7l!DGP)
zBDr+2t@600aTlT#CQ3@B2$f8cQJD~2%+h7nBhRpboFW*m45KwhxnWn&}F+S%jQfG1dTwBRtVywIAjwd}XpuWQtDZF^f!A-vc&**KpC{JGeL#rtnLSeo!qwRrm?pHM_!Mw*(xPR|cfmE~w$z>tck|;&
zG_!D`Lzv_X$+|bXAR6eZLRc6d+zz!&Kn(O_C+h&k;2n2e7ckOOoNpMo#LQ-A2;oG#
z6z3F(CA&=e6hbJbYZ}GxTexAL&4$<8YC@xmDUYWmK-F|^wedW@3iCxQ*W
zt_&Q;O*3_g{keI)Ouj-G`7P)OpEJGr;icgZSGwb7$iGj?qMPer5IfHXzUC6(
z_U|#Y-$RH8OJpbt?k}o;sn)^qTw+>i(y7(lYD`Y+Z3t&XVtmAl(k$&=?@~N?ymWn*
zPC7IIE3?*S`ZCxFsgg7JZx{&!5Y7vwpn^$~$i!~~NH41#tY87Gt~5w~7eG;lhjje%
z(2C*RfQGL~i5F>g&X60Dk0_via56u(CwQ5kk3J1jOV>Kbdh%sil6Y0+rTDkNAu_x_
ztG9PE5e*)O!T?fcs26pv7dIRBhlSv?_&T?6+;P$0%RKU!Ffs~e4CFu<9`I?ON5x!}
z&R?fGWU}R|-$vdqMA(J9j}}sXLB6KXuj0Hp)^YqiF2kqhCH*ns6D@Pnp5w76a1`+A
z|I8zvg)jqvUOaIr->;>TyQTU+;8d@5TSHFWsI9476{uzAo!oQ^ncbFnkFX=;pYfJb
zQs5fE(H%Z=%^*0ZIIs-ydwG;f>-NZr2_x22rx@C+W4h@0Mleur&ty14n9bKGQlw3jZR0zf+y?Y3dg!w8^!VsOW6?CoJf>s-3IPjY32nvk8XqKnKo7JQ_W?6Fprg8WARrCoM{P*vCy$3^c#Rs?@7q=ZD&tK<2rDFf*i(w)o
zz|If)xt;eL@H2iHAK_0Cyuj%;Mq%EHt|6XTR;&iF=00SDw7z6r!l}stI-v+
z^tDH&#+2bJ0Edb6Knr8?yO&$<&G2aEC>l16_VBs9+v{laiZY|<2DRhk*U74Rx_*US~Au#m1SGCq0n@Y~)QhLr4Y*w{d@nRwBrUHS2g
zNP)3JfxsnX$mhPnRggD$Fw7BEr?mrf@c`bYs1zIvW=TO^?oo8aFS%hY(3O{!2o#uL
zH=p7mZl92Hof7l!5dQl!cl}LzMh}Cv1Ej;a{q+JooCO(yFYx2>P_!T%VdVReu#RBj
z3I5kLORp`i;rmzywI{cHUioFS|LuI)t=eD2`X-XzhTko0rnmf8YL`6yD5M>{#F2cT
z3Ur*46byT#IKc2@S`dM!`02!0@O6l4Pwkit24)gtaW>uX96Y
zPq~x4m`6=YW$=X>RPu;)7a#i8HW(sp+qCnD6ohLdfa5k5;6+#AbV>kG%_<>ZqJR*O
z9T>o(%#u=7u_m6$OC?JzK0<#xD%<9+e~CHOl8^q8
z*KF-WJQ)jL803wW>!)G1tS5AYn@d#|Lz~#3B%-5ndL6$U5Gv*9RAEun=eWcx2;{#r
z8Tf-((+Ib6-+zXg1m_fqyETl}8wtd+|2Kqq$Hj@u;ntSvw9wJcWYGD+h7f&iCDTZ%UjD<6V%GBt3ZBeN9nM*N_WL
zFD}I#o_m%*iU}xHLsM<*F!KPvEJ=@A2zgK5PaDoVvrrHZE!?*;2$0CrEleidZfHYM
zT^sQGY)7{2%;1np2DyWucQp+e1b1@VUBzUZfw_i91elVxpOjU~*T@8BJzv`uCXfR?
zi(h(GHkgHlUrdL>t`a3FO4QON(d%UEdzqV#adX4U2Aj~XHa{L*AvKhG!hWGK{QpC_
zbcEs}B&~0HuRls)vCY(FwZ
zK}9Id%gk>9KlO)FbqdnfLu`a9pW;lh;>$n-w36FK7fAyS2{GyQjf`Nv7l08Acxo^K
zQqF_)>2V&>*VPJ}Z6v8G5n@CpRiPP0s&y#K%C0JcF}yGZU_gY|SmzvmxW8t4WO`vb
z0&k%8cR_fg3&5e3W}zJVWQd7S>ly&}a0CDSQ`YoL;cz2+H)Hlyw(x!=={V36#t@5S
z?m)mL*&(V+xg2L0mPdT{wf2zj$uU>zepnnxvazltce
zL|6Eet=oZH)(B+TdH+~a4^L_c7jmZ8HwDXXQO*tExo}=E&P-?nqk=m#Ipa;5df`&}
zMXgh-^9f0D+o39nVbBU@!US#~9ShMQ9oN=hk70FvnO84+gL-3B^G%57ZPJzZWTj28
z1?VE@W^lY_l^42pfitI^Pr!FLGyRP?LC@Wvp?>VNDKoxGyR)93dSAt0__lM|2fxM(
z9*%#@&!2jTljaivBK*ND7~SMgtiU*pAE#|+1b6c(quor>hg2mqhWjQ!IanEwuC-mX
z{On6z#H!SeZ0-w3p-icuw0bl<^wz-U7U
zp-2p;z(k%nI`RS~VwpX^!hb)G(y}Xrxa|SO?g=VbRJe{~0D|FFfbum79875lYQ=4R2YgrFQQs~+7!>q2$!qo0RR*EVWJc2PGt&rajt%OuaH>C
zn-@7M7T}I{u2%i6^sIn8Rjxa<&tcqf1HyG
zenWWGv4&NaFc?!+1;jzM5!j0
z@E40o%XOiSPcn>?4Bij9?X6DSEFWh)w-a!7tr}EdD}=Y9_J)Kji%DtIN*#7Dj~T>s
zI;Rj2#`D7!GUm#Q;Y}6FYmy1N1n|q^^kQuet9wO}M6Of9Ar@Mp^0E->I)R~1@Ys%(
zDt3dv${)Z;0$9y6SNCGxECZNbX<4SYb3M{@%r?7rbvwThG(BF^)yS*FUd!?XStxeE;cLGFBQ_+8E1&8n6>bM`^4blqR_wRhr!b~qZ!
zsy`+kl}7~s=-%Q)It+dllGuXq4nKDY&oA&F2aq$3*7czw4dBHG0EKeqncU&!0D%?FV{1|z+#=l=aYWVg!Pp*KKD`N=-^$R9b!qLv0w
z5Cr-|5_fP^cra9~B)SE8aa29Llx{s;c-+dvip7M!Xp5c@{gum7B6m{pQv%OQ=UsTr
zN+>##FG7j{l2iHCfaxR=1qD;JkpT~t}P)0;S
zRX!(_?-$laq7%&3b!2xa#IrAwZ0$jQ90XO>Nz39UOIc>(ruA^=QJ=+c#NpY`Ys`VI=?-SnBKf)o=a44dB+A7@sYyD+a~#_OFr!xc{;$|
z3vK?q*wchIu>ZmZhLC=|<_?s9jhkHFt?Ia>?0XpG6d+Z+c8XODw`1n}{kf=j$?z+p
z)J9KG$^8LpO+qLhwaWSB{)(#^7RQ`IrXQ$ut!-hN13mj{)_Gg~D??{=cb{lA&Mqe0&_CrxqS@3j&aSB`~X_EgkSUXS5zXM
z0oe?X;=XV?&wnnDkVsvJ;L)g0RkAW0>Ris8wEYApEDl5=*p|$xghE4I0A`~T;J@9R
ziu#2_r0pXf)aFBUjx|tu`MV8EFX9zKDHq6kT^fItRpKm&Gy2i44KoZV*B9us#_WuC9%X1IZIaeifFJZk;xlloOKv(u{HEFbJ{XKK
zX?g+d_$q`V97JcE*TI|J(fw=soXypu5vrG9$_{d05}Y;3NE
zpK|ZwK+DxHG)5YGV3trllDx(39U*vQiSY2mf`%j~vp5n++T&D%znt6RSlyXTJ57(i
z3sn?jd`>Sx-_5{M=3*T(x~KnxiLBg|(sg&OKz3t*@~u@DZu=mausCtx{9VVhH-<@Y
zgG*}}rF%1iE|BvW`3FuHITK86Ghlisj(h|lqE)ES2p%5mt!n$DvB@HR#MowLwwyQB{DXBT8h3g0;e
zn34smZ2VjvJs2W^iP@3v?g!6uDqJ*!Z49s%W@yfDm#Xsyd72+t=_b%m@;(Y}lKxJ{
zutvMR9ht{DM)v`jnOx%gNj!9E`uZ@;>g>=xdF4N{*qIjI*+Z03^&(Wr<_<0@Ke3yW
zXRo1i$Hv^(%g3Gs?Rz#U5wFkbBK50$3#7w614}5(uY!{CTd5q=Bt{jh+P9aCSxncOgx
zQa-Z8XV@_-!eu|iW{hnu5%Qu^bwGjkDfqQEdIZ?aP>{mb;EMd~-Lst;PN3Z6Yls6x
zg&`&KWEDA7B3qoeGx*AX+f)sA0SWNrdQOK@p8XLG(|NyDBDSRo0`-OO@okg%1o=i+
zROMs$8T1bz=TGU!c*0WL4|@Y3OlRUq$Zv(5vGuk-wbvKE2AAmAMv9zx)3J0_
z+V|PTC#T()_zTc<5<^(RTjCb|SO`F<;h>SdcKmV(W?4Q;5$NBwt
z_Md%?e)j!Wtf#O2UftK?k`?`{zlX17U;O>c%D%Sm-Jjz*8Tej5hj+tx_{{zc&&l9W
z0?*msdl#R%63+-euZR8p-ztvhJHK7o^UWR6e3b54n?kX7|7>yp`}cRgi(~d9cC>7)z!3WS?-+YezYp;D
zGkM$e68A1Z`1-qNea`omH*f_cvSNFUR?F>LS_?1BwLV${NzJktJpF^hb%F0`$1txY
zf#~0>b*V0Sx<#S-`z^0}K*c;)a-?m)#yi*k{KaxnymwxgRL;Hh{N3Zc*0e9#F<-mJ
z#(x*rb@iO_8P;!MU6vlgd6&NVuQ3$A*k3D7U28s5*L?7|^rUu;ylBC;=9ycf;qg^S
zEef&$F&+CB*ZN*#mQKPh&htz2xbEMgp$C`NM)q*uR^eZ|S(n+M4%L|t$K*Yj-B=oL
zQP|GxQmciv=ymcnM?bOMKQcDAmTC2TbY)#@DZ!GA=_bC!--qn%uxH`*9E--xIsQB%
zMuIFeF*+C5+9$_asXeOM58iK=d>y_}9DiK%guX%+ZG|C99sPa{
zEnyX2yZ+dIw8GPp;!0+|vby)>8+yI=cQ3!6?f3iBPyFZm`Tza*|HbFO|IdHoUH^e|
z`6}Yy%O^gcFCyR!zm77#UxXB`OL$Tqmw(ThZP2){MWtP{*iwpf%VS_4dY^OYmWJrL
z9K)H=n|AKFU$jHW;)XwP()ezUv7(ga1rpA`H=pl0Ndh)^Z7knNdfcB`LdOVL_w&by
z0!-=iAGMzQ-}l8$E~}At(!ZaNguVLNqVw8t-T)2&bSc5+vv(4J$xzfhR=_MJf1vuE!m
zisw0yfozbbg|1ik0KjO`JI~)i0H^z&=WqO3F?;yRD3=Xxi}LxVeB61`pSJC`-bLd+
z@;1)ll1J)K&tG!TvPD&)bCA58-{)ExVuAj%_y6^71Au6ILwS26gabW3%2lq}dl)y`
z73$V8YHc3qy>eX+mum-)!m+*A{0RWgqV-(%ro#Y`JU4nd->rS<{U$7^;@Z>jw+scJ
zhow9LMgj%Wq1|pPJZ6l4oa~wV`r6h^Uv7ujLm1H
zol-LRk}L?$o%Qy3^ziXySi8)T3z<>dLfODF(BEwr&YD2B`?+?q6SNdvjQ7fpm-459
zF34gFyq%z)ciE`(Jq?Vqeb!@-;Lkb~_PQ?fy9qeYchCL
z```Z+pUO|ve|`V;&-c&zpLSM)ww2sz$>Cf0Tr&BUE}*654Vdfm((E+4kx|x$8c@8m
zIjyj|#s*Xhvp7L+YC-tLCP27MO%rPE;U521IOx{aDWYG;BKalykmy
zyQnmpJ0lMx_a{;^~b`U|T8
zB7owO{qe=C*IIMLsuaP$p26>Jqu%2`
zQRD+BV5eV$>v}mB#sK3LJ+8cTW+i3w7>Z3!Gj@BzS%>2&8!TyE58#7*^~`J2@X3cz
z?&&C`hPlG){|Ll$K<&&Epw*yYGs?Hl@E+giDPlU`X<6%immMp}$9E9x1S{YB?D02q
zpDiYe^1nvSOKWOX-lJ#{&dw#fV#j9RAZY#7FbKlrxs730c#{dhMd-xkGNR_
z+6j(vqGCF1*?sBgk=-W<;I094tveWq04-Snhpn$|Gj`jBeSzEa;j?JBjm!87)p>(zgDB-Gb0D?j8Bfx@<}Zy?CU@4zM{f3STGToTuR
zKz+pi&p+`w)PMXR=V34Or={s_JD6M;3u=dN-%VZSh46QIBwK5aiUQsLexSrfh*(_E
zu{aD3EetZkVI|qb32`Dvq96FI@1NvyTZBp1HQK6WLm}am@SQO^ddllJ8HC!e(;@db
z5OTaBO){A!<-DRQ8Ku(tfb--1uD|OZR*(!^0Dq6)vDJl#DQTe)x1LK!3Qt$fxZIQ~
zjEe*_j%0X06up?0Nu~_A7NyYcT>iw+n1HSv<7MO+v2N~{yxCB&7~lFlv)NwreL_f%
z6;ayi%U*i|O;day48YrfugHirBwM}&0m2a{8D_6b5iionE&|OphARY5X;;#b{p>rY
zxaKoCv0L7#bLFvXYzq#O9wcSlxXp6n&TvHxGogRi(12PKe6I`$7#y4a?P!(2OHs~<
zbN<}hd&?+Wj75v_gFoLkUQpb#XYi+Gyp8)~Uz_P!+=mi_z_A_&_SDFCUNXWP;rU_2
zHEWT|!f5dLG#L<3Cvw$wUtM#>Fy~q^iW||J(g2#5RG{ava_HG>u0NDG7f?oLw0>p?vCT_RpmUSo(4W8=1S>iB^5S7P<8Nt
zefX}?p;8>!uR*~TQ-QiMi_B>`N?
z=noDd_i;e)*zG5;u@>*OsBEIW!$?H>V`?HS9hb7o$;9|LGNmE+h;b3~O4)jH&4;dwsuid?Zx`X6#v@Yem7g
zJ9?!@klB=7-ewc99=o`Po9QhsUYlB_l!HMQC$?L@9ysfud=yh!
z90x%1IBYo3G=MBB!Ew6#Pxg;&?h89~^k9)$Knz4CMQB-ugZiBUe$Ajj=Xy(z-Fk8V
zPp7e>h;-%Cd8I?}4k+~^m$vgf1%C9LUNSo6EAcNsE@%J#M
z#|Hm>V?IT4y)m1zL?Lc^qploUaZ*qmQN+JF82(*`izvAl`~kHIq`vSKp~ehHL8;;t
zaa+u^^!B{)WXFiKkzqj~Hl&=BISxekTh@rOaV{#d(3};SXmK2988k@Yt$P*mzcCq{
zvyiNW&Rp>Ff&t8My^6-pRI8_s`^yo?$Q6~8n2`6I0+?RAS4t`Rcz}nUl)k+X;ja$U
zIDb-ri^7WiD&+J^>&f*<5X!I1NPYkJNk<+s91mqvP85N5IUrH305KsW={F`Ng?VkVCb8s<<6;MPIL#E2
z*Rr@eXHUHqg95{w#=B=7_PVa6y!8yONK&bcb&TBF-lMH`7d^_g+FbAF1D?j2j3m{{
zwQ8Wn4`ng>nWE+iSQhOROxgm4UWn2*BPQt8ihs7`j6lr{<7QTpPW;KSW!#*iDlZ*^z5=`5lg3D(SRG|(VKBuZ3i^9MURW3!-0xUA
z?;CUoG|H4%(7BriB-c2?7}^KME3JG6CyG~S_*EI-#z}=(H{myoxkrDY0lXSh<#dg@^s>q2{1-fEEBFy3){6D8&&d(Cp(B{x?FO;7DTU^|
zDk){uu}Wo{OqMBEe~d{Rmql(sw>qCG}+|-2xz}gv}caQ0$4mtO>IWHdmXkQAqxC1DqsFuS@u3?qOtq^u8
z8Z8xU?y*5%$&ldlS`yY9PJmEiQT~5Yv^InU34pZ>`_a=>gzPrKv
zF-qe;6U6@!h`#AG$?;c)!*|(m~ONy=ePY)IUH805ztW
z%u)npB5{3kofm<}mP}HXKGLt=#D~ju43H`$T<}%bDAzbRm_qu%$xD8oJcs&eDbF_1;s)9JE61l{h3BPNh93P%(%cu+2P{f%kyv=A44LF5P)ZLs4;GD&}NHPKxn4(7CS2RvS4D?lLpdG&`Euus*
z2Z9k>z#-s|q$vQ}D;_FdIgK)N#xxoA2&j>o%pg=ILY`xQTB`)b`v#pVB5VD<1JX!F
zM~q;EY$O__(91xPV9AlM1a&vo}cE
z;bX6XN!A{%1h9ggdgS!xQb(XsO8=OfMEXgmI_lCB)ytH0AuUR!sMoTIg*CZCpwf?|
zj2|wzO$wuLFf?iB+Yv2sgHAb^A&t>hvz*DIwSc`kmsCDs}MLnil?DC963Sl2>9t9*@uPo+2b
z>sd%GG(g`N!vte{>_U=_4hCNYBnR_8}HDaxHD9+tw8bL~?;
zI%2`5V(HjUFsIdf{y@`J6au6BYuq}QcU9nbG`z$4X6ltG`pI+AiA$Dp?wOaP4}ucs
z`~;OL3HI+5bL{DIXv%}al4lzxhHgC9Bj#bTppG5S=%
zkj(ehW>IB%k6WkCGvD%F3$S*cbJI8^bMG%x4~b_y1VKJBMnj;rkaNNb83Cc->78)C
z_F1CpNy{!V8l-?}ELHzN^g)WX3~m)EJN55
zm-5lFOPr6ve~!)(4C$EV?z)A^NjJ%*|j8Qw5y9u&Zs19
z;#JvBrc6PF#Qcq6oZQ$ur<^0JjHTc*|{y_$e*d;yUg>j&gB
z1Qt|}Axb|vk5au5_qS6#FIfYog8ie(A3`Ea<{qsn#RA~E7
z{lca_J#2xTJoTib|5Vz25d<(;Ag}+BZG9+XH$xY4*K
z`|gQ!X@v1l=7u|022eq5!IM<@kq}5Z#uoU*2IX@p!*zxh`L3l0`HZz4{!hd{&V?MA
z$ALLgS%Su_#;D4Ijj773~X@FAP!@u3fZoawg1Y_RZMgC|gk*W0F|GokQ{l7^w8
z^04RyQRim#i!^`?M)7ol6ihY>0007VoJNa0_q<|MWAg+|%hX$(SXwsda|gHxEYo1?
z40|fwB_k*&nWASE%|UO1P6X`8`TORUl7S$|KEEIJH;5^6+%rVl+{-{{Usz*;oFbeZ
zSQZwX5wI-N-f18FSd+$EN8X$tqVBU6)2z^@Y;9CxGSc3b
z|DN-Lp%l=m%MK)&Mu9zzgv-R@Oyp>j_umA1>ZyUM>g7n;wuz@*I-O$25Fr!ioH*B_
z^UCe~HbtPI7<ZAwP69f|H=5UZffv{4s|`snA|wI--b<3fwtC|a)M5l2f>
z79Cs0*ivXXbk_ee~AuAu6eLlgG|LEuU0zoz;$2)o>N33
zFtKB(-QGJovIgI^47C3F8rh{>^G*YXrTuOVik2N20?m!7a~qyu?kxD6lyfQp7DgUD
zUZ$zEP7;n{&JZ&}(H7+#Idjs@i;jcNVWM4*Q~R;GUUYg}=Y+*Y$gwRt{a1htR{8WQ
zXv59E3Z}nb9dnXl;Qk&bFE)jRbXFM}>4^*|ddxq=X(_i^2DrS6x=VFQ9@*3$P*$Ci
z3yi<-$vlA!UJ2|rU>F)5MWgKNka5|E3^hc_kkf;UdI(xH#-wpf>Ezk5Ck6l7)$6>p
z^VUpTAYBA!yO}Gm81Nx*tm9saYl^i-s9dMOrc5HwZBxt^!2^OAu3IGkzPwhgBM-pm
zIhnM&Sb}HKZTrggFS?gI8wv;}s>f9`0Y|yx^B^s-_T|K)bxtKbUqLv6xcWSzoml6QXt)%S
z@aPL!Jb4#-?kVIGot4a|!$&O&DUIFApMtqf2ZKuNHIBCAu;k5YyFhVI
z5cfp;R&Gl~8eB!rk))1#AG(gc7-bWD(jc~~y&LnC#xCP&tmO^*xa466&5u0C<$R?;
z%W~D-Fci{>#Umfr?VYnt>QzYgdQ75Zw*id8iL}M0F85SK$Co!qhEb;^oOc{rkn6NO
z4hQ-I8Y`UtR}4`lV2$te@kFLU45F@GLvbJW2gg)tU>6)?*K({d+ZK{k3_EmAUgy6n
zW<0&U1hQQfA2kLwE(aeB7GBA28X))@fl@kz9v7JafqaQ^)6`v=_aN;v8)@et-2n4L
zP(s@(5!F5?sfV@`Md!D09IZ-q3V23i8>JB+t6#V_A=n*JM_JdfEf8Vsy;I0awFqQH
z(Ds*!(jKR`6;=P~Ibhe_BF3o~0fJ&31uoXnKLxQmUVFy)a{Y^A$P4spZWt_?9N%ao
zK*?~g0oecqAgYDo8gu_!FJH*tx@>VXngjtMaD2YPx+p7Nc-uoK#(eEAvplxrfq(aU
zJzy2KfWYQSf)6(Nlch6GixK03Yq$vm9m7!983cy12r4DcF
zA8!{H=~Z-^te9OcZs_|zi$O_X~v*`tJv=R~=u@-jaVS03>onxrjN3Ct2
zG_Dc-gJ=$pK+z^PA#7z^ai3*FMs@TUBs}hS5On^9pOY!Zda~eTDqXh)0TCpGQ*9}G
zOiUf9z_ECU^i;O0C`?X8&K}49Xe6Y7p+b@1eUr{E#%VGXN9cPG#}CxU4QmKMT+W*h)8g9bxYzRL!AIes)*OIHAIb|utiB?9`u^@
z_{t;t_^FHt*IzQ%U~nx1b{wCi@Rp7X0nkbG1W6Wkva=j3+1(Agg>*jI7ZnZDaJL*I
z)CUHPIqJFTzW*wpC$-=#IIs6Fe8eE=7>wXW12RuqJ5-pL
zss^XgJCW3M83cPE>T85DAh^e{u$Q$_sPg4hV+9;Z?#>Z~qF)+n+C3U`I*DiUZV*DC
zSy$Cc;b_bNKrn2v2C@%vMZ(ywFiMs|7Z7noy#xlhmMT3fM9$r>8esq;iYi{qbMMQm
z(DbohRw6|1BYa#7`f(KS&>f1dgn=f4EbP1K-E9Ui9MFFM*F{|~`%Duk!pJ}Y{$c!h
zl#i!a0{x|+z{g{J7)9=;Dw9S#jgGfUp=s!N$S?jmwEo6gKQf#--Mo67sv%Od8XD9^
zN-jC4t^E0sb)LJui~YDuLnGTkg
zKxC;fMzIDMQx`UoV(u42rM^o;T*kZDGL7Frm1|0X@6=Ba(Cuse&$Xu?yEgr@TZ{Gs
zg<+S>Ma8_gpsr^EiO4s&(2ieUWE^}XD)LnPGWeptBh)1iV4m&1*{|W;|3jMO
z{QIA^xWAV#*Z@#J9cIZ8=bFY2a)^mHtqeWJdHM2luK%!6E%KtTklHVUMSDWM~rTZY6>t7OnZ
zIh5ekSS*7(Azef<9KRGIufda1@R@iz9KG@mzCGvIPa9TGJY?Q`%|1Be?kJBrv+mNS
zbKP8P)g~+3QiC^PlPn4G{DFd%qOraoXj)$x7c#<}e!&
zY!(ZWE(c%-J&Jp#(Vtm@E1~T-L=u$A*&{DG>_u+_qnD1Ftz2X@WYnB`!bCrotBFjS
zzP%i|J%T_76VsGTp^dBqsYAUD(MYJ1!l%
z+%3LD3p#Ne)y&_}PAt)m2U}39ZR)7oZ+y7Q=#6~;W>ZU+x
zz^>!uLG?j)mHF=kyej7M(IvUBe-APY
zF-{RUYW;$F!CITZFTCu)SU>Y@_+&9wr90R%oL3RLtKeAcEhbrYhH
zV;;k7hKX{NlF^EAZjVlRGg`j${FA@F`hP<>A>VH?dS~RKXRLAfYC-B&BV|jLKkU;5
zvhS$BDy)H54NjXz0xrMFlFXYK!T>01gSkZbF`DK}V~U_47#aJVtCWRIiHiCD%!H7E
zV2BM7fCCVni_;#oT|z>Vt@bzPKI*EfV52(S03SixcuBXkU@=ZiOp{q%48}#3ciIF7
z6ApDTeVH$HdCa_P8QLEyok+&0)nfZIsJ3Wu0bJ-q0~fvc}fDdZIm}7VYkD4
z<#^MBDpE~wE^ip8)5)WnwM_LDsp2^sipiiY=Jm@I3DU;hN
zrv~&rMR`TJi);Ua4!@hte5mxnd{*pG!CjJLM@O-WpR>M;^Hf@d6{Fli3Kc#e5^R?`Em(8FRTl+8@kCu-h>Y^G>3R7Be7JxL9o`@gRZnt1;#
zF-J7eaYcW(ZYgq$QS_vvu`UZ7$iQlCQc>lal+{X$s2+tqwf&AuB|_0q@?qFc9s}f=
zZ;cODpsc_)3`u0K^%fY?QN{l1w8{0G8Iu|+yIUh0X2&Qma_~VK{urzd40^j>G#}QU
z13iUKsr36cHc@HiIQ*ZrpjBgjDtWxmP
zhQW#vj?vU`j(FDJ%TrF6pK{*QoT#|Z&<>AD!pZv;0g3?G0ryI2xHylie$v8XiQUG#
z&$kEDa(G5l=SlFGOvj`cRB0<*pied!EHTK}Rf=?Fov0^|fl6G#3m0(L1epr}-zSw=
zWZI_NvYWO|{W@Nt<6zu1G(T;D*f?gf@h*=FGhvGI)kme>eTmVHVB>_l!)m_njy*@*^EB~0h^nLWq>|@q8&3S8#!r$0XMiybMrrXWuy;N6bqP1gp|bQ{@qopq0a!C3}6Et*34
z#p5v%G3UmiXx&CMWjvMRA8KB-zE}gl=9fY$YEy_9Wqc=1gg=&kWM;%=kv`Ix2PO;)6xu
zDOy6|WSNq*nz`=vc)%{G
z<0{q!bp#*xJa1%zbIBD#IOk`0Im}e~2duxUESq}7!+Jfs2qi9~L~T?-#rX=~fK`s7
z<*aCAc*a$%fHvFtQ#N^MXGU5bMN9m^c1)U27rQP;!Li}r;S_=jgvQeF2w(!$`iNV`
zFMDuRon}d#Dttyhl^2u2BlXS}rom0I(6TCv7g8rVKh4bGu|2sVHCAp=uN_y4J>fR+|kOT#9{5sne1c`{y;JlLtoQ~?+o5h?W
z-nvzBaRpXD=cl=m>Y_MUo_k3rCz-~)<j|ylizna1-HF{CO!cKfo1zcd20M1H=|NMuI!yYQL>L_|;)O!2^
zT$*x&OxrTaQ@AE&(i9o)wDxXJyMDmAu90~SrE}gr4kLL`LAe%Fkeem;Bi(DqLt
z!96(Fyb*dqm7oFEy8MjLTmt?V98h|XH)Zp}%0XuVo4Qamtzgmt(wf-g6c7E87#J%ki
z0ZDKqu%*RC4cQcS^38r7brM*$9Lc_f<}$}vB80T*yg2OxKy<#>Km*|`9Ca;&t|DmB
znzoBpv?nEV?F7Gb?M0ef9ac)LhMXhbOsIoVIY)tv+{0bp!rPRNcCcJ{s$Pm>+Fg=P
zIn^3D5|VY(aX&6?@!>1N%vv8yWpPwryWyaB$%)9YCEsHnJWd(DmztYsKlbHtMN#9N
zd=T)QTfKH%0|sNjP0U_8N6OqVT61gRC?AQGWb5S$zf0Z)dh24*BwLKgF%CJJGP`9q
z?g|jvz;fesL`%<^?Nr{4DkVoP+2mhAF~2_S(lH|&2POBTZ^a{tJ{b_p&mE^J19lfO
z5=Mv_3+S6p+B9>Yc+;PpooHf%3ctNkv)4+%9=HgP`=h*~BI1c+t4{)bYLEWYurGq#
z#eifp$WTz1G?qH*nJGcxI-&4r3RXh_9Iv4~iB*Hz%v%B5{_IIV%AtyJY*S}|7`{$v
zwO}ezy1nrlNFTp8_-6fr(rCd#>094KUNiU9DWbNwr+$k}yOy+@G(FNGws=-e#aZVZ
za$arfTR$1Vs-j>l*|xqzobcHPf|yqh0Z6ZfmPKT8#6wYI5_+G!`T~G`DK>CWh-=C2X$v;O%|D&IT~XbZf1)z*E>-8W1LrtYAgek&ApnR!?4q@4Vd9t
zmZEc2rz?76iVy9p^*Uil!oz~u8lwP1T4_)(+|z$N69sGYXoiP@kldA=^5vPyQu)0F
zw4p_HgOlELKx8TT@FWa$5#c*&n<993+vZ(zshB(ajV!Wn+V{N>4WCbSg!p4NCkK%Q
z_MrZP&WVwzO#6>eV-J=K4u;8s1C}L%FgNZ$&wnob5)pbcr!86rQ?BSIaLZV!g*>0V
zUz`$fx`6hhfThu34r=pX3dglF#A>=W3vA2Dskd3p%#z5Ej
zCPw)kA*A5g4hT%xK~W;xU>$3A?$=Aew^ROozw=K*X19t|34npiaH`1L6XDvcH_^ug
z{_8<)3hV5OwodG6Hv$scH!Y$Z@Tpjow%A}|dqWQ-wse=j=V9)q2xN2GSSpf{3DIjR
zX#^$2iVyJq_4#-IUwME1TNB!(j5sbeP;j*^I}57Edc87`EA+njO@~-D_mK5MRL-1H
zqH7{hqAVCTbFuPX#wFyKA{{XZ5f;GuJ^EGVjAZi+M{M%)85!NC$Pz16UNGV}A^
z6Dd;KdAfQ%E&;#SIpCbS7*Gxuo`44*iiQy*<&7H%w(El~IY
zT+CtCyb7UIchbQA!c>-rAV!>`cOAQfPaPh5(MwjYxF2RM#D{yA=h8pHvyg!hz&YI
zI2A;@h#*S=Lm+V3UUMit`Vmz8^|34;8irhOl441{FAN+6JaqB}*Rm@o5!KR>8t%7R
z(_&rq4_O*7swH`%^{}(!Czd6>gf0E8(WTFAK4qU4>x5oK-@}4avSnV~TyKrjB^{nJK#!a?e!dlHM}&@MHHD~`99*>a4KihF>;$xt%q`gw
zI6rT%h&3b=5&XW6EA^GkgDyHjd5p!4aubwNj%5*r3N7gzY;?g3&{YiWs3NqQiFK6D
z+!5Fa-dm{8>rG)fy19#VQ`xBx)>e8Vy>5C3eLIOqXZK?Y?H1XVHsEPxi~tCEI7Yn)
zq(RUNI=yGIrl?DTe?C`Z5gqgDE1u-L;jgGzIYlxC{WCcdS?y~bLFbAA;2N!(o&l&T
zyc}mDQqoZ&9?8TTXvcw})o@Y%C?kAJq)dZy#V#%iG801er8B!Q*1}ce)N7a-HuFeE
z6sfVx$v*>xQw0-C*|v$KG_nR>ZxVon4XBL#dCuozSq#Exa8hy~I%-s7hztAUJ`*$?
zNF$I!53G4F{}plRccD7fIx6E|N)pR~*Q8BQSF(tP=r~CgpOLxWzgWq!U3lKP{#kPl
ze_wjCRcU>#Mvarg@I{AiGkEG*qN2zifvKEgu`C>vXpm{Oui^q
z#TusJw8a&n_X@kHFD}4ve)i#)f1JA@(;}Ly9&o=kj@`f9R726$Sv=)s6gi7tPu2k>
zem0w))XxhraK}{~2v9&O-+ASnye{Xh*URQ5yxeJmgrwc&m}?4MrqZ1U@4dc9jc}
z-Q*G_8&qz|w#Kb!Lf@Cv3t({<78Mu~c3m%KneREDxUVr6*IRm_Xe}Xk%3bSM#^uzV
zuq5qFhp$(Zit1D}pK@|)d+egM3%(Q_dd`%Q1&pB?KI}k*v}U5)JV6ot#A;dG#F${dG70Y
z;(~M2wRnkIp`nlD=8TgGY;SCxDk^ioAi=NEDn}Zq
zKwQI}JVZt{FATbs?OIu<1%i}bQ1(naWc}pY@STjYS!rPjEe=-;&t|@Brjyq8MKzQV_^mViW*sU#iU$u<3`7eSKqaH-rGexb|L<03>my~~mHwm6a$1HfoF
z(ICfZ8PT7NsL>NWXgrk5m*Pud2atu_>ZORL`I&bcnq2l99w%`q~{6JP>f85t{ZU9M!2yt&d0MOx04rpR4!0_69>ju6ycS
z$YMpfFfek+BHtGZ-v374ck~OoQ&WdD77>gh*9+Jg-&?GO=z9``qc^@B%+SXDcNRI%
z+foMsr`aDx*0g0WY^@{ln0hL`zSwT>MKd9@PN8&zmXwZo-SWp{>J~f`0+Zrn-dkWU
zC}Vm?NBXh>6&LFz6_;G9h6~?69djS1cTo__^hpDHcFx@F+0Th|^PxnaUcD7o!5_N$
zv2FZ|4z3DzFb_=Vk%G!DZ>8k;oD=N_fW^BB{!|>zwFuf`y?pJk#vgB{qp4!#xr`)q
z_8NI#()qno9E=9m3bex1l`;<}deyNd-Lve@ryr!q#-+A3UOLUkF
z0pOc$^9IcihsHvzNE@ck|1TdLb95C}XbO7H*;c1cXrvm;Dk`$0n-y|Nh<*vQl{)98
zgrO4OAjoPf#PgnxkYwVe0RLXn^U=E@75IQ6t;_{g7T0P*I_4)uf5MzFceW
zA7r!0RbMhmt6SV}Rnqs7C!tzK_Bxt0a#%j51(_9NGR9604~2UZt;2obPZA`caQV59
zxH+{&u)s)6ZyAH3<^mNAapP^}%B9I#L>eF!L(OQ|fo_f6DB1_nydh^=b1%4K7X{hV
zfki2%1WO)^+fx3%m_a!SWl;3%VQ&JvLK=O&Bnt1gn7QADzj(j8UQ-K)g}svONq?x%
zvHV>dYs$&GY)x_R`qUE{qU2uWQeonnwYQO?jIQNexT?W}dY_@qiaAnEn*@8anf4&)
zyJSEO*1CAkYE~DB6P!@-E&oOxObKGDNb;%+;>^q3~4U4$CLd!$3#nNuh>q5AH7}L8*Xy4~jCNfR%6h
zHVAORvX7m~HnE`@(}rp!cenD^(Lni0Y&p^tYGqwIGQRQiSh0qKI*@X-0xWvhhBf6+2
zX@1H14-4SHF^85ZiAPe9HBJeK=BBYo>7R5|o^*K2z=vy~s6N7!8Js@>gK8PL#8Pl8
zu;?9vh~6CKb#+Xs(v`)^;j9>k^|W&hqR*s}q985%zEJyDRoCmJB)r;U^cXf$vc+R1
zux5%Rc({PhjtCcBGA36VL!Tac^dY!Aj?-ssBf~G&x|CUe=^?O7m+>a+0wfA5j*}6a
z6b1~NFlbP}?_91bvFBxwlm||}1in`NjH&6w&x$Fa)A!Ry%0dofk#)L(+4<#mv|Q`U
zpo#PAy=2tpH^Ns_{6KjU#_BgQNr_W&(G(!X~=%>W{>
zFnD`q#%#s~B$mJ=)?z?#Zt##9y
zYj2reC^?9rjuK!y#X$>Zc4o}$e0G7wsB!F&pBlo1An7!0>iQ(fplC4NqjhP!en56o
zk5q@ob_88X{Dxm$&QuM+4h#=mz#f^7@Ck;}5cDfA8@tpf2{Mp#iQ9X(M>wB70v8cC
z(DIjF$#aBjo`OckGKl%BbF4v*#ON?|S3zGlHR5%|)qTV+m!u)D-WnVikD(W9Q)tSW
z{`g-1hx!naQcCmI66GHBcUM&e$BC2{^<%^#P?7KUXmI}L1NG~Fki&XLK@d=+fDrb-
zSracYoq=R9f;DB3VAVtQ{g>v5;^MPrpU%^(AT2WVp5~GU${}2NecUXk*-I!AFQO+Y
z*cj|x6M4d!*#puj2qoFlL(UQF2?A^~)^e`JqRv#xgcV6K(Y{Y@P>2tgDEfsAVZ8l38!m2n%*g>oK}pnS8=uf5u|$iq
z-Km=5Ic-@k$xd+oT9{&?lO0x!A*UYLcJyk%`6!j4W=DD5cKqM9{vkS77gXqV&C9u0
zM3pk+?zkPd{Ew`dsS=Rk%52G0K|6Zw4C=*-hA^w;*0ML|y=~lff#QI1Maq&14D|_h
zpR?er;DnC1wEjCTClmyO8Xx6Dc7t;auc;O>Y&(mxh1Wv!l4VuG=(I+3Mv{$>y3EP;
za0B!LvY)p##uUbB@?ITJcL0%k2iQ4ZW|YYS==Pd)fKR7T0QOaH^ZkawGw?7>&fVc^!?T|;4#T_*>+ZpPekoiISBSq_}*{Is=!EK?Op|%s?%WL
z)yyl|KRd;jhK>==VP&6pg8{G@R?YC$X6&QbT3@9sliMD><9I+I1?%DD-|WQ0xz6o25CBIOud<*bzl
z`^in)xUz{zmNx@AS{7&JC71*Tm^-K#7d{puWWN2^Im2Q>pqljH?7DJ_RJuTm<)kw)
zQ;-tev=H~41Q*l#(bxY|Yn&RFQdPP^IvK}B6Q>UN>d{o#c*bot5cQDJyr?jyD8uBm?Mp|uV=FgC(aI
z-os4k@^b+%+)w6d!it=GTko!Vsg2fE#X6JouXEpVcIikl3Ci$Jm1K0TosZBFRzZQi
zhd_DcNCS0Hu`jUJ*FYBZt^kp!rS53lcPU?2_lNese%Lzvh^MM37drnBDNF7t?I{b1
zi~8!<%iqy<85W=?#OAvgQUR(8u>+Aa-
z0G2=5M);!wiTO<>q-QT}juddEovaJPj5HoPx}?Jrt02W+K4&ZiR-_<<3<3Df1woh6
z&bDKC0T&Bc@_o#gbCzLVqhi~Iic;=HnsS^%*1UmG_vZf@^~!KasmdsDE8rmyfwz*9PMmSQ}WXgK+vCqt|O)^+~R%3CQ8{J=j+nYuINlVrr!fL^seb{
zE~MS!dgCQ9ICHuLuCz{~JoP5R;9Bs{2<(0Nj}i&?EZQbDtwD`!b&6MJHz?ebLA8&G6o887(kIuhj3W$G(>%w{1mS
z>^q;KBWD=HIFlZJpUc;i!%;fqLjWiySm1=@VpRlLoSgd=P~4<8I3GS5t5X>_6QauW
zwTG4M=P9+p{l%ISfJd@OU8lXu>i#Mm6hO}$uM5d1ot|om
zEE)!MqX((IDRo*crgtb!T%9eLkdHUm2xI)2^v{h{b(1#vVGF%kQT
z1p9E@X+}?kH4KY(M5lTZ<5cf|G0E*}Sfsl-xPCj}Mq#D^U*qgd{l)9oFH^f%T*40)
zBI;BN#_0NTr3g9JH{^35FXuTQ)1ds43YHWVyNIbEP^ap@%|^ciUXBI>Ze<`cMf+ex
zb!vuid5SBS#sxfSjkem_x%4J#U^kuRh+fyw5jcNXv-whB6OQ9WA+!WDW*Tx7zVD*~
zmKft7dj8&~k>I?C$w<>!VJcB`9ppI5K{GiyVeLux`46UHNPkNa6?rTP=Kw57L%3hN
z)@uM44e18@HvrmVYIqL)QSwMih7Mbz^_8{|+o|F#7p1+kfam(>THnVY6T?=`h-M)}
zi8Wp}&^9x>Z`G%0{H`MhE3{wlVu0G;A`gUQsDDoxrI!?>8dqxRCZT
zC-b9i@rjC59+A;MsZP8}E&MJF)6i_I95NEFq%O;Mtjposs8*;!i3?Gm{4tIITR3(!
zObjZQTV_8Oz$RUb=~Vy_<9$w>JZtj}hR15#==dn(X<{-Ml{x<5W2
zrwEE2Yq&#~9!u0#azFNZ&z1TBt1ag~_UV)Qhh;
zIV%(sr{=iB(R2v2m?0?7Lk_;b$ilAV&qVFq6f-uI<4-yMgt6!5h$iUX&^8XZPF)6F
z+!Lb8C1Kc~pu;n@rsj>J$u&=&-h>g}p{{9~Z!5g%RpyF$2NmwAOQ5J;``{W-n2h)x
zdERYDYv|L^^%P*QEe(Q@E25oQG6{yiyFu5U>CUXQrh;XlvXv5gQOv`iQ$)erCmf;k{O#q;*K3
zM$3?PtTNfsd`!;0CN%EC9*70QBN1a^<5g`O+~;zEp9Tp)b^!~nt&Vw`OXnW0YY%%(
z7i3a#U35sol*Gf=n=rCcJ;T1!m1TTvqoV33DCT>9|z5s{>WZ*30r_FhW|S@ie5F~{nl3!lrMfi{I)7LSn1`PTJpnWdj2g_4qDi*@Zm20_wB
zabY1HpaO8%_eCOE2M)`?>ioIM1}0mLgXvlv7lR$W7FaCJ0a+S{1E3!hZR)7U;`7mwOnD(M)QpXano6kz(4T<-
zg;v)|kY&DyXu$QN)Q@nguaKN?Py3EC19>Dw4-@_Am>TBJ&$PGabG=-Kk&VEe4D^0w
z5Hwc~j{{OfIA~tuo4wAGlS%KD47OfGvsCP)=UlY5*cLe+p>+J5s$bHsbhNrOB;gJ_
ziu~R0l)h9`$fCJrTw$9v_c2xxN=cMuOWWsl5jw6Z=#z04X!3WYW-boYSOlsx3&@{3
zUlLDQZ97)+QYN~h84%zeHn=pR{kn1&En+RY!0Ftv?l;!eXhpVLRhDaXbl2is%%b3IZ|Hw<>6!qB;wu*c_JUo;$qP$UpjV^95oqwD
zLeoauRqNak(_)SW6E^sTu>Gny>3Fw))mYMnDk2FK$qW!k+Y_tk66HV2g5KFVku%Td
z3bD(fwGmfyZJqNj8UW5+@&h^!*9_F~mV3iAz7-=(de7om^uYK?RSAy)6$=Eh>lMur
z3ZiomRL%4I_e0^xoVSk2hQf7wep;^EkozlbV3<2qod9X<^Nr`BfwNI^{TUn8NH!({
zZDVm6p&}=^snL{HpcpTm2ZvV7b8%ib>*4|7hzxKS&@_WPWzmf
zP1GQgtGh#o*G9)HTyxU-_!AkD)sQUq{_7zsdZ2^T{YN*bR$KINTws>dh)EN`hhRQx
zi3=IwF}jj2Aq}%-FVfh6L!l3WgJ8b}o&?d$aw^PVlJq2a3wL;(BHCzqk$?20Htbs3
zX>jgb?6>%K#!bBv!?u+Mi1B08NvDew^JG>43%9q6d!6dzI#1J{40I#Jn)~EW$_rBe
z%+_G-@%HqMk`NPt>H0VyXN3f~!BMkZ^a;h(PZtHg0*-}7Kx$E{ts2J=c1%SWim@T#
z>;i`pB`YDjtcNZ1fCGD(kh9VAC1m<+nl6}1x*U>_R)4L29%iMDrFN)_fF-79%y{2o
z`T#l~QAwvhQlM721~?v0{uJ)zH3|U+<;vtYd)!ajp~DU)(jWl@mnv4ILsy}U!}{DEqn4eL{d_4x<|&J&8X+qWDRcGocHqZ(3IIYG
zq2zrp70vW|7so~_-0gdHP8jBcWMCkNl`-8JB{PLr8m|j{(debtDekk}6324ojC(7v
z3l;1M?_=X$Bg;85;JwdleoK*EnwAv$_UlxWZ|FOb&8vG~QOXTUdOy>92SCgSKSb5)
zMND~tCGKzOm=%G?<~;+lF(o*pM$c!Dz%QK&%nD$mdun|Hb>2INU8H0)BOt~r%uUB6`e@sIj3+PVVHlZLhL`;I>NL=T;AtN)B!4)xbzV_f+9
zGFWhm!p{szt85?EkkMd^m}^6@AUdzTqyr7?f`BV8`KXX3RU??`K6&dCWrYq&sU?{i
zMSCN282>oiVP$TM0uVPedgSa?13>DBQ%)ME3+B?}_4$0+KY|ajO{ysXIGkWjFB$4z
zdOPaxMEESugu?6Slh46V*Q`Vs?YzZE5f3BStK3QYGcU0UCc=YBxQ?en}S$!kq+Vm1N$Rim6
z)h94zxMSpNJ=OIN@)eRp5gI7o46-eSIfrb^4%ctw_1=gI11HtAe3H3&J7L0~p=^m5
z6s_rMJn&6+8r6dr)_5vH=j_QEbe2Bn{RNU11)d;*#s>5w`jzyX#~0z>FjZcs
zmwC}22KT8~MmE7sJ<^PwSfV1B>sMia74Llo%#U1?;XX$TPeCK?mY9(4U&50_l9+xM
zfH1Ur_gjkOsQI#oaheiNk+S(4EZHHKw|riL!b=u6Io1dg=@Or-w7G#${wNrK;S9nW
z#ffkPqJppq8U=FVayB&9=h<{>a0>GjKT?scL5}+0$GmAQ*ytr0A=BA~Hcko>Nkh*D
z5;fQb6}+@CQk(O60m(g&(jGZ)z495<^mv8x5m^;ia=S=o>=D(`NilN~29(XKpvKW^4QqE4Y$o-cyqAhZ6(|Ojs
z-s)GR$PKW@cn>d@vww?pD~98Iqcead-77LHDGlZSWuUox4emx)5XDS{CqM;&O5FlI
z%6o;+tE%<`kZ^fUP{H53u5)cSGPg+_hKT$zPW53rNir~dDD^PWy6TP
ze@>VL?)x$?(bImFlBpjc9h4Ehj0d93=V_dUAOVGCe2bhq?(sf1P>q~>oI*T^7$S2p
ze7I*n{f&_iexzInf|#5-F9C6GaND(fjZy6UFiY$;!Q_)O-=ttGGWW~U%t(wH&W5kq
ztECi9Q18RCZh(FPJ*~~KF=stkl$}$HfPgH9T`j4TqGxR`n-*zKI|*>dew>g0>hgU>
zC5~{vh1QS~C}MyT!w|Dfsju_-{sJD_)Wf1LDi%Y#DDZjAJQO8+hk)mmvg1DYy8!gU
z3FmhysBh@TP^VVVcW5cdwWw|+$nFaH0Rpq1UaTWr%qyyL7syV|o9PAW;qps9V#tEU
zXoSvW@;GI~{PTQO*c&#>1;s?>8mNjrA}7x*wCDL})71h2DsBW-NCWriKJ&;n-h6AB
zRI7){T$1y)KXgBY=@B(g7Qkius`_MWF4PwUtK-Abq6~jJlCT@bId+|Ac1LhhTpQ*BR;QALhI#6WX7D8Mn!!)4Iy}
zP)YqF92X3gbvNnk1wJGruIpeitCjqguN;6=ujpPnQqr=E^W0HmR5i~a7+gHxE|!Zj
z<~Nqn*^agHmR)kDah2CmxJYf5)4&A|CuB&43aNwWHYp<}-%^q!fQF1=;WFp^c*{l2
z$W)XU87gYjkyIhTGN@p5r*Q`y735SvU|oe=l=B+HD^OBiw>^cB9##BQT632s-J%w?
zTK9yr4>EAF5m<)Rv^CYG@2AYkzYTKHp1HWeVlbUKL$*e9p;C*XD;;|jbGF7oYtO|@
z;uthP(q5kbJY0!UeLDAG-+Bq?)6b(bKw+uot>|_uSAJ@M9y-wgm?^OfvQyl1QT&}w2Sm)&RKDeLs+*?H(b*NlaJ;|tV+cjRzb0?C%#a<#8&FkiE;dV8vB%G4gYckyAP
z9u=FP>-W?YZA%oo9n||5uCqw_ILgUeWN_7!fQ648cxWDC7Gr<_BZt4hSZzDcA8*+H
zeUcNntM`)&u&?MJa+>~J4#a0yDJ_cDm%M~w_vLbFAfK9wqAn!sMsWkh@A22}?Vwq0blTZ%|$C<(=U(!u>kM-V5}^lVQ-j+qk?#4J-T?
z;P){t6%*wwlh#@5)k5Y}$}v_fT%;kF4Qf-ft~J&FDCCNgv?XR&A}1dg(!IKj4EMTb
zd`9x2RFME2?x%L*GAJb>l)5Qu%Cx20zps8Zx~w=^`M8ISA;EguT10Nw<2dl^10(
zmIlU~sR>VVl5#L=y5knn*Ux*;FH7Jnt{0RjM=ZBtMu-Z8N_xo2$3;?g?J>d8sDvGy
zfMSXci%C~$zf>%^O82D^UZj&#>b*jdmiJiC3vJ>!WEv-oMY(VSG0U==?+~0>6traP
zVpTYH1NIdL)ZI**Qtd<74MJJKmM-TXbbvSmc0R^dp((YaW#f
zIYRL~FiuWanf@IoUM#TFlBp_QOozdQzDIrR69bS0I{}?ype4Zr7E>5f0MdYZLzwgL
z=eF)GPWLD3dZm{${4~(-t>)~_9u(Yjzh_Mkd>5t!3|oJ)bsC_&_8Bz>`25UAc!-V(
zLS3wxin8M>7DUu+!(5YtGgZ74{li(}Gt9q2I)dN3S(z|X01QFp?#`C7IF}taAsaugS>Pl4Io
zzz{gv210V$l8BM7m{zoO;EOEGW1Z*xf`(f2kGS4k$Uva~q=pyg6nGV`mQz3i7wr)6
z$7h#)o9RqseXd*kC!#E~F?oSC%Jh}PFY9K6UgIm}HkqU6h4wdW)K#3T^Xz7r-f&%m
zz^y=zQwSCUZnQ@;L
z{U^KKK`-EcOkK-%_CmZoNyBb}ox;wL;m-Yy&sI#Pt7mJKT5+JR9yI#D1+colA4D$K=D4VnDeF4!qgxG51oXKQK;ppjDU>hIDP@X#W}0PLlN2P;>E%@
zHRUO-K}cBBpe8&n8TY#8TkJ$wMAAd4T7KQul*g&`vXQb~h}m6C)~$1ck(+2%qVGkg
z-oRo7kFJRhQ%3+%ag8-W8;)hF
z^&I6Jwm!{(iGeGPRi>VLh|9;js{-D&K9ytq`NfnM^01Q`H`c!-pR=*k?g0AMwQdA!
z*Z#t2rvZtZ4AGGmWyTTfs}x0r>K;d5uIEA`@f{sel+5Pk0t@$7TE*+R!S0|B1DJw7
zd%vio1Vr&e-E*JsaR^A;GAuhFG-1Ny9!1C*MzXcH*Q+SIr@p2bWc2Z49{Z83CD#Q2
ziSkwDBY>`kQHrYK*RhA*IWY2zZ7jd@V%QMbMD5j8Q**6*%WS9`vNR;tM2wzo`os4W
zliC8!@j#8K!5MwzMTE2Bin6r%E{RIF7%qa-c&{L@;?ns*9{?d~R~|`%gHbvmB=-zP
zq%{(s%^L)%b#a~X&RN8~&$ssqkWP8l7#UW73P?K=QE6Nflu+BJpo(CG$Oy(kN3`Y?
zL(stVF9}x8SVcFVyJ^F^Yatzw=w(Fy`8!u6*gImdYh0MC--5WwhMezxI3iL;ZYPXZ
zYyUFrLs74N87PgBeF6lw7WYIASq6O7um=(21nAW7!eu!h$wKN~0O(ACxpJltkGdhR
z57es8fOEp4ydP5uA4bSBY_wwGXp8soR?$H%4zv}JXTg(iCQC(vLefG)@&`0lE;!V>
z$ymzCGc~YjsFHgj=Z!rMq@ie3?_+Dq7aK>`oXYTzHsvIAHXQ%)Nr-jrcWK^N)S3}0
z-~~;7q9h(cT3m~pD$a?^4dP0&3J+6T4PGYt?`zd5j(eb5h9}lgt}%exdD{rn?xAO2
zPJbd2Hbk(z7m_-XMN%;q@>1
zAFHDiYga8{{x~NA`18+4b}7T1E_FIY#dPj~u>^#|dM~zQo*zFEwohtC`rlyPmF%>{
z$;)0NAgmw?yFQ~sCqSgpwhSZ}0w>7?@XKq6(=E&QEZY2{5G297?T&18;N22`28nW*
zkcDY#Rc<15=6_h3-JNm&zP90!nmYh0@5i3+?@2RsAizd&#R5w>e(FpKgvzmUXvGvj
zYsg}u2_-sdolDNUjR;Z#*kLH)o!@j0Rl>+ldQ-fW9zeg88z|owrGXOgx?klbJ?Wi
zxb78-)I4dFLk1-8p%3ezRn3^F^ZK4fa##HiBv+Og!dV{E$)6smYMw;exE0cib~EE8ue
z-P7L9n$p;bZZ8!{@dU#oW6wFJmnxiQp);-&NyU}zYk?LadzAZ{amkLf46Ioq$(Hlo
zv~7NFf381uFO1`J4N|tE3)~n%U#zf;V+}MdNnAIskHHh~t+&X{!mx17jsokfb%CUA
zr!W>LwP{y*)0yksc9;*_zY4u)uTLt}W$7s+H*+tCiN!-M{Rj88Bi}!3Nz95dEIu=p
z@&QoL5pvXdm_O4w~9>1mg2Bq$YBsaLDZJq&xKq0gfj^+)*yBpyOa1?UUL1y|=fC)LcWcmQt-RKw1T;x!y
ztBG;{BmfYtp3L1cXQujx^K%}_7^JnvpptGCYcR%pIigbKxxde8-tAvyL>?~Ot2YqZ
zIcC2MLUjZo!z^VcAt3~OUYH`}>pl$}gOTJRe2p7a$QhgTn*kVFOP`
zAv)6rpK|K;Ox3kqG$|EOO_Y#&kriR33-!cpkYgfR;R68%o#cGHgq0>MCl=h{jE4i=DX(Jh>}0V%Re{7O`qwjh%ja
z2uo-;VQa#{kM~7-pPpUYn?)O8K&Nzt3~oiZDU}=>|4ce>gXM(HwC>2uk74Q?p5kw}
z0S>+8v-p-$fRRt1H!rrdmEuA6aIZZDscPQ9dGCJlj-pcdo6;Hi%_?~Yc)k8xYH!+)
zEO1OAoRm>>6@6R+m##KWUP6c^?&(BVJm%v@Ad@Zda
z_I?fbf*p@*AD|LQNqh)Iqz{rM-RG>?!*Qw~N?1=~fVkL7Jx8{!&|8A7fFr~@qtz=E
z{ft++3{ey6zNH*BFLESpe=Un|I&X}E75k3(Gp3*|<`+Jr)1bjrBsi4ab^V%!GF8UY
z3`2ZGr%Z#;@llSe)sbt?Oe9#D>gIg?-vU-Asf=y5QZz
z732KQlh^$%A88qMISS__aD?|}gbRv-W8Zen<9KCpTIR^fA0ARU)uJ65Dj4)GsrF{!i
z(I!u2H)O1gl8VntAYc9Svdr{o{|
zv}H^b+~nOR9mg177RVX=(kUpKUl)a{Kya={6~5EbI=-3A-P7ss?ns%Qg90W+!P7O7
z&XqE3nn3L?B{2|tRm9KGWx&BIumU-c3X)8Y-PRw|K3={
za}%`-l;gzL=SV$}Acn#~kgcXiuGf`*5mg=o;&Al&jQG%SYz5nGzhRfyR59@PDIYO`5E@MfLc;EUAujpXM}mU@#u%M8W)f;Ey7(2S
ztA0T8+t>5__mZ<0c}}J~3-riCbWU}I(Jku~-qbu-3(d!>FMmz~=Z4TcO;WzMCAfIM7j|rF-?9Q8IALbpn;a
z2Wzm*nNvtNX$Md%S~rDc-w->EXngj7s_oC&I!sDPgFtT`bY^~%DT+3`poqaHS!~a
z-O3(`rRY?=Bl;EmeU0tz*^#MjNjpp*=aAW-&NHu^a~nCr+|jZSYG#EpRBp3rS_WB;
zuVsvajouZ3-h7@^u9O_OmXdVDkCCkI;r{Z^X8=S#y2MI+85*+~%P3=;HYmsUFlt$eTz4)dd%MaOkTTK4YK
zE@Lf6x=eS0k!}UQud_S^!Sc7}>d+Ip5XeAYLpl?lCKX2-gPkYuD8B;o9~#t2hms!d
z8Q0N+v^8ERRH>YI&Y$-9ly{Wjw-|>_z5D)wi2v8(Om1!MiNa!Ae
zRkO%bu=a15IJ}66kmZn+#X*HjhdhMCf(NXuxI%cL8jCQZb2yu~Vd$l#8BG4XSO!te
zU4~JGYkpR86fKI=;n5Cg}5izga)E`3zL6*-F{ufD$VrP*_1M@?5uU9QyMv!kTJe
z43EbyzAosGLAim8)`YcPyudnlJX<_npUQW@$r#F%H$&9tYKeA?-I|9UMGVi(`dxHu
z0tI~?I(-qA=a*pLcl(oy+}e{~bDx0X0U%Gy$#&G_I)T%CBAym&5N3}6bg*5K@z1*8
zb=yeR@JW-QE*8s?{Zf}acm%$<9i|xCJ35(7PvoQdZ_-phlw=_L;-5gquCd6Vke8-F
zjq@h@d&1hcDo}u2@_9a1DB0rxePaaGqWjdK|Et%u4Y0||+J=h11!yAeKM^*!j-vJf
zoe|-!gcLc(HHYJ@$>YU_M51FfHqA4jM-B~XSv-o<1Nb$#43qsu;c3EFc_4`txwvEok!(>h}7@%uMPI%|=sJp@u
zu8J!(l#C+?;itF5ErA5ad7lk%YccWmgb*`EaecSbAycTyHo(rKJ?52GUN#QZl6)%A
zWhj^*DG(c13A`f%Io=}ov_kFYdapYICV>*Byq+_Bg7F62o-TyePBr`?gAt@}=vaVv
zTYS~lb~L2q-70nH>j
zNh;Xmt(Xc@E!^Vm3fHtJ6GB=ML?;}Zz+<7#PM8D#l5QwsJFxXS_Y
zTBGdGt>;}`1CSPg905YH#`UZ^uY$|EX(Amq2ch(R=Lk!M_ClUD+HKe;oo?~*wJ0k_ogEr{>b=HPHjT`4vq2H}eBh*3l|jA<>1>~ozewN)T@^8HU6;oBJA
z7`zJxWDFPbo*oyEO=2%Ka`jLQB}T7Yvu4`ug5t*$eSh&e6T6GFFktZVnLD1M=&=is
zC7KHqfN@xFD@hN3U%Wq&gR_NJ^F+5WxmCY3CkhI0EF*^uG$#&Ht82zm#x{6`9R%He
zpA!KUKgSN^Zlfp9HJNsjFrizy`+G!impg|(=@h_Dw5}Q-fopP77RF5PQ>4h6J@nlA
zXHC~UBMQou_45J2=!D|vX~{Swp1ZUR`gAv-Eu|rllO?`C`Fb^-+A7dK`s6h0S<HV9sYI>Ietk^1Bp0rcQY15)
zR2|~bpEZbI3v%yA^$vp2B_a%aZ$pM9Rs1mpU&E;QRrbJjT?{_?KJnKh*M8uyjGo@Hu7@`b=h#;RAn3Vfekk0+t2q3(
zF1pL0I{(%wfMd`o?wcJvGA>x~8e{dO8Wxr4e)R#^;bN&ODuOQ!C>oI7LZKzU!{0@3
zhx8h;ek^Sj>r___ifFhDSwvI9j=#XbY(i*1#B2195W(A8hNU@j_kF9w(Q*bR%0gcN(S2DQJ>PZ+b`XVrYZO{Wd+p9Cfu@`I
z4#%!2x~s7>h3UJgRj#Kw4ddQVckhUxp`UvQ1UgC<0(2G2sy9f{UMRQI>R<7TWL2x_
zJ|l@Onfo|!>qP;ZT`t!aMD%Y24%k24P7mnnpHB_I9B+D|)p>>#Wa3WhGmJY#laf-X
z+k$EZne!YfK5s*j6FP0?OfS&N37zA8x1o4skY_OoOaE*SlCcV*Clrj&6Bhtd`wcPv
z7!RK)wZ}`f0WK*q*FOBO;XL20F)|R@U&}xi<7BosV5sYT=N$kxlu6jr<)Llxq&V_2
zA{c`}0dW0{Id{59iyapz8m!F`4hgPsk7TrIGzC5q6bx7ZIEGeu`FN$_s6!mjae!uj
zdWC*^jtpi4}LOseK?I^=|yz>Z0Bp8DozP#*lrXmScSYJF#?MMyCBbBw;851GC6L71k
zQQ1n_9e`k^SrdIkX)}(xC@{R84Mtzvk@Kn)&E*93UU4G54C0SnmX{Yxs{%8`4u
z-sWZTm%LN2W5zWHtz*^&P%frtA%g8=aRIc^k4`+5r*4q9wKFKt1Mk%$5l2ZUUbGcX
zzr#8APznj>3$Q>G@TRrXXX0})I$)y5(dmGN_)c8ngRok6MY;tJv
zETySbB)PCuimXLDsO$i+zS!ydpfe(S`dS1jUm|*dXyG=i(v*Dn6s4b_y&^jMs#gnv
zP$lDD=lCrgz%XM-#Q;{OR0Yc`VF6g-)b@ujvq5pHcp=TI96Ex~^O_l&UIp19>lit*4*
zXPAc!!uey*EU+IA$pbP*(8q&lgIyLl3^;QOz?;?(rNZ;yHQOqX!Asn#Rz0xb+aGof
zg|_ZW~f>NkqDfFn;&(f1Y0jfrh8X;2q7z
zbe@V}!dfzJROp8zKT>3SYh0{>Mm;4B`o8CYvVQrv8*?ashXT;tx!K-R(BZAjEB1xX
zOaAf9@-n)~_NwSyrHyt3yT+w$JK#IKZ
z=Ui?BIDEOkcjJ~ch_j)tFzDVgn)G9DJvb4SKJB+b%%dm)@1!L(o#@)DgFlsGNK>~x
zW_AXxU(X!lQW#P|(1Z+!hvekk&WjtYFy+g#n^jmm88PdV^wL}ml~(@A$Dv~9J=+ia
zG90Oxd$Wz$=&)Hi!sZ6P6e{u}kJW<`7ze7RjGR7?WpDS`7%?aY`4qT?>aCUl&aNpR
z_~(Uh7##YU0y~F}+6L(Zj0~2BC&G0CCv%RFnirPFNTkm^#NVxxIYBmg_ffb3Txyp|
z>T1<7{CWE*ykYk;FP$-2sAYP+gF&`O>0MVoT!(?Bp
z6|y9HBgJ^uZd=7QlWL3IW5ybVI-q8h(-cnHR;qRIRa1e}Bd}!q0K%T+`=hbU3g&4N-Sw;!;pn(xb)oPAU19msQI_g6oVKV{d+@
zOi9KImnu+PuMs4q$-Elsnx6%_r?-50D`p0}HFTu3itdPdSfLp>L^s9K_*}Rd!$U_#
zr{dK{-UUr@dzh2nqx5&SKO(0+4<5QQB40*7YB6obYC=V27hJZaiTq4PxJ$}X8J)*2|q^wCDZgw0UHea=oJPD
zZ-|?+YYM=qcl2O_^D+GBDPDT0l=3Nz;=Gfc7Lks_Hh*(nsh#>Efp|xS%H>7x7oFxe
zIY8rmt_$ii3@FNDx`3(HyqKdwt4vJcxM1VT>Ck23Xq;geRQf4mwv~^=&_h?#pj-kw
z5#6pgf~|(J=W+b>oI-bG#hThRvrNG
zdIet_`8Jr_LemEBgzmJb5Kp+9%`Bg!=f2j&f@=`*(Y5cj-@X>aO=KW!NC(wnzNvc6
z7zAcdNs6zl-TSlmWT*xwHP^lbc)cIlz~c!87Ayw*R878xuo_B1O5)zBa2N^1;<1G*
z@DDA&63R9DlDrN~$N9}x1QLdEtC&WKWQ&HIG14u$A+HpWqwX$lU+vz0RW90}bIy+nK
zBwrhPyfpky*m-!mX-|l;`2)jOqm?LH>IIUa9?$ixbzUK2B2c9Rgrbyx*ZlAX#RtgD
zqfV3VrYIREp9?>i#+D0Zt?|MC*^#Cm+VU1w>(E9M%Y(X`oc
z83VP))_SEq?qruzr1Sc>@lK^qT9*ydaP|0QAknq1J#k0@g(Du7e87r*c}3(HHnEP3
zBhk<;vXmLaDKtIs=p2nJ`1kPo94Ab^y>t4lyLrrgn(dbU=y5(LrBFfmTbcsE01XeX
zQJ|Vv4b0G>gOzpU$N8p+n5}?H4M_PAVSE3aqrU&%w|4`bcNdYPPh@d#{e;Cg79d?c
zGT`i=_-mu=4w?5gdU$RdESzif-j4^HD7u!SL4_G^+{5E;0~mk?eU5hkPe8E0jq<|c
z0R-CZ?W$|@rj+q&4;G_M-qzj%24US>`bI%}kG#U&Dd6y%JJGc=I^|i3$uM%_^wa{J
z5DjhLa)jj+BH5z|KhKBKWFx@ib3+F)yxaW}YBGJ1C(t2^M6_K^+ai)67j3#1PWX`R
zisa?`(o_N2a+N#p#sdePI}MSuRnx_1F~*8pev8OZ@;&AbT9Ho^r8L=L)-V|GAxr!>
zuM@SzJd5zK1WwTMM^l|^$&cyTy{#;-dCv8WoxT=vy?K9pNSw|
zbZ-$8KheWPH0#ULv~z4#JhH_%4dixAs)i^^ppjMk0v}UN7pBiCtd!m<&D#~-iDn;D
z+jGUWR0K6AINdsh2v}HG{qTPBuLg6;@9ee+Z-pg&m6k=_gqy9|v}df_3u&EV%TYGs
zyY)f+Tq-^@yxu~4_8AYSe*+<^SHkuB|?b$*X(5C?;t+|8eRCzAitP-3K
zcp^-tD3ac1HDcf_z59;&(=}F$1Qhn_3bjWL((tD{D=C@*r90nKRn&$7zu7oK9V4;P*;{qYN8%44*=3o=h
zte%JWu4|b0Q4B%k1c{Q+18OqtoofazPGT%R6ex%4TMCFMSzh1Y1xNOnzhMUBZ1f=I
zLACM`sPuIYibW5euy@PQUU&*41Z?@lap>IiQltW7*anJicay6U5CaG;(g$+reh%)K%5jpEw+S?9PEi+`?i4j!X^
zFtI&VctgrJv?c-~UB}T2_NdbxO>vr@bM6jf$MYR`kXtjJ>lnQ;`1^qqwcS88R(EKSUJmE@0W&U{{7>A`T?E8
zhMj0>d#XxFDJABp<8}Mvoh73qrnXA7s5ZbjQ7$~?c^*Z@q=zX1ltEz^Ozq2&*nTgb
z`vGh*Cv%SMT@gSD#7W8AHAmGPM!$O^c71Qn1V_~b+Q9K!)tQ897pO6EgzN7wYEF3?
z|63gRHEDYE`^7^#rEaBr|Ile
zTUl+#-SF!x=UK0tsHTft>-2YlHfd>m(=L>!$7NXpS-Ff-4j`CgLiME!xW)iCFiz!e
z3EkeH;zHtjAw@;aY)f7vg57{Y=}RlymZh{jQ7zo>nUFeL{Z1e0oOfbiS{g>Vo$((X
z({6LwC~FQfh7J)cTtRM*=*39>x;zpY%%pfrOs+gH>hA7W?JR>H>Fp#G$ec?N$T%&6
z*_puZ278sW1VgI}tekm>#%J1lfRGJg5LKD*+KF_iFp!)x44e|#^kDBukra@czBF%)
zH{WPh&vyM!4R5(IP9Zx6LLK_YsQ<44Kq+F3=?I2mgysb}f^H}xl;Z&U=+s9JkE?0W
z=;-vCA`ge&M3}BEJ*`wEOLft^%DLGFCEIH=eA0sNd1kmRX|J)?rSv%|SfZSf6R9X7
zKfh_BYEgAXGh@5BI9#j>K&UXt<)A2{+NoboBtwCU#d=@7LqZepaq_7xZ|s24EPja%
zyL&ZyJQQZ%Hriv&<@Lp2?F!c?o#+GxHqym$7{XH*fRf(C!ANsx6qh&KbW*72s+On4
zF_RlZBD$CO2Rl;p9nMVzdW_=gw88bU#o=sM%hkn?7wexNpacBiwLylganh0H!>s~G
zovNj1EnTdOv2uZ5T=dDBLO!iuL4&n|#yn?Chh>zCyc!ntciGzpHRi43iDyDJdRT)0
zzUk}Oy&QzXE3b}T0u=wg2wCYVY2KnK;0ln
z-R&^;ErhE-BbUzq7G2B43>Y6Nc`F|BHWgR6sTZW&FI_^3B1=5eob%=-IC_wjL}w&z
zKczGRx<|To^EnGRkS2ET$%G*5U8pF`nNHcFd3cy^q$@Uq?@3fFATm?WL6P0!Z|l_L7G;EPqu~BU04s{vLQygDPn#J?WROnq(
z-+R(q7+uy@Kb&W+PWY_ysNpynLI!cU?sv4AldD4c9qgQ!hU`U8Q4O)GE)Ex(6RPh7oz!BriU=
zbP`^1{^%fp$x?dcQpN#LovWWDLKUVxJ5kUtO>#o*J(hEPtZAvSU9xQ!7>454PR~NZT4}QA`zeD3Jx6Y&>>Vpm*^dbHiN4voCs<$cW`ID2+
zS<^04n$+=9e}?JR(V*=y0n7ZyugLzpBjD&Kc7kz@0FP}UrLph(p66V88Ur10@;#vB
zbIHg1hEtq;fJS+ZAy7(KiztfMTg;{}ojF-v{d8kx``_3B*Vn9U8ZYl}uo72~si3ntre|RvUd8
zA`DiD`J+@(^1X*tbQA17Tb{lVa|CZ{YvrIwM&&dfMO8W7vcNJ~^FGG9
zF6EdgKh5D*9)Kj)k}bE5(=O!`J?Hjh^~}7(tSO~W$hSvgAQq*3@6d@<(97xGE*JfL
z_x>+|y$v|)IdwrzdkZeQCrpFG5JrtHOR7Yb-C-d|SDHDe8&%g{?5;=1YbORv9k*@6aG{8^Tc-It7wGh~&*$pI*Tq
z`7*emfXnlg;hU;2hJ!*hIJi@s$UE{qeBIve9pxbYJ0u^y;W1wei*bbBEDQyEI8N_Z
z8{QCq1K1o=(paAsQ|+7(4WrD*Om}NeI7&|61z-w(EAY2?@a7>k3)1>c1wX?Z!g%>L
z=RozJIb3s#GX2?>2z6Fn>vs5gn}1SrKzifl(*roqftlY<%iCjbfDLV&QqS_3+CH?r
zx>i7MUe~w*g`Y!$Q&2;AeG2+!Z#JG}#8g@@E#74a5|6b%bNHUHLlO?wv58WK9R?Q=
z&^_y`18Oqty|%JmWAuEzWwM8Il=Cy!Z5b7Lf`<@NIB*BcQC5062yM@CU3F*TlRPbu99g>9B(D-6%WPIaM>3#s|l5^B4l`wE{y@%#YP1AGaz9n(vw#UfH;mgXdv%03&e2_6s
zokkZ}3NbckUGh>oIeN>6QTgm`G$8#I_Uv1Sh<`~F>+&h!-t%tRy}SlavJM*7H72JR
z4zPFl>=!7U3V9A_>h#EA`6N;hd=is>Y`Welg3La(FfJZ>r1uHc_xMZv0oa-}%;Nly
z%WO4U$;&qmmSPal@V3l%>-uU3^r>0*t`Au~KQ1lElpf1Db{^}aojH9-*ZZ7zuHV)V
z&vL4BD{bI|feI%}aPIy1Gtq)i3NZ1PLzFW
zykm%Zp4kg_e_m2}DbF7T0Sb>^`9}=))%zc!Efm#-fY58uj`F33{$-Nc13s0*GTKKN
z2E>)f!SB33FN4_G7@_D_0oStfzUlJH;M({6wU9DODE>@53B#LkAsMuGk3js2$!$
zF%^pM@-WF+R{vZIM{gU#H1B~4NU@|3yAhIL{-GSb6jB58r@hFH2hK!O?~R~!v?D;$
zL}&grG(JnK=ero{+|SYA|M0uy{(i1memA&yLz{cErhVNKusL9t{8Z^>>Sie15bWIt
zxNT^FqBSk1d01Y~NhTf=u)ht>#<;xn+bw^e)nhwdF=Z6W3O^kAFwbA|g?r~7C(@%y
zF26fp^n3ee7hF*S8%$r7+-YbEBwJ
zJ?4%S)kU-BpN?6Dz2gRL14!&3R974jVK{vOxL1G=1C4IE$MA-M=;crV!S}ashlie}
za~UKvn?hC>JcC|hLg(4?dVnY1hOU*Ch(|W!B-rvdHEJqh&BY_f6blt|h|eQHR^qpF
zC-&*spQm|2a@BKC!Eka~Z=X)cHo^w{B?sBsr*~1rW5;UCrlPk1*Drf~?gE|n6u09g
zKcj?P7?dcq@LqGgG#?#V7o*qBOhTQuzj=j?+vw#U!@E@MKu0eN72FX*_?1SQ=Tp@N
z#HA3Mnbvq+kuK_T(-iinlW!b5t*fI`7brFizP#Fu0Ya&Vr`~Hc8482+=se&zJ75Stpz2;Y>{)_
zLG2~8hjP*Bc?gI>x_IScFVRK!!h<(3WAWOm{(={IP2HcOC%81Q_unrChORFeMRIR{
zBusr21>rjZz45NXNdCOHMCXjs0QsGz12(6~W0V9(BGsbVI
zYF_eM^IbRkk+GKkAc}=fxIq*VhC~TOIK_%S49_(MeCOVfQJP|&(RrPB4qw}%jjZPI
zA_n%VWBD0dpx1Aw2W=AF6DAbR2mnBK;TjP$^&{|cw*(R-YxJ$C55_BHf29MHExzif
z!B*P`)pP-kBKC>gqkCKLu%&Qoyz&MDPvGT`P2CK-CNE%9G`1TD!Mly*-0joWzk$k`
z%bOCcfea}{zpseO`)BNFuX%UDtXY_=6XP{`&wpQ@{&y?Apq(~n!-x2TXol0uf3NwZ
zgymhhg#e!ge@q{TEj5WkC=SBH!&w%9TdAN`VeE75OPD~BE@_%Dv#tay&UU5}iWxv~
zHLrkiArn6s1c*_R9P@NYJA_yGJ+IF#_!JC^A#jQ@vMJ0EV21>`Bls3ADMO7VQ1*yb
zTNI9r^onML_w@#VqlMFZ;4edQjW!@iuM^`=3-Xc|ie3U8UwX{w{(7FJOl5yUK
z5aqM+zy$`Ma1)-sgyhEd*ol_3mn*)24CX5duRdsXDxRq;hp4$MMKDMmUBo}ze+(q`
z@lBWXQC#!=tr&fh8sKLWCU*8$G>3QQ-66pB5bH0=ib5PYNaF(0*V?=WTA|mDJhhZp
zaoQevX+@p{awzsJ!jmZel`^ht)+0KvehmLrLz5y%!LJpH(rLcFTVs=&qiZ1g;Luw+
z{d1qZ^bn0FmQ?51*pnYk^mtiLsp8M?ST^gb8eT@BmjIRd1euP!_Fl(N@wU5n@=Q2U
zv|&paU@)ZXmd1Lp62=5Rqn=@p_uVq6S^JqiX{SgSBlN%S2^Fj%0QXn|ogjXImLUMQ
zPBCmxd}NXJ5#LnK`;K0H4-?*4G8>);3N4)wn5iS1`)KddB9n}As-}3)8a>{j7l*xj
zMin^xXD-j9;=ze10LN|>Qi{fufbZE>oA^2HgyDsU>;$dwy}KwB!K(Vv?T%(ln5;wd
z)u;1Sz?=HIa4?sh3){DE1p3V>klhWAmJHA)wSdgXww-REIebxKtan&{{?h0Zw=gK@
zUsBSmMnHz?8nJYtLrBpl8e$~oT&UUo94sT$ALOsY3W5A>Yip{^(~OUYZ(*I3Qoj$o
zDTNAyAOJY{6d6J&-}jHQfb6+AqB7+bDQmXM#`nC!U;(8Hdd^`+D>W$*(erplyBUW3
zc(KI;MWX%)rK3`(u}CV4fbyi%RE^RIDvc$u6=wXB0xVrqC@dIP#H3h}BE|`n#g9T+
zIo_`gvY5IIGPZUEP${nR13as>X5C@vqjZ(&(o0=*gK-Y6bHAbu-!l=K0T_KD71N_2
zwJAlpz5%|Iwnu5gYd0SqCGZt}CB(-Yv0Pv@$CCg|MvcL!ufn#hD^5U!0f=RnF%pUU
zdvCGtJv(H|cDtoHFV%>=%4Jd|de9Do+g|rIYk7F!+q(dq)>3$1Ufa)_Zz)3zzi8e3
z*Q>u$A=2G2$`#fXUbnsBiLYy8nh!dAdJa`nq{oK0!tz0iB2mg!Mhh+~JhbdkY*LdG
z9f@o`I1W-kUi63zC!VE0UvLNVPF>Rg9b|BTr6Mtfj|~q>roC5ObJZGgyfTs5rXe13
z54OkY2UO*B?Jp+IqG!@;-d+Cuhp
z5$}pQIZzl&v?Bj&d9|3Iomc9hR%`F~A$E8!|FsrimY=-3+nV=Z_Ia
zL5fpRoaiq41k4sfYNW%=gb?=jsZl<|MiErbdN!%CGD*g;CNbEv1!MI|^(70Sw-J
znl-s!13;p6i^!BxK-sidHEX*rzNgs>%jAP=e&o$V6TOIzg;C|uXWnB_R`4msWa%P9
zUD&jHo_W6HpIVmkdx6q1I+ZnCihzoTbXWYIu!(mLk|-BczgPFpkS^q-pGF~EswYw6
z>XN|H7FvpGmt_TQa%kydoN($ZFr0@5s*z&+Syu>O4xIR+0qnAJI@yI?-etrNrZaM)
zmGWW%?Hv(F$}al(yMe|dUDyF18?Q_AhV5i2I!D^^Ojo>DGTO$(lC4Mj0SeM}G*rY(
zTI-GE)Fnp?U=dwh{^?Z%CAWDwhWAwAD$ekOcALSrF4(qkR!=3ecIhJ_Hp*qIgzxRkVR
z3zE9<#N?vGpyqcmkFh0Rc^ia8x%Yz?o^do9wF18?qDOH^m0(!#;$sGo*L4B-LW4rI@flBz2$W&