From fe9cfe7cb4e121572160f7e5319886bca91f816f Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Fri, 22 Aug 2025 11:50:53 -0700 Subject: [PATCH 01/31] Add comprehensive image normalization documentation - Add detailed 'Image Normalization' section to Image.md reference explaining the normalization algorithm - Document when normalization is applied (PyTorch tensors, NumPy arrays) vs when it's not (PIL Images, file paths) - Provide specific examples showing [0,1], [-1,1], and other range normalization effects - Add examples demonstrating how to avoid normalization using PIL Images or normalize=False - Include best practices and common pitfalls - Update media logging guide to reference the detailed normalization documentation - Add comprehensive translations in Japanese and Korean Addresses DOCS-1016 - resolves user confusion about PyTorch tensor normalization behavior --- content/en/guides/models/track/log/media.md | 2 +- content/en/ref/python/sdk/data-types/Image.md | 97 ++++++++++++++++++ content/ja/guides/models/track/log/media.md | 2 +- content/ja/ref/python/data-types/image.md | 98 +++++++++++++++++++ content/ko/guides/models/track/log/media.md | 2 +- content/ko/ref/python/data-types/image.md | 98 +++++++++++++++++++ 6 files changed, 296 insertions(+), 3 deletions(-) diff --git a/content/en/guides/models/track/log/media.md b/content/en/guides/models/track/log/media.md index 7d118e2db6..592ab0d003 100644 --- a/content/en/guides/models/track/log/media.md +++ b/content/en/guides/models/track/log/media.md @@ -58,7 +58,7 @@ with wandb.init(project="image-log-example") as run: run.log({"examples": images}) ``` -We assume the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, we convert them to integers between `0` and `255`. If you want to normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or just supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab of this panel. +We assume the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, we automatically normalize them to integers between `0` and `255` using our normalization algorithm. For detailed information about how normalization works with PyTorch tensors and NumPy arrays, see the [Image Normalization section]({{< relref "/ref/python/sdk/data-types/image.md#image-normalization" >}}) in the Image reference. If you want to normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or just supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab of this panel. {{% /tab %}} {{% tab header="Logging PIL Images" %}} For full control over the conversion of arrays to images, construct the [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html) yourself and provide it directly. diff --git a/content/en/ref/python/sdk/data-types/Image.md b/content/en/ref/python/sdk/data-types/Image.md index 384811b98f..3f798d1261 100644 --- a/content/en/ref/python/sdk/data-types/Image.md +++ b/content/en/ref/python/sdk/data-types/Image.md @@ -103,6 +103,103 @@ with wandb.init() as run: run.log({"examples": examples}) ``` +## Image Normalization + +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle common image data formats and ensure proper display. + +### When Normalization is Applied + +Normalization is applied to: +- **PyTorch tensors** (format: `(channel, height, width)`) +- **NumPy arrays** (format: `(height, width, channel)`) + +Normalization is **NOT** applied to: +- **PIL Images** (passed as-is) +- **File paths** (loaded as-is) + +### Normalization Algorithm + +The normalization algorithm automatically detects the input range and applies the appropriate transformation: + +1. **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 + ```python + normalized_data = (data * 255).astype(np.uint8) + ``` + +2. **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: + ```python + normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) + ``` + +3. **For any other range**: Values are clipped to [0, 255] and converted to uint8 + ```python + normalized_data = data.clip(0, 255).astype(np.uint8) + ``` + +### Examples of Normalization Effects + +**Example 1: [0, 1] range data** +```python +import torch +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) # Random values between 0 and 1 + +# This will multiply all values by 255 +image = wandb.Image(tensor_0_1, caption="Normalized from [0,1] range") +``` + +**Example 2: [-1, 1] range data** +```python +import torch +import wandb + +# Create tensor with values in [-1, 1] range +tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # Random values between -1 and 1 + +# This will rescale: -1 → 0, 0 → 127.5, 1 → 255 +image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") +``` + +**Example 3: Avoiding normalization with PIL Images** +```python +import torch +from PIL import Image as PILImage +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) + +# Convert to PIL Image to avoid normalization +pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) +image = wandb.Image(pil_image, caption="No normalization applied") +``` + +**Example 4: Using normalize=False** +```python +import torch +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) + +# Disable normalization - values will be clipped to [0, 255] +image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") +``` + +### Best Practices + +1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging +2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` +3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) +4. **For precise control**: Use PIL Images when you need exact pixel values + +### Common Pitfalls + +- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter +- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information +- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results --- diff --git a/content/ja/guides/models/track/log/media.md b/content/ja/guides/models/track/log/media.md index f3d7206d90..5f113e1ae9 100644 --- a/content/ja/guides/models/track/log/media.md +++ b/content/ja/guides/models/track/log/media.md @@ -52,7 +52,7 @@ images = wandb.Image(image_array, caption="Top: Output, Bottom: Input") wandb.log({"examples": images}) ``` -最後の次元が1の場合はグレースケール、3の場合はRGB、4の場合はRGBAと仮定します。配列が浮動小数点数を含む場合、それらを`0`から`255`の整数に変換します。異なる方法で画像を正規化したい場合は、[`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)を手動で指定するか、`"Logging PIL Images"`タブで説明されているように、単に[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を提供することができます。 +最後の次元が1の場合はグレースケール、3の場合はRGB、4の場合はRGBAと仮定します。配列が浮動小数点数を含む場合、正規化アルゴリズムを使用して自動的に`0`から`255`の整数に変換します。PyTorchテンソルとNumPy配列での正規化の動作についての詳細は、[Imageリファレンスの画像正規化セクション]({{< relref path="/ref/python/sdk/data-types/image.md#image-normalization" lang="ja" >}})を参照してください。異なる方法で画像を正規化したい場合は、[`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)を手動で指定するか、`"Logging PIL Images"`タブで説明されているように、単に[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を提供することができます。 {{% /tab %}} {{% tab header="PIL Imagesをログする" %}} 配列から画像への変換を完全に制御するために、[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を自分で構築し、直接提供してください。 diff --git a/content/ja/ref/python/data-types/image.md b/content/ja/ref/python/data-types/image.md index a38c29dfdf..36b96ed0aa 100644 --- a/content/ja/ref/python/data-types/image.md +++ b/content/ja/ref/python/data-types/image.md @@ -30,6 +30,104 @@ Image( 注意 : `wandb.Image` として `torch.Tensor` をログする際、画像は正規化されます。画像を正規化したくない場合は、テンソルを PIL Image に変換してください。 +## 画像の正規化 + +PyTorch テンソルや NumPy 配列を `wandb.Image` に渡すと、`normalize=False` を設定しない限り、ピクセル値は自動的に [0, 255] の範囲に正規化されます。この正規化は、一般的な画像データ形式を処理し、適切な表示を確保するために設計されています。 + +### 正規化が適用される場合 + +正規化は以下に適用されます： +- **PyTorch テンソル** (形式: `(channel, height, width)`) +- **NumPy 配列** (形式: `(height, width, channel)`) + +正規化は以下には適用されません： +- **PIL 画像** (そのまま渡される) +- **ファイルパス** (そのまま読み込まれる) + +### 正規化アルゴリズム + +正規化アルゴリズムは入力範囲を自動的に検出し、適切な変換を適用します： + +1. **データが [0, 1] の範囲にある場合**: 値に255を掛けてuint8に変換 + ```python + normalized_data = (data * 255).astype(np.uint8) + ``` + +2. **データが [-1, 1] の範囲にある場合**: 以下の式で [0, 255] にリスケール + ```python + normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) + ``` + +3. **その他の範囲の場合**: 値を [0, 255] にクリップしてuint8に変換 + ```python + normalized_data = data.clip(0, 255).astype(np.uint8) + ``` + +### 正規化効果の例 + +**例1: [0, 1] 範囲のデータ** +```python +import torch +import wandb + +# [0, 1] 範囲の値を持つテンソルを作成 +tensor_0_1 = torch.rand(3, 64, 64) # 0から1の間のランダム値 + +# すべての値に255を掛ける +image = wandb.Image(tensor_0_1, caption="[0,1]範囲から正規化") +``` + +**例2: [-1, 1] 範囲のデータ** +```python +import torch +import wandb + +# [-1, 1] 範囲の値を持つテンソルを作成 +tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # -1から1の間のランダム値 + +# リスケール: -1 → 0, 0 → 127.5, 1 → 255 +image = wandb.Image(tensor_neg1_1, caption="[-1,1]範囲から正規化") +``` + +**例3: PIL画像で正規化を回避** +```python +import torch +from PIL import Image as PILImage +import wandb + +# [0, 1] 範囲の値を持つテンソルを作成 +tensor_0_1 = torch.rand(3, 64, 64) + +# PIL画像に変換して正規化を回避 +pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) +image = wandb.Image(pil_image, caption="正規化は適用されません") +``` + +**例4: normalize=False を使用** +```python +import torch +import wandb + +# [0, 1] 範囲の値を持つテンソルを作成 +tensor_0_1 = torch.rand(3, 64, 64) + +# 正規化を無効化 - 値は [0, 255] にクリップされる +image = wandb.Image(tensor_0_1, normalize=False, caption="正規化無効") +``` + +### ベストプラクティス + +1. **一貫した結果のため**: ログする前にデータを期待される [0, 255] 範囲に前処理する +2. **正規化を回避するため**: `PILImage.fromarray()` を使用してテンソルをPIL画像に変換する +3. **デバッグのため**: `normalize=False` を使用して生の値を確認する（[0, 255] にクリップされる） +4. **精密な制御のため**: 正確なピクセル値が必要な場合はPIL画像を使用する + +### よくある落とし穴 + +- **予期しない明度**: テンソル値が [0, 1] 範囲にある場合、255倍されるため画像が非常に明るくなる +- **データ損失**: [0, 255] 範囲外の値はクリップされ、情報が失われる可能性がある +- **一貫性のない動作**: 異なる入力タイプ（テンソル vs PIL vs ファイルパス）で異なる結果が得られる可能性がある + #### 例: ### numpy 配列から wandb.Image を作成 diff --git a/content/ko/guides/models/track/log/media.md b/content/ko/guides/models/track/log/media.md index 8758e4501a..ef4668b075 100644 --- a/content/ko/guides/models/track/log/media.md +++ b/content/ko/guides/models/track/log/media.md @@ -53,7 +53,7 @@ images = wandb.Image(image_array, caption="Top: Output, Bottom: Input") wandb.log({"examples": images}) ``` -마지막 차원이 1이면 이미지가 회색조, 3이면 RGB, 4이면 RGBA라고 가정합니다. 배열에 float가 포함된 경우 `0`과 `255` 사이의 정수로 변환합니다. 이미지를 다르게 정규화하려면 [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)를 수동으로 지정하거나 이 패널의 "PIL 이미지 로깅" 탭에 설명된 대로 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 제공하면 됩니다. +마지막 차원이 1이면 이미지가 회색조, 3이면 RGB, 4이면 RGBA라고 가정합니다. 배열에 float가 포함된 경우 정규화 알고리즘을 사용하여 자동으로 `0`과 `255` 사이의 정수로 변환합니다. PyTorch 텐서와 NumPy 배열에서 정규화가 어떻게 작동하는지에 대한 자세한 정보는 [Image 참조의 이미지 정규화 섹션]({{< relref path="/ref/python/sdk/data-types/image.md#image-normalization" lang="ko" >}})을 참조하세요. 이미지를 다르게 정규화하려면 [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)를 수동으로 지정하거나 이 패널의 "PIL 이미지 로깅" 탭에 설명된 대로 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 제공하면 됩니다. {{% /tab %}} {{% tab header="PIL 이미지 로깅" %}} 배열을 이미지로 변환하는 것을 완벽하게 제어하려면 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 직접 구성하여 제공합니다. diff --git a/content/ko/ref/python/data-types/image.md b/content/ko/ref/python/data-types/image.md index 5d2b7e576e..add78b2eef 100644 --- a/content/ko/ref/python/data-types/image.md +++ b/content/ko/ref/python/data-types/image.md @@ -30,6 +30,104 @@ Image( 참고 : `torch.Tensor`를 `wandb.Image`로 로깅할 때 이미지는 정규화됩니다. 이미지를 정규화하지 않으려면 텐서를 PIL Image로 변환하십시오. +## 이미지 정규화 + +PyTorch 텐서나 NumPy 배열을 `wandb.Image`에 전달하면 `normalize=False`를 설정하지 않는 한 픽셀 값이 자동으로 [0, 255] 범위로 정규화됩니다. 이 정규화는 일반적인 이미지 데이터 형식을 처리하고 적절한 표시를 보장하기 위해 설계되었습니다. + +### 정규화가 적용되는 경우 + +정규화는 다음에 적용됩니다: +- **PyTorch 텐서** (형식: `(channel, height, width)`) +- **NumPy 배열** (형식: `(height, width, channel)`) + +정규화는 다음에 적용되지 않습니다: +- **PIL 이미지** (그대로 전달됨) +- **파일 경로** (그대로 로드됨) + +### 정규화 알고리즘 + +정규화 알고리즘은 입력 범위를 자동으로 감지하고 적절한 변환을 적용합니다: + +1. **데이터가 [0, 1] 범위에 있는 경우**: 값에 255를 곱하고 uint8로 변환 + ```python + normalized_data = (data * 255).astype(np.uint8) + ``` + +2. **데이터가 [-1, 1] 범위에 있는 경우**: 다음 공식으로 [0, 255]에 리스케일 + ```python + normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) + ``` + +3. **기타 범위의 경우**: 값을 [0, 255]로 클립하고 uint8로 변환 + ```python + normalized_data = data.clip(0, 255).astype(np.uint8) + ``` + +### 정규화 효과 예시 + +**예시 1: [0, 1] 범위 데이터** +```python +import torch +import wandb + +# [0, 1] 범위의 값을 가진 텐서 생성 +tensor_0_1 = torch.rand(3, 64, 64) # 0에서 1 사이의 랜덤 값 + +# 모든 값에 255를 곱함 +image = wandb.Image(tensor_0_1, caption="[0,1] 범위에서 정규화") +``` + +**예시 2: [-1, 1] 범위 데이터** +```python +import torch +import wandb + +# [-1, 1] 범위의 값을 가진 텐서 생성 +tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # -1에서 1 사이의 랜덤 값 + +# 리스케일: -1 → 0, 0 → 127.5, 1 → 255 +image = wandb.Image(tensor_neg1_1, caption="[-1,1] 범위에서 정규화") +``` + +**예시 3: PIL 이미지로 정규화 회피** +```python +import torch +from PIL import Image as PILImage +import wandb + +# [0, 1] 범위의 값을 가진 텐서 생성 +tensor_0_1 = torch.rand(3, 64, 64) + +# PIL 이미지로 변환하여 정규화 회피 +pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) +image = wandb.Image(pil_image, caption="정규화가 적용되지 않음") +``` + +**예시 4: normalize=False 사용** +```python +import torch +import wandb + +# [0, 1] 범위의 값을 가진 텐서 생성 +tensor_0_1 = torch.rand(3, 64, 64) + +# 정규화 비활성화 - 값은 [0, 255]로 클립됨 +image = wandb.Image(tensor_0_1, normalize=False, caption="정규화 비활성화") +``` + +### 모범 사례 + +1. **일관된 결과를 위해**: 로깅하기 전에 데이터를 예상되는 [0, 255] 범위로 전처리 +2. **정규화를 회피하기 위해**: `PILImage.fromarray()`를 사용하여 텐서를 PIL 이미지로 변환 +3. **디버깅을 위해**: `normalize=False`를 사용하여 원시 값 확인 ([0, 255]로 클립됨) +4. **정밀한 제어를 위해**: 정확한 픽셀 값이 필요한 경우 PIL 이미지 사용 + +### 일반적인 함정 + +- **예상치 못한 밝기**: 텐서 값이 [0, 1] 범위에 있으면 255배되어 이미지가 매우 밝아짐 +- **데이터 손실**: [0, 255] 범위를 벗어나는 값은 클립되어 정보가 손실될 수 있음 +- **일관성 없는 동작**: 다른 입력 유형(텐서 vs PIL vs 파일 경로)에서 다른 결과가 나올 수 있음 + #### 예시: ### numpy array에서 wandb.Image 생성 From 0f05c593651e702b4f8fa6c9afe6a24d6d83f06f Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Fri, 22 Aug 2025 12:09:15 -0700 Subject: [PATCH 02/31] Fix broken relref links in Japanese and Korean media guides --- content/ja/guides/models/track/log/media.md | 6 +++--- content/ko/guides/models/track/log/media.md | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/content/ja/guides/models/track/log/media.md b/content/ja/guides/models/track/log/media.md index 5f113e1ae9..21938b0e43 100644 --- a/content/ja/guides/models/track/log/media.md +++ b/content/ja/guides/models/track/log/media.md @@ -52,7 +52,7 @@ images = wandb.Image(image_array, caption="Top: Output, Bottom: Input") wandb.log({"examples": images}) ``` -最後の次元が1の場合はグレースケール、3の場合はRGB、4の場合はRGBAと仮定します。配列が浮動小数点数を含む場合、正規化アルゴリズムを使用して自動的に`0`から`255`の整数に変換します。PyTorchテンソルとNumPy配列での正規化の動作についての詳細は、[Imageリファレンスの画像正規化セクション]({{< relref path="/ref/python/sdk/data-types/image.md#image-normalization" lang="ja" >}})を参照してください。異なる方法で画像を正規化したい場合は、[`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)を手動で指定するか、`"Logging PIL Images"`タブで説明されているように、単に[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を提供することができます。 +最後の次元が1の場合はグレースケール、3の場合はRGB、4の場合はRGBAと仮定します。配列が浮動小数点数を含む場合、正規化アルゴリズムを使用して自動的に`0`から`255`の整数に変換します。PyTorchテンソルとNumPy配列での正規化の動作についての詳細は、[Imageリファレンスの画像正規化セクション]({{< relref path="/ref/python/data-types/image.md#image-normalization" lang="ja" >}})を参照してください。異なる方法で画像を正規화したい場合は、[`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)を手動で指定するか、`"Logging PIL Images"`タブで説明されているように、単に[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を提供することができます。 {{% /tab %}} {{% tab header="PIL Imagesをログする" %}} 配列から画像への変換を完全に制御するために、[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を自分で構築し、直接提供してください。 @@ -564,7 +564,7 @@ runが終了すると、UIで分子の3D可視化と対話できるようにな ### PNG 画像 -[`wandb.Image`]({{< relref path="/ref/python/data-types/image.md" lang="ja" >}})は`numpy`配列や`PILImage`のインスタンスをデフォルトでPNGに変換します。 +`wandb.Image`は`numpy`配列や`PILImage`のインスタンスをデフォルトでPNGに変換します。 ```python wandb.log({"example": wandb.Image(...)}) @@ -584,7 +584,7 @@ wandb.log({"example": wandb.Video("myvideo.mp4")}) ## 分子の2Dビュー -[`wandb.Image`]({{< relref path="/ref/python/data-types/image.md" lang="ja" >}})データ型と[`rdkit`](https://www.rdkit.org/docs/index.html)を使用して分子の2Dビューをログできます: +[`wandb.Image`]({{< relref "/ref/python/data-types/image.md" >}})データ型と[`rdkit`](https://www.rdkit.org/docs/index.html)を使用して分子の2Dビューをログできます: ```python molecule = rdkit.Chem.MolFromSmiles("CC(=O)O") diff --git a/content/ko/guides/models/track/log/media.md b/content/ko/guides/models/track/log/media.md index ef4668b075..e37811c81f 100644 --- a/content/ko/guides/models/track/log/media.md +++ b/content/ko/guides/models/track/log/media.md @@ -53,7 +53,7 @@ images = wandb.Image(image_array, caption="Top: Output, Bottom: Input") wandb.log({"examples": images}) ``` -마지막 차원이 1이면 이미지가 회색조, 3이면 RGB, 4이면 RGBA라고 가정합니다. 배열에 float가 포함된 경우 정규화 알고리즘을 사용하여 자동으로 `0`과 `255` 사이의 정수로 변환합니다. PyTorch 텐서와 NumPy 배열에서 정규화가 어떻게 작동하는지에 대한 자세한 정보는 [Image 참조의 이미지 정규화 섹션]({{< relref path="/ref/python/sdk/data-types/image.md#image-normalization" lang="ko" >}})을 참조하세요. 이미지를 다르게 정규화하려면 [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)를 수동으로 지정하거나 이 패널의 "PIL 이미지 로깅" 탭에 설명된 대로 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 제공하면 됩니다. +마지막 차원이 1이면 이미지가 회색조, 3이면 RGB, 4이면 RGBA라고 가정합니다. 배열에 float가 포함된 경우 정규화 알고리즘을 사용하여 자동으로 `0`과 `255` 사이의 정수로 변환합니다. PyTorch 텐서와 NumPy 배열에서 정규화가 어떻게 작동하는지에 대한 자세한 정보는 Image 참조 문서를 참조하세요. 이미지를 다르게 정규화하려면 [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)를 수동으로 지정하거나 이 패널의 "PIL 이미지 로깅" 탭에 설명된 대로 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 제공하면 됩니다. {{% /tab %}} {{% tab header="PIL 이미지 로깅" %}} 배열을 이미지로 변환하는 것을 완벽하게 제어하려면 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 직접 구성하여 제공합니다. @@ -557,7 +557,7 @@ run이 완료되면 UI에서 분자의 3D 시각화와 상호 작용할 수 있 ### PNG 이미지 -[`wandb.Image`]({{< relref path="/ref/python/data-types/image.md" lang="ko" >}})는 `numpy` 배열 또는 `PILImage` 인스턴스를 기본적으로 PNG로 변환합니다. +`wandb.Image`는 `numpy` 배열 또는 `PILImage` 인스턴스를 기본적으로 PNG로 변환합니다. ```python wandb.log({"example": wandb.Image(...)}) @@ -577,7 +577,7 @@ wandb.log({"example": wandb.Video("myvideo.mp4")}) ## 분자의 2D 보기 -[`wandb.Image`]({{< relref path="/ref/python/data-types/image.md" lang="ko" >}}) 데이터 유형과 [`rdkit`](https://www.rdkit.org/docs/index.html)을 사용하여 분자의 2D 보기를 기록할 수 있습니다. +[`wandb.Image`]({{< relref "/ref/python/data-types/image.md" >}}) 데이터 유형과 [`rdkit`](https://www.rdkit.org/docs/index.html)을 사용하여 분자의 2D 보기를 기록할 수 있습니다. ```python molecule = rdkit.Chem.MolFromSmiles("CC(=O)O") From f98861061a734e789e91c2dd979df937f8136d06 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Fri, 22 Aug 2025 12:19:05 -0700 Subject: [PATCH 03/31] docs: address review feedback - use sentence case, avoid 'we', improve headings --- content/en/guides/models/track/log/media.md | 2 +- content/en/ref/python/sdk/data-types/Image.md | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/content/en/guides/models/track/log/media.md b/content/en/guides/models/track/log/media.md index 592ab0d003..775affae2e 100644 --- a/content/en/guides/models/track/log/media.md +++ b/content/en/guides/models/track/log/media.md @@ -58,7 +58,7 @@ with wandb.init(project="image-log-example") as run: run.log({"examples": images}) ``` -We assume the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, we automatically normalize them to integers between `0` and `255` using our normalization algorithm. For detailed information about how normalization works with PyTorch tensors and NumPy arrays, see the [Image Normalization section]({{< relref "/ref/python/sdk/data-types/image.md#image-normalization" >}}) in the Image reference. If you want to normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or just supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab of this panel. +The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255` using the normalization algorithm. For detailed information about how normalization works with PyTorch tensors and NumPy arrays, see the [Image Normalization section]({{< relref "/ref/python/sdk/data-types/image.md#image-normalization" >}}) in the Image reference. If you want to normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or just supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab of this panel. {{% /tab %}} {{% tab header="Logging PIL Images" %}} For full control over the conversion of arrays to images, construct the [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html) yourself and provide it directly. diff --git a/content/en/ref/python/sdk/data-types/Image.md b/content/en/ref/python/sdk/data-types/Image.md index 3f798d1261..2dfd9b7a90 100644 --- a/content/en/ref/python/sdk/data-types/Image.md +++ b/content/en/ref/python/sdk/data-types/Image.md @@ -103,9 +103,9 @@ with wandb.init() as run: run.log({"examples": examples}) ``` -## Image Normalization +## Image normalization -When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle common image data formats and ensure proper display. +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle and ensure proper display of common image formats. ### When Normalization is Applied @@ -188,14 +188,14 @@ tensor_0_1 = torch.rand(3, 64, 64) image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") ``` -### Best Practices +### Recommendations 1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging 2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` 3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) 4. **For precise control**: Use PIL Images when you need exact pixel values -### Common Pitfalls +### Troubleshooting - **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter - **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information From 5e1e10c1c7a22d1786e024a46a41f94cddac09eb Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Fri, 22 Aug 2025 12:26:07 -0700 Subject: [PATCH 04/31] docs: address additional review feedback - use sentence case for headings, improve text clarity --- content/en/guides/models/track/log/media.md | 2 +- content/en/ref/python/sdk/data-types/Image.md | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/content/en/guides/models/track/log/media.md b/content/en/guides/models/track/log/media.md index 775affae2e..ed80f6f414 100644 --- a/content/en/guides/models/track/log/media.md +++ b/content/en/guides/models/track/log/media.md @@ -58,7 +58,7 @@ with wandb.init(project="image-log-example") as run: run.log({"examples": images}) ``` -The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255` using the normalization algorithm. For detailed information about how normalization works with PyTorch tensors and NumPy arrays, see the [Image Normalization section]({{< relref "/ref/python/sdk/data-types/image.md#image-normalization" >}}) in the Image reference. If you want to normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or just supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab of this panel. +The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image normalization section]({{< relref "/ref/python/sdk/data-types/image.md#image-normalization" >}}) in the Image SDK reference. To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. {{% /tab %}} {{% tab header="Logging PIL Images" %}} For full control over the conversion of arrays to images, construct the [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html) yourself and provide it directly. diff --git a/content/en/ref/python/sdk/data-types/Image.md b/content/en/ref/python/sdk/data-types/Image.md index 2dfd9b7a90..a385fe2aa0 100644 --- a/content/en/ref/python/sdk/data-types/Image.md +++ b/content/en/ref/python/sdk/data-types/Image.md @@ -107,7 +107,7 @@ with wandb.init() as run: When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle and ensure proper display of common image formats. -### When Normalization is Applied +### When normalization is applied Normalization is applied to: - **PyTorch tensors** (format: `(channel, height, width)`) @@ -117,7 +117,7 @@ Normalization is **NOT** applied to: - **PIL Images** (passed as-is) - **File paths** (loaded as-is) -### Normalization Algorithm +### Normalization algorithm The normalization algorithm automatically detects the input range and applies the appropriate transformation: @@ -136,7 +136,7 @@ The normalization algorithm automatically detects the input range and applies th normalized_data = data.clip(0, 255).astype(np.uint8) ``` -### Examples of Normalization Effects +### Examples of normalization effects **Example 1: [0, 1] range data** ```python From 437e94783d9757862f2e05c1ae6ca6338dd4e218 Mon Sep 17 00:00:00 2001 From: Noah Luna <15202580+ngrayluna@users.noreply.github.com> Date: Fri, 22 Aug 2025 13:50:04 -0700 Subject: [PATCH 05/31] Artifacts: When to use save() vs .log() (#1557) Description ----------- Adds quick blurb on "best practice"/recommendation by Eng on when to use which. Related issues ----------- - Fixes https://wandb.atlassian.net/jira/software/projects/DOCS/boards/28?selectedIssue=DOCS-1087 --- content/en/guides/core/artifacts/construct-an-artifact.md | 5 ++++- content/en/guides/core/artifacts/update-an-artifact.md | 5 +++++ 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/content/en/guides/core/artifacts/construct-an-artifact.md b/content/en/guides/core/artifacts/construct-an-artifact.md index f1628186e6..17021922bf 100644 --- a/content/en/guides/core/artifacts/construct-an-artifact.md +++ b/content/en/guides/core/artifacts/construct-an-artifact.md @@ -69,7 +69,10 @@ run = wandb.init(project="artifacts-example", job_type="job-type") run.log_artifact(artifact) ``` -You can optionally construct an artifact outside of a W&B run. For more information, see [Track external files]({{< relref "./track-external-files.md" >}}). +{{% alert title="When to use Artifact.save() or wandb.Run.log_artifact()" %}} +- Use `Artifact.save()` to update an existing artifact without creating a new run. +- Use `wandb.Run.log_artifact()` to create a new artifact and associate it with a specific run. +{{% /alert %}} {{% alert color="secondary" %}} Calls to `log_artifact` are performed asynchronously for performant uploads. This can cause surprising behavior when logging artifacts in a loop. For example: diff --git a/content/en/guides/core/artifacts/update-an-artifact.md b/content/en/guides/core/artifacts/update-an-artifact.md index 4434d737d5..eaba7829ae 100644 --- a/content/en/guides/core/artifacts/update-an-artifact.md +++ b/content/en/guides/core/artifacts/update-an-artifact.md @@ -10,6 +10,11 @@ weight: 4 Pass desired values to update the `description`, `metadata`, and `alias` of an artifact. Call the `save()` method to update the artifact on the W&B servers. You can update an artifact during a W&B Run or outside of a Run. +{{% alert title="When to use Artifact.save() or wandb.Run.log_artifact()" %}} +- Use `Artifact.save()` to update an existing artifact without creating a new run. +- Use `wandb.Run.log_artifact()` to create a new artifact and associate it with a specific run. +{{% /alert %}} + Use the W&B Public API ([`wandb.Api`]({{< relref "/ref/python/public-api/api.md" >}})) to update an artifact outside of a run. Use the Artifact API ([`wandb.Artifact`]({{< relref "/ref/python/sdk/classes/artifact.md" >}})) to update an artifact during a run. {{% alert color="secondary" %}} From a9f62908f63cbb90c20283c56f7085f4f1d4a592 Mon Sep 17 00:00:00 2001 From: Karan Nisar Date: Fri, 22 Aug 2025 18:14:28 -0400 Subject: [PATCH 06/31] Updated new view for lineage and dataset mismatch in keras tutorials (#1541) --- assets/images/app_ui/lineage_panel.png | Bin 120686 -> 504323 bytes .../tutorials/integration-tutorials/keras.md | 4 ++-- .../integration-tutorials/keras_models.md | 2 +- .../integration-tutorials/keras_tables.md | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) diff --git a/assets/images/app_ui/lineage_panel.png b/assets/images/app_ui/lineage_panel.png index 502f35cd8af2ad567b8236b5b79a8be1d8a8a1a8..a8c7e77cbc8e12f692a17b3599d626f9dcf51203 100644 GIT binary patch literal 504323 zcmeFZcQjmW|1T`DEm|VEp_jIK(Q_Abk4cMP?Y$4!HI(6OW@J7yfHdfrkBWyL7Ow+@?(7FB)h5;EKg*Sclu;bmR7B=$yhpodG zDH$?^%draA&p)}Y6w@!%O4^Y*!ecy^Q*6TGPsmspFwVaoOv>kr8%-1>iN*t&GEA!k z;E#b1CQg>^>4u*xOwwO`7W96(X1}}9+qQTj$d}4jIS4`vyllJdXd!4^L4&sWG9@rV zjy;jh*|B?|uwAW-;inKHhr`sib-bszP$$dV9JX`w8Jh@bxz3s-Ey4WuPGyp*MU?kdZKq}~!XfR! zt!s|5YOz>d)R}0qOyVB2XpygYj%VbLoPj5bvOHX)PjkW?i#BNrLKQM~Ec?-1j>lZS zSGWkxH`QSg!2>h|bNT_Yg5)Z}vto4K-L=SPDSziG0>hSrH}f8-%;|rkASDq2y>_i2 zG$9R}-;p6z?0?m7$6J`d%jb0^f5cC?VXrK|;$S{RFk?}+*eh$I;oS9JSKS-`5>028KOjt71!sM`j@|w?@m01k3P(87-5GEHe*4maM?KRXOsVIn zN!6Q@l{merQ{*$3KQNzzcPJtZ%^#n5svom*CLTReU!IBx-ns19nlT;nPWJ>JpV9U_ zayYA}0biiR*4vrEiFhdAIB}flIatki}biSSLZ(cw})lSz= z_m4$(N#$>Ue@W$%%0E`&-M@VYs=F-rSm;h)@Y3`A-<{yk{}%@!{N=Q|{pE<@jNk;O zA^h%j_;vVSetY_Z`AhR}PG4|oh3W4v^S?*=zqyS6H4WGQKYfD#^Sx}RqyF-Ld&SHD zdU$dtJ-l2+IW zIe@DE-I^C+6V>)jitRra|8=;aBCbp$Tlt?S2!AYjmg>*?s`!{|y~PHFOKb6*^b8E= zs8ZtdRN|&~SK-Pl&;R4y|Mc{wrgu0f{&X!=y81Gq!<107Qfj84@UJ6M5FaM3{4CVz zqmW4sB>Z-j`;7DvaT9gsmq^vD%#Q!o#J!8OEboyp$|PK4#r7}nkQkn;_es)kgGb7D z&GQE#Bjmqp6;Q1{AMi3~J_^$J-=5+wH9;kpThvYY{rZukNadOOgGMEI8Zud>JXU7L z*%W|HMurio*MV93x8pioUtd%jiqE)jg|@sI)~q@UJ2{ss0ekuGwv6>3v_a}GzGrb_ zVGndx`cwH!$IFrwqv8(b18R*P{p)24zP5ppN5RYh1Th#x*$6=iJ4uUOaTqKBQ7%%Y z(8=z=S!G^S#T(x?pw{JQUH9reIz?yydR&JiQg3W3q!sAPqM4s>)8CoaOE{VdUWG=;+GD zq@dgXF)?>pB6*ovB?EP{SP9Kb{D!^xIUE zH*M34DF6Ij5*C(EjCl6H`VD}qrP_yfW||znr-^wVA9|HApULMK$dct(iDOF|Ej8&Q z%)va8>QhtxbA`6+apX|nyklO_ZIZcW17emv+J9KJn4amCY#{0^es4O!`S|x&;q~FX zxKNXU=EwruQEBhBAM>Rg_ip{qUKYVgMWqV&tZ^L|WVoZ9A^AM~T<*ok8+35saRikb zmsbVd{!7ol%fcN`RhmILa=iV}zgbD?Sb3*eKK;?|lCIP(pS81me5N?JPjgMnM$*Bg z%v2DoL-!qSS?GaSC11`U$G`OX@o>-m#W;BmmNS2a1(uoZ@gPNKw>e7eNb$H51q;7| zK&eVi9&J}hD(EZBm#}dsTbA}PWBWgF->zB_+jZLiup3RI@vn^o82I0u_{{v;;Lz)0 z6+A73|9;fFN9OIV4q`T^8?a`}_Z&K7gcy=yMJ<){R||{2AXuy zkE-BwtvD;<(qFGM{x7{S|7p(Bbc0WrWY!s=SfAGGe>;Aq(ZeklHV5UA@$;Nd(=R2q znO=|zd;o$Q-ncLNQa#T+RvOcJ@&4tePENLrvO$NBozg}cS-h*~(Wec=h!Wd*KU9!e| z*~ANBo1mU1>wEZVOWyNt*?|iO=42+xtZD_!JNWJv>oD;;{LR~K16C}2#MtQVKUZ75 z`2`lj-SE%Qcdn2S3*|hUfg{51Qcpsa^WF$W|kag6MmTdLSME3jY3Rk@RcEd=>?zkx3X%@y>_#ZY7>22FzHfpX6?2#0FZ z+~Avs>VB2SzV|irR>eQkL-47vMT2!G!rDSQQ*fI{d&PBgVY@TqTNAsnbuALTCrt-( z(T4;p9JcA_Bwpz7w<6RABUedL7v%GrZAzSr{rj7BgBd^`b>oCt=p21NoAbEZHMVzZ z`c&T7;ta#O`B(VKB%4syWYPUjgaiF<3bep~=5^SU^Yd?1owgz$qNh!VBYFc~3?STR zpxt>8jN11pzXIPQb^U^;5a+Sxf*Q#96FH(&PKGR0a8%9XpKOIokVW!dry5^P-?B2s zYH`$rXQJhCw8(P4X|c$xWohoTCy6ZL`TM$5*m1pt5J})m60SJ-Jq`M2Y-r23cq zaf8AzgkB!9_I$Mok7hD}j-u|fY)@1g+>wB|Bdrd1b3f7lFb=C#r}(=*1(s2F zx1OE-LO!2y!Eg^a#X!(Ue-y{azho0okaAWzMq_u_fxC@hUgZ{~mM+zVnu_;cgy^?D zx04FgE$yUu8C_Btii|l@7T#{_h~ONzVHS5!EcVb&Xu!7eRIm15y=_Iwy#3GtgJ<>k zNXl3o4Sg}tgze2}zR5!Pi{PDZ_l%u@WagZr}e~FPRFxnT%^bX*@l2KPWxS*{Js&=9US|{o>tweJ} z;9jcFQ}v3Xv&sD>!+N}nxQWU|++l)F&>{{Ay5-DK0Mfh3tFP(aAe3j{pJf3yEU{hm zAKe%YD?Y7?0O{11Feg7qE)Uu#P7&HBRT#p+LrW=kez+!V6XHDD7;7}qAglEv`mk%8-fNMS zeitRfz3{PVxnh6Pjm@;&m^G+99TUoEp?Bb|s?zvfdKS(qFPPW7TR<{Jhj1s97#77G zk}^SE?2?|Jx$Z^1l2tBR4vGx$SsYO_rBt^d?S_Dnc(TZu zlWW?OF-hEtXiE{hmHQfa+PYXn@nHyz&_XE49=`|lo zxv)vC=M7%7{tPJ$eb+}VSQqlqt(b|*J&e`x(wCd~No8ahs4bEgvbJiIDXC2HbG%t> zRQQsdk+zgN!o@;f@H-Btl}E(qb#+%SA`>bXzx5^P&?W`#G3Bf4J;exqqA{o^m-IyF zTn&8++>ci`1g(b=(%8Pl8T!j9_pQ>_9p4V~Rq$P}1b>D$s-?f9h~xA>#777Tjx?nOIW zeJhMr&9j3(A)M~pG5pdY2e-Nz&iN%rXUq674lnbNKJgW$%-vV>k1KcX6CC^9sdK>0 zN%u{@2=N#H$ft%o4jdvBV0l%i>eEY4Uq5OU9v$_sPqnn({{ZvU%^=mJ{da7As3P+E zgpAfo2TbbRX29$M$EEN1J>PyHUM07MuoY0b_k9=}|VX|xsdQ}U+qi%w& zVdOtfrj1hK+S};of%+x0*eU%zr$DFh1&@ zddR5yQ2Xq#(of2{KYIUs`KrJXC0F%)@#80=aRRMPm_ztdOs^0lze)9y^32>Fzm4Fz zT48@H7}ta&w2h=a>=0MOr3!fnX_A=z5Ih5_J=y>1oGR>-?2DZ>ARe&ua;n6}w?GLX zUiVYrbxEbzN|LzS#i0Kq{bKRE+_T;CSXE9|X}_nEjqT)Ih3?6T62)D9W~rAdac9as z6CQemCA4R*dHwXBJ(b=3nsovuQsQj07w)fk#2!TMPSv7~;wVvVSvhPUPe5WM&!<`7unb|k#*rc)sTrlR) zER?V-cOnse2$RrAb8A?76|~2|n&QryHc|*NXmIhj6Kfkr)CZDOPGZnQ)vsdOyydwLxJF3&%)bZn2W;FtT7Jk+4_OXQ`z%~s<~ zzy%AFqPZMIxba~0VPCmE+SZ|IY%b8zeEmmdH9#t9pAhafY}55RB8ZmWQ?j2j5k&XQ zJYve#y}cA$w+lNu3VkIXL_<1wqXi8OV~SO6iavxK9E=+`sPl7YO8Gn;=!m&}k(Mp- zrFJIDwKpjKP3LsrEz|zQ2k7j>j*CHOlgc!4q}G+9w1Kth3tOj21r-IhjSNg^XLBj zl58*TinlyTKqaVuAmcGJ!_5xNq-g+-UsYTCF{qu1Po7S=S9e3mO3xT&7Uh}vKG26) zv-I6YsW%zr6z!|U7NRO0uq_yrd%3%u;Ft(BlAhL3l7TWwD#+pRO+0?-XrZ7_4TL?| z6H4ZC#rS@9RoDffNL9pO{kGWHb7z$(7QmRu}+h;v}sgYA@k*1ZP&|J7ng_7 z+uV)M=z47OKa7;009W$ldWRn`3oy)}rett1g|liGYF&kdg@ti#6WAZ2Lj=b0Rr6Hg-t?}b8+m*4$I3C# zy~TFi^ktzb%Eg5A@B^=3c*$KgMR`<}!7u+N2Qfp-IDQr*fVS*)OU#AXVVia{4LjM! znO=tG`&e}zIH>QZVL%dayj}CCJ&osn9oGNX9R=5LVSXuHF_H(nRXAs*9z1v>q9kjJiBSN_i~+Wu`x^ zG&_IrjsRoHh^k+eo(nx(+>33;K+VOs{MjOA;w0Y(ZEaC^`GKRnZ$A z$*J&D<=sp6l8WHG6~oNxcZ<71&A$y z4QGnMi!2ovF4vCFeLA?9khUKHmI-OV@gMzq6@*KcpuHpI@^0zPJMp7D&s_w9r!m_Q zWn}`Oh6_Q9)IFVKs}^K3K66>MKQQCjuzB??dI{5vLxl+5>hl?2(oT}h_3QztTMN67`7#}Lhin$rx zqt-~Kl^bap<4*{@Y_;;TIS`?aLrI4=06} z$k2l{FjXURgz?Rx5Q)u>kD)cJ=q+T~7$*{WWrJg|tuC;KhVB-m=qGuOGn+cZkT_01dT83AZJPyue5p#KGWIXhv zH{g3(U6sodpiWe6I-&s)MJ&KN{ng?5xB68foiXpnO2?U!0t~52tgGSCcknF@IQ19U zrs~z`S&IZE8 zgH#ok9fz5Xt43?!ve7)jl!C$Bk1yxX z*ES9!w9s;KaAb0)fBzuXWt@gkc;EZf%!es4Bj{~6KLcoz@KJn{xAu}s+W7>|sy`D^ zmPQ(kNSpSDJw4SEwq{5(j7+N2w>E}r)2BUE?r607whv)V zEF~K$=b7tJG66e@P?42HK$@}8FQQgS|A$3V!85=1`OB-}p(Z)UK~`>uJ`-F;0xU(2 zOvD_|zvb4@^zXK*|FE>@sjm{!punHKcHsAYsJdY%fpY#H`C^7PKH005! z`hLkZPE~5FQ;pWAfQ2h;qbp!V%y2elbG zi8)A8)!9f+6R))&IbTKgQ|GObjA*Ycq3pJEe1=FWDe z!oo~5<6nrB$Q0{c6A7?xP5t(Gw_e`1Z29AY#m9qy{FP(UGgMjiRmO37?^`%})fLs^ z-ln12J~X&SSyL4H<7(KnjHAQe!0sZTRG&k~71x%);ghG1uXySEVmoL@x1Rq=9^*lU zxYWv5Kz1>km2yj#*m))FGv-zbDjJ0xm_MlnvK9YP*a4aIxzv{wDrN>B)kZ(803BU` zx$SJ(q)!XuNgBuS5b}~` z^}G1lIQgviVdwrZuf3!7wEuRJ&B~z7broWRKIG%6ChXoQ@k&RsH(8_Av}(Uf-9aNo zdcJ53**-M3S=SyfcV;%l(CT~E5~_XZz;rhnDDuHf@S1kLismCNaCbb9Ibd@NW;Xgs zP(Qg4@%UcOv+G}S4I5~1$fln?@4Jxi6{D{8u`T<_aX#=A!f0pBK~ z-c*0!tXh|+Uwg+JkXjAX6fo-L|!H}jmn_xN$zxtRACR7w&YZG0O|^h=tMfh>i%2P$e5JGbIrG4mN(k@vA@5@IYBaSr zxT4-)je1^n^3(G?Bky7Haq=jB{GgdUa3#Hu3&i5gURsn-5KAa*UYUWF9YiZF^OL2L zo6%s4)?9MfZ1WNRNav;ij#?wBlv`@7W_Yk|N0U>iIq-QTCnH-heuI5C>c z+x#XuRntqE3^g33JB9d2X4IDwpUAgz4<3`Iqy3~?OUElXqpq=|w}TwNY4*_Fj`h6p z5hxKK$Fs%b(6yAEvAn{%*rsx=S88A|hx43Eq=anBn&VBu;|Zdqn1%p&W(_U0|E=1+ znvEH}ye6hx`oRh^8tXDz?19URF5q6w?NXaP8@Wvhhz=0S(6u@H7iM^ur9hQ_>&q+Q z%3-~TrCtlAM#7FGqOVg`1faQoKI!&}HcfCl3(W+@&O%RO#WQ}w6@vm)Nvq(cIKpwI9R1@pWyXXG6nyhWr*sJ*6LM=*gW@&PF(1P)G`D|E*&hsG+hK7UU zZN;oWg{goYFfH1No3<`Uv+-cVuBU8DJi0cLVod_=p>QPIDAsnFkumjc;(g9-XZ^`8=TPu7R|G=HrN)3 zdc_V?;Ghrz`D3hv3;b9#`1d_*BR;YJgOmoKF;+)g@B~w2~p%%X7_ZS?vhqt=e zezNvp!&OB61H-T1xNrUluBn%+8D!D*idL-7A+64B+lo|TUdx?gkIj#ch{*H1N3FtW z3oHTGlBl~ZEzjEZ<|mZAl3+6a8@8Cu^Vhdu;TXXh4?WU1gE4wJ>IYPl-Q#h{P)ejk z+`uvlmj_`?MZF*K80qLpH!3xO?rq0fV_F@O686(jYHQlf!+5zMTq4ZU(tN%s7|8BT z5s0jerDJb6)0|}<9e3N|_EH}{;fHiG!S^3djn!_--_=U56P?$tGl)zF6iW4<7t^7S zxTIQ;@}y%QQZvn7s`qMw3LC^Vnz9?2-OjD9Mt&XJEIg@ij&|n+^ku}lM+?L|eZbD@ zWRnGl4ER{PYhHhdL<+6MlF{2Ty%A~&Ee;@c6^~Yv5lr%9KD9VjR-wGEcVcQBrp!|J zum=(!dqZzmP1eTn+2qWTvbm9Gi1~R=BFWscr;v)p%JIR*3zcE-{!aw3PV@=#zH2779l3 z`V7I8ai|ib>_OPB3!_&JBF&gI?5Zq=4HLZH(gwI1uSItU9QaMb975j#0mQ%ICN<#R z7$Fn&V%%ui9lbQopnT|HYuGyPe!ltK=B9^c!lR>@$pD8wE8fw*r^QbshYN2x_A@&4 zv=Z$5UqrtHw5t6!RMk6Bh_mQ^h?$cY>6xoZs|v4*Iv!FQ+}YOKZf0!Xh@eequ44gm!LBn53@$B zr&BKN*c3(?3H(MYC-94TStg7wdYmxX+~1A6N@jx&hYm$x6m9(EExmseuQChXQh|ZG zVM0L3&P-I)@DM%RaQWJSM1b#;+hW1*zB`;9FSY%X zrb{s@_Wk_^j&hNt3LSpiul96K6~gzHP`mq45;lf(YAIfrqLVQ@4er`ULy>M!OTX%lNY1%H>*GmB z?@e+l*fY5F7Rz*-!QwUIR7l{l5<)ZxzM)}I0nrFLQ<2sa~I{^ zaqykRQI6WR%B}i4ogX-7#SF2v2i8F23m}IwTv6iq8xYyJG_Uk(nZGsBJISTDW!*=^ z@CyPVY*Dkn_H&MY&vMx97z`p-la9Z|rWe>xg1uFyGJtSHSPz|Pbn%jHRYHCvSdpS+V`w zP~s_$|J2Wp@o&_&@OZpXpl*HPZE{VRcW4D^cYUD~4O<2q;-mGRU^v~8YsB8p82hvc z)(Oy#-)#5=#66Qlo8aNWdi?LYIM?()T^LWbEe0NK+uJ}6JYQs9)ge>nr$apD+{IK~ zVi$W`7ad4c+-;ioYWgGI$vvg?{bzfTuo=F=w%_5==wLE6C}gd65pXCpiK;~+Z@CwN zy&?~Nl8xSajbCedTHK#4qW=Kkjvt$kdm|*d-QS3Y!~rvX{II zfXu~*{tQ`*4RNB2;=>(bxq)fj&`Mymnhm*}k zOxvIw=~!W&*OhG&9ZI@D-XSV6oEUuod?JrBUgKUk5->6C2eVqxgI~UF1kAyN0A08$ zC<)`vh+}gZof5iwfsIx=Wpu-6$hf?RCx6}Yew>8PNQ}l8{7`MYF^JqJu%2I5XEo)K zb4i6AwosZ;?A=f*y-ZYJ4$1UkdXpK~Bvl@3;nDOhaM6_gSZW~ zPUzq~!}F~crMlJC%dPzPqJg+7k|7atw(j19Toibb>v%>DCRmJBcSmkdR%3P~4nhu( zm~1R-7jFCo`F9!hMqGawa!`Vi_TPh*uOQF@}z&!QQvY+#Ts>+;qDo*va+YpFf_SlpMdg7vSsg=Gx-hUrhcFRl}h9X0S z?=SW$rCqX9(Cs)e<0wdoY)5(4n0qTTq0e_)?DRd`9PLO4;#p|yh`qTg#P*h=PjJkYzC1TC2zxyDq*_N@EA4 zg9lqR7<3ML_8MX4ti7>0`9vkySTD64kVT|XsVF}bpM(q+Gcrssvzo1aQlz5&FUbP$#!Jus(SX-tS)fjwB zXWp2=8Ht01Qtd7TGt!11G$o+2VDrkzdIroNVBDIdRZU=J8xbLhEq&qFk3pZTA}?;% zzpFoTUmvQj8#n&4EHeJoJX)mjy+@YCG@jpi-swv?Eg3S86j~K>3pb=pNVP`eW-T9l zqJ0b9o7|dxuGG^IpQ~@q$hq(5ktXo#j?yAUFf0bEX7hjipvG`5vi349FEHNu)m0C5Kl`-zV`yEtU_eqI-RZ$J+<*|Hww@eV)o3{ zmX85`0?|R8t;O?c=CAJanO5ZxWZBA?bE_DOHwW6|`UJ@9jY> zGxo|>AH5^uJ>sNsY_z*X4%)K^(qCr7NuUU=P@l%;Lma<|7tU2Au$w`@%I!j1c79J3 z`QjDCsbxaVz|&XCz(bBEG&xl2q|ky`W;V7BQm;pFYITfavm}FJG&4k%fh@yX?BG`N zL~##1sbn;b$(yXq+_}of7+KXO9OwFJ!$8NO( zl(yb@LXfwKd}UNR-mT_C>X}89rz|g`#FErf%T`4m|y9Hq%f)Cef59Y<= zj!;ZFTZ-Aw^7}ReHl_UrdmPk!H?pe-xP`NdD!cS zgb4!9NKFvn)jL-&y~nf49E~$gps(>%J8@r`kt@HMH|#k>r0q7b2@U-s_eTuN?g{Hj zK#IC#@O)>AmWW-Smck08v+=(r7rP`bZ2uW}i3ENRvYX;Xjf zp&4pG@zM!y%P#7QguDunYh0gwHya53x#M@X^s%$WddoV`Te;drK_J`e4;eq0&Js*- zlg?)2AzA^xlNqy%SaC#0;WLiJS(_v9Vp5GOiXtd?N$D@nIF*T#B#kC~bfTIdQyMr@ z7R1kIl9M)4WTc~>T+X4aZv5UwY(ssyq_#k#&U(1Qfc9vk)W{@M;rH@QAlzgV)}ZgM zlh6aER+{!kaX-Jn0NF1-9CaD3wdd8cZ{)NxPL=QoV=?10oL5bF;j{c*NPsO&>5izg zOOKV*-cyaXGp{tkG-AeK2en3Lh9e9@gJG}T8Z@`4B>)YbKE|Oh`G{V#WxG4$p?;b; zv^(oTrlAc|{v=&9>rtniMd6oo$(KTQwi=Y=s{v#DG!pJ18+5GZ0ff`|7rLCEg%H%xOs!xZNM_+48~t(4%=3lKWkc3PQv;S)|^3(A)Vry5vo`_Q}d1?PEgH4|n@ z=rUuRPc*r9L-yBqY66IlMaoRMO=D>L(0+*Hw%Cdp5LZtyEH_fAT8@VM2#j>NCAp|T z`#uBo4Wy9GJ5?EGU*%w% z$Yo30g)zLRT5{wWk?7JFxr=%Zf9C5(t1v8O@*)uJLxBBW(_klz(m9a9nTh3wfNn}s z3j`RR>7h##({;e?R|S8cI?r#$ts+UY zg@s`1G{$#i%aD%Wv-k^bOMlfbl4(6!NVHoi&j+v6QE)~A{LX7ZT%iNJQr=5%V8i~4q;(jy=`mof z1yQn^Pa-*ZrjGeo4(h;G#9y!QnN1U>TNAbE&3K|AR5HB6W}L9TFv#Wc#3pInfA%PB zV7dVaD_w)~zK)3?X20Xf3YnUkpjLJ~8K9Cr1b`>!kj zh~)x9+*!NT0W0IqHNssi&;IR*ya@i;Qy@yixF>tThG7Sg_UH{IhwWKIr}pPOnsmS> z<AE(+No{7j}?`W06EJMIVE2ceC!J7h+iio}x9(Dr!a;~^R? zB&?Ea0DTdm->&y7K)|f3LukUj=QiJIyh_|-av{Yjyc`p{decKn%X%c@^XF|X&F&yz zn~1(H^0RgJr&=pwnUurWmQ?cN6jLWep+>gYWYIG^K5IaD`p!MJ;tt&>f8M@=e%C5! zUei6t)?gQz_dQ+6E$)!Cf!9em3W;7xx^HH}&><$wU%5zfk=>yVzgrLQ1y@JHsp=2QZt$FLm5C4&g`XQ5fC5J4)$Deovk^@&d6~JLL2|ekG zZumMq^-C)5Hf^LGs9EcRhFQ(wLm;+^e#d;0QR8&*Db}d`cRnP05!f)X8VTkrGG&Yb z@NRy)b|iSTOhKneLA|~`9y7Q&dt9a?pTtXnbiG5ht9h+YwG@7zp=bDRxr}sgAx<>+ zeP{4%c6HE_{UgIOzy0gyCvlkFd`tl;D1cZ7xESJ~^?C;bbxEQ@t`wGuC4-8QodMO{ z$}an&OzC*h|3q?RPqkytBc9RP=AyC7Foacf$qP2hCBqcB8VShKK2OB>Bk~gI-~*X^ zJvYtGR7ou@5?ynEV*+WyqjGT*!w!8**t)_L*j$ozvXG5?n-6xg3d}gUJAVKufy`yF zY3Ul9S_Rj)#^sRPkc&JSlJi9Y*j8_y*KV_+hRQt&ZP`Y8wHpCvl79Oh0n*DChn!ER zUWo!R^XeH_m$r?7Lw=Ufs_qizb;&`Gaa5sZg?Cr6Ad1idmUX}l^bvJwlXlMX%W;Da zoC==Egn^AJN&PDV78);Y3-ap&`dw-*8Pfpy$H}kZ2YKA(>-;aU(`jn)-w(@I?Rw9u z+#Wo+N5P)d|3M!vR0mj3LhC+C<|5YUX}#oI%LtoSYAl#KkGk{4r=hVl&B=n%uW=Iz zSqX@UaB7vby9LpSQaB|w3_kO|GvAFMK2MBs*)h1t^G3y=-olJ<0QU#K^sM1FmZWj8 zQAm0*P3nqcPr1e_(-#Yv1U!4Lh{<1F$lpBkpc_C;wU(z2!!Z5!Mhf6B^<5TfVewDr zVSEyqcYW$BwF;_@@>>2mf%0xrz{@aU5;65UKVJmcm>H2*Y9d*Z8sV7_#1=9i9qd17 zco=$9W64-bwbQ(1#bIyd4!Cj&F6cxLr?-&|UPxnv_iAu^RQTdXin;h%)`Qt0nAth; zYh*TvhD|ysLWjd^G|v~x$w2{Qn;&uldme%PW<3ag^IoG8+jccItk14qj`^zng||rv z0gJki^`1BC2dBDX(6NAe(Jy!IkF*GON$uQbxm$#3A11?HAZ!4(X-G42Z%za%@g8Be z5iofcV*5x1^>xkY9AkP`VHZsVd1vbR;Tr&*`r=~Q_@(DeoIA@^R_(T1q>5rp8{@{h zB%iWx)HsR54*_(TSEY-i{ z&!}$6Zi2EJFTt#l;phW21oIBqr8^86U(S4kh7#PNHu8l=3H0s`y}hYU2cz5=6sKnm!|>~3NkP$y)P3SL{I|ntp&16xzyOa*fTPh-k!5A?#Lb@`_mu$ z)+ti+MfQ_16tHSAtf=cxhV*^t{_4G1S=)}2Cgn6S4`-u@%!$NAPf;LMxb1N+c&q4* zX`V4_sr|>*^X1dY;y(U-i&o>cG>H{F#aPYg$~+QkH3rzj6TI3~G2E)?=sezP5bnD% zt(8_6I8pIrO0Ia`^zSw~^1pu!EeC{QZ9o`qu? z!c7Nj=uG*wS!=fmu&4yXSCOFHWL|?Y`c2U%`p9XNnkP}|I8iQH&=|0a-dW@Zb@rk? z0BrZHCw+XuOXnY39K71&bwm!efG|Tx(`ysJVCK*o{xP^-tDR8bN2V2Zu~l2tkrOlg z*hWSZLa!c=cWf$1me4hNj+FSNOOD1`E~jk8A91P790L75D$q*%hY@R*BjnexbdPf`5jE{I-oUNw$A zvyW9Ocm#I8s6RW2=i~;w_q5U{u*pJ?O0RH_3;_SCCo;yhjRzS#)jDnj?FACeb7ma6pCBQlCq?dB|>G5P?m%uW$ar?*0JvlNhoE=o^6tS z#=bN57P9Z#*!OK_?1M4pce$VUd7t;byWijEoxe=+U9RQ4j&nJVlW<-;sTreDp1&_$ z179RTYp6nJO&@(M4UO~!H6my5J~89covWDg31NmG%3-H-8@0dM-jRlv>cjJ&_TH)c z!h{5sOKxLce<&X;*uAfqas!!pQPen79YvTyQ{o-jb);6MsvCVSjE#>Sk~J-jhmKZs zn;gGhzyPqAd^CI_V!Y6;3oR7P&khJ2^6}y}xG^)?tIp%F&N2gh6G&xp9Wgjtw2`(8 z6;j)*x^gFN!lD!jj|q-@8_Yb6+qkjLmv!T?eeQdHOBnapF1Q{eP-tc1yS0DjcYd)H zBRsGF?PTk9cD+W(#9@pU@_m$^vr_BsLNYyn5U65%iMVCf^lNOXct29wo??5E#an&oUuh|PMU1QHE~et-r3a;2)|>f>nE+RE27 z_xTlGb}~XYql9Fp#IqQSC5Xqq&|WsFtuaM`=_?pMrr8Ay_t-S zt+jI-v%Ly)?^C`-%x0%_Hiu^>$2|Y0+yY`T2hM_Y(R3qsN3=nK`wd`2BiuNojPFt7 z2|DHm~UFO>&>XJR{#>H+%PWI*f1)^82!_&`)2IVSB(WVtG z5B#qt`_{?6YjDS1hfY`9n{*iMun|xo!(8YM*e#xO{ig>AC#dOSWX2s)jUjgEaRF=T zWpDn7s%2&MFP{urKe~)zUWFL7Dx`Sd(bfmY#Eq(Nd9WSaQuAD!;bK`VRl0zXsO_qIV@&3gHV=1a+@+IeXq8v0S3dY^d|8_ru^HW6 zSUnUq0-l+}8I$1cR z+&+N_jORj4t#RGfdj*Yj9nQ#5V}$tFr2@x0$eX_-qsDWN?Eu}lL|#!gX4DqVE_fQf zTx9zrL@|&nv>!0Got~AlmR?gB812b1w;3*SwfQQqa&zioAd^~zez~meodvIR@U2ZD z{~qoy9Fh)Cl6}15AZNSMx;B_mz=GkzdkdpayeWMaW(>PlWvZ$6@gx5e#THLuwu(jU z)d;}yBGi)HVoPi{vqn8Y@#RLpEHDXYE&{|Wqi4W4awJCqB9E63#|#F;d6P@#TxIy_ zvY4J^>(k-)tJ@+9l3sspU6I<7>iOFLO}l}ceTvT=D69l9!r`3s29iD44{at7k^RV- zSVsm%IEbZq=mm1+Qq1uaEs2)QUwui``*Wr`Nw>5*sI>+ir&TN7_?RM0tf!hw19!Z8 zUkT=n!JwM*w_moJ&0G^Zt!!sY?|E>V_>nDDpSwtJQ=2F0T1+4ATC{f9-B&YhzeM%Z zRh2|worCsu@h)roeG=T>$$SZ^OTHP~KlO{}1z!^39FF?{GR}r#vRW^*FWoxpxLH#e z9ruF!D-C4&lyYuOFuTlft=cL%GdI-htwBt0OP@q$ zdULJe2aq0xm(YW#Nc1*rd&CuMtm>E9G?w{+1%v0b8En9DZn=ArQ_VEyzIKDkZ=!yR zx(TBKN@HmSh-YvrO121;vfaJh1n0iNz9cp1xv&Vps#>5bq-N9l46*k1xNf`@w2J+X z)Au^43ZOzx$R+Ko%(o}lWZ3;24BkbUYX!oU(^dMCXHL>D-LShO0|tbpR=%yom_SayRigm( z!*ZNsIu~>3QlR0~AdYoKo}+t*p>@6?y|XMk}$L6iW1j%K2Q* zn7iF6w93MDQdiUiQ`x{samwd7lspf}M+S!6;@7t|B;c>Agc;Z5%0DF^KaozZzc2fOoXd>f_aqO@Km;!D{q*#i%K8KmP|XSj zFDwH~+QfPHWSaPSe$ixn&#$F;_0SO&m9483lfT>AvPFaSbGqFww z6|TE!sd=E@DbIFG8yOk;4|K?P6-Z#wRF_AJ%jNu^B8=sL)`FVH-um~`YEMJLWPj;0 zzTfdBBysP|!aMM>M#er%+;xu_9_^?YN{;x9(JEKf#Gg%n-y6xJ@u@}38sk3hVZo}V z+pd%Kj+^|dlTmP~Xuc#fI~Z`A_xIhibxY$-BztsnSHqu}7^wh{M=D;<&wo}>_j@vb zK7>!S5Q4g6<&-INvHCgJp6R%c8dKlNyTIs9@1KJwdH|6GgvW%D(uT*ND(|Z()YGv% zI`yb`b(iSC-rkP~<(qOVqYJs^6QA0~#f%9`ltGx@S25yYvOIviT{kJx@D@agBrvkk zD6JnlunZTy_q&}kMJ`K4w2HI@Bi8tr$yW^0+7n8iwx*wlGeH+Hy)R=Jj4?JuMbFKk z2y=nvBRn&>?tJCr6=cedH!rfFS7o*DI#&+RQTGN_q>SP-3k+OZmZM@VPrf* zDA`6yvZuBOnXWFxN?er$H6J{=qIUb?B)fGB#Bf}ftTO|Ahs^dYbA-rGX8-uU(0q$9 zl7D)N|Ll_cTk!fcLUsqJQ4!nOhuM1QJQWn6VpTy668V)I048}jWuhUPyHg?aifOGhoXzus5H8boB6wBP^etP1p@YA-JY%nLEVO7YgcPUN( zmn&hd3WEB3k!g_(aNo6`Q9NEY6i6XXqk^nIE=WrD(R+$k>zx;T z9N{+o9!^K$G<2UXqn_jtE+CG&%=pgYUg3xYytePNuB?(pnqyD9n zWyyZ4=6Nd528l4ybWV=K5z4N6Zb6k60K$G)?PeJ|QNrDFboI(Gnjno4hnn#sPVOY3 zvQke^-MN}da(WMEOyWGlA$}KfT3MZEWQBthdI>5(cGr>ei{`=-<4ac(EGK_{CMF&l z5Fi54q8syZnuRZ_R9=lIYAy@{wD@CK*2tJL5z}K)Oru(rgF|&R>6dxwx_K>JfDYEz zy02)eOdSyv-uaEsabTFSlByswL8`VP*x@B@_>St&U|AE4yCi?o{D6qts7K`#ZtLlH zj348sSy*ZMxQ|TbOb?_wUJ4vmF7t5Zq)a&xpSNW8$&)An*BM{2Q_AWWy2XuZeJZyf z_r7X3Z-R3afwVob5+J7|0Up$YxofPvQ*DW2tuY|n8}&Z1YQn2_aE`)}zU@?$LnAXh zQ85j(QacEj$xm`^o8Qj6B6VJCxtr8I$G`wrPu@!MEB{X8?bV4K{03YuOa~>cY>9cX zGE0fIa@Fc-dlTVO;JT-nEwy!8Fxavf4X&NzniIkw?g3`s++@QAKLew@TTFZ$u7JBH z<&NWdHsOWY7Zq!H@LPKBwS%Ie7+{32qy}=T;j-iqUAY_KhX#z|0+8&O%gi+dfuY{L z*4@Y_*z2To=WhEoypb$A)u`EsaK&V5WM7i;%7Ga(7J*+DS}K~00MUjt$KF>xev_Av z+>(v%CeBu%c_-MQHj{LbjHuas9fnjTyK58bDg8n3M|rMYDTOGHD+7)>);a zy{dioetgJlK92N<%JHI0HP`yO(iIksMrz4KnO_%qzg@d_<@6QeT|}o@bDQA8xVPPB zkJ`kt6sTOF6cGh<95AmV=XldZNb%^N{pq7wE0qzY|c+b=(Y*)uGv8h z#B%s^amGElw;a!AR}_)n8u(_3^3B4}1rwj)0jf@BulY6NPMwNe=_-l1gI0}umigHI zd9WqKwMg0_jkMn*kYD-wGsRt&xpw0xXbR=bAK#s++E>eZw2G=m8mT79n3W>Y5AAd~ftl-7D-!8o!*B zJ42?wlpC^}36ENZ8B=Y;G z0%mu`jJV$+g~0i(lscotL75d4XO>)v4Iy_W!5U2;x>#X@`AM}9ggjv-1(|$mo;Kkp zcb4YM?vg9KqN}YIc9DjCk6+dcf3Jrq>oI#mB6`%K%K)1iCVOUkxR0NGX<=hdz_4Be zKj>HIpKF8`a`AJ*YK*9Je3O{)YJE&-RLj@(IfzJ`d^$i4@3VK2*(7RWm}>1|4uoWH zn<{=2S7VP`C8S{v!%Zlw$g7;~l?gdZ?+!Y)bCU9kfBjG>JG6jhTrk%^IN@^QoY*Oi zfI}hlvJ}y8!O+?wF}CRtU)}n2JAYxJFwW1=lEvuL1l_CvM98sVQUW?MlJYG$l}hgv zI|o`=+8|{3POkoMv=2LwJFY9@;x0osZHQ2!sR^W1M7VksBCS;+ z>GvghxJbZZ_S@45_r;eB)SVBceD%wd_n1byi@XU-GjEjHBIsu)`8$hq*lG_hD z2}sf~G zw+Eq&-^?2}b62>y83L?W3}ZT%Yjh^)kn|JyJ69ZL3GM!01nJ~xN64AQ3eT@6DFBUw zOP*~U<)8uqebyTQ{amk5VgU5UIVt--oXs1H$-#-76$h6Yt`c${%3u3mHFFv`Q%!XlcRLiyd_J}2Q$(8-43=0FuC^=v9(an zc#Qyg$n_}_6{oJeynGQUPEZ%FD>bRK2l-Kp-K@#xt*WoMzl9VdWXJe!KI-hAyHOCQ>bQ>==de9XW^Dax=^h zc4t5uuivJ}Qn|otQu>h49~VbzvDPs;{ zznHc6;WR4D`l?k01<8R7gU)c`o?F)?2>%B08#Ujn_V~GRoyEprQ6?Tu<`f@F8vQma za#8gl)2TAF{NnhbI#4|VvUiU8yi_(wbI-|URKrY|%=Smp+<}_WJ(Nd*F9&^~$(}ROmj630;r4bi4ioh3+CE?2+r>8jfh6qQcjhT`lw^?&f(^pyD8X%-Z~*Cn8rfL#DGW#`i4gga-QES-KA}GufRRub!kLUYg^#1 zOPV+W<~ZKvB~-Ja?@rgX=Q%Yuc(OA^N&a}Ir9TKZ0Qs#HgxXCG5b+^DYi;-{v7Ll@VZ_&P&LWkN702^m@Lx3OL7F>kQwFLN_^UsU$u zN63?O5xKfI0mv6w`Y)SDOqUEEu&jXVI^uRN8LcPMuTKV(-w!t1wo^parK3wUR=s=Bx zQq5ScVM(RdXnSFGVb+pakj-~UIm~VFpNjSqIj8eL=$et>{oqM@_8X;=bX+su@3*s52c=cQ%0Dk$b;N*44_-QztE_EYj&ZK?Bu zP`U7y;3y4;6_G)JP*N}n_1rG8lR@!`I{#Gi z#D+>t6zw#~9KydC(Ol>A$i4S1GgS5itK`NNT4?X%a%OjeE$2x`X6HBJ%j3_kSK!Ak zJwjqle$CO>PSyctq;?9oNSjW6LJ`fOMql7|>ID*bkqLBQELVzz@Sp zo|~00)ZB8MH3IqA%gu=>>2XQGWL=$5owC-qUC!hhfc+M5?K89FQ%EG*dURnwR$hQc zK#>aGn8|t5A-q3_^o5q!&}i~Od6*j2rj;M@Fy%{BayCQQ2OTBV`)>8^eU(L7C0=;iz3U17IHC=~o}-dMN%4XI?QOElZn05{#A z5Lj1UyRrBo#gDwlj$%mkLq3@WMPtU>V^W=Uiy7B-)6&pCA7GUCefE_$&*!OXdQ?1q zu1?0^p5xb#%`;M%;>%S=K8+u*6Oyz5YFH6B?#`jsbNVk296ae^6`o6_pnNH_&Ge=2 z8G>xU0mlSU-)v6=~syWpjGYG>*;R{$fvQA6DA6qCawxs3~FB}^@WF)6_Q^c&x?Qb zsBOt^vg*w-O8e`!gXgQcvuPfDWb--PmVnx-t>u&gArGQ~ZYj&&ya7J#g5-vOz&az( z2&0Bfq!w)25WR|AG2~{0{Ln1mns1{W^S0L5b+i6lMa|E*1)%=5#{`WdEIoEWX~+Lk zXTDB$NW(&J{D2Z05GLye`TAWWg>-I)AG^3Poi?dbQs2GsxK8qw!>);N-A_u#UPj3X z<;XXzNQSNN&os~>=aD?AugjBbfZ^$MK6Ey>G^(N>M3HOhE=skPLC0dIWo16qcMqBP zmD1)lbLuZl0B}p{n><*n*WC4e516i4HXY_!;RVWGvYgI)l}RA!T_Vp)d{<5=*X0^x=kKh1l|vH09}pP`hEW4u(e^0*Lq)b(yugEu^LFd4y2Y<@ zQ5#S1dUW(vTNuajx)_h1&GJJe^&z(Y7~5zShGMHqo3>DW^2)4|-mU@^;g@IF|z7p zGqt@|P3E%&p&hGlK{%16x0BA0m!-*;DO`_J&sxUXjbTAeL}~dOPgkffkmSG?|3py3 z>XJ~WGOz@HF2QF+(9WPksI9l!9K}TVga-oPFf*7Go1*4B01L#e^Rbg2BcuiIHVZ_ z<^qL5%#!i9%rYP~E0-iHGy}NBo@fQs`4sT-f@np8JAT;A1nh7v zj^$1ZlhxY$1_5DkcXpo(OTnCeE&-8YERq`sx1FL;tCyLjdX7!GFW1DTkbJ9sZ`&R0 zeutvm%GplpW0Sq1cN!1LW7Tq?_RW+RZm|$ zRVyg*pk3)lm3s@%bv4{Cha$iBUW;3(*{53tOV;H+FGRO8{}mLq+4#ZEbRX!EhYhRO z!H!%(OsfAJ@`aXqzy6zrtl&$YedG3P{@cr>NN?lbsY)?u|#E?XMvWIyDt&Rff3v4v0Mw zyLk~+o5t94$YhXQp$mZZ_H4(}zwu3VmKc%kaoZKX`!4iDF+52My|f)jEu}4GUm_Iv ziAmJzU~d68w5)dSz_;b2P4Y-yStzT_2#1cJQO|aU#)PkxZ0$)noQ`A9Y!wgrG?m!L ze{CsqzJFaa@RpGP^OMbmS^pDFLC?2{-Kom*2r-`v*Yz13 zq|l97R(Kb$fngL3Pu<(9BHs=Zvl*mXAQ=)UnLxGH8OnEV_;JQksD9pr#jH zAx!08%VWz9Bt};oFV}I?=|9#GyC4SzVWKAvKmE{qUC+No1ez7yb|`TfzKzl&-e(cz z*Q`o3a-Ye?vBf?^F}M_+P(FRyV6`_0LAiO*4l(GemfK!3tK+3xe7q#N=vr3|0&64D z4dMd#qGIYlQL)Ed*J%Lq7V8mi&r+^x7p-#G)!XTp-MW2@DL2&jVqsxniYi0i*QcK% z8XOVP^V?}V-qu%cKmsHQMLya&FRjW$@ii0mXE?3``GdPAFD)kd?!F6W$S!T5?}IByeuhQ_L+*%jmBtatZpM^_u`aKG3v%@Jlj3DbHo7 zC*T?%*0N6d^Fd|8wH%dJhYG(z#x$RJwZ#?M6&&DX&`Wm&G zq?4^bP%V8(Lc0xvWCW;@S{J`N)I{?xSGVvs)4>LI5U_UI-(+LC>dyF$V^71|rzba? zwl0c{dZxOz|B);*$Q#)1W6~LN82ibF)jxG6*g&hHDIumnbGABOx#5c6SU28FCjI84 z-TfX0+l|(Zg?9t#gU_sOcE!`nWZw631JyGVx!xj9x8Pi^(ylQ3x&)!5rk|ZC@j!cT zuQ0}cJ_Am|-`4&ze2xIp$yN0&YWdaixB?9i@4-AHi+)_Ll~REDGKRz~3P?2hW2R|q zOGW$+Y5p!6clSz+e3X0zjFCC0-QKMa^C<-K?q{$5#MM98v7n-Dq=Y!o{_J|(!d9`c z6EoZmH_Ge^o4U`Hf~3l?ISiVs$#{(kb5FbCa82K@cYtR6=}@4foq#VIqXT$nP6gvh zzvPwPKcHu|(x2#PhAm;T8qjtaT+XZXR-T|Pz#C;^y>aQ{i;z_={a~v4AbLafr3 z%3pC9{LpZ$;d5@)*wO@`vX(bil_~8(b?n|G-zHVH!c6Ge4IGE{o}aJ%)r57pZJ z3w4O%Dufc|=L|O15|cvWyqF~$!DOqZ*2T9`@nEI=Vzy6Y&#ks*-|g0rokvkUm7k1B zx8VUNy|Tz?JcDk@SapM>8&O=ovqcTrqt?BR1p2zswpAv0_0Q%RXSejGMX=*+^x zubH02M$25N3c;7%abDn|4SDG{$>*~#C0uO@q9tLNJdh0dGd3k5oEf}&ian4l>#;VO z_Cy`Ra)>m{tr9$qsy|DIwSfhA2TG%c@5CBgmB;e0@SjTOc1W8keLnL5%)Vg(g5iWk zyiSwa{iM(a?u@@4Jh0{cj*45#=ZuOHc9FEh=?Ko_@ zEcLWgj(mlx%~nC4PvO=y!+c&{LBabYfA0-e;hHuITFK%TfL4T-4J02G# zRo^CWYge|6-A`Z{88Ma7N9+uO(9)=cUcO0P13)$%8?$w{bUP8)WTMS*pH(k>SayPw z$<_-5DBO->*BOSC0<7$@jX%^VVJ~#Me%9Dk0$~~-(*AatgspeeM^Zl}5_DmhmRE>I zYL{E}W*ErJzhQgC`hJ9fS?=HIG7G~f>CC$KhcZbCu&pxIP#>$`{6W?R=hCP8bN}nu zrbk*NdyfY?&4LtzrG_lgFzDNZPXTcsMvX`8gr9Y^@IXqA+=eHU3ccn^lg;^ZzP}h) zMk){I<+XFt5ibSvGdnY1UjL5=Wozr87)?7?mt6+o7N_?e0`Xz=kXm*&?} z|AQmR)v*8S&`XrtS+xRkYK7OI zWmI`G@$HwvUSz38hak1k7*%R)FPd} zWXZxLDrby_d2H&9VJ2q9zOVFL-NzDzV0uh2C(M>*_LoJ!&1mdAjk4?kK*9C`9yYfm z5ES=8zp^s^ex$)tk$Xi~lnLPuxw0{rcw{a%Q2EaJ{Qyj(q|zERuXBVwP7og!If|SN z7BN%x+|t>;Ubotsr)p?ij*}Z2uOZ6c&jF1ShEIguogYW94`4>C`t#Oc)1;eRgXXKyS#~(lIA{xiVTP}w>p-jV~6ePyv zaR;Eajjr}IvzUogi~O!y!j%M-NM25nvDG`yx7j6WD3N7$t_{&~#lq@_IJpjYf3Cpb zmMD~I`T6VfZ)B#JqZ^0I?I&(wtp}>ji(Wrb{m3rHKkVa~U(-GR^JhlYC>mJE{z(KN%Znq^fkPgcABC} zxh*uc@wpcX7#mD{sZFGy0_0<@Rqp^uOGJp0t=w$7r}G%18ebxpwMrDP>umoLHaVUH z+Dow$JdPhbD>;3f<;ZMY{T+I=y{-R5zoMwE;{S}-LDXmW)XEr<*P_$7CIdpPRz0p z9z6VBmN_m&H;Egq{D8UCF5%F1*`Ur>A->!SJ5nliV;h7=F+g7A$U>KXp`jp9-ms6x7~rdI#Hg1BlNN_5aI`I9QmG`^PJDpmpqz4uQ!eXfgyyTepykh32Y3PW!P^g=!G7G z%=mgwrR%J&swKzE@I#->F)`hq-vXq8b74F6^@q55OJDtU zad*3%T%&R!>nD65x6<6zHB!2AWQHU)GKP0gGeQArB9FDN42x0dRScrlZqu zx5wS;<#`;x+R+4ZnV1HJ;IFv+=6f@s8v=ByQAyl+;D-AnW-0?9+W{Rd`E@2xP+{dR z1XXaf-azIXoc+IBOpWaNZs2Mep89Gt$Z4n){i~raR@mgZuc4nh=VZRM2&iK3c>c}r z*C%MlGH`nkO7X+kFhk!Ge%(*K*V&^!H%)tBqV+Fx3u429@>>|~yvmk5t|YG^<+4i& zig3(&nV1Zb%W2O_&#eSY*vv4;oMQdg;~a%ZPaW-mzkXhbQ^Bduu)er;%OaRUdfbT_ zg&8twLL-$wO}%dRCm+@{n=0+DbdQJqVW23)o9@{-!DvusR~!@?uUkWqEuf@5uSDmP z))y2-ak0BhtW-Q{wZ5A?GP#_pOkNDTD8s{>Ii?(T3w_GO;uXGp4Jz$4YgrYW<3QNnv+?j~s7Y3> znDPiP>obmAA)jOaW|iXY2$lfS($1mTcm-kIsJ|M%yCRAM={upSYX$CB#x;TFN$NJ7n#cA9G=fCWuIg>p6tfAN6bg&#pSfe&uh=IPVlY?9lU>#FKvN(1ZH^?G=e3t}vMed5IHL zXWa6S&<+WNBE8C-`t4I1@@9T&xa5~Z2gluTBXh=^&r>B3c6~`=%mEDraHaEtwW>4^poMXDhUQ{8@2;P4 z4uqto`?j?-c_U#Jc_2ZF+0V<}hq}Iyv^u1aurEHJYr8T4{k6m5f22Kp`3M4l#r!88 z_^UVe-SC{G^)UY}j~Rhed;TcolfKDs8b2x`xjwzjl5xUq^LTIQL@jBc zQcZ#|vU@sm)TXd-Gv`5jF!m#DUNMlxeoOTOeo?yP#i7q&TV~~+YM-gR1n;`DL*UbQeP_X?Wk=$-_#ULq3pRa_EJeTik1?twU=f2 zt3~Lo=vQX*&R7+3kvDHf^0yo16uMsEtt5i_d zH#_=b!L^m7AfL`F#PyYFj>sG*z`cr>8&AInNgCfK6Bz-x+k=X$6AQW9{Vz zVCWnht_nI&enE_tOq7ww+cBk>upK}qGl-qcH0G4`d26FF^V6haqA4^X+!k@}Pm%Q_ z$1<+pKNU2x1>2%6#h(k(#q%3HtmJpt`o$YMY6{^@UcZlUT^Yb6xeq!WPe248XkOtH z7wV|l@eK$YeH~KiY^Lx$i9_10%8H77SM=8EvpqY4NF}*Q8U<>}Ky!dT$yB{v02VEh+R>42@1r7U;szL3%{q`5?=;oU9 zjFE&mHzt0~TTJTul}_p2xl1hVP~eVhucTz=D79(}-Afq@|f32Pf-LmDe___w?rlEF$*2%T)3$v!7Kf3DU2{~A{{sXf-oYG-%oc41H~P+S4E*E zG|J$xpwA%62|TMslDbmMF_&_s#qE+49@OCh-Q+RayjAYPSyn6Umd!m)iK+cul=bp>KrN2}HBfTwG0(js-RQY)Zbq%aOYlXB!qSm9|>5|t1 z^u_XAyTH2hhK~4JfW8kUadjqUA7*0T&LhNO8IT;W{DcR`5llhNRb=rqG5JeWbeH09 zI9&6gOc`9RUC97FKad~=BP(EZh2Ns@9|wPnUtH_1YyhB^SU=C)z`kAc(Ukd*S-Eyy*g8=sz@ zHvICuGg;=O%vtAm`|@luVM_<2C61}Y1m;fjzj{OOJX)yXOkXRrpEwdHQ#GKw?@SEi zz(!)Nba;Ho7E;OoKt6!mr#6g0%=4`xMTbzl)6;95RBV{nO7`bTsR%zk+^G--+m!6GyRY93XLd*@0om!* z09_`Z3gmOC&r~f z`K(06RUawNpTm_lu|JxG3r(6r%Yp3YaNFl%+qo7>DYPPn_kX_a{wlJwn?{0fw!m}q6$Uil^fn?KCn>)h#KF5J zuB6zzAAUo1wCv`cuZI#v_y{la68^Il80k2&BSP}sN^BO?F@yO5jiKQ$w_YeA20_uTrJ(E{BdBzn3ci&)?S~98+wQt;Aja zE9?aB#O41++<%LWfv>=bpZ)K__}iF*ui$Ju^{42`Z%)ad&wkC}^|1dnG{>$}0)~WS{RyFe zcn5$YK`rpRGxI>V%oZ>cfj`Ac#FqDiN7S2-7qfu^Q+=d8BJH)sC$Z1YAUUlIp82=W zy;mTklhE8!|LM&Gls4&=fKFZ!Skxn+rXEB~i?zn=t7$l~6LoUKq%%c?y+&&I5c#)@ zfPV5SXWW5*Kvyc6XM*4Nc;TzO8*3|DX5r2KQL ze!E8Vqf_kWjGs9qH~K@tLDkVgdF*v0oS!Y%a4#X(xQ1t!;GgErN#sT!&EemLe~zwp z@0I+~xPov(mFtR;Yw8K=GX>7ak6l(CT`02~vsQ~0Djq|-^P&m&2jootwafng5U)XK zPgDHUuzxBDZgyc5{|(;o`kE9 z*Yf_krGC5SXFy@GeRK4Tvpe2~+SKrp>Sx0_qlEw8+#QX_>rvmZRwRI9&TEtRF?D`* zKn|G@bHF@3los|E}bv7@9HSpip390`iFGaJuJK2$)Q+7WkyX_Ck ze{J!-=ADLPwqKr`S31y_U3o`iQ*GkYW3q*e&X5ju$V`u(qvROyGB^Nj0 zcllZ4{r}}51^)7oZS~!UBq9{3SyzfXq(P-Ygd+p{f4iIZ?IhmqW_*aP=UG+c!E%{J1b?uJ?|d9;p+eOo{LAFh%7X4^z2q_^qjJZ z*Z;4pyEZ2kuc682@4&I==JAap!g?=KFLyjf>hMVo(yvMo@k z(C}t6Fo0E_?9Kjf`~COpg-ahz`0P&_We|}n1S!MePC4@G!Cb6~ZR6Y1xYExe7FA~> z_}M|=_FKlI&o4laz0&93f-2WOHL9$yub26eeWj?6-4m(=d-Go>*x&DPtn$Q>>9jk= zzUOLnvzSUe?2^n#&>t74eL>Uk{`11dwjenH3|j5W{tsG|Dgj-CIeKQG`a~!y%^^Zw zm4*kjhY2_nq0+SQlB&z8_0WuS9)8&Ac=yiHupg~rk4saY496;{j?Tyb@wtaP+Bc3! zYWRT{(=}k9KB2VI_?2dUh8Cx*#XiW@it4!+3HruXwJ3v$Eddh?enQ5$ATQ}q^&KN! zSs`3cWF=H2GD!AR4`27^nlz>cSHF9Lo%oUF?6CVH+yA&1_%ZYRk#QS+GW9S=x3Cce zL0+Xk2rc>*$%_toaEZ=I-D>ULY&`{CfpO8x_n{X>em?kkVd{+~KLD>RY3GbHPF{i`gd#_-=f9-GbI+6Zp zQA?bJi`$QYuGC>}(V<66!+U4dG2h~)L$1a4a=*BgZU#RGh!U`MK#*wMcXx_9;pTJsEE9_oIlhwb$Kp#kI{TiOKD=_3 z@*UGXmG^*di2-$?IbZ<}ltB%M^9lAj-UOV`9}&Ifx5~$6oa84L*HXR#!811l(Be3z zir--`1X5=8Ox}U+PF%_kgJg>liR_U-N7Yv+PT3g%nr1o%glO!zo_Er&(`Y5?Ioc(z zP9~e=>Y)&dm;bNDgzVBkm5ER+6s zTiCIysR3h+)EPH`q$K3g{@wxty0~-XEQRL-W>UKWEugi}n zV1y(RsIDyogYlAhLt2wBRRbZr)2ksDNF^aZqB;4HR0X{2r8De%Gr2SQHW#x&bcoucx7sP65@wK{eF zMD)X=dQAs=Nf8ZM0fUqxx(hdNZV}Emb1dpE+q)m`N5v2${7n4VtcD7nXOS22AzSo+ z+3>HQQ3JE~J==KdA;f@+{?aP)ocTcc|KsYtynxL&T`@48*z&vpLixZS+puh(-t9`}bQ zvDzW9xOr74gOf69A2*fY)T#WMGK3ZQs@H>V7laSzs3%)EPL!mA*t$O82xKm({9f>c zL?n3cfU2E3ZxKJ<7}xq=$xg=~~S9=(^E(8jcWZW(}_~x?*$KCzgq=FP>Acv9P+mL+HYW)rTM)MsaQhvFXRdED}k-7bf8}6%wgl z>HZUSx)PZ&s}4%ebx$bCa8Q2H=<+_zBG?8A=8EOt+=a^;CuSIBSt8v^#a}(}HlJcD z7cp&qn-?WXmlmciv(X%dO=5&Maa@(f+E^pDaL;Kye%GyC%Qv_L*JDP6y(%%2Zwf7M-+9h-cHC~$cAHppkUB}*ZrmSLEtM>RJ}KT;|Y|v z!E~}}=26D4AwWg|Ls9a0g+3Q{OhypLUbK+=%9J9AcqArsKy0NA$-2O?7}?x%>TWnn zmhb{Y)ITzNYYk_Vo)NAg5Z5_Ps6jB4(_u+VH)limw0}8!2)|8V%IXu{w$Lb7h1h90 zEcmyB|4a?0%HLtE`kGbGBv^+h+s9=>(q?!X-nDDUlfqLyyLBZ*ec~@csKv1EzrUi! z-LN2E<2$F^MVRQZ_*eSRwusZxRK((6o+h*VMz%92IJy>&d*=yMa?5b^ia4?N(03gm zqnt2g82=4%gc&WTbRSryL^f>>7gkFgnDmK;_N{-#z%WA_1#U>)hGjG2cf(5#lWWf% zCbtceYi#X`n~%}fp}G-c;*UW%La|eW*M{Lw^W-OIWhf)x3{>6*!do(l+H!60#tZrS zPPyGdMX#kq8!(J+{8B*|I;lqMQ*l-}a&5)l)Cs^P1_c)Sdy+$`mt91zPO8BZZg|25x)ABMta&BcQX*sZdffb*4wMuzYF%SI2%WLc|E{z7lloqX<>8`o?& z16^U$m%r(6NhKC+8(RZ{5YuP9s=d;YHECUox#N(LhV{+EU%5LI?YsA=#Xguecuid@ zZM62*DR>iCFyo{Qq!q2La>7;eCF!nTbNbhhjGZg|9pvVYjt;T}zVOpMo+z@z?ZBmV z!}9^S+{-L2pw(CM;qE=xMt|o@cw!pX`5*x9Kg=NYX&kBTP|>1KODF&J)H6P#zw_?p z9ADLbuD4ut9ZK1c%5k$5TB1=&tcK-%aj}1w*&2-S8i8z$R+O;WJ{z(5hfZQ5LGF&Y zlZUt}N;Rs%zVcZ@sU>jvzQ&H1ddJQ0DqkEL^&e_>>^LdzAS8!VMT((WUR2U6b_~jG zY&77#CUR}ke0dZREVL~)Dg(*G73M0XEUYjpxmizP@zSp*g8%8f=XBov?;4aUau-td zFwhs2=p{XZ4UyqE8Eva6Y? zVhGSfGq0){6ym!5_SaVhqOBIavb_Pg09D9mzif=G?==h+RaH~U=o?@#*2jI6P4;Jh*JBm1jUA`g5dk5iuF46GkqNYa`T zKZ1J+5q*f~%9%}p;bOCVEMW7se@p9!5)Ay%6RSh4ErE(In_G(jd+p!^_O9N|r2C@n zHbo12A7UeS_Hi=nr?7}#Q5?0cesG=j|MGv{>BnBDqY^1y3h0r_ z)ZxxDu*3RTB^LdWEEvxG(FtZa+MMag#P`&ZpoCg5d!ApeUU^oVZ6(*mbpq8z(4~7C zXY$|-B;avfj_&o>pk}XjTjUk z!>%4SS8Ko&dM}j5X1@wq+9E!|L#N`qCU&()3g{1g$MkStU13~L>Cw4H@$vaq*lAfa z$e5R{xx-DZgTri6PU{!@53?f<1+h5`llj`b1Q#J=C%O}6C&ckXB>Eb;-X#;y()>%) zr$x2bYt>|XDlYy?YGg9_BTRNaeBgdqeDtMS`|PvSr~=olD?vFsQ!7!3gYq%S#QQRJOSgiVeB9Ck@a3GY1CvhN#AWoLCvMhlErBi)37`1 zM;k*`H|6E!eW*UWzneN`m&kEv6MqAj$FzIkQ4rp^;Aa&8LnH&9nsF!+gDP~0pFJ%b z-IoTcPU}wcF0e&q%1f>oadEZ$HZA?WhuO5@G3|_}a!Q0C6gE@hYS)=sJt4ceea!-) zyR$9c4a+9F=9?2wOl5R0+P4N#O33(FC$p^AfJMW|*@}0jtyQE+X%V!AhO8ApkJ~H? z&axhpg8-3!_uH}}>Ko9Tl4r`t^iN>|8yFCu_RQ;sQW}fV0J^GG)U0+8b~W(m^O45E zE@C&vv|7Fvsgd+x@>B>WBZ3+mFFpA=m)3CGr`iCB3yNG{XW2Sl9?)$nmH&4%+;~_m zMu$SdXVyN0deM95(A|2f8n;1uM?qcZ1jD9U2y)y-+>rk}a7#{PZ1vIK;p|?m^_}@U=Op`#)NMUq=Q9Fa-p)rm-ivKO&+kq_vz;=VA-$@%Y|TE~)y%>c8Fc z)Y;iaB}aKBlJx^TcN`xrr6J?NcW=#H30NB1QzbnnDWIh0k#{rYf)B7QOCy-v^C zNF2`qz_K<*q<(cmlP$8<6Pf(T-FgLmYj@HN1p^J6D}s6$HA>!KJXFhY7+#;Om}xAq zEwR~HfRPEP(j340xfq&HB%l1B9fr5t7_P&B8A2j%^P8{?svw9(cMZ|19NXQ6Tuxe{YDhX7upDI*f zV4F_2%D4X%#3tvr3q*t;roMm|;MMzZ0EWpk?;U*m&_@kB6sS~XtZSj8xsqTGccb8Z z@T1eEg{sAljnbo%(}+;(wlIA&<|OC7j{))Vj$L{OXc%W?fqKE?*7sYAp|ze$m&t=S z<*L+m9d-C~fDxfYXm2nfz%+h*mJN6dfTAY?2 zB9H(4I{HfmJ{R#zBB>O-RV5lL-uqy0PPXot|^KL0Bv*@f}0P+Fj2{W8bl{#*;T7WN00+9AS_natWW!Yl0_6mgj0{JbNxe=A?UQd72Z`<-DUv5IdrU8I!POPq;%D zIZ{`OaA0%FL}%Oms?o$wb4YwM0JJ&dcl>*Je9OsfmHJfCJ9V~j4%X_P?*DI8MV??~ zFkyIbrQ(QGHS>6*YnkQ*f0lx(e)R@yo0aIaw>DU|+qnY-?-GdCCl2F{A#nowL+XOP zdbJ|}?>tYntfII)cPg$H@Wwa*&7kMRH#q6&q%2%qq{^`3B=212pwhCiX|Yz-T83xs zyDXHP3vfmJE5WU0X1Hp(I+I6pOMyHdom68B3Qz)u?=^o2M?Dn*1w`7n4VLPh>rU#H zsQi}>S!Tqlo#m>3rD7o3K}7sQYx8Mn(X<3#(WK0`<*HXxZ{7^J_X_AV#=DYl&4_A>9M4kGW@${ihO zgVnxbE6H26d`AOi3Ic^a-~A+V`P{FJl$~I#%taP6G)nT=+sbmt(Vvn;-poi4{RRIU0aRJdbK))4@neiD2^e0bdCaPj%{4VSd1~*S$gKx1p{2 zc*;}5)rsE|N-1bw_epMZFcajQYM`|hj6E(Bm!F@oLeJ(n&mexBYTW=lpSCzxYpmf^ zlqPI@!H{SYoA{0!5;YJq-zuIx(ESZ!&L~}?yV2a)e0ktu$n&-qzMrTX>Dq~Ky0j38 z%TdUf-^%=8K!SJdkQt3#P4BNYW2m2&X*@|gS6LXnp8w{ZobOGE%s z@kB8z2BFLggO=xu4!&?mZT*ZP!?S$BV@@8#%>F`h?bey$u{(`$q@+t4@YMbukij=v zP9%ewW$po`GThhC9MP6%SYci6yYsJmfh-ZD)un}OV7fMpQMeJX*=o7d!}id=w&h~h zuH;p&E8`gD?Zsa2zT*jVL=%aKIq_N}&@c(q;&zuqT!P6rv}>GaES63HYQ%)7b#x6_ zhX2PWuqfz%m)rob^Hw)b@!f?G;Rn%XmSeB1>Fy_qS$$$pDfP7Kd>jLg*~;0`<{8Px z+ksE$F>qq>C8~q7=in!NH!~C%Xr9=<3a8@XQ%m>cKS&>pILAU$VS8Wc;AZRQLi(cp zl^`l1KApd9Tns#FuEezld~*dTxkIYC<=$~S1wT=yk+ek$HjV&Y(aS4}8BBE5(KKmx zKfmRS$kLeF-h5xJsBkgl&gczuQyy^MKpbAZ&siPJ2)-0=%f&aC`8d295(b2^n`PbZX&7{iWlj2aaawix}C8G6|I&RU!fGqs^P} z4EKJ*CDvJC;4v{BS@kKvtojglGt0u-dN>bA zB{y3iO3qk|iXtk^gZ$kO{4+j9UcBfGaD11^SEjvxQKvtjp?-L!#<5*y<^kJQ`XzDsHfeOG z+7iy_f4?~VPCA|_<3N7ZW%NiJB`dB!DfCLu#~$>n&{!;&@Df+cD@C{u14;!s!ct$%*H#-97wa}U`ED^t|N zzldQ1T&tvNryaSCGlUF5Hlbbw?C_$0N|6a zmYs$QNx$y))XQs?ys;95L6r-EZE(ido30QDtMp2=n6MLw$F`<4s{rAWJ(xZNq_A%y zATDnrfV3}26X;N3n@24(vO{U^l`ycyXWk`JJ-wA#Xj1!hKD!1}%wkk>4uNP|c?+a_ zl2Q4tSxBHD{?WUQiIpN@gx@YJRWJ*g4(2H(zGZelYFY32)o#Khii)4xu>e4=Kf08c z8gOIP0+zamCH^K;_~N~ErV9hFPZrO7_jh$Vw3$72mYZO_XDQOOA6j*^=HM?V#bPng z)VFQo9-QpmpH$_sc`rb?rwI5|JUOI2{xsvGuOEOhVbEL>*<>?(wjEvRWTSC@LK#C3 z3&fN{=PUGsyVPG3RKx7O10k*NNbuP% zf1LBJ0&~gW2{JHNHnmY6GT!EAPx(qSeuUE^f&kr9F39Xx0Ua-e-;t^HwXT0xmrd8q z^si{30(rM-FXj`vHvur;II6`9Dt=9F8R_3%1q?aY_SKWCyUAW(4hT`P6i~RYG_|zn z2Ai69yv!2KHSY`Z7Wwpo4{|3wOm(FCYo<_zG+__9)M$HGS)aCc3!63=E%2!0*4Qrr zZpESe`NZBZP$8xUDGA><%Znxw)qG9{F)wy~4e)}NS*T(8SnVv;tNyz~(3uvL)B-J` z0xbSAT9C@*)EDM|wg8VHw8WcUZZ{gXH>R%#X4H_hv-V?eCnmbE;qr(N)B!#5LfECx zv|NY#Q27O{ya~u?x18sLbxa(IG;Tg27`;13S@d?XN3YSx&HrBv^ojZUzY}r+>I4qQ z6#}S(;Yf};?vr4NYiX(%+HT{s<3Zo)2JZiUCQtYHe|6|N4zEOaoq9VK1s}(xzT>8( z020UHhow{|e6{Hc7w6A1v>+J!UaBg*rU+qVM8Dj-PFFBGcY!WRr}l9$KjVt#{o4VU zBa8G2rrR;n%tj@VpV7@g`d6c7PU7VT!;wr0zQ&B75#zZM_Ako>+Q1R5Ku@!A5FHw% zM-y!OI7FxGiPNCRJO$6OgzaW!kRY35#*YZ|W-1ChIOV-cB%}RSa1;zVSmWa``0*S& zni{=qpHOYudy&f|p9FsAk556IG%dE$I6b^z|Fe!+wn6TNy$jft5EX6(8I; z2E2ta$tp)+YzT459M2#ZlG8nv^rW<8n8-u&dMz(ERYvS?exKDw4nE2uJcx+&tbF#2 z&E&gp6(ryQx)LzdtPT0PMom7*S`5(Vdofi4%ULWytR2`k?51J7C}(a9vL3(-9;eRg z4?e|dpLx>g?1w2(S0!j0xW z9&lSz=_TnjqlEDcch#wE!h;qxXPx>J3W=&2^fFDj-K#H=arg#lYiXzstc}ZNj}-^) z&s}t2Yg(!qq7NFuqviMfDLX6W&^2g%qsWqAqsu@DqZabs10@%*vKi)opgM@STyJ$v zD#*6;wW4dPrt4#hbGjfo+_erOQNNFIM+iwq(92+92&b`!%E6v{aIj;Z*_qv#OUlLu zoug;_XI%0()yqHeQ^uK4^Oh228U@??uD}XfYvr*r`pBmvak<2|zNFS)Wsq4nu}eQa z9SpH-GMiUsoG2JG?RROF2khd1jrCRh! ztln$wM$z)byI`uR5tJ+ojSx;14Q{cX7jH&aMD;+ztR$cO(yn{4$)b$Yzb?(zCZ1W! zJ3tQb8Yy}Bt3a>x^|CqK^FGN|fX(t&f$Q!s!P3b%N@uT54U+Uy>GBl74OvuJqb~08 zXJE4m=vNOSIf;cbyr6klwKxZDMy;hyfU9JhSYVLl>(R20qIC~yHV|3iE%`ni_%6f$ z`%z+$I*qE%DQnZy^o?@d>D;47?kF+($KA>sa`b#Z_QxiQom4ibOME3=Ubw;xus+oT zmr9-u&k)8vN{MhzvNAzX#Ojv%_-!VU{uf8%i zF)PL6dIx?(?7^R?8UsL7GjQCTR2;c;9z{We=WpB!(BfO`Ej5}N$g+{bWm@$nG{8@u zwuSDWOv=!2-YeLJ*_=%Ex_Hlcj#UZ%xx}3GUA5*VqmuG(VLqB5Ms@+c^Zd7mm9h&N z<8BLlyxHWP<$eBY{47vlI3S&dwoE_FrYSTGkb}~WM1qY zXtleUDHrm5TgzZFE@sDuO5k=%oiRDAeuYq+-YYHQN|tc)B76dBfJbd#Xm7hz+QXCx z4qXRn(Ym4B5EuK+bv|vo9>fbbD$P_xRe!ND;w?%bI0##zfrPKB$$VxJDZ@T9$M-pj z74}3QDGu@$l6qBlZBsi_-Aa{QT`oe*e-i`Kh4hNO5}1Vbrqk@UO=QK*9c3GaO#S)! z5nil4Jd{)BH)KB?U{zO=nwOi27S=(`F17phzB|lQ%(51Cvw?GlmCwdxrUOpY27Gw@ z(OHrlOnf?{WbcojMP;h*e(5*tVOF=D>&dc-^IekOLb-kJW&5Ob2W(Gh-qw$!2H#fk z_mW6JzfSeZ5yp_KP_eO?M;6H)%=Z2ptdQ$9eV5Y`ftGx|&vVyivMq)!a#vUt3v?Q? zX}V~nu^hc`n`XXN_B4w6hR@HYsiB=8X}MJ|ijAt)e=}%!?2w305zR2eBnc-r$@P`g z_eUywui2K@OrO=gEh|DsuuHzsf6;^*lx-RrYT#^Ae2L8SW}dyjSS~W_YZI`}4qwhD zs3swLg=%UEXUL}EYcKFZFCHy2O11zk(MFwcq0Yie-|8Fo~+oc4Jl z1~sf^4n%qXv~$~5<%5vFOpM74CsOn_AQj-5 zmEOz2uVi1YIC`Ip)HeiH6cjFs_1zKU;Qy(xbd~k~uS8>6nrG2r3av~uIUlJ$@cV#q zHtX%m2qnYl_ve@{OSZ*Gnsl#}3oyzqX$A(Ht1L#>@o@~UU^N#L+@4rx{b1z{xYsRK zW$VkTVtG$9RfUcjjSOd{nS1&3sTkOPuZ7>5mHL`jymqWaaSRLPE(UA{);Hff`78^D zMAuTf&NInl3Me=m(7`2%;j~aO{lj!TAC05tBg?r@)sNyv*uQ{AD`A6e8;dpJ|F-n9 zRIzHua-s&}MTvedzU#wyhMOX$r$p@N%r~XH)O-!iYQd_~EYe~Wq@%)7nGSjbd8?N_ z^b4oHm`BNINhNYc|BLPkx_JqE^gLdHThn+*R3I~9pemAk(z9@m-wf%mIuLP{a5v={=a8nH;&()yUBe!`JKfo4)&RV^CiCMuj zk*B+)G7o8H&6;JsQ2S>S#LjnhEd-A+z_QH=3|7>`V&(|B+Yv)5Sd81Usori9b{(0! z=uLH@s*2Cv`ym(n_a{?*w3Xx2-y1Xj!>~HjskLqAbu6CddlZ#SK_g8NNI2ezj^TI1 z9NP^_#@Cy%rjU{1xd&?MWP@fla#b9ke=yTMqdqHWQks1q%?^nG`J-5M!)-}mjGC3M zoZd>xugDqJ>pf)EDpwfgKrcTM7f#izevvX|f(A_v_hhQ!F!YqLbAT{0TmFSJ*cf z7vZVZnUQb8S9_*o@xOZSVeV2cU_Y#03l~4}8BQFyzSu7qwsA2zxcMKh~%A-#FbJS(_Ae$K_HKgbW z>tczB`H=jPZ>jra8sK{?pOn}4c;Cbd6|RZtp31b4rvFOF)5|GsL}?}0Lv9U-zvD7Jx{OIW@7w2v#3 zZ)`~}EmCf#BVrZtTN_}`;R9adtKFG|EFj$Q=nY_4b@nye>{J`8)6-cICmkA63WBrV%z21IIpU3@}=*rSc=bKOPr|6(wTV*8Ziqgt>|Di zCK}q!gLFkMpY%W~)=0*{`QX8037utu1X>=l!^rG!28o@n%8a>?Hc^i5l-u1tS7(}| zSD&b`QU%j*rCX zRs)Fll*7<4#ugat=_Gq(0D{>`L+rzf2{s^v*!V+q-k_m#QKfC@_fzd+^@huOg)g}M zYn?6pYX|1@y_#~@BmHlbNdkiT_(|=p+9mHBZmgmli$Vc2y=ty+n?J8}<^HQETX;i7 zgGlo%{Rqo|v|G_^t3V+qJE)dQhvx|u+9-L*?1X(lMl*DqFbD8L4S{BfZokfkiE+%B zHH9)YRlf(7f?mR-yVIrXQ8pQq4pm%*7I%ARU(vBiyr`h-|7?9p`rCY$Na;Pc_TK4N z2qNCFB%E9ufLT3>e^=4YUJRb+ledOW&ChWNQUGhDbAisIcuIj<{0F=SI+Dx%A#BS^ zhVZ2qmtj=JKxO@8e2xyfY~oSpiJfgz-=TUxkzAwtRb6v3{>Xh&VWj56>GmMV-v6a z;YwA``g5^YcI?mN??sg2rviX317~iN!eD~YU08}yT1a+n$@0wHY2i4EGiBOk5Hl!N z+iX4CmE@+W#~x`p9Br8OIHT}ys9uY+%0k%Y;l$gpcp?4fH$QOY!k4sa!|fql`>PI} zNO+{f$yy6(kLkvVC#cZjE}yIL2FYX8x1O1q&%U^T#wD0vj;MhlU_|06%3yssNWX!6 zThr@)SCRyO+t9N8Qa>0SmzrTzI>g3kKAwb zc(~K`7e41V(A(jbx;_43vfpUS5kwPz=}iKj2g}g`>47@CKFm32!ZgFf4v(~x>MtCOPi1Ek5jKKsr`>y z=FFKc=eNII|520${IJx+8eABV*Q0|L=u_VBo5l_E-WHmCzc6@md2N1 zsS=J-Uql}h^>{c%PHNQX#jYGkUriL{uxs+3_2qSDrm4|BX`p=|UlB>4vbAHmsd+O* ztmDqnmwITcyVRcmU>=OD#g#uSJi#fH21GSffi8Frlre+^rE5_$_+#bbZvX{D1@ zoR(`SE%z+nSH9#}Ne=#l+e3RBsO;)_iU(p-FPi36JD8FDyc{NfYcfT`umQGQniNLX zbkL+yM}_X7I;PlMtoj~jzaO{uCrOUGy|n-6p!TQ_?oA&wA4it>a@GVn1jGx*LFyZG zx?OV|tQGlw$_?$0sWi=ZQR8^*wOLsskMXmvSq^t6VK)4=+RremF}8KVfAXrm-x3=#9|@% zRY0B>+vOXCW#4v?pwdikk~Y4);p`yN<)d;iknEvwku>4KPLVQ(Az92jZ6UYff-7Bty!K(3$=jbQ=-PK+~4WDnxq(?(8UgE zuzmGdN_cbknPJ`BH>vdclVe>0#(er%HYqT3X-ea{BWZ$*P@!j?WN1glRnOZmP zP?T{ux+SefFSi?PbiKCl!_Kha)5_H zyg3a|@qjcW9W_0AdV5p!8D!RE?ek@~F;wIrnE+}}^Beve9eJRYMR znh91FS^gN;pQ8}eMm@IVn`WKU)C1O^o~6!%=?adt+wCl@@wq^8yQQWyM35gMIQu2! zihk<37yMqi*BrKYLLvslz4re|6I=sHqHo6rYDVQl*EN`E+ zxY119c%PX2|I{JDj*=XUmeH!tmsKBuBjU-4)C6Rm={l0H;;G(f1o>&H8ieLaw4I{L zhm@bmI(HI3F#7Oue`sc8i=+3tnEVP9BWuIi&+Gg6=NBz^V`3C@cwH*DcbbpZ1Z_g2 z6K`|3#S9FAp3pkS`pU;N<3w>r+FR})2h>MZ7Z3u2rk)w@D|Ecz!sNfHz9f|RHulle z4u(W&OEyi;mFwRVVW+hYC-I&};pMI8xOrM=0oGRT;B6O2HZ1Sz-_gVWqXj6EU+B+Z zi?r++V)6gV8tFl{yssVI(=jSCK|Dm~oKw@Gho8(2uJ&eNRjp#Tu&#quHC{L)79DY~ zbpxAP?aj|?9X30N)!4vBx|Nmj|khDdM?NH+Ka+`u=PJ} z6Sc3{fpGw=v@EYrh3OQO%t4>W94_tKbJ znKlqqy-P1a6j8 zx^zh$!BZQ*ZSIqQNeolB^Z1M(IN{;-8zL1!Wi_;#E>-?h_Hx^KoCx$Iz(VUab4f7K z{F5Ym@+8T#*_f1?jDQ_JYxSA`nzf)b*8PENFU-7+)krCvy(%7A;Kbr zA~bVvfBKSWn`Y;WR40EPTIkwx)iDLV97j0AL{@`Fmt~ftiIECbhNMyT<&L(tl-EL& ze}-KA_3sZL8N#=!j;C-kfO|4tCO*&vK*m0UH)trQiZSW0>z%q{@#m)TO%R`f#2YP< zFeI)#a0Z4AG?8)bhdZ&Fk?UyuAszh+Nk@q1W%g1&7x{jc46^Bn zDM3W-6OPWP_v#hn=uhOw&{VM9oR@m$iCD1LXT`T!_I&Nup1dQel$Lx8+s)smnJ8xU zmv#C(_+41pgSlYmI%XlG-;j@Ya;O&uwihi}&l!UyURv8;N-@&xs5WA@N~ch---^Eg z>}#J;l%We}9adlpZJ~64$q=yR<&Cz>twXvrA6h3_49eVx;3FhwKXU0G$fZeM8Y?x2 z^!v*+$v95emL(IwJP3RbW4stQcBaoAC1+vOK;c^$1i0aqoFsHVb8!={WijWVy-{Ku7=a#|l-Vx*rIh(Um>fKO}RBqFh@_7-; zW~Da-*=l0t6Wqs9P+Ojlo>wFe;venN@BEnWvT1)5>(?#Z7K3RHHxo-xOO<}uR-MWe z%8fWqe8{1URcz&w)}#F5lc{!4*o97LRuyE31DDeezK@Da9F%!I(|@DpUBx)Mv*srV zp}Bq623%ufKyg%Abe|*t#`n(XE?35ZH~*Fk`}sGOTLotSENITvjK}fwr4;j4@V!zF zWf^ip@nfL3>Mr#Ua1(4$-Zu!L@TF@!U-Tk55GcLN}-1_SV zv0*%deJZC1(3U?jle6#o9e!3@9-c^>HBfwKH!XCd zo^B%>M%?q>vHbheJUS1*i_l1tedH|gW$U(VBt3hiUas|ls;!h?MX@+XTz!UYXr|+x znE5UY$}=*l`JkM=Z7r>!rGb(lim4GYYrK;`uS!)_A>24EpjYx`^`5&62wswTGI>Va zLTcN|W@Fl=SlbmTgmAa2DLzhMtV}R>vPNFMzM^)ur@&J=rqFcia+JlGUYu?wn24qv z@=r>Ma<^$7-1v?IoX%-2GV3L^5}ic}hnb`PSNEiQ{1BTN;tclnT!{Cj5q;L6KHe`Z z4V}8AZvJUvW`+Q<1*3hM4twqtx5Cre$x>)?kSJql<~kcoKGJ1mzu4f6U+?&qqf&Rd z)#i5kX~ucg z9B-QBFp?LK-kg}VCg@USy>4qRKT?p{>ji`7VKbMlP=Cp9E5W?c5C6r^t4<=o z)%+zNiZ_t;^MY*sX26-B9O>sw{dqWH@^zyrViwZ1%V4%KX@m5{i|h*bYJ3KB;j~ek z$Msts>N^n&O8yZ(bT!(s3pa!~|K4!gZ|?a%(Ah5k2bMDI zH4;1<0=^{m>yhO#Kj!oNyIix}dKUKB|3omklmVvmaA-Ew-yW$2!*WYIt+t`jSB)Nu z?a9HuCnzfkqZ{G$qb19{wmWS8D-;Y4LDlslmUG|;RSCn8co%4)J!@J`9Ze-1-ZQXL z^(5JLoJK1aZd{{ZIA0EbPX~1_5~g_l!>qlhrhY+9;)SdM7*-LwUN!nU<(HSO(+U}* zQ|1nJQ5S&|HY}l{Lyll&E2X8vmza5tH8Mb@5AYI4$)0O<&N4kFIqQ-bM9aX8_iq5^ z#;swWy_V`(a*IhjB%Vj{NYkIXRUF*Ij@l2lTWfVDn=6lvU1~HW(`b;Hfc8U0_}DJ- zZ%1T3jd(*1hWhd*BDgx&rtN-xXsqbdox*9eexCEFA3d)!_wiHXB>M5UDltl-+@unJ z&31tr!%6nu2HJ;5-@?Vl_Esi)Up!j={%WeF>}>0Sd_aJa29F$0e=;_d3ikW&tD^7z z21gD7&B-OSma|HallzV&R_UN(JT8~g@V_SCW`25mCp43(RcdcsW}`9|k@f~pi3((^ z8KCR9A4Vh|mT=l8x_W(cO&oPuBdOxuV7T| z$STQc=kI5pK``mb$y%+!!2PGMua(kQc_K`Jb-1zDsF+~kM_>|a$jLX6u1>v0A@QHr z|FPE>q0mg&JouL5r0lF?;!pkSQ7tzQeC^XmDpbov6myTsmJO>*Q`7qD%FdTwoHj9V zBF$f&mw`bi&RsC#!h0cM>6gfNRJ+&F&VB!*CUU`&`~|<;Zgz z{<6Apiwr*m@XD=pnLCt-t1W_JYR3@#+n}1y{_ZFj0$|^+ahMBWxYY4BTdg8#^>aa*o?CPUn^-SL4 z7c+jXZ=)KP8KaJ}cSAFqwzgiAsOV-IfwYN3D=~omWoSt__yMmSRh+cfYMk==eCJwy zcw|_|Z(Wztv<{Fkm%a6|hc7sVR(Z&a6KI9cZu#8@MXn@4!qh2H8u+cu>8Apj$Ltt|#!Eon;6c0Bo-OK}&iSzQbk8ELZf&BnCI1zZirtGDrZyLYv`Ye@e;4)vcyJiT<#tyhU$<5Rr=lGRBf za1b;`T7y5-rWd*__t*IDu+#=x!eYFK{@$e?`xWx}jk$D$EjcKpTIw6bd!uImWwG== zg7@F0ea{?V&lguWd`r>k7s{O>ayRO_=r2l|uFbg$XQg0WdKD=?g|owAd|i%NuTD}1 zUc^m-14Xmz{OzyLzX#qlC(K^>IzOkT!Hzr0zzpDVp@u0&Yw zOwd*k5t9NX37=zl^t`I))q%*Ek`~z~ZKu?g={;`om=^f=v-|Z_mb9PPA}fVR#lAfo z`=<6k{lD+T=b3;hg3On-K18b&sOi4LbtY2L75m7k7PIjA3vE}kxRz~Y*P30;wwR6C zcA>yR@rhEk+CFZ;aTV_Xkm%m#D<^o?<+^o-Cr)>gJRgr(qCckL6-7A?NxVJQ>9B<6 zvPPz+ocEsa+~)yDt)MZKJn?x8m)Yvu`I{*Gy@U3tq3kROk0EVfat+%BbWKSI7EITH zOD>ESN;X*~`|K+xObO{)dK?r2zq?HTu~KNJE6?YWSz#_B>f zG5(WBGbMaleMVKj;{BrU37K2}cqyoA7wi~&T>mA+-BR4q(U<>bL>z1;9r&nh%4#h% zQ-5KLl=g&ol9Tc01iiJhJr zpHh7{+V_?1L+?9@pn2LR9$8`xBdXKwKvJm zLX+QqX%JRSN5Xj1Y5WG@H9g3THK7ehc<;_>H!DIe>)kkKbU`)_{vLvMay>GZdG)h?;I2q`h9!@ai7bSJ(QrKRY+`L5@< z7uGDS!2u8kvOv)wJV0&*TLEVuI(p_2-uY?X7WmXlQo8$;=g%Pf_SI)OGTm7&qc0|C zo^7`ml-_q?2uLNpq=nYs@NinW4)za8!Yh)6w9wh#ISza#lA0r#B_gcBjbK4%j5Bc3 z(n7FY{`PD<=U7baogJn?A34@Lk#Ek!#`mv=m0mhcv<=7Jra7J*zk~Z23G2EFPSstz*Pr|+RHDwuFHz2d5lwf zZN(|4@S_-vk9y&c@hF*?%jM%PWnt z5_8}{n5pmH=&EUVO^jn*?XPyJ+ry5zT$j}@dg>hkEp? zsGXx$6$?qT)Msq!prXAVThfAIH1MnrVUE>C@Ok(9rL8%}R2CK+%NiQP6bU~D&; zF~50%J<@)>$hXU4Vf_KeA+cx{N?H^_xb2W+0v|Zt3tlb(D$lyT0cmVeEx(womnXhX zTfJRt)HSPIcHh8px|DFKTh%%}ety)2ST$kx8FE5g%;b^%&np?2atGX^WUkuNsQ_8i zsBAmcjl26FzjiUSl)eZ&if9m=WSU;Ay8lq^B+8gl8>r=OVn+xT*ohN{+-8z|wWvjB z`-C@_wMa zIu+z;J*j8!Jd;H7H=2gHry=}5wOJE>on`O*GOxNMOZmfRM8mpt zCj6xyH%`Ow!>s!qe^#$>_v8PdWKmO4B0&~-Ak0~+K& zq}+dQa)1Z8QNwpnlv5vlx5yoE(^mNA(vr$DU8B;lpSUj{2fFu-rcwz9XSpDGd$5FB zU83jQk7MF9_}zmsd6brY(+jEYC`A)>Xid40bS|XiEmf=y(qI0@1AHQDxTIwV-^(9) zy=n{Mk3u@h|NjjS!} zVuGU6O#QB*j>LJCUetI#7TpRGbCUdEHw*PTJ5hX+fgppEF0TGaH5J5^pLK;D7=@y2e1yLBJrIGIL z8bZ1|hmd9n$)Wje^xS*T`P_5v_xJq4ix{3~KWp#3*4iuH@5f!QEjRpakJ0BWY`I!L zI?bA-LkrX5Va%>f%T7N^`Ln8Qb4|O9Mn9Hx!MI!-E%6zA5Be^kfLQa%WC zZ|$0{Vv!yG!Ms@$&&Qm(E+cgoJL6Gnfv9>t7}t23Kjl_pHpE9hVJ91+&_M$L9R#y4 zN)}5`E5DSa%V(&u8h)%f-Jd{8yx6sP-ZQ4KGL=VI=N+F_Gyseo5SkH4iHGS3?bd<|x&|dB& z8^s~@4g@Xt(Fn$>!eJIr^&@zGvSZRLNL@qma?xqJW#F%__*1zrDL%C4f!(0rQKPDi zzuYaf7Fnw#z+b0lzglGiVkKc5?oTRnOa_cb*=1i}4HxF@l?vZ(VP{$(^f-&@$z^g% zVy8Y_NRy2>VQ>>}1pG5JW>DGO%;jp3$28*uT{2KYVNM4Z?;c7R#_kex!eijZ@gg*x^?w45%Sex>$*|!jeo=p+V>M(g zCd&o+b_Y8r$-bmZtc8Ugj{gd`ve!w@kVi4!ul-(#H#g~_Im7aIqfdPmD!Q57HE@x7 z>Zn}3ph$81pmb_|y8#=vFX^%ANJ(5>ypb#{o#L20mN4}t6~*w|5Fcf^lG&fEZ~$qQ zlVp&%cLM~wa*GKfpEqkv0mIRXbaEANwlmcE3~Kp|n1aBDDXom-9xB(sb>R;fZy|S|%OUXU{0* z*Hl%j1JSJw$9-JI{@u~JqLIyjwy`J|uLKRa1FCE(YPm>8li4!liec-T6C*tY6kO$n z?=&wOU=5Ud!%8y@l!f-P_?W|3W2BhNY#x)t2d7cL5Rz&kf>YvwvARjk`j45|UA`Mf z@0e$IF)mXBg?Yt|nnUhjsqHZPNZWCO6J)su$aLToTOpBbL(Cc_CaC82bHQUQ=+=U2 zwqq~hqxWSx=YiTjAuX!{8;cT~hips&$~$iup?XfJTe}5q<6g}p1$pV`@tm%!5Y$vs z8jMZXodEYJ>ADC)83fVlzaAl&DL&-JUeY_E8qR?d@~= zI?Z*|jaG$T1JIzg1%G8g;x~j1Njo{bd~A7Ui~Rb~pKzR#b(^9JwE;zPxp4W5>7_e> z92fzqMQMAPYNJ(XK)R*n?)-pCwVgU3cJVr>z>P-Tx{S7-rVjnk8lQ9#7VzSvDL};9 zg?HrM6zXw}1PAKz97koxr7CGF=IQI*S%sc$Ys1{e+~(8oj1HO0JL6oA#@<^M^uXy~ z!*U|o^R;zqgKMY+yxK#*?WeMOPLB-qa9A6D0#rj9S;OX@Q0H^q&O^V6{iX=AVoz1= zC7IhJ*e&U$>m6jm&4N{4`@>UV3g&VxgG}YOgaJ*e0cMusogCZG_Dc2}STPRnoP_Btyk&jtI1-^A11XVv@6!8>-C(MXb`(Rv|Tgg<(~>Lf4yH zD~I(tDi1y-=AP=f>3v8z<&+ws9>k$X1-HKk-Mf_zBgbBZ>!9J&CfF~>?d2_`DDb{w zA137>uK2i|J%3s1HoBD-uNQa){RH^}#7twebey(QRJFzsxVf6+diw!G1_v1~DD&JL z%2k-sl}|SfEc0>8l&#UWjuxWk%41czjJ94VP3qj!bimW_KuG8+f;K~Q({Hf0IJl-S z)4mseT;-6kKbf(^-;M7yf<#w5UW^05W45@{g~L>qD?RNsG1f|`62fLCN3Sn2uK%h9 zz!~vg|JZQ>>RiiJ$>?U>h){D zMvj(~+f?0E8uL1-&WA`D3>6pk2u;Wd9qJny_aEcJ-Y@yly)`dY&QNqf+K*HijN`!G zACMO_)iCziJpX*F1VDRL+V3gcg4v1pPdf`II1*0DkaWjnqm4^UN8k|vc*r@o)0ymB zVfEgyro(_Ncgp_GXSTgFfYA7->keb)}knfKP-5+SBk94gf12`5so$X9dJ;r<OCB8-OC~R=WZ6CiM4{M0Xau<`%=n>I*X& zvG3AS`x+nuE_Eh6tuo?-*g^Do!TV{atL3ElnWf1yw2;U8Qy+)D zdR+~6UnYDg{dUyvP#j>D@UbT`rSM&CY1&agA3&0~T9?NhpJkcJIvd;{FEL#zeDWN1 zH^y3GVvzTzx~q%3-+1tTMjc7XX8wwt8M1N+k~#R=&Rh&JMbk}^@-z%^0Diw@*U*tp ztt>ty3fk7R)Rl2wXOayS!IhS}aN?=)V=smE#8V7KuF9=TdkC)P4R!B5m=;?Uh%~TU zW4md%Yp`wAOc(jCm<9~wDW7m<>fqvJjXsjkk$@pz+&~i8seE{M&QZMBNbY!8`@9#- z&Lxw(L`gg1o#|>LhxQ^uY(Yg+;W(_XW27ux)rs^;vZ>nMT93(SLaM{UM0KZVCA?Zh z!CbtVKXR>@i7CxMnI;pOgJP#!aDZbs1_3(214(G6lu{MyY+Pl|Ln5d76oT5&jhrwY zERG7~$_K106*~_var)5mke>-WzTPc0oAsjk5)KYN`PbQ!f^)ZmJuT?zyKc`}OJRY$$`$T>X{;r&Dc4$5rC)8KT;>>(ZQ$rG#wt%DYc&tyk{-fILcC zP#TVNmgkTu>xL|mehefkIrRyMlPO z#2)Z!56$#y^JBD7Lu$}i(URKX;F`7q0b%p=cH{kR#sb}CP=g??mZsMdPy`asGG%A2 zy7&tbhlYd3RoT6-R>B#aWv@lw3bjTJ$Nv}rjK|?wanU~|4^`|SFyOAf9c4N~KW zn=!M`#@LBt@d5${;ZXEh@0neL`n9sVwEX~2gjk@_?CR0+5^?T8!47}!X$J;L$B$PJ z-W==P^l#eHR;x1@k#>lEt&#tcouK^-37704%B9&qQUxKBl#BvEq0Fv29&0N-I!^0* zjkT7mhh2GDGMas7TIR*jpE64k73ecxt!RR>OJSm8ffN~G|lD+dCWnW6g06`xKh^=e!j z9tr(S_yr_lC1LCV+Q>OD0biu8-QnzTsmhV4&sA=dKR71|ZGAhWxusAx>U&`Qrp_ta zybS5kTY1<>bEaE10CxwQdcBDi)MoP}4?Ku+dkLdLuZnrTX4eafrmyWzZ!Px-ynk#?Jme z$)q@YwFsak^#zCwxzVkjH*{p3t9&{7nI}mb0@whm347 z1zPRh+XFke~}3t|2!Jmr?mOgpE5)`blDcE@+BQxBD_awV~n z=%1D(ZzVN<5Z%M{h^x`|NvyC%*u$oB@tZ@pSRJioO(<41!w4kRH*4|Z8o?1!ot2P$ zmb=PSY%EvkAV09{dUMJ+oI5ss`V~cAvDjSUJ{7+@D6+F#ckKskE>Ctl&w~o>YKA)v z9`T=bx>!td92E8ZKnMIRNp^e=fSP1t$S291u_qvF?YaA(tT~SRcA|2ju_)0R%XdyW zX{@q5jc<0HY<9o04j;xc=ei9Z@b}puE~6b{Z3VgUL0m3?s5vlxISvQ81N;8;=0*!_ z(QFS^DQ_%Tqc=EuyO0ox7<90M-ygXZEQxlB)e-g?;2jiH>3I}IL#4NxwviN-2fOb) zJ<~$McV5}E>cpG0M89V;4gkQG8>uZ3r=?Kt#6kc;lX#{$F#?q`c-enUB6y5aJE?V{ zIymEUnI5W=F)FP>zczZ$CbJ0EVBs%lawtV`A+dJgOsaA ziOp$Fd*X`gNx;|zJ;r}7n@?t}l&9q)I)*~JJfEUouz3-Fg{3dj);qZB>U%S_tsirZfJB#AB4OyQ~w2%qi|x?mG9 zR0Sow1*2%Tao@@67hlDhp&}UQEAUxmRCkD()LML^PNbV>uFsV(bbHz1ah&ZpjHfV& z8YqrHWIAQyuxNkA?Nw-g1PKZKMo&vCTX@tiHQ>4Lm~4KJe?lZNVO6x|3Ne*lqJ42^ z!Kg}L7`{Z7irW&K-2MIreFTe^+2@GI_G6g(TETv|!FhK;U<_7AJn~)U8 z%IxHG{vTV8mPzA@%`gvX*@P$5RkM;h>w7vz z)ZDI~HHRutC_p5Y*zie$uT`=ENxLj?PxM zjhS7J!NVH|RVzpie2I!nr?#&#m5*A=limDj)@Y6;2JAxVF;-uWbkVy z{}OE#IQMcN#$(Xg>D2Sd);f8j=aZhm=#lRP^~t7Qv%j^@SXjo4ZtHwn-s~gP=r{eE z_VnvTs|SUJFvG=as%$FCcBi(^V$sO>zO%rTnF>%Dbsqq%v|ww~RPI+E@9(dxTY6q7 z>}zNxUuW6-RcIO$#Mj8PCCSYQU(K_|3OZu5n4)!_s$s#H{rB|fWiRvNm+emX=RS4j zy3tRxc5YF3yWTl&mht7^=vc0%l8fe5J67#I^*dxKtO8L9!_JtnSyz&m(;Ms8Hw1h6 zH|?azpSs;2;iW{@3S=DzG@9TP$4_#m3VywAAP_U8hiEis^tIg_EHb@t*r!LpQIMM{ z3IA<-WwwDl=pd#Lbhc|)Fj@M~egxlriEafcIQ-|h{m5#A&qJ=+anTM(yb)uYXSTVg zXWx~v5oGl5bE3I`*nO59Cf2syLadNmbHzLDRr7&vz1eB>qAeTS0)V98Y-JO?( zlLD9B>3%zWTbIW(n_QF%?|4S&vhSqf_Iu5aTz6P*jgBh9_5H)TPL`|!?e=MTKp3!D zdX4yDrO8pk$ElG6{J1k@7kWCc2sC zz#jSpkUk2Cx5?>HvM?DMX}g3CqCD{ z9>wwbPjy|KJaNgG*TRR6qtt}kcK!Hq@+ANv z4wr8c=cr}iT4?(1?iTg5jMRhjy`L?~@d^Wd0Fl1@-K^?o8FElVT|_PD%rnrp+?Qky z8i2Sg0baTwv{V)V02*K`xqrr&Xp@n`7#p(GWmY^?v{I6RoG&YlIzyHB)Br&heDVSlMh!I(F+xznhy>8<0?`50jOm&CgswfVsI zN?=^`co!LOxueq-8aEler{E#_UZxkS;uGgTaiU6po z1|XWt0YLr3RgRWu1LOsmYF-d=ITkO{=-)3H2*L^KS*4ch?S4iNPi}M~35Sb8TCURQ zhQBz*PgAi}+L{ol{rbmJ5AO?>0UdPwF)re-8rCM55@fV6^tbjt2JO2HsIGODjccN~ z)_=Sk*f%#FT4X|Oa>qV6$z91iJ(#{n*P0pcv916vI#KCix~>cy#`^^7aFg2 zD(bjFrb1UEcNyag^w`*A zcG(JF;N>I3PGV_sbR3?(G>&(hUs490ifXi{Cn!|u^T=#+kryAFXmhIwU zb$Lde5r(&iurR))#{2Jw`KA8lEiot=^}*iur5V-)A#Z!1BtoGzQvgkDH-KIn(*r~o zj+mtrr8F7n=C4mK2=X7Uy;!U4=N{XI6!nL$7<*liF>OCmH|GgHe1E)O zZOPNmx-lW)bHUsCUso$ES%2>eY{o3U_B3WT-FS<_44fG)Gz_r4FZRW4TC?;^KneN+2=vc8HTHpAK;Y;d@4x4OD)cgA2b{Qv5#hnxsz!XP zBj+KsXP)QvU)j-`jEdQz+AfDf`wqu)7Pl|%>>g9qa*>r?n>_s7@$su_mpC66B&#s< z_gwwg=%1VqS1i;^ z5*oD!tcfZKFRhLEDtFu80-zcvMn}=-=pwPm=xqI!=op3cwY{+H-Mu6AKIyQmnZ=4j z0Ew0cJ#uQkRzglyWjBbKK(AQ=+u3{njYn?ZfF}d6DM| zoo4%{d_m~BW8vw7_ld>S$`VPwTfB2^%E_$qi^Fw~!?Xp&pj?kmot(g2gKiD+WN~OH zLT%_Bdd%A|r0B6bcUJwy=$yjV*eI_V@NKJmw3>xh|MQ##6Z4N%aT{Vqaa($@W0x3@ zHRsZ6G(f%p-p;odlTtAEa1`{TXG43YR5G=mA0rXng@#?4<6jHrkbHBVhanH;OZp8u zCe)X-4IPo19pEHm3|h4c0pVD=`~0L??Z1Wt3MAGqO066=CY|u3{ND<@9WIb@tAO3v z^7kxO>Qd5j8=Xcp05!CTor2SNx7d8Dut|0Yvk-vLl<;p>j&#wS?X@28xt`4t#1*r5 zS50=;q$kKm@~wYmo&3F-r5z~XGc)#q|FGVJDcLCxF-EtDACAk0W%7ykwZ!9FLGZi@ zziJoiN50Clj*mXy`)JvZ-QE3_cjD7)rDnM{M?X;hx$rf^`1fM0x3b$sfmGOkO*;66 z{sUgn2@v^-&*75+_t9p>hqYsS)FcOX$C3M3(~El$_%n|p*T_nMpllQejT|c9e8Wab z)JXsg{yeqs?DF!E;6JQu>5EorFF;2buWMjYKBjnPlgVjz;%`0% z!8-6x#zCn_px>uoOV%k2bc>mj!{LfZy}ze(%DyjwFCkRb2RdeR-Y)RRL>lj{ZccL0 z&__cW72a1U8njPj80ox43;b+lCF+eC@=1D?rE+mEcP;zC9ng|$(PR}m@IU0uNeSwo zw2kh++np~e1)Sg9OtG`ZTcB1|wNc^<>LC}U;M9y1zn~jt)lC0AE`!xwEuB7~9W|=< zp+-D_VTvw?9MFU{!7Ko0Og+rSkDdC%ChW`U9Nxuj{P*=G3>-{aacbcd!hO=AIWzuZ z@h*J54x3Z4hP*USs?5f4tM=eklQy&nNGykyLwAhj`o-}S-Lh2V7m_-E^r`zD@SZh0 zBJ!-7FH~L2LgpBv&UembflRkP?fUeqW-6M@1@}b3c$WtrmH{cxy zT$TGzO|O7wdu~Xv)iG-ne|N>-KT&iRQK5i;U+gM)EO9}82QY>q zsTU@}ME};*jmZfBdGvM~R_3!NE@9&0@sC@Clahg+i`^$tRC`1134?*}Y^(WZ3qiMC zvWx^96~jc3X8_T&tKZfJfl4;tS!80Ki1W__*{IU>3CHA$C6Vp`mS5flpy?9aUqf2* zyN#J5)}GO$jYl?MS}}hS8ses6x@XMYDtw7&nA3n3@skn1Mxj2#v#WPQ*MrAiq2mBJ zi4Pma8!o=v|DJItU)b^6xuZB8=AzIE@i9Mn*}_mC;v0OG?>PDtsi{o~OJgCXX~#}y zIDdXrD++i>cMAus?(eUS&G}29hI5KGYc=( z^A=Rqv+2{VVd2{c@3RKYqQYQKkDf`q8bPpibz8EH#kn|IWXHb!>-K8h`SrAkUy2Sf zweRXTL_YjWN}-HT#n^+-91G@t{GOe`&1EBh}mW;lR91XLO8ig@36qZ`)A_ z;WsU%n`=;;haE9y6>X6`xuQAoZICtoq;+{A|I-KKj^f9WG4&2~Mpl%Nb;hJhnJVN? zv`j(h5pj;3FKH(d6#(6o$ViIBVmAg8@a6g(NWgXVW}c*8TPah3P>+V!gWvSDKmsq{X}Y`MIUb_q1Ky2xDh(3j z`jZNgSh09xv{G356X=?RV16Pf@EcjEIGW_@qHZTvj< z%ipv1#QmQ;Ts_qQ-p?<6JrC5*kEYLaZkPTdmp^`>yavv>SAVIMFWPf>5P0}w^|!#! zF99w#&iKV5`Rj}R>&MufnSFJy>V9o9a`-<_P+qT9sFFwi1uscJ zJWt!v{_$ujppDC7=l<82|MAtAzdoVN(L{?17~7BJBDrjb&A@??#eC~Wibd2dzMCgU z;IveB;hHEaDo$?x{}5Ct9)aDfE1v(V1(1Lp!6*qew9n@PRdS!bGQwRQC__YFkO!7q z(`Sl_%4`m^2eG;w^ z#21ETsia*q);>kv2IP3bAWpz=z}pv|rxw{=yrZ4p`E!+&U6km%w4P zQ3lX0NsK1dE6pl^hu8)9BP{`3Sy{a68E-76Yv_M|;AC#(cj+82@XjvA*HSwCgZ0)vZrdChpNP zy6T*z4V{R1cW=DNF^_#qKsgH@y<`t9ftpVq_yOwRjraMc)2^q;>TjZm!a$)_G93rd z57Yg9=Y5UO#^yqYYWafO(TWkk&QFORKFC-y2ie)l-KNhGYF*>@18ed}y8|NDwu4=O zFEDj)xo?7*Me$yj4Hds5kl@=ZcbSBR=#6rGQOOfi(Uv8S23K*`nQ`||0k}Yl=6=>J z?Iar;E-2U>vjEs+96Kn%zR}@q_2_2z55M&H|4>rj^t!F)@t-yNROt={3PWZ(*#)7n z8-)7W)%t$1&KqGF;>p*!Ooz=tG+3h3c&BW~?8rZI%K!(Taz1rhZA%5Pmw5sZsDTRv zyVH}Mo&tOcg0~R3O7=$xd`cb(y1WR?LEvF)U0`}Qtb9vq%qvtXedR^oXu3kuoMD^< zb&jR#EryM7Q)FsRhXse#j4tELWex5eopX7l4Q@3R$NP`5Zev6HEAAo|6&DSi%oS{E z=xHO(^O>)LDA$8}(4D~<8nq~(*uQcM00MAAz@*1X&x)P&)|0?CWRpCtbWUcdBn@Mki$zf&K}>Ev){x6u~t?Ao6{vI=CL zUAj^+GX&N(|4vXUz{Mh+$iHidtEiFya2l)!CQ3Gqg-=gxGvBNzaj%yvF*`hr&tEcZ zYJWAHzcpZSR%tIhn%t(e=5LFu*49sI*4A%e0kM4OoG3DIw+X*Ty!x3_`0#_%{)Y%{b%3t0a2Ad!X)s)j zL+$(>@>xJgvmZ{yzgB~;sxUZmJ27!N10_H?3UfxDD~KXq+r?CX+jHS|yU)^BBtA3; z-$}akpdV_#HvBu&@eLfd?3YK|xXh=_!ZIffp64-{0)7KqOeWxT84(U<60_fh@jV&( zy7T#%_o15jz-G-kk!HSbQ1rf3Ccs%rVbUzEGywrSKkSBFjdeDNlh**~(XE&ct8Y!J zgKX7Hz>w9p%#YvFxb(KHn*xwQ6Hjw$m-5`$aD==4#{Y&;k2$}fXoXBFzf!K^fvB3OmKJXcNC-J82LxA;`^!W7MK z_<<^6dm8mIK$8owBRA`g17X#@e`{qy#}D7^P19``B?|60!LiO3cO#HS3h%fQfELym zi?*;Ak7B8%3~zqQVLxuiT_YAL4fa+H=vy4Xw78i2AL<^}TjD-RAo7N{cs|ixll)Or zZnY29c)$hg&Bijl;>PtIoQ6nrHcAXFN5ScRuH$`UpXNGWbK0rp`8j}q*{a{B8|w<& zw2Jhdw9eKt$m6zzOb^!(4wps|3g2Rv;Dog??KBb!kLHviiP^N&#|OrUGodi1K)a z2hnkIc!b-I`jyigo@Fc&@HV9FLdZZ&p;3^jSSo`npLG`y@J zOBM9N#-nX1da?~Ij|qyuhGHf*(7B<`_4>?;O1!<`jv^deYW1E1GKH7LZ}pBMOvtXDdHF-rIR$$tZXT}< z5c&N93H)2nrJbv>&$b7lKV&*-q&Mh1u z`1RDingilzQm38JqFiog4IDm6j+NC*wRO?WR>f?FJV#4yOQ})8&YGumx0{Nv4h5ba zD@IJ8BhJS>KYTX({?T%+l>e=Gg2XDv$LJ7n+!1YI#<{-icD!EHgTN0L9Ls!#pi`OB zaTrSq-($dmiFnQGJvRTBPD+^I z+vJp!bu(zTMPs>oYZ2_&0g68TO0-fkGHvC-{C25ZCtDlDIFEpZEbG#I`!M=~!p1~< zYCWpgN;%Ip=|BkZo?60=UCroX9lpJ)n6n^zzp8+d$$iTY1k2PuNWVf>9^I~epRS^P z5y*g+IAtB!;Ogd3QMNQi=q%t%{YcZT(%6LaZ(|?pj*LfY^(vjEZ=HPp!wNSg7iPM!8L?>A_Co@=%qHiX!-TT#1MPdi zb{a+11`gh!A!Lu4ab|VMH0G`!ndaZwBYMyfo28QHa3xC9vT8t1;8+ngu&GDZ`{ROAnHxYr2#vXV!RVbu+uoRu*{z z9J!Pz57$zkfOzWx;8K5=Rl2bXdI|EcQ}IuX&`OMjG7MxJftDqiF^cu`?nnAnibX7K zR(8Wq5t@E%q-J$>B3#?ieZxr-mi)9|K>!*W&1YVzDW}suvd@UL$zp#d`}Ij1z&8m> zp>)iRjXPI_=BV$yKMXsS?1+Q*g)hbImG*UELRFix9+T*s~`D)xsVxO0)1t? zg*4}8uGU`c@q1jBvcKD!FH6bc!+AQJ_#CfwTzDT-fB)gu{@%73`@_KFh+vy)0xG^q z5OAU*bXhsA_qBMvb1Nq%lD;}Nr6XLksok;s5F8mk=dl~E9UO4c$wrw0-lG!L=#r0X zjJ=jUwuUs8x{6n50hDH*t6QakFva$Bmt)-O_3<145`5+mWRDsiI6eDG8<=vj7(&tRGSH07W)d7N%zLD1}j3lAT= zNcXM4FR~kKj?*LZk?$k&)C{szU`ACAn+^l-+dll=TNi>ZHh+-|EF<6ujDy^-cCYlP zJMBJ>uxhTB0cch(6!G#M1!+UG!);Q)SpOI=m>UzYk#~dMT*EBYyi&moNbm_4iCQN9 zGvL{ktxuG-EUk~1bjSi%JT*uAxF)*RWXDhw(3*F|4d$JF<^!kw1r*A?VT#lyA;MWj1TJ>YZ` z&=apG`Td6q;br=}cZ1|&Kc#&Ce*cP#LC_23|Hs9l>5;TcsYZG3!_y$3Vib|h)24qP zFqzpORaqc~Q^@ebzC}*=wDq4N+ZqLZt1FeLAxP1G4`j`OZ{Kc){=vedSRSWOH*bb4 zI|hFgzxV4JV?EOo{4sfmhrjzaKqQgU@6-R|GUT_KUnW_Vp$wdLGm>M0Xg6*-@(gWF z9r3{tC49rX%X<&Pa@4#)352<0Zk8Qys$fkqUq{xQ(?-l(=UvG^4T@)9dFJZMB>~r! zeo_Y?ac?Qp%I9Lgt_{oU?#tq7gXEu$hX`##Y3I8XHce_G>Zj2$l(vJr@Rd0<+rdJ%1ZG`5@lnzzoc$W8=D zM34o3hz(e{Ti55cFuIV=3nTnJkGqRxaaSo%kFqd5rYr3~cgF~}^!5pI(~5rj z`0FaCE3-tV7FjQ?G&H>s=oT_ofETkLcq!N);wH~)L1J_|EJ5C z3rivn5J$(R4W!8rgn<&tg&U9Sf_^_k(xHb%fSBhfY{234$es^k8x;)rnj3zLt5Dtb zmEH!(u`%p-s>2^0+`*7gNwZ`svpkpalWN0PiRC4hU}Y_^JNa&slD4)xx}{XT2x``h z==*K9=S^~Hk_Dv}>MN;3qjc00NJvTdT&l9Mdtyg&wbK|?Z!WFt9<%)@8oacCFeUj?wgyT`c#^^%c89GfHy5C1_ z-#Nkk_2fh}l=_Yi;?n$FnXo+R%GFUaX_5LjBsXtA55gfFC?ICF+v+!{v|WO$NBo#! ze)MS7k4wEV9^hwFc*PeO{h&YkX^l(3FLvzN5E_e@iiDyIm`ck!Fi%cj$KPqZb;bKj zp4-%fzD~;7BPPjXSI%5-3Ccu}cH+@#V^9rJ>dteQNc__^mZx?p1nvEM6aM~+UT1nM zqyyR9z04y$Qt(h%(0`H3DtXfgNp#@aG471?Erh-^%EcoyUg=LMc0F|-&wq2NQDjV= z3vsfI2V%R?t|wa(8dVNVA3l5-TQ6q#b=My}Tuy9Op*DMus01B?9pSE{OE-T%j+d3N z{QVZ2M0jgYUC>Ylm90+*YXJ$h^;aycM&pq<3|xHNq(%w3-pHbll0lLT3?qQgdj+^- zc6;Lln!8CIB(|&gRQ`N#>%^sJUs6{K$9Q_u(m&x5u;ieR8Stn+*u+~X#*6_|BS#d} zXSccw0j|dQLTQL3nnRd>px4Bl(fq~HGPS}?8Fu1IF1gN z;>|aAZK#EUuLAzotD(_E@Zv^khh&O)HH`OPsXo}(Vv=`7a|Z$-z^>b5WDz+Fo;FjZ z<`kl`fOBF z=n{Oz%CJQXs|6eqqBl|>wel}M6P~VUMBQCt(fVQh?4$9w*h$Wl5VB8DO;U`Tq6(|T zPVj&L_W0ALdTnLpL1K*Fg9EWoe0HICW0BBHyyKK+jFjJRd|Quu@FNqdU99m$ zm~fjhB6@Cj+DX)RKjLEGmgC~%>w7XWUavh`U~}DET53^nWPV+J(94Gh4BZDjvi>Me zQy#3o3frZ^5}EvE&q~dLx7eUO_0x4qo;$yFqGY9BNXsROo3oO?WP8755%Z9gr|=o) zwQD{~{4u?=yD+W|0sdkml6CGH6YTaMaOc|QdlTQ5_145J>>u|)Wdn+5$!;{v^nF`S zS__S@JpS@)hgO(Ij5%Oo2vWT`7QNn2y?Lip_Px3XXUA^b5}Whjr37!}@P!5FThcwg zvEHg@R(R!^`AnZjFa0M$VhKm{xvWwB|L8*nP|-|lrs#q8Qi~XHC)&l0`7wC@7tx`! zwib(Skx&Teve!hHS!{_rkDXe)a8J^5xZIv|RRYulCW5LYtI!xwmuh&MuRovY7FKml zjY(VZ1AK(OBS-v+nlA?YWhn+cfft+dZY2ofGGg=GyjX#4R%SP6eoJS5HWp`IVn&j` zg!UcLy)mxP{hv%bJDadl5fMbtaRb^FgJDx)B(6E$+Qyk!8&!AwBrh2mPQQ- zsPFzDYWDY1kl?9#wD30M`Jb+VJ}BBX#dIWhXYpA?o^h)rCFY<0`rKASiQ8E+VSb(j zsN)E;JsP4J`Uoly8$2G*M(d|fF(leB%dwWP{`|rZuRY@FZrQ|O0-e>Q3{Z&~TYl7Js}(X(Jt`d!#neFxXOe^}VoMt;Ip z%etUAT3rv6zRcspLmFKiOYf8JHYd+1ZPGsYjM#=rZ;*q{$n}knnI09fPcPJY1-8e- z`wtZ)M@8UVWUvDs3BL}LDtw&c1l_uXH*q*#S0GR5@3DT6pj7E8RpmwgA1R1)dJ=|m zCI<1G{MqYT^uQ8$W;b?ldFrLSdy>oQ)lkhkqRR>aYGJN@FiF&cTuAPSpy9e zp`&&7#d#)CYAk^fDA%HqKx_S)2&hGomM3-MIM432jhtgN#qfn>nFdLr~i$XlQIaq^_;4b?F8z zi3jAjAKd5VjTDIn43d2t$SQ^>Prd-=w8=`A4ZudCfP4M=;`9%paupdF|CN&%BC8K} zGcGU!OG`_+9dI0BU}X&gNU~zWX9r0yW?GNXpp=e*gCjK0!xS4ETfxq*xWoCzZ{gi9 z+QIvm`lg<8dg_V={~{z%6we0>uZDT?zWokHV_UQKM4mUNzb**wDZt(zSpC-u#oP+Ee}s-1y9jz6=bB!|z1&WgfI2ec%}R<(}|DJg&kxoEt^}KB`#1 z?dJB*P7-P)Zvap#A68=6V~B`sp+RP3W%ZSwgkGJLW~tdXP~Wj-os2z0;EMBwzQ#_g z(#{W5@U)E#B3dj}fpwe4Jd%=>l;jp3g<-GQI}l)qo}Ha7vKj#fRk4ao92cHOqtxxA zeiKaP1pNAS3Do#h/EQDrw5{=A^> zQ6LL>3!0`!;1fSC(?9yp6;%WpJy)*w1_3EG5Ye-^(@Hw`2Dt`hzoDgc^bRNtR(0no zzwUk)L+qo<?|LQDuoZl>O@`t>>Ruj|xFkM|di=WjYu=^zS6A_Q`bR3P=dMI>yMEsJLKyRT zca}eyXgbNo&W?_8_rV{Q%dlAs2+O7uGNG7uk5DC6T8D(l$qLvnOdjc2GOd)E)7dr2 zxG!Hrmkd7IQGF6x43b;Nr^>Lw1upB|H#l7-H>|CC5T7deI9O0O(EHwY^Jz!h$wt|0 zDy;=`bdaBZLM}0`d+xk$E-v@5*8>w^)VN;vv@t!t**)6t6g3a&I-A8jbR$2?O!qBu z>rgw&--n@Mo-GppUVemm4NDA+auxt+Nx?btTuxH*$7VsTiHqBFjQfAGY<-lT@EJtp z5U2KU^PcB^u@vo+sh&z4T*m^4TF?zLuEIsqJKezVBaiAM3b?UOckiAj?QtEiGibYg zct;=G#na1+S0f0nc9};);@Y#&`w}=(bxNS(^GV(ZTlrYNrrBuB^E+(8@YKr(I6p(K z6B^pr3fRnF^?EKz*N!#Qv(OPSRzL?f^Zw>^&ouvMD}DRP!4VM+tDLW2$>nIagnpD721^9Ie~zSF&;^$IQJMB7$T4R8W+tr!AY%htDmA`Za)YoLzuU>w$cEA zdqzbQ)1H4=1Mbyps8}rOKU)z)ZacdV7RbujloIpb6T|{^1+CrXpAQ~9_&`g6xm%$# z+8U)wHG5T;fW%6{#ibg_OD7kjdq?8Z^8@dJ;=P6EHa2XHVtlJN`Zo{W(%rxB9f(4G zZ)!^M@O!JL#~XMg1XjZ$>uK=i5gak^9SmLG MvH$Xf2mXpeiWE#Z%W+KNdHNBp@ zJlu2lt~_7OFulaTjZ-J-483h|9Pu~z>ls)}ppDf?X>+Q^N$zdz=Er$?DQW2o>h|8B zvyE*nEiHOS$$End@=_5WC0!$N=6$b9e8D4Q9>^6uUT*CRiBXYXMX70tdK311fZuBc z-0)0~1~++s+z*D3wnTOI_|3YrJ5IT*_7#ztPt>1j;>44!6TNjDOyUVza*;f%F>Qf8$H&Ef(zd^eVrfQ-8TniGrBc_MX}@y?+PO zy=_UK>qmqd_t~xvclZX9H&wbeHVdF=oNHW8eFz4RcUX)Uvo{%dd7E4;NTCniq@W}J zss)Hrnrrk4h&Bc>eGaE~e(90|vI;t40qazck7jqI-aE?XwqH5AYronb*3t&p zIiwK8zYGEFD(U9gB(`V%8OE2ZJ#iCdn~5w(yGemU8zQlMjwx3Hq7*-)HI!Q9v=1P9 zQ=ZS~Q7HK`%TqJ2(93+zqmH`r0nQK1ML0HNZ3Fh-KOozrh)-FzG!qug?wPbC;g_c* z)Q@rQZH{6jgz>hjF|J11K4l1uq1QQ})Nd*i1AeLNo2ZdGA~8`33=NE`Y#l`0`IUn? ze0bZt*PK?D>5a}#4)Mo$;x47NfSg`FG2@-|-rkq|<>DZ&Kf%TBOmvT$m#^wxaPXDr z8#S$wGX^&{J_uZ*?_XMh-hth|-@w}vNaMJQ;WXQTt&y+m@nYsGqS5{f9ghv4SNKcN z|L}!Q3$HI^SMc&5F7Db_Mi9zre}R{H$LGTbRt$4e3G$UU=BmJcrGQrWFBgFwYQb@f z+K^cE&0a?IJP2p*E++-{1H^?E+n4fI``lCsH8mNaqhPQW292|X-u!sy-j9|2@emJM zWJ|#dllgjD5^|E}>sR82TR@t`kfu1NNX5!Y=j9*uG&IFSgg33c1>*SKg4R1`Y3T&osZqr<1`0e>j7FZ_QiTDs;|joYcBZW5-X{R zc=Y_fiB}Sh%E2AZO*ZbEd;Y9s3JM{?!3MQ)N_T!PFs`*kj{UMwhXxsR7!o&3lWCta zjbT%$@4Ul?1L{EuVCC2XQMNP7ReCMg*?%YkOM(3S3Kr*Mi$Q5VUfi}pxmKH(bZfR8 zt~I~5U?wK!eT^Gas1kNudU%tKk1tB(J|zeWW>H<%KlM&ZWuK8m*gdDlpu8*W;o-6N z#IIVX`hAe_mfN#uSnJ5Dh|LeSIZwhdBz{F;BB^Y#d2A5bH~g$Rzp@Tysj~Q(+~zWq z1jU>wAkWpGAmY_JJw0uq<@R-g)IAAA!+*YrE6{5hufKcmQ0Yk^lS1O<8USHlY&F{; z;CL{CZmx!c_V&fk;$q@qr4R4@aHVM{tRr%`J6_NcFBc=QUZA)3iqmLon(Rk~0(@WM zOW>v}yR)dn5Yh0Xar*;zK8KBl@d{J!wDA&6$9Kb^#2|(va+6TP(z?(ZK^P+l4X<f)rqFCu+tG)tRqvoHQ&BeVO7+NRNVc(&P*zVrCX~m5$GN7^ zhq)HT@IIXE5S=k@?u<0k)U26~%e@MMiQ@(m$6q4m8Xe~z_9|iHZk@D&t=e+6A$ReL zaISjRk*(7sja5$Pq-PI&Ca8K^s)~$WGdEnX1IjTeVMf=rr-bDRea~^s+)mIg>C7_UvpDvqkP@i**QTT{Q|2;!rrucP{?C03W z%nBM$-{ula-Ce9ana%uZ(jKVQ;^UcKE7_;%Ufec>$Kn&q-WJzF3i6OOUg?fehj>SR zY}E^3m)%d-nD-)$x~$rZ1ZwWT)v&UCA@JV)+eR@QbV{c1?_hzphbSa{x;1Vd&~jhH zonoS#R{dqiTzMqkS?U^^*2_MLyh?DF6) zIK|N~PPnoQ~P=m_CCmqav$UUztkGwIaF+5MY`rhqqk;u6roOa$F@lJLL7@;%=bFtO?GfN$%3tWD_nbt;}-gWl#u!TOBrn2%7KPj_Bf9PQ<( z$gznQ)#>g%wd>Aof7CY~5!wHl@`W%|e!_d=;SkP1XK4}ci;b1^Y6s)k&GLwq0@9uc zkvtEG00B`<&U(r8Fc9ycN+{Bb2KdL9fOgbZiGG+meK_m3{Oc(&S4ycS!M?Z93w{GQ z_MU;Cb)R0m$cBCKVnqPWBQ$>VjN&fXOfr7MyZC(B*;CtApC)6=aEWs{jChoNuSiTZ z9_9tGE1Z9hipqY{ukR$I_tT4Hu5NFdhD~*1?C_LPq;jrAtN95I4rAQ%Y=Czq>r}*Tgi;p zoKRlFDZ_A$y#x#4B9<#Ov+KxpTePBd!DIvXN8`E2TRpu;^1r^vd`J-e+=9)LE9Q(T zl0I{{Stuw<=Kke@Kl<&~cSxHaw0>cF1{KB>zH(&{$_8$Ff(KU5Aii^9!dO)65oDBl zNfT{Do0C;S9)A$b*FXs8k68)`E{0=dA>9KExb?T$qd`daU5%iS%N?Kk)R{I4=%a&X zT-4&PB`!lykr3iPuYi&Ui!+$E?t)np1vu9I3-3l2NfE&}7Z(@L6@tx_1xqon<4AM7 zf!iAuC8e2>@&DR%aeb(IET}opg8#jgcB#R4Bi~bPF2e>P0OpKC(*uLVA=Dg5Yn=S$ z$T*N!`Druev1c0ckh#k(fw7H(ZEEofhzeClhCclwxNJ>l3Qq(X3 z4AxU!F(-|ar4<*r)V^0`NvqG}bf!O24`9neAtM2?_zl)GtdAR!=r<1WKaLWLhRP|> z9^F^I0Rh%R&pE?yMrWUHnZ(1K<~a~T>PU7l>ZneNS*QCdO#ZwNRlj++P8Z6{5rq@^PiKgj0wzX~XiGwBKqnKLKz?zyDHZU@r=Sp(l$0Et zHt++ZTS@yb`@f*aQ47Sn&I$^$Hp4rH4UQ|JbB9em zc{O0^7C#&2_i0Ucmpo||&5Trxtp6iRBs+h`U!oTy?09(G5uzgG)u+d$b48+GXGI{9 zFCK}X&Ca&LbnN=;p%2>&INGAWz`O`pqk}(i{@pdX`%-zZM_c9%`pN#AwXL5U(N5F_ z9^XYho)rW0#);94+tw8Kzu=5tm;aoB5KBt!k&oNzpCOkUk0=m4+Y?tGB#S&293N_B zfIXT(qJ!|rG>b%bH{$Y)Gg&5X;!8(_cbM7&0QE41QPRu%x`Bdv^Nzaw@8<*=CmC0o zCHywc$n)v)x!b!iF=D|w2y{tOp>_Mx+F?Mwku;5y^8)$4b0CGO2-0BR?~&|tPbjD@dcW-w2Y^SNE&EnE)qBRHZoI0u%wiBSk)J>3 zQZXWO+GP~z-dDPWRK6S{qrM4Lx7!F3FsyQug191MKHBxYTjUfu;~`;`{yeji^YSe= ztS3ZZFe>Ih@$n{m6!`C__St>{!~g#MYZ1&2m2uli8ga>Lhr{8=zKCZA9KfU;v!qDT z#+*i)vk9b+-K_uM!#W&um?qjucafg~_2TjeCDhH4&+_xh{RiiUVQY&2Qw;^pNsde5?lhv-#7U%UleWJ$ zDk)Gj;g&G&;{UbXb&rF|ONyhYP17ymceD2t8C1-kQcP>%{5{L}dR-;pol;fB{><|S z9rsxnGzaPXkh!@&@UqzZ40_4gh>38cXU#X3#aC2NvQ$Ya$Aa$**>#+C@Da7eO&M1f z7+X@qdiJE3trn^Xq2`HFj_x;?W2!Sr2`*w`@XB{?G7C>G|UE`vn+;V(SzYSh(A*|4<@!tdRmqLPH4&4 zV%#n*E$4?JD7T$iMTht@g^->mb1b+QZ%W#+;{Rz^X2ks-co&ZVV*@Mt_LHRN?Giu0 zPrh|au66fNb&Z;w*SFv14+LX+Dv~v<8 zsn9*bwj+yKott4(&ej{{6k6M9xKM)CpS3+hfK8hALNgmsI|vau0q8GftqIp{n%i!` z#wn?chrqD~DL^2jRZeR*xWHaP(W=6Xla1e?p7~~JH}V2@gcdJlTiogS!-u9q>obnk za6=iec-$5bDcYKtIUR%J+C`GhHbh%_tKA09_s1k05wtKgNN9 zPLPjQ=QfAur}aI!^TQ<$%qvS{$cG4;^~r?qMJLP>dM-4~qV!|DD%;>&nT^Tsz=JH8r=C%vc<`k@9JpnC^95Gh zTY^!Nr;h6J+~8X)ZUd95BCifq%|x;Ej|EJaWgd^mIL@0_HQfgs&7R7y7dI zwWzZDwM64&(&ubq96F0uL)nVV`xks{Y|H7*08yeWVa_b$_nF<=B<|nHF6nTz6+;0< zh$IP&a7CG=qQcRldTR$8wLb9PA=K&cy=&W^wd1wIJCo&C?RjzG#YB z{xC9}>!$y>OM@5{0}g$ktK=L1`{*xUzR(*5mmGiewk!S7&GF{n9X{&Os;8S+>dn{% z%>#ztjXvATto2~%?cQ+dn{RfyUbo`hZE+GS@;U=mxvGg1eeV_;NuYCsu&>`MohNuL z!b*Av4!Ik5?A@kZNqU*ZoZ=f??~}I0*aU`)&1%5EEn_b8_(RFf2VF)mR`jRrqp}(>UdRX zIC5Pn;e3vabmsP@pA!GSM+o!TDtu3t4HWSluRybi^qEm5Pjn|Y6b95p4v?n893sTK zhvGCO20E7zHy1B`=FfH|E+t`CjVq5nFTj~46{Ng;yIQ*&Xoqu|wgKDsSBCRCZ>GL0 zWBJ#XB?T^2nCI4__v7?2LTEJLB%EYHP^BLkU`{T~1N|+-mAls`79p>oo63pb-!A6A zSqEl6NzYfK&>V=hz+z-@7%uFE45#M0>eh@u2<|PHw&5f@GwaV9UpVHGa!qJg%Q3Ty zM*!<+;^MTDX}(p1*o@ya^C*V8Sct-~zBGuiy*HIKx*ll=!-q9=>`utAYD|@lpLo^3 z0st0gz{zG7x+e8kbDyZif%WVr<^Z=N2?Tb#Fo;mjaS*5m$a_i4Mgv}!^E%kXU@YCq zc0`D}djX3`z$3U?FHN92Ros=!aB<^gb7b8Y!30BmRQmrbS#h>g07(F$0rKRzKt;+& z6;H@7Iib$~lu%$qPzSuOADU<-{*#@ar!Ve50Zz|Jzv#k269g-dOyN>4I0K+p1t%%# z=&Y1vMp0UTfs_kaV{ZYmy#N3|g`C%4=N&Wo^0Bb6Bm=+3VAhEwj``u~gz;OT@!>-M zD~EdS!NGyh=@+=X-SdocyOAa~MNpr#

pG~KOxeZze-(;pjuQ$W#7R?x#h zg01dQNUwWjd-!)<90|}eh8Rv^cXDBsxf>A~i{-a&=9#H}@hfWvL47P)t}Ed@F_!UB zS}C1eUSCQ&Q4mF(UnNdo_Z@{YKpnpodr`Hw?+A&(Wzt95S*vHUST)g`X=b)2h|0 z&s+du@VAz0XLRocoz=m7lPBLH#|f?P@hWWDjtM+=eMGK!+c&%plS`w|C$Cz?nchM~_^X^Wxgcm`9 zPFLp7-Jp#vFFO|3>8}e}3vDZ+XS`=oZTmji)7BuXT0IQrKk1q2G>d$p5fL$%9h*cx zOI5L!cwa5oyeX~G+|IZ$smT=`vV$*uwq?Jq5ffpest!XFbZTv1l-a@I9(e+SEO^lL z6*@o1HaEu z1Xlr~=s_azXC4Z_DmkTNK8(mmN7r=|>_tVwHw7-fO%G@BN(p7`)&7NJ@U;G#IjN0Z zsVK-*pC<9r^Pa02ZzJHj4z=)P+V83aFIf~;h9@P=Hw88Rcc8mC$NiQPtKW?cVXKO* zn*C}ML`A?*FM7p^`j-k$GA8C&S6aihY;{VDQ4%{qL$n-5HsB504PkoM!C zDD&#Z*(XB42k;-D<4J$~4d}OI!RTR#6oD8xGCM(eVt=unG_9VqFHrQw>tJe% zc?f5ht|rEKec$gQMDvYzo6k1aGos{cpwrz8raK*7khJZ#%=X#`L9&a`Ix+y*unZWm zj$n*wx%stw!iQ3bf+EC3fi&^)1ut6jh-R=@iIgjeema{L-qQ6(_(@_aWC)(Ibsa*T zSxVKNkT&c7ggtFNjhfzY!t#|!0L`w|RT2R5y}B^;x#fpY1PMw){bbx#KK;pxn5AL1*VvFX{>+K1j^XA7HV`HZ;P`BmUm92k&OMnD9gau7|42|kG zR$DH}Je=}-pSHRC`-vpLPd}(KGR7+palb?Sh7pbU$hBRx|8ZT4#n0N}BLz18P;>yr zPHotGiYGK2p2=9}t38Lb9)o;Fq=tAznS>4zv3vt;3U$aW>9@H?@3_X*5Rl8BDvM8D zdrbU$jdv8u3`+ncrKEbV<=u8GZK#d|-Rpjyxa@VD7xl;fAx=~%f-3z4%Y_fjjFV^Z@U4dxdxG$iUUq9RJ znbt5adX3K^)9}U;`i&@996rs1<$Kjje|hWmEyhSbt)cYm(m1=qx@ylX^?QDk7zRk< z5#{e!Uwq#y87I#fh;Vk&S#HnECSUSs2j!6_aDKe!nanKqp%p^)^%x6fTT=Thy5i1j z{Cv>E6Hgf3n;q@=C_=JHvWq;Nh^{;mtam&X{j0DSiXrAQXmgS{Di5QjMNt$|zty)! zTz#ITSq~)%$3Nkz|Lh@br7il11DR4F+cmcTxi5sVsTJf$`#Juy+FZ_Vi*9Wb=?H)E ziO3z$<4qS&-|G$|1y^%zet6*@k$Vv`>XY*HYOkNTqHHy&(miiD^Fi$sz zI4>_F=tep=WU?~Rob#of5IycQ5db(0yXAjSmf(JD>YPZtdN~_>&|N_5?*4!lV(^LC zxDxTS)9f!O`}>0G!a$^hRj{q=GpFFK_}tu%7j%k zMPUKSv;;R;5hjwMeRsm8G{If-A(TR)mwO_?*37VDKDfJfHesJJL}za4P}R_P=w^ zeyhJ7qE7JDJ2`;jgcAqXWXatFXem-!lCI>#KXyY$gQ_ve2U0%J#2;V}xom?#`mud? zPJI0Q{M9QfD@F*yIyykg2__&oii%45$*!Cd_zCxD$LAHG-T}2gtR$5`9x8++0mFd- z@-B3&BY56pJ$}~x?4sMJ(d_G8>#ovjc_>aPmz2@xIm55G>k%Kw7f*R?vfNenM2T6l zsk0pB>*PsParVRD4K0Jlfil&bV|d@FlNKyldKK+147XmN7g0j3xF5xgt?=`Y+DrYp zmuNL4%$mk%DE@34oMf!hwcU>`@?Iz8IG5xT+v`lY&leAIc3Gdg6LieeEM>6#)*ef$ z-DkMma>VXhS1jT=7v1U^aDe)+P}1mt%e}KI+dhK!29h08p!uoN^I@pIMpe$GBAb3i z0t~5qs>aP&N%@U5R00Yk5%o?4LavulX%r+PT~hQVyXEk#|EC48p5CZBN7O=j58a9c zG9!^SBKBp*Wi%`Wa-hfL_ChgvRA`b*?i74iS3JBor*uYH2Hc!N6O)^LfBz2;fRT)g z@_hqi%~%$eWATdNTvJ`~TJLotaiQ>{^7-Q-m%y4*H(qcE>7-(-ysyJiN24WS;1ICp znq&l(NYut^+$!raj(_at>}KRV3t@AuZIEF#ZNz3Q#**rn8%BDBDeU+u^J4I z7l01MST2^-zG*nSWVl*J{$4Jkn@4g!6*x3$=>)SwS|e8%@XI&Ty9+qh%+UX?fCoC$ znR54rGQBnLNqst{h4{+Ad`f#| zui|FJl#M&?H(4WQ3yOz-J2ZYf7i!QNz&UVd>eR)=W2+lv#itq7=FxN1jf7usxuC4b zaJv*(UM(3lmKT|Y9Z!g+nH{S|+=2)FfMlW@*6{9HdH_DMrX2C6KPY%<9Z-mY&-Q*_ zvwfTn*3%|U2S0Duk6h}YLFAEls)zeN-DU7mUq#Hm`SI6;@}7qckw|)hqU6 zaIP`o_?pM35q`aKfOm3T<$`VZS>ARJ3PgN6*(H%CB@RbFI*8m`bsoF#ZAlliY$5x* zcwqI-d+a6xRTZ@Ef)83*dDvfo-vF!_<@$xN9acjSd17xJGbYkG+{Ld7wM-3@CrV@_ zli}}O!Zz>sR6tIYjI*QNnEmtP$x_YYEsl61lg(N+7EP|tbdKD5@$yyj@AUsjb_5&% z*rOk?`QUBrb1|(ohleNUUrJYEI_A6EhlhsRXsZ4VU2_+{a|;AbKMwyH{1q6C;B>nD zsa>61yKL|Ic=X3nR3W2I?k?I5Ixb1Y$@>;JUH@)T0|S!VGob*WcTD8rneUPh@+Umo zcx2v*=j(aAqtT_O5}@Ui_D2AwlK+9O1nb3UIvzE&^)U4E66I6*umEh=o6wc#ys0Ej zx~?6N7h~E79HKQfb*ELBaY^w)*zACD`S(wSD|g z6HWiTB97n%QuCvbypFGl^5sJA;4Bpd*Z8(F4i8vLg6#}*j54`XTT<3(Ej`LbF4#~w zr!57U?&o44znf!9-~T-b4_o_srgS?9kv^(CL#Yo?C`5YfBK{=`A(>M*owSS5+HufT z7-v=9JHm^uZI%8xa7z83Mp*WXk`ag|#Iij)Z2~V)L5r3gEoTy(%$J>(pXF69KZy+p zUZBu+%iC=bj-=KKBdaH%l3XUvb-2WRhue{zwerL=KvW=Q_hWmxG!xA?94XUTiRS-p zNM-0Ek3^dE6xAoI*=_Oq>#Z-Z+u8y!u1+4V(un$=!g*oDh)!ghZ2gg1V^EpK62B(P zkyU<)jM+rwOA>n61Ca60o_U2a<|W=Ivzr}>DP>y0l}C*|^(Ii^SUFJl{jSto-4mo6p;`-229=jkLdHYi_)1QdWG3=rl*r-#p!K

  • d9`zTg^wq05ljSBB4=;RUVL5UDKz7Y1 zF+4?=W27hmhW=o`Qh24Gv{U5T^8BXd+44WmtRv6=?y}tL`?Ge~kz)`vdJ;IVchmtO#=;A%YT#c;+B#bQp} ztD~c0p9tV+hHGpW8uI4nT0khe%Y6-Nk-mNfi=Pi~CWast_I2dXn$B-tYW^rEyZ>JG z4d2cmKvK2%&X-eXdqa&h{XWVOn-DJ!pu+OT)+fNW*-1gu1a%xiOJK8tic<3Rh@>=k zXXN>GUQZZNu#nRc*K$(ZwJ`MRxTyL!m9L4k8` zVVKtUcj={QID`SV-nW4?pjy3)Rqoh6Z#U#8SkQWWB`GRWXV09F@IBcQmEOf%Wtvn# z%NH*gkev7RKx!oW(9QQOZb!=Jk!bu$#kvNUfXnZsH)~U`HUa(J+#l9ynO~6{>j2?; zv(WMUh4XL)pVOYSXErnsrsv7>300u)PNIb3lOo%twD^yHg$x5@p(b3OZ^}bBNF!@1 zzvy5~PcYHzUUQfW|D;CqGRDX}uruxE=aQ9c@Am`@Hh}DsI z#jMkT3LI#MfFF}q>z0^lQCmal>`>VYw6Y1W5ah z&_z2v1L&|1itlKS_TMnVVf;h*`#j_mVpB$d{=tTpb2k>nE#v2Cl&uNJSVruZ zXaxM)0ybVCSHLP!NO3&^%r};G6y}!RIPB@j!DVVHgpeqjxA#U3KBM_Rw=A;E2J=<$djO9vCuQqYcu6rVVSTKfG>$GyPL~h5+>Jr+@ zFZW}H*t~2Usb%(gE~Od^fh{yvH&fs1Rj7$&gEO)W3~ETYY?mns^trG7x@1wc2ne2- zj?#u&gu*JC!IvcmGPA?AauKC++OpZNPCt{!O1DS|K8D+qm?c-2>?t2Dj&rK*k|e1+ zsJENnw9XW*LHS%gqrK;~4Z-Ca7Z7@GcYpSJ@X@(Fj-8*^((CZiR;XZPVCbF5BfD?e zjk|B_zdN6=Em}S+KV8>w6A%SFB@3&8zz6+ntzIwQK7$El@5l9J?Vin@X2t$lAg!yK zs&gwjUyT@S2XpvUGwBWp^Dynf0Qua0hbE%344?>-b47S_9Y|3L@MDW>yvp{6_eGHj zRIwUN!$W`Q>NmBu*#Lt4%Kuo5^5x1-btF~5(7{ne^ZGd_f)8LZW2VKxw@@e<^`jh#2NWHrP@J9J7UzFDxhxz?h!8rRyWuzMsC0x|E zF+@fWb?(Rv+70j`zsT4Kl;mtPI9lTTAyMQ+zP&Wk&a@8y=g67r%>OZ!2+R;p@$E{!dlXK#(o~RZ@ zC?+pC3CHteBfpyOQEYD0Dc?@9sRUJIMnohjr(0GPb2!#X!2kmH$pI>#=j`aBsHwN! z9O&-HUqdpMSIY0uwGlW8+(L@Rpig=^Z{z+S*zm7eLn=ptFfZ$VGaT2f8wI#DLJRF; ztNpIr1zsg?@J0}M($=-!tpA=^=R2_qpx}%vJGOoIUQpr@9&QQ_{>`u@2f^Apkvp}UgA3hHE1V99 ztOU=ut_en8CGF>Yq>CdiJFEfU5a=G;bvmCDEDRbmV&#yL^>~(fMe6K*eY+5fmA!m7 zJXxgfL`sGmE79%UoxxW+Or`8Vij5k_<`}j*+3gjD`|xWYeuwpa$tXux4mI6Xnk_1E z$Am8BEw;csPvq}1RfYg~eZpivY?j{VzWn7F%Pujo8zBbs1Q`Z$@QFkMeU^%EaiGqjhH9R4l0bz_2tLr z6htNi4n+R_3 zE@5Wyyy~`_F32uI0ySspfjb1-PB2 zV!iJQcB1X-p6A{YDK|TGo^}D$c^=U1m^-03$~ivq3YkNWc2&a{&MWg)n-C_E2Qk<{ zQW%%M2-vRHs4rPQ6sZ01E?z<`wQg76X`5lKu4t=K!?mY7^OA)T9Mh@Xk+&_kZP&QA zP0_?H?RY+)@u+8Ye}W5!{3NTOeMMu`ZS3inPe&oCwQFtl@uT6D`&Mi8F3=fg1_3Rk zR;#n^e+zMqAD-H}WpRD`szY##|6r7xjb^6qPtK$dsCV0d8<@W*Y~xXrXc0F?|6Osh)+Zg}wazzvK2CdVh9Jmeu8n!wd2 zvb+s|HVq8u%3I_g%k2D6IcRkXgC8nf4mh5t-Ic$nz4JR9N8u!J?>cB=AKMoFxkKh? z5fUrVf#C@N<>N9BJ$mHu`9~L)^q55dBAziZ+(MMZmXBNYU0U{#*$;tAWxqh0-1ooiM4TmCD-@@ZJu#6g z2*``XftAFzD)8iEkICNta-Ml#hXLC6yk787!XkDW^*U)a)P#$o?R-#x`PkB>JU*fH z)XAKN=xfCvMOCB(*C|U%!@<>aHhl9RbJCBID32|0 z_*?At33vcQ@`~UXouF>_$pbnOXCUd#Vbv) zH7;5ForjyIHRT^F`g&`G0D1f@EC{ixC{k8`1ehN*jOt_5m0ANo)F35V{Pu1+>)P0x zd@-)hQvY}AD07fSHHe}sU#~FW-_YLrF#+TQ*{(rD#(nu`iI%Q+1$hwyu5BT)P5$wK zsX4Zsfc(5~b+8{aJmL5e>%&Ln`=9)dYRx{`#5~SJx8b@5h)?#EWQXp%?4m{gUSE9z z#x_`TJ~vny4gcahu4!H<{Lit#^=uN~aIiF?DgI0Z?o19Be|@FZFT5y{>4s=3TxP@4 zIl)3dl<1Bgqv`TE%O%K``9%s3H)z-Set+7f}c|P?Gnmwb;L0GMg-QbH7ageb|2; z7+H<~9>jfoasvHkUp$XctW>8hR3A)LM@mLnEv|}4=2=E#&rX$dVAir*?s(0agmtK7 zbDXXarJ1&y-$ru1p7my19J>kwS+cNGQ)^U6<=CZ0hN-EjuzeO=czSy3vjF1HCO@i! zfUPhxOqm>%%@Y_v5y*r2KH~h`6nvqgWgQ={fvyPjo|nog{NUAR|K+%Ao51xM^>#`o z^yg2juQ^vSN}tFTBsSYDd#Bo2^3%B;$OXxu6B3HH0~`i*#oP$+JS2826)U6o^PjeFsfE%jkQhhXN{B`HxYt0*nND37ZBl~d4-#dloTe-4|i90lIIh7HDx=}yJT6RKo!#AH202^ z-ZsXEBY~nO1J_DUc6vgQUy>Wlpj@xWIK}Z*3(>jHup7o;RCYINx^aO#@6S8={QHSu z#N-jt=X8P%>A}GW5GcWF##;(YY~S++4G}$d#q{#Eh8o!N3!~tDG37>D{vwUz9_vP) z&#pqQM^PGa&wB&ec`ll!$^+HTY0-=Lc?K2+ejPr3l1rdpeVa1|DEfRmZjY`RPphi< z!9l?oR9QWIeTk8z&)rqTfU^qN@*Cz*czB)Y_4z*KJ|UqV@GyZ?U;Sp)*9^?pE$8=< z;AxsC;F8UW8Maxy?}$I{>;VOAIR1}qNWe}PcD&4Aj#lz`yWJU^&5z~(QkR>tEaR2c zmB3*52G&sBYAy*dEkP-s>`LqleW~Mm#L}-NSV&09OpMmh_cr>C$S2m9F}wRCGcy># z{c^2zv20g5WZf|jqI?w6d~`cKEvBs(wNW#Y%I=Dq*QRf)iG}cApIVRrWuaB$QOd(+ zpPcK_C-jrpYYL)?dW>WVA!AaoI+T%W5QD-hlx`ppj~rO6j~d z=A{Jn1$#)kEb6k}k~A}E$x;Pxr+l}1ZZBZ|)dHYOu>qx0S`D61w)EiMHWZqEH1GLO z>#f}rFiES?#ZdjCNX~m@fR5G53$`Ntgs;$!J0rybCH_!YSoq!2P@vmZSp%jk7a8?H3dUZtCuYryKbn4bjGw!^QGoID;^PqEW_ugJGJ~b50VRL;Z|4A90u^84 znY)NsK8hz5W+Yiv%Ct@7Z_in3vv-|#D$D*)F3COvLvkG@$L$ONJAL%wX&%XQWOJvq zKGC8Wq-II`@sXn}WK3Exg#ajNiF4T4+_|6g7!A!Ppb1P*wrS-5iIxkiI5s(jV z<2(9!<0uPf&FseK`uIqY4{m0{*JG_(*Bb>0Jvu9w^u=3*;V?R>tcQbwYG^3b#HbUJ)eiX`>0tf1cth|P&cj5XXm=-|Je_G} zmK$xA{e2kz7~5-Mqn*VhQCK{cgJeGDlmc6H=&4z*O_4V$m&L1(+IzSPdy{BH_wlTi z)Sq!9_gmsjahZlMLe6entg@08Ey5qoN`vSzo9Wq!q`tv&isTlH zh%zAGE-8#O@$ObSvyJSq7ty=VQw_a%+e6HE*Tefb;gyDxvdeAN$8&7PNAP9W8z_@| zb8v9r+t~g#4f;`2}!1=9%0$+GnQTQdrRNCRi zDl3*HpEzbD6N}3uvgznncAgftR z-_o=6@*-IyS^Fuem9e}zswpL6+xu3%&jfBOB2cg~FvNsQLRU*^VPq;N`Ltfw%6wC| zHXP$&!F4zOSvN#HwH9Nm-nT!rz0F^5)A&sVCFD~J(h#4<6m6(5K{aQ7!Fl@&cX%=B zY?Az`AB9kV_jVW9j5ps2L*rS&O6F0I^saAgiTjV^iaS~}H%`s1JVwv#CsoH!8Yo|C z34ePDz`sf%KYko_hF}^A8^vE2p&9!e#-v*Qq$4%-k4fjiDqr%ztBQjmqIX|mJ8Xwv3*-!!NSd)}?+_opi%`9{3% zKJ4Y`SQP zcvV<0X#!P>0L;3>m}XBKZ;JxA^y5GAJP>?CNT@3s1bT|vGR`8V1QV4js{G4= z@WHpiB(7%cd-MG*Il30|#DBzv)Bq!ZOaE}aL^?#QI^#eRN!um#?men3F*2dbw(-H} zSk;$4-W!iP#H&Slz6r>HaTOimtMwR7zBvD)+d*7&v_^#Z*Ir0r_vZ{j7mb*7t6!+2 z&)7mgvWlceJIbJl=pf<}E0||L_wiMYm8&CA70&2K8Ewz!Qiw6IuHF{|_u*d~=9ms} zy#OXX`SA#HyuLlW4HGeje!Zr*Rs(cWrKqC3w$N73tWrw~pC3mNNi*yE{P2XJNp^-I z+uv7laumIwVI?_RwFGq6r2;{`u7@!oIO6ALP9h=ue1nbF2E!EnHp}o?J0VdBYUcXJ zL1=ZS0{YIuL@U~Cg>D?{R)1+%giuF>;Ip5KC`MW$$$|LaJ7@-~&%)3R)2J_36duOw zp(olQ>I?ymFmQk0@}}BhxhW8&@0FvO{j$GYPb2ueN$+~PinTU8aPWF%LIi zU=h{@m!Y>QGI6@<-HtTNaLo?=%niiuh$}U_wo9Rz;e^je6?ehBg7+^{I&X72!8P+nI2knsdnM9N77yeRAB?yEYakdVK3j= z&mj}buWBerbfpm0wbSwDRstUK`Erl7?D4B`Kj&9`om1|$U`o}ic zgBqScfX%+|HLWwQx0lkV2pa#H=NxY$9AE`$!Sr-%XREc9DpCL}B|r!~Dzj<<6kbT5 za}vyInBO@V21%~kLGPoBiN!1Lx2x&-r-Ft*VesJZQQO{e+PoiViRvV-t)MvGAciKz zNL3U1mACi?aN<-(NITZzRg}5Hkcq5fqnOX{OT!>eTyI#y6c;2Pw?03c%4IeS%k8_k z++os35DWNWiVn#2%w~F@uW{)cIjy$Xk|uZhi{85Y=FZ_yo*AeMq(GLnNuaKJ_qh^z zeLp-qEV^Z47-n=!wRJpX)$;~pWQ~dyhHPeC zr;cOdjNVtFcLd>X>F$UD$HfUhz5GA679^v*oUK%f`R7K6R-9&a$0+{TJR*Uu_C6Ew znP^74CE+i-jZNNCA(T3E#qPp{)|zc_-ywIMl*KQB04CraKht*}Xv|H+zQ3$56KguCSL77Z#h;M|4J~GaMtJFI3k5}AoYhvaeJ|cYPln=0DZ}qyUwRo zF{S-CgFUFjvX5Wq=L_?hZ=|5JF72K|E%5< zwMetzp*Wik-VnG)qA%R;vP1lHeoNmfO2lS|QI&G3BN0s?SKJ96fkb)+j7R$!tG=HV zVb3UG$d6b~9eA@EQ_|5DEMFXGTRpvWcXT~7~mAe z8Q@W%2e;<4Ql5S6KN8p+n%q=rQ}GQ_cEh7H!|22mhm~rmR(7M;bo?npZ|oDY`pt5+ zP4;nO)7#F|lIZkh370oS?Ixq}Pdm?`LrcXpw*thBL<@$lF5|87qatM!nAk!gg+2fN zWvcs^B8+{=72<+Nt2X@@l_k}>zwbP7m)qhb26OlxFNmrG?(9z zy1cIvEAGvvzd(x$oV*Y0^TPedUjOATxf(HLT`EBkW6#O}M8 zfrIbs29fC-q5j2;91tW5p{?)NIzWnjE#*%HzREGoI_+kthuU4oNO=`7rZDKLp8lBp zeNoF|zA1N)5SB#v`thb{Ps9bcqFYh4KhlXkGK;-zsSSnHcSSd5u_`Z(x@lg03qY!5u-{HF#t8ZP@)GTWZQ6BFp93)S7Bt$Y8jo+-%_m;WS1 zmRbBK6siRCkLNoADs1c?UVeD`%(jck3FLS`KIns-Pa1}|!8WesZT2)qVeJU{rnB+x zX~K(sZ|xKoyJ^?qi5zIi8}royWZj5Q0DN}^NX?g$)Nd{Q^`peVa!w2t1iEx;L3G^R zF)u$KoFm2{@It9}a)P8+`b8)NPKZYN`C{%+>NbkTAl0ac_mym%? zSi!KSnwDMC_OXxjaj6(Z8mmh~Kn~(at_b@tH9n7%U+IiGDbZX759a|I9K_syfYC?C&#zaOC6kC~>8AAP zi@`!A{TiJ@>i-rfF?SC30?dB40M)Il3I$5Sql3eWZCHXIG#9C(O$MpPii*cYC>JdeUGTJ0K1H0V5Tg;Y{~6=kIO$2PtQd z(`UkW9m|)G4*^Gtk@NK}nHf~fHw1ZfA?HoX{JF*N^1YDCd)t-V$^h#ThB7rQ>R5%J zq*3vfH};QNB%3ymDrsF9A{0seJ;SXE)YB8eguH%N4f@;F!cL*A3D#6M&n7s&>MMrI ziV}J))`?opdacVtYugs#Ycbe~sAE9T;fIWz9CVE2Bm?K-oC0nPFUGQrIH?y+H~Fc~ ze54{-4g2W`u?LTt&j^VY(Szmbjivo$gpH?+nwkVyJi$(uwO zYDSMWGlavju>3M^<>Cqs|3-Pm>6Mv1OJn0nd$mo|Wej8Sm zVUj{+U&~y0onL+nB?C8})oz;AHSJo$#t;K&SgS1V`fSECZihjFSoizN$Fq}Gv=rR} z1=Z;_k5kUB?A5b;1x5Oo(#OiT#V$^6l4IFPQZ4>{*-2bBtI+*LW8|&JRDep10mzgB z;*gk2l!c!tLJEciL!R!mD>2JB{(O$xk&b^{3*unHWge9d;b0+{Y)r7^u&KuP%FPLcnsi&5T??{#NOUJsPCsR1J601uMOI<@y9~eR&J$RSC@ZQ-r*G;8$U<|wlbX3_w-=I zT-7}Wj3wKXjF0>t`O>0rSJBmhMU9}7gVfHiZVTjU+(XHnv^;77)^SVHZhmx+7LJzh z!(!5CU&aRM&hqG@87D#*bfAAg0Sq)=Z$1@FhZN+U0^<69>IHu0UL@4BpQzsq&uwSl z-jQ8P_w>kT*cg8zyXdEBw{(x0AGixsg0#F z2yy!2V@#--0Knx@Wr=Y9KsLylNBUX7x9Fk$R?ynTCqok4-?6wv z|7>VBynR{zt&zBS3njeGk&}N%72sh2&5clxr$HA+eQU#Q60{_?owhEZL%&gFFz9%^ zSf*T{Ib)71@XusJ4!le?m|JG^5^6L$;PtuNa&+4^(h;+th`QOoUJ<@Wfcw65Cf$Xx zQ@_!edEX?;Y`ffog@6)gyp~JB^I^#^nBsCR?6PSME*wOJXIH!pg!=#GkF)Ck!vP^F z$9{7`7W)G3^}zv3ZzizLva15?o+wEB%r*8sExRu zht$nz;au!jQ7vZHgzBsDM{ScL?EFa6EROmL#9n9NmF@xBH4M3lt=6-PA^&uMNl!kM z5~}|MJ-qr$nJ5~JJet8NDc3Tq;qK10=rc>gDD((7oKiXd;GFfj(fTvNA_U9l^G&|x zr=gsii|8?q5#2g1TOs#usV>fzM<`;Ku~pv^X*X31HB_R|5R?7$1o9PeK5y0y1$JU+ z$!9;JfH>OCv9ni@D15jZs&lW*N`*an_KOgjn2}QapGOQ=F{}RA>Gu=Bf_vJ>C$)L|{NE2n#ox76<0VEv^@3BlIYqX18!Syuj4GT)WiS z&l&6;;=_i&!`f@^Y1DUm7{XzNu6g>zByra9{AT+4a_P)u36#0HY#n95qnb}LLDH$9 zB&j-Si$+5;)B8Ja(GRF&^!N#>FNA>&9uFk87EPBg)`!kb9E{$B{vthGWY{b`m z*_Xh;{nx3nuV+16TCIM;J}1TZ@kZPGtT2M^H{j#L5?6u#6IR4LDA%xSm~MfgJBAKHy3>In1f-D$MFBxlN>TyIp+!1|?nWAE>6Di4?uPHiz4!Ay z@B7|AJP!GR&fM2^#yZ!z7Wy$08v!2Te0{G%=ERg$W>Ic@X^`XU{t{cJ5L=9df;zA*t&aWmHpqCo9S1L-gV;u~ML5d^p$=pVuk6Jq4S?Gd$)$mf- zYr2gfzcvy~dBi?liKC-n;-_}e^OuJ$=KZKBk9Kj_#%niUZERoK*eHd)xfa&{Ex|+n z?K%WLX5lu5gm?T9izlNOE7a7Ze*0BYn| zg}}p$%nx!qA4@$veOlP)!VGF@pq{x3-zVEjLO;D%cerAe;D`MnaY3&Au%Oej=s|^f z%(E+lV6V_4;J|%$d`^6!!A`pOyczM%WctJ@QQ^5HEHtTM*T9fq?>Qa8Rk50c+wuH^8QwBODFSy6M&5LHHC1T>c2o7} zcj)1t8dy4GhF=NxC(>PP(o+ZWvDGm+S56C%9)0FC=P7@>WiGH7$CB7^-bD z{@s^>7a;8YeE?S%h>yvymH=##jAy@2yGL-+JgmCkh$RtYG|cVTa|g*wtIHO*$%kVK zXF85e0ptjZ{QMzW;ZbA!5;kP$r$V5atxS^Z6EYnR!u+8@g8J&y?u&z}Kv+Vc;@HND zzMk8(o+zQ%1�c?)!54p!H-q3m=bW#=F`k1Q>rBH(84K({d>n@$LVnDmlHe4f*d{LK$}jfeiPaFf!cN4p zS4^mVNZ50m4~WM~2VkhnI>yV)O}SH7zuR`Kp&mb7XQUjB4)_2Iire@a;HGPG^yRJ? zg-F6Js4=?XBLUF8z*(Te+o!g=&igae58f*hfOot^%r(9}r2G1XAbTip`mjHzRd8dTJDA&`MpYAeSRzTX9z~wOXJ8GOATtdUQnSaC5?K^?Ko$5B_Ou9 z33$dn4Pu_Mn*6lc!(%zN4=@hv4ezMqlD11@)0hL!AFR?Xq1b&V+vgdcsa1*fZxfQa z^XaIN0*Tz@Pnd$f1eKi#d7i1b9%&ezPz{GJjJ($7kWJ+kG1@nLd9Nhmy7z~pm-X(B zIA!q(zzb$nN#!jzOtLP?1UxXT`F$Q3Mw)x=q+d;FD5|L^=5ZpU;}NTWt-1*%Sjt=Q z^#c@;*jy`hM%CYBlUJ_i!OhuXWMlVGs8j%$&qm61FjLk&u7(8^IljH3M&EQ3Fdc^f0hD6^)%6m?wv z;Z?11b$!uN6obcFyLq`aP-~V85HAK9V;L#nwc)LEpbEc<+m7h`&`n1k7WZ(#P6y9U zm4DW>hSi5+qQ@A0g#mwee^JeQ zu_G23uE)9ebqVaT*1>)psm@ z(LEiM$vkxMvAUZmenR%==v^e`@-9Y$teAS}Qrg(ELr(_2s?dID71r+9FrbXPFQGl( zZCx=`d7NhL=u>)My>6B=e)g(5V*f4`Cw(&{Qj72k6(tx$bI>CRn~IA2XZu*L{(-3V z=+}=oZ}Vzsb9Q$YoiX3Od%9Y2v;6{U+8N|9_RQ_F6ltbZY1!lE&@65vJLmpss5b35 z@XtXbXpo`j=keo{kHEm+heE_z>s;|rRl0ne8t13@888s_>kM8_`98$=qg%*%Q6T~Q z`H8XR5t_1cl4!7fR$aQ1N`$g?NGqTo1Sl>IZ-?#+JM?OKLJ3O*?u_(L<)iPvgd5Zp zk|Jb@D7+g=(Untc-h8>A%z@qelIYD=TG}^6Xx;Yf)XGM>bND-&Ye}K>PS?N)ZA55i zXQw#bn@25!;(zu(oLhxavoG+IzXi7&|J?jwgKc$Tr68t?x!41~WNurvP3X+-iF;a{ zt^lm^MMbgVAw^&v%y$ojQMq+9nn$OIWmQAzi=_h3b8X7+qVaa^Sm;_-Kh)g;$!pjA zY}bzgRq(n-S?Knn={GpKx6;EQ2Lb9lS5~$odXg)HEysEdK>y~~VF?8?m9SrCiFBxU zDur@rC@9OLYaCPcUJZsr3mL8pv}@pP<<%dz0|w{4X>1oqF>U5V>AI`h;^U+HkQioWkCrUrJ81@UwFN|Exr9>Vk8JVi zXT;Czj4YG<5Gs!Q&ghX~T`3pS9=sI)V^Dg zT#c16tM_FH!OG$=kBR=hSVrHxc^Ixx*lLm3ow@lENjQa^m0it0GRma_`Vgnz$+GcFt0H=>lbe_+ULG}a1S`f*1n2p#=4I=RzB z5&+YR*$|;Ru1F(i{Td1lXl7*-XbLwLCRPik{U7=KUnQfLtjmK}jSFnpz#dUKb!!WO z&plWN=UmzmX2Tt6zKr~FMsx``lh6h@T8!(>02O6T*E|Lpss< zR-V0ijfFc!jIkM%NsTImw3sIQ z{l!x6&PM=^NTg$9tLqvXNbM>ePbSoid6$}%gGR-6mQ!)B7OMK4$Cc-UKQwmsD+%UQ zl`USP&kXVUMseSO2^I!U;Hk@}{^Z&hkVCFo6K_P91EG$-10QWf;d*jV02?1Euwrg@ z{C)jx_DKbQ-{uNw5NIXGd?wjqLSG@`ap=hD_0PgaBAu!Xs3KGe-;o8I4+DS349O~q zCkuI9&*5)xiDC%>h3#AV$cxoA}ud)k)8_rLv_ z5V`v7{CkcnrNM_~R9;R_`%*$1An%1x_URT8-EbS%8@B3?Ltg_M>fVD$+)eWYc7k>R zr7c@nctgi#p#l-l_cpfH0yJ}9py=_xnZh`;gHPYkF2z06v@SFCt-7eVzin|P3C*5u z0c?JY^AVs%FXHp0;Z5y_%fVa1X`AQ|qY!@v!pZ&YkgC%*$jw-(FrrMhB1ThMNT{pq z@Q_E7IE)NqYx50gD@Z2{UIpw#o>fi0K5mIGnxngcJ_X1ebo~a61i&(2J3u)&vdTRO zSa{>P{h}@}jww640&hKh>|MW^mBmvXgTh`6y-YQhjY8ImC1?eZ?2JPF;JInSQK_kZ zg}U=g%gZIH9qBf9c>ts-J*k0E+k-x3`Yjwclz{F@aMx@lf&HSnBv-(^oGFt12@LB6 zmTJ+D7KDgZ@egki5FKZ-<)_IS&v1=Xh(a$`J$2H0yE-F36|h2oJ)s4xW>Vt~=fuyB zpVQvJNlmRkuw_ySR84qqH9NN+3TI79wX>vnx2|rTSfqKOe=FAeYAltX>dIx(fC{XF z%<4b%gWYkIk$Ja+Ha`@gBm`2+=AkI@)PmaRKooG8sXHs(L z$9*`Wa3<>_V`7ZCxg)NbnNaEpSOyOS0-=hDv7n#@=5@DEc3VQ!_yV1P6&Ky24<9UEZ2v9fNZ}@<#k=ov6YntLqwCw&p9~`hph^1q|dTT z`&;2&=(v3gEkSWR-h_uy1-9BZYik>?cS^j>TRtY$tc()tUS6aB9GE8zT|4sZK0sUU zmVuWIE#^@OYj|36ycctC`Q~W!Z85Wn5D-KlJ&dX?mBOWx#CmUV;}{BIVPt+iQ-lgk zh!~?_8At|i+Cx}pdOEd`rPmc&4;sjQWL8fHU)qZp{(hG{wie$T@QbAf_&F)x- zdFeRf^`B@#2_j@>DECCBgHs$(KUwT+8XsqM9B1g+sPXSj%4o~<^z@O2BR zN%!~`Ux&YqZBJHSo&uzW{!reJl~rSqEoKZUsJyzGc;VyCKDwLxO^}y03e{er#Ns9} zaOv{<^3}O?pEb!xn(og{-jnA0oXUDt?dfmDvmN*R-+3jRaz>r`-VSiN!ogS9*D}|A z9G3Sol!n;Yn(=F+Od%MjEtky?Y6>r2XiaL9+B+X*rg)qg?=?FqJ)u$rMCVllM5WMR z>)X5N?2DTi-Vddd&pdX88FRC=gw+^Pg|AD1rmo|=!?uScz&)w{6d|5oOeI`VC0^|^CASdGJJ+z&%TV!Qs4A+VWX-n ziMIuiZf$JB@`_4_i3#c!+l9~(!j1Iz3ufP&ESX^BNkPO@vGF8{%>-91rwt5lK{dLz z1;3XmZ1G_IL4^2<*k<^ zmHb{KGT)H7I@jkQ*=X8)(+u`jrrsS7ibaPVKO4OATP@ryb)LwsF7wTEu|t@SoFW!L z{_Gr8cJ>+>!t=608Qk&6nZ{3r4{vt^S@H2{ z2&r@PO^~X&ZOI$PiFb<%Y*=dmF(mEH! zE+|@;MzuY!ZKe7RD&!$Nzr~rGd>GR6mDA|KUX#r&=|k!2;NFuK$t^_m+`b>Wf)8rb z?~X#w9A|XQ&7$j*SAGGxyr>J%aRQbTM^RwMSn9Wn^yh}>0acwCqaY3x#+;19KVDh! za{L8sWRRQeo-4w#*hpPpl|LDg=V_f{=mX@!qdVqjHooxGre@uC`Gdv>Pg4dBU@0cV z9@07@A%Fs5XJu_o{J0}0q3<gTUmroG}6%(WB;Ah^& zoPRt429tDoofs{@1|Ymne7wL$I@98s(; zsR*?%&i4EypU4@0@V{!b0W2Z44JJ3{=ESFK2MO*eVL331Xix;I42PSENWuA}h+UuK6WXhp8GS^EaZz;3dnYJ?!zJ4aq!dF*0 z#!=IhlbA6}iNHNsIqI5&Ac6w`eh-l24mst{jMMP}K+GIT%8P`EMEmOy=+2CUaJDTM!SP_K1* zTJ8z;)=i1=FL*OlIJ3?mI9L&oL2rLfy~eN*OR_oG-&;@?&aBbh(V-*i6Ex#j6JXRr z55~E=ybgWMHWbbA9xpM1l$siuRM*w8c{^nGc%iCQSk6>QNr~j-cWj+!;+1%0ge!}L z5e+Hi=bxX75F1!f7({<2#qp$?e`D?H8g=m&)!@kBAK6yHe{zo~n2P!w7xZk{=Lu(7 zI|qu@aSeEyUL~ehGjY33CN}lF?j)p(idNTmRKR7~G1u;EjcULG zcGvBEJjd!d5^JxssBBJ)O!_6B*k|<&4VjEDZOF8=o@b^E^4?ru<QxVMGKJA!ZEn5(4&+v-`!Ep`$Y)BU8`OEO~CDvp8N>9n)k~ z?D!t!SsZuY@0?({5mz${dzhf<#5h2E!d5Q-47F$}FP?S5z|WgTx9yOAbt5i;DXgTT z(n)LLLFje!;NO#%hqa-5mDF0l#rH1s*k-1xZch24>E=B2AfD5R3rwx;HvK&%oVvlK z&=(%}NLU}ze8irfEaq{FM%`W@T`+)oM_zulC3|y088~J4^Vv-?nRP8{Uz1uJpA?06 z=aI*nvgznGos+^%Ap9=>3fvoIVrED!F9Pg5mm4J&HW-DGr|XN9D7CK;uRkV#$q+yI zOp>(B#n;)|*QoZ+_l;Ph8W2KDN&`(8qk0P}G?5hByuK`~1aCK9Q;E)YzwDDkR%q73bRw=_pmEObI^;-w!$=R?-9 zN!@Z_C}~OulD{+Y`@Cj)K&j%NM}=BAlb4i~w7t8VwyWzc5}CRLOT$BQZ#tkbi6qJ^ zAn{c>-f)~SP~dJj5SL1#l&9{TipOh4w(@D=Cvj1!UiTvQCg#RRm3B@YTfp-ixCQ(& z$)CZiJ%i;Yj-vYH)3P&eIHfezZA94b$Swou$H zjcd0)RL?f^fs{szsFCJ*&-AwcJJU9#`{i~HxqI*^`xmEylMBW%n*OM^c02Rd#f;2d zlw8(5LFXxj5!FO1a$yYsh0q}#kE*Joug1Abchfr4*Hb%hbFg(dx!nCYPY#Ig+@e!_iBqwD%}u;57qO8E*$*~U zjY-z3WKaYsNwe$^KRQ@9Fy4O zWGqTbbi{8P8>B%8d8$J@bu^xSSPF@_hj0mI7B&>bh&<}?!F-l3uH5bZ)aYC*eca5Fp}|DrPlt-z1{CtbtE zZop}`Pli*3F|jZF^wimAs_NA$kJCAoyEIuTO+S0L`mw03hrH?qX_TNOy-Z3(G#3{a zO(+R>XFPTDPZJaOn|xn}-mza-1jaW9pDt^ZU6X6;)9>$PUfuD_B<70P^?GZ?kLBcK zdIQU+KSi*Wk80PhPL}2K6Ke92p)olxh-dM7H(qWj_ll?=^YKcuT^y)J8oRyvFjTmv zvpEb@#-1(PhCJO75GkhZfy~mcI+CgW9r#7BhbD0c8HGcY1oMqziH1 zzJ23h2;iU*@hD1O$J49vWI~2={&pe)VucE-utqwN#36_@Tf8mH{?+KHFZCFjL< ztx@KF!M+!J=+B(lTl*>r>CZww^E0C`4(ZTdRq*QQ)Rehl8Y34PLPiTb7%N zUYDx2nRInaClW*mKCOM_&)Ve0xqicUtv-)GCWO|I0sc!@j zJq1w>q-W-cZNuL(tQuoAVw(fpkH=84!8KaTL@T0Ur#oY@kCd1t8PfP=c~eH2ZB)QHHVL?ys>(de>MLKeb|p9)xPKhjS0s@c5!;MZUebeA0+{<*nE`=vHH zy;nvMRn-zzpT`9G&jS>TC3cV@YHlM(&1lVte&(QzobsMnfSGA*zI=m-Corgd?YJ$>Z6;<8zNsiZEi`#AH z0pnWji8J?KBfrQcJRW>g7cOKhH>;I)8Xd;$9tV>BDMBVqqyCw*BO^#STTB}=t(|}# zLit0hjLp3*_VwuvsI=t*k>u&CkTh?au>B@)0M0k5BPC?SQ&WrTuATQV9bL_zs&q{Y z3aN5k@XpG~$&ZMOiTUDwLMnTGhrKyf$B{GiJ)n5D=Jj+)+;O=NiTm6S2js%ZqPT~9 zN-C<9@h=uCz*cXN`sSf6DV3ySRj*)fo}xtaL1;_U?KjW<4|>h!#4fut0bkI*mOt(V z7&mHBxt=GxZ^t>j7yVAmEj~Irt`5%|U{Og8%*tBDx?4FaV3_2w*mf4VGeMQkDH6|2 zfgvPIO5sX}>J-aqvDt=GTyTiG9s~!3lIV)IN&#{Zk^nvCMJn$Tw*36H-Yn0q?(WbE zJSu(&BuL{!R|xAUzHJ(7+B@4wr8|T8zFLbu!6+CQZEX_))^#t6II!A0alA4FsW<6X z!}o$+w~t!%p&Cw<-05C+Q$Agl1Q}a>Qu{7Uo>K4J-PWn>oGRqVx~`E%cvs2t+iogo zg(VEd(-~mLTAts1-O-Fv7ar61!hZZ78NJN0+7o8t_T)Jf z*7dEppuXf)11=j|J|GZfTAr8*T#)xltsps?%y|0hCs#jK<;_3k5how`2S0di7nM9- zGN42~$^$|KIpxOsI=#5KY>~XjoyQMf^C|Y)_mdxj&NJ1}zbXT|mY7vQVDI3=+fxLe zJrM?DeD$T9ZPn=K?Zs~tC#4)(!@3M=%`yGg+S9}O5!rR+ zSN}{$dn}ULV&aMnY$p3hL`VUsgZ){7n2jy9nYD-Nt;!mU!0N`w|B3xP_F1!*l4Nq# zRW&v3bxx#Uh*I(ZSl~OVVYe=n(?N3go9`FLP-ReBQpDeSqyhJla9|LA^s^kB?EN~O zDPw7tGs?oRF0LW`Qz=`#tw_|Yrniyad~GP7@IWK1)#(_o9F4b*-?^w`)&na_Iu~>x zl`VJKM;X?VjgxWrLTn^wx|7ME{>=Q*QH>WVBG9a@`Pvqb#Uji^!;cP~4~?faqTHPr zFBBCNItG%2CKl@I>k7Gtm1s_D0|G1p?gZdBtlTIc=Ig_OrC2S6$&GHmV+SxKaoTA; zNTqd6d?KU{=U(x?r-`fjl8W*zxqo;BgcPg%|`n8sAW73-uz;$_&wW zR)0qxcK{H(AP^QMWqh0zTB9vig1G2+*5sU7`ICR{UfNJTJ0wl9fP2;+OtFo=)GF%8 z%5pr@mW1~q@bhxi`BC5NdIQ3s&==ee~AX_@cuR@E8iX5*1 zVue>|je)^zkb;iVm4Iz;Rf6M#vfJ`O@yCWCxtTQdC;6vy>duUm1_rk>@Q8;6pFbK8 zcPEi%;EE`MJEfBgnV5DA)v*aXj6As4+}wO$uNX4Wb+66%1FJk@|Fg z)PH@NQ;mU1H=_sM#@ZUo4Ink&Un=(LyivG0AfS4eV>$#gO^eHx_WwvVc=lOA;LkVL zXd$fer$tAXM`LY5{6e+fADUVHVxic z_YUnsEFtM5gJa+Vb-iP15EGlSOFbrgZ)8M_t_NvXM%3|rP9$z3LuH=7ADo?AXGpMZ zK$t);ZK%t8^sWQre0_Q!fBO;vB%4f+=m-AK81%1 zf|9=Om&Ti$obhukm1`%_uN1wAiF<_P**UVu!i2LFtO%JBBTJlK*uJ=V_ijTgC@U+7 zrU;mh2E|*Bw^V5x8gkZX(qN^-3#m7$G(|xe#&R4@&IqK9QVr#850!4}WunsZ)C}~J zp%6&cCcM{fsbuagO54cxcw99{XlH@f;}IwFWjOdu<}dN%%g5#%VX8Hp(l#>0jd4Ne zRm}}T{(mYZofUO$picui6H!xh+*DzLW%&%&NGK^;X4u%-^^2mF2+8{UIpsPSjA9_$ z2L5o<&-SB!&qHyDy5cxIbg7*_g7S(3HD%qehO z?LHZwMVcWbY?bx`Bg4@Myn1U0rZtUm$MxV;YojDIfA+>ev~f+m z8DkHGf?HzO7qhamf;XW<*L4h6aObe$;hi7$OW?t@TH;VtCYD@Be}=yYUXS=6`j|cl z7P2qZDszCZ@M&oA2{QI6Pem8=1zpDmNv8Du&zJRc1jKGcX$fB>yv_dEbRigpX@ZYI ze)94%LZG-P2DO5j7vc~xW11XRaAO07K!LX`jdn}T1Q``9|b!{%=)ZL-o5TUKzQClf$Bj6aK3mgFlJi)?US$xMO7avSwstY>qTU_*%`)OPorU zG+VJNUu=p{=Zfy-rKqylzj3F@&USrYeIQ=CU}O~U*J-`|$a6VDhL&Tns`|<$N_nKd zLb7^qM+OnalcmJ+^F)@F?Ir7}B_p>08)%4gH^hM%FrxOD8z_Wz!`X;^%Dr79r?m?;z#7ppi z;7ITKfNy-uM|a1NexCl;wZogBeqT{pb!l-FY3Jhp%nC%O1r3}(v7brCR_MZEKRGH1 z%FaAXd?Q(p)r}@&{z*^*%v;!066dz5yDM8exmzn6`S)B09PF;f*EipG(bctjfEjhi zjryZWwvGCXdEkFsfWL?&4wZo9VqWIMX_&_pc>v$K`yc1Ojt)P z;$y)gr3h@t4@4uZg+H=z6xijYm2wcmMaf@17j?!REg8EU@4zJy$^<3F6~erMm#a<6 z0rdYKF1-VN?duCQ7?E#x{T0EUWyeTgvnl@$CA~*FTc7|cMg16TfxCxFgvGAXCLYfa zc^=Ub|1N$}-hHXENI+Nr`icYwFk$xHen>H-vrF!uw_Tk!7Lzv+D>~Kt;fL!8jRf! zcC4g?Djgf0gcmccK+IWnN0dmJ+I=DaTkn8|HIr;D{*SgF)cSkwOKC{J#s2Fr?-3-W zK0F081}i9vshW)tOLbUmFlbB4t2(e%$PocDp#{x)z&BYy&F00TL(@n>p5}kVM%bS! zMmzuMoPaJoH(uz1=P4T=9vBSf`8X2>%*=qXGSIH*sFbt=|9OEAi>zq$u9{_MGwxNq ziI`S9u0NtdFyB!|5Tat5IIGeqWTR=*qMm3WiHk5zLX}%4UK9Heer289LdfcyRv+&C zZ?#Vm0br>qirK3N@v*UVfWrh8HTTLmo0EUQJbrVcB%$m!Fqg$wg)p108GHsc@S4(k z0CK^gFeE~npB>d0d!$U6ow0hJyd1bEIDBU7npiun4Q1F(5yju?F5fDG??u>jS{0=} zD1FhK@=|B%z0n~l8{wo34XZ>gVw7h3pH>IJR~5!cGg+fnl}8wdv$Ea*$Fb|5=W)wM z?McnX>M9)_9og^S=Vet@GD=E`Zn~vvN&okos*;BHXA(f(tS`SbMJZYMMHCfAaDo-d zKKm;YLjiq&-c~GFimC>3v7!UKIL5^cfbxvQ-^<Bn!gkid2eUc{63-Ri9Sc)@IcCu< zeg6ZYtcw=}&2Z_lqni#JB1-A2Q;hD$cw= z^dIu6uk{=i#($TwxcDBp#02UL0C4knw+ish0l4FdEzkhpa9})$YU#WdLaCvl(JAOP z8}t7(m-$x#to08FQ1Q=n#bP0^@0S#wvXZC=VGsa2pF}@5A@=?Vuq>n8qtCa~_d&!i z*ZytKE}2P0Z_+2!Ie?M{V|bH~P%OEDCLDl<-t@wRWE9_^Be%&_Wz}?Z1vLWH)opWN z_E+d~nrBn*Uz#Qr2&4J}?2s=0h~imZUaoT}_(1{~8ZPyWtbFISW%l&C8|MyvL<5*7 ziGX06h-42@hk0XHcVI)AWoD!ZoW_UMJ>#Na;wNwi)=|Z&CDwQSXX02}i^8|lfRf7d z*Ux6W(UGBNmlV2VUOThGxq^L~v^(9vizlXafnG`R#Wp_L4j<4 zzG1<1=jVUC24IEY%&PzUEn!lv5ocche7@Vfiw)?D+}iB> zXO|%Uh)duA6M%>yW%zeqoUBw5Pq}17-Bf*{56{Y4lLEgLJoHI3ob1#_eUFI{vxo#y zGZb=Gj2x_;jvDjvi#fym7@D|rfRk*k>kQ9KqI;4>o;>2SsAlSWeCP+{J8O@pkCyX~ zSqAXkl6F863`}k}!C?Vl!h<$4lsHDrN0#RnuZ2%YkTdTe>QbtNFdy1f)^rAd04%x> zOd>(Xly%;7cm>#Rt@hhrl7^n04q*ae?&^(3{XG#WX(+KqN{UfrnhD)ZSq`D*4z@j! z;ae1Haw9fitrM4aD5x{uY2}#LghWGr4Ua%!-H+`4oc8 zuWyqaZb!nj_=Wt)(VK^?fdzSAot-YfWc!NK7oIp5e`&1qa2q~zD6rko(soZ2Iw3sq z549Z3&5id+0$p_`&d$yjI4*u{L;raegPuO!!|fl?E{i&*iukt|)|&ugpbcQTfb4)3 zW#U0f3ST}P#H3WX8Ny21-wfVN_FVlnJKEby7vd7Dt`+vhTc~V?fvl9krfP*~CZ!&u z5_lJpN!+9AiO%k6oYPw_Y6|VKq5Ez_hvEQ>vm&LFxfDMNkm)b@`&u3Y%mf*56bv+x zU_@>J{LI*nc}!-CgJxjY1wJ5LmQyrU%tUal8famZu`$@Q3t5Ne%8Ck;jAl4&T9xY$ zz6Pc#XNNN+#Kfq&y1H-PzWu#f?{;-;Vc6tRJ=iVUdj#Bs#r1XOZMl{5+<~FKK5Q5a zwj4lLcY zU@4@GN+V+A!2B|Qpnu#xbyt9B2Ny5 zF1z<8Yn7Nzq*mH9L_t9=;-vp+UFa)Z8n=m74O9wHPt_z2rNncs^bse2U{LoL3lG(f zf2@=NM)+C=9AbkdiN1A+;x7J779#Gk(L=;}BuNh@OJH=v5Ys*vSzRWnk`zSt!QX-| zqZfV@rtm_ooPW}Jze458XRj4eVl_QIt=Hi4lu@xQjzjkphL*o;XJr*UB=ta5+I-E< zWmK<~yF#74`{&P}Oi8Epe$b@NfbGOU-}E&3eEmnltLd4UkoQuB=2CQvyZs4#W##1~ z99TQyltPe}BwhR}kT6{k<- zHr9Scc6BW72O)o=s>%QZq zegA6BnqRV`YXAt09AsGPqHs}?kNihE2hUaQBvuW@}+vN3VDCy{?3wRf&b5{DEV@ojRSLdhuU|&;&$9|kRD_t zjO2za*s~GK{8`^!+2sc+sIbR}$-?Mp0tf_x5y>jiblXn27Q?JTuqz||uzH;{D|0;o(`U(cbU_{hFJj8b}l%yc)^Uw#Xz}Ak^MS&?z>q zu_6ECw(`XDbOdTSmNP)eATx65_NUR^aVU?0hKkz;A zD$jG5T#qlF=5{TG$^}o^7NSy2!0R97-R4{|$iS8f`)bamN(u^jrwJv;^C@Dk&))Wv zmE+0$sCLI=aO;?#r>3W;M+&qK0lQiLE_!7-TT^VgGF|x&J54shKC2iIZMyY@ASW^FUXl>@} ziD^9#{SKRYBCf~M?)}#Xl2-O=7v$3Z0>%B<({+k^D68VoZWlV06#g~|zc3tV6N7#`r zNy~O$IS)&b*T1{|LU#~B?TN1CP!GZ2@8x9ogUV7c8##YN`+mtYP7_-EgeEYKNUou$3GqQ1srfrE<(bv8?m)G z>aK{p+o6LiFSFo{UW~udldye%ef5LCV_rj5a(3ni*f8|W3f5{x<6-tyj%!L$fjXh3 z`zX@@;JQ0oi`U??C#$Zm-jpX)ysLVaAn7Fvq~|`{!hL}>X89CxcY>YY#m`k$aTjm| zW&*qFr0y@_9&Xtis#Nl?coy)XZ(}P*Uypdm>iDR+ zgbuJ36cOXM4ZN%gg_Qu(@r~LWl;6=USvz$l$3FV{i9brUh0L=gfw3yV6)vUVi~H~e z8^@veFJHdM@ehzZ)yL2an)wWV5*3kyNyU~1d_3!gtk>PJHnoqAjs$7cqR#lk_*Y;= zL_}iGCu??}KVk@GdP?4%YVOWA=wALp$ECt|Z@gMhGN+Uvb^CKq`tuK5KAgZ5^z-HO zuehB+aVh6qT<5g0Tsb63=u4CUhr_SE0mEI2Z@}mCX;uI5fyR3GP1$2@(P>=)?g@Cb zO#b^f^NGLZ&!PezLlA4}Dp!B$Lf1=5-9{_R{J-}{O#N0Iu;yX;5erfACX zi<)vtGc@G6AyP~bzxl6ZExM+A3^)LH@w2m54n#6U8_LET0<7kQ#U50Tz29Ha)bl|o z$U&e|kEKn-SHy>{*UTWnwqG~BM;12M_U2|3bKZ>n-lM$tVtC~8Xr$9D&Ov>MaYSJ;Z5GJoJ$(R!k3F$dKsm+7P_Ytq-7kWD%fD^jB7gYdu+ni3B*y~3(+vmJJRBlSjjbdX z7nkjWxRvhiZaARh3Lg0O)Bqw@F4bu+AP!t!A3uiXRAA!*cvS_!@%#_O(V72eGGBna zO+-Xrn?)uNezk4t&Chp26oB}CL@h&@tCrTa`YQ(Y&}q2?^GZBEmXaY^ z%7^A;W2ie%9bV?J@~BP`i!6}m;~yAk{_``cu+w_jpe1+jf}*10@})uT2jZAFTbKNp z4qKzyKw0(myS?!1E*v1Ts(Eg#tiV&npL6_tj|>I2TZBC?%Yt~Hcjwj*^EgkWz~rt< z3~E(hd@d?VVVxl;CXuolWmJF1&OxpAiny6~g#Cnsgd{V(3q~CZWW)C1Vf@Jwqrztn z6@q>Y+1|h$?HMmG!Cqu)Y6;)d6obU^daE(!FmgV5s>#+?>DcEdKVn%@-YK~P-jFFP z?Lc8z?(q=C03Ek25t**a%ftBn`?p@RXPt0_(mekVx}RvFPBFpWi^5yYmaKCheKxN^ zoxa@jY|J{P4&V6(SGfDhdYrs)z2h2|_ruLXsN!sI$7JtJ3eB*3OL~2UUeUJr>R&Z# z{R%X_^%-CnWH5ze|Kq9?tIwE(F&`MVo-5f3%5HL42^1XyG02%l2nv}Hu(3MHEObDgCWEm5DC^%yek^5mKg+to_M{=x8Xs{E1R2n z=e2+@cMTn_KA|v}n3S}zHm0SpD&Iv6m?q`H>7s_Rs{AxQXliPfhkNQ+^BK8EmS#c> z!1xC7LzyqYW9LutsD$YTGb+fWEhe1{gsPj6Ok;W`a~z9}oAkp7tGo;cSaY>%N?Y#r zlQjpN27r7J_`d?m3czdsx0mWH_yh!su2yxG;&eR-7XZnH#?XVNmPiB3e{+>JlZT2O zpI>88uzlhx%gL7e03huxX2%;7 z5Z+&39B?lZC+5wR7>P9=by2z339_z~P>;p_X4FpzTFK>3%zy{v(a9-OIY$;IM zbe?#bbzn)5e_n1iai^Lh8oXEUOU2IpQ|%$G>BKi8Q-AjI&&~Y(l~EHAcuSkiW#r1} zb8t(?OGKmVY9`mIOT}Cct;Vz&yKK7fIsl|KJicUxrAK zx^eLdQ_Sh@dVfNt*g=B(!^Yj+fe*LO&FsQ>*hz%=anPLS$$)i|$JurQV$9%crd`U9 zP=l>=*(7i!ZVf7t#lY06@}fd~chQ-p?ylNFpT2I@#~5an@$1ey_10o?I2!eD9BcXIk zDc#);9SV|?k`mJ0JqXfL(m6XoBy}$Q;{L9ZcsB_Q0_Fil4wfDtw?8v5z zUWQ?!QL1k%pRj=?DJDMmXlxa_1214`2E8))>%N)<>Yvn|Ys)4E-^d5>KG+8bGg4{$uszrKN zk9n=eow2AF$uV`CSK{BQ8FHBRKzS`=*|jTPwp<^|dz_4E{-U@7zuvH+L-lz-_ghh{ zcmQnk*-)qIiRq=!>6F&aFxfLCr!M~%dCS-xcZ2Mi}%b?pnW7n$J;I>y1zeqoOeQ~6Y4nb(Hdv#+6 z@vS5#Wj2xdG2DMaa_IaJYt+|C+&=?d`~w1T&i!|N8JhK_*g^i_6hj|?7pP7}fS#hL zp*sg&HQnAqx`WfRcyJ)%Fg_Yk&wkGH>oBvy{V64M3i>ce1Sn1j1twUiI8$Q-Brkns zE`ir2x?vosHxVyT8O? zFW})-KIJTlQZ;%VH3QAPPqi`7yY);`qLEen^WD9014revEQKmk3FW)|Z0n_$ocG zLUDh!M|CCfkcu5P#EhhCe$)CG3lEPsfQ?8PYhwu|YGblIOJ@!ERk52GeUoh0^6U&e z^lEUnv=4-oV0x6Q-bLB;iTx$7<)}RtHNDaUiDG`Y)aIT7V1wlY<4piY)n5ZTm8l`8 zzhg9SFKfy|rgYfoWefbOfMTb3O>YDdt&bs&j!+&eQ2&)-P-r@k^2BY{Aq1Ol+$K07 z6cZwSN-uKNG7t6laEV6|VS~%a1WTpz+07S?-yHUFuDW7sjJ34_cqY6#ta^u#DJ6uxPH(3BfQLXf5Z{drG5;v zZuGN}Cl$bnO=<~YriidNkcrP&OZU+SP|jOqUEST1J^)HQ&_?+eP{1rkt#-{P@rPF_ zL02U&5W@hwpMW5u2kbvxaa|LnL>}4UqAo~ZefCXF4LbUbtRh`L2nvNnAUQW1z%do$ zGZ&Or21K`Aekd6IC+1|j239wz@K?=)XgCj9xaK!ewko;>FGl1!^A%*$eQxUm?83Q$ zQmURLX4ap$czC9J2i z#ZqAX@jeFbGg@XzGbZ&gCiPrKyw|Ni?2YI@>5awU(|;oc6anyM?~5hqow+}=R_+AK zzW5OGoUU+g-E>txqDIiL0w|@cXc)kz2AfJIy;%65YNFt8{{c@#!qe2-SE@dq>`4$9)~y@hB;#A;J^9{-?1K#9 z4i0A%onC{pa%KIN;3IqDwvx+t;ZDggY3Y2UM@|v=Bx4!!vM4YeOhQ+67zaCsRsh|5 zWB-x0U~BLs6p~QfHL@0H6^yHOHtQ5=z0fi@3!Tn3X!S0@P$?btXqF;146k61nZW`D%}ED2nmb-3djod%8C@7dT|~X+C%ZDe4cr+i~cP> zH;vAgzsr7gRr>czHj8w*D=|%M>J402ox$6iXKX`XqM|ZD4zW@)FtbD$lwb zB+{d(>bxmQX;gClPXK-3_*(~2BW7=&nj_cqbJH9t2tSg^c6$dRB>IJSg`Xnscpo42 zFsJjQGC?pm0n(w9%k#$^6Vrak7mPA0h{~%2wG;6N)$%iILW($*gwY}{ zk2kV<4ZpjtU2YBfM&F+K+-XuiH%AiLqiQX;Tf%rR-#ue!8lYUkvPY%o(!z*ts04Hq!72x30J)-d+xuDti_bbNb@mK>lTxPv zM}9L9zTR)I@I0{$x%IsnE7VkkKp>;sRTcn18!I}-)Ojs0uWEu$Kx=KAi(&!JpihoI z@!Y)0Y|hNeGCBD3liPBXCGLr=8Nubu!tHl|6<&K(i?Vz!fEk>Th_6}-Z@ms?m;)x0 zb9tDxz?>Z4M&_XiA_ecvswj|Pe zUO*`UxB=KN>$lmU$M&++KM46S?r%9E;&(i2%(2o)z1e?Tj1 zd$fxP_rkiD`e2$!oB`?rI0;7&469oMzZ%)9ZiBh6CYh1&M_iqZK_YGs6I$!gB0y|H zQk@OB2=Z94FmAiZ!W0sv?d*3_)N#v)Z<)Rf=VT2w(HtXj1mF5Bk(4gdH4Gex4a)Ih zo8|3Ft(m2g#$87oYEs5?Uvg;E(SQ*C1o0FLn?@1K)EwF`qN`GL1$^2nDQ^CGe~d$3kNAe2JtH(kO;yFa{#Vyl*b^O@K}bFVtMC_K49wn}-@KegGVRS*Jmfvpj^V zukt)**)WhVvQIBI3rkrsG1ejtCxPB*L5)t8d9==5H$$X5E;T)fYzj`O6?#eE7{&LY z(=oQ|WCSPy8+`ZT;HabV-o|)wvVVmV3?ZHOU>bc3^he8&!8&qKBM-f;=lnDKn}!mA zS96zg4&WgPL{U0MZm=OKf$TC`6HOjY)rtrh4%t(+t~QT%ty%Hk&&2u!2Z91z?#X(PGo z))`@Im zY3!eSomxM54`#1?*Nz_7Hf(7%UZm{Oc9+1eomnAxAqM~hK;l!WG<{vIZ-!TK={)WP z=?`!OHMalw(SPFZO~69EWe-0-9X}o4`37e@Kd00^iSqn)2b_!O=o| z{|7*(*!3op!{)C*M}I_4yai50g3*y=G%TNzPU+#+Z1&j!OuWZV)j-Xdf4}OT7@R#D z9z^&d(BTXICaOXX&+%ud*!&9L1@ue z&#v|6I~_x8x}b}tUj@L*QCoHT;-1^1zKg%OFC#$wFLSjwoy-z}1m1UGlA2=;VM3O( z%no6}G?<)AMaZ5Ky+*YpY*JQtG-~a^g;edWb!>ir(Ecb;dpTg}ba33n`G?v#gfA_B z*=JACBe7jz`&_3dYhriWCNknoAU*W9rb#h>vy&5k1vZ!-dFiJYx={*Uie6NP`|j3d_S_xEs* z1qf5Y z{Tdbv#@z)-uze&4F8@duHZHE3H%{-yPVsHzQw%Yp4EzrY2BH|8z!69@TIvj^85sVG zg(ZpwvW}ICXM*E5a5z7dOXSUpZLz)liu7EElf+?-q$0f*0CjM)!3a7ieDt~L{7=~T zyRBJopvkAyWj&PP*WQS3mX0TyjgACt9r@FtdyFgv&i^}XgHYvA`V6L>Jw^QIh(7!u zN0gYqB7!X*NF%++J2RTUx+CsTEQbkrs+?*5EG?<8QpZvWyO*I5s*NCsPQ~1Ib*r!a zFld$hiUcQ>*+fM2M9+~Z>gay)lxko^fQenxK`=J8%fn6$07`wiyK22t1&xml0D)ra zp=-wuiz;=>c;fFu6dSe`!43n4RyRZYF)`hDCQ8$j;tHuxYT-|fci>jye27<{M_Qvr%#{I#y{zDV< zWGC5q6;J<#v-Cue-JYevd#GXc9qU$yx@F7t!s7*epQUp5C8VV zmw_!K(mIsoXemCq{bYoROupNP!Yw-ZorpG5tju|Gxz5EK|nit`UF zL9U@j=%{Fq$j5+PLO#!oM0yQZovv@Lj|OQfw)iKu0IB8>kOK2721T!m6w-vXC0+># z2$=5U(h=Y_SjHq@ue#Nnz{v?*ANQxS)ug1P!p4|_svkh2 z(mx!^x=?E(h%O(~F^r;}eAviIASntymDF=EKzo%NU_^wwT_~tB8cz4G__}z=f4Q1$ zN6r1RA!cBn-*sC{z2^PT&{e+itBH{u`CnHvKwmCa%G2~j|1?~;1RT-3PnJz0%U&S1Xl;4>WHRF-!J z@RPVcTo^y}T&|({*%Qz;PZbKc#SgNvt??4Qty5v38(U1L}QNR2zQZxaG!9TdrWk(wMQG?Yl3jmDt&dU_mfvve=U@C1N!hZutw*LVfrFawP z=G6@u^&f8_TJu$NACi+H?|dztHK2fA(3iNB{D3(dXlG(8Nle{q`~=)7Ie_<81#WFs z#%A~!6C5hYPvVy_a5a)19gbnL+pzVtXJCMDVryHF3$;sK7$yvs9G*L0EC=i+|2 zJ}4m}@ze6lk7?P#V*;myMNG(Ou(d|X=EgM*z4c4wdC`B%kHESB#H=6{q3?UQLC2G3 zazg@N2RO8yZ`_MqUs!wZ%Rn%@pI|$bn^bG!a=h5^hWQ)F{4*163QnQ7$-e^e9p|Yba~#Uv=zn=cI!EKO$4ZDey8| z9AX%4DZQV^178X6mq4L%L`r{8-0YDi?K$kY;B(!>#6qxRT0|T;5|C~ zV*uo$ZI+5LsJ;P?mO5QQa9l=N85ilxnPx9H-p1xaK1fDj#%o{6ZbU%R{Ile%k3_P1$r=`T#fyIP76 z(JpX9!;G&Yt$-N1c?rMTsbyUvoko+qv7hg5SY}7nqt~hit{#^Bh3D{oO(WH20%JRb$Da>_WY6NecqItg z#h)AN<{7lM^gcQPC{V#(%#nZ!$;!sRRR=vR`|LdnP#(DT#|ud&nBp~>n3x#FbMvT% z+ILiISLCPR{XF|WkvMH3Z6XvygyUePHaJyOS)tc+Iuql&c0QJn*6zIZZ6O54k7vjP zZ=fO6>*nu+{Shq*-Jvu{45Rh;rcWCb#SqU&F<_RZlfDpAr+OiY{!n)Ko|tx0~r z%)Bkl|D>M(i@W9!^N*x*&a5N4T~WK41fUi5%HV90zKdMSbRMUpWq)sFoqj0h{~HwR zv2Bn&ayJCiq3q-M%g+50-`lM1^DvnI{Q84HFjygno^Go>=|{w;6U?K|FHZ^`t;|+z^uTOf*@DNsZTb4FYPLY%*R2uMUP0#Php zE3Tx)Igzn*vX>$;{QvpsmjA%a>;Pr#WC`6;Hw;ze3-netmL>c&i82NR+2rO};jbG{ z9kFW{{D_Xb@;7(iJ-i}3y{ww`Qj?GB-)^amJ}ch&CiXUR3wzYTw(rmNs!=PCX@8XE zSV2A?%(v{w#X*$7ibkY<*R#2COKZ}j!HQFJXRXJDbz{%>NWh~ZC6h%F>Iir!?nTOb zLuyNO}ltN?v$myJUsM z@s@jYRU9W!VMZb|^ZtB744+_8)4X|X@62G#-;f6XYUM(GBf!G)3Ji-Bz06Z9Q0rl9 zyg8l80dTmb*w2#(DRAJEW@t7LTzZp8djQSw7wGoRz*ZU3eShSHs`k_-`E$x}nXKWG zk~XuRo*w1BcLV8XH^y53Ieq^s{NI6PQ)ir00B-s#C0tCm(gCUe36h12Rpok+>Epr% zIWXfF(Z7(Am`Ki3NVHu&q5lQw;Uy7pu`FP>?kE8Q*)Ki}YiA?^EcpzOAZs7lJHsI` zdfEGbXl@dqps@fL)gCW?%M55)XF&L0bBZeFvTVwb_+?e#QNLTjw0^^PN<;I&>^wr5 zb)At_Fdq1Cn?IlC<(ds+c$CHX?1{DnqUFP$%$Xeo5rd=ynNWVp2x>wHOmc!07~mCHOc+^_U|Oe)p2$eVacpC@6o0^7FS3Y&VA% zK%a#hz=%nFHeUgQR>eEuw3P&TLG4jF%l~0x2wi z6zTpsj~?GfN1d^7vA>K~0d6)ETzwcodcVZ@$1iB3-LLk3wT>v9XUNB;H{=J_)}CwC4%QC^m>{l|pu1dw;f#1Uw>Gy0atjO;xZj*7cDiCg`iaS0Bvs>nd zd0!cCet3w^wz&G`yDmK4aIa{feX^0e{OZ4IOD~%=tV>qOz79L|vAr4}Aj)nz9CYFO zh`f1|*6-{-kg<>p#2s*I8f(6z4PlJY=R5 zMYhDZ%t`ljcgKpimkPOVC%(+lkMVtZf874|aClP1d?ft++aEuFuF}>ISSiHp18*3a zXf-AJJjH&_k^L*`nYWq+nAbjcbpDjLwr&DiJ4`;!XaDnI1BRF?ZZztA{15r(oKod; z(&ExTC;9gNF!po{F_X!y*RNK0JFCR|G5nxblZBf-6%?7>|L(I$FK=$1M=iL_3D~fe zeT+U!@LgG8>6gCOX9K#^*R%=%X&@dN&M#ue;-}80b%wS^JrSzNN?iH(nK<^n(vpnn z&HLib9Y9c5iXsKVN`pPpItP}36Gyp!`tJkT*q|UbKX*Ae_>{v<7l(*G(XP0$v(#yL zDc#lk=|TQB@~_Mt8mIsRVj(8QQSzy7lTs0!@ZSxPzmlg;5-8%s!Qqh#eZ*4|r$3Eg^JVOAWwsE4gi1Awnsx ze;yPm(>XDwB^ofjIWe+{12k{gg~epJErUv z(-`+D?Qg#R+wqC>M7}zk>QB+thLTul1)?&#PkeL9FKVhJPL0mF-CSX{A3j%{38MNr zZu=2;4w$G~<3y2uoAscB&o+7z{SE0%Amd?wQ@G=0ksZ`EtiI?zXXlQ7BS(*LGOiWawd5M{qs#=c^ z0dOCbTyInyFthyXPb8!4TixjXi_V1Sz3f83QrJJ+axdjYN$fR>-UWB)$Y zH45Zc$nu9eIGo&{++XSkVxy!XEq8gR74aU78V;nJtxKg&n#>5sGX>aS6IW^Ad=|qpu^$q zpVu*-Or_$vop+qNP1z9avg1t3K;4+5izM3`>CR6!tE^L`Uou1un#?h<1}HI7eYjrf zJe{SpYJX1xa(?F1nyLw06r$*qJu|I$_*OUOWp=ts7h&J+iGiCznf!oBYQ?vgR98(N z1A<9)oUh8QTMZz)A^voz*V-i&wIk%(jo8EJSt4aCv^q)fS>Q~E&r55R=rRJ0H_c-qDdkwR z1!*9+#Pm7_ir%k|Hf`@h*;>ZY#tx;w&zbk6Quhf1gK7jdo`%F{Oq~oxmsHfeRSa%w z#W&IIFn`9UyNM}+PRjlfYf`XJ37A;__+T&=;c{A9*{B#HvaYUg@&`^|A=e%2mt!BW zW-un~_5J-m)vyO=h#!1;3}AzS^rKrnptj} zL&vFyHX*^!ua{&O^f#c7Y6fM1nDYw5N(eKofKMTE8!3()#cg1KN{&x5GjpqTi;==P+ViIhTr7wtfNJzxcz6Q?3=f zAfU_8Tia?@UtiDdviba|!Fl5{=QE{i;kOWHIMbW)euC;#G8Mop<~!lTwEMSTiYTrImvg3;DzDUC63BZcX~l4 z{c9Beu?_(WfrQ8W+RKK*-~M@gE>pGbULASs{AjGqG>RH~L0Hsr;kYosQA-T003zeu zhvVIsoj93yf3{;EDkRdfx0Nq|&=|p`UU8cY{F18{ z|BasNw7VC{61QSL?F(f3#oJ8{dm7}UyJ z2Yl}i)Vpd$Ox!^09@WSM{XKKrGrVeFZb$ z@Z_1cy6D`wSI4BcfwTWMZWxi6(Tw}jc}c9*=ua_v=~GG*CFotCv4F)klp^fFn3W=Ztl}kf*ImZ*)DV+G(-dXSv?t3d3bSAywJLL46~*KWU>>azY(-@{qP8tyt z5*iQcd|C_7nJi_nch+_{@M5QzMax>hxi5BrN*NpPtpe_F?_Ap9!MKN?c~)YljuGz~ zHH^f7Ouab`;QeAXz+2`$jP?G?wWtUaGj3UR>iv~%=&@dE4A7KBKtF;5P|tOixK?y0tqG9o{FGsb!9EkJT)c%siA(Heh!BU;n_rAnt7|FIVm8d+;+Pp*OGnrox-o7w05F32ZnyrQ= z%7Mc@V02Afe)24W$G)KAOvCz1_m|U<8nRNvMgtKyb+bo3`6IS``BLNeHBBz9hE-!0UCuGH?Uzie$h+!gg|D9Pg< zSo>4*X1=D~8^=D@Ejr^V5CJZ!7pd_M`6suicRV6D*{EnCeT(gFte_Zz%F-iW^rg#G z(bTMrEGe*`qwo{i5ycg`_uqk~6pt|e?38^L`}Xb6+dqr~u~7unsZTEJiSM(O9)UN_ zZF(wPCo}W?O4OY)8qLZ|D#rtDS}f|RAUL)BtJrfRtko;;dmxZ0m%h*bR1OR?Q@9&Y zz3;25S$1-~*spxPMUF_YkC}(5sl{&3H@`jIohxK#WXyRvL$yXoP0bFApjPg3Y-vfm z#`n+>Q)JqE`l&@nW#V&YUIjpRs#I`*osFK{XGO=;6^hnbt9|j^MlrVhwd!l7WHZ?& z6>Lvs;{?+0#GQfkesVHYN*C;J=#?c-BrM=8tZtj$ zA}3=v!$}puU}9xnRi$gVeHAh_;?pHjxw{trWuw`nbTar?)HKgT_ay>zikg^%p=LMX zq?6A#C>|}aJGZYUbSJ@M#I_B6$}|OK0A;NhS&+Wgv-MlEdxT-JVu62Q!#l(TISwaI zPlEp(*bWC#k68`9O7qG8iYD0bcGUoForQCKX&T zNJJD(Y|P(}g@xsKJG-L2d2q2lEdD7h(haL$96<5J?{w1<2s(gc*-xD=E)sR#O7-_Y zTUP~#5ELH$JP~pD6WHfR4jH*YfpiR#Kc^OUfA!->J0tU*L`U3!`S;qch3;;aY9F!I zX2jRywLD4RhtZ`uh!Pb&oX95mEW}Od>e{86;mF^6^=#uEL`PzWCb36wu-oU zRQcv7nZ62Qc-^3@;7IE3)hmT6;g1Og5qtbL+BR#Q{ z2e>A`M@ITDx17Y0Z`9PPfUCGhbqyx>ev})@q+{Yfr}j0%Px1wp--UxGl*b0EKzeAD zCSbE~Y3mG6p(GS81Xx(=C$Hrd-L%oWez;)sT`FHcn;>}hRIe%9G*?EX=p-yW{7db; zZQ-TxH1!BUwlf9w-LG+*&)j!W85tSGH#8Y}RO$eU)g%?>zNGS}kd$P=TiE>gxPw{N z(loeZC0o9uUJZ*snWb^Cqqii5d}_%$SMU0+D$G&9wzvM$=Uo?z<()K*P*?j>%G2yO zxnX%H*`RG3vcdN}55JiV-i2eFUJV`25jbh69HXF?yc+1oZsT|E4M-_0c*oz7dY zZMx4>|NiaLC7joYae`il=F27&*-8e~lhG57CZ2Yo@a&~0mYGQmFe$Z%DU}{(eJ??* zehhupfT{ex>X=C0i&YlS+LbpR{AuO0nSShbx~+T=F>Ylr`f^m0*e<75#2)GU+8}GMGT2|A)s_JDheM20=tCILkDxp_S(lZ9GGizSwio7C(Q_DZ?o6E{lqh+I5{_& z;f~vC5juQl;!%^A3Q^LxZJU+o5qxy*!%-IRsTyCIQkoRi!mjUt_SDbQ6Z0t=mm+Z( zw0v06`i8kTGm9w3tF49c$8Rx%25sww*YQX-7~;`!hiu58YD8dP>JnKWimK*U7XtpY z0`A8md=8(A@FRKpOa)B7UU=AL^fo+qKJpG(Ij~^s~kCfWOJitsCt86d0^8%=j9vJVk#~E9}Y}j zVD*R4>ZbCT+RoHlB`75F%g$fzGnC&PiLIW-v9_|Kv_gwOgj@9}mlw;w@kbh7zsZa1b=Fy<*8j0dI9&v#u%MN(C?UGMC|a)29UW6}ux`OSub!G^d& z(Y;5J>h8)lHVfY%E=4~wmI5M0<5+nOl#@)6DT7}{V6;A~9+d-EspzQ8Jlw`phopzM zFsKkN_1D&ec!eav*ThGDDSz}C=9t_qc--cNv%J&0Io!moY1ek#rhBGcwnzhc%=eY5 z*+oU?7n^YUNV3HktOnRls6uzX3z1aipw(@)(ujVU(D#u#kvRUT`B`*ovhLvtWiyl5 z?Qx|x9}|;|wNIn`udQT}xurcWys zf44RB2RpazNqYDtUQ}0FiFs-J-R+Edm`nvQhJS>XbTMN_T`BQ@Xphf8%P7%*Zgb zIJ*rgz%3ENP$gNpH<9Y$tYb>BkXNAtpspP(K@3e`!Fuw9IMr462Y00WE7us*)HgAR zl$QdY!K?L*Jw#c*ejKViX5)kntfr|Oq5n&HCg@ZDl4aOWP{slf4tyxq!#j{dT6A7o z{*Wds99-C~sp9 zd!=Rj1Ub!)&Z#5;aky%$Z~x5`Hnz8ba2I9N6+NtQK+}Sl}y4CWwc*D zr;YrVP8#QZU&xKaNm+n4z<~VOV8vryFNBQXYCraZ)*xv)^!E0a6=(R~4D~Q`baa$W zup~4*JTd{8hRieaIr{aY!ldtE5|3Gg0N6_wYmbx`AS)i18@VPx6Bq6MZ$cxteWZlI zhjV+dlu%$`VBUGnj}~T6pO5gDUO_=Y0t){6G3A2%VjUj4z(lX>?reSuRfT{+*8-4x zwC}s`mVOMy@N(IESb1qX$x)u`w9=6ztW8GAP;>jm7;!llljb$f-uQ$?#k#pucA~oR z@k-U4g-A$9s!W$ylG~eL>uI){G$Ru!`&iv3|DF!-QF81 z@{s?nTr+?R>T*8+2vx|XMGhp)wZuccuVsPE6Mem6aF(~S=d&N(dzcwi^n8W2iLQE+ zk89wBH2?TLX}(5f#He?*->B87+RoaQgp@)T6fv*ZdYySXcH6F~Rzqo(Y;N{|!f)Q( z1mmqniOx>c2{5fTnr*>jbhDu5N(o|Qxsn`9$T7+t%&}=aqcN4_PfKAB2*KwWEH${> zGoC)Cz;=Fex9w!9eMT$ti(P46pFz(Kp$K|^?2z_i^Y>k&-$B1e6y%`H-XqgZNp4H)F{)zth1{3v6B$9Trt6(yD4K2bE7h9td^p<&6gaWcKu zDgy2VcJ-4Xm(9uiO31c#>s{`qlU9v)doV;?wXk+R9s}y7GH-R~*S8hA^n`BjQ||H( zpA91uUk$>Ip@)33S*bamwbpR)QJ++iGuo4P7yCpE>ihMZOELyy$YEeoN*yZQ2nn>e zMG4LPa7Gw_Alk7ML+yg&N<4AsM3Po6&--F3841LXYOcSTa8|-*nB~?)}=6a5$sbl@>e5Zk2H_7oF`}@V6%@ z4X9ngZq!fnc54s2e)~t|=x4!pZvFeOiOqkOMRmW#ir_c>5SfX6V(_E6X)Ty+k-)=o5bx}d z_wpwL%H13&(V|}@@|Q!?30w04G|Y)0bbsH|({XUW1LL?e^7!<2N!E+>tc6y&gZFGz z$Md(H@Xb@tr3v@LW{%BQ?p(VbqAHtqK34@ya;{ z_fSdO2(6dWKF}M>ehILOMQ2>g{_K-WAu1s=B=68x2Sxy13)+N3+p`S-3wsn`<>yy; zjK2kp7y-WDDlMCDXRW}BsCuBAxf1Dfo#pS5V)RB%4&=xR_}xia)uMn+1wy*2DLizK zQeps*tfmi{G}lT8IqB&o1YNhyBYS|!F~s9_MtX$OQdeXG(6Sa9rAP-m z0Up%tf~seLOkz4s%uZX%YI#TobG5%kr2b{1B8HSIQwir24mt& z=K~<6etm@TK&2l{M3=j9711m9z2y1kGXu+kx6II|GmmN`-e!ZcoGxb0CL0Xk-gXxn zA)Fb!_M&IkXcR4M;$RpG-=)AP)?ILpX*c3mu+iBPb==rST)mZvv|*rey3pPU-7h`y z{cs4i_P4CB)dtF2FRW*cZI|SW-?mB=jOQ`lvyPB7ca2@PyU&?Fr6eVRc4N=E1YVr7 zQD^s(qF5m)4_M~^&|Y$J7V^sCP{(Pi$|DT&40%j{gs$$V>c`xeAJu!h%GduH@xd@NV1_mp_&5`~(zVXkRAa@c>m%da}jT{Od zQ(1AuJC|$VLeVLU(TlOm@NAu5(<9yMjg$(%ugFDAi-eD6!$zvKoiO<24N4ClF!eZf z0=cZ62|Mv^%lo(~0vA1XJ{I5icGc0TjjlqY@X}}9T@8)2o|4`aygJ<6Po=)Pzk6Zx z6|ThJbw04+x=07));lF^;8ZqVw&j!kx}rtcoHJp4>H_lgJUK~M0fhqtgZa;Lmh15B z`=;j3Wd+O9EiILo0`ZraF*x+>u5YO+?1QhKjPGGe3J+_c#K-~3r!aY?Hn!;K;i~sK zvD~-bLw|qpXzTs^KPx~p5Bp8OaGZ&UnMZ#Ag$rr~^+M&p6#KjppvQ9@_CS2 zgE$8v0T={)esSeSUA@4PEfZi*x z=l1#*m%r!C8P&8D5xPC#Gc0U@F)6GuhZc zH|DWPYW&yk`u$_eZ&S$HAtC&Iu+bJ&j{NT11#P+COC64O(Z%e<;ULe|8X9&5G zx;@YIEOr`BHnm6%iXNqZn3nLDlEjXc2>!;Lp8q&fQe$9NJE!k@^_Ba z3$KUN2qESal{lu{>~76Due=k`tAX@Z-sV+KQ=#m@l2Kaibkn+Avc3| zAAKDL;U^KHVdzR%y0->?L!r^`oK=YA*QWVdRGAbsacq+z?*tm#!IJ2bBjA3ToImno z9j6mB#vR>IH7`M&fJBES=5EzE{e5h}ZLIp))%Iwh@o;}>MIE?e!7wWbSB*5Su*f&d zgiWJi7Pw`;;t5O1O$d~Y<(D?)_wsgmUXsc00r^#r!fQP12MU?gj7qnpD@%L_w^7;- zSmk3x6Z(=0RC3(XTMGI!rRAyhH`}n*4}T^&Ucn>t_J#|_3dh_gv=;~{i!MdEDFkwd zeu^J`>q?K&k)e4(fud~n_PzL#XJ0MJB%jCTb9D@3hYIi71Gl494ojNNd>+RQtB6yr zddoPqLUq{f(U8%I55lctsdho_dl9cp>rYpk>1Jj?nWx431GFGp?u{F+#b!Sz&cZ7@ zo52Lf_KNpL7|L)B0KK>gU%Pro_h3wzuo0deR;I~!Cf+qU%nerD{r>wFrkjRJ$J-wl z5bwDl9FvJb`hVLkdPDC=Mk_84GexkldD&@P(fTBtEn_D(8avo!6i_a}glg(=!68(>;|t-JzgK>?gDGfocZ^R)yVbp0H6^LI9=*URLoWSlRn>t~l0Ol9kf z2n;Kp;J@(BN{2XRgiE+0*F$zN7~8S35=Rb%7IvertYzDoNFh4Wc?`3NKQ6{VO%jAQ)C zGX;#+m%dqRIo|D^ooml)*KYK$BA1ioIJ-3se2kt-H7mcs8$$S%PUPh`sWpFR*KX+l zwp(H=Ilx0s2Qry<_To!s?c{}-GS7)#URNJQetzvHdL5{Vh_+q3CtG8~r>$E1(i@O}qVipB%SmWHVw7otqAZ zyOWIFB7JFbY&i>|7-MZL->d7wPTs{{5)irfM0TK`3HnkMfA6AK4}p^L81n5BC8f|p z&%7eR2RyZ3D=RtDpTDk$A3m@N+=0kMP{|T-jgdxPvvyk@U_9DxqkMHbn5hswavu2Z z-A(_ueF=f%mqwKRC{q^BA>SB2M^Lm>y^|Qf&GY71@AaIR+vdyCm}Bum9uHv=xshb?rm2vvsii zZ(@@IOiVJep?AMv1A3oWK!Lj8g-A!hS0kN}-I+@3__lpk_A4<(iqg{lbx0CQI{pQ! zG#{*PJP`lvBgM8lWhf;~Lr2H%u-V;0QO{GRXKyu~s7zMbvH|-KHV92^zL%3z86gQt z{$1pEe~t~x#JH>Xa^rvg{8>x5!eFZF2_$h#9rS*t(=^Z3IX6jgQ2eHFx1^Vm@3ix? z0AbAo8qNWpw&>xXg607u$<->@hncx%mxb5Gp2(gp+vCf5GTG>=RoM<>a+=WAt*QCEGInk-nwvL_Yv4{#Rj!R}$aU3(7WMoi^eqQ|iyos1x@eYr z?=lvAXZ)8zW>rr#vqn<{ z$`0idXt?3AuFR^g9n3H4pWie48F}fqSX=>o)EQ&{_MWKW@{)`uA>2UB4oL3E?~Uey z;O>Y5$LdG6mF_!0kOjZeF*|zh9EPVFp4zzBe^DnVi5{8AVuC6&eUoD1Q&dwcImFDc zW%!-FLO>VHIbI6V%g2+P;eIRuqS5!_RhHGEz^PdFLqPJ`(EQL znkTWIWmfx1PkYFz#g#zuQQ3S2y8gR;P??)pT*4EpwNDGGiiwsTB51q{mm{azgVG#v zVz|q_1?rJJedF^L`ir>4a`So)`fJxL@qv<-%dysvoMzO1`sG-G{)$6WosGtjW1dld z5v!2TF`2WP8He-G-NVzpT^6w#zIibbU-5}A$PkJn#X#obynV0G)|^?b>Yt&Be3-mf zQpY2`=UM14{u$mMY(XwSMBG(J0eaUD1&lpLSFiMUS7F31$FxdzQ=ugf7d1ncys(Ff z4A<68PI{Ub5mlI%bM@l6FCx1~t-F|(*jP%s)?-!!gyWh=>%tr<(mD@GqZji&#g6*Y zdu}KkH7IH?)H}8O@{(4Mu?I!8Kb+A&b;rNAHgHez*fBb=B9KGgxMSerX`kC4z7>um zk90f#sRr&>0vYVk{su@(M~aP&&*&ljKooRSKAD{NujJgnU!@!q~!y)4F=^ zbUnI$`?mSmfs>M9(xeppR-}~3-jZ;U%+9RWR#yU#$;;b}{<0uzZ8C^`j003K;X9t^kVoe*NVi7oPo9rxE64j+t95C zH6w3?iqYR_4(ud*Z&GMbs}bN3E`d#0&(H#NH0a(h%i-+DtBT=3mSn(!^LYb0MpVk{ zrui+qy6zt@!PmksO&vRj($4YQ&l-%Hbw)56b-V#?4A{R*q8TQSR3Aewz3%_SUC-x6 zSFUH+o8!o*s@O*ah-{9TN)smi*z!HKlJVJ9uDOCKTXP8wA>=8X-sQd_9isAunMhe! z!ow={-G}@?>C}o_zG4H-)&s6~^^w3;mJzXca*D3%heVc%aVqR_xSBjm8P6_!Nfl`$ z)vF~qy7d7+x`e$yDvMd1LuMAh3ebni>wQ-Bg%6Yggx9KCRrd~lo2nG?8&q^P(VfsV z>%_fpr-musnW}k?E8?1|XmqR#DR|mWJbIk-Sv6ZsOjHQp!SsFDxd*E4a)KjW#G!`P z{3ZX8dgo1{*XgH6#%_us@U$+Gh4TG|!i$jIhBph$iM+6HE|@W8qXMlqm!QbPJdLGz zvx8I#3gk&8u)p#B%MvE;hWT_h=e=@vS(sYbqJ&zWU9UptHdocu{Wg&CusSazUongs zU-+bj-jix|nU@I*n3%5J1$~mG_A13#4#O%X4uX=X4BB~eFhAX< zx9_Jlk$m&=_BZJLlxBTb3_S*GkYPq%h5?eBi{uV_R&DKNPNP;KRVXvOsS*2WrPyR~ zvrjTUE{U4$(k2@Re6#pFT2#NF7RILu8czetG5K1j zrICx~^J8nGxK*7xP0c6c<@{zk#U_P5!IVsmh7y^eB zRoKrNot**KxtkC|#%1*_85!=es^?mnU;<;_J_J_q+XrZEGjrl7 zHti>PaYL{|wJ-0K(j*PzxQvVZ>DdWN0(obT+acN|XZP$I-i-B87n=)M_bw!yj`R7y zJT^$_Cq>~@X3OVq5^89#ssKa1%onk&PB`7TB3I)BaGV!4>~yo0dF-0suzJeOLeaS#=yZ5R+VXJw*D`O5GS2m@H$`Na5liN9cZ?un z$RT0Zt@V{srcp4N(m352VaZ+OoisdQ%XxYFD?C2HDW_DA5~4z2i`JlVk;bvkA zs~6K>4|te*2xE}cv7{>Ukv_5#5TZom0>G>NE{v5mk?Q<*Ga=zVG|U!GBgPr>_BxIE z(|6a)R6hF|L=^~M&)%?jTJ`{HOfe$2cGzn(ar-utGX-%^U5mT|vdnMdt%~_KA5O|k zu|I=pxv|ln$Dyir$3mAK7BAhC(D!`Wz2$=j|)8gh~;=cDxldh68Y%JLbQh2+dKS zeTy4EKes2nOu<0hoOA(nTW zKFEEC)G97bW2562KY9Dgqk86qwUT|k-W($eD~OajP4GMG?>^qYhY9pvI1~|qH@}BO zlx((zO9>Pyd}!nO-PqDn<99xP^4?0|Wpe!D7UcO%E+};UoeTfl-X(;KdlY*K_%ors z(l<6}6cqa-=3Ftc98(fh40&1wiH)D&0vKWU_u&g60kyIFe=Zj_UY$U3yK*i%wT^ z@C{9zbp@(b)<6wl>@{)>421GOc3o=@satdMw;fp=ZuHF4;ToYl=kLb*t{yE%NwW$w z=;(@mG+G7H?xw_tT{A&z!fqpeF1{&mmbznUc}3TF-VwubCj6Gb=dtd#h5e7fU!tq7 zBx>)S0CPsWLspzhnuq3wNhT5@mW&3K;gNlRy!ofWkEP{>W?Z7{Zfku}E}`|mZ}VA4 znuZc`^)H=>C&1?s(?KPSZxIYD)US>cabC_1)d1ezY8NDn3uNTF7va0sQ)6`cwYF1& zl}ZHi0RP6w`HeY-9~Mk`xA@~(QSmx`=}T!Q@!u%%9m-UofH}gIl~q)5a2-Y!2hT}v56VCPQ=P*N1nw|Zr_;a>O_(}D8@K&pb4ti;xe-u;<*Z%P*NO@|`Y6===nURBl0m$D!Z zH|e>Iy<;h9lvCIz=uCx=?K}Es@a2So&A)F4-@L`Ndo_x?{qD>AZR&+}`=&c1mY3T> zxkGb1B;!4mP^ucyM~e_@*pu6Iy|1rczivhL_G!!;HOD=RbjR{fLWm7OE6STnSMc>g z)of_bLgeabA+-?y*2?Ec0Lk5+p!X)F5wO`Hau`|RbudLVLTq>KyiUlj`g$EC!4gEg zs(jM1ebrIsWIqbe2zPEd1c?e+jXh-r#cncM$@_i|h#0>OI&+c6jUGE=DuQ+!fo+*F zX4v^~w#c4a!97q=)~oIuYDbm}LMToShbg{I19pArT;ufhW=ge@FlHdI`FAY*2NKVB zMjzDSDztOgXfF%jA2#QR=^9??@@w8;>zndrhUt5=3mBb9Q#T$NTMkx5yPSLalbu*L z^d>bcQP7hM_V*ZGrN`0f)Nvgkkw069@G2pXRP*W1XZ+VG@0Fx*b`5)5bImOt3|jrv z=uA{A>I4`XjQ9d6^UBrmAYoB*nnL5|Su+H`!`@=F>I;y22HM+)`@MKc_1OSV4)`-f z{;?jA0iNjTwzn9V4M`*dD)tX8LIAkV+Xsv;it{SA^5agR@(N~z&dXmAu@lZ)E_Ozm zpD>BtWZ?3l(@OIIt{%+K2;@5qmAX<+6*`(P&b>lB^L3h*dY0bbF;YVpMq64lsuTbq zYb8;}q1#V75o%YvbbGW2@*#yNmy{|$)pPwZQMxV~UZdp0VW=Do-<1o^xAU`BqCKYQ zW;ls~8&8)*Rq^IK>uCSFUrKu%u|JvaBT(_)85TOIq)GMFz)X5b4>(32!yzsN{vqdq zO{Ckn!~^HT^};ro3FiEh0UvMSy>Q+gE+Z!gr-C#?(KHn_XYEA13@zvFfne#-&DfWA>OSa@I{F@CA;`h2JNs!ylnU601+ibQHU zv@V5>UI&wWA_7)=7Qq=G-UvyE@(FL5b%OT1DD zccTm^tsmIyvFeISE>(j1O?x%l{anb|EkT1=4OcTf*Usy~51!JD|=dvqzIBu(c zgZ1&!h~UY^_bG8MdE_j}F~nk|(upOOQa!a;wTtu0=tcGp4&6N(m-OG(uTkK* z$}&^J)dv8!j)=IdpT7S=FXsBZS>)#P+6!91-ue$`!euXTCh*c^Wg}w3^xYILTBUkO z5KXaMo)mFgA>olnidujt7x5!=!D)NoO`@=EQ7A%OU(^-NHRDMUWq%SL@v{f@t?LqE zPE7r$m)q+tDep%>hx2L-Cj*Pz_cb1BQO60QdntJdL^_(kknuv!fCXm+fN&lzbP2h_ z+}vtS$Lnl2OyBWyzXK*E*e6v~wE3n8$kB^{3?PRT_7|X*A3R^lRmYO{fV#^DR$SC& zl>z5+oaop&TmuI^0Ykry;mb1&yKB#wARVG{j{$sy74Gv zBnmJ^e5ZSB!*V}$>a6aV(&36Ox5Q2NMW4H(fufT3Io+w##)DbS zDg1Y3Mfzo?A4TOhhF^(44hnrWX06ri>l16)E#e#VmE{Fs*5GK_509nMo<0Mv5xfZ@;#D9$nKrQEvqfbDCgy!K%0s%*W16wtxCGVu_Hf- zNv3nnT||{ed3>7>^qn!%+|4`8wr^O2LVd7xi+LYW;e}hY58ag2t!TH(p=p|!e6fy6xaX-YV0h5UqrRk z#j&-`-Z5-s!~V)F1o1oO0HCS)&h>f&j`>TvE-5Au$I~Z) z#9hilGY`PwgRz}0A~Vc)ePLW+&=JtK|BB{0)s}eAFM~0 zg}?tUrI?5@_~~vrumt=JBY;&(*Iz$Bn+<+_PJzXh62T8GLU{xeI*d!-zJI5wW@H~2 zDg&C!=K8=a+!C+*yP&^4V8Z64Q`=}3>Iv-Y+AomN;KdB-<{JEDLniRZ#I_x+^@@4Y z^6tK-MYf8ZD{ff>;l4&~nFL%lJa79B!yqu5X9qKU+$Q3$h6iuRYRz{}#+S^#GJo)+ zc~wcLQ#$iKdI}hA5=pJ0^WUHQ?2=H_a@bUtdQPd&f^7yB^C&Np0aZu9w$ zbVpLt=quxWt)94IZiO=r{p!S%vMMG!o3`Nam`7#j$!?hj8^HoXzm1eFa;a;IMCZ>9_*)`^eFTQ>GvFy-^=n0 z6CJ9W@rsH|-7IMahsp%W9+NC}U`ZP=8}U2*0-1l$-zXCjR?;uX>lw=9FOD7v-=DuR zR|p_}X7}0QaDP;)lnk-)Q?-~aT;=^gAd=h?Fd`Not`_zq({ybXn(MN5jv%}6_zDXDmW09!+oVA*gw4%?|JfrYM!YnH$Z zmU*5R*rSCU)&osHsBZT#Mu+h^jASX^39%7)Pc?tD(a+5+TLocNrJ&vtneXdg!wb>3 z;ypC)>n(U8DyT>xhZ|^=Slg*N-znKHua2TbL-=d!kZSVsA4YQ3WnzaNBNHr6ac-&G z;q^AEz!Y$7LHHydQs*Lt_u>6qBj|WfJ{k~Q7l$HgfNf3~emzbu&m;$rso$(qY}Lh? zSM|Wg9%*x@RQlXz*I~eoD|i>HvZTsS$x;Xmr`O%r;_lXe4{GvvZLQQO&20-d92EvY z`$fn<)~w_-n<5Z$QV=8QJW7iLw76RqY=*ZXYCpSQEe$?O)fxas-!mWk z`+8m_+DDj?p5&Rv)b2LPftCBNth{_FtXfZT$%;#{QC2w=s|QWFc#m%eERMM3!ihb7 z&Ny@KkjJC+n;tK*y^M03v*yyTwsjs?b$5T}E2b4YS=_J=q0!}I=Fq8;2M&lPJ82|6 zU6fL1?1O_mfE$jz*k~kizp0&6RXAbq23=0uN}kvsuU`;`n}jp93*MCrE#+yQ?-cfT zB#w4gxxbR~zEfl+#TkA-P?`n*G6SPc;O`TZS0<}lwWxv>-n)~gq7!^rmS&EMVon`$ zxI-y9pAY?`E?0;f+p8cbq%q&Qe-Zsil?P3JPx9$xW&g2YCWq!AQ`{e% zkAzVR5Ampiog^>?Onk?JUh!YbEJh2refV&xyg)TgQpHbWuqBfAl3vN{fGPKf{VkEi zIt@f{7kbfXjdUuwHxYcJBr4dzC7Y4<^B279+Au8RD8=Tg)EP`y-s$|GMVfFj;AxjLt&cSG%-9e{mQx?KbCw?5{8D za5!EsKNmFL3>_)Z9kwy)hiCG=k~rBYH~BAI-A$-l<>w=YWp*5)wXY|Vb2OhVM`a&t~rT6b6=sWmEMYc8NJ6Z>s+aa3qF=0j(vKdcBZ`LE)bn3B0Ub6ZV4&x18K zlXiG)mu7Z%?qEiib@k?Q&Hg>XougJiZp`b}ckgu(2ai(Yy3@SP&AX{2JVVK-M3tJ3 zuwLZ_dtyrkW@+aASmz;3yGq+DRS9jSSyoN0?idbx!~E=Q8^8dG2K69$zB|d)HV?_E z1zeLo$Go>~XG$7vX=MUA%U62$Cz%CKGs`6{Mfm-fJorrJsSuWJOnOw-o8sUY)J$ z=N-zTGLH_hGoh12pT!Wn*4ny~0y@oyTmP$yb1c@{E`)hEnQ0UKfa# zBIYDF?mG~wG2jli5C8fQ!cWPcJqCHgrrQ*JX6IF(`A~zcn{KIWpp9^=pRp%{q7N4z6J4f(W6)BwEMuhgYc_sNz{KzZs+yCo`9@0_?7Ai32 zD8(1joA31X0me?|^}s4rD=Io#;@BvW*R=C$xdAlOkya@C4Wrj>T4O#M%TD?DIbfz! z28OoLn56M10N0IFbW3ETAL6lmJn+JkTwD+Z`&m3tcD6lYZ9|lr#fkjwd%&tkeKwLa z6m*!EcJ^tA_34gzyt6l~+QZi`z91O9akP>aPeSM3(u-Xzq(#T4TU*FMGj{~rB8#~m zsqwRCF0F%^6^UyKJypOzR~L+u60x_9KQ2BmD5eDcw#}La3f%M)Ge99Ruaz(#-aSC9 zNQ8YYzqa<<#r(nZ@O-J_--kmm)Gx9J&;H2;*jY|!6*%#(XsELVoUn(ite+(LAT>|s zFOkwbX@M{F-4@<~rJ}Bri;U;FATU4g9dFOq(_1COUc_=PObge~@`EPI59^(?Xx03L zC)3pCWybpDu7$f-CI$S#GI~~gQUawYgWeW(m`^9=^e8&e=&Cl;cKxeaWyW*45*(8v zwB9+i2#FP&jPg0c-9GiwG3AMtCo3& z-4uWPd-9%L0Mm&+#0r&nZ?B@FDp|rj6QqXQ_QOMP+LcN}tFtLI(3s3%h+B-bn+scv za_r}P?S%B2iCmES`)Rkvj*q9d3Z>>tjqF2XGBf#he8Q5Dk|%o@d%JJqJqP*;k7Z^r zwoS_I5X9s@vjdk$oN_54?hzU%{^we*DV*)op@0C+il%w+0e*5flP&IYuqOBP= z)1H4mtJ=C~m^r#=zQFh<3i;u^Ve+>g0A@i!|62U=^PO|jG4C$XtU-s!^<3kyh9_O` zve*n$TbFIs@{w?U-yQb6V;?3e{v zdxW$8U?ciS3o&9v!&xYX<>7l;JC@T7?Q4oNY+OIP_ZPwFtCD?6hyx!;wm|fr`+kqd zA=5VBc?GFb^OG*$BAk5KH;-5bcV-og0V`KqzJFYM4_d(lb==O554GgT;uOPIf?uCDBFD>q&c)ep}R1ZNQTkPVfFaiqtVMny$s zf>y5iRF-1A=t@D*eTU2^8Dd)vnm7kUsCN?Quq_7W688mI)oV8$E#Ir=B6J6abW5P~ z8V4whb+*vr45+;Z|zLFb~Z2s@2ALsU50wSHjG`9JF#CEoK<& zY}$&aH)D&+n;G1#t+c9f86 zK|**$%QeE&mK}3S8beB?KB5fpSu>z4ZI$Ph(K!KE8{cfZB9D83?5Vv@LGPhSN0`u> z%S?{1qW_G@2=ePX)lEloTFXYxYP$;Aa4I$s^fau_mV;Yj)F`G6F=p z_XRZ{%yKmwR8Cq&=^liRYI*w{+85>J73=J_^v?A~ZYrW*O(DaLBy9$EENYdg3eVPH zWc*nLk}*Tvv9pKk8?uqW1Y$l~C8R*8A(fE>!jf=K6u<;#^Zw(5FLzG-H3%S@6WwOV zd~d`2TC%&Mthue;t=_Xfpb)*rXOY=pTnXRkPJQwyhVe1VD|$jLZvJp#@`%ZVl^+{d8F z`)5UtEJwY3|2b1{`A0SJD`a~Hyb1g6%%l(@FfxIg>`9foiJJAAOG3|ERNSIZ3YJiRSubXd53H-M70#!WG{B5E~3GSaps2+ z0+xz11;d#NH=q*Q{Jn3RNt}ldT*zrGbuo@9pYL$R0T0$>-Jp#p3q@1w=sHBOD`5LA zj+95GwL}`aiDEH;^1$D-KRIzymB?=yb(70znjLlQ_8j`vbGwyfX;6+n7Bo=2PLZ__ zfGs$0K&Bj>$Fufu`Yl&lsncy4DxSANR9vfFpfL7FU?7eiH_1M$P5oSa3I9@$yabc#(tmMLzPhfw{K22RvY|a2{y1l4^LsNr0dT)C|g%!q+}lE#M{$ z_#DhY&&fv)Nc{CF@#W-!p|vXwDb38`DwiQkk@!XKq3+pZkD6 z3`RZv{fTUeY+gzGLP4{(GU$ApGxoSov^ys8$|(hYZ{NyHVq$!6feii2LXIBoBCd#5 z4{WJRy8V=jkLgb6QLS%LR{xpX=Io>VG218~^^gY~Cw1;@_D|nNk>V9jl{^=;n|=J7 z6zWu9-6YN}Xaq+HuZg>fIjrUu=#)Q+Le@}z7k3`Bx*~SaJsGBN)=}hxjK4l<1{i^# z<4nns4H2n~^#`DZJ6CFLKLDHbk`Ah{K*?|#_3r6~wB3KC5MpOjM9x$NeTTm@c=S?9q0SBczjTlP9QhubAU)h$Oekews){&&0ZWpBrt# zgyT13gDsynjiD`6lHN5pa!aA@r7{xObHE#C10zIIb+e7XZs;ChFT!a=-ah58GXgDs zr4mcv*vd@EXEV;7r&X9lI<#1|Wk~|(=g#UVi3g+qS{sVKq=1B}JN-#uuZQ(Ou3|i= zMz#{wBSI5DNLr$WcBCHPVf|*n$ z_G3I|1h^MP&((@|w|o!8Jb(vsufQ;dpF2B%W2OjDN;L{}cyQi!hNj$SGGl1qWKCUOOvVAOn#?MDE;{Yjbq?6RKF z>jTN-XYiKD6%it9-=h6Wy^g86TEh#yCK9Kqs#8M}NR8lsfZ3!lFzwc@`tsIeOg%Q5 zL(DZ~aUw!u-Z8nd`BU0hRHR8xf`f8yn%KmNsH?c1s4G2##Bo@}=ZV(aUh`S~E5>p6 z1Z|^J&w>@t%n_NE^{E=gmU2|$Z&PT*y*H|TbW%lW*)((dqaE_$>jGO31nyPF(^FSP z+xY}1a)+TsE-WhhsGM(i;XwDi3Ay2@-H?&waarWXl;1!>bflyDSi#hT8Bfa6I9S9w z`Q(eX+lVad;?}Z@7RD8i0%G{y{;0K3l5Vc2j#=TG!u$>t6@pm>VS=w5;6P4uw*Fg5 z7b-$6;9Hm1yfDG|nD%tmzSDy0I**aY@bpxkugU4*3>%#=5sy7Gp+m(~n2%Fti zK=5b#6a2Bz%c2f@fUKJQ+mQP3w;)nEP}Vj5$G&)lPZ*uVBxy`d9e@`i?%U}{e!wsh z!DEK45}kr0X%=c{z}RJk2|835TWUxSh_M}GEuYUD{4RWR#6KY)m=@oor{8irO(l+M zplwUFl<#M%_ED7(nmJ#~NsG(PWvOn;LAINcyloHxKK3i0`whX2zqSa&_{WbQiy*r; z>8%(PP{UN24ShkGuuMD-@`1xTyHsP*O-~8j*b$RtKQNM}%y~uS#@C@VfdaE?E{UpO zIBxdAZpun;;({ykd=nM}s$?QlbYSj;ZQ__RKz4n+MhS@-G>*SLNEgPnuECNqPPEijl)XqkY|82M?&}O1e^LjKg zp{mCI0?*m*j5s_h=m#INE-~IW<}u&0;!0|uslyNVH9m(-8e#WwoA^ixcMD9Nvc~>N zJMV+H@2FB-T4}^og5fC^>%Fp2KC7wPVwX*xfjUV7Q$V%U<>S>g+nV(H;H_Ezo<`J( z|26GV@-G&Ng|8KB;~hb18&q*Ff?v1o2(q9TzoJss&&f|bDcao^y+u7V?#~|jq2^&Mb9^0GvoN2Y6UFLzN*!7j~?~DjbuIf7d z$)?Qmd8Ho!bQv5#w>HjB$CneNelYktSR!KL26@q{6Zw$)*E2iXj&w$Pa`@_nz}}lQ zYXgH66qU!`1=ZM`EBQmqbQvh$5JPYH#EM06zS$58Moa$DLS*CMy=&{)0VRI~E%oTah(P$Y3CBa?)j;T(l> zt#4knj|#ZAx;??E+{V;R{eJzPO?xd3L%MtStY^eH1W*`Yr@wp{Fz_0lnvA?J045eV zH)7EF9b@Q>3(wT@IOCqI)(40uFN&*;ylmN0S7>_AKGMk*m(sp;nlXyUQ7Bc?Br(qP zo=WvIXx~#>ebi``-{JO~8>d55Nqdu)X9&02M-a$%&_C-?q%^ICpM1)B?a#eE7S~oT z(cnRXO8(Qjq@+2sz?TaQ4kw-1mGrv0`77g5raIbgsLsS*Bu>H$IeIl#tRuaFdjrFL{ZDmLoNJWxl(vH&TC#bCFkBbY^eO%ip-_^qhnYWjCUtRW!3|bFjPA!yK`w zx8g~M9h8op9aK6JC-`{M{Zt|#)d*QNy$lKA6#2GGF1~0WJXAa-f z!MgTi<=?teTN7?Agsr0cciD3m{+byZb+(=@s_!+d;GLJ&7Xveh$kq;tX@7ViS>uVz zWkXZh1qyc&jjt2+lUIkla`$NYH8jNW<*42*-QA%0PN z9X4Mw+Md+){U=+WKl^j&v-MHGpwPQ~6G5rI)(hH*)?Elhi0@ATgx_LWZBb_#I1i?! z`7N~Edan+Fu9_X&KRwy9ZU1^2`qP2dURUjmtpChPY;eJ2Oy8)VqnKdg%uvYq1XV^- zZ{Hj==E57Ihc?049jP z+8!5(GXJdf-oGxqbC6O540(-5t?bPIQc&#iNj&vgMz|~&knZoT3K?9#9)0=L<7=wo zNPx_cedpagy~d9f7IQMhl=Rl$MT$0N{G_R?*YE4QZT=Q@FO{jr<9*Y;;U7o^!=}2i zUNBgfwJCfePB8B`W}O?ee0YMTcR^0f_9wyaWsQKjJ0q=ak|d&&Xe)+Xk9s;@zh6$Z zQ8T_9W{P3PM-N#&l!9Ej_cth;{DX z?_0vuV$bV52;m^9U;IqnSLkfKF=SM1&ZQU=_p4DEF~JG30igb9f_tX`u0 z1euQmYH?7x@|Au<0^guwB;e>*Wce}Y+PcA|cGy1G)xWzsG_sS}T{!G7OHEm&n_;qgDb z$y&NFD;Q4|o*Waiy&$R`TSVMYos7M1lT)E))q6#hd=FN_Y`vGL%w|+nx6Y~2@o0U& z%wmGt5>;DFMlZlqkMzr5sR??f+u&W8D3K@E{O0;>Wg5r=-?LXvW?B0nQ|fi7tJlnl zqoz>GgQY}#P_7sNAw=HL+6OQL0q(HijM>LZKe4N(sI;m;&PHP0r4*ZX^20U?8s)@V zwdG#@x7}(vj$W2?S!pSXDvsUIEiwCmjA!|@0<+%rVekL$OBqa->69sQGz(IfDC{&+F;3`_^_z$oG{HmC|}{s}Ju9Vmj5~4J7cS zuIcxR7!vx;fQ-~YvymE0mHK|8=g1TZsBO6cHUE|k!dhs}dbHVyFI9@5rH=w`&j-Zsx<<#(%Q%*IZ$R2r90bg@f4=2~Cw^k_$8#N0w` zn$IPn2U_vRlL;UB`pBeCzEUKm3FwS|li`}EvMxS7TnLZySc{_K-uO1t?lQlXxwf0> zm|s-pdA3~8f>hPTORK3N^!T!Ir|j*h=!(1RygZcI3Dl?37wP??0E$%rWoJY{mv%8R z1q}z5TKH2xz}7wH&cD7BX|~%nh9oe=Sf8zV8Wy5X$U-9)V_sG-n421qv4TomFjM_Z zebxJ(_~HaLR5piLCahytBge6Z%HikTOOTpGoD8wEvKVYd2qb;`PZc&`PNnJ3HRS2c ze<&cTk-#f#rrb7TlgKAK{lAzLIAyJ%Ys)1d+R%1DFKlqx;2G#`U#Mdrv3L!bkj5#; z6_w#au(78%xJ4aSKOA|0z7_~6V0NgBS&L6j79zgbx;jQ-czN+qNlKDY-H7FeutVlj zNwsdhZ*+sv+P#VT3;!uzlqgqjOQb|kOdl>1zWJ=e>w~>ty|8}w!?~9J>Y!VKNJl|c zd?h>IZWLKq%7dAcWk7EkmT*|i55tUC_7j$A6-+Y37aRP-yAGMI_go+iDj~~Y=Yj^`u|mFO zr4&Iq$PPxsY%Sf-gQUiJlD==eFX+9075n}{9BQo_&PJSq{ ze6yo$y$^Q!1jJt7(>16jC2bu!cOW=WiE|vIp^9jVMl&?B(Qd!d^~^kPBE4{r_U4+F zTPrbqsBEecX%Q2Vu$U5W-T6&x!9lmuvoX&E;-7pArwli^XaWPCAP^RBaNqcHSGQ#p z!Du81r5X6{RCWp*tw8HE#>gxDTM@_^Tn(-?I zUE~(JGRkg@XPMoAQ@;umbuA-DKHRs>Y_Z02c|4Vm#>X%}t5?6Gw@Dc3VO^4Iz4e9H z{JYDIc@5W%N$YjH=|$aa3<7vko{Bk)-#sd_v;HV~E4qiK?hiF43wdr4`x~Trut9c_aM|F z8!hcKmEAn*Fx`afxX2^y6k=<7J+*ek->$c$qoYv zq2OksHTGWqsjk%=q&%|L)UrnKRooy;GtW%4uyI!``?(YYzA=nsfK31*pdrJ zKD1UKozHbh>BgibDyw+JVnBA8L&yB7JZfFH3R&cG)P3v5ClOHa7(%8ArB$DZRdK4S zstTby9z96hhF2#dd{2MN@m&cYYn05XCZR@4+oSgVDtq2EZgTNwm<^eviCD+(|4UAjcq0G}R zqeDemF*>8;YoC&Qz;{gmtgrlCR+a$S^ZzyuJ|TwRydjz-sY+eB`k-&pHC~cHuDi<6 z>!FBnf&mWVHVigx+NCxse;;eZ2(M}12ZhDg}+zL>QH+>IZesN|_oB8~N4j?)%%i+>$ zmo)($^FCwN9_`}zbD53=)Ff<^4a7`Av@4o#gsf&%*V??_@V*qk4I3;exCdmhh(v4m zY;p($=N}$|@2L#0bF7g?CZi(yzy=-2`NdwD)f((9plwnfU8f)K_tJ`_8 zpxx+mz2Ji=B1J6)J*>#%V;UdqX_iL0x>(*!W+VH^z@Tan9Q^5XK(>o0)YL5Ocgc!Sn6_!2>N|5p|oz=9G5wi*46m_)d* z#W-%G+Y)O-4aHJz2_P+nSN`l0AHZA(zy6QV^glnYn{l9xS8HZ!5QC^c=$d}s^MLU; zS$`EC$Ra}iWD)1U75snw{x$z>$A&+(2lW4!xAymQKK*mzi)sP7%>VXr5aQm}WvPGU zp8tEO@MjhKf2aKa3^e}%!v6od#jOAqZA8o9tH0mqZE_kgeIudeg&htO;?A2QaO?ld zCcrhr$bYT!dV5FMoPYeR8ne2pi984YB;Dc5**iXHzk-rd%4p7!2xdKAl(Y?;UrvJ>D0Y9F;>m&N- zk9z+PyS64GmGhA*La;yILy`;^<5EeXU_~x;WITNMP(bzMzuYWg!x7Veg8?bePIaKq zKS)wB#d8_|B5=j6+--5;6JreJ17<|zjloAtZ_Uk9kMVH}W~SF-CfJJ1l;wa30+|QL z`u`ZM@*N!Hvrj91Got!=a^x0P@d9rhAL;&hfe+N`Nv8+E!`2T)y_>TMQT-VbgEQqUC zU%&7gV8cDl8dLCGaLhVTqJ}o;8Xe>Pxe=T3vqfOwq^P5l;DV7*)6f{+d6}bKY+l&x zIIQK7|I{q+)Qkgt*50cQ@S^AcTRKJ5rYd#u0L&6`+HkM%a4k$@1K+nLECfq1G|acS zm^W-5{PfIEi=M&$gb6IeRp7wJqF;@PXujj2Net$BRP3(ozRbZMPq}VH6*m>L?I%^m z()%5MGzRPBs?$G;!H*yQ=T~@>{n!?RRgx<${cR4}jqTA*&vT*T#wVjo2R0@jkK!Y4uQdl$TiAv{Bpih@mg?r#98_yJ4X39NgUSjlL=Xj1RlaEAv^(5yAjwC`(E9PF$1-71Z%J=*{Iq;V3Lpu&#E+IcE z=&#t-^-DHcSUmh_!B?*Sc8Kc6O$s910=YHf{=Nape3Mnsc7plM!L0)-;mgbJxP>=A z8TRWDF0*2g&WUNcHr;ER!l1H=z6@cw`CkLc&eP>A#bv50qs;^W`HNyq5B zbkLNh=BeFz&3nMGJ+y0#+(E0>shJ+}?RX^bu7c8wx}q&J2;~2ypgsK^sv4fv-{_M5 ziNZnl*|Q(k`J`?Jb*8CDbr+2kedT62fZ){jvsaP#?c&`;F~8DNo67jMknkkwey=oN zw!${A&aV%93SnjK_Li3{XTLtw$bVyfRPqA0Henx;nAL--wPdlJ*m*US`@E*%d=sjH zIXULF8{)rFS06QVuwOMpWnE8nC0-s8%&}B>&>XQ2OX@C~@iGNbWvBlyv&^mZH-PBm zMeiw%#Ba&?ZS z#8A05ef7Qe=@Rdsmo?#B&F5ovUnEn|XDJeHDp<^%{aLG&ZPOD12t#u-6Zl{M1rM(a zcW*)aKmc=C5GSG7pM&C_-n%LLv1ZX51CNZoxr{Fc|jYikHodu8EPdAIWIMmFo6aIeseo zF__I>JZ1IXuHB;(9sz##KC6KEHB|A6gtR_C!DV$srX8NlLO@I&3pMiWh?%0>unl+n zgT{$KTJray(cO1bxU72Ep{qMK-xPHAkk%fgj?-kQ1FauTV^al!S&THW?T9btviFne z3-3(%$ulog2wYMjj(6YkDeX9E+L$` zmWNfo=Qt3~C+-%gm&bvw6nE~~_>!``QcMyIort+mUlc*hif+c)%RnM7ngK5O_mb@U zb!RAUw@TiAfC~sq!Jq(j)TAYox92F}(P(3)TV|7$m?#gW8k%^=s(qyLBK|^W_M)L~ zk6!pCoKAu>F`lz)3{#(}i>A={y})JqBbU_99nouCaqTc=&ct7#hszV|Z|( z^;0#oTU%gh{tFYb#q#;dZsRd|WJyU!=3M$U^N1?qAeYl6IKnH_Fvo7^Aje36kGmG> zkV#Mns_5)*{5-X?`;wsMw!BNFaU90AvM*xKGZfM0!D++X+7j=ws)Y8(2nwAQ}Yl%gO2e&{ z5BB}^&lAY*QvV`^$AJEL=MAp7$PYX>QWk3H>{quV&Ff=EUakn>gA_q+;Lq`gOKheM z0x}9H4(s;O$_{@uQcA`&Paj19l2p$yH9-=bXJS#RHin z2IY-8zfoUQ)q`IbHje>mCfX8J*Q0#Rd;vjuFoio=$U{B)Ya55}fhk#)LG{eE6#)u9 zUjtJ*)|N3Eh?k~4CN~UhNUK|p8HhGuait_g4BvCra! z_M0HA-KJ&pOfVWe8Y(HE@J=2t#g|ksRk+Stjg5X_7sY7c7rjJh%J1Ktp)5ATY?!;| za?JZw=fzA=oZH^6t?;RQxl|(WGyMh-vv<6FW7Uxyp-4O4ZvO3#-9XlLWTXIhqq*l8 zO|r)r3R$NO%P%_lb42ZC0H<||+#Fy!%Rbg!KS<33W@V%(ABld|x0Au!?6um#|%!txhW)Teb8;6YPrYX@9CZao%*iS5hkmybnZzH z%sH`l9^9PYD!!j^TO3{nq=u}eD^2nP(m?xu#rZ zy1p}>jOfowBfw;d)c>Cnx4UX6O?-Rs8X@XX`HQ z*(szHbg_3@d(t4>va+%e1mUo+3qz}4$SUP#SmB1VT1>S@%3*}o8>}h@<|BA#KEOxK z^A9iD^`vMez9~I2FY)JqPE-TTI6h^sTG#MJdDYF!MwDpbWM>%H^LOx|N9 z5l?aW@}81Xoc)T_!+aEBmzs6x*sikwL`C2&D?Zgxd&HnOPSlK4E5Et_wQ^4n?|qi{ z@1!0jEeWGGhJHZde*8R5^Esnl0)o04?Y2uEJ>|>zT>_1>=zly9lI-tJazT@`e9pIS z#sqLC{rsw#hOyfI3{NO63oa<^GDjU`D=UIP#&oE}+I(B0(zQ3w!*6dQQ@u6XQF^pV zV0gr7xbXh>BePLK=dGfFCeEwPvFoP5N8W744u#+9|H5Mo2w;DF5$!xs_+1kD=U{|g zf`7X=2P@-N4Cwpzuyf`rn-}xiyC~nj1zqwcSr03duf^>DC=yaxD#dY0Ne$0QE_I9o zrX0)fUsQF{+}X+xel{oJ)T{P0wm5D+IG2a0Os))4WfrsJH}@qZRY5ay&wJLCMfzq| z=^ahGW^FVYY`E{e>qVLSxkfzU@aT^%*w`%+Qp3VzlOBAWbGQQuVUG$5PEdP`oe0M- ziO+sT0Y)p8j$wGZ#-18z?7k{bIE-jdmNiHg+o|g{1bt;;SFRi6hl5JL@CNZ80-%&;&QNbbHSL2r3#sI4E3tHj_QDq510FMksHJr~=FvLPD96%rJaJv_7XBH1h=CeaTdhPFGHl@BX$)Ge{&S)NeMe(&X4rMWS zW22o~ND8il%W<-72nwD})rFBjY7fKVHF0rqA|fu4GS<}`%!W8cMbmAmMVlO2nYACT zy{Ka%Rdw}EubNsj!ax<|0E)}+eI1olNWjC& zH9akG`*?lXT3&NM^9Hiwjm;wpf0mp-aK95U&>ujKAfuuM-BGqo6lXXDg$xf8q+CNh zJ7BNf{S0dFNG`8D=`B@c6fFw-tdL3>kr6Mf+h&mK-jrZky`MvQ;aa8pYScIBWYK`M z{^HF};Y`hxsM|rXkw3=MjpLU96m-_^A0PqjRb80%{JfBd92fP) z8)$9D)F-JWtJcZac~0C7P;80Eij#`Y-bc- zY%?(IfU;RXrwbX!h>EgSA0(HI9|HnE-MM_LgyY?bgLGK+saYw)O6^K`Euia>+1}dh zq>)?UNHz(lVE0otyZ&(4WJJ<+P7=OhiszVKa;l9v$hCLhAq{QeIC8J0WGgR)vp9hY zBeN|6$iuV>2yNVb2CamdhIHPoJ%T!Zf;A{>j(|o-(;6QoFqqHV74>`|cjrKQMHE|N zk$Gcd20I~TnGD#GbdTr4yHw*gR?)g$HJ2aO?Dn=x%y_X0eaUQV<5o9rONoij8;-?2 zKTGbs`5%qhU(W4H3Q}u9>c9qAL4EfIacCX#Ep^1 zm2UIH#^=s;J^x^Q@GwIof%~za5D($?2RCX~$B`QK@#E?1d$X@k=~v3L_3xrOj6GWB z%n6rEm`l;ZW&r=YeExsJQhw0L(ya{!`M8%oPzjSt3O;MqPZva7S!wsrU)v5poH*bJ zOUKv)zRyC~Ma{T1O%%KAbgOkrj81x6OYwqeA`5w?3=rGwEj29zlOm=-fFwVTO~qPB zO-)TCfITT}_U+ZKgw6gTe}~Q#c_D4!^gew!byvfgw<^tHw79iBy4cYeuubm*!lc3QH;3|qPYT5m^-6rF?4~F7YTcWPeTg*&# z)rK*-&dwQT5HPRb*IaS@E^Qr3U(~E&8b++BtY$bU|HXBu<^;0*kr-0cc}01))7Yl^ zNIYxo)2km%O$m6MP##7o3DzqzYUGk@9uWiCCnQl2m zGDwIJY}%eNWW)*^RqxY$Bxq9r(6RL^yw;(SKgX}St%}1T#0d03T9NMk8j(83$k-)B z2TEd+e-FQ`;r>oZbz#|}N5DfSp0I`Yezvrij7dyHT6kHkho{d1!5lx;Yy>`dc`(kE zx&OzHdwb0cT`f~xGQ5t%B_1#V6V(D(HPww zF7nxmeP%`~y&b&oAMm2x>$V-HRk$*kL1MVbK#owA4R!kYxv`Tr(tc8~P)n6WKNp}H_3`uLKQ%_PrtvWM7xPB$q7YzrkT4kWWyr#^q1usXE?2(L+H zyc#foHMr>h^w%D%jBDnrY_zP^y7_@Jrp2*(M_6k~D>|KGI`qS@Pf6s>efHoXo85Ii zXK~%2AqdFp2H5hicOOoeve`bIZvB{*oR&yasUnj zb|%kZD-kEwU1sab>%29hp`>q|12sO{>o58}+Kb;*_PHB)0B@1&S_?_i7)0*2A)##H zxkNFgMHC!};dT-r3Co4VaS%~?{8^j?M~es?x2w+QM}4`ybF~Yr470v z+K^5WVHUiz60cvr zX2K)-U^@QrL2aCj&gUvhiTkm)Iw~N$Lpb(S&4y@9;lcggRR^HyNh!;C~O z9-cj!)m8ic@Du+|%Pi)8SI!6~m_VSV1&zKRv0(1kK&{5s>Um)s`+NgAR%DLH`KeRU zGk{yIP5H$+hIqGQruN3c(6p!hv0vpehlt}a4a$AqC^;xNKK62SC(+DM&28tcmcFj| zZk>ZQI?3(F*e(E>-H+R1o)>tSjBBK0MeHDGhfqS-&z|St5W?q4i4~c^=WQ1I;nJ0? zH6##s<9G?%!t>sHx9#_iZ;QgqqkC%Dw`;1ZJ*K7@00c(J6!pc=4~$faId`Ako9k1$ ze_2bK@ntX(tme-~nUhmdb5iE!6MqVtZCS(|!hz}YWV1k7ulj=So4kRHkQu`ai->3e zQy#$Dx9R1}{0q=2B?y?58J;>5{1KEw&|XW1>y)anDYOhIvfZ4CyZP+h2X<)%fL+Lt9uYdyVU*+H#eM(kaqWN^}_(y|@=ekF@Qg6%s4{?$;o|p9o12|=2di}&maW;7) zaeZK*!zU)5Apucw*HXvZBZt>=8j{@hY8VFI&-Y`MttWdnbGIH_4i&h@K%33vd1i}m zIgIuTa0rX^?vbxB(@^4RAW`GT4^F(8Xj-sPv5%IVfrgpirlV*=sLKSR7Mp8ci z1mu4Wr7OakBnD5)X*oBLd(6mLFhS?jCQhW)P zjp(RChdmdDyR9zeapksvPQdEamv190;XgUAwwLbzyt@S?<)5!V0o2MRxrPP3Uzdc; zs(wWF91-266mx!xOaTz6X#|}W0mwh7SR+!Zm(@!Yla<HGBC%4vCP^JIkh1xl2G{^U)fET)=(*aaQn8JkU7Lij#+5yAU|k8}~)^=bARtw7AofUfUlAL)u?F+w?W- z=FL;|!{iVLaUVMepG~#fJ5}LBD7l74_{ynRX{k&n=fePH12C9-RY4%GjP{e+`w;oO zH@EBh2zD%G;j&;9M@d_Cr6{>~63W)gV8`E=fuUq{)xN zwmJBm_4booI^nyxhKLCgh^#k$d+0`}B}E9j#<-QN$YiAQGZH{V+dtW+OLUxC0Ydmm zX5%TnXVdTwV54wxVBmB90?e{1;Z$(Dh$9k&vYx)?04dZo^O=ErPIaK}@A89R=Bpk6 z&ik`5_BWgd`g8))Sm5o7ur7I_k!kmlL)MdeF2~=Zm6!Ygp*%s)%3lVr2p83ycN#u zfO?bV)2}7tE3)2C!6bj7eN0Yehkz_|IKcdJ`fhCPH#Il6fjPwXLNIm5Qi5|BE{2|H z9nd%!1k$n&;jBQ?rz7BfDtXXwLcH4(49T|e)CQ!vT;TUi`e`#%m|Lg)>fLIqJ-2$R z`KqT%(+4(QPEO8_rM!{SHepw70FNRj2db3xC-ZDKqcum`aI5i=k>~ev(To_+i!tA% zSHcX;g}8My?_Xcrxk_YG`(GW&A1xW;mXkJw{bgrmrO(0d1g5*N$RWpW5 ztxl6FT~h&ma-7I0pFWd_(NgTA77*i+N>#x2)Hu|h=!ADojmHM0%|I>vP~m$Xlap6w z;YmvK7jnQ*;0bFzPWr)Q3rrp8qL)MyRL5@^^nix_97ZRkttWd?=jLYiz*BU;?M93= z$46`P=TJIUiCE;03}`+0vmSinJ$+6Zm`>mc6TkntlCw-9;L|Bm9&DOoZ#jIKzU}OW zUqp7^A8i3tB&lU8iA93JH$l=`f0&Q4FFJKXXH736qc1 zPJ<-i$B~CwKX|#y@-oG@(BMYQG=v^U0kb*NX7J@lj+0u?FYYheQ#`Y`5g&i;y;kzW zSS;<8t*9R#w+WfCJCu!*fe>X{s!w;qm(WoGNK^_sd55fxtHP`$_CdrQ@BaYsF1ZqJ zO{TStd!oVtx;Z!`HVAlkdXKj2(#~x=q1t-VnyWDHt;Z@qKQ1|7N8Ei^eVD!HH`}O@ zwObF_=)I0q2RPOYWNBZtjlQk}`$sqS$%&`~{p!D4*X!{A<`t5qNVE^zor-x@xIgy-0CN)2>|+q&Qwc-K4sZ$kPJ`z9UuM9 z2pQ;#%&{ATUiLZl;UjfnCjcC-K{^&L4PW55o$ES~*$%!WVN@dn2KDmcv_PfX-s}L( z8FtjJudnxh(`G2D0;Dc3atEW!Q@d2GLkAP!n)4|k=h#D_hNGOy2U9Mw+{lMj+BJ7gT^WP)oTbqqD-;qyPUU-sZ11^tHlV9xZ^FU=r;H}bD^>(vA!9Zz>Dgpw z&a0K|lCyKew6axKzq-nHpbfZes%3@qL?$VJ+NZs}L?&Bh@zx9!y zTC<{JEr%^J50^zh=2RkOvuvxq4tTU80bAXk%-oGJFO{Arzb~29ozf-G)sfl>vO8`U zKOt-OGuz4-_V zzu+U`4dk#$!Xm?S#EUfbmW#TT!jaY^)X z%>gmA&@FEBGAOX<&Ahh*#k6$pt1fzV_GKOuzMi=%!uAu#SCCO&>R0m|t-9juziSj( zbpan96gwqmH`>3~6#1y2OMbDAcyGEW+e$t`%xQc0;&KY$$Kir6-TwHk_V9j&W`b2U zhws+VMI83lG=6CA;mpF^yH?EXmZ*2Q6i`c&Dpu2#uFtty@x0r!RA0{b>vhs^kMHtXT z)af~F+4FHE#@sLFHUowOCSIg)zhHQpHAgU*Mc!<2z{w{1f?s;1C~5D)prPm7m|6yb zmQCzxvm`eUck2tS3sER^G-?|wJk6m&T!+W~0j5M^iQBzs5WWauQ3kM{$f6h#p|}F> zR8XD#-M{;f-6tJM5Okg{C*V)tY>$<^em(ZFEk%G#&ufYaU-LdnvLW$KcTd0ajrSTU z!hJ)bIhJVskBEcqHeiVpu|S?|=`hfpHjV}eZNfR2ZFJl&mlfJ$VV*TVawIX{%W=P~ zcsSDaaTT+-1!Jm{`a(LB;iU_M<@OJ*2qg_1U$I9r; zvVG&b#3GXNqg=c-GADrd;P4PR}4!s zPET#FW&;z--hLH=Tz~UMB?X(~8j}xbZLbt1m9DCB=!+pI)^{K?hkKi5OUl;vg(!~O zb{|I8K^x_tUBGCZb=!XXEpG0I=gx(5&AKHf%D|a_FFy0#libjvT|`vqL*nu! z$OZoW)tMqBUwp1JDFU<)T@o@Wn-nc}S=9Rge38ltts6cOZ)HTvx;ZO!vp!YnJ!4xK zCxq+Dm3jfJ!svcce021zzQ=a${(ECKV;oIC4K$)KU?zS#}A(pN7|iv-))GJs-0anw$$=C@rfW7jz84V(cy}{7#m+5Y{uz* zJpLv)xM=wM<0UD!4*I9P@?VXtD@>Lq3=XhMmn@TBvv(yYq2aQK`!O8yzIR1LM5TM%s*&R?*Ok#RhH4qW4IIPf}29SHlXxOV*w$m zeq_6TnOD5+g)L(%asK^JQ>UcNVB7TVlCznbx+MLsz9R*-BD`~H)d$kBo2x^SmpkRf z9RbZ|8MguE-`?D}#j+zqzl}oM|z|z5n#X-jkn)Tf`@&v=ds! zCD8x<_om;p@9^r4%FsvK4DY^aDMX)vC5pA;Nr=q-kRY=Cr%pDipiNuG>1jap=(^3? zl?p$V6>$xu;{4TWrh)` zB8Lx_Jr|w54(1L1?9Bde-Pi%$1Q0_<#P$`)Z~_F8@7|X$Urts5%rxpGAN+0;2HD_0 z1)%?B#z6b>mgrMj*2YF&2~sog9g|JGg19QCSTA8w(Jlh$lKkDtG`X*M(u4KCOd;}} z{r!bhyeFZrC(HFaCmjm^|HF&_@7w-up7a0yl1v_3ugcctW_NBQd{vLGnWGsQ?l05d z4y5{g^)g^H_{gL5PUXUbcq&5$0e_W(2Lk?XXNebj6}%wXe3zAJKYjXo_I8uc+tvg5 zu6SqyqH|f&`jk~S)5=^$vb~kNsXY!pn_IHuy~>zc#p};Wm zeUdNg_hVD5uK9ENr+qaf{{05|R>7MgWljC^!soAdbyrfhhDhT3-{0-Y`}R-M`|GhA z8AKsOA)lX+|LsZt@5BDNv)ANFA#yg%C%YT^dv}=+h*-g^yPjG^kV})Nx!I=Gi6Fms ze_@@xj{cu_h?F&5$Cot3hwUU?1*uCZhyh2!ICMPR=?dd6`RoC-ca>i>g{-?;8v zY}LV>qm%yb0^fr=5NEE}ja6-#Os@TGz@$OW7|a1i!tk+c94gpu`MFj*|F}{>aCN=V zTNB4v;`H2%x$mAo9s|Tj#(7&GLWnrDm}P79S=+$Y3YC)-Bx zKP;3(?2{4XbGl;8ycW1C=aRjbvoe#Zj^y{@fK09wC@(+(M;B)v97tQGz!!QOn}S&w zvGVkHC53(;_~V)13;$r8p^pBP#NBI7gMaM|UuH8S*qeX4qBPEtYDAE|xpM#HJ-}%^ z1_b{e0TIBd09E#?ARt1UCusPs&QuXUcJ~!kc}kI02UTF>xgAOOSB8dXm+PIXaWzH_fz@k}Ed%azE@-v?=5j`&~9glFIB-*^AN-BXuQ(H$%Dr7E|niU{oQiTzfN4^zn(bKK;r*y*waT(g+w|kb^icz$DXMP!NH02+)zSZm3wzjPe00(54NEiXvu zJivoD^s}%~2H&si-X@4L+??%Tb65vdiAtE6?o>rFpWU@jHz1+xB|yccLgSXK*Os$m z+#<6LA>Xvpq5G43MlXlF^j_-fQl5V#2(uez|P;byFP_09f7&(E5|v{ zGWlr=>Sn|N(nT)CB!X5+Xh^1gca|G9DiH8rO-YST);peQ`BvU)`^h>ZB&QKH>z z&|O)94}9g5@Dvjb$P?Z0HAnkPV-sMRs6TPqw#N#?%ImIJqSp_yk|JrF)l*X6GVofJ zDc1-_;hSO}p8N4DRM3Wxm5lqERXs-=N^y5MTLAo2mnHIY6{>J^U=qKj46AqR0#hGY zbM=jC^BOMI)!q{<@f4bI#g8Q4cm2J&gv5=+YLyY7v@iubwntWD;&0hL0mswb&)lVz zTj5x;=g~9OZy)nr_5IV<;B@L;Zvl1kWXP3o{60Pt<0Or+T7o+&xK0k?&To+KFqh)z z8!gJ3b>^$nqG(C}8}Hv2@|^fWBEiTch~)*nh*()w=Lca|l}yCA0;JF?Ti>aL$_R*zWIgPit{G5Z zSwPSHykcn&8Y;4+g9>MCpx}C#@??s3N;D|kv%8+RI*=0sG=N$hr|Rj<=Kvv3f4!^` zFJE&^^G=L`IbtA>p`Xi885%6NJ4tHbOMEX7^&Nik9@lea>@u(X-rF`cZWT4$`5*(C zD(=C1xK#Tsz)C!P9a;5jSak{fv!=}c(;TDGIFQST`Ppjjq53Tnla;qxwb^+&s2=MI zn!$)$rXP#eOA-!t&Y(XD$PQ~+YLAh*y*dc|@L65y(1Trq3;)qc;;K4x23fdl0){y{xT{O?-hO=gMy5y9-6Y4maWard#OkhYHgP zfZN&m(MQZ{jqf-P|7dVYr9*%I=B$LoF#h1s5T8F%-j$BYBC_z7*BR!Eh=NUr!5$t19 zTd~i;=X9S67Pn%O&>=(;FTT5;e>$IXo#%8C^Y4%n@%mpv@h$UKrPwg!9 zQt7jT3y=`Q6gkA`1_-{$XH?+#$t!I+EN!S>=t_k~do1=}GhAtt^VxN#7;{@0m6QnK zDpPK2n4wZ13UO;cSgn;`c|Jr~etJP-V`NFR2jygrsz&)|E54>{bmZt187Q{ZMJ4&) z@?$NjBr3=yfxHv$5sutzYoI!vs{2=@vP$+Z$zY`TJED;8h5iY8TmZ0$$L9Z@OKM(l z?gt9+6nQVPdnqJg#ld$XfD%-t<%#De)h;%95LT=F^A5A>$nh`6+HW9GChw|92S1yH zpX|+PeH|-Cg**JOL$0at(mbE{+y_T$RfIt);DVhxoBEwl)X_zWU!e@3SB94Pd_T8| zSV}#F$)$#pJnUfq03T@i9U2dA+Zij*N#%pDjzumw8E3Z!TJd|^w1NzY+&3Gs+y(cy zP$Dbyqd<$$=_9=gTtA3MkVEFV^S8(g{a}duGLh;%1-4>O$8z7#| zly3?Kh{yKHsW1M>1C6wKIaB0}CLgab=G$0{=qi%w~8V6EoXXzp)FSYzopdpC39*!z?akTzqoy&JtFr<&+(bT@jKuRMN4Da<3!({QJ6pfJ4IiVX8jjzW+u4ekc#zH=rgGaJ*tT^xv?!F z58WM$m_Wj7nK0QXEMH=;@BS7O6O2EURd!ewNQ=!iIh}~*R;J#yyuQ}w>S(b4g6w7Y zg5-aIXZBNxAj7jDw*Fum(Flus^cIBk6*C|?qa?AcIur(r@)%G*=Zvf*d5lwHcu7O_ z=D=K3CHy3e>kjj$v#iwi0~Lle;}{x=cO>#;Fq!@NG)%3ZpZ8)`W+FM!waSloE?nac z;IzD>M$W2==I;ln0m=ug!8N2SxGWCh+8@KwY1@1{UsQAk?+K~*mq)4E0J~Z55+DW? zHcL(dSId}QNjJfFj6&>@e#6DK#Qt0hQVXjyZ7~#tx)fc)4|!L_azXQG%@P*r;h696 z&?tW?;xub8lW3DAv5xXpj>y}9UR!!8w`g?x@!%|BXTk%2;d15&xh}Z_hc8j8C$iKE zl?&vNBncpGnBRZ3m?e^Jl7iTVnv5jTI8PtTzeL8MQ!P#;>~0i=%SC{+fok9H*XCj% zIrVTOqxo`vHAT+%o92e-y7W>8z06XD(@92u^H2aGV+c-=9$0t1gFb!5&Ez+jlU}7A zlj6@=n*tL69N#_eI4|jjiay+aatPHpU#X+9vxQ6lyf80WZz9_MOV_9I@yLV}r3o=m z*mcM06sSG4$9#?gb0FPDr8NzuP7}{`;6QRceIOU{-c~yM8{DiNEG^Z;o4>;1b{PzY1Cu8FNEP zLS$YFBlg{-fyR;PEURU^YGhr+)4HAk1M{Rj4bitk#$1gE<3J*YCN3ayMy<^U zlhctPUWsQv)jOf|{v!NxccP@LNaPievJg~1X@Ija=2fq)1HyaJr{o(4k$l?U@uZf= ztsQeRP~wpAYBG$uXi1XD*#!wZzW}qqbRghl4W{M_fj9=X9${MHRsb%nhtN1r1vEH=v@6&&Q!OMK|gV*Cw6c$}`&9{;dDDSiIROHHT0=VaQxBq4`fD&N9j zhJ=+WbvB}~q5JuVzK%f5eYN22q_4EsmSM>Fr{mN%EPQX{f^A(-m#>p)vSO2}tYuc?=;;@x>`#Dl>dsDQli06L72{=|qd*+zntSZZ#>4CS^-24yt-|xVt*mkP!vvb}1n8Fbo7KJlthiQBhi&NPQfGWz^ zBwe@$u@s8xf$B2d)Ad(XL7yrA!_f**2_=I^ouauk8LgEpBPo%L*>k_}^4Zn8u-Ac~ zmq(`bNFnSk+R@yCPC$>D)tTYK0f^|~-3c4FQ@o$t8G&@4wY-#tr@m6LEn5h45F z769a8uBg>&B}v^j6ndl4_B%~(Jpny{#iaqE*M3Ke{DdR#H*|xVQRa3ufZJk-C8$tY z`>_;pTRx4}L{!$|8`%I2crh9&*|L-KjUe3M7#&sf6Sav$w(GG=Y@Ce>i8~#X6>?=s1X;Q+e?Cu9`xv!bT4Zx!igh|2lG9EYKe3rw&Fv(|f3H;jr0 zOB&t_(NZ4;q|lM7m>-i;oEa|aqAi5&j5#^MhBjVYpW{vOwNsxkv;x|X!KRs_xphuQ z^Z5oyIjtuATp}!nP1?g~Y9u~gz9lyNr!xNS_^pr8NW;fbPT4TO{mm$log~fv2gLux zP2Yx%O0<}rj~2<~D+((GGF)yCcvP9f08V)6X`O0hvLm%l?>@-s$pd;F%3L>R~ob>81#V69;=ADZeaBlg`P>hfz+!Db7EjC(uflY!x~^! zmdNL;_7`*Ntvgc8>+_6|F4VVrS5F+twn0DeE}1DrOVebBq)S?AY{Is2YH>x|$+A?=5s+SemVZ8=H1iWwFX4Lp~-g&_k2%<*kNiK(FVCs$*GgP=K0>y$!mKZ)<)w| z8C8mQkd9gi zSdk5E?M*i@aRggyMNZ9F5YQ#Vf=$tyHd1n23PnAhQqD2SQyw-he3DYR#^|k&RQx42 zL>}_|1YopiNw zk@;A!gv!#sjTFMNwf+|sdN4#_IJ$W^uXY1{JrCgZyyUEu$s17b&&jGc%xh$bl0oQQ zYX*G}fMDXwk0<$P2KYEb{lxgcLcWC#>*!0TJZoagXRG4TOq|b$g0m=!7-xZLl>|Sb zbEAo}1Tn8Z06hP5_|+n-`c{*gSaL_JT=u5TXvQXgUp5gui33=x ziR+*;2kLx{mpRQzAUS>hfT7xckoE#Uw4CL7r3R#}Hd6z_s}^c~!Qn0J{qf-S2s1eD zCJ+zYd9|jPU^6qDaMpJ|LLOh+kuXtwP1n2#n{gjiFx6C9v)}xl3VKXT%IUN+r$+!3 z*$&v4ocaBr+Q+_yIF%sImk)j&lm5pHVgPAA0aO;w?`*p{SxO97dAg2nN}6B9A!!5y zp`W*_H#?<41dm`!#w}KC&p$hyc>%y*-v=k)B>CkJxsM>n2OqB}<^s%X#UtR>&$(A1 z{5w-B%`~YJ=j-+XAV1`AI&*2uqV+ityKM11&s!ZqA*s*f&n^F&cu)&;>S!E4KY|3@ zM4L=4J!|ETA*)5%^n5Q#ck)g#g=lUEs*>05FDq(aoSq{7`a>>b3PAj#gZfNCJ|FXt z$b5r-yRZgL|4 zdbGm=E4$LN8#!RTVeY$^E_=Av5Laq3h&G%#=mW}>CyMgVmdL61CZgV|4rPBwTq~NJ z3x9O0)`@>|zOe0v>j=P|wg>=Eoa9@S=B_1%>?dP(@|z%m;v0QC!HcuC9sqB0@h0H0 zQ%yYCAV894dvKIvPT=45J(>~nTM&_>s*~*Tce4c_rT&t!5c%3vKw?OYdSnwLz<@n0 z)&Tg@7_4}!v}f?a(k;k?iyfe(P*D(d$0^FdxyVH`j_q27=0{wf0}F6194E;#LwhV@}(c}5hUshB?)z<~DXlCUjU#l3MK z^QVKIZ|Y31J;N$t!{M`*n(&fU-7n`sw#!$geVrz$MNI$r^z!`S%f~fPkt|o)C z?3ZrIQ*wk4_OBHbgPM3;IVvnzrF;fyZ5m#<5(w+5iYxck<7kKbdIy`g<=rGCoPIsp z>(lBRkDp%JT=nytgR$la_>I+00Fm_^8bTMo9@qiImNHk#dmb*inf+E1P6Mlw5eNr` zI46VcQNwHb0(O>h@#!k8xL;wB~!Mf44hgI?r*2jsB^1#R!n7aftNge-Oa zJR*4Jw!a=Nk%~^I4B!8dUpWjIcZwar_=-NycAT@*q)~Zq)Bq~1W?nwn^h}bXAH>ol zeAKx730xOgFKvB)&VeasuY-+Kk4>Xf!*g*$GsaP`9_;qT<8 z`vkHP0pn7El`^mB{F&kHdn0P_epk;Zb>a-|E(`Lx(;s{GWVg30JyquN{38$Rhb)&Y8vf1c)`%CV6<(m+32E+5*n^u6 zn3H)dh~~G;SGc36BX>XS?W%K%m)qxPD4?^aQ(~+$fE*N@O+;@e z*5gb$&%h6aQBA^9nm__7aJQhn$%ofy?yJK)xrQ*J1yJ?nUa3N>Ocz=6lV*rK;AgM{ zhyx1!kR>Ww|Yz69{ zw(p{k$J0R7wqwdTx#xNA{i?9;#Gsc3P}) zI%J5h27*v=j2mGh{)}do%k(uE8<-;$9p#ce=6w0LhMyW1P0OIPTwg6~ z`~jbx##N$5YaNpcw}l-kG3g%o&``4w!#6=bvmu|0f=1J3>?>0vkqQu^n&EpNo%IqQ z7YfIk&lbU`Kx{bYmLzJ7bhSLidhJM?iNkLrB1Ph($+JN8MzKBi2F2Kthb7fMPsFkJ zk%V!nLqSoO+=%7vhgM%s1fPjY2}f@?gBhyb1Hbz;=~=SiEqi>@z(B}0NcwxS^|F4= z-ucjW_U_xidZA|Orrd}^&ZVYN3kf~}G$Jxt3LYXIDYbR^HTjWahI>3QJ@Y-yDJtr0 zzV`u39s*LCZbo`A=E>L`O+d$^9B}@@KSXSAf+S^vUySRMX$fu!BuJ%rT%qsQYUWj9JQO zfKX@0T85~3r*$&R@*`x!ukitY zWz+dW_5AJJ7ZV1pe2;U<(z4s*1|&^p^{}_r4{HPqbz;`Upjkyi22Z-}^d?$0t6r_L zw6c;%C<)4k9mOXtt)Qd_KC$2ByZ=`~@jmb>lt-A|8!|8*nfONr^(gk|_w{`n7eMXi zWsh&*E5u=p)0t!0e8VPCAe9z8vO}Mzh@YSNC8f2S-pwC@zKGYcbIin*m`iv+eQvIB z^JlKwJJeae@ZCbYiBs7*rIay*doLQusj1C0Wn$1ji`;lZQlkjEj-Nk7aUiy|O-8bU z^=?$Z8CA#S(a7^x4x2aYG;y(|dGD;yHPL@y_gyHFPe!MS_2SWXNexV_Q62vubMG0} zWS6cDBT_5~Dj>bss0s?wn~I7BM3i2Y-a;pIA|fIppwgrSLRkU5T4`_a8q^uG!Gh~~VaDu=Baqa^M?mC;*gjaRLbDYrN$xP`^c8eNw8y3J zrB~!aB4cvs$2}@Z$o@Q!Y`KV=@D*ne;mZRST%k*Jt{z%sretXw9G*i!u$%jMltz06 zA-q@{jBctNwFg=`D3;J}YC%kGPpIZqF7x%F^EsQ_s@rFF`aebtJ?z~!}WW!$L+gKIZ9n}T*_fjbPFg3oI z20k5_TC%*KjRLnF;X;mUUx7dxkO9FatE~@S43F*z#Z|=Hga;d!>0h>|zL2t+$jQW1ngbEy=*Jw-R zjil$2r4nt%EQ6Vv`*jUt%GZ3VpT`mV$Ii)+_t$e@)V-k(2XI$d7^jSw=dIF7t`fsI zZz|-aMvD1T*$hOh=WKmM#`fZn*W`n~*m%{R+ zL5ZzJe)WO($fKH~i(J*?Q6iZU@Mx4p8B(tD^E>U@eU=cAOU|D@yPWH35iGVvgtf<7x@hi=`Cl|J3XaNz24y2Pp5 z)FW>)twlF7z1?0bM?OS$r~8($r`{R#+Wf-eD!+G=T6o2s?2~UoKGl4|K!(e_=~;J* z6Mph95o?r3NB0V+`0h~3`F~a1*ku`k66D6XRBwt=sK`NwA*hT~HF#0oS+R=^xE%5i z3uz&jc6aS@GO2}oZdcd-wEq>vVtHKtlxDtGlW~OF;MZc8ej$`vY~FT*nl(bad8A|k zf-ZxOZqTVLT);{{JhVd`*Q{;Lr{6PY{q>-AV}slyreHzId5U|j!iWQXFn?6${^nhk+$gnL>1LgK+-WA$xA9_DH{EMBQ=5lSl3k#G<=%w;8n0Dk;bB_uF+{yPDcR z;#K=vLobu6*T{Fe-Ll|?IPzTjgXSyx5SshBI@0ufa!cVFDfE%E1 zh9gE`JYvL`2ync^!Ib^sTsqLqe}9TcZdT>P+hVYG##ur06UKo>vw}L1P(@sFU7(iV zyYf_c~*e>0x7C$)Ey!9w}&qgu4B;#^78)4moFQE*7@Xt=Lh%A2Lv0y*@ zA@W?Rq?BhNYO=>I%ok35^|*ZbA_9;XDMREJqXfc;)L(Y}`VU@t?&B!KJLK2f=xRRB z&;#CX-h%S0tWTQ_-#yU)C6jwQKWvxpUg-QP!L91B(0kcV^>q|RCLOsmZ6?WbRp0y& z^)$%yXLW#)pFysCe}wfE5Y&UJU)A9_q^L={&uf})oxaWXC)G4yYA$Bt`+Eev)Mm5NdW7C<4RVwHe|kv{2LeGLhlaJ61vn7L`Ydum~7#Z zc+$TWaun0>DCV1*tSpX;$X_Ytu##slVYCFR$>mvH6WAWV0P%ZU+ffB=MEbUe(Lgf-!G9o#9{j8GR8mtx|Yz9w`{~I>av#M3-Hp zQITbC^yB;~Kz4bxu$D~Tms+X-PCqVepHA%W(8--n*!O2&;NX%?SF7r4sv})zbNYt# zbLe`G9{==1JMXC$mMcNkLh^|6vFYUhVF-1|lk*{ywvXp2gO%)Z|G!r9;T^&!x`uP$ zMgeaC>AUN~;wbRh43ncrZ(ar4 zYabE4NjX<<|7F6C=fLQBXd|YjW~!E}Bq%>W`6BqA(_t?KEX2?9rbGAmw#c_Fbgmrc zA+s5(SSUQD!-AyabHxdrzxogsk7x(KbNSulwJ6(bR`;Jos2^`U@d$R(^sMwZzkU`f zzf1)w`@wOs`*Wp-A+$y`-<)z1JC^bEvUmV<8v3~%W-pg1MzXHYz;p~00e5lkdhq&) zSii9N_% zoqsR2^oi%G@1ntksiOo~qfiB|C34@C85e*m!neVEC8R>hB`@holX z!SWs6c>sJD9$9HBge@&&h36QPXig;zVj$QlzDvYyo&tw$rCCh1ysg^`gS=`5-tA4Q z!MsEPa58E1R^0MThor6JIcsR`Oy}e{dcI`;Ee<-V$+#gK$ARzUUBU@L^ z=uwGEqC8v05#B_FSt2k=cgsrlJ=*ikWZ^{#Kt^`iOPdj5^_01U@JPk0F)YHc2a0Y0Dv7E*!= zeiVv@h=gYXx-k6(Fbt=+XrBW+TCbY2sN=^6B2y^Dy=yNT2~8GrDz1L}&Z>1HTTLDUvLU z{T9LcF-3H@8ffJgks33CJgNA!#rsk?UsprS_kyiuoTG$opVI;$fU-$Nd=|5K2ZtT( zYFBu!etG;Nv%?DR9@A2&G4GR$i04TVLF7wGQ|N<9ltvF^CZ zu6&aGz+CETH0Of!uWKwl-nD1j-pf|R)*bAZp(D=#@2v>%ih&K0g&h9ge7ug=G0COH^vHr@itDI zAlijju}Ux(5c+o#MMD zG~l5}75)Q{sn5`cmv6PMpD6F_Y6!}Ir^f1GCCjz}W^vRfhmuBpOb4Km%vTKSTN)L8 z)Z!&W7gCRMYcMG77yPCR zFO8zDGP#F!3Qut#dwT975A5=Bk=2@RfCmP;Pax&=#=nE99h+(C4^crDUzb{HmO*6( zg(w>ZD&a%k>#Yg{?WX4DmlsX^Ei6tbOIam(rH66JNh)9YntV|}`~%Vs%>OBZb|rg} zRmM|I5gu4Dp=*u4II^Tj)Ci{Mnw)twjB~tLfCqz$A~0iXXc#;`{?uo#v?`ZGjw!A> zV2IUvsP_TP{%Wm(_dHLCgmiyz*|_BWA7!i|5*jL&xJ7+Z)ownPs}jGVKJXwAD~2|&E}zKl1c z4vi4DNRpBPo)$gdR8i<6KU$(}Wztgs<>P2)+7ld^vo+nZB4KL2V1p;$du&)&)4kw| zZx|-_QK}s5EH7d4E&7@B?MTz)ri;6RoOH{ zSB=<6hZ7UM)5I>*q0q+lH&5^#v$R&t$1kDF7mUy5w%z;P?|`r!w#i==ilmKaETU<# z2I`-$)k1_R?l$@D&4IAxoauT@1HcQ!5@g-84#$CnUO+!9%~7Z_ zDu!c}K*M(AK9d8i8a_NW@Cbh_zgH|bLDA3qO{pC03OvKS`4j!M_0yk*OYVf66={v$ zyRWIF{XXe=>NO{(q~KZE*))8})AY;9$6J^2*p*t^Ueg=QWcHY44#4`S=S}-Ax#gT# zb!CqP=+Ev%A?!0xktwVtsIhNVMmAgo#J6ssXpvt4>>=~_ytj`%_=C7bcdVncMx%dT zBBJgdd0mA=%KtXS+h6DGFIg~1%i?6AXPq|kOKVeUYI@T%EV2YwLUCBY^>7)o+^Rx% zR_#olM71xC+Blv*tr>Z~t~=l;W;4KT-1^rnMg0R%djd3Aj5-|I#K+z8B=W3u`J|AK z&CXlZ{FIw3n^{cr>Uqfnjzvm!Y`EHn5gIbLmtS%MCWc6MVK<$ynb@PV?x3C;vdzhH5x*2FpM}< zH~uNH(+zlbobz&w2JB?VZhyYN1zgUf;|as&^@h)a#E-Bt_e$^093ZH}fYHY`3NiC~ zYWXp>liO67&2>ZW`w!Y!`*ICXSE~DbX4>LvSH-dS;({tZ9RbFCfxxh12v7)_%z0%2 z3)&Da?{veu>x=zS6`SjClWB)uGYPY~!!I{5#)z7XLGZ0#-alk9a{FlHzdL=)fFGcJ zgU{Ki=^jcN%! zVlw?XksuMmLYjR}jMFdjMZr92|D$ELdt|0Tbqd#g_33!MH>m(^Hb~`DGsBWc4&xTiOJpi7 zb>U6iUOs+&RK=L`)w!$2lP#n>%Zgii_J}mG$Lh(074j}ob_?G_tcAvZlJkmd2urF7 zPZspu7kwj$euD*PY75U2LxAarpmQa2*`DkGin`E7Ot;+hGA)JT)}4G+emigwivIIH2E#QBVu=ks&SrF+yFaIei;Z2P`XY#9OR(dmOxDUOWou;LJC* zL6ArDHI6S9W{z*(f1}H)bnMDi-$Gq_HpH<&2lT>v_-#W|z}ZWCPM=y^+Lx9H7-Xg46GPtACp6v#g>$bIYNy{ewg8Xh#_! z3`D+b1|TZzZ)KIVl9lvXif7|Z)k54|*p3XvAEFcG|cD|Q5xMqNfwd8rV#pR(R;cnr;=BN?5o3@Z#PG7zR zPTEHaZ|z6f_4tmcF!ut^alJR0#PA9NJWOq2<%=g`cWty5#ar&s{B^DMFH5Y-ar~#1 z^jgiw*qgkB3=-olrpm4%t3@w>LQYLiQx)u7k&Uwe&rQ|JkCPBPZ~5qm%1=~hK~*FS ze8rRwMX-VIZ^>{CoT28t+wkG?5@LmIl#p>c32$vV_s}6 z4nf-`1_l9{|8$gIlZbZC=fH5P=)0R{4YM{%7T?~rI59D2d&r-2)}er9o&qhrWhB2k zv?jhqD1t+(qZZQQumJ>q!(h4FU49tlG~hT`@Ztzl%nn&(-nXS{+LT=<=YFUP=d~Kr z)z1Q}m**ycrlzg#o^9|mX7SL|E$F|Ezpg7$s(>YYibjo*ufpZwi;Hvh0jb5elE=R! zMJOz_kT=9_7}?Zv6m+LHhwW?xWrTIIsl)OX)S6F_YhOP9n?NP~dOA;ts&3CPD7OShf3XKL&D=Y1!R_reKsgFK=Y^+%qCI7WYnf@kS1 zmJY;R0Yo>La*P<~9O1Y{i;gom+rC*zx8#gXAi$&`q&UQF&p>FyCjsg)1Pm9&RNo>f zfgZC6Ix34!yQPe3@7RF2iGRWRF$|`~vXtzFIh?#%>J&%Csg?D#T}1fl70DDsR;e?+ zg>vP?%ZW+Haupc)F%r?UfK4FM1)!?IzB&d<1+k^y5^UIUb#6_^Nnu z{+XhKHj%gxqsMdeSV}S3`Hvdv*s%`v8jU9z7n6swBQpl~cTZ08(Plj4KgmfVE#n4V z1K$khcb3}bKzRRVW6qK3hjhSlC;VLMgeKnyYLy2sxqUa@mB84)^p07)mQ|b}qSmWNJ7GgG zGNaN*G`Ti7=gG%x!mut@cs+6eXBtUQA|NP88y z6M?7+d*;5Gmc6jk2|F-JcV}k(^>mM(-Ts-W-|cGQRGq$GmzbU0IL-G)`n%N*-hg)U z)~7tJO1INChW3-OJ41|4n!D-q(Au;hF3)J`TZbHsE#gTspxfo1UeFP(FA)YF4}XO# z|HUfQEokBaSydaUJcrH);+#GJHO8c?ad?b?wvSJwO7-*y6RGn|{LC?v7v!JZ0LVhA zJBYh*#&1x(?aYf}24}nGIry7Xy{08p_kvPBPZhpSwh)#S1EZkm!?6Yih&izZFO<#s zxGcIfX1EDxa1MTfr-62A-(5T5#HF4+kVarBO$PmU4w#ELROSKk#5Ff?HVy$j0Rf?Y z8AIG0FhQ9dti>LB9wbORn-)BFiW56=RS!{LDJ`#&mBD)%?z%Km_FziG&reH|fzxki zBDhx8M13>(P?g%S9E9EVq+2{=2Sm&i6#(n_xk-(~uuX3DyhuIN|Mnj}{~96m+jQkA zJttssGmS~fACv2s%80$9XRBq`hp?5f!*;+OE%fmjdklw+jYG;dPproMv!!2;JWtE& z&sA~|LhcSh*#_Zf2FY^kTpQR5P0SGoj$;g^Pe&+++tM)R| zdi4|C&Cr*r*aTDof?0Uzr$*{C<2tp%wInkB!qOn68>mK?o=5kC-e67b^R?^9C6BEq z+uTD{Z~U<+v%s}Rl!Pl@x7uIXyIPuFkGU-U^{6Jnav%xJr(ENb_^QO3jUS*6v}MjU zF)~}<(b;|AjQ}AK;c=;T#0-=YFPAmWtX@2%YcEXlw-U*D~XM0CwAi${VxqCRnNYn&EYEC9FHL=7ZEJE7> zFzZd?C+J4*GPlodQhEJn2)rVuuApo}y6Te2VT(i}g5pFBjyLEZFqv zWEGD_v;g)wC2Ak2L{UJui{8>^W5Q)hqJf*j$Q}@q&^U$XN48A03hD|KKK2(m-*pdr z{Oo#-hO{E+z-?~9-B0x`2?dG?F^hsQrR9NqlQ-pax3a-45Jfm9Jg` z#t<05Nb_g|lA?wg?c~OiXk6Z5QNyuGQXciz`+X|H;Y@z;KcLh5G@O^0$bNM1(>p)2(cP3v3%nW9Qx_<=q#Ec2au%3W6aBn4 zw+EdWC;24CEM6nihhTD^*})?^&4e6Ki`-+7|NzksU$ z#TV))WPJ8e&=e~5G$vkPe0I!d`P9o#P6|{F$67+(wVr)>>8*10NxC&N&FXwH99ZQN zb^fnRDgY?Fj2OJ+kZOP9u&$se7l>4oeryMZ$F)dVbxkaBp%}D;6dNIkn_m_i-37!y zZ83i-vKz?jvRdas0$3Wo+Q=cUfCKg@%kwj>v-o2`LE*aKvoU)nhhd<(i4zRx`yd`* zDtFiNHW=#Ub8cLG&Og6TiDs%h*RKN%av-rKg0@Nl(m)Lw7OO+D!fp|2-P>%5C$)!=5cE}(!k#w~tP=?Q6~p2| zwNF$XiL+GNJro8_8YGed#s0VZj_(J5jcFJ5m!u${0hWIPZ?PeEFT|K)v_Aam>?1;! zU22ZHoLS^;hR#_v8ri%&6VvqbT*{7D-zwpeUpl|5I_P=(EUCc)?^r@oq`2)zO1n4H z{D@8KYY% z)zkvl!^KjWXFAxXcJ}bcg8p3MfA7zK6E=nl&4NR1p5|yq0mF&}?$B-RH}3Du^ouM3 zYm+KoU2{HU3Z*J#Xe)UvlQUDR!gm?|NG5w+147v$pG2hJLb z#%fL><7~;zO(=9k(b?Iro=DBTl{y9^`=K1YB17nTivx}Q_<6Wqg|3*5$X7`X^qV3& zFF~3ogxqxjpT2?@b%t*e0kA})=l?^rXv_Vq^~D;#)ytz31WlHP4M7q0>mTA8{IVQg zq1{INkS!*09H3Ywvfs0OD6;|Tq3OMwmILaNFBb`y15$Y6L6l}yE!LJtZIVyfc2Gm)kB`0Hz@TH^*L2W{_ADV zfl8~NUws)l8bY42CtfyyBBr%*?ThV4ls5dMvv|yri{?biWXIQ!BbEV)`5$q&|Ly|E z3r1Xt@1>re-WW*PDO*fbDTEwc#>GLhHz!!D>#sXNr;5LZr{buO1$9P765>i9uR6<` zn4>QHiB7Ay|Q!>#O^5T>OS2?WiB3y^qgH#qK;^s$YP^OUrVMao6ncs&d_)C7%7m zH;Yz+LZu9h50o?FkGCjWSu$CV}ziFyz&oR#^m@&>U z!?te4>jysEA)B1kw=14)r9G^uhfchYBP};Cb($Yf@%<}N^S?Cmz>!*ZXt1PSR}_ca z7gN&iTr7=ixOXWOeX{A}+)ndkG2$6{vADUp*ajLKyburfL@oBotd?7NROj@`6zMiD zBAv6O%j*w3#he3%cH0=UgHJzX`u&j}(B3%vV_Z<@Z(IPN7%kO~yLG$F`F*6qKy%D> zQDY8JZ>;_W2>&m3e7vZD)MYu0urM*{oh~Br?dARR$)fyN2~@tv)D?p z**{GU6S z!F&H~^7jAsf=qjMgflIAS1IEcTXBA!2^iteiJPE+?FYWi*=*Wmt?{8jp(Pi?1@I0@ zyuV-n-@K{2f^{GBt10b+tK}Q>L|ls$w)d|Qv9bJPipGHAUc`5G36Y*FJGi8iidK!kpt^l5W!dOISUA6px0(lFWVAbMU(l{E2Tt(YuH+rR zY^Y62+VRjgA}357c$o6KnZg6FsE~IsVCZA*AJmSM07G5xe-g z8Obmo?_>|2mY;vf166_3*Z=*|DT7bkuPrj-@&6$^w*In`Tx(L~8zJYchd@}+a9KN+RlUw1DYZoY1&q(od?ryJnk zN4Lm&4RdlxEcNShv=>Ha50twCGHtCk2*FIW!hhAhLE)C%SY*11a9jZ4?IAa$K2E8|*+=IcHcoi)t%{@Ba zNwDj8bJ*%;K6KB~C_+N*dL^}`w5utp59}Ys<{s+^0;|*O&*u5x+Or&vsF3#tstG`4 z0IB_Y8R|+phe1HB667K2tq#Su?Xtv$fa2%+(WJOqQQaarS-t9vVEtV1xHg+Qi(wH-- zNDYMeT7_G(#`y#}93uVdG9t!f%nUU)hT-=7^}wLa8xH!J-lSRLd#&neNnwq`$BUfF z_5Jv+6m?E{&HM-1ueXRD-do(yOE_*6`S8rz^e;$-_PfSZ7WqV|c|p?qD@xLg4D!+| z1&l5) zLQMF$ywH>JvpOA)W`6AhbD^T)`oTUt$b^M4hgR*Ja5Z z2W)(xL}~fI84(}06?*^+$-NzR59)ymhTwdbtpqai8Vb&&=H3i=1nK!TfbiP~lpHx} zJlq3{t_C&(ayW17Fm%>bP2U3=FdVXtX{r=!5E`&GPi<5`F{~#`&sh@YSUo5ip|D%r z4;g7WSHGvJNrE|hdHH-HPC4sEQaxc7&~4;i7O$a)d0#q4^=XS|=jX$*z8V@u@xdx- z?n;t;v;VZ^0w!j)ch(0i?R2R$W@JY?Ht*pQ33He zPrCs}x1kUk#qCe-rT8PlPsUNm7gIA|6rTcl@D_4=SjPm$Rh-arNi9KPz!peF) zK{_Vtgml)xx00)BotjLvnhoZqlySO!J97}(YjUOU(Lq|e4`RiPQ|td1i! zJ4;y#QbB3af+(wS|LyOO_8%?PjI+`wtfU`Og`bG6k^$c@c`YkFCnd4jwlB|GOU#ct zgc%g0_-W}1X}vdGo_B))nW-%`;8+21jaZ-!!N*;Y?xHw)At^X_6TYp# zuw|61565!qN;v#r(s4|}f>Pii-(OpK?-z4~{LwSC4E-y?m!CG={AKFMk?@Lxt?5|p zq8D~{*urcGu1q*Lw+PH*5CYGU&Y|?s<)aD^azJ410Wga2@)E842nf5lu;iVr(f(ck zfrF(uKkqM6hs-19KhKfGr5cTEjqk&1q03$afaL0QO;H};AYc?!wS8bpG#lWPYHU5+ zB>rq8OaFX_|IG*6syO&QP$)q{&sDBgQm9Zqf6OvIo1I^UL@zJcNDWku&a@eEQdNAT z3GZ2U2a6G2?uIg*wjIVmib2EKo}B5rsur~K+Kc-8-Z#Rv~WFG*Qs==+?wQvLmvjE*R2g(J@)SShwoGm-n}%pcIl9jQu~lteth)i#ba-FVVl^3*514BTJF80 zZW5hEm?HM9$Adn}y+#>ZL)Aq(*vL~!C0QjL=xa~RaeG}K1)+|+^O(GKewm{xXz?7w zv?jv>0V^Wuxg*jW*P@eW!I;q@Jq=t0q(%%H*cEUaQEZ@WeUy)G>Q8&Lt%xb=q*&=7K)jH^O`I(m$k&hoM}x8T zA-(nkd2TvMpV+0(1~b6SAWHM5(I7*{wqzbVEfrm&NUCyLwkg=)SMb3?EV75&h7@mX zn1%0I*R&G}ve>|AJN^D9Fm8$8T-?=($1ZKqvG59M%|A^*J%8bOEsXWLt?KM%^-!Ch zs|!^)*v!oU(JZSUmY$j5Cx~$LGjE~FHqH)kpF7Y45o+Fh+&+uV8M)9vD|NQ6-WgIv zt~UIhMdy$<>zI|ob3$b0lc$a*iT}BnWqY3h|&)nJe^A{()RFG6k}6#1#f^S@U-N3NYN;MS@tt_I}9J z%+)1a{E+qRN;kiy<6eUjWG%{OV*Myc15qA-Tcc6D@V-a^5HTQ%hsjX)FZ?g>l3USt zkWVe%l95lf#LQv*dH1OTEv3AZd-k{&y&0Xz+Q49JTb%3Y%MW*G_V~=k1+=>lqHhje z#LQ&*n>wsP|9fyrRN zHSl&s4g*!*LEAMz&@JpL$9{Dlzq=oM?oaR02)+AM)%zg?l5njzxp-`~%j?QH5o6o5 z1AKXDK&meBq7z6IzYA9AB`2!uqHEwAaM1;37eK`L+rNJm9+d<8i*petWVBp?^Mv5M zSmI}bb=)a7u^&OnJ-yqd1CJ)@RGcVJ&2q!VIz+DcC9A;|7k(ACmbC#=J7oOJdGG;X zLlxAaKaK{0*O&iu0X+E}a4`(Rd0nW%vJ<}CQVD~lhp~!x7&~@jfRBsSVH+`^g758_ z7A8!IROZZaKo-{nlt*L|IGVz_P|?{>bJsH}3DHv6K{*~bM03i*H|3!r=hkBLITvDq zvXoEBE*g{9n}Gyk)h>xOF=9S!%#z20z%AkQ4JK+(*dhxhw#8SB4UakE`e)hVy};)zf%cs-ZJ^8^sx6zt=;t@&mccu3j$kvl^^Po4Ek;=mmSGrPefZSD-w%UOoTU^>~1Uh zU;_ea%YqpwC^ByTUT%eyL1(1b?haKewc7SW%sZ=lMuI}JFt*Q`v-^=vN%qzr3Qrmw zD)T>uCEVyY!Xo{v14j+2;uSRV3l$7Ej|;mJ12F<_ql|>|djkl9tXom5Lj}&Z&%yc> zd!=>60K|S^WbdVf?^bk|%wCDXQd8u@j=Ojfy>3&-D4|nl#1a?>7VV{M+A8iAN5g&{ z_n9x#$u&%i7kT;Xs7h6?`zS7rXtfowuypb&t}DrJRl=@o=Q8>rd=0ZU3Wd%-frd{X z=R87j^TqE3oLg<7AOrK*rz3tVvx4+s09tg&7lcBQe%qw&8*OagN^u(+yH&(+SE_MU zqkq24JX?B|vh7cw@(q8vRcAP>vRzk2KA;^y=`)eCt2l@Ax0G!Ar14-Z?8?P?d;GgO z9bde(7x2o|!Y$>ld^hqj@bdYlt(>im{y3_am!$=p6kH%S@M4qW8c04A6mL%)=0tTj zo?x1xQz;k8#I8_hYgUGjZa#Nk=&6i=?0Ba_21}b|VQX<(-}hcC80I1saK%9F-he^GY1v)kweH{-2y?Nt(;y$s5IX4kC7rm87J!^xUH5+XD~EB?)h9NyfIcHa4Ueak!$PiWy-Dv3aE(SaRu zXQIBGq)6_0NSf2P!)2lgiZbZ${t~Zney`e}n-{=U zd^H)KB!8XHB4^aMX%tG>l)~afszx40?*)c&D4pL=p}N~zMT!G-a$7m>omU1U2gqFD zurcg$P~RiIptMxhy4=kMiuo!3B%xS6ACO^wTJsM#g(U7Z9&9Z(M6T;Qh(ey?-@#! z%e1H7aVa=_+9aWoZ#pJMjl+jVeC ziYIQ>!BpX=K~eR`7b61~muwV)Qt4TGTLG~7EMNHuvhp-PN`w7r+qM2J6NijTPJ%mV zzG)2fhh=rDo|}jUq<-p8=2CrbGepfq$;jaZpRZg*2o9p1Vo|ld&a$d#y_W>h0 zUsl1=HXxWG=-^Ep9g)6L6$)Wo3&WKW(Pfntr$_(IqXJpf1i@aV#P-6Sr>_eZ#L zMv9GvJz(eBn~U8(?>9lIf;V#HLT~;&Bh7x@rP#ocP`)XHStn}kpfxkeB^w(XE3&`W zb-^DYZVt2;0#Yl;FZyG1?uNz=f(gF;ukYOrd~V-Faff!-mmkIIcN!-TC6=v^#HQ*( z=1sI{YkmN}O{BWd;ImU42BF%H*q>doiCl8f#O`%hli7XW(x+NJ-yXWNh&?nwcl3*! zb#iatDY?@T3&L@Rfjvx)PlPdMCk~&xTGWY^TduMK$G7}+$hoBCB*Be5Imo993hQR~&Eb zh|M_~By10(Mdz<#E3ABuaUEfr!#V!rsT&d;zE27RP*r10^1?g=3KJ8gAzP~Y^r@& zw;Vf+Ci@!3Eby+Ty9A}cx_P7-UhZefpy}I2bBfEW#b++XZulvzeAw8h$YQ6Y4(Dqo zS~c90)o=T}2ru&U`pnM(lZAf&d1{}vOgy9^tB-N|XB(m;d86Pd*)O*pUcK_GWUX26 z-M(j0P-t#nnT6rFfcQd9v$G%0AR@_cWP#nI#BJyVi)s&I{Wi9njP;mzNb`pUp%uvk z5&SK`^UqFjP^#P?KW1wD3lt$CvChEKNYZ}iSoXk|lu;L~k@i8#J8eRD{8FWOb79v( zFD9=}pW;30 zJm9B2lP+g?e?}Fa4A}Pw7_k~wc!EJ`@`urdk1t~nK7SWC?HDNo9qKe8vD)z0yf*r= zf)C=cpU)RVYGJ>+tJ*21%m(my_RTtVx4Gf=gSG8PgagEMZd({?SYlw*e!!R zZI{Nvw&DdhDkqu)e@5xsf9oCYw3)2Ts!}`NaPpjKE0gqGaY~T$Nb%`Qg}%j;h4JC^k&v&+$)XfedB!%QSKX@kCWB z^%~n9DR+?9y;a}Y#pcUt)xzddA9ja~($ zW5EM=J*v#|D^tG(}{g$d-{DKDsdTms>MmfHod76gdO zbN4!#h&37o)*oM_n%D5uRT8(Z_nxU?E3M*ndj0o;+S7X?^=sF%V^g>hxQ+KYWZk}j zD4P0HHB8-mo`;1K1;w9rNn&%2240>f909hs^00W_cG<9zE6TNno)gZ|KHGYAc;gb= z2&v+`j~>4cy|JJHY(5rWiEU$lG(2*Bud#Psx~|(Kul0Q|VDkeYZpi}1#tg7d0( zx>robZgi2-VojXIFJ_uu_{1h|r8h0qPe2GiM69)UZCFaDGo6BT6*NZZlZEY6KagBA*II2>s_vS{gdM1n=lbx-um-pMIStg{_qCOy8hD* za6ZKjFJ*~|cCP_0BsRb-5S97NO8;JCQsmLg)e$lRuYaQ|e?l=g5mezLJNZvWflk{h z?IGN8xoZ0(M^m)fqP}%pNMw4s^S$Ja#VH!IWL>}u?_*oq1V>eIAiovODKVS18^W8( zI>~3H<0Z{vjJ#TpZfR$4kCwPy*)P;9SOJm4tB)LC`>R*sBS2fmBuy3mq2^3|#ZM4Z zf7VeH{$+mS9AVg_V2xP_z6$oQ1W_Ff5b34jtx`lK$Wj9}AIY~7mTK{pBj0SShj7Ed z0RC$-514myI7*^g^Q*?*&~qe#aL1j~LobrIW^KZf{c1%j=V{QSCE{XS4JLFq+3NM`fb20zNbcQi4AN=hE2n#NuFOt|)I>K~LFTa8-Rrm(6C(VFC$^jI+ z)h<(?H%>1*R>@4Nf`Si882mU@?l~P_qK!Tf*Up#-g^rKkI@kdPaKJybt^$rbC-=a1 zBO)1)`aHqrTuM@SbN;t5n9A(Q5rY)jg-2sE?FMIuRVrLnn$o3d!++cgfM6kuI}apfEqrIUU0jnE z^f_1Y4=piAe5#F9JPdKWUn+gPO%Ma8zUUrsfuj5GCzkFNfTnt6U~2;EKn{Thpm6p$ zJs0RcF#Mge{Lh#LTC|t?-NR^E{(0TX?P2#4@*e&wX1O*0ai%iB6r~=|W;TjWvvTku zA-ET!)x#rsrG+_09@`f&U)AKY4}OVj&s)U~o2g48z;G@Z5pJmxF9l1{%GvrBEsJMI z=l4u)>8~T9y`=udkL;U^K>LLhLn)bG9ozjG_UuHs!SKo^-}XViNa^-o~CsyP(MS){f7gX=+d zzr8J2>}Pq4cK~0zI^lYGSWMPZ`3-%^VFhC0P&L}q)AOTs&n$;m6Qszvnn_N-RmJa}}^M6NF&sEoc1A>i`hS1odD8XJ}t>VtVOncB;MOIwB#rscXs5kPW%-012(BtEzTpQfrI`O@{XUR?M5DTMR+ z!X>Yrnw^(^-KOkT2QCYLMDCGL{g(+WF%q*4xgppQ_3l%Ay%b-MisQ8v(+|ftpOp6k z8Z34xBsa0e6RZEzDlQ+)W|XDMufsMyL`?E(PhUs&%vVs5gw9m742YzJlQd?hdl4?yEcic|Ff^A$JmQzj>&{EzUE$96mgHyKZjqkb+w7(N^n zT;fanZ~+Z6x&KP{U*qSLPB&!)&01f1iE;SQyYDYcCkEH*Gvm6+;kh<>3i=Uu-DUwn z8^OeN-_GvQ_}Ut&^=PXftF3$COk5?U#PSf$^Zh}4Az%qR790a5`9r8JJ6ZMSsl}#y zBTuK>Ozle}vh)GHEJanXr5CX>eRXZxtZ*)hV1M_SHadGqwh)@GRSZp+wXwGK@fh$Y zaV=0dByHXgA;{a>(p@M3VK?NN1R76@=swU za4E~_o@NI5yJZWpRiyU7sd)Av;WBYSN zYIJ+zOO=mYi8m(eDQx#EvmxIBnp>j){@gPx00-rFcptz4BzUX1B(4+;I?Pp)<=&JI zt}n6>7x`xUd}7`5xPN1jSR!CnmIT+@zJ?Q%!C*B{C!o#hgSib(`N?=wR>oq0=k{i-JxDDtvX1*@Te!Tedxx*8s@%Sy))y$?}goa}_>v zhR{{|S};gKZ<9tVwSmf#+bj*sn*3QvoV9hx)dSPGG7j%9!N;-5B2~c64&H}0oRXrC z6EtBuJyxqEg6SC)&G+h?_LMzJLq9{J`!hf|HZ#Mteui36p!oyg4}BxXbteOAOaLT7 zOJ(yja|b{?d-9J^>M+~h-l9dV1Zf?}do&gALDf-+h>eX+Evj!l3jGRwb6AKB=}EB2 zXk)J$w30Y2X;bjXry#}14a!0(@m!Jwr3+6T@hME&eQY zgZ>;KuV4-wFl~67EbcaclM`g1tJCdRfk)Z1`}s)~IPn$vQodP_FbaLooqb7>(bozA z(9u`Pd~ea;fhHu(Ut{ehx^* z?}-CAr! zT%2R06B2)bj#y^8n2T@~t z_j^Y2?C*iyzP$jT0qlc+Rge-Z)F1ltW{It5v7U z<*t((M}Bv#RonJDL4dkys>~3N+$IiFu>E4Km5%|Mhcbu&FrDry6yofg=<7d5hcfob zFQq~^Jr#E{xL}6hOE6g@4~}6Nn_aa^?;-?ODVRP`2smNyjA;T>c3G~m9^ij@ytI zA-1Pt0yoO+_YAgT9=nz!uz{9oM?y&2dP%6dDsRI#&dU7z;vM+|Z20|VW&Z@+To^Xi z@MT!k9QmBhfI1l2fQ9r8Hre(~*Fn@S=P+%v_$KcD+z&-+M?R*j+$F!r!4z5MU}h9J z0J8F%2h!Yp&Jh~}5HLWC^{jUWUGKg+$RN`Hl?~!3SEy>g+8-NdUyOH0%V?~s6FQ51 zP)idx04vI=iC1?1g@hixcRI#}=xLA9R5N>JvpFh;TxZ!O{J3o%pD$(4bcming zUl{rEuxj<2d)2-3e190MJp}U1#-DR;jH@mFlbHlyE4UgpaY0%d=iv?dI~=m^_1c3p z^^B#=rq52`Hi%pSL0IkjqgLj&Eg*%TY!~la{!AE4A8*Q?gHzWOR2-1x@n4 zG{@eM+wxezrZ=`#da9L^cbwL46U&M`5H_eUu`NfT>0g1wFbvz1C235)A-^S?@&Ikx z&672wpaYnRchfX^SH8Oe@=w?C>2@yI$;%j2`2aBtfFnhojTf(VKFYZ@GX3l-17SQ| zSN_Y`f^>RY81wD#42}D4)$B5z_>=WI4}V|3eevH1c&`xl4|`yz?jiTE?D>vtZ#s5~ zbfeTb?jzprJq;P;*HV4)MIEx+?4+j5%Y^o2F~QZ>a{8$)Ms}^kED)MHMG~F#%y#he zQs=4uui@!^&BepkU?d%TwNur@qpZ+wwP<_!mF=8=$UtC~AA~+6Inq`L8 z^=?ex{H^wk^OQCIQI=h4&|K%d%pjpgbdnD~jyHH0VYyqeBZUZ=u~~k=lVAr8DrdC6Sw@} zQ<4IMDe_yApU5spUF=vTY!@GB+U=-#`!WCV2w{WC+WR+#yvk*?Gm0Jn z?Ro9glmP%|!lrXa0V+IM&Cmqwf0`=N__ zp-5=Vzx#GT1?2hopZ)Uuy2vA{>Qc=+&27D|-u(Rh*O_+rB7~ch?R;)d0A_1XmCwO9 z0(;50hlf?fNQOp}5t=>im+#0E#BD5Mx;KG>?I0 zo0uS&kZl6OIV#)=VuIs{-3*=7Nhdob$0|b3pXDqE4gZl|fI5%;tX6$_LYF0~@91fdXaOVOgQT@AjUm@l@x`7bfQ%yyJ*Y@fU z&9#rcqVs^o*s>+Y0u*+gaP$2HG0@di(QJ}?W^nNs@sM8+zSKDt(8RUShIDjxvIqy6 zOe#O6P$PL*7eJH=3yx4iM{+bHhg8=^wU18>6il3C+8c8PwOk8vxud*>rE)gCHO+^w zBWn*D6gYPFeQGLtJDbLG&fY<+TDsj+>nLbeWhzc z${6%qn=AYp#0;GadVY-N);Qk!e;YD=8@TJUexk51LTH5{!T`5PJcwsT=3V|xRHGB26iMc3wQ+w%l@2S+I zsrkY*)&0-BK~VF*@{80cpVs^T|Fb`Z3jg1IB>YUiyx!J8k0T9`1zf`WWP-(yUpM@v zRp)=+d#|uZq0W;{d4N`Lk`lrOuEC1_wbSf>OfT*@9_E^< zLeu=E8T7yPQ9t{x{%*VtsWou$hY#ESC!2pdsKgyBUMvItvoh0%(SM^2|IkGF*SzM( zPSa(I$4^UXr}1fP7#+C6D-*_~9f=j)Ho_}z&TDj{#f68>i zX0guVjN(j85p*>C4ZpkHnCee+cHrt4-9Qg5@KIhtj~+k6_7Ghc9=cW)?_uoWBXTD| zlPC4!zc$PM{(Q9H<2^{^f~(`hRRTtWo-%}Ow4}3T_V#+9 zH~Q9rHyRomnSco{A}mb(JVo~1`tptI*Smof-8kxwtLvI|x3t89Y#8h|#k;w+YWa(O z0i5Olc3%q<_4i3lh}EcOI)icoj#cW0sOVZi4rieZsO4Yg+3Dh2_9qU!$Y|8nH#Vpm{Pgns1zJ?&-Wb|besYFnceA+urHV)JUlC!edHsqq+u99 ztQ{x%N*v9qTEon1if5+V6OCdsfOyTo8M)Q91M|yYyfVGdmt{T0)x*kZ3)N_pjj2Dl zaa(K=dat$bXA3jqla?0Q`=`JOjX(9N)yG@bR$HOjzQ6&$vwuVNJ{8Iezx8`pr0GBT zA6i9A%M^xEs_;h-;4r+A*an1wiGTvcb)?!CPBSx;ie_Tox|HdQL6!q$fdZgoHId24 z^(fo0A9RZ$^jBq{EyOc&NwfiA?*x;`cN_9c7C!xA`=L21(+|+R*8@Epb`5VYNH(&Q zN&vRiKZW+O2eNR!KrUA9<|!d3c+~nD&HY0E`R_go|$|K1rme(ImlH(P9P88MWbRm$oH@> z1&B*z8-gmHKxhs2`^S2~6`KBNnN6Z&@wBvSCRH@O`R;0n&(>F1D_u}Ro(1+K`$Z~i zhI14A0#&?^UH9&I2}BM#2sZF><6;vGGBtB*H|t?^BjD&|pX@_h_IJ2{c79uZ{8j^~ z=ldq!4p{_bUW({PxWNrSP}s3D4)i6@4NX<~ob0j!ZO!N5TD5csS*SJc=YCaxvu}5f zOK6n(SuF3iW>wUX&4eO*5hwb!jn_E1y2Q-ukFl?P*2ytmFB}oS_M5>0yJDIzjVH^J443+Xuf6vmUrPZEOxm1Rkjk2t50rEZ%I9SV)%8q zat+YK=zEmOS01-juabvy8TTx!F90n1QE&PI1&bElx_!rAcmcnpdagDdCWb z<_BoqvOlwIE})>sQ^no~tbsL2Ws{&U@<+v@tTU^_OnO>#SsMUr_E*zu_=vCxN!Pg# zK)1R{s5e7L#31*vFQs^PoG?eoa@LdQt%SI90hAc{xjo}S!~CxUC-&b(2K?YFMD9^7 zsS^LqL?@ul^Uk&N!Hv@1IL~>>8yT;y00bmFs1BqvC{)SLh*CffMRf3?M1xML2GHA4 zi(A9#JXhSuE941?P0P?4xDrwo#csI5-M+M|ON%N7GU!QfcYBO5@6RP>)pGx{OyLlI zd*01;CONXqX=j_{Q_LV-QVmla<@oftiI9VtHwO5VLCMf}fGiIZNR{L`hG~AgC4$NB zK&31A=L`QE3M9GJE=||P@9aSY_nnO?98iw10{u--B_2-@5^{slm!O~6x&86GJLEOM zay?k-3P>Jt;X1}MweqLUdHPRQHPPR?#Fg~XZD(0-x3jV*RSi?; zZc?0PbO(Vhgs|>-F$WJM&n?G53D@iW2D!ai>ap&+ET^fV;q;zwodoO)JHJMGhAI*R z9!VY$Tj6TuDUTZP~v;#`td>n>@gl%xu=b+hLHO7!v;Bbo5K-07!Bj83WT0)tzRodQz_ zz&8Ey`u$+G1CYZ=B0=8&ki8Q3HszxgK*r{tx}FYUHmry8xSzjCXbplOrxlT^ z)#F$p7av`c5u*E%a8qJGQaek3!g92oP!>dJzbxTF1Y)esSFvFt89n7dA!Se$30J-8 zlA|xS#D91NvbzXzt>bnpc^qPSro-FzQ?>fLiw_1}*--QmKA>ZN9zh#>O*)i0)@!MO zZ+@p3z3;BraX@|BQknBS#Im|wW)}R)KrC0x1r`1`2@Ao)<-i^csfC*XZn5m40tn=c z2LF%9h8)t@yEY=-_f|EH=#HK`R>fxw7^RBuQw78A?g2__7o2tI_5XWfd+aWl}~+sd!{ori2UhFke`f)(I^*2O%D5-2bmTG-Rxz1Qnof=JG3NQ z!W!eKzm%TUVGlYHR@Vms5w3aBY(^SR{JYd=GnNuOJgN-^NP2U$sL zNCx6_2INiSeSkiY@Kz$ndp}~b4*^*TD#y(u{>p4$+TW|P$@ST!4E$dZ6I4B)k9Y#vxl5|QUm{S0$m|>f;T(@H+I+bC9yuMK# z*7zy);*7Nh^=OaW|8?94B)1_6^jyF?VieOeYfoB+#EV<|%{g zdWn1)1W4`C)-BefjI?xY)<5`#5{p*&?S7KWfeP7YT+h*mpy#xy{o;-cE$Z5Vw|$z{DFl`9%<7`ZR99=lYIRqS*L14KzX$EReKP&=NIhEQH1DWYtp?8e zO~4NEfIGV(+?Udyrpd=%f^`_$cSe%h*hNh)&-C2@MH7$_jO?ocF=9tm*YUI~D5;69 zF2aV8v$l2wp;`m3tXT*ptOOUuF>_H197sLHG_9i2EKPT3%u0}G`=xn?vGqHLQ1l)z zcrlQ|`~gG6qbX|@`s*7~>0fVLxbVGj;thw`U}qWlIv*cE0RLb8D<>pgLDgwdU{o;= z9)F<((tXGGs8A=UI4sDt)gOCl>DVqx&6&zcL~2Oz>;+Hr{&B0Ym*NXbKzR78(C#K7 zXR0K%0^}X=-19kP{z4*%#B)(E52CWGHf@010I11q!ltWAB4)ZXB0c$^=6Y{^ku_&I zszD3(dhP8vUL=M3@VnR zB<5*--yd(e;FbvNnmtY`#2?Wh1WzK`_6RiJL<}K%`+*{lJ(b^lw46Fvvq&!$=_ht( z->n8}RT{wFG|wO75AH=Aesl-fyN^QY$)6SP`m)jLES;93SM|wWy)INC<vLPj~HFJ3Qu|%-V#DpW&U{&EdQC^e_V|?uO$@& z46@Sm0p-Yxds}aWOuQEFJiGO#E>gw!O!M=k2^YDZcd@ZBz-8!h&1NQklLh%eossAkIjBRT9xxvVe7);oVAL4D5% zbpi;A3qb};qQ$D0Qt$k9X}rNcf`=A!4U6iUlhrp7yFg2r10D0U$~EUrPMU~wcdE&5 zoQICo=McVHb4$6vnN=8jE@_V_Pj6``r``=0$R>8#OQn{%k$9M#?!n-RB6Nc- zpW9 z<~`S8uJD5DcJ)0dF4W<7SV{Iw$!AddM$5V+skT-Gpq4nr>4jN1Bsy50ZBeS;_3nbf z`0}0C5pm;SWODYyGWYwyjX-QJcW@^rI(T`Gxtqa{@Eglh<+)+ZTpKOJ zBL##~;yv8qdBJw^B|^IFhMta%5r?f2ek0Afv$ZxExUjxpge<_)R$4*#<-a>Le*^df zbu~{POkDUaLOu85Gt!l3Ia;3>1I7GmLLR&OlM?|2ip*ZdonOaaePYlaRC0`5`z`EK zIEO;Iq@os;uuz7$^P-rL-GRD{h-^TrOmB2cyAUL|E{!$iqx3jf~$kg zf%N~xD`U!UJk2j^S76l|chkG{i0c9hNZU6*=}c8a!}sz?;etdmtGGH}yi~Bi+QTJo zQ!u6v$1;I@u3O%hH~c|97hv`MT4~Z|QdVGsP_lstB(6zYHk~2|2o?s{-a2sTsKf=( zH8tvEqgPnjnC7ZO$>6ckl;giyQX$oMJwI%}QhH)ePS2w=P@MWlEmS=d0%4d900PAQ z44Fca!g49tsz+wa(=2pY*r?Q{Mrvf8NwG|$)3rGKdrLPDJ^_e15c35KAx zBNyea`-h<~D3Q4mVSEy=!{#UA0AOby3o`9VUE@y)k;Wbkm2OlM23gxP!?wNi?RGo=HwKuQnXfe&FmVDW8I8uVk-qAYtPv8zDEB4vTT^i=jOxA)x2WyB!M%zp`M%39TMG z_8{!OHM~AwPEy2+75Bf$xlu}r;oDR$X`yZjORy#AP+L_jjNCH_h*X1GAAdPNM12!|GRh9Vgv!apdr@>i52zsEwEnXn0 z;Awk;=34SDWvzRcjfP&K@P>1B&!1*VHd|cnjb3JZeh36M`H0Ms zm*~bbPhEKfBN0M12yWmfO+ngr8wec8fkR)Mt8OeE?eEKrpchb1co@bUK8+s;Vdhmq z6kYml9Q|ZM`UIP*`pBJ4i&Y0qm4Y;Fw`q!pfL9`uf?VR;#`L1M>93h-W>)K~!yNc% zGrC?NC9_ibWUZATc&34>xTU0-IWYUZ@tLXSDG&n@I5sh@QJLLs+`ntv(RT;%kfqaNfb_Q3R<4M|D zZgnFU`;)H&)P_*2sIN*KgUnAW@=?Ll=(9gatt3q`%6lP$9v-g+dG@u^+4PKBvdHq? zEASPd@QZe0ek9KmX#a#3We-TiHTOen#5=aNAdMZ{WS>fnD5I=7Z0 zb|{!#++-`76hUTo?~)mPEnHS#nG@upbSQQV*nYeoiJISK^9Y1Qbmq0$J*ui0P|aBlu3;13pcNkhu6gfLy(So-1v;7>4I4Q|brQO%_v>4*-~{?w8bF zSQvG%`~FG`S1J~3D+OTk6Fv>CS|jz5XGLA!hleAUUR&1$1p6Tvsey>(>`~3BFH(dq z`Om8I5Q+Ax)ui~5H}bQOwNf(`u6dYX1-cUW^kyn~`7p}t2W=R_yIcDb`(Lo8FT&zm z9wN5VdGs$9bD&m|ALe9uUtgE*(~AWTp%#jqJVWch)tT7k_}Pb`J8k_)wjgOCq%noe z+V?UhO*wX9AVn&C``T=ktW<<-XwsDsd?e{}w`S)>?YR!a;i2mqH=T@0sAs+x91z7RZAA)Ud-}yvS4^mK0qzM+Yvq=y(A2;dT zuB-RhSoQz(0n|b!D5F6Gaq~W#qPKh0$!1graWWFHWz`aefghhV#;lLMHu}u3TlJVS zZ44N3jEFZd8^H!Uehk97?8eV(;@l@WC=~qeys@Tt*KizgSo%QrYtR!2W9Lz6s@HJ% zogSN>%o)pT9Rzx=OGnu%@N?Lmx%}2Li(wB!7TC)>#-$31Rg8d<97Ywl+|>@v0Cm23 ze7a|Di&T%tHZ`m%$IJorHr?|p_ySk#^iHV{K9Z_tBK%>Ew7WwZ4ITK(EH}5~4G_%o zV^JV>c1&YDu|3o#Dn+NYHsa3SENamh(A#-bE6TARLI=s!jHc}35xqwC;#s%r45p1s zdph0O4C{Io&5lx~o)*>+JkwqhK=`({?tUE~#)qATnq_M(4FVDnMpjid{8v02q;mBv zo@?{@$ygm~|Gl$Uh!WV53WpA&8$gH_#LP|oIl`<7d?bLw?V|1b_;qz45=*}9w1Hmh zKSUBkUrDtUIE1UHnbEt>Z%%kALvbCVwEg>pSMKPK@?n_6aY5#Onh_vWZ!zhHiiv_O zR}zy1kx70x;GpO$D90Nn#ppea1x#!~CZiDiMS9jdI4IntUCKK!Ze_y%Uiv$!JQ)GB z)X;Y62%6`lSCq!MW85o@X9T{{T-+lSIX zZol}qh?dG|Gep*+QIL-Z(wldLtl%lULt;Q+{p}+V$p>Cz9eDWkhupFD$g2{sTY3PunhEjR)rIlReX8WY2jgRmp|DseFRd%KaK)agr83g1yP|YIE zcIKF1;N_GhJiwK;fl@a4R^3zbxW7fO_Y6oTrPw~VUw%Np!EZN?>&?k|_{x%xiX9s+ zC^2(p0cBwv8krvK;xK?#L&||Zw~PUeQS{)vj0nh;ctC~=z?~R|m)08~q3;|@Y^yXb zgn{~Ht$%EQf;#-HJU2> zxd@oU0FS7fVF_j%fkse@r6|3t z%O-mEOpm9)J@XO%o=A7tJ(HeQ&CVFXk!0`HnNGgJ$-Mng{KV+%^n4jw1<(VFg$K-ym$&rl13IjAev~FQ2Hac zikum^+K!Yw^m?u*c{5MGKU$6F-I0MY`52+o$dKz{=j(8#j0%=OsUIE zhwx$6K{fT&pXe5kvZ)MsGpN0R8L%P9FsQ~^NAmqvl)8AR-O@^G$orerd>J}(d()Bb z`{-cz0l*YyWMge@ZQmp>U*v30(dxS|CP(|#Y|xC3e}nqa(}@UBeby-h^m~&(SVS$s zFdEGX)7_o>247r~ftGa(f(=A9)~u=3VQn=!Rxt{^wAb_1&5ith6I*2|m$#BW*QZ{v zUgMaDONHgbk?D~ww2b;X zMu~_Mga&gd)hebaj=Ql@^gPxw+g-bltb|fw>6fYPa|3`^4e)o(|1wlVZMtB+0|v%J z#lvLh>z%&o-;lT>)sDCs)u5DL4=}!q^Oe1~v9Si6KWzLvVUS(@3K%q-Gh?sf%)(*! z^71`ucLan$YKh|5-tZNWSU0u=4=b6FN z8{cWUndE$DRfl`)OM1{6x7Nf%?t;uBHh0u

    AO*`CD(3x}l}StDfFibf{A1s-Njm zncM35oR{D((`}!UDj}h0GvibF_?QL`#4DVg>|)E;{7;ND@lSypamram_bqTn%O?R@?;J1Jw0e0FQJzWvtvqUown z8Z3Q>9zjcf0PwH)5n1QILcWfSTx?31QXZx>3fg0OoK44Yv{tE%&a*6^lUnCK0#b$h zh@(}Mc~4xPb8(hoxFw)2da@=3?pcBs5DVk0IwOQkxHn6&D9A3#_7XL!yMzLeO!n4& zLaegUyR-1|VsvaQu`xc?;L`nkfIMQ=kJPCv^t(L}Vae1%v+$TAkeldn@w7w~ zXk)p=jYoO{#0-x8Ozi5Jm0*@7DoiYn#t-g?hKY$$+oHE^-Mi(93Gh${2HGIj4;{$l z*{R+~e&QXVz#b?q;74INw64_!`lQa#sN9y2KV1{DFE`W1EzY3 zxC(k)%0t608x&=N#^Rx&d{F#XRsy%)MYsu(+H5JZAF|KTPA&b&aIJ@GQEp9_J;p|` zh;hOgjq0p3iy*cwtE|*)iM6*{-L&WWN6@z5wVun+`#Dqg@lP@&=IvmsaT3S#1X`4+ zm>3MGmX2O&2S)EJZezx>nny>C76v}jx&c#^KMR9WUdP;Nh*ZgEu$I=h|Ne?ivS<{i zn-BR;&*kn{Bjz0zE9X;X>)icF>=Y4K+*+9mkFY%7K@h+&Tip>f87I{EP(TGHnIjYx zqm+6W1m4KM^_I50s1D1C-`k zyc1-7t3gp}x%NK)>*%eLA?ibMrTr}daeEGr!6T||WtyLkhoGgnyDsHyrZL%MfPhj7 zrn4J#aq0S}qGd7pv}R^*sc#T^n&)1~$-ne1`{dqjVCI_W5kJ~TOY_|P1yBY*F4a*z zwz47F1_^&NpY$Mvw~cxpPk%G||hO6>_dypf`%UAO|x%XL-E?QIj1z zGfKEotE4LpD3sy40Q=O7oUNA_8BidkgWkOm zx4Pb5O&f#LTU7vG=HlkALeUGQacj{Pdg|*(3A*J1hw@LLhXIJ^fvQ0{Ql;Li&_}tTA*dXbt z_gb%7$6!heVoZS(vpSsbPANXh2MGdg?7#{}u4e z0!xK$dVkIawSC945SR(BM^t0CN1XM%22BIx-Q%D@l&_`TlIaz^s=#af3smZCCR`ov zy{kE)Sy_N;6fWehOzDpTyE^Z0Ve7x=FkRM+b{j@cQ7eCCzIA$ZOuCkqCxN$!IFtW{ zDL{6LxwO?2&-^WO@8I)*Rux(lP+792DsK5k0@SGww3#HnPz5Xlc0b;52v;P+@4+oO zF>fqa+b_#UNm>D++Ekb4lu%XVo7V`KoqUG03MpfKD{uxovETB~!dVHgVdI4F#Mcka-GOL8j9E(T?M6;RC4xXatCvuF z3uQ!#e)XxeYy}I9gVv?sKAe)*$Vq%X4LrjfXvA{`@yiG7M3Hw^j_AQ&wH2?22DW&p zsw*&Ex5m(5;`yHkcMdxr3I(6!oo`$PBYID6>Ku1xmKc7Knn+m-2>Fv0ZFp`C( z<)mJ72`k4TdV4X^^Q6N6>tnd8Pi_IV@0_SRHA`8=T}2M?}VC9TU=zn z58oRcK%#Jx4dJe4wS9XQT)+K4yFQDAz%!~6G3Yca08M}aKAS#tyAZDjs$~``A=SO+sPk}_h@*!u!nGv0WIV2!2xNkwCv@)D_V`_&wU+(&Q35br zK7Jm#l#^R~j^TFwiL>+bhpvvz-@O@dE2iWavm>A1L4}*&h7^A*fBUlhgsCYWGc4vD zH5(T<^D&9evV-hwM7}O&Ad6Rtk>6v2qM->$Jm!Ap)ZUwC zYj(VR_u4Ok0R^w;a8~}qP(a|VJa)8}+ED}!F;`&yftth6#T!*241lHlZJ(UUnLW0z zy^q1?wDh&dsUJxA|8b?m(6dXouI#<3;J3Z8f0SAd{^K>u|L_gxx;4L&*F@~`i#RsZegu{V2fhCcSGy7k9RQU3s3 zi8X*1aIc?tJoLwR9z!XN`##=#GvFX)nf8yH+Pi$Mdi)tKs$kShu|K{O-2OS{gS>lh zp6^4890Ld7U$0~TZ#_t=0?v`ODTY68fA9JMiMQk>&Vf%(SpDleNBM8_99G(UQ#tql z=_8NR9<7(XvMphD!f0-t1>EwHv;WyGPgt{}{4Q~9pC<5uU-yLNKU=e_Q0{QJbq*XG_`6p~iPx=A&-S=E+u$R9pCuEYGx zF4p?*9_L*z2B^U9Twm!u?QDIsbS4>lX)}cPVBx&_&xH6;nN28)p1o)_L-cVX4yvW09p3;Qx$AZv5?-EneX2j~E1_ z)Wf7tvji3c&JNs*8 zjGoE)_F!Mx0PpFA<7$wttWL56F%=`OR9q2@XP+MeOT5Vb%UZGc0j7y|-u?>re9%ia zp0~?#7d+(9faSj(&$05__OH;8Br`IyV>4FOX|o)38}wf_{JG}+fHrY@3^(7mrr1kx z@Kh6JY8T=(!i;rjgM0kO{nzhH+O-DJL~4wLqaQ;A*gf@{3xEHzdlDUp8UHp46W7}p z4#kl#ng<%wuXB#CFTQhH>2oY~{F(Amjgo+qMUEv|zn&aR#(pvF)b1}QJ$FAgphQZA zAR_QpKMndz+@t5VTtjYpt+xR+u2Y63E)7}|WytLpI^|Uu9}^Uh%>0`nwpQDL-DZR0 zXatX(qrd=VhK`v#>}ybnQt8DUEJ2D`REQr~Drgu8n1k z(GT!UlDIY^mtzm$_PN@s>&T2grtq8eF9(V5Jv|hZj2QQf-_m=pzuKqw>nBk2c#H^y ziBqWXbWtdMH>YxGBO=*`n4(c?$vZ_-`r@)7AciMOyO}o@eC#OSJ*K%&<-(^3qSReo zUHX}JtI)wRZ(LKp2hoE={7FIAXL3Tq|th=c_kMDL_p`T~>d1#M%ha4@Q=760XF{kN9(LRCOwPzX=M zt4FpLjRhPUDH=v;>Ny|1uDK?7?R?h}B`k|r?A9p81trrlahye*E_|z^X?Zs}Dmas&R|}510UtSHaEm)YHF&Hro@uJ1%9ptp@@44*=szx+5l;NTTZSbuRz5p*iy@Z z^ln1M?bspDHI+b{)mWIzJ8|x^yMa*;EGGH#CkV-cNkIDJZ3bNLl2&?|VMwNcx${o# zg-hCHuij_tVy(82?^TB=D*cUx7T=~9-)7&k=D0G>YgKM*vHsa{Nxx|9t!!(ws^{d# z4VZz8i8ru??iZY|@~USU=c}rscT^`u49{J__P!N;G~1YDEoSk(@w(Gem-YN8IL9wf z{Bo}TTR(Ziq&5tOoezk*i|So5wGO7gC%~0uRNSuDs+MiB zb}wI>bT_5Kp>br?>^{D*qb(6CcdFsDS?};$VJk;HjN|LxXT~MZ^aQjW&U6#RS2-|{ zWb$*R)@hfk;fMRS+-d0FjQvvd)g*B6dA3bY>6#@i_v<-dni`IU7*p04fW%} zhjXT;Y35A<4NDF>*Ax28TXdF#V?djm>2fg}YO*Q#7z*s(5g{%jPUL}p1uvV8c-ZSm zVW=KH$r`QPuJOoxpsBS&fHw`URb?ZSreB^%UYA^6nU>aVe3UI{BBaY@;azNxkH}@u zS4p+643@BPZs=co&GGu~)Vz1odUO#~+e(TBsf|Pz#UF)8D%nf_tlRbkH@4&VpT$eS zActu$UD8dDDnO>b1|3V07zEhhs z;UBjhtXuFiI@TDg#K%P(a=UdJGo(6u4K}s*qr4Zd_^8z zld{=9uo}I!CZe`@w_|Nc*lBjfEp8&)uv9?umG;}J?Vt}+v@S?Nu9+Qvjg*w*DEq;R zhP``1{)e&O*aL}ITWPTT+V;Gkd@@8A2Fb)a)%Py$_ZonKZMLU}4CWMdsCmL6 z@$|#{EJNs*lWAuV(S?y0g@(g(bA1{CBEky@Eq2ms3Ztkb%_vMv1IEn@=ooQaw87g z4(xo7DK$2`1w3NOe|RIvBcq>hS(E6YI{}GWrNawahrev+6G2qKH{{9qK1uKOJ&LM~ zMen?f)8g2LQqWoAwK!qOJ7qUi^wqXs_XyN2fgSZi?ghNSX>}9)zQ9CZc*2 z^{Tx=pHm;~eydk}jWHHxhR7}rw2WQAHWk2LDocEUby+v-yS+EO(|9jJf$aixw0PXj znn8tZ3~8JpwZQRlAT48D5cBT^(Z44WXzfw{vFNlDWLYQlsl`0ffy#)` z5Cd@Bj!?W$|0NR{Gmy|oOLfFDY2+y>>g8E_zn^hSU_+f70M^&78Y9lW7kBy_`J_`NTL>8gnB@tJHts8-9GFW`|XP& zqsqm&^dU+|$%Lnp>r!P=)#G+2VDMhPIW5-utuHttYzWhK$j^Uz4@{ zCsSZm)bmwjqsV(Q?WTn@d03`1$M1qb{en7+dBf79FJ2R;g9=+b#HF_T&w?@2u@pKG zcs~8*oGow){0K8ttEzo+m)*geCj$_Q_- z>+D-@)(6BpD)67?6pzKRJDoprb1@A~dgy1XzFSl{ive3#}R@HN*5Zm;@>0* z>ekdN+V;R!v_!vbw>NE#J1=j3naHS^hs}f7lT}3Xktln{-02&Yz3h2edg#{h+{pZU zt@{)!0}`F5L>f&rro!L%zrMqPi6E9YI?vG8CnO6d%+U=g>FRo39`35i`rj8r_C4t(yb_BXM zHa*G=brOIup`qX^EG2UZj67dR_ujA#;XWI%6-AQ&V_)wbHjtT9v%Dw(E;lT) z@CBbL@FdG>mevnLx<&jd+zAmN6u&Y+7VG^iIDd01Q}1HJwp?Q{=Z(b6 zp1V8V%bN$#y%3qr8asoCi*OF1#z zCqr(AQe(GC*zh}UX$)f1RksMX`=cLSZa%aboHDDP_vQM8+ok0Xvd$q9AUb#WB{@U> zU+leiRFmo2J~}gwK}_pNo-IcuG>4u8xnS@81aDR;f@>q-kpX4|HnQq@<_ z=4+~U8B!ItP?N{Sty91_CST~Vd;wS4MD;5$Rq~_CRnHSpzH^Bl3EGO(lQSLoi-c3r zw%y6;)CM^9r}pP(<1bIAe|tE6ML>OEA^98vjy541B&?2ks^4zRCiBX`5AupasvSDL zbTK~iYfq+u(T!gaG)|>dYocL$hP#=-sCxV*5coXE&+hu1389f3su|y$q?%UeA_T6D zmJjz+4OPH;8n?+q^XANmT7Ch%=qhU!H@k|sDmliME9R1GyW>W4U^L?u-doL5SN>Y0 z=@JoQkJ#CzbYzonk^t2@-yJlr(y`mGzeR=evaYvsfw2{c6sFhH)PV8P`1Z-ZRw3ZB z5y+UuS^F3Iw!FHJZpX_TnK@49)>t$<C>qm#wsqK`^|)@uRDnWVu6dY6 z1af(p&9do0n`EgdjJ5Y^)pH?ViuP;sv#_pO0~P6OZzzM7r%GtQo#~YZlsmt9dtViH zJ=I>Z5SB{&P_#L)OY}tK{9diJ(XA}Al>K(fQ^rnV9kJkdB(uLF^IGE}s73t4@P9p! z;9tn_Wa_cgX~9uPo`vvU(+ypc(B}FA+umj9w5)FS*;KeI3~M}4CVf7BCHyk`x8-SN zY}I=VRvT?<=C#=Co`+;|#C1;ynRNnk+1y0u+2GP_W&WVah0sE*ige^^Sjd{JLSw`s^D%pk?3fpCXajAnS_PPP>FdX`r!TrEVzaQs9PVJpazTcrZ z`6D-G2NU?WYXlPWRg3qu!^6KPqJSP{pn=lVws%U-fi1_ z&Dig~u7ZMQDu`3QfE4rV=cd=^+CCIoN_WLw!(7pQ)VggpesbH7_$3*E$d(o`DAZ2g zC$a>=`Y0U9#S^BqvG!`f8rvDnA*Ws<>6%|R zxFIm8nDqIx2gY@*(lt{@-h3ja)eV_V^XQj{Dq|vy{#3eDG`W|00l127mGyklM^;F}onC;4q!pxS8 z&}mggkMWXq06_DL4jpzy0o@?hRXDEveeiZ)(`M&Qd;E%GZRjV9CO>V^=@NW(RW}jT zm^!kvbZ2`eZYfe-YPBc(shU9Nxn=SS}9tn)i9uEB|zg^C21a@D=Bo3IDa{ zeq+bZaHY*T7kjgZe5|S?G;IiSM1dqnrD^hxy^MmNVm39u3NI2YEL>($WA-$bUBGp` zg@FTt)m-VT)#duKpB(BR|2Huj65jVIK{Eb{hXR-Kzw|Lo{GZFLdYX9MMdc!!Bg)S}R# zJrIIS8%}cny%KU;S_3E@y&=^=br0?}?vd{Tc_D#NCH5-@1z}g8o#v9D92uaOt^aw- zudqjid)VY74GCiERf|F~9p|iY1CcpUhPz!_(Ba4 z*2=|~Ju+V1qn>ju;}0nCx%N{0yhVe;Q*{EFRN2q(-+AqKF=Il3@?VQ-Y~)2Lx9J2hvMduu8@tZ7 zXVu@)#4$~Jey*`}dNM`EJ853f&``|2pD!#}CL)}=e^fnFv5HB+7-!hs7_#qW=-4ik zWFNMb|EbtOmw{Pu=6bq0%%^<~CDH;x+4pd#X81}t;j}=9W6_@-aG?ptq)8;t^5Jh_ zJGfwKkOyG_I{Nk<4-whqQqzaGGleg!ZY-dOET5h^^GVtumoDnG%cPtqrBXL(4o+zPWr3_N&$W-*||vJS3{ z&zz%VqY9ER(pc+Dh14LOPmUkeP6sPXZ%#J(>$bv)(T2tPz0dVQi_{yAy26*?N*Lt} z6^e0u#`$DQarptSA|5`5_oWd6@4flLrvKJMR6SiF-4S}uo?}0Q^yok(vh!MYh+X?c zX!CM~=r(d-F{M92DEWX?CBn9Ik^lGW$a5*ok&_dHCmeS#-MC>5U)gayuEe%+FxAnF zU-m3sYVLCO4Ajj~v@S5k0}I*XjKoGuP)D{#MvUUC=4$6x^YsT={X_c)t#8)t7qYx! z)f0KL_mZQ=DR2INLtE;?5qQB)xY9^;Mi@ z^?8Z5!=;v_d#wX(I<)Fo2%Bxc!Of>9*@RE92uTh-D;_Gg-Ye>pvRVGce$7plusQC# z+T&C9^ai|@SHO5y+nzR`obw)fak$iUiP78Vq-rcD&x8I0gTuq^lI>fYij^QN8>&c=<5Mmi_Bo+@v>l-OVm$K2Iskigo)PouO*UOg1~~g(A{F z*dsF!ut>06%XhhLF|Q5EA2NGqtcuBAwQLH#OsKN)N2L4U@=S>iHYLI5_OMHNYv#;a zDn(ukf?Y~h+B`yLwMhhV`&6oNfyGg8Hv;+qz{6X{T}BFr@~g*n?u~NwRJ_5)7}Icj zfUealtSZt} zhk}ZZ1|azC1;BKt;mX8=ekGWwQI284s?z88zZzA9;Lj@^20S&*M$|FfBL_D`Cfd%w z5j~B}BnHK#k^rxOU7Cg-pGIOL{K!vf-gioD93zlBrruqfhkLq&(W0n~*o5jGyEvJC zDL1jDN5U8|*TxT+0wqvGG1 zj#s`m>3)X(GuICHNN@<};maj=(>j+c=H!Po9t&^?S=L&3Eec&W^M~+U>6@z0xFKmx)fW;hH8DdwhCbLBOh+MDX~lB<8$-noi)D-sCUT4P&w8 z=uRs8WpGu}1%#DZ+viTv3lkTe7V_U>TCR)mYA*2Ejh&!^(dR+U;kOg)AOt8Sl|DDVn>M z{E_&o3ZawPbcywYqK$j^P3>yYpl$rEC&z-v?T|y^ zV78Y^+~RDoRB(?>2pFuk2a;1+-Wrti$!y|&({ote6o{6Y9={3tw%6y6S9H17F1k)u zcmAwfyYAO$%&A`irL1f;2^ZpY+Dm14?a_}b{iewLOPLJN{;RsSHgEF;tddtGoYS4u z`Yy?1T>}E@=YpjgTe*{S?W(rD$}nC(iZOroLWc=wGyMka^C%sRn-Js=$GC;cLmDja{Q0> z;YfmbkK#us6UfHys?u*6ntS!N>o`*6OkZ$#*Zn>7rAced{3=R2y?3kG@s&w-4VUXg zjntrGwt_jQa@XywuKRs(T%?^Zn}j5__-=mVSXOZMYIB)coy6*0(Z2HZzqhY`{@O#o zC4Ku_V;!U*%6#n1!e+Ny50^H@T)uQ%Yy6J$l}EyRLy0aS9@UU}(ri708U)9UAb$rHsJ8T(O(!sP+B)GRJP5 zk|`UUS31hBYb=d}+%U@_crF~Vno;9>2?%DPwL~SoMY-9XEB5HycGG%DYzL48Z2rI? z$i4tV9Eb@@;F@mcHm0T9Bsw~exUz?iZT@9e4i7NQy^8P>4PR8G2%0h=HBxSsJYO`J zW($qwSmh&Iv;;{!%ua$>&oh#$Q`(nz2!ot64u_AAexH1Tmpz6<33m87M63!8*#4EP z)|W>ZE%ENtjeMzI^I}HDPoc)^+sue8&A4_$Bd4i?Obc`fUzhAosMBJpuGYjQ2{)%= zrG|bgIOttKZ|fR{)*Yo^W9K@Cq1}vLsx#NeX}}s`3dSKzbzvo=NMeYz!}AfiRIA>4 zW4vsr-O@d6=e&|@4FYx%$OK36F1(e_gT6cF+CA<{7Dg{ZCQ5%WEem1{r8sR^^q;O~ zSt-!UbSkLX{EW8nnHHa1>MhI;OVq7MF=Is!|u&CKr-{xq#-wR5b>{)sS~=?p&zjqHL!X=F<-))?YA)>wA16_JN4 zr#4lbM{x6fm*`lalX*dO9jV8;^{$DF5hG$&EuFjF+q@4m()+-^y~i;LW1zm-F=;!vsOh~|f=Hy(!lKSKdqv(%=eE;;K>>wvh()mOQEnlhUL z-Lu!@sUfE=lVEX`AM3Zy&-YC=!L0Xoart)R-Hu{iTOFkPa>6`Hk!MTY@Ms}`*Ac_4 zQc5T-E5tD?HQ;O)n3r0(g_nN#R=@sMGoZsoKveaoxau76B!xF0biR~$#(1Rl{Tx&h z9N%g$4WH1%*Y+a&_^@Vf;D;qy25ED}`sZGG_$+iW1Lszfy4RNM%}qNG$CnO1EXgS~ zYZ>p63hKERD-V9iMr7W(sm4it+t4v4Z^-?(n=+o;c}02W)TF!8#8G+O8A9WMzDf-f zEA4v&Cej+t;mH2nM91o>sO0OSL}SnK7JoeY=b7baT+QLgJb}UB;=Ouzvr%8Tgn!n7 zxG6-y2%Vdu6gJKCY;2FP*{G8kvLfYRGaLx$gB)UyW&+oIM=b0eH>+ewc(ub&JNq+% zLED{ky=$VPq3v4gN!z}iNrEWDH-;*UecA3dvDcUpJ{Cwe@4Dx;)>tY17CVlra!~tH z@{52>kbt;n_v45t+b^HvRNr=+zkSYFJnc3BF>Z~+gqHn~QHwHc*zSW{F8R8xO;=&H zx@Xy+fc%|--)hXU3iGA)dLec{zdKpGX_7)+B09!^NLf9;u={$Q{L)}pPee7ylDdiN zR<9U!CTJYJlD=BG6K6K&L~l!{IgWqI8=Kgk^0$9Th39v3g5tmhA=qs6oP*)dg;g)MM0|z)4hY$0S~97G&RNL9F>nx{o<6_8Qc`ahX6*Irj{k&F~xbv<`f7K8!m^Fo1U(g>v#cm+G3DAGY9pZeEfmp}~wa%q#6-oj2$p#7(2h*}?K?+Z* z9Z5O~esA82!e@D_@Ci3qtJF>}0k=e*|2)G3qn_erA0t4J+9}clTK$Qr*Y-U|hXW&Y zAwEWxq>vUv5PaCdlzhH?w&tJ##)70cq^jm4?zRli$Z`LvWNo|~r{)a*N=*KHohI&Z zBm+1zEDc%N-1$`~o?aktYXYRz>yRsCkNS#-vN} z_#(3-kV~Z=^UzPY^@XmIx$L2;Gd`K90v-El!6ad;L*>>u1}~&)Y@(D#q6D^(Qz@3eNlXl>9bY=&n>%1|$zVGvyeWUnZgM2|Nf3R0o=OZsSO! z7Km}#Vzu0}z_3V%r8Br;8HBHC**WHsF0*e*+1EpJ7V#C{cJ`me$(tZNAPqr%!5!Lt^ljgCrIeG$0KI^}7a*Q|ZU8_F1i~;QW`6AOg^rzyH%M7P3Nr*md z2eIsYP#*7DHEFNSCz}}Bv>~;~w5gmGt2O4OUTYDU=d-WjCGQoca*chT&3XzdZXe36 zqS1UMzZ-1}LGkBzC5Y=oflLXHyYRt9`vh}&%e z!+(7}vM6Gm1OE`^`oFCHGl_d%) zZ^hx%MG;k8;6;I#^h`y}w%>I)fA+3RKUA(CNNT>^>4O_;BtUrRRZ1{5dhvTrl1kb5 zZnVST&iaL0%m^1HrTraoaqWJNvg^!jGs{5L8HbhGa2d*5>8pShX3?nSC@BY=R1l~}@~(!j0XDuW8%S#yQ36TDc2 zVeyPS&{RST@Qpu1_qZ&3*F?}a25T6G#;O>H@OFJ^wlIUOyWW&N4g#(wJ#bwsueyji z-%#Vz$Xd|xKl>Q^_S1oV^mFm`fP4p>4gWg;H|xynd}+FJaJy13gaK;&u-;?l@eMa} zZK~u_32OF+onQ11$DxOn_WdxRDH}yFe{@@g&$%zeYoTjM`Z&~6K@oJ7bdd~DKSd5* zLG2qZpFSC!;xBKFbK_Z~DYzH6(u20X^-M9}G@H&;Eb9|_Rrv}Q<4GxPh;g|a8T9*Y zyg`d?->c6AMAoMaIQXxx>59Qyl~IzxC!i+|G)y0D8{~~5f6y%wb5HdJqYXlS-Ku+` z?a`+3LME`z44%4^tdbTRlPzp2(cnO4$w)pHVjnwno2(ww8f3UXQMdWrk7GlURnLj+ zQ#E-^AvpL&V8zsC1tVgXoOO|HMvVr;z@Mb*#>#Q{Pr&6icUPsf61g)A3jrdG;5P{d z8-$9)?Zk;0vc6qft4jA@?@uuJrh$+)Wn?Gi(=tVO`f&uG`^>||dwhSnfM9v)b;G|# zj?q4_{5NGDJGyc`9A3x&O)GxKz_=tB2 z`xBa-`FgMHyp}&d=nnyZeqbe5Rl>HW)(a&*u5caVXzG-)!#j1HPg{m#lNkQ3_;J1s7Q39Zj z@>wlnZ*qi6@++yeTT?Chbtm0E!`uC3(X`69^6b}>P4|-k!}}P-$#g>3@4llu(Bm-a zu2d<^r!ss8@$I3AZ=5VTf`{60f99ch*^W80t+NeZ+Gw=73vgoQ$>w+Onl7H{zb5Jx zyT?}!8b-0Yg6#R7FWY$7_nnsxsmFNNMAUnl2Nkn62!&KufLXw7dhMC)j{Aad+V@2# zDnFR;nD5XNu&ixntQamg-}FDpMg9B)L^eXe9X5K7WXK0X^s5C#N=8bzv__thyAx10 z1)-qR39Hnyz?!#}^;tG`ii_|jYBikb00c+YH5m+$#A3r&N-Zf`;7rc>Pc(`+eJVaL z(i*2AWth~NRrsqo6=U=J>swx6HsNCbDY?DzMVAI*`YvEt6gWM1L28Cyc=H7~_O1O{ zi6>=m>+Tb1OOP~N>l4%)4g`gPB6HNE zaG)JYFXkH<@hL`tGQ4>ATSlJfi!)8r-6rJ$5JhHxa-j|&=H+U~3vIhi)DX=_xdCX1?mi}bv)`l?3}$PlnOzVz zTdP9_SvW3mHu;vFlXOt~AesOQ{XFW*QmSt6`f6z==Mk;%`Lx^D`a&?t{Yu&EOVZR> zR%B0~C@I!NzvGxfP!L$O3c#M5Gd5j+{CvJH{^vM{O>`)1^Hie}=-muz4U!s$T_;lU zyVOV!8hpJSU;jf76aF;=$@7>OVzStKeh#zK<`{Zb$$t6=dYfUa>RQmu8(B*He^Of+jXjTN{hBsQj~9;8C2;!;zs21C{xQ_RSMqA@LBt$0%Th@ z;&|6&Yh}98Mmd$H@P#fTI~h40&}@m*Fe*-WwCc@t9?#{*0{!B7+(Ngi>RkcewKKF| zykdjHPO(Z(b{K z29FN4GjA+CG<^5QFtx5{2hgftmf{v9Ic!We@F00g2A4$(DTC3)-89(Y49C;UvPICY1xDZh>~K%sd-pMgohv`c1#6f};i>^a7@ z6OTV!=M%V_my*{XZ3bCk2ZBU*bWK9=c ztRJ$zhqb2Hu+Uyc0uC;%P>pZte<6R;cIl>>=#a zW?nz1z6qsL@mAY{>|>r2)qzn(r;UJvFVZi2T6B(?6Fw}CQb<;f<+%2Mw##B4CiTl> z%lws^9yHJY!ix&AHzc2V+RQX^9kyL_u<5Nn-!H=Z3NT&oXD;OoF6me{va!B~J$af% zlb+avVLb0Zj1e#4K*ws9uYWX<4#WofZz6ot{Cos4dnw}62x>NPlNPOqcyGqUM>yN~ z#$C1`NbRU^#^HvX<9Kc&qZ4VVdWllFd8(k~ci%=U%r z+*~o7tBis!>Eu&i?!t*9$r<}_7N(iy*|sRV^o1^ycsGS$b6> z_XGNk7Uj1hOFfAh`wU%ComNq2{o?Zy-r8GJajo3>K zIkWFOYmEdY`P5$x92|CG^feb&YP5C*1m_%%vc(vi;6$b3d$X-xMvrZ$+0v07OClD} zYXa<-WCB+bLVl0-2j%JcO_9Q1R)ZE3nX>$BJzyxamI4=0iuFF}U{ol7lp!4gEolAN z-FHzw`_loedOlsuXVioB^UQr>9fKDDcs2R+4AZQ_d=c0jIGB@q>~{{XaIesw-1np#m0BUQmLGx?l5~LOuTH% zzJhIIbo{u?FuiLyQ~8e>4?)C2^~ zkeEwSiN>zuX->Q>B4m_9rOTL?z@D9>fC)OcnKDhua#D~4sKl$AzH7X_nP|fj*kEbh zE#%cczwxfzHmhJ87E)uOPP%Zfd9|34X_k%BxGf6S0%-#`f$Q`omk+FE^|`J=O@SwP zn%|TTz1Fifee0peMEEc@QZ`W=*uDUWo<$j?_k3IUx3_j!`)bY7itjK(jN&fUtfY8u z47WThs-8#nLQYG#jrUZeyoOc^(sW~CEw;{i{M8|C3`O|2aYfwo6Q|cVKhE&vJ*iT< zB;n{(YuaKbB#tpok_}p`w_=3Wl`*5LLm@&0zZ#90NNdpPh_3ON7j*93Ul9}^9;X2( zp}yX@o98>>hpVG~*8m#K38Lvh=K97Kp(#=`og1rhG^bq$L*85EvbAztc+)m|!{r+m z0N~8hZS8t8m|fF%Lvua0;`=W8-k=QmD^i*?Zd$h^lI{8`Mt?I!WoZVLE^o>2&Jk2S zy!zd-h7lQzTrpzUx13guXR6b5(>vhK5!dh)I437Fg*T6?va8n*m5b_It{vXD0gA=W za+H?r3^{+{t^g{QuLOYCr*|qq@NI`#y|@xs={a{*ydSh)qdN+`UK*=JZpecKDC0S? z7=FiFXDp0ma#I_hDKpnBwJU~I=^gitncQ_oqO)fR-LD}Hf_ZxF;6ByJ zi-+SJ269Jgru;2sv*a*TV}u;`O;b?JX3!e`45$1{p1zdzv8EBjlC3&S?wV${FR-`0 zW64bmU@N^gr&}(L%#<5@&(5olq377_v$OVk*hWGZCTOc8?j3db4Y5^(8~b`Kr(%dh zUO5}LVf~qM?t9r~drVVXvlFTcXdjtG7tTUg%zW^BOeT@^aprx2 z>9+Y`iQd6mAwMh!%iVgUFuk;n$av@&L)1_pRe}e z-zFxYxqctht|Ing{lL?5Q)Fx;T)O<71AYz>7_wG^RVY8V8W{I7mh$1o25%t)FolG$O8}Ps;%l2op6_4g7ys=bu5^x}^!WcgmG^Hy4;bpp zW6AH&kl`^Yw+LLLh0})(d1wBPwfgr{vGfwQ^>zoU8n{Uq~&^9krU z0!*}a1Aowiw@|wZe)$~je>Wa{SQCK+>;j!Vsp7VP)D(CF1)^C9Oluf!J{ueY_=xEH zdwl(WH}*^$Zyf;RrSq?J72|+#kHs*`N(7*{B~V!GQCmE1lKThx)3%82LjP!o{9#-F zn;~lOv^apG6jzld1B}Y7wrnLebZ0rvqB&r|KwCgDm!)0AqWW3Sdw8Z$1DRE?0>**RuAX{czm0!7JZos!<|uLp(z< z)C|z2^QB90M5$dHCz*9_LcauV>@Ge5>?;k$SHj zVepYRXg=-Fd~A}3qAEa>^TxT4fVbKf#i7RRR%5$Zt6`cDXk+`6h@!D*=pE4-mMzqn{gPn|qkNVDQNr+X5wz(w=lFU7#nz3z#^G zX?FM{JM}=^#oZS0el0dU2i=XXGzsS(Z`OAK=qG36-~MSD1*~(<&rmdNKRQ+e>c~f$qN}D03g1NG81zD~`;+j(OK^VqM$e!O>sDd$P_mDDrfdz9 zu*E13dz&rS7iV29Dq{W4#ye@nhs2EAKsvBlISX*8-ckfgSD4kI5#w?TLMk9%L01BW zxv?gH6hNe4iH(sr@lc+hMlnVqlrM9!y$3;)wp|J8K*`#zU~JBK#JnZQS3EY-qr?Q2 z1X^Dyk!pbOOB_X78Par{ZhqmT3{Lf*I=yWupFDgtG9PUuKmPW?lcn3v7Kccur6Jo` z4!Nsf_9zLk=yjKw*FDUjfypJW`3^>dT#N4LRVmUken5yN)58bUB8lXP6D&P`KnBoB zxC=-ybP#(L)#zR2Z{47;Ms$Y%u-pn-7n4%!N?{vWm?SDoNrJ>W0hBP<)Zf=d#wyV- zSep=kTMhovp1BLDdoJ45B^=4E5*fgGJJuCbX2JX>UiX)CYPW+XDc8-))CAd!dgz?k zfj~ve253wS?5Ss$I1Cjhu8r4d0))i+?%HF;&DZDdKKkKkkN26wao3R&vA90-8YiP7 z)2f28V_=9qW+`N7*_%x^R`EnmzfOHWqbo{7O_mBQ9taQqih?=NgqmG(0MJY^??o+? z*OaH_2)M7dhX?zrKcz{!rZp^Im{PH3j!L?2`f$-`i~xL$=$KIaSmmmYaa zySW*#(ZLgs$#vzL0|auK;>+F$N#Jq*oN)kxj%J6N3yC~CngXsnn%qWR7$s);M zPgMmkj7d(l1e?$um-kYi1S#EUygrNuPoNC@ofNigv{|gMQY=~>rnF_l$wfs@c@y%X zJQq~$9>O_wQhC;Izd~&MwCUFHxmF=c7;WBqattlz`k$BPF6i) z2;jImDT+^+5r*xR)}=;1JFu2L^NBe@$9*}pVbB1)guM#*EaF1m#`W32z$=!5x7 zQslH;P*B|Dbfz>RBH)BKdw~FpU+Fk_wLe8(J)dU+1EdF7AKCY$h>Di?Z;0S4KA~Fx zf${dZGb%N7IkNsC>v0|6c17=o<$y8g&P(wtXlvC`@~n6S_1QWdo5Qw85BT zm|9YbSf+OhYi|`2BJNSIzPlA0or6%}0}h4S22ho5}!?W>L0ig@vG0N3=*GW}7- zL0-ccm;LZG@(j>E5e?H#yP*z_juXrP^gx9{|53u~SHdP|bve}bL?%>leRN?ugs2@z zn(|j*wg@@O1V%=QMIq-oIo*CdnJB$e1NfoR{Q$HXDrVDhSwKJUWrBc#1~`z~HAoPa zbORbzc~`v|e)V`>eH2WA=a(3E8JuBug)3%E+IDLHmsCm0f&<Cb{P~z*3)YR^yblx73x`ttJYw-&Z|)xdn=^bGOp1;m`yJaRJCh ziD9#v7zR3DUF+QeR!$@5cEriR?Rn8P7|%9d1bAQE(03 z_Ts|t0)Bc8y8(C@! zCCOOx2{NO66oN7+oL1p~1urlZyX3J@svF}pWcDWFXUW-ybsl+m>HJmwEiCBxo_GK` zd}EWxVzOJ9bTEA45m>Gp-Oei9g69Yk#-TzlJ#0!?2`VNn-xaYu-km5abpbcMbqxuP zmHy@H92T9c_a9%adS}axoZBuib2$wa^8sygDNM4Ui4LHayAFt*KL5;p=jq9z6MAdt1>K~P1_F>Ae}p_BmD<3puIyFt12vF z&r3{f^jml0CGj zwr(l`gFg%$hlyDkbW?X#c2}Y$d;P*)L!AmBSLGnLO9E;IY^E9|mmGKCb+zXCWx&=1 z)aHZ^FMy4+J2WPAXvkJ1+Ln4G$On4y95w?<0z1py)e$Mwgy-7{B6DeV>&wjYN(7=8 z6quWn42rsiPCK?`0r!3>P|day0CiK*&WT*E&xDaykQJCJsU`BEY#E4lRU2`U1`rs* z8E}ZdFu;de$36mGUca=>hCFPA!k%lyHMv+z)dev*7+L;_@ehL-qk69sQ1rxhV#ba% zE;P5>i5c=YcvD-6i%rRKM}fU%VOX1yv_5MG*kS^?OxNtOrO0=y5k^8~8x-xkL63|U zF1NfaB`!y(3$X%(YaWZ>ZhGRRZRhT6IgGz{)sqzGPWsvm{-YKAb%%e3zGFntg&_BP z-Qhfd;9s2NwVSmG0#&V}+=54nZq6mC>trjg1JfMe4fr;DLm#LJUNbGq1r`p>u?AK} zmwnD9y>OM_pcb`0bu*|dTODy-4FZz@LLjUYXZ93^H9IU0PzLvMoB>CHJYX3vJ=pNa zyaJJON}x~RdUj%H_4rE^Q>FhVEY>4`WX-q9_@v&}ylxoz6=SN8B8lozre)nm7m!OA z8UCap5!oVDUKwK?go8-SW*al%4hiz(Q0j!S3XoI&WPKb;kL(H`#v5O*q};?@H(U^! zfa85jenEO66uDko=TnD0^@8Bs22UmiXwTX;yZVF` zT`m~h(bQa*^iddZ2Uj=M5iqiTW;cipX2;gF#;H&3lAQ88~-mmcZ&%DF?@4>o*Vz>Zth^sc&sr; zW%tUResWu!EM04rGpB=tS&(k97Y9U2eeg>)-9Hs?=aNca@v3vTMX^X(-6NpR)B+|8 za>bC>q#mnsok7Jz6>lIAQb@`L45mV1GzuO}h?r*WjZq6RCH!;clXIZ^ubJGpS#LVi zp#cFt9wBDQcR}!u>4UaC&57U70N@?ihj>wIRUXn8RY4&_qpyYG%}DZDN+@Ri^qDhK z-Hv<$iFFSfmBN1i?R#-|1iOq4(13#JkI{qw>gkr?U}fI*C%~k4LyhhYTI(L=ms2P} zAG;%sn_EW4D#oAYkWGoa+n8bOh_k;miA}MLMZD5bm&HG4j!5mL2WtC1|c zPC%*(PTt#kZ<&haST8L5ef|OhfRmkIBYR_L@}c33+QfXhjWrfv6n==(u>!1}&`-yi zno4U0fMT~Q|MrMGbDGZ-kRjNWH#j>e`o`7Zof@ey@#DdZ8-DoZs-wU(i4s>1eNa|Q zdM`8!$GYyJ^=33n_zChGBC??wbJD`#NSG8G6e6Gj28JOx5sRTE_7JQi?U0%^HxC%S zvqp9~V~ta~fLnIKqo=&ls{WkL)%r;a*=wWRY>aCq_9 zae~yneabM8WisrSLHA%>v2u{7Con!RmMftVNML+aL9lF-M|8+e7BXM-%}M==9s)8N z-@d|;*Ulp@-Eg=Jjpf|csv0z^r-#U?&`q9+H}wX&%EyU~WuNEDZ(E0r1b4~mPBr>! zFR-(<(41ASnJ0EADAuAtlZn!CtTvSL`vt^7OvQBiuq$1;9cC5oD7gRT`3;AUms!PZ zp3}pQ*BEDJ_Od=}!HH4DB7W-#9=WR<=%xrB^BJ`;f88n~-4|(@h^~SNV6lc?HUesW zTIs-7*j)q?k@P%h^{^T2ab>j9fXdR4Z^>Efx=lf6!r_R>*&%qX%a}d`6aW3Bz{OPi zy!-DSFYbTm-!wIp(;my^zQ`f%@j9+@I=A$SGOQ0fqFlCNpY5&K6&TEu_ zC}CwF(N3hIRVZt@1(qoSq6yegM5#yK1XUuO#}ZR(Zl#+pS!Pnh7WQV%8XS9%8=#u| z8bI~-l}Ans?$t*iTO&>=C3)mo-6-ZxvGfP(A3X~2(5p9T!r*gOx5h=@lGO!znhO)| znB5-hCBWXDVu*rS!hpE>uhu2Y)N?T$CSjv1cR;6fKT7dh&TMMpqf*2CeywGo`eG!0 zE1V)nB49@dr4>V%eH>R4{ZgoXM{n*!Zr`#bighpUsSK>AUHXi*gM$_1taGTYX zj0BUaBH@nt+%2Eq4DLQeZmL+a3Ee5T#qUnMG=$*1e&0CJ1Mp>>3PCmyK0-5{-Kj~k ztIY|dS$}E&pGClBW;!fq9Kd1L%VVQZfS555ZCv@ZWj+IvwHvlwOg!g-zUTF2W<+0x zj5W+1L`@C4<+r1hs=Hz|RQ;v^y&XZpFBmoDHbONJ+^#3}g2Z7j1-EqCCyn72Fy)fg z&Vn1uRmY95f|#UsUDnvlNPTI5@&(QPY7s0muFPCYJa&3+t4`>J~5WE?J+?a&SZm8-(R3e!Qy5gu4mRnur=9F@)mEu9Vh$E;F zh)8w+4HEdfiijQIs7a!Kh%NFyNdjb`EZnYgJKTA&j9vfl^+ay31R3(<_IDpSk-{>7 zp`Z`95V@8pc7d5DVJlL6M>Q+|z;%8z6X5?d{2Q%t=xLA_(fX$|Z;qJ`@8|Z(Og9ZH zbfcHbz55zlbV@?qCzh(G`~;oqU`yX0WBTxY&aIn29X*!!u4t%ssNKzhS$T%PD#PqPS#Dj@1DL~$Cwg}#g^!96J zfO=Z}R?+NdDa7737j}RR%a}ggt1BEpLb)nn&9xQNK)~5lfIV@mT}we~%Eth%Xm{|! zwZ<27lfC=pl{=6Kq_s?Wc?cI*_etow0>*K$pgCsLIOOShl+l4yg~v=QT{x_3BxOJV zp5t^d`w8N38K4Y+{u2P=TlMPr;DZJS ziv|2s?6H*t6-)?_*Jq2QwG+3@10KpRG>ob?e7xB9n*V&_!t8C|)u_gy2qaK}FsjuH z6%`X+Q>_q4dJ%81>jg-K>CXgT{Op;dt$H%;`d7&yDoj_h7oHPZNKQFQ=(alag4uz( zn^2@+rV#rn4%r(m3qtOl-AS!_taw2>HCUZUre{0regj>qXTPF8wP|;Cl$^o7*{WxM zLL4k!V(t*0M#UF`3c7K-L3=fjYnQZ_js&UZH6{N(nWa1s1DU{xwP+eBjHM(~CzlkH zot&Lxr7=VJek>WUx*AHCH4L<>oAlDO^~)KiwR~{4LhG8xUKtjs@}z!#rsN`eN6(xh zOw##q61o}K9}K}r*N`c^Xw{sOgqW*ZDIaIik(FU2#pnF&0(9;V`dXs+J+FKQJT84K zdx(*?0t5qe;jwGTk#@sb&j(TrH!Cdmg%tQa9|#FQPZApdFB@C`+gPIq6U*j9W^k^Gm9B!B=!z8? zD}~inMIeFXWx<)SFYg1#Dr24&CQ5z70t{n64z=p>!e?v+7UVzwb_80=hSiIdSnyq6 z7iVw#GPhONvhKQ{{LJ_o`dFXbX7kR92}U-Xw89!VOQejnZ1@IFeOW^zKP=H`+ak31 zWZEn-ih=k;{i6&mRLd&-h zwB3t)+K0#zuB!(lpFZU&(TGa@5y<*%WKib4t9Cabw#dG|ZQeg0$Yfk5OKg+AbTRCz zFeLVp1jNcC=EOTsMhW4UZ7S8^(l5r}O^xGcQ1Se}uI%e^95vSB<6NbS%R6=@2qfb4 zpS!kckD+#MBLR*3KT3@ZP@xl0(erw#>jjM-h9)M-K({o$S*vGQUxRsJ|I5Owtx37* zEkyGqC^}`s_o}kyTs>x5KG?k!W#F?aI_gA*QUv<}RG_IEWbO{vB=k^RRKr7%((ZwWYzU+HSSz$|Z_|$Sxr|=onmRdMBs0-hI zC4WRe!LTbR|E+ky=ie{zmDVyqK~YE(h_AcBgbybHyB34kR!E33yfgv{w2P!#K*?Gc zNb&_WMIdQA*@p(grxU5r)12}t{lPtb#6uEBDfzKm%kJ>bx_g*hvKMT(*d%{YN3Kj| zhCk9|#M0lAbM=_w=SMzUZ7hYGAxm=mA$0Ls0vrB|DtUYaopiWs1)x;vdN!bV(~xj} zH8bpRzajj>5?$ZXh`cC;N#)8`3?+-=rM6bBXtMoIbZOps(PmQ1W-tDTr!X3blP_Vd zE%JNldLsw2Nn2JygnR;4>Ckh!kA$I7n_==S6D}_D-&Nj>SHRdak=zrkv&-lP)a zM3iWoOw4+_v^Bq+a=r#_hwOo*V~3eGYQ!xoF4hfI90^<^XV z@A^LCSAM_`@76Pa5!GnBm;2?C0Sa5HZcANR|DXscV)9&g|74YX=W15?DaxYUcC!Nc zX)CP*ApaZ$@3+;g6;eLYXmQ>YWJgcHDBJGFtuc==_a4Q;;e7!t zkn)(^0c+9FR*NjvW01vw6dNMG&ki_hq{4_&fmhRqDUsgt?$gaY?N%R_ruFU#tSZmE zVTzdbUf*MYii+c!rgzJmEmHZ4e1ZSb20MI}lOmlovH>n&&uV+=9ccQ@n;=?Z@a44r z{X)gvA|Ax43QyN6?_*IIJ^kKZ~ERodUqH%1Rj154nazsap|O z?mRVw1(WXl*`l?8Tax*uS8&%H=o^^G!0C~m2lwTg)&YFU> z>$mH*TOt&OMJ8m^_WC?sq$16>F0~&%#B8PcE!3A_N8)NVT($xxSr}2?AxbI(!tH*g z)7^G0nw|u3fOPzqaZ@^`y@+G7>WN{rwF`QBKeg%hydu}H6W$^K03s8lYZjuTmIu-ez9Refcx?4=^3sp(1Fu$cwXv5U)ld* z9YZQlK+S7i)IpxCenw&Eu32+H1&)53xB~v;R(|djW^a?eaQ)s(x`E#L{mGd{oL+-k z+1I}*_vgJ!K$}TNdJjxx`*Ov)Mm~S;Bk^Cih5phSWM{$~i;I4q_mq{%xIZ5a9Lou) zB8izX?(pn?ZT#eUJ86jMApI5be|be6<(XyW!)I@ zB%I!Q6&sEO53nRBuhMfbMUxBQ)^h#%*$mLKu+P!B0njm`_z?IJwqQt`Fl_Fw?LQr8 zp9We4@MaMUH5Nw3Ki{gZEn)VSkIlW>>2-Pb!)3Y@EA$m|fE8CPTY>wRA=Dt~?BRSz z@Ckc2{HO;BTB)LfRFV`1)I09V6Qox@dm9#Y>N-mgj07# z>Bzv|cuW0Mh&*Vh>VoczF5NVyiQDVI#iP z1h62|()J3u4(9C%N4S)(m6;K>7QDw-Dryg1swUS$J}eb2H27tSWbdlxGWpPsoGILu zs9A>#Pk4oiIBv3U;KMG*gofO{q~xhV6ch2QZ&|Va5_+wC)r9L+f2fLbp7@_kutJf? zC-SlX!`^#`HGOvf-9(Iqviwj*O{uUCZO7C4H49E{YFjpX`WEb*?!WIG$7&`t05@tVa@kymcMZJ_)3IKs#(F zIX$%T0PRI)UY-DBx(tlA5~8~?vKr%GfsTIIUto&vul4evJrX~GWj}L<7{PC_qS_5O zbW>A_n&EH31k)iHX!CjE5NJDcr8!KjyIA`nozw5LfI;c869I&|xw%#0f?UVEMo0MO zji!NSXV}@VuvS4Y$vDoDYBwiYR)@osjfJ_+?Le=F-f`E7@+P;yg@EUZbg#dkRGjV@ zPs54CN4CA?5?R0GT59r?`HcXZ(^q9yG|e=!!=s7NmmTIlV&#YV67Qq4@; z=qomBDs8lQYt{$)CowxGO>ggxDgJPwNSCIbUUP(Q^B-FAy~gvX4a|w@{EfaNZn09a zgm#_rIJ=?Sk)<}-c5ZRYed`G_Zm}1kjc#L8o`G&?srArdaNcb`(VDenkIKlCQN4M%X`bt42-_Dl)p|FKEu`X^M??(_|n#E zJ8rL| zDs#n`_T(Bwnm0#T`31MiaGc$mt7jv*q-@t3_VOeK`HHg<*qMt*a2F zPK*0oHQHQ_U1{kAf0vq7qJE@85Jzn7=6fGypUtr%q7{9k*csn7q8=O+dyDrZe!1?; zijYdY>=h$&mtC4~KtwT85W`0%9TA_Nekh(;CgGH*MBnIO@2FCjl8X2Ec#=fM($d_p zbHT_25TvJD`D`>fQw6<;vYxM*oN=Vfk<#ccRy0paw2P?a=qQD_FOn+f-bQG>IkKIe zm;x((f9_R7s+52rin>wG9Pc8r$)@{Iy>V45m&n9XAC+hjIB+n^uuBm^(y8&(t+2Ek~j z6hLrgmh!`bu3HN>N*2%{&cr9Uj{7xAVix_9FwbquT#k;vqk%qW^?|8@P(!cox}(sO zcA$)6%7pH```yscKfj=^_aWlpJl7%Rf@bhC+&sR5x zIE`avH@B5LxRN=Qm*}4>NKLxl+(V-F_5z2g?d-#$!?d^Rss9R1<5=EyWQmm}dgd%q zdi5+MG8nd#Y|E;*_9%8Y8*;nk4{6*meaWKgR9P zFEppd^2j7DIVwE;3MwOFGC7QSGK#0`1{+T)CqDMa$=AXIxA>TS1X+Ev%Y*>NHcaE) z_e$>p42N{nN$OQ=KFwm5Ec^t+$1d`(@OPThoF=GcPlQ1;j-}`VUZ}xIXs;$z#Supe za)FNIaKMPu5b`NJ)cTF_^8U*RJm1Y_&*mhS$ya)^N@&bGe=0#mNoa%T65auSSaq_z zsb@N4~EH- zNlYClb$!J7^@=Yn3{@?q>IWFpnkxz%>=*|=t*DKzLEdQ+Ic(EkeR09kI>*zSHvcrp za_c=;$I8j2>9<%lO?ACG-rinef7PZTN16%q@nvdLhc11L zv%^D{+~>UhkXEt1!><(Nv6<$jO6l+q&+gD3p0>g|!TsXi(M#8?v0w$ez@5 zJHC#USOR&6>~;lHhX;g)w(eq>vH>@zll$mvG%>MmU5?bBNKL{|H+XC=DA(<*FA)>t zSKc}={V^6!m}pqKa}1}r*O;Dn)vUky)J4cEz9nLcV;Si1prl^r)JUGm8<~^g5kbIS zy<`$EsNrF60az8*qpZ--8g&TX6^mn1pr(PrKvE~OpRhWMkCFthl9A?_64OuvfiPfnjtzPj^tjS{G|4;MN8Cr#Ar^6PmbXy>t(Na==9%C;fKP zRIusB+l;W6CQ?F+kPdM=2uxJN*it^2!&-xvpE~Uk7cA_gX>AHzBJBzD*Sl)>9sbUO z%R)KP-jJr#vg@c%JT;x?KEG>Uc$oejoto=5l`IqFF;KHOr1SKs2M=oih=Hk445wo2 zPva9^J!jXV`Qf`5e2T22<*xa)8HFY2>pInMczaM@t-bDOySX_;w-1O8NwQz-a}vBc z?LD@_bqL*K_|Q}UaisEL@9}|w5|4+gNZKXpmlz-QBS(C;q^t70J;+s>sMy-|AxYN5c{E3ZKYtZ8v_l_3KX!xBXQ?jI+WS_cv&iJH=kB;e@yx zB+RQs?GbW%vH3#fjvmWzmlNMF4#nD5EVN!7Z5s~1q#CWWjVRkW@v#(V1x-0B_*f=i z$NEsK4Eb)C-*cFB#pxjZ*>D2%esLlmOML)|&YRbwvpDd?PJ%ycf~C&B@qgiHdYRy zr9m7njZ<|-S(0Fa%xA$Reb$nR)j((husdriZS0%e+iLW1D13Hy_BP_7zJR%?|8t6`Fb)w3Dn`>$A*m`?cvvJLZ)8=k|6|mL2^kubB?4A}6gGPpS^6i`jqJRnFaH zWAIdB6-SA^GhV!7JJLv7!p;pP7;I6WK_GIYHhtH!M+S{~_qgaGD;HTwIjaJy6Wr=j z<1WwH;$%Iiwzs#kEnGn%`LL2Kg;MhI8Jo21TM6ect+>D5FTLUM>?q~^###{6_kV#| zyk^NfEzan^3^Ee(;b<63zijx}k8w+{Jm-ASGy*wxtg-|q#&2A6T}3gM3C*Cs916-L zfnLmmIO+j8cDD3GVc=mF2E8_ifSgk0-_VTkSoz4k3dg=p$h1N0kZ!JE4xt%c(82(6 z^g~Qa(@&3pNYa>>-4_I($(!msmAEM@YUfR^I_7nWI^Pi6Au`vUZqq{1tF)FUkBQO~ zXo+Rs7lPNrJ=Fvmci~^RVXy$ZHs>^y7sL-CTuy%GUq;_=tTA2}1YSU!RCj1!vB z%;B5^F047~+vN+30ZHoCyY`9-d#7SH-0)VOdNB_oeA?8Li{or)lic0|Bax8ecblLl zI1-~0Jqiby&}A+@V>6R%xW?15P$&vww};kz=q(3`*jkDv>9L@3^!}ZeXEB@0T~j?S zcicg^>@@@PqyK!m_K$)H`92^YhK$6rWLwE(@>NEkgq1*yi;IgT&_*OE z$hTda_+UN1e(KkU41CvanMl1(CX3fb>g3!u;w6Q7mFObfm+_IMKblsp7@!m*Jtt?i zG^<_|}Nqfb!a6^bY!tVb4 z`^%odQtW&h&`G+dc?gXW$g1TT1op0I-`QMy7Sn%0ylKHFbx(8C5c|?6B;I4(yct>V zJ(f1KUp7l&Rx*UEB>R*J6O-w4tSkNh`}1zt<-Gd zMQp8PZoXG3zS3oVU^nWq223@3W$B*8ZCn*gTiO}Lc08=5PZU4IYF14PwLW10M&EGE zLul??`(gl8)_@ccV0F!EQW)|Y6E0?>!he=k@hQ}Sul+D|^CQ*7eHJ)dwUaU#aBzkrt*df9!`PLdZ zS#$;7Luc7(;rRL3b{6hwA2sOl?Xk#b_nN=II-BVX} zrDm>E1fFe2RbVH_cvWTCg7@>F_w<{!*CMT|BIYe_+ahhKFzW~N){3Z_v8k)0>cbKH zkJbw2>Q@QmqfDpf8FiI1=T?0M=`1I5;`zJw^ z8@SKItKP-%c@gnp4qyYg5@`y%A@pdC^!m;+54${|_>oLMo29*i;)&B?i~*2W++%JU z1d`7Z)G6U+w^?x}70@#-4Fo;Ky$XD+1GwvI^Mdsc46xFPS1mP$6|Aoul27gJa35TD zWr;PB!l=an&ryrqdB*wX*yj@(5~X@muFyiMm&G^vv=fUzbV7buzinLgz{VIPM1S%a zyD8LNn8G7k(v}wNIr|O@7<$6+##{CDI@e8ZBAvXjuw4;<#1(ejxBTYcBhl3r-a+f3EZjAI5?tv;-$@QPr9CevREOAl6^4vd*jA= zXhfruUQL;>d=rb(aT%t4kysd|JQdI1tV?C!Sp4S5{$e+))rtI@dugk@=(sxN{4KRr zr~Edv)b4yaPh-Ty)`#fn$t10Hi+2vj!|Me_9l!Nd%TOn))YUnqCkEGQgXvz96Z6W% zlOtQkm17Oq^ggEql>@+z3M7sJr4BlWU0x!^zB9A_u&kLhVdA=(%WHEeY26}Tp*h?x zu$m+tF7l?5t=~T0=VqS3({w|B2JC40&9+)K@w03=6L#RbRH1olvu_&DK)tL9wO5Ug zIW>OkZMmew9f=LrIfCOIFygr(*u|_rh@iBwJjT_o4gxpBqo0@@#yq`U>1+w2qT2#& z;nNdy%ZwO4Q62V9+xm*RM1Q4=RaR!^NBQj2SP9hEu=sC?A_s1do_cG!O_`sigAnU_ z@RKdUoU)!s8Or?AV&}eF120b6*=SdNj2VREAPyNx&HG#4HyM`*TviMUP%^|fxm(Vk zJvMkF;AaDr4s~;D;xSlr%mM@YN= zv0XBh0O$a5UNtz#s~T^C2NFeaXm$5`0Ol`%{jtY&s85ZY`6+&E9wrU1cv`{=Z=D9b zO<<1zkWTm4lN!1Wx!pf0`7vFFfgFmELN}1+6pRdO9JjJTBafDB#)Ak5ROL@ zPc9pJU;t_X&a@`7kcUQCpveH{zCZESycGasnVFfq3xgFR!K4TS+nc5hfj0q#Z=8z= zpb0r?@NrTqK|i@Vl1I(rVNZs>Ay%)VHpr;^6fQ_t2=EKc(LJZQ$uMYpf=LD zi^>t72@D;?RhL;U`dh!}Yej$O$R=;JfLrT)yRx`Ar_D}o;uTXHyCEEh-I89gSj0y6 z+`S#M8dP)d8lcm69Nj0XWD^qN%3iHVhAo6OyPK#1-Jzn$I39mR*!Ap*;D9vDWw82b z-mdj@OE}AO2q~1^KKA}uuzPD3B%~;U{CboGZ+CE#D$5HKg(TQaF8BF_aQ*%L(oY$g z=(;IFs9~1c(Bmp?I||LiDq&dp>&)g~SWuqmJ`2cf-C~zFXV)jnFnF?1X_N~uZMb_G zU(ngrH92WQnUs4nTVHA#4RejAh0qRJ?tIIl-N%GNLxES!|J$R-Z#Qqi?&-gBG=65? zLupD@RPe~@FFq+%`$r-lh#|K17kBmr0K!x&>9w3aq6i^*iol>|QzXCUWxKp4>J;ua z{RKNKUb$lW9&5nBWrS+f&#&drrSEK{=NUTVT$FESh%|8o2Mombsm)DTfiKGOwGpt3 zw58{NF|D9kkq+_Y>iL3o4?4rXnZ`+s9vlE^8V7Zp=9bFkgME-+2-glpT@#?>K(FPLpyo7 zl;&OR^>EE|&P>utwqI(h)mjcnf~mS(#v{R>UD9;svzfZ7Y0P&0{8})oS}fihioDFk zS{|ddaJw|0+NpHQ#!gb^ou0g@9x`5_t`z9BvcD0P6sNI_?CBgA&`dmKaI$or%U{(~ zz*r)DlM@`?Zvpevl0UssT(oU9F4J9dk9VNZiT*G5lspV zz!{E}7K2N*T(?X8Yu$jMmfY@Hts+;$t6lnwGX6g49cEfnz!62;ny~!z>YC+;7w0Tt z0`T_r@_!2<;LFG!Ld#54dRxI^2da2iM(~v*K1-1p_!|FjmXwrOf(A&S5-7T2a+${e zg{c1j8EA&e;s5Il^#5DQ`fb@C8S)1o-Ea5d{~|}2t;?r^w7LWARzxs@5FskZpN79! z>SOx<@D~&FDd5TOV9-l{+<8x|*&L1yKs#K+LO_(FeHDo?ZsC6re{9`2i5oXrh9vt6 z>np6~2B23^zK;1yH-s^P6eM4r!LYl3d3(T#bZ@-g&k`*CMzsfjI)7e+YWMdbDYvVp zAbI_b4dqUM@l3ho~2M&69)E zb@WlOdCM6FHCnkavMZc`sOOvf4|)sr&co=`3oIhD(m;qA37qZf>6uBxydnX|PPHg^Nwo(1%U4Z388h?dz3IkdrdO%;p#p2qjzG5Fb%nv#CEmpci ztmH1VXCd7u@k9iy-;NBzO}=yDV@JJuTK=PP>osX_Q2tGEKHd~heD(yr^P^$6l;@Vc z@f3)cw%Z(-T|A~s=edi;8}#bjrOyUc)KFd4*Mk@E4!@Ufa0Zu%)~>k51t9Hgg03_S zkv(0eN-j_MFA%KSJ3GUA_*qX$ZoaK&a*OY^%i$cz^p?pgDQlTEou65^jn@-wRqlJ} zF|B(eMey%(o96Y;QA0fDimm{17~KV`pR~`7w#`cHws8TBil6l5^?RdX+}F~elgaj= zY`br=BJWvww+xZgWZm0hmkGB)D9X4PI6USHitm1WspMy|JdP*5e#W5$RF~ZuN#}d_ zt*t$`4}6~P0Gm?>==VFo4BG*~PzNAkk@!UL{l+KS#q-40Be_+gUR(lIGL(ZI=C92_ z(7(N{XWO16k2ZWBY-Tb9v*@K^WMf{7StYrWP!982V!AA|*H85mb43^f|7DRxIs&e6!ok=35 z)z7P8!-RrS3h4R#_5f$|v=i*7vmBIKFUras8wJdw{X;m_yUjsFqZ+#j1=&{^t?fuZj&IAv9Byd2)+&C1-KbL5&5r{l%FY zXZ#CqvM4hIE&=B-NdL#mVJb;l%%~3nhUGt_PBoz&F4T|61@bCgTrSBiK!3cDa17IL z8PC%LWfhBQl9pD(?FROy8||6%BL1=j_xXkR0Aw(xk~cMBvMEwm7Qr(R{P-3dHY~Z+ zW*{Oso5iwmIns%BAh56t1lp?-0TTgcChP?&J2I^g=Z+yvd@!zn)UQ|ICQYjfY^}K) zFs;|t$z8sM!ozypwfW=g0{Pm~It^zCI}F7doZ3?Zy@h7V4C~FX0-Fz;M-Hn`aO}*0 zWI71zn&$ArW`nO0>e8%}qJN_$jT=FA93^JwZ5fg5MTrU9Rp;4{*S?C1|08Xx&rsU^ z;?j42(XzT)uv~>0K(Bec>s8R0Kwh&jo~l5oZDh7ThceVR6@e_1$r>ttC_=L+W{aV^ZFLDt8s7ZIL$dtFu~RPlk8#9B1~`NX zIitriL_m1&nq}d52%e->E)e~2`DHM8(&7x01RaNy2(*9T3#N{((-ZDvhP7&B3t~k}Z`7yCF z6cf#ZEWX93J=U#G5sb}q{U2+&eB@KQIft^TTI}>RPRG=7I+wK{?Fi=r+j?P{KcYzjVzF}vR73;cu(x)`%Pms&yGG>MsAJ1cT zMh~npI_;LC?gBqjMC$jS2){@Z0(EL`htu7VIHO-$` zijIXZWv+O}%{d+xO4i_I4-a4qiPGmibl>4{LHz8p8hGY36d*~_2>>F^p_bUA{wE#r zTP=_ga~!feh5~1R{}l!85${On;XQGNZDRD;YKM-gVjBZ5K#Bc*Dyk-n%>LX@Gf~QD z%+zzDVwf@n{Vo3{1)Gz&M%kpT6GbOoHokWqyn0qXVrxvooc)Z9yEBMgK~Tk2dVDfg z<2tYJYeg3AB$bns;f#u!?YE>gX89<`N3t8tbMOh+vyAecZUKl;pHfgsaWCp{STx8o)m~wSwc-I&H|4O*hW* zwvqq>32j2>HeDAo=pT@_v~0ek^G1GTK<()Fta-quB)ziQn0$Y<_+c{<1auu!dQ zHNrIRnO<$B>zWc)>aW;EHW{1w&}N4Hkf^$7bzh0H@!cM7%|ZIgpwqJ|>VK_DZ~0O! z#W6D_?0RMPT?V4=T|rr_f|*nX{WV)g906Vg{Cw2VrN(Qpk=x{7^Nikjn*yT$qW)qI zi?H97zPs3X)2yms1li!q*T6Cm^u2QPBFnx6#jqF?R6%gmca}iV4LQE9&Gt(o4DvXPG8u1m!Z&Tpsaa2(zqO5#+=A3ehc^em zdE;hWo2x)6ruNyydf?I{WP*nE-#F`xiMeJd0spFDfx29kpWun9Z7Q1gGNm3ch z*pX2>?@;6%W^6JQ_h|2dsdvgO)+S;}!dB|gyu9JKX^_5)(s2u5VcZnr^wOHUnfHZ9 zedKAp0RqyGO_22!YJB!1B>lpSe-)W_LAJs-S%a9VW}Oa9ok4+`!E_Puua%XRh8uoP zh{@MTTJOWSbUjozzi%?g&WcZ4<*-j*jqr2n2G%bzfPH8}JoknhZ$hwX5L;lD1&9B6 z?dG(TZG_4KmFn(X{H99uK3Q65aDe1`}@@gf1*N1G#&FAKMhqDGm{1oi^UW}vr4znG;+T>czStJAz-t6A7@SLL!){_s7gC7xExa}ldDN3*UusF z#nB=bQxmue|NLdifahz^{)Sh<%R0aCCr+_X{D;lrckRp4-K{LR#>C{c6Y}9a`SmIi zsVXN}KQ+m(@dP_-A^pEs<3-4~Upgd+vG#MQZ|O~qJ5jVvg3`XHF%H$!S@%(R8vmS*rwkPS6irIPXOpF=T{d zM#;rMhC(*Jp@!>^8Txiy_14*<0hI)4307RvL1y2O?xy;~($M*<@nUrHN!){_VWVgV zZu36qC}$>c0$7OjFC42CH-}t| z8q1vA_jGTCwAP_tA?7pWG>O4Z<>oChYKhMZy;)w4>&Z1cvJBi?@{9~XbomJ_4OSF6 z9?jOoVn>o->j<1DM$%if|Fk#&5_Q~C@EZ;&Sbk1#EMtKUm!=Z^GT>T)bq=#{1kb?* z=vm)}Wbu8?Q+^}lGmOh{=d7IKknNs zrj>JsYbn5PY$PGO3-cLxsF!cl@yq5>{(aBJE0ELh^}#*a0R`OZX<#=(vFT}PX?(_I z_E6|~PIwqhbAhWS!09-}ub_eNCgya2i>VYwc%aC^6jhKF zR$wv&>^TL+d7Q-hH%3el=+ySiwR#B1;<`!dMOc%pCT}!RRMF+ zTxYhGY0XRutjrZE%oCOC-3Cs_(-=NAHHYdk@w#o*1m8@Tq!ecP*j}~gE(}Rt!41q6 z9Lit2%hWF)V+(2-iJoQAh@kH=ann)zc$I#6IOu>D_yyn$P6t|+vWWw(saMWgkb^1p zjn*NHE&|*?u0y8mjN8>T{>NZ+G>yHEJ*Ez9m_iXqw1ad({H<2ix#b48D+JC-Pz1=( z3=Iro+Z>vCSVR=?g`%RO;l~8VJ;PQt8r8dsSn$g?-g29t^j|mKpAOV~5k*NO$8ul|GW&gaHr(nv;M9(H2P9++doDXQO=~Z&Ur+w=l*rHW$9c?e_{WYU|G*H18_N3kJ-^)+)=4g(5Do5O^zB?o zNOXLgELlL;6ZGDzjNM8&eq5saFN}};VC)_N7jOCYT`J)CwyXq(g)0iBIo#73rjzG> z2@cyupS*AN#QGb~VY2dAM&)S@XRSS%3CA;1+-tfA=H|Tz1|AxKj$iteh$#FG-Zmv? zJ~6xDI2a)OiR9GY2YHpm!c2Kq9k)S#$FO(f3g$AS@=gHw$0SSOX-O-X*PO;?#-4Gu z7M*1%3pFWuvYhBEnI-F<1|z^J{rg7$Wkdi0UJ)sN{Yqh*0iC6ypCJHCOtpnQTo?`o zQZ#;cwsd-vwT@o1sn77xV}dDTxu}!4$_dZ-`Fa&8#$IDs_$w!HQpDV{;eOt-5pNI1 zN$r4((7kiH_+ehK*{L)}4tD+!q_}o8W7*!t73eUl%CLh#t2Jzlu$fYVZ$|QJd-%H- zFGHTMZTbj)3zUrY9H~t!zP~P#S^B_~A>c9m39zjYQ?iPLQd7Vwes#vo*Q&2ZJWU^0 zP2{YRAA4C3*!mgTYU$WEXbafKTQCpuBTYFuxqzNJLC<4^Ky{(_?R$?&zfcG~9RiYp zNDF~}uvGxOr2-8$K8RfBzg^4&-P?pCgJ382_un45#W3;FWB>eKr95o%efLO(+9bFx zyrx7H8OtV=$lKFts~8ZOld)rRXDOd?9Vc^ay^no|SyCRX_JVsvdZ4|)pZ>)8d=F zwGj7ur;pv4aozvMm7lCIJ6o2%k0~d7vp7qR>BbLy~9o zGvj@$WeMXn9<1!=WPI*Sc4~6@rD_)gquM*=dA3ynpBbUi*4T47we97@g?`4C4NchJ z`KatqEr46vC7)U{(P`LVw|jxi%2S||NLC{qLnR;zb*9n-)bEaEgawV8tN~>2w9RNQ z%a4GjHNG1hs8eQCus93-t|6--x$B}J-yh?l(l)m%?+yN3?F>`klpiq?;<>QGc3{Lo zR2MTM&Sww~T%)`CzFl+c8Y2Nq$@jvtPSUl=bx~8;Y0GSmo)2ytR0;9x z7oKkm)LSGiCmh-(3ErZbdH=LxA~C+n!c~|sk0fgr-$d|*0?Lt>^ZNA|e9RyqJE_ih zBX*O$g$0A#}|0zYYE#dV}16C>3no<)xrb?~tB8yB7j50ZooeQ>o=VK)%i zZM}w_1n(UDocNDC2>FlPjS848Kc;TXFOuyhb#N{}nm(^1lz*i^eyd&F6vXBS5!;EW zS5~djLQdD_+FmJdAKWujm9_eAwPoqp1F_E@9T9v+A<$y?CG1(XzHbg>MpEdT%&du& zbe_4DFvRG`ux+)jCwa!oAUJY5E7G#Yb8fnEB-x$?pU(5kZl8)hd#G<2=wg0IL&>M8 zp?eGcF}pi*BRU4~p4QIJViXFt+IcG1a=;n(GY?Y1;XJwdLsMkF8U%pm#Q4kXyZfbP zhOuAZ^BD$M?jrkQXb{_YV1Q7VE3B#EkTXYEDHu5|7xvXXvgbD%PhMz)YQw# zYD}jDR3QZIfsc_7h?w-5Vy}%42_msF z+v^L9J#|;;N4hxEm%BHAo5))-ql3a^@QPb)wrTFtYr2@r*a&Gs>I`vl$YW)?1Tf8< z9U@`ta2Ex(?l0g{XKcahx)~Qx6nZD?C^2Lrd-osC$jHt8R#eP99miBW*jWj`I}JM1ZGe zC}UyUoQH8ESD`pstL&)}XtP~w5B9zaSlP9&$Cj}=yWef*ZgBc~XC?gnUy{h({|EOd zil1>zzlpVW|0Dc^@!KUFxE6uCVi@2J0#)JRzFCvL{R|L+&wP{N>P!dFAw|@xdzTOU z)}7_s&wwX=gbD4CytK2u+5sAmUEezkt}@kLOG(x&8nSU4ex++6Kntbz)NYhi*sp~I zdtT`3?w%qfpC0OV|7KO+{cVi4IqeRLgC8g$f82F>1Ho@FmLWaT*Bmb;P9-JMECmqp z5Be>^tC-h8*v)PJ`W`9YHnB`K(22MlY@O_oH)Ua$QUW$c>K({qJD~3!QMAV}ls4x> z&4=%jCSJokZFRHS`AYO<6pg?|8yuFj^InZzU4mc1H z*hix39;C3mkX(uM*^c}s;Mn~^#{1cQ|HOeqXh1L&?Oflp6Hn?Ih6vQG5xEQOk(@LA#ZdD#lcqjR5GWJ!hE^M#senWdXhcKq`%?qEWv?qNwy)4E5CduPR2$IBfWMXG)8)upq(qfE%8*TXZ0zLcyD+-W66QQwE{3X4 zvg3m{M%qGKnheRQO+>UhYlIik(kRZ9=rEPufvuag>)2*B z41?59U5}Je3lFm-$9dCdvvo72teP563RL$I8}S^p9m~k%jrSVfCW+D~f=CV6q0U^` zjq%XWXwqqAJ-aRh%9$K{Ef?Dc9D5{fM7)}}X7fpUTG`H!REN|swRWxaVYP_SM%NV$ zx%cf%=%R$Ecr|*b1<}W?Wv(L(IOhPHxDJ`1AkttP_QkMk<4Bqa%1dW;dt{vc8{B3Y z2cY4kdSU`S_wbU(2&MU}8hv+n?d}*b`g5YFp(Ghjq-YJv#!fx4ZZt{Dz{5ld4V8y% z91}WGaxTu$qyp5nr5@IxhMFE+?MU3&#-1&E*Pik*bX8(|s*GNs;R>yNNP{;pGdV?V z%B)y-eKK(pdovL^q3QNCOAzZ|mJWcwHKmUBxvifMftStOX((G!(WRq2HovV zoON;8to9NZ98O4BJQSbyep9+_I{H(Wwo8_~IeDWqKmX|`+IHK%ot4)-{%+*fD%#p;J4Kh38o`2FoM5@E+hy71;^p?~|A-VhepWogbf0TUqAR<#I zS3Sf%D0cDaS?Y_yPa*qpS3f%wKiZh#8^ejY9_pF;`LEZ`x;hy?;-Y@1Uu7Gw61Vnp zaP!R*M^bqF=kcce7mShV2ROZX>h|L%7AGI%`%Jf<-=0ZkTiNM#AXeF-B+7{ej&J27 z8}}YN8s2W2@Dk0ze$f=`iaB)X(BpxvdCtTpIWA7ovN=3)j8aAUejauAUHwgA>-zV? zrc*i3DOqKxGkWyRk$fTS^z?Mor#oDqs{;hXmGjCq(o}uQnk(0@FY8`X4;ffG%{Oi2 z>}LqBW8n+-iEh|db$3O+tG|iP`XRqZpF5Y=W9(scM$F3cd-NIUBp(0s_F=R~$*#x0 zf}#IjoS*D(jPgAB9<5|?_Wq}o3d8Q@%TDK<3$OEuL7&Mu6#a^Rt!D+o%kzk(Q$L%q zQ}?{^!1@W~?V<;{(eW7y@%xE?STb1>leB+jw`mb`zLZ*B?Pb>c^0UHwx&EU4p4JN=+9Gwi)kDs4alD+99J!#4L$x;}*516hv+_AA zeX^Z^{>U_;)MmdnzV)w_9=R(`-)Uz_?Z;IV8ZAn_A9$j}_zdD!SK-9Vf*L?=73KG54xdy@wNz zMl3U5QaEgyiXNTOV@47g<0Rs`#tfPNkeq%T@Hsv4!3xte#)9!4LRfx6%b|UJeJ2&l zuv%I-{&Cq<{)G^=>{IGb7*>P#e2I(Gt=TXtZf5uV^`<;Z4z_5%I%2T-{+Q#SUr9@j zxS(3pt;?h1SKo&W%EqvGi}ch{1lEk4kUTtPPmz3*c%M`2tw5M694V8@KfAnA;GAGlz7j#X+ZaTb)S^Goi`xD7l(0b<4ap?qUsPBT}J7UD)|)`cqYd55N0`Zu!b#SRN&_ z?8B|uHRQ?o3^Yyni|mJ~7lf^~ zF57n&`&Lt1i$^yD{QXl&H*HjY1Rn*L(mtHPtZO)ebOL#3M4kGm;11^zHp@1&cE0KL zpHugu>pnQ0dz1{5bHeqp!r!A^&WMFL@5M>$EY@z;W=vo2 zp*Lo0Hq+*4sRpANMwFu(MaFUxOOFFmwFfjOw_;o-E2arb(*i@yrmxm57w&?Ne}}kx zVfpVxXCB?2**N%B+x+im`Vk3lROnH%4skwZ=uYy4RdThZNS%{IkNj+)dK%Z1Nj zBCS`w0D2v$XT~Xf z3=(${45djMUluHi!oDw?FNR;_bWH%1~@YNmi@7J??nk4q)ta~$*N@9pr>ORc-`)lTX zUUmm-jw*T>fI;Y2`entcqP7tIqUk86kg5ku=T71>&IjN47XoT$j6ls!;~OgNf?WF94Z?;YHWrZ2rnHXCWQ6+x`OsB2(D64&GU7x_e_K82bl zT;5Fd)7y)-NOVd-IBRKd&rGl#-z(3|U2{Qn_Ub|DQ|cjQ!`|F^1f)EKh<^$Qa1{vc z!^w&~3OPERAc)q=74RY-{aK_X<-GOhi}oocmL29F2bK<2Y!RExAl+H14>fI;MXmHs zm~ChZ_I3}n$}WM)zN4*qb8Y;-$Lo@AAo%w-A;vAACI=mT#C3$AC0U!M|6m~Dw+N$F z@6Leqs!P8<C#t+2b(4g4p)U`ry>eEsA&esIXu_XDI zn~_8jM4!vfGQsEWcztBEO(fQ(B*$P=F7r3lnL1^S9WnHZC1W<$>+=Kf4wpEnNTfs{rFON zPM>jWFHw&vuJ^CEIG*t=nEQip7g*n^((0OHFE1)bDjc!$T3bFWQDXEfb|dRob18B~ z%7yYk<4-V@=t_!ZQ|M#{v)Wb8iGLw1PqMN$u~{T?J=2jjuH2mGhpfWL5b*&a#^z_s5e>`NnH3005 z>&g~synW0OH_duK{D6Bf|3XvEiW`g(dX^P07RoN|(Jb=jV)NXr0o~tiC!r!+^n1lFEAf|&+;R8J&kBnmTDC_Hqu=T$$_~lQ zX&M{vbF-b-W6acLIeGFU^g3#m?=J<7&1Ivi&7Z%EF}3Org??b_E32z!s=vkx)~#ge zqep0aKWAmz*=UQVYBy{Q zQv=f>Ce_4#pGc0H&u?G9hdRUVF$cYWDH7W%KCp9sG?8xj(-YeY)uUNj+LA^OYe37a z37qY{i0$de``#KaP*I6ex~)~C=NWl=i$m}!BymzV_clLaAI`E#$0h|Xh97W@r00L6 zcY7GoEcLQIi#vVgptJQ*sf~I$w|bh2*H8)I7qce}jV&%0-Z@vbgvVEXZYL;wJcs7w zG!*C@+8pKz4$!sAJR0bWs_m^skENT# zwD9kb3Kr#Z)VRkoWpcXOKi?*oN-YMSM&5U6Bu#DLbRVoZ*Ite4s!-O=)Z6}jv~)vC zI%T8|acOh#VI714G1ud6(q>Qe^6Qor>P^q6ZUv&9WVdfzK8x)b+!#Z1RxC#DZ)^sK z*Q+-U-iiOWJ*4#m?s3s*nC1&P!3Qvq>i5HJk}9%^6yJ$L@+&OkIJs;NIYPg<#yPuc z{mY+BJU#`aC$du>AFiC#A$8ZJud)Ap>uK=mbAok}`VSWQgjo&d4&Q);EEGfwk++{w z%VAWOo5@-zF`xcRMW&TFWDJRG^Fsu~A8pIl2?=JXtuOC~24>rv_%^?MKsRr!pTD0A zaBD0zUV2j0>70uz=>Z2qSS!KDJ7;>L)cgH@H_uLHVe2h>(m;OQ;HT|N6;zy*S^e{8 zcN@bpUvj$UJmrkl`#n$2;wq8aNz1wO$VuF}_va$cV=~f2&!u((2NPoHnab9cHAXWT zLj~EDj6tU*uR{Cw^8`ZNu#TSIy?3`OTZGhQR7amPih`Sm5x!47=W~2M{_>C2z~O%( z%-&b{t+_Zj6hJ5L%Z&3eP5FJlb@A0##U_(V`ZeT%Q5DqdrUA3{Qc8 z0xIx^K>6tR6~M&b9ewlYU&(V-Ff6P^?0;GG=Soiqc6_XM0|9S&{gX20k0JU-wl7}1 z_%bok&h`o)pUPSFpJLn92!%kDnJn~u2G)4s_z=6?as|9($vWfb2+ zMkEHKC6`_{PGCrCO+G9_VyRe(sbMV334eydk3HpY00OJ=aHS zs3~xouT4O*?YiHWBh&9K1Idc2spR0 z4$p60oMYhv!zZ@UGt1|HZ7$&NI{I%m|NL8sl&El=Zg4rzhTYyZd$;N7!S4+hR?iRN zG?q4;8~a(;O^2?#F`&)AxE&S4C;kUJ1|Rp2OA=FWx%9x*aOAm#O%zl%)Irra=Tt|i zwT`)r;gYPZY!rty0-dvg4Jp_G&VUNfM3Ck7{F`Ab@BckL=c@Wh0ZAwPV~!>i_jNVujT=C4vx!)7eDb zS;rjLgpO?{|K;YLl=JV$h|3}pM z?VaQ(Tu8OQ(u+ss^-rEWsYG%WG-}TZk>uO8Ls0OA_#B3*_OI{e<3jq!PO*)@_@}*# z{-Y(|5L1`@@Y_$Wi#&lh_8pOK(fp&cefBCOt!W8wF(tz|o)|HhHL&8kEL};rlMZZ) zINy^DS6gYus5E=C^?!Rb;H%GYAulh4l&=)hY(H&|Z0)m@b_uAswo`6qILug-*(!4^+^eY8r?DD;PVd)#+&1>oO1Xmcq#9zBh zAD#08T+u<@K^2aV2?)11?U@lfW#)aanT2k}`jR>H7AQoK%QZ=ogtY3s4SHb*=!>z- zEA(KWEwT^;1nqidRM!d0-WJ78RdD@rn;LOk#g3rGPN%1Myfj~&Dqewtu6cDPi*L6Z zx!gKZr#y6kJ-SC@=A*#i1nQNBLavZOvj0%0V{+IQ(Ws`NLZOq2mp@qMluUAq9J%?y zwI}cHR^tv5&=w)}2}iDo3oD;jC>>C~5i~a1Qnq@`r?0d%FTa}gl}BLkPPoGJXj6?d zTe8nzfV4|~cO&swssCzEKWTJ2xIKtfcVEa>?vRQ5!5v$2CC`}NJmhAodrlmED_u}X z$fukW;{^2pH65ae&zEVVCat6x8ppHAt(Wtz-w#|E#5*PePWa^aXC4cld-4R2Sfbx# zfO(<8Awb8d&JQ1$hw_|?2>$D#U5Ei2=i`OzWE z6DZ!8F~2e6$E9-)y)Umcj2j&7@qRlVeEx#Qo~tA8`(|2AwKSYKdHIZp`XVdo(l00z zbXwtl17#QMJi;9?{)y0u!qPG7xGh>M#i&Is&bd#!`Eh`C|>*Y^o7P&~M zqYkCJux!jNg#$vRmgYX63~%Vq27XRC@JrRuFK^88MexiS6QdIYF9b}UktaW@{9^WC zR=H}I$SEx#&nO4XxHuBg*GgAT2SRtUp@a+X%q(mI{Jg)f^#P>Sxl?^{d6slHgi}Iu zV4tU|ea(|zP1Ja-#*;t#9+7S8V|&hANpd}b{CV?dG)Nz{Lk(A+ifgu1HlK1ciE@-K zjIofeIBM_R1)l*MbLQR@?3yug_`j?!a(YAVS_rL}NXS^`{Ia z?xOQbbrgJKWDyrB=(&fx)bnx4VMrZGzg?bRYp#f3kg-!hd^KH`GmGrJieza zfsIKy0;R#dnz8>??DHEf{jFf~icM_UohW%^><7N_s9!YPohaO0l)LF&lYexvj=^|%p(r|U*{nFnHNRN*ckACee8Vb^|O+qhT5;=;!17ybYPFuDW zj_h*I77rq)TD4|N`PfA1iXRPo*p_y`9Xg{$#h0=;YH3ZFM39vuWl{YWaOjS&Ub-XD z-z(!c)Cz>a+;76Z(Y<+1=wA<;ho)PljCyOZIS6aun9cj9zK7T`t-kc1J%3&voQV`) zT+tJuZ;uuI$Pm}L$emYAJc5e|OVkb9Fw0Cby}mplxNpBQhMDk8d;_3aS6;D(*mF!& z7FkVly$Lcx`1qe9Yuaf@Oq!w^)qL%SbMp5;Opy|WYzI3>TF=RDHHt2~aU9?bjx-!I z+HBnG-2du~|%Eyz*5(QLFutEe;r>m5;2)7gd!=?|EAe6)62 zx=pDslTw}XT=k~^(*-N1t2aA`RR5!TMkenGSlTVS11Kh_Rz1l}9yv*|F7?L~xFVkc z_p{Ckv(dW9F7xe~nlb3xa8w9u8`Ln@@^Rsm`NBPc4F9_aW`?eK;4fq0kluk8KNo+f z)JGgg%SXPmF&+cC>+xDvJfo#<$ZEfjzr>QI@R2c|>dGi<_o3`d-WS7W>+|1Q1dI9+ zfp<7MObczaHbPviHhm!zj2L$w^=@N=t)_F>R;HM$VHx`PmgQvd)dvx~*!G@|+yd%oFT%hDbt|R@ zw6f}b9Y5|@8KrDL^8P_c8tBmWC)7%lx!lD$^!Sq;3I?_b?ud%a=>3sThkt&1Z2tDR z3P<8d=h=ym zc@1<}^T6UK!A?HKK0FU#crf0hoJT=I7TP9veK2R4?T|bFB#7CGRH#_?o-K>}vjY`V zH+M5dNO@w`1OFr}D<@}iYs8QXDP7pPTODeYUM*0IdxE5^&*#d*mK5b}?701~l^5w> z0oCSG>mwW_pwPv6HJRi9H7vP;!Xfqpt(oPh8RD$UVh zbALaMyzm_Sis4GqCCL*kTC}Myg-B;}*YN9RD+)Bq=2kY(DiW?X4(I(Gd!dv1^JAsb z&&ntr@d>ZC@D7ji!HZdefoDW+%^ou2LVo@F)qC}>Uj2&pb_eFc->H8T@ci+7yP;#? zY3Rqllb3g*u-#_Jt|9J-1K4_GNE&CF`6kiMrJEhE-@kwVqsQToh}prp_jSV!Cqt^v zQeA(VL44=kr#sYdc4TLq4rZ}?%lU$x)5We3V|)qu=MLBW+Nu7?w^8z^(`rYnPneIx zr_c8@+ERqB>9V>leK5&uxExxC8XVe$cBwDU?nTy>SC?*xT=9S5)BD;c&5=`o5ipaL znYJ!)<>2akU`}UlxI&5il(6O2Jkp)dCrTz48}lu`+6_S0Ys1SU-#mIpfYQS+K8~-g z%`Eic6(df!ksXVq6ZvkCizYr*FaK<(?)gbmeewy_BXC52E;l(iMmkB%jyX(S65rUE z{%qt@3GfB^N}hr@It>Y~RkWZ#1N@MXv>2$iDdmf=DQu?*E8U82sx)ie$*Fd@)3Z8V zotz#S>Iv`!DUKnW?;S~vwD=yS5F(YCgqOc`6P;sTP#&m*a?c_g8L4_QDnue7IehlC zGAh)yabp5kVuo~AK*ezR2K(f@ku7XJm4P}_=UPiTrNDyX9UtKOo5ra$a}7y5 z|HpLEEOb3Y`tzX2mO$N+L+h)7dyhe+Tl_0o+u@Da`B6)a)z8NEgGo~lK_7O=bji5> z{ANyw6vZ~e- zu~C4aha0X{Up#s|fn$*&&)jF&`BirXJ07$wM*^#etNTM_LOO;f^ z3a`ZU)@#6%{gEWo2hmTz*RaZfdpLXx4Ipmp64AKZLAD9Tb4dq5wD#pzo zt1D+p{}Gq(t&befYnn10jxVikuKxLilNPy;b;0~xFGAq-AbzGUx1t85%(V!M4n$S+ zxVn?N<~(@3N3eB`*E*^%wv%+@gER*tR5Gd-gt*%}&uOAY9nqWH3o)h|OUvN{oyR1Y zcP3}Hh@C-?I&iSaQ&f`~IHAnsfX4cJKl-|7@AfRQDuVHt$+Dh%S&K}GvqMTZiY?Bc z-&fq@h3`x@GmVJcWst_ny3T8QrZ1%wXT(hk{4w~pXnUrs{B_~2wP}v28Ry>Zor*_w zJ)ex*X1xJ*P(j7cbal<$eI6B6#&FwSGBPsHAm=okJ0>Ay?HwC9Gn(np*q`rEx8EbV zDwR7ji2~18rSb~rQNP)5A1VdWIw!M1*2$24puvF=ZTFq+^x31=xr>1sjHt9bjsF>5 zm-ds*zXgPNYpAK=a#Q<7d(c$2OfxM$0|OE? zd5c}>mP(mF-G8X;KsUcEJ3p2SDYHQvTDkS72uney+)iDYf z=7NwB(V$@fYFlIrq4}p{{F04l^L_veEf1>ES92g(eg4B>X>ZA-l=qpT`DZK@|l=Z9?Ek1 zzAkdyyR|2X)ZG@`+i>8upsadE$aBK#K}1!Qbv3$MnYlR4lTl1}VMBmAl zhIN*oUFzE2DKDplK^*aG@CMpoOHP$|G-@sEqQQyabZFYDebbV5mtXg(T4te2ok{Fs zX~|hl#yplqVRNV0iZz* z)K19Ml3i?ppb!F)>Pi;57>Z4XZ3AOF@c<)@7W!X6h8Ycf-Oghbv`ste%18YF9HtF` zhl7bvVK72h4li;Il&-y;niZY_Ag<6q1{rGyI1=?rptA!=_?XUI^9v#iBX$mV_Jai| zr$l!_c<0g!{SGZh(g*O{0W&@|)mb=iqLD^CpLOvZh{~xTi+9e|$IjMHR+?4LKNL9g z#6a}md0scaj}_Ve5+g;|BT_>3B~ zbZW5Q#i+sH45u!5Xsr*iX}MO(^i(4EfXm67pseziFHw7gm^!JCRNcMu7XksrmN|7Vut8=Vqj{_= zUll4qlr2woMs2|xb1l8Ora&=eQ+vzk4~5k=3#v+nm_RBt&)piumd}jYmoR1`&EH)l z_A=uVeZQ|3S@(I0_|~~OhGyum^KjitgE0iMlr)6Sh6~EodzP0{RKz=NVCaD~)H^&$ zd((Ef?);mJS|0<`hTaX7RF-(X$z3TqW}-DKpB32sa7YKaj$V=b0YT){Cq#EX<^-*0 z>6K9wpc?3z*Y;kpvT|*^f%#J@OQl)phRE*+o$ZaC_;#s}>>OMef+vJFMP|h!ebF5i=tX}98#i{q%G88lTsqH*g2ep z_aRlB9SWYikmc8bqhaTdPhxtHmkf2$1|Pl6i4k=toNBC?x(=t2J&df+zLT&G?&kxr zEjswdq`l@yJ7o_r^kKXb3lDfiZ=6X|{yI+Y@an%)urh8_J&7w#)L}2yf!CP}>@MG& zuv$hhv2foaRtDI^OkwoaeP0^2Dv%8|<@wNF`a$)9nCko1)!ikwhIE~v!8Wox{)#)X zp{c(sxbm3D;VDOW2=9iV970TN9{nWtf?Ug@@2k{p{0RA=MP~HDobi2$WxmhJYpcfx zEo|mKy!8=m_?K1PhF({!ygI9T!W*v>;X@^$!N z>zn?mQ=sd6Tzt1hTc`X`5N3O$>B!1xd2q!L<`TjFq2PF|nT*KseEoL^eGzCgPUMVf zxKgA)-Y~JEnf48){|rFmZN%RaLDVYH;NBq6UDWg+7v|kn=+R?iqB**+6MWX)o&u67 z;a=z;cN}O4GkFvrvwY&mj=B=hE7Zp9t>P_tLPyEDe!H6^ZPWcmSdxb%N~wmlB3mTc zwwRn4sh9r#{QSy@oq^&=n@E^HI=yZzi1CYVSWVrW8HDq#ConyjQm&UWeNO%6NXjdt z*oiA}Vk`cv88oUJgq9~!x#f#SFhmv+t9s+9??l(nZvUPjR_tF<_@@6$HGWDdyWhLgQ~@rm+bV&^E)+-I)DEfH2fZIJC`&1hBtX}T^lfeTqmShzB? zFg!D?yV|+zb_)ho-|Z)xIaEA&_8PKIxd8if;IwbweWC9VSMQKcF_=&-F*^r+M(dSfKADv#(WGzh zA5x3|PbZhFf~1L{=K}pa0BGF&AtNgzbGWCVZ%Bx;TIg6=?^a5OijF8~=~Pq2uj?A4F`VFI5A$$_u)7XC+PUYC z?&u#~TUiEB^pX5GFfwV3X1GOXzPEQ6>P@3T8;m^`b@A4de;1!;E2AdwQZk`~Ykcon zh_?eT>>`ztCHyXTO((pguEbAb2x5l9S} zb^N!d@ZFBjbcVWOIcqx9oj9#7($48#WHaOFXtjPJGB?RBze-$&8 zxFf&&gi49dRu4&F(u>syn_Yb<( zU|_FI+sh~1k>!hHZ@o8*?Z1wh&Q8-VHl;I6=p{cf1TyIdRUj`2tyocM_Thabwg)ma z*Iwt&dU<=uDOs zji|k%)~ZEp#SG0Qy(KAnm!Og?THxQDZv{kV5XMFO4U^7s{>k~Ub4Cwq9QpfeqDtsc z!2aRH*(5}T%kf^Bw`-BLgr-Z1I^iFj{;35J^oQBwgsSzi!tRFWKkl)(yBn^bVe%xL z%I;BwG|TkW9HthXFLB6Iu=GrEW%J%JvFkd}-N@VkCw>UJBs@1s)Lid`cq9np9BxFr z^o|TC2mAAiDb=L0ti}-nI0WZXGEM0=?U$&ldq_WKlX8gCO9bgL3CEJff+&iBXN2i{0>qUEXnT=uYH`IIfLoTOAf~qxg)#V zMo~+7^&L-Y!V8I1>1Z^o^<2?h(#XWHH@)^&cQXdf5U0~yELR)TUFdn@@%GC(XU@}^ zMbJ<}qtR1vQNR08ME?iZ^hz@*@*1vehI4>WmY0{hundz4`IspovZ9*#A^R#+#BP8; zIjHUV>@_u-jmt2s?jmWj88MrQW$zT96PFgM_rVj9#71tPRuQVxK?n8K((_oe*Pr&d zNqaBFwqa7m<2${nJ}E5yireuwh|)9k(`t;#4-u_1CEb|RRdMMh%6^p;*7- zqJK10H`G8HS>L{RY@@AC%)}Cpr@HmVqQYaJd4`c= zaG~$+%Ze4t`IlbZ{P_k5f~KIBl_Rc#;_93DZ1%rlSyF#0w0*+0er;+l>idv8*o?{M zYPA}n&~JC%*pnDZte*Xd%Rm+Pn~4*v^Vz4z5oS0ov?-k$Jy#Y(?bo`FsPSC_-oWxGi zm-%%T`X0M3N(AQoHbr=GD<*%1zHu#5L#ZmNU4*|oW@)b7)&VAo7~lds84ph}Dav)YGGgZMZP~-7H@%I@Y$g{?#lzz&g%uTidC>d@hPQinxL{ z{n3Mi7m8F%o=dXb*RLT+%2LTiHU9P=!!yFA+8EN8;foK^td-Nd^r)Dgm{B@_BBi^uX?BqvihxKH6dHLA}!|fuf4>MT$>oZ3X zN71H(c~fO>Xj3VVso};jc*|KlW?4pDj0}nIrgu$ zND(3RM@31_LdVD?Hur<=+j<`iW8Ie$Yp96|hZRZGb6(WwW{kE_EgnV3ilZ1was5xh z6CA9^Zy_#O2Ny9Xot6fFVn%q8eXt0v&kZn7Ck2%^SsyU(mFeO|svS;0E-5N$E}h_c zA}jajwXoe6hcL~K>YMn*PR8}gG!wP>+W7FGJqjY+C0?Rgi&@k9;UKgb*(X`U+JP~QUkR=2>;D>1XiM-2W?83 zHEPs?FF4P?_^?cakiMbc^o&E%(mRrxEtGz`Nmg4-3qR|!QFabli=0XDWv_9-9xU!Ez3Qvr(S&rE637xsOFVxwtNG^>R;Ksp) zecBp@eJ26J-L(@201uwZU3gPtgPFK+0M7BaQy{ayKZ$og8W<`?sj##i@f$;2qQBT) zK6$GsF?%*?gqlXIOyVPAg78LYR$E!STMNc@^@ou?9cwdRTSTp!C1ft5+%ywer{}fw z=+oB61nay~S&YfV;pY#7ol1Pi6K6S5jJAREUqlMWtLSA`Sv~^WaB3CS#8*c#8Z94j zHmisp7?S)+I_E|0{M(h>*M+2|Z@BkVt*SY<;+tyS@$~R$DQmF_!JVF$i`(~@e+W;b zwDlJ%qDcwV#LsIOlPlm*R5DSS6Vl#lqPcrtSYlY3G=8F!2|Xo5su(3PJ>$OCr0(Vm z8CB2omY}A3et;7IC4Hfyr?0+->CPE<^wq3Bz;Rcj`?7!gaHp9#&}~>H1T^aMw|A&ww*zlbxj2d6R&|DMQ)EgXZ7w4# zkNzmbSLDSul_d`gF(%n17YIbmg0%^~mQJDjj9K;^ilnNPvFveDb`uEE>ZZ{yOerp< ze-v?{pT|b`B?`^)yq!Im)at0tHjLyq^O9~3mLdI6Qq4`AT3nMWF{OKpqghjBLB&aM&M)A#Ol(Nn&mOFY!jC!3@qDtYRI@@TQzb2`b$mP45&dxR zvuh^X_X_eAV-n?vv!YXso%h91?f4X$Fu%|!jB9IIamBgX5eh9EqKFO=8P``s7YN7< zIq{(~iuVIe-nbI007W-gxjxe%C%T5-aR&alfKC@9+lUUu9HO}#8nQvHD zB6?2O>7%NV4DV(0pJmBPwnap2rF#Cdt!~Yr(uZWK%3SaiR^fmX*efGR+|GGr1Btfb z;VpLomj^X2hT)v5Sla^cFWU|C>xRi9k)PKhZMgKC9n~KAe=5LlPDz}keVA}o>6q6r zX?EOfYjVt5lubEEcf4BOu6vDN+WZ}{QK#JZb1%)h266V7MlI*V> z+|$HyUf#j&02;N6Q&Hc{IcTDlt(07^dbb8Siag>EX{Cc#&_VNS*`x{~8?r?5tbBt1;LAA-nZtAJ)lb3e%@fY%L>W?ezdy zRHm{z(5R(EWBTNp7c=7&`e{$Flv}-q6~lu(#jVfi?jce^;%RG=r$i!8IGt!&1krsE zI>1Fi^7$QFz}c$CRvFcBZvS}mQM+1QBCNv&02scOuwirn-MeXgu6%qq*v|y$~Sl? zGFh&o9M}Ocw}9m;jGW=bSt~62cAg-l0jqa8v1XY+Y4wxzP3=#sz0vI$Nyj5IDhH$5 zgU(z$45gU6?nMx|9}}!zi@7a0he%bB{o|s(9(v#B7nCPY^N&yV`=%^ z!9!|6Cd;`%D`1r3b|6llH90)oObUYGBeyuwnLv$RKz%c&FWM`#@FHzqS@`YxOPw#! zw8t*2oqs%5gF=%y`xI9D4Eq6TQSti_dC|v-1ys(u; zYBesNYNh8yE$em3r-)^9Lxap)7?c*4k(2}hbGZdG20O`W1mS}z z4&n6ad%j6*A=apE^yJJG*^eFNsNTYUs9NyE+D6Z4QK@Cv-8PIKPKlhjz=iz&8@xT7 zq)a64O?Z_zx7fJDoi6js?@hPs2%^o+{tDfF7!i*l4bCP-P$}j9T59lGdO4b+m^*y1 z9fOL0r45(GCJdR=Z6tDSmZrPP@DP5Ix+_#ly32Yh^1#$;#}T`eEncOvTrnmS!*q-q zE2*yE@LLljX$uF^@mh0sHH}F4W=_CX3deEG4C9VVQ7Htt1BT{hzsSj-hM$)NQlpPV z4g^mz8drN;Gg+2SGBdkA;~Wf%TzQ2dAys;NHalg1<{%US+G6KE*xHm23*PgNf48^m ze!Gp6+XFR~yRGYg*f>3%Qa*P|lQHSw|JHYmGJ0ppJ_)7Rmkr^luKwmI+(6UGfVrgc z9>3VGGP0}nAf>Fmrx1%lEe1tDB$9EkoZjd)4HJo6sl@1G5Sdzgv~q(t5aoqN)!ERc z7K%LF&sBeG!D!!g&TVZ$i=@xTisPx2o-e!wa#NbtaVu3DW&ek<{p(GJJB$3|y?WTR|ZzbND&7r&*KS zh>@!nNMJEWIl;t<2Kl83vIWDOOWJHV;tLy=;{k7q#nus>#sCsYjUp@{sxNo(Kv)Os zT0YqutLWiPtC~IQBFm2oeVJS-FLpzg z&8xMEBoNkVr@Hmf)Dqih{(JL!sKj*}yJH|H%oWNbUy@-gF~byF{7-lopQJ|>i5S!U z5zR5Aj9GCP@SFNxmpyVAlSfpeBdF|p>~Jy55<5g#$vNXlcUM^tz(yCh;5$Mgk&IH& z#UWCPboE}f!Pl(Fst4yWXuYkQ zI=YuSp~dI;3qi-??zWRh5C{xwW>U;1b1!qzC%Tb{K@@j(XbmG!q_TFQ@dWCq;yDtA zz~K+>wNoi8F<;=ZZ~hLh03RN#4Evw1Za`l_vldBdaAkDnISNFx$7ia+7(*MGUiOk` z(x8{U8Ug#{;PBs639-6M3!)N)D{!3X-_OD69DP(LG;&bmh~3Ee2%`O46j-*BujJ|m z(&}}1YFe-pOu!yQ^p+^qx00_ch}7*HQ<9+Tyz+T1+s}o>>K(UnOrJ*U=Bnu5Q(!@f z3wB!UiFfI|cGuVMCPLDflFA-8{|&i&L!5;!EX78BHibw(pw(a%RRfP_4`JQj<#i(0 zM=2!N-6lku;&w%qg8A6a8dFc%M3NTK2}{QUjHDLP)wnFwl34aEcYyrZCATb*$Ylc8 zV{+x^(8rMW!Z`rPm^3(gQ%fQpl^`}bRN;$hUZtk#bQ;@rL{`=FSq+t<(BZ#W{kR>K z|Ip9KKlfZy43chBOh=(_R?rJV!Y$4xk>*>Hgj@Nq6I)K zS(OG(13RflhXsV#bn6b)2LjP9uT8^OuhKE?*G)Wntk#ARo)Mr%m!pZj=Z5+2u@6vvR3f-=oBBV=Zsw_QA-V=k5 zMO`07s6I;*GV*|idus08&I(W3Bz(1IK{WnAgmaltU0nd3=BQSOch?25TFPSFZCPJi zL?j8^QlIgy;RL^miT*C;vqJzX(fM;qru0eT65s8i>(EL)5*)A1N+_C*>G#)`S|3Ru zM4QN)PR-GyxiP*5d)efSU|m&4T58aw@0EG)X7_enR16zV2Bv@fXbGpN1|S<#SsGUJ zjDStabwH86u5ZvsfJHzd~Czp`9*0Mrg&%U|743 z0nRqHr!!I0pc~xbI5q#?xkNEZ$}4iyRUCKVxIA*#VjhEjY^1%p+^IqWdj-C5Zxwkw zj$~u9mQsS-eJc`9SmZIp>%wiZuT)qwcVgS!q!;1WFOyAu0cX$G-Gg3bWsyC=0^W+@ z0g7(zR$}I`9hK%)TS)C^Z^&%5(`oBDI@(>d6hqN~1_!`vLn=e~ko$yyeIZu5tt&wRa zqsr#V#`MeRHH^$;tXzLJ4v)7%pIVv51+OFF5Z?}3Xkwa$ zYy-^>cru`HqlT7S;qH96w$bxhHg#_8L~XQm{KRNAoNmW#dUoIA9BXM|IQb~8>E#bl zX@$HH?m}(IIK&Mov7Py8VH?|-0{#2O)tAhH7cwMAqC`iRU!X3Vma>u_OjLJaBqKRt z9>Y)n31mQs=n~t~Hr-QHY~TxVou->>zf zmfTi(gqMA&rKmgSQ^>eEiq$g27E(4ME}nmziLk5I8O736bZYyemmUtNj$C`G@UoNo z;?G%T>V15|XXIOpT(yFR({ZY`>#P|ktTJDk-Rh_cy~_CPKwQ1YBTz4|jy~@5;p~}I zM{ORWrB+~3;?(APcBmb}m7G;S1BK$VvNcdJNi)&OPx@Hk!hv_Q1bt5I0@5OpI~1>FY{+=A;Qq6_N?xyTwdv?)+t*rDwd%!i5q8w`JPUya+IotiH;vb(pa zo}klIr_TDnN;71599I|Jw5bBYpgF^1ITWbne^IKK&9__rE4i#E_J$~?SeB)Z0EhZGyT*XJsEY1W=^V)b}D;s!rcr0vDp)vgW7#Q#L;*HR~Cjp z?Rn68vMU8zM<}i(CWprO?i3?F>2%wPp9+7SxK=w#e5#L#TwD|Wp`ppvbup-nKt4It zM=QLBH`ML6xoZ@KeW}NV+|>;VV@vEE)nx&3b#EgFgLt#58*O};gd?8mIDS5j8&if( zRr%zC>L%k=W?m{=Ki+GfFF6X!gS?#F282h04wT4We7COV;Sm58H5QM%ldHDDj&7*AsR(R(IKj-3>q2$XvQdRos$TYkKDTC3(fC`^z}|Z9dCVA^%h+mCZ;VC5Bbh9D~gq4k`6C8fNm7wjoyMAEN+2P=Jzueh8Gz>O;Q$(uO-HcKPL! z{WtTRywWBsBS>RGZ_ALBlv|GaGb(oG>7=xVNWnw<>=idsqipxQy%;22p$`gBXHt8o z3y{rS4kwViJe#3so{j4KFW`>@tqL4(!%%r33;dtA3Z>@CShk&D@u*J7GGgHGU~Yu^gZt+XBr zsg2eby29W@(;1M=YoV3WK)7Q0;ZI2mv%~Rz#;O0Up2vn z>4f@SqVQuTyB&MI<4+xonVobxAf#Jd)*vHb>VZY$7+T*aP#5ITs4Kd}%p&mRLy*;j z1hXb28D%oFXnIhQ?%jJdP>@z6b*6bNv)Iy|wTR2MXNl8_1Olj|#AToZ+(S8*loS9d z>?}Vq6GhLXn$Ran2Mh90xxlj$S*L?-L|o{dQwedL1ol6Kj7m)E-%`uph>a81IltyN z642xP5Ml&SW)p}m7`_}I00I4AbBa9VM47jvT7#tK#Ef6o)J>QyO?)_g5%99$;Tg@@ zsB?xjZjO5^3U&wZ9P7+2rq>(=2LFBU=?{NF<;#O}H8O$ZU6u0?Y450fy+`sDlyzl5 zE4uYn21SgJVgGG~d_m5hcAxL@9H>`z>rxeHc`sx@QxtdygA3y(2rG-U0}t%f z!X2gF(t{yb*V{5uww4{@yW=b~1OsaEM|%QW$8-5Ejv|R9msS>LkKY-V;w zN<2Wf_H5Jkw3LHHHjHN3)c9TiEp`I81Qk;QclmV=FWtZuPTtKiu2NKStUpXU-@F#t zb(f07H&5KSZA8xv5iK*ZUJaAL3Xj*3vUjJbcyw@%v64dg1_>|^_?0c?6*Jd29_@NU z;wV$>9c0{MND3=pdJeNIzlG!% z%$Rq6A+xOH$T}4q`oWeW5tKa^l(s(kI?ug%4LY`JYo2>gUC=0E3BwY7Gg~m>tHlH) zc=_d};8j3J%Uk{;luO+w_twz%9O!bJE|=*YUf%@_7E2-oo_}lyMtgO6OND6;=$(^@ zW}pa21_$N<)dDrjoSZ8Gp$dvYidnt^*eGDHV}Ghy{Dbs4ybY$mX>t%Npv~#GIos6Hb1;HKs3X4(%K8 zi;+@1o@QCTg+F7wBT5ktger0}g2_ywsI|x_M~^&}-JQqobGy3U`ob;62)p3q)qLCF zR!7a~^|7YhiBi>jsUh2rV3+56$Cdp9tW8+eq3R5~^uHc9%BLOYOB^)7w{Omx@Sk&G zmrv$WZJpM)Tn%$4H9ks9*~vHDVWX$m*cQ~XH0ljetbawl%H#1itmK+5BlW ze_g+wQr5UcB}3gYP2Xggnc(DJux1AZ+|MvK)U^k@LgC-fi7}ng{L(EBtP38kbeG!) zd}ETErYzHODNAl}kT1dY3G%K%=^08Sf}80t+SRw}x+A~ru3vM0Mck8+D)ypT{G()uQvcb33DYwPPde!_Tu+O--}Wq8 z6;~`bI9-nl5*??`WOTf9BVJ@`8hmkQhc#ZqlhmC)P4Ev((O|`${bzZ ze9hNH4$dZYc;|^BM1CbeeX?in_{BgRqE#(FI>>F@O!@(8(&bvQ*tp5pq@eFMg1VXt z*;~x7<<>o6&yUCW{L<)5i$B}!G5TyczPYn-Uh4go>38u-`!PdtapU$9viMzzi5jXt z&O=f5PFoNrTN+llhMMZ7Y8%D6oajop`f=1yQ~!HRU96r;*a592{ujUh@8e9pYl{@G zUEvB2c8qa3AzsHLOS?S4^FaG(hK!@5^%>r0ei}E~?uC?_buDYtuefaA>99NO@;#%j z(pdE*yC=iS0}n&7KI)|vn3;hrO_|b@mg&!<){rg-Y}@^C^HzPNDc0}uo5aU^hMIm| zy8F__LTIB;?{==f4eKN%w?r{NC2n+A+|EA{a+oVeIfKT3Phu

    YbPRLuumbe;prR z)3@o1rsr~q^8k!givCOgi^eUX4Z9lpHlEn?z~QXW_s;=5YbDhuZm)6E6K%ejg)|{Z zo#+_R#lbYkneYMzY&UmAGGJwECuM4QAnoxhZ{xk75I0`w{bJ!; zC$Ex&povCw_!D)tPir=kC1$@;c$OrpH!wL3W3>3Pt4lwJPRNtR-EME*{(Tu-=Q-lm z=!edmB06+4)D4+w4<_EPkUc68g+gw(%LJHQ*YQ@bH&37CcFXx{X<=c}>a~UY*}Jcn zRGnG&Iat8V;|-|l$p#%t=jFN^MfkhG5m#;#NYP0+&(aWxZ{e} z>!KQWiuM5vdG2dmXWD%A^`6gj>pN>rnA08n8p*5#q;eS3XC$dZW$A?mqoRsXfMn`< z%Q7VnhlxJ~lj{sPbH+bE`fuZFvL8m=qTYp1V27N~P2`Uoav$2=+;&3^8~(n==fL+I zm*=d6d)j?b()+$QvR>Ie2*0>r{rg9u!bcA;-!l+79#vGnQ%Y**(>0$gqI*Y@#yrn! z?fc;+#o{*IxNk1T1i97X*s--K>(^;!bkumOJ>n+gH@sIbGZfosmCoL}`S|N;U>(m0 z=;%G0ouBSUsB}GY*y?g@e$5)w359iLJogU%+9jp}^lbCF$Z_6n{$x;5*;OM`XNEjAylzZkzT3X|IQ{z0!{ z*Yc=hUggC{HSf1r`R}PbUiRK?Pt!yBmhS?O4yb?D>ayK(QD_>svD!0`04MZYi%#xt zaev^c`f1zPr_I#@ZR>^fwOS}?d!~2u(DizF&_^e*FG?wbS%Rr zW`THem`kw{*d>!g!~!J2o@E_WLv4%{TH)_I{lMoYTkzh?*A!0vmq9pfIm0Yze?G2zRs0lF$!BDzxXrm} zw%GU~NqM)xvf)s{hl!G5soCnaK7uo1QTZqK84C~agyw(V@7DSBMEV7FvE7p5H@8V& znsYuOS$CnvM#5rrZDjyPMmgy60Fbo5P$5<&@K$84;0`LrYAGoZG~1 z$>>IyhKQJ)RgQDYDLJ2yF&jCb8HJ22{NB_3{XBmEIXn#4`*l4}uh*_ikLM$s?&odB zRY;!NoUD!P?0Q@Qx*br^6;6D@c)_Dk6!>#1%bMQz6B$wBs@UrWV4rEn8zlKAF|JoS zsmwFhQtFMqT<*HofBJwNP;PnrsnftYL|G3pZ()xAPUoRj_3S^t@L;-?MTA*_HKX{e(|a#0Dov!>M~CxZt`0 zad4F+#~_bWLIVW6r0Qw#Qnc^3?4N1NyNe@Svi8uY_tTtP#pyMa!X0~OS4ggrOOoF? zW_U%jrRgj;Y$dl9s~NaumP?@WoKxr~H4zG3J*a-3NW<|$R$347k*@`sSk*!p*EOVl zDXTi{e}q4!!_MsCb*Q>M#Lw;zA6&6nc^2t@ex=M;@wL}whdWMREl$H3 zgbDW<$;yeP=!dCJeO@tg_*ZED8RligZ7pCVbpi-j{nLT-88&8Tow%8GkxZ9g6vVnf zK_U6@)bLew$J%O=$Em>qpf{=&32+2`>e~c%0QB)|g2xeH_tYYFud{Rur%JU#Xa8jO z;aufx=Vwbtsxez}-q7V`nqhZxA75ROr}OuaL5a;=l}t}A-?P{G<3`O)1erw-%$Ci` zGtPp$%FRkYqbF@fKHS2w*Fm3fRSrq9d&obx;=W;^EXl-WUM=VzF7n9KM>TGNY~<~N zm1K1pKvMmXufm#{=d!19_!`5E&Qy!KU*0M^D61lMuV#V4kHe7a0eDiQ# z`i9!7^Cw zq1Cu;uD*A^ghv5xzTNQ+M94p~8&_m4gw^LePa(AhFv#7B8pvqZh)c{aEfM@gThSLo_&B_@Uz5_c+4@ zi|e+!lsM!R_CT8!QY|6Qr%A2MyXiv{MzcB{p_L+i_FnZB{s!1TSfFQoRIz((bvY%5 z%y7xjooSHeDcr!>j~8^6u=_%)+%w#Wso8G463nf10l zg#$pqr~6%%_F`5&XX)E@;4ZHx=e$9iyiXP&JY7dXV8zXb(NA$j*m z{sm_YS^I&EX1ztRB0j3p5W*#k)X?iJDBUMvDvi1;E5(1-`Ht)NgjR-@p;e{tz^1{X z8&~uB^aMQ|wZ@{>ar-ShC!s3Wyu$*4;{~JmD&gyrSakFGlkF^(A#8mH5jS7FY5Cbj zpBQ#xHccIXUd=ypteg6Fc{W{9h0@=RNH^_xT1S2@0fMU*SA6Wc{jHGJ{5p2~VlK(R zGm`uyiHy{i#V}|i{Y!O)3#}T0k+qN+m&UkIyDXRY0pQrIG%sx71eh}(gloDOsJ;sDYWK$69Q@RHp) zIPu8wlX2`Z^gOM>e78N3AjTbKa5Zroe=`@XctK4jpa@P_&NF15XOC9!G)at*;y!2C zcl8Y~xYtK7Wez-U!CfQv2`^^wdcdVq#mrGIybBU_UMo0-rr9gL7?h$=G_ z5)tQSonkab0LKeFCBxsZt2?ELhUM%rt4OxlJr!&NO3v03>%pTsC` zCRd4_L+_m8gASV|L+nA^rE2=@E1b=4Khe)S68qlCgP=JW~z<~kieyNq$R03xokF@Q2?~9bo zfWr8ts|h|DTdP}Px}>@SV<{PjF0;Y?DiTz*D`DhojMA;_0zdpM*Tu4>e~|2va`MG! zKL2mmGGE!5S)GIxFcmxGLWD?-o>CKCG(Wv3J32ft5n}e? znAN!i)#GKmB_t^|&eq5^b|S=N#amPY;&TGqjQ>d0@eT_hfHf(|$#ZtR2n$Gi9^ z544%#`AjHVtcbT`6FUDdxD{V&FelIK1+$~Qc5ND+dwc`9b(YSKgFk7nJaKnrnW>Sh zGKTGu7BFUB|1v*15+$7S?l1U|CmT-4tw|*xf}6iBX{b$JB43F(CCo)?R0$d9V%G81 zgn|roZ{yZkF5jF`sq04bG;-n6h^J>wsT^8+jxS_3yPR(_!>1)7xasKhDQJMGTy_7Gye!8#+U+mw3ss-~wM z$~!5P*ffimL1t+AqzlPGQPsT1{uihw2Bc&y4N$co^yiR)!Wye7qK2deh*AKX41D`o zvw(Vmdduq(p#%@?!uI?Y25grk`bRp(N6)&-S`mNnSWWWZN*Q#xm@`olriHMTL5rTX z(~Ewr(1@3~%&HiuZ7Jx2j) zk}N*c-pqUwTt5=lLk3vuIK@I%fc`_us$A8XFn!KK$3Vc`?qah7=}z_;r^o(Q@!=f; zzt{8uGh57_{Oev1FgLLOr}kv2yyw&>*k$Y)Y~D~L8&;$wOQ77889rGWh7xXSqH$M= zT#C8`2mXKpz44M>A;g!b7*6qfK|bg8Fco|_VX7;|lbj^R+G0-bchtE9@L2cSt?D$P zz^2l$IX*hq_%hJtPF^(Hkv2vtmKs#*TGhTxwUfYCZX{Cb_bzBt z=&n_>u2naSsXJ?x~u?1CP0BOl*@VA zq~E&-Z`u7G5CMh~?CJy89D!@#G|42VqZ7av8TraFRb{}Pz3Ockwr9|H+3HytQf?v5 z>!R?@`GoAqm(a$zg)n4foYG~yingzjbnDfz&~@dxXVvB=gUxFqmj_d~Tv4h-hj8MQ zVrcl%>1Z6wcO!tvyx)O;*`L(9Dm=6>GO}@a

    t3gv1W~bwjO~fUHi~A-zkoohno> zB)-1%mhvXNf5HFbn&TfzEl^Hu0-$GWtK?;%B0cS{A6Gm46FxXQ2_$xlddN9vohBub zQIH6c?n^VZZQ`ZZ&w9w8f1ZaZ@`xhB#A?l&E}}9#Jp@f)C^e5M13Gbjw6d%!tt;r3 z=Tcxby5KDgc%RF}sh$O`&r81JnFTfyd)<2|wR`P#3W?tn>BO%^`FWbTs|U8qBIYn1`kArAdxr^Z<%T?hT^roP6&_S~IByx1G)Ay9rq`i*TxlqEUMz z60-~BTxF|PQ~B-ZHWQ0q-FegT2|<(8qO>FXTO=@Yj1OX0$1l+10DE#&ZCeLejac;m zWfK5<#HRrofgvjr_2F1vnqbM%Ga$X71nV%&Dqvy&8c6|mey8VBOhvbq4KHtGY(b3( zK5n7hIw}A<`ab_3sr*3)ypr%sakdvN`Yspyg{9`MA|pdI!?SXn(_3ejUpM7+R0yp- z8d!bCTP;p4Ve6mwv+2iV$>ZkQ9zWJG3#K7b+js9*fwr6q-NEcxmGj)H$m3=07!bMb9 zD>P@iF`3v3PxrjLDwU|_&oc4VE5D5Qrs8C%f=jnT4hAn_u-_ld$Q(any!?ZLJfs)xo<`?vU8jc_N;mhs9X#v4 zSS_Jv5%5Tiagmwm+ zrg`bJ9*QM$V?Xo@C|QnT&=BK0RTV0@u;;n87xm%Gq( zAq@(=O9Xx&NIB5Q(JSMhWvElypx*<)>>Q*9h|)PK+xsxzf!h0@#i5xAR1cp!Grv2uzNHxiaa)v3!Qoaaci@r_uO*L=;G{fd{ocN`o zuG>u_7S{8T2^Tjlr4)v`I<^8Q6&nY$OuQfa`E7sL9z1%UDPFi?HE*3%vHD{%DZVY^ zRphY8POKg&5@h{hW9zMafF`|!*w@1$%2Vr7?P`+yYP-@h5!P+pFF{u!m#vqb1D>@2 zrD7|q1dyK~I48JqQ!?;<=qvCrLIHSbR zaGq`BYAc~1Z>E)O}cRK|RK0U7&J3LD%bjd1lUwTA5qKmdWNU5d(^YMv!G6tP+no$RE*PRJJ zL-1~w^?SI=Q~e0WWkxjTHVOJiDqieY1a;x(ztn@n_R2}-)tEcAA$l~QA{n5rv~(Hp zL_PLBMENO(_ytglk>W3+#DLdB6SoC40XbuePow@M7nUll+5edlonJCCBY8uPThyif zwH&Od;%4nVHQ@?P>myE|At*MhL#@M(1ch#aW1v2k^XKLZd8(N@>+Ei4M}{4i@qzEG zZ#)h1^1q`B9$9;GRk{DKMvY}6t1J2v3f5q=YxrJA2A3;Yo`Bs6L_m^xu>MV@1p`F% zBhxl<`aaMO8I+-O12*7qTFzYjfokdI;dEnYeKUR+$!cO%AbJP$r-6spMo@oCWn7MP z%jAXH9b|(n(S9waZKE}IF{TND&W!$C1Y7?XQ2lf=)e?f7(yHaVTZ>@Sk>7Tta+#n8 zAp02-GdVf4=ZC+i=7fG8#gy4JP1_m566Rt9j384OVW58DAx8NfmKT}!b3 zBKG7=vamYG)*7Cn*Gr>|kNy5arI@X^tJc3e4XwdUf8Z9SsR zr?u>mnDBT|H~r=}msWU6<^WOB{mey|M7Tt(nnbFahXyM>V^guaQ%fAPc;zy8Z4N;k zNLmm5zaPnCvm`7CVQFqVOKNGC><(np_)Xg@$)qlXSGa6h)vyUoi(>zj$4~uprddT{H9*{=~_$%fxLvcg6ksuy!JBaG)=(w9Uds1c!t1TTC@T@HN8Wv zT=``JqoNqt#GuFjCHa8!`p<|r?7`L?hNYohEp9$fhi2We(C!(3933BBRnU4hJX$90 zLwW?WVG2#e8Tn<_HzBQSt11zQ=2Ii=?!=U%U=m`zX~sM5gc0X2t|z_k7CDW`GhryQQ85U@oEOInF8=gu&L zz6s!-h%4>y6%gAJDRq4U6o$v@^9lQ@aS-q`^3>!lkrMc(J&XBHkH7>k_<8-pu0ERR zg|AnReN^4{bxV1t16UVE9jQGwm1|`vK%^@}=>6ATJI*SA9?5-x0e%}&zvc9ls%veQ zowbKu>Kd;y-_|(XCx)5HI%Rh(NNRw_F5I$HKmTy4d-)V4S76nRnE~H0y*!EV7HtwP{A3)txc=Q{ z!9mpMMT_3xwAUJ@ty(xsv!`cww*OslU*r<+F-U0xAVuJEcv&Beme-=h^6>L@=P*Fs z(ux~Jc1^@Ust{PeE(SMV05pRD$`dFBV5vq|Ue)j9erKa1T%Yt)2V>Xd|HBCWzBRUIHmQ{=nqRlHz2!W=^*+l=cZN(VQSOY1ej zI;~tv;F1iOg?6vxqg89_{wb%g&aMGbP5er2K6y0|dJAfSzaJU7>N^;l?cyC|G?N6; zER?O{jmT-Wtu(iVYlPh1F_0)qB#)`f0tJCcj3E~`Zf2E?t_xurs0fWfN3R$Q^pUEa zI$BCKE-FA3;t~%^knK+5qoH8Qc{;auAd2YfJ~eXTK9E8`@(={p33-ML1gS`9iM2)#{t#*W4H`T?Kxa5fja!`BM0kz)d+oe=P(P8l)6 zea=Nm0DoWfy(y1$z-(^x=X@Ndq!}-vzU2^EzCc*JZd8C&<9pbY@#y@um%M7+wzI-@ z0tm$Vr<%Rq8RPNG`0E|gOOR{=wg|lQV=NQ zo8Y{d_vq&%>c#F-a@NnRGRA`A?zn4HMth4|B``e*|F^bl9+9wZeqjs&=Eetn9nmg9 zSD+3a{Y4q=)Vh)crN5@amX2dUhIcPJa-@I**BF2WkDCyX>;#bNcXqsp|GilQL>awp zM)T>c&r_qNrGf-8WWg@8#XEu_-m528X+8IL zr;^lV>1ouUlByO7`i+AN&M+ zy!fxk)JVIpFEshG9#z|i*YMfuiCp^_kJ2D6+HSxRY8J+#ApMFpymCJe|Hv%Vqn!kT z-)=8!88R@m6-%|5d~Zs(G+>%{PfG96p%;7!6I<;(E!v-IK-^S3*_jys`ev{xP+S<` zL2K7yPcVvWuEc&^n0S!dyrBlRg>a6=2ivL3XSQ^%e*clH?_~eal2feHPRmAXzW|d+ z36MORqrY7to}ne~QH~<;l!~Kd1V!MX2gk~CNdinqVP-B)-NKIb~C9MS(U9YeFZ>Egx@-*MuEF`Ev8_v>>j^0 z!0F(-2QuUg@P4Lkxjp=;8-Lp+#Xon+NK}gis{^MHx5t`0g|{G>YFm!Gb1E3|!wl+9 z>-;-%w?_jS{m@lbwM*2_25U*OR2Q!RJxk%QJ^UKz_Y)XK8B&E}?!UP`9s+p1(M^og zuDTJh0F#up0Hk2lomqQo@(HHjY{?&FI&Wo@s!KI3ZdBr zBt4#U0o_XRucM8!3EJ=kBzw5sL=3W~B5a`JP12c@fR4apERD$7h0bbcN4Cos&bMdz z!Vea`f9!1RFD5s0frPx;lh*86b9X2_XYV2AwGw=g5#X254D+>Tn>~@j{0%7YCyOn^ zO_S=Eg4U^&N{<*AUrq*Q@-HWl~aRvsmAam7JBa6sInsi8WUwn%VQ;#>Ni(~Uafo#+Vu{? zug|aVj4{&e-*WypR~|dYZ}@}PYL8V^{FXQO&Ho$9Vfg3&%+BOf#&8ax@qcAdEj7y5 zqN+VcnbRvC7^lW-Hd;j~#x0?T)f4Uck1izF47(_*>u4SKi>3?{|k zem{PayM67#i(4CRqGfX13Duv-3UsMr8Q@nBk^xQ#Vg}^DbLsA z{r(j|r1UZx3wNQ6ybjpep9E?6W3D)EDIs_L9(WuWld~gVX)fiSm9Q~=$r~ANke^*G zVK_(b-glGLr)6vtK1*gzbg;#0u17pDUq#8Ly%4|N6%kdg4%^M5fZ*rxabkL1s(b;N zxzQbxGSb8_eLQEUwAcD=V3??RaISsyi~stw_oMD-&f5akLMF_{#;^{&Y5ODvreR7Hyn?^290ijVkp-_sbaPSeEOCp+X&#OUf0MkPcw}XR9Ocf1$+#MgXn@ENU`&Zhinp z6D*$x*!Y~x@ggPz`d0w|d+P&s6aWInaL)!Cx85APuPF+k>m>hfHL0}+^O1?=;dPi< zsOd9gWDHTxn6%sROx*V((6HU{H8n%_x+i<~a`8OGW+rZ>>>1nVo#i%!<`_F>v&SgG zvq&-dQgy5SP}TjJnFUeR{i*iYbRJ1VY$i!+{~j{TVDj^_8KvjhH!8cZhM~=)1!9hq zosE~>!K4C%OOR6+YR{xo^PGbRXO_o;wx)!6R_m&(lf?_4_Uy)G=sTO9{_xZmEr%}l z$aB# zapu^Q|L53gg-ZkVQ=tspW6Vp5YZS}G1{48Til*X$1I^00K5Arw6UF*8nTGzT3ae6xH1h_P~j~+cD(0HL@@aPe?{i8?d>)4p6-*oC? zj-$Tdc)c(IJbHvj^4G`Xnbpv{M~|2uX{ac^_Fp>4$M83w|Mc+lk8iGsHjRjIq&-oN zBv7KBURZdTQ@GJhP}ta7Q%KpCLnl&u>5$^%k#D-lFR|B4EqfHcc+0E4%UAM<=9EAm zS=NacMyJD}kh436q$6{g_mDhmgGVkrFoZ|M+CSp8S}Z7(1Ow z`9B^-5%p^;s+Gyu|Cc99ZCt}t3r-*Ymxs@XvFVuzssD4hC?XGBA*u!y2BH5kkH3F3 zC$q=LPs)hF|DUt`>!~;fb>uClecXRpMAUu+el!p9e;Lk8KUI)Pg<12zjz(^e|9Y6@ zzm4UG8VmWqJw#0eO^IcT`G46l&=0w^)71a+uCYZ<8zo>*z!8xKdzX}@1utN;b^~~QOyfGCDAFAPk04~tRF--wjD^;lvRq3utdNgq9o=aTlMP??Z zZU@=|6LoutCTEhtHn%IhHcr{hyH3l7qQ@w`AHF5b-2k65&PqfiJ|_8ZcEl1ZK?k7n zK2ay4^+QKqBKkkij-{nirJPscA1t?G3A1YtwiTlZ2I@UyPozmsL+cmhsp7|sz67x< zW7!}I@`t&{)g5)pV|cbGwoJ&#w(a0G)e>Z+KaV34`$Di-c$6akn=P%F2S%SOss##N z9xekzf?93nC!G%#JIMvrFKyt-NjV}O=xVCKVyi2@-I)%ITi^XCbxd{Rh(yJEq>Zr( z6)^fj*}vAq-))@8ZMBUgSBwW-elaG)GnBPUzR+!l*5Ym&eZ3I5JKL!{fFp!?8Ta2U zdgV2{oE%1Wtz1l_u4TOhV6P#jLNDNfMJMoKt;e04fl-tUO(~G6L>a(CBTZzsqoXua zzP-?5{k%LS_Q&y*2NgAxvJ^`#@Nz#nX}`qlVI)DqXNOFuP&!E%4%z(GE-c}5DWX## z8_mHlzt?`W1thk||1JFEjUs4{P4>7ldw}PH^2W|qm$R|og^W=_K$1~M`4d`oT549 zU+7oKNr=FYXFU##!jKoY4;R^K{?FNAZ^(5kAveEOLQ0x8$wr^+20Rp~)WZ?=`(bCA zDX_ScQo(0Us12lwuQb6r?rS8!hW;eKewaPIQ+xhg=gp?Ss${Z}VbRK2`{p@U!b=D|LDy30Md8B{t`KI_i~ox4TxA_g zhUbucb?*R-moj99oa!4V@3yk@f7Fk@TBx^^?(b|?Mzj#Jh8dCd{JGQUx;o$rKJE{= zTbE*#^LxrD8^*vW<3}Oye}pd=upweF&;O&SAYV|TN!oBcH@t!L3n!oKPBHoyu%^DL zB~a*H2|mYY3haEb5jCYTTWOb(q_soo?*96BYtD$y}7w-{@W z_2Q^sv6oDQ_}NV6JieX>-1wh;Gw2FxULmo(TW@c?{$QKPZvl&}tT=hdB=twy3PL;d zVkd;b#bIzUDtRU{Ns~QJc^_Qx7`Inf0||M>E;H=C=!{S!>efH^o>vAPveaNl9my)! z1&kTyos#WD<570KJ6e%ESZ!szA7GvJ{!c&jH4Oug(xhdFltYJ!SC$?QDO6XqVearn z%-R>&Or_cww*(Vs@mPEqO4zaPyQ9vL_mtY1$WLAbl-q%-zUABw+U+iX;WZ68;VW}D z3m4}jTkjtGay^~Rzgb}wKqDV|q7m$WK_?e{%M*Hj$98xAW$b$R><^N1?0E8qMJ6}k z(SWiPOWEOou?#?{L|nzj=a)g1RVT^xzLCvOkmy2J0DX)9LWcudi)RvOy?9`i#2(*k zR$0xW{mnoMKCK`w+m|nf%A7pU`%8v2t6$DmtMiZ9fj)#Q3jpTm>>V9Vka;ozB{T#cFMJ@!j&&4J&duEY$2dI7Bezr*#ySjvSI$i`gIij`QzOwP6R z+_UQG>%UcXbAAp@f8 zyU+~!WYGNBPkrK6EH1NoR_`%2bFH|cAWH7zhiF`x+_Vq(m032hl}`&|p6m@R)qV2W zW!8I9%+3Lj{2U3HCozOagOl{r3oQ=x$#l@jff!%bohb%7*#<^$Z!BWQFsen#Pe<57 zm>1MUWDG(cs)`_17`t%+|73w|0_sRhwL6nVkNzx_aeb|-YKo!;>`WM`vqZHW#>Udw z<1GGu=hj>IB9Th6e6GhZ_OnVxRb=b$v*^o1&o~9@=e#_&3f0TBxvuX1=e*NElWd7{ zX*vM<&vMA%Id!Ag*06p?wvqL5>V}J_ob_S~?Vqor9-%A3d>+d^>8ETjikCHe06{UA zc-IS;^ipPE;N@Oz7nzbX5C1>A>=OlW;xVG!I!|M9R2+)d)-W&Yrf9xwZ*x|ERWNa2)5(*HI~8`)3%zaf3A>;`-KakE%4!F|g6omS`9bo# zu|M~cHGIsPzXPj5AJWq~bTVM)d%{$5@7_O3CnDzLQ;tb;BOwSqXN9Gv=Gbc& zd_^M@o$hw>j5^;lQf+1EroHS&Liat@oTphj9iI ztA8gO{W(o(mq_my5?;m6r^+jn{M0=8bex|~lepIl#HAo( zg0i|gfv@Z(g*j66tm^Xb1}?vT6`gD3vuH8TFZw#WFxg=XAO9qrwULtUPT)Op9i_4_ z<#W-YSJ=Y*Y?$asc;+lQ*memYN9c+D$wvo6q<|n|{JOtAALw@pUZJ3+rJNG|+AkM! z+r*~!Ra4rHM-gO~qN`?W(7>j!A>B_9-I%ZOs=TDCYP>Qw{6@#&_8KPzBP8EO^7ZT2 z#y-V-8|PB~$%VaMX>4Mb6$JvlZHGVlBIkU?RhfW2_x-%n`Y}2dvxP5HiA7YZdoImf zMrO8s{26sTNg!?ZdqTRsxm%>3r0Ngn`gm1x>@)jU=l(<9Y^q-Pgr?c9D=)P!yK-~p z$FgeO3?lF+ME~t3{2)|ZT;C*otEkX03BGyT{QBIyU$yq;T#dF0qRdYgFf=sG$Eo0J z^LPdTNXyEPZmQCH=2XTiD2k!+>C-}&J2$&J(Dp^jQ;EU`yjr7Y8+%cQ5S>s3~g-}BRBOI6}fuMG)QIPk0|NRH5IR5CFeKyrM3 z>kzxxO|A}0wjP_)=9CPjQym)0!Y*APvCSwxAH<=nw(L1Kx&S1zpUQcqr}trS zuPK?%Il>OAG0GclcAc<$8Lv7SAEz1Z{mNK!k$N zBxrxA-&!*XX`QY#$Ovv2G7qt-GEg< zHe>i@q7hcrKgOcHui>Z^H_0;Yf}6s2zr~ryLWoh_){cmhF`+EuS;An3%CibKRyJ7- zYLibdUcCHX8&+(Lv>$|8U&twqDlr&ADzoMZx17S#3+oIG>E%OLO#4RO=$^;^KxFZmh3Dk3innz^TRr7{4_4l|(_R%5 zd)y|P%o@U`GNmDA2Bt2uDYbQ!A`AitY9@=`&p0ZqLF0?FJ@7%s6h;~Qwr#msQk!6Q zokYPE)grh(eB^ARKp5f*9!31#v|K4*Os|j$|J3N%DK%}FeudVN;v^@%Pc#<3{jng; zCNuh`;A$xDO^?6Uz4e0>MA{GfWM4+-YU?43-^#<>1S~b)V4a`eYhB#j-O{q5OaC;h zp?C~VFr8a(_3>U(lzY|J-JE8A4#Q7`mK*@t z5X^n_GkH_UBc&xG>W?(|w=gr~7WzEATK4CZeTul2!j) z;ZbGW{(UK^R=rk_do9O_tVa-X{q(%~9~C6-h7Y-Y;>j;nC_;}v{#YHPQ#V*m)*dkIQl2Q76HGyycx9`Jll2Bc4DyX^yfrO}BIlpOS{f+>>W z!6gukq8FWGWyC^?&gk194^QdDH_eBf$j$!Fiwzrvdl$oMXv zd0e6|0}Y%we{|*FjI92~zm-#~+k5;0ENsY_ez0pNgLuUNS<8^AJ-cEUOGX0d5OjY& zpWs49uE@?JdcT=h7V;%62L(b`gAI393DyE`(L%4^RV;RHC8C- zw0T1GS~%;{N#-{~0(xFw#d<)QQ*nUMx^Ie!!m*e8cYu1=vKSbOb>e4f1xD;G713t_ zsQWHuV(gwWxy+JOnYUAoW%K`tC&#OI+mqoj6JylO9`E}&)L8G;hd-V#PC&}Y7|A*Y zUmsdnZTr~Hh%lD|UZEV?ryO31Mpj&1#>KcZEHT*77MlO>Dc`P-zUbMPs<6AsH~x(* zc%JzaU?a_sCu%lQwv<{!D{o1Ox|wll=c*Nc#bX%vdPugoStju2&RPEiak{NGfX*aG z8{G*uVZXEe{P3~8A%Eu>mZ~Hxmdo{K+2=oMEuS{7!+~9f%C8IzBKrILZ)@%2ByF`{ z{PGi8NCHi0xIys<>hsXq^Uim!^4iw@L?%Dkjf?5Lp3<8b3D7C07XK#gchtf@^wwS^ z9kv3NWA2Z`)9F{OwsdVLhub^rcz9$!j#1;}28EdCUC46mND$kxr@yt{xbj6KSe<`I zTi-d#Ab9#Y4G0&RhezQUMa^JOly|BDCFucuA#c0zTjE+iqf@IrTCY+b57H59oXlB( z@JQ=?cw<=5^gBuPabCWAX3EbO0~(G_(xzZNSuzegMXj(=3Plox9V+4wkZWUC8vdnBd{WC&CAX z6NXxiVoD@o7CDl1r|?xO@r&;A`pA3oGvIcgjKgAm;n$PZtXHj}HaGE^bAYac}rK$~WCT{lwd zY&4PW^z4lHB(>EKYlx7?>=I7)az=S&#$l}Tmg2$Tz9H!2yE)PzE0uw{%R_12QgY;< z-y!Qlg&|BB;`2*ZpJpGjczdA_G9h?lg;zfw62J}Gr+)(G2^vYqbA;x^%6WLtxb)f< z9R}}U5>v4q<2k2I9JS>WIS$kh5||Yd(e9BUITk9Q9?rRR$c}}AJ_g3CakQ-BM*fP} z^(EF#*@T+gmzZ&r6)?%%&%nUqVtuK{%49FCoQ43Say&gZgAwpzh8nbGT2&$GMvyhC$G zt%;OoxbUM}lRh`Srp$W0W>J;k20P$3maniR>?9u!j`*i|gT5npRwr6Zzw6yJa_pBJ z;%y;-Vl43A;U=U)ZeXjOUQ9azLTB&N@(O@uO|PGomov#p635u4Si_+fv7be8{$vD@ z@iS@;8sQPbPB(mp*8~_L4jk+4T_R*2RtXv8!rpQo@|MqsLD`ysv{oNBzN4i6d0e|= zQwMgrj`vS;E0Vx9hqP%Qvm(yN-`m5tnQ8?XF@{tbl8e7KQ4N8|^h~|`Sp>fnFbJ*S zo7hH?fPygo+;XQD5e>vXN0H{Wvs1_Feu05%-H?G15*?9a^*^1oHx(flhUgu&O z{Uf|VFfmed^8j1+ml7s@IQTjRIrq*y8=)fC5*&zwn}HE0q!`kPwP*cV+^SiG368`p zMQL7od>)Dia$%QBNaBp|I_FQuPaI@B*GKJR06M z+2=0WKj3*%AwZuhfKm+UZ;$pfzo3$dY^F#S+zL+cFXk4Gis^JxhjCUUE@zf3L4K9y zWu_L&+EAez$GiM(yCr3J_kltwIxPn~b_GddcNIAq{rMmQBe$X5z}Qv z%yaEIvrCN2H-+Xtpb^bZf%t#{A+I@N%NX~$9)q--$v(8CB~)o|LG93%eNMxpA{AIa zdZ=dU;UsTYSk#PA_77WuSQ=H{f(Of7B1qyb+Ct7*XN`njy>_Iu6b$>MPUrv0y=skz zL+5U3zp2?pim%5q#L(hwV9l4pIkYaUQkM}7w^=RDfgEW>C_UF1T*jadpL0C|`%#}k zpoZFiezx`o-7A$<&^PcbGA`|7B^h;L4zW%KpNZSYiQ91zd>Ch>@ji4Pw{+}5rw&F) z7N+@pz$8F+&vV%0j3oitJDE|7@WWOfv2aastvZtqdmnX4aK?T97w!MXlI}_Pjg1V4 zD7ilHgx%($B5luf;Uk0P8hv|#T%U;tJ9=EJ*AwXPxY|@U6ij?3f1|lSNqd@K(<%=M z8Ztt%1{($8nKX1}2|Cne5}&1PpS1CMGIq<(jbFy^9w2_875FjjXWr90|KWXo zzuI!M)Hz|O1`X9bxu-2m)q{2R2QtnH+T$#9r#M*%!9_Rmkhq~eYg}6EzN{b)vXUvg z?EBC|#Ng6WDmANw(b=_hfuJQ__k|#x_%&|W;Mv03Sveu>8U%&tov%A$^NNEH-2hPZ zdorifp8Q_CMe)xK>xgHq=+p!^L4#I`KzTd$FmUNZGK{Ncy@h!|?e&?GZFBxc>64zC zf8!n(u8@B=_ArUz1$A7Bic`FhxdF$gea=s!gLqzE6^JfaFPJ0dE1`1${m?o|x$XkeF_$5VK8?)PYkEtrgD{Wb#%kSg8RzTDqltvW zu2=xmS{a8XDNgBr85{$iQF%(+h2z|6VCo>U}kDI9(A7jtQMeMijS;XbB0x;T# z`A60J>YtEhkrfw3SuXp{{Zv+Sj@ph#Iy(7Vpi}qZ8fn%k{<@OMvo5MJz^xs;5*-M;5=o?bV{`cBD1i6d*+FQW4eT3v&PW$`1PS~~z z)!77}5>9!X7z$+@fSL4TXp(nA*db|a`BTZlXrD-*AwJBnLTXyUcR+IQendkhmTkTM&6p%GVwQtl5kAl^HPg0HzB|4M zpFt`X-6zoAYRFQ!vQMU(yY)42mmKmKP11!Rs`k-9PB@^zv)J+>~*ES z-CmORGCxGz?uTUFukN^80U^adzOtwIT@c&n_{FY%7(ag2MacTq#%e^6X)Ri0>t|5n zrys=%5I*CUqt>c%Pr{IMHO2ZL1h&fZU|~@lq4DcJ5A0U{r*(K5!AbPuR)6l-dFhrX z7+L{|paz)q$dUFWSJ4iFIl1-V32iF|b<#=Dv($D>&wt}F2|V^Y#0fzQ0$;TM8;aMK z0&@tUWURtEViQAU)}N3;h{8DxgbQD3d_JkSbSP z-OgMETgg;D@e?c7=C^YTe=Y=5!Pr8=$FD(Eh>eynbys`;J%vkG=*HePCKlf5hfR%!7u5~nT7 zZRS75axHGlbmj~~zh4$HYerDRPFv-?og^QFMsz!{dot0fdkoL1ih&{9Y_bYo4Jm4m zbtV^JAesS+ZVM{$KJz}fd7LUNK~V19Or)B<&JwL3a@;b-yqX47~2> zbfS5Jo%E5+!RzDLTc)!&F>N#>Q9|wB5e*1}0G% zF}If{4?F5i_s&H(`c{gV`&7$cTH7vl{$?$Xs`I(oy9M2EwTx~tvcA+3>Ahb)I!FKO zb!X8)L(iL?6ehbLKbYsF6M*mK3q1~W#-Om zz5iYVO>Lhb2F^LuxyyU`kzu1?M<@b~HL*w`8q-;cRcMSCL_AYbE(Fj_^{yt&TmM4H z)7bpx&r%`3B4$!-a}rnNx5TdP!acwI2z+53lV0YIrO+6=WPe$|aQF)Q9gywEdDmE8 z0ZuTE&;oxZ@+A2!ry|hoFQ_H94RH-08~TNGF9V;_1`KpgEccN0>ZkVIjy7zz=ik?g zD`Ki@Ty|ZF$Rlz}yya!#kS#_#E zx3UWe`x|heG`AP#o z6tj7bNK@FM$+H71fc|!X#Z0vQE@xJNEx40K$*4XJ}_V(p0+|mO8s{o4JR*&x$Or^T^dj7+0 znd|WkuC`$#4N`@Xm<-6Ry+h7n0T=!nmg0@+Vs`ZXBF` z4U!sY8Tbj!j`lDHH0G0nAjn4N(TRE9)|Sc_@$ped*9MTO>hBPFyFoma)D47WTwI!) zZ62T@gZtMCUs${eyNuNI=MP8F(jz5SD`rg>)3mKBWe`FA&lj$>6^T}dnyDL}H=pmF zrz&0l&ZIRLe;K9rz4_R~UNaHZ{;k4%8vG|Fh04N}UT|BRHKnzoAxN)%F#^Aa$swq` z??WXCGFsL8S5Ex}iM^LWE>K_xCu#AZc0qSWz=qO3^~;!{F)29ss!m>bLp9O*N0Qvw z+8uToqr01;n!&pUhXd~#{44fh>@rK**0~2N&+oUKH9-WZj+=h}aPQef2CI%HHv;5Pu^N8jt%+@yJ&f^skdx9e%E4s3~` zYa(WxjpdxfUi%>mrIoeE+y{0C+(opesWunkCnyR|-Tzj>Qoipu+SvD5L8!Vv?;#|c zsq{8SXmY7yL9iUZT`BbVuZ%iUMtE z;6A2dflWY*ScU%mP+EZO0j9&AkZoH|l| zO|r`Apx0zR5224M-ez@qfzD9gvg!;;($Nr!bR9L6CYy~c|2i0DFpr;Mdev256I zPme&Lot(ii2&R>bz18uLqK_CGe?+Tx$49mxVfJP@mEm3F_NNBC)3l9%a#Y+{xyzed zpTw*w^M|9n+>}mRYY=)LROfpkBqJw7Vv)O=;!C)i^5^Ok#CmiX+gxKUxpFb+Wa;E} zaOX1ueh0dtFHr;625Kxl02V^e{P6*l>ft1~&&D2M9PW6vlH^M$_kp3>%-T66Io4EW z9yfuckSZnPT%^pgoK=*Se9vTWczjH$0N%q%wb=HhCjviP)?TE%BEkCNyW#+UlGI(X z2Vwd;|J`GVhM~@TNmvflBLY9DASaHkCh)X{kRw1+-i#lnLWBePu^hF;kIEe6Kkqeu<%$pYJM=5h1|*jo!v8OyM>Z5 z*G|oY&%a~2JR6@k@;l#0e zyP+(-32K8cJt2lnPsqcTP3-W5oT;wZu;1Ydp@O^}NvgS_wV*xgeM$zLI;({apH9(5 z#DuH_!Klvs(FM6Z)j*Itv^Nj9$*f$P3H$hEc0gA&imCKDPejG)bloIiy2ya-^Jjr+ z*8rMkfU#v#Te8*w+g7+>ic~_dsc4*e#J(<-RM$3 zVx`K@Hw0Bq#Zz)L%$k35Bm-8|&2OIE3%7L&1bbg@qax!2NKS8aw;j#P)D1%htOTY~ zQ+ZLq96@pBZ<2iW=d56QYDE`g1{_4n5`$+Gi=)yz(}q4s&6C zp}r3cgmZtJ{k1d$fjynBK5baonXC#krfirXdMbcnV9mSk-^tPwN2p1cGHZi=8#1Yf;z*L!9P-kfNMvvQl?yNXnEjjT#btx99a zp4cGyFv_F$eZB)LISly6)^%I>>4&NjSo2V-p z8G-!4myNq~9c&#<&cKcCPTT$ua>gyrEisbD?QV}t4EaFcQ4Dn8g+Y~B>yv5Pk~*J# zOq6)C(Gn5>E|NntzerW|Mz>6>P%^-lr^xILR zZb2Za{k2&SduY=c!vSB8s|#)O?1uI8+*@3sHj={1Y00-oUAj7k6w9`=w1lE3$!()u zVyk7+``BZ-K3!tU#}j1sRLVoAOi}l&cb|2M*uDV3TGjze^YPmq$w_Ur{4(5r;hK!5_lEUw3C~BF;cnMAkzxGBQjrg5Z-D zF92<=PA1c_y-pD5@B5W$V@L?)tZ7BWo79aRTCslj--b3sX-1T;`T?p`fk6 zg-1ZWua!)Z7df_%kT*9sri|BqifYbl0=-LxCv%%3YDk(_S=mPD+y38C*F>P=G5hBK@fda;Z%Mj{O5ZCZEQkuPE8iU$n$6EkfD9`8P_LyH`ckqo|+4bbw zRq@+&(RSB*}ZZ`tI75NA@5o_liZ0+x2UuR#0?DZX7E(Uq@AR`80NckwM>2+`{)n%L( zS)0$-Cq9|6UYV)t|KKk-3Fj#G`r~3AYm$@+iz&KSCS<8Hqtwh0tW$T=qM?ev(M?ny zpH_aMm8hy@EA!ESBrWHoecC}gW8gR^7dH+cA+-;&#-PV;T%k`BTjxqcM0I&lDYUb~ zdX!6`8t_6kX%Iy^5KR1wWx#k;#b_?53PX=Z4_@>>{mGEMWcGLFWA;~gW>eVZ*tr$O z$#J4Bznho1h2DL_(;spp+r3-|mEK1=o_Nt+izuLHi$lj|OKoUHcs_jVhQ=(_#xRqT zg52aN$prM5*zR5jb*WdSJ4V0#bNpf~U+66S9{4J?@6V^0s?tbI!p2=nih#(pGC7n> zLM33!CPfkWjSX2L(q~bJ+^G$Yg(DJI-d-HI@!;*SZP=jD`&X%oZKj{x)3_B)D5(r$F_}A{mkyi^ zqPL$rJ!_=4&t|@MIw9n^|Kvv`-Zt5w5%h`BG$1b7(FF>$bDX+bxP(1z>OMy>4RjIi zw4KYXcQ>q^iei58E&$34V^(RTPIH%~H1o>}>VZB!yLlbQRIY;8zk=D z@lrAfPSJ21cCLM;Ma(OtJgFM!IhtKN!C!v9PXmD;*ezt;(68y!?R%~CNx{$C=BOeP zfmE$#mI*TV%!o9YcQV_ZBnZ`LqAZi$>aBp{dma+i(Wt?+#a#Q?Mj7>?@510Eu?jn+ z|EMoZ3?USyjP*Og`3YnD5jWbZy0ktP=ez7fVD8icquBG6CuNRsCP`sZ4QiBfW9-ND zmi@tEJ$0p{&fwtj5!H03u#$9(YQkT>@#s!{B?u0$d%GiHt`Vy?^P4NyM4mQgegfP# z3#j+p)}k{PMT?_#EVV@#$fVa%sYe~)yBQnF@`c7r0cP( z)rYGIz@G=&DZB?s?Tk7B=r~9`MV_5EPl7d^8yPKRq-UM-ewg*jEq&e+MotsiTZK@3 zWf~|1ED3VcOKu!qqSK!BRZDH5qI9h$3No_S+k}v@I6*kx;9qg#>Gl@2olx%mbv9Js zfHVcV1eEuY$QcEp>}(^74GVMU(_1PBGoK@M1bZHt_|0f=iQq)rE8ioCm&oPiB+Qwt z^@`1Hzsrb$aaWx-t#P9_S(t_w(I zG$b+2)aq{IlM%vyp|?dg9TJ6EqOdPbAYIkFLr11N&7?a0;qU6AW4GM`XHeLWcO~nQ zYxPz|ihqEieVTbeVMUL_5xTe<<2Mq}x&*ovR9-_rk&1q~KfU0=xaKcR_Va6Buz-7r z^0}I>{N(~c1U`-QB2ag(m_S-K`PY`9%f!;fn~TCLSgm@4HL3kV(iEBeud84ls~t^k z=&G?%^$F-eGBiN67>g2#{8^=SWt6H9K*99OJ@31w-iGg&PF-bm$#LIqdxp;#_y6C-_)V6#4w^%tKbL<|{cezxskeGY5H(yHbhJJC!$3Jt_(7b)x-19Yx?p@}eG+ zYA!0pw9EeO)b?M9tSjI%??QI~rxtINB&yG=?SK{ZHAwV<2~dP=n4z~J%ZLVEy#!OYLp z)V$cOrq^W-<;<+oq$cf>!;=u@cLV_HP#hr^WfpS+6B164Fp$)vmz)OI=)<(Jy|}qs zEYzObO+vrUcNNohyWkUNiA}WNr?&?uZOHf^;BGn8obv02AbF@{=%Sw>f^;jGk9XLcmsK zD2nJqH|-I5?JaKwKbaM~`!>*Ox_JEQrt92Zzw+@^dSPDMxwz-Tr0a96R|Vp)7Dt@? z?1pMz{ZJxS)Q%wO5qI|Aa+;IuS$?ZhiWk=MnC2r*2jsh_JWVX~2goq87Q&Q)8I+a1 zeDN7M6NHCh0X>~N3=O|0T-I0a#tN}ODUnh?jE)zcHvAqf0(uSv17jbD&laLAT_n>_ z6wEHtfMHLi#;jI?x#yOw7v4=UMo17~ff7!K|B%j^~YUpOo^dJDdxhY?p5gZuuK^2aXY1-6!Xt-9x;b$jn@Z9&iB zxuQ;Tw&@1M*;)o(BFufbR_~67Z4Opi9#7*{SZ|ZE3I?rM3PlG6fTti_kkU27c=^QO=d0ALw0Km58mf2{bn%3yh>3Ol z4Iy>-rgoxDTP6*@#nDey@uZ!}OfqvH&8wa~{N~zGLPLEZHs5YU4tLy&$fe$kUte5P z%vi$O8nSLk@oTbQizSUkL?`g1nQ8MJllXOUDM46zez%)Uo2F;8XZm;vg!T`b7Dn1k zd1O9}C%wC%$SNydzi=@PKT1%Jc{;x*(M@!04p*jpvfw4oukta%q;YE4Q-|5I~g882-1pinznrzakGvcHuSHMl+sH&*Mh z#l78&)zwxtH(+@>qYT>ow$NnHPN72=zJUsrSCy|y)pIF!4>vaWSCYUiov^(OUjOX{ z{Vr3w7r7T5)5vuN<*%fIx{817usKhCE2bEi5bhHK!1Kjwv7-{0a zMRwSnY!PP>v2w^hZC~UX9-mP2P^^iO$uP>8M^YnV59qyLVNkbr#NgLf9$*nPue{bd z*ZcWfHcm)I+pAf|g0-To%=CBLXG6$Z^3>TAqrrLt!NJ(mt;}jbK6JFenOYN2zVTiDbnLd2?Jue^2RVx#tS1BR*SNiZ zHO9fHr$yGU6KuSU2@grLR&)>*_}Zr`^aE2k|0Es!yIddBD6|${Hcl08C=fD8CFXbf zn&FMiRarPiGudz&08<@bm;fa=O14W;l4_zh_A-`ytgW4Glj1LtHZ=i%`p}F!L+K5x zA!lMaI8P_v37@(`%#=lKhr+x^#QM`D{Ief|Nz+ z*k6S(aetXXSp8h>{Pfd8w~JJX(S-W1?`+1Mj?{D27U7XtSF={AzFCpeav)FK>eNqE z(zrZ3^=y==N0NrtX*OYI#$YVALgriYTk+uf1-n2zZ*C={;q#_Jsj$0i?|tDR(?pcg zL5!kX8Tq_0%Ik55Q$-sK(Qnf4sFqogf@yy>biz9e6I4?Ou6S-*S{$o-vT3ihq~J?^ z;OUEhsrRQ-|8g+#!vc0YZkiwJefE7(<@kYN{S+oC}=OApxx)Dj0z$rI||FDjuln189}`aga-tM`8p;mMPt8FP|}n0=*g zYx}2WcexltO789v@ceiDL;jMjbL$)@w3!OxbT|MoL7upQLW2*DNh~L2BN9%QkKdm@ zHJ503tWK#txWdf5J6}w`Bp&q0>Uu@qGz`nfJ*m$3$}ad%a*j-$G=(-p++nVf-SQm4 zLW`YnOAwW0U?ObOXDSL&Z7pH~=j2A2e8z25DY~Kk@vhauf8!e@tkj0qZ8rtu zlI$(J&hoow=vK-d1dx$sTSHJr*erz` zx3?!jMjL}lj_n3UZx%XSo{B+&$N*~5tWEC1sMaHobGKRiQq~Nhcl3Or>xgg~qitOL59z;s9xg0c*$&va4FuqN{07O+4>~(Y(pj9YYIU)nDODnB+ z{S(S#+JPN6a)Mf6LP*S0TO@FMq5hqem^LHOWI%bk@j`j`*TLhc7SCvASt(n+imU}u zKm5r-Rn!W*{hz<3)nPPE5sndGH|RGsKe@AN38H>FP8(BSoIkbEha{TmVMBHGg!>Z8 ze>kenwHy6Gq(wVJOA0&(@A!rN0edF@)&ihx#Z0Bj+)+L*Neu3$MVbj% zf<*w=aoSnR)umSDc@O3sDR*|?!_~-zYxMbo!{k_~O-(esG5$>9JYbC;QuTb>O5 zA}!_mO_PN`TM%P`&O$tD(x7?G%@dc&$9XH1(R-yaCE5vMyK^m~x4v!(V5}%>FNhA+ z&t6JQKhKmp9p0(>5cMXvX{fUFc#h!wEWE?c+PTaXwee9rX>I@S{QkDAw)@>_ZZhij zcH?F~nh@lr`WY7E54il5q}CjG`G z(<@)t?8<;Y-~A)YamfhJ40hPXroY=KVd}#r-;*ioL$g*z#8Y>{4o`6{Y($(5vxR_Q z31vs~+h~f3OUxl+BcP|KKPi(=VDQlCR+4Q6#lJ5{ywei^e zlustT2@CS48=qqmZ%NLUShzOzRa;TUhqke zqw{=;BEgIk4VP}Q#0GdAVfPOFM~(;9HIP{q%QZl%TIjz&+_&-a;B|EiN(t5qOnv1B=#zo zxOljVK&7nzfKe1tt2jQb&`CX)Gc*Fqd!&duUA0_CLRc|BG{KNiK5>Uo91NEaBRpz{ zhpza{qcW5uWWD{SF`0)+_%3z&@p>e$HynP{?<9|ug!o+*Yv)cn?@1b+Il51kJGeHv zKyj~D=kX_UTfwBd<0|t<-82U#s7;~?uVW(mZ~p~BDRobHr%3gz#d-02DY;=8H3qqR zXOVlShx2p;8bjdMFYwOGwm%P!S;w}MIsb>S?+&E0{r|715Xp8(LiWnBLWpB!WQNS- z*yB{P$0^wv$KFIpX7;8K$KHE|B6}Qr9KXx+{XE~#Grzz45C1sl-1mK5*L%HQuj_r0 z{F=(As&_#;OuF~rR4z_m&EI6Tn{blN)lTj$Fw`NQ>a8Z~>7(Gc=(gLmsAK|L;{KDU z{Bdexw~1WM^{K7zuzapt_QNfh3g&g> zXWvB!n^MHg6sc*c)xbAVzt)~Oz8+D8>Esz*IK!+I$v@fpjmC#AKW=v>iPOXULZlL% z3=8|}6BlcEhG(?%#nNngu$F6MkJPVNsa(7&{;YI8m;okdy0>~cfckEw`o?i68g}oR>x>hm*ZkP?72QX6gOnyv$WHXb%=Rn{)d^#lQA30kg*NuZ>I?-TamQsT%;^Y}Relr@#KWc?Sc4 zCH|$2Nr!iEz>RD?g#|ABoahDn4PV-J>fPz$?t!2ZEl8GHao z?Z*}V&n;MDFGK!i4AXBj)#^}3<$T=m+OKKHWoBm1bjt{nIKn$SZaCxo4|Mmtqiwe9 zpu>@7+Nx+cRT+liLjkK11@Mx{+Jz3M_{0Au25+Bje8BJqr&3l)PBRn-JWK z`R`p)o0w9GCyU4pSlsK}M&WzfDxrl^gIA>^|~K4-$LL<7A6f8job~bK$d83NG(p z$+N?rgD!;EJ9_MnAQkJAdh+ArEidmG>|O=OP3BCN2)34| zHDF4zbj3J24NRX@>Zndp&$F3t607F;(8}VJ z76Z4{_fk5Z*%2B8%krc4V_svSKiU7((5Kms7mc?uWK4-mbs6Ν=kYr-84xx`u|e zVxTAGG0&dKE|IauIYDls`k~-RjX5)GEX)A&5<`|NZu6P@k4t{8oDxK%>yf)(WTmd? z`1bEMXghMUvkMMUtgkkW#!eARwGbl@m^hUKfBTuKu+MZQ|M(j_)*e!FhT7u73yMoe zcu)!K0ojvhz<3foauCMaI#MlS8@0KeIvUloq!T4qCjMkl_BK*-rv6+n3f!SW5&5SiF{%&DE?>y;$yiJW*jQD*?|J^y#V`;6GFvhw-trAI0E11n~{#NOMyG``hj>Uw>i`$L+~o&WP!z>d#ZA|Vol z4AuEFa2HbR6dHY_KRal(o=Nmvy+j05`%bqy95xN^f5hCwhvs}FmN>bglTsBFQB0=r zu;+pL3;dK?kIM+p{k1E^$LYU~m6*{ z&ZHGEg=ROfn9-_9!LwDGOO*h{wY-1GhBJpbxDX{1pk z!w&$seROuRC^^46R^qY{Z&B;kxXE04dwz4pdofh;WtDGBY!?gN-xnhMq5#&$0N~3Fcr~r@`8_@wp#YQR>^{?a5j>mBo5PT*%pC zTVZpQmJ@f8# zFd~Aq_H=KejDV$+kKXHmK|@nhU02s~D(O@Vkxod!>Jl<@y6!BoUkC2G$YW62Od;ZW zIp!lF(wBdsvNLItg{})u6s8G;eqqvGV~S2 zx98syOP0z|xbM+7r(ze8mv{;Xca&ES3QsX3hE=I&Ta!&;pbLjm)WD4;f5{VuX}1;b z247Yc=F>T)b*Vp?ZRYcJ-!Cv|OY@-$v8g1-z_#T+C{^`hm z!RyKo<9 zriDgrUWxbN<>hT{z^*h5b4+f^e^%F=6Xf9FFsR$lVnMU9iJIec)SsQw4=%a5L3D4G zUAcPw+hFBP1IPgMUFYAaN}iS~vOQoDAu6tjygd(sNC^c&Rzc(Y2bUposP7xtuL*2y zCiSkIsx(c9!JV}3M~6K~d%H8Rj^3Up5jLNzAI!F{(R435k-^mn@0qp`o-|3G;H+29 zhgwg1?oa8KSyN1VU{s!rM*k(7z-0&KXm?-2OfUbW6?t;?CFa^)$Az1z=4S$E@Q7avDSVP6FrCrZn{Y<1-kC<0ZR_h*dIi*rYuBK^Og^b3 z3VFTbBRoCc*NMCN*KDlfHE0+>VdinrCqG)ZalzWg;O=T<%bnN*D>;TYoC#X z<*0(k*;PqAZ7Dh3WEj_hBDtkM1S}yq8Vo@yQZ2JI&DcX&O zGp`{mHYS>X>%?Oz!BgedR+UTg{(ZX2Q_;PX?{%BH3$%sKxxPYGLzf|Wh534A1iHZ{ z0RaJ)cX=P!f3!&Tq%1M-Djxr)6>lrN+6=|lUv#%8!G>pVfNk9ppgg;|EV>B7)o#q z2n-z_#1tJn?-0`fCX|3?c%k8=i&%%Vww`XEK zFa$l&Wr^9P%y{)o&ah|Sy>tVSe%(nAo75CQTN&2I8cF3Nec5;|=GLW7uLZs7_NgDl znCfi46x}Mz+b(J>>syIGsQ}xnd{_(wj zvgJr&){i7w0Y!2GWRo*}ALNh8^-JxiXvW6I0s;fSitkU(J2aHpj0go5Cka?ZI5xWW?7i@=9N_nc>+Js9ce;w_aS?rGReAK3s&FG6pIik!}7rxqmR$ahMrMlJI6 zo|AXl*W+`0o{eV9Z!Udo$Se0m2lP?%ncrl2l%f~ns}%EiBFyZ|{Mo664f*DzZiMgQCv}hWb%4H z>U}n`@oNLqp1ZetxvNoJ%wuG~!+?4E{Pj&(R(qua6crEqbLMc3PZRa|$;zq33AKTlGmwtLU zWa!3aJyx8lbIaxX#oSfxJYDrg%t#a~x#;Ezxha=nHF;NpAnY9_ciz6i=Zuf3Ue!6* zON{>9R|EwOYKPKsp>e@DWSiBJ zm+PpDZMznHap7tzs7v>|@Si->W~aqjTq$e__4dBBlC^pZYEAko?v$>~R?Z)DrKb$<%cs3>FV+8RhIIxy7<`*GRxp&6&k5dQ%4_#e03P z*>u$ix*FK?vrq|1z?Ho*aHMh29T+Y~;w!x!{ZMs-Zu#xFXuZ*^-}RZzk2^ zgHL8a0!QWRS4zV^J3TJ#AYYORP_>}h>VEb(iIQ_~aUvmqVyDoZuhXsF`z89>L5 z`J7fq-@AC#)3Pb!6BEb&DSW<|Ab@%lIrP-OE<9}Bc|tAuB0>;Ze>0`>AdC$m5eOFxkVbXoaD-YSNhMgp8 zP7hMg=BQkbmP=yRMkYMZCTJ0Qt(jRCbKB7$xeFPl_vS^W&!+eDd+t-g(lRW}=QDX5 z<{Zz)*E?uybqILtkID)kPhdaLpN)22uv<~}u z&-X&S5FfsEUV}@e04%Y$a3xDNV~+h<>?=;SP)H8*EkgV%nVLSo_dH6FpWJ zaCV%HW9p+OmN$^N5P|+9D0-q!c9VwdDq{Ec4Jkj&E#*Tu{ULi?iu%T%n{GReU#_>g z>~6}V)CWoqm?WOgl_l*PM4qdLsHObsR^tvD@ZINBp=&5cLbvSiMO>N|Pt@Y(xX*># zS;gDTGWaqmll6CIUYxGqD(Qe!4;i>UF-T_q%b=>J14BLZja{rt@{C` zr2qD~ zl5YNnj*gBZvkuCYQBIu#tzF~$#38#LEnD-g`mJOL&r0!Au8>SbsI^RGk9t$TKk zj*hDP%KLX=XP}v~1i!=b;3v(GIk`5=wY#1{F7wMrnR~wVCH_RTltl%5u-%hXcF23J zbwTxC_sXpXTJD1~C(EoO#vBy1@u78$R#wQP-4!Yqx&hshLgSh9JY2{(p<-hOq^en_ z{hv)>HJwP}v{Gi##ah#kru`pPQ+oxie)XmePkz9oTqQo8oJ{0x zaQ*z$B++Re>rW5|7P?&CnR5POmuAl(v}Ccf%t{RFx`DH>WMMOAmaAP@zLI+iM9-t- zx3G3yi49+39Dfap3rVQ-KBH%BhF{WXhi3bv$ULI|>k@6eA><>(dY z^eJ#i#wqQ`Pw{GHm^-S&U{^3LAwHhTQh&)DeIu2a-sFS6azMEK9m8v}`=+d&`I!WQq(w+|nIuNfQV@5quH+ zJ+Z5dx;dV@>EIHXPumqYCn%1F&$dY0)ks)r2&I^Y+rs4Tx3=5yrtA5kf=|=r}QFd*!#j>8w%Q`omG2I^A-j2gtLHvu}C4u7!s2X-7~!w8Hxt+7^sa-7mf0)mDY_Sf`I^q07ykonG)Sx7+Tm0P>fH)+U$5Y0aO+s?PPC|)sk7jnK*)YUElE2J zORXu#ANl6flfU>8K?7Z{oGn*oH7fC*t~M;SR5wuP5ePrrYNele_;Od4?D9I%ac4S# zC9NBoK>QCuSj&ifXxyT4o{gb~e!0Dh{&iHPrSKcpb5j%A69kuH-lX;9Pjr{`NZ3^HUe2W3MzRz+i^`Ne5M#dhahgho_2E*KVq zpyz1;H?VkAP%Re~%IF+z;mljyR8Z}E?4E(axS($qzP_=UBKzEo`Jv)tTV&aqtXVy=gy#D#JSnAr-8E`fKJwCLs(=&a_Yfxi|RMzx8>S zgi?l`Lxr&cQ4dCpk)BVh-%%NR^u@SqhTqeC-AV0*1+FI{x1gE>EW-Ym*<9K?w;(sQ z`p z6Mcnud@Gz!mc*cTCOK#?ZQeLX>(l4mRF^q?S`*la>l+4~KeC_*lTmLFUjuls12WeEucKBrK~7M?Sb4N)(s zBA4wMt#M90)i>S!R=Om_9hB~BJ9Fz6JB*jZ`dE+E83M(+$!pNp*Wi$a9pm7uo{m|5`h0wF*$3l7ly(fQ9}pou`zv(GgmV;w^CUrx{cwFXC(WU z8KJncs+$SIFUL`7r@ay`GwGqRmW$q@I_Z9dRGCf4JTgA`otJs9lGVn7-pz)m(-F@u z9~Q!}yFMC(*~DAadUxRlqA)pgcBPEP6n7p#$Qtt5s9ZupG zd=AK`(`EkXg>G2zqfu%2u9+9ZNo3^Ew=w!$m^TvoMwdNxRzzQP368|++gCBK@_(of z8orHrHym*f8Ugm$)`Pfy(P@Pwk*$jD2FdB5XN#%R;`VYYPTt|yg8HeMff9wog2~Y zf=T-arsxNMNU2QF3ZV?93wdM1cz2)o-Z<(IUhI#8l`cEBpm{U^WZ_@9y@*VC*mBIa$~!z7$(^+&jo9V)-eK`2Ql5sJVM9Tg#E!S4N@8H~fS@IMwcTj(!G>8Eq(C>bCUF zzc%{5*_QJNFhf7@PttC^F488LV9RPCe zqt9!sSwEPaVj(;9ulmw+Nj%q)-*{{+xUL3(F^d2BBMeu~cuWPl6Ub4b{Q+I3Mq=tyOm=3KQSF3Sl$&5$3vx2;Qk7RFn36zNBkkIeR5i z5{dmHv?FrzBT}{Vobd~Yx@{Cy0P}li+4ovtLGdl411wq8aTUAloO}%%AY4QUSkl{Jb-meb%cQjqNz3{Ha?E#8+nxSLMY_36{u;~a4c z?Gy@_T%OnEhnT^j&a}ndt<^$C4;QA0d+ZP1w4Z)b=Fx}wIXB86>NqAV`SlN&UgE|J zp0HfosJ2$J%URPxjeyjXZ}KDC#oUp&F^2`c#i38sWG!2A?Px23Hsq&)2bTTN7?H2W zI#j{$$uEeqjK_8|s{O#N-s~8C-xGe6=k+iF9qoEYFCqH<=|ewu#rez^@0^n&U*t_5 zW)=ldvMU^C1wIpADRDF z+flFF7UJM8ikRek`RP&ZIXzjiB%jXUFbSwSO$=bUbuoU7ui2vIdlyBAeR+RWs{0ju$JaF zO`a^eC2~P-g%2%=F*An+^IKmTVi-&Z7P?5xHYyI$!^U;yDq1i_!jY$@L3%Iikp@!YoXPCRW(uzY|R_8pPNj47(g9Pi?s|pg;+;EIIokC1j{* z&aUxt67@o#fCys0tr1Q^?&UFp)!t;$FgEh8<0k%LJ&*-g8d0zLaj1`-yRK5I}rx?p)$8-aCDE~=`dE-Pk zw;!P=_OvB)w#k3O&2FN+42h6H&Sx}SqvVR~9X)3*g_6*hnnR&~kQHJa}q zua=Oo;K)Yxs>a{0Wupm61s@o%#%4=Z%OCWdov1pD7&^PccX2psV&@aNgO zKihhb08lWurh^jL*bvV7rHfsu_sCMN+RpN9?Jl1oV^z5>JE@j>y5_NIqOV#1EZV@? z$b52ZWJ2?q^`?Az@57HCyXQ6j5)tWG*0MDnn5oIcq`qlOP`lmN^gB~AuJk~SNiVvo z5zy4t6$Ex^)d><&LwPb8VCd&rfvYb4BDfF97M?s*kE8`572m*l%5fnp(Z0AZ>lGdK zeV5NwyTkhVhs}^Oat7%oJ$Ql3M%{jA-jfnrfk55tl=92~3VP(p3!mDVQzBZS(gmE% zv0mpI*hmKb$@i-Zu?;7ghk(c+-7_q#{}cwQKq%NrQnr6ZG3+JCyhhn*vDuW0FEX)~ z9v{-YvLDsC?b;?nCD7#QgMx~e=a`4%EN6f^@|Mko@K0s{a2Ou^P+sG>LOJ}15iBJh zf$KSvX6;b{N6Y&|I>vT@+s3lqjVACT{G%+!FL;m^nxR7@P-^~0l(&wwEL6j}cBege zm^`go<16DgmDX;}V)uH>x$=eSnqDa_=!Bxk6l)W+!aZ$FUEA zuGh})MX7pJ;ZxJTgs&UW&ZZkTcsUiDjvPkvu+jOpmnu7IRR9HulVMqS4w{-8%BTXt zxk_weegyrN$rUbSHx6gb-i_C9@Txl{yuol8*7DO9eIuy$ecP@~7j?yyQ)0;7aEg~o zKEa2W#R&yU#iH506(CM3Q(2soRsRH!d+uoDwv-un?A$~o+gpMq~vSWzM`y!^hG9!1l)#F=LqG3QfJL&d9Q^3m!(Na9u|628cI>sH*$1yLCi7I~nYH~em_GNh+%VTJ5-jia-q8#@$bLMfDS<&z})Wr+0 z8JHl-$)ZD1?3yfG&wPXx8wXS+?N~==x1|?9g%&Blmm191AK{5r$GHpt8pdRevg z*mPRw)yZ!Hc70G2LB!BF*OTht7qvd)CC-J1Af;S5*qX=Budu?HV;3*gzo(L%K#bgZgNBUjPGoGz^QW1oZ)=)ZO_x<&fw;h$29vDfFaQz=&HANp>V^@l> zLf-&sCOX#oUHxTBBX|GdOYK<<^TQ33jloh8M3C#gCo|yn#!9XDrSX1RFX=^3 z2{2PbFMxslBM8eN;}d+74~$g?=imRCRtg5wb)N2jXY|5fC+mkCx^l>4f6BR&=0B7E z_m4oJ3Qs_wKELzdhBm?meZc7lZ2bcG-rt@rN(kfL5CV?(G13( zU{3o-&_>No9$ksm#!NZS@Y#T&ym0%IA7LF`YIZ1{Sj+c|1}wiYNzp48{`B6js?aph z^GAA6Oa$)!higf>Wi5j%@VEc%o!|Qp*V3bw{E;c;c2fHD$1T3}P{HvjYaF64)ieR6s|NKQrMMUwPOaJyiLnP;i@4p%QKRoW* z$c*&=@wlwga=0M3AOg0;4>Slxk{53y-kwj^f3N=JVA`_-9l5XaM*Q&31i=mu?BOCq z$Sv`4?B~NZv+A%xt@^Y@?hMAq|4^} ze16N`+qqxgUQNU&V~GPrm;4}|gojI~SLMU+RSt_SLmLkxSza{aWB|MU?-N|OfLe!7 zI;_wIdxwE?YA*2L`O%J*!^y!G7Xr3$&BA=s281#vWIX^1zc>=#ymvzz`PKWfNyj-A zS7uj8$!m18d^#Ay-xwo5A!%pqxs$+GyTYwk%9dAOVUH5B9eujemO-l8&>BM9icgHX zbnZOdJ8bH3MgmVLKh39Aef~W#1p{O(0+$W*PUKR&^Eh)`MxM zUv6OAo-)oi_ol(RfG*S29=ZLa&8e#8>Be-O(**0Iq`NC4g%y1w*XcybwPygk72Z_i z$-35X&v98x_Yj=m1Z!>7FR;7#VEgM|m#kbNjWd)$BEkgS3vUo6g`wD+;`t2dGJDZM zA}tVTmP#U}HjhISfNe>v#n2vi%{!@U48l3<Jp8GbI+PgLYNeBh4rm^&FBR^#%d< z+?kKc@ucdirw$~1-Yt_hgfH&FT5=xelF~3`-tBV(NACT=y^6m0w?XV{m^-`s+jF3b z8sizfl_6gIa%$sYE`fl>^CzYcRal%=+rf|Kd_X#Omh6`WhMm)pqQ^^5(f^)b`nA?dQynuS?aMzHu^&Lla&u(0cv9QgoAtS z(VRFTwAV-WYF#m%1%@?^8Cu+Z=Ql4ocY6Q5UcX{Bg@5CuzD(Z+g$LZ$^fMmgGYtSg z3}vu9JYRn8$?Ad;#{BT=!i!DD-Ya0xCeo%B?0g|{L~lNzkmM<+vBI~K>ybZ#ij&@M;AT= zuRnldcoq*WwrRN$4}qR%A+bG4BK%rezbFcb!E(ZjR*i=;-%api;)nV1>u?!V-qk7x z1AywA1FZI&74}A7WH7vbw93IkR~%I6v!-A@ngj`63}(Q&ZSfUFa(eu9AgjyOEy8WR z_HZHJIWa*cvEumyh8Dr6k!?EfKU-n?S_!W_5L%Lajv1dx_;1sl&lDx1t}$IK>Dw%GGLUc_@jTk6PR*Nu0PvHHq9K zNj0UkBT>NS>(igWwo*xsZeDFO#Hl~s)7Lzi?WgLb>S#ElzYLRmdC)3}HNgTIH1B{L z&;elD(nf0s@Js_{em~1e10CzZj@8nW?UI(HFo4F_9mwy zd8KxzI9WOO@$t7Pm7tH_hBeNK>K|I;Ok2Yoo@;)QPxY|wdFM|+A$axrO}sqJS9v>F zOA1KM&iAg5r1Cjv2NhfdiUI+bQzFgN_yjNly}L5=_wA)OPeTYHdRY5`1ww5ez^~CI z9u?NvyX`E~r3FT^AkP}wf}+jmey|TG&3-|J-vntO4<^_6lw-gRgm=EFYR3}wEB_mE zNf}IKzU4OFT9;ZTgh;al=RHc1VM<_tO}xB&LPSR_jGX_-mX*k3DR?y>B5&q}BsDPiqI0TuMVg*$tc9qKsAs8wy0boS$#>wdw?;DwxD_yvE3N%Zs|Q9~nOwg1fzUgt&d<@7&(_-^od6-p5%$MM{c zFrF7E$5_ujO5Wj@c$(otWBQy1km;(4qTZVp{-V<>*kSfM2fCMEd{PGlELG7o5VSNY z--yt_vPPG3JoN)~bM>h}T$#P!vgVkh@M+nh5kPW=^z$Jh7lPIa$LgH8P@ zk_+N0iv?3_FJ|gJ<+T9GWXJif2eA8HAHu{vxWFt9sJds}LIL64xJyk0s33`df zl#V4iCql@P!IH}oHQUd&t$9{b32mcSST?uNZf--I7UTQ&5#19qg->F|0A;=tpYUBt zDKee`rY`8VyJS&FQN4F{iiuNn;5`Fu&GF;R>o+E;MXwl)V18K1RUqWi{ULE$0FCdV zz*k>505!vifK`7pmr?D#1@V5M#5G?YDMVB|tjh9$5;;D&K=XvSr~UtZB9x2uXped?u%A!Ffpv6yGt6I7lxHmQCA`tsi<`wVbhJnBGAp`&c=vcegI$wx zxjR7|@KvH|8+jC*YZ;C}P+=?ZM#HQ<{*r8vev>biuxo|;{w=ZDLN5vqy-B$(%agqt zrwHUmy(h`U$xg|hleo&B%fy!IC2Q$r-|AgkvFKcF;h&t3i*^&rZ0Qod42g2A3})(& zRs#aN;-_FLk|EC^($`v*taizpMebdU%V+|N0LwX0st>%4xzJRI1`RF*IwwN^YSplv ztZ;9w29=|{_-xBxX#yJ0I7okLROt0Kpv@(o#WC10A-ezPt%`=4sQX?((PL*rDw2)N zpFudxig|*q0y3F7<;qL1Z@KR99UsqAa3qSSwDrHb4G>z&{OOXP%!<)`s^Xd*DNJEe zo|80Iq$1ka^L5q(gQy1X4i%{z?9@m!h*H@MdhpEJ4F3ecS8(;NQIX@@f+eg?Q83lf zWYv6@rRb5Zgl6vE;qehH%?&zIHLt(6WT712+CDMk*rm!kCkPj?8SXT29D1B~X4(w0 z2ivjN7jZh9h$U^2XG^L`Yhr!z;60W z`kWfGICzH82zCmN7?>8+A6Yv9RB3c?&>m<_8OI0DG_4m|B@YAH0YY^8sanstp#h*r z_;&QaFSQ7zme0S%=Iv?**wVu+<@jfSSEc4tpAXS0`X<~q-}}}8BBIqQN!bUI9qDWj zu$f-{_!BuB&$6$(k}8?V@*>b^H13(Kjuxf0N%Df?Y5NK&JJjaFGXP!o2 z!jnBDJ-76LphhoCII|Aj`SIbfMXEbPI^Efw_bgUr4IK?95^nzdddf*2Uk%c;uY6OK zc*bg5Ctx+)_$bS!zt?8jfh?wEF{4K&oDf2DH6%LJ?jw&OtV5v6sM6&j|I0z@fi%-a zz(3Y5zGY;%H}Ezz1RIm&s!6sLl=S?2I=iNwfZcTCuud~6Z%|O$w(yMC$W1Y17(;0? zKrn-?(daz8~>l@Q<6`uq>Zf?f8lU11;u4lBP`bwOQ^$W83 zLeH^mn?ioUz$=9fCWV62zwR9I=q?84#9qJCgkmUZK1Z1V2xKkk5G0ZIGM=gSypW5L z#B}nfvFRS~ZwLdiO;{DCTWrnZtLMIUq0!XN>KOlyX9JLWP3S{-@UK9)JL|4C!=NAh&Uo@#ZrT|9dVvdCT+937R*$U^v*Yhh5Ruc;15RPIJKdMGrFB|wzs<-WB1z{xVXoF5;WADs z3zQ5bWOgtDX<^cbSg~$Ynk8>F z*8S(%BoK-8?!H$mA|1XrFCUJ2CGOagWi6Oe))(q#!esg?Iu`?P-QB!PVNM7%FBfjRKSfvG@cn`)#gj&a8Q--ZB#u@|feJF(1uE%4 zhcM>V8tImedXf8H>{@lxYR;8yHLlkvarai>U?G4&>smRxKqK-u?*3jM= zVp(V}?bdawwg*X383E9mhq8(Ep zmo_JAU|Fj$&P%PfAoq*2Qg4uMo}PaqEM`m3yEVZS!COh~n(tx%`&>3caJ` z6#`^UQOhleOn+b=khNumt1B)U!DF5;Wap(^g(@$J_7(3Qo2;;jM=bBr-o2}|`0HmV zKU~Xen;o;ATj;w)W;~8^A%axb=h^~*nq+>ZppJ(~1Ds&ZbF*V)22WP#p*)xZ%jx^~ znNfl?659cSGL=2C>Ez2BK(afrQlpcrQ&_(CYI%_TSq-#i2|0j0Jx&BA@VFb$pqW@6 z0^ZQ}35eTG!z)uY6;mmUBOO=UiTR34 z?;SrZ$%aAh=()K2YlOHA#8kdwPPn;4I@k@$n$}hSAJdJ~p!PY{N3>sq;27vp$$Iuq z%t@22<$PSrBm>DyFFaqF2>7q4>(GUO+b}uCMHr~BSP+i4wDPf~$(Ds$WeJ`&M!V4o zEoo?Wp|!S*wu>XLp8~IFXFQtv%a1UmNr`u)LhiQ8kB-~7VSeMvtVymm&Dni#It+1d zA#w=P$SDKn16ZMEgBZlziy?X;T zMU1g{hLO=P6;my~{n6Q-?^qrPKB=}njD0g+>Z+7&55%dS{_WJWj2ibopewh?8kaAZePE+b zpf;|weDPGW-k^vvFC^%3D!|DlERZ^b{@PTm?<{)aX@g?Y6mCyiOdVbch)C* zM-+qU+IU&H{qFKm7pO@um6^6)*9U25-9o6?`rU@ZF2w3tbpg@o#)>dJ z05JDs+l{{0CbXCVX#gEDa?b~Kk3#mN`x7(kO3b*wd$P9QD#C4Li`{ZnPzkT%FsI!p zn5uCWP83FzBTWGugQf$H?KQvtt6!rc%lio|cdt=$n=r|{sMB1DO^0K>efOLpw`$Mn z#4Lp^|9lfBet^Q@sbpA_?gBU^3mg)Z1>CX_os{_+kYounV0>7t1p9IWAE?YNb0gF2 ze)1Z02GfYQ_^rm`e}mlUzudrJ+LI!_7sqzgBu|f&tfGUX6@GYtVr)qQ=BLE)VSa?I-`Qip7mgLs zcgVO@g=+$9EMzo{;)-|G%LJO;F{fQdZ?Nsful#X^y&b4zg2@wpJYr9C8`&)8AG6nD!8cq<%cq;VuJ@Pczu|e+MyR2DS zMAp=~4d4kwYTUXK)SjTuBZmn%EUdhsU(w}7_`?$t{V1nKNfPAFFLj#)b=)veEQg47 z2mb&ey}w+onNnriy;b@+x0=WZ$&|Ma+Szr1<^1y=E0M!ek#2qm0>)0PfkP-aW`3TJZ7AF#x>c2EhAcQwL6TGsb0g= z)hICh(d?qW?NMI*!)G;6z`6GayM&W8A<(^NUnP(F!j#Xjr~IdqqA6>?1MLsC z+o+#xwcqbkwdRE3qHT2SOP75(LkVDn*Q;+B=H1B|jmU*6O9c$~oEwR$uM~}`z2GDucJGqT* z^}o!_QV{kggaif8g-MhAw#ZR|a^b?{X|g1WvyhAm~S%ulWP!+F1e1t2`CIxp7ag%K%sg#gCOQjF$Eb33BfSp&4gq_PMy2@oST96ANI zAxvH{zk-Q#xlI?Ks_ejT!9Q37_+~8Y_E;(Vq%PUUpEBs2SC9J{mF{>?)5OF}6URXl zD;-WJ>f9@q`abv@B06HY_tG>t&f#yA_+7Y9iSvvE#=1#v&km{b{vq^it*0j@ynv3iQ^GuxwX=k19(0wzgPM_4~jVBg85P zHgno!-d@YYx%AX@GLh@{p7B@uLqJvlwugb3bJZCn&IaP~>^{n{?x+uehAQuYDrhG2 zq4Swc?RuX8L!MowDNji_%Z^#z77dit&b>7>+3DrzZlPS)B2Nkv=bDhg45zkd|6U!4 zJyw7mg=V42=UWsbmYxAwEQ#Yjh+_#t2#8ZCo1+?NW>@Ty2Or(EpWrlHsa=X86_Mw) z0H~Xq9!Y$p7LTTi`QqF3TkP9`fGzZNs&#y2OTo^bjADSX_*)(wJrjtBQ$w2p$ku5Z zCT?C(wUE#& z$~3T+8}dqt03yuFj#=bf(%B+1K3$X=J7&=5s zK!$E46{JC=m97ElE=g%cKw3arQknr&knV;dBnKFV?jgQs++E#w_kDk#AAjub2Nvc$ z=Q-!D>$?yk34Wte!@Gk>mZSP)XC%OnOP_^#~YE^D?L`I&SJMkYD*#iwfDpu2pTHGg=PQ z8tBd~$YXx`Dd^xpyUxAp0LU4^aBompA~b7V%LV~N&}r8#WO9v&4ZbcG^E=Z3{@K|- zQC4GyV%Ku2ZUXdfjY`39V&M!I>zxNfgS`quhW0m1Zx@t-NPye&lU7V_p!6>+)q7ji zjNP!B*Lhv>$7WVazL89;f@x0sZC z?d{p}bRHYPHM=9j2lYB}PY9Nve&-AFp=h1D9sR{k)jgs=Uj(=^6rsdiV~ zq_-KA1!)|v;oL=fp+FQLPkFp3ZM-q_9lw3b>(s-dg_z%-!S}+SANTT7EAbm&e5fA< z8RL;Tz&qs%b!$67eZ#Vi;CM1if%3CyFAZso!)#NSF@PwIcUML;0bRg&b^MA;=WPh> zAQm)gp};WwLt>N5}q!=|2Y z(5^FvM-2X1-rkiYK%}{_ocqZO2;G6+$#ToR)rm}C+>L=@P36zI7aef5g90b=`oF-q zS!=I-hc=SE1o|d`*5Y0zz9}OzqqlkjdP82zX+T}KP5UdX^*Hn!gUDXb065|gOZt0J zI5+_MwZZIEd>mN$@QxM>P=JCN6dP{VIHF7D<0Pwl3@1m54bZM8U5Tx|53$fI))@cu zPoNxlGCys64x({1t)x{Ad96Yn|BIDW@w|mV#Ts+}Z`U)(Y z16x1%y<8j|fPK#HqO9dKW^0_*xElg*@Q@2Qass|L2T;5zna7@XR_lN7%<-Dlqb1yT zX>$Pce&N*~nVQn7$kajp%Od>kBLKUCVjrOcH~{0uExa!5C69EG+i@J3kGEnnEh;Dy z4RK=mu)p{F3v8n4&YbUKSdZkV{RPnZ2cYhikBOx=vrRo~t?xkVluJdDj%^gR!Aj-E zLdu@-bE$BWB&0(5pvBPpn$@~)Y-22!$=z?0wU~~h3s*V;o~&Fg|5T&IK5n$cfmxl< zJLF%xV9=sc|JdADhL#3g0TZl`cimom0j!cS=rKrE&5vH+*yicQ-gHJC`n>T|#fMv817THZINIlb5=-v- zdexQT%-j4`bo|X-8Tu=TpdyQ^XqZuk{{2}B z2sBvsKE+RtMdZ~7oWCEUrQI;}lk@HLAG8%D(6?r7&+1?JLU>-J9qR=6vl}!gl&6$( z^qX{C&G1&ry^(RG=n|ppVL5P1WNU#9BwDN`^%e&~O6mFRFItcdXZzop3TQTWHx+Wn z0RSp~h_aO3H_+Xxd~L)g)w(%+R9NU#U#I5OTpi}l4s#vwkTp!gk}?s4!r>{rt%;Ww zw^ADdBl~6B)Cr^Z*aZC>vD|wLZVUsSfUF${NDI$=zly9@O^t7T z@i+1h5wQfxH06+dCsOahV6tb^lSIMqndnL$y%R+Nkfc-M*GkNnnb#9Py>}~c_c0xV1eb+P(2z;j{0%sWN3*oj&I-HiRSiDSD#NBcn@JuJQm47J50xK%zMk}L3?Q0%d-ELr z!7&d&r&5s;JGj}Vxr5F_Lr%RMWFS245QRPm4Qp^d&!b9~D2=pJ?}rN`#j}o8V1h$m zyW<3CI6~AL5xC>b%y^aEi%k;xw0v6HkhUyOb{73(>XaLw+v~ z{A(c}1t7G{1k?=*R@{bn8$zi8atErXw+18M)GI2f6w8zqm4JGWiLxo?aZS-Cm3Y&1 zVLf*kNxhLAWURh=XIV=;FqX>BU-`{`C)1$CKxkF?Vt<9#^8x(AJFJ?PsJ0)y#dbt? zQ@$50T|j2jS$XMJLs}rX{r_bh>5nxC9$JCSxjmE`n8=1=1EhR^fyc!y#Z;Q1MCw^z z+7@jaaNJXoXCbhD~AavIwZ$*5ZD=duk1C5^hYZ2-7mxyL@n4>G@sd7YZQ zC(h#_8Zc1#WO@uVAp+4T3`}^Y;oM97e;=)EXSHXC7g}@zLW^gc=qG+YkO3x~93REZ z_2}f=cQLQ}{*Q4wP6x5fGn37xs6=6pF{x^XXQh>V_KS)g+^-IYGSSWD6M!#2RkMWBN=b-an7fC+c(GTIB+++H4NweB7h0Acjgskei%1=m}fczOR3ts!`*Zbnr=> zUTbY4vr~L0z+HG{rrMzcesHDZ)2JPEqWKwW(Hf#7BdFXIO2##|>jW-%rc2)56Red@ zV$rIIciR#&)HUnrd6L9EVR>X8CFhi8qqrRoqM=dOLk+kp1p9KIq|_aEZqDn)1EJyd zhp$Sp`N!YBu!oEt&`Jj*G_+2#Nb_oZx=pa^SN2 z$nGuSN0CFneG7I8KPgglS0SE#UnAd;j@Ft)Xw({q@j0$I5AeI~u5vM}=R*Mx@pa8a z$It<|+{oBM?!fvJesfYybq7E^zgZT3l`43h`g0R#7i~Y||6iUhok`aWFaZUmzMcCY zG0ybu0z*P~?85#C!+-A$h!}{K-n^sp#6~Lrj!~b9<1)&A`yWvrdoUgZ|9|~aolD{Y zl-Ps$Z&oOzA*HhK|4sjFVAgJB*#94K9y@_Jmc#j9>8Mhwca2{E@wv0gq)_a(FniQr zXl1$h>zTdhzSwEtct3(lsltPm*8jkUfWSaNO|OXELMY$S&M@Gy3mF6cvN5PR7=s>x zr+}$1;Izt#ePO`VLiJJ#-R4^G?Z7#?j{@QF0!Tg=K#J&yIa2v6I>EtN!s#|TgQFkC zUP8ktd-TUfVcQ0Ijn`<5iJcn)P{4+Et|;b`-vR zy?cD_NA$@hFk1B2>b1kyweR3i2}7lx(un=J!?amfAw*US^<1Do`vGtnMLgrH_e6Bf z$n$ILhFR_QU)oxqWLS5divN|k;NZNh`nd`Ft;D+1GGhCGZZ&9Pq>lY;mFu)J@5jg;*KFaKE;(_cIe9B()tc4r6xc~DY0X8(`lrEJzx%*R%2OlI5=5Egq}46cw59^{WJ6ah-q_WfR~ z;}ymL4%=OTs_`7lk2Qguu5-FUkN$W8OepF10c4(^@wWZy7z-Wx1~KPIxrogsp|Jh0 zWiWapOR5r7MSKShtH!JWiB0%wRpOCbzj2gCi9<^5U`pgUK7HrV9?(>}lC zh5NTm<>Q^?_Iu%OtO&0!AKqDC5to{Lh=2d-C+)bB-CJCSq2xT1$yuOpOkX~QFDe7Z zWZMQZ1vGNVYr?tcIlkK}))`uAV8WAN2aCnC9Y7SU^Q&*5GL(WIyj%|f`~7t~kzGjG zl_Sv9@&|96b~GUL3hRoU@!qP?!ukQ5f`4=SB{MC+RILz+LN`6F`<__2F&x$z!d7c( z-E!P_ObMUB zHA7EF50CPr`+6rC5FSr4ze~BwgUgvOHv(0JCxOyWu?I$fY+KfIg<}=7b7TMfpBDpA zG<%53ziGn8O8J_t4wFjd{AXqTNwrRUv{`vhVrNo(e0;HC#a~~K@C9Q#Nn)b_iHiJz z;oU%heaPidr;EOsu_1+S^HpL_E}~^4zEiNmCY@19_hDN=B}jtHP=Z$PL@X18%V05f z3tjC(0ipG$OBH+s$@FZ0*r~gSNNS>808ezRPSkV)Vw0`v*aDVMM=CJNb7E3LjC&va z8!OS#3q3j>Ep_<};c$eH7TaYVYo#`9?@e!=o@bnz8DPtd0}W#X)z62mM@)W!w=YBH!3d|7HE061gnjnjlhPbw zvfkvctUBf6!tFv!A$V!q989OSk}sF9?i3mzSr+<7@K&ma@oOp+(;fkBuwcy7CE8qDt6fUw*VzI|WA zj4KUa;dC&okQ-J~&!$y3UV2=b8IEMDoqLgtJEnX!9hBeQK&oK8STwy0?F3AQQfbO;r-s66-dGxht$Dj88XOET%UAyMxz2yn;eV+z864nyo!J*|>x|o78yLH66Rp-57-}?xoSmULc`&TpKmWeOmYe zHNN-!OcV&sYihHSb0DTIJz>v)ZUf@T$%wM>x%H~*8sJ6ii~UtO#0ff=+-{s?+HGVZ zeEKQ}6#jNy7LIS+=GONmbTM3Kwf;P>Pzle+zDF-h5xmg`U)=cR#+1eiG6 z8K!H2hiR@%Dy~9t32=?wL?m6%Swd<_VFch5s#RG7#dH!9gb|Tu0*Vf-?i@K zQvIad8J^X3D=8s|YPLz_B-;A+9}PLupqa<1tC}V zm@Q1W$5&AMHepbpsk(TmGtBCfpIwV<=k?VET^b+z1tkazt0}?B0`=*WT$4Im}GRp{R8YBx?(az0YMt!Axn|63MK7usL_usq(m2 z?iZr~r!fFoqCpRC?5*8+%M@;pfUw}=7eLBEM$v;vs6i=fl6ki(n+B%_k0kinhf6Nf zB>-mv95UQ%r2ZWg!AI+26Rti1p#>AJBbyJ!0|MHH07?(!(a4NB@zMa6vf`Pq`?=|a zg`)*aUT?LLRC>8sY4ORh=OrdJn?)|Ztf)I}P$ZD!x~aZyFM)lc(I-Wn=yuUJgC}{( zx{B`(9=;*k?n@D%ATg2Tw?PhoYD+$!SkvXbL@kHg9@8M{MMIBSt4wBJuOPokhQPci z3ZIQyj#dxGm?45UKd^^2N2Wd+2hWuyZnLzoRjuBuQ~NV>YUMUey=h8Bmm!LkOio)$>o`AQS{VvfIPlK4IwIgy`+KgLumF5m!M%dH1 ziZ24Is{Ga_FyvL1CmkJc9l^%g=w4Now3P9Sd8gHeIu+p8heXpKVl zlO-L)=j+zu%8kZCdyc%b-ZadZ-p^KL(Ss+50jA^~U&W6XB-xZWaZP9xKVD}8+?kEl ztS2-i0vWq_G`iveG*6i6!uI@Il?a&xz!~~#s zblmMtd&PS|Yxh?Sf+NZcf)IWnSMZKtOgQRz4r5ZX%9q6h6}{Q|z%V<7iB!PKT>=9( zZZDK4hH!(1L_tKWo#ZUM`61i-X!f=C`V?t3^N?o<=ggtKCO;rtRDXu5_XM@3S*CY& zoT!#K+hoZQl>KhVLIda`^_qZ_I^aFc%6 zL=taT&y$#Ui5Klc<_5?Th+0MMV+GuvK!CA}) zQe}TBJH*pRbsrvh(tw`Az2e;$lSdbf0Z9knmdEPXzTPCqI&y$3i*-vntpru&ga6c4 z4VtEvmMo<-Ybh`|8u++YZmS7FZk!;P#`5{ z^SZZLG}_{+;h1DVH-Xrw>K4mzks2X7Z&W^FU#+Sm^{tpRdkeNOhMKl?f=zv-+)%j} z{z#GOyc%|K_MEyApJ z(gX9|!H&(iD;^fRgv7;+jdq+*d?9#-UY9MgETeE@(khc6J5@V5s4JopQmRmp9pf)J8czNok$( z^9s*;x#u#WNZo9M6ryI1toK3Gw7n~qBD~BYdvFaBzf}uTlzbmLjznq$ z^Q+bx{+3)aHKiPTW~%%kzJVzYbvO9l zo#-3E>w7rUAYn-*y~gJQnp7iac}~U_GFh}N&j%S!?Wf4toc{>b^*x4bkIw;3D@n? zo!ihZYJ;RU-E1wk4$)i~e{AJm3c;~&HYs#s#Rm2(No)aSgHC>61zPY@s~$?KDmLC^ z19<9|h7r7B zfYM3G)=V3X1if)Y05zaTjgfCmPtze#&DuR5opSS?_#)fAMBwk7E1&RLz`+Ot%a}ct zC0$GNSm2nVQ5nt|PhJBR!Y$_{m1|%a_sfIuBmp*wubw+)l}&04A)<<2`blfRNC6r; z!v=}uyL%HEs~dIR2LLB8w}d0d9g!8omR#l?&$JYlUc$4(brpBgK-}JxQbMO;IKG`# zznMQn3rfqZ^eR$MUtZ6a=(w$+aS`XOIIu$Te#6(Ytl&6w-4nnOuFCDFbLof13ka8a%sDCe9me^ zrIPz{rLqQTWcHJ!fDzv76mKQEQ=5sdFvApc*<+(0nvyp)Hc_&UfOyR9`p#|>x2kbM zgcSOkfoycF4AlYfBWz~>M1uCOJK@cX&F3EnP>V0d<+L%YyS3GkeTp3(D_b*7udp6> zo`$%C>;M#JOXr~);MEOg4ggaA{hli@;H_TXT7$}(YJn6O3gSULA>SPmsGnPt*326y zcU%4LLnYSu3fJ!K#X+FvAh`2Hjx=9m$G9`uApO>T?b4CVg>_-lA-??;yE62y(T|>! zLhZ7#%A0Yx|3I_P0DQU0#OfXfbHtD z+dlFS9>eP+5g^yg@P4BYfhJZNzv^!JA=#|gbls!_>(`Y(m|o?~BEi|BM zq5Zbwbx~w3sca#jSWeasiEbBSHb=XC7l@id$*sGYGwN*?=jFnpm1j*;xSk#!)gwx) zxMml>_kdQSIDkpK2XXW$8fe6@(hd8CYP}_+o>k5ic|25Zmq$$-ml?W4%IzoCwK&)R zI`>NLz3Rk%c_JZ?MYV#E2-aoK?diVPtE`9BAG|=ujyZGnvQs+Rlw72`W%=93BSnB? zQnbC!2njvhJ_>O_pXBM3;WJMu2%=0+lkAq)hxHKrFcFvu=p_~d+9&;`Yljc&h8M&+ zK#JG{Jw3^Tv0=3!jLg-o?(2{toZt@*mYiD8Xcu7ZL-$AR9&EF>%ZRP=*a9e(<*RUm z&?@Jo%bw??_hZ5x6t3qdI5s86?|QjMK4*K89|Q>Xwyo8U_P#>}9)Vwi2v;>fC5W+I zzM{UnaXFRSe(NLY(TA5lKdX@_`>nZhdL%lQ`Ki=31pQtfV21CO6&$9KTKsk5aDuL3 ztuV+&mb30nOnfZ;nB4@rF$nK)0ek#R=<`lNR;y>{#BDaUdqC$6=CCw0A?)VRtpZw} zCP4?=L_9x;_Nn^=@2%H%CmP|s2S`E3Mz5@0Z?Xmviw2V=g;IARmsi$_TZ*)W9w#O8 z(EuRDVlBqDpmruCEMl4AY+14vGl73ow{?pm=Q7~c(v;@6DLhqh7V+(iuP$Oqfv9Cj38Ds4;lAJyvAs4YHyN>6vOi*M zD=#f#+w z7(2pQxeLg3vX7BZdn9G=T}MZir{4&H07I>IpLAsbA+#jsV)=~!Kp zDADz*3VT4c^k}`!`n2!lFUKt}@ut6G zaBZpHL7`=vH`0jWZ}(CsJ8Xxl-@}6l684HWA7HDms(!-9xFg7JD9e7YcF{6y5>B^F zY#1$ZNa?pww?pGB_(r{yaxBoayK#0xA?X&hy^?60jVo9))$|tsREci~&DfrE?Uia_ zMxFLklQsD-8TY#dr`3%Yz|4X?lgyg>rP0DAiLVT%u640v~Tb+~U=7EI8ySfr_#M71_%6Ab$yU(hD z?t=;-eJKN6WNVlhbmMoXJ5P?`+cn!9ZZAb8yKlB~W2nb^&p{*^(zz&ChK)eiPIre7#sW_ZO^ly89&m2%f> z-s`-Z0wR@z$bFNKqn}NX?9B8o##s-2ap#q)KCv_bG-!(*$Z-ta z8}51J6T1IZosmhWag*VSN}O=8$VdGHKBSIfnFs!(AVb}ZY%Bcsi{nmSgPUWbQXji) z)xv|5g?AF&zDl~(q;<&2ezH!PIZVGRFZ!>P{9HYmu?i7YggzeVLE_P8KX|fF^>?H+OQKj>#=cV8?+^DvYl?kdY2z&h%5x_ z9j22%eb=Z3B-j#8`ljCAdHmB1lfU+gy)a^*q2oAGC8zQNKR9TYI$G_)*k>^*b_Fs! zt?P!RM#QxCjiOR7HfEgqCov~f$BRX@ZSBES;zwu3l~IVSDpZ6X)9&G<=Tc*of>~di zG}SrHg4(kXYR?7f){=cASHN*SK%N3o8c?Cnhp_OCuGu2*t+Uc~g|k0_ z#t}`hpWbpRV^u1#+)3*Jq*g7pNy`wR+(Ut;4u>>bT>%@!u0NN}eZ#;noUp~&ZGy7A zEL?ckuM{af#U@felbg3yH#-&IGh?AWCETgw^^w)!k1Kyv>rjtFw%$N_#ZdH0nF&?; zX6)&{zffO4z?o(HGZhrB5V4Le>g^k(jPP_-J$ha-&vx9ahmYN%=N8B*qchyPOUy|% zd`Zb42L$?o3RvfVv3$)YoL zoQb_~8DNG-MgnF$zM^Jxco!3rsBV%)+BPjv&Nymf3m5?Ic zlFJ}VL`%_J0!Jr`z%b1kpEuc#@K-`6{fg^YJeMNMW%`D(kMyOyvA+GRWn!OtIU-fg z>%k>M%c?3KDd|gw5@PIK^EH^&*x~Q{V?0iBu{t(ttaNHs6MRTr#mRfG@HZ}^4uN6Z zxd3g7@M3llnBR7%DDM0kNRgcE=J{lK5Oc@ps%dH`ehXf&rfFw<*#Sn#v-q^1Z70Mo zu0{jp85f}&5Oq&M9x&lLXep!*YT{5`)E&!;C&vIL(uzA1E=ZjkJHHZ-nju2At^nho4&#}8Z`|=-iB&`OzyXLFxTvyP zpcxik8#EOoUWww0qAa^arE2SOBuMx@QzHzwn<>5Cwj<|@S8eWxI24EJ?f>K{a3_H) z_pH9}L+z0WWXUd50?pMGL8x*Jv+B>iHC|NFv{+V%Gnq^4u9`3>TQ$?F%5=cL!-~1CE{m>g7 zu(F~4KYnn`^Iv~}yLvMGUw^=Q+*hv|^7hw^0Q3G!7og?i<2dC0f2|~}!2ZY7F^W4$ z@y|u#{mEDES(KTYJOMJfe)z%f!7Le1vKOy`{0@BLEF{{lnf(G3hacDH&x$OG&Jv{n zdc=5;Nlb~Fc0v2IvD`^DpbH=U-u!U<=9S4>w*+lp?Ee4lLr#ktx~lplv|cM>Wet`K zNPpUq7oeee3CNh5M^&AN;%EX>g}o3s-css^ZP)e8`Y(a#Ax0{c_{}*2po!OI0JOIc zP#+%pZf0u_-Jv~8Wz$i9*$sS3Y54C49Vk;7=1zJEw7!=>!@v_D;`?WN6;twGJD(Io=kSl`Ed6{LBKObl*&z#-wz@@}=$ZcUlYz3R z7`n25Z1VD0PDqWSIMuli4&Wz0zf6JeDZOtd_W$v{qzlma|6HYxl?E6WqUn?-mVObM z=HCgX&AkK71M{6YmIS{ZMF*H}Cr2W4OPEh6yFt@Otb%&P-jrAUDmO6y%+}ccnQBC1 zUAlCHJP`4e2da-92B77}3}{jDeN;k$kzWkm5Cc%CU3yKQ{D`go$oAwx`J`|eQp>vW zWG8$B2o$^6eR~QX;pY>1OJx~7CWk>6;^ zJLnydt89sr_Y#2`)3UtV8mFXEh)1Ut4 zf55BsT_1Fp%+CT&;{ANx+UnH<;hg~$vxCifxh5c3hN1xPvYg`ADduO9T{aC7l{$Zg zDNlIdiV@+9Q2hkc_FnCTh#khW*J-80JM@sGv< zHa*j;d2z1`;QJ7VUouENnnE|hG0aVgE$0qeOnPU1w}Sysy5?EexYW1(Fg~_W1b&^*ST38 zA#%c77^2veU|lzQHxi#kN|V6Z84L>8;x2@5%D9XMd?NtRIa?uHSJtLGrQ1sJHb@Ut z!`^>>D~fBG^!xl*i8W>7o+(#@-EP6#5rdSabG-;vJWmGvx^lrQC3j+y_TVBZvMlaM zc8j4;n=TRw(F=rX+}QkYv*emT-JSc-6m|6=r3_salR>kp1pnA5*(_iR!pg52`tA5b zHA$PmbG!hPE9P;;$=C!>_GjmVAz-Mfa+GKjXpU${Zi0a)TyCLR$*xS+xAd-%UY6}h z5vhlJ0xdhauD#SiincQmZR}vwT`u2w=yMb>B|p!gXgV_*?m^^~Dy4N1Dpso~rRB#i zW<*6BO2!AhM#9P9vHta<%fWkggUwA>pU}214Iny=7iubc(J)kFU0c*xYS!O|QqR{M zoO4fcS%|x)oA94|gpE%}b+UUYd_ol=ZKin-q_VFWVl}BmD7nm4PY!|7A##4_ECdX5 zH{)CH8F!GL2Pf}+O;Ku5BQQ@X!$O-av<}_s;I!)eI;~inTXsKZP}e8)CUxiyk)t9x z%qNq+0$Hv%PC`ZFE?{Z%8G zBhMgMf7=_g*e)i+0rPHADL1f6zchkoKv!&UkH2DgU!&wBnQKrk$%f!7S};QH2}pjV zTqb~!L+)E`s|%d|g~h`88@?9VWKOg*q&Q|F+|- zNi{?Ho;@ujLcpbI{_O`{nlYTz|8R4xXaDVRwsv*E9Rtz3P4eCD+-ao#pJdAd-B{Khq%PY0Z`k&5F z1>B8)u1}}j;xvX%$p@=o;Lr@Bg15=8%Q*vudi9*ozjzCRfHcIR6Fr=)CiI_fG?v-W zvNeWILHNM18jQB--(Kq9y6B7zU_%_V#N2S9enQ?-RDau@l7=p{XXI_>r`V!N>^UOf zlbA&6fRsLHSXPN&xz(<(Q|qB$0Ce3d5C@n46-(@1G9*;vnRG!WalZWV#lBl;?;{x9 zMm6DedR)cw7{D0)=!&UiFo1Nm6dXhNF$iBagPM!)4}ItUGo1de&F(Z`JQn=VcTutS zcBnuHVRbR+K@;foj0N2_aS^woSwS0E60`Ov-CBN+!=)@2ms)(3e!|Eu=&tI*G;O|T zUI?v1{4?MKK5)z`O4^y`6(QEGvxPJR_aH;eavRYkT(;FXj>f+Sf}y{KI65^Kp&*!y zX4djEWg!@3O*?K`pKjQ?=m;boJM>Ut`Z|`RFd6-+G@-{G5IYcJ*$e(Yub^5{2;v)` zP*V|th3=G;01!87g$u5KIfo)Ima`>Z+(@}~A3m)$VrlR>5NzYGfNMd%nxd~t!vu(VX|Foxu6`Te+d5D%@srdeu@Y zE7)}Sdb*pE>@Efc9KGlPb{_;z_v@`MhOLmczT-gZ@xF^ zk3%LVtWxr3t4+;<3XuEZM48v4W-=!+{@E3~{)uy|azIe0pau=7vweg02RetX8}u1g zS&1=LprCfRm>zM&Sp53;TSzDqlX?bswpj)RxMeV185_`tNP$<|Sa-oEbm+CH6uy}A zQES$+)zb3T_@?}08ul413BxE40SNSeua=$*fhdpreU{j-WOM-bkgD!M1cJu z|Bl=X^L|t~>CxAs6p^8IIGHSduGrlp)d!{qT<)aX>9PIMWBRN@FO&e^kC&7FM95)| z@Lo(zom@HPW=j!#wK^l;^-(%*0Z>`IegF1&qe=KJc5`1ct$-Ixu+v2 zaxl=Ja5N5FwejKZs$0g|YK>m??jdR40g=wz?wv1`UOcC%el;xzPCGKlE`;D|@^`LnAtp-y*55BJi%y9nqzrSqWUcR2~+=U!=VHh7pJ z8#_M_C`jhqkv3f}L?2~YkJYRmD%YGgorhA8vol${b=U4$4{%w^P#JWMR~=N(_@0Kj ztTpnT*bIF-;{A5UMo&uQ_=$^$wu$O!P?0p;u@HW!88zEPW?e5-zk1Xi;X+-1&mYWS z%$TN?%0$&ny+ei!k#;6opQoT9yv3rIR??CdECcn&E5GXL{AhBat7i4_XQtdCLT~_ov=nCjV>uM9sjw zOHjGZXN3X<=A3e8JuW`DwM_Q8PdAc<7b|du@{wd$=JGNhLcxhX; zD?uz7&R)cpyP~Sn>7?Mey-#T+WTQ5xrj9@6Y z(~o$Wk9VW5RMrhOB)#F(G`iU^Hv}@^bDxQqS3@Za8|6rw&PDY2Ue4utWd--rs6H)s8))O|%h$%3Ks4 z+r8VyHz70Zs)8`PfX&+9sGWPs|s z^zq#O5wCDj_>DzcDy5P4qEbQTqgo=F7(#Ah{%^W5}MO(nQ`})JTrXUX*;^MWF ziwkZ9g_6Cicg4<&2VU!b7+hpHf#EFR5W5r2@<3i_c7@C~=US3PM$r4yDrEB<(&Qb;i=K4CPiON}sG^Lx^Nn~vafmjd|IsPMtCZHx8KaL=dW?>(q z7o&@gCjU08AQ%@b$^2HPmw>jQ!t@LQOj5rlTevT=5~(y#!=&pz^lFH@%yXl?eUqJu zId|ZSVdL@Xkd<4ep01~46d$X$)1=b1ySYzG2e3AAm}F2m_@x?kvh;i~Zy???ZHG>4 zyG2^f#(`@63C9ioCKMrchI9Q8qQLP4>kpZO2J-H1wG>TzxvrsWtq+eKhO^dI4kpGg z2gv;`>UKa6Wc*Vu507PgFJJwt2QjrXgzQ;OypM48}v!=$tv18{4+kLfy(D8R;x`!T|CI_ z);zo-nT-f%Fx0tfs=DH;-a+19Q>Dq0l08-rs@FD%^juxib=QEtaZ^1Pwb1r@4_Q#q zF}-}ewyqwqRWLby4fu_jV?{m+_r+=WYQ$&bvVAhvs%>?b6y69;oBWMxX}5Q2oieJ; z!~&~3_Ygog_x@D7v)l|Sy1vfXw>dnQm>rqqG%;bQf~oU?-ydM|1m8eLEA@Z4p`|EkO^yMcfd8s&fbZ_{!p8jrb@kMLp&<-nX}_?L*D< z5R3}^b`JrZ%`Ebi0FNQphmhtGz8X}21=t#gbC9fLo2L_e>A?~}L%mOUYD&Z_$g9JK zehy}kg23z)j&YZV6iJ>}U|O^Gd%z@Awd(10>P>mV_^xQxqC;i0d^Y*a9#_tn-o(ms zJsnS@0{dc&Q+=$w@Xoe{^_IH-ip8dJ`&mC8F47@E@lTIs@js5F*G@FAy3 z2L+c=lr=mquc%QdY>2;1AMu`Tl2VD?(1QktVAyM>HH-7otc2qH#f#*LTJ?m{Dqz%> zqE)&IBsX>bZ?`t7e5&{{%DY^-P)v#}Z`W&VI9$J`O-Z$tLHNP#7SmiVpU{ojDJJ#G zphA!*x3`0B!!yuzX169)yO4JIp~YE$##X(V0-J85S$O|9ZCB zjpr4aS7j;&x4TAtO<3*WTbsiAH!0F?5j#~sW>S!CHz`^4bj#owpcwb0cif>=RrNH; zF392p#M{eOw@}qdwweA{9KCwavyj`G zq%92L6Q`>p6=5SI*lL&E=EeEqIcZVi|8FzFlMWT`kSDKHxn06+OBXQ)ODmK0qr*}*a`&Ec1?LnB8@3E|O*eIexCL#F(l#%#(Em;q(V$3EBd5z2*-Ic9 z@^+sel$3s^^Z-RcR_TM_oqh+6;0Yt+E79q4ll?QweaS0!89&l=DF|&)KT=P(d;tnn zlCVCI(vXOoI}$%_Mtr6Jt*8{dOH!03m#lGCI^bG&s&{ti&Gw&I{NR0VuE>q)zKo@h zDarABLWd)_lnW;-=48mQxzNG;SUm@aL18R+MEkUKb;X@6g_Pr-j_p$6CKoqROOR+I z)*4kT1}$HUrU}BQ_NEON=Dy};FHzUzoS!z|QsIc8?TizsfTR1n1rre7joQ8U?tQzV zR7|UBCow{+?Kof99?h9X*BtaFy^p;)<(Y7O8(8dW@e2G3o4l;>Q4#!TUR}2{)i0Zj zU|7Itd0-nva2Xb!PO$2&0=M|uZY7;r0XI*cZWlnTIM8jxl-l0cD!*?n06W|!OAL|$ zHPj;+*R|TcKCI^kXWDgAq@Zkn^yOww7BaH6FQdJ!8TeY%?LC^+ht6xqds`l%9^Lhe zZwElEQk5$;Up&h%cK(?@0xzj?M^Y_pxG9`@ss?4fvZ@d)_qS4x18AL^xHkpdD1H-O zL%=oJ2;oE1la#~W&AT!6N1N?AK)%Qf$oY|SSR%UL)JDV_k{M#gOolh=<2Hyqs?8kN+Z46BHS`9{S-uBt+gu zLiVzs!abEd(neT!s}_;5G2&cGHZ;jZdt91c5WDg8lM13#F?sd^4G)l=OH1(9g8v=}8u7M;kt*&o&5dNSz z1;t?tZiLX>n-}xPL|_*4n#%lP_6Yvkn=x~`($HVEje^qtU4b&hoFo;kW*A{NN{v5T z&GHsP1XOM2MWvkq{$Wduoj`(ioYaCw_AbsetSxaEe*_l3>r7%NHIuPK`x{pl74BWQ zx>_l&Mhc#JW8ewLDR15=jON)*HL^6i)&MbSe0UH{y+B*T>3DOV*)}gC#{Kj=pQL}V zE6x5*pEs2zXgbaGCyAVl6;@`g2jv-fr;Q$=GItYD#})Yme=o)^_gPW84qY)Ul}JbhdsLm{kD1y%n)!oEBn>izq_ zn{K)-N|Eg1whh?xfNw^K@rLxS<1dnB}-%*yRl?9G`7JQv;EGy z`>9m-`}pCOs5WApF zCa1g}f+9Ax(ubUSNH_IybkPC-t-V$60=eR`kG-izgX+}T%GzyLVt;^CwndxSRjiPD zQpuZYK#}6WxjZuU7`nSV{6M@1KclQUjC!iD8Cd>1;)IvC(pI?M9I|#EF-P6AIp0fa zf3L0?KowD=ERYM@bUlR`P*1<+liPXY(bEK7 zzp&TB7T)Qn(knz?lfM=~O{{B0eNcG?>w64;fq9|c`yXH~v?`eTk7wZB&f0cmyZ%-l z&2Mv)E8TJoJoVZl^nce!MX%Y+Z;{NFV-!OgG9^V7MW9D-9MHYr-Zdr0Vzicd2(Iyi7X0-Hn)LEJFdguPFK=|T} zYm3OS0!Hp9@Hw5Fy;7;V678oXA>03UyeFTcW+#S*vk3h8jw1A z$nsHI)?YFq%pGp7KAB`(XaA#6Y|v{h=N|%P5d3shWl_o_Ur(D-z`-~x zW`%MpFEF|EInPp!@CwD1WX-W@D+jK{5oIY9ZCig@fo)gnpb0r}e=VgZyDSa7Em8T_ zg`a9UQ^Uw{oGb>{oU}n4$)erLw5!xY4(HGZwlkfbF!69g06xGH=$-ButYb6rT9W>c z$PvMyM@8BvH~EJYO*CsG&`9Yh#;m{!L4!kO!;Kuv1gNy_U1?-Sruyb2*XopLlLuNd z)KA9mhGEs~8xCkceMq~ZOmsO}q>C9dFzFw?z4{#!`BdjdAH0@_Vx(KMRh$t_Z5>x`f(z>RK?gQnE zz*f~hkl^X*z<6^+Q^C0`>ODkf7%yNwT0+J6BdOPoK10Mn!M4cxKKVwM0s{ufqY=J0 zDe5HGobmwE3PJM7pF80OME@v$Kf_Bi(sv&DB3)!Y2>pqQ!|+x(3UhI zKOfCUB*23M1GhE`9~O+BU8rD}~Ne>sY2!@ckHIKGkDIO(_l? ztNeXC-h~`(mMNMkK)g7n-MTHNVtEs*UEMA6jM*$T*f#WdorBkmiRm&h7{+{*fH06T zjj+y816ARx>LghuK+(z%pFB8UKHxONbHg!jLi%RA#TXut%M6=-=>9dnd;fG@{s{z2}QBxOvtmv@mM7hpPul79<2ekiQIx0xlf%T(_^ z+yp{H1Co!I+x1Q7bXCyOFEUTCHH68%HX>b|_hiAUMnEZ9JRn*y+)Cs7)Nj%}1v*_vYt(@^;^5fkLZ$`?f5(Y=CKl*r%b#&EFXP1r_>DS6Y zB-1d>|D0KnfxSB2CzL2RcUeg~oqtU*zLLTXrVl8C%#tki-H)NJz17m#<)i3L*TAs` zgJdM9ET`N{3U6eE4rCutY^KIimvz|dq!;UwoR{wQ62d+oe^GK=^$u6i)mHwLBeu(5 zG79&%HYIn9WSH6oS+U)G`!4YD1cEq>xq?SdE0_O=Z)jO==2LU<+xq4v^bt|Nk;eHk z`=ov9FVe-CuY~U7^?eeZWmq7H#M3>_8LvJw+h4X(d7ikboaIH>rVti#AF8WgADc6) zxQ&x$g)-S-Ca98|bkU$Z3bv z$nVT2r5pQpHaBS`CHU?SJRz7RYIS7tT_6UqjvNjk7FyrLM+zGz9@Dtm5O0-p{*@e5 zeh6iHmAespprAKjbz6K* zO(YHPma%#~CJbuv1usH6feZmya6>l5wI6;%-NC4Yo2yU9#^40+NAMR4WA4cZX?rAUb*aSj~4^wzy$c`rZL` z1Yq2~>(?Vg$upai;3pvbsD+VqW1UO$!zs~b9~=B!Ns(h$yWG*%N!c>*8LKPVOc#Q$ z6;1mQ6Bu@OQD%B61=vGrW3Cg=+)w!Wc66W*88odl&gNwG()08)>-DeO7{#7SPlVHn z&Ml|muE%G2X*nq&wF+w}@@PViBWwb|YpngpH!sL6S*nS$r(Xq=f(noLJv)|J-fVrH zRw5jFd}Hj|@W81|)b`T(BmpSOo>AKZ=%0d|LBhR=6Zs7HCv^Tw?#{L{!a$90?CPM> z=Rn-cR_!#%9I>C5m>_uWZ~{?4oX|xUi@s?`y=Ua)4CnA1AnMmo&lKl@>|sw5k6+@{ zMI`l>t;A|O+eX~o!Zu!bsJc+cC+VZD(`15K@chU$S}CaOD*d+aV`%P||8BMiqK_E` zugEU7Vu5V>sIIoCXt1;BCd=GIYmZ4$7s9-#8Q=cZnYv<~Mt2KC+CAW~{6z*gJ-DjX zv1;YG&~rT@iwI>m`rB(0Ird~`R)L@?w(339#iJ;b*7wObPaO+?zeTh%e_(`#pchon zR=wKeJl?55Prs=nquqDgkGAY!O`P==<-WP1gA15Z$ms0`FH`}^5e_gyVLNX-7O013ixe8B__}m6>yy=1hTw3=JnMo$BBKx? z+V>N}b5-Fgz{%w zjtE*DUq<&IbA#pp8)VnsKJ=Hl>a2{b>dbFF5O@U0luwW(a=&(bUux}+y5~vkl(`75 znuCH@^}?#cWcMq~w_8cwXHxm`uB+H4x_zeX5PyUd=*d%%T(JUyI-t2NLPWH8L*6M7 zVYjquD?OSi^SfN?^|k^xe<1!Ww8(I>n2cC+GU?aJ^x}><7rhU~)MacNCPkxdxH{OG zZv+VSc6lVAAFs{FR;94?z^vg@N#d#@Fi132Znper1xzhNks!p43E0j%kXj2l`&v>7 zZVQ%fU$X7A+J$E$LOJCRXV7$7KF~}g0MmED+83n&V<*QZw)2;dz09YYxvxAr{Lk*z z#AKlUS_5g(8hAn0EzeI_!=N*LZ2qr&rOFnZqZ~zj_m$uW0&1y&yuQ*(JW>wErAwVg zZ9pbE?{FP+0e<;Uzvun3U)u}q+)akS>Qm|mMvpa&Ft)A^VkON3Ab=f1nCEMWRgs0$ z-NG6{$7|mDGqgdIV8RGY7nFp6$j2HcuE0yN{*o=iOVRPU2Mjg{!!+0k7(AGOlzd@G z!84=CPjPZZHnG2vnb~Bi_JJ}_>u*#ebUOm9^Z4@LIWRxYR1vgZ5-=XeO=0_QUxR93 z6z?<`Ls+!h3>&weMwoxlwcIc}Y@SA*o$%>Rl2z>no|stk@R-oHsw1-P2fvl0fs11{ zxhPe_Q5|)jJ>!X#qbUb)?6Pmox9e)B0GN>sP-xnvylZ1izzMwJ2@E8{$g=!(|E*O= z;Hdo@zgvZ3;JwS-G9_N`t#(oZBBg{)dwP43RsP)~&He za^68|wMcByNmU(IfpWNcCEGs2C};u!$n6IDP4}4p2$=(BYcfm^lM3aO%9hvmVy^4{ zP``&;;Z=+ApZBvq_hCoZypm)r#*E&E#@aCo^**>0ZR&YwX}U|aNB}h!X;9w~yiPU6 zyaz;>@tM}%%RSdgD6hNX8^k4jM=$Wg19#Xf6m~j0-kz5Dz7^fxXXs0ueq9ca3<`M< z@u(GlJrhNPzFsk67OKdZC>;vm!~l=@k&xw`O%8f9y51(h>tS*#2T!5jfQ>m`vwwql zXN4rUVtVcjS5fb8&-bG*(c81}N`flb2hbcb?GJ{HkFk&MY`S(TB_!Qf%zdVBVPVrX z76AkkE|R{Kd-(*)$7@Gj_nqT4otP#IAw^=RY1)-<=kM*f?&>g|B8nGxx+9M;6^uK~ z>mMJoUYQ$u31c2e)3B1!MTZ2h+yUZu@?PbDWN7G~3RhYjP_2BM>>d7Ln=JKa26;A! z_wJ9FA77cbhCC(>I`P}P`iH;)9(PVqvHYWfC?OxeYN&e!i%Fun7cXn8st%70VydD(N4xOKa77 zbtrth2B7+V7uf0`{>g5f7$7doZ4lr3{m@AoT9@I`-Ko2F`-f;NQ6&(jATq?FAKK`tmJc+x>f7jV@RRisH{n@wIY9 z3ach+s_U>@e6>}2VJGcd%wbT1OgnR=q(jMAR9|oM!XEqyK8vd{;_fFva_7@lW9Uu69* z#A=mJXO{t-f{shbxs;b`6AQJ+TPhXLYk*H&-yzm}G=#aZ`gqMJ_W@19fLX64Nsd>f zd3xYdX`&5@9avsrK)`bObpJq+Y{Wc;V6hOG7p^pK>FLlv0{h$1?+dj2ue+_L7IHQx zlslF|VX2T4h&;yohO_?KRo{S=zT!EkE{di_3Gq<_9)w2u#4$~vNkOLZIA>oCFs(D1 z{OUmlgE2FSCsu%~BypKv8l%lu%)Jj59k+$n|K@o_{zBBk%U0{v5O&UE=;HxDR@d|; zflaOmF!XeD=#`9vfT)r816zZT&=tJ^alxwHDjp+CU1r+I^;-p%T4tf6DAt^Bq$Md# z^T4y4o{z>VW{ZVlpmAfc3XU%#bjyh)xU(x6$t+rZt<)e z|5(DUuNbH?pm7r^A7vY}H-F|nRBhe6h#Rwy_Z{y+pf!Da=hjjmgzShA{?ZLiLpFKj z>+=b$=e-{9?9f38MyIl z7U?~AY7anA^WYiN?8PFO-vuLtJnO0QKlK~JWoMFII!SEQ68(?QNY+Hi2^3Zb9whMI z92J8fPMK^GxHPe(x_=~At4FtMaYA6;pO^Z<1>XOl}l2nbe&_TtiA zCk89jnioHR4s6&PffUlv+91OXmznvnQ1F4kpUWJ1m`Q2sfXoM{s!*vhnI_qZ@^buJ zAb{B?PN{7;!vA(>ZSZsuYKHyB;Kqry3p9c(kQUzaR{gqaV7#|w6%E7~g~L?h;+euo zteV_C@L*{Mrl_lV1CL>G#8Asvs87!h_>W!Pb*)L=j9OFF_F0WrNK}scjL;@_z(r$A zv`^nP|B$`Pyf1B}&g;3|i{g5Hc>SOA@)_ECGrW-!{LglrxC2V@aEr590c4MYozBrG|z8CD}w9>~uC-Omwz2(_NqZ$Qys1 zj_ZXG;wc5D1~*w3Ue$0WY_Z()4XkQ$dhFJnty{*MEN+Q#@DZ(!3y?cfL6Ni9Z_>D@ z^6yhBp@=IBlp40hbK&icwpxCW#^oQ)?2S6r@h>bwefNe!83^Q2AMEwXSzLY0K4#y` zA2AQLVtrvzLHSerLOgtR!%ayfzyQS!{yC`1Ebjp>k>IAo%pYG%)3LW0{-7RvEP`1e z!wDLIO}1ycCoBe=nO<0wF z?;Hmk{Vjf(TsdPlP^tUq_$jZKTlEnog~DQs@=NUmMPyoY$!1mWIWFP-=1^;TBI`P} zyW|kohR@>d=e_#&8i^(!_eE0=*@;=u68kbn=PbfBx|SwQcn154jJTF0--o>DV-{|) z8^nVypS-he7}Xk|Tp~7pS4ki9M9$qITSVs$Um|Ng!cMXC4}sBeM?prKUtW&Y*499m zEtf737&kKYUcQ`Y)pg$SLjjAhaiWdt9qmL8$d;+9VV-CCs=k4!x3W&ZBrId?I~gtn zhz~pIeYz*O2SQW#uhKhU?0&tZZ>~1P*r7}8GW|n(f|cr6kIa<&or{sMHFYpp=*|-z zb!9G;hJt-;hzYSz4)67LJ7L6NAf|l^fEwOI|7=u=4L6k{er$G}91sZHX`_ZOXYMPR zp7FvUR3f6+>eK5|8hj&ZY1EerAK{WjHBV1O(=w|4L-PFGOigK%_Z|qBHl?q-*}o(; z>N9MZVh_KZX!QN+jZLb~aMqPaX~yWJB2L4Ln|y`!^>qgL&3>KBW;m>;yp1RuP1>V4&aYF+&m~k@g1iMh^Ir z@fZFS_x0;vLS-9~2WX;Mr^d%8nD-wty!RIvJCuZ83M*=;eKhNpBWTg=qM>qD*~PN=H;&gD(2bkk}%P9)bco z7!~i_RKeE|d`%bU1f#YgHR|7tf17g3Njefh%3O~>+k~EbXpFL`fBZ6*rS4vY+NWLm zVx~U!cKSPphZE@HOO+{(CFZL3d`Xa!He{(H7Cbv$M{ijgiU+8Er|C-7<4HhKL1gN} z2BdrSyss4vqdtrb(rBd=`Q=#j<#(RH+OyoG&Flq*W5;X2OKW@K$+!EmEk6}YVHF~r zz5sdrN)@lB8n<&FS}#XhJn*n!=1G>0{ho{EG6qElPcprxKWAi&uG-3+D?L1E(J&1r z_71d-70@R@svhAD0;T#XuK&?iOODmy-TO#so4J}55}M@6eG9IXp@5l}gK_T1{<8e^ zj8x0D|4`FVdt0+$POLoYX6d4V!DDRU2L#M^b`8G>2;X101)-em(tz~$^Y9<}R=)l} zA~M+k43kME)N{I{7})X0Vr=7ddZM7-J!n@Ld3wyvu~J4+GgdNdX4MKOcIEvC36O~G ze-UQ>L98Z>6LR&ZZ$3g6DCSEsvW9-?`VGU~oTXu0l^5M9P9vtxS044czCLUi!GhPe zsr?7Y^$@+ce(~vU`87;3sr*`IMZ)6YYR;_(?dD(qXSf~bP9rZuVpq0{6(aw=5%V-g zZl(VtG?$~@{;D&TymHv2b?EkHr!b@GlA@2U9N&+ron(qA{eS<+>F}jaC*{r4P4>15 zM9B)RAEu-+v7IDDuR=!3Cn}1ISaZ?D=uevvj5?XnGk^F@+rRdz@^TlwW1=r?`#1Fr>F(XMR8) zy)-(dEB~vq^{cM@Tjl_nE0)^krO?j1^LsR>`km^T2_OWK9QAK%!r>02TQRbBdW2U^q1lu)_@miKt<8g(&k+JG*L zcUrPNiFi$W)gM1*wVA#+t6zsGcY*x+Z_o^SXTQ}cw1$_l!i7!Bd`Sp33Im-qR+;{g zy#Fn-4$WS4r%&YQ!AB2rf`-ARTcu(5`C*BVN6$so-T#Fr2&aksZetvLK)cy`Te+WLnCY_ddBxU5%g z|5!rKfa*u2OoY3^yY;ie&#CIwd)O?tRT|yey{6Dgqbkv^_7VV;-h=hv8>vg+ibevE z{SC8FFk2mWUp~$2v#8Tgt*M}SGFhIDk*{{&(qg`59njQTG^lL;SkKN}zOq_W`Dg%W z^ciEdaoxfuH|CqCXO21!U~%2MN>ePt?$U7#tqE!2LjU^wCu?>q+Kkk- zI6iUv8yCPj+zh?{PU|}V5HOTkHby(~XR?c@rj}-of_+&GHF4K=@9P=-Wj*K>4e0ZG zlTNH>M?_1=4oX1SiU1wpSD-9j_>PK{38viYaldYb`2+yb5~ukDD8tdwDS5tbrW}i( zm8}P<+Yu6ma1MFo)g>HY9@)LVILuX<56pOtb%txM-Dfs%!%W_EWRpMC_XHY0#S{+@3!B=Q%; z%9y}*URLVL8jL;;rXd&qoys)8x_K~ouZCgH$r@tjog>R7_BUkduXS)u zB9Ab-6=`YNRpY4x7UOAu?Akqgi011v`E1d9@F_0dAt5b6QA`$3oqnhJ5Bnfmi;v;5 z(RaOZg%AtvrKXMFMDB&{7SBj?2XGJK+c3#84GaEkqGpd$eO~j2CX`5oK6@j*?P>9m zZL4-v7Qfo^*GHpLpKiI3I+ocvZoj|TB%a65(NK2j#RJZ5-tgNf_TSha=VmwGoh?wu z&stS%W)JK?D(`c|&DSu`%_OGCL*Mho;3pT7^9bgz?Y_^J%Y06+PH1jb+FAAT*+Ol2 zTZd?NtR>-WV9pK6Jo#h+!xhakQ=UGzuT~Z ztnHThH|~ZEANkv-woyBw(KLz%gC7{hC0T2_FY*a2Q-@am3=KOJ@yA@zcOGelZO3MI z$M7f7+r^iHS>dgtINpq=X@_X-x+dI6_MJ{b30khRv-oDJ(%tXApP`Esl=x8Ab<-U{IBDx#a;-RsuP)R);EhWq-@qR(ywhh(})6K5Y zR(B}4E#m$b!m3#HAg|OXP9AoH0fn>_XCC%QB0D(@I^3))6#jAD zPF^AU4Gt?uu~acLD49es?25c{zrs&J+GJ()iOVc!LK+*n@oaxUqoz+ylPJ(wXC100 zQm-pMc`)PYbH4QR>sIRo>_#c}(!V=HE#RQp8U0up$bk)6#rY32>u%LpKs#_fQD!Na zAv0bj3(lCqna6&&({SdOSHiO{>`f~Fr+Pcbcs0vV$P@l^I?ssBjb~A$fF%pw!NFY3 ztA8>PmgKgD$u+gJhUZ;(7j-@R}2oPa5!ZDA>#{mRwh> zyIU%T_e9XJ-quSjX?_DoBAdzY0$eS~D~_l+{$+01dn@*Eg?Y=~BC0>Ggw6D?bqmY+zb=tuaOds1#fu4iXphn$!G9tpnCD zBQt*Y<-9d!*xrqfj@_ow&QaO8S`liErCQ%<3T%C!@WVoK@%*!3=3x`m6;QChG zn{X1(r@hN-%NM&g!+-x>j2>%fO1lqA9%x(1I zE5~I=v<2EUxfM12a7sek;){ir6XjJHnoeU-VsLp_OgQE-gi26HgLrJj>o4q0B4o(@HoG5eSCs2c6xb^p1 z92LIw=5Krj{g0Ox6J3~BJ$Zy1`xqFvIlLK&cu|FkCl_kC)0TO(H|2h(Ng}UbMvvaT z-I66u*;npxBlvMmvbjp>c*lCT2lxC))uT1=DbQi<}Rz=GYjRj0<+|F{iccn ziZBEFP(ZAMx)QxQ@?}+K90DmtBK;7oNNNJA$v#-#C)p)40s9-)Nq+BAvy=@t4(&L< zS0iX|@XD+X(dY31C-v&r$Ry87-?Oke#AQiir0tL&MY^EOqE0nXUjk%E6>w)g>w7t97QcomgB@E9#k@6kK7@JtkFC0U~syj!esCjB>Tvw#1cYHmm z&^@c|1hke56kJ9e#`T_D5F9szh@*iOVaXV z7s>`AhT{s~;7M>6z2392FR5|LcabuJ;ddp%*Ijf>`5Js9d9m0Uw-wuue%d?u;v|w_ zH>~wLWt?vydVhZP#|iIb6i1Tdbykop{rleSWhdeDDPKPvye0(i{W8B@&=Rh7{=x4* z(FDv2+`F&&y72dW-LptB!l`!0Uc`s*{w}eq5b_@J3OEd}V|D4rcfFTdUSUL%?RAFV z8P-c8+b<=4{eF$L-=7`=Y z*v2Xzu<6%-$z<7b(Kdj)aOrprI2&F19}I1p=+a{dNyr-JY&mSY9QXH#=ilw$cImve*H+7Cp5r|@sIb|K(-GhL%>%W1`)_@|`}y@Z)%~zex^)_dt4kan((N^ROO3^m>ztYR#saoosdMY`!7Q z$02rfqo-a-&xL2d|II>6UV_d@>EB6G$nhjPrU1#@=Ds-AmM_V+NQz^d(ax&KK2I#S zei!S{PtGUJ6d@~eEaJ80$A|eu8e=jNp5$F?Zt@TIzrE4G0Oa-BEekN;k*bazzUP7Zi*I#Bff7yN*h;9`+OT+~6hCg(VR z+^ZM4WT}Fa2wkW5ha=s3={&6Oar+++L_FW^3+Z4;e*SRTX7)(+3d-Nc_i4R|)d4*93pe||;eE=2vu3;oBhS%}yt z0rvN&7jK*iIQ$Y4^fmu8ZZo?+EJ6DZ>tbL<=tpZ$cV@NiL}aRKek5Eh<+~wgety%x zEcNzAJis7Q0HA91Z+&}V%^CpFQ?uQf&!QtkAJ>VUIZZB9IzqjX7r5JsK5c-Q+yvy#Li;ozPB$#p#(@S0y@|60-)}p#tXbGk zKsqxJ?U}Aw_e+572>1N`@*b#Xs_3DgNX-VT z5}Flf|8eOk=$b&&=US?8x0IP|-{3=w`eL+{rj4aG=Ju>TX8&FR9tL*j@lPN9LzuYb zBF7t_h#jRigfD9;crMq|X&v_N$s<~_DIW`oee^N}Fe(vS>tTghK)v4$j~GhS5Ve)C zQ`c+kroUCjKpP+eS|GxSqMjz%Gc!Qu1>^L~nuAy;>R?{QM)6WNucX%8ePYHR5t_Fl zk8%{3%jA$|A%H_rqvZPKt>3#tKcjs37{VW09-!Z9e@(Ep?)=h&Dt-tyB{~YdU{lN0 zjWh~O!Z*p#4ZDSmB7y7}|Zkcl{9frTRqU^4Ex@^*LtVthV=Uy4oxy6KcLt3apB@`p!l z$?6GZhknl7nCq;}!R*;(s|4>(0Ucs7`>m`snVq*qxTE!5M8UPzE>zCU z|K)FJRYY3mZ?}7BwvsrQ|EsauOt%iFs|kUYYOTG?o-?E2u!b*?lEhj&=C|VKj@gWu z%C*g|&C))k)QTznH+XFZp@DEz*^)ICu5`euvUWdo_T?yVk_PvHR81 z_qz?~|Gt`Waqb%b_0$Ict-pAh?76r&a3JbGnftQ$?r!3i>M8cZ5peJp@QEe{*@lr0 zWrXZSsBk(K5t~5qgISUqc;;C#SSI19U*$SDA3K;@o8VnvrPBW+j9_)cevH@`S(tJh zdRbP)`4)Fo;+C@eE0^)Io}Zp9iYoQW;S0>jTut3bQ(SdOiL{^-Wlx3$k@v`veiGIT zcv%&NHe~4;f@t|;kCL%4)X_nh#+48%;I@VQp4iNd`zPbcEF zfz_ioAeqVt_HQ?@7cgaU*P|)&=#518aELSbt(^aQMps3?Iv)~yc ze$^(4G`swz<3@j#ym32dD*qSno4mk>**gaw^(zqGC|;O$uBLNkB!O1059+rvbS!Og zLaoi-opLRcfgLd-MZ!7f7sjH^Qb)99kHvY`TneE0Bd+)BA4m#K-1}g^yDu=X3$xen zQMz){b2@H@G`mTre104&^Qwh>TM=P`qI|cuF<0IBKuJsRQG~&=Fz;mJR<6ezx50RT zurZTTHw_|n=G~%AT9ba@#J!OZlF-!E&fIhO<#kLyu|I)fhi;mtk)J~_+k|uJ%0%?U zbU3v}ali{r+-)-Z?iS!LfJtnEP&}>xi7>a?*5qcC<-G6z~4_9ch+F!E20=Sx$QrrZ6-rTsoi$ zXsoPA@mwO;aZ*tQv0-A@1aFwa{HV8Szx+Di$?548e>^7MM086?JRqT`xl zz9i{Rh*-~3x7Xplp?^ta?+}U2)aH%D*4sB%kai{s&hMbTLm*caU8GX=`ZFJgBsbrc z2eoH%oQ>NOczr%qH5#mmd%8O!_G#Kp`-hkxOfsX>jC^>i!&Wz)JEb3xUOtCQOBlPM z@=0mo$;j7Sh4Dz@R2B4&F*#v3rnQiuSx;o%i6Y`Wa$h@a#XY8&z>6KcwWQOeu zhbzE$B1*gScf%7*UDTQMSekPVcJ|~@PvBNL9Yq?k50BThT&dC+3Nl}MU^Sg!a`RF= z9g_ag&hWRV?}(~~vp=F&+#IWS!EcjFS!U#pfnH%<$#LOwxW~plNrQn@XSoZR7uRR4 z0X4hr)A}Jv@Z4-!K2TtkMAE_WkDEDDLqju+KIHBfqXT8ZCf|1hyE(1>WgACh_|$GO z`rT+#2MM$|(%p1baFW0JSzNi*+jrHtl39b(DO3t0HR0&C%%l3YkWmi{rX%P4+%sHq#%JtiL#3gt|PyQW= z=W2^$!mk%z^Sb}DNmJH!57RN{viBT3$QzJ!@Z|bPWp|d_TO(C{7u4hFfq^{me!UU< zn8okaZ%2F0ZAStW{pKQ2s%`kYK5aas!p;Gjl-<<}vf(-qf78>HhxGveLBVG~}~@_!0*|)5kyV z7FRap6bmxS*6Twfo~lh_NT_^q$x})yqV?Gxp3yP1%V~5eKFo!F8?CJzz@!05Q`3YO zP!*#)v30)YLRa$L1FAJ^u*<3<08z|k8uMzb!W!kN44?%C|F0OQrla^k*? z%{EKDT~+xXB%2lKC#?s@L9Ue`FO$ReN4U)lx&JnbbRbDJa$GZJ5tV)Z=BN1sW7OIs zqODr)gYJHh@PCT07_CVG-THOF`2z_xX_Y2?!EYrybP!MuJ5v&oc6OUhMUvGho`9 z72QRVu?>9d)K8tzOCVB*8~oJC+&6CYf}ODjyA;}~G<$$8m4Qj-a1LkJpx@iK5mpV~CYQDm3Nm?2i6059^W4JNH;;mQi$|)&0*O?);SyHSX6^X0cpRVcf3f0l6x zX3IOtZ?vado(A2nwS^^wC+MAZmo=&*_|YZg_{9&p&B-uhcMPRdGCbootG!);GJzbe zm~bSzZ7*q<#?=@(XaWfQ#--6!YV>kVt$@<}`dk<`^6!fS-0If<4tD0}LzT z!AZ*AXUokbvLX`;)*=(#_7mWt{{OIzi95)*vW(J@)%ocGR%KNpza=Dqi2#`O%eTLb z(D@~nKZ)gb)A~0VxU+w49`zsCk6XW!!LQg)8CTsAOLdSc|C9~(&Rnj0_zR>0c~5Te zZz;|TJd+&KRRRFNU!LKHZT)Rcg}xNF{YUO);%;YK_wd)0^uN9jfkl4(o__1!e-$jt z|CxE|8fBe(acn(1G#zXvn?O%%gPYqlqD3?T#OZ$JI`V9{u;F8*U|G#*c;`C`l${&9 zS0ly>_qDUx`9LOa3?kA46+^{^Emu!%N`hWtxPmUzhT0$;(AUqvOo(j5KljO=9R#f$ zrlewZDidi3&^YwK1_mX>0ySDM#?+6x)DDhP!ej2U1X%BD%|cCS-evXbwZe9O@?R(J ztXl`bJeC_SbvwpC1p4Mgt>2l&TK@#JNmG3#&PgMWwW(yHXclMW5k<;em1;xqi;!WK z3FAZma5v(WQoFAgWNr zPuFn}N#FIu{2+S{>7b!N1hb*QV`l7b;xW(wYG(85iv~?MLzXlp3;aivbzF(_MAW$B z;Rxml=1A!veX07tERR!vvIeiSE~K|W>GYiNW3kI`Z)FV(f{(q_+YHQjsS;l?y;j%? z`LpWB#8}jw5@05g-5dLYtxC^mh6CtQas~u;x-6QUBx^ z@FkVURFB``+{ncBqzjTz+q~&h=LFVsz8<*8EO3S(#ag!q=FXUQzP6}UB?SA2*dWTx z;egxth#x1i#VybzChPIJoJuX_s8fFi4_AU*vi#!`<2H`RU7?j9!`@M7IK@ldkk*E5Xo$Qq2Z# zTgwY&qkSPdEaYIS+f1KKppqaclgstCeFblk3pWiqS+Zt8=KZAs&*`jkN+WQS^;9e@ zpoQ*j3AV)@b+=Z30OCSl%Bz|TU!2rn6o;V)4`f{?TLiC!r9pdH4$4No3vCC1 zp0%{*oUS<1DMZ5;@4Jpg=&%e0lL!-O(2&}pb5(5d^c|TXR*7{s-r*e^4O|R<;k4<$m)w?+0gTW^MmjMxbZmnSS$wcjMgpbbiCYavPZB;MD$#C>F|n zdG))09>c0Y)qFr(E2%a_9WoEwH+E3uu)b1dn+FvLAcpsgZ7fL`Vf2{q6E)12 zrWY+RqBJlJ zNa|Bd17sYG3(#_`Tyo$-p1|h4{1n^ z0|EWO{|rn~p$D!U1cz#E2t)=&FiP(Xse9+ZS^TT?fU3ys2hyN6=3FS<1sGv6GPXyQokIfH0 z$u6IYf$Z%0d8{g<*J1QnC)SyDTl%=RpYPvafi0?e(=!>PiUe)dP%S0J;Yn}NFpi_w z@?h5Hwpw}WE9V87MNhQ5Vad&_Nn^}!JaVqkPADnzBlR%n&xtV}HEN+ej~r|~5|HvX zN#2|~?3YtEpSsuzIb|NpbPc)3lw9EGp=Fw060~F8x^pJr*P#OQy`ERdiZ$dL$@ZFQ z*B8dt;7chwWLwV(ud5T8f?!+P{U(#?tGayS`>u0M-3 z3r3Q>IA@x+^F;xg*X1g7(g+FaCsr$G#l_sEMGecPxn2AILeX-Es`NiT&p^ZIpib1} zkkZ76*>bQD;fV6y0p3Kxnr9`>ZTG^}#);2);agO{R17@e3!5>njm_~}w@YE9t`8TV`?^}RTG~GXN znn;^@=kh_URw~1#73pVRpD#BVK?$LZ7ZFEqsWR<{7&cJn{w>`vBe?4c+F%MAu*R-< z*OG@#0^te|0`(xlHoO2^wYEBd&6ymf*#X1ex{6LvIY6S2wm`ebN0>To#gDp&h1{l8 zK40$5x3*q%d8Ap;##jrCug&+-qMbO|jinW>2-_=YLcl0m+^ z;_XxYLL$+Ac(7H6V`6s~w>@*0?pECouah3Re(H8A@fr3^${OYvZzLyvbUL89DnIM& z)N_orc^}Qlj_`VMKy^1hs0#Cc6t<+E(x|H-#g_405pL-3h!o8n) z2nWTzcM8?PC-#YwuIPimMr0H)4K;J zh80A|_2UOSH;9|BBrY@3r3dpMz7}ajFD?k(yu}|de*dMkgf=vZ@^F+UmnV#l0m9@N zqu}M1G&aw1sygJN@3krdAlCypnnP8dn}_Ub@vb(Q-lM<^vt{0Imh&^>OFD16K|EE; z`*Se0b3usXR?Zd(FK*$xpp}F(iF~w4QIv7dFk8=0?`>E5lHgE`i*gx0ls3Us*FE_1dVHKi z*@%?&ikiTgcXtyfr9+rHp^7150^Go#nKOCfp9jrcuP8$T3K zC~OAfOl{$6VP|g#Zq;;$`%xmCnCH@w(LpLkfJ$@)ktYA?cs}ZjQ14XnE^YPgl%epF zwz6{Nr2|T3(!2tKS6So&6FgUYyhZ$c7heaEZezAd(FM7k->Z{W%@gK1DhUqZ7}3dK z{?og1NRG25N^zZYr5~Ius0qd{>Z?`CerG@*d@0ESP4t~D&w)ajun%c08b@L?tB?EC zI!}cv4JZDU?=*e}1e1xc0=m!*)BWeWqz29!n0WcD&OfLf?K(EomEL^>S`cxG1a;vbw>2MDARdsS zP$^qPCI4FeV9PP|B0X&9tlWJ3&gL0tY5diHEH2CSXKvibFTLy#9*^stcC@)jOfGK( z%icUetk2{INp}B(W&7eHXh;XcwdYQAiFSsPMN>VLR{C%Iola5>cUUwn1jy>5?|`;n zRP#g;u`gQrpvruun@<&asH3YtBF)3`-m*1JX)YTo50_po_Wq)zwCd4e!&d!S{fosJ zWo+OkC6zq_g~1b*;*;P=+xKt3G<@620)l3kDWf_CiGU}lnxWHTMc43)5Tt*sR-Ejy zm{K=<4ALtl=PC}90(FKFwk|UC&(565gK6_9Qo2r322UoE-d8^P3#K$SHCR++km}@<4u;@U-@@; zJ4)2JxP?L01Sx}d`&Q@l68qfgeJEl>W=>;VOl_RY8CgKlF%N3vBjhTQ+UB62R|B_S z>mOc7Gf&+nX(9Lc8aFW|K|>Zf!=yGsb+p-`ZBkC4P#aycxEOw$NkP6jIQJBhM}gDF zA$ar7E@M9n+!LK;r6z%t6L{j}~8Ns8-t^U%n>Wj({T{5Fv82z)2A( zFc`{Gq2JxYaLz;$0}T0D?!BSVKj4wqK2StiT`mt_IYvcP!A!JyW?ZsvvsL?m)0awK z-|qAZC9We!5*)LSJ1shqTtn@Ha;ERHnP4VeL(OQlDDI9MX7Z%PqbXW*(yA^kj$g>> zRuj^J(x$I5DlRN$9`pZ?t+$SA`u*O=Ef8rHq@*lL>5>jX1u0>GGy>AyF%S??K|m2u zI;2KSkxm7qy9SKz8ly&yjo&?A-_JXKpU>+*Up%%wihB(&)BiZm`^NOk2ej$bQDxh; z{_JavMa1kqJk7UoKe`Qj(o%oc9PaB%)ph`iA`&?z>RZutXsaK_I}hJs`s!!%s$q|J zz9I*KtjBl|g;w2Yj(uGJi)s=sW}kNR9^kYkMd|31W??GrUvDa+KTz)V#`^_auus$F06RZsEn_h-(In|O7pW8k73~1phCQ3@tS^qK6vKr2hB|%d|0AtdLs}Ee^d}EDj|6` zxc$)K^oSm2q2sssqiruR>o(>W;HS`XDmpMV8pNo#$1_I7W*87}Y^ z$<@X&G)VY!dKX@@-Ijn`VELA4M1dH}{FqokJ+qs(dc{!yI9lIG)`X-@D`As_a{z&v^8+;q?yAM>sY z$IksW>~o#?cc`pVHCK|f2Nah3s~8i?u$B|tWlHX_7+9cQC2%9Fi=mwfvN+1LLu+5P ze9x~}bR8Y{XTcbB6*m+efhST#3`lwn)XnpPCZF=?00>n1&dv-$AOQHZDqi~oZKW~& z4D4ewY3{wXQhGoOyhIeOeq|&85&0({_&&TJ0-%{{+j~BAXJUL%!8O+J3lg`@27p;% z9iHG{$&xK-bhRyt>fjIRCF~}T2e8qI85E0rcK8l{^1z z)5uQtU%@ULI*Xm$q)81L$fWPex>#{keFPY0U8zw!JeXA-`03^_ptXSnslwl;Irh3~ zZ@MX|Q44&SPUjRRDI+$0T8LC(E(Vb0s@#WT0SARQ?45rk^cH8E8^FY@IyVvC9%Nus zo_+PcGgqYwHUj{;^|daCjFv*uNnVqE7NVlvg|zi2(Y7tZ`1 zyA4d`d(1<}yv82iP9jgFF^`_F{(M>RrGDx~#%cIGbh#OSF-Zyz_gdGHNH?cAG@Q)) zYY(HdDcgHI)OXLxjXoWqchjixLJww;#lugElps;_$4YWV`VW?aFP~|W3aL%;f4&0J zOF@uc`db6A9l(N_B8m)uVq35To#b&K3b$8E&zpT@^;8wihXEUk=SuI~$}b6WE4^!( z_q`Ex9?}E!aa7u()Mt-st^3j-2uEI{tLE=^x7(H-vd+HR{O9-@zL&4;P~c z0&&>H3nV){dv>WYJN8)mTiM`n9btM8)IMs@exGif* zcR}ct?yJlY5h>=Y3B>LI376DD!_SeV;t`Jduw9VC?H#^a%F5~h z@ixV@k3Na>kw}^k`r&521+e-Qg6Zn}0+_`u?CY5GWMWqWp9ZLGPfrdL4V}9neB6Kh zdta;=_gezqLe2HQz5)vLi>sUkbt`S4n(4cHfNP65OG)nlG30NFnVEDe;Q0efEhoU1 z(CWzcV_Sf)AOQ786)pegyJ|P<(Dr?@`{&Qeb#awV{RPuOi@e_;GYTx7R^fo7NUN*d zoUA;cd9A!z36is8G(eo=6VAXG1i|%uBg*lJ?Pg;%r_)bBHo3PUN>}n_YJ0As%6JJ5 zG|A1n2u)Ro#?FfjAceOdbk*B@6GPb=59vhHMT~M13j*N@gY|$IQ61h zA~M^KWnYWelU$bT06rh_k|~iZwReM~0DY@`o%#yg+dum$TedijLRnC`Xk-CBN%az< za!C{%u`Tc|G4J-t{7=uxHCkeM!5Der3A3us`L`fJZxcUwmzS2tMH=tJwYAog-l-zx`4#i>i)ZtT?)lVXjh z6?uO`MgN6T#sRl|2Pd zPartb>8D0CljX z7h-}=+~X{YY{erQa3^``j`Y)a!-K^^JvjI@FYRNY?xP4-fzLmxfjCdEbfew&ZSFUO zm{~g?E8rTPZ1)ox;k^MAQkYE=M6Dk$J6+hd8Wi>FPI zL1_^mY_{<4{J)Le03(y%4w$;!KSXjP}- zxZWn6OlldRe?z%Iv_<7DwX$ofeCQeD8wO#^CoH<0%Z)SGP#Ag&yvk9o&AL!IwF7_b(WaoMCBskpziIe1#%-v zbSt_=GS3U>OY%iM(bM(omzRnG1wZ+f^O4f!;fsIq1U=An>5Rx*-StDfD&%~`;o$K| zdBtY4a>(9TWrXfk4-D_iBXg$AVc`oxtYBmt7RP$_8OZ#9lsxRpqi4LMa0M=s(g!T> z8g|)LVH86GKj(y&{?;(;5+BeFXz#VuE%dTn7oBOq*VYf|nl&W}l$LMi;p=_Xvw0xx zO0IbE%$O{o_F$qe^hy5)2YSxeP zo{pa}>;$HN?Q0PfD?!T-S61e}!lw3ijQ;Eio@U4F?h#C_OlR=PrvWW4>_T1K{klI3 zw$0e!q2)vr4?S|z3;c;f%K2w`Wr%$6Og?p zq^IIj=a=(lb#rra2_w}a?<(F62M?0DEq(>lYUrGJfwSH}sNa%4X7w#^#bba(5h(G; zb$oyg3t0c&~o;5?hDN)*HcwC5(WkIX&!7+6sM^_58P|YH1&jyx$aM4Ju8_x<*x@)y1u# zcjF67Y8yJdWz1chs{sz|dz=QM_oU0-gNH}%_%OPqe!V5AZqe|yx$mGY!dM{;(gToVUX z0f2!-7%-$)0$KD2Ic57W0R@3O1I31w!a%Qff8a6e5>VlbY%CzX$C)bnD&CpI8IVpXxI`qVjRELx$K z1A<9z=d-2*pnh@A3`=+rF2G=5QN8uc2~<3TCWJ&l$_kLN+^qc^4p*;hETh0+gQ8c9 zZ;Hpn*JH%=AN46Pn6^wgOrk*(A;JY_zeEsIUshPX76ohp_?oj9W&f@(tFhW2YIEzt zwVz_-iT%%mqiWH$LC`idXu|x7Ok8y5^=fjuAG-r;l*E#|ff&=tdnF>_9d!WYRRs-e z>Ku?32Kg?HJkMZnVjaLMU2R=GXf@11aE43i8!7^mP-%&%1^vdl<(k2V%CDCQa9m|G zFl|gKENrRMH+cwYP-d2F)FmwgqY)D^aSay;Tn291KiGM|FjFPSohcr_Ihu9^4Bert zZ8y_ne4U#2BmnwaSEL_yYpc&2G+kG&!GoUdc)x*DPHHk|f!r@%pHj*LxASr}i~iF$ zpP5JWx4GADo-W+p-7#_Yd37tl4D6*|D+I`49y39m$933l16UpNuuqRsymXp?MpaSvMGek0(8}xm~VZX@O8Kk z-uq%?yhvvo4Q=~B1Bc&nc3^p3A^{E)M|FLxvr*=JS?-wZtE_ndY27-rM$3d|*X-8M z6t3$|)tE0&8uq#OMscJZ9)4i;Uru?K28Q|7#aHCqnpz(s>G?b89b823(K|ReShx`< zr-a7FQa@3+5bTzq*&BR@n)&24WcQSlZ)iPgUO50xyT6>EENMj))Dz*sF?3S zGo|)w5mV9@7eMk)#ol9iVg>AbjVg_63_OJ=Hj`tgUan>TP{yV09v9a* zYUAi>94=mNIH6eIHV38AcBsduXhD|5@(4ZDV^4ImjKZ#_plG20UceYkH@a%irkLix zW96Mfj2&HrT+^cdgX{wyeN|@d)AiU|2Z4q@L4J>&jyxfxg6LUIMVITf{YE^cC1*Zn z*qyH`{US7hHBhR#{?O*-C&$UUKUYpP0>lY z2UWd?l5N=WE990{7-?*sVLEI*0EN7V>m7Os|K2K%+8FYup&UFOMas24ohko$Si6d4 z74>|<-aZ6Wc1n!>HxrO(w>D_|cEnVJT{1?}b5${>D*dtAi&Be~5=skGFdy~_JMQPx zvVdnJ14qPbygXYr8IHMqyXh+3t*8O7A0rQAGj;G_^lEA&De3kXJqm>(&r7njS=8Z` z#Y^(;ZeFzQ)TlUc_%#HT`z%_ehrL(vB+h4f_x_(S0?d|Q`1l4zHppFf`|D&qd zMxC7U>D-&ZE0%AYXp>r-$(UD{GwAX)*%ZaRJuk|9{~YM(n5yAV)>RI*O&G<#B)TneD;dWs;QH&!c)NAR!pa^a(Nog(#4Z%hz)o%+fKMOW7? zalc@4=E&Iv*h0ygL7fF0w;#fdm~_dO#+^pu^Nu|y6W~Tm33Eh(4QJ-rL8cy?{!Hz$ zxXW3NRcPsHfdg2QPn$-Y{~7)w-jV8VY^LVBX@87o=;?VFj5IzJ}c- zXTvjd4Jn*MA~p&y)_o|nNW^i%*sfXin+)6`b&|~E%nPZ_1`Tu>gU4B&LO+)SMs1<0 zi4VNb6R9CG(HL0uDlRl&gWjXdF*W!8NBx<;K9OnhITMLf$43okG7wohCM$?yZno+W5knvTcp)Q2Y2o&&gG6wPr>a zS@63+br$-G8x=_c?p}Eom%&S56V^KO?QrHAqW{)-k%60e9g?z&<`4IvZFivy?&!e3 zmV5g2X`bV5!wpte^vt~;zH0Mk;=>a_L##E{b79LaD%wnm=Q^1;3MMx;^9#PN7Juy- z-S=sK+OlmA+LHd8$tNt%NeKW&^wv)}WKp}EouTTo$6IpoA_Hl;2hjGyjbEhW`PXvr z2z^e=H3+mjgmb@JZHO`f{nUefcRsHw_N}kG2Ocy2>0a@CA@GknrKWceX#?q{Hrbg@ zW}8J{F2;!l&RceMcdomgDs&w!Bhiy;21Mlhck0r11MxX}lW}t}H-VRn+Qz!6;?&HJ zeDxBYEWVpp^lpf+qb<$$Wz~c}<5W}q6(}8(iE+H_lRUsWRJUQlJRQi6dVXh1-b7cezv>WOaF~1;r zb73!)Y$a>M4sy5u!(vOz9q#~L_=}Z<73NC!a=phNcF@VW_We2(HFI?i$M5&1?>s*H zvfJz7o%X=Ep|n}}k&_WD<5#NCNl|ppH;A*cKGl!7PTCR^qf(4hdS}ITq%kU}hsnd#f`fjhMtlPY>+ce!saeGUoRF z6HE=v*Mozd{WmnGqC@r2o3=#1-me^Kt>&J^ed9MMoa_H#czxVzj*W@b5_af#O79Q%$UHXThm|U`%L{}lzJ`6I|R7-iCKC1QVSF7+@6XgCp zn{_Y{rGW}hb}k1rpH69U&&&eG*0&!u#(OGfY_SLEN@bQo!dwFH+x!{)dgaY9wo8I1 zF)$~G$*P??N9+OW=6$x1zf194zo~BV3w!P9G$#%yf}!3w#Poa=mCQ+t%&BF#h_J8= zo2wQNH*W>Q&sWG(_znG$7=Fj6=2{=2QWm+;N}GE8#e_6FmOUPXu8MbMIgIKmWvcO!1hV z1QhBd|C((C-$w2M#z5r${o`|+F0V#jJ6ZR-P@0)(3(V~%-Qy{;DZ5}Q3DL2(4J%Jb zGPS?*Vq|OO&Ruxxp#gH^c!^>Qt5TR;R1n6>Mz}s}aCyN|>R>s_*$}{-b8IQNL#E59 zp!z_wUlpydax4D)(w-T!Q;PrKCDqtU`3*8*jdVlx#IK149zSK0;2GG$E#^<11cu-O z9m=o|tIZASrWX9w;4vnjmK)I7A6`^QS9gDMFMQBgH(5m)T+RFgf-e30-OHrLN+0S< zi33Z(_q;y;o=f$=9tdC$ibY>NrfGRA4Y5TR^?YZ_YeOz&^L6(P#m8AV4rdALJ-of* z5n8@$bdzu)8Ngcc!oMH$o1e3Ct>4F=ib{nbt+KCJn$2>;9V>=U8Iv1$n@iH%+Id-- zaoLwr@xu6w`je?Ekq;rZcOhz{3W5Yn9(rmLOAmrD;2R+N0{-3f!iDYj;6>7ljzI$a zBVrX39ZeG0fX5|MyCzM(JTHwTV37!un62HjK1^B(_p5yFwz+QZLPrOISDy$UzhyY= z48HI`*2f&UyEgO>qrD$Q6xVEHEdyHzsQcJV%m+#>bL z&vlFngv?a#5RT z;SkfUR&mtLZSTbXnXvsHeFFRZq;hnkGt7(azt;^thk}o`RTN$0!|AojhB4&Dbe8fr zMcmw0f3N$_x@-`JN76l9J@EhcITA~@0PuXq$8wByTIoHnhVv_X8gngJnm@?=`@|F; z;g)C4F8|j~d?R#$;gTSA8d%H@_+F{nmoJ~u4TU{2^$3pIs<7=`_uwgU5WMyGzK|bDVzz zZ|2|V``SgBEtKgO37?gS+QD|p9V~7vQZ`^9*{iwCN3{oDo}k-Fhy>^6xa}|(|L?n( zF@WZ6G(m|DDxdziO53_Uv5)zx3qf>WrU0LNQ;=V-82HNiOVs!OwZGW{EaQif_31au zyB=SDc3#jfXpERYq!QQJ@x$v*;d}1C&+W^51If*mzXAP!qA7JRi6wQroR6kwRE*dw zr0Knu6i~)8e^+g;x}~g##7}s06qVJyqati-62%e)VhMu8!hhVa9C1@9%wlqP+PpTv z!{d><-h4UXQC%M7E~*au=Vz;cWHix#FD$S?q`JpGI|fm}qIViiV&1qYdte(<9N~QG zKX)fy$#{UDW9xoY%zEGs)|ZFp|5{(O^LpS-Qw1d1B*w^44~z)Zc`YsKcX5`8DZ7G2 z{d)Uac{nLUD%g$xy(Q@yo4<~DVGJWe0C8NOY$6YRU8OiUU4!s@MURRZPTZpvsQ>#~ ztL7pjXNn#?JV<=KroziyE<1mL;Wu`F6sE2FqOI&M_@tcpI|HFNT&n*&0?;YTdxoOg zIe&D9_bU)dJZAlx?9j-{B+Zt~voiq#35edo*Fg{Q9La61jc`r&+XUs(|93R~vmso} z7a=1^>j*)I5;)fvt*NoWO8ed{iP~3oMw({q*P0np54jF>;>*i+U6{f5Q1fPuQh~`3 z5YYaKmS$`=PEHkyt~(W;NUyH;_Sx6_pSVTRWS|qRQORatPx_G-z^JNc1M}kDzBV zH2md{7j^%TOjlMT-HUTG!{V;c)ukt+HXqh1Sa8sJx4vF#c;}s}LhqGWC9(exS z`~{j0?2Y#|H4?~b`$|ITEb2HtIA{Oh!GjW7xmQl;6sfSKHL#i;1TvEUTuM1nD`N!ma{+DSLj4 z-&+A7Zf(9!DrvDF8nrU8G8W>v&YS=KJ*i=tWxy62G^e$HA()gPCD1&!m!ZxNKoDwGkg5U-1eh z;Xl7-ZOpX3F*76P^M!%||6Lk48F_%IoEcvmvU0>i^sEaq9tKl96< zy%+@E)&H!nUviodV@HsI=HXE3XQkurD`92#!z5XlkG}yZiyG^+Gx&Whh$1RbctM-) zEj_OF$fRD+Ji7VHsr&9ZX2cQC>*k4&(o*-n7jU`N$L8lutC=I>SX1pt=eCzD& zw35g8Xersb(Y>ob&V#Jscoe^T7*wOekSSGt< z;+N7p|B61pYg!7Styk}Nq<=$Cn} ztpbIdk}FR8rfo&8_128tzrbWkJC%`>7A^jKz(}k4?aEQZOWx zp< zaz)5+(r?*?|2U`L(lByLQA7J>@Qd31_?$%4KwJ9vL{tB%1YSdPBBHP}~XJ*4or(e7PGpDLu6uc3aq{W)92huwJ(2 z)h2yd4Wmu+dxR{4-3SW{Tan6@5Ooh|+vsaJET}7u4m;3rcd^gVmH#j{R;7*}`fwi7 zp?RdB(c#;6mzQ_iv6*cKQG?{ms3ZoohM!-eOs0hM2K*+L_jjP_ZB`Tb%^|#5>+IKe zUdO9HS2lC5-uzXykBR;r$CUdAJSSSgrjJdzJh~-bxQs7jihxH`vteB?VHX7=5m||I zxdp-r(KztI5kiu@xS<*yaPq=5ee*7&PesP7sLM)EM`vug80m20&pPJL*D#+TwpLtF z(j;eMg23HYV^93}(cCoPIDR7=yriM6S5*~Hk=`Y8%2rl6j?<+h5DD#|{XcppKVP`Z zxZ_@7I4)wWZRES6BV`>3P<65PM&Kci$}!S;MkAefFUV9^w*q?9K`!>;vG+!Idv_V3 zL|qQVhuyu)fyE{~d{|95<@_-ifrFNzzHt!HCiWUEPCD{_Znh7!>}@3hLB)tXVUN`*zS=aglL_Bd6pz=Cp7(MB_*`N50`6* z-qt|n^_>}-o|&2+!u!0tDUVOOYL@%UUnk^k?iLsbR#p@@*~)1Ad!?t_6%69Zvb? z-N^q|Hq30wtZKZlxV&7A7^zGLbeSy)+Yo2mGcA&HBzju*5rS4v`Tr-udHp2g_n@z7 zfUWlw_BC$fgvpp%F7yp|d7)rk446*vXM7+Yo7s7lk-lLs4U1w0;%mfDFb%Sm8h>27 zn7wO%+f#7rLLk*^qQY_on!ybk8`bL*8>oufE!VE0U*FSi4PCcF%y<+UHXha+R8*ic zAG7}a_U-4z4o0&;4}gWPCLoReTyZXWRj=38|JwhRZbDG<)U3Nx37N!X05a97vCKhw z$_tf|yqR`6e2i|d(G6L&u-loBm=fLCnD$zoAwVdVVx&?dBO;?L-sRB0 ze{}u@bF+PQ9@CK$bGS~=Y>u9(H?)qd+sT8k{mWOj%5vWsGJN; zG#;N$)Jfl1-yp93_2$Brw1fm*NzlUu{3V^r$wIjOM*5nxA7GEcGBT1+nNO-Qm=ZLm zKRc!tAMfN-2+t2pI0>X8_wQ$+KQYUbQ#O

    YDcRPclrec(JiJ?R^$IC+dD+UDCL^ z2A2z0ASS(*{q_pKf!U}1Bffe=b1P^TyF+5zw-_jV+rPU~o zIl4l3?a34IdA#)MGV(M%vcedfF9XY+7_0X0dwAG`O~D)!5f@35vT|}}dB!mADM!el z^h|>+yn)h4V|sDlVqXw80vq5|zvo0}lfRFN8yZxfi0hMDLI3ne=G&7vS|}^$8oqLU z`mQ!@Kwz)Ly6o8l4a6mh%e0%CvlF$^FB<1ASKfqFjg?!+NYZ(vO$T68onA~0>}a1; zf(EGRW%a+ibMNjr%6n!bWRtYUL#gf2`5UN+3;aZ=^Ug#Hx+gLmXwN3Zxz{eV?+b;aK8Uxde2^cMXz+d=N`%%EPa5Ra>HwF_il*&g2`)9xY?0@ zVik2Pb#63VY+`1-1j)H9eVnsS?Z>U9k?9+XBGn53pi~bopusSgIEs~Vr^-K_xIV8= zXo}Een>qVFFzts;J&0`|5%n}M{Q6S*VLrzeewgLlS3EwV!gcSJVdI?t;kR#xxCIXN z12jCXReOm7+BhLRbHF*djcE!y$Z+-tDb|C8`-WQwu9ykW145p#kO4PFOR7)2qTr=V z$yNv_ZpL65k4IU>F>g%}aS2HckLKcL^Y}4nka-zc*M!I0Vc2lg$%zOWwU_AiQv=7eRD)q%i}Ce*zZ8T9iDfXUB|up(1>bo?#P7u<%5II5tSx&9$%4kP+n5Me8%?okV40a zS~PYh@BIx@^D6t0*XQ2kUK{pW9fOfZG~KQ}*hC!&4W2aMUuF8kORQ<0?UV?C=Eek| ztLr|0KC$Gl@hzvCzxqH6qLDOl@s>%uVwVmPXa#%kMqWBcosgwVg=fe~BVeol6s>P+ zOLyIBHyVCBh9(|PI$wDlU4z&!OHVV=9yf&=hlU0RlS?~_JIyCM@$WRSHyh=U)@y{; z!A`yhot-~qp=fdj#gB1hK}#9&JCR>SRt1-Sj%IR62p<*yenydCatyb39N&%C)zR5h zV04-l=V+r?w39?L2~IkO5*LZg!Vvp+z7jiN>L-P@C%?(!X4KHExl`v=NMS$4xp=oxKbC++fiqG6J$Y}XP?-NA$~ie{ z(G*d=JGim{LHC>0h=~_nn_)pA-1+JjD>O82UBH=xRB`m%B64DZNZ6pU(yGI^p7LdP zWY_np7diMf70Z@MOMt1Yo_}c7k;-~c-_`1RV}5h2ceV+R0rE_{{2bggF)%LZq=bdoUZ(MYmEo7|(buir}T}?&W7WJIGS{dtdG<2O~ow)h} z11S%uM`nY*p-9L>Cux}556d;TO6GM%m?>;FoLj)g5X2OU1`H@En@gBIev++o*L^hA zoUJewTYy}d>)~>)g}J~4Gi?>`41=A}WpDruji|u|STx**V7IdR zG;u(s!F%fPH8zEIOBX|>=9|^0ip+4SEMcIj>aWh(yTuOBhJ5FW>p??_jQ$cHnGLtGytuC3E`;U>KVcSzFIf#hp zA+i+Y$F~O$>10#E)-oXyIIv7K@U&3PTdF%SHptua$|z$Ox?Wrb`xwVXPNdteN&9+t zcb5|PfUgXOPdQV z{?@#>z=U1~L@1=|DE zOvA;2gFgEb_Yo|i9yt?ntZkxM&(?sh7@uu8vFgEhBAoWmB-DdxCN(h>j*YM*uF324 z?V3upddeTXE8N7H=&HO&bTjSVvorc2CF4~$R zV0zMoSpf;>z}0cwM^~)`9)3Yxy{{|<@2TUKvSnX9>@z|8oT(uw;6Doo>5 z&vAEgo^x=NaxwzHpVxFfSnO6v3_Lo0)Uyxg7_UMTJ&^Ikgi1!n@405z@UKeZZUKHr zaV;CNYo!hg9k(Q(>)R@cP>dq{G6>z7*``$>4`w7!-dAPCfudR~^Yp1d&#=N9Gkp{_G=EQCp~9<6=J&q{`=gP3zD4fy-`_skC;I%>z| zv%=p~PVdvND;AcfTp)o>PzQEorFT76DN1vf}=ia)WE$fouOAOSC3#CTcf7W z*I}Ee$#P`Fu~%lHM3y}Xv=J7^`DI~^H#XBMpY(l&ntVzLNF{r1W)67n6x*>UDGTt& zx?+tFYuxkljLt1*J~le1JYrCjz5Q6O0G{Kt8<|aarTLGORfu9XaK9rDza|ipes9mFe}DOaI*hJ8Hfh2D`)bbH>AP;| zPsZ+Hwmf1L+WV`HY>FGQNne0t!c9(V8NYj9cuAPbtBftd~TwGgNTDF~pg$8sDlY|IG(Z&X&S?cz(;TDweM8*P-b~x}B z$caC3?$g>|xksSIX#cu`=QIulB%~H^VO9=+b7ktOsPW{bnZ#$k>9y7^x}C2Tc%NYA zPCjq35JNw4E1rMQ*|L#nA)|W4y00;niO<6qV ztn(3^I6lEq-xe*r&=(b_?AIl^t8Px#;hkoC_B1at1Z$hyR7(W@-Wd6OAC{Yo_61<9 zaeh-Ww??K3(GvorBEegH?kz>ll2`4*7S;h5j>c_h@!vkyf1f;|1l{t7ribP>Yo}x0 z^T2mC=G@pXFl5ZsLK-nIEr7)gQjGWSb=QzUR2l!;Om1&~2AV)rAf#Lup<$3W3r%>~?HUN9i zX4`lvkG7`YZ{}>)f@$aIIvS1+_OqCQV={UdRcF08U6> zguK(v#&n6|+ehV&!dXRY`rjoK{x+W?@2{Jhf>Qd$MCTO7x=vBxONZ!v{7x2eu+IUJ zy<9)L-y7IFtV#IDv4zSJ2iBzzOd27T%N<+fCJJGHWq!O z9CB&hc9vW=6%YXPWAc-iVf*|5OEoUhhCWp@&1=YV+Ad7b!GIu;QF7s&u~t~vR}}3H z&=wFIc~_w&HrqTYz)_0ukx4U88k(iWLBs06ow91RvzpcuA0yEOOI$2BQF8 zz2 zlvex;6Q7P$@8I(U81d<*Iq5IxPD!su(|}Y?CJ`R!eRBak{?OA6->Z_gm4~_kX!EGeLMF~y|Aos zT7t%9dMjp@TDc9H+#oRZisczv)TKfeTb6oX5{o6LivgD0ASQC7xM;#!!2yCaTgfhFineZbcP!dB8yL+N z%Iu1n2Ax*ZnN&QQ5g7Iit5Ag+ik-D)|5#u&<&5Ed$OXn)Cey?6J+ zT!r)JW?$YYPvTPjD-YShgU_;-fVyG*z~wijT5ka8--&_5W<|W zyzvXR-(C9}a4V^$k7(JY7|#VB{rKo)rmc}8BIIBVkr^M;`4_Q%2H0) zfDN)~F^cE*vnDx-n8lP3TE;JXQtSz-c3v$q;owb#1eh@^{!O3Au&GWF7&6aL@kfpzh!o6POP5zk?J_ z=Ptb+7O(q+_CvZ+&}BiIq|}GC#qYylk*Kli!XZIWACH3>rm4Wt`r=z}xGALoXzpaJ zA+uLk%f<2pIswIIqbB6Vz!u?5$&i7cM3IW6NcqCF(*^J@Asc+_t*1%l8xw~-WuX3Q zOb)===`A%Q^rot8-B(;WXJqO{{fY`_cL4oK(qSmG%17)$ILzK={W_)^0(EkhlYJq- z@eZXO5EZcgRefl&+E)9X`j!@(;_rAucYxWH=+%W#LDJW&?hW`{L&EZf-4&bii-6*K znka`(3Y=@*4_U-8Dum<7GN{ot4q36`T$j^~343f#2Klp{ds&BC;W!XrRqHzp*?p7*|A09&O>CRXx0XWhgHk~OK zk>Ck_=|!M?CNk#N`9(8{z$DfQNn(S>eC`f_{GSeec9^aisbr1c()S)&-{?y)_R6H4 z39KM`sO+*!=|Zag=31-x7o@H^7>y1lH=>74XpZ;G{;t-F9F$0{{W#5%l4`kU@pBSg zSPeau&YPVunL&4!4KfzmIbiRx~6D-&iUI z7VcFw%x4%n{p4#fnw`$ZFLgrfjab!jylP9XfJ3>_To6?0kCyTOF1-!vihloWs@5u7 zk&=DP?GbB;)n%v}bgDVx`e{6?&VZc!+X@`%0|0GhU-+XTdss)ND!V)O1li!8=bZY% zNmfYc=O%$@I`bU21klIfIo!IC2CzBC{H#~;BK@XQd71cI#`pf*dMtii(^_8j`J>eHW-yebv>u8p_p;@!1-aCA+A6WRe+S-(q@7Zl`Ip} zp*`(|Gb}bukIz;01&RPuhxS-ml6<@QFj=FKPl=|O;@b8ZRz3R1h^x~yLa^;jR6k;u zjMONsdV_dsvzIEB!@!yo#L2=HsLYJ%heziro<37?Ym*Nqg;K&^W$C_nF=Dk>QY|I1 zQbr}wNB7`GC*QSma{BKLgces;Zt^p|{Z})-DV0p`rkWds!-#VsJlSZ3m?CSV(XUiJ z^aDVw2IAf`jt%?o&s^1zvg&KxFJ-EN*qRWr29P?z_QjusNOgTNmDfh9FGP zXBq`i4OT@kZp%Qc zdHMjusc7`jpD(*^^yO}+qp@8u^lJL}N^C3Q%6P?^B&TSe2Ka1_Y!ejCp(z(+L z&hnC1*7f|L#(c4Jqk6ip+-1+2idCVCnCloBj9U4WK{~f!cgo(^4but~IB%mW8tm%H zXn9^2$$UE*S-%cT3ga#^UNOu&`A+(d=Ee;#A7`_25)mkzTlCWOzMg5oAJ|AqOwp1} zW7;m}j|k@NNtFTopr>#*ccH-cQO$%{vGdUKgOZ@4g>P#|EWSmKhMrM0o0>Bs{IQwa zEpOY>J<=Ot^7&NwB6uJC(3X^z*}hoMRF4(E$JgM29`sC)d=$mtNSI)8BS1M-O{YjZhMM&y23EHTSE>N6X zf7-YLZjd6^j;-H6Ni8%#z3*(Wq;V|)_}@`HavHDckq%piHWFJHe|fnY>g1y9{?wnH zSgo9$v1k&|S(uCth$BVl3++W9-@Zs;94pEBd@`5tKcOb;t6%8Fo)p4uU+ZaHfi<-P zQf!snJG$k586RY}Z>xw+5$644QIhYOu~(o(eI)o(lQhm zC5;g_PyawFMCz-w&snHDpq-TiJ;l4?cc~r`N(0`&YAXt_g?*n|hcfT2t(adhdh5fr z##Uv+?gABir!}eQLOm5bWT2SSALi47a7)vILZ5s*%WacYC_Kq4Oh@)T-0@bJrdq=#!x}H?F-m1(KduePiODe zwr1{0b<*j6~BGOT?fC$o?ihzKkbm`I+5)tXrONdGn2!bdeQUw7i z3DP?#LTE~nCLKg-C{jX;A#cOa0*N#4y*tB<6Yt%-d-j~~eCO=hz4BV|+=;{3^^T4N z?M_NNr7{bzzL^R=0v|trA#r7N_OMaZtLojvN)nlr>wVs}#bbvTY!Yn0torGt*!De* z~uI4ReU|RW+ ze7w>Wm)@;CVAf^fh=;jTs$U%mt-!68VGdWM*<6aQ8_?$JrN-vyh+V_9i75}SzixER zxb?cSEb_DohkTR!&jt&x30=!<>9c;?$1auBx>Dsj^&Bkn)?rvi@p^m_9SD@M*Vete zRQYnqdr+o()KZOtqWCK)hTBn`_VC6CRrCE;-IQlySJEHpMLg8KAWKkHlfLUW+wS&c z*FKiRgB2DDNQ1mCsb|E%o7K0wJ1AJWzula;nne}qA$lb9!6$<#15kaw>S4&8{Po+n z(b)ji80*@`&O%qPg5XTG!lV7n{`*^`K}A;tvg^|F39A0I6Eb%^S7wXAD$3VI`b_Jk zNSnC>#j&wQ1Sg&o+28Vg>Iov zho!xbvA(H#LA`yej$+8B7>tom-4LJ2Or-NF-e-VA_>^++d_VVF+jnb^58|Yet(BK! zJAGJ#rlcwtiqD+uHqDheLQHw=r)lB4aa^!HvLC-d3(7-h7IM;8@oW}49s1TLB}f^Q z+e`9qlX5stHhKTRQ$KAIDC*Df$hf!sY~vm?b2LBOFc!LNiYtHP(zI(jJ7fQZX4O2{ z0F4@_ooLr_St7p{`t=5P7_t-{FhI7S7cRmWWR~7qe?|rN^Rr3m@Q3>qwP)8Q+&XIa;<381?F{db{!!pH7rAFLD z$yaE^(5q*8;cR^T%9V0e6f$l2hmoD@s~`1&U(EIv`G&Yc5S zdrFnTO?&)*|3getcM5^3bDM`SSFGCwLeE}FCOB@YX&BJ2WowPO|Y zS#-<F8{pQb?EQvwc`-9LO-f*JIh`jPCfDTK zWzxgjUY83urDqj!W!y_V^lH~Ah+2D8iw1gOoZl^h{S21j zdMDng*xJT#6-2~>%9}TJCz-(J5(N#+%iK4_W6l5Dr!efLJs`KFo2i5*)>~j!GheKN ztgbCMClrb7d-v!l`p%Vvn~FBcoD4d>ky^DYpP$W0`w2KD+t=n??n{96erdXItiO3> z468a&sYfrN7Ap-eaoU=AX>HC?Q?1dN+%}6$h!G2VBzV|{ke$;e4o}N4&;Q9?x8!8} zJmrOUNthwG?8O3NK14FFLsL0(VE*nEiO!1jn*fuk0#b|Q*KXe1c>jT^3_pjP8KFHr z(7|1i(s<@o(D*A>+kD#72@kZ+?4mf*oTMuPRXqD=$IAQlw@UuwsV5d{v}`xR2QOn& zSK}`{yZiBkM@Oofr^NU5dWSorMe#ZPjUKDtv8AJwE7!d4CfH6+!91!WlgBJeCX%wJ z6=zZLgZx>RlWm1`TSfk_d6)%Yp8cDWiAv~{|Fy$>O*xWB3uS&xY4E0ROjREa;X}^3 z+8KG5WIk-`obQeo*BSn5pVW$%)`L>>v0nLpF8l2}5X$8|Q(TgJSXAb_|F0t&pS5HJ(gzxm}JiXn>Qh#J-c9GW9l*-@%8$^-gaoR%*u|68vDrs+wo!VqqERC-!@CX1)E>7_@9 zZM`_kYhp9AzUDaU!Cu}z zZ=o}KLD$GK#aY#+JvN_!+;=1u<~7ukozat$W^qf=(lUD}JJH3#AqNVD(v|O}xatc^ z=eHL)Qar5JYiot1b9T-t^d{BBO3Mf#V;q3V%(__?=!lE1HWeNR%P9hI{K77fLLCT^ z+R`QORAf_rmCBRAmxAcAQYuUm4jGT!iTVp7pMI$KyUs?8Zl+?s^E z!*ccQy?cdpDV>E9WgWU_igZ-(TUBHyFP7ytOT8`^FL*CKFVE~TuQ)+<_8=dBRlnaE1MfF676zd_w#)QeAfgU-M6C?R zLU02whmW9Z*%~LyZG2MM3+1PDy84e^eB^U?Xei?G^j>A|x9>im4W7aH&VYmL)~Z*F zuEArDl`0ElJ8RuNut3}yVuS$Y>o|!oHQTm!_(t`nGDzy{ntSFc*|3#W+NZAN&P>`%Mmf~#gLem?*598ae6 z2l zw(19pd%1*%7>AFWz%7jZah06`K{i8y4>E1;jc_P14Ge2WuWch$X+QkpjZEUg?b=G9 zkvbX{A$-p~+vABD7vHLN$)y?e)B@wwGG|=T$h;l8w?m9sSEKJ!0Apfm#pCuzvBIko z+%I@o4u#(B>(Z(P6-+uByr7KG-o zRU0PVtLalx@^`s<*T&A;s2b%!qq-dJKZZR%dQgS?!pK1Rv@(VK2P(=7$G+c8Oomc~ z9N((0u1{2AUR~eha*3~m5Y&bJ8d7F4)ZVbx<&P^1sce}UXOFoLN~_E0wzUq`=~Pxd z$K|P)rEA@Po3!w}7oPO(RF>J&Xu&g6w$U?yjLBUCmWPpt^6h!+;RQ~9&Z9Ku9=QmD>*ba zf7`h$`LtDk7Mt!WVVf`OCC-UJ!Q?A-HH{Eo%2`Rljc zek=7B4g9{eULplDF_AF5q{GVhpT`W@O{+f;0;gPBrvMw`LFFQ+YB$gN;J!qjo#+|N z($qUxSs3_U6>P}qXHtMD@@6@Y)K;km6*OJKBXL5Hnd7=&Q;faqX^0HdxB1N1KQph6 zto*32-G2LeFpo`Nx|xKIBDapHF542m;*+$^StYBijGP|ADic04^9mYD+YgdR6oU3d zr(;L5Bh83Ol@D1cHkB_`#?eU$xZ41puBfOg;Wk}lLau~fYS|C;QBy;BamQ1LP9NCt zXGo{_&BU}g`#o&vYsy)h*8DYFJ1*jGN@Idrc}3x*@2ooS#K*N~lMVgl86LqPIsgb~ z;|+XoHKT`to6Y@4wcjK4U7&Gd1rK6>9%m1+q|f$MP!hMiGI1nzQSX4GRoZ{r%38Knq!clj#So~J;H)FCnwzd7qwiLJP%By z3wB;qaj?8{bMyM)GHuayb>J47L z@MK|CF-P3l*`ITwNvwV@{8cpR4WYVKVmACxgMX!vsf_=0c zU)|nx#G8fzw2B$ttTYkdsQ!8R+z+SOz6D@UiYT6lUE#$SI&t2s&_7}5aJG_IQRs(y z_AdvG{W3uesuvx^!!iMO3lg`pJ_FkUhZ6v0{JZaLnMPG3cj{jHz>2qAz?5P+2l_?l zl*^=cMCN0Mlar~k-k@St75T0$xV^Nz;Z0)wX8u9xtavtcD^C`QoVP@;QfT)SvU<+z zHLLuq7H{~9rLsV#vqV^H)qaS*+MU}s3%Dsoj!AnBww7BG^$4;GN=X##=F;^le+vC|+CLyI;T@lQ(dT^N~+M6+7MRUTfEs7iAm@O`(E} z*BqOen|Q6>Q61nEchWUWJ33PYx38(}jcW;D@ap$UP_>0t9RK2lt!ru48|q74ox&SD zjqLvP^E4wH@r6uo53sIt_USR2 z&~6fo8MMSz#(*+e9}BKtOmMq3_O+QZZms07pdEX0{?zqC8KfA^u(MeeMmO7 zMbE0GUnMj`|LOd^XR=+-?vfzjAGb!zlX`yY?=EM#WcRT&WYasi%F|iFo45c8>fY-z z5Pe<63NdlQS~B)vAX`gzK?AS4^Q%i^y+xzeD_yaV@FikJr9t!=Mz0Stz+jJcBSf@i zvaQ6M^QQTc9=CGT2cPx(u+hzSbR^NhbpeoNQicU7Aq8i>UG^v*i z(aIOBT@Wua-}bgn8wZiQP&~W=s;nogwpYp|&L8twgJ>1vZ8WU|-hYe$G(f_05t&~) zb#_2nB1BQ&=W}nkIh@6{Fz7urndWP!7B$sdC1!dm)kp))S=-q2_XW%cqf%o8{&MRj zd9Y=YNh{@$PJR$TU7}TRB8VJ_@+m-I<7-c|8cUy`l0aeeU-MYMWdEjc_n#fsiFU6H zddG8VvJkHmKZdlA+O$7w*VT7*=+Elr{600!W$68`esf*7Zn&PfTN`e0tp+Ckiu298 z)DK4w@>-i4JR^)N+4YqdA2y5#x7+tX%}~J|$qS<q@B(h)Vf z@)6ro?TJVrvLbMUPRAuM>wm-XitaBSXzCQbo4Wvj-QtT&<}&NcJQW&!9cE{n z?ptSjJ`l@pX%F=R>$HWbjE0&H5pgIh(Qk|t6p3efSYFLuwJezZ%wIrXSy^wC9fW>8 z7$C=ZcKqv8#8G9QjQsGG&H7M{Yoj;rq#W{F244yM1qBg4a8NbzCwP1~sF4yl!SJS( z!$*QU^WjHV+(G`vMy3y<5sk6Pi^umy5Qh6U4>So3O3UiQo*(B;&z+sBT^VSLmEI>9 z$u$PPNvR0)%HTpYU^i9wY^BU8JEu^P++K}g==1^nQP91MW4Pj zSs^P+DuNVb?D8@mTf%bOp%o_Le>x{7PYBtbT9NAR_;O!{?OcJ(UhS7_AHjJI$?UmC zttZ(g?=&SfD#Vm8YQJ~l1x1hz0oHJXTJ=iUE3T^&qS_2Uartt%MSER99hSZcTRSUN zUYm&2*S!`eR+#@_7ceY#RUPxZlsHrl>>{*q@m4%01`sxMQVm;Qq=|LlwFq6KzI#0N z3`+D=6-_Z7*K9C=DTyho|0Qt}h%4LTk!1wBESQ(oW-{w4;r7gY-PJy=C9UfFr8IV? zxk}TLGfngP0;Z`yC;O5TeRt6vRO7z#I))y0;I&xBoqwf+ez75V+dS5FC zqr+;5>eNSSU9+9e2Zp$InZLd3&3NCV*mjqUtS;4w@eMbr{>a6RRJA6f$k++FGg8}( zGb~{{6Hc^Xd1W|@k)m7~vdh=0EB@4t^2UVKLbd(IHuLl2bB@#t1vr09Yq;zFbzra-CwijF=4jWOxCq0LXxo}Ka3l?PSh%%h`>=MEM zbdRrAS!U8R1A_sZ_hLmEo8^6~f1m;nVm9I-F#nypVef7GDl|x;Y^>tr=A&DbBB|L; z`RMT<$eKh;T?GNPrTO+T09nM1mVK>9H-xP0Qk+9j|7rVsf_22& z@d!f~sB!kVD45tMXAUN%Ms=4?4RocBmw9WN%Cj7-vz69?n67f1DdWwn=`O-}#Qg1^ z=S-AwBok}>HtBU@KfBDIy?#lj?q1--#4D=u!4)~nkN_U$(9g9J`#Lybo>*1(7atunQFb{2RHJgu?Up62qX@z?rs8~@h~A)E=QlN=& zDNY%;S$67T3-$LGJ*(H1iJtG_lVeznsnNHDG%A}vO50lDpQ?*X&+vZ{Z47qjWiCt{9( z;>nRR5#=%($XYLvWkea=PHSvFUc>o|X4SsnG$w}UTa{o}RzXF;%tDLSPoHd%_ZZ!q zcv=H+Vng_v_4?FMXh%+(LCAG|4K{-}qy{`0(xLT_k#T zV);NSNHYL@pFlz9>YSA$hPlrn>r+*B6)T-hi&umhOF&xNJyDhrg=9Zht3@m}^2}Io zb@IgtDoCur%Rl%&tyfiC1HP>G@!9@A2P8e5I~V5AUKW4{{f%QX*H6LHmSpoGEzWO1 zpaECVPuv>{SKR&w8YZ^SC)cZYpht=t;>7sr56cAe_A_X^=N-s=Yg0(@9!&((X7}6- znRD@h{GuOy$R@19+9}@{=VFw-Z^m+~w{7c$i2eE90*wM7zfe1BAP{j%|HGo3?>c7_3J8%}zOG zXH?d1UcA^Pv*rOTgE!q4D3OhglzHGqG0-J1mHYvj(YLH_za@B0A%;%whZ8*O8x6g`N3znS{^Sd;6{smHu2?VnptB}Vh@gjeqB96DZIf_^MjZ@+)ATu zj1hiPZilvJWD=^D2x2tvH{HDuvx}eCD0hfGpMPW zegerTIk(T`g`$mAT~9}7DV$;j6khaoh)vNJgc4`vzQ!v!J3(+R5mv|-Km)cjP=I4? z{_E7K?dgzMLuFv%@kTC%9yOD@UAnJdKXsoybNbb6Z@R-K>?cq;$yoi`02%=HvIFL8dKFz$dw4hJyv0nwrQ4X$^67}xW={QfyKd* zy!ME4ht|QNu!aB-bUoz%!*e&V0`S`H{&2}kYP4*no=Appgr2f)jz{sF zh-P+9&*XXJ^nzUM%)mR&(TwaF$uyblHWvh8xwedb=;=aZ5iA{rN9#?`;;3;+yD0=C zIOKuW80z$tlz%+N^7!yWCfU6dv9GSvB)pA+FBVc63VUx3c~=Y)CUYDYmzb|{4Gp-y zo%3$NvKE?q{m66eXf6Os81MV5nG#4HXKL#aYrKIqFu<*I0U#7L`-LfnI}p44W9Ba3 zbgjUZ*1?%@y}T=3G=vt(=7yGu);3fnPK0&$^Mv(bs!=+?sDD8AFh4#_@in9I&>rR( z@Vwa`W(xah>Jt}*s0B~p?#%G2OWS$PW3fvTgyGnY%TJHbV=Pc**_E5d>NT9%C0=%T zyg>h0@L`JgT1O$^`P-R1QAX$M?;ZHv)W-9cI2tyToi~9>{VKB^P7~iVv|&&&vU0sS zuw`T|u+n=ZhNi1&aJ42qJi!fDfSo{PIJ*IUNo@e7D=h@%d;eUj%J`V33Dy}ra#Rb5? zAy94xj}HT>TkAmU7N7Py$0@4(X;9Ms`&E1>+-udJ{Zr*eoF_(x&@z;aw(8$G>w1Akbq~0 zRh^LLLQ{b0-2ba##+k2{2dQNDUO|9ZL9K;5q6+ijG?pDdhh+7>f_v4p+67-*Fr3!2 zgMZI*i>p&##)jGrn8HU+OV~USQec(x9@8r7#P55GGuuUB-}rmhn}aEt`F=NQ+P#W3 zTxH9uTAQ~W{2||);8jk%z0g zZVumo5jx?z0-i&966Gkps--5ZDWWBVZKZZM#V|A7e=2npqH^+Ai~aYs(5`cm+A_-l zt_XQVT{7Z%3+o_kU@f3S502gR_;F)3tMch{f#f+Ew{q`q>T~JviH{Op-Bp;J3LYUT&pHFRQoRJ!tuO-XSyYB>F^9E*fSlepFU!D8pZY^J%wlb(U^m8*LO-|gGxyn@;Kie{(PZ%m z(j0!-Y4R#Q+3xZ!8T-$(*eIFJ5QirV8PDN$--p&STxXC(MaKpg&1DCKI#=ERfz2ikWlL19Emp!t)XP;i5mfRdJ1a`e*dr;(rr-|p{>fqWS1_M2tQ1o-W z8{IOrDIw8SMQDpbe?U)mhgQIry^~d$01|D~vBsTcu01htTvg${gxan4f`Y!Y7)(EV|^%Owg<2*3;2>1Rz9I3*?Yx*-!#(o_}f(} zuY4C2;8Z`()BXZeNq@$eKmBA!X+YqU_avT**jC_f(|Jgqv;LLN(&NRjs%>Hp1TaVYHj0y@rH-!8DG7I95S6rY2{s`4yroVL?D00f|Q+d6!gFSg3 zuSq+oH#=y8tCR>l@XMWVJ>3(L*BO-U+Ha)-?{d@km z+UF>H(3$a^^EYS{e*H&_p$2*m`l%J?vzxC$82@)HW=D!v%a|@{M=3v z=phdX19!PaQ?~BuiNHookd4b?_LSDWU$kNCHi%x=(3wB(|JP3_ud(%lc~8>5xHqlW z{^3HuUZU4>ko3LmhL?v(Z(}!pl?@w;soC?(cM+ckVmQRes7f+%>-+3BEJ^!_`e`UL z1yTHE|68vHx)s>OFYZU89Qcq!9NcKMj`!#}G3!4~HCH3rzJ#YDo zjcwfq(L)&&oJ{82p)LBR6|JN)&UBU3|8n#{8ngA}Kh&hVyp2?4p)ec6+0y1STswz} z8mvIB!0bq+rk8CaLL#*>k<7;=W&kxNv!7Q3eWX*ZZ%PNr-#IBj;P?L_r^?dLq@$H8 z12eNzPAgx@-a7qnx!cO4xk!vfOjbd8=j6;m8c1%+bR|=nu~!mg27v9f`G+O<9~5TqLG zC}1o~q7o_PkOtB>d{DL~>7N9(a4AZJ1dc;|5S8<)W{-EkD)fiuS!EJP9MVQUg1_XO`0T?O>p!Gut1XT) zO_TD9e@MK~B*&Umy3GmX9_o-lyep}z?J4|uniSH`^Ba@Rmc2vzKTxhISe{f$k&Ju( z#1Q`t1I;QNd%ow!cJ(L2G}L(keo}WLY;;SQWagp87qZBKKK6X{FY@?HPPyfRq&fq2 z*9dF-zrdGrurjGA4paV#gSXm-=tVM+j<50CPqTkOod1$CxtEK3&Z)H%3H`}M=uc{q zWKUVhqmS)Gn3VNeaz~Cj%;--#qVvrHuSh7wKLkC>@h`3Azd+&wzzax~w((o_pz-H5GI)013^lqF#{(7P67LzuBn|9{XjVcz<$S9fSNtME=w-Rw`|^7BrbAKbxFNKIUlv}(x|J6SU+`o?jPNVl=y-Di7dXRGac)TicN5-b2x_6ga^ z6&xMOJY`-U8A0vQC-WTS5(NObK^7SFkg~W@D(2srlC6G*kqu`LGYNK#18pXu{#hL6 zAfVn#3lUoZEO?3kLJO&{@mpUfmy0dizvku;BOmG9Rd_6RFv(w|3S?3d&LAvy_lt*5b00F!lY4f92MzwN9vIx6`6sT z<(3P-CC9Et``U)^ov|d;y_58(ZMnK%8c6P~%E&Zsku3fs)6+;xGF=NUQ<}vMl7=GwyrfRzEGh5G&CevWf&72UdGQrlFitLd zwaDKI%>Tq}te<<5)N1XW$(*A-^kx|up^bpDk}yuxVZ9(y)jVX!O{UhtQKTq{GAj4% zB87^v&PSxJEoDa||Dj|3!NH_=LQP1g<;4V9CgB4mnPuBoa*%|i(b4pz(2e^nC;X%){jHgh^O}tI z4UYwr&`&a{L4Pcg1zL?w{{nGMYZAIuChS&AmL~LJFCrBexfKR^yv zO8<(TGRcO(;ChjyHNaOc=;uD$8Oj%bgD8?@TFQAyGXv84)e1Yprx5i>a#~78ryCc8 zNMVV2j?Bv_myyA5Q>Hj6c*&E(Hd(B|YF9r=o;Zut{f7^m0Qry|Ig z57{b3lhqC#3w3PzK|#Ubrg&9Wn}Q;jH#mwckvKxG&8WloyPZk%$P51e%9#~UlNnmR z6`7mKC5t5^UC0PPoRuTV?4v&6{WSlfb`KtrdacIWYNQn7(M?~K^Gm}Ni zSVTlbNc(bl@9wk8zl$`sI7JUx@J5>3XH7UiMP|HYSsPML0V308fSEh8{babn>+|6~ z2}uL+MWVPMm?ULr9HAPH_FVBu^v+IVjq!3w3en zs>;~DzG6T3`u~?41|HgsepLvvH)VeZ#(P+D5o7Gy*)VP)IPY&UIdP2&> zpr!p9i#nr4EKP?DU%!@Q=c8YXqdJ1PqF zpbpFX%S7vI`O%5=!HI@lA(js4lA};kaZy{Gs(Hz9N97PdSAx$z8tE~S*%p;UqN%ma z6G|`tE{SiA%WW=B3E{wV60LQODY^LD`$xO6&#s?Hr;%Q}4I6&6Z@LnOo32a1Bjp%f4>MiXWVWDa=c5h$PEM@^!dnTo~2f`#JG*Stgsp7359akwSKYz)@DH<*$`#Wr3(f26=_p>jxs%M{nc0DbY{_V~rJ zs0W4HGd1y&X>{f)wzY1#N6e3@^ur1G5sJ~N__iUTjWRXPPj0z3%og%5H1>-8$*(*l z3=Ixc)N9NgV4IJAcqqX5mTB`N=p9$PRhgT2x<^ZX*!>ukxkB54II4QAX7~_o4PE$+ zu&Z8y0?FU}_WNHIJ*arK-Y>D{$Wf2jXcdpU?|DQsMGg4d zx05zqD={8+=(*VorGJw06`k61jgnqN@h{ILdcCkbKpvJLX0EiSCB#`Z-VF_m~7 zlxGmH)~q{bh&#G3z}MzgjGt&okDLf@7gjL7^F{l`9jGEDNbsx%$- zzlQvNwyv(ORB3d*dU3a4L-KtCg>3HRDJ&<_&yRAb`RI;^_SQ1higD@)Dcsuq#kR;* z1@SDwzGY;vsa-fD39espsN~1$*ZgcRKx-NZC#&l(3;QWA4&ESmyk#kZoF5( zU|9K@bGx86u!cICPlXmal9|kX!|tTjF|C6dJE>&E&-KhC#3j9`d;^d|fbMvNiq8C4 zn)#<*2ttDk`dt<+sG-0!z;W7(0#3Bo-_q=`YRw3D4;EUai?{j8ojljciR_Yz4%6TQ zOP8Q-Wb1m{{Vr2!k?Ayzuv}KFZIOj|{r;B1iVD2vi)Z_o$GDRN&oLvr5*T1w%G*bU z4UB%8Q&=dTC#2x5tEX3{fqs()$u}q(G=nMkli+KeFVr|^X1(+d1twiDWX}%fi;DdOFHXnQp zCOfKw74mZeS(&p#p39(}OL5KpwM^?J-uR80RPZ(e#`O7&bTr9*2ZZIGfHAWk;_%*> zN}KuK@VsYn@qr6~%b89+)6XmOLqk#{Ep<S)fqJQK$52?GY9D(?@yM;M4!H32T z4T&WKza@DiQxE7ab49ATwXJQ}2`Qrr4>3u{aTOPUf!3Cm=RChZyY7E!#Ny+;?+{h@ zTg+xPnC(=^u_4PK?quvl8R4@;(e|N+broZjbai#_hX_gIX7a4(2P@>SGgYV+D-`P< zUu7Enu(IMZswHdy-%K!sIqsMqOY9+Gxtli+3%`2#bm1e73zVhw&`r8*~SL24>UIRWMbh=hru9WA>8RE>Lj!T~&lL~O9 zO@)Y-6uxn;s|c>F*db3Rw7%ZaYM)XwM5!ENQSms`+QtSi&6ylnz=`Z5^e)cwZMXC{ zB77L7<^-7p@PR@x28eIV!Gsd6?#O)GGQFJ&dc)#om_`8#0)t6eA{OwgLl){%SZ#Bl zc-0pEEouvgN8%8ZSM+~_LRj+!j8gBc%NLGOfc_!CvfS6lqMl!t zcRPEsp2^jw;cY3MFnYaKf4fE8NFH)RWN@5T@!tPIwvlOcrYDaM*pTB~f61AaTp|+n zhHE@VX!Hej&B6_TQ1R~=mXK1|vuxf2ZL*CVqd<6o`<*8evHh^tQx)L~eP>*Hw#Jc<~(bI!=WV@-obU zmITJ-65C-=9oRax^=!v*j9HZE7(@^jDAGR!qgl&s*zZ2Kn{~)@k0+nzz&r|Th zM2IX}S)f8xOcwoU_r9crv*j_D>-W!R3?5*H8#Ap(-HqD@0URMfmV0)CF}i>Xf%>m` z7A+lnm_1kLLWu7vEG*o+m)ZRVEj^^)BQ$~&y5ohQcbc_K?O<4Ze4Os7=*I}jtn($? z0fr;Q1f%rops-&oa7)`J7%P+a;%a_a7tlOx#lz zA~-c<(Kf633Y#hVXSW-&d29qHFE4MyV}$FVy9K!2yzc2WqBuj4=S%Q*T#y%_&f%}Y zFrBTxWo)d<0Ev}>dxN_{@%PoOyD6FBC9Y9+_B%Ta#1Sg&LCK84V89#fO}rLr1M8T; zSHPQ4CKW9B^Wel9J3E@g!Wmr=2sTLmCUC4f)g^f^bJfp9<-(ktXQK5N+!w#s_#o)c zs-1C<{9(Juak@Oy!v_JkUEdqE!>mCs_w!!V+{>(oGHQtToGS^eWm1=3hpdc-%5b?y zs-?Bx4s!_RNLJZyKj=h`)3Fc2Xw7_H$SPc5ii0)`0GP1S`k5~oecJT}Fo z=yTZi%#Im&0qoDC-UorjV!^W)=Bo(a=0fNXKKt^Tq6E&Ct^QDx3!~+F0{*YF^r)-? zaV6@H%bh|Th4=>2oD~1s-=D=MW!V(G7N(f#E^!dwM~Z``U0R2A4R!-lSxmzO??G|UUR7&m<7RW2q5+`kU0h_S2my9o~CY*;I*i@GL@F#A|L!QfCi`}@Q4XcmlWDDO1pTKd3MN(#llAM z1T?>IHt-s=&O~DbZ_l&S*Y_$gn|iN9mo5A(TGGsYIRLVfj@Gg^i!b{73xxMXus~&f zJaEMEvgp_H0ZB5W8(2(4yU?_W-qHQSJ731?s-*CQ762DuRQub7!6~!R$1m6qr*fFD z&J7F{oEn&W{Y1e5k zpfsldwY)2h=FaA!g{aVOhisTx65_}?AgX9Uc=A*|bQHnIUVRTO?b~q^KnMVfWNQGG%#P{~eJ%{ty%vBy-QfM{SuijxsSk}8r=tORF1}-Rq&b&kF_~|1M z?H%+oo9*bbq45`i#{EL_Wdnl_Cl!h(j=VTbXMVUW<7$|O$oBDw>_zgQQ(!)+Hc1cl zO}yjm>|9sFG-47EC=}E#)F`A7*Dkb+1hO496_{CBS+&^BeLX+PBP>IN9(R5FX2!=j z*s$GRK2H^uo(UC}dm+f($c++=Mq2_7W_CD;$f!*b8p>gJpIH{0L&*@`^FbiwDilE)nyNB*S+OD%hEn4D{W4*tln{g^S z=&|mo?{^gTGkd5u?FWGn1h|Zl0HzU9+DJQNBdUDV%( z8kp|F8UTHFFi?;dI$c9+Wo1=BY$QXD7;fD$N2CIg?>CD}82`MZ3wi$^UbU@zzo#>-o0v*1UZcB%~^&aC<9~ z5~TTU4S80v+{CDUcm_G9LIxVF^n??`d}fx`Y&B719@YqhBPJp34I z+XbqbIBQ<|Go7APs~-l5>Dwm&7;RnS*Mqj&r50^W<0^oq3)v61? z#sm*ZZ?28bJlc0+(tWNHH#Y z|0%B4Cscx|QZ@Aj42+DfH6h}PVDP(th0+vw#L?@b?k+*LVIKf?yubXl!qau7*DAa7 zfz7X5-p?E{LtGEjAj-^#D6?*qapqt3?>SS=6u;HCKQ8UI_Y;AP0tN;KFUcK|^rt*i zI37!Qj7WfQ_VjUYM3r7hslCkel{i$92C;_wcbs;6kT{gQ z>+G6yHqP=6xT!uY8smf`D{^-}%_G zxzbEYVIZ0-=YNZ5i6@`&j#k4wP(Amxk$}7*Z=B8CN3e%#TQ{da>|>vyMz8G~V0HJ& Uhf5N7_){oeQ@e_iyZ!k80D+g(u>b%7 diff --git a/content/en/tutorials/integration-tutorials/keras.md b/content/en/tutorials/integration-tutorials/keras.md index 45062fd088..fd08253492 100644 --- a/content/en/tutorials/integration-tutorials/keras.md +++ b/content/en/tutorials/integration-tutorials/keras.md @@ -63,7 +63,7 @@ configs = dict( ## Dataset -In this colab, we will be using [CIFAR100](https://www.tensorflow.org/datasets/catalog/cifar100) dataset from TensorFlow Dataset catalog. We aim to build a simple image classification pipeline using TensorFlow/Keras. +In this colab, we will be using [Fashion-MNIST](https://www.tensorflow.org/datasets/catalog/fashion_mnist) dataset from TensorFlow Dataset catalog. We aim to build a simple image classification pipeline using TensorFlow/Keras. ```python @@ -167,4 +167,4 @@ model.fit( # Close the W&B Run run.finish() -``` \ No newline at end of file +``` diff --git a/content/en/tutorials/integration-tutorials/keras_models.md b/content/en/tutorials/integration-tutorials/keras_models.md index c59b17f5f7..3f21a69a40 100644 --- a/content/en/tutorials/integration-tutorials/keras_models.md +++ b/content/en/tutorials/integration-tutorials/keras_models.md @@ -63,7 +63,7 @@ configs = dict( ## Dataset -In this colab, we will be using [CIFAR100](https://www.tensorflow.org/datasets/catalog/cifar100) dataset from TensorFlow Dataset catalog. We aim to build a simple image classification pipeline using TensorFlow/Keras. +In this colab, we will be using [Fashion-MNIST](https://www.tensorflow.org/datasets/catalog/fashion_mnist) dataset from TensorFlow Dataset catalog. We aim to build a simple image classification pipeline using TensorFlow/Keras. ```python diff --git a/content/en/tutorials/integration-tutorials/keras_tables.md b/content/en/tutorials/integration-tutorials/keras_tables.md index f5de511a23..c56c86d6b6 100644 --- a/content/en/tutorials/integration-tutorials/keras_tables.md +++ b/content/en/tutorials/integration-tutorials/keras_tables.md @@ -65,7 +65,7 @@ configs = dict( ## Dataset -In this colab, we will be using [CIFAR100](https://www.tensorflow.org/datasets/catalog/cifar100) dataset from TensorFlow Dataset catalog. We aim to build a simple image classification pipeline using TensorFlow/Keras. +In this colab, we will be using [Fashion-MNIST](https://www.tensorflow.org/datasets/catalog/fashion_mnist) dataset from TensorFlow Dataset catalog. We aim to build a simple image classification pipeline using TensorFlow/Keras. ```python From 6093944522a03f649a8cdebca314d29e4992cc01 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Thu, 21 Aug 2025 12:05:28 -0700 Subject: [PATCH 07/31] test: temporarily use mock preview URL for testing path matching --- .github/workflows/pr-preview-links.yml | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/.github/workflows/pr-preview-links.yml b/.github/workflows/pr-preview-links.yml index c99c7e29d9..871c434a6b 100644 --- a/.github/workflows/pr-preview-links.yml +++ b/.github/workflows/pr-preview-links.yml @@ -216,9 +216,11 @@ jobs: core.warning('Branch Preview URL not found within 2 minutes. Links will be added when Cloudflare comment triggers update workflow.'); return ''; } - let base = await waitForCloudflareBranchUrl(); - if (!base) base = await fromOurExistingComment(); - core.info(`Branch Preview URL found: ${base || '(not yet)'}`); + // TEMPORARY: Use mock preview URL for testing path matching logic + // let base = await waitForCloudflareBranchUrl(); + // if (!base) base = await fromOurExistingComment(); + let base = 'https://test-preview.pages.dev'; + core.info(`Using mock preview URL for testing: ${base}`); core.setOutput('base', base); - name: Update PR comment with preview links @@ -557,7 +559,7 @@ jobs: const modifiedRows = buildRows(modifiedInfo.files); const deletedRows = buildRows(deletedInfo.files); - const header = '\n\n**PR Preview: Changed content**' + (previewBase ? `\n\nBase preview: ${previewBase}` : '\n\n⏳ *Links will be added automatically when Cloudflare Pages finishes deploying (typically 2-5 minutes)*'); + const header = '\n\n**PR Preview: Changed content**' + (previewBase ? `\n\nBase preview: ${previewBase} (MOCK URL FOR TESTING)` : '\n\n⏳ *Links will be added automatically when Cloudflare Pages finishes deploying (typically 2-5 minutes)*'); if (!previewBase) { core.info('Preview base URL not available yet. Comment will be automatically updated when Cloudflare posts the Branch Preview URL.'); } From fc5abb0c174b4b399e766a3f77b6a25bb7092e88 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Thu, 21 Aug 2025 12:11:29 -0700 Subject: [PATCH 08/31] revert: remove mock URL test changes --- .github/workflows/pr-preview-links.yml | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/.github/workflows/pr-preview-links.yml b/.github/workflows/pr-preview-links.yml index 871c434a6b..c99c7e29d9 100644 --- a/.github/workflows/pr-preview-links.yml +++ b/.github/workflows/pr-preview-links.yml @@ -216,11 +216,9 @@ jobs: core.warning('Branch Preview URL not found within 2 minutes. Links will be added when Cloudflare comment triggers update workflow.'); return ''; } - // TEMPORARY: Use mock preview URL for testing path matching logic - // let base = await waitForCloudflareBranchUrl(); - // if (!base) base = await fromOurExistingComment(); - let base = 'https://test-preview.pages.dev'; - core.info(`Using mock preview URL for testing: ${base}`); + let base = await waitForCloudflareBranchUrl(); + if (!base) base = await fromOurExistingComment(); + core.info(`Branch Preview URL found: ${base || '(not yet)'}`); core.setOutput('base', base); - name: Update PR comment with preview links @@ -559,7 +557,7 @@ jobs: const modifiedRows = buildRows(modifiedInfo.files); const deletedRows = buildRows(deletedInfo.files); - const header = '\n\n**PR Preview: Changed content**' + (previewBase ? `\n\nBase preview: ${previewBase} (MOCK URL FOR TESTING)` : '\n\n⏳ *Links will be added automatically when Cloudflare Pages finishes deploying (typically 2-5 minutes)*'); + const header = '\n\n**PR Preview: Changed content**' + (previewBase ? `\n\nBase preview: ${previewBase}` : '\n\n⏳ *Links will be added automatically when Cloudflare Pages finishes deploying (typically 2-5 minutes)*'); if (!previewBase) { core.info('Preview base URL not available yet. Comment will be automatically updated when Cloudflare posts the Branch Preview URL.'); } From 48196ce43815ca795f0b73e7cd3feff788c671f7 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Fri, 22 Aug 2025 15:25:39 -0700 Subject: [PATCH 09/31] docs(weave): fix broken relref links in Japanese and Korean media guides --- content/ja/guides/integrations/pytorch.md | 2 +- content/ja/guides/models/track/log/media.md | 2 +- content/ja/launch/integration-guides/dagster.md | 2 +- content/ko/guides/integrations/pytorch.md | 2 +- content/ko/guides/models/track/log/media.md | 2 +- content/ko/launch/integration-guides/dagster.md | 2 +- image_normalization_test.ipynb | 1 + 7 files changed, 7 insertions(+), 6 deletions(-) create mode 100644 image_normalization_test.ipynb diff --git a/content/ja/guides/integrations/pytorch.md b/content/ja/guides/integrations/pytorch.md index 4577f8cb56..320800c3cb 100644 --- a/content/ja/guides/integrations/pytorch.md +++ b/content/ja/guides/integrations/pytorch.md @@ -49,7 +49,7 @@ for batch_idx, (data, target) in enumerate(train_loader): ## 画像とメディアのログ -画像データを持つ PyTorch `Tensors` を [`wandb.Image`]({{< relref path="/ref/python/data-types/image.md" lang="ja" >}}) に渡すことができ、[`torchvision`](https://pytorch.org/vision/stable/index.html) のユーティリティが自動的に画像に変換します。 +画像データを持つ PyTorch `Tensors` を [`wandb.Image`]({{< relref "/ref/python/data-types/image.md" >}}) に渡すことができ、[`torchvision`](https://pytorch.org/vision/stable/index.html) のユーティリティが自動的に画像に変換します。 ```python images_t = ... # PyTorch Tensors として画像を生成またはロードする diff --git a/content/ja/guides/models/track/log/media.md b/content/ja/guides/models/track/log/media.md index 21938b0e43..c650429d2b 100644 --- a/content/ja/guides/models/track/log/media.md +++ b/content/ja/guides/models/track/log/media.md @@ -52,7 +52,7 @@ images = wandb.Image(image_array, caption="Top: Output, Bottom: Input") wandb.log({"examples": images}) ``` -最後の次元が1の場合はグレースケール、3の場合はRGB、4の場合はRGBAと仮定します。配列が浮動小数点数を含む場合、正規化アルゴリズムを使用して自動的に`0`から`255`の整数に変換します。PyTorchテンソルとNumPy配列での正規化の動作についての詳細は、[Imageリファレンスの画像正規化セクション]({{< relref path="/ref/python/data-types/image.md#image-normalization" lang="ja" >}})を参照してください。異なる方法で画像を正規화したい場合は、[`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)を手動で指定するか、`"Logging PIL Images"`タブで説明されているように、単に[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を提供することができます。 +最後の次元が1の場合はグレースケール、3の場合はRGB、4の場合はRGBAと仮定します。配列が浮動小数点数を含む場合、正規化アルゴリズムを使用して自動的に`0`から`255`の整数に変換します。PyTorchテンソルとNumPy配列での正規化の動作についての詳細は、[Imageリファレンスの画像正規化セクション]({{< relref "/ref/python/data-types/image.md#image-normalization" >}})を参照してください。異なる方法で画像を正規화したい場合は、[`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)を手動で指定するか、`"Logging PIL Images"`タブで説明されているように、単に[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を提供することができます。 {{% /tab %}} {{% tab header="PIL Imagesをログする" %}} 配列から画像への変換を完全に制御するために、[`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)を自分で構築し、直接提供してください。 diff --git a/content/ja/launch/integration-guides/dagster.md b/content/ja/launch/integration-guides/dagster.md index 89497aa907..dfd089d28d 100644 --- a/content/ja/launch/integration-guides/dagster.md +++ b/content/ja/launch/integration-guides/dagster.md @@ -181,7 +181,7 @@ def create_dataset(): W&B は複数のピクルスベースのシリアライズモジュール([pickle](https://docs.python.org/3/library/pickle.html), [dill](https://github.com/uqfoundation/dill), [cloudpickle](https://github.com/cloudpipe/cloudpickle), [joblib](https://github.com/joblib/joblib)) をサポートしています。また、[ONNX](https://onnx.ai/) や [PMML](https://en.wikipedia.org/wiki/Predictive_Model_Markup_Language) といったより高度なシリアライズも利用できます。[Serialization]({{< relref path="#serialization-configuration" lang="ja" >}}) セクションを参照してください。 {{% /tab %}} {{% tab "W&B Object" %}} -ネイティブ W&B オブジェクト (例: [Table]({{< relref path="/ref/python/data-types/table.md" lang="ja" >}}), [Image]({{< relref path="/ref/python/data-types/image.md" lang="ja" >}}), or [Graph]({{< relref path="/ref/python/data-types/graph.md" lang="ja" >}})) のいずれかが作成された Artifact にインテグレーションによって追加されます。以下は Table を使った例です。 +ネイティブ W&B オブジェクト (例: [Table]({{< relref path="/ref/python/data-types/table.md" lang="ja" >}}), [Image]({{< relref "/ref/python/data-types/image.md" >}}), or [Graph]({{< relref path="/ref/python/data-types/graph.md" lang="ja" >}})) のいずれかが作成された Artifact にインテグレーションによって追加されます。以下は Table を使った例です。 ```python import wandb diff --git a/content/ko/guides/integrations/pytorch.md b/content/ko/guides/integrations/pytorch.md index 403e106a1a..d4e107d6ea 100644 --- a/content/ko/guides/integrations/pytorch.md +++ b/content/ko/guides/integrations/pytorch.md @@ -49,7 +49,7 @@ gradient, metrics 및 그래프는 순방향 _및_ 역방향 패스 후에 `wand ## 이미지 및 미디어 로그 -이미지 데이터가 포함된 PyTorch `Tensors`를 [`wandb.Image`]({{< relref path="/ref/python/data-types/image.md" lang="ko" >}})로 전달할 수 있으며, [`torchvision`](https://pytorch.org/vision/stable/index.html)의 유틸리티가 자동으로 이미지를 변환하는 데 사용됩니다. +이미지 데이터가 포함된 PyTorch `Tensors`를 [`wandb.Image`]({{< relref "/ref/python/data-types/image.md" >}})로 전달할 수 있으며, [`torchvision`](https://pytorch.org/vision/stable/index.html)의 유틸리티가 자동으로 이미지를 변환하는 데 사용됩니다. ```python images_t = ... # generate or load images as PyTorch Tensors diff --git a/content/ko/guides/models/track/log/media.md b/content/ko/guides/models/track/log/media.md index e37811c81f..e9fc560072 100644 --- a/content/ko/guides/models/track/log/media.md +++ b/content/ko/guides/models/track/log/media.md @@ -53,7 +53,7 @@ images = wandb.Image(image_array, caption="Top: Output, Bottom: Input") wandb.log({"examples": images}) ``` -마지막 차원이 1이면 이미지가 회색조, 3이면 RGB, 4이면 RGBA라고 가정합니다. 배열에 float가 포함된 경우 정규화 알고리즘을 사용하여 자동으로 `0`과 `255` 사이의 정수로 변환합니다. PyTorch 텐서와 NumPy 배열에서 정규화가 어떻게 작동하는지에 대한 자세한 정보는 Image 참조 문서를 참조하세요. 이미지를 다르게 정규화하려면 [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)를 수동으로 지정하거나 이 패널의 "PIL 이미지 로깅" 탭에 설명된 대로 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 제공하면 됩니다. +마지막 차원이 1이면 이미지가 회색조, 3이면 RGB, 4이면 RGBA라고 가정합니다. 배열에 float가 포함된 경우 정규화 알고리즘을 사용하여 자동으로 `0`과 `255` 사이의 정수로 변환합니다. PyTorch 텐서와 NumPy 배열에서 정규화가 어떻게 작동하는지에 대한 자세한 정보는 [Image 참조 문서의 이미지 정규화 섹션]({{< relref "/ref/python/data-types/image.md#image-normalization" >}})을 참조하세요. 이미지를 다르게 정규화하려면 [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes)를 수동으로 지정하거나 이 패널의 "PIL 이미지 로깅" 탭에 설명된 대로 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 제공하면 됩니다. {{% /tab %}} {{% tab header="PIL 이미지 로깅" %}} 배열을 이미지로 변환하는 것을 완벽하게 제어하려면 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)를 직접 구성하여 제공합니다. diff --git a/content/ko/launch/integration-guides/dagster.md b/content/ko/launch/integration-guides/dagster.md index 036d57afb2..a22daa7844 100644 --- a/content/ko/launch/integration-guides/dagster.md +++ b/content/ko/launch/integration-guides/dagster.md @@ -182,7 +182,7 @@ def create_dataset(): W&B는 여러 Pickle 기반 직렬화 모듈을 지원합니다 ([pickle](https://docs.python.org/3/library/pickle.html), [dill](https://github.com/uqfoundation/dill), [cloudpickle](https://github.com/cloudpipe/cloudpickle), [joblib](https://github.com/joblib/joblib)). [ONNX](https://onnx.ai/) 또는 [PMML](https://en.wikipedia.org/wiki/Predictive_Model_Markup_Language)과 같은 고급 직렬화를 사용할 수도 있습니다. 자세한 내용은 [직렬화]({{< relref path="#serialization-configuration" lang="ko" >}}) 섹션을 참조하십시오. {{% /tab %}} {{% tab "W&B Object" %}} -모든 기본 W&B 오브젝트 (예: [Table]({{< relref path="/ref/python/data-types/table.md" lang="ko" >}}), [Image]({{< relref path="/ref/python/data-types/image.md" lang="ko" >}}) 또는 [Graph]({{< relref path="/ref/python/data-types/graph.md" lang="ko" >}}))가 통합에서 생성된 Artifact에 추가됩니다. 다음은 Table을 사용하는 예입니다. +모든 기본 W&B 오브젝트 (예: [Table]({{< relref path="/ref/python/data-types/table.md" lang="ko" >}}), [Image]({{< relref "/ref/python/data-types/image.md" >}}) 또는 [Graph]({{< relref path="/ref/python/data-types/graph.md" lang="ko" >}}))가 통합에서 생성된 Artifact에 추가됩니다. 다음은 Table을 사용하는 예입니다. ```python import wandb diff --git a/image_normalization_test.ipynb b/image_normalization_test.ipynb new file mode 100644 index 0000000000..0519ecba6e --- /dev/null +++ b/image_normalization_test.ipynb @@ -0,0 +1 @@ + \ No newline at end of file From 585403fa04a208ede6a79890cacc0237d23e889c Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 11:13:02 -0700 Subject: [PATCH 10/31] docs: address review feedback - move detailed normalization content to dev guide, simplify reference --- content/en/guides/models/track/log/media.md | 99 ++++++++++++++++- content/en/ref/python/sdk/data-types/Image.md | 103 ++---------------- 2 files changed, 110 insertions(+), 92 deletions(-) diff --git a/content/en/guides/models/track/log/media.md b/content/en/guides/models/track/log/media.md index ed80f6f414..192bfaff44 100644 --- a/content/en/guides/models/track/log/media.md +++ b/content/en/guides/models/track/log/media.md @@ -58,7 +58,7 @@ with wandb.init(project="image-log-example") as run: run.log({"examples": images}) ``` -The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image normalization section]({{< relref "/ref/python/sdk/data-types/image.md#image-normalization" >}}) in the Image SDK reference. To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. +The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image normalization section](#image-normalization) below. To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. {{% /tab %}} {{% tab header="Logging PIL Images" %}} For full control over the conversion of arrays to images, construct the [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html) yourself and provide it directly. @@ -97,6 +97,103 @@ with wandb.init(project="") as run: {{% /tab %}} {{< /tabpane >}} +## Image normalization + +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle and ensure proper display of common image formats. + +### When normalization is applied + +Normalization is applied to: +- **PyTorch tensors** (format: `(channel, height, width)`) +- **NumPy arrays** (format: `(height, width, channel)`) + +Normalization is **NOT** applied to: +- **PIL Images** (passed as-is) +- **File paths** (loaded as-is) + +### Normalization algorithm + +The normalization algorithm automatically detects the input range and applies the appropriate transformation: + +1. **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 + ```python + normalized_data = (data * 255).astype(np.uint8) + ``` + +2. **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: + ```python + normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) + ``` + +3. **For any other range**: Values are clipped to [0, 255] and converted to uint8 + ```python + normalized_data = data.clip(0, 255).astype(np.uint8) + ``` + +### Examples of normalization effects + +**Example 1: [0, 1] range data** +```python +import torch +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) # Random values between 0 and 1 + +# This will multiply all values by 255 +image = wandb.Image(tensor_0_1, caption="Normalized from [0,1] range") +``` + +**Example 2: [-1, 1] range data** +```python +import torch +import wandb + +# Create tensor with values in [-1, 1] range +tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # Random values between -1 and 1 + +# This will rescale: -1 → 0, 0 → 127.5, 1 → 255 +image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") +``` + +**Example 3: Avoiding normalization with PIL Images** +```python +import torch +from PIL import Image as PILImage +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) + +# Convert to PIL Image to avoid normalization +pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) +image = wandb.Image(pil_image, caption="No normalization applied") +``` + +**Example 4: Using normalize=False** +```python +import torch +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) + +# Disable normalization - values will be clipped to [0, 255] +image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") +``` + +### Best practices + +1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging +2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` +3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) +4. **For precise control**: Use PIL Images when you need exact pixel values + +### Common issues + +- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter +- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information +- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results ## Image overlays diff --git a/content/en/ref/python/sdk/data-types/Image.md b/content/en/ref/python/sdk/data-types/Image.md index a385fe2aa0..a7f4bfea11 100644 --- a/content/en/ref/python/sdk/data-types/Image.md +++ b/content/en/ref/python/sdk/data-types/Image.md @@ -105,101 +105,22 @@ with wandb.init() as run: ## Image normalization -When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle and ensure proper display of common image formats. +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. -### When normalization is applied +**Normalization is applied to:** +- PyTorch tensors (format: `(channel, height, width)`) +- NumPy arrays (format: `(height, width, channel)`) -Normalization is applied to: -- **PyTorch tensors** (format: `(channel, height, width)`) -- **NumPy arrays** (format: `(height, width, channel)`) +**Normalization is NOT applied to:** +- PIL Images (passed as-is) +- File paths (loaded as-is) -Normalization is **NOT** applied to: -- **PIL Images** (passed as-is) -- **File paths** (loaded as-is) +**Normalization algorithm:** +- [0, 1] range: values are multiplied by 255 +- [-1, 1] range: values are rescaled using `255 * 0.5 * (data + 1)` +- Other ranges: values are clipped to [0, 255] -### Normalization algorithm - -The normalization algorithm automatically detects the input range and applies the appropriate transformation: - -1. **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 - ```python - normalized_data = (data * 255).astype(np.uint8) - ``` - -2. **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: - ```python - normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) - ``` - -3. **For any other range**: Values are clipped to [0, 255] and converted to uint8 - ```python - normalized_data = data.clip(0, 255).astype(np.uint8) - ``` - -### Examples of normalization effects - -**Example 1: [0, 1] range data** -```python -import torch -import wandb - -# Create tensor with values in [0, 1] range -tensor_0_1 = torch.rand(3, 64, 64) # Random values between 0 and 1 - -# This will multiply all values by 255 -image = wandb.Image(tensor_0_1, caption="Normalized from [0,1] range") -``` - -**Example 2: [-1, 1] range data** -```python -import torch -import wandb - -# Create tensor with values in [-1, 1] range -tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # Random values between -1 and 1 - -# This will rescale: -1 → 0, 0 → 127.5, 1 → 255 -image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") -``` - -**Example 3: Avoiding normalization with PIL Images** -```python -import torch -from PIL import Image as PILImage -import wandb - -# Create tensor with values in [0, 1] range -tensor_0_1 = torch.rand(3, 64, 64) - -# Convert to PIL Image to avoid normalization -pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) -image = wandb.Image(pil_image, caption="No normalization applied") -``` - -**Example 4: Using normalize=False** -```python -import torch -import wandb - -# Create tensor with values in [0, 1] range -tensor_0_1 = torch.rand(3, 64, 64) - -# Disable normalization - values will be clipped to [0, 255] -image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") -``` - -### Recommendations - -1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging -2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` -3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) -4. **For precise control**: Use PIL Images when you need exact pixel values - -### Troubleshooting - -- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter -- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information -- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results +For detailed examples and best practices, see the [Image normalization section]({{< relref "/guides/models/track/log/media.md#image-normalization" >}}) in the media logging guide. --- From fdebb3437ef68196e713e6efe5fff3dda6b38f65 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 12:39:29 -0700 Subject: [PATCH 11/31] feat: add image normalization demo notebook for testing and user examples --- wandb_image_normalization_demo.ipynb | 322 +++++++++++++++++++++++++++ 1 file changed, 322 insertions(+) create mode 100644 wandb_image_normalization_demo.ipynb diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb new file mode 100644 index 0000000000..7ccb821afd --- /dev/null +++ b/wandb_image_normalization_demo.ipynb @@ -0,0 +1,322 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {"id": "title"}, + "source": [ + "# W&B Image Normalization Demo\n", + "\n", + "This notebook demonstrates how `wandb.Image` automatically normalizes different types of image data and how to control this behavior.\n", + "\n", + "## What you'll learn:\n", + "- How `wandb.Image` normalizes PyTorch tensors and NumPy arrays\n", + "- When normalization is applied vs when it's not\n", + "- How to avoid unwanted normalization\n", + "- Best practices for image logging" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "setup"}, + "source": [ + "## Setup\n", + "\n", + "First, let's install the required dependencies and import the necessary libraries." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "install_deps"}, + "outputs": [], + "source": [ + "# Install required packages\n", + "!pip install wandb torch torchvision pillow matplotlib numpy" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "imports"}, + "outputs": [], + "source": [ + "import wandb\n", + "import torch\n", + "import numpy as np\n", + "from PIL import Image as PILImage\n", + "import matplotlib.pyplot as plt\n", + "\n", + "# Set up matplotlib for better visualization\n", + "plt.rcParams['figure.figsize'] = (12, 8)\n", + "plt.rcParams['font.size'] = 10" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "init_wandb"}, + "source": [ + "## Initialize W&B\n", + "\n", + "Let's start a W&B run to log our examples." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "wandb_init"}, + "outputs": [], + "source": [ + "# Initialize W&B run\n", + "run = wandb.init(\n", + " project=\"image-normalization-demo\",\n", + " name=\"normalization-examples\",\n", + " config={\n", + " \"description\": \"Demonstrating wandb.Image normalization behavior\"\n", + " }\n", + ")" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "overview"}, + "source": [ + "## Understanding Image Normalization\n", + "\n", + "When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`.\n", + "\n", + "**Normalization is applied to:**\n", + "- PyTorch tensors (format: `(channel, height, width)`)\n", + "- NumPy arrays (format: `(height, width, channel)`)\n", + "\n", + "**Normalization is NOT applied to:**\n", + "- PIL Images (passed as-is)\n", + "- File paths (loaded as-is)\n", + "\n", + "**Normalization algorithm:**\n", + "- [0, 1] range: values are multiplied by 255\n", + "- [-1, 1] range: values are rescaled using `255 * 0.5 * (data + 1)`\n", + "- Other ranges: values are clipped to [0, 255]" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "example_1"}, + "source": [ + "## Example 1: [0, 1] Range Data\n", + "\n", + "When your tensor/array values are in the [0, 1] range, `wandb.Image` will multiply all values by 255." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "example_0_1_range"}, + "outputs": [], + "source": [ + "# Create test data in [0, 1] range\n", + "data_0_1 = np.random.rand(64, 64, 3)\n", + "print(f\"Original data range: [{data_0_1.min():.3f}, {data_0_1.max():.3f}]\")\n", + "\n", + "# Convert to PyTorch tensor (channel, height, width format)\n", + "tensor_0_1 = torch.from_numpy(data_0_1).permute(2, 0, 1).float()\n", + "print(f\"Tensor shape: {tensor_0_1.shape}\")\n", + "print(f\"Tensor range: [{tensor_0_1.min():.3f}, {tensor_0_1.max():.3f}]\")\n", + "\n", + "# Visualize the original data\n", + "plt.figure(figsize=(8, 6))\n", + "plt.imshow(data_0_1)\n", + "plt.title(f'[0, 1] Range Data\\nValues will be multiplied by 255')\n", + "plt.colorbar()\n", + "plt.axis('off')\n", + "plt.show()\n", + "\n", + "# Log to W&B\n", + "wandb.log({\n", + " \"example_0_1_range\": wandb.Image(\n", + " tensor_0_1,\n", + " caption=\"[0, 1] range tensor - values will be multiplied by 255\"\n", + " )\n", + "})" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "example_2"}, + "source": [ + "## Example 2: [-1, 1] Range Data\n", + "\n", + "When your tensor/array values are in the [-1, 1] range, `wandb.Image` will rescale them using the formula: `255 * 0.5 * (data + 1)`" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "example_neg1_1_range"}, + "outputs": [], + "source": [ + "# Create test data in [-1, 1] range\n", + "data_neg1_1 = np.random.rand(64, 64, 3) * 2 - 1\n", + "print(f\"Original data range: [{data_neg1_1.min():.3f}, {data_neg1_1.max():.3f}]\")\n", + "\n", + "# Convert to PyTorch tensor\n", + "tensor_neg1_1 = torch.from_numpy(data_neg1_1).permute(2, 0, 1).float()\n", + "print(f\"Tensor shape: {tensor_neg1_1.shape}\")\n", + "print(f\"Tensor range: [{tensor_neg1_1.min():.3f}, {tensor_neg1_1.max():.3f}]\")\n", + "\n", + "# Visualize the original data\n", + "plt.figure(figsize=(8, 6))\n", + "plt.imshow(data_neg1_1, cmap='RdBu_r')\n", + "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1→0, 0→127.5, 1→255')\n", + "plt.colorbar()\n", + "plt.axis('off')\n", + "plt.show()\n", + "\n", + "# Log to W&B\n", + "wandb.log({\n", + " \"example_neg1_1_range\": wandb.Image(\n", + " tensor_neg1_1,\n", + " caption=\"[-1, 1] range tensor - values will be rescaled\"\n", + " )\n", + "})" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "example_3"}, + "source": [ + "## Example 3: Avoiding Normalization with PIL Images\n", + "\n", + "To avoid normalization, you can convert your tensors to PIL Images before passing them to `wandb.Image`." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "example_pil_avoid_normalization"}, + "outputs": [], + "source": [ + "# Create tensor with values in [0, 1] range\n", + "tensor_0_1 = torch.rand(3, 64, 64)\n", + "print(f\"Tensor range: [{tensor_0_1.min():.3f}, {tensor_0_1.max():.3f}]\")\n", + "\n", + "# Convert to PIL Image to avoid normalization\n", + "pil_image = PILImage.fromarray(\n", + " (tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')\n", + ")\n", + "print(f\"PIL Image size: {pil_image.size}\")\n", + "print(f\"PIL Image mode: {pil_image.mode}\")\n", + "\n", + "# Visualize the PIL image\n", + "plt.figure(figsize=(8, 6))\n", + "plt.imshow(pil_image)\n", + "plt.title('PIL Image - No normalization applied')\n", + "plt.axis('off')\n", + "plt.show()\n", + "\n", + "# Log to W&B\n", + "wandb.log({\n", + " \"example_pil_no_normalization\": wandb.Image(\n", + " pil_image,\n", + " caption=\"PIL Image - no normalization applied\"\n", + " )\n", + "})" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "example_4"}, + "source": [ + "## Example 4: Using normalize=False\n", + "\n", + "You can also disable normalization by setting `normalize=False`. Values will be clipped to [0, 255]." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "example_normalize_false"}, + "outputs": [], + "source": [ + "# Create tensor with values in [0, 1] range\n", + "tensor_0_1 = torch.rand(3, 64, 64)\n", + "print(f\"Tensor range: [{tensor_0_1.min():.3f}, {tensor_0_1.max():.3f}]\")\n", + "\n", + "# Disable normalization\n", + "wandb.log({\n", + " \"example_normalize_false\": wandb.Image(\n", + " tensor_0_1,\n", + " normalize=False,\n", + " caption=\"Normalization disabled - values will be clipped to [0, 255]\"\n", + " )\n", + "})\n", + "\n", + "# Also log with normal normalization for comparison\n", + "wandb.log({\n", + " \"example_normalize_true\": wandb.Image(\n", + " tensor_0_1,\n", + " normalize=True,\n", + " caption=\"Normalization enabled - values will be multiplied by 255\"\n", + " )\n", + "})\n", + "\n", + "print(\"Logged both normalized and non-normalized versions for comparison\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {"id": "best_practices"}, + "source": [ + "## Best Practices\n", + "\n", + "Based on what we've learned, here are some best practices for working with `wandb.Image`:\n", + "\n", + "### 1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging\n", + "### 2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()`\n", + "### 3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255])\n", + "### 4. **For precise control**: Use PIL Images when you need exact pixel values\n", + "\n", + "### Common Issues to Watch Out For:\n", + "- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter\n", + "- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information\n", + "- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {"id": "finish"}, + "outputs": [], + "source": [ + "# Finish the W&B run\n", + "wandb.finish()\n", + "print(\"✅ Demo completed! Check your W&B dashboard to see all the logged images.\")" + ] + } + ], + "metadata": { + "accelerator": "GPU", + "colab": { + "gpuType": "T4", + "provenance": [] + }, + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.8.5" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} From 71f10925c4ff03c8ef30db5fab020a23eb6d0480 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 12:45:41 -0700 Subject: [PATCH 12/31] refactor: move detailed normalization docs to dedicated guide and simplify reference docs --- .../models/track/log/image-normalization.md | 110 ++++++++++++++++++ content/en/guides/models/track/log/media.md | 100 +--------------- content/en/ref/python/sdk/data-types/Image.md | 17 +-- 3 files changed, 112 insertions(+), 115 deletions(-) create mode 100644 content/en/guides/models/track/log/image-normalization.md diff --git a/content/en/guides/models/track/log/image-normalization.md b/content/en/guides/models/track/log/image-normalization.md new file mode 100644 index 0000000000..dba676f60d --- /dev/null +++ b/content/en/guides/models/track/log/image-normalization.md @@ -0,0 +1,110 @@ +--- +title: "Image Normalization Guide" +description: "Learn how wandb.Image handles normalization for different input types and how to control this behavior" +--- + +# Image Normalization Guide + +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This guide explains how normalization works and how to control it. + +## When normalization is applied + +Normalization is applied to: +- **PyTorch tensors** (format: `(channel, height, width)`) +- **NumPy arrays** (format: `(height, width, channel)`) + +Normalization is **NOT** applied to: +- **PIL Images** (passed as-is) +- **File paths** (loaded as-is) + +## Normalization algorithm + +The normalization algorithm automatically detects the input range and applies the appropriate transformation: + +1. **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 + ```python + normalized_data = (data * 255).astype(np.uint8) + ``` + +2. **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: + ```python + normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) + ``` + +3. **For any other range**: Values are clipped to [0, 255] and converted to uint8 + ```python + normalized_data = data.clip(0, 255).astype(np.uint8) + ``` + +## Examples of normalization effects + +### Example 1: [0, 1] range data + +```python +import torch +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) # Random values between 0 and 1 + +# This will multiply all values by 255 +image = wandb.Image(tensor_0_1, caption="Normalized from [0,1] range") +``` + +### Example 2: [-1, 1] range data + +```python +import torch +import wandb + +# Create tensor with values in [-1, 1] range +tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # Random values between -1 and 1 + +# This will rescale: -1 → 0, 0 → 127.5, 1 → 255 +image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") +``` + +### Example 3: Avoiding normalization with PIL Images + +```python +import torch +from PIL import Image as PILImage +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) + +# Convert to PIL Image to avoid normalization +pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) +image = wandb.Image(pil_image, caption="No normalization applied") +``` + +### Example 4: Using normalize=False + +```python +import torch +import wandb + +# Create tensor with values in [0, 1] range +tensor_0_1 = torch.rand(3, 64, 64) + +# Disable normalization - values will be clipped to [0, 255] +image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") +``` + +## Best practices + +1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging +2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` +3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) +4. **For precise control**: Use PIL Images when you need exact pixel values + +## Common issues + +- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter +- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information +- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results + +## Testing your code + +You can test the normalization behavior using our [Image Normalization Demo Notebook](https://github.com/wandb/wandb/blob/main/wandb_image_normalization_demo.ipynb) which demonstrates all the examples above with visual output. \ No newline at end of file diff --git a/content/en/guides/models/track/log/media.md b/content/en/guides/models/track/log/media.md index 192bfaff44..5bc6b5ae11 100644 --- a/content/en/guides/models/track/log/media.md +++ b/content/en/guides/models/track/log/media.md @@ -58,7 +58,7 @@ with wandb.init(project="image-log-example") as run: run.log({"examples": images}) ``` -The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image normalization section](#image-normalization) below. To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. +The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image Normalization Guide]({{< relref "/guides/models/track/log/image-normalization.md" >}}). To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. {{% /tab %}} {{% tab header="Logging PIL Images" %}} For full control over the conversion of arrays to images, construct the [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html) yourself and provide it directly. @@ -97,104 +97,6 @@ with wandb.init(project="") as run: {{% /tab %}} {{< /tabpane >}} -## Image normalization - -When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This normalization is designed to handle and ensure proper display of common image formats. - -### When normalization is applied - -Normalization is applied to: -- **PyTorch tensors** (format: `(channel, height, width)`) -- **NumPy arrays** (format: `(height, width, channel)`) - -Normalization is **NOT** applied to: -- **PIL Images** (passed as-is) -- **File paths** (loaded as-is) - -### Normalization algorithm - -The normalization algorithm automatically detects the input range and applies the appropriate transformation: - -1. **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 - ```python - normalized_data = (data * 255).astype(np.uint8) - ``` - -2. **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: - ```python - normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) - ``` - -3. **For any other range**: Values are clipped to [0, 255] and converted to uint8 - ```python - normalized_data = data.clip(0, 255).astype(np.uint8) - ``` - -### Examples of normalization effects - -**Example 1: [0, 1] range data** -```python -import torch -import wandb - -# Create tensor with values in [0, 1] range -tensor_0_1 = torch.rand(3, 64, 64) # Random values between 0 and 1 - -# This will multiply all values by 255 -image = wandb.Image(tensor_0_1, caption="Normalized from [0,1] range") -``` - -**Example 2: [-1, 1] range data** -```python -import torch -import wandb - -# Create tensor with values in [-1, 1] range -tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # Random values between -1 and 1 - -# This will rescale: -1 → 0, 0 → 127.5, 1 → 255 -image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") -``` - -**Example 3: Avoiding normalization with PIL Images** -```python -import torch -from PIL import Image as PILImage -import wandb - -# Create tensor with values in [0, 1] range -tensor_0_1 = torch.rand(3, 64, 64) - -# Convert to PIL Image to avoid normalization -pil_image = PILImage.fromarray((tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')) -image = wandb.Image(pil_image, caption="No normalization applied") -``` - -**Example 4: Using normalize=False** -```python -import torch -import wandb - -# Create tensor with values in [0, 1] range -tensor_0_1 = torch.rand(3, 64, 64) - -# Disable normalization - values will be clipped to [0, 255] -image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") -``` - -### Best practices - -1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging -2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` -3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) -4. **For precise control**: Use PIL Images when you need exact pixel values - -### Common issues - -- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter -- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information -- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results - ## Image overlays diff --git a/content/en/ref/python/sdk/data-types/Image.md b/content/en/ref/python/sdk/data-types/Image.md index a7f4bfea11..ca500ef3c7 100644 --- a/content/en/ref/python/sdk/data-types/Image.md +++ b/content/en/ref/python/sdk/data-types/Image.md @@ -105,22 +105,7 @@ with wandb.init() as run: ## Image normalization -When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. - -**Normalization is applied to:** -- PyTorch tensors (format: `(channel, height, width)`) -- NumPy arrays (format: `(height, width, channel)`) - -**Normalization is NOT applied to:** -- PIL Images (passed as-is) -- File paths (loaded as-is) - -**Normalization algorithm:** -- [0, 1] range: values are multiplied by 255 -- [-1, 1] range: values are rescaled using `255 * 0.5 * (data + 1)` -- Other ranges: values are clipped to [0, 255] - -For detailed examples and best practices, see the [Image normalization section]({{< relref "/guides/models/track/log/media.md#image-normalization" >}}) in the media logging guide. +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. For detailed information about normalization behavior, examples, and best practices, see the [Image Normalization Guide]({{< relref "/guides/models/track/log/image-normalization.md" >}}). --- From eb9f9a52dcb8e38ec8e930f34e26e589efef0e17 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:28:35 -0700 Subject: [PATCH 13/31] docs: add visual contrast effects and guidance on when to use different normalization approaches --- .../models/track/log/image-normalization.md | 29 +++++++++++++++++++ 1 file changed, 29 insertions(+) diff --git a/content/en/guides/models/track/log/image-normalization.md b/content/en/guides/models/track/log/image-normalization.md index dba676f60d..276941ac86 100644 --- a/content/en/guides/models/track/log/image-normalization.md +++ b/content/en/guides/models/track/log/image-normalization.md @@ -64,6 +64,13 @@ tensor_neg1_1 = torch.rand(3, 64, 64) * 2 - 1 # Random values between -1 and 1 image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") ``` +**Note on visual contrast**: The [-1, 1] normalization creates higher visual contrast compared to [0, 1] normalization. This is because: +- Negative values (like -0.8) become very dark (around 25) +- Positive values (like 0.8) become very bright (around 230) +- Values near 0 become mid-gray (127.5) + +This "stretches" the visual range, making differences between pixel values more pronounced. This is particularly useful for highlighting subtle patterns in machine learning data, but if you want less contrast, consider preprocessing your data to a [0, 1] range before logging. + ### Example 3: Avoiding normalization with PIL Images ```python @@ -92,12 +99,34 @@ tensor_0_1 = torch.rand(3, 64, 64) image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled") ``` +## When to use different approaches + +### Use PIL conversion when: +- You want complete control over pixel values +- You need custom preprocessing (filters, brightness adjustments, etc.) +- You want to use PIL's image processing capabilities +- You're debugging and want to see exact values being logged + +### Use normalize=False when: +- You want to see raw tensor values as they are +- Your data is already in the correct range (like [0, 255] integers) +- You're debugging normalization issues +- Quick testing without additional processing steps + +### Use automatic normalization when: +- You want consistent behavior across different input types +- Your data is in standard ranges ([0, 1] or [-1, 1]) +- You want the system to handle the conversion automatically + ## Best practices 1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging 2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` 3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255]) 4. **For precise control**: Use PIL Images when you need exact pixel values +5. **For highlighting subtle patterns**: Use [-1, 1] normalization to increase visual contrast +6. **For natural-looking images**: Use [0, 1] normalization or preprocess to [0, 255] range +7. **For custom processing**: Use PIL conversion when you need to apply filters or adjustments ## Common issues From bf500a736129d2732aa07b12c08c0c55e37298ce Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:28:43 -0700 Subject: [PATCH 14/31] docs: add improved image normalization demo notebook with detailed explanations --- wandb_image_normalization_demo.ipynb | 74 ++++++++++++++++++++-------- 1 file changed, 54 insertions(+), 20 deletions(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index 7ccb821afd..6892f9d57a 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -2,7 +2,9 @@ "cells": [ { "cell_type": "markdown", - "metadata": {"id": "title"}, + "metadata": { + "id": "title" + }, "source": [ "# W&B Image Normalization Demo\n", "\n", @@ -17,7 +19,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "setup"}, + "metadata": { + "id": "setup" + }, "source": [ "## Setup\n", "\n", @@ -27,7 +31,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "install_deps"}, + "metadata": { + "id": "install_deps" + }, "outputs": [], "source": [ "# Install required packages\n", @@ -37,7 +43,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "imports"}, + "metadata": { + "id": "imports" + }, "outputs": [], "source": [ "import wandb\n", @@ -53,7 +61,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "init_wandb"}, + "metadata": { + "id": "init_wandb" + }, "source": [ "## Initialize W&B\n", "\n", @@ -63,7 +73,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "wandb_init"}, + "metadata": { + "id": "wandb_init" + }, "outputs": [], "source": [ "# Initialize W&B run\n", @@ -78,7 +90,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "overview"}, + "metadata": { + "id": "overview" + }, "source": [ "## Understanding Image Normalization\n", "\n", @@ -100,7 +114,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "example_1"}, + "metadata": { + "id": "example_1" + }, "source": [ "## Example 1: [0, 1] Range Data\n", "\n", @@ -110,7 +126,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "example_0_1_range"}, + "metadata": { + "id": "example_0_1_range" + }, "outputs": [], "source": [ "# Create test data in [0, 1] range\n", @@ -141,7 +159,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "example_2"}, + "metadata": { + "id": "example_2" + }, "source": [ "## Example 2: [-1, 1] Range Data\n", "\n", @@ -151,7 +171,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "example_neg1_1_range"}, + "metadata": { + "id": "example_neg1_1_range" + }, "outputs": [], "source": [ "# Create test data in [-1, 1] range\n", @@ -166,7 +188,7 @@ "# Visualize the original data\n", "plt.figure(figsize=(8, 6))\n", "plt.imshow(data_neg1_1, cmap='RdBu_r')\n", - "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1→0, 0→127.5, 1→255')\n", + "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1\u21920, 0\u2192127.5, 1\u2192255')\n", "plt.colorbar()\n", "plt.axis('off')\n", "plt.show()\n", @@ -182,7 +204,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "example_3"}, + "metadata": { + "id": "example_3" + }, "source": [ "## Example 3: Avoiding Normalization with PIL Images\n", "\n", @@ -192,7 +216,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "example_pil_avoid_normalization"}, + "metadata": { + "id": "example_pil_avoid_normalization" + }, "outputs": [], "source": [ "# Create tensor with values in [0, 1] range\n", @@ -224,7 +250,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "example_4"}, + "metadata": { + "id": "example_4" + }, "source": [ "## Example 4: Using normalize=False\n", "\n", @@ -234,7 +262,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "example_normalize_false"}, + "metadata": { + "id": "example_normalize_false" + }, "outputs": [], "source": [ "# Create tensor with values in [0, 1] range\n", @@ -264,7 +294,9 @@ }, { "cell_type": "markdown", - "metadata": {"id": "best_practices"}, + "metadata": { + "id": "best_practices" + }, "source": [ "## Best Practices\n", "\n", @@ -284,12 +316,14 @@ { "cell_type": "code", "execution_count": null, - "metadata": {"id": "finish"}, + "metadata": { + "id": "finish" + }, "outputs": [], "source": [ "# Finish the W&B run\n", "wandb.finish()\n", - "print(\"✅ Demo completed! Check your W&B dashboard to see all the logged images.\")" + "print(\"\u2705 Demo completed! Check your W&B dashboard to see all the logged images.\")" ] } ], @@ -319,4 +353,4 @@ }, "nbformat": 4, "nbformat_minor": 4 -} +} \ No newline at end of file From 6804a5c8f6a1cfa98bb3f107104387cce3d02f18 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:34:50 -0700 Subject: [PATCH 15/31] docs: update notebook with detailed explanations for all examples --- wandb_image_normalization_demo.ipynb | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index 6892f9d57a..cb2767f697 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -120,7 +120,7 @@ "source": [ "## Example 1: [0, 1] Range Data\n", "\n", - "When your tensor/array values are in the [0, 1] range, `wandb.Image` will multiply all values by 255." + "When your tensor/array values are in the [0, 1] range, `wandb.Image` will multiply all values by 255.\nThis example creates a 64x64 pixel image with three color channels (RGB) and random values for each pixel between 0 and 1. It then converts the image from a NumPy array to a PyTorch tensor, changing the format from (height, width, channels) to (channels, height, width) which is what PyTorch expects.\n\nThe `wandb.Image(tensor_0_1)` function automatically:\n1. **Detects** that your values are in the [0, 1] range\n2. **Multiplies every value by 255** to convert to [0, 255] range\n3. **Converts to uint8** (8-bit integers, which is standard for images)\n\nThis ensures your image displays with the correct brightness and colors, since most image viewers expect values in the [0, 255] range." ] }, { @@ -165,7 +165,7 @@ "source": [ "## Example 2: [-1, 1] Range Data\n", "\n", - "When your tensor/array values are in the [-1, 1] range, `wandb.Image` will rescale them using the formula: `255 * 0.5 * (data + 1)`" + "When your tensor/array values are in the [-1, 1] range, `wandb.Image` will rescale them using the formula: `255 * 0.5 * (data + 1)`\nThis example creates a 64x64 pixel image with three color channels and random values for each pixel between -1 and 1. This range is common in machine learning when using normalized data or when working with models that output values in this range.\n\nThe `wandb.Image(tensor_neg1_1)` function automatically:\n1. **Detects** that your values are in the [-1, 1] range\n2. **Rescales using the formula**: `255 * 0.5 * (data + 1)`\n3. **Converts to uint8** for proper image display\n\nThis rescaling maps:\n- -1 \u2192 0 (black)\n- 0 \u2192 127.5 (gray)\n- 1 \u2192 255 (white)\n\nThis ensures that negative values are properly handled and your image displays correctly.\n\n**Note on visual contrast**: The [-1, 1] normalization creates higher visual contrast compared to [0, 1] normalization. This is because:\n- Negative values (like -0.8) become very dark (around 25)\n- Positive values (like 0.8) become very bright (around 230)\n- Values near 0 become mid-gray (127.5)\n\nThis \"stretches\" the visual range, making differences between pixel values more pronounced. If you want less contrast, you might consider preprocessing your data to a [0, 1] range before logging." ] }, { @@ -210,7 +210,7 @@ "source": [ "## Example 3: Avoiding Normalization with PIL Images\n", "\n", - "To avoid normalization, you can convert your tensors to PIL Images before passing them to `wandb.Image`." + "To avoid normalization, you can convert your tensors to PIL Images before passing them to `wandb.Image`.\nThis example shows how to prevent automatic normalization by converting your PyTorch tensor to a PIL Image first. This is useful when you want to control exactly how your pixel values are processed.\n\nThe process involves:\n1. **Creating a tensor** with values in [0, 1] range\n2. **Converting to NumPy array** and permuting dimensions back to (height, width, channels)\n3. **Multiplying by 255** manually to convert to [0, 255] range\n4. **Converting to uint8** for proper image format\n5. **Creating a PIL Image** from the processed array\n\nWhen you pass a PIL Image to `wandb.Image`, it is passed through without any normalization, giving you complete control over the pixel values.\n\n**When to use PIL conversion vs normalize=False:**\n\n**Use PIL conversion when:**\n- You want complete control over pixel values\n- You need custom preprocessing (filters, brightness adjustments, etc.)\n- You want to use PIL's image processing capabilities\n- You're debugging and want to see exact values being logged\n\n**Use normalize=False when:**\n- You want to see raw tensor values as they are\n- Your data is already in the correct range (like [0, 255] integers)\n- You're debugging normalization issues\n- Quick testing without additional processing steps" ] }, { @@ -256,7 +256,7 @@ "source": [ "## Example 4: Using normalize=False\n", "\n", - "You can also disable normalization by setting `normalize=False`. Values will be clipped to [0, 255]." + "You can also disable normalization by setting `normalize=False`. Values will be clipped to [0, 255].\nThis example demonstrates how to disable automatic normalization using the `normalize=False` parameter. This is useful for debugging or when you want to see the raw values of your tensor.\n\nWhen `normalize=False` is set:\n1. **No multiplication by 255** occurs\n2. **Values are clipped** to the [0, 255] range (values below 0 become 0, values above 255 become 255)\n3. **Values are converted to uint8** for image display\n\nThis means that if your tensor has values in [0, 1] range, they will be treated as if they were already in [0, 255] range, which will make your image appear very dark since 0.5 becomes 0.5 out of 255 (almost black)." ] }, { From 6c5b58bf587304032c6e53ce458eee828239ad4f Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:39:34 -0700 Subject: [PATCH 16/31] docs: use --quiet --upgrade flags for pip install to reduce output noise --- wandb_image_normalization_demo.ipynb | 74 +++++++++++++++++++++++++--- 1 file changed, 67 insertions(+), 7 deletions(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index cb2767f697..e6c5461230 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -120,7 +120,15 @@ "source": [ "## Example 1: [0, 1] Range Data\n", "\n", - "When your tensor/array values are in the [0, 1] range, `wandb.Image` will multiply all values by 255.\nThis example creates a 64x64 pixel image with three color channels (RGB) and random values for each pixel between 0 and 1. It then converts the image from a NumPy array to a PyTorch tensor, changing the format from (height, width, channels) to (channels, height, width) which is what PyTorch expects.\n\nThe `wandb.Image(tensor_0_1)` function automatically:\n1. **Detects** that your values are in the [0, 1] range\n2. **Multiplies every value by 255** to convert to [0, 255] range\n3. **Converts to uint8** (8-bit integers, which is standard for images)\n\nThis ensures your image displays with the correct brightness and colors, since most image viewers expect values in the [0, 255] range." + "When your tensor/array values are in the [0, 1] range, `wandb.Image` will multiply all values by 255.\n", + "This example creates a 64x64 pixel image with three color channels (RGB) and random values for each pixel between 0 and 1. It then converts the image from a NumPy array to a PyTorch tensor, changing the format from (height, width, channels) to (channels, height, width) which is what PyTorch expects.\n", + "\n", + "The `wandb.Image(tensor_0_1)` function automatically:\n", + "1. **Detects** that your values are in the [0, 1] range\n", + "2. **Multiplies every value by 255** to convert to [0, 255] range\n", + "3. **Converts to uint8** (8-bit integers, which is standard for images)\n", + "\n", + "This ensures your image displays with the correct brightness and colors, since most image viewers expect values in the [0, 255] range." ] }, { @@ -165,7 +173,27 @@ "source": [ "## Example 2: [-1, 1] Range Data\n", "\n", - "When your tensor/array values are in the [-1, 1] range, `wandb.Image` will rescale them using the formula: `255 * 0.5 * (data + 1)`\nThis example creates a 64x64 pixel image with three color channels and random values for each pixel between -1 and 1. This range is common in machine learning when using normalized data or when working with models that output values in this range.\n\nThe `wandb.Image(tensor_neg1_1)` function automatically:\n1. **Detects** that your values are in the [-1, 1] range\n2. **Rescales using the formula**: `255 * 0.5 * (data + 1)`\n3. **Converts to uint8** for proper image display\n\nThis rescaling maps:\n- -1 \u2192 0 (black)\n- 0 \u2192 127.5 (gray)\n- 1 \u2192 255 (white)\n\nThis ensures that negative values are properly handled and your image displays correctly.\n\n**Note on visual contrast**: The [-1, 1] normalization creates higher visual contrast compared to [0, 1] normalization. This is because:\n- Negative values (like -0.8) become very dark (around 25)\n- Positive values (like 0.8) become very bright (around 230)\n- Values near 0 become mid-gray (127.5)\n\nThis \"stretches\" the visual range, making differences between pixel values more pronounced. If you want less contrast, you might consider preprocessing your data to a [0, 1] range before logging." + "When your tensor/array values are in the [-1, 1] range, `wandb.Image` will rescale them using the formula: `255 * 0.5 * (data + 1)`\n", + "This example creates a 64x64 pixel image with three color channels and random values for each pixel between -1 and 1. This range is common in machine learning when using normalized data or when working with models that output values in this range.\n", + "\n", + "The `wandb.Image(tensor_neg1_1)` function automatically:\n", + "1. **Detects** that your values are in the [-1, 1] range\n", + "2. **Rescales using the formula**: `255 * 0.5 * (data + 1)`\n", + "3. **Converts to uint8** for proper image display\n", + "\n", + "This rescaling maps:\n", + "- -1 → 0 (black)\n", + "- 0 → 127.5 (gray)\n", + "- 1 → 255 (white)\n", + "\n", + "This ensures that negative values are properly handled and your image displays correctly.\n", + "\n", + "**Note on visual contrast**: The [-1, 1] normalization creates higher visual contrast compared to [0, 1] normalization. This is because:\n", + "- Negative values (like -0.8) become very dark (around 25)\n", + "- Positive values (like 0.8) become very bright (around 230)\n", + "- Values near 0 become mid-gray (127.5)\n", + "\n", + "This \"stretches\" the visual range, making differences between pixel values more pronounced. If you want less contrast, you might consider preprocessing your data to a [0, 1] range before logging." ] }, { @@ -188,7 +216,7 @@ "# Visualize the original data\n", "plt.figure(figsize=(8, 6))\n", "plt.imshow(data_neg1_1, cmap='RdBu_r')\n", - "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1\u21920, 0\u2192127.5, 1\u2192255')\n", + "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1→0, 0→127.5, 1→255')\n", "plt.colorbar()\n", "plt.axis('off')\n", "plt.show()\n", @@ -210,7 +238,31 @@ "source": [ "## Example 3: Avoiding Normalization with PIL Images\n", "\n", - "To avoid normalization, you can convert your tensors to PIL Images before passing them to `wandb.Image`.\nThis example shows how to prevent automatic normalization by converting your PyTorch tensor to a PIL Image first. This is useful when you want to control exactly how your pixel values are processed.\n\nThe process involves:\n1. **Creating a tensor** with values in [0, 1] range\n2. **Converting to NumPy array** and permuting dimensions back to (height, width, channels)\n3. **Multiplying by 255** manually to convert to [0, 255] range\n4. **Converting to uint8** for proper image format\n5. **Creating a PIL Image** from the processed array\n\nWhen you pass a PIL Image to `wandb.Image`, it is passed through without any normalization, giving you complete control over the pixel values.\n\n**When to use PIL conversion vs normalize=False:**\n\n**Use PIL conversion when:**\n- You want complete control over pixel values\n- You need custom preprocessing (filters, brightness adjustments, etc.)\n- You want to use PIL's image processing capabilities\n- You're debugging and want to see exact values being logged\n\n**Use normalize=False when:**\n- You want to see raw tensor values as they are\n- Your data is already in the correct range (like [0, 255] integers)\n- You're debugging normalization issues\n- Quick testing without additional processing steps" + "To avoid normalization, you can convert your tensors to PIL Images before passing them to `wandb.Image`.\n", + "This example shows how to prevent automatic normalization by converting your PyTorch tensor to a PIL Image first. This is useful when you want to control exactly how your pixel values are processed.\n", + "\n", + "The process involves:\n", + "1. **Creating a tensor** with values in [0, 1] range\n", + "2. **Converting to NumPy array** and permuting dimensions back to (height, width, channels)\n", + "3. **Multiplying by 255** manually to convert to [0, 255] range\n", + "4. **Converting to uint8** for proper image format\n", + "5. **Creating a PIL Image** from the processed array\n", + "\n", + "When you pass a PIL Image to `wandb.Image`, it is passed through without any normalization, giving you complete control over the pixel values.\n", + "\n", + "**When to use PIL conversion vs normalize=False:**\n", + "\n", + "**Use PIL conversion when:**\n", + "- You want complete control over pixel values\n", + "- You need custom preprocessing (filters, brightness adjustments, etc.)\n", + "- You want to use PIL's image processing capabilities\n", + "- You're debugging and want to see exact values being logged\n", + "\n", + "**Use normalize=False when:**\n", + "- You want to see raw tensor values as they are\n", + "- Your data is already in the correct range (like [0, 255] integers)\n", + "- You're debugging normalization issues\n", + "- Quick testing without additional processing steps" ] }, { @@ -256,7 +308,15 @@ "source": [ "## Example 4: Using normalize=False\n", "\n", - "You can also disable normalization by setting `normalize=False`. Values will be clipped to [0, 255].\nThis example demonstrates how to disable automatic normalization using the `normalize=False` parameter. This is useful for debugging or when you want to see the raw values of your tensor.\n\nWhen `normalize=False` is set:\n1. **No multiplication by 255** occurs\n2. **Values are clipped** to the [0, 255] range (values below 0 become 0, values above 255 become 255)\n3. **Values are converted to uint8** for image display\n\nThis means that if your tensor has values in [0, 1] range, they will be treated as if they were already in [0, 255] range, which will make your image appear very dark since 0.5 becomes 0.5 out of 255 (almost black)." + "You can also disable normalization by setting `normalize=False`. Values will be clipped to [0, 255].\n", + "This example demonstrates how to disable automatic normalization using the `normalize=False` parameter. This is useful for debugging or when you want to see the raw values of your tensor.\n", + "\n", + "When `normalize=False` is set:\n", + "1. **No multiplication by 255** occurs\n", + "2. **Values are clipped** to the [0, 255] range (values below 0 become 0, values above 255 become 255)\n", + "3. **Values are converted to uint8** for image display\n", + "\n", + "This means that if your tensor has values in [0, 1] range, they will be treated as if they were already in [0, 255] range, which will make your image appear very dark since 0.5 becomes 0.5 out of 255 (almost black)." ] }, { @@ -323,7 +383,7 @@ "source": [ "# Finish the W&B run\n", "wandb.finish()\n", - "print(\"\u2705 Demo completed! Check your W&B dashboard to see all the logged images.\")" + "print(\"✅ Demo completed! Check your W&B dashboard to see all the logged images.\")" ] } ], @@ -353,4 +413,4 @@ }, "nbformat": 4, "nbformat_minor": 4 -} \ No newline at end of file +} From 3bb8f8963385a5098f7f03371c634d483f8f258a Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:44:08 -0700 Subject: [PATCH 17/31] docs: add reassuring note about import cell success --- wandb_image_normalization_demo.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index e6c5461230..62d79e869c 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -67,7 +67,7 @@ "source": [ "## Initialize W&B\n", "\n", - "Let's start a W&B run to log our examples." + "Let.s start a WLet's start a W&B run to log our examples.B run to log our examples.\n\n> **Note**: The previous cell imported all required libraries. If you see no output, that means the imports were successful!" ] }, { From 6b427465e3687d7e12a3bf973ea3875926f91f85 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:47:49 -0700 Subject: [PATCH 18/31] docs: add explanation about expected deprecation warning in Example 2 --- wandb_image_normalization_demo.ipynb | 32 +++++----------------------- 1 file changed, 5 insertions(+), 27 deletions(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index 62d79e869c..f3ee1983de 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -67,7 +67,7 @@ "source": [ "## Initialize W&B\n", "\n", - "Let.s start a WLet's start a W&B run to log our examples.B run to log our examples.\n\n> **Note**: The previous cell imported all required libraries. If you see no output, that means the imports were successful!" + "Let's start a W&B run to log our examples." ] }, { @@ -171,29 +171,7 @@ "id": "example_2" }, "source": [ - "## Example 2: [-1, 1] Range Data\n", - "\n", - "When your tensor/array values are in the [-1, 1] range, `wandb.Image` will rescale them using the formula: `255 * 0.5 * (data + 1)`\n", - "This example creates a 64x64 pixel image with three color channels and random values for each pixel between -1 and 1. This range is common in machine learning when using normalized data or when working with models that output values in this range.\n", - "\n", - "The `wandb.Image(tensor_neg1_1)` function automatically:\n", - "1. **Detects** that your values are in the [-1, 1] range\n", - "2. **Rescales using the formula**: `255 * 0.5 * (data + 1)`\n", - "3. **Converts to uint8** for proper image display\n", - "\n", - "This rescaling maps:\n", - "- -1 → 0 (black)\n", - "- 0 → 127.5 (gray)\n", - "- 1 → 255 (white)\n", - "\n", - "This ensures that negative values are properly handled and your image displays correctly.\n", - "\n", - "**Note on visual contrast**: The [-1, 1] normalization creates higher visual contrast compared to [0, 1] normalization. This is because:\n", - "- Negative values (like -0.8) become very dark (around 25)\n", - "- Positive values (like 0.8) become very bright (around 230)\n", - "- Values near 0 become mid-gray (127.5)\n", - "\n", - "This \"stretches\" the visual range, making differences between pixel values more pronounced. If you want less contrast, you might consider preprocessing your data to a [0, 1] range before logging." + "## Example 2: [-1, 1] Range Data\n\nThis example demonstrates how `wandb.Image` handles data in the [-1, 1] range, which is common in machine learning frameworks like PyTorch when using normalized data.\n\n**What this example shows:**\n- Creates a 64x64 pixel image with random values in the [-1, 1] range\n- Converts from NumPy array to PyTorch tensor with shape (3, 64, 64)\n- Shows how `wandb.Image` automatically normalizes this data to [0, 255] range\n- Demonstrates the visual effect of this normalization\n\n**Note on visual contrast:** When data in the [-1, 1] range is normalized to [0, 255], it increases the visual contrast between different pixel values. This is because the normalization process stretches the data across the full brightness range, making subtle differences more visible.\n\n**Expected warning:** You may see a deprecation warning about data normalization. This is expected when passing [-1, 1] range data and demonstrates the current normalization behavior. The warning indicates that this automatic normalization will change in future versions of wandb." ] }, { @@ -216,7 +194,7 @@ "# Visualize the original data\n", "plt.figure(figsize=(8, 6))\n", "plt.imshow(data_neg1_1, cmap='RdBu_r')\n", - "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1→0, 0→127.5, 1→255')\n", + "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1\u21920, 0\u2192127.5, 1\u2192255')\n", "plt.colorbar()\n", "plt.axis('off')\n", "plt.show()\n", @@ -383,7 +361,7 @@ "source": [ "# Finish the W&B run\n", "wandb.finish()\n", - "print(\"✅ Demo completed! Check your W&B dashboard to see all the logged images.\")" + "print(\"\u2705 Demo completed! Check your W&B dashboard to see all the logged images.\")" ] } ], @@ -413,4 +391,4 @@ }, "nbformat": 4, "nbformat_minor": 4 -} +} \ No newline at end of file From 59d90c7a26b1da31cb343de35558352cdb4febad Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:55:52 -0700 Subject: [PATCH 19/31] docs: add reassuring note about import cell success --- wandb_image_normalization_demo.ipynb | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index f3ee1983de..f33bda7afb 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -65,9 +65,7 @@ "id": "init_wandb" }, "source": [ - "## Initialize W&B\n", - "\n", - "Let's start a W&B run to log our examples." + "## Initialize W&B\n\nLet's start a W&B run to log our examples.\n\n> **Note**: The previous cell imported all required libraries. If you see no output, that means the imports were successful!" ] }, { From c0bab04ed4d02c6f7dc4191f4f7e14d1b1fd22a8 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 13:57:09 -0700 Subject: [PATCH 20/31] docs: restore --quiet --upgrade flags to pip install command --- wandb_image_normalization_demo.ipynb | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index f33bda7afb..2b74479601 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -36,8 +36,7 @@ }, "outputs": [], "source": [ - "# Install required packages\n", - "!pip install wandb torch torchvision pillow matplotlib numpy" + "# Install required packages\n!pip install --quiet --upgrade wandb torch torchvision pillow matplotlib numpy" ] }, { From c1f8928e20270bbe3d8d557258c27964fe8f61f5 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:03:19 -0700 Subject: [PATCH 21/31] docs: remove --upgrade flag from pip install to avoid dependency conflicts --- wandb_image_normalization_demo.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb index 2b74479601..0c08747dc4 100644 --- a/wandb_image_normalization_demo.ipynb +++ b/wandb_image_normalization_demo.ipynb @@ -36,7 +36,7 @@ }, "outputs": [], "source": [ - "# Install required packages\n!pip install --quiet --upgrade wandb torch torchvision pillow matplotlib numpy" + "# Install required packages\n!pip install --quiet wandb torch torchvision pillow matplotlib numpy" ] }, { From 6b14efae3d92ede9ae8545fd93bee8144aa68f38 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:15:08 -0700 Subject: [PATCH 22/31] docs: remove notebook (moved to examples repo PR #609) --- wandb_image_normalization_demo.ipynb | 391 --------------------------- 1 file changed, 391 deletions(-) delete mode 100644 wandb_image_normalization_demo.ipynb diff --git a/wandb_image_normalization_demo.ipynb b/wandb_image_normalization_demo.ipynb deleted file mode 100644 index 0c08747dc4..0000000000 --- a/wandb_image_normalization_demo.ipynb +++ /dev/null @@ -1,391 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": { - "id": "title" - }, - "source": [ - "# W&B Image Normalization Demo\n", - "\n", - "This notebook demonstrates how `wandb.Image` automatically normalizes different types of image data and how to control this behavior.\n", - "\n", - "## What you'll learn:\n", - "- How `wandb.Image` normalizes PyTorch tensors and NumPy arrays\n", - "- When normalization is applied vs when it's not\n", - "- How to avoid unwanted normalization\n", - "- Best practices for image logging" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "setup" - }, - "source": [ - "## Setup\n", - "\n", - "First, let's install the required dependencies and import the necessary libraries." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "install_deps" - }, - "outputs": [], - "source": [ - "# Install required packages\n!pip install --quiet wandb torch torchvision pillow matplotlib numpy" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "imports" - }, - "outputs": [], - "source": [ - "import wandb\n", - "import torch\n", - "import numpy as np\n", - "from PIL import Image as PILImage\n", - "import matplotlib.pyplot as plt\n", - "\n", - "# Set up matplotlib for better visualization\n", - "plt.rcParams['figure.figsize'] = (12, 8)\n", - "plt.rcParams['font.size'] = 10" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "init_wandb" - }, - "source": [ - "## Initialize W&B\n\nLet's start a W&B run to log our examples.\n\n> **Note**: The previous cell imported all required libraries. If you see no output, that means the imports were successful!" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "wandb_init" - }, - "outputs": [], - "source": [ - "# Initialize W&B run\n", - "run = wandb.init(\n", - " project=\"image-normalization-demo\",\n", - " name=\"normalization-examples\",\n", - " config={\n", - " \"description\": \"Demonstrating wandb.Image normalization behavior\"\n", - " }\n", - ")" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "overview" - }, - "source": [ - "## Understanding Image Normalization\n", - "\n", - "When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`.\n", - "\n", - "**Normalization is applied to:**\n", - "- PyTorch tensors (format: `(channel, height, width)`)\n", - "- NumPy arrays (format: `(height, width, channel)`)\n", - "\n", - "**Normalization is NOT applied to:**\n", - "- PIL Images (passed as-is)\n", - "- File paths (loaded as-is)\n", - "\n", - "**Normalization algorithm:**\n", - "- [0, 1] range: values are multiplied by 255\n", - "- [-1, 1] range: values are rescaled using `255 * 0.5 * (data + 1)`\n", - "- Other ranges: values are clipped to [0, 255]" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "example_1" - }, - "source": [ - "## Example 1: [0, 1] Range Data\n", - "\n", - "When your tensor/array values are in the [0, 1] range, `wandb.Image` will multiply all values by 255.\n", - "This example creates a 64x64 pixel image with three color channels (RGB) and random values for each pixel between 0 and 1. It then converts the image from a NumPy array to a PyTorch tensor, changing the format from (height, width, channels) to (channels, height, width) which is what PyTorch expects.\n", - "\n", - "The `wandb.Image(tensor_0_1)` function automatically:\n", - "1. **Detects** that your values are in the [0, 1] range\n", - "2. **Multiplies every value by 255** to convert to [0, 255] range\n", - "3. **Converts to uint8** (8-bit integers, which is standard for images)\n", - "\n", - "This ensures your image displays with the correct brightness and colors, since most image viewers expect values in the [0, 255] range." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "example_0_1_range" - }, - "outputs": [], - "source": [ - "# Create test data in [0, 1] range\n", - "data_0_1 = np.random.rand(64, 64, 3)\n", - "print(f\"Original data range: [{data_0_1.min():.3f}, {data_0_1.max():.3f}]\")\n", - "\n", - "# Convert to PyTorch tensor (channel, height, width format)\n", - "tensor_0_1 = torch.from_numpy(data_0_1).permute(2, 0, 1).float()\n", - "print(f\"Tensor shape: {tensor_0_1.shape}\")\n", - "print(f\"Tensor range: [{tensor_0_1.min():.3f}, {tensor_0_1.max():.3f}]\")\n", - "\n", - "# Visualize the original data\n", - "plt.figure(figsize=(8, 6))\n", - "plt.imshow(data_0_1)\n", - "plt.title(f'[0, 1] Range Data\\nValues will be multiplied by 255')\n", - "plt.colorbar()\n", - "plt.axis('off')\n", - "plt.show()\n", - "\n", - "# Log to W&B\n", - "wandb.log({\n", - " \"example_0_1_range\": wandb.Image(\n", - " tensor_0_1,\n", - " caption=\"[0, 1] range tensor - values will be multiplied by 255\"\n", - " )\n", - "})" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "example_2" - }, - "source": [ - "## Example 2: [-1, 1] Range Data\n\nThis example demonstrates how `wandb.Image` handles data in the [-1, 1] range, which is common in machine learning frameworks like PyTorch when using normalized data.\n\n**What this example shows:**\n- Creates a 64x64 pixel image with random values in the [-1, 1] range\n- Converts from NumPy array to PyTorch tensor with shape (3, 64, 64)\n- Shows how `wandb.Image` automatically normalizes this data to [0, 255] range\n- Demonstrates the visual effect of this normalization\n\n**Note on visual contrast:** When data in the [-1, 1] range is normalized to [0, 255], it increases the visual contrast between different pixel values. This is because the normalization process stretches the data across the full brightness range, making subtle differences more visible.\n\n**Expected warning:** You may see a deprecation warning about data normalization. This is expected when passing [-1, 1] range data and demonstrates the current normalization behavior. The warning indicates that this automatic normalization will change in future versions of wandb." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "example_neg1_1_range" - }, - "outputs": [], - "source": [ - "# Create test data in [-1, 1] range\n", - "data_neg1_1 = np.random.rand(64, 64, 3) * 2 - 1\n", - "print(f\"Original data range: [{data_neg1_1.min():.3f}, {data_neg1_1.max():.3f}]\")\n", - "\n", - "# Convert to PyTorch tensor\n", - "tensor_neg1_1 = torch.from_numpy(data_neg1_1).permute(2, 0, 1).float()\n", - "print(f\"Tensor shape: {tensor_neg1_1.shape}\")\n", - "print(f\"Tensor range: [{tensor_neg1_1.min():.3f}, {tensor_neg1_1.max():.3f}]\")\n", - "\n", - "# Visualize the original data\n", - "plt.figure(figsize=(8, 6))\n", - "plt.imshow(data_neg1_1, cmap='RdBu_r')\n", - "plt.title(f'[-1, 1] Range Data\\nValues will be rescaled: -1\u21920, 0\u2192127.5, 1\u2192255')\n", - "plt.colorbar()\n", - "plt.axis('off')\n", - "plt.show()\n", - "\n", - "# Log to W&B\n", - "wandb.log({\n", - " \"example_neg1_1_range\": wandb.Image(\n", - " tensor_neg1_1,\n", - " caption=\"[-1, 1] range tensor - values will be rescaled\"\n", - " )\n", - "})" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "example_3" - }, - "source": [ - "## Example 3: Avoiding Normalization with PIL Images\n", - "\n", - "To avoid normalization, you can convert your tensors to PIL Images before passing them to `wandb.Image`.\n", - "This example shows how to prevent automatic normalization by converting your PyTorch tensor to a PIL Image first. This is useful when you want to control exactly how your pixel values are processed.\n", - "\n", - "The process involves:\n", - "1. **Creating a tensor** with values in [0, 1] range\n", - "2. **Converting to NumPy array** and permuting dimensions back to (height, width, channels)\n", - "3. **Multiplying by 255** manually to convert to [0, 255] range\n", - "4. **Converting to uint8** for proper image format\n", - "5. **Creating a PIL Image** from the processed array\n", - "\n", - "When you pass a PIL Image to `wandb.Image`, it is passed through without any normalization, giving you complete control over the pixel values.\n", - "\n", - "**When to use PIL conversion vs normalize=False:**\n", - "\n", - "**Use PIL conversion when:**\n", - "- You want complete control over pixel values\n", - "- You need custom preprocessing (filters, brightness adjustments, etc.)\n", - "- You want to use PIL's image processing capabilities\n", - "- You're debugging and want to see exact values being logged\n", - "\n", - "**Use normalize=False when:**\n", - "- You want to see raw tensor values as they are\n", - "- Your data is already in the correct range (like [0, 255] integers)\n", - "- You're debugging normalization issues\n", - "- Quick testing without additional processing steps" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "example_pil_avoid_normalization" - }, - "outputs": [], - "source": [ - "# Create tensor with values in [0, 1] range\n", - "tensor_0_1 = torch.rand(3, 64, 64)\n", - "print(f\"Tensor range: [{tensor_0_1.min():.3f}, {tensor_0_1.max():.3f}]\")\n", - "\n", - "# Convert to PIL Image to avoid normalization\n", - "pil_image = PILImage.fromarray(\n", - " (tensor_0_1.permute(1, 2, 0).numpy() * 255).astype('uint8')\n", - ")\n", - "print(f\"PIL Image size: {pil_image.size}\")\n", - "print(f\"PIL Image mode: {pil_image.mode}\")\n", - "\n", - "# Visualize the PIL image\n", - "plt.figure(figsize=(8, 6))\n", - "plt.imshow(pil_image)\n", - "plt.title('PIL Image - No normalization applied')\n", - "plt.axis('off')\n", - "plt.show()\n", - "\n", - "# Log to W&B\n", - "wandb.log({\n", - " \"example_pil_no_normalization\": wandb.Image(\n", - " pil_image,\n", - " caption=\"PIL Image - no normalization applied\"\n", - " )\n", - "})" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "example_4" - }, - "source": [ - "## Example 4: Using normalize=False\n", - "\n", - "You can also disable normalization by setting `normalize=False`. Values will be clipped to [0, 255].\n", - "This example demonstrates how to disable automatic normalization using the `normalize=False` parameter. This is useful for debugging or when you want to see the raw values of your tensor.\n", - "\n", - "When `normalize=False` is set:\n", - "1. **No multiplication by 255** occurs\n", - "2. **Values are clipped** to the [0, 255] range (values below 0 become 0, values above 255 become 255)\n", - "3. **Values are converted to uint8** for image display\n", - "\n", - "This means that if your tensor has values in [0, 1] range, they will be treated as if they were already in [0, 255] range, which will make your image appear very dark since 0.5 becomes 0.5 out of 255 (almost black)." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "example_normalize_false" - }, - "outputs": [], - "source": [ - "# Create tensor with values in [0, 1] range\n", - "tensor_0_1 = torch.rand(3, 64, 64)\n", - "print(f\"Tensor range: [{tensor_0_1.min():.3f}, {tensor_0_1.max():.3f}]\")\n", - "\n", - "# Disable normalization\n", - "wandb.log({\n", - " \"example_normalize_false\": wandb.Image(\n", - " tensor_0_1,\n", - " normalize=False,\n", - " caption=\"Normalization disabled - values will be clipped to [0, 255]\"\n", - " )\n", - "})\n", - "\n", - "# Also log with normal normalization for comparison\n", - "wandb.log({\n", - " \"example_normalize_true\": wandb.Image(\n", - " tensor_0_1,\n", - " normalize=True,\n", - " caption=\"Normalization enabled - values will be multiplied by 255\"\n", - " )\n", - "})\n", - "\n", - "print(\"Logged both normalized and non-normalized versions for comparison\")" - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "best_practices" - }, - "source": [ - "## Best Practices\n", - "\n", - "Based on what we've learned, here are some best practices for working with `wandb.Image`:\n", - "\n", - "### 1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging\n", - "### 2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()`\n", - "### 3. **For debugging**: Use `normalize=False` to see the raw values (they will be clipped to [0, 255])\n", - "### 4. **For precise control**: Use PIL Images when you need exact pixel values\n", - "\n", - "### Common Issues to Watch Out For:\n", - "- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter\n", - "- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information\n", - "- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "finish" - }, - "outputs": [], - "source": [ - "# Finish the W&B run\n", - "wandb.finish()\n", - "print(\"\u2705 Demo completed! Check your W&B dashboard to see all the logged images.\")" - ] - } - ], - "metadata": { - "accelerator": "GPU", - "colab": { - "gpuType": "T4", - "provenance": [] - }, - "kernelspec": { - "display_name": "Python 3", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.8.5" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} \ No newline at end of file From 394761f8da22c90d6224c995f94a790110e0baf2 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:29:47 -0700 Subject: [PATCH 23/31] fix: correct GITHUB_TOKEN parameter name in calibreapp image actions workflow --- .github/workflows/calibreapp-image-actions.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.github/workflows/calibreapp-image-actions.yml b/.github/workflows/calibreapp-image-actions.yml index 9abfec4631..5dca04c26e 100644 --- a/.github/workflows/calibreapp-image-actions.yml +++ b/.github/workflows/calibreapp-image-actions.yml @@ -22,6 +22,6 @@ jobs: - name: Compress Images uses: calibreapp/image-actions@main with: - # The `GITHUB_TOKEN` is automatically generated by GitHub and scoped only to the repository that is currently running the action. By default, the action can’t update Pull Requests initiated from forked repositories. + # The `GITHUB_TOKEN` is automatically generated by GitHub and scoped only to the repository that is currently running the action. By default, the action can't update Pull Requests initiated from forked repositories. # See https://docs.github.com/en/actions/reference/authentication-in-a-workflow and https://help.github.com/en/articles/virtual-environments-for-github-actions#token-permissions - githubToken: ${{ secrets.GITHUB_TOKEN }} \ No newline at end of file + GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} \ No newline at end of file From 750840265d0e02eb0d37006cfea258cc498db2c8 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:45:46 -0700 Subject: [PATCH 24/31] docs: address reviewer feedback - combine sections, fix terminology, improve formatting --- .../guides/models/track/log/image-normalization.md | 14 ++++++++------ content/en/guides/models/track/log/media.md | 2 +- 2 files changed, 9 insertions(+), 7 deletions(-) diff --git a/content/en/guides/models/track/log/image-normalization.md b/content/en/guides/models/track/log/image-normalization.md index 276941ac86..a293311d87 100644 --- a/content/en/guides/models/track/log/image-normalization.md +++ b/content/en/guides/models/track/log/image-normalization.md @@ -116,9 +116,11 @@ image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled ### Use automatic normalization when: - You want consistent behavior across different input types - Your data is in standard ranges ([0, 1] or [-1, 1]) -- You want the system to handle the conversion automatically +- You want W&B to handle the conversion automatically -## Best practices +## Troubleshooting + +### Best practices 1. **For consistent results**: Pre-process your data to the expected [0, 255] range before logging 2. **To avoid normalization**: Convert tensors to PIL Images using `PILImage.fromarray()` @@ -128,11 +130,11 @@ image = wandb.Image(tensor_0_1, normalize=False, caption="Normalization disabled 6. **For natural-looking images**: Use [0, 1] normalization or preprocess to [0, 255] range 7. **For custom processing**: Use PIL conversion when you need to apply filters or adjustments -## Common issues +### Common issues and solutions -- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter -- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information -- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results +- **Unexpected brightness**: If your tensor values are in [0, 1] range, they will be multiplied by 255, making the image much brighter. **Solution**: Preprocess to [0, 255] range or use PIL conversion. +- **Data loss**: Values outside the [0, 255] range will be clipped, potentially losing information. **Solution**: Check your data range and preprocess appropriately. +- **Inconsistent behavior**: Different input types (tensor vs PIL vs file path) may produce different results. **Solution**: Use consistent input types or understand the normalization behavior for each type. ## Testing your code diff --git a/content/en/guides/models/track/log/media.md b/content/en/guides/models/track/log/media.md index 5bc6b5ae11..730b711c1c 100644 --- a/content/en/guides/models/track/log/media.md +++ b/content/en/guides/models/track/log/media.md @@ -58,7 +58,7 @@ with wandb.init(project="image-log-example") as run: run.log({"examples": images}) ``` -The system assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, the system automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image Normalization Guide]({{< relref "/guides/models/track/log/image-normalization.md" >}}). To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. +W&B assumes the image is gray scale if the last dimension is 1, RGB if it's 3, and RGBA if it's 4. If the array contains floats, W&B automatically normalizes them to integers between `0` and `255`. For detailed information about normalization with PyTorch tensors and NumPy arrays, see the [Image normalization guide]({{< relref "/guides/models/track/log/image-normalization.md" >}}). To normalize your images differently, you can specify the [`mode`](https://pillow.readthedocs.io/en/stable/handbook/concepts.html#modes) manually or supply a [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html), as described in the "Logging PIL Images" tab. {{% /tab %}} {{% tab header="Logging PIL Images" %}} For full control over the conversion of arrays to images, construct the [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html) yourself and provide it directly. From f6c3a693372e02ee4498a9db7e1bb27df9d2a236 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:52:18 -0700 Subject: [PATCH 25/31] docs: remove # characters from headings in reference documentation --- content/ja/ref/python/data-types/image.md | 30 +++++++++++------------ content/ko/ref/python/data-types/image.md | 30 +++++++++++------------ 2 files changed, 30 insertions(+), 30 deletions(-) diff --git a/content/ja/ref/python/data-types/image.md b/content/ja/ref/python/data-types/image.md index 36b96ed0aa..2513cb77fb 100644 --- a/content/ja/ref/python/data-types/image.md +++ b/content/ja/ref/python/data-types/image.md @@ -30,11 +30,11 @@ Image( 注意 : `wandb.Image` として `torch.Tensor` をログする際、画像は正規化されます。画像を正規化したくない場合は、テンソルを PIL Image に変換してください。 -## 画像の正規化 +画像の正規化 PyTorch テンソルや NumPy 配列を `wandb.Image` に渡すと、`normalize=False` を設定しない限り、ピクセル値は自動的に [0, 255] の範囲に正規化されます。この正規化は、一般的な画像データ形式を処理し、適切な表示を確保するために設計されています。 -### 正規化が適用される場合 +正規化が適用される場合 正規化は以下に適用されます： - **PyTorch テンソル** (形式: `(channel, height, width)`) @@ -44,7 +44,7 @@ PyTorch テンソルや NumPy 配列を `wandb.Image` に渡すと、`normalize= - **PIL 画像** (そのまま渡される) - **ファイルパス** (そのまま読み込まれる) -### 正規化アルゴリズム +正規化アルゴリズム 正規化アルゴリズムは入力範囲を自動的に検出し、適切な変換を適用します： @@ -63,7 +63,7 @@ PyTorch テンソルや NumPy 配列を `wandb.Image` に渡すと、`normalize= normalized_data = data.clip(0, 255).astype(np.uint8) ``` -### 正規化効果の例 +正規化効果の例 **例1: [0, 1] 範囲のデータ** ```python @@ -115,22 +115,22 @@ tensor_0_1 = torch.rand(3, 64, 64) image = wandb.Image(tensor_0_1, normalize=False, caption="正規化無効") ``` -### ベストプラクティス +ベストプラクティス 1. **一貫した結果のため**: ログする前にデータを期待される [0, 255] 範囲に前処理する 2. **正規化を回避するため**: `PILImage.fromarray()` を使用してテンソルをPIL画像に変換する 3. **デバッグのため**: `normalize=False` を使用して生の値を確認する（[0, 255] にクリップされる） 4. **精密な制御のため**: 正確なピクセル値が必要な場合はPIL画像を使用する -### よくある落とし穴 +よくある落とし穴 - **予期しない明度**: テンソル値が [0, 1] 範囲にある場合、255倍されるため画像が非常に明るくなる - **データ損失**: [0, 255] 範囲外の値はクリップされ、情報が失われる可能性がある - **一貫性のない動作**: 異なる入力タイプ（テンソル vs PIL vs ファイルパス）で異なる結果が得られる可能性がある -#### 例: +例: -### numpy 配列から wandb.Image を作成 +numpy 配列から wandb.Image を作成 ```python import numpy as np @@ -145,7 +145,7 @@ with wandb.init() as run: run.log({"examples": examples}) ``` -### PILImage から wandb.Image を作成 +PILImage から wandb.Image を作成 ```python import numpy as np @@ -164,7 +164,7 @@ with wandb.init() as run: run.log({"examples": examples}) ``` -### .png (デフォルト) ではなく .jpg をログ +.png (デフォルト) ではなく .jpg をログ ```python import numpy as np @@ -184,7 +184,7 @@ with wandb.init() as run: ## メソッド -### `all_boxes` +`all_boxes` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L610-L631) @@ -198,7 +198,7 @@ all_boxes( ) -> Union[List[Optional[dict]], bool] ``` -### `all_captions` +`all_captions` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L633-L637) @@ -209,7 +209,7 @@ all_captions( ) -> Union[bool, Sequence[Optional[str]]] ``` -### `all_masks` +`all_masks` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L587-L608) @@ -223,7 +223,7 @@ all_masks( ) -> Union[List[Optional[dict]], bool] ``` -### `guess_mode` +`guess_mode` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L474-L486) @@ -235,7 +235,7 @@ guess_mode( np.array が表している画像の種類を推測します。 -### `to_uint8` +`to_uint8` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L488-L511) diff --git a/content/ko/ref/python/data-types/image.md b/content/ko/ref/python/data-types/image.md index add78b2eef..cf9ec50f34 100644 --- a/content/ko/ref/python/data-types/image.md +++ b/content/ko/ref/python/data-types/image.md @@ -30,11 +30,11 @@ Image( 참고 : `torch.Tensor`를 `wandb.Image`로 로깅할 때 이미지는 정규화됩니다. 이미지를 정규화하지 않으려면 텐서를 PIL Image로 변환하십시오. -## 이미지 정규화 +이미지 정규화 PyTorch 텐서나 NumPy 배열을 `wandb.Image`에 전달하면 `normalize=False`를 설정하지 않는 한 픽셀 값이 자동으로 [0, 255] 범위로 정규화됩니다. 이 정규화는 일반적인 이미지 데이터 형식을 처리하고 적절한 표시를 보장하기 위해 설계되었습니다. -### 정규화가 적용되는 경우 +정규화가 적용되는 경우 정규화는 다음에 적용됩니다: - **PyTorch 텐서** (형식: `(channel, height, width)`) @@ -44,7 +44,7 @@ PyTorch 텐서나 NumPy 배열을 `wandb.Image`에 전달하면 `normalize=False - **PIL 이미지** (그대로 전달됨) - **파일 경로** (그대로 로드됨) -### 정규화 알고리즘 +정규화 알고리즘 정규화 알고리즘은 입력 범위를 자동으로 감지하고 적절한 변환을 적용합니다: @@ -63,7 +63,7 @@ PyTorch 텐서나 NumPy 배열을 `wandb.Image`에 전달하면 `normalize=False normalized_data = data.clip(0, 255).astype(np.uint8) ``` -### 정규화 효과 예시 +정규화 효과 예시 **예시 1: [0, 1] 범위 데이터** ```python @@ -115,22 +115,22 @@ tensor_0_1 = torch.rand(3, 64, 64) image = wandb.Image(tensor_0_1, normalize=False, caption="정규화 비활성화") ``` -### 모범 사례 +모범 사례 1. **일관된 결과를 위해**: 로깅하기 전에 데이터를 예상되는 [0, 255] 범위로 전처리 2. **정규화를 회피하기 위해**: `PILImage.fromarray()`를 사용하여 텐서를 PIL 이미지로 변환 3. **디버깅을 위해**: `normalize=False`를 사용하여 원시 값 확인 ([0, 255]로 클립됨) 4. **정밀한 제어를 위해**: 정확한 픽셀 값이 필요한 경우 PIL 이미지 사용 -### 일반적인 함정 +일반적인 함정 - **예상치 못한 밝기**: 텐서 값이 [0, 1] 범위에 있으면 255배되어 이미지가 매우 밝아짐 - **데이터 손실**: [0, 255] 범위를 벗어나는 값은 클립되어 정보가 손실될 수 있음 - **일관성 없는 동작**: 다른 입력 유형(텐서 vs PIL vs 파일 경로)에서 다른 결과가 나올 수 있음 -#### 예시: +예시: -### numpy array에서 wandb.Image 생성 +numpy array에서 wandb.Image 생성 ```python import numpy as np @@ -145,7 +145,7 @@ with wandb.init() as run: run.log({"examples": examples}) ``` -### PILImage에서 wandb.Image 생성 +PILImage에서 wandb.Image 생성 ```python import numpy as np @@ -164,7 +164,7 @@ with wandb.init() as run: run.log({"examples": examples}) ``` -### .png (기본값) 대신 .jpg 로깅 +.png (기본값) 대신 .jpg 로깅 ```python import numpy as np @@ -184,7 +184,7 @@ with wandb.init() as run: ## Methods -### `all_boxes` +`all_boxes` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L610-L631) @@ -198,7 +198,7 @@ all_boxes( ) -> Union[List[Optional[dict]], bool] ``` -### `all_captions` +`all_captions` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L633-L637) @@ -209,7 +209,7 @@ all_captions( ) -> Union[bool, Sequence[Optional[str]]] ``` -### `all_masks` +`all_masks` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L587-L608) @@ -223,7 +223,7 @@ all_masks( ) -> Union[List[Optional[dict]], bool] ``` -### `guess_mode` +`guess_mode` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L474-L486) @@ -235,7 +235,7 @@ guess_mode( np.array가 나타내는 이미지 유형을 추측합니다. -### `to_uint8` +`to_uint8` [View source](https://www.github.com/wandb/wandb/tree/637bddf198525810add5804059001b1b319d6ad1/wandb/sdk/data_types/image.py#L488-L511) From f5cfa3f431bc11968fc9ee0c3531ae7d45837248 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:54:42 -0700 Subject: [PATCH 26/31] Apply suggestions from code review --- content/en/guides/models/track/log/image-normalization.md | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/content/en/guides/models/track/log/image-normalization.md b/content/en/guides/models/track/log/image-normalization.md index a293311d87..44043fb052 100644 --- a/content/en/guides/models/track/log/image-normalization.md +++ b/content/en/guides/models/track/log/image-normalization.md @@ -3,9 +3,8 @@ title: "Image Normalization Guide" description: "Learn how wandb.Image handles normalization for different input types and how to control this behavior" --- -# Image Normalization Guide -When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This guide explains how normalization works and how to control it. +When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. This guide explains how image normalization works and how to control it. ## When normalization is applied @@ -71,7 +70,8 @@ image = wandb.Image(tensor_neg1_1, caption="Normalized from [-1,1] range") This "stretches" the visual range, making differences between pixel values more pronounced. This is particularly useful for highlighting subtle patterns in machine learning data, but if you want less contrast, consider preprocessing your data to a [0, 1] range before logging. -### Example 3: Avoiding normalization with PIL Images +### Example 3: Avoid normalization with PIL Images +Normalization is not applied to PIL Images. ```python import torch @@ -87,6 +87,7 @@ image = wandb.Image(pil_image, caption="No normalization applied") ``` ### Example 4: Using normalize=False +To explicitly turn off image normalization without converting the image, set `normalize=False`. ```python import torch From 7b3b451778621f61ccdf2c55847307291ed3dfed Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:55:51 -0700 Subject: [PATCH 27/31] docs: remove remaining # characters from reference documentation headings --- content/ja/ref/python/data-types/image.md | 2 +- content/ko/ref/python/data-types/image.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/content/ja/ref/python/data-types/image.md b/content/ja/ref/python/data-types/image.md index 2513cb77fb..d725996f00 100644 --- a/content/ja/ref/python/data-types/image.md +++ b/content/ja/ref/python/data-types/image.md @@ -182,7 +182,7 @@ with wandb.init() as run: | Attributes | | | :--- | :--- | -## メソッド +メソッド `all_boxes` diff --git a/content/ko/ref/python/data-types/image.md b/content/ko/ref/python/data-types/image.md index cf9ec50f34..74061550c3 100644 --- a/content/ko/ref/python/data-types/image.md +++ b/content/ko/ref/python/data-types/image.md @@ -182,7 +182,7 @@ with wandb.init() as run: | 속성 | | | :--- | :--- | -## Methods +Methods `all_boxes` From dbd68d430db0ac0375a700ff2b398c5c245dfc14 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:57:24 -0700 Subject: [PATCH 28/31] docs: remove # characters from English reference documentation headings --- content/en/ref/python/sdk/data-types/Image.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/content/en/ref/python/sdk/data-types/Image.md b/content/en/ref/python/sdk/data-types/Image.md index ca500ef3c7..f7cda75d46 100644 --- a/content/en/ref/python/sdk/data-types/Image.md +++ b/content/en/ref/python/sdk/data-types/Image.md @@ -9,10 +9,10 @@ data_type_classification: class -## class `Image` +class `Image` A class for logging images to W&B. -### method `Image.__init__` +method `Image.__init__` ```python __init__( @@ -103,13 +103,13 @@ with wandb.init() as run: run.log({"examples": examples}) ``` -## Image normalization +Image normalization When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values are automatically normalized to the range [0, 255] unless you set `normalize=False`. For detailed information about normalization behavior, examples, and best practices, see the [Image Normalization Guide]({{< relref "/guides/models/track/log/image-normalization.md" >}}). --- -### property Image.image +property Image.image From 4a45f33b7f615430a20efa9b9c6f17e6f76e5986 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 14:59:56 -0700 Subject: [PATCH 29/31] docs: improve normalization table with matrix format for better clarity --- .../guides/models/track/log/image-normalization.md | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/content/en/guides/models/track/log/image-normalization.md b/content/en/guides/models/track/log/image-normalization.md index 44043fb052..65b881044c 100644 --- a/content/en/guides/models/track/log/image-normalization.md +++ b/content/en/guides/models/track/log/image-normalization.md @@ -8,13 +8,12 @@ When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values ## When normalization is applied -Normalization is applied to: -- **PyTorch tensors** (format: `(channel, height, width)`) -- **NumPy arrays** (format: `(height, width, channel)`) - -Normalization is **NOT** applied to: -- **PIL Images** (passed as-is) -- **File paths** (loaded as-is) +| Input Type | Format | Normalization Applied | Notes | +|------------|--------|----------------------|-------| +| **PyTorch tensors** | `(channel, height, width)` | ✅ Yes | Automatically normalized to [0, 255] range | +| **NumPy arrays** | `(height, width, channel)` | ✅ Yes | Automatically normalized to [0, 255] range | +| **PIL Images** | PIL Image object | ❌ No | Passed as-is without modification | +| **File paths** | String path to image file | ❌ No | Loaded as-is without modification | ## Normalization algorithm From 0bca154bd32f6c560116770eec9fa603976f0899 Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 15:03:07 -0700 Subject: [PATCH 30/31] docs: convert numbered normalization algorithm steps to bullet points for better clarity --- .../models/track/log/image-normalization.md | 28 +++++++++---------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/content/en/guides/models/track/log/image-normalization.md b/content/en/guides/models/track/log/image-normalization.md index 65b881044c..d6bacebd03 100644 --- a/content/en/guides/models/track/log/image-normalization.md +++ b/content/en/guides/models/track/log/image-normalization.md @@ -19,20 +19,20 @@ When you pass PyTorch tensors or NumPy arrays to `wandb.Image`, the pixel values The normalization algorithm automatically detects the input range and applies the appropriate transformation: -1. **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 - ```python - normalized_data = (data * 255).astype(np.uint8) - ``` - -2. **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: - ```python - normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) - ``` - -3. **For any other range**: Values are clipped to [0, 255] and converted to uint8 - ```python - normalized_data = data.clip(0, 255).astype(np.uint8) - ``` +- **If data is in range [0, 1]**: Values are multiplied by 255 and converted to uint8 + ```python + normalized_data = (data * 255).astype(np.uint8) + ``` + +- **If data is in range [-1, 1]**: Values are rescaled to [0, 255] using: + ```python + normalized_data = (255 * 0.5 * (data + 1)).astype(np.uint8) + ``` + +- **For any other range**: Values are clipped to [0, 255] and converted to uint8 + ```python + normalized_data = data.clip(0, 255).astype(np.uint8) + ``` ## Examples of normalization effects From 90d2d1d8273396e63e14166af5dbdf5c630ebddc Mon Sep 17 00:00:00 2001 From: Matt Linville Date: Mon, 25 Aug 2025 15:04:23 -0700 Subject: [PATCH 31/31] docs: remove accidentally committed test notebook --- image_normalization_test.ipynb | 1 - 1 file changed, 1 deletion(-) delete mode 100644 image_normalization_test.ipynb diff --git a/image_normalization_test.ipynb b/image_normalization_test.ipynb deleted file mode 100644 index 0519ecba6e..0000000000 --- a/image_normalization_test.ipynb +++ /dev/null @@ -1 +0,0 @@ - \ No newline at end of file

    • RcWvl+_9Io%uy zyUraG!$pXf2Au_6XWpJ{PsI3R3U4!K!IoLCg{k~6kId5<&v$1M=e_r(S|)(96%+9X zVwA81T0W!Hd7op|bZ_}Z){slV(z!xy?o!J5R*AKO5u&_wz#%)njbWDBav`<%RRX2P z9J@}@p%P3_=G+&i*4n?bldl5Qf{^!_-5${&i5`&ee=@!*Lkv``<22lD&Bh3I)!j@n z*LO%SM>Ng9MUUC^J1y^wAd75k`JL7nyUE{I{R42i#bwo^0@8`0n}^Hm{qL!09K$;j zmL1gbP_Z?FmV<9)P4eMb@bCk$2R8s3U0n_^g=$m^cMLUMZi?HyXBt>pS@G1Z0w(d; zn6iwt&iYdSzI$MJk~8){erB)bOkdo)v5=u*))uabwoVcE1G?Pqq_v`PR-U}hE@kmk zn2Yz5@`~!&mv*yqOyC`zdS;Hk#lo^$KbO&1@H5r~>P9w32Bq`(ffe);b9sr0iYniR zPrg@iu8%%#q1pRsd(J%Y{+cu`UuA;Kx@Xkd<*p;d#T4sLTE@PHtSoCZ^YJs+&GEDb zE+(-pMsc4D{6Tu@`)r4Xwm4op1hB`qzZ2ezp}%H2y(ktHww?Mc zeX~KyTShO~nrAI&F%MJK9{WbQ%i>YE@q!ep;_8fjEUL^`kjHg8a4;v!t!=im1hCJ9tF zsiFi9ucQNqjiEngDt%E<1_ZZ~4`*FvAzzMj6JvcI6?h(Ytha-@ZDLteJ|}7Nc>nVX zzQ=g+esB|z$mKefqOwJHB{cZVA$k2te-W|VDW>+vd1hwTr3{dNd+(n%4pmK9-dz6) z!rS%Bpg`o<)7kZl_TW0q)JUn`Ketb)eEc;x-?C&=a5H?DzN3O=)0wQmFI4&l_{-*jRYwEI@E-HDa{^@vyczPUN+?;7K&almaF z)`j3|Hlot<%~Q@V_c!e*Ei0uuFj@(N2jR+s8RDwtZ?CZ+32FDC-6-BU>71-i_&*~D z7QpnYG4wleSo~9Y)h%rB#DiC8TK0j238}6<=&W)guX)PIbf-U`L>6?E$y?6N@0Rm> zpjzeIGmPb9OM6^Td=iosA-jc&uVn!zrEhCkF7qey31&Pd)tr6aPE2p_s#cA9M29QV zp}kIS;%DSu_dTt$b1{aMre8nzGnFiRUk1LtEue)N%JhBjHIts6loQT{71_pjC;yZ= zq{F8pISxE590VB+m%s7yU7oBk6LKCWJNcWJq51yQdnYi4pWrll|NhnvEUdnf^mOr$ z_`tsAV0@KD7W8VmXGt(w(wqSh;KWsp|E*CldWZJ9OdTbx3py(+a}N6c+aGd2d=1T- zp1Q#ys%O$ExfI#wm2{i+(B7-6pM9=`zTeztpBByLQ+sp1P|Np^m(MDUXteH{S3*Z? z^^oB*qj+5@N%VvCh80Uqp`VPQREGcbHBmyY+et+whx7UXx~|T7bNZ#B^rRHcuJ=WF z+-QZfZaQtY|3#5lmXYxbDQxT=bcjt$P)+dMS8ko6taTH!DNhp5-+D1u74wvyh80c; z$zKb^-prhhzB6QOoyy#zvDf}TjC1IOdRvlSF$;OjKmL4P!#V7+N@CSWCc3H!7uC_p z?#bI~HS{JA6+FGfj2SCkh{XT}JYvHGwICe^A-h-pUh9wNn^{s6t>BKZ-WQG~QK@t# z7~B3yda0oSg1JM6Qva_>PvP@@xxHh;!@Ullavq<$cikf{<{Mla-&3F}jbtLM_fn`g zSOw43%8KV2?`~f__}=Sov0paWx^PD2zjn7R)Y!vb)uySl-BWZ7)IhL2;7NRi%x8>d$xXc3w(r zsR`BQ<9;8{Jlw$W-@Q#QZ(A7J>?uxdGip8jWHfR{ZBP6f$x=J$OM8a28UGZtq|q;O z7M+ROiikR!iEQCs9*|9TQ)E_%MBu}d@al=W>KGzVepa-E9?}7KRjr(p>z&)E$jwAu z3ws@%b^Pfwj-r}!Gt4KO;bJj33=2Q#EROQLw@-4o*=f)(#a-qW_@-WF@z1;azkRb3 z+$*l*7^NwP^F6%cP~TSXiRNfE0PR}b2mqlF_Rt_K@s3L!j+e7|YBey_c3TB4pM3Fo zo7lx2P;n*H{$p}dPXbR9Cb~>F(EhWyEzYkXFfkYm23G?bsx%qu_Z75g@dtiOD?q#& z)-Q@nxbJ;j#6J%lVK?&nU$7Y;5$}1qQ#F{cSMfos%0kt7rs^l?aoci_s|*e2V=Xi%bdd@hlPl-BwP9Rrvnf;|xpFF4S6D0lD|4ViY2k2ocQ7i#U0Qnse18BcT=gQ8&?Ou!pl#>=9 z1q^&wlCWylg+JZz9h!e56H(bg0`x*wWf{e= z$4GV@o9@b*r3rRE#RXB7k}TQ)%9vJ}k(Afzm~OwgqL*g6GCOO^2@*|6&2=D2ZQ}j= zGj)H{_O>LTSrf}y<0iEYvacX3O^2m;LEfL9gEYDLv4I;~`H>`Al9<(R|S24)a zb2!}dq0M(*c^@`1PWREgQ};<8&;q&`R4>*G$0w6fEl-+6MF*lD$`#?n}`G;om`pZY=&!<)7LK8j z=7pk4vzob`x#oJw(aH3C>WyX{`>D-5rQHUte7%C#V0gpcHy%SJ2kqiM&tI-$zJ*an zZ~a%y$5UQ$YZVM!t|p>t>ZhoM$yk8Rk^W?s&ak%5;%pVB|BnTrziS<&&Y5=IreD}U zt(mG;5i(onX(25=_nP|bL$-?Js%`W$ZoOnCS(6yrizrB*S^mKxo7-ulCiHRRIjRSF zvdo|XQ$_m#3tZZ5Kb0cUiAK-b7jf1tn_Dlw%wSTs+#CEN<9Rw8arfq{x>vH;=X91b zSAZD}vl)0L=Cg`@dBt*|+!*yo<>J>~s2B-!a|MoK7G%#|65G!Anzx-VB%$Wf54^^= zqqWmpD-_OOYJA2iCMIU6;W$-EP5ju(8*R^drz8U=4!p}C{3;Hx`Q@^SpI98PfabT2 z2kGITKQB8oUoIS5+)vzU&e+Y;BHEn!Xe$_54$h4;d;+p$kdGc4dCMor%?SNRmL-Gp zsJM^6eOIqBEKH5<-jnh=^pdk!*CYo~_0QY(o0oL68oHGmmYbH-Uz~WYww`?TS|0+T zarwXtG(0+iyM=0ucYZap6bzTAR&F1h>>%Bvf=@Elw3T^v_^`HjW?MA>w2Ko{k>d?Z}+s&bx-eKM5dwA=< z(v@P5%lhWFOV!g($#bJYL7YDnkmT_H=ovF&X;4oTyu&c5u~1!hUXCr4niZ$n%~@cX zHTb;HW_i)T@3(T5|5c#INwK+D@8|U6w#aSJMv+1Dx4F826Sldu1{X91duuK{!y_F- zya$38u->VjN;amO^5pz}C#IFJ^A)2!?uR}O`3YhUnlJui%OWS&ev*M9a-5N`zgsIv z^kl*_5$wS|&FtNEZSlh?EIbM^GhL#vkWta_^r-#q4ILvxqzZq2aun&*HW{RTn{u1LQMO;F+?Ai(w;!`W^f9d6d4nOWKs9*3d zS+L~PPyPG9n6II&Q;&8nlD%y@I=QKLpmbm{Q>bm|gKnFqSI3(RzxO4_{}Wmv-r z*A+lSiKK=vWg7pqu=i>*p#V7dB&Lfr2zSJ$_!1yJq_*sgWF!mu$1vrGkutt)bl-Qq zlGru4|DAYC7px>63y5>;MV9r0`9Hke=%tr!r>nZd(Bvx&F#rB!*OMuX@tcuvCHco; zbJoY-?0MaWVnoxBwRUp@nWE!9<@~Q&%W&V6C%?^_opokzeFY54S$?;8|5|+IQfd0~ zD%v%=2j(t3^Ct2#it4>I8c6e$SH$PTP0N$df39Gvz!1EV1`@`R=uZI)Y)sNEccZln zAe*9|nvmo~5_|0p1qm}#ElBu?OHvgC7DcRFKu5x4c*nGu!gBKEn-d1pV|wP%oD&&kK`f`Dq~wGXbD zQoiEffIT_=jNF+_ED~;k6S>cIyPW+@{`sj1_F!jlsQp4PULByBB`v#iVS*O@xd2e+ zIO8}GX8;jwX#qEhy#W}ag6}=vWdPm*cvsG=eCYn1T-wc;D9K+buT4f!p{gmzN(7V~ z>gET7-{tv+qcJM7x9}Kd%=Mn;;I&?iUCCCBb)?S8>4w9m#tB8(QH3}IV zJHANz$y7%KUGkr7`(8LuzBiTs&e<%Z(0j>)h5eGOWrKE?F3l-xO%a2=P$ z`wDc$Xd1_8o{1#wCgO_zTQ9H2pMPj#291gnpN^77?VDoCH&q2@YPb!Ql&)Xr8Wu*C z2M|j;vlYg&BZA$TtaJya2{Pkf{8lMwnHhj~BbW!e`HPFN50bvNnLWIs>e_nxHqUv! zK6=LPH9!=cvOEtL4U@m`f<;8DLASNBD{ zH|A?TZ$us@J8o{K%(WI0w zPQt1}s-r7Q?C^wV$vtkk_FzUP)YTQ^=kO>e`U2%a>VIb%EG$KCTJ>J4g<2&~9D;*S zr`zR24<-ADbUvsBs(KxyblLFEYEO+;aLFLya{=#8noZxDILGRq_|JdT=yydlyC*cy zHrR(v>1yznyG}OYITP41jJ7z_GJAK98ZhqSW0nM8M?am6C0Gj;%9<@U&zzCW9?Q!3 zApgt^6>B%mxJqkVlyNLsh%-SgqW-ICUc2QvO}~yA$(8u}t0Z+Zo{nq7Yp*|&h@4(x zzXw*W9ura5z>PU#*L-k3=iTa|YHxHl27;~tnJK2yON+`TzDwv)#754AEXB7FgPQt}lv@n=1F_?O7EfH~t z&lo92-ecRAnI%cufomxup4s?Wvayl=D+u5}rO5wFGvF*P=4Y|bGLmM^&(Ak8*55G~ zqG`p%`8(b*N;I(dGnioIV|Obf<;6)DUR7ej$9*1_jUQ(LFLw)$QNY41{JSFUM)iAd zPp4b&wuAI?tnI;GfRA$g|3+`YC;>j8)YmLL8|cP6_x0~%InaPiX1VPG+rbE8pRYgp zFG~}?O6Gq7xi4X-d&lIYp$^JKL<2kEQ_&!BpzKHbsPk|b_5n_8Z|mJDUwWt2B>pDe zD$J{w<3K9a24rRSbeSF@uIQA zk%UML3l7BAFe|`50g(y%VH|vGpfYck^k1$}`DqV#84PD9mb97(Y4KmrY^x*F9Bv#Q z@-;)#@bhHuwikJ1t|5P9lv$1}($3$@ECmNEE5OL?|1m_rgun--&J7qAm1`xcQ(AA< z_=tPKxt-3~Y+R=4BuPYTM|m>0Py26&EJ2)LVrQxomQmViFp|Cc<(y8*8I~;la2UQR z=;thsy24Q_XzRjlm~7PjfNr|7b4KU9wGoH0->~VkxZiMwPRZQIj)$!>AKNHsTF<=j z3Z-41ER0@8r_}x-ixZNhf-Sq?gdTala)ZG2PF%W8jVAcIe!*G%Bkt}_RG^A-5|NCg z7=k7w;f&b&OoauN`0iq)6)6lChKt(sco%6UjnGk4Al_+YBW1{=RSztpw5_@n8jRyoc|2#kmpmKLy*8cpi}G{lI|w1SQgc_5%9(K?{L@GgK| zDlrlz{klZIRUH70&;U5>R2`5nAI2#u8TjONa-Rv2U#s1eQ#vC|JhStAqrsYu)|)X7 zBfqF>#n_Hs=O!qqEu?AjH!lbZkX3HWI_ClGbrTC2t4AFw*?OnmeJLGd&`Y$s*HCXr zVr^R6M@*h{C-=&P|CvI41ktoOwS%t-{a*fH z-a2nP_ib1FiV(?k8?`#gyomkg=x1%6uc%k8n8Z7eqC4EifNikRPSI7M@Rc5LM*Mx3 zaU>T11mlk_sC?rQPHzXMPXHq-Uuj3BHBMS%$9enUAcABBF_{oo2GY>CD|ys_qX=Pme0WQu8BD)&frngAJkzU2mMTc3;@R1;?kU2PZ#+7Qa<9uGBrSB;nB0ncn~> zW|Dt#?1A}Yg)02%Ya-}vg!J~USgm#r-T}gl(Dyv(iwJsD<_=vyNXc zXQj!XzAlS2iDE|1dEDO;51y0X;@kR!Acu!<5NJPV@$f4d9E?ECRT>w_%$+6&H zSe7!d#~)sI`xr{RmT+!z2BdjZ41mThV?e1y-ExwH7S+A=Zb>I+XKVHRevzkm=S(3h zp3fjA*z3D`5)djwHK`O6G(&Nj$vg~z26pVW^+*x%zohtUEuVU2U4q}vU)M4$Anlc}Fkq2@JH=#k^C zrP0$q(!FVl*@O|X(g?3MXPV&#pWj@klJ%(%or*`Sm6?y9Bj-TWsbZs@^)!P9Wiui! zZTlK(7pg|#Y=a_<3NbQuhkc6<`@hcZdq%hDu_t-a^Ino?%$ zWe9AM)Iv&YW0U7B-Gcbjio)8hld>RI>GZM1;+Sy5O|f)_feKs$NIAi=#VBQBRRdoS z^faUfpX1Z+!Dpe&-7LG42FFV-?f!n^atko$?}}%c{p}U^OsA>Q!uqf_v}E0!ZpRbp zgpH{{MURqEL`#iBMjhYIi?9L1F=wxL=x0wAQ=W^!uGq_rx1~RwgnWfkjBOB&?Ob(? zY#EKWAQv*#Kzk->+74FPy)fIEMZx>KNE6#KMq zlp#dAr>t&e1(SG|_|yKxWOZf|0a2UsAU(S3G)ql92|)My_>8QVoM;7fY$ZRV(`zW4 zmV%Z!5-(zxsTVo@IxYWpZF4xRrTK%B(IAs$skTl@=^uq1Z4bjXffpq{P?!bIo>{D? zU9-S~KhuBPj2nI7Niq1*>O8?d^MT*_V4m@Z6oN3j6EM1kIGkvoICjr^t5)YdRydX! zK59vS%x26~0tsLki9F~lZBg6^XxdhpK}7A(tVeq>7gFN_P^SJx2YWug~cROU>_399Q7wv+;}&U)h*#Bc`s0)uD0>v5!N8Bro4b% zQ09-_n$R!&pR%L--j|27e>fC>8C^TYi%tcDtU1X+tZk2h1`Ic(kcaWfpe51s(Ezs} z*CLtR032j^kFk3DK2UUGYWHRCV#N_dAv_UuU8 z=D?!}^OlZsThy-SSW0StRFjtGsTbzQMgWlQgW@>Ly0JFvMAu3d9;T zy?$qO-Py>G9QD^dMQ*dPaEj(o+!Be9%1T@j+Og~i^e`WP6>!jAJQbUr06mTw%6u3) zNWIgvl1k@F+S;-|dW#q!7KRJ%WDgAviIm?FUw_^+9c3|G&ug#Ds0%p`Ynhe2`N=U$X0JJ;e1p4P z86$a_O3=4_yf7#|=u2xqMfj_}-uZm@Ia{M0-JY`Vo)LPGFGXJ@A&7F>LpOqR*AjRH7(M|$mSMICZf*6n*orEcu- zZh$h85i`PTogk-7jRa%2^S_ly6BpO&7^0*2dLW@|EtkGHX)~uu7HEic6?!rYAuYZ7 zo&CS{F(CLJL6@$A16T|4e8sPaPK^p_3Cd*sG5`RIJ6)gRrPev`Z!j{KayTnZcg@r= zplnoqvmh^h8D9{XPu9Guu&9&FShGp9TUmPsDb*$0oyoW-K}lGwdIWl~{prPj2;~U=WWa15!Y%coS zgW}y|mnXGi>Na_RFy4I7LF5KV)b>_ZoB-?md9dAJj-g*`-y#}NZ$YQU6feLPn}1Bk z7dEfaeA#K)ESU~iCY8C2t#XXNnt|=4CQ3BD_Q*Gn9J0_AUoGEe$SoxOi>tb-4T`2I z$x$O_v>%|DNu!Oxps=xZQ73mC6y3ZV3ionKCmF5 zO~SroO5n+Qe8B|^X&YTeu^Og4#>12N`a-R<%Oak3_j%0zQVuzJ`CxUuS3evA-ygMA z#!-o1!k3>`N*VGSm7=K6Mnnv2Why@vB>Xvc$ZXOJ_~TPzBikq7wt%;MP%DZO*+JLh z%WfjCrjvFR2yCR|a8w+(ZJqrUmJv1(HAvgC<+orsxqx6SR)1J+idq5vk~{JC&o~Sq zq>Tt%QKOAqEcp2z{X_u!4zT~@;sP(N0Z}ipPEE9w;Gucr;4)rA^CK;=qK?>-xhdbb z>Vr0*tO#K~iUZA^iq1yl5kMh8+yDf`3qumqi=_J1!cG`vH6_O!W}^UD7`vpVKS&ss z2^QLMyIim&)Bw6tpkP!mlS#3A zbifH0pIchegg)bxQAisxQ|c?ITHR&bsQ~ELHt?yPD-SS+nC*n;u>Tcun3|qFcrkG- zW8A9mhwh3momP_lud=XlH zi+<9E@i2mkZOH)VdVP0gg(f({<%7EhUs>yc%4?XAUN@f)oBE#xT?k(C(0p}&*S;yl zv8kRR3}g%Ic>FNSR12*pH6w@!W&)=DJeBhpu{T81AZY9dv%$%q0mrnw|KKd=nSdSk zCc5FYAERMwh?Uc>VkN5%lJL77Yq2BOEZb~+c8RmGBj&tYI@mTeytXrm*B#4Nm=2=5 z&+ano=Et4@D>OQetPA=6LIy`w2JX$)_x??N7|x|tWUfntDkr}S1-9%yl#dn;<-1N| ze5p6Xlr7frF7v~w{}YrC{`4ik*_$-`~O+W~tZ(p!_S^3$qlO9JH@RlhDl zlJdrJ3{_qQTjxjB;JEtt(a0sU8&0QRg2Pv)9r7<@#%T0=KiuCzc9g@8pUYIY#K<%A z=fVp-zf{#{|8EDFRiOpdZU^BsNF$W}qZ@z~1)rW0Ejj~{x$c3Vw0wpGE8Zw*I$jXe271TpWu7BFfnuna^-G0wI9LadwF*dGK>>ok2nmx8F(x;$-4B z1Y6{bct>Ek7OVoUpiuFZ!)<(j!k;w{!xtmdZBmbmI^n&gpun>q^1oM-4@E8J;{#1* zE=mbfnc0>^)TVi#WX%u?-3|Z&SFk?xY(qUpW^kAE^y7{Wiz-Bm0eCsaKWT^_NxO<+ z!ZavWe@WmPq=>^4a|jPbMu#8ppz>uC?f)U&o`j)!9qhU(ym18cKP#^mZujGV=Eh6y2qhF%<|2QLH=`o5mNp zrbl0eyUoJimCTe!9OGfk<9F{M54Y%#AINLWN$*)ZGnaL!4TygT0+Km9M3x^UMYD{+ zw3+u;e$zABw;9|0t&)42PbbsNLr7d?ZOhE+C9G zTuZJ4YhyB|XcteSNR*oHPK;45!QS-$`bk1(@2+3SUFrNFp@4iC{8b?+A>xK)uULw0 z8d;|Df9$)+IghA@FUZrIePvUad?C@{x)gsGlV_YwHs*W-bS>?IAh3kix2^mTop61y zL;!}}hfKNBZ4Q@s_y+Ui1z`A>YY)=$blJ|K+9=XjDm-b&k1F;eq=Ws!ufSU2stT_P zB$w$atS9~@6rF{%5a#)kyy{Pv$Z*>|(-}PT{-Zu7(W4Cal@BHWrXD~IC#S#~yaqx6 zT2rn1*+Pl?H#h~EA)#6{Gns65kIX|JX<$NBTIANav`ia|0p^GKWN##f-NuTwgW4*< z$s)GYQBd5Et&0~66bgzpY$tbLP_QE*E6t6_xyBOVlz->8H5+XoG(gnT zXi_nJ6V)(yCe=ePLUP{5K$o49#kVYO)o9fTdIXBijLx3h5{Xt<@My1c`N1VTnQnrP zNySC|8{jjjvqF9MnpMDNM@inO`2s9#*(S8?_k(u+ykMZJDVqk(tlAsh8=D~ECB%eT3OK08Y-WMjL;){HPK&KdV= z`WGgV86ly#6Ax#Kad(%k2^cbWC!kd2nGuB z){T*jiEeNlpL0`{c(8 z#}Rkyfms~8BHsLC)BWt9H-;eFZXnW8Q^pb3m zL4@p=4|xA&uWrde3Rg%N%c|!D@Bw_NcW|HX^%u9$NN|^-2eglO$V`5VeXI|1N@C1{ zn(S9tyf5x!G_V-A(y8l$+w5UCipuu0JhxPGVXwxC8sg+zrMIVaZn~E~7l1T|JLLqt z>ND=e`dDfhBVsqX91 zj|-dbaHMS`b8~dZA-1-h2Ra*`?_le}xl`3k{(a?c|5@=?+|QvIcP~`F9Mz5>#KI?17)Q1n{lS?LR)G|yamno5KRZ{nfIQ~I97m zslA71(BV;xW;9hZajhQm3|twDq8E%x(<{_AlkM1Y3HHL$3tB;bE7puX5njX(wG#Kg zSz+0%=w?=dOl){#y#gzT*lyzE0dJn{GuAiJZk3KA7&#Xv5HnKYd0B<{WCIqec$X9r zcKF4zWJDl7DwOTYdG5e=AmU+=RYk{> zQf!Qn^>A=|N*p_U+OU@}d8q~Dvq+B7GX_)$XlTB!_>roYv$pOnoohRhWqtQG>X~oY}6}F5zi}<@(6oJqsMiTvp(^PvntOTT?o2Q1h-k>@& zoIRIh7f+Bvhhu6UyVwsr=(}D9ia3`$;n=lF9ncZPuhVndGm5Ep^cdF;d=5PK%<|u6Z$EEoB8v(PEK0 zZ!pZpYkD8iH`q{9N2hH!md}xmv1V(rbAigdes_41w6JLwId9)eT3AOfSrhp$B25^$ z?XN61w4CB0gfD0*b93V+!m}$Z1udik#63j{uEE2XXMnhC+`NaD9Ko5Di7%BW{l7MSJ2Q=xv8%PuGuE*+d zSln2AAsJUPH0c{ZJ-FQwBbGF+RZ$>xj2g}wgj@tC-sVsAf#6Dw(9-&5&GR`I;I%05USi(zRhorUS5{0@H-Wl(U-Ol9zn)Y$7n4PK z$n~}rY~dvxOKiKN7S@CiIT`AUZO=`YKh{JjzH0uS3vl?qN29Q1CQiKX<*Qk?R#y86q0r&(n(KpH6akX_%di^z6vKY zZXzQiv!7H%dx0gRDBf^r1-HlWoB)M5c}aG4y)t;LVt324JBb_1w^J8x`)DeCyVxGY zr$f*7epPD{#~#Vp=uy`lSY^5+d%J{BD!E1UohHDO^E!TVA`zW3qmm zk(cD4)rGWGuxohYkUBM|!mhZ_A`N2olNjh~e5f5*7t`+i#`X>b1%C?+>M+jmh(Iw* zeaq!3#;{i=;G&?r!PjkZLzYYV=4}(ZhDP0EseK+s_4(_}=9A3N^qVG2Wmp;5x=eG9 z6%%lcjb&=PMzj{KLI|K~j$mk3hN)z?=Om|Q4F56=D0~IJlEqXfH2c@%Tm5dIxG{7a z@KZm3$pX{gAw{fiJIp}$-iv!4xirLvpgXNxPBc-;mmijvn>LA|)BG`^=;Df%Rw<8< zu)8OM@krtL+EGb}rr<96`=iRY!cf_YV_51JN2Vfl&IJ4Ya@qz3)8e1Fm6`>{?(<#2 z@!vrAfx#k@!1CqpMdUU;#iQ-#Cqer{JYYNpXwb2}vU1XHft-gJT`HOy~Sb?f5G^`L zC+9dPavblxU+?RBUeD|CvRUWyYppVoL7lZxHBkx{(E1h`{NnnHKrQLrWw?2m) zSCuXJnan@-=1m|!ZXE9I?p&hS^w&lmt-vm^F5LC(vufal? zh2t#thELSyG7e{puh)y_n59rUd*ANQ^qz!iVU3ZqHi&g_$WHog_;kNnO;>0bWDKtr z!F<3e=^BAH4O-Z%IR4-%eBNq5$H=24`G^kv7rDYDcVe_n-Tc^q+kb#gsCE0kZGVT@ z%Fk_)MpA@jr$+kA;)EXqzJV5>BEH-*IN}gZl*TKus-$e^N_}QvMKA1E??1i14zh_O zapVPOa(%rTDW+QuKL9G^Etw{@@^6~J&EE%l{!x-mMQ5;zWi?7_xQoH%ER?cIO(X)c z!(NmDQ*B|MFXi|FLG1qXVFi(KTNz|^fiO(+?lpb97MS~K3B9BU-~laY4(U9~LSnQu z#khJR&yTE)+Hb{)nTs0p3dsk5eMj3L>6zSaNyKf`_6K@4cohVLjurJRqx?g&yqzvy z3${o+kP0sirgsf}im`JOchevvy{hH#2-=h12 z8L%DhR|c?n`Gc1h#Lq$5cJP>sUt1T*BSgT3pj>ciD45Ggmw(cn^2268P**tWvCnea zD`0WuS?Mu30!LaxIgC1?=Def*gJ$(NDoz~zg%m}we+v``308YvAwm$TP&qG#Q<-$l zCU8_HK=c_^iDBiWQ83Cr`(%=I=FVx)4Z9%ph%?CvR$Gl2t@#=&cis3hODNJsCA#ew zE!Um5(ObMv;!3CG$7TBS#|sy_vol5Azk@#T$@%E)TDc zUf7T$=|r#qmxB9eIe}@_qIHCvW==n$$l+h{b`}}Vd`J$$m|T-#?a8PiWDgh-D`8 zbjwb3t&pM;yut73Leo#nFgXAWGXb!mWWL5)nz9k~*B)JUURwvZdlOUSBV{^4KZlf z)oz!&_mz%bwKUvzgd9gvf-%Qxai2e?$+~YSiB&zN0u~x_2Ym36T@>MbKOY6=TP_JD zOK^ge8hh&YIDk)@BSv#5_W!+uh8g}+f2w4Xual$sKDAO~$rMPJx2@R$Rzj=QF94T4 z+?6@gar2xo(z>zPi3HRGx+il&6cc9K(D>_osfOQe10R&Nn-24rmtnMkZF}GeknU27 zV8ki=LtJhgC0yC=?3|KyhM&RQVWO~;e9-8f+_2F)W?HjNbLocww2pdu?QyuKdn-eq z_Du&Z+_&~A|45ab+f$Ol(3q^gGMjx3Db--{H@-IX1~SxBsNClM z!ty2TcCa9cAw2%?Y=J3(gn-r-`toBI2s)-;U>x7^=+!9^Xw;Wi1?bm{%Ui1OikK#m zsWp}H_nY`vMdGGUxFKKYv21y3xu($yf9m^xPBvAqspeP?E!|6kper=UqQh|)t;`Ur zf${TT3Gse8Bxie-<*dHV1Ee7GsIS`6QDjgtozJM^h5mp@^VW&Ai|IIC?($VLp=q;4 zmcay2uZsjCXIX~Yw8gxBO{CMdxg%YX#L&gn_W;RkID3p`itMySa;-0I|652!4hR7v zJM8(FEP~2E1e{^~^MBIIVE^t#y(&a07 zRWnUSu&*vk@(|ZmjKJ^Bql}T%&^Bm>%ZLqyUB# zrc5?}p7P{RmDmm=IZ_uG;a3HeCO%@b_2l&n%31)v@t9SO3ChFq>yUoTHx8SDs(iCv zglul4*likkZ+r)oz-N~d359%=GDOXaFMUZeDSvZ?23tiS*?Kc!(cvv zsCma0xyUv$S&vTtww8KeBfg=fH}2Pnqc-KRX}$H#Ndv1aw1l!`DZ(V1w;&5!Ikb{HxSFcWV?!$iH6$(5Eu$* zMoPxz7aq8-;*V5~QZY*Cr^*qLhX;z8M&4JuE@)veue-ed8snq9c*)I3Hw6nl@musP%@e(*v?Yx&Me>|E+Gg*BU&g6^JxL%TNb@6GzoU5JY zfk1?l?}I^tB}xR$61D*Y>#YXRZL(Kb{`=37@xZqCA8o$tqZRUI+z^NV&sHn^v#<^f9y(XfK; zo_xL3$7UDkDgWz&yTyVTXpiLKK~7X9eqA`x@e#rw@;+jD@;$k>{ZUP3i%u{2NG=EE z|Hycvh9g~^SY@=7l2vL7TTEE*UQYrwvHn{ZQJAvb<@~Ci49Bwnf_CmpmZmJx5E}3I zOy#>-=^-LP)LQ`Q9=tW~pc1HODaPHM)5?4B%RH53;)U>y?;?F_52T}74a%ba$$@0rSQDe&6p#2pWWv%>RwN8UF%iT#>fOiyuZi*_3`VRlspt4faPSzHHl&Yk4e_x>%NwF}t2UfX5=s-i~ z35Ex=zj}x)e@O15O131iPoCU#PN-D`{d$$?zHiEb?jTa)aMT+hmqdnM2T?X6=X6<9 zWX(F5LKZZWq~Zn^x!OHpC^{Hh#bIh89&3LAP8h0_+5}4ye*8nT0IQVs^uK8v=nC}oj5B3K4MLUE=D>klcyq?Qiq@8 zy%wy)tnLH9^ItOde?9}76%P214L`fzjiuWjz@FePMt$TCJg4H_)bQI-*!`z>M0zV6 z9iE7|5uGmabQzTO)1E{so zrSLf}Lzbrf+1m2jdf$NXTby4dEQJ|zJ2X-rH!M z-rJA0XWvbghUeQJk*oB#>}i7HO44_ZyKY7&Ywl43k2nqG1>#dI(ByMgN8IR5F>xfLe-;+To_3r6DVkBvP&DbKj zd^8!YsI8?ap)@yBQM-dR#779Ot~x%)<+{OLLoLFR-M#U7l{uLwTbN3&#keK%Ks$DZp$m` z*<)R;U^;og&6POt9JhgkqU{JE788ZMB84F+@R=ADLQ8(iY|SRnp8%?G9EmTvP689a zaA9_Dtedh-gL+X(-lenoUS`0F`Xfzg=$F^lkbX zbrR@{LhX&a#&pm%V;d9lBwRuSu{><-WtFXyU`apD5so6XM zd01oh=j&jA%iW-?u7)d4%PA7ehjiG_*Pn7*_e1Tb`FE#|PDN(d1rBK7Vpmk=^i) z{easm!U}l~xqPIUOND5WSU6))9$)djbN1rTrp@lAKF%x&^3A6^S^D;3Paw5$G1B5eTq(%NCwzKx8<*Tl?F1p!j(e+GAaL+ zJg#dhg#+x^#QkH%fVCdanj7hgG!IHy5>`2v_1xn2!Fiz^;HlRaN0+i%)e*S32h0cV znox9aX?C)$8#S4;jZuE0OY2{m>pES&#J4oj=-cZgePoJV=DAksevME^M;UAppF zmhSisSxG%9J)9z>s30DM2CzzWpF>_^}i*ca) z09#4`5t5EhS0_+nwkgtkPKl+$XSxdRwRyqaZBsok$G7p!5}l)wVzBWZoZfn&T(>n` zS^rG-Sl-MmoG_hlAwGYcG*WbDj5hB_zw~MM?;icb(A6n0VpL%uVXEXarQ~z}q^|wG zkFTgVo77}R(*8926g~GlCNLy^sG{U=+t#sVwocb835fom#1An+E?v}6=G-B8(sRte zsEKc`5U30{-VUpMAB^@y)ac7}GAFR(+Q;%IeD>TdurXlphKRfl{9q?IuZi6mOb+CQ z(}C-LFPk4;lW@qSCpE2i1a-J=j5Ir6D3MIu^YK(MLK8!^72`#UGPaBZvy_}Q6=C!oh`I zk!=9Bec_?Tif<_=1p)Nfl?`pj+uK*Z-g{>-aoDQ2;sYGRftad+{Nnz zANe_O8~3k8*ksH@*5uM=zJUmWJr9D-?(dBqw-}mtl`Q5x|1AJ{kEH4h5_!m3nFWAz zxwj*54R&i)$HK@CmKNq_&p+$6WYE5Wd$=w*b=BKbrO>t*j`8%Xd6#vQvTs;Dp2;dE zMrPivdH2sXAAbB4K1qM5h^zo&QK{QSY4*4|jjyh*gppR~Kf!35?J6(N(wBMY)L78N z%fC~}do?Z=@p%wDxa#}mO8X)o1^kep1oc7Qw%FtM7_IfPnaJR{H_2*mbjaM{96Nog zMP#mr9$U~qq>9A0iUjrdW96h4-K=V>w%X+Rg-mgPDzoHk^cVi0>?^MrZ z_Cw0I9@yj3fn0az-2TS&iiXcHtuP8`<^GD3?!OOm3{Lp1PfeE<agOaBj+%iM|6#{$%{G+%hfdrE&40g3=<^!KVE?2 zcZO=sltoi;#$~Ez>ye&PS;r-Kt&E#S+YLBn&tc4P4!PX7)WE29gZ9&?pMxM6;AJhu ztnW&I7a^oo5RBFu2hpv1XD|kAyFtwqb-EhILY|?_TGMa42?gsC&{IO;qOCE!rnrsD z3DX`1IOOCP02l+XmBu7+06-?}LVyUD<_NrW(JL{3q*v=|J+R7aNN5ka`$*Sjm{|l# z_yr~xpqKw`*8`OEufD@WC!8Lx*#G2>#!O8q@B8C~KDm83H8&u3aR$GPZQ?|#%p<%1 zP{A`5BH84|2|nj?;5f1Dvq{fW*a3?4_wNFGuYu@@_@j%KMDL(1Eyb%*0(9gTK$NoV z4@!On6p3>QjFzzW066*&l|&Sa56++gOS_vQbxVm!%SaAjXpGG`H?wIo9?9SyHP!u6*&lNV;Z7l zXb1v5fV}A*m67#dV3$0OR1((?f`#s8JM_<5IkUw$V26Gr{F->ulxo@X{%47Aa^9<~ zZ}K$Oz3+XoO8hRsTzUX6hu8qDn%4j$=Ms4)8&Fgm2l)juWE zt5UZ1=!-g*mu@*D|Ekwk&syJ#ryd(9GFOxif{n9R&${FYAOrYmH3=t(8_ey->Qc6i z|9+PqIB$UxK*1qN9(om39nLO#Thk`RftrGxR4k(~h=sA1w}oHp6w%nAMKSwrBw>x} z6%6L~(^4`9SYxf%)`0|K($-V;KROB7L#u_hs6Ow1;qoakEZh1#lKIfcljpgb8s%*D%@06}-QO$mDNZ2qQH(2n+|C$zrYBrPD`CRB zCRp&ySK0lM)6dEh+Y_pHuj)YPp ztq$8TE}bS(M~TLdhFJv_QJz?d`$rA$#&?%_pVPU3utO_<7E}r^9)r=(WnM7uFc{{& z)bp6#T#w)rEQfAwSN0}l;OMbKl|^uyn0m3NnIJ?*xS`4`#!Lhga|t{ene2nix7@c~S&<_hvk1U0{Ve=IM{Tu^&{Q;|uQ! zTW^mSC*Tb>x}LB)Rt!C1wM3N5y%)lT`%$|T4mlLsVNqs@VVvSJJ{j_lBiHJJ%?C+e zDqdWwyA&VP^Q;iciZ3bd88w~TW9&W6)b+AnIRb?)zfWpVNwhncIfKRF+p{sAAYsps zsnj5p2G!PrZ~Z!ih73CQYa&6X9E2ckcQA8HAyxam$Oi)C!m8PsBo452rr|@|a#pEt zScA;(=zBf-FRcYk?zV1-$LHm~Ja_i-FYb`-g~xHmQQWEg2u6^yOQ^XNr<~`7S6}A6X;Ee_0*YG5;p(Z@%3p{S`ROYH3 zg8qlGY-X1>eTdjG+gwuD`*kb6JdOaX_12(mK7Y8q&#t;cg76Ry`70FXN3hqEx^ok0 zxK{VPot%}6{RwWSaHdV5xrt;ki9LO&uX>i}I0A>c;2?-mfgm*6iMVRy6jmhz8p>;F z>sOXLbn`NktDhzLy6L5`F71)aD|(y*4#*ZK%1h&#D858cs`U%HWHf0_(l8OFkx&|3hHgQ(It=;a^~_z?kkeUi}tO z+_fB2ynDuyn9O8`6iV8zvt&NyR(eb@eRvZ^qX2`tugqVnUjt$T+V^d~nh>VvX0G3xwx4XU8z)^o`q%zC z zO%-@mmpOVMt@HDb?aH5;k6^T@lr}GKn*M8$DyPU=wRceaYb@4=_svpO` zjnh98@_`4se5S_Ali8mmt@X_0Py<}o|0!19m0*Vy8Q|PN0E$+;u!$jOFvfI&n%2cK zNUnQhDC=D8UFier`>XYj@x%pczA6%B*8x))Zex zfPj?TQXoRbwE_GV+7j>{J5)aKP_PrUiA>c@k^NvA@_;5;(&4A`y!cX{C1JbtoO`Si zFDBtV*V?0S@qh+=YbJm@+X|t682(g869seI0X%cM$6I4#fE6u&YrHVu_hg6vxNDsd z`ZW@?^#YaUgI`md$T=>K zKypiTKEcpczB{wmqD$Ne3iFq+&ABUtY+r~Qy=Q792V@m8JLng@UVJ#))bD?G8gYJA zy+954a@!O@RU-d z>8NZXb0C)-x zU+6kv+x+XuI=?-hpDQR!rr*r;FOL+GZWGx~)!85~^r9g2T>Iv_1jx}~gCZZCvD)g- zK<@RZ5(zGW(dS40x=jQhSAQRen6!3Ips76>EYw&Pnhv=o=Ur=>J&g3~-rpNrx_-43 z@$CFpbX**rhw0|6^XZ)2#p0-NrYvvT$<>TKfxw}qt*V)Q-4ynn+itzQ8qr-ny-{I= z&_Zeu;wmlRi7Dz;eZ@~>Uc)JxslJ4La84Nu~rh0jYM>9@dqPSVSz1G>L6NkigxxfuEO4t~D?-K7Z8g6>;(~DxY z5gXt4Oa)Cz;$!qzR=Fp2vr#~A2VY7Gk#`%}rS;02B}NBZQWw3fxi8gFA43Z5Vyw^9 zIZ<3Sm;bLE0cw(tz&Gu0xxT&VgOh@%Mw^eRz~sSHya8UMgyG>QG?(8^2HVoJDYvEb zfUPUTvAT0ZmwS_b;c~vL8FfKVCg<4TgxV4iwe}yDir3Fgbk>PIUM(&OyVLSK>!Z=e zpZ)0WCgxwcSB=U#;=|I)jlon3Tz^gZ!uS7>B(FYCmeMKu^yTE_W0daDS8KwfB{tVt?9ep_A%n!K zy3AtSP$R5X$=%TXiykS2YKy>I&{u-|VV!V+~<@>u|m=dRX*bF6KOg1hM?Az1qz@Vp)PzI|a% zc&4?{<)iPdG96oJD3@XVH)5biGVY=N?ZJ41{bb&Srgmc;1tz=b&q; z5b(Dt_;QeV-gn(3r6|^F_Tp^aY%Kdvzl4Ob-u1sLEfX!Y;O&8VU^PZ$pfWS+43u26 zHs_&r$7j5asSAH^-iR}oQ96we?@0sCOzrjFbbid!bkLD}NCiPi6YMbIHrn44d6;e9 z0mL7?WuuXE!a9$}_c_K&XFM~N=Mk*?lRM75&5a|8gXIAoHt1pFt96Ht-YLlC2N`G- zROZ>Lh~cw&THV@}mq9;JO~ON#&)1z4W@}yabLEUh=e>3OnQwpuKxTUQ())z1($7-j zZ&p~pIuk(WToI2+;Jz>@9MXU$fqzd3!_ux&9VypaYf`b^gr{(9{q#Nc(Ov5nqC$j` zLJux=P_SYRYwzu+Tgj)Pz45X<& zh&Z?A&NRA>Jjt#Rnu5O<4ZCtU(7k zG>uw*mf_uI2vl%W03UQlB^P?E@bnSQ6T{L3Nk5}$sn^T->ev%eT=?l9px`Aw zUby_QVnS8Ic+61s&0MoZf%1jqcn1~99X0(1**pJl5wHMg5D!6gbvtD}DjPnctK_6^ z9n7}g=`Wa&<|+(3x(=LdSR&Xc@Nf%v(=B+dP01w6>4XcfXudw*?k6b7Nxem6UM8ro zh#UoWS#p7Yv`Z@)FDcThTS&SM@E(r=Mjqio+P5v>^@8$#*HCf2{9AmMP6fd}T~u)K zW`nACNp)L!`MRHO-y<|N0Q8P;mLMmz++NLNVl-R};0fsg={TwzYt#o2M4JDJ0U8Cn zIemge!^WQLbCT`oGza*fa?Lk{yG8&PTGS%A4*0WH{g%U#qRZ9*EcosX+=zA~tL9D8 zjT?pXBcGCEEWzOvz-M2*x%8Z^adCmxfYr$L=BrdoTo)#p(GmTasNVA=B9Y|b#U=UFP^`%o&>rcA8fBa* z)VqF@qrlQGa6R)kGW~b&7f;{!{NH<+B%LtWxSv+NC>3B3Sd_2BAeasncNzN;CVX4{TYmP}excl=CMvFrUuDVrQOC~o2Q_AVP!Xh+IV?4SienFM zSX$4_O|Gb<`Z|9?jZO9;kmE=zr#pN-vx#Q|ywE*M8z3~M!Ew4lwc}1loeD=Z*(^fX z7*k}^uc_JPl63Y4Se>`uKAmxU5WxNxs6@ucM4FY;$xQ85t%Sl!;C0162%wY^o5(l@ zdbsOsh-S>b3z9MwW(vIZ8|YW~U@w$LYvyyc5K85Ge->|&lc_VNDk;opLBHAZh$uIR zG3?rp)%+djT_20Q5ypLmBCsN97zBC-*7f98%zAMEeYD8456mQNLdZPjyw-*s#&C>g zaehKKwMa<7;_n1t5)uV?yS|+H{{|$$ zOH9ng{*JL3&EVpotIgsYZa(0)@L~|dz_{m%C(b`()(N$!X&A!_jhd;6&@>U;kd8u5 z8EXq;h#_nVrZI#`qH+4#RNTZO@>lllBitX!DI)f8zRi{LvEd49Uxhou$oI#UF?lLe zx3-)brDk^{;M8GlwA?N#xE*;t)iD>fU`EQfo;lDpA$-?=J;~wL> zj>wmNM{=AMX>VXPR6VS__rGs{3<|az2tQBJi0H81UcnxNqn>S?fFetL@Lt-etdwCz zq&PNZ8Mq)1b1ozAjQii;OYmgSc)Ea0KVG_HKi;>T|IwjkhS~vZ2 z4(ZBpl0n<@Pb|2X%Qj)p#86qLv{>4S=8`4Vt&hIPKbrn{Rk$;`$OY1F#@bHzxKFng zDbVh=fDn}Xb%=Yba{It7;2t<2i0h)ZO_I`@BH(ln$gQ8q@!<(Fm!EF1P+|eSBs`W^ zx17zooT)!8t&nbpE3hD~l*Y-QCEQ^WlaU#7qx&SOnS4Yb)Fid!p|5`UaU~_H+q=IC zfOz{XNOgM1Z{3WJ9A5FJ1?-V-c+AYL6L@DvQC7Siri;`xd=pLk2*=mk_uJSOi1FB( z1w!lgBs#50ph;)o>e%eqftZ6;#<^-XPC?%@ZO@u-3TLAdsTZuf$+aU#_c(w1B|iYZ zMW_H_J{|3~a<>T6^rmxP47{wwAZdh=M{UY#f-(GbyR^`?a%<~=_dE<0o8-g;iP&(7 z^X#-$QA#uNP;8;JvjWBnGg?QJ7az!%%3@wTz~vq^_PI_i+SS{SjQcQLsSzSq`pM20 z3hp4;yp+NfrmB2K_NY0K6>W0VH$Mx-a;PS@rPlVcZ!p;tX#}g~DPo**1B87@Vxt{N$e@lhskNr?`t-W`p~ja|1s^g8LXsS~ zMofXbE9TkxJ~PC2`Q?X^WFOB?+0hD)Eib^4=zBG#++^9C)7PG78Ot%79Z|cyh_F=D z^P4@sZT-|A-*hlp-hbx>Kk$eiKE_-(yiRfz=LwaAIwpv()xJF_TOHeWZb*XR!Y@sK z^n^6Hbll{bNIMG~=E2$q>gEZT8jh|Mt&ayT>1h(3vD4FbaZ81|m6* z(z_YC7TLVu^&N`qdxP&n86OQ%DwA>2`}`BNP!28GcnU_8LUx~cG8dW75xnS?d;>;s zLz4I8|JARj`n`!T306)P#gTThv^Q9+2B{C+OEBi{WY!PyZZ|t13^g!LE+ekV3^1&E z$Dck8!0m)ZzcxP)UIk9~qp14Fr$8i$kkV&^sAHXb-{$ zG?TR9gG_^TN7JsmR~J@e^;(L~QOs)py`F*qC+MFW-vxJj{%7|ID{O?`xzuP5v$10Wnh}+kK_f4QkG&^0tq=9g6Q{ zfR+xL<&aj~B2+-+(Ot{AAt0M}>&;G^8sAWd^>7Hx)V2GiPs`i%U_j>wII3lN{$rKg zC=gwwWZJ)G7@&rRz&cC`G!heHuhM*IE)#3rck3#* zzROPJm(k~iqxcK7!i{1da>%_*)5%sb#>ViI7-ML}BGL~(Fdom|_>H?xHqN;HlQRUy z<%vpl$VXO=3_@NbRI={3yx@lAPZAoFi@n|#Elow`7@)GYO4-ntS7NTc)BFx<-% z<6n<6iO}!;b)$~eiEici*! z(|L|NT+bd1l1*$J;T|6vlaUZ2N_-PhSyzf(l0i+9e;WworysN&+4*TOtvO3Kip=3znMj%vU+|f+Q`zd1>pF~ zw46`dw@`AC>Sd31X7-=J4?>R}xSKZ*Ur1}ci+;>h(ktn))twkm_U|__`e~bn^nw7@ z0G9G6r7CNVgzB=em0Vy@c2H?-(;7US`Xrss_?cm)6K}HZdj+Rp=??ClfgS0o`b$-H zKyl1^j;JZPa_<_{|K0b&odv|jfVAqc+gco3gt7T|qk@_aXBakSm`xYf# z6684fuRa|m?I$(wjlBg%a`E}b8c=dD(NlseRg9y1v#^`CB(q57qtxqvy(n9~+Gl!t zu|H2$BmZ0FR}$06v}NXC1aH^nB|& z3jO%mvtdJhvYlr#!0A3&@#$0}!8G7UTy3C6rw31_a9MIIj#drElP}X;4&Tg4 zO;yB*@w2SZ0du(Vp-5`4?b1s#TMX|$IJ5WzcPEQPj+i3TxufaA(gU>TUaoFs#PY<87!(3U-5Am+3i9`0Tml}bAql;=H>w- z&cuJUJteAH%|> zNT}I`q#0`}{&CSMD#3zNwG@xj$vDXsv)!^$S$659OB9?8#$Gk!!9?WTML|mI^W%MX z6aV7{@Y|1fU^<00Ti&t$Ipg;Xt6%V%Q#eNp!}vS+50u zQ{a)s_8v&MfI#PvZIs;H0~wU2a0o}_v&eF{ICHgAjhKGIs0 zmWz$qY|D0TZD(G}iy>WQ5ELcuZ&kH4T>_R@krhDxxtJL{d!BKggCg?5(y{b=4>|_D z(ATk8c%X4p>tR_uTWRm*>*L`b>r3dvH&zTmyg<#2V&Uiim9+Ho6z&!fI%>-(%+=YC zk?gsEin-?W_s@!+EJrK2rE=xEghy`sd3j2<#1P5PJCyI1S;q=lLx^s z#GP2e-_RibitnXCpfOEEe?D6msIPbc^9{-crkX^f z<_+yP?)$SlwnbnHoF^ldJT-RDp7wftXf{c#vv?otx2gPl3}_r!#Mf?;ep^USpkVy< ziq<*y#`81`CtLVj{ga{YwPHnBOSyLtOo}o_vSo8U?^KZqmX;(lxNj5A`szU#Fg#h$ z7u<$BCPirlJGP8&=eL=SY|0p<=rk0-*=07bCET4(*W>r6RMPkQj;;HUbClAmM`Sn9 zUcW}(3&s2OM{9&|2z#EZn#NQHqnPn)nVr$oS^{D!*PyJ2fH9|3_%UDCfxJZIcn38- ze0sGGVwSaSS5Uth9w>Ylha8;-J`a%HS}$Yfcj~Sc8-?q2?i*L|*$1Akx2(i^uD+Xh zjZpeyv8sG=te7vhF12p%A7*|y3Q^70ypOc{&vpoO)k6yxzg-o(+?%F=vHglnCs~4{15H}Sy$Bi;8}B8V_762>q7!%uKAC%ZIMj~HP=MIIfk#gD%fhrDZCmN< zO)OGlg=g)(g*^>+9aM03o^+ilpoy<_(sSHOEy5GpHaXjg5`m8*YLxT~1Cq}FEe|wY zvQgnkY45h$4V$-j%=Q2p6+$J4M})_0&A4<2IhS0s2Z|>}Q=jg+42TzFR8X-sl^+ddB!}0XPj9y`k|Jnh&2&xx6S?-p zM5Dp9F0ctJ^W>)KqAl9{3wU5W#r0tQK5SH#i;p{pdZ?*ew_xj`W#YZno$Lu2#!fEx zD^t`j6JLcU6_vkgotIjT__G4vOw!(kKm9FR1SI{q=rAJ5L{fyYoncr99&j6=ufvXY#o4 zE@J&rIZciZ{}wEB7&&}NZo1vqpTmw*RVNxm=Umt9W8EdnJX2QJXZ1lik@_FylIyvM zrbMJklDP)Jjf<_~YWqlD;_DFd5%~94Hr4W$J3dJrQ6lWGNpklE`^>BOGuuU+OANNN zX3gL3{2MgzuvpgJ_r;vc`W!^?@Zuj7M6LN{j|9C@OeI>=$yP1V8E)P4YQv=@RUYBJ zwFx%m2&V)lust2Wf9+`;|K4pIEeu2-5YWS!!FKF!He4<&1Pqq*5B3n3*Labp+tuxW zPV3ooaXa;X5&d8*{l&K1a@A3wY$FQt)45Y1hgrg==YnUxWsp%L^?BOb<$c4dZvkBr zTDF+!JsA##;rwC}IvLApO=gm_`$$SWxh3&{DU1zGmFwJ;lo#<$D8GASTPsD%1jnL7 zOakNyJA61{d!LR#feISiP`^j}6Fgkles`Br!M???E%35Wb zP0y^AKk%t=A=XB&cE=O(a7Ulv*YyE$y?dDYm-@Ni#mR1?lx7;`&pBe&R>D)nQr)Lp z(6b3h`8rR2#Zmm0t$y$K+(BAu4% zeLrt}$;my~em<`fewKI=l4aVdN!=~;k%EmuOeFCEZXT81WPi+Zm?T%_*SjDyG5$5 zRz?O^emE^I{aCSpZyN^UeB*5ASe?Ix&6BQ9Z_}H)rROoj|MO$FW}rb(mb!5MkAZ_g`>l$%_JQV(SQp>~ z|2Ee_1DORjJ070y-;mE0wqp#Kzghb2%|I4K)t>tKL_58Q(uh9UFQ?+1% zIn9N8?P7ad_OHqy7aHGqT;+z&F28jvZnK#~Y!&=Mdofe}2>ME=oj!n(F;>g-f3fwR zVNGmp`*-LlAR}H3z`9j`0ln%!0el3}O$*XTS`7sim5VUEXx$$6 zuvYD)=3$T5Rk;9@DlbzT9j5)Zq!@eDTdajJfAn(H)PZxnQE8B?nc_5)j84Rbz+=#7 zPXSZ)Y_a`f{!~Oq87r+I|E_~Y%u{WRY)cXbuZYGU8|s`=G?)?`2O4(>(%etpM{oeya943j6e-FO}AVu<=~xah@i&B zteFE4qcmJ3o|?OmAl$FBQ*H0YkX0#Q`nBUKA)q!Y?uQZ1CC)n`6d(h4lUK7)SMQ0ik^bzaT4N zpL&Hwl06~ftW5m60e#YV=|rJ$=Xy^-@%9$=;|qmSLCb-c!nVeR2^*I#?#`n16PEi9 zE5shu+(xWLeH=+{DLuM1??WC^6ycInQ2hRre$CVW+>;dW zNyYV$@0Js8x|{qL;{N=}yutx)eJ!_b3=mFQB8mXVCO9o_X6BY^5J*;fd% zgTT%j#JP_6H6ivC(k^QF5m&!=Jrq?i*A61Ub}jZqyyBD1Xwj3qi&l-qW?|o&Gz}nE zPPEo0WGTf@Zy3J(tRQbY5_o`Cysk$LNb084)3TICD=ArhipD!%ULW^p7 za6JMb@`Z}venIH_RmHgeN;1v&Q?V{T>?qVBGnoby4Mr~-zX#|W5KZ1WYVhpVnjc8; zaS-!AO(+<6C83kylhDCa1_vP#fr17!`EebhUJq353>vwFC)FD+CT&Hn)L6?V#K*6J~ z!x(~dJn=`G1r>Q&oyjD}*MoH@^pYik2;aeL6Cf9Sa!3y#jAtYgX_$W#W$HCrB*)$w zNKQ(6$QO-_mt{`myk+);Y2^y6;iav_+J<$Mb{=R3D(oy;C5a-l?=x_i zs@yM$x&b!SS@-T1-SfmWe6i?|7a?w?Ao~(|lxFaNWI$O4?;+pf2PXN}UDQYV?VsOI ztG|2P0fGFTNaNZnF1KRGzgx1c$Rvbm1EI;rOUb;4FKYGNCeV-g?wzL>Ywq`+%zb>M zS*gOSi;YEu)Kx3PoXw76Nt%b_3Br_-WVlL%Gg;xJS31~<*>PRtNc6=+d+$~3IZ?LyikS1;oI&43v{_3DnN&PL`M<^ zF<=nOZ@Lh%8zJc(_jG&WY&&o#D6ScO9*Zc}tyW3kAMqXFn2K$Bt2Z9(2qS@z!z)NHZ_O=($ zB_t)T+sSIx_;>r(Pm5J!4Z^pf?~t0jO|twUyK@rF713AjCO zYNyz28`T-L;69$=?0;SDEHzODN>Q(O9XW4{*63{@4(L*R*OyFxGGlA4Usv$KDWUTQ znJyb1m^N-pfm(bY$e2k#<$w@6#T3nWc=EA330V!HQ*3P>I32gBP&7vcKCreu{`pKN zUq#}~YNvC-pi}?Om#P*8>iib*V&h19rss&O#HL>`rG*^J55~#Pkr&HNY@2b~Cv^p- z)jZjfoa9>vLF6)D{Yz$(N1@bSZ2h7W^*RPrryWanR9rGesg$2X2_lZp6n+AM=09!^ z;N#<*k+I(5gGP8!}D*jl`09_IRTJ8^)M^`&8@&9^2w&)35CGcTx6z%nHU74(i3 zWh8G0R6z~-c3YhJeY04W>*mN-5^Nx98`O@kKr^NdC~a0xixC*_7mHo8e<}$9E2+;6 z5-$TX0ln(MO{Wje?Q(HA_uIs9zMmx>PXO9Yq$=;J|GvkX*MZ!6HmT-A2&9!VD9)8x zi_oY$mNyk-2%UB!d(V8u_?8beE_FIbi8CD6;J^DRDw&AMG0ZC3<%dzY=lcMk$3VFH zkJxg1hwFH+-i2#Y*6uy3AAp?{f1>8)*$n6SLI{qQC-#_kTQ+E~eJTM5>T(l*Rb99? zHv=R}B)3t;Kn)j~STKlsE7oNxX^g;$9*KkdoO(DQ15jJ^dc+#Igc7y01hD&6;H0Gp znD-FJjhC7CYVv)&$b0egYaL)X9#3>^qMhNmZXP!DnEwU=L=z;#+@n-{D=oCGC(ZpLK1t0HuBk=tu77-h{^c*$i03|KR|7`gYKw}zZ&UZ-~8&h z{Y^r6NrS_D-PT&7657(kXquGC6 zR4@D(;B*P}9rTc+5L?!4dh~pWRg95>6a9-&iuX7Cf|lH~?r3IvY*p8Q;U>70l7sh? zgv^-mU7CYI5@NXp*gS1tU81*Mja8{HTyE2)YL>vP;>pq_E0 zlQ19dA_3rC!JvEdaMz8~Y_^WoCt-sCaye7AiyD1RDL!Zj)T+m)VsBjT1ALII>Qo{5 zPRR&wef=0mbH(67!*vdY# z3Wz42&yU^PuF4@IjS*tT01lgY=JuPG)Z>|`v1|KT<@0y}S%dksx2A61OLxl3_kfIw zf*Z81L`%ndRI&NX_Y|IWW(GhkskNH-B9@wg+E7co-?)T1eSyqL-s=dr>D-UIETrxl zNawczLH}-Y)%z}HT!bcOeV6D##2t74*($$aq5yU};%DA~V&yqHI<`y29??3(f~Q16 zomVG{+=A9C;6%(wFE+CWJeIiz-%2deM!NAN9TYaD9!0{zVFx+!4oKykdzzL>x~lT3 zP^RHXT156{r7&*B5w#&#YP`EfYW+spBQsHLBSnG1%S)oS@lA-8y6|G|Jx6X+D2m3f zbU!0<)43?TA*mzn1|eZ;^4ws)--k>1^lO z-p9>63xu$ql({poR(;>R zv8bqq1K6?g4~K_ko?52*eCcLFS*?cgYOhm0Vp+Wcs%&<}ahqbK<|mcAQFe-rjZ{@dAcrcka+#*b*A@ijzW$HCp$%om8U%uUvoV<`tn2Tql3K!%4RA}3;em*}k#ponBHdTgvzR1j^}gL3Y}>L+7~GMLEdx zpd<(I==fJ}3Z3ksLdd4)VR9~IEAbV}MJS>KiX^blONws@D}J|LAg?x3A+#@@Dp@#m zo#VI8vsL?x*X$Mb!BH({mrVXltHI98S=A)D&C%J@|NPrqo(H-bYr#Hy_H?37ZO$ud zJ#y>cD#sl>7R;Kceucii1<<%bD+9RT{=}RjK&@{V2th!Y1BSfTV_#H@)Tqo+^}~pJ z+A`zC&Tlm`KGZ2nuL$G=ob%O;W$vbY)^@EyNr2sjeHu#h2a$APl}^F zQA<{i0zS7>iE@_o5?Yo}6sX^6E}b|6`^;^9V1%q9}`Y zzY5#u5HXF|auD4jRCkw8pY`Q=G{Cb?yxuo3PrQ}l@0}*TR`9Ou_hXR%Vlz2wTYo5D zg0M{5ol|C30n8xnz~u+n&HS4+8g=`00mst{^E>Yx-xsYch4CW3ud{@7Bn$T&ma(rQ zB{iNneLk4r1q<;I5cVimIJ(P-ra#3mYVUcel|$Vi3j)cfN}`03kl*rFJT05U`x5EO z)C0>?Ug+p?KW_bQ_Qz#`Kbm;0XRCtpJhmBdH4dkN`LNtgm~TLOxo$MGHiQmPO>K>D z<1;&pw8tf96;Vk4(GP(rkCHTtUro;&rc8e;WOm?$N^7Rb1#*K%PejM51B(hyi)+IT-d!MO%Dp2M3^0H-V=0V3LGpf)VVY>lGl&7?3xne8vd+DSFf!qhIV;A z@S2Fk<%7NG7ZEau7X;1}&%`ec zPQCWHSvyXuc_H38k&8PQF?+4`xDjSxLE8xf@;U(?8b#9ZDfhm(Qe6Fmq{l^Wz4ol> z-!K=cqhqEQj3X2U`!Nl8enXhDZ4yEwxu?9%3;Z1Qe60*y5%qMngZ{RTi z&Cr>q>ceg>DRr-q#%#&oHAJ*q470v%A4gctg%gw-6o#MGco*hpUXKov<>HdE?zg+m zreBrgDkUmBb&yR4tfL;tOWEq;F!S;#K%(>&^I9nq8KAvFEhn;kj@^ z8?2t3jHG^0FzXi$3W>h~=Kp<<7iD+9r~KpM6!8^{Y0#`lG2;AXV6Ez9-iuwWKox*NuM)4cx^d2ZdjgO*xgl#$PQVf2CSJ+WdTmcj&<&>|q^x>|0&o=-T$e zPQi`bVBO%z<6rCXMLhj`ww#-%kxCVjA4sTuig@krmrLLR#w0X7Qjka^;rR2qe&<;` z-NjGfP%F8`W_uQ4~p2fZo$v0QyFH4C2T|ACS*KtJk^c~a$8 zvPZs%H)C69R|NsHYELh6;f(uavLo3vq~acGkNQSGbQKg3{Ce;_i^sJ?xN*%kXS+t- zaNUbd^PYprng>w2`HSVcq^6DIgMxeD;CyvPr7|7_)1{fRnEkYY^;q zQ$1A~E3TY`3>-Bnq)e}LPT4(55?Q#E`#%b8I;g;8l92URS%s`;-)_P|%a5g-p%`R5 zg81IYTMT&)b(`5kNZg0QRztG=Fb`AFWfC-X;NbPCgs2$cTsmxv0u=07u|`LWcd_ai z3~Y(i3=9u<0{s2B1)x5@0X~c;p!m+M-v?9;DUxD$Gi4jzzM;XZ%dBa+1>>50UFZ4} z%7VifZ>k}A*cM`oQw!LXsnNGVn&Gg>(3M1lt=~H~Ty>oQ4NL>gBqt?~J7A-w5xx5^ zAnqbFDZMfjJs_d+M_b06blqfF%z!gUw_g(}gMpF&(3bvyv2;UJC`XDWfHa~e?npi5 z3xAq%ZQGb?vG3ao0OqT6Wl5<}pW@_uG)BaIP?Yl>lCWkupz3Q+nNPXcAPH~7B7@I}GKcadSvJXY;p=lG z7E_I8&3O)?z_2~M|2D{w=)M2|MmVe5rL|dJd1W)cWkDZ*@*gdLTRh6_og0VI#Xy?^ z1m`UTCt{6P3m}`s66MHQt!bL9%{kePl?s~_)t0;??sn!{TrDLA^y=!@OE;6QAFrGD ze)mG)bsV$`6GE3+xawnttvq!nDxdu_*_r|LNvDhf327TC)K41jT8LgPI(ji*j;Ivp zQX6@{%9(y8adj6Ltdsw$cUD=L&IzBmpFbznBQRv7-`9FZ0NwH05dkq7*Cx5!|COqZ zt>G@i)v}7iE7CAfgut&1Udse9cJgf+8TQC6uT=JZkD`A6{F+ewxDd_~eDA4k7N2VR?rEHV)m+ zevE!E%y*`uvKrL9{0~=k)vxPYa0hIQevy#6=WI{+QY0OKkOhNRfV}+kw}-~BPBt%d z^R_>gUmqr=vyIz(2n}sxBFigo;DIZh=br>Dr{gHC^d^l$8}Exd9S2`sKVi$^p;%`Y z%)UP`XEBip&s}{qk!$Tervl^;a3a1E%$Piucrn!_ERTYO zc%L?&rnj)~!}%1Ad=-M!l6(rsFbbOMnk;=l$LK0iU2fQ^N zdImNFu^;&3Waqu}Je$w@3)ip`E)WSV-lG?dZI^yg&0^DJUl~kbxF%KB$mfOzsIpv` zg}))@#1SzXly&un4z5kj^L2iP(-NJCThyD_Cn0W3J=yR$D=SM0L?sK}QRJfc*nrXJ zvc@^Oukq}p3?hDhsFmKtdG@0yuQpv`85jpSU)8(sMwCQvCps$B$qIHv`kKb9zo@i1 z*~&#f*#2l@{;37Va{VcF|J=q(F zGKPfAE5`E<9Q6mc_n1o;rwzoNUPnZ+8*AriP|I1bfmdu{A9%j}RzwP%6B+?Yw)0+` zkcbQ7M?~@fxXgO)m$=lF!uwYk@6@dICE(EV0aL>%cnK~(=)&qG+oa`!>6_o&{bfv6 zp8kBYo^1Y+%vTOP z!i3*F==e_8b0fdqlZTsAqvg`DN4m=MUX$}SO-81m-UOkmqitTBGh0PX*jI>Am$g0#!4k>XN9xH3T9SOkN+U$X(40Kt{CUs^2Ejf`gVR zzFV}y{#W-0&4~}0(oNbv@AO10o>2KkMjh{K6R@udF7U45i`@LwsZ7EQayoosSl>$r z4rIPX1<+w`m$WKx4{%9*g%sgGcpPvp8ytaq8%XBFLw|z|-is}P+L?vWJQ%F#>Q>Fj4 ziBtLz*yk}0tuKR#O|O@Aih@n?o0aQ{#Gl6FjN?H`z4yFo6(yYr)_*aOJby5dlN3~S z(ve%8S?G^xpJeS?mF}tGr&+%#;Az=4ti zYFnl^c%3D0Fi>_j^C+9N!FCH}!iW4t1lbV*oKBuBej$CEYk>4$k~ zjw}UM#&yO(m$d8CkzMnJ1~W04JU!n(O2f^$1qhQ_-)+~s%}|tfl7J5}!N1+$+ag~v zv-?&LWR+cd9(P0bNbSI*&m$>_(?6qiGcU&_b&4^C%U!$2?W!EY3*-hKt)I6H4cYyb zSTWZh7@tgUyyI^^w%$5E|HUi5_`V``*SVjzf}I%yC#+Uk_C(Ehk`dNPTO@ol`n z_{uYZY^BAc=3Z3i z>?33po5{gBHC3C`V+4L{-Izi7fRF6B@S$K@l4-s!kaTHJ*uL}e=165?vk;~ znkMB>MEt~%j0XC2kpa<4vCpj+G>%vm2_(_hrmo$b;fWH*cfpO@oM{XODXTY%r%jDy zZQ3-aS&j!8GJZX(QHxZmfUMcRp|bSt^_S}9hq_7qn0&XQJn^p1NYC<%1zwa^1SN-m z2IIta;t39#-Tq@OvFI;mE81P}GWENQMTr$|2HWW;_h5!I573m?M~u}A){at4n^{d0 z*TXm&tdq5bBBl`THPs`bUbn+f1jeNfaP`RIzTOIr8OW^{1-hQdHOI+_%TdqigHL4n zeKSEKEf-&;RlSa|i~4cyT~kKb-x6k0_DTEi?Dd$r{Uq(dAt_Zid@a^g1S3m3%o%VD zB;^KkFw%BCxT1}-4hZc9LnzPJ`UZZaokg=Bp4+o}5nBjm5c;I9b?a`)CtmHOLLfIq`@I3+5vWu;B8_C(dtzL? z_iERV$riV`R;>nb3?5ITdRm3`D{a``#8Jqfj|MXn6Wku3gf=TJJ8G?g9{OXWt9MEY;lCJS8&>j?>B;mWF+o84VP@>UvH!WbhR|KAnvPa-8LpFe zk937V#5v3tME3YZB}vDIaQ|AW-68-l#W+RD#iU(b!962rZhf?m$i^a3osqhJ7spZP zkzh`|0InGdWc$q>e6?3mBC@=Y3N%sA?bAm+Hv(+Oa-aXOuKqQhjH~t9FgLF~OS%|NJak;{~AF$m2-a*}suZK$RdkIos{H)S|$uxV9upZs;tr z%w`Z7?6c+U6FO;hWnv`b`hFvjsHj0`W2Ue^1wq*3aPs~S$+75n(j|3j>-*5mR9`Y$ z0im%U?*rm$cJX$zNS!)w^ge2td{w@_anjbh^=rUzC5sAPE0Rk2HCh`#?E=ujKom{_ zCG7!Ehxgw3udh5tH<`!MNNX`AORcNs-h50cdLFYPCr6K-tgt-`yL{cPBJ3|}w!!Dz z_vSiO@NNLk_JrqbCzzOH<^-*8z}BQB2@p_h8N96{aM5T4<1xp9L3n7-Wo+ItS_Rr`Q1;StR#o>HX& z1^@bEd^M?qBkI7%mWCq4f~vDOC&YFGU@S)2QqJ>M%@n+wb*G+s5vJ?P91IAK2Ivlu zN_z47c#hvkJ&*D7CHoj%XCK+@G^Z;j{k@y0i8REo;Y&A}M+R|D@#-UaSEPKT^%|yw zx5v2JKgb*KQ?4j1t~fS@It?DC5<&Ne{Ud*h*w1>MhQEagU7NLaO^yT7+^~R%*x`mQ z!YLo6)7J=+@OZcrjm>y=VSo>D&(!>A=BN)<+NGETE95vG;o{H%xez^l(ZEcoxGHI3 zN&=TbG}NG@DJr%55PM4Mf|~v4`Qq z4bZH)&m|#Z@!m$;3)GkJR>{L<72e*sn+4`u4~W5PiPBFxWbhwgeEQzdgYNq5Uz2D0 z$JJZmt$c7<{eaz>hLy{*kFo3p1CQlUxeY2?jlc6n8n01|mPlWe;mNfqdf3h*B2!Hm zpfc2e61;eDd&a$5`oD;q`e(q0s&K+MTi?)@iYzHem=A;D|iMaK14geX#GO2jwY#erJPS)-{v;qT4bPB8u4JcfBc$iYie;Lp{ z79VO+{3oIvSI?sq&)T0{)`WN8giPmTBw>$dPI-Cf=9QB}X&EE2L4zFu#WbtUwRM|i>F_C+Nx%g=dm8U3y*NkaO(b3e;^pgd|ZtIlLk=acHvfG;`DlDM(cfl(*42krJOgsS1o+$(X+vj&9U?2b_YDtyXx>hblgSDu%yLx%5NM+V*ZKi!&oCgFy@Dc~a3((PD>u8)9K_Ujs1XunUk6oW_VEm26)J6N&eHah_k1P#NoFFa54~wMVZ9Q!hgv-ygFNB#q1hv zpG+g-cri%`z2+K%aSZ-QM(z(#JzJ|5*t0CJ;yYkRfwnn&0=*K{q6N00j~WIM*1ZL$ zXILYTeuY}`lX_k=?go`L|Ds>M55LU(4v(W}Xw#Rlj7XG`CkWao9zkR}ls}%2x6Q!A zmG5`!jQ8^~opfHAd2o42^-6ooG;97D)5U~s?^_s!DdVQI0ctt;v)-WADf}v`+y{s= zt{(ROUoXv-i3wcRslu%L&a@vqxW3n$>#T#ATor`7RK0VHXgy*|<2lhl&2UX#q}_s8efs8rb^M9vUv1LdqLtL*3gFg{qe!$a}ZiHCqT2iyL-70Z|Dy{)2;oUu4kIM9$kjvY^sJl zS7st&db}QKNFw{O0{4LVrjT`7!AW#(xElS(O?LcN~DY> zY7!@u6cD1~zeXU7n&QQkva*WrSgk!RCb@dMJ+W)EwiAO-5JB~+zB4?YWF;_*HrnOU z+RG63Dsn{xyenuKoQw?7_M+(-C=ENQb3kk}Tl-Uk zdrh-=7X)r*xAQuE7w#EI_TG)jcCU3ZY_R6{Ix-M(n@gJ%i|MS|orPTZR~&)vM{RHJ z$1?$-s_4~(!%RpVe1f~OsOe=A+3kH+A>(@YWa6dC!q>dF=>|xkvN$)Aofpnon~^on zrW=xU_60|+H$yTi z3DP-feSDLE=iNW8mtBspPib!m7hfNtd2xT@>a%A95{Gq%yOHJHzxv{bXLKrjzIYBq z8>aR)#J*k@BZ%J){uKxPje7}tT1t2$Wpa!IB{sl=w4bJP9`L+~gzuQ`lGj`$*LN#T zP0?*jV*JVM>YuJh$Ns>$UNI^eCh*4@cC{C-c4La1mV9c1S3XFiAh0_DoB@iT^(%Eo ztO@Y-_)&U#bzSs`%e{JTr2?`q)N2fjq#dDpD09j6$oDYpT_v`epg@Jz4!3St49tlO zjVl`kxFX@S=YvGKZ3c1K>y)VXG4qmkxC0@*vAnu~?1SHCw!eF3*7#QbWI>e68GpSC zae%N`GEB1Mpp!}7vs1FGLxAFRGO)>$Z#zrXbJyb)oNILQ2$?~IrrHH}uzhas5Xp>n z`-lvboT9pRYXOr>cIhp_Z_zeL4@NtET$y7}t|f^4n}2PEl40AMua@EYkxfC$G=&Y4 zY&G-ndBf^nNA$Fl;K`hV_bCN&YH~de!R^&}h8@_G`=0?=c0CSmU&Jc1KGmETyufbF zum1Fk{rA4WeZ;`NK)OF~r`;^0PC?)4a8EOsQc8pNFt&d}Z}kg{`FbWBWVQlSe7hfu z!nxm9tH#)e?e{1>LdkqY(zAF~l=X5=hKHa~%C-lR@#;nrkRB=Tc#BCMli1JE6TAG| ztP(g4h{mpS)6Di`dWcnjLELLsv% zit_Z`GiynhRfM8DO2K`1*mxp-GqTUo-Bj*aV1er9SMy22Yrp$BD7zBl^FhFGLBW0$ z@q2b|YxEboosn8_lzxATW_aJlY0rr$Ej+|^jlk~iOX;`#_2>MTVwOr&n#Y$R&Q?le zhs`lzAiKR%OcQ0Hy;3$O#0CMV1N&x*%b8%;n!#8Pj6i?UwS3Gkb<{$fpYcH?+3?!z z&V{F0vO+}XaF1xpRa=6f$;lIKM|N%ryxsLTYnt#{McBFkIym-pznbCwrJkYzO7l$r z0j89FDhFklxMF4yX27)fX#-K}A1kJ-lnR4YxexeDf421PF@WrZ6rk*DrS=FehE8YG zBYfn}xE4gnm^%eWlWbJ}at^5eALoGBqw=&g_6HMPJnuuL-mwPn6ftaz`wwey1+eYJ zF*~^~eV|sHW5yr-YUtNSQhfG-mPniDE$f(GvvS4JV5k$_+kzdT`8=H}ghD}2DNt@o z*R&%%^fq0)cFyq8CN>#s;LlI?iyXYP6J#o{*+8%M-nC}tjrxVY`l%TL}FfKZY+GfM;hNm$p#9Q9TXf-t;U6a900(`r`TJoL3mWh(6*iN>1u5%(B13 z3wDR>jVUG)D)ik+^i1L%P6Xg2;RqNddq+ToCMnIuw5^zk*k1(W(Ka-?V z!Es!Jx7XPt0q(WSog=fN^}$$*sRWn4y=wRsBj=q{?0_-TVYA{`@4=*E9z6M_?3@TO zNNKnwQ#m@zVc&MR2ZjLrK=AVVDlv9_#+Dv|JJ34CkvGkiA>N_o_Sb8+^`*a(k;~%7 ztcpev{FwgNC43giYi?)mD2u?)G}t>7xXHpC3fxbPxFq()Ye(%jc65ULV%{6OdjrGej%pK6+MRB_D244WApL7%mk&10)n<0Lt-B zDcXbUm#qU-NFK)i>9pF5ZD$}%lA)*q94cvsk%sWdSq?EM7^_Vj0Ia9TT?{bQB>*{Tl-9NHdt-5@SB7%rS z*JmnEmuwuxq)vgu`f4WN0c(ZQ+g-Rye|c_!k0%bryg;3qgH=oUA);s=biI5pB?~c_ z!6pH={BDtVGN>@J)r$f0;MaIApmEFtA`cYnB(rR>m=Cf$AL$)5wX;=2IUn@)yDH&D z+yXyKw@kd{$KZz+gKDZ%#Kbmh1!mvs ztbR(K3Bsm^xu4zuDP{HKudLiX?<3Yh;Q*_US;`;2E`r>_E=7F0UxS>RO&CT1>*bhv zWwBNU(g(RMp(tc8-@Jwfq^hc0RhO+&Qn4{ll;I7C&jjG>^L3^ht98aCrJyt?sA~CR83+8MMjKWX zE_(_EB;tm!IrP2VYsva?bBmbLBZdbxmJxL9q%2}q^K@ksi@-lm3c^!Kt?++=; zoUc$)gwA`dCvO1>;4pplL{Z0=Kz6nUJX;I1ocod?5 zMt~mlo~|)HD8kNAkv`VBM8D>NT7TR(E*@9nO0+8v`kH={CMk&K^v<|jgbTp^=XBtB znTZ(;FQ^366`ZE3K_5M~h5J04TE`Rzk3`jlMk0LMV?v3c)~ujM~(QlJM>$cLEY zXw(Z=r!TkbonJTqhx7aw>>2QC@V!z>{{?8InN&6<{@-cm#c-3iz|xa=3JW>eu~Wb6O28h}$O%rqlkA zxt>0+XA+qyTlHE0g+!7<6QpLHp#TzDVf-CI_fh=4HyjSB$yfb%za1Bk>Wus(coCV> zIM$<+?>aljDsj8qeWu1agN~%nug6QP7{HC_!~GBT6oW^q7JRq4-IV)#xt(Hyx*GML z*(5Zl*-rXze=|4w-2EP)*7qhg*GPI8+2FP)k`RdKH<-oCdREO}mCN;|!ZEKZ?Z)*x z;^|_?Ujp#%7syS1=icvy0_*}J*;!ct{kOiG|BtODTRqWYUHm_W7+}m}KMs;%Vtg_I z9@{@akG~%TfN&}4%l~9={P*7cy4`IbV(&Zu|C5s0lM*#jMoB%LLVX>B2)lSCa$sAL z2b68e1NI-R_#|PI#_&cuA*(M$^t3n@MiU(M_`kQtdnpcn?=y|RY6g_uXOhzoC>c>Q)LG#-Y91rb;Y}G$OdF|(plXWfN zS8DGcSTH6_IHvU30tx4zTH1o^dXFZQRh^~aNQUbafB+uWRXGy6E`P;@Bi&4XE6U3- zC)>Ue`-{eHmrN934F$geAay=3dRnpscu=m#y~T*N>6&mpd;g<5!Z7zwjPz7&6LFFr zKjhcK6~bck{Vtkbc9?BD97je1@o}v`$$_G#CWl4FqZj)~fD>JK@!LnW9~c+|uz%)? zKT`*$aG%pFIPgjJhyU(s%Fgi+gp~;SJ(05JVT#+h|KbM!%{{Us{ma+nX7{g{W90ug!nHd!;h6I zSASte?H9h@puhh7em6V9xsz7Zq2U6ki&JSo>kb$c9w)aI2;wnj@xEr49Sn&o&R-`& zq)n=6i@joK&cO35uJI{I?PE#6f zmeS1uTb`GV^^D8OALHqSPbdO+eT;G*oGCLSWLuH6)*?wTtdY|N=N{r1^H zwav?YIkVdaa9)j<=o;(%Dbhp+J}#j9+1cf>;H7(-wYjC_0XCU8?d|`EBL4^5iVy_u zFfZbhPoeRY#s5x7e|xQ%p%#ftQWxMHQ_95q=7xBUAE5L#PK%H;# z(ENy#xw=9R{^`h|EqAcG8gB*9SXj#Cg{{iZZQz`p32an-Zk*=m{IxWv(u zCkLJ>zw8zo#wjWX)s>-BPe%kfbei)4D8l$D5`eGJ!Ij?n(CAqZT!k82$}LXsyaAJj zY(>8C1LmuV($)#%feW@%d+W_d1bt==b7kg<(AL$pxT~BZjy=lSf#Z0zIjL%QG7IhA|EW?f9Q~;a{UtI}Nhed&&e&bKjd=4V^1k z9Q@zD`tyySTrML#*=8`sueOM(wD~zBNnk;1)ZaJ|;Hqc0NdP{`ZPM&p5$_>`!&8x} zuV{biO<9I&=&H1Nso0;%))&nDie@1{3v8(ASvds#` zUWD5$5f;m%J$h36IYcrFfA^|gDritoaCa7A3{JZ;KlZq1qm3vqE)2kKgpN_4tObAS z!>akQ-vpZzV2H?r01#y{5G?M0DCQQLAZcFpXMUUk1nL(65i_Gmp`1SY24YKmy50k! zU%&TyaM*tDN+DE`y&ffyb`Sl0tHrhQyuNTG1ycmd{ah;;8&7l9H{ks5r`L2o;@FRE zu*QAQ6&&d)ZHaULU(@KHab*y`C-uD+Ia6vHJ5z31Y)9h-6%noctHbW#K+~5`ccwG^ z&z6W@^Z#cwiRu{x0-Oq$Z@}9PdU5gXPp7sH3m7pJHa~aj4`{xg{hTm zJzhX-5Pw0F^Hd1I2nImMvRYu5fV7ZNBWBTnggPtK3Ve75Y&GOn{mt5TNFZ%HSHOyW zTsaXF76(VKaT{R@pwk^0S^G0p7pLy_)0MW>`b{REW#uS22DLbUS_7T8k57F4JhBpW zuXJJ{Da>+_=++f;{ZIaEL3V-P4A#0CA^g*7L`CYE_HsQY$rF_Vv2m80RfV%woeKZ% ziauIbGm#1?P!@)w(K)i#v{T)2-Cq9RefigBevJ>8&fX&_+$q)^HtBbHt4l%*CgTp> zL;uApD%~#n6&s124=~sFu>sLSx z`{Tqk0@&JC&{SMhrIPQCYSCIxGO+a}sPyXz;&HX`-`953d{x|XMm?9=MzR9TgFM0M=1v=f^0n+li;njTY0 ziT;^yvv`ex>L13r*=ngdoxkqVMy~f8@5rVF@(a!A(TOm2(#4uGIHnCwd#e&ECNDi! zEO(os;?ZY6zt0Q52GFVSS;qf*Q#TAQntc$*{9as|>zPlM&-bHF9zr6i>8!SMa| zTn^7O?akho$|(W@?p+`GvUv(~=Yi8#HKU1fzIi&+kF*~@^|&>6qgl|tLlWmphlXp% zN-;YWb@X0Yu>GF(QHg;KBa*8)hX<+HK=Af6wJoKkzZ*$shHXSmn9AOfQnC2198t? zB*M7xRlU*fOl9ZGxPQ86#bxmq#qz$RsKeKOg$Vb>t^yWJYU3rSIIrhICdJueVN(#^ z`5gQLARSRYW->!FBdH4Z4d#j z-6+{U;3qjYqf$!t-wN89@gOs)XMxx+pjQJaLDkKE=gY_xlHZ6VDhA3&6cik6LdWl6 z#VWPDOloc#(qC)$x@S^Rovj7pX)gsv^1K~#d!$6)DT08XgY7G9{GTrY+XwVlSZDkQ z|5c)QH^9SILUf-MdpF?5WNUBhd9W$ce?BK>oizkY5*d+^EeDWU0q z9p(Elwy7&T5ORXRUbC(bV!0?dAge<^AXL)JNX9>BkzqcEz)?^B^qnagQR?h`hzAr|$6W91;q)Man3cDAUF;(r?hfr!EL&&l!sKh4brW?-igQtI!O^uA@Z zwaK$b{OOSj`m0$i!WhF56_ry^_gqz#>mJ>|8osRejQ5GaaU0F-3OiEJi@1yNz!<7b; zB`n|0Q!Y^)wMGWB3GKYU&%vQRP5kdR@gp2!{~A$yV?--4(pE*q@l~muQ`H@SN_4y=HOBJk=R|GE9B{@YwoBBy{umyo4aKrB$Qn=LSuFKxW{55NOZU znX|am`H*Uej6{Nk2o5#-F`t`jCFb*+>+OGa4sa;|KjzYiGCqqpUW`})IPdBwix1qs z9Oa;8lzktkCxwT9HSK(o437o7V?G}T?aX3>8jV;RPm3$&zx`fEpEteO2PR{(&RQ*| z)61vZjY87MYAj)d2XnHiVLQ~onY7pfF)6M=ROGVr+gMyfcRcb%D)0}<+iNzLP1uKP zJKi0;H<9ze|8$-|vuytCPQ`L;lSgeErB9998(D(( ztaZVc*FV<;KEoPU(ATjM@H!Ms__SoZanPvW@8ah6TEd=;Ze7kkqQ!R)(X_39>AkoZ zNi{%5MiQ%?V{)I9vzw!>7e0pj@4FBP!Bqn8MD4kSfs`84$65GdfbYr4yK+n(%Qf+!U*AzqHc0Del` z4iYYV3v&F!d&eLk~mT=biz)t06{426|Jlv+xSS^)=D+<2NSgDp=IgS}Me zU{f|RNe6Bts2CXYUt^N&uw=iku_|Ygul{|yHQcyL7tEUB{b(gGjsD5hmxF_>&6h9?cm-$UE<$0Sd|4@Kf^75YMu*qpbQMe4XF1d zW8S@Vd@^{)Y4C}F>)H05dv06V#*g-zP|M5aQwO4%=J_H=S0udM(HfP}&M0=zC;Gsw z<6aK{9PqNPUcdkAVSRMnzb<2f^RIJUthxcdrou_iolzZQ~de3QO@H0=(nj|rw z?<9$e5GPmd`5h;55UBbg26gXWmj!ajZ;p~a*z#&*ww)-7E^bQUAfe*=dQL*sQ;ojf z^qK5MUn?A4uW^^T&AKDf@c!Hy#P2Jc64RrVpGTWHrmH9o#ooE%rtdXpXA&++^V?3{ zBUw}_oW`|zWBJG-f1@hjb2ihQZI|{>-;S1YpN~}fkmn1<6hGzs`6{PL5+F1mEnG5n zEw`%(H2go}-ZQGn^=tQCg1{0CtfivTMFgZv?;xU70YPeLBE9$CL*7$BXD$>)0PO z6zZ1WKAFCOw5BfC|IVawH21vLKBoRfdP4O~j6{v|D))TXzj=SmF5XBvTv zn+rZs#g@l3!vpCv@zTC)9Ki&dxI|923O}c;H?f)C(*?%9YeS_7FWiv2l=qqg z29x3uI#Bjmlm0{d5S&Hgzip&H{%Ch*mhgB(E`NHI*K)QZeeJdhX7Bfz>0}lp9 zuAk*Ex12?VE>bjC=~j(@pK?LSVx;wSJyW{S3}{yE8FPLRf>*=%D&B^E!2X(|j;kN6 zO})Ob&~_|g?(Cf%Q}%0do0FAQ0huXmurB6w*PGF@>qFv5krv^@bQmdx)u?ig%3;tG z=b7qdE~|wcsQ0?+ac@{c(I+f1GhD5^n?|`u;8sXaocm9Ty|y!(?8~P_c^$Uz*SEQz zl_l*nNvPC2WB3`woZL=)epm0M{%yI^Oao!`4k}1I-QF$}BmdEtl*(EG*&~OA9RH}9 z72jLwd(%h>0yK(aVp~lw%eMc+ZPdr-+aY&~uwy$o3J=Mgb49h(ihIg1Mh!{*6Gy)h zj=_@w<1cBaLz5UxwN(wK2FrI>J%zuoHtgb+{vGz;j%l#=tajVu3G=1;BdenYk)f5V z0{&TO#@&b8FLkP&Xt|)(uP@JJ4f+cdU7ktqTlD4HR$CN35L?GLI_Ct$P zEk0Y&W5*Q6;ZNGvxr~}x_P3|aH8UkFa@NF<1W$rz!H$aKOu5Ld; z^LAol;_~l>4%A2KB?!ZDgn`K8e6-uxx~S78Z@ z6Wmo~v$q?Yu-L$Dm$q5~sO%T8!F@J^!lTs_ZTPGfJ2echrZv(WnB&qKi5>`F!fiS_ z`>?#`vn(*J(eo+9U9a@-yR_UCQwWX zZ{PM$^g8eW4X)=^)o(|g@9-Fu=Y9k6#{7S~lEEv#3@fi#dsCjh&VYeFxRdX49c&wo zNu*^Fqg>GUlGYZ&+@Hr<_1|`1Acys}7yazMocgU+GB#;8VL<(ZYE`wuwckY)%kS$4 zt(qtnU5mg#0E6{LI)b_^-re2ZK&FtvP<8g#uV3tv$NlWApK1?lLORHe@n9rUN}1%Y zD&~xoomFtBFlR5hWczz(XTTddB(8|LC_n$6YbEIY2w zS4H0W>VN}!x(9ZZNyX15OZIiY=xDrl9E%V7(dVaVhW0oGvgPCnp4g&}*w!i6(Ytb$ zy2=vI^5DU2aDH9=uf^*v_|{UUn$est&U;4#V|`50s_nDqtxofHyQB6x<7>Xf9<=cD zj&Vs=!9*jk2j7S-WBWi`rYEw(+6)nh4wPc zPJ~4H|2{&&pMimCIrbFQEsaCE(o`UvM)VI^Bvui#{YE+J?edm(+I2+Tmg*P9t0BNp4=A z>rs$R=vm{_k1{`==E!vr&QjpU)`y#e)}zWSMs`n1SNqcm$@kdV%@12TzI}6QE*>fD zQDzCf`-Z%gyCbrVKTn>hyV9S|BUn|H9rc8Z^*ZDRpE2^0-1#qcx(9A?r#tdAQ0ugp z1%0Z0`p%h)p_^sLM!RciRp8VYbA4uWenyHCER-f@I}w?QYquHDqYEh&$TKKM&|&)1 zbg2Qln&Rpg3wPVvqsVn-2zbpHX=N$pF?#=sMkCnS zGEJu7Y`300IY@}R);=^SdVj16Q0DTsAzbE?KT8fziEWaZ$S224uITbvF*zynlcrz- zEa~C17U2lY#HlxKQeE!MuQuE@h2PkQ=rB>+Sc56isd5l4bsJq7>T;5iXA)M@uP|Da zy@mNSpea3mgHHD3Ih~B(yA~p<8Fj3Y1SDFh-G;|kM&E517}^Yq+>-uglTZ4p7%?F= zgt2`2a=Gqv$L3iv_9GZ#J0iH4uiL10vvU}Xt7DSt6|^O;o{a54AnJe9uebRAs=frz z>w>AaV{oZ)Q{M30cv#Q#K{?JJYRSphP&*en*=5*&EHjW&M6d*Wkf7rN9Yw1sR+M8- zyBV0>SkiJN?4vH9a>iQqJcoYM%={hkVn6JNp~`;5=D07|KR@6@0B9&uZ(fklpLIjR z13_3T+7oh~^2)$4)LHcmZ-DlVL@sFIw;#c02O*NnF+2T6H3j~CF2Wrpt4i$HB-w1Ki+w(4ADTt#L(QRL*ZjM!d}q-hM)m(fa&ShwDN2 zBL@1HLzPMo+i{HNm4n{rJw4g&Bd26kt%TN%xya1}!axQQg*!dYDex~93n7M{Ok#;8 z`ydA%#qEv9%39zY4Z+ywk~zHR63<;nG<=Uik=SaBlS5&Bi=6PBU6@U5sMWx7(73^c z@o}ZsWKpK^>&Imt{Q$!694G%Ruz1fmMi@e(U^~*G2#=o)J2|SKe-*%c39>ni z(mwq-zZ#p+Y=><5`iI7Cmq?exX2M%=@-s~zzJLF&; zZ8@%&IXOy9Xj8dTd`*{awta}erUd<>L4bQ@Y&3@K&jjFE?mSUTN=dNhk?0C}BzJ06 zD>GG}mE;p5umA=X15NfnqLM$t@)$x^^+;J1BakjMgM%N5*7UUY$fQ?oD@lPK$tfd6 zi9(x4#)z#r`ME^(lCdv5(!N_DVZYA92sdw2*d9pHS|4@1BkDL-_m|2uw{F@sAJ)2! zNqkgt`F+qCx@AGod2;=4Dm)0noVr%WigK)a^nSjUughE%S>%8Z~n#xI%B3HcQtdUQtvwlk4MQj#bV9;R6}IERzG%dxHQ;+4yZc$x{?gd zAe5h!^WC2Eo~%nfAzE}`;x>;dCmY6HC%!+@-fnErDAmz(M;TyEyNX@q;pn9`t8|{XDrvN0-hTjfxnquMJ3`C6zjeMCa!o6B zJat7Lz%Gw}u^w|NDazb27AO42epC%=-T zlkFtYrvK*NrTXTfi}{Rcf4Vdpg?Vnu{qvFG_U8}cY|BMBs2WwezWIFf1O;R%L$SU#;sf54uge}zZ1DFLf8{nN)y)X%5wkd zKW3Tbv0|=hqa{Zyb1rg}n8z7q?rd>bSXf+Jf9JEJkpVN`sZK5^sp%{o`@-%kAE#&* z(3mG|KF!xsu<=;Y=%$K?T!lSr(I3Y?XC!T)BWZN1O1mtL;ZKzc3W&Ds`8V}y+d(s9 z#i^p6=Y>)^6w(_F+qxcE|Az%=IpAt-*KRu_<#o$f;DR*S-@kun-h3`_fn|Hp0v!=i z^DXv>` zo91h^DGb96%-3dpKgsL`Q4}w=u>2+vmJ3p)r)_5g@rl{31apL0FnRYUgUDW@WR%L- zx~@c1dRSiVBjrY_l*$K6P!MxE< zwwDbLg#0BNCtOi8M^JzIvCq?aZQk-;H^v0bf-p!j9zZ$ znPUUzycWI}jR{&r41Dg01R35|*-LGet$=zZGX^OW&8((|o19ON`VaHMYCm#)tz+V{ zT&Ta!`E9uoYhHWj*$cJh&$T4VS~8QU*M@cLqlmL}3|JuD8dyDaVtI1rF;!rR6lkA| z@!Qz=tQg5YKZu8#*R31urNH@KEw6Q4!3KuzQZe$GBfl}|0(;=!0^TMt@h9?X4Va~L zYFDe*_wzx7e$)`fX_r>oJv^41BeWO#{o5ZWcY;lVTE2I2n&ey_+=xmpU3N99rPMG5 z`hcWc(n%nYBNAJ})u5k>K$F~`@tjUKG8ZiJz@tOgo|sE}Z6=j?3<>NmXEth{M;ZAk zm3|Yk8`TwanpwIv|9fk2vc}dL%+eBaAGY#8fAhD#Mxp&)%1yiBV5@<6Ewjy^LytII zqG7Oj@wuQrW4e{kGi3RJIJs-gf4i$%f6Abh7#92~ha4`BHh<8K;#)qgok|Edb3i9w zq+9R+$P)2yHtVSvIwAXB-E=AErl*-QMga$Foc@$I>-Hc=*wE$Ding8Xu5#1nQ5$H;No&sKRl*yz8P;YRsoxi$-njmg6$(vG1ah#;<^AA58%!@@Nw+7;m z$=vq6n}j1==F9xjJ%&%>ySA8%_{8`%%|PTqFyT>8<~GpI4ybPxXT&mVg@zV*-m>k- z`EFaYre224cp5glwB8Ufojb%EPWzysS-)b4hQ(%GH7T_p4Stl2IWg|u;$o$`6^y+_ zzHy{>i^05AN%WbR;~{o#u+AYdkokuWiL&E(*+8XZs?tB>-*zT?iGYK;g~Rrfq~ej~ zqCG6JK0_?H@bW2p(M(Q7S$LdKiSx`+>*?{F%(ch!s|KLe>NP27weBu0mo=@uh*!P> z1%%#yId3<))Y@+1J~*k~3(*e$$!Hqqx)RLONO_~`cZ@z)$#i8^nK5Ec;Lm)GH?H(L zYjZWJqh-&{qf+=eLHKk3T)XqFs$Z|hlO&qd^)tr?pl)jRmwZsf<^%-1Aib5f&Qn`74 zWCq`GN7W?N&_J5)tV`yT)MSKh2X+rbzrGGXG+WzEW_K(n;Ru}jcGGDHFYY?$F%ZUv zw}rS25Z4`#-A22FR>xQq*tJ+QntP!}t2F2JA+2Iw6YL6tnD+Px{}O$w!-(EtV^`?b z8c0qbN-O)GOH}}wH0d-vI70ijweB-8{M`>1DoieGl5(7aDDj=RR9!yBgz|P5pcR3*#rv$@;lRPD#wY7T zsEh4~jkF}MPN8nPJ#eqy1(gfQEJE{0g}i^pwDXz}M##=(m|TX$lZ1psB6rx+Y$;Qc zZ4?`RnV*kSOX)<9rIC3XkA`FJhc?Ex4^_}fsPT9uwW-*bdfm`z&L)EaugsawjEQ{< zOl4U6y<0#FTM3rN2)OB*lPYhZ31qS&wL3vz9_klAqbo;vQXoNF_oTPQ!52YeMwK^8 z8lZPXBeZgf$K)xq|E8*IG6Tq7;D}aQSoJdhec6D6_;>FZt}tdH+J`XZ!2b$kfQuxJ z8yXwhEj*%Xwo7}AFVb`ISmLn@qlc3eU{es!wDT+r2xCZ4i8&j{Bw(_7g)w}z+_9^6 zKB1<<0SZ|oCsp32AN&3E(W6{F){_@Yk#p?^bhc(*EPJ`F0-a#8*wPu7yJueqF?S7RYDbho&uHMtY)s)f@zlyOzDWf8xaxeM97 zR(@!`1Dj~3W4#MeNx!ED!sEXp_R}a2vUFa(k}#PQx`2S$i_rf=90nfeVf#=%S3+h` zr+{0&UM$I3mhpDqbLd?*HeuekFT>cl8Hwxvb#+Vb(k4fGU=xLCNrYRfhrG#PD+MeP z!zpQL_ur7y#J>ui6$n!h%?vR}xC;h!e~Fc^QWK-+r;~Ph4Y!9_F4a&f+h2wSaWl%M zmA=|v^=zb6R?mNw>@bdvOKKll8OUfQj8N?t<$!YtZ`)SbKC~)mIl_dy;3}#_WEPo? zN%?TC@{i`uUo5fj6MZoZ3t7sg3AGk)YJ-*p*FehbwKm@2wKX>~q$uUR!B0WW*izUY zk8^|ukH_`b49W`NZTc)N#F!8UQr>kB^-sd`UQF&2R~V=n?eDD4;xVZN=6*g=a8HC< z^>7IIZr@dgF}LPF8V6^33oXp6yhLKdc`#e{a5A1vxxf?cvWq4#SGQ}UPQ$ZA@g<1f zj6j)IPUOrE?Qw@JO+JH!M=X}1LCxE}`&HN*XI>HBqzX58`)w1>rd9@tVehgq!L8_s zJg-e|eMt3R39t$VeUHXP;o#1zB+CE#HC?!^t5b@~#`P}Ac;%RVs%)!$6DzCv)ubAm z*-!7t`Jz&|Il-ZING=s}cKkZ%>__U{x~M%F4A|AnS@=18L;~dvK30L2qiRO!kQs)b zzu_Px{GoQG#AsSl-)P#b4FSUTw#Xj`Q~ulGx0b;&^w$W>(-OEYp{DZG3wA(JY*bnf zZWt8zrQm}D@p?FUMZ#d_j*6xjy%%tV%FNpzzmbFbtB^|DyALinkSK5ae73Ndvup@s z(OT{GET0WLOlUqMz9M@m;x*dfuioa~ebhd5-s0Kz;_D;VWo}c;u&=Du79C2sgM2@gGG|ZN%^jEUkYs= zge6LWS)j!PS#iH5bz{OUX+K3p(04k?c1qqAw-ExB-+g<)X4*zbB}<;cZ8(PAW!EVg z*SK)x|3&ZlFGshWD047|XLRk90dLxFaIST(JwC(?5yIA1sRj;!tIR^W)hoR`qL0>1 zu1O4AUih#Vn(=jsVm;*&^q1w+Q{})vW)O`E5QsH^kd=0CaIcq5{>257ihULd6;!6| z#_VB-HjC=QHEZ6-d3v@&ciobbGI?oB^%1>dO@jo_l~e$vPe~Sz&1@V(GXk?OvXq6i z%E||8vxppEsr7L`DR2QGc09)bRORe0dl{uxICt1MHp6J48!%k(ljT=lgvLja$vJe?Bl@G^oK$|}ZFCX4dMecV9wgWP zHm~=*O4n1;7hqCjm0)yKRMS$+c)6*q=SD=hS_=Gr1yXEOS|Bo(F&=!q-C%gf))+eM z4TbgC`j|&f4Zjy;qiJ}4B%kYfRv`LP=}~g}HvD3B|BImic>aDO1uP_5aNu40(9YuL zx%DT@nZz4BrkW>^+xLMQS@Tc@6GB?{D>}lown=lnJ-)gA7C+RAaHZIEu)fc<@{3nc zK@G2Y)gy3$n1B1G=@)yv_*wD7dhTtV+%>FYfDGA5pYXKwiTNI7rV)Ca?i9G9|tbPr80C4x<$WS z%q?HUumFecHh4T$9|(@@`;(kcpf@&i`TU-dtYwchYqiqSCU0B_ZmE{cbZ&B$my4xEbqgk z2J*@abe4IhK92ONKouO9x@myrSy&JAMjed0B`rIr> zph_-(mLYSx>b2P`1^VbO@#xIFenGzi7J~sdMWl*}NhW5;MSs4Vse_b)S}F&_2f7;| zj>n>Ok11~<1V`FaicWMyagrg&%bDuq>ss-7(H)$kGplz)m! z`bK+_n_i{3TheBKx)@w2A6u+&N4Cq!pGs|fhRnIl3qG^=!i8&TbMU$$Wj*og?fnu= zqlxMK{a62kVsr6Yn?zZvU5MS5`d5@lg-)i5)y2=-<78+djl5&kEA1 zUTF7|!M4IZWOiRAKqxX7Mc?Yr2!b#Yo>czqN6DF8oGKpSh~{_odxA%-Hsdm@#@+O< zTq!r2KKVsnw%@YvD~k}KFA26(Fq27Hnx~%~nszB!innnbczQ92L6M2&NsAvDp|3HYIQ_zs4{&C@?{; z+8b^mHyin63O0(X_k!Ub1)iC0hIIC8WouHXlOgXM^bnEBD0z8e%^oW2&G@@kFE`JL{ZXORsC#;6YISZCE8$Bo)g z7=HWq?UY2aFNH4^B)QVl820e2#km~eilu&oQSJr1J2t7FF+ly#XT--2mlO9wXkpYW zk}nnLFoL&t7QU2zG=@|)be(sF)~LgR!pqt~s#E#*F_#dEun zIP9Qx(&S7yRR6{kNzIhr$xndJv3Mq{a!zk^3S8h z@K-<&NZXzukO!CrJAc_?3!Hnc(`^*%YumTh?8fx#t$&`?uLkqtcT72iYF zQA*JNK0d}$%?$<*?#wji@E;=s&C2(%gXRZEjfB|f=)?~hVn(3|muZ{DPuP%!REAPJ zPxsBifI+ra_mPkdF)5$vdR!&)qB;Jfj&J+~Swi`syNLm1*XmT8M13~8^vde0B%pHp zXi%d2Tc_drd^}=e<0qkY=)`SBbO-RXvks~zUO(NgdOp~Q%oz`v-I;3k%?VEe*7e2= zh4@6U9_D;m3ac5rENyO^Wq>~eY~sHd8$*0xv|^=2;^{I{2Jkut0D)1RU!}gnI9qny zG*_tv4a+W_HBe#_7S8ll`e15fo+Cch6+`doD{nG*5e=M$QT_M#|LQ5dL{y4_x%w4G zjZ<%MJqxjf3U!hwfB0D2?WI6}5g}hdp8t51c^HK8iBeVqFo=!j?S_|%fil8DxD3;@ zc@dts>94M`{ki~z?sJ!6fBtU-_aBP-5u8+cir=K+MFsFFUBdDRdnxgXk<!Pjr5yJLzK7yFux4-DnrWm)ryXaTo|D!^fW(~f)^HIkB8gjh5BFr~biMSG02&U+z_ zv8sN(X_*O$RmEMvp8FSRB&E3tw!ZF8d6X{{GSyg#QeMeBX=3Yl%kIE!->WeQ4a zfH(I)MC*>1VPz-e3h#NEZc8=Iv~UD|Q=-q5P;Fnfxg1+NrET?cmzG2sX_2SWxrO+o zAlvp+`vfo3TBMXIn(_Zf~ zoOK!!v@Z)iM<-rU&F)o-LA!Et_pY!K1NC}Mj2_}N=@Oq8mv~)}@BhR9`jJxER;MKD zdw0*D+&d<;Ohu~`x_da5`Q5iBcZ*&2dEPbLJ5An1)bkzvm^$04NKZS8zSJf1IzTGc zcP8z7r}>lif4+@}7jGym&s^B;v3DNB7mpjsO38UvB@DzAJ=kyqbFoWOdMV>ECUX1B zx{`t3^2I;?Pi*q93s!%A6+y|!o|W9);0K?nV~<~t)FTsDGL|2y|iNk zW#RUnoWiOSSzuE9IyEHX|Kr1gau27L%$Ps5_AHLxz57%CEr*W8A}w9&JYAaziE=!& zaY<4~mQpXkk)iEO)N`_Z$hxF#wjR$an|k@;s`-x%^}j!|@qM>^ zn!_9^Z?#`9E!p|`HD6|-yZhdf0IA&+r1B5Mte=i$NfinT4$TR^tB-nyW{`+#dNlv; zlgY{zw2Aq=3Oce`|K-b<7ir3WLNNa;BV1|@m9=^NXqeV?jC4Kq3ZwnReQs_w&|mT% z#&wPt6#XV_P5Y|Y+1W*qSqT~cNN`*qgj|MsT}3!e-p$CYYX%Vb6&jks+Ru3zqRIit zz{Aa|4d0PMoHYg+jP{QyEq1%%KCSz@={cz~5b-KEo`7^j_<7`h`Ia1b&VMX?mo=~| z<1%RNSTDXVpgRS$e95m#@Kd; zk}5;5aC2BwZ;!SQ)vgKWYrg5!X%1ln_h_Z5K)T?oug>4c zx6Bf5@)@hEY@K*4b;mztkU06;Snu`pa0?><{A{{wQXZtrEHSsa_?TVz20qI3T6V%} zHlo{y3Vd*rmW~$yqc0#o!7x!DFgwQPAqRxff1Ays+7}}SDA-6q!KVGoj+@a8%LCau zAmMLygMp`3le^0RCE%WHv5wxHsCo`#p=n)I1Pec;w0($9dL^5!i~Y`>`D)OW`o5}W zs@8XQ8EGJB(~;#A}9olhh;Sx&U5B*ubMcTzqGX6qQ?PD zN@i%A&ZyGub5=9#>i{KV9q3JMKXqyWl2ia(_n6SaWpF1!-wgIyK@w5RzL&R zZ(UY;!;A!r2{daSHtD=3p=JP}TY$lv0p>n7B+>Dn=Ve%-dd*Gv=EtE8ks31Cw=Cpp z-3}3-VSU8l+A{*c!@ZT@z-7^YJIaN8=eze}O0f!Pt*#h#@$Q|+(}&g01ZwF|1h=xY zAl7AHOr*^$>{twg*U5PZwMehk1PXe%ax~}vVF3VEFO{Bl@9!(jYw|N$o3kswNY`ww zI#9C-d6SEp+h$b8e8bmkYyD%kd}x@=?!i?k*$04CvQ1Uo8>>4pdM?~T%& zZoE0(4iRAHM7>tGm+i+z1bPKxaNoNKp~`Qlu&As=(C}9!opLPSZ;f(ypv50GC1tB{ zPuAw+SftPcxBSo7WNi@@SGX$HvW(Z=5x9E|*xR_QoOfQxB8+?2lrL)LXCjg#OU@bW z0^iH%vaSAL7EnQ60ZP;YyNCgb$_CdZh5--!zR3u9rJ3GWPdnMIse{l>XXQDK(~k?P zklwQmU@AppGx>}=+Swa5<6XhZI*XD;mCrWk>uo3(_($stz4-Y8*W?=;+!Nr2Obr;= z5Nttfmw4j0Dv(Xn-`A$_bb(v3IJwjDQqP#Q(&C_GK&pfgDFSe8A&y5v*fn3R-|6zc zGP89#XTeLq>XDYeV9i)L4}B!j2U#1kBv=fEN9?)2AIxe?)J|CS1m4R(u2nZG3K$S} zXGoZy7Y_V|IqfafFkY(S%&s5w*_v|3{pjrMIoy)A8u(auVWBlUdwIn;%{FO_!>3<- zzKvL`-$1+W%RCF*Y|y)E)Ee-?X=cAQSl;`E4Q?*DI;2Bs+GYfwu^h<~ayrnV0?G#D z-SVfELBL@e3G4pIH*MVkn=L+EFekI``l?tRuqQUT9ka)U4TM(Nfg1gK@@D*&t6mK6 zi=?zn7rQkZa@fqXcT=r_vG5Tv=cA{^AJd!r1S2UgkP32Gqjq*7>o%yU-J6<1Fx`=7 zD?3CV)HzttnxycWa+m5?q+FCvp|YfjC%H6-8>GaO)`F&U*S6#z(W%|*qK-=xw%Z35 zXQLLEh^VT?E`WaCf{^eURX@I{VFNeyKKUPqo3oCDtlyK5OKWf3V7$vF&&bXI3Vf~m zQl)d8DqLem_s7Lcfw2HMqqSgXY^Wwf&{ya54$zdRQ&@ z>Y-#Lu@a%k1OMszEiuk`dq(4~CpaF^z{wL<^K*~MHha9ncf3y6jK1WjBw!(D8PQk7JAmdT<*=daBS8j9CG$& zym-B+Z%SC!1RYaK&S%;_eOfDw&wloP&T(AA(H9g_p@_k=KkDK=uF6INNoEo&)d_9w zg^(M);5te2gk;8D6z};)-;zf6$~1TGGkVW9E9AaG(VHS%TAyZ#&3;qT$Z}slQnSU< z@5O|hInwLyK_*uHH>caqe+9UeU*GD$qcGH4ZLngTcdjK9J^lJA)0ES+v&VM9cB7NW z8AmYI#B%SJU#A7QSLfQ3kuDhIY-4sx!1zY)lWIr2%e48D^&>@~kOM^H^G)iJz_!4) zAwrm0S3t-bT;0^u0(H8IaWw5{Iy^uwmIwPX9aW6j5uV~Sd|2F(HGWknG&k6Vgq*os z2ibxiOZSaO0G!;@jB9n>P{pd@^qO+izeWG+*+5> z6m=iI(H7|v1F$k z6CS7IT8i`D9x`?qEf!cTdYZZ_EU)TklKV9Kxbd_tu%`DFYe(6)c$bYe^@O)woY4Dg zAO+H(VQkd15RutCK%4i?8$@<@z*&-iS#x)rZx8UZ_v$8m`sls1eC(NUUj!RwPnr-du#|3_lX{`C*Tsn{U$O^8 zci@-M>BFrMTK(L+c}w$OS$HKzzZg;Vn3u%_W51e&7uffW3!NT_!*hp#tIVhSyj~}6 z1h1U$-={z47O8JWS!`-&_~aFXF)rX#(bwAl{8Ah)>y^1KWf>G&{#vKOPm*wa zAU>A_6e+*bzWnm1d7o3z5!5w5$IpqJk(W=;XqNT3p8$oJC|O0iOk5EKb`10TLudxL*`(=M(ieE=8i+kZizyazB98Ya*u^d$8Ib& z2H54Rj}_^lrp1nkY@IfX=b-ek37DSCg?bWG2-h|;RJH@!3?L85&F2})yR`iFUE|)Z z?u$wc((q3}xmL~zQUiQEpO&=e7so+J$?9zuDX$6c->cY5zP5<@;%1+HPt!h7Y*gdk zZ?218HrsH*lDB=vWamBBxvNEh@>@obZuZ}pjdmKrg20orXhi5I~Ua1``>rjf%dSRc;( z(owKMUltVSR1UmncHEw+$Om;!D1vvvD^Ke1sMKVa*+8QgYfC*Dz`PIC21X?B!-<7u z0{}4G>B%m7m&9-WggN+>D_zpFUi@A1#|Ww}jiY|L=B@(Wa7uw54=tK zL%UY*fmcV;$=nxh7U@L{)MG`!2Gf&`_%hIEj1n!yE1LrCanR??3|$-eN!q+bwYN1t zd=Cl36yX2wzumWcOKqO$8O37zngFd zCaRYI1NnCDhz6SBb;VP%sFXT)^%Uqj(*`$%s1%x%;bl^5 zprY!3e);v=li|?$7p3>W{H}30$S~38e!uc-gTB|uA&nb32Dp_8u!vqo$vw3LsJGPy z@{Rt&>SL4K7fq@?oD>+6mlHtvY+3#}cbK+REvpH#`Sn~bn9v7mfuIu)ORoVi3|+FjD*A=y ze4}bE7)RP;-jK9q+(AdApiThW+x=wu%Ad%CLuq|`8_+rF=BWMHineRJlH$X3T{Shb(J7q#jW@7P|2%FQU1=+RWUr9>D8y-`x* zX%g|Q2T?A2m;&o?qq?tcNPsl7?EMn!F^Re80g`ve_L9&n6sYuMq#rhulq)?&8hRuw+Dd2dCS$rw4i$Ol*byD$PcIZS@ zSo!j{Y=Gnj9C?FBTOTU5@odoioO*3X^u-dW)@;zJ(UrSY+@l3Uf~2V<*dhL|h+ zSIu)rX~MAgLG2zwwSueF1+W9WIOKU4oL%dPccxq!!V2+sP7AJ3M{f(+{k{Msa+i2? zdtn9WzPwm@F8?fyatxz2ZK}v0DWb{>#=cET*>2-b{@4@X&VSOQi9Wn?ZtXO`YC%P| zPS0}xFZJ5KsHgX2S$9^*3qVPY1IXX_$%E;t&xlZfw~ zKm2(ZWzUvI?8#yD;;!DnIk_@@Xm&eWP5BjUEdLSQS?RI@=#ZHZ3%mmE2N9pwEk&|1Uf@M+`U88b0vgbkHk@}bUfDZ!DF6&5G#+BG}~2v z(rk?wZyV&$tUg=qI0ryn(l_?r=_+YZFisf5+11^d&C!>eD%Xq8+&b`{YjL$s6Nb@W z{rRAaz!ForXShG>TkUW_FBx?7j!fD`ZGLq@Ob21cmM-Spq{j*YAOEPPC_v60?u=#}gYDt()8 znH5h96WRPs7;P+D?k|myIoaQEy)^jcOZQa(@vz>2h^|6^R=$sTEY&j8V}$U`JZRP* zuP_%_o4x1yk-Gv#l_otYDT=Ylcg@#G0B5{`@S>$Rl294$X{$g<#1T5n6xJD@+o!VZmi z{#^5}nDblYE5>|!U+cf`E^SLB6kOsmC7rspRfAf%?mTVu%pOdoxcRsRWPP&Az=K~F zge*-DldaM-J$k zz@ChY4LZ`O(*&D&b^Bj1qXsfk<&na{%1_T8w#)07$yP7(aph=ysRfqtT+3a2VZOPzS8N+(Aq$mOOd}1)zx(6EZdbjwdH`?Fx#rwAZ0>HmvV);;!+a zn|~K^>z5KJ%Mx+R#|*i}N4#+uhYMU3`{!v3a)q%Cib3J4Jw?;%;oM_ZGf8WQ>69Pd zr@~AFyK~OHb9_z3oW^xLR`^$m!AH}$9vBO?u?pASS&B(l8Mh_KjmMcu8Yo6a3BRAO z$*)}5Nk~h(V~fhHieiOcc}#lqq`fV|NM!XhVJz;g{;i?NG(>ii~B)C zN`IzgA9wMZVW)K0RVu1PHs#oWm2tYf98^|K`;(4BzWmR^pYFi{q?xd`m@E_iz~aG8LoZEU;?0-pdP5jquW6Hjvt`Xk%9QCfB`A{nc8J=VAW?ksPv^&?1S?V{!y3w!m0M#JUndbQ6hP-}vR?a?=QU zGO6^&YVo&&PGf!1HLKn#7aJgJF=x47m5dE|aGeWAqqrgDe6yu_Z|3(!)du>jE7 zS;#G)lWUf=FtucwBxRPsN7?>|lIbG$Q9-!<9$8KZMT5d6@1<^Ss%;>X@r6gHCJxG< ze)smfa-aV3RJC^o zBaN(!1uK&|{7%7Gey@JndQe;4(<7(cteS1#v(w|2pD^zq;Cx@6U^|jknCI*Ns9J z>aS8Vu1T!Vpwa}NbL(BHcIyVYie%SLP&vAh;f@Eo^eluejBU1Sfh}_RN2*!~^03Nu(6`n6nm7F#x%Gaa-cH_NR=UpeIP zv?)g)&Qiy%86aZ{8p#C=T$d=(ZbpWeCnP3j5hDUnU5(k8z|gnhc!2Z2)N2!YmrtNp zFIFj7$d1x0RbO@xU5W;*<{{rAsWPAQEDgQPchpc8r>};69e1l)ZPU*Sst9AYm8XAL z@~rVAS-e-)n#2aUL?lmSAp)Q)OZ2RK9e4a$7MfULj4+#$Ims9CMWVOY4lg=Glv1v= z#^#qnWRc9un2{IO;k{!8N&`}J>xM4p!!#EE;_|&Rz6)|ACr3NW;U)VyiWiuG4WMvL z{6=hkeu;3NMn2EBBlNS=!o-5xo_s_&CA!Sp{pcMFSx8%F1@(1L)K8h4WIYedfZL&t zJXvSx(fd>u>^s_(6Hs~wK!f1&qK3=dAdCzOjT#wPbL*z?x1LM{LVu{>B`wxLP|-Tt zSgOZe4f-G`4P!*(@+)?*q47XmR{us84|EZ*OQL1;s-W_B`^H^&i#@zNNtByDaK5_6 zu`ynedeI^R@N-hr2lWp3=Y5;)_7R1isb zH|$(5BZ-dqL&WX#UNKS9XR{&v%0;J-*FSneMiOF=c5Y$~i=ku`b!G*l;j162tb3v< zwnZQ3>N$$I)y~nGZpO~z5Gl=^zWaj68e@-3g6GQEXDbka6IV+ zhkHR}&A3_;-0tgDIJGF4k5;-~h>ky0i2b~n>)GaX>_Mdoo$sa#g`iC&|M2=i3(MRa zp)>>o+*4%!AHLoLs;R7dAD*#c1q($XfEDQoQbJb|kluS0M0)Ri7z;`XDhLQdl+dL2 z4nb*3Pbi^>9zu%{AcU6hWaj_=M(6eY7R$9<8gA}Acb|Rsex7IVJ$i7In*Q|!`L4L^ zH>?*I@Bsoa(q}O5`6Xe~v?+MbfMYw(BIw~=<}*5EmnAle!XZVMR}Y$GL8mA_y(e$g~+bRWAx<|Ov+f3VeN!!OYff5qqh;iqN=ob;$myg}oy&fr!b4s>OzR>DY z-X)H(bU3G}`)6AAbuA&|+4sz0xKV;hUJ2*P?0dRj_D#ny1tWnizV(hep?TQNSz5(U zkyQLbuXI)wZb+~1 zESV8;E#+#v8YxlQFPwH$e7><}parvR93ZE{fY14d^d4+%MCmYuAp*g4<5B3qruLh&>enPi$N1R%mcKzBD8qwU!Vp)Tgq?=I`^~;q^rU%+1eQbywWNJh=Fletca4=!nlI9oKmvrIe~b&qatwxU0EK^XO%7co*K=J- zl`n9T^{;yTjWBimVTee_;}8aG(Z`)be)UT%;*I4k6`RC5bhN?hK`49HG4HiDG+UKAFHtv(z3s2uRTwARtGz?XQG!5BQa;$A%(3o#TRSr)r ze2Ek=vx1--(8YB_)tj;#*cgCneQG$4WWD zn;R$Z>!?feWb)Hq?Mjut*)2}`bMF9e`<_*xZs9mS( z!w|O{#v8SkL{|{*vGvnQj)fhpFtk5===3+i5@FE}ELPbInk3BS8(1^lp6iKAQm7T{ z!blxeh_KWjm`gYgc{@LOC7ktjleze0>HSGN(^OIim=)ltnZ&{#y64_di6&?6XQ`V? ziVpGl`fy8_5A;uE2PAUxB})*}cNe}QV9^uxyZu$2`IY160uu_>%+-?H_DQm@2rDP_ z4lC~R(n9LDt!PJqa4=^O{N`k~Ik5D-JQxyq%FDkg-M_2*t%ViV(A9{#p>h<4meCpF zN@<(8VG!5N#cC@-@+!>obnV;vP*pvXzqU2%imNu;3{fI!0)B92+!t%MSO2PpBiP8F z5Hy4gZ4l$HT6iP%91j&8b4sqlXTk)XhKhn9_Ms|AU~CtDHuq>ch@RAp@!E$a3sAO4 z6X;S$9!vy1D(8frga zvXhJ>nfq_Lsywxh*aV%9TK0B9@{%UM^dw=#C9)!XPP3)*Qo%Xs4*wxVPY*lJXHUTx zw8H9S)G(9j`=D5WCBPVLSmr|P;xr)5LWhAtLo}BAjebRal9>(;4yS{HK zVyJWEJbHabMK%j2+9N|UEprTYqdf)DAMriOobIVgR|$r~>=&ENNL!ckEw%LZNX1qr z7m$RS_;YW$>3T76--Fu+V{OXC2UfT`U zz?LfAJyUPhoBXu0>7=&!?m`P>h&#s27x#xolbLS}576yDr)5p5eAi0(gE8N8a-qMk zhj%5{qM@RLHR0`XNX8#pNpcwTP&wK|W&y(t;!bbxNs`_csEpuBNCu=vt#d-Pgaewi zZ>egY;U8;PYi`E4N>olu=$?6>0q;w!A1Jkj_-_>ZtI|sEWkVK}+Y%(0GUl|~3RP4* zMZ8bvFhjX31nt)#YMvtbVvoMQf+R|23bSk9%uVriJa49gWWwxDDYE6ChU~A^5qWCK z*YQ%?e#;Bi8ecs#obl!=VZ&qYwrl>nURPKcZ@%|(pj}SyRXb^r;i)P2&&PK5bVABy!aQ2TtG1 zZC03-8Xyr~I}%r5a5A!2 zX#2h;6LF!e#>`UJcSsaY4)ynC%`!BnFJ1&5LMe(}_TzfJKu3o)^+*Q8R-1uU7`Uq< z+JXunUJFC$4NqlTrNqdxumvucy7rHLdPqx?;g4T>KIl&3j0e;b{&=^o zwmtwCzNK^agCU1^q*BTw;|dq1{E@OqOl`yKm`@IF>BpfAACorBYt*?e>K%ch*KgZ) zf1Qe=Hyz0u4nbAye@1~`BI{e|^-s5r_E~+qe*Yl3y5~vjq{;;x_@fX?wH2d z9Vcgv4_wO0Ei~FofqIK$flB*FHF%z27~(A$2QgrP{eE|JZeWjWF;Jxuh7b{R(=ES^ zTFXjhWO1re@;Vr0DOQ=0wjup@cUbGwcUs>`ZfCchu{dn|$Y-jNdhCo#aVBJ6_zDlF0r+cH{dsgd;}G zgHu5IK4rb5u=z?j+iUh&*MZd!d#^%Ij@6SwkbcW+H$5=%^r8-^67RjWu?Qf`%$9YQ zlnx60qI44nF6e6NLxy`~y`ryE`5Ze9Web@t9`YnSt`6)tr^p(6_#9VJti-o;OXUQ9 zftHz!V7+6oR*kP8VSU}u+pFZIdQ>THbPb;cpASavwdcC;eao7ilwZ`2mDrx`z4y(r zu$~b)?$=@G+xyvl`?X@*?H(x>_TWADp!IKmwDaG{m2iuzQOKC=9l)5fn3)HN@5Ilq zo49U9$x`MD_CgjFV((oCp;f_{f-xl)y|c^KD->>QT<*dYTD`QoFp%rJsv7s@71wXC z4>9A%`nwqM5{A2QMjh(Em;}f+Ds|*izE5Mq=M?z9kh>3$J9%DRi%_@ z^rmwOPUf!+&k|ox!uFVX=J>=J4+FpOKqtw>cRaY~{KF4kE9kmmqAL(aE`0kzw^>a^;-WVUKTg6on*qb_okLFgI2}SaFSDe`&HBQL?9t@6$G%Jg28kd zcV>Q%l=HKN2fM67-%2X+tv-Hx&j=40iEGH}mRgEg3&0eJFObN%rgfKr(fPA);-u&`q-M48p1&n zq0{TPF}-xd?fMTGZ6h=APjq#eX9nrbKjBL^JO{9SIK6~?EZ+^Ymbxj$Z{yV1xU4BF(>gM5N&EK-#$IHoZjp=S7DBX8o z)4M!iwM&|(Rb@>(P_LaUk{ft+Z%6#uRrBSs8;c38*S)=!eh;UQR~m;E)GaPTS2k`F zx8dGf?4CncSp!dMh)m=fIhnY&Nx|@A=*09T@8;KS_lf%^&E2&t?+%N#RCD-ZiI}?l zYszi@mbn%F_0}=6f;GMb^I;v^dwW7fW--7{Q7 z?41DaTpbjgFoA9NhH|qT2qM80s}#FBoJ{qSUhbGuZt(ly?rQf|Biw9oRv`D$d^9w3 z6YUN>U{mq1{^ok(vKDh6J-_(;(1mGaXswa`a!3&L^!72lMiQg_iA`-pJ&Y9aE!^^ z%Y9>Uytc$seKD`hyZ)H}L8goDr&Bsf8eC@+SD&o-4BaS9xChZAe5P-$-AP*F?vZ|$ z6;WUz?xrh2pg*Saxo}0LZGQxNP@y}y5Jtn^8CFI=rhFxz9%}3(27h}BQ0UFE?wa_( z>)yI5Q@NDw7J`SRpjoaQt)x)+@jn06T5Om&MAz%B@=_-8!ROD{U{a#Yt0T}*nsh=* zM`4u^e3#5*8Yh<;xLWUOZ%+D(i)-+8sa=5u?dxO}3*&U!(l+X23(R&3w#QpdK=BOr zbpk80%0%yEVg|o*VEnwi#6-XN(lp3tJ-ej^XyQR&dG22s(FsUCtX$$q^bweT@Fb_S z9lSztxnqs|dF#t4dyuY8vW0-;qy8bky=wL1&k82GMh9^2P?`C3otbeu+3rz@nDt-; zyQYQuWcKl|>6-TzbF{Z&WII^>(IMu6E3V|vv8m=!!m(0Hg)7ikb+lneR4J z4)BF4qkC$53kJxp(E{uTEr4gGfF0kRfBZ(eHP9wvtVs3AbLxtc1T<81A}F~o24g`m zzEolevN8~f;Vg|td`_OI-)QGS<*3yH=z}-<^Hwogf3&5$R^;2b7CLbNGQt&8scaoO za>=$uA!BB5&Q%mo{gdBgAJGNr0=LO|1keArn z7|As#gw*ZK9ZODVZ7_E_2oG56Ktp!r?NxRBIG?>i7c*R`qc5vn&v8_?E-_y{Sw6lc z*&;x(oq^7DmU<>GHvmlo)>U87SDPl)tH2%(nh!?-UP)?%)ypn!zbOW>94;FZN-3mp zt=KjVl?li6`K7Bv^_nE`7yE+yrzV@$>zsPsD$&wKO8BIFxTFaKNr$=5WxN(Q*>{Y^ z`4Y?9R>vl_%5I{}(L*|JyhqJB_+@a@6Kqm`ZmU*$LE-6>=1EKPlbmKV{hK-4Qr>rx z1g#Y&EQ~q$0v++@0y0Pf=-=-9(aOm{Z*pNZaFwrX7(VxXR7Id3;)@N|*qb$AX)Qdk zjU9Sin=);H#PWv!i7w}zw;p(99-zjGLMJ+xj(4vmW1UmM?D0KZC~n53tj(NU+8=F) zPJc<3d60rYlfJU4k)lKQzfpi)A(vV-FuTFu2vz_O?zmgFA+2rt#7aHG#b9?Cb2X&l zfFonqrV>DUPwHa4So;!}emWQbXha#ahMb2Q9}Poro53bnpwOKlsav}qBb^ST`Gz4N zBU)?ezl1^xuS)N30er`$eyMOx6_?=AYuEc#P|&^`oBK?l!Dur3s@JWn)C6CWPI`U8 zTKwso34xk{^j9t5gkh$W*>@I4dh%h6OzD7(T<{L;7*4^zW@;rYVbt zEYV5od*c)Fjdc(>2)IM+tNTUfn&@Psz)ZfIsR289zYSCyQ84)XXPUQWCJ}cuPWq+8 zgzd>zqjaD>A?=~`-yTT7b3SKXeeeb*kWXC9Eo_malZM1dZ~4`&DO5ssjJWRb1|+lw-oYo)e|$Z?$1O2(ns24n0@N2=V+&p`hTKvDxpV8qTE9?$Gep}mWBWxo z_;wqg(8=TY1?DVOi}%Y^&FHVZ>i6atM{@Csx=nt#(JB^hr#!0R`O%Q$nWVnnWcK17 z|GoRcDwvr81B>WzWzQSW%E4I_%xIR;K>Jp;^I>)6#G+BsHH}&E}e^qx5jSpDp}pX+AAhqJmnC?o-*(R+On4JsbMWurp)=N3aNAaW^ihS$)-8aR`R^1 z#DPq3Lu!x6vG4{kQsO8AEv13wvj|$Vk%4anGB<&^2Al=GJTx`v;^@MM`(6e3k@QXU z9s%&l;(Ko;gf^FlZ-@UDN$WI#R7BenwiX$Bq;{RyC>yK^ThdmuU)u?v+sV4d0^>}; zBm5ya9&0b_>a`S)ypXWCcl}?c`?~RUBdTOVWAvS9=tI?A2d$8lTaNf5tpK$SwL}rI z%!A;&wCvF-LehY%fXwyT&XY1bcXOr_A9ogvH{p1WL-fwfT<7A{5fRp1yG*ItvkVDv z)x{I(nG->Upp7QSGf1X-N|Hsctdqr47<$LP7C&`B@AjZ$u>TTgJv?6R%FF(+N7KzG z{6`1m#O2RgITQ!kkK%_SopsL)^q0kKB*hmFwKX#*P&mR6CuVNy!c-(E?K9l0o+GXj zThq|?gbVRXqZ&AaCB#p`}i>gTuZ)A~-sC3{z4-ibGcb>k#DNIV3nx=qkwG zU|6>@ZRoHJE%S?ttTv2t-_+E0PeGG)kxcDG$a`IzeWeb_lIwXW_q_8P7#FLe;B#wb zq7bXKnx`F}P*RAz-AYdkU8~4_9v;8VW+Wwx%$^EN&kWY3uqmNsIP^z>66MEIMQYA_ zNRh`c)F))2uI{>Qkc_oYIUu6YGJQhTx6ygkf}gw~Tu!R_TFCh$n&5_~(c8^jz^XZtajLybjFD()o7ERqcW z(*nB@XzP!09KOQ)GdXtz;A2N%@qkp@dNSTo1hp)C)NAK6&L-2d)Yb9%o%@UfliSqN zhmLLg+uW^%0Jh$ks2aoDtU&tu^*Fqp{3-h(?u&Ta7XDfS0}W(|V@vwOoJSc@qL1Kg z`u2@IE7RTzKb#f1}5)kwccO8v!_#;P;A}S+WS$?`upT(LBFrB7?|Ift8`>G zqPO+Lu^ehf>XUYZ#w?GFX3`5RM7)VytM8OzF|o{KYOm7h4>bGGl9O4M>K>XeAMA|2!m`- zBFhYi#_dXj@W99IWphtW=kJrt-Yrcgsus#O{`5_MFLuLH<7*HbLIH#+ui|u5^v~#kNDQQY@)?Bqk_H5FjX+z0o)5&+`H#o_HOE-fSCl; zvz+>sxKwCj2_EQ(%2+03SMg|M2+6-~lCL@)&3)T_1yi&nlS z!N(TM5Dr4>AMx0;An-m9-I+M7sKS5SWhXD@ z5dB==Ic{`3mwdxNU^)Z1qX4~e}XZ@#A`wp-|yZlp-Xd?*JxMcL?Z+U?nad8^( zAA6GgK|NF?CZ_DhnlH*O5wv82^;bP~&4pZw*SQPl1}1Xt-{o`+mYMpWK_&>;!e<;b z8{T6(mWnzfDojQ<^~*xYLapwmd~Kdaqn8ZA*Ni4)+L52dG2T-8!rgP_kQjGwNY6UM zdlNEuBF$WRoyMY&>!L*KoOczCzHoTEqkB(0W}O+9+u|$dyk#*e$X(d9uW24$ZEoH{ zG#V6-E1V373w^Si7j&cHl);uvL5FCE!E^&#Wkf@)gJh56$SPCc?!0v_PFwkt>lDTK z-!E}+0nozbvKiq2f3dNz{f~b~43y4yr?BW`s=hpNfukABnZY>VA$`@Ep8CpMH-12C zz&oJ6xEyWJ@FxZgua*bfH3}5VrM&4DPYGl`46gt40|J4Nw~Sj{>uv8{Bc>(=Jvl}D z)1Yf6?y-R~5SadGfS?$-h6QSzAjvX7FOdXwPg7U?#djO~f(5Yc?!<|O>}^x8oL194+1D11K&Ooi*`d;80J0D8FL<}*we%4lO_<36B=nI$pu8dA4k`mbIEpPO*- z&IWXE!P*+5pf;_D9JyRjdvJ{E3^L+=+>lh4hwv73i{m&T3e;Uh9;~#we|@MpPpgPA z4IHlo5r)7!;FMI8gatX&lXL^}Z@Eu5&mHqc;|3CPhq}|5Wpmkfl;Ah)Q*sD;X2|Wl zoN2>^^8K=u9GvF<*ijhOK7W`$6?W~0+eQlL%?hj#yxzRirM#{Hwx83!oSn%m=<$g5 z607eQCDfW`S}>k4G1CPl_xH^#%ce3>n_YdYoa^j-H+-5b+2R{M-L{e76B&ehu`kAwSRZw!c7HCfopTaZu!@? z!L|6&_-)h42fwgFK6}{RBhV`pmQ1 z#*_%2!U%uje9Mv<_A3%VKhe3oZX0t6bAtBh^ld;q60~qmWyKNL zv$$#8pHMm3UEW~k(7^poJ!WWo${!r6FJ2=Q=3w`M`a*=}3DBTLZ7yDOoVGOGYrxCK%)=EAylhyKyxhszC$eO1J6WE^a> zqt4u)4E?vi0RaGr-zvipRw+!9zAv9g`^0eUF0)X#l0!A{K%5oBA#K|oX1)T&<|DqN zTjBTXS$$zZmZd393)0XE<->}2Qypkl-zkCn2q5Qb)I2G>iQhF}Jr-AUBE#h_Y1}t- zxs>obsvtNVI{nq!3h=Piv6?oM3lTilm92sEy;RM_`zD!Bj_3`vEkZE{=Ba|K#{SNS zknb3D?<&W%CCtqMl2!tu7c|IW96yL8&zns5X@u*qA{ArAtb2l!aGGc>MfPMlW8h+n zvoIUIJA9QBY-(Fh%+bEC%p}?Gmb7*$?t|i}s&D3Gzks(}uPZif-mU@><}VJ5o$f=5 z@@nrSM2st)*=;D>a6zvHTa}}K_}EoB(q3o4>G$hs$i_aAPoUu24Gl*^PcmkNmsk)+ zyk7#E&lTYAbCGhgTxY%7QbJ*7EGy?WeMfGjhRhE`jw{1sSS3X{R4=lFwj~jsJ++s& z#tY^5LSpmNeBlk^nz|xx{f~Nrt$ON{TR;bDi|UxNkdsHHH@+iKC3>SnNcmhS- zq5&O5M8H|nTYhCMtIc(`x=G-+^LkgIpi@nOmJVyvat~j^?e~t8o9}CD*?~! zN^0Y;UAYt^wPUE7DkR>pXZ5YO8 z8?Pn3;H&t=1munV=a+)FhH>J$UW@j&%~kF*3ERtEl@smo@v*gd33d(837US8JRAGF z``V2J>i%!0l@xDjUr4hX*92v>qa9!rot;^x55tn2Cecu25)CrAR#pXSg z*q3I$*lS5Z9MK7vMXnS7+5!1U9jTq6nYVYxX6$2|!?uS}pc}5`5g5$SJ^{ zU=i7^m|z-Ur20f(TpuJ_2sZAq?1f!4S4J&q?K3}I=YiO({4BtWo1B{ny18(f5PgU42*%gFjZ zaFAgG&|No>rN05a)|dOy{y{tVEL25sxSav|;DcNmiY>5>m|Vd?5X z!CHgg9`%J@KUe{JJq=k_<4XtjLw2Mg|H^UC#B9T@6yLzPJ6c=FgZ(A=Dav+0h;Vo zzv*vv(MdI+C;z`|BXutHio1js+|^yN}>)TLr|1I7e9fudD(`S%oVkwSJ$v z=HUGvXtubX-UiXtMWIIzO#KH$sy$Jnqhc?p0>qTJXlS|(S|XN1Eh5=v^hmeh;fc*I706tRJkL=ni>R|i#^MPLDv(d09}BF!X$QvT3C}CXf^-GW6?o^ z0aZVg|1A|2L^(_(a7~ta*RulpK>!nAa!O@%5lA0bov1;~0C~vaBxG(pN{b#@fp`XAJlCz$xJz~mqX^%|IM>*L@5Fs{;V9X69=HtaRy;|S}5k28&a!>CD z$%Y}6{8{isAa|r5I|_Y|cUzt@nVq+8gYmJFdZGJKrR$=xK%7-0l>Y0GoI2*Jr=BF= zq)^)?j!R{V_Q783-kYiRHU?unjskSt1;mMcU93FVT;~CLXpUy;VVM}IbvvgTn|iy3 z&`9VcxZ66+M^l0Jv8h=iA@Uj^`JGoglLAP2Jn9OfsR7}wA(l! zGfjVZLKh80)oqt-Q%WCAW^2G93J$|L&JfS#HIZYmZr=RZc=6eyYtumW)ts&ryDpgC z!;hS9N3T^x=x6l63H@BYkCc~ugApe7btF#bO11a?^TTVP)d$%y_c?$yei`4>z#^EX z15{~GF$8P{-CQie)AKK`qgDLJL1bY0_9#pndh@8C42aaJA>Ch5Ov~WrU!ub2$u0~g zuC;Q@MMRgiHX{EEm+=~b;^+;}liTTcG>wMo){CWy%bq|i$`-iy4f1$rU?=(h(iV*X z(3S3{_)g=H6PyRw9?=&-N7vo~I^@!GGXu|E#+Ofw#5zGtFpjS(MZ6c|F5Z5uDn%7J zz$xa&q_jwZXb?8fIWs*>hgFU-+fcST>N;;En6WjfjOH-g{W6(Dkw9`(u{IrXcT~M%OcmvTW_li0*{o+^ z!Qwma?Mrp_f`Eufl=5FcAC^*P9#9B6`ueA&19F5+w5s0NaVeZw12deO0hjW=r`}NZ z64cF7YP(l%ZSn~n2nS`&^P5oH5MJe_bQ^7^OkWiY$BHL4s&w1rY0=icw#8Oy)4T7W zPof29Zz7LN&skD?CuH+-Mx`U&x#zujsmHr-k(m6N6=9B2+aAVW=UgmM<#Nl~SiN%EIdV7QX&WHvMlOv-5P|D@x2@n#AU;2_2N zJYV#v_m3+m{VyviYtPO612uBq>MYc_7AQ<{-|PH|9>3XcEe|`Ed%^Igm`uwLKbuyg z13grkHCnq1%uAbCB>U8DRTiMK2!W2{hd&=u{CEKdFq5BAC{#r4m-L;Wq3>0nX7aVR zT#?ZwB_-Qr*UE^886+y03D)(eUul$JgeHjus=p7Nx*=|->zdVCj160v5kpes-ut3# z1YIYvRCVls{aVdgsPa^TZjAK7{k;vbbO*LJzl=@uOh~X!XjS`PU&qb%YJa)3XQt|c`b)lw z3Be>$S60BKw-aPB_!hgOrPN=3pBYd;_p0OoA9*c2;!TX~B~3Ermm%_)jpYFr|e^jXZQnK@{BBKB2$A4be8Kz&42k_&a@B)y4H{# zd@posJ9Y-BMp<^3u^d##jfB%uaYq0S{ER;30W^u;M`<&mcD4Iq+AKgVbVJN_Chb7a zZgTz6Qw}mOMXoen{eL(5BMNreO*bDcSM*7S5@c8=xzNkGi?NVz8c%Kk88#1l65k=k z9^U{#ANw(38d!tz_koj@W_dS%ah*oyvrz6scj2`AyfBIRr-q@3{51bapEZDtV5-JI zD#Zedp{}*Jhd@$l>mX*>1vZdzM>9u*Im6NZmqB&!U(fU-LS~r)UUB%tSo^-LHqc_? zlI&g~cCK3<9wcP>j;?zvxX@x5?ME1L>vOg$e9V_-UwWA6=zYd#oQGC&I?)fvO8#R2 zVb?+@|3SiwIF7*Pd*E;ji(M1Ub~xMxlI&D$COqPaOcb(Fj{ZGvnq=Zqe{FE|Q8+~1 zxFPiA1yBw4f48?z!~YManj1LDxpE!M7^01)4ToH!tj}!D{Grb}P3#Hh!%YoGV8F<_ zw%KnwR@2Qd1s~q~*37Mp%xXkCS5m}7Ko2-A*DMgQ|83iiuVvl-LH7+)K*TOr1R4=W zstD}r>=9!yF(oAw7LZToY}Rz-g87yTxuP^}nI@}eF<8ISzrjtgA7|-I|G7PVAVD3u zn-jtwbdjb^g0sE7J#B`svMtZ36iE8yicc|KKgCU#v-6~EAj3unh%GbL3T5}r_FQ6} z(?AU$6%YksCw0UAh$C;)9#)jUwNTs!|N0PNtsy1Cp_W>w%$Bz@HaEAu|9I1ChHtTD zF0@8M){+O9Nm<&=CuRqp5GI*_`HSDMpsoT`P4t8zq)8J3>QeZAsxWS*IaVV&R$ZJ= zU4mRj%`=F*31;qkV}aNw;%7_i!W}G@_q3gKLNP;rkt@~|eO9aPv1-RosvEu}2b2zzU&-7m&15E8UIStwLE4AI*7Bjg&Oc%mtVg-F`7dLlcEshkLqWFIT3 zz&KIjskKxuv3nsMeu&Xz$Ns;rZc&BX580otS#o_*+I?UR32D}HUv|IQ3<5k>VAAwd zz+V1}7eIx70$$r2a%mn?YYrv~!AUS!mBM7c{6>-`zw$Fxs!fjJ9tzRP5W?KmUUn7p5{dD7dcpsud&>ZAT_r8-c@ zk!oVUs+uY#p6a)b=iIG?$IthRt^urSZVnT7Hu+!gx{;Y3YK&?sVKNJ{r+sc3K-L45 zK@q<_NUmU|BIHo;@CMb6T}j(_iokqR>=dVTkTeL)wx=Tn5_X(gJKs^Io3sADAQAd2 zHTls?fcd}Peufpt{Y?G-u4THL8PL!dSxg?3Y$+VXN)^=Y)DWb~K$g9K6c#bJ8wUfk zekl;rIHgfY?ymi?fkiMM6(C#yxmNy{2Le8;Z*b!bp*2E`e5PfsU3`D}*&ivFJe@Fv zZjUtO9@Pq|Zf+VV=6b^3iEM^h0?9ha={DBdihDw6;^x27S&vn%0gRSob2X4~fm#hZ zu`uZTZkK05g}4lpq=2sicwb#}knVp+32KKG!`VV~_mFVcMA_uqwyVth)a7Yk5l|b{ z1=TGcdy|UDD4;qiS+CF8Tx)|j2O!6<1LQ$ZriWq0@%!Wzt*s}aG2;LkG7nyHRasW~ zuMJ@*0vh!tiAW+z0D-W|4LC!s_Y^K@*m^Efu^ZKh5<3N9Sd#iaDW`?J+hiB`{UiC~ z;$p{`muK6MRdRs)rU+)Bs?h5+0f`kb_n#Z6+KZ5F@sSjML4{U8g_g)eDjoLY+7;Ct zgYnmGukST3N2XXr@Kxp>XEE~}j}{C)*b_ORuAeI-lvyOeYt>wLOG7qd-F+tBkV9>O z1FYT|A7}bt6`Vn^g?(!M#yU^#2{YEl=v=dll6n3S_xu!Z|unSrPLkzmT}4nu&t zCjOnx-CI*4tb znAYcYBzfDij!xT zG-O9yp=Ne%n_@muppvLig!;I_#C^~aj?cKviIuV!yO~xOWI88Gau}5I8S=+mlzQ|a zrGf<1?sNRc;gA47YIR?Tz&L~&eEqnlz}H$c{2r5F9wMSHKJy+roftcvpGwGEOQDKb zohP#mNuYu=8tYnNLrt>C^;CSOV#FofZgi?w!qu@ojtgZMQ*XU{taf%)>64fEzx>5t zB&hL|LyA-=@)%fVP^K+`^u@sm?O^l4#az*qY1|TpKU4J`GYxByol~XtlZ%pDks0yqGC}vHcE>zB z#y|r;L|R@RFM#YhCRPRl-~E2$zuqrk)5?;L!L-dho1%+bL`SKp6t$52d^mfsQ#5X* zPQGLNTPTIUplUrQ5zu&@$zlRXI9ofLjMaU}P>yrl+kI$V zh~frO;3i1;i}CTJ_lIRE*AIvZOKqh+-o_5^guh`Ihi`)lDwbqdGXY_IoB4$&Q9fM zz;PUBhiQ;ouEqY20+W59dtVVA;ptvcUlji3si?vN&cQKS#qx3WfjM+KyP(yhAfQ$R zP*S#Gk>%Z#%B=sgRe(P0Hvv}eGI!pBAzEmFRG&wB`(H;ESOVsYP~+hlZq9$X5i0vy z2BY~anfNg_`CH%tF7y*;OufVZjxPU}oKPLqPdoQ-oyOnqc{J;2FZ>*$J)y=gn*K&Rq)D8uc$q+aK_WVj8qcF0aq@mTnyhs^raS z+BB^?PYTI+cE#`MU*tM94*BO@{9$t>nz0l1%(KH9m8Vtgwo_y%Y-7orwDFKPsk8mX zw|!-F&%R2bg~xW;?0VDTV7XerW-jIFDA+p$jX#&$v-e3VmKknI54%R zzZFuj5@LNL!9+KlP*7uvVHod=j4yUFD+n}0`Z@&0@a44ZWp|PA?GvEqWkqXo%DPeu zg;=UEPa!e!C+e*E-q{Jf6E63Usr=(~mi&h7+GxfOHq>r2{H9NXr)2ml@9e{0T)EkrR;;0@%4)h3Xt{`CjbZfT zKT9M}Dp|gyx4f7dawn=T6IrFb0KabdvYpC6dXf1icLt5~Y^Q z6QO(Fiy7Ox`xHtSAz5e2bDURFqI{`eU%O$XfMr{VM41CD6nT3)l zVt&U^|AuFRu`!Cw>xea5EFyH5U+frxwMDcx5IwelC-pUN^)`Js^J!&U!1aB6xhZb0 zlRQ_CsgAcsSI1*Q<#BqWn~Os6<@F6bNlfKrT71c$?k1@E;S5^bsQ*!tO2XUi)g{yL z>Z&@z1b=xIp{9RrH(iuu>P#SxI3b=_@5|%8H~KOt6xWv*-GW-MztwkGp}xa_#e9fB z(lMT3Esh7tsD0xjl<@|LiT>iZe(?NX6hIS{9KRQS6Hi|VYTU1*^xRmIkE*&oAB?(| zsSaUAWQH)vaVQ2sWv1%|uI+?q1!Qa~TMybgHTQlxN0fF*i_>=S&VP3M+T&2>NRGO5 zf_c_>LKr;$;KJ@hUT(tq$a#&z?ZA)6xtm59p*tLOy`TqLk6$rS1&~BTA`;pENwb4Z zD43J~(QGv)WnOhEv)g)MxPgCGoULKh+#6~2dArgI4B`>=GH#}^zBa+miVEHMvTfJ- z=e8d&&rB0~C_1s~a>I_p-TQI8hyEB)9WcHnZDm#YBFt9TOA6xe$eSc-o^@?7wNY#B z`x3RNmLbL_db5F@;7iKB;Nb6lojm1P4GVOV-t@f4%3>IMj# zIKrP*j5N5fYU*x&9@;S8x6C=F72$?@c{9a$q_rolEm@lAlUrb*vlN1S{N;!2HY}{>^YDUpR?UIekj4<9^NfEzXSsTK3|ixqtzy)UBfwHo zq?Rv*j}4y2_hBy_6j-UzK|mat3!ozb_q~@{ivHxP8-FVB1h`cg&SxWHIzDE&jAQdQ z_hFfwr4dX3sn!2msjEjU^kqu3x|5aUs%VR_ zj#B_3GtnWSjb7(;CTSv&#yWuELE6iSbLZuSPh8i!k@V{RxeuF9TOn77kXj|MLr%1+Ts@}f(rq*pkZC~N>c<;k& zu}&0hiRF}43~_a$)L=c>sNK4ju7pvz1Rcw0b{2LPcKpiIvrm5zyHPQaKX7lGR~3A% z?tkHUu^g^*c2BmK(}&XSIfHEFAg8OfFuIMU*~^eLey zZeCl<;-$-jE2C)FMu0l_M-EQoD%C|@C$56_-eMgJF_#3^*47HHj?`>2e^n{0a~=eY zu}X!@IEPMF*Q+sd!>x|8zSj(K9yc)g~3CR&SfuZZ0 zJ1&#}CgaRU8CCdqh;o@7@MQ6pxgs6m@{tKjug}kQfin7@n)uMZ55=Xss9AdtNv-?Z zMp>pA%GxvH!I`v#i_p+=k5<`*T2@%FSUw?bpHPNX@vZ}otKcj zE5%GCrADI_7cV-lzkj7#quA0a0g*U!RkHs26{pmh+GnGr(+#el`}iS7m%f^vG%iMu zf=&wN7f#T{-0H|wPktC*?m$erFK@PfHcrBO)d9SBamZ0d)-)i;=$eHiokBa?&lm$5fFn|u+i#Y5JwrDcd@v2 z@Aj8-Qk!42wEbrW`La}Jz(|R6p-JTqJht7bA{~!=P_&B3gfBO^mHRW`&k$d~a@r5p zl@Cu~TvM9c0)3z(W&XKbY2_a14_B_uuF!crSaD4Xh?)XA((@PnG~){mZ@41i9m|pd zoqna=W`_3w#rgDFJ!NNaOf87^|6}Ys!%BfY~Yy|>Vt z1w>jxN4gLK(xsPBMifFvLWy+ggb}Hb4Fb_3Cl3drG66Z)6y=$%_BxvpgR2N;Qql%bumQ4_sbS1cPb8luXEuy#qJr=PEUsAB`stvWEQ&d~_A<5g%{3J=VXF2716(Oo0O!!v2= zbcLFis@$TP?b=;o1u_5rFl64ojEv0md^6S9-$jp~d~daBJ(|1_H)Zazy&^B}tITq5 zj_=Vwj(B|V>`K2dCLFAcd$$^Y!}Hec#jN1j&Pd3(mQwq>GBb2k_KZZTkYCzxk=X#x z#D|YFHO-MCnT^MPB=4_|7z765x)smVvH8wqRV6+%FZgJnQFiIVuB;nD89FLS#Z24= zrQS|xqt}Q#ykoLGwzaaJpp6zxnZIF=twtS-xwbe>SE_r1=3=BG!qaIxzWTV9xoyys z2QpTbDMQBwMr)t%n(6+moM82_Ev&Kw17)JTAj*qn1E=S&AxG;nRP-gg%AX2x4?~$l znIBzbT=+Q*8TuuTNiK<%kqz9h^B>pzY4FdFpQ3+T=zqufiIq)c)X5UvnIRpZ(HJ-* z4a|{8*I-_cDm}im&h#uvX;fO;#_^iep*tuj+!c#-%hT81E^^`G=P{a66oVILSMYeZ zYA|HXexmGg(ZD*oM3vhtXjoeG zkHbsTs>nSp9A0F90&bNRaQY3?gxQwse83LTlM{Ps`&i6_-@Dmkw+Jkkk{@;b%F)u| zXnl72Q$w(Hhh;k2CzC-DEbjF94%Wujben)3=Ims@yjobTaqNQAKmvzLqTujk;8p9| zTI9AS)Dk_`zHVA=bvJA;_Z{NOO!&d6N(Lf@ilO1b&6JoMa6v_}#+29=e&xL=L>4r} ze14ugd#TrXA@}ggZI$nXNJ_B8k?i_!Dk&T2)J1R<1+bGt-%CJRWW4zAn-EhC#3H2y zmG|yJtnbcto|wz_AAJhIi&5{!rZ#(X%&#IYG7hAriB(jN9-#b>~bH39a1n z$JEBF9&hDFi@Es?DLsbQAll@Aipz4GX*tm@!y25g+iLJ4)MAceW1oP_6m2z-gzHoy z--@RAyYKNFdU(Obtqnfw=Tjurz-(nC<{kw6n7^cipTGe}!mh~u{DRDSb;n3> zqaciumX_2>kY6hveW{TyFfl9(et>#^mh^@eqPWDYRm)-2mCwA~Ms#Ataayj-9KEOH z2<<}l-Oa99GA}bGZbxk(OH=4%{Nu%!wvvj~Pbkhq`nNdn%q@Rj9}=MtMTS$u&qqkN ztEMCXLnq_lAWmnhCHPZEyyOgZC2VZVYFX9aCG*OZMk~iP1_8<+D5>q=_c-2zFhWph zZ=~6>`P^<@$yRXLvr7A(Jqyzm`~8@hn44$nCRUT7oRon(Z&lemW)vl<{rsAF=h+8f zBYNX^m58B3?-c?L`&zd<*BZ9fp&6db{8Cyu3u|MZg5Jk0n%PK>p@18%+Dcgeq1U z>#oU@KU~G4&)mobIBNXK%{HbhlxBd<&(oGkJ#qePGuc(oZq{HzU**ApFT(lR{M>?- z(tA_%ZK6aAmEvgGehQPdqf&zYjPmEyDtq(rApMQ$=4FA}o#C?jjg--r#^Z* zof@EkpC77+ApH1?gfWvAoew&-!16S6`@y9Nt7ozFK1(HITUw(7d6vr?LsBVbWo?6! z3|0xxSXFwu^``evay9=hy8YBf(H1v7!OGo)QbVO-ZNIq{>LOnZI}WSxnMSMmlv=6m zM#}n#MJ83;EO?WGqXk>v`gQA6om26_I~Yyt@n~^Mh7k6oZT38geo~i!{fKdX?fSvy zO*r8Yd6h<_JAJ zbr+3*_XMuw`H7Q8(!qsZ$jNEvn_locwAfIB!C;ldmZHs?1>WyrC7YYg8x1>GcCgkx zebnV(k=|SWWrT;!-{#Zvo7c+-v7b^Tc9fz z^jNc}Oq#UNcA?-c?DY7;Sfy9^LD|$FR;*Ks=-Z_;)q24z!cR^QwSG>O@GOH|mtjNi zuc!2+@T1{Ar)>(GGfnDkZOm%ywCZo>56E6#?|OS_Xd!bg%<9XBib_kI`4|X2q0kHg zGuO$*gaG(HJ7jsH(z)Q^z&SP(z28+h^b}8tDpRmXahCK>uW-B;`krb^Ko9rEQk*wkh+BqjYijsB`y+ZsV!rf1#Z7i=!d_;6nE{dxRl70C985U2 zF1IXEz8SbvU*c(soO!M8@!jpiWD8x$jHkR=BTsrQzIlqn5#m`LFc2IO;XYb4t-QrB z<&$kNb@K4b)2i(L(rL|r7Gas-!pl(a{`onV&aCF2;iQIXcWGMdi85&q!f2^(9aN0z z97ckBw32!oKDqBWS>15F1o-C6xEK9vl0Alc(nq$N3ry6%Un;VZ;)bDiDxYa$V(m$$ zX0VW5PALe0i2_3O@_)1;{R6^XgkFT&FMsa3(WO(LQSE?hA9jOCKzPoM4z3)wKxN7O zK1*}GjLw$QX(<{pY{Hc!F@J=j+@rj;%RHRBHmw!L>du~YM5?NQ>E{i(M0A>nuJ##~FaBnRa%2hb-` z8N|ul5v}RfLa$I_^}ugcZRqFKuN5~zr1i!K`-tLVTYatVhf7MMX+#xVrW>u7yw;lz z!OBk0!~3qvRV_w@4XcCBO)!C8dStuAbe?<4e@_^99AVk2xt)@nHY1c^e)3vueO9JG z^c&lLJxiG*&Z_f``w}FqN7pl>D%28A+x46g8EshWiCe!F#vH&Cl;5~d&*4Q|n~0F* z=N!9|80Q~jqH2Kgq@ncZYHxV>UvHjc=;DvPg&BU^s@tP8jboFS1TB|%4#fgAz$R0J zqUDO7xw$#8K(BUgcP<-cTRv@H+r;vBc<}rjbvD~r9w!=US3bIyD&<*br)_wyVi%B_$KTVcXr|7Fdj%c#EcD!qWYP49{FqemKDr4x15 z#3OY<_f;qRk(S~G9z$b)@;hv*Ddq90Rm(kvI5?m{@m zE+@x9E~xZ;A54&vpc-47Mr(2nzvmC}<1ZJo8P!s*{Cet`#?+^lZWQ!t`>3}RbPNM;c&w~5B)yM(lONej8Lbuk0u z`9XfAX42jRi~4F4Y!bHtJt0L8Ai=*<)AI;31~?rCquRc@yI7odZNcgp?oR2%9a-zw zcr_ufBhwVMxvZ~W4bKDnEq)XANQ3vB2?YZpJ-Ou;DYT-lX_j`*?{S^(y{=Xr;E#7EJ|P3E)E(^+x1t=G%01wYIQ60ZH^y-#YxcJULWV579i)M z@Hmi6%`gVB$ z@g#Cg6H|&4aor>8D zxOt#tUG>J~)UX3vY0B_*=qHR?yliul$?sQbX73D5=@R71%%Vi-Ug&-S`+7|xvC(Jw z^0RiQ18s4o;ifg?n2L4jAtQEFnqHCypQ2*p!p(h8fC8$FD-)pNNF|0FLHrKtHf{#kR$ zWmCGwJ&}Qdtt3a&9%L?`GU56Zi%-rL0Kxq26&tcQR>WK%ZyrsBSNg}46EeUqPVYCk zRuAQk)5kF-OMG^z^A?RfL1fp3w;AZ^XDj&s`yQsMh#A-0RV;L!g#}KOz4D6<$=FY@ zI^jWFy(=iBK4#a^YqW(ecI`}2i0bLjMAPGUZ~i?UOFWc-?5>FFs^}e$!w%e20(FlL zi}Y_6Y86yC%``}+?oiOu+Dbs0lS*TG46B8-j_MPhNxzBYM0TiA@|PP#8!qtm`x((WBb>Be4A)vqtp9KweOFO3_N?Cpx>lX7+gak)r3 z%dJ~}*BP#h%oAnZ-WMW+zVypeGb34qgw)#Fm=>yOk4~*RIhkQo8xq9H7<$P;f#i_L znZ_8tP1XG>0(x(apJADk?=-#6A2M({d_2LY-KecT_5rAXYUo?ScGGf2qYieB_Znnr z_wM)yGQT=G`gJI%dugq~b1OIts|b_uJ`Tkg5(RWs?%uabO#LL>5==jFP-f^f1w zJ>r<%>J~9(vs@-^Behrm-!Rtt@@%JK%#jBbSC=G1@H1l>5&Q8+9t)?Fo7Ou@MWo)e z2hh_Umga<}km&zR?OX3d-Hl#nwY zNg4B{UJI77?UeYZNK8N{lm2HK35SfpJq05JpLDcxBV^1!mKz9qO3dxYi?dOYy<6K?8h0RI*LpF^pKJc<)8JQJTGk@nr+2PBjLYav>I?Wn4IT!0szgc1! zeAlltRJ0AcHKcx}!v_na%C1XmrC6;U6l%9BJK7JO2G}1uvs2Sa_@uO17p2VuTk&ux zQeYnD62gl1-be~?vHU>tCI~;A9?%ALz1g(BISU6V3RP( zBGXdEA`8mp-V?29-{l}q-ICDxD|<^Yx@;hl{4Jkn-ZzfXlW|1(Hmv4HspIH_7RboX zan}mTrnEJIUqxS}oU5xI=W962u}p1kJyygHK6T!r`L^@tM2Tz0N?C1-bQnL2Ci8x- z5K{Gr1rA;1$i!2X$mBwKjq8Ww_-1^B z17~W|gr$(%(z10B7>RV_BJV0E_1pb(v25 z`LUwvi@78z2?+kw>u!x89sHM9*9#<|jm-`4P;iWM$|S~jMiW;BYPdClzVOOW^O$_@tF#36%N_%jAu(4Lt(noB1~lc6JFGkbONz-*Sm$vu6L1-kcNCxpEDp z89>*y@+skK0xmP~l(WG2zCbx`Ks)1|3oO!F%5(;5`c4jV0>}~klK^0!{}2EFAE}Kt zt>?c|`6NDnvuq#+U;rO$4=y%xWVNedn>u=BPQ%nRgTnhZuL5BG7;TvAWZw_%w~BjKzVQE?zG;t%-k-nM9_0x zvutr3rDR=`>@3+6_sx>tvSL{_gi-s5oW^>|Ee4T|*)1`v`^d9D)3p0y`IrB3p+Q>t zyqW^DEXL&dnwl-Os2jhZD%+C;Wud=gm$FfcFKiXAo4I*>KYqYMZ~XxbPuP>Te(X)HXQcDCplN#2j~URO)kyxAAm+DuA37GNpdY zP`A?l%Z`bktj{GKS^LG?lF3nVS~fw`ZHrbPu}So8gCz{qN~tQsD8FCVQ?0%<*H@i;7s-m;sgZE$5%2c=Y4p11KQ1-+*w zB4F2{diXugv|GP$KEBo|$bCXXdNCo8)hfcWpJ+I>KZWLhcsco4<)6q+dVPPX6pS!N zw8na8`gewXZm6k|_gNP!Km0QLhf%dtQfO$_mKfjvfC}0_3X})IFaDj9KX==)lSlp< zkvkZ9QinPW<**l#(N-*pj{6h`-^^CMg~+-&t3(i56>{Q4aRu@^;z$LC$brBoEEsrVj@&gI-SJ3DL$3dJKp$zl6`)kC|VsnN< z9*_ZmI>C$JQ>3QD>>agksP?(0^`zC%W+^rIAwju!mco`bB%(Y|d)F1>DQn>})j~9; zUVOVR2cjy*FH#vtf3gU4czV+M*9Ma&bCQ|^&Xv&-cNkD2c>|`jD&oLw&4r^ZH>ix< zzsrEwG;ax10D~Ht`CF1yU&A3AZwVZTlRBpf#=ddbt@75T*e_pn=sjQj*9E5dQLOyC zQ)wTmSnC-VsmT57T#Q}f8QSg8wrc1#xk+|%rL25sr=EZZi#E%5``ZY?Ct?yee4lXX z7IW*hxZHD($PB{98;dwi^0)?TKLH_o!T0x!Wq3qP(yS$$@(ao&>xMeb$hNou_0E2? zfX@WY>`t6ly3)4Dj54}kT0P}%x_WxuU4V|;SY6uUwtb8BNpEnp;?{eUhh>|MKF~C& zpt3;y57G1ftEv<5DXZanPHgenlfy>yoA{l_viNb&=eqBWp|{!aZdQ3|-ydk_d(aIL zmC<7OU@iq=`!Vy##In`51^!?+xC~C5xNF+%uU<6jVgolRkh@~P_BD8R&KgluJIp$! z_IB+pnx`^gpygH=l0N8n<$nHdvGtGBw=kn+p5%oq^%`fT6A7h~feY^~U3$dJs>*;m zWOO94l<#x)6#HSU4!(#&wbx~`%=M00ZF>S$*{IsLp=rmP-V;pa!!w?QN5?bxU%MZV zy@Qc9{$Q8duq<<4dY%vn4}&U^Vd^Uv`SD@~CipSeH&^LpG#Zb;KpldI)S6CH&64>{ zIO|+70ln>^oDUXCox-eY5dVYj{t%h}8guVK{JItOg^ZIp?gmXWnwq-?8#lFFN@9{8 zP!=EyYs|h`{(Hf~yvS4SO2^RXEsa} zSmb35UVx1e$PrkdFBdnDpx2F>V2eRvNHD6W6T3LA{Xt&1%jiQjzO%|}LIdYb5Ivwh zv;sl{3!id0lC?)a%`2gxUnnza5>d>5ROQWoR0$LA+QL&S%J@q{DZuZ3|GB;WTnt(q z*a+qW#7~_KVh^VykE_k}PYaUm)iDLgO?SFSPpyS8Zt_SYQ=>+I=b-C}J+?6+&^)h+ z#$x8hV`=$p3@hXsz07z<2nlZH!GB@)y;;FJeAhMoPy=tgD(cLg9c=z+)MYYb?&QkF ztMR}0P4u)@1v46LCkdHlo0Ar<9Omr*ap{tHQc2e6X`nQlqD}vaa^-Cf=IkY2J|nv2 zgT9chE)X5neF|#`B5&OT%J(*JB(I(3z#BFOeLh%q8F#|$u%x37j72BznsFCcZOI~Y zs!`c+N5(%_e5sOZ`oB>+<9W&EuXV8JZVU3LjCY3CWho&p1voWP8_s`iPDHbez(l0=*>{AH(*XT$|xQSHgXrG6xniQ4eUoX;( zS19gRP==5Cf5u?a7x7e*J2cp!>fg6PjgOvGc>g?XRq@!{=kt;L%pMhO#gxegtT#bq z-Lai~wU4#^I|?UE(cbEydc|1qsV#H+CPT=Kqc(mKRLkLn2B(F5c8cZQDHW{)kJi)S zQOA^f&C|o9>CO;Ly~kqYZw%7wZ)a{qz+1A)NM-w`tQ852IUaLXZ|0W`8`3!*;SZ~S zKh)q_9x*AMFa^L92-H&EYoX$Q%4qu(^?js4T1N7|@oVMTehR@+ zcL*_lM^_0-!|~&dXnrvMatH$DMQAGezuXpGD+9)Vaoj+my2Gqsbz9z{&r$eIM3%;^ zlJz~JL$$TI;5y6lHl@s8oVEdYP+ZSdp0!Zs4cOO8obF)eOV0O)4DyIlIFgO^s z1lE+8;+Zjs?7)lo75q)OZrdQA{CwauR$G#rAF7LUXykq0a8D&k_oWVpiky%N-wun5 z?Pug~hu+=RetqsG2I*PJ+OqO@Qk8LJKgx-Wt$F|Pljw!P1P(w_Uc|BF27`XHs*beLxblh2~J$K~AX8x5p zgT5Q7m3;bpge^~UM%o5d=yFCTd|a)C#r&G)Kkgol4)01JAX%VNV|F9j?H@i3s~ub3 ze}}Mt&L_Z=`S-zA;U;VFQA;!Hpvp#4y9A9?0Zo8GqQv;)9K8J@6sd#tUbP;|S$Tnm zu(}kRCFHwCCr~!`cg>_{#FsdKvMSLmfi!{&GbkAyw>iiyUumNN_3rhTe}Hn)xf6i} zDeFk_?-kSn0c){w5e6O`mhNiwQ(ywYi0MxOXz}(yzHkAWqv5uJ$d2dxqu7KzLH|d zDAT{$YyRBkx-*syp1b4e)Umtg%99T~W{p@@y0HqWD{XNqkH<4RIjLCBLjbbLa-m0K zb}P7eXv<+<2m7>xDz%2Q--+67$#(BGYv;~2_c=XL-P6g1#4&RPWYW(XgGz`xcAnWT zZ_ytb!GM#$?76{T{aPa{DV8U3ofu zI&NPgJ1MTXEPYW^P3j+CU;ID*`qxUKS1w)@tQ^Qy=;B-CyiOYZPSA6i_HO7bs zSs!}1zPb$MVzke)ndP`wfgW7_Q&KzEB-D50Vw^y6dC&ULB@KLM#c8bN`%ea@DIZnK zRpH)G=k(4J@itcGgs^VMJ`L5DjW+AgMR%AvjEt1W{O8Jq=xJHaxwzC@d7w<7z|0sL zYhm~wUPLMK_n(oMu1WJ=p)R|hJ2+Y@X|*x^lQ@jSkck<&usY)CvDiHVg=evIt;R#7 zeC2QRn=^9jmk9ilINHU^K+^d-%VD};m#4JhXRuh4{8W+6L5feM7ShU1f)IWg5Jw?j z!QHQ0&@fv!D%{x$p>{D(k3qy!kzCWAbp^ExZEZWHX_jv?kD)W-s8_>Prr{KVUO^e+ zzyFBMTT%bwNvy*&dnlvCUp$7Yjob{925*)Qby%Irq1M<@tJ@L zVXenUl&rJp827E3T%)ZBfgNvD%SB~*+Z!|E3JI=C-`m+aX#a8lr2gakcaK&Emufp+-a=bNH@RGE#dsB35wROK zV(UXWzL(%0E+qMKigTNY($f3?Fk)qq&2oqxyzNb`da}2EE%xnO!|jbFgsM>GR08;( zz*oX!REgbe{HQLX;s^oZ%nFqQZlP5`r?j~5G$*~fF*UpJQD%}9U7LCx6VZ6s1Ivo> zzEo*}W>OH7yzMOR4K-G4TA++kfyKmXJm8Dk_$cYYg5^ANlJUoIa%Tp;ococlV|%s! z!-H+*J+{)FR19pla=>=W=k5+_gYyK3#Cza8P}0*s?f4MbQ1{#K^d2Lhr$QjWO_{fa zP!W5vEWdiKl_QA^+TDTl=?zGVg_~2R){AI3kc+X4kH79Z6Sc5IdygiJnCD%`q3=9O^zUj zT#Xz5#N2-N)r^@8FX6Mv0S$LM)a<#xS>&PJpSz2K2`^jc77)l7n#|pR{a%ij%H$C zEmOcd=gSK#?o(fCwL7T_4jSf7P}`YU`zIDTc@g9GnN0M5fqSt=gCSRNp|7*zH|GnK z4Grt7ruPI+Y5icZYT&f?&X8=&Abqq#7U+mcxanJ2)K^6fPL_(I-0B|W55*ab-PE>~ zPozM2hl;f$zi|l6#kG`OZ?rTgw}3?|2DHf^zsk%VlpR-mV^DKr=vra>`zwOBJ=dy; zHxtg?etMv$ZW=FBu90Gc zn5#P+z#Ye(lao{FFv3M6X!YiJXGjxKqX--n0CxxmANyXqeqFvilwvQH3SpKuy;a_&5fBgEt;PuppBG?exR$xh7n?hTF6vax-XBxQH zHaLauuaG(x&(@yh(2~$*gf70DH^`XNB2-Z4y@`ZcNVX$g6`|cn%$t?O1?izdC5w~m zlgX_igs)4*30vj2pK3RsiR0_O6*-Cd8^9GM6!M>6^t)Wf?|BJoOq=_tZ~Mz&$7s79 z5&!MutM;Cro)69wBDk5K_KCVOmo8t9r+Hwb1m=D%?BXm^x|6tjI^&qFNqa=aN8Rt= zcRq0Y_7L|r5lr}yNxo|o5ZIh0_v-glPkzG>e;$_(e#J#m^LK-RE-3IUv30H%ps7Ps z@MxF7K85y40p47xSfMB1E2W_t%Wv!J}_zN=TA#$#?e2 z5srjQ(Dk{P*49=q3sxV_Ae)6J2oP)R!*jvB+`AR|`lZGmJ1bf#uxE8zmmIeiy7&&4 z5h2IOTm0sxGKUK{!GL`;w2HjE{Of4vu6WiZu;+70=urz@#MX<_x9SRzi_hWOlkqS|_^KU-!?=M|3ZQPhC(ksRsNsT*Bet7yy{H#3Wgg`P! z5`VF_X8mt6yiP)q9Y*|(-2UefFT}oYq9usTYA|CpG zt&cBG+}4W zaaleLI9e4`3E1n8mBW~cza=q4uXm;JXJ2KIF}3N>-~@w$1k4aT;E1JwiP1Ff(6j#h z0u5!fLi2XNBpCn2uRPlp);9pU5rW?zecS0*!1`5rzKc&u>H;f=lh%%IHM<0{^=HeT z4TH1xgXsS@yTW-3?0tBt_IGy0dl2c3wzzQLTgmsr>4uuuJLys@HebUse6=Lh8-YBX zdMdiG!b%mhfWH-!a@m(P)(QCR+kSE&4#t%=eg<|#+6blnAp_!ELiMXCMahHr_bG4O_~SQn zDjvXmm?|nNmYO8k(Aim%f`V9x2|-Ewrtexqf_F_sRMdm>{+q5EHH_EZn$ab4^8Gt0 z=z|1KEZ3= z^wn$1JJ4HZt=_BOi+(3Vg1Rx2UNr?T9!bS*BcGo0I6Q5?(R7)!CQWbuaIgVJ{d|3& z0gEG*khs=Kt6LPziYzzbYr+8m0qI`N{$LFlwVCT+fbCgwKJ;i6SDxJj`opyI)o8X$ z3VU~!WzP;<+O$+~D9b@{+@X3gIv%dMtY0_cSUZ?zm6Ww}wM;6V>#+oE zeSxnARQj-E*VdC-h`7gR%c(SB$6oh|8b@|f*J&Xf4hbbczq&L9bm=kX1qcK}yr9fL zC6R*>BoaXnb#1>@Im>;$F%?89Mo+vGZiFK=M%-jqadZ1XaD7cWJu+8F%90KME-3?u zq=k2LKXkOjh%Ur-FEtGZLZd1On|*s=zPP~tqm5U5L+gqR!3Q}_n;o=Uhs)WjH4YVBW&gUxv90Pm`DddsgL zp4#wmQQ*A_*dV+nwY??jQht@i!J?>f(D{^q}2XsL!`l@?Ce?L{`c z_tD~qJ<)4s7Iw0IajMfKmUQ4zhWr58|E@FrsI=O!yb43EPoaCF_*?82vYjVt8yqP0 zgt$!WKM;aI>M%2VIq68FCvq3S8AnMfzSYrEZm?t*#yeCd%+>kzME7!c0y{$d6v`09 z5<;4oSpY*PA1r?%>bBLdiLF7yk$ z+XB2}A7xPyF4H6)BdixsGe)mi->ca`cT|Np{5oBH$mw==nR89CHcBPQbc34L*vcwz zIF#9}oSH3pf_$scjS4(bgEQhmGPj!bxsw^`DgFUJkjoFn zj#vjrqg0c5k}h1K?p_o3L_M_xQ=2W$lz+cQ|MFbL`wKcbx61(6U<)3EjAIZc1g7+L zVPIltn$u+cppu+1ySR1tt;0_Yr+>&(OkC(1WncA_3Oa&SSq$bXScBPbmDWAC#RCsM z0Az@R<(_9fqgJ=sM+VMsM$K0@o12@>1fwJ68bQAxQL*I1@{<<|xZ>EnT^3FteICB| zB8=Kd68~7>TT#oD7>S;`{N#kRpTeq|+zG_`J}=``g$uW+++pdzT)2FvK>;mdPYsze zmt?^=f{#tp5Fs~_bU-mtsqgrX;}y$sL76S%tz^nb&w^)>-htQ?v=?DXtEXtC#apI& zKG8b{OUi3YGEX38M3)EM z26N%qanhp!?*>w)aTaGsIGbbtv$YdV;^|lMks#TV9UW#v`*0pK@k9brC{|$k1^(oy zXyA2>jB0Rju#I1VMkWx3is|wRDBFFdFY+oujKZR6|PaK>&hE9`+G&!WP{3k4&oq_C9mlHT0h*XVCK&mt~E&$eaHEEles?57IV z255jY6Z0lf$T+)*sna!tWWs!jwRA{rV56C|J8>Oq)(BH(gEUyAtH|kp(^gg~U!F(E zCWF-=sp2C6u`msoskh9|bpD7}r=lcXq1^*8EZ>Sy9SlY58of|f$jQ@k?@)PS5wx3B z?3kd4-g+NxgTC9~vs*F9r^l6jvdACe?j@<>pRK0?_T|z8D{a3peNxOMn#bNJFEJEC z@>EiKqZ&FnX~a2q3*CTYSs~5GTRn<)U*J5w z_an#uVb$+2M0XprsOgF^JUh~yORlwF|a&TFjG*T{cpYJHMY{1RdyB3#&$R}8dxcqE=;CYAl?x2QY zZBNp)PXLc3!aoFtuA2#-fX!75o^gbf8rSqTg88V89nS`yhS8Sx)ZpwSA!CF_Z(`cc z`qV^?c^D)}q#7A_`N>h0jAgA(iuZtuzw4Tz-mRlP*I>&f8GG=C?5n+fLDB^@XV7Y5 zMQpde)-ffZw49BDqqD^E>xUft9}B0XciO3C0eDo3S{-LI_37zpzKn4@N74reT$yui z;gg(~P%^H6@f5)vw_7w+tRs%YLJX&%(@1XSUA#R`M5(8X*qne39a8e?P)3N60GB25tKRAIeGWzBOmeOUlV zKSG|*kW)c>#0hqgcLX0rNRHRooSmNVjdhzF_)9{}8aXO9oJuYrXvqRY6TnZRxJ z2eRwiPpugx%fX%(l>k21byd(i`c?@&A92b(zyqc*?a z<6MWzo5~RasVEMR>~tak$aC%tyY%ll@(4%!8~mU#;==?+k`SZ^BvLDIme-tB25 zM#k;ibBZ&VB=M^?LNkRKke%nC%#|1s%Dm4{GZnBGkEje{177$9&{o7InO(QEECo6H=Dg(3lcdkdG8xJ?BnBtCrjd~6AP+mwaQxK9`9ln1X%*Qz2S zm5cRDRoJXh?SKaL){cL~2e)Zq37&}c^6mFbBPkZa_V9|}6MtN-|J6)1z^0=mL+-{O z2~R_rrPl%7>aBp$a>t-pAo!3!TEV(Rn+bph8b%2+dxwK?`NA$ee};hFRqI+191Ahx z)#e70G=(tKa>alW766JMy&~N%$PbQwIQsso0l@sYO`XU{_ao9Dk`R>s%)<1~T&$Za z13AwU&Vf?%PM;{DxTs2NohO$3^}UYwvc?^aKK${5&`@bs21-c(G}5sF znju4ba)MH-h>vG>eQ|5~gC9X8{()<7>Vtjg|8);&_#lP43C6m=DRV)_Z1vF%h7_21 z*^pr7plYWv?YlyDe#DRl z*WTxqFOSEf`W$2K3isAKF)=fv?1%EP8av5N%EN&c5#f9)fH1ZhfgT&AIWr6S9jLzqNwO=^mHY-A-Er!Y&3o~uQe zLC^Xt(ArOf;DH9ahchR2?-+(%q4xITP*E1cUVjAad!iFXVpcaPDW4_V`(%S4hvI_R zH^2kdVi!~^RG*P)n+hW%iXSK6}NxuF5Xp8ww# zdLD^F-FE6jnS;zr)pUU_bm`(nC3g#|=P+90Pl?HH)$a3(2(P`7fq?;5;jq`Q{fEuR z4X9o_rLa=@(d&v~T|^trB_V|?TO5REYj1TZ;Z8xT0LY$?ATQ$+Qzpoj8X`euSPG#n z=FxfvptWb0;5w!nm$!Y;&0YT1nO*JeH)GJYzz5_#ksfDTjuYCqJN`Q|OWjF5PIbPU z&k&^dyC1=9C87A(q`G(&RsSjf5;&bCpsN_p3E=e*-sdy5&zValkpQX9=V(Of;v8q7 z3JS`AR7W@EwX$>3oot+)mZy>hWn+W%D(nFZT^jq+ucC}_UGAEhnGx3H@&o|#pM*mV zT@&u}I`KStf9dJUs3B?dm3Rnn=p)AAz` z>rfzC@+m6}JVmeC>E0b{@~fKn@yR|Mliu@@WL=Z+QDzgKRYE`VS{?2x8CsVR*-M*p ze9W$%#=_t9y^Wa~z7yQCH}1H#j^0Qslk=u0HaJxW#AQ7Q*jod|W#ZS1d+Twc^fmQxzVkoB$uxr|Cgp+YvTbPC)OY&K9V8_EXf-67a#5@Cz6Ta=)hSraF2F9(zFrj8PW zcZ)4j?{bI!R|*8mB6>rbY?oFjex~(^f2>aS&)s$*+v#KN_lC+1-n@7JuLP*W2*G>) zfu;`H?h{pPQ|7Yj&6LEIoDF54vVg||5I5c({9OB5`{DO-5D6HE(Y1)kyFhCYnuvyo zV@IYs6w&la{1(%!MN%bhRu}F~>7%dUzfK}+cmOx?z9c;UqT@sguWsBHm!wkiJ{MO5 zy314ESt4;cf-(JQmxBHKY-~$$m!ST=y98YNyuB@5<`7Wc{f578>7#PTE(%*z_(OD+Z|N4JD`d2pJ{(KS7Ya%iH!>BCDfgj+KYlwQ-=-Lu^mZldI3{r2p!_`IgIj< zW?r94UJLrt>IVcqfr(ojPyU3pSCSA6veT?(yTl3Hamj$3_N-D^#(5sz^_L;O%-es~ zlgz7#Ei!LzBp6{2y;^10Gpn$VvUM`ar~m+2$7p@*oTj!0V81Ktg5)}oR?|ty%pBOn zc#UJv$5&)JLEa@#tPoSBCZZ2yk0enzG0!D6PV=UEqk>k0BT~4NJRnqAjI*68ia?4E zddGi#pePVPOM$kyKG0UKT_SJab!JWB+0<{=sIq73XT6Q20Kj-V|LR%#f$O+a9tCJ` z->y2#=Y^BEKc>8MCvi>nbJb2b!z{;qC^I@F;dbeQtI>2aNYToJbErI}F099PM<0yN z@;Ed?BOqjWMHlh6^BT@aBY;c$ej!IY6s?pf#C2v@kOayrY3q{81j?H2Xz!=Me%L^v zeNKZHn1b}H`?b-v@88Dh#(yYZ9>^BWbO2_=vPVz^z1PU%xB>At(sJ0{t{??uqPvIC zFj~5q(-V^Q$rQKE!I6No3bXjeeRg0MRkc?~5N$qwpE44Y2fKhgM=G=m_5C z!B^4Uy_=SwVUn7(5SGTj``Y6vkyh4#z;aPt+J4TZjP5@Er z0UhC1yQ{*2-jl!Ce|aC)beavBTing{JzI+TeYr0WqA!b+Eo(cJJi!tapplss3XYlzd_srIW0C09<6K~2oktwyw-Rcwy9!k zi4tNhYffH)v&P)CJ3Y?~?^y8seevu3-CO%+Lbb=*n;&Qy*D#VBqY3m>?pPg1(z|7q z#uLzI6McV66_cIij$*^rL&Hk*u*35u*VG4&^SmtL*Rx?W;eNOI%$w}w*}2#(Q^M%-L&AWvKw<@(SsTh{zWOXzN!p?NZC2u6O?G4?tnb>$U(rN62*yDz? zEqizK)hyh{*1*5Jq!uR}KYA5U1<^f=_yO9>mnnNdb;_c#Ro9iHAa9z8XE3k-+OiJs z_j~ocDP*q1>EKJD-GJi|+u6SWhVJToUp@Iu1Uuw;hcDK)x6nh%=6kBb6#M|vs-J&( ztdj6N4NN&~GlzCL`90w#1rJFdjq;pkRl z9Ve7qp%WuMeR2k6lrfMr!Z8j?8`d^}~X&rBZS& z#Z1isdK`93UzhU%8(K^lSKTw7C8?)#bfJbYo_g{g?Sk5^TXEbqyw6#VEcE?J|5l)e zujcBr?^2ilJb6_6^S#14ymU*&_*JpXw-c0p)D7(9y>-)&<7%n#+47-!Upzrz6YqT= zSRJBhY1K~oG$!UEjTcu0{j2tSN^Ww*_Iy68@sw;`oHwT7>-Z^S$rR^}r8mwqFyRc3 z+7XyISxdS9&VIc$Iq%4Tc#KShO-yIRG_WmgU$8;(Fdu8uqDndB5~t=l)1BvKVG>}O z&n?e9PltF*1kJ5$X1LnD(i4%wrb0xbOjdnk+tBmx?>JO0z_ErYlorrSm7P_8Om$A@ z%SlQaf-q{GWG67|TB?Qgyu)u=aHC6BQ@oNuP3GgC2x6!D`@AkHNNkF;JHC+aKC~bh z!nbXk>X{t1%fqbobI_!rc;)1-meio0wl?yJ!zAX z?Cz58=L1`Y&XEFn`L0yXV+h3foxQH&l%rDBJt^?76WLS+HeOl78o&3 zjo73`iqaJy^#tj(b=^g7r!U$*E5x}iieqH4{1ngMQd@3ji&NUmd2AOL3Lgy5X2cYQ zxG!fH6P7iG*;IvoKECS+*YUKFrBp2ZHR(|v3!Zz69)0Rc1U*9aWTsKrc&T`A6>jPE z%a^MLqC0C%7fBwIVjsmed#{LB*V2TNhn{MfAsM40aOX3nsifcpLy7rKC-T@O4Aq4p;~d(*fV z*U^2y=Q6O1B~ej9n%#rgalxhx*LD?QrfjKrOaP2t&2j9tARCK>k4|V@X>kyevtz`1 zI-TEk;A0U#bKZlJZ^EXwBO!0z=(JFYL<#$!)Xmn@zkFHFyqWU(zcu7PwA;Dv0wQ5A zG-B#;z7m(B3aOQEt6MVh73dw@SK5E$diU^RukC$!=(oUapHtkS;lgFby`@)-AHO27 zzjqAB5N<#XNk2-*3g_?hK69K!yXuT_6u#A%6FFwN)~Q ztjru1tl2m&A0GWn%;hgeCyzbVTzJTU36=L~405n?XE$G|*D_S*n4@^w5X1TZe)fPq{*i9gglQl=#nJz~#AVz~9!lG2#8$NLc)39kFzX>k-AFM^%_B-SPLWh*hU#lOIocinO z5u%R6Q*L8(4bM^6?w8-EkqL)@W~W}_4{FBrAHTVQ70w%O3*0M;FWt*g#W^3`_o2ph zw0R%cQe~|CwI5zr_$jRwTrkD?%5yG4+N&AeopO9D-=GZr%P_Tp2+yA=AK1}|WQKK=9^2M>2n4>OXVAvFGSpI{W>?0Jxi{ZIM2Z{Ob z9~+XqN5eyWHkPeSeQTB6`mzYswtbG7aUN48wmr9}zxgLjiB@_pOgrf8xcwriIuuH>PD$jd2*wWlg-Urx2q*)*B5R{E zjb%%o)sk2j5&tDg>|llR55nXS0Ex+P;DhQXN}yZAZGcmi*rxT&+4scaEwj{Nz$U+-D{RKW5gJ$nmm$NZ-vGHeV#5!dU?g>g7{$*^^o41#jfWL1vY}BZA z`j4llQVQ{e+FQF5q_G1uktC->HI{B96#lyA0v=DI_+i2eD1GubB;dQLmG0Qfa&>aq z8WV6I?B=t@K9X+DFsjaEVF=Qbj$W6RUZM{>RkYgz&Co5el$mWcl^0nQ-J-jHqleJvfH@o)4?)p4FY1%Kz7r0 zdfvPhRCV^t=6mSMvDr!wToz^FO#dB?nf~lPJK@>iKVD(rRQLre>U#xcZ~#LSuHe1K zcWdtqyl;N}ncR6LH6B7%2Rd44>;^X+qaY5$22LdO%-acY?u(96XMY&JUb=*byMys_4KaKBP5_25= zMAOVF{p23MVM$S&<7jcCaADUv(Dh1r51=}u1_W*e1qF-Y-x!ixhml;lAQfbV=}83O zF%5TX=GV`9fbdVM!+zOQ!h5~guw>z3v~h)PsN?u=U~BX0~X zf9bye2qF(i|51AwlNiF`2WHgr$!JnNG+!ypdP-dmjMf4BWSpW&p0l=69e(`UDbHD9 zyr>sv*rFc0wD=x1xLEveE z78w~2=wF+zY`=;~qD6qVj%+ZaJll0s(&viNk<(0TFH(?5=-7S|plT4S1%=JCg6e65 z$X$Se|G^783<92#@||aDUx^u%?izt$ep#foE0nH{bEO#Bgji~P?|{7#k)f);cU&!! z&(uXXQ%@s2JUmO*9&kKQ*e&lvL#7q~@d_vfOdkG)=+WNrk^&k7xCt`%2gQ%rZ4lP;;k(y+)gz^XSj zvbzLgYSCeEzhXUFGf6(}rd=o3ZwwH*xbzp8voI=#=M142j7m;UR(_wKIw501u8~(c zEPSn=Tls-~+^VQyE|@ubOqttCe$sktRO2@enLMWamLe~6YHPojy7FEvA z23wo`bF7-!MkNN)Sp6(l-OaDU$D%SG@{}y~stvg@=#I;$G5#t|`Kc^_+;}|6ZL^T6 z9O~n9%?NJcvavj-@$u24S93~x&xQk2x@7%5mb;-tcskwwgrF#PCn2?Dp%HlI_8!Gh z6uYTQ#~#>G{4-cN`yUqOXTblKmKri1l!7aB(C?58jpQ&oazVgk9N~F$3rpDOA0*Jd z*OXP>x%`6*bI9w`02oy z=gDUKkx?^{r(IZYps^9CCbxC`S9D4~0Zs9pC}oIMXF?BP;`C-Ng!pV{`GNO5V$o)- z1FwVWoQ^Nm4&cn&0Cu(SqZ^HgLzjtx8-4gha`N{HwvMdeq*l8&MUGY-z33Gfm=ap6 zO1O0$jEyoO6Oa`rt{LV^ZeE&yW#uYG&^Cz@Ff~Yqzh6`QO755bFCR)izj6#UL@5L? zr=7kRFBEOK5|T=tbuyoLlQy%MgEs4EEXx$G;+=%wX+2A^(5-!ZnjE`TZqDZzR}sdy zz8I0kTy4hsd4C!2?emw4XkVtWzKqXSw;hT`v&71-4TlS}t^XQ!ek}PHIUsqtB^_d) znx?7JpvE#04BN^3*f60vF+iE6R7AfxRFw=Syr;;b&h&0D7DYt8xiMmW6a0Kt5b`qx zr9f%jDdjQP;@tOoaQ$A|Zu=vRsuMB5%)|7oAPDLR^?_=3}``c!35+}a4%b5>4U*Oi3kKoi;%WX^a!VY_szS-!k z!VTHXQ3il~|Gxrg^4)z}-{ntyG*n#9z0>Ih2zC!CQ@U!NR(GAF{%Qlwf58yhY^qOwR`)KdCSZQw10JjMXn>>+}GCbh#M7u0s};wWIu6vCYO# zPo9)Z*7oWKy^iiqX;hJBQ#=Vo^o!d8vMMSF<0oK)&Mo3?}1yhdq`YCNrJt2F~TRsolLBsQAX`4&2{g6+h>H`!K&abS*$8hrfO*9>~5x zo?#;H8cnqQ@@G_zj)Di1sGh2-Y9Ue8CDZv;_m1`MQBsZfs?YeY;AWM-Z2ILp!K7=< zz&F|e_`NjpCTvdiFs1y3B7bnc3s8lP`tW#_Xu%wp9)=l$+`;!OXhsB!K9RRi=cBgn zv17-)sUXu_ez{GBv;FgB-?pzYk5zj^ll^vL`I%wTHs=mq%P;Tj6>!7vOg8gVoxXVU zzQp>4-qaT8xx6LNo1?#frE&A|QfZRS+?S;mX!^p(3|H&kcqbzG@N$llt+Wuy(&tU`g7!?Y_mr)3t*pyJ6?t(byCe4I}&7?gt7g3_xX1%!*}<$KO8^9 zI#uLMci}=BL2mmyJ;D+ltMZQ`Z`TAO4@DsKisl3t!)PqKK-4PJq*I~DAC-4s0jz^i zxq0v{G~PZo6GXeR)Q|xb#CrAZqn#zBfyhCKAP>1g?U&k4`_VZvmI}h@6v~*htDtPX z9+~Aiyu~7=Q*6%YG2bgObOVB5EeS$KOL}zyT87M$bKMKms`Q=ExXcG>*@m@KBY`x? zg3;n%o2H*9`o6vkOmosHu=N1Az$p9x+`t^5Du$m{jXGV}z5yj#jjFE*anZjlKRU zlvSYSU~xh-8mCcu-8|R0O12MXQ1=L#t&?Me+3R5+(y5fcK*CY!+F8$L_eF;OS_Z%- z0{qS~$HF`PHl5%j9)It=Y3o?#Zx1nDC>AfG|dq_BsE^;DtLVnYLJ zVKp4-gBc?@cv}s>G3cz^;<>*16yyjpZ9%)M0iM`#B-lKg4(hHFC^prt)m(#tB$>{_ zPSK8+cP(p1N*xBauu{u)YL6d3UaqX0C{-ItSja#pDCyZ&_=r5%-+u-w`(>qYLZavL zV3x2|r_US!X}iHld{9KT=A$IAntCuw9s&D0)c)};P~)!i6e-o)`j2P>oHp$faQz41&v zWIkr%8z4<#Z!{rT^e5mH=W4dh^UrmPMsxBamkBh^CT-a%7Pi)FQ)#c}P+BB!BW+j0 zKwpWGlM%wzZKvHJlvzZwOIk5H+7~pL^mryYjCc;(T;SE!e|E(}8ea@Tak+0_N2A{T zEg%1vyC(nZ=M^H?Fu>v~qkKk5`#}CL%Oe=4IySjUGc=QH#M93do<8XDJ@@5pbP&66 zWq9m3Zu7QhG4lJT>-;8#I%B>Yg*_7o4--V~2QyDIiSdBWw%YMZ2P@JrWv5+ynSEc1 zf=X|>jhmFi-V3!Pt*9*=jV4%#tD9nffiyf?k!%Yfecg=PP?>#q*$aYD_K(aZ_klZ2 zdDrO|X3>%1jjvl@PG(F4R!~VTSr!^E=F|lsj@~Rcv8qu3APss8m{g{sF&$o|P_+u` z_uVlCINP3dwPCOntw^xIYxCI;Qi*f1S<{(cRw&5PY3xVm;wLRPjA~b^BEf5<%8dy8eHMB4RtCQ&pC~%V{s={f^ zp|w{ochN1Mh?disPZG#T^a7VOMwH#Mio>i7-=`F_?Jc8OL)+2AcWYRiX`LP`0n{ z``0%78Z%ITbtX}J=X^Wy-{#67472SmE}@Fb%p7u)8vwM+60hXCN_U(YwFGEXR?`#r zsqjUCHq(xi;B4zbs~iSQwk<=fdgQ11F@#Oy!^5! zIeh&^gn#R}Aam@LTFE7kzKurD#WiKNn%=akMUF^6S!S`@LPBywVwnB?Bc6s{R!P$) zLCDZ}QC}*E)p;rixxh9~Ila6Afc$1`QHH@<3P9(uT)L*dbm= z=UWc+d~l(z>GywYMVG-2K@1V)<>8tuDN&I*4|({FOY8B^zJ?mq5%<;GBnW1Q5#6uy zNAvFbuA)t*9JSz1+g*9P68o{+9dTuSP6tuO9frk4X>1VylygwZm{t*7`u>qUiP9L5 zkD_v&-Dhpf$6*I|q+ok5LOt)y9l0NI^87SJ+=sB|0y-cnF}H`*j7qaj8Z5U;e*B0UZaz$ajH9#RIjC@ zQ!JZACkGp^7CbL=Q403W-fE|z(}!1bqUmdB^BW!lzQf_Ok`z zr^o87Kd;v>!b@lL1;nt&6D86XNc7^@--KkqPQL!)X)yVznzVa%XS{_#+w`G%7b%XMu!etG~D*_8f+2^0dA-=m_e z_t_y~!vm_}8IL0LUYQ8b!PhI`Z(gnCnl*KfB@KW;O1ejG&%V~GlJp^O2*)a8TGW$0 z;JEC4g7AHky|Y-3Yq<06NA~mkd*x#qUgH34cIutAF6eOxS+^+!G7{Q8+|{e{ls5F%ohX^()OJLlOVa{KXmlJ%iQ#up=|$<>@Q_ zpC{wa{lYPVBnsf#-RFo3gp-6b1ChUSo<9?<`;*PS-mi~eO^Y!V=alkX?qxzwbbr2m zhwj|DceXZ-1OC#@^KuxkD@6w;)Kpa3pgI5M3UaLz7j^zsJ|H3eGBul#1)B(`E8Puf zX=w=_U7z`zH7nOy@JA+Q;uS(bJpdA%45|SOcve*`YwI#cqHn?7t-Vr*EuZ)KnhF4+ z2jz=jA7e9MAh=Q=qYQYy@t^_IhqQ7^l<^hNs+fUS)y`}MAr(~KbE_cyNnOF?B%>kAy$YDl zG1D~mvik*Q81=mg_Mu|yuEz=!<+o{hf=Cu??>Ky}SnZpGB=te#E*+AmCp||G>%$BE zSjg(nS;h+I&3GNnlX0nJDcdeg-+iXJ>y(G%nD0YOr*WstUNZlWSXds>o+WJYq^VtR zHj=l`Hrq1M$A=ff+7TxM0VUwrb-0p=cyJa|A$FRG?T}5qiI=Gkh9w?=L;OZxa-^V#RllOOPp8;j;(~xdf5thBwjA5d zIL~n~$HQVc`MfD0iT_rvARhO|(;JCq677wnvNXxHVjU7faAB8t@L-|YH!vZ!r#j6G zR%X;M%&VdL2DEfLp#ylDz`z+FgiOhvGEOKB6XU;7FqUUGb%W+h69eKsCm(YKDmFZH zcYobUX5C!i*wN#|c-s|Wd|yJ_EMIL2lER`TNd?*TX7iet7%pfdAvzTw2de=C-S!w& zT2(=Dul(by2+omdf9&RwO3;N+nA}k1A~%sox>f=MaLupv7T_Nvbcn^wyRqx!Gj^R7 zO!Z{hFG@phJP58TsU|t-=tQy%H75i!3S03=N=o*OBkQ(y)WDj|>L5RaGpI1C0b|D-g?z9BnK^S&9z8 z^#0CVu4{?Asz55J>9wl165aRxR#Aqx#9cRH0KY38RcrDi))Syt9q|W2CQRb@9ELr< zfB15zWNIX;Y>WnP`2FLZosA;&=C^nxs9P|2VEc{P6>C2xR(WzU(awP$D#J?miZS@e z7RzR|zm---r?Cip3aptBf=++a=xx=_xe$AOD!irAV@YHykpOsau0Yc`E03=XynTBX zwndb#A$$=B&h_e$NTmzAVG*Xi57N1#Hq4`p@D|V0*%5eZ~iIX9m zKg}pQx`N9=)#TgsSS{YKuUS<=g;)9AYo^)R=t1}SfBp;Llm-;XPxo^r!!6kYP0Qc` zt}}e^G6B`lj|UqdwbzEK(`(EYNtivbMf-Q>PJGgeNLX6iBaAE)EaN962llaE`B zKKpk4vD2(KdRa-lm|Euy-wEj4C`=YdtrjP(JAg6`t7{5l?z)EYA$XT04}171IZ2M} z`g{?Sg*gLgWCWs zWg_{0o{;0^rRNNvRPfB^rJE<2fQ{)9%BePb)HLFw}`%eC%VGT4pT=37V(lgT?mK`F)N z>c5&$6!uNy6DG&4jFo(~Qt(sxUI%dhn$^cDQpr5Z6`0{fL9l5CtWAACVanDIng|>o zWqR%%@04YL_9u8_Ys-)ujw1;JeBfsCnoSj%EA6GYsu*k{T!^K^+3rEEcjK z7QDM8>X(XOl}Ue}H=J!xRQ6Qg@Ax74m)j2)>%@`#~Vx8B4%UIH{Of%HrMdL<~N`H@lijqA3#8 zT_BGxxU6`&`N!TE?aP^zMfc%DYyiCGd4l?iggYj-qP(c4xR;eitDjeDZ%oSj~(b}Pg9?Nxg2F3Sru+@)v9|?eC-!ED0A6TjJ^^DiG{WljNB9!+$y{E$I zOJdM3DI7X^--<5XG>@s>U9|JH2el+9zGb?6kv~K=HYxwC06KEuo@t(}IBhXPpNYzx3^b;EqVPL=U>EwY4R&Ql6 zVqH68j(>Uo^raiUh(KWwk0g1-S^z$QeVao1)phx{{?f1Rs}JAy=NqB+DJU0A!>piv zi)i3?O1yOFT`&^>e=kDU=`O5a(9O12kSYx4nT~lJM)X=(kFqyN*TYtB4XURjoo5`$ zjUb&R`5!Xy^xr$;Tl^8A^3~&r)GmqRH7$6}e{pkptm%Ct$PhnjIaOz!v)~)V>-Kw( zy}3DCLsR96EhXty!S6_2f_CF4JIe+t!rJ{j5{aI;@rWsJw+xn3RE6x8u@+k^W6_Vp zZNPS3+wR;7;QWJ)z1D8b>SUNu_$KEW)_kj|b=ydBJFOPmHcH?(DpL{d4mWT@2G(mu zugEwWiD#P>mD2NX8H*2>NbUPYy20PK>1C&UX@FR_6k32xFP5L_0*~%D`VHmx{HVui zZFXWHI{ZcUb69Oi(-XhJtCuHGH?iCC{vnv(<~g7Ou?%E~dUAtJ!Ti0aSR`6mkyWSh z9wY?I4hd=CK%5yPj1vs54Aq)tuw-#a?}E*~?-+S9tg~FV#6cTmN=!`hxx9uq#XXq5 zf#oG=Mu+CzyXGm=?gJKcUepj{6LnZ1a*p3H%~1RX2eptP%;VT zf{_5N{vcmAhhQpt>a|eguw`y_7SqMYVT}VPAZtjzsKT_5EK*(^RrKkr5oOLoGCS@9 zo4$E_cxpX*vQ^hvywkO3|8hzTV%_;AdDYq0{jRnea35u948Tt z%5Wd>ofVbFlg)2(EvTH$zSS;`YS6w~VTTrjbu%wn9Oko_y^8?~e0Qmkoms^D3;j&x zR{7ouNMB;IF^(27>fOPux_L>oH0?{A$Kqj?yJ@D=?Z=PXBJhtvZnY}e?!?xA2$Lee z)rU_U45f6Sbj5Fs2qA}6V^#~$-sE7#wq$WIrS<8vfNtZLFJHgz2DKlo%htjO8XD<> z{t{bcB@=`M?t}ZrjYT>V*LED=|Kh1y&LF_zPCu|bLLK#v@2n=OQM zy(BE%*}U@=&tqeNy*DNw5q9lkH%Km6GAeRR#FR8Q-wS1u>sm>!&<7Rqgj{it^%))+ zBh<6RrZ=BvVScYvwAVE&FNmyfX+pg@;L_&my9A4se1pkvnDKbZ3 zK1$&vZPPMy4l?=3Z?E-XoeqY&XkNA)YE%cxn&2jMTk%&(+B*$nH5td;bF z84S5$^dl?6I%#emtzA;b_zCeG^-BF3JHk1KOMB;2*H(nbut% z%9I78l8R7e_G5YgXW$vvSP^m>{G>%A8-53bf&&GC$yKR%Uj{_GPsfgao*}lP$*+TC zcyNSQvpAJ1iaXJ|)>KDac@P8oRpz5iV9TIlcuLh1<%0`meDvxc+zl{PH!P6S$3ik| zc$kiv-n$8)R-mmGp^*sh`BmDTh5&;N*Rqw;ml54A@e6a$9H93AZ#0inEh*asW>V6a zSg)XxX`cop3xmAiW|oP_c*?038~3=P;mz{K2xVcWVcxLeTy9eWC3N%^0SawT{S3&LlqBxfpL=&;XSigkQ%0zm*_hJ&dbbn|4qID{Eincae zI%Vc^0u}!uNV5@OjxE&6|1+qs*38Oilqf=d!iJyuG#l{+{NVGnPlXW;=~gLfx)!iY z=;8R#OJWZB6LOIX3d--bf_VFSKWLd5t+tq!$0$dkasdqr4d_E*TpCJimsn+dR5hd{ zj4b%{v^=d$>L&J~aQ`Qo@B`9kG#sN!8ewmZs{1rHPRWMAzo9CzC%Q`U+jUlV4%FSJ zkq#eIXCie;$~4F;qz77HnRPFU9KDXFL@EZjqGkcV`I5AGOya~F z)>%Y*WNCfZq7F5r2-6ANcGU8w(7=u(Jwo6?`9o9RjfDqsQUrid|2EaB#Mf)ZQbmCd zhua?8G?Aez)y{O&JVz=d<`TjnuJoE96kIN-9>#f)!8;;}l6$V{B6gx~g4GJ7xMNZE zS{TPV_@p)HkScwn4ag-`wfQnW*(wD6jY1mZ#66q2c44P|h9tLAc!OdTkG>XG%Dp>; z-^jpWxJ18#DvF#SD<~>(EaTvywDn~Lx3kIQXM*1VaI}r(&8*?gWH$tR*cb#FsLa2> z!#pdfr5I+>mL~?On;mLOW*e6?BNTYL(P#7B|R4( zLp+HZbTli9%0mj+9G%VavOmU=ZxoNrv5v2t_Jq|*cH8b{g{nGJ{`)^-S87bbYgJ6JNx13=Ru5`x$VTXw zEXrg6#JOxNqnI9%Z(8R__z2>aohRKh=0V7|PV}ORu}LYuJx->{O_oa|(J}*2V~ku( z21L`6!#tJ-r0@qyb>3HV&agoX(ZyqYcZ5G-TtvWRCS1(-$ zU-%IXSe4!1%O>)Ro4$X9T7f4J*CKx9S6?106P%?*bozuk-*k9!;t!ydNK@wJw63QpNp&lDGdYLakGWwRFqerr|f)dHCEL z&_v-t@EV1jC-?F+)_1*2Ym4N7)Pk^hr4Ed9&G2NATnNu_%l4aH^gNnfd^HbL?Kn#MXl<2)R95MF0h@&Vv3-s9cxdTugkKEz&P2WT-1kTaqRN%EN42{iKDyuYGBn zOZBBcztXGbF$}0e4iRm8L7@0sZz?eqsVUB@ol-36v6AhURM?ew(7ZNk187zNu(@ju zfJdy*8!&IfoPDf(sMNt|_nx|}qPQd*bMAgX@Zi7%9Aj*BLfNp=8QnChAzOlv;x;v4 z5D2__iKn@v+oBEoQVVdJ2_c|Pb)ZVS2m-F&LR}^_d<;t(ezS`QR&0106eV8(K;NHr zh^OTSHN8M@M1>#d0z##K>Mu3p0zDJaym~qLpusxbJNX@e$!EWXGWozh-?Bg_nu4se ze>{-%t;S@W^k}O{z2^AVxTM?Mo4x%$IBLnfzLvz>U*0xR?hS=zsFk$=7VE$`7tdZh z5fo$Up$TDKDWmT+LJdnm%6Z%Oz~`Vk*4uQUuStSNp)b>=hYlC2n={aiEAuaFOaWT! zrsgZvL?i?!7n7zz#YmQht}Ufnv?NDUEAJn zKqA97OIk|JF>}u5IO}|u@QTXT8MuuB165^hx-17)WOfK<8cVON)`y?p?|%NKO7M|> zzV$=-aQpRM8_%zo@i??>aga}!qfMidB(y|G;1`#N+*@?8HexP6!cRlDmh3wT=e4gQlaXM*+r=x^s-%#lm^OLFLpSLZ$u7LDiI!RaMy#N&8h{nM@i80TEL4-PW z{a7g2V3_46&v+Fyv;uZs-&duGjo{!UO_VIG4o9dfK)ubuRFt}%=F)xL6N;7(D3 zrl1j%<3i% z+9YWFyXJ+=mQN_oNe2R4-5X!)yj2#nPIt|=O?#Z3xnhKxI7#W@xrJtqW&~0F8PI6z zNap^4K~Xt?F*As=>;Gm{VH+h=r$h~DRF2XOe`hfT$-6{C@Xy3d-wa9?taHuB;MjED zFIU(XG$jyv;b+L)9pf+_Wsk*C1saQvtW74-bi-V?FwAHLKB>ODJnY+uRzSD5I61AB zp|oTRiO%`}HtY>rbO^!Z#a=GK2!?{meMRc-glJM@ypN)ctDc^P|H?730UfbOxk z<8h2dGL;~Ak*iKl&Ky@%cp-;OW!$p*JVxVnkpG|}CA!kRr>9~a^uARrZ_4<~4vr7Z z#dRj(U+r%-_jIo7>eRaGMsXW;K~q2zz^}34X4nJqb(4L^$ms6)w6^zcXdDgVK;`CR z%_&-Ac(2_a< zJ;*6_+!N>u*7Lgh@px$b0-8qVQ8aO)sYt@~U}d`qOlw-D@QZ77%=@q({}eg(a$a%oO^P*Cf3fwin>S#(||8-`niUO~TId znGEA`6Z=apH3$2YTl=fBVhfD3&5>Ti#Y-Q6vyfQYg{GJZ`w(EdfJK!%*C}#+B3T@a zNZ7f-DC!rOxF6P(^nL=jA8YSz>dDDi(?o|*9m;%Ma_Q1cLIX$=t{Tm?kS(|elf!7S zM5F5)-(4^_!|OdU6k{0sPW|4!d)b7ls@e~Cg8Oi!;isjYZ`F{hv(*JX^g{F;mItB}HWO4YwqW$?{Xj=zPp(0cgG7^k zj$3rJh~L+HYVoo?h!f^FVBb!aI859&Ds@y7eb))lL(lzj*v=wId%OMi0S(Uh&WqCn zIeRA1_)=_NiiKTGcTO3<7@2QOW-D&{RFbR}Y02d5%_nKmlJ0&=;L2+s9HGEwN`h?! zo#@XNFo6Gx*z7!K;cxoNuLF~HN^%pZ942;b-ZbqHS+)57?tM2`qaYCGxu6 zHsi-aK>BUOEa9^pV^t|tM{rGFeMH4!g-%i_9ydkiPS$kBcgi@=uRdatzG$w(Q}<1D z>hZMu#=H<;iEsOWH<81I8UkQem78F&OONCaW>m!i42v5R7)&EPsRn^xrXrty-c~MV zXY<6dqeq8HJ-)HU_DyuXe^StJgZcu>V!FE+J&tMKgYAYzm@cD;N z`IM^HsCxJVjKi?R#)=w}10d%AqL&?t1~6RyB_jEizZzbI^#V(u1wiE3Lg!*HoGiI4 zVI;n@H4+7ox&M3za!c0FewRbq=nPoQE``h1TRjxJl_aq# z{ujsj@QJ~<|NF^DAE`dF0@QB9Do0rQE{n?{ZMIN1B>Gw(7#L0@Uj5I!_uo=DWHJAO zSyrA*4Ov4U{mS8^2)<<0_nqwdW|G{$e)_E`&m05TVzXBF-?bc z1*RwdT9C1E$m)-Fw-i`k4^XRvF63F=Pl7}7C70+c-cspm>!%_v+$XBKydI6kks?-lnFXMXqS0GaY&hz zSeem8;y9aE;r_<%ZeTi|Rb#2-luDBbjS9wt;GvO@@sbB0Cvr2&jV`X!%`bysInR>R zFtV4l%~2=V?RG}KXWMsA-Ji8*y-MchX*t1o|2J3szu>`U;nrso{i~~Xl3ucR!D_Yd@B@nBGPLQlBhRQ# zNzb476SUI?22h>2?vJ6E!R^+F`4iBT!Q8vUA|&`{n3FRqZ1Dt8xL=ZMxP#1g7cz`W zt=*?)7_lS%w~sv3T*tegp`ia0x!bZc}h=MU7{I$p_Drt&9vRs(qgc@HK6R=e465&dy_qZH*zv@N6c$MYRoPf$~9v^+VgOnz$e zT{?9<;LF47WXlLN7jSO=_DAwN=c~gE3=EWlPNw$_MjARhX-!OIS5KY@Fm2|05_ClH z`4K+3&!bGAR=oqfFJ1HNMxhslavX$51dw~Kg~p1r7Mo?>tWMDpb;5jT=?IKd&)jG! zPlj1syqT50ApSuyj~FaGVmhu)A--c`GNFBCkRIopHfcPAyYkA8*78em9Z98>2q+>NJ&(E zZf!jyD_eUa#D`y)ImVl#p=A-NWA7AofFTx=6O7 z|D{a8uI+$H*x0L+cejbhKG%2CkDJt<2vdJvBWJEhVP~pd$_f64tVf{VwH4~+$dRO% zgI8zaUyLf8J_RTHf2dA=_@2yzA}j4ben)d45+Ee^N5nrjHlBW@h5B(ZGg0RnCxemm zI}!3TX{WLa&+h+FxPF%Q0wouQ;Zu>9qPe-wvswlqC-C58Mfma%R8{}``N_>GX*MC- zjuUNd45Ff2t28HGXdGQ_mpLhV*r#*$9euKSjQ5eGkGJmKyV%aZ`~!>47}dXumks}y zJqGs-d^*T3|3^4<2-)M4_wK5yUdzqJPA<+p`_X(i#`o2!<1GoAYGmmHs=o{B;S$Yu zQ0?>h96n}d(^mUe$2Q@hoA5FC{rjT@jb1N4{5X5xV5q*0>krmPPqVBF6I9!k`pB=> zqAGb~^XP*+fgxxA6n3*adh8gKTumPL_%naU5?qIWiTanu0g(OLJ+EH9`gv@OTD?>7 z$9cNbDs1X0>Q7$ryiQw(O{xYoT-nrSO0v$t5z>bFQWIAoC&Kh zB8sRVn@!Kav>#u=F`NuSK2WO^5ia{&eJrFaPan+8agj8=C-RTVBKv5PTl9>cUa9kw zh0_^-eI5-dBriFZcYRU8@PjVq6hqh_;*yf=l9K7S#?OP!O@HM~u^F^l$M;kVO6!}K z{MKM6pFHu={{O7guha%$LCM#c8|#OM={*IjA5GKTf8w%!^YB{)m?e12ZVQf5q>&pA z)ehTEqwF_EPjwi2+#OODNpp5z?~T=2B6o=YR3BjdVBMk*{>=p-%lki1XL24*hpvs?Pd{dC$lq8E#(gOqK5m75Z{_(`-Rv|jrbbNL$uIvX?!(}W{7^tnliUF<1Ke)Eb$Mo-m_%9RP9Ly=CMWXmXemUayVJt)buim@tDr}#5lku{%K)y4` z3y9wM+2%~{(F$K0w>g2&Gdw*)Hm#=s?};`V^`K8Jrb8Lt)uM<_LVCw{9#j0?)4wGM zUd)PkJE7Wntax5-fD5}`Oqwx=d6LXMA9z%?Z=vT0Z3Zk#9rPSJJHYZEtNvsO2~WMf zcSHLIn;z;*?+o_bZtM?5MGQSj<)|v1?JJ*6M^s42xs9LWmL(NO($%6HMN&97 zyib&Fg#^EF#zC9wPG$6nn+S4UTV{Y{bZ{Vdm>HavZ*JnfWI(dDU_DPTm-bZphKwej zO*8rC*iyMuOS1qR2$4GoDJItHGFXNC<7V6e7sq8-?RFNkdV*wx?RxI5gO0Z`xA^L} zV^10uF92|EDYaQKW4*>6#E`$#8XWYd5BtRo9GNv!6dsf=*X_6L8J*IBbF`JHoljX zVVWVJQ6+#C-DzX*W|6AeGWWZNb=$KDw`o-#^IF^rXFMDJWvyo=zVdVyL8VGx^~N<{|{YX!4+4xEgL7ey9IZ*#)CF4jazVnyK4w; zO>o!7-CYt~gFA!(!7aGM+kEGqcgH(tJpKW@*P5$l&8nJ~%afU2#dDp!X19(UC5nq@ zphBNywN(j|dApR1*Oq!P_9oYw7ASdbTOYZT`(u9lEQOipRzo)-wO}G&h{46tY<+VQ8L@ zRSI4BXX5FcYVH+R$@)rrPK%7_xQv(%{cO@hB9?u?Q3td#P*VzInfUVHa4w{3t1+Q6 zy+&o`y9OnT_)2Zue#Ok6HElro3HwU+|GiRou~E~5R7Myzm6gJ}D+0gIyRcE9ECzPR zj1C{{!X`sdegd~mszdZX+RfE9D54Y&xU}dP7{P&ofzj#${|ce=_N<9XP^`hSwq2yz zNR}o?A?WNQgNYNVlbkW+LgC_{19L;qaKcy}6@RMybMA@)DK5n1eHMxo0?djml}IW2 zJTTdL|5A@g&=)c!na@BlPy@yN zumGIlfSAq-F#xFi#~b>`c+m0%E8C+?$5%A~)U_@n{(ad+O!7&{{_LVx){hi) z;}zHVo$Nc}r%xTv|D{kD90|!M_nS#%?go^hCtt$(1>UuwQ@4*F4e0Sxo?uwEpoK{Z(JreST1egTLMti3=r*mMNT}$ONNhG)nI0E_lf!IvD0hdpM9EDF9fKsh%=Ux*?B(F{h4fHN&jVHB}* zbkPpOBE{WPh5Eb`brZUbNG?I+M=C)?%uY1dZ?S>@jB zmyaj%4$k{CsGrC5z5`Yl)8O1J8Pz^$hSwj$mmU_6sHS2JPZ6hV*sN&~b4?)?hTb=EWsN%C#?XnDzVtpa-nGLapE9 zQXaZ}shCNCy#ss`n0;QS@LL2)-w41)buizI2#6TIAMl>B+C7RM2-~AMoRA(7sOpnu zWat~LKEBqhbEj)-Z}%7l;PXGU(fA}JG@bE$985GvmL}*x8xfx$l<$!iY`4T^>$B7Jo z#X+iP9n&z`iT2Hso%jY|cb&hrYRqwe{&+r5iWs@2XwH|y^6@o4W4I`Q;z=p*oJw+096u%M^W_LSs6e1QiCR-xC#ttyRA zDa(B1J^H$f>1n(Lt!i#Mvr}re*Rj4zyKDQgQf1g3DoL(o`MVkv;E%Kp<-c0xZ_yib z551`S&8=QPmM3w3`2Tn>8|rV&+2l$!E7 zXbJGW>p+k4WIPjzVLQ4ui|4mgyb$ky`)l5jOD5@L-~bahIxl-OA^mW^cxlr_2dG7M zRj1N@f`ya^ceH^(z2)|ea@q=mQ&x64*pS!64z;sd-97G;>*FUD05nlTGL%3_guVQM@ISBp-0*VQX zigg=&KBnx2&4W(FCI_E8`21KyIl)-dt=v|q_C(PV25CIr_DlK7q#}V8L@qNg*d&sj z`Mc)aBWFb6Z`yaPoDS^I(bP z{A-=UiIXy5{#7aPI&#ba|FQ@bwHCttmIyAeiI=O`4EL14SSyqYr2`=wq(znuvj-?`G+V}p3L*2E9z4l zEj}uldo~SRhzbxi4dE_TSoKnnO@|JS;gA*`<^FQ@*IomM3LkIa;b_%^dAEF8&8_*5 z-@}sN@sBGFDyJs~Hn|j*&$|4rn44^f$?_>8yuW1(b|H>5iN7D5>xyx2+mL$ZGM=HR zPUF!)*g1CT?2lMe22)FQE+V1~R>yPq^SuW`Adi!|7O!|xIYOQIWK+>psiy5^!|`%C z!|_b%8zchmcYaR|lt9L5Rj+J88Or z%I|t4s_M)c1QX`%X`Upi1L>Iw{C^p+Z*CbTgt(-rxEgMNKi~r= zJ;G0T5qG(N@W=Bzf$y7C=Hvaqht8t6Y-ng00k{?sz*$Z9l;W6d$_u=j3`6?Mq{t6BM8g1jYrNSNbzpl>di3tw^3W|SF0@Ev>*K6pcuQu@9@W3kX|EHmqoL!}X zgA3{Eo6Y(7nHPBVdAgtj2+m_{=n;y6Ictn8=D|%OQ-@T(dbpupmzDKhk)+NT(LtHw z6g?D}WF)+ueR+apRufH@3EyZii#bNE|CEm6arzDGnfoTcldq=& z`7XVMl6g>^6dkr%5`3LP^~swzS<4WLn4H2r8IR9EmJdp~ZdfQ$?TT}^-HQptCp;EW zF=gfqO7Kif4N=&D8m>4n(8^^B`MXdOY}}$#V2H?UX;|Pl&;F8IoJ}BsElO z?7)Ho&VvLSy;{r&N1e|PPEZ1n3HZ2GB-q;c#jJ0?VF0pn#a3E@7KIt7ZUe4J){@P*n z+M%EO<(Ob>J|m2UAV0yuc|8X5`MGg+Q0Tc{`El*qL^u?A_VMoRF9=`F`QRNlymY_N zd(hngTt2_8cyS+M_C#cVEU}v*{az+qKh|XIz(-E&pLAGu4Op0donhNYVKs`Am%Xzb zvpYgJlqjqA(o_ebj~-j8U!J)1Bc5ipK5%?UX`=eRW7@{JKNjybA`q9CUT<*5WRA3D zymKGnet^Xw!H$T468J3|z|cbCb$W3>- zIjrGgb8>dp-z&^)xOrU4=5mX4-{FaT^s6-s=Go@#@S_;s zfvHXTlD;6Id3X=(RE`h}i8i<%P#*Wb#O{ks`{DzKvVI+=0ugF?`b`ET-8gRwd*0(M zxgB(sFS$7nU)0`&zE`?ESiV67#G5Ri+T2rYBLjas@15gobKqWHQRS5uNN?0lZPKZ+;eFCfj7^adWd?Xds?48|mpOp8CDFMV%D!$Zx+> z)wVU|c)i2Md-)ZJOWpJJ-wZyz|5H45s#~Vj`0^H+fPAU>={uQj(=|BQhq)pGafr9p zbl-I(g`f0ni>0UGkl)bKGP9!m*5+6HXHpy%E&8P?>j<}EDaL6-nZwGTf}4> z#*E8313b=3d!5@~>#?GS>NQ`<2jQE~`{6PZ+{h}^){cna&@K4%ct*%2!jA~V8WhG) zIUV?#KXStdZ0QAtEm@U+%wHuTwZWk#5(`Q?^732g6-Xu*1|8ZkZchD1e@qiU_|mQ7giA zIZEPAnkT!s9W!W&G#1OoPZPY$49>Brl9iDb#o)69=*w?cGIvy6~Tv((8d1Pdk;bgp4lh?6-%|X4cnT2UU-{{iXU2chU9YcciwB5MW{dUMC zt#ZY?vUtMdudyo|D(FxEFRP+W#*poCS?5NoP(>25`%f%ERHn1>VO{ik2Q{{H>@9kARecgP@{3*8YU zl+Jx^4-hNi)?S*=1+J<2VhvXZovvu&!var`a=Dq_`QDU9Gs~SvVkEwQ0o?U>bAtmUhn)w4_>U3T4MWylO-gTg$SDo@gNb+I$c+J=0@z%~u zI|+1`*yJ&^8p5qH1LQ8l?5f3c;&_+3ArXE4Ee{++@syk8@hxFKy)f;9AdSoQqzb@3 zXBK-X&FF=n6FJjjCG$X?eKoD(eoIb*BI4=u&CI(T!meGs8L_PnT-PWx(NmZ`Wnk9E zSX|YK=bO~1wl`5fxV=}6SU6o=KdLzKbJo}0_uAufu9o$ludo2)gVPQUqGCJMy2%Og z&3ButMQWAsIE=dC-@YkjSWj1|pqj4WCbLaKfA?~K)4J{~z`{HRVtjxUTF z(vr;U#SA4D_NN+7glRU*&?+~O1E!?xI5u%X=ag_!v58tYeL5Q~I<=P@)N))f7RuB7>RdBNF8(waU10gWa&lqb69uz}@T z%L{_QU<=_U+y@?yJtW{QR}8_^ndl^Lz5E!(ccGMHiG1&{TXS@ev;Vf_&1|)m+ek{| z3m|ZWaey2+!X1$KD793`Sp2);PqC zBv?@A#e|f`8Rvpy3gYK4^g@PLx!r8^ef{SaEVCMw1=12C6?7#6T1Ki`rZ6D@8~(>+ z-dmtcS;_uABa`(S&L6majShE$LrMJq3nU*Ev`n~b-0XuJ8!PWWHD{Xw?ZA0O7y56k zHzR^ruSV_N@Dv>S%K~!Tk8-d-JEp1ilS`=c8Dm7F>%3C-s(}lSa_C==TZ709%92S) z36t|r6?>74tLjDOGHGnwPrW#}6EoqrjK;UB+=x+Wi??B%G$F$18lfoEv=%ZAI4QN= zsE`lTbzQH(-Lm!xJc@&E{6)s4`&A}r&)gD{*TCg7(`%9^ph26r=ys(2AvBDaO!fCZ~B z9p9_#z3(~34WMKggEQD}*KA&$_ze7x9(KSYmR|aNn0CfMhsO<`A~U8lLwO=U0m-s!p?WDQ&Z$yti5_cuUa1A=c|ukwv`A#W*`=X zoR}_J)Oo`EF&xmx)3I19doUA+A$=5OELJY;aI8@M0U$p#Vqzu^i!dk2x~m`EA=f2J z4k}$1Sj{84{+?=F#~**zcdl2xfd4}=(cs47Jz_q*Zss&@-Vb$z~o0i7vUPt z$hlYD+fdU^XE!~A1g7YVl%+8&cEL!on??fTBZA3xrke?e)GC4%Ry=9vATG(NDZ|{Q zE*fGc_ZABkuYor`6i_@5mhram(E(`hsL;H?Eq&~(?%s0pF`TMMxKR9-{4Yx;80 za}tX{o<~mGLCZ{y&!HbVHC6UR?Q8~gH?Cv07X$g0SwvZLdVj)*w$e|mba@C(+Y-@i zK8EM4$H`Hx1;&J9sbbVkMg-xm0S%%m-qgO_x}l_mj6GM#IKPKI5l>v&Z_K&xUjQ^` zo-U<$rYi;NZ~Y4#!{u;7?VElMEe~JJthw~)V*>N5C*_4FE7PaW>Ew_nPR(#^xJ;|f ziug#C)+hUq^*{HOa%R}Ql|PelCT#s)b-~RQkCQ0+SkLx_lA7|F4eJilMR9AQ)J1ZxCw+xVqdzk`b?p&hKR3OD(fIx`X62bb#mejugGe)f-x40 z&g&VgFeH+vFG)#m1+T)tK?%5RU~$j}>lLyjQeICY?2_4-;Y77W*o){wwkR_q>3*h`&Gnc^9!MONWBM09GQd%TV}8?Y6PO? z2TD;c@|rF>H??}@?Y&lHVQJPaTWxnE1zJ!J0k}N-IxPmGD53}3u9yLPO7_k83eC4X zb}K)t6eS*jyn|@J2PG_E)$N?=vNWNu@8hW^5ZD%vAHu$umHr-lLC6D+Gh;H*DhZ97 zXn=qS*pav_=i-*aBR@As@t$qLEJyhuoQOyWLsgym_Ic^JKq#A3qmYmg36L||+ZZG( zCB3&!npD&7yV;CJ9Myhz2^^Fdx(^@jM)YpyWZET1vv2OLtW|#ojE|2aMb{gwk>}t0 ztu^`w3=umEwQ+ ziZY|Gj}1AYn+vR9V)wOfuyikvTLh6pg~Jjsv(L0YqC+CSIRkG-&Rmc2xOuQ~zm{xU z60nzaiWN~8XEmt)dV64U#am`^zq268;++V z;x+sYAAy|D(D1=cat=7kI*8_ff@#d#FL1c`kxc^sGVn#{yWW#;*6V@R-3z-<;3avgy5A zi?z~abJ!|sn(^YTSJ8|UBByCVj#4Styw{*I1B9+W@lbiueU2I-U^_i%Yn$#`vAx%v z*O_$d*!^o*=W6U7z~hLh7~&L^w=;3j!iPx>fcF(O2QKRg8wgW}YLzO-)HmJTOjp#2 z*#oL&zm}zaE4ft^M37%tfQ(?cEc$H*Ef@61$H}yAucf_1WCOfPGQf5ENN*g$R6st4mQVpQt0aQ0?d<0yQ48&a4%dh-cAlScz zgm%T>;A-$#uQr1f_L3%J623vT*9zAY{`yQH%-0^j09HFjJ0@7=NDU6PplSjRUhrDc z%RL^;uRw%XuGg3lAP*aGT9Cqq2F_DLEo^OL*0@EP@daH49?!&-Uy|P%I`PR5Z>R00~HUZAR~I1^)o1#^S}O*ZBtmm zhtFWsx@y83-V>^9tkmGhmCozMJ#)P!39_6mDK9$%5{)_HhtGpxfRkl?brri0*uR2K zgb@pK5?gtiKZ$c3iAZ3M=65?X^Ypb0f_ZAzSbKOn(9$CSW_OXb*iAOCpSmk}n}6=H z7OKSg_jxRX;f6eNg!>ls7)Es(xCpuEz!mW%<=MG}jjc=uMhDNaFMm-4Fb!=@&@ zh+O*8KU0GLSB!Q8By?MG@N2$lS=UO3llxz>DIgwz=^}vD=_bFW6uM_d%4_5*Ij6%3 zk!}@?LY&EoA`~R0P6lM`Aa%(fd@el%-q|62)H1#ELw4(Kq*!q{kS=4C9Thc60464% z*@?u&xhZhXIpfB&kGa!5b||PQTR$2^jCgXs2r$d~VY5GgwM%Qfwo|+3lugHrDC;WN zH-%dXVMnIgZNSBk?0M`|5P0+42g2%a^R~AteGB0_)CO>R{p6o5ECH!XC zbtUeQ=RzuepM<~P__if&+?xkK^_Ic@Z))uUqOzNc+aY(U_`AZ~+Uy%w?w`V!v0Aa> z`k@!=H<4T#CT4}1af<8yK6OMo)t^3G3cL1Wa89LN8X#XGeP=91+uPN9q)&n z{+f21$~;lB4x^LheSZ%w9Pxh}BB)foZDCGt+x1PYVE*C*iH@Cw;oJjg6C@2Xi%s?i zJ#UR04RP;2mE=3^E1K9yNQ)s=YJf0}WOTkqkBRYIACL!ne=&tV4b!)D#9B+=tE3sc zRhrUAtW_Hvnc|BzF#%$a&K~otFx8sWfaHwjj7apS&aTnXhxP$UK=?9B4nCPbi~S z6bW#fCW|;^u=~yZtcf-kodv~`K*~Oa>{)%239af=H;*=eH{a>g7`@H2t?bOw*6>*? z0wQ#4kS6T#rS)zBOYDU4p9a8`69>#V^)Khf0TWn4gPX=Du@a-p<(m|H(BcJh!d!tFpam#Cjb5Sy&xO+y@$&NK?^C$6+AKA0 zBu#FfIHkz6FiPD5W22Fdr~LR+f~Ml{Qc3?cF9Wi<{#Sdtw$qIW7DaGGdY4Fo+B;q@ z!s;j;%MMi+c*U(7E_!ovM9bYr+zYm-B@S~Trx?50KEZ2ZQ!*2!=UXHeX4u~*Mf9T} zxUmdMtth>(K8lv2E7$y!@F73mW<4>~w6zB{$gRIko+qI*&V-sG<_2hSc;c-(QBp~_ ze@_d#El866%)^HAL6={Tc2XOR14Q}`9z!oVS6VdvNa$hEw<)(tGV}1U?|&tdTri?r zq5wMML?1vm1Q$&;pMCvWDrfQq!B-@zWVk)L{oeWpjGA@5c8 zynlzcLnL`3^}H2|?b4U^>#qLj@t9jO2W<`S&U<&s&7su$8nuz|_8Q+$Mhrc6S21#%J(P>5y zxh?ag1pN?*g|{~FD!G33Z4e((w=m(6c)jnoQt`GlmX#Y?K0k?gF9{8+Ny6Jpq456d zZE1@)QVP*N8 z)nY-Xd45_@Ff*fQRObRWiL?kO^~Ezlbz> zZ(6pTj7fnYZ}TGqVSrgM*L`7#R<*@wt4nzTGNfrUsM_@32fu_uxHL*7$Hv|iWtsQX=dd9qkW5T?McP}jvm|Nu?unbi$3;11i@Rtq$H!=eK3R-6tey8Z|Ji{v(D6-pz+OI8OTHTV5y?dkV&v zLI~^kI7RC{nINV-kgF7aKQ^F9kpum6P-Z6zvwhsbbKwr@ z60@ENi)mSpgZLA`I)g6*>g%J9l8`KUp^ped3LprZ zXUdnO4HQuo3zUv>9@2L!JO=kAg!4{=@QUdhkR-+DbEw4ruc1)GuW-pZ{41rBJN(Ec z>yY{+vFckR)>{@IDET4EMq%Pzc5x?WfJVyN_l-a#cH(~~twLqae>S6(*{3Etm0=1L zSS%{YPFqZ(z-lPI!H)xyPq#TpKnzDjrVkvz>0f4r+f$kVA(TktrDJx$xI-4d|LV?4 z@R*#a*yif$3S94rtlaD=Pv?|e7!T`=(h9eqZt*F6MM3m7B=F-|2gxS(hO2cBqYBsX3)c=kpKGdU7ItA^jq^c<&@%SsB0tCg2$fu9 z)OXo~8BQfuH8v*g&$K|Eoe%%f4FU!cXF&GlbPp^sDO;*Sj?!_F*v@=CUSe!lcQg&2 ztPx{5Kj3-gg!=@o#jPCReS5s&wet0}JtSHGFWFMXT}5vLSg%a`)^LMw&fmB-jPqpYS8x^W?o(#}FdrX?iixmKqaj=9 z8PXZkWy_j8b611OZd{4!gracbU$N1J8%S!uks6ABbE>5940K?L89HQKF`HSY5mR2_}l@vRNe+XYgbyc7wt ztT%lEh8HL8k%MR_c7B$xtJ&TC-fq4TYKoQlzlOvJ=~&49tT(=@ilhqJT)y zY?B)TfV$Qi(QgvB<#5^3^*%AikFPNJvTj;x++D!*_BQF0knP$|4i@cm<1f^C84xiX z5*PJL0tDHo`I7*o8VqrlX*SbOwWQJdP;>Cs?B_9eemxK1-XY^7pel~JwUGRJFJh?K>0mt6T%}-^0s*vp z|78aN0B|84QTb{UvA&vlsH{ds;oT>Q+$+EH9y}P$PtgQErYeI4%j@eKKa8*89rksA zyYb{F>0eq2(5;da%Cn2^xr0{}1)WYPKydIQde{zM`)4#7jhiIm%0xSW(StfsTzZ@I zez&St>n<94_3<}z{;F(_$&L#jLD69Ns+xVx0Iu9DZJ z>ULveDj`YAv@4$2vqao+_$fV}rC7XD_dZ)4X*KeS=gIa`v#SccE2Hr?v&3wW#4GT2 zb%^~Sr4%&hTP>rCDea_Cc(sy5A%Jc_r2x)e67>!{6Do!*9+xGTL42ln$3IXnaA^;- zLP`sMj5vzFr_;oT%AF6(F7#biN}ae|}qk z**yj@pTIMhKQ#CK$wv#x@1X?rc7f9f~AcR1PY zmo#A|;Bb8V8yXs}3#6n~OVw~oef8sZlOy&oZ2As;;K6$hqUU?d^{uCT z$|Qh}5iBbD^G(_Ta9$q))@km;BPw!~ZW^1jH-Q60N0$BRiZ-s;#&M7`4xfKl~iO>6LXp;=%(S$?lP5 z_L-|NBORBlvX0i#9e#Ze=KLFGRl&|0hBGmrIlwFvd6CvA2D^a=Xtu=RY50@dC2Hyd zyO3*N%3AK7Mvm6w?irh@#C+ie!N!Ne73PPl@sxV1M&nqmJChrH@2Amc!@Dx*mJ?tv zUh}?teTEv8-JkkVz*#0RD$6zFo4V_Aq{!LACwAir-0o#o8^8Bv#3fP&V zqpR2R@A7br${^G>h$7nu#=>5wNL`IbzZa=w$}eM6Bn*Al57gpzZXf4a!frPo`+~QLcVjvfQYtdLdTZRLWYK@yqSD>hy5`baC2S+QIHwpIYRqvt+lt|mP|5ySrPu2qIRwtNjlF=Q~f?qV7}>W*F~>O-othtlndx!a%7^k%~v`w%gU_(^zIz3`9Hv+$V>oxa>#b( zn;9N&Hx+@^n4(>DD%BDCR$^;xrH(g+3te4T*vvOB6L|QCY0;RKpsQqqk9c%UfGp73rIJU~6w;XcMb4$Yy#>txq+EN)AI zFl?B08ythXlsg|^pBPuO&o@~hKKD=mW9)i%mHuP%1r~K6gxa6m1O;DjOx8QEa+NNH z9T%e#3D9>cwmtneEIAxuDyuR&#A$POk*$w@bSvetUFmJmgZtwazUWiGD)Fo3>P4E{ zK|oSt^gYSa&?1f(DslHRRVTynRrk)q8Im96shD>^*`oK^rJz=)L)>SW?G#kA&Knt; zAr!rGhdO-xc?K}!_YAq{7+P(A=-kKgJLY$O^%^W66c7+sp?hIvNT}mQB{U8kZJHYF5NkT?q(}^BM5GSbQT*v>&zl`ihxbMPKV9t##w-AReR1o(tUzdWCJc2VNi4hW5;*n-6 zwPt?ZC2T*&tUbj4br}KdQ%HYhq_&+4nHy#G#AkU+QlAB27L4cu66tyS5xL@FC8Tuv zK-imd@dR=^_~N}T6@BjoOJYv!&)FLTJ`fPbA=-oIrpGsd21$;04~1DhE@!M-!d#^> z=1V{JnOS!_q|HPZ{RS1F3BJKFt^EoSV2H~ZrM{MEn!F4q(m!uSYo~tjeeH64)EBsF z*R5Hc1!|yo58GI-4*PhMIGkN9q=6v++;QCiTqAxYM)mdtgLAU?>lHJ>9%w6@;|&=F z*@yP7oX>+hymv1-k*|K=nYB)sPa_}rqD5?sJy(}sK@<;z=f8dlv2rwiiHdT>G&WA7 z<8E+04T!Gmgov%t`XPe-wlslkYHVR_mWVE@;ci{tes-&~gJ#4(aJ7I0V5;3=_CXtB z&zf3ii3lLAgqk6#KyS*1IrEP1>CI&ceg|2hIm15}#^`ToogE)vfg!_Snm^c7f&o(q7Eoy?knOLcgX~XB@ztt(6S9hkb z+29?d038}-p$@C6X^;fwoF-bRq=GV6WX+EamnbNBAorr>%ky0hz>||O*-^JX%547l z$mGkYE@K9ooaJyj7skUey;sd;1k{O;-C>E8%q(Cy%f1wA7`Fw8GGiIH`!P0^b46Wj zgw=}uB}cVos^v4Lg>Abi>&$1~1<9EA<_3@I610&M90{p%`&LHgz>yKq*RZ&H)l3U#xtl&p^H zL4aYY?ar(hn&%ZERGs)PTX~o{x$(~xDm`)W1;X7n;vI zB(PdHio304#nSfdVQURBFvR{bIJK=@=81T|$Ak35tPE#{W9OT$dVUpL-m2W8z)*KE zF56J+mGwhL#|s9@NCdebsFv^&TI%)5X9n#dFVa4kWRz3@+nL3@@;`lM*6oXl`>a%} z>~+4wuS(MDQYchpIFEvuR30PjUaX&q*6B;guCQPW%G`OFufPLhx9tWR(mUOy!*hvUuLADb)9VOuO}2FJfc zVg0?RuamlqIkq>WZP@=&ikN{MkyMZ1e+I!uOi=G_LK!4&k$a$N6VKBJNH~^yyzPoJ z6+-DxaPFC?-E_*j|7@)wo4fG~1>uOfFiE4D+AuHn%hf26skcJx43K(h6311y6uEK) z!2*&&O|R>e)6N5)^DJMYxzTPF@q#}{iv`#g4H?y63%b0`OIZ4XXoMiy7%u2t%2y7e zle=-{_LQ?xI=*Uj#zK#Vthk77LcM}ZQUY#Xh7w~mTOYh z6oI*EIAk(<)V~XZWm(-H-|BK8C*lVMF4t|eF(EWAe9_K`_iACM16C=1XwE_lCi?mX z(``B1q6s6_XlmSQ$bqPJdjev+UT!Fz3x8DZxs5t#AfQ8^te6WIM-1gQ-Zh7mAL)xJs z>u>VMDQgyl`b9YT_7zRvSw&vkSgeHSHuYs7tT91Jb#2X9`uM*1J=jlj8PHq+8*Tc4 zm0bo_R!J6VJl2bm%qyM-Ke{CF3sl`4fL;q8a{^eEm7-Vl`B(@K$kZO;qZImbhWh^d zH^Xi=DJ?TIvrHKrs1iwt*k~fFO9QS1Q8LNwuRUUJGOz5QLIA2#M2!RjSeX)nX6)} z_|%?w6kz4TKXH|7H+@3P?wm{%04#~OB2~#mCFT~jg|V!L3Skc4qT*4$&5T^ke`I2p zA{KNNgO(9?< zu!Nj*l^Pl&gbfy8>BEd$=(dVDCB?>Pd<5`1 zkye{`RL$tvZXj&NiYHgMd<&1m0!LCY+26SEXV>hkh}qsk`xC9_BRA6MqHBLX!(8n{ z^>7=?$ws`bElJbWZ}=^yPeG#C$PR!1I)Fg!Uy4(xL`7H6JJF&zQ!((^T)tR7+NLE%uRDQ2k2S z$MSa1lsY9oYicg$mB5OUzJ2LaOod2aFoF`g>xiNpKDm`bl=JNHwZpr*iu9c~8oG_4W(7Q6|$?#q{;34BLvep_8idoi>f>3n-5f2v8!M|W`}!U=yC4!MTd34eW!VLTjJcpOZoPU>(3i2*ZOBZ z*Uc9%BNxmA9)Rv|gpw{w6;KlZ6uCj1k<(B=A6asj@*@c;AmGk)H#2%=2jpkuH;??m zKB1-djX?F3n59rF7%YnX8~Pq_MyvF7ZDvrlFW0g0QNL4HpJk^|H8-F4I-a^0y1aG4 zUVvq$rFJbW@OJBC%eJ;(+s^@l&Au_YKb5k(vgQk>mczU3{~dVo%q94=btxMS0b+j+ z)uOqrr&>3X-^vLVcUk*)kFCD@Ev*%;^y-X7*{{OCIE#OLlP6XAio6{mRO-0>>X(G&c(gIK<29_f#W~KV@PpmAN(nouYgpVK$X?A%s9iX^eZ0Jl4;xE z@v3gu({NP*2sU&Q9U2^o_nGl*(x+{cEgmm}N4E#3doclfZ-tYWA|M2lKaM*k874KT>eX*$<3Ie_=Twp;_ z{=te-!5dswAqup*A8VX_8HDLIa2g*H8dnP8z%?p1;Tie-gsXkm!t-m1uZUI&vdJ6X#Z3v(j1&s^fGgqv?7%w#FuM;b}tbP1BBq=zEV7F12fKeTO0NYG)J=+jR za2L1Ww6fkD_KfPfX^JXVM=f>g@z*=lr0T@bl+h8&kP$LnN3T;5 z!|(>PmR~GwyAtaj1&EFF1x!HfHM<#`?Lq~(egR3Bf+{sTVeY&# z7>}~0>g~FsVZ;`%XO<2IvvZcH{j6!_ipE=37`6H?pUT7W+wZ-Wkb4Zr^LA0n#h+ICG z@|>aGZCx@V)P6eFX@}_ZjaKKsQ+~DJxYe{MeprZNvl(P>qfKZoUZk~U--ygnc>=-x z9P2=k>REo?rJDAKN}h&Agg%m+Oqw^27vIK91ATavv7(!HWi<{btn*iykq7m{niG4t zzn6Es4mrx34fAb=B(<;LMCr--Bes$XH&Ha&HFtU&u~MMBw|RYp2P<^kTL#Ch$Trkyt?bwv@)hpCsNs_ioV2nqvBmHNz`t1p#ghpOvx_Tv?6%N*dp6MFIUnA{s|uypX^+ zJUuOhU5F74t==h}a`Lm~0LunkwrQxh>)5bCnao~Ggt44I(%Sd!EBf;Ml0+}44P7=J zuTyPJ-v)2LfKBNRWh~3KgUGip_0YmxFO7jtq19E&Bf)jyuw!yTr_QOJJ05J3ci5?! zh^}XG_XLi~Yxh@=7ddu4pS9W=qgO>> zun!5CV(g~1M)CcpKPL-w3?jC-CO3WSFFYfk85&of2t)SxOd~#$A&r8=g9sv6 z)nUBulM^eq1qLOqr_IWA&zp=;85f@zJGUk0W@?-?9~L!6hksmJGE5^Bp$vt*$+WKD zKQ_U)-JU9cfC)vRhl(HV#=kq;aFRQDfp9CF;92{1L{N*Rw2lcyRCweVs8^fkYm*== z28dS;PrERSJQc8BHK$ZKp@qyfUKNB5j@|weB;ug!;i%iESu`r*|&nh6BIv!HCW1te+EbzI2Zc|u)af`hT_HnC+ci^#UaD|vF!dhPdPg&=Y>{{N z)irZ3RGh4RVL5j&-NeUJJM{%t6zWmo(@Wy$@xbrSzEA0NQfR`-U-O<9SaVd_9LmgK zLq4VA;W3{i;=<*Ug8RSPmT|#&r>eU9b_kEyzT|NkPvfZ_c$?8eB()`*}@|?eXfnwZ(q{3G6!r+gY@xgaLi6VDMEuw5vQ7%A!`i#)2k8WG5AIZuhhuK~8_w_hwfTnQJYNbZ-CQGATZ%Yq2}`6~a_OWZ zn(2P8g1G<6fz2ggpBiR^mNms}DwTid%5?9tL)s}4tu?F!z_ddXU^AZf7$zkdYn+$Zm zf1dj5)W3qx60+F}bVW6V639Z873h^K+;eP^w)k*v!8F??7uLtg(RuI&UnueFrDt5r z;6ge>q+x|O<;^#B(_iVSa-~8`5fjv?+T_;=5eUjbE$;PxoMUryD9JI!^WK@nxnY(S ztiO&}x=yX9-F-1{Y6MJ`DDLXCCG2wN>9K$QJIKvYz?T%;wj09)$Mx^W1aeuW{aW#P ziBGAj!c3vE)|yJ`5s0cp`a7f|-hUW1WCsmJZN8hktt)`pK?AWW!Gd1AT@~1&6ux?Z z{)p6yF@lzKo{nnxQ|eFevFBD`aUO1dovO@R6F;S!q~oEGs&jwDdO`}OR%18pZP^i9 zy85ep)yB14!A;9!4GXHl+v3%tIG=|LpRFK(Q-qR1jt6a2#qVT zx^Ee(UCni7w+zY(D;p4kC+3$L{O|x?8VL0daVEVcogj9`@E;($zW!p0XFW=-U8T<X1G+(vp7uHr^0@JPH-0mGB8mLHWQ7ol z{9fX=={(KW{oDeE2=)r2sg5FRff63aWG_p}1t_j#=9VPpB?>~=o@8WOh%Iqtb1oB}FbOEs&WF-z z9gyT903EwN;myD-nH6j47pPLeEALFx8#b(faQYQjvTFbK2A8O!QVM@8Xz%4t6ZKkw zYz9{P-_Z#?ng11psDMhAiaH(o^Ge*uj+Fg0*k)mLd+v-MZ8|8IjQ{wv zETyuaaQMlHVAxmDm+sL($ZQK2A3p>g?5>Y~T>f2d5o(pB)9dCdzhtJKBwK-t?@8tEldP$ap|II5HBicKb^c zRf*i_l&nJ8rHe$reoL5eGKMGiLM?xvzFQ*|PwpeoNZ_OU8UPkttH zc>3I@xCcCQrhh&28&^fuDG60t?_xP8z5yCn}`!Q14SmTK+ zfd_osyTq(I&>|@IVIRZ^Ju%#9k&4U z!|72Ih1yEx2P0<(wa@szMlCKy!P)nEzkcl4h3)tEkP(trJr!X2l)C{{dLgL_?Fj*K zB8iCKz7AJL^p+hS9C>T$+XQc!`M5SyrFoyb!MPB^+3+oyfXjURlx$?Ap<6v+56mjd zn?*I9xx(`BW%BJeBQ^W`$0b9*=Tk==bsKMGE~1HcoMnhp8pIMPmnUT&)h(_SFJ?^^ z;bTQ-y(^cTD z;`J=|__rEd?_d;mW?@NA=1(aa--A-@OTdy87ipSV%7;S894-6FI}WaHe;^dLXN=BM z3pxG7;oRGEYab=-4W{YBa=bb0w?t)<`3U?eDMr;kg~U&@+OA1ZKaQn-%5%a*$o{S@ zTt6WM`zw~rApvvqC&{jYfr+^M`im90!J%+84`kzp zw`{W4N@rT{4;%Nje_?x~tjnZT2eI47ETn14Q%)jgQiS_$?MwyG$4qCi=o#PYQZ_Eko=4^Knj!nBt3Dn?S0zCuEdw9vK@2=wN^|{d2AMAL6V% z(aJ@2DEld;2awKCRQi>dDxepueE(?d;Ri(QGd%TSl4!fpy_37!gCka)H01WJ6=|)u z^^b%50#ad5tB z2^sv}1SW=(C}k+WF>VFnXSa$765w22Y;6h03Y^*}OIJti{u-T5dbr-jNQOC|kql4c znz5Tr;ankiZ{i-(kh|6l5?9~%5xhyh)`$on118HGkM!-Qx6gl(dX&t4SF1K*fJrgy z$S9#1k{jKUaK?wHwG95GMCip%5xT8Z`Ww9xSUjQ0YVc;wAanLViCVl@?G zf~MLgtD^Lz>E*Is!hM??$wAS13haYT`UbWOOE0u)on>K8;dXRNFv%R^(X4TZF3iR< z*SX#9`GeLa!+D9ur&40oz8Glp8#VX@%v4H$sskh`?$gnJg?WKY=Kt@t!|&6RbaTvm zU3eD1Yv)hk#W9LcviOvmiqMZ8yHnV26|`zxDZcE4RxD^O^3j?)OG@k&k0>NiW%MhA z&z4In0Q!Mvk&V(m7p4)Xj6Ih}dTYG&gOtiaH_ecuiZX9cp2;20v1`kt2=9k{q)&pp z--x#!$k~lpj$l2>?C$`htHs?`95p9ocX~J56)Wv5T(@U^<14NADf9@%sWb0G!_(5F z_mI;e+O@|E2g?lR#D8Vjg3^EPbR>Pa-F#Ehg1hhDIwAPax{%64g^%Ck7&2Dq0;VHr zAbj>R4_Q&A#C^UPzJ}0wx@R!nINV~Q=N_c`?|GQ5TiZpZgb?4IAfat_Ci4 zrjF=+#enhmz+1xZy;Q1}XiJTSR-_b;3nEW+vIxsuiqH8*O!&r<`Mxx5+DXw%*?1_e z)18V7R#20@j<8zPBhn#Dbm?y@=7Z0wCv93oV@*8e9Km0Y2EYvAxt}OI;~&@52^^aKpSK}~y@3xW7YPU< z;b9Qv9~xLR((9V3v>Xll-Y4wDIrP}<`>yy_&XjqB<#vU zk4O`fsY0%4<$lV&nti<+J?le85nA>}m9;rO^iTSPCf$-PVMktr+K7v1k}@(0?$>ya zALn&61bk}yH1uv)#lP=(c@aumffuf4g`-_2O?XzgKslx~!b!xjV}-Fv|3Jd%UBGmnE~&gWoe z>~M9un6w@LeJgX*=1mbf!b!#I3-SdXe#u&_Q?gk1%gz#*?>3 z&o=$;2W-S2V3P(apox^m6&1Cj&zm>pU3wrY64e@KA2Ksre+dYvxA`jJj0cY-bYvpF z{h~OBU2(;VhxjHsoq60@N!t4aeaHLJy3HtO0^A7jmq5@{i+5=85~zqIcX?2lK!h#p z#xXHl3$#6oopLd_t*7(oNyJ1UuauO+x4jhQ$P}di=O*DdH;GtV8DSbg8Mu!?d_=|=aHsz}f-g9IOP+~XOq>kn zt0s0F3s_yo@+R*P#K@}Ej&Vtcc>&qM+qtSX$`{T~Gk>_8Seff6D?6H#lor}j9mv0w zdhpSPrs<*w^V5v*{W*(R2GYAap<-+K92=Rs+*wP1eWTqE!6oB8d16uroy?mpG(5VX zFxibz$)9>})T@$OUqBpOf42d}>%3K!DEh?fybL$8KS{_Wo`zbU5*YaP6`F*kw(fma z9GtMd{Ubl0m$&{v^Bv_h`SGheLh;_cdo8!g$qlNXwRN;3m|U(}?w(!*$v9wji);G& zzL{TL#c1@s`ofRzmF!5heUG(ban>~{^A4=L9| zeP*(HJ&@iT!)DrbkIU;g;jE)QAuj_;#uvCzPLi{(fmh?W&aO3P#hS6KILviOz}T^^ zS-E}ewD(=4p>Xb}bv0bugRWYFWwqKNCT@1i-}xKARmM!oQGA(_73+Nk4wJreO>sg# z_dIh|;~w6>bA|uvBgvChyU=4kc+eb(Z!y1KO#+2O-VRujHr_jw6hj}jMyzD#kaQnf z&phh9O))9Ud(cSci`K2UR${}DiL@E|`jz!`skmHL!Ut%{KN`d7#4cihwro0$qU zasp;Qa31q5cP;Q!@EKHdy*%I6Kbq#TY2KNcHAOU(#;vTc8`XYr-#W!q)9ES9V9_nc z^*Gt-1j?boJ|h0)-dlPXO^fBdB+2+pAtwg}uRgpsDhA@Yt~PgVGFR(KNImS;{OB(!iCS>N*`?DHFvfe5cDU@ zm(>o?pQspB=e@R>n*cP@)jx9R^L6TRG%H-6adY>NJu>lqQ=;ofnd`XZOmh$9XnXD9 zJKsEpp4@)HM)cHAR0bC~v*%03A4zZEGDN7KjeS{8#^2}8PD_LjM_z#9<{98E zOSgxT3F+7QXoNnYmHVGKj|HU?Q05Ki*#UKE*7*18`~&%`0V#tiWGpX)OusIy(z`W0 zl0K2wEHI7*_p9N>Bka1cLPBCLv+N9VKJ$F~GS7DUquo&pt&!t_Bq9IgDMtzIm2Xya z-WX^KrOsZ;xif@7I`!?a`_WG5iSRiD7-Ol02{8Guje2%46&OWJ*mAqezE$}YuN`|& zMZ!I?lH@5TVnW1;>aerhuHvxwroiKIo9QAZ2K+d7&Hvw<2TQX@(hMxR4S$F<-U1Hu zl~C#!8e3AE-h)YhvR@u6Kh<5`wwqi07?eGhq**h?{W zbbj4NVKcYR3#`sNaM8cn#Lx-#0|}z#na~4^1<&8%zcFH585j(X^-t0bZgNN%KcJww zLA>QPczfg)s=QgKbi2;A$5#A+runRUEk9GmWQ>0FCX;`TtH`RAiy1a{5lDn)$vrK+~!4VP9y1a2J|~` zK~jHHQv6q_s@@RKluP`ayZMw$0=APXoPE1OgFC#ftxbP8gGtIA>!Fb4#X=v(!+YIH z`0#hd*FBPsGaLlb>$3YZpHn70pNj^FHmoc2DU~QmJ@5>d*WXdZ9#In^)U8E|RhKuW zEX4_D6%~*RYfqvj3T7HyRKLmjLrFZB=BC5|mF+Ywgs*`-yx99dM+$v>CG8k2Ti>8> zAYEl(fUNT@^DFgubv}}Jyf`DEUU%F$;CS~g&mJBcnjCk$X@kNst|KA?vvlsrOMDWZ z>cx2TayExCh|@5V$Dn~RGZ!b{Y4Sb69Q_x2UDfh#8Sbl9uf(ZO!wKjWQ7IKPRZH^#5IqO zHOhCspka89?w9^!?~V#cqSDHVzL;D+-Or+>fvkJ=?y}!%8+FPJPZISGUo8(0o*wQ4 zML{%ty8mu7dpS2FhffnWf>zI`=fx=V+|pY!EGWj?cfU<8g@rsh7)yzXHn~3=Ip41n zEF@(dj$80CDf?(7CHs@N?ymn%g0zDs9=v400jpY$!{g^GS3 zxTiR7cYxN%tNa?-BHMj-un|sRF;YQ{yG2Z8@pO>sudG;E`ta#Gx8ZL|sxalLqUO<^Qy?BUzql9_@FX+KfC1Kk z^~fJOfhqmZv5!xl_h6sz%V_TSB0wnRE%kD?R*$y@?qQ33H0m9D*ZGJAwk$D+=v#aF zb&C$NE+s~R-S=P)t2aT#)yY~DDNuChqA>CGUWhBCig>wtsJ0b_R%1CvSvv$;BbwVHLr##aNj@sufRT z!}@qjCsF0XX-(2>1=-5=EqFR;))_UPI09g!WyVXu(+lluxV;ecQ32?-fba=eO{m5J z*Nl9p-u}=sulGXQ3^Fn2&#Fc8Bb_a_ZH>1M?A@O8yzu*u0~Qwgz42FDD(%CCmruWL!YR8 zle1{MlTixLH9^9l@G0weq9mQ!5#LS);wnthdK(Y*kBDO~3>p3^Fj|4Z{2N66HQ~cF zN0h6_Zf7`kuH9ls$OU?*NxrQk3Ha<36btumv|&ju_gQ^kmRR^v;`skj*h%qQoRQyc z;IomI)1?VG@F@Z89fFTIb}Wj0C)LF1Ie2ECNEL-79gAOb5?PG+wKr~`1p56eQM$@w@p!dDxw#>Z2;R3N)8tOMb; zj`F0w$-3`zATJIStslc%`ZFFqldFQ>+)g;NJ~{A2~j`IHl(^-Ko1P$+I-b}4T^Q2FW)Bzf7+)!~tsN|*866W-BU8vTiKlCu< z3S^78X%8vLwj~5VsUT>Z*-JPX-x7a6&Y1s3SoO$y)%UU6uFI!cVc$0{{m-}h)qz$F z`b=Qwl&h}Hy{Y$dh|CuDj6WDlHNU5?8Y=4w+=`BsWt~&GQSEX5nexWJ$Z=ii0l6^b zAu(rDnq&l-I-y;p?2jKme$~3zTplBAc{^Op>Cent_*^47{6f@gwB1hQ#sMz;P(WpE37q5XpM?e`9-dbh3CX zZCtKiDIhGYqy2>MojN)bLu zBYpAWg;@cmfuEs|KQ?(q{$lXno-K7f+XKJUE#K(s<@-b91x{Q#90SW(pw->1k^TM# zEuf;j)A2{xDk#D%6+rhqa_@_TOdPvx_e=N#A~vkm{$$l&DI(YI29$8cnGo&M>=fw~g5wQUF zD1|y1quOP65JL>Kh3Si46AK z$H5QGr_M7k)o6%&kGyX0GH(P^)4CXlZF!G4lA`w_KYw#+dxI&32T#%hqv?ShCjD6( z{Oe+dAvTSuG3#JI3Bg!u$XyH!LbGEMKh{FfE(=RnB>n3=R&SSiPy_ zqoNYs-mTqjcr%duG&cSItO$e|~&m1KJ;0J4TdBexozJFJ^!TZ4fv1&R`$GF7` zPx#U)x!5@KWc9i!;Qq;t!+6aus!Uc#RItT~)(uK;> zk-VK@iGGwKuw^;`$}B`ZE4v&aQ=a90JeNm)N4in4D470!Z?LfxMafT7%S+2!y%SQP zMf!!F6yWU8a zq8U8u>!#h-M6nd!+114^&eNwEIa6Pz3 z#c_>toAkw{w~^u~rD4Iz4ol2O=_6mfMVF3dbX&lOI|_Jt=II_VC=CK@8ZAx!^yv;F zW;t+ew8BcWJou?Y(&q=MCvNLQ)QmcHgyf!kA7sDgXjU>8XxAF^L`#sTet9e8aa_j4 z0I*%yLlW*XSpX1z%Wpbm3l6EnQ)p38ktJ-U6U9t zs1LJ~mu13BCWbYO{tlg}G<=g@)C)-c7x9qz_~`nGl`DvixS;9ZVBr6X)nK!t`iEbt zHpmzs&shxSOewzqaB-$*$ww0G2NfJN*cDWs<0N#NhW^OeKkOupU~O{YqxodrR&kiJ z*N7AJ3t?i*%0~;<28It%x>N$>K7cF8NrLZ^5g*A%D^d6_9&217-zvZ4Xf zQML`$g%VYBLoO~ADKE&`4Zmy)t~-~T4z|$P$0C>)OhBPC!Y+bBM(`IF7Z*z@KM`er zGAO$eLl46OZd~r9ZgYKqf4^eed1o5TIfll$2vlobCIU7W{EKzb;&jgoZw>};eu$P zubxtvGTxKFt4e;fgq0RQ>F-b%{CQ3k8#C(+4JJ&XZyPDl7MQ8%WheK!WQ%6jj>{ep zV1LWYezJUiva)c@Kpr0rA9j| zk!9YY3Xu@|Fm3>&rqcD_5ThVJBRTwN(`G_4Nmu%bvikV->({@Rm(%HMvmD%aAjeyi zzDB*#FS^C&PI#nWD4B`fi3H7(meEoCi<4bd-?k&CEkdvQF7~jnustMFd9z0z=S>18 z3%m2n3Pgr1Nge7KL8seL7ycR;ItDp8dASToLqRkTnyGbRn9q9o;g6h_kS^MituRWc zC8FML@$){<>TMAb5zTb6PBu(-z}B0$TZKCb9Jra=dp*ZR7k}yK`@6f!hulZX!7P*{#&oS6Ehw8bLBLm zFu+pNT|Lf59;AE~W)Nv4;JEf3?uJ|&y!Ysl%Ej01&2c7h!J>bA+_TgD+&X z6szJnSVBF#IA*oT%(7Q^xF+Fk?d@W+7Lz5^1-gyaWQfY^x}@wI8n3S|FtDJ2!T{FGs8J@NjaCv|rufw=9%J2( z@nHl4ATNJkXEi10yj0-B?WLG?8-kzEhG0VnGfS7>|DFaclCv+tz@I+hN_TzZf9bRW zpO@)ri2rr#tdv-3s@Ejhz!fqtG1o~DyO5det@Ok9-b)HnE}K~hELR0K{FFr}FM61L z5EKirQ9>ME9rtzW`1h1iM!;?PR#ovL$}P!k(~^l~;y5uwh`CDgGjn^t{G>SSvx>;! zWWR337HRX?tX}CO)E)V_!y*}*}84-LH?QxAd8MQCj`Tok+a{FsV zZyAkj+(aAUj9k6Ey7(t*b-6UiVX+8U)Vk0IfzO=8?>>QwRg`n%#ZE092sLrupz1;J zgS1DbWb{|c8>`nrqIZ&hhy~?%e{~Mv9WFM2BJ2Q!hbZu(PzxdASZQd8IplU&k%_m= zIC*{${_}bvh$XU(8U^xV9nXVqTJxGR={C;p!Bl7>7L)?!b=b#aUUSbJZt&6(k;X#8 zZJ8B4yzM&CRUtZpVaXwRsmO~*7ZF;;Cm{}H!$X}4+L{nx7W5;c>y)kIPAy4yG+ z(qX(%SERO)luxAQK*_x&EP-6z_~Llmd`m`S{BdU^Yc&{pznQwb>01ea3kJ_}8SSG( zs_oL2F@CF{H4+W4uk`fc>lF;i0dsQjmx`}_|L)5EYX%cwMP5KINSt1^AQ)tJg=<6U zLE$Q*?vEAIz$=(2GmEey#o=D*i#LgYVd2w%x{ZcD(e^Awz0_E0;n#TSY-sT9wb}Yw zD?S`kl>)5*UGKAYZdp-UZA>V;-GXSjTzepX)YrpFljIb@cDgp2OMw0ek`*1+BO>IE zR2UYM38#lD$!56h&aRE+lSH%VS|et68?OtFDcPfxCD}tMi$daAr zxy!8(PmeaoRZGIk)Xy7yuCJtm2l0cbY6fbYx6NrG-2}4@`yC{%N297%KLVe5QLp~; zoLxF0=zk_o)Y%@LS?7 ztD|1kP_-WN%D*s}Dl;>Wr4S38nA$5ENHTIW3!TlFZFmY_B9@hulys?uAH0V*Ha1Rm zSQOT0&`ZT4c6iCe&b6bhrmNWG68Y0Jy33Ucb?U+Ew0~qslueO-0i&0ZY^TqDZ&WbW zZgMDN4cIOIt}K?|e@-#5>;JZL&Z~$(FhBr1CJCZ`8<{asihJfegjOHhx!H*^^Zw_y z+b;{kLblKPkU8D|1Ql-J53p;c6qfG1J(Z&b?|KBrZd681MMZJTZc&_^%}h^QehgF| z!D)U&-zow74hl9$5ViZB5J@bN)h8LQ52c&y1mIFfR#6`K-;<$87Zn#@$SrJ8I?R+$ zw$9z5KoH!Pq^GUFI^WDvN)MbWkl|4+4x1Ch3panI^Z~SP`OJmPV#t0|4dTN^CBl0o zR#xMU=OudiK(`Dw%r4D4grhD2_i6rkSBl&HY&j+>fx>`f>GbD7x>!OuN;3W@F!dOD z5tMR(T4CKmV0eKjE|}WLoBC`ieG*!brCjgsaTYoBO0eCwzt$vMvm`ONzHOr&f8L#W(3q*8epNpmTv_>nU-s|LH0o z6>V$iJ)1bjOSaC#M@#WI;43Xct->;Pc-cv$%SkI${9$?C(+Dbd^bAlp0f6zG7&lg$ zg)%C#ERD+Z>TK0?GzULJF45eDlvTzW^S`gt{tS3mD#$waD@Dc3U$p}^gd*M)0Jj_5 z`Mgr<>>b;AEw4aJM@P6hX43>lT8h0}(YKIOMwck6mGi5L8B2dpG<7^d7kB;9oKagR z2dJLjuVkK0w?4K9X#d$$xxv=~l}ztrrE2k=b%7uW>U8NCm0nQd^{hJ#-bBb3V5#S< z(6Ff16Lr~wQN>j86zy7P;LQoqvB`qL zI(X)I{+0H**-)CKxo)X(pVGxpn6Map?X# zECQ-FE#YUpJYaIGV<@TKBR@TV4s{O={ZT~}nU4xR={mNr{geGwvijfK^_|B>AIa~L zU%B(^YsCQ7($}dpeS6)jVx#3pbCriF;`#h*gQW!1q}D|%9l})rq6O{Y)=-i1cAn8C zMC{e)Kl}oJ`$y;RX*OQ%39Jp=2T`~7+T=f{dv2iodcRbVKjxYwg9^D^h-3t&h2&WE zx;V}@5>5D;!_!YOzTXz`WShQjZBvxKv{o6kY+ix-z z8W%8Olwup%i*aWatX0g7oD~>b30|A)Ajw(RV#DE%*VVN)DkxTmF-xk769#Rh8#n_CC?bb0V< ziS-gfPD4T*36a25$qZXWQec<-qhw(}Y*%tuTWXqPp5~JRcQdbrh z78oh?@vGMv>l=&|;oowPt7>X;)`3Y1*0}5z(09AKyHf}YUwyFRybm=KT8+8Ap}`w| z96LTBH7`Li>Nl3J>961V`Nh2X<>@|mng+bza_Zx*bF)*jxu9GU6FaA#h=;|28SG0JFMCsZ>fWn zHQQX1!}zDdW+QS)dTpk*QnrEvyFbQ6pH#rPEcXxU3p>!o!FmTDPei74xt7RCNPHPS zmh_PEd9<1hK1#&KD|M8GJ~V^t-a4Sq!7qJI}g`Sbk4!|w)iHbR}= z-ND!eL}sN<*T<{UEF@m%oZgq&Asel8=f`gxk!DV4s#hG5$(aW4ew}Albe-5e_f*NF z>Q4~5_=SoaD^o^dVfJ2(8E z*S3J8BJAtXl(V?Cj^Rpq7wgYqpZ}gk1<=G66J#{|ZKoXX(5eax<$!3WOze+n95%7G zBv6|Z-V}NVzRwP%!zje|x@nWMmPKs`?E@Ng#tL<9RQWPq`(ds9O7X?j5H%h04TB)Q z&G=Df2dDbQBY%}h83Qqi#_I;$fCRBmsW-{31IR_-ZveSE)}JOm@3joOOD{EruoNv zHUT|-TjgqVr%1ZDkJ7uQf+Ym^sk*HQ3HJk0@wPzdWCe1)#V-gZ!l+R<;>>}R z)YZk-pz*?43zeDi5ievo?5hol=hh$F;^XzX`Z#Rl<>?8e4XxC4T!{eEPoCQsIvj(o zajbuxft>5ifm!U0U67NE4<6pba=r?YI(TTbV)LSR=+J3*KQu)D8|~paSZw%Ll6-?e zf{RLEfyl;l%Q8{+9`DYPxB6pug2Dj?BZuCEA8P672`XaLmU4kZEX}BsJ915*7VddG zX<|BFFyqzYw7+aG5W>u3ZUv?31C1* zGKwLV#%c&{5V_wQqbsGk1mQi-6wfgiT2prsXMn^O;+rfI9H)CvS^-oTwoIA$>z6{j zA?;{L+6=|_ZYM^O^JGKqCfJjxMQGt z@93~sK>+k6i@f|Xq8YUV_n*+}FQj#=71`pz$%Kq+o6|02;IdTToHn_?JH%M$kb~vB z`xaw0O?_S*pxN&WWFnm+6g%fcZuP3>wSZ^1l_4REAp+i)4coJIZud1Stlz@Eb->9(V@r(z{~`xST`#&;C2TZ% z5(}NwZ4!cHovEitV}RCAO9{x!uA6SBrI9rz(T-;bCzV8@7Y`;jsGVe3 z6+QPreGbWK9Inlhg;eFuz28QFouOY~eq#QeR;^C0Gorjb!F5rKnd97~cWj&2i~lnZ zKOs$g!~ZVpCHkMj4fiUa^0=Su%epu_biI{WxKOn=F(^i2JSnBpA5?C>RIoR5ZZb|_Q)Dby%^Ah{&{l_C|8ug`1U8x;j0JV~GK)b5Ffxdc z4x+8~q2kqB!a_qshlRr&6bf~O5a%Vj>n@!RG^>N2vyh7|yHAw$TX0f=y_I+y`h7`J zNbDRaeaqzx4GrT~N7QxAf1<+1s&GJdQ5|Ze<@ZHzWpj9PA|e$UI@Mo)&jC~iTg7n6 zY$dVnR8V7>A}W}sDs2ky)PV9U+g#UPTlo6&&~&Qo1;5Kq-Y-}p()Es+Khj4-*E$Dr zf;ySY^^!{xj2-BV)q3(j>oDDeI@4h_~x|NoimnE?{ zUp+iwqRwePErUWJJh1s))&r&Hl)0vk5}HN!)>qs2j(FzY9c^%#RlG3-=1r3#TYG%E zg4j6qNU9r)SNB0-2S(F{(2rW=c*TaIY+ZQ+J=KZv8GM3pje2qyV+>;EfMo|)iqO5k zLPO6c=`ybu*?BjGBr>l}tUX~Z@E7$bi5pAye?%%iyuvhMPeH~vYuNx2O{Wd6wz-44 z+)t0I)hD)B$+~p+9LoD^$?=t~IG}gZ$m-ow%|kRg*nNs;vKVx7AfV5BCwT;SFZBGO zh<&N5)nTK}^QL+9%(Krwx^z|8HVF`Y^%yW+cAH%3${o-h5B-(vLS8GSv@PHwHvfbA zSC-_VVV^qkIrvX9C~hNCqT@W&;-1 znd>Qbvjb&3CUx<|0hW+Wc}gEJl*SN%Yt;B!XR*d~a080iZ2BBRvM<6?AFxs0&{;6! zBL(G7**P#mZi~DuH(M(LL`*z3j;u1rp8pVt6J9`Klc7@hr0N;3e9@&qa{-k`;8sDe z6x792f~|cUx~oYKS?-R2VnWR%X^3#(Y3IR~*Pt$L0h!tMaf(k*j5>fYbeL#M0%t3z zgyz>2J2(u>ZK(#d({sXrJekMIgJkSOhow$@_)jmERmJEq;po+gD0*0nG)sNjicx#% zECwcYXcbiJ;)g4J^E*4%C&APET7ICfv|R_H_1^JuMc)Y$S(ZBz^|Y|zoE|_jhCEEd z!m|~LR)~rhl9EeQo#yf@4isLr8U)k1BGd+ zV5dU;;JMl6m=&ECQkwcMlTHcVoTsL2dFI=gogDT-8`SyYnd;#}-^oF$)Df}xHF@nW zdeZdLdCGEVPJW5?p-O%f;o`~)e4Y~eB20`bhFuq?jPLGfXj|DuX*K24R5@z?J7S>W z+K9b@6YH7geSNj>K2 z2CLxaBj;q%jVOW}QA7I0GILZ!PS1$X?`MdI>XR9fgoS2h(o{L2jdbv32G!=yact(z zu~hY574sSQhs3!XFR?vZRah8czLq@RDAaCkb9OABL$pT5n@nbIoXuC`pOU=JhKv>k z%UgPFR$J@N?r&MQNVoqO2;TfXJP+|OlMj8xE7DYBvE7t90*9>BMWLFxo51qJun*S)5K^&7tQ5iSH))=m{U#{kRo|a6H zZ+tFUe1l%k)aj5~$uyeC?}AbvOh+U;ZcRUwjYoYyPuB~3N62AaXQf@e)0nG`ZBwuN z64OGpK&M<_=l%PaqdL_ULN4w?2YsHO^-Hc$cmEn1{a3njGlJe=>v+1SUk9Kt^J)MA zOeInGl$^V`%{JOCio>2u&^NAao{wa|RV&iF?TtE8JrWCauPrI?c4iF!)k*HN1r|=t zXl!Qdq^4{mwH=3^z#Y>*%J^a7j|i{)j5V+*t*!~#?q!rkg}wEo3{b{a(mhtM;~gu| zwirLDA!Wa3`STkaA!s4xA4DXb0EFGmjgO_|XEc*0tM{er(M5~jeJlL@l6T)V0jLQ% z0X@BuOJT3`EQ@AALb*!KfQ^<5z-G>_1C^Lo!N+`jk$FMOm z6*s%ixUUnt=foB}#_awdWA6dgRJwML&p4tYDvAOE0i-v90R=+0(R=7!CG_4qj0z$O zq4%bALX$3CL3)#3LXlnrLI@=x@O$Ijd;fRto$vpBx|WOOV!_Ed?|I68_OthGz-930 zcuMta_T1iT4T4@kSN?foD#PCe^MSWc@OVVGZa%y*w92Y%wCjA zP(LWFMvthl)keAZ9Q|<@A8T{Jw29C}5d&m&o#*;QZunF?pb*=VK1_Uk3_yjb(3O?( zqj~N3@7Mi)gjV@PzTsObKX7T)V9j-qKGeck_NEqdQ+Hhz@McQLSaGyV7W1!D`BWSE z`lr-nyGlyoa<;K2gd0LQ#mkJ~no5vXBMmYOCWv#dK`FE!^*E8Nm- z;!AB@=bUvtG*mVA-otU#>SenqEt==rcsx z4mF(KFvFM6F`LH2hGGy#94_`<4b}yL_h<^9w1K9{`|n~ksAyTVpTcWlIf~D~ObMET zGtzLhT=*Kwsb8LLz|M*75v+OkuQ>=DYmIWH)Nj?W#p@D==(|oquR^^(C#rw|_u2GG?Jmv>aWK5MJPus($nUwu=Yc=6`5@1I}TBmvr;83Bj& zKqWMj9*cj2rKjGMVzm5w&_i0OB`O-B82_-Xdv4)aclK+ihFN)bchn`h@>1qDrRQHe z2R+Ke5e@I-@kr)bUiItESoG?CZZ;&M!hJOSYoV@vT&U04REO6GL={Tb&+nYi>0V*7 zbl?}}yt!-H()oaYC0V7Y+8&g}Ygi*niLeZYCly42pXuz1pnS7akZ*impJ4lg+^!lZAWj(MA057#{fYNy?7LQ)21mTr0TB! zDxa~lAooU;(cQWwW;3+csM1GPWEw;E*X0{eSrrn`g&3xKmn0&qaiU`_$YHDBcrYK zMg}prOZ?3zf}kn97w(^)so1a4Vu&_VQi9?DZWk-h28#qv?wg7kcpNUbURqxV zr^M(9%8gOyo?6>)?_Z!B>}DG#s~u;q9aD6bnH3Mbn}KCq6u&Ij5y7ZSu%zniC1hqO z_4u&J>NU$_V$`HG#$PNtTOK?RDcE4e+oGzARiG{n)nKMm44~#honnKvk%E{y_UwTr zpX-wv`421ZVrj!DtIy{Q~-<1Qg;kF{_9nAVnS`i>~Z?})w9H{zZvN>PwGOWFY{#tp4J*;sVoQP_V<{@-h9hqqrA7=lc1|@w=E{|c&wRG z(nm|-R{t1f8K8y<^bve~eR=fT8xV^aOG$swRIf{UCcIT-z9w4z4Acrp12xc~j5Gsp zEF0h4(sa~?i>UaxMZXctI3N0gvFsX-=MfdF7$UDEd9Kdgwtn?v0>5c>aXlCpUfkA% z$1^>C4MqDFsIrcxq+;t+Z;FbPr*?|39Bc|J2I`cW81u3WtfXNZ^~-(cgQ-W#)6>)S zaYec~+pcJ~F##5BWWME`wKjV+hOY;@$ymT$>ApFJ$FXOv>IZ!UWk{iYXzFrN;J)xO z8N4cXIgCp)H~ZMU>Pk!uQCICxX({5o{GS!`4aAdQxQ$P3{q*bZBDG4b!HoH?I0G-F zD?o6vwQ_3+~P0*UbF1 z*9{FR@8q=%o{&^i|0ueJu<EY9BS1tNuon&& z*vQ?7g?RuRXUuIz7^zO9D6$zV$XF2}?b_8Tf?ds^2FW}oZ_56$Lp%@yxiLUER5_Vm zR9o*oDArI=;503}Oay3F=)hPsV3{sM7EvA0IyPk9W52nv=nGIlG=4$mWGb&TS-0NA ze>4OD3G>O$sV>Sct(p!NhiK_vNG?-;F*WfYX8(~B2PE#)^Gtrbb<%J$CQTzbCqL|Oa%1#d)ATXL90gqo_V ztg><#h2s;QD9hI_^wMjSX%dSu1ce-?8v@xa(aOHx?mc!DUtCzQdx==ASdJZ8O(cIC z&`(1_@e%-Xh5Rie0~^B|aI+2O zIFIHz6UZ+GPzAB@(oynjU(hZbZyC($DBt($ZPAdIw@U7i@LF>^!8SW)0tMb`@_xjs zoC5?siw-~i?IJJ$v&c8f;g&ktzqfdF`npEy&!TVkVr|ctk(!tSj^dgWV z*6Ja(pQ_ef^_kCX^ANdG@d!FZuJJko65Jxw9JHKK%DmSZ)y%Y1sJu0gZm}xCqBP)>+j_Zu`t+9HdLL+Yl)b&sQ}%}$6-xl&nS5;& z(fT^#E2gxmxTNaGk4t$SfGj5JUUHhm!OyOoS~RA_!_pomBejUj|Di&OntSRKxarb| z_#L+$(_;F)@g#$G{INZ%EZf~*;7W}D-{oU4^-jV{d{-##k!Zvmd=wyWXS~;rGfyTG zt;#h$8n2c2xLSEP71*S6nx6{*?_mjg8*=1sV%wGMUEL5`Yv zGZs()}v^!mx!Lz_yKY8dL2 zeOm$qB81FV?sn;Zd~=K6@QXuWeus`gzqIb`$q>>Vr4{#17X^{@n6^4I&9Q6ihJ8@f zc0L-em-S7UNabf@#h94hf&BW*KmWi3=R1!x-x-naM!S;xFJEc70mK|O%}%dQwzfV7 zP1GGW#n?zdj9F_IA4m~VW03IrVK8O~RG(z8BW4-q<%`c?+qEBS`78p_9DpGV2q+eiJ+RefhcOukhDb1mMB!j$CEdQ0`i(=m{ay_e$#-V`e8Wo z);@P&sQ1d2z7i!X)*l8NnD)l(3#Y|gNA!AEfuX|8wvl&WJfEbGBG?~CqV5Bi@nd0K zlY;U5(9>fOnPH#2thr3;ii(VDq%=@PSu2q?rsLa^m2YynP5omwIk@zz8^9!C8Xzs} z(aY0nsn8_(A%TH`*tc7xasY&mDGlsosSV`Lk4T9H`eA#wy3$Jj&??j%FLAm*;hgW6 zAYSTRvi!9NC9X}Nv@n|Rvxd}b)m}E(gSDjDSXR+i-Cccr?Dgn6;-63BmtVkLZE#wm zSmB^%IPOWZ;)mZM-ofllfVeaK^Th0{E3<>k(ppXy_ zuA#f8O^+OPk&nztFI}qSfzV61>48W4%`%9uXUl#VMAZz6AG$!-E~k@aMGs9?=pnQPuHZXFny)c$ebanb6w$sNozpkuMu?oob2(M$UDZr94m&Q|nRRMU&=2eG({S4B{@-!8xpn0|_cOCO@g zO;54wx`Lg~R?Hah(p@9@ccs=2KF> zG{4a&&Z?K9&*4w-{xnV(IW$_<_$lx1Ww-L%eUgEnuR`g8d?MPr;c_d+&W3f zx8$yIwNV7&^TSs+!K0N{X+EO{PT&b2B~7zh?vC=u{;0b_vu{`Z*F64%@T1owM?C1AS{rz`OyvORQA43&w z(uU2zn?XTN&QE&jO7umoPb|a@mzgR(Q27Ql>nuc-W;6$!1tft`+@c|T_9GrR-`%re zE(q>tV#<}+^BLgaZ5hlc0uQ)hl#wb{dExsHQ&^_5D3JdRUv4v!+y)Nzu)62wTv|mz z&S9Cb_f`t;oZ!Z77Om~!2H%{#cFc@vR@*Mt%f*v?=LXAm3E(wZ-;e}jqewcDX)OBT z!-wS#gbQ4bkU~XlGOTHjzpkU)4C*wR+5YqINZ{?5#_!-S-#drI`{y(NCus*SN2 zVAQPi(Coxxdn#=!Rm2f^@Q$15kFG72Qx?h>>`^X{tPh9c+65Urce*E>Z!W2wojh;> z(PqqXg0n2b3_DR+eYiCv7tm{}jvUP|9%$62Go0_7&WmmhD4tWFJ=^D5p9Yo`s?Xxh zxT4Efk@4#zaq>|SG zroMehj0)g5z|KUL#lL)TX`7Te%Oezd*@rm69VLy5&R>{`GOUrrCV$OS(;t2excKlD zTS069dVZZZAmCRYdrzNNPUkPlO`fzG0_~SsH;BN(H9RY3u!D8UUY4?8vK;#}s|9?k6f z_^cHxd=Q-VdjYKKb;H^5IH-N5Z*}}GxR<`aHY#`d78iRhhVq-e)`D7vj(s+~_Htcv zwAk3w4yL-&B0Rg00RH<@m!^-jjAHE2s`e)@E-v-Tpj*P~HKq+-Akj@`HapHVWdV=V zTDyHym8|B|-T75nX{xcsq}A0&vc4Fll7{N9GrH;^>6jM?#{I5%u6x(t!FY0gE@L@T z%sns$uc1-klP(K9a^HV>yntWk7qmON)MIw@;GOq2z#xWKb5%8; z^Kd`wPZP%;0ys48^XEG$ge1I1xk0U>wi_17818B}R1`eoIH5OiRVd*2GG!c{PF$H5 z0H`f1t3Jj@^)_h?klK!*GggjNcTl6FDyBwvO=}fvfYSE`7wksWk;jA-|J%6Yrv(Su zKpW^s@ph`RcVNou;YzE@B zT!2+&kMt)QX!jIwA)LEOLBYOVBf)xqi|_0~A{Z5QF^_|;1gu!Ze~I&UDMR)uAbz>w}HO$(rnX&)^<@3;^1*0?SLVMqanXM;&d`NzG$UFxW zE`p#(A-pW&Zq+K$dEBjAf?U3VX{~W*L~=vfdbDFHa=LHhIaz?VL+pUdN)&)f3{1XZ zB`%Q;j{ecqm!|uc1{Ke#tz$DLL``{FaKJE%9pW|EEl!sc3fC3tinN3 z6up$aBki-5=$Ep|u|AShc%EcNVPxC_RZ{;+F~MO5dz9}C!^2i*YHdA9uU%8!uPYz2 z$`BhEYWGt@&W)gvj++}u0}*$bVzuUeww&(R1VEcPa8e|?7-?8zw>4`5c*tqzp zbF=xcXQK)DMb{+#@mbvEw1twCl$`pogULm3>L$~rpJXP5-JLECjX{^icKcd)RL@?- zn;VGoTqn3g;aIu1kGGP#x>K}m9&yPh`Z%-X;>mC4u>O%oZ%d&V-0AR>rAB$)UkVge z#h%@z%!59ahSS^RO-P;H^KmT~hobDpZW!jmUU zN>xnkD=&;bc_}bTa|drI2SwK4ap&eQ{VNII!Gk%IkMhV}()$1QjU7PQ;iQ28mtK)952P4($H`qX8q8?LpQx4me+Rka4cQYpY z`AH8>!6)6{z=;U0Mvb}pr}rMSJI#(DhFK*f?fP)SEizdShD&8Z`bWkt(mm5>6TxvvBK zcO3j(X#WxNA47mXNs_dKgM)!Ew_Z{9)|dtjr<7TKIBqeaNRz zv$36PR<)FL+#FewH(p);sXF>3G(3D%+lkM(iGBV?ls=Al_jH(Do81kWnYRVHNP#fv z-MW>X?HQ-tgM+yIwZ!|<4D*++Bg%K*y}C@ijC&j29~2Ep&YzbD#OSs!^8|mom#cJAv$=gp32EZ2bUIkVlh_|9XqSu^L-l*u020;#J#<;aFJQEG04L=QnNCxy#cddr7XEw4RQSfx(30CgzW- zpC^o}4`7p6d@a%~cI?nKh~)vaJNF@!nblRcv1=O@_y7WdWXp#R!RqWttj99rQpM14 z;1l;q5RWBkx$&=){u#vr;TqX_ML0RP|LD1{h-c`fh}cNE!{v?EUqlbT2YbQ4934ZL zVadO@@dgo9r1ZIf6xYJ`Hj{2?odKN1`_w4iV%L^2D2tkujEtIG?dLbK^2m9Edf^YJ zCw4ELp2)9+`7pUTyT<`OT+hqkcgvuZq+nohsEq(!eanAb4jZIRK~9Yg3e^Gw=ac)RoUgTFbmg86F!I4oWh?uz>zh|YAK6_9bKzMC* zPyM+I2p~<;;&&g&Kz(lfc}<%aiAU+p47n#aD5l}Xt9L=26Fklx#PYAZ@kfms6td}KKNyUmi|scmfv!Vq+?&(EqUvDGScdb8h;mQ zQ~BqnU6N9@9>(%ADKRKVMf(H)>$%i+(gAkYZQB-En`)ZS&M)zd9A%8Y5fLiSGOTC# z3NkcN=|QVMpuzTxdEc5Q;MEERrHpycniawy+g)D7K!cr?1`@MJ61dXjY?{@OGZC7i zjtqH=W$S4bTzxeoy$KUq{aGW`7-!lQ6N7Y1>$HS*e^pkI*Abe>2D{L!uI$Bx^l8jp zVakE}N(PKGw@Dsj-oio0Iy5gKErP2-O2Y5x+o%4}Xkt-_`gBcQo4mAs`rBxD%l2!_ zdRuYuP{D};;jBCWAO`bqSZnXxo_`bBdwf!PNi_@$`+xcKbp|-vV4+?@aH-M=1ms>}kW_h%Na8CTbBwua=t*Z^ zS2!*sRT&+j7b6cYm6e_ty_gkDg8tWEVMG5iI)u-pj@vA+H^wx@Jx12vM{)STKYQSN znV`?YT^gZhxo+rtM#McWts~Us;f*QF z#C>-N2e4L6*enrYm0%X^>UbR@*FCyfBIvi}W~JpGv>vhQ3iz!#CkNCMlIo_<2kT%@y4@zSPI)@}S!XngjFmYkh}_-26rdFgO>4YU;vm9{nGGv7{o@LqxG`<@He2}A zK@^Q^0`0(}WzjRXpa%DU+&-B*|I6*;EEYhyUGE?*o?zI+rfIHwzgjWyw9R`W#K?}+&0&fS3X~0>}DQ+ ztoJ)z+QopwoOaaK8K&LUYIb3?+!@6uNj+Sz#&(zFVo3Mm#Ke=s+D|jFj*25O4*qgh zU%IB7#b&pHd9V_ii%SbDd^HjcChN31?!E8dKb^S-+A`(;^9PdLyrw0+E$wFnkyDL7 zecWL_7GzSHe8DXZa1Q z(a{AJ)CvS0Npr%AalBB<0D7x6IXfaQ-aZ30)11990gUW}tTE=>EtEdfeFK8llwW#1 z3WeCn;!C2o39^ctC9hcT{LVRpZgyIKD^OkW6`;t2{x_nc%KcXZMCaX9J*&Ov2zZU5 zlKL#75&3hojBKx>kyJAnD#i7vuIK9dlw<7C5e_w-p1#Ww6d49448#D&|M_Rpcm02k zgX{XP-_0w{JM+1Qvd%bbS6m(+ZuZC)o9b0O3vt_&hEbJHzdt-ij&UIg4s+fQ0_?|B zuF;FLMYUW0q7*oDHQ~1qL>F)#WfK!?UulT3QG-2Kx2f?@>d;r>|M1My{AA=<&~pAV z>D8-B<^Z<-IhFcW+>B_Zl{^eOx(CZ50ub$N8*3VlXVW$f=Z{bTmC z#yX`l&)rM7GUTEcca30Mtz-$QIkK&M6xXjOJ~LCi#6$y1oIfKds7r|{<9`g&LlhM% z;1~|>b!Jj--gD!9_TWL}eY;`J++ID-FJC^;H0caZ>pIEj_Zn6zUItP_iXrJn8H>ETZ(Hg78V_is8ufMc_ z191M01-YPy!YsYORXZ-Sw_^=`WyE_{06}@(OHO(=9by3pTCXm)A28KP_j0SR(CWK0 zgj~yr_2Kn-^KMhc`^S$D+dC5+)uVAOt&8oZ=nexNfS`ifFo=ct7u?kI|JES>BY=Y0 ziNU;8dn(&dHR{e}xoh+GYM7xiLA~yPjw>dY|GMYqvv~+CU{%Y7B(55xE!Stw!`;W+Nf%Stc34u|+QxuK;EaAJsr zxI&S4czg1m+nnAuTT8VD-!urG2PbTe{t++C%)cJdxP0&2y|wznyFU6g6`vtbj3u97 zy2!xN1SyC0P)RA3Jq0#zaJ#Ktz;U*&m?RBdf4J$&AmtEm4^H_KBx`><#} zJ>o4mczX`K_J%8!6w$c@#@zJ>^koo3?Q|Fl3!J`nhhY&Z&2HgTWh2g`W)Te1Y?|S< z6Da4GZo)Xvm4@X7zfqazW@b05G>RiMC{xDlCM+t=GC>}PwbS_LAoUBpR{;`97GD$< zwP+zf_r=V-cS%&xY*uhx+ArD*-(IDOov9^D2(RuLc=0EW^=GCbCO=}n`8kzr88qa9@amNtkjaHaWK?S{cfH9MMkGn7>9%OJgklBm zS@eBX*X*`JPv*6Pv6l!`%2Dy z=}2GDbN+H8A+C_mhmMV|FSTCFz=QS<&$vTWOuKWO-93vTh<>@ttDKDp16wic_D4x# zwoxGPDw1i}A2LQ8`(Epu=&H!kSuAmBu~KrueBN@bvZxyWEZ~Cc^(Iu{?fGdWo=0zq zOj%u)T#;&M@OQ^+XQ*b>E3>Pn@cV`jXmO}z#H4I9l$(u95eq3Qgig^UE6Q_wlt6KN zj{L@r!)yiAtbwxNcwxE#&J*2UJXCxpPa|TEm^Ddzhs@5Fs<;Xf)xK~w1HOg-V>_Z% zPOURBv?AM>Tn?l;szbjBX=*mRwE88x{kaD;Vm5EL4f86f;h3=*gY%l@QM55>zT|67 zo8MgD6zSD#96+FUVaHUwq6HA46<;%BXX~_7QBRHj+j#YJ!;u=AyrW5fzLTJe`4fo( zIquH~3;z8Fx(TesVbJ7euekd=-_(qZW45~@i^NXmOTjnJyw$@*rWbn zrT)6T2zS+VQU2dvBNDKN9w^A-(6a52?RP60DF=pMJG1`D1*qGa^pw{<@h1-=d!#A9W}GD8SnIYfsjO2hveG!Z%WqW2O$br_78eVI5@S-y zR*o6OTprO@Zu9>7;^D3tMsJ1Jz1AV|?)_Wi~ z{myoL*SPje?p4sY!|JU?^cw>T`wqw%q;9R+mc(=GS&Tfw#iw1gidrj(yYB(u-#YT! zIp)k6M>aCtHhp10ZSkvw*V9^+P^Dw;`_-v6SZMFP75|Jci($K=Wyuv0Y@N4Xg{9{K z{OI>Ti#((2VSn~GVNHK&4*z^455O!!F`_7o8zQv%O}Esxx0BP`tZj;S0qRnoQ#Vkk zS7?Xy z&a#}t&x3;g{S+zg5NdwYxc7R*5HLnR!$m$}}55l2Q2OJM3}( zWjSQax3urjNlszm5?mDWa~Aid@FWS{bBh_|r1-HD{bFt7O1to~3=-kJVTVOmSd52% z+ko_E3txUS2ShqliA=^KhtWMFw0nVa6H5RgyKW>FTgJk&v-wuvhHew1kR8Zfo=``6mIi!a&yN3~s8dq#416#U{EaEt|3Hj>FxeJ@CjDfwM z#$)8#Cy<0|v;vbQ_hl|Im5Tk}Z*KlQ5Js%$>_37N_m$ZOKJt8{-(PmkMl`o~A34Tt zKqe04JBBC81OvNl3dS4i-PT25{i*d?Meu1UiV=*28p=n6i~ zI3?tw@!jdab=ZBq?F);bK(wo*bc!El!I!7{3=#!+eq-A*bTtIvlkrvS zffp<|OcX^6bA;Inbky$9at$j-d(_WeUd~N8=J<6bN96%MCtf zjJ0cgXZdIfe7*Z>`*^sZ?U&2MjIjO6s?b>^IIFGQqfCc6I^3yJ&;+`t^vcOiS`}(g z)lWoO>J6MqgYpMsEP9;Gu7F~}Hg!&zm?M<$<36J}wxI#PK5Regt&IA()mKQ}GU&{u zS5^RQ0a&MW{G}SU{8SE=(7BRfWf=}(0}6{+wjdIP|4VVDI^XHH)z&i(-w)Ckd^+)@ z$DY4FS6%Y-YXN4cuIE8gg;Q|&`90*Anw}Z&9t7(4$iQQxbonMFh-~e(s*!P<)oV5~ zA~TNfP<@6jtu6M(l$^p%xX7B)W@4**IMCLk>rR;-vYWxH70klcHgOZ&DJg4~>3odt zIjwOimZhVvdmh1AL6nGCr)(VmUEmjY?C<6lcGsX<_QI7k&K_OIXrayO?(Ql>cw2H` zdBEbeBq__#O0=+^U(r`4P2ve@%po@kce*9Q2%KQxdtt#tzTa;fCn;**gHa|hhYz?Y zzNQnmMjRcrlRp0Sp(lJRwpI|ecC(5~WY_y$cyD(5G{B?H0}QKlL#=HXiclYoX=~6d zH}m=uz8(4Bd~~Vh#%DSzM6c<7E{U(K2$4($P@9*a+WB{IiTL$T0!Ov`MdZ1G35ALr z{g5V?v9-mpH#VzZ*D>zP*(#`mi8x!6QL)RLqGp$6K~>Uokmif6yUe!6sbIBSS6nc) z+G4($4I-_9gyL;pNO%oDL?77Pgh%BiM2t90$y9Gm3YbtSfLR+$WgZUeV#=!P-~Pc?;dsbS>!ILa+$nu&6!foTgeQy3=YN7Ee5eWA+7X-wH z>dmhO6@qsATGn%}`Jd(vl*gNjMdm8)ODlKg^9z?a2bZ*qG%Jgee(c5(E=IfyZ19Tt zJtQJMzOJvnr%kEfIl5=qvfOG?Y<=4D(_q3Z*LGlcZ?M5#pZNC#&zFxY-`ux334;f- zY+?{^-o$SICQFAmg5L-(GUGkU?-u9hfpE>|%4YEyxRoco>GTz}A7gIJ0KwNkMHj%; zY_pVR=tx6lHUWlT2(>>}?$fL^^Mv+W`sLF6ob#51+E}qC=z6+ouZ6Z{V#md*D{0=HC zyd8ct^%p<{iC+`{ zZ-|dxC8hQaBiWwXpYn16SKcsz}^NgBC z+8hB*6Xgf9<*N6YW7Z7f^Q^Bz)M7s@<{DDD7v|gh?9*}G%&EB14`l0$J z)kQ3ijiI9N-YWn2^TFL1OSjPvU`Ds#*VhJ}H6jMOQ=1v`1U}PY!Tt|sd0THD3@rE7 zD#7xsb(3nms;2AETue+KDD1ww9c0?g)at&IHGm2tvfANahJ@<;#|XFS=~*^~(UH&0 za^D3$i0pIjEwxFZv=z2aS}Pe{>?;#|ccEzVld*v(ja8~^>c}D(;pC*47m9FjK?+Io&bZj(ezrv04bI@{&E*I8Yw+ z-!O(~rSPuI<<0#7;9mSJY?Vw@3EP%ULzZn;`oVym?7j*4qsOgo^&$=^mV~s@^=<|3 zynd5NL}LgI;iEo(zixD*m?OWV=}alICC29j6i|yz>XXkZA7b0qU8zmGQrU=s_yJVW z;yH)%$K7^>+n~gj1A3Gn;9&qQ zQaTJ^seJ!T=giyr+JF%M>G6qcimRrEU>`tys_ZK5uFy*kpyz!Fyt|c0> z9Ggbfp>oUvXNZtC=yy2HZK|X60ZzcZq}QXN7)C4Q@cO<3O0Hcsodp=pM!jg|P{60R z?w|Gs)S~CvUdgFOv7>R_N7fj4(C+OHf#Kz)byq>Vtt;!_W(F2VI_^*m}RU2Ql z|H^p}%osZ77xS`hn>E^{@SPy^ng3Yh8p^jf)vpRjDG`Gs`+ICN>}C>NVipIDYBz&Om9kcZ?|wAVE$X-5 z|Eg~wOAAr~1DQ|&=*FCheczbz1v>k^ib>rE-|3U;t>Q);hEy(GNy#V4_GpK??TEmP z96!DDP6n;nzPEPaoPw^#hBUk(wcbfz38q+Sx|n*qS{g@#;#RHOF>*_I z%4cDT#u&}XjSF7e`*w7)5YY2D;;>aN!tNmF>mA0$i4uqHx@cgP7GH&A`rxv>;cnJH zt&+L3(1d*Z1XLNDNVN@7O8T2w16{)?j>#(cix-|tCv83DzB>;^UAD51+H?(I1EaR% z#%(=491l~>mfvEqPh8Ku8;smnwqztPSc3H)vUP34*oLJIkKt`U-E%QwXJ#C#JpdkX zH_sCHe6*y0BD|t2#Ls7fNV#c~I;xj!n;fT=cRo9IF{61s0o$c0tHNxXMr9CHjy16;6E?u#nZ<6fA9TWX67oJ9ojV#@;um?U37R#2JEsB)m%~wRtFegOV~! ztugI5T2jIMjNQ)Q>pMzFHt!%G0lF zPI!H_+-*`&|KrmSc2fm|94drZ#Q?L2MT`Q}e!HW*KLrHH-IcOjtXm95+SF-^b~0X^A$?Cj(v2MIyM z0;XwEU78X{bC$tq~i@5`RcX$d+>YJQD_;6_QU+Q-NRrbV1at1xAD-!-n_Kw&fL292E zkp%F&x`5u4IVPUB?nl^RA-{0l`UJ#W$f(EM6!Y20uRajpE zBE+4^4vPOEy#G9*fCYV7&ccu-+}2tfZPhl_SMGFduznBza>QyvLvhrK%iN|a9)Po5 zk%PlaKL%r5wt`}&Bi@!zsu5$cm~WYk*kwcl0;Z|3mNTyTF?&g<#JCHGhBpDmwdZxj z_hM>9FIj)#c!^Tug}w!%#&?$5Wl=vb1hFu0^n(IjN@pjb6XrA!@fR>7+DjzZeneYeF_{3iko` zc@?-XXr%0F1U}gq%S!M&Y7-eM^qe~qnY7e*nJ$h%ukP?$_NT`f*L(K7cy+NV(prid z9r{?K-PU=YB8RQRvvw)LMhL!MzKj!k23XJ(R2~VUD)Ll*Xa9|-5lcNEO52!NKy%}T zor6J|a2c%^X=)_B_eRQ1zCrl!DT*Zv+D5spi9U%hp3UdsC=(97{-_hhKrAudHbZqr z!!}0jv#HMtgFmkLue2?LLK41NO+S=3cgK3J7C;eWJ$%OXhw-Y$KAnI;+%`jqUk){A zkdxGxIunan>ehX-!f%wy#V=ox$`UOGaPA5!pEftue0DADKA zGM<#Oa_u*ijNcqh$I=NamP+~O0NKTM(X)qAeYpqY&v$Btub?f#M!pBSf`ANa)+ZU*(`a z1fb0xo7I8n>Pr#1*(Z&a%LYX1F4I~36QAsd;Mnb+5;&6tT;my5fh1?splFH7P0Rjd zUdDhp&6?F${7o?>lD-PytnjvIA9rgnfBy1Vc=i2Mz4@l_ao|p1Zbn63TK^BG^hmoO z41eE<0vw8+Rn8`pN(+B`6%vmK9wn0IJjjT#xy=-k(d?uVKqxLVIVzz#<1iXI-O``p zr$tMC7?A=tm~0RrBqO0|5@}&OzxI_>vONw7{4&`TyLL`@KZMcXaoH>wX)e71UW~B2 zUChX3$^(YNkK=02jLGTx+KPTX+EqXi zz@zONO`K2sD`o!i=hv6y95bB2JrX1GKZu#(52x57CpeFNHX@tv!41UhdnY#uRVcS^ z!P{GR4IU}U1)j`BEaqgbjWg3vrUY8w&?sVy+<$Y|=EY%jPLG>3aqHnMby6xwQfXdV z5guz79W5XC@oE>Tw)Lc%P4p}Xgv`t==8=onCHuiykspO-A0Ed|Hyxm*??^B8IW}xB z$2Q^*x?1HzP1Tsgt1rydnt9GCg^6zbNbt<@ey3vqV-AAP{CCAi9O;_}W3R3oV*rR5 z1g`<#(xO2doE7%zwhUC!-z3gS>jRAPxK>@`o!h?Y8SY7Y7!8NUV5yGSsrGDCZNg_yNXjkx3~K$36{FF=OLglt=|;B$mA9WsfueVTP01b0thV?n zSu~_i3bd-}vpe8uXZsGA)G-1DxjW(>+wqS>B09GveX#_v&s4c}_O_zJZlb*h1|cQzHYquI0+0jT#28om1@C?tx;Ge*A>qv} zhhP|9t#BHAu3zO74gq^fj~+k12M_D*kdy0JJ17-CC`DAX-wpwYJ2g0 z=w!;P6^{??7(j+&sY|nwrVH}`tQ+X0N8P8o?pc34@udnqNJ#}S3KG{jd5mPx)NbTq zW`5Mnlp(k>E^Iap$kLe|-V;A=xB%yF$-@yZ)B5d9#!~Lf_H$*LhlNOCxPL(W_=`w& z$&PBf8TJ=1UMMt~fUdgm&>fRouac8>ElsS)vlgQew;{yq%E=`=Rj97Cch{R(QG1%JrPoLzw$|=KK&tyO`VeP)1ljC-*xYy=c+veQ) zJLxsVS)?t*6Z~+Wj2Yj9vNuvIwV)`oq6JmE#4gmnlZ^i?O_>8o(Hk&-VZw16HfoQ6uM$gIvEcv!ylP0WVGhY?*ta zvIZ$23>I)lGm{$g&GL7VuqxRTeh7>Ml2q#uW=#yF7{`FvF3sF^l!LF=n%Q^}ws46b zcG3o=^Vh9@PU%0FT8r7QCT6OE=67@=nB~*z(u;G;;Vda&+(em*42D z<=e&wOo0C3isuWX!qd+#J{*2>X2Yet%y z&n8t#K=KsXJR-_?e+r8vPIdk}mHh?pndj67Q_|qw0+M-7S28)$oUA~d3sA``Z@00(j=Rk}Rha-#jq@!U(-<(IS<5f@teVsqt=>&U1@UHfLHY_TE7O>G2FJ`J7qiCZ89Et?r zW_`ZO?_M|qC=5DA@APzu`!X&!>_+NKSv=io@av3PqGL7K8M8_=W^*mgxx}?b`^if5 zWN1+dGY5x*9IQ$SK-a&KSVhd>DN+uCJ}cuDF)!fop|=`qi2IY}S@YIu!CRJlOxD|c z*SzEzTJJhlIqqsG85MI_#n)cHC}@~ft&9x1+?|8Km$`3``vPFZzYhKAxUyd|rvPH; zq7Cs&K1^f~js>V`<*7Ckc_G81h~~hdCF#9G`l?aA8Gx4THn>K`$Cc;IyKM1axNu$b zR+0Oi_iJ8b+r;&cfz2^MI<0egM?-4ZjA36TO7~)uGMCcByD^8$Iz?aZ{NJR z%m_0ogJ=e+YHfNqke0RJbgQVc!2$zV^X|o{icD?4hM>@s4i)OJ%z<%8qD;{jyq3w8 zk%jX9dMyP|LJc|@lHquw%Vt6mDrfr|vgNZB6e?l%E4zHGI?;%~4&3Gn0mdWyD-*5HWZU@;GEen$cz z#_|J>wO$`Y+di0Ae_oHcJei#|>C&*B?)gQy$|))m878x0l#t&6FDR)sZhjcD0+nVd z0&-%+qhAAzt-DU6Ba1laR(bAk1k^b=Dm-bsD`-|3u|IruWvj2ZS3D+O%7=Q;{9>_R zx1|>@Ir*wxjKUGpx+&U|;QrxuKwCP>z?EBr^F&wZz zz`K(dGHuS<-}uyP{rzRkqTho%k*+zY1lhogY|9Y%r4nlnNm6!Jb^PX>PhAe}=?(8( zcK)baqrN3!_2%u6B}E`caD_^x%8-v8`y<{0vA0A{a9m&Lvug0T?o`glE{gVCNGoxL zLoHNwnbR@=(Hrpe_;L7Mb)#q#%mb@SbzjGEiK1f>lzm;J6);zPK-zC?Y}~3Cye4kZ z7=L-OZG2_tRowK?P}-~Fxr%?W6Y9V=u&?1!`NTnE`EHuE&c8VW8 zqiTQoN`6A?Up&giG3wsE)QXF`1bN#o~^XIW9iDo_{Th-W>!;` zJrZvZ>a#(<>$;z(70VIBeMNQC&mPJGHhNWW~navc?X%W>?GoY^m$%5oL zRGhQ-`rntI$(j&7!Yar@i-8$`Zu70Xueg2xG++w0Sm4Jxb>*$A_3_^TjVsrIG^tm9 z<1FA9L`peL%HK{wza-0h@AIYJ-aH0N)`#d-#uQoW=iCA$yPWv<@837aR_M0G$R|!O zk9Z_n-81Ue#gErJ9o=mW&D4y_Xo_XF?-G(^81tQ;RYzoiH=xcJ^Mt&k7G;6hqyS&; zerCipYHd7Esr)8H?;*;#@A8zda7^BO`4%5Z=dkt8OWj17R2Px?^Qw~#0+H;pab~;g z1Pz_}ofIeSl~0W<^((OUk`WJ2f5M={g_RWmas6Ch4tSsZpt%_Uo~yM$*F3?SsIWu^ z#qWK)ATR$Rqpgp1Wp144be90HH`NTVObybG@0&ffb~$@Xpe!8?{9gD|^5@?_@%0B< zo=)S{?d}!~t*h$^!g@6i$Gt%hislq>ZeAIx&4l=B&>5^J1NZaqq)FX2`doSJ{#UnA zpO?jEUW@BCsoR^213E}r6jSw1jqBQ(cGGuiX;+5cRM7e9f#q$F#>?hST0C@#_uf6O z*q_SeQH#s#2AfYzz=eOm$;t=iBP*M)4j7i^VjU{xL+Kp76kSaD`sR9!eKXPK)3G3wyi8 zb>>&rZ{z3D<>i@k`}w`EbD`far2l(%JqFNR%-+ z%FH1-uf6JSH$7g?Lf>xRCEMn73w+Q_ua>;C^RhTKm7WidCtl&bal=PsY*~8VGq-4H%lpBD2QsaTqoob| zB>znNsXG%G`yu-}e^S}4+qV}rEPNu799P?Pn>*8ujBd(X@lP$vJ?I(<31{h~>9*)f z0#C)Cok#y|?!3Zy3Qc%D2p>EkfZ{aF%1dvd8eOd?^O4E^ zql*#C<7gvrR@J|QP^la9kUv-LmBjS&vZK-T z6Xupapx+g%hzA-59W)9qDo3SN+vIbFKzsp&ZNq&%zOcW}K~{7N6kI(|Mu5qj!NYT%PoLr@yhR%vrC=+k9fObgYJ7 z!4e2(itz{lQ4QpHwQo2`O81((6O4?q*sr=q_o&(lyBnYse=4{ z&}!u1o<2Ks`t(I8$41_*LW%XsN0yz-avn~jafVZEeq-QVED`Is77B&$C(rcdy9B!J ze1h`P3l1lg$_ta<`6NMn88mopQf=o_7+}qvUw-X~h7`Hj?PO>a=g>_2)un5oTFVkP z-ICg41QpIdm&SkoDLA_XAWO=F7?1sN#u6GhvnS&`ajF;zdf(`hv*@qAp9VDdDTfBo z-mFu{Dw|xb!&v7@IvI*Xaf+hfP3gH%4(xZS=LB_4!Y|y`OdOi-Dvpor7vS_779Sqo zxO>~+6=0I*50;%3=-HpbWp;jFJ8QGrOsRhu>62~5#CblBsrGf$^t?=$O1etDd*2539 z0Dv_4Y&sLwsw=%;&`_TIqj>ep)GAN=HkZPH;Z%XCPP@^Y=(1BepL3ZoJHrqQ;G(DE zTxeGM0uasO3n+F#N~Ks@4zhn+nId1C+mci-3hPmnc=Fwq0mb9bPb;7hC#$fRm?F>1 z36;tBcMQn{?u02|Ue){Q7RwN0rBhpVUSGknj|~-K0rBWRJZi@aoP`B}mCGwgdRn@- z@r$`$v4wo_UJp<3LC!7s58Xu1o0OGY?3A_$_i6hq% z5{8b!wS4-Ns~&DR{gaH*Og^6AIrkb&p&!3hQFn7Y^Nksi!5mnaaSyVNK7(Qri7kM` zA@*a}?4}kf&!E_)`~;`e<^YF#og5f&0SKK8COjRKD_oKV)L8GA5-Z`AY9jWTQ4z@S zG3AaUUa6x=a!27I*+tTXM$ZslFH0aYAj%v#xHf~b!!(5jGa~_Nr{0rL2<%Iz`6Flc zhj!{x*Y;+`fUK`1lIkLQv5;~(o>w0TmRc=<)}-6CW_T<^(N_;wV};kIY#3@aryJYLhwrts%HoGbD!=12zKq(K22D%|D_8j%TX z|5DR0@X+T|zfQIiDw3J@^g`3U5i{1$qdfGn<%WQ~yecsH`+^n!P^qk>L?E3;T}Wkw zu?0Q|cinsul6@BJy3cUXN%{O1*^Zl)EirPIK-i|R_pL>cDt&g*K)#E}6P7b5Omqy% zKI{!3C4o?fZO$h1f%#Gq9^~0DQKPimd0w zm>4_>Q}?9K2aMD^H3Y<7he)g{cx;zQy@a|L8DTxqg3>LzZy)l{b-DybvwJOdXpv0! zlznE3<(i`JvV6OnamB*~7J7kMIHm?P+Wxsg z`twhK(pm#jbJ_4CU-C{^c_xj+JeiXObJ-ZD9RYRW8WPlYRw%DbH4{E2@5a$3V>g%o|+LXU*-!&t{my470(x=E~ z57(WzeNakgh1m`x&{lY{-7BpmhT!b?#Q6Pw^>EF=#M9Yd)qohugm1i=zdktK-m7-w znL5ZqLly0G@5(xtwOGi-0_3O5UkP0XopP^RbsrihETK1Lz_v);6UAV8h(sJKujP`q zE#Z{&yL04(cf1*twa*};YvV3+EA&_cx~fiY`3` zBuAGSCDSGs*m)f0&YwRZU|qULh)8xzfQBU$T{s*TUlR?@2*x>tCC1$R(xEYs$mPf~ z!pqsjHr2+7XZ#8&uDEk=WWj6BCL|juMMzkGcSW28{eL2!f!g*aOQ;)ilI`+pAT{@# zxW00X>p2#d%QIhf2FZ^{YSN!P@iVdbaJJ)?!*CgMD72}#wh4$7%}22dfn`Jzz#g{i z91d@TLihJ`gL%Cgir6X!0ezKEwi&PSd9#UN z-l(3D$+$#RoC)DYGgSowMT9FrLgQ2@j$+;`N|^cX;4&JLZLwZUg2n-rYW`T?=RraI zmVJ7^bUmzEBe63M=H^!}T)1n3c4igNk4(HfvzvBTqmr9IpLpVwcW#>0I1B6NX@fb7 z{USy6mg;C%CeUqIKyWL1s=bIhUx>lgIV8<(NlbY-uMgV!DQX%OIPQLWz2!UJ@x16?im})b-j{)Q1ne9N+-$Sae1_0E_HNfitgICqa#qQLV z`i|aIH9z1P+cB}LfbGjKDc5ga|$$_jM!xJd<=Jq)tH5Fc+Q$!mq#a>5>>6vLH z6kk}~Cl3HYt1ge#p>hn%YoyRUI8d@rTsR?o_3B;5Ye!FVSNfjJpW7`fqIoX$Yc=mu zVeOAN9dcrO^l8G6q;x%xPaMk&|Hc~q_}WA%7`(@tQbCNn|~?L_uY#g+mhJR9^uEX&8i z^BLamm27y{jms;=wNX)EQ>WrS*C`Gj`}>rDN4eKp}sTxKC8f z>{4E5Iu;H5DB7^l{VQAz-m;_2&LgS@SiRQ$HT-kEw;kdp%9Vg*`nR3fE~J^vQiF$bjZCIJbt zn?i;_nIsj=(>1YWN)fEvtURZz9&)$6jLF{I$Z}oN`>I^50R#(j3@32l+gGk@m1XjM zc-;H*H1DR2MvuQe3|Qgim$5j^_oQY{pD!XNEXM(XpmmR_hhMD%6J)?Z6Fk~npc|x@ zZODZvX{PKEisTCm`K(vv%J=4i#62`7_d{~-6-ww;ZI|~q+BncvhR=za+zg8svEnRG z7^IXoD`zAnxLP;yu(BH2eJ?27CpPKizF#{(BwGXpvoEIzEssJtx;|h{ON>oE;MS!F zSNMT&>&&NKOd=Ske8sdYSJ|mpTjy-93|LO?k_7t8R0QaA65gZcb?&!o&54+4@u3{k zC=0*+%i1Z9iWrr%45q;StA8vNvFtsgnpHQ0BqUbbjag7Px=DJ?KuJosx$^cojiQ4I zaRY1sbo4x6jmYGO12fqqFcefHS_a0M`EGPF9m%_%c=wc9^^;ePq;}T&*u)eEz~BGZ zT@yGg!{<<46u|Yo4(x%qniB$YSRjXf`|ShAhbKIS2VH?y8Zxk_O}hRV?B_6`JvV)4 znOYz+qu__%L}B)|l{?qcfjFaFy9YPq$;*szQB!n{@X!_GLdP#c>rQ2g>tOUWOXgaC^w>kK~#of1vBbqq%XWa;PwpF$NFoR z@wy1JRRQ&BrEjO4}W^$dDx|c&zW^|WSQT+y>Re((?z{o7dI0RUgXOO<;#ga zrKf)ELRPPAZ)#Wlh$Fkm`4-AdeR*)0_{z9?3sUwI`o+H)Ya4NIIL&p z*f}K`C@p{;{FRtgzD{c9CmT^JSkThM?iS_1%6oGH7{wJ24_6;?h|5!5(1EP#W5H-9roFVuWv4!R`80j_D<0 z1cuZ4W&i?n3tYJa04JLX;3$V;_(XtXddwN)C2=bl7LX;(Kr2FPayG;HqbO|1w zgylG5iBYNqn|i`dPW*RSyv#-R)Bkn@&8+eZ)vNXe#AY~Tsvt@Q5~k=P+jb0Qqj~!* zst3Qjmt70y3K;PclO_}A;SP=5Q>dQHRhp+z9W!Xq8Q}N+!}6c5?GPOXNdE;ufd+B5 zM0OiEU0!?m5u9v60WixE7f(CPBa+xqG!*A=FBPBLc?zWgP+Pm%$PacswXU08RlOWj z1+emFGhA@=ah`u}!|6Z4e-Ji<&E#h)PNUwm9D@l8!ff_C%h@Y_^cqcy0h!{&-Bt1U z$kduP3)*jMTNAv^NdI1TjFSIcpMa`*fmt^f7^wHx_gfFnUNtIBnZiH{nG% zM+m< znD@{BvGDs1$n)dEoY&}DC#Ue(Ac1lPe}C5Sr=UN}XvX|c*|h&=w9)bJzi!F}PJ^HH z=SP1DYnzn+DYf(8>JGn7h>$mZj$ZJbiT>Wfr@zc!{SU$#;JTdBtN0}y!5nb?A6h7i zh`@gq5c!Ma@n5&=I;7~wp?`>fOfQld|MNY8FaLsN9On54bMKT6?w`b^{ybrSovNIN zzl3yzcAnn zX$0vfrYTdx)4;gK-5sUrEd!p`e#|T<)H&eU0A~ME24_*!UulMZj{X&g2Zk|cvaJqF zNK|#nfF0=Eeb0#7BEP;0XGB6zfR$wx;TF;*unn*sdLtC zs?B0|lluLDJ*h*2^q}v~Lh((vr=+1?MaJdi%zBi00(jUl*#CK| zJ?)Occ6pC<-Mdx#Iyxif0oBEY-X3%?dH45A@Lk@H&dJNYNY@JbQ(goB-0838&yPQU ze}U6i8&a9n+pTo5B6ONxaQEeYtfk;Hx2N=ryBj&lSa9VWcxd@0gWFmcWibuG=Lg@t<#{sQL@(dR;De7b=@-fV4<7y1UPsJkGjI# zCVn&0tCQCrySl;IJ3GG@!WS=xYd8Z%UKAfczXFoY zp>-m%nXsLXuWNYg?c-tA@10H}j-Di6M)WlLSbuJN#J6z$-AstVFyhI>|04u?GBChk zzg}2$xF{UNag#*DHI?zN@st~7>s{fQi3;y2&-<&LXyAOi1 zKZVBS_^*aW+?z$ac{~gE99lx#?ga($#6*94hr^;@qFyn(xbUovRlx!%+iJG9i8HJ< zDT9OXwW_>UZh1bt#YJ0$`D80r2av!gPQFpHn>cm!FxM#_9?P|@EoK_xKx}%t{zAwM zzSihq>6{c$;IQ>Im*M5r_a<&9%14fBG-yxQ>HR}(q=Anpsjc9l{SGT3Vfz}QD7C#L$h1e{=Glx@+?JbFa# z_sp{b+HEyNg&dV63)_tDob~~JQPJ)S3K&l(FGeY@#otL+F8)yIB)=WqL_J`bjSQU1 z$<6I-%?z;Z##DZ&TWD)wxLZ+Uqvp&?p#c=2$Hv+}kcLH+o$T&^<{VB;e}y z;YaKGUP?S)uM*Au8By5Ml#otjtb6}oi5LFQGgo>bpfk~aB~mH~Va#_R9p$docHvgd zh6hb*_VpzZ00%X!HTY2mq{Tm`K@S; zz`7eL=oLO(*-j7qk{|U;(i{|v<0lO{*$AB(@g+5d*m8)DdB>gLXtc(cnXmu+_P};j zQ(Px;0F%~(VSrv75Q)@eqaums>!TbS*1#62(y!YD;HRI%i#!`nd%_p`tZi)CQ(c;^@=7gN=e`?) z+pf4fIH0%a;-e0SXM6{vFGIVbvT*p`hEwI(SZl>{VyVI`bp_Ea0}4B-fvjw+9DVHM z?%2e_z2UJjBa<3pTCL;#yHu?mhC+82M4`JSVo~<~%dJZqS2m~n=Nf>VIQ0OTYXBHo zT`q5w*Ak-OWk3l?;v5Lvor~ZEQJcW(?)Iygk!$!KcwSFY=^l=HqtoHpDfbGXbM}u= z4}SUQ+G7q1uHf6?Y%wM7c;HcYUbi**ZIrSYlv@2!_th1!QT5yEEPk=Mn4cM*iQ^st zyT%gA6pT%BTTW6*L}Xl~!EJ!WF1tlH*MVPJwEND$vx;P6V~Oo#e0L=X9wEt8B0#}z z45xQhr=@YWR$?f7shcYbcUKBbOy@C}E40h|!Q*~LqJ_rg$@}k|n+C*RZR}+~jB(D_ zj8d*dw2HMdYLXkA$l8+SR(S69k6B;xg!D{I{3fZ2S7hLxpzMEEa-)azbT6z%9F&uT z{w)8R*#5!zz1abl1#fOLz)YqVbc_rVWpRGkV=CcEBGM0g2COoZa&oL&((o>im(o1YXLR^+?y`?T!2ObxV2WUj@)N!2Uv z&bS%guWjI7+#_^#fFt~XihF0r!61LiObyw`8Yjw097m6>XWxrSO0Vpp5d4 z3bct04uze&*Q0;vxp0>2Y-4jiiuGa{*2vZMr6aCcrDpZ})Dzs#Tm6y})-zwVD*bC~ zeOEU=JAhm&w&~CJf4|)cf4qSxSUY4Vd;h#&UHy1$=#DAqTCCE4{J6U~>hUzWdAcFt z471nW?Ue}u0O?42`jk?Mg)+Ue2AYC};0KmW`%hM_W;WzRyR#rbrXIHp`nW zE^3*0f}37HvG(vCf6#$Lp&eX4*u>gNU+AwNt6AL(&_InYE0l|4NS!RGdQ!j#_*IE9 z5Lc{WhqJXnYq^l0>GjeeIXRg-j!CG<)w;ZRtQrHI`Pgxnb?nWmB~mWPb8hN;tH=7_ z;z_}8iKmoe4t@M+n@Sq8b8N!-G6NgnUW^6#k=djvXypQE%YFY_Y(Dbeu{pfNWFsiq zdv8;`ay!(^_CEF7109XQkB=oOY8|@Ca%3+_N@sbq;PycrSSeL6e(M%VjehbS?0!+^ z{l&dOwtjHI*5y(VO<0Q|NJS`X6I#}tt*w)5-Iv<-Q{_cjI+dm z{g8{;Y7^}lkh0bE_9u7rM>-bvrD2}|h3wNx@Q-eiEHErPKq2lsx)8Vo@EJ??}XDEHshZ@{cuxU zB;kJ1kc+6OsZTuKGVQ`x&1&?!cdl|?OOKKvUN5>*C@<(D_kI@`qNN=+a*ALj2=1_O zsPOI^kjj|VA(gpV#NoBBV#N&?!oK*sJ@3j7`(fgrXSp!nzk}uDn!uU!=N}9CoYvvO zHDYlkSc}?8D|U6T73>bG&Chbrw!oPCZzrCx;ON{>@o99qEYRtLD~<04FkDsDuX0T_ z$1diStls!v@&$&TFL*h>v)%c z@s=VYagh$S_W!jtj4TKN|XaV>lO9UIp%+ zIDvqFR5MU#3S@x8gf_gvjWYZ9Me9Gx1SBXESOIPTc65VqVIM?|*tCG04a5O>L46kLF0l zUs_wS6+GA*`=F#d!+GWs0lBX}O#PX?jN@5&a^DrWC~boA2WYxH$+#gvszUtH8}W1dyjd;9JsqN0R5^7A6&h8szrCrx9`*N#2l}sMWHYWAY>)SwXN-QjK$c7fqP<~BltkuZK;mHr?<;2+0nei7{ zJyA+H-}L?(?AH4P)+Mg0D!Ms>+R;+ge zqPf|(rLQ2r6%x$r{{yqJ=H=5=zdd3xIe_(c!+%+8mCd&tq-z8GcZ&YU@Bf+){F)BT zh7c5YCMFcIi$XkgaiM^(Am$|R8K;pzy}Y=%7?S8Ufoiz1{RLeG7ngA0$k(Ci&4$0v zv3lk@s|p#-0PvWF!P@{;o*=+QA^7PpwyxkqN<86d0lH=iX5|%?CFKK2;sTe zmoR2OI<#};LiyGCY+?*G>^0rUg7!#tY+HnJxq7!>WR3=UT9`9o%daLR8!XN`_fA+m zLO2n}YQ-%f?7|ETuWXQ933)tUxUw5Jgz=t0ghQjC z$`42ovSf4lyd%1pg^-Ndb;S1-b+#_?44}R(9znia0INypv6^E zv7x3}K)_ofb&uczWQ+u$#*|&}FUO4=wvXy3h)7Bf@Z#Hi5vAwm(Qer#`^bcndv$s; z_gW4V>j5DzgM^PnGe}BpF(+Z5kCQ_2cP`q zyf&;Quz#vEq?9<|>e6dm2kT6kJQ};Dz4*SsS4?cUTA6W(N2AUnZBj-AAZ)glq}RJ^h3||#e||E44_!jsUCYm0FLNR^7Fg%DnnmHt zCK|dO+IvUySMX+&Jaqonf1mwhu)wR2)=^+7^&1?32mWeG=mYDWPr>f^@IFRXZ1cx>zcvga)Z{w&M%^RSbFk5dV|O4ywnIU?zi$_U6I!iy9Mk3I zkQv{%oQsm3?@9ox&SY;vp9xb_ZcfEpHM5NTE2A;?=^V_;bdCywd%>lY#MHWn=8%dJ z$Yi(SyNquReepsafCj z*rxVkHY;=L)ws5?KcY|h&Q=KI^h0R^796s2+mCvC+uGu(K}R&_?9rt>^00`jQjb;? z>#9sj_ho2f$L{sW1whp2F{I{Aj;r7L#B;SmdnBypOY|6x`OTSDm>+SiO@~F@%N6!n8dltT4SzrY{P>7%D+n(E$8nVUs(ZDnGiT^asv_k2{oZhMEQ>f#afx*0b!w`(+71n9v zkGm8=#Ra3W3!KdULB;JI^1y1f!u{uc&u3xcNgXN~^LnRH$Rv~r`{IJQ(S)@Smz?IM zj4mbuGJeTxDHTz4RpR~+l{}>;aC|;zPw}dn&}eaH+c=I~ee6nX^y_;+J}Rt^cE#c| zcpm-EYF>+uakcGSpX2G2>}|SWGRzl-#Y$GZn3w@$vD`p;wnV91=nRmG6Gy zK!grHV+!3h6KC28jwacqrfxU3V z-u=GN{u--{#`T?_r%x^xl*sOEuZJ@6B)W+=iGV9J;8>;JS(8Tt9luDF;`Vzhn#n{v zXVgB@4xLw%Ufnug!T{PP~x;1Nw z#R+91sJGv4i*}SN59Y#j)Ya5PuJ*E^{I*wwtulQL?T20B1%jI(+APR-Z7*$UbyS&M zELzqS`_a4-c*(Wa3;Anh$ZL&G;I!m@Rxf%jNypARjcP_UQ*vl;LUVO!&m=xC`BZ!D z6AL!IH_}NO{)DY8oKC9pr{&cL#l3zjlds8n+0*gdNg$2QrC0gjH33F}v-Q!lp_(KM zy%rA>z7`p1^W2z3qpW)GZdUd3^YPuHV{vs*?4q>y^OwQe zsI<D7xU?;r1yAI)3quw7;q>OHS_J3`xT~r>_(-1_TEBBVO8&Kyi0@tFV%6Iev=0Do698BG)LoF1dF%Q6c;^RV z(o=25^7epCl}sast#V>MlV(pW>wtL-bAo54(Re_9^@=BEG1c&uocHqU-Mu*@%XTM& zB6Z5;?mG#rMlmW&29Qw<^zISm&3DGm z6H!p+BG>rDMCad=0X4IeQM`HfRGeoZs3v2l;WyN3?P`fr7o*fbw&U#VkvS6zMQmFo>awSS%&%II(Sr1L9uNDUwJ2c9n%nD&XU(p)c zX&^gsF|c`_Wu)G#ILD+_Oh5qEKOh+~SoOFsRIm2IS$(CY74i}&`S28-42kX(+*kNz z`E#uJQypM=J-d{;D)Dd#^aI8Pfgm>*Y58w#eM$PNX=c`Fr&2TIoDk^k@2{y(kf#z1 zWKEn9Esi*(=;A&Y3-A62V(Q*g2~`qnZbs%EH_+3olxwX|o9F?sa3DbvnF|al!8sAB zH7vFJ%#+YxJ#CQZm9}hogg8=#O~dc)4kcDE89aFsY^)unW0KOGc}tC?%=CH*)RFG! z&tF{>1{&f)MS9fKp4t$oDtI(-&V6TZ$w(YRdAq?Buk_>CuoozI!b{UB!tR>9f%s;t7CN|49zjJ1U!lxbDS(qvU_;gbGh z2aU~msX7(0v<+|TCvN(_$qWt-t|wXo0WV|QpV6=gR|A&CHj5t66M;FHB<*~_>`qW5 zK_ysp;5#QDS@) z2+2s8Z>R_dH$fBbCmJ0nC?GL{*?sB8uw~f}&S}2BqFe#HJW3SpHhGNWXixKXVOhvt zAx*uF0@eQs;4WzCO1{9s#%4Dvv9x$i^&9WsHfQFXh<##hkC zw;iU}b$z(eKO;=?=eenc{?jNSd(Zu;<&i3z-Ss*5ou5-Tt&W6q6hDemP)PkgX0V8_ z>=f&yU0qW=$q8|_Ag86K3J(<<0cU^)Z7^90mxxqXO@17ys-iPi%@8 zM_ch(0foGg%S{u=Qu_g)57!GP-c*$AtSGeZqD#pKGcW&`ANZ~MVc&BmmgH@3!UPsR z_(@ag%?*8MG79qZ zV2o0@Sx)Fev%L$%j1eY4hvY>RqnDpO#Z7@^d85<$4&}|ZBR)9)@(WSS?a(pcr8Icld>H$C zXmBU>!~Iv#_$x3;$+>T9QVc!n@gK%-F+kjaU4UA(+Lfo6PTItjWtS99FCb7R-U~s~ zs~(CY3lnQ7N{U3svH8Lcz5Z+B_E0LfymxN$s%6XW7BrOb4RfyikWTu*G|I(p&sPjv zh6(NIYnI^V(ZL*1+@;Q}zQw)@Wb3(MjFn5mw$`N?(W(@YxlKnXn!b_#OVUQe*k}33 zjB{9ARt=l$&A_ibiHBOx@-IE9}=_1`YV*6=DQjQIub4R2mlRODm}qZh@;V82flK&ILf zBTeqX%Lmi~5R~Iwj5>_4ddmj+D2L6cqb3*eF4%DqUYHWG#xrk2e4gk<*S%;(rLR+p zh9rEKq;?19u$rOd24h*0|M6omH@7tmBP=}W+^MH&hfnJqhbj4OT}$wvZ{9VmV0vIQ zVR(DgYe57GOdglrg)s1<%}2XCaXO4?xPCM2+8V?9IOEJ6e{`Vn?(|(Z*}W9RfdhF( zv1KICKaLyTu1rdoBTMt1VK$g*TWO8ciIf^D1FpAxRvEt9E6wGJU9Y7ZqBm#3WQWkF zDf7fhL=UB*sgq|vKqpxUfSMQ^^OR}(nv$yw4z#08s;r}^D%Zi98H#V_nbVxh z+mO1msOr7A29Jm33O1vE=pWFqWb9#@*h7G-Lco$a^3UesVioj3~msuLEd6F(8neK4jl^lt)Z#eg0l&IhB;BkYkX#EwQ&)A@K-U zNK+!^*1x9Vk|ljI=uDIse<0vgH%jR`wIu{&Dd11)uqHqnx3tC&aS|=vDQE|N zS<8Kf3g=P(O}`J-K45pqpW#3r?F+jp0tFj7MXmqmO~jlWk6X*afbOoaRra+UmNum5F3s^82K`^fJQxZS3x$IAFBt z^z|x&Mj_dzpw>LFM|{TI%)d?`&GHqXvgf@12y^f^&bHe1E2WYxtL*+cLT0Tn8ehT+a~rd9ddFL2hKMmdfY z_;wBhKghFdYjT)MUE`2!j0GZFQQEz_F4?=eZ}_obcl*GjHWel&xAA?g*AU}o)7_@i zYK=0^QUjx2D?$xTh#m__%BtMdLYckvEG2~Jy;2Qkj#1cPrBqPMq`IbNOSU%i_a7aN z`(f&PN0Z7iyEScd8G&wTb!)P*!_5f2Qp2FeP!7U;T@HsgIA2z#7&niq42tVrI6AW* z5Fex#D;(|m@|vpcu^-|4Y^09hhhtWZFmpd(%DyOPUg?=^On1y4W?K5`!GZ#&+u=kH z$=AI4IKDcXQc+3{7I+v*>>BAfxRsP?VYk*)Rr*8k3W)l z#hx=ANZrLaZk{i&wh@q%GgjXD5!84MF&x1BB+Ndu!FyytnY=7n6$u~%kgw8-xt9D} zbp^YlJf*I)+K~Q=6dfp~+78QV84x48(raQd_9(koiT~A=z?*+z0nGh{9xuwgIKg3C z#DTUO`08gECOB9YJ6l;L=d&G}N<13mvb9>-u@@?J4{ zqGR`}6HT_nA{^n~Qww$0my&fQBqh7GN}B5&4hB8% z+SfMsKgWWao|#K#T@T4-N>bAn6yiL=V_fZe4is21*L!Y5`WNyqJvPV(@Q!vP%RnpS zvp0qqq{MS!dGj%#ueRO1c~n>i7g5?M&+JGiOeQ)0!`+YpEwi4Ah-*mAF=3dEihf zkS-N-oBrHdFYvRXISLvlxR{)LhJ}Uy_ok#xe)5rd}qbE6| zZ8{U%v6g|v^DZ{xg1BndX$TZy`v@YRV*>C`kIDM5H^w@H7U1Hs%?qMh(c2Kl70Ogdu_*y zkjF^USTn0Kj?xIJ`yYO@>&v=*ajKx8hJb!&H*i^UGVi^cFA|BY^xd$P{Mv7xcPIH$ zE!asFR_CqZiuIsQ-y z>NSS+BXLIg&!n6uH4+1=OS-TQc1ht;SSp^ul(A%cfyX&)422**F=etrY=oqM^(j~D zz=TRWSPvy!;oK#!#)0XJCfNM~#s8;Zf~~evY_0<|0j;Pz(<^6>J5CiikTNQfDt#?6 z%GOHC0l~)JZ?~@BPfGO$AVy|XMu2_1m!#`dNcQPrFY_jC#Nvd@(=jWcv(u_qY@xFF z4GnvyVIT`Hy~q=|UYKpd%p~{sgWF4etfxP~f8pxA>6tV96gVoo^7Nq+aAZGx#qV6y z-}>2>>?rwq>%+6FucDxiUwh0-`hhuM$XV2~cTrlFRk)N8T5;|3S-m=U8%g3y!)l_K zc|E|!pqciXG@z1?dg3ptyih~i4{^@K!e}MK+#x=z3jGA$LoF{3$_xH|DTPYm4E{3yUga1)in%12V%jN=?ub?k~es zgodOXhPy<_wHQiwz)nro%k`z=a`}1TQ8A?>LWH#L_a~WaJQZ}5w%{={GQ?{y!X})F zRb`~?ZW`zIbd8y(iA+AVv=??g4pBT87Hw?|5rCUTF5iVVJNYvq>Q2s)w;UK1x_dlhFni6lEMPBt}+ zyM8(m63olo!Y=D9LL5a)Qr?ksH^Pf=S+DB%PF33tbTz{x^gk|%2kwT~$)ttw`YS*w zn{kT!KR>DNRnCc(mcaA1xdsfG-;77j!0oijdE`Ei1$O2b&kT$5`p62a)(%jw(FIlL zBI~@*>3Ko#fr<@pf|^*oqo){{A4?X>M!PgoP>Y}93kPOBsVC|X8_uMv_dtd4)hi$~ zC~Vtoc(tReOK33AU!gN8Fs(LyrRmP$OZd5EA{Z>0)^9C&^t8k{mp@uPnqxb0WcX9Z zm5snJyHlU*GFO}jCqxMmmk0cQ_>t2aJG4p_CV@5CgIzLIf8p66^-IWQKB+q|l&+bH zVW@__55}L=wrT~Ay!A!&m=!+Q%GN-)w8V&&GhJ8VX0t6~*Gg9&xWk053G6f{X_Wd! z&d%s43-cWSt2)1Jnf|zqI(G}@uBAB2_01t8OIC;_$fqvN(i=yh>)>p%Pq|G!2A0z( zt0_NcI$#We7t_nXVa292pL}zS@gze`$B=yoSS)L3YomPc4;*!*6ft$?uR{<2a&-DD z_XEf=9C>tPDL;!RLYeR5$EnzuIMJB=sE(eQYk!m+0iPzDfI6A10ZWGBH5~7xEi*zc zv{qJ~2CFv>QO>Q!I9YEeV5Gi|htfvc_*OBr=77yUnCgW-U$g^OrX|{AGXsqvZ`)Fp zNaI#HXG`HZ1aq$;gOk3htyh!X9wjGNKJL5ZySJF13l5a(@Oi~2?@oupC-3Thq1%j4 zI1FR)thK|RzmF8NRCH|3=9l{SDQ73M;1I2+#Pjqhp-X{WN`sA(RV$YVBEfZ ztB1aXo|L}%>bRau^KM>pp3`85ybTxj?M$^NqO|2~BDAK*3)HMn6ixN?rZlhv8h5rO zwjokVW%4edX~}DeLY7kk=Mw%u#=bhN%Jth;1ZhP;Kwtq%NO!kViw^1TlI{i(kY02* zi%#hd>5%U3F6oB*;oke)^V{cn?mhm44=xw1_j_lJIp&zkA@-O+)h(YZ9h1uIhLp2r~0)QaJt!D<=%=lT|DuA2 zwQJdM!1lz8e(FyFg!Qc?D($^Zd8_)v!iD8;N^B?5*3l^`B$(tfj2`ih*bk{V+@UXyk%d-_-sA;W*$ z*m}n-y=4*PU@>JmJRP9;=@SC7#oj>8VuX63LIvc!p>y#ohu12?zqA|kk`>v-CE`48 z(L}{<3p!FHF^}$$;ZX~oj7Irly{oq*`E4Lq3RSOZJ_ZKp{V|gN00N_S$e&Px12IPz zyguotFTUN?-vm+IeCk#^-s-})iVov{sQt))mTc0AhvlxYI%1=B;WBFTvYt04Zn8vk z@2z(BNLh&EIYIF+{jix)H|JYV6*7_(ZJ=T&i3gnA&B_wh_%>7Z8-uVTd7C`Tg?b3` zN3~;71XF+<-tC-U1U*$08b!0MzRkYZsLN7msn>hrtR-a91b>B3sZBs zxVZEJA!vZ>hLw_X=h8<1x?3=AzGe!_Fd_M%L)(YZ|3a@Xmbk42}eUfs5`5*BhbT&j zqxe^35^=o%M+12wFYbbC*rNGz@wpu;FiL6;SM2L6)NM;%bX}-a${3T>c0&Ek^r;Q> zL^?29_&(a8?Iq9H?~;Vi=G+?&CX4}cEV2H{gL&Xe{GQL0a(iMVaK;UeAbj6F$4fkD zxf<>FyD>%qTg*B{(`mZ(0Lrh(?k6Jq8GKVIGz_Bpzs&BXP_ zX?rNHU(**_k5@Qp*Ujy^U8Xi|R84J(0<0D9(~GWo7p0aYG5tc1p{C!Ng66jKDk^h{ z%tumBGk+1)`xx!noc|WYgr1D0)m{l9A*05u>*Hda*p=@O1~OuyzEyY$Sdp+8SncI; zPp>`(0RcI}w|0>J?4#^ox zQzHj-DnHV{zRn^Jp#|Mvt%X_68K8iv65o)4X4=BJpJ2|)*~khFJun$%u}&6ojHY65 z-EO!z2UXEFn}Ucf0siJ;y#rx;&Yo#Df$~bv10E5PRe@msv++z}l!`l`B10ySQC<-R z^gzK8Tb~Psc9_O;W!N#|PI z>9=DAqOPGFHGTFv&PA9x0Gj*0AwY9V(g3o&!`K{<-T^g4*;7 z3d^Po&Cp7VrC!%?M4IxUG2_A5hND61qBg(;#3JMBlX8%sW?Zn7)d56Meaf9me z;AaC&I^vs-R5t-Q+(ZCl@vOre0JsxiuMpw5o_aZ2?x5~sJ*<zht%#JdD#qIh7~xs z%I++7Y+Scj-k(HP-5-}qGTiRZHpM06P6Q-Nq|>S1L@3J6CPRJ{n_5PqnKP*cK6F=(MCP|Ap0@Du%A*Cuj^u3q($@Hs`f} zOrE8+W&}V}ux3dfWY?lWbEnYH6IC?}hBwE*hrvK!O)DIyxOt}0oM5ka^88`(o1g9S zN9%-a=BJ?lU-{g#wh@;i%dzB+{+}gh|#s&+2V3Ln&;nS5EV6A%&|Sy zGBbz3WW!FlGfQA#;H#CM5>JD$gwtpMHMdY_mbu)RH*(`w%LZ@lO=V_cS(P;3d_h7+ zrW&h-0NB>Z$jGY|kvf3imaX05nd2JPaIsrAvT>%77qG^)*{>yCD7#||BwP-ntQR&3 zvHT)(h4?di2~9+9>e}wdei7fFfa-47L;x;KFzt_{cbX%*m$fMt*#PEi=f+P<^k$xl z@v;$^Q?3uPw`JBV0%JLIRxJzfKhJomwsknNg%IYueckirpZx*N*^i3A4543yl@v%9(h#kO(q7*aRPg|evAOTA}o(7T=O$V_c z&DTbL`ZK_^)*F`x|FE|)%8@mmBw)k$QnJG^%=Zb#Y0MVdWhZVm7E^s9H?ub6`ZJHp zs=1n5jl04z51v)?wo|;Dx3ucj9Orh>yCv3|gq@3aeK|CCD$qGT1H+>{g=3=G{bs`v z?Km4&(!NX@#qRyqfzwtJvh3lFk@%cab9aw~n#Uec2nlojOQ#RG&PcZci6dkiFnecxM9+6E(5+%>qd^P#|OxgYIqGMu3`O0TPqQY8HF9 zr~McS=(AzZpv4FSu!){;C?Gi24^))qa99h6d!;Q}AKh)RA^Ux-so0R&N>lmt$gsjz zX5BV8X0O(-FegC1mh{+=379)o*n-HyC4gkF1KGVp`ra2TLLwrf?lKB#4qL*HC2Ldm z<|%|>?2FFrjMWm<1T%X;YDOhmr?f>W&QVT%K0Sn|R@GBNwLyc)X(eHp`avo&exveB zga|@yVF+}>+rGJByWF*C9tnJLd01d<%9?q+*3DfYH#4%9Gv<1+Z`#$vVGIl&8EFBk zMP$I%t#Q~3z+|+_4aB_rl&QvM^dFPwEVRk171opGpdz@yBn%)Q|2HJmC#FC+nk3(O zo-=iQwHI1Tw?2xFYc`2ib?Yv@efJ=JPD+*)M%fgBwmtgoWpd_efRHNVO*0)0Jen zoEf;hxcS}olc#Ux_Ovg{MqG8q*3vy}d9er4EV=mU=oX9C4i3zCm+w3!?VA^jx6G7? z+73=XC%>~B<&wUz9qd-9NxDQrYI*b)dnCfcGtxvbD<#{y3%gS(hi!j#m&UjXU^a1YCY*M>4c^=XfX>6D&^;#Pp`EP zwC4fBZht0e)X~m?oHWX+S@S(mCB=x)s`az7E3(B@18vZ7$iIe&smR;pzNppjqansl zj~gKI#gI0Hbu$G#X$rDWh9agr;I4z$W#XG%~mQ_gRDWxaqyeR z)Z9_4l`4^4g9qmipRiu z)5OYf`sTac|j)n+}+Z6guUP>gLNnibPP3~MU6M)W5NeY!$NLSMiLIYiuf{Zk-! zw2lElobiR8l7X^wbGe0}I9kp>5kWWa$1nUn%oO7zeW3u1flUp{Uta|}5Edp58Ie>V zEHV&K2lf8a{-NV-U~HVUg<#RZoy`ALjsMvK!61>fX0rAAA>NN%Q}r~LP27r-AjXTU zAv-L{zP=HP8kGyst~S{kOc~9SBdM70z1x$jkq3ap{Lk4zS2s7YsvJ~e&p-EyU0#mM z%OvOnW2*Eu3y$h#81YIPqlT4{u@C$V3_nz6x-ETPO?q9-7^5I6y8g1~c74g6298Mz zGHVZIUo`QB_S3h_-?dmRHUGCRtncd~3AICCXn7z1&8!yS$FGOZYkHZ7__SB|7Ce6( zSbMR{f9jq0{ryAmKT`@hPIQO>DZ_dARZ0*}+8Sy)P~(0b&B(Ov=q16xbC7-2`LyFM zg$Ni_Qtrgkl%XE1Ey=}z^zi_FUmzOLC~YO&Gf1tgfzg2z$m3vACG4;3|a$3M9dsO>Zvw>~VU?k_%;Dq|m9(0! zL}^4%kEqL2tK>q1tSA!hnFnv4Zxw(9c5VLIHT9Hap$pKZTUWYB0GA{*Nbs@Y-qLiu zC>xltm^p6UPp+(Fak)*o{+z3;4@`$fxL;39^@#in=LZxG0F!?j@NARI_ah;w`2BEu zlynY{Mo2R=GXrs1AFsR5;-uZ4}(r$>?coXd{Lu(m9=II!q4LV^Dp4}R!c}&urZYEhJeo+V75%xWr|J@0MpOAyD44D2;9xwRT7W*5E zT9xP)tNHYB0OwkF7<){3WnWl@N>=}Pq4E$S?Nhfm{|9vJsT=5@j){k-mSt_f-V-6v zz4)LdjN6bIR(+YDo62M%s|pSaM3$dTtH6SMN5%b|-X;gwc|ytDgr^xR6aa1GJ@RRKSSD)Tz_y0DF@DJQR$E3&g*j)BDRM zdc^_-##;s4{yL&;bU>!_VD~j-->!7(noXx{zy!gtH!PFQ>KoJ1@~U}xCI^SS&3&nX3!;zlF^I40J!CCUh1nqJ;~2_A4738-T3OE46+p`jw4>Fw>c33#u@3l_#v zXB#b2tLnR3Dj{B{0kHz!x(D#P9%Bxb1dem1O&4l)CwMvO1dy z7u*9N{=l_?h#Q*D6!5wSKQ}_ZO)c=&~Zd$JuOZmoJrb zQ@?o~6YmVDNOEr;{|`pvUz2-JEWp(_5RGnJtcm8x&8^K%wXOaoh18Lwl(>o^Q#rZ! z20&I{4(NUg*aqw|aVfR&RRc-zBfW*&0&VG>oT(8&d(7>L4`(~dGO?LABZD$hT(h-i-gosiJe2=O~m1O~$YOQlHmEzyRv zO-HPrt2xpf#h$I@9~Rr13e3~(2WD^nsO|jwJ!G}u`2aWmvD5#S@Nf_K+t*vJr+Kd@ zZty3jlY!8v^BLW8o(Y*vGOLILfA+qjj+0K*;2D4R-a8WRjiiwKa$a&^x_c#WT>r(N z;neV8`ah+RIOsq0ntl@bY1!Q@Cz|uZb$4_n-Oh)HTO+-C=QLYCZAGE?|LQ7NEod%nxeMv| z4X)R3og&f?AH8o2h@TLfRx%2_4Pd!u1SUzp{AEvNyA~7gh=dVnQROfpvi8#^ zsYQ%Kw}63_Wg`^$g#i*>C>B7b0|a;e%uBYRQ`-Eek_*mF-1QAOpz;ir`86vAg@S+p z2{2FlS%BtSd+|~GxfxvYi}&aelcxf2fM4pGn5XYXmB`R@6T#nCVi;(32XIjA;Nsq$ z)IUdU8yFmnm=-*qba~0T#Xdeir@F;W(|61oWfvju{soHSC;yrsqSF{=;h$t@FAZf6`%t|^ryl5dG$Xvvgk-}((!Py^x^ULbQJD|=1IWv z8$HL8f!CZ;0@4KNkC;fSZp6~=H0}$_cB``9F2WvZ-S2(EFk57t78f&~&;6$S(IjYD zt@?wvm*G?vM|6j1^+VN&Z?jnI6mBZIzWVLn1lE(os>{If&`a*-FmasmO+v9ETHFsG`L+B z`+_V8PQ9D%_RSFi@TvN=J?{_LpR}(H3=Fv7Kg1XG18kVsSn_CXb03*M^~pOKjv%Z@ z%;OW)S*?V`ZAE0pJa;58+pJ@;{TUR28er@FrA(LQ{%6^tXc_x8-XG)nTT6$ogZ}rB z%GA_Iv;uH-nbW@F8nq;fy`d}Nc=8Bo)70ZhhxAj zPxS5jIw-ETR;W97H8KPb?m5@Ck{3ANCzym7P5g^mj=^S-5CPUq+&hf$EAHEj zU)zkRVh>V~l&jENt|#pjzFg^h8nIr=oRf&49;873fg@b6+m}zaXFU;+3D0v&;FLnN zbw!YYrU%?Xjr25ms;#^%W_FigNvV135oEC1j)`8IO!<0V&GiqiG5#kzXaASg1~v6g z{k44lFC9ATW-pi0DW%V&=t$EbudSOQ7wyhEPYCm@)kXNRE0ZYv(gk?9pGDc~YXSaK-^hGcaXB4z$8p7< zH*+WgGFD)+%Jvx0WqA$xA5VC2iT>~uo@HAHJ|^@PVMxrcWOFf&C`xWnrX0wt=g%g2 z+rX`>V~xP!FP!hu0cL8BQ-STKvS+>~A%da7);dSQ=bf* z`#IW4vV%A@LC;9d1UaJ=iv)C|gr;1K&}3g{Ku(YFku&DjO}MF(a*>Hqw<#aPo)HCH zZGUFbQWL#%?jDuQ%RNQg$hTRi<+k^o}YBs|<$fVi6x3u0tF>Jd^md zP@*4k(~R4|rPrce!d{cLX%OS?tw-u*9Ru^$uLNehqsCnKL6jN_`m=?-AaCw?udnZR{*~kgwIwD< z@kRIY74kE%!b84l^Rx>m9-~wFhFhFB2|Pr65KCt{pi{3*r6^7Y6;#G(-9a_gOAdSi zzD0vwSbvWK51CF?&b@DPyR>-_z>AC=TpNv{W-6E44t*oUXx`C^9{lv*;q8I_)*W3R z-m%gOI>xq`M`3F@0LrK_BgorP#Q5lM+T}vua{PJE{%{kTh>x}>1SX7lLEi>iN6cP< z7DWA+#!&g`2;nc!7;vk4Hr6dEf^R-0jN|puFuU)tAr2_druPPom{iWIJuH}9Q`{}q zkw@FS>|pGQck08`Ri=NJeDgM#giO&)RoD}bIi)^HHAl|(gC<4bE!oWU^t2YiGQ~a? zEQI3o*kC^+6ELj(lGE;KPE=ffMMnIS~WC03SZMv%vE4n>R`ImZ7k4 z9J(9VC$`V@yv?Ed>|-1N2Mql53cVQ06Czu?Q*@^fNY^ifZSkEfAG!7?i|0foxx%tK z>}owOVK8iQS@ixsU*PyzT1U?4^6E?wCrPh=+76jhV5$q)9D1{v zb+=7gkzscq_$y13UKpgj#)c`~kD9h&1j$1z1sW9fe_!13$#zZpGtIrRWOS=)m2(i} zC%4QI%cDH&bMe)pwOE;wC$)d$!H=JNXvAm_+~lr{dKmBwjG?-4t6bd<6%t6%c`>pT z9n7J(oVQ>1{O^x^eiGk~PIrHhx18VAvAkJqgiJSZ z@>~u6eR^0oIpbScY2hK`Rv>Z~3S)E9qWE2U0x zvATecEW29Ad-$gZ@hL$!}MPw zdJqj8oAd$A^*L+|3}^u3c5@xRPoHu@BY{qqT+{F?^qDGMpCD`B*lT;>{=D zAYFJ;Ps zZ46DTt>`dO7jo6H)y|Y2M3_d?7(P;WEv@&IAytVB9bLhXFdvSLdA<6q= zn#ic3KQ;-BVWUGHR!32cxQ%r>6MW*`^vGj7uHNNQh+SC4O!$k=QKAL;t}DL|lTi-& z%G6A$R*LW6R3{fu7f0)zz55Ij>+71)HEbzc&-Sxt!j93Vn<-?j{_R=~ZC5RAFoU@H z9&a~cnu^2s(F?E&U3^IEHzAi4vVIM@v^>|4j5Tq~Dc))(mqNKX)a*4ANps+jA}oaE zhuk~R7@EFlk89<^YQWIuD?g@5RHDIcf39GQ&I^b5Sl5rRXgdPn&WKP%2ByczhF39PAhElxt8P8UsNi1=&3 z`r#o3T*_bnzLi>hfX#Nbq8DnAr~PyFigqvP0bgEtIOvLJASx!-eSMnWQ)l;LrDl|) z{vW@nsOEIED6BWinU!md-15z19EY7zacyp6bECfpPB^ZVgl@|iV)@ToD(-rbrqNOE z@K`Q2w*wIY6+}S3+>?zEO?lv3<*m2$wl6O%eR}*W(~Z}5nK8sS5b-9Tg8z9~Z0FVT zy^BZ2FP5&N9~KyI3M6O6VZiUT@>mkGfQimGxZLNy7!>F@3}0oM^pKzOvndlyFL=9cW|{D|An()W9iVX5Jx2vw#X zZy+QK4Ky%-)>g}b7BPtERm0esWIu?WqPxONR3iL7zW2%H$gC7H{}rSs`^kY24Lp#T zz4fTSV{EYE!ov$7o=q&MdL1o#kbX#QpNVG4;;r>vvvrVzhy4Ar!blGpIQrbku*455 zBq#KUp1Rt5!jh8rxH~w^=tOD0`~8>baXJ?3mPgJDZWiPn)6dJ&aD<}7$0;>5gyy@O=9zQr)dg{3t^ z&MF5<)n%%pvp&0LK%wAo48AwbB~2G>eFGz0Fj6*xW;@(&`20a<5yK)ktgUR1GN3Vv z@-$N)a6F>a)F&9464r9%JRg8qmx3sR_pzU!p74{OSh;i_DjesF_3n1|Tb4W9H3PhF zF^JMsNN93+1w*28GGO{N!H=mQ2t44m#bY-@V$TQe3!dx?3kD2#Qj7gPy&){gH{^{O z$+%sGNG>g8yWWmXt>{NLf4BNF@9m-NcgOJGud@??g=(CJHyIb-wvQbeBNI%KnrT4@b|Oa`k&8ouJ@k#aIwK3D7#2=bBK>Wc_#SmKBX*AQQA-x zSg;5Xj(+RmWlCEcdc7e^#IgG- zdvJ~;9z8T3fiwFf1y?LFs??>Hzf2L=C&p`}_i;m}-cVRvx$R9Z^i8^p-$SA7ug@tl zSEfGeG#O@a7(K@Z&5*6U-vyJ&ns~Xs`2smWNs_Ys-OuX-ivwjaokFfm<$CV5JBM_> zkD#{-+2D8wkuOy(=Eq{oY^?5pZ#AmG$CG&rm#vI{$D+&dz0~37rpe=KfMfzH0T7PL z6rX|6Sy-t1wME`CEgghZC*|s^&3^jT@P_IW>DD)$N-ch25t^GV6+3q}HwGgOo}y4& zmi6fmNC@{MlW>oA`{XdB;^6S`g@gs?r3~_ip~*Me92ardgOehhRFH6?-{C!GzelQM zZ`itx$8It&?&L5&e>RtVdW@o0^DPB-FaEZcy$#I2GfId5kx_m@fcTc;IdP!}wrPhH zw;8}ZrCd+B7O^Q;ot**s^eFjs|EH=MO$2&|+CHdn*c~feiPO<*|&-Y0nQTu}Pz!`DtRjb#hd1 z**;w#TTl$N2v;!P>+m?d*+o5~>n|9;#Q`-jBkEYwks2|Q`o~kFqh?q=$&m|Sz1f-lfHk?1YNK@u1?`xtYrK&m11YmLELAo z#Q4r&bq6_{;|sblIMYO=!X7AlmiT_6cv)|Ivs2cO6056R)A{R7pju5Q0MV>9VzJgs zaSNH@|L;DBP&k5Q>b~~iFt~U>1xC;OrQxZ3+AJ;lnnx9yN&5G+*3rLOM@o@6g8Frz zG8gTxu+ekwtm7f#+bJ6Cz3j=2epC>nwSAcy4(aH6j5p@$QRm2_tU14*Sv>bD5ss*H z{_IyM(c;ga!Eyo3X|$IqjR|x=lQB(D^!Xz!HXjH@gs za`)j1-iu%!R2Z zWjNc6K33xFnfK_jKC0@juWSTPN?b`=rbsbgHCi7I^6tyBulE<(nLN6`rNcVDzx4{4 zS^tiAA4)b{$uy2Xf{|C><53GZOH@z{3J43#MfKfp`?0Rhdld0v$ISrifj0x;!N9`z z(&&sHVx1KzIj*-rin|c-75YA7kCj2LY-H(G;3AXSh`PSyhl^7EF@chg5B7WOtgop?k z2G*yAUc+!_jNd-1Dt6Ny==HJX*5ntawLQ_G_Rb7E>^B%E4?+~)3dP(&YjT9C@L+2* zFJQ$cnxWOr7IpEe4U73>TejZNWIb65Ff0LQ4pG;-9p77=EpIit_HA6u?`#k$ok39&*Uz9I)qfwAOdfK+))c zaLsN-a@0kuTaz|w8FR%_DRcR#4*%e()L@tnx^cmoZ}jP}Ib7XMajPY39)^3bh~xgx zdlv0t|8c2|)(R;tWTjDr0c=(`ou~#zhGqy(S`KpRXL=$Zf;x;H__%l9EfxNj_PC2z z-Y_mJka)7fj@FTWV3$eeCHH}Zlmz(uZ|c04Uqr`Kd-?_=n!zGQv5r6Xgqw^f%=CDd zmhPj+SqzXK)$R{;%T!9AGN9!E9q2GVB~X&q%w|C4YnTn1;Z}L58KBR>!Gevl`l9R< zQhe=Q0fN(M*pw^8rFPw{X}v2dpmp!ZS1jL0%Rlh#Tfvp0pqG0|ALP0zE%yX}g_gc? zAFe@3Mza#+LGC=pF@Pg|o|Ko&M@gO$&)%-iJExmK!R##y$8){asozmhhry9p=|Kru z7zW8W2+1WXAkyrs1@%y8s~sT$*zH{5E6_k6@>e_0*WdjtgaH3IY#$2y2LG7v?F82O zW~jhr6UuV*3ON*C?u_`R@eK#URT=*2!6PKp+P{V3Y=`{lfm_U*40JFh>FZ6 zOXJNo-Wt;WC?XP@oF!o4GgT(_@qlT==lizah_*C~D?y3Zb)wW&{U>mdMvb3#I^b&S z0PpvdsJimCM(mc)H#(r;eEy71fEjEyRq~j*baZwWEE~N9g^F1{Y4cmuGYkQQr84~~ z!b-+NTJl~ht8oz_1BLuGmD1S3RBrksrB*f??!DjmgNbCe_<~7U_Ks9VLh^#A!R>!Ww=whN0$0|t&>1;0ml`=8#`r~z+<8_h zxZUg(fz3hO6PQUXH}hm6dH>FS=xMF(FIPz3EKCY9xe~xYPb6Q2Sq9NxRS0qaUoQld+tN7vDoFF^6s! zrdDT_i|3H#`#n`Ebls{pTn=s=I><@;5aH?V2tLmswQ)Ng@&L%(^iQ3#f=_?j|0xV) z+zk)p2wGyB@su;BM_yCO8K4#T@%M@1Ww%6kewpdCWmQ(b4YMHAnL+MB7^{5=bOh`}Eq6aD|^H z19-5Yml{6+Tvhw&hgYuu@}#(sI{sOn0U)(E3axS-v@cEMXA4S`>(L>61UjhMX8QIC zT*wQ?VS+UYjz8jAu?ccRgo?d$u8New@tk4Zho{|St*2Qb6NsRDrB7SEBtN#=kX-J1 zFG3vLTZKvIOEBzZwFue z#9+ho=+j)EBRmQ@_&zw7ysYttm;`qK{nqW8&S$g*Yw-s&xYQ@O?6;9G4_dbVQwmiO zV(}lDnU?tS_GZ&9MRYoPy9kyrZ#`A=y?z;Z-wr8CNCgAKVnHZRZ(iZB${6-+zZt)@g2>eFS&N!j*W| zwtGHD)tDR8w75nt-6uxRyhRtDqL;W7fxKHDH1u<9(Wf8_>OCR1mK$hx{VZ6yWXiEi z@k-+^DAg2uZ|cLJ(t3IO`WbjP3)qlV165jsY%}X{6v59d_e~UTT%6*awA(L5)nA)^ zeg6VmsU5W$b0@yt(;M+6cg~BH@q6JoI{o$d-fe?flm?#0mpAzn$9%N?ON`NvReS6@ z1|u?BV}8+ogcjU{Y+UbSZ9t_SuPE;nF342rBc2Srhg!b!(O7+X&BcR}zE7SUL)P#UG?=Lk zO?V@)F0PpMhJRPuh>;;m&Z2cB@-_YV+tHWeQA0l%`_K@W^p{icV2+{@_r*B9*)aV- z?u;UC4z3w?ysJu<;bZ<$5!osy@Bcmq=t%%OnoRQQL`zxRo`VQlB9z<-Tz(G(`iw1} zzpZn}-VHr`nZ09-ca``3o6#y>EWDfe$3S9YMX)ag8yj1-G9kxjD7IV{w1^}=(e~Lz z;Nn|Q`77Efj^z|aBiS34Zl_bQuG;RWU%fA*cMzXv=7FKomOeX>8IuMhjKtF9XBC=u zj2z}ZJ^e&|zah?JU2x~vGvawBi9NUKaVySKlW&f=&!XQ&9rPaj;46DU5@sQAJ1|DJ zeg@`du3X}HSsgLQNH-{m#sxSy0LrYS$K{}EINF)HwW_7Ufy^`kc377~W)?O=5M54b#sfUp=MA0GS|A?&fdo4oTO z_m!jtEURYkIu?usy*M0*)no?i9$pqVo>n%0GQ7Z@m5J&j$Qk@tEOHQjN*!dNBBvRPL$E!y>7;;#T_~M)Y`V8&+JDS* zYUCApZU!WIPgSiodcx{FLEk)~Ntq+fAkVKj9Z5qSXb(3oMk}W+C+@9Xru92^C6&s% z`6LfyaA~see0$@Uz)>g)XoB!VpZXr{m0?=!J2V&>IQC9{y4K8P-m8K%^yt|my&XB3 zmqPaBf@mDJGN?aqd1ob-O|)+)eOl$VW9&>!_+MM8R!+h|Mhg)^FKJ(ZMi{9Oh3CG} zr*zg+Xw#rh`-#oPW|nF4QxNP0pFF+{hIPV`DIdz4G?9C9rloJJqdkOBUIEK@R6CbP zlIKY&;URG7d`H1YUzmY;{CpkDSoV56~N4K!TV(fjVO|YHLGQ(L{YrkWgum$Yn zj^)_|MkidiK71vo<_0&6^6i|&Q1|rwYU>P<)BqN2HZ(ONeH)VBfC1NgX*W4)Mswqv z?%>8yH6+Gf$Sq`Pk7q#xrRp6blH#lnb_JY8I zE0NUwe78zG8(Pgi0kgneAM18`h%YI4E!QlcqG14+w_(QCb1Ufp-eMzlt=@U~{NQr^ zf+xINFJ`6}IJ`f{NN%0AD6>TJNtQ1^Op{P+y4C@QE8(K6ToJ#|-O6-nVRa$rhAE7Y zH}*(8Fy6(X7tcc!>qFR zdCe)iW7~w_%hbbeESogc>%22=$y=-WM$;dbI}f7|Ni~+d=T#%TIkwoc3w*9wPD^2KIh`^kHvY9^6O$D^^ z&;|xjq8VO)5_uCNT7`>)rk^^!=z4EJ!Pa#Z6^c}UnCO#`@X+Z3*K-yX zmv(pjikJNDDoOXlcA1kmTwpZB=&iFN&{Q%@87 z9;k&cK0gPeqAM;STP}9JO!02&!s#Ut8uq4KkQEPL|-72%+vrf;}Kt}0NrrMS4dp<9M<471O zRbSohAD{gEx}pzo6vXp+Qiac&LOEQJrR{K9v$kGWGFE@Pe?#sYd~3fqX;Gxsf<@5( z46>~nz#a8fagVLdb>D!+a_OZHIJCPQYwtCNtEBYDNnSRMdH%xpY;P@@(Ji=~1fAsA z9We<;ijNqaU5~~Z=(zYOuW735j@sgp#zqPx=+E#BX!km%9s{cBBo<*4Nt46leP1S3 zi(4PwriF}nO~0klY{rX_W6XvR9k*x6xbV&MPZ+dhC6t89A|*lnK%OS(C{5VK&9PbmPeAG`Kbj`@$VYaU zxKpvId;Z;Gl>Zb3v4e%&m}PHR(-PX=GGG_1yUNVG7%JN^ljZm<9KI!X+h($MTW9{`g`9D=)p>(PQLHQ$`-WM8A(1j!b5H+ ziN_3wsl{zF_rB}k*p+ncuh`?BMRo+c4e@vIC!e+kwTYKb0febt_r@Gp(t>Ziig~^p ztb=|M+>^_QwXvf!E_efHKll^cweyLvMPV-G!!uUQ=I~OlcXZm#LG8_828$e&=%tCD z&G-K2B{3f^iB8VDX;#MKFYsvG`)Jg0{XqfaiUN8}Ms+kD7%L46jV=`mO$T#c zqE~**E6r}D$+Al28unh~(n++ZF0*DNCi#y6Hf?G1SuY9{mdpz&PHV-a144u2hIYyo z_lCihtdcYcKRYcFejj9RFpEAc90x?j)0Sjs=aEydFxS(l(A>=gd>&4`aep0uSml&f z)cq|72uY)}7H<~uWp%b{VNqp=^4Kv z4AxKtheBHOn0+gC)e7^@(TJVml@tTBc~S_ELPDOH;&aGuV<=Eex-z`_DI#}SUdDF9 z&`hB7Jlny7<4}{nG$Ne@2@?sWC6@T|u}+=(7utO2_U%kSZN!J6jOT)UrV+ASlkMER zvp*dx=9po@Hkq$XykvDT97#c)7q#!;aLD@g^La)~lEKI&`l`PPDbD#@YbJSD$WC(i zN+1+}rB%e%LL-41C}DbfkvQM|tU5x1n=yDY1BC5Pzb${Zeq+~R+)k#S9>Ug^RurmJ zW|aZwwfUx994JEL1>{1J1r6|!=mT5GKe(+OVY-Qm?zB@XS-*ai$9o)8bQ}jl*pnU&IqM0s8Ae1TQw=`+xgAVQ9D>g|@jHM$lmri3;=MI1tx@AhO z?AU!I6fE3}8t)ATtYrNnrf^FCev@GJ=!n6wjfF<75kpV}(&@j%>UVNfkN677O}8a& z1CX=u^Cj><(eGMW>ri=HR^GTr5-U$0;^#)RcCq zu(D(_7be{)HdIXA?rwTiNV+FgR%U*T{gY%Z9>KJr*XMnJnYBf-ow zrBG`hT+otTEb^w`t!}RsgV?qe7%r`>Y8+)e{vWEof-A~!3zw1*>5`TP=|&pq?rxCo zmhOfDBt)7arMpW~O1itdn<4I}=bUx#{R6Y6zW3ex$#GU^9IurY-FoxH49AtcH5P3l z06;K$l*2Qw;34pm^JZ8%1Ge)>pANAr5H5K6XgkC~TXAS_f5CwVGn(dr?S9PUAM&m1 z1}k~kN=0_SAhD<*jGSkHSxh#;7rA;@L|JFd12Y1>232=DE&mvHv;(wOq+S!nNKaWH z8~%~(Ch-fD+~IR7zWM6Pb`Wac4*yHJZH-1gJyw3n-TEpYX({S5A%-9+A%U^c%-kG_ zv>RI0VniHai-hIu6Z#g-D7uw}bn53fpnYBwg@N46--Zv~cJBxT;VZ82jN5wNO;O*J9p&{t_; z+~gvCP1zVM!QXNligKyP*U>PN);|!t3%%>VEPUKZv`g_v!*(0@|7{8PCfqu%h*o6# zZrrUN-EV+l?;(}bLU1t(qTsgXpIRQ$;EIQwy3V413<*DXiLP@i%KHA)&?4l9EvT({ z06BcPxjX&U__R9KzUnQSouD^Smp?7pV|t#%Wm%=_h&#%rVEkVSe}t@O-}|Jk zaiuAV$7g__Ffv^f@8(~u49%Wy%iHfWKzES^b@z;yOR_hOleMB4COA9H`h&u`_bw}XhxpP(C^PO0@uc>dk3x_ zBcEDKm6N#tuEzrc4(kyyU4_No+WQ-a0i7NWuzgQ7L?ba1;RUDWo9PJx=ar{+6ElOj zex4T*kY?ezzry8hA9e3}55_NF(|7yBjx z{9)pCi<>#*GBikJ*kTA#9!}XQgaSZZ0Kf)T@?#r2;|aP>c_y5hBVnp%lacE-mz>mf zSkHoQRqSeD76R_p8kW~O=~j!#!5XG}bt~sY!j~`kAGETU!?=!$jHU<69i?sqdpt?! z#hUXL0HH)_EK*W}{*Y>F zdggz7T7&#c3#?>GHW99$Uc;b8Gt;DmYK3)X$vGzDwxlmie|uQUp(H@)0LoTgVDANH z+scHUwYFFrkJE$lI;|gpjhYxdY<1Su(S46Q`F>}cTUd@yy-3Gv{!o%2IC5TU2|h;+ z=WoHes3?DVLciDl=K#kM=TWJ*PvlFh=MG*8vCqSgQa^`#*YM_`#D%#OD zQ#S~qNP~kYuQXI03cW)>eb{&kRr?fp38wWna-?m!!;eZK9@&Yy5;~g3M?tk)+n}T2 zS8&Rp*OV(l0=)?puc@^RE~B- z5@J1INU6|^Gq;P~={-E+HOue(d8}&D0?w0WW-zRk&Wf7^QadHR^;Eu@2c3UO$RZEM z$dSAeo~v6LmOQN`7E(vna9m2gY56ZtRAR6$)&3ohl2!)AcIhCHurI+D^#K8#r|ob7 zO4mIU$oHFsOt;N_xIPtv5pt_ZTIYU1)?WCuXvwfWJPvVaoD;0uw?s{MQr+fDpyn!t zLO%ciko&v4QOdT&##r8nx=xq7SOiz0sS*YsJB#A!&_g zixjJQKf1YTKf5b1!GM5o46|L;%47h`P?G(@s