From 5b1dc0fa93dbbfa36cfc8b646aa8278413de9f99 Mon Sep 17 00:00:00 2001 From: Simeng Liu Date: Mon, 16 Sep 2024 10:58:02 -0700 Subject: [PATCH] Remove deprecated files from quickstart guide. Signed-off-by: Simeng Liu --- ... the Tensorflow TensorRT Integration.ipynb | 664 ------- .../3. Using Tensorflow 2 through ONNX.ipynb | 1275 ------------ .../4. Using PyTorch through ONNX.ipynb | 992 ---------- .../5. Understanding TensorRT Runtimes.ipynb | 107 - .../1. TF-TRT Classification.ipynb | 711 ------- .../2. TF-TRT Detection.ipynb | 585 ------ .../3. TF-TRT Segmentation.ipynb | 480 ----- .../Additional Examples/helper.py | 111 -- quickstart/IntroNotebooks/images/tf_onnx.png | Bin 357902 -> 0 bytes quickstart/IntroNotebooks/images/tf_trt.png | Bin 364204 -> 0 bytes .../qat-ptq-workflow.ipynb | 1732 ----------------- 11 files changed, 6657 deletions(-) delete mode 100644 quickstart/IntroNotebooks/2. Using the Tensorflow TensorRT Integration.ipynb delete mode 100644 quickstart/IntroNotebooks/3. Using Tensorflow 2 through ONNX.ipynb delete mode 100644 quickstart/IntroNotebooks/4. Using PyTorch through ONNX.ipynb delete mode 100644 quickstart/IntroNotebooks/5. Understanding TensorRT Runtimes.ipynb delete mode 100644 quickstart/IntroNotebooks/Additional Examples/1. TF-TRT Classification.ipynb delete mode 100644 quickstart/IntroNotebooks/Additional Examples/2. TF-TRT Detection.ipynb delete mode 100644 quickstart/IntroNotebooks/Additional Examples/3. TF-TRT Segmentation.ipynb delete mode 100644 quickstart/IntroNotebooks/Additional Examples/helper.py delete mode 100644 quickstart/IntroNotebooks/images/tf_onnx.png delete mode 100644 quickstart/IntroNotebooks/images/tf_trt.png delete mode 100644 quickstart/quantization_tutorial/qat-ptq-workflow.ipynb diff --git a/quickstart/IntroNotebooks/2. Using the Tensorflow TensorRT Integration.ipynb b/quickstart/IntroNotebooks/2. Using the Tensorflow TensorRT Integration.ipynb deleted file mode 100644 index 7eda93b6..00000000 --- a/quickstart/IntroNotebooks/2. Using the Tensorflow TensorRT Integration.ipynb +++ /dev/null @@ -1,664 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Using TF-TRT With Tensorflow 2:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The Tensorflow/TensorRT integration (TF-TRT) is a high level Python interface for TensorRT that works directly with Tensorflow models. In Tensorflow 2, TF-TRT allows you to convert Tensorflow SavedModels to TensorRT optimized models and run them within Python. This is a simple and flexible way to get started with TensorRT when using Tensorflow.\n", - "\n", - "This notebook provides a basic introduction and wrapper that makes it easy to work with basic Keras/TF2 models. We will take a pretrained Resnet-50 model from the keras.applications model zoo, convert it using TF-TRT, and run it in the TF-TRT Python runtime!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Use this when:\n", - "- You want the API with the least dependencies\n", - "- You are willing to give up some optimizations in exchange for more flexibility\n", - "- You have a network which contains operations unsupported by the ONNX parser but still want to use an automatic parser\n", - "- You do not want to write custom C++ plugins/optimizations if your network has unsupported operations\n", - "- You are okay with being limited to the Tensorflow or TRITON runtimes in most cases" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For the TF-TRT portion of this guide, we will be using a wrapper included with the notebooks in the [TensorRT OSS examples](https://github.com/NVIDIA/TensorRT).\n", - "\n", - "You can clone the entire repository and work inside it, or you can grab just the wrapper by:" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "--2021-01-29 23:37:25-- https://raw.githubusercontent.com/NVIDIA/TensorRT/main/quickstart/IntroNotebooks/helper.py\n", - "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.40.133\n", - "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.40.133|:443... connected.\n", - "HTTP request sent, awaiting response... 404 Not Found\n", - "2021-01-29 23:37:25 ERROR 404: Not Found.\n", - "\n" - ] - } - ], - "source": [ - "!wget \"https://raw.githubusercontent.com/NVIDIA/TensorRT/main/quickstart/IntroNotebooks/helper.py\"" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Checking your GPU status:__\n", - "\n", - "Lets see what GPU hardware we are working with. Our hardware can matter a lot because different cards have different performance profiles and precisions they tend to operate best in. For example, a V100 is relatively strong as FP16 processing vs a T4, which tends to operate best in the INT8 mode." - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Fri Jan 29 23:37:26 2021 \n", - "+-----------------------------------------------------------------------------+\n", - "| NVIDIA-SMI 450.80.02 Driver Version: 450.80.02 CUDA Version: 11.1 |\n", - "|-------------------------------+----------------------+----------------------+\n", - "| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n", - "| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n", - "| | | MIG M. |\n", - "|===============================+======================+======================|\n", - "| 0 Tesla V100-DGXS... On | 00000000:07:00.0 Off | 0 |\n", - "| N/A 42C P0 37W / 300W | 125MiB / 16155MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 1 Tesla V100-DGXS... On | 00000000:08:00.0 Off | 0 |\n", - "| N/A 42C P0 38W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 2 Tesla V100-DGXS... On | 00000000:0E:00.0 Off | 0 |\n", - "| N/A 41C P0 38W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 3 Tesla V100-DGXS... On | 00000000:0F:00.0 Off | 0 |\n", - "| N/A 42C P0 37W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - " \n", - "+-----------------------------------------------------------------------------+\n", - "| Processes: |\n", - "| GPU GI CI PID Type Process name GPU Memory |\n", - "| ID ID Usage |\n", - "|=============================================================================|\n", - "+-----------------------------------------------------------------------------+\n" - ] - } - ], - "source": [ - "!nvidia-smi" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Basic usage: Optimizing a TF2/Keras model with TensorRT in FP32:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Remember to sucessfully deploy a TensorRT model, you have to answer __five important questions__:\n", - "\n", - "1. __What format should I save my model in?__\n", - "2. __What batch size(s) am I running inference at?__\n", - "3. __What precision am I running inference at?__\n", - "4. __What TensorRT path am I using to convert my model?__\n", - "5. __What runtime am I targeting?__\n", - "\n", - "We will be following this path to convert and deploy our model:\n", - "\n", - "![TF-TRT](./images/tf_trt.png)\n", - "\n", - "Lets address these five questions here!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 1. What format should I save my model in?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For TF-TRT, we need our models to be in [SavedModel format](https://www.tensorflow.org/guide/saved_model). We can load up, for example, a Keras model and save it appropriately as follows:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "!mkdir -p tmp_savedmodels" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "from tensorflow.keras.applications import ResNet50\n", - "\n", - "model_dir = 'tmp_savedmodels/resnet50_saved_model'\n", - "model = ResNet50(include_top=True, weights='imagenet')" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Model.state_updates (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Layer.updates (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/resnet50_saved_model/assets\n" - ] - } - ], - "source": [ - "model.save(model_dir) " - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 2. What batch size(s) am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Here we generate a dummy batch of data to pass into the network just to get an understanding of its performance. This is normally where you would supply a numpy batch of images." - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "BATCH_SIZE = 32\n", - "\n", - "dummy_input_batch = np.zeros((BATCH_SIZE, 224, 224, 3))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 3. What precision am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will start with FP32 precision as a baseline! Later in this notebook, we will go through and look at how we can reduce our precision from the default." - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [], - "source": [ - "PRECISION = \"FP32\" # Options are \"FP32\", \"FP16\", or \"INT8\"" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 4. What TensorRT path am I using to convert my model?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will be using a simplified wrapper (ModelOptimizer) around TF-TRT to handle our conversions for this notebook. The wrapper is bare bones, meant as a springboard for further develoment - not a finished product. It can help us easily and quickly convert a TF-TRT model to a number of precisions." - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [], - "source": [ - "from helper import ModelOptimizer # using the helper from \n", - "\n", - "model_dir = 'tmp_savedmodels/resnet50_saved_model'\n", - "\n", - "opt_model = ModelOptimizer(model_dir)" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/resnet50_saved_model_FP32/assets\n" - ] - } - ], - "source": [ - "model_fp32 = opt_model.convert(model_dir+'_FP32', precision=PRECISION)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 5. What TensorRT runtime am I targeting?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "TF-TRT essentially yields a Tensorflow graph with some optimized TensorRT operations included in it. We can run this graph with .predict() like we would any other Tensorflow model." - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "array([[1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " ...,\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04]], dtype=float32)" - ] - }, - "execution_count": 10, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "model_fp32.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We now have a finished TF-TRT optimized Tensorflow graph!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__We can now compare the TensorRT optimized model with the original:__" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "array([[1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " ...,\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04]], dtype=float32)" - ] - }, - "execution_count": 11, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "# Warm up - the first batch through a model generally takes longer\n", - "model.predict(dummy_input_batch)\n", - "model_fp32.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "53.5 ms ± 423 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "model.predict_on_batch(dummy_input_batch)" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "29.5 ms ± 117 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "model_fp32.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Reducing Precision:\n", - "\n", - "Inference typically requires less numeric precision than training. With some care, lower precision can give you faster computation and lower memory consumption without sacrificing any meaningful accuracy. TensorRT supports TF32, FP32, FP16, and INT8 precisions." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Reducing precision to FP16:__" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "FP16 \"mixed precision\" inference gives up some accuracy in exchange for faster models with lower latency and lower memory footprint. In practice, the accuracy loss is generally negligible in FP16 - so FP16 is a fairly safe bet in most cases for inference. Cards that are focused on deep learning training often have strong FP16 capabilities, making FP16 a great choice for GPUs that are expected to be used for both training and inference - such as the NVIDIA V100\n", - "\n", - "Let's convert our model to FP16 and see how it performs:" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_1_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/resnet50_saved_model_FP16/assets\n" - ] - }, - { - "data": { - "text/plain": [ - "array([[1.7182514e-04, 3.3864001e-04, 6.3493084e-05, ..., 1.5010530e-05,\n", - " 1.4759685e-04, 6.7664997e-04],\n", - " [1.7182514e-04, 3.3864001e-04, 6.3493084e-05, ..., 1.5010530e-05,\n", - " 1.4759685e-04, 6.7664997e-04],\n", - " [1.7182514e-04, 3.3864001e-04, 6.3493084e-05, ..., 1.5010530e-05,\n", - " 1.4759685e-04, 6.7664997e-04],\n", - " ...,\n", - " [1.7182514e-04, 3.3864001e-04, 6.3493084e-05, ..., 1.5010530e-05,\n", - " 1.4759685e-04, 6.7664997e-04],\n", - " [1.7182514e-04, 3.3864001e-04, 6.3493084e-05, ..., 1.5010530e-05,\n", - " 1.4759685e-04, 6.7664997e-04],\n", - " [1.7182514e-04, 3.3864001e-04, 6.3493084e-05, ..., 1.5010530e-05,\n", - " 1.4759685e-04, 6.7664997e-04]], dtype=float32)" - ] - }, - "execution_count": 14, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "model_fp16 = opt_model.convert(model_dir+'_FP16', precision=\"FP16\")\n", - "\n", - "model_fp16.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "13.5 ms ± 20.2 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "model_fp16.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Reducing precision to INT8:__\n", - "\n", - "Whether you want to further reduce to INT8 precision depends on hardware - Turing cards and later INT8 is often better. Inference focused cards such as the NVIDIA T4 or systems-on-module such as Jetson AGX Xavier do well with INT8. In contrast, on a training-focused GPU like V100, INT8 often isn't any faster than FP16." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "To perform INT8 inference, we need to see what the normal range of activations are in the network so we can quantize our INT8 representations based on a normal set of values for our dataset. It is important that this dataset is representative of the testing samples in order to maintain accuracy levels.\n", - "\n", - "Here, we just want to see how our network performs in TensorRT from a runtime standpoint - so we will just feed dummy data and dummy calibration data into TensorRT." - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "metadata": {}, - "outputs": [], - "source": [ - "dummy_calibration_batch = np.zeros((8, 224, 224, 3))\n", - "\n", - "opt_model.set_calibration_data(dummy_calibration_batch)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Then, we convert our model to INT8 as before:" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/resnet50_saved_model_INT8/assets\n" - ] - }, - { - "data": { - "text/plain": [ - "array([[1.61497956e-04, 3.58211488e-04, 7.12977999e-05, ...,\n", - " 1.43723055e-05, 1.47045619e-04, 7.21490127e-04],\n", - " [1.61497956e-04, 3.58211488e-04, 7.12977999e-05, ...,\n", - " 1.43723055e-05, 1.47045619e-04, 7.21490127e-04],\n", - " [1.61497956e-04, 3.58211488e-04, 7.12977999e-05, ...,\n", - " 1.43723055e-05, 1.47045619e-04, 7.21490127e-04],\n", - " ...,\n", - " [1.61497956e-04, 3.58211488e-04, 7.12977999e-05, ...,\n", - " 1.43723055e-05, 1.47045619e-04, 7.21490127e-04],\n", - " [1.61497956e-04, 3.58211488e-04, 7.12977999e-05, ...,\n", - " 1.43723055e-05, 1.47045619e-04, 7.21490127e-04],\n", - " [1.61497956e-04, 3.58211488e-04, 7.12977999e-05, ...,\n", - " 1.43723055e-05, 1.47045619e-04, 7.21490127e-04]], dtype=float32)" - ] - }, - "execution_count": 17, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "model_int8 = opt_model.convert(model_dir+'_INT8', precision=\"INT8\")\n", - "\n", - "model_int8.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "code", - "execution_count": 18, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "13.1 ms ± 29.5 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "model_int8.predict(dummy_input_batch)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Next Steps:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "You can find other Jupyter Notebooks demonstrating TF-TRT conversions and end to end workflows for many other Keras applications and models, including detection models and segmentation models, in other example TF-TRT notebooks!\n", - "\n", - "Here are links to those notebooks:\n", - "\n", - "[__Classification Examples__](./Additional%20Examples/1.%20TF-TRT%20Classification.ipynb)\n", - "\n", - "[__Detection Example__](./Additional%20Examples/2.%20TF-TRT%20Detection.ipynb)\n", - "\n", - "[__Segmentation Example__](./Additional%20Examples/3.%20TF-TRT%20Segmentation.ipynb)" - ] - } - ], - "metadata": { - "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": "3.6.9" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/quickstart/IntroNotebooks/3. Using Tensorflow 2 through ONNX.ipynb b/quickstart/IntroNotebooks/3. Using Tensorflow 2 through ONNX.ipynb deleted file mode 100644 index aa8f6328..00000000 --- a/quickstart/IntroNotebooks/3. Using Tensorflow 2 through ONNX.ipynb +++ /dev/null @@ -1,1275 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Using Tensorflow through ONNX:\n", - "\n", - "The ONNX path to getting a TensorRT engine is a high-performance approach to TensorRT conversion that works with a variety of frameworks - including Tensorflow and Tensorflow 2.\n", - "\n", - "TensorRT's ONNX parser is an all-or-nothing parser for ONNX models that ensures an optimal, single TensorRT engine and is great for exporting to the TensorRT API runtimes. ONNX models can be easily generated from Tensorflow models using the ONNX project's tf2onnx tool.\n", - "\n", - "In this notebook we will take a look at how ONNX models can be generated from a Keras/TF2 ResNet50 model, how we can convert those ONNX models to TensorRT engines using trtexec, and finally how we can use the native Python TensorRT runtime to feed a batch of data into the TRT engine at inference time.\n", - "\n", - "Essentially, we will follow this path to convert and deploy our model:\n", - "\n", - "![Tensorflow+ONNX](./images/tf_onnx.png)\n", - "\n", - "__Use this when:__\n", - "- You want the most efficient runtime performance possible out of an automatic parser\n", - "- You have a network consisting of mostly supported operations - including operations and layers that the ONNX parser uniquely supports (Such as RNNs/LSTMs/GRUs)\n", - "- You are willing to write custom C++ plugins for any unsupported operations (if your network has any)\n", - "- You do not want to use the manual layer builder API" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Checking your GPU status:__\n", - "\n", - "Lets see what GPU hardware we are working with. Our hardware can matter a lot because different cards have different performance profiles and precisions they tend to operate best in. For example, a V100 is relatively strong as FP16 processing vs a T4, which tends to operate best in the INT8 mode." - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/", - "height": 377 - }, - "id": "IJBfZsGo8yaV", - "outputId": "f4c4e20d-fcfd-43a2-b10d-c6978c25c91f" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Wed Jun 9 19:47:48 2021 \n", - "+-----------------------------------------------------------------------------+\n", - "| NVIDIA-SMI 450.80.02 Driver Version: 450.80.02 CUDA Version: 11.3 |\n", - "|-------------------------------+----------------------+----------------------+\n", - "| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n", - "| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n", - "| | | MIG M. |\n", - "|===============================+======================+======================|\n", - "| 0 Tesla V100-DGXS... On | 00000000:07:00.0 Off | 0 |\n", - "| N/A 45C P0 63W / 300W | 5572MiB / 16155MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 1 Tesla V100-DGXS... On | 00000000:08:00.0 Off | 0 |\n", - "| N/A 44C P0 41W / 300W | 9MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 2 Tesla V100-DGXS... On | 00000000:0E:00.0 Off | 0 |\n", - "| N/A 43C P0 41W / 300W | 9MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 3 Tesla V100-DGXS... On | 00000000:0F:00.0 Off | 0 |\n", - "| N/A 44C P0 39W / 300W | 9MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - " \n", - "+-----------------------------------------------------------------------------+\n", - "| Processes: |\n", - "| GPU GI CI PID Type Process name GPU Memory |\n", - "| ID ID Usage |\n", - "|=============================================================================|\n", - "+-----------------------------------------------------------------------------+\n" - ] - } - ], - "source": [ - "!nvidia-smi" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Remember to sucessfully deploy a TensorRT model, you have to make __five key decisions__:\n", - "\n", - "1. __What format should I save my model in?__\n", - "2. __What batch size(s) am I running inference at?__\n", - "3. __What precision am I running inference at?__\n", - "4. __What TensorRT path am I using to convert my model?__\n", - "5. __What runtime am I targeting?__" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 1. What format should I save my model in?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Our first step is to load up a pretrained ResNet50 model. This can be done easily using keras.applications - a collection of pretrained image model classifiers that can additionally be used as backbones for detection and other deep learning problems.\n", - "\n", - "We can load up a pretrained classifier with batch size 32 as follows:" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": { - "id": "iVRVItvR8quS" - }, - "outputs": [], - "source": [ - "from tensorflow.keras.applications import ResNet50\n", - "\n", - "BATCH_SIZE = 32" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": { - "id": "cKT07xPV8qua" - }, - "outputs": [], - "source": [ - "model = ResNet50(weights='imagenet')" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For the purposes of checking our non-optimized model, we can use a dummy batch of data to verify our performance and the consistency of our results across precisions. 224x224 RGB images are a common format, so lets generate a batch of them.\n", - "\n", - "Once we generate a batch of them, we will feed it through the model using .predict() to \"warm up\" the model. The first batch you feed through a deep learning model often takes a lot longer as just-in-time compilation and other runtime optimizations are performed. Once you get that first batch through, further performance tends to be more consistent.\n", - "\n", - "To create a test batch, we will simply repeat one open-source dog image from http://www.dog.ceo" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(32, 224, 224, 3)" - ] - }, - "execution_count": 4, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "import numpy as np\n", - "from skimage import io\n", - "from skimage.transform import resize\n", - "from matplotlib import pyplot as plt\n", - "\n", - "url='https://images.dog.ceo/breeds/retriever-golden/n02099601_3004.jpg'\n", - "img = resize(io.imread(url), (224, 224))\n", - "input_batch = 255*np.array(np.repeat(np.expand_dims(np.array(img, dtype=np.float32), axis=0), BATCH_SIZE, axis=0), dtype=np.float32)\n", - "\n", - "input_batch.shape" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "" - ] - }, - "execution_count": 5, - "metadata": {}, - "output_type": "execute_result" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "plt.imshow(input_batch[0]/255)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The image above is a Golden Retriever, class 207 in ImageNet. So we look for class 207 in the top 5 predictions to verify our model works as intended:" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Class | Probability (out of 1)\n" - ] - }, - { - "data": { - "text/plain": [ - "[(160, 0.32290387),\n", - " (169, 0.266499),\n", - " (212, 0.16812354),\n", - " (170, 0.07066823),\n", - " (207, 0.03341851)]" - ] - }, - "execution_count": 6, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "predictions = model.predict(input_batch) # warm up\n", - "indices = (-predictions[0]).argsort()[:5]\n", - "print(\"Class | Probability (out of 1)\")\n", - "list(zip(indices, predictions[0][indices]))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Labels 150 to 275 or so are dogs in ImageNet, so look for those as other common predictions in addition to our correct 207 class." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Baseline Timing:__\n", - "\n", - "Once we have warmed up our non-optimized model, we can get a rough timing estimate of our model using %%timeit, which runs the cell several times and reports timing information.\n", - "\n", - "Lets take a look at how long our model takes to run at baseline before doing any TensorRT optimization:" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/", - "height": 85 - }, - "id": "eMu3dZlM96bh", - "outputId": "537a88e2-ad7d-413a-f815-abd91f010e21" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "46.8 ms ± 514 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "result = model.predict_on_batch(input_batch) # Check default performance" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Okay - now that we have a baseline model, lets convert it to the format TensorRT understands best: ONNX. " - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Convert Keras model to ONNX intermediate model and save:__" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The ONNX format is a framework-agnostic way of describing and saving the structure and state of deep learning models. We can convert Tensorflow 2 Keras models to ONNX using the tf2onnx tool provided by the ONNX project. (You can find the ONNX project here: https://onnx.ai or on GitHub here: https://github.com/onnx/onnx)" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": { - "id": "aG3tXUEx8quf" - }, - "outputs": [], - "source": [ - "import onnx" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Converting a model with default parameters to an ONNX model is fairly straightforward:" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/", - "height": 68 - }, - "id": "QxLAvWp68quk", - "outputId": "d750962a-d098-4a63-c195-c3442211cdc1" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "INFO:tensorflow:Assets written to: my_model/assets\n", - "2021-06-09 19:48:30.462380: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "/usr/lib/python3.8/runpy.py:127: RuntimeWarning: 'tf2onnx.convert' found in sys.modules after import of package 'tf2onnx', but prior to execution of 'tf2onnx.convert'; this may result in unpredictable behaviour\n", - " warn(RuntimeWarning(msg))\n", - "2021-06-09 19:48:31.938818: I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set\n", - "2021-06-09 19:48:31.939684: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcuda.so.1\n", - "2021-06-09 19:48:32.010614: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 0 with properties: \n", - "pciBusID: 0000:07:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:32.011850: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 1 with properties: \n", - "pciBusID: 0000:08:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:32.013128: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 2 with properties: \n", - "pciBusID: 0000:0e:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:32.014344: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 3 with properties: \n", - "pciBusID: 0000:0f:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:32.014373: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "2021-06-09 19:48:32.019097: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n", - "2021-06-09 19:48:32.019146: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n", - "2021-06-09 19:48:32.020281: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n", - "2021-06-09 19:48:32.020567: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n", - "2021-06-09 19:48:32.021254: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.11\n", - "2021-06-09 19:48:32.022280: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n", - "2021-06-09 19:48:32.022445: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n", - "2021-06-09 19:48:32.030879: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1889] Adding visible gpu devices: 0, 1, 2, 3\n", - "2021-06-09 19:48:32.032680: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n", - "2021-06-09 19:48:33.010741: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 0 with properties: \n", - "pciBusID: 0000:07:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:33.011970: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 1 with properties: \n", - "pciBusID: 0000:08:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:33.013195: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 2 with properties: \n", - "pciBusID: 0000:0e:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:33.014389: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 3 with properties: \n", - "pciBusID: 0000:0f:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:33.014428: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "2021-06-09 19:48:33.014458: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n", - "2021-06-09 19:48:33.014478: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n", - "2021-06-09 19:48:33.014497: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n", - "2021-06-09 19:48:33.014516: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n", - "2021-06-09 19:48:33.014534: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.11\n", - "2021-06-09 19:48:33.014552: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n", - "2021-06-09 19:48:33.014571: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n", - "2021-06-09 19:48:33.022970: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1889] Adding visible gpu devices: 0, 1, 2, 3\n", - "2021-06-09 19:48:33.023016: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "2021-06-09 19:48:35.609734: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1287] Device interconnect StreamExecutor with strength 1 edge matrix:\n", - "2021-06-09 19:48:35.609783: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1293] 0 1 2 3 \n", - "2021-06-09 19:48:35.609797: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 0: N Y Y Y \n", - "2021-06-09 19:48:35.609806: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 1: Y N Y Y \n", - "2021-06-09 19:48:35.609816: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 2: Y Y N Y \n", - "2021-06-09 19:48:35.609825: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 3: Y Y Y N \n", - "2021-06-09 19:48:35.619000: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 203 MB memory) -> physical GPU (device: 0, name: Tesla V100-DGXS-16GB, pci bus id: 0000:07:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:35.620513: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:1 with 14206 MB memory) -> physical GPU (device: 1, name: Tesla V100-DGXS-16GB, pci bus id: 0000:08:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:35.621962: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:2 with 14206 MB memory) -> physical GPU (device: 2, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0e:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:35.623398: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:3 with 14206 MB memory) -> physical GPU (device: 3, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0f:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:35,625 - WARNING - '--tag' not specified for saved_model. Using --tag serve\n", - "2021-06-09 19:48:43,221 - INFO - Signatures found in model: [serving_default].\n", - "2021-06-09 19:48:43,221 - WARNING - '--signature_def' not specified, using first signature: serving_default\n", - "2021-06-09 19:48:43,222 - INFO - Output names: ['predictions']\n", - "2021-06-09 19:48:43.250962: I tensorflow/core/grappler/devices.cc:69] Number of eligible GPUs (core count >= 8, compute capability >= 0.0): 4\n", - "2021-06-09 19:48:43.251124: I tensorflow/core/grappler/clusters/single_machine.cc:356] Starting new session\n", - "2021-06-09 19:48:43.251388: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n", - "2021-06-09 19:48:43.252059: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 0 with properties: \n", - "pciBusID: 0000:07:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:43.253259: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 1 with properties: \n", - "pciBusID: 0000:08:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:43.254444: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 2 with properties: \n", - "pciBusID: 0000:0e:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:43.255627: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 3 with properties: \n", - "pciBusID: 0000:0f:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:43.255663: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "2021-06-09 19:48:43.255693: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n", - "2021-06-09 19:48:43.255712: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n", - "2021-06-09 19:48:43.255730: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n", - "2021-06-09 19:48:43.255748: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n", - "2021-06-09 19:48:43.255765: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.11\n", - "2021-06-09 19:48:43.255783: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n", - "2021-06-09 19:48:43.255801: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n", - "2021-06-09 19:48:43.264001: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1889] Adding visible gpu devices: 0, 1, 2, 3\n", - "2021-06-09 19:48:43.264071: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1287] Device interconnect StreamExecutor with strength 1 edge matrix:\n", - "2021-06-09 19:48:43.264086: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1293] 0 1 2 3 \n", - "2021-06-09 19:48:43.264097: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 0: N Y Y Y \n", - "2021-06-09 19:48:43.264106: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 1: Y N Y Y \n", - "2021-06-09 19:48:43.264116: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 2: Y Y N Y \n", - "2021-06-09 19:48:43.264125: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 3: Y Y Y N \n", - "2021-06-09 19:48:43.269085: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 203 MB memory) -> physical GPU (device: 0, name: Tesla V100-DGXS-16GB, pci bus id: 0000:07:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:43.270297: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:1 with 14206 MB memory) -> physical GPU (device: 1, name: Tesla V100-DGXS-16GB, pci bus id: 0000:08:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:43.271732: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:2 with 14206 MB memory) -> physical GPU (device: 2, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0e:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:43.273448: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:3 with 14206 MB memory) -> physical GPU (device: 3, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0f:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:43.293134: I tensorflow/core/platform/profile_utils/cpu_utils.cc:112] CPU Frequency: 2198860000 Hz\n", - "2021-06-09 19:48:43.355209: I tensorflow/core/grappler/optimizers/meta_optimizer.cc:954] Optimization results for grappler item: graph_to_optimize\n", - " function_optimizer: Graph size after: 1253 nodes (930), 1908 edges (1585), time = 33.193ms.\n", - " function_optimizer: function_optimizer did nothing. time = 0.577ms.\n", - "\n", - "2021-06-09 19:48:46.008484: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n", - "2021-06-09 19:48:46.031017: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 0 with properties: \n", - "pciBusID: 0000:07:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.033674: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 1 with properties: \n", - "pciBusID: 0000:08:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.035311: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 2 with properties: \n", - "pciBusID: 0000:0e:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.036940: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 3 with properties: \n", - "pciBusID: 0000:0f:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.036986: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "2021-06-09 19:48:46.037035: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n", - "2021-06-09 19:48:46.037062: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n", - "2021-06-09 19:48:46.037086: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n", - "2021-06-09 19:48:46.037110: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n", - "2021-06-09 19:48:46.037133: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.11\n", - "2021-06-09 19:48:46.037157: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n", - "2021-06-09 19:48:46.037181: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n", - "2021-06-09 19:48:46.046998: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1889] Adding visible gpu devices: 0, 1, 2, 3\n", - "2021-06-09 19:48:46.047077: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1287] Device interconnect StreamExecutor with strength 1 edge matrix:\n", - "2021-06-09 19:48:46.047095: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1293] 0 1 2 3 \n", - "2021-06-09 19:48:46.047108: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 0: N Y Y Y \n", - "2021-06-09 19:48:46.047120: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 1: Y N Y Y \n", - "2021-06-09 19:48:46.047131: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 2: Y Y N Y \n", - "2021-06-09 19:48:46.047142: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 3: Y Y Y N \n", - "2021-06-09 19:48:46.052418: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 203 MB memory) -> physical GPU (device: 0, name: Tesla V100-DGXS-16GB, pci bus id: 0000:07:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:46.053664: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:1 with 14206 MB memory) -> physical GPU (device: 1, name: Tesla V100-DGXS-16GB, pci bus id: 0000:08:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:46.054881: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:2 with 14206 MB memory) -> physical GPU (device: 2, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0e:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:46.056098: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:3 with 14206 MB memory) -> physical GPU (device: 3, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0f:00.0, compute capability: 7.0)\n", - "WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/tf2onnx/tf_loader.py:603: extract_sub_graph (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "Use `tf.compat.v1.graph_util.extract_sub_graph`\n", - "2021-06-09 19:48:46,541 - WARNING - From /usr/local/lib/python3.8/dist-packages/tf2onnx/tf_loader.py:603: extract_sub_graph (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "Use `tf.compat.v1.graph_util.extract_sub_graph`\n", - "2021-06-09 19:48:46.600644: I tensorflow/core/grappler/devices.cc:69] Number of eligible GPUs (core count >= 8, compute capability >= 0.0): 4\n", - "2021-06-09 19:48:46.600797: I tensorflow/core/grappler/clusters/single_machine.cc:356] Starting new session\n", - "2021-06-09 19:48:46.601148: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n", - "2021-06-09 19:48:46.602435: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 0 with properties: \n", - "pciBusID: 0000:07:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.604322: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 1 with properties: \n", - "pciBusID: 0000:08:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.606193: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 2 with properties: \n", - "pciBusID: 0000:0e:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.608049: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1747] Found device 3 with properties: \n", - "pciBusID: 0000:0f:00.0 name: Tesla V100-DGXS-16GB computeCapability: 7.0\n", - "coreClock: 1.53GHz coreCount: 80 deviceMemorySize: 15.78GiB deviceMemoryBandwidth: 836.37GiB/s\n", - "2021-06-09 19:48:46.608091: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n", - "2021-06-09 19:48:46.608129: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11\n", - "2021-06-09 19:48:46.608153: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11\n", - "2021-06-09 19:48:46.608176: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10\n", - "2021-06-09 19:48:46.608198: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10\n", - "2021-06-09 19:48:46.608220: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.11\n", - "2021-06-09 19:48:46.608242: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11\n", - "2021-06-09 19:48:46.608265: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8\n", - "2021-06-09 19:48:46.625482: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1889] Adding visible gpu devices: 0, 1, 2, 3\n", - "2021-06-09 19:48:46.625560: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1287] Device interconnect StreamExecutor with strength 1 edge matrix:\n", - "2021-06-09 19:48:46.625578: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1293] 0 1 2 3 \n", - "2021-06-09 19:48:46.625590: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 0: N Y Y Y \n", - "2021-06-09 19:48:46.625601: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 1: Y N Y Y \n", - "2021-06-09 19:48:46.625612: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 2: Y Y N Y \n", - "2021-06-09 19:48:46.625623: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1306] 3: Y Y Y N \n", - "2021-06-09 19:48:46.634557: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 203 MB memory) -> physical GPU (device: 0, name: Tesla V100-DGXS-16GB, pci bus id: 0000:07:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:46.636578: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:1 with 14206 MB memory) -> physical GPU (device: 1, name: Tesla V100-DGXS-16GB, pci bus id: 0000:08:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:46.638422: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:2 with 14206 MB memory) -> physical GPU (device: 2, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0e:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:46.640290: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:3 with 14206 MB memory) -> physical GPU (device: 3, name: Tesla V100-DGXS-16GB, pci bus id: 0000:0f:00.0, compute capability: 7.0)\n", - "2021-06-09 19:48:47.379855: I tensorflow/core/grappler/optimizers/meta_optimizer.cc:954] Optimization results for grappler item: graph_to_optimize\n", - " constant_folding: Graph size after: 560 nodes (-640), 1215 edges (-640), time = 399.986ms.\n", - " function_optimizer: function_optimizer did nothing. time = 1.17ms.\n", - " constant_folding: Graph size after: 560 nodes (0), 1215 edges (0), time = 101.728ms.\n", - " function_optimizer: function_optimizer did nothing. time = 1.017ms.\n", - "\n", - "2021-06-09 19:48:47,938 - INFO - Using tensorflow=2.4.0, onnx=1.9.0, tf2onnx=1.8.5/50049d\n", - "2021-06-09 19:48:47,939 - INFO - Using opset \n", - "2021-06-09 19:48:52,720 - INFO - Computed 0 values for constant folding\n", - "2021-06-09 19:49:05,218 - INFO - Optimizing ONNX model\n", - "2021-06-09 19:49:06,920 - INFO - After optimization: Add -1 (18->17), BatchNormalization -53 (53->0), Const -162 (270->108), GlobalAveragePool +1 (0->1), Identity -57 (57->0), ReduceMean -1 (1->0), Squeeze +1 (0->1), Transpose -213 (214->1)\n", - "2021-06-09 19:49:07,076 - INFO - \n", - "2021-06-09 19:49:07,076 - INFO - Successfully converted TensorFlow model my_model to ONNX\n", - "2021-06-09 19:49:07,076 - INFO - Model inputs: ['input_1:0']\n", - "2021-06-09 19:49:07,076 - INFO - Model outputs: ['predictions']\n", - "2021-06-09 19:49:07,076 - INFO - ONNX model is saved at temp.onnx\n" - ] - } - ], - "source": [ - "model.save('my_model')\n", - "!python -m tf2onnx.convert --saved-model my_model --output temp.onnx\n", - "onnx_model = onnx.load_model('temp.onnx')" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "That said, we do need to make one change for our model to work with TensorRT. Keras by default uses a dynamic input shape in its networks - where it can handle arbitrary batch sizes at every update. While TensorRT can do this, it requires extra configuration. \n", - "\n", - "Instead, we will just set the input size to be fixed to our batch size. This will work with TensorRT out of the box!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Configure ONNX File Batch Size:__" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Note:__ We need to do two things to set our batch size with ONNX. The first is to modify our ONNX file to change its default batch size to our target batch size. The second is setting our converter to use the __explicit batch__ mode, which will use this default batch size as our final batch size." - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [], - "source": [ - "inputs = onnx_model.graph.input\n", - "for input in inputs:\n", - " dim1 = input.type.tensor_type.shape.dim[0]\n", - " dim1.dim_value = BATCH_SIZE" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Save Model:__" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": { - "id": "jFT6-13f8qup" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Done saving!\n" - ] - } - ], - "source": [ - "model_name = \"resnet50_onnx_model.onnx\"\n", - "onnx.save_model(onnx_model, model_name)\n", - "print(\"Done saving!\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Once we get our model into ONNX format, we can convert it efficiently using TensorRT. For this, TensorRT needs exclusive access to your GPU. If you so much as import Tensorflow, it will generally consume all of your GPU memory. To get around this, before moving on go ahead and shut down this notebook and restart it. (You can do this in the menu: Kernel -> Restart Kernel)\n", - "\n", - "Make sure not to import Tensorflow at any point after restarting the runtime! \n", - "\n", - "(The following cell is a quick shortcut to make your notebook restart:)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "id": "uZUnHVHE8quu" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Restarting kernel in three seconds...\n" - ] - } - ], - "source": [ - "import os, time\n", - "print(\"Restarting kernel in three seconds...\")\n", - "time.sleep(3)\n", - "print(\"Restarting kernel now\")\n", - "os._exit(0) # Shut down all kernels so TRT doesn't fight with Tensorflow for GPU memory - TF monopolizes all GPU memory by default" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 2. What batch size(s) am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We have actually already set our inference batch size - see the note above in section 1!\n", - "\n", - "We are going to set our target batch size to a fixed size of 32." - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [], - "source": [ - "BATCH_SIZE = 32" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We need to do two things to set our batch size to a fixed batch size with ONNX: \n", - "\n", - "1. Modify our ONNX file to change its default batch size to our target batch size, which we did above.\n", - "2. Use the trtexec --explicitBatch flag, which we also did above." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 3. What precision am I running inference at?\n", - "\n", - "Now, we have a converted TensorRT engine. Great! That means we are ready to load it into the native Python TensorRT runtime. This runtime strikes a balance between the ease of use of the high level Python runtimes and the low level C++ runtimes.\n", - "\n", - "First, as before, lets create a dummy batch. Importantly, by default TensorRT will use the input precision you give it as the default precision for the rest of the network. \n", - "\n", - "Remember that lower precisions than FP32 tend to run faster. There are two common reduced precision modes - FP16 and INT8. Graphics cards that are designed to do inference well often have an affinity for one of these two types. This guide was developed on an NVIDIA V100, which favors FP16, so we will use that here by default. INT8 is a more complicated process that requires a calibration step." - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "USE_FP16 = True\n", - "\n", - "target_dtype = np.float16 if USE_FP16 else np.float32" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We generate a batch of repeating Golden Retriever images, as before. Make sure that for TensorRT the image is resized to the size your model expects. Tensorflow and TensorRT have different behavior for handling 'oversized' images - so this is a safe way of ensuring consistent results across the two." - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "from skimage import io\n", - "from skimage.transform import resize\n", - "from matplotlib import pyplot as plt\n", - "\n", - "url='https://images.dog.ceo/breeds/retriever-golden/n02099601_3004.jpg'\n", - "img = resize(io.imread(url), (224, 224))\n", - "input_batch = 255*np.array(np.repeat(np.expand_dims(np.array(img, dtype=np.float32), axis=0), BATCH_SIZE, axis=0), dtype=np.float32)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Only we must now cast the input batch to the proper FP32/FP16 precision:" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "input_batch = input_batch.astype(target_dtype)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 4. What TensorRT path am I using to convert my model?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "TensorRT is able to take ONNX models and convert them entirely into a single, efficient TensorRT engine. Restart your Jupyter kernel, and then start here!\n", - "\n", - "We can use trtexec, a command line tool for working with TensorRT, in order to convert an ONNX model to an engine file.\n", - "\n", - "To convert the model we saved in the previous steps, we need to point to the ONNX file, give trtexec a name to save the engine as, and last specify that we want to use a fixed batch size instead of a dynamic one.\n", - "\n", - "__Remember to shut down all Jupyter notebooks and restart your Jupyter kernel after \"1. What format should I save my model in?\" - otherwise this cell will crash as TensorRT competes with Tensorflow for GPU memory:__" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": { - "colab": { - "base_uri": "https://localhost:8080/", - "height": 34 - }, - "id": "h60Gmotx8quz", - "outputId": "065384aa-c848-4194-c72c-cad0d80449ca" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "&&&& RUNNING TensorRT.trtexec # trtexec --onnx=resnet50_onnx_model.onnx --saveEngine=resnet_engine.trt --explicitBatch --inputIOFormats=fp16:chw --outputIOFormats=fp16:chw --fp16\n", - "[06/09/2021-19:49:25] [I] === Model Options ===\n", - "[06/09/2021-19:49:25] [I] Format: ONNX\n", - "[06/09/2021-19:49:25] [I] Model: resnet50_onnx_model.onnx\n", - "[06/09/2021-19:49:25] [I] Output:\n", - "[06/09/2021-19:49:25] [I] === Build Options ===\n", - "[06/09/2021-19:49:25] [I] Max batch: explicit\n", - "[06/09/2021-19:49:25] [I] Workspace: 16 MiB\n", - "[06/09/2021-19:49:25] [I] minTiming: 1\n", - "[06/09/2021-19:49:25] [I] avgTiming: 8\n", - "[06/09/2021-19:49:25] [I] Precision: FP32+FP16\n", - "[06/09/2021-19:49:25] [I] Calibration: \n", - "[06/09/2021-19:49:25] [I] Refit: Disabled\n", - "[06/09/2021-19:49:25] [I] Safe mode: Disabled\n", - "[06/09/2021-19:49:25] [I] Save engine: resnet_engine.trt\n", - "[06/09/2021-19:49:25] [I] Load engine: \n", - "[06/09/2021-19:49:25] [I] Builder Cache: Enabled\n", - "[06/09/2021-19:49:25] [I] NVTX verbosity: 0\n", - "[06/09/2021-19:49:25] [I] Tactic sources: Using default tactic sources\n", - "[06/09/2021-19:49:25] [I] Input(s): fp16:chw\n", - "[06/09/2021-19:49:25] [I] Output(s): fp16:chw\n", - "[06/09/2021-19:49:25] [I] Input build shapes: model\n", - "[06/09/2021-19:49:25] [I] Input calibration shapes: model\n", - "[06/09/2021-19:49:25] [I] === System Options ===\n", - "[06/09/2021-19:49:25] [I] Device: 0\n", - "[06/09/2021-19:49:25] [I] DLACore: \n", - "[06/09/2021-19:49:25] [I] Plugins:\n", - "[06/09/2021-19:49:25] [I] === Inference Options ===\n", - "[06/09/2021-19:49:25] [I] Batch: Explicit\n", - "[06/09/2021-19:49:25] [I] Input inference shapes: model\n", - "[06/09/2021-19:49:25] [I] Iterations: 10\n", - "[06/09/2021-19:49:25] [I] Duration: 3s (+ 200ms warm up)\n", - "[06/09/2021-19:49:25] [I] Sleep time: 0ms\n", - "[06/09/2021-19:49:25] [I] Streams: 1\n", - "[06/09/2021-19:49:25] [I] ExposeDMA: Disabled\n", - "[06/09/2021-19:49:25] [I] Data transfers: Enabled\n", - "[06/09/2021-19:49:25] [I] Spin-wait: Disabled\n", - "[06/09/2021-19:49:25] [I] Multithreading: Disabled\n", - "[06/09/2021-19:49:25] [I] CUDA Graph: Disabled\n", - "[06/09/2021-19:49:25] [I] Separate profiling: Disabled\n", - "[06/09/2021-19:49:25] [I] Skip inference: Disabled\n", - "[06/09/2021-19:49:25] [I] Inputs:\n", - "[06/09/2021-19:49:25] [I] === Reporting Options ===\n", - "[06/09/2021-19:49:25] [I] Verbose: Disabled\n", - "[06/09/2021-19:49:25] [I] Averages: 10 inferences\n", - "[06/09/2021-19:49:25] [I] Percentile: 99\n", - "[06/09/2021-19:49:25] [I] Dump refittable layers:Disabled\n", - "[06/09/2021-19:49:25] [I] Dump output: Disabled\n", - "[06/09/2021-19:49:25] [I] Profile: Disabled\n", - "[06/09/2021-19:49:25] [I] Export timing to JSON file: \n", - "[06/09/2021-19:49:25] [I] Export output to JSON file: \n", - "[06/09/2021-19:49:25] [I] Export profile to JSON file: \n", - "[06/09/2021-19:49:25] [I] \n", - "[06/09/2021-19:49:25] [I] === Device Information ===\n", - "[06/09/2021-19:49:25] [I] Selected Device: Tesla V100-DGXS-16GB\n", - "[06/09/2021-19:49:25] [I] Compute Capability: 7.0\n", - "[06/09/2021-19:49:25] [I] SMs: 80\n", - "[06/09/2021-19:49:25] [I] Compute Clock Rate: 1.53 GHz\n", - "[06/09/2021-19:49:25] [I] Device Global Memory: 16155 MiB\n", - "[06/09/2021-19:49:25] [I] Shared Memory per SM: 96 KiB\n", - "[06/09/2021-19:49:25] [I] Memory Bus Width: 4096 bits (ECC enabled)\n", - "[06/09/2021-19:49:25] [I] Memory Clock Rate: 0.877 GHz\n", - "[06/09/2021-19:49:25] [I] \n", - "[06/09/2021-19:49:42] [I] [TRT] ----------------------------------------------------------------\n", - "[06/09/2021-19:49:42] [I] [TRT] Input filename: resnet50_onnx_model.onnx\n", - "[06/09/2021-19:49:42] [I] [TRT] ONNX IR version: 0.0.4\n", - "[06/09/2021-19:49:42] [I] [TRT] Opset version: 9\n", - "[06/09/2021-19:49:42] [I] [TRT] Producer name: tf2onnx\n", - "[06/09/2021-19:49:42] [I] [TRT] Producer version: 1.8.5\n", - "[06/09/2021-19:49:42] [I] [TRT] Domain: \n", - "[06/09/2021-19:49:42] [I] [TRT] Model version: 0\n", - "[06/09/2021-19:49:42] [I] [TRT] Doc string: \n", - "[06/09/2021-19:49:42] [I] [TRT] ----------------------------------------------------------------\n", - "[06/09/2021-19:49:48] [I] [TRT] Some tactics do not have sufficient workspace memory to run. Increasing workspace size may increase performance, please check verbose output.\n", - "[06/09/2021-19:51:05] [I] [TRT] Detected 1 inputs and 1 output network tensors.\n", - "[06/09/2021-19:51:06] [I] Engine built in 100.683 sec.\n", - "[06/09/2021-19:51:06] [I] Starting inference\n", - "[06/09/2021-19:51:09] [I] Warmup completed 0 queries over 200 ms\n", - "[06/09/2021-19:51:09] [I] Timing trace has 0 queries over 2.99006 s\n", - "[06/09/2021-19:51:09] [I] Trace averages of 10 runs:\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48546 ms - Host latency: 6.30948 ms (end to end 10.0032 ms, enqueue 0.539108 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48946 ms - Host latency: 6.31468 ms (end to end 10.9038 ms, enqueue 0.516052 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48004 ms - Host latency: 6.3107 ms (end to end 10.8822 ms, enqueue 0.513507 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.49315 ms - Host latency: 6.34006 ms (end to end 10.4643 ms, enqueue 0.512753 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.52059 ms - Host latency: 6.36953 ms (end to end 10.2954 ms, enqueue 0.498505 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.50788 ms - Host latency: 6.3551 ms (end to end 9.11696 ms, enqueue 0.518701 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.49774 ms - Host latency: 6.3454 ms (end to end 10.9278 ms, enqueue 0.495056 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.50585 ms - Host latency: 6.35638 ms (end to end 10.9322 ms, enqueue 0.505725 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.50247 ms - Host latency: 6.35249 ms (end to end 10.5564 ms, enqueue 0.513574 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.51249 ms - Host latency: 6.36059 ms (end to end 9.63242 ms, enqueue 0.498096 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.4911 ms - Host latency: 6.33875 ms (end to end 8.90275 ms, enqueue 0.474237 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.50072 ms - Host latency: 6.34651 ms (end to end 10.4826 ms, enqueue 0.498499 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.49602 ms - Host latency: 6.34083 ms (end to end 10.92 ms, enqueue 0.486401 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.49089 ms - Host latency: 6.3358 ms (end to end 10.8925 ms, enqueue 0.490247 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48907 ms - Host latency: 6.33452 ms (end to end 10.1912 ms, enqueue 0.482959 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.47534 ms - Host latency: 6.31992 ms (end to end 8.9359 ms, enqueue 0.484119 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.47952 ms - Host latency: 6.32281 ms (end to end 10.4421 ms, enqueue 0.481885 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48701 ms - Host latency: 6.33408 ms (end to end 10.9013 ms, enqueue 0.491455 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48179 ms - Host latency: 6.33092 ms (end to end 10.885 ms, enqueue 0.505078 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48776 ms - Host latency: 6.33756 ms (end to end 10.3106 ms, enqueue 0.494629 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.47145 ms - Host latency: 6.31754 ms (end to end 9.37426 ms, enqueue 0.481995 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48057 ms - Host latency: 6.32472 ms (end to end 9.55609 ms, enqueue 0.480151 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48557 ms - Host latency: 6.33252 ms (end to end 10.4543 ms, enqueue 0.486841 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.50972 ms - Host latency: 6.35627 ms (end to end 10.9478 ms, enqueue 0.488062 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.50054 ms - Host latency: 6.34517 ms (end to end 10.0418 ms, enqueue 0.483325 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48201 ms - Host latency: 6.32832 ms (end to end 9.67512 ms, enqueue 0.481812 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48279 ms - Host latency: 6.32742 ms (end to end 9.18972 ms, enqueue 0.484082 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.47712 ms - Host latency: 6.32109 ms (end to end 10.879 ms, enqueue 0.482202 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.47788 ms - Host latency: 6.32166 ms (end to end 10.8823 ms, enqueue 0.481006 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48203 ms - Host latency: 6.32615 ms (end to end 10.6967 ms, enqueue 0.481055 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.46802 ms - Host latency: 6.31384 ms (end to end 9.47229 ms, enqueue 0.477344 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.4967 ms - Host latency: 6.3428 ms (end to end 8.9686 ms, enqueue 0.48147 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.49275 ms - Host latency: 6.33767 ms (end to end 9.57681 ms, enqueue 0.481714 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.52278 ms - Host latency: 6.37007 ms (end to end 10.9759 ms, enqueue 0.493896 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.49238 ms - Host latency: 6.34084 ms (end to end 10.7861 ms, enqueue 0.49917 ms)\n", - "[06/09/2021-19:51:09] [I] Average on 10 runs - GPU latency: 5.48333 ms - Host latency: 6.33235 ms (end to end 10.4963 ms, enqueue 0.500806 ms)\n", - "[06/09/2021-19:51:09] [I] Host Latency\n", - "[06/09/2021-19:51:09] [I] min: 6.28442 ms (end to end 6.327 ms)\n", - "[06/09/2021-19:51:09] [I] max: 6.66431 ms (end to end 11.2405 ms)\n", - "[06/09/2021-19:51:09] [I] mean: 6.33588 ms (end to end 10.2251 ms)\n", - "[06/09/2021-19:51:09] [I] median: 6.33411 ms (end to end 10.8945 ms)\n", - "[06/09/2021-19:51:09] [I] percentile: 6.38745 ms at 99% (end to end 11.0925 ms at 99%)\n", - "[06/09/2021-19:51:09] [I] throughput: 0 qps\n", - "[06/09/2021-19:51:09] [I] walltime: 2.99006 s\n", - "[06/09/2021-19:51:09] [I] Enqueue Time\n", - "[06/09/2021-19:51:09] [I] min: 0.413086 ms\n", - "[06/09/2021-19:51:09] [I] max: 0.796997 ms\n", - "[06/09/2021-19:51:09] [I] median: 0.486877 ms\n", - "[06/09/2021-19:51:09] [I] GPU Compute\n", - "[06/09/2021-19:51:09] [I] min: 5.4425 ms\n", - "[06/09/2021-19:51:09] [I] max: 5.82251 ms\n", - "[06/09/2021-19:51:09] [I] mean: 5.49097 ms\n", - "[06/09/2021-19:51:09] [I] median: 5.48969 ms\n", - "[06/09/2021-19:51:09] [I] percentile: 5.53986 ms at 99%\n", - "[06/09/2021-19:51:09] [I] total compute time: 2.00421 s\n", - "&&&& PASSED TensorRT.trtexec # trtexec --onnx=resnet50_onnx_model.onnx --saveEngine=resnet_engine.trt --explicitBatch --inputIOFormats=fp16:chw --outputIOFormats=fp16:chw --fp16\n" - ] - } - ], - "source": [ - "# May need to shut down all kernels and restart before this - otherwise you might get cuDNN initialization errors:\n", - "if USE_FP16:\n", - " !trtexec --onnx=resnet50_onnx_model.onnx --saveEngine=resnet_engine.trt --explicitBatch --inputIOFormats=fp16:chw --outputIOFormats=fp16:chw --fp16\n", - "else:\n", - " !trtexec --onnx=resnet50_onnx_model.onnx --saveEngine=resnet_engine.trt --explicitBatch" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "\n", - "-\n", - "\n", - "__The trtexec Logs:__\n", - "\n", - "Above, trtexec does a lot of things! Some important things to note:\n", - "\n", - "__First__, _\"PASSED\"_ is what you want to see in the last line of the log above. We can see our conversion was successful!\n", - "\n", - "__Second__, can see the resnet_engine.trt engine file has indeed been successfully created: " - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "total 508284\n", - "drwxrwxr-x 8 1000 1000 4096 Jun 9 19:49 .\n", - "drwxrwxr-x 5 1000 1000 4096 Apr 5 23:28 ..\n", - "drwxr-xr-x 2 root root 4096 Apr 6 01:13 .ipynb_checkpoints\n", - "-rw-rw-r-- 1 1000 1000 34748 Jun 9 19:46 '0. Running This Guide.ipynb'\n", - "-rw-rw-r-- 1 1000 1000 502649 Apr 5 23:28 '1. Introduction.ipynb'\n", - "-rw-rw-r-- 1 1000 1000 23645 Apr 5 23:28 '2. Using the Tensorflow TensorRT Integration.ipynb'\n", - "-rw-rw-r-- 1 1000 1000 210995 Jun 9 19:49 '3. Using Tensorflow 2 through ONNX.ipynb'\n", - "-rw-rw-r-- 1 1000 1000 334050 Jun 9 19:17 '4. Using PyTorch through ONNX.ipynb'\n", - "-rw-rw-r-- 1 1000 1000 7052 Apr 5 23:28 '5. Understanding TensorRT Runtimes.ipynb'\n", - "drwxrwxr-x 2 1000 1000 4096 Apr 5 23:28 'Additional Examples'\n", - "drwxr-xr-x 2 root root 4096 Apr 5 23:28 Getting_Started\n", - "drwxr-xr-x 2 root root 4096 Apr 6 01:09 __pycache__\n", - "-rw-rw-r-- 1 1000 1000 4085 Apr 5 23:28 helper.py\n", - "drwxrwxr-x 2 1000 1000 4096 Apr 5 23:28 images\n", - "drwxr-xr-x 4 root root 4096 Jun 9 19:48 my_model\n", - "-rw-rw-r-- 1 1000 1000 3228 Apr 5 23:28 onnx_helper.py\n", - "-rw-r--r-- 1 root root 102169836 Jun 9 19:49 resnet50_onnx_model.onnx\n", - "-rw-r--r-- 1 root root 102470353 Apr 6 04:18 resnet50_pytorch.onnx\n", - "-rw-r--r-- 1 root root 51398352 Jun 9 19:51 resnet_engine.trt\n", - "-rw-r--r-- 1 root root 161081907 Apr 6 17:38 resnet_engine_pytorch.trt\n", - "-rw-r--r-- 1 root root 102169844 Jun 9 19:49 temp.onnx\n" - ] - } - ], - "source": [ - "!ls -la" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Third__, you can see timing details above using trtexec - these are in the ideal case with no overhead. Depending on how you run your model, a considerable amount of overhead can be added to this. We can do timing in our Python runtime below - but keep in mind performing C++ inference would likely be faster." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 5. What TensorRT runtime am I targeting?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We want to run our TensorRT inference in Python - so the TensorRT Python API is a great way of testing our model out in Jupyter, and is still quite performant.\n", - "\n", - "To use it, we need to do a few steps:\n", - "\n", - "__Load our engine into a tensorrt.Runtime:__" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": { - "id": "dX2jFwrA8qu6" - }, - "outputs": [], - "source": [ - "import tensorrt as trt\n", - "import pycuda.driver as cuda\n", - "import pycuda.autoinit\n", - "\n", - "f = open(\"resnet_engine.trt\", \"rb\")\n", - "runtime = trt.Runtime(trt.Logger(trt.Logger.WARNING)) \n", - "\n", - "engine = runtime.deserialize_cuda_engine(f.read())\n", - "context = engine.create_execution_context()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Note: if this cell is having issues, restarting all Jupyter kernels and rerunning only the batch size and precision cells above before trying again often helps" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Allocate input and output memory, give TRT pointers (bindings) to it:__\n", - "\n", - "d_input and d_output refer to the memory regions on our 'device' (aka GPU) - as opposed to memory on our normal RAM, where Python holds its variables (such as 'output' below)." - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": { - "id": "q3UJcdWy8qu8" - }, - "outputs": [], - "source": [ - "output = np.empty([BATCH_SIZE, 1000], dtype = target_dtype) # Need to set output dtype to FP16 to enable FP16\n", - "\n", - "# Allocate device memory\n", - "d_input = cuda.mem_alloc(1 * input_batch.nbytes)\n", - "d_output = cuda.mem_alloc(1 * output.nbytes)\n", - "\n", - "bindings = [int(d_input), int(d_output)]\n", - "\n", - "stream = cuda.Stream()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "__Set up prediction function:__\n", - "\n", - "This involves a copy from CPU RAM to GPU VRAM, executing the model, then copying the results back from GPU VRAM to CPU RAM:" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": { - "id": "6R-F8JtV8qu-" - }, - "outputs": [], - "source": [ - "def predict(batch): # result gets copied into output\n", - " # Transfer input data to device\n", - " cuda.memcpy_htod_async(d_input, batch, stream)\n", - " # Execute model\n", - " context.execute_async_v2(bindings, stream.handle, None)\n", - " # Transfer predictions back\n", - " cuda.memcpy_dtoh_async(output, d_output, stream)\n", - " # Syncronize threads\n", - " stream.synchronize()\n", - " \n", - " return output" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "This is all we need to run predictions using our TensorRT engine in a Python runtime!" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": { - "id": "AdKZzW7O8qvB" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Warming up...\n", - "Done warming up!\n" - ] - } - ], - "source": [ - "print(\"Warming up...\")\n", - "\n", - "trt_predictions = predict(input_batch).astype(np.float32)\n", - "\n", - "print(\"Done warming up!\")" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Class | Probability (out of 1)\n" - ] - }, - { - "data": { - "text/plain": [ - "[(160, 0.3112793),\n", - " (169, 0.27026367),\n", - " (212, 0.17321777),\n", - " (170, 0.07165527),\n", - " (207, 0.033843994)]" - ] - }, - "execution_count": 11, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "indices = (-trt_predictions[0]).argsort()[:5]\n", - "print(\"Class | Probability (out of 1)\")\n", - "list(zip(indices, trt_predictions[0][indices]))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Note that we have recovered our same predictions as before!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Performance Comparison:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Last, we can see how quickly we can feed a singular batch to TensorRT, which we can compare to our original Tensorflow experiment from earlier." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We use the %%timeit Jupyter magic again. Note that %%timeit is fairly rough, and for any actual benchmarking better controlled testing is required - preferably outside of Jupyter." - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": { - "id": "XAtWnCK38qvD" - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "6.41 ms ± 846 ns per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "_ = predict(input_batch) # Check TRT performance" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Next Steps:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "

Profiling

\n", - "\n", - "This is a great next step for further optimizing and debugging models you are working on productionizing\n", - "\n", - "You can find it here: https://docs.nvidia.com/deeplearning/tensorrt/best-practices/index.html\n", - "\n", - "

TRT Dev Docs

\n", - "\n", - "Main documentation page for the ONNX, layer builder, C++, and legacy APIs\n", - "\n", - "You can find it here: https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html\n", - "\n", - "

TRT OSS GitHub

\n", - "\n", - "Contains OSS TRT components, sample applications, and plugin examples\n", - "\n", - "You can find it here: https://github.com/NVIDIA/TensorRT\n", - "\n", - "\n", - "#### TRT Supported Layers:\n", - "\n", - "https://github.com/NVIDIA/TensorRT/tree/main/samples/opensource/samplePlugin\n", - "\n", - "#### TRT ONNX Plugin Example:\n", - "\n", - "https://docs.nvidia.com/deeplearning/tensorrt/support-matrix/index.html#layers-precision-matrix\n" - ] - } - ], - "metadata": { - "accelerator": "GPU", - "colab": { - "name": "ONNXExample.ipynb", - "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 -} diff --git a/quickstart/IntroNotebooks/4. Using PyTorch through ONNX.ipynb b/quickstart/IntroNotebooks/4. Using PyTorch through ONNX.ipynb deleted file mode 100644 index b90f9d49..00000000 --- a/quickstart/IntroNotebooks/4. Using PyTorch through ONNX.ipynb +++ /dev/null @@ -1,992 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Using PyTorch with TensorRT through ONNX:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "TensorRT is a great way to take a trained PyTorch model and optimize it to run more efficiently during inference on an NVIDIA GPU.\n", - "\n", - "One approach to convert a PyTorch model to TensorRT is to export a PyTorch model to ONNX (an open format exchange for deep learning models) and then convert into a TensorRT engine. Essentially, we will follow this path to convert and deploy our model:\n", - "\n", - "![PyTorch+ONNX](./images/pytorch_onnx.png)\n", - "\n", - "Both TensorFlow and PyTorch models can be exported to ONNX, as well as many other frameworks. This allows models created using either framework to flow into common downstream pipelines.\n", - "\n", - "To get started, let's take a well-known computer vision model and follow five key steps to deploy it to the TensorRT Python runtime:\n", - "\n", - "1. __What format should I save my model in?__\n", - "2. __What batch size(s) am I running inference at?__\n", - "3. __What precision am I running inference at?__\n", - "4. __What TensorRT path am I using to convert my model?__\n", - "5. __What runtime am I targeting?__" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 1. What format should I save my model in?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We are going to use ResNet50, a widely used CNN architecture first described in this paper.\n", - "\n", - "Let's start by loading dependencies and downloading the model. We will also move our Resnet model onto the GPU and set it to evaluation mode." - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "Downloading: \"https://download.pytorch.org/models/resnet50-19c8e357.pth\" to /root/.cache/torch/hub/checkpoints/resnet50-19c8e357.pth\n" - ] - } - ], - "source": [ - "import torchvision.models as models\n", - "import torch\n", - "import torch.onnx\n", - "\n", - "# load the pretrained model\n", - "resnet50 = models.resnet50(pretrained=True, progress=False).eval()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "When saving a model to ONNX, PyTorch requires a test batch in proper shape and format. We pick a batch size:" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "BATCH_SIZE=32\n", - "\n", - "dummy_input=torch.randn(BATCH_SIZE, 3, 224, 224)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Next, we will export the model using the dummy input batch:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "# export the model to ONNX\n", - "torch.onnx.export(resnet50, dummy_input, \"resnet50_pytorch.onnx\", verbose=False)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Note that we are picking a BATCH_SIZE of 32 in this example." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Now Test with a Real Image:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's try a real image batch! For this example, we will simply repeat one open-source dog image from http://www.dog.ceo:" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(32, 224, 224, 3)" - ] - }, - "execution_count": 4, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "from skimage import io\n", - "from skimage.transform import resize\n", - "from matplotlib import pyplot as plt\n", - "import numpy as np\n", - "\n", - "url='https://images.dog.ceo/breeds/retriever-golden/n02099601_3004.jpg'\n", - "img = resize(io.imread(url), (224, 224))\n", - "img = np.expand_dims(np.array(img, dtype=np.float32), axis=0) # Expand image to have a batch dimension\n", - "input_batch = np.array(np.repeat(img, BATCH_SIZE, axis=0), dtype=np.float32) # Repeat across the batch dimension\n", - "\n", - "input_batch.shape" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "" - ] - }, - "execution_count": 5, - "metadata": {}, - "output_type": "execute_result" - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "plt.imshow(input_batch[0].astype(np.float32))" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "resnet50_gpu = models.resnet50(pretrained=True, progress=False).to(\"cuda\").eval()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We need to move our batch onto GPU and properly format it to shape [32, 3, 224, 224]. " - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "torch.Size([32, 3, 224, 224])" - ] - }, - "execution_count": 7, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "input_batch_chw = torch.from_numpy(input_batch).transpose(1,3).transpose(2,3)\n", - "input_batch_gpu = input_batch_chw.to(\"cuda\")\n", - "\n", - "input_batch_gpu.shape" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can run a prediction on a batch using .forward():" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(32, 1000)" - ] - }, - "execution_count": 8, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "with torch.no_grad():\n", - " predictions = np.array(resnet50_gpu(input_batch_gpu).cpu())\n", - "\n", - "predictions.shape" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Verify Baseline Model Performance/Accuracy:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For a baseline, lets time our prediction in FP32:" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "31.5 ms ± 72.1 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "with torch.no_grad():\n", - " preds = np.array(resnet50_gpu(input_batch_gpu).cpu())" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can also time FP16 precision performance:" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(32, 1000)" - ] - }, - "execution_count": 10, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "resnet50_gpu_half = resnet50_gpu.half()\n", - "input_half = input_batch_gpu.half()\n", - "\n", - "with torch.no_grad():\n", - " preds = np.array(resnet50_gpu_half(input_half).cpu()) # Warm Up\n", - " \n", - "preds.shape" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "19.4 ms ± 5.42 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "with torch.no_grad():\n", - " preds = np.array(resnet50_gpu_half(input_half).cpu())" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's also make sure our results are accurate. We will look at the top 5 accuracy on a single image prediction. The image we are using is of a Golden Retriever, which is class 207 in the ImageNet dataset our model was trained on." - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Class | Likelihood\n" - ] - }, - { - "data": { - "text/plain": [ - "[(207, 13.121688),\n", - " (208, 9.614037),\n", - " (257, 9.361297),\n", - " (205, 8.777787),\n", - " (160, 8.557351)]" - ] - }, - "execution_count": 12, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "indices = (-predictions[0]).argsort()[:5]\n", - "print(\"Class | Likelihood\")\n", - "list(zip(indices, predictions[0][indices]))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We have a model exported to ONNX and a baseline to compare against! Let's now take our ONNX model and convert it to a TensorRT inference engine." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Now, let's restart our Jupyter Kernel so PyTorch doesn't collide with TensorRT: " - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "import os\n", - "\n", - "os._exit(0) # Shut down all kernels so TRT doesn't fight with PyTorch for GPU memory" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 2. What batch size(s) am I running inference at?\n", - "\n", - "We are going to run with a fixed batch size of 32 for this example. Note that above we set BATCH_SIZE to 32 when saving our model to ONNX. We need to create another dummy batch of the same size (this time it will need to be in our target precision) to test out our engine.\n", - "\n", - "First, as before, we will set our BATCH_SIZE to 32. Note that our trtexec command above includes the '--explicitBatch' flag to signal to TensorRT that we will be using a fixed batch size at runtime." - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [], - "source": [ - "BATCH_SIZE = 32" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Importantly, by default TensorRT will use the input precision you give the runtime as the default precision for the rest of the network. So before we create our new dummy batch, we also need to choose a precision as in the next section:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 3. What precision am I running inference at?\n", - "\n", - "Remember that lower precisions than FP32 tend to run faster. There are two common reduced precision modes - FP16 and INT8. Graphics cards that are designed to do inference well often have an affinity for one of these two types. This guide was developed on an NVIDIA V100, which favors FP16, so we will use that here by default. INT8 is a more complicated process that requires a calibration step.\n", - "\n", - "__NOTE__: Make sure you use the same precision (USE_FP16) here you saved your model in above!" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "USE_FP16 = True\n", - "target_dtype = np.float16 if USE_FP16 else np.float32" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - " To create a test batch, we will once again repeat one open-source dog image from http://www.dog.ceo:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(32, 224, 224, 3)" - ] - }, - "execution_count": 3, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "from skimage import io\n", - "from skimage.transform import resize\n", - "from matplotlib import pyplot as plt\n", - "import numpy as np\n", - "\n", - "url='https://images.dog.ceo/breeds/retriever-golden/n02099601_3004.jpg'\n", - "img = resize(io.imread(url), (224, 224))\n", - "input_batch = np.array(np.repeat(np.expand_dims(np.array(img, dtype=np.float32), axis=0), BATCH_SIZE, axis=0), dtype=np.float32)\n", - "\n", - "input_batch.shape" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "" - ] - }, - "execution_count": 4, - "metadata": {}, - "output_type": "execute_result" - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "plt.imshow(input_batch[0].astype(np.float32))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Preprocess Images:\n", - "\n", - "PyTorch has a normalization that it applies by default in all of its pretrained vision models - we can preprocess our images to match this normalization by the following, making sure our final result is in FP16 precision:" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "import torch\n", - "from torchvision.transforms import Normalize\n", - "\n", - "def preprocess_image(img):\n", - " norm = Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n", - " result = norm(torch.from_numpy(img).transpose(0,2).transpose(1,2))\n", - " return np.array(result, dtype=np.float16)\n", - "\n", - "preprocessed_images = np.array([preprocess_image(image) for image in input_batch])" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 4. What TensorRT path am I using to convert my model?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can use trtexec, a command line tool for working with TensorRT, in order to convert an ONNX model originally from PyTorch to an engine file.\n", - "\n", - "Let's make sure we have TensorRT installed (this comes with trtexec):" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "import tensorrt" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "To convert the model we saved in the previous step, we need to point to the ONNX file, give trtexec a name to save the engine as, and last specify that we want to use a fixed batch size instead of a dynamic one." - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "&&&& RUNNING TensorRT.trtexec # trtexec --onnx=resnet50_pytorch.onnx --saveEngine=resnet_engine_pytorch.trt --explicitBatch --inputIOFormats=fp16:chw --outputIOFormats=fp16:chw --fp16\n", - "[06/09/2021-20:23:03] [I] === Model Options ===\n", - "[06/09/2021-20:23:03] [I] Format: ONNX\n", - "[06/09/2021-20:23:03] [I] Model: resnet50_pytorch.onnx\n", - "[06/09/2021-20:23:03] [I] Output:\n", - "[06/09/2021-20:23:03] [I] === Build Options ===\n", - "[06/09/2021-20:23:03] [I] Max batch: explicit\n", - "[06/09/2021-20:23:03] [I] Workspace: 16 MiB\n", - "[06/09/2021-20:23:03] [I] minTiming: 1\n", - "[06/09/2021-20:23:03] [I] avgTiming: 8\n", - "[06/09/2021-20:23:03] [I] Precision: FP32+FP16\n", - "[06/09/2021-20:23:03] [I] Calibration: \n", - "[06/09/2021-20:23:03] [I] Refit: Disabled\n", - "[06/09/2021-20:23:03] [I] Safe mode: Disabled\n", - "[06/09/2021-20:23:03] [I] Save engine: resnet_engine_pytorch.trt\n", - "[06/09/2021-20:23:03] [I] Load engine: \n", - "[06/09/2021-20:23:03] [I] Builder Cache: Enabled\n", - "[06/09/2021-20:23:03] [I] NVTX verbosity: 0\n", - "[06/09/2021-20:23:03] [I] Tactic sources: Using default tactic sources\n", - "[06/09/2021-20:23:03] [I] Input(s): fp16:chw\n", - "[06/09/2021-20:23:03] [I] Output(s): fp16:chw\n", - "[06/09/2021-20:23:03] [I] Input build shapes: model\n", - "[06/09/2021-20:23:03] [I] Input calibration shapes: model\n", - "[06/09/2021-20:23:03] [I] === System Options ===\n", - "[06/09/2021-20:23:03] [I] Device: 0\n", - "[06/09/2021-20:23:03] [I] DLACore: \n", - "[06/09/2021-20:23:03] [I] Plugins:\n", - "[06/09/2021-20:23:03] [I] === Inference Options ===\n", - "[06/09/2021-20:23:03] [I] Batch: Explicit\n", - "[06/09/2021-20:23:03] [I] Input inference shapes: model\n", - "[06/09/2021-20:23:03] [I] Iterations: 10\n", - "[06/09/2021-20:23:03] [I] Duration: 3s (+ 200ms warm up)\n", - "[06/09/2021-20:23:03] [I] Sleep time: 0ms\n", - "[06/09/2021-20:23:03] [I] Streams: 1\n", - "[06/09/2021-20:23:03] [I] ExposeDMA: Disabled\n", - "[06/09/2021-20:23:03] [I] Data transfers: Enabled\n", - "[06/09/2021-20:23:03] [I] Spin-wait: Disabled\n", - "[06/09/2021-20:23:03] [I] Multithreading: Disabled\n", - "[06/09/2021-20:23:03] [I] CUDA Graph: Disabled\n", - "[06/09/2021-20:23:03] [I] Separate profiling: Disabled\n", - "[06/09/2021-20:23:03] [I] Skip inference: Disabled\n", - "[06/09/2021-20:23:03] [I] Inputs:\n", - "[06/09/2021-20:23:03] [I] === Reporting Options ===\n", - "[06/09/2021-20:23:03] [I] Verbose: Disabled\n", - "[06/09/2021-20:23:03] [I] Averages: 10 inferences\n", - "[06/09/2021-20:23:03] [I] Percentile: 99\n", - "[06/09/2021-20:23:03] [I] Dump refittable layers:Disabled\n", - "[06/09/2021-20:23:03] [I] Dump output: Disabled\n", - "[06/09/2021-20:23:03] [I] Profile: Disabled\n", - "[06/09/2021-20:23:03] [I] Export timing to JSON file: \n", - "[06/09/2021-20:23:03] [I] Export output to JSON file: \n", - "[06/09/2021-20:23:03] [I] Export profile to JSON file: \n", - "[06/09/2021-20:23:03] [I] \n", - "[06/09/2021-20:23:04] [I] === Device Information ===\n", - "[06/09/2021-20:23:04] [I] Selected Device: Tesla V100-DGXS-16GB\n", - "[06/09/2021-20:23:04] [I] Compute Capability: 7.0\n", - "[06/09/2021-20:23:04] [I] SMs: 80\n", - "[06/09/2021-20:23:04] [I] Compute Clock Rate: 1.53 GHz\n", - "[06/09/2021-20:23:04] [I] Device Global Memory: 16155 MiB\n", - "[06/09/2021-20:23:04] [I] Shared Memory per SM: 96 KiB\n", - "[06/09/2021-20:23:04] [I] Memory Bus Width: 4096 bits (ECC enabled)\n", - "[06/09/2021-20:23:04] [I] Memory Clock Rate: 0.877 GHz\n", - "[06/09/2021-20:23:04] [I] \n", - "[06/09/2021-20:23:20] [I] [TRT] ----------------------------------------------------------------\n", - "[06/09/2021-20:23:20] [I] [TRT] Input filename: resnet50_pytorch.onnx\n", - "[06/09/2021-20:23:20] [I] [TRT] ONNX IR version: 0.0.6\n", - "[06/09/2021-20:23:20] [I] [TRT] Opset version: 9\n", - "[06/09/2021-20:23:20] [I] [TRT] Producer name: pytorch\n", - "[06/09/2021-20:23:20] [I] [TRT] Producer version: 1.9\n", - "[06/09/2021-20:23:20] [I] [TRT] Domain: \n", - "[06/09/2021-20:23:20] [I] [TRT] Model version: 0\n", - "[06/09/2021-20:23:20] [I] [TRT] Doc string: \n", - "[06/09/2021-20:23:20] [I] [TRT] ----------------------------------------------------------------\n", - "[06/09/2021-20:23:24] [I] [TRT] Some tactics do not have sufficient workspace memory to run. Increasing workspace size may increase performance, please check verbose output.\n", - "[06/09/2021-20:24:49] [I] [TRT] Detected 1 inputs and 1 output network tensors.\n", - "[06/09/2021-20:24:49] [I] Engine built in 105.672 sec.\n", - "[06/09/2021-20:24:50] [I] Starting inference\n", - "[06/09/2021-20:24:53] [I] Warmup completed 0 queries over 200 ms\n", - "[06/09/2021-20:24:53] [I] Timing trace has 0 queries over 2.9909 s\n", - "[06/09/2021-20:24:53] [I] Trace averages of 10 runs:\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35326 ms - Host latency: 6.18286 ms (end to end 10.1932 ms, enqueue 0.460231 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35654 ms - Host latency: 6.19131 ms (end to end 10.2018 ms, enqueue 0.473865 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.38982 ms - Host latency: 6.22551 ms (end to end 10.2071 ms, enqueue 0.460098 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.3761 ms - Host latency: 6.24244 ms (end to end 10.2638 ms, enqueue 0.456512 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.36218 ms - Host latency: 6.22775 ms (end to end 9.37773 ms, enqueue 0.441846 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35991 ms - Host latency: 6.22073 ms (end to end 9.77996 ms, enqueue 0.443829 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.38082 ms - Host latency: 6.25148 ms (end to end 10.0299 ms, enqueue 0.44693 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.39341 ms - Host latency: 6.26748 ms (end to end 10.0738 ms, enqueue 0.456384 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.38766 ms - Host latency: 6.26089 ms (end to end 10.2009 ms, enqueue 0.461377 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.37385 ms - Host latency: 6.24359 ms (end to end 9.65547 ms, enqueue 0.442078 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35819 ms - Host latency: 6.21615 ms (end to end 8.21369 ms, enqueue 0.436646 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34844 ms - Host latency: 6.20999 ms (end to end 9.77367 ms, enqueue 0.433765 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35132 ms - Host latency: 6.21758 ms (end to end 10.6213 ms, enqueue 0.435864 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.36421 ms - Host latency: 6.23065 ms (end to end 10.5457 ms, enqueue 0.436438 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.39054 ms - Host latency: 6.25834 ms (end to end 10.4534 ms, enqueue 0.444727 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.36874 ms - Host latency: 6.23105 ms (end to end 8.89895 ms, enqueue 0.443665 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35729 ms - Host latency: 6.21859 ms (end to end 8.51741 ms, enqueue 0.437866 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.33851 ms - Host latency: 6.19753 ms (end to end 9.1334 ms, enqueue 0.438574 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34199 ms - Host latency: 6.21041 ms (end to end 10.6064 ms, enqueue 0.44613 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.33002 ms - Host latency: 6.20233 ms (end to end 10.5858 ms, enqueue 0.458911 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.38256 ms - Host latency: 6.25411 ms (end to end 9.77722 ms, enqueue 0.460205 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.3837 ms - Host latency: 6.2543 ms (end to end 9.4882 ms, enqueue 0.448364 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.35146 ms - Host latency: 6.20986 ms (end to end 8.36691 ms, enqueue 0.434412 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34351 ms - Host latency: 6.20732 ms (end to end 10.1922 ms, enqueue 0.439209 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.3502 ms - Host latency: 6.21951 ms (end to end 10.6236 ms, enqueue 0.451489 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34368 ms - Host latency: 6.21904 ms (end to end 10.4949 ms, enqueue 0.462231 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.33777 ms - Host latency: 6.21189 ms (end to end 9.99021 ms, enqueue 0.455859 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.33193 ms - Host latency: 6.19707 ms (end to end 9.02058 ms, enqueue 0.445972 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.33115 ms - Host latency: 6.19114 ms (end to end 9.11257 ms, enqueue 0.433862 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34673 ms - Host latency: 6.21465 ms (end to end 10.6074 ms, enqueue 0.442139 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.38572 ms - Host latency: 6.25532 ms (end to end 10.3253 ms, enqueue 0.446631 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.36335 ms - Host latency: 6.23845 ms (end to end 10.6406 ms, enqueue 0.45625 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.36877 ms - Host latency: 6.24153 ms (end to end 10.2023 ms, enqueue 0.449341 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.36023 ms - Host latency: 6.21748 ms (end to end 8.45557 ms, enqueue 0.436719 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34392 ms - Host latency: 6.20728 ms (end to end 10.1899 ms, enqueue 0.438428 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.34636 ms - Host latency: 6.21821 ms (end to end 10.6184 ms, enqueue 0.447217 ms)\n", - "[06/09/2021-20:24:53] [I] Average on 10 runs - GPU latency: 5.33555 ms - Host latency: 6.20952 ms (end to end 10.5899 ms, enqueue 0.459546 ms)\n", - "[06/09/2021-20:24:53] [I] Host Latency\n", - "[06/09/2021-20:24:53] [I] min: 6.16092 ms (end to end 6.17383 ms)\n", - "[06/09/2021-20:24:53] [I] max: 6.2887 ms (end to end 10.8184 ms)\n", - "[06/09/2021-20:24:53] [I] mean: 6.22352 ms (end to end 9.90214 ms)\n", - "[06/09/2021-20:24:53] [I] median: 6.22021 ms (end to end 10.6108 ms)\n", - "[06/09/2021-20:24:53] [I] percentile: 6.28583 ms at 99% (end to end 10.7902 ms at 99%)\n", - "[06/09/2021-20:24:53] [I] throughput: 0 qps\n", - "[06/09/2021-20:24:53] [I] walltime: 2.9909 s\n", - "[06/09/2021-20:24:53] [I] Enqueue Time\n", - "[06/09/2021-20:24:53] [I] min: 0.424072 ms\n", - "[06/09/2021-20:24:53] [I] max: 0.49585 ms\n", - "[06/09/2021-20:24:53] [I] median: 0.445618 ms\n", - "[06/09/2021-20:24:53] [I] GPU Compute\n", - "[06/09/2021-20:24:53] [I] min: 5.30127 ms\n", - "[06/09/2021-20:24:53] [I] max: 5.42108 ms\n", - "[06/09/2021-20:24:53] [I] mean: 5.35895 ms\n", - "[06/09/2021-20:24:53] [I] median: 5.35571 ms\n", - "[06/09/2021-20:24:53] [I] percentile: 5.41693 ms at 99%\n", - "[06/09/2021-20:24:53] [I] total compute time: 2.00961 s\n", - "&&&& PASSED TensorRT.trtexec # trtexec --onnx=resnet50_pytorch.onnx --saveEngine=resnet_engine_pytorch.trt --explicitBatch --inputIOFormats=fp16:chw --outputIOFormats=fp16:chw --fp16\n" - ] - } - ], - "source": [ - "# step out of Python for a moment to convert the ONNX model to a TRT engine using trtexec\n", - "if USE_FP16:\n", - " !trtexec --onnx=resnet50_pytorch.onnx --saveEngine=resnet_engine_pytorch.trt --explicitBatch --inputIOFormats=fp16:chw --outputIOFormats=fp16:chw --fp16\n", - "else:\n", - " !trtexec --onnx=resnet50_pytorch.onnx --saveEngine=resnet_engine_pytorch.trt --explicitBatch" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "This will save our model as 'resnet_engine.trt'." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 5. What TensorRT runtime am I targeting?\n", - "\n", - "Now, we have a converted our model to a TensorRT engine. Great! That means we are ready to load it into the native Python TensorRT runtime. This runtime strikes a balance between the ease of use of the high level Python APIs used in frameworks and the fast, low level C++ runtimes available in TensorRT." - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "CPU times: user 15.9 s, sys: 556 ms, total: 16.5 s\n", - "Wall time: 19.3 s\n" - ] - } - ], - "source": [ - "%%time\n", - "\n", - "import tensorrt as trt\n", - "import pycuda.driver as cuda\n", - "import pycuda.autoinit\n", - "\n", - "f = open(\"resnet_engine_pytorch.trt\", \"rb\")\n", - "runtime = trt.Runtime(trt.Logger(trt.Logger.WARNING)) \n", - "\n", - "engine = runtime.deserialize_cuda_engine(f.read())\n", - "context = engine.create_execution_context()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Now allocate input and output memory, give TRT pointers (bindings) to it:" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "# need to set input and output precisions to FP16 to fully enable it\n", - "output = np.empty([BATCH_SIZE, 1000], dtype = target_dtype) \n", - "\n", - "# allocate device memory\n", - "d_input = cuda.mem_alloc(1 * input_batch.nbytes)\n", - "d_output = cuda.mem_alloc(1 * output.nbytes)\n", - "\n", - "bindings = [int(d_input), int(d_output)]\n", - "\n", - "stream = cuda.Stream()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Next, set up the prediction function.\n", - "\n", - "This involves a copy from CPU RAM to GPU VRAM, executing the model, then copying the results back from GPU VRAM to CPU RAM:" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [], - "source": [ - "def predict(batch): # result gets copied into output\n", - " # transfer input data to device\n", - " cuda.memcpy_htod_async(d_input, batch, stream)\n", - " # execute model\n", - " context.execute_async_v2(bindings, stream.handle, None)\n", - " # transfer predictions back\n", - " cuda.memcpy_dtoh_async(output, d_output, stream)\n", - " # syncronize threads\n", - " stream.synchronize()\n", - " \n", - " return output" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's time the function!" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Warming up...\n", - "Done warming up!\n" - ] - } - ], - "source": [ - "print(\"Warming up...\")\n", - "\n", - "pred = predict(preprocessed_images)\n", - "\n", - "print(\"Done warming up!\")" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "6.28 ms ± 1.07 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "pred = predict(preprocessed_images)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Finally we should verify our TensorRT output is still accurate." - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Class | Probability (out of 1)\n" - ] - }, - { - "data": { - "text/plain": [ - "[(207, 12.44), (208, 7.508), (220, 7.492), (160, 7.426), (226, 7.383)]" - ] - }, - "execution_count": 15, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "indices = (-pred[0]).argsort()[:5]\n", - "print(\"Class | Probability (out of 1)\")\n", - "list(zip(indices, pred[0][indices]))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Look for ImageNet indices 150-275 above, where 207 is the ground truth correct class (Golden Retriever). Compare with the results of the original unoptimized model in the first section!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Next Steps:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "

Profiling

\n", - "\n", - "This is a great next step for further optimizing and debugging models you are working on productionizing\n", - "\n", - "You can find it here: https://docs.nvidia.com/deeplearning/tensorrt/best-practices/index.html\n", - "\n", - "

TRT Dev Docs

\n", - "\n", - "Main documentation page for the ONNX, layer builder, C++, and legacy APIs\n", - "\n", - "You can find it here: https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html\n", - "\n", - "

TRT OSS GitHub

\n", - "\n", - "Contains OSS TRT components, sample applications, and plugin examples\n", - "\n", - "You can find it here: https://github.com/NVIDIA/TensorRT\n", - "\n", - "\n", - "#### TRT Supported Layers:\n", - "\n", - "https://github.com/NVIDIA/TensorRT/tree/main/samples/opensource/samplePlugin\n", - "\n", - "#### TRT ONNX Plugin Example:\n", - "\n", - "https://docs.nvidia.com/deeplearning/tensorrt/support-matrix/index.html#layers-precision-matrix" - ] - } - ], - "metadata": { - "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": "3.8.8" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/quickstart/IntroNotebooks/5. Understanding TensorRT Runtimes.ipynb b/quickstart/IntroNotebooks/5. Understanding TensorRT Runtimes.ipynb deleted file mode 100644 index 05e8b3df..00000000 --- a/quickstart/IntroNotebooks/5. Understanding TensorRT Runtimes.ipynb +++ /dev/null @@ -1,107 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Runtimes: What are my options? How do I choose?\n", - "\n", - "Remember that TensorRT consists of two main components - __1. A series of parsers and integrations__ to convert your model to an optimized engine and __2. An series of TensorRT runtime APIs__ with several associated tools for deployment.\n", - "\n", - "In this notebook, we will focus on the latter - various runtime options for TensorRT engines.\n", - "\n", - "The runtimes have different use cases for running TRT engines. \n", - "\n", - "### Considerations when picking a runtime:\n", - "\n", - "Generally speaking, there are a few major considerations when picking a runtime:\n", - "- __Framework__ - Some options, like TF-TRT, are only relevant to Tensorflow\n", - "- __Time-to-solution__ - TF-TRT is much more likely to work 'out-of-the-box' if a quick solution is required and ONNX fails\n", - "- __Serving needs__ - TF-TRT can use TF Serving to serve models over HTTP as a simple solution. For other frameworks (or for more advanced features) TRITON is framework agnostic, allows for concurrent model execution or multiple copies within a GPU to reduce latency, and can accept engines created through both the ONNX and TF-TRT paths\n", - "- __Performance__ - Different TensorRT runtimes offer varying levels of performance. For example, TF-TRT is generally going to be slower than using ONNX or the C++ API directly.\n", - "\n", - "### Python API:\n", - "\n", - "__Use this when:__\n", - "- You can accept some performance overhead, and\n", - "- You are most familiar with Python, or\n", - "- You are performing initial debugging and testing with TRT\n", - "\n", - "__More info:__\n", - "\n", - " \n", - "The [TensorRT Python API](https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html#perform_inference_python) gives you fine-grained control over the execution of your engine using a Python interface. It makes memory allocation, kernel execution, and copies to and from the GPU explicit - which can make integration into high performance applications easier. It is also great for testing models in a Python environment - such as in a Jupyter notebook.\n", - " \n", - "The [ONNX notebook for Tensorflow](./3.%20Using%20Tensorflow%202%20through%20ONNX.ipynb) and [for PyTorch](./4.%20Using%20PyTorch%20through%20ONNX.ipynb) are good examples of using TensorRT to get great performance while staying in Python\n", - "\n", - "### C++ API: \n", - "\n", - "__Use this when:__\n", - "- You want the least amount of overhead possible to maximize the performance of your models and achieve better latency\n", - "- You are not using TF-TRT (though TF-TRT graph conversions that only generate a single engine can still be exported to C++)\n", - "- You are most familiar with C++\n", - "- You want to optimize your inference pipeline as much as possible\n", - "\n", - "__More info:__\n", - "\n", - "The [TensorRT C++ API](https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html#perform_inference_c) gives you fine-grained control over the execution of your engine using a C++ interface. It makes memory allocation, kernel execution, and copies to and from the GPU explicit - which can make integration into high performance C++ applications easier. The C++ API is generally the most performant option for running TensorRT engines, with the least overhead.\n", - "\n", - "[This NVIDIA Developer blog](https://developer.nvidia.com/blog/speed-up-inference-tensorrt/) is a good example of taking an ONNX model and running it with dynamic batch size support using the C++ API.\n", - "\n", - "\n", - "### Tensorflow/TF-TRT Runtime: (Tensorflow Only) \n", - " \n", - "__Use this when:__\n", - " \n", - "- You are using TF-TRT, and\n", - "- Your model converts to more than one TensorRT engine\n", - "\n", - "__More info:__\n", - "\n", - "\n", - "TF-TRT is the standard runtime used with models that were converted in TF-TRT. It works by taking groups of nodes at once in the Tensorflow graph, and replacing them with a singular optimized engine that calls the TensorRT Python API behind the scenes. This optimized engine is in the form of a Tensorflow operation - which means that your graph is still in Tensorflow and will essentially function like any other Tensorflow model. For example, it can be a useful exercise to take a look at your model in Tensorboard to validate which nodes TensorRT was able to optimize.\n", - "\n", - "If your graph entirely converts to a single TF-TRT engine, it can be more efficient to export the engine node and run it using one of the other APIs. You can find instructions to do this in the [TF-TRT documentation](https://docs.nvidia.com/deeplearning/frameworks/tf-trt-user-guide/index.html#tensorrt-plan).\n", - "\n", - "As an example, the TF-TRT notebooks included with this guide use the TF-TRT runtime.\n", - "\n", - "### TRITON Inference Server\n", - "\n", - "__Use this when:__\n", - "- You want to serve your models over HTTP or gRPC\n", - "- You want to load balance across multiple models or copies of models across GPUs to minimze latency and make better use of the GPU\n", - "- You want to have multiple models running efficiently on a single GPU at the same time\n", - "- You want to serve a variety of models converted using a variety of converters and frameworks (including TF-TRT and ONNX) through a uniform interface\n", - "- You need serving support but are using PyTorch, another framework, or the ONNX path in general\n", - "\n", - "__More info:__\n", - "\n", - "\n", - "TRITON is an open source inference serving software that lets teams deploy trained AI models from any framework (TensorFlow, TensorRT, PyTorch, ONNX Runtime, or a custom framework), from local storage or Google Cloud Platform or AWS S3 on any GPU- or CPU-based infrastructure (cloud, data center, or edge). It is a flexible project with several unique features - such as concurrent model execution of both heterogeneous models and multiple copies of the same model (multiple model copies can reduce latency further) as well as load balancing and model analysis. It is a good option if you need to serve your models over HTTP - such as in a cloud inferencing solution.\n", - " \n", - "You can find the TRITON home page [here](https://developer.nvidia.com/nvidia-triton-inference-server), and the documentation [here](https://docs.nvidia.com/deeplearning/triton-inference-server/user-guide/docs/)." - ] - } - ], - "metadata": { - "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": "3.6.9" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/quickstart/IntroNotebooks/Additional Examples/1. TF-TRT Classification.ipynb b/quickstart/IntroNotebooks/Additional Examples/1. TF-TRT Classification.ipynb deleted file mode 100644 index 4fa0bab1..00000000 --- a/quickstart/IntroNotebooks/Additional Examples/1. TF-TRT Classification.ipynb +++ /dev/null @@ -1,711 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# TF-TRT Keras Classification Examples:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In this notebook, we cover a variety of classification base networks pulled from the tensorflow.keras.applications project!\n", - "\n", - "This demonstrates TF-TRT working on a variety of model architectures out of the box. This is a great way to demonstrate the ease of use of TF-TRT. TF-TRT can still optimize parts of your network even if it contains layers that are not supported by TensorRT itself. This makes it easy to get a first-pass at an optimized model - as we will demonstrate here." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's make sure our GPUs are properly configured and visible:" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Fri Jan 29 22:55:18 2021 \n", - "+-----------------------------------------------------------------------------+\n", - "| NVIDIA-SMI 450.80.02 Driver Version: 450.80.02 CUDA Version: 11.1 |\n", - "|-------------------------------+----------------------+----------------------+\n", - "| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n", - "| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n", - "| | | MIG M. |\n", - "|===============================+======================+======================|\n", - "| 0 Tesla V100-DGXS... On | 00000000:07:00.0 Off | 0 |\n", - "| N/A 43C P0 62W / 300W | 125MiB / 16155MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 1 Tesla V100-DGXS... On | 00000000:08:00.0 Off | 0 |\n", - "| N/A 42C P0 38W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 2 Tesla V100-DGXS... On | 00000000:0E:00.0 Off | 0 |\n", - "| N/A 41C P0 38W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 3 Tesla V100-DGXS... On | 00000000:0F:00.0 Off | 0 |\n", - "| N/A 42C P0 37W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - " \n", - "+-----------------------------------------------------------------------------+\n", - "| Processes: |\n", - "| GPU GI CI PID Type Process name GPU Memory |\n", - "| ID ID Usage |\n", - "|=============================================================================|\n", - "+-----------------------------------------------------------------------------+\n" - ] - } - ], - "source": [ - "!nvidia-smi" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Remember to sucessfully deploy a TensorRT model, you have to make __five key decisions__:\n", - "\n", - "1. __What format should I save my model in?__\n", - "2. __What batch size(s) am I running inference at?__\n", - "3. __What precision am I running inference at?__\n", - "4. __What TensorRT path am I using to convert my model?__\n", - "5. __What runtime am I targeting?__\n", - "\n", - "Let's get to it!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 1. What format should I save my model in?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "TF-TRT requires SavedModel format in Tensorflow 2.x:" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "!mkdir -p tmp_savedmodels" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "downloading and initializing models...\n", - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels.h5\n", - "553467904/553467096 [==============================] - 73s 0us/step\n", - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/inception_v3/inception_v3_weights_tf_dim_ordering_tf_kernels.h5\n", - "96116736/96112376 [==============================] - 3s 0us/step\n", - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/xception/xception_weights_tf_dim_ordering_tf_kernels.h5\n", - "91889664/91884032 [==============================] - 5s 0us/step\n", - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet_v2/mobilenet_v2_weights_tf_dim_ordering_tf_kernels_1.0_224.h5\n", - "14540800/14536120 [==============================] - 1s 0us/step\n", - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/densenet/densenet121_weights_tf_dim_ordering_tf_kernels.h5\n", - "33193984/33188688 [==============================] - 1s 0us/step\n", - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50v2_weights_tf_dim_ordering_tf_kernels.h5\n", - "102875136/102869336 [==============================] - 10s 0us/step\n", - "saving ...\n", - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Model.state_updates (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Layer.updates (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/0/assets\n", - "finished saving!\n", - "saving ...\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/1/assets\n", - "finished saving!\n", - "saving ...\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/2/assets\n", - "finished saving!\n", - "saving ...\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/3/assets\n", - "finished saving!\n", - "saving ...\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/4/assets\n", - "finished saving!\n", - "saving ...\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/5/assets\n", - "finished saving!\n", - "saving ...\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/6/assets\n", - "finished saving!\n" - ] - } - ], - "source": [ - "from tensorflow.keras.applications import ResNet50, VGG16, InceptionV3, Xception, MobileNetV2, DenseNet121, ResNet50V2\n", - "\n", - "print(\"Downloading and initializing models...\")\n", - "models = [ResNet50, VGG16, InceptionV3, Xception, MobileNetV2, DenseNet121, ResNet50V2]\n", - "models = [model(include_top=True, weights='imagenet') for model in models]\n", - "\n", - "model_dirs = []\n", - "for idx, model in enumerate(models):\n", - " print(\"Saving\", model,\"...\")\n", - " model_dir = 'tmp_savedmodels/%s' % idx\n", - " model_dirs.append(model_dir)\n", - " model.save(model_dir) \n", - " print(\"Finished saving!\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 2. What batch size(s) am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will use a batch size of 32 for all models:" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "BATCH_SIZE = 32" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We create a series of randomized \"dummy\" batches to test our model on:" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "dummy_input_batch = lambda x: np.zeros((BATCH_SIZE, x, x, 3))\n", - "\n", - "dummy_inputs = [224, 224, 299, 299, 224, 224, 224]\n", - "dummy_inputs = [dummy_input_batch(size) for size in dummy_inputs]" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Last, we \"warm up\" all of our models so their one time start-up costs aren't throw off any of our Jupyter magic %%timeit timer calls:" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "WARNING:tensorflow:5 out of the last 5 calls to .predict_function at 0x7f59d856e488> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/tutorials/customization/performance#python_or_tensor_args and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", - "WARNING:tensorflow:6 out of the last 6 calls to .predict_function at 0x7f5b3820d6a8> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/tutorials/customization/performance#python_or_tensor_args and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", - "WARNING:tensorflow:7 out of the last 7 calls to .predict_function at 0x7f596ca93950> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/tutorials/customization/performance#python_or_tensor_args and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n" - ] - } - ], - "source": [ - "# Warm up:\n", - "for idx, model in enumerate(models):\n", - " model.predict(dummy_inputs[idx])" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 3. What precision am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will leave it as the default:" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [], - "source": [ - "PRECISION = \"FP32\"" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 4. What TensorRT tool or integration am I using to convert my model?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will be using TF-TRT through the ModelOptimizer example wrapper used in this guide:" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Starting resnet50 tmp_savedmodels/0\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/0_FP32/assets\n", - "[[1.6964252e-04 3.3007402e-04 6.1350249e-05 ... 1.4622317e-05\n", - " 1.4449877e-04 6.6086568e-04]\n", - " [1.6964252e-04 3.3007402e-04 6.1350249e-05 ... 1.4622317e-05\n", - " 1.4449877e-04 6.6086568e-04]\n", - " [1.6964252e-04 3.3007402e-04 6.1350249e-05 ... 1.4622317e-05\n", - " 1.4449877e-04 6.6086568e-04]\n", - " ...\n", - " [1.6964252e-04 3.3007402e-04 6.1350249e-05 ... 1.4622317e-05\n", - " 1.4449877e-04 6.6086568e-04]\n", - " [1.6964252e-04 3.3007402e-04 6.1350249e-05 ... 1.4622317e-05\n", - " 1.4449877e-04 6.6086568e-04]\n", - " [1.6964252e-04 3.3007402e-04 6.1350249e-05 ... 1.4622317e-05\n", - " 1.4449877e-04 6.6086568e-04]]\n", - "Finished!\n", - "\n", - "Starting vgg16 tmp_savedmodels/1\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_1_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/1_FP32/assets\n", - "[[0.00022801 0.00222478 0.00050746 ... 0.00011863 0.00026599 0.01312881]\n", - " [0.00022801 0.00222478 0.00050746 ... 0.00011863 0.00026599 0.01312881]\n", - " [0.00022801 0.00222478 0.00050746 ... 0.00011863 0.00026599 0.01312881]\n", - " ...\n", - " [0.00022801 0.00222478 0.00050746 ... 0.00011863 0.00026599 0.01312881]\n", - " [0.00022801 0.00222478 0.00050746 ... 0.00011863 0.00026599 0.01312881]\n", - " [0.00022801 0.00222478 0.00050746 ... 0.00011863 0.00026599 0.01312881]]\n", - "Finished!\n", - "\n", - "Starting inception_v3 tmp_savedmodels/2\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_2_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/2_FP32/assets\n", - "[[0.00043102 0.00033233 0.0002535 ... 0.00012701 0.00023254 0.00082577]\n", - " [0.00043102 0.00033233 0.0002535 ... 0.00012701 0.00023254 0.00082577]\n", - " [0.00043102 0.00033233 0.0002535 ... 0.00012701 0.00023254 0.00082577]\n", - " ...\n", - " [0.00043102 0.00033233 0.0002535 ... 0.00012701 0.00023254 0.00082577]\n", - " [0.00043102 0.00033233 0.0002535 ... 0.00012701 0.00023254 0.00082577]\n", - " [0.00043102 0.00033233 0.0002535 ... 0.00012701 0.00023254 0.00082577]]\n", - "Finished!\n", - "\n", - "Starting xception tmp_savedmodels/3\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_3_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/3_FP32/assets\n", - "[[0.00022673 0.00034859 0.00021873 ... 0.00012943 0.00032854 0.00086526]\n", - " [0.00022673 0.00034859 0.00021873 ... 0.00012943 0.00032854 0.00086526]\n", - " [0.00022673 0.00034859 0.00021873 ... 0.00012943 0.00032854 0.00086526]\n", - " ...\n", - " [0.00022673 0.00034859 0.00021873 ... 0.00012943 0.00032854 0.00086526]\n", - " [0.00022673 0.00034859 0.00021873 ... 0.00012943 0.00032854 0.00086526]\n", - " [0.00022673 0.00034859 0.00021873 ... 0.00012943 0.00032854 0.00086526]]\n", - "Finished!\n", - "\n", - "Starting mobilenetv2_1.00_224 tmp_savedmodels/4\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_4_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/4_FP32/assets\n", - "[[1.8110585e-04 6.4528472e-04 6.8695762e-04 ... 7.9570833e-05\n", - " 1.3486181e-04 3.3463116e-03]\n", - " [1.8110585e-04 6.4528472e-04 6.8695762e-04 ... 7.9570833e-05\n", - " 1.3486181e-04 3.3463116e-03]\n", - " [1.8110585e-04 6.4528472e-04 6.8695762e-04 ... 7.9570833e-05\n", - " 1.3486181e-04 3.3463116e-03]\n", - " ...\n", - " [1.8110585e-04 6.4528472e-04 6.8695762e-04 ... 7.9570833e-05\n", - " 1.3486181e-04 3.3463116e-03]\n", - " [1.8110585e-04 6.4528472e-04 6.8695762e-04 ... 7.9570833e-05\n", - " 1.3486181e-04 3.3463116e-03]\n", - " [1.8110585e-04 6.4528472e-04 6.8695762e-04 ... 7.9570833e-05\n", - " 1.3486181e-04 3.3463116e-03]]\n", - "Finished!\n", - "\n", - "Starting densenet121 tmp_savedmodels/5\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_5_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/5_FP32/assets\n", - "[[2.3581024e-04 3.7533988e-04 1.1308040e-04 ... 5.6219425e-05\n", - " 2.6299071e-04 1.1581751e-03]\n", - " [2.3581024e-04 3.7533988e-04 1.1308040e-04 ... 5.6219425e-05\n", - " 2.6299071e-04 1.1581751e-03]\n", - " [2.3581024e-04 3.7533988e-04 1.1308040e-04 ... 5.6219425e-05\n", - " 2.6299071e-04 1.1581751e-03]\n", - " ...\n", - " [2.3581024e-04 3.7533988e-04 1.1308040e-04 ... 5.6219425e-05\n", - " 2.6299071e-04 1.1581751e-03]\n", - " [2.3581024e-04 3.7533988e-04 1.1308040e-04 ... 5.6219425e-05\n", - " 2.6299071e-04 1.1581751e-03]\n", - " [2.3581024e-04 3.7533988e-04 1.1308040e-04 ... 5.6219425e-05\n", - " 2.6299071e-04 1.1581751e-03]]\n", - "Finished!\n", - "\n", - "Starting resnet50v2 tmp_savedmodels/6\n", - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_6_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/6_FP32/assets\n", - "[[0.00082353 0.00079469 0.00060477 ... 0.00036948 0.00069747 0.00154858]\n", - " [0.00082353 0.00079469 0.00060477 ... 0.00036948 0.00069747 0.00154858]\n", - " [0.00082353 0.00079469 0.00060477 ... 0.00036948 0.00069747 0.00154858]\n", - " ...\n", - " [0.00082353 0.00079469 0.00060477 ... 0.00036948 0.00069747 0.00154858]\n", - " [0.00082353 0.00079469 0.00060477 ... 0.00036948 0.00069747 0.00154858]\n", - " [0.00082353 0.00079469 0.00060477 ... 0.00036948 0.00069747 0.00154858]]\n", - "Finished!\n", - "\n" - ] - } - ], - "source": [ - "from helper import ModelOptimizer\n", - "\n", - "opt_models = []\n", - "for model_class, model, dummy in zip(models, model_dirs, dummy_inputs):\n", - " print(\"Starting\", model_class._name, model)\n", - " model_opt = ModelOptimizer(model)\n", - " opt_trt = model_opt.convert(model+'_'+PRECISION, precision=PRECISION)\n", - "\n", - " print(opt_trt.predict(dummy))\n", - " \n", - " opt_models.append(opt_trt)\n", - " \n", - " print(\"Finished!\\n\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 5. What TensorRT runtime am I targeting?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will stay inside our Tensorflow/Python runtime:" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "array([[0.00082353, 0.00079469, 0.00060477, ..., 0.00036948, 0.00069747,\n", - " 0.00154858],\n", - " [0.00082353, 0.00079469, 0.00060477, ..., 0.00036948, 0.00069747,\n", - " 0.00154858],\n", - " [0.00082353, 0.00079469, 0.00060477, ..., 0.00036948, 0.00069747,\n", - " 0.00154858],\n", - " ...,\n", - " [0.00082353, 0.00079469, 0.00060477, ..., 0.00036948, 0.00069747,\n", - " 0.00154858],\n", - " [0.00082353, 0.00079469, 0.00060477, ..., 0.00036948, 0.00069747,\n", - " 0.00154858],\n", - " [0.00082353, 0.00079469, 0.00060477, ..., 0.00036948, 0.00069747,\n", - " 0.00154858]], dtype=float32)" - ] - }, - "execution_count": 9, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "opt_models[idx].predict(dummy_inputs[idx])" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Performance Comparisons:" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [], - "source": [ - "idx = 0 #resnet" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "CPU times: user 160 ms, sys: 5.52 ms, total: 166 ms\n", - "Wall time: 148 ms\n" - ] - }, - { - "data": { - "text/plain": [ - "array([[1.69642386e-04, 3.30075040e-04, 6.13506127e-05, ...,\n", - " 1.46224065e-05, 1.44499005e-04, 6.60870341e-04],\n", - " [1.69642386e-04, 3.30075040e-04, 6.13506127e-05, ...,\n", - " 1.46224065e-05, 1.44499005e-04, 6.60870341e-04],\n", - " [1.69642386e-04, 3.30075040e-04, 6.13506127e-05, ...,\n", - " 1.46224065e-05, 1.44499005e-04, 6.60870341e-04],\n", - " ...,\n", - " [1.69642386e-04, 3.30075040e-04, 6.13506127e-05, ...,\n", - " 1.46224065e-05, 1.44499005e-04, 6.60870341e-04],\n", - " [1.69642386e-04, 3.30075040e-04, 6.13506127e-05, ...,\n", - " 1.46224065e-05, 1.44499005e-04, 6.60870341e-04],\n", - " [1.69642386e-04, 3.30075040e-04, 6.13506127e-05, ...,\n", - " 1.46224065e-05, 1.44499005e-04, 6.60870341e-04]], dtype=float32)" - ] - }, - "execution_count": 11, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "%%time\n", - "\n", - "models[idx].predict(dummy_inputs[idx])" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "CPU times: user 30.2 ms, sys: 8.3 ms, total: 38.5 ms\n", - "Wall time: 36.6 ms\n" - ] - }, - { - "data": { - "text/plain": [ - "array([[1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " ...,\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04],\n", - " [1.6964252e-04, 3.3007402e-04, 6.1350249e-05, ..., 1.4622317e-05,\n", - " 1.4449877e-04, 6.6086568e-04]], dtype=float32)" - ] - }, - "execution_count": 12, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "%%time\n", - "\n", - "opt_models[idx].predict(dummy_inputs[idx])" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [], - "source": [ - "idx = -3 # mobilenets" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "CPU times: user 105 ms, sys: 14.4 ms, total: 120 ms\n", - "Wall time: 63.5 ms\n" - ] - }, - { - "data": { - "text/plain": [ - "array([[1.8110899e-04, 6.4530974e-04, 6.8695901e-04, ..., 7.9570033e-05,\n", - " 1.3486811e-04, 3.3462986e-03],\n", - " [1.8110899e-04, 6.4530974e-04, 6.8695901e-04, ..., 7.9570033e-05,\n", - " 1.3486811e-04, 3.3462986e-03],\n", - " [1.8110899e-04, 6.4530974e-04, 6.8695901e-04, ..., 7.9570033e-05,\n", - " 1.3486811e-04, 3.3462986e-03],\n", - " ...,\n", - " [1.8110899e-04, 6.4530974e-04, 6.8695901e-04, ..., 7.9570033e-05,\n", - " 1.3486811e-04, 3.3462986e-03],\n", - " [1.8110899e-04, 6.4530974e-04, 6.8695901e-04, ..., 7.9570033e-05,\n", - " 1.3486811e-04, 3.3462986e-03],\n", - " [1.8110899e-04, 6.4530974e-04, 6.8695901e-04, ..., 7.9570033e-05,\n", - " 1.3486811e-04, 3.3462986e-03]], dtype=float32)" - ] - }, - "execution_count": 14, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "%%time\n", - "\n", - "models[idx].predict(dummy_inputs[idx])" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "CPU times: user 19.9 ms, sys: 4.48 ms, total: 24.4 ms\n", - "Wall time: 22.4 ms\n" - ] - }, - { - "data": { - "text/plain": [ - "array([[1.8110585e-04, 6.4528472e-04, 6.8695762e-04, ..., 7.9570833e-05,\n", - " 1.3486181e-04, 3.3463116e-03],\n", - " [1.8110585e-04, 6.4528472e-04, 6.8695762e-04, ..., 7.9570833e-05,\n", - " 1.3486181e-04, 3.3463116e-03],\n", - " [1.8110585e-04, 6.4528472e-04, 6.8695762e-04, ..., 7.9570833e-05,\n", - " 1.3486181e-04, 3.3463116e-03],\n", - " ...,\n", - " [1.8110585e-04, 6.4528472e-04, 6.8695762e-04, ..., 7.9570833e-05,\n", - " 1.3486181e-04, 3.3463116e-03],\n", - " [1.8110585e-04, 6.4528472e-04, 6.8695762e-04, ..., 7.9570833e-05,\n", - " 1.3486181e-04, 3.3463116e-03],\n", - " [1.8110585e-04, 6.4528472e-04, 6.8695762e-04, ..., 7.9570833e-05,\n", - " 1.3486181e-04, 3.3463116e-03]], dtype=float32)" - ] - }, - "execution_count": 15, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "%%time\n", - "\n", - "opt_models[idx].predict(dummy_inputs[idx])" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "jupyter": { - "source_hidden": true - } - }, - "outputs": [], - "source": [] - } - ], - "metadata": { - "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": "3.6.9" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/quickstart/IntroNotebooks/Additional Examples/2. TF-TRT Detection.ipynb b/quickstart/IntroNotebooks/Additional Examples/2. TF-TRT Detection.ipynb deleted file mode 100644 index ed694d7f..00000000 --- a/quickstart/IntroNotebooks/Additional Examples/2. TF-TRT Detection.ipynb +++ /dev/null @@ -1,585 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# TF-TRT Keras Retinanet Detection Example:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In this notebook, we are going to optimize a Retinanet detection model from the official Keras examples! \n", - "\n", - "You can find the implementation here: https://keras.io/examples/vision/retinanet/\n", - "\n", - "In general, detection models can be tricky to optimize because they tend to require a lot of custom logic for sub-tasks such as region proposal, output decoding, or non-maximum suppression. This makes them a good demonstration of TF-TRT's capabilities - It does a great job of optimizing a large part of the network while leaving the custom logic untouched." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's make sure our GPUs are properly configured and visible:" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Fri Jan 29 23:17:01 2021 \n", - "+-----------------------------------------------------------------------------+\n", - "| NVIDIA-SMI 450.80.02 Driver Version: 450.80.02 CUDA Version: 11.1 |\n", - "|-------------------------------+----------------------+----------------------+\n", - "| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n", - "| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n", - "| | | MIG M. |\n", - "|===============================+======================+======================|\n", - "| 0 Tesla V100-DGXS... On | 00000000:07:00.0 Off | 0 |\n", - "| N/A 42C P0 37W / 300W | 125MiB / 16155MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 1 Tesla V100-DGXS... On | 00000000:08:00.0 Off | 0 |\n", - "| N/A 43C P0 38W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 2 Tesla V100-DGXS... On | 00000000:0E:00.0 Off | 0 |\n", - "| N/A 42C P0 38W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 3 Tesla V100-DGXS... On | 00000000:0F:00.0 Off | 0 |\n", - "| N/A 43C P0 37W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - " \n", - "+-----------------------------------------------------------------------------+\n", - "| Processes: |\n", - "| GPU GI CI PID Type Process name GPU Memory |\n", - "| ID ID Usage |\n", - "|=============================================================================|\n", - "+-----------------------------------------------------------------------------+\n" - ] - } - ], - "source": [ - "!nvidia-smi" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will also need matplotlib to run the model. If you do not have it, run:" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Requirement already satisfied: matplotlib in /usr/local/lib/python3.6/dist-packages (3.3.4)\n", - "Requirement already satisfied: numpy>=1.15 in /usr/local/lib/python3.6/dist-packages (from matplotlib) (1.17.3)\n", - "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib) (0.10.0)\n", - "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib) (1.3.1)\n", - "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.3 in /usr/local/lib/python3.6/dist-packages (from matplotlib) (2.4.7)\n", - "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib) (2.8.1)\n", - "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.6/dist-packages (from matplotlib) (8.1.0)\n", - "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from cycler>=0.10->matplotlib) (1.15.0)\n", - "\u001b[33mWARNING: You are using pip version 20.2.3; however, version 21.0 is available.\n", - "You should consider upgrading via the '/usr/bin/python -m pip install --upgrade pip' command.\u001b[0m\n" - ] - } - ], - "source": [ - "!pip install matplotlib" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Remember to sucessfully deploy a TensorRT model, you have to make __five key decisions__:\n", - "\n", - "1. __What format should I save my model in?__\n", - "2. __What batch size(s) am I running inference at?__\n", - "3. __What precision am I running inference at?__\n", - "4. __What TensorRT path am I using to convert my model?__\n", - "5. __What runtime am I targeting?__\n", - "\n", - "Let's give it a shot!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 1. What format should I save my model in?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will work with one of the Keras example RetinaNet implementations. We can download the implementation code for the specific version of it required here:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "--2021-01-29 23:17:05-- https://raw.githubusercontent.com/keras-team/keras-io/cd6201c1bfa37625f503f51e8fd3c572666770e4/examples/vision/retinanet.py\n", - "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.40.133\n", - "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.40.133|:443... connected.\n", - "HTTP request sent, awaiting response... 200 OK\n", - "Length: 35046 (34K) [text/plain]\n", - "Saving to: ‘retinanet.py’\n", - "\n", - "retinanet.py 100%[===================>] 34.22K --.-KB/s in 0.002s \n", - "\n", - "2021-01-29 23:17:05 (20.1 MB/s) - ‘retinanet.py’ saved [35046/35046]\n", - "\n" - ] - } - ], - "source": [ - "!wget -O retinanet.py https://raw.githubusercontent.com/keras-team/keras-io/cd6201c1bfa37625f503f51e8fd3c572666770e4/examples/vision/retinanet.py" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The code has some unnecessary setup steps, so we will pull out just the model implementation itself using sed (you can check the end result in the [retinanet_model.py](./retinanet_model.py) file)" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "!sed -n '1,40 p; 71,820 p' retinanet.py > retinanet_model.py" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "!mkdir -p tmp_savedmodels" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We perform some imports and setup:" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "import tensorflow as tf\n", - "from tensorflow import keras\n", - "from tensorflow.keras import layers" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [], - "source": [ - "img_size = (224, 224)\n", - "num_classes = 10" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can import our necessary RetinaNet functions from the example and initialize our detection model:" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5\n", - "94773248/94765736 [==============================] - 2s 0us/step\n" - ] - } - ], - "source": [ - "from retinanet_model import RetinaNet, DecodePredictions, get_backbone\n", - "\n", - "resnet50_backbone = get_backbone()\n", - "model = RetinaNet(num_classes, resnet50_backbone)\n", - "\n", - "image = tf.keras.Input(shape=[None, None, 3], name=\"image\")\n", - "predictions = model(image, training=False)\n", - "detections = DecodePredictions(confidence_threshold=0.5)(image, predictions)\n", - "inference_model = tf.keras.Model(inputs=image, outputs=detections)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Finally, we save our model in SavedModel format!" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Model.state_updates (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Layer.updates (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/detect_model/assets\n" - ] - } - ], - "source": [ - "model_dir = \"tmp_savedmodels/detect_model\"\n", - "model.save(model_dir) " - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 2. What batch size(s) am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will create a dummy batch of size 32:" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "dummy_input = np.zeros((32, img_size[0], img_size[1], 3))" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "CombinedNonMaxSuppression(nmsed_boxes=array([[[0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.]],\n", - "\n", - " [[0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.]],\n", - "\n", - " [[0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.]],\n", - "\n", - " ...,\n", - "\n", - " [[0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.]],\n", - "\n", - " [[0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.]],\n", - "\n", - " [[0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.],\n", - " [0., 0., 0., 0.]]], dtype=float32), nmsed_scores=array([[0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.]], dtype=float32), nmsed_classes=array([[0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " ...,\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.],\n", - " [0., 0., 0., ..., 0., 0., 0.]], dtype=float32), valid_detections=array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", - " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int32))" - ] - }, - "execution_count": 11, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "inference_model.predict(dummy_input)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 3. What precision am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will stick with the same FP32 precision used during training:" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [], - "source": [ - "PRECISION = \"FP32\"" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 4. What TensorRT path am I using to convert my model?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will use our example TF-TRT based ModelOptimizer wrapper:" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [], - "source": [ - "from helper import ModelOptimizer\n", - "\n", - "model_opt = ModelOptimizer(model_dir)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Convert to our target precision, saving the result in a new SavedModel:" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_2 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_1 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_3 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/detect_model_FP32/assets\n", - "conversion complete! prediction shape: (32, 9441, 14)\n" - ] - } - ], - "source": [ - "opt_trt = model_opt.convert(model_dir+'_'+PRECISION, precision=PRECISION)\n", - "print(\"conversion complete! prediction shape:\", opt_trt.predict(dummy_input).shape)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 5. What TensorRT runtime am I targeting?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will stick to our TF-TRT/Tensorflow runtime:" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Warming up...\n", - "(32, 9441, 14)\n", - "(32, 9441, 14)\n", - "Done warming up!\n" - ] - } - ], - "source": [ - "print(\"Warming up...\")\n", - "\n", - "print(model.predict(dummy_input).shape)\n", - "print(opt_trt.predict(dummy_input).shape)\n", - "\n", - "print(\"Done warming up!\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Performance Comparisons:" - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "109 ms ± 5.53 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "preds = model.predict(dummy_input)" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "45.1 ms ± 106 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "preds = opt_trt.predict(dummy_input)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "jupyter": { - "source_hidden": true - } - }, - "outputs": [], - "source": [] - } - ], - "metadata": { - "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": "3.6.9" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/quickstart/IntroNotebooks/Additional Examples/3. TF-TRT Segmentation.ipynb b/quickstart/IntroNotebooks/Additional Examples/3. TF-TRT Segmentation.ipynb deleted file mode 100644 index 5c091c79..00000000 --- a/quickstart/IntroNotebooks/Additional Examples/3. TF-TRT Segmentation.ipynb +++ /dev/null @@ -1,480 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# TF-TRT Keras UNet Segmentation Example" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In this notebook, we are going to optimize a UNet-style segmentation model from the official Keras examples! \n", - "\n", - "You can find the implementation here: https://keras.io/examples/vision/oxford_pets_image_segmentation/\n", - "\n", - "Segmentation is a great demonstration for TensorRT as it tends to be very heavy on convolutional layers, which accelerate well. This is especially true of UNet, which consists entirely of convolutional layers." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's make sure our GPUs are properly configured and visible:" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Fri Jan 29 23:22:33 2021 \n", - "+-----------------------------------------------------------------------------+\n", - "| NVIDIA-SMI 450.80.02 Driver Version: 450.80.02 CUDA Version: 11.1 |\n", - "|-------------------------------+----------------------+----------------------+\n", - "| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n", - "| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n", - "| | | MIG M. |\n", - "|===============================+======================+======================|\n", - "| 0 Tesla V100-DGXS... On | 00000000:07:00.0 Off | 0 |\n", - "| N/A 43C P0 62W / 300W | 125MiB / 16155MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 1 Tesla V100-DGXS... On | 00000000:08:00.0 Off | 0 |\n", - "| N/A 43C P0 41W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 2 Tesla V100-DGXS... On | 00000000:0E:00.0 Off | 0 |\n", - "| N/A 42C P0 40W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - "| 3 Tesla V100-DGXS... On | 00000000:0F:00.0 Off | 0 |\n", - "| N/A 43C P0 39W / 300W | 6MiB / 16158MiB | 0% Default |\n", - "| | | N/A |\n", - "+-------------------------------+----------------------+----------------------+\n", - " \n", - "+-----------------------------------------------------------------------------+\n", - "| Processes: |\n", - "| GPU GI CI PID Type Process name GPU Memory |\n", - "| ID ID Usage |\n", - "|=============================================================================|\n", - "+-----------------------------------------------------------------------------+\n" - ] - } - ], - "source": [ - "!nvidia-smi" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Remember to sucessfully deploy a TensorRT model, you have to make __five key decisions__:\n", - "\n", - "1. __What format should I save my model in?__\n", - "2. __What TensorRT tool or integration am I using to convert my model?__\n", - "3. __What batch size(s) am I running inference at?__\n", - "4. __What precision am I running inference at?__\n", - "5. __What TensorRT runtime am I targeting?__\n", - "\n", - "Let's try converting our segmentation model!" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 1. What format should I save my model in?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "First, some setup:" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "!mkdir -p tmp_savedmodels" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "import tensorflow\n", - "from tensorflow import keras\n", - "from tensorflow.keras import layers" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Next, we will download a specific implementation of U-Net from the Keras examples:" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "--2021-01-29 23:22:36-- https://raw.githubusercontent.com/keras-team/keras-io/cd6201c1bfa37625f503f51e8fd3c572666770e4/examples/vision/oxford_pets_image_segmentation.py\n", - "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.40.133\n", - "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.40.133|:443... connected.\n", - "HTTP request sent, awaiting response... 200 OK\n", - "Length: 7422 (7.2K) [text/plain]\n", - "Saving to: ‘oxford_pets_image_segmentation.py’\n", - "\n", - "oxford_pets_image_s 100%[===================>] 7.25K 29.9KB/s in 0.2s \n", - "\n", - "2021-01-29 23:22:37 (29.9 KB/s) - ‘oxford_pets_image_segmentation.py’ saved [7422/7422]\n", - "\n" - ] - } - ], - "source": [ - "!wget -O oxford_pets_image_segmentation.py https://raw.githubusercontent.com/keras-team/keras-io/cd6201c1bfa37625f503f51e8fd3c572666770e4/examples/vision/oxford_pets_image_segmentation.py" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The example has some unnecessary setup lines, we can pull out just the model itself using sed:" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "!sed -n '70,74 p; 108,170 p' oxford_pets_image_segmentation.py > unet_model.py" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Last, we import and save our model:" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "img_size = (224, 224)\n", - "num_classes = 10" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Model.state_updates (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Layer.updates (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.\n", - "Instructions for updating:\n", - "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/segment_model/assets\n" - ] - } - ], - "source": [ - "from unet_model import get_model\n", - "\n", - "model = get_model(img_size, num_classes)\n", - "\n", - "model_dir = \"tmp_savedmodels/segment_model\"\n", - "model.save(model_dir) " - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 2. What batch size(s) am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will create a dummy batch of size 32:" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [], - "source": [ - "import numpy as np\n", - "\n", - "dummy_input = np.zeros((32, img_size[0], img_size[1], 3))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We can test our original model, before any optimization:" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(32, 224, 224, 10)" - ] - }, - "execution_count": 9, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "prediction = model.predict(dummy_input)\n", - "\n", - "prediction.shape" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 3. What precision am I running inference at?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will stick with the same FP32 precision used during training:" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [], - "source": [ - "PRECISION = \"FP32\"" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 4. What TensorRT tool or integration am I using to convert my model?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will use our example TF-TRT based ModelOptimizer wrapper:" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [], - "source": [ - "from helper import ModelOptimizer\n", - "\n", - "model_opt = ModelOptimizer(model_dir)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Convert to our target precision, saving the result in a new SavedModel:" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "INFO:tensorflow:Linked TensorRT version: (7, 2, 1)\n", - "INFO:tensorflow:Loaded TensorRT version: (7, 2, 2)\n", - "INFO:tensorflow:Loaded TensorRT 7.2.2 and linked TensorFlow against TensorRT 7.2.1. This is supported because TensorRT minor/patch upgrades are backward compatible\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_1 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_10 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_11 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_2 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_12 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_5 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_3 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_6 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_7 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_4 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_8 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_9 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Could not find TRTEngineOp_0_0 in TF-TRT cache. This can happen if build() is not called, which means TensorRT engines will be built and cached at runtime.\n", - "INFO:tensorflow:Assets written to: tmp_savedmodels/segment_model_FP32/assets\n", - "conversion complete! prediction shape: (32, 224, 224, 10)\n" - ] - } - ], - "source": [ - "opt_trt = model_opt.convert(model_dir+'_FP32', precision=PRECISION)\n", - "print(\"conversion complete! prediction shape:\", opt_trt.predict(dummy_input).shape)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## 5. What TensorRT runtime am I targeting?" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We will stick to our TF-TRT/Tensorflow runtime:" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Warming up...\n", - "(32, 224, 224, 10)\n", - "(32, 224, 224, 10)\n", - "Done warming up!\n" - ] - } - ], - "source": [ - "print(\"Warming up...\")\n", - "\n", - "print(model.predict(dummy_input).shape)\n", - "print(opt_trt.predict(dummy_input).shape)\n", - "\n", - "print(\"Done warming up!\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Performance Comparisons:" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "182 ms ± 41.7 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "preds = model.predict(dummy_input)" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "22.3 ms ± 95.9 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n" - ] - } - ], - "source": [ - "%%timeit\n", - "\n", - "preds = opt_trt.predict(dummy_input)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "jupyter": { - "source_hidden": true - } - }, - "outputs": [], - "source": [] - } - ], - "metadata": { - "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": "3.6.9" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/quickstart/IntroNotebooks/Additional Examples/helper.py b/quickstart/IntroNotebooks/Additional Examples/helper.py deleted file mode 100644 index c00ed985..00000000 --- a/quickstart/IntroNotebooks/Additional Examples/helper.py +++ /dev/null @@ -1,111 +0,0 @@ -# -# SPDX-FileCopyrightText: Copyright (c) 1993-2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved. -# SPDX-License-Identifier: Apache-2.0 -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. -# - -from tensorflow.python.compiler.tensorrt import trt_convert as tf_trt -from tensorflow.python.saved_model import tag_constants -import tensorflow as tf -import tensorrt as trt - -import numpy as np - -precision_dict = { - "FP32": tf_trt.TrtPrecisionMode.FP32, - "FP16": tf_trt.TrtPrecisionMode.FP16, - "INT8": tf_trt.TrtPrecisionMode.INT8, -} - -# For TF-TRT: - -class OptimizedModel(): - def __init__(self, saved_model_dir = None): - self.loaded_model_fn = None - - if not saved_model_dir is None: - self.load_model(saved_model_dir) - - - def predict(self, input_data): - if self.loaded_model_fn is None: - raise(Exception("Haven't loaded a model")) - x = tf.constant(input_data.astype('float32')) - labeling = self.loaded_model_fn(x) - try: - preds = labeling['predictions'].numpy() - except: - try: - preds = labeling['probs'].numpy() - except: - try: - preds = labeling[next(iter(labeling.keys()))] - except: - raise(Exception("Failed to get predictions from saved model object")) - return preds - - def load_model(self, saved_model_dir): - saved_model_loaded = tf.saved_model.load(saved_model_dir, tags=[tag_constants.SERVING]) - wrapper_fp32 = saved_model_loaded.signatures['serving_default'] - - self.loaded_model_fn = wrapper_fp32 - -class ModelOptimizer(): - def __init__(self, input_saved_model_dir, calibration_data=None): - self.input_saved_model_dir = input_saved_model_dir - self.calibration_data = None - self.loaded_model = None - - if not calibration_data is None: - self.set_calibration_data(calibration_data) - - - def set_calibration_data(self, calibration_data): - - def calibration_input_fn(): - yield (tf.constant(calibration_data.astype('float32')), ) - - self.calibration_data = calibration_input_fn - - - def convert(self, output_saved_model_dir, precision="FP32", max_workspace_size_bytes=8000000000, **kwargs): - - if precision == "INT8" and self.calibration_data is None: - raise(Exception("No calibration data set!")) - - trt_precision = precision_dict[precision] - conversion_params = tf_trt.DEFAULT_TRT_CONVERSION_PARAMS._replace(precision_mode=trt_precision, - max_workspace_size_bytes=max_workspace_size_bytes, - use_calibration= precision == "INT8") - converter = tf_trt.TrtGraphConverterV2(input_saved_model_dir=self.input_saved_model_dir, - conversion_params=conversion_params) - - if precision == "INT8": - converter.convert(calibration_input_fn=self.calibration_data) - else: - converter.convert() - - converter.save(output_saved_model_dir=output_saved_model_dir) - - return OptimizedModel(output_saved_model_dir) - - def predict(self, input_data): - if self.loaded_model is None: - self.load_default_model() - - return self.loaded_model.predict(input_data) - - def load_default_model(self): - self.loaded_model = tf.keras.models.load_model('resnet50_saved_model') - diff --git a/quickstart/IntroNotebooks/images/tf_onnx.png b/quickstart/IntroNotebooks/images/tf_onnx.png deleted file mode 100644 index f08b3c708e5965c8db274c220aeca834864f617c..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 357902 zcmeEtWn5Hk*YALUAl)SmQW8o^H_|O7CEeXEB@NPD(%s#Hba!`mH)kW>KKJjO56}DM ze0%3NvuDrVSFHSBYprXKzqF(<{41PSAP@*%CoU{KfH&C74itFh+s$~ zu~5GTB(K z)H`T=;h_$%@upY!9E>6iqX(6kPib{zMEQp>=p{w!3)R$iObVUEt1APL?$U$3qca|1 z-Yq&yRI&TTqvAWH05ot=m_wH`pHC|2#0_+W;wi2U4GM=puabk+r5f-sA@u{}i_T>g zZqK9|7_o?&c|++5>qrjL!AYST1mmg3-cc6(`g%+F@{Nvtt~41!>l#{rRE?Qy$!-SKY;Uo2)buUQQW&an$)r zrLR<-;24o1FYYx*(;pD*WdeQOD3DX-7bM%!)1?#4X+$e+q16vUEvpYBz=528ZB@g_ z&8(6?*c0u2O4yG6RWbtYqoW+0lDze)0CxIgcnnr%X{$*<$J`VcZtZ(s^kzJrs*gCx>BphE zI=$_v6&GdT9Xw4+aY}80<5mi3tdz2H1SWnS#UYCjRv90AMDUMDWP<|ny96m-!H=$m zRQmtuo^dI4E^7;e*q_2zc59fxCl5R!k$1xF(hEoz*jI9~A^IGK^XO1fK+h$DN}e3o zOMqJJ|B?=T*aMai43rB&Dqffk8;Y^?CSgYY(EVhSjC z9u%;Sr5YsLVrVqp`%T`$jUA}6@kYTNWXGj9iY9Q@? zKZ)^6DZH0P>6JL^9vaaL9%S!=utbAMeNmGv890zGe`PE)VKE~n2HzqG+{SP&ze`Ix z@~{XV;;ms4ktKe76OAwNct-4EB%#HaxrAgcK96UEIn^)@_hJbJd={bf#zWKsdaN+ zInqK#Q{~SM?c0*kDa*aLcF@M)Z}%uBg49|N@*OPTIc1B}Jf~5{j@Nj(zq|MSnjFZ@ z`_b08(Aaw)QH6*Wkd>Pp>cOtetVY|7j9JEd)J^y0_{Z1fb|w4+`~y(V0!jR@G2^V8 zdk^}-`^<=<7cH26x8&<|clBC3WtVx6(aJy^paMp^`v^xu6Rw4z7vgE-~E&H;(Fh?YPVyi91ZD}mjE zlvqMyMQ{V#Y$daRpYz?#M&)?PxkR(YkL2w=gbej1K#Az$8}*83NjMKFoWBrC7$Yfa zH}8*cC~^KSe)5F9VN6{@KM<~AK74yEJV3nPMYF~=;?uy#m5wFKA1nb;XQ3j_EPz=&lQdf`@*-BpT$d;$M|=! zo|5$BH7Iw`9Br(82wGHWLDC{qqO3fukoI!q%QWF1bx`p9xhz<+kgfe0zQOt%``7#b z`nLbm@25J7zmBLC8D!YXM;BS5+@Om1iU^xXo47F%C#eY{)o{+Ph0ni7sW%C@;{sMPCbkjFVIrpB6Vv`~1U3_9h2PUf2l4D6 zcz26@>vHSR7Q3$9*0GLn{zKh(U2t8Dt;Uwj+1@#;6Y6=uIo0{aw%XY4MD`Z@mi77d zP5F8LIpWUZ&>@p1RuJPi1`2v5gOc{8f>t^H?`d9Iv1vmkV-^w1FztSDL0Y{_?;(N? zu@bPev66%YNQ6m}Wr}2chG{dHGw`+fS7kaWJ4b`^gAsA<>EC6&i1^i~PpfyJtFG_% z`pY|S1_4H$)-Ye97{3MvY$FsC;`fv+)23GaYZ;WijD3uQHo1lo+iDDReH|tR)AHrE zdf9z-emL;w1XNfmTsDV8i+wWIn5Ph@*c%Cbbg`EzN-6?MU2i|BCYEef>8L=dZm1NM zaD188?I?Lw@m;B~gw5zYNVjjLv%P6< zmC3BhEP7R86{p=NyLWaB`#bw?<=Io^`yChgds4H_))qD&ZLeC>x|ee^{7D8&r%D%& zLr~`_y9@*53P%<6(-~$z3_2E`l3|W@^a}3D?pqC`+n(D+)fCy95(r^!u#2-5*iFH+ z#N+-d>?&N?f3Vm%+FB-SPw;KV(qbWY{W*f1E$m`>X*FoSxaa2lW4Y@ciI`ATkjwh= z3ei!{Ma_=LCER7=>h#{VRpc(kaE@(;nbW(5FAZ)-M&~zs`>QmUW=97nrKjCzxyQAw z%wK+$W0nnGp}6q7Y+m7?+%IkJD{fU!9!j~hJaRnZJxB?&hhPLRaNKBlNPxICwRwaI zsp%Ok_%(&YY>~Kgn74RKOiSn;jE_)57|3JFBTk8bath#E5w77)hO!4bzeR%$hNeL@ zg=K&fLr_JvqnmrfLEgaCK;B0EMd)YgM?M|-wUI7d4%lQ5#b3MFs|fq8)oQW9}>=JBnJ+ZC{*YFTu#4?kEN9D3ngd^Lr3#dqUQ^%hb( z;wmkf-?JOr?Y^GJvZXFDD0x4k)`^E zuKw8T5pj2O^_1*%7Q=y7sc}iHw&d=!kF%g|N4C#bpG~*d*V^Nu_km@Yl{6RX4Q4Qw z{x;p`nY?1vbPP(L8upe?V%LMIm8g3Z8YQgY ztuRgI@#oo3#!kM?ajgDSb8%9)WDCDbGoe`USi@VkywUkBIJgp_GAiL)LI5-8LSotV z8OB9)U7?;Tb(Pz~oB4@SPOaN+wv@U$ybX4Y$&B-!o1!FacDnsn^W$?%6D5``*7hCv zCBn6v(^7HVYKyp))wTrA8trbk_e?LbUiRRIcWwuBN~~msY*cJqcxfY4Rb5rTzH+=8 zoJCM3$m23`aX3ca2_GT37Y+=)z>mP&U^3dx) znm8&v2$A@~X5w(K)q2lbVzY6OMg6GGVo9-B1F^!fno0evcFLvuQdaDC>w+Z4i2Wud3}{G>X4@y2Yhk-tXHP4Ureo%7VWvYGv{ zslL5__2_=+789S8tL9$%=;)fHBSR`FGjuzYIz@>y*LmmoOj^-lGE;vw#g;3E^Kh5; zwtXoy9#fL;;t~69(~sDmdA=7^zNt$j9IvgaY{aAN2C>We)if#MMlTj$-rfJ$=~m& zsrj)I67&ND#MOIydpmJ)>AnEx)d<;AN*RBeDf;cmq=eVj4Q#TSHWs!i+#T!$*SQky zL&pe^Uj*nXiRej4fT)0TXb=P#4hRxB0|P!>V7PzJg}^9)gyiWuI0)ot2!i<2MiTgb zdW8X>r#{c$FTVSNUIM>RfR95e_&=?o@={;?bN)0K2*fMLFCqeb%V}Hb>Y7;_nA=Ff z7c~PHU@Sf=TZ2FtWKSP35n0kBVEjo#c_kYq32`=Ub5mL^9rG``v<{{gPvd|%9N2(U zQ(YS^LI+b5Gix>nPU7D!*nsn=%XGwqznj%q(K6D~6LY;HBqZdp z($Qm+6%hK<9r(pbY+z$!!A3`CZ*Na)&p>N#rBCHdxlbme%u$|h~-plhNeU}y@+448xK9Sc1($L|jR&#nJV`P@~(TGxu-+!W|( z!}TAj|LOeioBwym-*YPeXHFJo)_+g=uRDKwa?m}o{x4BH%lY?JK+s&TIOzT=jq8=M zP|!M{$Jd4eQu4qzU}aA)Fc08B@%#;(gOiXMPgVj}#RC!%;FWg(+e?CULHmXIOCzYp z4+1fI3BsO2HkF3SNIn;qhK9!0K7%JMGt;L^Dfm?dBo!-ZO+5sxXPCEl4f5^a>ql7o z>hkRR?B;uy@AIzO=1uj^?B(Vg!HP7q(74bZp#T1!iu1sPQ?Ppzvcdy`_^)q92+f?f zU=TRre|;|@fz!)>fM9F(W`#%sXAXLu(Dp8=p{|fbgh5G;7 zQsT#bxj6Csq=et>#Dxz8ByidZKQ8Z%}=Mg zJv-SAgiFI+*ROi~Yo~Hf_KQ=PRKjNHS32&k58?5vVNUPRu%}{)z2Xfn?&f-X#kffl zJpq^Z@I#@fjg`7y*Wcz7VOuYCwd}aAvtu#0)-bYeTQS;IngwGysPbRIDED|o#U_U9-(4HW3@B6AvX`MhLUpT{W z(DnF*nzM;2aPFmA#%xdA;Zv3A+i>4MvrMl^X>Z&jo%vd94dQt&FbE@~d?u(|IY;(w zn+HkJ`jUuOR0EDY{W5~j1fQh9pa6jV0=^Mb~@w%+q(?|LLGW=I%#v1(fC3Rf zqXmeopxm)j%EXQ`Hh_d>T43g?QF7sP$W5WUinPBYMy}{0A8<66GHqR zx}_cZ$Tssr_iaXF+#`L;r7&l~MGoyVIoZhHU`2im6pYU04pa0yDJFk%AH#&;X|aTO zMLjU8&P)-62q#82)sB{HoA{mr`wyvmJgFi|zJCtZ#qMLoKY?t ze&(PxF<2?KE*|s*O_T_L4*F&Wk8e=@U5NUHH;lRcTAt_N&8Ps`{_@? zLu(M@X@*U;d^A1kH*L#x)XAtjko=@nVdTxI5Rz>-`g> zydK5Z@m59%`)s%53X|$K@R@03|8(*22#p~iW2y^*Vs1`Z;X}ns^;`|~PU|_I1mQfd z!Bd_*JJ)C`u6JUhy^C6YBhnCm`qm(NUVS@YtQcR7ly>*Hm4C$qID>RBk5j9w8)XLhcogYcxx43vacP8k^f_piSrA67cga>(256q1x7f7pa1MG*t2KXRx2n5rvJ zP*>M~ia%;-*D3}P#oJN@VI?X#>@J)&;<>k8@VPIwjM@e5xx>;0#LNVNsA8pK$QOQ< zT3a=(9IM^lz@Fb|yNOJZ$13%Rmn6zQNUIkmEIfUB02;%3b`LIyh8)$^VaE5C#6C_} zZG#uf!LmFp*U*s)7{fRlD$d$7E_u0yR$P#2$p1(g2F#pJV`u&h0Vn#16!O}qUZn6W zAB;bW3-LF)gP?0DJoDm3XIe0f)vaCLs1?xK4y*8(K1&bDf(HED<*UKRe9N&En!CN- z{Y{k~epTRRKJ?hw#Lxvp^GHC(XwHw5uL_QY|IF#(u?_+5+P*Ei5m0YmKHELp`|AWQ zkrfh7LDI!C73Vi)9*z(rw0l1~UaDsRxV%Be^Y+h-ATU9gpHS85!*~lBS2<<@8-aa* zxtOnZr)nVd0J-rR<4I`NZAmmDFe z@8Mia=a$>siTvzI7< zYpTS6$v)}y_dAD=aIP6_eq@`(2z^vB>Mk5WZx2@(9W?w@jSCLML&%GckdbE}e=HaRqQL;9 zT5Tk-B2b{V4x#vh%EN;R;z~uno4h##JvT*G!ZeCWneZR!d3fMKA5AOpt8(&ssdDa= z87sC*>-ynB3O>_J!B1F!#Y!`G`ZkH0?T5H;7vvvQ<>65Phv%Oij{QN2hC3oUZWZ&;hTXJbtRk2;kzAr@&Su!-^O37fdeD6zJmUzGg(LbE67QsYDH_v;jq#FFS?o8;P!e^=M z^^)1rk64*y0$hOVULjEwQ|T2wsQ=*KRHO>S{X;Q8w*?fq|2su`mbnPZA-X>7v3Ged z;1q<836Y-hKn9{LFm!lsA5qTD(Kkb$r}nR@QXzeVHW0WxkL{l*G%Dmf=t2RON?Gvh z^-Y6^3t@`#tDH;`WP*Rx+0qR&hG-;mSyp8)FO8Z#f#yq<-NYe{@N}OeqUAWT|E&Gs zSYBOWEe<4Y{0}x*gM&4t*(k>vppM4&fflFo0~8OJ^rj>;21`-w2Q~1%_fzF0Pou?Xqf{n zxPB|b%g=(TgOHW%&uB(;Po-{q_jiT_)S9e)YR1&-Ml*FWc|-tgT9Sh@&vu(I-?8@9 z(!y5!6QwD@IE1qVA6(&|8(5X+tt&-*fK!-|=Er#!NC!;8hYPYoR%W$hmvExLqo1N; zEjWS08{XHPQon-dZGne25qk`NU_D|Rmkq`pgHrfQdr6Q7rrPWQ(+;M4wQje32oMqX zc!YEQ9?8QH{1HD+d{=6K84vm0U&oq@Zb`*J991w)ww}F7`$D(`2f<8_2-S<=-X))F z3uT-7;x8Ru&_UuZAw>q)C0?R27_j_NkuapsQn4cikwoEXWl;R1SSxV)Pam!!zdF%M zEPW4uWmctjL(J#+=dN^EOV>~RxY8YDNb3qNY9LV6_ltZUh#AGBo*Ng87yyIcm{&st zRs&|!9vDFNArWMMS$)-FERNXs!|JNT;~%qF5`ZW&y{wa~0Ry52f3sIP;7@mw1qgj= z@vd*_g>0Ngn1t2-iK;$>Pgc14KruE?&TV}IOiR=Ix2Ad*HcQ_`AqxSIB|H;3NVX)% zV(XIP#8^9>1`QxIg&W%MxGGsh^H*HQ+3iCNNBKz((_i5zLzo?0A={e2pDC#M21NaY zZ!Aci@o1*bS;xmd`IGQGyYvNc4(&^2 zTSKh)OvZ~R^eQNKpuTGS|&x0>S?= z8o+eAP`-t3rCAEDtnO3ez61jdLZGs;^6Y-#*kKEGp}9{CO@rmkj#{k`)jTBNCr3yKKT2mB&x+d z!s91^t)kPtfBD47AlcF7awz5cn0+!Vi0SvGr*6{0Egl<)LrkU`@-*%`aNtD(n)CkC zXN$w#ML*u#{4GUV5P`R=1ZQi(Naf|dgHy;`5!`rU3lM3@rb)Va_2M9v#_qosBM?{= zjKi=O=Z18CY!g+C*B)<6?eE}2eF2-<|} z{U1RhZ_V#X)}`Gi*ij=wdU`M*E5+Sl7jixLFyun0J!=g(&c(L0Y#@$|n`%gB`qIni zenpRaN=`iVQ9c4iiKYc7@_dfZ4~`pn9x8wA-J=nZOGcyH4DjK0p@E(f-F&e9^&hO+hSp-xh%yC+m@sB)~48TiU z3k_Tzxu48rI4{)&sRY(Deg@~p=@c(gy=0blJq_vaz(_+V252+Mo#*#aE)A}g#DW*} z@|0hZ|3q#DoDWPiQc16n{>qwA(;l1zH6@c;h3m_ul?)*?>(`cvGoQ0xo-D6r7<(Z8 zQ6vZqXZ!+vXYo#5Pz1mh3Q3*c8CJ4LYR(FWgfjP6+D|b~|8~5VG>DyRh~aXtXt^AB zSHB(4HeMc6aNQ_D#;D4AF)F$oHMEnzrV0Wh0-T5*a3WryNaeje0S%;szBnV#FtUYw zWxCi~ak@v$i7ok8K|C$V5J1lEZ?=7w#u+G}{KSbn!9%R9q^tgKPX1sJ2zm=;3&_gq z(%Mz)h!aA9kzoi>o^_T;GbN>AQs?8DbFPHt+4O*hfyArz@D8m@C`x|(vrc($4hjti7b#>3^T*gzAu-t+L z*F)~2{2(|gDdrMbcK_;4XlQ82NzeO!&c{(ruo5zfWosn?em4zw zj+kZfNKqA<>D1IJudd9~fk{`=>~f*lcGEpxr2N4E9#}3(TNb*`D=S!$JK2E`EFoIz9|RT)M87DyN4 zAe%GJePp_gL1>i(3F5OSPpmt%TKkl&J@iPU{{pqFe-lR4qb}3cC5qCQ{mtpQ z8)QCB-ikjU%vcz`VNfAcO?rv))q@1)%IS6};vPfpK8)GP2(2`fe^kzQwLJ_UT~`vv zhdJx`UU4>3+0BqGsUb`pccdvTD7~C{eLc?~C`pV{>&Go1o+nDtWH_rvLznUbSwbQ; zhd_$?>;4Lx$7%jzgn1n~)dUUE2bnvi0izk`1N(^$^+BldG1hBzJM;4)JDcX+*>_be zYLi-JdKZyQm16zMg6;%v31tE01tJF*hDUuh_hfeYuec^|^^Gv5$2Zu3{F8xC2S9ot z2ozh|?YEc3VW@Lx!@zwlmccHbewEV{AkUJ&MJTb@!p*Gt0&J3GsL9Q~uKqTZuOkNE zC7e0=fgcMh?HmGF@3y1e&hB2R*neS5>JuABufGQt5;J#yCiI}tEy>Ueyp8F$jR z?yKE&hTGxWZS>HEDqWRtN+7SjFrRJ5{l69BqpJYsPgjd;fIIU}M7IyIEl}+b= z!C~bko}uKYPTbEoXL{efknh4){_1fL^8<4)GcMuB?(Vb(>TP6BMOuJy(%|Rihmu|O zjj!zUTu~O+vb4?43BvH{l--Im6-`JpA++xN22?G<@}XxcA4}6r(F>7S-VDwjQz*zQ zQpFeFzn8mkTy|QGD_tDt7wJKAsM!QJLZ0u<>eKv-XC9j?z+U&zFE{ur6o2PKXncusuaU_d=R@NDrJS;Y z56*P{_iU9xSVfG%AgUO+xV(=%{D2qFOLGkXu$PS+=;+LQ-p&cTgU9l6(O!iufr%q+}xDg zT_Ej@9DOpvA5W>2Xz)J1`gJB1@p1$*&c)lG<6}-|y<#AusW1)9`7O)7;-^>6yTXk#F8P9SU% zfgxpT2kRp)tQTX#TYXx}~m@H5b5Ynr#0k}Ai4b`l-kM(QbVG(l=b$}4Kz zN=A>&jbiepnlHGzDZ9M5@_`>Znna((GiUwa6^(b@0N8&FOwRE^W8!TlNd+FfKkMZO z7I3N86+#SNGC#I8UM6SuZ;J`?@?o1tEX?my2k3jv3#cmVa!H8c_i#i-e*2mvsu#NH zqlh&aB&TLaq3!*YW^>wihz39#7?4K0Ai*h|m>~VGCjgJMl}tx~W303pNL}o+<2s@?dSU4C zC(hTuQEyMZzx}l+mNC<~qC_Um2CXEbsj#!zN3#n@Emv>$UKL)UVPk*yQghdE0JE;e zmg(}#MZ!G^`_2P|HLhiL6sni+M#;{r1r5pI6dGJ=nO)6} z-^HgZU-n|q%rg|?9I9WEu47lOKxM}GnjjZVyM)xcoES(NX`RZ zzP18+W6%#`P8h4@WL|O%|I7x>abX2Cr)2scNVD50RNCK(ZHe)m1_q^Wuz($5~}ix(~s$_MFz88%@^7_LKR%|dQi zb4Q%?!tSRAV0QTYl&x`pJ4zDb;(E6Yu1ZHPkLIF;$T0oz9(DQBX5dvjl`pBQE8P>A|My(?|>`sLKG17fcc&P^Nf zP0GlzEtSDsdZQf-2n7A&%xiW6+rT;{6_`vog?3d=9)1+{NVLJpTqkWu@_gu;NXqnfDiX0qY`B_+sP;qVACW7k<_o4=kcqrTKOk2sv%r=v9; z4P86i0F?%>TwfI2iI?-;RZr1Hog=q7(~g}z@f_8Gplft?I(_pvxyI?om^Y4d*9B9J>t8D4q#+WW}@RYyeIX!bLA)FH4c?XXCikG>*_}iI;;=d z@v9gQRFP8^DX*U6!4}j%U-Knjen=~D)6P}fV?x6lnW3*aXaMS60~mX$?bDAV(PxX( z;_U%-^wtPC-i)cv^4F>GQHkdv7j<@LGF<+rhxgnY3yj~)*fJp(8i*oOs( z-lY^9 zP@8E%kijchIa)?x?A0V8BrTE+)4nH6ykV>lMET+?^V#5K+Pf6;mGUwx(GXdu1v@sE z2aTB&(+74h_t#Dpmn(H;UiZ7OlbFl2ul$X6R<7=@rZklwiF1GS>L@A7QX<-VNR3CA)<^Q!im`GsqVg}Y5fq9(8v~ARzuD-o4 zN-ohk^sTH>*eOs8hv`I_b7eB7dg}MVAT7wf@x}KIZ!WIj849RprhNroTQ9KXadQrd z-90d6A;T$jZV|%$u0y7V@F3tyFx%dLtFyM}SStUnj$7pee&0qmzDhvOEpafYm-;!( zdzycUsEYjb^p+L~z)Esh4~i_sX5LmcwNB~*{~Njs>&3lmIB1ZYEOKM1%N>QA?y!N6 zS|t#^j#uwCY%Wh#Y1izp)pDHqh$$*LvbL75J8HO!eQ^ED^Ve-5J=$nT9?Fe@DSuj{2g$9keBaYTJnSc!lkdKNN{SV`Q-7x)!GK`^pAIvOw{<~l?R}YH_5dg# z!;w8@1bsLR!+yBslT>+#eQfgFn21@o<&4#h>V*I1k9E4`MrV5%8IHgk+lO%GlM%I5EP7aX1H`e4CWveKV2PgJ_`~HZjT>n!3dT41(mVO z^#umC2fDWPZqTbZo~f51pkLGFAfW&Khd4SncLhwUT(m)#sN2T4QZ0(gP&~;Os$AL^ zr-N$3_Z6GS5uN60{ZVerwx=7v?q2kGgXK?|+eB)GZQNBih7Jz|F}2ZOVmOuhEInpm z;jJb)PhwVD+k*R`ZyJuwv=&IntnVn4in;@k-Ix%Pj>{$PfxULxMpy;n zNN_ExM=v)%2D9eX>z;U%#V|j+swjfOrK!1oT#HisN%Ja1wjO?M#>D8o6~ZvL3E;^xYwKZwn&@AdeB_A4zf?Y?P;MI(tF z;fjMXDZT1z+tW&Wcm8SYLf`%n1)1jLKMgxkn*CLZ?7!!{2r0Sg!_p z)Wyw|2~M)rv~xA?SOY?LsJrDPTW$+xr3Ta1f9YZ2x#Ucoww=qK#0?i_lb=@)4NNqi zgm7Jaxvx{$!yBC<76g(=Xw2V*dmxEaa@`YCaTMR9>QT#KgfkV8y1;aAnN7JSzpuH4 z^pii&^AXs}XqvS|#D!P>@}8i5fKaLgNFv$Xh~j8O8(VaoWW4#6dktD? zE%AaZQySOVm;(=UD8!VfgZI@_)sGQH+p4o%+x33B5{FxZ9exYmzrbUH-X(=&QH+hQ zj&qwkuTchqeK^E9Y1{AX*3=2e%l&%8aWflc^>HB^2gQNR9>MqxS@V9jo6-2+gy`~y zEv+tZXJikuR(j{@Cb06_VwaD}u-jKz_Tt!&AN51rm+G7D(Wa;6vTP=Ntx1m4U8PEt zof9gEm(}|RnyYw*4{;HDDJ3&wk?v*AW*9p!?!3k|UcMQ*D3@KW(PCptmS}(PTRMEz zj-tVYHko9teHOW9yfeVLt?94}v!}hWnk=t7HtqxSBQahWs6@8-7;7f%h+^jznX`xA zl*Fz3s&ZMoa~R%7`F^ot=0{hQ?q_Bl)F`ws{%$HfT%S-KLv7rX;~se2fP-P4AnDox zMVOpwJS$bkN+U&xL!n1SB7DXozOD1NLGCj_UB)hlR^*@@6oY-&rngq~EB0SNC|N&% z{^r)k)tyZvo&*kc9<%Km26V0B_Ybp&(RiAffDI_f12*t!I^ssvK;O8y?sQL>l%rj* zKA;i3+Ysho&-k1{i%ZppTyIwOX{N?_!X};>I?<>tY6VD-#w7?zJMqTTi%TqTUC_x6 z>LF6dLR^+|8TU%OaV_X&x~UEPU=2rZ)ibcj<%_VnhXKVsV_al0HY>SNH-w^S<)RiH z<7b94KpNjUv4DJ?y-9n8zjd83O+x8|K7?%5-in!E9%jp#6-=e?z#3j367PMP3Y0@< za>|?^EM2LbnJY=)QPbIm8adr7x7qG(my|c&Tji?9iP|S}tT#<~8Laxx-G2#A%aagC z!QH+Ol+M_u^qMkvjeb+0nn7eEw%J=`>kZR~X7#wQVTPj|Q%w;UpX6@pXw_^!V_|&J zU?w+#7*-QD(l1u1FINrc=As`XVsbGH@4HfGU8@XXi#&HlNJ|>t&#b!A$s{m7=ikyN zp{-OeH>YQn?bI_aNXb7=(zh>dz2Vknv0Bu$aS(>Xnm8=oVXIgUUo=s#*j<%@b7-f3Vts5cE!)TK>)D*VtVA4N zTRnFXlg4Z$<@u0uTZN{1t`MIo7>nf?{U?VM-)*6j#Ks{}qY|u;1o*8#fwfS{)Zjr^zWu>`L zaD8)1ML~jT*UEpc&B%5toGZnR`>5$czD#j_{LPb#0Lh>f{W?@d}@;>o2**9DduF3tG$Y+$qgvS$%8{z zD1*n}!~A%88QZ(C9HV#45_wD;re+hHH|<5iu-h9J4vZcwM>}G=@DjrdU@ccQGsvX<3G54n88pne0H%h zhw0*haLBJl!P%jQ*l9dL!2FAe?XcIie4i-T`Qtp_Q*jDgl#3yGVv&zs^^(@MXjLET zLF}AHsR`VdYO~I52Zu+y4co}+t4DB?o96evW;enqTWtRMJu=_flS3Cf$2 zX|qe%QL8iqz3*l^fDN0yc5AqpBCohu<7gakufI{tb&jatifN>q?5>qr?x(hYb%B@O!`PWtoG z*3koF8cebatINPlU#bH;J(tJzHvLF<@AxWScV@{+U#~dap5om+=}&P|xRL{@)`GX& zF5QN8Ef~ogorpluF96=`qt|rTECPH{aiF=wQAa2Bwux3|; zw5Y|0`ZUpC<{Bahvc<+Zp7upgmsQhW*-#(#J7bO5Pb#t&=I09cn}ru5`^;7jI{e5R zfPCRPzPhv8brWOsWK87y`M2vms{?%Lftq7->&JRB*Eo}nkOT8BDm@{OEytzqi{#s2 zw}sO@><_U7f&&y)yM3W&yARbq*x2=SRQbJz3Olz2h_&9Lv^&MBJ@9;ZE51Wr%;w$d zit0;?e(HL=7qlqRh3Hh*4M!YN4tgAg%|3Qj7iKmC_fdW47t*kw&4Kb&;v}-l?(H`JvF(*j&|H$;q787#V7Ba)0_AMl5 zlDc{qkE_Y0-ix_aG%O)&Z&EOvVyA`55F(!1aPqb8%J-$qr=JloK0)^`SYPO?)JSne ztoD4m93AqGfh#BBN;kczdP!!*&kIyvBwkmDytn!uYM1!krP5omT8KYRwLH#u+F-yw z>ybw+)P?jN1dfno(18Gp{R5xFB3fA-7Es66E-(3+@fHZcX5L?k7ja*33$imz7zheo z&SievdI3r~KIyJo5Sss98DJzlS$X^ILYMZ7(6*yc@3Lou*JaO;+INR6E@p=nR~k>9 zOr8U`(x1{D@NJ=I@lp~b&aWz$nrXxkq()&^36$Od|0XxVc;)UEgLkcTWhupioHFfT zyje^(NE}K+#m`cBP&$(2o?KnJ$O3@zX5%Vl@mmc=_{P<%qv$i~L$W@l^?dS8a+k&+ zZUhl-+S$|TH3HT*a7wgy5-V&E4ZB2s$HMsT3DxYGO?b58qj{^yJy=nu%nlz*`d(&Gb7Lq(xq7mnQ;| z23nUR^Si2n(mB`Gm+C+Wu9ennuV!2G>l-dx1G#b6nrp3yiCzBZl(@~O2=BD&BJQ~2 z0b^$R&dhS5U-r@+cyQ{l(p&HM-ZdJ;tDagJevZ|jlxXtO{`Nw9=yp6+Kl-z{5hXsA zkgh5p4-EbE#PuO?tWHT53q)`;;P86%vm)tnHQ6T{uClm%Q9*9)Me(tU!QqYj%l-wn ziMKH_lcXW~T>Y6D6TqU-^;z(nK9Nu9K-;OJNw^DWgBCZd;4nX~mPr2Uu?@M9AicFP zB}M=_*msqr_zl8ULOR6g5TfqA1gsq=Nw=`%5LF%x=hUv0){#_w=SWilO&dJ`oy;e) z%n8P;+uRk|Z!Yn)o!_mlfs1f(z1WynI5_~3!So?1jar_J-Oj)z>IsI|YxFV61NF$^ zuAecw?5m&sl@o-AbMLfKT;qMJhx&-w;`ATbe5%PCrIK(Lh`R$HLtJXFCH+5>N@twX zUCOKiJ4^AE`}pO0M)Ok#kUQvfOvx!-YJ+n>Io{?5aHdG8vo-@7Vo0^8OK4?MH4|L8XVrlgf=rB$Fv4t16>ib38|aR5?R ze>&)aBQJ-R*gZvBFE(_%#xM}tARSv?t6(iXz+@upwF(RFJbe+c(vO*W?OI8%&i9^H zXYJ>Viyh4ObAt8Fxv*O|>fbr$ZmtR!BhHc~BAXuUrOKEl|9GJ3G*-x)#kKa@gu4%B z6Qto-s%vIB38NnekV}gp*bUQHnL1KcYFo$bJ#NWW7`2oEA>_Q0%OZ zw(E!F7+_m&a5i4Q6ZIBI7EbFsYO7~do8)`lU=EdHQ&qh{Zwq!aO8$sceKYAEhyM+J?>bU zfbMf?d?DHBM>_L^h4qX|S5g3_Cd#NZcllhJj*a}q_#4F`$*XN$BI+2!-CP#!ZrD{^ zjG#(% z`R4IxxEo>NNZludj$-8m6{5BZlMYTl)G+12j2>Z!48qN{e15pv@;QzBYF_;D{k4og zxopO1@y|n5pt`w{7l}K5f41{U+Rm3tokd)#Q)=(p#0d6MYc9dNLr&a6Y~$pO-@P3{ zcumuA=Xu5b7Q&(2V3sUNUbZEVYf=R8z@=`5#i78a{6L6GH+G(h%<>moN^4%%dFyGc+H`k=ba#hzN;llK!E?_2$9=kQjKNypn%(o;2>Gb$ z>66yC^c(P-92DcS^c-6+ceU`!g~+zjXn8fri6(Xs?~wxA4dJg452gbSP=B+rt;i*` zB4#X*bldsgsw{uL8;WaS# zcWoX)(LT_eHO4xcOf$Q5`{-#v%&0Y(1ooTtpfUfuMd{YJNr`)hxXaHu*%lkom}l(*R@E;L5^B`T`GI+g1#Cfwe^fN16) z40G&h)#;e{ZWM`ou+DFc)_#U^MoE3oWc8|yTt|G%y)d!OV?k8!#;bNVNOuWmv}O}; z(X#I*71vUYmnIUMEb(j*&Ndt-N~~+8`7HCo!UssEyF< z8JE!@iees%8iFCOjB22qBGwPcbjpX&bvW0%Do70vfE^ z&Y@NH+-pjc>}nPT#q2KOI%2R+a8caPl8iSyu<^gE-jODIU?1;hbJS0 z+jAG4J{^5Jt6YLSKQJ{CG!lIp9X@SuRF5gOo&&=@ToZ8gHMNevDew2fVJ|LGE%6z9 zypuljmrgVF9ZVf-GIq5Xu|LHUKwKCzS_MWdkLIbW{Ns14hLt@F1q9c#xIMz_!@Z1d zM25xu(FR`=uukP_M!m~0FegGxXYZ3$QzW;fY`49MwB-sC+M*D1C{MROs{fbW2TPXe z3du950kqK?i~Hy1k{CYxf+XLHYL4eh8QLSqqk2$)EfKiC%YWX^s)=`*uXyKL0o5b0 z8SM`Lly>FWh#S?k^7_DHm#l#WehTvwv#Tl(dd7{}DVG-YI9tRdLM1O;GkbFpcrw;kYOYTI9PoZPhBfhH(7BH)UVo3W1RE?t6OYWz?Y`4pN2}hu$ zHnu&PjB+19t<+&2BRXsnW}{8Fskpe(pK~+}+`mCA4|IU0KlrrTJmqb)kpUEu3>=g6 zL9Cda(}MG^o5aE8Ca8)0O-TzK3w@NO4ohnCKU=r^HR*d9^dBd43byUrB%%$#Pl|mI z;5aOIboHImVTE~o>_kR<{lb-v@|6$?1yl3#S5pisLKM;vL{iw9z*d)whsTM&%a&Bd zO7#ztIN;j0ql%bSw3j|*l0W7Kx-hqPUK;XO>YJgDCk~!Hf}->tfCkvfl}4Zjo$M|l zL=g5eAlBc^WIi%?)KU^MRYlni($D)iWm%Ek@_}WQk^!r!NyAHsF-nA5%RCh9yWde> zLowNxbu_+S8P+6m1$WcGBRiZ8y#Qcwf>&anO_hPTF3&;qAHp zo&Lkc2m4QAoKnRe_eAzim&6pulW2_<<>FGw@YIbbT)3=f)PG>l+-Jl96(y$({x+7MZHCwk~YQW|HJHZ=(Ee zPWA?RvnOMR=iHIh$=7R(#hnX9x3u$@+3Q3mH-N6a&xJFaKM83JN^SMx-UOr8?wS}H zDl|$dIh2el_7uS{+W;`o`N}T`C)~}xfLk)WJijn{Nu^fiC=o)T&MS0~Tr{276o5M1 zs)-1)J;GP8upZ)nN-M^Iv?6OU#@UHZHxF;#BFS)2rbuh9sYiB${vN3)Y-(SdxH=X| z6h59}(a{xI-FNz?%_9miai#cuy6pGY(^5XbzOVq_p{E?#7BuA{!eWSa~d2VNCM)hO9xQbjX! zn{(Y0?%jFNf7T0&2#0Mx zAyJ=}==|80g6W|^oMPi$O7MGh%1q4=!qZbn2k0n~w_idDnCpPU_cSJyCX z^JHI`=D)r%t9bBS@sLC7LIQnm$ zwM@m+5OT7QbRGOKffGP0ME2=1yIXv3u5V}3u_cI`x4Y-^C>z5{+_|JZe%EKq4f(H z?4?!f+|rl}$Hwk*w}XfUP(m3hno`vbORxz@=iB3rFT`OXj+^Kyv*t04ycK z25Z7kj;@`-Rz~j6i32ZPl=?H&YNz}4!nB(0)X~Oad}EZtnMAqE^z_)6FEELU# zk$0!Y>t1=&Y#02l=8WQ2Dhq8>>Ei%j5?(pWBL1>8QU{k*(78D<|)ZFm^pg zAd!#(++TwOMm&OBw3E_!oULfF`E*OH=Bi&`3=SgDj{cPPy3$=f3j4?mI1L*g2gCqZ zA&P>HCWijH5S!}WW=mx*y-5To2N>!dLI^BltxM^Z^?#V!n1?yo@25}M*?vmfNrv(y znG|umeEFYdh&7|qZ49_Zn4TEiXY5v&S^pJZ?KQTw=gs> z(`tgo;EZJErZLGI1NRxFX^x(}GsWhT+{P2b4?mhFN~Sl{82?={A?tsuCN*@^v~Zu9 zld=17sNGX67OU+TrG4fJS!&PN%-O*XF! zCSRI`yyf)G02ZcJk zr_MAj|El77*U~t`6u7C8&z8z|dhCHMpY``WkUQK*y&N9TIcwP+a(FvoTwr=C=C!BG z5LDm9Aq6`jV)(|`Xm8B=?zm~uunekF83zQ+T4p(;JTJva0jo<^+bvn?&KTbUGD5d| zAI$t5mJ%Oc-aD=k>Lx_BTZRUVF{Bzvma}XMenFe%C)pa4EgmgW%vx3wip5T@e{B5C z?yAs)hZ$oq^2Wfj-a|}1e8?=w_V;(^=DS-#%z7}zlJ04*tIi6cIDR{m43xpJr6kGp zBM0+Ma4Io($5@5})7Z4^`03w)Eqn?KJ~{{N!H!Dc3?@aBqgSI#XhN|X^V3WBNOC*_ zv0X4C>wl;@)32wF!vg~h*aa9c zw(!h=$tnvf*%quZ)fAw_wjC@zEqB}{9}O-r$SdFA0PS8%=i`C`P%J#ykWM@5xV_(p z7U#}$x=no}aKVGMVeuBZ)y_UJ^0ceO`=ZdSG4Ie<&nx<7O!$q+2kftA!2XrCJoeJ& z6^jeg?Ug1K+&H{n063WAtj+ZPB8RPcp;`K~QGHE!NtBMl`?y$FvDvZL@wb01`y1kY zODhlZ5M3k$P@6UiH&j zo?VC8&c&~HqQO9FMneSiG?u!Bc16G8+NW<8?mS0zX%lDn+?D)!XgtmIh~jF}gG_qi zr0kee4u;oFNMZWOej-Z!boPZZkvRRBt#3RQ5r*-nH@RztbM$lDcc`lqZ%pKZKEVIlpUOUyu>Lz||9z5w!QQazmy5<1r{h z4X7n$_`7Li2iRov0IQhl(pL0G;;dDyj-xSmGe33t3I?FZL+-1pNz|l>MeGZv77SACL0PYtf(>waS9pUnYx} z-)xN(#H1L`8mrfZ=f{gGo$DHidz{lvZY#ITc9AYj9W(5Xts}#|6%v395J%ZEpiAxP z<2bE-%f>*D(xJ%#GLV^`rY#CvRu~9>aKX;0oc|_Bl#YOQUtYP|{@frCoH4$*tU{@n z6$H#1k`Si$>vLdpr+m6S-}@S)Nj1h#EA)*69ZjKs{| z;Ig970g83*p0`DBipxOw%PHTBLlR%QJf2x`_NMW#Poaz9$a0hhS;mV4b5g8He_k!m zvx^^1ojka`7EFr0k=_JPS#JVAsu_-Ft>L$BH0NK_1e@|-_TM(-ktG{i)Op)VF{*Us zDri{kF0B)(m7cx>$KfV&xe|!dCP}-0z+x;sESL0AF3F5uE;F7@4x38C;i=<;}!Bq=A#O^y$@^HUb#MKXQF&k zHazfEcZ!o35r&eP>j{4!Z7)`U>bOn22ZOLlwJebSb;we(lQCCV^LbPs8&)_}Ps%!Y zT{6;SJj6zwd=4#*r5Hh_Ils-R&gh#nQpg2k91=^GwGAdD|3=tX9r}1vCdQshP1wRs z;V*c<=O!O3r%X*DYx^O+sAFs%j@2$jvKlC3OhLAggqpgTsjl7Z<^tr}9k8rqZoYA{ z$24y>FS@=YY_Xm(Cka2h`LdRlMC}ONF9eK3YdM+Cg>e%u11T<7yX*$gy^w;jxP9ag zPn_U1%wviMb~kdK`471C~TwbpLV! z`wS$2#{?X$gwFF-TC!p*_DkN{VU*6D9k>T+sWyuk z&rKdlIyq|c^DcQ-u6Ax!?r$UNy6UCwz2dUW720xw8^?uz!%8fqT;ftUhNqt*_srBcsyn=KM7bNPG@*dt67()X@0l6b7{7~(OWMcy%upk-m5D=Ha;I{PIqj6bsRu#pI z*wtTt%#*zVmN((VJIyZNMEI?4Qh6VAQXIEm(0jae10XMJn{N!iybi33IU<`d_#KcD zBWJIK^V{E)hmqULEDcxY&?+N!@Ca{fnP}(ylpor~-l<{<>$V@8a#)|>3 z+X_d<=|uR(jDw1ahHLV2+E``GF~09wSr$f*(VU z@%MAauD7{~{ZTU`j;XbBL>y%Oej~b|ILl6$3yRNP;p6$u6tLx-Fj*`nDXZ}_%O8O z7iyr@*wpy)NZ%j3P)v1fDkJC8F4>K}1ECVc#eon-gGyRpHf!imM$e3vf)$5fuijlf zH~O&6tsp9|vRp%1(aBu8JCORfhvxT@wU^H(Uir(={U}hCAB8LjpKo1u1B2kXwWY}I z&i&zyB4#oV6o})P7pAtXi!(M!#o~4vAO(TwsS90Ze@RN+92m*_W>wR?^684+&|#T# zb+U%P*C`h4a;C)-MOp7G5Ak?&PBE(4UOV9iBf$`e=`7cIC8Fy4{K2qpkBS8Z6&HUP zQB6p|GcpQKX;0qE0ZS^Lm_7P$iT?S|+E#VSI60VLppbS~1QgQBo`yog>>uunQmN#o zf3x?{!V5umCn1~ZkU3koJ2SO3&Ph9{(xlUHk72|s3~$K zbm;02uze!kp)UWuR&FA&4h0Ui7HiEChHsw9NAkTSw$TC3eDdKwn|4iJy&i9q`)D(@ zv5cW)>gUVWnampQrKg>?kjZ+9v#<@k`}ADI9J{Wqvo$;}xiQ%;{B%a-FBvPYa#3X0 z-M&C7r0~V0m|`G-gp(#{5Ukc!>=40(x7D~g(_Q&6G5j(0J)={^e z!(CS8Hwf}-0IsNC*1_)kbe`L|8wLEao9WclhjL>5t9k-pDc&M&s$FFqBD->*-@m#P>smOg8v%2SyK{%!@*3g*P1tc;K9yz+kf3d_Y?P=#jxlbcD+S6`^o#>@|2p;vRx-I5TC+Fr@aAsd zR2OyfqC|XO{Htt8y={`gAMX@JlgVu7L*yGVL%oe4)>W#y-X0k1Lm2prN~3`-8Poea zT6<)SJU!6I7zll5#gFFwbTyna)+tioElAw?7Gc||w=Pn%gzj5ZzO(@*?|^KDW0u?H zn?}}hBv56Wo`m~jtW%JCov{Q-@{$_#fFy4LeJP>^m|b3d17|_BaMj5-WQ0xp%jX0U z{;hE**0!b4xi5NtnpG*A+M=h|Ch^&yL+rhw4P>A@3q%#`cfU@N@gGU-s&%mKcy^?a z1i9pbjjtWHA2`lvDRb|sa_c7vsnbXg%~T6TUjjP^Rfr&quTk4eL@lrmYnSMkv*}u; zzC7rvNz09&ETkd+tQnBoPoN8Rlp7X$@~4NE}tX6*vjVOPVuvH4~u* zgY&1zj<-EpuFC8!8kBjGplFYrAs1Bq04~O|X(apd%UeEx_95~vQfnE0I(Tgov&hO7 z-wt<19)m!TJ87o|R(DHyk^wSFH&{3n_}0Z=jxHEA>h4DZHOo~9ENjp zvl?>DSce&0a>Zx1#Y{xkyl5Vw+Weh!+V%Zig)tvUS^Jj`n>7pDF9s19Hs-m@KRfIx zDFXdr?!|+WjwL`Kg-{7hVpKlF_&&Jzp`PvzoSLJX}YahQOb@8HZP3AC<(t{yOJX15r$>M~7H3&zxJv$g7{d1FukR zgTu?Jx-B!JBzF+sFK$cPzjpDI9Lig}7Abd7P9v<$AMzlJKekx9P~Qys9^ILPasVXz8%j_?%-a((P1X4-X2Z?CSI{1-AB-p!ZAT4xpEW@7= zuA`+f$|GAlY)ad2QM)Q@%mgPC`7^5=b$ogpO#5WrWX%*+#A^Ra2(u(>>_{sum%%~| z$Zf({6pwGXz%6b4O16+D;Njs#^zVoq{pLE#?j8KlNbt5P=uB)Y!G#TDtjBtXW8jNN zes{6>Sq#Dj$qy86!}niWj@*1{XMb?$sYhtPSX}l9nGZE|ngQ%&!EREX0U$UdrGG9Q!91bl^> z5a?1=x!j?#xm4m2vY_R9S{0+3V!}eK&iRE~Z(&;W82;tPAvRF{0hV9i`r;1Kub;v z5p?0s_Pxe4BB0V!K;Ei|q~(l7T8&Qw_&JDMzin@OuVE|hw3jbF#kDpbgqK6Hn-LcH zMaLHZ$_P@n5Z?Q4Y`wSTU&S3$6@o;NoQKVpca2jIKHzfwXeV=3xOz^_o4v>BPA0KN z3)s8bD1dG0|JVkkq^~6}*sna;thx!Q=Jy*lvTH8R-A1HSjbInQt1C?HG!_lucYUHQ zWLg&Q>W^$-EF2X`raHfq-1N}*>E*3wYj!L-1IGEPHWF8Nc(Q>|f~{q)xS}d#SLp3| z5_JL7|68|KfQiGTURYdiJJA+ z)5>@EC7g>E94mVW8p{uV>d$lN2ZQll*m4ImV=7-6;=%w4$rK*=t52qhbNgx1EXHR2 z%t(=*>x1LX?Z{q?lC_{RJ$H50pHAj3AP=RZivjWy7DXF_%$MY|?I+_jyyfAAmH|-M z1(L>#f%%meoo@#~hBV2zH`MA!C6T%;gEvr*Ii3uW`0!UpM|o_+rOU0~k$gkaT3W!t z`i4J&L?sSLRIbfW5_4tBsxE=eQ9pA?i;6Af_W%+@3gqU}K+hI4nX)%f-%%3kOc3p| z>*q-gxXdyn@xfPa>pKtqG_YEW{S6bE5wn%hjxJUjN&zV~Hk#0+sQ~%HO z>2Kttq?9g;&vk1-r)SHN>?Yq05A^%r5cVx%|L%)H_)|ew=k0ReOPrrnhgfN^aO1MN zF@G{(e17Q%nXj2EIKaDoJ)n7~@N&In8MyKv4;U+cs_ALMeb#cK)TEKBF-t6#SHZY7 zo^Kj8-W^$St~~}zKV|hecRY$; z@D-Nbv!*f*e!!b%l4T)FFEKmmozog%4Z0>@45%WI0@j@-`_pdW(;-(c#A0vv4)Psm z8ohZxK4AGrgHLT-Mi*O2Yara6ad&YaYS8s{4kB-(g57SrD6p8C4I84}8GiyY1PdX%N=h^Lg829R-jC})O zzTTX6j6TyN|F89kW&hHC}Zwczw74!9DujWQp-ar(}`l=XFp#hnvXc#$2G zQ}<$wdR#6xD1W(zE@}A-`TXv!a!FB9c7E?YP0g8)R^lAG8+USJq2qZ&`D`pRC3z}i z(L8HxSLq?)ucGf<2U*a`<;$zFhl-RFU9aFitMWzNTc+3w3{EOs`wP%fTHu(?W!B?S zi8|@~rgjFHa5$B@i?%zM6EV5=(gPIoCGWgle%uvYFkv~85S)Cx zd-G^=vtRJ|lQ!FqG~`rDr_pnf;BruNH-d=Z%51KfpPEOe0ILyBI z9x*dMVxBp=k{`dhyjnr}ft-3qpwPn8QK^sueaJ&!eb*$m3|z*nwmc=3)NwvE)XFP$H0Lg!i^D ze(az_CP)3YM5yup{Z#KwWz}BqwTI4w?#;uWqp5{za{t=3ciN?mwr8%cmoLQY(9hg%tQUUHv?)ichX`7pZ5 zj0K-k*Os*Sid1-&{*7Tc-O$v|19`DaOtt)-TP_tDuo9TLo$~eajU3zk_HoX~z50>e zYtKov$>BW=p2!OEBQc*@%W$Y6cmPZA8zGw(hbvr*H(!)i%e5mO|H?|?t4okul+Qa^%9Wjn{&D}$q4@9-BcB=1piZjUH8y5e@{-4b0X`bHM^QMMKDRTf>K!lb%hlq@Lq)w=T| zO_R%=$?ei%)!Wh4T^vWVzh`{iPR_ML3as}Vi@*#@-9QZukrv`m^)63)UMf#(0}gNZ z-dd}!QS@-7bym+%+T*2g7z_;u5k6BP9ij|y^*~Gn@MaYW zf7=!D!yE17gg8HN;K;TQk@Z7SsbpIR^!ij%>)^61dgBqNp%<0X#=d}#@>|Mv@3 z%QrUkWUkE(FY-1a7(jsYYn>t=F*(S7JhXcBQrDINSvR3x0T05>ES{Fw)GbxX?L2`p zNJ5eU#!ZP0M#;nX)_$DVZmIm@v`(5q8!;mHd=fm2=Xt>qh9$>;KnKn*ErPwv(!Qy% zC`#I`B89!%RbCd6sIk)=MKXcLQ~+p*B+M^3`I;>$h#ux!E8krgU%Q(78@;j!Syn78 z7kCJOj`&b;poYZM<`$i4HxEa`nCC ziWL_0UC&oM4-X-j`-x`_q2FLnEo)j#8UGPiM;RqTwu+9Ssrfu+v8mZ2D+*v&X)~;N zp4xJ?T(4g%3qVw9A+hOagjUd?$p+#{BESEG19;XScy_K`bg`7S_i(B7wSo=sAr%DR zts)WTJQP74UEog!#6+@Z8-x#qy18iSHL3L<$`=DPHC0w~_1JsY;Pwo=ppFMB0MrvO z-X_5DNdaNSPg3O9N(m#Jz$ifW7Q(eB>Ci!dJnJ7?uM3J=Y|=%_Spz6miUmO+*ie?# zueBZWC(cKaCxP8(1K|JvNl1VXNQ9c`&`H>5ZL!gX9NIx;=`z`K9d)6`g8zrx|y#Q zt`UsB)1K0^o zU}KN}{AjBn05Bs=JTwC2)F&?D2SG5hsAZt7-alk8{R|TMFCh{j|H}WnoeB9UKWqpY zNjrDYkD>orAPb20uvPCGPD(gl8dk+is*{Zn$5_?@2iI!B%oXhoYS$rny8sknWt4}f ze*R#kOf&H=zi>pNkm2)+81LItV-TAo@`DTDxPn?@+5lR4#H;aHT)}=Is`I zEx75VuEK}aODa1cmZ-;npbAGj5$r1fC-ksDMpV3l7iU9lOQ~1KYh--zAwLU`3aWVnW#R z4F$&+031hlo$|kMn-c=^TwN@|UMyJNz+T*|xDo zYfl2Ki8wYMp^`X7G6dM80RiXI_pXaZ?adh zNC;S1vuO%ih@=0f63`Hyy-=jALxC=D!1j^yo`wbBLST6mZ48TZ-zd8t^y?K~MECRFbn?MlFD&S2D!pg>E{ zFCu}YX7^t}uUSXqbHLCL<6?~83%IYs(t4I7zRb09^y_Q)Lf3P&+U5J&g$m*cDcnt# zeI@&scx^n7gq2pQYF;s)=Bu=2yhZa^<$dL#3fQs$VmeHZLAy^nSS0k*q@6yh!l5RD zmI6W)h^W1Lrh#d`q5LmARDc1+W!sXb-KZ$v8{A*Y`mKqMX5(^wM$6OqajdAJ zFJUnbcC_DZK>+`3l5a*(PA#NlU&_U1I0Y4 z%Y%aQ{nRn2241VKNP!&Lp|c_#wQCMxe)QW}Q*TJHe$OCqN8$9QH5 z$&-NA%~=?AQZp0Qcc?5IW~H1HELPVwXE{vPJ<&@Y)%Zpv*bi*c6wQ{a1u!kyH$<}j z5cWrh3&Jbxg0P?9`qb)AP8li&J{a5`-G11o*|HmJl5Si6wCPllrl8Vhd2m5k5^W1* z3Oti@X|U49_4~h_0Ikqm>ik;2LJTn?qyLkW+UI_z)rMuAoIFOpx!Z6zR*vfsU^dgD!n6|++ z+kX=!lnJO=MDu97u-;Y;Idz^G*b#W_@&&o`6v}lCS~L8jm01=`Z{sP)!StuMZA*9< zUbILS{qX540Z;b99J>Xd1-2_#IvRl06wPAe$=c39L-ecRvu#|HW@bE5_+97kH2u^t z{PnYWQN0vD|8c=jr1WADRr4S+>>u0z980nv|2U%>g`&mSVp*zz#oXc27YtCr9={GY zhsNKxYX_FLWnh(!Oi<;Uju{p&{~7r(gdisFq#m-2zE4XFf*?%(m!p*QyUOdW$E0Ee z8-epZk;&ERqN~$&C<)kr;W{PDwtm(%9at9W1gFS#di#RN<%p)0p5C{DzsQbtt(d9# z)5t_+>z>m9&JPU?5>OUp<1!X_?38JZ?%CfkY4A2do zt6+D!EAeyj8ZV7jJWKnJwIuc(%4?qmW9_&dMNthsmUT$7)S38 z++pa#Nj>WiHyp%S;u|wu4*k?2-&+=znb^N*WLeco^H&n3l+68uk2vE7WJU1u}oz7Df0DAwe_=v5mgC#WROW)Bbq^ z*QGZNir}1sL(O(22Tca(`JvSwj;mEASa}_-_lj-k(>}Slq!keha%pPj{QHAV)f!MU5`DI(gRr@(m))FHa$_Cqx-*Lf7f%Gj)GwgdD7vK;)w!Y<@Z6)tQX}KKTJg4Zz8aMMpwIx%d_Q4r*sJm&}r3 z-Z1gWE-hkv#2IbKR>d-^acdmw58Sn}eqo%ID)<93tKqUV1DGVXL zl!j(5o9@)DmIB%nP2o{j>4)W}Lz~g2jK!%SUUO^p`wUYIZ_?gCf)VHle9p}oWtf5l zOPh+?uj3cvG`~5qp-&;Hq}DHtCwxd?5FwJo7MUVqMT)mULHy-M3ao#!$*T{UvKv&a zVGmZSqj@=NXqi$dN!z%F-qps~6ox8!WNZ`>uLNJQfdu*%vsE_MV#hJG#{UV>bojFX zX?i6exxyt?H#iYmO}6mDRC{;4nkJ)N*BN5fNH{Tj{~33lWOn6E^3kAbM_7L3WaM<@ zT%9pm;}FMyqh0KZvotIyjEyvxWWyQ%w>cE7vIg-0*C)yh~G7lm1LQcdXOW7tqU$Pr17VD-WRapDzhv^jDou#B z-3Rc;uJk}qHs6l7O$kAoU~G5cDY8z`I_V0z#e6QZsAAV=T&hfS4jB`*ysvv4MlgpA zN1seTjbp3aP8@IGbq^Kcnt1@%737|a2vVv$L1I@ognaiWGs+5ybsBpqj1qGYzOrx@ z!3@{gfZ;5O-Zu;y-B~uuVz&1?1~O(*{C49t{Kz4QpX&ai0=|qH{)ZhkiOrdul5qUP zWLB#2KM@NBY%nA;31uz%nDmW*-0{-yYmYiCeMoDLZLRM6I59chAU^rP6|5Y54kTZ8 z!o+BU7XP*+@^&VfNLi7d8Kr`V*NKj8nD?E|Kd&=LYnRIG%!8Cx9gA_?BS)u9fS6G& zx4wWr*&krVpebXl5`=)(2M|4Bk6ovqlLg6R_(*X)a(uSNS(}iU(PeDkoE?dbMWFd} z5+lo+)`=MT78;u@rq1_UgmVme!B>j8db9|e14h25=#%Xq`1AIc=7ySpII*Ks#<~)s z&afaW|D_8ezH-iOlg}f2^RnQF7Z>NDYHA<11*V1W=iyG&zi33HVfJ&)7fdd_)ppUw z;UMXI_Ts_!c3nCPVKsk{`9du9RJ4cT%>=T_rpkDds?&e#((;iqy#<24O#unr=P2_h zKZF}56uB=?N`U0~zz z(MtP&wZ#H~f7WR{Owz#<20&qpF;NiSdk5AW{yPFx&>bxdD&I+pAShN!UoHsbl*C5p zM_C*<5Ek^E`UOK@CknP$BRZi!*zt`W)Yo~(FK7p2zhpOh;%`kGoVjpLVjT)vaYGpm z4{Hsf0VP)mOwkb3`2YwMLqNRXQ}x4<2%_UT5rB6HArZDY+o3q&-cITHra>bfQ9;um zuxdi006t*|%Z5v36wf2`Y%S=djmRf}(?$`nYb8e)e9`08k(`-+`E{WCGkGqGW!?Hp6k-{0`3Wn$0gp{S zLL9R&>vJUcOQUttAu){?q}chzk(CJC6mwgg{(fx5{cWV{pmC|_~xl=%An>%GF8KiXhu$8kXLUCfe&&&>IX7h$1_Tu;O!-K^ zRvehAM8Jb&U0{&Gipjj9&a&5^2A`AhsRxCE(ChZu>g6O!uv-0)f{5<$39}Ve$q*Zn z$$B=_Cil=S>=y-NS9#P}tp(A4%_vcXa(5Fa^}>iW#C52+2h;@=ncVK+f%E8VV1Q^( zeJlbT0CD9%ft-eFU=N5z+qe-exC{k5iQFlKua_5mA80T$)v9WB=`#A@wrP?o=7u8P zuQ||7@2Tbr<+8#CF|C7RNXedONIJuWK&^T;XsiZ|YO% zk!r;rVEd7D6QUKw{V_6_Ydee(!(wuV{WAL79C>Z?`KNvl56V|-_sHAWp2UyT3+AsWH#Ovf=a6(yp>=nZWxYvv$EB{qM>Heit%9bA~ ztwkU@Q=+0Ff(`u&daq-(HREo%R^=}fA17ZH7q7@G`0o|W%h+kN0@_<|vXq^9ulu$q zS1>8Njgvq5ClZv48#HE$kBCIrxJ<1o7tv8Cv%d(Q!|>13$NcbH^K=H0`@xhl5pKx1 z!T|*=8oceWMuelE0WWPBilE%Wh@`?+|C53RAk z2sO3T)wqom?&*40D=8my;6_xQZ)Q_cpXF_S!GySFnH$Y{B|KbQkke#&{Bn!r1MhU> zADP0E@-pdw8j=`KdT|Ok@iR}B4`!d@hW_QaF8{*!>uUJ{S4feeICnx)1ol@Yj;aUm zix>PaB!mSNu?WJTG4Ol|NS&c~aC5cT`I6h9iSC9QLoJ*@ODG3NcdH+RfBuxaYHwo> z!aDvSkVPv?+O{c-G}TN)s39DIAAYt)eYaAW0y5peKH(LN&qvlwb3wDfCMcQyV@|PT z=ukB|Aov*$<@}@HzMsX8#p^)RJk8m#_0WaPrd$fm`E0%+a7$o~1~*EeHIyAjlDGkq1d!Z9EsAVE`7#m;V8 z1hU7u!wb;C6OXDO^4@G2RYCKOMPaoCpm3RT}`g~f2as~^g zIJg6o8^Jwzp`QzWle~>mx#H!;?-u7nVS4l1Pc5eE7;66LFjB}EP)1i-vZ(G*%7T@7 zVbyh}5B*)2qmuO5UJ3AJeN|;IyHyPa?p9PcNIXB*8@1*`iIQV%nJZLpa3X-9UYg+? z=6#dtO7F^F*fck?wXa#=;Mk%Dvy0- zK4k=+1S;>siT8W1+(E6II%Pt)q{ik5x*sU6<_=e*6J8GdW+oB(KB^cy{=r%T;y&L* zhkI0*3x81jYxE3+K1Ys&i!AW9WKXPBM51i)g=HCBl}y`6DZt3L9(woC(=~!z%?>ZB zy!n;RlsFeoANDt%v$YMa47z3N+$Yx{FWCKbOiV+ia*C|Pflu$RJJ)nS@}q;94A#03 z{5o8W-p+xrzd71gh8Xz&KV*INTT~CQw{&-RBOndZE#0+rcQ;5b9nv7uAq`81(w&mh z-Q6t>@1oy(f4J{I0MB#IoSFI5EYUY@*s&ST;j%itu8KK8Njl(V<4+#}tcLrd(3DL5 zV>xG~cN@4*gLVKCd)*lA#tNc=otXQCkuyN8fce#*z9M%j3OpKK0xcyHaDN zALj$9L@lzL7Yx~6Hnz3#I{>*1PZqSz#t@LVsLzKHbYBMcC4HM<#;{8l$*G)A zr8OZS68UW&jZVup9YaDYIPomL)ZnxZAN!L9VquR|CfF|sZzWw& zzR1S*56BnMd5N%{5Y3_dQQ;a}97NGjjpW5dk>^nbslxZ92(7C=@0EUjp#GsmrtMwP zMnq52o~e3^KgY&Sw;1%z$mP>Fgw_w%<0>F{_MwY(-TWa2%s#CLD_;!H>Ob3oZ0V6C z(!pfRS$CNj;7?(fF*yh!P!9YpG(&$8_^(|NzOCT2(qr`xiY>d_aAEwC3Zc>f{qP|L z7s5Yop@9bO3UYoA9|@ZiHsiHa@w?zfxFKEqK-}UZ#3?!VHccrRc`Px&kA9du zXp)SR3KVIvGNgmV@83U=^lTq_b&q>rhbzI1F!PS{buyk5b%_g85}hH zH>LY2!EM7KXgj)MAA6i!lv%w=8h}ZeNXgybL;YS3QAgag@jNGSpj<9HHmyxUnXo-y zvRdI8E}<9KZnU5zqS`!G1#_xiS%PPQxra!`;NJN0`&B{06nP0UnQ_?Gq5Qx90{XY2 z0dV}9s-WIhfuoZA%VkAqjlPW_ON@^VH~Br^U8@s|FIpq*t7|}e=EH4-TMQ0LX~Y2@ zLXGo-*vG^G4nNzXObG&b+(@(38ms51>2TI>1_IK-R|BOhKCpp@ka$|mzU0!*haIAO ztUvxZ7Ikf8O)myp&Fs1bWFH0)4`wcyBCB1-Pu=DdSE1$(4FOJFAhZ;{V^{g7hPjVG zxt&B1L}*-j3`);kANM4A#s}<0S*0l6Av^P6GNGw`K?>qi@Y;o_$rub$o9S5DYB_W1n0t+q)Y)xHJK6} zOm&GR9q@W=52m>Y=V>n2Zza0VO2 zidiM$7_YKtjUH$D=upv}Xz-FSm+C+-C{p<&ZJc+j))XDK+O^k+Tz-c)lM6Lh{>v#| zNsK{E>Qrn#wMB_v6E`bYk~!!fTfS*3XPV6l!j&lPM{}Z zHSoe7zw&7R>c`V;w9ilEfwpV9YR3M)D3@rK9%&Gt{hM5a0#RW%3;R~5KS~=BYcJrZ zN5OR(%Gzhjw!;Tp_7d#wG`{O&xpxn+o`v$DRaGfDbOJRv?AmhPAN^#Zo%xxL)_vy> zlh_D!n0lfq7nKME^;y3>6yr?U8G&XM*31`L9mRgH1Hir*&W)M(YvU)+T&1xAi~+`qI-`l$t;wlMmw<8y_;aC@b+ zJ3=DQZJ?h=b7k2KwTHLZqTj7=)XvH*oX&gzzwoQtca_#!rbF#fC}3}QE9k|XqI@39 zK!hX7kfmq*$b=}qUedAU1|7tXt?~sjQ9f2gPe!ZtbJp!ov#)7uvP;Z}1^MV&aL}MN z?`FTPVdU4L5W@?Y%e`ort!WMn(>twxrmgTO3aXn8tr>J@b1_(F?SX#uH)%jp{o`!? zkQI2$EUjQgQaDmnvqHLqsd#Wb z@54G-mp_wr0Ld{7*^)Wq6f1xf-JjAkY3Y}MezheCrnUq$F5greH?+SPF$8P`0573O zw7gduT*&H#MQq+fAt1Z*OTHKdeoB@1a1I5qFsa;C&Km-SkrNp)UJpr>VNvQ1W@t$2 z3A8;Q0f9Dql+g#BGt- z<-a}1Wp^eJKF%2eOf0&2d2T>oN@aRw_<1VE+B2vW z*^odM@O~};JRWS{>1!K&czgVFI@;##fVcXWJn`k{W`>=&I=39oeINp61xy0TQR)un z?ZpAb=yV!gp%i{|5)!HE4lg8^8jUDw4% zUEo?bUJl~cDW3LKx)8dG3$0#ajCV5I8cD=&`TEa46j%dZ1OC?lBxZ#9wgxoIj1_o? z`dddgs~qN+nAjjJ7O_fNSF;BZR4L znZ|urNya^{6Do!}UY0)z-s0r(B?+3L5D7GNlK7yl*j^%1a4?_}s~M$io;h5+T80Q< z^dsXz=J{bZk|#D25%apGWunp?HmoPeu`M%+!NT|6MjU`c(G)831za|mNN-M*H%S!r zjXrSynkfU|+`_`^LH|P)Gi_I%ZZ1ppjlNMvXGKZ&v$j88Cb<|zpfLqvkwbKQS06e^ z3!Hqlqe4$6VXcub^U7zu7LRKs{elLsUFH+uL*>NRM{4SP7`~nlo!H4?XYisW5 zS8xivz*)|c9E_0m-AL>}xFEmtq*#@KE*A4e8X-#rW0!19s-X~x>dXN*;ZID3F-dSk zouOMEvO{@$M+U(RW$?N|A}hqP9)2U4BDy(GUBI7NnFKanva=9kb;Q;A&l`XI=p)O= z+Q-9?(a3S{k>8)%pakfiK|{`%!mFc~R-*!1KBImp=I);qB~HYO8d^*$Rp{m(`AKOh zzR`FmdWa5RZWx*X2*)@6;ZKxb5gRWgsn144Yt?c=ALr6q1`inN=R-(TGUe-cNXde- zCYoG3`t`)9v}}o8akqhj_YxLzqV;rKjop71bi@|3et`+EKIOCPeg|M^=O!OCGs zlS3khjy7TDa`(BLGZbv80+Cd3k{F_ba!6FQD}F)#J)^flyU~u^?9;D&d$G(sCc*6e zkS>9~O~ZPTZ|A)|@bHPn!>OU>zpN=Eup-RPWWaw)>EM_i*#-XpT=;gA%7W#a67u#FA7;Y=k5SGLbZzUh(Rdvk(u> zKD`54*TyrlBNE0u)Y7b|*KNcU!cNds5kf0D@R!pwgHu=AQmMNeT+ocd-3gxW&-$LQ zI>H@KcO<~JVehc5_9nV-Vj)69ZV<^Z=UEGc+(Pxc?53X55YPtKY#KT|xGRADAOFev zdVhxpHSY{!x-&u?xe&9BIt&ATDW`~W?IMNlLg9gqLWMyC!7o;ed&~TH>qTrwZ5^Jy z-@f<;n#bCke21@|PH6Nb2qbf^Vbs=XWo-9M_AXXZJvy!&ca2RMYz?|6Wo6;@9w%vw zJQ#BBjWYJ!{A3+<^QBhue#`fTvR2P-01#bLySbAmfv^WQ6UT?4EhSl*v8js9xQ)2E z6Je}#q!>a9sXiP)vZyQ8e7{{-`0BRkpFB2Bp}H_w>MCIUyXVLF>NNZ>Y24!8?a3)9 z((A2;@|YCrCpY^f0QbqpLj9VcBHT)bwg0Qe;y}<^7H*CbM!vje<H(`RbjxlCL-NN|&=n$bj@WCL z>z3KUwP|#Lcal!q)O6Td3T3HgaX*`*32spt#HkX%o4K&?PZmN*=miUWLJd2Y~zvzi~d(P1H!=K>Xe^9r-)>F)5>)o&^9GSZ=BwNdZ#T0C)P5^Dk0&b zpdgc1To_-)uHK&QS41)tPda|(WO$T1ICx$V=SQb#{0y3%uw4&N#Y7jFD1tJp9*JT6 z=`4Ww4KM8oM$PhjSSc!d!Edh(S<0ag~kQ@HClYSHt8GDQzO7rmHGGyRN?rui^iw5 zn$%^1`PD!ZNq-)^D?)!r5v-WN{s2MC$E10Dhz_EBwNdDfgQ-rC>(4P_G3oUpG)nJK z^B{kk=R^Mqs}&9++)yp`Vk;(16TkCyD4t)SXML~2Za>V7A1O`cnlBBtXWo9j9d0Z@ zu^67J2${qDg2vi?7{kfa3gv*G4;HjAg83^&y}Lf<&vzvS5vc{PDW~HuS{@tKJ7=Uo z?o^yJx!9+>PkG{<2|LU@#-AR(uhdnxFe9A@*h%gNmLpY4jm`DW);~{+XC)1IytM-U zCBTn=#)CJ-`>jMelc0-D^d{#E0W!g^=KG^5)dasomHV;{h*fqZk2#^d+o#KN`)4@I zZNj$PH%(|5#$+U5$pwO=ic?RYNN005IWga-N_OLInwQ%!@U+!mxL>_H>bquJ(IB$b z{DEfJ&xm-;3Av9N`%&Ajy zgfDe*28Pq{x1>u8k^pd~24xvFq}mem@Rv{Zp}`qwfSbI%@Tq^D8O;%x;yl?SmsE_; zdvm?*z{l$k9T;h=n0w0twaD>)t~PpJ?}zJ4uU|)74k|@C*{5B516T?E&n|Cg|#R!oO+Jan}I~<9>W82wJ2irmzk2dU^^_^YPH{@57 zJ&wgksvCrsVP?iIG$z5fGAK_kx2YtXa)_CiwXm z=n6L>UJ_nhREw+2JevQM)WaXlH$eN^+htj~8rx>Z?b~dQM3|D1zVSV@Ebh+g)|zxthEbD_lknV}chW5d*~IP_gYF z3|tKcn;(89PH>wSW{s?EP6@psVbNo@rFmDm$u}iMi3tJ?4vzXTp0OGNNK-fqPhq$d@=8iLlLID9xV7E+Rkqif*f$o< zK+GyPdYyYWyKkYC+lH3f)K-rj89=xi;kn;_*eKhY0q@lA?#SP|Ehv8fujB6D*H4i< z1at^4Q=f=fGk!pM#a+_gkbvt8-dKB<{^I3*XnZvAZDC-r7&jMj=3;@-NRD7&vm|$Z zP=>X8kPlt#yLX`V3oBn+-vZ0X_lqHBJ5{gx$pW^euw>!}I~q1W%IyNC6eda?ktGRR zt>Z>J_sb;N#X(h5ln;tHarCK6#QHExAB-bANoSYXa}OZ}vZq9KeIo&Ma%m89Y8C?h zI18GgAg8#+Wz#%70}sdZmrhIdYDz|pSUv2pnhfNcwD1rRq$KZKegiyH|2IFuit_fs zXHge3sD_GKhf#m_$}nG+Yy^;6i&PnuxU6FIds|OU#iL@A|NcF~-DJEyLU>3FwXh359^hVl>J1jE(h+kh_Lz-_s_?!1Vvx(F>-V z2HWPW59Zy_&4G+w&aX}n2vBrRVie)0#S8PEf(b&K;O{zSt}Hg~a`e4e< z%`M4ct=;~1l~TN!w>7Mz^hsl>zp7kNZHz7GaO-;i{xS9N%$To*M!&?#ls{P| zdb(aDnqb;8H8EXE%t@%U64o0`PHcKIpTl0u{I27*`XO zq?1LL%-^0(zqrs6SZwjQ^~f(>{P;yAi#V1`W&Xatt7rCh?GW^ltu5@&hpf-cir;Lp znod;u;{@o@n_61A;yzr0BuesV#PUrB$JURr_VA6}Y;2*?ej``~OdQxpE8a2#9ll=c zrY4e5gi9L5RJ`L@-@Qov*3_K=Ba)0)jQ<>K`q>eU67!x2s+y4pa++cz;@3`0t{tHj zztqrmH4}gIQkVT}2?VdvEqoA8FN2%KaRZusNpq(WNXX*i{H8F!Ucj5LF12(HwQuZ5 z6=p4pDylvrOMm%&BLNc7%Av#Zgm0?8S!dHisXv;28>nEXu%x6gE#qbP9?S{-tkVa4 z_^VM5y)nsiGA>pEm#)qmNHxtH(w}&V21V?FAXfqt!tXOl473M3ooc?RsG$t4xF0uy zU>Rc;B%*ief2E{oY1O4~4E7G7sURgoI4DvDmNq?Tg)?l-hlzm6Mo>my=dd=FXuln0 zj7`a;e%9*n7nm5Y=z{J=*3y<7oky*fL;N7XBdu(}TqiZltRjz2+Hie$-=FeC&G%q3 zOC!r5Tbe4Cru3SN+>4%jyNC1Y+3Y^Ch5~6EiNe+6YTLG#XHH2B@a3|y#goPFdP3?T z?)5Hi>foSM5J~9ue!O_x@6^>JXYNIykzu`cZ_5kv{jejcCv$6~XDX?XADYiQ{?L3xtHof%BGS$r7&b|HfO8-b9|tuiMZ~ zY2bt#B`=bg+s-;?c?K5fDYS;@!eN~#A7VioTs&FAJ)HiZMP*qd5e5G&vi^*bGg~yh zh(`GR&2R`-E)5C+6F&+ee`T^e7P~gwz!hI-X9|K;w7Of)XqT8e(O8l@5$dONyykoO zG)~XMd==X4I79#~_O8H|(3d9=7(>a3PMTQ>j0`Rvc1KdQhA;%n^OJWz-Zl6?HcUM~ z-XaOa#>GvsJWsL|3SOJ&x5PBG_%SV9;jdY?TJ0Bqz207sI_Uc4bFhtUyGYYIUw!sz z=W$qfE$`vButXCWd(QeIA%#&ZBqh+YNh70}1}Si)mW;lplj$^aKssC9?0)C%eZPxV z*mEj(zQO0S92vQz?p6W?Cj8yMmvC|+yb=Bi-IC_|j?jnOPW4qitIYO(520E5FJk-* z0%$+u{l6L8^2a&le#b8HR1EBvcOEVtLE1#V+9m4^pU^A%5wH`f|BHV9;`dx$8)J~X z903ptzi_uzJKm}Y)gfr}8*0kWNTozRcpdhjOkKM3`TESPrXI9rU4^8N!V&=_troYVk3}enCYI(X7`_duh=A8!%Qh1f&L#LI*HHl0Ox3rNYjro&JL>|mN#Dr z1SN@!$S45?Bm?kjw_s06>FMQ%ww1S}S=WH86mK)uI`$sGdk>TljuC}L{fogMeoZ^2fqCM(H62bsMSN-r|nMc zfG3$w@(nTXr|ccpjSMACp9uEvdMR8!S`tlKWO!wK`vBh%_df=07oQI9BJ#^nm7nw= zG_-W>|0M4X)y*;Mk@11(!K~b*TOI!O)dVL)Wj`)o;1A`!3I46n@p<|ju z%@Zg@jmJ{g2o>eyb1f(NI6V4+iOAH9V%DZ(V*ZvvPIzyy;5itw9d2H;?ecY`@ZrTdVYFSlj5p?GM`N(K3jR!4HP{@}_Mu{tp13rB zq2}aGb6QX84L(x8SxVJFI@0{%=nFX|m&9^V7A6?u|Gp|!t2{LsO}o>7;!jMeh%{vw zn4pH1o0H4D8od9yEUE3e;sVDQ%~#1NzB-SyDOZ9h?!N7MSZ4Ta<9!Kz)D`&qa`nKY z<>2h~xx;b5UBW6r6E@D!H}zEb9;_K#4#H$p^jWNt`Ae#|%iftJS-LIU;-H3Y5Af~d zekq^FF#qMKkXNrE#^1%9QbC8%dTv8?(hxiOfm7*Tx8t`PgnOJ3f?QcSm>{5T5d`W2U}Fp|jeQ*CfmX5TTBI zY#fh0iAc1{-(F!P4j7Q#F=zZEK~s)wK(9ku9bn~Zw5=Y;p=VUFu6WiImq!>m95khK ziRv$eLdFN#MfCAzJMx@^c%2Qq1QriHpIp@$fgw305yMRI&^+}4&5~=1yn-w_(PV}^ z>x=?91oWJd*}vk;m}Y+eg0C#8n5>9#zmY>>AA>e3^}Zi$K9qDZ!XMzUax~jASrG^^Bi# zka!zY#Q$N`yb5gCXVO*d!vasT|D&Q&^BPm5A^^$-k^7CIxW+7qzFX*ZGycS1z-NLW zQD|soAMMjSdXp2%L@Q&7?iS{v%@Sb}@$OKnQD+wNIX!{gKfnY$5*7K7isDx}qBA!t zadHc7lL5~CfovddGVpXOb%V0U0vbc8QioA*3vepxv6y;_tOn_U8auV#Zy7KFF&W}7 zij598r5vwi;FO?$WOESGnP9}HjZP$=gSH$Bs%;?p0^J}}Kpi0+zeG+cPz6kd-H;S> zdyw&%3O`weaLy&i=pF2~Zg7P^Z@*s;Mw>i)eL2g}>N|L8f7)RwZdd~@ou0_tZXO0+ zo)!2PyExpfUOxJ<@yMX%__GLhf)4{uw>Q0@82Xh~euSG9#`}(B{LD40(&&0rr`3bF z$>;sx!XWzE`i5mf7pC2>UI^tqF4>PMu z5ErmGg9mX5QQ?AygFI5yz~hdM@&q=PwtG=6_-mZn(Y0w~PEQUAJb`WFa< z7xZ{(eOg9>*GRVfg$Qg3_^XdqV<-4cK(lrrzUwCVNu6`;Lt~+>(4e%#F`s<_fh!w zd%_-x<9lNHVz9S_UD8Bxlf4u{(h?ft-buFO>V4|&8FL8eF;x=sNSPvf&b&P_yhdqk zorr`Q#(B(ly~r5#Q%?90&A@vB$x1JAlR-~-XgxS;nw2GTk*d78Rgz4yO@^ln%af>8 ze?K6(1{=qV*9UT;%o&$&G{hb?Jdz9)%J+1|E({SH@jaHJLY`*YAgSh)%)U|w=#UHy zu<;%>?2mZB?eCnIgPfu&zZ2U7teiXXho+pzrl4s9GBIbU6hocWIlSP#15@y{bh-?7 zTNQ37HM(o7KWm4H@to7VoY^7XQoOh8iNc59;3ijX^LPxOih&X+W(}n`kW#NjV}$G( zqTwz6FH549@d=LzY+7$=31py;A`u0z;&@zcfP%w`vHq$LF+Qi(uG>?LkY}!YoKRs9NnTpJHJT}LwiFwC`J8{NzQS}H zI&pq0-ev0aW_^z~C=>`44LV#M48nn@bv7fKKQE?q|BOF+wauW}-Q)6enNe?S&goP5 zchVxg)6Wi+9lH2V?pRzUrKZy1`OR9giMrcaGDZc21lot$Fo-eyUqp$Y&ht>ul)%u2=sN0^uxv3WFk!HRu>4)u{Q+ zBBeI%VKaw#a-VRSp^>1lP`OtzMv&d~>ZopKwj%{-`BS9?#c(cni8tt=w$b7t3+!-= zeyy+hBtPsIN>NVtr>kRExD-CIHbs1E-W5r%lqyD}4~)sHZRY0^Y9n*8V2h>`FUXW6 zSlNsl?)_?dn7W*cf?4~DiF@WJ_=8P}xA`ovL?nW`v^S9l8x$dEJ1jgtb#-|_JLD^+ z&q#oPAWWeSBT9h6e=0BtAJ^BW(by(WmuQdN^B7L_D^60K0Gq#~?h%sC_bOvezT?)X zBSq+@RH@MLF4U0jw0Lu23SRZzg6VQ-O@F!Ni8<>gyTB)|BY`x>H)-qwHuwmKHN@|V z;F}}nPft_(wy>sE&-RO(QPDl{wUSoeNlb=28Q!N9tSoqcg>1}x4_k*gJ@v` z-P;Zpm8`T8+DZMkgBfdThBDMuS=;}P?Z*ugBbI^*^*Z)c8nXYq3y4yN(J!*kcWEw- zLUqz$PHr(F^0E||g(UG)Boq1G&DUia*ERk-u$ov;4d+mVT&n?n`V#lGxKosFbV**}C*gD3j!g_#(P%&C5-mJm8#H|n2jZeGwy}#d z`(&)X()VLX@1nv@qOgV+e6yA)&l;f(`6{TWdOHmsXQr3@anvwKJ5$5}?XmeBld|G> zeHyzWIdnuO$X@^gaTpqOpLQLcqw9-EdfMYyqk3a_hKa%cv}F2xa0~%VEz8~p{tMy# z6Lu97KZ*5>pGPp@=lL%1+2O?X^&)E>%LcO(RdM>%EfGiB-#{{*6+pPTu`MZt^e`7H`#Cj5E% zQ0o4Kz*y-6of~NafCmf?6Lz{10<(PSkGg|-p>ucGgzPMv_f~?m1}08LSFyZEK3qwe z1P1L*O-}n?LAF9|v|q$HrYhge>N*zYyx!-Gk0&32cshcg#4clZK6f}$&a6m*1$Jhf z=6Ca-zm1Ck6}H;jD<`=U9Dxi3v>JxTBWP=`Lhh(b+>=Bs&sNP|cL*aPP|j`h9uQf1 zwLf_O<=KB0sfo!^z7dmVY&Z(=dSY+w%fpRAT~v!g_kv9C6833x({d)xttC}Cd&=?8 zS*0u2p7RYnfR1W7D9&BHXWHK$wN$L@lhekH244^3@wjR?_Hi(W6Q|-;ihU`0VmtFl097CVB$_6Qca9F*eKKwlQ6onlV-Jbq@%f#tPbxpH3=FD|66_QuYy z+RAkR`T?E#2}%Ce80l=|@4-nA!A zNzFEO>R;$u_8;EMX){fJaLtQqM?TWLT6JKo`nsZ&35yN{IPi?O^<7uFzFEk|niOSEY&qTO`Z%MK#rOIjnJg>r2q zo6TV6?XKh7I3-xx>cUH3kE!US%48sA}o<{pYay|+N{KPN5wK_h$#}`qy$MUUfywW+I z2lf$wobEvoGvxXG0G`@>MG#d&L9WuWlM z6~VXO^;EH5j1IDrnnJ6Feg8p9*$a-N;S9HKC!U(CVG6A3f$154lJceLgLpZoEZW6) zbi%wUXA(Aoujo{z)sn1t+GS`o5kDbhFdOt}XQ!S{mRVw!mUgO{JiD#ZuvC){4V4ST zNHcfQ_K*p?28zO?c+(EV05mjm$WTO0@iWYGKXf7! z#6c;D+-$w!R6WWK21Ha11}}Dp!B~kw4c0l7jJ89j$5H#H#7&&{h=_3>h)<>@rY=8I z-Sg|e4p9H31qc0=Y*p=XOrhS~iGeXilz%nxO054%t6Ui-3a7{(p;Wutp+7xAAwMGs zM(mE?wsFIWopV_?ZaYj2ux)ArJ=xBci&$7hHz~GO3$e?6Cj3ecwqvMQT9)7wM2A+Q)kG!IcW!98klO>oLVAo{suO2W+Y% zXLEP`s3or%OfMvD0QeE~4tbR1g0J?RN*S{T-`Nl}LP4nyj`G#tJb6)5zV6=Ml)v(<_}!%( zV9YA5CuO8$6J1PFv`?peT|gx(v%JSfND-vc{sjJ2I+f1>Kd!-HBmM(?sR-h4bf)q9%)YfXRAfXFnZs{Cy=Vx1k!B#djMB z!(+UMPPs}k&7k+4gka}83eG8|;IC3eT^*?Njdn6*+|Nhe$C18bx4k+izHvqt*GkQZm)>w0wPbBI7vWqF>jh91Z#dKN>r?T_ET^ z7vPyjboWiMbqXcsUs zXBNIyg1vuED^63&_EtS-*L}t;r$is&O=emS@WuGHt$YUi9TnG!8VwcgulypQU>JaB zch6gY7Hz8v-@u6?x8rL4-a)y}8>25aQ;s`v1E6{v#D@=b&ny36M6D|$JXUq?^KIgv zM*qR_eajatV2Wk#8Lc;cguZ@>iC6pwFQ}qtQ>l8?V&j1awP3aKPwo$F`-@0-xt-N@ zTF3JvA(xQaPv?gF5zzM3SeDJ-<9jG1dB1+g_lR6!?C)~IwE!vO6q;Je@z&NMX}qt; zDw~0OwKYBg{XsEu3CY#)X-a3*nqD&T;&V%^nCi7t`#O|IRLm+)6PODlz30PCq3*|Z zgL>MT(~9Z7_aiS%AOoNq z;;n8@e-ioenJ$5KMN}ptyDiBrACRnh?sD(mwR;;!gbO*+ECRX= zI(Mb}jAXYK>ae7`C{r!x7kuk^Yqc{vjCY>wc4YS+$^oQBb38o{falW> z3Ifbj5q?h0xRvwubPUTlQ2#PsaCd1)3Ssj6eC99wJa+b^m;mqmI$|mG6?pmHAF~@d zWbESt*nD>{*@rxvKNjQV6y-**+IBU@e~|Tgt>y>L-DoFB{b04nvA}1%J%d^2?sSVE z50b+K)z7Qd?XG>>QZCC}Kczk|b^@C~%&7Z=imQ;TbC~5+%&FXg+UNEpRDPD8IHgbJ zrB1-?mvyukazVd`X@ho&70D&4rz4@*U9hK26MVpu*E@j64>7`+$i6dmX9&3jhf5aubQ`SCj} zGCseLjv*2tpS)AO9gb-VT;3fC9@c3SdEm+$;w^m5)7)xV*`l8ku5?!Qn4LPB?y)Ks zQh%=qHQ(jke*jM$AA}RtGq@d)4ewXog(|T(2)+6Xt<2oGW-H-V8jS?Ur|OBd=*0(S zFV@n!4AVAw)`YE#AOoknTO7G%agsPytjb(RISr+vy>9Ht^~$aCv$f?p_o0nNJ~LFkxJdtzVAchAdU*9Vfq{GFL!KHZdHA+dZ*qfeBgATZSahP;i7-rD7QGdWFjv)?j7sf#1JkOH zRVI89{GLs~5Tyc`ZgI>Wr&Q8-HD}g1{0;M8?B;TO#n#qTLQ|dLx01vwqvz(@lrb^` zjgDgFk@cQ>)LAkI!>{a+y-z-Cj|LB#uZM;UeVu7yGfL0XN;I*Ro479??8XE2_^~XU zcbM62=Dgnso;_oh8r>*YL(HNY*Pc&5+d~f(IBDM8@_+M5!druSTwTz9|NL85IBE5n zvZ$;DOw_R1uxAxNDs_B46?)X*JIm3fi!$NQv>}Rmmm}4uh8VjP&Qhq ztgqU{0`?^cw_asz;f{{}G!F%_e`>xv=oT_bN#fe`*ca^D%)V9lzS$Hd$w-Ql;h3i z)e;C?!h)*zW-tKD62V&l%%K;-QzS$>a&2=wA-e$&a=4PsTU)o|h=^@}$x)I9o?k2V(It$aO3c)~BO)gDS{O&}I_b{R24KbuU7zmAgr*4xkH zb>&KdHH#-#<`woZRD&ILjF62Bf^p-NwUym>T1?)B&QTq>|H}%Be z?VX;IE58q?vYA;csLt6Z>(F*Hba!?2Wy_6B3gnQf*3EPW<|AeMod*fh@xkrI{q{PC zkbD)mzyPdMV;5UF)=-g;GQTXk=4%5Sc4U5IBAKfU+U)C+k4^pd9%d9;(cM^Kk z{4rvmJnUj3TK#g~M=wt3BvADcILr}ufY%DQV;Z8z;lkKJWSJE4hcztihl7!zsXW%& zq9x_IY1WuFcTQI@ecTfMgUJ~WiopX07N}DLg`_~b{)93ABlCk+2fKh;7ny+XgFl&y z1b>kYOPqokiU7C)&wd|b=ni`;*-u1@eSeRf>EXS_T=-wipxdo7Eoq!?eN-1%WhU;7_-WH7;7837=e>>^*7?V^z=+H4r`lN`K z78zx=0Tnltyb*w=Z=MazIe)o8JHx@m$Ha5E&iTCcjI~60K^YqxMd9@N;zv{=p8KU6 zbFMhjBa@p(NE3K=3uN(oTv}r`$YYIZSoI#BT}ewu?YRKnZ{l^N_Q%kc3ORyq`|CMV znk+ukwcA2H8f*wjQY!A#_cuLWZ_>{eKDb5z$nNVPUM@_`(*!6T7ydILmyvvFX{f4n1$Chf^Vx{$|j!=t^AGa=sS_E+W0pQBnWIM08< z={n4ecfbvq4eA1U#l6$cE8vEf+sk`v$d?yVU?i#u@Yx{?yq31FstBY&;8f` zta_1w)uAwBtLffC&>o@;^BhV7tzXjZ$OXL8sZc!35s+DVzorCNF@Uar{^+14D`H|I zsolE4B2X+xwsM@3LmS8&R>j`fxBjufoGPYsoCl|+0xvphYm!*eyoMvvrrs1bHKS^7 zO{qdHx?H#?{6;YVzPiEf$$5(snwV>+_jKP8cVIwulC*=aiZZ7%FqqY%AOc&<44K~ zj2BqYzs~Eht3MG&4=9CX@Z=3)8e>1P)zI-j@kWCDd8ETxR^(6eaT5a;B_OR1ygjSv z1?)^?YuK=_q@H8wnmQgK`K|0MEx`vNyB}DqfTVRAMv+`pN{kQ~y%^?37yb4cs%7Mf z+;*-&c)MHlvwYvR^FG^UrLD=twHvDQe7O@P^aZmIR`0qFXDNCa&Nr&9R)n1O|#p% z(fAu6yKr7er&tm=D5#-9q>@BG8a=SMhS&6NPPGzAqr9(G4_mKsi(a7=(O5^H>ZI`X zZy;&5y$3BWs!=^d69xNDA6MDSl9;WM>i@rUNG^#lYBn zOOC>jg7xhrcpaN6{2Mk0`%>Nl#b7s6yxV!tM^KPcyv+GH3=A29sl)VwL3+#v3=U6N zW`c#i*vvaI9F0}`;NR8q`+9pCvNZHXLFO`>U~y170qrbp-yBa2XGw9%n^GkAQ0Zk`Nxo!rFcvHV`)mYV`ng4n-wIiAdS-8j~h|@RUg9hy@oGN8j?fr!rVNGX5-b zzoW{=!7YTZq4;QDnfkY1&FC#L6kot|+JBB?3H1%_+;X70^(W+lqE?MTD zV*u)7?mnzagg32z4KID(XqZ}ch1W=3Z1tN~T*-LU;>Lsc-qEWb$cEMH=^9zuB&dRq zzJJ`E#>_KWq7;ZI8eJTEAAtYEk&f8B=mjDKG$)+XJn)%y<;l9a@fO(Z>BRj*X~Md8 z07Jt-CpPwy9$i~$gcBm(TZn9(Si%*X_d;S1{i!@gktrZxCL;df-unb)@VBmYJ(+EIJmpJ+u#t~-F*gk3lcOq!QBG{3GNJoJHg!p z!Gi|~1b5dndG_A#`{`WgH>_HHcUM(c2Tb~7@&voD`l(W5(F-tg%_v8)xLoCFun4jw zA^cESGGy-SB??VM5rI;xlkdX9BU`q=qTn;xU=uEkF2WZ=@clC1Nvv%na9oBFp`bAO zy{ngTX*WWmmZkR6Ppym!=k7+|da%gxp_qKXKk$8v*TU*TYI};Ntyz;ueOO`a2?!MwKf-e*clk@1`+_AqExNxv21H%Nk2`f%$ui?{K#BvF0b5>t-ym25Z~N(e z465&Z}!4MLI!w+u?(y|ac*9r%>?3p*B=hX4o2=@(>Ar< zvd+=MzH~k`C&=~Yv%bb}Z(t*udiX{Kd-Fb8cxE(0yq(ifw`G(hM+vrp6{*sXiiJ&EC9+NBo*X6I#!UT&Z%aQu zs8c3&yC%|!PRS(Rv1|Ku@(VuZ&ANKPXO^&=mM|F6&=@xQiQY4CP(kv1}_8#*;A z4g=ZXmc)^Mhb97RSHi=OJE#X6tr(nFII=SKq;GXgKKDBFi%IChMM>Q;+B zf+Ffi3P7~*aiK!}XkGLl_Y_fzY~!KH$OyPx?OQ!A^h@W_zF0n4C>K z>wd8ALxRbmyYETy^}*5`Y1zBZALU3^e+KW5vtG5Ckk-D*{1--hGFjc+Lh#RpYeWsK zOOSywBTK;DT3Vn`?UqYR68)hz*zcLkQ{o;m9HOu<+ltgJMfdJG{Bud%cOGZTvMHEt zbHm;Hhw@x=qfCP)f_s>t6Sr!{b24H(XBGV0P=&2Lsmcw}4=K)Z=iBaNa|JnzBJH`t zwOts6Cs0L^+M0h|I1#&?Bkyz0l*B7n>BybV-{*PTYHH2~>B4uqEe6YQ?^-`=YM_EL zO~P_{a_#KH;s2wl0{;&Y1*lRiW!AiIA+T^` zGWu6;Rr3MqytCDghZ0jL^_%iZO;=FC>sgHWX}W=J$o)|ffR}pDG1r+xe35lCpD%?J zirh@?7s9#xJ}8GY$=_*z;WZu=A~>%JtY`sUSQD-HTw{0QCgScgt;`DLk;ivv4Ea~u{UkSpU=Rd{kpJPX&35Ro9$ z=fSVV5+Yq*emi?u!`HDQS4|1O5x@)gMr!6P}fmS z#Um#ju>Rxgsyb2N{7&n7H~^WST+8zk`5IlC;8%3DkZRyS1x!7r4id$=@QD9@bd=Tt zv6I}TWzQ=cI)$$nGG0aR@+fc;^1GQk+t=tF>v$&igs&T^MEZx|gSpk(AecGg?%4La zzXjZE#+GHw{?O%NA0JMW&(~58`Jw~ygdcMHDC(^9gz5bM)_p_Ut9TPEimmX+OvN)_ zX8p=3KjlA4AG@@Kv)9X_ZR`y9d>Vnd#&0`zdy@9<^si+H19s&6e>H8RBo z+zXH2r}Soib&QYSf-N83c^{spfh}{TFrdSp;ix@9- zi^n0Tc3XMT`{yLA;%z1Ku{L3uMH+mG8CYjj#dX{i;g=`YIs@uTl2FWT#GBtjUqbu# zLwQe*iL(BC(gpBxP3*UOG%r(}Efoo}O?b4Of}h#(x@uTPAh8a9x25VQ=6CLvn2nkc z#P|I4b=ir?P#$h04V&QcC_*@sQzEx&4;D| zo?cD{V4tY~4uqQD7Ph=E)&JoB8HwvBjfqxxjFvVvFZT<~i@X4DRb!!QwmAAp=pZ$$ z5+;g<>XicSpC%X_I#|5!(7qoTwfOUhFY4`r2xGk}4=_Xg_gA%WuyZ4xdy{NpLG3wJ zljB*LH@z2rnX3U*UDzVjo?@y7qcgztC{RjTTW6Z|!e1Jo#X9SAy=tC!8^=iYp8?|Z zNN)U`tzRL6&6UywJxGJ*c>K^7Qup$Vs^>l@Wp4&*vp?OD(BS;ZyVyLrn(1O77KUxjp+Iuq72&jo3BT8k+j1FG6Z%fxi?3FPU z%ast*OY~1vVBbYfDSAdfN%%-Hw#-aqka01D$Pp+7-hS`9vrSad=VjY7#6Jj`f^n)B z4G_8}ZN!TW&Y;je)@r0#C0h7Nh)Fr+<-_nq>>>?|Q$WhHhCHDD{@~opq38BtUOW)P za~IM_8S^nVJOA(L*Gx=T*Gv2(y)NiM5~`AyMy5!Ht2@uiFQqnQBv+^++_ojRz2%8t z2!(kXJ~8S1-LFDEDIw%;K>Gf7Xk&t<>pC3;L}Z?+%4tXQix#gBv4B_aLR)Y4&#vMZ z=Z`bay3Vq{tJq$XHFuHiIpGgwpocgD^a1~)9;*KvP&VjO@aSzfa6p|Fb{?%xO%F(M@reY zr%v4Uw(r2NyyEUz>2Q6>Z$g8zRLrPn{Ci%tqTKIl^abtU87Z&>)IR$CPrcirH6w-5 z&n|J>lL9;beM$=q_OD08R8?D9~l+v?*r+=?|vq`(E~B0 zituPOS)o6(LxfHY3JmmbZlQv*!B>pQ67un0n& z6!DvKHuUK*=ddFez4_$Vk4bOZT-cp7*a&bNgCOT2NAQ0N@FCSlYh4A?M&X3#YJyDmhYdD33(JaEji6TE8+i(_P1S|h;#O!>*-*7!=IXYnyYb3 zXnbHs&y;CTeY1@kl687Yp{PIlAiE9UiK=(B%5dmtHI7djVc!HVBq(dlelj3D_de}*#L+9e47To z5{W9iV@^cDOcfuSekw&$T7FmXCcj!M!m;IHJ~oe7wi2_+^@#XRC;3iIHicV&69Lv_&w$JXr>L{@=N; zl>(sje|GM5=&0J?sYE{#Z%Et}hYYFnYo*-#_C2H@SFBeOeS|V zb>e98?@ZdT(X4DY4kj8F*uJ)xv*NuXQ>UyJ#V+&vkjVjaq&b5qAhK3NxAgHVZ{A4R z#JZgRouQA0xH%m}lKfT`EbNtHPGaE@Nt86lQWmr?@yzV0p`lGCM)Z;^co;LuDo2(r zFySmrF)@)Bg?*uT938O~`0@P}G@hI!+qZ_;UZ&Tlp5jhSDWJ_a{>a!mGas@k`tTk{ zUYej=;(N8vpVU7G%j?%$I}$1y(=&4f(?n0o7-BS8cHd0K;IMf56OqcqNDl_#&hc7< z-gkN79ZCH9D*XEX`u001$E|;IP%-mY_Z{y~E-usEwHLWvh8&d%?1c`^_kf!lXcT;h z{|mNe3_n4L+O}&*{Ws}*xf(t2qt}XrFLGk2&}|)w0&Cz6-G?mL=dLI*Nz*VK5F9GV z88$N-6xn&qLl!D_gPqUo>DqYHAQ`V;E``~@cq*rhF+f52f2=k(S96WYIxAfIFqj+FM z!jfoU^r;4V^-TvbJL>v6?aWUMi%}Kr@n9?e?ZG*?f*}6^rTU-14FyV+Je$N7DyUOw zGJFpcv9VeV;V9>&cS_>fx&30R)gozbe@`VyAl<&;?)T|&@!Qgm0FjouKmrBUS?g8po-k1X$+JE^3Ax6Ir&*M}GY8^%+)|oN33ycL!HF5f7NDa8PqMKx&u3R z{P}(>>L7dbxoq9vrSl?k{jSwzv!>a!Qdw1c0rCaZeiZv}9lU+Jz}<-wHhXXj*5p~$ z@@muQQYdhLWK-iQzCJSETOalreIZmQixXZZMp)N|4H~;(K7F&<| zhoQ1x{r$E^p|KdYt1=YkW_1I!{;(@HadiB+Isl29>#AuTHc*GuZ@^HaZ5tD)t;I_5#lf1fln|U>6Fy zqKpkt-|jb-dWu}-f?~a|?-Gi>#_8l@-Zv%0Z+|ztxgBPsVAQ=ZH5a{f-GCE92v3*+ zzDf9>EqqjPH%{2Qee!X6H))Ot$D4;QktlyFhZTD-Jr74WxakpuMPeBp`!nEom^6)+ zo^~`n7cX%b4C1ZNSxbkbjhB;m=h0l32EXeHFvDcsM5IyIhH%$z+?J+mNO42l`GLh3 z*(YH(G#C$L@2hR9q0o!t3XdBqxrX{F>uXNIV9r-C>UIMrd{}VNJ1iE`ntW32kfWZ< zH67$G>_8S`D@GCoCsmq(Y)Oj_M*q5+@r+CDcS>}|J+X%9+lx}_e2SV8u&#q91O+VS z0%>_+8x>z7?=tt}v(w--W7x>4@`(tT}xArJnGFvouj-GBo;nXIMKw2 zh($$wJGc_d(8CQ5ob-3v>DUlA?{02pz@ifW`e8b+g0Z6KKP8ibMN=JlY;20sI9<(;YKt}o@0q)i5@RY!Mn3Kkkqs& zWVNLrow~=!rTFwCi9O|Nb%t+Xwmc)xrjH%-x03#if>lJ7#shzb*4Zm+*_8u)O-X;U zy8lGt{SGnASh%eFI(}DJGe8X<*4pH`?!n!&$%>!Qerxc`0luMCs6|O zl)K^FXZiWz(uUbPN2UZ2=WS^V*c3O5bHkQ)V)Q-HP!=U1cl>Xxh;uTGtz&7{@YY;l$EU z-wEiIcB6Jm8Y(%Fnev)*FHC!ahZ&5>(@Cqq5Fc?*=7{QTA9)l&c=OT*`7y7%Nt%@? z1G$3uwTDMV!-rN=YjV)o$J%-Q5)ZDPI{S($ zFCwkdyJl+skr+s@-fRq@eck`&Fm3Y3j|jlgMepBLoSn}^pB3~G*?%hPu6G9*jX|z< z{o3d0NA}Y%*1b37q}$k>j>oo#b*eOOFvR(u7>o1_CAt>%T-)76Z?}-^o^}@1f%Qn+ zo-A{mXH&2E!wpUHLw^w3++U!EAMfj1Ey!Mh#^}(cu*39SqN7fAO$}sKqC4y})<)oy z)5C#=7GV|T{|CCJ)qyNqupu!$wFas=l>>g_9q36o`l%U?JiC3g$;#YLVJT@!de00| zVaOl9j7S1^>*PU$Fmi6RT}-&c@t-X}|3Qq)lc((2*)kvVP2U^@xCr>4AWT^UqUIYl^t8mE7H2|pyq{vS)mM|50ZS=Ri z_{CLT5smWy<_gmB@Z(aECjkRd(aDmVor5SQ z3SUaLVLR_LMppE{!2+p3lA*A!=V%bX0F5kx+-HC$q{5ZzyYo-%nmz=Ef(9N=kiNdn z7A6;5&~8(WC9ZDL>!un_2nPOjEiBFN+I_J;WFPAuOeU(wVj&o_aW zdi1@`em;K{^znUSjr|9GqCMA3rh&WM#UziHyXS{b3`$Xvk@I4Ija&-JyOB{gw<+!1 ziU3=jgM47qHTOHd;Q4G~JXrhLEm(R+!n`zagYoqpdAV13RK#JTwSb;EOH=3$24m3*?iG$~eTB-60?OJXthBTOhhj75QUC#@jYKtTOBZE$4gpM!vk?4uy z&h95g^fg7`jFafJK3=4uaC%>mwQfAPpM7d|faarLqIDrV`gFe${4+_BYNj7Sy*hqX zcFFrlQLe&BJeaNeP!=||(H}M6zAFprHv{s}x za*(vegF^}+G-txDKZ-(D&<7oV99lVj5x~qrQ7R2XRn!!}CTUE6_pOn@pc{6Ij^C8m zN&O}L?vxa?@~dSXy57N%3wCgnJ}z(i`0&Feuvm6s^X+BR3`J|`F7u2(QqJ1ygSu&L zAj|3v`;iRxcx}j8(+}4KiXum?f?JYB82g@53Od-Bm-*jn-zZ*j9$Dh4XU5g2;PW%q zll*dYwgx{;dAr!#D~-+0%xIC2kl5e4m5(2z(vJLl1{VlKtvMUuLfUpu6Z>EKaRib6 z0Zf{=Lgu3N=3~v9G=~@ZW30MqaX$=eA>t@x7_eBqVw#~!aJo`#=8>Z8fg@!Y~wy=23LCT{wvX_qwQq>4xb6AI2rXU`?%}ocRl2p8Y;EX}kio_JQj82`m zTD#QIc02cDpn_q7q`bvEft3c~Pql;aMRZ|GGJC$;K{K{1XG3Kaj9u7moqK|>zXxxRX=%ql9 z0T}9<5NHvsU_HqhEPrOK)umKPLnA*YTj&Z5T`B*CEu*+K5+@z};^UeYhIbP4E5Cp& zgm0l>?c3w0>j-@ntd~^8?J+yjYRRMHqQlhd&;L`M4qRlpF#m$0;VTx?x9J$zo>-tGmJ6xTjnv=k(!8o%H?`gZrLT-w| za8mv=IoIg2n#J$}J}=Z!h{FXLeOagK?$Y%bD{#r)Y2_L&sR2o3S~fVHzZY7M+B zusE9BxF8ZXnhCm9X9gM`5Z6W&rHlrts5ZM5CIH_daFS#eWsGH%b-5p`=nJo_M_oJL zj(U(uqs`4Z+)T)G>;a|29r*NtA@w|QfW>d_sf6y^=mZgVJVm~}b)FD)Wx@K%+mn)m zl{w=f-?hs*P}9uBF$VNSJ52byUkz>5VHBwO|Mx`$*P(CSNJm4K>cECRsLQ4xSWz=d z5dB^owlju#xt zPBu!U`f)6J%Y_alB|thOge5p#X7`toP5*YTzbl&aPWv>pSga8ArP89{$R2qo4G0=% z>E$si9YnK6eNDy)JWw%O;G|Qny!uNW$TuNPu6|n_CsCu6BNt$VMrG8|q|`1GVIb;= z@!NyK{0oSr3TDzXcq!`gVsWB{$aVAN!OEeL(NWle#(Mhty&hgYY{+AW8&x#Sr9f+5*9n!$U;3e zrMRD`WUnn6(2Z%aWk3h{y{8@nD7I0*ZrALCk&r)J0O|q{WzC#e z#p21u83u5`EVScskHcLx7hE7;Rps6)O|epD-bys|H0U7Jk8X>Nh+&OGOIa^EidQo! zh9nMoIZ_Y0YyKt;aJ`Cpf4Da-zs%>4dCRXd%k%L*qmjUVSB@bTmB{lwfVNT{P#mc)aG3%KXoUUu~?qZZ!^;F%EZ$`wpaO`i-!z8Cg6K!5b4Bu{K37GjKc#R z_;C7tPlqC*E55Xq-Spppk6^9ZDM6t%l)t^7A!J>w5P2lFk7s}lLU0xLRSgcy7YC3) zyyv|aOWRxM`3&`)ppNlMu@ZNOfur12c zd4RT&4+lsexXWFyF;Nss0@|F=NoWPpoS#ZwrX&MMfZl{VMB zsgop|CGQ3Fe9Iib1yo%~dpxLrh}Sk|AfwNs;n`j#!J9RNn%sp!pFpHuzoiSMw>DEq z->f0i;=7196KR^l(SY1sB7(WxAm+3qNw+;SufrN^6=-_zE<*O=5YN)lZD*4cP#)a7 zRf8*?o3`Y-VrAnkP}*=U@8B4xTwsH^RUBZ4nxjkgH8UPq_|wiPY^u0~h#*_Bq7hd= zHn0A~@I!c0G$h>yI8lU(Q-@%a?)hE-A1tuubZ+=TOB$z3U3Q^ZA?eS-TzUIz0u6lE zipd)S^DTuE@MTWoWiG8o%O=fXba*0tuQJB&+>PNlbrCMs;;GkqHx(sd9VK z_01(xE#=13F;UM}`adr)Q;4&Ac4$)fM_pe-N?~E(S7{A_GdgghVs?VOWx|N-h3)l4 zkE#I?&@sHES1Du{mT_ZQ7?=~D!9I(@wU(xE8NL^i1LkGX;(Id&Xly-lUSeCUV6bBI zGnR-B(0LH$g@KD&(GR_<=+>X>yMf_`m*)O{69b0Bq+9}Y2|W0QZ8S9GvsM)B^Lo{l zns=~k1pNTj<~6!82-t13HGtP-A`gijk^^>7_-p+aI#4lHrj5A6-$Kcvg+Nu&!GH5Q zjKXi^{~M_o_DTpr>s&g{E|T|Cp}#AnqyQW}UrE=cj%$w4MPiSeZbL{z=%x#6r~?s; z3+W;5FPf%`URrJCUN8|MQ3cv{qV3-Y-it9~%21aVM0fgDhZRoFbVli{7c7>1=R)Bx zjVNd;xaN7~8qi3~V78)cGk7@$^X_rc)oLB;P#w++AP^B`f-*|!Jakb{ezrGj^NKZd zzV4{D7Pkm*C->3~w6wx`ll(&64FqjK^o=I)Cq`$La0CIZFL%HE8e_K&bv)jX9?!9d z0HSxdbf4Zg0G6Q~<(0+BN>XeD`PWECAcZW@%>?pvca}e1bXf3=RjHCSP#Xz$2ylTB z3B0IW<=kb6hLEbV7F!^8_AX{vUuk+pn@m>|i`B}1g`Z>5ZM`V}{C(-FGdhjEX+ zpw@H9I5UO6;=M0csq*xwdRsx8z$3aiH)F!&^Wnsdcv8D9B0c=e-rpE9&+3j+3t5J4 zJ}K6;4Y}^>FN@y@3nKJiJ`p#u>^?C}UryI1kbr*Fl((JTcedb`QT4Nxad#vayzd`wi(r$+P(uTLY3GwBn23D6RK`Bu| zt%bQ-9RMj&>sl_VW@k!;B(<8b@jHBpiC^~}fx<=O+7ZUb#DTdLO9;84vT26w5#TcP zWr9D^E5j+k^O5E%9=K~L@?QUv@e>x^Eq0)}|#I7AJxrFEk0{4%Qfl~yU1bOP_X zKEdwB7uF@~98qKuHIzFWBe~g44_j+3bTx?;E{;$ch_r%*DVdVCkc|nTrO<@n(zFKJ zZVSS)$Wc2-2&pM2qE7sDol*u~q-k16&|;zHEJqqZk5~$g1=o!SqmLenX)DB|5zuL4 z@27K$ylZB3afqnqECob&+tcSCJ~Yff$zx-6|9GnRntOyC2Sk1IrthHsQ;C3p=%cn! zux$Z}N<*BDAq4$y6VxR`Ow4YFDRK~A{1$ZXQiyhL8Qh8~2T4APaztTfE_K7!rT)54 z80u7aXgE>aVBpx;H4C9?vwG5^!;h7)EqZu|Lta@YjV4RrlCKtHj|1v^+0+WhoPavt zVY+$~V-K_rq&q8x8@kJ@&HS@|o}QeXSj9jrapKqo(hgIb+q$Y!OBFS${$PXn3^tyn zOxc1$i1-_%;p54p&8aj83&!VrMR;cKyx?oPj_9ke2OI+j4Jzin&A;6!EU&HFd0}LY z3iGSf!lSFp2pbz=Pwd#rFi-Z@kPxPg&AFOQ`;Pg8?RiljxjH|rFA<8xRucVyuQqaG zXi#m*m%AfCw0799xh}+N6h@$>fD&T-m`9OsldaumETO!GLHnY6wPLm z@yuB!$Jr!Iw6i&VPU7n~HP=Ul(=iNQx1N7pt6R%`;Jxi3=%GiwSQ0{4@{FWeLRY%U@sEdN=P;R3Z|vO zcS^Pn9#mCZMzC4(qE-NMb7QBPmxfPp*z8-K_{M#*I(s) z^aKonHx_Cp&svj`pyFy0I5!#L4gza_({EJ^kc@O?lK6@uHd6hkwpnIn=i?S@q^@#HQl(kbwN0`Qi4qK;dT#sxBG^W~em9D2Q8!oeJt^ zio%fY0-%4VWy}68Y>C~te4bHS`^k>P_^`6Je@TywN3VsQjEbVy`o zC*@3V&T<}tt>1T!TvWKXX2mu{*|>m(JxXORy*JI~2d5e?td5#P2Ov*$-_ey(2X-YKc<} zxkM=n4SeXr8JMeKrA097MOB>t>E^3cfaDmU{dFh2+sXr#fLHzM%5*i34Z;PJJPv)v z^|YjuylmW&&$8%4OSwQn6NFl|;-f5eRpXNRypR8>V$ zwO6GCvexIXK=Mx0F$E~+^+y7_iS%F1jFJBT)lBS6D0|#I{|;nMElVT_-S^eNbiO|H zU)47FmimZGxVCPX`q`$og8vJ6eC~z{F{{|kfUOrzKIW9RP95lMhr6ih+BNdcmpr~y z;+g>$lvIMO9p!*920(rC$Sx&CtWjGo;4vozE#scfxCpzLTj~C6Qyq(`0$(Fu9%$O{ zM`F`79gqLSkfoFbkr3+Iv;~j$i4_=r(6^uD9p7jWHm?PsRK`S$e)tUV#9-!J$Vvw4 z$pIq{Z%ONMS3G}SK%K;=Z;FPwUXCMmFV9U6^xT0w(SU0`SEJUZhptZ3H`2|NKm)bz z$W2$wm!H^dG2;<#5%phKlqHPYzPiiQy)CCE+D|L%<+P`4s z>Rg{hhUx<*Y<+1^HJXC96lyrs**Q}*f4qh~a5*8HTiTWh$(47;h8idwnHlbK3BP$v zD$X6IAvzFWQoidG=Y58WhOp=-IF0X}h8|lC^E#$X!zOJD8pC12Cbdu21|oYgnRv5nxzWXZ)~ z?>IjU9-?RfauzcR{I6rNv7|LA`h;cGpcK5rcU#|n%|NNt-;0z?&JPt`A0@ou)x)4F zc)8TX&VS7OZ)C;)L~wG{nP(9eEzi0(mu(RwelqH`-WNp}dV(~h6>e8EBqfjmj6t(- z(v{knMWg|vUMj3@zr{)yu)@H#Z(STc1#_#`1^-+BNmh3ceX(mK%O2Lm^+=Ulw)sb+9WAf)*bAx{QZ_K)A zDSd3mz3b^q7me1pPLRUJ94p9j@`P$##+Xm5xo2t&bE|`_WyjW!$25AnO#uf)Uja>Q z@wP02h;6zCX8A?5DM>I!fT7E92ETl0KN)G<1Mdb1F@*Oq);PlS^GF^vBcE(-YP)G_ zGrGSlP+Th*Q(@hGF9YG|=Ex0B^grW@6{QGw=tsrK#@j}5lzy2GJ0_LPPfuPS=V#+? z)SlG zY!R?ai+~;UI6q#>!#cfHrQjz$mI0H4ctfDt zn5~V}KLwPGI$aY5+gIotxsMSe;;DJzb)dnrFH9wbQ!^-ifD%VcfRvlpPj47;W1$c7 zuuyB`cJvMC8OvQ%(7$q4G-=DJ%vEm#5&G_bRqhHSIK+zKC|2%DfnXLUvj)cwZPUUs z09$JRP$@`m+J#T^50$+RMXqu9h3X?*x7QyX?MF2(Y^>&igIfqn^ zD3o1ZqtGdXew=UuF$eve^ht|$OC3vY*=U^WTEv)&XmAnpfJ6-_?K`v)yT{yaxvUQ# zIg9DS1C8u>0OQ+s%Zm@U@5n=np}ffpR%|6HQLArzEklV21zed`L=YUE`92Pa-`)Mx)G5nDi2%?^D(6l9WwZK`d ziXGN1zUE*3A#agPEC8wx@l1m-6L;FIA9}zb>jfM40MLkxu19Kk&-ABdT7wN)6Sp^; z!0ttdn)~WLJM`8BjkKr6#fMu1QJ2>TI?G{><8_`LuOjw&G@D;E+Qii!Z<~}vdq|5Q zWGI5CmCh5-kF5=LFJXbj>)c;oW?YE8&wt$7B!Ee^iX2Q#+26Euy)yWmy7N-EE(H~< zNLUCI>M7y6ZN6-;rg<(0mo5OHnY=T1T=^E*`#WK0A$J^Oi&L?G>LU$s(pBi&9jJ4O zgDkE{)=>K8DElw>S%;s>ZNQ@JL;ZD_&%artZU_~t4`bQRI>V2;_)eHqs7K`=%d886 z5b5!+!i#|^X8scmEia_>a{%VLtag7A*6DHmHxc4@75${2$QCQ9TO|eER{Pp@;=piY zkKeZLYzp4P9hJ-ZcHL^OlR0F>J3HfZztU4=*yq`K_81ZrK#GpXjRLFBs6gL~$V9=@T^v0` z(Xfyh7}<=%gQFrGXytzp_(eDZ`W-1bp(D;<0#+#J2!P{H7z zG~)Id#(&j){${k&TyRtb!*$lGGs(T;xiImyP! zMJ@)y@&z~-hUGdkN|R!?;`UpSDCp`);~bM8I*9~N$QT0IC)OUV){JKO|4LRR-Ssf+ z)Il$;EFRvXVH=o@VOG_@hc$|1bFnmutcR(Z4!(EMr3@o}rgYS0l*m4Tb zqtlyF90tY*dL3Bb=q3HV9BUYwINBL%5-Tg~<=zjntyXbSQMS3cIUX)9*FFZy{9I^> zwEDyChxSjN+mN0%`jqxxA1nARdfs0a^kvKEBPx?tao91AiM%v&1kSUT2d#k|REggL zZCGbBkBoNZ8*0s@C@<6Pet+LVJx#3TNks@ci$uSERr3BSbFT-Y;GDhEM~JK(^q9@5 zdQ=DIl5-{<{J()3p60PcmcT56y0^85kUWT9tG_wtlDKJ?w$>xgCRXZeY`+i^E6Hex zK055DfHGs8MJ2_hm3T!NWQwxnN#APpr709RLa zE>MG4*pg`b>b?(&Y7{DwB}V6V8-;+kd*94n7R;1z2<0+b>m@8s?;~gOukWxFE8y?E zsI9yJjyy55En*i z?C0{yUx~f^Q3cLc3zW0i?i`Z28pc41q=@SJX_T+S4nuc<)Z|ko^-GIsu$_zq{+1WE zRx?-@WUVDPx{TsJAoZjX$l3W$>t>4h{IthP%k`6$m715AWO&5iA{0)S5%25e#t)ZJ zOT)|X)8pfx0W68HcN<}HEo8e4$7l_&;~#wN|{ z_Wt|Sty3Q$HdB4{A>yly-9lcU-x#bLtI=@#SY1td?05vriii+^HN6^af8dL9yL&iJ^w-45?wJQDgdDsp8g=s`U-B>a4Q32c(U zlj<=t{dd*6dr4C%RDAq3I1B2)F5`!=s*ju#nr_&qC*p|qyo?iMwbsr%j#Ws_3GR(` zjP0Do0WO6);EB2Kn-lV$IH$K|@!;{V$&`aLD|8Iso-8o@$YgbtV#@e@pPx7T7Up+h z7VQDHWm3#eV_2ibDa#>bQdKGuugy|N>?b<1k*LL&8-M=V86wipsrW|nHM#0NZgHG1 z{$iR%>i_u3MJkU|9R3@y?@^l`avEa18j{^}X3lQIfo` z(ygstyqiUVDwMAmg`Ivk%J~%wrK~du#+R*Z*L|P$mNm|aTcAcml%9zhtv06ww@*a? z`MNH!EH%`Bnj0k-})H4=e>Q}G?mtCqq-Se?o+DX2>BZvLg(eyyfDzO;AIoR8&2ea^beA0+^bTB8H3}Ft3it+A(Xj4%XeQOkNsI{(aUVyPGZrM z^p)HAChgCU!wjCD7uo78_nV?l?<-znr^JwGL%kwMdiD4!-dWgjdmN+dk4jMUM~B`n zQw;hA!j-!Jhp@K_tMdDzg{4bLL1`qUL%O?5VsE-bP#UBgq@}yNHr?IQAWBMyba%)1 zZv6eft8{6Q^vd{S7H{k2FPA0TmZdzoRyL0aHM z^2T4(lu$Desn4RE;zTsbF(i$cHsGUI?qy=d!2;(}LR`WZpB^8tPEUK~_9#B`^Z9b{ z%=bAvT3E=H6^o)%LIfYqRq+W4(L}0<{*suVa6pV)OZ5h@poJ-^GH%BPavCc1US%7$ zDS!gh>Pf+TVX&ax&Za@8m$hsYFs`F)&E8`PcGRb(BKT4ibS}!MM4~JczjH!w7d3UC zK18yhe9!=W_}-jprA%x$3$glIve!xAgbhx^MAhX)bcK}ARq6FH3iZx$GnW{(evaDz2nG&EeYAywUo zHWN#heOlML0CY#NYKmkKy?Z2`+k+tkL*zYufhfVo<6-hw+CT1PtbxE%8b4ZC-T3&@ zf@gH#U@()$x32X5QD}i^l> zn;%Z<>}XV;CtxB>0cIQu2RQ>HZ&Ps7Zi{`_Of(zW$w<2!xaJf+S_VkpZ&FL2qT6(k1D|wCbh~IrTv$&`kB=0( zg~mpWCoBALY1%nZoyoL&KP}@UJEQaj+^9MR zW<%vV;89$l?>>-K7Xc6~m=aRECu$LHPA((q90Zpy(}7w11i4UmXGF0{kjhGyaGdq2=6JA;SIpL*oljsBDj`B=)eDyV9P;DF1c6 zNh_9UZI>~7Ld52M^MvP-!VD(?oCrSd_l5=DxyUmpI_7UIG=6h@Lk~S8dFpA?7ihfp zx9?IwayS5;1?946fEJOI5#Hz*F2Ftpww0+$NJ#YLrev_o(Sc_fW{@q)aPX*$ZO7g8 z#Q6dEnnWOm`eVkn&-v%mm*Og&ff}T>0WL_3d<>-mADiFat;c>u58GWWl+3zF+wQZr zez4H|aVxuUJ-UlD4A&5g3FaGS?Q~45l$?6OkfF9JXk&e00mRdt#6%=~@DQ#@sG%-x zscLWMU^^iz-|f4B0@z?ke|6z%$t3@4 zRQ3zM-Gk)FJRvR=b)UV{#geCwr~xspd}sIr*=d%CtqG)f9Y5$$-HYB#|01}1D?W!j zP#~e7pkXV(3UHo%5xp`h|KxWK@6iw5QWGNoo3$WMs5Icm|`H zbA~CHyIQr2F6=;nO&L5ql)m$^67i2!uYxACCrh@}beM({Q7y%1i4y+HNcAcnWk=j1 z_hj`D;I>?W>!*ef8K22F>k8l70AQXKjM$M_uNdEyluNiy_BYUe%=6oVa=5fvG)K#UR$?9`?uJvDejuRZCS5=f_ZUgIiH9w0Arbqdan}PGy@8tk7tsk(fV4c>mnHag0 z7l~U$#dP}+11s5oR}vyFzJ<&-<84L_DXN`RLM)?oSt|ckiK}eKFJPbR=BOJ?*L~lvJu3kK#3`&lr2IavIdVu;}Ysrb){x^lkK90 zLDp(5{np)TE|o&)ayaI=ip$}K$vWvAHNcEQaZ_znj6uW^=+i4MoH?ih${ZBVHuvJU zXIwWY=8OBZ9Ha@lda(eIJF_n0&Xx+zRa|J7I8rNV&oG4o*blB_q2OcZ9o%_hmYfpF zMn|mmEq!=;G>EK(<#P~@fIZJr)5`L(DmGZFnKK{!W@014W?4iWOi2gXH`&7%^Ds{JSG&3CLS-4cs9M4chn2+ z1;j1Q5GS{$9q2IaTu98u_qmR)BL6q!l|nF*9&Hl`oot}e$w!^Jf)J4Ug-v(rmRN8V z<6kpq=3A(DRfqcbOrCQiWv=5*&?@Jg4QNWN!?j?4Ed7vd=SG3=_(n6xQbX) z<4H}=#;f9J0KT;D(+Y$+sFhHIIm|-ev_|5~`DCg$(=hjE{+D2B6zxFx>K$3B0XV?_vc+m8~Ph3tK1wT(b%C&69tO zzs>*ZCuQHaezp~uV5GKP5np8q1T6XeNO-Tdddx%oH-85TMmiy((a8~%c=2zn`B|8d z?2DPeQ$aVngPC0ynVW^EsAUjtwCeDpHHF1e1Y%UV#wjP0FfG}Qi=@vp)Ay>G)>c-< zcb$bFGf_hzmCy@*1Jck*TA zzqz!0&^khVo%2caK(Fw`(e{@8&oyTvH5Sj}B~2S`w1xrMtj zUm5e{yRYrpq6F|?3TVS>Gf z@Br9$;Bp&XZhB(%|IW9Xy#9WM10G=xR$|NortQ+ZWzf8I z@#}0YA@h>r@LfdbAav|y!tKiNEO8<7MgC8fFymWPjWzyPN(SFw8!-X1>5<{}YgLW6 zYFjJ7D9HD>rd2CaKB9j$c%*T7q}_QF-+2{3NdgyXUO~jLt+@72pu>cC?%o3uHRRIR z9yDMlJ_UfClg)qy;r-GSnecGlM!4Cj#|7G*;)krx-U^w{^kgxov#g~-+lN?H!hr|x^l+z@uE{U%#{OzEZw{p_Gcn}Q zRaP76c)4$gl6E^ZLi}~+6133a@c@|zF3X=T_XX|&72lRrfxLGyfZ3V z69bum*^xTLX&avljeNbSua<=S=TZ71ix_@OeQg37?p>o8gx&JC-4jIPHRZK|kXUrF zWNV+)nz~dPIJ-S z*pJ(xQLida{@~|63{jHz3C-mUQSzvO(tI>Jd2>B~Ap}th*~zi?%8Z_phtPd`I{>j& ze28?*#8N9j3pBHkV%H6!qbN4lbo}i&9{Aq;4S?}v6}W>{0)fd9t-FDq?(@Ny6a07X zzLYz}VGu>De4eaO1Hu7;d#jhUA4NgR(b<0D2t|P~+an__F~puP0#=XjBb$xKy5bU8 zYM5vkBz0Cv0)zco@pk7jsQ0iLAa08n`tB`bZinJin$az{U+?i6KTU+?2dRvzoovoG zB52}y5q{x#4OT^>Wsid4UX_yV)8vEKlfn{1>e5CYH`tH+BSPSn`vp9h`iGsD;`91; z#**-hzfw2VGy+q~h-s@G>4Ah!Wu8*fd9JKAm*Kwh)>bA>YDj1eZHqH5?}Sxn70t+N zJU0D*l%`sjW(5YlXsZow#j=8#QIl?*a3x~+KA1izBq2-Dv|pYuhkufy^DNAr;&83P zUKjx;9`ItTU^)C};%+US>Pq6``_@?kFz?taE>OoW{Rhmu72`8!Vroe*;lj>uZyea1 z!R1f#c-i&1oI~DeALHmdKk)f|D3oT&l%B_5t0e^^n%v2N;@=5%H3|#}#%0Urn*KiH-&$wy&^nU$hkZciz?G$gCbNiF+3GxWZ)w$E;w4EgGQY-J5d)f;B3`e(!E3c#o_b%0 z7Sc=Q_oMr7XHQWn*wy&L45A$IO4Y42Sk21pojr3cQ;{gsh#IN5B)z`#Mvpgm;M(Zu z_X9xC-t}aKb1Fczs^FSix(ySD5>?He<-C!tf2`7Iz@4DI?4!GN#Q2>{v4Ikc&+uZ) z?X%&1-R61&L#dO!24U>h)^EFJ)CM^XH}y?@>5W-qTEKq0Z(MWP)X7Y2Zf;6mt9qg+tNC|}1!L3LrGDF;*1@1TtR<80UXfB9SX zzjjs!UKw{12EBQ+XmVSgS=EvWRQ!1|ulZc|vi(jJ)mFLJo+p^7FT6M@$sXqw(wQh1 z8dqCB{#)xqJ{6cD@$#pl1zSLkO97Ez&E&kK6cjAaP$*@i>*)^8(^ zbefmq11;6nHb&~2qMBmA0HP)nfzVJ(vJG1zjhd96+MB)&AH;^AFx2pp__J0mWITAM z4XnBCSf&ymU?T%ChkM2wb~|0Vs%6nOIdazmNunp_jY$VIZpxR??aJ~R5#2>c*LX;- z9)1A2LOk#Zzccd&fD|OOU{r)G+rcXaLA7DOSYzSj&5+BVBB&t@b&r>O7~JJ{wH9}y zM|x;G-5WmukX4IfGHhJBcC1H&v@};*5CK1c>BVPFaD|UgxA(+^Jon%vqjdr4%=uv{ zHu^fr0NBw)#eaCi@`)Q6q8UBT!cZ^!-GpqrYd6aaF2p_+>2*zJQx4?4+mY+WtCy#TuY#*=IV+dL0yafHc#r4WgVM7+g#q%VXxt9~6t< z(dZ{2`QmPVSlrm<#ckPtar0aW;bwd%tpK!<8;2!d3KQqupL@CO`LW+H;_M=-%1O0<549$-*GtuWQZ83nS?t&e=7!@we5>%QSN!6*nT!Z-KYz5ih0#H9MS8PtVHv4L8vXCW7~*gt?0Efg zgldcR(u_PKIjF@9j7Y|YJKr>2_>Q$R7%$|oEI_71Ng+W8jUsaNd+T?{-(i)fYAd@( zc&$aES6wZl7v(C=jbQ|8b)|Abk5HF0ZQ5@nGXy;$2_}?9OWseOO^$nNkO7FB@y7(t z_C)oPWoiGB51hn5l&}%>U9MYCG%bXDej^?fp3_Sv>1~hF@V73!%hpB82~Z#9zrY*U z3dH*`#MO+9u!4P4*N=wp^l4fi56mEoGZhd-6RE`4HrdK6nsQT1B{D~&7(DlV%F$0= zEG{ZfAaFCps~xxGfk_lAJC1RGU1;hH&~(>Pta- zk_?PsL9e5Xw16{824I+=+Z}HUe&0$ro@G*cPqG9#{A2!q;BK9<9IKP(w+f zm@0l9>9xv~w_R~3Az4T>u zrcn+!e`U^XNv~D_*|-%Cw3e}Ufkit)ml4$FwXdu$sZ`ZPq>6KF2yohINxMK{+qc9_ zLLGKI)%BvoY$4YNju!J*%&zy-mG{QQ-puVIi=~l~sNmf`8$K!NmMIvh0j7bAglf(j=VK9D#V_0x81GS6e-H#!{*Qa#&jUcPr~>_ zQ>M#4K|=7+<{LHxUKm+d?w-cGnfIS3fx!D+wpzS7k|kKwWlDaSWqmN_4|##rVoh3Y z8Br08WZ9k7Rnq(0p6n(p6@|d4)$zs!k|(0Q1P2f@m4pSK#-O~>%?U%e_VvSx9ia0r z!v<^wOul{vnPrkr7Cci(3Ex6E(#3Im{4y}tI3nYibI=~0 zRGpE(1(i!ka2+aZ+_inW4ddl`qa5JYQz8+SAiiW{E zLoMZZU@Gmn{gt=K5W>ow?iG`Ab5wNi~$TV@E=S830Qr&9Gq2kJroQ7SxXI8~=RpzW>PUsgRQ)aLi*&auO zw4yQQMlD&%$)7{Ta)H+5%bbr)--=MJ*ph!BmvFjq0|wx<#u|#K9<`(jVX{(BJoSYCwjZWK9C+kdVuI;G@?m4nDqb~wLL1tjWa(X#$^H)ub?xJjy7d+Y(5?nN1Ud#vA(f$Jd*BEW% zKE5lYjXK!lYr0CDv)QBgkVa;xr_eH2JDR(Dt;=3|_yGVdM*&KpO?N>~&iCGDK$6Q> zsYQ0zb^s>Ab~%t&9zt!{(AZm%7p-Uhk^C99dT~!_$Ojxm^jrj^1ueu@NBb?oORu4u zmQf+M{$PpxCd?d2nc?r^ZmbYG&_?jbI^+kdHS~n+lphN{qLeKw@8=ig&CJRymH-&t zZ%Y;(fTq;PrI~A~p)+7=l3cFt=XT}#|HsuMCNRMwK80$zp|9|XU**}uA)&sNN3S_$ zI;YKyEn?lHnM9_Tt4$Mi(y)BvUuKI2BOqQ@(#daEJB7Ef(adFgr351rKWq%_JAI|0 zq7W=R8!a!AI&#Ss_2wl?<7>52Z5JnoxUH^c*?9TpgGR)#{S2hzJ!udYb366dyQ_Yo zM`gS>Z%p0Xv;-fojgwQlnriNo$;* zIPieYUmYv2JhniZ!kRG!szRMZWVN;qTflU z*vBIB@FghIO{uPPFRx!V1(*$<`3Oa}6Mye?q=H;$*s}Z;dO8!@xI5}dBz?N>x4$~n zb$oJWQK7~#U8>ChxaD5?M`UHSbIV=FEoMaE`ZdI#;o7I?}EfsjcB)tyQJIh zeBX9|I?QZpWu;Kk>HalU<|>)8w7KD*Zo&)Jia8?$Q|8{X4>)|wR>d=U;Sg3MS% zosl8ThNGwZbQsUH6#R+dmd=u&;ioxg=RL44tfVy%*QNTvx zba%^B<`X4;{IiY|%*NtmfI(nifvd6r6q*sZ6mH{rR)^_u0^GBa&nossKaMNXi&Red zRjhe4Ug4(DyeK3Njnq8)725x4+cB@DD9izvWVjM*a9n&tMV{|*gmAl-H?1(mHE8K8 z+rv@Vo;l;}fy0FhSG&QLIYv5*$UYGhYn}3j8Tp|fzuuvui@bX!p$cQI5=|+UmLidv9sn(AN~Gt=+t@P z;Ox}x4Rq?fN(+CJ+0ih$b=}W$Ka=@|BJX&ZC|ZRvCe%>9h(0Z>NOXS1D}8MsTANRq z!;zUBLL;W4w~j}N*LJMi4TK~wS&g$cURo&Y>9qhu1>|{JBP0Be^%OzN8iWZx7(%N} ze<@7^VPg#Ff1S2s6d`4uSa1%g>O&Dff_bVwsFm6LLFk4g(04`N%*PO?d~=+7;h|*+ zqZb!`6J(h#5sA)-5aM5_Y%uUEGDLEPB`0)!vd#T!y~`g7shlfzTQWVzzP9KV2>#6R zMYp}oUUs1b=~Q&Vn8ky0n+^bvUE`7fS0^^u(yQ2PaB_ku$Orq92&N{uR-TyljBPiAfwppypuQAz}^BxQVa&_{pY*GJKx{NaR( zuN0)cZre?o@Nfrsn!k5uHzhv^(=$_)07`S%xR}_{Q3coY?Pz}YOL?R;u{xd=`JCN{ ztK|+n@hRvg;-QwKc6VAlaH0apj!d#MW?yZ_6h?e%{4&imDM@5+o{AikS8?t8nY;k~ zk-R5^f4;WHk*@__y<<|MMtI53l#a}U4)?{GY_EUKX4KVRFi=b^c!Cek?cFh;x9MOT zV6Mc{_)$4jW})(??%1H*dOGDm9FSs;;Nf>)*Uza9QjaBk^#ESjUA z^*j^@bCG*31_%SFmHLU-4szOCz|KhVWxUi^s1HCV7TJGMYA%HC%mTe%^GkuS(R&*1 z13ZwVQh<9^yJ2`UC{b@{$yr}ITaV)unPS{R_bb|UKdrR zhfLCN7GS$Q_EJ<6R-*}2X(98l!e~!tQTP``cea;J8ywp16bQ^EWa9MPgy+(&^k06u;G(-Z2m=&Ug5-KVwNPDOIk1Lj z;=}#W*SbSlV#fpVDdp{J0chr5&*FnMgH8==$hUT$Pp`4yfBmkk0vW_*^tTQj_aZ#SBDJt@g$9w9beFE{&f95^MDpai@K+;7C>M2PihbCQO_>vvF*S_S*TUpXF zC;)So_9J=K1vH8?VCq%+2$yAXwb%)_P} zL+IcF>>3~LqF*~axFBl2h0N;aBFAdnl@zP~Q?iXZT&ie8o81Q4OE#$$59pKP)0e9d zbc7QJJ3a=0G6WJ}TR++UrF@&AERJ@HSh14S=9)gpMX4>RL1Kqhteuu?@>a!+_3P)% zcsEyZpkkonV)2ESF1pbiIy`UO(H4EqS<*Fa1hmlV?X@6#MfG#ef)Z?=y@Z89C(_k) zjiKg3dK)!+H1i=4bnbh=UJV1bO$21;<0ppy_EQ!IO>j*5T;Pad86^II>+J~(9liK! z=QnieqP>o3%mG*1R_o6cwN%RkgY$ng2ZVP5P1tnCMg!ftANB%=l@G$74=6Vya5?PR zz>Rxu-BbqydP4aRA?y&`VuseJT?~E01H*CE<29J>(3q6a&snANDn08mY4~6XKgc_P zYWE5ojOsV&RGZdN>wCC{`&HXSg+5#kY9@o!m$kRaKyt`hU^56ZR=($|Q4ut;Txwtf^YPx86uuEoI5vug{Z!B&u*fwEMo$!hh;v=_obn)QOG(9k# zW#GFY4&2Rj?W*bm2Dz)m%ZiaN5JOm4*{8WEYH~@X1^F;dMfg1jw#fLOo2A{X!WBb1 zAi*sVo6WbDE=iJE@#<)0B07%AzY@n~gm}3yNa#`|8xd_64g8>uQj25!UQ|7em?O&~ zB^g5ZW9wlvP<75g`L;y)y2nw05pq~85wB2gxF7YNt}@v-lM(@lI)!<)^Lx>P(1^0k zc1FMlMW1Z*pI@W$V0JeqXAC%G0r^b`Zgkt=b-0yVNhH`@!I5^sa zOSe1>AYF`KjkNatD{5(>>7smHXZ&wOKv6d@{R2)+rmH?{Z+6i;P=uxR5>?$ujU&8~ z6xo9~F49}esO%e0Npak+1qj3N!rYtc_LiKgoD&RmO)~sTr*8UXMc)_18w*5j)D#=e zEXhqphepZ#{3ye_`~HGm5H3&%)8~o*jCbdjlg?*^tef~Af?d&K{CxIoPxw(NAeAZG zYjU#zdmSomp?*8J7;DL>mXmGAAA3W3LLf@ovZx|IYgl5%?4!_V38rr?!yya+7kHZh z{4QUHd37jq5D?1W&}x=l&E=s$f9QE_Yx(mJrF;Q2RG-xCnRc=rBp24`XYU^0c_hzd z4eVc;M%=h)xj5jFe473p=~{qHh0sv+c4vX#oQ-I~+F+Ul_Z@nTQL*ln{TDbkGRUD^ z{h!F=?~^PHZ zNO9V4-K=T&lk*>6`I@mx{o^2J!!d;I@0z0VxBYF^OZUEzi(n$@zUNq~OFq zug9AE0YMF^aVdNdndSY$n4eCWQ^l{Ku70p_n4$zKdmou8}M`0m8K zeh5*+!lLo9t3x>^%*^T<-13hXjk_iLGN!JIGbKZZmeOj}6xAe?osD*+12j4tADS>@ zF$uY~Xl5$OWqPmiRN(NZ21Kdz6uD4iP%S_ltO%5WTQ5g(Vr{!aaIph!-^oC0UaH~) zy~MBYfYjQYkvKE`{K@H|gnsBJ_g?2Ml9}-yt}wY8OjIccXGbM{5*Z4~EvgOv!d^I& ztbZm9`kgaWFFpG^PuFYY!?x^2!L$%c&Mw^GYs67OxxXr@O>e_um2dzqlO4q`#kaq6 zb_ZjrwE~0*(i#Lug4SUd>*{bYw??n>cI5J7JdV4N01b%kVn`RVuY_rSd&mWZWO)42 zf}$kX7ezaBC67lx)8*hVCuQNEQTy-;qk@+)yTTZPynr&SUd3rB<{mBC0x(*a#ahB|@PBM}}5~ma*u1I}O;wS(R@b9~#eEo$jEAM#2 zq4Ol*wkMCZvA4@9a#UONl z>_2I+I@gAf!-}c6c-74zRWnnnbD=VJ5j#Eh3O)l%mgG?F)x0_m`qUl1F9Pg^34YtjU2~m`rWCM&gn;mC( z*a`}hQ-$DP-vUE+3RZr&FnX6hpFN>wW=_{AsrVOQPIi=E>;elXjK0se78ABYMHWi! zrNS60*?Fe+sz4z~C#tQO4F0&JxlIC(WFcHlO&?|lI%cvwmZah@NWrCe$+ek9<67^G z92807a3B9VPDHv|3!nad`#nDUDV;DNFTkQCj(5fdfh_OeEor z5AO`VY!l-XwlW4V43gIrG)tRE3r7E1YdHhfh1_hW+MG87o~n*SjO*jA|DeN^<%Fdo zcj#x#+ANcXg)Q&teR!Sak8eDg7A)$TrB^2H=&8A@k=p2MT;3cs7?So{tcwy(-Rsww z{p5Gvr%aDVEJEE?r5gqqE!6qsTL`nzMjcLF4L5aB(@U1IED`_fDAUl6}JKUwJ(AAoYLjYt*0~5HXs_5#*%&qyE)K6!f9drW= zkc85@zU67bs4aOK15i%@^>?hnm`atIu}>wYs@U_WrxPIPfFh30wzWD0iYkcD1cN_ ze44Z_^@(amC?deQDk(lfqam1m28v*{>Gp6VkfJGhi#p8`;O){?+9gL4E zsp$zROEQKOJ>}d{sEb~vPm>6cR1W+7BHc?xka5%>{ATdue_jBnmF!d{-kRddZ?nd` zXyX?3ekxySIVxUiK6f-!cb<*tbwFp?j-KR3_*VqT>Mpi|GPLj8L^6Ate}tF4DmIES zvT`J$TdH3VP5R2sLetbGgSBmw-2LJ!w}$eLP~NwZ9IZ@xJ8ECvDm#FBh>iA5o#6y3 z`?;1{$9)D6$TXoPZ%d4W=2#|O5SG#qk1|46R>3W6rZ>9&X4n_209qGb%2BOJtPfH` zn=6@ddEI&_5Qc}Os~(9TuLc2ga8q%i(gN0FV{2D(IPveLNSKl{*yVeeV#|QH2iBk2 zSs6rST<}D@p4ZI1-wt^Jc}eP+sn=eJmRfSt`fsadMd#U>=sbdh%&47gB$W|#sdWJ* zrRv?a-(>i>1I128p;V6g`W`!pna2OxyhpYR>j|6q-|xlEOAEPc~kFKuS*mKpwaCvt}s4QO7Q zN9tXc+Msq0Z4Vx8j{k~e?*EG9Yl{7W<1F7{g6X;9awci3!=?ei3AKL$@X<{G5eM7`0QWBq**6ahjHKSo1SPkA z{;NG(Zq-ej^;>hB>U-sTl!y%9?0qhme1ED@PYZ5w8!9`qc=V}hfprr<2Gbs&)Sab(@PVICCMp@uG9)iIR`9Ei!j*a!SlO`@oSd(?BicfpLgU+Zv zcgm7+`8gH=^V!(0W>CQdJ5r8rd{b2a-(($Q_;C>s%0Mxs)TRSg^P^jG0m2*Fme2v71!~P+s{2zG0pcDlPh*3C9YJHw=fc_(Vq>v7TfaGRkTyzxZMoDI+t;B zA&|Oe9rLYRd%^HJJ?p$^*EqL9UDcI&3Tqn+eq0oEl4|3VF}I){wNYnQUDXgw{|jvO z;DhcOHT#?KrMLLZ7t_~^vb))uS3gm!5n6letHC_`vMLFK*$t+RPX+5W8nLoL9-VqS zeXmncr3!)1qP6VsE#yMJ1nd*^=QI>jkgqE z6}RSOAAH~l3E)({lNowOE%5pE7U7|Vq#sG3dhc@~$qpqk?#uqfZNufAi0CQ$oz{F? z4&t1~REqTwLe$$i{fI(Wk~vyKq1kJ(Z?3iiEtAsFd}{%Gdbu2ijojt+!Eb@K$c(SKM|%)sSn8@ynDmi`?Rb4QbSE5$pmidh{F>a zg=6=5TFwLPWci}wpHkZ@_(ya9wKEg^#F4-h5CC%YB?(vJO9oZ6;-8 zQP;)tP59ox57+&Bn1QOx&?`QPoClbR)_-62(xnS0)|_>DM?cSO(ooM>7kph$;_Dei zWngJtyiJ=T9zP}Y1)I)>QU}Z8@>Se>6OUtJ+BcFSD!9Lx>RD<}$ztd@%FHBis{WM< z<>F9;7jd4@I~%!n&$d9dc7p8QO1`(^DZnV^&j`8)k%*UJ*@Gt4A1y0PaKn5ujL!B(riUgI3@CmR z#F#dgu)led=kMPO5QYRpHpzlsN&UuM={syKEqx+OoiZ(+E|HvPXAIfbEXdCMRZu(& zDp7ywX@nORm1ytDI_E4Vd23@R0P$jf>2Ej3a+uO7&dgk~gBZp1H2&+hXby!3O7Rm1 zi;~=PluJT4Qv}YXdUYY+U-yk2;~*1@c<+upSY01jZkuo|ac>=wnT4}&pbn3D70^ciurt6OrK*UlDT~>A%>6%oUVcx5d>Ra8>&|l(1fEU=s+VY76@ISdQ z{I?cCGc&HR1^C(tEzFRmj7R`%rIUa?^lhs1_#oatKAv;1`%)|K2>4&W zfY!S_+Yz0bP@{>`0axBNc*3+gAdW6z0+bsVxL7yr)7Oc%-Nu(h0Pv;QIqEFo|Kx|{$VtLZ z2qAuZ`295rgtT(GYo8w-J;N2CV)0FVN|HPK)?`u7dF--@&HvBy&`i_-pfPz}wJZ+~ z4WVAA1JO}}tR$bd?27Lz?}dk29**=h@8$zNlSu<$!p35WJ8PnK74ayr z8+qaNl?0h_FC#iPAn5Tni&N`6jQ|B%QCSN@rtgp zc|6SE+`=Sn`ur&&H^KM*W`2|v9E8>&iY;`>4;zh_5_OJCVGYKCxu0qK zplKk8!F9`8mDUnYw=ee9)+SmzjS~B7DZ_&%iBO4kjCE0DCjwx~{Qm|hq1(>VBtD>< zD7*OLftp};m=y;Mo@(RH{B70Z`ic8amcx8>iTa-h9-)7W-SDNJhj$=dz_!hn0A4=G z7zqH9^g$KOP$j6l8gyT+&rri2nw=b+Oa$_wDT>(F`(egAerqh#ndZ;c(Q@LL(31}C zam|(o0_|*_y^m~*#@G5kyOqIZ!x5H8ePQWzfzgewj-Npf zuIWTLRm{jJ4BNbnu+l$A3a9rq*XyfWH8xpf`u_;ypDr;NR8DPfZhqKk+j)T8iv~)I zI<_ocuS~+no3-3de5o7n?iNi+PhZ@UB31QO<*}T5oG4~Dnw$4tIm5_PX~jbu9MYn$+-K}wAtK!J@46MQ<2SILGTOW)$B8dqsh@kdRw_KCjWu zii)iK+b9s>XNNn@xj9{hdH?W9LhHiEN+dgQB1-qmdz(z}k2Ct_8`3-Xrc-;~X6tpQ ze9c#>#?`Ufh~l-u-PT+!#q)w<)03|4D;n|+=?*5KQQMgX2CrSyOL)_!4Le@Eq{}zJ zR5|76otBzw8$!W>bBA86eWAw60mE@n{u5k|Y$;VdeEXc9+@AguV;!d&3gXGV>{HaO z^AxL<4Yj-dS83{YQ*H;&y7FcD#o-wl$nK`l%m5<~IClEhwY*nnczF1;aNk6`+8yYQ zclq4V7B-XWr?mi=H&ePc_nR(lLc;5)RPE0V4Y86v$|cRFhYg; zo)=&(O-+;w%A%1XH~lAXdGE1{BMtR(?;17F$HPrI`muL;=5lj8R#BgO@zoyEOs5DB z@?bXn0!_4g6Nz3rZ~l0i#+i3@OuwE`a-5-;ljle+o6ITIZ-L>NN>N}taQIv=2VcGp zviONr`ghsaEn&Eo&fDn3{su?;fA`)?EbRLKW9%!t;#!+1aS0B=-6goYc5tT~3&ABo z5}e@f?rygfqc8voy)y7kXb?feRNCKjvSj{o7k!Im&iXz(BH0UHH0&Z8{Y#>0xw zhG(`q)9M_RLg~5|<7zB^ZgcOV-t-eyM_4DsBH__LpkXy7>@rm_ZqG<1P3)M7=ThVf zCIo)@0ax1m{o%V!GHT*!U284QE(JuU*9`KeL&wr?34`~dLzxLkpjC_@sc(~e(eVt- ze@k@~KC5{pPx6kRd#k))dhINAyL(ByX{_ANFMli%Ti8S`EuUFlm2LG`>eyL`;;KM^ zV~*gTANEc{8GR$(cQt9xPrGM1cg1}cSCCM2s2`5O{2~OA0Jq~ zR40_ZRE&Hry{FVez}Z`$nK{?{VWuGhDd|Se2JH;ygMXkh$i?k8H6mdo65`RPZ+J!nVQ9bBY0%Kgal$jgynJ z-RSnS0w*|4fxKuuk(jZ*=KP%Dt@M+#i2jR^0VQHkNev->gGb(&9`Wk z3G-%8y>8BPO7aNIE492J`}n4wj+$88lp4xt-wzBJP+FjZ|a(_h*$8)J3v!q9>Eey z%Q1&=d+~BSv0K+QAa`fT#Yp009GPSIN7pDZ=y-3$nPuKNu};?L>8h@}&M%CBS?;yj zn*MZ1D(ZYvYDsYKgRwDz<^o&ANe&y=)h0E(Ud(f^=qZuH%HDGcZcx7dy5!@tB@`cr zO2G48PG3pJyg0x2$rdG?H6cmu(#xOX__himc~ z17X%hdTmBPYL`AQ(e(yGhsS5+u?CE@5lnU~rqZ)1#a80Lu?|=@&|!Pym-09L0Yt~2 z6*fidbuUF2gW4Yxam;-sftgi&WyPQVN!5QGAhFfwA$p!mjmK&$;(>POak}n?+*^KC zTFg-LL&pA}gn=vv{_l&vrKFahElCAQJt52C4=rVekcwb25%4Pu&IYoJ=0-hGqnjy@ z*-I^4btFj1g^4TJKom-ig!Tt9TEWwv(X zu(nfo!P*-=HR6~&7w`&m#~a5O3d?+!+i!^tLk50`Zy1#WJ9amvM~zOSqWogf7ks)b zjaeYH=%U9)uT`z+T~`}^d>rQr8y9!x<%gR9d1*3L6xIoi|19VI+DuI^@jrPcayTI! zVmU_894o~V@ThwL^EkraDrEUaHYiJUIC7y+=9-aq*xEMk)I9p4D&ORr<$#eyGgB$4 zN_khwm^7PGMc%bHhu;Q~B4m>sk5~F;MBROja5;c4&Ev41qYk$%u8e^2{*QW>F&y%< zmGr7CiHI{l9Nx%!OZ#UoZ=cwP-9ArfyLcwbXoW56e5h*FZpX0!)fz`QfL+~171mTv zDm2I8zdQ|skJg%01P1py`Bfq;##r9& zNX<6@gy~C@6|l|84c`OF=%`~}uL0{zqhw4T4M8hlsF5CgG^=N9Y*4**=#tX#<|tUE z>3rd!or4Jy3!WWG>x23G=oet75k}+9sDELs>&XduKac}@?le#gx3v=d_k4 zFa80_k#Qeb;k;nU$t|(QEN7v0xf?6AneY2RvWMVe21yINz2{^3sq{DDG>9f*M zrYDo@wMn-#Xk%&a91hokL*6gN{a6KyOiOS@$shC*;8xDO;jx|@7*t^)f?KKPx zjdgYX;R-+>w>Tq%s(uQ>~y8Lmx>C z_cyq6#5tG#k9fyV@=QAklLZC;*~A%0Tsl&%5R!+Dlj9Ne7I3ohmUiZ*eSskljqR03 z8voUti=kfH{Sc2UdGQ^dPCL(j{xf5CW+G3?WDNB&IfK{e(4QamgU4NV+q>){Aod>c@~n8YnRQ6i*d=?*&XiCl!!^!&MsTHL zqEM}JMmkcyUc}HBk`ZyQx;8`A$HmGhJQuahrJKw{fY-D>>}SV`l0o=XEs8`hpGGQCH*eOND5iS6_vBS{&Y8cdjsX)+af4p-q@bBClRN+x%7| zkJd8O?~EgnCQVl9cF)gE96D&H)k$2<|kBq6`aq{7RrjOAT=XAx_D zl4p;!5L$^F$B%tJ#(IHD4HeNUDIT#TaWb}GLm^s=6(}kaN&RbN#RQ5WzKt51+NJK5 z$@GVc7_~?{(d0E)`yE$11JlH*!~%pX7JYvKnHW}IrIg7W=Fd2zmbY46}Cl;iMJ!uI)N1p}X zB~MqJ%goj-MiMKWyF>w5Xxs5~hwqfPzOK|^_*SN_npC^N$LUmQMJghL+3}v1(fd%&%Rng+4sj<3C`C~ z>j)%}{94v;icncC$*1!AH9g91XfW@xveOm1=H`f}AqA!1%&jtIP@;ClbF=XE8kPO> z#x#1oj+*{l#RunZ?=)r%{tYw|arZed{GzK(Tv%{P?p+CMtZF;>5uVk(@d|$Wc36r( zl1EYr(mpA2_~8W7&$X>#BP>qsC=|h0p_*&>R^7AHO5M9E84h zA^AZ)^7j2k>NcViihnL2YOQkAr= zYfY3kHOGPuwxyR46?dkUN9Nt%W4KecrON*uGIVqO=o*sIJ&vb1$a`$&XSO4tV=)eo z#j~;`+mm_9?`}P))5OVV$6E+Uk%%^&*kc9>>YDqOT%{nzGKYyi{ft6bPETaMS(PxD za|l7;o~+@L>WdcenrszP;1f@|Sy9tYGNR;&$RS(kIpMx0R|6UMUd8+lAD=>y_+^M_ z1~#b`lH!BT8T}CuP02sOqXX|%ck|Dn@!Iaw``;7*k9?vV?uCtv$`LfG$#)Zj(#7c;R| zHGCgO^8Q1E$a-B*^mz$yrm9M&$lbz(+-jQ`7Z;AhafkneNJqBq=X6dFwv4>zp}ltP z`2Dh?;~66EuEva3tHoZE2FoXv0bnWW2RxTPn80GUKAM6&NKp9uAO*}{OWo!2rbQJD3*jkn zTmTCw_#jvnG04*CEPXci!*SD^Z%t?|ngdKw+n#Rmz#T+C%Rdl}Pxyn0nMV899ViB)6z_A7?vx5BY1Z97|@3h+)$89GL{bsyv0%O}`GAn+3 z`~W3&rGjf|MfD4@f^oK$zx!P*X9fj0p<+clQb6$dZm2j7jx!1}8I5;$d6066havbV zkazSs1dQBsQvUGW?L_GW;Pk0?&y}L3h6uQHP#V-c53h;-?kzKd{Eg%&hEi8SN$lvvn4+jsfRt z9~zbO!#r-c>Nwq=RqdbaSqiJxjhOIjt{2>B6t=pp=@%{1`niqJm%$eJ>eic*M;V=Q zX$gtJQ%IS8%KF?>mR)4-)@I0a=n$~!PZ!0%ecn#biDgS*eX1{IQtzQ51bG_uFv!7bC3Epf@ToOH|bu`JxZIguO;kW zONT4}@Ma3BDr1a18d_7)bms#}?hzX%2nvy8ND=s1#Q}MNW=Url@U zn5|MnGy!#e%{s%ooqGKeN5KdLYR3GV9fevHTvVJR_9%otlaV-OWeCCVKE()MM_h=S ziWm>2kbl+*ZhQCCr!!#LnDCLocMUD?A;dUBosg26+gJ%Y!p+pP7XIm{iOc-8*+H{Fa~ zS@9SkkAL%jtfb&3d5T8PD?WsKGHrY)mRY$D!HK zEXE7!TmbtGSc#q#&!pA`9E5&ZIImCE}{OE97=zqbKI;pW6$ zM>#RslEO0c$x?xFa^)ZR)Q>OGI34U4e+j087nhY=Rmk&CEk0N(+oNSrR*V>hQobwE z(b~ zt{d;5hZT!{DEWjH-W)P(&b0CQC$VR0o+>G8q|J}#Ia5Zb@JJs{>j?Kld8%WfbDeAt zL()H4;Qf9+e?|J~#0weCIk#EOyv@Z8-rbD&ZfKkrTThPFi6U!nuDf&Ir(0>7|w^x!_WV@#6M4tn61-nOy-ckI(I$n3#TvvY!LEbDt&2|8T(VM1;uJ z7XC;;=KdXpB+#}#BPf);^c=B-$vi^|jIOq#hPh|plL#Q8{HEsM!B9;JW%z37$s^A| zu+KH2B#Vhfq z?3D64;tZ8DJ%EMt{&fmK!=)uaLl@Lgg8PA98$+<{%qq$b&Svqr;K7Ow7A?Q&I)qW5 zB&2Zo;f>~C)~S^n_Hp|e>~MwEUFe_9v*?Zp5+Zpc9o9iYb;C?@hipBFXnC(qV8-Xq z5)qVsZ`=R;ZIEL`Xs3Ta#%XW=VTJwN5QVfV7|WSw0dkYVV(9&w z-<|r_SFhKv5uy=0f)s^q%>VkL+JC+XDLXVuPX79$Wp7PMb{|Y?@zVLHb1lGo!_szJ zeSfn@HX87&CKhO=2eP08B$_vT0Ta^oty`4?azyyVh)A+9L;{Xf2mC6mbwBV^?*6$V z)$@&#paMhMN`keHGCr*~20qH{4^+gg`@N@Cr8qEjUVFrt%m>V!3)GtywVgn+S8#RY zL{L14Zj$fJxOp_LPXFi40qI{_1%r49>R#UbDvxlm&4z>X?$(8Du+H zy>#C`&TBME<;U8M>f=VLRuo$zla;G4?jko?D@uopc{2$MwiSEoBBwcDc~tpq{!kZA z_DhAd>=l$QYzh*pZb_eCk*8`4t$QOI-T7EUYH@8Gq0n!_UlmgIcHEW2O3jyO(@#$D zy`jQ?{DvJb##UJLI3E>lsdjuJD0oZw%x89P6Au6#^k?mvTjyGNC&jsZH!>VeXyu2pl_*7h1T7+#vSFG5OZ-#(VkeSF_fibk5$uiL6c^ zhGo=Trrz4&m6cS-(X~%a2E56ksT( z6)tQx76_H0@+K#Zgdc5J#j{=G_#ML{eQ8xhyk}!kW0K&EO;qbr|{R72H&BxRb(qU9#&y1D{@&Zu+JgP>FQLJa8MS$KrQyID7 z*k0@c`lWqVX%^>h*}6Z)I@#DbuyU3A0E!AxC8d+OVv^mHTrSKPL`-I>5SkIGx=^~W zR1>HSBT`L3#@%FiEeV!PIXxt?WtJjUN`{ue<9{8w5BcpEDFK0=5X72ly@JW))PXS+~` zp!Ee?y`J)cV6?k5e8}=$<<|9Aiz99GC|>N=kJ69}-y`o(ogc|D^tmeDp6!T)ThfV$ z*}9C@H0Z8IjcZKF5-CZR-n&b%x}d*z#PE!fFed4^w=w8cGM~i;kuj7Y-f|My-c3Qq z%y}paCB*|>gQ(e2Q&^mqEs=sag!)Gdj4-ys!0%a@q~Mld%FdeJb%!~Sd&mw`Tmj`Vsk=B*6ye@s>isah0F@8Ss649R}dh{ z@dT6+i_=*K<8=p1e68k#g&pgq!iB(5eeI|kCp0F#6uY^Akg!pP!Rk$BXs3CbLiDD^ zVG@b2lveD8?gcGk<4_n?)lM-auaVER}hn$EME3gr22FEW7PQZZo3 z$5lG+J3>?1t7mQ@qN5|XY{-#rGSH-!?2$}vOz0#qS z#~)Ry0VKgWR3wTf3%>Hjn0;#c2JuY08>a0rcywKeXCktw>;GFYKw#w5U+&Q+tbc@6 z&soo!?>BQ6B_T-2;MM&|n=4~gmYEi*AXVj+F~gEiJ2_(VUU;0bQb6#I`?rbxd za=9EvE#UcDjaHqMH@~t7a8R998$;c&zdcC}sSAJHU69ATblkOE+FhhOXl`mE&`}Rg z^v66A*4D4MzHmUa-L&1dp0Au8VGMlJ&Uk}kzB?BioQhb%0*{I(eQuW@yZ;>q|Mw~X z{2xK8ycv8D@XZpUT0rJfQM-d}=!&KX+n3Zr4z*cLZ2J4_1SO2s0@g#W^3#%%V2fN{ z)wf4ItW@Vzo@c8(nqNgb8dB9?X7>}$+}|N|%M=NzuDkldC5vw9F~VF>G&}+a7CiYS zfX%GhqfSpOgZU`Q1mPz-oP|`xVO?UJW{-y4W{t7ms&=3Zd41TeS6ieu)`~V8qZ>4* z5hN08;k2ZiC2l#Y;)2XzCJYpn1}j@kg|QqgtJ=FoOL~oFQN+d5ncin5CU)h<0SPBW z8pPMJcHx9J`fi_T zcVxd}`MAzUfS)#0Peu-sYIJfC8M!Xon8Ci>-EM#rtP@#0!bGme7*K8U(4^oG0Z09c z5T?^2`z#BmK!=JJV3F$l7u=3_)LNgj)JS&cJkdG?7owzqa$(=pXnU@HIyaOgsgG>x z7+d1yenEICxAn6>P{t$Vwf~y%z6=kfG*K!z@gyoF1dB&pfS*BLuZ-N~hEw6>8G26I zWqZS^oD6ikJ#lqwzO5l6Qm@Zzdb=!8F!j6W2h;A4(jnelXUEH~DRJh6viroU)S& z29p$?OltW0bnT6|9!kD~v3r+=g$s~PA}7O!5c)-;JI3(crKR}l#kf-L171b%Qnj7w zcO3XdJ=c&jRQA3D$RYZk2eEg&YoISV31Oa+kuw(Ija?(D?-NZ@x@{PR?`)vuVISfp4_9csv zgS4(ua`E_MT)eWdsibj#lZkjKCXo6J`aX}UalLHy^=S=D@XxvJ`jHMi$b~Q-{F#zz zG!eVq)vwzu?49H6P1-)OxggYTbAhg5PJRtPyZo!wfv9!DnAuO+bRoQHn5ZzlKNgLght_XQ-50 z&r`;;UfD-==tWI91dq`XK`t`*511O3i_t#~VxKeK4CY%&0_!VR0vrXTq;XUEaMfn3 zs1TI0Tce}b^M&<~nHJ!%i;c`dyUj9*Vf7Zi<>qsY&?;ODlU3P?thDzVRtLEH56L7=nVX@nj*l!Y>0V6NO*Q z=d|BRA!m`_4*ylo%>RYDl^lwP|56GNQeow9LLGn=SX>-E+g`>5wWRaEv1J#ny061& zPi2qRk5UpjwlJpRV;M($jUnWsFt+Mi)Hu!BH%P??^pdK%ou>Z?`QDmJeM^zYn~m+{ z{5CSv@E$74yim{f6`pzq%EqqCI*qmljp{4hJ$t&P57{XtPpN-`l^QFiS<-h^rmaSa z6dv1qzK(~9nUB2HX3$m%^V%EYt9=S(XoS1V(j|UaaEpfiS5f!XG}2T{-}c7$OC%ii zFj?Z1(i5IqpOmaRpmKIn`Fo+rM)<^=@@q5f;2*Gjq7HiMn(=~n!^CN-A1lzr57%1a zbuCz{stiMmTW#Kwh)jX?G^6+U1@tj0zAX#x@gW@u%DlfAW2@7hZ5Q^FzS~RV#`>1E zd)$3r3giy&kB@peU0z7KfSD9IQExwWugu)wo}#l>YRC&=GDDJwV~5mz&B{en5q@|I zBHBq3oU*DEE{k*hP+D5y(w*C*Y1n^??=-0PdGAMO#D8_ra^^2dJ1ZA|dAq#qsq+_+ zG^~f;*LnBLp$hzQ>nh)C*b|+7_!uWr=6c`Bbe=|s=cn+$Oj-sJisDmv5~Q1-7Cj!i zItVE|ajc7T?-4D0w~hE`D@fc#mZ(F(aG`!E8SNg0Xp}{6BkQ0=U<`H8l3j;7PF#T# z+KJ&ua56;14;(+GtIY` z(>I0M*q=0BehmIOX)n05ViaY)X1}_+G!M#FzC__05Vl$Ly_=sg*Br>Xb_-{&D%3_s z78z$~xYiFh97AyCB8VkUWZw1N?{a<_7)iKsM>3MJ?Cx?&sLQJ}yqhb+RNZP-TdFP- zTBtBWNZZBeZV{1fxkY2&phc1ZTNfLrP`^o)0tw!6y#Wft~p5%zHvc!BtFJQ_jRvoV?L?QSm3&h*2XDph& zps3i?h~H%OTj&?~R&#zBu58_~Mty;Sgir9u|mvvgcIYE&Lz66b7$v^0(~V_9aqBovmJi&Nx%x;m4|bsmuC}yu4DK%pFs^~kzeK@I(L%=#5+w+HbPrRL6Ncy%W+TOxFR-L z>%vIdn!R5?s^xaOr#f zUmb_%O_T7W5+MjO9JtXk=4vn@urU|sSr4`AW?NyQ>QsE1)r}TFZJOgF^~X5j(<`84 zn4#ajs~pQldJ$RP4f7N4?B^C~0i-V31KD6_;JN!9l&ueABf+vgOnFolPQ;F<5nqU~ z0-7M2qBse=*QBDXM~evxl!E1oE(atSAXl7CA*t=*^bmzv8eB%_@oHkFm*L5<|+_L-2!-)9UYoux{F zNy2>V^@f2T_d4RTXnK~iy8Jryqcs?TaV_VB)Z*gy;v8{g-QTjyB0~VT0JbEbgJxSE z8?m49_;*>UconZ?S`3*G95i~z+oz)#5N!!Jlr<`&#KPk}Uy1uGkwmo0^xpbgg85iN zm^U1hI_THc&n5gyb3e3*!2OAqG-yTm@7QcRA3cp_SQ(RGexP~X3-JyX9LvOLOAQbF z`0j;WfGh=9V=o0xpQSo+?|K-0L0s_lk?V6WQgPPVzuhm9I`rr=TzYu@P<75Z%g3x5 zeM%BQDyZ;d5Q_RfPTx$rS(5fcXB=+R`w|QWt+i-z06lpgsOJ_76s(v&T722invt1K zjS_z8u~ANT@9FuCdU$_!AF6diV4U`2C1OpDXFIaOK(}9qIf$Z4tqBqk* zpj19Qm5~;D*S!R*-Nv;eMk(3F-s$G!;NcbF5A6g(Q_5&uG-uSO!-RnWPPhPS& zs0P1hj(-<~`zl;nP_{(h{LP!%@`)0CFu{NGW-vj4gI8twfNNVpG*b*ja+0B+8Fx@B>s-a@f8gVX2A@ZvIs;>B{q)8#Ux{^s$J?ATAV zQ$~}YYipPWiLi=jj#NW#%a*5XZ#-kC+P~S8r;jsf(~cXxL#}H0hZ|s!mh!1fM@*->l?`#)7%?Q~Tv~WbVnY^4hQlm(gAj--` zQ*2UH)Fsn=&elEiOKs2rL6w}IzbYS!g)q%EDCSt7VTuRYOyGbQPj60%1vbJORhKHez<^}(;X8BRPM1_te-d64bVI4b7qI*59o7Chej zAOR8#oCfX*y8HnNv`@^01TB*ma#Gzx%(vWODw0F~wo(*#tdhoK6Q3qTE4Ls0$_AoE zq&~n--RFbuBKupNuKL$n#x>6T5i?5Yno8z?dwt;#F?+h5mwNv~-L(IpZknhFy#w1{ zs9TsAHbp=olu|{_Pz6w!bdJshEtIfE zgL@)8+N`Rh;JSU7@PyLs=B8~uai1I0h&E`x3r0ME9y&`Kb61u(y6-Yt-LvZ{zVxCb zOjxhHCNL2-8WJ%#SKw#e{Gc6x((Uplw%lLtECkr*r3xYRp&`#$uRJ@YHM^yY|Jn!U z+`AvODydd7RGk~P7VP<8TiG2jB9`CpZv#bnZ8d@p zH+&PVv45n*o}8+6VNk##*zeu!P}QE~EKu4c*AJZL40$Hbb%3p@^e5#PXFT;IBvu)0 z`Q?T7#~|8pxX`)j5z99yWAP;IduE@jeRI178{h-RnMQ{wtngWj!9etY^gbnq1x4m9 zCm6jMB1r|C>RP=B>&PVGaZk&6SlAk^>TV2Y+$8vu9KSxFpxaO5nX3!@$jr{7g__bS zE7ITZc1((|+zp?|t!}oSujqg5NGgV?$|VpiB0)*1)D%FER*6~c(>a9m&zBiYj`MCy z13cMSfD-8o4x>|Rt=g4)#ShOZe~BTES*9LHs{E@h(&vyn{M#7&`&{DxSA0V5GOsl8 zVS=(>(f8H2Lg~pP3b((4SI;}-NvW3(NU+CdTD*d@qX_|)LcyCvb2$_wAG`Wlr!4Z^ zGbC`G3RCxGePqHJ+UTP8c>|QgpU`AkqRMr#XCrMDsDmay;HH>eZ8>%ihhCp?i_|l zi2Nn2xV3=&1RXGYa45z3%QQVC0*;3#MEC^8TpCK4s|;q6f@hiMyiYlA@GMqswh~gY z)wp^(Rh1;an|-uf{N9)g&_~-lSel1K)|F5$@VKiqhz>j&g^hI2Y^BbI(apm(%n4!X z3{~S5VX(VHZ~76jXj4F3VK!Npp>c}Xpo(+WL=nU3R-SG1o4TDto2Kx;QaRByhWBmvGN4h`#6%`1a#LCb9CW*+F_cX&Ji(mq<3;042->LLi`fvh)%4y)&TBGyL+$O2G$=yXwv)AmZ0am0 zM_LaaXQj?_{6)g89%H{D1n4_9^&rgWG+Fvuwfcjp#Kltfrug~>s?Muf?Q zy~yOS!y$Qc#QvBHLF-~FTrpV|?D`6NTl7yz>QqwtBX6k80muB`E&!6Dh7lkhr3qMW zV-+0$N)W$3VSK$QIWeM*<;CTpsbNMYf9sMp3f88nJOINACXOIU9%Om2vo@yLGi#7f+E)sRE=}ASE|kfBtl31`Z&=eI3ZwCaTVdZB|0GI zMR-y?)_aV=fxuh--m&~cGGpt+QhWpW#%uZt(z)b2REISB+lKyXTjHID+yBKv#=csA z2ZB6=WAX-y!7V}j&cvmbPa|SQ=vd~?7`Ct2H_aBJ52VjA^}U4SlWOFc7x%&30WOCd zsD3gGV~sIsjiS_XXi&M^L`sZ$u+9hx2gG&_@%wT-jDhnzia$>nV7?FHS;h0r4Z`g0ZIKVY2^2JX~z zX#(p9dHSoA7DdE$_S1Hu(FTYouy}XgGFB9pM@d;=LvnXM&vn|x7Ypw3LTR;TsLcix zTgd|{y@M}txMuX`3uk&%-z%7Z)}zy(AA#09TmM|h9q|71_YO1|zxD>cZZZR(Snc`R z499B>>v%xo9W8i|J_H;&rsM;I_(3by_H41oiM9rP`1)o1Xh#ORHyodh0eRDcGJRYf zCp-f*TR^gY6rOyzD;P{dy-b(|8N2z7?uVA%we#wUPCzHGqPo>=mPC~B4DODM4M?JJ z>|Bykw}!V@=wXTeftc)Hr!8?i zZQa@+OI4!{vVYfvk}+TFDAbfM!T4h+PA&KdQ;Sl&_?Dgh9(9zLa~aYBL7VVJ`z5zi zP-6WkJ#06_K1Af_s}|1y&}{d8HeUi0rm>^fHdLt?afbJ@>1Ro z)49BoBf#AzU{%}nR5+>J<=;h^ve^;~IVk&br*UY!9u09V)SQU~plzv(Oj-0@G(m0BD{PDzo|<-gXxa9Oxq^oCnsya$Z5!vyL%Nzx0Oe{q0U7-?rbPG{OHM zaZT^{M|vYyA!z{iL!ka_go``9d3a!pyTy(2N-^s?0H|Z-85|)XM>mnmSv=h*{B-yH z>w(Gr5R@7(QqhU5oN!{#k}l}nZI?_5o*CQtH4x3HgZw~v*L@Klt~u%RCE)$?O8$Fc z?WxPlB%8*JpAz9o(cZlen#NM!ibe|x3Sx}tl6k|eh^OKw3GLCInaq~!D3PZ-k z+&=pLR%5nR{5w;W67)B3{`>wfQG5D}Ywnw{U-3Ic5R}mkA~a0!s%W7iDV7~9MsFYE zxyQ9D|1D)Wb0M_2ag(wfKl{$-344PA(pxjp44#vFZN+#SW)8b*R;zuR z=ik(Iqc=q6zIi{wo82VY#T~p@De#^>~ zV2>@~dAt=XJrD|IAOU48L+dEz>Burs^`s>T9@sZ%!eGjE$R)sgQWvWYv+f(QcxMY@ zbee!IG^)`ptdTQx_e0xig(#PVZqE$;PdOQ-riu?)J69!|mk+?% z5FO5vit0BJ_*?U&@c*x3lD|=KLa z>~b8!ryWT(_{s2Rlzo+AP??i4LDY23t|f{JENo)sgWux5HtBNiMs>ntaMC zS02(_1x~F*@2kiGL-$w0XF2TQ(GXR|%^K2KS@k1cMBHB^E`2uu{c_@&2KnB)y*nD(cFHhgAR4;4lan?3B2_bMV4`UP%5J7x)cW58{D{I~%&{psy~(qt?=ij`rH+fjSoMj#qNuPUgdnGc5D1}mTZ}H=_zdsuj-#; zwfcYtz@cD4VgZ=oaAfz!#S9hp!`|*dcaQByV-AQQ+68HrB5@1i3$~<2(k|{UIkn2;CDc!?7UsvRIZ$SEB!#hC54YTu zVtSdI3d}x8FA)0vxLa#~&JqegU1@RUg3<&muxP}S=a_W5w-czjlO0$uciydNk^RfF zDR=zWmwdI@%Rm1+O)b}`HTBk0-F3m9@pv@)7(xB437ACy%<`4UVH(s53Vo5K$)HJ? z3{Y*0(b2MtktY;4?Gped%v!6(<_gdY@W{+z>xjI{I&vtlyZs=3zy@W=Jy!r`F>}K_ zivL{vplsH~1|{qe->Ahn1p&cdd`fA;oV7|$#OtHbaT&Ls4fl3Ei<{>-*kjdCE$|@H zxtp0NuINu0(>FAnV5{7)lt%70mqhz3Q=qFRuB6^r5|*SFY5{nJ(t-zr7OY7N?uv)% zfChl}Z!4+utd_4bhKk*60J!f$U6a21*cAbQsk3k~?}RN0Zxw+$oFbk0MmbVs&&)d! z;90%L3BrS{P_QE4Vd*Mgze8_C$0X!^-sfif#4rK}Qza3(^n{n`y{*PRl>KczBROu4 z4ZI2S9vx5~E4d9+zC^-Qj2UMBzQ$OLZ_@hI(U=fmzo!Kv1O0#M@Sm!>HQjLXAsVN4B@nn z^;@s2gHf{{(?K7meP|lc{4E9P!ypAUe5gs@EKr>QM#l&BNDG4*MHaY< zT0g@QY(y;H^{x86`NraavYI-_H)k8QQkuA`pP+uZbHDBFAh0UFSYDbS_YO&wKch2c zDS@7w$c{J{ESGsr2d7F0ZaH`x2rITvKjXo3*_B|AnJ6B{az|Gk_>?hq-SV|WD}dy; z9f6wJAskoo=$fuSv)k)0dwd-ZiRnQ8fO%P=@(ZHC()z4;HM9H-9W{BS;jS@xRlR1RcS`QJd-`2Td+^k>xeNAI&Bq~#H+Ht%(<&V~zt^39)5 zwY5)4f5%*+JCu}$k9k;{gK#oJnF0=}5u;%4Fni}Id@6Y+bX2x@#`UaHRsnOMp0fPn zhE12mNsVk8u$ND9ix3@{p$DH3QlF5eso`>_s2q0MNhhk8e4*?QC{8PY=__XY4tpqp z)R)@1(M*ZU7L=aw&Rg!!5>dCq@+VA9DRwX~Hcc4NbMB4tVD(DfbkcC_b0f;y&l|UI;R4^?mQR_ouw%VgIk3~E0J-Bc)j0xS!^wYP z*cx#k?{YR>R;Wv-CoU+1ayGy1V>V}y(E5wW*jLlsVc737$sv=EJA6L8gJsNyl2UUw zpn^-)+#^K!H!LHDNwpj3eFvxS8n>#s!(Tt#zUf04!Wp}_l2CB6Igi?J36z#)aw#B( z$cb@@sCCS8L`|)z+l;w|${5~0P;Kd_63gY&9itPHD)J}$^Sn8U%#=8>BpsnW?3@d&oOdmy(e`@neOJ=xFvk`ulkk3+|vXKinfH<7x^)%-{@4>stfi!ZM*n%Zk3vpZ0?bN4b>}DW8B> zRb}qA7?iT^y?NXguD?d8z^O%JtOQ%)8%B!R^f@@87m&XB9h1p} zli%brAZ68W8`NWO&t87?ADOjV{wq;F0`=pfkktLPVBP$-|C-1ydH-4P!mlShQr>qB{VI`PAgdj>Ncb&(H1HPMc*psXnP`&jx2gSq90Y9#J_v?_Q>j=K# z@e3_BgLhaWxl-iz6H<;Z9)vGH65}wM5d;CW+r&_#wf~2*zYME#>)r=Yy1S$sghfb5 zw{*z@)}p&RrMpu~KuS6mi|#H-DM7kR1OY)sK%~yYySKmnzs{#~E}a~OIwoTdW-NXE#FO` zR1isZ)lu0eSYemT&A5Q?SA=bPx~b`Fdg*AMTYYJ_;iF*mZZXyc3ahD$(gQU~NM99&` zJh&U~z3%GqQQzAFfeztB9-eQSzj-A93)3OM}%8 zjmYIW#mbt9MD%A)hlmc<`kUeaPnDuQ1v7q0LObi{j+Va3xQ9Yh%fSrOMUO`zSS~X0 zKfZrE`V=mwz=$zDxO&T9_JClDL|U?o7Fr{sa-dvzU6<7Bxlk^O;2kYde1^b}?3hTb zDYR~+!6ls`2VG3!3s)fC{4LxvZCaXnEo6@d=6kP9tGF379URwPWzbW%Tzb5`e8WAI zC z-L6d5we!tJCM)-2lIgKu50NAbo)2`U=ykB)Om*rYPiX(sml)Ud??8sQo7glWHrj6E?xLtHJ{$ZZ14*B}-d{`>0!sjg(S(@+|`EFlwzsNV1?|8PHulr@< zT~JgbZcoAw1O}iY!pXMKZIha4-wOPfzY^iCwvZlYg#~6!`C&Y3{2BuWP5E+wGV6o^ zi9R;_B6bH<=$PY=u&_AC^q{v&28IkN4Rj@wlvKvO8KNmrU^2po*xhyOauKg(w6+&~ zUT-CfJ_oXDy$C+(Fo|jTN-Z7vjXQ+S7GG%PX;rzgm5)nEPUl;Xj2NzuuYyg1#!?9d zg$0sWP#hyN(bD2#RmQLe?liKbst)NvPz#P0&aqvVuEFF}rg2Yrbc*AuVxI6vkr8YH zW0SR$t-_-y+n3pxEUm|BUM%*c2Uz^?Jwg)_64LXf&I}Xo1{MhHKNe_bZt+)x2YZ|H zpxsHm`XGIsYV%BJj0*L5(A&o%0^?Vlp7E%UP&|I{PcI?RO<41nan(H#e6_wQ4RJ>! zF7o09A;`L7kKT4Y4^+DQacC{LkNc^`uu?u6?4Vsd(D}uxX4#>Qr(Sj=jGw*V?OM+m z+P=J#2#*d$-2IT$!w*ja-3bBo2@(lVkx+3%6$@HQ%t-95lEK9;2B)YH2DRlz1Fanu zi@pVHz0->}y)k&%zZk2I7HunS+2heoZLJvw6^J~Aq?EZ3;=0_W?uUxG2_M#vd7Pz? zaW)>%R#w7JQ!^IyfT28fLFpx@Y#hs?GPV#@P5C&959N94qnizHHN7*}hWcw26+{Ql{8RiN(jrV5gwa!*6SCA~7oc5j7O&X4ec> zi%Cek#vT6O9IVWSQOF#BtaQ7dqX{gT)>$TU0kw>e9~J$7eTi*Z&-ZPQ14JTot6&d^ z|8s+*ewYuw9&XsjzP~@(YrgpwQvTBPfjlyzPOK9lxJb?qNKmh?*i6ryM{}vRsVF<~ zj7xXR3TobnMy6zqOwzkkTJjHrDKID+n1%S_`4iZ6CXnfq{LNtZOmXlCTgaWKgGcaO zku3z1Xwa6<7P9Y|LWr-qKyOwbbn|A}))M9ZeTZ&fP9XSco35pL>ze?!E^cq^$xc~) zw?cb5K6~HPb4P~0=)5-Z+e_C<+)M;!3v|BsWWp8kd=cNSQ9lp6=yOOVU@_fbb=;A{ zg+ikY4IK|%glRl0SwCI>*3GL4grWkja0YPE%;F9A-4!u0G}Ghl^E4}ipzMjWxVq*S z#u9@LWOa_c&n$#gufFA|?;oT}Ho?RXeos=st3PXK#k$ofCNpT@(!ZX)KHo2SXb~}W zNcFR2==d3szweyV(1P@``L;m)r>X}>2&I-Ny1TWv4(o!4D=o#~|Bsn=hn~ofGZa=_j`Z@OXa44XVm?YdmdH=Id5Wr_^fbe;{d*|Yml#iq z({k^)C^v2|RYWo*>P{3Q3N>LHB_NYc-`J8P!`Drvxq#xPf9e`~`flrXU}2f|c*Or? zZZPEQ#2>$GuLSc5+H9x>qN}NSj2atm9Eh60c)901w_Xz%U{Idq$TJlYO>C<-wN5{t zJQ&rq&dE=0FZkpJWk%`rH7%kJo5o@LJB6o5g+9-=h6Y?xH7|sUPdDcP+gDiTy3t92 zzWwv(yHzsgPj8-T_-_4tMHAFihC*@rk*(F%3NlQgwSZ+SVmf}ah$U5R9cuo|2#OcZ zSwUG%ZtZh_`@Xj$O)CwGEM4l~j3`2wjm|)EMvdR*(y3v3Eh+Q~oAE@)RI8*{Ya7S< zzEv=;Jwa%tskt*fyxlv?zpU<`+}!+Mntwhh^5uH;Bh4?2Od2JnhfV!@2W%XpXGZ+m z5iaw~R9r=D>t4cZfVwqtueN+etdB)(41uRU`Ci;AvFwLGfpW>{70`^8enLFwrl6gB zo@*eE?8LeVwvac{v`pSgPK&stXdM$FI7{(U2v|SA1--g?C=6f(*fBD z;&)%9kvWBHB?FOZt<>dMx|F@je%^WQezf_A8OYEtTgFJY;aP95+K$!F;w&>gOn4xh zGP*{J^B&t>ug$MV!z?yr-c(mkSgowzKF6(`nRy7CYxZK|iNg7IP5K zSL8?jiL=yc?;m;_|L@y~!j5XaS(;un#SWK28TZ&k>bL2|#Zf~^av6>Rt@Enm57|E; zA)5x%mUpHqTfUD`%zzC<@O9WxE;3w07;k7;L681Cw2*oVM z2D{&vHFllgDm)%fm&RP+#|rB1_9&S;r~)&Hmx*vPW=#%}U5PCPHYnUkxHZEr9k@6I*4!+aSI;k)HbTF2J-xu=9Cb3RQ~8LWN1dm3C-8;j4#G z*y^nL0d4*n0a2CZXPTCcCeC5&7$d`#KhF@JU-hcjs84*>r$xGhrP6BpJW1@dJd=TF z9#j83Ggf)n=tK9@Lu6PYf#1yo?m11hsmqY@^A!2qN!F(PFL4&ca@33B0pmpY({p4$ z?)k5X&^HaVEIjh1?XL+ni@QQCMn}vpE9`1NPJ1rS4-Y%MG;y7wE5E^yEvzQP_w!37 zw}qH+YhSAa$AE61hHg`Yi}}#qrbKh#vkQEr7$Y!s zp1yW|!6wj|!f}`*AAvxh9-Z!eCQ*#C(=@ux(d1RHoN&@uC6sbi+3*(fsXZ4>r)*6G*1d3vfU`A|z-`agNY>%ABDN1-u$fgA`E zf@KdmMBX+|ylzkKMALbkOc|6kk&;Z-mhrB=?tB4;t3DsFalXayx3p4-A8%-73ga@q z(0k#KSI2?)-{EiuD(gVEetr!spI;a!X-a{jJyj8vr;u{dvj~bY(~hfg&Ohd_0E>)& z2N?P%!gcykWA}p7*_HxX$y#CvIq7o{vRwg#s8Q)jLD~NCB)hBFXbY_tqe(98?H)3R z$NpXJawRHv(7T`vPe?8>*C<_bKe(a)_;G&al79KGzuiUD4vz(xOfq=+d7&)G)Ek|M zEv+A5e~zA2Vrw!*rbbf8K&FC!Y7k~r+`L7^o3#-;wzs`@EFQ{qKiQ8;ls+#1A}|JW z2=+WoqhJdp5a_?H{*hB=61U`(0aR;M<=T#nRN>I;o`ZK zG6+TYA<5^#xW6406-YA(-y46F^e>krCpoaV_(WFrdlKdlae!o~rujHlPuX6Okc}-l~#0mtrPb6$K9AZka-GoI%ez`Y7#WTtPYVX_Q$u z%txr8Y_sP~tGHN!isw*j0-F_N4YXURGsW^Km$3<8XbP*HpibpYfu%qMK0;$}|oBB5wQL2%wi3kX)^aW=i}SQ5JJz+65(nP3h~30 zp52laWVwvvQGM-g!4FRxU{e)|S1>%Jjuk$$43jETU+;82nd#}(J6yjuYt3**aYqFX!W zfTa?H>6MdUM*?$tppqE>_}jpIHS7NsJcm+)(Su*1x=rF?6(4Oez{LwN4KNj5<#5lk zHDZV6t_Q!mjQHj$Ogu;gr2m;0DnLBM5I{?Gl?Eqro2Wahur}Ku;u2<${FY>4WqUId zaQXaotC03v-gi|(l>?b2-(R_~)NtM{^%uKsq++8G4 zf0YNIT33$4AMBcd>|>}<%%EZTjVOER?ohW98Rj0;)X|DyQd40A`C2G6hc~o!Qujd7 zA@}^JXzPSSjKI5}0ag(8%xz+l^`tc4g@C~2A*MegWe=q7Njb`elamSWL49H*k4K*XqBFkzeu4suTG1`Xr(0V;0GrZ|g}x z%>HQ;en80Y?+mHlnffGO5?!e-e9qBklY&k5T>SnivJ>9;+x5OtQ1Pa|3m~VS4zS9P zzA=AnKA1(>bml=fl9|K_hy8JshwSXAh=&!}47krv*_`hR-Ef>e=?esk5oh=4#)SP* zPI*hl>R+5i-1G;_Hfi>=_ot769UAdspKZ?~=to=+&+Gre$V0o2BRWhGG$epdY$>v) z%4Ik|DRuzHxNzv#fn`q0M7v{a$e%B(=5F0(g!ZFcjr;bWCK?^vvWDOgGu8 z%PqPkm7yAL_bo20CGJU>2d|Ieb2rshZngh1rFG~ppn&pAb^h9#mCG21Mfid_Z8GPX zgJ0aJoI5dX;a)-B$cHT+@?p&KC*pGcG70L-JZt*4!!%94*`ijQ=6U>mML_x|j^}94 zb^ezooI57BuCpM!orAd6et(_czZ(t+HP%AL#W<1yopOD2279K)6aVeXYyKZMizFg? zpC+E|`>BYy`ZGUpUon;Nm&C$A*~UOx3aIA$P~M%KW3ycfJO(dsx}dj2nyxg(R9zw5;-T!IZt4|p-wroxKX#Rm!>!pxx_L`NQz^OKCn3BlG zt+Uzqn@O3C@qPhp&y zI)-8*NBCm;*K5u>aU$Deuz-LYYyyd< z3V+8C@LDGEllLXXz*C^5U1YiYnaoAX&k)IjFIbrrP8BTEIjhL!#BYRV?5(3@i;v4` zjSLre`Ley>5QBfXuxA7RQVV^N8VfN!qk!VjwF<<0KmCBDXx`raq{o@YG0>jIU}9Y} zvj_dd+}E@4j3FxmsOFzDAuDC~%yj;67`%E%Aj`7-2jd0iC)M;(M;={Q5%Dh=wN>5z z`Eta(QI71)AN?sc$LrKtUgOysWc!tYBpP4)Pgp@1nxGk0Z(&CsU)-tEFe)UDYbmjQ zeL!c82>$d%D1@ctNQQ_omKcquo)X9*A1T^nca+%TYg@)}R!aU>y~J=HF4Rv8C&4Cb zj#$IPGFoh0ut6j$;kPG{QR{ozIIMSo~Ts zJ=oQFO)47(nFtOAMq=H2@sD!|D?Cr7%XoD%bH?J-@|1Kg@}_AC~JbZUbqa=^~v z+pUO<>j^~Xid(`YZF$>KSjqNHogL+oDF-hpULA5BBr1MRZAp?8Gi|Ww@@xE5-AL$L zLkq#ITl$e7Id{2;=h0*pLHz_a!?0sze;uJtgQJo@c`kaX`PKVk6}lLF!C4)_m*g)2 zuxM;S<98NeOZty(aZQr}r*$Fr*C)>u`wD-GJbZE-dIC{UIJ?vG|9BjNv3u0BcvI)R zoBy;AA<=Z4%_#YS;MFtH$JyJ7k5ifhna5@y=Hl?B2vn4KkThG)_X~D!S&TyRT$Lzd z;rgLbOFM_h@I5ny50WJdt-2+>9W8x^;qO9uDL6kj5*x)d70^sO$bd?;DUpzDGJm4e zXMdU3rm0i~kHupJzt3I!a0-7k0u$ie1IBp2kF!6 zD+VqLF(nG}q3dtL{tou57#WnTXNFKjkBRY@`CC1Icl{Z=u}+x$ZNQJ@F}~x@ zr+kG5xZeEFPm{&;1IIp`iqXu@yy>%+_*^!Tu9#k$Q&x)B*@fQJ;mr(Zvk$YTFCUW( zSIWJ{r7*8{)ql^!%)*4{I9{pxumGOw%@uN`R@tRy-K+1tqkwj!9-ggedwrEU`0X=V zaB*=VTnE1wgZ5|#SU9TwuEsrHL=E+KtJ_w3cPCQK^@d?#Eed1L`?kh1+TE6_|0p?Vw-xNO|-FZwF!T{`==S7Wf%hF^5q8qP?UM)v4H z@XsFynuev#Sg7ayyk&L!>Noo5FrTSK=K4LNTn0%5>CcK;+SB!XIbBfpA2$7ekqZ#)X)Tq!V1B>kifv$=&I0Y2-hRDiiy(pnslO zHYOg7|H~E3qG9$Je;CILiRREk7{^D|3=66<_Rzybkk&zuSAdD z5KuYo-^{;NKNjQ^rj|A?R@OP+e`@HhoShdi=SOWR@0Gio=-UiYU+l2Gs--!)yGiJ5 zeza-$bN{y92=Fph4GQdk^}hdc0CbjnO4VBp?K|2}%la~~{q)ni&!1Huav#O^S+{e1 zHwo_|j9V;oiY!%uU?jTxwf7G!a2Cy+`;hLF*kDjgJrr=OJzB#g$3T)cLP$~(7b#fAi+2FB^XyTs)o~s zpfE@QjD<5`$q&|REp}AgF8nF9#ID(PzE~nPZCUu^e zf6C$yh5=1N6&eG4s}(LxNN7)x#TG1 Yr$Rkkq&D?yi>3a1__f5>DC{Sjv0TRuiV}cxD5_H6(WkfV^_b{9<~+Cj z^ECM2x9W_k@clf>6nCG}vF!Ej&neSq05+4v$7v*kno#b zRi2%$nrGaRD3h>{BYxQY$Tj z!xyO=i9T6G{6pko8O8Cr28AtIH$;&yGV%~Qtw=1JNL)k)T)b6hnNL%0(djYgHuM?F zu(93vhRHIu?bx8R6ih*cndUf@$yb@l7G(GN6vIkG#!H@Ev69W}W|^b$qC2Mu0d-q< zqOm}YLe|zE@^nW|#fkaxiJzdOQz;Xw&rPo5?|=tMxs0SBG;y_}v&EWRsUz2L;AZ5f z8Wwa08+~F5BkDGgeILf-QttbAd|l&C)jj-8GsDS-!CMNIgInJ{2~0A>z&3}}bm4aa zxq;Sx0|y^XA)h7w&n|IPe%i|+WCLhx!g~62jBfkZdB%~690$1$|L#P7oM&$ z@^!%zsA~M$Vzsi;%;iRn^CSylOO&$L=k^VAoK`>@>#dwJw%hC99O7<0Zav-wj^aC$ zoz|ujggior&gEgYA20dJe+Dk2{VwDBDr9k-hu9(>6=;@+{aTpL__gD*K6^hB8FVY_|M zr0dLGHneqtyKaoqN9~CJ+PG#(u zlDm}P-_|-v&*ONrLvJPxt&k|@j8?B2WIUf=a|fRH?lbKz zoF8SXMqi&%pN;z=Gk)XTDRhzi?X$lSVK3X&U~GZOoI0d4xHG}vN^1tO6|hv19qX?**}EjRiu+-hs}F&8e^+qeC9r^aeVjS z%m@d1GP{u^P9#pI+REl}K3R3%MB(8h>T#X$<)^Az1t}z+K@6Dt8;L=DU&j9|3`Frp z>kFjg7A+WLq%CMiMOesR*p^|bHY6G}BsHD;dl6LBd73IE5w`f7&t>hO=siE<_>=yX za$39f>fGg=!;RzYWj<6Rc3>glPD63Gq&m3f^Zf_KZg`8QUhGa0pZd1Q&G~WA>gmSB zRYO`^HWlAVpn5`Ea34%pQRl^OCGAAQZ`$|M1_QdbpS0CCyL7Rak9{i=fYHNerB>Mx z&QO4RyXcid3;b~K%zRGab#G?u2NY;9?Jv)eyM$GoV*V6}rzd~l&czpguFW>MR7|(1 z0Ra2Y4e=$PYejC&2kP7DK*bv2tEG?wM@){8&5zFx3M@ZbPg#V`**x-iCDKW?EiRs+ z5#Nu;)Zbq>4TbL?2X10}RTWRZmpNEc@$RnjVH2e}K67{Rr(p0f`ktfsUFF{Rv!z@E zD#oX{kD|hG%!W5%xRktGTW*3cb9ZZ7CjG)fDVK#;#dGv{UTvrO&B)%qnS5&&--Sqv zf_t=or^7~yk26DLvILG$J+GFQJc1G{U)m^&Cr@w78}2X3NJV8CoM#b=jv|lyiXpc< zIGbQ+K5wCdY_U<0a|G0!;MjCy6@Dp{HMqO_kz&|UW%5(t+*IfpWY1=i*VHNv{dGG~ zvtc>V!Y#QrVhIwNbqvYZ#$6MtrEGkt)M9*P;ZS4P>v5IRM%}2Sf(evVfA>aCIo;p`X)@B>JX*p6iGNJI?`I4KU0LBb>Q6%8#=kuB=S@ zhT`dBK}}`Ml}CriVfxV0lvSZf7`&w6&SK((Ar6FZ(so-Abn(RGOt}SLB~cFgPP3)Q z`C95zZiDKbbzwH-UTNF9c>KI-vYxLCPM||KpR1pKJ^XkROh;&*>4{>EQ}^HY1U<8Y ze=FfHv-`S#QvjtSeK)qt7WzNa4Yre29}d_z^jCBm`ad-F_i;U#KGpbI{NvfD#~wd? z*DhvN<22iulJ2e0f+K2>auyoWU0Y^p`pb1DNl>B$=Dd%?{V2cFR}CFQ}=&C7A?c1rHsa_pVQOJ3T@z$sLxqK zm&C31$vuZ2ZZ12Xp4(Vi=_n}3jl50_s$x8SEta<h^S#iXcJM&uo8fknVVG z;mzsG3CGb~B$KdV zP5!=r{_Gj5jtA#f-A}#V@_rR8a+rZ~Ht8a9JVC54Hv4X0t1mDgHf~4x7&D9^<-?LG!vf@BS@fE=2>T;>n0lS@99#&48SvF zYGL^xGyF-3j3pwWzY9|fi)AybePH}VAZvuvbw;~8(gx2(0^Y+6D+@`sJTwe0AlQzs z9(>9L6EFQq2DzYI2kI~A_OvSF2tFOSj+E_EvGjBQc~%~ZN(Obaef#l%pu}N}6Rt9h z0LdUZYM|)$qI@etcsPskL*=#*-8yD(HScjpC zf#|=_20js3qvAm1JG`PxoV>RrsJ}V2JpaFCBN8G*9mTORU>I6(sGf@XsRbB{^ftuk zV3AnG9gd!s_(viefpXmjTtM5bs9oqex1x@u)&=yQ`8=AehdMFGRj;#J$VGRjS z0-P0!@%&a!csMy(3|ga8YUKT+;#Tx@!m_=f3Cp9i;_D@t|BECQu(_~1i54XxGx(FwJ%BAtF|(wr?>f)Lj4LQ;{Ms+OA2AGtp& z?3AN|5GsTT^Mt{uW~g2Ce=1A3W>nsU!z1+}mGV$_x4SkZSlWmOjRdS%o>B^vt`8In z#i}MBji+;Bn%RwU9=CMnLsbi#CaFjNGX1Nmk?XNa~#;qEc z|3}~+*E9dUF#a5s1JqKAlm+L(gCbNT)H(>DH?-H$#)v$9@&?O(Hi(9`z=ngzq7d}? zk1G@`PxB;|E<6C74P}A*azORPyP_@S zNHDm0M{Nq0%iUN}zsOhxnOO!FRXYxfc*nd|p~JUAcuQtN;fDy&3pNAw>y8*GyMaEU zo}W*s(Z@UChP(DYb>P3x|I%g1$M08v!Oz; zD$Ssz?Wn)VbZH%~_sF!HSpZaQPY0eI$R6d?_209d&;XTpj@F?*5nT?tGHWZy*!=I= zi?#jWa1cIBJkTIkmNXP}f2pIh{`SocZUP(u;!e|6Gs>ta1EuDLn=QYIjM9)Sc;h6< z_4T>Y{6Lg(B9-9IXRnIhL=CU6-5uR#dKWMoc=wR{_kvNg$}-}qHuwhn^Rb8V#dUO z%k8Y`2Zw?1-*uys9^h%YC+R_6ouTXPBkJfm;p^utMad}pt~pG&hr)ctD+{d%A=U`+ zG&J~p+pfbA*Vc$2hU}nQ?#%XUe=*%qJc!6KoMds}2rE+uCE|xLPev!==RiN&))C-w zWTb$U%X2wBLc+h&Bv=O=p`hy;p2`$}Q(yzopZ&o1r*yA_8EPW`xAtn;u)2c#G0~(( z*_zLL#*zA+hpNLG4?+ybQ!}Jj0nm5_AOu4otboLgs>r z_r)M=a0VKWyL+bSBq%K=K)uUea{*p#4wlCjMn}UeF;ka+8#=Q-TCZh?id}`mStDM( z`va}YfZg#}4e=4S*0LWB%)k3MSP0k|bGv=JrPW_k0$E_<-BEcg0KjBA0s(}e_v^%A-L7t`(D=sV&_$o!U}epwp3as@Ec44b1Z z*nKP5(dIfANfQDbcF8hGhxLC3{C^vJ*a~cGXAP`usTRq5Yopep_2C~=O=0K?yiG(c z@ES6%1H(UEMZx>#1XYLT+RMYHWm-BDl^`?H8rWFz5i_NgA+Mg+*Kjexyv1Uxch3mr zxWdy)Kl4o&k4uwMTGG@$=s2gM-TfA63#LWT3gA9}Gw%!iz{sqf-ALKT6);F4GJ=D;&XS~=oYEM{VXMrz_6vY$gjqkj6|5in zYnRgt9SFe#8FXeOT)yM`jU$D{@}$PM0!Q^(j3f*tNRwc488y@@kd)UG$ZQ^ZKT{|x zEu?HGnj0CMF|p5_p~8hgVj4a5ydwYh*J`mtoXv!E2GIx9??c{Zc|7k*xa@W^!@V*( z(MuDgqYS2r9F;pS2T8~+01NLIOSJ`e`Frv>sla|l!y1?jpulqN264<9a3Hv5CHkoV zV4JBwF87TZa>HCXBGiGTj6TwBe7LAMq0eBd<9g0Il-)qrePb8JSIegm3kP$K5oWM% zTcfJfPxq30c+jq6kmcR6nnnGmA7;#`P zV7=YftMd;3j|(tC$^&jOA~tMlzybhvMXHQ~FsAXS@-XSHMdlRd&lbSuHo+oq!ZVD&KNtj{?)mc@X}}tY zislhM;=?|H%N-^o?rwn9!db6M;;*d2f zESxvMeu}HrID7q{TOX|pVAJ5*rDkQvhL2(y3tAsFU^@G7YtVo8C>nbPYkr+_Qo4l=%G&2}Rz=|P8`?j8TtLK@nt%B80VrpWGRJSD0MP!$g z+e=jTc7(QASj5ZVphRqL)V3khZ|a)J(5W(?LS8&CBWjLT!o9d6Oi~4wopSd9B0S#e z<#P9M4veDF%Tk(LRr@~muzL>RJnmePIl}%k%)d;4m+4tr_Rup%>^d_srd#Xu226F0 zF~`~*AnIuUx}8-S-t^KL=n&kI0=O8#K4(R^FC#m#SxU{&|fNLtA zs?aAuwMorLcC^cOjc%Y-8U4d_0~t+u@TIVqltv>l6Tc742Z#L?1tQh@e zGPla%ckQJd*{n}8PX=nLUPfTS0|Od*<$O^f&81ioa5dJFnehVRDlBs<^)N(T!1SVb zBi8)>fAE0yQQDMQ*D7NG;PSyArfDq2h2ZK()OB>al)6oIx#rNUQGHy0zOY$J$_u<3 zr63<*``3xSdTkE!p`%ync@eT*|gEG532? znei>La+1Ya^U!f9FG}z_%|&GQ0Er%$tG{;&6CUWSnK|!KlKh6Kw3! z<*ab#LG_0Lj?fj@NOko;&yBGJuo+{|hiAx!PP<}Szjm3pBSVefzA__zwO2uYHTPpa ztVTf*)n~Ai`k?3@;F^dp^_N%JJ z_NigmZf`&Il$;caOa_HtB4;^n*z%=iK+wY9a%+#z3N)Tg`FpsH0j(9;HQHa{Q4!|gOH)l zNA_L#rH|N}GG!Ab_}7P33lju{bfMg=HDHOf{^qn+RkRtC5=OaNZmRRcfMx{YA`Nn6 zo!FrB)|oDv^oQ_>+n)WM+e-NW5X3E*wVAg63POhB!8pl=2id1F`7q$~X&}8?LahO! zvEgDyZ}*gVrKmT9tk_?q5-@%2bq{IzrT*ZM zZ{1-25O>VX+xlb$(SD#4t6WDZ#us}6hHZc_aa=X35DIF2gRV7oK#sL}J&yukKD7^r z2X0COH9iJd$i;wc;48`M^ZpmR7BT@k;Kf`(vyO^-(%Xg2Xg~xH4Fic0<{{y->)2Se z`o1h5uPD`fXGpzEMHy$27NJ}DW736U>LkUp`zl6$y zA3t*Eud6jhReO_#Izuv!JT~8NNaR^j>ir{jo;y(6>Wic@iN=5jywv5s*{p%Ea{p4r z@_3J>W$%a9)}QCfwj(Xok;akGT1mAG?VAg;mEY!FWC<*H=_)KyX|Eu9mL1pnqovle zsj;7Khn|zxpkDvEpqY{_jQ=XU6oLXxAcKnQI`N?jHa$8x|M)Kd7fU?aDHF`xE7eEP zd;2r@nJF?fw@!UYw!+BwJ`Z&8ysk?_l0NuxlM?g%d) zutW8$bx?a;NQ;&=&(Z}cWY)m}keYD#xE{pqO1DbIcn>^aqOviAsJdXUxi9|(i5KM_ z^^Sdy`=ZgYtc8RNp2mY&hMJMAoUSH{&5?-{w1IWk#Zkv0q1tho)(Pr58{SKOVF#7n z0~3x5bFdZClgsT0A)5XueqK(Oe)PEq3vLnhCaro>rBQQ1Yrcc-mDP@8Lpp?*iHL3k z6YN!0+G@=Hp(Yvm^-8?Nr}Xu2RoUlk5?XWas%lSRrU{@+Gzj5HgEO^~Wh)4mib=2h z!8qU6IQvCm{6!JeSd_xF9#$N+z*u;ql9U)Bk)VN7WiivLk z_+sX4NH4J34HaSqAwsFQp9&XRzxF&+R2ue$2LZSI#7X@8j!!YLNk%aMV%eHJO7TKgU5XfxXwu_HO1nq{Q9HS#H;h(KTZY!Emq&;BcFY!-@~ z+P19HP)@1?sPILHZ!NFu%PCLem1$jKm`eGlw6NnoBkrn$$S}C~P;(GE^f4rx&usGe zKMu8AS`2U?92B|M+MYaGu4fRM>F0puvh8|S*AfOw!lQ&41l!>Jqy!?~b$``~-km}a z4hKf%_`U~16+$fr$#-B-8`CtAOa$_CJU&jYKO^@Zg&<6D%i(46>BHfEO;_z^tbX1_ za)fKVv9umWBy}lcdaD8RL|&UA96%-d?66t^-sS z0+n+V;}t4|P>vA%V~|{aIww9fej}=>)uX;#J+}S@OIL$|Oa>C6-V6fxnjE@;1|8C> z7Bf#k*4ERaAnD`OK_yf7G7vD&Z7Kz+AWPa*xt%Ww2M9?o+!1jUyz!-Rxi92!di`8J zrl-5sXvsf*hkUnDv^oRyp?lu5YX^}a6Ix%dl1TmjjHccM31*o)8;{u^yPWu@n1e)Z z&+;qf)josfl?9~e3knngl1Q2{@)UI`z=`9_@OiaP1;>Z@e8WFtZbDxXNtdqUR)vKR zmzuMLB0zeOvZ-Qt6u8xa|Iy$6|Dqzd0o%8qQSaAB0+_yF_P*<%0^pYG)mK3ReBj|! zU|)~L0*d|9!{D%1wVVATY(S^68K4b%5U{+PCKEZsuWghuX|>(Lc*BJ=S69u9^5ee5 z2Qekqs?_(!@B2emB~PT!6_BT?|q-~(G*Q$zR*m9WstL$U@wCO#iE z(7{>ecvPLT(GW$lt~UfZST}sF zP?}L43%46|@_F(u6`}x_)X$fWWtk%M4RW+Mi23XY@2Q^veIs~C42q?5#tESI+Xzx9 zT0dLv6ugWoMA+{WYF|o77S%0DFxCXDCnW6v?T2>!Cz&p(v1OK3`R?k65Nn7K_}qoBBeF9*v12#v)8*f=nV9)@8`zuI zshZfOm{O<2q$uEaA@Rrp^AMr0ZOd{t*?%;QsybZxC|itrX(@_ViE`#;;d>UgX!|lQ ze`SSz3fygmsnEZK7v8inIW~6D;7v%mJ|XK)xf46VT2ur&ceroD%D- zU*Ft&3#Z)me97OdeiH~BZ}DT-J!nYep&*xB-VO@0J|FsE&N0plwzX(!;pY*cq4(8GC`^Z6%YI$T>7lU4rfFPt>>t zmE{8ND?596VN`6+_oS$t19=8)ExY@Xy2RC{*6~d;2t5w3wkrNVWPMdwl^WBi)UZbT>$MH|+6W>tFlY*FKvw4!FN3?s#5n;dFKi zgRnv!azgfb0_sHUk^=Y0Baa}u+cnnk>K+GdfgrZ3W+`-*9XvZY{gH{h61@UhBgH`a z4eYM>(K=>NWsaKap*3`XA9&hbc@a7wYb;0{rt3MO7M7y*gzU#{Lzqcks_$~X&whTt z&BlGHhU;?(B|xwypStpf%(OKa(!Z$u?XES54F~_1@PrP)5bDo&$e^W@7TRedI4~X@ z2yQjNmPu@&68SqGx9LcXL&PvfX9+Mj;bjNyoJO(rDpM)OEUW)(GG*A-9dYac5qWo% zwrJVp){sJF!Is-{MV#`b0hxg|d;J%u8&^WSWO}|xNPE5191p+xjH}{fk_o&7hIeRvPg0qq8j4pJs!0j13b!a>V8t083Sz<|+c99$>UQo3z%d&I^FtncATA zdgT=o@IZR{vZXkT1z(z zn~AybB`qlJW=FMDBNO;~_uKf_&kC>Oc-P>E{3)azNs_D7jEwJ(CoO8K83NSMR{)cJ zA8&>%lgSMMXTghKbnO0RAE;1KUA)vL4u#MsMU^Y-=6_xBPx(3*HOoHR7fd}YoRbzo z|83DzUR%yk?k+^WaFNyUaa6IS`~GcN(N2L>XUI7_xl_-)4truUTdHUceS<;GARE)k zB$w9hus_n75?l2ib}5r(fvB!v<-InbTEOv_O4#Vrd}>AXIw7Pljkq|%WXVOH4h}^A zm9Sd+5fK7qxtm|iGl2T_AIqrFpDr;1`odSj4#}ifRyTLf;$h+S@TmRbicm3CIp3Xp zt?X=O@R`?7Zb5+fM1Q>hQzRthWR@#%D8W={_qun2 z@=~A(QXP8K2iZ7T&fT&n54#7Kc?Q4KC9SV>iAw!5$5cjG7Y*-zNU$K>#od>&fPXTZ zfQq1u66c;4)5!4S+NO2D=UL#|s_Jw&a_FBbJ8JtqDgmm* zUv0dClJTK|fl?@S>+ew+GeGDqAQe_FF0QFlC+Or?{7~k8INqi*AU&oNbnWSLjC6@FUzW}|1E~O zx;azc5nwa{j0UT$8N##n3&$7X84#Mp1gdlC;H!4?O};`g6CF{P5;U#yA5SvRgJ z&uj^K^>s(CGRc<+v0N59a~lc4IfM%6*#zVi8hPpA`8m#GZ*!{ShxYG*iJ`2B+OS^{ z9<@4xL!NrTU9Wq>_psroDq6!uNt=Pln5lOaaD@hZ;>EwN{Y<`f!KKEy9oU>6#ONQo z&q?4z@+6VBzVe4MgXPJJnx)162t>WZXUUA8Z1OH}dQgr}@s-frR_vtHV?Phdx4TW7 z7@X8EuG67P-4avHN9YW4DbDOq>&|$zRRQW_)0-%V;)jzA_ihE%n1(mmj4-Z9?S`vN!{TvhPGB{N&<-O-+2qk9 zy@YgJ9BRezu`=c7r|wG2gboWY^~6Do5X$a+E0f*_J~c2FEXyfooU{I;r#h&#>j8D< zkC@+PjT%RI>tq#lHpr7rj5XfAt!rXl$43^oz+9!}>+KdWP6^ zyXtK@oW!=6efMyF`nuT{S@U6x;{JH;<+JDP&#Z0!h2{TCPcCZz1pwbqTkqTAwV$;{ zxiErMK4A0*XmO7ZtJ{0aChXS;-CTHo9mh^^L|M-O@`fCa`H`6Vu*y?DuJoR(#<&7W zy6^B`4)3UUg&<+(3FjdNoN~!3>$qb|h#SJ(0oqT;46PaA{*EyvW`?oZ@SA#$C;ik$ z%`Z#j(t6t$73L%H?1;*M56*>(snA*G{k^6T`1h{I@&$Dh)fz??!UI(BG+FV^$P9!~ zptS}L%42Eg;!ee!;$#vDTi@pD@2O1e+kA-NUrU$7M2%yb~YtU6(3PuOY%hoN_)%ZxT#c2+uB2<*<3$$e(DV z0b@opRAdtFY}NJbk=R|ZhccjRrL5V>m_Yn(lv?jN~XA` zvapVE@ecO{17S-GXutwkm*P&PA73z2YO@cKNf%RRF(E$px(ejipWYmp*aW~ja%l`8 zSUoIhik!=SE1`|g6otX^q<1aYy`d}a4uz*|3us(GB{r!OW|9z9WwcRJrniG6nsB|1 zHi;lS=j->>ZZ&)d6m=E75q&(*w1Kp(E??ZsWcS%M_mtG86#vdNE1u)LPR< z8In>@RAlQf=8gnb{uoWi|9BlI2o*!h4<-%QX8h!R$HM>zKeL}%ANrM)??zrey1D68 zbi8-^JGlnJlBlfZKw#`_II z9EjI{(PLmeVV&moQu)gJ^MwdR$*7GKcE;Or&dZUJ*t;H#ttAgYm}b z&bhsu(Cwc4>7&;aC4l-$O&_L0x+{nq_tRO)E8Put~4H4eM#GFGBCu9xYy|y?Ht719caM~ z09BPye>XuoGKDj0I+q?MmqLX7sxvN)YrHWG6H6y#i7wXNX2^Fa=BOvbM*h{sq_3#M zg%l^uU;K+d&0hGx4lgEXe>%Cq@5WAXEb}{&&?ymXf|&~pawMI;nxg0+oCDK!4m&u1 zHgua>Xy1;Ba%Z{%+#hOI!8EgwM+fs22SE>H1Sm^`EumRXmsq)p9sETBuQfue+_;yo zES3hUW-D%Q-rkbcSb_zhvTZ2mD=&go;p~dxs4jG)YSFat37eEqPXO{av|nbTHIes%5;|-#PQ9q}go{cJG&Df%v1eekub^PibS@VQLEgJ7fa4;24cv$xw-Sk zhYWo5U)MpkP5;raJF8g6_eEtZuOlCz+sNTs<&`cJV&{J$raK9pTaNEG{thlVwSwSH z#6P+8ni}S6?_e%ewh&J9D(uht4sw8G4_kV=BsywMe+0mJ6n%Pnm=1&QCFxCI5&@s` zZn0t@W^^jAQ=*iTGokn9Rll9aevTfjW9!B`2B4W&gy>ZLqJbZ#seCJ;#~>Kty1ygM zmDLaOm>{)7VR7xBNDI;}jC4TqfVpPc*JD%8m)-QCW&9r92Q_AV%exih_$$17A8@!1TIkiF#_%{-(8_!@KUh$XO6Fe$ZQ9<59=p zc6F}VQHdGE{Wi%3Xbq#v<)=#j6#xl_x~c?6O!!yQOCOr3TLBZ9usR{q*=9sLDpF*-;{! zXbxA~dT!`R{~ar?I+jYq&?UW|`bKnUQ+%(oYvW8lq;*7Mz)4CHbWc;(0GWs8o>wsT z^pYo4I}45hI>|74pVhjG(d=+dZ<2~PKBIpR-ZsquCP{DUFmsuauwqRzjKg|NtLslI zkS4qKINIcf;z5W@U{vuhp-*=?)oIJwefQdVqZB7CoZA{;j@itm0hInMXq;i_chW$4 zb{*VKCgPR^ikx-(eewJGrlbrp$wT8hyDBsbNq4zl1i8dRl~ZlUK7o0>th11U8)-lH zQ4ywN5=&=FHv4dpuYA!Bj{W>{)2&dp(1ev$AXH{4e2R^7FqX zF8tpT$KeV_pM;L(Bw(dT)v1Q9u)j7+Y$F94|baTGw3874d9(-#ON|BikW8mbCpF|I+<|HZh42hp~>_HU8H9c}}_W}PvE3)|E1O|G@?W8~| zjqg&j*zz&UzCb%E^KxfDJ#-{Ft*Ywol(kzu7T86!_%+GU11&#Gkm-8)T65HhrEnU| z+;ad4s(o+l8aA(B1J1FD?`#&h(9rIWaHjnIEl}84Td|it|2ot8aSok-;m1kvMN@EU zXG8*$4piwch3s59_Q8MZQD>=23LBC1+yg5c>DV4IVLfCBI7)%ff(U>0MH0@WMO?41 ziM_6(g$AjpszNNMS-|J0p53@v06^ch?6b(tJeQ#NO<7Qx_i3B%Nz*O`g09Q+z)RTv zo-$^DUX<~6r9sI0vfpvt+`%psuQr~tId#zY1ZN6rKmW@QbxQmPsw|Y-e+Ry{?ET8^ zR-*5ZJF2dRkp!7LrUO`Ee@&J1zpvJV36C}?D)q9y3%1UK1M0L&A_Q(6Uc*c)Lc^tX zYOV4j#rq?Nerz0Yz;KW|3=j@Pp4-VdnBM+C*j0^VN4R(LB6h5d*k%r>w*I4qc|1&} z!O&SvvV7xcfRUu;5oN^Z2!iGKg`37jphM9KMDo~KhOtUM9>e#cV4vH0J3H&MUpCTZ zQx;O=VCc}pV9~b>Ws*`hN#nl-LNLC0N^|cE7jJe_BE4)#C<>!(BEaGoV(#^lQPE`N zdDAyPIy*!Jdr|0n{(57DaiowO2<#=&$H_;1V^N;Kf@FekW@6IStVTR*^c^mXa*#GT1uYd)oY(Y-lZzD_OVebe zKzaoz$b=+0h>RhQdTOR@>?u|CBwh8HRhaX!IZ@sw2#N0Yu5{{vX?+g9dC-(aY6|jj zX{02|dpLf_@?Ztm&z!Jt{}ftvILM|fGg~&ovBny)BtA)|s-1pE9O~EXJBTuuy#v8A z1>f9Fx8bNP28ZBkwVFEWuB)q2S;_c~N!rFI46w=~@@Md4-X$jC;!Kqh2)Tz)%Q|1SV%aBd% zAuYJziX+meq34QARxf3i0`3ltN<(uN#Y!Hj9TTx;>u+YIOe2Bl6g52H+F~0l?LL%Y z03Wx9Gj%0bxVqS}$HaY2QBT2jKnb=}QvVZkiID(J^ptVp{2y$^BzD+O)M~upPLB-_ zvjyYlmwEXsLb~w#xbW1i--FHc-W}4{Z34HVkpfCKS=mKGxuas%f5cLSgJd#MK6#YE zDYAdYf&dAj*NOQf0HC)!LFa64bV7e#hyrAH8dVgX&{n_C^U~0^Fvy(6`_m#Je|e)b z^jHiRvw>yLRd|9{x1Njo5mJV~7^sbfYlK--#>)742v|*424N`p)*%(qWKfVHNCmoE zkbT1jwlbZNt&%-t>!vTy`LWeO5P(~=E!JA>Ow)O>J{1UTjwx`>a;DyIcyvkI6>%5f z@@0QqYyEcS`QUk3eY1r3ahICs0BhxVx)3xh?>HVd7jfJ1`h4-&@%o4wi~Q~Gb|je@ zIV|fkgEArAzMc5IupggcqB}w&;_)ur7x@vd!ud?7WRA2MH6yKi$*i!(u+c#hxpa@) z!RaC|L(o}&4%=yhBywq?=*;7WOC|CwqPkq<7tLGqRJW65X6Pz3j0N)9P;xg^tWinuFWX%!7p(D4Z<^?Ct)- zxyV;gOt)gRC+{PUWAVrLE--fIN&uN$E+uqvM|dELEpvY$$=Hz$8RT<(OzcOM;_WwI z#|lHVe5#xC_6bX?eoK8S*8D8TR?x~K*^El3LNa<1BCFl5NVLiJ)T4X|ufq$d;06#* zY1(oG@FpeXdkM7;7;30Tr-nMVkPQjNFJhY1ky^*Dd+OX8#v^4rqm1BvnZP&M$15wd z!u!HiK2CcRztwy}J3b=;(w&Y{p@G^Tr|`%L3r=$&GIRbOxS4jk3n-kUM2C69QkqUkalR zvO4=AoeEd9FPye!IO z{Ox^X6rHJ2BH3t=JoekX)y5YGV@gm*zE`IzjA(Gw8%bLY znRAOfmG)x2R^d4YN;7zABhZ&_9iuNMl<+217(LWEs8&si+13U+BkcXA)z|S<-`-L_ zf6^ClvyNBxyN9vBD(NPjbMo=mOrRS zH8ZGh=6z~rX2RA$Z0R+A?m}NA`zn4Q4mH7T;jrNkSqE8V*57T3gp&B;-nS=-;o&Gy zg(3ZArzSgP_8aY5e9LJ&>Aq1wosMd3$9X^L--*|cKmC7JD$*bS@Qmyqr%&E{8Lbr; ztetpbZ$ry5?`Pde)ytn-{C{Il4{P;4VrBdg2k7ff=^f0e(iz3?8y{J8mC4Sxa#B>u z#w-*f>B=(qc4DiG_Sq3yj!Sm&YB<*;lQiQVj;Yw?tcI^+mTP?byNmI z_8KPimFNA{LjxOtlK5VzZiZwbS}efdZIOJeQ`=dwpuV9{(*R0zbltRlS90y#o_GB< zM9azw?Ag#xv3?Mk#S_Di;ZVI!iZr-yuXH?}sVRJJsW>9~r5VtlLs1KFHo_p(m%o6+ z)-OZC4(w(}?cr)rW?Y)P&{bQu2G__{+X$C!{8TP_Pj?__(hi?O)RLudN2oA&P#S?t;Zq;XxPQ>@?7lP#-ro-*e2utSE6C$&{5Ht9_eJFlPQ)&BnxINNgtF z%uu#rXicS)2+MCX>9pxJOYs^PCjQ}xo9#NjDqh)C^;UmXGj?}fzes0yRX?`w`HOPk zwTOk>b*{IlD@udgG@>ts-<_%FuJ9Uj!|(YQ{Wc;zZ@7S0o_PAkA+rX>j8UJXa&CRFhcPq%C~NI zP_k0m1z4HL_Bv2!JXH~V=@$@6>W;iH-zTq9S!AKz|LpJ_M!403mEiT}wvUL!Xdm>! zvfTw{zuSnCwHShSauuf^$Cs^$l7tMYr&&2a0t3=#BojB;*l@u{11TQ@*w@C#C^a??rB#)#=1W z>BC$eqz}z!vZmm!wTHUXQ;VF0ZK1qqRXdKOwB|4Y>-}nYYqvz@Dfa>?chdSiW zr3XLZPK1iPm(U0GVQ<~fiu8T(zG$z--O*;pA6(^g_OStb@(x9C zQsGhYipIpAJ{T4zGIZ1X;jx+;Q>ab%4zZYYebGa)XB#nfaH^-;HdEYEq?gRSnpf|B zI%dqyda=^i)5Fk)A*NAANMQu$jRqca~1576incoy;2=2sux}JE3~yLuk(L^rGa> zzZ-8`ZT6tgf4e8mbhUchu50u%Ar&4M5x2VxeG%R%ZH7jCi+GLE-e-$+29#YX&9T#7 znRkspEcsE@!F8j%pc$c{R1&s~!|y!O##B~ddxPLn?__vwIX3(L-t+i^z2zvt&~3re zrn>K;W^5}A*B{pfTj#mY_$f4~89Y5Y`M(UkKJ#BzwDfID6T8wa>q?pK^pyUoCwAZ8 z`4(IWg`@3&z1UwkR=z_41Bu3fIz=#^XBqFm7(~v>@G**)(x*~f*D82gEJYf6!dzL} ze1c!f$rvuZ55|&=eRc}^2v!IPHtHQPXF($khAb=&W~gb51Vs73NT=9tqm7SKqcu~=9Nh%pDNRl9^}$I-2>pdef1>f)?_pJ4UB`w}X z5?KP27U+4g+)9W26Wx^Bw2qfX>oJ=;k@3}@c~D1jgJ$_F*Dsnr1feNCkct|L4c?s3!B96E&>_qiIY9sIM-sJ|CzuxN0O#{mvby@pK z^Wk*RfjQ-4ipa$;TASRfqYvT+;D^a%s)}iqrQ_9dxOt8V*hs85tc>L)F_hvkEG6pUU&Q0g<$6Pp4##A5-$K8>+bOat;=ZoFEU%Do1v$wQa+Z>j?f*8wSb5-$ zF3VI}1A--#%(v=8G?Rzg($B#%XeQ z+c62G|D}2<2be{t5~(QmEos8VP;=WHDl1ZQ`ycSs@PYv*T}ucK?t}xObh8R9dU-7! zM6CW~XQm%UG5xnyjT;cjq5C~``mcx~PHek;bR$uN49c#;e zc-rNlM;YDrP<-|shN1pp$MzLJ3OgTBZKcGsl#AWw2U#UIkqj|2T?IZ(02^csB_QqiS zj%xeqc=crGk&%)%R_E81;O}Jd`7y4nndjTJ$`jb4qN44+iG0+R`MJ8Z6wGpn-_wA< zA^7!Su9Tqs=I8%xOcD+FdskU`OWW3}9Kc)F(d|V!3KDAofaaT`jHLEahp;F|p3B;@ zwSg-{w&;H8JpC`fO1hS`dSH!7WnJT-y(K%d9{4rRX7JCKIl6=qvoh=?ReK!aELjXf#GK-lajn7Q>{xzKY| zXARBuiYk{llqrf)tLv<4?Fe^b2Sv)td^CsMz^4%swon^6K4;nh)GlY$QdXq>H?5&? z88_4e7m64q8yZU);_MUB#_AO>76iGa7k1bTT?kr%9BIc43r2R;K!BDZ#p&>v34Pwd3uS zgJ)8GM}oZ5g(ePyrV@~G88*V3OEh{t=u~*mP;HOjiNdGFB|&47w;nH(9HnpaXCQbF z{KnqGr0K_5^sGUED9kj3Mn@MPpOwaK`fimCrv*_0%TXf(25-vEy859Ovs9HvY~fD^ z-JA5#b&L{4e0wh{#nBFeF=fZsTZQ96aW?t@;0{RX@)-Ni`(R#z8ynW%E>zGBPNc~0 ze9S`^0(WA8EJ)52f5p_XO&p2i+w{0yJ<^Ytx4zmG&BC8wnb053=))V@ugfn2v3w** zon2hCvmTdshur;d-G?b(&nbsfc|N=qn=Wry1y=fOv+$na#)dEUwnLgaXc{-01@($o zUFVE0y0HAzp?=C-^>)xb3SR-)z$d3vaPZEi;SRF_@O^KUSxx%}?{xx%jI?6uPP zA>4~R{jl}DpZqpd$P_((+Sis8US%=K8DgpCa*i3{x|u`|m_%nSvBDOH3~>^_OCUY@ zlS12yM@=x3bR;GL^uSXZU&+p17BI%wQFfse2&>&0+U1fFh!i21`ooi>&$Y4b)qFMF z<7+~Y1H@2}y~7LZ!_xwStkf%bq=42&l;&vuimR|9{4@!U@CuQ%E;XoR8nKwhT{Lte;>*Xa+5~MJa9ySy{X!b4B+`^A; zY|y8b3#6zJ$GeGCo^%lnQ{|iF%w!k%U%y)H#wa-Ms49YDy70-db@UJB<;^wDiO7ay zQViuC@w~N;krz;K4*>D&7|0%Q)4q+A_t6GpiOoEyv9Atcy6plS*HW@uqFkYNDy^I_ zNQ$j`xE}1^5Aa5yzW>UvXy!PUq=9uyDh!bq6E!8k5|ZJJc#;E6<#ieuB0^~+l-*jV za@4*^O9*>-@DjayAKy1NfXN{DWRO8;q=z@0WAUc51M#ni`o>KYHp5pK?PK5OX?w4x zkuKI$0&_=oj_bNe^IzM=#=^KUn&F~Rn;8|a+yHIQM&|-swubUd?J#>|Z<7P#boq_k zmZ_W5{v5R;wzKxzHno=<`&S(buRZGMVHYS1Kcx0NtyZ$s!ryT}`d36hORG5ZdnoeP zr^c!@(8zc~qok$vP{y<`6}_1^x52-utMq-h8wHYm(^mo_M~2*8{QBz$8ctumVsLdQ z>hsc1{jYBWdR*Wi<@1}}@=uwzn6}xf?0xwSXrq2ux}Pcua%X3{8oL0=c6sZ<&@i zHA>%=eISQRnQ~Zgx49A4uQ+=tqhy@rLb5Fn{}$|k(1!bn zeK`&S_Paso#wBsJC(Mcvf9lupGzKT^U&j01iCegDqXkK51NdYcvx7tNmgoA+ED;2F z<8YtLGdq5|OroK-NVYLSt=T#t^iWz8jg(M9O?AD(tQ$b+L0=8tk2mnu>R()S;9A3$ zQYGS$lNJ%)D)&mpxTbuap`$Nci26RpOUEN#lqtTUhC`ri{V;-1%d!EBLJT+|(HnDu z)8618?(t`N*J|>E^Gq;1Qqfe~Vrip1ektN<$EpBvm{VWADc>_(Pt37A5VGAYV;L>V zdEG$&@I1a^{pfB!UX(x};PmI~Q@{Pw7oS6ewMjw-OiHn`IGeS0Uy{rHl~cclAc#?m zTNnB2<+lIJ*sfCU8;l13(Q7xZ8}5UL*y}um88|GF90F5DwXErRt=8E|w)GftVGF3q zFfF9xEvJ-j`PtcIpur(Q#3_MRS_bXbx{-(rHqws26gFn&ab{g=uH zT_?^#crq0sYCoKYczbWQWElhn0O4xp4T?Um<$u|&@BKu9qaI$4Iook3Ggw1D4`LYu; z-vNdwIg-3seuy-&Y~0J?XB44jL6fF-gKYfq^bp1sz6Szzh}Y-`rPY%4cH%~S_t~vQ zT(T|BzR0!0AcsXV(U|6fnf3g;(BBWm?;s#GPt%7zoxAy=Tj?oRNJJKzWYz?HqSs@c z(X3h?`ffR#Ro2OgIxvNU!88o=KhGJ*IsBLdq*Lf6I1Vg%#h4s}t`bWfq`_i`p94!u zm_3)4RvEG%riSc6Y-ce@WsTp?+|-(c`q<2}2Ck zagNvS`uW%6@x^{&C&yx8dx-|r;X{3FL*U27IY{XHFxlE3+C0Eg^6>~(dFVbEdUvD~ zZ?6)z{d8r-KQ>BsgphgmqIK@CO~~LU=`E{9rFkW9l+?p>tNlssi`$C}B!~R;pG@9- zv7j8mp=@K7FmTEtT(@kL%{7t`uO}Fa2r_*x-e6!%X5t3w>GF{@I|Du#xGVYs>%X~9 z>}LV*~t|rO)C-mx?`aie%B1>dvljM!G3}WW% z#`x70sEO8k;@ChVcnOxn`cR|b8ZN#k+yE4isG);Lxc#8a|VTxKob%9Wj__5Bc3gD>7mEm8fo=Ud+2!|glzX` zc|vEK=aAq07`jT;Q?ZKoQ?I)9E3es_l!^*Aj6x@+^*x2;BbwO!;T30_yb_u48yM@F zZ-y1UCH$DnF1{irlW3BFVeU#gk0%W^A^_Uaz_T|*>KS$jozBm>G z6|M1Q+?=C;Kz2c`r`k**Ijr zo6rt0`9`V#Fi}S^FA1p9^KpH5Mvzsv^3LtNx9v z*|rm9&lj5KS&mLln!mF|OZ`hl8Wv_81XGAndxscZkwchXQB z;U)CkY=5FnDC{o;8jrb!2A(gJJ}tNYSPwlXZ?#(U(^A?B^H$R{$u1Ck%w!2Q__JOIO|e;o zK-5?kb0_4pVCz}R5)xsMrz=`9G0m2cREV|>LTYH?JYJ3_HajM<@(}sa&1M6~l82ix zYuiokr=<;5&is9jlRn6sgG6k9TfXrF0lz7ms8E#Cg-}a6MN4ReKynzF-ZTd&6+L09 zL4yVnDob|z3T^ZHk~1dOGOv&dev4SWo^%w&@N6-LZMA3Ph0E(P1=@sU)Q_63WeiQj zm)te}hGy@_30-!sEfixz-RE**tw@gCTz_m&@;aQ)0dma9AVV&aHuSx@W|Sj`7~bzr zC`Mer{1GA8^m;dTJbS~xWqN)b);-T2zIG-;?0XSOpS3_UGbwAG;iJdyvp=la{wPi- z8Xh)c24k-V_>}7~6Ecfhfee#wV!rKHp3&AUMVBb$hXLP3f=~?E0cyU(_UURR7DMnQ zPuDWd#L(s~#mQW|>BP&d-y48#*T-wolG)d_0HaN=RXn6i!8i8>jG~HJ*JU3DyMmEi z9nt}Nf43N}aW_qI4*&b%ZAQOMMwd8E`}ny8>#%_{>Ojg{L_Z^|4xe}<5#re7qm{pYzk;gVaJ`U=c)HvW!j8bSWYa5&(p}*(?~=W1RA`sn=vhPX?7gLxfZI>z8!6B zr~Iqle%w_nLR~!n=ki5(bUH2*$i($6S|r5z600=yja*L>K2^*kdVk>}YAGcA+wev` zN7x66s6qs-a&X`_&hiF+>@Z!9F2-#TS_1ScHuZ^1gnkU zS))=LM4~n|X`Ac}^YvjGP@?2uEr7-E&*L!4AM2s(!yI!x*+TD^t5Rl(%JKXGJ2?Vi zNB0x1!x?oZh}ETk3;9ZEhFDKCGyZ*nuyN*`V{Z(iSWjw5a&QfCF={0D-S9t`vwSho zwLR%MixRbrydO!LmE!PN=S&p8t1{5R%=y=;J>`?N-cD)${*=hg4j&D#V#l)=H^9T~ z!JbEgHptSAQq^M(Z)fj%tEA+$ng60de1 z$Q8Q3w@FfXX`F!u_scd!3}5Lh{IjA&&-(zM4NSH}BFQO!uuKJfknaFa89&rcpSEam z^L&6QdczUvf?`Rx<7A0~Un~2CYg)XY2>h$^TZO`lgpN%UNdO@SL^YNmY#%tgBr&6O zaOK|y9k4qvy6NZl{+*3B9x5$U&-Sri*0Ef265+1NLVQ*)vl#i89fDy2H;OK)w>J$P zMo4o4Rd+SXZ7Ko^gW&XL4X6kI zlc3LC)lGHzkoV4rwPzoTA4PDDJE{?>Q`~owr^U+WH0V{n(jWhj7UtqN=5V5C%aD`c z#u)X)BvzMGkKIi7eX3jE+m#~qgYHl?efn;?tf^1J?(OS~k(XPVxMS8Q1J1vC41(D7 z4Y9%4hS)n|$JT0zr4`LeY#(M1h7L77-JKtgj-pymTsvqe?>dq2veeODo^iyg*rVic zTX#hX{W(gTB=)Vso6`x{s9vF;h(fs}3LXj(3Pex5e!yV$g1~c789ce7k`BE&OVepZ zsmP!A(n{S)Q}&Q(e2?!rq?I-4GHvqor>WcTnzzz38WxzvGo5&N&+i6x?3NTVKh9jM zPVYzC`{esT0|>LV&uH)=V^6|QtEbQ4HnGoZm}(esLg#50=^gj!-7n#D3|QR@9pkWf z_Afj3(|zoZr8_0iPi^O_`MKr44URX`%Zt=^ZMlTv{}}}D?#TjvFD#q_4pd4D$7^ul zW7Ev5L>9%I^ezU=zZ%rtb(@lp1-sBH5PI+KPzJI{+@kMIb?uReNLpI3{#1?~qmzI7 zL2&u^3dl28KfV$630WH~`(rU)q!0S+d)W%R#jy|Lp@NfE7tl+UIhG;u(2QJe|6Uan zA|wfUPXd%d&X$92t}wI7E252~O`dST_^Np0zcr;9c#XXViWyM$h?5R|v;PPTAaz5dAsu?V#-6AGn z!?s%EIo@O634BTpm&Az`oQ@ZS_?CNrMOb?!PCi2C&kFhn9@2Syo>Hi{{-7PPS4Bm! zn>f0iq7B5Vn0cdOq$#R~=HBnn>$bF6AW@^F6I?8_F>#C-&cY@BA7(oB`MTSMz;3sCF3~!kEtvkJMZY+1WOLb17K@xcqx#TOIncDtfRU`=vBdH&*$L2m4be)>A;x(qTo5 z-RQnuZ;X5l*|)Xll{HW_G}usR`JufGj@O)g_3R7(uVGVa!LL)hVz+{Iz4t@9?KO0M zWJ^ykkEhATFWbgR#;?~6>tI($&itk8r#b&;-D7d+Plv?_0Z>yHF0FjpkB0^-Czz{`2Cj2g{c7_|hckXbdI7|7F)7{5KT3 zzjoKmb1)UyDtFN=8EFL8>bc$$?7psL*ovfdAzl!1g=W4pzzzQHTUqo6u1ED&!Evi{ ztI&98Sf+R;@Z*3tR891x=W|mP*fLzqj4PRBPURSDg!L%yF(v|R5$Umq3TUOaeZ{CF zOj0mdMtt6R7HqZsJ~jlsUd)Tp+&@Kk)+cCOPx&r(_;)o6zDr#MTOti`BgjPZQxC8MSdE{0ZKF-TYGV^ zT6ioji0rRVUI1-Y${$QhKr$38Q8;G21E#6bR``DgzAvT`)KYvyk*KX+rWou#&U+VJ z4o$9i*>yPVV`zs?t8<33$8x0-rH4P8W6X*}?{{g*Z#FWpc5jew7D( z&iVOyeN;K+^2a2I*=|(HOCBtDyF<@~dF%1j7~ud;0`zU@)BA(V07*+9NC5A=jI;iW z`4ywRVxc;!?bTc&nnU5dVod2`gf^aa3~1Pc*@Z8K7Jh-_#HKpUgK+By4@yL5 zh1~2GW_@UL*pRX$wITWP5Q2q@`9aCD6+v_U*5zBepNteg*1E{Ps3_yMUw9~ip`^P+=|(`L z29OpQU}os<5~Mp+L`kK)bLf^13F&SGB&ECSyM3SierrE#pRg7OaCBes|J6mu3c2(< z40ycV5I^*a`XFN|W!>%4Wz^JZ%-&4Hjlb}{H<;ys`g77Ei`eyY>TT3W$}M3mq1@f# z$bIUf@9EGVncKR%AxD2rAjIZjf3ZPxfR-EA1np7Pn(!=jON^!2h9O?O&~#7i1Ie_lNImN#lp4OheOX^ z(a8<^-KVW&$HFFj2AZw6zYpW%7(D)YGt2CR4|K8(dkyUgK4k&FwEnx44k+3c_?0BuPN-6kXIYfK7M_ACTgH zCR{a-(V2_J3IYAnhhM&a_+VFbl4ohh}t^fy|mVOlLzO!y~#x_r&( zipbB?7E1ObPieLE!$rBHuCt%?0l1eNUA{@_^>-`h(Z^r^l6^bi+a{~Xn8<^W4ycmb zg~QLo+OEQGJSIm@>QaPDx2+>q0rO$gQ<(VKsLYLr_kzPC{Y~Wt_tgB5UEaNoy6lWw zDTH){;IK222Xy!E&8U_&r(~gj=uLoJI?EU4`XOkpTS8txmK~ZIdJGm#@-$!^{WE}i zC%Unm?x>uvJiy(!@VumPHI(CZpY6}cjq{km#G4~ub4jA~Sz-0w`j-9r_bt~Qq>Pq8 zjs6kysa%Zki7%X~ByYr1N~V4fESzNjy>J3)CxY5cDns-MIKgIMi@zHRhdcMBEluZ7 zN|gHWyNc$c7s3)%uU~vC2-Sv%Z@9k0KMoGtN!w z&xG>1KTWH?aMt5jT{~muIYr^~{h(jWuB?)yoI$cp$1hCjXb?84p#4xGS7VCQ!mz}WExwy4NjK=>_hl<(7RWpGyj+g(;~wMn-BTKnN(SO%h^%>Ms0*nE-R4t@6w4V1MAUBmCa-vD(me&9 zK3wh7)U2Wj7Q8t!cF1HeSFZ!3~P0*@2dY{lhlbK@Q`*JwztMzQV#e;#0b^53{L!s}vrP)<7f5pxwV&2>FdU z#dz%hR!{$Uxfx{(?IG z1_Hsz!~TL{pQ%7h2t}7=3w;LO$;f9Ybj`(9ION3H<6aoIWPUvKFg!STo?_0Dlw{6) zdvGw|bGlTz)#7pJx7D;PB1j8HTcR^+dG9! zO??p?5V5tKH=E}_$y&xnuv<8mt`d?OhVN(evo_qc&Q&{K#ERBd^Zu8vve0utNZ>Y_ zxb&fH>zmBlbLG}pki6}(b_+%B?#mWL`X?-PGJ zzg0u!*5mNJb~#M3p4xo-Ajv$b4$qPlK{wNwWw^KrGgVNc#ahcPY_&3nAdnp?D* zpUb*gm+Yr;p~q6pD=ICmy}9wG<}>{?H8tfek(MBXIBrcrS8cN>eP^|r5yu>4pd~u( zZm{;F3RXOO8!(;z$f4ZQXyyIeZDt-&;M?%DUS&tniO)~#If{z{6RWhP<+t;hCU}XZ zSk3cKF!>(9M%0+|+Z?t%wx;=ATsl+_3DwGmk-3f3Ew{;JXliL~*}zkWQm^C1FBe@d z_Nqs`=XhGqqKA?N2dJp1U^9Gm{p=hIH@oGIfIp?dp>aFEZd{4uylF9=DAo6*&m91F z7Zx1b9DMdFL(-f&v4x!16$Y$*%}c~~ORDz(uS$Hpa z(Y~G4gYWI^irYA-9h)m@Jy(J*fVKJ;t49Ue$nIt9rT8si$#EcNQQhNT3Tb4(S=#97 zky{9Oyx-4_Q3WTk`#+8ZJWOymtP}YhbznyL(5K^Uik&(1(t4!&ofj9_JZKoz)h)VC zs9eOA`*v?Bo_!a`*pXdUGLn_7`tzCk;u`lNwilYJ_rodc{=7V(c0x_^cgFJ3QpMK% zdSA@`Kg!c@M>B(6@%-J7KTp~+n(Y=xc*(wt`0*YdvlXP8(%;VZDrp!t@SCJIU(DO> zf573^-f7v=S-G{Ff@_`c?%jgBhuLoWn@DDq_^gV~gyN-Dr_$v`wuZX`~iNJJ{dDsOKs^?1)u`d^s)xoUp`>kG082>X}*sY+t*} z-pP<$pK!)IhMi6CX5Bw5MP7Y@$k;%n4hAaRb$m$}3_&^&L*r^D6JtS13T%y@=-bNv zsL{60J4NJCLTTTx@gR96ujgYQpP{JYOwHAU|0*U$P#9T#We!AQjc(U|$qPh&M2AHj zf?l@MmU~q3NzGXnz2!FhOe{Js zLs&uDT=5!yp!#3RSE;B@#?o#v%`wxb}2stFv`W|(ZjT$v; zOC@FQ$!0y#KH4>UoDX^yYYS}kH5_x$9&+Mel5yEaN($Qz5YQN>&sv1=K0uXp(Rr;C z(zgA%CuW$QEYfu7k7cU^G{f_N(ej<%yN0~j5yBtvx3to_^Y_P50gFL=ph}?mOyEo2 zIbrTp&xM1E)h_^|WezYG8DSTC4t2B`xX*RBKHeWTrAkX2x4IHtEk8b3ZDjXDLK@P$<^IoEGz)H+dg=8Dez;^WVF?Kgs@>fCY|T#KffoY zrWaWu+Q_pti-*Z?TeraSR5gXr8;X3r!+Hw{N=-m|A1T2oNNGPZPVE)rDZv^2wG*5z z4lg4v2>cr(!xdu)jcG5PPsydtAK$zKOzOf_gV(ov)%yUt;QGA+%HFKmtj5!JS9LkT z`ONseY~E+lgC9S8pY-s8R-XQ?-W*Qby?nS@i>mj#m}Al3aWmRW&7KKYcFiJjiC1zo z`}zicJ{N1Zq+epZx>cZ!-Lw4>9k?r|VEXrh7FA=udI;5asp=Aggv;J%c?*R~qUjU< z_B|E4P=up)AJqhLsl?QJBhRx+5;2y{d>tDr;Wk?wJu8lpI#%)POwox{)IclypP{JV zVkPDZ57pbb4JL(*Guhd@s~K6yB4!Gx>|^__Nh;!1BdTKCbIaNjvvKt8`Zz1DQ47gx zS~xNtvMew(zW0thM1hM`rgb#;wNVsf(3>M_=ikU|3ENEm3Wu~^=vUWTkT-IIT9XA% zUxbC9?0xEM?3;2yWW`bghu=~|29|+bSZ!-cF;GSa-9VSt%dah;I;q- z+!y8^iQ*+v_;aE!ca&rr-#o?vAt5 z0153m{zqZSeph3`&5eXkrs60aPy!Y2`^XYJP0Fd^@P2OJZG9o?`^xdC~ zT;Aw2Iw;Wp>L>l{S6IYFqFC!|HR-hJFDNuK0Jjq@)~aVkXHJxaZGw@1!$)MGSa>}r zxQ}3xK!|~X%6Li&(?%6=7Bfpu_|uSsVd^aDj-6zB_9k1YE|Jw*H|mq$-joVuf-^~v zFcGAFHf%wY&WnM-Q48yznrh+}vg71=;2IN_fv+YMaUgTAy{k0()ozfN_v&CS2(LMQWJTMU1 z+>BDFfmHLW>XVx+!VQ9*_kBm^xJEaUp*jp9{Q?!kV7*?YhgcP;xzXp$&o%Tg;XE!SxZM1EC6feCuD|1JHk0y0g;-z0!2aZg#E$DnV9 z6B4t&a8h`ijpf?UQM15!q!mITEw}@VYf8aEf4=inyr@Ms{}@!zmAw;~n0tzyG~i;) zo>m(RsiZ17D8wKBwG;1gg$??w*bIW1-0*yNydk(!L<2kNA;7FXkq|7+9;#`6{snnH zb~Ocr%6NeGfF93WcJD#uv=x0--AsBdpUZM}DoKy(6V+OrLM7zeq{Z>}n3^XjFqLT` z+vwUTqk$@7;i|~WV(_SKjr&H*cDuf*#y;^1oJsRzAS@p~&u(;w85faDflexaB0U$L zogl9CJ@dKCH*hAqql-H!C&xz62fPgi8xx8*+f(^>C<0jAOFZ$AyN()d68L-Q&i~HE z7VPv(YPqLieuNtu$oREmdCP?^zxt5!pbtUu#!hj$`vwD8T3fDZzpfbZwK$4s0ofLW z{ocW$Ux&4#fVbA;{(QzqQYJt~^5Zjt7$06UUbfn`NYRV>ME~vr)kH5DFD$et)%uB$ zk8q3$4s>mJBD84-S=@DV6UA}DCwY}ae_!~FG(T%;s;;-@S?Cf#1Pv!?m<)pVSNu_s z6(;WCR{R;YPgn7#=`2?X{EyI@9v^OfN_foJj5(RHQN9A3`Emy`IR&X7zug9`uk=mTFeK1KNO@$J@ zLm>2yci(H)B;=Z`4BkmcoE+E;h#*G%2o)|{!@3#5ITAzl>#EG#)IJ#T9 zzt1w>Vn&xiCjyYfpr~k$4DTRgVvG+eL@P|y8-i5_=w`~~AK4?^>nzCnCjZ!5eG?p* zWVB})_Q4{Q`S@DE@hP!AOxzNOB{WM=ga4a90j)(8ijD@)nGYrAAdOIq1+<_f>6qiUkTMhCB=6 z(!oS_6*^~vU3_=yr`Q?;ikh8I3AR&$#bp^_s>ZnP2k1%`WJNLL3US}nkZ%xPwokvD zXoP<9%#m}vW?r&Ri3V~%0z5=Pj_9ToU)K=?ck?c zE~Y2&-XBw?BI1Y|gMO6f%HKFXME6OSGHbEzKS@yv%(9K&35?4dKTnw=18K7Q>$rhe zx@;X0?|B0^Yc!Z*zRto}?;6n%xDQ<{275W+EON3{g8a^Pm|~^U^_aolJx}m>0m~de zAExUK;1b&nm00pfB&NUqC7=3zocM27)5wl{Ap0sYCNn615G#yjT%2@vpXOdsMn-op z+GDS>m(Ha<ii)~>fQJP`Py>UU}#EAs$H zf}T7CAKT%*E*1ji8+dP9;FrCByL(CceUjv~4L-kc`z};vWf5<9$3i)>w7Yoe#by2B=UJ zGE{Rd=R&zmuvhEpu;|h*$x`KXY`O%JT6c*~*1|3~+5Fe8WY;Pv zL%GSOMN!DRD-`wJdRvPCE6#RzZbM8%6NV}K5UKIhBKKpzK7;G#hCh%{Y>P}`xR`i# zZDzZ3Mj9N2oDs)c^Gr}98cDg!vHun7yR$CLT-v>Xhm>!b)wMw}O1z2b94hnBlsc;FuFP(iaSJ%Z%*8#m>603$g2R z?-5EwX&NZKAfGV}{)H0mH&8ua)&qgp{;piw1=d5aDwM0fB9ABjieT##v$=L<4O7@nXY;%k4IFG}(++W{KKhn$yKQxFfOfj3L&D8l^rAp}JdGOgQfZj$ftVSNW2MW|2= z6*AWT$D*Y%N!xN1tk2{agvDGi)=sA#Ytr+J?SO?3^7J)bVUM+uA|ngG zNHyZq#O8HbN^IOTh3b+Osr8iit>?azlCfBl*-2U>jAKM7URZ-GJNEjwGgUw8Y48tY zkqm?)QpW>-1wMgxV+!<(krZR!8F{^cyU=E~zjF+6$8A!YG-k`0n?M~j_>P~aXFzMi z^ev3JC(8=Ei~gE`nQchewgh&YPg^>$y>Lw;+ks-IGmK-|&p$2qas8rGjhm4^8PliqNb{{(sb?{B)NHy+6q*aMV=ic05Ls4fK z_f2J+=(XFLpY-*x$dagU3ZC3w9w_M!dhzkU5%>qkIxh|ZF5q})6af$~fVnT~)hgCc zE%m(5K&5@_1s2k+d-hfoAm?#G_6=vM@j+pas-)6et^4fL5H$*L>Zjj{A8Q~mO!ZMwVbzP{1EEZBuRPhXLw z<>789p=0!*;iFhMqjtTvr6$qpA4sa=6V;lWvP3aWO6`A^-Y&4prTT?ZED!!f@SGMQ zDP?&CLV~hipBA7oA5Exvd3}gKV-m1!elfi`H%`6h@O7-XShea_iSu4X11!v*| z#a&L>uROW8<@iarA8Wfjf7YgcG=o?pvr;5Te&nJTc?mfD$+?`pd+hU=Qs}e1v5hx$ z6WF!Z*ST=j)ZJk4spUHPdJR|aq$MPYw}{z&gJrsjxq(OGrg%?kI$LxS}7gu!$o z(}(>awikhROK!2F73#(U$7LP_&4(qry~0y~CiGF{BgW?mwgO*S%WFB2%-`Y>YK2oR zxfAE8h2z5vUOPpP^TcXm+xaKPDke8y;x<4Ab|sacQx?!tS3Lo=1Rv=!Qoq=y?hU}1 zdAc2scK?u-R15cCj3KXbit}XAR17l--dFh0xH7$QUL-RwDVyt?Nj`As9Zzqm>hK+ zZMXp$w7Y2}HPCw#a|t`$4%DIMe-cA_tFMIyl?@mWfV2^!@>?CZ11N_*OS)$^FmB!d zXvpq|5EV*4B@jCbOal;M3ClNRDDi_vUJ1NLIHMs&(^t}OCje4W&glS&1uMLyy>vUwgb%<{rHfVdC9`$9K~1HY={r99HQH+Gw$tDp&=DAEvRwxw$%~7) zMWk=|fn+})_M~r3cpFkZ-y;QX!i)+nYc14^eP;(sOp5)j7AK8H7)Wa|!<~B)2k$zr zLIlW;}YGgde=oCN36snA-yJK(hQK2f#v+h{{b?qU=hy(FiQ9mS>J)CgpKFK4qu zt4$PFkJvm#j6xVPL!(}GFrvOkzjr_&7eJH2?0}KCvpYtGo=7Co&l6q6 zDk;Agwas-asR(^h@XoaFI7LTgZHHe-RPH-U&dXPDoFSR3|(SERArsD;Vog14c168`GWMIjJfw~CRBwi2Pw4Vwt5MAK?~)rH^|?B^!#GwBSGWpKJb&QqNd>H zgYF@dwEo2zacH+6(-POQw`h+m5T5As%gm?Q~r-e#BUFm;pu*$A{q$& z%H?LeAb}PHob8fs4c`y!3BMd9J+Z)d!b$;Si21v*^TA$2NjI6K%A`K0ZiyJ1b%5>( z-i|Bb$l$?_m#TlEwy8CrUK`W%Z_HWHG@bNw_*MRd_7MblC-!$;GTp;|Ebr;CcK!${ z|G`ziShWo2bXOfaM7>SQ{LlE4l&#W#!?V%cpdHjbqE<;1K7ond1%RDzHueB3!Q)Ds zP2^k9YB{(~Nn+R?nfr3G@qy*VQK91IglZ6ZswV~a0T5Xtp!heuG! zA~P)PEL^!?=Yr6Au!!3M^*87tS#L|>$>FvqVfHqa>jxL$yce8v>Ve}45_yl$T@(PE zZ2BHdKV}jf8U8TYq~I&NZRF?BPh~ehWGP|ArFK7IYj{p5iq%Q&lJjg&2>N@|gBBOK3~&OO?7sFtzS)_7v@jP&>Wd;*o{D z;s9H>>Sa>S7i7^-%09h)D(#FKyW{jc6ki|rISTxpr&>6Bl5_34f|PEW7f&6iaQS)P zG?LU!my%DKgCV)5?~{qD<;{G}p|x(e#+q<#S0Y|_zwX%UOfoRI#Lu>x?<-eGv@(Be9q}dCkUZT(&m#WY=ZVSsReJRsL^doziHRxzNl| zF&S|YcjLGH4*i>AZUhW)=nukjBALQN=rq@ZF_qHxW!s)cfIYd-qjJ_+UjnLs=CJ6Y z7yx%H=x2p2)8D~5F8&Vr(jOA`OAemh|5LB+$%Eay0j_Y1c-&EbG=M#@3V zgFSG1pi1WSgKiK!5SBxoKd-PuuTlYOs6 zX7W~;40Umf#?E6LVrO;EPmF@eVgncpS<`m(+e_*HV#FoJ(EKfrPudl&QrSdHS=a;{ z`}C6i31JVD^g1g@YakS_$q^tpDN~-FFFXUI7BcrOPqFgV^17*dtMDO7U?$$QsCt!O#)wc7G992t=a=Ns_#lukHCu+*%ll|{S*xzCR@24s2Ub=FeZHe zOuURac#=|K~sII^)d2mVnh?X|O%$XA523fWJa z>~Qq_bv?z<6y1#}Mx>@H?fcs=$J2}fDuNu2CQOzJSt7Q{0;Rg?zU@mb2lDa$mnz*e zw#GUoNJ8ia^A&|QsF>pAodVjYC;4w!pmD!2k=B@&pgimSLQNdGNj;>ZC5Akfgc2J$ zid^EhSv%UIZ4s@7V9s>i&3kG+)5Rz86e4p~Rc3l?pp>7|?8^PhCWvcDTCw&$?1BRAg1ON<$EEKV%1DOfTer`X@poGZie@&n5)j$|^Iv>wxVU z&@`WIT+RMxy(E7AC7LjvpmC@biM|-^Aq^l<#IsVrRYR~ng{a` zLeF7Jv+7Epo+vQu$r|WlxyELKW)OQ!U2ZMsNq{22e5t)MIKEX2xS9A&l;ZJ%SE|zo ztpbO){;>7-MpUD$=L`d8qD1Iq4nU$Fr|F4T=$ZIIji?NrjjvY%(e?nIo$BORaFT0} zgR;<+PG0f`NW18#%uOW5r~r>hTceE#gUE2eLJRiQPL0nl;y^G)tRB+0{^MF`Cnxrw zApy9*sM;CI2oNi@H$WS*7XIxT)5dbmK()pc-azX<+dXbHc6 zWA|2#=pfO*uwMPmZ?~x=Jhi1@FmQ{o|JG6K6I4zd+G1YB?U#_G5kz2>Sy2;yJFea3Nj`4p(QyT4X6Q@xMlI{v22P$ zXF|LbNj|8c{kyzw2%O-Of9JECR2t{eTMdw$hIp!sAQd{5{7bA7qI>uX^xZ{@&}C&) z>HhKUprP0BIF1$0w@!x4X09pQjf31zQ<19JoY}%KUlef%JhM`>J6lT*2V1;i`*-KW zTI592g}Y9%<~bK%efC>+higB37`-1-xky`FnsYPt7X zi6`b0zaCaa9#JzM=RB4ff!i*dE0%9(I73i&A^t~wguAVt`9*mH{G)z>NO}i4@?@6^ zuTj0FbA{sK`j~w?@il=P?HWP8lJ_(3a^(hzy0DhVJm}3Q#Lv%-SF2%ud6z)Sx&dQE zAD=dqb<7)lUEHCr{|E>*1VIIGErThk#4jq7o5rspc!O-Q-4z8tp7?wc1Jw)7nG<6c z{KL&rN}nGV`R7Tc`?`Rz`aW1=(-HSk2l^${;U-u&ny*56a()53+$mdP@ddMF; zcU6KNWQFg9ieOx!Y1eO%HiD2o)bPH?);%j2mPGSF1#()~j_Ve1 zRN6KjG}`@u5#9nL?MJVQipuZ>iaAPR^Yvy*O8~Ah>4>oLlMx($MsA-03aW63cARZfSk`faga%AqtX8K8ZWPu6X>O@M-9KLG+z2zcgNx z4?JW+5s8v@pFM;3!qPNJH01c+|3`~L#+30Vwod^;NN+?v;eiM++)byOB`Yhk8-)M32^tK$gK%`h znyJ<}3;Wc|CTW=1DaP8aZqIIHzM190VgJ#m-Py6#vS<^dYKuK&m3WddDtYMtJl*6Z{AhNm}~I zDC-m-BS93C_%KmRg#u?Q2@20nH&yN+2p3u3Mmz_#{~2%CRkH?jaI3y`Mur? ztjr4#snrZ9ZLy4X%k3TTh)aa82@x*XaUFlu3Feg4L+zuWX%h9C>@JgXK$*=h=}t`S zAFh1sJ)$YrFmvex4Y@ZXP zO%Bw7jYdWN9H8g2tM~$ahxW`^_2~`RG?S&ex(WQEACJpaM zHKdduZRm7ck+v2apTjg)=GuW18zW-X6zzEDAwyWeC7}$XGAp^$H2z=<6!?|dZ6|pv zLz^Ab?-7E%r#GQ@?LI0Qprdely1-o}>^OWF;|!1@woI9l_!*7D_tGG#46*!unc!3m zWbX&C0QD^w^8>ewO5fhUK!N{4i+?6TOxb%9L~kzQviVjx>z8Oq(wsBqY_vSh`%l{4 zIH@Y62P{78agi+3+4)&)?99Q~I%B-{Bi zkXNTN9#l;BEByCusUr@qba^~E)X0YO$z9Qu^o=@yhOkoDz;4ZzIDb|GIk^h;>sLuV z1K6++L7$SuD4=sCcO*FyuI;RWa^EGsy%OPToQIS9Le*LBpt8N=$~*LpqLEjJJKe*k zAHDG+%u3(h$+%Rr@8*KztN+e}8cc#9h2XW_>@7P5Q6g{pR4L*GdQOjq0t70=X3VBhe_L`XEMWI zgBV8pZ{0vuZ(=OgOU9Vr84Ion*UWz}?yHZ3aOPp#v*qn&hp*I!h-HT2`WdTB>9B3< z>XC~;r{<9z8n_(v&Y!Ug!BY{J!O9E|#_akC8N%vC`i1VLM-5g#D0{#f^>PKEmFiAP zN<5YBs0Va^JQw0K1{bU|+NAi9V9J~o(O^$jboD!v&cFXeN1w4a+d&k(-RQRf9)9T> zymztSc&Oxff)mCE->-A`-vjP7nrTjiP_(nZdImhAU8EY2kPHuL zpoq0ANL>f)6csem+zc|b{w0Ep;pFTC;&CTp>+ldprvT4sf56t!AeraXstY9fHv)fm zXOKmE9dt?J;s^|zu=&|3VVHljX(SGC!XOG=Gh*vXbon|G4JEM-_mIbDr(fhPPiF}- z7K#U%0^ksU~;2>IW2z^Z=!G^M6)7 zw74NBMXdv1)pHL0?^SQ%tww!fMLEufVKz%9vdcj76m` zsoKkgp)L4V3}?!qIAKQH=Jy$%^h4n-e9|(W9$-~B>raTq)Nj#*GTj%o6RAisjb(XDN!V=uR zOsKMPo<;#JvvUVJ;>GEPIM3?`5W`q=1I*gHrEvPF)Sp+(5v5 z^2SAF)#VL%DGh=9r46nZ$zYm~$wsBv=$Q+;Sn8fu6U^ON9aNhm>NX3?jB5pi>TYbP zb%uTwO{*6NeKdSYvMz3GRUR5K)q7YeGNZpDfh01}hQYw?NYULs7$_RY*c$FNxZQ8d2{7D zvvnkDab3kc_HJmNpbWaepoG`whmIYa#yjjh=V<5bQiJ?lsByclwZr%OS&G7)XSKFh zguo0rFrW&L+W(I_QS!P^ZLe{+?38q~f?nvitL6frbTS}RGMW&76Q~>_kP9l}&FKma z7{;^^w1L*R;N@f8x=ri!lFs~dn&UDP?V-GVglj^Iz8K{KdL$)qkNfRE>k{-3lJr_n z)9?LZ{-ZSYcXgo@7veAwQMJ`b-9IVd(4>haAO1jmdl*ma-%#JXQCBSqG}C_|yT4yg zIOV`_NvEm4+Z6mlIf(;ju7u3t9na_Dn$NR({2)XoD8`dyDa5dfmoUyEyF(u`H zKn4BJSf9@~IuZdKz5Ty&^gD4g$vu`tnbTyWYOm7juIFG&(0OS}fvrqbNE6YWgkXOS zT_D^X3?+-tod|1n``9`r=)%7&yu8UVGB@O?t=#0An3mATF;3^+|4Xv?w6*jrx&LOK zRPUF_81Kme@YI)B3fQ?BMyjCAaSPcP+K27ekMX3|O^?K^@iDl@?-@hl>n=+h445&i zaooyUs-I=?5A+KjHb#GA2evA6eq4M&wM4PlHOvg)kM%y-+N4KJuX!5PgZ!pJ}2O2;VjO1LQSA+E&7iBPP4fi$!OXwn0YVEF&v z|36Z?RQ%7|q?_H(IB&)7$%f>8|7YrV*1%=$@{bBF*Lk6LTo$RdAC{>|CR3UGuqq0m zRoT#9Osof(F)eUvAeyZLx-5OkY8r66uxrtJC?-5}2E-Y4C7BeLwgid-f60a0IQe#_ao^1;ksz`nb^+7xbgRvpL%oI*4nz%bT8 zFlG@2vHQ0EDc=FqyKVut*QfUBQ=p49|C?h6HAK z3AM>Wrm|f;I^JqNK2bC43Jh*_T#$oz7 zVh#9Rh5z^P>3caAy@U_yd6C)=MYDg$t^?QU(+tk|3XEJ6q2?AdwNGM#s)ncxaDI!v zk3U2H(TzZZo5BqITHijlFwWb5QKP4!ZK8x&(4!~^s?2Mk<|m`)MmScj$Le0ke}KP$ z6i-~n7GSLJ(0pK>3*rL|Uak?Ne;f#hwD$Y^L*Pm)Iz&&)1kL6!N~gu-EwiWR=f+d~ zmNd)|T6Dpf^S?enE0hJni|fd)IokKWm;E{@oeybfM-`CW%sWztd*ZAnN}&Qo>OVRG z^YnjsobYk0rRK~1PsFOO2z2JZ|DLSEeV7%WP6XPF@H|3XK$;*_-WR-&Gn!(nKS8rV zOq_mDV>g-UH>Ixz`$6tzN??JnG-h=K#sre$4i5ZG>X3F4^uaIx{M(;bX^@>8zDYnlnK6G^}yG3<6TT##;6Fj1F=;+Js$el2cba7P$CHnr45CbeX9*z$7J8V zBdkDcAo}a$#mzg>FFp(KqIY-D`=U3QU>)Ndf&iO;@e`L*StAhta5u4pW<8b+UuQ+x z-do?}Zt^Ct9^cbMl1^XR=|3hRw40Y@u)S5Fm%ku|5G0WbSYgk>y{Toy6AJE)hf~7h z?C$PJI67YWeS_e_Fce`EhBy(#_zT2yMj8OuWJK>JNvrGOO$?v zz|_Dae?La!8+`RU(&qngMrSmJf~0HH%YOtR4LvgZbzuF^aVZFAbC=&$sm=WiRWirm z#y%*Dn<-JRMTmNRdchoODyqw%BhL1P0Gb;V?21H4KuNnt-8N9dLSK7e6)M?Uy!R=# z@8q_xUX%_}Iu-ol@S`cO{vkR~uDnd4i^o3qNb1WJ!Of^c&soOc0~ZSN3?-b1=?N0- z`z%CA_pak*&B>wRjq(3bA(+TBDa8vQRA=y?BM7ru=W83#p+&1PU$D-cqWR$uX1KoI z?-v|gHi-;%5kJ@K>Rn&9DM({}@+;_|oLREXR$Pva;9H;uT_*Y?a|IH#$J?T7O~IEv z2EzEqYxn>pptyKGVFUfO#9Q!?qgSZMkl@UZ@`UukwT{SFfIOPUNtRFF!DMqVsf@F? z9=~-p`&w-hOtK@oUlSu4U{}^)sgo@9G&3|5w20c=X}DeH6atqGocbr8yE+CwN<=2O zg5cUToAlaUxcZk%6twH){IXbzMtU7rMmAi8Gk*6DC<_RtrSX#bW-i~ zX#4FeY*TE68RnlRTPQgiE-dih&+I8GN(pJaci$wwPEq>$B2Fn6<4s3&X9V`xG#l?-kJv83cfvIuRmF`vHo+tBJqjHGrT7IqTuVielU|kumC1)KF3jJx?_%5>aI| zI2F@hh4YZ(5AB4R#ApP#g@x5KSTEW`1Ihv&M!}!opQD!4befcqpquL1aH_YXn4%*H zmqq<7P@o&vy95w;gzL*Bgh9gB)OcA3Q&``fanqeiZ6DKgzR&!Z(nwN0mUuVe=ChiCWQ17WTXTjIrM3Y*b0gN zA5~u&73CN0J#-36NrNaMAtllw2uMiI5Yj1%Al)&vBHitf!hm#0mo$PR-QC^Id**-N z`{AznfVf;P^f}Kyd;eJ8T-#`$7$R+i04)X@2gLC|d#`M1rzpQI0|p ztJ~|N!FsB}7XQa)%9e)_`vdcuC2^TR{Qq=oNYfA_y<&L{wqes>vSlYp9||TzzVQDu z8Ge}QOZ6)LJ)y?62H_(NVU}f5=3Hb|XUC5_^d0zPie|;LGscz8CU_Hm3}3%Aeq#Y0Q!bW$ z_LAb&_S3TVec|}OO>fq8TO!^|qMz*71>#>!Fx|0PI{;4z(~6Es50G)S*}i@-0`^R! z2MO{ZkS+b>Tx{~dp(Lj5JWVIT$hlEwJ$!KVRRF)~7NIpijGMeiss>|qfqh8Rl6d}9 z4xbc0IW0+ENSafLmER^w9w1$>OcDBWq&$BnM&ovM63Yy)qHWk%uU>2JQ}y@Un#-&a z3M^>PHiHkca*<*_^i<}3=Xb5=BwTtfgx|_Gnl1VYXnBFs*Wq>5ojcJ83hwh|RAeA~ zS2WP9)a0XammR*uGm^f9Q65^GTB3H+q#ZmmT z?G|A4wGLFl)MvTWHym7rzjNy%RG;YJm%7$rFhO7s_3ivHV5N6RuKsk;yjHJWsgCahk zAc<7*`3G0+#FEnw9~);lBm9DJaOj+(!~t7Ilo#UAP23V%{xcHAo%Q@S0gN!ERVu+7 z!6;#+vj&dF@R==9eP7}cP(j`QZv_?D6Mf0RrPf6|-&QSbs?*59GNg|tc5Ww4^V!hm zOT+`U&Sp1(ck@Y-8{4 z9Z<@`WG$cMN#T?B`cVwONsPs#;e-_J#u0zcBzliAH;|~<0 z!?1}MoL)0x!-{oVqN}%=?lDfJjeG$+1(5l#91%#fBVt_-B{-4_OYxJ!a|su;JY$X8 zZ%D0p))lO^?D$jwhP-+k?@8v%bMK~+s`o#Dz&BokCyR$62=57moy;nP*5UKj_r)X$ zW=UgHe5v4(1-2Pnw)4*+L$nr8S@zdMKKJE_B;Cs?`7YLi1Yp3V#Ca>R7IPnScYqH$ z3w2%!Pdj!!{?QjumKrC8J*hAfwBPv5L)Q4xH-zMuSDc^g- zw=>oRk%Zc4{mVn+dc4o}6lV#&a9+SLPU(;3@q4H7k5%o^8RlZiB^DL)9zJtLRmiD*mktPfehy?dEX7ZqaQE zh9nm$Y~(%meFUV1s1qGBgA!On$BhF{W}J*}kr(I;k}H~fQVfKdz=G>Kki*0PPXaq2 zT&3o!RdzFYCeV@Q#Eh+7kmC{$D}sDVcq|NbH10Dn&i zUEC1$dt9!e7?a&k?@L$oHh}5ZSPcF1C9vObeP!o)kV5gL7zarWxW+Q~pCTE{FIaLbzta3?R=|zeG&x=QZ2&{k~_6XTq zQjnScc(;|E8N`Ar4Xtpqf*w82egy)Ho{*5wdCvHNXW}7N{{>#t7Zo6#$uYsaEE;%6 z&afhd^AX9bprop-*EkNVeKqCMzi_6Iz* zst@mKgZ3z2sdM7Ukpqts?C*RpLUx5pEHQoU5Fh%e(2IGnoEdgW31-?2i?>!0@%v@K z?Xnj+w1)DmU{T=@94vKB0hz?Kn7XEBEXTH5QD$Yp#Xf*jL|%8``&?2DpeeR4pYMlA zZV$|DZGKo`^ohIKQ1vYbAq9WCecSH|R+a!0N%6Yp)(^p*VNKNv5QsizhCqIUz+Zg>O9a$>)e>&x|R!W>wr{G`XboZkbhbBX6)bGkxBm$m0-C}0v2HpH}~C7YB>Am*NtXiA)n3g*nFC(DuL)d z3dQ>ijG7}vLi2E+3+-RmOdrSJWw>`f^bA}}6)#77A_^v%?Z*rkpXai@CgL}Lg3hd9 z^RJ%7O>)I?f_!dGS&avA$Wc5GkmE=EA2~kn^y$!pm=6wrToSVGZWoBa6FDEd(S(RCkQVB2@7CxsfjsX}X<|I3kQ z44SmumP!gIpKld=^t;w$){^cIC;csHy_(WZ2liyKXks}9Fc45`(5LZ%iX|m{)S952 zH=H@9n;w_7;Z(^?;;)p`!uG;#?T0e{fqIaFBs_;2Lf~8;r(#4*iDlVw!Nx{!FS-PQ zlv$u59-x^gbZCz?73y_hl-n?MpkCEOKRO?Vgo*Y& z{rGUQ8^;+FT6NK6fNv2;`2Zs0jG>lmVrQ?H@XYKzf_>)Tr&LcN;BLJ_7;bvWaU=q!3v9WWEq@VTDfGidABuQ2}`77Sj>yPtF z0cQPE3&-9-o8#yxo706u4*v(v4XdEz@p7!Hh~DCm61Q{bgPVV#mD%171^CL`tf0V) z7b4vqp@%)Oti;#YL@yPW;lkg+NlKlS@JinX?Gt)zXWcgTTjp_H{&!6_AO~b`2SkxL zol=FO{&J=%%Tybjl%3Ky4>4fzK)^ab!z^iC>T?UCIBF&9uP<=Cfev1Co2N{LM^q%q zLJNR=QsX_#ehX)x>_)j27w|=uA!C0zPH9(=bwIu()@GTx-+Q5giurNlcayjQ(0uU60gkh6SP z?lNqW{_r8^>Ytpk#G@aX(2(_YrWaaqa_&No?53K+vG@ckJk{r8Jw>5jgRb#q)+NDO zK~*`@Ly=j$;eJSFCMqI%aN=lpO^1Ui^V#bci z`#bUGtg@3!g8QHE4F)nlgxEtv`Bf1+oc%{bls{7`GGFI;=_WTu4tD*lw1gT=PmeVh zXWry(@A$$!=m|^YxtE0!Vmz{6@pAlGHCGV3H{e~kL!LI-FLh3!3<2NHf3bRv2y*Nj zc^$~2U<`w@ut8uNI%c*0QB|!2O(o!&nA9@zcHVoq?)Z}e@@yO`IK7V}xK%ahb0r)V zz(;-Lcl`L=xejEeYy&5HmA0XK$8Xq9X*-e2;uqFQIT{=21hJmsj!uwo2f6k*kYYL} z0$!QdEO>}F_Wj`3iDsG3Ux5_-5Rwf0PO6L~@*B8u((63(2VmE%{qZnlJv_VVsmaw_ol$`Y0%lnr`eN)tC@UD(}Hiww`7LP6N@qq zwmh=L>Z1b3E3U$J*`aNdqOBV+i5R4(pd2UQ+1qc1w`#_N9mJ5 zA*BIM2~*?03Hbs_5rt*Tn8{dVR`HNDlKpscs}4jgiQI6;qW&leb$p>{g#HeLG zz$Tik3e>I7aU5<#b{6?Ah32zG3LA{ym6`tYWB2AcLd$5r5Fz?8u+Al<#?139S+TDC zv1u&;Qa&Aq4if*s487vf&9TW|LuL@CI2aj%G-#f@*_m2#*K$09=*2XvgK_mzT znf3O+PmIN}0{Mj!PhwjIF{Pg0>|n@qH z|Fh54)ZD3im$4YU9S=alC^rBq25EPV=pXWB=P}*!NVa^WOQ{{GG>-0xdYvPvt#0Gc z zGpq9UThtO~Tt`uF5EhyaIX@D=4$D60WkQRTBp=6tnJ2;qkI5)u2hQ(cE<_3NY_+0btEski1y}BqAn3et4(~m?eI!*o@aqqdM;^z2 zPj^YNX}X|@(Npv;Rj#FY@L~9l??n@kWt`@6d-q(~=)l+x58TU($ZurH44f4}u@Ql@ z&wA>lH1IcE9dloJzlId@sMeqy%or8Z#7lM<$f$=>1UKy7Lj46i%`&6ibXxcO*(pD+ z5RG6%9|7^tPXmb&pT(~NAoXfW)+eby9TYm0XKGjb2kS+vM)3Qqf8TD0TZ^vWiS>5@ z@8zd=dIv;tqNuLj^(H~!-BZ_#KI>Zy?`Q1xv-;cOwE3K&T!Jc+*C_d2`(6f7r7Tau zDnWiqb0KeI3Ht4!)`#8-N=M@%f(_ z4rJJg+vPyM&g$oHpipxx7M>g0YqV9kho7|5USvecaIE?0^2f>bfIm}2*x3c|?RST? zbuKapbfE+(=Wf27$Ppc_`!d-r>)KK`6vXvZgpVH{CD2w}E{dFJL2unBDJ6o1cZ-5C zv>sTIC$Anxnm<)bf6E!uI$nWOwr|$-rHsx18-bQ zF{6+rrFIKUXWq_W!mAf<1-aJ*)l;_h!b!2;A|!B_*?WhZ0p8>S64mO=+_h3886;=P z3$FH>eIJsVTx@VGP%;Y=@@REotb%PTbB-?5jN}C9p?KXm6@|xh&F3B5+dvx~%iu_? zevF~j@hAq6H!vKvj+w;^tXCHUu1!mh6clZNEfoXRAS_LE+`qZOA!Gp;+KPWUTGM0! zlG!3F-fj%aAdJvH3UB1Zrv24anR^|l8!knD(gf(*oixV*`-;l`02$oFfRnzqz>F3N z8bR91wn1`tD-GC!;rm= zvKR%rAPeEuhxlX`Fp>S_CMwPkQ0eV4HYgw1?3jGH5Ym&M6O{vLB8!5RsGpZU8RjI- zZ_^`d$HnhVyOxU%4#b-XC7)90nv`wIcd8phxG5;#g=-Jt^*)VMkB{DHhH5c1;!R}2 znhJ^zM>CFozl9ssw>}%K%vHX_%g>s{u<^Cx_pFIeCFfzm##GG0EefU7W(&#^LZ3e* zE6kkEA)|ee&3>vrvUL-|O<0S0c@?eO!!UJYqvX1<#^LSO>|WyDH6eL-?Pw}Qw=#c! z%~;Ijk~mgBsm%K0$K!2CWc93vwIFW|SU>OX0skS5F7@yIckhv8T^?ZnUZqSXnEp$i zF!N6n=h72BFTofcF`mOosf_}oD!xE|v-2XBCC=vqX1X5c^$&=jiKX&cVDXON$BvXw zzLF%Yz&FwXF6FwbYY!0L^D)uD#a@;lVn*8!CJqlcdxmyD2?H4GX65(8Le5gnR$0aN z;E;wpp0~YJ!H>c4Z`*w1a@C`nV2NdaZZh zCCvc>wkOO=VFpSkbV9&0PBu9}FYA2XV(VG~kAB}$B=NJhQx4Jia+5c^SC-_izz*+t z?QzABWcAg*gbZ6HwMztJ14vCVFu4HWv?%Tgct5Bm^4GtQA70^HV$*his-Cc{2L0&- z=st6>Fr0~O&!8%w)@I$;3W=eR8x1gag*IoAuFdIWW;jW{Sv?3FY$B)*rABeZq_(Oss4q*v_aZ>!e+$`JyGLL=FUwf)vi=?|H zNK9F%Hw9=p45Y|!1^^nc%kpy0^HX)eipgzj`)i{>|GPEmB&)Wej8su1>oVXNU*R{ZQqxpJ$AoaP$DEg?M{~8 z-fe<^;5W}SQ+HGy<5G34FR2FEa+kk=Qelg#pvfBi@ZM)dHPj#{X071|8;QWeuJTX$KVdK&}YRb zwl4KG#yko*>`=MLJEaMH!aPU(4zZc;YT1K1lS`y!vMK=gJAPWKWm3IUzt_}sAAVay zo0;OrZtesG(r%IjTP5{jRy5vsD0ROc2>S9CYvCpjYuO#-Z^Aa2Gc6}?jtcm1Sg9`t z578FyEX_A_P*$Kn)zj8w;T-|$_3z|O)c)`F_gJg15k*H*%x5Dn?pgPMFP>=&H+t-{ z)~^L7$B&`bimKI4JKYO2^D78vFw}b49y(w{O{AZ#6&kC5PH9QE2tsr_%BD%gX!hU| zwvrA1fxsU4hF&@eQDPY~SJ3qojYLvtFgyCt9aVkM_yzwM@#4c4#p|u_qJGFvAzn1y$y9>-wGrsEVo` zCpKSqObA9(jo{`l&YVan?g=O=N+AL8F@+;h8UO-SN2@146pFCXoe)1Lozd347?UDAf_wmTZ!U+Qx>nZy3}8od5g(zF>>&ZngwY?jc8Gm{{skOC%|I|Ezhzh|R1CBR z^+nU%CS|Z#80aZj)v)F-Z}xoWyob@Uc*YQWkM?PNCFJRx1~aq*^?f~*X(vWi00Z_MQ&JOLN6BX=aMkmVvccVv zym3pYq{a!qZmNI2GcUPT$CJuY8IEFBJ_U!Lq=D@Ax}n#e76Y~twOM||xm=r{;C@uD zH{}{8ZT{Dl=wL(#JsOo{=tJUAQJYDzvt;~%MnpXCY z7gWiY#D`zbMp%LSAKBzNsu!NfjbNX$r9=NzF03{GIA)ho1!5r0$|WU#w_dqXLwc#n z(WP>NUO-K~nOoW-KVr;}q@~_$_@*u{DwX)5A|WHH+&q>N*A!~;aJ9%Lb8-S{eaOp9 z-xBiU z)u{Sde&tDwyt`J-vjVnIL^myCKN6ymX>9W;vVRxBwm3UHiY-SDG5s+hTi!FsArZmM z6Gx&mDsL&^{gJznt1sqA-Fj{_)BsLKX5CQoQkZJ!&=$k^2e_V5bFQ%2BMnnI&c6F4 zlAgUU{{oY&Z#77~JS%1t2dR-eMHK^lWJ5vNI*)K_bGUaDu)#RLrUm9 zw|ST630KL`UKiJGvKzsC4|`baIe8OxaM`r;#iYgJzLg@M)vYttpUqx1+$3d~#!G|= zc$tF%XZ3~0@p_bb?F1;@&gd{P1Z46rzjsNh27x#E4c_Vqhyj5^kYv+<0Ajl+BKK^vMoz#(gTEev)Dr|UKIr{L(fV1N;CV-9X$npC*HE8h29mp8_=9nqb zcdAYfXCF~b=g7MdfrM@T%1?PR{~juh5gykR)sb?7RwuP%!1fQ0UP4kS``z(!X*@yd z={Qt?luYuGqX&ASllhmAy%g}5;c2E-g&*G)z^W@Jd$#o6EV-HJ{8fB|;J%I_l!Fxc z)}VARwo~AgB+de zOh$8Z*Y&%@e{DH!X|o{G=RI#h^}!jvBufY8CD!F0DoY0*=Si234iX0ass7ISV8r@X zSpMcqG68)-BS<@lZSwa9!H7vwel z0$4kI&g$oC=g8vDA<%9*HHXHNQO7x0(j)y0X;5moK~#u8-kaH6l)HyOr(+u9PO5&p z{3}YSX}eoKJa}S(3NyB?L{=^O?QX7)K9Nigo$MUPOrEu3 zqRb=0Y?dymk^S1_zDH_U{H#vB;&ZPh!m`7OIMB_{!yUK{5*_5o8OdqRcQoVZM4 z)NwEB!R+4kF)H1A2c&em{CSBpL2X?m!N~YjQZizrkcL{W`+L9Oe>vPG&q8i~?fF4q z!!*X&M~bM&mhK1=msqZSi<&MsLHA%w{4Yr60Z&2D;Qe*RDN1pCRpM zryU6hB&Sk2@dqD8P$@B@8q^L|X-+>K^mZ6X_7w2i-9E&EF2p{LR1YeDCRrN8$i^_v zTDi%y!SdW(L)GPw=SZ6ylm&^MH8eTt>uWD;2U-bw0)UtQlo;}aAY&k~p4ac0LP0Wa z4)bIcAI!Xb+rKCPCY_C`lNOAL#DA zjepN~wC1j!E2l(X_2&_x+1fj@8-j!qiMWc~B z8IwbL^`0Vk3wa!bETs4fT5ch?+`m@5eE?E%Pi_b-`Fi*7dkK>ftfQ=;JJm~l)_YBK z&e?*VkqX`eVv2_V+@G1TT_CFXQma4`uuz-)7jACxeB6O5^bpmV`mtvugYLHX_aMs2 z7pz)$Kk3svs}*2v{)@5u{XNJ!cbuDBirgC`56XMc#p~^iYhq7wvkq8IYY;E}_9yh% zM+gm;lpW-Jx}F_L&H46ndKR6~!t3GCLk7^SG|JM%9ibfC20Z|LstNQQX#T#I_4fZy zqT=9I7`E@%L03ZISxV*)7&=HkGxIe3hiR>7upV z%Zd~!i};lxeb60T&Kr;Xmr(hZE+jr97gNSR%Y{)YUQ9*z4QqCYm#|67h|1KoMxpKd6Vi7&@*c9eL2t0{ z3H7pl_eQtlWF02J(LS^gfTFnvak1&wx=zRLo6w)ifXu~d9rO#5_1kFO-wabTQ!#S` z(h;h&hlZly3GLLbWjH#f&s=HD&Y(!&CV)KEle#JU21}GjU3-2dy$nRDoKHcUw`IBC zv>(KMj(>A|p1=9j_fW1Zv{qahFKbwC0zay$gF;Ve3f~fTR(%?$T`*wabFs3^ZJq8s zT86T?_+^%vb;Yq?Y5k_GnTFEf3KqTfq4^#Bi+?VjR;U{bXpF~AkhjmK*nzDLEXCI+Ejl>1}7HfjKfs|XJ+zZ={C zbOhr6za!YY)rTH6)UHub_*B<(;;T}1g)#p@z0o5u`hyZ3^fJugNK}JxKHaxMl)osG z7&(h5UDgVoOKT?v)tteOCtvLqL|N7txL(nV{j+!_BTX5;>_3`i0GXXTuZO!a4MSRz zjW0`j7GA(Wh0^r?P%?=*o1^ogXdH%~?{KIW|9)iiv`izdxUno(d#ZOc&*o_}l+=Pm zZGg-skhvp+#AVGYG3JY-EPYIR+OxJ1jaTVZlkjQ(?RN-pXDhxGedBa+BEjURnv0Z zMfV-v2mhaL5Ez7+#{<ABrz?m<#mD^X1X!2V^Xw*8+dyepogT#)|VCI4S#+*xuo7P5GDUk{o@dgD$ z&Lk=A3ulK(!=tYSR(O#bfc%ocYt1t8 zk2fEUyhP{Oc%EzASt6IOX|0iWoA>G68@+Aoo-fwA&qkiu5sb!pX;E3;qf0pS!0PYy zb;Ik$m@o8iGo?Z{+J9zNdryMPklemtCR%BM;FH@3&^QAgNgf?Z*J=BZ0M@0hNY`L; z7BtDT+QBbxszOr7P2;`G{3J3YKwD+RLkTObl>nx;Np*R;m8KZG=SHDYRAE$Xl_j8A zze34$EU|dWu?a9b=8M5V;MOYhw0E0}7%tVqO9`8@sY#D(?(&4{7_Q(}RXOiwZ|Til zCd>d&+zb1~R}36SeTxz;`IpPAXI1R>No^hY1Y?$15P|&h z119l(z#hS#uEKrtqm8oToQ#6Il+8T2wGCqHl4US~ix>ey zE2UtA4qpa4L77&&L7~f=%^{gseQ0VU#Utmxo0u5>`Cc&8qgk0oW<3JJY~l`d8Yyy` zv2kgyKy3vn7|u2eFLIML6+}uumQsv&rij6NX*w9VAmu}((&IHvz5-Pq1_#-O&en6AlKjZymZjAW|3y_{~zwUm(GHSJHf2clHm?aqg*W7Cu6^-m~Y6TX``$=R5yi86 zZp~uouw&A>>s7%W6l4BeGEq!MsuRM+$%o+P8oiqXL{xFZA%6YkrnQ%Thm)zQ-@Bv^ zpa#(SqgF)d$(}8_hLT2nFY^0!&oRL-C;T6wK6?B6Suwk(uby(;&^@j8w!dtR1#Q;7 zjH0C&Q#^p8=%e7F`QG}p+!tP@jDNl{Z20^y$U>XYpc^tmrfRZz zw=q3zDmjedl&TzY7G2Ys#-jJJ z7)5eY%O5|uWCJ)%%Fv`nP#a-9uXbwaa@w>< zfd)l?L4Px~{e#P7P&0Xd(84@D)`B9(4|A|d0&(@;pGz$Q9?7YEXJX15huT0>z;<$k zlzHz@A@M}hv^94J)bSW2Wc~KHyX(XNINM7qJsE^@r?Ik1E1M{ojBNl7g`XI513>d& zVzrKTe2xQD*goL4NBs)&Sa^p{!uXArsaLr`)*PGPP8Y>Jbk0c7AKwBvVD0u6h5dtG zYSPq&g%A7lHxXy#Wi&Vbu_SyQg9y;k9?Q}4)5q*Ll7r^QH|3R9l*A@^Xy&NB9=m58 z&CP})vhIT@zVOpdzXys5>KF2@Sa=R-ARjG|+Vk;h@i+j5zEF$lb}|3gUj0P9x2^H6 z!b&v2%M-#upfiE5){*djOz}`fR$9m4o4+_SoE^O&!BRb5^Zddcqvigk*=30!15Ecm zQbH|<`r=0~vf;E-Awl2wllAyDqta-2;_Xl-XWu4<%NLL6hQ<3zj0sYle#|6nrvbIB zxx|XIAq+m_x==O`-V4m7_jfmEv01JSl&ZfQ*=G`li_WqkP*D54U0Sgo3c;P)@L0#u zaA=Lg!^4~E8qS6>>ZO|_BFQ@+wIGdd4--%O`3K!5*mTb(-Z8w>dh<=f+ZgLOdJ{Mt z5n5dJwRQXXfQD?cm;|!MkCsPw#;$4G3d5?FoxPZl%N03 z+Z_uTb#O9fw0Cm?us1q4Jnx4$#6?J892;!{@hTHQ9i?)N7PNp4<P| zaXlcQ!Go0WhMz!&=fT&(l1$X%pI(E#j8)E+XZ!<_2~tg%>^F2hCE!(_Ks~C%nQyZc zA;(&a<(=!ZL+UDLm6_YSYo2}CHAmhn|Hi5^4z~(2#5)MqE@>olIM$j-wx9#2TAs52 z``vN1Q@BY42uW2HOcA53Ag}|Wj7!rSM3kAye@oUC>JlH8^`&LAi$3 zGc+VorTGo!aU%YCWz0qMp@6`EFzRWtCrXtu^fv-pR(O8_bKVc#rVf$@b|9-w`CJi? zw*?(B3poG5AUwWci~Ll#VObCIuEcz=yIc6OiX5;mT{guu z1HeU%XoU^Icwoh^_y@F~3dY8<6--frHgx?5&Wt;p&MDx)`+%Tf+(C(!v!g-DPiCEA zxLw7?{fsFAYykcw15Xv{3tNDuAK2b^0NSvp4We1+U<=(47wY zOZ52Lh(M3dtKLLVSmRIWgMt*cudck~CFvde6C_n!OP*Tr*j5m~ukYGuvO`{-T+j8> zje8(n^Oer?9@DR!lAW`JZPFR+bUW*?Ou34C_J%mVB=UpaA2^}-b~MaUneC}3?K2!H zMu4~c!2GxUr*anVV&1!L4wXk1o%V{4!8^|$PRsEq(S!i{NMD3g!W^xD38B~&Z%j~- zjXzK3`C?$KdwNFAnJ-flsMZwSxE4izfqVT=A||T9zS+PyLNRI8AXMcat9KGh8|MSP zGM26NHME>(_1oVEY_kSKcHl9w4!v}gddAu)JMgST3vzYUk|C1EFQhpSgrWEOCT2Z5 zcX?S$Fl7ehUX`eA^2*=qvdZ)Bg?-!Eq0(+^C%I~8xSelOQyXee!jJVd7jZu{7qmYI z*%ay{^#us%XwX z)Z50G8|+`OHmJ9rGSwX$G}JE3W(v(>Vq#(sSvSzj`rljG$Lys_^>!I!P9Y72&DovE`>< z!s`Ci6cQH@LfHtxJm@qXdi7l;UE`f$utOe)rupZv4}}FPmIc-D=6T^L6Qboy(U^tZ z%gyp`G-9YlM6H*6884H(v7=M0w1$n}p0k~qG*Mkvs$09DNYWp(kcMMFlqjLWoX61U5Y_LkO`hJ>?(Vs!y27Du@;y+Ptv**Okx5%+VT3$=t-lKTSV{;1$1sdCE&pzN~Q>e|0bu`Z6rAQD2n zgtN;FqD#uPqC(Fh{DGc<3q*7x>zDeziL;GY=C8)R!9uqc(7Kma{lu;5@(}s)mRy8y zvy3=73odmF_C%#5wu1#$Z-nFv=4ZEoiPvQg;9{bBMwVb^jRWcrqa&w$dC%X4pK*{W z+-d^Wqnx!sBQTh8PDNg#r@@{7Hos^^swlEGbzUO3Qi z7bUmLIKkdZNfA+#y}L*Vr{eDn;~7A-{S3Rsh}@rExiFLd1r~Aq-X~io^;((wT4Qe( z2vOF~!#!oD=@RJSoUj@k+YpZG_S5&3C)qe@c1*+ojOcXZ%ULrnQzrk?%)#80bGiEP zz2`KC_dF=9m6=4*Zz08hh}@%}CfoVbe+eg5>tQ~ON_74b{Utw=uZLs>Vx}XQBmgVu z>nYyO%ywKGcwvtv62%PZjuM0vVM(#Lpp{1?a#Lt|(38lGXc-@82$p2c zcWKP?IBpcR3`m+OxT~E=6iNj4hmlfj^#xpCqm%BV;*&@b$pR6qzy4&((2(W(`;g&< zWux?YVgO@Vb@qez)#(S;`sX=WS{^jkYcvk=h%lw?f%eM*jLRc>aJfI+>>q4`l_5NE z_PG02%UAf^pbN9o4BMYhk?^j&ueb^1(Frod;+0J`MY{1vQm`W|q??+N`wBTt4dlK0 zTkSrl%UwkN9%v7+D^Z$gXP;tmHRcy`D0QI&K;2Vt$gv5#%$B{DFa0oZs2RU)hUvOK zPT^Av@*`}ze@L87V&=@lJCSWYR*b!PpihD%QL`RM6L!h?6r@CH<}z(BPtc4?^QrrH zuUu6-ulq$b>PXm5jqM3Ta2#&v4qOX1>{>8j$@SV^GH7-_*9Yp9t53yK zf^k2?j<=l$Y&i?jmJFw9ojX z?N53ETtCcmhmaq7CUu{*;&al2qKA{AU%%$Q8_XDb>nv-4H*x3%#N?f^b^DKbxjk6u zP4`W}_kxBHv6qF}iQ>;+(mZGSCp-~s4YBEdDeiI1lRn-P@x|sK)%qg=1t(O(x4BMY zOv{w&5FMZ7L6G?2J7;&&zRS6Z1vn0}ynU=Qze*PaMo3QxrIeVce>r=P(RPk9O!Xt_ zTLt=j#pSTO9br%ht^c=R=ybnTK}BBRWrY0;z`hp))VkVO!!KBr$3hrz>v1OmGu|-L zm%=+-3(CZ>ZdtzSRV&|dj~^^>5aT*o%)K0Mo{fq%&CWka^p(0FgY%8n!EZAIS~xr8 zu~>S9xXby6R`J(LZ@L~Cl-w8T>+4%?CuZ$7ixjt1#KFJEKdP3&^KrLgP$KX3ygDBk z(FtUnfjGJkLZ<#N#?MnXGu&S%8)PB|C8*7FKnUK&M9D1)={(Pg1aMiAjLf4W6jRi87&;6_BP_Y ze%MI}l;-{!j%87?Py6(y=&$pndyd3$jM38UY1SZl`-f;nDWa~sA|kx*kg#(fe&JjZ z-OnCL>*#6%o~IuR1-r>vRMFjdb;*SNzHlB#^Rz3&@mRJCQR_odc{SKjSSLHzAOJQk zo?}~t8P`H1{`zi^&O-MC)CS=yY1sxO?&}1ZCz;V^yI*@`oZ_5)NEf7~S z(X{HF`-e~*TD-xN!0`WzjW9HDK26y?9orlLRb(XnUG3W}a5{L>(O89ac->0vB@(eo z1V4Ks*tLzwRTJ0XYw#thI($bPIfk(JZW0gPMo<~0+Ege1aDMKz-n50=9g#PAid*^h zA$OUxm>N53bm#AUKtYVwkVzW+%Bl^1)5H30+(QE{hC7G5gu~!fil$Y>4IA5oDfHlD zEe!BSazTVIL$#@Ga$~HxZ9hgP@f`y(32rKZKvuMe!Mx<8T@^c$oP-8 z`^5Fhznha2i%8VXLi;INJWYONTG-KIqdb?YkXHP!AHSQReg7F8N0&XhrMyyBml zoxNG|wy|wNb4pKg2&g;BbTo%fN@50$!U_8D#PJoNcbjk>%6NZUCU={Yu%PfCO(nxp z1?ANSPfE@*H7bwxm^FS%jJ!w2DW=OdvmA1+c^%U3gO5(TpA-2iG)$t%<`tzFrd6Qw zC{hHhz=ay?dqxinKw-XtQpzd3OEd&@9Uq^=$-+AOh2+_6iZB^}^;fG(-VBQs0uwaX zQ72;{gr3q(`($&ZP-Xddx~TlK=4*UC{*Wj$xKA@ccVV@)L> z(?9Jg`0!=1gRo7~_chqK14&=mOWU_d^*5$~J)Teb`{%#^Mm!SM_2k+}17}Q9t>Z|{ zqz#kC%*}}BzEEwfD?tGIUjN+fsbH#m%I49YAF@ig!^7Ckm2Y=`WFriBdx6)Q77Sg& zAEI9Weq8kw9rouM76(v`q_&n@vx&v3y0f_HNDp<^8yhEK8;{0(mj0(WmmCf z>Wb*A)1AUh_Q9CefDk+Ot)~XbOjDS7Ocmy-*hap6t5U_ykGZaPpZ}Nzuc_nNR{+KW zi~}(hQ?UnT>E66%}PFo>R+H*>lgS>KQ(6lx}<+<<0_#q z{X@ix^g+QNipTi=f8$g$Fi@W!p{U}vw9vmecFTY`eE|v4H3xTS6v4BfQdoPtg;<;B zmCT}v-@biEx2p}KIm@70gCM~fayWX9j#=2ZSktqqx<+dz*~U%th#2HE+a8rl+rz8H zPse5C!YEB4#WQ{j?I3a~e?F2;+ss%+q z6_QC>vsq{hT0IEW9tLfDQsvcg@}E+8-abb+D|JE&#uoi z-VhZnQpaZEw#)yVoi|r9hrn5yrXX)BQ#muu(4Bs#3J&(l5PUi8CR_JR{aF4bi1-yt z_Y*ZNZ}^+_Qx%^8;^S{Waw=a@+|+ePkiHL*a58HnTr>!>#h;JILxg$;e8gl7SrabL zTOzPQoH=m+)b4&wq~my2tMih*g?-oc&sXs>;#*f2)QGfwSo6sGObRCBP#;`<&Qlgd zp3i>cmB#6@ZTa-={^QbUKP7S#v2?QEk+S#Tp_}oWaAf&w9ls)>&mFW%`?osE1|OXn zv?8B&6hA?-_Dxw?rTEAzrqV^mG)RXR^vY|Zm41%jmw*XOKlANzM94Za8UzW$vIha) z_0jr0BO3|E<>$u{QDCKwnX;5JOTV_O=KVmt__Jg+GQjg@zvH z$UeA!fK*ac{BZ*y5knnOf{aYs*2k8IC^JXI-DJ>rqkTp>X9(#; zhT|f0!mIwo(&vAmk|K{34BgFF&*pqWvxl4N*~QCtUpn?|IZB3zm$|-LyA1=j<**I_ zw;FpA8Kpn>16twsq=;+aY>Y<3a51W}YfH|}W`^m4}(1Qn*^k>>f6w^q+EU3b~YzdEoB+I3zby(91cI*FHAq zARM)|_qwl<|4qpnY9te1^3+S0kH-jyKp_!-WG|Xx*t$9~SsId(rSQhvJ}q}3BKpoA ziicHwhe+q+2*)V+gQzpow)%5&D>;=lc__y@~JEo;>?{ zJ^sI001!6JBI!Mg>^E)-i}Wv5%lu(~;zEmUv2MPFIfBuF>fQTn*t#E5&g^FL9=}j| zkm(9OwDBROJ**S|NzOcZizZft#-=(Pa2R$#NFO zH!&R4heQ}v>yUrfKqh@0@ac6y9tKJTzLiFo8+km87xJZ&`6?Vvg^ASq!5#~w(l1Zv ze0@pn9t1i*I=E-dLJ_Qac>~1K|BtP=4vNBy`$l)^Zj=t`?vw^;>7~0v8e~C`MnO^$ zDPid@De01skdTm&7Le{V-n0ImXWo17+~GfGbXMlK5G55MY(y!) zr-lGQoO{y}0HYotv*NMGj(G&~Qlo^IxZHb>kUnKHdNB?BFfTAxy#bq9=HypO1>)fr z%rKiMgo3yArOdKHRQ$1vI~X&{f&09D5W!? zf>4al)KPIR4fx`iks8b~It(3uWnuVB#+Zy5WIO>LrC?!G`BZC8wPuRp1_zlI%uy|m zGq>LtVTqjC3N7ECMpH%qo7omqw#!k=C*DAUuJ2t5XL9HZp4a=xD!%EPhro2@~jTM1`Z13Zphn-21@g zkVmY;s-O3F!U;%;+`j&d3?pcmOhY|d#hiqZ`1`@-Y2hI~M!S%G4W|Uog_v!4dC&dB zGTgG4QlCOTDdIoxZOeUr`}z9tkTej1T{|3@&$9f6A@{fUC364RKWWMX$Pi4?w>|Tc zV)PrW?Vrpg<1?3FIGO)MA7Y%~vYIy}dXD@rtu-Q2)*M8CH{Fp7B!o8I( zv^G%XWRl#e|MIqJ#MCG+JxJ5#p^b=f8ix+~fp0C@AqZ-SYjt^8zDdfbV^ll!vLbb2 zjBOkWe!?67=g?t7Z-H2)BAg8?{|UuU5Z&S4%Z=jBiTwk@cnP4U`dV{}mFmu0PW+PQ zEbOBeR!tO=D$~ZP_*z2gb5DI{qETJp{dY!I1$d-nUnI6^Dv~@S zG^P`XF$&06h#awjKi85zGpPYa_qBo&HTl zDesArh&S>PWXDzLWYBbZdQ>bMJ!^gmNNiBDyLlTB7c}?~@OwK<_fwAyM`-x@ATnStVktII{Np`0`23(;B4H>v6=0w`5#E= zU}Pnx@;fRF|MP+Af?%|Mc=_9vVBRy!+Y%+mV9u%$b(WRm2l(41-NbHsWqq~LJAE4J z+YL3A1;c>`9CQv8`UV|koF-3(A?5HLS=!sjbqNE)#eZ+jG}Z`SsIM8`?VuRFZy>^P zg;7iPE&tD1vj2xdyqd)9quCQ>q<9dIGKEJkGuoZ3)FVY%m92;P5x{%C>%4O|5ucCJ5#^*{UGkS*3213@#xk8Q4Q=!JkOJ{`Q1Ha+b%Y$cflc-2 z{xy1drN9rhyy)9LEUk1o$IU!`hSCDGwH6WUZb;UnEi`W(2snj&3{lat+op&bHrqdM`+2Qtl3}4|h-R_L1-A{I<7ZBNF0X_rp2hXRn8GTRo_7 ze9lq9+Qmk0)ycQ^$d?@PfZK}w5I_3so0t=(%G~HbUst03^c8R}X*oS`R+U4oICnKB zd~h2Moi3kPaWJjSPssm8?ZFo3xK~;k&&%v-j1o; zwzlmmYI0@WcCq|6mDK(F;;EFgb5s&Og%^yfvUA6+LEs!#O0D7Zfv7_A^^IYD=GP?= zer6fsC;EYa7W)A*l|BTsOR%FRo$_A@Fhmd51n3KS+X zRywJwrZN1bcqs5-mHc+DYq=C7eb()Mya_Bvm)UK!@;_Jx2?Eju=5FT?zK9q@!9xX-UD3++;ha&yn6F)Z#?* zQn&{hmG>-^az`D4c<3g6*84mRd0AJThH65ETAF_}BogL2f%`Hw9g=i=vFEf(98$)o z6K(k_9-Y399FmMK{jpYBoR8xgXB;wNtWOojo&IjGzrP>ji9q1+XV=(u~FBI#XT{HTRX0q`ytAXRA%)Pmme=@{IDR8rLD* zA>}WydZC!g1On(CPQd&RIaA5<$L8BWUGtlw+LVb6tNN>MaAU;LWQ0y1hxjvMB*{ z?_&t-`y_t=P!!IHEi1793;c%~Jkc1Rn%kQ9O-z>QtK0x14Ih{W)7lzlW#+;;J9Q7K zhpzV=CB9WNYxQm5*esGt zzogNDRWe}7k&~)JRF5Gj=JIKHTJ1_!%G%7bx3nKnvE>{=z|>ab4}0A~{8P_mJKvv8 z@?vkG$$6>?=mwMYy);6MBocfAP=h*x`+iIX)AvjBLx;k5<{>)ASS@jX>epzIMZTsF zmoQI#bC+(d8xBZ@pOd+0{RBXZ>Dp`a!2@`?4Q^@pps7;OBT{H#`9J?K8*PuZNb;+!b2P6k+(xH0&z zDT3eeXA`eCvg(NR7tel9tJbX3il&s&HOgS=8@t<;!epLXAr5_&Nq{M(^S0$#bag5R zA`V1=a_I_}3Z=h>!ZfFqI+YF?f6;%4*BHVMV}{skHAfoCy>D;#?y?OrzbGsL4t z=~L&;WxB@ckRMliS=4U~pa9N^a|=1uvPBGJbyo_lLmi18IHr%7n|xYQzhvi!&9sqZ z$*9@?xgxP_R5%sp6NgpuvqoEp_*p$e&n7F2~*d&K;i;7e}m~T!;lLYWhiAzP^7PDh-b8><=WJV38m*Y90jW)qgoSn zohW(yccHZHuOm3+%ZY_cw0I+1eO75`Q2H2n3NU zE(0EdkO*RVYaoOPp`d?$ZG(!X`!phs8s{g#SUf=LYJC|CS<^oTTodo7JOW2C$lJTl z9AF(0Y<8l0)!>_xadirs9fj$jyRCOg58`dxe_t+A!@oStN(aB)p%nknvuvbB!w0kw zn;PG{j`I#n@J?s|@d%&6T3)D(hA&N|+Ja{wAOY1b1JTw~j@yzSe%_CvToz;HfRLSq zsER$#5Pa0f{AXq$K;au2KKC(k?Y}^TUq^rc9GM;kwjKZ-L*e3c_aA$l_0xvAU564n zM#wQxZ|=|auY`a6aN<=4N;%?XE2JE&iVjWoV;Wv5iVhU65c6q|-Q=U(re}NJ`(^o# zsLDRyq0f0*hPHZ?If@XqhV9P3;Im!){H`LoqX)S_FJFZQ@!SfZ8YUkN8HHJE`t@@0 ziU$q`KFf2{pJjn7OUMbch6|>=T#*r>0jF!mtNqJU4S?W?zSz_4O3b)Dc{%PFK#%>~ z6wAr*ZDsOXtN3@Y+zy^P6DzfzUV49g&iC@VI*e7V&ZpQb>osf01P!yTCkISB2{6}k$%>M zy)f#?Wr*&u+8xfy46cz83?FEe^@P$fs~s*d}sY2SO_69tzPr{_e8FF0XGlHopw&x+5 z*e|FyL#t6>@oP!am+66KtP#pys2eyCw&f^_e8IQrFXh-Dwfj;o28ya3r+D&NwpS*J z+Ycpcie#KEKe$LS5geegO4$zcVX$af6q@wy_6B2jBI(bg8yUymf??fWu}i?Z3TZzG zNDO(P6D<=H9BD-EkQ{j;gTp#Z6hUA?y0*fzX-Tbg8bRuvuLMKicQ;fcUqg!Bj$9JS_l**ZkH6 zA_VP$ghkNKfz!<#y>)r#Meuz{TKadSlzAAGS$Q&$4tPWb;bl_T=<4h=n_0gzu{lPu z2?BaXK;d9!-G3{zTK#8WdO~rx_v{6HC(OJu+P@o$Ijz%c=r_a#$e&U4spVu z$2gB9l=+^%x)7&4^&jq;rj zd8w@r)Dj+5zc%l6L>|)X9Cd=mx3<>zqOIEAemnqm^%vvCs6!==%qtop>boJPq-3O| ztS`tTIYspATMj`5CH%M3Dvo7&l1L={%Rdn93+*X7Rl+|IZEvzn5Ziz_Eo7}=h5g2C5L1654IXk?4O=ku+f8zmI9SlH?KHR zumdQFqDC<-F$Ak=p2vPptJFnzHkg|TJ)b;5`=hHe!3ybC`b2Uav5YUv(HBD#QyC={ zrG;0;YtP`9kIPu#{^5jmDzM{Ag~Q}0nd_B|fBW(H*Mag$R9Hg)4cP+zbLOEhh*G

S!8eEI`%-Y@N&DG=7`U@Edyew{rn6WPX_lsg}hDjc9Km!$XbsLs8lk8Vl z{|fG>r8)FABFb2ScJI9A>%O98wIJ*naP510rMi4~1ndN7Q|_GMmz>qZ=kbE@!AhbLsfZJr(hoTEvJAX9k$*Z;r)+ccY$@Dr2BW2&Ri?1bc2h- zU34WU&1})|MgM}^4Zv9jcbhelpLhNBH7JI{_i8wf1dh>}jvV=!oI(e=!XU-(+Fa+O z!1-ij#S+J4K~Gm@*dPPq&M{@#_{e#&$%F-^NA4$+gAHINC2l@c*uAOV0#_dq{Ea(x z`H>zxK*dnsqK79!Z&bC?y5IvG`hl>`MptzR%{m2W$h>n4Ui3mNC+9>>DIZN9!@9K!9?A~deFZR5#02jPbjbMR6 zTp9gxXC~@xnI)mLb#T!saP5qVwYh_O<*aDCFAQ)wURNZ!$^ota%V|(>NBHQfhu4W0 zP_wd*l>=kG!VCO?y11lQgv=|rTtwChuO9*^P6|}216PIOGmnE{&aeA97E+F zfC)R-y4u5E=M1MuN!Hh+v$M%8N#HZ9cXpUf0JBgzmwFchl-DXysI#*EVEIFcVOWAx zU-;lWDFA*ux41Aml)Kd@{AT`mHFY7_9k?DvIKSNAtkU$s*WopQ701{#bBi6d-UM8J zMKwG8O_Q*6p7+9TU+{BUuq$r9Z2yv!zFnI%UI8*>MsG{|Zm8ba5Ao3Eka+jzZh>F1 zz)r!gT1yEJ1d{q3`A?zEopB!z{I3&R_u;eOpq_TFFWUWrtz~y{(noN9EjvF-^Cvch zo(;UViV_vhTApO8#&(S)5zkDujw1^gBtOzS*%+oQyoq-~pLwtBvlD$3;aXfI1fCd4 zCh^Ds21TYrcS-QHiINNMT%V`_Am%+!#Bzns##2H!d7@roY4Y@JSUeo#GJ=nx3&7jv z(y8iB-KS}_?Y>61HJ!v&s94JEOE12+r#sMq;m&+ad*tI@stB&E;)Fo+uZ7Yh0mW~> zu359q9Iac@TyWbC%L&Ta?S~=^fh!mf7vxsC#5cT4Zz?01{-bF{$n`-Znfo_BlA2XB zfWS}J0_-K~BrFi_oq1wV z>i+Vl?xrZ%&KYEuNjK)Dp#_|1E)LD7qTVN~6EZ9hF5!M!UCvLaN-4gf{-~wN(m!;! z3rwE8qg>ZZ@WA{{1Ah6~YGkdR-7oMR8RKXQAc&Yk2qLo(9=L}$zsc&9ZPpJ%Tar30 z?*Gg5)QCJk*Z%ktMLklMc}sAD+I@YkHz2(8IrVE;%rsMB{Vyc4$4=K=i4YsU2YP@* zR6!Ts>wZg0cq{ZCukR%*PIkUCY*UzD{iVSXQoVMF+2tMHp$yT8&ij=^Pqo^2B;+$U zbp=5E&c*MBfo%1FB{I1X4a}4yCeZAMq$M?2uAL-fi{A5uw@R2gLN&X8FLL1X{YvLZ^x4GZMq>Dn>rYWRw$HEGk`tFF+3GoJ zB4hOnj)0h~BsNgyl%u^AunxQ4n$%;(DfZl-PjSD=I@-OhQljw{am;=!yp=sj{WWD!G6OWUpjrUeO0 z7BWgI*mTa*ho%2KI=YzIY8?guSrWif_V(tQg&v8o-pVYKsRDC<6Kd}H1A7e272rtr1=l-aR(+d&k7S-H1W?o9iX*_T{awQ>Iv}gqCMZgZLVErr1w*LD1&ET;^tmTws;asBN5bCdy-ZjcOlHDqa9Rek-wB6W@984SAO6|gW zF#ww3sl&(=)60tI0C{*`&NVWPA}wp}tHpgjwH+&6*wMIp9HdpdCk2p(C>w#{@QgD} zux(RP)Gq=0Z2z8L_+%h1jgT8S$f$+f4IZ`_8U!Q&+ck{r_!!BgR=9!dZ=_}>Zf@{LK+q=cP|$589&v8j`e5iZx`IHU%b+;A+7E-; z4<;QRWQ>^A;uF6Rjy+#FC=B9$1y4O!^qm6;MbUC?-kzzo)x?kzXu_c9vQFJEXr^TI zc|2w6K|1QKLcf!9xZ=1vfw5}bT+hG~?(^~W=-D5$8F8A0pYsP+f~?Ip+5mrX4?@Fj z!OS#Xq5nfrTiWE2DecVZ-!VWWC-V|Q7-Nw$fV%sYJS2Q&*!0fdk1p%CtFRPp-Q8fZ|iNfs=5A36!tX1RyWY*0hrI%R+^- zWObbJYmW!`1QVpYhML^U^*Y%7-qg9B%Xh69$LT22#%4(I%kN zeb*S;P^{y=r)`*FxB7SNc)>O39SQC;I)fbhLuK!J7$Zm(pmb5r3yO*_2(pf86UM4L zCFR=_d>3-P^RekBYb={lw;P3eaUgFnRw_N)o{oHl-PvKKK5o#XN;n68OE3=R`<^8I zM!gcsMCEcFt`1LerrKMcb!ZV&IKpG1FlOgZG^t!AZ$w8pV(Dg6an}cy~?jCieZ5RS}T{om)DLc zu1+x5mDh-%0C?$ZtM(Zsg7Gsu$RH1>%-XqVWB)Tq!8fLaKGEd8lLA3&*7Ps#grAB- zU7cg!dta}!P`hP3gqO&21AH}xNa~WP1nYZ&zzzloX>?&fytr}eb5?Fiq2KfXa22M$ z-)bld>13(2sio%PY!$O933dJ_>m)-ZAV8ONFuWlPlmjHFes-2E)yyZrFTO1?m=Nw{Wrf`~+pH}B6<-Qp$j{tXz5m@?0OVt0*O&sx zq2@Wl{nOy{OB4(ME5DJP|AfQASM87%?2-fA+LSNULjrRkwt8N&ks2}8;#Rnz%%}=L z2`iL}$QrV10n8x`PvzbezyxP0?hIceD+m9u327ANWoK082^T>!$IRZa*hY%>l0KDe zX1_rmpzomayu+W}O$)$iU`9~M1M(4yC#$xIr|KCIcpVgpYndFSKbp?}IJyb4Gw#Ix zs{c*3Z}}V5@ccj$s&HZxYvdF?ZlAffllXc5goqNhGquHT{>6Fce%)5;*uXIgMC&!` z3w&3|sf4$*@LEp~aBz&<^&_zukWrqG9NRKBQfuyuvFuPEA6YtxU{HPp|H!H>fj%P>pYIN5-M<@G` zRIpD*6w30A=T=&56td0o<1w>+>O%3AW6aW@Du((7)AjHh#n)NPXXYdUl4=tTz!~tr z(@GA#uULHOl+N?5c(A*TeBFJ*#Sc014DD$(aan>E_THA6mIGM$!q`?g3#~eyKkeDh zRUf&vH;hu2nWS3DYWd(8OI{P_dQ~_H=aALH1yn>mjciuBqT$diN%G}2c*%bK#x=_L-kRK< z!;{&8yGImbA&hQDZ2OQx?(A_dK5e3oP+T@?v($KKMLu&8qoIxjE%%PB&9@|H%0Y8F zozdxreKa4T+j1@|NkV<`Q>8y|=H1*355Hw6T^jFx7iOgDf8zT?V9QUbM;cGtT6&l; z*_QWcbdkTz951KQzEj_4p;S!orQDps7ZyZc%q!2jI2ndu6%V?PSwBK1VAU_ zigy=xh@o%jJ4|gpCl{KZdtZcuuP0rm+t}UBVk{mATBRnM_PtVcl2yu#rb{bp+bU=u z?3)ORE-uN!qjqGL2$rM{3H8jU2}UwzQlF}>^WWB*ntOuzoz0d#SyT&Y3aO~bxe8exxps+)HJ?-|;VDKxiN)!_HxxWX5#{dChF;@q*<{`)p0+MjY2 zVMqk}=>qjB4XIWg+(w~N3KhEYvVc_exF8~cblJEh0)y1bP?u%RZjczebS-P<$B=%* zV^2Hc2 zcd59K#*XFhe@Mhse;#e7&d~8x;2mowL%$wOliv5Wo{2ibHhlbWvlqpRyI6}E!n?`B zsx<=ALonYOKdc_`t$L`O+A4zCWlSsH+w0x(!8j1dv6vZ5zH42^PQ&NZEcs~566>wl zs+nMM-3i+uyT#Xcmb2*G`pNrx1z874*y)PBMX(-Ejj*7LmdjjTZ+2dri;>{cRco}J z=W|k)XYr*9*v+BJkHf~zey{a0o@U0^ZlzS`D#_ikBL872xVr81zR9-o924WF)!`D8 zfbWo4nMOO-{7%HC&+0L=u$Bm5BwW)v|85|!4ECpl|^v4`PDm_ThzaCY+Vj-cj@c;tSeEdK2Fcd zmJfKt@=fa?Cr2;YuPjKpN)nev_uDB5J~;nDu|J$mRD4=psJi!QNU$fTxC+e-@rq%y zG)KK;3ElQkCgPo&J4p?X&gezesr$wqqGDf2Tr#(mfA(qRRSw&Dsy1SM^digt4= z;*tDirpC0Wa%KwM0rIcyz>g%P0In8`n>qW1Jvf_wvXi}$7tkV4@W)9@LQ%>297=qw zfq>ERLaT@wum(;gR$iG9LrJgtBXC+-uVsI_D?s&dMPSh&#^Z1M4OukcR(f^&v1?{< zc>d45dbc&?tIuW&{Afo(;FL`BJ3{;6`Pt;5oU)u1#`j31S=a4)O;`DlQ6=U8^2hBN zl&gVL89d)`{pitus_m_9x1EcS)*t<^hmbzp)mc@J9sNHZGRY8i()MXPK_KI@D_WEf z;Xc>z$+fcSS{UApFYip}dE+UzFgoRFhgYZLe9!{qCx|PzEmQ;^uJ8W@}g>+fSvw)wabP(;2sMfTx zVO|j9zpymXH|q?MX^%MMFRfL|#+=Ab6k8uofB!5Y`dRtE77IH!s?MG1{Lg#3NdsS7 zIz6qLkZwOb+BH7lgNS#X9N6M|P2r3MADTml72iBCGy;*k=o$jABMZ?7#b-VKH4Za0 zC|rL_`-2N~hbHeoc-|<$M$0`A>J4f)Hivb*D8*qfC{o$K#*YRBS>NCK*y?(p8#$k_yn)w*YY>C{(_UU5q z$$saeNjv&38hF4E3h;p8_u3w^e@%`Nc(YMeXHUcbtP40&y@9Lf>V0J=nnP0)8=VmO zvseE8f+O(SX{T^!>wFN|uoVLe%9=_#b$b5w*|S9BfbXkPrDaXO1-^er_VbqOc?*C( zfiJ`WA9BT%fKA<=)dL5pNAP9Ts^ovErx4<7ljK)OM}dIBxkLR@6!?$7c9H=fc5>$Z zHU0o6bcb~vmgTJU|JW-LGO$Hr5C|9r;84Uw>*dSO97u zB@k15wywW+4t@a`0Qg%RPV*PQ&OKgwUml6B(yrXuAM=9Yfwy0^K)YFa+-!=bF&=fb zgqVV5`ZD9ubGTXq-s}_a{k_$HDGn8)xEMFB%@2)BEx$_kT^CTG{EqBChm$(PM8iQn zA@dU~CiAHSM%QrJa`e1sB4{WcEflCM{XNL4f5qb*E%PE zK|R*{p5?c6r~jXV#)3bz6$~Y~Kj%Z#d;Ef417WHe5wO7}STd2rzpIxELEBZ^#6*xj zXWhT?t49pj5sRAlh_NlOqdv9sH>7aDj!1}FsK(2se(YX*eJ`dC<3`~8qB?9gF&)7) z>kozKdhFJei*7dArdB)`zc)JHb%k(a|NE-`W^5M{zc0!!6aR)g3qakPJK)QK{|XAB zfZ1GXul)of?%hxA_H$N3hy*bm;Z(lYSqR}t)%jU|RIOv|kTXKQ`@)03RidUMZ%q%w zDAAM!)AYW;nAiU;f#I|TXDx#Ix1&yMIveCi!fI^Pm2_cXw+f`jqaDva86wdhl@wy5 zR+ji5Q+r1Sh?e%n>C!mx@FtgELS2+n?b%J<)-SqohdYB;M5>F}Br@Dq36A*|*#TzCy<-nS^qT3fC}1F>2v)n`{~tw1w=bcc;1;-x{~A>!1Oik+RC$4j zHNuUvJ1O&_39MTRMJWlTaX)zL-%{e^+hAnU*w05y8<LozBCYQ#wH$Ho!9~!7sD5V=7814B_7-uWnnRl%$da{|%?8hYJH!obO33_P%k0kg z1gI|l@nuu;Rc;wg8E@gLK+(DDd(h;@*)7luK%f-s~LhTBQp)8%!f5ur(6@kx37-wzVp^H^wVKkn=u22PC`9`ij12eui6MB8d!%hdga9>n8+9Re%V ze}OFI?e44LSJ0*_=g!a(xXuCS4o7v)8x)=8tMemZbE9YmDK*P1214aLTTH9UbHiVo zYD}pH)I9!?@OTJ{y)bZIzqtWOVEF8z*Nb`Kg{h!h_g#y|x=pDL_q8AN1-Cb6; zELyXYe2BkuTghzOggdHd)IU*@C?JSzmHXFQrE_aAen92% zt!MARU@IP=)riBCVDzi;FANg&-9-Yy3=L2c>Fa6L-5;scEL&;1@g;D(Q);Hv+NoD7 z5!6;|q77>1xECGrcsscnlxVb68OM;xps@!vs>4}ccAyd^D1Dtq-|l(1SozjtyP#Q; zil|mK|3;7HIaHew{e7{!CUUm+b2t|Y=6<$q9j&Y6Gn>AV8G$J-{dGZ_+l|(LA{3hl zLOt>u5!)?lW;n$_!f@HsV_h zc^R>Hs@n?XTH0R?rgSJeiU?tgYhjAtn^Wxs=)aDvDf-A6&wyEiRzMTNxClFKb?T8N z&&uBiys=C2O(PSK4;c_GNBf@R2CVoun_Sa5*+D})QcC7j)N_+Wdsk~TU2-Ct7|jNN zfsXgnoWkR_l3E>wNZ#>$+h+mtHK2^ULU$lExmx%oufYT{)Sjee6=#^MnD$ENwr=Bf z3xD5*O7rE+A;jr?5RE!>W&4b$Y75O8ZTN7qA7ITF-X8IS@I&;Xeqe#)`K z3WcCQVh7?p?MxI(wPTeBKTJigMJaee?7f;hVYp$#4x6u`#Rcpmk7;8xqtf?CWOXwf zQY!P89Sd8-{D-WBnsxpO3IddX`Halor1nUBJMi{pSH931#)Ka1epqF7Na!}UDEPJZ z=leuJ7x}wO1&e?QN0w853C71W^6dBgdc~8u26?{W+OUkoJOMP}EcA~#pbiQnT-ai_ zCdzjk2#nxA1Soe6MRGsZSusIQw2>lRw1lpUFjDRTt0}mp7^9nwllixtV9y1=AaKJa zxat6?rp%i>3!j52uM^Aww^mYEHBqf2d#q->(%1X+D#iA4yP6YC67Nvvc#niV6!w5f zCHLq&J;%b4i%?3Y4@)ItkI!kg9RuPQ7@U>rj))LVX}`H9mR?0b)1WzmSC@rCNq{Nz z&GszPSVt%pCPw2Bmj1NPE@lxM#?o|3fD5{!)d*lce1r2n<%yAntK@tU!?$54lk$P* zi$JG%Dx~10na@go&N+y|qH|J%Z{m_37HQ06ifL1rtR|D&VJd|b>JjI@!|RpNu1QIJ zwq2)~1XZUQy1G(xrVnSJw`S{@3)o@oIo$eN^ax+p>N#pE^49Vt(YoCp&Xs;o9(gv? zEva_f*e$y!Q;-i3u-_ORxMnfK;AGxyNfuRS1{eKW>t2~WEoEex*_TbkXzS`B)nSEo zzAZ2&u&)W3_;G$)gQ(_>4ho(<8HqD99(&!0oE^=AW6eEYREn-%k6^;Q#;}nQG0ivb z7j(8lv%&)DK?X$U&i|S^6gm(RHC@J?erlVn05Q1n8Eh-O*gUeoo2`oR9>>#Ba$b!$^~pd*tSX3RM`^I_wzK=ZhN{q%jo(1Ai6> z6=l$az_1q|L0#31t8Sfil0 z6;&)nX3OuR1gRy9mCr98ghtcMa2cgQ=?*)TqpKk+WECr`Ph&=BIcnv6!KDSh4Y~NqA%OO%0I~Zmy)&Ujj=w;;~g~0)l ze>UJMi3Zp`3^x?yXHpU%8$HG-P@p=EU21KD#n<(|-i&M@A*W^6Y9+=u8r0A+?_NIJ z%r_H=^W(%-9~AUsbgU7`pxpIek_w&Le8m3ox!Fv$8P{umS9uw$_5pbj+6 zhNp;W>Q;T&VnElQMjhzd1Or$zr!{|6c-TGL^+YcQDV_l3>hK>%f^FD3REDU&B>&{l zXcDNpx6!T0%`jUON>@Vqc87vI(yQ-baH8(0qvN|7>cgY*E78@)^1RnQzM*^S9DL)I zqIGHta=?HoF6zeGL0UAowdDeVIEPjF?ItvF*!?Lx4b+JR9kEGU?7K%I9l&dBjnONL zOXoLLnOB+|pH)>gIyn+FzD^U#FADfA`JM?E)jtxteEh{~-wO1r_6y#_ePJ@ z(z<#n9Lg=lOl20oX@g!Q;jvSfjDTVt_|CjZIE+)CQiJ{Tl-Yt8^P9Jho#F;50>~%+ zO6Bh*C`!nk#uQf7FpT{FeEwS>x5rTcQ8BLbIF%F%VwuuPlB473vw`gt4sG2+8kA{2 zU{AZH)Gs%QgErP3!nUEr8tt6C5k>E2zYpr>F?0-gC{n*xEB5SQaO&});^R@qypnva zb~~`$!yqv7qW)cdp!#|!iTg$UD-*&ohS%y2?}EjqZ?Vjsfe3F_rHj+>5!j#8zSq*T z=z;nQvOiI+s|BJk&<|Gv<*#0w)D;5=(OzFJsRs-$Fi4aAncf5hQQ>2nzAKOC6_070 znxe)ttf?qPSHl2R`-iPjM1EkiVc2-(h#7zvp;qtwtLasw4F!5aVszQF3>{zwB!M96 zbb7TIO#+EMxn&i}26N~s=10l(+M6T@szGt)k$Q?53P%g6byj5G8KTd^i1$}rcIR{E zvoBUzZlk@Tq*O?F&_G%=dXi%%C?6AE`inh8kQj1lMhfzLbc^$wVE;fle{!UpV4y1f z6Qj)JXcO+7ej#H(xtwAI6ruJ=$T_#yD)J2Xad&35{!JUE}k&~59}bNhHdchumYhWS1*;(*DoQl z$}!?`u;eMx2q-@J3G} z3h>~vc$L_kZiCI*j zSv!Phj@kcP=d+1HV$-tBAi_@}JA-$|KWRR65R5}t?7H*@V&|w*i8wtK@P1Ld-x8Oj zBsQi}E}4&Ik; zZDPN?_Iv5dO(e6>soM||Udmo@wgUi+hn-YZf z4x@liC%Vd$D}^M)0QNqNavZ0^-q;44&Ah|ZK(`b7TsumVX`WIqy)+xv8Mg^$1Bxi&wm8lCUF)qG(L5O1k-FB8d_K7kKnNhE$S0Jo6)2r&XG2* zSffEk5~@Jqho_lWV??2gz|WgG#tXMG``B^%5jyoKWSrcOI5ssUrlf|kD1LFAJh2)G z{m2b3O|r{>3g)|Xgu~?tfpn3z?#h=yxAe+0kBbDq?Tun+MZY#$si2 z1=Jkq&2^9EN$?9>P6A1B0)de&vU==fvM=RbH}MIqqKtUf&KM$!Q2^?hOZ0idY(gpg zU*s>q2xw1!Xb1|`%t=G4`788O3>JRqs(}ZL6{9GPo==_qh*EnFc(4qf9cX2_L54cl zQzmg;V{Y`BUA2^p#=LBB9P8vn@`(38te|DU;n94FRFSJVY>p|CVAj=l-fRZldG!C@ z1rRgaE9HW_vg?BMWGlcyQv24eL3k}vz*j5MGw~2LPOf=;UVLHL%ZBkv^MiI}zA$~J zZY9Hg;u9P;N+GQWF5nW4h@!s#Sg?UNB;T8T1(2(er5LGr%rK;!kLu2PzJP413|wiov|~13iOq-J0BtwB+ZtJX2m45#1C|NJ z)}i%Z^T3CgM}upN-2yp~;M;WZ?^7t1XrQZpJz&5G6;LY|&E!^*ldw}H%lCy@M~F`3 z<9K?_zBxY8Xy!*M?d5SIp`5LI@A8FwzH|m%Jf{w!KAah&WRE|_Kup7AM2v7K+@OYd zdP>d(S%y%;v53K!eyk|9Mzxo|A(-*3X8LA!Zjlh+#$HT zyK8WFn#Mi2y9al-;BHOh?(R;4TW}}1hOnD_|DL0{W==Td;O$zgo_gx8S{(xJ7(wyM zDXu)?Eg+UTI%pn!0>YHoQ+L}e@&Oed-Kd)P+%kU#7Gh|$W2P*FQC7!MjWZ{QVCgS% zc0&`(C5Ih^r(Z}q>3-fM>^B)#RfIMg={Xik#f`d@tfxYfoo%q-i_Dw-EIx4mzZ`t` zuSfzXqzVRbpwS9j(M4!1kTa7DmP_fkF`yL+92t3ce*z%@!Z5(9-V|#;GW4S!NWx&Z zwaIU_P5IG8w@WCCcmDe^2_Us7M8j-dTvr@G43XcDk76To9Tub z3_s!a*D2jLbJ3JZZh0xOfl8I2fy&E-OI!(ca5gmkZ20M)Y{y^jwBZTi-Ylw8S1^U; zhd}OQ`C+$Rdh%Ma#i0f@c!^3A#zBA-9Kc2$myB>YI%UY29A|07ZJ>go$a!CaqNL|C zNZO{J3OEQKD_rQ^Lu1uPZ;7*_@@mDs#${E6rK@%wLuyz%?YUAdAA8jOWryp!(&^37 zA;=QM{x2|*A;wASG700e{ZPkYFbU^pP#rIbBO`Lu1I7s>0gBN{rVsO^|5Nm`mBvuz z0xHC?nXhSUBh&l{OTVM97dC1BxeBJ5!^_rJ@QK#G83kgEF2FHJeNpK5Q@~ju3DFrb za@px=wX$8qVDf*YA>905Qe47@i1||lrnruGu*tsD%cSm6^$6(CNDbo*tx#)K zf65tYt-C6hssKY(@6kw2sAU>C3qBvn;n5y+e_3tJBF?0Gb9^!Ml>S*T0nm~V4Y5}y zGp?VBE{0LauB4r~3VRYl9r$rpW*h60LW9;XtL#{j8{K>Iamu4eAx|J4X91_pnjRKV zkw0&>Tx&|Pncnw@kQFjt6!qgX_#1ZQck0i-#=FoW`2@E(rZw1&QH-FhU%dbKsd%6u zpNimvK(`Ciztk!h-!X#_B9gww>{uf3t^d7fn4}sKg<|@N@P7I6sx-aMN!Fy?;F>xa z^dm=ffSCXTJRl)y@5knkCZ7Hx!xRtvlL1~UbWW;4x|ZCSpGIY|5QVMuNJOCuzRA`^ zX5UTg!=Qs|K{_Y2iBf^c`2Rjg|FIAaHv zNHr}UZCc7PjIH=dA;letHE3%a%fw@+_ee$U=X>PlSOnyQk_=>4A1tM@l&mNKrsT=g z#IVR86~nq4sNH*_g0F0on=0L40b1rGOI;4*UKXQj@YzbdlHaXX>3m<2*uKB~ziG2# z`8lTU?xieqKsG{e(d|{h`%Sz45oR|aoXBAxf|RXIGKRX;~~(U7BHy^9U`3= zKn~~UR#eq>{UK7uuX%hn5@li)5!#NWR~#5ARXwf>qieTFD!B<4wplKwz(;MNyb zSLW6)Bt$D}HUrn0$s8tYi)vLt-Jj>D>zI)rIR#zC#jQ4!fRs%0^{5ZAU=S1RX64Th zd8XyA_(h>;0;(bmt;oS@uEsD>gHJBOAX_dbYhor@Ty#0P^>z9OmoI zJVCKNGz3w|5f1!zWBT7A<9a*7B{JH`b8ZobHJ`g#6HF6#YwUNW(12oYdEB%`9CLra zETEzXGmDZ@gMS!zjZ;RG03<$Yo_`c~P>t(wnf!8olFU|jT7m^EHIlR4a^_; zYV~m?A9QE2GB9$*TYV<873o#v(lk97_Ya1s>40TuN2=>!?C#)Xe$Fw_FY4T{5A5M+ zG05?zoO)*H&(io%J5Q!w8U=;|W|Mvwa@{EP4- zvU%2&Zp^SJ$oTJ$lQ?p(AKzy=zDg%apWuW#R_(|UGA$6_8luro$V}JW{7m=XT+!Dd z_WUkKO7-~&_J*m@tqAVWvDW60p~q^Z`T1+Q9P#w`eew#xdO+aC3Yvdx!KB~u7ychz zE;Q>)OR#5d!@9Stjw`B#-N+bJkJ^8U_~qLQt*N2tPriZB0Q8k#fjT<%&jD=msU|MI zU+^I%D{6esM9y1heOlR5{C|NY@a9nbAYpubIyL9VEl>$A;3oP27s`@!Ifu{8G+@%} z3l;F-*=r?Fc9-#5cDl zwvF83C#-zXGo@YX!v|5xZClIifeIWLqOc{5LE>Og?}%WwM;aWvnkc-YIRew`FGPhH zkXEhhF0$h}nUf~RPL|H2*m7XXmD(4M61vn?MbknL2iI?(gs z^!;j%6~m(bw_VD-*UDvCub*Ptyv!{c01_GO16vhOJMVI-ES9}~6F2N|Kve|0YAdC} zZWu@}Db^%LJ#zGc7_eUTi4%Lio1Xp*TZ)Q z^mQT*DM+JQSj90`b)^qyW~d4kJkhChl%($JhoocuCfV8bLj?n25ln}?j5v3?2(h<% zrymuEZv7w;-IAa?m1CG9UqpMswOslpIaZSQC5%DD?%*FVKsiF(sp}FsvW2((P zZZA5#9{_nN5V2{`LK9@K|8nt|T@s9j&yi@?`Gdn~&`pCM^BDf=`=HoFH^Wx!`j~RC zBnc9o-X*?g_@P~B?=5p!?|e7x`iWAg!*f*rN-Hj2P?MytW42zAbx%FISISqM1x-<& zzap@uL=FA{EC;{Tkda`~h2IZdN!;nq`eKNuTPQUnhi&WI?8o{Ox)cc|@Q z9gzftsmw2| z86KPJmCakZO+)`@b`_Qm#!9f?s(u za|DjkwfBl6vXvICsBO;y@3${bYpt7Gsl1uaQiF5pu>mx#*WRlmoR-3Otatx|9#Nuy` zfXf)~D7Yf5SC`MP^@ZN~nWg&aJ@Gg6YYOD;XgB}?-Ueb&%z^0r*bgIzy$I#bOPjLB z>fo`wLJo9{=%o_OZ9Xbq#K{@^9JnAgl(6IU=8&i5=FpXhXUbqgH6q+N@ZA6r+6Y*^ z@7oPck}ji8AI1t@yD zgp1PX_{^V73vzXdZ~)%!8`nue$dc3fP$GhWJ&|wreqI36zV0*gX)JHHZ%~dm!L#nP zue!@Gdq`wB*RXVz3GYG~hA_f!Jn!PICg@A>f5b_RuxFF!gcXV#YT3jW4lG?mnqzMk zU&Wn%hM9Pt^m*9XWa0oQmao2dr)}F=MK|1fpFMDXLN+`Q>wHaT|3lTh_N;WpCv{qQ zN@y&OIyUEhYB?Ak7{mri8d-<%A4v3q{pKShrSA!nlCoLWS_r%rUsajjy^`J731NXw zro$xN$vGwz<=pgA_nXWXcoqCYaOSHUcd3-GwX(g$%Je-rekxfF%&%Cxb`gEzY$+V0 zbVegGD@+F$tpLz}v2CYl2{m}n25JmScFu#7NQo2<5oOO%!B1^rPZ@>uo%quC`r2v? zO=^bxjsq?><6f#YsvFUDQO>~1jHrqz)fYI0|&T64{uQCF?%L%{g)r>VG>_+AUaH8dWxK68St;UkM1C)C|G|fBx z&VaBL_rXs`%@C0?EV{EBl;3@YCx=anq+^#C!ls=@Ru9)`KRo0)Qzd}_aN!!@|5TU> zJk~l8#SositY1A@@_v7{hhyE|t)ANOndcMpbGuWAOs}kGFScF{UURKI<69lf__f`u z@}D65&i8hOlgW(R{W9c02U36hq5D@RYb;75E6R~YU^(_#*ZPxhD)O@mIO1sKuaJ;5 zz3?j>2rUyHumqk(l5)xLv0V0m0(q7x7Qn zdZ6AKyc3kkTDVBoF4xd9uoP)82l}I`b|Fr?U6EU+MlapE!J}I(JRB^!u$9^ZgkEqm zYl;v~xm(3j5I%xUj(+z$f{F+>=u}Ijdv&ZP0V~{G(F^Dxy?VzYgYC@3n6>EW{XQvs z8gMs9nM0{M;pGxbK2WpBtrIm!iy{@-Q?NoJ1rIV8`tfE{OmpDqw-k%K9qgAFi*L$} z5O%_vj~w}F+ne_sCT%OtA0X-kw*=-^s0qYw&{lB#u*`tbH(@j;0IIEF2~Ax#;R+m; z)NoXHzDR`oVDMWK2kcLKQzb<7gbo66cVpt6Rpisj|FCdBZ7TDP7{?_}j_T0sh$RWT zT;yInIt~@ra?$Jd<8p};#LU3fC+=H@{M-@V90&83wG%uC_LxKJOl*bU0MhUvF|F0l zj6h1wbNcjeIIVb(a728HckS9<`HNE(Vy}R_C5m-?wAcac*0394ISD)-BhGAR{4UT~2 zN{`S?780d?G3NNA9WIW3yr>T5I@0=fcYc>DVeDi2No76$Y9eRq0nJ^Z_X$difQffX zytoQ;Usxy^VhyoQ>ibLeo2W?>UD3e+P0dPb%CnPRG@uQXhvJp)paKmFvd^n~;~^=l zNLgTyIec>tiqnLZ)b4i;%}ZUoumJtMn_FW;mKl{+hdp?qZ34du6nEWenMr6BOBzaj zmy6E1z90+CtJP#$9gl#z>4RAsnN1pk0ExOU@G!#|y3%j=JT|&`^-tZmBpy^hlOBTx zz{j6(T;;<4jd;3Hj+PcQIr=PB(+N&-3$nMiTk{S0`dSiITe6UUIP3_c9$?g_p)jxSCw` z_Q`dQ8fRFvK$X%ANzLAQRIo+Ttbi&VZ;s~oA9&5**6pOCg6VDV=N~z{Ov1@%2TOjf z1zk~&D1HuhA)%i+HYRVf2=yof0uw9QKaceouVj~Bx4a(Y*}JL z1d#H1ooZ)NQWXifN*u~6>BCWbZHZlLOBSm8cME@Y?S)j8^@G$o3`_C}l`5s!=S;%) zw4DwLX0p!`$rf3X1F;DjKf9=`(1KdS+V$8>LJuzX9C+k_Xx?`Dv7kT)7C`-1_j`tJT3V=aVz3EE~jY z9QT#sdtz{a%RU+)>!wI$Ng90Pe2q3gRxIaK7zRCd|7fFkgD(7VO7C1k(U~41P`X^CWm@`EGmtm5*f;LvY2ypfdks zs&eN6VZPTB5@fTFWCZI|lwHz9MPp+U@gg=YFhS}nG$361Kop8zAm>|3iVhb&p{{3I z{09AnUBKWS?s92nTpild2D)TmifRyfBz?PKBQ)uz)`z$F2$Z~|`xhT9NdMg#pD(08 zk*!7y$m1oCjtn$EibyuZ6ea4({I1~{gC}RlyTTcj2>Y2yhdVTC&nsDnJ46MbWVOoj z#{@4a4SPIl`>-II#YcsVgtc4+{uJT*)ZBlufZab-k&Do8?GHL*4r`;@;9PBY6+%?( zp#6FNDB0@fk@PTVsRvUV??Gv@Cdopw5)HfacT6$eLJ79_iBFKGu-Ys3UIXCONys6t zJFI!|JO6U_qum!t626~$2-~I=T4@zwoF}bmQ-mz8-4kfv@u6;z7Kn2NdfSXTuT@)o zc8(mhk~|GI9z{Y;B|=SK%l?V4s6xzDNX*t=J`$CHy$#V0HVqMwP$Z9FM%KCFl27+37{8n4XSR z5~-Nn17ZJv`G;6ke_KKfl7G@#h{!9rxVR{Rqi%*w#<=)P2qu~OZctez@gi-q=cq$s zt$`^})G>C$se#^r!J|Gyc0@kz{);m1MHAF|UuqJ7qzmeZ z-SXdRoxi0&tW7PI@Pb2+%SPAKvCxJ=JcQ-!1{wyU>Si+u9zPJ*x3mu{*Xd`EYn26G zEy@VJS=8}L)uU5W24$1C6jw70DeYQF?Un-Yt%|Ywn$)8KZr5PxCACR`Xby!I zC#PMq_Zuzj3^_W%^>8J1p6s=BRXfk2rqIK|WKu4@V>1_VlyB}6rIk%J=wItng6zM_ z;oC<^x=^i=!i|0Zby_)5+Zl-k z?%`xC$AtTxY{2LD%T=+4EM^o95T2~Fdf&qf8C&c*CF^kpbb8y{bfD(6-H$`LZ7Gne0kijT=PRbQWCFk6UbS9Lr|NKhtS_4 zV!Z3%0GWjOqVK<&vp1C&JRyNkpyU>Sih5~zv^PGOa6+v(21KJSo2D%4(hC8p>mbg9q3Tq4ip1SvJIw+npHgjWw1cK-&# zBu=?G124M_S0GaoC2t&j1Z4ImDg5C-A||#c=dB}3o7h0nkCs5k^fF^q1+o!_*9zDA z-p@ub8E+bey7)M4p*U18;b;6^R2>i98I=Xcwx6cLOYp8?MX*|3-}ubN<=_X-BCTE~ zZuiQb1oQj80AT52yVE1}=6*_(qMKNyFTIO_ueRv?aR0B8HF29RKu22&mr?s=LLXk6! z;wE#JHPl>Z%7|4L8#|@x5Oa7Jw-5D$({a1NllEb|@jdS}0D0&DVUte(t4mmOzh;#` zosX9By|bneMwQ#{mHlN9n->P_^M`E;QNQ!%>Q z3~Ne^{$O;UXa`@dmrxe}<;umyvU_OQR2P4icB_F>w@EFFK-7&uJ?Jae9&L(htj-ff zk_!+2Nku;?%QU3PYIUy3O#fy*jz`undb^?0BtBF2M)HEWxcBsdcJFP2qx`N54O7Gs zj&nEnhrQE?9hTqEN)qGK%nk*RJJ1r_NLh>VKf_DF0&1aZX?{|qnC20Vg-ntkN$n#} zNM^%4x4e3*6qe_}zN85effhNGQubl@d=MYic@kCS*BGM0VxZjan0u+BwAFHkkXq5* zFX{?Wo3>d~zFeveE87fAjOWSA4KNEhk@(Sa8b#^xg2BN7`!tor!WdR*8^oap8iI0J_$~ePTFb%YCq|vLxDJJ? zCn@+hNcA=i>^=+ya*M1KexwzOKmK@83Z>xrk0siTjBJiU< z0F&Ebs-XFWyUapdu(SPbMrIxL>iT~cO3KL461>F6VNf7H{&=5%u5P-fs`ZlUMJfT*43;!CAF z-!>M?Yeu4m^E&|wX|E(oz4D_T)4cd6cnt9O`%kI`qplcewcgUZ1=ZdI`<$7m;JKH? zh1Tt}4%)h8aoQ?tme#sfsGmOk5U4A96+Nr=nbG@`V_cDM4vHD-Pv0+DUB`EG(1;3n(X@Y5mwtuUZd(qKFZuH$B*N(-3-GTT#u_sXBemqNZ-O z@QT!B^P-bAc@}R8co;R=v3J=n@|}MpR(IcYIov(q?*WuMJv3?_Yzk^H6S*QY0j9RG z$30D9qNv|B@r&ZeyvgJu-ndlNz|)i*TByivKXp1dXqlRoWJ5D&;1SHTVlN#ou`Vy0 z$Wm?1{9WLQxHp5pz_OJa?`M>35_TP?;`%gKw7`_TLOR?`_P~-^wF|mXx+5NyJTE;8HAA)sDZ1)Jb~u&4r2$u(I2T6M0FpO7wkXeY>P*KyzS!A~PRf*J z{m!r3%;T)K+GV8mn5x!b0_3)#dz1Qjz6An$Arr+@#ZQsyCc#+etApdXW?1~mRz#Jy zHrGDwrO$z!fwVp;oZBgIz-6I3^Dd}I4y&gBKhr^&KI%^ z3yamQvwBD-xvAJ*l!-Y9%g1D=HXw)i$+&FBcT&(*OLRUrv{e5AWvhbwl}&fi0;F$1 z&c*4?U8t?WddoMDor4V{APPM3_X?07!UeL9u88-2x7M4^Gilt}`g8no%k^LrW5Z*! zIkcXdXtn~~l_I`o%DUn^lx6wU4% zG}cyAUkt8=ZA6+6SEzhLgB=Wt-mB}-fH#|#rgfBG_6w#X(+;J5W|dhE8p}Gf;;M^u z4!2mSmCX7RB*yM6;LrxD(;!#1{t%8ftBSwGl0k-I5k4NHF__(O1KD z{)h{mU-a|E$??ma3k`3e07#HVPusxb*J54a!mMBycG2{LsG1E6iCp?}Z3EY($h6k= zhtH|6Od@tU5(%2rq<1OKHfwV*Axl&Z^#l)mg69c=jLQKozXe|ZTyHI#=OW#&yBm1y zCPt4!zOsLZ=P>dU#w_yh7HM?OZyz?l>JD zh4j+IcA{>HB8KRv+Sim!OJL$M%lRn3z&QfhJ+8=8}G|aZizo`Bo)UhAdYIios$IzLcSJ z@&5Teg$CK>^U=8d;<=mu-IDHONCKCJV~7ld3J<(uw<|98^8Rr z;Chdl!a+8#l42i{Gf;kxkPF|BEsXx@j6Q$49XF!{%5pbRK25)}kba=E!GvpSf$Xa{ zZrV>Gol#6S9(Evo5$&A&NlK>C3}I+vQt9qcb#bh$(;zjeZ5bO9g%Fwz6i#RwbgUEt zyVuIM%wwfBnY;`mNxudJ%b8_MmSmVXEBc$bP3&WA$hwbvAGJfysnAtCBUf+VvhsIa zX5LOpWesgrpt?a-`cXK17Wx4TiE_|((Ly-~zFCNcslLkLT1%9*2>O-~iPDFAGwnp6 zgt@ft2RA;)8?52$#qWk@Ovf3wL~79BpFn7Q?;<#U3g@0avx-@-Mb z>$LUs&}bt~E!(Y;U83pm8x1L)^|&~kG+dhBQHBftqFX=7cfTE-@q08e@_3I3GcPR5zQG$3S_ObqU`to8Ks~x*eR}*v?t2Y zc3eIzEtn~wd5>WxbC1b6ydI9cFOqlB>OBEnLBQ3OF7>Y9X$!@l=T9H=A5o9J6wSWN zo#B2Mp^JbE_t*1H&P1J0GLvcMi?()-=g#XJcK|WwkEdG31RhL+^*^x=A z*zh<5S<}lv98M=pJJ6S<=xzCQ`Jk7v(!LU1wnckiR_O0JAR>3wNAbVX)4CH;n74ML{KJj0x+Cs=Llw5XXUfAJi&=0L z_c^fh8e3FbAmSL|lQ?N98Sx?&&tw2*5-JER&X?AR9l}1VvZp~n{Bx^z5r~j& zf)X-R>>jcb!5vbuZerxCMxc1D>6R)3cYsDu}TIqI#kCC9${Zkqn)$_ zL)dS;GOvt(a@X$gf%pFbaNbI}$OekOM5?tj7miC19Abi@A&X?~vEq)46ibQAhaKnR zz!VKNV)Rs;m!5(8Ad-*7y%ApJ!#b`GY9Hng#C{J7V~u*SsdMYk?zY=@no@-#T8F>S z8fXcUMjE{|w|8TGbniuP#*D6-kLMtGv!f1KRCtNi=@ZjSDhzA^-6$n9bz=*mLw*sW}5> zAW!oRo?f3y=`Vy{ioEFKr=ejRav+f|AAz^bVX+uEW8BK4m>sf2^e<4atnjDBYa&I( z(HR-GiG=RxfyFnqSi|pQ>F-azO1lnjF@>F{Ob6dN6T@jNo6>h8VewtHf*$gWBcCmc zlK6{xnBSjfiznU;WxW>X-*=~9cRN4Hm>H;Erxw59I=7I4NRGv)qv{{wwz%0L#Ol? zUh>WJ{%~pGVqjXJ5t%kOUfR!f){&VN2*6+b6RTCIq zrkRO*2P@6?41f7b1oBp$Wa!)raH2;jtr;} z>BI#55&zJ`k2csJTN7xSF4todj8*i`DIUUsGFOtfO5bZxd|sDk@oM>ii}CaF1?x8g z`2ZX~9H5+<=n{hPZG~#}^)FWGA)p977`Aw-_0^uo+=%X$m*&9Rd-UT=-v6By>%Z4! z#yQ0wJ%LU*tHEX3^rP}QuKxS`Vq)AMz3QgwBvxt=Z=f`YY0EWGP|8K->2IdDYDIjP z!BgPs8kF``2E@UcP-!?vIaj9C@?Jcy^2<@w1XUUZ@?+d9$6}urundtjFIrz(F z;>oZn>x=gv?ljS<5@}{FIOzsF&(PtK!*v<2}j%Y2O=2(N5|lxz+(fY*&1)%Tbj^7WzOWY9aw_ zJoCj}?^cXZS2*8e`3O>AKR4;8d=1G|3JG%ao!qhXSbu`NB_9joS!-ntyaAQ2=sY}i zKwr$hZ+_cdC=1VBV}7SoE`vr)UqR%OWoUlYJf)|d4CqlL@kqe>^Uf(Rqzg+9rk0>z zq0-BpQ#czGg3OO=ST+y|B#G!-g5F_+5z$|PB5$36x0NaX_CY9r`bPjs%aJjd^hR#T zv00dT7!ydEpwX+33-#ND9tb7X9g^?3gP7izo4eeJO!J3}*_P$lIxV`M%uk12%nkRA z#gv*JR+wkSUPmspe4B+zAdrrZ9q6*ddBS@!rFf6Vp!By%>Ra04jgJXwF`g){y}Jb7 z&u%5N*&JJXn5`LGJ@~Dkt70bs_9%@zC`Wv-#uV_hVcSsjd(}isE=9r0ystzx&j_^H_J*-Sxb>0_z-kM5wwP%k=v{M$P?$#Olxp?D;$pu^ zoZx4n*T?DAm)=h2)1}&r?I`9IucjWEX$7wY%hklMoVzs=mLX#PvKCi(8*xrzqM zH>_paGiTrcrtw#Mk)Xab&4pz_4Hu55jJl*oDu%INDlN1M%Pa%E0YfvglVUekVaB14 z-gLL~t#@Bx^$Y%1ElfA}o2xc^quY?PsAWuXxTOp`QGw)CCi^@n7xT}Y2s%s2h5Kht zd|HKj_Mre;-tZA|eRB95jZ)8i#TDweJShAC0f)nr_MqT`10@S{mVcwMC6eXN>W*^t zfvg{gJ;BBuoqj~(H3dRq^UY+)Cgh2xUN?3=Of~|q)J6}-rDE1E^ms=NRbHbRtoS7` znehEWm$;4;93b>)(6%Gl2YO!DsB`wm%**9FZ`Y#6cB_$hp8t5s zQ1 zRGAQ8tPTuKS{1&_CnFPI^0kGcw)LsDD-71-S!6oqqw>=HkS$qV`)fFv%7~%`2}u*? zCQco4Af*$VC+@wbwyHhPe$)1bC{B#v&i2c<@g8wNi{KoWrKnIwOa|2tJ$84Lo1*KQ zELex>KU>6qjUHx+UuG{(>!(B2N)CzH2Tj;?tP~ZsglVJkUQ)Ek!;zr)5M?ueeDp{=yU^Hu5S|Sk3Y>^WKc_L&D+M8L zM831uhi%&k4L%?zGRL0am zjbt6dAk+q$nqq9vV&nlnpq^1WUG+8+j;{pX)za)@ctvmRFvs-#YMCt%m-`p)!@QA4R~QM1?-E3Iu=p=k)a<@NyOHpY z$DW7Jli_}&T`ao)B+Rtmd01HFfB>2UQrJ2u?Szq5Bc?}U;;^Hk z&5f~+Ytk`OwES{vh7pLPKyzT)0{P~l7GB3gsGl;|^Ms`JARY%;%5(-{Yig7`Y`=38 zRXanv*`VFxtViy~t+@iuwLQN+zqpeWxpX52l|YV#+>|04e*8{6kLCsxj7rgr0A9|u zv1TINC%^7JfG03L^ev z=<)izN$hv;?(fc#tKI5Qg#GSlb(B4+l=vA>uk2v0^4xiAC-+s@whLsvej<51IiDLb z=(D-g+_vVFmW0nN>ftF)=xJNGVjl{lD1g`XXzfq3=C7UPtEqGhEu92Y5COP^8+JB4 zsn5%rn{sD@)TtBP$NEo&Z$lz~X-qB&jY+z7us*N}Tm$y(4`!*~+uLPiUL}A%}>1B7?xXEuphS`5zaW&jb1CHmG zuh3F4v{7S$F6jYIpV4L)uhMX;onw|t0MEi-#!L=VGBS*f0YRDG6+VB?6h4c4{B)^n*?w%C}lp++zkrV~e zkZlv@LcvKi4e7^7*({!_Lm+SVWedzO=;sFJv~QWWa`~N4H60!T=zP|3!EVat(sJsA zWGP>j0z9n_wr~fy+w z+(yCt{hbH+Vmln9Dn%5+DcE%w<9)gGvZb@}M{KM2dJ`RBbr1|wJrq7{|3ULYpq5#l2Gagi{jM#UHHXpt?cGPBPxXOtqb!2`Jzz{-XN$e-`y7kHAN?mptu-OxuL_I_%m z6}lgt@!gshy*nI*-Hm}!p)F#~TYf%9nP6i}@CxgHN2EPka3Kl#wNOg|E9cBoZ4 zJpGk3diZ9{hDi?xF#LuDiY9kgA5|b6kBvL77|>u+HE&1;vqzZWg_Jz?{7BboEKRWP z+)tme>mqlF?Iw58WG+KQ2h&8sD@n--zzCHu zGY+K8_#H*C{5d8L?j4HCU+zT907&W*k7OGtVBFAoDPYW>W3s$NnCu2Nx=o0~NSv!dY@wXmY*}o7jt0SY~JL~lR(YOrZ5*jf zijfO5G9a&_%hhhkD;!Gh$Tjd>DGd@y9cgT}r>pc1Y>mp{7k za^^ZTPWqdshSD&77CnZI3OVDdeyUBTm%>d85z#Nyr>@lve~1(5yn@F*z7Mm7BLfUu_^XJ$?xJxr%i#=Vg53 zJf|btByF*xC@eE9-PJe={f;D#?u#gnz-w;buQBRAhZo)mv5YvGCJ8M43jZBdK9SaD zgrTiDT0IzOE&p}j|GNu7ETcg6qPQ!Xcc+9;2y^&&1GT%2lKFGXYtS}4x!yYk_NqZ- z%Q_NC$F*Q^f?N*`P{ubAN1OmkY@3!in4yZBRfYgmqFvS4_v;H%VB4!GZ0X>0Xa=Gj zuCqeS!*=V?#mLBqCM%%D~D^ z@AA+8kMh}5?1guPA=exqS(>LBY7UMN<{xlCdJ5h3|zj zPL8&SYC3`OJcW?Rl&7f4>obyH%Ixswi&RIl1H3l%9_5!F#{3cIGjkEF&#cAkv=bvq z^Q^?LeGo9+za}p4)&g+$`SSQgxkB0gv~$~q{pFm0omT{Q({&#W1NtMvn1c5`%K;f7 z*!RfE`U2t-;i9VI@up;70+k1vdmDnC*iH;I&4728)L#h0$k;erU4gy|7P?;UWx1X@ zEi3$3cJ^bxoA>*UGpetvsQzG8O@xHd$3z!>yLGeum=bZ8g$9HJj24Tb*YQl>(#9+L z9_I~S<}U(Je0rcNHXpM$ao4MVFyTLzKC>wZO#OHtXs?8s@Ms+k_B!@QQL-|f z4;`S_m#=f=<-8J|18O`QzqnN-L52?8-j;*-HP=C0 zY6wjYK?IUHmp5fT?fFEI+v!W}-|Ckx2oU z2>{1Oi}8kyb%6OmYthROp5J@saLBYrW4v!$h37hym4a6q%{g8#6eb-h$14;uUGWN( zKbM`{`A|!FPufLL;k(6rZU>M@Nte(!`uID{VX8#RfpUS8(vt$BK*{S?8cCV|!3kazsG7Lj`)dMZ_XK96~LVqdpyB&0TayMfGE|7x#+MjvGI?nJHvU=XG_a^&oSAFeUfF zq~Yi=ulS0FeXQB!9s9bPI@LP6>7ZNIqO2YdC2$rFJvK~a%r|ZO{QEgtE~L;VxBEsp zu>rvXZCtAB{=))E{%dWGVx%S*7AG*x0Ze~>6$C?}3bxgJXSHSR_f(OS?t5>IIPoX2 zcGNN2WCY$6*7)asfc`k76zJ$jkAs{t*cq7G(tLLCJ!=HAGI0cYx1^BB{e*i#Ov72a(uMQG@e@gd z!7^fhl~-A1xcZv6K$g#;Pgyp4726+5TAAVip!n9r2o!DwccxJE;lPLZfd50-TL#6| zHC@BO-QC??f(4hs9R`;{f(H*0+#$HTyF+kIaJS&@?(V#k>wdo9Uv<^g{5e%qXL|3w zdadrYuVsrc=UZS3QRSBMZ}-|ev2Hot&rA;W-D^A>I)ztHbI?elbO|kI3{^?zh_=C# zuHlHOG_NMYS6GIeOZ%prU%q`va#IqhLeN24l?X0=-7@6vsl?EN2gegU65 z-h+*ay%JaOQ>c@IHwuqY8;?{@{s5fc@AMLFZnQ8J^&^t+ zL19oUjsy{v4|l6AGw-&K33R?Rf^Vby#q=Bm` z@u=kvVu2bTN*ANIn<>wutEmf_QtLidO~yJ8h(cvDjnFQRma!FvRFs0hmjb7_j7o&y zt8ApS=vS>r=bizF;LAmdB)WGZV%opQ4%6AK2&6huDI(z->0wb-#vAv$-yYgQeJfst zP!C@slbSIykf@vQ-wC1HtKC`0%DYQ9J%(;23_)o#D?hJSv>9Qa|6Gj46_ZmxpdsSu zu7>piCJiXM?lhY7_?Ws5c$I|E+hX3&Y8{UCS11bkn!p;DPX3Mg!}%#gFcnhvpqiZjmWld5!>1zpnRs|4 zaW%u*2;ng)1L{=C@0d|$zn08bX#*4RtUTfvqnrF?B*~l5^1>egIsCPGmyhO?)y!IXP<=$=Z(U_)ZSdD~CIb)1Sq@M`HE%Q2UHq5#sUwG1 z{~Hq$&xDmenxOfS_ODdM=)Z;(%-MVL6%nl~Di26h4^^d!?>Gx952Ia)8mHhFPVgkm z&rm?_+p3DiFUE^1S{>QTDj)YI8vIM?oHS$1fv16|n6l>M^;yQ|yF6ITe@B5c@(ab> zHy*VpWqXDolyIdUoQ1frn3XcloK=a%HE7YF>@JU#eFa{3mEX-C52Zz{Hw+BZKi;}O z9(pI=PbZg!9A_6xUg!~k7rq}az1@#^A3f)I6Gbm}BJ>UmQ@x7#I=_Pz~9HyF|D9}~5NDCp>&Ik_PQKj|MCWIV7 z)lOadAJ5$$$8oh+P2D8RDeL`x26D>l?2*a^&qqeTi`xb?+}<-g+{!U4t#{cjm|bKF zoimYuUv>^GlK=UoC%3Due~9f}A0;cDvsJ zlL;LQy?R?98Zu&_!PcM%4kI5f=WhFiLR)cx#@G{LJiQ}T{bxMdE^hwKwlDID(bGp2YXfHi6kr#<)r<;v|e{W>P&Qm8RJqeBWKpiwI9s4 z8I%l^NPQq^?1H=o3>c<37|%BvUgTJqFnObQnh@)pbk9G3K0j{av&|Bnh0snWjn>aO zgDLSdO0El4*NYTsB=-29fYd>H2eQcTwoo%kkbUz*cC@69Yz}CJu7*QngNh5v)pRS& zBFhmxyX$f{U*keUw=p)0db-~iJ_Nir9=tc`9lH*EPwd~SF6Fy#<+o;N6fY|?0`3LN zW3NAwjGmG%d%-+!fd%^BuD&jwPS>L$c|^;uDHj{_F$y;jX~9On>%4!|>)CK06oF5P7NO<7i=5qPv{^tt=zgx~uE#S>cp$OE_#5r9IKe~4 zU?`9@BRlxlN1*%fRcCfiWXip|@YC(mzQ=&&z(Mwp2_L3#vTW^6MRr`b<-^2=v&r7tXw_+) z;aV2juDJ{njiq$ZXEj7U3`#?Hwl$bKwLWuPmh8P>DEC!mBhXcHsc3rHmf#Z4BwS5qE>zX*R2;WqmMDfL1A2y&{V5orYL=nNiNDuK zbbHu)1x?kHvW88qzzj;bYNU^L*n>(ya7*+tBE5nE(=N{h| z%~k+6ZCQ9_o9wTUtu+9d}JTa9&3iG-F%E_`k3&iRmdL;*f zO>v!0n7=N8<-oD;ce?*^yzH+3$8E{-$(djlR6d81a)#AKjn^j6r6Z5=L9i8A4^UO0 zW5yffMCn%(!?}w$uW;2x+W_V!H!aJk7+9IH>TJr_5fgl}o41J(& zgqQ61Bz$mvy*&_2N7TFeV$~a;GEx1MlAxYDf$w%-xWoA> z|Ci*6gb6d<%nU(5I)e&nkNw-ceQ?fGT29?{k+W3IPBW|{KyGZOIp zsOA*0c)_XM%%q^6ytaWXqyMOCGM*az+xN7xZ-Erm4 zFR$Uxl36(Yf=x*kaYgvc=8+vo&Le>xR*7|yA>VQf?045e!+XW)7=l?TdFi5-DLptx z3ADK1ihwr@(@rLmJxSdvM3dwKsD*#d%MXqcqvO_fn)4B%s^ zJ#G#j4&j@}LC_@1$jg(Xvb*cIdnqT9D*0>m(kGGkaOg`Skg`ttEaXa$X>oSs)gQ!p z>MkHh+vk+DW|wdXIepH%7pf-;%>>Zi!evJU=5n`#=VHD#+u=5vfhWDKM?E|LH|fp! z&x1xP4%hHqoVp1S7{o(-h*?Bn^cA+z~XHo)?lMKr5Ve>X6t zHT6AEbU1{96=X#w2N_FF^1O02fF)07MetPJq-W;Vd(6q7KpSltW;SivU^N1%Ah0X% zVnMZSETH2VW!AZN?J%fuAsexTtQ9>VYye;?)WH3qp@P<8os=C&NjlP}56$?-hse}5 zMV68o>BvB#pU9s$5EeBuvmn0}HIg=^bl^ORm&ci%Kcnyuz(ygcJgJ}pxXV{@Dr z_L#iiBKKRhd3Ig4!Bpff(i8RaK1hVz7Z534h&Z)dF*0E zg>X<}g%Ek&LepEx7H}N_w7%~6zQ$Au4cV16?dV%`Egyx`Wy6!bZo>8cTb@J=Z$~-r zN0$X6&jnj!XHsn>q9el}K$N5*LnKA$8h$s6rBy{_Q|_9Y5GA$ zB_rnv@B82nHZbq^!P(Jr1>-$<8Wk`i)U}Rz6+;Y(VesSoQFtFc^QN6^#a0ICQ%%ng z!w-=l)A0odzzqX(FP5?}!M8QTg=|mm6Z7acnnRsJI^A=_AsWNQ{oOb2RwHeI++nlt zOLK3{$8$|+g;tGjeI)JMRH!ia`a{=)dBNlNNjq?_%$m)f1&M#wKeM`xR_cGZ>%jf1 z+=})dxL6#-d^WD-%T=G(J4Zn!#){>Uw>9m=@G-MF~!{`CSdu7oiARRc7$66>9h z1`_h>-qWVmymU!0YnEinM`{^ZX;6z=+!5k(kbiAf_G0Mao9m`6Qju!>nozzpAQ{DV zN6o@iER63Qd{umF`Wi43R^BD}Rb$ieS^W8;7l{1N@qGnm%C~9SFsa7cnG`oP zoII9^AT`Z6KNwccC_>Y06KMVp3ymM+cB7Dyj5(1|jUA(kY zAiCxKTF}W6ZD$Htg7dDA%e%rMtvZ zKzIq2Po1YcoP|jEFd>$oW0mW&WrmBSNs*Iy8DyF8$jXrrbUV*-t1;@r`HvHvgDS91xJ!0 zACTppP!`drgCfaYoCJM03T)IYiS>`MFe zNzyH|fV%aH%c{^5d=5i>=wJ8k-#<<}5sVGATv=b&Jf_TsYqb??Q3a+|5xt)DDrX@af0$?sz0u9ZmFp zc%Q7A8ZpH)FI^{9cQkun5S*{SoOVv|>z%A@=-PZ_C<03(&tLwjP@t`zV4kkWJB@{k z1Mxx6Z;}oSuZFH_Z_o5G!tZzGHwvp=*=|C!^~-ZVs_IlpBLqin608T-{Ttnl_ZF^R zC*V{!7xNbkm5Ze@l=bC{)fLm()$-vPWJV&v-H+zP194-#o6T&`^h#&$R4m0r{O{Mi zZ&4ifYHMf+M!`>MfU@rH4e?L!m!TpJ2fCB*8SA&S87%~(+B!7?L)>k_)l8q0MXPe2 z2;C+PCS*6wF;nSJ53~f+-!-aIo}OaNMn$f6Xe-Sg5>WoKhfR};7nOP#r^p%yQF81y zS4ugZ`Q{7VLIzY7{=f%PX{fot8-{^ae z62vLD`)iRo4Eg1aio`2!)z{4nB-4F}{p4VR$cq-zT<&tQ)Gi;THAmw?dsg?{Va z@ViGO1J6A)1}Q-Vc=;nRRI_|mg|ZU;wH`*{wp7WS!Ip(TDwb1kZ9b|G!ARbIVjjr{p8I&Kq5~?Hy zeFieLc?}jo6N6j@*IgGZ=pex1j}R+b<$ojTpPYgL)X~+`glA%i1qf&~b(KVh!RZJ; zMST5SdPNssAtf*!q)Cr1ok;;f6fa9pO>WAST0{UsVIwL^dI#(_3kUZ+x1&$kZktqJ z8)MFit57w`APrn=TXENZ?Uv_AQDeE&{|zuN%^bZ!hUQxGh8nsnLLd8ktq7(*jPIs{ z1pXwLQPCKZZ4vuZ@jQsi{DvY0;`Btg4}D7Lr)mz@!309n`pftkhdg}!{^{35X3XzQ ze9qiv{H)5^rZ7JOk$TqKc#&yC{sw)+4DP3IcGzmFftF%n$P1Il;~;WJ+U*1HPf(zc zZ>B>iG|TG*2&L#h0XzO&|yXl0-4gqHKRlT(AMi!)2S%L`2-BeH*|!q&W=eWRn~!P}(e z-rI(xw%uYagn@yI2T6RsA|fsLTHFe1fiv&L)bj3W>;6O&j-Z(lW?Z^G&;GoG5>Y$`$3bYDzgo;nYr7|$e{Z^HW-4UR*ix25B)=EGx zgEP8y5r$#B0OVa@JONw0UtU*nckv9Knam-LwZ(v1c)*tMWQOpc%;M&wpHVc@L69iu zzo^T9v-=@tmBb$6sutw8mkN)bYn>SwuWoSqji@cKObC&L|MJ0mGKh4)9^-`OiI0>& zqh3lZMuz`eijahT9;?lvP{UOfrCWSJpmmuKq6#m;h+oF%C>>PspRKzFdFUlgtzuy5 zCx0I$PE6V56n6@mgAdrUdf>G*TWT?ij{2ADw!6pHu9yzsNWI4_ZJgm$%{8Y774^IA zEfqAzNpedbY37BSioE7@ha|b&=Ox%st?h=>R`wJ@o1j4S5kiG=+Nwb-?l)q?pGsG| z(Tr5QLp^o5M&_n)&dBI1`R#k^uE@{Ax)GtQ_8q-TUwfPZ>;*r?CHFl43=eRrE0hnF`%PC#O%>b{ml;7U!0pjf5_F>k!JPI+ zsP8dPH-()si9m$1UD*o@9Cl80Je-mz;g!BC4!q#nP7JrswhtQwYc8_jy|&hJ=y>RoGUATG*tfw$sC*?f~c@E6+f43s%66Iho{j&`*`+PPn(2b^lpsG^!R0EXpGv z&m69a;AIIC?n%yg#n`$XEdFXNCctrRxBwPMddDOsvTkf977c|=$E0UsVQ5q$c-M|M z1Ao+;t+EZuK}OdK0OMWn1Q7VZq2P=#OUS2;ZFz;jgrRxFz2H6aH$7^^c?upDQ&XR9 z810C*euR!%GCvaMw*oh)pWGS#Z#wHTx-{qdkcuxfH}dG`miKPe#TTLhUb0*WpIAhp z$%ITyG3;ckG;AkG(kisiYMd(K)Z1|iA+{Zr!rEH_?N6j3AEQ%pQ3u$Od4xUgsyKXm zx@D$^uI`>1J&cmE<~UmDiR}3hOla!(M?jOz3*?=cg94g{%+=`Ka4v$qn0dcEK|BiD z+MPEhIa(k<>o&eP;3o1Xsrf;Gz9p|RQqLc6@7wB;z-bDZso*-_iQpllM+(SNR7qLw zBq$H1DRpCJ^zBV{30*A+6|I!WwRygbz!2rbd+n;^)fEWr>{k z+Y-k~p$Gh8bg79usM9)GVLpqs3Ndy7@|-K#!aKsMKa!kwxlIfZ(_? zeM5w0#i#EMD8X<*Zxpt3pG>qC$m+xKz21pP&wVK;8)CdB*xQj@X5nvzr!kYA0waHret#RiL$ z=mA8?XgEYSG86Wa^}o{Afql%#blMR^he{APUY!bh$0W-#S;myqg6G`gqu(5F}ZoejWZEHhDJS||LlyojInf>fU~vVRyhSjg__8oZFc z3YJD+`31fdU_k~~r+=Un1re!K24AXGE-JoZppK9=aFmnEJghciGJ2VCSPF1~&Vd&n zyblky0HJ~pDOuadjWgfz;pf402%4*ltEiZh{S6M|W%=BBxlu`fTt!ACnXp>U3r0^t zZFpWxWQuCdy&7J#^2(YHEXhbZMQ#vI;b$FYda&sZlAn#CfQFhnAszE+_M(b zSx|AioUb4<&BBFp23LZg0u6nB^ry9AYJG_~@Z(2%JVQZ(^k{FLO!t$W5xp=MnFj@4;7qqByzgvYwdXQDXdb;%4onU-QbCd3E9MO zSRrq2$9xuviA(P>?1$~)Io_QuR}s#s8ly#m(Gmk9N!vkh6gr56ln}0wnJ}Rdt)Qgh zWNbgBk$c^-x zv6g?rZ&f+t{Mrpw9#j4PJGvhoK}jx9RnpAj#n_u0o*?S@&$99S9+*HBH*Yi&A}vZH zyQ{D&a*1LlYGZX|MI>E}3mNj*To45U!O|)M9%84vKmzs zxE?~ERu+z-%95gvEAwBjZ9zrZNE7c3q>|||v?A30YQX{kU$`dOapHvuL|(91JE4#oTi zoZHwccMO_yP`jFF?G~dN9e8uHrQU?EFs0Jcsz~qy$M1#|(W<26a7Y$4_s<^8lZ+6E zzK~1v)%ht&WYwdIP62Qk%$d|Z5nQWbDs^lyn5hKv6b-%vPI?!T4Yk>QI6@K3$dcGF zU_Ytdry(mb0uNz!Tx0*Q4iJow6h6CT9J=w>HSZ%MWXK;WXNnwsNRT+1@!%QYr2);9skrJezv}VL_D!)QcD8=H-=z2VP9VY9TEOGXJ?0t8>+FXZ_T!Ni5 z$X&=3&GXX^uOA*0CJX+#-g{T~b@?-=exX*~W-m<0qKfUKTsvbU>GvHGWl9?H>sm@s z(dWur@OuOG^I8!^C)cX=x?bEx9`BHyWPT_dW`1OtmwKJ8G@~74I-^SJcRp_aOd!To zxu7s2?aBZU-#n!Vouu2LZT(3y8i>9XU`?%h(DdGtFMJrgu85SUlKq`yJdMw|mBiO# zwa@{QI#e%qiik4rZE@N9av9OjOOCPsxR27rV~9#(A>B5A<98oRUi-r}Rw0d4&@ zrtF2LAEgAjlkjgZy)w8vbW%Sf;swzVN&%Qbo?p^Jaln1;wrZS&Vm@=6eHX0+b{Sfn zGm=!UaA_|b9V8qI9N2Z1Si@<^B3-$`Azld3Shu+*wjxL=teuGzp`($Us`!Oh>*O4& z$u{D3QB8mvxK0JKvhzl6fg^e~i$R3I@(Pb7pEz$d{CBmNqp+@Ri7*!<$x3Vr;hT=9 zy%QXCjD>D2s8O>1-x6+A_1;hcfiK}dtTq!$29pSVQ>oDyK|_?ipI~LQ`}fg5V-cNX z##&5^zF#Ks^;o;N9#*~(ao5{|0RZ410N`ZG8}9q8jAo$Qm~4 z0FUTle9|A5Na-H;%<42W>wYioH>^I2;3p@wQ-A8_kM@ph5(33Bk(8293`VmnIokr7 zQfNb^5d{j63aXfa!QPR2>sS(*({wY?bl1$5V_&DqsGunb5##zI&`V)PBDJg0eW3}Y z;x+REm@fb_rxN@MX)MO6zY_>(YlE}(8U!M-W!laZsTL-^JtGLV35z2H4G7hKS*|tT z3Ls0SFQN`h0=+aYC{rQjPU$X1IV`^ z*mv(Ifg0r{;!MPj%rm9k=3`ejcYgA9_iol;N%r@DS*AOp?)leU zs8SL*EXmGj@AW>OLJB13vn3Cq?|4KOe^?Jlu3{<+oKA=^)8v=br>n8y(l3NlqyM=+ zKp)RVG_^YJ8m(7aCN&!iNoCF+$gO^iz z{io$5?R{R3cP@{lhmJ|`dNnnWzdy&8gl(jLXzD7L*~Iy<6Xzb89AyQ;{TG)Zfw$K2 z4ndlGaXMbU%fOe`i7OA}KlBkl`4bgY1>JMsoc07Kfe|mRhtMlKn$=JVCDEUoHz)an zf-#s*h(}8~B%$WYqA}Js$lJ5dD?-3$za~oX8S5VPCOFu$s8wu=%qhiw$}rLR>M1H| zUXUHVf#FV3EV9l38CrNPV|?wlr-D?q8+N{lp3jx)*vSWsc$8JPylULRQ~uEBm?;Le z8&o|03@cgmUiIa(l|<_qdSrfctb9RSE1EgIktUsSj%1Ixylb@=I2GzvEKo=-rz}-X zVF3ia`9uTmk(I_+KS%ACu}f4adJItncKXw7aiK(+Z8HK$2~-nMlO_d~-0e3^KQ1KE zZi-%8DwrlAQqeYyR1YGuO`Imx+{tY1BU;d9* z$qw~7SF)ep(%PExnirAeUxYCE!!LW9RvbTm@`AL!!EMN{-;6x!81$TxRY7yFJE`n` z7z+m#8FWW6O};`C=Ig#FXe14MPbsHSnG}$56I-T_!qNoGmpk_{#&-nEy586_#%s_Bv28i1XiRiU1>K@hd&Pmw4f@DP)W~yiD7y#RE_#V|)G=Vbt$}ds0 z5jeKYRvM` zG_Zr}Lfd`lLo)1!1hgkVXqN>YlHx~I8Itnz%gKK!w@K}Q|6hSbZFxmEgZ8(d>fHkj zU0A8nDjXP;vqU0F1-%Zhl5$Pjcg}P2cb_Rf`Q~{Y$txIrktkED7y6 zVyiHY)0f(yIrdJD3o;NxwcVg&xV&a~N5d&G1zs!zILP6ay zX6Q0t)vt!d>SrGQq^u@}ta>m76Aaon07Ye`H!!)4Rs#B1V2}Kzz>Xr=_$!$j(E{bo zCR?2D2PEP1caD5Ox+3P64^=CO5v_hG?gs8GcbOpE1c<#{dRAfKYsa|i z0&j0(5RFyKE;V}_m?!$&=CxX*1o^+9lF6O_-I|A6< z?8=q&Qu~YU4~s+h*9}XYfz%y4F>bbR(zQ`Q3u^NQfM#s6`xCpVi~36wI{ZveaI!*1 zaSEm$qvm?$rb{lG#Jkn>+L&m1X?dTD*K?LNpYNAkjS9j$wL4;|837fB0KeCcpVOZG z(&$#FlKZV2xewXC(vB~L)UZDmX(8Z4ddL?Ur;ypDom>%o6<>Y?8^}m|w`=mZApWQ_P$A~at|!MdX6YT5$n=@p9X~>(7Gz_e&yL-S%oEHX+L7LJe^Q_+ zO=V#Q!by^qM`{U#k;yk~e^`J1$-uBr_Jy=s6QffxNnw$kQGE@V=Rbq4>r;yXqG2jf3Esq~_5f03SY0qSP2yBq-l4Gh~cq)W^kP zr{8`|P2N?hDX#|)O}N6}Kg>)`_x>=vk`vLpak_mlJt zbJywDJtE)}FE5zUZc?b(r7NJ5q-wbQH~Kj5y!0 zN#0bvBX1>Rbu^YeyO~a!b_1TnbExV~xd>qiQZj)y{J}2dzzB3>7%R|VE^2|H9*7J3 zLZ5_1`o{89i5zb-Pt^3w-5`=ZpBE08UNK}f71Bfbt?`x5WJKhk2Je83sateKLzalg z_IDKTdf1oq7O0%(*%F7hyM={HDI>{+@{TG*Nv?Mi*?F{ z@45+}K4XnvIp>XNS9(zDz}cLp`=_}d>}~XAKJpV2b$({?@+UrPFZgl3;ilN2LAGKx zg6+xsCXBWwKw?%Zxq9USuF`Uzj#!l!y35%7J$rn*xarbs)ntAjt|23|6bVr0`ZgZ3 zAYk_lk~^vjr%Qp2dd3oJ)xumQoK4@`m0MqMw*|P|5kRFbb zq!~UG(*BMeeG`j#>x~!o*UTV=dXzL3C||b4wrNB*du+m!cu(ToxI%{?el|*0B#ao! z&tYJrgms3vT1+d=S)cyDBMYsG(w@nBy>jsm!?MOM2;h>z<*CIe3p(?^0Q{48oQ3?8 zuP}l2Vd|$&RF;s`me>>ir3JwU4}YN(TJp@hgQ_o#8Em$m3Fjj?sGcFp1 z+42BBVLJ>BfFJ1nE|q#U5<@XNW0y^MeaM)yO@FEzo*jB2`o0v3aj_I&DXb{ZAAt2X z0gFf+nVvd0e@mIpFmP>24oZr_s-2=%BDZPSg|C<{1a zT9mVp#!%rFAkhBR$iPksO@glgtEC_OqS`2S-JiVzx~btXfCpJ_*%-hIV3oLq#R0UD z`R#G}%aI~JyU1WF$`7(0p^B~u6z>y$^r3W(-eB>}NW+|$J9ugpg2RTZ@3MW)EENlx z-%(ikub{$U2%jd1sguh$+NoO7ai*!|1hNL+-(KuYit_RxSXfwit~fF7$gJ82bSI9; zO_LHwc@6#(TiS-k?+ScqRFoGdQ|es`5w2>>gP;Xvw5yC8Tq+K)2X)x&HE z<(1X@!X+(`&97Odg42zCdc6t`?E(8L(^QmIUfPjnin&879+7>ja$I9;pN7k^Pezf9 z)aWR#u)fhg%_w#BUj~-1R~NvdBtVYUp#Pw(+}7`9)NKq&+RUDhHlIfiNHRoy@#mdM<=HgL^W@GR^ASr5n%>>>Hte89`Q^EZ`jG8wN}($e?%zC3O^1 zV%#KRTA4ot+u?#}h|S16^p4b}^cjpALz6C5`Izll?%a=Q!Ai~XddAT*r!=^m^vlWf z|EdLtNE&dpP?Rk?3WmA<(7|G1u`=-yl~M}V)cU%m=@;D@q`;V0rlfBJbNk(*?DG97 zPZAtvd}W!rntHE61Yr{jl&RD*Te?aSTO@^)WU=3}?|VMX;r>E%mK%`w{^08fpAJNe zxb&NlH6TTCAZB)a?4^Q?PbDnouX`8>bDknZhN!?)}gXuYGdx$bLsK_+|8h((7{E@pp;QogbOnqQIbEI z281J(cKki=BnJGl9Jy`jgbEdE^2irDFgXh7iIIQ!oACIJk-K(dJMU`4aga9T!d|Oc%<%;BL%`xRlY;V!KBop^2B?qGHmG&EE$`YU7)idjuL#G zi&;gq5ZIMoqxLcCVL*xWu?0 z=hh}eN3eZVaOop#h%PK3s{Pc5(=1RE6My?H2}5{>j(X`!&xqMhS0=O3e?~59-l;wv zY4>I7T+q$lta(XbDKGP&RTa&U*{%u7=Wmc^5l%kQYfI^J{cacaMY zzxw0*<3dq2bYNkS#}|`U9w;c7zFrTbinKp>xSjxSG@h?wd~LHO-U~gYr9*?4+P6Z8 z*q!$kn6^LmQ*y?tl(pYZEkt_fw2Y#ds#|(Qi)cv_Uz^G&=LWw!U>lWd1&-O}8A ze>&-Yx4Cg5(A|?XYHsss%&VwCs_~qwm|C3en@Fy8GsYp+y-^hWb>Dy9>G^scA7W@aQdkod& za#r2O;B%41`;UzmQechYxMM}+{{Gk3E3yxcuE&11<=tWJle9`NCAc)^;r0`yuNXl$J3@K`5L1BR5Dlzl%0 z?r%AoVgRL-CD(mB62nYK`aKcCV8fe3cqXi0FHV?zlK&V_N#gT=`q6(HQYB{TXNPLK zDNe&4xrlCY$dPC!4URP7D3imG9Vpni90}s`lZbC1H=sA<~AgB6O#c1q4TgR;C8o{TT*m?J-1|cHyxt zZKV5+R74Hc`{gpY{}2^a2E4KiS~?FuAtR&?IfxlF#gJ1CxVE1P&{ElNf!Hi5QQ)W) z4>N8UsV6*;?g0XexO2VPzb4RGPPiS7Xd!P{)0my@n)P?Ph__1 z_{FnwrDDpqE&Iazdu5H5LAU5V(Z(ad>qT!Ppr>b=4r}t_73bXW$^3NmY}dvZDj-Ug zkxu4frcds72I(}u#f==YRS-4e^xq-1L{GYXOygzP%MvR6`0${xhOlf?WW z)nz={QLG6mdpLWCRqXu1xuyq(%|||=1|R<`0;1!ft*aVlu`J6zgUod~l5Pb}KkX;z z50?F$w{|R_gBifBi4`vpfmM4f0}+X74qK6nzPv`sH3qw@9Ck7%*)u-n@v)4Wi0P(> zqn9TVd~?R{k-H-jMYobd&vig9GD%Lh(J5@U>)GrS8Qd%{nI3A{jf0dJy8=J)*TWl^ zfd+{Wtrpv(t?Wy*Z7=y2`pZO{BsAlZp5a$WqE?vz5R{(kw=V=oFj~2SH{8`|1ELq# z#-^OQk!QFKNHa+Cxjhf*GhKTa4ixCqk~K~5Psd=Ebq)>G=yqYK*L~~{N6NQrc>GQZ z=KUBaS>u5A-Y0rCO1cb*!{#XlA#PneBwpLC)v8ux*{hcltIfwbvYBRRQ`5U$ESfxn zy3B)`Z-jpmZOag%1*JR}t)0RKMpH@BW86AJKoGqpPNzuPrQ;7B4{OLPrQ!5Zp>0&p zkBz0py(R1GhDd8=Dv5;6+*{Yl%^fgCAI}TZ%gdgRh_w8nf5m8@U|U8P^#5Tdn{_ca zF*^VM%k3;*r9nmQqsiucJNMMERQU;&N_v24zwL#OWKJv5)d87koDEOE^Beazl_5ez zrW$}?cp5%fJ52rw@su2ElAI&PfxB4Z|MLRaf@-Z!d}3fwhG82XQeEIsQn1_mJTeXI zX*w>+FTqxhx4@Mtc!%z(6YN0}9Iut@+Mc|{0-`#mKc-*ex>B450@H5=bG;OXj1)gl z?}3$~`l*JH0iJ%HwXjT|&y+VtFx+DnLVd=k-5|m0s3ORHJ9C&N8sJyhgk!g`mdL@j zj_X6uj`}Qh4-Y8@FC*g%+}j$RXzu3eG!{#32S!LPTypg=kTLsDg=mWVvh5#vI7N;K z4;}*Dg#%2$2F8M8cW*R_avuyJ#>fU;L=|2$#YLG!m7B?fulR?X)xoJl7=p+%OyNl9$w$+Pwow$TGV(m0fgZTu zsPuQ(gK$dpm&E;{`(Z5PEp@cnf9D~rP)eAXPj$bl-imk<_2_9)P5lZb)zr8yJ4FiG z<0~NAuZwn~K1y%LhonlTUBr^P#GS|SgrC7yHG~hgM4I7J5^sGGhyKb2@U_LQy0zS- z^XnB%Ph@j2s#1HaU%aVY@0ddCdPuy;+Qb4vl0MTv?P``{o`c##pf!x09 zx#1@%OrH6^kdF8tA&B#Qy6oU3pPOBq<68m^V-}V;z8^bm7Q3~f6&#e1eRPse8B?C% z^64)WBb^-s)mIEu1{kWay9;sK9jDzRbDU+#7^dpG3RfQ}L@nvcJev_em7t_~yOEh* z#(5H{@KFhj;Y+d`j-c!zW_6s>g>d&_X+jmtV@dZ3%8QZhlVstqY~wAJ=#mDtq+4*3XBoCij5A4PhVkVT$LYs2 zdrOIaVps}=8tVQ$%$plf;BMiT5SzqT2|gpkQJx0$fPLb^x}Z|!6}Rob>ljwh<3{Lt zofaz>o9cnE?`DEAhWr9cWsLBj-c+$7FiP3<rJkV z+_cQ-``^deT-$5pqyoY`Hsm~{I8RS1E;29^?xYnS?(V{4X%$RF@c?K zYY`LD(r5_apV|=@$BhUTcM29}D+B<Eq9x^1b zo9;Sd+Q{yvx2vW)lZxN4n(>9{$iSH*pNWn(UHf@(I?NelBx>bA!j8c&LJz=GV>eMy zivEd-_a}x_^H=ew@<(U55oT8<1@HJVp-Jajcbzw4rg!{%X@`X;x3{gFtOJ|x%+%%Ayua_hYGtrhyE)9bI3R1SFn`dp45b1XfqgT;&}TW$vI zCC2bDwcJA1cbXx(Y*X+4{UZ-%0S8ukCxaK)3UM%_9HgD;NF4?8ywCPLW?0Gg0DAxp zKG2d#YbM0cfw2qaceiQT#Q(6%X?c4?f}iYvSw8w`@nbb8$!NYru@C{Dzi&j6clZaL zk}W3KIso*>Mc>(i>_2{ju5^c4Q;R(pEWlfsr-o5I~(Er-8CtRYi|= zoe*EIip=%uRi9-en-#M8dV`VU2Ady_i!Au|>Hcj+BtP!i9}hN$RIyK$#w0JFL)kVB z`8DipR{1>EmJ%dYElUbC%#4s*DHCJ#GKw2h<;M^Y%DaG!qP8$m?JXYB3q!Br`A#+E z#3Np1QC}V0FoWSJrA!m2X1=-H}u=k!-4h6o6#- zCPB_zmbO^dP}T7N5p|YfQ9u9JM?$*0TRH>*>23rRSeEWqknWNO>25)K7g)M$X^?J^ z?v`%2SHHjibwBmddNH4Q&z$o*XO4p7!+Wwe6*p*g0unz~f#u}AhI$bBIvj=4mDjxTq6)+1y)2cM-1tAqcoX_hN)PL%l zKUzgdPAPxZhP#Vvu35VeTx>D8+9xk~i zC?B&+MmQU_dU!#Zp^A3@Yab8)f_e0Sk}MH$E0QNF{leIO()5@u)a-D!yt@J>s({39 zyRcOk_`q*2at69>2NM`~O!!EHwOm)7860w#Z}ERuZDcBh?3E7q=M8z7t@GX8dTlAV zvNZVD;z4LLs`;RidJ7Nt34p!qyJ+;Y9A$e+2JWRkCLJwrf{s2Af?#YtsjJq?K8>qu zi@q6?#NU$&(~rRlG^W6x2@;|v;vGLfr3WMTZJKPSioykYNX?Jvr4B}2<3uox&(DAJ zh%g`SBODNUVRqtrCJQuJ!eM-c&*k{E!ZWwb_=ZhZBiubcY2*GA&u{do>HIm(2%@mw zTZ|-Y5t2P2?1_^LoJ?OHc>Sq{q=W7?Gxs06vzg=iztZ~MgXP|_8d36dwo7=l-tkN-~a0V6)Do^IL$ zkFLDO_KOvrL}ex2QUT;B*sqnUn;I{6B6Kp2+>Z3EmgX#N$8bwT0dZT|Cyb74MxA70 z6Il$_pVFj?F-*~^kn_Xq2+%<}cTz0dQ+cjG>z5fzm_k`7?BrA1iwo%Q_jPM?`Hw08 z`ag3g9Rf+&{D{G`?88nPVtNVl1p+pC^#1F28a94_wnG{ON!pGnZ=5K+- zBaIIS!J9wAl7B=(wc?yED)L^lR2g0eByML=F{Pq zqb9v#0sgJFX@JOY@@oZO)q_bRn7G6|f1@z3B1lTeQvvQVu{Q8?qZ2F(+4bT#}X z4HBQ;x5Io?ER*QCx(tYJo7YuA2;x>$VBjOC$lav7<4YO!4RhTHSF! zO+f;mgP@8r6OT@5n=C4W3Wfvg4&Y*7gPeYM)eixkloUml%tnpf53g?EomU05^OojqeF|^)Jfrg^6 z;p3_*M(ZC2CpPQOhHK9rFL#nDZb zBL?3Uz8wNcIrA&xPoxj3mL?ufjbwQh-Dj;pN~6I7C{uGpwEIj7U9DK*XKLt9p#~BPKKVcSEiNY_*#lr zQMWF;OGt>c{F_J2OYmy2_TcIF9A@E&u`%n`v;)h>8!0hmY|y(&{ivq6N*n+h>VW$2 zrlPzOP|L`S*`{!n!bFRMw&9ck4kk2;BK@C+FAMI!facYP~Fzi4#T2UX$eK<*V53`<)xPxgRmQYzM;rf^Xi z4Oj6hKN?fG4$G@CR{>H4gu|z6+<)Gb%HuBdT$R5T$V;bv<;UQirLJzeGNgY);!#tnoMSYIn!`5nEh-3~rJA@aJhykpm+viA^&EeEveZGei{o@unMRBe z>;LE^G~nn`W1#)v?{1kr4`U?s;ID$8d~_n%Nh`yW!)b8Yl)i~Q=ek@#HXPIAX`%5X z843~+wPlLzGX$12i}AL zQ6?XXd&bdU>3(CBmb`qvm;B|o_ApS4O!7KUMw~&n=}Zx?Gp?x8UkJ^aUuVcKW&ICX zO9D|nY})UzB*s_qh|qQ#@9HR_pRkG`+K?f-W7DiEn=*e$D5UXc+M#29J1oS;U@PA# zw=f*uFr^o62P8x?WhGp6c+?4V8&c)|Ol4M>+B_<|F=>Z;@;(!eY5!?drE7@2v9$*= z95KSBDUSVFK_z)Vxxj3&mO1t$2hm+;iv-g7Pd6CgVbs=hDD*(bwOKaZo3LqnZkS&2EUXu+bs z9}nZUn+4mRMK@O7Bd$Fjhm#&Yh7Hn#FU7PB4NnG(9XD5)zRq3v$9_NnVdH2@t{`tu zdD}Pj9@(m$vfo@<>>h&>w2QAcSZ#)(;I4B|E7tXRPs0M#6J`SjoizFcy7}4iA$a z7n@#@9S@UfM16w&0TyQOrI;j7G5dq$rC06)f0upuHt0Ra_{2A8t1JpvBizSOIE=Bq z0>hDUpGoB-GAdR)R87@m`Rn)>xfeBiDTHoxrI(c5#A5Bc=BOmh^M={tsVMhoyUTZA$-Bx(_RQ9G6f$|Op$T1O@#A)(-T;wfQVX49D@r;XL6E3ig9CNSy?#ua`WF>amN)- z(Y=|RX-WV7(4644O`iX0B~HD92Ls4jr+9EjMEY*9TZu{@p5T>gTnaE3KD=KDMM42+ zW5ZI{l^agTU9yJ1^tu~Sq_&oA)D#$oq>P*6%tE{@8coQ()^t#?i=PRT^2?MgEX4co zHH+bIkYr2tfgSB4gbHY~h_mOUmnRE@z%7f~k&_Mn@IV(Ln@~SUh^wN^VB`r6fsdzw z^cYaj0~-qu;7AX4(a(OYw_%Gr2t%E2?@T&>$k9Q=e?OD3`fL)o${bCptJ*JSa-_y;x5hew6bEA>uFow9 z4WjP!pSGXm{zOg_#)BdGA9XL>7_9PGc*gpW9IgXTsr;wle`kDNd)5E#{W%7xsm-78 z^@Vw-mMB+m+4~gh$B+CU)NfN~V;{#*39@-tLhf89cl)G#DzD4}*R5oB!euUWkF`zf zJC1_-Se|z*h<6QX{gcMZ)~e=cwx>ER2okTul%GA5O;Ql0&wr`7(UaQLS53bbivdc; zM0_om79Lh<)yvYmwnWE3Z7G`Rjs zGe%raj>lQ;|Md1%q*5+<*a~gc#C%?8VEE8`0mMIj6y$()<6be*S}jQY>~BF?MUkJK-;*(x z5*GGSmLEfL%@uzqk?rvC-`!zgHYPcawd|wzabtt^o@TNdKD0We)OXg`AR`ma5tgWv zp{b&YjM*5c?3#PZe#^iZ!?(gLsz!Xn5(_aE!=2*9-64h36qk>9k&OBSOkm@O2&^l+v4#-p2Az5gQN z6jGTj@=o^6hywo_A^K`ejb%sSdC}XW5;-tN^~wVh$n~lpVC&{as&+1xmXXGbW+(cO}T3*Gu_3b zcHLKEIfBR=<)m;c#V&3fud^Cxa0^JCwhiLLioJS2CXh!fopX+?m%57YkeP~aX8&Mk z(T%sWNI2yoN+TA>KlQ!3!C@1)SZsB8JjDexr~djJU4FU#e*Qh>*89f6il{IcS334c zOYDxU3ZF>x4_fm;)P`{YrC=9~eWu^Qg!?>m5;~GRZLROYgmVz3KHP|SNP?J5HWULt zy;A1!%_iMGaqqPK7Ae2z>|~&EL7=M=_NV3hl=jXoqX5hlD-GCz8kve7wlg~9#lKfK z{7n_n)2jKJ&=p$AWc)fnWTfod2g!=CnE;>_J$KMtxW{68YX+FBHOAebE*e?La;c5) z{T(`QDR$B6AA#Zji&X{{S{tY zrI!pnflz!>vsY<~lAZ<1Sf+92Z>&K8aTxe!hDa4OBpii{7!5%K;KRywuYEDhfR+Q1>h!M&#kU*X zUAZo)?bfzFXQ1z;WD>)|JR*ozoF>*?s-X|`rM-v167hPi07kVn=*G`Ce8~%C3E?bw z&KQHmA+9wGza+6iW(vSN%s+#KU!G#&3Qu+T2qTC(M*OJD8h+NzaTSX^w~K(=|AcV0 zeusP_>BRQ?6GVikzLk6)3ust<-C;d5U^IlCm1-^#dJa65c< zYJhVW@T(RvUy@U6y@rC*+uXHl2p=oyac4}w-6}ZVo;%x_2K7H$f1XS$I4aZ7%f30R zj=-jx7o4dmyPCkUL|(kP?@jnD93~_h1qj4|!zx!UbWt**n^vSbhe2FvTw*d^v>BP*BZ2hNF zG$ZW_enB2SBjTHbnBtNdRJI4X7>M9l*i+0&JL}SwFJk>Ys`WZ1F7w1G7hXTHbd_fo zH|YdnX~o8oibSn{^UT$z0T!-h`8Uwe6@>cZaR7dIc}p~oTEJjJQ80m_{^3TSg1+Kd z>1+t-%3>cOCr7lF|Mb>kyFhHb+jecj=tnH9-9lMsn3+=XJ{1GOLdLq8q(cNO=c^L> z-NaH!z0!v9>q7$w{|!tvgE+eqbd%8pgC79$-*CB@WG`Ed~L|~9f@++3^0|>&?q(E ze53RE*eFzACsEp_wTG|AaIqy}9m#^?2P&2C^w&{ji;S<`r;oh4K^S55&`8)l^0tLU zuqnQULqO*1LZVIyUS|b9F-Te`R5#f=_yHf*?X3zWl~% znol<^_f`*joKGe)P$kJdLLXpcTru`NT9`Gvo`2fBs;xpvV5o3S$>A?r&lx!t68waG zOGZA~W`gb)MbHFlIyI`sfx1xUpS<%ykN5ygMqK4m5v8EZ_hYub8zgkv24Pz^r-2>Fx z@2L;|V5*u?#_(B{)r{h%U2WtlE85J6wA`63v6t{I#s*rNK@s=anEp4&Ljg*ma z%Lmd>uAr1A{UiR5P~)^EQUjD+=s^v}QmxfPRD`K0>-sCG9^J~ zuPD!uwhE=@mpXz9dY&s3A1OGS?V?rX3@%-9cP zTfNq(b(yAX%ENGZeCzNLyyDQnUsUzp$d&?b>qD~sG5Cu{5g#6?yW7}0f?d4 zOov{HK=dpPDn|b1f|e~Gk=kR=_^Ouf5thNVyR}LBJl2*>GMPSqAqJA-55wlf^NBoMBbGVw2OozX{i#wxhoL$ zLz2<72*)IyY?JxfE%KJ)%MTBO&{ivkm_J%`i4d<;PhizRTH<|)cS5BHPz!dWjsk6o zfcqGfVVJgxsZ;tJb=Ww{kT7X3`yW@t^goTlNhUa%ii-8H{ho8%AUOHMFJ258I--J9 zt*oV-$Y1P^%P6aN~FiYKb#ab~kVF%}@!{j!sCXwuL7jv9QchSy zHKURcUlux!VjeLMgr%-=JgK*3Wz}^EFQ7<>>C|Uf-ZAZ7WFUQBy`jkQ+ApSPcJesR7ngPK2UR(~;5T~DL<3iAhS;O(wz|_P zi8l{qbD?0;C`jJTtKUrFezP6EZZcV2ZXs^K z8-Icv1v`EWVSY7hbVI}RDM>?4e0NJPSnbHz<7G$g<)IWjm7K(h6|8qH;qSy!h^rv`-^79$3je6^m|)1{1wA(Hhu3TonrK7ak)1SB z%09hpYH#rBOF$zi=u8994{(76A0Q6IueHSj+g;V1+CXro=z@wqP$Z0{xepEqp$iv- zRJ0+i!;JwbmBoG`z73BEWX>$At+w#fcCz6TT?lWeObBW6nX6%y+1|Y^`ydHZb{TFw znvcEcB)2%~AFFrvb>aEoxzUL8q!YeK;+aZ*LqJM{XrPe}_+Tu`kD({%W`Q{yjK0Gp z7?gMHV%(%cj|C;J0|zGCR3@_N6#BqaD)YoW6D^zhq*W9MCI*J_OqtevW4y_-hdzgc zKeYKi>-y=!582zx>yOdPm669*ynJIQNiI{KN)e2n)_0xJYP>%HO-xT;R`azz`aIX) z3$a#S2)Xx>TCH79sm+y2FIq7jT!LUuV~r)|&Swy&TB)!!)S;`^dq&|y~KA$fVQWr#+Jx~ z*F;|Ib@>7<**Wq8>vxpyc>{GjDJ}+*YULf08nxZC{;U)50(uJZTamPy`SO~^0w=n6 zgd%lk{4uT;^eKiNTR*K)+}_p>!-;zff|DkG&&ySpBD;g1c~ZXkp71>leqO$Jv1wCSJ%zS`S`KXL2kkc*(!o+$^1yC5rWgMVsMle`J4)h`9`c z4`k;*^23M)KM`X&Beh=FVkbhHDKfqMgC1smvS}Xy^NU)jct^@De%Q(_t3gHP!Vfpj zziSRM=d)K>R9E^s6gUIsa_^|JmY)1L@8WuV+1dm98=i0w^tqr+#g~2ax0UW;FN7&t z+tN)aD~=OJo0mHNYcOy#z|wP%MGvIOOaSq@bNcOl+O{LaOH-*L&=LOl4zK$dI_L4J zY^oaz*&n$qlVz`}+`M`OPwmWB zlLA=f{er-voi4Bhl39+;@$qp# zxT*QW3P|(KU3Mot9wzUS)p_)Q$;~q#LqdDB)y=A&XK|~y?lMW+-1TTx{zAlp9FQaD zLGKcx8)hVS(*Ufx5U|GLUuzg`hJA?+Qmj+4+nIi$91pqD8r|&r9#G-&m9J#{H{zi$ zeg`%4i7PSAEu2iKGSJBNo7c{Ikk3=cduk4{H@~JFhmPhPAMLtkyO13(+Nd|@Q~Rmg z&Ygkt=PmkvmB{YmSF!U(E8pQbJgw?S^R?{BVjUX8@X2**!>raRQ~hvjep;E)^z+s6 zuZ9_ha%L^NVG)hAWqf1axGQhixVC!1mErQU3m%$QNCO^^57_+~QZX>O#@Rmypaqzv_+X zv_|I}Sr)?18Nhocb+w1Op@^S5McLXFu|d?vaj-O&kIDuaBFiofy0zMhmaYc33O1;L z^NC?J=cfO%eWRow-fphEZ^ChhFGt7vKW=*5 zw5&#x@{$bIg%0l1@DbFvOeg*0>scmj=E661|C-g0Y~lWT&NjZeOtsJkXphgM=lWD~7f#_DAZuGBj?w&{-C~c?Dm3 zhLj=DjP0+`ZIi1m`g6#qDHpXZt!;dOV7UXkXcYCjb;y_5HgaVlDp4>)@z`(AMPSe7 zIjp%Z?Qh`TB;SSbKX#YG=zpmWCk9wElUnQ>6ut>+)WPmZG_h_##5+_how>nW^~(bq zmmi-5fB&|)LhLY-qXGo!ciX;jy!1=KqkXk**9?$!Nkcl`jKE1}bdXAhiD}Mg=p?qL zB4?u{`OXVPecW?tAJfF+t-!%BUqlk~XDbF$iCe6>#?uhUQ@dOVKQ9#+FvpBB-dqtY zqv220#kLpnbrxBU5eyurReZv0qQyy?{nn-ryrD6XHRg$Vzuk{-!bn@8*MMbo3nl*Z zbn3rq?h$Uxv?nq47_F&S1K?H@Jfv%`C>LZNt93hZM7*V`{gqH{LFO)$^p)ZBzr zDs0PM$M!EhOAjw8TW!iW=}WlJ`S|=npzdVj?8urMHrsvP-N(JeY#o0XU`{omBs02= zY2Xds)xtpa=*toSDs(QT_+t&bi$o_{WreQ_J#kGexQXQ~cpy_5`r}_)x{qTs%bn&Q zI$r-4sKNAn{IlNCbKlOiZ*qkeO7zH{K~L!svyS4RcXy-D%NpW2X#ZHL?SoVv&r8B{ zisW-ld@B1kp&PTq7k}|vrEz#y^zD97uC#+S?Obr<>eB?g5@3z>V3((F;`-k1RhaZ zs$0}G5w(ZdXdE~yOcEv|0cBXB;YgLO|Yw1I8 z80tU}Sh@9l96NTW`Sm;mS+h=j$)ddf?J<@&y05bgF+uS1_C8+Fd=_Xwi@t?DvnibA zd%fW1nSC)#{l#03?bIc1qF~c6Gv@Df6Q{J;yVry*TUb^_8Q_0f%w!__D7QU^!uph* zYHAu`#qJ~@(kk?I<;9>~a}u37_5B#1(}-FEAby7%GnnLc+{$U5Nc<0xvJ$3bf~ue^ zMNrjvP+6?S1+CYz5$r#-EDy_iE%z{ic`mo`Ok8Oge$>UzfohlMLLR9wmfN!+`I0sw zwy$HPFH}>_277P;Q``2lYT_^uNc?7~Zu&n;SP52I^p_G=E=Og1=^|0=qOw099~0&= z*}gL$lv8zkK3~H#9rugKK}9<&25#qBqvTyPe8a#d*$=1d`5G*lLmbCaqY_{TAZ=V1 zOy*PnxjJBrgy=)y7AK=clPIU_IDBU!!W|k+6LBTPS>b*E0QT;_>N<750zC0RBTcA3 zmA)Hfh4tYKTh;pF0{;3(_U(5&Z*c+7qG!2`9riD+{vyBKtkc)e_p*#qbNz+z08dX( z9=pP;5!u3|LNPtzV*Q+scmepY^7C6-b~k@gKjTz*jM&{u5kbsW(SEm`lLHt0&4XvS zQvHe8_99$X9?`${4(4uUe2l6xnd3Pk^9h~8dScL_OIvc|-{H$u#uE$H%e^HBFM1=3 zhH`c#N7>7mxx79xJkt_s-ZYBeM3j@RKsxn(`mh(yx$ULRwQ(r!&*ilK>hW3DKeEQE zBi0wzNb%X3x~=;BvR-Modj59{yq%l4YGu(YcH0)j)7CZe0ip|F{T1?UZ_R2x6VW)4 zO}|!@SPbv(Gk^V!XrPhUi!7e4p5xi0IE~5G$0N$$!N_yI=xPpanbMF?ckCP5)atvNHM5@-q=F^vADEbHCn2ea=$ng=7 z#<~x-g9(?F7JkMj_@O3vcn>$_b_&V?t5m&;pHDmP*nY!;hBos*J_5`xtCYQnufX;& zDF_;Npyb8nzw7RY?8TMJu4CH&EbM)xyD@X*5@*I?qWz+o_?K@ko%Nr~*;M;CJ6h-` z&AjjcR9Nq{q=lkgfsHZE0h9_x3H5ZoGV^1@Cc792WM6u@``;njV$s>+3bmFHYbK}Z zJ*X5pdwQjoud*D6&RFgxQ?(r+V4&VsIN}4!0L`vdKyH0z=#WvAMnST!tyj+C*voli zH6K zdYd5jchFmaUQsxPS*qFBxb+9LT8P=(Z^;x96@L!xVSlg{go2O?pT+_|)TDt(RVu(t zaI%zmQdZaugf8j#Y^9NZXZMi57_4I@_*V% zekzyh%&R*&^I+F}`>qFHEv>Ft4S<=vv>x*$c7OM1W;BXCY6oLixtK8UA{bN!?-+jN z@MbbaGWY{B!e%bou0Tl;4V;dT|z zZ!BfG7)SiVT-YgO|7MCpi-eRB1iB1VC&vXAKgLQ3pr=`7D1%xS{!0M z3i?xj(II!v7E)aAL*2mjg4HNPWhtloXAxigyL0mk&eFYqw6NhV-dy@(KooHtMCguw zo)_{88~h}c@|J-n)0XPz^P&$Ef36JlfQvhq3efVPy|$m~_kGzn^XoOLL+8fm9Pc|L zl;PRc$gtTLUN=Io_xxVlUqmHh{N<&AA#79t<>rdF`9}7DVQu#Q>++Vpsv_?pN3A!% zv<5p~w2J#`59c;hZyIzRj;+KWNbOsW`LPy)_KuEva-SA**B9+7W$o-hFvm8)$WMh> z7^KkvO@q@_w}ZvV z)l_Vo(K<;CSB0nck)jXnv!*AR(2z&GSE&ofF(?!s(^1AWBq>+OO5C+!UhA|lY{oln zFlz0iZ%Fs?VY4By6aMPB6&Ij8r4*b@U>2`wVJQC+Ojt+BGWfzqfI&}ijmjn~nVv(t z39?vweys&l3g@K$0(T)v%Y6R>?(+Tz+@-6+l1sp7NJ4Vv^>Nx*&8#-Anrz+5Y~H6m ziENap#*E%xS7BK9UUW^jzFyyjYTS7xyXHmRSFN0`n#R2^c}9lnvNFy{578dIU)AeX zX#TwR0&xi&qz56xSxCaW*$M+YDj1r&r>6d}6kkvn zVI!_~wWukytEOwmMklA|?_Y2qDttQkbJYjt>736THreXTv}3;>Ij;8hn%ih6?g+HH znXCy{%#nOyUXO-;PLL!2wq4cE#%=_*G0|uP@c|-llYBkNFbDo;F{>nYbR$tV+lq!DA52ZXPxm zM0IT$_Bctp@VlGLewPiG)r-S`wIaTwj@m0FL>=312CA)v-S|;lSY-z_7`WhvWqkJp zK0T&8Om;`gkc&1MHr$RmE@rLvrmPe@K16o(xBrsUcY;!HoNI{5Ky5{-fH!;<;#YC5 zdQ<=kKKATsA68V5c>iOJGQ?(;UjO2dZ64yyN939vHJ5c?oHpgb?**KzOnE^toPQ#b zo+n^Skfr3@e&vd6aKxNO;(xKWMlK^1)jgx z@xA^1E|%h~Z*%R}bwG%5SKv(>;ESV>LOH{61+Uh>Dw49xrMk`A8hzA98Bt?^f-1!N6MYsvvB)mw8`cPXeOj`0ng9wNzoW$(UJ#wOnATRBB*j5}5 zl}!NrnpN}Va@uaeF?#enAfNUPXX(oYH;#n2_h<@lnekBINtyl=01t? z#uS}Oxv4mTnx4iwI_e4DQs;i)o!<%Yl;i7Y31;29NW?qgg)!*8dkK?io@@v%*Oa=LOC z)f^NDkA&uV^}8V*ro#8=D}Yf^L2aR1puFVla&whDR?6j5~k^ zTwA&sz(QyB!!LLjc&T{;4R|l771(BS7wnG{Y^1Lt6#e(2?wcJ{FL*qGzQA!W))mOs zl(pS^eDZ5C*j0Mx;_`8l4B#KGokNEWa&7cq)~@=6!WQFBkcjC@v88IO1KWBxff7*L z&FB8Fr%^`G<~Gm-j!n-@v2Vco{mQ77Sj;Z}6LMQ^dx+>l8%Eb&L{mxEF^V+&L1=I_ z?)C0VcWJr96v(hEw5Mz>;J6>Tv!;Y`{h&I(xXn=O(D{_C}-v$I~ zWI9)atc`ys=%`O+$u7pS9ynNo<~CXKDsIv!I&k3JQSqC(8NnNk?tfFs|wkwhy-~Q2!VO!yeabPOd zQvnq_CJhS*a0$sEzkgZq#)ne`o$KG0x0T)G?C&p-8WW|E0$wec;Jk{SAm_79VY~au z40|fu|0-ikEx*sYwWE(j;2kK!P%()diC6|x_#4fHBOGa(38|$~OL4>Eq1c%Bfs6Ky z%h=_oCs(a2Z_BcdWANBr%8lrl=PG@nQ5r%VGY(45iyY!-*!P(DE!X5BcXMA6$O`-C=-u{>IzwgCI!r%@knsDM*E~)O$>bZubmpWps(&q-EG)0)3 zqY@v*D?(A^H)CcI|nE-$Z+c3SptDi*mV#xTJShvZ7ZA>4`8=t zivwVCF<{}k5n&bg+S#l=9_NK)#h42v9`36XP!v>z#-KaEf)7h{XO5v&by_@E06 zIiIS7peeqiOm#QB=rFf9;5#A_Kn=(aydiiQl`ZiJ4dPrikNbh!aK1wp_~Z1h$-Xk# z0cASzWY;??U#kvS85H*6PqQ{Zpr>{^1*IC@Yd|)fv&uRhpwR2sObM!?0;{wx2W#e@we? zsmZz62L?hWEje@;czJnUl5M)Ux(*7j-l;ahs%i>aTJ&ntnbnN=A3)WgPnbH{zs8%< zgZr*~G-EtZHx+l<)Szg3zg&~M9wqMvqh4rU?tZ7PW5wN_wR)?{Ik}^CFo7!suMnK> z6AlXmYP8zB`M}p=dAyEldM<(bF|I`^wb`^tztz%@UTtXx$c}1UE!xLqA!7D+ByP4{P=dS7C<8AF zr^ATn4&$}&;xw5HGD@*`z0b#-!@H2$_4el5;D!MUyA?KMWWyxy20tYZ@OzE4cS5Ta z6a~O&KCofiNvs6W;0>2qc_p>z1?qy=njm;k_OOb->7+Ju`8KCCzJWh!iyc|%A zT4Q(uhZ_+i&@)hCJknW|g(F%9Na)&_B3*}~PL-v)VKr_7M+7fY$wSrfy*Tku0nMa1 z_#Ms#<0K&A_7hOQ*Q6%I`Fl5#n_zT|&3o?{@sv&^w+Cq!MSbU=@6@^(X$WyyqZo`r zQP~>3f9Of16n6g!@AkOV5c~Kim=G0j)cUg~I_PZiG)XYu^{CS$IWYL&@o4Z%!*UUk zp5_6G)kMCm)6@rn!4ZLO_(Bz+RuQ?w>ReaqN6cQIR;Qg+!Rz!>JuLRM^iJnAj_W~u zE?=$oo6Fq-l1&`M9Af5+dkkJ7zpM|#R!JQ(9>w}!Hyld5TLX{WJhz%Ehw8ia{wNSL zYAFpUqK_qLq#C@5(o;hwtyAaSQ$9_mM9xjIx_*>^DJ)UD@3@M&M72ltuJmZKgtq&~ z{Izmc9^i6@TNy4soHwI8#9!YICU6|3RIWa-#23*D`Yvo&drfRlCJT|)L@&9Nm$S1s zVPPI6ls}xt8=d$_cqzAewR23f7jrg$n(ex`k-R{QkY6kiuaS4V5n&lS@r0ZG*b$s5 z(MbKB$-H|b+G#XWplkAT%j40u@OG0ZIC$=oGc>@Rm2trr=pWr$q=Upq=VS^V~wnZRYAD(T$Uz< z>hsjn#A^xNq6E{e@6TU$U0QCv3Um_S{@vi3+1T9iJ*q?Vnom zmyzs^GQ6mtt+6+;>&JM4Xjv%?S=Lj^qZL_ zW#AnF1K7UxOz!S!^~IO5^3ii?77N}J$olkH=rGXWcZNFHdjaHq#cO=VPIL#p@26PC zpfo5C*?-W%GqEbs_w_8kRprBhtj7C(YLq&6lOYH_w*^1nCEuA{J5c}9F&rJSb=HZhCdSYk&MI_ zxbzXzQUM5J&DRZa>|D2B7lvCrrl}#?0bPp%P+uD=@LUyPzpC+jugs7>>ePn~hHlIc zo_kl`dq)hR2~Y)>fT^U%gOk}@?O!r5cp7f?_PgkDE=8$*(930{jo-mv6+GN!TzD_1 z1t%v4X;OWXjIG!7>^j%|Rw!JEVps5$45}TEh0vdB6(bAWXOJJrru}8e9{4KCm}83h z4HstLQkCC6k0Pf5(--1)LppR$#OWDYu-yHx8Fv|Zc0pyfW)wt|-rPCb+^lFYfDN$s(Rq`!rP;2C7BQ_DBO8doKG0U+A3sK{3nys;T zT3+3ghyM?1lE2+b+N7cy#F`J$vh z>|yUAueJ@RXvq`U*1&w2RFWfMCvvWFVGKl8n7mAn#Yjx}d%mwfWgc!$97J`bt++R; zoKQRz{8;Bh)}#74T0s^L3ggT@=bM#8A1@ieb~F(D{iCP^LG8S@s1GagYXnkZk-DE? zL?4q$FaBp_D`foIuZWI~Q7#oCsq!d<^lrb@4|QXt=YsH&ppc49{k!8R`(8LzWJ_v# zY?&Rk!{h2qW>4uH0n!56mKicvjK6zO>_@^Da}>mH5~BNlR5-~NKkUqTF3;mqT;7y7 zhyOpizA`M!uIXCo?(XhZO1e8GF1ouLq`SMj`=Sx)6iG?x?vfU1cu#!p@8|cQ0}l4N z_pCK**33|fRk)DRKqs3()m=M|kuJ=}O6*LDNXL{=#c>7= z1eZb}L*WiHg`&}La~c^dOT`#xV8v~6!x13vrjDvB%Hr2jNHC7$rbsVQ8#XL$6;NR} zAZI)CP2Nz(Cb4_WBa$Wr9_%#q7Tjcq;sR5cCkCszNBhB}}4EW!FtcO|O5 z*acVmYV;&FBLph_ZydP)!8r1V3%O*xTAdMKM;KmQTXemQ^l;;y*2tP25Mj$G z9Wj*B1JtC)`$s81G~Rf+j2QxZ@VIVT!poIg&bkb`Z+}UdDY=&m9)P%n_1v6&+qMc! z;p3L}?k<;HJA8y3QTq(`dtqSIrD@osjjkew?zo*_TR&!@?`#ddheR2^xStEEpA_h} zcNL&Z8@qBnDVoZ)jJ!mH*Tt{iIv#GkP>Q7PcFBnG+L9`Y7>YSx_mJN{)|oKkscqj9 z^U|iK5&o2p?}4|nS5?k?*AK@poSDT)Smuoo-J$yf>?2c0dINjJ`svc`Dr;6Sj@!ZY zE?4ebtZLM8Z`{~@MGCTaziC3)skY4`6daWAq>?f+pB)@j*uR^Dlf zDunAA@V%nPQ-c{LV)%F{qnJ@_yW>WxO*6O!!EFdsUU|mHi%=OS z2lAp1)|t<5!~t*V$R-_%>Lt~?Pj12XGc1I=m!ZQV;-65F)L<>cm7e1bTk=!Yh&!|7I6uG_F-OvtG-{pp0@|Ej}6K^LU>A44&Zll zWZ+H=(NHR!$aHv2?K`UK?Qgo+SgVkQqLMHquiby1#%N2R?b|ED952!!yVY_xOHlm_ zO$`lGg1FJd0OiBRA1zdy3`dXeelr2Pd2Tb z&9z?Gd+M{)FYWdw(u?DNVI=CdnOTouZ?AiKrYC}Cxwf99$0RN33~-3=*x6Y1e01ZR ztt(JRhs?M%7w*I;a^+)Nt({8G$DOsy#LV#H7Z@O`_u)I>V4VC;v0v+Q)_j98Y`xwM z9(op2%w`y~scwwwTKjMG@G;mNZ_uwQV#&i~!r}96RJ`3iZicw|+6R$;%!#3AZ3j~$ z?g{Ta`(XAeTdFJmMB`Wpn~KK^MMM6Q&K0d3R+Y6f&8PBatK3~R1e?Not<^Zy1((>4 zQPv0D8nHhr*~`_~gdpz&%;s5_5Nj#YLQDT2f>f8OZhqaGajqYOY>p_9FeHDfBYNYM)hFZWs44t6n`uNTc#wiKUcESjb3lmk& zpPu|wvzK}^vnQS-kwmsnKFEPJ66#>Nugz;XRS)Jt!cAbhH%-;wq+`6_+ITmTk^Sm! zLaKk%gze);Dkc8Q5ZjGzJw?HLQ>`AquIA=u`8!+rm}6Z%dKm-r@Ta?boxQ??o=CU! zi`6ZW&3Esa!FT+Q!hw^N`KSV0e?3e3@QK;u;+V-0UM@+dKMYrQid}j0Y!b{Pkszuk@;`|GA;hG@&74(W&^2_-q_{YCl?&zNMY~ zSy*CaVxar;9G4p@0n~Y!jPFuo)pR~b)LYIhD8vc#>Qpj#J_H7TU0WD>t}*%a`m4Um zgZi`?;+J^G8K7Q&CPEzZKw(s|B5pI3+jtg9uod(#7b{pUT*ZWBH}PPmebgWXwpiBB zUuI$m`WAmrw@)y>Mlk&qGN(#zL2jT}FfN8IZwWtZb@XQzjdGM$Kiu^7YUuh?1D0>U zUt#y+Fc6<&S3ABgaZ!S;n+(7j|B&q$|9{A)7F@rLEG!`oy@h~Sy$R1Kg>{t0sElE$ zAaRSF2xUIm=ix;MDO%2RY5UCtT>~1V8p?hW7u-X%t$@>*6v;L6qFgPGn@5IN_n;8T z)TWUinu($D5?JKCg^gP77(jK6qq)OWHRkH~m#AjN7MU@=WCcTRy?m8pTnZp(wp`h% ztiDHb&MS7`M*4zbON)!e?A!Kmyq-@CSncfW)bZ6EBPP!_Yzw9GJhvE~7iJBTMET3b zkwzGj=&8&O`;$$emsf~V>25nvlsD@O7Nk|JJ`!tea*`^I5Rk;f6Tlu2J@2q`VCr95 zB(jcrc%=Z`NV`IA7r`a;x~2$wj;Wa?=DSDim^!L(-*pTkzmTejN1DGca%-Iv-dcL; zbVhh|UAY%=U#`>R%lY(z(~pBaN_E?%u+WB5=h+{1(x2fVd4|9C;x8^n2W<}p=`8O2 z-iZ=-YdP5<1-JN@Uv%|b*c5s6jF7p_k7%snzg-=n9RwYiuwEqe3fi?v?$nLnDk0eO z4WlAGo8MwLKHkP&H;n1Pk>vSor+p89420_2&S95V+ih20$0qBE7JSLbDf6k=Fc?Lh zr3PCXK~vJbH&Rz7PgEyE4)u3L#|xv4rC5b3d*r|%2>PBHdg}MVG_91K-k71^wDXKDGlendxd3kKavRdcK^ zb3aT{Z8i!_3QZQwApup-98!I3_J0~iz<(I`fg0RdKgMjaEYWZb))AvGT_IxzyZR7B zA?dwy7H^jBXUIubvw*NCaEyU=kAB7z4!sc6=V37l<#icU#R#r*gzOA0= z+)qu?bB_8bfn(W1l6*OuF~Q3~ta#}dRcf#hWutSSAOG3Ph?5j&E*By1dxY-*^MDT( zc=1djn`;DC@Z$abbi7}QEYsxMG{vh>Rv(dX{I%iAOvoe2eCJR`H}pr0e>eLMKGYF@ zg)$~woUH|_h&&TpH5L5igIl3nAiwlr6zO3e-P<@F3{`TF<;);yu(lyPMi5^xG0~U) z)jgSLR)wm|Kd9IHifzX2lR|rsN#*wl#WrGkuThs{PFBoS9RvoIr28LFTTHZOyX$6N(3GPgZK&R-!~%2@?9 z*S}0`$aIEej=(`98M1#-(}Un}_%({%H|B=Rb1~3+TmMZ}_nUpL?P9$M?BNgZ*K8zJ zk2PEDJaPz;jq4BKT54`6^O!eo)!l%_CyB0+8FP%Cw|Q|H_AN!uakJY6M9#B4KT&$t z2k&C7B7sQqdlH7e=!$PIyb?_#qgDESSaQlQYBIQLw)xWSbY13s@+qDX1iRo~9G0Mx zQCw=&?J7^a6S(4^M-mDc`c6U3;4V>PT78N{XgP_bns-H7d*rFphI*3tQJB$c>w$#c zeGeKg_?_JYH5(QiUc7@22aRz}>Mg#(=7?Hj9u!b9^;rNqJG@E70pyPI|D|vEav(Ej zd~9n(=!tqtFg9g!6bI~cWo>Igm7LaHDGGYLpgM^+10~#vg(va4`D6fZM z7YK8n5J9GyY(zr2Vz~kJ8*1o#P{yCVQl1cZLmFb-MKjyj_-QTia5>>-Yb~`XcDTFx z@$CRNx|ERWYOrZx*q^~=s+I>`E5RFWcxFK<;}Lf{I<;kU4;o@8Lw}XTlAz3?{^5yj zhIo@-m^j?w-4_U_Irtu;1xGi#5ZLl>G1 zIukZ$PmbPUlbieoe^+C~8Pf$GLi8hNdRv^`^N%l!U`^0Q)0n@Ch|s8x|#6oHLKl99vUCkvtqW# z`LTWrih5Kcg;OTdi#Ota!cF-%1B1%aH|%mk^M!5^9U4(}2%=+EO=p;y|QpDe)Rz~;|+7U*9YjC=fA!%aK(Jr28 z107V2Mb3CHip3o^}EUWPbIZxoIcK3e88ZTETQu; z+0OFJwF*-znA+|KYp)%`jR-#`S%3R-{-=K2y}}yM5JHpbbl2Tw<{(P{E_{{B;9_3` z+SDy$-(EJl7zc_qC^xO7W|7>X=OKB9xC8GwVChHe#>*z^i?E*3G50p;e!h>hCD3lW z<8Q+dNwQvt`w{yFJ>VgaTlKwNo0*$)QP7rt?#ejJ{^l9xkA8e!PP znEOwRxzS&=dWWgsMmD_h@d4->J3Gb}|1?N8sJP(&Flv?@+<7H37!sUh#L`mzNG|i8 zA_+=MLLw{#J~-Y#$3E+#kruB^7UX2F(-2fYDp8SFAm|sk_7P`MAPA7yoiB70aZwy0 z`=H%d3M0VMay(&fPE;(ur6E*V9n{TRoJl7+ZLk1klu?bO=smuQ>bDThJDs<)v}aa6s!xvMV-UbCV332DO^&s-PyIvwt( zj<50QngY2ykHJFS9=uusw@p3eR;8_8*XCL+wo57-{K+E6mFYFLbHZeY-rWRlE&ND< znPkdJ?fDM0gGR@wnerN8VKa;U)MNr&7Z83g*+a}`ymPE6CC;BDnIsqdb#X`PKJU#a z3mR_TCR$Zv|M4FCZ7vN8-K=DVHI4$zj{Gh^RSe{0WB-8BMb_|X*L<@+>fNEpfLo1egzAp;*GOc`7|4efH2 zim59n7}JDc>p|cMgDPR+=+KorsB(tl5r6k96tL7I`tKF;r^L8dkdGOw2FUg57Ks_3 zuq3C=WoRw9wMau~qpTXV4!RvFqK(3R&F0nCO<%kidaw3Mx$jp-d0RaV`}w$@Aolm! z!My_DT`}4$_^!acMrCk6s<{iyKWgpVAwF0B7T{xVfsr8uU#(uPam2SdoBz^m&W-bU zl-Bq7P?D^Okz~6%IUsZJcR(l=_M4Bk{^1h%gXJX;PZmNksw7AO_sFE&1(J_41p5GY zXvWCvCX;)tt$Lll(oyvTkI!Ab571UH*Zs*o?yG<5EgR%UxsRV^;@$X%$wTk-1upwl zx>sP|Md9T1;(P`Y2M#))z3ZFIM@Xz>V{l6O_Q|D^H&~QFrL$s5iY~~E;02vDYg3>@LF`iTimwC=@Qir{<#9yaw{2&$)NYr^ zNg()shE*sl*jGmq7gdPZV$!`ur(I}Jwzn-tt4w`E*F{?H(S5obK#Mwb1*GO|1x zoVH~pZK)ZPQR*PrNy&#O)Ju!UgDS!WxzH3F3*QY{q&lxrb!S^mN|ls^E(h4vMVB0c zg|oyQkW0`~zfGV4Mp+hKF#4ledx?izqF$+-^z?4VFmR2?9LYS*Lw36ac?ROvKhexW zx`I9w-iNme0jMoCJk>ay-0zMJj2xQrp68XG zkwUTXFq;m?Ez!1nU~+=MxmSGqt@XzzeiUv%1AN*MdMRH0a#G_AY*wo#95F&OuW#~y z_0wuF4&9eZx0*xoIY_{g!Zo%mpz>W>yGlOjp^CTf(6f}tJopeL%6mQ6|C_8}>PidZ z`}fRXu9|y|wik7Sq}pG+8D77`obJGht5(gu?gD?N#IEiq{q^&F|7y9F-oM=T7){M(vbLdlb`P;H5ht;-=i%X!$8-{3H(`3U|vAbiKz7_>N~$piVi^ZS&$v4 zD9OII1;JphCAA+eP&Mph(0{E>QtiaA+@t_twrK8*MuMV@R7*EKxTW}DxhuvU?}=-9_pnxYXz=z(-!0$*?IOmUXGz1 zZ83c#d5+ar(dEx13S(cN0LEHyaPSP^nu*3?o^jal)QF-99=H0YwdDfjQMXS-AE%=} zLP?UpJ+nHV9;;vXhYVjVEv3If%WoE+ZQgv!P2jOLlE4i(d(b_^1+zlMRJK9oB-B$N zjN3l3u^3?orX&A)|AZR^;2)(Xp~Weo8sRfrzO_-voX@eHO*J5A<3Xs4|9AM2K!t2m zGYG*HD;1A{&Au%6v-Oiae}1jA;beaHnFcl%RUU$ogeFSye0`Ce@SRpT4$Lj8Jyy2= z+zdjw41XVFfvK2^3~Pt*GV7~ABnuDp+3RY5C)KgElv6<3FRWQni&^2pL7W9Q(66hF zhTGl0rz>;BB=FUFd3jq^v;9$LS3sVh7S>Y#Wuar8`t<_!mCWF28+Y&2!QWe*85~3g z{=?24BGzQILZ#ss9+s} zZB#L5xKW^en)#z99bF@K<=Z^R|72h6|6m8OiuU!9kL=~j2@ADO;=AU>Y^SjR)Aq5$ zF(8MO44t0d<+}7qXos5&&V9C26A{S+a_B}C?u4Jv?gMP`+ro_M*xIs*tq8>v_|5bk zSE5!uzXdRkK+%1U?THabMV;s%v(zx$1&PN^rl2T7+>JSd-r(lB?|8^ocWgMU+ES}d zL&R4X1E^!Yy#X!xx86EB*}A?ra}PV+Jw1fbWbtBMm>Q!5!5RC(IV^uiF@zcsZTDTeIzmv?y0U*}sAe{!9-3;Q}8~T7^ba%UaXe zxYdd!4Ek#OBo^6HC+UOX=$QGQD2^mJbiT}Jn|}=5_H!`Y#ri>v>iX*oi3z9I)(#{yCjwQ33K zh58)`SLtGV$T4=?54=#)e;A!T$+yvyOu1NkzML^T@6T-XfAOPlHOU#629I!(Uw~+W zp-|eed4}Fd^rVwZji;FK+mZcd8STICoPBOpT8k3`>|Jw7M7AO7W|K=V4R2JpVXDEV zDG;E#O@UQ=R{y|}U--YL-4LQKh^qMoZ}$cP@!JG3)JRq9MEwpE9|Jp*2pvbBJX}uw zG0^-55GlR6egt*#)gW@V4^RYQ^%N$hK3EHw0jcWmH8gZy)z4NQCOPm*Q?b`Scb@SM zZTXfB$OzQok0VWarDyGKz=ez@SMSx}`iXwRJP~JO1jaK5&lMj%d%=E|2v&fpd^g~J z0R5IAUPGwam@2&S2LtMa6!?3Wd3DNU?8*%MB5u

X8uot|04n@H4v+QH6^dO0c5VwwQpu~hvQbZLcK&dh<0k3Y6&|R(5#?)C)HQq)*|e&D>3){|KNSK$3b zh`mskI`%s0#~;QV80w!63;AM~e>N?!AQ?|PB45{1Cq|Ff^|l}X8gxihE5XLZgJV0b zD{9j8Bi^^=kS2D|QumVl_wgzk(*Tj=LEOI&c+;f`pW!x%#d1+&ei}{vm881zPiz zs*)^Xs6U3rlC6gSMEmGHlSlOFr*CMkkcR+3eCQ%&?jmPP2hJ(tfd)*Zw+Bnlqx}Aw zL`E+E>Ce4?`eXj1H$L30(YrGKoXw}jqYH?&cgKaXALUQDN}jQ#tcviSc`NsHy>_5U z6I;GBparSWhh2Y~;>?4bv%;Pyi}u zZ#yQ7S9r$M5TE*)sNP>CzK|=iB-E^M=D*+6=(QlA8s#$2W1$Sq@jh9EVY>eL{vYW3^!{1~ zt-M(|HCLn+b@}R0(_rEQ>L23%tiRA(g)%-W(=#-vDqcBm=!f`Vq;|@NQ@b6Y*!Uc_ zm(&>B6@{q7hp6KOA;8zt)rM|Ow8ClL4k}|ET(v3(lxpY$)#@;ueC9|wBcrd!M4kkZ zJ50O8y0OY7AAeYHPh{;d{zcB(LCFXoBsAhg z9I~NNOj?#5uBi+tbu2;Gthuyi`vzn28^-Yp0FKzwk6^x(y5PPylBl909UV?)-JnLb zy$AoEJWYfcp-<4Du2&;xs=?acgp&ODE4Kidi!SK%10;X(eSOZ?C)qahXV?;IzaTjh zhIgM^A)w~DB0dnPB#TJy7)iy(zj4(aN^54Y8`QDFZM7T!P&MhE<{Uh%x7p^9hwM3b zH`%Gzr?X&abskz?PsqX;T-H5w)Ll%|el{8oc4YW!%glNNP7w`L1~iR=%KCDQ!YuPp z@?LQ=r83HTg7D7dBYoD(Xa~R+M+e7!%s zsres+0<5fsrRym;b?+SIf<|AToaR0mXW=)HAOZOoX0KNUSmSI>J7nEt7A_=pDBU3f zO%g|zs)mi4OYe6oY~Bkk;DN4_rB9p{L7btN)iz2H(Hml8|5?K%4T zgW-xaCQbxM#{J|k0jn)koN(8XUm&Df zCXEVeOlDLLsXR;A-sLO^0sbAx@tnbAt$r8TGnoa307l$y823NV@@L+E zF|Ss;e;!jvhT~qKT_bNXf$L2|=2UKnE)~Vu6RsH#I(6nfUdyOTKT|jsi`_=a5)=uo z=5^b~r_Xh^*>enc9$DL1;sai5FUWPY3>&CpmD_Ba@iJ|}JKHeQ6pkIBXt?jv?2qlQ|Y)zrJ~Fjz4qRah=vS1MP?}>mHXBW zhK^feNHMb+emDkduvHjJ`0^uAU|Bv%maIzz$mIIKyx{-HBs@SSaXxETtt1ZYLaqw_ z5qg2-{nH~vQIpv`5g>r0Mu{EA6C}}Vo8Yu$*8Us8|swRzY?OLrziRwJfrGXoMxXWtozBmWB#1CZ9pU@s{7Tg0;Hl;+VgGs~4^=Zn18gp0I%S(gX2 zKZ31LAt=)LG*yV{s-k`DG&2c%K~AmhP*g_PL7~{Re9m2r_A>mT=6)dlaI90^&dD=s#NgUFT7 z;nXlR-nK&Sy^Z>KdZ>GCr%IMN{PUn@iOF~yKWDB)a(4%dN|F-I(k{Jy*~)*vo1B$$ zvg_o_Pg>3)l)s)&$Qsw9(ROp><3rLE)6&9aWo<3_D+vi3bu?I~n<~EZ;v86T@yBan z3!1o~V*KadA&P=?%|ZBJ9=tjD1Gz@9+t;V#d(sp$1&fpM9X4>a``Ie}RsCXbD=65by`F`nI@V?4 zmW}+THdz!Sr9XUH0uyunj-+DA0NFJwI_Bcfrx+Cx~TQ<6(R-hkO6$kSa!_m zlWA1?^E=CmgtV7@Kc&B~=RIk)g(8t6j|GBBgJf>UI=AH{jBm#G8Dv}MeU6;uXm4-d?K=0Zd)gxH zFSUq)a@Bv}U>y#g^wIEZIT2z<>l^$g(>uixF?pjVn?dSyuejHtYJs*RthcwRw~VgO zR3~))nw)obv3glG)8zT%lkX?sCooHC`!3(4xY|`u=)4n% zyv5Gnzt@;<1s}DRoDbng3R$R3*s=nJp0p~&g_)}_1WVMdGVN2KG^*Pl#0s2$SE zE+9^IwqM&OeHDSp77o&%a!F_seR}Tvs>Z!LeShr4A7xCsObXCu%6k>Sha3PcSU|tp z8pK2E&fe#-19_Orz4eCnpD=IvSkp8NP{GeY0KS<5bWP@up99uWnDASN_}^1|mt`CA zN4oFi@IC){tNJ{sME)g#Ku;y50Uy4wu7+XN*CWt6^0V za+5w{t2&QleH*@~?NZ3X1j$w54T1NU@XNNcEJz<0hjyb4o{BJX4{h7we2$mAJ9p z1OlmZT(a9}Xh!d0Z*6kVU*xLfIO_sy0;4(MTq_0kH=;%+{oiZxJ9kFQ%0d5@7^r`v zweJ0Tim}w8ZhWDoC>kj^);5nIRVdb6&w!~BIY^eXEODhW-V2$Eom3%?*xWeG>%TXh zvWszwq1~jiJmC{7&ZpG(y@LkHIlN%172h>0C6AH`55g_(T|Rl;>ZIi7zvmmQj2ebz z4D;WQPfzD4%sa+M-Rqf{l{O$LW7Jpx1KH_a8UMr6#>jg28d*Cry%-bTMvz!Qt^M#% z%tX53N|*H4DBXrLNh@kC?5N|CciPPl$_h`?o-?Dzh2bPgxcZ(V6Q53)cwe8cZ~RKl z@>p%S&J8wdt?dCFf+j0XQ{YXLGC<7$5fc)=pM4HGH())#Pm*DqCY59tA`69N9obA4 zO8uimns_lbDNW_^&IYu{hKqjotg3=73U;dslwJ;Yp>S^MOlpk3p zDS+s(sJ#~V%1ajacOmMDCkqu=t@XCpPpKZ#R+m~c?OOygwfv|!TEt}4j_)sr zm8L18VYCP&e{_%0+M2-q!iA9C4Ja_G;F8%Hfl>l-a&pQjDNV8o&#^f0?6C87lZ;l@ z*R>hv zT6+AadkRsJG4U==gs1qcrs5s?3O1KpYTO~Oich+r=r+g&>UvWOfZ3#+AIesjptj;J5m7z4`7%nMnj6e>0 zlN5q`&3{&y00&u}JMmdS#f*&htC5Hzy*VsZI!TJ-N1AE1Caj~E_}bTt;+xjm!0||4 zeiq|X^)D-%yAp_QZ`}X5apkyu1|TACw_1Duiq08)|2w?QweDJ~8czC9DYp8AjB%X{O=`qXtiM+>FQC>#bDvKnoi)jW?G z!Jobl!#I`A_|pW-Lo-LO!2PpxK$y>mJxBuRw0tBqD(DJv4nA=T!RI~Zg)E9z6c1Y2 z!O#`u%Zj>Eff`braZ19BWU>xXbg#-y(^8;`22O`2I%TjL=dI-tQD$uDUp({*`|kn! z2YWFx8eqs&Q%I(}rB;b5Uh%a^Pwse(^ftxIGNq=MVW#&#Z%E4IIW}Nrw-HoAIM0G~wY~%%<-b8qk26uYJnY zLayxlc9jInK8z${#-+3rtwOXhL#0nKzz0?Yb){mc<@jKh$%iw_f~+Wrc7)3JEPc#P z3eNx~UwO?v#gQ-%)`Ypb?das5WuRtqKDP>FLods&pbqgNRePdV|7 z<4Cw63#eEDh+_L`dwWeQx`x||n}*ZB^lKmIt)eDdYG$t^*dtpG(eDq4X@MwTVB`Al zRE~Oq%t}WLFUP%Uzq9A+%FvZQ-lmy?4t0Ej^$`3uOjsR_EYN4O_r2agRb{0N(B>U> zz4tYX;N9jTl)tA|iYV{Vl(SzS^MX|$Y(*(y`B2*4p8DuJp?>CeYfx%WM7F_ zy_E<@oN$qr@7VGGLXNb$gO5C}`hmf9ihj{dM?A1xYCkP;*CU(3Ga^I@cM%6y!rqys z`VF^Kh8eDG_@*<;THj&b^tx^l!zS}z>=2|KXZX+h&H~cf7s?$G7Xcdzha{_8J^0Yi z5jGkx@#aJ4jm2hF%3=GU1AjIC;+G-!n1b4?X#LsQWoZxTwM6e13R5CVFjoGTXOoSB zFQ;t}HQ&%RZdN=rT>f`tZ{8OiO7MPJN2NJWDw-`e?9kSJ6bqTOHufAe)WlXk%CK`Q z{%8b>V0yKUV(os}2Kd(xQNxXH-$vs>1i*JWr_Xr#7vDinAfnFZ_$|;wwV$=PEOOh(iY^N zo71*^Gn1tY&%`6V17mf_Fc*5YnKkilcw23Wv%p+C&NzFA0*K8{)Q_(kbFn}Ylx(D3y58Zwty*UDbl7Py>BfAEjodf{;No_n* zxB=AKa*k{`_U~l0E8W8>t9>I$Xx;OuEo#05gAY4}lM3C#->oNNP}rS^b>?I;5I)h@x&Oae05U#{4@7=s zU|HQUgbo~<*nknT+Z@7@^)Ec#<5Z@(p=Q;`(hA zr~PQfhbg$v_Laws8T|cEdMK`e6X#7$8S}|cTr0O;ql?->Kg8Y+YN7n=It9=k&=n5$ z``;v@+@FJnBn84$$f{}IN_ zfha2b^PeeJ-Ex@Y5=hLb65tijOl$nqmLSWY9P7GWXRPx*wV@`7AA1`k@-~f@gt-&7 z4`wb-P8fx%4Yv|wvA$-TutOY?3W+DNbu<*#bZBeH^m=R~?X*j2xDXg&%{Zf;=WSy7 zFkOf5yVB8p>6+-h((1C;^YSaBxp+I~OJ4_+ApTBUNhIgO(Dz$7LV~%95&P-kFW34t1 z7ewC}nNFM@d51OJILE)td6j5v9gT+tg0%gN-bdszzDcW9l7?%LjT%vk&a3946<-0Z z&(~eg&)RnIi8m8!<3xI?IK;0}?>oG6!)zbLrdd;@Y>mE!$`A7jVqWheE?AfKZjruC z+tL`aw{Ab2mVeV=fJzng8jnPe`&9 zS$DOPx>e*a@DtBJBGUhUv6CSVlmy-e%KWChxZ8;z>)3dV=cwAoC z1%}oA+Hq`+TUYRzgJ_g5)YMDOuI4qvR}~Fipk@=mte8InIU)ZTSljd z{z$PV=_?{@>TFt%B+kdW_afNWbB19!=J9n5FPZ`U>>lX7{h?JTFwt>qrc{3cAe*U%qgYmlgA`DX4V^qzpm~tHdBHch@rFdv9xK?nU_q zzs)A$fPQnHLx6DR{;Od?y+KX9{-a!siL+r^a0LESfnthrl^Wc0Mtauj*F*+IU<(21 zco?lp=@(R8p0+}jB+Wfr$=yYn0~_@JQ9yc#o@>4y%5Q7#*;oa_xX}fs1>*h}!K5ju zzK3G%2AUs)=I)XpMO|sNQ~I9ByH zLgT+f2{t+{VIW(*P37>tOE_ScfeOLs?!w-F*F7i|Zen18pr~BZT|V<(x*;Nja*EXm zUoT)>1i!Z?%guG&-TW{5g@z(iBnd3VVSXmmI_mN)Rf%h?N;+Yw)aYWStz=y9jXSwP z0yHAd;c~+V`ntrKW$uc^FHM5W#v$syf0E~B{&B>tx9VDG)9+Rd=?us`T|RhOaS@B} zybJQpp;`8h2qRv++VoNa-Y}`eu*P43Em*i%-eS2_)_Y(1x8)C*Ibi*wh#>}g!*MkE!y9Ge{iGjTKg@IjtAET3_!lj^t~8-1ceR7@vJB6UAmg zG}OQ`M-MFc_Xrw3FrUbfzx4%G%4$)G8R7up{G>h%8>vo?to#bp!@$2t8Jta;EWII! zClp#5vEeus&t?{!c>ZXa+3UOW>?VtxQ&Ax#&V+p-*COk>H-cf{43RPh^+(=kzR)J9 zpMd3udyrJ`i508e28yaxiNCt-4!J)=uu=H9xmG@@`sY(W+)PY$8K18$EwbL;9c7dI zjbv4Ea0@YO9Uk#im@>Sl&9f!d)i!Ee*Y^NcT({jGl$`<)ihQO~^^dQ*0RRcFsO5LVYZ22>zsnZFJ*1TI9(_#I9U7r4?W{*DeSG{1^sMw~_eY}8E29RYSlVyVzGLg1 zZLV#9JnJxH6Vs}vh>7soOCC!8kwHcakEROj7*71G#4Ev(3GTe;M{G1XK;iYdV>2C1 zV$x|>{)c&Q0Fq0{NcGUrI+{}EGKl+-@_i=npVg4RS_tcI#v3Mz*>Cs?T0{KCpKMd z|P*AEk=P2X(i%O1v#1ii9XFF&_GbX9u zh&DeQBsQbx%L|5^kqUhQBcUykW=!KyF)pdr5gmG~xRd&+{w4KNL_z(X(!J`3Kl)dW zMks((7nb7F#zJ$P*~@N@TYPe1YRr$cLU~EXwvdS3cGpHoW2t`7dv+XHnK@uM2Q)xM zDWdCX=C=RNn#w$WvXWz#1oLo^6KmvI)ST0&MHr8I*?9knd4`Ah-wO>oRA<})y(Pt@ z)hs}{8xA5cKByjH&zTHtm>3%wwbeB_x<2VIj=^dVVI<)ch>5)A1nC^?_jqZifg6+n zO$Vv$%_y8tNxoMGcBG&%MRU6i=gFWOXhxUC&`SJG0n*fa#Ifpl(oCuEXCgqrzT=G) zfkq~>o|5^HB24GVraU$^d^aXA3YVjNRO@6MH{sZ4D$NP8t*$?N9TEPtDkO_JJ_Ma- za&f>=cL7eWpv!ffOf^s}I?8YWhWPKX@3i-(X6bcltz4};CQX;J4@$cqD;s=Jfto^P zh!>$TCTfW#ky~A!l0IlMUSC`wFf_0MFJuI~CZf)6kw%I#qvBy-N!-79oTmx4{;vLo ztre&eqN+vmD=PL{>JcVSTa{i4R3r54%DcKki7I$k@COyiIF4bjzn)32Q4N@;i`rlH z`DFW&ly(eWv7u1#?#JU!+p@-dq+5n|7f^81eCF~W1C+&=zg=Hf|6u?=IZ$rn0IuKn zll5ng`g;`_<1{rA$hd@G>Qgru4Y(Ie(%A0dIwv6?{Bt4bQi)lP$2iq{wPh-s01I+W zD%?Ed$mjmBO$3XO0p7)l-xQnJ@x*-Y!kR>sx{qS4CZA6TSr6U*axC=2%gM<_H-G(l z1m*@b^xnMM>g@(Lvt1hVo`9Qvj15lUL%77fi_}>kc&m+Be*DD0l8jBMrOq0H zeNr7?Cpe48KVyOA|McOf2*T6Fr~M6Q+ge%ZZvmy`yd18=C?y`ohFfMw#9w3{I6A0v z1(=Rj-|52cQ|pY&YKPT>+(|CC#NfQ@>B19TjxdpjLVLtpWiS$P(9}<7a(*YYl(8IH z>eJ^c*wpI<+7REB1mw)TXVnxTPU*L`*O*ez9bje(r*>kCKBwp^I`}>L8}NO-MW5e) z%lG_z{zV@H%4j3FaTW-Fz-%Amzz1K3;o3Fx!t8wLLYj8*H(w4xL$lCFt&yw|K?P*6 z^_h4}iu8+a944}e^Zx$+{14Z)UTdwr_S#f~#r-}C9EyFsO#RqsLx&k%nv=kYl zWPrnxd!n`-`Ia;iHT(Wj(n61etn#=JF1ve3(#-U!KBv}PT>Vm{yd;$9hx=&i7>-*m z@98GJU9Sm_d3HbZPZ6+&&|)a7^-|M+jUJ|iww%%?v6V_7BbwFR_hal%rDmLZ2Ko@K zs(QcT_neW9Bm63{_*xxBV{GoBiS@TPJ;471x;{lxbow$LHmn%RpJ&efr(dD>Z^PN~j#wx9VzbWMu?@KzS*#`y=Tw0im!8d@tJz$I~ERR|Cr&q?-8uhNUW|6Zz^A_u~imgir}nK#Qt8Kg>5 z1g!Ux$6^dyRb?J#M7ZC*CzWj%u|BK`!=bjSO~f+bhI{G#yYO0VPG8pe(?O7skhRvU z?8_~dMR(tn^t9EdDEm{)GU1^kLRV^T%8G-|@gcxTw*dAVSsQPgzYifwCy+w7Z-cW? zwzjx-^9muE2|H|WU5T#0aubsblbK1_qXis0Mngd=0gGT}VNHPcf`N8?vE%nrsi*q; zEXv*9@zfR3XcTHlo5A`0hGGffO_D0*$d)ROk}9<-7}rkio0!EY2s1rY6P{b>u>P84 zle2B6XQGxO^PS4<#Zn*Om(R=`3g@V%#y*1b`6NV*t4!vtRKv6#Ssuz8*RrTr)_Onob zSX7t=%TkP=D+D|5)m0=N3|;ute2A<7qP!6&lVZ4Rld2-AQ(d@(3Xa4jzuzCLGNqg) zcm<6VIRwk>P2@&W3zvntI8WYPbd6V%7DnET77@WUUtKNQkVe6Oi7LH-RY)u7J)%WP z8IgcbB6oC}BRI^l?KiPNW5VJ6pzKjEF;05=o+Lx2Ed~DVB3*DXW%FGsN)cLf*)bcc z92W*GPWnC8L|nMA$w494jyKm0ZveTSJa0}j6U=Hz+}ey?=(XzRfs5gw*U3(bWdO4c zVUndEvi5LoZ?(UOf#(a4KtDwR>HH#^l;6moGf1eaz7}#m$2*dKN~aPN-OpAA3caGz zB3Nv7R6hDQk*hV|PDlUw9dk7G%1H@?FMiB+xiEiux!2??JU>5LpNljzwa+H6Ovh%n z#s3i5;y*+l`2vN7zWkhg_Wbw}ACSX8oUJtM4kaz4H5H^OM;Tbx?%;r)YYzVPnyeH~ zaj!7e=L=QL{HpE2`-#4=774FbOC{KNOF%S+qJUr6tAK$DrVg`OVW@hLBWy zY!SjmrfI!&M+3jYcex!t?7^-5{gvIxqKYBo0L`vscwQV%6$h{NxnDhX-k0g&w%_@3 zv70aD__76GhHPdf$lhj`{RzOD!=YVDLKE+dpejg5^eUok&q+jAr z)$Yq5o2<+v|qN|0Q;^!VP{a`Ou~6hL`;{PsR57`^<0N!R|w?O|m8DO9uS|o}v9R z8>a}3iQ@JsIWDj9%pf>=rCX)g(3kd+thQXYeb8|lUI4n~#c9oeC?6=UCBONbQR637 zgJJHqXmp#q5g`8es|=P!>3wd06Zk}Rx+3a=e^;{$jyA#P1;Ua%R_?hvOCrT7lAY8t z230v>eS?G}C{@wCthK$Oj4^mu)t(Wt7*^?TjZdlECo9T;Rmo2tI4nwQO^y3%m?I)n zdB7l*Zs(IP&o4rjJT-vIyN^VTZEGePW%mqWSAu3Ki89?9Y#v&=>C)owiZM&h??{2B zE^k~QBO;q9+F?W!=z&w<9}-L9N%QOISvxgyC3Tizba{*0&gViJxhy~K`|t0fvDb+T z;Rg~UZYnmB3BY>8OAZ=b50G>ORaY(?H>xGBYS zk4AvhfE5`txcEMYhMkb7QgX<7<@@Wl7?g$SGfo<>F2u%l$E6JRshkU6HQZ<^rg|n& zt|VS&^se-4tv&<6L3bAFemNz6h_uEcgjQ%n>0Iwg4oMO`ou+1vuuP)-pWczA0~tJ> zQUL=QorM;Rl-L`<|C&z2iuS+%XS=lj+iv=Y?!h?$IG$v%{D1k&d1k0X z2p6h-(Ri^Y`#iwR%zD_M{I^(TpYwoCp|zTSHT-)aN3aCMb9%Drwqd-Vwez+K=kHM2Ufa0_T}iPo7VmW9<>W45Xiw|ph9r1e#OH< zlV&Al05FZih6aI5hr%i;90t~YtqDxMi6i|&C?!%l{uQb~n!ea5>P-~0_spi$eEvAW z;RWGK#RwtC~ImSp`j*+zk%ak`szRT68&&vs8# z)$+uM@b1CN4G3qiDRjYD(RevpNPKU`Ty`%4 zGH#njvYPZ#_=)oc_c%kn9GyDC;}yXQETOH~Uoz5viUt(p26h6hW0^c15M?PLP9SzT zd1%i{_NwZ`<+Fd8+Cu=@;M64?QD2puorh+#jJ093@wuE`n<3Pt+4zsjKs8C^_PY)k z7|LtAY75}|KAIrk^ci=w>0`(GCtsA(Clxb?8-b5n{;`NPV)A1Cpg6HD^@9~KQpTAI z;?-RTHOe&GR(~kyooo$F*v5J21!%5qLK)f*Tjr?9>E!zIUV)QQu1vq3P|81w@Re$z zSX0hKwf&h3KdmSog)|^6MHa8dvTQf+GcKjPP_FL?M|z&WYv?&9&}#=f7TW%#P`!vs zuUI3`5!Z{YY(ypIra+}NV)Lh&QltL7KFSgU@GbK&#e1i}{1RaHm`jDrk8gAM$gB%e z*oa7it$I4;-Y!K#JU!j|wkg|4f&Ky$5mcLzCil-OlXLiD^WTI4ANQYAru35(Sbxy; zk<@-yf+{?9mUQ5b(aZdOP7DEKe2~;#2?#3%Jz64qP@T_27sfzJ15sqGgeTn%jYV(c z&`1nrJ)uwMzmt5Q0K?@QY-SZd!O?^22ew`P-`fYpq z+>UKw>iA!HIy?1qFToGjNNs$huD$Sc{P*zcaPq?($;X21idj=L#@3e=HDg?6CECWk zP=q;2xYI%+pL4`jzEcB+eVs1oecz@a?*lkji$TTO7?SSTLjQhw>8Csu$(pWv1bLgS ztAGsGsDL|c)<~f)SSLw~xNeJf0deav9;69u=ncsiRT&4`r$-^)u$MwIlX_0svZ7wZ zq$|lV97U9qkBU_`Pn5*Q&nX>jp%0s}08OLO5h4w9^~zhIHFl&~jXDdoNtK)n(u6V8 zp_{n_lXwFlPO!B5b8`%kkx3cLZ{uNGf}8?!r^t+L!_uw#JRuIRqRvoPL%b-Dn(icp zpF<-tM3R4$ELGF4iFgzv9t6RN&b)DlL@($X9nC7qK8N)it@k&Q@++CV?{zH^D2~R! z-+-y8A*f`T8p^cr*^Z4M`RKkUy`ehv*N4+(&=9>3)Dk$+A zNuWN5-q1d^2}b|k13}1tyErU6*bvf)md!pH49Ly*4>tEn%K%0ilN&PM5 z22=q|Y`q-_1bKt{N0J)sgqE0&%a0BVi}XkZNrMAkk+3<6Fh-860-;D(HjztcE67_a z3Qm{4IsHpK9Y{{qvUutW0wNv~gd{CYN~cy9EwC=FxV(i&%*_7BU_p3x2g>XQIZh_{ z#67rfy**f$_Wmm>7{?mbUfy21M*E0p`UJw9Jw2>|HO`#agrdpn$hj!iFa|1N9#?nb z*AH91C>C8|EQ$TemOjIulFBGpt>iZ(o9~e()q(TXU=h!i0_P4Pf-W<)$*3ojTeCN+ zrdv;NSR>Gg+*{8%WwM%lN7(~^uojw9d5!+8xl7TD5%gx@Bb6U&vRYK9s#E@lLxrC$ z%KwzY>=A{Gvf4kzTc+(YI);`E13}UOW2q^`mm@Fqv?|pwH(HFkjU`_^%6o)gkEGl{ z&I5!@yG(*Kc&f)xPd7}8=;=KfqcfRa-K3_})+DVc&(UiR^s*iBY>>Etb(Uo4FbHWwgUi>X z58%a-go@+olTnu}uS4W^(dA1}d_aVBwskCqyg@WIwOwSVCE^An%1qSNbVK6)ZM82Q z!rml9d}z;T6SuRGRIs!eBvlA36V&{}tu2;V5j%H+T|nJk$`wKnYUk8o5N|kCyfU z`+5xHRJZ|lkRNrErb6ZR7|HJKfZ@XVx#lEW%ADOmqFPGSW|K{{L~o-IsiD7`qNEPj zf5T2z+<#Nyi7!ytdI1Ny>Ri>C2#3|KYDgFwnLmL*z^q*nRBvdaTyjZ*hk4xDOWOx! z#*33L)`)SSUf2+qhGF3P2igzFBlK2(;4&n%w?D$ZFaVqD6!GhzVS(6}kP*ZHo`_Lo zh+|Sp8C$t^cnkhiU9&c>KM;#RT<%351|q31ClC}_igLw-W8r%k@Z{-RfkpMQI7M&; zqU+@GAq`|^ccif@p~bn=d=RrG_5xH3kLBc(b)Mx>Fb%Ie;5)XEf2fh@Wi(@xFA#`o z0`znp5Mgk(Sz-#3ZVzO|=Vsv>rAyFr;nxJRO=v{Y^$IJgi@U3lu*0+gy=YTPZjAbM z+*UXY(yW9@m+RNo$K&6K$(Z}i6K@MS-tf!Im&{a9$IKY%%P-op`~i0T*Cudu)N6F* zl?UU>m*^h6k(&^ zZ8_&yTq4P53;rDZSbap6^5v<6V8?{_+8}{CQc_HMfc0H`AAwQz=sAl*Pl1 z^5u0l*5i8|mcF}qzH^l1cf}BF*`ztUn`y2>UNIwzRm7$v7bysw;4!*rj?m3+40~9(JS73@*S;kOYkZ9LYU6MwUt3TJ1BO11 zu5G7-laru;nOD4Js9XH~0axL2t7DlyIjxwmRbL=7F#87|Cq_Dt7+TcSpDfR3+W0z+ z7g3!k_X*Bj*7EHJC_MxSw1@d?Uk|^?plH1+_J3hB0;SQ(|11la&y=5Vm_6clcYzKZ zGS&u-!8NWY>>NONSVvCC*4-uU?5Wp4^3aDUSR*L+PEH(I&c964gZtE3<) zSLO6bCw7R&P9ROV_jPUk+Fq!JH&@`Wp2>6B))|}(xM;7F!CV5z1_xtO2mj-7t{O=Ir7Dyu9lMc)KjjmljEOm2SG_; z`0|GnQFBJBK2HDgZ-IhDWS{WxITt3{po~!4|IED(2v2IOM8Uc5wHD>*C<^58iy@B; z&tN%~cquODpor=PF6Pbku)MLdU68hqx%2w{mitL@!$0Wn3oHMyFA&}8*Hi!$G z>4D@9{ZzF(VJO7HTr&(`?=El(J}#Avb%_v}HG9-x;4bwH@eshvWbAue28dI)M^Z6LvaOqxF(>r``dy}`pm&&VDx`dD_R`f>$ zvCdizBCj^{@i@KSU{FuePRKN<)ZF&Z{%JSmSy~?^x77>@K(IZ0P)>-_zSeCN`)5MVbe&JMl+IourfA+@Y`q z!T&^}Y#}M}orKL6 z+6}IyRYn=Sq$nWYRj(NGmCIsGWs)<=$&8D&a^-~MQ6Q^fjn3>8EVKb3T2t`074KD76R9ToKT&!Gg1MFUcIa$L9FgCk) z2HNT4(hjcCuNNLtZPwURL(S=*L21-g#u&yG$u9W++Fg5U!Rd>W9p+=hI2)$Uvg=Rp z3&X{M`It!Kkq`+)`DMTKt4WQEPb+?*|eK0<9_F`O3!@Q|fo9oloxsJ_@8yqT)&t)O+_( zf#oo3PZA?&#dzpwOk@MTU&ENKe9$572bErq4h}Yjo1wYT0HIB=#)2f3^w+>D$S9i%$ z8!~u-8Ox_|bbVT+pfKU59~BxA%4kgJ6~rPfO0-%ITzGJGjOL;QR4?F^{1DD{HE;dQ zro|L0U!BNIg(#3{B@7sb3eai+J~??<`0VoAp+-wdh~8}k$M`J=6hNdcMe76zFh|0T zvb$+=X^wPY_D{i)NpGaaeE<(063McIOWyE|pQg;q?RCL&{AlZ_3$UQ!t zbM*_C-B7f5PR-i-#m(bsm)`Hd^-B$Ob@yZ(5BuH`mWi`z$zPKM1O)SfRC+o(L2>R( z-iAFL=;|&eU$oWJ%!_gG;h2#{b(WB`EG#FB3ksI`Y@CL?MI7&c?Az;c2v+Z|_s4fM zjkQpW`$y#0Ue3?Yn;hdwJ&qhT_8io0=q7F-9_*%|2)joT^Rbwbdk+v~!t$d2CO%Ri zK03$kkAphk{bL(^}8sy&B(tSwdmHCkWSdpDgPO;1V7%IjWU}aJtYf zDr|eNLg2wCO#!k;`{Vel{h|H<>n#mOBg!4UA{kR0=*HosnN=)%yp00ApMgv*EfudN zecf8)LEW5qarWdF4@xa204?m3v>nlrtRl^2w17kXO5dkNI$jg9{y0h+X(BU=xtfE+ zP~kZ!;LMpJi+==~LP%nKuIjOg$IwuTcT_l5jXhcaJ4G*eI|iC4__x=;MtOkNZR65( zXCii-mi{K$Mv;<^v^n?5sCX_f>8GKG=~skyaR_qDm!+?5!md#!5{*BEDU2kF6^Wpx zv8!@>Pf}M^J;%q>)8k<4-!+d{L@d^?WD1;l3j~ehgcsTnZPSEq_{TM1zK?>I!+GyS zG`ZTP)KpX7HA{Mo0i9_q@#37}pDO@^W@3boI~UDQ9d@y%sC+QXsmf!+?CvcV>uUz*CoTv@F}8 zLTN^)eq_E~77K@6P)ooS%5AcZRTbXfZ=xnfMn2*|KTg+vI1PYl5$o2jS(6nwwX(~1 zovpQBy91Up{lDbpr|8G512o>*KDYY=h`1-7N%zR*psB_K`hb1d`Ze9wU!C~}?7FNx z`eMp7s+3@B6B`{x0;pU#jh|T)l;2GIn2rLH=ZM_L!yb5!*if!L-6#3B(A6(n!aAV; zeRWs7Qv!Tbt3sm&v8HBwvCz)gC@U{(KwN4~#MocoXFRFMx4cCI(cbzQGTx>QN{Y8R znTdrs{S#~s8bEEgQ3(ak=sA?&HEQ_i!I;H9q@2Qynr<1|5Bz0If5$Ou(R!%r2jN5Y z2Jn4a13IOG3mobNr$fPS6gqvu-n5~EqJ(nLyuGdA%~XO;GPR1vQnNs-?#fTg{@Pz% zO55+{8wP=$`oG18?StYs7cMC7SyX*ekYPaIT=q{0`K0!fCe+>&NX+!TkzALka?J$k zWE<3SJH@H949aeZ6w1yC*&zabk6d8snZA?!wU|Lqf$J{uwc@Cq6PDD#l20#@+*f2- zyGp4rLuY!C9@L(5IrCm|woS==amg1DrerSNT)F@86$EEy8r^P7UB8@AmWpW(zMsZht)yLD64Muu4d_5Un~=nxJdcEKR+?TM+@VBhXP*~ z4vAJIpRu52f`ufj$)1ib`j>EJjlCsCn(E0==tGg^GwED~=0iV*oJF4XX+B2jLIJ}N zRlK{(aK6k5?vXOYyB$$wpQV3hr<$c1(pD$3*etOV42@e4#yn0FT_vez!73uSy@;)U zLo(tfW@Qzbj^fypRfls~-9d@2No|B3f=@zb!fwF+2TbJU8?dNi0vt6>Sx=r9@-}Xl zkf?XBA9|l>Ms^D=z%T4-TyjN-coxg9E4wt9{S$irP`o96zRT$9TXbRTTtp_2v=H4x zx$@|%HTvh4UbLGvNkAuADq;Y9>;uizl3)PlrM1ECciNBEs50Auea>DiqVB1PB7i}Y zaLqj`$e;Ban&7s~F6@|#h?}%vhBP47{rd($uc4^;=q+hsSSl!7L4viX{x>y*0 zcCMDp9_+(1@{Or+Y1@O_4PRT}{b>dXq+=m%r`Y09(`84MYf)tOVck)a6i;f8MjmNH} za5*s_LO^8osmJ|T;&>^~qIX$$+nAcFX89=2nB#CXN^ed0`AQ*|(^<3rN}WW7p|VEA zU=P$%moKQ!RHn*NwDhf@yC(i`6WDH6m2y%MW%~(qCp{$%V{I=)UdlZGA69 zcTRl!84o{Qx<5awFAA==|Bz_QWsB_pxKN#YRY3B@sQA*}#x*%}3N2f+EWKwvAC)482K4}Rb<$E_P629oBZ_!Ek?sxEvZX{d3pb|*7BhKB z?QaN}ST8J457`8^elI^y+YVZfga>cR;xk-RLU9;1e#}v_1_n%lM?(#(@^o!0iIF_x zJf&KK6Px?Glp2&$lR{5dLf}6KdJtV%L4%Q(twE>?o^?~8$H$t%UdL&S={9)^`5iEt z_K^8HkKW`(mRG1Ff4qIDcG<_l7}iCVRskszMGSv;yGOHJiC3T00W-F`kj;Hep3tDb z8{|o|+{Y{U=`OJJg5m_dUMPdBil%XkYC%}kXegTG29jU?UgW_ho6g~hO9_SMFGWOC z%@xy*bm6Cj9wSnuLbSO-qFoj7IJ_5aDw&L0EQtS|Gy#V!xKT?iL+*|bx)TC7QCorS zhDcAwlHkmopWeyAfAR7R$987}s7ng;>r_YfjNd>8{%iwC2|_|L3V<jb?paX9kSA`jXXJ zW_5Al$oWl$YpT~$NvTFWk}zl5UCU3eEf@MYc-&<2b6=Hb%MkCkxL^c6{PEy3`&#>i ze(fQcP#KAbzbxNvA>XVPtZ1d4d&J9fh9#XHqgB*Gx$~dUdFfPrK7q9S!ET=f?dB`# zCe*aeKnczL*R2lM&oAfd{Xx)X5XkqjefM^$Vgu{33i(N&9;|JCFX$l=A3c4v*@$Z9 z@NUDEd3je0FQg|wI{E`9?EL|Zc(#(yxKQ23540^FUz8n)B3Y$^6lJSN3Xg9A_*`P( zJ8l9{h_XK4sH0KV>eO;ltwZ~ja=fpOneIbrZ&jP`@qP964xH##iV|)-lGd?X0&|3-q>i1p_WnE znt{1n17yq_4smGZ1S*b_rR3kIB)pi?x+2~Uh|b4LwW~E zUCkGXiMU+D*}8Mmf53;!B+=U?7*i2F-u!9zxGvCJaoz$Ojx!WZi#L|Cs^eF`Jsh70 zKAY^j_XMG}9OUu2J(z#^`29p}OhWVD_l3S?^M@LBW@+CPy6o54Xt@l8){4^m(g5;Fux84Q^)HO*#OE6E2uK)q zP*Oq%VMVLaTHe7jLitPzi7+ZeCfJ1s`c*z8@4Tg$CiykzvZX2|>pFAop|TyjtQVAU ze|KfbAxh}_#E%$(MAtL#-0*^d4gkgx-Wf z_}O+(wb;K9i3h&hOn#vh8D2A#!80Hvdzg{Y_f4-a(iu~IAug&>r+WOss- zOHoIw0q;2Kn)ia5`>P~(M5;Aihjh>-^DU+k-xOYjJ6HS+F-j{T#BPwYa);DpurlBq zX-gq4wK_00N3a;8O=XDPg+5b`q(tk5uTG^A!ON>*u(a)aBB16INCZ;MAO32myUuhU z%RyLY_4&Po!_&mjr{f)gx81CM5~B(Q3HPpc$#EmO{W<3r`L3mVS79(vTe#g}3mT81 zrugoM97kEYM-yD{9aZK(R<-zxgyG$nxqjq=Iv=TTkTdqGl~MF?5As847qTg8gU;E8 z)|~6hY}ZTx;W=IasgNTY;U7e*hr187HAV>tawA0T(|C~}9y=BcuSHs3qrg|yNN6M( zKUkFz^*^T;b1fcmwUgGW^%tDYoG=bjbn2{vIh0LhMRm#GYXRY@-r;;5g;JGB-a2vT71Ftc7`d(EB$PcE;myx630O`kNzD={ zVfz?&zLHEUa(~p2`Yr-dTZZ*0)c<%R$Q2C6d7Lrj#t7ns6T{Pex1vt5{qPZ?U?H}n zh~8!=Y5qK6B$_7?HM1oIsnnEhQy=4BZar^sPDI2F-lAns=vzmmJINM!Sk9rwwzs)A z9Ct(sm_ME^YVa78ckZ+%lXF~{*I>xroJTXcvUfXONzZh~g}`asX^k9bgf(0>I|8eB4seYU8c(Qv@8{l$EL^uf z!1?vFZ-!Z#z);akO~e(1q-_Fy_W@k+9KM~8HO42Lce%M0Z0TVWoGeI0cMYKS3rtWx zb7dA1Yn+%sgL#8+S!Y%Itt8Npq#Ud4`Jh3M9$#!znoC`q5AoaYhPWSVsLpaweG zMs2*kJC?&C?>q(dlOdre?$vlAZgEKd=hdOdUzzB{(el4iPa8iTx<(bmN$6U^IOQCZ z%*$U&2fSdmeQ#)T7PVcU0+#cpWrh!Kj9@RH?jS zF~J~(60mVB!SEc=O-Ra_E5W0=M>F%fv>vn@BiZIM9EW9*i8-g();xZ$I@KOp5C&J9 zP2>wCGIWSjjvdO@L$_YdVk%V5yOd6aSJxUL=gE1wKi!y?(g&H;%W=oYj;7a@x1X;% zbJs3(rzem^p`QScc4kNOmJxHamg6CdHdytVoRj)SnJ z^E6|!%~VED+Tz!r>wT$)C3XIJVG2DJm$AXGWdU*L@NdlQY$ZET3Xc`zkU?2!xwIlq)_lD?t$&`x2em-Tqwt z?@tWq%s-Mbg-7RjKl5)5bnlogT`%#d8!UuqBDs8Aj-4X{_eG&IUZ75|FnP}hXSQSz zy^MCf|9B*6oa)oQ4GrH9dgTCdV_a=8doJ0%BPLTGrN4qyQwo%dKq#U9!(%L#P)2Lj zmp1|B=8NhZ&eRcKff6PklFn;2NaxOTi?-ep~y&>}!zb`!BHrdUdp!i6dU}I?VvF zY7+NM6K4M_E`fps%V;{QL-HUWJBRd?ZxU<72JaK-)yf?3U#*tV^nP(k1NC+xejG)% zXI9^k!gD_E6RVqpl21-1%pWp6O2r(>f3UdHn{!&8Eit2s&4&UnA8u?pk=vLjYS*Ms<=OtC z|4D9awh1ru9{N9h==;p(`Lwyu$Lmw+zimQx5Y^uyX$QM+f*U2&AkYPZV_O2#$c8R} zZ45A+a0S{!lMspiur!?ZkvBLS+1*paP!G|PVM3WXT70N85U3swyU@N#yk?=xwo&@&)iHdVhN*HNfZt+;oC>)TLk-IBjS?2OgjcKqSxFS@}K z5NnUl4x1>-TUo9@4Q7kM2^`%?5UjW6Cjju!ObwCjjql+dM~fv*`JosLy!h9-Af&;>wsq$v5{$b7&@bPem@Ft(ZR<% z2Q*4M+^`75H-9h!sjgnrEOoejxU#dFp$$opVnSF?1;jv>g~n zQONhaBHH%XypgmH#SCP_KN1&exVcnmRwL*vq^?3=n&l}uywpK>V~*L(_)ppbWPLxW zsBtKUY&Krfxi3E=Z=@pY{aa8#Z98}D46jp*s=WJle@d>n-~0sTyc6xEd!e)(Y9eXbs$Hs86D?WF6MkB%AHtL-d0s1oKJrBO z2GQIkfuNPtoJNcGNjPKoYH>QJ{iTf*}T2ie{+36 zEpJ}MwhL?~uTS%~lc)*nug$_45$Urf?EQ4F8XLvm;-!51sF;dkjH2(3p!ok^7XUp< zQ%WX$=f(HIG5NA;?G|n85JCor@7^}y6C=z0iI3DAcFiMKQASzYeL+*0APG&!zQIb& z^(I7kx>X}E^&*Qapu|ahQW96W^W&9L{v%2$Ik7YD+_NBZ0fYsJvnqZ=Ur#aY{!Bkv zkKM|nYhnu4W%C{KPKp1}Hf>o=HeohD1e48TPV&MozSP*ZXrEW$mx?TPdF-jE&-??i+Zpq-E}qjt6~U6n zvS}R~MEJ`3?o!G#`+s26C$&L)S^7_D={Xj0r2J255&w_lxR`y`9b4Ot^49XJ!lBYQ z{V3A*yf=o9F}Qk?n*mA!W+23D-u;Dl3(?icn?5gUZ|Rbw%EC--I*1=;J+`ozEG4i{ zk0o}u(rV}QqJA{y-X`W65UW+(S8HF3OjPV{SKmjc(UmIx2@)u}vXTs7lKT#hE>v1l z3xn?t|3n~vw1oO0+!xuIFCQ5|N;wfD#dc*Wpt#}Rl5d%@nA7!|A2^pqhAnOvN#EKS z^Of}o^M@h8wMvA(p~-v6>PMhC*GWs(aYZ>AxA4>%`nlTXacis-s$oI98V_! z6<1QEwE-Dl*q}5m#=AKtoEslpq(0?bYhn1CLF;k3_LsDu=j5U68B|*lbb$ijR1Qk- z5*}1nB!r%Q0*2N`4gI35I=q-&NLs7_(2)3Rz8mm5}V!x|0JRS zRDa}15a-{FTttB@TZ+njFb0z$_7d^_`hh8F{1DAZ@WgtiSG%Vj(ARRt@$EJ#DC2g) zACQu$s&T&Rh1PM|VU-e=vM|JaD#KIv& z%c&RS78wVLUrKoF87cr^oRDPfX1H6*dVJbREKUE9;u!oN#W^vzG4S~FAH_-VaC6da z|3`5;?0Z3J1~A0TqCylA5TdJ%gYz{vU06BFLkdV4G_)Gtojop#p0YXMdBKw*eNQ8} z1Y!`GRhS78b=2{Dul-6Gu*ch24!Edi0bJ&i#aFSYAg}3)lOF|k=Jy^W9S``i}uqrX`9V#UX$Z5CCuzlCD&=4 zrLoYX37j5^W}eU1fQ$8TQNu_ebBkjFsyg->euisH+06|gNl1B00|8?Ij=2o;D+|p7 zW9%=(w*D5CPywf|TFZC8Jz0Jht^OU0030t-=q;guFkmB6(6Z@-{e(REu;F4f3~%lL z^<|t5W^=WlG|Rh`|1DO*486^^b>)Y~E}!AJTohe6lrdUBWTA5o19WB#pLw(ZvFPgRlnpp7fBk}@4CC;Uedx)A?ACr$pW=WmwrpcJm@=1J z;u-;>C>@T5&|?WuW14@260Wht$^;=3l^)L)a*wr|Dg1o@O?F|Mdw-zsa6gZ)DHX-6 z6IbJSddqE>A9%u%DRwcCJt9>68_Mx@58hec(=+p6JOw#M+{Su>bTRHW%6xU={Gua7 zTl?lPx2(!8LGR~`^VQDVXI;?5jl72j@4snRC_#tR|50VzYnSXMzr;+TQ6B2@U<-JC zc)hO~7G4+VY;kXDk_k$og4|lVxvzgnG#rKp_e7+pAP(9tIM9p!GRanQN;4FbfS+kW;Lhw{JLjId_;`}?SR=Y&i)>+X|0m9pB_znX&_ zeee2rvUI-_?I1H@6sf)bGVGw_R9eRHPW7rNU%s zu*T(Asn%fB;gs$CogCKSnt?t5e~l-AS)rdRs#;8ev?v6PHd&Ki=B@X?vz& z5Sd{aG-=luyn2TgLoK1xxng0GIE5q~GEM zfRd^^D!!uvlm*nL4_;jcGj43hyouMkJto*`uo z0?*fYE628lX7{UG$@-2tBOG~3MBVOU*iQ0O;kpM}+#z}P_oq>jFbdP=n0d~#ekxB) z5Qut}EY~qMET0W7Ml7PM8&369_F9CJmus)CG$3+yi=e4{>V!Rob*IHq%grQD=VfrA z*gMHx(t`;cc$%nPEnn|aH@H3Mn|!1a39^P-g4KH0nN1e1luA0@ad~>KvNvRN?hNt7 zvtxg0A6Yh{h@-W)+GyAMoVBW#T5iTs$>wKhs_Qh_*u!Te&T-itp(TB)HfH(D+6V%6 zAM;6L@&(%rH9x&hLxuz$N-lv;Z2GPVo+`pCPuPwf#7n))-=9-0*TKmR(EpMh&Yxyr zc7fT{%>-PTY_K#QSR^{rapbLq=Z*d!4YYERoqkf3nR(0vuTvbX6o&#HYMe|0yb$4M$55Gsuo1zYK>fxXtSaV7;Bc^R@w!pGS_C(M5i z;*-I1BRlkNNfBlc=B^Aj)eT2gLZ^}!qmYq(z9o3TpWhE_3`@jmk*`>7ns@pdM(I)! z72>-9KkUC{g+i@ zn9SR;8e6I{Y5vq1yVUX)r;S9>qEs0@_w?MAuHoOuk}p#CR(RJf^^=8QTeDqDl^EOM zmc)XWrZuF~h_t`ZUMWR7T}l#Dq+dAcmvUR46*Jb|e>#DWSO&jL>OJ!O@K=%(|7Cf2gzZJyzlW zzM_OKAq=hKkZX~B>be)$C*h8qwUb6Y!{R}%%7OE3_zUPZ^VJQxT+W>niB=AB+KvlW zhSPg2^p>!B&du$UDBa#58R~i>4fuAq_AOE#P<-TP;EU^DJ;L~FFkI2@cN*S3r;9>AK0iUE$B$T4(28#wwW6M~6vU&|CO2KrQ@PDlqCdeOTkabK> zZz5K%Ns(DW?@Wo7YU0P2vT%ihB!L&RqB@Q%41W~Y0X3Hrc&=8%^Gd~CrGj+fe-@^pg{4tD~_3a+Tt{ ziv<)kSN0n0gW&2OYUKEkcu2*#oS;aJ#k~TC)z2WVeK>cku7L)-WuPQ+R)EHlvUxo2 z)!w+vZ{YKT%q|LH15fMmM>^INT4PJ+Hz;&pE>bCduGLBLFC&m~*wU~zxV7u0uM8!#HuM0XVVZ7Yr{4R3oA&=msh9nKNhw)QFRw$ze@Q9M zlhxXtt$#?VaWqQruiN-a5qK;L1SAqhww0$j6_hS#9@ z5?0gJnhoA)!v_pnI8NYpJ@^?cHY-lJwVE3?r*TJ;QiZfYE^`!Ibu-s`+Fok+SbQw4 zOBauj&vN#*K2RefY6`0TW4L9!bvi#Cx4j$~EhG|5kQ%C&`QeoKYV8sCYZ&fLL}-pT zX3wE10KA{0WU&5o#l&op0?_=T8LP06@RJ2TZ?5Bor6}7^$=8!jR;~?;X2jl1L#1O3 z!a*eXAg!RS#lr1dj!|M<=8XpCky2kW!#7?4_lNm{#hLEnUQv3uGk**r5cxvyb~vea zR<>56Dt2k<4zYYlyLWR@s2C5*lJ0jzk8sB#~8`FXTI+g0Dm-++ilGIvEC`W=D zKR%;eeOvRi$Il3kMRE-WmL=G%w_8+Mt7IZ@h7raq)I<(Ijjp8F@&2~h6Q1D>y;1k1E zX~FL;uy8h|POWQv1n4G%uhk7XG$az^(vYh#&Vaq5!qaT@jbZAXEN7j6Wy=?l)I^cF z)BeR(ZlJKinC#tiNYyId?0_Z*JM|`LCP^(ouNO(#r7X`s3Y5BrM zi6%MDw;l_oJbcT73|dBty^YQ;4QO1++b^{eeEuVd20CL|&{h0_I}}7{ytvnyw;#n& zmjj?y@%H&NUWzyU$yP&)6?-N8mJmioADt9mgoe(+wv=uG^N;@qc!@;C?6ID*=O0 z5gBVfx4Kmv!m~2o!GVgJT`XDw<#@rH0g_>56Q1Zo)E7;@9BD2EGBHz}RNjNU2lc`s z3+wFseq9gL@IUmcx8hQ`_ml{}?naf734KdMJUL|)YHcI=kALLEy>5S7==U(cth^2r zY>8g^z&?;B$@mxig)x!@J%>z}qED%`c6J=LDs&IAT<9Vu=Kg}zlB^;riKr}+P%7LV zRzroeT+^?KqPF0kpaC^|ZrzR=4P?xS4eQGQbS)+P`NmA)^5lUt^=@tup*OoJFMiK( z{COA90hYh*j<0QAT<$Miv2)$|`Bpv2@bzO+H(ccudLO3WDk;OAgU)oeUBt5 zbW=YX7S)hzLm2hhw8_L+AOV9+l;&EdIUv;q8_7`DZkfOAAiz((PY;dtTor-z;kXD> z%l-$%Rp2(jKNH+mM9k&vdFmb3;GC03;JlCA=XW7&erNAgsKbSINWm{F-RJ|H#4m6? z?VuWNd|0IeV@;O&Y+i*9e?rO&rtEQpXD<>;4&aeZ``RnYhw-(E z=qXy(hC`2xLU4`ShYvLIw7od!!SN(`+C=u8bAN{)ncc#lB^k z_>9QbW$SX6sJId7-rOiJUG!iI21UF15)peQA#WX$ZzEf6Ko3)u}^yjt06Lu3c;U@1C;vfA+ zWu$2AD`(kGho9E{1`bTDg)_l=z_tB2EIVP`n8foQjmB&Vdi`VAd>)@3@q)txx*H7_drA!GI=)pjer(69<`ADs;&<3H$ z58G(!L_%Sz0Gk>!|M=c=qD=Vv)P@OOq%nJg>0W94AGnK z513&B{$JrM%_sDdl1>))0dx?y-^%rk0N-)Z*tDi8b0o$fxbhFR6ZEgf92uw3RXGG$ zW09$+AZe*J>)YJnDsM#^SzYyYnX2-}hy)$5?hXLl!hN8t(YWpllCX5`Jb+6~9bNq^ z;eWXsJgZBx!UVb6=$2f1nBssVZ}<=gWK(A9wSWJ(cj}+9&|QDZV07d2beko2Poc?W zK8~+LOFZI`o&0;|AKpFe*STd?g$Bh33QpGGd)(fRKfJ zOdnUDSz-DDGo2-AQO~UbO?lPKbBcScF9YA1n#%{fRg*VREn|k2LSPz5DWpu>Mw{TU(WHgm>kmH3$%>zWat1OW8D&6U$ez$>m59zO< z*a(QK*)|=FPKjSk0X=NM^FANu06jeUo8s1*^~{zz)2eFLA0q;HzP=xV8Cv`}l)3M$ z>0_&zVqL666yGfrZ*#^yOB%(3#kcXhhmNba5sfv>elurX5{Oa8~KJYb`N3OL1dee@qtjS%w11|~QG z@Dt}ny=H~qVJ2ZJq~@e=n{2-Yyut) zl;U>sZT*&BE-8uV7lcgrW7nsNq&$n9Hs1xPkXxH0ylKcQTpVUq+q)i^9rX#9N(8*h zb806GT1hQ_ulzHGO)oz`A3;ZbPJS*E9y8(PeuAAJ5MGxDw1bI(cWZL*EPna*=m<|va?52B6l9V@t}SKT+qBCLcjW`rIA-v%@L(q zU5AS455Svz%PZMLW;UHd>tKn3{(v;OM^-AZntYMBw7C|;q!*uDmBB8blnGKondO34 z!f>s-Z+V*FSf<)jdd_6;AvPK-A*j(0e&h)tOoaEFM4fK z=e_nrzVt6h`9eE>TF)m28t~K`4r#L2D~y=TA;Asw^IZ9uR^`htv-+W4aQt|zFu>k4 zs0R17RYH!iXD?Sk;uWknUuyPHLWJUy-TXR$q4v*ZM(#_Y>X`!o-I6Q7p`-@`wC)U3 zOYxf54439w%e<@)7SqYb48_prGN30<`dMH8Mko8|jO3E>)pp@cm+CG_XqLs<$x|6j z-(>@{RW%Ny46rIF(!_ond3?A0u_5RC5&xlq58@s979p2)_s%D#OfSb*cvzzD=24T( zFS0Vy>5}rq4g)?meCLztqv}!mCm%pz%r);k$5i@2juuuCL2dW@{mjMMYIkOXat0nU zZ5}TdzN_Ma&1am$N$EB8vzmV?QMXe5tp6xc;{VU!zG;m`f>vS}uk50M2y6W1;S>Rq zi=q%Zao%|3HTasN{8>#)N+J1fak`;OkXCN4+KXfL)Ld}VF|SRJjZZ2Y`Dh`5ib$)> z1MW*K2FSi=xTbsL+HY=_>xd%Xz zK*yUv**f7t6aj5Jg~zU58!z4Ed*4_j$_5#M&|HW#%9x+1^&`K54-_lsa>+%(zvfa8 zPoFygxb}>(*nnhl1;;reIWcG@0s|eq$#IaKz(*`f+l2G0G)81pZ2G?CG1iN#vt0R? zQUo*WK6Xr}ErX=BTcL=kw-wxY+TD9p>PWQv^A(W8LeW(7iNzm6!AU~ZN3fzzSE|~a zqr=1N>lSQ|)@Z6S84TLI6cB{#*%1U;xmo4xyTiYVYV^|K zgG!_gM+Sn_+#vmEgY=Xa6iAZu+iM_h4Aktg^`1T+RUv*ty7~Q?XTNz2P7w%9s_gI< zF(_8Wjvn=dmer5pq7btVE5^AtrQ5ddxnjhALDGpW=SEac{%8SnEH`?`+OPfNZh(0` z{jUF==7N-V;Twj3r#a>JbkW(#6E@9v#;e+%sa>{h!?!>|yM7FWvx(V9o0~LUmTlx6)GRqVF}O`Jn_#C z_NQb1CBZseS&}TYVlU88u;$;QxV|5pC^E(F{4zMhUlEq5|D?DxNwp}~`Zmd;Te`8_ zY!a0SPyy>b&O_kT(S3Nwve1amO8;7{itkU-hj+ zW1dYLwBJR<FmciFMI!;kzo&3{_-wCRM`O|>S zK;|&E(BA-5$)ydI7Oz9!9~qo5^KRn-^I4ty=s&8d%I2wz!T+73NK*bo+n|rZz=1SL zYtMaI3_C-Rw;Ylv5rD#ss(uew&2msqx?f4#@;6e*1T!ve)n0+T9iP%`S*6&dd0L!D zC^G_Xz|I0`_(f}1hkY-Zr>>82Ug$iYHg_u;E20omN^|#93M<%}27KyBq)k)!@}(m@ zeww8G5VWqvC;vhiP=c?<0s0jg-9#|zmN05&KB*7zH478GvsH)A!0$||@yCm;!8g!h zL*x^2HGdq0C5f6>j)~sCYT@_~N;ENWo6RU^7HzFR5K2=I+>B8_$xOjh}&N%v9BHSIG%kx{a~PzWy( zm%~`{xHZ>vv>1|oaz z)_ixL3g3Jx-~iylf}f+__b|XtyGe%0%H@BHqOCnPjEMwW!hAc#$}9Bp>&uDu|oydgR!N5rBx6$-H zQa|Mpd?hy<6?N53frMi5BJeCg9;J>Wk-WAHYA|H==1Z}kBiwO%KH!f9W{}vb(-Ai^ z@lYV|6ouCAyS1WUC4LOAYZgf@h6J#J>>OqGJ^aVWOSQ zQ7t(i1-ULO-9BQlL|y@P^%=4Cd!{s zZ!{2l0-`&?lw0Qe5K!99}$(IcfEL0e>rj2;k?$y?D`VZ(XKNqF)TVl7he z#_tkcR##f{!*68F#LNb{qR!Vwf5K%tw4oV8%VEh1;R z&ojmjjUqnM$~{fH7|c$F6%p4qPE*0K>IBn224q!>m!?2iNTt%U~^I% zJp0yDSNXfxrY<(Nb>nH*fZ<(|*C(eHUFD(}U>B?*uPXQ+g##;U!99U+rR968N@nrl z(AeM9q3h8c_cRnOt!<8-??(+Mf*LM8=HqE>y}{^jfRPOzLjFNvZ)2y*^zD)9Y2R3$ zPxPk`WCzZa@9{aN(4dg0o7UeRlo*d?M#|8(;1jS)F=R)tP`Eg>{~vfH^e=d{-n_7^ zM)g1N$o47wBT3#r;8EzE;~x+j2!{!M!F7Hev=5B;#{^ysiN;<=dw&=X77wQoz06;t ze6&C|UP7=}ET;$vxt?$W=t7-{{gb6b32bwy*R8et}IMN?ym8>A^2c@(42OFYnJ&Sl;GuT*iR#GLw0S z9(kJhu{O0Z&@l4#=3<;hxMezd2#MA4~}QR~&F&j(Cr!0zUP)0PIyv>KP>^7}_s z=H@fxzAiiD>Th{?F%hU!S8w(}*NDbiuGjSD-K=~uRJ2hJCTeKhzlsDu&sKX>#9tp` zWG#+=I4V-g0oTNFz}J{O0tJSumW*Myr|=`w9ZjL1sUcL*c#N*44)4N~%;(*SOfTan z>%%u79Z`Vh+CF2@8m7G16)i^*?u-cl%-D*-s=4vL#_H@P$XbG?_kAZ-jXj)mKmyG# zu&R^&c2uM6)JfP?l&heO!(T*HoYKh|-mc!cu=1uRHSPmRUR!42DtW16N(wwp)e2)= zbZJ%IYzEM(3rq;7yI$6l?>$=BVtqfc*??yJ>H8n|YdLITPRvGFn=fGj)9KV3gG`#Tst=27CKQK6u7ZROy7p^=7n4C45{tadzizJ1HM6e5V5tpJLKTu{8xCN-m&!qc9v(w;o?dm|7`!83MK ztp?ZJMkp1ByX?;jc5qV<=)2UTTOTY}JYCD*MfC-5J<`Qeuy_JP2&sg07;^5(NZq|= zJ?{8{u`~&hU4kmOl^Z?!K)30=IJQ6XgiY_s)ZJx))u_v9odtR~pwsw4DCWT7YBi(Z z2hZ&s%>yDTHUHu|*v8}|(gu~DDm1~hSH01;RV<;&Xqe(IuLz>lw4FJ<8^rHLyDTZi zyBvZ$3fc7Gx2!osk-I%!Et$K;;Xr1-{1ip`Va2?Av3|AS_K%+3Vv8~(uBEdtBXV3g zQxXc^1I-}xlj1*RC$yodEoRTR$7@xFUbk>^CnAcle5L6FO`fUm8|rvp%v)}kY*sAX z{A^Q;{{PRWvd4SpsnYFIZMRkOX-w2=M!HpxtsO9d5#ukKB3V&g@l34y?fE3+JvEnL z&tNKpuofiwNcJPYyzR%ZSO-zb1?X>U2$y)F>v+r@ ziLCWElz5N4<9OyF2b!fp-W4Yf0Bcak_W;bRIt|U|{X!LTRN_vGU6vcmUHg@GqEjs) zA)L!u?PM&dX??EAzNzwVr4MOgN#55S(&H5B={}6T=9i;-g(%obfD}iRT9%?_^5t18>I5t_4^VCr z(DtJcAuayYU_9 zBGRZ&T!g!zHdy|15FvRo2EhLV<)S(3Yj5Fxh2YU2Jx>*V58|e=V=5kqLCRQW-X)i3 zzTtZ??O|EH>QUQG^_EY6jaJ$SWi+9H(@K-|{D=S(on`CTNmJ!%CZ*pO_81j=LppRj z3c3%k8&6(tWZN>>s0`h@tQl zY+c+_{eitJ5Nk$RlMAaQd{Pz!1-g1T45UI7F%<{TOiJy+-PxQIT5yY}?v1W#2z@RH zV*Ojm1Xa)VzzP$1+xZgTH{=+J$n%2)!BHST>aL-(JCm?s!ON!Oh>o1KZwY(QQ0jn} z=w=|!bWqcFTZ^9MMQt&exc>F2QR~O4z$cy06g;-dzuJ`;%w~w9axi|*X61$13)K*i zijsc+-B7`4s|n4^Vb-Wk@aek{X*nyR(}GS`F^|_AI8Gnr?4f~512Ca5voWmbG_QKP z8Rf0C^t+Vy1KsK`NrI((hbWB1CY?v|{`j7{(W5sIE=eWr&7xH=pd=#&rVD+6M9mB% z4W__1CEnC-tTE)fRky|O#Vjvdk(^Wa!~{bgITpx%hI~&N(K44bQpI2H-gnjP{$z>^ zw{%uCqd2d{lGo=1P0rHsc+C|xM+w2zv(HabPDcREWGG3=h* zXt(N7kN#lEy)p>$$dsf8G}C26uinUTs~oaz25fhw^|_r;tD* zN&kG44_oJO*2J(5B4M}2%WT6kHH(ZS;j(rxM$e&5gi=dXYCD=iQskBT!cmVDBk6!3 zJ;B`bo~W2^g_aR@o=N;DF_zY2C~nzxC@1q|l;GfS4omZmeBsf+kmxC_dbV8KY)}r} z98VVVT{Ue?9--mWO>CV@P|t;o$cIj9^dCFefTsd}* z0IWpWAB-ohFbT9^x`F8BDQW7IjP%uiJBsYkHliS$fcemZkt*Bk{lm3{4}Qz?vP5Lb-J4N*wK(zD}G z9PMB&mKwF6eYxi>WfS}xUMR2~?7*n;!!oQy()2Jc9>ZeNmzr8NQ?&e%S5)2YFCBm} zRS7|Rt8XWQ?Q`W6s{=AU7K+xkr))FYUIT`Y}{KBrIADQU_zBiYRo^(OzUr zvVFoSwavg;qKcXZWoc3M@tiMl`6+Bt>utZ`lXM6U+1QXvbwiD=SZ1LIds-`cxcBnR zqT0E|i?zJ_AiuwG!DsjQS>79zNu&C?`3E`?;q&HRMKqkbwUB764|7+wDN@fLX4ju$ zhL_vpZgM~E1Cz*v4b5{Ne9Wj6W^d08FYWG=E}yd^=qB-SiAVTtC`r9Li^OJiYwDdv zdb;r}BqT0;rEhKN|5hb@ER&;qi}Ode|FlJj@SQ;BZkK;TYi9+uFgQ@F!u-O5 zJKi5?{NfVQuhbiYw2$iWy zoF07tsO+#Ci9pyNMR7t8e7Mbt?O+18TmFVRbD>GUd{D(mSE3`Hc<)ow;7my z*&iobyo3`jjr|dXWy4MpLBH{;?K<`89GDkhJYib>$ykkgVV0TcMw=0gIs0ahg_-e}6^9O1eE*-nF4l5FB?|_E=+0=J>g#9$u+U zN6=esv)B_5zF&g;B|IrceE7MS3^RkYZIky(9DPvucqlY$mK?2=lc-6kz-oz(Mbo4)LN@7X}elu|Hl*Ex+CvE z`M;HAVD>L}%{1|)yCY;6{>@s$rXpVE8BEXQk_0Eh=E6?O*lXWk4|s*yb=*mu{riUa z@sPd^iP)?*Xxnu@k|gnD(!4L63SU!OoFcHA0|Im*1|C77y++$_Xb@x4RYB%wE(W!F zX|?r$w0gBZv!7)(!$UM?=;uM#N#(0e1U(pb*ZrCprjl<+ZHkQVzwVHl5;2R%2kd7K zewdIp^RGjbfLCW{a=jk!$c@J-WCR~+apjMtU1?pk`eBxaAI)_7>Bo`eP-7q;W93wA zm8F!%Vu`6^Cu4v9Ysdwjk{g@NDO8;W^vzcQ#Rv4}qyV&8Rkx%%Rud!>&ayc9K6}&F zEqheJ8M!LX7Nm?!L#|FSU#^4vV^%>YVz?rSg9h#T)5ZMTy&ScT1rcfVU(W(_W=u;F zsAb!d#u}5yHs|W8rSCreZtVo2#rxBkF_$81YMK_I`0j8LzT*xeqQx_WA$qLTV52x+ zNJQSU!_z1_sN~?;EVrv>pRKkc15)$VjXXU)z0PKyGJSroMd3Z>v2xW4!b@Clm=&{F zfj^cqB4YGSw13doUucnO*ED-e?{2>jT@UeWS7-}FiM~ZqnDQq+(yu4+=oCi9VeM~s zH6t1i?Y2)Cz3)<<2ZgfbkgJ&1P9o?*-Ar8QpPC5pDDs5+!N1is)62HI^Trl2#uRF_zK| zYK&I4m%BJw6rKMR2!@>b%EjA8DZC5Mg}Y~ytI=Z{{Wv=fY=Ra5Qr(&Z)d^n1yO&kw zOHU@B2t!NJ8L7v0o3VS!w=|7)UJ4(;ak75qy;Y=u4oI+>WN)oowdis&w$|v$Ip1_I(N8t5ScNK&n%ngEeDa<%X=jR@`WydKe)f&3Un>}!(y!Al#CNTefK!6tIX`ny4 zMjR0>4FN*mqbst&x~G>OD(96({2fJUf*I2v^`ZC}=1)$)yiLGt?Ue`zO=Ig<<1l4} z1ZR~@SXjlG6TC1ue{kYSIS{|qIXKBQV?G9ELPXTV(Tt;SvR#bVxk2}RP7JumV|G3t9J)oT|Zf2%KnT=Y1ikc)U|!9C+~#M6*vu)_G51}nb&B@#{1U*r0MZje6AzHCu~r{@P}zL)txwa9bJIv z?}&mEL7_xj#;d9mFve%iK~xR#qF_OcutC3@FyQyU6yUHrgJmn;3d6kH0pGlQSHv}X zeKzEY0Ei7Iy$!Bj6y6!);CrM=O5Fa>@YdVAoA*4yKo^Ldht7Kcizp08GYNF)qTAR$ zR_z3BTxpxmZWi)?#sB~xk~))cdjKXr4GZ7g4aQsNqls9JXuxc_{1?CU9oipuFhH1F z*Yk!?GN?^`zjQrqdWN$Z4}SmD^bAm{*OvPziQLfPqxU@p$kU9{+~8NU3hn`*ayjSr z%9V)+q0mh8&+;C9C$;B`B%R<@a@#`Z8D19pVo2&J!)vY;tn|C4D3S<;uyc+dMJju- zZV@joV`TBSydGXf$EiWel9fU#G4ys+%Q`#|g6lOS53ZAC@`i~h@ao>F^4ee(vNraDeq53QQr$(otCz1|w ztk*5-E4Jhx;cNycYXi&k2~n5vs5QN}?+vvuSZAd(SB~y|)KN}YF|7 z_ApOI32&68^qG+-Tauj{5rq3@SZ5gr7(VfZdcGVObw5$eoS>h1IsB7R=Q%Rnxhc5n zw_S7A-?U8Cd))d&Mf>`9+DQEZhzjQU(J>0+PWAKuHdEy8wkT4j}a3TNj^fQdh8*wCsG(6~Och(iwO!s}5Qd|#-P zTFB;?+y&SIpKlLI$&ADm*L;fNS64URQhhBaL(= zr?_CoXgc`#R@r3S>~})IE}h^xNzS}U`UA?s8+OoOu;lFqt6v48`>R;TC~o=h_3}2A z%382sadkVAQb$aZvUDcQm2Z7kI~-Z9vrQpb%4I_?B-<-;;e~@u-XjO1sBvH56sTTv!+IKV<~}Vc=;ZO+XuX^19eqJL-wyrmS)O5aG0$rh3-L5`lNh_2M zXYhRojM||(i#m7aw6Gs%uisuS{GA&pCpyz7ru5xStL|^tEKzbjqgXoVh&{WIQo#Lm z4!5@@n8G(yv3K{{=5aMR6U5xMVQ^OQp8=y5_3<;-RF2QdNrl6|;RqeoHscYIi-^Le zqF2#zVxd36qkaZ;6aVDvYjBup&heF%w$tkNpBALss8Fn$nGoub2_YILr9UDiW3NK^ zzbt^3>L!Z1eMcNPMjdlksJ~`>cxRJ*cP&#Hnze;BUUILdKYJn`)Fkz7jT1{@Q!OHZ z=q7+kt^hW+zQs{Zc zF)Y&w(JybtKAVM6z1u{Wz902<41V4~lNB)2qmgE|4Krt@x8aN6CL4?Gg?Gn8lJj4f z^m4CgE_v-d+b$a}uY|q84BSUTtBt3lt(iN1Txacj(6PHL5*~=U3*P}nXW`!wsy1_X z-{&0N8=KddHVrg;=CC^?aOF$bW!(kKs3|a+A#v@X&j*?sAIBYqvZnf8R;UUzy#Is2?AW zNWEnjD^v9Xo|wha*yQMVi4z0&I?hYBvtMH5V-zoCI9Cg$)7Rys6U~>>S30U*x;txz zGejqT_uUcwqOpNDqIzPsz+Uu&u^5{;top6FgS%WC`OJ|r6Z+yJBF z`X~yQ{L|kYk@#-z9lyzmjZ|3jPn?(a-8CpnPfj^$V3(Dgxv zjb-a!@AJW?KFwgSW1Ux4c+kh{%&0GpUAZWjm2boM+H)43*B?c}H>2y2KA=hDt(63` z5rn9kU)ynYV)nVDD5W1LbM15O;&w3pf(>fFM`oQP^UNdJo{D^YGJg6F8fGw&^oV^U z(aLd$8w1TUu5FQdBr11qEomZ^FX?<6f*w&8m!}os?)ydb<>~dOFe7GJz|;ZpyzR-f z-&FSf_$%tsYO~ed%enC)>UpX0gSZyjK5fDJy~`96;$Sm6H5Vzmdhp7{1e%&%9`xaH&`v@v= z-ZUy0ifq_e=%IR&&Ul*kTUGKM#Fth)82{9428e>-RLjYi{mIox*#;MFDA3?bNH}#5~r8~eNb9G zC4Sk!?J#xdy!th8K)1iWw3XubFm6S-BHdTgHW9c={Bq`Z`|8=TFxmv(s(6a5)ugdidS(+heEd z`(5$2J}3DVi~F#qi|V>?Tb)uq?O(Y)s6Iy!%>=}_qeIj@3-*Mw1FWot%8)}dTDzb1 z8y+9sk?cC`gK*b^kt{h%i>FMebGt%5DY}wSPIh(LG@SQX*qt_$RJguGY|-2IftoBq z;Nd>#^BI(!6&9BTYS=C|Tr?ng-^w;pyrfB`Q!9}znkX3Htao7edh$9w zXY|~_9xj-KWZ~(IuJf>!1G&mc0VeK-OofZFR#;{i@~_v`Q~C6!r>mAb-nD!(9efI_ z)TvcIwA?8A6m`8rQ|Nf%1hRD8S&8)fvrFkGbHZECXVwh74M@{ff>bPZ{5;I&96GDc+UZA)O1oUUkwxchIN!W3G)A5Lk-Fgs$w#3!Rg%--qI^jrjQZ z+@N+`Pwky#x(@k-JSbxb-40>E5VJv z(7#m6E!6b%;DV-{l<%`|oKaIB=>%AJ!q_GB?RSk|FOJ-Pe4=G~eQ#ZIra$Tb zZG7{WG&mn$MI!)Co2l&9clt8JKroG*hTkd~%eo1FO1|%djDQyJ=j;d8CiESZwr56O z<-HwL#Lp8y2Ll;DCHlK1`VktaH}rmj6w41U$Sp(thENSDB)1B;*2JDI`#d(#99e*#MB7$ zDDKAOvhgeS5!CN>;;?ajKC*SFr~KQz?@cG9FX!3gL1Lcy6R&NeVKaC_;i3U~6;7L> znaTU1^NI5WdYL{tBusP4ENBe-QXkKt>8C!{TGzaOGg)`G2T$wMpFVfmgQz;}>StT| z_$~FM3((VoK%Dd8!hP$BoL84c7U0LLnu*7)oH?^Lo$Vt7``7Jnc!VKIZ?PjP*?JV3 z#_wPFkJrwg74@e-0qb{#c{?YhD=DvgjxBNB$q#OQbCbu z>_nSr{+9z$$P~Zt>TuQ*mpAM35>)zWfBm_tq7pw`JN`BT+nv0ZvDh!`$|dIMabHdl z=koGZ=M}&KpJV|r*rdMmsBHHATfD*sCHJdzwlgHwrh4AzzCBR&`|Nb7bdgSX7;SO( z{G@oBz`v$duli8)7`%`nHu=DB-lbG55AC_}S16Afv;>>gkYt;+N=%WrEwCX9nNtN$Y ze>Dsd?@`Et=h-cDbAL&O{@ByDQOK!oP`D^;K+drwKdc*zuRm?KIO4Y-%$L8oK~6Y4 zbk}dPj8>cnd%x6RpC{)vVO&nd?e=Npyt86H6|veUV=LC6wU*_|)!^79McNZT3iMAD z$~FmGN=K}}vWy{obSJ{T`deNxqobFo6EO`WFVFXVKP*KmxWtKi_RVJaEQ`Fs_o(G& zE_o+H<#d_%Ov}!6@(_jy>YtZWC%|D5f7_RC)jLrUBb_s5lU9{&t$(AuN@V_U^?Bn|q*srxq$q}}{u7VM_dl9GhD!Z3xSvv!SpE+bUuzlS}_zDp!skb;zC zZ+xXYNvZX%ZS>dRTX&97Dy}RN2{PJg(%H8! z^D5f4Yk93S3cL}lN^uES>lCqgZXU5#I{!S1Nj(uvA_`9$K)R;aY(#QD!rjrJ-?#=z za$m1LwbK*&u1|)pU0ZGh-K!F@Nn)J`i*ouG+@0m0jz(1;+05vo`|U*(%y(49+1tUt z>njT-#EyCS$om%Q&yUG#l%LZ8g1Jo%*%Ea6gPpb0w#uxUnJKdl=R=F27Hxm@|NT_dyO^C!aDN*u9>E@y`) z1=4k>sG0LVwf`r$!&dLgl!b$g8M#Ovv8a)Go>Xa=wIYH?kSD^9CyGx=kYe|A$(z&y zRH+`nGX8NqZ*q5Y($a19bZ`m0s!ehDAjUgQ&0-ZyGmhhTcgOmZhxZqB5sdgL(&Kcc z)qOoVIi*cp;?cz!)p@f=T7Lqq<@kkD)Zx;zFw% zP4*{~Am|&>5xpNTXN9K6n8*?VwV*Dk_MpsF>+txTTm+`)EYa?KX^p(DUl_-Wx6?$f zd(-4nB_f&vn~Oaa!86PScJtFhTtRT{&(h@Jl7?SK&7`d1Oc(}p2Z@9WWp~cJFKz6) zCL(4c!&xm<{)F&K3wxL*Z1qm5qLF+mhiSBHJmXbs6VqwdZv@{=67_tu_8Y7jo}S{h zRy=R0(oV=x_EJc@I)daLHuc4UlRmj%@o3pwf>Nz02-mkWRPs9! z84u4vPAcA#G9FN?$ijJHMYtZ7o4yEZe95ooGl;o8c^KQIn(tMeq5`<;zbn`~tvfjp zHC4_bu3iz&XUL~xi&%El zhp`gzWERnfmPCq9h1ME8k9Ln!(}pcl*SE7<=}v@bCcM8q2b|am4f$#9my~GHV&!uI z6>0Qw&zZg0Sdp~e`Fv!DpDGT!OMD9WTmKGjS@p|X7@hJMq5 zD3rp6XKbDwX->{wk<(fjutV^dDCpgcye)H*2Gim>BbiK_K}{#;{qMr|?Yx3}IeaFhn%60goobK_z8=f4 zakJRxzU@d)YHA4=33SBHL0NGZG2ZWWl@?)Kr6Id0*vw=oW$GJrIJ}+29N3D^lN;wM zo{piaWMKAm7U`t#)a;JZ0U8Ve?+uWN3#l%OU|G-~2#qj3x+VVXT`90Lov@ZGwXAId zrwlAS79*pI-%KDTRs$uBtv-@L84QIkUB*J@EXMM~=}!j2(;d5>gfxl-LN8*u-Izr` zj?TczJ$#`Wu{@t%JQE5_$5N%Pbj3b!WGm|3oZPD&kT<%<{eN7&cRbr~_dg!HC`ygm zsuT&j>^+L2wnXe`?Y)&6p;b!N9<}#Kg4mnXs8ypz)K;@bX(?*<_i}&l_x=7net-GH zBV0M>dCqyBbzMGfNDMUMU7h=&(B!y%-^nct#n)Y3&0I?V36pkDLe*cANh*)2>FKD4 zCSDHb=s&nKt&30&ji|d2(|#AD%SA8YO9Ci;hfa-f7EpSfR^0e$npDHYNl~^4rUa?~ zN5IAd0tsa~I{XmvdO?gts{Ix%H5TPWFY8KRumRu2NtKVWSvG^v&Q_C2J%l(4WO#5A z`yiN6!GaL>*)6;$S;~;ejNP*5@n<{z(8OSex2l65&eK)$ltY1kehsj1GWtkApz!{d z{ZR!@J!HeqY?yz6p8tkRhd^mz-|`1Nu53lSViaF-E&=9lE*L3r?478N8Dv7h_O-}=Ev_uTYi7zQJ~KeH~UMa$dJPh>z|=`2_F6u z;^GeVlqY`CkBgF&btg~Ts9nMsQ>wozgw{+~`+P*fC=ubDORLP}k!t({`6!-faHg(R zc*;hdQl+QAS1?3-E#2omX|(Da zNTPizEg$^;&Kq5X49T1K;h&;;DyRV|-_ct_TKtLS8H!wWhTOnyz24@OAiUMjXrLgF zQvh!H+m-Lwc%M@JjR-SJ0yB1Gs`abv(aDu4WXjpV3EK_v5-*C3M~kS%biE1G6dp>RF@!#+i3=uYp;*F)badUkS!NwI=Pb4e!8P*GjSFzPryqcP*TdxL(xPVH6*@UNB#Nb2w&@An#)Y>cr zLoP*R&#Ga6)Brny>sy)LM8CTTsGd|k6&y-z1~Tpm+h$ z<~1F5hVp%*0WKez+&BTFL_fNQ38Y2~*8~~U!0RZn1BYQ1{vP6?p5<_Ge_j3@aW$ws zVGySeH(k5u2{1cz`p7p2!kyi_|J4=nBTWD(bqyK%XxmsVx4i_Tg&wm1d!PtifM4Ez zJR7Y3=C;3ljxk8zbOW`u_2|IBCs8z+i-7={fK0|E@4WtL8R0VO`Y{wLAc&dkM2UKn zH&SA&KfaQ(Met4+wHXVHxQV`oQCFmGM5*+GOa&(drc|%|S0RA=m5eL6{}=UkihvUB z@>2L(-T}~>KD^m!ozqt!uhf?$;5?8)bxulcPAb6ADM4YYVhE zM>9m7@#eH+lx)GQf&dwm8f|SO>dllUgnc+SFy->;mT4YtTJH+EB^3pQ|NMEJF$@sC zam7VH|I1cC9aY)K!vE8Y?$Vzt^ST!EsC3|l>ktJ+dz$l2`I1n7M)Wlo+&eIm1;88;eINCgAN#DOly?Xa z9x)8e5DKYZ1MAKP508Q=e>QFg{`|K2i8&7^$Pb9exU`2~`Cr#0^Au17jRF6A&IZC? z3qf-fYUlra1Hjh{T!kA2E$cr?LMS*NyvM`~`t3+(6#NSFrSy^Wbh{gU+B=R}rxt!L zs9*|vnlm9f9SN00CZCzjX||EY>X8-vbg$LEk5t}U<-SjVwswJjku$dH-=|Q33sM+0 zy;Q){6w|&`3NlOcNXkRM;QR`8BPzH*WN|bB9q_8=H z*P(P=U(7VP?|>4URDr^R4*?p5Q%{d@sC-d;2*>~C7sB|CJ)?(P(WZiK>uTVkjTA}i z{_i0XMgZpf_rc1L3L5*m#Of+H{FQ6KOxh>${flZ}eL&^fsP6Ve!>c3f|CrW>ASa!t z+>(UpB?bXmrf1F2AnKIO301ElBX(x)nQB?7Pi}(B(sCS^4)m=Y*6tVGcZc`X1+J$T zT$i*|C~K9#qE_gzu5~6u2Qn09?;l=_zaJD$so*2Qys;!|Iwc~kF5Dx=Qe4zbLW|6* z998fx)r}fw;v=#yL);g-z&(z@DLqiE9ufaIj67T+KVVM!j>|jicNSVkT<1A``ibD| zb@{ZG04$zH0si)G(cMoYY9O==i!Fcm?)5BH^S?e)E)9iS6ed_PPFxNj41tlMx9*!cs~k%WY@ZdSq7gNU%~XmPr0! z{`+13k>darqOM#xlrke%L->-o2>ackVOCC!ff3et$@<$yk%TQ2O719;1TyL$0FxcD zo8wy5mY!eJd5A>~$eL^6P+#ofMf%eq0kzcK)$XM2^+q8~ebOHS#F{lQjHpNR`Odo2 zBnhJ%gRyDTtb!St)SO0&vFUOQO&;$0l@&3UcJu!~+WQ}CkukioqGWz;3EQg^r;*4h ze<8z4DJQe5P>~F${U`c&62Ymq$6^ZEU<74jGSJ4$?ZbpK`9;Ru&Y7s748=H94l)^@ z&?yn8LAyHm&ERnaFO6yYQ8I&dT_S;i-Rz zoFyO`t5!ic&Sr$Vxn8Z$uFHq(O2FGR0Z&w$lh{Jgt=V;_Z}&500SXX!v?Zg#_D}f8 zRJn4|JtjGrhX7@=p6nIbxQm<=nX*`w1wegtg=UMbr&Cprs4&$e^iF<1O5|~1!545T zc4Gr_-Kvq>Dfs?s4KLXA`njQGwg;Lcj^Lczy7Xg!w?GO7A(x=C6Gx7GS@pLBVM}pW zi@U#0#|dE4ka!>MZ9_G3vt%}trtst?1`yhWv9@zYV;1@8i<5xCWviD)l{kHx8E{;g zCZC&<|NnsY#ew+&yLRY{J-JL*kI7d690GI`zupACR81b>GNwz_G(2oWp>LittMck5 zBcP;E1qd?v72R9WfFXYDj7T)gUi7!6XS#EduSW!#JwY-9Hi~f+uHWgf`;tku9rqtr zj;Rtqs{eM#h4ab>!r`d9w7tGoSp{8E{_wtUIX4=O60B&w95CJTe#!QdhbclS^sNEl zzit}w#^?R(za&VndSb6iY1cyOBae59`Wg84uFfah{Z<(O&Hc$M9B*ySCY(GX$2|Pg zhXorO7B!)o=W7?R7&SH3pZTPw}vqQCGzT^+g8J=-nN-l((a0BY5hQVI(LWJ@z z9h8*DKO{7UUU_DD{9|V!0s2Ri4z7ydXH3Rug7zS$jDlSwd9B@}>!G%nnO=zsDCPA} zr`RHnhK+}E5rk?b09A_B9*h+D*FWDi0?ahgRNRi&4v2Ly$J0N9%($&TqVMUiD6%(@ zlKMzBbE={22;nxmqKIHrFrj{bf}?~bK;3eIyoJd2L{wi!nU1LgE9o|A1@0pFB!;LX zF}!9*Pu3p*4n+~8*}ShSLVal(*P6Sm*Yl>3=5zu5_GL>cwz$&E8-@hsIPDJ_RpB+= zp&A0EDe+01cuoRlF}(+tWkW#`W=x6Ctm+Itl`#?OS8W~^jWjxUgDc>-hGp*}NFsl$ zFv2r?e*rXE)#UfX_8%}MZOi>UqVfhJ^mXxvft{Wd zxq2)n(5ADF+Y`9uD4XBuM7+UqiEWY6NnUY?cTLn$_bcnjU(d9LO=pfZU5azwvY?(A z>tVKYco|~VjD&){D@>Q5;3{O{RxKDolG~|;P`OXX6pB<=s#v)wqq7SqEVW1T%-Hg9my5ws1m(FDHzIju7_UFwT+EI57-+~F3i12PmrKw( zLyhvx_LMm*`%aX#f>8+xd|#0#nuZM0w8u;-Ugx2CmN{fLe z5M4u@jL`Y_f@Rh=lMIbp4ns1q-B#BZ6Gyefz4nOP?`5E*d2}4w3l7di`N&g0zC7036!UR%kP+5%b2_U(wE} z{3}xu!MR8CfPjC5mAzga!Nm)Hww;f0+dc{AgzICw%_;Ypb*qONVTS?pNSFd95wmB0 zz)YqT8gvyIgCglm{+FHP;U<9m-*F4=zD0RsZ{EN%{H6gC1iNX#P?8-#L4&%K-eZzI z+{x4+)jt-zNdCg)2WD{ZeoKm#dZwuJp!ov`Ms3XVjFbg#_<)lE@rq;{l|6Uq^HS-4 zmn25Q=quQKg@KFmHfZPpiECI7p}4}_4K? z^83g~t`WuP#M&LIdyr;s{uTc(dtns%%%>k#;G-G!o z(u~j)Oqm;fk^wTZkIN*hUhn!ikg$DRi)7ezE2kOoz}+W?b$2b}Q2N!9it@Qo5SnO< zxg-Y#I9WO9$(>*SCm!6s3bt{h@t-Rwx?dhwjCF?+U8VlWu%VxuZSADT$ENtrhDoED ztZxas7PZgxf~tB@_ng(zT!GJN`CY`U2RnvB2_YFO{c%yE3Pgyt6lm8E4y;CsqC*Av zr)C&5mqDqMk9A!kgO>9PqkBG_9z$2xJ7fusqtFefRH(12)rg~rQ_SxQETZWe*^!p` ze!KpHbP9UlAAfz}f^>IT2Sf#$Y_pqBX2qKZ^fVuL*wp`ipKRg;EB*W->#TA5PEOr^ zkB+g!FArY6T3#o}2MYm?`>FL%#z|y7jnVr=5NyJL)L^#b`M{q)R;<|k1vCiV)C7X9 zpjjVeVy`4(_3=*A|B(p8l|-U;;=e}%CN9gZ>iqtyOjEU*krz#d-w~YEI7pdy3$Y*C z-Lf^Bzn<*5mrswm?WVWkXylUTJH3_M)20(!aI+Hl%-hQRAUk(dD3p|f8D)9ecNPUBu4RkckQb&9 zc}P8DEzUTOt>Ykobtjq;5@N)JFn=l@Un4121TtwIWg!)g{}nBD0k%yDk3l2MP<+z&r-yMCI}mvKU%w9>7j8NuWjJ1co;M~Y3q zG@(I=UihSG-k-!#Ku}t6-e{Dt7qp)WQRx(vQ=)NXu$!&@0u>t$6&7P%x^}_35c?r@ zQ^IWsu44sET)h=wg!&KXdrBIc#Xp*d`1#8fo)^z04qj}CFi2j^w^}@Sx5M&B28#P& z&A`jULnIJH`df5w8nMZTZ#y$5gM}Z`VCIag_H(zJzI$bR^BlA(^)wgxLxjy}v8UBg zj+4gZdx1|4*TaTOx&5W4Zf<}(##Yo#yB`TIQme>tAV`u5T+x@Nd>VI^)s#LQwub@1@g3~^ z0lhSG@&Pchp?8-Oy@sc1sX2Nth{3nH3155|AY>OD_mO!BK&Y(o>=N}2zmP88Z)9XJW%??W(s#V?nT+24 zy%u?2UayEp;r5d+qnRog%_3sWdTYmC8iA6tSI{QI$K7IsH;Ai_x;j-%N(#m_I6<)R z_Vo;K?bFu-0gyfT9{$+H{rC)K7Gx?+km@}I!W8im$h%O#QrjAJ(G@i9@Lo9G$f;jW z9x?sZ84XJh84#IHrV}hW2(`{;C@Pcv%(?Cr>9C>TdXs#X%#TeLt3Ln_Ccm3Pa#Pa; zsgiWPc{5kzWVuCk=JgM!MgKVYStPVczb@U3P>jXruQm{FBhx%~9Uy>TAFl{~BL9EF zv(i;~&hsp+O1%otsZt(0Tvy>a(xfKP6$A-50g7aJKN;lY&cwz90b0+V0OQy&t8CAU zB!Q0-BECCu+(jxAijs?E+Sc(Rpp4iVZ?Z^Jh1L%Ca!Ltz&1zDO8^gb%pHL$BKZ2L& zssBpQJS8b4WHDAuKg7FWPqWucPxI>bf_g~@D~*f%r@(F0648kg23Gg97`$|9ERwhJ zWpd9*fD}mflo^tBR>;6QMcpJ2=4NjczTU=0m{~&zSTt;%)Gp|Bmq2>!I{I$touAAfX;U7qAK0B_|K{t;FUaW9;3b2T8$L)igleh? zpOiwzmS5txSDsa`EDHUV@mBtz$Z+PVkU>m-kj;*t4NVrrpscLwUg=gjv0PCekqQ0P z2mb=)6BOV#jTD~_SHJ)RSOMRfvICr%4e#)&6p}$U4ycQ7`vDZecgqVL%D{LJhqA5N z1R6VEh~hx<*TSfl5{Qk)D#kA}gfOT^Vhpc+IIF~Zz0`nYv@VjQ*Z1?_+C-o6O~*d{ zzSyuFoRXBmeOu&N$0VoeF`mO<_<} zvPflGE>y3AHJM5X1ew#SpC5?YW2TIw(3Dfk*<;}=Xsq;Fzxeqoq@DzhiYpG>-eAWcX#)W!;0NglUE_tQQ!d87XWc}VR~%M)!eBmF*yICq9!I_W`M^d5ZU zsG;0z!xRPwOv+pkH4{9O%jMoBPg`Q>yJRrE!rm8@tF}un_lK|PxJzH%z-ZHKy?twY zMq6X-%1g_MO?F&vNm=lb-HaySr?H%8B%&u)8N&Mt7&@+uhCv|9E2WZ_mOc%F0WBEBWEBV!a`f~5=GHozWbiJ;DF(`* zlU?zrAN7Ib$*WoUIOrs2x=@>g%13kp?y27JUu)mwIg%~!ceVH9mkD|fUF(W!I8Jl* z15`e8AYxbps0gSSO}ZzA@a3j$&w6@v;A>mJc;blxek#x?$XF(xS^#k&!UB>}Bge>O zaWlerhweb@R13c0C>oH|-_Hp?1=<5wsfi19$-C+C9`-dAIj}|AwjWlL} z7Mel&*Ytp^>dGPX7Rgmt;lGp&;O{uctDK}4rO@ifiEgz6?*q++vgXZ0{hwc!t~VCT zL$hCrJZUQ_nm4WS_0yV7Csn>cE|hJx1$B9S7!S&AzsG)jGQSxMm0witNL@Y@kd=uS z7@s^l0~oAvxvWSF0fqzoWg_$b1rC+9jwOT>P&IlKW2xxb@--qMrqCEyPHu1mO2DC_ zaFy<7V`wsFkQnniD+zj=Ka%!m>6Dn$BL3X4629?CyxjdSAIxR3N~yBwz@sN(jUi5h z(O7|zOsGrLM51lkqkH_x+3__M@Z+E82vhl3>FF09E|w{#L9&0WoeJUWo!)*CQ^8#d^QkW#WZO!@0N>_@`uMYEy>V7jEvc*Kkljyqal z!-CL?^(Qa;{I7d7&7nvJ%Bo}$Ci)apo{7|L`3TGQa<=ad_(V;d>xDW*Zl4!zepbz} zkv&@1IIOvQ;K@DVp@XwmL#)LK$|DXKS}Mhnlab-jHlVWz-1FRbZVW0pXGc(R0dG{6 zBcc7vQG19mn12?I;Utgy%}HEj(^IxeV(d&?nCLJ`HZ(?F|+QMJ1Zd$ zVDF_SBy?|VEWz~XSxo;!dHD{4oRWO5`2K&g5rh)MpJB4Ufl6K})eN{TZDT9?0+3R;C`(>5IqVzu09Mtvh z1g@Ya2R9hz@$DnO1^%)t2{wH09EXn2M7^x@njPP~W=s41bbq&Uj%@uU-_O`6m43iJ z5cBWii$JjFk>XaV+Co@>EiHKT$-v;vf6SB!LGO6JUY`pd8}eDLZ}NZPJ*XHa_)erl zx0sS#{{|o5*2w6L?eFRfS55}(0B>> zvE0qrco@hgAF(?7rq2df!knR3W%$#ZG2+ zVa2BKV1MTU?P9KTP=pf8THYK{6=^Xodorc%p(doV}pl}_Ceg558^A4Qsn?lVcv=t`1`F4m260`{a=mT5-#%p6)NP8>fV*Kn(8 zo*Y3~xSvp7E)_^7OY_-}WR!4}7cJkOA<41DCko$a>Bxjnn|eI{@+0qi`C>&XA;uvGAU{g1kVQIt%8 zuThuGVk|{9_l8#RlZ^m7DdwCS`ERdYfIKJ{uHZuwC*}ETYOK}JB@{)ydp?&=;TPPpKKfeB5EFgpxUiD=+Sb8B*X|si z+m6nSz zu0&Vwk3LSjKIM5X*>&0~`$95Vn0=&?_$dyxI~p1Ex_psVMTey;PN?jA>98g#9AiAE zWfSmtV^D0FYxg>4eSKZ#Vn0&4Glufn^J4nGm?2{%tVuAUJl6pPt86<;Z@V~}rAF^x za4nKG>c||=c^$>a26k_B44%=T)y$aw*TaBb6%i)W7hC2|%)J;FGLD5eyJBsHID&06 zX(@Kf0Y=q!_~A`KZo$q`CAz(%9;s(~Fjnca>wq|BH#QdTh%b1V52%O4(vZKj7-oC_5q?xrX9K1R;vK7%PH;>1`HofKC3 z7p@H!V&pmBt z3Z2DX_5)Y8ip)Gv=i^DdNf#?mKNsx#Vp`5fjTB=@pR_ zj*q}MAF^+o{cTQ$1v@7nnJe-dKm3LB^b4FOSsMP}w_ToR{q`?UjHjI-f+8qT7+fy7 zxPj?JmGaDjfAAmn-JGI2C>GK`l*~K7rcK#Da{r!2sm`;{ZbcFvWD@*%%w(f=4`mKr zSmF=iq~BKU3jkGAXDHy|e{B$x+x;w>J*$~)Ux$AsC&QqOFc3r#)2AtyG7@{} zEO3A3Yv$3XV|!o^Tr-gx1B1bKOczg6=3D$mzU?0^-tF*FIJT2JV2*$_k`En+cYUXw z`d@X??y8z0rD;=P5xgP2c%zgY_DhoFFk;Ix zzFqHdq$){dgkgIkCvABsWL>L_|@<*qoq}cob&p=h75p>TeU*L^=xz8tJ=VWMEEp zNRIG9)w!>}1)6}K>N-N5(?l?3>~Cxam-tC-Zky|;W{0yOIMz`bSVnBDyY9MMD@(z7 zWa?`w)zObB%-_q7`h|T{nBzii`sXkY28pFS*cj1M zMx8)m+wn4X3Xt%!)jIO6PjK z5XcZ@{7Y9e8D=7UwWUurs%KEkQ<`LWO9hy1nG#os|sS>O5SVgAX4Se$Y5QqdvviclE<|D4nys4WE8b?=xWMW zub*|e`Tk`6l;H zsy=fmMTy1HC`&23gQ;Otk9a+%F^vlHDgkg_gP8lD)G7w(<6wOya#)e!8|IZX`Hye> zQ%I%}T<*~14BSZxCh+6{*8C@|{P(??{br8WAeF3G%Z2<3A{yo_I>OVf%0%93=Xd&DL+3eaKZ7>>}3M|&GKr= z2q+vL1D48|l@=8%sr%JtU`(y=1UL$d;1ipc;EU$J77dF;o{}e8Ir1S36=qG(4h{}j zD`cBuvF_~0QLV!r7p=p;x~2Z9NBn0BLu#%mFRSgsvnB;dGxeya7iQ`Q-E<*i1oK3RjN-ssDrbkg9d0Twe0i=>tgFWBTp zHG(0BRd*7_Q0XE4R@bw zIg25akw`eQi|0^~nq*vy-8!EY5q3MR=I(dhTp~TYDpK7dyYIL41G0?E^4=telZ)p8{79WD?D~OWo!oiz zpu6#pA0m`u?~niNilH^fgxhLnF;q(Jm^v=OGh_&J^=U;d?H5&z1z&S5^e zIGH3$W4pS#`HsDhWj95ipPo%iy*=>!ZuKWgUV*8g#;vZ;QOycLep9{5liNyM8lv%D zE2b8f>O|DP4|CTH2)2qwCzY!W;sHoGrLSx~!lMJ{ObCs(2~FDa_x0dYO=E8!LV6+y z&}Q{vK-|%*jswACgag&~kHZ3GbnnfiZogY`w!^+nhPK>^Bx|Ibe;zhKT)~quRt6i` zrM^Ll&ithP>!#c5O(R6Zr*#prc0b5qu#>}B=p>}F_1RwgS;b%1CbAvIKe9DHTh?Ue zURpGJ_bbADMF-MfYP6^RH04`eJpF$0yutff$Ul|k`)DA>q)(=^SDLl7IccxFJb0mz z_M^C`@z&lv>^?PSIow#|zlC(7; zobf)Uq83H?tZqN-2r3z***8kz7oH_|;5?t&Q25gG<_onod;c8freA_CP5Z=hr5*9@ zeL&4N_mKQ9EpCgrM{d2)_8Qzu@zvVE4c{jc_th02hhRdx_mzwulQ; zxt9@WuR|J{H z&O5^?sjJf(-Y?1xes`1WtuipFy7!MCO1omMLykl5jVjs}L`#0O`OENLwRkMeI<96~ zwS?}vMRPnPU0Eqt3yQlCHDy{dc9{2V2)LeK(jZI*qUS2ZeEbDB6dho4sqpP5&5gW1H*?~q+(v_3$q#>*~^%uRiP)j zD7|(>%$C2g2YEuj57#bP|J2b^(DVJM+6Ck4qIDl*@A>3#COj;i;;~0S^XxsT;SumH z8z7Ela_0oj0Zz%{Ub>2-9)@qO@m!f@qm9e$=70Aj>kS;%uknP> zOU{G0akSp{CSr1;k2I9q@34MzOsMGKXNUI_TS{~~27SxL(hc!mGvMS%2iWqwD+|(B zytER~w7dh6YOt&yj!7+l@t3R`bvg=e-n>b2U_d(#8hdo3Jz;Vbx6zQ0e3MZ-OrGS3 z2y>1H9@Qb)=UKgDP;q3U$p~dR!Om1aL*MgzSvv4sL#ker8Z&cjZ9F=85@q^aq44>l zi4k$KFu7UKqT=aABme?ypnL>JAe&p$A{R^u@&(UhYGO)mQdD5OMj{|7-mFu2l#7#(t#aa_rFuN&Gu^Zk+} zbxpkT2Z1v&li1it-dqlPpi@{-yK-AJDHK{nj7gwuY!-BMY($Q0jS;+CqHFLbd*n>u zRC^8kC#I#ry6sH9>SK(2?Z+%Pa&Gg6C}0xnOqPAhPeztFx{nO@B#J<+%S?Y-h~o^T zsG5Xh3%F#R;7z;rMv8X(dA;Wc_6s4FEE{wBj0gq7J zQ7CG&aXmR*>}Ekd#17nnqUFWhl6x7;-L+jwYy>91R8{9f`$PW2Ad0T?aAqE6=*FU?PcF{j_pLB&rMU8G^S=s`wg#r=86{8&Dd7Iv#2>LhQM1aVkQ1N#(!Vv zDuSk@OHBvvZkT(STov71CRTYfzeWwb_!t^MwL8pZNGJK#u6d$B%#?)tK0+{bm7q3B8HMJ6s& zeW@SuUg~=ObUfSrS)9EA)vpI~omU@^n5q*kgxY;h7UgrFyzyAmm2kVD#MOEHjmt4C zNVuX7uVtFdV65q%7qZtKmcR|{_wW_}dCNnO+j?v*7yr1J1OBD#$wr5*%YFt`8Uc3s zx1#t*;wYryXQ41w9B~l+?AdCnZZ2^l|D8qf5klSm@Fj)0y8Fa&? z;sGC`XvCF-cZsKuSb=WPvKw`v5hT6+Y8Xg=EkKTo}_|v`8Hmw7#lcLax_-Ar%zQue9&g&g|V^yU`OrP+rYbZ zq=t?zCGY6?{JH7&w==rfwyu%Uf6YMXFAh>aF)(pe6=VuugWyv#a7W7{aJyXrQ?x}@ z=)P(NMbLT^VOSmU=fbv>_O4Dg@m~;M{^k2Yp5O4c8jvn6Jh|{M^G;9gnavj#G70RL ztDe=v?6b}_t9#trKy?on%$hMi`e(t-*cbgpofvlhp1JgA)?Q$5g=L4_>#=6OO}~LQ z0{7QW9Xks&X(FcpY0IW14&Fe58M6(p*Me-={LSjKAHsvjoUoBD7b5 zWp3Gr*cmU{FrIbT5GW@wF`iTWV_l*q<;e zgGbg@>Q=7^))B5by_F`@4yb)%unGaVaGXEHy)iG`vt3gA}VR8va^xh9f#uevixIb;F zvh_q))GK6ypk6986^Hfj;@&8{<<_{4qnf&#o3uQPLwV__NvChC<0w{ynGd!e-{9V0 zdL1Pap!9cOLv(prucHa!cFirBwuvMZndlThcRV`6mqQna5j*1`A%K_hmZ@S#6}Z1#D?URVfc zlHkTqeTAP=hO%QGnMj5M)2giXrh#5Bnc1l(<+R4>;T(V3euwqLBd3*v9e($?PwR!^ z?KJu+S&r^wNQMq8#MrD`hYN5yM%3T+3ya*(M0v=^Qx)bl(MT&y;p-HJ_eE7tim=(_K1#NsNh!I9;4qQ*R?74l9 zKx|)qFVk<_p=cd_YBhO`x@>r}FJ;-pDEuoUXAtI>h?>0G*l88Rb z7mDVWebQQBu)*gSpw7O0C|UIT%P|eG9sq*yl43iD@|xx6J>UHf)lC;0v$jJVc=|yv z)1;s_E>MfoXiWk(sH<#;g!nIsonlW!^ydv6Q5ur@5vp&;I70FH+h9FEB2(9UGc)I@ z<$u0dRu^-`X>?a^@!JK(?2l5LX9<2h$MtdElg|>jh1-`Ai;j;!h3T~&lfLt{pRVS z>{Kpg2{Y23Z#8XX9~{0Z`f;T6$TH|FekZx$%Y%F6AUn!WtiNpjCr0d1yS^8aX;RM#bCO^~}sfFwmQu zb;}w7t+vD7u{i_JwFaX;7-DE4N_5h$3Eu<()u#y7O%%`9wrZH$^Agm*Sz|5pWA^Y#br=yUQ`rFNB^aBv}*d~oP`tKOMDxY$!rVYeW1C1f>nbjOtoa6qs zeTQUbFhYfl6Q|+F;u`4P+5~0T-f5LekOQ7e&4BVz|s`?T};uH{1_ez>|U}E}yyUn#7g3-?L7OGB4KmD3w~ ziKoxK^jQ}p1P1M)+H_dL^`-p1!zs%Kzr`RwBGR>&g4k<=sEwSE;1!GeCMh2E>e5|2 zV?_LGhXVsqL>T9{;jCtz{H=ud*TT0OIc=VhM0UE|b~^fHW@c`F_>d42nOaL=VN_vO zWtZ*$X=9Y7M8?DApR$a*E&xb&!>SsOOREJ@zQZ>EITq=6j4Q>=QzOM!#k3BKQ2Ec_ zYHKg^gc@AQNBdBbn+n&OUTT<|{q}^4{0yoJFiz2koeeysZ6cmf@Y^Fgs1d)kjz-xU z{piKB_dvvht4p_B8>bDNyBa60y>Z0Y=G+$}lblSA)2wjG^?ex!{$|j)R0W|8L;YC$ zZ8QW^o{(DfMMNW=rro4s`eX*GQvgOn2gWm;S2L+V90xXa42$e7ZKuV^uTeAIT?FgK zi9%EFPCgFO;T)eb@a+O~7{Gc-#PH|@5kxiP`5OvM{&+bi!{77ECpIn+$xsfl^Np#| z3;SNyGwm%8B6$FcdbDdwLnulT^zaIcZoz_!h3l`?lDd*#j3o}9~KpVWl}a+h&q zM&U!QN4z$rk@w^;>=ugV3FF@O5F`$-1;pJ9_Z!V83Qv&|H!{VW7&rLVtQ}}2s}!8i z#7CZFD!(6*znLm_L-SQDhu-3v~(|nj`kvIu&vMT56ux`t_h$Y4d zn?MWH=fIi{1)7$^B+>LwJ$6_|qoug8>(LL+KX)-9!q*5HsRXB&tya>#ZO*Ak#ONiQ zpL@{(s1wp;e69~fXCvGYaEt6(1?Ysmh9)wxJiG6pAXd&%{@{`P3QWD}dT-M(S9Ghj zrr>^;*ytuObd$}r8#}!m{p0+@!BRKoNlLySHerj zOSxPMv~Nb$l5PLy3n|7i7!^LCVv3OCH-ygKtKeG{ z+#Z=QcD^W_OtUwL7@``_#``=l=C?xQHp)Aj`G`o7 z#Jg|Rs4>LA+$P+$PbE=`inSVxs@*lQ%tc)cSDGFVAG@^I6jvK!AO!`XpVN>JiYjvZ zZ!BC-TXC>AuQQOO;{sq&+$AiV(Aqus?!g5>TH4M6+=ZMfz(Zfc<@Eb<)RLGxyG4w5 zxP&((1LtnV z%{Jef_fO_L9QhvO&CWb`oqdzs9lVmRn~SOw^xBvdg*a(H(vlW{RQR`q{LnGkIQSb{ zw-W~aeS?!et((>(d^AuPvOK%!_w134d7fqz2qNU|Y8U+m1SVF`{`K1 zMlnO30ET2mOwxq>>7rut-xzV?HZ?YE`+Ipt9f358l|4$6J^1{H6=Yv>D!Usb)3s0N zyiCFYN}j*=FmmNn4C@;k#to0*qYaDp;FlHzFi_RK;9t=eFiaJpiHnf{u+#LP>J8h~ z=F5BK$<0`=Oo05o1jwJ*r!UXuYFAFDuLg=7ChS-YzkXrU-4267hG;~>d)O6uK{Q0I z464mV-+Q`>7Mf-X>fs$*DNq}&d)1;mxbZIvKU3y@mx~b~8i>2x$4lDVg?^PS>bNMA znj1x>rx(DDSoMNRtMRC$)vSW*()J0}EzXH^1750nGDoVHHyiFs@AsGIuLiqAbFVkN zNh&KpFQxr3pq!IG%=g2w$&tyrpQM(%Ela2zd*0cGnBA@HaB7%ZU{S)B2Zi1-mN;L( z?*@e2d*660m~oDDPHXP|=~4)l4dyIIe+wGb$FL&>Om-O1KFKm0H!mu}ND5ten7JwBm~F|SiuTk|;{hbvAJ+;jy|RugM1 z{eNV=byU=U)HOOohk&#+C_RKq_aM@Z%rGfYObCgrqP74BahK5&{F#4bt6p z=l8t#UH4h*{t181`hLzid!N1ciCjIJr13fV2#RJJ?oH!cHEdq9w;fuCflKOY9nne{ zbaZUWO9?m8)aT|oTn|N&<5PmcdUUdpSB=z2?!otjjS3j@9Wb_;&tb%qDFg3ihGpcN zvgGysj4S?LZ!y*l_GS`MS(5GJ1Dm}&<1N#^Zis7XS^yh`ngA(}9elQ>VbEIL_Nam|QZ{edY+RvX}VLoYg-%gS` zRdY@6bF(IsYd>7SJN{;5GF8}!VE{wlKkiK$oio8Xa51|5PgY^fgt6+g_&*T?|Fc{D z{}6`E0lKEOiNqUXuGzhEym7hzj~v`t-+gT;TTZ~aU>r2JmAo@h^Au^PsYPltku#`F#YbbA zyNe&uX0Xquue5LxvWi}5wBY=dVx9xPzQMaBz<=p6ue0|sdm9REz#3csl~cXpC@-8! zB5B53W|U5ylj<>@e{p?ujhjzhJw9(ZWKu6-fA}{OQQ@jf^3DG*gIqb<;0)~?Xypq= zSxya^y9%XtMrjD~ApXIIo_4JO-5|CD*)Cs_ozthX|51cc;weQy9Ao9A`KcT;7Or2| z6Y8RH4)wDoD#NnM=Kl_;CwE3`E*0M2SO4~lOu%j|u+0(m8EvMDciH9iD&-JG=_i&} zvp{19O`--$<&D;3p<5N&HGVC|FWSl)m>77Ibkwq#s z+FU91Knkp3+b)UV)YV?@vaP4Y_I$FrEb}{de3cahe<$a4wisoc!nd$5s5Y%^i(8cw z=&5`A3=n>Lr|ApEbxE5)F*><~ckOlL#%xy9ZMcyNNLc1G12Koq`n4(A|h z+^LQk1&^`|7z6%b*6aN0F54UID!784X$0?0TF zIP|+1kvR0W!pN&MZrKdpLFt)#M8C)p`F(h{S2Nm7sC58LepbKabz+^93KkpuY~{Mx z(e@;TP|BI`3VoXDW?nDZU~1nde)RvPpeM>Z_0iU#b`f$jtw+wz+Sx-`2Wl; zmQ!cQS0WSDo#WL!aHwE|Zo2KudhE=v0|Y9Ik%D@t>l^Y=$O(mRH<1}}X-5SJ#6C|@ z!Cis6aPT-U=ZqZ|wLNTtB;yRN?Kq(s)m;oIq>GzbSDQ-r=j09#$W-JB=g$-dNVX}+ z_P|0PJUkF?+UBltOKT_NxX<*viN;t!Zz|IWldrtP=`xdvTjmUxcvNe~m9Sa6bnb@u!z+KSpUxB^xLL>q$}Fj5MY-L@bebyvTlBSaBk1BL4ayV_ z3WAAyl#jna%Unu%+L9Pwhq_{?QYvkQ&vCtz{0IsFNqG#=j!M7xG+{=}LycUiy)_HK zf|?K)NCg(xA+aXUeyS=M2=9ArJeRp`aDaRmfhADNcg#mtR)maxUy=>2aQz~X@;;WX zGZRed)}sqgWkHH7AtwAB1;1%j9_4dQ5 zO^*BwV6c77Q&G_WS-YV_iqJc@5>unL28`sR#a1t3ra*VJz%#Pssr9%2@uGcHpZH~W zp0?k!!DMSvY`BA2*s#+Do~Giv-Na5G5n)7!ltRu)9J6=7w2N-~n^HQ859b>WHtRKVm3Q4Gmfl>6FOc z8w=v5M!U*o2W5}9d;_UVk)TSyc7@Rrf2_g33FlM2*Cp%EkMO$$n4r1Dad$-ssU|h7 z-~}iAvjMUv_7x1xj|vsyoy8||X=tzmuHh0D3Pw@rj!V~wdUQs)Zj98oE)o~l2Q8LR z8EUM3@YOQwo!87xay&xoue^O}+eqY+vqTM>9nM99hAKSUKV<$;SuW;kaIgNFWZ3fN z`#y6CN9bzq{YoNF?p?@Fhe;;xH_*(Xvb^6w1A05#@@HuGKCX1v_rMy5-Akc?RbMnj zz;t1XBky`i=D%!{wR83SJ)NZO>2st}jCL zwsFvED3<^0T9w#Gx54wE%ua$lx5GuLhKlAO8TCzZ8JR;Z0_xnx5lqV!l}adlCRtN@ zJ53LqtNbJ-G92s^tlNlbsljc8vC5cW&3;`X8GNzN!W8syy=9C$GB7%`zx8WtTgU9y zPkVCL9c8-iDNNblF{J+>#aI}F7HFUPNh=qL?_X}Z)IGaw^P4|IPhQ&S$qRqFXx40h zF>?Xtqn@THf^t&uJLwAK$!1UYp_KjBH&40T8vF2hQya`q1 zHQb)B+Bhe{`l_p8G_)1t@_n!`?=Ew*x|vg}?sJP=6XX1_FKGDE!_-Z`jRS_(Gu~y} z3XUwiuFbt4B&kDh_+;LzkBXr!B_$P%X%nT`&?ev+5`*zVlq5O~12o>#mieJhD%}x~xLt<f*>t9$69K*trXY+Yhr{mm4Pzam*9!*aF@b5V}ziyS!tJ%^4yg^Hy?)AV-=M%kO?O zHN4rUF}%MS682t$z-NIgz<0PxTwKq(~O>^Hil-u@7<;t*f@CkL; z41JBlo-^19-iKHYK*{}sfG7or7H#Q|3L!D{R5Q&e{^Qmu;G0BtZY7;@jTEoBUsE*>!kXnHKaBB=zpqIw$ z`!&8bjSY#NLoG(51YGdC_;AMr`d==6@ewIsPP7#2eWs3lDqo{n(}T<}SK7;h|7qSr zT$Wn;)MYAi8U7>U+>Og3kxMD|Ee#IJ6(^Z(sPco8h?x}#rJ08|GcLS%zcWf2S#Pwy zVBqs&SieYib$dZtF|RJC%Oj-dwD}>iDZW2@`5@uT$WL4z#OJ6H(GBcIT}a?px1GCr zp^dFXtoA!B4c)x#??J}H;dP7pNkSCP+c{u^xp;?fgZ85Lq1AVp=~+Fhr%Z0J)be)8 ziWR11JyT}KayLyrib^7}ELP*gj~3Y3@bkef{~CDzqt|?$IrMeMjTP zqEICn-$|5ISGLa16JFvi-xCSWAbJBU-2z3*+6WFj)e@&{RIw0t8IG~}^Qye(2RKb? zkBuu^%Z8EZy#?kuWpU3)oblrz!~#R@AkoEr#W3ya@J) z&FYe0E4J{XD;Ipr+(M%Ke<~MkXH=Y^_-|+Is{*F`e$$gb49|M!Ay{!8Xwc|bgoWjM zcihLY3J7q;Ey(p&3383(_5tRt+>qFFZ`vi)#?S!P0vsE!Ef_1@wnnavpp_jD z&4!5^jhre6W7hY*{DFVAs#m=igH(EBEgQJN+3THNsW&Sv!9vXTX|1_Jgngp%4&w3G zCYIBQ{_IB?8XK)2WNfQOmO@%bwl!C2y@!k({(%==&h;y&8&WLDPSKg3+SHRw8$&NM zZ#^p^9V{;JgWP!dLT;#~I)f7ZcNV{~7bSBGvTPiJWQ|UIrwf!YS}H%h6Vz!#2Xu zMov)>3LTZY{MN&F_SyM{KEwU{Eom-P*^}VI)>uO`*VQq%i&Fm!c#|O~6Dh}|co2@|ZQE3GJ1Ae~7xGd9 z84Ww32n64V46j!pk(8pymiNy2`QlPyR_mspr`{2~#506XIV4b8_Ir5d{M9S3R0M*% zk#m+M+rKNP)9x9sPdE3D5#>;&VzNCrwS)tPN`m(YH9Z=p4YdmmN216P`9?Wx51g7Q z)%mh_9>a#%pR+2Br?w(65$0s=tKyv|@D`q-G7<;*;|h=IRvBjm@$Wxx6&fOa2+f4J z2hW_)#^(p|-)wkhsi}YKi}^S6ItkzsmT+?1BB^=BJ_SwIjYjA(B%U|y1FcN3xmhox{ z!KSgI8H}EmVcL09TgUyN_ZI!9L*mnxZl%-iSSi#QMOHDuX0{z}T7{0*&%$urvN|t} zEE0b!p&DxA6f}|6{y@3gaia(|Z0AxcLJRfP246ISHmq0+Bv@fu*Mz^1i_UORal=yj z4JL90{i>2F$&3<4n=PDP0QkbBSo8fKXK``f_~*sRj{fn6j~qmZe@aDikC8HW-@WWk zGvWkz;Cx1n%g89-yyPUe#?i5^tKR~n-qsu1lYh+D`hC79(}LwAnHkFVa9J%yLxdx6gHOpbO`BoKYyF z#toVo@aPLTWE8n?4D|dzjCvC{(#>tdqbr-Of!>#yqet(Dy#90J%*P+Ysr>!w%bVc% zb};Dywo99r{eFonX(P&5Dj*W2GfZV4e{)`Ua)+t>@xc3YIO^9IaIIi2DmC1CWAO0X zsOeIgy<8p{mFaK=&Y#_Y4+W9@PCq2UwM}J+Uv|}qq;oxpn;qK7kgSXcV^EpyLsd-q6&myx&OYPk2O>QG#MfWQhERIx1Va+(>Xx6l zshvkj3GD!(hp4|<3f_N>3))X}^}-|1@D8TP(FQ2YZ9!{nJ3>xp7OW)F?j z@Nhvd%6S52D_^SfELsTfWl%P{>O+*LTOg=@qTDRLkUD|q<&k^Q3x>Ib?OSF+&VTyC zs%xB4QT%s8_N`|Ph(c28-wX0(m~aEr7ree+OAY7K?Y8iaOFC%aB@w8ccR(s{zVF9T z!0KPcAU*lcvdZg<8&thePRhBQT+w)4r&yjI-|ht&QYIMX35k1f6Z{ruGT~w&Iq*NT zOj0Me7gamjdADe;6v@xf0@DD(UAJf$Y`_48&Kjbh=DMO=V=JEHEMY*N+?kllGj+kD3!RcR@5^Xn127{L>ORS+$8n`` z9UYKGpcQZPh`!n2$06bdKb9(da8q}w3l^j>m)Wuq;zglq~q~phJfAoH~PWpxmQ0q_IjyJ=Nr=* zJoJ?~mShxVv9|NrF4=|)8D%0ee5(8-1Kq&dPsciVtRkS z^mvauUzMcs5O;FF-X8rcQki4xT5xI}9=PQd*q8f#zC}T(kVn$6<0zc}3j8^Hkh*h6 z(2jRpasUk`QH=i^Om0uIC9pR&`z;+bNY~a~xDNJjf+D?~|D>aS*f5sOu4PK>xahNI zPv=aY$~)0b8G!b)JpInq<1y2H7{ecsACB(cKrqRuWez&+OMLQmx^J<7!S9qzQ=l1@DqrnF2nljPEP}RO)xwjJ9$k*dh!mMFM|PL*hSQqcd&C#PK{I6{$cs;c3o1Hfgw@)lDMR{ zPmvcHt1EFkm>7|1mDr>27jxpEB~JxcSwmV>Rv$$v_cIUnn}?#SB$C(7iMJ?R<+f zozb6ruT#BOH|I~Z@QEn0ULhGFbKEy9eG9M8A7H;F3vTMV!*H;uOREEu#p9GyRKLy2 z#`fu``17TSWh7Jkmyo{g7|^7=IB~1TMzTWxft;vh0I`s4_kY_(%-!N;`xZm8J28e0 z?270#V?;=N{@MLVw=hFxCw(4quIBSGvEfjmXW2)M_XTp9w?x;47O8PE3Tqb1sm3l= zTJmDThsGqEU)K)XE^|g#X`c9O(# zbYCsedT8{Y4~p@S5Ps_sx^Dh+s<`%d+gnc2}WN9a5E{S+PT?K9)zL`o` zvx)j;`f!qd&zKgkev?@pg8|oOR$3H5Vz3@d_IG>cKZ1{vTDP>)#@KBF(>{pUv2ZC@*Ye=IY>F+BpC^hc&>|#KJ@Yb8+Z?OM8!KN=Mf3 z48191PY>Rp%k0?^c}O(IfuZ%Iot$_;>XCFOFz?56MA7j3x{s-D5V#47bv7N9Xve#Y zsMPOZYvA9hp>?G-E>Pnpa-5yd3z+fN`eC7bdl#Pr3#8(BGRi30rZRn!E_c`KJ18eS ztdd**fU1k}{5N7+e?l09+A+QID(6hLQqVH9LcPOmVK&>^9;#z;X-K!EJlX_J-HjaT zpQ6&sJx?9r|5k8nSN~i;)_8ditF!7obF@7(*H4S>@}|b11E#VS>hR_17&)g72K7T8 z79R-slDJ(HXTd+h3z_9(hVZp49ai~yhHjpL4krqzUa-vkvcT7$BU+;wR`bf0UYL+N zZ4b~`yWlT(cL~G$V!8207RuCUOKN`ZKSWVM^*dw=Ng$6=Pr+Ys2NQXcKSHdXQmU?W z+`>K`ojTMZG*h&cRSvEJ>ry}cGM^^wQA(T{FW+6{?i#b5lv@u-A{0_2tBMEHU9u2= zPN8(h6y3xJovrmW+L%!=Eh)=5H-^&F5`cmhp!6ZnDM^oC(R`yqo$I^P zmAtZ8!2=(Cl~{6)%`r~aha>1=?_}v^VJw)AfCXQnexPhP8zmO`J@J;2(JB+A_kgLf zu~4N+TlJ@BR06^1gM)#iIZ`zM(J8=suh2!o^S!vFprlHHVzU*6`x;qYoy&>Ies^<3r&IxP{}Xd;(y`I* z!GZ9&1ZuLht<00$aAyj0`CMSeYqf3yDRc91Y_gLx`*#(?uSl0K_UGhSMkBOMQ+i3E zjqFGdKd&mrGkV;_ zduKFizzhtksWQ(*?r-ol(W%Cp@rvak+N~lSyzj;@4+^huuaXOZny`STX3IN5tL)mg z%c+mLriQLKhw>LhOKqDCrbPnV2>mf}s9U(I!Kt$S;IICi;;HKQ(fNV$d>L;ZmI?b!AYzvo_ovvKpW z&-dH*Tbsh7@$5$JJ42R9P2BQG)@yn4YugJPS4=q+I2J~L2Y2p@qHTZK5wIRR*)3+Y z99EpU&{Dp5Zv94coxwHQ+pgkqZm+WSI38Ki~D6;zVgS zHufjbugWt*h)ntGUzFx-HX^RcHf~W5|C03|*QJ$fw25D1m~0Ie-wYqBIiKgusE=|# z)|b`}^4Pq&Hs=K<^u0NG(!OBj+4*g*W!K*ubCVQfzdSq@U;)5!GPgplKE`~(iMT(` zKI`$J_(VC%9sGE5Qq5*d4~Frjm!ouB%e;RzQNNCq+QkMU&1J_)jH3IbwyQ8f10-WK zp&0h~!T{Y)l};(1RsjDDGyG~FEQ5=~sE0F%@6)H|eS+E8mqf4{4M=FNu71Cur|k$Z zI9u*<76Kdgt`ZSGLF^!r&$UMXdmy)YaX%zJbNzbJu=O?e%JD9$M9zQuml7^>wr9Us z+{}bD$qNp^bkl(nD2MXoZeExseV4T_<&d*G$~}=VK(=$d_|F_NXFoZ`LW|yg&j{n9 z21!fZJ-Gc$s)mZ(QQmfXt-@lKWS=3)`>A>ABwHTb@Fc&zKTE$tpB4!Rc9nWJxv~l`xO8BINzS6zk&v)uqDR1cGFsjML1e0?c3N@9}bY{%TUWhsZ z=}yLLjRRhE?_{&J@6qb)3x0B(@hI7(IGjJDRL`GoD8NG=IQtXAE4-7`A6u+8-%40+ z;5ds*@Hc!S{9r&n=k|OKBUpiMXrTyyv%0WMxc;qpxhj+4O}P`(cCj9baKs|l_&9C_ zS@g_lad0F@QA4G7qyD^;UA{GJH_$q2cfyKpsbOdZAKs`l@=(Qr7zt!(CiCtdw9U8H zl<{%xRuN#g60+h&@R%2!znxm54?`wB9enm0XQwC;?Iez_t{2zb;!FC}ZpcvQ8Hw|m z?&D`>Un3_H?8)sb4_Ytp_r2?OT4_2EpeRQ6Nf82{;JH6iGIWxbf*HYT4ck+Mky2lC z;qE$psU>4^XT~5&(nzmOmzfoTe>h}X%9>qzd>`8ID40<;Z!BcUUJ%3uWay{u9^)dm zrkhF2xSvTmNojI3Xig|fLGiyO7q{T$C)JLfuPq%m68^Aa=!Wzj+8cfx|D@l<_y>@o zL&KH6ECi>m?a8wi0pbMXeB=ZC)E*!bnc6+i|S86wk{XAm;=zy}A5t!J=*So^3(D z+cOvJza$w3eCfkQswZ{)845Gdu_YHgkb8*HKdR>!XlWc~*J;n0@Q%Ip7i7MA)vGVK z<58Zn_s)(+m?yDO$Tzr?9BX9uNyjnAzki*}(QVN0pAM#ipbp2oYx#=5p5G9&%a{pG zRd#&mY5dZP4nWPeae;|RnhWs>-A<%gd_5di7Rownldye|ept8ZB!`+d^MxzhzPT}R z8){uL7N*pHc5YaSKh9OIyujA9V#sNd?ANA$AIDSU1MlwsG5hybOSRhU*sP|<{3+0N z#{Ldf=rq-xd1Fuk?5L;GTS$X)up&O<2>Zm%0&Q6abX?6vtv3l}(p$IQR&Q{Hwq>NTKWQ(O=YLDc$b+HsIkFm>rlzeOY)-oMRI5)d>dhW{t2ZIuj)EM_l;o3s|snMZk0|Y-~F$jD6?Dl z4Y8kF?LJ4W-yD47iFiBycUd71=&E?xpMx*sb710Yj?uwVmZA%%$=uQWrHKv`g!c#c0Db67_lfHFOHihE;!fq_+F&V(UzWAJNj~ znPHI0OQsUBdB?AI3TGoXxtbDGi7w7e+ODPg_*;oliEo@tvjo8VxhVCFp^(=b5WE2p zlm5ZrJcRzV95LsSr8c9u;sR!E;i-@>j{*_2&!+mk#!0r!$)UlvuouV*BEKA$dGXYt z+!2GVj7>4zvk$|}2jPHDk7s6d(@9w4UiT3%@BrK_Tz^H`nFY*bT;CvM*{kBHmWy_y zlloA77B%2u_vBk`Sx)_EIm`faG*&{O^=7|0Ul9*J2&@ zc87_7Cl+NQsd%7qvdMZ1gbA)h#oyZ2@HJKq-b>pZ&T{yj04<9Td>oz6!~O3zTj@B~ zY{^L8M_Y$m<0LR-DA`kl+{LTHSV%^hHyVQ;1N%}}IM*d@8T);qJP0Ei7U`iKdlxPh zKL&`Q^L5Zj&V_(VchsFtnA(%h;pZXkOV||xAt77;Hixr3t1(6wMX@albH5u`ds5!Z zwUWGnDmgF4a$(vBF+gxMVX^Os9&4>=ru^J=%Gm=BH8MNmWyEOF?IFEV)t-eCS`KR6 zpUo_tY}jusTd4!pep)hzxPNy>QIzn_dgOH?@!piYeGpWKxQLWJzDk{qK*a25`#ShBG zEt8?e2fXgLk7qF^>*Ti&i607rg5nL^Zy zd``B|z9dNlOgLLUH{uifWIGB>u%Z9-3#)y_WQ;VJrSD%8E_PA$ow_~@+n}fHC`%I` z{nKfuX<;V{eC-d&RVSZ&)%0?_GnGvB!%)+CL22t%*PhJdqG0;t2UkOky&XuoHum>I zrNU&X1XXZ|>BLnn8#?=NgsFeNF(XT0c&6Hx?_suu|cHAl*c=xxN3xg*dFBA~!4^)cAHQ5wvsIUJZvWh`fEmqAL_~;VP@T2d( zaNwgm9`|R!JO}ANxqp#({o3f+3dB4AK*}`p#m^w5eXlDFx4t!jp>aR|8?lxGP8^F? z0%;pKZ=uPpkBqg1EmD(X902AI=C62WOX&ISP@DL?%V+j^C;vemAk=tHbogSWnq6j^ z%BBXbav%Mwv}_>hEW?Vu{xgejCyU zhc!yD@0vHX2!U-=8~(L2i`h|5e9%s|xD~y#J!l-q5A$Ztrw@bbwavFP#eiyIR_Obu z`!S?0*P;p~YM&du4o$56w|On@8aNQEvqycV2ADnGP8tTq|9do|jc@)b{sVoFJNsU! zhKrEPXx8>+>?iUm?4#bw80VUdc`*Ta9YRIS+&L3*pNKau5%ak+N~zMJiF%$=pFE1f zF6;x9D&eO#*{k25kGMD2+YNu}Fc0xM&N}r|wyiquRpmg1^1J(Qk;~Gm9!zg3D{xM` zVlqa!ND?8}d8{-p$`8keA=MGwOoih)mh*{p;Q6a?1L2<^s7GtErERSRPFPt~%ijme zR_BV$oayQ{YpCIY2aTy|>h+eAP^bvDUx+Ml_F-$m*}5Dw2kC0?EvAhF!Ct_~B~RA9 z^9TR{9v`AD1Y$7jqlymfQe&Ih zhIOi|Xa)>~nerTbH!=UI^Cx8pX3}KjSo-SL8Tc{{Sn>`O>k~G-uL3*6-@5?usr0d` zUL>j_ZXWM8WfxmNe~$S~M@mDwO*|jd(sX;g78u`>+mXALQ5>2z4j$M!kAUlPJ32TF zGocKIwby;y4uXt}tBTHOdTlLVS8UVrP5-_=jes|~@*TP^nYf@j?!QQ!b#-{1kH^HP zPqdt45oW2d&hj+It09eGuuyms^E6!19ZELexrGaX{70QXwo~J^$9*p-GhjK75id)uA_H@;-e)Z(;Atw-W@nzCfNhdchqkE(7kt0bKR(q5}SzbFNAJU?1kYV(`!ll^tNkSp#n$ROtWbK}&f z&ud8&oR+Y-vr~49=ExV&CYytP!kI4Za%kHVG53E7w1K;!W z^WO@qwB(W4QQG*?$D{S_e(Q;fcYd{{+BmbX_Aho&@Pq$WtJ9UHUK`anlW&)$W~Xzc z{3P3I%L6;I-;Oiz4_y9dxe#uz@obIZz%U;3z1Y+Dy1R0DWFS6miQZb+6>EL$VR~4e z8$_2J$?`NDs}hDk(m7}aAb%x}G4Tnr|Mf=t!D?@GJMUkqf&vu(i9=xEqSX21JCk2A z%?tJUr6{REjB?YU`#Qg^Y_IyG^vi7$o5OgShsIm9pm465#5NV()Dmj^Z#%AsomwS} zE5G0lr5`^0QDiy~P`Qg^u5~}J#{w9uyn!MYpUH7j+*1PE;Kw> zo{LT!Rah0c)^E@fFRLYKat$<*OvoWNwip8NeVBV6KaWyu457Z8UE-@VkYlY7gR5lN=RzCr-Q{)*z$V(HG4 zR7!QgODWwp4tZCXS`5cSl$GOKmV(qho@*OKC6>4RQd^+JinH4MyhMYR@@$lW-OrN4kh5T{~@#MWcbzVw_`ct#8S!JYVjVrhfOGcY1f8Dzll`o zyHdFSJ+7XSu}j8Yb(TbhJE{^Il*GlR7M->n{j zICEU@Q?#7e7n_TDa&3A@I1D+N46e*g7|03dACI7|DNp=5;D|9`lBJ z5d-iCw~nBe{N}5G^~lu)#*&vf@-r^mtrf6)Y&wd45XZ&?Z3tbek5m`9+oME=vd_Le z!~A)|8)F~*OBoBX%;EZUnGEdU=vV<-em{8*zMaAEjPj>>SN(Fi?Q-#SP3xc+)4d7_ zm>J5Cc1dUfI4}jACElJ*Xx@jmvyR~m2kTCXMZZlTU-$%odGr(=jr~ zGO@|Ds!4*2VJ5`ru`Ogt8(lGty{TJ4D0tJ&t9S7B>)_N%aEIQ=C|;-7-o;4H)8LxR zK$etGXNKdQsDtK>bEgGfpdK7WS!sRB?@*XZzWi=vxT$GXR>W|g-}!?ESVP{W!nZR@ zG)AueiM;=8;bEOm4)Hjb$`>Q5F@YE2D&Yb7*70CL-gde8df5zS*4SH1fL6=aP(y3r z@B}P1F4p>zeB(Y=RpT@>3?p!ulB7whC@f$9rh}gFHt+rI(+$J)K&=SeI*EhDS&xS2 zTFwEO{ij^ub%h)tayF+V`Ah+TU@xr~pP&W6h3{nlkzgS0K3GnNg>LVpS6nKl;x7uMG~o2MrcIs(7Ua=s4v5$K#OH^ zMRITT`w6szIoMbAk0XV_{1L7IY_B?lcUb^rnV{9sYSo__H()d-d8oDje8GSy z?WfBlFlbct9ScQ(UlDu%;;+WFs_j)TmiT9HBYS**&~oT2PQobpkBXX_p#vREz{WZS zNyk<*a|r+V7T@RL&c~E^oZ!0*ai{9+ELx-F-;(!`9CFe91`qKLAY|B<93v2AfBP%& z@pj3tNNAr~nQrOHq4_JkzruT9lY0PkOunVBd*=&B&e5`g1kJ0x*X}AoNkC9?E)3rs zD2x_--;WV1IGD?4pY9`J81`oErL5k=m%UvWk^gBca=q+8J%KoF{`UQRH)=PvmU$yp z3WmB%Z)C_QuKJiC~p*7bRj$zqOTvoj@%~6 zno|6+_Z|o;twzKVkMQC5i>;qvxU5ns7QVCPX)y%pSqhz(yIWEYTZmA{z7=mooF!X@ z$lfk|w9Sq0-~(^Rf#Dwd&sUDO1DADuW@S#SyS%goe{V9m3b2KkH2G-aUbcm524%tg zm6>4FX0;~Nu)07w@VrNe&E=f{TO-@cqIIZ(#gj8PEB%ChUMBJ%o}F8intrN39r(tS zr|2LJtrxK^TtgM9WmE}+@hhu-|%L1R}8)~DgWN;t?5W` zJ}psyLh2IF!BUR(x~Le$kTOouNdUHc5r8Ecq8c=`RYI&pT^EkYcYBrFeBvhKInx#& zRb0LJ2lIZE0Q`JxWyMvC%hzB6UsNXn=D8`Jr-zdMEqd)3 z%cyBrHq!Ic_BGs1i~O+mT*ju8K%Q1b5h3FY`IMCIMD2_RjyP;iU5)V$#%z?hMRyX( zmtk}uBbsCNtnNC91Kr#9+phY8VxmWUCODfs+9^2IzNy_<}MGHc~GbH%fRo z74-j*AaT@HRaJ9}3!i6PiXkyVbH~)~CQ3B@07z`?Tw_(wX;b*mfw{b3R-6 z@oA8+=u-gP1+L2LNP+w$G-eEhiaL_F(VTDh@!|= zpxih`qmESnlXYS_bbf~cqWHWs3cuWpG(-$Tz1r2b&l0-T%)88X$ZZh5 zOcmEWpV+s4p-BY{x_oS$FZYu+kuH$R!djg~;ZOFHPW!Z%kfftR*|LC1(-_(K2sRt-9vwom5}7^^B*d zPGPIwuN_8LFdjTIJ}L&IprBIkx9^@D**jurH+~IG7u5P#pp|#3$^CQHcK=Mw>KR3V z`D*P%-HhKu*f;l*@<&+=Xszu!APtPT6Gq~Jads)%k3FLV+LuyODJz5JZ2P@h(Jlj7 zn(?~BrN&{Y)dWTrKWGM~ZX(At=bm>2850{X?v%0L<-z$SkK=1&)gV~MSnO{M|L*3< z*ZjP^*8zzMrtG7u*KSMz;sio4U=f+m_%!sSHi|r+X@qGVAc=+bsie{RQ{OQ_(A6NXrv!9!vQqiq&FKoomPbAX z-aB-ER{@+EjS5Kjm!TzU@!*4G-K@5P3&qgH|5vGsfYJ*UYhm*3-7C14;EpS$zA$g4 zF}a!hJMdp)J!?v4)Ey$35o6~l_S%+@V=6PizErE z|8e8}7~EmJRd&E}pnh&(KdVTnS``JZh6K&0gi~I>eu%h;efEl$lVm%sRuRRR+S>y7 z&Cf~VZkO2ZERefDx@;j=tBchR1XNTkQIJ&Z`eRRhBx>$z1Hc@th6+uwJMG5pos%EX zf1bP_qQR)s>x{yB>P?TeUPWEC$5s?n^k#8?z$QLH=5C}s;K(igdOaZhanoK8L7+9h zgU=J?x6T0%lOuK&By*@7i^Df?*$JrG-46*}Ewkt^Kprlk}My z&h{QJ4?3=DIZl3@p>#4Lmt=#pdid)%vH$#s?_9P`$|Vlqa=?Q5MEX}FlcHhiS=6<4 zi5CH*!OiR{ZFxErb=~iwAg)dK{yKBHgIn$!hmmaeYfS#eJ-HIAB9CsNSNzbeHKHuZ zxWeC7wh-!-J#K8{soojOw9B>c#VSW5q;({zBx6Oj$YKM1c(c&HLwM-34HLP;aZ;Ub zM@)}#UNiYj1&*!a&o7(~ZbQP@4>%c;qFydCm6)9ohZetdp~a$BO#hX~v0%F!ln_iw zU>r;-T&Y+Uoy4@k==eh`?qs;0u$6EO`*T=#>>-UkA%JFNe{O4TlUAOga}vE_#--Pq z!hCq@jZ4U#vIKy@bX~yF1B9xT3{pMKB>Q84PC$GflZed`UZ%kxtBt-k7BKY&W}){= zIJM-{Klpg`P%(C~1#gDDZ;l%p(HGcbDhmMjxGbTl4JPPltdDMR*y#hr zy(`i18InDSRSizdMx!H4!I#Co*}PNjwWHp2MI>R7(b3WEv2kQ)VZu9K+n!ZbeeQ8I1o)eW^WpRoc;n<_L) zf5d?kG&~nJ;%(%QSj0$v57ou7!_k+YQnug8g|>?ObqdwxzT0e}p5GhyNjDWe;H{iO zBY)%Dtj9m8l>>FD1Ud;LY%e+oTQ&}Q|7syQkI?aQoH-{!i_T)(K`du*g3srB(z^4Tzg z{DsgXb%c4toTW!Qh9mQhH%bRE9MmfS+Rya3NzEYCP7y>f8iq0O@-%UUHV-Su(L7TF zVAzFCj`pVeCl}GB0-e{hg%_aD|p>tVY4ITE!Th^xh*#{iO~ag$Ytu=+O= zucuAdurh1Q*mhOX!4;-ghoN43j9V#0%)#{-&)&|gZfs=BVx8<^!GKt1shh5-kSGjGr!u+3&oxotoeBQNgo7L?o;?EGt`JI*;lr00;9bR^AJbR+H+joy3j-TV08^{#dA7hkcS-#KTWv-jDj;Vbgh z533)$pT&_f#NMI?2ZabE{yNorORDFGIPhpV2>tcPoEs1~UxB!B`af~wy%BSjo?oV= z_jA(4t$YXAIWGkH^}&*yl%|Sy!b;1>1G7=bzs5D=sSa>@Qk`| zpe!15DVX6hb{F|&uQ^w`w!iPUKpM;o*>nQMU#}O1Q@rrqjtpBnD9y-5e?ZX1=o(Z zaBdKJ39z(H%CCV>wJeXTm;E(+d{*!To%ict44ztF#u&xky@kzG?ldeH-E=oPJs%Z6 z0a35rTNl=4J#%?f0!GI%*6Pg}l|0cSDl37Mu<~Mijda(K3@l|YtO)b4on@S{qqSdL zu?s%<>job8b03e}(AZw)_2iW9q;xU(j#wPauRVXL6YTVk0Mlkeh{ep4ZPaEIYFh!G zT*gDvWdZb$+g#5Y4P!+y!>+C|8V#8oG;Ai6xx*BAmwGEUje945)?TW-*Q)`-g&QhI z?uYmW_;C_8T_TXC(-1t^bs;(&YHK05-X`kDdbCft zhX?#qbDYmMp+2l+z7qACxOC(oD`F#b)Q^Nu4=BJDp38tGoMze^R2|(s zTMs0^f!L1Y^@>+Pef_NGb(0s}6epd{Ysi`0(8Ydzy}>ogqRF@(|>zX*u2MPCkClN0k?z0XTI-^fqI8|7l^n}(N_kNd$#j8!|P5f}QUWC{B@ z&dPSIgogS7SO^9Wx?^URS#6~6h7xd{eGe(()Hdv@EWREd4)IL)*S~M=2k!NBBpl<( zt}4rzk@u`EZ@=>WCKT>PA}sh<-fjGx3D+ce{|??El*xM*%pPaK(F2UX#2kixemXS* z+c;Q%8YYHysFP;iX>EqE(VsX{f8nEl>c4WF8r!?K&|ZJ>5dAZQjfYHRsh+$ygJh#5 z{pNB7xGak|i`y|mSP}roe%+^l!A`q{y_;j`6W4(Jk)p59rw5h> z7Myk_cu)RRi_vsn4lsgVSsTj-zpwx~nC>tb(9>Goi%dVJ7ypQZ0Ox=>aPUpIA!8|b zI99-ryw6#b@ck)uLscS^U{_ibCOc|J@{qq-Vy!q^U=Q$Yz$JbBu;I(ztsdouJ;V7J zvZQ(c7gGAqkbYw`Y=NbfjLro<1GFKm~eeLfv*@dr*4_!&a}3|X|{T2*kJe5@R8yvGse zpc6izZiX+J5kU2ax1(@6fGUi-;+vRLAvJFDL=_mF zhg6ItEd%kg92Vq}4xF`?=U`AR5wWJIfXP9QLFGCI@{rA;s>CAZ(d)QBF7W6pJp5UTBNx zAh-Mkh2d&@w(-PBS>;bP^wbhbNq8P1PcxkT05I&|Lk^$Ui0JDYobkN>41%d{@LZ8U zR#`CIYf1aBWe34Ix_At07Qvvnj~w6BlS-k=>}-rgCUO1n_h#zdF_II9?m((QO?fn;X4uz}R^|myvRbxH9b({1@vHcvGjyiaavvhmllK_;q zTSesjMYrJ0p)D`rr@lY42i49ill?#BL~E=!c`EzsgYPGQ^am#RR@p;;Na*2T${D6< z%#rp}z~6NiO@b_{vUrXEh#L|0r0+}cV|Y+tPDQ$tbnhbsf`xjHrmTs}DJrp#iENj` z9;jPB$XF1x;!vEP4AKm-w00sJtToDYV{liRh}uD%ZOmF*t8W9YrlN z-?ZLs0DdKx1dOxzCy5=QP5!`&(BQzg|2^mG%u_@s*+AmpHkFh6HOEZxUKm&I>EqBp z;jHcgDLreyN)5&zN?QFcH(oh_6j{0e=HQ*zy5rBo+1~ioH*CU3N)>9hq&TQtl3nBd z`|mYhLypWwJh*))AJ>?Cq1& z{P|F|3oY&Zfplz50V}LH6gmXT|E5}@vfy8vNgab@as+z!`23bUsX70BmkCtBy*XAF zl|p{^zOBPw1=$GB_?MnYzCdH!*>W1Gcb}jh@0krBUPhrna9w$2k>~(Wfrfkr3izqM zeE>>y*X(t#g!jw-5xkj&OL%roU3G8p=rpcj+aqe42WWIeuIsw~GC6k#h%jcASS+&b zf71HnwvnA)o$PmIf5`~{#|_uMi|3GSteRS;O=DBj|E;rXAOwI>z|Lz5T$>;-CM2O3 zGKxIuYL<_~d}^a&NVOVG({~Zlbu|j;aF;JOwPhatlv&KmRMCAVhpT7k5FDu5W%FJh z|J{LSC;KYOMZ70oA|58#$u5#FP%WnAgYIIRJ*VXX17q@F+lO@oSYf*Nl7L*jwPt-TZ(c390W(+Y>wi8sfyi7J!fLag#IQ_ zPWL%ZyeZQ2(d-Mm6zIy4>N4KXB|R%DC9+_V+ZUay6?d3L-l&`w4fs9;mxL;QcM!KB z{rYw?<+sxRhB!$nHCgs;v_y)Tq6IpWsYeB6Tzn_UfrK_K%hiRkh#;Tu#~n+I80|$? zwY1pX{KfX#9Kl0cz%WHk;L%cVQs-voxh9YY;joIAmb1p04 zw9#yFl3y+X!7a8f=9Vg2D^%Q>2)GIe-#Q;f6ilBsYwHx%&NqW~#{4ENLJyjeV~SO~tE`$@N+ZH$lIBd&9MJ9c(PcQnR$*z0*z6Vca+SPwqkmeLZtzVzbRG z1TRwYR*tRNK7^Ei`Bmx^;Pztwkmu@6WPK5?dw)oj9UR)22E|5+74p?|*bK_N`T-P- za33DOeW=}XwBI9Q9$h1;*rU+v!fuIs;U5bP4tY*MX!xiDaUP2DywKKVOmyjfAID^P zqv%VSS-U$01aV6Sl|-!zG(}r#za3NKq>M&Z87R`pD9%I_r3)L)=cY9AJ;o)Z4q zD+DX`OQ_FZyh z{0sky;YTKT5+1$dc(#^(Tg~gvwZ>Kqf*2xQVl^2$3w7E1xH(Gr=S|M%xl;R)m!<2^ z1xxnqLB9Wd?K`-*Ht?g}Ka`HmOKVYos4R*Bh1qh&J`NNv!b)e4T*(Gg`3) zZph9D1q-kEz1usDzWc>X_jga$g*JWOzFwIY%h31&dMwG#-~rZhh76JZt)6Vb+Y$># zp^$~)^*_B0Fh$`!UsJ>+mj{PXY4?&ExcKPc)r(qkq-0A(_)qQXm)E;C?2eX0omDS? z!qP7k6Z}@yv-^EAzS46fj^Jt)ZXjas-wB4Q!BdqP9UqQaWbS(3^+LEB2dsHLS4#O) ztRg4;U!nyjDPf5GF<%Ar_8$>EBq2>dPezb*XFJxfX991LXuC_T#H<3Z3VBHrq)}v; z)>3DWRplCR7{()ukB@(e`OF%#|B-yvP3>ii1iIo46a*3f?!NsZ)owrOx7#cOl7+}B zu3jFGzs}s4iR(;lTJGxH+`+#LtO7FMSAGnJVP)x>Xvr{EeKYsYP9O-~YY@b8PnMRR zYSXJdoVi^E!WoN;n9J!gf9kXX2=%uCU)=o*y*4Sb)35Y$A1935;kq`Mt5ZIp)yFeV@3*ZzwHWEMnqhlg`S^~95K^Xoaz^cvJ?ZLGLQDo+F`T5 zdPK6<_k!VCAk7vI+&THkeyi08IV$qVql(p8BHGzLue6id2A95o^!@ySGof`MLn}gJ zt@|`GSyHlqrh2}M7=tD3z$)@4c6Y2R+;ARIR?}UMOvMEc=YtBmqf3TtN{P`qr=O|W zZd4mnoXJvYK{S%ux)XjSpW!U}TH-R{7K_y(?i08Y+m?-73{}kNt|V;jb?h*;B!h(^ zbzwj2Q8-1V>Tl)VZ7^r|1 zIgD?G8fMgV#0zPQeC+E&P_rP)bki6*%>&J06YYW*Amb5fwR@k64g(n!6Qa%I-gNXTrQ;Am4i}PhS!`-RGR`uT9C{ zCEm1@^);`kYjjg=>&i)WcEHdzAonH>)_BQl=)3d7GvuUp!`hqu4=0gwj83+BGsEm- zlkD?>Um%1oF>SX!(9?bDom%Ewzs;LF(2`ftpE5TWA|;No{EHAQ|M6whz~_xOzIIEr zPWaihj-`JgK@a^5_^-6e|j`Wc(L%wzFG8I1UDqN*1O#s`6Z zXhF3YfsKaidI-Fgg_#&+GACgx%y>ugyRDhEd`fCwf;=92mfDrXPvYj%xp(GKFbd{!tByOmjbBP%uFXjT$nY1BK-bZNnj`}{0?s)) zm0D4Z$Io4GldGiBoOrF)cn{O+!<_RK8l_OJE2!z+fg(B>ptR}rs2ys0 zh4+)Ov98?m+H-S()lS%Ywk;!he5& zzXIfrWTZ2jGU*+>0f8F;wfr#|05z#H@%5(D+;ru(i_TR$qK{!wZ_SX+pK6-45&sK*NNxxT8a3amqQEj&%|&s`3XrxJ2^`jm9Ic6mGZw^DfwbWZUP&v zI@ck+qj+S_L^Pg;vPtB&mv(yS7hh5_8kT`;!g38JqYf0_OPLG#C0stohNUHr&~$b$ zhV&w_yKKSC@w#>C(!v}L*qZ|?{i@O5dyPvq(NQ-cd8SJKJ*2SwB&$M!k@BU_Ty2N? z(G?H8G=>lL`vr&x9M9yKf9JLl?7L}Rs}E!`f(qii>3^*PqrP86@=_|dBV@Abssp4p z-Mrr9?@n_;Bq~xxZca2#IV4@Zy@_%cW!$cI_l-aSiUNWTE&Gk#86^r8u8(dP$RKU> zqb#5|$G>CJcCGEgsa?$%Co?>*Fua7zC6FKRE68vzphN>zP~_~ljpe;^)_H@6=w!$+ zfMm`}>nfZ*i?-}7%wvDwl-0_uz9DM#cv`9Bp$}iUjr;i%sJL z;2^k%jLAO3baEExYL^s4WCHfBI~-7CT$PSErd4}tW*B^*vVrw^=YpKXc|Q)(n$cj| z$jMyfmK;v+n%BJP02sFVu93At8H+UqXTKq8*17=lqH#mLlUlkUy)$8g>KyKX2`c-ufG^CawhP~=?SE@WmA3GB*Jvq+tEi}xZ!I3lXt42pSXzRr+x7|G>{q7L_meHc ze%N5$%1ZB6?zo6m@3dM5NNIoOy%+6g3U*x;;R^_UGVtnZMmCGj54SURwfxX73Gj!F zidx8o-9`ORP+kUc(y!32Gc}yWCxIXY399R%Dxju$`0lKa?+)GC;TRjp3Q~{@*LB z0eVou@g&`4EI>TJRP|DJWWObz8>unaxTGqMX4+xqBqIDX$C28;&q22! zqrsf`iM#zrxSgtN1Z&cJ&4xdNlKvqwco*BVHcx&&D);_#l~BgQ1m1A?CtI!N*#Hih zfNp;}Fsn|qL(<5a3vJc4IV^n@rE&L!8?W{ZtI>|`~xPYWI5!nJL zvsQ6ewh=kQq*$eqMK3!t!)sb2gj=94$^OUK>N3G(GWXxA`s~kB&3~VP1&ObH{robu z&bnSXA)j1Wq!)&+<`ZP(W0#mrC`YtG2!x)8q~gMzYmZE3%$Lo`Z`*He&8a~J+hYPx z>;V;phK-Gv2#(3{C%KpeO=5(MB#JU6&E$u(dZI3FQxf5f?rO-#un>{EfF`_*D`vT? z$N|yLk_hRgHpsywuSYHJ7uWn1*QQMP$)TdmtiiKE1Su?O9Pf~VaSXAfUOWfG{ZK?D z`C20tte%D#ZFAElov1vsud8sD8iF|TLCeQS9I64q+SE#lb@v~u5Ptf4>XiH6O9Av8 z!&e-l>Rg#Gab7p;1!~Xc2uYW=gY&{GaG?q>w!c4GF`O?IplN5XUP*#n zs^knPoDrn6`=q-41@|3VyXXBkIYgI0RTsm>zOKg0rq2Qbs*HC*My82QXZse#v*yLT z?gv*pelxijAeP!o{G+Gc0N7!pf_tmV2a9Zx5f?b(p{pFw&Ea(yq3Uv zbJKn1DwnCIsRxD1d?t9tv%K&$5NjNp9orx0Q+Z)9Zs`DRc1m_pJ55P|MBf*5PB`SU ziOJo@|2-*Eo9f|iz3YUme6m~VG|t_Hq?#|#ft(%(HpRv!(Lhq5PVmb~d&-k;9##p< zs}_#}7{7j*dOF?cR)rw45#KAN+sLIT3FXRv$tS}?>99nNCpL&n z(HDwzMSU)M+CF?$GyN`LytQWidkTz+vo;1@Nc*owKVhJ()W}(wD}P-<(s4a$b9wk) zX(Oes1%Gjo4OxLQW<8qY2jles;d4_#Vi+1-QW7fi4(|EtW$Mmx$Tl29WJIFS50s>d z%3r4S&q!~X#u;uCypEs=;kYX#4^oz60YDNvwT*mubP95DcEIj;E!hW_=W3g!>nX!? zz|HFGBo|NwYGhy_#!WF;M!Aw`(04z*Txb`V8ZB6x%IMG6g49gyU!d{9 z`D?t~2>ln8ZXf+98V5m${_8oGFoKdpe&0W5pD!~QdjR5PNnmA9vl^qA#nVI4}4W@ofrgnz6)vqgULgA&Ny@Aw>9;S%b zL}-<=Q?8s8=JFftnL$5|WwyX=*wnw6rBNDykklJlGaKXm|FWJdNR`Up4 z#^CYctZlxGxM=J&)@J;qB=q!kVJ?~Jd|cY+tE{Iy|2-HIli;$iI}B`Pg`6vl7cM4* zGLng52ACxV3|0JHj_cW(@TR71G#S zkYml^UI&r0rlniVv#X)4r}Jt??vSUS@$M*?9E&u?fEitW^nH+{wpN!I>b?cC;HUe6 za=T(d%A}YCfW6{1kx^EcL(U_>V515M52)+R*TE;`D~-%Id<^!KI0M4`<>CzpC3Ncl zQihS`U36Q=hI$8LouTG2pvI3ZS2q`Cj@A9GCmQ6pySe%A7o)+vFFPY^%NmpT#pd&w zUJa%_NzYxVM?|pse)5;by544BkT`p?w+D1`xq0i^1$7BJTcnq%G z8-$0+C+WJ29wb5#+Kh7utdiCT6D6y^o7WL}ElI_@%a}phBr%YlGriz?8r#{Gv@ZE) ziA+J4E5X0o^UzMZLtEBM&k+wXe77n%8*aofqM+<%eCk}e(rk=nGXW;GF5mcY*zRqE zgw^=Ip(ul77=B%^r}UH=e!+nUL#XK3is=(SV)7XzngJt8<}GvQl*d!nT?4?g$fR1@ zv#3tRdC*j2eRF-q_(3Y@pv=Lvm*&bc$jx5O}tkBEK&TPZDh z)sWG;Ae=$DlhM&SPGm7X*YbJl)(IW(_F16|&lXIg8uH?a!wY)`V&nRrL0xHCC7boq z8U(;7oRYhEoZuyPlLsDZ&>?k#dk^k|3ij~Pr|<`wyWPKZ7)Q(P0ewyRyKw;|Xe8;! z>2|eSa>9&Yy6FO@IsP!Y*@dfo&kKG-!E+JF^efXAm=KZT${DhO=kFVVVIt5FyZRJvnI@ zGSP^Nr{2;eRsBc@JdcVAo0R=bwgZX&YG_K9W^5M*SW2se*FSt(z`B~B;YgFj7k=+B z0-9=2&VBv@6hNS4EsNbA#YMPi>Wc0H|2G zjoAx8qkL^hU!TTI36Q!TNh!@{YPaY~uSrD22dQ|Hi=~z8jJQ9L;xeQb1hWn)UM=4r zP=_B(#)eV5ew5jISD`^(<+?3rcbx6Fy-ytJA){(Q<027B${yXE_dTTI@<8Dk);y_G zD0cgFZZyyeC09EOhWR^Hdylo`GekCDH90ePn>$YeZvYE|(c+rqg7HCJ9vp#38j_lpcjExVG4m#2;udwN9mwQ*E4=+u3D_YljFXlC z@)XS=Iq(mOxzgP7C?R*#j4+|E%#Pxk=4NmdK#eO}xe<@ozXS9)75twxLfsY*{Ur_6 zrmbwE2FNF6O|;ct8Wzg0IUK7RoGmpvRo6$#Z9{(;-OH>YRlgd_&R8CExe8_tA*_HF zU4}*}v$<;glUn;VjtCQTr4)k|>^i_whxwb>SYL{rAYJtpf( zkNm`uF$W7HR9juBiT2L>JjFPx4nzV|N!iEH4|B;SJt-f4WF_N)sPf;y!Y}Ol zb94{=hq;UFX`t;mPV{NNgaX^3o%YPv+ZjDLCITEPDmo%pK!iI2jGX29lFYa_Fr-@m zhIBc_Di_@M;m7KNyVwgOW5U5R>vTJZSO>(PUC z!q8*bF_akg#W#G+?C$Bqo2_#xm%2U?uPB3=ofw6m+kx92hmPqO`wR~%>XsOQPM@EB zi?@V{f|wPwAVif9#qmf=Vw}P;T7Eb2bPpr}GK-$edPYP8%%a>$f}9s5CCs`V83gqc&#$&VQ(rAp z&MOPvqJpWT37<7!_p#oKsYI`^@EmW;NIoHDz>Xo)HN8!!F+@-E5D#dHin6AyiILhn zFX=`f0EkcX`6@ls!gPr}>i({kQx)#Zj8iB|<8Sk=3FBD}9BWw-{bQhe6!qJ; zx-tK%leqWlv)#5#H7tsbcC_qKH7yjqNw>t%Q<`zU#Ot>v#jP584bTbT@f>_XWfhol&jafZIUmSY^o@opsBz7!37J_8+t%L+NOM%exdwVTJ`n7^ zMEO6#Y4?^cg7*u@cXV6cKE@zrd~Q0}w@DrJ#w%b{I??kjYh*ES~j!AehR?Tq*n-o2<{(D%;zpYz_iwu3qmtibQ0 zFi5L=>B;au;s6jHo_GN{;PTbw+R|+fsI>Ef{exp{l#!^Rc=YBp19&P|0a!^|%ijj- zfT|BZP^}F!ogavsXb+iRHBn~&&!>De1%^jgkjrXmY;jJJbUoDlD!?B?DZ>@6BR4$G z%}8$e$O~>Y8i+(4V6+-^DOM|rPLhfqG0WE?sujc+;ssW?#V*?IdKGU5EKbB&!iuiJ zFtabHRO81hN@(HyX+slpJoQn5ULgV(r4h0pPYJ2uY20(Y1VVjXK+$HOs*lU2d=Ew| zsdtFea`ZUBpcu$*)7+9WE8omBtPDD{xL~axN*!XtzPTz5rws@MvoL( z`|Tn_t=joN!2V7pU3{xaAXd@#Ia5E5RUj3YjD!8BSL7aU!1`%10b>^ATV3AguClMS zPD`Q14~xL$rLWdnNAH!I$^G&NN?j5cCKSKLNJ`>rO|_nq2tKxFpX?m*>s_y>L;NwL zpLia=U?EVp{oi?0BY8R;iWdBRL1&l1gjCFgjlz$&TX<&q5ik``{O#ZNEQ}X@^Yx$^ zlPkyQ(OO$laeOwSZ+|A9a2>y|Kb!J`?5-mHIdPcT&YcZG5WHo6(>!;g3*kK3-fjIp zliDb&8`hd`cRL%X*{Hl2z{3#Ln}E0lD(y@9{N1o`K-Dr;n~$Fd$f2hxw8x zkEv-}e!sUbI|%kT2dP3wKJ?XKg%BOHmk|3&wzN3^|9FuoL@5mON6%652G1*~;Zx@1){l)Un#x)E$V3!g!! z)u~?st`Y8SoK!MFMpmJNf_ctXZ3AmzybZgv^kt-Kq)WUjfw<@lng$T=d?wLTva?#i z2&Y#cJ(|}zMVl7}ymv+unR;-JH(<)qV0)0-j3qc{Dc2Vzwy*pIZX3g|E-vu47skv$RdaNyby)g(}lwT zV0Hd#?G-xh_PDWq3&sREvp*~{_?G^gb-QLC_~4%TK$mB`e!uA`Bbd?zTC~F2*Wz0G z8Kq2n)^=ndPVLXyZzL}YMwl8>_`g;rQ}EaCY*Q%OR`3c&baopb6d}_oQH@~E?1%R* z&A(01A+?Vi6eW)_8ZWa^!}8K_xQ1~}zzZh>SWoqO$WHCRL*ClI^E5hRUf8AQ&ppKv z7zIL487P*N)a8fL(8E3=;B_^Wi7YZ;TV9I;LiaLp$u-(nvzVtseE{Cpcsw-~mrYKl z%5(+9Kx>1V_YZG3$3$yK^>2IKtb+d(aj21-jpj4f2G!P%zc1K$c=E*0S08?MrN#}FnFyZv{=^-nmWU6DSB9I1^A>p374FpEbN!7SAf% zyW@@ttHw8os`s@h?FFfT|6@>d>y%e1UihnIN@rm`T9!7XF(KzCjSo(O7lUD*#tzMpqMnEc&NCThgZ|!8t zhfDwjG>$r;yf{}xzIG6OY@WAyXc8XWEL-UW%_99 z=>{C%g0*#zG{*S+nfR4i>ZiK>w;m_Q@~=%7aZL;#{vh^SKl$9WM6OT<2mkw27LS?+ zNg_7J#2Ipe-c#DwdU3n~V;u~kCP%I0s_vDph=;CpYKKyEp%Ne%DF!I+Y?4VF1W)Mp zE!iPU=Dob|!G(M>8Lo>a5VKU{5kp{B_)GOP2>yJ2sJEIVz(`EF$$!D4J$iH9l3^OP zM5Cp*%L2R=C_??ldr$|k5uVA3MH0;j?trXbp0EDExtW%yBc>fEL#oiLeK%wY1Wh?E ze5#{46qI4aMXmzJSK2Vy*_CCQSZ?O24bX)>mnb$&uo5>w&p z%I>F|19e?V6mtY4#^cbei*SejQ<{05F8M{`bNVham~QByEtfLW#`D!ZRS??Ke2mA# zr?zzYQF!to{NKmAH_5(V8oh2erKdDUCs5@qQ^wY597psKd))gZzll2ZG5N9Vs^^q59W_!&iN$wF zA|xf1I+_#uazh{^FFVr2RU9sZu=~Dh7JLpkw}^h+{FZPLm{F@U&cS||v5KD(2tST} zwWnCp@MF}a(ktCtO}vE+XD_O-ujM&X&UxcY9Rn4^M87b0CGPouX+_=5dobPax*F;i zP43tix>A1dP%gq9W7$^(KC{cADVZ0|`npTG!E~Ck$a{tBEo%n~brjE&)QcrW&`3lb zB=8vSZwZioXof9vdB4V=xcHcU9TJRt82I~}iQ47Xi&S8RyzA%Gkx_2xau)FUM-14F zvXn(zXAEyGD7N!=RI-O-GQvH+xma=#xS4(_Pg%9i{&7g#YuFL=gwqLteYt!9-;=`w z0P^Sx74{|CSj?*AtPJMPs4n>B(%Wf$#It*qo$lTP|Nr}f;TiHts|8VI*&3CPWE_8P zhsrH3*^3uk<&1eL*blw-Qub0Wv>IM2L#9!JH3YyKAm#APuEl*GuQx=p>|>gs`XI^sz)wx+*^`w~5UdAG znuyV42inlPgVKE5!hLpS!XdHSA1$nPcc}!aN7D9)E>)0}VfbngULKyBvonuktq;*! zs+CdI9CGUGQw=^7`s_e%0l2}$SVkxOKB7_64w``gODnlfEcWAKHwTVOVgWZ&xad~n zpi1rX?E^i(a)%~-Zz1$<7lk>@Vue|it7MiK&Vtlw$u#R*Qix^T;`yIXyz`a(G)LlW zk=!_v>t!hP;(5dKETI@(U4t)2MJ#O{h`9`sqE?MsaxEd=a>2{9S17#8I za06KG1@f03N5eYxY>Z0^`+d{Dj8Ekz+E8pG?~(9UI$UG!NIO5^8*%4;9QE4zlN2jt zr=st?aO33BQpqITihcZf&=|XeA}~$2>`GHuv$M$ridUUhUqR_;jd6eD=qS; zPuxdDV;)7$vWI9qEwR^r0n_o~<8He$mIne^E8_T<0@`6R1`$d9qI8qABtZ5Fm_r&; zf(0X%SPLGaTV7Xm!L#p1IqLf8!sQ`_K?;ObF&8}jJET9+(5xUWk&1c17r+O*CH(P^ zg~~OQr}F08`yQ?@FsGZ9K~_ck!KK3PuR3B>0>c}-ouoKLr$W*|IPnH8o_CHYS9ey$&W0FN)+-dx1`w_Sz9sCC2nS_lm6;+DLr{ zvV2a6_W~8X^PJ6i{eGI+6#mX9nXZhk5qr<_+ih4L0Qe0?oHZzaY~@De8Jp9D`b|RK zQi|l2B!gg!kZo=7HCSVQ;h8z^35dAyq<-g21{4@YH#Aiag2&DY}>1bQ^+sk%6ysqaU{#B)hL7px7deuZ=y7do9 z>H0?WYB||{61Z#Y;rEi-J{AP>euv=O7yVMA4>bRjeB^yh;w?L;?uqXg`QfF&3OJ6V zv0S|hu?ZfTHSevsy%OJCqF^xqyzr~j;F$lV&$-LQ$>x>m4Iot9xH=Ddis(@D6bbI%cxLWlgbRLMabeHz70ae3PpxL@?P}tOkxPz_k zks;WD&EpPkg(t#qHaPbsh2 zk!SE{zLZft(@H_SK;Lec<2wv~> z!L5#o>AlK%m9TAin@9AE!n=NMx1XxE_F8%k2bz=kpVDF+*!54Y88WD1$6#@^7_-L9 z^C2ai6eUIDxcT3nIz4^zLBQjKbkFZf+n9gECTy&@^wtu&KtVA=QpT&jD#Lv-x2U## zok5);U)hxIu#<4_ygV(I7}F-<PJNQ0B&MSRYEx%J6V29E`r z{oQYYV*8+^Ez37$^-ek9<@cdXm?YK?rr0R3f7#+Uqt9fz)Y|a$cM@Zz2fk+tU_Ys^ z!y>1Sn@jT8O+5gzXhIlVD`ntbIzyq*0IYp~zr%s6yB47@PLTi|v!zB|g!)QLCcCmOK1~82 zK;!NMUni(y=5{z0xMkpsm`z^3SP_NBEhWDhYzS$jJ1*h7{PPGpGXGk%=9<#JOOC-4#x_lx?mhFYo_w2u1U<7nM-NP_%ZmvpbSz$dTH!RWuuSWtltqohJ8I2zS|t&N1kAzayf zko42a28S_=^lLf0?M=TqJ_!2Da8z>RS0n6``$%nyHzWmkY(e&z91vgHzlHEh*}dmh z4UqxMQSB~{vo{x0H;cZ9V~yWJUuCBX;UMls1#=8VWrN-z^uLd_%P?bK9G5_>z(db; zM|p9Sw;_nRGSagMS4s>aQLhy8t^7C#HMXxn2XJ(p?@ku<5w?8yZmKhuUCc_rD&cb< zNVSvL=1(m@mHv?0p@$g2-37#u?A5#6i!*0%2Y?2B8b#| z-5-#3+}mz4#@zVg%q#An}c3Bjp>2fYdiG-H&vN>wHQCjnJ#uUe|c@G|}u;M|69A zECq$7E~-4=CiW6zpB#`yjme*4gu6Drsk#ah_~BEyKCkN9YntG7+-|pg>9^+8h6Bk& z?zal=WmgG`hLxf&g3v$i24&awJ@2~cxiZarh$%xwR0jFDI9HP!**=O%D6P`g3r883ZHt z4{k2bg45yl$Z%tyBa|D?n;Cf~LXt)4Y9aRouQ-%v8wsj@Q1v>N;&U00&1wMBor315 zM|8)%x=ldtW#6_cF!wb~KAso}siXtLPCHI%6GPhWIt6dI&!$z2w8@xAgpua`vIe6$ zB)q5w+V4fROyG@kV5qhz7@H%d4#_5#lD&Zlc_i2DdK(5=3~u(y+y+~gw*lc^kxmb_ zOVDh*vMp#B-`^Hg{!yVRlA5M_d1?<6S?G;jsc%UCD)Dv0G0XbFQc65O0AsiX)(mdZ zAKNQ&QbZ}X(I~+c^^}GJ`thk_M4-^w zI>^@-MHBZxLgr=zqa65%>*%3uvWQ0lwHw*!w(uNLZaQGm#Zzt<-4LLaDbELRz;miL zuOm50cev2W8nI-5UneCUjp7%Agqrm@Lxh`a6_$hZIxlcdzlI%u(+>Ts7(y%xW&;e0 zn*f5flRI<6*}1Hwyb!QpsN3K{eUB!lZ(EyXKP-h$(G#sO$-4SSD+B>5&eM-=SrB{v z^GWMN;*OaIL)%<0%z>hoJt^wDH`ugMaaNix3AEUJ)h#K_de)o%1W^O3I_7dbd*dsF zv=6~KWHiso62Vdj#OMbA74Hii^tntWP)H{KV;=283vOY`{afgT#`rI?2HQzsW*QvX zYV8s`6o-!hooi+*$FB3D>a549G?qQfM*J2^BhQ@PjmMTNHcG_*-d~?y*SnhSY#!U= zn`9@atHl1=W9fZU@y!6`nsVmfxt!iUvg7Y+-?(wG@LXmhT`+-sBxbH-T5%GV%BCW# zEN)dNq1Re(0sjEm)X(XeXsAaN8g^9Jw5_*2xPjXaf8Q!H42X7&ye+2S4~-S+12PAe z=2TTJ%Gi1ioT_pi)`(G6!BF{fLT^X49~>X1QuXW_@nFX7Jvp@+}q33Itkblx!8f?K8Zc(YL82Q1#!!;Px zhmgJL?H5c&lB4BOk`CIH2rCQ#SSaFKO!@=BSd7knl#(2UZDw8D?7PFhFyabyjc~0z z7j{#*!fLpQB-}^iEYRBeoF!ap(#f)02NG6DkfLZ&TLAjr8dA?+1{aTurgt=ISFh~5 zld{#xS!K1!$g>#P{@ehvc$rv@i_Fn*K!czAm{6PZUP-=B=6>7 zq3g_WkPVmYvK>?{I5^~8$&|TteO$y-61(%w696Qy)=tH}fchqMk8_hEnBeWRJKO%A zEl!_sE*~!4kT3PQ(R?<%QQ%g>YJ1B7=DDPWpEx?v;plLg+w13whcGwgpStzwnPWS% z&9CL>VT|ltWuIF1K1v*Rpt$*9eLX0&+uTb^1ymG5ABL;RotLfRV{Ir%o{)H32Ons< zh2Xw$#yOZFPlALl$E1EgioBN_k4xvCYyTHrLHMPmZ;{nSVz9THbUL;Hg*HeLw&Epng)USl z6~f36)(JVYK_TF76TD<-hBb61p%2o22`0TX1GxZw{9~^8#osUx%W%W6`C~8Sm>OHy znA!L#k$R6j6?h;SS>3dYfP7rRotYzLQ7g?WrPkJyUPa+9+ZGq~7vT2-zcd`eA*}@v zXTH%QkdPf?P)_j8`Gyr)!?T_4c6P3zi0WGWs0 zG|4cGCTu@w!3!AX_V+q@;55`J==8VWZ~JiFtV|s!o#nCpHU6eX8R1uu=@j19H*HB6 zQcN^7VDwoSFO+|ChD+Gs>tDTnxt>TPpa^M*PVntLeO?U)yJVI^sY7_LQLOzht=gdz z^FY07fgB#y2NFFd^e~ShQP)8)$5@32OyoaACrD2EA=fT$F86~pAKl;kHpZU&Z`J6g zTSYq@w^`FVl9PYF4=WndbL(69%SCKgI^V5k8eF@bZM$^+yZ81I|L(0Zo<9LD@uQ&m z%Ie{3DNyik-lSdlkyiebzHks#0m~*dF`A z4!|!UqH85Dh4oosSGy!yqgW5RBZc0o;f?;>y0;?-%UkNd%IR_>H(nOye>dbXP$BQe zQ-VbPOO!EJ9VllRq+&bec3y$?eR7%Z z*$~^WbPV_u$mUKly>QiaNawKuAL^Bdq(AD@pQifm_qD!C(*;sNw>5+oYqP z4h0bgs*Sa(Kw`G8DDSnN^sje!#NLK{oHMij(czkFxF){g3Qctpta15=Ae zAiXcw?@lsTihuTrl7A=lSbXbO>XpnQ z+z~wMj4(v=?WNaz)iA%6x*Bn9OeHEkM*Q4a5fOROVJ+P_>(APbtjk`lGm<-jDoeRW z$gIzeGk1;NGNp=rmuee8Gs+9O1i%zbF^U{|+NVQd&Ig327lP7oGG+p11_tPaVIiOc zOmjh~OBmm^=|u;apoodA@iVS*S7jicC+A4)z@Wr~!M`sZRJ;+)I}iBT*vT_Nyh{8O z`AG2@Bo2B}RzAH}+Tz+m9$b|NuhUJfFR%Thc}~Dm(Zh`wEe{QDXio<#Wb$vv#6t${ z2GN3WCx=h450IdNhDQN!mRN&tvpT@h)TEuBhIUG7fo0(?9LrLB-`m8Kgh#bW5(w+S z2)39srDMWl(xfp1^G+t-T#$Kfy?Wg@1Mz@(PC#v-vm>`nIPN%Si8G;H8wZM}psUgO z;_QZv4aH3q7}TltO~|GdYBa&=fvoX}&=~O}wP*qMSkN#wm^b@2aA#WVsh%$02JX&- zthzXJ#FO*ogjTQ9Q8ZO$hfG>b0Q1K2+VBvGI7gEK*XJG+vh<@pRRce`I%}E?K@RM-9HZAM8_GxNmkKVwUu&7e-S={>U!a-ny zrKzfg^y;&0&1PAV>uGeRfm(y|f{fBU@Ro8!^jj z@3d7K#_II+F>^x=0^p$JJtM0ql5>M1B+~BSsYH6vI3di-?n>oP6Gs}vNR?^|5Mm>6 zJOMic@^wF|hh{8+E;sc;kTPk#gZ7Z!+FZp1CBCaqcA{>hR)Y!OcHqgVz zeG+<@d&aM^M3`Vaf%LRf?#ve#U+k+VvakFW7GTT67*LwVXEd=@*`6V7aB&Cz=|o?t zmv%QRIY_Jx8&MKxz&QqbSa*n&WGaceFEdM-*|Jem;)+gCc=Hj$$!?p6<=ufo{TqVvh22fZA? zV|wABXz`kft>b6FmC{Af`eT&HFdc{c6`f5k@?OTmKJu05gS0~_zTQOW(c?YLD>XpP z1zN;IMEfuVhqE5qTp3tl*#cXPP#ePYcJ1x_UvZ|}=(|R5UkF9asm&_zLfE+bKqsVa z@|~xC54Lu=%igSa_^i2f_yw4`;}JeUR|J|8LBR+^-Y+90gZOG@3N>5iT8XG3oGMS8 zsvptTcN*Z6K2HYeKT^DPT#ZO$=y!?vbK@eWDL?#1LTXJ8OB= z3|gif|1n~+|35G>FyLgg{r&NnwciHYHyB`f_K7jrt?vbX3W!67uuI8l$Zt{Z;$xI<`DKKv(pJ2eYH`k7==qY3SKAVK(Nzp^u%nf2}B z^b|v?hP{Eu{|0eE}-Zn*^(v=>@x)lP^NUqL2#w75QEB=s8~ZF`O&)oL9`RO72o->SN|vrT`}}{GNfd z(MBLkY7VkEryBgO2oQwd6~|_KO5L6Vm$1Dw>!>m zflfYK!-O7&tFmEi_@P%-Rdw@AS=Wmr?%nae2*zGnz)o@zznu)Hw%j#uOmf`;U|N5k zsk4IYrto}vo?1YDRu$9ZnSw1$7B)rHf5L7fljE6kk|mJhlT`1i77zraRDnKx*HiJ5 z{kLx2lJ8;-W(!Drm(z2o!{X4e7ykB3aU-NkQ2zXC0n|YAi^=fL%lplnmNFom8r)FA z;rDy;e+nx9#s50$Qt>&@2>vv&MMRKkf$nI5JSQ;62ME-pv|B_97);YdtNjrRJJ;*D zAcpnbiFCyO^A(4K#68~j1BQXZ_8?p`Rc+H(b@VNQn4r7#r4H^O!b=mq$V@>{8>c3I z*YB<+G_O3R z0<#otKG*&yJs;ZyfYMOFI!Nycm$m!AfL*Rz6sGv#xA?%CqQv>+mg15k5h~kp;aUZY zJ7l!VeJJ^k0oB1PWpbF;wG@g|NByA&&^kNm@fXxC+k){umF7eFg{Jx68v_s>gG~Z# zGP?9G%!sj3HyLfl7E5$-K^PJMlJoj$%ub?jajEm78b1bktAyypra#>D0+!>r3?(%j z4@%o~{3yo53iaY&H#*)(a%_G>o_e}-yiyU#6m;P;VH3`kXvn}uC3d6iEpcK~Onhc% zYrJ>FdXD2=SHk0E(EFOM#BZi4p+%;Yip~INKYFR*#Y#R7eS*5zAl2a7s#L{LGNZ0L zXfG|XhrIILSr<9&$rig}h(2ii<`6&?i&u%w3Xr%$y=fE$5A;=i$Oz^k_wOoXEcQs_ zYvV>kHqB|A@C=6YHU77wW0s=d05tXO&wcuLz3`>Qt;2T>pX`dUT9D#BN9H^AhUF)? zKVE9)QpVwd9%0-8b^U8nVj)dRIZZeo*ldI(s0Znho$Zq?OoOZYJx-SIXmR64%1q&O z%(9&y?>$;vB{!QkXKMKF4@&i0Zuj=eb+-Ug>`t}>X_{+?HAv~v;gHxR_^Bu8SNt>y zdId!b(Z@d|5d{JZs^J0StH2meN(%v5)ilx{;1r;?$<6()Gy7b>0d9 zMrG4)669OkU}rO(UBU_f?u$*jB$>!>bWf9^rpFn3?oR#@uF*m$=RjR3_oVfl*$U`R z^#hsy7(G7;y*Vlgn#s7bI(I!Q9=TK6jjttz)6U=Fabs-x3r!Sgj^klj{%Ccq%o-#>LqhzwNI~lgX4N~0Od6~;yZP2 zqy#~Z(s01YvcSCGB%4$UZf&`nqdo?n><5F?$qCByIqdZnrdzLjRH!W~E#ssu5rP(l zSW5#;Yzcm?qjVq2=C1Q3-x7aei zj4SGv{BT}VosIVbybJ_}8RT%v#p4`Q`$2`@&7)kQk=?fHXf!=Ll!saL6@!0@ZD|!r zM1CAn0B<)d9K_y@%9CsqRPg{;Ey7gUs3UHNu+&)S<$UKA*U$1V86Y_S8Dgn}eAFYB z8RyT=mQ_eH3;Rh86;B5%IyaV!u!hMNVO|0ybx{EE^aZZfpZKm^t_8(IP1EbuvuNLa zY!?g4^#|8Vx<0?*v`u@AGVS*Uq5V~@%BECXBA0mtI{rM_j=#vKK_@7O`$nrnHSfd6 z9nnw_@0?))C(n)HRbl-v^%U!W7t_VobBWMZIlz<+=sbtOQ#x}0r_Iht@fOll7O9j4 z|GtGZzmJ}^%EBId1GN^Fx?vmPt~QaGz6|AS;SdQB=?V-)*&BC&M!0)OSR_5Q)jGUh z%`^&bvkPoxqx!(5+7kt1*6#h!Ex5MJ$@t-mTXOatT!`!`!SJMrUaYLvgLNcMjpSDRyt~)%2W6~v) zrhtQAD@bToYJt8QY@*D~R~^h1JcJ0W*Uj3m`tFC31v;31OmzB32Hy=&!*x${?iXLB z>d=wbdOtP94Ki&cJuyI9Yi9}eBb`#AAh?}Uh+XRmO<0N7M5)`=VkCK6Iq!^NvLpZ8 zrr#F>+59nIfIgW`(O5cn-MA&JT{o=x|M3D_@M$2i(eXGPc&3C6w{a0_kQGf;`raVn zhZP;E&|bxg$Q`AV?BxRG5_!o*!t3V0)7`ek#HE-sDtezJ7{O|!Jq3~ix^z+W^k6KeDQ;EmJ}o#eDp_uJ$Q&KY|6F z5FoOmO^jty?lKw;@HJ2TmN(tA5Xme23+ArC>w3q=*bPvoIX^|CbtZ@ZpY^VX+y5$Nt+yPs}6!EptdA`tc)pUTy$ zag>~1SDF{CwsD$mZGS0u`lMtbg2lqjEH1)d%6nLSU80;GrnPI~$qm<4hzf z-~GP48dwyq2kIe76-;iW7?j9cE#jUs#E<5-Q;A%!EDqehQLGr?#x*n6_goa)m*C(K zZU14i*X6R;L^MS#ZNjHdb#Fgb<*Ljte|-D;cPP0sAP6}Hn@%Y%&|l(;AFTgc3;K-u z1Gcm28aYks;((p~z(HoO>UzI8B>eCS5Pk;c(}q!7VQlRf;WBG_S1c2%d&v#aw9@xe zrmccTPOU8;&j=*4{c|Y;6Vcjz-6fvzz?6{|x8 zL)?-Q3M4_jh)b5npHYtgfn~;&lW)!3wNmlXlj`kmL!Z}*=amn(stx&4-Y@(5nav(N zH}`$slcR4!qccFGM#&Az01-UJ2UFn22lyYcTwLy6(z*M)FGK|`w8^^9%byl1Ree2C zzO4Gx{`Ex5%mBZ=wLhMPg+*;&!}?4g+As5@H8376Xh~aoR>w`4}bjmIUzfhPs;-|Q2_oTp>jPH@nPWo8ZAEG8i6yS zG0+4sT$9mz(ay5C$wghDQ9?jQ#OU=7Mc8yPgtGR%k`bR^9|6LEW}~lku2g>zL-<#M zPyGGAF6W)#Q=gq*9g_)t!n^7|4ac?I?1YU8x){#z2My9Y>T%S@ro0c_ zM6#;`Op0Lqkv$xzdsUHHv)?HR{f+g|zdfYnYU<;01rEs8zO9raqY4 zi4z~KQpA0BhN@Rr%ERkR3Z&1=7dh=^a>G?k$Vc)EF;8WFu|2bMU_jB~e3>r!+hWd{ zzDAc%tM%3{D~eSXi&I*LlDI6Ei7%6cI@J&$M&RhdAV>_D#x=_=tmz7d51nU2Z|DY* zN%(4_2U~nD^AF+%2(YdDeKu$k6O_byn+|2 zn(&A-U&tSJqd?UkcT2dGw=6%a?f>cs3mfN~T<}<~*+$qW(uonVQyk*klmq2l%&Kll z?Uv7xRB4(UHHz8460m5|P}Vkr#6@WLCrp)K_(HqbX2Z`Ksq2-3E1#2NSx^*T3%xh{ zC6dIToX9)vSFA%5z-P`}QCi!~TjNNVaa(iz+OfmR)W@g4YAiw*L>tIcEUI_jtMcp` zzuq^}LeuHt0JeXq$_aE|{6MBA>6hr1q2axNpWCLpwJ|&|tE$18j}DHFo=IVUpnc?D zoKMPdki(x^&XasvTxA88wFbRa5LrH$}%Ycj-$n!m7-AUXQ(;=q9(0-{=24)k~;%gz9Za#97 zbd3QM)MCG z)XKXln{vYI{hQ1ul0u92V)={PAnoDD_j(g8CZHkt8#nn+#y`y6HT9p1J<-&DP^eyHeMrf#5OF6agT% zTp{x_P*7F8V4B{lU!;x^yAa6acy=zTjegerBYw)luQ!(`z>mb{x|tf0JF> z1$(>k+N?Q&0EH(VYp`hHE^ey!5ljH$U}P;Qg*g#!P&@!Q_5gD^#8G6d%T;;dDVyMP za6q1C{?E+|$wi_IX{4TosjBkREo*PBKr0c1I`94Bt2XbxiR%@SlD+d<(vpy!fi+=X} zT(P(X`10@;B~g3lFu`z>s)1YqB9jSi-puT!x{VkgwqJaU+xu1CPBk;kFyu4k_VRW3z`=D(^9{Zq0@7W(WWkA59kEM?jt67N)-WrRNr#hF(l%)JvX# z>xoT{c~H=$!CEKYq<6u<855@B2pHC=A?Jhfk5i9GOL9*gCv%UU2(OPw+ja56K(olB zp5b~-lvkk5_Q-qmqFiaXalHpQl9^^Jso45jUo0`{X^)G7lCjFL=2J9;Jc>uoX{BtI zJ>Rw7(zai#Q|sF+72a&A3%qO3qxBJ+cLNru)Pn`XDB2fygKVRaIs8aBs)v!G^R}w~ zZ8YG&{^Uw&GB~EA$%Bsm{mtFO2oi}ll>ub}mGfk%z|ibLO2Fcc9bT1xE<@jzY*LH< z=X?0`Jj>1Bc+G3<+S08u=~oostA(TF%AVNW9@}Jq0s8=4;R(}(0|&wX4wB%))30tD zs<>6I8v{i^TQ7nbop3lnY@Qy^r9%~p7}L}i}Np7`HK5;fRM-d_$I}15BM~4O?_a_=yfM8hm_?e z{o8bJ^W?AZ&Pq^Ui~Yx9kGIZ(ez+f0oxGjUPr?`RtFN8kZsbOGWSs>!$h@(t?bY4= z^)cWt+0e*~rYhAceKn#EPaIsXF$PX3C)7dozsLeXblUeyC4H{xxCs*%HL!*h(@Bq$*1S{GdH0R&m1D=gNc{(FGEBb;UCYNEt z#NA)!yZ^u?0Wx|*i|xRVjBLDS)0UwHtff6M=IY^kCNx;^2WU>7c*8)jNF9DU4jGg4 zHw!2M@wSnp#|RMemG~+c+upX2%M-j%N~r{NBxxrjAEDc=W_r4;dvbClt+J|M+l~R6c9#{xQI=bmR_UT!`PtPPS5Ja&zsU)_=q)oB z@{UO%+lnVlH3amAAAf|R5f}L_@KZ;2Qclo9ZF`fS#d@JbIo+7~%2nZ=kr(yt=N30c z^AATpl+@2f_~eXyGS@mc|LlZU_zu^6t&0=2CcLuHAKo9HKM?+mcHOS<+v8gwG8V0jP^IX*m!djYZ&P7kL%skJ60jr7>9cE19VhwC54iG7rFyU_W0Mpf z81311ufbI5PYL-&r`RxZ=Sf-6)v@Q2xPejb4!az~ zjI`c57&Kux;C6O*T(Lm%9g{8((MeAtz1s(-5J=C6Jo(MyuVd7Nxl z+94jgqSop#QZOEtF}8nP;2Gcu9spG{d}tHP;F}Ikgq?!_%}Jda4jp+@)_NLtd#$S4 zlRXsjGaJ1#Ulh+AhkCSY!O~KmgQERl^XGSK6X#hLe%0IP+%Zqz&6fTZV>d~X|M3E- z)?kN}eMl#}j&M`FWS2UiC{SYcq&xuBT6nHJdmHAH6X}I_CZy-^y)tl=@7Y_38vP9#s_H*sp-ilxAKX!oAz}eSE+4&btNi zaGFw=^ZGTJu+A%^ev{%bcq-{FRFMfTin9cc3Z*4mrDO>E$uh&kJD>55gq~RNF4%@t zlFfc^vaYYDb~6u2GJi)-LvCTSiQ2?fd_rN~!N+ed_7X&&QUXk}? z_eA!3lv?xz0AztJ1HMAkbCMO;sXj!L|H1N0b-ULws%Tsa^qJp%rfIIEy`Obl1Rt}P zjk1HL5+Qie`94#W%A|B>qR3!7X9%wIath5`PNNBoMiz6FcT&Z0x5hV}FfaU9HgD!* zr-$bt;vemT-|Q~(Xd$7Ll|(gfnGzW16a6K4!x^^GL5UOVm-9()#*(9;SwLq4axkjD20R?krVq&&%1Vl>pOOnzIsxmmU~T6PEzg_U@|0O{K33x8 zFBf~|0;)B}N9RCSebHtux7Ieq`tLARz@1wE^AKO6WwSU$$hS&4kcuEr*9T$yxI`1F@Re%QyUkQvuPc~P+s5hGt5Te4^A1QQ&V z9F4W&tDFDCzb28qKt!W6P!e|m{5kAI_V9yTDqU^F48sA%Akga+jhxaH$F1t>PhFCb zZxEP~V_3KsD@)a@G7+-wXGnexO%eT}lnz1feP{D%Zd8 zp0{Zjdy8jv`H{2>+sH61Or8%PFJSkRW#_>1M(^p~Cd1Xl) z;85NGeZwRhkWy0Yv00K_CSmO|i%mW(NqT(Gw$>}$kEqtEN)mK(7x zoh|J=3WWOJK22GB_#=6vKYS3#$o8rWUD_0>Zx?p^r&V`5{AsV2 z)IX@X5nO&i$^+%~E2HM?nCV`6oTA_D_q!8z$WGzc1`^7ROwZ02Unw?nevPKEVhKLx-+5+((MwXqL6hbVp0DXHRdj(L9E~9tA!>N%ZwBY*H@)N+mLW%h1 z5l>5hj0;J*M~rHI(1F`3dU#d!Q^HXZ$dn+KAev!ZT$Gc6^NJ&itC3UMl&D)xuCJ9S zoVD{PU7NM#Me?4mGvATYc8^*S87rJ&J5OeqcZ*LHMWBAy79b@4rm- z6WY~lVHLMC7F_{f0#**8sj|h6n;4)O1y3$cf@^@@$!?6#pXAJttw%>Ltt(3LOFT1s*v$GRF4k~ta{;iw&zk8cM{#e%{TME5 zk3ve)Ef#~7Xi+8ZFW!Zi_T%}Vu|MEc#QuUv;$@cRAtI_Uhk-gFu>6z_eU!}|T)W{4 zeO=`#4`I*;O=5a;FBkLAr6jpRpdX6@hfr2i=USqI{NRCZmR0NOcmJH=RcVfCCEuQa zn9MmipQIdN4aczcU%x70SpAYbTJ(i)OTMLH)~d9Jj&R!3zl1=Z0XRC>?8HL@Zq- zBl3~9wzhbIrx**x^N@8~d}|cgVcHg@$0#wm#k=CR9l;r1O6uQTZ-R~Vn#e=M`=W8s zN{0;2>_+}~&H2@_OkD5wSxjA)LMyxtdi;+KPA{-B`ad9eo|h-PYBo`04|t4or>(=3 zj@6tDPc$ybp!hFxxXbpxUlhG+dj}Xn^Eq@+u=zK=)yQx=rVWKut@JP`RW1J9#JN5M zl<-TaX@(aW;K&pv#WOm<!8^;>zUa1NfnO(TO1OohFkT< z=y6ILZ)YNH|9gHE(Xeh%=_aclC}&me(V#f89vA7)aLDsHX54*5)uNq9Wu@d6xWNW* zftw<@s_;8O_8%~Y%7x%JK{-JgLAwse+zJ-~XPe{)k`*R(#d!jPW>Mk5xdDHNxgPfc z80LczX<@!dthmuv%B1+*>P+H3AR!~xGCb7jgH_FAQzXrqSmw+0LGRHT#mUkoorfhl zr%Thu#rKXkL>_-e1O)~4_!d^Rm6P9`^{J4thB}L9-GrxKmc_DObZhl}sv^D)5X3P( zhB~C|{b~|x>+`?Qv%oj+qDt5{=|8=t_zm-HRSA0Y^bDPuTvrl$&l;#}XocCg(9O#y zaZ!^ZBvKk4a_{Pwk`15{Bp*?Z{^l)eys*lgY3xw4$2gu!nZST{@e$L$0&gl7K9j1g zYlK^+HTaIf?Tu%tMN80|Ly;esz5s6C_AQ>_KxSD|q_DXlo^N*BOMj72_TW=li zW?JTGX#In3Vj5#pr9E~!zN>!7H0HM4x)%@c@sVWLs{>SVzRz6ip1x!7)><_^?rYQs ziJB9|#+{20E7NheQX65w?_;!Bx?^8J6+DhksQq?1H=Q>-)PPvpR+ z_O+26AT~5hKKN>dj^sYi%b5PRB%eV|bNBCm5PtEDFV?K_F0zj=fOGhLlHQhYoEFd+ z^J?o1!S>i4Fr8vAndRG7xn>QA4ZVPz(J``Nkxgxvx%NIy03crgM*ay-k**r z_c)W}W`dtO;O{F&F;dd4i@HS=^AdxvqCm`bvc#w0-*S0nF|}h8u~UhWs@D`LlzH!D zBd+i5kdUs%v!`}-x@9qvY!V#66Kz(ROASE$o2{U?jgKYAWlf%KzU24JEb2IhUQq3G z?9=HMoGNbjNNuBA<0W1yYgkwLpiU&@Z*LCiv?N=qpV856Tgc)v6=~Mk&mZ)BA1)KD2Hxp@?FvtWlVj z6dI3t4H08#DYz{Qx<|)eQBX>6jgfX$R~5!FAA}Iw52o@9+h_|&%n3jTv5#OQUnb#C zzud-%t?vI~1YT8PEJ2W4`?&+f1qDt1rzjI|uXb$%-Ue*HTf7`W56f~fkj7UuPAzM?HGB;xetsXS~^j7Xjxw9knSV)TwjJ<71nu2~X+$CM{B+L`6Vdry+3%W`@^Mix^3u>O$dun!PXpk+IX_D+H=|gJkgtN+1>=JcX?ISpwguDrP20($KwGPEiTw`W?+0}Wy%z> z`FlE6)2yA?+)h|aXX>j+?x6ASZ*Zv_Ke^Cxn(qM_mLLRXj&D`A!r9@k z+VO_OpENEv%Li_FsDF-n-tXSdmNWF6l^m9-^dQ8CYBuL2oa_)&T?{(GfVQT{qQU4QWA#?Z?hrh?Ck*EEEAxc>5-Z^t(G>-Pugjb zu=~DW=kxOC82Kxh&xwQ@5F!5i!MjEKX2b}NaS3XcB^Vbs5qQXvqx&gB4;poU?&jd` z_U0x*0!00(#asPKrql(3R@S%lf5AUh^Lk%(U~(tJ3-$eqonx2$cn~ zOk?rDuv#L_&OL}dMM*HlI_08CGext25YZIZ+A7~lr(&Cu@fiV--j10R7{3R7bXph~ zS1D>=ow?NZTqv=7(iS%E29kgL_;KcdOCYJp1Hdwg5@fvhRsFIOOubxMfqBp?v%M3( z1H=zm!IG`qq-cLJpkoy)V@%QBrQK`RJMYavjwTwXs!giPIWhWN*P+K??4SEKbwO`Z zA*Z8^Fp?xj`oi*cXu<&=xh07boQ<>yz~Hbxf#BHN1?zdiCg$Fc=(M4+=h)=hd*3=IqJu_FX&Bn z+@rhTb6A=J^l6o=*Qda>wW6!j#t3q7xksEJf9-K{cN%q4xwj<&e>!B8D_-*ENPiUS zF_}2;U)y6BJBB}mSXetbEkDyt?BTav!$FM5i&*#ZiPO6B7a;m4TIyzpYoyon#`e9A z8n=-{rWIyeIKs8SPGL^y(2?@c4!;mB#J)*uABZErF~QG$83iu;vQK3?N`+I=6qHT$0RIQ*HJN=@=u4~l#RaEJ>LLU>ep>M{9lfVFYXnT%!^H& zo{yZ-11wvy+4-LNY^%&=!Q2Jx-A}aY~%|BJct?pg@Jc64=HjZ@jMh$ag_M{Bd>$L8B)PSL_y zDbNso{RpjNYA1H~eXx2u8DqK$=p>K(?iKe7OItk@+a}NrXg-RR=GrO6a4~_Sc);Ql zr5#d)=_e4}R~@P~)5*#>Apq_YCrjN61<+Do%>Z+dCFp}P*i^nuvn2uq33!}cKWn6+ zcS?ySc5Kzq5qG>2})3JX~^Fc^wYdSnKx3`snmlUcsz6lvpAk+1Huaqgz z%XY5q#Lf0nZ76~KmS=Wd0f2|sCm^;-nKuc`KfisG*ld6P{ej1ZysF9|JHGGy zS6wN|t=8ie7DeBrmh_@Jq#(q!5f|>m1zE&fnYxGLv!Rp;g ztJx#4 ztYm-|BlP3jhMK21sT1J&3Ddn!&5I=*`?AwgYLC*Pp!15?qK)(~`qV7XFr6XZKAXDJ zF3p}RM}xQgr+vGAPwNT}f>zfbp(B7>yBtD~w1?%mboC|ptZKgzC+l03L~06`zHdVb z1|Xbks_tvq6R>|U5WtvFmY6N8dZO3Q@8RF*o|E{J*S_zc$h;`Urlw8D;=Ile3ib!M z_c-RaBDckf4|r8?NxS|6?%2B(4nZ7mIn$O6aGN8XE8s0e>kdaby zv#a%xhP2m&hNA65VT||HS&%m~DwW^8FW!QD zWoCZ-G5NMS+Y$QCb?T)ssd|SVFmvP+JoDc}Xw@S0uj)7Z8XncR>HO6d@jweuAa&c* zq#lhp^Ny(WBxhSZ#JBcF*%gpfM8HBO7(=}epGV(n7VuM7kK(`dV!O`pHb?Igw;K)A z5q^jNm2aA6JvRDSk9 z16%68>K!~I_v}Dpu~ME|rb3=;+n#0BpibS3BMOaVfOgyFGw@vgqovS!9KGx4#2GQYn8$T7rbyiV!(9+(wlV z^XK~QeMI!5h&N`gY33i@Sn+Y#V;@I0#gxCaZ1>51eDi^s3q^g{J>aCu71auQ_@Y60 zT1K1w(bu(r^G7lIf*gidQWDI_GzSVYHYn+M6KVha>b!nr0D=?_&EY zVRTmhcRm(d$0eelqM2yIx@2b8Y%GZZPY#92LZBY%`9GIK&cAS9nwpr+mJhmSXc|80^pd{_wBC>j@=Si~KZ zcM8i%TU1=n6q`s`_Qz(1s)3M-0;r#kJQpFA;8{6@ZRoGhFSu}?z;}y^mo~gXRf|0P zLh#;8QBvf?WpbZX)FYlzl2CHi6^^?l;bB|Kg@yyo15J1!IwQ0Bp=Xe_jGBE%?Z;kX zFmcoYtHjMk(j0KM+G0%GDN!%_=aL=_EW&LM34&jVVxoa0r?2cR_zl1k0jctl*pGj7 z02KBChF~{)=7)%%Bf0BiCM6zPt**rO)p%=>u_fi$1&<{jxpy8h=|gok?mSeA5lYh9=s1MAOJ`>K0h z+>qn(g@gWA9UiRYY19gp5G0=1JK}v7cKE&x4?EX@*RnQRa3FoDW~-^al4G$g^?T%N zZ|v(IC|b0bq{IRwtg;Bd7r&iQ7}9SSJ|J__rQu-k0D3aGC>tUHJDV3jUj|w`$%kgi zWUVxTKlIg~Mc!TUHte{KNzCWJyX7SHUz{WMft2z#Q8p-%VP!+>o}jBp7!J2+kUWvY z+S-4P_pg{-;3qk2jCM)s8;?xqrJIfR z=%!`;^6vznR6d_~4D0n2A-4Nt(p}5F$VV#FRFDoQv|Beuv?O`vL)N0@ARYpUf9H=v z19t@TV=MJ@DC%{tTmD}CcqFJBUBcRQ6^BPhXytIe*MF1oK(c56QQe7X3XJ1#m=Jk3u(Nhr6 zZjPF23pm2?#txf)C!sBQx^Hl3J!(d&Wp;vdJ^aEFF>#d3DROczpy3Iru`f zY>xZR;+B$THvMV9AY@YaWrBgR1c~H>o5_+J?>I9TF7~x^Ifw|u$&3`9t->@zxx}Id z^$C>!$*<|6;G3DvkfzmUvzY?jOb%Yz12f zOR`qJsKjHN*{>N(xxOo~>+akHJ8P*6ihyymwvMxa#~uIyH$G^rFOL_V%I7~QDSRVQ8jAD2vM^}ZHn3ywRdYqZK=I##ojwc zi1mxl_xpSPe9m)nj&s7vIsQm)?!E8V`+XCF+-CW2_}3X(Lq?#A_?lA8zhgKD6AdM) z&EgFdR*A4qKZ4j+)rRcXm@l+3HZT6IOmg!9Xt14-e^mZ9 zDRA~a2Zu97KE-9m2KDi?mqQ>2oWO@8+HifAJptqme!FEvC{0!ZQ@=`naV5K?e=xO6HRrWI;(%=Q9F)NX^%^wk zb!0}ib<=~rFne=8OuzbEy9A*97sY+eWL&qED(fIE1gD|UN{LI7C4qm9w5I#|@BC)l!sM{S(Y|OD#T@+oWEKDUo$;PW zxI0}c=7v{)1Q|;PU%(BkX2$s0z(v@qC{Q_;z!(Tw^x_sF%Thj!2$n?-Hg|7jVt@8~ zpYo$CQrlm44l7*|skOmBe-TB{>HYwj-iLGVk-S|rUbW)S0q-T*t&<>LJc&p&dek20 z)}Y(AC^Z(%46h^{3G(eF(jgNT32@PRxCU8lnRo$$9(|xAMH86&Crz zfM1fp2rsI#0I*BA;or5yIOr1HYGJy|l=n>YwxIMq0KdF!T*7u7^{Qm3Qo38bb$D0q8mFq6`r)5!$(H%%dZO> zJE^3{L_CFAXS^hSR5`?eyxmzE0}Y-6mWcuVxC{3*kr?_jTxa}wm@=jZGuQU)9}4AY zKT&dLxj&t83^t2CUWrg(TbAvThCg}V?`j#zC2Ip9j6DC&u&vhpY8!q?6U{=FLoWy1 zGbl4p)wE7EE%53Yhy#_wMPly0UY11 zRW070Ys~M^izLYIZ^}>(Fe?!1Hmsza=4)a}uKvmV1Sm4=oH)70(GCXskkSeen@0gE zdoOy=tlVEV|Dd*9_5B~L1Sm>Wv9Z7% zwKUPLrNrkgLkJm4(<6#omZE7gvCBSMnqlBrqlFIovobQ@e?P9~6%X_R7lBBmy|hxj z;(l9f!~UO&?NIv*fcBD!7D_#zzhHMq;@WBZY5^`r>I<6Xedpu#;kP1eN!ZBKWgJfajnfT@NFm9S4Ss9T0=(95DH908~#5P`ON4 zW2hQda#_jKeV)6K3*Iuyi1-VO&i@9Dzcb504ORBN;xCWmO(9Lg6U6BJ@MQtryJ`s!33gQf> zKz07ckwSp}?rA2cR)D=cUF3V_o#u{zAZTE04AjxBgX-JU`(HzRR$dyDxzl4FdxI!| zTp}T_zOL`WK_R{t({?3MF$se%aDiPSxrMfSf|(?R3n-CZY{T$fVKF!<$lUfXxZtW0E~f-d3c{5SSGIZ_4v2Tfu{f_ zP}Q?^ids+!)P_F#_8RZO?t+X4VTRtsT7O#mzcj+?xKBKzV&Fi>txil#%nIkpTT=j} zz!C00s}V{J$(Ow6uS($?cdNu^{uR7V?ftLBI$l9AxL&gVzh3f`1=mY%Xg)5mzUYu> z$_c-k-04yfWEqAj486>?GvR+Ee)SnI?^oi}##a)e=Z?rs*r2zC*T9;ad9~kW^f~tZ zwn^OtBil`r5mKOYvgA3#meIosG7(-*L=l!2DKwJ0AMp-cWKWFqdxt;3vZW&}UobE} zU@?@+b0e@+YaKSiv*pU?dd zxbq+#O<>FyXgT!AT&V(d4(xm7_Dqq*h*qaoPX@X0T_}+M9|%54#vD7og-OooxOT@ zaM7>pi_;@=pE5|Fp|bHUt?^h5I0;I6hF$YZbh5Q+r5^b!xTZ?S#|^2CJ^gP(CkW(6 z`A6>mMR$$fvkWuZf(r>0;X?>)gA48=XMr6er~0C)BRIxPI-9DU5aPkaZacq~w|NR4uu?5YF_p_qOhOs7R4`8t4S@Whn1Sbbi z38ZTevRBV4#Z7mWofyT|;s|A31?X|5%)n_#Zmw`RCzXrGH``G~I&o zTn1bC=AQ0MQFj^p@p!3a9Ndkbz=bk~$olfE=Gn28&kwzj^Tow4E6Hqd_99$cWX7UC z1zyLMS>S3f^Z#3;3sElD;o;1Hf0aT|1>7|P@>}2UqG5!SSh75FKB#x@Y&dOwo9ws0 zEsp6U0Gf+ZiPsq09Z}bvt>2_fl@Nq?jIcNs5^+5WWcjuqD5=FBuKz(9_;vWn6(#a2 zm_^+E1}wn&Ml6-7vKb1$akHJ*EF}y&JqQ@_DIl`4399> zO@h^Bg;|pWJAak`CH<+Xa8sol)*c7z7``H-8B!*UDK^#ZiQvArjrvYPY{YBsHDBc7 z?Z9oYR&ATlykWDcBq+d@amH%-uC#&IX-hxTp`zJor>?zQoA4mwY4CMZ>6UFh59@rH zg(s_sTC?ZI0l<=*^~**q6!QCzrt95Hi6-GEVSF2J&PwI*4xy>oPe)o|jDD}fU-`*P z+F2>1G4g2VKVb?BlBy}QErNjjAG)Oh8&y7-Jxh_!_;=^% z)wXAZzk-I7o4+3rw#|fL`sOh&W)~S$BqW+1cg4tED1Pb+Q_o^vTIvPq)Z%VC>QJ=u zY&GYZZgdhytTU)UHh$tm?$QMcq96cndS^E8jcmZL!;ieIua``k3|WZ2uOACf697W$ zHg|T&U|+8{GT1pC(%is(^e-38vNPi;4&q@8(9jJ->Z!qIiXj4dQcL&i%kvrVvyI8! z-Cf|=BFl->&w|XqHd}=IUdccdi@>q?NyEJ7JnW4eu0@e6wr|ssDyc zrinW=NQysP)?BmS>gvEt+RXi`ju<^3lxcD`GCt8d+eZAexEi-0fJ@A2!6gr9$}JyC zQQc_urP#_Y7j<Ba+} z8`PS4D-|GXufc%dar#XJKE!%WK%W+*4$$~KM-Y}pMqP1W4vc6)q&2)nk#uVD&;5EI zY@crYxI)!VKjZXS-_v^Ixt;Sx(-BQM5{mQKXaqxoK8J{I;kT2*mw_5{;doFyvLDLV zTZ~HIKehr=o8YPT>!q14Zs~WaUQf>2o>Z?J*ExKxvLo9BRBQpZYl!HabvPz1y%Mu$ zu=%33^9FV05<_XNew`%EW469O72E-NKn)y*+o$8@uJat(hZpGX%Kk<_rD3;qW!fK$d1eoL< zs4>%a%8XQ6ZB|%tN`D_KA@&9HrieaK-RdqoJ3d)MJ=r{1P)?VrXJKKO=s{?FvT8~P z$C+5G(mbhWVzRBqXdh>4lF$p&=)sg<*FNkK&~RpY>RYjRk`Y&q+Yo~SUXfofyO@BS zd{0^rE8cp^HkFAGqX)R>OS$5q^O7R>t$Of}fzJ>JQElH2A4W2WiY~PgPz4=k7@Z}2 zc|0=T-PWg;Btbiw5CNpEpqM^5d;4D6%?GYJc9&D`P$c!OW`_wL-tP0^z2p9s>gZ*t z^-%d{JM=Cv$;FusPXZR@vBDlz5a@~W>-7no;}Igd;`AEgVGheL$HOiasoiA#edo8> zjenW#*Octkbf_Ydq{N14l7+)|GDmUk&)KSP3P5+iP;KkaKfH~Bs`Q#Kfj|l_(Xzum z^hW{E6?VhgFGh zr}R>8`y##Hd$Fg}E;wDWcGc#u;4Q_6cE&x6FA)5h{|>iJ|7cZpL6oJoM#XTKohiPB zjB}A}BWHe5PNR3Ts&z(ce$HDNMGa^b3s^gn%LkLtx)Z(jPaiH=7+@A=bNvf*5B@8^@alK_5NkP zuwIaPz47~_jLlc0`9n^FOC6-l+%uj)m3r&o2Wu{%x5oIN=kj4~N%QvoyTzrBxZ2U% zoA&a;u5yPh>z)WiMROB;mEq2YvT>gjE)JP)RkpUF6{NE$N#9ndl4u54zI zV)_eeq!DOFtU`4Ev5WXD<6L&1Vqc_DXYBKBrSA*yOL}A_NiPA43pr!9R+4x6 zmmI>KqsV-6Begm|JX}_Hzf`oW1LyrRzx9-siCMn(!=QntZ>P8=#Ic%=WYKDwAR#}|0R`ui z8ePx%_`YQ&VRm8w8{wzOAJMW{TL3Ru3#=_b)Z{c`VaJJ}!-`(T?h_JWB(2W6L^N-5 zZ$?{`2$m$DDc=x$>_q%!eZqW9w7(8FFujO#%S#uy`Y!c{=hPur_$Vrf(wq(T3Ty?{ z=ArIas(p!Na8OUfrCw*bzGTu<1svsTBong7o-#+spSRrS^Ic;!A932?hf~2{4?Usr z(!2?91fKvB=;^#N3)!v+S@l{MQAOEM?duvQR<)HuvDFc|X7U2Qn9{|%rK547 z?ra4W`02-qRYX)M$$(H01V0QA0-(C1GUC+#EJ2Y7Itp8O-N;K>JS8!pF%?S`58$6U zdz5w&abpS}H`5gW>b+yu?^+4O7#M7i(U8C-*SbX2${fJWr2YPA!e7tiyMmUnrK;vy(+!#@$myvNag)p9R>vdOi@4o6 z*boJqw2)BM({}`j@E1W=^hzd$!Ph~*R#(d%h3-64q%1Bg+p|2A#BgZ%n#BSM_rIXm zPL2(E1FqXTj>@yWb3_SqS1i4ris!N9#H8$L}Sckp9v#cW5a6_40IQ zJ>y=KwUy*%k90^+K;?hZWGQa$)|l@m4l5uCfuRJ+u4dO6m5gw(_JR1HvpvjJGdDS* z2I_I713itGje6)ID(Vc%{|4S3ALt9L>gv-meJRZ<>4FQFpjTipG$=Ku+2GdR==w&S z6^x<(bfDI$f!BhFr+-LEWJ|qYuG8rm(xSyARTv(S-rfe8cCPSC`5Vz9fAfQKsC$$D zyCm}p2OB!=@6~5-1U1TxsHBi?pMQ{WxpT#CHT#hQ?VA!Hm1W72ghS9rP{k4)JM??@ zTo0;t@k9tj;_~u0KuDHW%bitE6~%##!84G3ludh;dAXT(K(&xEulX>G+&=auz20$G zcZ7#N{w!%ko56m8gIb1otXdZb7;u+$_~7J%TRsoQUjP*wz10mOWzOZOh2q%S8Ti$=W2B^fo3!1v&Y`pZjnJ|_XCf1Lg3~vhfxJ;2I zjNp@A-$Tc_zfo^Mqcjcg7t_>wojZ)CTGfx*3bm@ymldZ=)wC0XSU~UIXd@F zVL7%*{dkDY(oCoGuP-MqJM<=+RzDS|)jkEh;oc=mY;{4EHJ>F<9FIr{j>6QBdq>q! z0_4RK(E+)K1R`6xmVUB&)Z=XlfXCKvEkO_a5M18+22Vmgw9Opob! zo0qtz&)%+86Ba+!M?Qfr5{8WaQ0+fZ)bkUFN{-isnVp}&E-u$(M!)h^mJ z;}SkP_O?k*B9^u;5Q{$z0hf>f)H~Q4-+46@a5dm<+a&TThJeA+(w{ zH#ci>|L@F2ftE>Y)0ND);^Ny|yRs1tp{z zHrCbFPUoVpf0>vYNCa*%*&5KU55PVOw_Ymgh+4Nd#>J{mfG&;;JZGOO#(%ZB6-vZz zg_3TPia)jym*pnqY7JSwCosGWnFac1pBuS{;G)S0sC7ZHEYxgMnsHT#Q zKH&G7rjL~K>pLCn6j$(;eqdqy{ySFFi^qv)FJ+pdmE_)twie{B3OZCU%k?iCPYS2B*3Y{wvAH+IOW~$A(H8R9AH6vAcXh8mF|7!;8spLUfs;S@8u7>xB`{* zX@IUntltY$Ol{T0^`#NVkVgNg?k>^G&^e!MEw5QH}*U2DLy8&8lhk&72&x{ick?HNDPYUHP(P(y`8<|Gzl0-?(O{l^ABPZrzysbY2gjZ4+JtyUdw03s*e(&!K}S>GM@6r<>UE-dq9xmc(#SNR3;T z_#ugde&6JstWklF;C6h-m%GNFE9a#Q%c_-IPlZeUI>mjC`ej79GN~koxXfv5LKBnl zeJ>Ap+nznVM@076G?Co-yQBysp!0x~>ZQMX0v_uNfWba-Mc+>c=op_dK* zS3D_ul%j1Cim1>*$(!0zzkXAc%XYaygA^G}?_;{^4h!!aVGyu+?X*!dgImUF7lJ1kq^EGKf4f%n9 z-b)^&*vWgNhe>KyfAPXx5aHenDa%OlMV0_h8dk_?zq6Flv=w+W3ns%=e#L5oqNeP% zZ>1Z&RpevwmV0q)dVsSYs=XDkS=VQRK&T~ zsVl4cA4aOQkrNQJ3s}9XVOu<*kRTO2(D|kG+z_#OeYj;YpU3Dhb z$llh%X38}1q(@`>lCs4vO74tW@M4l%=Z=)oP`RzlxeEV)$O_SVAD=%HTdENzQ))JMdj(l+}>IYb9Y?;$KtwyD%EVi2cEu}0smZ~gNv?R(figC-sdjm;-rLeegmumBsV_Z_t{R)btrxK{w?*H7M*?7K-idr}eO9lPyC(Ig<|h`2$Ci9? z#P3F{IaQ;lT44LeWf`k%tFhdiPXKhsUARdM^jr9}f$h#;3P(p3qzs%!$%}vZW;9GgrYzs=>U(UFVqIV*gXvz2kz!Npy?Mx!z%> z&RMR%;7l5~ANZe#Wyx)uRJrE-*Z=1&Y2ov8Qf|aT8aKN9^k|Bdyrr5ng>DBKms(Y^ zF-;=`2_?TXX;KZ|3SR>LGZqxnsbSx~znB0XZ!#zU(|K$6w7vQtyP3G5d%H$>RDgk- z!1UO*w~3T)x9fVcv}WDM!2h~%90^Sk9y8j{p|FRp@lTwd(bYa(P~2|uMGdrGu1}l> z%GF@)=RlMPs9w^g5rHe^;*#@9n^JWj3v0SMWW|&~1Gd!p2}hOXs9r}yq4yEln4P-p+Lr_6s_&SaFR zo8XQerBME^{dWiV;(Ew(l4^!AFnf^YejHfv|7QUt z>CHO!N9wtmNp+pqH3o3`|Bb>1{PwPbSx2_Scz|4@HfE(`XYYA!Z3 zQdk)8pIQJH#I|!#*vhn9PnR(4Kimt20vY&PLG2&|xS6P$MT@U862n4_+$P|}LBC|b zT-R);6dej-m^spRb25Q|Fa@Uah-g{a@|RWGL{24|w$Y*+9aZq{eG_}qwZZx$;};)q z0z4DQJNHd0-v&E0 z`d7oNjT#0;gM&)Ubgpr1^te$5G>04!2cwk5?AV+%GV>p3R63zW@}m475Osyz0(Ch4 z^^mnfqKCh{=^h`2^AmmNp>T{{CBQc6YMJE-V^>%cf6kC~-$yT9%AHTZli1R|9ISMB zzkRuHe{k}w^V$nz>l=l79sgUI)8T)r`~PqmEe=ff228hBjmePkZ%DWQW^^U~E~#4& zyZ#ccy7PUq{(wqrfR8f+4MdOeAcqAv*zSq*t@_ga)z+%n^Y{jy`}uxfSND9z809W7 zD*-iPmqH)T8^8M&L0q&=42{;7j2f|h^(FiL>m?_bsuOt^>F9@iGRGzT*pWe#_HcJJ zq`{KT;*!)pHuQDfQ3J_4BLzXp@qZtxWig`@=(FU%YxB96+pC}QY(T~&eC=iF)f&RR zqnkedjz`{7r6#P>41&r~_g)ectel@&h8nuB0UOJ5F46j4hodyutbOU*#hOW)0;>eZC-Xxs#&o4OOJ2p9(%q@abNuGKrDsDvhk5RP>y#xG z+XVJ|4L@s)w{%ci`L!668ET8T$%*zPhm_oWQYL*$FKR&9iA8%E$kA4)K(TMx?tROX`kN`n!78pSs6YaDX5wKrETaE>1Q&FE z;Vm|4-}Z;l%Y5;W8}LJC<6hPebWQYYrMdx+-8t*-uNn9Hp=w2U$3n=L%=F@$#_`Q7 zkJ%-rr1|M1fv#=^=H|-LxO-QC7;0y>pUzLnC?S>IifvjMV3yOgt&tYJgJjWlV&=Dx zO01Flm9z}5c_FgtdOZ&JJRV4eHxJ1_d zNAeV`uLUl;&1D7BRa3bRwijq@Mz{jyG*D?|lK_MD!OZ&60_{-ldqqsRT$*?O!byr; zC8obe_8RwF7RCA_}d5lZ4Ljh17&i}V7+CNGFVN+Y6}U7_V6oQu83s# zi#YQ)6_TCjBA|(G1@BTVJzq7w%Scd`qVV6t#Mu!wo<22iSzmQuko`pLciIG|u!&C2 zi-PuELe~z#(JyR95Ymd=*ruizUfjr|`O<&=! z1%FFcdc8Eljq3nNB^Nt#;l$s$r|f!{&G^$Kaw29w0TU8E;&QpxWP?**{Evz~FrAdRJS7d3m&RMJv=Q>}NE|%pF;5H^LjFbWw6Vg;p3gqHAO=HTEM28tgCUIBFivZVYB?qqH@SGX zx+YY!4I*glTyLOUjs1b8tu=hTKkFlaIW}|{h?leKG$JiA4+}RSx!;Lp?-?aNpgMAV zWyy-1ZT{_(%p&9Q!jj=+*mazt49FTB^2opV4@(&D2&BRGV@Z8MCRX#nkNzxU^SU$! zN7wDe`#=&u2YB2BV0K^(&PpwPY1V+5+R%yS=~e<^{uj%X)QD5g`Y z**-8R?anKMfm++O9kk}8)i$-$i|wkrbAkJJ7VhmQw!FNUFPW4zz?2!$5c=JVoy5KD zU*kSYT~pa%1t^bSC!^{w>59+C3(jdy`^3IhT!Zev6?v5oR&fE{bQl5cqfQZZrK0kr z0#O_;mY`d&8kiPN^Kqfs`VJUeLKV41Ado!2+NY^#3!3F%wV|M!Fs)c($WVtHfjyNY znZKmr|595HaFL`s>ynnI_VHR&?;P>RA&q*qc6VOM3v@jN_Z^hIJ#YV#^{L709x~#8 zKc3&d-!d#KACuL_R~B)$Fw9DkdfvMnNfBJ$8Vj%@_{yoV_9d{DpCSE9R*Kr2=u^dT zMOEN7LmAKS>mWzv%Xfa}6uyc?Rxa7=%Tg>hgZy4m*Gf|>y<)3XAHJ^lI=M-rxmeGA zWD_QZO~|Z2#bm(zBzZPDDGe4mgfKGY?_^h2f!IF=)TshJVm@4YLUxP%)@_F9MR?0Q zb1gX^5$d7U1tzzcuMLv)D%YEUtAySuWZ>zGPmhw(W|&op)zOfQF1uw-mMrdm4O>p~ znolp)lDnPOE9(P~5-&%*?y(s!;CaQrPpaEwr=(0+g(X&f8FW1(p)5=zS|_-~G@Mtj zq_}*~c!zz*T-N;2iMr3jTvzs7o)l*=NYFN=P)wDZ`)`)6-zj+MmcE5wiWjy>eqm(|C-wvni*JZ?Be?N zD76ay`u)IzDB^FsG%MU=unqUd)5%u*DNj#igsR4Z7&6!93U2}(Y*80d8%=V5NIt(% z+3Sq>;VWK<#rkW0Gi*=8*H-IbHvjppueg`_!88eeCC~~~Hx&5wrDk3~BIkN<*n%;f zKrdw)Rjz*?CtcZ`EHmT=zHzwe7dG+805Hi5Avno-m01+IBe_?xLf>n*{ja1AeAQWv zPohA7NkDn;O|Ft}h#I$x{NLqo)9rRl`j4)qgW16^gzI>bX{T+*rB}tQ|FZK5*($Ky zyDkbNth(Oc?#k=xvtk7j>YeXp0e`&eDAqt3!W#W7=lrq$kh*oN;Yae1p3H=&Wz>iV zXq_0}qk@J<&VJ|n#szrjtN0>)edo{7e8mX=d?lfmOlw2!Epc$ZVTUz{`Qm3QeB4hL;G#M zw~$$jF<}WlE2?&(EGFct3A++p|JvLnO`^n*(O4nF6-FTKvQoH#UgrTojSC6^b?pVi*oSan5?;{7YyR zONAvTqF#Y~V7Xe4s}r+aC0)u5q)L8vsj}YoOw06?Tt%D1^tTJQX5BKQk_@u;2}5lD zK=a2DMY`7E+MeV}W z9uXA>%V3yNmZQ299eV$>d9d8G)qI7<&hdz>@fo2?dY>wn}P+;O~ zVo{XZq|FHmOk!eU5up+whn3_#p1c6np{_G|DX1td2W~amHvXm8wGU(Y8;JlM)gSNlcZMX zuJCD#BIs$tIP5Ze+}%7j`Q8)q1RJF^Yc|-#v>)0~z(NiwBqNO2T9q3|8G=fl+RpXV z8jTKK{}huuv@W(crXR8U+Tx4{RBijjTTV+OKkPHr&~lNvvt@CeQEOM!){vvKE>5na zeI4F^Lv+(@+xm{~W}D4;uTe@gwUc8g0#;L!J%yzes?_YhG7xqupC#EXu5~UFs~eyb zNun;yp9OZh2`vh8QLA?0%Xn8B$R)yv2U^oq98+~q1r1D_RG zvdNYy2jG9L>VGo)$YW|%M%d{3_k3*^We4@#AgO$>%mITv_Kz?llFwt$7lplE!wCxQ zoam&jYV)QE1G%?He&?CSb0cD)Ieq`tL~80qJmx$3ZTh|8hm^fau4nxUU=2jicf#=l5>Ex3MIH2EB;Aq z#28Eb39sr2`(e$PAmKDy7@&9V5-;CI0x-Z{ZWK6%4Vul#yh*8~tN~k?vpvUJ!~^mK zO~j2tn`+7|wrlXv6;d#jD^llIf6a2PetYad6$7D0fWU$E3?HqF*Q=}jh86F zD3r9zKS$dbp8k=}h>zPz{l-5&0`>oOY8dY9dY(=Zv1Pnwi}eg8$&hd=J1zHCQA^F4 ztqN4~Aj6FgfBqXC@Jpx)bx3eo^mFX+2fPpTId~7heo`S;`KI_wc>(hT#Mmks_nEmuc;->%JIXbtfKR4@CRJST>j zIFDBvYRK_Rck;iq&s!(<1EPoFwdu|XmmcwpGsVrYQ8 zgj=psR-+}EC*ISGC6HnkvSOq?mQ(dJYe@F&$=wkvw{;1P>gwqF@-u*~g48i>$63?>Yu=lnhY;#4 z`!A=GL=}jK3Bw&9k9q3?;19R(VI7}fLR^}c0jevFw*a#4*%s}PCQBh5vNoRXT%I}v z26`G#I|NLpfSuf0)BuFwXI|)cYTC?wJd~3dHA>3DnGdB8@pbUB(73-7P5-7G9KjuZ zkuzAzawu9qT<*rJz@gWy8b)KJi~w`M+aW=q0dJ_!L3-5!-8q+CE_op-<{Mbui@+iN z)5bw|3}Nc8Rt0J#%K`^EtZOFJWr3h6C*$}Lws5=U@`q_g8}&5RwIYvusL)w>KY)r@ zCqi7Ea7%5TuqQhc+Q_vmPiH;+_1FZ!w54L6x1I`+E23`tEQHX<-UTm794^{dUEw4h zJGp5B=ZLdo`#otVUZM;=9<-B_6XmgP7tZ*M2iv%u4W%5?lmxG*hY&R_cCnWswBUM?Y_Ln=)AIZ-VXfajFpiG88s# z%?Rl1I7MP_`n!)YxvY8#eIbw=?F>_SqCkJ?e9~*TUwTxq^Jo4#pntdu7KLWs>Lm^v zku~a{d7T~UJy89y6LUgECETQHnJ6Lt=AzZn@Sz3ISOo{ z?BFle z%RB@<@d#T6&vNcJ4&R(_(BG_dZ*FkeLKH9P2fxeuv;MEF^M8?c5Vk=6V|=$qZd??g zE(bUs*~*3x{qkd7>SwOeZU8y>V}LgN9$3ku4e=GoBtuX^Kox-3{wzb10s@r%lzFxx zjBg9P5BRvi2q)43%G{ajJK^0>qI*DbY+1-d(V8+AN}+andU^_|$fovc2H4Ib23+A3 zXl0_@fw4R6p~sxSlB`J}|0xCGWmZ`Na0vJ-P)HpYxd}3tgfHSOxFB!tEYRk#;i51A zSkp*E{AG0v_VVF5AH8KoI73IMSJR!G^a0l+qzQZzG)!4)h52 z*pl}Fb`7ZkOPC51$!&f1)Sj$zvvx?aEv7m+y`weU2({+9()Ib@uvx~w3Gi>dq}TG z>($y}AO{Vj8a51IukgtvYxhyspU%9SaB;H>)wh66Jv$t`a>ak|@9 zncutM@9<#S-@|48J&ZME=7af2bdJ>9{39D&!0Fu9N^Iihc9LprWSGDz4i0)M7`Oo)IIkh3^YZv2eogQzyh%(RAo|5pL7jWM z(YFnIjdU`4KY5hYdt{0vK~+yMww~EfTps96oZ}eIdj~hwgu8D#heo?mM%!%F=b*c; zp5B3w%JO{x-ayU}v5%V(W9Z zksa8!gY+jf45+(Gu6l9d{V^@+z$7>N37f8L#>FXHh0v~*vf;OUzgoK^LVG8tWU4s2 zg_Anq_o9p7=*a>$|FLOlgMXyLS3?7vc-^X5S<$G#nY!t&sBNX;?ys2@hMs#>-?5L_ zrvzXkY;)IHN%vMb(CE3Irc^vE2+}jX$uc#cH^R=|M!7_-a;13y}OVJeO%tE z3r4@hHP7qgGu%#y?n*63g%nrxRXmUjzzbkk2>k>j zu_S4K2B-7BV@~Io&|N3)uQ1h4QURBr+IS3C20SpuTd>sUu{eHk%)a;J(_Jf>L6JQx z_HH<3KY$dhq>V2R3RAFI!i8|dX$FUivm(|vV)A&LGOWI>OL*ax>3HM=WH#O@0Syk| zzCf`B?#nFjRuBbg1*rdkoP`3>3FL@ zv6Ah0)9_26AC5+_k9f+E*`gqm6)K(w-oF`AFthdSvw0wRoU?DZrJA52*~D~gQg%4z znh{bLCzG%)`zCQ69$FI-s0+<|MG@c;A{yD2NTG-z-@Z?Nl4@kpg~C@sp5d2ZE%u*;{lb;| zS({iNXEy!|j}Xhn)BeHDFDT((WNo}|+>8jxu&3L(MEcv63d2UX%8|`_Oi$eJJWs{& z-}I%{fgL(0#My;cx7m2j@TzQe+D3mK14l+5PWQtI-!$y|ChZrk5fH@U6M zHzu3BUk=2vCpH3& zE`b2S{^MsH+Mip`qCF;KB|g{dXKOpwVH_*!{13G8Q8L+sGHi*W;7*2TQF9f)C;+r$ zBr)4pYyK5DP@pankkSP1Mr#-t9nsxX)_(MVh&yC9e{Kde=w7k=kx#d(Z1Atb%mF

CVI&2^N zPHC(E)F52A`*$Kcg`V!a{<3kdQj{akl)xFhb5UB!tl+l@`tCNF42~L?{x-2ewdDK% zdg^t(5)uUlk=%lz{2w|Ls2=L~g^Io+eg&nHKgT&p|4%u2eoJXlcb-G`^#5Bi>lfMN zx$a2n;9!D^JGs&!A2CgzSV+zAW!5_msZ)1+9)O9U@;CAphtQR_dwcW~EXRM&qUpZW zWCI|Ne=|r>9t}#t^-~s@aQ&{Wl1+(ac@c~#xB>$o1xA&S2$ot zk{;FD&O-*;iJ^qwbNA4W!CGcM)f5LRLf+zoL` z*6)J{!BTkgIugenI8D+nSg>ep5*FfIWZ&z`p^w44mA^@BYeK?aG!; zibxCOP@j&o9<~5kp6n^?GWkD*KKT^w0x(}7^3A4DNAxA*<5L9@q5(ZQGMjp=8`K@y zs1>}{UxXE(9n0pnyK(S#uTtx}4OHGAUJg)BgmaC=w{CCeWmDqL?`#pze^YyrF2jW^ z!-q)9tE2ug0E~cpfZ*>=v!N108@o!a52&1OuCKC9G4I=RkuoJ}%L{B-hn6+3LhuZl zBq$)SRH`x$28R8H6qwGXHJXfZ@j2t+@zj}v2?Ql^tOl@ToFs9knsmN64~ZYL$_5*2 zl~bMSjwI=W;=TW%UGgKF0vJ!S<1w$Z9Rd>fj6Y1}gRGMJUZeLs`1@Nl*b~haJA2=I z1rQh$LpOh8D`A&`H*X#0Z>n*6t8nS&warS(YlaFwJN6NscM>fa9&`m(FpbkW+g!ZO z-l5cB`np6AyXyZY|GI>oRq(2SXv*>TvrHK~iLblW1DI0?F1*aX$>Xtr$^S>yTL-k+ zEbYU^gIj^N6xR|cr4%b}#S1MZ1oz_Z?(R-;Efye9yl8MKt^ta>yTccr^PJy#zdr-X zeb3I$TyyQr>~0E4h;NU|vu;E3KgMa0aKYp$AzjO?8&mT}>kP-Nv9`0#Rs;L!7Djwv z*=}jm>7`ZmtoI|8#&~0o3dN)_r*7LfZ6#HfN(-9HZ>sl)(?Z(Ki+7DA&R#uL$f}0l zYyK9fA6_td{ee@u!pjf65Q!=+bjS{B?z!1@>&neO_ha$_4pjfy(B?HSqP$C(+-?PJ zqjvorJW1`7*hZUOJgbg=^k{sjfFK^_w=MBG`X}&Lj_le}OT8CzXOX;K0io$T0exIxa~%NFD`U)P_)!fcpEp z6tsaoK~BMObH{y^hQi0z+Y?Z0`A|vuPJe&Do1>L8sHN$!{P8x){@^jwez9;8=Q687 z5!MdXPjUGfb#Q((=J9Y%rW}wc?}4IsUat+n)a#Nv`K?}Xv~gY6A*51|Bk(g(H0x%K zspe3nqvfcJeTTmi+PfKkK`eaJ!Ogg`S60V9bIRUA|6Xp|cZh17QbkedxL&({_2Lhi zHUOlCMwrw@e<0*Img%K#fViy3%}$*8O8|ycHalPsWP8A*H%TVK{f!81@WmhNOd0?L z6?O7O&nY%ix^)7-5TGERN$V~qS&=o(SpSZ?h-k(By;pEAuD5$?YV#4sRj?<2W*B}yrdo?#sM6=hpi z$De|?l^8u~GlfL$M!8SAZcLb3ua(G7O=6O(W6vUp*D$VMv~HO}$@KzB2=l276hD?T zy^HYu$VayOGx`o)^=S2lenR7zuhVFY30tBNS#-M($7*U0MRlTGeNu7o-7){hGCVAR zj02;MP=va!MCmyxBj^lBIr6pP13X5OurT+-`_wEU+nvrv!Dp760_UTj4`dGb z`}6jjaXy33iT>QP0m%a<{WOsu`mEJ%qrG}QmG>8@SL&YUf*f>36UK4B#i_(uHFbA> z0cP@Au5%tHlT7+A&oD-Twec@K^g7jPQT)t>d9ejYu9B5Cjh0jY={2F4aHtYD9$py-7R|+X>-Tdamav<~;t;Gf*`D3vTG#39Zdp*o(lmo2X-XS`zt6G2&(ECo` z5c=KR{siStqt)!%xFuf@Vu(ilzLW*KyprlxlKin3AHKp#3wc=dw(2-+F8{a|eujq$ z3_!W?U(xFc-`bm+1b2>!;|mg?qS;Bwm2>@2;!!z_EtVma zSDalLtGd5dZ7nGwRt+&eJ9-q_S5bNOb{lQ6TKHkyyF!II*@%xpr^fBbd#kU7>k6`4bUBR_q1KHJILC_L-7?zxDi4*~WoJC1f!Iuc(kRO6^OQzEu^prrAif;M zLbvn7bm^xTK2F=#-cnmz>}Q~+CnCP8MzZs5vj(7_eGDNHs;2P6V;x0$i2|2;ETz(*AN2+F7sW)$ zA4YV_7o<_`aDG@T(_zp;4=!9@>@HSX&lYya^pw3M90FLZ?k6!VN7>jCO z?s2K|h7X_C3aJP}mY%)rN*UKFSL;qYi{M%9aO~sAr+M+7R2K#w;ob>Zqq6+&0mY#c zoF+N{uw$lSj1L_0mz$T0c<6D(hNHh-TZXTAEq(lcx8jH#RUKSeuz&W$Lz8Kufvg>~ zv{}T_yMGW}uck^jT<5wDbbDeAzHY9A3{rw_ApLPAPWs(J!vcl{w^X$>YRb3@6u<6@ zF4A8mCB_5Nxo?HHsoeQRLHQYLT-0n5y&pD=Etdx#yZgOU_Vg0#QRN9i8tPzE6(3dz zHKg0<9IaC24MpXYl&~e!IuVp%gPy%A=che^2;0I4d`EUlL0wWut!C<^iR*2%Y2tL& zubS!1x{Y-nNK=!cxTBifdg3QW!*jYz@N&Gc5i!+u{vWSn9*Yyj#~NwovR>;%h;8!4 z_nWgtXw}X@wAp!p+e;G`#92Xs*%*KxOG4GtO`WQZ)@xzNS-(!Nr{13u*=d)xJnrK> zVRJ;QcoCb&J-#Iy0jQ$5xgNC>wv*>fkAIQ@Wi4VSmV*0f znKR@k2UhX(VM-8w$kocaM7_=O0#tP>PMT5kPC?-_qDb?Yt&Q$RvnAQSK2@Z;Ppl|t zbh+T3e6cRen*VU7_&EQ!@hHk`AgO$`x+m+(CvpevuBO7Xd7O4!ZaYNqJT!W@>oq1f zJRJ-Q=)aGq8(~6qD&?gjg_-ep5C*QKaNqM|1s#L%5K6{;l4J-xtk7(I_Gz?2BhZ<`dm^toPOc(pWG6g`_P zhjSqcn(MWL-~krY$WU0$TgF?OINsq-0nZ%-Mj#>K#Sqp@=0{H8&k~j?k1SKV?^Qa& zu>AaLN4Uk(qgP%aLJ6||Gnnr)B-CY5*CpIp0^7(&`p)@?jZJD3+NqeRHMDpz#&ugf zM<8!#KDMceJH1RwA&PrQ!HZtG`xu%SmTCh;6m*3&LYZ?dt3%&Yf9IxyNK1KNTO>pM z67*P2dG@JCgpTNt;fV>l+g2WBnbA=QIL7Zyy2ZS;y>0x5s`RRDBEf$ zqLF`5SDbPvuR;nIR!J>r8MgfU#32LfqQ{wA-%I$G&R3^v;^5p%S(-VndR-+DDnpcgYd&^9ay9u4j6%;B`Q7+T2uP4u*nOW*R2Virk z(0YOwhIW;BVxwpxs98gL%t|<%|ByZ`>C7Ebc@JU1S4}!p!&deGQykBIc1-o_ zr0}6}U2$HqXF?e8F-iiq<)c-0^`q3BxLf!!x2E-qkG%P$QH1E%Ul#QTj0cpo8#!Ui zb^GX9zo80G8g(#(J9xjFfx2vIx!d3B5KxN2kl!RMz{+lN-o`oXHO|r`@L_}P{P=7^ zU4cOR=BggI%?Dip+c%gMAKc;^>Lng9?Vem=vKNr{Hor1f&t1%L;$;H~>kf^4z1w+e z^HrJkTmew^Ttx9AtU=@ZQF66gZe!!m6ZkU2MZEVqlc6UAa*a5z;N>5(^Ak_4DO;|Z zKYKLuI6Dldxp)1fF0RF{g=apGIaCkrO8DNenX<&U5_OKrXTYE4?p&ygX#cy^=fo_-J3Dc{$a=kS$~I` zrF2^XrM%mJQ2D4BTWI)f@Pi6IRX=ZEK`q7K=sXC4j%niO;|IvlL#PvJ)q?)UGeUA1 zh*71gp|DqIcBpa$4$O?VZ=x|Te;E&>uRQa_5-e3kL_#D9OBj1lAB#2OvYzdx&+kNEdKdg{Q&3c%iSo-SDNP}~DKGFI$aO`Bi9uwb@ z_!@H@_zhVK`hBlY&J8a5%o^Sm{GLt^?b!D>|IOz%tk7fBU}>{4jF6^CQcwij^z(o} zp8>#r<0cwoK!Ui&H7=5b#?~Cj*9-9#>VAJP3dROMO&E1cF1c`dmo&J`JhCxFF<4s2 zm%>OCEc*52PmNSs8*yBO_07-~2kEgJQ)X>Ua3gcL`3oXaNPD8kL{~~7W~r4W3OQWN z`-OAA95HAfXo?`dVchSuV}`5QhZOT~;iU#l5Vyd!hh+OBOyIkTze zJh`vOy$4<$HXD^@-BVc_UfWSimY@l?RrgG+ML)*jwjR1mM2NbHfQyN)pH0B;MrK9S zOxd9g!`?%Cr8>~%9mmOqTGs|j*2i9*x?-vub8Zi_Zqn+yNjTz^A6%VIhNX5|Nd33+ z5D^%|(md=dbs9A!%A6f~+4`t=t@53&=gJS@FOqal#=>qr=%M4auA)J@onI=GRST&2 zy*GNHMi+&=$X@@^qKIB+9O5_m{ zqE_|-CjUl~30A#;=xsx1dR+ioYcL4A!R)kZj@h?)+CE1oqU z`1`b3tsnT0NLd?r6KpZ*kt`gDQj6?~V2PaRPQp651Nwu7&S=QVDaLc{EOlwL3djOT z97W!7&3V(PDYL(+77u8J<#(kO{O+{PSqHY_n;%9kyKRm3tq(Slk% zxhdZ;*C&4cIoh|KwwAWhHj1`~wu>X^H}Zm-dHH#s;G?cxbGo4BtNj2KcBR)5@A0Vh z&N2Bf?7ipgl%3IFfifPIN}$%Wpt3vfN7&wQEB$URiN-i3Xp+uZ@VN_F*xJXMGko~_<_+DDXz@OYWda*a!t+THKBIWo6#38=t>z`R(MJH%xiYqE2IwVXzeP&3!*to(e260X z{I{YOi$4)4!i_sDaEj}$gmdHK0CS6N9tOgAWyhi;Q13hJ`jkVVHWA7W=ZVpdQ2)FGEPVT^j6=DWLrYp zFHLG{>l-%{tE#n-Emg`)xE1ZKu?Y;=^-r3oo*A<|GLX_`C;YERe|~O82;&&ooRy-c zEYSb{oe^+-7d70_Y#y1s_mn<*T69`OBFL)7>Z&otQ+SNZrXwJn!H0#(xCs$}%vi69a&wy zdfXE)xcH;AgDBi3gr%2>H3*QPAk_Qv23HN{2>9Wbi4!X>;3!r_p#RtrJ(Oi-o1T03 zA&5gOqR>+J^V|z2n^d9q%?D2SI8_6-r@zxfcp33m|A0mGN6Btv<|7>L zrQdxaHKiVEIWMzD6KD@|`to+&M$i|#bsT|e;g|S^@bA;0R4B6XMc?3B362g+h(m19$M~vH@-69YLlBp7h1&YqT+%q@C-K6ANpS~(h&Y_A1hd|qQ`LQV z*4@uYkLx-}bv5$}&{!_8O5wiHyQiPZ4Q^Lps`fqfJV;Hy^4Z~=#I)BREVeHZ9n-?M zt#Fsh^7@~Ndqr&CD{uI(!&_OuO0RVVo?aXrnjIGRWI;ojva0T`5}*X-sW}rqgEY)UX|)D=LObrc_`3jkR(r#HyjVSsQPHIIW#0qt#{WUS-cyE+7}eD7k&QoqOGK_ z5NI$r<`RKNW*Bsb^ji@qOHd1FN^u!!Wm;aDHO~4~kf@S~u2;|s;GoUwtb~s%dd?rT z>?VY~v15quUNdMdQC)OV{_@KG{yU^z*kn+qSz74YUvbwcS3bkdO4A`>wViN_3i-E6 zdIl6eS1FrG<(469BRMKMBgHGBU3oc1@f7ZDP@RGh=)m7GpiCu%>I8JA66$Q~ zNuwFDhV9&GY@+x+t3>(Y69rr23-=vW4C#@@JPGEDUmyYvvUB9%`yy=Z>kVj;dH zn|;{FL}k&^Wc^88lMJq$C1xCa+wHODz89iJ7W9zbWIV!X2Tg&W9su$d8L5V|Py=KM zl4)E^Lmb4hm%#L&j%8)Nbp54z@;2j&-ZXcOT2f2s53$e{ULw#03ZpKH47MJ|XuorM zUW6&M(=U6K=knDh>iRpY%4Tv*Ab_!(UBRcTPlv`2qX z^kSLT;ac)H$C~5{_{r|#^U-rMkozhLIsa+4g6Ce3j~#fbNQ zL%wIRnRSbyocD>VxzH^ac^NjD;34_Gt%m2_8~$^-$AZ~jOpW^~Rm+FR2NiP9Lw?LY zhP2E^%7}gCOSSFFX!UD61K+Cg+J>DDBan748+y(kSr*X;Y&Q;lG+P^{648#gPCF1g=r9jhEc^B zTSz+wmLCl9fjf~ju1!5Ub&pX2li2J{1lmMszMQGF5EroD+RBQb_8t4&++h=NfLqJ? z6+2wxE&iNLSrM=O;`R~p>hh4k^Tmz@bwjx(NoDb3g4mbHPXm;$i+t{_7e7|o<@h{Y zz8sHeivSR~ez?^&v$kHq3vP7M=;?p_LG@Uz5M{n9Q%-hY=(*;-9GJYfKCNBXh?@2y zWb%na}g(i02Rd$iTFl-#ef4vuhWAdekSJA7HbS3MrQDI zAvBQWlq2RVI7109{#O3)YHgH^9pebyv8YBa8m1%X?9kdr*UE!X#Of74pAc1{x|$Xb zNCk0qlL1T0kxCOlQm%Z6k#dcI*g)yhIYJPDm9C$=a7Y+0eisi_FX0C4>cdW?h=i^l z+A;|yaLk7jEGln{GVi;8S!Uq*uH`g z-RmFuq=ca>c?z7VI?cD4oouv5gcw1(Jt;mIMs^GDzgIH_eb7S_n&*koqKAxG=0)qu zY9fC1p08&{wB>zk|IEmu)d8~sxzp)>z@*g(Iiv~U{9GSa@$^(C0$Q-Xl>DgX|h!vbR?S@%8MYDMY_R!C`R&H2J7FZmzBy{D5}9uD_zHXq2+GzBUi@YN*Q zM=oaaYr2haxoPBzj%9#JL9B~K8gf%wxch}FDbxN=4fK& z#H2>%Wtvsj!id9kJg%KB43D*dy=TD1i4GFE4}u9(XU{B)qi8rRTm`F7256T-8*J=# zKH$(A2{2pjxyZxcRzqUplW}f`4ZQ*L#809x*ge)yLDzQ`vm*tnld7Y(X7=Llh7`Aa z)8Lfs=xH1Rw!c zOtST;5OgXvAO5Bjt*&wKn^ioBA}#MnY#clqIjLmF+C(YocRM7&yqrH#5 zl)bL6L^5E_B11<0vuoXCI_!1oih$PRQV*~~`Gu|Eg!h@07jIJ&k3LuFeM!slJhqdh z7-s!0E&h2D_)Tu|ZgA$EkC(d*>ID^&YWk%&P7~UH>&A$@*d^dDPqDS>GSLM$s*My- zKi`bT>t!YdIP-(*!^l_sWtFY5s#F#e@sOv%#wnudAK`{uLM4;N#**J7yeX%%wc;fs?OQ32mF&R@Ud zK{4$@}caagM^~*N1P!t`l*rr?em2B zl03;iOBm>cbgce}$`ieg1^%W?xDnQUph))N-Z$$rNFVlkW@2{gYm5 zN_A!J%~T2fh{&(EeINqqiry0lPhc!#Zh*nD$)^!-*6>%!2EIqMhc*Sc2#yM$h%CiW zrWXU%pLOqs2rMCyrejn%P*Z?cU~%SgjcvYe4U{Grw$(gwKQ-;o(vE%hQ0U5q`;GU; z9F@sTX~vo3D>vY4ls-17{5wdh-N@K1H^_{fjajrZy{-p*2O2wh&ARQLWEK9r4@*^* zh;3J`jWulQI-}m1m=WUDP+CDMc&B4ejfBPvo!<91X)TOnGk>=$i9ya`^GoaG02s?) zkB0c^1ud#N7#zxMvXygE=576}^W+`8bQO`egYVxU3ucekM}q;+b>}^4 z+pd0y`z&j2 zq&Yg^M#D~Ork-ATcJa}&)brliGT%-i>Q*A4W({%EW^tOCs*m{gtbLe)yi;4Nsa)&vPS2=Sm z^{3C2{DzR5HN@#5ou}q!M9Ds!oS1r!<~#d6^J3t>X-f743`$<(XrWX@iBC9SILmvW z<$fXAQtj=ni}V9Nsc>(mcXoXeob5V3lUEiLD95If1N{^M}9z)B*<4MOjyorm%3J*4ck@#e>`sxnsb(wuv#}kQ_aGs2w zXbRtZkv5Xedy$1*hySGV5E(sRFS?}wl~%&b8*bVqyUR}!cb6HoBiOL8KGR+z#=;+i zK}FIU@*)HX_HAmHF5|wz%ZqP|isYM!8Gk-l_Ac~KK6m__e5aHc07f_T>zx|@F*1$V zOLEN85bW(der{<-#yeo(J?fyeTnJi9ezuZ=k0dP zOhsDeyzLc{Mr-+vwoyZ^NKHe%LMwwj6|wnIZ<2=H1Z5wY9qMmG$4-y!S`;qS(_#T3 z_5%n|Yg0RKF-?>5D&gIS1)y)5@&sT@5YGr-uf^}uE5Eju2JRSn=<7z9lE3IOClH{v zV#ben)2iVE2}87N#An{p39gb?$dq#)Eji=p0^Xo<$Y!ma57+cvjZXd9`H9n=?q$b5 z!UO$zCq?D?;VjrZ5dVHzV48TE{C+)^R(M*%f@!mun6=)iVQ73BXS;Te+a~ZszBghW z(2oHW(nN{Cxa0@>Ga0JUZ_|RBubf~dWXyP2%N8i8Xxt5IEh9ibTB!0@Slg# zL-6%Lt6qm49XR3T*yKuz!sCi?^$D>e28?BdLa);2p_g1&*+>jbVyQq1%Bl$*ez2y6 z7)t#dKd%y6ynV@ko^@AOd($6Ydo@UAvHfGQZnSjTQh;4wI^fld5so69i(h7>0ntP2 zDxY#N8$i|TO(8^TehOY2g_dXjirY;g49y9NngQtmwpiIyoqH{f>l-qQGq@zV>JNpo zfDvwg-|P-SnlpK9;D_^HELCBzf+azbcSkxkuntA_mJcK$EElK!rAzC=w|)q9fwkNe zj=5%|`zDw27Olr)LaLyC0K-Qh>-ZQ=OSTIb4YqU6Bc^p$QJ(yuCde%OPp5gV-6UEk>V24Ns7tmru1ry2jS5s5wZTSJLj#PdT|xTstUUMYId|t3-W?lc8)(@_#&-!<%2{sk z&##3^-{E>uSoarrwp3qE4iL?|*dCpTy<0)MJR8qa%FMqfuewr{%02|tugp95HjDkU zR{u|i9^)xPk1^8RpGIKQaLdSh*YlQOhD}d$QeK4f4PI#bJU0C^;!}=ykHmc$CdY(j zA83)BH-X?!$Z&S^@)%sF2DKK0aEtUL?ojElwoh&B-;t3ZdrRgt74 zy`jKH6V7|Rl`L*z#o069kYd+|!P61+e!Z~_?(Pqj4_6A5JKx2!F2s_$F1jxyBUVs_ zEa|C?#6Ii36!)ygTtO!Ct7?lnjX<*X;9en^BLsv`a%m)N{``+8hleN@hX3N|y#9D`tFETdkJxR+ zvmZXq;Td*3p~!!&q*zBV_gUB!V}DIO$HOL1!M3S^vJK7SMYxOBS7B3qem=zv+lY3$ zBvEqdy7*c>j!%y(jv{mQ*7`bc;%jX)YwZPnnX)1sgUw_cC-*s)+6@sBlA}MRcuwrd^y}q#*07e&g=K1@nh0@)az-5vc#9U0Vyp#e^Qy> zEZn{S-MJ6<(?+id#~J%8pGfuy3!8pBu&ynR&p1gVm+K^lkGfx1%_5h*v6c`ia2FCO zaLqyPy5-L1eHZbgmqog$sW-2GUd8<|`XO??}+k=Je#KCoePHpI+tL$5( zLWyUg*@VS((Jt*}+WswDE=4Mpr8;*C?w30E#DpMAVC`C`wS=X|&bZSe`yJ2J*iY~C zf&q~Ctp+Ie`ujP*PE8HaZndt~gpO@gwEa$|)|jPrb_`pzi#zD-y^!rx7}ZSxh3APW za8N*3LCuoHD1$NWZBEWZXtsH6b)4RT-(<@@SE?CdL3A698W~Z!nkPjajFM#^#d{3@P)K-bZ+6(dx@$#%J2ZoP2UTjna9G+oBEZw zWaLG19nwxq5|?LQxBQDWn_1p86CL7Yazz+a9@LNs(*ZFyCQVq|7P}?wuw%oL`zGx7 zh4wo%Z4f`hi;#TTJ;agbG!sQYV&YY&3hBJ@!u2^2I|gNWPGtH#B8}v>tFd?!wR}qU z{)=Q^qCq?2p46(EE`vQoh|*ptspnE*TP5NX zQ+o9FHKW2!8eezh@=rU5_fz7HV4cBV9F`G_R|-O6cYj?n7jqzZc*haG)#Enzby~EpmSdpb8}{9i$@uWD z0zq=47P8l0Bhup$GA=KcLHJdn-n33TRTd?!ClMs(cC$=8?vZgr6p17gD2m|EUe*mW zbFzw2A``mm^i`;Kl1!|sWnlZyQ)-9J7BzAe8Y&yHW$TiavRnh97R+Wkr7LHG+zuxZ zxEVqQfy7K@yb#<~LoRyj5*zk*y2nu4EI>(Q#kL5r7&@87*8RH+X1 zEOppy4<5maLYhBVTdJK`8)hiKCdm_aoxtib)OX)O8dyB_!By;Y;3hEQn(WG8 zhlLpX^fo0?ib>XRF`+lxp8Kg7k*NH>b7O9;tR;f0K7XTC({P47utJHRZaj*o@jCY3%%Q^qD-RI-8xK|c?iyg!a}6kZB%d4;c0lslE|SB z`*4-A9k5=AO9?jX{x-eS{X{V_4Yuhift-UeMh;CWA8WzJ=H^Er>V&r*KaGCKB*u

D}xPBFLSBB|Y zeI^BAX%+{#AC5aRtzh9yv!~iHUjM49VvV2O)iXkM{DS{N?<3{AgC^atYIh@qTl~o7 zyq6Sq6(j#GPa~GPNJxmTA0&s06Gl|cgBDD@$4cxczu3=p8{T;;d9`KIECozviq2j7 zVfaC1As!!W+2sR)N%M=M^*CHcsi7qJgCQ56^!r;UVJA|-se9J_>arKInC*np zAu$2p+PgHg>A!qV`VaE&F{!Vu2#57+L_dCzWLc~98FS@*ec?=`n%@~iL@OL|R>e<* zNh{?TW)9Yx(vbY|sdhb%y`sg`UC8A9ei;Lhl~DG4Qc_J<1+oAM{z<~2#Wt{Je-<)N z(9imkV9fx^FrU30W=5t`)&n+>HF~T|Xj!&l$STk4atz31PtBuDpPG)Y8Po>=Kdrg1_on zML2?S&gc>=6<^dKM!7;#k=0Dum!bFB}l4`09)h z?sbdq4;{-Fkib^HT?s%w_N`Q!QI@WhpA$z8qytmShrwLOY(f@fvr(SM=F6LqpM)r3 zp*al2e_7tpk=f5@KpcH^(~HQMqvd4Nkvu;YUG8ny`C2aY?Nq~VFxxDF^{EbO8@KX>x?s~HJECDip&^ZB{bU3FLXO6j$PDt&BMm%bA}ShU5!OW`{&xi@&BAeayee51=oLdP86TMn3s< zbYCnU4?I2U98nN%eNtp=%%$7^QDoMWA{DEf%dHTKT)-uX{!NU~T_f>?Kr|u;{2CSU z;Ugfgx|tgfRj;P`ryI+ED5kHd7-AVI`U=-N{J*JUMBg&WQ{OU4mG*%Xp-qGPZ)w-Y zxhF5%CpbApge)5+LL1mlcGP^?z7uQ*BrnkYAAi*)e~N3kF#@6gd(>-}rxzcIyDae| zR;0dEFyC!N0Sbll`CkHr09zc07oEQt@njvsEcHZPEOmNX%PvvVErp$mQVLrC{@;Tv z$PhqK8=g=&2SfnDU!WnrPQzy~MGMCzB|ma78seW-oMWHCM#ikoO#g({GZI9S>=vs| zb>qL4abD!x0L060QJYkMNBIma>?&QgzeNniEcuz6)04GrXV$g@LAf~$GyiRq zd4{TKd$IdeJ^P1|+o0m1AjGB!`a-!U)Hp9VvuYosz8Nqn`ijlE6d~{)Dnb}1zbV~9 zN0=G3r$=M+ADn8VMGUG9FOp~Qe-CmCLC7PBIv(Cg_k?bAGXlzF2pM*e!=(5Sn?V>6 zn?YEJJ6?QT4p4FG9&=xLdXR_?p-opqg^jz>|JVam@hQ8n*ydid`{YJ=fmC}$PeNd3 z8#^P?0=|-Mn5DiHsT4een57ir8!!JI2O@)D+=}<_hbP6qqejfK-i!72gS>$Se?cLP zVn@V4^G8o=KO&I!GeR6QCKFsIbtVY9yGN7bKLX&eB2XKDZ^WhJe-FwIezHdKcg;oo zVr}aO{CO~v1=SX>r*;1?b`_^J? z=8EfzMip_9k=2wW5LPn>4E^0)j3He!Rdu-myf_e3*;B^bfnyy_Yf)kdSf{da-jk92 ztsPhHa2k%S!c9H_DFs}>2@jsDxX{1E5j%4PRyYK91V z{W4uPn}Vc5lRNjIHcwL z{Q?v?&Zdq*lk7yaxi_TN|4FM#HS2ZT$xsU^30o&=cqvn(0(_+qtLj_NGUwUT4@w{b ziB5t0PEAVFy$+rxedn{%sX#EMFKhoou<0FoaRoxIF%uUJgk0Il07Fs5JqAjQ2395r z-pQ8h+&`c+BZXR3#QPh$E}Cotuhi;LcZzq@a1w8nVsGsqMeXc!q5I-WYFSoWI8}b% z+?+zT4J>V4K-a_&`}RjZdgizbv5YQ!qdhWUhF&yhK3mcIe}Njrzgc>jt1 zPYo};$S*%3Yzqg!-1*NxGu13X9Uk_J807BSwb!(l5%V=I`7L(o!mygPc2m|i{}bd; zocNY2X~s>yvP{z`o|SBP&?x!;ezZfr?W&eYqPjz+PXHP-v8G2vZ)<3>Z{g16+mbTZAH4>^Utc;cI zf0vwOdy7ci;WIvA$Dr-5W2x2Jy<1Aq`qT-Y{KQkxoaE7z!)Ne9SZEI~vgwm7g}0kM{}W{iAb~)z zFzKs*OOI_)1ik+V6ZO_NlS!NDN9`9o>bj}%zWeDL+f8Z`U&+ogOz!VTix^KBj@K`J z-jcD$P@`@>^LyId?q2oRy4kPICt1vT->{Z<2V--qyq(?7V<2ZE>G-QQ$JQ~vAcB~E za`06w-3acJi7wo1+x}m9c=gt&|6?anZ4@+5PK;T~-}wB3S$R%96(=agtVVsASh~b% zVC{V1l|t9{vl1-xa7~&DYtNJ1!m~8qd(6!rYa4|c!h54i=M7*c$PLcbvu|Vv#tVnm z!yv!gR+#z|AUdJHVV>6WNM;C-$1s6572o=akggwi zB8>qZ{~JmW;4Jk7&Q!G4+5gzGf9R7#BSM{bS&by^59YjdKrTow{+F-77;2Rlv0W?$ z)*@z~gh2eDo?D$b>v0-Aff5+_!9lZhx(rkXS1&MA@th{dHSb?1ds|0cnL=$-!WqeQzEnis#;fm9#WUX48I|*Me4M@;vjW4(IMoRqNF0Y*k2Y z9Usq%GCsbC7^lJq3_3e3pQu7DfvcZ*)YI6-h4TM;{?Bh_o_(s z+V#ewbQVC|6DYc(tL5|8nz@)yp(-TT(=r??(ya7fsmmtNA!R0jq?h7w^Pk^X zca%qLzvQd5e+rW0{`z*8)3D%caA#a1$izKG7j%6?=oK8bsUr0iUwD=8W{S z!SF?FxB2r|AFYf2E$G{%o5GYbu14vUg=LvpX(E@#q;m}jH%SGo%#e`$fX<1fc^B$| z)zfFe6`-0O&Fo+RX8K_@Xu(6%?k7pgXODpUvuI_TAGkzK7Z3Mj{ND+`W07~ zOT)hJQOW-6#&uFt`~7^Zn;FoLtaSVLP0CLwvG|5GbNGMEr|PLoah6i&FykMqcmF4h zJz0IjyO(KT{_SVxO!t1BG8-n?p(Kgljnin0rwk=k%6KdtYrcmM4BTiZa;f~;sme9u zwmmL%NX>uC%$WpM5^WcI^G+<(NcsJBT6OgEiR=o3xU<)cggLm?Z}|XL=pAZ1O5~M| zvKL)(8lnY2QY_kHLtC!*7Rm;B+H;fshz=R@G zO)kRhy8iz5kevFKxU_>01b$P$XrVQZw{t+nMB@rf6P~|8D=qnKUldFb!g3<|@ioZqJ>xY%QlL`P{WR91m z{t$UIqb-k$#+_{-&I2WBj7{Xxa^BA?NTJJ!9!y1hpKT}qmUK_2LuAjM0#UpDOM%Gl z2P*3gLF!tIxr(hPdS-C^AlQa@10IHUa@VbWY^9C4Rl#>_9i3lB`;r5&(RbcZ(d0rd zQR`%^3RY2PD)#mL!BN6AN}!QGp@uiXq86J!eKfyb&F{*y`G3O7YxRGqO;ywVz>gZG z7!gZCAy?0y=DfAX;iUBgYb)EXa?P~5UF`52rJ}q5`7uSz68RSUr53?!aHmlkm()BEIF2cg}362s@NXL>pix8}ETvhLPj3Q|SWddK%t ztk+ejroOQ?;9Hn`m&GwA(m*~_2!9`!Vubxwt;fRI&&JIlNW8gHK=^_zRgb`f=jOj! z?Y>goW&7O(t~vcxx|~TBhnxnfX^LgUmfA@wF4cQ%`C*XfWM&5*ey%U*(?3`FNi^^fCU@ZNv6WHnbTHBer4+S?*bBl1;AD)q~7VX=qpB*ns2 zBuJmHcR=Xv+3gf1>Zq;5iqsPwNvuxs=p1P5&3kIWiaGW~%5$lEM!UOII$gb%v~h2G8kicw?R@%m7GSM#WN%X{V!!lL=0 zm7?)WPRrOa(yF=6&mVr*ut4}JZ*0JsXp%3d;)>tH`GMY67P79kk|B0qjagDw zR{uS?$jpJYf_j>|>Q%n#$9bYEYZ8zl7(|hgSry+?xSyhl7KOI}tqu$LcJ~ucY_F&L zY7O$(j0gdU#!(Q%OXQUMy(C~LEkE$mn?QxRKA7z&rPOS4i=zQk;+m43%#Xx4d_6z- z9ZsCzJgyfLq%mj3_8!6K?VMi;C%2EA;B?z27#ucn^@;uBA3o&6AoOMJU0!kn8yz>- z|IB^eS`UYs4{*6#=dgJGW^=x^cs-KlW2GG*0Vxby;N-IC>5GHz?iR zT|+n0C0)|p9nvK;Lk-<6B_Q1m(jXuqIrq2^|H|aISpyKII3Tx3Q0?j!?&Gce+Q-RiA&115rEzA#~-7d*{0N}mbgMVXJ^e~U!?ZVzt| z%s{QkL}y1iP;KLJwX_21_%jC&ifl&lVVF~sH^%9CulW|@+G4+(gjKi6p{o%&Vq+)5 zq|3y?fC_z>!sMskwBu8aMbEHuToK4F0lKU4O1}6mtl^!OyVJ|lI3h;87A_f!he)l< ze^7?Pdv*9(-EI7kvTAN_4y>1^|2oxcsS>*7`8uD^KOwjCzT~<&xm31a?z}$5V_TiQ zSz?9{qeB8J^<0>$CaYPv{Jl9fn60Xwv+!Ru==$vA$z=heCKUr{$EBlqxYb*fmq&!; zh-$}K3B9BYyZ@asjwjCJ;7#@C^8={P*Tmhr_{iz$DiWinId8`f9_Ga)pEGeY8G^1K zTUGTw*XE(ws+Eqdhv8`l^8GH?D;4Wdm%Hjj)-=tY$8MkFK3j&p?Rmxo1C-LIi%7O1r(cJv$c zCd;i@uq@$RKJq&*)L+-P-pn8ic-zZUe_45p(C=t{+AO9fR=6ss@ZiLTITT#<#gc|d zDt&fV=cnCul|cwNsvf!%m$WegrUXwVKB+J`H;-|7n5y)bz^EfEA{2#j%l{V-?=oMT zX|MZ}Fo5UE5mSTxhmeTx&j&7r*i4f26gdPdQ(b^TnWxa1FTOX{ulhj6ck)_Ur0Bx1 zx`};6Va%gjzAkAlB%hqvA`4pCyR1-`5*jH`=~lPytBx+LC(Y$!G(T~d3Yne>ahl0U znsD`9+Oqu^a?E;3HEF9y-+f;c-|oB@5hl&Dzk|IYHjg+}owZ0X6i+M|XrF6eI#t87 z)?#sH*hy?~P?*Y@G!meRsYO*>%xy5$snv{_`g;MWV=wP3@-FLb-Q7dWvv7m`Q=io8@&iUNa0zq9YRWV8MABOs z&WZCEaWPyx?Q#9&>i{-52E?lo?{nCY z*`7SmtV#T15iO`iJdEZvoyNBMI*GvaeI08tK_a1+1%8)`#3=)_kAowwzUm{#(UZ^F z-O%>G803yOQ94;ADJ+SHqnC!2A~yIdiN!2=0wZZlTH*tZ!qzMDhdDC*h|P|ILI^Qs zk?jS0pl49Da04-<`r0+EUHRS6X7gnDcS<;KVj{gWU}!xf_25k5t!YOWlsTJX>Fc(N ziCGZ>OpZR^DpT*OYL;*EqrW-IR3euGy@f^X>D6O@x1P%Ow_# zi*mY&W_|l9Vc_dZdWZ+v#r}p1PYL4pdW)Y~+f)~t8;-D^u1aXhf~TRg%F^T0P#yYN z%K=k8>o{qaa$fXB*dqF}g|dEk8>h;MrfgrcaTjBBejfJ@*m)SvX(bBMmmK40B}>iT zE@Y-j>J`=_*um&BDcNYu98_@R5R%kl-wO?ML3hZ?wuaT*Md9hy#elFwGr){7OmO3G zZ3K$g-N|vMTQ?jN-tTl7s@~6QOy(=+75uu;r^p-RX9mb07B2938qy$#l~^X=SLqxz)uNZ)3k%fh-gr`wPURfr#} zH$Te0m^wxWl79i-?}+I{Ym$)BH4?0YlQSqFBTweiZEq_6l8S#yvuaMT#{>ZaLxqZY z`yP3QD1`ABG>+D&EZg+Sh*wX%Iffn=5*pvUSUAAol~MUWlJ2t*a%rlQjzgi$DqM0( z`$6SGXC|czz|edLmM!!@8S^|jdt}2plLRkrL06wVue z#7yPbPHD6Jpt}c|G5EpT>h}9XD#}C;bzL|?QOr~CGNtw>+&cq;mYI~@D`>n z^UgLy(lR65RTKOBk!cMZl_k93TKVLj7+10?@1LT8u&w0mD}t%$M1Xaegj7wYgSqpR zEfrWkor_}zAcIs+mHshoWEL|NJHrkt;xDO!#{qG~nUoT^~pY22{ z4ISF1$jldbmO7n~PBNRpYZ9?6;7%`@s}1RcTE4{i`>_p*8dNyY`MP zX6xQpO8N#yX|z$x7%kVBHX7+ux~vHcsy4idT9fNSRj)oOD=%dZy`^mo0W$1z(I<-? zq>(cqXRd>m-r9LoIA&BErQ~$;&!h{3$7fDer`l7rir$&FPz5^fRTUXW_4-m>RYu#Z zlPfXWVz8zfYkP<*!rz3B5F5T`t*#2~RJc!goPIh(ewTBTM_ zV|%;30uv~x0JwFilJX-_&eauDd?7~b?Ao13NOnT51=HkhgQkb_h*+L9JtSe5K?MZs?P}trSTjpdFX2>A45S%q*Vz9!cWibA-$7ICu zPEX%Xx}@Th5jyyEU_S1>AoN^-#VAlAx2H07zQ|rmUK%lztSnkM88^AU?Y@US3A-*~ zS_jaF;6i@3f`Xn1U9Irp`?suhE`ECoPnMZq&2w^ytljNY;dYb8sPjbUYQ1#otkXdQ(kXE1st(T z1?`aiA!iyknCDu}uQo7h_`{{deGq=UBHu=4z-GFP^b0UewHP0ZTjH01kfAVK=ddh= zrrG;ka$jwSYVl|aR^>y<&8hPC9;X*^TxCL1l=oB2QIg`9p@SXkx@<95PmS^R+7i8r znUks6BM*PaHDb!Dg>oQ`E&-Y6A&zEG^YECPoE7NT*14M+uLR(L{Cc4jj09@ zbwEp0O%g;wZKBtQ6~3Y{*WgC&S-8i~%I3zRPZ_XjJ`)eHss`Dd`1;!4MKV(^gh;m{ z?9B*YgbC^y;7fp0&gkm~ou*0>%t`-tetM=D)Au3FG3$T!SB3Y&j_`cVJy=$@1_YYO9+Q-q%;6wR&b>hgv;MC=UP!A+fx2t>Vf$E~e(~k9bAJw3P{G)TZ z&&5+Q;YPT+@TO8e5PHUD_|O`l?3PqG8&JtIi@*JBCH7rDsg4iBR(F}GE(IHLYeXK! zvv~^`hBxo#k>YBTb-?iRw0G`VC^ zT$`StEi&jh#Sz_ew|M%t1uak7{K!1-44?|{4REEuur`~_pAN1rTk+C5Ub=#B6BMYv z;jX&f5m!Y8r{Y%Rec6161^)7zHKM&Yr@K|>RmX~MJ>GRu@|Ot_+0q)B^r`bF6$8ZZ zA}{2ykhoC1t0~sbMK;i6;;0zX`_eMJ9{=V5=R8acM=tD9Pd*AH<{q^ zpcO2=5^hYhwcqE`!M59HiIaauE(?9@ z@+$VduVx+!t#o+rTd6XE_K@*^iZ43Cc*=j^)yn^?zV29$IxAvA3XV5mKo_ zox=6Ewv6D3`vPit)Bd=@ot@OQ-(yB9za;1*Olitfr|alwkH2cmr`MZdm!|uDIor?P zEZVx6(ql>eynFXW+hT+zKN4s%T;c>f0v(Bpq8>vTb9+T0bt9r)Ct=K+^L6;pc;Ehl z>1XdDQ>e4voN2O!+JX%miCEoMsz3NA{?AITi7wSiKTH8M(>lNK!4a>-7=UB5u=sa&mBpt#Rz_;T&c@7P{YHQMhJc1`=jt~TuOyd| znFpy&+T0jNO?GTEZM*CB7!^1^fJX59a6~2ivn%%q;<)7>ww0Dn2Ko)UEq}g^B;EF5 zZhgAk-%rdH5C#fCmyuunJ`ik0e~qwu$18i$3>7G)hqnLIa&nA;_6YX+s+bY~D5~q} zDJy8om{FQGSXUc2TZ8C(zNX5HP*d8vx^fD7%yK_6s;axg=NaYxB)9i_ z@Bbrw1t))}Sx+t9ez5IhWd+M6IlyHI9`!r2Axw|%Gr)qmtgXNQn*4k-6R8xpB^@<; zB2;OmfI35yg!Xh)DvdOE+EoK_@NjL zKY-?y=LSFxN}~0BVAFNpB??0=PKzY81y3fjT=xeT`#M7hqC(! z$xpmAS9~q{Px(3R`}88CSwCZ1>~#=U|5Ke~u&y~@TCBWq&-$D2@rbMDakJk!51FRA zoL!V{c1D<%2EHol^^yQVA5h7rTRnAIFG&+}9);X;N~onAnJtO@=INp|azuNCc2GJ* zZBwH9JQMjg{495qmGG&8Ht)1A&9^>LkX6D zUyG?VAZkDZhks$8XhU;Z1%2%J-4@YbG!$ow-3;cOf5QoTtBvLxHAEaUD2@EsSW5ij zAjD5BdZ&GeiMcYkAq~H!LWS8x;+yno_fkyYV0feO@kcIyi;+`)ew)E^X#q(R)B$Wz zO(O3josBL3w*=onQT!ce6G98!cEmn61+2~Snfz$MG6Q2EViAhx!zq_o0&QVfc`z0W zIIKCFbpC0DpSGV)|JDWR=^t)P(Y(@NfbswDj^})t(FK= zAfh{G+|&(yBoMTj+5+c>%2cqF0j-GMKT<8A9Si#PoDBWK*=!c=o!~*dMgNvgUqglo zZCT*3>tvaX0(3Yn*`N+z+@ShsNy+<}%^XGitF5(en$hzbe~ADuVmgR+3YK+eQuLQ- zH(Ze{>%HXo%$1^XmF;@grPK1tig=A^?-*IofblVn0A;eRMTvbF#Nu+R<~nE=&Yirr zrdx_S-LGmy4_Hp)8R*JRPxOWVYU)ReD0>upZ);RHpLinlAi(?)hLgv%d{@s(eKO$)z@{_LNHg!j<6eJP%2Uut>}zdF!hY2z&Q*_N1_ zO1%Zq*Ft)4{3>&s4^0|Mzr@M@QPy^#ykd~Ugwda>`b1ghY~HY%=y4ShpJUy+>gv>_ zNQ_g4EKYH3s%6mt6fY9o##Rl5-dX6il!Gjd7{Ve1-rdqcQ(^)Aa1wcEyJ~|%44yua z8pp6|@cKOXR730dQ}6pv?FKa|kY;`%!?`-1EWcw7oJO!!BJPgkoBs`Kg2UE&k|S_< zd=%dwwXh^JFyB!IA!U|D77y%l34A_i9h(@_*YX`f@YalSDU5)hRh7H18M^z*oG}aF zIR}%aE9&EKzNcKD?JF&;p;u5w57s{|ot1YUTMb14geapm#`Kc+tof-?;h-v$A-R&M#isLJ5fR070(j;ZK{D^?QaqWt~3|<314Ihdqe&IH%Cb&JrnkS zcLDszsB|NSuE?u%GzHvl=*;oS+t|*oW>X_UH7Iu_W-dBY6jddRv`lY9pa!+r>Xb`vbfFTZ4hX@_-yE-4q%t4^+0erbS}T2wJ5w#TA2C2a zl~7$%TlI9a1=?K-DWJxXv&o&8AH9UmRx!?h1}|8V4oNg(nIyrWz9z4-ijQBws#JCw z!sfo(Um<{b9)duCj?w&o-R9@8#9{Wrzob+a#Katgj`QRk5*`60s@a$ZD;IdZQ_;V>@euDbX_M5f83Mx%C&3MR2faK>BNkJ zAl+X9T9)e6j@MpnaI2zNVd$oKEAaB9YpkXNYZ{?H6p@J@`tWokkA@y+LLSH}n|TSQ z_9X$gw4N(=kjCCw-jg8ql%JRevH1d@s&n~WG<}DXMS8`eFx_)~=zmB|Kt<8F=!7BU zF*o+A>hPNsg_PCEXWtJy=hl9V&i2~{J91!n2Eu5(_F^O?OSe0LM1sth5`1BV@K(g; zTdjy*j9DeIlk`#nd10F*{-YjxDgp|T4nIc;fop1dX$r*El_5nq8B_;Zs-e5hK%B8 znFGm&3XX^sO1@KRqWHOZoa5t&-Xs}%q1NAI=vc106sjO93%a_NlN)4_Tj(2M0Xv3w zCUik35Y8f!zY7gNdf0T@I!$Q7@+BE8`@a>ewk^Qia2br4z5cjE`KyX*X)pBSd1bYUULG@6WG1t)ecj;{` zm}k4Wgn%h^G%V219&m6h1^BpZsUWRW#YVv}M(8!F$%5l+>OHRV2Wi)pASMro05wNg z0VCYGC@RaHfgj#{zoHAiI)o|bo!X>uN_hL+m^?yrWGJ)3A<|vYNOvo+Rl7E=L8}D? zw-6tHedf41vU%F`T~nLNTU4A&#c|NbLjgxrC)(B4rXF+8=x2a=*G>kxRjyrA_BoZh zE@^E9)LqV>XLmooL2l1o$+Id)>X#v|f}J2GY1mHjV%%pgS+y~Jof+^lC4uk@_M8^_ zDgpZb1M~lCxbEr^$O?v`4@xVEzfgIomD)bEnN)Xtnq+K-qozOk)fzS{E7Kt@E4?TU zy#viKXc|^Lg{@)}RWq-aspg9X#A`8hVe9yeC(dW-feL1QL!v5k0MRdU`!Xf<=SjO4 z_|EU3E<)?1F;r{okXTp&wx?wUo2g_EX3O$TLlj!~j@5zSduWf^kQ~P6HhwmRC z*KZOgx}jW(6FDwMHpF9nsX#xJsz&#xyn`rF06b41BeF=~Pzy788^w9_hT6(8aOc79 z@`Abs7u+r@_ipHyQcL_BmLsI$WdaZmve3v}eFEn1;GLRFodZpJ)PbTa{2h%#34T*8 zEu5srb_g~Y>>BN4m%jp`jXO(H=+3%7P0|;={-sWPcNRLUw10ge1GU<(1q|KxjBQa} zgEOt+tLiBGJY$E@E%!E`{nxPPL|&44D5QcMWq&Yr!KcOL!A9=hFZS}`0vg4?^im?A zQ-_S>h;q8|;_8_4IEI*&jS8Y;O3I?7wM!5dQxqxgw}b5!UwELJ)J)K@K)Sq66`QT) zo)$>fCU%6@7qA&_;pIj6hBvjJ(Vs??3mt{V1_xeK_)^1t5>gNe5(+J_4liDQh=TbF zdsxs~(2OV!#;Nc!FwhbZK6BDA<9&C7ydfi2B^Ry!1vwSNhzrYIVN}6YWePJ~k$Gq} z`VJ~cemw3^Lhe%vR6i^>f@cQo$+0zI?IUeA4of7Y)JObsf@a(RhT#?3mDf%Ny0tg> zpv^lA*1-_y#P_S;_69>M@A>Tal;drR5~UDU)vtvepYv_Ur+*7MvYcM6s~v*Uq_pJz zh{$>(g~^HQx|%i3+D*@`V4W&M9p-ecC_0!(#vSJ!ZFwZ$A<78G^@HLRj|-%UPw9CB zQ+;AG7^;R}e)I*@PNCeJkBy#~C?sJToAtZNSN3o!ScaF#9VLr5-YEq56r8Dg?)gKW z^bXohw$4N$2Qm+FG2QeL8cKDOHM_gJCa=Pm`-5dGf@M3~;@{PNx>;wMs)C!HAI>iz za!;$wKG^esj>PUb`>e3(`hDDYWaGL)x|w|zYGv|pdOL`mzMSA~R``U%K;ElV-tUxd ztQGI}YltO`tR`l=F9G_Zs44(xTo}s)Hk(STYqg1uVrk%s5HHutuqIe-?c(z%|ptpQ;{tH<-QRGLu#- z@sUOn3nZu;Rsp;4fe4<1F{(aD1?ebGVHGa-li2vFhj5zPBRk42W4c!AK#_Jvd^|u` z-dDk)__3bo@M|6vA9uXV4W_5Xm(ncyPIW!n5jRwicTKC!!(UjA zY7>*FLy+HgdwVolQV|blc@RKSF8T=5!aC>FF9hj&reh^rjfZJ%t}#3?5?r12~nqu z-s{v`+e@_~HB$46d~YJrW9PQS_`YJ7V%|4G`xs&cJ+Bw&YK~kn^6)GE1sd<&Q$Sy2 z59wR%G#UUZhk>@UbBC3kL)Q8q236dl7sh&(HbRnf%cT`I`Ft^Z{=rZoUhv*IFG>+1 z%nK>~U`aGkNdYQ{Ex}eyp$2G*X&Xo}AbWa&mSnw+9FKqg+;i}@vCOp(cBT{>?gL7{ zTURx~TV$XEC?v%v98x>*8kKvx72&8N^JSnXHOmh=_$cNsBT;C{{FQ8ZN6Na~-ya%h z7sk-lLtd}3j;YN9KaeCcyjG!NY!6sv;H1XrJPTvm-9QOb&1jfA9I?%ubcn98{h6)J zntlRp_W}U5*r`*?-hbwGYhqFn+RvKAR>xrM4y%A(v~z)K&7>d zcA+Nv8B9AowNK!ZxRD!7TAV`XNvI@td_%v{p}<2zSBeLmf6!AzK+YPUL{bPUC&v{F zE9=6=VK#F&$~W;Z2B^KZz2kW;GN<+tp)wrDVNyga!+(daCkHk>mX2&36JC-_iw@-W z1gpIcSC9ahd^BOK-=Ltt)zWQ|w4a-{+j5QYDzBIeOUF|w)&6JmO1*rzur>d}%ZT~O znEPSri+GXf{o*IomPB}LnC~goRjGCPh_rHujLBq$AG*;CWe$Iai0VPb-+v7U270vy@jKCjJ#PR~cu5zm%Bv2dx5fxt9 z4OCJ|{7qQF=E=%>husiHOl=%JQy@(3x3cf{{HvP2^jvNkpK|?%b;>8dWL&&SH-(6z zcz26tk%PoZ-af3LF!zLiX5F`I$8=RN zp7Jx>I(Ij-C>Ur;fz(83thgn z$kb&Zs}eR(#KNCt@dM;F{%k;rcYW(l!1>jud0<*BIN8@_mgrPtT|i$9nsgs+hnyJ4Nr#OnzCgxkd!1jKwx>!u z6ZeKf<5Jes*&21Lu9R-?c2Z62Vna+tF+$PHhy7+}(m9XLTsPc~Jc`|%gE|U^VmOei zi6TPi#={q*iJ~pGCe&cFevF*6W+w`Zs+w2($EZwNZ6pW1C`}3npCv**&6JL)-WNe9 z7~+ce%9(4(SyL|4*FvwjPrlbTyH)6kfRJ@diL|q`sCA_=-6+Ah!S?A9pAOpRv}t;~ z)AHIC#H20tQWZ@p{}AZX{r%%0?Qwno+DfhPJwhh;2;R*{V~2&JIU4DZEk(>;`9E6X z&WJ-$nHQ&rI>SY+o#;pKax{lfjso82%BlIE*_5BW`7y#T9EgTndEA`HuBA`=Hh**o z?Jpu7Z3#6ykM54)k+@eN)CWWx!3ygGTjR8c8-PzzX8ZbVW4zl@%tS-elF%ilA1eEs z?`K3wT|D(ZUxxNfyRI(<5a$rX5M0B(Cgwy7olUwwAMy#HRwPgb(AwUk5jAI*X=q&wtUV(>TFU?JL=O0)e z13bHo5v;V*5hp8MMd~Qu3_EfRUW3Xu7UqDKGU1chdrWyA z)cmtCIc^5+0?}Q&W5W3Bz@4TJpk3)jEd?~V0^j zgQcH@6aWnrVD7=tv|p80fLWYg`Pb8$;d35%d3dI!(_q)I zBP6Dv+nnKz-T3wY_`tm;)Ca<$J`hm-Iucm?6y@6yGY=QJbMdw{{r&^X4qX#F1Oe7w z6d+`rRvs*+gYRO=rVW`NXWl%Q1z4UvMPf_6k%WHV!w@d4c_(se1-b28%q@PPzmkYJ zsk;AjC9o344vR$^;~B#}7*Q(24$wV9L2U?~)U%o~n>29nS|3@<=#XYJv(1jPH`azA zFfb17yq8G*!12xAdf6IgTsPtC5XZiPU{zgcTDrm^pMuD(@81kSk2)Uw)2O#*{rEeM!X)zPH)?; zmWCpGaAX;c@D^gZrn68oDNOP+;|cI}Kls$5=)nK^Hi$>Wp~t)oyO2Fq%DvsrK6!)& zzsD>`ingo5k!(41nD@DD`IEkk|9flq_xn##n?(rU9^pi^Pnb&X>B!!D7nZJPxDj3a zxjYxBd{tLFSsHsPGZyXJDE~QaCF8)2XK^AqNKonYfB1wG)F| z!>T`HuxB~c@H$WE$v8G}2iElNi_G1b_;4iF`)f-`WK40YVX=|#oLUwR#cYVb8Mhrs z1KK6VzKI`gw->EKd;>4;C%Z)wE zpgqF%gkRHaxtE`}6KP(1Xa7p7$u3JH^vow7qB>PRa_VIF(z{q*#>iZODJL~^E7L#* z2^qjfN5bx*U_{sG>4n$8_JPnIkqMyXL(cm<|5qOmWWM0|N`th@PQBBf2CXK;4I(^?=7Jk=Mp{(@V;qGL43<< z>5qorg-Ubul)mZ@-7=)~>+xRVEV~YO)ptFtRm3;zB3$Det@jFJn@pn4Xv0=B+0$_^ z1$5n=7d05-EF($u4^;T6`r7-$m%R}Hz8cLXwnY35uo=Pp!(4>_&1FuLBe|zLj4@Fq z=iP{wL7>TZ?wmIymGtW0QdnSzhEdo08^W=~01dAMkLK#@Q)xoxa_X0|`Lg_uyyR=B zwfj<2^6^WFR8fMwP(#yTx|S-vIwDCo&y3h|OIf3;{Tl{Xo))<&>UnMMiZ<_3OpRG- zL-GF9cV@V8=9zK@*Fd~hz|6l_r+-2#l1LrQHZicET<3DfdWoRZ4uUwS41L#@209!8!ie}!%aIdKCP(s8WQ7AuCnm6%S|=5-#87us+{p$rM7@p5L!=aaG4Fwx7Bili^#U= zyBsF8JbY-_hDoq2^yT`;mt$)SAcBhz&#MIKW{$qs1lLUUjS+0-8&i6KT zXt&=*ksjIoP-|j4d$Q_pJWU?;RmCe(??Up_qL_YpcT}WUgO2&p(s<}A!oqvvV^mX` zEb+uIGA9y9pu8j(MlNh^F@c+=vDfGsjgHX+XX1UmH?Paq+I#GJJ$zxtExM_?ohCWdd^6M6cPp2!?QA zX{g)w)auflPYsffd;~6Gu5i%$;fi=`3_VrkIU(K#rDeiJEF~u zA`&R=Fmb@S)aPIUrOxWT)96Us(`ZxoQX2JKLlSRHd)$1sd5o(DlIbkqubt6{4Ssft zO3e_b$E$|#zR`k&BXz{XlzNiwPv`#24xZs2>)2N&Rabv}wYtJY)#WgbuCOItCb|Bk z-Z~6*W7HLlWBTj`1n;3&W}|lHQ;WMVp*DrSovuX!bHi=1&Rb**)SU1{i#|yHQOe7> zGkvR2J{4lrx!?M5$WbS@m~m=_tjn2LJ`7zkYoizCa#Il`xOH0p^b;kd9KLJdF$=nI z(La#}gXC@=cEKoQO3xK}==G3u@o^wZ+@+OOE@Rxs&_-t|%E zH^Y0cLDw&Q;=+|&rCBC*aKny}B!DZv%Vp)q=wQqX3E`K5grnx>b7g3zky|JhT>@@w>nF`RPHWydxTwk& zT0u$tAdp?6mb-qNe3K=NXivfK_X(*%DKOM37BtopYPU73 z;65R?MojA<&}7);S_tHDBY^=5?5rzOQ8lo8IrcwklOAjK>#d4q$+)NGcN5uhKTee~ z&G0FY#3nK&C(qSvwI}a?LVGp$4R6?)x5j5zyjG8}nT@Lt_@V!i&~e}vm%dotrggwI z%!jKpEm4zEoCz*SGq=X6V`VnB#{6ybkWk?qX`&c*j8 z30>f@d-_`jB&g(AHZ3u>;MaNKj$>-PPxBIpYBzY2@kjD*2tmzg0`riQUAanp^+dnE z?94s)R-a4WPzz8)FN%5@{0KREq`!chJ#M^9(&h2vKQ3 z2mUpSH2kn=S`6lv#ftc`SF(gO%f2p~Axo0V$6aT87fv7dcoUkuIxQumsjqK8%X89m zTrm_M{>EZ7!}jQ|F_&1#a+`6cD zct;Vdr*Br;LJF5d+o!74x@Hwo{4h{vu0UdqfA3s!2TH5CJL@dSRWy|TU6@As5mZKf zbnbh>_;9j#*broDhbrr)>v(7lwn@|Dot>zWM`9WsGLPUCJcSbhzU})>I?p6LL-XAl zkfagtONUZIeY`#{gHCKdk_B2 zTZU`=L&6_!qU0h$4i_1eFqusU{H>2PJ$N1H9=)-o zwS_Hd8tLeMi&V&LDA@GUQKyRdqw??Flm#&5f3stSG(f}|oVAEmANj0Gs3hC0 zrE0>p0lwDI>TcIE@#MF-TBd+V*T^pgzt-ya#qzA`$@$j?UnTsJ4J+?qY_5E&2U~S`F-zZUpSrQ?xd@0Kw%UwU z?~Q8RtQIqdDR)&oo8Qj+f0&;WYJQ^Y*`~OAx9f6G2Yw1kxh@%mcko|qLO>5T3OJp# zYU!h{AZH!-VutCVl=FSX}87LN2bpM*PBeavkBOa|IhgjMUFdb z$^Xb>u9eSZNS6x=XL4P6-f+NrV(_0Pu&p((_HmEu*8ULsW`K;lV#{r4=Z)MHn3B;! z36gZJXovB!=nwNBh$xO?As$C~)t+P(BZL?d*Pj=NO(eZIL(!-At>y6T#bhU%pmEc9 zh2TB$D3xy6-DzWcVL;a#tA{LrWR6v88IV8a;h|>`Q=6_GlYSff*8>^|+4O1w4;Uxs zOOe{nn!Gf~U2Ww=AtY{H$=KBGKGlHj!bRK~@NwQf{@PjYYnGv{5F4A<4fzys>0yd1 zqqArNMt!8a5D=3Y#y?f0E;!x1z0Jo9~NWya+E2YXg5Kn8Oz~GAR?Od+a%Wy&`Yt zylBGi`WQ4<3WtL-IgAq&q$NZdI_s|sU-KoBHjM|Cs-CFZ&25I3?dWT?x4O}kb%EZi z1XNV}kWmq~oWa%JE!r{*s`+{orH#-Gf4yiq14^JeiMC<)DgEY{*y~-*YO=R7jKo|; zj{dmJ#@1h5?L=r&4$+TnE54Xj*k90REwV50%A8Q~>|Yy-@w2!oIVGg)tT=T(O|gir z5PfWd@h~D(Pak6S&dS8}PB!=Xb>RaKXGh;sA2sC%?y{;&nmbC1xV&8$0da-77BT8QL%E`*X_U zeRos79vt+Lz6#HdLfWO-V-|PWZfBtS((G(uDI2M&woN&HB1=2nRN7$`o;FuQxr&^{ zND0;hn}*qmx+qau$(}EfXPLsA4(Mqc8L86U?Ola2%1i24%2Ba{$EFZ6qi-cT3@AH7 zH@x~4L46oi{>`JYAmLLK7mN_C&*EXbhtc|af4IUQIj=uuDb{CGdP|x3&@cSL}Jg5;2qfb5k$nx%G4hE}IJyDHrfuci$5Prflmmxt@FLX%9+Mpwe!FZvTeOpYb!) zGuy<>NIa@Ev~}b)LfUuUa*U^O6p!=` z{T>`tIqX6)f0tFCQpC2n9jzAPR=H@?y91&9oiU`+wlbV}go;Jm{wOK|?GUgU6346$ zl)l}Pej;0UUlGp4m{xAmKvR3oGZ)yZ=*A7q*zztlmdRbukH(OCd1m0%AG-UJyT zMi;Y4n|TCV6s(H3uV>d5=7QwYnFBo1$^KL<8VaGwqH@Y#?Mc;q<)CdE@rT%TDdoul zU0e==Q>Jwetki?oQ;^KP$MY56jV9K(R&FgKF&)%sn!EwcjpO#(zRL(q$b@R9zB817-^Z{&x%1)C&Qgq)-7y-^*Siqu!tP}v8N9KW@tG3MjoGIU zNpzUmYV+`R&D>B|h&L_!lr-MtUamsPn+=sNn`yw^k56naI2Wozz#spItha89`tQ0x z=@5`^k(Ta~MoLn8=oq?FTDk?KyBlVP?(Xhx1nEXn`i%ab`;T+ZHE&_f{_a)#Gp|l* zkU?Z*SS>QoNsTOgFTBrpfd31+eFy|SMm0Qprt~+YIn9@_-LTPA4y6{TQI&V}^ba?Z z&U0~D|67{=KWu)K1p3wFwKJ4C4t~+a`a->lfe_S+a4h}ph7xr9_HFl3nEh)k4}&*E zo^f#cdqwdDA}=&B0-2X*f)qqryg| zndlVC&br@HHy5V`ldNjfv0&}mSW{!mRUWmN$3m4pc1z(SkQ%&A?^YKRl}=miQc}Y- zL>&Xq>(4|?RN{6$Ws*V+&{4Y$4lqVPc8<|Z1>WqcnOh?Fe225DL5e|TdVDj3qqo+m z4r8$zZ}I8=kEM7NJy?n>k$!GP`~*Im5h^NG2vS7%vu$9LlNWr_WJ}X{weB2zE(j?%=$01qWot*MKZk`jUow${ybp*ZSsgFo&b~TNWr*q z%*V4wLN+J1U;8r$4PYR*h-cjvsGd=pN(>2VtfmYx+E(JC%h<6P{XwK&$T!WVT7fF@u}%KI?ESiowB9SO_`4=ZORc;g|OX_34cQ3j@6 zuFBKEONB#zd!Jqrx~-R;_^6dsWSUHm(&z%}@n<%DAIHX-7?Y$%yq;R*kL8&%p}n^v9` zD~nenD3or&8TG!U9Hg6k7z9V`L|e~ilplRohZx5>aSXVXP`+M`IE3?bm)!oLcE*mX zDDkhlAi~6T*r&x39Ld)i+z-k<{7N~|u)StsTtXh6p2iZaS`;?p5t>8?q1i(HPE;mh3co~S5M^;KQs;B>HX@QAF`xv2Eo(&Cvjeg->%M@9kt9ufv(R+xaJ-UV2Yk8$JP7?u=MASaEAZ6 zo2vhZFn%I41`VZk;g81B1A7hL_078vzeK7kxo3(|Rchy))-L-oQV0*8kt(A ziF0$n8q=H8j9*_oL?aq?(|A@ps!4(AZ~|0GM-fo|7?a$UKb;Mqc#9^2iJRiDy5te6 zVbzCbf&Nw9V6#Z7y?FpahXSraVT#ir(H`xYZNkNNGP7f+UCwR{;wPYN%B@Es;AqLB zE9V}Six%bwt_Ph32rl_5@-qezzSDCrHw4&y+Z`$2uTVEj#`jBI--+9_b;g4(c*xYv z$iJT&D}Q)&_q&cQy9)S?PCtv~=d@@nlzOc}n&_C?w02R~G%mWI$2_zh z5G~6*hdHt(67#=mPU=>us*3LFK|i8acP9TIdH@F(6y{mdklGQLTg>T;RYEvz1`f$8&&BBDJ1Op7CV3Sq-G!r}qit9@G z6&&&-t$xqKoGc?!QWSof26x9T%YISMBqK`|hBhdMoF!@9U5V75_*-=qjJuK}smRRA|#tztq29bT|Jnd(Pph zl+T&Xz11H1oo1oQ$1B~}#5cwpr2z~iv5G~H_T?&H4au`x{a3A-=&BEkTsqLQ`M3|;!(XR&k;nk=e64oY5WjM1_NL(&Z0_jMObso80K@f?Np?{D=i z?xIQER@*q@1UYzzR3MejE0>F772e|J#HV=q(uWBJvB(nI(lxBFnCGi!;~fh$ql-KO zeytcBpq-cV->Ut>x)n{1?xfd1w79B~t@tZP|wRK{bP!xTJ9 zx5s1mn+Xc0R|Nj}&LW3t+sA4&!&YE}R7RgFUCHZ2F#ZbsRsvq3wlc{dbE|MQ(&u4W z#ctMVH~&#dY!!NO#8%a`W%y!p70HqB@?r~BM|Y3rt0=gU4Z4*%O_hsZ^ubgR{hs7$ z{B8jpC$o{H+;0_c)Q#{q^DnsIt-teBDB*X2-W0Odvh?_PX0lqzxN2B6i=;q5uxQr| z!5U42m;%&|i1yy@AFD(*^TrZimq_(QC45&FkzTz!d6$EfC@}H8LWS&-)(Ze3u1`-7 zfZBsUoV!vzN1Wq*fBrM|N&}jw{|fU8M57c@V8_gfXd)kalRr7Wz9wwE)&Ao!5a5$Y z!?FfZPgL6+)e!tl`YJlH9|Cde#sJ)#50l|#!X51@7Qp4g%WTTr`kuqS` zmx!y{P{5bF`(jKRRc!6i)2&~&gNwbIictzE-<$_rAcrMxG?=q((PK1}4 z*w9Q!dC04v2DGQ%5yJ{v9o%WNVGs{Oyht`RgWWhsBdR-hG zkcOtN*V&Q$Pvtk#4b41pPFV|~I1}zeL3zy7&FkdNes{mRr2EG2@SF9#Tyd$u+;A82 z3oP-+Z`OZ&kJ*AC1sd#p*~33hvr4Rx0k=JUv&)5l)`}sFZiR@>3>4u;wl(=(O?Cid zrJyE?g+rJJ9*^TEc+bA|Luqq3s4U;&M*fZ2)-LNhaO-sZuJw#38`JIWN)yn48=2Oka^uVtB zuJFZEfL=`flOFzeMNLRWi>leO3`A-oO>ZB~gJKYB0ko)-OK2SeX4?Jc$eQhjK#B`x z)}QbE2+zI^jOr*W;kDp<1~U9c?Id$W?jZkdbt59){py>fbWY8kF+T`qn-xStR1Ay) z=EF9l+HQWhKt0wim(byHYGw8D8dsbduOUPQYuDk;{wL!}Vwmdu74o=%%WYUf?~MxQb-|3S&LsJ zl&9XgJzR~j723kOOagQpZjSk}Afo`^PZ#CO@(Uz4$^)M{owrC^Ac`zO;C<#8-I8fU z!;o|IYf4zx^E??Wp()Nr=RW{fL2Ofwp=kx#0>nz?DjHHjB=xKiEUiN|us-T#IKo2Q zu&x>mN;}bw7_sQHz#{XC!*WDJp z^a_0`?_=K~8xOiazE=yo#$o8Q@$-@qg{sAahfRmg_V46Lg*x!uZgZui<9KYUmDqMzQVW-3AhHCzzVV0NRoqz1FM4@gRRYiv`xVtUo3yau8K~@ zLg%WyyIPN&R%@U{mo*Ghb1Shjvai`TCA5Btu#E2bN>K z3yR!r?%#MV-A=D!3!3y_;5S3BiK#5SjO1=wA3Sya@_wEw`X0ByIup!Ijwt5p zRaRDK|4b-~Zy!}Q^iIG+1!q_?6-WQ$J>eung@3~}7Cl#&f^?A>M<-cSs2U>6L^|TA zgfMlnjd1o*r*}G;aWSzZQ|cf8T^SC6qFc0%klm=K*v$g^W;F#7JK} z9J}N#@&EWYbMIfDwpnem$A5=3iGL;jEam&TbFH-1bxplW(4^B@Di8A@u|?iBWVI>v zem3mKxQB)xw?nsq`3x;o1qjN1=wx^nES!J6$ax)k?tOganH`S|*7*O!mCYYb#n=Pg zed6+m-6eDpO1o0g`b`>d_@ga7j7U&hc3u{@PYg&Tv9oN4Lpcz<>qlsqLBR4X{O5 zl8H#I;Q`elQhTU6P?gjVU;meJj81cZR5T-;04v!kFu&=4urKhCsi z%*_NeD`Z%N0y5gy5_9dnWRiHgZFM;Eb&=C&FN@rtIzja^q)CBbuHKRnf2BANd}^@d zmo22E?Wx8krEd_r*GKfQhv>)HlWJ6_2bS+qDt=VIK$uoV+q^cgk&;5D7}cy!M|=~w zz`VJzn)!Mat$?N#%S2-J$VP%=LW|={Dxld>q@fD@f4?ipUuiszJ$Y~-Da`WWcKhjT zBfDHyUSQ?V4)@9ix3y0t4yLhBy^9IU8qv~O>LawuaH%E;@uxtKL&VaqB0UH62y)I^ zxh}R?b>25lyvzxSF_JY^sJ`2G?E(8}IC%3YpF<9%!oG?`vvr>v-V3)^$=8e}Yq=7_ zESN>9)PD}GXmFmVQg!s-S|5y_k0sVzus7$`pvYU$ZM84oaB=+>k@|>ziP!Z0X1>z# zxkkKfjy#-k)f$5dTPt3kTw`FskJd>5AcG57S!8CKjvK6SuiN!PCSSvXQ*Rg6N3+wd{L*ssv!&D9+msr^#aVk`JodLvXv8BKS}q?Wm5s?C zJd7dNo%5gd#44bafE?AC=e;D|x%ZQ)e65dT0!5vz3T}FbPvFk`k@mwE@;jzY44cSc z=+PX4gr6<{LLhP@2uMn3Uxz!`D=^-egvKmzAF*3n+EkvtT+iApRQ1f42?%ZpR^86b z#M{ttfw^7eQnl`Vlm%i&>+!W}od@@mM%**s4{B2xHYnLICmtyf7^dd?}bs|y*E6wREw zs)&*o!|=_6BULW`cIiBoATL(+hru-O=`FM7kF{E)V2wC0nz#It*-5F0wC;O-I1*DI z>Y1h`raBf5lPtcgoq!M>J7osd+MjpV@(o(9R7mCKKb$o}|L;g)EXfy^V*1icDybvY z#$}(GY*EAZS$1Hoow$gOZNucRHX(e%RV-YmbNBNx*%# z-}^H@f$Mf#-kRk=Dm3)ms?O5HmTY5hQf7O1?z%{W186LG)sp)^Ikd|H8Ue=Gw8oWQb!~{cub3~%VJUm9{;pI0F8?E~i5drxsbG2(7bZSHYbN;gqv+ui zeS^~&%jus6ea+M;V{8E2>Wm>Y`jzWonsn_+Ws{nMhlu{${BTIvK*SNlT+iw$-r3p; zd^E$D!=|O)nujK7a-T{nc4}jq{X-FTm`6F9Qf7m~MOrN@GuzcNa6bvEnDc z5dtIsb+?aP{rwz|rte_U{DbDUyshWuMG~@vXVD+S)|4jiUK8nxz=v!BYL?fBQCexu zlJi47o~sQ~YSHW0t@(gv}Q%mzCcpHIR9jE>~z{Z?)?5bohT zgpMiMjCch}B^m|%#FZOk zK0KOs)x-d(!M5pC{iy_jJmU$$D~(>#h#z63 zwtz;?e0q2oYiOh7DS$NK`SaLjY!>a}D&yzdcw5@$x0Ni0XUag4hQGTrjwUvMswfy}V0%XalHLPpr*JqdIwuRpqbfZ4K9R`A=;Bj@B$bEj zb5JPJ9+oFmm&gh^P;oe(n3++Sg9)5&-q%!*=LS+Oss#y>>nL}Y9Ot{g14C?XZoKB_ zDM>4xafwjDVQwjXx6=RF;Qvy4wLoU?AXTI|6xwEVv_EoABzCXQm~3l%6C!k}s5CTi zcyWSVpGl8H3vZm&Op~Nt-m)Kr(m>D33p4^W*p1D)KzK>Dmb=T{^M)|Jv7^opaLkjQ z+nR@8m~eakcMCwAbReX7+6;%m)7})N@aSENH;^uMpYV%bc&t-?j3lnp2hW`eWt#ci zpIim2(z+lXt-gBWu)|yAv8tnHoPrj)A?iDNG`+gLBNxp9vQ0QPt`qT zFok$vO)t=y=P8dn?8Xjy?t!QGyGhME3899&pG{vLgNTRa{CZs378lnp9}MG)0`z?) zvOuv>VNk6fZSa-0)2tn4xT(g0A{&Xa(g^)|d{KTIJC$$feS$#&Y&+hcRYGL@n2?TI z+pp(j4eMtrlMnRA3YAub4g~A7Mc;K-|EwOX`Yv2SRUODtOj5z$_|=D>gXqaTRb31iFa1b>0|R8IT5jZoFc)+}6nYablBPzhU%QnFdTQ|2)pq{pkBoTbv^ae65 zw-<$a?qZl3ugtStf6cSACkU+<5N>n!$Pvk(y&9Zczh_n{%84-4ooJ?Oy7jHmbBcDJ zMw%H>ED1b2XGVI|r%rv)#pbGM0SO};QwxXWCyEO!n7Eo?)Q{jU4-tc;LvM+P`AIl_ zsnRKkMy(<2u<)(e*E{twGnMNXsWqK2RyQ$6+VG`x^lQlI6Sq;MYnatS5|wMgGxTxE zTB}LiLnIBH*J{B@LyTlsm!Irq1;WnPu*svZF{JQpAZ87}_nA@J<(jFQ)a`x$(z)qP z!*s!kHa+}y6J3YoHA5i>shXp=Y#R;ny-%3Nda0$C)8hI2mPKSNjiVO( znSFX;HH_CMuwRo()?L;|9Kc=qZ>^u6|5UQoRD7sPCi^Izn;EmC_3^oM)m^m4Z5}+1 z*;S4cjN}pfQVA$;DM~WaTbZZe45rigVc06W#6r=A+H-e<&~|=Yf9}FVm&LSnuWqyw z?@=?Hph4A7pA&;^qRvcco_@o~eP|(o+L$3`!Y{rdp1$qxXm?oEZbT2LxAEr_8KY1%idp*fq-G;;%?VUKLO(aaE-7 zbHadgj^I1q->x=oZ3Hz5)O|cZdPjzG#r3g#yRW_#>#i)*lR=9}!D@G-Hv;xMj-#l<91Zi8OcVwC z^W@gI^Q-#dGHZgMb9|0_tjfxk(tRe447(zCnU2A9y(Vas|10+0f63>3q&@Y2VIpH@ zH0F_|nT4)IHx&4&i95m1XJMKSzf0D6O%$XpLsERGAPb8N%vnPe(dpACAl5VHm1oM} zNc0tic$hDec!;X9NDy@BO{>&7GerISqMEj-{e+MjM7VmxnFSohBW^)m&V#5d1a=j7 z&M#Nr;w)ltiCNBH)mU4_P~!%L5Suj+v-VlnVTgXm5%{#Yw0YOq=O9INq% zYd>ZY-o9Ej6apn4jBBw&7lv3WMXF{rk)ME^cWS*h^78XL?8`MVa8Nn$iv%{@%>*Ox1YSEt*) z97;19h)B?kNskjCiG^Jh(3b>f2W$Hd_&mzqjkxfb+9)xNp zVAhZN3rd^>_zg6lT7Zk;7S)9iF)=DwZ~d~G<3BQnsrC<@M$Bk3sNB{}u?Mm=V=8a! zau1I0yw7&`8dgP*Qn9}Ub}tyw?2}OQYicIT4305wP+3<$zKs;ULzcKDrj)U2z+xmY zVox?pO?Q$Dq2?TxL7#oHXa&QcWE28C8{8CC8_?T1H&c2Kw-m4~-WHcA4Q{Y|3LIbi zRKiDpNqX$ta;d9+@{l+yE}*>VT357;~bw<;g25$Rfy z6=-nyfeU0xzbXJ&IZ8mF&ci{}Cg}$|J(S{!05O>4N2>`&jQyrWzn=C&1EFLYH?%ko z)q1WPPq!(gon1fdQy+2O3@bu~g-qlgIrox>bh~uSl+OG&gX0a} zwLcqw-kGf*>xYBi%TD4pr>#4d;8J=V?(7Q5bZ#F~c| zs^T-rU;FSMy`qA7tzNBrdQk<_1)_lPeb9yr~FaWxtB-glfjK0PDv{+ zCjmHE`ka62TTPLNZ6u%qBGgwM^z1o$et`y)u25E4*RVWWtWz5MKOj0E`_=1&bY0k- z>1PWVS{T^7Q_`m?mgGG5wwyGuKAdNBYe!}fmzF^ltwhdW_Wxr_->jj5ZWn?mv*j-} zDRoKJ4VIxSg($^wC}6_#hV=|gZ)Tbmu7WdYq@#J+$ZZ~^J0Om~*Fg8ix^7{bNML4OJtk)bhRk8+i+tLJUYPeKS(~B>mLCqBzm8oyf+Oj{duW7Jj zR*mu@-5q^|h3}|h<=YA4&1aGK`Lgy%%~Xlh)Y|6d=%$Afk8fHCial>#w~9uuKko-B1q!Eglj5{Z8|bDB?g|(G&MIazEe09~2eRMr@QfoX=ZFmz=fWWb{=l ztTb`0UO5Bo9XWOcCAmWxm+Wx9F*dz}?O^1PX(?yapN~)IpnP(GA+x1T5v|mh+zNaw zP;(QHtry42zCeyU)w!vXC%*tG5K%`^^>@RElYUU~fR=Y~(7okr0SeP*O~p@b=+;Lv z)w9WkElVO~&N{^=vxQtiJdJWTP))&*N^4PDm{>jtQw_f;yaojP#%62Aj(>$?MP{nU z(6o78!3SXk9rxVRmBi|E??^X{Bc|05#z8Ad<;AFAi3Q~6Rn0J9p#E*5FEUOU)6C>4 z@o~8CUFnzA=iCLsQpQ@Tqr0+~Yo&8R`zuSM!g-eNm9YkX7NQu=J3hk(y-1#LiSS^J zDap0}xDK?oSpH90?bS*y3%jAl8eZqf2KM{rq^XK_EJH8H-HkhRjBR2MEMCsU>+0&e z%zKf+TW^N;S&f6TZz-F5^2hfv9FrU<>rLwE2Jls}B)`(4E@@>(3-fjo+|<&QJl+g#V2vdm{9W!FZ+!wq)>`rc&J(UGy9^ct1R=e&ZSD zA}5Pa)O|Pkb-iNYrv^XCB(# zMsL!le9^2JNKAeB85GX;8qZJBflACtDU~=u2r6$m9Om!Gg^0+T-aJ%Bd=#0U|7l`r zXIA#!u#$!p*-hgP67%M#FaAeCAZ?2Os55FQF#vkh02rDy)Js)SmocjJ`&$SrfgG>toZTBUz7EwFbHQ>)mG~u)Dy*4||2u#g=out6W_1a>ltvz*0 z9Dt@3b9Uta`2eYC|CbA3Mj=LFQX$CyU`lb=iX8hQ8;s8dshi@GPZQJ1`sHOBG09o* zXkD?oI7ab{!&%CaOrp-nOtC&meedH^0QB*-(tZus0bQ-^wciYb(}gLr^b+cY>aaQI zBtDfs#B*|k6(uS8ory@saq3^t$ARMMOFKhpLNr`G+x(SIoKu|Nmn6V5%Z>Bs*i_8u zx%L}JQ^7#7pDRy1@P)N@03SRcA9#6~SkWf!Z!IrEJj`4BCTMjfx;1Y04Ii?A(4H6h zX;aYY;?Big1iHoW+}axj;nKjNE5%d1yIRjG!DS6J>nc#5B1gmxsg-OJ>t?h$CRWHaPtP%*_FUIy&nt&xIK^WMHmu=V@`XNW{Ym1we04W&YEn}M~t)2Zn4yVYkY z9inrZL6?rd5qDZFV9{u2^P=nP&}NQ+0rwM|>jc{r?XP2ihBwpFi6KONtrox2SA}Al zZTQwG|K_bki%0D$*EyokgTZsK%k34J;pByZDS7^xdkpAoG&;E+>SmkZ_hqbr^??Pw@H?rI&@fMg zbgxY`HnyZVdAGUTPX|VqcVBbkLRJoNFR?!1Z}`g()>a1V5Q6>V%x&;+fvRZO8EfCD zAHz+sElH9}UGt<1C=~;*DGiW1+VN;3QNSjqsVoXM)BTy&aYMA^Lng~l)yTQuWhQh5 zcCfYU{o}rWE9$4D(HJaWNYmGN*9=l1Hds0)!B|ehUbCz)!V?Xov34dwaEe0#3pklf41})=dspUn+}MKlxll7G|$}|5mhyF0xA@P&)va2`!M1C zJ{Bxm#Tikzk%Uypkn`ZuE}7*0i4>1_U_}7;O(u$UxS^CTUTJ2B?)~Y<`^){gqA=kc z!&LtJ96U?!{iCf5+b=j-ktks!WxNK21BbZUJT7IbthyOyx(Ka>k*!z0m67(`rV8*a zsSK+8&XXC)yuEVApRM$nGNAE-`4&1zaDVR!BdNlNOGcJ=h+92MQY)y!SLzpiFY*S5raf0&?j{)iAr z`*i*K(&nH^ZIN;~MpoSMxwE=nI-(wAZI8-ashz>0Hz$Vu&S#0ZnXuJpllPoi#u<6L zjur4tiNi~}R9Ht`+$P)MjNBTb?(zKPzOe$p&`%7w`$}fa^FRN}z4TwTgFV^KG!Ki` zs0Suq+mi~S)%%@TjcvnzJ2aS>2SRR4{o&K1hP3kn3C04GF)J2QrX|I_$=d?K@+OA6 z9^zCGy!?U5czC@&@x%$uf;mNMA|TYIOudIb(KSq^SS131tCZtfx*(49BEJ^1o0`zG z#CZ@lpdd{ndkWh$o|16Qt~80t!yGlG{>auKi9W?gwV3#qfTHf$dkyS5px#Vil%{hP z1D9c$!P)j~8fGoX+U!>me3vj|sFvEQ&4CYUkK;K0h$0DVygYt*RXh-q(nfO|C&1yY ztZ#Dp1!o3{Yb@>wx^qwC0??We-qmZ*+CJh7rRy=z@R~9?>mQMJnk=lFpfE+9zWYn` zuVl0ieKMqTUXGl4S+~G$!YhJD4`Cgjqy~56b;1dj<5Jq*L`Slwfs~XSWB43*v zU8r}Hg~@2GR|lH>cB%ZgDM$^&ArT_Y)s7F$gIBrei>XhsV6Emt3%w!*yF}YOdb?<= zD1D0AUGUL6Iznv@Kr#eKF4LGffv_>%J(tX4=5kvmyZ*wm1zH%avqlvn&*exSvPg)J z3wftS4@Pd!%rr60?O=h3{ojzutU?ujSx*4HVX~a!*5s?whzRc(iK;<6kHR3Gg z?$kAx-n(Ww)VqaP3E(rE@WPj_O)3ad@7v!!%aB%=o|T%U2Jf1-Z5`TdP2N)UD1%;} z+e=){iV8>JuKi0rEOSj-R@k6*PJ%?6UqKAn`iC-ZZ-O;0zQQbs@^?2Jk~I~qYubN; z7Mt;?Mt;P3dOiy-)IP%v8SIV>cc^gSl09DTKmHX@HN}GG!C4iLTM~B*7Nr)4Z@`ab zc!X@VLGR z`Pzoc;zH!igL=BQMn5}~<_L_k6C)bYbQJvDXaB@kt1X4Pw`19tg57oTdBvvV&4%ERFo{$w=QoqHeX=_)!`cqk z1ucVm-Pa=T1Rbf3ZiDNrIUH@1ZorI1oL0gULO}my83LeF@AeL1H_b^|8$41zDw9oN z>qUe6bIQVq+2i{R@H^Q9C&(@S*B=f(w2uT?-=ZTJFGJtm`tP@0Jf3VN$N8H|XL-2| zk9(xSa*B5uNEC2wjg{GE89#wxI`5tg=9bwC5;9uv>U1&>Tc@ zf})z+e2H%l+i6<0gwy{w$)(BxB?(Y(>Kzs2F43p*Y8T%s|(k z(h2BMiPk&&^D4Wu&xsY`PI_j~(BsVZ=JnFf=|8fWD{W$Y+5XYW#gjSkq z%iO^!;PCdFDIRMkr;BAVJCjl3)7XKyG@7+)z5#KSa}M2%lAj_9NE$fA!5QbvG3F@O zu<)YREN0uZvO=v1;*+JuLJW{>ZUEpR)tL*E#k3^K@FO9Z9r#Kr<0{&^;ru?`h}A#` zx&~&V5p4{vQ@b}k$JLk^Sc++_7$i>x+Te-<6@{`&^P0He?4 zQndLap)M*_%FT`)YtI;VMue>J!{k++aXK9m=ZSCu2Vag7sLJ*8oW8foXfd1~6z;b? zIyhl8VIuGS7BN}WYp&n2F0V{mz6rLjwh6yl49j=Mx%BUKW$>+ks9b-&`qgDlzPmG% zX*B?I@`Atg%%7uEWOODyNfVCDWhGW?);cEUg<*9QujDx)!-T=Ukh|Yr@T0l9o>1b} z_m|$4H}##vOYQlZ67uU~Gj7f~I}RX6>8^V3zfHJ#;qM+iY0o^`l&E4RjZ5$r2=&#O zF?;XU!z7@tpJJn*M<1H310p^}jCHI=y#SYwX%{gNF|HdxJH+Q}98fmogtC_^bs%>0 zL&A9>N~n|VvVQ@R|JE)DyR2+C{t+Q)T>87Qf^}#TvSd>VmGJwYPPKfdK1r}#snWcW zWdI?>>#c{ND>N!E^%TjDUB|$TxAme{am2O1zCoG+jg?j5Ns%()EHkp{i(!Ol4Z9+<-uOKX{$WiF zrd#I?p0_!R$!jOVM$+HemG!wNB$}XtBrEWCIAcTG1e3VHq8sjHC)hu&C5i@5L&8d| z_62&p;_IO;kFV#ad8^0hhb0o9Ioq3g1#)~JSzeCKqH=tY6ee9XtD#V38Xs`J%KtZN z>+es9^m;D*XLF0yI_6TWH{-ojB!V`jlH*m~^GD&L4M`>Fo zO;;^t(q#^B3G4*iLXA&(>{)G$QCtQ|mU}KBoxU7%O)IRBh2|uBs_NiTDRn42Y%zL4 zSm}xWwO!=X2(WOQ{}lmFB}J&vZa!f2_*BZ`*Uw{``i^V;37u)YLJk{X-InfDyq>4; z7MOqvOd6}GG0|5VIfT;N+cRBvp_5gZdfbY8 zbdOJPTJo|qWvZ_LRlna=Y4@0f9*lSuJ!@#pqJK-c2})OaP{3m^U{%jSxAFIHsSVFB z*cvrkZZRwNjgyb4?HQ=+D|IVYKSdwi)?LwF7wc+sXj!4il{&4S`474Gi~o>o$Lzh0 zQzMBYBXNns86U!F#tG$co!&dX9`%QAm~Y3c)YALKp^IV@wTi&~9HF)utISePT2($X zS8CY?HHjU)3X&k_u>af&y>ueO`)NtzvZpOaqk%#cFe1AAs8ZWPIpa{c=?9H+q3ep< zkyuT{A;$C}qzDFNYPIr26|ao$bYsH=)sT7dsWO;fW^kjH^IZ&ApJSfl?W_CXR{nIc zEhdlk4+Rng5Cjz+_gG9z>%_VBDR7`LWu%1|*zOdMG-2m!?BdWM40nAl5IZd}!4cEq z$obyVWFC|)_;4|BBMreE$T_L#hcdXIpR{}qzbDCPA=;Un39hY4 zZ>N81Pha;YNWi1sI_h2uBi;CGm#4UjI?KC;+Z3^|G-Q5d%2CsWTK$y4aU13Pt1vvgG123o0YdjQISN8(<~|z!?z`VXXGZt zcW(vE1+^-uc$a5VDw>~2@eDbJL~6gxQki;)V!{s=Zp<9|^>4l(|1nxs<3OMOz_LE> zci(1qJifv1tsoMF^tOfi?yeHETys8nWeic^U0Xs6d~(dv$>u{$fYpMN!6LTP`L}jJ zfB9kOq(TA}FeyzQ8!_Ua?v={+SM_>X8@OZ+o?ozs?j~8Q*uW4DW_8A%5k0t?A4yTP z7P}uI@DAc?=JDN8PV=9F*D{CZGof86)xABXP#S8tP?8%I zE7zV5_v(^MN*4;q`)%D0B%lV+u9T5HP6HqU_h(t#fIyaC5Xe(5Wj)`Y+frO~`%ru) ziLC$%{1lY9jS*5lsX;$$zJSD3Y4lXVWtj zc-8mdd14leEh7VH!}Z7seB^so_r=gT!J8jV$G3?Sk8--xZ_(-et(NY#f?s}zW5D#Y z#GQ^LJ{Q73*0Tjb6KbAbA%@Re82tUtTNU})qrW$uPWTeU#eE^)#Mc&sCn9oyzJ|%R z(XyDKYhPVj4-)E_8v~V$Rq4>Q1LfG47>M+Pnz)xBCG{XA_zWQJnhjZP6;&%bQxzzh znKP4+rrFy2wA)`xO@`4l}QzB{~L zQe;c(aO67l%+yAvd9V1LUu6jyMbU{O+AxCo8;^nEY>%3 z{%qxJHt?OMlo3?|Yzd3KG$NTwxPUbb-r4UgB{9|Bx9)zXT^o7hC~By5^d$>kM>^n6 zCp6mT9YE014*gC@l>qUR{32m*Lc*r(E#e)aVVxfMd+S4H9h1%Y)Z~4Y`^n4Jc;2{a z!Rn;KRBPUQdBnDlfu%Dts5i5qB&QRA01sa%V^szM~FSjP^2aN2`Z7;Lc0R_HpCcc$ly``F1+-SkX;#`c1m`(?J}@5l)BjZW_a#pKy3GZ$&B+7C??P5Q!8q ze<%hq&+&y2{WU~g(~dTw5d-JjTI!wgdDG5ES5^}>7w*%h8(GElRinri=Of#r0Jqbq@8>f)+Xg_43tZO2P*X8;^ogn4{GnBZO?izYf6)2UZMQeyf%Pf- z43LA3z2|G$mOdN-XquK{WK=yBliIWUc{w3e*Cwf)Z8#Cmfj;WF=Xvs5CfAbl?UkVW zmrmLKNtdZe;bx?#XY-7XyQTWoF#A?T!~BWIZ{A!}Q-7}j&;1sCKK4r|AluGII#c{N zb9a6IVp!!T`Pxqs0;P);l#8ZjpH1HZvRs2V*0BlfPD#sm^072AybE0x@6R#1TGvHP z(YPQjklMl<(>Xeu+=TI9PGba!Iebo=pzYQ@w+{NH!p|?+%Ap|%5~pa*E*NlKPNl zE%#T_`UYM5w|W6}I_?0Q#_T}9zqKP9Te0`XxWsONjB6EQTb3`W0z&Eg6h`UNjPrmq2w zwZFDnIGNA;7!W^0R$~FLe{P3d)5fp)WteX6NlIHrSN6zPyd1C|Dq|KU|J~COc@Y@M z{J)&azb2PxV>NR)kh}}&Lph%#x<`0C@V7ZSXY=FZ)?m(BMmgEjFRoUkSgN2h)iU;8 z5U@bdjNOFPJ6CNFWC(4&!uM z@UW8rYuRit>f=bI&&V7)E5j)C2{~qLZQ7h?J&E!Z|L!*bj?AFV*$liMl+KqhEP#Z; zkEJozsAOE_E7fB{W`XB5{Cs(_%2IS6i>xZu*XoIM4Zr(u5h}m-&#%xKE^$<@ox=^e zJEi0PAF|$pDb8q#)@E=gKyZS)hhV`yNN^h*g1fuB2X}WH+}+)6NN{&2K+t@XoO8db zdv5Jt0MoVKwR^38`u?rZwTzw+v#RIUq)PnFQzD-nK^Eah))vpFU`N|N_)hVa)!2Jh zW=*@0x5qzk4$EEJ`xR~QN+UUQ|6rt%6o#B~j2|xq{~2T8&BNiX1nNAVyN7$?IVkFF z)*M(u#QT%iY99`jrEnZJR~R>ctBDC)8<(Gm^a4ihub5V_uH>NJEd~y#X4nOltfk`A zD0NyMT?MT!_bloxxe_dI!9;?|i(EkesQZ^ZrZ0>yp(_LYfe3h4U5g;@>9Lfl!Lz%7#cQDAC>^H z_~&~--G~f8&`Menoyfc&p=HnxjShb_Imwig2|z}-ezx#fBx@RUQNqV;r*w23M(RAn z&Er;}4qFL0?hivgfg1H9{@oyN?=Srfe9Q6GL7M=W$sA6bsfx@OriO*c$7FHcOmHKX z1!qx1%SXhAMdGQ7ll!NOsXnE^YrE2&|6Ik>oH51!8A#;k`DfZj*P*U(ER2;r9)ahh zG}1NRVa74A0( zUdedr3{~ZQbwnlKl9brY!4l=NY|)Z-L5ig5WgATu7itNQCkLjORE9XC-4$tX=q>(p zkNGk0b6Gt0`g|BQCgUZF!Hq!X9UV$h08G!O!?V}rb+Lk$@oqtz(M449zJ9yoEsALj z#9`6X`1v@^YgpWIC-mw$mJFO>ei5uM>^e^PQIq2MzX~oKJK`?$}K6+$c;4=vF+(;`{^;e;Cb#5UAf? zbDr&Iixm_%v(NvPcFF%;+WNR`t$Rv;Sm=rk28Mcl18P05ZCATiR)z#xI#K+mA>?g? z%cUh8cqHb47KJs%86tQW7kiUsGkuormxdrmkd6YXra%#%-m+)i!H#xF;3w=~mRMy4 z5&Q{}Z1QAdF9H0Nhy^G3_)^(VfrIgPTT#@MFzY|?ewv=nif&1vTcK`tQ$@Ar))W~g z^qH#Va}&J_J7-S_Lm({T)WmIr?_XbX+~m~WS{?6YW_7j*vah#B00qh3@)z1<l-`*=5-&!*OWV=N9K)pTvwrS%@3E|_du~( zfA|g`9^zKN{=$HbuCmjv(>&X=4=}WzHaQY4lL`Ofbyi;%0e%0|rR_AJFpH{t^vj%DN{3?FN_}73n8t%R z$jRiKyIR>VbX%u1`Xw0RoN3ZPbg*}uj!TcbPAhR~Ng1+g`UvmNNmJnq2BLm0#2 zp~iFXdADPYyxb`ha%cLJL}TZZR%nZ*z}gCtZ_H{@w#n&c6gTzv--O{COk=v%Y&%Fd zFFQ0=-)&iTZRkvkjQ$0T>;4Z_4F!L#mQXZ1@G`iY!+3yFB~%5l>g)Rq=(ij>4+>a1 zB*Fe^4(TSYP4)Z* z8}!uZD#bfS(O-doQ%3yua%Nah7W?y^_1#&*nwYWCHnkmT_&aDJOT{&~wrQEqp9z;a zBQl!J*iQ{*p2ilWW;6sVK2IzdBymbd?1j>uCj4Det?IcodiFmZylEmr#3&6)w>>{2RRViD%Had7e( z>i;o_+Qm~>yX%^!^BK?cAZyeU0Jk)K}<#W>Svx`>)3Z2uK)HR zEZOL<1NHvZelN^FbOIy+LE`shC-P2#gW3T0SK%kV4ulJ8!C(69;|tzLL0WiG651mb zT9t+>B5wx7V>-!oppaaX#bQX3!D(irN<6YL0Yx~TK}P@Ec|`(2%pR?%gOC3_lm_hp zz={S39zO&>avnK%UEn{5v*GLK*G>w9-tv$kKNUbvTRPpXDc^00Lu&Pa(Y#q{6Pu^kMtgl?Q z3w8c<5xQu;rW%jpl4|lm{!2SPhPayxjaa)EkaRPop#%})|YqSqEnAlQ|Ztcv%rVQIO)FL$bOMM3B*1`k1 z$xDu!&(t!L0&qoMfJA|IW-l(r=Eu_QQY|ZwYkYuvHE_5GiiC=+m;p6Z+f;Y$PI9#= zl2W7YhCh4J$Ur87(>t>YXITgSQl0Q9($l$aQgU{SO~*=s8dSm=dswoMO|twYL0WN# zvvB;rn6mBQz#Wr+roBsG&5`unhUWAX3r)RjQC-Jqw+0{hgRFxiJNNGxd*#QP;N(l} ztY^=vgYKVYCbm#&z8AkTo5ZJ4eIX%eR@aX1KWHD;rXt(ow0QKb=I%X88$P@Kagz_EgY#)A+EAfn#i18E~A<;6{ zZTlA7SFL!eVPUna1WSdgL@rK2Fj0C6bDS{vcadBQ*H{4v*W0&;)Y&ZEdrrDSTI=3f zYd^ED@>7M9mGYQNw7l|nL(65$W1orVW^IyJ> z{iD;<)qFqJOYya$JRM9gaAlZV>ff|AD$FtMoBocMggUCn73@3>)UTAc-9_Et zeX=-OzO${l+#va-i`U}kIAk0(bvpO~?lt=EMyWy3YG2}^QO!X?gW*2$!=JOuvI~nL z+}iI5LGjNWxk9ZFFBT`HZ5NMVL`Bo8Wn2%h~_q6hBR&1WPai;-xO5G{PPLvmc>g1XKq0wBp;9xpXam^ z`s_4YeQ*BGT~-6s22;QT%9eAcR3>V4{q6Xd^_zM!yb#9_FX#(?L9kXSp*Y+d4SL3e zV1_`Ep%F^+fmNXNNT6=>r_}*)O7R(>KR;F=T9dKGZ@?<=F<(H z=%J3jVH6;3Ab3w}KQ&^i>7c;`L9&M^Z6a2RK}c*b(MkM45Tt9R&=aHaAX&xzwfpk-QQ z#KvS|Z8@BRSh;J6Fb!w)yR{z#qXTGezi{X}Pr-jZ+AD!6#QGEb+g=~iDPkq>u{8ZN zcCQUwOqJBxkr`NvA5Bw@>Gq-d(WDQ^lcfZ-9VZ6{b}#7>?6UXDa#@MrXm>UT@YKKN z5gDf@TZ>pUCK7AwebA=T9wVM-&?C=+?0fogR`IsC2n4G-nyA&9Ug4Y|Pt_*|)XuNy zw^DpG>Kv8&*?ijAa-HjE?ApXrGb9rMm&;9t>=%L}39ezMm_!~M(txDgK!jsYLoZR6 zX}bHHWF|Mh@z!Bo`OgdLsV)^SM=qx%mr7dC7Pr)zki7uU=erq|*B#jDE~FxNmK?97 zdKayJe>|7LNA)ckxt1R`!>&>nO6 z;^W0nqN4VZgX%4X^@Q9-f^pc~`8zC8(H3G%{`4=1Vj_uPZRIY3o)XQXcGOBuugrC+ zsx;{|{Bcq7l*8vg`5XB^Ul{Hb#EI?PQJE7VS6t`f11_Dwx>#|oQdHMfg8Z~uF?1Ui-R^e$3Did<(NC^Tr^vF`pl~ByS{gft zC!48jyF75w-jr%azjR;u@%cHd_0PhmW6JS%k8{YvkC7v$ga4Q; z5XeZ{zbJek5FK8Q&1zDzi8ZETQ7+G7o3f?J>*(#b=fwF3VuL-AadR$+FIqLts~&y7 zn4t(=N6}D%txbvXeDV*TFdTLT8X=~}49K*IGJx=uNxq1OpvsfwW=`* zyKR@CGpD5ZI6{~A44wl3sQmq%&f>otEmV)kC9*6(%ruyXV znQJUlA)c}!Q`l*IDqCv&U1E=Z16lP+HH46}0y8@bzl%MA`Tf)fgJ<)JyH%gY{8ap` zgJS8X$Kpzk9+SGm_8Km=#m-9zaDTcW9`sYK=bP(N!_rxMoNCXn8?oirq}J70JF3@D zZI@n|`t&^-f{52sGE+V-Bh{LV>Qfaf8<(wr&T}_z1iyF7`egEgKSQDy8Mi3-Lp@Ia z1}>9R;81PT6h-BzL0MiMO_zF|tB%@U2+ev){Fvq`3IuDqaDI%rrhm*O6W0L9(~pP2 zKBbwby9*!38U`y`+-n<6f?y55n6K4H!@&(Y<@IkhVxPP#;lz@i0nDS3l?N}GFdjWZ zt=#XY@l%BCJ!P69r8NMdSc81lDSj^>G1?y-yIh{O(XX%S7q@liS5@l|UtZ@xpY$7~5Xi-d=f`bNtyKhlz{~@J+&%9y%U}{89d= z&$h-kiqOgwWj#AZ=Onwp z+WMNji<^_$b7e&`R8=F6L7qw7yPF7jML;pgMMJZvzui!IJmCSbhp1_a)Cp5G9@8-k z&lcfJFYtlCF8P)aQ{knFe6sb*0AR2_1i)bth|kL1?e~eUEK0skGEA!A?Pz6>(i>QS=r@1C^qsf$jK_Xi6V{`FtKT78zxT5S zQ=80gij}o5e0N|>N-z&^A|Bab3K6wI@)IqYX)s&)cgnnOUMVQaUM@yjllhWvb8Mrs zuJJjN)A~0-mx))hQkuADXTd`y$@v7JN79Kx(iq-;Y^$(5Kx)Br0PcJt-MLzeN*KFF z3hVKkL+UDdmhZAajq_IZ+jJtAur{%VDFCD^xYvk1gk2E}#(blTCDnEB zaZ60RSP@@smzgSIO_5Loce$opp*BIu`Qvcl0EX{6IO_GP-_Yo;_vkV;Wk5-RoQv_z zUH3=^>A!t<$9xn-@nVvjN!!tw?{sU(VfOY^vTRW@u=Y(FjNN+o5*YNT8fjS6Wznh* zGg=`tPh~a8pP4FYR8ASj7kuVpc&!l8mYDP4o^s)@_eKGfI&{bzr`#>NPKV{EvRvfZ z8Olr=y015>)jB<5<~#khotKlmxB&Qk7`2vxAs|VSSAAz7jW0j6I-bvikxxAI-$Rfuj0<-d zf--S$P0XGLubW~OO{TAjtBx4{`9{hW&mL?+if?WbMTq^m&+}NO z_}q4B#kMQK+(}uA;Cy7|>AAYoH>y)StvE4QQSGZ0L|ZL?O>0ak6kIm=KN(3w5 zA7x;e)=>d~`{N1%7jQO-L~$0n1;{^by6BP#Z=MA>smKSz@XsSMwM0Uk%e&@Sq*Ehv z{P4%oG+qUMndRHM;MHlA(yT4GGU^EpTD>yZ9B4`4T1;}(`l_`woug+!IRdrzB7q}y z$rV<%yW<$WO)Q1ZTBt2QHBtjMnGnY(Xs9sKB07SgKtq{{hKA)i?*lIV-0Wsxf0=Ut zX}7jP0#NsQn!$~$^Su#HnCh6?#1|q}4E~mGMhSOQqj>3)tw@Km8X%VwWY4)hROZL`1CmI^9Tm$dV>O z80E3RheY^trKA{F{0F1a+@vA+D&1x1+7l!yaaT+M6lf5OPuTrNxc^cnznI;}%lhjb zS{F6{$Ev2t=NkwXOh!E>2%& z1}1DF5R5nhK}JkvZfk3lkGLa!{vfKep6YP8546HpfjPm9IW|8ewPHzfs(F_Qr~2}A4q6?p8Ib@ z2&dJ3yfB2yn^gR(D{ewhQ=hnJ3+N`yweD%MW%qSy0H2pY$A++=lA>?~lmcPg#|sX6 zys01}VDeRIfFG{P5l)FB;x{(GI!KY8e|o_0tssVCRKXqSw?so>6x4cxH*W>|!a`AT z|GZ6;9U?&*Ph`vAc{vjk%^n`UR*HvC_fMzD?EZ-D?x~BtS=ChGd37Mx%OG2GYHf6) zKQOgHD6B?cRnde6R@=u^SNCV2y7DwsuT!H7bG3Z%fWI4hN_@}DtVcSc*d)a9m=8r7 zb%B%1cc*Oi3#=5_5yn&C;jrL8)s}PpjeT%mG5d~lnh8Gr94ZhFDcxSE91Ok zK<&Ar&`{=(ppEq?F%l871ZDDhOeQGE2Lrf0*AK-QNx!q5=ljT4t%^_!ZE)*mtaT(q zY8-&Ej0U9cmC=)X+uB9*Vs)o50%9Ud&23*^TwF76RZ|Z{Ww5{jVh#m{FYv0Gr}lq4 zm|lO-{&&$v`H_4>p@92))DxgJ-)G@5Gv)4!#7PXwH=K93d2Mpejn6%ON{IM^=)RF3p3j`Xz}ppIBwn2%6ZB*tJm)=O3G6b==hVoFh-LiK*)H58!kJ2$I|!f2AQ zMf#>ye7I2w04B2;mZFKRD1Pic@zK!g_|8N>8wHQ72yK9TA&CB5$>k}@f^b4{ATP3W zFbJz|shp)L@4;rt{tck)z362{mCS6Yyu8S2@|EiYBL$Q`uMo>*MF2^v{Plv2yo~d( z6}BHKsLHZt9mSmX&x=;ACM1FcR06?1$=|~HFA1_>>=%-bCv=hJJ_4JUxCF`3XO(|p zlYe?&Pe#X@7{~O|t~Ay`+Sy|W7MZvyJDRfrsK4&E$PIra*~RhvF6QNT=mRF;p)@ac z)5I=B0M5Prdc#d=jCRTJ)`n}mKqJY4Xo5o*N@(@lGOS$|$iH&0W8s%J2gaLBFLI7` z$CAS)KWwdMT}TMdE<0&O78D~mv0{Rn>~KePe_J>c9G_5-HhlMU>+`xQ=!c+@9(Hrg zR}neci{eI{?Xmq6?TqIe-`UGLJsj+u`f7GxeNs{w4aGkRl|uNvt1&iih44LJ9}koj zUy2z@UwQ?iUYiyJ!sIzm2gQ;=KpLN`Y1Y5Su-x_IF(Z7V(QBQ+Lx9m__aafa= zYjrvTt-54?5!fRKGlvla_`l!{5vBIU?7>0f`BMrrK{4LVKqqI=B4%qZI-uQ+4b)f{ zGCg9Zi>Q7}M5TDrnsB*9X*5RMRCNq;o%nFZ@3cNQ3$*pAMj1l(UvdM0s=xFyymKGB zpBRkF70(bv{C(-;{}2ry8CB|)q`<(U*CC8-j>vlyy|%frIp74@JVKK>u{))H+k5V& z)T&#b+5Yv#9TvoLWf-F-fh-O=#u5L2p+3fqE}s0dNBX@e;L&~F{+kcnuB3IOsjlAS ziRSkNGNGnH%-7<3Tv#sQteq4vs1UvG`BvDePESEZoC0gKKqeuDJ+Kd`Y;LJlhC!rc zDzsI_LaAgw)xc!!wX`&B3JqB*5`V1^o*$+drTOh3Ra~)b_mtcUiljo^8E0X7iH0x4 zRdmG$+ZZMiiC*UFi5AV+^rD&}F-kRlryB{JuM&P2Z*x1?1Xr~0DsJVCRGWZ=Bl5r! zhNJHEKmbtHgV zgs$O-?e_52a?!U-YYSadMXrS;>&^)r>6bA`5XLB8<~9hIwbTT{NiY@oD0Wl!EmKl6 z-#Yhl3X-}@ULhM5Q@rrlw)=RQ=lZ;-t?t?swTNiaDp%4TA>?(oYRl=+yKJ9u7yeOJ zFOQ>%0J9>JIcT9R0Ntx-&6t0r8aH8qE5cb>NqL^(R{+i^NY_;_{=&y0@JWxK4L8gn zna&!OZ@Kh^uU9|48L=@TQ^;u_0iQ~E7SFCpD@N}cZDZpizGjng=jVRN6GH}MPt_W7-1qE8QuLDQeLyL9j7XIfUA zM6Vv~QLYo4wY~w%cef z@3p)xG-YKvCp^BSMv%t(FF_PYCqDhVf5)B}uAYvERSPKi9elepwDZeBpsvGjtNZCd z5UkF8d&vUu@V}yFWJH>gfYm8Zu4NM8Y=!0MjpPYXoIgtHq{**~n?yyaO}IA)Y!g{z zTGgr*ojXH;c%i-!XYS*opzCv|`ajzWK&d14Ev=Y%y}$rEmBX{uqt--lB<>enr7CFI<{oNaNNaGSrkAai;jBN{R<$Aj zYFhRz`Q$0YnbMSgOR)sQmGHNS25_4#JDz+DhrH?yWd{g7npS?2| zSt!#BoQ(}|!jwH#2pNGw*7pj3jxs}xbIp9fPzD~nzbEdkIKJL2tggYbS;#Egvvtsx zs>LtvGRop)eAm#@RI)(vPX>|h+3Ju`(Gd~uN0Z`p4dybkTt3<8Ky$UMLG6oJVHZs{ zgf4|bG3pRVrux$-u+DAQ!>D`iQPuppKGb+I>J)GES}&RG{w{_FDkr;SSrGp7IP>-Q zoL%$!Znkb{DjNxl*>cLiN_7;ZtW*2XYfDxkfC+gz3O8lOabc)8A|lzU*X}~VIX0); zUI0H&r_LHh@DMY@sud2jTtOq6ktad2j|~(3+tRkw8T9Txfgl;o1d0Qr)K81LFSJ3i z0@Sl@@zSRaSiHw#1q_zVJyZR!3@C5ZPBl>f?$w~Ow~Q-_en(duDYTHq0TpV6HJdEX zUkNM0L6F|PkMD1}CJDw7rfG50!@rLhnq^3I1vC2bzb-ZHwYPmq*8^_A2=PH@=T})| zr=)2MpK^u`(`M8dvK#VWw81zb{BWh@Fm!IO9d(`-mk2L!t!ppvJ&KA6M{7Gj z2x(7H7`SkGW+!Pm?kq{p`M4Ht%UF`Ont)%F?N!D#5Ng;; z?&-=X^FN)Tl_^i)G?r?#(*7bJq%Cc}{iQ~5@aE-7pCm2*G*EsD-t@plK5qVslF=;x zvHQG{63WuDFhe5UHV{SczF9DBEiKaxF+k0|L{Sfl^BP8hZ^lsPJ-(+bCt#=Lw?Ou`{-i4r2YdxO6pvm6E zZ+00AuH21(Nhro6hTX`r&0I#`iq2k57CP#6HB+YjSJ=rm6m=)*pM&_0o}snSuV$(I z|9OK^`vJe*{zulPB>zEkS(_s zD~x)k8<6Ue5Sp|}>Yb{UTK5VKN+nqWf(Wk77;%c2{M*sgU9Jukmf1OZ+kSi;{kGP8 zai9#}nFJ5UN0~bIGiItfF9x_pK%)4l6)u{J6y{xXq z4iigULdIA-k#Yv_(^_XKx&(7?OerDv?8`;r8o(E?_T=E;rWMA!e${DHs*Lih{_XAQ zO-O2yo7Q$>rHk7s7Pby-OjSLm&+mfc3xQ2HeY>&9q-&*%Kl4GU#lGCWDSvO+>;Ji7 z3Eb_4UT!A*Y~gFMa+#(d9;0j?uC6~lc;{ZeC#830vS#I8Gg<51-|h-$6V&NV0D=15 zgkSGmj_8oe5qIP07f_?ds7ThJ81I$ku^3r%I%igz)?t!TY7s^4B32x$zdp%c?WS7V zv`j-rn+p*9;jf+?3G{00+AZ&}_;&~ocbH!vByB$BBs$S29bc-6y!4bg;Lm#ug6cKj zdv&{g2^=b2cs5!(Eyl{GUA9;=%#uNR{GmON=u|1{@X5kC(lJZs8t=1cNv+Z9X}dA@ zKF-xh3(zZ5N`sfRqs*o(V6F-)lD57Y>-CqwVMJb*yw^x4$q^sRi-IE%-NX8J{6V+e zHN*^B0coMxsda}O#koCC(qd|K8`iTBoqYmKG0 za6W%LN?x=&YC+;rnAX*2{MoF&EAXE5wL(-Lw0*Ak>pZtrzaIBoXk$IF$IQS0Q8T3W zpTGAoeDoDEZor4M_L%$T+F^CkHb7ekHL8iyXPZRfh((Ls8Eki1Ssj_4ijC z+Ai1C%IH9UH>9Y%tE}dQ15v>$rz2UdsHCgq!;_(o7mUM7^j6G~95(iGNG7XyDdE0` zdY9ieLYbedlPv8$gn2MOEeWZ_OkgKy<#)eeD`&rA`6Rhy9xxsjH3&99gv?;ZeeT}(5V!b={wnte9XX^9FYC>~U~ z0@SFBK^9oU!qpAnlqI5z)3xJwOO+7_qt&XLcjwQ$11cw5X4=}EGfd=L-Vk*wPu0bC zhke_^@ABSCzjanQB(1dHZ8KvfT584Jk@&=^cda6`;vnVFN}K4RsfcPJ;q`k%tT*yh zy_%5KvKNo>cfsp*}fX%=awG9EnQ{rwnN5c|qIvvtg9IyME{W~Gad zGYC6V`)^7=J*(II`;ohK8cx40k3i$snRa-eBLFo-Aas}Jm;oFNNT=2XB_#we_kqW} zt8>%GZ(SGiCVja<=sJK&+Tz#{*f?%0t;(5F7#-U5@Q}?jXVicBHPsa9b?F$=^m=o^ zyQcROE|DHpA|urU;P*%0?>YL82;WZKnt8=!XN1zAX)BCZo%EjouC4fIR-pFZ9Oojy zeukxdlfz#VI%@wi(+sF?wzgq>Tw3OAqc2y#)`*2-j>?4YufYxBl4V2+qEDBfEE&^7 z1j9SO%YSw7Pu4V=f>KX-l&nx;0hpXGw2F#R5!2w5W2sfdz{NOGf{PMPg}^1_hHCr6 z^%csFO;jI5-;8TyQ>dgh zZt2mBJ9K-5))8ZgOeEwBj5&|SW*N0Ag8OvF0s+P)orvF`8^mQtf2FK+gnw&*bPs~! zK#vXZ7l++WmD?w0n;R-V?WN**(*imG-;PZQOo#hdAzgxrEScTwJFUL+drv5J8(dw} zS2XGCpQk5Q>!JTT2W0=#Ie;g5x#BzOOWu-Sg}1z}2@Y~y7Vv1^_jUPG2bnjMb)B)& z08|me13la({bf;KV8EG6u1O!W1sVz>ir^M^4cP$d5(m6PQI&9w5kVq7cKlo6c4K6v zO_lZlbQ#Cu6$kFq8Ql*xa2_-rCLWwDg~H7aQ=DMCj&H>xe08Wht9Ay7?wV_!ME)>h z1{b{&{+hnL7Ax`Db_Q7c$2;_{X$-9oMCl=kJ4>v0+iYmb4h1oJCK)W&q zD~rOmg?D}$w(%%!+Rx~aJ5J8$7^dC>QFuS@TWyw(!XJlEDvm;vR0K|#`EfadO9Psjjl7g-lpopeXW=0;!e#MzN{x6 zH=hm0-L$w{pQ)|z?4DX#g|7IH?jB{nl76c5uP0{K?ekHrNci9Jr;C&5^B@S`R}s=h zG=2J1EY~XW(dN0Ty2I;F{wm)Q&u_`g(Lm>f7RU{{#kdt|V`%|rRO28GnKA;}0J$X4Aa*RvI3oQRTyEsRE>9h78dig<2Pl8rO(G zmIG)#C%MyoHrV{oSSv0z0DeMs(gUgL_?$O0R{2PEx6er? zQg577lz%6mkofgxvltI20{J_Qkh)Yrc(1GS~G4xxiAH&0R5g!rwyUC18_I?9nG z<;=O}LN_g&x?cu9k4C?AyBMcZ<gzjRro z=e}CoVbQW8?hJT=X!%H7w_Y_rMOmFDcXxDexGn>GDqw2_MXcn;jXbJMypCZsY-)2r zObLSu74Tx}Ta*Bs;e$EZA)i=S9+PPxKAf1kFu4?aPZH?<_WBR{6+E*aFp7(~6Ay}=YuG*5R z#wFQUHb4vkBztGwb4YUGf+otMSKFWN&lm$wN;b3hqpZ~h+%X@O=38|*_q%h4!*m>O zN50{-+qe1f?eV(1#v2Y?!TTM4Vlehg?@4RN`(4U^ZCnaCdB}?c?$G32iUO7ANdf-H zk5Zx_SoU-9v>mZYY0h8d=oRU|DQ7YH#{) zTy(xPsP7oQFeLUc4wZEDLC>73;nEx^pjsQ9v$Ft%>_6))O0!LqQytD_O#+}Dk*eAv>{e0O%mTu@%HH)Q)F%g-hdVCRPU|wrY>-`x=10@ zS$2Ig#>qL^%kW!QlY34h)3zDjiFq@({NtWV<9D$w9>5TKn~$z+>9s{MYU$B{f}JB3 zO1UbiI3=%$Le#4yok!7cC>Nzx2C57x2em!>uoZUANViV#A@ZK8a7&Lh3= zp^cez?py;@2-cZsOfb7uZVdYX43g#nUpWb()Ki6=J)juLu=iw;N)QWxYWtySV@8_$ z!;r?}WdI(ZJ{oY$ji2m0(ptXIN<(=cjZ{gRsu57-kjkSluVAn|v?7*qa{hyxgC75) zYC6_IzR_&x(o0|7j28DgO-2`#`d4o}%tjaeD7CdVo6cBHqAjvO4#qz~y&hnz(tJzu z4^S{7FncC@@i%$z2}2Zqk3{-~YK%>t>(b`>W`j$Pe6ryfHh~nXPcAz=sWZWY9;Z5! zhaXFU$3IDH_dk-Bn7`GX^}1jzZ>ro~*Fh!Ve1XawMFGCYqjNu}>jymIH#rS=Ghbb_ zD4Hr0wYYlxYY%_MDrQDTMx)dD55Dx!dLK?<0pE6DCW^7J!8>`_E-V>~r}t2xEf_#` z5h$4yO^0*SFgvrkqfXFuDGO{;Fn87Gp#xQI<2!<5I^UE;z7iMT=#TPJd)-9!fCbP?f5FgwhBgVmf%>+9RSbbV;{+ zJh;y3=vXqI4%tUFR%&7jhAaa7Pq}VfnGT=*r(t%<@PEp+omOTi5lIazkSAWak_^)x z(G5H(u!Z}!=;rFS4)$HW6A_3+1lEI)19KqN>!j-OUUhK-%+sx-E5yGe9!T$jf>@|A zF6U-d6(pO08NH3aEpT|!#0U?>A#CwiEW@^E;FiR8X^h%ph*5161JpSVOCy`i(-^7| z<~0rmtGIlv+F4W;#b5@}AcoNVNRY2aUa(nC_cW16)Ib3~Z&%@q!r5jeGV~?W zG{Bu&NPqn@EvXbg?<)PfAvu~x``r9x8zqt5+2tAZHI>J%9s>;uyTTdAKg+`=i@WDG zk!Cu4>ulFt8THC0EQ$X7vG$m?8DEm|-QwSmQfy|IY&u?Tg^`NwG7TBcO9@3s@!cJM zc#nCp`*ys&OADMskB&{52~aMM?gwWu?&2zLHbY`oodEQGUH`zr9!;bLQTmUMP*Hez zJq$I7&@rX}#;Fk3x@vY=FTtGwCV{p!q=-GW-Jo(s!V!J#>5 z?9F;f_EY+q4zFIBNd1%aUc7EKOaERb3Ax8FRvefNqA}`-4kW-^0<~(;X!ro5zor7g z5e(~hl&gp}bS(Z40;v}>Y8Je}{V>e@@qx6xaOZl&1}w9Qw5cNJa&;ZG{0PdvrtSK|fLPyjj|mtvGxdmRu2hBs-~% z9VFNj9rAWdmrsGYDK=mEF_g#Zzfvx)(|TQS1Jwjfe22!UY(v)nOchQ*(qxd|!7c2Y z&VQ9|FC!$z)aUUzcwZx70}C-4Qw2B4qf^M8Vyoy~`n$L<)ltp*9jikw(nPY1!E5rd zPC5}=8H0m-76t{c^h#4feAu`$IE=13iZOz8UBS!H)yLJU85gF7yM$`J3`pbV#9+-1 z87tua&%+@HBc)`bbFu-;L>Dz+4YI8U^f#<&Xdpr9*PnV;s=^pHsURdw{<7UwMz`QY z5_!!cYCaP!r**o3HpPP1`TFyEb&2M5l1=MSW~G?q)N$z(bs#;z?Rm1qJ7vgO!0}au zQ?=;ZM_6s;OAhTEE{r;hg?EW*x)o*Tn(p|u;!rW}$D_w03bRD;+zpWn6CWH|{^!9) zO`@}bnq$Mg{u;K~?gOggF#C-A2Toz#{7QFdUM47S%^tEV%KTZ*;5%srC(|ik{wcii zM_+_1nHTF2X~immQ$vD2Ln!sD8(k*Q-Ue%I!aS6{dDjmoZYyRkrgS=MsSSQBx+CDp{cB*Az7k|ox|A6+~*_y{WDRDV7MIl-9!SS_fxUyP*fImbFscGy_T@u$G{cp)oN zD^oDn88(6awpX6{Y-iPb9>W%4gva5l@wr+2Hnok)WzCMk^x7Sx#((o(JrcJ&4G)Uo zv|L@Q)(2LjF*BG!>&T&;9^tyPOgy?iLWAx#j3y>Sll-T#rY9op%45AGg4kUCO(&sO z*BJq;Lg7fWa1cOk`&jGW6KuPB(C7cuW%|D-SaR?_o5Imgi6HauYeGX({5S((K|_qO{oV!pGS1_H2Aq(kwX4<=qZVI6LW)R7N=le4hLDpo$c|#|m)P-t*~5*7G7=z- z{=z-i+%o}ljwUbb`ka4fKFNYmV*x(c99HK!QYc9`;xHe7PSE`#mJL(U84b+qG$+^W z-mu$c*-fg#*}6xsC2ksPYZ0y9McMb= zXjgXg`+d!sgu`)D8pLrwav}7#+k!fcN@hoRbe0SWF|9w$k<}pUEbH>4I!63j))#2H z8?_F-yPA*q^3NAG{~v83!`}+lT^ma5<9;AvyCCgV7&xhq*=n;~A_U91X!O@h^JnY8 zMOtz6BAc)W$nmB^z*_#FQKtx{UF1Z>T##ZBp+S3P+SHTK>YyZJ|0T)tOIs>`98lgC zW$H-3l^n8MCX~7d6Np0|MI5?ud~AiDpk+uqvp>__9>9qW__zqaK>sB`a2DT(S2dVo zG;?6+!8o`DwxLW0iUII#?-tovrd+;``(VwH;Q>kHSfQ2QYm6z+lyY+~>uE$$Q#c#( z^EsQ~8!WKTYNA3mEmzYmzH{00qMdmSa4(@<2G}URtF+7{?JxBEikd_2yepMi90$2) zAB^TY-JjB>^rRnr)&P2z;Vm-yx-ZwarWlnJZ8wK%i1p_?nJ(W@UOG=Y0mGRez(2xd#V|a~C z;l5H~LiXMjQW!FErdKqI7?9)A&9on(#S!y z)$E=J$mT#9Fbi}VXdAt-*#UbKK9hwzPGSi_#PVtQw}&Lm0tXlMukrw7XdP=iSzKB) znql7Ia;omNeA(V!rtk1R%NCr+ViZ&=kx+z%XkIY(bll|eEoo9JEJD*KpRmiK>tXd)tNrR@WoUj{~>XKcvZK6E@7Z;#j z{*6XKsJ(6b?eKE5`A*r}=!I1b+URxmHLT|)l3><}j6LY3N|y6sWYc2?iqS`ZUgYle zD3|1i=o;5RMz?$aH9OMiZ3Y3L(nB@E0Z%%B>1N&HsR|3D+CiaO|5C_1x_F7j3a20~ zVf06IdM7$Ss0N@;>wy>d~R-$RIK}a zGVQ4>xu3s@b}Tu=fPO^rC_`fdsb?ckGhkDt*!$Y6+HT*UGR(pDMm9%p0W4cjs%^c) zy+$jq5UFLY@|Mzx8afFeF6P^Vq)gGWgZ=}O?U}LeBX7;XW6OAItJ|#GpI>{u8qpNMn63;cxzT0T2k%Oe?y$0uC8Zd z+3)|JWTPB^HEvBpEhub482figFR#Uo#^j>veX=m9HLh2`8BrqBA@QRlaa5sLUgKTP z{BiOZ5dvkAZ{@+wGT@FahVtEP&-J^uTulCyAM8RbO#U z3;6Mctg`Tusp&GyqE!$3a2Xcwiay-}D(i?SS})IQkM80)VWSy~P3#vetNjUKYQmQ{ zJ>cR;gu7L6d1A9z&2aD~Hro(R!dx<=Et^VF7z1!Bt0yT|l21YzZ8j);&` zs^*m#PaWu^?sPw6=I`$WofsHMzxjSR4@9&DBfE~(9S-7pqzpcA*q%Rsb$LC0aE2Q8ItqiD}MfW1!O$QhY8na3T{vOuon-3c|#P851IV zcY_`VaS(G!oM3G-^C*e{Z3dMa`|IeKmC?oU=-4rCPv1&-_Fi*^nrn_n4@bX zZfhqtr49Ej_2{&5_1UPK=hpvYzJ&O)0$B5uFFFYh30;ETc{iwhnE&Z>ba=4;ow2@Q zs63p3V;rSXrhyZ5N!&(hw>dO-uNZ;G(LW#ACJkUuCUwP*To6 zEEk9*nMFWlnqczzo34-LHr-HaI+f|wA#(H0@l~_1!hG@_@4Q)sOBd z(!Jg|-XpQcL)7wUNY#VFPk|RHHr#NxQX=ze4rFlv>O2KKg2a3ijcZ=X4EO@gwg;Dy zsl?1#)faQfcTT$KLmRwQpW*QmZ{JgR#49>4HPl;I-0PFV535kZTS3fpgeD-gt|>-& zPfc4WExI(W)b5r8HbXp;(7I@%7>Tl3dz z-vBw1>kn;gjmZDFc%XW>SNc3PaGm_d9RTaALpLTzs^U4y)YnD`+EXWL*7|wGDy}`V zJZ`VaZFKK^MW*p;}!$ zG{;94U*V_~^D;g@j~mA0Nip^)*^Hvu;}z}5j~r$n+kLs5sjN%-aEbdlF+p}~*tBZ0 zoIc6$$9*30*A#B}xL1|jO>^uV5RsmxnBdHA^-ig-JI0yp;#hSiiqvY@xqyd zb7NACKJ-un>q;RAuT}=?x0*au#DFh~M_OW*q5f%C2k%q#Be97jzj7yLIXX6CFG+jw zeL&Rlyz%J0ly@*71C&exN}^bf>o8W&CWbwZSKqi^Wn*=6=^>XT9|>TQ}m7HkI?< zAk^L)A;dzTxke3yyl6IPunKOtbd@mN&3iYcPaHkKsh!!k_4{*PO1S33TYonfFOM_`Eq(gXUT0b3`pZv|@@9lTmD z^(U|N3;;8{kpSrbl7Ck6=fiv*Xjc^FRw?j$ISoLm-h9yXH+SK8+l(~-5Jd+a&%BS% z5*TH-XANik!(j&}@pb6qTt|CsJeRq`5PNzljc$FqKG;%qb^OgiNp(uc+Ea?a8?T|e zyAa2rwdWTtu%|%q;cjN!-T#{AUmg?HG=DSXGcv0hzoZ`Bfon6zl0Fr~a84o{oMs?j z(c<2>xYp}UQxMd-z=-J!$#jNz{ShDs+cFhNu5wL8!AeRRNgN<5N+OJ%-0-}TNPcbv z%i{@_^gfE#>gQBEKhde3hI`I3`DyJaiD#;9>LG-`z0(R1Rzzb|4#Z*}+5CdxzhNn8G2N{gHTtoY z%p|d?Dt6egK>b{pu5S0izhQk8LqCe-BYyuPMd}$lIdNj2FjxB|Bl|D9$mFHTGHTVQ zn{A5SxR#CygkRkT8sCq&*2t4%Y$)ROq~F-<;Xxq6^M(X{l2UMdJelS=GcR)fO^``1 zV~V9zG8f!B z{O0)MGa(Ni6q{7m{Qt2@oF~iI0rWY)j^6kV`Fk8b<$XZ7Mu?7MoPCaSECG9ZOjeTE zvMj#Nuu+Nu32ABw@!uz+xdZu`S-8+k{@nlS-DNkQc+SN|1S#hvv+WUgfb=x*jNV9sQ@2MQ<$>ozXH*j#f-Jp$}aq-`A4`a!ea>%_rtX$>$4y zRs)5~oC4nEBr#>rOOtojM(e?bFP~Qlph~o%QQ1^d!zqTwlXq&qV9NFUz%1izBaF6H zHtqKo0BoqmrS>gi?l%JfV$$9o+J-ahEZlhD9B|ViZ1lU;cIPiQJ@0bH8VMnLQJS{dYkUL;jh}e~Mbf$UF2RrU{|4Sy)F0H9z z4>WL!G3Fj4uA`W)+{;h&tA-eA+-F+NsA<);cTHF{;W~I2)*SlA?js)J%>wHj1`rLZ z@_7n4X`umteZ}24>p-T8q{;Gpd%ScOzEimPa{GcZZJEwQ*Wdor^*74Mf++1_VcZLK zM1tpAr$2=b!_QYh4y1+3qz5DY#X^go`Lh~tjc1)$iXH(w3+HDTKljg5rDK9-9IQtV z0g^MsQEs%RkU zr4MMuroRHEkaf*f7VIxXwv3u`8hk%xoH|iHr>%8%Y>>HIkN|+wOC=vAYoJOnB{0C7 zfQ#TIwV9f_azD^DU0BZ7Py9CQnK!i5)V?n8ynU`4^&@8o_{8BFCz4w4SH5_^n&kENAtAfEM>o&P>m_*i*R8v zO77-Dz#h%z4jQ;rTq_M%IL_9cf3?Pfk_JxxcF#+QmK}0%A3WzRwQbCdtwlT=U}Qi^ zLV=SKT}!+8j7M)B*O<8Bi{F8}32hMZZHZh0(FL(TUTl|vnyOTd9V}o-7Za(JV-8-< z&t4nL#sK(8!*WPzn8F%ZOe|i;SjvK7SL$h`&AOKAQTmiE%mDIpNzJu?ouG;AS7p$= zq~;Ol$X@lz)g_|BzP%M0Q~4ojG}&I?CG4IG0t%tg4poo0e_uxG)@+@SEbvb|AMrr7 z^SUx0ON=jm{S6hhytvzvei8EN(Tn&US89Jw5POD$idh?6S#UpVb6OvepAeI>NVi{S*N6zUtJZc^@1G zvf1KonU^xpCo4#587*w^u_RT40U)^Ba^{-o;9ss!3!4;>cn4ngc1g;oIB|qiw7hJn zi3nZAjhkTrlS{${e1_K+pJA`&*Ph89P(i?{$b68w_ zdz~FgBg4hce!o>67zE{lj?(eIfr{7zQ`Ig@ zy7~RXq2GYt(9-F<;fE>F{LGsZf{F8pq4ZA%uWr6oJ|*gqk(m=E7(=bXjw;Jh;_V#|&by}5lpBU}>ApBk4tFO(XP<4cqm}C@m z;wjJS$Qu;>hH|&|OsxK-vZ2C1mzt;(Gw+|DIc!*3Y(7lK2JT(e09_uvE+jG4i6>C~ZvWg3RgFd#%0wrCFYD z$ziY~FF5Q=pH*Qu7tH8lxP>3o?)Te}gaU!|tpD?X6J zXA(LHS_f>_N7tFdg%{m7&WF^|UOEhkLwOlrIc&|@?s|LPac@`!5$Oj~^h9I(I6N{) znL?i|v**^17q0wMqJfn{KfPx8@ge!AKjNyVt5O@A+8c0?#eDz#0H`y=t@;jD z=AG@FtUVJEXWFx*)rKA$`^I^DpI+C6kl9f^ASRR?)3p+hIpYqewym%JJrv% z#25G6J=uvtjNE(741QmoOANV1d3(-7P2C7B<2A0q= z##Vc@$^EFC^Ldf+(rpI-mE0s69;#iL*?)zL??6LRP-94%U`#@-gV{7bAhuy8%!wa- zf->}Tn1{I-%PC8=AGwsq>D9`-muZ>8=y);p#;lRwu9axyYs=hQ+IFjt&HY852i5q} zWuF$l-I(1+Xe}sZx*mNEZV*L2+e3Ax@de*S@hZREhgSEkC{(28)L;>-*(U(3^=`NR zrn9y#Gd&H*fECT#w1CO{|FK%5z)lS2X7{8rSsae;WOw{|=LN5W;eNxu^_<6R!PH+d zv^yP38_H~-TNy81{hZ$o=u=8?aE6^c${{)IUgqXk-k|0x#K7<-rL?12A))c~r&gXi zmGUf^eLv;k#mAWNM;ndZAw}9HEQY9G@Z(or8L*MwN8^X;a=3h?+#JRbHkjPcyY12t z9mC#0n-yhZad-Q|78=f<_Ui;o93S_vRmsxIRkaK zxjl7kL_l}=-BM0e|67y4#Cn=H{LQO>ArNo!{{WD5--0H(#*QDWPCkY-9}xqd$J6YK z9m}MrL6j9NXLBGa3mcv|2!11oRLxg@qDd`S zV!G86S>*EU(Ohsh5KhcGplD9~IpKXx2Dy2@VL4}SsAB#Wb5XA~&JQq%zjQR`ePYS8 z+uvxe^!@3ZDQj3J3ND%a$Bck?ZOoqThiQ8PXnZBAAkbU#meIO3)jbBEOU5Tf{QEX; z){W~ClPnte1nOoJG%X(N}ykkMU;f0 zgMoSHo6XZpp1yQYTLP!PKfwZ#N3B3eiE>!#$$wb@*=**DGT1nWDB}V;07kS#Xnt-9 zfSD;FcXLgvAzI;o-$H+1a16F+mVwNt@2kCU)0ot5G8dA4TcRg+1Nyi%@|W{_M%xdg zo_yP`UzG>^c`UQj`s(S43xk(S^wVr^vsZ2l*jf#*w;T47el$)Zn;9fwi8`wJQAVwP z-DDz%07zslnXQsh1bjrW&0|^FoXiQq>V~asOwvK5F4oVqB(M6}s3}*H7N<%la$JiV z-SW3Y`V*3CvWsH2jxv17#D0tX%$kz2-M)C^As_><4sSdKQLWD5-c3sJ0Sa_K6+C+u z(v_brHSMV}uQf!u&R1dJtXB&MM?myEK9AHAK~B^Ct;H{u62+w5G2pQU8cihsGd>}n z{ndTix1HEABryf^rp1n(;Uo2?VYe={y*R!@x|KQn!A_nJL3^HA%is>Ji>L)&Egrat zy7mD6*LWBjhrsNdyVacN(c|56A#?=-E7E-Sylcg*MIC3TZ7Sncq zzCP!r=WU6F2*%F*C|aX*^?~lB`l$Pf1t-{;qKYLBbBjJ_r6~A#u|P$|d$mNvT3H;6 zSTwwrVOa&W%%8aiwLea_0-kOsOqLcqH*In5)fyN1q^aV?JHZ9-H0|8dclFsdP}AHO z-!DydyHk|$*i_elR0kGXj4|J7W=WVdE$2)XO;9zo22S^fu6*g4eP7)Er!gc0>@7D^ z$Sfump1?tj220o>EmQj6as&?eOEIS-ZX4oM0GaalE~U-$F(0$>_mVM!W>kK(nbi!kS7+JWjS#$$o@z?H#P_Yr- ziD&(WdTA#ZkbOr>i_UZ)XDd|VGtddE7YgdUcGPF*AiAsUS{cc5V*eEuCET4Yd)gk0 zp)|KuzJ9yyQxOeKD(H;5=hab;g(fXF(;6oGU%v4is*Y&VU&%M#+~NR&@tY+cBMjNE zEI0n*fa|VO5+dFY121YqcZBv>6y11{xp79IAn;9H+}x~$OUVZc(SV8oc@;cqec6QI zQR=$yZ8&U(!jy4^hZG(#&f`7WG75aVZ-=o2y}p9)97kXY?rXbyb)FvUs~o2<%AJ_cHVXd2P5hPUgt^sxXsz zbjcfwYd=vji>2g1i~e4m?cVf`^4z2~BHx$>Q9LKy$GN~1_~42OY24gRG2ntniQ;+( zAZXZ!)gsxUnPp1>LkSr8JX}EVmO{XaE@)lRTe5DPoy~3eD$##ZC#nf4(->EP<~2at z{jHMR@8^#2kLPWB}E&{B040g_>L2nv2-U-&68Ln%RI0@SpZXh9YVdJucG z0^ren!IoOCTD@8vfty-`h!>&&a1lP@7RN%Q1E{@&^lgbU>g9vvNGYs8 zQOXqiQopFtHL1e*&C*M1_xd)Km

-k+M2E6658YFf_B|+ta>thGYwH;_W3mWP=tk zj$x6BSyk;)L+&&|4IV89adk#)x>7*kE3x@LbH;@be7$VYx0aJv;+4A<(e|;5>?9Q!%H%7`I~BY)c@XQc?@(^l3~y3qsi!($Lyoq9QQvNtL_#Oz zdlK!3_ZZ`>Pq$F}& z`3*hCTX8s@G?s<}^*eM&87i*NIA2xnjBPeOf@q-Z$d25hpQy)$R<7eGeE5{VsLHq#)b6PAFRZ&?Z={nzLaZyex(s6*#QM^|;g zM2bk@P_ThrWHw`AjA}ry2qcT_PYHtPI6*iE@TksTkWqgy@8V$YY#gJ z<_HA(I>!mf4;^8G{01ks?-KDRr4lh~%HeGkjKbTPl3P)loP^KkB&MhjbU=XhKHQtN zFE<00iCB+T0{VBP%;0W#|Uw$k~h8oEg)F( z|F?h`I)Pw(R`Gr>>5RV-lNXklrxv2OW4S@34iS@a2?=@Sgkhe9KL7a1_XpMHRUSA7 zdf!s{pSiZG=7#;) zXb9!N(BQ|p)w!}>2~~QlnXK!!_PT=_j!@dMf3oxVh%UWw%{6)qZ3n_E0Wr`>UFPjs zjd}|J-^^(_?gNDtCs%Z> zDjg%gb|ySG82{Ji5Qx#r+0MAzxI^KkYijU?lK}YFL@yff_bCM*Wt3MdOHWvit9)5X z^H`Ktv%}MY+)DuXVea8i+MGE7Z03tr$;8_X2DL5G_lu!vU8zqGV(+g-wdNPL)Zbv0 zXiXxnO-d}zdxY^MRHMfx@{NB4yW|O2+1`t``fmn2o7}}8(^!!1y=(pP^JflAP=-j7~7pafyyVR|t^!+*Vu>9%T zEaG3O>bdDOuTeYoO`IG}Rmn-dt7b;-N~LszGLc2|pM_zp8rLv10tMK<3Pu-tQH z1gd{X=3{*vnX9jdVD~#aqTQbOQlWLLF{4RL2mCcS=32pz9!OiORcyQ@R^)N zdVxl767IDzCMB+MSGjMndCGmXl8ZV_={mvJa;DR~9^hg$G<;??mXA*mZ`8VgQhLvQ zA`sy>wmo+?PIEnNvTfAhHf%I3&gpowWx&(hg*^UjAPDdEZfM z{g;_u|1vC$I_4~YWjOhvR&~Uf%of9XmgJzf074E*%)f5RD2dv|JgsMP`{W@dRriOW&LvxDk z%af_A4N`rUG_WLa;Eh#rinbdkOjG9Qh|y5djZeM+u*A@@<4{LPaWyAewyHY0o>O@^ z$OYF6M`x!6(_F2awVy_L1JfbMpdrh22sxV9KtZ@v?f)-p9lypq6y zKJcu0uGG}(cN#%eW7J&SKzLP?O}Biw?5ZgvViv=mv+NfmOIWtdyz9^f;srh<{pIEu z&E7$RQAS?sQ*tX1pk~tbrdYt)#}3ZE0VydU>nVcoysm6;Yc=Vrr)=`55u8hduez?k0xJ%6QxPN?}UR!K0q zQw_URnk?98DyMXAE?XjwT94FgK7+s~rwo>eGV?We=uy=hJ!%d1LUG^dKn{kCmyGyGaI zG>h0?#sr=Fj`Q{(*qe!hy|4f@u>Mr=lP(EiU1g-N$fp z9}Qk;V|Z0nTfRtQpja&YH8lh`!ufc{CgZi_|LC)jvqnXx_0*kAps#)lqgiX#=Y$67 zApM7w?8Q_fh=BI}_7>5?_huXjW_;)WRi83oJ_0%Y+}gbQ4JBk{wPzDMNE0jJglVXTe8%4P=EKJ1&%V2Fv;z_r+5It3JL4nAoieVgeNH zVhn;ffG`rzDO>kx;U#q5*Uo<`4TXbB2MvhL9`Tu}Pr8y7CT)oWK<>`ns9e(e&T>z- z<7T zB##dL6fkEoy;Zi>RY$a0teO5X3S~l=`AVlR*!E5$TSwCJTlW`!dF#B78BuE4p}ED1lxNN(&IK zhV+<3yrJWA?_v81G0dt%|GG@BO@FUJT%)2>XW3UEKlarv{YstTX7>DMH^*;Que3_3 zx_bbd5nNhF2=RzenV%CfI8L`+*3Tg;C1mjmu}6FJ6B*aixxZpLD`sArgbS5iPC)}J z7>9py#jWe44^M-;iMzlNHbqo2fMLQ2-2KUC5X)cym}hvt_S7CihH@0iN&Xe2+~ zoAflAMtE$b%Fa2c!oGU0ddL>=zsq^;NPzLe6wBMoiG~s;{f{n9*h%^N)2p2v*KQOi zc|D5`eiea}trxi&HdCeAjsE^rX8^U4tVLX;6G2pD0}vrfTTO0Tov+`ar6pN*wIG~*?4Oe2Bb12g4@}rB_gzU z+kP%tNK72reYXK{B^4>=Oh)# zh&?d(Khrlg*@1CsgM_{VJ~}+_W&PN0V_?7sXy=ctnGGUp&kXilDVq{b8sp^r5ubA= zbB|J@;fJdN8HMuKQ6x4*@allsZx6r2x@rGo$NMz?tIdGmG33iSawtnCHp1=#qLpSW zpI^qMF+F;`V;LmAN@OJ_OfO@Txxxd7LE7<*Rk1YkRcAm`BGUkP2aXFNFP|Q(q0_p@APbs z|0K57?w)l02AvgsS4e+P-_C^VI_|t}P=wc@pA7FT?<;P?T*AD^1spBp2dsBON;O2g zn+}$?l76uQxf;tDZH|8WcWiz;AU6TflW8>dOrd6xNvp}h-hmZ3UgphY_~iO7SWh>0 zLIkmd78O1&Si9atk{i{eRZj5;_Y{0n*x*Oo#xR_rJPxDKK+3Y z3sW;+AGMIQ2+{08S`+I>)4pkl|jgAFfPs{&RIbxS z^E9mPN{5w{V?}wX3KO|>&wp8dqm`^misW7T9lBPP&@O$2HFHRFSJ=U!#l^28&q}US z-*z*YB(qG5mKYX4LD`{4`jSmk-X8$Z8Auh@M=)nRy4+>}W3nQ~0!9gLq-twJtYH8y zHzK{K)M@pr9=*+9^FYTDITMn^yZJ9dKL*b~VU_p3Ziux}@WQfQtNf6KGWLFvugJod z;>tFIHSW>i^c1i>7$Z z#jFuu0GxJpz|{#0@)jpmW`!!-okF|`Y+q%e?(eb8Syp~vEpkWEcC)C#{b!uoJk?!4|qj;@z**r)9+SY`y5f1YW5luC0JSs3(th?EV}2@_Ve~9 zNS|1g6wmCb(g2DPDUB5P&-<;}`^@?`o&QBs zKAcL*0oiLLM^eULxK&}g``gdIoa>wIH{i|o+Ynr2eJ+}@Z^*QsJHG*dM(fT(1`MJl zjC}0+p5u1Oy43hNb9tcDXyjG}8>paK)hI6n<6ll7!EM6(=)IYycot^SsGU|{d!&{y zI{Kd-QvlqjksahcTnOQ95O=DFRdXF>r~gt%2LmKd-1sRFs9a8|AK?JV47bRVczqlB zVTR>%05kmtlAehLX;`sGr?CJ2wN;hw zM;-e4afK4SRe3<;4q>{l({kjSce1!C)zt%s7E~Td?gnvJyRjQ07~3nmbF;jQW==$y z{l}S?-s7z2Ls4%$->FE;g$GumTIvO=rA|FJ`#JZM0szi9?*xJaAIz788{6?MvWHF# zKei{UH!nFgW`-=^JBuwg3WDJ}t`i@V2>nq`p{$!b2M6~JSTjmBx+GJ7*bC}8rXKRs zHUlNE=cWLWKm4lyd=W7os7G6#_H>3fGWfX-fV6dnpC=JgCp^sYG1PanV-N1u+u!c0 z7Q7v|K(MC+uVqu)iM}K=8o2b=*0OEyqUcIl6+dqfvK^#tgznn zZ)==a{`$S1#E0fyF;l&bI#d>}%|VlK#_>q=XZ*e&9?5ybd#p-GF`<~z*11Kad6-D5 zuIT8p$2r)Y>2;BviF|=LIUTQ$lebM5Yp$_bBeoTzIBhBfLcJft#$VCj!dX9e6i{QXY417c?j-L z1AM84?j33{wVM{z-&ln8|1TCH301H%)^X^8(lj!Me#sFMyk)CbpDN`?hXJsIdj)uO zb68TMc2xc|aqr3|IkmsZwUE9&`+&)nG<3pa>B6r0l<3jB1k4oRNpLj?D7USD?fWCZ zsLk_~?H{$|Jc^*|iJs6|^!Ij;IhQrVqTxO0Nw3S6!Z_Mz!ONr2KQ0&E0nq?ndM=w} z_uVnb-s*g_`&9mL^Kcf%z_4}J7PrM@N~}RKtJw7(cZ5K90(oDE6M1; zRs~dYlH%XwBr6MpHiv0oVP3QRi>dQ5+KBg0OL}o!BRu|8DB(n1sW7CNI_mdS14?{w ze}#OG?10EQz!%NGyuocCQSB6?!eid!L;>0o^M2x}W(0ew%6t-}j0&r7HB#D=|7&kC zD2>w$C}33n$Tp1MtUmh+M&Bw)%zLIWd9 zMCfQg$>#3cS0>HU2G{aINi<%)ynGen=hVpJ8UO3_*s-nCQ7o4}Jk4H0kI*{!!N2;o zNp}Ij@*m>8;yB25T7;uTT$E{Lw}i>%1o%eJ32nbQr`p%o_uz{6aNQvmcWcqYz^Dhb zT>fmWe#)y#_F#*xR%E{Q>i(uQ8XCy#;wv0^X%l#5)YU*$bf(;WZXUt3R5@OGndg(GmJk+zeTI zbVU4NtNrP@``cT%{p(?}zzHJ|Ncq=gtZ*4CR%m3Pwr0Lcn;JLf6n89}v1<`#GM>=j2@a&YrSsl)!^o6O}t?k8>#em!FkQe^Q5 zf=T+?0F~3*gYC4=FCcQ#`xiT}bHd^aD!4y8p6nYPIsdc6botWsqO1rFywnqG}*QVXp%AFj^&!aSC*r54@Ou<{q1F=eh>e12fx-(~v^+I|aX zu~aIyc)!xY?|l2&j|sW^&>6O5ip?(#0DrOz{D?V0;aS?$;S7t{pU4+F-Q}{VK3@tv z9lyJNyc9!a>W{M9Y?Ekv@`wMoS$MAd&j`wi--o6=DqZ(St$POo(4!8;>pngCs;9)R zFMo`K?2#D!oHRktx>$$NSByz^&s9H$i#<;$ z;h&+dliDBaw(-}q!~U9HuTbvRmL~|-4eUIuw~-7WD!pZ{@6L*1V}FEt5s+v$@*PU%Fv!Y_HX$yjd9kl!#PiIL?|nSRDMuJdvMh@`Zfd_< z_qikw?=n`{zc2ip(`ZCS-;=KMNw(2_vqFT3BuTKs)m_TrVldS9na=)_L z1RSx-6yRFO=wbhp*!9+>-$Fk%k02=93-ce1*Z8qHUPYIx2!6tnWYuR zlGt}I{?B^0AENr7-S*@!{y`O=Tp`&IYOFX4SSN!MiE z*q3Z2L*M3oIWTyK^T31c(j7**y0;(acs|o`arB-A7guSNe;v)nU8GSi+Jk~Xp@yNZ z$Q;}ja8skaU2c?XSG;;Iw)L?1Zx6?j+?iS>X2G2| z7e1b*Zl{hV=Tl4NOQ>yBgH`l;=Q$yb`MGI3s4uz_VaRPB7VSH{VOqPF4gu}s6dH(J z2A(cDhPR+*9j5-PZbf|9X@MQLz`yF;%zvF)ZBclpCjJS7o(iXR2R#F^TsYr4e4gFMCLA4BXC+%$0GMDp>N|^zZd3ku^w6#kg~VZ ztFq)V%hO>YZV(zRBl5FKxZDe&pd_zd!P=)aua2z0?R=HxCWZPZ??khu>15m$G8$-< zS5{R`&3}hre8Krjf@HzTko{d5gZ8884=*G-pu%(zDh#_0k#^EYG_+~_q)Z97yngXE z*l>XM^^nWFd0aI(wiC>uk%My=cf(Omhurim z{pLWk)u?ry@3vdfd7B4vLwCRDl6hq5X1`r(8BX~&HafjJcsLW>sW!PVwRM9 zwPK-e_R2*eo8wEy8uR+Ul#a$RX!GaT-HRQFU$y*nMAeFMQyoO#H7jcD!!OSp_l=R- z*U6(~iMip;6izNjIEt`jj!#A5z`ca=DFECBb#Lk1fHVb)T3s&bEj??gTm6Ds-7Rq+ zcNMG1aZG=TAphR)sc|<WiYX+|809f-|DfLm6D;8%|!21^&5wT+^NSj30?2JwNS8R#f2&f>6 zNv-b3hy0N1QU`idH8XFUSd^n^v-a8q8@?`VA+%jjF@O$jFQp+AG5`>l49GgrNG^UI zcK(hE$Y~}h9^`or1iSZ;{9o$YRPir8$jy!YQv|2B5_q6;+&~V1X~&)naakG?0`uU7 z7{8&jjQYI{0Sr0C=2cMP_0zSCez9#Ecl!Xam;Xw6q(!K3oMG1sm zRWed(%lUE6HzA)H&GoLLm8^$kuh9bHDc~wL!QTbxNZ*Y@v)2|z4tUrgLI^r&p+(KVgZhnS!OkhTklbH@)=fml- zTUM2K1$?iP!A0Or*(L=Uqtb6#z@psL2$BsS$m3qm`+k7M6PLnkEMzFUf#W&D+WbU^+#y|vhL z6I~p?57QAZ_O(_U6@2zFAw4YSn2SE+*oa-pAS$WUT@yP3*z9!!49xes7{Cf;tluOx zXQ7z-akg<~vc#DRI?k(vCbd|wM;(9Pe4ME1CuBig_JUy};mc$^n|Ft&QR3E4(xfDR z;k#a-X|BD2x&q(2} zi!zG}ypOH^=Uf2lU_k6&rKs z)}3QfWBSwrwPyPr@+ zrn*xa2gG?w4YqVEyR1vZ3eJy(mZyh10?L_XvRg7OIZ&dOS8uv<1^>F|cWA2JBK-b)f$&jx}SCBEM$)1Xez_*X;ei$14?p&XqlVx(p~ zeJs1uxQJ(pg!fZ{Uw@H$S#FGZzN3zJxDdeqva=VZChaoYLmCq7<(81o9Qct?=>$as&qME&r3wY($24y5)TKwiIv#_e*GS|CVMmA47G|(@qK1!_d zbkcOkG~T(ZId&L4S5y*EzemMPH&!j($5p*tpvnm?5b!Z2rt3|G4w@!YPVS%(>}>?L zQ-S|`0@M}I-wih>-(jB~Ua9GsxTQ%aW?xzc8}r&{?$1mX5(FN@*bY*baw!W1_{shn z1wgx99qs=O|6_lL|6bg_qB)ju-KRKBKi{ajqUGVqJ1c+puS<^g1vPC1iK0wweVp}O z&lz9g5PoC4gkGUii?M5I{q?PTZ!-{v2kH^!E6(&ci z-)IQaUW~jvPky`kkj-8EL0$~K=hvC7B=6IJG6i&aF|A=X&M%S^cT5r>q(~SDABE~t z!%5b}n}0};d3Mr8Y89xIVlt}vvv_RRZ0H&%E1*b0D*&8%$+(nk!3J(=g)hx_^lhd5 zgFkmmz{!N}<+%y#Vt;mAU<$=F{In!Jti*Y~?l&j&oakcMCFO>+D5%@~2?XT~Sme2C zz27g}QtYl%4#NNbH){PD9Ep8-yKS-8Sve5?o|-O#=V(EdZ_zt6Oj|i)5j%@+lZ0Nk z*W$UAi(Kr%ec;6+hH}Ds!(m^`VV1AY!9v>14HAgY?TW)l!K@5@rWj6JWy?idLHF}B zEnWoWw{~U?0d>PXmA)rrZqpU(ISR>b$r{n*j!}7xy3Cqr_}t{t3o|!gGaH9?ZlrG? zto+L)8h>V5ymWjLn5U)J(x;mLSZq|S3Pl0g(-;-B&e+NUVqf8T6&8n5N52Tq> zpbK~6DOtt1oU;}nI{QwNn@%8>KpK9FMQ2GKjq;M*89AdHvXfj z(z0dc_VOP>ynZUk>B3n1t1?HOS6*}UrT?CoFG;THRk1Xkc;#qJ#2eRZ`NxrCQ9XPt zt5QRJIikmc`n2tLRmKP&%<7``buH}VRZCZK@nrPLZ$j_1Q1< zM~jRQuY-MBntNDzp5gI)B4^Tpk)zTVE~vtxnAXmsCcDS=WJgai#NTj8+YwJa2D-30OLp!Qg-o78LnF zFsaW_tADr2R49`C!WIW;Q@J|fY)`wM(4QcUI?*-W8Cmce`0r;! zCFf(fI}=r(fqu^;ljq^0RJGUQbkt22nfwGM&*R%K^4fhIbVS5%Wz1Ml)vpWU1-DXr zI6kYs(mf5{dzUrpIRi?|N&QASrXoeeJn~KGt>0L!9oAdl3}dfEC>&g(YHM-9lELD; z2H8!DQ~P38Z>5^q86>Zy)6Lrc7^`!;Seh7a-#H!y)^T+y{j_e3w=7usUG;hLt-9JH z2d&wm*B+L`iLHO$ap0{}~<;|!apsYR(1m^|C3H~i>QRrANJROSdQ#oyUj(Ieit zWU+;&V)RDSN13xvjVagMFnNPBeOKSL?(H>94VU{=N7Z{c6KOP*8DGgBl1UXzRD5+1 z&OtDkexP0^lJOYH6q`9*Y!*H!CDf{?@wCxXL)W5;w-^hFA4~}P%5sVWu~K2#oX`O( zjyEb?3C9yUU6=0|lh1jv*P44Aj(i7opV>#3#QYz!zB(%ExOtcE5RjHonx!N}8bw+_ zfnB=0yIVx0I|K=7*rj`60Y$n&nx&*;k#4w){@#1;`@84!x97<0d}p3{=9w84QT+L$ zd2|6GMPfrey6GNcRFH1iFFeEiJ~I>$@=ggXpp1#}Gp>d1x3wt@kdLrmVSw=^)nFX& zm6&mwk)yf(Bd2IHw+#Mz03caXwbBH0i-|XB!JGkN{X@=e*Mz$Qs@5^-B1Jvef})j% zY#>PiSboXvWArx{p0A%8>s!Ibi?69mh|1r*tK*Z}v|KYKP{To9yu~uEN3QXAZ%3(? zS!8M69(k*OEY;%eF*@Tw`lYbkkV=CM!dBEkpkM4B3`%%Sa^hSeFX7?hnB*yePK;9} zI<6s)6ch9h2(!X3kYh@I&l3XR4R_Sz(g(Juzh~AU*;d)Ruy{=-*hFjN2S0{N*rSD@ z%<_vzI49d|!_C)f)|n{o_nJ*&LI-UZS>W81(!}v?&{1<*a_|kg*S&Pk^;RM3^f%a# ziAs9u(Mp(aC!A*5F9?vyncptH{hOq!zfHx|FmVX!&s$kg=;lxBiOw5+RMR?o zQ7>3sSlVNfzhEkmXCV=ZI(FP$^FlK9Sv?lc$EQ06k;V$0EIHXd27bpi9VI}9 z0cBXuSkhid%2vhDU1+&9&EXmxEj3ypnA%btd17bvPb~v*idl;hqJQ)qpvMuT?2be8 z&3pW<>fxvB*B3^Z1&N8_CM|j5X9Y*~LgSfCBX_(Q!vVI2I-CmHF?~G%{&C*`-U=)pX!x=9 z8CCzOU}C3V$AOs#bUJIx4E^{Di(V`mthEUHZRlgfuq3Eg#rNtUlMTV(tfntnxT(WH zc6u}1zrG?SGl4kUN~WbDb1uBSJ+thklu`2GkFRB>K?QS_{%qA%%Khx?#ojJA{(UT> zGkyy8)~dnxTP$7CHxpkWkMd}LW!iF;%|haDgj>n)+`U4sIXlSB}U5*LoA1Vx+ zX?7{KL+!Xxz^rG1li|u|Rk~wkEb;V}%CBXLqazqhiSSEZbUjJK$%WJGB?WY-UAC9T zqo(VOfNs|!+)$cFovx)Zv<};Arb^)_t``R1feeh($usTT2SJ@QY`%>PPRFRAU~s-xt$>a-(pIUqWhFppm&pWL#%#rBa@7 zud+1TDZXOa(M}Rir2EoygM4l}EyuE|0r{G_{o>zt_9is5HRy8d< z5)*CH6>maM7WTeKR+7+5II!iW&1Ls#);|@u{jpojAk`vZJlTwCTu+NNIo*{MfY?@M zYs7~9+3C2h8H{HO5Jv-97}jZXD>43k#%(Pg=( z`(#6p{&~pExQWELVjF=Xfl~8NJw1ohm&6^T8A-Gf2hZLy#(CX-t0}^yP8O}>DV41i zo3Gy~E4P>pitBDVK^-zMAWM+3mn9(5@?oE}e3w_}OBKP7<@CFlPDM7DNwE{Zb=5b_ zbmN2UJ5fE0D#kt8?BLVAGro+kIpW(Y5aE^vxV zJ|tMJEa_q;O`Y&_p=h_*_MbnelifQVp_BD5x$x)J&vV26$0@JF0wAiV9!*hfPy;Y_ zOy~GqvCh!T#;UbDIGCa7dqQb9cDXmNjZtujSpZ^0+$BdpxK~JH8}`#}um{gl&yeEXm63`k7@>9H+G$g>^zw;LKQYW z8HY)f8yeHY+-;38ni@**bSJ8hQeLcPFs)V-s*s@fw27GN2j zJ}7x_XCza!Jj&DhiEI$=%a7j`b!t24L-g*9r;C3AjccVL4VGun@u$D*zW^2ddruo` zUbzITvPd)vwwkAZ%_V1jD>cc{pIoVg~Dy8~E5zPYGTeW$^;?+|T6L$_j|2d00E7 z6PSf)infX+B);G0=!o@F$vf<$W9$#YKfUDykn`IKq7*=+R-LifHdjm>dhsWs<66sK zzdYODahQ}mxRm8f-gRJ}7sAS4SM|O6@eO)lW+8s~&B5MvIAGsTYhA;LCz}7JK@Yyr z!>{irItRDPhAz=bEo5E3`e2o8<@vT_65BQ6qI)0rzHJiHXwA8S8kfOOtuNIB4c>L>5}b>E`C2ne!6gJjH1 zD2YPt#`~Sit8{e?osZ`_Z9Q$f=SiDZ@+>s$;y}(>q~4R|5zlCoel!_A`F1yy%N6Hd z;>Kn($L3KYPcJX}&hydVe60cVaLJAdX;VF4mTm~tj`EAp^E9!UG-wtngqbTJgu>h* zUfJ1?giB#-ay_g*pI??6=1Lu~KH`IB(CwSzd`8}3Xn8a<3`Ab9sEsy_VGsHD2>!2w znY9S&$f;}h5Ew-bv9l5#=UVxs=A>AqPtpjy_)>^V_gHAQ^XH&*og`m2?%HKQo%!b% zHbSXI-u-KDQ6L~x5I0)&nsf&(5+Pw@we+0HyeecMPJES;#H?*A8@QYxaRlFJez2gU7h{IH+?52kyZw0{EMGo487_ogu)P?5U*iMg-mp>UAsmopU48pq+ciK=hVEV+s$ zziA#lu-q8;*<>B9h{+h(5k(WO`TnnOVg6DO_3mvfRM`}w%kWp$)7iR>a=3JDUe9)q5xSz$dbNteZDpiKE(2|GkSbpK&b}KKXR(H7b zN=gpsjyBbubAwAp&oPzC%eR;=`m2Vk$wvExzSvtVa`Ox}T{crUG7GoR#OvCAJo-~L z=igiJNk3`rvXq4a+Uopf2s04nlDThyRf-e#mDShTD(^YsB~wZ~enq@VGpR?3tuHY) zzd|oI4+bidd(WJC|E^c;xf`Cu6$Zqiw6q8KI6<1fY%4Z5PJIMOqG{4XYFwW28v-0e zT*z>wNUCbU8VN7aK;$H5%%p$gm-8RycW*kIzeQVc$Xa5!o;z4Enyiry;=IV8HWf$@ zD0J8f+t<=Em-ekeQM1n-Uc~3+q-qR?Xj?dXBl4^30oM9EJ;uf;v^ySDrfBrdX(gYlcNnzK1 z^-Cj8nu9=NvdeoqsP`&(1wuT!C}E<`JZ~707!vc%!C`wzxL8{rM>A}@!0`1%#ka0) zU;bK&@{eD*@1Erm#s;wDH_AqR8sPaKN86(6W2+CJTu<`Fp~+lG7hrDvr1dYFDf%=(tQ#!Q2gYhdYq8Lo4_?-jmq^8p7{pIrP+ zN7*vvOCqNwBHjoUikucLTtQ_0=V8*}#YRQw(G0sf@lV6;+*p1#UO-O*IjU-Z|~R&u5*?%7X9A4px;&IwD0JLomD{vjs3LiE$3*AnqA@WXg_V> z#cO%j{}jE=Y0@Nwn{vXbjV(5OA8;)RiYIEePK-886r1z1=@yMs)pxYy;W9ltEOjw~ z>O=i6u8N3vqr4^deUz=4L2N`4>6fM2JW?6AkzCX>{H_*7{p{E{0?)2zU4fc zXcwMXo-c7XSHZID8uFYd0O<_b28OfxE1p}x)H4HxTYRq!tmG?m?CTGiIP(ID0o#)( zAhfnA_eJ-S1M+sGGg?a*@5z3JU9Tq)r4NtfBUA}hPEP76%J(e8fV=NGei*D)lZCz9 zR=@G)CDJ9Fs>>tqc~XNdz61a``q=ahji|LWUy3wi{<_oJ3IBBAsBO5Lgz56Dyw8FP zT%8y=sqZn`LKs|=?oV{${9&@zA1wbiH@mP@MxjTtHa93TH{8u;r#z~bsMEi$B<#mgzhXRV;QHA2i#-8>JXYbW_X)WHZYHJ8s5XzzvOk@W!yhE4MT7+Dv zrFP}5ZvvcRiAu@C`1i*Dmabdfu2(b`S!R2rgIo1=U2Ea;bLpznzrQ}3sy#E2v^kon zUT3+q)(DpGi)J3&f8#RcuedZd{KZQ#>qzk05~Vcp6-SIN79T@fNXr(lB3s6j3U#Fb za2s_=zrL`cBH6fwlip)?P?$h?B;l9)2TFTLp4Z&mu8 z8E?(;lpXij??Ny6x9|3k57rs#R#TX(cHBnvUD12$6U2>}&ZxbotTrhYrk7~knaHk; zwYCW%=Y2+n8JkGtPBRg7941^_V`i8vJ%o(qFtNvm`Mb+S*x|-xHXpXTOunp>I8D9) z-};JmP@@XH&9fcbDNQYMnVyj3=GKLB-Rl4goEqt}=#ZD%)`h?tzN{6$Zobo_4EfC{{qtCJXmb`J=v);c9oxu`G^{fA3I#B-OEzMr6pR&T- zY3{Eu285Y1+{7T_;%c;$Q`s}t-akOcR>>1ysHcDw>R1x&iC8$*La?Ug`AhDIW2Ga4jLa885kNZ_0PRBsx{sH@cbT zuaH^XRKKv$5YRz=qBma|`8=a7^Os^jciha8ic!OHRcv7Z1EczQ^z8W~uU|uG zUozmyl7VWr^(n1npIL*Gb1GR38dShUqAe}@Ek-tlFrD=unGBCQzG4ilgciqIG@qh1 z9iiU76xVdzH1s$NN-UedbI<_Hp=omj=Uu8|YG1mQm)Rx`yha5v}&L3+!tCJ?-ZGR~tVDI8EKjM&&&9y8YFL0zKy#W_T% zF_W~cqdj>Z^5u(;@U!V4@Ubi17K6zBn}edvxk=e36i~xznWu!XznafD^0wUMpGJ)* z|Cli8fWPQzEM{E-jyt|-K2;&yrNC-TMKvB=ED-f829)Y8=#a#MN-aEVpeD z2cTYYAF}y@H%Ue$p*Gmh73O+seyF_6n@;Isx{%7QlO4=W?5q>eW!ZFfR(R~5ar~$w zFJ|(o6j!Z(zMV%vDBay&5rU|I28zUSlg<87x)!H1rSx8{YMGRqe9R-!Zbo+qF~mlR zWZ5BWZ)lWwUrb}aPJpxb**5mcBAL~l0nq2^Re;c&v;ON&M4{f^=8Blcd7oJEk;GK2 zX*R(c=E?$bjO0RFV+mc8$+CPI`Lwnbq_;?}u_8h0C#80lHi6p} zwy>NR_j+>Fcs@4mx*VHLPwkH?pvoTP<@ZEBy5^4<(WJXj-yh;f$_-*XKN9zrg>W@r zpE+Q>oo&;IH><2zU&q|vetD3ZO7YRC?=YtW3fy;VdB!ho0)TKQ0Xi)9!)EI??#e!a!8~YcYnm(8BRgx(frrFpUWTM|az9?(NHCf=!ok?bMk^Y^_JJ)#O?_ zr>2XnxNC|J_qU6b$<+RnaQajJgLj~wD?JnY1RUT-jki)dp9C*0HT6X4oMgpHu~F|b ze+(B56kJACzx}!KG9B|C$i3pO>v~z37!AB9mPvZ-u3-9;6;M?o%ZL8UoT59G zXgAEJpBhk2-LdT|#>Sd%pIGa6|0!Tm&)3-%nQAt_BU8E8(cl(uM3a|_Cca6%0v7(9)bgbl7QDXY zIB)|T^j+wWSB*Gzug)~GsMk@C@TpIhCOJjC3yi4LV*9ZR5pArgLp<4- z7#^Om8SX^CQZzVmDPXTeM@!JHO9R?Xw0hc3j@7WED7Q55Q!As4)=1|Fg=ll{msjM2 z;U1nI(5V;te{2%*02R6Dv2T_;eIjz!tO^Sl@8Qa#*U#)C?k(#@ zfeBj_esejgftc2){q;DlufsR^miydo-Q5R-AN3DMa69b(+Gp*EbbT#)_674+14H7g zauT+|llUETr*OXu%|--#&9H3Rp_3E*9Ty44njB$l8TwZ-AZh7e$` zQ6I&hGH#bBU0+;au;Zb^U-tkWA}U|z{VVP>6dqbY`LqLi2Hw#N+;{=#Q1gj5D%$cPZjxlS3> z^l3eQ3(Q3u)(#7mw}H4c>vA4^20bbm zn5t=N$j$PAP?C+12xZH89#s`&=>nD9^wg${N=TYK_qn3h9V@^U!TgmFtYs)c(L3=h zEx24IN_Sci5T_l51`+p1$Go3kmmt_`L_hc9zVe=*n_9`xBX6v`0&2BAVeZ{Ui6r}h z>vSi0328$dptL3R-|=QQz_OJMZnaHt*Ixk};|?1uGh-c%2tN z@6__9!m)7CwzTOVUr*hX?~Ox_o15n7j7mKO6hWWSuI6%L4;u!+KGS?DEas%-mxTX$ zD6Ut4^_%NrTiKK)PU&-R*Uagx#sC2n1tRCJz`!W9m4bFkOZ#gI)z^$j{E7kLbbrtE z<)6ERvW7gP8IEz}a`GTHT}!1MT!t36DWt!$wUC=OJ>qt#UZ3@$(wp01b}^+%>OMM! zj6C5*@;-K-yzDJ(epO?RF}VIr0vL@(Tb4e|BExnV#e|~f(Z?7Q4^;xi=&M# z0+D~3xc%mY@8Ch^)T?)`q213Pcx5?KE5^lp(^VmBEt|CZ-Uw_?=yGw&UGczzM`mU?U zlO41`yTj<@J)7B}xExpL7gvS47%SsRa-GO8 zE<+CPO*7q4t@z~A2T&=b#B!JcDK0k(0#h>JiBm%gr=oEdb@#3ir*&;aWJU zDKpO@lmq;A7Mi_{h&?&nH^{dpizX}hl@_kebH3P71?SwG9L2NwWOH@+$$hjq#c3G( z(~qr&q3pu<$WwpE76hP5A;3`)!1z72k{~eSVus*()&;9%l6zwDYolSg5&rkLI_%+m zDo-~(FZv6`l@Hin%t|xC4+6I+N*zcD*2Ss%^x$LzX#VunTOMr0Ht1W+&G527*fFX- zj#sfIynMxmVTr_BNQ#@1H8~6$=tE~u4*BYZg}t0+K3X0CQ$0Cv$`61e1|tl;-aI$y zYS(Lc25Rx2=~h`jS;+IdA1?4y?wI5PRA5N%$f}FUosDTcfE5$u(9r)2R&Yd{w(Eap zqOjkBASk`;OZ_$D+pY0fHe%=3Fxz!>DJ!u0r!RFVi?V*)vg#Ee3Y$}M7rlwPGf2kd z$g8Ar-4I!@`Yrhm*@;+Rf7y!=t(#O7Q`G)CoH|>}O0Xr-cJDasQt?of&rhTLKDQ{y zO36KdkTo5(8&#d7kle@)fJ;?h^J9l&;U>lWw*eh%joDkt5}7!{%O!pr#Hx=@0B?oI z4x6qY(5AhNwJ@w{J3fq2ND>u{0kx_7nu%Pxs4?XOVsfavRd)My(hdm_zkXIS)n(VAClR&TQGqo#>8;vrxTG7iY~!z# zzK+%X>lt>q!);8eV*gsTk*Ez+R+RU{x382;KH%#-AVWp=lpzRWY@0mFmA&E>Xwny> zG7Rn8mKA<3c2H!wxsVPr59IFGpd{9r+-$q+A|1f@6f0v zJ^tEJLeBlH!yp%Aglc^#0J`EnmWrYY?SU$p*^nZ)F4`u!dv|aKbD?k@1!T*ANZxqn zc(d~MVkp#gPlM$=XzpWo zxnO!FWqF$5eF3d1@Ufp?`4DUIuO>4Ptev7mD+ zsmsqs5;Tcs^^^DqM14aSg360O^3cMJAMN8GQmbM2^UbA zEHGm1SRPJV@rQ=}B2L3UnREo-6;M|)o8-pdv`L}0W@|0`T%bj0o=!_fbKLF|Ok8xm zr!9th-8JSHi&?$0oI}fz=-kPxc-u473RLs&WdarRA3^%J+q;bm&;0b(#UI+3q`?2x z##F*S7X@#Z1T(HE?43X3*lc7X)MaUB3l^OXPFdF@IF0YlkWWqhWq2)A&g!S?o8~*$ zf;SqO@)lH-dL_U3^-sP}xk|>w`3y_*y#WbI*tPU$a9b>Eh(4}UA60TaeIig9clkHa zNYlcs#7!sd*~!zA zPviX4qQ8?oe0U9SQYXHz;0l|vVmzr!Z+a+3Zw&coW_t0?(@$+0q~2XH}X!KJlXBUknw?SZK! zSZ}53+m4EfqSeQM)6sZF>6V@}#~HW>NElqDrjYH54E-V6Tx{FOe|&lp9%tQjc+4>I z<7p|^EI9vUC5G6x?)_~QjZO@ywu6vx<8t8r{6T*}V%sNcr$ljLRwNNIB^3MLz!2FW z#OkHG_uqOOUKJanLZQ)3lAXkAzOEB+R$o(hD;W_zLbh74WD9a5Ny5HhKA&AAH?bK< zOToMqEa%=~b9uF7hFq=ZGls5J4tf0n+em&=uz)pH+yZk8`!et%btTA77!PL74v6^t zdEyk3xQCJp`bZh!&?gP(1c_mSJnXo#t}xNP$K~C=yuo zrmq^e+W6cg_Ts=;MWTjf8=6))uYU(-FcZ9;Ka=4;%`zMZk~x_Wcp5W0N(opTg0jbJsz z^JYt3d_Vt!>zzsV3GyNm`BrzN%)h{0QusgW2K@m$y6Z^N-Xx3%VAiLwG52EEj!13} zUj4i|De0OqV0krvvhU9EJU_?3=_zPyZJT}wo!7wC=puXswvE-ZzW9@N{wi;J7;Em* z-6N3=U1ax4_P>*9g8LtMFew%v3A!9LpT*@<=`ishz6G7)cE8Wn^hrvXI{K=$Qb77X zj}*WO1a9h1SAWWZntMZ>k8$|yF!`cz2^I?(D~>UaNiRQ{RROiw>I$rQQ^t0^3BE34qdUwj4t4Lwfry? zo9D^-h>ysKX=W?;Ta}?^r#7ioMDI|8Ws3<{%0%7D?!Q8=b4k?&{Bdw&1OSv41$LNth@%W~8-h0QYm+q@JR3bqYQcoq(Ga+Xm!iB!cf9Ol@E;Isrt3#Ha zfj%TOMz%sTiefOfm^*2(L5iO(!eOaa(4tgZ%Up9T}fm@;?rMYhS6>s{tj^UWh z&D3=$qU-z<>62M0e{w_#&Dx_fLo2n#pdS*xKGM>;i&8D|fDxbYg{OwJU!8da_;%#h zp1mKCWILUEh(*YN*xnjBgN`i9B9*A6hbF=0$VHsw6mUl9^kH5bg4ow&&SVNN1 zUREa+dg(@k*&^&rr_NU4erMZJ6~!$~mpGtKJGz!R7)Cl!QlMqI?FN1hzqonbU=s74 z=iNYkjZT2%Ya|NgiuhkBbW7A5NB6noFky5*r*d#u9qvQp}{l$-;NfrK5pU6&1t5-foBxA@wb<*jV z?j@{=i3vw*dwr~|Xz0FVoAry8+26Nl9gFE1aw^r$+dTR%i#$F4*UN=`Mv`wb<|&@_BKAmb_NVHktHwS8jq5( z)q|T>geMTZ3;W(s#di==W&r~3V^}+jC{Y|wP+~9SXB1Ms0%skJ`A#DDPVDZ)``t-2 zqpfwW77>T&Id`3Q5vn5>vPp7H^&R~c(U$(Vh?Y)}PW%=%#SOhE9)Hu5>ZOH4&CWP0 zT(Y3kshQn3BbA}MK&w73<$izJtK0%`x+l} zbkP?-<5@2ypewYAfq|2=PzA#-O)@-jB?^`Hf4HJI5}4PBZl}Lk$I1sf-sVjLKF|*f zm4@@QO~5(~L2)_TZt{LRQVHxsVEFml&%WERc}IU&RnIrED71!b0|10`P@3P)xQ*8t zv)}6WP^hH-w_pNo7a_qH>x*NyVB&7eg`=;*;IlP2@SEd~66@bmc*hs;bNC;$q0xFk zvBVkG3r$`|Yd4I_xyY_AcZ|7{aYaQ_SeJ$Hi|2@qA`4^Nji(IW{X`jOv-4vJ$-@~D zVP}XG6pKsGxgPu(&|yM7Is)^AtJq>$>lmB&(ql=S%SK5iCZ8SRM1iv&3@kMDzYMHN zCw`NzBpTpoOkPvNZ>sOElfABIAcj<(PS(#HajPE95(6%qz^VN;#4rv~dVNl{2$Tia zAyJf8e=&_4D(PgIE+WxHAcHyHdL5p)-sOa)+86HUENZ zDVS#afn|7u`-JbIl$0##K0wk@*hHC0 zKD%4J84*L_DkRf|)Y zxIYsI96#Oa4T^e^A3K53e<^N!6%qa9APB_SXr-q+O_^X7vpf>va)A_I@EM*Sk;w5WTGi#jkxJX{t93R=EnYb{p6=25@Nh*bV-gDuMnJP zV!Onh&W^M2$}dqP;;yRnYjq~RHmsh32dS%W-C}l%JnyoDj*D7+4Fn7>jjNVAJTMn9 znoK8C?sJ=r^5z>ykbd4}^WTM5WLfCJiCA;W;ups>Ni}q}UVdazz3L9Hveyth%ix3^ z5UwdPmc~+3q4jk69Q4BQfN%kg=%oG{21$=ko{q1YDhSA; zSQy=kzFxd;Tv+(|n%|}@WwTA`D^sEc$@Cj&g$x~{lX9dfsmE8}$tyevdumSb(Djj+ z09u73PLtnCnY#nt0>ytTXk_K!{0W+Nhx&kl6Kqto^$8Y zv`zltBr{jk8{Hk;p`rBWo6dK6w2})#fW_q-jpL0az9j7G!j)N&$}8kf6Q6OJqqNo& zs&a8~Ak+~=XBQOzfY|igZMRobnuIAZd;OPg!}YDu_hNp!{qVCi1Z2svS+21VbYjrR&Zv-7r}}lzBTk>nxrgYqNmb?7YF2z;pw}~UPhJTvwls081*gsE4fA&zC$_k~@uBQ@^5A8?4*>TXKk4=Nh zdAYc}@9z65ML|KKpQ(>^a(>>@*49?D9ad)sm3&|z^lg}$sij|G{r)u=y>PFB42%ahpf<$IT-gk=QtpR;TYV=p>}9+^N#t7Bw||TU zLti4R)s87&el*Z--scEWv#+`_q+D`(t z8#YqOL<`B^bNSBYmdV{=Y$-xzMcRc+7;Ps$E*kHAu1DTW{ge2}acU%8QgoLqStk7q zsJy;?bJ9EsS9#L^!80YdIrEOT&=DzmFgLIcN?ctKS0@o1t~ypJv2t zRkP_T7HQm^7O6Z2YLAXOY?;p1=7yL>4Y6?08iSO<6&iki63%DoEe9tg&C2#kqR?5m zm4`e?(t~F@a*a<&;Gpg9tSa+svdR=nM}7?7-}Jdx_VeX`6!lsU@Y(-e^tyq;77a+1 zHgdb7^Got}&HnJ=v6PU|dUsDxKx=F39-&7^8FIT$OJ_Lo^~M+N!M@!Y`!hUt+ksz<&~Tux*Nd+_>trR~*8C^|)5|4!%T)xtN&W-G^zj^Bdxo6`6 zLFWA}N6Pf~-1PiQ@6e|LA(V0F%^{sIMAO;m{l6bG|Nj{EbovEl2?YGTu|mI549q_?%9CzPToWu;p)2ayVJd8r;ZiN{ z1zx<@c52GjE^gL?X3}&gWhmt$LopioBaflV_umpGHjZ-vJJq!;y6e|6JJ=1OP6Wvfn&fJTJgNQBj{-P54UrV|^%F8v z4xX$w>p{x&V;SAli*Y_31xA}Aj0Us%`f!B>MUqf}{v=0J#xcB!*-UI|3_1Qc(SFN6 z6c4Zpx}Q90LO6%G0s&xg?J(L7-B94SpVb}nb6x0;Xu<)h1a>JY5Ad$b`TsW+Q&B#~ z5g-KVDhggwP;a41qN?E<*02}?Lg?(F$oXK>=Guqrf<^=ItP{75Uy{XdwNW9@Kxf@x zX3zPS>2P?X;Use{j}N=`?r6!5NnHxt;NP-<^B|*3_dlXZjt7Jl+wHpZZ2IDtGzXOM z6#zN7gx=9#$;S~Y57MrN+S=O>R#xTVwq5I(cpR-M66@+Pn6z8O#{zf$mS)GtIb*Qa zYjQFNN|0oi^-FoPW>sO0(Q;JzS5p#w@L;9E zaa|o9R~Z@PO@p%<{;J*4Z!(CV5oCzo1Sa~9501@@GZJsRqB9g9Q2%lNdCQlsMErJs z=5W1uAgX?TCzGSAF5+Grn1A1lcT|6$Fo2GTz*g-^0Rfx%f`OH-Zu78<;Vkpx9nO<# z-bBmb+BsWc;o`ZhEgLU3)wb431Q`CH#YuH8Xh;74kfKhrSW$2GnDkiE-bxH)HO>SV zM+Je8u=tAutiutiz0Q*#8=OyDqh}{>4G^<&lIcsC&s-Y{G1_i^%p=a;^m*3m`sa2? zF#7#KfSg{?03(q~cb4yRHUII3`lJW0$(m~#aW`99%(dJP(i-y=UM2-Vo`JI>98TGEk`w;sw(py6m21q=ZH@|=0Ev? zyd&pZ<7IvqH_y)p%_z;C4gxV}V}HENC7-F?xty+O7OsrjAcH6wrM(z>;*d&Ws2V?? z8IFCWuRP+_Uuw`Oiu8rdEwdgL+?zBb|2G`*E2f`HLMe}--S5fO&>k2#PNcr3%3`_z zeQ0Rt>ef1qV$Ez`O<~~O2_*xWv2L#GERSY?AB z-~vKN51dk5_U**ykk-6Ejnj82AM$pd39udy&tJ8D+ma@4Px4RFp3D44`F;bpv4*v) zO!_I!z$q-fgnUof0+1U`dxV;mx`r1MDrl@>Wo4yPL1Iq~kOpsWZ+l$b-rg?KF#;f- zwYCEnkuFGS7lzE`>!r9REk`vwjg@jjGL&DCS?)p6Ebu`MK)MPN4vBd*gM*=qb zt}Q}^a7IKZ_97+U`tSDxz5L!FOmu1AWNUx+ceml~RBM+W90%uU<)0QIt~`|sT2MHn(yPlwwRI1_bP_B2d==g*l2t;WrK7D$D+u!n(!cI zy=2Cxb~!<2BTLrE@Ys{d?n(F8R_~l}Z;lwG_9?UqjLF&8zH6(hVVb<^QVn^bb82s> z*lZj<$yOATb6F^OEivoUTsJOG>oA};QSBPVQEY@#77>9fjD)xoYxQ=OkM%E6b;I0$ zG)xiupy=-zVH$@iOY50*(St@s{4f4oT|f+)t4|dJlxfu{(vhlqw|-EwnrU68*=Q9(j4`i{KTmX`S}QIGy= z?36L3iKjW@L@3}EuLJ1V*n43Sk!@7K*67{E;qTpGa-S(-n;#%Yv~sE4aHvVPlisJn zQmgNC^UX0s!<|P}OYRao#k({y`>R!ir!`xNITiAYAQJ$M-Xt$uJ2LTV6bCK90`2-^ zhIstM`v(*~`^S$GJ7z@=Xx?Ok_QmIGo`g*16G^v7gJT;&=O;2&NZEWi=5W2W!dTPd zwyW_Lk4@qBsn~Bsixb=xUwHsmd$J@i)M+scN{N^?_O%$w5_@KReCUQ)pAV`^vfcqL z!aZkh_ZhB0@T)jFd0fU2quV4kg@ZOdZIiWE*t`(u-yL}voa_&2`g4A2Ffi(jL)ReN zi|Mz3qf&4RHgX+Je|kD}dtA{8>VFh}eG!gGdbM%y#2L$mWb$jp3H-17u@Mg2hmN#r zGtqnVQnfM$r zjQ2b+rl@Nw?etaj{7)t9d?)vbTfIOil6GnZiUF3KkkT*e4qJ8O$#X~5<_Zw%V_n6U z0st9LTXpF;lt?*!U`kun!kw#8Uihd4QVr<0_QPd*DXHo3?V|HG7WVZMm&3icwl6kl zr{xyrh2?rj2leP=AEhHgPj&Jk&>vJeMy@&6y_jCB?;8C$dV+yQ*SP2SQWvM80ioSv z(*|hmhlK}TN%)~2oB6r~Yp<~@rHe09DNK5(#@Fbrro^UtywKy@zc zMAxdPwlqWmS4;(-E9dwHj)cYrPiD(f(*Bdi2)1Da9s_dR%-p7FgBZW#99eP|hoHLu zI!RMT$3`{=9jrk;!(Nf7={R+@4RZ6tUU?*`=*}c>ZAQl7(UJigae>b0pT#=9e}8rV zih-tF_t_GBVqY~|e4C(3DO9+dNl?TST|3NSyYj2CLw{+D0bLTR$GBkmpedknW=pp@ z%W4d70C|)BOh<71^tI%V8a^rbwBbmUF*~|;b@0w4Z4zQ?>*HBjM#4&KN7!`ehcNZr z4nMmG#%O$(Ues$49&}cc?V=xeMX4Kt1wH?kE?%+t#LWr*#3SF#^?LGqWC3hq&_cOo ztsD(F#i z9<&=B-{bWz4V({}GvyQwK(>aF_%?4YnEb`=hm-3}FE)52Yf&?hLsCcXk0Wo? zdB^CUc zcUV==7P2zDjBd^{Bd8*)$~dX}U_q#cby_6Xd~mOuc)y!AGDH&!#Dd!0a|AwLav>tF zRSrdj<|iRksGGII+3GmC6Ex}x$DvAE2RM)BdBu<8#quPEaeS~K!>3Z)cRR&_jTO(L zvgC1Z)Rt~+yaHDJ!pPs`Z!gQXK05Q$S2UPDd9~Js@q*~Y^(4^!DmIqrK;l(KJ_YlZ znf(~6?c+ak97y)jqLjio?7BUr5Yx+UuR_l_jPKN< zPxTijj9!plu#MbD`}v*N#Yn{)WBBl;4rvq2%suB&;M5>_q1*CY+(srqfbX)RcQ5F5 zx*Aj;+8sXj=jTb!Q8_PQ^U2(v9;Pu83IuDd{U&|X(GSu_{XByQE(d)|!PllXqL{yk zRvQqO+!xEtn!p&>w{laP8Q37GONT@Sdc{2upx43V`tr1|DM)40Rufwvendc zG{4Wtj|O>Qf-_%hiPp8UwQ$yZPC-X_h@N>|T7_!qs7&k^pfOwidLbzdTfgx|tRY%~ zh;)e88}Onc88&k0_ z*4j4a;K^#edFq{MnDf#C&vLxq5@<3U<0R`EXCg!AFJ5x)|B2&yj#pJ@Apj9W@7Hi^ z5saSHFMJaF-7!dd);qrUlRz<~3IyYDSURSgPi-)RKqYq#ef1YRzF6`p{NpF))WGCg zmU`-yL-wlal0##E?VOI}$9x zT@stQTS8rL$Q5-g4gmX`?sJt<-%qY?@xo&5y7D}p)a3xJU6TqFll#X9{hhn}`)>K_ zS?Aoqp0e{HelL^vHHgHcha=_XiMoWo!~;YqC4|Hr&4>~K9;q0H zmx;4Rq4gsllINxa+V-y*o0e+D^v?#A75~zpytYrXoTYBZ?Dy?=OI;J1CD{(^wEY%* zFa93v#Vh%xZOGsbmbh58rEYoXHXPA4U{WwOrTRhNX)&bB0%}Ivgkv_5d+%_P z=uf{|ous8!r_``xese?K{2&>z{D2csSY|PJm70okOugrTOzP?%a5BKsTco5jTCN!; zG;gzd6q)FPS>>W^Pc@4k<+F{XiYOpISI9sa)Yq~AlW@1jC#>Fk$ z#cIr*>3Q`(0W?vphYI#LpRTXQ{om#;SS_l)cRid@{-z`GjY>^bwMN{1zq+)PPfEna zhdmpp#USq1e|COOX+4AuRmG5~+;4sW3N^QoZCGX&7Ui`#Bsobr-}N&jhIb^-H?ae> z_IAhvta>_Sb|YSemN`f4hfLa6?*!?+nf%dUA=go-kk{y-dYIRpww+mrEHAYvjK{;2 zPzIFQi%kk(LqY+MMyi)~c6SZ(6&OG~y*bK;x!4i|%LI%$L|Ncm>stYCElrcP-@V~9 zsbkD8uCAWRh-#egjF*R|E;&7(ie>c8Pv<+=h#BGqZZ!ticDk}vU1b=$jUcAuAYGFM zFB3N`s<7_x#?7?0&aSD=KGQXe$yA@fsL4J#LZO3J%iom~&nJPYZ9$2o`ch-o!_~7h zB4qEE+B{Av3QsXqniHN5GrL{<5k^5wmoDU+dqvMW^Z@8!UnErHH?JOJ09)qH6wSL< z@&*WsICRkEU}r0FLl$c!ziDc=+0kBvs?r{CFA#@!Qp&2QeRj;xrvXWx< zEm~aE#TU6VQ&W$whrbQVu?#co!*nQsg?p^YgTMc2R@|B}=@4^X>S_zF@KrtCwguS$ z8l6#2g(l~|{oNkPi_$J`5A9i@y?cb4Fh|wGqVMXbI(B>f;yN{qRNDpHU>sVL@_qa` z-Ha4TIW(OdFIv7k7q~3W zE47hI{m8=m2TvM^+l-AM;*-W%0&bDCBPe_MmBIsrkdiBaF`%W@WM}%8WSK{cN3f(k zTQPc*Tr%5rcav{&dVK2|VRW4HDoY%D8;f$8xV;$?IoH@I*+fMFNP~vHd(caRNkI3T z)%n~b0HRu;A@>0$IM{vSSq;nu&s1_0zr5&LYcMuYD!$3nC{P?DdcAXnMM zd)cbc!vPaSW>k<@cjpy{6rp@5Y;&fYr~eG`MPmM)3e~ zkQE_A14mGo*E~noP-sGN^qD){LkUh3WBsl8wQkyYf` zRI!cvD&|qtIkb734c&i;R1+0O41dkcJV#J?iHNZQ4UaIh?VWV~heW0Lxu2SP7J`hX z2BP>FAHD*uGC|gG(Z%q3P(KfMP0liN)i2-gR*k=&M!a!R4$#`bxtbe| zNq29U-_+BkfG_y5A)=)u>XzPntk%xesE^y5YkDym*zX55U};+|K1u$RJka<6_Poj68o<`#+VFeZ=Nqd*i(imADxFx3L-*OyWC4Z^kfx=+wcZkL?Tt zp_)-35&qb&NbyKHYh@cjn!q2rmu&$@7HKr_hkCxIo1~zpI{F}2yZPGnz8nSdIoEW| zgdfDBm0YPCAiM=2aXR^{3Xuf{PE*UWjBmEIt*w1`r;{}Vy&}9p@yW96MhJTTLwy~F zSXhIk4te?cSAd*rO4;flq{BL^eTZl{kC5BwzX54 zwk|WSEdOd=ZbPk_T;Uqy%b$zzB$_)Z#G14tj;gD&kKTlTT6nJa?sO)S* zs(UWY=bgp-dS+#Q`%;=gqoR90q@_{;x7^Lk=Tp5xtePXV%Cbae_A*H=U0-mg48MZ^ zOb4Te_j`S>PfCp3+u$QdJL@d28>!yAZ!_>F>f!wh;N(nB8-u`Rdn+g1WX* zpAthhzLxL2+yq_l&Bdcb`tr))oGuSRSuuU3mm|i>I?vIPpWs zNJp8%e9da;Tn`+p#7t0lXMyD}vC96!3|{Bes$BE0OyEk+@DsyA{ENG@;dVuvQEZS6 z65~Xtxf^KU{lKsheNn6aQB%{~I*oFGjp;%w>V#0|w$-9(|4K~Fro*d0hV18O@CVAS z`I6hMF6$9(39*7o;`cFpLV8(wxUE%NZb?+z&70nv1zHw$j&E@#$gi=)+bnaTSXz$R zZJk3R81b4hx;&=*Hnz^qePLBi|5ejx$X($v;CJ;dyE&(tsuD-?LP45;A^X{%&?0Qu z_wu?JYxTl1j;s1;FA~jRrT2L4;SG+?9=*OhXV$;ds4{kj+!GB5wE9b{5^b)))iW}W zzM7rqI4jIKbvcW_L#iJ6AMI+-Ee~jT@BiNpv>9<(a))lyQ!1g6$HuSVPl{u}m8XZ^ z0+j*WI)-M(2;UoA1h2L=fE-5*^V}H#sg(URg?J$F7|7qxB-kb($ln7(0UrNDWG&%% zQpY2Bud$Ij71$!6tWIY<9o>cAU(F;nI)GpSV{l_^L;wrm2@VN{?gW<5Qx7JzJil}s zN-8yZ%mM|ag%wFSh2nS)g%B?hlrKT#819y%rqDS=>*e?s;a?!O1m#uEGmjvvp`~xt z=m0#;C3liVZ@2f4;ej~o2oHolePr9r8@tKyL@nyO!%X5+^~FevR9TQ-B<@vY%N(K8 zq9ZYbnQ4GmHIZO;#Q9K=uoF`rgE7S9+Wg`?w{)cN?o8{B-Xc=oVxMhy``KD zE|V|QSv`a$c-by_c68p{mmM2Q3G=PW{L0B0!6k&eWq4=y#yw{(&^o;_TKFQhYoLxC zi;^mt>_T09KGzc)sh1``UvhJJF|oKP9>tQRfd5VdtLdgoy{;qC?1zWYbph_4-qz9? zwSMmD=l8F8eFyH(UmeH|%3ZL?$GmuMW=eL}i@%vBCnxuwN+CwjkgrejN$7i|hS9$l zVqh}Fzs2|B>O1>5lGAr5aGh~oCtfaUsQ?2-%ph%;ecVGelH3iSVtG3DTgjE^LeUw1pME6 zGv|BRe6(_II%eUy)~qV{_FEkzc2)i~?f6glHC0+_y1!Dd79X{z{NRA8J3at@`{Nb> z6>~P|5Ni{F1|Se7ikURA6GjGhlewfipOmL-u4IIt+2EJ+!OW~ zmINje>Anp@@^X^1qyp}dOF0tRcr7z=IP@~?Yr|ByokFXoKHYc@p~ibZqc%A_okQcF zW>GSNd#1`=$sZ2uBzzrgW0OeZ3GTGG1y$4Ne2}KN9i_Y*l};Q+9g#RU4Rc&X0=0)8 zy(X4e3w=h+AcnlI`dx^`nEI0mbeJ`?NGp2wy2YKy(Z_Xic9@N>w4Ih80OJpQM22jn zA;$QR#XP^yGVJZ0`(1l`_^pLD7~+UYQ{pBzF`-1DnMO9GDkMs^EgsHXfP)1>v*ck}qw1!C@=i;EzPr>C5L zdW`}lHhzt|w$8e{b^1iI^8){!A=dpNoz%LW%KHK@sk6;&EitqkT)?nvJGhBFiK^4& z)@6RLSC-a<_BPyYh~F^DnRvA8ahaIOY{p1=DT<-QM8ZYFMV@^;$NSsGH=b`L{I3Wk zrY%GEp2gfzlV3gZ;)(AP$xpA`Jl;fZV&084jor!B9o1^s?1hPD+urKelM``Pox3$^ z+W3B2H!=GB^I)9=X=FEZ^q#K!5^32wqN=O685S}Mm z;Aa1xZrBe{3ptJcSl!&vEJ5hNCN6Qp#+kOw<8utW=pBMi_q4_=d zdmn{7EErH&gunzEYU;yM``1_Sy~yZ9pdgp94ov`zK0}nEP7tWL+Z@QD zQ6}ULV`W|-GB!7byREH^{Py0PmvOhXeCv0CP3FWUt33M@TArefE1&>W3ljyiR=8Bw z+zn_=&K8V+@tV+gwXfXHRIs@d9lhGPLk&Imo^Up2eje)R@RGp@PG zXPs)^-R{Ot({eJ+lFQ{Vqtnuh@=Rs2$BKNS@0ayDWiC}BIa1P5>wYcAo77ENV0Y0{ z-yajKlNue`*uBO&AUfmz**Tgk|M+&M zV%B|rM>}*bgU|Er(2HtWuwM*%VyM3KLMZ;8QPjEH_l1`arN;s4x@mOhyIyJ6AI$fr;Q>h#O@Sw0aVVeV&)!$=UeCUxVi@H+ILausvk9;=m6 zYNYX_E}%~biDkc|$p$_KGXW9Ch3~ToNU#uqY+3zH($NX&6fcmLu%Cyi%^`#uD-0(F z?D63lbuR)o#zx}&-MqghK)f4_2m$y47%YKWj!f>k3gWruo{c62MZl=`1UbkOF2=~w zi{P;)zy!*+fSiO;;ZDtuiiHp~Fx{r1a^uN<^h!DeZzvFV4bE?>atM&{Q1So-TOA|V zp-_?Rmoyi|wgZ_GJ5n!nn(l%csZo5?Je{CRcKu?~RhFS>P--h~IVTMUau2XkvY+m} zw!{`P(w*(!LmekWIp$!9a06U<09$pR2iGQD5N2-Z_3~<e?&WvJ~>4ZCaxFKREwmYGkdiu#vo zE;%oKS4TI^D+sQo^ItT*XcCxDt;EqWQ3vSIKqnNs%eLazv9y@gJ1x33bnHGBT5Mc@ zUY^|4wKN4DAb@3X>`Hp$WE=l1K}e%0heBrxa9}DVTC|Hr=H zPW|UI#LU8n6Ni&?_gAuxl56i~bviMVZ(E=s=X+ipAXJ0{I)vqb{~AZmffLUA-TYLi z173&>6XlR5PqyDxc=`2eh^_N6whiFMcs-V;YTUpCI~MpDHw6d*cgmqkQ?A84emlO# zdEFrbFhi$9-`DvkohFN!7kXh)nyW4}S% zeSTbg^|Pa4!nD8e5cP&S)>U=e)3R$E;u_`rm|B6ZGoKyb(!~bSV`C-@{A;PpP5Zh~ zH6QlE$A4eqj~QScMEUjIo}c&^{Jov{)@HS{RL3Cq^IAv$cK4~xrgUBjsl|AH<3R}HRnAtR?p{>mR@oSH)IcP0bnz=`+}(s(D|U+?Ku)6oF)Qnq&bQpg z=C)DL+%iKO+sZ->JKZUJ#pnH1<%#+#P_3kW9%MYXa#pf33o<#c$xyepsUaOxP!@BW zE3nZ6VzoM|g}QaEWSuvVuv~ACRpEt*u;b}Y^(CkMcm7q^LqYlT1*^9Ub<)R}Jnp`V zweXFfnuWji3Ds+p7i;$RRR-ASsqm{+2of&NzG!J1imhkX{(SXEY%oaj4k~pkRTrv$ zC;fHSFaz^KNqZ$c@jvTCI7dLwCG!QH5%)l7x)ojD2kBtG&ZP`-_lbaqAU=$iy+ks> z7RCRP!E^vf2%zuRS5C8zJ**?-L(W|i9YzDS6tln)gd69DF|0fto7RlOPL_1VhV`r$#-WUmQ6(%7## zW^;V&6Zjb6_c))cQo1@+v(`|Ao3m`H7r7W4isbK`RaSup?_E8(RGNhJxou0?Si82J zo6LCp_A;??`I%)$J~69YtB9aj&5FwpT)a;@Uy~nv!Y|TdT$5OnwSIRvj2Y@f8Aa9& zKWHl@=SIjAF6>>ceDU7e6W>>@0lHX^*m`bJ&8Lq@#O$;|eg85~=8kIy!!}o?E_Su& za|Qx@=av4y#?4f5sp)p4%+7S#jTuLS3qBo*vohq+jD}Ax+q%B{%*C#c z^4Ij7vuJqn8Hhmx1L>k=jiD56l}@<|DvVHmjImhnV~+6!u}hrLS@!2=fIt5MllB9Y zY%y6?vMA04!=CNgoVOQrkV|xuxmmF0UE)Lk@2*3v&0SI0MPqwb&RNypX=-K)ALMG4 zVBJgpOc24qQ+}7z&(-qlnn6F&3FxXRDvQ~0wdGhOp9ib9;K%1-p5d%N1Yp+` zFm|^YCvvC{e@I|K%xk`F=l;ceq!oYh*Ly3Pvj+6d&ubCseB~X*jI(+1$y!U3d@jjy20jLn1GI z6`Br~1@>DDSWG`zR1d($tG&I2i~BS>sut=CJ||b-x1oOVv7Bv2^9?+!9rQVsaktC% z&j|9`Sq&9adv$TwD<H`C!TmKw$iDO<5O_}83JF62kZ?XMRFa@?WWT)6bcx#rA~<~ zebjap@o5Afz9z0S`mcb7UB_D3~quO%n6qdr^6b zbiS4|6)cck`h{w{A&kqxL1L4t`9W?LfD0|+gbovTy&B8AEqh!giR9*E?pJprd+z_- zZtgHUrAF;LVfKT=NC*K~^2Lyp&j}E#5f7m}Ce$1bL-%cSa)Gv;Vq^vpFSm!NvKg|i z4NO83ghPiE7Yu%12(0Z(5pq4NfnY*{k{4ZWy ze9ZJ|L1X+`tTE*bV;nIc-|{(QFGocV z8)ZK+)*DtP-CCQf4dG~)I*hRWT*cyO)%vrlV>x%SV8d-D6)8$@!cVQF!MUs9YM&uM z>-QotnL`nvvo-*yJ}qpApNE9DTe<5hc*dC9&OGb`9SJ%YFk{cY@7w&CR=#*j~W*paUQxHBn7+9woRY#(e4 zDnluZ@i6*+s0AH}iC|T!uiMOoAH&up>h)7rI3D0&`V&$*Uyxsa?1mSuF*aVOlIZqG zlc7a%=x{cDGvOi8&|?W=BgY7-PLOT2+ipJgV|gR3&ZlC{0$O^c)X(dkGDQ1%H^gSG z9y<@G{Oz#RliW!9D@J3CZ7ff7LCbG!LyD@sc%%hF{eEA~Z7{1_6z+MMp}!ww4cBAH z9oOoohDpo+>|yEjjs;xw?0%^`z}hz*!24-+b^!5)>;|kvt8UveS@5ATBJr2&G2i<) zT5tLz$bP1QGn?+}_AlPwcKlZ&ML`E4_6wfp3v$oZLQ(V$xDtIXEY=EieacnpS^9L> z3{{mlwXffE7YW)9s2&K^4n5e?y-$em6aZc^lqR~mxFEd)5Io1CDw1n7syi{CRoAP* z-xrHyr>JgaX3-G%3|)zQAG-9U$kf_mwba?#{&pynzlJv)az8Q-A>0sJ-syO5SAAh`Z8Q~4L|3Ed)K6hJUuS36ks&^Mkm-l6-`Tg( z_m6h&u)t1PuA$BP>de#Vs!=?aUt#lZQg8BJ8ap6?N*(G3^;}Ee88Z{~-eCF%eiBC- zSd3k>_cukR{^KBQ@aMkz{JDiJqEDtnh?4Ye;WXyWkwhl646R7J3pOy!_cxPzu)Nk%|=oq(Lia`_BF6s-Fi zxMTNfU;F1vH$^7DGIWdDt4*sf;-TZ1Y+qAk43I4HD?7)tn&~*m?@2yC{WB%17~f@b zTuv>@WN+32lsf!)vn#mYkc7bxS;{%w6M~5Sj!A4iA~lgozh-I9F^L*{ml2S9*0y_$ z5o|B|bZ0*In%WR`DspgESzs-mnQ9(Il)e4K-ENx>UwRyv(O>l#WwjK0-JOfj>{**R zTp81cmL%+BM*YKIoi6|R?0|{Q+cV+U<5HK^E@|!Kx-1bnGoAdsWW6ZtNR$##OFo`D z7pE4%3NWEII<=OkEVQ%1jvdEDT>jO~%Ylu46nN;4u`FVRTL&KXR4mQFcd^(@zt~A3 z68{*4KywErh87l>pZn*t7$-I&L`6y{jLqc?$LluccUbnt?dR4wlN--%>O+xgk&tR| z(pjABK~#(cCY4aq-{3lW@|>}tDj>b_Dx=7E-~N86iZv8;-tv?y%p+4AKRf8=Dp_(b zk4nL*Yu1nYR~~V6zbpg2M8AuuhwwbTFO9^5hVURk+-#s%S^&q+>X@}gLW3U>5ePYS zpkt^O=|^>LUY{MW-?V81P~QBEcyapC>(TxNAdE0Lo4aKoii+q>%YErHb?8?Q8bn?X zpb-Vq(S$TF>xTp2d6Jc&TRIQ9Ec&%K&{(hoYXZy0;m7lrI-P5x?CmYMf_Ou`RpZ@J zjVgaBS=^6{{4M$IFsc_55qv4zKd@)=ii2IpX0ml4PMkkf0lL{N9HH>E^L89z}Zb+u}F{89up-s-8l0F!5t`m%JVx-~bFcrSv z23f&Db(lc&u6jF9{sS4%(*XawmBCtVoTjP6+>*S~NQZt0 zUKx{MW9m=WOm9Brl;$c+`^4`UC6NZ2v_&d9e)nRTWMbD@t=sr z9@Rd=fVJ=t%J)qVf#+WsVZsm4>F~}U#_V*MG|{FkR60u98T$FUEz7pmvQkG7IMb*f4IA#gbk<*>8voDf-LbOzb;>nwisH|elLBuWPXei1AwF*?qg%9*%~_8?T45Xa_6MrXyS2Z>7d3x9 zEx^$%`2OYH7o{M!>0?%2YB(KSAmsyl1}^hZ>?8?h>lgbPA8cgnl&+D7H3N)hVm_3W zwNwlhYy%yCjRlwpdjRNcU`i8asz2Vo!hG2{vdVvB;!9<;o#yoG|S%!ZD$ z#f|qxuJ;xPpzk*+{?%1d^`qMykn7*Qklm__Fl7J#nw6Wln_OtQZR!l}!|6I?cun z(0)2q^*zP&3+D6I;$&}yuf*ytr`}$zUY<6*)G-t+swKgamBt(6#v1>f<;NUkzE2DZ zyisFbdi>BNae*wB5B~#R@e3XEOm?f6b!cvUF1PoHGGs zgV5ixK~jIF&|}@Xw>J1RciD`7e9mFd)<&WC(;M`4iem!m*}p>9ej--|&2()05V? z-s-cT1(pTC%WopSD>%^D!hCqtJ`EZi(p7=89M8fhg6mg&wN;l5^P{2yZqyo$@m>Nl(nD&e;u60GHen+J`;yfhlZZewppYQKgPts z{pRaeu+WdN)vIeQo~{+X^h+gl7SVtDL+b8VhQG~oo12R1vA!C$V9kNo1vw;`Ldvcs z6_If2zdOAiEbw3W|FbUDfM(`z02BGZGN1k{0?r(qYdxKT6*S&K`a)4v}_o)`! zx&>Y0->HLMFgyU=riLjTI<2%e@pBFGt_jpZIM)>!HO?XE1P1E?m?ZaUvH=P13{1zdQ$Vd>oh{=`CrFvDiR4Q#+;#kwfpAJcyVh zuX;4atq=W`6QhxUQ9Eq^IZLQlr~7R8aEq-mtkc}*FKt?Lgj1)5iYez`n&Vzab076` zxL;h$?cW!P*ZWg@xaSCLCwh*nTu$e9^ql?LDaY`DtBn7?_!=5k!x-_$It1mEo`lm1 zXF~q=LQ7{chIGsO4gsa%!hxlw*OeU-cn+Ne-L72 z{ns0?@;p{&S>PoWyCnS8gXvS05+BUc!o;oeLkDN# zGy-13=p+eZ-_chnKEtB-yPJCFq4RCJR_l{{oUN{{t%>V6bugT`AmI*uAa$>1ZIm=k z!;Y(IVf(1nO3UNj6L2tss4bjVC;<|uO=!E~l^iiZE zo-v*oUN93M-Y*#5cOIDnyux;g%o9d8q%#|`8EPYn_v{tpYqBXaWAe?nN!Csbmg>y} zChJm_Orb$c6#S7F&{#a3mpW{(6K>QC@Huf2Og;{l?S7@m_OgVg?Q+jl1!wZ?Xr=4)@x&ZxO1AB7FF@3Gwsu{6=3wFK{zrr>7r6I9fcnCj zp18*qmHCfR(}XAG5Wn|TYh5FR0n6>!cl4|PKQeGYNm2mHp|#nUb8RWeQ6vtrNQ56i z@59HDVX5q|rMn!{0an~S8P04(J1tO6U$P>H1~r6R!^Ns1oNZ|7smn9r68Lu1Bc8ZN(tvDII85og*+fw~+_vVTk^#VhEh9ju|2$a^ZEwY=5>b6d=8$Xruh}umY<%~|)6JM}jZ402s2dn;J=TKlm=~R&aX8U7eJv3on`3zZUrAMok$IeN) zW9z9CFb@7qu|lj>T)O^fJ~{YAr1UBj_h3uDWKFs*sqke&XVeu!7!nIu13o7Yz9UP5@lXbnj*KkTQ7e|x#QGh1BgE+Ou( z&k%vxYu-7ohKC1aehtF&vA8graXk7tQ*GyRmC*r^`s zSvz;U72)547-WII+Se|o zYpa{Wnw}hy-lfAUbXa_@vN*eoCgJG*)q~)Yz==pQhCN{>#QAGWHidysLy|4^_DM2>c?@{{rvht7&kcgAkgW)c$HDHJpQTXLDv-*|JIa>m<0W?%yDUV46fOe%Lw3BCuj2A-SeuVHet zbbHN!bP*G(9*{DY_`nm6ST?3JY4OSo$!V8I2$onD9d^YtrX&O<1CxsZG5=qsVyFkE z>V`8BU!ad}6i5SKpGxD}Fzcw{<`v`K!Xtfrwo~;vkSW$2$mY*b04hc_U+Va!j_9u+ zfBHV7oLKQ}s?~|%SGRx=c?Lutzs`&f)7U@zC zrkwbH{!^rceg_lwQ#0$cWts&lcnL&U`2&olQui%;GFivhFdp!C+3Tgv;QKYnhoYQe z(H9KrJ_047ybg<#J>(_1G1nUD;>!X49YT`SevEhQe$X@48?TECk(l+RE*X!ZlCfaT zZ#L-F$|V6YzTl|M@FrSvt#ppSi?i4Q$R&4g=8jvpoN=KbY> zWG$Xlw{Hl%+F%j5omN+TB9&F(%as&FS~%OfT~=JN=ho3VURhlBCn+S+CgSKsc|GRd zzRGuWqWpbQRVoHuobj>fmSKAh7xvc{)k8 zw9b)>1^rJC{;hu#?DTkJS)6} zzBtse=*mZh%2^SY-SWL1eEo+Z`HEa%sRykLIsAhX2qLNPE+HJxs*3f^>FQeJcTB94 z-Ztf139l(pFy*^$>ovA5`0m`RQ0?g9C9q+Qmvb5O`8fE{5c|+O`PkpQ_YqJ>*O5>k zJ>cA~hi>e%EGO3vzL1za3a8@WJ1o=jYy-j3`e?;dmi!oU-i}AeA)u5`C50Ax8NI(G zx1X|HMXoG;nBBzD1*V6IN|I=GfA&>38V|lF5T-HZ7#%fl=U?S_0c`}~LfG$wCKr;X z9{DjvJDoN|EHZ~%dS1jl39A_jvi{O{%qO5{c#|ylmngN}X-d&_;Fj@`mg7plxoA%a zBdbep!*zcJDf`rzgP5JB%XJB+ z_**RLZ|I8}DA*3X1h$L}`13s#kMpYLeffbDMW)D}gj*(_3@(-}ikT$WSN2QrRMb&{ zd%7tD-ecv8)(1YrJ@_(M=BMA&FCCpmj?Mr=z;LM+;XsY5h(}kZS-4Bagq6`y;PKx- zY2=Xxj5TPHn9el1bD4Ai#VYQdy0z)iSm-Jfnz^T?B$S8en-dn#%}lP`7i--5fR0iGsPU)C-?=f1GcQa>ffx|CfSxXr4thn;s_l4-vc?Ko`V zBa!NH_FJ;B0CFJF?hqrIW7E~ns8{0lJ9iX~5BI9w?gwt4%^W*Wn!VUfexpX`W5i(@ zD?{VtfqIa|0Hr4j=HK=SA@mdPaaH7g>;F@kqqAAE`w_>x$0wmbVh>n}d<}7IVe&ar zjV!M8rycekNmF)LX7NCIMmR|pQ!;;yD^pXL&l{61?Fa$fyeAfL#0f;W1&WH@zv!~awCVK7`HX_@|^?2h8<{C09ww$43hmU zDsMm2x!MJPoB2IkNkz{QN)C5`I`ibvwNsra6K+(Ri}Bo*;1x#joq9rv58lnVq6Qh~ zC)1{jxotx7S;5EA(jFX`6^*5ZRmcj4G%bB+4LG|0WIjGCPH8!&X*2v{>ZG!XOA1Ft znbagvTZB#JUvRs`J0*gmVr~vQT4G_h5xQoI;iS*9Ai{yRjD%#>?dlz*NNF8X!Y};9 z-Pn&fo}}N&0fOnB0@a)Xm9eM-9|uyFoHQqM&^%F~a=IeN3=f!m8u05_bzf!Y2;0p8 z7^#wR=(8&3^iL;|j}qcnyo%#q0@Px+Igyq@w%`xoH{Bz^=NRY zc_?g8!l{>W8XD+er(B8u?kNu#BNFXuCKpxo)oJVMTaR8IOdP)ztP}o$7t1sLePH1_ z-CozC@2_&B3BZ*zPcF0{KZpZ~d~^C?3Ed3;!B zmdItkEs?*Wy!agD=;=taflfzyn4Ce@;xFc>V;Wi=jk|p#DZUmypkIb3O!&4vUh)vY)5acxjIUuFo?N49*%Rq za6R&Epvl@zmUH@0+uZdm6KnhyS%Aa=2(ZzsRlGq=t8%KWc*c@*hQBGNU%h*-DB>Vya{4;+ znn8bjPNK@`#FDwyiqW{B;NSL6iv*qKyC_$$)2(E*S~<-zJ)kjl2UdvT>^Mj`iu zyI5s92-+Y^BtMghEEQbP?dpsA15(<}og=?uN(PYoFqz?D`N9+_GWA%i<)o6gL>yK9 zll84yd-sl^e0Wdsxl2QXq!afnQt1Bnkh@MTg(lZEzzlmEzsb9B`DejT@pxEwYkL0XLM!j}r>Cu+zaO?` z*yCFEjyo8et4&reT*2*5)@yWKS3RFdlXU+Wq7*Dm!*Y0W(J;y=ahJhA!TWst!B$XR z%;%Uj{H7j!8Yi+nCPJn+I;ks-tgfd~>kf>4Ct^N2K8Kl^nXK+u0hMSEbQ2nlF5#v! zqWK*lna&bXdMi_Aig6NT5x7ZjnlDrZA8h)uLmC1gZXU2WqR{s0kDl%9FvZaOonvOC z&ODe>O{c@DzG>t1#B1QF;In%e`C(`Q$lH`)ZH2(wTk+}Nl=EmF*M8TFPj>n>+pRwB ztCux>%QI&&JBJ0S;-}e}L6=eP!Pyc{H<|E8LrO=0v#L#7P1V`ecjK#=9kNMM4MfTH zdAW%XM@v}RacX(1m7QQ}P;SHRd4rK8ejm-bQF0sXxs??k3l{R%gL`(^{gX7?+nMN- zMFxRNqZcnrZ;$%4C1w|VRu{QsS9c9x>X>)E{DGlz@dCceF9om|Rq?>k3*qhAu`T=v z5(AIy7NG#muc&XCOo>dPh}h;3M0j?r*ug9*!@y+ePMEz zH1=vVlriNCc{VOV(_zVBnWK)N`1Iu6j}FNDQn%=9STUcyaL6U> zzvK%Xhj*JTz(g|}tq8w;&Q5hY&Eaa1663NZRM%!#zUH}Eq2KA!>26_kIn00gvF(Oe z+n7MDn9~c9fybE@H}W@8UZ=y-QTsfY>E;^3Kxkyjj2KDc;DdEYSZHeAl6fx-s8(`= zQB;26T*b7P)J0R6AtJ-*dJHn_O7}jlNXc9_+lJHwd2xMSbA3{msP-h1s}d4@zrF^n zxmHHQVbWkv#A2|0qFTbRWbw%zIuPBcrV|zFF1s>X6Z45D+UMd&oMC40xY>%p%EpR# z9h8n-3tcjmNQArlE!}S^U@>B`S7`l&4H>(EkNixE5zg*2sRM#{Q&7j4@51v>1tBF& z2=t{^%0`7J*PK z7f@Hm>yOIK)EyCn1bYz{d8h%#Y3DJ&(rcZ=uMg2-i3-=d9iRy0_DY z-L$y&|JwWNu&DZIT?u6XVFoFshENa?5EP_aQ4vsq0qG85NXa1_Qo000+5qXHhVBLl zr9(Ot7>1moxf}exbI$XebMJlb|CfJfHnaEot$f$J)}BJhQG{Wl8aAq<#??m}%;c*3 zMxF!=+y4Ic7uNU;diE@0JlzWqlaPq<&?$~Jy6uX3_B@6&AHWj{=?IxYyiR^Rel>3# z9vdkREUSUr&1g%k)G`(W08h@ZN-WjsaILkAX5#tEp{$NvcSab=(=s;E`L|5-tfG5F zPM4FduuU8RWOC*}5`P756~=UUZP%3E)sq?LFfikisn#A+Guu_A`js<(z`xatN?XF? z%CgglPKgcdRHhiJpR@DCbg1grx(|^%KyN&sz zbD9P5T+Ur^w%$EF=k4jf7=nku)_vSb`1tt6`lSOC*_O3&2H2Qy^-13*WH*q6;wStK z0K-usbDZOFD>DthRHJb3KYNO;Y1l%U(!0C-y{^GJ17+VnYg%beZ;x7N%m$)o2&tFs z1O)tklq_$REq^nKZTf?ymj%9SD|tLNIk;_VrEe){@3hR^Ag13d?%tWf3=>*&rm|RK z3vSuam13{0Lx?X$8gnrMXop6oT%9Bi(%gKKdu62nrLK|u^TaY7RoWJ$ z$>+LIo(%VJR(F4NSMm!iK;xNjZMkFo$?1+SO${7jt= z)tBYs)S^jXLvlfoyVQ>zBk!!%L7SLcL#eH%6CJJ91}M5(n>_Nt3suwB0JxYK=;Nd6 z>-Q|}d%z_QHh(UX)Y&GHjsS5=Gk5u-m)Br+VQT>@syc#a()Pb83da_9X#2mKDkJ<% zKvHHu`hAI!D8F@+JImpV3qMr@;%8{T_J+~kq@Q?duq{D{Q$PD5fYS!ciRz~5?f<6r z+@FY3gD|2mzoI7`OLns>983ur)!aAy1@5tk|j z6+lQ?s-woFMWtioNDf%@r$iR#i^P#ga47U0xSa@=UpU`5(<1X3IFV&$*hZ0f!o4M< zD}_~sC=;I$xYuPGrz9>Gi0S1a6H;C;oB8+Z+3rf1%gv!;(`D9ipu8l`lGdfR-L|TV zHUECEd~CQkHpX8>nqgYv`o_Y}qGm|(z^1vMENf1)%+-;~{H8SDUd}b0Uo7u(zbiDg z{X~5@-FHUUw;4Ds9g&-S1m*cP@?>5(^KPZt#Z9)OO|T&i9iJu4M zgz#PVMrWZwr)wqFSW^JdR@?-I$V(mW?A|`~WUh|2>c_beud5bTUTxoM=@~%MKKZTi zT&->`+^byyS`ut?4gMoXBY<`RvF$t>afv>e$+Cj9##&+Hdi0>oy_qDVu{BmI zwo6Ht9Y=bVTw{Kms+}~(v>Y$1cTBEywV_JbY?Vi;{Wk4`vz6fYz53Rgz0<4DLNk?y zBMG=&_`T8+M<&VL?9eVxF7@DL_q6rCUlNPZ*}u_so^g)nCPe(<(I-cS${HD!^`EXE z^;dK`)1xKwKy4hi!Z7{|&U3y8-HiAjo7`J!mcOpul}w_l9o~`FF?%$hE^ge09@*cz z=_SH~HldZQuR3b@Jr|&P9C^!2#7hUA)ZW06`T5A_iQI0-UQ}o-I$C^7!Ekp4DX_>6 zO`TtK+I+BegV$SG)|TVStar8Kon>4MwE)VTMf7dYv$6JlhYcaoZ1B$2cH1tc)X|;0 z<3Rl<5R1h#iML?}kcd95>8~zZXoJ4;bQ@(^3zO3vntD;NCsn32x?z;Xr=&lqgUB&3 zY9+65**?fqgNP8KIX7)(ZHuybP=wCo*R_#!)d>y$8M+iDkYj_wOW|koKZg`h?xZVf zFG1Cx(46;iH7E$v&kkxZnWkqtFL383e?6R(*AGmnWhh2WCFk&PmfV_%ZL!Orrt`XC zpn@-<+B=>^IoYcrbC^HGl-hX@L}d2GTp~Ul79E0$+D3~ z-TtnA4qW-ir0R(+V~+Bi3qsOP10<4-eohk-RW0tPi@AwxVd&XstRX5xF(As+0q0+F zp<<_%k*a=daYDNU_#NygG8s>I-Nhp@G4TPGXN^HbHI-Q@ii{JU4cdHMs~~i-KOzt} z9W16Y*($v7mWTvT*`<|tAdUayQ|8G0Ev%?TpZZN7Bs4nIC+br$$&68KzXpW>g8yN3 zO21eqXBcZr%|fwv$9fZyw@AA;(}p3H*7!hoaX3`4+0xG1{nBiXuKcB%a^7P_osXA9 zlp^S-^0k3<;If-*U8}*_Zj^c6%OuFR>%{fj1JH?nO79$#T@_v(AOgsAbrCt?ot@b8 zv%JOeAtt)fYXP*UY^S^-=zZOd>?~5=Hv1C{F7Z}p-Y3(Oh<)c191si{SR-%TE ztt6$;!lD()y%YILlC+k^N}E>+Ngpy7N5iy{iXK@n59cL+ySToYV)6XcLCFLz##!yB zfqcrmmbYuQ$g*gCd_+Gc@|E-1Ou$!`)|x#&t87NjUi)ftumIpCv`c1u}GP|Ju5ftVOiK{e6PQXnNu% zCSs*60>@~|Z&KPj+xs7#yKf{4f_aM%q`_r>Ur{Qa;alk`fe+dm?pVP zYGR}>Nf_DkQ_LK_W3S}MbUG$?v2CWbDnJj}!)B@dT01hKq>NTk6b9{4*LDetY(>9n zR4)EBXQxlHd(XCc>5_ED6Gb@M`{GN9o`FNG3D!<*_GX@By&LVmQH# z)^h0kI4`D(A4NGkVtWUQ=nB)q=#ceF@QH=;u98~A1AW8848|RT`h#ac=nJtpHKz&y zM}?0IC_qllv|WUS=-$2X@n?r`VqfQ zh1o0K@S*X!vGo>|H+4*WHyLc))?)XaM9k+kXW;NvzFtz}HZl>f5xuyhzUed%3l}DQ z9M>}4%UCzFEr+BTBAe}}Kv@mD2P1I;HH4*yAC>;a#aM%@^v`M6-i_p76 z+)l(jtZkp@f*z*@Fb%EFS1;mEeYg|ErB#cMwC*(;Jo#N&_?E7^1=NV~DvC$;rv%-F)%@7G8zrO#dE5RanyC%{eb~Kl^ZDS4Wj*mI|sP3wET{J#~4* z*hy?vVsB#pwvV;gmgsjkbX5jc9*C!b>{_)x%KMUm{hIdI0-zP%TvuoVR>*wamEwhm zCP(^~*xu@!+9+J=p-WdEOrOzf)6l5_4$)c<{YJ87+I44Tu&tVokT#^z(ii87sKp{^ za}dSe4#&J);LcxeuYQy9Vx#mxyr8S^t{%JMNw%;j$ouVE#KxDAKv8Z19VuofSv!Eh zB=2_g5lTvS5W#M;RO_PixkHrP&(`DtQ=dCEt{*4$vsq^JkbK)o#?LanjcCM*QNg{Q4CedMYQTDnnrFBc~{p~HwCwc*z(`@mwfISDuy0D>+_ET|5M?QHb;XJOWt8zd# zzV8SXYEEx_^NA2b>M9AD|LA;eY5eE2rSXZELAB#%hx^krE{-n)J$GGhl=;7z%Xx%r zj3VVF6t!&~8GDo{rOwiW6LGvD{sj}c_UE3uclFd#_D&gap+%l@&(4}6zdA!!+)kPF zmWv2;NkqJj#3);$7iZFVd;h~}ugQ0)j{+uNZPXyrg3{I`@{rTT;yBlRZFkT2`?_-b zU*qI2F`i;KUYGW=z5P{!ZeeJSOI*6vXprmRqZ~@bYB{*C(!XD(U%WhGs|moGw84!{ zgWw4+?BQPMAY&j9%LNjE&6DBwt-y00XYzQ7!lpL+UK|^;NVC7-Z_d4yH z)f|eB*2pjP)9|Y$boT9xz@X?Av@@yAwEKUsNJzwbDA8?ufUbTD4F-=NAcZKYVA5e< z`+EXWgrw{TL9#kmDWb+=7W{Gh#?v7pG`RSyyk{lB2VY5^GkgaVm>mBg)vd~HE1W{L zSC67~ZF)*Jb)SrFnfb($O;le$-PcV&z|S?+GLpXz2C}wh-nomcZB;fS7WRr(Lc8yg zHr4kpTH20W?imWROQT&+kf-~iVj=0AuLfdh@V7vCdMX8ND%O_`#w|gM-0FtN1@9Gx zesjb|<@cff$(R+PNBr>Xyqte57$$pGI0pA3Fx?|FlLyxq@0Ev5TG%WIS>)}AEOv%r z)bI-v9nVJFoaj9Hn_p*k#%Dnd$tmWXWkm0r1S3e+L(=Lu-az_Qf{VfbAK23zMZkgWNdh@J<*E>)kQqO9#7bY@-##0nd#QR&QDiE%L-c zV6^R{N9*cshu`mymx)ayraYtYX^CMeXp!k(Wzyrh{BO4sUvvqIKB&2yrd8C85`T zlqNB|_8H_lMS72{wDdxW3)-jGN%zMHQ~eJAY2h7el>%VrFix{#%>LJYl3y}!bM7p} z=F){3FL9votx~uO8E=J$*MY#Cl)+xtOqP*?V*eK5Z@wyNH0PDINbvpHUTJ)dS0eHr z?IV@y=PP?OdRB3yA0>^{KB~!3Jc{n5;~IMcQH2sJ;@+sq7=OGf)B%M;#eAfd63w33 zuT|mYA}8m}ce-K{RXBx~c^~juN_&2{%-$fRB-c=Jft3Pg1^+?q5G{MrR5A8BdL;pt)q#;SD)Gztk_aIK|iUiMESUFI7wfe=8 zEN+mu0cq3glzq25r#|gb9e9_sL~*Y4P4652{!UJvH$btp7~baf6p0zI;W(g1yBSge z{zU^w6njd_Yd;sT&$v@JVU0yl4@l1JxcR5y({8kRPzC}7;Qdc<0Zk+ zwYT20ODFm)-cO%s3rbhZv9O3a-DE3lY!YmtLtF3SpH)p8V8^}`{jxJ*glVTrI{tjr zA>>T5!B@bCHa-aoo6ZZn+OYC2Y#@yINgGptBI>C)2^YeB|7CM%)$iZGA;$3p+-oyV zxVEiCu9g!=3#%atcRH3*lSvy7WQ1f)H_ul?Gs;K z!C!i~dGfq((u;lq>U*Se)*@2m<=MpbBtWx?ZRFBF9%R-Y>-`xc<>5BYsJ?C_#%TQT z?m_^q9G_ei$&t~FWV`UTpC0@X1lUcXfsHa#XEitn?1M02J%e_@r9>KD z0BnZQBwJ*8Nyj(rvF#jDzBAhTOA_R=3dq+vivup z)habmGYQbST#15`$=nP~d{!X>vxoRdzhwsN3YJd}7EM|8v`p{CUu&d0EaWkJ>y@+N zf|KY!EO99s^RSKwAbncafd^$X(w-|lE%p}Fw5BRRLgvNK1qN8d(xPD47FHvOy4syK z&OcNjpeXFlvF`B$5EVMuzY_q{#bKOVr!c1pw&ax4{f=mP2v7YjR#Q>qhmwsg25EXl ztn5z;KOb2P#dVJJP!jf5a4BjXJHU2;FX$frWbr;uf$y^BHwG&jqOqn+lr=XYMMese z{P4%`<0|=ujRW<6&k52WX52F$G6p*ndC;z5M4vyBP@7IXtsiDnyoC{Mr1gp>_gM1o zI!U`|ckmYx1H>&$cFFajQ5iEC@7*rh@Ter=V7xAOJd}`wdSqzI>VaX+_{f_(;F+`1 zGxQ6)D>U6uFk*cSD}7`$GK~Iwn==f>L-w6P@n2%aLE>50hRkt%iAdI1laO2{i5jsu z%O;@0=t)?xeYl=stP4H&DOk?egrQ8OT+@7(lR ztPfP8Q)xa6xct5whkYnT5SH2$0p+YO<^V|3ZS>L|8qRB@M=V&9NSTRUfKuAVa zC-Z?~{3LOE+5mI%amLh=U<1T6I(&8!J(yl~CH;fh7XkP&aGaOfzp4No1Tmfc?@`e8 zQW7_2(jJBux5q7)g@z|@3lI=e&mr!6S?ZOR_Y~IKM)1LFH}ziiA|DxG&z9`@AP11@t3Vo=qzBMh z$&626f|r@snwYqmzf)aDJil^7shQ4C-cK~`?yUip?5Z~rn5Xzu%prSK>yk$haUDN{ zsfhXV3%8IuIs48g9aZ-HTsIt7b7-F72DY>V7YrwSG7Je&3ld1_;;z{);YYk4BYdu? zY4Li{mmFZDbiC4;F~NHGRpi58SsZ1*sVfYWEoXIb^=g#9cD#|f6-R6Q^1MjW!mFpE z2yefm^U{1w;^gJ@ICVdh2alh^x1_t3gvP(zZ!ELC>SYMJd+hm+9Yk^HVc5RNBoxR; zu$A3-1-JN0krSkyIt1E)jRPu_bwj|TSu+@~u}R6T}8PXZ#l4CzxiQz5pdOjy1c1Z3F=i&|MzijFsF z)<4qry6*J#qCXYf@>;R17sx|gp|{XyBV4roSnJ+_z*J58Na)6Q2gnYyIbfRv#zt9A zWq6tkIwbu?VrM4UiC0d;+DXI89W}^eUL{ z&s#Tl3z|IZiwTrw>STXGV%=^>x~oL*keQ;`AnK@384$lltU9agHNVJiK6Z(JZ1MV$ zUYq*>9pZNw(`MTv;2z|ju@iax31Dhi0a`mhxKCYfQNuFYlqk~=9I)0k*y0vf&zrW? zou7LD^qXu^6%ZVeO@i9O2z!6hVX6D8tT9D@D!t=T^t8>)U8m0)=E zN>dEJODQcTdCPgwlylp$G0Rspi%*V`WTlLYmD7RaI!E-v9|hYll*yn=qz+P6{o!uL z%J1_{ms{?3ed8fk)Ycq+V<8-b^*coMb|E>qnvV)1CuH~K?jCI91hA$&7CxS#OAGZq zj9-KI1odtH`f!l;{dGPWH8B|#Yj&poWc9BNcbv?tme|e0`?t>d$InueOB(8rTwFJI z1-hpvpX66}AgY<=nrjBNA`?z9*eIL= z^8R~r-23ScQ>IN9`Dmd-$Ebnec+F@{BmZAa)9H?4c%ThXI`)sWQqS3DM|XAlL}C$sXvJ8=}*H`##d zVm`3{31O~vI(6{-`Y)oOr(|Ti9UTgiD@@@Lg~T4j!^HNE1Cj8XY0EdGIo>XisuSvV z18VB05F0f&`XuudguDZKhw3JL8c(;ic>4?{3Gv&M0;W(l-EWZ0y|{!JucoXyUERp= zWZiCYOZ_jm4_&Mo+sw?f{F-?d?nRuz$$y_(eUU){i;i%gYKh>Xuuaf25cYrM9 zzbfTy+VPs4rG8FA^(REUVI>76&$$zD75&O=dS0-uXwB8&)T3sVuwH>Qr<1A5=Q}AF zbT|^h&js}m-k+O&&KD5hZ*9sN9vt*&ovVg27#n5Gc!PZApo{-9dm+>3hf7||%_1`@ z%8YPxrLbSuV4CIV`{V9oe&ruuN}p_P3gn*j@fns-U`FYf*#T!AtVi@8y-&}@Gc+RY z*lYZT`fnpKkq?djtN`zUkHDijN(k&Uu9V70ASD}ebXY~g=TkQ($->K*3``!9ii@#k zgM7F-M7IG#uPfvY6-wdd)_pEDUQZ*YDB87DmIsA{RgglJKFwtBYaP~p{nonR)C z$sWu-=d|U`2Ho`d6Lkladdua;yq{~hoZ*C11k26DF0uOii;!XhQwhgMGVL6W8w=Hx zdiPj@?aezC%sbgE?@VMc3SNP4_B`Pyuli&2CQC*%lfJU@410nKCK+^}rL5v7@)(=N zMmRAK{BYA}mEOd4busxyKe^i6E%1gA+8rJ7?9JP7=U;jvaM9Z<4mAVd8qw(FH0s03 zt4^Dbe0PJaOzv=z#H>~Ci#}QTBNJ_Egq#jn(4D`vOni+>?B-!R_efL!y*(FWJE6Qb zRXS7exIa(~dT)W4#ewsiSD#*J{Fag-1dwT=5cmuB_d%l?iJ z&PJozeQe(4Rhr`4NA9v3-=Q*sH1@HIN-x{kDnMxgTXA=AZ_C~e5SmN* zd3wZ*6c+ZYy{7p?p z!ePj5Ez@Yd-$&ut#IF?V!!!0^EVguEPD*P5qNOZQ+wploT#w!N&M^p;&n780W};

+H(YyXeNYpkhOuBHbb>Hz*8j<$^ z`JI7|5Z=}6gWkUDN_0z$j!b#%lBkHbED$qxHdG52Nnxw}Rju}&thRobFr`zzR(%Lz?O`M*l`Fd?a2flVlS>v_y z5h#<$3!2>?m1lV!EaZKBFj-@Nl0Kq6qjBw^F>d#ldHU&ivY%j5 zVp@1OmIbQa!Ui+1SM$#<}eOV zD*pATh)Bnb^`>fUF;)oG6d*0pYf6Q7MngB9t0TZl=Crp~_^V&{k4R=xR4l^~LW!d1;VrhG{ zYN@aD^%~oA1rU?i4$6DJaCB^_>Q8}Y&vzVf?EA^qzW0I25Aa+Q4>O4Q_$599NuJIu zGvck){d=LldWC$?ItD{p#T35uh4{223TujxXbZ{zE^Byo~6UuHr;`G2KPtvZ1 zNiI4Wagdm1Q)4WY3kBfw&V@-0RfmK8s)Y(EM`DyD^<=g9ddY_7$W+;V zuLpYjk%f9hmZOr;Q&C-ldQ6rts+Tui`pD>VKCQ8)tivsxi}INq|?n<807QboV+=D&Gbcjfl7X2U!bfoad$c7CjFsm0l0N|U|djQY|y?YGOY|4zBOii zzuEI_E@-(U2hxZ}3r}jUX^k)M@fMVOyjmt{ZH~~Pj@<$J97L-lJ=3-VHLFAyBh0TU zkw6)td$e3ABcNPE4p0_s+NC+o6oDzd;6isim^a^q5LsUwKIYbY#Yz%}KD0IeeP3M$ zI&)OTzsvR_stDuhA;EHgr=TN;tN4ToXhtS%HI6?st-jhnSTeZO`K0G5KB+)je`R~n zLx*$kPyIBfep&*pp+tf~!p7Pt|Ky~{#-5LgbUNCxVZ1NqkJF%OR*_{~=|D9=-Y>3$ zc{y%`Y&N8dJC9H@t%1RaR8-7dv03n2_q`Q)Y;f)>Nyr0nV--GSIKnL(PHHyXja%OI zNIodl)5MXOZUgLUqa$XVAMdp0ODn+gXnZLCQuw`y$388_xt&xqGrO1F2?x^b6iS}s zkCF?D2Kz`__IEVYl8&2mO&ZmPY?t0hd3zO#rR1|myXjjp&JZ~)R`yO$cWl}~dy3-8 zk@wGC>*&+yl(ZUv$UfsQXLUCF4ud5<+JlO+pe0Es@iZpMjf>2e_FORu!sw(+t{&>! zCsGe-^2)4;w}CV|UVFy1yl&x@VM)-fgI|=~QjtI*{FSwFHAZdbC>f0%k}}VB0@%wB z?qhuClPx&%-7kgdQ)(M8`4k>yjUQ_#reh^B(3;~i=B`JygT?`C^c8_9lSu_v=gn4$ zY7tlavadY}X)}CHYZg7GsU}iP>2D0}p%p3>PYnBS^4bMUCZ;MM=}oZCnBM7Iklqq= zkrr7=`w6d=9EmSUn&{PB+3sA4?YdsLV7t&JwBmf;jKGb2Ja>mv3tWf&oYFgm3f^-T zWU^Mtjo&d)09=+9MSGiO>(^%v!UFL>w-MimG~fpZVS6193fw=|VV3)YS4fmA%@H47 zP{EQWb$kQgo;Qb$s6e}v8^sNWU3sudM#zKdO2=h<#-IdKnoM%$hL+|P${0n|!rt_r8hVL=1idRDw zB2D!!chLVcdTrKm60xahQLZ`W=JUhaQ#Tr(EDrBz7e;@IaN8ap)w@(<#k2xJja?Gs zt9N`7cb?G|m-kd;=K*xQ*9^$h7?z{`DBt0rj-wv3V4!ln&Tv$sXUUOo)k4M#%G+BT z<&q$2^~pIZ*iGkd!&tt)r(V?qUSso9NU@H0Ouy$-Hmbg`@VvbWy3!n#cf7I}ty$X= z%6!d=oV0_iB!ck)C$0_L+506iF}owGs7^)PP0A|S4(pPZebncI*_OoPI(6TDFJnr4 zpd~$1Jn!h`qaU>59C2f`V|(7IZvnxrUoo)u*Lb|QZq4lo(s?bbTLxk15uSvZ%$@7z zNIlS!aMC^`DEOziMB=2&y? zvt9<`6@+jY2`oIQw=;C#th@Cf3*ho11xb(~5(0LDE9T>VZMID$CwfLJTB zIs6(DC8GjstFAtEOen;u2CApB^Txl{Mlla6fP=E9BY6BFdE>TAZD?6bo87jy~+s{O;y(6V%HcUbNL)FKzUoyy8hB0$@7f{%lt^RD|^W#0-u?U}bm?{iyGXc5mIE=-c@q*J2Xl zF`m-ZeVlI6P9=W?wO#>L-PmnI$py;NSP&i`Q>TQ<)R#(f(_JzR_N%h{@^YjUA@6=0 zz7px;+Kb-nTC;!ULHE=~nr{f|dx~?LATM%v8=vKB&b7^w4+SQyFQ`miIiyFEh#4!A zLtmJXDH-#f#P4@0?hxmK&ixHQ6~zYKgg{wtRB&{tzEV5OV3wwZ>BsONXVLq3^w78z z*%rGn_YgfP|Ebg5oxiA*_bmRD?m9B?N~d)>`puM)ebrX!T-_atb7CO@8-%S9JbiX@ zDzW}!;fyXmVWVt7S`^nVvj{jG)bxyTTf$q2s3UYW_4eaLb^65Kkw|nEXJVXPzqX_P zp{JME)D;pEY5*lQy9pAVHY|TVoO?6M{3r0rWl8C%j)fE)xk}ido)5+Rg$mNnV@X){t7LjW{_+-akr)jpO6yhJrgoAivkyE>D@JvP{3cs{}SL55(0e;e_T$D zp{106&C^1=Oz+M6O|iVjm%HLNBPtwYnz&ymQ;tjy%81`{tDs(@g^onsjVf}Y2tqB33g5&{d_rkE8Qu^BcygWHcaep zwZ`BoZw{RdKzhL;;`(I`&!WB8nushoOpSYLQ6Eu1u?Xo7>Nz!q)Vkk?7Sgp3mqbYA{>`6^cDy` z%u9EW#KmGlr|%ll@U*OHKsV8+{s!{GuNoy5q81(GSo&3fHZY@G(p17nj^ z&UVMjef0T!DgkOS)=ZkjHX8qX+SK(d&D5g);q28Jk9*2IjuF&-)atG$OKa<^aQ&v- zaV@n3?S)vd47547+)S%0{!SMx0ua6?ptN_~D4O_Y;B!9dxpLhYG(&-H>Hj>Mt%sN^?qoZG&(tHBWT>la}&l2pRPS%AFLDD7LA3z zX^E;nP@a<#GH#+_KgjuTCh+ulQ(S*p+{iyYt~8b`nRTAEpz@I#OiK=Ctfi!+ByYf$ zEdAwpGWOkYI?uxlonr?onDb1=o^jxx@gIUm2MfV)Z1Fu~!ykVHKJDTMr#^*F5(*#q zNN8D3R!?C5Xt@Ss&Jrff%6Q5v$u#?{fZZ|_${%C-Dj&E>7Tk%THr7lSWdHhanDq{> ze|?5kXfV_;C8NDVKMbPaB+dmA8q)W9ULVcUAh#I^8_+uq6Zb^wORN^GS=pXPfF!4n zh>niZ8%lqJ!^9E-aZPNNCO=6_NKCSeiUhiHOxvz4QexdFDjiKrnt41r=60h>%f!^S z%{bzxK1i-ebDI440l&e_Tn@a%aH6yGoug`B6A$miCM7?p@Z^DE z+{M!OLMUz$=U%^%fP1Lf$8!k)RSb9?C}*_Gw(dtmzF82nh;F+a&HH1)Oud0_9JD>t zCWo^S5moK2G%lF7+}LJ4&opLeeB9cWv=dxj`n77+axJ-lvhJ~URiQv$mMH62ZK<}J zT$Df4iFX_A$@oMx{Li))i{P;mAfcuuoV0;ifOCg+*ta$+#{+751~V zvg%X?_9$elralF+K4yBy0!TxDTM{e1b3!yn9I_aUSJHy10PU+? zpo$QFPpzSf{t-p15p5T(QBY7oY8|LD)P^w`Pch|!zYk~xs5N0wfZX@1)6Xd;fho$*0#ltCW+)Rf7}FtGUHnUF%oAy@3%g zv*#N)?)et$qK}J8yKe_SrGhQpCI2aR;V_9n$7>7F66nTJG6PD3^8!x2FcN2g-dkW_ zo~UML~fp`nDx4Z7V|nt>??&LMF(|Xq#xw`jXILO{TK2rAoc-X!4MU*No#E zGi1bAS(|xCqH}9WXP^|udoK;oGB8qJ-L6mx^Ek_n?+L;*qua$#^6X!i@s7BTCG0Ve z43FDU8$Z-MtSPH}s)@5IrGg1eZy6vq&@+V~qKDd=t@ocixj+G3SXqDOiZ8!_U1S)* z(`jENWY44XPdg+-PC+`OVZPTG=qb9MD?p(&z=RW5 zP@c;#6W+@Z%1G^-AL)#-psr)fcxa55TBE21eNPwY?)ncpzu-eze2QTv zO`OEgmTzY4T8P)~+(#njPVc2e>KD_T1yOisV2RYG>k#7Us7I9(8-xQmphpTTte0Nh z7ZIvK0sC%rxNPdm};CHYfs~m9Q)~dmfXM31w!7Sav#_9 zHL5Vcwcl9NWCO^tr&R8 z00^Qw*e+v)1J;<436v+7Oyn3JMe{Cq3Es1wGg%GCHEtgBP1_Ee_p@=^ za|35~Y6X6+QhH{km{RrTiv)f(n5PQ3@AX}^-7VZS7n}=T$BaJ6T|q2B@Xx!Z?Tc3-FvfJss}-(_)iV-uh#N4)Nt80u(d4NrKB< zj|-gT8cFklZmHlVvVA%W_3_Pm)$#g_K)&_h*2}*eLL@-ze`+RA=jh$WI$LEi*ZGwH zf(^iuK?tD=1sunyc2o&B^me9r#it%9d%)(f_?-hcel%@C^ttZxi^AvtS~3aig+(>F z@Z0K{Rb#5jI;~mHvh(wYfjVqtP#xC2RVQ>`(gFn<5c>gP>w-T+fD_g!!qU!`>;n^4XYGomagU$j(1q*D_eEIjzSe1x9 zNzzegxY7!pv(6YM#J8mLTJP_nkb>Uahofd{=j?4U+daH;`hRW{=du8p#Ah7vRwVS^ zRp5o&1%aIT+l44%G9<%8K(-O8$L!yOJU@Q;P*$(*?^=cp$4CeonJyUX!e9n%-1qe+ zwC!~zCT&kvU3Kf6s)UC|xp;HuEyM+w2$fW2Ozf}LM zN;&S{ov^%^?~y(0x3K@FbMDJ7kem#5$kCnXZ8f;Gi#d9nf4wzGeevMWTtxkz@~zI#yT4Y>JK?+%ejSKph7;I2DW@HSmuN z-b0Y*)mLK2*^rOZtn;lQd4-Fn>c#JdJ)AwSQGqTCL7w4Hl=h2N0l*juAV$gA7gPxZ zUa2Mdp?gE;JV5DhGrdhJpK%>{078 z_d6GV#z)A7P4j|JEE&@oayU!P!u~lHWUf+%VD0RIri6^uIRY^}$wLG{wXCQbYzHoo zWZJZ4;Ur8%DPZ^$+xtsw&uM_JJMCo)3U=a9KUsB!n!4p2Ukj}QZ=VZyHUFV ze1vCilWwbN9{31}b00rY1-Weq{AN6M>hV3cyHH{gH$D5JstpBnZWTiM7tf~xz;S=z z1WpJL;x{Q`&a8HfZjH2w+s4_myJ@3dJQsA63jr?$%clPF98X?uCS`pS z$kJ0T+Wt_?(VLQovqM!OVw09OfG=j`0f_WV!N3^Gb}}{a#by7-*yA6kvEYOyGw5 zdDJPsCIBa}BRAWb6pRT2*Uoqb{)7b#*RrXTusy$a{?r%EmTf;-%O3MCy0^*V zg*{=lPPFXJxq6n5JpVpgqX+RLuAp^8H}B33k8P+z5==LhK`wyXhAR<7Y)mRU%tzXuv?a^V-`u?L@JP$~vR%*L> z&fSk}FRR1;t^!UV>a6OVtlT?xN8w)uq9H=U^X$-od+=E^C@#I;Vtp03q?-v@pLv4o z?T|46<9&Foa!J~|y{f;Y{WWfl>x5bbRlm*$VF8DWq0RE}8KyOkHkGT<_m$G>pnZ7)|+Vyfr%}I=VMz>U@_V7e73B}U((2Rt@gW z^QQ%L0tC!uR@nPypAWA1btay7m2=6F{@Ktn?8a$;)HMFY=mShob9B9%;h!t+pa071 z3PYb@_^X3md9daS;FmOq-aL8r_Y0l07qfa%xRX>Vuk}ebue9&w$rJxNfp6TztNGJ~ z7|+aK=)GQa&O-Ia68(0-9HOB>*usyc9?6)>$Q-MQc-Y;Mk-zouUo`{D7Q<1%J?_)i zU`igm(wdCqDRbE33_SgyK6{1 zkx*F7C{jORH?hq)Qg7SqRWquApn?et{}O_o-{JF`iwy%UQZpq8u+I-%YW0IlnQ-OT zc)RGO3vJeUn9+wWto>m(waZZB`YSBoxa+6RWqDWNaGuK&XZ5K}6fVe=q+e+`+Ny^F zt3zeN6hLK+ri{TvZ) zHpUlih&v8rW}MR8ClJ1fKIiImjeVfpD_F|>mog@Y(ZDWG7ybS@H3fj8$QH0USnO6x zx$iRKdK{P}Ph{^Jf+L0}q*xIbs$U+|*pAB{0~RaX5flG7GW_u%46XG~G`f)dH?#AM z14OXfa1&d-dc&(WRwZWR-<$8_K-nF{ch%&im7IP>yB;j9^tiBJs2=V4qPr(7*G)(L z6|$jMhlVwBG=5b6d);=BY${*z1}?RW@`~h5bgsGSwAtt<01GP@T`PI@S1n4|9XuDc zOT4=_>Y?M*Tt2oO)I_f)bDsaJG*xT^Ljx9xZqk#5pvlN}ZRccQFD}mm+OJXP5}TDI zY+1;LS~Fm{+f1s6=!jdr>VG)yZf4|2yQXW~N8OeL-+t0le8EWd8}q6>J;B3} z|IC4b6MRNIhG?(^`a-^|K8TVt9|?!;)6MPQpW8I#O#fB<4^&9no{`A}&C$u_$Tdt! z{ODvgniu{=|0Eo9p7aB@s72hDDq-o}a%AhfS&_cp&`(KxR%7q(Tb0XO_t(fXu2B89 zopy@B)@8ZW3?w~C2SnJT^ zcmS9B(4r4?q3^gmb8uPZOIf+Q7hZaif&gZ;m*Vl=q@l*m zswhH*atVHw^x8ir5&g)>+B>Yk`WM|X2B=eA;|sv24}2kiS<# z*v>6}Pnn^>xa;LYZE=0qM@B7e<%OpgRj-ZZ_VAgAvtQ^pDw{PACc5d4u@UT9-LsFVj5V*svC{q(;6*b1kJtHE zVgJh~&wGHhFn4J@5%WKM_}?$PO>~}UG~IlBneD%S_;2d|pPvi~5whpM_Q09f{I&ak zi~j%q&5KXEF9GGB`CZbN7kJ2j{+s{JW#!>OzNM4qJ$}*X{@?u1|9DoUkStIl+<4}v z@qha!!TGbv@&Pq=Zi1yD`aiSo|AUA!u0kOHU+n*bWdGmAzIjG8PnYM&`~8RdCE)ML MLsi8>c_Y971)bAL7ytkO diff --git a/quickstart/quantization_tutorial/qat-ptq-workflow.ipynb b/quickstart/quantization_tutorial/qat-ptq-workflow.ipynb deleted file mode 100644 index cadd4de8..00000000 --- a/quickstart/quantization_tutorial/qat-ptq-workflow.ipynb +++ /dev/null @@ -1,1732 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": null, - "id": "b861c182", - "metadata": {}, - "outputs": [], - "source": [ - "# Copyright 2022 NVIDIA Corporation. All Rights Reserved.\n", - "#\n", - "# Licensed under the Apache License, Version 2.0 (the \"License\");\n", - "# you may not use this file except in compliance with the License.\n", - "# You may obtain a copy of the License at\n", - "#\n", - "# http://www.apache.org/licenses/LICENSE-2.0\n", - "#\n", - "# Unless required by applicable law or agreed to in writing, software\n", - "# distributed under the License is distributed on an \"AS IS\" BASIS,\n", - "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n", - "# See the License for the specific language governing permissions and\n", - "# limitations under the License.\n", - "# ==============================================================================" - ] - }, - { - "cell_type": "markdown", - "id": "c6384192", - "metadata": {}, - "source": [ - "\n", - "\n", - "# Accelerate Deep Learning Models using TensorRT " - ] - }, - { - "attachments": { - "img1.JPG": { - "image/jpeg": "/9j/4AAQSkZJRgABAQEAeAB4AAD/4RDuRXhpZgAATU0AKgAAAAgABAE7AAIAAAAMAAAISodpAAQAAAABAAAIVpydAAEAAAAYAAAQzuocAAcAAAgMAAAAPgAAAAAc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAEFubmllIFN1cmxhAAAFkAMAAgAAABQAABCkkAQAAgAAABQAABC4kpEAAgAAAAM5NgAAkpIAAgAAAAM5NgAA6hwABwAACAwAAAiYAAAAABzqAAAACAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAMjAyMjowNjowNiAxNDo0MTozOQAyMDIyOjA2OjA2IDE0OjQxOjM5AAAAQQBuAG4AaQBlACAAUwB1AHIAbABhAAAA/+ELHmh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8APD94cGFja2V0IGJlZ2luPSfvu78nIGlkPSdXNU0wTXBDZWhpSHpyZVN6TlRjemtjOWQnPz4NCjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iPjxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+PHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9InV1aWQ6ZmFmNWJkZDUtYmEzZC0xMWRhLWFkMzEtZDMzZDc1MTgyZjFiIiB4bWxuczpkYz0iaHR0cDovL3B1cmwub3JnL2RjL2VsZW1lbnRzLzEuMS8iLz48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOnhtcD0iaHR0cDovL25zLmFkb2JlLmNvbS94YXAvMS4wLyI+PHhtcDpDcmVhdGVEYXRlPjIwMjItMDYtMDZUMTQ6NDE6MzkuOTYzPC94bXA6Q3JlYXRlRGF0ZT48L3JkZjpEZXNjcmlwdGlvbj48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOmRjPSJodHRwOi8vcHVybC5vcmcvZGMvZWxlbWVudHMvMS4xLyI+PGRjOmNyZWF0b3I+PHJkZjpTZXEgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5LzAyLzIyLXJkZi1zeW50YXgtbnMjIj48cmRmOmxpPkFubmllIFN1cmxhPC9yZGY6bGk+PC9yZGY6U2VxPg0KCQkJPC9kYzpjcmVhdG9yPjwvcmRmOkRlc2NyaXB0aW9uPjwvcmRmOlJERj48L3g6eG1wbWV0YT4NCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgPD94cGFja2V0IGVuZD0ndyc/Pv/bAEMABwUFBgUEBwYFBggHBwgKEQsKCQkKFQ8QDBEYFRoZGBUYFxseJyEbHSUdFxgiLiIlKCkrLCsaIC8zLyoyJyorKv/bAEMBBwgICgkKFAsLFCocGBwqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKv/AABEIALgCEwMBIgACEQEDEQH/xAAfAAABBQEBAQEBAQAAAAAAAAAAAQIDBAUGBwgJCgv/xAC1EAACAQMDAgQDBQUEBAAAAX0BAgMABBEFEiExQQYTUWEHInEUMoGRoQgjQrHBFVLR8CQzYnKCCQoWFxgZGiUmJygpKjQ1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoOEhYaHiImKkpOUlZaXmJmaoqOkpaanqKmqsrO0tba3uLm6wsPExcbHyMnK0tPU1dbX2Nna4eLj5OXm5+jp6vHy8/T19vf4+fr/xAAfAQADAQEBAQEBAQEBAAAAAAAAAQIDBAUGBwgJCgv/xAC1EQACAQIEBAMEBwUEBAABAncAAQIDEQQFITEGEkFRB2FxEyIygQgUQpGhscEJIzNS8BVictEKFiQ04SXxFxgZGiYnKCkqNTY3ODk6Q0RFRkdISUpTVFVWV1hZWmNkZWZnaGlqc3R1dnd4eXqCg4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2dri4+Tl5ufo6ery8/T19vf4+fr/2gAMAwEAAhEDEQA/APpGisPxl4gXwz4UvdTKSs8cZ8sRRGQ7sHGQO3vXDeAvG9paeE4db8Ta9qU7X8iRt9tgKRxSMCcJx933oWt/IHoeq0Vx1v8AFTwrcTTwm9lglhQSCOeBo2lUnAKAj5+fSr+i+M9G8UWN82lXUqSWikTpJEY5YuOu080dAOiorjNL8Z6LpHhHSLrUtanu4b6R4or25iKtIwLE7h2wFI/CruhfEHQPEOrPpljPNHeKnmLDcwNE0if3l3D5hz1FO2tgeh01FFFIAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigCvqGoWul6fNe6hMsFtApeSRuigVzWj/EnQ9Y1WCwSO+s5brP2V7y2MSXOBn5D345wcVn/GBmPg62gwTFcajbxTD1QuMg034uRpb+E9NuYAEntNYsWtyByCZ0Ugf8BJFC3172/L/MHt8rnf0Ui8qPpS0AFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUE4Ga8hTxrL4g+MV1pg1PV7Cw05Y/Lt4LRlSZwXL+axHCkKAD3o62Dpc9eorh1+L/g94beZL6ZoJsZmFu2yInoHbGFP1q7H8SvDMutwaWl6/m3D+XDMYm8mRv7ok6E+1HkB1dFYHiHxpovhieG31KeRrqZS0dtbxNLKyjq21ece9N0vxxomtaLd6nps8s0Vln7REIm82MgZIKdc0dLgdDRXOy+OtBj0Cy1gXTSWt/IsdsI0LPI7HAUKOc56+ldEDkZoAKKKKAMfxd/yJmr/9ecv/AKCa8y1OCO6+G/w4hnUPG+o2QZSMgjNexTwRXMDwXEayxSKVdHGQwPUEVUbRdMa2tbdtPtjDZur20ZiG2Fl+6VHYjtinHR380/uYf8H8TjfE1vFL8bPBzSxq7JZ3rKSOh/d8/qfzqBRt+L/ikLxu0WInHc88130un2c9/BfTWsL3VurLDOyAvGGxuAPUA4GfpSf2ZY/bprz7HB9pnjEUs3lje6DopPUj2pdLev43HfV/L8LHi2n20N34O+GcVxGskZ1okqwyCQ0pH8q7fxUo/wCFueDGwN3l3gzjnG1OK6yPQdJihtIY9NtVjspPMtkEKgQvz8yjseTyPWs7Xdc8L6Tq9q+vz2sN/DGXt3ljLOiscEqcHGdv6VTd3fzb+9A3d/13bOhormP+FkeEf+g3B/3y3+FH/CyPCP8A0G4P++W/wqRGzq2tadodp9p1a7jtYs4Bc8sfQAcn8KwbP4meFr26ECaiYmY4RponRW/EjA/HFeaa5qyeK/Ft5fmUT2ls/k2mB8u0fxY9+tRTQR3ETRyoGVuoIrx8TmkaFX2aje25hOsoytY97BDKCpBBGQR3pa8v8CePNN0rSJtL8Q6kkD2cmyAyBiWjPIGQD0rd1X4m+HIdJupNO1aCW6WM+Um1uW7dq9aMlOKktmbJ3VzW1zxpoPh6UQ6nfqs5GRDGpd/xCg4/GjQvGmheIpjDpl8r3AGTBIpR8fQgZ/CvGLCIuhvLgmS5uD5kkjckk0XsbQ7dQtG8q7tD5sUi9QV5x9K8j+1oe29ny6bXMPbrmsfQlFcfp3xN8MzaZbS3urQw3DxK0sZVvkbHI6etWf8AhZHhH/oNwf8AfLf4V7J0HT0VzH/CyPCP/Qbg/wC+W/wo/wCFkeEf+g3B/wB8t/hQB09Fcx/wsjwj/wBBuD/vlv8ACmv8SfCSoxXWoWIGQNrc/pQBp654o0fw5GravepAX+4mCzt9FGTVHRvH3h3XbpbWyvttyxwsMyNGzfTIwfwryCG4k1q+n1m/bzbm5kJBPRF7AegxS39lHd25DDDryjjqprxqmawhW9ny6LS5zuslKx9BUVwXhf4maJL4Zsjr2rRRagEKzqytnIJAJwO4AP41o3XxL8Kx2krwaxDJKqEooVuTjgdK9k6DS13xdonhsqmrXqxSuMrEql3I+gBNQaH458P+ILgW+n3w+0kZ8iVDG5+mRz+Ga8dtmfUJ5dWvT5t3dsXZz2HYD8KL+2EkBmj/AHc8P7yKVeGVhyMGvGlm0I1vZ8um1znddKVj6DorhtA+Jmgy6BaNrGrQw33lgTIVbhu/atL/AIWR4R/6DcH/AHy3+FeydB09Fcx/wsjwj/0G4P8Avlv8KP8AhZHhH/oNwf8AfLf4UAdPRXMf8LI8I/8AQbg/75b/AAo/4WR4R/6DcH/fLf4UAXvFnh2LxV4budKllaAygNHMoyY3ByrY9jXLr4S8Xa9f6ZH401DSm03S50uVjsFk33cifcL7gAoBw2BnJHWtn/hZHhH/AKDcH/fL/wCFc/4P+Iejw2eojXNczI2ozNB5xdj5PG3HHA68ULR3DdWPSKK5j/hZHhH/AKDcH/fLf4Uf8LI8I/8AQbg/75b/AAoA6esjXPFOjeHEU6vepAzjKR4LO30UZNZr/EnwksbFdahYgEgbW5/SvJILiXWbyfWdQbzbq6ctuPRF6AD0GK5MXiY4anztXInNQVz2HRfHvh7XrpbWyvtty33YZkaNm+mRg/hXR18+31lHdQHIxIvKOOqmvRfDHxM0WXw3aHXtVih1BVKzqytnIJGeB3GDWeDxkcVFu1mhU6imjvaztY8QaX4ftxPq95HbIxwu7JZvoByfwrJ/4WR4R/6DcH/fLf4V5bqWpL4m8TXerOwliR/KtvRVHcCtsTiI4em6jHOSgrnp+n/ErwvqN0tvHqHkyO2EE8TRhvoSMfnXVA5GRyK8CuLaK6haKZAyn9K7LwP8QNKsfDy6f4h1KOC4s3MS7wxLIOh4BrnwWOWKurWaJp1Oc9LormP+FkeEf+g3B/3y3+FH/CyPCP8A0G4P++W/wr0TU376/tNMs3utQuI7eCMfNJIcAVzEXxS8Jy3Xk/2iyAnCyvA6oT9cfzrgvGXiCDxf4mWK0uFudLskDR7QQHc9Sc1QaKN4zG6KyEYKkcV5OKzKOHqezUb9zCdZRdj3mKWOeJZYXWSNwGV0OQwPcHvTq8h+H/i6y8NS3+ka1frBZIVltPMBO3dksoxnj/69dv8A8LI8I/8AQbg/75b/AAr06c1Ugpx2ZsndXOnormP+FkeEf+g3B/3y3+FH/CyPCP8A0G4P++W/wqxnT0VzH/CyPCP/AEG4P++W/wAKa/xK8JLGzLrMLEAkAK3P6UAaWueKdG8OKp1e9SBnGUjALO3/AAEZP41T0Xx94e126W1sr7bct92GZGjZvpkYP4V49bTy6xdTazqB827unLFj/COgA9BinX1nHdQ5I2yJ8yOOqmvGqZrCFb2fLotLnO6yUrH0FRXB+GviZokvh20Ou6rFDqCrtnVlbOQSM8DuMGtX/hZHhH/oNwf98t/hXsnQdPXnukf8lb8bf9g6z/lNW3/wsjwj/wBBuD/vlv8ACqqeNvAkd7cXkeoWS3NyqpPMIWDSqucBjt5AyfzpNXGnb+vM4rRrSCL9lqZUiUCSyd3G0fMxbJNafjKGOH4V+FliRUCXenlQoxt+dOlbyeLvh9HpH9lR3enrp+3Z9lEBEe3027cYrX0678NeKrAQWAtNQtbN0IjMWViYcpgEcYxx6Yq27yv5p/cT9nl9fxOL1jWJZPi1qVhpd5o/h+7t9Ph87UdSQyS3CMWIWNC6qFU5yc9TVX4Yalb/APCVeNbu61yHVoo3jaW/EaxxyAIMkAEjb75OfU16TqvhrRNdlil1nSbK/khOY2uYFkKfTI4pZPDmiyrcLJpVmwuY1jnBgX96g6K3HIHpULRfeU7N/ceL+Fxb6X8QLbxRf2MkHhfU7qSLSTLJlLSZ+PNK4wocggHPGfeveqp3Oj6be6WNNu7G3nsQoUW0kQaPA6DaeOKtqqogVAFVRgAdhVdLdhPV3FooopAFFFFABRRRQAVQ1ZIo7Ce7bT0vZoYyVjKAs2OcAmr9FAGLocuj69pMV9aWVvtcYZDCuUbuDx1rQ/suw/58bb/vyv8AhWdpXhuPSNcv720uHWC8IY2v8Cv3YfWtugDxXxjo0/hrxVd3RgP9mX7+akqL8sbHqpx0rGfUrYL+6kErnhUTksfpX0BNDFcRNFcRpLG3VHUMD+Bqha+HdGsbgz2mmWsMuc71iAI+npXl4jLaVer7Ru3cxlRjJ3Ob+H3hU6doclxrNrGbu9k81o5EBMa9AOa6DVvD1jqOkXVmlrbxNNEUVxEvynsa1qK9OMVFJLobbHz5tn0SVtN1lGtp4TtBcYDjsQaFS4164XStFQ3E8/yu6jKxKeCxNe832l2GpxhNQs4blR082MNj6elJY6XYaZGU06zgtlPXyowufr615f8AZdH23tb+djH2Mea5X07QNPsNMtrT7Jbv5ESx7jEuTgYz0qz/AGXYf8+Nt/35X/CrVFeqbFX+y7D/AJ8bb/vyv+FH9l2H/Pjbf9+V/wAKtUUAVf7LsP8Anxtv+/K/4U19J090ZTY2+GGD+6X/AAq5RQB4FfabceE9Rm07U0ZIRITb3BX5JFJ9agkumu2FrpKNeXcvypHENx+te+3dla38PlXttFcR/wB2VAw/WoLDRdM0rJ06wt7Ynq0cYBP49a8upllGdb2rfyMXRi5XMzwr4WttE8L2On3dvbzTxIfNcxg5YksefbOK0rnRdPubSWE2duvmIVyIl4yPpV+ivUNj5/uLW48M30mlaujQ+UxEMzD5ZU7EGmNLLqb/AGDR42u7ub5VWMZCg9ST2Fe9XunWWpReXqFpDcp2EqBsfnTLDR9O0pdunWMFtkYJjjAJ+p6mvKlldGVb2l/Oxi6MXK5n+H/DNppHh+zsZ7eCaWGIK8hiBLHvWl/Zdh/z423/AH5X/CrVFeqbFX+y7D/nxtv+/K/4Uf2XYf8APjbf9+V/wq1RQBV/suw/58bb/vyv+FH9l2H/AD423/flf8KtUUAVf7LsP+fG2/78r/hXI/DzTbZtP1j7TZRE/wBsXAXzIh93K4xkdK7iigCr/Zdh/wA+Nt/35X/Cj+y7D/nxtv8Avyv+FWqKAKb6Tp7xspsbfDDB/dL/AIV4beadceE9Ql0zVFZIlcm3uCvySKT6179UF3Y2t/D5V9bRXEf92VAw/WubE4eGIp8kiJxU1ZngT3T3jC00lGu7uX5UjiG7Hua9j8LeFrbRPDNlYXVvBNPGmZXaMHLEknn8cfhWnYaLpmlZ/s6wt7Ynq0cYBP49avVGFwkMLFqOtxQgoLQq/wBl2H/Pjbf9+V/wrxrxRpUvhfxLdebCy6ddyeZDMF+VSeqn0r2+o7i2gu4TFdQxzRt1SRQwP4GtcRQjXpunIqUVJWZ4BJqUGAts32iZjiOKPksT0r1PwF4WXS/DKf2taxNeXDmaRZIwSme3Nb9l4f0jTpjLY6bawSE53pEAR9D2rRrDCYKGFvyu7ZMKahsVf7LsP+fG2/78r/hR/Zdh/wA+Nt/35X/CrVFdxoeSfEPw/PpHiAa3Y2pawnjCTiFP9UR3IHauXbVLMR7hOregXkn8K+gnRZEKuoZWGCCMg1mxeGtEgujcxaTZpN13CFePp6V5uJy6niJ87dmYzpKbucX8NfDEyre6zrVoqm8KrBBNGCVRc4bB6Zz+ld7/AGXYf8+Nt/35X/CrVFehCChFRjsjVKysir/Zdh/z423/AH5X/Cj+y7D/AJ8bb/vyv+FWqKoZV/suw/58bb/vyv8AhTX0jT5I2Q2Nvhhg/ul/wq5RQB4Bd6dc+E76XS9VVkjRz9nuCMJKmeDmomumvJBaaSjXd3L8qRxDdj3Ne+3djaahD5V9bRXEf92VAw/WoLDRNM0vJ06wt7Ynq0cYBP49a8qeV0Z1vaN/IxdGLlczfDHhe10bw3Z2N1bW808afvXMYOWJJPP44rW/suw/58bb/vyv+FWqK9U2Kv8AZdh/z423/flf8KP7LsP+fG2/78r/AIVaooAo3FnpdrbyT3FpaxxRqWZmiXAH5VS8L6pb6xYy3dlYfZLZpCsThQvnKP4sCpdd0aLxJp0Vs93IlsXDuISMTKP4SfStOCCO2gSGBFjjjUKqqMACgB9FFFABRRRQAUUUUAFFFFABRRRQAUUUUARXNut1ayQSFgsilSVOCM+9cl4RudQ0fVrjwvqqyzLCDLZ3RBIePPQn1FdlTZMqrOib3CnA6E+2aAHUVi+G/EkPiG1mIia3ubeVop7d/vRkHvW1QAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAVznjG61X7HBpmgwv8AatQYxG5x8tumPmYn1x0qz4m8Rw+HdOEpQz3MrbLe3X70jnoK1LKSeWxhku4hDOyAyRg52tjkUAVNC0eDQdGg0+1yUiXBZjkse5P41o0UUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAGL/wjkUXiz+3LWZoHki8u4hUfLMezH3FbVFcdphvPDPit9KlE9zpmoMZbaUgt5D9SpPpQB2NFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAVDdXdvZQGa7mSGIEAu7YGTU1cdq2iaj4n8WRxanF5OhaeRIiZz9qk9T7CgDdl0GyuvEEGsy7pZoYikSk5Rc/wAQHrWpSKoVQqgAAYAHaloAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigApCO+MkdKWigDB0TxMuqavqGmXNs1neWb8ROcl4+zCt6sbUfDcF74gsNYika3u7RiGdB/rYyDlD+NbAZWZgrAlTggHpQAtFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFU5tVsYNTg06a5jS7uFLRxE8sB1oAxL/wAR3c/iy20TQYkmaJhJfzNysUf93/eNdPVPT9JstLe5exgEbXUvmytnJZvrVygAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAK426ivPC/jL+0LdZbjStWkC3MagsYZsYDgehArsqCAeozQAA5GRRWHH4kQeLJNDvIGt3ZA9tIx4mHfHvW5QAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUVUm1XTreUxz39rFIvVXmUEfgTU1vdQXcfmWs8c6ZxujcMM/UUAS0UUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUVHcXEVrbyT3DrHFGpZ2Y4AAoAWUyCFzCqtJg7QxwCfc1zPhbw9e2+oXWueIWSXVLolQF5WCPsq03wtqureItTutVY+Rozfu7SFl+aUA/fz2zXV0AFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAY/iDw5BryWzNK9vcW0okinj+8uDyK114AUnJA/OlzjrXkHjrxnbaB4yTUPDd6Li9MDQXcRJaHH8PfqD6UAevbhu25G7GcUteMfCrxZfaj41vE1e7eeS+iyGc8AqegHbr0HpXs9ABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAeZfELTPD0+qJpth4fstS8T6rnY0kefJXvLIewH6113grwjY+CfDMGkacMhSZJZMY8yQ/ebHb6elYM3wrVvEmoa3Z+LfEFjdX7Zl8iWHAA6KN0ZIUema6fw/ok2h2ckFxrOoauzvuE1+6M6jHQbVUY/CiOkQe5rUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQB5/wDEbx5ceEtQ02KxCSMxMk8bfxJ0x7Gtnw54u0bxxZzwQRsxWMefBMnGD29DXiPxF1n+2vG97KjbooW8mP6LxWn8P/BfiS+1KHU9Pmk0qCM5F0w5cdwF/iH14oA9/hhjt4UhgRY40AVUUYAA7U+moGWNQ7b2AwWIxn3p1ABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFACEBlIIyDwa8u8a/COG+Ml/4b2w3Byz2zH5XPt6GvUqKAPl/Qpbvwx40spLyF7eWCcB0cYIB4P86+n0cSRq6nKsMg+1YXifwdpXim1KX0QScD93cIMOn+NWdupaZ4bSO0gS/vreMIiGQRiTHHU9OKANWuL8aeOzoF0mm6Xbi61CRN53H5Ih7+9O/tzx1/0KFt/4Mkrza7lvbjxVq02r262160o3whw4QY4AYda5MZXdCi6kdyKkuWNzci+Iniq1l865S0u4s5aFU2kD0Br0vw7r9r4k0WHUbLKrJkNG33o2BwVNeN1reAr3xFZyaunh3Sob+2NwpYyXAi2Pt5xnrXBluNq4iThU9TKlUcnZnfeMfGEHhazjCxfab64OIIAevufauCPxB8XCTzv9DKYz5Hl/1rO8T3Gq3XjcP4gso7K4FsBHEkokGPXIqGs8fmFWjV9nT6Cq1ZRlZHq/hDxbb+K9NeVYzb3UDbJ7djypx1Hsa6GvEfClxrFr40nbw7Zx3krWf76OSURqBu4OfWu8/tTx7/0L1j/4GrXr4er7alGo1ubxlzRTOqv7+30zT5ry9kEcEKF3Y9gK8ru/iX4g1KRpNHhgsbXP7szJvdh6n60/4gX3iybwnImtaVbWlkZY/MkiuQ5+8MDH1rn1wFG3pjjFcGZYyph1FU92ZVqjhax2/hT4iXF7qyaV4hhjhnm/1M8Zwjn09jXfzTR28DzTuEjjUszMcAAd6+fNULLHA8H/AB8LOhix3bNdv4xvfGr+Db9dR0uzgtjFiaSO43MF78V0YHESxFHnluXTm5xuyrf/ABO1nU53bw/BFaWYOI5bhdzP747Vo+GviPdtqkWn+JYokE52xXUXC7vQjtXFwBBbxiP7gUbfpVbVv+Qe5H3wQU/3s8V5FPNK0q6TXut2sYRrScj6Forh7O88ffYYNumacR5a4LXHJ471N9r8f/8AQM0z/v8A19KdZv6trlvpF3p1vcI7NqFx5EZXs2M81p15R4qn8Xtq3h46hZ2McovwbYRy5DPjo3oK6X7T8QP+fDS/+/xoA7KiuN+0fED/AJ8tK/7+mjz/AIg/8+mk/wDf00Aa/irxPa+FtIN3cqZZHbZBCp+aRj2rzp/iF4smlE0a2cEfXyCmfzNVPG769J4j0geJo7aPCSmAW7EgnjOc9+lUq8PMcdVoVFCn6nNVqOLsj07wZ41j8TpLb3MP2XULcAyRZ4Yf3l9q0PFPia18LaQby5BkkdtkMKn5pGPYV5HoZ1NfHVn/AGAITetDJkTkhCuO+KueOG19vEGjr4mFqo2ymAWxJUnjOc9+ld9PEuWF9u1rZv7jVTvDmLT/ABC8WTSiaJbOCPqICmc/U123gzxrH4nWW2uYfsuoW4Bkizww/vL7V5jS6OdTXxxYf8I/5P25o5B++zs245zivMwOYVa1bkns/wADGnVlKVme71keJvEVt4Z0d766Bc52xRL1kY9BWJj4h/3tF/8AH64zxz/wkf8AaWlL4mazMW5jELXON3vmvcqz9nTlPsjpk7K5K/xC8WTyCaJbO2jzkQsm449Ca7PwX42XxJ5tnfQC11G3ALpn5ZAf4l9q80pdM/tL/hNNL/sFoVvmEgBmB2bcfxY5xXhYLMa1auoT2f4HNTqylKzPd6y/EXiC08NaNLqF8SVThI1+9Ix6KKwfJ+In/Pzof/fElcb47TxIt7pC+J5bGSAysY/sisAGxxnNe7Vn7Om59kdLdlckf4h+K7mXzoUtLWPJIgZNxx2BNdd4L8djxDM+n6lALXUYl3YB+WUeq15zTLNb4+LtK/sV4UvzIQjTAlAMd8c4rwsFmNarXUJ7P8Dmp1ZSlZnvdU9Xe4TR7n7DGZLloysSjux4H+NZWiQeLo7/AHeILvS5bXafltY3D7u3XjFdDX0J1HnXhT4UWWnSC/18re3rHf5Z+4h6/jXoiIqIFRQqgYAAwBS0UAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUVHcTrbW0k8gYrGpYhRk4HoKNgJK4Pxt4Cn1bUBrOhSpFfhAssMg+ScDpk9jXY6TqUOsaTbajarKkNzGJEWZCjgH1B6GrdTOCmnGS0Fo0eLw+C/F95N5DWENipOGuHmDAD1AFen+FfDdt4W0OPT7ZjK+4vNMwwZXPUn9B9BWzRWVHD0qCapq1xRjGOxynjXwUnieKK6tJhbalbZ8qUrkMP7re1cB/wh/jEyeQNKgB/57+eNn5da9qoqa2Fo12nUjewpQjLdHMeCvBsfhWzmeeYXWoXTbp7jbj6Ko7Af57AdPRRXQkkrIsp6tpVtrWk3Gn3yb4J0KsB1HuPcV5PeeAfE+jyeRYQx6tbLxHIJBG4HYEHjivZKKyrUKdePLUVyZRUlZnmnhT4eXv9rQ6r4mMa/Zzuhs0O4Bv7zHv9K9Fu7SC/s5bW7jEsEyFJEbowIwRU1FXTpxpxUIKyGkkrI8dvvh54h0SUw6OiatZD/VbnEciD0OeDj1rQ8O/DrUb3UorzxQscFvbsHSzjbcZG/wBo+ntXqVFYLCUI1PaqOpPJG97ABgYHAooorqLMnWtBj1m90u4kmaM6dci4UAZ3nGMGtaiigAooooA57xh4Sg8WaWkLym3urd/Mt7hRko3uO4PpXnEng7xhBN5A02C45wLhZwE+pB5r2iiuethqVe3tI3sTKEZbnHeCfA58OyS6hqUy3OpTrtLKvyxL/dX/ABrR8YeE7fxZpSQPIbe5gfzLe4UZKN9O4Pp9K6CitlGKjypaDsrWPF5PB3jCCbyBpsFx2Fwk4C/Ug8123gnwOfD0suo6nMtzqU67SVX5YV/ur/jXZUVhSwtGjJypxsyYwjF3SCsLxb4Xg8VaP9klkaCaNt8EyjJRvp3FbtFdO5Z4vJ4O8YW83kDTYLoZwLhJgqn3IPNdp4I8DvoU8mqavKlxqUyBVCD5Ldf7q+p9T/k9pRXNSwlCjLmhGzIjCMXdIKxfFXhm18VaK9jdMYnB3wzqPmicdDW1RXSWeLSeDPGFrN5A0+G8AJAuEmCgj1IPT8K6/wAE+ApNHvTq+uSJPqDLtjjQfJAPY9z713VFc1LCUaUnOEbMiMIxd0gooorpLCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAqK5nW1tJZ3+7Ehc/QDNS02WNZoXjkGUdSrD1BpO9tBrfU8u0l/GPiXw1N4stfEX2RpFkktNNWFTCEGcBj1JOOtVLvxP4nf4eeBLmwv9uqatfx29xLIoYOGWTO4dwMA4GOlaLeGvGfhrQr7RtEv9LbRAkjQ3Nwjm5t0OSUCj5W9iSPoag8M+Hb3Xfhz8PpbSSIDTLqO7nMrEFkAkU4wDk5YelUrN+V4/rcNr/P/gC3KeLtJ8eWXhyLxRJdwatavM1xcQLvtih52Acc5HXpVzQ9T1zS9X8T6BqWryan9gtFuba7kjCyLuB4OODg10Oo+Hby7+I2ka9E8ItLK0mhkVmO8s5BGBjGOPWqreFL8+L/ABDqm+DyNSsEtoRuO4MAc7hjgc+9Rry/J/rYatf7v+CcgfEvie68N/Dj7BqZS81osl3NIgYOPJY7iO5GMgccgVuabPr3hv4k2Wh6jrkus2Wp2skym4jVXhdCOAR1BzRp/gPVbWx8BQyS2pbw6zG82u2GzCyfJ8vPLDrit3U/Dt5efETR9dieEWtlazQyqzHeWYjGBjGOPWtHbn+b+63+ZC+Fen43OmoooqRhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQBx3j3XNSs59H0TQp0tb7WLgxC5dd3koBlmA7msSQeKfDfxA8MaZc+IZdT0u/llErTRqshZYmIU46rxntjFdP4y8MXOvwWV1pN0lnq2mzefaSyLuQnoVYehFcVfJ4pf4qeDJ/FUunRgzXCx2unh2QHyWy5d+SfbAxz1pw3+Y3t8ifTz4q8U+I/F1pD4lm0yz0zUTFbGGFWf8A1aHbk/wjPTryaz7C/wDGmt/DiXxPL4k+x3FirlLeCAGOfyzgl88/Njt0ru/C/hu90bVPE1zdvCyarqDXUAjYkhDGi4bIGDlT0zVDRvB2o6f8Lbvw5PJbm9mSZVZXJj+ckjJxnv6VGqhpvZffYrTm17/gU9f8S6lP4W8PXi6vZ+H7TUYkkvtQndAYgUztjViMsTx3xWT4H8Xy3HxFbw/aeJ5fEmmz2L3C3M0IVonVgpUMBhhz+FX9Q8B69CvhTUNIbTLrUNCtDbPaahuMEmVALqwGVYY4OO9WbDwr4tb4i2vinWrvTZFjsJLU2dsHRYcsGXaxyXyQck47YFae7zvtr+Tt+hH2TD1DxT4p0PXp/BHnPd6pqUu/S9RkxhISfmLY7p29a9VsLeW006C3uLmS6ljjCvPJjdIe5OK88k+GV9q1lqOqa1dwr4puJhNaXcDEpZ7D+7RCQDt9eOa7/STqH9kW39tLAt+IwJ/s7Foy/cqSAcfhUr4bPf8Ar+mD30/r+uhcooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAbJGk0bRyqGRwQykcEVFY2NrptlFZ2EEdvbQrtjijXCqPQCiigCeiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAqtcabZXV7bXdzaxS3FqS0ErLloiRgkHtkHFFFAFmiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigD//2Q==" - } - }, - "cell_type": "markdown", - "id": "f5454823", - "metadata": {}, - "source": [ - "## Overview\n", - "\n", - "Deep Learning has touched almost every industry and has transformed the way industries operate and provide services. We perform or experience real-time analytics all the time around us, for example, an advertisement that you saw while swiping through the stories on Instagram, or the video recommendation that floated on your youtube home screen. To cater to these real-time inferences, deep learning practitioners need to maximise model throughput while having highly accurate predictions. Among many techniques, quantization can be used to accelerate models.\n", - "\n", - "Model Quantization is a popular way of optimization which reduces the size of models thereby accelerating inference, while also opening up the possibilities of deployments on devices with lower computation power such as Jetson. Simply put, quantization is a process of mapping input values from a larger set to output values in a smaller set. In the context of deep learning, we often train deep learning models using floating-point 32 bit arithmetic (FP32) as we can take advantage of a wider range of numbers, resulting in more accurate models. The model data (network parameters and activations) are converted from this floating point representation to a lower precision representation, typically using 8-bit integers (int8). In the case of int8, the range [qmin, qmax] would be [-128, 127].\n", - "\n", - "![img1.JPG](attachment:img1.JPG)\n", - "\n", - "A quick rationale of how higher throughput is achieved through quantization can be shown through the following thought experiment: Imagine the complexity of multiplying 3.999x2.999 versus 4x3. The latter is easier to perform than the former. This is the simplicity in calculation seen by quantizing the numbers to lower precision. However, the challenge here is that round errors can result in a lower accuracy model. To address this loss of accuracy, different quantization techniques have been developed. These techniques can be classified into two categories, post-training quantization (PTQ) and quantization-aware training (QAT).\n", - "\n", - "In this notebook, we illustrate the workflow that you can adopt in order to quantize a deep learning model using TensorRT. The notebook takes you through an example of Mobilenetv2 for a classification task on a subset of Imagenet Dataset called Imagenette which has 10 classes. \n", - "\n", - "1. [Requirements](#1)\n", - "2. [Setup a baseline Mobilenetv2 model](#2)\n", - "3. [Convert to TensorRT](#3)\n", - "4. [Post Training Quantization (PTQ)](#4)\n", - "5. [Quantization Aware Training (QAT)](#5)\n", - "6. [Evaluation and Benchmarking](#6)\n", - "7. [Conclusion](#7)\n", - "8. [References](#8)\n", - "\n", - "This notebook is implemented using the NGC pytorch container nvcr.io/nvidia/pytorch:22.04-py3. Follow instructions here https://ngc.nvidia.com/setup/api-key to setup your own API key to use the NGC service through the Docker client. " - ] - }, - { - "cell_type": "markdown", - "id": "06b37d07", - "metadata": {}, - "source": [ - "\n", - "## 1. Requirements\n", - "Please install the required dependencies and import these libraries accordingly" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "id": "0a068b12", - "metadata": {}, - "outputs": [], - "source": [ - "!pip install ipywidgets --trusted-host pypi.org --trusted-host pypi.python.org --trusted-host=files.pythonhosted.org\n", - "!pip install wget\n", - "!pip install pycuda" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "id": "4e2e58b2", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "2.1.2\n" - ] - } - ], - "source": [ - "import torch\n", - "import torch.nn as nn\n", - "import torch.optim as optim\n", - "import torch.utils.data as data\n", - "import torchvision.transforms as transforms\n", - "from torchvision import models, datasets\n", - "\n", - "import pytorch_quantization\n", - "from pytorch_quantization import nn as quant_nn\n", - "from pytorch_quantization import quant_modules\n", - "from pytorch_quantization import calib\n", - "from tqdm import tqdm\n", - "\n", - "print(pytorch_quantization.__version__)\n", - "\n", - "import os\n", - "import tensorrt as trt\n", - "import numpy as np\n", - "import time\n", - "import wget\n", - "import tarfile\n", - "import shutil" - ] - }, - { - "cell_type": "markdown", - "id": "0575e590", - "metadata": {}, - "source": [ - "\n", - "## 2. Setup a baseline Mobilenetv2 Model" - ] - }, - { - "cell_type": "markdown", - "id": "a83b886f", - "metadata": {}, - "source": [ - "#### Preparing the Dataset\n", - "\n", - "Imagenette is a subset of ImageNet and has 10 classes. The classes are as follows in the order of their labels : 'tench', 'English springer', 'cassette player', 'chain saw', 'church', 'French horn', 'garbage truck', 'gas pump', 'golf ball' and 'parachute'. " - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "id": "50d60fbe", - "metadata": {}, - "outputs": [], - "source": [ - "def download_data(DATA_DIR):\n", - " if os.path.exists(DATA_DIR):\n", - " if not os.path.exists(os.path.join(DATA_DIR, 'imagenette2-320')):\n", - " url = 'https://s3.amazonaws.com/fast-ai-imageclas/imagenette2-320.tgz'\n", - " wget.download(url)\n", - " # open file\n", - " file = tarfile.open('imagenette2-320.tgz')\n", - " # extracting file\n", - " file.extractall(DATA_DIR)\n", - " file.close()\n", - " else:\n", - " print(\"This directory doesn't exist. Create the directory and run again\")" - ] - }, - { - "cell_type": "markdown", - "id": "2e25dc45", - "metadata": {}, - "source": [ - "Let's create the data directory if it doesn't exist." - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "id": "4a4d8949", - "metadata": {}, - "outputs": [], - "source": [ - "if not os.path.exists(\"./data\"):\n", - " os.mkdir(\"./data\")\n", - "download_data(\"./data\")" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "id": "07d1fc63", - "metadata": {}, - "outputs": [], - "source": [ - "# Define main data directory\n", - "DATA_DIR = './data/imagenette2-320' \n", - "# Define training and validation data paths\n", - "TRAIN_DIR = os.path.join(DATA_DIR, 'train') \n", - "VAL_DIR = os.path.join(DATA_DIR, 'val')" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "id": "acd3cd99", - "metadata": {}, - "outputs": [], - "source": [ - "# Performing Transformations on the dataset and defining training and validation dataloaders\n", - "transform = transforms.Compose([\n", - " transforms.Resize(256),\n", - " transforms.CenterCrop(224),\n", - " transforms.ToTensor(),\n", - " ])\n", - "train_dataset = datasets.ImageFolder(TRAIN_DIR, transform=transform)\n", - "val_dataset = datasets.ImageFolder(VAL_DIR, transform=transform)\n", - "calib_dataset = torch.utils.data.random_split(val_dataset, [2901, 1024])[1]\n", - "\n", - "train_dataloader = data.DataLoader(train_dataset, batch_size=64, shuffle=True, drop_last=True)\n", - "val_dataloader = data.DataLoader(val_dataset, batch_size=64, shuffle=False, drop_last=True)\n", - "calib_dataloader = data.DataLoader(calib_dataset, batch_size=64, shuffle=False, drop_last=True)" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "id": "a2f8914c", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "tensor(0)\n" - ] - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "

" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "# Visualising an image from the validation set\n", - "import matplotlib.pyplot as plt\n", - "for images, labels in val_dataloader:\n", - " print(labels[0])\n", - " image = images[0]\n", - " img = image.swapaxes(0, 1)\n", - " img = img.swapaxes(1, 2)\n", - " plt.imshow(img)\n", - " break" - ] - }, - { - "cell_type": "markdown", - "id": "4b7441e6", - "metadata": {}, - "source": [ - "#### Setting up Mobilenetv2\n", - "\n", - "Mobilenetv2 available in Torchvision is pretrained on the ImageNet that has 1000 classes. The Imagenette dataset has 10 classes. \n", - "We set up this model by freezing the weights excpet for the last classification layer and train only the last classification layer to be able to predict the 10 classes of the dataset. " - ] - }, - { - "cell_type": "markdown", - "id": "b9577f2a", - "metadata": {}, - "source": [ - "*Define the Mobilenetv2 model*" - ] - }, - { - "cell_type": "code", - "execution_count": 29, - "id": "c29ae7b8", - "metadata": {}, - "outputs": [], - "source": [ - "# This function allows you to set the all the parameters to not have gradients, \n", - "# allowing you to freeze the model and not undergo training during the train step. \n", - "def set_parameter_requires_grad(model, feature_extracting):\n", - " if feature_extracting:\n", - " for param in model.parameters():\n", - " param.requires_grad = False\n", - " \n", - "feature_extract = True #This varaible can be set False if you want to finetune the model by updating all the parameters. \n", - "model = models.mobilenet_v2(pretrained=True)\n", - "set_parameter_requires_grad(model, feature_extract)\n", - "#Define a classification head for 10 classes.\n", - "model.classifier[1] = nn.Linear(1280, 10)\n", - "model = model.cuda()" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "id": "5c03df98", - "metadata": {}, - "outputs": [], - "source": [ - "# Declare Learning rate\n", - "lr = 0.0001\n", - "\n", - "# Use cross entropy loss for classification and SGD optimizer\n", - "criterion = nn.CrossEntropyLoss()\n", - "optimizer = optim.SGD(model.parameters(), lr=lr)" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "7095a995", - "metadata": {}, - "outputs": [], - "source": [ - "# Define functions for training, evalution, saving checkpoint and train parameter setting function\n", - "def train(model, dataloader, crit, opt, epoch):\n", - " model.train()\n", - " running_loss = 0.0\n", - " for batch, (data, labels) in enumerate(dataloader):\n", - " data, labels = data.cuda(), labels.cuda(non_blocking=True)\n", - " opt.zero_grad()\n", - " out = model(data)\n", - " loss = crit(out, labels)\n", - " loss.backward()\n", - " opt.step()\n", - " running_loss += loss.item()\n", - " if batch % 100 == 99:\n", - " print(\"Batch: [%5d | %5d] loss: %.3f\" % (batch + 1, len(dataloader), running_loss / 100))\n", - " running_loss = 0.0\n", - " \n", - "def evaluate(model, dataloader, crit, epoch):\n", - " total = 0\n", - " correct = 0\n", - " loss = 0.0\n", - " class_probs = []\n", - " class_preds = []\n", - " model.eval()\n", - " with torch.no_grad():\n", - " for data, labels in dataloader:\n", - " data, labels = data.cuda(), labels.cuda(non_blocking=True)\n", - " out = model(data)\n", - " loss += crit(out, labels)\n", - " preds = torch.max(out, 1)[1]\n", - " class_preds.append(preds)\n", - " total += labels.size(0)\n", - " correct += (preds == labels).sum().item()\n", - " return correct / total\n", - "\n", - "def save_checkpoint(state, ckpt_path=\"checkpoint.pth\"):\n", - " torch.save(state, ckpt_path)\n", - " print(\"Checkpoint saved\")\n", - " \n", - "# Helper function to benchmark the model\n", - "cudnn.benchmark = True\n", - "def benchmark(model, input_shape=(1024, 1, 32, 32), dtype='fp32', nwarmup=50, nruns=1000):\n", - " input_data = torch.randn(input_shape)\n", - " input_data = input_data.to(\"cuda\")\n", - " if dtype=='fp16':\n", - " input_data = input_data.half()\n", - " \n", - " with torch.no_grad():\n", - " for _ in range(nwarmup):\n", - " features = model(input_data)\n", - " torch.cuda.synchronize()\n", - " \n", - " timings = []\n", - " with torch.no_grad():\n", - " for i in range(1, nruns+1):\n", - " start_time = time.time()\n", - " output = model(input_data)\n", - " torch.cuda.synchronize()\n", - " end_time = time.time()\n", - " timings.append(end_time - start_time)\n", - "\n", - " print('Average batch time: %.2f ms'%(np.mean(timings)*1000))" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "id": "02a625c9", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch: [ 1 / 5] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 2.315\n", - "Test Acc: 22.93%\n", - "Epoch: [ 2 / 5] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 2.177\n", - "Test Acc: 35.09%\n", - "Epoch: [ 3 / 5] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 2.053\n", - "Test Acc: 49.33%\n", - "Epoch: [ 4 / 5] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 1.935\n", - "Test Acc: 61.50%\n", - "Epoch: [ 5 / 5] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 1.836\n", - "Test Acc: 71.11%\n", - "Checkpoint saved\n" - ] - } - ], - "source": [ - "# Train the model for 5 epochs to attain an acceptable accuracy.\n", - "num_epochs=5\n", - "for epoch in range(num_epochs):\n", - " print('Epoch: [%5d / %5d] LR: %f' % (epoch + 1, num_epochs, lr))\n", - "\n", - " train(model, train_dataloader, criterion, optimizer, epoch)\n", - " test_acc = evaluate(model, val_dataloader, criterion, epoch)\n", - "\n", - " print(\"Test Acc: {:.2f}%\".format(100 * test_acc))\n", - " \n", - "save_checkpoint({'epoch': epoch + 1,\n", - " 'model_state_dict': model.state_dict(),\n", - " 'acc': test_acc,\n", - " 'opt_state_dict': optimizer.state_dict()\n", - " },\n", - " ckpt_path=\"models/mobilenetv2_base_ckpt\")" - ] - }, - { - "cell_type": "markdown", - "id": "b829681d", - "metadata": {}, - "source": [ - "We will first generate and evaluate our models and then finally look at the performance to the end of the notebook." - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "id": "411d0ebc", - "metadata": { - "scrolled": true - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mobilenetv2 Baseline accuracy: 71.11%\n" - ] - } - ], - "source": [ - "# Evaluate the baseline model\n", - "test_acc = evaluate(model, val_dataloader, criterion, 0)\n", - "print(\"Mobilenetv2 Baseline accuracy: {:.2f}%\".format(100 * test_acc))" - ] - }, - { - "cell_type": "markdown", - "id": "71fdd581", - "metadata": {}, - "source": [ - "\n", - "### Convert to TensorRT\n", - "\n", - "TensorRT is an SDK facilitating high-performance deep learning inference, optimized to run on NVIDIA GPUs. It accelerates models through graph optimization and quantization. This notebook uses the trtexec CLI tool to build TensorRT engine. " - ] - }, - { - "cell_type": "markdown", - "id": "f75ab9fd", - "metadata": {}, - "source": [ - "Let us convert the above FP32 Mobilenetv2 into a TensorRT engine. Before we do that, we need to first export our model into ONNX format. ONNX is a standard for representing deep learning models enabling them to be transferred between frameworks. The average run time of the TRT model would be the 'GPU Compute Time' printed in the logs." - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "id": "e24451cf", - "metadata": { - "scrolled": true - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "&&&& RUNNING TensorRT.trtexec [TensorRT v8205] # trtexec --onnx=models/mobilenetv2_base.onnx --saveEngine=models/mobilenetv2_base.trt\n", - "[07/25/2022-16:42:22] [I] === Model Options ===\n", - "[07/25/2022-16:42:22] [I] Format: ONNX\n", - "[07/25/2022-16:42:22] [I] Model: models/mobilenetv2_base.onnx\n", - "[07/25/2022-16:42:22] [I] Output:\n", - "[07/25/2022-16:42:22] [I] === Build Options ===\n", - "[07/25/2022-16:42:22] [I] Max batch: explicit batch\n", - "[07/25/2022-16:42:22] [I] Workspace: 16 MiB\n", - "[07/25/2022-16:42:22] [I] minTiming: 1\n", - "[07/25/2022-16:42:22] [I] avgTiming: 8\n", - "[07/25/2022-16:42:22] [I] Precision: FP32\n", - "[07/25/2022-16:42:22] [I] Calibration: \n", - "[07/25/2022-16:42:22] [I] Refit: Disabled\n", - "[07/25/2022-16:42:22] [I] Sparsity: Disabled\n", - "[07/25/2022-16:42:22] [I] Safe mode: Disabled\n", - "[07/25/2022-16:42:22] [I] DirectIO mode: Disabled\n", - "[07/25/2022-16:42:22] [I] Restricted mode: Disabled\n", - "[07/25/2022-16:42:22] [I] Save engine: models/mobilenetv2_base.trt\n", - "[07/25/2022-16:42:22] [I] Load engine: \n", - "[07/25/2022-16:42:22] [I] Profiling verbosity: 0\n", - "[07/25/2022-16:42:22] [I] Tactic sources: Using default tactic sources\n", - "[07/25/2022-16:42:22] [I] timingCacheMode: local\n", - "[07/25/2022-16:42:22] [I] timingCacheFile: \n", - "[07/25/2022-16:42:22] [I] Input(s)s format: fp32:CHW\n", - "[07/25/2022-16:42:22] [I] Output(s)s format: fp32:CHW\n", - "[07/25/2022-16:42:22] [I] Input build shapes: model\n", - "[07/25/2022-16:42:22] [I] Input calibration shapes: model\n", - "[07/25/2022-16:42:22] [I] === System Options ===\n", - "[07/25/2022-16:42:22] [I] Device: 0\n", - "[07/25/2022-16:42:22] [I] DLACore: \n", - "[07/25/2022-16:42:22] [I] Plugins:\n", - "[07/25/2022-16:42:22] [I] === Inference Options ===\n", - "[07/25/2022-16:42:22] [I] Batch: Explicit\n", - "[07/25/2022-16:42:22] [I] Input inference shapes: model\n", - "[07/25/2022-16:42:22] [I] Iterations: 10\n", - "[07/25/2022-16:42:22] [I] Duration: 3s (+ 200ms warm up)\n", - "[07/25/2022-16:42:22] [I] Sleep time: 0ms\n", - "[07/25/2022-16:42:22] [I] Idle time: 0ms\n", - "[07/25/2022-16:42:22] [I] Streams: 1\n", - "[07/25/2022-16:42:22] [I] ExposeDMA: Disabled\n", - "[07/25/2022-16:42:22] [I] Data transfers: Enabled\n", - "[07/25/2022-16:42:22] [I] Spin-wait: Disabled\n", - "[07/25/2022-16:42:22] [I] Multithreading: Disabled\n", - "[07/25/2022-16:42:22] [I] CUDA Graph: Disabled\n", - "[07/25/2022-16:42:22] [I] Separate profiling: Disabled\n", - "[07/25/2022-16:42:22] [I] Time Deserialize: Disabled\n", - "[07/25/2022-16:42:22] [I] Time Refit: Disabled\n", - "[07/25/2022-16:42:22] [I] Skip inference: Disabled\n", - "[07/25/2022-16:42:22] [I] Inputs:\n", - "[07/25/2022-16:42:22] [I] === Reporting Options ===\n", - "[07/25/2022-16:42:22] [I] Verbose: Disabled\n", - "[07/25/2022-16:42:22] [I] Averages: 10 inferences\n", - "[07/25/2022-16:42:22] [I] Percentile: 99\n", - "[07/25/2022-16:42:22] [I] Dump refittable layers:Disabled\n", - "[07/25/2022-16:42:22] [I] Dump output: Disabled\n", - "[07/25/2022-16:42:22] [I] Profile: Disabled\n", - "[07/25/2022-16:42:22] [I] Export timing to JSON file: \n", - "[07/25/2022-16:42:22] [I] Export output to JSON file: \n", - "[07/25/2022-16:42:22] [I] Export profile to JSON file: \n", - "[07/25/2022-16:42:22] [I] \n", - "[07/25/2022-16:42:22] [I] === Device Information ===\n", - "[07/25/2022-16:42:22] [I] Selected Device: NVIDIA Graphics Device\n", - "[07/25/2022-16:42:22] [I] Compute Capability: 8.0\n", - "[07/25/2022-16:42:22] [I] SMs: 124\n", - "[07/25/2022-16:42:22] [I] Compute Clock Rate: 1.005 GHz\n", - "[07/25/2022-16:42:22] [I] Device Global Memory: 47681 MiB\n", - "[07/25/2022-16:42:22] [I] Shared Memory per SM: 164 KiB\n", - "[07/25/2022-16:42:22] [I] Memory Bus Width: 6144 bits (ECC enabled)\n", - "[07/25/2022-16:42:22] [I] Memory Clock Rate: 1.215 GHz\n", - "[07/25/2022-16:42:22] [I] \n", - "[07/25/2022-16:42:22] [I] TensorRT version: 8.2.5\n", - "[07/25/2022-16:42:23] [I] [TRT] [MemUsageChange] Init CUDA: CPU +440, GPU +0, now: CPU 452, GPU 5848 (MiB)\n", - "[07/25/2022-16:42:23] [I] [TRT] [MemUsageSnapshot] Begin constructing builder kernel library: CPU 452 MiB, GPU 5848 MiB\n", - "[07/25/2022-16:42:23] [I] [TRT] [MemUsageSnapshot] End constructing builder kernel library: CPU 669 MiB, GPU 5920 MiB\n", - "[07/25/2022-16:42:23] [I] Start parsing network model\n", - "[07/25/2022-16:42:23] [I] [TRT] ----------------------------------------------------------------\n", - "[07/25/2022-16:42:23] [I] [TRT] Input filename: models/mobilenetv2_base.onnx\n", - "[07/25/2022-16:42:23] [I] [TRT] ONNX IR version: 0.0.7\n", - "[07/25/2022-16:42:23] [I] [TRT] Opset version: 13\n", - "[07/25/2022-16:42:23] [I] [TRT] Producer name: pytorch\n", - "[07/25/2022-16:42:23] [I] [TRT] Producer version: 1.13.0\n", - "[07/25/2022-16:42:23] [I] [TRT] Domain: \n", - "[07/25/2022-16:42:23] [I] [TRT] Model version: 0\n", - "[07/25/2022-16:42:23] [I] [TRT] Doc string: \n", - "[07/25/2022-16:42:23] [I] [TRT] ----------------------------------------------------------------\n", - "[07/25/2022-16:42:23] [I] Finish parsing network model\n", - "[07/25/2022-16:42:24] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +839, GPU +362, now: CPU 1532, GPU 6290 (MiB)\n", - "[07/25/2022-16:42:24] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +128, GPU +58, now: CPU 1660, GPU 6348 (MiB)\n", - "[07/25/2022-16:42:24] [I] [TRT] Local timing cache in use. Profiling results in this builder pass will not be stored.\n", - "[07/25/2022-16:42:28] [I] [TRT] Some tactics do not have sufficient workspace memory to run. Increasing workspace size may increase performance, please check verbose output.\n", - "[07/25/2022-16:43:21] [I] [TRT] Detected 1 inputs and 1 output network tensors.\n", - "[07/25/2022-16:43:21] [I] [TRT] Total Host Persistent Memory: 82528\n", - "[07/25/2022-16:43:21] [I] [TRT] Total Device Persistent Memory: 8861184\n", - "[07/25/2022-16:43:21] [I] [TRT] Total Scratch Memory: 4194304\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageStats] Peak memory usage of TRT CPU/GPU memory allocators: CPU 8 MiB, GPU 624 MiB\n", - "[07/25/2022-16:43:21] [I] [TRT] [BlockAssignment] Algorithm ShiftNTopDown took 1.67234ms to assign 4 blocks to 59 nodes requiring 449576960 bytes.\n", - "[07/25/2022-16:43:21] [I] [TRT] Total Activation Memory: 449576960\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 2512, GPU 6760 (MiB)\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +1, GPU +10, now: CPU 2513, GPU 6770 (MiB)\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in building engine: CPU +0, GPU +9, now: CPU 0, GPU 9 (MiB)\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] Init CUDA: CPU +0, GPU +0, now: CPU 2521, GPU 6724 (MiB)\n", - "[07/25/2022-16:43:21] [I] [TRT] Loaded engine size: 10 MiB\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +10, now: CPU 2522, GPU 6746 (MiB)\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +1, GPU +8, now: CPU 2523, GPU 6754 (MiB)\n", - "[07/25/2022-16:43:21] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +8, now: CPU 0, GPU 8 (MiB)\n", - "[07/25/2022-16:43:22] [I] Engine built in 59.1433 sec.\n", - "[07/25/2022-16:43:22] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +10, now: CPU 2289, GPU 6696 (MiB)\n", - "[07/25/2022-16:43:22] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +1, GPU +8, now: CPU 2290, GPU 6704 (MiB)\n", - "[07/25/2022-16:43:22] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +438, now: CPU 0, GPU 446 (MiB)\n", - "[07/25/2022-16:43:22] [I] Using random values for input input.1\n", - "[07/25/2022-16:43:22] [I] Created input binding for input.1 with dimensions 64x3x224x224\n", - "[07/25/2022-16:43:22] [I] Using random values for output 536\n", - "[07/25/2022-16:43:22] [I] Created output binding for 536 with dimensions 64x10\n", - "[07/25/2022-16:43:22] [I] Starting inference\n", - "[07/25/2022-16:43:25] [I] Warmup completed 34 queries over 200 ms\n", - "[07/25/2022-16:43:25] [I] Timing trace has 501 queries over 3.01732 s\n", - "[07/25/2022-16:43:25] [I] \n", - "[07/25/2022-16:43:25] [I] === Trace details ===\n", - "[07/25/2022-16:43:25] [I] Trace averages of 10 runs:\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88872 ms - Host latency: 8.93236 ms (end to end 11.4268 ms, enqueue 2.05089 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88841 ms - Host latency: 8.92661 ms (end to end 11.4266 ms, enqueue 2.07079 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88647 ms - Host latency: 8.93378 ms (end to end 11.4245 ms, enqueue 2.07513 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.8879 ms - Host latency: 8.93474 ms (end to end 11.4218 ms, enqueue 2.04516 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88708 ms - Host latency: 8.92472 ms (end to end 11.2913 ms, enqueue 2.04477 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88964 ms - Host latency: 8.93073 ms (end to end 11.4241 ms, enqueue 2.04273 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.89016 ms - Host latency: 8.92474 ms (end to end 11.4283 ms, enqueue 2.04633 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88841 ms - Host latency: 8.92583 ms (end to end 11.4307 ms, enqueue 2.05944 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88973 ms - Host latency: 8.92712 ms (end to end 11.4225 ms, enqueue 2.06941 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.88892 ms - Host latency: 8.92521 ms (end to end 11.4224 ms, enqueue 2.05708 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.92097 ms - Host latency: 8.96465 ms (end to end 11.4841 ms, enqueue 2.04125 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.09852 ms - Host latency: 9.13358 ms (end to end 11.7906 ms, enqueue 2.04748 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.44015 ms - Host latency: 9.47874 ms (end to end 12.5498 ms, enqueue 2.05565 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.25358 ms - Host latency: 9.28981 ms (end to end 12.1605 ms, enqueue 2.05262 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.14546 ms - Host latency: 9.18715 ms (end to end 11.9508 ms, enqueue 2.06964 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.02576 ms - Host latency: 9.06241 ms (end to end 11.7147 ms, enqueue 2.04923 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.92704 ms - Host latency: 8.96814 ms (end to end 11.5024 ms, enqueue 2.04821 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.01957 ms - Host latency: 9.05573 ms (end to end 11.6706 ms, enqueue 2.04988 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.94579 ms - Host latency: 8.98406 ms (end to end 11.5354 ms, enqueue 2.13973 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.89835 ms - Host latency: 8.94883 ms (end to end 11.4496 ms, enqueue 2.08344 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.99513 ms - Host latency: 9.03672 ms (end to end 11.6076 ms, enqueue 2.0929 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.08859 ms - Host latency: 9.12035 ms (end to end 11.8224 ms, enqueue 2.06177 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.9467 ms - Host latency: 8.987 ms (end to end 11.5444 ms, enqueue 2.06372 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.89579 ms - Host latency: 8.93199 ms (end to end 11.4334 ms, enqueue 2.04498 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.91914 ms - Host latency: 8.95847 ms (end to end 11.4744 ms, enqueue 2.06753 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.92528 ms - Host latency: 8.96528 ms (end to end 11.4935 ms, enqueue 2.05543 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.92607 ms - Host latency: 8.96593 ms (end to end 11.4996 ms, enqueue 2.05464 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.11134 ms - Host latency: 9.14991 ms (end to end 11.8276 ms, enqueue 2.06058 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.24971 ms - Host latency: 9.2879 ms (end to end 12.1685 ms, enqueue 2.05168 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.1583 ms - Host latency: 9.19552 ms (end to end 11.9784 ms, enqueue 2.05416 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.03793 ms - Host latency: 9.07539 ms (end to end 11.7194 ms, enqueue 2.04376 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.03723 ms - Host latency: 9.07742 ms (end to end 11.7207 ms, enqueue 2.04446 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.16055 ms - Host latency: 9.1936 ms (end to end 11.9269 ms, enqueue 2.06987 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.24443 ms - Host latency: 9.28486 ms (end to end 12.1531 ms, enqueue 2.04836 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.94749 ms - Host latency: 8.98728 ms (end to end 11.5623 ms, enqueue 2.05354 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.91284 ms - Host latency: 8.95781 ms (end to end 11.4716 ms, enqueue 2.04207 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.98567 ms - Host latency: 9.02083 ms (end to end 11.6108 ms, enqueue 2.04358 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.93113 ms - Host latency: 8.97266 ms (end to end 11.533 ms, enqueue 2.06318 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.89543 ms - Host latency: 8.92844 ms (end to end 11.4434 ms, enqueue 2.05273 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.95349 ms - Host latency: 8.99211 ms (end to end 11.5469 ms, enqueue 2.07312 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.94853 ms - Host latency: 8.98025 ms (end to end 11.5569 ms, enqueue 2.04573 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.97844 ms - Host latency: 9.01548 ms (end to end 11.6017 ms, enqueue 2.05762 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 5.96038 ms - Host latency: 9.00027 ms (end to end 11.5838 ms, enqueue 2.04302 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.03623 ms - Host latency: 9.07041 ms (end to end 11.7005 ms, enqueue 2.05886 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.08901 ms - Host latency: 9.12502 ms (end to end 11.8232 ms, enqueue 2.06831 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.07283 ms - Host latency: 9.11433 ms (end to end 11.8008 ms, enqueue 2.07654 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.10923 ms - Host latency: 9.14961 ms (end to end 11.8509 ms, enqueue 2.05337 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.05793 ms - Host latency: 9.09639 ms (end to end 11.776 ms, enqueue 2.06641 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.1678 ms - Host latency: 9.20984 ms (end to end 11.9751 ms, enqueue 2.05627 ms)\n", - "[07/25/2022-16:43:25] [I] Average on 10 runs - GPU latency: 6.05857 ms - Host latency: 9.0998 ms (end to end 11.7799 ms, enqueue 2.06199 ms)\n", - "[07/25/2022-16:43:25] [I] \n", - "[07/25/2022-16:43:25] [I] === Performance summary ===\n", - "[07/25/2022-16:43:25] [I] Throughput: 166.041 qps\n", - "[07/25/2022-16:43:25] [I] Latency: min = 8.90146 ms, max = 9.52582 ms, mean = 9.04623 ms, median = 8.99896 ms, percentile(99%) = 9.50714 ms\n", - "[07/25/2022-16:43:25] [I] End-to-End Host Latency: min = 10.1021 ms, max = 12.6202 ms, mean = 11.6584 ms, median = 11.5563 ms, percentile(99%) = 12.5932 ms\n", - "[07/25/2022-16:43:25] [I] Enqueue Time: min = 1.93103 ms, max = 2.48816 ms, mean = 2.05872 ms, median = 2.05432 ms, percentile(99%) = 2.24194 ms\n", - "[07/25/2022-16:43:25] [I] H2D Latency: min = 3.00195 ms, max = 3.14062 ms, mean = 3.03002 ms, median = 3.02588 ms, percentile(99%) = 3.08609 ms\n", - "[07/25/2022-16:43:25] [I] GPU Compute Time: min = 5.87982 ms, max = 6.47681 ms, mean = 6.00728 ms, median = 5.94946 ms, percentile(99%) = 6.47375 ms\n", - "[07/25/2022-16:43:25] [I] D2H Latency: min = 0.00708008 ms, max = 0.0134277 ms, mean = 0.00893093 ms, median = 0.00878906 ms, percentile(99%) = 0.0117188 ms\n", - "[07/25/2022-16:43:25] [I] Total Host Walltime: 3.01732 s\n", - "[07/25/2022-16:43:25] [I] Total GPU Compute Time: 3.00965 s\n", - "[07/25/2022-16:43:25] [I] Explanations of the performance metrics are printed in the verbose logs.\n", - "[07/25/2022-16:43:25] [I] \n", - "&&&& PASSED TensorRT.trtexec [TensorRT v8205] # trtexec --onnx=models/mobilenetv2_base.onnx --saveEngine=models/mobilenetv2_base.trt\n" - ] - } - ], - "source": [ - "# Exporting to Onnx\n", - "dummy_input = torch.randn(64, 3, 224, 224, device='cuda')\n", - "input_names = [ \"actual_input_1\" ]\n", - "output_names = [ \"output1\" ]\n", - "torch.onnx.export(\n", - " model,\n", - " dummy_input,\n", - " \"models/mobilenetv2_base.onnx\",\n", - " verbose=False,\n", - " opset_version=13,\n", - " do_constant_folding = False)\n", - "\n", - "# Converting ONNX model to TRT\n", - "!trtexec --onnx=models/mobilenetv2_base.onnx --saveEngine=models/mobilenetv2_base.trt" - ] - }, - { - "cell_type": "markdown", - "id": "0a079b97", - "metadata": {}, - "source": [ - "\n", - "## 4. Post Training Quantization (PTQ)" - ] - }, - { - "attachments": { - "img4.JPG": { - "image/jpeg": "/9j/4AAQSkZJRgABAQEAeAB4AAD/4RDuRXhpZgAATU0AKgAAAAgABAE7AAIAAAAMAAAISodpAAQAAAABAAAIVpydAAEAAAAYAAAQzuocAAcAAAgMAAAAPgAAAAAc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAEFubmllIFN1cmxhAAAFkAMAAgAAABQAABCkkAQAAgAAABQAABC4kpEAAgAAAAM4NAAAkpIAAgAAAAM4NAAA6hwABwAACAwAAAiYAAAAABzqAAAACAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAMjAyMjowNjoxMyAxMDo0MDowNgAyMDIyOjA2OjEzIDEwOjQwOjA2AAAAQQBuAG4AaQBlACAAUwB1AHIAbABhAAAA/+ELHmh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8APD94cGFja2V0IGJlZ2luPSfvu78nIGlkPSdXNU0wTXBDZWhpSHpyZVN6TlRjemtjOWQnPz4NCjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iPjxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+PHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9InV1aWQ6ZmFmNWJkZDUtYmEzZC0xMWRhLWFkMzEtZDMzZDc1MTgyZjFiIiB4bWxuczpkYz0iaHR0cDovL3B1cmwub3JnL2RjL2VsZW1lbnRzLzEuMS8iLz48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOnhtcD0iaHR0cDovL25zLmFkb2JlLmNvbS94YXAvMS4wLyI+PHhtcDpDcmVhdGVEYXRlPjIwMjItMDYtMTNUMTA6NDA6MDYuODQxPC94bXA6Q3JlYXRlRGF0ZT48L3JkZjpEZXNjcmlwdGlvbj48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOmRjPSJodHRwOi8vcHVybC5vcmcvZGMvZWxlbWVudHMvMS4xLyI+PGRjOmNyZWF0b3I+PHJkZjpTZXEgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5LzAyLzIyLXJkZi1zeW50YXgtbnMjIj48cmRmOmxpPkFubmllIFN1cmxhPC9yZGY6bGk+PC9yZGY6U2VxPg0KCQkJPC9kYzpjcmVhdG9yPjwvcmRmOkRlc2NyaXB0aW9uPjwvcmRmOlJERj48L3g6eG1wbWV0YT4NCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgPD94cGFja2V0IGVuZD0ndyc/Pv/bAEMABwUFBgUEBwYFBggHBwgKEQsKCQkKFQ8QDBEYFRoZGBUYFxseJyEbHSUdFxgiLiIlKCkrLCsaIC8zLyoyJyorKv/bAEMBBwgICgkKFAsLFCocGBwqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKv/AABEIALYCnAMBIgACEQEDEQH/xAAfAAABBQEBAQEBAQAAAAAAAAAAAQIDBAUGBwgJCgv/xAC1EAACAQMDAgQDBQUEBAAAAX0BAgMABBEFEiExQQYTUWEHInEUMoGRoQgjQrHBFVLR8CQzYnKCCQoWFxgZGiUmJygpKjQ1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoOEhYaHiImKkpOUlZaXmJmaoqOkpaanqKmqsrO0tba3uLm6wsPExcbHyMnK0tPU1dbX2Nna4eLj5OXm5+jp6vHy8/T19vf4+fr/xAAfAQADAQEBAQEBAQEBAAAAAAAAAQIDBAUGBwgJCgv/xAC1EQACAQIEBAMEBwUEBAABAncAAQIDEQQFITEGEkFRB2FxEyIygQgUQpGhscEJIzNS8BVictEKFiQ04SXxFxgZGiYnKCkqNTY3ODk6Q0RFRkdISUpTVFVWV1hZWmNkZWZnaGlqc3R1dnd4eXqCg4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2dri4+Tl5ufo6ery8/T19vf4+fr/2gAMAwEAAhEDEQA/APpGiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACq9/dGy0+e5C7zEhbbnGcVYqnq8Mlxo91DCu+R4yFUdzWVZyVOTjvZl00nNJ7FGHWroXFol7ZJFHd8Rsku45xnkYqe01pL3V3s4Y28tY94lbjdzjgenvSaXodlZJFMtqEuQgDMWLYPfqcD8KUWsw8UG58v9z9kEe//a3E4rmisRHl5n19dLd7Lr5HTL2Lb5V0/H73+Zp1V1DU7PS4Fm1C4S3jZggZ+mTVqsDxZps+pW+npbwecI72KSQZHCg8nmvSpRjKaUnZHFJtJtEsfjHw9KIjHq1ufOfy05PLf060HxJbW95qKahLBbwWTonmbyTlhn5hjj9a5rU/DV9LB4k8iwy93eRSW+No3qMZI54703XfDuqXjeIfJs2kF3cW7Q/MvzqoG7vXfGhh2/i381/d/wA39xk5TV/67/5HZabrmmaw0q6ZeRXLQnEgQ9Kv1z1npk8Hjy9vRb7LSS0jjVxgAsCeMV0NcNWMIyXJtZGkb9QooorIoKKKKACiiigAqjq11fWlp5unWkd0y5LrJLswAOvQ5q9UV0jSWcyIMs0bAD1OKuDSkm1cum0pJtXOcsfFlzLpA1K/04RW8q/6OsMhkeVsn5duOOhrZ0TUjq+i21+0YiM67igOcckdfwqn4csbiz8H21pdRGO4SJlKEgkEk/41L4XtJ7DwxZW13GY5o0IdCc4OT6V111RtPkVmpWWvTX/gHbiFQtPkSTUrLXpr/wAA1qKKK4TzwooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKRmCqWY4AGST2oAWivPx8Q9TbTh4hXRI/wDhGDNsF0bj9+Y9+zztmMbc89c45rv1YMoZeQRkGgBaKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigBshcRsY1DOAdqk4ya5iDxHrcmrT2L6NAJLdFkm23WcKfT5eT7V1NYNvZ3UHi7Vb4wMYZLaNYiCPnYDkV1UHC0udJ6aX9V5nXh3T5Z88U9NL33uvNEdj4lu5tUtIL7TDaw32/wCzuZMv8v8AeXHHBroq4/Q49Um8RNf63pVwJ3ykUhkTy7eP0AznJ7muwp4qMITSilt0d1+pWMhCE0oJba2d1+bCiiiuQ4gooooAKztV1KaxltYreBZpLhygDPtwfyrRrH1zTm1G609TCZYElJmw2MLgfjXPiHUVP93vdfmvJ9PI2oqDqLn21/Iauuzqt1HNYn7VAVAiibeG3dOccVa0zUZLyW4guIBDNbsAyq+4c9OfwqKez/snTJRodmDM5wAGz+JJPak0CCS3t5FntZYZWIaSSVgxlY9Tx6VhTdVVYwm76a6affbf7vTU2kqTpuUV6d+nS+33mtRRRXecZlXvijRdOvHtb3UYYZ0Kho3JyCwyP0p9p4h0i+kuEtNQgla2BM21uEHrn0965nWPD97d33iqaOy8z7ZbQpatlfnIUbgOeORRf6FqSXssmm2Mf/IGW3TeqlTIG+6QeDx68V6CoUHFe9r6rsn+tjPmlc6C38WaFd29xPbalFJHbLulKg/IPXGOlXLbVrG8umt7W5SaVY1kZU5wrDIOenIrhdK0vW7O/vdQuNLurkz2HlLDcSxtufOCuBgKncKO1a3gXRL7w2LjT7y1QpIFmW6jIwWIGUPOeD0oq4ejCMnGWqtbVfP7v62JjOTtdHYUUUV55sFFFFABRWH4uvr3S9AlvtPnWKSErkPGGDAkD8OtSwpqJuFQavHNmMll8hQVz0Ix79jUc2trG6oN01Uura9+lvLzNeiuF/4SPWrTw7e6xPdQTiyvHgaAwhPMVX28HPB5zW34gv8AU7OyF5p81uh+Xy7WRMtcMf4Ac8H6VKqpq/8AWptLBVIyUbrVtfNW8vNG/RWdqKatOqppk0Nqdu5pJU8zJ/u4yPzpvh/UZ9T0dJ7yNY5w7xuF+6SrEEj2OKvm1sc/sn7PnujToooqjIKKQsFGWIA9TXO6TqlxP4nv7ee43QRhvLU4wPmAFYVK0acoxf2nY1hSc4ykuh0dFM86P/non/fQrg/iJ4ovLO4ttG0ecwTXCGWadMZSPOMD3JrSpONOLnLZGLaSuzv6K+edO1XWYNQnlstWu1njJI8yYurkHowJwa9r8KeIk8ReG7XUHCxTONsyE/dccNj2zWk+WFV0r6pJ/eroinUU1obdFM82P/nov/fQo82P/nov/fVI0H0UzzY/+ei/99UebH/z0X/vqgB9FN82P++v51xHjLx1Npd+NJ0FYpLwIHnmk5SAHoMd2PWonONOLlJ2SE2krs7mivJLH4h69pl0kmrtDfWW797sj2PGP7w9celeqw3cFxAk0MqPHIoZWDdQaijXp1481N3QoyUldE1FN82P++v50ebH/fX862KHUVynhTxNdazqmoW975SpbNhCoxn5iP6V1PmJ/fX862rUZUZ8k9zavRnQn7Oe46im+Yn99fzo8xP76/nWJiOopvmJ/fX86PMT++v50AOorkvHXi2Tw/Z29tpmyTUb0kRbuVjUY3OfpkYFefp4i8TwTfaI9cmllzkxyopjb2xjiuSvjKNCSjUerIlUjF2Z7bRWL4V8RReI9AivsCKXJjnjJ+5IvUfTv+NYfj3xdc6Q0GmaKyi+uFMjSsMiKPOM/UmuiVSMIc8noU2krnbUV4nB4m8TWE/2mLV5btgctBcKCj+3A4/CvWdA1221/Q7bUrb5FmXJRjyjA4Kn6EGscPiqWITdN7ExnGWxp0Vj+JvEMPhzQJ9Qcea64SGIH78h4UfnXlT+I/E91N9om1uWCQnIihVQie2COfxoxGKpYe3tHuEpxjue20Vx3gPxbca5Dc2WrbBf2m0l1GBKh6N9eOa6/en95fzreMozipR2ZSaauh1FN3p/eX86N6f3l/OqGOqG8iaexnhQ4aSNlB9CRipd6/3h+dG9f7w/Ok1dWGtGeNprmnH4Ajw59qhOsiD+yTYFx5wuA2zbs6/7XTpzXsFrGYbOGNzlkjVSfcCsOOTSJfHlxD/ZNoNQhtEm/tAxJ5hDErt3Yz29a8+8UeKNQ17XLy2s7ye002zlaBVgfY0zLwzEjnGegrKvXhQg6kyJSUV5I9iorxDSPEuoeFb6G4+2XFxpu8LcW80hfap43KT0xXp3jHxL/wAI5oRuLZUmu5nEVtGx4LHufYDmpoYmnXp+0i9BRmpK6OhorxFvEPid5PPbXplm/uqi+WPbGOlej+CPFL+ItLkW+VI7+1by51X7rejD61FDGUa7cab1QRqRlsdPRSb1H8Q/OvJdf8catq+qXEOjXbWGnwOYleNR5kxHBOT0Ga1rVoUIc83oOUlFXZ63RXkvh7xxq2latbwa1dm/0+4cRGSRR5kLHgHI6jNes7l9R+dFGtCtDng7oIyUldC0Um5fUfnRuX1H51sULRSbh6j86Nw9R+dAC0Um4eo/OjcPUfnQAtFeUeJvG2q6jrVzZ6JdmysbVzE0sagvK44PJ6AHiodC8b6vo2pW6ateNf6bK4jkaVR5kOeAwI6jOM1yPGUFV9jfUj2keblPXaKTcPUfnWJ4wupbTwreTWsrRSoAVdDgjkV304OpNQXV2N6VN1akYLq7G5RWH4PupbvwpZz3UrSyurFnc5J+Y1t5HqKKkHTm4Po7BVpulUlB9HYWijI9aMj1rMzCijI9aMj1oAKKMisXxHq8+kxW7WwRjI5U7hWVarGjB1J7I0p05VJqEd2bVFNjbdGrE8kAmnZrVamYUUZozQAUUZrzn4keIb1NSt9B06d7YPD59zLGcMVJICg9uhrOpUjSg5y2Qm0ldno1FfOGlSXtrezz6bez29yjblcSE7j/ALQPX8a928Ja7/wkfhez1J08uWVSJUHZ1ODj2yK0naNWVK+qt+KuiKdRTWhs0UUUGgUUUUAc746EkvhW4toLa4uZZioWOCFpCcMCc4BxwO9aGn2+n2lobyysvswaP51S1aNyB6pgNn8M1pUVHL7zkdHtn7JUvNvfe9v8jjPCek2l2l4+pabcrML+W4jW7hkRSrNlWAb5Sf1rR8QJaavHdabcaTczXCxkW8ptztLEcFZei4PXJHTvXRUVPs1ychrLFylW9rr5a7GBfajPpNjZ2L29/dymFVmuLW1eXGAATkDqTWjo86XGnKYbO4s4kOxI7iPY+B3weavUVaTvdsxlUjKFra9/+AFFFFUYFXUtMs9YsXs9St1uLdyC0bZwcHI/UVxOm+GdGv8AX7zTbywjls7YN5MJJwmGAGOfQ16BWS+ieVfNeadO1vNIcyAjcr881xYmnOUoTir8rudNCcYxnF9UZ/8Awrrwn/0BLf8ANv8AGuJ8deErfw3e2uqaLYeVYGMxXKwqW8s5yHPXjtXqd1qFvZNGt1Js8zgMRx+J7VP+7mi/hkRh9QRW9SNOtGVJv1OaUG46rRnzdZ3KpeTmMGV5CREiDcXOeAMV6V4H8F+G9Q8NQyajbW97qBLNcfMwaNic7CMjkdK7y20bTbO4M9rYW0MpOS6RAGmf2HYDWV1RIdl0FKsyHAfPqO9dUlSnVlVaabSX3K34mFOm6exk/wDCuvCn/QHh/wC+m/xo/wCFdeFf+gPF/wB9N/jVmXxRDY601jq1tLYxscW91Jjy5fbI6H61ug5GR0pSpyhZtbmyaexzP/CuvCv/AECIv++2/wAaP+FdeFf+gTH/AN9t/jXTUVAzmf8AhXXhb/oEx/8Afbf415prujxeHPGF9Yww+RbzhZrb0ZcYIBPoa9xrN1vQNO8Q2YttVtxKinKMDhkPqCORXNiaCxFJ027XInHmjY8RvpAlo6kbmkGxEAyWY8ACvS9E+HGhR6DZLqWnK92IE85i7cvjnofWr2j/AA90HRr5byGGWedPuPcSF9n0B4/GunrHA4P6rFpu7ZNOnyI5n/hXfhf/AKBi/wDfxv8AGj/hXfhj/oGL/wB/G/xrpqbJIkUbSSMFRRlmPQCu81PK/BvhnSdX1fU4NQtfNjtm2xDcRtG4jsfauv8A+Fd+GP8AoHf+RW/xrS0XTdJt1a+0dVK3Q3NIrE7+f8a1a68ZVVas5pW2/BHbjqyr13NK22/krHMf8K78M/8AQO/8it/jR/wrvwz/ANA//wAiv/jXT0VyHEcx/wAK78M/9A8/9/X/AMaP+Fd+Gv8AnwP/AH+f/GunooA8l8eeEbXw7NY6tpNqy2i7obohi3l5xtbnt1H5Vzj3UEcXmNKuzGc56173LEk0TRTIro4wysMgisODwP4atrsXMOjWqyqdynbnB9cV5mLy6OJqKfNYxnSU3c5XwT4Bsbvw/wDbtdtH+0XkzTqnmMpRDjaCB34z+NY3jjwzb+GNZtr6whaPT7iLyZGLFvLkBJBJPQEH9K9hqK6tYL22e3u4kmhkGHRxkEV2VKMalJ0ntaxo4px5Twaa7hhhLs4PHyqDksfQV33hP4daePDdtJrtpIL6bMsieay7NxyFwD1AxXSWPgzw7pt4t1ZaTbxTqcq4XJU+3pW5XNg8FHC31u2RTpqB5z4z8AWNv4de80K0ka6tJFn2eazF1U/MAD3xz+FcPFdwTQiRJF245ycY+te/Vg3Xgjw3eXjXVzo9s8zHczbcbj6mjGYGOKs72aCpTUzgPAnha28SX19qOoQu1mirDAyuV3t1Y8dR0Fdt/wAK88Pf8+03/gQ/+NdHBBFawJDbxrFEgwqIMACpK66VNUqahHZGkVyqxzH/AArzw9/z7zf+BD/40f8ACvPD3/PCf/wIf/GunorUZzH/AArzw/8A88J//Al/8aP+FeeH/wDnjcf+BL/4109FAHm9v4L0d/iJe2Jim8iPT45VHntncXIPOfauR1jSH8Ja7d2N0rR2ckrS2k7Z2ujc7d3qOhr3MQxiYyiNRIRtL7RuI9M0y6tLa9hMV5BHPH/dkQMP1rnxOHjiKfs5ETipqzPBorKXxLeRaRpamZp3AmkQZWFM5LMe3Suw8deCrTSNHg1TSIZ2+xyhp1MrOfLIwSAfSvRrLTrPTovLsLWG3TuIkC5/KrJAYEEZB4INZ0MHCjRdJa33FGmox5TwIXUBi8wTJsxndu4rqPAfg+31y3u9W1SKdIZ3C22yVoyyr1bA7E12r+BfDEl0bh9FtTITknZxn6VvoixoERQqqMAAYAFYYTL44abne7Jp0lB3OYPw80IqQFux/wBvT/415S1o+h6jdaTer5MsErbA5++hOQwPfg179Wfq2gaVriIurWMN0E+4ZF5X6GujFYaOJp8jdi5wU1Y8Rjs213VLXSLMGWSaVTJsP+rQHJYnt0r1b/hXuh/9Pn/gW/8AjWxpWg6XocbppNjDah/veWuC31NaFPC4aOGp8idwhBQVjl/+Fe6H/wBPv/gXJ/jR/wAK+0T1vf8AwMk/xrqKK6izl/8AhXuif3r7/wADJP8AGj/hXuif3r7/AMDJP8a6iigDl/8AhX2i/wB+/wD/AAMk/wAaP+FfaL/z0v8A/wADZP8AGuoooA8EurBvD+tXmk3QMZSZngZz/rIycgg9/emG2bWby30iyHmz3UiqQp+4uclj6AAV7dquhaZrkSx6tZQ3Sp93zFyV+hpuk+HtJ0MONJsIbUuMMyLy31NeU8tg8R7a/W9vMx9iufmMgfD3RgoHnahx/wBPsn+NZPinwZpuneHLq6tpL5pIwCokundTz3BPNd/UYaG5RgCkqg4YcEAivaoz9nUjO2zTOujP2dSNS2zTOE8L+DNN1Hw3a3VzLfLLICWEd26KOT0APFa//CvtI/576l/4HSf4106IsahY1CqOgUYFLSqz9pUlPu7hWqe1qyqd22cv/wAK+0j/AJ+NS/8AA6T/ABo/4V9pH/Pzqf8A4Hyf411FFZmRy/8Awr/Sf+fnU/8AwPk/xo/4V/pP/P1qn/gfJ/jXUUUAcv8A8K/0r/n61T/wPk/xrH8QeGrPRI4HtJryQysVb7RctIAPbJ4rtdQu5LSFfs9u9xNIdqKo4z6k9hUC6V9stUGslbiUMXAXgJnsK4cXF16cqEN/wOrDyVKaqy2/Ex08A6W0asbvVQSAeL+T/Gl/4V/pf/P5q3/gwk/xrqFAVQB0AwKWu1aI5Wct/wAK/wBL/wCf3Vv/AAYSf40f8K/0v/n+1f8A8GEn+NdTWdrOsLo9sj/Zbi6llbZFDAmS7emeg/GrjFyfLETaSuzGbwDpSKWa/wBWUDkk6jJx+ted+JtOsrPXftXh6W91O3ii2XshLziIg8Hee3t2r1rTF1G/02YeIba3iM5O23jYtsQjG1j3P09avWtnbWVqtvaQRwwqMBEXApVaVKUJU6mt9NHoS05q3Q+c9Okle6ljsoJLm4mOIoo1JLHP8q9X8P8Aw1t7PQraLUL3UUuyu6dba9dEDnkgAHHHTPfFdpDY2ltIXt7WGJz1aOMKT+VT06qhOtKtFauy+5WJpU1TRy3/AAgGnf8AQR1n/wAGMn+NdTRRUmoUUUUAFFFFABRRRQAUUUUAFFFFABRRRQA2SJJoykqK6nqGGQarx2i2Nk8enRqDyyKzHbn/AAq1RUOEW79e5Sk0rdDOs9Uea4FreWsltcY6feVvcNWjRjms27s79LprrTrvlvvW8wyh+npWd50o6+9+dv1LtGctNC7cW0F3CYrqFJoz1V1yKzNcutXsBBcaTZR3tvHn7RAG2yEeqduPStKW6ito4zdyJEXIUZPGfSpgQRkHI9RXTTqJS726GUouxT0rU4dXsEurdJY1YkFJUKspHUEVcqtqFtLdadNb2ty1pLIuFmRQSh9cGs7QrnWfMlstdtlMkKgreRH5Jh9Oxq+RSTlH7uv/AASb2dmbVFAOelFZFBRRRQAUjKGUqwyCMEHvS0UAYnhzSbnRfttoxU2PnF7X5ssqnkgj2NbdY/iaW9tNKF7pztutXEskQ/5axj7w/KtO1uY72ziuYDujlQOp9jXRV5pr20uv5/8ABOmtz1F7eXX8139dyWiiiuc5gooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAKupSXMWmXD2MRluBGfLQHGW7VU8N6WdJ0SKGXmd8yTMepduTUVtqlzfeKbi1tSv2Gzj2zMRktKegB9hW1XRPmpw9m+tn/AJHTNzpQ9k+tn/lf8/mFFFFc5zBRRRQAVnajJfyzLaaenlbhl7luQg9h3NNkiv7zUPnY2tpC+QEPzTEfyFadYO9ZNapfmaq1Np6MZChjhRGdpCoALt1PvT6KiubqCzt2nupUhiUZZ3OAK3iuiMm+pLUbTxJMkLSKJHztQnlsdcCqVjqkGu6fNJpksiLyiTGPHP8AeAPUVV0jwxbabcfbbmWS/wBRI+a7nOW56hR0UfSteRRvzuzXQm7exFqFrr+q6lJbJPHpmmIR++iO+af6dlFbyJsjRSS20Y3N1NOopSqOSStZIaVtQooorMYUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAR3FtDdRGK5iWVD1VhmoLLT49OjdLdpChOVR3JC+wq3RUOnFy57a9yuaSXLfQzrXWIprgW1zG9rcnpHIPvfQ960aaUVmVmUEqcqSOlUb0anHP51k8cseMG3cYP1DetZ3nTjefvei1/r0+4u0Zy93Qz4tF1LSdU83SL3zrGeTM9pdMW2Z6sjdR9DW39qgN0bbzk88IHMe75tpzzj04NEMxe3SSdPJZhyrHofSs7WvD1prOyVi9teRf6m7hO2SM/XuPY12KpGq05v52/P8Aq5g4uOxrUVm3GoRaDpMMmsXTPt2xyT7OpPcgdBV+KWOeJZYXWSNxlWU5BH1qHFpX6DuPoooqRiMoZSrAEEYIPes+LVoTrsmkmJopI4hIhOMOvt9K0az73SIrzVLK/wDMaKa0JwU/jUjlT7VpT5NVPt+JrT5NVPt+PT/I0KKRHWRQyMGU9CDkGlrMyCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAqpqep2+k2LXV4xEYYLhRkkk4AA71brK1TRjqmp6fNNMPs1o5laDH33x8p/CtKSg5++9DWioOa9o9P60+ZftreCFWe3hWLzj5j4XGSe596mooqG23dmbbbuwooqvdNcPak6eYzIeAznIHvUSlyq4JXdiZ2IRtgDMBwM9az7Gyu3uftupTHzcYSCNsJGP6mp9PsBYxuWleaaU7pZXPLH+gq3WXJ7S0p6W6F83JdR+8KKinmMdvLJEhmeNSRGp5Y4zj61j6UviC7vhe6rJFZ2wUhLGMBic92b1+ldcYXi5XtYyb1sJqfiRob5tO0ixlv75fvKBtjjz3ZjV+70mz1aO2bVbRJXhO9UYkqrY9Oh/GrwUAkgDJ6mlpuaVuRWa69RW7iIixqFRQqgYAAwBS0UVkUFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQBDdWkN7btBcpvjbqKhsLF7Euv2qSaHA2JJyU/HvVyis3Tg5qdtS1OSjy9DPTU7G9mkspxtk5VoZ1xvH0PUUzTdFtNBhuRpkcojkYyC38wlVOOig8DNXZ7S3uWQ3EKSFDlSw5BqG+kvoWSSzhSeMA74ycMfoaI1KtKDU3deX6objCcly6epU0fxJZ6vK9th7W+i/wBZaTjbIvuPUe4rXqhbxWmovBqMlnsuI8qjSph07EVQ1PX7jRNRLalZsdLcgJdw5YxHvvHpnuK6YqNZ3ordf1b+rmLvDSZvUjKroVcBlYYII4IoVgyhhnBGRkYpayKMPw1aXelrdadPG32WCUm1kJzlDzj8K3KyfEl/eaXpYvbJFkWCRWnQjJMefmxUieIdLlvrazivI5Li5G6ONDk42lufTgd66ZxqVf3tt+3lvf8AM6qkKtb99a973t5bt/maVFFFcxyhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFZuua7a6BaxXN8H8uSUR5QZIyCc4/CqhCU5KMVdsuEJVJKMFdssandtY6XcXMcTzPFGWWNFJLHsMCqvhyzuLPRYheyO9zKTLLuJO1m5wPTFV9O8SRaxrTW2mbJ7SOASSTgkYYnhcVuVtNSpR9nJWb1/yNqinRh7KSs3r5+XoFNd0jQvIwVR1LHAFNluYYHjWaVUaRtqBj94+1UrnSjfXwe8mMlqmCluBgE+p9a4pzklaCu/63MYxTfvOyJNRtbi9SOKC48mFj+9K/eYegParFraw2dskFumyNBwKdJJFbQNJKyxRRrksxwFArH0zxKms6kY9LtJprFAQ98RtjLDsufvfWtIYe7dVL5/p/W5MqmigXtT1ex0a18/UbhIU6KCeWPoB1JqHVba+1K0iTS9Q+xJIcySqmWK4/hPY+9Ivh3Tv7ak1WaIz3TEbGlYsIuP4QenrWpW/NCFnDfz/AMiLN7mfo+i2uiWhgs/MYu2+SSVyzSN/eJPetCiis5ScnzSeo0klZBRRRUjCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAjuIFubd4XLKrjBKnBH41Us7S7tpGinuVubbHy+Yvzj2PrV+is3Ti5KfVFqbUeXoYWsXF/PIP8AhHr23+1WrHzbSbgS8dM9vrV7R9Qn1HTxNeWE1hMCVeKXHBHUg9x71JeaZa3pDTR4kH3ZF4YfjVTxEzrpEieVI1uyFZmhk2Oi46g1XtnGDVVKy6q97eaDkUpLk3fRmL4/t9cbTGl0udjaBSLiGNfmx3Oe49a878KTGDxbpjjjNwq/99HH9a7TQ/Es9hciO+nlubdsKWk5YDsf8aZr/hSNb6HXfDu2SOOZZJoYjnGCDuX+or1smzjD4ilPD7bq+33+v/APqMHW+r0nha6S5lo/VbPzPQ6KMjOM80V5Z8oFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFZ2tXmqWVmsmi6UmpzlsGJ7oQAD13EH8q0a5T4hS+Jv7BS28IabLeT3Mnl3EkNxFFJBFj5mXzGUFj0HPGc0mNDPCPjHUfEms6nY3mgrYx6cRHJcxXouI2lPJjBCjJA64zjp1rrq5PwVJf2lrHpMng660Gyt4vklmvLebe2ec+W7EsSSSTXWVTEFFFFIAooooAKKKKACiiigAooooAKKKKACuB+KsuNP0+L+9KzH8AP8a76uF+IGmvf3dpJK2y1hjJJ7uxPCj8vyrrwdalh6yrVXaMbt/cellbjHFwlLZX/I5TwTompapqnnWNxJZwQkebOhwf90ep/SvXYru3a4azScPPGoLDOT+NeZ6RfNpMhW3Zo4nXa4Q4OPUe9d9plxpNuYrWzuIzPcKZApbMjjuTXBLOZZtWvTjZR++39b9EdmcTdSpzz26f8Fk9rpUcN093cObi4JOJH/gHoB2qPWtSvNPhiGnadLfXE7bECkBEOM5c9hVPUNG1PV9SdbzUfs+mKRsgtcq8vH8bduewreRQkaoM4UYGTW0KdKgly691r+LPBlOU3qUNKg1H7A665LDPNKxJWJcKikfd56/Wr0caRRrHEioijCqowAKdRRKTk7iSsFFFFSMKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAKd/q+n6Xs/tG7itt/C+Y2M0ltrOnXl0ba1vIZJ1XeY1b5seuKyPHAB0ezyM/8TC3/wDQxRr2inU9WWazk+z6ha24ktpx2bceD6qRwRWLnJSaX9aXO6nRpSpxlJtXvr00t/nqbEur6fBfCzlu40uSMiIn5setVj4n0UMqnUoNzZ2jd1x6Vn6TqqazrFt58QivILWeK6gPWNt8X6HqKfqahPHXh1VACiC7AA7fKlPmdk11KWHgp8k072b3XRN9utjWbV9PSzS6a7iEEhwj7uG+lWIJ4rmFZreRZI2GVZTkGuf8Qxzwajp82irG9/EJWS1cYWVDjec9jnGD7471c8LyW8uihrcOrGaQzpIu1klLEuuO2CenpinGTcmmZToxVFVY9f8Ag/5b9fkbFFFFaHIFUdW046pYm2E5hVjliFzn2q9RUThGpFwlsyoycJKS3Rxc3gW4H+ovY3/30K/yzUWn+GNZ0nVEuYLtYIdwM+xgyOnfKnHOO9dzVXU7CPVNNmsppJY45l2u0TbWx9a5cPgaNCqqlNuP46ejvc6auMq1Yck7M8i134h3R8bR32muTZ2ZMaR5wJV/iz9cfoK9b0nVLbWdMhvrJ90Uq5Hqp7g+9ed6h8G1OW0rVSPRLmPP/jy/4VJ4Y0jxX4JnmjNiupWMoJ8uGYcPjgjPI9Dx/KvqsVHB1qK9hJKUV6X/AOCeVB1Iy95aM9Mrk/EfxB07QL42EUE2oXyjLwwYAj9NzHgfTmnHxF4nKnb4OfOOM6gn/wATXlFm8k3nz3OftMs7tPuOTu3Hg/SvlcdiXhqXOld7G1SfJG56x4b8e6d4hu/sTQTWN7jcsM4HzjvtYcH9DXSzzxWtvJPcSLHFGpZ3Y4CgdTXgs0s1vdWU9kpa6juYzCoOCx3fdz79K6/x7rPiG58KtDqHh86dayTRrLOL1ZPl3dNoA4NLB4p4ij7SS1QU580bmhP8WbAXJFjpV9d2ynBuF2ru91UnJH1xXW6F4g0/xHp/2zTJS6A7XVl2sjehHY14woCqAowAMAVr+Cb/AFaw8Raguh6WNSWSFGmiNyIQjZOGyQcmuTBZjLEVXCUbGdOq5ysz2KsbxJ4p07wvZpNqDO8kpxDBEu55D7D+prP/ALd8Xf8AQmx/+DZP/iK888UXeo3/AI2d9bsRYTR2yCK3E4mCgk5IYADmvRxNb2FKVS17G05csbnYWPxW06W6WPU9Ou9OjdsLO+HQem7HI/Wu7VldAyEFWGQR3FeCTqj28iyYKFTnPpXX6RrXi1Ph3H9n0JJYltGEd418FcqAcNs25yB2zXJl+MlilLmVmjOlUc9zY1j4m6fp9/LZ2Flc6lJC22R4iFjU9xuPU/hWt4Z8Y6d4o82O1EtvdQjMltOAHUeoxwR7ivHdPC/YIShzuXcT6k9TV3S7i9s/FulzaPAs96zsnkmTYJE28gt2Fc9DMpVMT7Jx0/EmNZufLY90qOeeO2t5J53CRRqWdj0AHU1zP9r+M/8AoV7P/wAGQ/8Aiaoa7qni5/D9+tx4ctIojbuHcagGKjacnG3mvbOg7W3uIrq2juLdxJFKgdHHRlIyDUlcJ4e1LxcnhnTEtfD9nLCtpEI5Gv8AaWXaMEjbxx2rR/tTxr/0Llh/4MP/ALGgDqq5vxL4403w3MlrKk13eONwt7cAkD1JPAFQ/wBp+Nv+he0//wAD/wD7GvMbuW7ufEeqzarGsd99oKyRq24IAOAD3GK5MZiHh6LmldkVJckbno2g/EjTtX1GKwvLWfTbqY4iExBSQ+gYd/rXXXNzDaWslxcyLFDEpd3Y4CgdTXgGqELp8kmdrx4eNh1Dg5GPxrq/G2oeKZvBlums2NrbWc0sKzSw3BZ2B5wRjoT1rHBYx4ik5yWqJp1OaN2bMvxasftH+h6TfXNqDgzjaufdVJyR+Vddoeu2HiLTVvdMlLxk7WDDDI3oR2NeLqAqgKMAdAK1fA9zrlt4h1OLw5b21yrRRvNHcSlFVskAjHeufBZjLEVXCSsRTqucrM9krlPEfxA07QL1rCKCbUL1Rl4YMAR56bmPA+nJoN7462nGk6RnH/Py/wDhXlVm8swnmujm6knczknnfuOa68diXhqXPFXexpUnyRuer+G/H2neIbz7C0M1jfbdywTgfOO+1hwf0NdTXgEzzw3VlNY4+2Jcp5GTjLE4x9D3r1T7V48P/MO0Uf8AbeSngsS8TS52rMKc+eNzY17xDp/hzT/tepyFUJ2oiDc0jegHeuTg+LNh9oxf6VfWdsTgTsFcL7sAcj8M1zHjSfW5/FFjH4jitYikDPAls7MhOcE896zGAZSGAKkYINceMzGWHrKEY37mdSq4Ssj3aCeK5t457eRZIpFDI6nIYHoRXAfEG/axUXF5Kij7trbA5Z/Vj6CqPgK48XHwrGmjRaZNYxzSJC128gfaG9uMV1l34I0rV777frSTXVwyjKvKQieygY4r2I0cPiLLE35N7Lr5HVGrUpe9S3/I8Tm8QX8zgrIIgDnCDFe0+B7LSDoVvqGnQfvpk/eyyfNJu7jJ960bTwtodjj7LpdqhHfywT+ZrUjjSJdsSKi+ijFd9aphFT9nhaXJ6dfXv95lerOXNVlcdRRRXCWFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAZWuaCmuxRRT3t1bxxSLIFtygyynIJ3KTxT4tIkj1Rb5tUvZGWMRmJxFsYe4CA5zzwR+XFaVFTyK9zZVpqPJ09F1Kg0y1XVzqSR7bpovJZh/EuQefpiqd54fW81221VtRvY5bUERRxmPYoYAMMFCTnHc/TFa9FHKmKNapF3T6W+RRv9Kjvp4bhZ5rW5gDCOeAruAbGRhgQQcDqO1S2NjFp9uYoSzFnMju5yzserH3qzRTsk7kupJx5L6BRRRTICiiigAoorK8Q69F4f00XD21xdzSP5UFvbRl3lcgkD2HByTwKANWiuP0Txhf3/wAMz4kudNMl6EkYWVsCSzKxUIOvoMmqqeIfFWi6toqeJ49LltNYuBaqlkrrLbSFCwzuJDj5SCQB9KOtg6XO6rgPEnw4lvNVm1Lw9dx20twd09vOpMbN/eGOQT3qHUfGniJm8QX+lrpMGnaBM0U1veb/AD59qhiQQQEBz8uQc13emXv9o6Va3ojaL7REsuxuq5GcGonTjVjyyV0KST0Zxvhj4dy6fqkep69dx3U8HMEECkRof7xzyTXZanpttq+mT2F8m+CdCrjv9R71aoohTjTjywVkCSSsjyyb4X61bz+Tp2q2slpnCPco3mIPQ44Ndr4T8KW/hbT3iSZrm6nbfcXDgAu3sOw9q3qKinQpU5OUIpNiUUndIK5fxh4Li8TpDcW9wbPUbcYimC5Vh3Vh3FdRRWsoqSs9itzy60+FuqXdwqa5qVulmG+dLNW3yj0y3SvTIbaG3tEtoY1WGNAipjgKBjFS0VnTpU6S5YKwlFR2PNdV+GN5HeySeHL23S2lYv8AZ7sE+WSedpXt7Gtvwh4ETw/dvqWo3AvNSdPLDKMRwr6KOuT3J/xz19FKNClGbqKOrFypO9gqG9tI76xntJ9wjnjMb7Tg4IwcVNRWxRXsLOPTtOt7KAsYreJYkLHJwowM/lViiigArivFfgBtZ1I6ppF2tpeuoWVJVzHKB0JxyD712tFTOEZx5ZK6E0mrM850T4ZXH9pQ3fiW7hnjt3EkdrbAhGYdC5PJHtXdarpdrrOlT6ffJvt502MB1HoR6EdRVyipp04U48sFZAkkrI8sl+F+twz+TY6raPa5wslwjeYo98cE123hXwra+FtPeGB3nuJm33FxJjdI39AOwrdoqadClTblCNmxKKWyCuA8SfDiW81SXUvD93HbS3B3T286kxs394Y5BPeu/oq6lOFSPLNXQ2k1ZnB+GPh1Jp+qx6nr13HdTwcwQQKRHGf7xzyTXeUVR1jVoNF0uS9uUmkVMARwRl3djwFAHUmiEIU48sVZAklojN8WeErXxTZRrJK1td25LW9wgGVPoR3B9K42H4X6zcTeTqWq2sdpnDtao3mOvtu4FdL4N8W3viDw/qupapp5tJbG+uIBaxjc4WPoD1y30rHk8W+L9N0rT/EOtWWmw6XeTQo+nKsgurdZWCrlicMw3DK7R35qJ4elVkpTim9Px2FKKerW36Hd6Zp1tpGmQWFjH5cFugRF9h3PqatVwvjLUvGWgwX+rW2qeH7fSbddyJc2c0kx4AC5WQAsW4AA7iui8KT63c+GLOfxTFbQ6pKm+aK2UqiZ6DBJOQOvPWtlqVsbFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAGV4ii16bSwvha5sba+8wHffRNJHs5yMKQc9KdpEerxaGq+I57S41ABvMks42SM+mAxJ/WtOigDzzwl4gsvC3wbbWtTbbbWZnd8dT+9YAD3JIH41jeD/ABl4X8UeK7PVdb8R2V3rcmY9M0yEsY7MMOccfNIRwWPToK9cooWjDpY8O1CDwrPda4/j9Jh4tNzILLIkEpjB/cfZwvBHTp3zmvXPC7ai3hTTG1sEagbZPtAIwd+Oc+9atFC0VgeruFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAcJ4GvrfTNG8W317IIre212/llc/wqrZJ/IVyWkfEfwn4z8QWep+IvENpBDDOG0vRvmJEmcLJLxgvzwvQZ9a9oooWlvJL8Byd7+dzivFa/2x4+8M6FIN1tE0mpzoRw5iwIwfo7A/hXa0UUdLCCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAP/Z" - } - }, - "cell_type": "markdown", - "id": "bf3d4397", - "metadata": {}, - "source": [ - "As the name suggests, PTQ is performed on a trained model that has achieved acceptable accuracy. It is effective and also quick to implement because it does not require any retraining of the network. Now that we have the trained checkpoint ready, let's start quantizing the model. \n", - "\n", - "To perform PTQ, we perform inference in FP32 on calibration data, a subset of training or validation data, to determine the range of representable FP32 values to be quantized. This gives us the scale that can be used to map the values to the quantized range. We call this process of choosing the input range \"Calibration\". The three popular techniques used to calibrate are:\n", - "\n", - "- Min-Max: Use the minimum and maximum of the FP32 values seen during calibration. The disadvantage with this method is that, if there is an outlier, our mapping can induce a larger rounding error. \n", - "\n", - "- Entropy: Not all values in the FP32 tensor may be equally important. Hence using cross entropy with different range values [T1, T2], we try to minimize the information loss between the original FP32 tensor and quantized tensor. \n", - "\n", - "- Percentile: Use the percentile of the distribution of absolute values seen during calibration. Say, at 99% calibration, we clip 1% of the largest magnitude values, and determine [P1, P2] as the representable range to be quantized\n", - "\n", - "\n", - "![img4.JPG](attachment:img4.JPG)\n", - "\n", - "\n", - "We will be using the Pytorch Quantization toolkit, a toolkit built for training and evaluating PyTorch Models with simulated quantization. \n", - "\n", - "`quant_modules.initialize()` will ensure quantized modules are called instead of original modules. For example, when you define a model with convolution, linear snd pooling layers, you will make a call to `QuantConv2d`, `QuantLinear` and `QuantPooling` respectively. `QuantConv2d` basically wraps quantizer nodes around inputs and weights of regular `Conv2d`. Please refer to all the quantized modules in pytorch-quantization toolkit for more information. " - ] - }, - { - "cell_type": "code", - "execution_count": 27, - "id": "f1520afc", - "metadata": {}, - "outputs": [], - "source": [ - "quant_modules.initialize()" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "id": "ee09402f", - "metadata": {}, - "outputs": [], - "source": [ - "# We define Mobilenetv2 again just like we did above\n", - "# All the regular conv, FC layers will be converted to their quantized counterparts due to quant_modules.initialize()\n", - "feature_extract = True\n", - "q_model = models.mobilenet_v2(pretrained=True)\n", - "set_parameter_requires_grad(q_model, feature_extract)\n", - "q_model.classifier[1] = nn.Linear(1280, 10)\n", - "q_model = q_model.cuda()\n", - "\n", - "# mobilenetv2_base_ckpt is the checkpoint generated from Step 2 : Training a baseline Mobilenetv2 model.\n", - "ckpt = torch.load(\"./models/mobilenetv2_base_ckpt\")\n", - "modified_state_dict={}\n", - "for key, val in ckpt[\"model_state_dict\"].items():\n", - " # Remove 'module.' from the key names\n", - " if key.startswith('module'):\n", - " modified_state_dict[key[7:]] = val\n", - " else:\n", - " modified_state_dict[key] = val\n", - "\n", - "# Load the pre-trained checkpoint\n", - "q_model.load_state_dict(modified_state_dict)\n", - "optimizer.load_state_dict(ckpt[\"opt_state_dict\"])" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "id": "b8726956", - "metadata": {}, - "outputs": [], - "source": [ - "def compute_amax(model, **kwargs):\n", - " # Load calib result\n", - " for name, module in model.named_modules():\n", - " if isinstance(module, quant_nn.TensorQuantizer):\n", - " if module._calibrator is not None:\n", - " if isinstance(module._calibrator, calib.MaxCalibrator):\n", - " module.load_calib_amax()\n", - " else:\n", - " module.load_calib_amax(**kwargs)\n", - " model.cuda()\n", - "\n", - "def collect_stats(model, data_loader, num_batches):\n", - " \"\"\"Feed data to the network and collect statistics\"\"\"\n", - " # Enable calibrators\n", - " for name, module in model.named_modules():\n", - " if isinstance(module, quant_nn.TensorQuantizer):\n", - " if module._calibrator is not None:\n", - " module.disable_quant()\n", - " module.enable_calib()\n", - " else:\n", - " module.disable()\n", - "\n", - " # Feed data to the network for collecting stats\n", - " for i, (image, _) in tqdm(enumerate(data_loader), total=num_batches):\n", - " model(image.cuda())\n", - " if i >= num_batches:\n", - " break\n", - "\n", - " # Disable calibrators\n", - " for name, module in model.named_modules():\n", - " if isinstance(module, quant_nn.TensorQuantizer):\n", - " if module._calibrator is not None:\n", - " module.enable_quant()\n", - " module.disable_calib()\n", - " else:\n", - " module.enable()" - ] - }, - { - "cell_type": "code", - "execution_count": 28, - "id": "da627181", - "metadata": {}, - "outputs": [], - "source": [ - "# Calibrate the model using max calibration technique.\n", - "with torch.no_grad():\n", - " collect_stats(q_model, train_dataloader, num_batches=16)\n", - " compute_amax(q_model, method=\"max\")" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "id": "73e6d51c", - "metadata": {}, - "outputs": [], - "source": [ - "# Save the PTQ model\n", - "torch.save(q_model.state_dict(), \"./models/mobilenetv2_ptq.pth\")" - ] - }, - { - "cell_type": "code", - "execution_count": 18, - "id": "c7dadbf2", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mobilenetv2 PTQ accuracy: 68.11%\n" - ] - } - ], - "source": [ - "# Evaluate the PTQ Model \n", - "test_acc = evaluate(q_model, val_dataloader, criterion, 0)\n", - "print(\"Mobilenetv2 PTQ accuracy: {:.2f}%\".format(100 * test_acc))" - ] - }, - { - "cell_type": "markdown", - "id": "efd5ff11", - "metadata": {}, - "source": [ - "Let us now prepare this model to export into ONNX. Setting `quant_nn.TensorQuantizer.use_fb_fake_quant = True` enables the quantized model to use `torch.fake_quantize_per_tensor_affine` and `torch.fake_quantize_per_channel_affine` operators instead of `tensor_quant` function to export quantization operators. " - ] - }, - { - "cell_type": "code", - "execution_count": 19, - "id": "3f10f707", - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "W0725 16:43:50.537823 139848660895552 tensor_quantizer.py:280] Use Pytorch's native experimental fake quantization.\n", - "/opt/conda/lib/python3.8/site-packages/pytorch_quantization/nn/modules/tensor_quantizer.py:283: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!\n", - " if amax.numel() == 1:\n", - "/opt/conda/lib/python3.8/site-packages/pytorch_quantization/nn/modules/tensor_quantizer.py:285: TracerWarning: Converting a tensor to a Python number might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!\n", - " inputs, amax.item() / bound, 0,\n", - "/opt/conda/lib/python3.8/site-packages/pytorch_quantization/nn/modules/tensor_quantizer.py:291: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!\n", - " quant_dim = list(amax.shape).index(list(amax_sequeeze.shape)[0])\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "&&&& RUNNING TensorRT.trtexec [TensorRT v8205] # trtexec --onnx=models/mobilenetv2_ptq.onnx --int8 --saveEngine=models/mobilenetv2_ptq.trt\n", - "[07/25/2022-16:43:56] [I] === Model Options ===\n", - "[07/25/2022-16:43:56] [I] Format: ONNX\n", - "[07/25/2022-16:43:56] [I] Model: models/mobilenetv2_ptq.onnx\n", - "[07/25/2022-16:43:56] [I] Output:\n", - "[07/25/2022-16:43:56] [I] === Build Options ===\n", - "[07/25/2022-16:43:56] [I] Max batch: explicit batch\n", - "[07/25/2022-16:43:56] [I] Workspace: 16 MiB\n", - "[07/25/2022-16:43:56] [I] minTiming: 1\n", - "[07/25/2022-16:43:56] [I] avgTiming: 8\n", - "[07/25/2022-16:43:56] [I] Precision: FP32+INT8\n", - "[07/25/2022-16:43:56] [I] Calibration: Dynamic\n", - "[07/25/2022-16:43:56] [I] Refit: Disabled\n", - "[07/25/2022-16:43:56] [I] Sparsity: Disabled\n", - "[07/25/2022-16:43:56] [I] Safe mode: Disabled\n", - "[07/25/2022-16:43:56] [I] DirectIO mode: Disabled\n", - "[07/25/2022-16:43:56] [I] Restricted mode: Disabled\n", - "[07/25/2022-16:43:56] [I] Save engine: models/mobilenetv2_ptq.trt\n", - "[07/25/2022-16:43:56] [I] Load engine: \n", - "[07/25/2022-16:43:56] [I] Profiling verbosity: 0\n", - "[07/25/2022-16:43:56] [I] Tactic sources: Using default tactic sources\n", - "[07/25/2022-16:43:56] [I] timingCacheMode: local\n", - "[07/25/2022-16:43:56] [I] timingCacheFile: \n", - "[07/25/2022-16:43:56] [I] Input(s)s format: fp32:CHW\n", - "[07/25/2022-16:43:56] [I] Output(s)s format: fp32:CHW\n", - "[07/25/2022-16:43:56] [I] Input build shapes: model\n", - "[07/25/2022-16:43:56] [I] Input calibration shapes: model\n", - "[07/25/2022-16:43:56] [I] === System Options ===\n", - "[07/25/2022-16:43:56] [I] Device: 0\n", - "[07/25/2022-16:43:56] [I] DLACore: \n", - "[07/25/2022-16:43:56] [I] Plugins:\n", - "[07/25/2022-16:43:56] [I] === Inference Options ===\n", - "[07/25/2022-16:43:56] [I] Batch: Explicit\n", - "[07/25/2022-16:43:56] [I] Input inference shapes: model\n", - "[07/25/2022-16:43:56] [I] Iterations: 10\n", - "[07/25/2022-16:43:56] [I] Duration: 3s (+ 200ms warm up)\n", - "[07/25/2022-16:43:56] [I] Sleep time: 0ms\n", - "[07/25/2022-16:43:56] [I] Idle time: 0ms\n", - "[07/25/2022-16:43:56] [I] Streams: 1\n", - "[07/25/2022-16:43:56] [I] ExposeDMA: Disabled\n", - "[07/25/2022-16:43:56] [I] Data transfers: Enabled\n", - "[07/25/2022-16:43:56] [I] Spin-wait: Disabled\n", - "[07/25/2022-16:43:56] [I] Multithreading: Disabled\n", - "[07/25/2022-16:43:56] [I] CUDA Graph: Disabled\n", - "[07/25/2022-16:43:56] [I] Separate profiling: Disabled\n", - "[07/25/2022-16:43:56] [I] Time Deserialize: Disabled\n", - "[07/25/2022-16:43:56] [I] Time Refit: Disabled\n", - "[07/25/2022-16:43:56] [I] Skip inference: Disabled\n", - "[07/25/2022-16:43:56] [I] Inputs:\n", - "[07/25/2022-16:43:56] [I] === Reporting Options ===\n", - "[07/25/2022-16:43:56] [I] Verbose: Disabled\n", - "[07/25/2022-16:43:56] [I] Averages: 10 inferences\n", - "[07/25/2022-16:43:56] [I] Percentile: 99\n", - "[07/25/2022-16:43:56] [I] Dump refittable layers:Disabled\n", - "[07/25/2022-16:43:56] [I] Dump output: Disabled\n", - "[07/25/2022-16:43:56] [I] Profile: Disabled\n", - "[07/25/2022-16:43:56] [I] Export timing to JSON file: \n", - "[07/25/2022-16:43:56] [I] Export output to JSON file: \n", - "[07/25/2022-16:43:56] [I] Export profile to JSON file: \n", - "[07/25/2022-16:43:56] [I] \n", - "[07/25/2022-16:43:56] [I] === Device Information ===\n", - "[07/25/2022-16:43:56] [I] Selected Device: NVIDIA Graphics Device\n", - "[07/25/2022-16:43:56] [I] Compute Capability: 8.0\n", - "[07/25/2022-16:43:56] [I] SMs: 124\n", - "[07/25/2022-16:43:56] [I] Compute Clock Rate: 1.005 GHz\n", - "[07/25/2022-16:43:56] [I] Device Global Memory: 47681 MiB\n", - "[07/25/2022-16:43:56] [I] Shared Memory per SM: 164 KiB\n", - "[07/25/2022-16:43:56] [I] Memory Bus Width: 6144 bits (ECC enabled)\n", - "[07/25/2022-16:43:56] [I] Memory Clock Rate: 1.215 GHz\n", - "[07/25/2022-16:43:56] [I] \n", - "[07/25/2022-16:43:56] [I] TensorRT version: 8.2.5\n", - "[07/25/2022-16:43:57] [I] [TRT] [MemUsageChange] Init CUDA: CPU +440, GPU +0, now: CPU 452, GPU 5862 (MiB)\n", - "[07/25/2022-16:43:57] [I] [TRT] [MemUsageSnapshot] Begin constructing builder kernel library: CPU 452 MiB, GPU 5862 MiB\n", - "[07/25/2022-16:43:57] [I] [TRT] [MemUsageSnapshot] End constructing builder kernel library: CPU 669 MiB, GPU 5934 MiB\n", - "[07/25/2022-16:43:57] [I] Start parsing network model\n", - "[07/25/2022-16:43:57] [I] [TRT] ----------------------------------------------------------------\n", - "[07/25/2022-16:43:57] [I] [TRT] Input filename: models/mobilenetv2_ptq.onnx\n", - "[07/25/2022-16:43:57] [I] [TRT] ONNX IR version: 0.0.7\n", - "[07/25/2022-16:43:57] [I] [TRT] Opset version: 13\n", - "[07/25/2022-16:43:57] [I] [TRT] Producer name: pytorch\n", - "[07/25/2022-16:43:57] [I] [TRT] Producer version: 1.13.0\n", - "[07/25/2022-16:43:57] [I] [TRT] Domain: \n", - "[07/25/2022-16:43:57] [I] [TRT] Model version: 0\n", - "[07/25/2022-16:43:57] [I] [TRT] Doc string: \n", - "[07/25/2022-16:43:57] [I] [TRT] ----------------------------------------------------------------\n", - "[07/25/2022-16:43:57] [W] [TRT] parsers/onnx/onnx2trt_utils.cpp:506: Your ONNX model has been generated with double-typed weights, while TensorRT does not natively support double. Attempting to cast down to float.\n", - "[07/25/2022-16:43:57] [W] [TRT] parsers/onnx/onnx2trt_utils.cpp:368: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.\n", - "[07/25/2022-16:43:57] [I] Finish parsing network model\n", - "[07/25/2022-16:43:57] [I] FP32 and INT8 precisions have been specified - more performance might be enabled by additionally specifying --fp16 or --best\n", - "[07/25/2022-16:43:58] [W] [TRT] Calibrator won't be used in explicit precision mode. Use quantization aware training to generate network with Quantize/Dequantize nodes.\n", - "[07/25/2022-16:43:59] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +838, GPU +362, now: CPU 1543, GPU 6342 (MiB)\n", - "[07/25/2022-16:43:59] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +128, GPU +58, now: CPU 1671, GPU 6400 (MiB)\n", - "[07/25/2022-16:43:59] [I] [TRT] Local timing cache in use. Profiling results in this builder pass will not be stored.\n", - "[07/25/2022-16:44:20] [I] [TRT] Detected 1 inputs and 1 output network tensors.\n", - "[07/25/2022-16:44:21] [I] [TRT] Total Host Persistent Memory: 75056\n", - "[07/25/2022-16:44:21] [I] [TRT] Total Device Persistent Memory: 2367488\n", - "[07/25/2022-16:44:21] [I] [TRT] Total Scratch Memory: 0\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageStats] Peak memory usage of TRT CPU/GPU memory allocators: CPU 11 MiB, GPU 184 MiB\n", - "[07/25/2022-16:44:21] [I] [TRT] [BlockAssignment] Algorithm ShiftNTopDown took 3.69334ms to assign 4 blocks to 87 nodes requiring 131661824 bytes.\n", - "[07/25/2022-16:44:21] [I] [TRT] Total Activation Memory: 131661824\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 1674, GPU 6412 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +0, GPU +10, now: CPU 1674, GPU 6422 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in building engine: CPU +2, GPU +4, now: CPU 2, GPU 4 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init CUDA: CPU +0, GPU +0, now: CPU 1665, GPU 6384 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] Loaded engine size: 2 MiB\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +10, now: CPU 1666, GPU 6398 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +0, GPU +8, now: CPU 1666, GPU 6406 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +2, now: CPU 0, GPU 2 (MiB)\n", - "[07/25/2022-16:44:21] [I] Engine built in 24.535 sec.\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +10, now: CPU 1435, GPU 6312 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +1, GPU +8, now: CPU 1436, GPU 6320 (MiB)\n", - "[07/25/2022-16:44:21] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +128, now: CPU 0, GPU 130 (MiB)\n", - "[07/25/2022-16:44:21] [I] Using random values for input inputs.1\n", - "[07/25/2022-16:44:21] [I] Created input binding for inputs.1 with dimensions 64x3x224x224\n", - "[07/25/2022-16:44:21] [I] Using random values for output 1225\n", - "[07/25/2022-16:44:21] [I] Created output binding for 1225 with dimensions 64x10\n", - "[07/25/2022-16:44:21] [I] Starting inference\n", - "[07/25/2022-16:44:24] [I] Warmup completed 64 queries over 200 ms\n", - "[07/25/2022-16:44:24] [I] Timing trace has 967 queries over 3.00851 s\n", - "[07/25/2022-16:44:24] [I] \n", - "[07/25/2022-16:44:24] [I] === Trace details ===\n", - "[07/25/2022-16:44:24] [I] Trace averages of 10 runs:\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.62079 ms - Host latency: 4.67811 ms (end to end 4.69463 ms, enqueue 1.64643 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58894 ms - Host latency: 4.64457 ms (end to end 4.66023 ms, enqueue 1.64765 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58884 ms - Host latency: 4.68113 ms (end to end 4.69763 ms, enqueue 1.4498 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59182 ms - Host latency: 4.74732 ms (end to end 4.76547 ms, enqueue 1.01564 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.57819 ms - Host latency: 4.72507 ms (end to end 4.74366 ms, enqueue 1.02484 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.57656 ms - Host latency: 4.72242 ms (end to end 4.74165 ms, enqueue 1.02861 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.57644 ms - Host latency: 4.71519 ms (end to end 4.7332 ms, enqueue 1.01613 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58525 ms - Host latency: 4.71598 ms (end to end 4.73434 ms, enqueue 1.02659 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58402 ms - Host latency: 4.73148 ms (end to end 4.74992 ms, enqueue 1.01769 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.5875 ms - Host latency: 4.73852 ms (end to end 4.75818 ms, enqueue 1.01811 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58987 ms - Host latency: 4.73746 ms (end to end 4.75689 ms, enqueue 1.03277 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58997 ms - Host latency: 4.7413 ms (end to end 4.75951 ms, enqueue 1.01619 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58946 ms - Host latency: 4.72262 ms (end to end 4.74041 ms, enqueue 1.02238 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58925 ms - Host latency: 4.73135 ms (end to end 4.74933 ms, enqueue 1.01594 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59028 ms - Host latency: 4.73451 ms (end to end 4.75285 ms, enqueue 1.02201 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58268 ms - Host latency: 4.73112 ms (end to end 4.74874 ms, enqueue 1.02508 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58301 ms - Host latency: 4.72178 ms (end to end 4.74047 ms, enqueue 1.01762 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58886 ms - Host latency: 4.65172 ms (end to end 4.66926 ms, enqueue 1.51528 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58914 ms - Host latency: 4.64406 ms (end to end 4.65896 ms, enqueue 1.63688 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58699 ms - Host latency: 4.64383 ms (end to end 4.65996 ms, enqueue 1.65472 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58555 ms - Host latency: 4.64166 ms (end to end 4.65729 ms, enqueue 1.63208 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.5912 ms - Host latency: 4.70112 ms (end to end 4.71844 ms, enqueue 1.32826 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59344 ms - Host latency: 4.73959 ms (end to end 4.75857 ms, enqueue 1.02899 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58987 ms - Host latency: 4.73836 ms (end to end 4.75505 ms, enqueue 1.01709 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59006 ms - Host latency: 4.73572 ms (end to end 4.75276 ms, enqueue 1.02136 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59048 ms - Host latency: 4.71992 ms (end to end 4.73885 ms, enqueue 1.02228 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59038 ms - Host latency: 4.70565 ms (end to end 4.72057 ms, enqueue 1.07745 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59613 ms - Host latency: 4.654 ms (end to end 4.66982 ms, enqueue 1.64631 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60891 ms - Host latency: 4.6658 ms (end to end 4.68058 ms, enqueue 1.64453 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.63901 ms - Host latency: 4.72241 ms (end to end 4.74214 ms, enqueue 1.34059 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.62897 ms - Host latency: 4.68709 ms (end to end 4.69999 ms, enqueue 1.66216 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.63082 ms - Host latency: 4.70751 ms (end to end 4.7218 ms, enqueue 1.45334 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.62992 ms - Host latency: 4.6874 ms (end to end 4.70267 ms, enqueue 1.64911 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.622 ms - Host latency: 4.73571 ms (end to end 4.75325 ms, enqueue 1.20652 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59867 ms - Host latency: 4.6564 ms (end to end 4.67043 ms, enqueue 1.59722 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60093 ms - Host latency: 4.65856 ms (end to end 4.67501 ms, enqueue 1.66334 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60172 ms - Host latency: 4.72034 ms (end to end 4.73595 ms, enqueue 1.27314 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60275 ms - Host latency: 4.7422 ms (end to end 4.76001 ms, enqueue 1.03055 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60154 ms - Host latency: 4.75237 ms (end to end 4.76968 ms, enqueue 1.01521 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60103 ms - Host latency: 4.65785 ms (end to end 4.67402 ms, enqueue 1.57283 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59702 ms - Host latency: 4.65447 ms (end to end 4.66899 ms, enqueue 1.6537 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60101 ms - Host latency: 4.66719 ms (end to end 4.68365 ms, enqueue 1.57606 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60033 ms - Host latency: 4.66338 ms (end to end 4.67982 ms, enqueue 1.52695 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61044 ms - Host latency: 4.6709 ms (end to end 4.68477 ms, enqueue 1.65308 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61915 ms - Host latency: 4.75122 ms (end to end 4.76687 ms, enqueue 1.15017 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60728 ms - Host latency: 4.74371 ms (end to end 4.76132 ms, enqueue 1.03044 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59255 ms - Host latency: 4.72791 ms (end to end 4.74779 ms, enqueue 1.03347 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59315 ms - Host latency: 4.74182 ms (end to end 4.75947 ms, enqueue 1.01835 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59058 ms - Host latency: 4.73859 ms (end to end 4.75806 ms, enqueue 1.01575 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59283 ms - Host latency: 4.73408 ms (end to end 4.75116 ms, enqueue 1.02853 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59253 ms - Host latency: 4.73284 ms (end to end 4.7496 ms, enqueue 1.0173 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59139 ms - Host latency: 4.73563 ms (end to end 4.7526 ms, enqueue 1.01703 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59039 ms - Host latency: 4.68552 ms (end to end 4.70142 ms, enqueue 1.15013 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58688 ms - Host latency: 4.64351 ms (end to end 4.65852 ms, enqueue 1.65355 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59246 ms - Host latency: 4.78259 ms (end to end 4.79854 ms, enqueue 0.765063 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58976 ms - Host latency: 4.79293 ms (end to end 4.80812 ms, enqueue 0.447778 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59243 ms - Host latency: 4.72633 ms (end to end 4.74291 ms, enqueue 1.14955 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58927 ms - Host latency: 4.70409 ms (end to end 4.71877 ms, enqueue 1.46211 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58922 ms - Host latency: 4.69727 ms (end to end 4.71404 ms, enqueue 1.4674 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60126 ms - Host latency: 4.71378 ms (end to end 4.72882 ms, enqueue 1.4665 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.6146 ms - Host latency: 4.72861 ms (end to end 4.74229 ms, enqueue 1.46687 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61904 ms - Host latency: 4.73428 ms (end to end 4.75139 ms, enqueue 1.45776 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.6167 ms - Host latency: 4.72507 ms (end to end 4.7394 ms, enqueue 1.46343 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.6165 ms - Host latency: 4.72825 ms (end to end 4.74551 ms, enqueue 1.48093 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61758 ms - Host latency: 4.72815 ms (end to end 4.74431 ms, enqueue 1.47295 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60859 ms - Host latency: 4.727 ms (end to end 4.74077 ms, enqueue 1.45435 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.6021 ms - Host latency: 4.71687 ms (end to end 4.73274 ms, enqueue 1.45869 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61111 ms - Host latency: 4.72588 ms (end to end 4.73958 ms, enqueue 1.46362 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.6085 ms - Host latency: 4.71299 ms (end to end 4.72961 ms, enqueue 1.4863 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.62021 ms - Host latency: 4.73657 ms (end to end 4.75117 ms, enqueue 1.46689 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61001 ms - Host latency: 4.7217 ms (end to end 4.73774 ms, enqueue 1.47329 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60894 ms - Host latency: 4.72175 ms (end to end 4.73774 ms, enqueue 1.45996 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59602 ms - Host latency: 4.69124 ms (end to end 4.70601 ms, enqueue 1.48582 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58879 ms - Host latency: 4.7061 ms (end to end 4.72107 ms, enqueue 1.45811 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59341 ms - Host latency: 4.7093 ms (end to end 4.72632 ms, enqueue 1.46155 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59861 ms - Host latency: 4.67756 ms (end to end 4.69421 ms, enqueue 1.54897 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59929 ms - Host latency: 4.65381 ms (end to end 4.66875 ms, enqueue 1.64392 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60144 ms - Host latency: 4.71389 ms (end to end 4.73044 ms, enqueue 1.3313 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59775 ms - Host latency: 4.71812 ms (end to end 4.73245 ms, enqueue 1.02263 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.57788 ms - Host latency: 4.66929 ms (end to end 4.68704 ms, enqueue 1.26707 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.58318 ms - Host latency: 4.64211 ms (end to end 4.6571 ms, enqueue 1.6553 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59839 ms - Host latency: 4.6543 ms (end to end 4.66938 ms, enqueue 1.65542 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59526 ms - Host latency: 4.66873 ms (end to end 4.68474 ms, enqueue 1.57432 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60022 ms - Host latency: 4.74575 ms (end to end 4.76467 ms, enqueue 1.02512 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59861 ms - Host latency: 4.725 ms (end to end 4.74438 ms, enqueue 1.03474 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60442 ms - Host latency: 4.74048 ms (end to end 4.75903 ms, enqueue 1.02407 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60613 ms - Host latency: 4.74568 ms (end to end 4.76294 ms, enqueue 1.02964 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60151 ms - Host latency: 4.74846 ms (end to end 4.76499 ms, enqueue 1.01465 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60273 ms - Host latency: 4.7436 ms (end to end 4.76155 ms, enqueue 1.02131 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59956 ms - Host latency: 4.73704 ms (end to end 4.75496 ms, enqueue 1.02078 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60122 ms - Host latency: 4.74536 ms (end to end 4.76064 ms, enqueue 1.02913 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60396 ms - Host latency: 4.75247 ms (end to end 4.77131 ms, enqueue 1.0165 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60598 ms - Host latency: 4.74436 ms (end to end 4.76189 ms, enqueue 1.01392 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.59995 ms - Host latency: 4.71816 ms (end to end 4.73706 ms, enqueue 1.02988 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.60974 ms - Host latency: 4.7179 ms (end to end 4.73477 ms, enqueue 1.07554 ms)\n", - "[07/25/2022-16:44:24] [I] Average on 10 runs - GPU latency: 1.61443 ms - Host latency: 4.66958 ms (end to end 4.68594 ms, enqueue 1.65239 ms)\n", - "[07/25/2022-16:44:24] [I] \n", - "[07/25/2022-16:44:24] [I] === Performance summary ===\n", - "[07/25/2022-16:44:24] [I] Throughput: 321.422 qps\n", - "[07/25/2022-16:44:24] [I] Latency: min = 4.61383 ms, max = 5.11646 ms, mean = 4.71056 ms, median = 4.71863 ms, percentile(99%) = 4.80322 ms\n", - "[07/25/2022-16:44:24] [I] End-to-End Host Latency: min = 4.62366 ms, max = 5.13928 ms, mean = 4.72723 ms, median = 4.73462 ms, percentile(99%) = 4.81934 ms\n", - "[07/25/2022-16:44:24] [I] Enqueue Time: min = 0.337158 ms, max = 1.83459 ms, mean = 1.28084 ms, median = 1.0896 ms, percentile(99%) = 1.71924 ms\n", - "[07/25/2022-16:44:24] [I] H2D Latency: min = 3.01642 ms, max = 3.51599 ms, mean = 3.09767 ms, median = 3.10742 ms, percentile(99%) = 3.1925 ms\n", - "[07/25/2022-16:44:24] [I] GPU Compute Time: min = 1.56671 ms, max = 1.6599 ms, mean = 1.59911 ms, median = 1.59741 ms, percentile(99%) = 1.63635 ms\n", - "[07/25/2022-16:44:24] [I] D2H Latency: min = 0.00561523 ms, max = 0.0314941 ms, mean = 0.0137833 ms, median = 0.0134277 ms, percentile(99%) = 0.0292969 ms\n", - "[07/25/2022-16:44:24] [I] Total Host Walltime: 3.00851 s\n", - "[07/25/2022-16:44:24] [I] Total GPU Compute Time: 1.54634 s\n", - "[07/25/2022-16:44:24] [W] * Throughput may be bound by host-to-device transfers for the inputs rather than GPU Compute and the GPU may be under-utilized.\n", - "[07/25/2022-16:44:24] [W] Add --noDataTransfers flag to disable data transfers.\n", - "[07/25/2022-16:44:24] [I] Explanations of the performance metrics are printed in the verbose logs.\n", - "[07/25/2022-16:44:24] [I] \n", - "&&&& PASSED TensorRT.trtexec [TensorRT v8205] # trtexec --onnx=models/mobilenetv2_ptq.onnx --int8 --saveEngine=models/mobilenetv2_ptq.trt\n" - ] - } - ], - "source": [ - "# Set static member of TensorQuantizer to use Pytorch’s own fake quantization functions\n", - "quant_nn.TensorQuantizer.use_fb_fake_quant = True\n", - "\n", - "# Exporting to ONNX\n", - "dummy_input = torch.randn(64, 3, 224, 224, device='cuda')\n", - "input_names = [ \"actual_input_1\" ]\n", - "output_names = [ \"output1\" ]\n", - "torch.onnx.export(\n", - " q_model,\n", - " dummy_input,\n", - " \"models/mobilenetv2_ptq.onnx\",\n", - " verbose=False,\n", - " opset_version=13,\n", - " do_constant_folding = False)\n", - "\n", - "# Converting ONNX model to TRT\n", - "!trtexec --onnx=models/mobilenetv2_ptq.onnx --int8 --saveEngine=models/mobilenetv2_ptq.trt" - ] - }, - { - "attachments": { - "img5.JPG": { - "image/jpeg": "/9j/4AAQSkZJRgABAQEAeAB4AAD/4RDuRXhpZgAATU0AKgAAAAgABAE7AAIAAAAMAAAISodpAAQAAAABAAAIVpydAAEAAAAYAAAQzuocAAcAAAgMAAAAPgAAAAAc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAEFubmllIFN1cmxhAAAFkAMAAgAAABQAABCkkAQAAgAAABQAABC4kpEAAgAAAAM5NwAAkpIAAgAAAAM5NwAA6hwABwAACAwAAAiYAAAAABzqAAAACAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAMjAyMjowNjoxMyAxMTowMToxNAAyMDIyOjA2OjEzIDExOjAxOjE0AAAAQQBuAG4AaQBlACAAUwB1AHIAbABhAAAA/+ELHmh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8APD94cGFja2V0IGJlZ2luPSfvu78nIGlkPSdXNU0wTXBDZWhpSHpyZVN6TlRjemtjOWQnPz4NCjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iPjxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+PHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9InV1aWQ6ZmFmNWJkZDUtYmEzZC0xMWRhLWFkMzEtZDMzZDc1MTgyZjFiIiB4bWxuczpkYz0iaHR0cDovL3B1cmwub3JnL2RjL2VsZW1lbnRzLzEuMS8iLz48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOnhtcD0iaHR0cDovL25zLmFkb2JlLmNvbS94YXAvMS4wLyI+PHhtcDpDcmVhdGVEYXRlPjIwMjItMDYtMTNUMTE6MDE6MTQuOTY3PC94bXA6Q3JlYXRlRGF0ZT48L3JkZjpEZXNjcmlwdGlvbj48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOmRjPSJodHRwOi8vcHVybC5vcmcvZGMvZWxlbWVudHMvMS4xLyI+PGRjOmNyZWF0b3I+PHJkZjpTZXEgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5LzAyLzIyLXJkZi1zeW50YXgtbnMjIj48cmRmOmxpPkFubmllIFN1cmxhPC9yZGY6bGk+PC9yZGY6U2VxPg0KCQkJPC9kYzpjcmVhdG9yPjwvcmRmOkRlc2NyaXB0aW9uPjwvcmRmOlJERj48L3g6eG1wbWV0YT4NCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgPD94cGFja2V0IGVuZD0ndyc/Pv/bAEMABwUFBgUEBwYFBggHBwgKEQsKCQkKFQ8QDBEYFRoZGBUYFxseJyEbHSUdFxgiLiIlKCkrLCsaIC8zLyoyJyorKv/bAEMBBwgICgkKFAsLFCocGBwqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKv/AABEIAKoCGgMBIgACEQEDEQH/xAAfAAABBQEBAQEBAQAAAAAAAAAAAQIDBAUGBwgJCgv/xAC1EAACAQMDAgQDBQUEBAAAAX0BAgMABBEFEiExQQYTUWEHInEUMoGRoQgjQrHBFVLR8CQzYnKCCQoWFxgZGiUmJygpKjQ1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoOEhYaHiImKkpOUlZaXmJmaoqOkpaanqKmqsrO0tba3uLm6wsPExcbHyMnK0tPU1dbX2Nna4eLj5OXm5+jp6vHy8/T19vf4+fr/xAAfAQADAQEBAQEBAQEBAAAAAAAAAQIDBAUGBwgJCgv/xAC1EQACAQIEBAMEBwUEBAABAncAAQIDEQQFITEGEkFRB2FxEyIygQgUQpGhscEJIzNS8BVictEKFiQ04SXxFxgZGiYnKCkqNTY3ODk6Q0RFRkdISUpTVFVWV1hZWmNkZWZnaGlqc3R1dnd4eXqCg4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2dri4+Tl5ufo6ery8/T19vf4+fr/2gAMAwEAAhEDEQA/APpGiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKMj1oyPWgAooyPWjI9aACijI9aMj1oAKKMj1oyPWgAooyPWjI9aACijI9aMj1oAKKMj1oyPWgAooyPWjI9aACijI9aMj1oAKKMj1oyPWgAooyPWjI9aACijI9aMj1oAKKMj1oyPWgAooyPWjI9aACijI9aMj1oAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKzfEP/ACALr/dH/oQrSrN8Q/8AIAuv90f+hCgDz+iiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAO48Kf8AIDH/AF0atqsXwp/yAx/10atqgAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAopCwXqab5qeooAfRTPNT1FHmp6igB9FM81PUUeanqKAH0UzzU9RR5qeooAfRTPNT1FHmp6igB9FNDqehp1ABRRRQAUUUUAFZviH/kAXX+6P/QhWlWb4h/5AF1/uj/0IUAef0UUUAFFFFABRRRQBqWkdsdDuLiS0jeWF1UMzPyCe4DCoHhN0yRW9mIZVUs4VjtK4BB+YnHFWrCVoNBvPKnWKV3QoBKFYgde+ai0+T7VeTvdXGHMBA3ybBKQAApPvTAdptiY9Ztob6BHSbkZbII9QQcGqj2U3ktcBVEO8qCXUEkdgM5NbqPEdQ0qUz2qrCm2QLIAFPNVbo28kttfo8Y8mQLLbiReinOVGeQaQGZJp9zFC0rx4RCA+GBK56ZAOR+NSaZYi9mk81isMMZkkK9cDsPetC+kaEXTxX1q9vcA4WJU3vnoDxnj1qpo13HBJcQXDbIrmIxlz/CexNABaCyv7pbU2ot/MO1JUdmIPbIJwf0qCfTbm3BaRV2eaYd3mLjd6Hnj8at2Wny2V5Fd3u2O2iYOJNwIkxyAuOuafJMupaXcASxRyteGYpI4X5SMd+tAFBtNukneFowGjUM+XXaoPTJzilGmXhuGgEBMiruIBHTGc571sXQgmubt0uoXzFGEjMwVJMAZzyOnpTzJEdRjlW5ttv2DyyVkAAbaRj2oAxRpV15sCSKsYnbajM4xn04PX261K1lJbSXqLBHcJGpBcuDsGcZ4PX2qzbOkWmWQkljBS9DsPMBIXjnGaeqotxqxM8GJo22YlX5stkDr6UMDNi0u7nZFiSNndQyr5yAkYznGc9KI9Mu5f9XEDkkKN65bHXHPP4VejIs9NCwTRvd3Q2u/mr+6T+716mreneRaz2LtdQSKhYO8ko/d8nhVzwD6/wAqAMgaa50sXm9Duk2hN4z0+vX260XNtO91FCLRYpDGpCIeox9488VZdEbQ2gE8PmR3JcjzByNvb1544q601v8AbGjeaIefYCFZA4IVsdCR0oGYrafcq0Q8vd5pwhRgwY+mQcU6bS7y3ikklh2rEcP8wJX8M5/GrtrKsGnR2sjxmRrtZB+8BCAdTnOBVh3jNxrR86EiZf3f71fm5zxzQI5+iiigAooooAKKKKAO48Kf8gMf9dGrarF8Kf8AIDH/AF0atqgAooooAKKY0qr1NJ56f3h+dAElFR+en94fnR56f3h+dAElFR+en94fnR56f3h+dAElFR+en94fnR56f3h+dAElFR+en94fnR56f3h+dAElFNWRW6GnUAFFFFABSE4FLSN900Acl4lvnae3tA7ostxGrFGKnBYZ5FT/ANgWn/PS6/8AAl/8ayvEX/Icsf8Ar5j/APQhXU1y+zhOrLmV9F+pFk3qZf8AYFp/z0uv/Al/8aP7AtP+el1/4Ev/AI1qUVf1el/Kh8sexl/2Baf89Lr/AMCX/wAaP7AtP+el1/4Ev/jWpRR9XpfyoOWPYy/7AtP+el1/4Ev/AI0f2Baf89Lr/wACX/xrUoo+r0v5UHLHsZf9gWn/AD0uv/Al/wDGj+wLT/npdf8AgS/+NalFH1el/Kg5Y9jFutNhsI4ri3luQ6XEP3rhyCDIoIIJ54Jrqom3Rg1gax/yDx/13g/9GrW9B/qV+lOEIwm1FW0X6gkk9CSiiitigooooAKzfEP/ACALr/dH/oQrSrN8Q/8AIAuv90f+hCgDz+iiigAooooAKTIqtPcCKeJW5DOqn8TXa/8ACO6b/wA8D/323+NAHJ5FGRXWf8I7pv8AzwP/AH23+NH/AAjum/8APA/99t/jQByeRRkV1n/CO6b/AM8D/wB9t/jR/wAI7pv/ADwP/fbf40AcnkUZFdZ/wjum/wDPA/8Afbf40f8ACO6b/wA8D/323+NAHPQajNDbtb/LLA3WKQZGfUdx+FVmOWJACgnoO1a2vafa6fbwtbR7C8mD8xPGKyKACiiigApM0McKTVnw1Db6rfXcN0m8RKpHzEYyT6UAVsijIrrP+Ed03/ngf++2/wAaP+Ed03/ngf8Avtv8aAOTyKMius/4R3Tf+eB/77b/ABo/4R3Tf+eB/wC+2/xoA5PIoyK6z/hHdN/54H/vtv8AGj/hHdN/54H/AL7b/GgDk8ilrrB4e00f8sD/AN9t/jXL3apFqFxFGMJHIVAz0ANAEdFFFABRRRQB3HhT/kBj/ro1bVYvhT/kBj/ro1bVABTZDtjJp1Rz/wCpb6UAcdeSLqXia3tLgFogrkruIzx7Vof8I/pn/Puf+/r/AONZEH/I8Rf7j/yrqa5I0qc5zcop69vJEJJt3M3/AIR/TP8An3P/AH9f/Gj/AIR/TP8An3P/AH9f/GtKitfq9H+Rfch8q7Gb/wAI/pn/AD7n/v6/+NH/AAj+mf8APuf+/r/41pUUfV6P8i+5ByrsZv8Awj+mf8+5/wC/r/40f8I/pn/Puf8Av6/+NaVFH1ej/IvuQcq7Gb/wj+mf8+5/7+v/AI0f8I/pn/Puf+/r/wCNaVFH1ej/ACL7kHKuxjrZW+m67pxtEaPzDIG+djn5fc11SnKiubvf+Q3pf+/J/wCg10ifdFTRiouairK/6IFpcWiiiugoKRvumlpG+6aAOE8Rf8h2x/6+Y/8A0IV1Nct4i/5Dtj/18x/+hCuprCP8WXy/UlbsKKr30L3FhPFC7RyMhCOpwQe1YDa5Orm95Ns0Hkhf+m23d/P5aKlaNN+9/X9fqDkludPRXODVrqy/0WGI3BtVVZSySO0jYBOCoIHXvU82qz2018xjjyphCMzMFUPnlvTHfFT9Zp/d/X6C50blFc0+pX121qYngMgvGiBjZvLcCMnOe4z/ACqWbUdRmS0EHkxTfa2glBJKsVB/HHH8qSxUHsn/AFb/ADDnR0FFZsd9cnWDbTLDFF0Tdu3S8Zyvb8OtaVbxmpbFJ3KGsf8AIPH/AF3h/wDRq1vQf6lfpWDrH/IPH/XeH/0atb0H+pX6Ul/Efov1DqSUUUVoMKKKKACs3xD/AMgC6/3R/wChCtKs3xD/AMgC6/3R/wChCgDz+iiigAopCwHU4oyD0NAGVqR/0q3/AOuqfzFeq15VqX/H1b/9dU/mK9VoAKKKKACiqlzqljZyeXdXcUT4ztZwDViKWOaNZIXWRGGQynINSpRbsmK62H0UUVQzn/Fn/Hpbf9df6GufHSug8Wf8elt/11/oa58dKAFoopNy5xkUANk+4aseBP8AkN6l/wBc0/maryH92aseBP8AkN6n/wBc0/maAO6ooqOeeK2haWd1jjX7zMcAUm0ldgSUVTt9W0+6lEdveQyOeiq4yasTTxW8TSzyLHGvVmOAKSnFrmT0FdMkoqrbalZXrlbW6ilYDJVGBNWqcZKSumF09grhL3/kMXn/AF2b+dd3XCXv/IYvP+uzfzpjI6KKKACiiigDuPCn/IDH/XRq2qxfCn/IDH/XRq2qACo5/wDUt9KJZ4oI2kmcIijLMxwAPWo2nimt98Tq6sMhlOQadna4HFwf8jxF/uP/ACrqa5aD/kd4v9x/5V1Nc9L4p+v6IlbsKKKK3KCiiigAooooAKKKKAM29/5Del/78n/oNdIn3RXN3v8AyG9L/wB+T/0GukT7orGn8U/X9EJbsWiiithhSN900tI33TQBwniL/kO2P/XzH/6EK6muW8Rf8h2x/wCvmP8A9CFdTWEf4svl+pK3YVUGm2otxB5f7sS+dj/a3bv51borZxT3Q7XKM2kwTXZuA80TtjeIpSofHTIHWmahpYuYbnyR+9uNm7dIVGFPqBxWfpl1PdxmSW7vvMUuceSoiIBOBu2f1pui6jcXclptu5rnzIyblXiAWI44IYAd+3NcCq0aiS5fi9P66kcy+8u6dpDwRxi7ct5MvmQIJC+z5cY3EDPU9qsS6TbSwtGTIuZjOHRyrK57giqdtJdtqF3ZLfSMywgq88QBDZIyoAGVqxpj3BuruOS4a5giZVSRwoO7HzDgAccVpT9m0o8u/p0/4b8gVtrEq6XELuO4eW4kaP7iySkqDjGceuKu0UV1Rio7FpJGL4lvfsmkNJtLlJYm2r1OJFOKpx+PnVAP7Gvj9Fq/r9sJdPGf+e8P/o1avw6MhjFa0qlKMmpwvt1a7mcoyctHYw/+FgP/ANAW+/75o/4WA/8A0Bb7/vmt/wDsVKP7FSun21D/AJ9fixclT+b8DA/4WA//AEBb7/vmj/hYD/8AQFvv++a3/wCxUo/sVKPbUP8An1+LDkqfzfgYH/CwH/6At9/3zVTVfHD3WlzQnSbyIOAN7rwORXVf2KlZ+vaMi6Hcn/ZH8xR7ah/z6/FhyVP5vwPO/wC3P+mD0f25/wBMHqx/Zi0f2YtHtqH/AD6/FhyVP5vwMq91U3CGPGwN2JqGDULmD7khI9G5qzfaQxcNGOQc0yHR5pGzIcD0Fejh8ZhIUHGcOu2/5nNUo1pTTUvmK2oNdXFvvXaRKmSOnUV7IOa8fnsxb3FsB/z1T+Yr2AcV5FadOc7048q7XudkFJK0ncWo55DFbSSAZKIWA+gqSkIDKQRkEYIrCV2mkWcz4Y061v8ATHvr6GO5nuJGLtIobHPTnpT/AA8n2LXNU06JibeNg8ak5257f59KZHYa1oXnRaRHDdWrsXRZD8ye3UZpnhi6s45pTc3JbU7x/wB5GyEFTzx0+teRStCdKDXK1vfS+ltO93qc60snudVSE4FLSEZFewdByfjW/EFrbYUsfN6D6GuT/tz/AKYPXXeMbMSW9sT/AM9f6GuZ/sxa6KdSlGNpQu/Vmcozb0lYrPrhKn9yw+tZc1688/mK5UgY+U9K2pdLUoayG0mZJGEfRjnOK78JisNTq80oWXq3+BhWpVJRspXJodYuEG2TEi+/Brp/h7L5ur6i2MZiTj8TXPQaOVG6Qkmuh+H8ezWdSH/TNP5mufF18PVf7qny+f8AwNi6NOpBe/K531ZHin/kWbv/AHV/9CFa9Z2u2k19olxb2yhpXA2gnGeQa8jFRcqE0t7P8jd7GLe6VZt4Pjuo4UhuIoEkWVBtbOB1I61C8zazq+i2l780LW4ndT/G2D1/L9TU/wDZmu6hYw6deCC0tI1VXZG3M4H4n09qvapocjG0udJZY7myULGH6Ovof8968+VOUnzxh7vu3W17PXT+rmHK2tEUvE1jb6ZFa6hYRJbzxTqv7tdoYHPUD6V1AOQDXGahPdz6lap4nCWdpGfMURKWWRh2JGa662uYby3Se2ffE/3WAxmunCyi6lTl0206+bt0Li1zOxNXl+rasYNfv4/Ic7Z2GR35r1CvN9TsVl1y+Y952/nXqU5Qi7zjf8C5KT2djO/tw/8APu9H9uN/z7N+dWf7NWl/s1K29tQ/59fizPkn/N+CKn9tv/z7N+dH9tyf8+zfn/8AWq3/AGalH9mpR7aj/wA+l97/AMw5J/zfgjY0PxffWmmiKHRprhd5O9XwP5Vo/wDCb6p28PzfjN/9jVzwtpMb6KCR/wAtGrZ/seP0o9tR/wCfS++X+Yck/wCb8jhNa8W6jc2M0MulPbrIhUsZc4BH0rj7bVL7TvmsbqWHHOFbg/UdDXq2seG1u4GReNwIyO1Y9n4FggG513kd25r0MNj6FGlKLp79N/vu2c9XD1JyTUv6+RgeCNXv9W8WBr9BhY22SBdu7jmvTa4zSrNbPxlEiDH7t/5V2deFzqdWpKKsr7fJHVTTirN3CqWr3ElppFzNAdsip8rEZ29s/hV2myRpLG0cih0YYZSMgiiacotI0MxreDRbOW+Rp5mSL5t8zNvPrycD8KgudWv7COVbuO3eXyvOjMe7bjcAQc/XrV2HRrOEMuJZEZDHsklZlCnqACeKauiWaxSxkSP5qhGaSVmIUHIAJPArllTq2tCy0/HXy9DOz6ED6jf2/wBpS4S3Z4FSY+UGwYySGHJ6jBq7ZXbXjzum026vsiYDlsD5j9M8fhU32aL7Q8xXLyIEbPQgZ7fiaLe3itLdIIF2RxjCgdq2hCaer0/r9CknclooorYozb3/AJDel/78n/oNdIn3RXN3v/Ib0v8A35P/AEGukT7orGn8U/X9EJbsWiiithhSN900tI33TQBwniL/AJDtj/18x/8AoQrqa5bxF/yHbH/r5j/9CFdTWEf4svl+pK3YUUUVuUVbOyW0sfsyuWGW+Yj+8Sf606ztBZ6fFao5IjQIHxz9asUVCpxja3RWFYzYtLmSSSaW/kluGiMSSMijyweeg6mn6Zp82nx+W9408QGFQxquDnrkcmr9FTGlCLTX5sOVBRRRWoyhrH/IPH/XeH/0atb0H+pX6Vg6x/yDx/13h/8ARq1vQf6lfpWa/iP0X6i6klFFFaDCiiigArN8Q/8AIAuv90f+hCtKs3xD/wAgC6/3R/6EKAPP6KKKAEKg9RRtA6ClooAydS/4+rf/AK6p/MV6rXlWpf8AH1b/APXVP5ivVaACiiigAqMwQtKJWiQyL0cqMj8akooAKKKKAOf8Wf8AHpbf9df6GufHSug8Wf8AHpbf9df6GufHSgBaTYvpS0UAMkGIzVjwJ/yG9T/65p/M1Xk+4aseBP8AkN6l/wBc0/maAO6ooooAKKKKAGuiSKVkVXU9QwyDRHGkSBIkVFHRVGAKdRRbqAVwl7/yGLz/AK7N/Ou7rhL3/kMXn/XZv50AR0UUUAFFFFAHceFP+QGP+ujVtVi+FP8AkBj/AK6NW1QAhAPUVHMAIWwO1S1HP/qW+lAHDwf8jxF/uP8Ayrqa5aD/AJHiL/cf+VdTWFL4p+v6IlbsKKKK3KCiiigAooooAKKKKAM29/5Del/78n/oNdIn3RXN3v8AyG9L/wB+T/0GukT7orGn8U/X9EJbsWiiithhSN900tIeRQBwniL/AJDlj/18x/8AoQrqayPEGjy3cqSwMVeNg6sOxByKyTbeIv8An/l/75X/AArD3o1HJK97E63Otorkvs3iP/n/AJf++V/wo+zeI/8An/l/75X/AAquef8AL+Q7vsdbRXJfZvEf/P8Ay/8AfK/4UfZvEf8Az/y/98r/AIUc8/5fyC77HW0VyX2bxH/z/wAv/fK/4UfZvEf/AD/y/wDfK/4Uc8/5fyC77HW0VyX2bxH/AM/8v/fK/wCFH2bxH/z/AMv/AHyv+FHPP+X8gu+xvax/yDx/13g/9GrW9B/qV+lchZaZqdzIg1C9mkjV1cphQCVII7eoFdjEu2MCiHM5OTVtv1BbjqKKK1GFFFFABWb4h/5AF1/uj/0IVpVm+If+QBdf7o/9CFAHn9FFFABRRRQBSvbZpdrRnDKQVPoRTG1bxNu41WQD/rmn+FaFJgUAZ/8Aavif/oKyf9+0/wDiaP7V8T/9BWT/AL9p/wDE1oYFGBQBn/2r4n/6Csn/AH7T/wCJo/tXxP8A9BWT/v2n/wATWhgUYFAGf/avif8A6Csn/ftP/iaUar4nzzqsn/ftP/iav4FGBQBGt5qN5Gq6jctOFORlQMH8BUtGKKACiiigBGGVIrOU6jYTyS6XctbtIAHIUHIH1FaVJigDP/tXxP8A9BWT/v2n/wATR/avif8A6Csn/ftP/ia0MCjAoAz/AO1fE/8A0FZP+/af/E0f2r4n/wCgrJ/37T/4mtDAowKAM/8AtXxP/wBBWT/v2n/xNH9q+J/+grJ/37T/AOJrQwKMCgCpDq3iQOPM1OQj/rmn+FWVMkjtJM26Rzlmx1NOwKWgAooooAKKKKAO48Kf8gMf9dGrarF8Kf8AIDH/AF0atqgAqOf/AFLfSpKbIu5CKAOFh/5HeL/cf+VdTXP6xoNxLei4tZZIZBnDxsVP5is7+xtb/wCglef9/wBv8a50qkZSaV7vv5JfoTqmdjRXHf2Nrf8A0Erz/v8At/jR/Y2t/wDQSvP+/wC3+NVzVf5fx/4AXfY7GiuO/sbW/wDoJXn/AH/b/Gj+xtb/AOglef8Af9v8aOar/L+P/AC77HY0Vx39ja3/ANBK8/7/ALf40f2Nrf8A0Erz/v8At/jRzVf5fx/4AXfY7GiuO/sbW/8AoJXn/f8Ab/GgaNrf/QSvP+/7f40c1X+X8f8AgBd9jfvf+Q3pf+/J/wCg10ifdFcro2iTx3Udxe3FxO8edvmyswGRg8GuqAwKKcZJycur/RIav1FooorYYUUUUAIUB6im+UnpT6KAGeUnpR5SelPooAZ5SelHlJ6U+igBnlJ6UeUnpT6KAGeUnpR5SelPooAaI1HQU6iigAooooAKKKKACs3xD/yALr/dH/oQrSpGUMuGAI9CKAPLqK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDzGivTvIh/55J/3yKPIh/55J/3yKAPMaK9O8iH/AJ5J/wB8ijyIf+eSf98igDJ8Kf8AIDH/AF0atqkVFQYRQo9AMUtABRRRQAhRT1FJ5a+lOooAb5a+lHlr6U6igBvlr6UeWvpTqKAG+WvpR5a+lOooAb5a+lHlr6U6igBAoHQUtFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAf/9k=" - } - }, - "cell_type": "markdown", - "id": "d3e676e7", - "metadata": {}, - "source": [ - "\n", - "## 5. Quantization Aware Training (QAT)\n", - "\n", - "PTQ resulted in a ~3% accuracy drop. After PTQ is performed, sometimes the model may perform poorly by not retaining the accuracy as the process is not able to mitigate the large quantization error induced by low-bit quantization. This could happen if there are sensitive layers in the network, like the Depth wise convolutional networks, in MobileNets which are more susceptible to producing larger quantization error. \n", - "\n", - "This is when we might want to consider using QAT. The idea behind QAT is simple: you can improve the lost accuracy of the quantized model, if you had trained the model with quantization error. There are many ways of doing this, starting the training of the model from scratch or fine-tuning a pre-trained model. Whatever method you choose, the quantization error is induced in the training loss by inserting fake-quantization operations. The operation is called “fake” because we quantize the data and immediately perform a dequantize operation producing an approximate version of the data where both input and output still remain as floating point values. We are here trying to simulate the effects of quantization without changing much in the model. \n", - "In the forward-pass, we fake-quantize the weights and activations and use these fake-quantized outputs to perform the layer operations.\n", - "\n", - "![img5.JPG](attachment:img5.JPG)\n", - "\n", - "In the backward pass, while calculating gradient, the quantization operation’s derivative is undefined at the step boundaries, and zero everywhere else. To handle this, QAT uses Straight-through Estimator by approximating the derivative to be 1 for inputs in the representable range. This estimator is essentially letting gradients pass as is through this operator in the backward pass. When the QAT process is done, the scales that were used to quantize the weights and activations are stored in the model and can be used for inference. " - ] - }, - { - "cell_type": "markdown", - "id": "bcc10e0f", - "metadata": {}, - "source": [ - "Usually the finetuning of QAT model should be quick compared to the full training of the original model. For this Mobilenetv2 model, it is enough to finetune for 2 epochs to get acceptable accuracy. \n", - "\n", - "tensor_quant function in `pytorch_quantization` toolkit is responsible for the above tensor quantization. Usually, per channel quantization is recommended for weights, while per tensor quantization is recommended for activations in a network.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 20, - "id": "dc144132", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch: [ 1 / 2] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 1.806\n", - "Test Acc: 69.88%\n", - "Epoch: [ 2 / 2] LR: 0.000100\n", - "Batch: [ 100 | 147] loss: 1.800\n", - "Test Acc: 69.49%\n", - "Checkpoint saved\n" - ] - } - ], - "source": [ - "# Finetune the QAT model for 2 epochs\n", - "num_epochs=2\n", - "\n", - "for epoch in range(num_epochs):\n", - " print('Epoch: [%5d / %5d] LR: %f' % (epoch + 1, num_epochs, lr))\n", - "\n", - " train(q_model, train_dataloader, criterion, optimizer, epoch)\n", - " test_acc = evaluate(q_model, val_dataloader, criterion, epoch)\n", - "\n", - " print(\"Test Acc: {:.2f}%\".format(100 * test_acc))\n", - " \n", - "save_checkpoint({'epoch': epoch + 1,\n", - " 'model_state_dict': q_model.state_dict(),\n", - " 'acc': test_acc,\n", - " 'opt_state_dict': optimizer.state_dict()\n", - " },\n", - " ckpt_path=\"models/mobilenetv2_qat_ckpt\")" - ] - }, - { - "cell_type": "code", - "execution_count": 21, - "id": "0d801c67", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mobilenetv2 QAT accuracy: 69.49%\n" - ] - } - ], - "source": [ - "# Evaluate the QAT model\n", - "test_acc = evaluate(q_model, val_dataloader, criterion, 0)\n", - "print(\"Mobilenetv2 QAT accuracy: {:.2f}%\".format(100 * test_acc))" - ] - }, - { - "cell_type": "markdown", - "id": "70bdaeed", - "metadata": {}, - "source": [ - "As you can see, accuracy recovered by ~1.3%. Fine-tuning for more epochs with learning rate annealing can improve accuracy further. It should be noted that the same fine-tuning schedule will improve the accuracy of the unquantized model as well. Please refer to Achieving FP32 Accuracy for INT8 Inference Using Quantization Aware Training with NVIDIA TensorRT for detailed recommendations.\n", - "\n", - "During inference, we use `torch.fake_quantize_per_tensor_affine` and `torch.fake_quantize_per_channel_affine` to perform quantization as this is easier to convert into corresponding TensorRT operators. \n", - "\n", - "Let us now prepare this model to export into ONNX. " - ] - }, - { - "cell_type": "code", - "execution_count": 22, - "id": "176a6bfd", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "&&&& RUNNING TensorRT.trtexec [TensorRT v8205] # trtexec --onnx=models/mobilenetv2_qat.onnx --int8 --saveEngine=models/mobilenetv2_qat.trt\n", - "[07/25/2022-16:46:43] [I] === Model Options ===\n", - "[07/25/2022-16:46:43] [I] Format: ONNX\n", - "[07/25/2022-16:46:43] [I] Model: models/mobilenetv2_qat.onnx\n", - "[07/25/2022-16:46:43] [I] Output:\n", - "[07/25/2022-16:46:43] [I] === Build Options ===\n", - "[07/25/2022-16:46:43] [I] Max batch: explicit batch\n", - "[07/25/2022-16:46:43] [I] Workspace: 16 MiB\n", - "[07/25/2022-16:46:43] [I] minTiming: 1\n", - "[07/25/2022-16:46:43] [I] avgTiming: 8\n", - "[07/25/2022-16:46:43] [I] Precision: FP32+INT8\n", - "[07/25/2022-16:46:43] [I] Calibration: Dynamic\n", - "[07/25/2022-16:46:43] [I] Refit: Disabled\n", - "[07/25/2022-16:46:43] [I] Sparsity: Disabled\n", - "[07/25/2022-16:46:43] [I] Safe mode: Disabled\n", - "[07/25/2022-16:46:43] [I] DirectIO mode: Disabled\n", - "[07/25/2022-16:46:43] [I] Restricted mode: Disabled\n", - "[07/25/2022-16:46:43] [I] Save engine: models/mobilenetv2_qat.trt\n", - "[07/25/2022-16:46:43] [I] Load engine: \n", - "[07/25/2022-16:46:43] [I] Profiling verbosity: 0\n", - "[07/25/2022-16:46:43] [I] Tactic sources: Using default tactic sources\n", - "[07/25/2022-16:46:43] [I] timingCacheMode: local\n", - "[07/25/2022-16:46:43] [I] timingCacheFile: \n", - "[07/25/2022-16:46:43] [I] Input(s)s format: fp32:CHW\n", - "[07/25/2022-16:46:43] [I] Output(s)s format: fp32:CHW\n", - "[07/25/2022-16:46:43] [I] Input build shapes: model\n", - "[07/25/2022-16:46:43] [I] Input calibration shapes: model\n", - "[07/25/2022-16:46:43] [I] === System Options ===\n", - "[07/25/2022-16:46:43] [I] Device: 0\n", - "[07/25/2022-16:46:43] [I] DLACore: \n", - "[07/25/2022-16:46:43] [I] Plugins:\n", - "[07/25/2022-16:46:43] [I] === Inference Options ===\n", - "[07/25/2022-16:46:43] [I] Batch: Explicit\n", - "[07/25/2022-16:46:43] [I] Input inference shapes: model\n", - "[07/25/2022-16:46:43] [I] Iterations: 10\n", - "[07/25/2022-16:46:43] [I] Duration: 3s (+ 200ms warm up)\n", - "[07/25/2022-16:46:43] [I] Sleep time: 0ms\n", - "[07/25/2022-16:46:43] [I] Idle time: 0ms\n", - "[07/25/2022-16:46:43] [I] Streams: 1\n", - "[07/25/2022-16:46:43] [I] ExposeDMA: Disabled\n", - "[07/25/2022-16:46:43] [I] Data transfers: Enabled\n", - "[07/25/2022-16:46:43] [I] Spin-wait: Disabled\n", - "[07/25/2022-16:46:43] [I] Multithreading: Disabled\n", - "[07/25/2022-16:46:43] [I] CUDA Graph: Disabled\n", - "[07/25/2022-16:46:43] [I] Separate profiling: Disabled\n", - "[07/25/2022-16:46:43] [I] Time Deserialize: Disabled\n", - "[07/25/2022-16:46:43] [I] Time Refit: Disabled\n", - "[07/25/2022-16:46:43] [I] Skip inference: Disabled\n", - "[07/25/2022-16:46:43] [I] Inputs:\n", - "[07/25/2022-16:46:43] [I] === Reporting Options ===\n", - "[07/25/2022-16:46:43] [I] Verbose: Disabled\n", - "[07/25/2022-16:46:43] [I] Averages: 10 inferences\n", - "[07/25/2022-16:46:43] [I] Percentile: 99\n", - "[07/25/2022-16:46:43] [I] Dump refittable layers:Disabled\n", - "[07/25/2022-16:46:43] [I] Dump output: Disabled\n", - "[07/25/2022-16:46:43] [I] Profile: Disabled\n", - "[07/25/2022-16:46:43] [I] Export timing to JSON file: \n", - "[07/25/2022-16:46:43] [I] Export output to JSON file: \n", - "[07/25/2022-16:46:43] [I] Export profile to JSON file: \n", - "[07/25/2022-16:46:43] [I] \n", - "[07/25/2022-16:46:43] [I] === Device Information ===\n", - "[07/25/2022-16:46:43] [I] Selected Device: NVIDIA Graphics Device\n", - "[07/25/2022-16:46:43] [I] Compute Capability: 8.0\n", - "[07/25/2022-16:46:43] [I] SMs: 124\n", - "[07/25/2022-16:46:43] [I] Compute Clock Rate: 1.005 GHz\n", - "[07/25/2022-16:46:43] [I] Device Global Memory: 47681 MiB\n", - "[07/25/2022-16:46:43] [I] Shared Memory per SM: 164 KiB\n", - "[07/25/2022-16:46:43] [I] Memory Bus Width: 6144 bits (ECC enabled)\n", - "[07/25/2022-16:46:43] [I] Memory Clock Rate: 1.215 GHz\n", - "[07/25/2022-16:46:43] [I] \n", - "[07/25/2022-16:46:43] [I] TensorRT version: 8.2.5\n", - "[07/25/2022-16:46:44] [I] [TRT] [MemUsageChange] Init CUDA: CPU +440, GPU +0, now: CPU 452, GPU 5862 (MiB)\n", - "[07/25/2022-16:46:44] [I] [TRT] [MemUsageSnapshot] Begin constructing builder kernel library: CPU 452 MiB, GPU 5862 MiB\n", - "[07/25/2022-16:46:44] [I] [TRT] [MemUsageSnapshot] End constructing builder kernel library: CPU 669 MiB, GPU 5934 MiB\n", - "[07/25/2022-16:46:44] [I] Start parsing network model\n", - "[07/25/2022-16:46:44] [I] [TRT] ----------------------------------------------------------------\n", - "[07/25/2022-16:46:44] [I] [TRT] Input filename: models/mobilenetv2_qat.onnx\n", - "[07/25/2022-16:46:44] [I] [TRT] ONNX IR version: 0.0.7\n", - "[07/25/2022-16:46:44] [I] [TRT] Opset version: 13\n", - "[07/25/2022-16:46:44] [I] [TRT] Producer name: pytorch\n", - "[07/25/2022-16:46:44] [I] [TRT] Producer version: 1.13.0\n", - "[07/25/2022-16:46:44] [I] [TRT] Domain: \n", - "[07/25/2022-16:46:44] [I] [TRT] Model version: 0\n", - "[07/25/2022-16:46:44] [I] [TRT] Doc string: \n", - "[07/25/2022-16:46:44] [I] [TRT] ----------------------------------------------------------------\n", - "[07/25/2022-16:46:44] [W] [TRT] parsers/onnx/onnx2trt_utils.cpp:506: Your ONNX model has been generated with double-typed weights, while TensorRT does not natively support double. Attempting to cast down to float.\n", - "[07/25/2022-16:46:44] [W] [TRT] parsers/onnx/onnx2trt_utils.cpp:368: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.\n", - "[07/25/2022-16:46:45] [I] Finish parsing network model\n", - "[07/25/2022-16:46:45] [I] FP32 and INT8 precisions have been specified - more performance might be enabled by additionally specifying --fp16 or --best\n", - "[07/25/2022-16:46:45] [W] [TRT] Calibrator won't be used in explicit precision mode. Use quantization aware training to generate network with Quantize/Dequantize nodes.\n", - "[07/25/2022-16:46:47] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +838, GPU +362, now: CPU 1543, GPU 6342 (MiB)\n", - "[07/25/2022-16:46:47] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +128, GPU +58, now: CPU 1671, GPU 6400 (MiB)\n", - "[07/25/2022-16:46:47] [I] [TRT] Local timing cache in use. Profiling results in this builder pass will not be stored.\n", - "[07/25/2022-16:47:09] [I] [TRT] Detected 1 inputs and 1 output network tensors.\n", - "[07/25/2022-16:47:09] [I] [TRT] Total Host Persistent Memory: 82480\n", - "[07/25/2022-16:47:09] [I] [TRT] Total Device Persistent Memory: 2413056\n", - "[07/25/2022-16:47:09] [I] [TRT] Total Scratch Memory: 0\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageStats] Peak memory usage of TRT CPU/GPU memory allocators: CPU 11 MiB, GPU 184 MiB\n", - "[07/25/2022-16:47:09] [I] [TRT] [BlockAssignment] Algorithm ShiftNTopDown took 3.32319ms to assign 4 blocks to 84 nodes requiring 130056192 bytes.\n", - "[07/25/2022-16:47:09] [I] [TRT] Total Activation Memory: 130056192\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 1674, GPU 6412 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +0, GPU +10, now: CPU 1674, GPU 6422 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in building engine: CPU +2, GPU +4, now: CPU 2, GPU 4 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init CUDA: CPU +0, GPU +0, now: CPU 1665, GPU 6384 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] Loaded engine size: 2 MiB\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +10, now: CPU 1666, GPU 6398 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +0, GPU +8, now: CPU 1666, GPU 6406 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +2, now: CPU 0, GPU 2 (MiB)\n", - "[07/25/2022-16:47:09] [I] Engine built in 25.2523 sec.\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +10, now: CPU 1435, GPU 6322 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +1, GPU +8, now: CPU 1436, GPU 6330 (MiB)\n", - "[07/25/2022-16:47:09] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +126, now: CPU 0, GPU 128 (MiB)\n", - "[07/25/2022-16:47:09] [I] Using random values for input inputs.1\n", - "[07/25/2022-16:47:09] [I] Created input binding for inputs.1 with dimensions 64x3x224x224\n", - "[07/25/2022-16:47:09] [I] Using random values for output 1225\n", - "[07/25/2022-16:47:09] [I] Created output binding for 1225 with dimensions 64x10\n", - "[07/25/2022-16:47:09] [I] Starting inference\n", - "[07/25/2022-16:47:12] [I] Warmup completed 63 queries over 200 ms\n", - "[07/25/2022-16:47:12] [I] Timing trace has 976 queries over 3.0073 s\n", - "[07/25/2022-16:47:12] [I] \n", - "[07/25/2022-16:47:12] [I] === Trace details ===\n", - "[07/25/2022-16:47:12] [I] Trace averages of 10 runs:\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.92225 ms - Host latency: 5.03344 ms (end to end 5.05219 ms, enqueue 1.40172 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.66963 ms - Host latency: 4.78574 ms (end to end 4.80028 ms, enqueue 1.39754 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61669 ms - Host latency: 4.73438 ms (end to end 4.75002 ms, enqueue 1.40104 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59776 ms - Host latency: 4.70923 ms (end to end 4.72325 ms, enqueue 1.40551 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59918 ms - Host latency: 4.715 ms (end to end 4.72859 ms, enqueue 1.39258 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59212 ms - Host latency: 4.70311 ms (end to end 4.71815 ms, enqueue 1.40127 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59171 ms - Host latency: 4.70111 ms (end to end 4.71709 ms, enqueue 1.3924 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.58884 ms - Host latency: 4.69999 ms (end to end 4.71507 ms, enqueue 1.38793 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59119 ms - Host latency: 4.70641 ms (end to end 4.72385 ms, enqueue 1.39411 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.58546 ms - Host latency: 4.70263 ms (end to end 4.7179 ms, enqueue 1.39454 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.58618 ms - Host latency: 4.69799 ms (end to end 4.71401 ms, enqueue 1.38189 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59365 ms - Host latency: 4.70694 ms (end to end 4.72247 ms, enqueue 1.40284 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59426 ms - Host latency: 4.70533 ms (end to end 4.71981 ms, enqueue 1.40167 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59406 ms - Host latency: 4.70507 ms (end to end 4.72038 ms, enqueue 1.39868 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59302 ms - Host latency: 4.70604 ms (end to end 4.72096 ms, enqueue 1.39022 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.5956 ms - Host latency: 4.70856 ms (end to end 4.72499 ms, enqueue 1.39016 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59622 ms - Host latency: 4.71029 ms (end to end 4.72501 ms, enqueue 1.39351 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59784 ms - Host latency: 4.70826 ms (end to end 4.72278 ms, enqueue 1.39263 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59805 ms - Host latency: 4.71088 ms (end to end 4.72592 ms, enqueue 1.39367 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59795 ms - Host latency: 4.71144 ms (end to end 4.72837 ms, enqueue 1.3975 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59713 ms - Host latency: 4.70555 ms (end to end 4.72311 ms, enqueue 1.40206 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59601 ms - Host latency: 4.68881 ms (end to end 4.70304 ms, enqueue 1.3645 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.58742 ms - Host latency: 4.69799 ms (end to end 4.71174 ms, enqueue 1.39108 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59344 ms - Host latency: 4.70665 ms (end to end 4.72214 ms, enqueue 1.39278 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59734 ms - Host latency: 4.70482 ms (end to end 4.71854 ms, enqueue 1.39332 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59714 ms - Host latency: 4.70997 ms (end to end 4.72628 ms, enqueue 1.40047 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61176 ms - Host latency: 4.72535 ms (end to end 4.7418 ms, enqueue 1.39706 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61494 ms - Host latency: 4.72816 ms (end to end 4.7448 ms, enqueue 1.39434 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61383 ms - Host latency: 4.72913 ms (end to end 4.7439 ms, enqueue 1.40642 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61697 ms - Host latency: 4.73928 ms (end to end 4.75625 ms, enqueue 1.41578 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61782 ms - Host latency: 4.83635 ms (end to end 4.85382 ms, enqueue 0.316187 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61688 ms - Host latency: 4.81012 ms (end to end 4.82694 ms, enqueue 0.524707 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62682 ms - Host latency: 4.69824 ms (end to end 4.71261 ms, enqueue 1.44248 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62582 ms - Host latency: 4.68247 ms (end to end 4.69834 ms, enqueue 1.57075 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62538 ms - Host latency: 4.68074 ms (end to end 4.69913 ms, enqueue 1.56764 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62548 ms - Host latency: 4.68276 ms (end to end 4.69795 ms, enqueue 1.58025 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62765 ms - Host latency: 4.68287 ms (end to end 4.70229 ms, enqueue 1.56355 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62581 ms - Host latency: 4.68279 ms (end to end 4.69857 ms, enqueue 1.57596 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62439 ms - Host latency: 4.68186 ms (end to end 4.69902 ms, enqueue 1.56841 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62468 ms - Host latency: 4.6818 ms (end to end 4.69666 ms, enqueue 1.57666 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62562 ms - Host latency: 4.68257 ms (end to end 4.6985 ms, enqueue 1.57379 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61575 ms - Host latency: 4.67201 ms (end to end 4.68948 ms, enqueue 1.58751 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61467 ms - Host latency: 4.67125 ms (end to end 4.68734 ms, enqueue 1.57214 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.6139 ms - Host latency: 4.66783 ms (end to end 4.6828 ms, enqueue 1.56377 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61342 ms - Host latency: 4.67017 ms (end to end 4.68673 ms, enqueue 1.57308 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61005 ms - Host latency: 4.66664 ms (end to end 4.68411 ms, enqueue 1.55513 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.59465 ms - Host latency: 4.65076 ms (end to end 4.66672 ms, enqueue 1.56719 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.5959 ms - Host latency: 4.65466 ms (end to end 4.66882 ms, enqueue 1.5709 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.60471 ms - Host latency: 4.66272 ms (end to end 4.68046 ms, enqueue 1.58149 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61157 ms - Host latency: 4.66888 ms (end to end 4.68478 ms, enqueue 1.62261 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61403 ms - Host latency: 4.66865 ms (end to end 4.68436 ms, enqueue 1.61089 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61339 ms - Host latency: 4.66898 ms (end to end 4.6855 ms, enqueue 1.59581 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61229 ms - Host latency: 4.66919 ms (end to end 4.68688 ms, enqueue 1.57114 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61361 ms - Host latency: 4.67148 ms (end to end 4.68864 ms, enqueue 1.57201 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61329 ms - Host latency: 4.66671 ms (end to end 4.6823 ms, enqueue 1.56505 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61117 ms - Host latency: 4.66793 ms (end to end 4.68323 ms, enqueue 1.58344 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61322 ms - Host latency: 4.67312 ms (end to end 4.68901 ms, enqueue 1.57474 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61351 ms - Host latency: 4.6689 ms (end to end 4.68566 ms, enqueue 1.57411 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.6125 ms - Host latency: 4.67083 ms (end to end 4.68839 ms, enqueue 1.56761 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61216 ms - Host latency: 4.66829 ms (end to end 4.68427 ms, enqueue 1.57145 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61221 ms - Host latency: 4.66812 ms (end to end 4.68464 ms, enqueue 1.57742 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61724 ms - Host latency: 4.67236 ms (end to end 4.69009 ms, enqueue 1.58645 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.6342 ms - Host latency: 4.69334 ms (end to end 4.70886 ms, enqueue 1.58391 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.64475 ms - Host latency: 4.70205 ms (end to end 4.71633 ms, enqueue 1.57148 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.64463 ms - Host latency: 4.70203 ms (end to end 4.71699 ms, enqueue 1.56494 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.64092 ms - Host latency: 4.69741 ms (end to end 4.71147 ms, enqueue 1.57456 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62642 ms - Host latency: 4.68474 ms (end to end 4.70034 ms, enqueue 1.56938 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62737 ms - Host latency: 4.68528 ms (end to end 4.70254 ms, enqueue 1.57288 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62422 ms - Host latency: 4.68096 ms (end to end 4.69629 ms, enqueue 1.58088 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62236 ms - Host latency: 4.67939 ms (end to end 4.69592 ms, enqueue 1.56531 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61946 ms - Host latency: 4.67705 ms (end to end 4.69207 ms, enqueue 1.57915 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62383 ms - Host latency: 4.68113 ms (end to end 4.69565 ms, enqueue 1.56628 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62493 ms - Host latency: 4.68076 ms (end to end 4.69827 ms, enqueue 1.57712 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62881 ms - Host latency: 4.68533 ms (end to end 4.70332 ms, enqueue 1.59106 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62705 ms - Host latency: 4.77595 ms (end to end 4.79063 ms, enqueue 1.23335 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.63042 ms - Host latency: 4.83225 ms (end to end 4.84863 ms, enqueue 0.584692 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62356 ms - Host latency: 4.80049 ms (end to end 4.81941 ms, enqueue 0.722852 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61289 ms - Host latency: 4.70488 ms (end to end 4.72126 ms, enqueue 1.16353 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61414 ms - Host latency: 4.67012 ms (end to end 4.6865 ms, enqueue 1.55625 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61272 ms - Host latency: 4.66924 ms (end to end 4.68572 ms, enqueue 1.57039 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61147 ms - Host latency: 4.66743 ms (end to end 4.6821 ms, enqueue 1.57139 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61204 ms - Host latency: 4.66624 ms (end to end 4.68369 ms, enqueue 1.57068 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61245 ms - Host latency: 4.67002 ms (end to end 4.68525 ms, enqueue 1.56729 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61497 ms - Host latency: 4.67256 ms (end to end 4.68835 ms, enqueue 1.5822 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61396 ms - Host latency: 4.6707 ms (end to end 4.6873 ms, enqueue 1.56724 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61487 ms - Host latency: 4.67173 ms (end to end 4.68682 ms, enqueue 1.57334 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61299 ms - Host latency: 4.66936 ms (end to end 4.68381 ms, enqueue 1.57117 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61013 ms - Host latency: 4.66755 ms (end to end 4.68381 ms, enqueue 1.57551 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61992 ms - Host latency: 4.67517 ms (end to end 4.69097 ms, enqueue 1.58848 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62769 ms - Host latency: 4.6877 ms (end to end 4.70227 ms, enqueue 1.57029 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62732 ms - Host latency: 4.68355 ms (end to end 4.70088 ms, enqueue 1.56836 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.62974 ms - Host latency: 4.6852 ms (end to end 4.69971 ms, enqueue 1.56511 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61794 ms - Host latency: 4.67524 ms (end to end 4.68911 ms, enqueue 1.57212 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.61436 ms - Host latency: 4.6708 ms (end to end 4.68591 ms, enqueue 1.5667 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.60833 ms - Host latency: 4.66543 ms (end to end 4.68132 ms, enqueue 1.57961 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.6125 ms - Host latency: 4.66885 ms (end to end 4.68545 ms, enqueue 1.56494 ms)\n", - "[07/25/2022-16:47:12] [I] Average on 10 runs - GPU latency: 1.63328 ms - Host latency: 4.69219 ms (end to end 4.70671 ms, enqueue 1.57573 ms)\n", - "[07/25/2022-16:47:12] [I] \n", - "[07/25/2022-16:47:12] [I] === Performance summary ===\n", - "[07/25/2022-16:47:12] [I] Throughput: 324.544 qps\n", - "[07/25/2022-16:47:12] [I] Latency: min = 4.63513 ms, max = 5.62218 ms, mean = 4.69772 ms, median = 4.68481 ms, percentile(99%) = 4.86353 ms\n", - "[07/25/2022-16:47:12] [I] End-to-End Host Latency: min = 4.64392 ms, max = 5.64146 ms, mean = 4.71364 ms, median = 4.70197 ms, percentile(99%) = 4.88013 ms\n", - "[07/25/2022-16:47:12] [I] Enqueue Time: min = 0.310181 ms, max = 4.23633 ms, mean = 1.46804 ms, median = 1.5567 ms, percentile(99%) = 1.67847 ms\n", - "[07/25/2022-16:47:12] [I] H2D Latency: min = 3.01538 ms, max = 3.23657 ms, mean = 3.06713 ms, median = 3.05371 ms, percentile(99%) = 3.20923 ms\n", - "[07/25/2022-16:47:12] [I] GPU Compute Time: min = 1.578 ms, max = 2.49139 ms, mean = 1.61667 ms, median = 1.61377 ms, percentile(99%) = 1.69678 ms\n", - "[07/25/2022-16:47:12] [I] D2H Latency: min = 0.00561523 ms, max = 0.0319824 ms, mean = 0.0139259 ms, median = 0.0134277 ms, percentile(99%) = 0.0289307 ms\n", - "[07/25/2022-16:47:12] [I] Total Host Walltime: 3.0073 s\n", - "[07/25/2022-16:47:12] [I] Total GPU Compute Time: 1.57787 s\n", - "[07/25/2022-16:47:12] [W] * Throughput may be bound by Enqueue Time rather than GPU Compute and the GPU may be under-utilized.\n", - "[07/25/2022-16:47:12] [W] If not already in use, --useCudaGraph (utilize CUDA graphs where possible) may increase the throughput.\n", - "[07/25/2022-16:47:12] [W] * Throughput may be bound by host-to-device transfers for the inputs rather than GPU Compute and the GPU may be under-utilized.\n", - "[07/25/2022-16:47:12] [W] Add --noDataTransfers flag to disable data transfers.\n", - "[07/25/2022-16:47:12] [I] Explanations of the performance metrics are printed in the verbose logs.\n", - "[07/25/2022-16:47:12] [I] \n", - "&&&& PASSED TensorRT.trtexec [TensorRT v8205] # trtexec --onnx=models/mobilenetv2_qat.onnx --int8 --saveEngine=models/mobilenetv2_qat.trt\n" - ] - } - ], - "source": [ - "# Set static member of TensorQuantizer to use Pytorch’s own fake quantization functions\n", - "quant_nn.TensorQuantizer.use_fb_fake_quant = True\n", - "\n", - "# Exporting to ONNX\n", - "dummy_input = torch.randn(64, 3, 224, 224, device='cuda')\n", - "input_names = [ \"actual_input_1\" ]\n", - "output_names = [ \"output1\" ]\n", - "torch.onnx.export(\n", - " q_model,\n", - " dummy_input,\n", - " \"models/mobilenetv2_qat.onnx\",\n", - " verbose=False,\n", - " opset_version=13,\n", - " do_constant_folding = False)\n", - "\n", - "# Converting ONNX model to TRT\n", - "!trtexec --onnx=models/mobilenetv2_qat.onnx --int8 --saveEngine=models/mobilenetv2_qat.trt" - ] - }, - { - "cell_type": "markdown", - "id": "b5108ef4", - "metadata": {}, - "source": [ - "\n", - "### 6. Evaluation and Benchmarking" - ] - }, - { - "cell_type": "markdown", - "id": "2e5362ca", - "metadata": {}, - "source": [ - "Now, we have converted our model to a TensorRT engine. Great! That means we are ready to load it into the native Python TensorRT runtime to perform inference and evaluate our models." - ] - }, - { - "cell_type": "code", - "execution_count": 23, - "id": "790d73a6", - "metadata": {}, - "outputs": [], - "source": [ - "# Import needed libraries and define the evaluate function\n", - "\n", - "import pycuda.driver as cuda\n", - "import pycuda.autoinit\n", - "import time \n", - "\n", - "def evaluate_trt(engine_path, dataloader, batch_size):\n", - " \n", - " def predict(batch): # result gets copied into output\n", - " # transfer input data to device\n", - " cuda.memcpy_htod_async(d_input, batch, stream)\n", - " # execute model\n", - " context.execute_async_v2(bindings, stream.handle, None)\n", - " # transfer predictions back\n", - " cuda.memcpy_dtoh_async(output, d_output, stream)\n", - " # syncronize threads\n", - " stream.synchronize()\n", - " return output\n", - " \n", - " with open(engine_path, 'rb') as f, trt.Runtime(trt.Logger(trt.Logger.WARNING)) as runtime, runtime.deserialize_cuda_engine(f.read()) as engine, engine.create_execution_context() as context:\n", - " total = 0\n", - " correct = 0\n", - " for images, labels in val_dataloader:\n", - " input_batch = images.numpy()\n", - " labels = labels.numpy()\n", - " output = np.empty([batch_size, 10], dtype = np.float32) \n", - "\n", - " # Now allocate input and output memory, give TRT pointers (bindings) to it:\n", - " d_input = cuda.mem_alloc(1 * input_batch.nbytes)\n", - " d_output = cuda.mem_alloc(1 * output.nbytes)\n", - " bindings = [int(d_input), int(d_output)]\n", - "\n", - " stream = cuda.Stream()\n", - " preds = predict(input_batch)\n", - " pred_labels = []\n", - " for pred in preds:\n", - " pred_label = (-pred).argsort()[0]\n", - " pred_labels.append(pred_label)\n", - "\n", - " total += len(labels)\n", - " correct += (pred_labels == labels).sum()\n", - " \n", - " return correct/total" - ] - }, - { - "cell_type": "code", - "execution_count": 24, - "id": "f3fd416f", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mobilenetv2 TRT Baseline accuracy: 71.13%\n" - ] - } - ], - "source": [ - "# Evaluate and benchmark the performance of the baseline TRT model (TRT FP32 Model)\n", - "batch_size = 64\n", - "test_acc = evaluate_trt(\"models/mobilenetv2_base.trt\", val_dataloader, batch_size)\n", - "print(\"Mobilenetv2 TRT Baseline accuracy: {:.2f}%\".format(100 * test_acc))" - ] - }, - { - "cell_type": "code", - "execution_count": 25, - "id": "a5ec3a81", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mobilenetv2 TRT PTQ accuracy: 68.11%\n" - ] - } - ], - "source": [ - "# Evaluate the PTQ model\n", - "batch_size = 64\n", - "test_acc = evaluate_trt(\"models/mobilenetv2_ptq.trt\", val_dataloader, batch_size)\n", - "print(\"Mobilenetv2 TRT PTQ accuracy: {:.2f}%\".format(100 * test_acc))" - ] - }, - { - "cell_type": "code", - "execution_count": 26, - "id": "eb95977d", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mobilenetv2 TRT PTQ accuracy: 70.31%\n" - ] - } - ], - "source": [ - "# Evaluate the QAT model\n", - "batch_size = 64\n", - "test_acc = evaluate_trt(\"models/mobilenetv2_qat.trt\", val_dataloader, batch_size)\n", - "print(\"Mobilenetv2 TRT PTQ accuracy: {:.2f}%\".format(100 * test_acc))" - ] - }, - { - "cell_type": "markdown", - "id": "20c82807", - "metadata": {}, - "source": [ - "Compared to the TRT FP32 model, we observe a speedup of ~3.7x with only a ~0.8% loss in accuracy. " - ] - }, - { - "cell_type": "markdown", - "id": "52f311fb", - "metadata": {}, - "source": [ - "\n", - "## 7. Conclusion\n", - "We put together all the observations that were made in this notebook. Note that, these numbers can vary with every run due to the stochastic nature of the training process, but a similar pattern can still be noticed.\n", - "\n", - "| Model | Accuracy | Performance |\n", - "| ------------------------ | -------- | ----------- |\n", - "| Baseline MobileNetv2 | 71.11% | 11.92ms |\n", - "| Base + TRT
(TRT FP32) | 71.13% | 5.95ms |\n", - "| PTQ + TRT
(TRT int8) | 68.11% | 1.59ms |\n", - "| QAT+TRT
(TRT INT8) | 70.31% | 1.61ms |" - ] - }, - { - "cell_type": "markdown", - "id": "91dfc2c1", - "metadata": {}, - "source": [ - "\n", - "## 8. References\n", - "* Very Deep Convolution Networks for large scale Image Recognition\n", - "* Achieving FP32 Accuracy for INT8 Inference Using Quantization Aware Training with NVIDIA TensorRT\n", - "* Pytorch-quantization toolkit from NVIDIA\n", - "* Pytorch quantization toolkit userguide\n", - "* Quantization basics\n", - "* TensorRT Developer Guide" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3 (ipykernel)", - "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": "3.8.13" - }, - "vscode": { - "interpreter": { - "hash": "b8290132a159428f0004735847c0b4016c8a5153e62fd80cc71ad5cd485f05b0" - } - } - }, - "nbformat": 4, - "nbformat_minor": 5 -}

`%?~P7j=3*XYVFoPRSmbr(=I^8jTKfd6Qw8Sur5Zjn)@kCIb(XaSlESW_QpemESR$h)h?uzpP(&z}@_*NgjC z#2T~j=Bu`BQLlT~jDph#pO)Z6PgG@NNEvV!E!@g_jFs)~C{K3p3`w$5xZLDGp4Qdy zq+mjS0T)qCzn0VNJSazG1jv8Rl#Y&>cc_KLabh*@faK%)nGO6RPH=|4bIztqfDU_2 zlP9itifYK{NN#hAdaTG0)UccHN6(%17Kr*}sLRSAuya~e=-xnH@EMTpCrR6pLhkLI z5El~^rEi9F+9?i)=7jzU#6JnBKM}_|J~ltCMg}TfO4+tE{s31*UYtxV9V4S5WtYWz z`t~+bDB<6{bfbAF&JuizPtA(y}7cW>l*(z*3ou~Dp!Fy39ovt?y(rV zsX#L4Z~7^bB4b1K#|b1{7R)1ExkZ*RU6T4I49fb8G5rYKxUPOu!(sR1+dq>;Z0*ml%yZuYcz7gg z9{AVMEy5a;9IV`e9~;UOG_?af~Hn!dqz8o(~SS_15Isn^h>B| z#1WbL5{QMgbz2|`Vx}(RfWjznKmZRa?t2VA{_D1=sZrl>`88zX{jL`yxL0V@Lqb8Y z6V(u^Y^2oIJX&RM+a*%c9jRFjg;i!59A~}Tlh3loK$cjgMQ8E|Q0c1L9pf*KFVg)9%To-Qu0!5-Uw;dnd zK;k`lAVtULK*^Jb={s4d<%pJQ0HLNdKL72XvfJ!i$ulT6M{vv6j;+P%dyD@=Lb(`o zH9{sJ;naNa55y|t37QxAxmJMtb;uTyTpT@XsB;B-hMl)e!c#8!azEU%7Y3!SOBfAt zKM_Mk>e7^Y$#E?o;rf(1MRTLMoQ6F+WrM(KHD0J4scAJ0o*qQTzgwbHnSSXMTkIU9 z=Vj=;>1`m{{8z1qk7aH) zNnerE2OsJP#@1^&k1TgQ6O%9`R)6E~j%73CdvBlmh-D80inIFtb;5KifQAV96)02{ zCQb-Wcpv8EvHJ#tK7`B{v~v&?ptP~XeMW_bqML8>>jn_&X&QVTN*&y|^*xpqq@L>C z^aD$X4+;6>@!yfntnIrWl;vo-{#-7a>=wcE_+>r}$W09O1Vo|tfvnMBN?~W6FM$T$ zH6~ecb!k@|>HGuyeea&tfh`Y{&*4|PcdYkcq``wo^h;k-48OTec6bs<`L!zg_))y- z5|74E|5RBRozV#(E|tge6f0=;Nx}lrIUb6i zf-)+%yd_gS-D;f1kgvH@f}O%lo(e|S#pcL8otY{PF%DOKdmnyxhPw|c+fgMDPq+d4 z53!waM|<5_Q{{Kt>3`<>3vlMCO^Ep8RWWz?Vh<(^C9y4m*6P`L?yldt-@>uO91=mF z{h_8Wy%Ao{3g5J?Pd8^K6%kjBkCTCHdKLq{X-Kg4_Yp^PUn; zp{0w4U-}-gACz3onTD%0iNq;h+*-!xAb9h zv#5Qf_Z5M{;+s$!aqA)QmP{p!eg6TNp|{A4QAbTLJC0{^kE9q*Xe2>JslI1>js&)B zpLma{Sya7ugAT2qT{q(|o(5m(MQm-;b`~qphjC^<5Qa`W(Jz+10@%)yrKgMTdlb#P z3T6N$w^%6GUcMdXzHk|N2ad)760CeZxj+7c_215Z7m>aAO^)10^j1I){{!lZBiqr0 zzQUb=T1s#fK0y$H0OdfSx&P0p{>}~gmdMj{fj4xpdnDS@?O0{otQ&IF)D29zP&$E0 z{0z2l_vh&IwTY^|fMc}w&&*6gTo9yWf_@eI7j6?OVX`MlpnS_GrIyvL>H$rm2s0*6BKeoF9Y=skTHpZ^-wb7y*RucnmlnMQWS1Gk@~gc^40WYg7dWmv1Magr))1m3VrBi zI&3S-`t}heTQKP$wIbKHj-;Mim@P%1NeLP{mNn6agp^-|Kj+1@lsCZ=i1<7O!rn~1 z5o!O}XT97*z>H;>2-L)FD~G1MM}f01{bd#U?kRNXi|1vywOhdZKlkW>EBpOu+13Sdsv)&R}?HyJzr2 zxaHsD7hj)@)!w%XR_mDJbvHn`-kdrJx=u@+PO8w)i;txrcB015$NvS8SAA`YW`pxf zr8W>w)g&q~aPk-{@%QM&))MH)a(ts4F}JILWK{m06+ip5+lK8C0NeU7%&nv-1^qr+ zLVk+{13a&lMMZUfY`RM&PY^?cBCN?)9JX7=#TI{t4Jp8HA;qf~U(58_?C`cDV}E6! zl_St;X<9}fclt4qpXa$fD1e$yivGrB^|Mf*T!#)YGoNp^f>?=IkN7{mN}7;5YwGh{ z(QhPy#D#sdeQwUtPj~%eKMZ+?MZt6XtmZ^-=Cc#WJ3CL`xQ!oeAkWHKAc-DYIRyi^ zS2mQcnX* z{6m=xgS}h%51^h-LH)@&`jaYz|C)4|P?5Qp)>5WGIYGEb^c#7UQWK$UHUl!hI=V6J zxEo1pU=PX5%NSSJB?**1XlD{fRS>Ahvttzcfpu{K%|=wnw1TCj|Az+yg%NaIVJ6 z3f)w@_Gk#`V5sm*dsRX8MAG6}_rBAt`6?wt-5{bT-5bz7-`3a8tMW(=HZuMdFkgM5 zOrvq9aOC$R{z77>8KMWyPsl>0Muwk5T5H3Pvmhkw4-|TzoiDe2-u&|ygsk0t19we_ z{FY6`*C+3VV_p!Q2DoEOpwzJTGM(XLo~xzp*s+Fl`k9|kGl(lv50MOY&=|c-PD-zR zRI1>tLYLPG4$P3*&aB5Nn}dadXq4!=xqbD5Jx{fwnEx?NZGd2F1!+(QcO0+QigVUo zWHW^nq!u^gX?6C>98!o{PQT()XAmd%mVkoFlHln5sZR9Gt`e45iFEn&3WVOq?2J8u zcL7m2p-v%($G}}3HCcR%4~woigYNB)uZce)SE<{zdd2q>2*rU^97s&6Ctrv!McLXC zUPnQ$fsbA-y8B`-^HB`M#DtO*lgJuN1}q?;qMiPUXo%OAzTFDU`O!DWq+|_MlwMw9 zoL!KOF3iV?W*uc3Qq66MeKGWdFg zZh?tyIm{pjn@sAR4jNR_q|0$Zygta@XtEdgw+?DM+WV53s*3`ck!H zQjqbl+6$f2`>kF)1ImDJ(%YSqREU|75~Q9m%ehAkkw=R0_7!jI&}`u5V(%5SHSbNp z`?>>QV5;-vs?Xlb1VX&s1xt&w60Z8`M(NxYUxpOqNW(m}7rM>+AJ*J=H37!#05UY` z^fR zo454v>Auc^WXXBTbLjr+8vQ<#yBxKf(?XE$qpBi3wjrUw1fq+bx`}(54SP+ifu}7z z@Ep6-uLP^af*7R2&{=co34^axuqZ5U!(}KHqds&?0Go`G#-g8vb!if#UHKAsSA)m1^kSt1DMBy$Qei%x0m@9i zLSKj^tJY1Fi_a2Q=%p%f54|gyJ8Mx-lPvi-Xkf6$LF6Yg^;1K8M^a%SW4L-|VfoK= zOOO>3)A@(%L?UM;!iIHu~N~t`|)AUW@>2jdc&&B2%oPFwr z8(77#Jy2Ah0zuy}xKA7y@pt$vC4o&s5c2A&o*LGd(aU_gFwUqwo4)5YG1IYETtYypJL=oZyLdAQI?GB+|e78O-M6bD#m{F{c`a<%^Mz3&t_$#E;I7I=T+a zkq0w2Yynq8!PVdG#xOFm=g`_dZ8=gh|95c5c9pjq$`{Y@P>}YVXzfp`YuCecicldR z3BqG4O><8yL|iNc8CV2@CSEzh*XYiqz0T{L+6N0b;C~|iSd<^z5?X1rDZMkIR=|gk57;O2Ed;7zX#rO~sglL0Ke2j`8 zXeL775U~Lu3WAQ{)TPVg^=qY*4x&hYjRo{3LDX{h*ZzK+OHJmQkWU^z(L&L|-DU3i za6Spn?+wj9(@nAlVq{khT8cCKeplf4BSK|*2Nc$z7iWjXd?HX7hO46Qr9@j)~T^eRjb()sz28pg*dmAmix`?zc-7<*k?K&tFKRMQUk zl&=zEg4N6^I3s0PSv^iKOor*&cCyrP{)W2YhT?h?+_wf(&c(u5NKwLPtJ}}K73lqA%A$Mtd&ZAcmWzRSwXXx z{QHECy{5Ddq0!V&ktm8X!fi)F&w5kq1{dSJTt5TXq2BB?koRzD@H8xd%CUN?c!E<* zCMbul%~msJc|>;k9}51LBBe5S`$9qvNY@>vmY>0tqGm! zH!xSKfXibd-Xhy=RB6brGte-Biuo?hR zG?=%DRZ3R2W&XS3qV~J(DcTc`zhFm*1JSU)X0aw1sqDo8k^HK0r|*<#&9Hh62=quxR(l&vBT zm$u)(#~_dt34{rc@li#|&EFj=Xd zZ>CayIYU_hjJs0(easFMWwHUar7jmH+3Ts@X-^H`fxQHmtcAGE{%O?`Y61&M%bQdo z8r8L8qiD~{w9TkH%1QzF=MW{m8)JjTG+u4it}4G63cfq5)j7hQXU(AkWh^u-#huEresL;7Ld@t?g#h9=STP&sXJY3n?%r2x>RNPUNakW^u;@2j2br=W{eCVP0YxsC3rNGC z=MXVqX$g0sjR&M<;7s_n;Q?TOiag-ZEC za@vpG0bFB^aeHlH@h3fILWRxZgJ109a=gow%aliDsI};{uhD4XFM8g9QH@LI0+Z@n zDJdb}t|FguiS;qN}OyBJRw`kU-U?wr}7$-xh;;DLAL+?08HdxO?Iaya<{c? zP-7(O!BETeOCW0x-~=f^MlkBAM{J3WO1+`J{xJ>TaSB7BoK(T(et~d~eogyn98?2q zg`EYfDoSt(1iBOfT4PHd2x*hQ1c;Kv*0jJ#zo)Xb)6nrXyb_yvNz4?_R(4L91I;9U zN)(=zaS~Pj>T9w(zKpgzO*(MwAa~GvyIjPntIMY}h;w{uQB9!HY*;9Fo7f9esQe-1 zu_XSi3qD}JPEnNV^s=!Dv@gfd~-wge|MI!RZ7 zMP^>%h$Ie6(jvm6iByXMiPLvdSz_T@Jp&kLX-D+AkG;C1S`v;`D~m63$xI9O%rsq* z1{#8gF?c;+81$Lt{|N$lR>7&Pv3vU#wL=jN5!vSu>%M zTUOHA8gU?R#ZM2E@7{#Yk)3j8f^=`jq2cnsSyi1r*eyhgJ((bRlKbmdj4b@IE?*$r za_9aPd8_%~H#0}$MgPxnEjoa7?aFp{o(z_yRBBtKbU&Cc54PJx@kQU04Ub$7-wi{e z_p+s7i`!|b22Oy=GJVEch)VML3yNl=J6lQVaKJxbVg&5o4X@h@VXbAX+1Ya(! zS~ecvnWla;JM_@^r-Q-f!kN?#QElnPsv?uP2cY*lEkD3xIgUTB-6EgjWa@uUaq1){ z>IFbB*pZoF;o1~e0Q}-SQ?;RyXaD`{jH6ZNGwCVVPc!7MPX_6;bkfcOdxUS1ro7k%8Zi`h_s%gSmPy#&LlTl# z3zPvgw2)547`&#F=71xaL%%Qkfm#_v(9+@fCs)$ZPVflg@_X9mvr(GJv9EASSu#+N zrM^%xl5g?De_o8vhmR|+0F%G}d;M;^$zJ5lLX-U5>*AxQYpu+#x7l&INPt;GXCD`zy5pHOfpNkOx_wE zZ1zir)(Edg=P2FUR0wS#X%^!{L;JnnNcHco%nPmcn^)gq3zy#?^`ak;vwn5k)qQj= zpG%d5idT+TT8VvDPi1_7OTs+p5xgpVLiL8_8QM^#Wnj=!ouZ1_$J&s#OGnN3UDlnH z-Y0ma(-#gwrONuJu*|S#D zl>C&4S*tR3ThdH!1{BtLkm6jk^7r(HBssqk;WUQ-ju+B;r$`_noKl&6jlcRr>fN9}tyJwCo z@`YJIvMdXM(*{0;7zSlL4=U^JXE61{IW4^4%Xk@}j#*FqOemJD)G9v6G5ALRwj>=n~h_1UEHaZnfbh1*-{&g!g@g=M5dt1r*gFR`E__}TDHnX7nhbw`_ z^0mW{Oxlm46&$JO!XoG+EtxHZ%mJPef$K&QQeIB$&*FL?JBsM=I_BN<3ZA@)J-rw; zS>k!v)tY(_5no?6U|vumR4K9A`AN9YUdi((Mk|m>ixfZIot}?xjjA-Q#;A#l^Kn^h ztr}=qnVNO+#%LF5yk_id4miDj2+Sku?ae%pJ>>m&=}DXkrxFZ zf?&+e1q5o9<8Yi+&#Z_qRD}InvyKtD<9JF`-Fs9Cjc0i`@az z(R}P)`Eyyab(15UMotD8@lak z8_a;mz6V=_?%zUJDC)22PJdfVKkQ5rDFT+W%-t`a8a0QxGZ2Gwlcz{!sp?1@)8pnr z!1K0VebGMzkBW{E*g{}Jp_x-NfD4<+5?;!Epvw3@x_VhZ@5X8*qV4a4zzv%XzN| zVf3gdl}m?we%s4M_8cpg%wLd?OHtO~M!*?Ot`LV9o07y%#%h6JxIO6kPphKdUp0P@ z2!9tjW8o)@6^5A=#oROS%r%Z$_1!IV@GlD>q?5}`sL#$fcNRgQ+i$owzM7?K;ZWi_ z$GA55JtJfQwvq0yTJXLU`~Q+}i8sKOqSCTm^DAV%$OJm+s!#z4Q#j^9C} zvi6p4LhM%GzDN-wDriWNlC z{23^RDmdSp8}H`qj)hei?ixC-GyU9wOIW1|hF5gS>u1I0=u%%jCRYIAf}I1a zlx_lXHd)~mHC&>@vuuwF$T5;_%`s?tDqd`-AE!hOe-a`bCtX-I38y0&gyVmBL6`H6 zQ88xvMO4GEFdJf6Ai{qQU{IYimi~#`hmy|-4bWS3o9$RdiA3*d%d--8k9_*%``!f+ zcMHPVx{TW(?d|%9%Tsi(5cKfX@GbOLY$S8yxfMTwfT-3JeFL(QRzUt7<%ur=BJQ~E z(DXxRA(uXLh0vRAL5_O+*SuCnBWwSxYtXIQm0xg2?pAx;Fe*7?t{bB2N6lIrOT8kU0?x@tNYt$11B`Gc`NSu(LvW(>1Z z6es=^sCqWcIoi&h{(Iu3ICaT;rZCR;dF5~W6{2#!GP!?CtO}hb!py3SpV5#Gvw??j zu$L6^aL}rgC{30(+9BRquxXP^KPBTsZ7hV-Hjv8Z7hl(KcB|>KRxQ8g2ngDDZTNbN zdi;`gUG8mKQ>PBows7Q>2O@8_ovL)oSAa+LJBAGGM*=&-82SMQQ&BpD#c(w+-=tbK z6kKtTe0dj3AHpgG)RapFs_ojDGx1$oYjYYitb5Idbh@j)kscQQO>n-Sh&^v>uguE& z$4pX2L9z?G`v*B@Pk1|;zME$GWLtf;Y{aLyPr`vP=X9v35%D!b@TlfWt7TLy&uvO1 zKl(RjEbN?tcRa&}kO52a4lF^!_EU_a)bHn)!HDrP+K5{8N&1*U0OQ4O-ver=iZI?V z`uGHjOId)?+<-GZmW5p)Q-3ZHwoVMCm?_cr1Vn2HeyY>=vbS7DCLbF#sUIvhx^aPs zp3?Tqg}o+T!z4QCO`!8i1;8;Xg5*v!bQKe%j9_o91K5WN{N<*=A?B^CuKK=kAY~nE zpNBj&xr;q=K3*Ia7^k)-$zy79it@LYF2$eW3M@!y%__0P#4p0Ne)L?5s&vpjHDJ7R zaigP7uI~VpGH74YY{>WDnnb1UEqAm9>a;YLX3NtO=;S0!cZG&h0;(w-_B22IudoUN zCrNbn3c4TjSDsIF(DIP$DmLH(7?CG~FHkuxLPQQul#sEX2Z5_CKCvfBb&33Wa1YU< zb1?5ON@^UZU#v!qi~U}u9cI?)zw4!V1{^u29dHeFHbSrWSAS#Ka*_ey)uC2P5{5G_ zEcNNQ2~URl_|2alvoNFCO#DQb9s;1XMcW%5@Gf>pdor=J_hk8>)-BMtEAZkADr{`; z`Ry{&>`{rdh0Px4m(drY89-e>L9!c~i7%AncyhR8WuE6&-mc+t=uKejFi~0d!(x>x z9H(8jocr}SjL+XR%=*U+lhI?nIJGCQAAI<9PNbxKcS;|Xh1}Xru#LAI7IxC?{Lp!ZWe$7rJYt;!tFQ~|>Q_?5m7q?4FggA)Fy_ph^_>SiiKA4J{PSd5+)mQ_DXk7Ax(^DU>bSh8y8M{D(SE69pFtvbZYWjmE}behi=0COAor zdsi?wVMoSEWm5Koa z8%m4d6(@ENOZ=N~rTmk`EciM2TUd5$Q|ul@R!L zX9jL>y%q!8R@Wk~h%wCDZTP?07NTbN4vkX0yc${UiFEJC;kB5Z^C~7!OJbBTLkcZU zx3cV>Bv5y=!R4-5y3w39HP11jwqO&6!e2MpR#P`+x`^#n&YqZ}!lSXj*+li7cpQ&T>}Hjj$%)z}X-RlU#9@ihF|`k?j2)PtCzu*PPngbEW)u)*Y6j zw?X+RtL|8#vtPnaH@mHb~{w8B)Fv znZsqoUn$v*T}MTC^)-?!P)J~tea=!b(D8uQko&FA%pm9KtHLK6RfS$xzaN=4fBkiX z!^UoHS4)`ounJ3uv#T+x&7C*;dk{NB&zb=bqE6%K6=y+%7OSX`x}hotdyPjlwXze2 zaj#65#S+KsC6oIInUWm$a~H3N(nX0@%8$uA$p+hmAIhcnY8sBd)JYwSSch&I_d}E# z@|I06sZH1E^s*5zFch8XDlp_e%dw7?yrh5-`(<2_B52tuRdZ%N8SejRuw*C_zeayQ z?1mNdJZPzM_J6YgcnB#}zBZZAlNS~+EsTq^m%K0~%yEC)P$bd!ly>=NO{s^Wv74SY zr|!5wkbJ_g1nSBLBOKA^g>D8RLVJtlvkfw^+P2cumN7%PbdN5)rU+rC|BKIQPU|o@ zCE^|J!>1F1j2V|R=MWth7#*w(dD=P;i-2{Qh4qUr;Sqi}1)XqME;0=uP*)FaS z*O9UI;rZ9U3`o6)VNb#qj=@;2>`HrxN9#{k?>w+@jQHkqfaL@}U78~3GUasX(R!63 z3a`0t!lU5t?j=6OcTxwG9s!Frj{lt?_11HBBB0GkHS1@)Xmf>#Au?loFY@Mi_x&(L_gHL0vwHXU<8WocR&g3cJ#|BBtPa2XbaEsm*on(8zA zBSNb@j{=bZwP;OxmRL5IAwq_kv!1BCkhs;)#WgBvs-twR(9QqF?T79`JjUXrjLXQh zHk~wmK57sV%2Hgjy+84AQXBCkMkUVJeJdkQOUnZBftyPINrH3Ib)D^BT-r-iRt8n> zF5k~NWX&1g$mFnFjc2+e?mOiFApa1UWJ~$+S?F)r-zegXil0_SP}ndNizImxX9#g-knExRiWjl-Xd#*yFzI*&sj!@yW}1kdcrzCt-wm$Up&x$AZwN_`h3x) zK*HGQSLNS5iR7$*;9tz-3z59z^xoQ+FNkvVDf%s+9f$;~&ehmf6$7SS6)A5M!fW+@ zK>4C2=Hxr&>WNH`+a9nSih15&U|7qG2v0Ft2U*2McvQke3qPmJ3Zi#?3RE7uv$)B= z_oa}#&ui|y_Y_3N&Gz!ZFbz_!z<@4v#04S2dl?G4jO_s|XX5rm;{}~n(;?03O{S1j zV)W}^z|7#qcQ2x#)|zRvU2CEzS-2f$FSeca9BCe>@5OdWY|mDqkU`s-xtWiMY})xt zA4Wb5q~=Us z(C!wRHG3$tAK{Nnt&k`T_u46x&WZ^HMi9M^+OxNfY|RZgw#&tQ%cT4@3f$XlbrSDDR6>}yT5WZfJMJ`kh8Fu&G; zd;B`WdJN<=nNFGkte+aDTrssrrq9?t_X{!n#KqsYmR)j8#^B z5dU4~w^fomqlZnW^HHBqFZNA#SxOL@ukX#wBcy|#l*`RP+<7jP#`kB9$W3rd*Z=I^ zG`S0(=kl_@f09s+fF|1gAD-SZsuJ&i{|*zTCfl}avdu}8ZF{n9yJ@mdwr$tRwl&#x zpZR=$_kZ=Mwaz*ZTJ62xxUScgY;?IXrfO=%x2TxniTH7NVsocfnf&+?{kI~dMUx!1 z=8MA@`(3J*(^zJemUGn%+oRQjA2;#qEZtcoVYmA(F6~>_oj?fwm*E3*1--{iZ;%ms zANQYAtKRUgm-t_lmIk%Jw}4;)OrAPII!9v z#98R8fp*L!4H$m5ZU;mj5*Q2;vq|Jiy*b`YP{KO+bj%NX-8J)#{C_O=o+Dz75GtBS zg7{0&rn1z!(Paz?)*WQe#f9wxg%@W=L&RO*t%5dK2a$AD@t_rJyS&Sxj9IqdqM7l; z7i>y7`YmwKFF~N5%hw4+Tpw*(djN2bdh;@NJsEj`z$zD=MF?F;>!OrAI<70GPRoX2 zr)6`%y&!)oLqTGef77TAJ+3DUBVgy#5+Fz|pL~Uxe?a-GF_m0XvQU?5^bDv=if5_Of63qYj?l>Gc#V;(I8Oz@+x*e=k=zu}tm1q? zs`>mbN0Z4eO6&KV3}f6+?VwtvFOpD#6t1U1?XR_u?xS@QLt$x4$-ax00{HiPy}A23 z*ejUyLUb9aj5?!Bb>)|P^il05DX&s%9gDR*Mm*Vez0M-@j-@Npf^W*%?Xv=-A_iiay9Vcs_UyH^hk}?~r+lz5Y{wE%(>9 z{}-h?+SNxCQ{(-wtUgnG7xieG=c&@DFQAX0e_C58lybKz%0uXu z0HJzT&Vqm}{vOs~blR&;S8;uu|L@;(oL=)|5Js8f;OAGr&ls^(8mW-x!~o9VjXCye zk<&>P4RR*dOHl^SpLZ!c5Uq*we*$~^j!=ti2D z)(pCc+H4o~%=_M@h`8y&sCmjjFO{WZ5t{d7~_Z#Y+ z9rMPt6>OK_Y0h+tp5s`b#Da333abjZbmhL;bSV_$!MWcvQ5qtN%Jrz`OxCxM!~na>1${XKZo zJPaLTV90Gzg(`2Eh94Tud2M6Eiedczph7CtaO`bDt6ez&z$4h;n!$F;L=Wl0A)(d?NUwR8Vyg%QSc@Vf1 zDS!N-)Sv!Sq(lD21T{BOei`%eJCH~kt#$p^019;dLf)tHT`4UM4ZG>nQO+IVXPbsF zAccocCr2z>xbhE&#g`xPE9@T+Rl*3wjlya}c72l}swzv-ZSt za!n@ml2go5hmC9#FLI|rp>^z^rOgJ%=P!WgqG4_#cbXT;Vlo>65=Rea;$zFEeo`&x z$j5AvFK-bWl3C5jgsc$L7!?dLFNb-%9tG3>cG61~>0(TZ`Xd=66QEF5t-Y!V;=H$U5P4rm>8n&MpuH4f3^ajXs`ms1SyFzTdN*X68TdT7#CY>p8ZzWqC}FpPA;U}z%I8T z5XL7s?5@5z86-sdxslP5%8)^U*6_@u1^u583CAtZ;|>TqwtE6EdjAa*$3eo_%JHOP zPfChhxb8CcnPgrdW4R$>7&Rj7FpRgz6zsODW<{-^`ww?M|4h~AtfG8LA{tySvB;MTMvsd&JTT0;`)pdw>9iu_>=ZS7ewsX>eu}_7 zz4BE`V0NeJO1?79i%T}T8e@uLLkjJymbFO{BT~_jfNdPI$7HSV}bridN!= zbP=u$sc%_fQb&281$nREqYcr);0l5~p(bBcQoBF%RlKh=#GJFL!%I1FQd9w9aZLnY z>0pcb=$9A(2C5;Uf8h_BQCBL`8aDShfq-c{Qui!NpWvRTQWoQkMfC?@H-+D(^F_&x z1pl~*?PAi{gKeP-#_i&1?V{EaNRzNtipJF3r^=F%wItsb*CDZL*BvMKW&)T_DR5V@1?&XO7mQUj$`lm2Pwx%O6e=6JP2M<6>c-^^Q}gLAmHy^CfWHrn4Euo8 zSyeV~F!~lD81^0I=%fyV0suRm(8|;9-Rh|#4jx~cbb79!>g6dPb)3QFzEGv>b>FZo z#<6B%Y`Gy3z;l3*E&p@g{qM$~R%Llua`wM2Klhvk5uSmj@Avb$STGyWyR_EM5rIb4 zT_3-btxJVhD`#F;n{4q`ejpg@E^IgRSo+17RqleMNwYYk#=^sQ*?+5o=eb(4LHZS9 zHjY--f8O!GuR5U05bjq~v_Le;=hLv@{=e9t+R2a?r){<&)x}Dle{oEyoW!6>mVrut zlvQ8wSIkW$VifIE>yTX$ZbCf30XiQ@d}c( z%=oioi+=vYID3H5?F{s2p#xz=AI%S~#*6}X*~L5#{^p4@{@Cl_N@zI>n^4ii6(m7a z7(?#=XQ;(Zg8cUc7V~R^xV0w__nbukwVmEv8*g5_ptjR5&IgGfMvL};rL4D%km|*T zs(}m(_#&_Jz6b)DRXP)03b1VNV99Tb_+lPD{1(LO92EW3CC+=m(`fY6iUK?$Z(|b5 z2~H(roIb0t?5H9rlx&KEXj6W*W~!}K-RN`7me2>so<9GdH`5DtwjNGSfEr&kuu5xvf}~C=u*}00*2}p0NA8UvHIBzy#I^FZV&TY zSpYD+?501O$Py=ObE+|_U#rsD(ooB!O^AU~HGHjpTzm{9Zwl zZ2+f2L-u{%sI_9*2Jggm>>#oG1o71NZ`l;|yV_^~<$DAmvTB%_ya2nWr| zOU=L=+j*(Zg;ed+#ks!ELCcbCcOr9dHd%qpn^s13%*PtiZ-+AM$w-79C3lL*6RYK8 z}nx{puClbm{9J0Ti zTws}dI`hvDx9-=h5bb|q7~mZ4CEF?f*M!JnG4bZ@7mND*m`%<1C19Ou)LfhBsPAT3 z$Rc^OoE@Y1icnGkScoK_fAOgq8B;+cn3bOP;u%hbg8%$~AP3;>t`Nw9)YO^)dQ#S$ zUtcnj{(Djc{@8u+UxUnI-NRjB*^mq=ARCaCfSW*p5Qd%r3sk@&m?{yqqDv=)Dm;QS zEa)!`1ps3O1Avujdr5H-hR*;sxF*G=OiZ^xXZ^veAhW|8 zXXW0VNXx_s@u)gB7AaD8KCgTG5U=Ql0Hd0lE1t2)e`T)ue8HL6EGAk(?M#NtV_yWp zL+AXXtSO9uEb9@M@gSI#er|Azs9QU7sC-xr%0HY@f)UQ&Wkm(9`GgLtE$#q|+Ch~u z1uju_;!|m?>r+i^sRphLM!upY&6($ooMySL8(5V4$J3%x>#Z}n>mgai_b(unPbc>X zTyH*;zJgHy%mt${_taB|pt0jd(tN!7<=N|CDGunF63Q5T46zHxoD#Yw<-7F(G~6mE zTsT~lrdC)=Rb)0BeZnf*KM14vl)T7f70kasw7gfkkJSQaF4}=4!^LDOmsM8B`2EOIgA6lFzsWn~ zt<625LYa(x$O3if?13k#LqBH|sM~+EkF`I_S#LIqHCKElcZZ<@6F}d^^2w~&)83mV z!Mk(GGo~gGOnc@GB;WY?pGue`Mj)Duep#OS`;ijHKAi!3wu+6;uyv#(czo0zI3)B4 zq1U;gbr^DL(hF?SVlt{>?@%lO9SD;cD3{4GDH5CovhzasfDiG%92eYxM4Nc&%}3c@ zV<3LgIluL9Ai_g~=YNE~?_`J|QG-b)#!9kg-H-O_ZV!ilvf9JmgVb4gB$dYK*K@BE z?itj7vyH=FOL!ogUjbZ4;EapE_G~&XWOug78I`J>&sN`iSyX~p)M8qPz?yXV$E%D< zfvxjKk!z$QwLz`2=pKUx^>~SV{BlXP?8louc}~2olI~2GHWJM;NXFCtYR~?~0ENHDQmMRbQcJ{Tr}(TmA+gCi{;T(l^q$`<2NxnfgcqK7H%I1x&j^#GD4 z>u-L;hi@A??Y3(EL7&zq(06UzjLbgarb8s6;z`wP-7~;Bi%scfoIP2;MNTG?M|ReC zSSmB?aS=J}j0QGLJw4F6Xmytqs>k<3LT%zjO9pEqDSxs#p{ko(hW2*pC$%X5FfE4r z%NxJu!&?yWG)xl?>I7F?tEr-sx&_4Dq%7*5n7{9FW~iPP zOvx{5BTR1#7hiZSZMgSTZur~*+C2u50}at!IwMCCktAeKEKWaG`P@n2Vtc0!#BB@f zmgI3cYP->0)}&~)>1GOgbXpg?_m5V9cJ-4*4oL!u)o(d>qr2>@XoYo+}SeR5(8AN^@_CuXoVOVK&}TPgt^C%klFs0 z=s3*KLqvk_e0f!svyIkRx6#D&BtjU_Cm!7sOY9bhzaLdcE@4at>EE`RwCyR*%7%Nr z5z@ZjM={yXdSVt1b9K&!;*AL1qgz+{XHdN4e^-1cO4CclC|3i_X})pxXk`Q%X4{HQWUkFAcNWi{B8NRJXm} zo*oZC3%uHz(27bI5cF>OMbuq`|9L;@=J_aR%~O|5GBz2c?G?oFdgBJEbpH1fPyu4l z|BHt(lQkRV*y#E=?Ej~&{EbwS>8Aji;k{keUgaGL$^J7k|DMjW=gS2IL8w}7!<6nS zJlE-)48Af+udGdjKyNnuCCAdfFcCJLndp~ZHJD%vcXCt#V0K_d!wLF5^!sz4nB=dC^_xuP6R6Q7X z*{BFn?sDkbcRME*$FtAtn5i2JQ-wDlkDNu4y({1>EH#NpU5F1o-9TTn;bIy7|>2Jum4E zE6q>bp#XZmig149cR4p2vni1d3CZY>Nr*6rN#Rv%s>#&>U*YHH;Ep@0qi zx+MVOKU73NUZl)?*(5E8c%|(FH~S$Dbjv@@uOJz z23M)uzR0N42;^xvS0xUd5wR3CiwR#2E?sT%38T+!fLVF!4aefs!sJ)2?Mxtx~7z zZkPc;?se7m{?Jb(I02W|3#d>%W3TI2{=vD4Z1O@RX;4MXya0ad{NtsjVGEQlX3_n= z){U0^9wGZlaZPbvXFA@{u{CLUMUy(LU1K0jT*TIJZ^@g#fkj$d}pvrry-&HPPUu8!neKnK$C30-^JkzKDcBL ziiPi&;?es$^VAMpiIMHREQl;iz~-& zRu^qplJlWX*tqj5U~fAuXjo|+Ne59;ad7D*cuO*_`zsS86z1aTUF)~P_-9I3+%fn( zWJPC2Xyf^2AJRwLNX8FOe_nDXCav&LWmh@+D60IT&!RBfZKBWm$)rYgwO8X9tIyIr z57~;>z0b`_R{%V|LJNSTV3mm3Wg215WB!1+b+b9hzMDylI}+{$vgbmt&VM{xH_l@?Vo{I6Uy4GQpeuIboyMIbMmS$8sdL=A%%tG#X9V{^ZQd*OoaD0XTdf)ir zv}s8!F@b19JzE$7NR2LcuZ!lbm@j8|l3!Ulc-eMq%b`Ec8QUGV?+5%r4nTH0B~-=0 zn^6v%?*3bMk*F#ZPq#wn9a-BUE1pOU9ArcO8;u^SJ4wRi*g) z3q@XRE_9r+UbaDBRl%)<(m}rwOW#v-AhkyQjgbMWN%&Pox6aPiq9r1@0;>^eDvGJw zNL<%g35r{tls|&r3gV4{bEA~MW|8(iNo=8N$#t#!P4V?-#2s24>HO;Xsx?PM1*-Jv z6iT`%tei^^pKg-CHwa=gfxiJr;n?;p^ijtaKb&ht#nbH=z*h?nL~6Sty;OwT=ImxN z`c55N>Dfu@G;rgYxmGcEMMG<^e%?2yiV}1WUv&uB^Z?jS0JcEsoEfnb^Gc3ulhR-9 zr}JH2nVf<|^d*EocS_oICJ`Woi%XN)B9jT+(ln49~%b{*-5p5TZ1VHr{lsJxZ+ounM#v+WVrG$|vEB@d?!7-UJVKwLDFvJN%$4QMu=Lq4a@daI$TMM8nN*ls;q3LlJF` zlt{;{D6-CEh;5EBndyYxWJ9(1q*WuC)JcDjArMC>5slqV#3Z$9s}ebOT9K*@gf{b| zH2nQ0v^&!HlH}*9a0GEe&(dA#(SWfUi2_bDQ0pV`CupgVoG3(Zu(J4L_?oqL5r0NQcQt9vP*0hGenVdh-R@ z=2U>wh{N---|YFc?DiSi7R6@=CK%uA4wTM554SX52B}ON3yLmRO?&Hiw9ed*hAUj& zCrKgaXQ(YSG5PZ)_hnLvPV2{JAvrqUap?GO_{Z)1W`R(h-^3eKFm)|fc(?2?+&c9U z0#>fCd^dOqpXyR1k+h>=^8#HP0pVb0p3}qhokC&=`i` zQ4GSz(;COig;#^*0^+E;G3lwk<^MRjjs|#nDB=P$5wx}r>;x4j_nx~1}NjOzQ z38B|&)8#UW5wrrgJdZzv1y;v3teXpy`v|af@UlZzY@y?M2YM$dcF1`=RsFnml1YQ7 zJT}PA)pZ{tqQSAhC~tPnZ`+^${qC2m9FrF+Lt12$$qs3|_nw3{4Ll|8WbSm!xfFT= z;=R#hc4t614`UQH7gtZWkQzps{c^64PU`Ae^;IpFm>I#@oX==3VufCUC3A@})L4 zTII04jpw|iVI3P&gQ`YE$t_kWi0r?bK>lUM<9#Mf0KFpgxOCP-#x?!en?)nHvo~kU zIeWK<*0shC zgSJZFh_YUC(IXfdbEHMU1I$FfRtY!p?)Gq)?=iBADAkP*i0mZ<7SF*7ATm@)UnZ@=ZzFY^&gKu`GGvdFsv(At z9>LpFG|9s)hM`-RU#2x*Kt5o@$Y*#9FJ z=4KYkBbb7Kx$qme5e{h%b25>qBx`sbIN5H3WS_!3f=&<8h7?jW$O#;;@P?xwBzUqT z6?IMLvIB~v&0ANi)N;wUA-xuTMhWeZc<%vwm!d+Yo3~^>@F$Mq2^|>v%q8<6W%T%x z84>%PtB0nT;0QZjnx9l~0OscyxF3*;$RIc~9LYDhz*pH^D$`qOv-T~~6vcGFJoK+Y z4X7U@4!bKW8HtLmi(u__9gM^v`Su~bF+)bk!uXO1ZJIpZ2;bkEi+4l02nwM=X`2;5 zMv^P;=T>4TD4mL(&?Ey?m}iV2;r{&bHjZoYx9n==xU(*Dm8OoSfQh-GS>=J@swUR$ zJ+hHr1EVD@5w>#oxo%^}@?{iGfXT=)HaHd#L_$cn0 V6?QX8AJ zeMg;SD_d8!Og2(9D}CxyD{7jiLP!T#L17OBta3+ii%O)^L(oWTGoo7!6U`rstXg)} z@u{t5al7je0%z`LuMF)DDN?8j63A9B!)NgMGApzCo!`g$UC+y4D&4QPG}j0XNk8+$ zlO`JN>(_Y&jp%c@?*96uewIvYj(1ho#g46Et65=fS-=ZCXs7Dd%9~`~Flpihz#_gB zi=I%k3oUf18^xd3Qe}UYJ(lL(5cd1Xg@04Kry%GcZkci90~?3j^jOli?R|r~@hK=! z5)3I#jIWyFXxCVj$K%w&1bXN)JyT_SuS$Sm3E|1kZHh6cDXYe2J0hclXzHq>j^9tp z{kAFM`|JFs`EH{xW?8+(JH?bZsdkS+(&-xZx-r~iqB<-SrTZ7V1tux+fOef%dypQe z(ki%|fcmZjL@aD1QdL^?qOeYD+!`9^SRU8buN7to4rn^|vaUHVq-e2%n#8i@go5FV zl_L(ijZOgBijWF}2DAR_&y7mL=s=Yf_3v&+u;2M0eiAS%JQhwjcfZ@dPM8ZAx4T}z z5h(MBI>)=ao-Eg|6J?P@GvhxEBT}04RMhrM4M(>)3W@D(OM3IPJHK%qyQ3lw^YipS zp-~nq@bT5>sq~ zh@|HF@2BeG=Yj-0%ll!k-|#mf9O-sN2` zG8Y>K{$Y;x2%mP3+VYkf{@9drXDbeNpO+vUn6;an9B@S1j!mY{GjYIyF$MhR>F_Cyu3y4m4d44T>i55jgG7!z#BDQejR@28ON69oPT(!jnLC zNA_*6LJKiEzTM;YP)NQ)zgwU%VroCYyA}vl7at@ajplePr!xZmfNe(adB30zvkURf z^>Q6&7cKNILHN&ckADZtrm^?eguZ3$jqA9*3G8e(?7%H3EFI*b{NRl)A1`7RVq5Hj zM9gOq7+5S=Y}oIx7{Fenb!qqx>^Vx8f^Q@6zBU`eo)oqPwFfB-B#55d-AY9Hl15=% z^BEcfg^-vA;sRwq=an4tJDz{EtwW-XA$!jzbrfK=h!Bf>2$?BSN=p)!ts!$IG{|ip z;dV+zITdb-%*S(@pyXKNyC;I4J*GpVQb~gW`@SISy0@nv0yKhSP|8t%=dIt!I0_*T z_rE@Oc6JKu0}WdJ?t%*Yi124JoLi>Z5TN@khAu`Trailvi0mnb5>!FA!5@Dd3N;YMI3J*>{`pWJDen z3aVy8gN|G2iTg-PA$*E#BwVKn6!W5+)HZD8^3cl$ zJ89OT&s%ea?c5r}9EZMTL%)owNutNr(fM-P#}E%&Cm zd<;=T2&vo9SdiJ8Uftf~Tf|CLWwBZH>c|L0MZ9ILh+RMm@d2Zbtu0n%V-rT2rpH22 zDp0n8qKJvhqT~BBwb@uGVy9VH_pP8!cIuIJQqhcs%$MYcUaQXxB<;0(H*JZ!@9Jd| z5P_vK#KM>-Ua0RFlf~b|^tX(j?cXv?_1%?8#f?*LQ``1&tv!~gl9#k?o@!ylRNmFC zSGrY~(;9~pD?Apr@;ly|ic)pSxf3byiEoJkfGuXJ1plwmk>LQUoI3U9Gbnko5H9=b0wZ#$-La7sm90i|tE{5vsz_x^JGl6uK_&^K;Y3gCA`e!7K!^kz?Cd z2{QBq=_wRTASILu)0sGTu|!tH4u#cUd*AVxJj0TThca;ACn~|Om|Ewijjs3Nt_LaV zt4Mw~4(`__b5=7cOJ`LZCRo+FXwDu!kAz{_6?Rk_2+z&&@AD@HXG$3SdD0EK5<)9bWe}US7Xc3EXZdB14t{Pq#A! z_pyf4Om9qC$b2ft|(&p?u8^QIBiP7}|MyhigR9Wj-YDwvhw@KfJLcXknp-3gf@VQ$`E+ zVi|#L^n#Is5$tg#vpJ3MI^sN{hBX16Wn`%iecMstGzbBO$XK1(AWm@?gg`s(5tCdUkhDK#B-(eg zthVSw&3Tc~6(JgWKK3v8M(yStyXUt{K6lh6{wwMBFp7GC-6J@>KtMQKT+Uc?TOlL< zN^e1XUpSX#Som3|_(Fb!%5w)+!6(;-CWzg7@yk02fVjoo7gm91T~fkXxcRzmko_0U z2!!7$rsLmXrKUph=N%YB$?Vox!~FHK?5xt?w1F{bev9FNOko;q>9(1F%y&=WrXW zj>iZ*g89HG}*~+{ngXHi`Z&LNieR|va7jD z@Une^LDD#@G8Y$!u`(2_C^FHdV694DTzacpt#v(Vwg^dC(zRW=SJxOnjagi&R|$?6 zp?%-sU$fa-`T(Zd_zq}^^I$}c>0g|Dt(0FX?h(r}o(y&i{Up&w;@VG4#tbPe zDu^X$dQ%s)E*rkAS4nu)5^y;1wDwx6)_O&7Tb)5evO4MhG^<=Kh%6?X-qgU`G=di= zG)vV($7RGhf_I>{%w$OC3XP6RN(R3`f?tS}%)Lx8Ob*=FF@1OjziMWN4 z+CPMLM*U)BYEh+aVqy!S2x&A~iMHyHPYv`?HS`4ugQSL?dPdgbu#GTLC06~JilF?+ zPvoN$^Xx_fUbl#FfE}#XNZcyB7QjABuc`yqm+v<+Tt& z9eM1zNIaqP-Zy>}t8Le`$zFc(0pn-zdLNjN!gNyPvcaBtelyM)T-a>8Whm(T#4=u* zEfVW;7UFjMW65$<2-h)Cg{HT(B!?hCLc#z5d6@D0F7MZa=mq?+e-I*O@D1fT zDmXe!mNt*0^T)BpY?&$806PI2Pn+|Eh5j-9P^^y4o>5r5mvj0_gVSLTQ)s`Sug4%e zj3*nIPQBTW<78lKagr=PPyRR7-GG$Co}S$I1^xGwPEb|}3BIAxiwS4{({oSK7f#PGGh8RxD4kvBpeOdI$S7#wkNkWV4!wOtYaqZdN&hA)xck8>%raz1%Ou{o4gws#<8t zn!Y?X7x0Wfwf+!sk%uRLc^&==QUCf&LijA*md^MbcpgxiuNOe!67r^KPt5O-jAZOE zCLpbX!l`vDgfE<)K3Wz> z{-qNUFy1M`1Rc|2JFwtSujzOyL!O?}QT$R~oDrFlcvg@K1Y7p)+=+y4VfJ4gfACv| zLbU5$O-UqX((ZY*Y?}Q+W@)(+rwg$Kl^N@bmUc=rS-qR-k zs`g3d;;P^G^B|ScG2GGu$=gQ)HJme~0Tr`vbhzQjKOQWNx(j`BXfOCvFXv4@-{32} z1G)x!oGhQkQ#D)ql(TLt#mW0EJ`H*AW|k^$d#g3Z{U=x&uYolq1yDqvkE!jajC5|U z(po@(_)iL;jweJo42a=oPvbU*{Txy5(a_T4rs{UW9(@qrkhgU|XKg)UPX z6c%yKNQ|`V#s|0e-!ALUZo~kIu6H+5K+|ZQoGksARYk)!Wy)v?L%m63DfKAB8HH0K zbK2X_4N#I|QPC$r48si&<5j)kIj${_;Qm{G6w82B$~IrdO9d+b6gV?uk0$)^N2<_> z7+{iK!i34F-7U}LQmIsqC6Xd1dOC__$BF>-JTxe+Xt%DM6+y1S7n*(QF`zHRFzX+!DCG|H*ZzesTY-U%#?JsB=Nj9lFeBB zE3MY$k_4QqPAq0m3U3Dug4y(?9DhhABa&rl@B}+=A6E=o@3J9_KodNfP-Qm-1Ti<% z6RtK)X8eo;Irz1 zqEndj2uB%*WHn+ezMzNG3WS>ZQnt+|%yPz4X<)Mu8chroat(l)K%+g%HLzL6W*1Is zMAjlF9m>5FV8J}%2(lmz1oN^=VTpVZf#F7$-sjeeosTN*tAm!4ien)e05fLa7wyYM z`+{IdiK@DfqC=(xg*Yqzr#FO8lrNrhsA5Uy*W}lH+R!68{JuR(BLr~p#ck|bGaY!E z5W;w0gUP2~YvpS;p8@Xy4pHP%%AEVd4W&$=#o;x@1ywh%2Gfot4>IF6IO9Hvw8~)7 zC$@b7Br7%37+NAF&*>M#5l*L`DC=nXHaTERic=L`d(m#*KMH1{2&Oqj0|ULK7H2D^@!Mdx0pj zqHVLD2=G;)Pe}`o?9QE}k^|iBNR`BKM?L%4B%kwOOwn^R`xeHzl~kF?&Gn}$P*kO( z0B8`$|ZRFP_I1m($5xBdEZKHf3edvTA&2b zT1m|(FK@eb6Gs82SNRg)7_@`|-Q{z%oShq~KIe~>%2H9Qv#5S;34{sKH=9`;Ew=Rt zoo~rSkC!z~A7q9eeGMb<^c4^!ls5(GQ!C6~w3Xe9ZFZmB;)kNKneW$bMXZ`wh(`IVkF+YGzl}ZFnYPyk!$Xl}dJCj9DRc-ZKMhekKQ;|5p z!fI|v%|drI=UMfv?V1tK{e}eZLoKCrERqjzj@9n~fpeuVPCm9YADjWxa(!A3lxj3| z>U{YpoHjvhS^nOlwOjTcqffkC{p4Vil4w~7Sek{%)M=ix@dtDjB;thw%I9;{BWU%U zn9j}u33&D4UlnV=Z`E>Av*a5BBYGf+~&kevS7;BT|3CB#`WzBPr>ZbmwVkNEj#(l{t!S zBG;R18bUNVhYcH`e9QYK7xs+wv5j;dMD(^D-2O55|Bkrk|BkqAW+Ui`x3s>!kjxxU zOT9atCKlrbc%rDz=EE6nPlf|7c$?S8Fz6KMkw!E^YGI9WVX>vkq$Y$gSulS~sjx^B z4r`!|CKZZf7?TW1$D@Jp4Akt>~h$YO143xzrF_65|qN5~bPlBh-)oXPa$m}2#> z$ry&(;!}^M+{YqVn6Ih3TM7*r@*|OiNM5PE`jm2dn0bKvdOvvyw*P+?fLtcu4W3SK zoHeWxP=@DDq}zfZr30(Fj4c_*1)6^bgL~c-NaD*k%q5*4`S$Rj_L%%TT|QbCN7;-I zFTULXs#Z*|H7t7q%^$pH{h32AW0eDY7_3q){<;X@;=`&Lnk#5X?SE14g86C_m3 zM=d!@A0Yf<7O4{3u)LI--_qQ=@=Vhz8rDz8P1s$eTbeGP zQ`PqVOf_h0k_&0{AyoxNh`h^=+BaO)5cKAsdg<4k8*H_Rt1e=C%VB~rkSZ>br2uMR zgioL|1L_gE(!KcEYhR95^MiBXPi(s-=rB?MCm}b<v#id#hfckzJ^^QTz&j}{#Zf@SE#!n;o zoX@U@KIM-n(jHGvjiqUyiPdWN@}fHKS!NA|-frPJA@vh*tgoSQ-{x)h(-+U)$Mm&c zB3>IMQz%8mjbcWP%5Qavnyb#XtlkyEOJ5%RnwvV;l0$K0rBPKo)>p0~h9a>>f3urY z0t)xPN_Ipbpv5fXrG4;PDv67>8xPBSoo1MuDp4pU&Tdw{YCT*fMlGF&_-k^tY+G@9 z7QWuqPg{NM)iuna6T!MQcCCahayer}Wb!nU9Ir&l;7d!seDmNZO|@(O2oQc9C5xjz zIZX^JUvS?l>@3w<)4C62TzT|+BJ<&Q5=vFpN|i-;1Wiid%(~(PJ_->K5TQEV9c`B# z${rTlG@L{vf(J*dsid|*x4_81Tfn=FtM}nI}?y>-ad__ejAT z1*TT~=y%W*k$IO@!5lVk8%vgnZvJg#hu~Yk;J5KlimT_w|JOOp{cltCl6~cqdNQKhD;LGr@xU zX>E94g{ped`CEqrM41jR{vNfZ7I>4)r{fiECNkB4sAl2()3B!gANT_#KStP{h0JfH zvSf|Kyl=oAVG-jMlewRZrj$2>OFC$XPAQh+Lky%e{lA(9*cGBk3py8j*L)%4yN`5^ zggW8qlEirIrMN_s_$Z^oKZa%;$g{iVQ;ehO)==m{@uAjsrkWlo$c{gB=Ar@=LoTc& ztYC{R#8d7PeiR=NJtZDIT-M=xshnV^9;o;em=7(>;AE09rs^7#(D5IzA82rDj?RCQmNgjG7r#8Z~NHC2RFTt>&> zMXVcKx>D@DOBfb_ex@vxmI>cHHiDjL&I0Du)auNDg?q;}DKK%`j-QBl8xIXJ9WNCr zh&IjZ+y0H{CIl&J*cs6nht)L`Yra!$)qa@-tyi*(xBY(T&*a*rPURR(Gn=wkZ@CA_ zdTpZm{M2ay*+3{UEe9*Gma6a?o~~ODI-h&{iu;t|206pcA@r)29f>pF;jqfRF}%vF zYZzN{WYuaJ{nU^C{L>+6ZIuLp7~5(dtmX|*SJ_Y16yLqARWa2}sZy7KOF+m+rNjehnt-y4+(d=3#+yBn1Sj5>AR(x9qwv)H*fDYD1?Qc{zn~KKEe1WB*$S zy%tWlVo+yI)UD>#rScv8ZesPA@i^eTSR?wpCJL)Au(*9frE3Z(3@6ebnB{)k*m~FX z0`Vz)?IPVRKvK2;3iRBme`i_A++~w^>XqK!e}LZ%``=+U!$wrFgM*y_O^JUJVcUC{ z%3}L55BIf@b5lWn^>2hmyO+D%tJyX1V58?!5rRSdM8Oor0}sV4tT+#vkpz@?+4ePPKj*+>+!|CRrI{*dHfh{*EEPDGu=tkH-3 zL8R{Iub_$nLU#rM>!`LRw8uZ)U;<{=Kj{!0Ln?=kE(#x1Cfo(>Kv6^1tX_66B92pkcARFFr>EnfPMluCkN zHYAYp2lJpKG~TX5&hw&_iV|k9WC=E5Mxi+VDUGQDHgmxG9`TQ{$hOd~5n;D%;Sq{E zddm-1RMq3qu6UafN`_O|Hslhw4!I|S2-$1 z&(TypF~>a^?C;JWZU1ggYr3w%FH+v*mBQJbRX43)T-BcsE=8uF*4kL#gG19!EHgl=7e{VLX(+ z{Jc<7odNoh#%m4AZ#^-dJvaK5U~Cfe)|lbWIr*G_T_7F2BZkp8jfH7orUwGH;KCl3 zw|~WHeAL3wyIC~5{-GV&mDQgtS+%{3Y+gz9ySL_e+DD>)uk3EZyci!$3@q}HtvpzC z)4ZJ0Z@5d08cDx_VHr*t!O{)?|Hyjluqd}KZdkgc1(fa%0qKUJI|hbsq>++NX{2EY z2|*ATx?5VhQF)Q&bI$v`-}PPd4{&kMz4zKHeyjHSihia9`>c!?;F1GAxD9qE z{Y4+&6Omd2^&k4_u%voly*X>-=+`Xf;#6A%J5c9IDR>+=yNQG;P7gG)0{(pOjo9%9 zH?P8a1#nwt982mROW@I@72Coyr4Ho9cW)BklxBLNiPGiS9R_`17FzupHT`4BN+&u* zuWS>c#unrpjzv~K!qIAdi%qWa!QgpT$R?NmazzFC_b^%73Opl8vD-V_?d&}F;-c-CY64h`@C*E#3@VAo-7qd3-qsuxv!I1r&i<`C^ z_=CakYlY|IPR_%C=xO>`NJakkN15s<-gp&m`iGpG>B2vA0rzuKvhQ!+I1d1&Wq->+ zI?{2|Y;)FznUvTK856zM8erkzv85^IRI(}QHT7#dnGiAzdibhh}ZMX9|s@O z7N|(eW!5Ieiz4_Lqo)gajSCeX&Tw%8vI{D9I{tw8AQ2ZDX=g??1N-U(0?N{MZ_u!u zmV{?zDVbNFsXNFkSc;Z=00D6c6wN#c6+}{ZWd{+QFgeM+B6S&0S;i6DmRwpIPEb90 z1{WfX%Ja@7`$aG+Fj{&}*6dmV5a33rf?cP#U}`$kB#O;8$YbVB^v% zCcoSE(DQIx8DaLbNx)&!wTiL*(PobA*Xtr!?H#X}UX59(UwgLv10U=MBo^I=Y+=$C zT1++L3Vdx-&Hm??b=_)98g#gtiRrmPZ=Fhyehp+R3fgv;eWpM}GVV!@C99o~ZMW;b zeEe|!{11NI*VeBVqh15bv!_Fk-=lmyejJHB4lJNxBXU`=enFP=kvng_O`3hb<4)Z} zDMSfNj+nifYQUzd34%=;8TWlVMXHfKsH?B~qR?X2SJz{5P#7@rd+{7&w!U=krq*Mo z-{WC);iBqn`4=^;h#5ti6!@{~TaY-HgcjSZk#VYFz0n>;&4ii#iWC(N-J94f6FR zApcWn`I>#tdMet^cL&InU;eBa6LFF~3zp=K5#vC=$WIXRw4i74&OK%CrlL zd5VbVuh&;Zy}EMsr{k6Z7t(Ssj{qML!ySH+2?NVtD*h?a)*tR^z9V)fDYhn~f$?J^ z2rVCtmQh=(UqZA8BOl`>J=yAnK%(e?4}L|$ki-Ljng@tIp8#@9@|#q)L4l4HmyLO0gmDX^A#d;Pj4>%Cg#KdxyhC0 z7$rYdCFK}niySR|B(ob0>t(U|Hm-YgjZ=115Ju9aSx#wXKnv#m>GG%h4S!vH%pVaKMZ}-l zf4MZ=mAWsVpQ5*G4IS7>8iOziK@&)m^f)@+MeY^w2?a!ksM6-iwslqCzO4Ib-G8X- z5rNfiEC$Eb$j%s67wuxLCK$RWjs^6l8bUFbDheJCt~K|DZFlHI{y?*wQb?9fF+{JO z2$Y0uVr+NBc;wi06BMeEzXVYkTkC>QvZktAHKy!SB$s3w9P5JuZk%S1CuzLZdV}8z zI+R{2=pU^vPd;dy^PXLg={e3R21^eK$rnr%v_V_eKg`}kJs?BD(h?{I)c&eN*1K@) z3QYxSk?P!**)4Uiq2J1We1tCR7grd~hGVIs@;@-__NS-LpH|y1v^ce!sk85lHyV3g zBRf2JM!N4r2j>60UU2le8(^4ej8lJ=Y;0OC1MQ`+D$=XDpx&2ouN!PrkWt9*=dPi6 z*fwQ*Y|y{`3O4TOVM?Xc8RAtnjb-Gq)-&FM~-ZRyJ}1r4}XJ!evQW~ zmIs3tn~+9+u~JF0!CF4p5%L=+`}jrDvC@s}YjtdS1kONT`ZYU&-veo_VpHVcZ8)Z9 zR2{xV-hoIq+*SxAC(n+NbKzrg95!%IkSBx9^T>x4*C+)keJJtAcYu+;nA&fg%UO$W zBl&5*P>tb$d-^gkB#{g$NfN@#abDrzKn4RbKr<}*#%Ny8#133D9sfjPnZ~Mu&|d<< zojaM6_>(;QiI~c5=Na%!+e-R+_X&CcDzh8|YskBPdh-0|*I~4|)V&5O9Pu2-fl3sp)H$k!3LAuDu@J{b_KxgFUdB?l^oAbhW3H&K(dUtL-umOJf z$yZj&ihhjio{9z?*S_^0ylPYg(~B+)4n*NjDvGQHEg=Nmoz2gp`lsF&qG&pHPUrT=?vk*ibKQC$+*Yte#!*G7LSE&Cu^tTlJfcks5B_vP!uXA5tXF$VloZQnTwH!oOKL=#$SpK)9;3 zUTuKsi%SlsAG9c{XQMe9+@imkT_1Y@)rZSv?vK1|NsRA}nwT6LN1=Wj%hooXGZOTO zhWk2YlZiejkqizc#0H7qCmJhu_UnF1Yk`YPVC`>&8Tfd=a=@SQ^!(o9)tBMes6E^J zW8DzVp-TdNkb>xHyR_6A?YfSMQk3S6OLT&`>g+0DGxhJ4D+FI82NWg4QW1+B(;6*@ zSg0F~rmB6bB{i{znB!vWKMsdgRDy5tLPqgoi7+tFjoh)aeNSviUJ{38z;33{aCc}O zXATRKaarZ^_j>}XWDQ*6m+QJl!HCi0O4Q=OGZO>oW2xldwU8|7rx~V!Y`b67g2vv~o*(qabSIHM z{Hc;UX(**H0~66IqiCVY?ky$O%W!d=*L+?10uB#&ieH`I&rA&ja8((*H^ie5-^cfT%9*DF?U4%@ojlOmLIQqAuM= zISS0aesL72WtxpHX}TU3Q*V8o$A<$Rg0f@WSe(Imu*og~f&;b3LD_J~`|`10buuri zI_fqEmmEH(Iyj`03oDO-odH#J1zSQQU1}o;mv$B;E4Upf%<2y$ZagE%4o>0dchs`u z@Y9u$?e|f9|imZ}J5^|OIEq-q2or;z+3ao~v27vnp&lJZ^7n95Tn?bb71ur9EVz}xBmSZ6F7y538 zCqErJEAGk*2(7PMwd7Z+Ts-43cl2ChBv=!q4* zBCrcM8iJAu5~CJBxD}2(ZWNZ0<-w#uDa6k_^D7@Utsx+*%KXn~yN}rsFpj$d_d-h% zIv3VY2US%KzicqXu2Hnf$*S9O9u@7pZu}M9BhGZ0@TZ5St;<^$xQ6B+1#KkRl8<$T zi5FRehn`Ew{zqgiBG*OkFG8Xh_RI=~!YZ199M{Psmz}X$H;1rU-{Y3CUf)tGYy)cV ziEca3Jfy9=x5=k>FxUXu+jWNy{w{sD5r`@Lh)?2ma|`uyDA}BTnGe!Xr$=x&Ej?YN zsLuYmWS*Waed1IliVHShzV@;Ec^Q?jP>^h4-?{|t&tPF2Vk@nMh#to%=flUfFywer z<8U6=5d!sg7V7SHei1OBm!AF2$*Z8L1WK6oAMxt;c2$l8HsBg~AB%yn-i0TU%Le5m z&|SrdnG)?VBAgg*%&i0|S@`c7-}M)TkbV1Iq1MJptCrM&K(+oAKxw$oUB z=|tQDVo}EYD?sR*f;+TOhDCR{brcrP-TrB_e24gzD5&^?>8p+q_hF3^Ul;qM(i5I{ zp})WPU#EOM1)7P?5`xD?ta>Kj0TNBRg!GHQ;Ee0p;IY6xrHe<2mm9`bv5I5L?IJU=Z^|Cqok}X3f_(a5tw$o zDy7VC468-J=*M;oe*;%hKU=RB3t^GYV@@Z$?xnZ}#Az49)#cX*6F8qE$~GV&7B1_L zW&3!&=<^;$1H-Tb{l_-^L}>Zl0Z{07IUdv;CU#8qhU%aQ#d#_A^)zl};InnX9aJKD z)m^OSV5SXH(+T|GT^>J+xG&$-#wOCeQN5%Ld$9dz)CrEMu;$f>62Q?#)XE15*rgZa z#X6_eLpoZiVjx0)6V&&~Xr7cw-iqnDs&4Xm=)Rtg0G8LHbEkc0pC*$Fo9S|H_PV9k8To{}x1IYz4Skv-NYfd2m*=Z! z-rqe`=pLZtG)zaVMMMaEt;7(!Xqm0rQ5pEevB#z>#zZl$|0k?~0I>uyBmyR?>vg_- zx;2~hs-_kC$YObx*!Cb&f3==dAJq5Q8|%M1R$3TPc(QEv#T!mr(5Lhl-Cl*GuBiJt zq`t0qvNt%zas)i>ubaMZ%ZMV&_URR_AKmHtK?<~af6M-Q%pU~VC>5?IZ=ukj3Dl)p0viTfogwAB7bqs>-K9>?*Y4{vN{6x5Jfyq;y>6N zaFbeQdG4*aw-@#C-GR@Q5bsSE&?fOvHdtcKzhf{>4v7}MaT` z{A`(KY+o-t#Nt?vgpf`6YpIIdXOd|Z!v6vfNHv%FuI8q$1DUFECy&^UU(y{Y!>$c~ z3|#*UL3iNt`ygf(0aO*k8~E9r+)e&4{m!DuuO`d%tY zN~**Sv{Sxq@f(N(=fr4dCYA*O{$4yJ9z7l&2Vbd2){edOD=Ky;(vN3Mh`M-agmRH` z+MErDP2x*cVy{vG(kf|uaM;d38gb(MKX@O$S z{Uj$$I=g6=#AK$Ff#c~uUi;XOYl`4#*i*n<38+LN%*S4p^2W(BiIpaBZcpn2}0P>pcqFGJBzVk za*y@ZB#-ws0o_Z9TBa&}szt?!3{x7+Xdd8xHi7y+AE7)MiBNL`Ny+LT3WA|e9pC1ED3-!v9XPkTY7qEAx z7F-o*W66!|vmw+Rq34N+DL0o3M@g4IP>1IC{byYCsFn3QF$9#aO2|f{1LbT4T0bT# zMBd4#ROwCN1xD#J7C2b1GTXJBZrOM4AJ)}hD@K~pP>03(9g@u6tcd9i9ZB@g4Ji&G zu-~A9W`44!;8(ri?dZ~w!Pe3hl&^^*UQx20juksX61oT#Yv@%_bPL^dC4)9UymiPe zjj#8A7-Ks>nse}f-|u2hy_iq$hAN~{Q0r!@5QFlYkATe;0@RNV=JtQC-aN_EsrDkV zSC^{0@YX+8Y`0+}azx!pt@~r*+^(eGe#A&*u8{Z+lHWGgWU&(gy27AZ?2BocrVI0An%t;sK0?OLA;gDt(h`Qx`Fu4SVvVuVU$+&ata z*vg8kz?&!JaZpWUykz!4FC$I{%TFt^OTZdSrh}T|L3^xA^*ojiiLaj(8jBBwyvdvB zeG@-$5&%q@Y1}-QMkDn2WjA2ajrT#v*uANAPwN|uL0;Atn5kiH9bXdTU9_+oS}E_N z$7ZNu$j$evuYxS7DI^d(ZA-(ZsrTa4tIGrfUunu1i~ zYo-xeUPS3hNqDnuOi&7aZ+hul$B<#(BZ$7V!IXk)ch8f~ zaM*FE`qSjJdhzIe4c=qAL`M2f()pDdrQIugGdtcQAu}2>v;GIT%|Xrn`73uS*a{q_NnYrSe2s$F7NH~9KaU@e zU$NaYL^CI-Uzf55mp6|<{jW3^RBh!cEh%AX^Vo=3L@vGSw+?5;0advH*Ix^O;gy0j zjwWQ9!(IJaJE#|EU3o?MgV2$JXTN>yQ?`C2x`$Uz=hyk! z;TZyFA6LjYmu2Z_@)&1=R{lJFBJBPkz%u>(6YHWN2aeXCxfJM1C&e;LG&3&v);J}= z=)z1-pF!6yW_uDl~-h=5_<&~3u!$HSd)jj(&{GpfT}GA4R}m>?IjdU{}tG=8Di^Yd4hD z=*4TyJ-%!_#amAsrxLbZ&xxbclU(y|jNMovM@?YZH-6rgs;tQ<y9x4a=lL?|k)?JeG=$E~bK5n$+ zTvT+GAygu6!Igo&;)L_gIb0oa=G0i~apGFf;~?$Yo(gzO$Ns`J7zX1=y7Hw|4tWRL zVZ#t36Il^)lnDHY5{L{Z2Pfv}mHL4?z5fQjSAGnsTVx|1c^i}JV1x@?G{nCrypi_# zjtjk`63Svfkxd?jX)zILNP%NdZG0#VyQR`K9+4k?GyhnhzmZfoSa$_)N~RqWH=udO z;CHq`nWe|Rt0FpqAX3cIkHY5`wW;j)jOZwr)vS%yuL~s7q?%Fnv6E&{9HlEO+7S|w z&KkJ+OCO_09*-OsI1G>Dn+DS`8*KeXl=*&2Q#|lNGk`BE_*R5N&{HciR2&&K+0DW&57pVzH2p%3$I-DJc+q?aYUP#?*tYnR_@7`5}YEGhg!^S5ZQZP8G9QTc9oHxs$ zx=pjp0_lZOnS+eaw_(clAH*KYF?|&kSq{9^Ate)X4SL?|nDx6f?dCr$+Y)>5BH}s6 zIT_Fhh!u$0i7$ZZHPc2vZeU&7x-P`|(Y*`6Sj)1L65icR2HuOSw)8J>KF0;A`@jG; zjGb|7vun2jI-+n6J?3gTMYL44*Ei8BJp!z6<23MTwnaqL%(*SX*k5upr%_ zf`><{A!D`mrb3xS$@i@9oZI~)Q;k%*cEH#(#qSHP@ez0hbR^st*_U@8&)qx}P~7r{ zRos24s&qevoOG%2Ja^n*;CT;BUdSXK4-kNN#et)EJ=Q7dnnLz;h0))-kd;i_cX_*+ zRHvx!G99;z?#O4tggV%HFu}fVVVqOWz~XD@ZZA_m9{VVYbAVZz-sKnL=OknDot#Q? zO<}*LLg|c>`vi>oj(+v^p6+z@+=i49AV22Q_4^k~t@NYw7hVUYP- z^?5+X&qB{Ww?yYWV=j}%u}sq<42E;L?}}k@bHCzF_;c`#%(Mho{hyDezaHRrO~PMf za&hDSD%F@~^pTs>{w&9no}4|KVQI@Fo+nnSm0~LPcugQoHQPLML*P|TqrBgBR&r`` z%E#1XqV9JTI*KySlZ0JGCg`U4bM^Dcrzc~6=sv8gq`p5t*mx9F=kfYO%x2{m!OTgQ z%V~NoEJyzkInLCdSbQ{y4+SQqin0~z*w3_uR(m&J*(%!j=`q_$uSVjc5-UZ562)QL zVFll1%G{J@AwCTd|3_YuBmr};N@IJCB@OCV)e8;GEheFLtcDJ3EIRsD0~$m1Hd1`L zXf`vN{HZ3#8ZXJ}6YhVq)dV!S4Jg*nj)05m6)S3>y+je8C5bKyCX*1bg9)jft#7y{ zY1-dLV;*|W>apm#R5ISMe5x6ADX4o89D-WRB;%rKkycebn6i+0k3`8{=sp~gXI<)9 zt~7!zdNrXuD)mxIzGIR@tcmh}s-xN3*39faF=)<-$8V+(s+LDkXPYK?TJliq?quMd z8Uk1Bf44wN5vyeD=ZJldbYD&VMbC#2Qu4Z=wWX(;sI+QOSobEsH&#lZYJ;LTkM03- z^3JXO%*U^#ARJ|8KGKt_{j_JY0*H?7Q@3;&u4+#5?W{#VJJ~*3mAWA|kD$&26G9eh zC~3)nIrbb(nLVTt zSilAJj(W9A*-pJm!uxgaVUkoU_~%eS;AQW%p171l_)c9xcdkcmkj zn@#1PTM}_Lhpf7Wk#fw>FNfA_rPQFPAvNNI$ z7hY>SxdX^HQado-A+Zh8t5 zv2SFs!k6RIz{aCsz6lt*@fr^K9{aFiG&M?bDl1aN=8NB*kPp(bqQSn>rC&WxE+b#( zmp$fYZoGQ)QZ-zI+J_(ll=2AD>5YClNkxl;BaraB=LqOcNmE3<3iM)6f$r`8mo(sq zPm=D4gbh${9Rnr1mo0mt`9scmgL_YfId53tRkV0i2Y`&$q=hgdhUU#$F#Kwev)Vkj z+&85;sfQ4jc?zY7F2AmKky#3#;WpGUbqnkwXsDS`U$(1)nluBznhdNDGYHkPL)D?i zIB3m@pI>6{%3xfBE%(@c45`vB&&I`D>|-z_B_c@sY46aeCglPWkeWW&Luo>ud(&9v z{=_Hc91rQeeHLi>fFsom#tcQ_4{<92fh(^x1oLmLuzobcjKGWHe3Z0aU^nxkNugR? zxs)p8U@rv;a{P4^F|NGFnduTdn5lbAd?l%=hD$5G%w*QIz|P)WY=?o=Qy{0T$H4eM zo6&)r_8cITGBQt`rYd6VvZ3mmF(+LWt*qtE!zK^Jj8_0q%qVo*lhoqDCe6nG=dmDe zDgaY*Q*B)S&-i?Ph4U^C_DaFoajWxo(Y57{Sl)*LMqK>HfCzY~hyi%X-vahc(9EB; z!(ySQo4HQG-o(vg%5FnCC&qlsX6TlsM%n^*9>ZRd>x7?3?ev<`# zl>yHCHUP^x-8&w39hp!P~%iD$j*1~@XY-zWZ9$ak3832VBagxQTf%~kpq8^J21e0tC=>z{;uW?Jg^$a5RG?a|23Jz$(l3+hS|0m z_x<%-Nk9p`mFz#vVFL(~#il_%3cT$%G4RVs9iC9zH{g%tB<<_}aJU)*VC3hX^|t-* zt3K~Mk$ES|ZPc8d=6Y)bPGB2p)W8=h<8~lN9e}_H1YsqA^J^Uj;!nq( zn`|Ka|4@ke>0b#X;>5tfv5SBaMf^bkZ6iW*fW3YG@$?oih3O|^D~?6})s@`AKs?|r zqLpgTe*}7mX!XP&R!z@@y-xibbb`H~NZS_gCIFBEXj9`?pRuOn^6~?YLij!T-`F@tj(xXC8p~SW=Q_XlAti2^0@_+Z!iZ3Aauz5EDIl!E&#;Px{T zCZGi=hyA0u1I4+q()?CtQF-=Q-HWtd#MA81Km0aQkZtx>KtA6Sh!*F5E?#c*kS!4X zXrL)v`{HIwEVChZA`z1Sh@^Z`gF$i)s-SLzdR$C)P6H4IS$1;~vNEW~l*HbuqZN+l z`{f2J`ds<1YoHn0pi^sisn4$f`+?Ahl!Xs3x+K8RC|n6FWS9yu!!(mWd?i~O4D_07 zdm1+WUEWCezsr+5uQdyb_Qp>c^~MuG6y6(mO3YifMFgID27L3`EN%8hB5{0`?~e;| ziaHTLbF(ZBwOMFTL-_>h_)_wq)eQ$$s~O8y-H)hi0XsWoeHmQai4{b59SVSEFzi_N z0xw+a@BSXAP;K+A;Xa`lL(%f5>#3B;gREXMsEA&U5KIAw$gDaR^LJfSvHz@#cYnr8 z15FF9M6)o`ZkSs&O)hp-6I>FZ$32xa`JHBj7k=dRGeADPZE@qgC75gF>6^}rvI zu$>~Nq*O0O0F3?tGw*v-DuLP^0mL#*56BB0$6l4Ny3~My_@wB(w#hPbd7~*_&sNq_dN+Q0(ksfCVsVRiPb6r$RvdBN z{@((UB%X}t=iknM)eCTwFaNMg-hAe&2mj7rR4zE7I4%5o$||IRvT1 zE51#Tq^wg3nc^~LDDUIIBZr8$a(48fKKPVBF`NNY@y*_Vk$U%a-o8M_9A1DI2`S5r zUwV7Z46z0}22lO+ab&1&VKB68b|M4Q$2_a-IkpELZ~!W1t2(jbY={|E%Z4mDz}1ld zTM=M0j8Ohny`yOVE=L7yl7XO8S}JdXTI|w98Y`{>9{f7fRtibXVgFoISuSTeDYz#W z4V`qlJ7f6u*5y-<CSgoJ65jI)*zdQ7{B|nsn4%y*)}2fAgRS)b z)SIlYEg8|~7~Iiu*X`2Pal~Q*c1Fs~s^hMW0|@zL>9d_k!6$b8hhW{nIe@!~5?KD8 zskk}S_(udQ2H5DgSTKk@y(lj^BG*MXsXFM*WKWgQF(v3=*<}tc|+dVBv*3vYH5xz`Rp zKABgU2;mMdKxJLDjGa=S|M`m={wcb)d9uxKIvN}PC58iHhI0oH1KqDK-_&O(=niqF zPr}5nvSiHyZJ3O>4U$${*;Iun6u|Palj`#tYK66R5w!+Ry!JV`zjIoB(VMg@)vC^m zDE5^I_a{;a#oy2>fY;Tk%cFik<5EIf!|;01&%6`pP7q}hIg!_9p;x0z_R8?Bw6?1G?*(Eu~cHaR_%&RDGL@R_)tjn>ra+g8~n+oq}gGB8W&Z5GLsH6t-{ z<#BsDf9ncz;_4uEREmX!?_-$C@M8fU><+1{JGHUA##l%83L}xQ8i|4K%&{l=z4c{R z9c%e#zfU}nN&dvYaWTuz|5(zZpeNl9RAj1;7ap4le2eNYu4PSQjo0eLlFuy}I*=em zGLx;Thlp@`MbAt_s5tD3^I|@3I#^f?hsny?L$<2!AeH%D-Y^CC9jDiyOqj$HR92e> z6)`6WtZ712+!*=`lDHS>`YTL=q=&UAOI@0H_ogQ%vN2@sN=nOW+m$}au|$-VP9||` zznjU_s*uW#&lB!_ky)AnL7lnVnfoNPs-A^3#awc$fP zdSxEY>=5r++v-u;E#Ia?(9E^QWcJtq_4vp8g8{SE}D73BZw6H!LQ^7H~57cdO30S5KML)ZXZTuD$w~Z^-VsFw)Tu) zMXm4WE=%rL37XZFwoWx(S!A7i12EV9H7xz< zMQ>=;XuAo(GXa(d0-}6RdMXRFf_>=Aemzw|@+xSY!ip~RIF`4W{-%QgtRt3=y>PWM zY&b=EB#Ui0^TMmi``Rain!=R{pXNFd^WzH#>rxA;((F8*$I}d!jyhY*q*?nxL7s;L z`yq&j&+LN!jDO=NbwuJ~u>(NK6Sgt@e$t2?z^!nAbq1E<*T!lqKA4v5W%0f!_y`Ru zZI;ZuQ%X>)v>X9F+an>ZYeT^r^tOG&x)Tm2=+?War?x^9Xb$!bPBaq`gC1gNYpw6b zW3l6C^?a~j(>gC-A_3+*oCV)$6axJvg%VsAFRQC2yU?HeMOW8S3%g9lgYk$rm<>EU z3fZ#7#t6^cY`i+O-80>>Z4ybKlv3Xj^O*aIFQtk-#j=lG`P`pizhGDleNdl=R?iE> z8B<*FAGSi-IvdjKq3BAzUq8(3Qr#dbLvi#OQC|nPw74ea5Z3jsui_4WX$<|mCk=Y_ z7w5?zWfm#qao)}i%)qa?3y1rk{`i*PJyafPv|VHTrlvdyFq1}zt_pCX^Mbs|4Rf)v z^Pev=Ha|?f4tQ#+yO?PTk33uU>6dd8 ze_;=@=WhOxX8V?8P}su9^%OPhMt31rhA$%)kaLB4C!$L_dPB7fqd5poq`CKX*4qBV2j{ z7u>7RRhW~CsXpl%`ZzcbH+X4P+CEuFPp zo;xs@MJsWIdxvLA@|S68(Z|31HRA6?Y5uc_GE#$?qs_k#9f_C}S6in% zgD-;Vi?!oI@7JK;?6lgt+JVFIgVpkYk6oNoa;MD6WFgl)!;P2YEGkIQy_13Gt zKop(1kPdCsQkshawxp(^c>*|u3P&T}7TMC-k&SnKSh!l<)nCoLT2FgT4QI;l%5hBz zQh9S1^X>(W4QVhx+NA(2SEuIOo(_6~lJa*?{5f$oH?vuZ?|tH;DtNP;yIfA)@c0OXCf)QZ)>zx3zrIlZamTsR3;2pFA*tZ?4FofMgv_4DcAg!EMOw56$;v*i(pPB|yqz2_cP2%lIq`q!Vn4_B7$&7Bjw&n_aM z*B?HcnfQDY$Y|p9qidEMg5%mg7__){btux)0|`zS?45P&uI8g_@ikG-<^KpnC*|&)0i3oTwXN&%WV%hb{`Ed;D$r6C zCKRaO%hW+f!|-L?*`11!FhbItL^UoeqKWtnAz|PpS5Hhvl`Q5cg295@#Pd%U`sQA zw84fZnd(%msk(SOKMUdhbUw6_y?i;f-kX@i4U>_g|4u#`|1i|7?;!8-Y2>h|K4DcV zFl^e>JJ~ir$hnME@c&Z!|5CTcm27}6wR_tU_f!#f>A&M!mOhG`YQzage)Jt22}2QI1{pmC)ADT4=Ym@fam^uY;)iXj6R3 zUS`3c#yI%P%HRqO_osja0GJs5p))ghiFU`;tWv(9e96R1c%BHkgyzea_bcO^LX<49 z^=|WzJUrKu=@ z!=NsrESBRY*~r*j(6e6*RqTEMJ2*HDPBr$Xb&1;-Sioz~I2*({JKSsi>+N$>NyjNs zhTZZ6hGq)pRj~HGX%vprz811eTm4@3%K*j!`q3UG(V!<;NB5S^Oa;XrAUYc*fCxjQ zZbxXJ+;Nq)Ef5$l=0}s2fW5&V+3;^}nBb*9bz1kMtQQ}Ep~fc$g-M!HwLT+oPnHxR z^bG=Yb%m=DMJYLH|4@{z5(qq@*L5n25Xpje(}xtZyj9ZPnqQ)v4=%Y#RYB>pwR&Sc zs`VE2Dsy6nFDa<-k{LDH}*Z&Ne%6M@76Lp-m_W4nj#Lp_b zhM1^Iv++T^fXq|s_j=WA(4kG%xh2QhS*ATSb$)_<9lAYQdh^^^$XXOn;gFtL#mgc_ z=u}eRYwnV>%6F3ysX?xK)g@JsTRoRFxGnw<-L?oWtgp^$5v(R>9%0+b1RzMv$B*;# zX;6-M@;~X)25z9o85W)S!A-@w@Dx_5eAy29hwaEER}=dja6|931+6&S?=_la_bB7csIY1h`%5!d!;`=Vaq~P8@l=ITA}vGUzKeN4A@}P z`sGCb*WJYvV}K=E50CQPHp<8br1}R4uN&W^vbd$F8cflYH+z1iFuWIk9&iJH_Gys^ z`$>nZ>Z;H)IH!b2-6IKcB{zK0`%c2AUqV~S1?QLxoiv&*oa}-UKZONnNc)_nZpb$y<($_qyP^2FvqaO+`xQ6Y5rE*}l>M zlYSZWUU9ty!Vd;vO`RXb;z2mznNDY*3)PjEJq(@e{EK&aq6)(utw?@R?T=LSfd1k< zxrVALbuurOxBwgGQKdO){m&vmO<;Wwls$vtM0w~Hau1CEvknrDl<268BDQP&lLuHY zH=@FiT={K&=+jysZ||mB{lhLD--|! zwvzL7?wG)E-*2z144W3cH<)0CkHbYfAg#?n7fXnvJ;67D-EsT0rk0O|A%uqgw9H@3 zXR@A?9Y<4IjPg018~LNyY5LiVy^6UVT>;A+9RTxw1C;vQ?`I~2GU!FcY78RSv&QNMBoT~C*%K_+oX zJV;t$SGKY!Pv8RSWmAF1Wh;^QB*kD0Px5Gq@MEt9g9AhKmLR#rt%83&KBfkl6ZZu@ zXNhffsaXlXc4BSaNL7Tdx7C7Cv&%%(^+b6=8ar*XJR|l!c0dZnl&KmeJUCpUcWWwa zsgS3%P*fip=gnq~1b1<5;Bcj%GF~Q791%K`@DD?xt)Op9(-SfcN&`r z-`($VImrPJZkwL>F>bXXSrW_|$Tfm;HO0}fBl{G`7-*xhYrV`KZV_5r0nR&s%`wdt4H5LrD4-eJ>S zytVx56gPRC%dnF9(|{iGEAhVO?@WXr$3{lyi<<<55XKBhyVv3Ghu5g>S%r4}LTj)Q z!VWL;{XMq%mc`!w5}`;nI-7Ct=*OHel{&tY~Ww@B!`vZ(gSfSrl4SbL+b$8U;x^sbT)(gx8r140a#MwPX02QylE`F&iafPCfD z&OlxafWDQO;xNS9V_h3<=H4gD%$@*2BG0K^2qbbssn|XhX_lwwd5On?w(480fcBo3 z?-gu>7X{h3chizHQ+d+#WvalAlXtuw5QHED_GjmIbENSK0xiD42BzLu4nN4YgxI|!pSbO5uUcav)2q%?Tv1QmObW! zzoq5kEWdFOLxp;f+M8I9olmrCr8Gv6eOFc^oz{xeiM0Z_@(n`?g}3;OMI#@H46xF& zUpDMarj014{OKyzZN(e@F?`mTf^HtD@2TH!wavjcd6}?J@6eGel*IlF!wW9H48unE z>rg9$f_s|>>&a5vH2>W_Kmbu?lsf5vR8NMjIa`SOOlf3m`J3%E2c4eFG+m;?llQSX zf{RYyUTQ707x`o2iIE4vh|XX1?8%^mfDfrA`tvUxzLRV)_|q_xXv&u>Sx0;kp@BMb z(VKeRqHWG|(Ailyr)H=!YT=hCd1+x7ym*Qkw>wVM=AQqHqu*UhxY5i8aD$6EAf=}9 zq(HRpmuq@R;&01>S%Nmv7e4L1?qdF6!n``i!-QbY8n{;UlN}d(Jh%3=6p^h(nWjtk=ktxFCpu>?}BN^*$pcw2$f}`k$V?s2ws_`V?ieqmb zl=Yx|sj&ALe3T)1r@fVL9k^!?RpE3YHeufYu_`J2p7!?rR>cEPtm*+J+V^|l<~^9d zR-~7#e30ozi_Z=b8_c`7>hhcCq#z;qpe-;DOYYbKXHFFU9o-UsDZ?EZeICz3!2D29&QK|!qJYl2(Y#j|*89dx@t zc5g=g1hpFq6mMk@3fN_3jj-wRj$s8a8>(vge4seXBb5 zV|tq32wj0!H=}YAGKx;A{?e!klvlzyP*87yJAl$vDeSD^_wJQU@7<)jdZ}X8R&}KK zj4Rb%2M=t{6aVNYJld1M^kJN(uysAT7V)!<#45@XzsUEyqqW0eG<4CcGK}tq91dnwD@^sU@j;rYytnY5?=+Jd9FgFqEQ; zdPT*bHO{K=4Yjiryc}H#16mATUPX3xB{gL%VXdVzZHm&|I9|h0u0CItG1b+~iPaP# zlzauNl8JL3Nh3&Bf~7WGUb_w9o9k}h7vJ;s1!%5>yUmME14)iWOW%_+3dF-6q!sOg zKjx{A_(7hUf$u8XT?k$~4!`mwTCK}dGRG|vTv7IpGQ%zMMmH1b%k$RRD17*Gt6y$| zNl!L~t1gZa@5xogH&$H~j{Sf5ddsjVpl@$SI>dJo1HwS&L=3PNfN*USnVk{BuW*$w3&8g)=FY1>iM%2TWJH zc5;+HO<8VM|kJnZBjWWtcvMSAMcW5QLT4^OL8Rx#a6W47Qv*4BBV6Of+x+u#9E$ubZ8dl=qKcxXe9QZ2| zGSK$JL#E$9iOW~iCiJECP$@<-(1U#6`MBeZ`@GTX!S3FQ(qRrLqpt2^qDO{C=WA*> z+*H?AhHjHyNqr^am6fYucfT3W=YL5>sqg3c!O7=G^3@W{t+zVvo3t%0)(flCQ#bbL zt8Tm+EJPKNfvaAYspYE)A2wW`8#ic7UMa7nYik+RIwHs7LkTY7eI&g2-y=a@44`wd z%zgyyI;8iqx7(itl1)Dv(0P_znN=7x*U)U@donIwZR@k)c~C@l$JgQ3zlAHu+WFPo zpIg8)n35s96bsuRgU) z+FtA-JP&{w7|02dPOnmzNI(m;3qkvHPkiI1Anq{MBD_B?)UMO}&6j02t5Hp+#ri_F zjwG6rs~2rQ9#7a1)8O5D^M{;G86 z)|~p6y4KPGP+e8Yk7PeQ%`;5GHPw2tT!LSK0!Fwv_CXLXK1XXD1eOe|vkCJrgZ(81 zAIjnuzR8}C?7M0>t8B$=G2LsoieL!7f(Gn2%c9mJ$dG_ulZE3bW8X7MOE812v@_w< zkP`dtToi~XcRETk*mpkF_nURPi?=J^Mc$;Ns1;~uNpHXRVY_!4lnkvk3(-%N#-rGt@3Wu|fR`)x4eFy@WaIf9oc`DcOUx5p!&`;Y{+7hm zJDkQDd2#m1`;@!M*7wMCp$eY^fv(Vtp&v^aRHJdY1Vhp@_~CYz)+ zPL4dl=wfnbJ6?kYz{+{5k1an0Y1cVsF{#C1D=^!(mE(Jj2TBY>j=FC-3_xRW0)u zo&tMW7LA{*KuG?!CH4(=--&pn)Ho$OdU2!I?`#}$PR?H(&T&Ejueu2I>tuq9uNNCy zEU41g^vC>0etu_Rr9;1%PM&t)loJ(sn1|lN(4<*`vl;2I$mp9dY0>gC@UPEjzktf< zUy|D2mw^II|9j!Dhe>@sn~U>D#>z8mzA8$r(R|@*FZirAcyi||8AC@?%h9+@WHH9D zIV|IxQeBg$)IX3kEv|(XoLB1AI&%3mUbgHTuJ|EpWSF|B&q3?ndK4a-oIA6pJMgf7 z?@H%d90@9}*h>DSG2ug^B)@mNb@~kwa;aTBUf9>%dyqEHYg4)4 zwi$ENfwY}g^je~;ceo{u*8rGR+?hEHW>&wR17X1_9$9lgI?n;4irUe&nzb#6YQd;S zhS4OwniZAnJXOGq-sN>oU_!2TDGC=L7O{WV(Ge06haeHZ@3O{dg#E+#3b?-zQv|! zkNBuOzAoSo7L<=(?InKycsjc5jqK_>xO7A%6Lma$5~RNV(>n8E9>e=^)w3oi!Sy)D z&nHp4qg=7flPfBOk%}wwFvoFn%6Kk%ZW*3>=2=tYJHsjdU%+=WGmISWDzh&fDZ4yn19522crqh8W0qE~ zD%6L;$0yW(v%$mQz+yzr*=Y6qBCK%*v~2_k_ru0U_Rhxbmg1B?GLOEznme^J?}3-l zZr(tEhA(!sioiwAiBi)>hxpJvG7`l*)$`IlJa3EMc9W@TR<9PetT|>@d^{lY9<vMvvL%G8HXYvT81pxP7lLzL&R08u4wU&l>d~Kct}4+^x(cac>BLs2gw9` zr(4FPGSg?9*J0kLo|8N5zK3YYiwRKp&~m-;VN@qcnxaR6+1KR&GP|Kdw=#!K0j9Jt z>(ACf7eSz)>4TL#uR(XW!~GeS=-x6J5x*ToMaZrd;2IqLpiQ=Dxqd4Zy6;X*=;!rV zOwxJg8fR7{tELP4z2Mv5^C^Zy6PA#CpbCVI1u@Rf;%_~*#njzHO9eEo(8*UX7|s37 zo-i~M)o}j#CX&0!NZ(`oDS&1V+g_ZjIsjDgNY0PWmqQr!_AmkBJxf0&f@XC+zY*U`R>{c9ihP+u@5;m793H!?wsmCJSlfH=_Sg8K@K*1iDfapD4@UA|g^W6XR8(baa(STsHz zsV+#?h`x&f%yP?Uo>d6PF~{xb*H=Td)FMIZtcuf#M|u0t5%)K;n{+&dA7bORh3JNM z$aIVJ&9VtzG{VMs_S(2IopiJ`6O4*|LniC;tnbMvGaOEVj;^B8($y%9fplZK#& z7|tuJ7*JmB>0L>s$&USTC|&*#cx1NX+xrLfrY}KKvr((E_~uf7df#^!4S7XqGiA;3 z%*u-PF}5WDR`A7S;3zLXa^QRQ@~mJ3HL0}&t-$QTa@m`7oQ*{LpQfAy(+7+0vpb=a z)_n*W0|?F9b1u#yfFVZ5O#b&OP=;h`gMUSvIENbv9&>Hugiko7#;&z)F%D(+RHZd3 zYv%P+IhHQ0hgp=6>sD`T+-(G>j<#;KaRWy|Mu&QlTBA6ql0h2#&($S&gH7|HtM4G! zcVQz&pp6QQL~1XV1UwO<|_4*byXJjUEk+Hi2C9r5t*I& zfyV(Z^~iuofi20V`kUynI#b+@SUGHHECl}9)enJ54yn04HW+48geICPS29Bgksnd# zGbtUI7Sp-fSQ%8PPF(R0O93luMHr?^&0{3vO;gH<-S@n7|nFpYfJZTPsG^I;VtaLP)iI#c?Qv{7RF4q+F`J>tTtNL!)vu5HmmI!vOiIcKZ`m`Jwy7}ds z+~yzB%RVII#4t>t z;;yFIOOv5dtuv|w(A?)?7c6a#*=F1G{kd}a7Iqw5t!bH+Pnz?1dQEQAx81%+3)3_S zCUbMUEF027_^Tm{dP>iW%YHu>C!mP(B@9PZxk4PWVs|bl1h_@sfLE31?-*@yV(QoW zgtM7d!8R3ZGKO`D_WS6LD}pAu{sp2UwF4^-CVxwOx`Z62vzPy{9{9ejb3{kDP;e1q zqNamjw-TjUK*5m7bXoV2Gp7N$76j(ExT10F_Jp9`SoO^Mfnx_^dB`JvYEGHjd?B;+ z3kBSIQ7U263ddpoJjK+Qk=Ars0yQ(g_tm*inOc~hBS=-*%`~La`f!{Ad{K0oTZaN{ zDe$yzZXx%#NhEATWLw!sRD_yPR<39EXQN4$*R+Ds2j>skrSB@va+bwrLa@$S=S5-oI+N*}l6rJ=-O0vXtx3=McgBmd;_I z@~Fka%b{=3(3)33!1VO<&v+I7?grnmnm|)~Rc{NAS=pM^`2trfA7he)`EwM~-PaW5 z3}ANr{JuQ z%m=rNMR=ep`e~K27M1qrbWHc;W8#>XnDu7qWO*~h`h&HWqb<(fI`@kUMX3WO$7Cp= z?^$bBo(TDO2e{BIZ2aFLMs)L$fBt)jL;eqXb9_Bbk&9R|%w)grJzeOy2*CQNZcRbJ zIFVku1|w^$(&3efAa@KlhO4(jbc1Rbu3eJu6EJLi{Oj|ir#^6K#o7C0?#vl~2Fmki_vPt7&Nb-E{$7?oI)WF^ zNDxCL&{5(K%u!1$FcSi>rgB}i@{<#154~=1Et6lY7p01v?pM%kt!IS<80XOrL zK~ExsaddmHs=1iY5@ zRRNOb9Dywki|*BjN<0<|R|nZQGYiOXH}&EdFZSx+mcAR|w*CqoEP^&#Yz(DVALg-E zmmwx-Txv_KZBtXrxEqe^l2hrbODm?~8h{HMSg{9|s~=11oc`~!g{xS?T~C3%pm5gklVylMlK^Fe5Z%R8;(lhg81BXZvN7VbOf6fAi_4vH7z6Y<=8OQw`_giqNo} zkUePYQ~Iog2X#Mf+)w@A?~^}$t}_YI%ebkT#gabj(@!80um zHSPgmL54LAsYbAA&)?0pa7)R574y#Tem(oq!h`#hDECv+Zni%F!llzjikGVrfe-;k zR?g1BrRm-sc?aw;^m2P+x6nF3A8K$;e|hNh5X@>WXD$ByQNP`KX(;_ITJ?OL1B8!niiXxzQ&t}*-dHaRT z3g@eQ&pq|47H%7hr@E`jOD#YgpiFB2yOgeh1$aL9@72{zYqWznOGs4CRhGcu^iZQ? z=U{3|rHG?90@7}m4@_+L;U8kFT1<|E0OM{?Q|Bv+eGa}1aB~0+e_;X2m7%whcKA=k zXTsoXr!b6o_hYlZjv9_N=2DE<8 zLaVmHIQ*Pq=PY@q=4C#>%($Se)=>O(Z~M%$`G66*mU@|v1_^6r@!%_E>0LkK(!ao1 zNqwDh4<;hxf=8m~-*kTWDEU>A#C{Y524*&SI@XUWZx+NQn(gQOczR{#F}iJm;3p9=`nsf zP<+ySP6gKPa(7?WAX40VY(B$@=#b9mrSS7}9jw#Qz5I}Hi1v;wq!do3B(-1mRhh=f zE6kX&#KPG9cad(hN&oite6fP{=-}s#odo-v^M#vs{Yuv5o{uFi@1D5wveNq@{AYjU zaFh5Ua&mqjBc9}U3whn-$QkS8S6shX()HBSh}dN!eFw;A6(MQxK=eQ(v?VBd6#Hnq zcmM6}XGNtQ$A5c^_jhl*adKr7pP9ZT8O~-u5xv&L?$`{2sl`YIRy#o2q%SQfT_+jGQ0;c;T%djvDOB#x2Adgj{H*F=0 zc?@}`TR)(lEEodMn(EI}0{~F)q!Rsj6X-?0=n#W`Q32`2Wg0mIn2494kRi_|ufSNz z?ZF55!(Yu`*a?DNufCQExGQKpJTeyM<5RFRk9=-cW-Y8aiLu-O_r5iIEc>)CrB}Ma zH1vZQDs%fU=r%HWU=?x4@6KZudAW&YnbWFfg2i#)HO1#A9+L=qUON5K6*b=z4&%Mg z_J?o(19la#H`Jm*4t~-`_JCv-$qW>0C#ncHCU^C3>+h_Z@5XJ32f`8+U_ybCeu-*1 zyGLsY1xjolaru~G#+(-xCxkm1O&5rp)AM3tRVjsQgXiX`rz$Rvm>R5TZuN@LR5d?d z8`Mh7&`-IFHLTlUTEdjgVF+vJA7=nUB^gs)&aZp&^xRmJ(ec$F6GoMib~;RqUC1RR zm2MX=!cu@bWuwmw;2axa8u6Sh9`3d$TWd9G3oXCQBFo4$u1?~&aYoj1H(B28(BGle zd%5(z;1OJ2n>*^C+0#pu>u`ES8Uh|)8a~w0{>A>R_GD0Yz-EP2qFj1)KQ%BtwYkX6 zbig9vK-SJ4(;(HoyIjmSmb+s8U%mDtO0vQ2U&tJ0+@}Ln$PgDjJLn7)PRuBg;+g4P z`T8k*yfi?`?>#;*hb6WjKKkg}#!VeoA(a`%x6ic=9WV@giDh`kS+pa{*kS;A9&n2R zM9r+3N-3TZ2JGEqBJo;cB25OMra0rNgt+2dgWd4WjOa!dGTArR?))*Y{EFe&uLqsF z?|$%@Tzbu8c%15clYboB301BfjY=_z$&Oz; z#>yVgQAHJ%L0vYI3Y0j}6uisS4$oiow-@FxXTS4nWJwcL(le^)IiJq^Gyu{z)@<>b z|Cfj&|7RkCEQ+2c$G2PA$ChQwvNDn4qU%J)c_Xhr-AwG?hI(J>EI2a8Jy2K1n7vJI zv}-MDI>Nmcf)S$x?INM;+oLVSjeql5)y;tXDFnrc{tc{h@h{3|>{*k;%D9Raf0U~lF7hIi2&EQDXDUlU6mg?f2V4G3A>(M!* zxx3h(*-lvhBn9uvEN!jEdmx8H7rr@s&tJcOwzgprB0V#~j$e?=rf%)>IY1f-r4#)C zXK?MMTg*QgjKx?l+!zuPr@$rrE8lN8gd1i2W9ZudGPO}8N?FWu>ua3o3bh?d(MG_R z#Fv>L;C9juo*L#AR`{;Y`IeSdm0I(_ujssa4PkPLB;s%w>Jfu#y6Vp=f;TnUeOI>x z1?{0WDl|fqNyIG(f&EZq-8wKPVe18q_^so`4|;m~YT|fZiG#hfSVs`zk?d-oPwsr< zBO08tIi|)MkT9OxsN0O>$vgRia<;Xw;?*i3VmhyCP5StZ9}MR}eDuqjsFo%*eoQn_ zKY2iYb~Q@2=xr`pZC8y!QA)I!^{sAzAKvA$`(u7@e9ta996TPopA(S&){f&?!2dgXb z_04h7g=f?ddeqlxOiszAf{90^#lCk{HxqYAGUVv~bqm>GM-YV^Nk?{_MRDU9q8_MhV*^U-(v2Jg((XPHmf*<ds)LX;yWX1T3jAy z9kiCM9k+?1K{6#1sd#adbF;sk0sde)ohR`>!jI#x6l4!VFlY)rMuRv>H7 zLxK8`%y5K*vv{${c8(4H4X?130QYm~i`GW35-*2s7IU5>RR=p6+UHWSfii0t;jELM zZ=esr?;na};jj{^toi@R^9s#wT}#y#1l6B527= z0HiF}f#`b~?jz21C*%@Ld%oGncbcbap1C-88|z@${9OTzM};p}moD`usUM9 z{Wq#wk_@f!zGu)-wgQyP05{A;`wuC7-u^jDpA4%Oq_un!h-kWMq&X|TT}ho1XhQy| zBz=v$mM8h+qec<5w0hoSbm`F{=qOz&f6@vk98tY-bf#LYU*fvpJ}@{Cp#u!`kK;=3 zkm@j(tzB!K4!TJazZttdx}sy(0J?R_ik+`$!wi>2J5NU`48?J5WKTLYU+m)m{_Q~zp)gW%sTb8_dC0X`J>TB+7sTtY%=DJdFV~}U(SM( z=dau7JB+P2k!z~d`MGL%kloubMn>zt@qupx+{M5B;gEb>HD$p7KAa>jVuzhHID=#i zIo;Z*?(IS6yILhG1&rt}xx-+O6qc1wmgh=IJ|K%%$!c4yv#tUFeQ(9n-W_YmVW})k ztt>YNT-;Rg&mqx{yrG%?B$~e4jp_pL>#hT(R(|76KuIK1B zPn9Y?M{m5>-RBD`cUnzG*99yjNRZgj%H6Z4d&%gy6@dIz$<5g}^XpwN`BUgmYgxbQ zq{1=pqlrfDKUk4PaiDs~A_m3q&)UIo++0~SwXfc%^^?1I*K^`k;>bNh`^E%U*D-ey zTxgxw?|Ba-X9HfukBYKg`;YxKn|fGYY!SS>%d<5%Q!}?KdmtY>k@0*IznC!8va)+;2Y<;jA>j$9e-XR~I{ zc}20qsI8u|u&P{U16f=%LDH5ndgnagh0Dr1em$NNgvq49(C+gOp0iQ$vy_`I@hdXV z{pAQ_=|2=Ay!_@Yad|mgmhXe_gNTlaFLgFP1xbCl&hO4_Yaj)YXc;>){OY@n+wG1)Rhdt6R4>}`l|0bx$nvKVLg8x_$O%&OWvA%+(CRp{!By)zM z){gzIhkaY8s-)pVxMfn6#Cul3^XRwfF#-CV{J1bG3G6Y+>Hu+y=bf)E72Fsq*?nHw zb-Y?v!Ze;`l%_Zt`Q4P45Sl%hF|5=q#HE0EiS@4O6(W6Jze5?2{`~r6n?f6)2D&6# zT>a!bVlv`Wwi~bh(vVgM@hmN@@DFboQ(jmNiP%=f>Q%9JCDR_MF!m*zYQUz~+H7L- zx1L32+DYFTt-+;keX{B`T%Hq_%{D!q&`}3Lvu9dWFui9r>_i9IOwfZfAX0QLKly_g zN(?!8yG2w~;!ueVuI!PFy&j7KDbZb<)2!z(eBO0e0+&aTORV|dow%jYA1arWc&v^d z)^ds77v{kK&_lM;z~jF5X|i0G2p0LZVhoa^mx>gKYF%dUuhQ-h6fN1_@n?;fOW z&2cG_U96$_g+}+_XcYK({doD2{JSbEfRgpNyST<-Oi5r;v(X<(o2Vx#xvI-wXbL!g z0@ZXA<4r#i_2ZTt1;SgOb^a}`sT($)ow595y3_u;nozgaiQS1snVe%O4oQVp(X^=< zHCJ&$A+As7x2I{C{9?L46n)lTUf2u~!?b*6L@a|S%5z7l@su7s3rJXY=bl{7iGJ@O)Na^Q2)BixYqZ?71w zQrXsaW_gbx=e0YmXKMBLxn~XhM$(Kz8`uk z3s3V{9NQi8S$pm4jISMxDXgwajn-B*=-Q+4o$WIB9OZG7CS4c0jXfpZsN_!T+a6(( z0v@k!Jfe0`Fj7i^abqE{hiNJ03h>@iyV!X3XLlcLMhUkN*rN+Q0@Mqi`QuXV9bzCh zCF{h)@L-k+Urs1B$K#BTUc$nqKI2$%=I4`qQ(#kCi^A}YKT!6i09Hlt!Bgg(-sHN> zqc?8d>ahoF$RHm@KawjKA=+9Z@6xNAu#Y=0sjh(oV)B=1G z^uv3#*OJf+cMoX#(Jiy*kt4qJO`5v9DGyX{_d1Fj11KmI*|TcAW5sB7!`ZxD7<1mE z_^7X)oQR(M{?k+rL>(Be6W&T79ggtM^US?hQgTOYvbeSl3MiD3L(sLMxM9)c}E`Wv+3-*wp zecI}wL?g$nszDlq4_lz)l5LOsPI2m3>hPJ-M(UH--@eP@EXD&Mh9`VBYnH&e@h1xK zD+T*gD3Us={`}aQqGvgCR1U;SL-gML z`aJU-61v&40niil4IVE#qOMUZC|k^V1zdZJg~va@YN0wA(E>MP(3bh5{t%h4VP`3M z%{HK`7<*45PwxK*iA+bEd)0+SZ>S7+NvTj`sgoK{vTX`|p*rEmeAo=KN8cF}(!4ZM z?j#++0OWpib~9{Y8{-~%-$OBJbCpO(C%PYSp1hk zE{$X&(3+tyE!?vz<)9L^%TS=c<_SNvfwhilB(< zSl*{;@I60>DS_IwnD*T;+M2dyKAO)mltX(1MD--_;C0}|$+aB2F-3cJ=dAtz2aV*b znR%Pi>2Lf}uRXvbh$L5XD`%tF90GIFb!h_yF}!ztd>@Q33LF@spMH}?f8ZC1M!^Al zd)eeBm(1`>()C=~Mlg<$Io5HmjL;VK@ZE3GUjv)fMBvOjB=dB75^zkpISYr)j7x1I z-5h{;_}Ts!^8h1GIVjI7-}_a^F#9;pB)+=As*va>DF|k--Y(`lY<4`3JT~}?X#~C{ zYPOewavq+VJnG!lLY#kT?hf;c$77t)$0T4sJFs`Q{4<{`{H&mjJ$ou-mI$OqX1ait z(|=CzGDDhd_)BfOE2mPcet5Qc^=%5l5x&P~@6q~bm}+{v)e|(Z zhy|k$6=BU_eH#vN;KPUfnXl4%tA)gdRI;4L$UUv_EuHv)X2+g*5THgu)!z8Pih?X1%?*Z z5R1XlZc2y`&F2$EbVN$>?6NW=J6EB=&Ys5r+fhIpUYsUU97$$~*X;+2Ib=BRf7wa} zHBbH8HYeVVbss>KSpS!-1I)!uBPNVb!hXWC46G$~z|_!A#t21?fCmG)@h=P_K(H)tV~j-xkuKC6E@cW?h&u9idt!fYe@q3a0y+-J zV7`Oce{FKj7lrab#a7+@lDk9|!En5uYkSv_xP)!-&YYOj&_7G@kogaxm`^S#J-Pp! zj+J?R&j*R+eSFho{TUCu_U8R)z$B9wn&4wt{lc&*+c_gyRZ|UsBJY)NG8pa7`xLC z#h3Amrj*(jxWqjIWfF`nFOlNj;J)l-N*l*5(q0c`eoTMjAn*t86Z1ikpwvYM(!to| zC)dFG*I?I;S!QM5d&$E);(s!Y_d#Xrb>PjM99JZpbx4=SmP`PKA6EfHwbA)V^HmPw z`)HOLS_5 z(8FkRZQQOiROt93{kLn!i|f9`5%~iiCwP}7ybdP&^b|Zk zCk4GFO0sab_Qy~G;8zf3g(eQITYI>WYdofc*hRmV%=f2D$q0J20%E@8K;RDJEC!Kl zFMW&cw8u$_c~!rhQ{3EU#=KoJHjx(`laLcJjg#YK-Tc_*-INRR$gOjgjgaqrSYmi3 zlNSGqn<}9zJ%4kGoAKP25aPWI)WtVC83d{Idf@o!>EeeEqSs=GARd0N`Z1$_Y?eRZYWHQ2mP* zj-gRL_mi^A0vX++fMv}tb>Lx?#<+J+B!2lHF5!KK=$8sf|K}|LAmkRG1l9f4{Adtu zuOs;s)#`=D6PgrDrIrCNEB4v1yXjxJG(rY(APlHydO$0k_Vo^@h{n&`;y~pqm#}l1 zCqZy0&A^gfH7x7_w}tByFn z6-8RE)j_sd40g|-eB?MHeM0ctU8CuSBuZ0_ubv!IZ99YplL=+6yWNhtdM9u%|E`kz zE_`|}+rSXuD!sqGVaYu~&4v3%ZgB{GrsHKRi=;vNM=!MuItCw<8CKePa>e;yQ(#a4 zsGWs}@P*Pi`^6?v%wy((ZXs>WNwIp{<9rX%-Ku{Ls~@5rx?7PYxC`a@6c6n zoQVxxXH?}to^_v?KBdD5a6)~O>GrgCM%z!nm+u)D)mvk;9G&U@Ik8Y1yPMmtIp6@+ zA&&`^cmn~uGM96R8-4_{a28S3yt+z=QC9FaBDRt4^l8-1FQwPb&Rf4>Y`;AsKAiIM ziivXxUS$ ztTw+JJC3SJ)wq5e#dbQ$cZN1p1Gh%l8bXFpntkpn zZWhvl9?!vy>jR<^3kGqWwCJd8Q)V%m`l=As$wH75TW9N)z0>w_UQY;~^Xr!I^xn4i z7S(0<|Lr()orSRtSPMG}A#ZZ{lNUYHV%6-by~>VE^~vuB9s3CMMxRPK|IEqg=fZsx zV=cnGxJf-=jT~)mu(`Vu_w8&tZH(%HllgSZ*4)tX0j%1XQSS$t4Wm8uDLPKpL*x)(EVYSZ39wEvQ*2x2=;R;%=2R`f4 z3zc_E^f%ucJ}*u>BZ$fj0Ig7dn?al~cX{DA4_ zXK>y}{Yjb64BSB#<)fkmZC_`{@+z2du@+?>Q(8OHQK-O&d{o;C1Srv$btU*J3b!p3 zy*Jq(d(Q2x)h8ueo|}x)RUVoh<%xN{*Mppzhl@xP z&aGQ9(jmt+7F-D&9Od5Q52s~gMHVf|wBxF=CLW>0=oWm#?c8Vm;kdP#GR$jTRP8+-Q8`QN!Ap%uPPpLHk1itbgnMWMwM!8aO;$B?9EV45dFA4E-}0TgE8I>SbDe z*!OT&H+Yu$?$ow?>xE5izJg<6UN39C(p-Gsmr>MRDA6!_XAhF^>;y#)w`?&70&5Y(LVP^7@gPpbo*S*o$m_x z-t12&C2%iGJ&l|UdVy~XR3Yjg$30O^VLX$Al2V1yY&yjx;9ybmDfmo>l=C+{G8ZHj0wal2`r%lON{B zrxgALFV0^(k3Y7WpxwyE*O&v1w`Pqm?^+7;`mpWx5QINoCD1Vf4g;{B#78r7N>@Gu zYTHN>?@Hqo5H~RR{QABhEV{}wv_UA`>jSdQQ%4)j3uVYlR1c%eICBNCp$ncW%IM&1 z)#rJ3mi%CJsY9~Gp1ec$o>rtfJ>3j9{bP`0qbP6t;`;&rSm<2XJhWLnW+{1I&xJmm ztur?r8Z0XagsgPrwqd04WRp*SA6a-n4go+q5V$vR*bLe5rs_z0vP4PZ4=Jp4Xg!!) zLbS7#@)Jcc)a6WO5Z68S*za|V|M_44u2Ka|fb`OJk@vh7aGZm!>_*jpQoTy3Ts zE=$&|RODi+>GJh)ze?@a-oCCaKO3B})hX3Hxy8#RZm9?rp2Y-fJ=M=`zxCcw>@fmd zC#1)7l#QJR+;I-ae+&`gDxm@B6{1>9XB#ao@htfz>&KTtPO>vY4aXSB0pA1mvx&9x z#rNix*^9rVY2pR^Wa?X{9=$$liZJ+5iLVrhUs3sHfwW>H;tKs5BRWuSf&4M$V+d~b zjNgT4rO@{c6f07KWP_n9U^1QH4afIL^g~aq3?&Zdhqz`qS{2R%B5wv=%-Kd=pW>cusJfxM<5R2V&HL( zqC15T${0$FB?Jjf#EI?d6u-nMyek}wY=YBvHb~!SQy0xvYyU>j^s9H3v1MTp0a)Gf zXz2!kVaeTPTVatV%pn?5^&!6vE$IFbU+}##N5v+vTYx@|HBvY_7No+cD*vdZGGNg; z^g=b(3>S0QGvv{PbWpnBTYL1$Uqv@o@W<1dn&V&Pvh4F6nV;Yk)CI7|<~zER>$_t8 zrQeY#Isbp&`n8Ao5WBo?Tqmw@Ny@`agJ@9~E3(dvtbri!1OKSqB=L(5pMy=hq-Ng* z{yrTTCyyzNHSBpfK2gK>na#BK6IEXIm@Lchb7Pr)JK7}M217mTkEonyW;Ab=z-6CW zXuUP6oxQR8UXj}>m|(iX%nKqszU?w12b(V3P8VpVdv}}t3!D4Vh1kEQtFcXI;Zf|| zoX>abE(D;dfe>t(UT?JvT^1K&Lp;}p=M;t&H4h85qg(gNmNk(?MN0DsPl-&&NE=%gG2N5 ztXJm79QHWu)J5!@hl<#!E-@4y|Kh9Jpg>ai!Sg6`h+_8o;j40uhAC4jztz|_KMlC110POng<{KnzP;Q}cQ#Jwm|o`0$PV02~LmE)V` z#}Abqjm_0s<&hRM8|rJx0Sp2#+Z5K{o4s>^P50nt4s|>IH<#xMx>{%y)5o~T7Q!6S zq8RAKbLxv~ZO;xHPK}*@dM`uYi&OFPecx*;0OZ+$g{Brxw0Y}xZ@V9u&2^}ItQo%~ z)_mOFm-*b_zLkI)@1+pk)YeTiC!4*J zkJfqQyyXf^vsAMMHK(NUc(9(Ik8ji6LB9rrXR<};eOk)7qW3DZGeT+36E8LFCs*0_`>*L*I44D`;S9WZvu8z(7V+p5)JIk?Xm93tJSIyF4q+O@r9*VtBq$UgF^ti zxQm-$BR>wWP}7$QCVol+mTgVRy2Woy7;&g0rQsd+`nK30^qu@VCGy((s3@hU5=_!1PNHBL>1|I}`adRi$1DE_EQ@ zbJ5OEQcZ5XWh)feiIEVpyHDlB(|i02raj#+EE8b}YS^*sp4H{k!6*GxXs))EpP8Qc ze4|v$krNbtnKeb!I9T|7{PTD)15U9z0Ul0bZhi*b|9pz zAVxYU+9|ugZd>~n|AEdSHeU!r6Kc#+x}~WM@9J7C_1|t&E*O5BHrIIRuJh(Ludfd= zReI<&=yP_edbV)aX6(6P2oxG>`Ot4zXt6m+`9%L1s=V;EvAfuZ4@_SkVN5Dlq?Qe! z3<;zlH>cBT!MPZX@>QWDv2wn**17CA%nSDR>Qu(h=@m}ndVM#X#22Qomh-DQf69_} zsWLS8-hPZreuZCD5h|Hd|8aLs>*}8uBV|pYtL~q5j5+!q-P1FtS(rPcSt2k%FIV28 z1K)$<$!|?^n61myjuzXA@pNEvAhB5t>~7Ha7*h0(J)+-5mG9(iB-Q}%>(%~mVf?7z zWDq$%W=F6ljMiE2(dHK@1YEy~Q*ERpkVU3UG1Z?~S9{L!*JoXuoqv2v?>QbU=fZws zbH@1)H+WC^74!l9u(|#lrE> z036I_edtGj=KMsrSPnqASiq56sLJui?I&wzvLr{m9fw5iT!EJHX$oKdMel}Ko~`xO zbM3^^(%yc}(c~WuUfZJ~m{|)CvpxV3C$IGNIc6H2J85Z(JWu|tOuR{qNopP?5fm2q z`daRrRBFok<*y^=T}>E-YEzbW1BONJ)=Iy1P52Lpnw?n!C53@4Y|W z|KSzSIOlP)ZO9zmsit%OXG28#SDPl^ecu$$Utd_q5rfTiorgWbXnjZh;KTkVf~p{G z=>(4P`pjC%-Cpxo8Wh|MGkjH2%nCS>1M)1`&zWd1i}{@-!fhKn>OIO}KvzDQcen=+ zEv8!eHwBhfgzlA5~QutY} z?o!CnO_qe>nZ$0vZbr?NFQ{`fj5+d{ofAD92%p^!(u4G2%hoo5t0>Nj(zMm_f6f*# z_lcpU5}ip~NTI3lOZbn|aENzw10W^U)ZKmvk9v;1|1#<@?Usj!rx~1Ysi=|wE zCQA?1x(nM^uD|oFUG?LOR!>xsM!TE_L4{n3x@<-r?O%Wsk+~LcSle+;#w33atl)dqVNdyCb^=lHOeIQ6etleIgJ8HX- z=C#Syizf(^#LUHOeMu+jH&n1kt)~+tna0n;gGiAnNv1E)>*$Q-d=Lb88B0LJ)F$gL zd&BM$bh9sx*ZFhNjVg6wwK0~c1Gse#<)`gA>ja5Zr95^T2)8da5IqYxGE6rtyL0;0 z@U$N*DylB%zWp7YHmZ324D?U)Z-r|G$JN?-%Q=WKj}3p=5dnGoqsC3CQpqm3Qe`OM493Rsjv;Q3^14?S&L1ywuoVnQm?d z+b{6a%_2dHh9*bX%p8N1nYOOSq>z51qG4&Pn$BDAqd!bGSLDT1h@GzH))VW8$5+^_ zEW`;_W6 z(#7kIm2H=eP`H3;38h7|6x~KT7A_t^Z*k{k73U_sndKLX4TqT@pf>?a02+ckB%26&W7fVBDVB^XPGgvWJZx-Rr$NM#J zQ&$rYJe5VY3^21(rQ7|Tv+bTum9;)BW*qQ+Qz4<)!Hb^-1MLHoF>984IOReKz@O-P z&WGdpj+bYCDed4#<|v?BzKOx=t2z@NCjMMG)aCYdS^<;PS-aB)wSjFfvr>=5GmHP?9-$~eC~ram1hQrcZe8MzL%)(E;CPc{W_TC z%^u^TZ7DvM@yCCe++Fe>3K|f;2WCdR4mQ&wwmy=_l+eUGfyi@x^4{|50G7Lon-X20 zat2R6GJmNDR|5bTnH)t*xD6li`RBZ^@;zUO*3)S0LNilV*CGK&VZrUUI|(4ALz+q! z_E3XbkJl2v+n*K-ukT!`+eEjyCiPC2a+kyHJvrvtmyZ3O#rED!sYSD~1TY}$e44dZ zFl68A^7c8f5b@nn5b!>bDdV|-S`>_hQZBdG=8TCIgOWUdyc>sgNfTPoCprM2ztY=J zDUJdZL~nIgGelj2SIvNoJYP;gxYrpobd?;_m_-EyVupI5*GbI1^{+2t zK|d%iq=L;xwQOX3c+&|pTyeK>RJi6oM>c3@4czF4?ZNq2k-*9y%d`dl?@ut!{-fday-iZ`f6YjQsr}vPrg%^;)$&EGsl4$e9!Q4lB_v+GKtJc+*9E$dm$d%okk?=MZTLeB_ZBVR17ErGs@F}c+X`e#v4p+9 z46(Gse}?$!v@y91BQubo z=RLh_pA4#8VMK5l*fZJ3HFw(PiK;;|wCn>c0NCi(m6-HZf7p@otm+G?NVl^g{Z31W zOeYMun^f2|GC?vYQsU)JuEje~qp{otzLda5@Lk(dMw$(dHrd!tzVl7B3y$NW>ByZ?!x@f5T`{$wgM_pIE?v=yes5W2rNJ z^4qyZ!i#K2BnkV0b2Y=Sf4^<(2}zF))>u6!YQHk2$<+M2jAS&vA30gxkKz~XbZY}? z@}OCPNJEd4<*I!2ivR<+l7c9KrQ)Ug>8yqE)`#z{Z>DQh5pJV)f$izjeo#Pwl8*{p zi9S!CXK0j;VO>Su@0>3%g46-U{+2&(vT|C z*ez8HdPSmk7Q!P=Ebbm7{x|0TpNvU1(Cqzx+!l1yqWGVzIZAT=OkJTNmb4TR9=oK+ zV^fj#8V3U|rZbD0d%EhxQK&;D$+{aqQH+l}>1;OC6(g4)&~r;rn1vEjPPuP6!qxutYYaLEbHNKjstV-szZ9RCqaJXcH1OsC*v z8`E!wTbb0s$y147crJta@wz=mV$l5g=^baI$!*b12zlD z@37^yRw*i684AQTIKNq|IT>|qwA^r7PFhiwwHmGNxSU&{53)T?YV#I=iE*NC(VutU zTky#er)*i?>7X1=0M5pF*XRdBDkli@D^jo#pl&(UeI%FG0`&eAAMbw zM-f`@TEB=SE?X@bUzR8?-(()Is3krm%grcy7gSj?aI&$z#1qReO+;;$Y+YY@*{QC)OVlUwcL zQ*hXn%r%i`MlG81O^6yMP7H%>#(TBS?fB0A7B&~W48Sd1)fv^2Cp?8GIKN4uy&|jG zytjDIBv>Gy7Ux4L(B@C(AMY`^#_JGlYF(o2R+nZZ04SZSa#{)S(EO2LceW!REZReCdx}-2rwD?7BRSPW_`intJ5uncfA<2Z$a#*-s+Fz~_!E%m`a=}Q zAb9VtOvK&adX!?5?3No&nXfjvv$Hb+>S2E4Wtziu5Y}LO3#0{s0u&>TY)mr5X z$o9nu5$xClxf4v6v{ww6klk%qw$)+%ZBfW9So$?I=`u+}olE+hu!k;HC{Eg1Rq$)4 z{GnP<^)=JT;h%_!R`e#*_?fl5af58kWBtsq)5E4at#Flu?W|xFHhW7p#Q5&h}`gc40ox_&e%8^g=#i%#rdS=zf9b#Ek?-_Ei6*o$?>Wp*6H0 zSVX|HCD$QCczCBj*>)bhbiL$%qjcZ;d}2G+Xkol)0o3R^--(jARdjO7>5bi7+BRB%v;iJm1a10n(un@bfxg{1vwrfj-q8%yb9mG7t8xkXzR`1n#+SIg{>#oo&snV)~)pH`~d zFQ#yhuT+;`?o>4av?+F1TQ0}eI;7S()`=5I?#mo*oTNko%^ut=v#6iyi?u2ooouyA zbqs3{4v?30R|3Y%o$dENG21iBWL5>| ziNr;o-zI6Zc$tAkQy0uMJ9Ez10g<1z$C>J7$4BH8$T)7B%gJO5`(1EUUZlojNdVtx zrV`O)FvmuIS`Bs(MA2>RCC(r*uEiM84)d7v9!UJM8NcJaLRtn94E?b=&k`q#)mRWNs(dai!qB$oVxE&{ir;XS6C=%LN zAY{d9^LjA?8>RoX6pD)6Qk5V{lC7SaAx-9NAN_rp_+Ru z+g5!G_|Llw4J6PwKdrqVTsv2U?1CdlQX*-dKH%;ziS0yXpk;pPGz#XO)jMlS$}(fk z_CP7{C@;na`CV?u6sv=-^<9t!YWNV2q50TBDm4Q_m*&$;tr}c_7pb`_G!_0TrXWA%dkF$VAAh$DSC3FSnMz2( zzirrKzS_e|VRq-^%KG(%qX`$ahU_#$kFri0aPpp{ZeC!xF7l8)OBVOs=VJM1#=yeJ zvd48!N0b;f0++1t?CJLVKXUNNVKy=t_pI*Jb*U^2T$yt#*#3-tU$*v0C`w&w>Sa`4 z#eo$>_!po-dR!~GRvjT1;7OFcSbm0^icCwM*uAwucltw@&s995K`BkZz#3C2xh8;3 zhoaL9Q$Xh9ei{5gZlgAdgcX#5!%L^mHGkjH(v2{6=4oJ>?O<<}$vLLY|m+YPHwasReSn z!7nKb_>K_OKS$$jbFG?46(Grw7&y9^*%=B)n8S!sDHwtL<{tkzi)`{;&7+x3k9%Kdv+c(t`Zlb@Yb$h^_H+{YA8qm z2m`^y%dpvqzrVHu2x#rJnx^<`#`&+Nw}1C&DTz`Rl?tiY(<%)?PLS+;n>@&ad~q0S zZ$8zZ7?vwj#?=*@%|1VXX8W%Zgt6K;Rx#`_fX3smQT<1HIGW%x1zpW95PXYnhmkGpA~^EyE{TIIRz85A*95tMfugbRhH8HA%#?1}!u~VZ|eaiI+A! zN#E*@lNOL*o4l1)NWF1_ulV8y7j68A0R)1vX@w}VG#4gfR|&=UCAC>t>R!!3Zb>lq zaNK;vWKA;G^FPAgKg!O`H|{1I2G!0`AT`rDc=MD zh0bQtM5h1RRic|MR-0RGQ;DRT#a{*of@#N7j-M%9j z8KqEPxfNDBIi0pU8a?iJ2HG;fKG$`1YU5v7 zFdd*ccrD622&efyq%Vc&{%eTNVLM zu}kA1@qc-qQM#`!p#Qp z3SbM=V9CH&z(kQ3!|FbJ^Tr{7&DN=+4?yzf2V&r)$Tjyn@_Vg8?a-!M`t!#8@mMYG zlwx;eZ?rw+_;@qVpygcQdWiCQhBqH;?>mI_)dyy59&3x?Ld;7w(TCGK6IAHs*fLt) zon8LHS#sEJ?6Cwv>wzvV+55THcLtR%+AA!#hgJSpX6hfNQ>R4doCP)6j+k7MKWC*% zVC#<={-FM?u)HhtegH!wBekCnhwpzzy_zo+Bexj>=HtUowv}i|DSC-a(S^B5kc8@}8G~LfJsmp5PA35CsDj z@r$IC)C=092&F%*D^moTg#IUOaVi<1`$?w@f?_j0AJKiIBoc2EdugFdaJmUlvOK|E zeiC#|0I6=dEN^0B=1(8Sh|d?h;n)2K0U}>Y= zqx%qO&1^+FmCpgez%YwZ@U-%9=R-UhSPYZ|Dv|b`9!kf;IAS-zR^A)$lJz@}wus{X z4N{|ew|0YHwE3MZAg@r}Wt9A3wjC0S+C+UZ0^D2v=JB3OUc!K7sr9O&%`YCyw;I6R_gSkhGX1H+-AO{5^t+Ff)%HSZjceh-8eHKCcuar3+Bokivoz2G&b$k+o zQgNqpX^Bv{)4RQ^)?_r7iA$$&+6pPX&d(B2fQI`LFa(Zn&p-0gY(W}J9W-raiSi}x zPv`^GLPndBq_XF9(W!pHwDTNAV~qGys1FMTlCc6^!M9+QME`-StgkzXyYkWf{dtJh zO9mmF<4sms^V6Y2@uJNf)Sgz634Uc{ug39#zGp^piO=6cgH zKl2l-Y*zQ@N88Vl;^8efh8d+ave{4OIi?o@V22hjluKGe{Cg$8()RS2e%<~&_iQ22(rlh2kSuoU{`d5& zkOS`}^;h+`Gwu77@m%75R{PxR{9B77^spnAA3%ID!!V}p zY({WLeR0f#1WxaHei+LbT_0rhGKjtPU97nOqRc;=!Tf1&Rvi^Re1*WJJ@TWg;hZGr zPM~cRvECmWAq8()yW^^C0tK4u(5SxxpQ1MyrLx&_B$xk}1cS(e5&5@I!T=?qytXbu z3~eO3V%&pxbYy}r`dXRNjSSI4kQg_7uoSWeBVMHc$n3=R1Hu)sZR449>I!^IE7hFz z@8Ct*W!GX?M>n%3izsBM;;n@=+oI#mg?{@I0MdS>d~j8aX4kCGPeeWs zR8Zc|gM@%l$RqT#L=r$FQjkerzFRYj)@p*@Q&1qq-euc?EPU)&X!;;xvwFb=sosm5 z=?PYo#sQmHnNjA}?AU=mosTCfGX9VgwNU$3-3N5FZFEIxxR-9B5i`9$jxlj93IEN@G; zQ0^{2>ug!pMI-gd*M_OR)6_`Q`Z7Han$4nQcPRMY*pxkP%mFY(@j?NT-!vSfKft^qoJXLRC zS}A7D{0&GH^8ND>NV8ekOE6D7jyWOk)3RfrW16&an7*Z;1G=bc#x-|{b^F`jAc*{W z+b>fyYAmwxV3nY!;$5dOoss?dTHIrZcSbMK_z`J~6TDJOsDTk~9GP<%5m~cd_HHH< zAXs5UI+$Xip)7EE9(R*l07ur~460NjC58#=r+5PO$aM;c+x6ED;_*GPT76UZ9 zR(&>&P>5_5HGEh@(Xjtu*P_j}Zsk*Z=$47ftHkU0y+_E^l6ElnGRa(< zpQBOpxK*&t{^aoz-BWv=WRgCk=JJMI0>9pWjgbLC2|c>Ufh=A>2f{Q^jIi|b5{Y$$$`)a=&`f0*M^&qE80*@pOqhr!S?F>T+ z8wYWe6f`6ke$4}BG9c^uoKzgol^E6c-J$U!xt$&@^@L|MAUqwW%p zo90`o9`R+Rc+6i^yzm)`1E?sOR~399`=k(Jl4CfI!OgWU^I~|rpHo5{StxOr4T;q( zVOYZBEWGcvU{Q`QBimK^UYumUL%V)2IEYJ9^ATf*8pRU9^MUI8uCWZ=)+5MzHd3;8Y`3ED&>D{^2h3BI(PXuiffq@K4c#ZPVkqydHxgYJWrD&%(OmPZMIML zOHYKxHd1>m`Q)U(+t1cD&#mt{Nja(gv94YJBks8ek3{&p@Z6C z!%>1Dkk#(?;ibf`yyQqRjlVpJ#%VuOKr!1zSqr_!ADf2Tgj#evW*^!P^ePpG08?jX z>d->XbDG1Zn@u6{TWpn~Zpjx{`lN;BjrzMt-+R0-)@|xCBS zJUQh=rwjoO;fu7e_jVwOf?_%T*(L&kk-ogr*m`S*p!62q@quPpGQKr0y0uvVz{m7U zIN6x1i@MfTe;8jy+iCZZ!&*3naZCs`KNdt3sA=l7TCzL(kHcCnpuOMvC4WfMI}E;6 zUdWGo`e>^8$UFX?j?}H;Ugiweoe+i~I6hyV$6)&lb)Fb|W~x8SGHxEg-~CX)i()g& zJv|?#u~{|Z5$*dza0qptb|~*ePSO4y@H#CxjLEuS){|&S<;^Im0O0hN`eCPVOz(A4 zC0-FlGBb57Lmo&r@Eeju_`kja@n>gdv${`_Gst;j*b$&_-<|BV5;y-2LB$TL2oaD= z*>_#!ZKOK;I&qDR7cX7#z^QNp9)X6RX+pi&_j|C=8VQVBkslyR>?i2a6rm3kDNHY! z@kX|%4SkOSb%sdsW+i#b71pJ@lG$5ulGUjLD0S3OL9yCQKpui7{5D|m*MRETd>L9% zI}O~`e2VQfWwRH_v<+k0GiF@Ek99VYC}~NHavyoqOE-T1z-EUk#-~Xlr+$%rnO#uz zIs2Xkw87}ivX~snu)UR#&29UJ6Q(9Kpb>SiY-5X6qmIk8g^6w6JuAmd0RZFcKl3Z} zaOgFTehAZNM{?S25 zEFH?Lp$GNAk|DZ1J+2?_iA&?fJ1(4&T)C`<#(~q!qWq^Jh<`=Q1Sx!$*83mL_*n%F zxhUTL-TLK$DdO2wwNySVIx0{b42&Kr`t{5G!bU*c)kYXSc`}70Lt`$gZ&M z@QxxTI$*)FrBrV@Y1Hi*I{bm{+uQ2;5r^*I=Y(W@cBb`^eB-siu_4%FNqUIL<^glX6zazNSl z@!ysRb(2Uj>R71g$!y=NJM6TSKB;IvCLQ8gJ2ttca1>FMYH!Sn>28tsB-X%Hi?C{7`-%4W_ycEGsqNf)4c`^ zFw&QTk>x^N-emWieu()DSLHiBclWx*$rt*Ju5Wj~`R81EIJ|U6nzv3jb@u zALHzt5YFn(?$6dZ+e@V25mkg=InKVc(9r8p?EWvJN%ezs`kGf zPik-Dqv-LwNUoG|IbQ&vB`Tgo=1P=?qEZW#WDC$|PiI>C!?9^doX_CtZe2foU{k@k zihy&Jq3*Jj6hZ_L9{5k_>xt$7)(BKgTisMwmB_>Dx+(h7i<2YdHgPgL* zv(N`}o(KKzS$87wcdtYFO7rM-zRb{}Enqi4ouBAF&c*kt9Du?{hO|rcea-}rw?dJq z`Tt)Q!1h|5B#eA3hhsck?@MV*8buJFB52mgzzDuJf+w165KgCzq27{G=yV|%yyTD( zZv9Mz2WrzOT3We+mKhu5U;It=$^7iVzOLy`sI@wq7VyP4p{_q+Yw1)ktI4>6^I^Z^ z<(dDDUIpl4h1fxrZuLaV%D9ivo$-Hv#w*lJo_~&$FWK^Rv19`v<0%Foyds;W)_m+{ zi2vjq7Sr<7t6MtlUP6>JCUNk)bT=iHMwm*a1p;B+344s*M;2bOd<-j-FrUhzQP~bE zLorEH>LG+(w6V^UVSEex7LA813oc&Y3DMSpwFN)9?G?IH8{uYaV}F!<$D}PR@}z(*;Y3UNDGqk zK7jc?P&d9C(Xn*yBUli%=n<`Ad_UO{gGEP02QJackv@Q+si@4O%ToOE2n=@6dRW2p zL{*tB$cj9fg;ow!IY)h_;czOFYcv~)F{+mIqoE^InmEQ32g#rvCgT;od)R#irz3FY z9<_Wlka(a8;e>@F=?*m4X;(E<&h}3|%mi6rZ(M!Kd3lWu%5U8L_Fosm+u6MnOZ{)s z>W>kOtkGyD)`0=>sMI^vQPm;0&a4UX-3^zhlDNCWJwkU}t#hu0Wg=G>*Z6J{($;^j z5VaS+f#?7R6+rp=VS>T9p%4-*{m;`MATLhkGx6uvw~2%0xl()dhtDIX9?iW5-dNmJ z2pe5^jDL`CDcZs~JB~+P9x~i+A;F>fvAzM{N9{ONH@la;r&IT4q}Oo33ga=b zR6E)j>3#+X_~s^)Cfiw9%t9N)qBqg1s?%!YM)hmsvb={4g&J*`nNa2S1{pd1uUVT< zH`nUz9B=~UX>h)(=Ks;a(SUG7v1OJKsu9II{RI)j(L=Twz$SZfDB3t#)KJR!nO^(e7t zkq>aBUB<~h;a+u$Y&(XUVUVB7Ulgk~%ZwqehzvFoz4O#Suo_2l^~IMeyL1+*o33va z%17DywbuT-)1*hOqZP0QC)dJPDsHYdxOzDM>l>>6QP*W~G#Z^2l318{zSjk^{oDj1 zAzsx#-X5fOxhm$b;(29@-+*#o(t%$`S>WzE-xpRBsDXoP*pS~V@;elwODbld#M{+_ z;83hzomD6%adP}+RSYMOsjaovLsJs;zD4+2W^sdem8TSPiiV|$I%UQ44dm)+2a zO?XH{Ea3_7PT}6Kp;K_Sluw(^BB8eS%j8`N{eOtKtHw4D?iSNHmyLo7z&A$vVf|_( zI@}AStn6)3o*K<4*!USVOwl0jxS2wWu0ZwFr(}(Mh-Q zsWTzhFU-Wts^YaT6xBst$3|39Y#`{mQ0$0$)wYMf)x6$+Gb2)$AHVkU(mK8U z9B%itxmI}1;eg&~FM;c(OplEamjR&#@$HuLkjMT*0S&nd#{}B_?H@l!qc;!YyzH_h z=5u~^zsKoYuy{S0KibsVs^L65j$W`DY;KN!@W*iasIS zEi5Ite=1LjqeqN4VuKD$(*IWg3HAL~{@fM?LNiJ7urhFgP?Gsb_v7oJvR8ZZp6ZdC zZxMm6n;I|ltveDg`B;(XY2&-}2`tn#2uJ@P}8gaN@X^8C;Ny!X&hT#i)I1$Q6L_L&!57?gKGX&j`F=sye%D(1O_4;MB!7cKw0oe zRy7aw1*^!}n?yZ(osy?Ri9473?h2n_XaF9PEs@7n&q(6qLB&IL%+^IY$6SFOZ$_=@ z3v72b_8`e5fZp8K$9mmAHENAv&6b|Vn-S7Z=zI1^4onD0@=UeXR0~&mW6Hxw zU+r*V69!n)yv48olAfMkS6yR{3TR7KoIY&>Bw8?U3LZy|KhI?TFsT2oHr)qI*9ClJ za49@Gx@b3HU^ZH(l$8#?(SXG_W2p@`$&-LHB^4*Ln>{ugCGoa|RYKt?%RcPVNm6b< zw;JoqoZxA#XMfiq6=>uW9VEUEeVO(Om^RD!H}=sKvDy zz0~d(m-JW`O1PJJC$V{EE+GltD8tR^u*|ovSYy2@9aQ!^CAj#JNGK5D-_TJnjcF}z zB%w&R7ZXr?b!5JH{#Rmuo9b2^anK*|L#Y36R&@63AC(bUq^rvg)L8L|83_>)mq9~_ zIOqK_WC-|~6r0~-ecI|a1jS2MX0YNZG&img3R9~RjQ!fu{6vNIE#6zAF_BP(iDzi# zR@H+Ln_5TpMX8NTX@gFaL3V_G!HA^*C$dTs{k3|U5?#`~vU;F&;gAjCUadsn9ICvV z6Rn)RTlfWNKa<2-*(!yEGuWjBKdzlvf!EU1pQ)cSkO|BfQArt`N!^I|K0*B%bSQZo z8zk1uPi%J4d8Kps<5=utw7#qt*>xAjXhAc>&-Iqha{KIlWC;qT2MUsFmz zBE&$qI8rd7`xhHm6cn!r(@Fs3ABUyF6h9i9cYLm&+_(i<-3HQMkDF+12JD_Jb1yxP z9^519biHb%28IA==g-ZH*-1jg-w^vm9wJ6&_pjSkK0`e2#wUA;rNvN`^zn@$Yy)gh zZchWQ7Kr!gQ=I=r;dOq6|9JcY#8UmpzQK<|0cv!=&E3WQOOmmdeZ-3X^cy>Xon0Hi zNoreH-4xs0Zki3EGfd;%oHby^G#oRNu1)<7??t~5gZafH8){N{ba1;{NdT&g^s$mp zLm@gh-K3EcS7hJ;Ts6@Pf!VsX?23;n=pL{i`r?0nCt}K)m^!W+4ml1slwoNB3UN$mE^U({?lq_J8uGWx_uhzaL|H(^|2MRY`_8j`8@7XzKblXfq z4b+}e?*eDLp=h#xLfAxO&(87iz8_W>oEH{zoGm`Ob`)j>t0O@xOXJ;=HhbzIVhwes z65iWJ3kRDb$r>DQp-_f|QKLI>^TU)x&Z9!&HtJ*)ucWm$2XygjTyL~B?qz@Xq{?Np zA8h7p4B+m76KA?p9VD^)q1Hm8RnKX-_B~QbjerIdV~0|k@YCd!e%;z__%Cvwa>G>r z$MuM7pzTXUza`snTK5g!0Ue6)$eiQG@$8GD`upG}ys}hpE0{MCEnjUg zW4E%9??^0>&>7F0BCm(43Ey_faJX*NMQ-XZ}FH=8Bvi+P>i7nSHYXA z#0vLXQWmfjL;HJ5=c_`%h_ewv=dm&GA|g&WcKKm+S5SC2sJxgOjmfhg9)I9NrTb!C zWuEpeYWaX<+#jF&r;kwpL|6%;%C7=@yDS5Y*0 zxrBPEiA{kHsr)jF6)^p)!`<6A#tYkUc2edzlRp|*&#Z*v z4^*=UbWI%~frvq`u&A7u2m)=#V$jN;KSZ$=!a$G02;!c3`pd;F?#w>>*!FZ}sjkYm z-U!_FV4S19Zew}xTVLGNES+9$uwo)+lzW=@eL=Sv)O$1KWA;Hrf{@*MDq@>*3wx5` zl?5d&TOu-5h8WE)aFzSum3^=ZKlkCM?UXiCquZz9WJ+>$S!-idqWY-MJI7dA&s4Cr zDI7KaN;9@>w{}2>vy0a#f?`vms_Br5LNsVmug{02Ve>|?xwiVoo+uNg?&t?2smdpg z98aGV0@q*G87H+w!$^K&rq(-md?U3j{E|TPU-~ZamD857aSyLAD@!O7ll8y)hhcNL zXJ|;pKeN`#I*`2%h^%SXC}Kpymb3fz#gISl1o|Pusw(pqok{{wN_jzE;ELQ{2a;w&@xu z#^(Iu!6?a4#H_B(Q6-& z?VR|&9dbshfF%E&vH9KpV~E82*`}^Hx!I^TI;d&(Er&27VFBdrrlDOsP2U3`So#WA<)6>NgL`gzvn|3!(O1=@` zW0upiyi~1LY|PHIL#VVW(k}u)re^H)6m))wYk!(^STHe^4g82$upp*7J(NV*^2^Bp zCSPQNQ(M(D^P}O>BHycuwaT^oA&+B`iWaHmDIy=OUQ1@_V+Q4{dZU}m!b)T1+)835 zW8Y~9?uw7?9^K<~>Jn=%n=vEArcc{Uv-^9ODo5uDUPt?H-WF#%8* zx`HU5U^6_}lWLWEo85Co^tFkOwoFX_t-&)2MorMH#>ZvmPvH3_gK@ynDmyBy3eMK+ z=DFj`6lg6}O9BETQqYSU*w}3~Q2@~SZ~Q8lQG~JrfyyNcEFHerTb$n4Q_%c~)_5zx z*arC5kC=u1;jAk4&wib}#cxIHD5?pjA`(ZIe?e`*+YZkt8WDe@X|}l~=4{$Ipg(yR z*{=fFHq}|BlaRrQj_fe)Vc6;od$Y^2UR@hPd$An*m=-f8VP7bIrzhaNBT=Kl*8Lm9 zh9bHBl?uPMvRXyi=DwIWZ%y925$4>5TomoAoz`Dqo<=duO8GT>)}nI!W}n79H{{VR zTCV?iok(C>m`<2S_`j9c`flt$={=50sB!eb-g&}XsvuLmHAL9t+{xk<{0KG$wLPw7 zWh^PwU<;K3j~QG9-pP((b-V(CF?yS;IYoF45{c%e+EBEq<+tK@ET&cx>8$0}^Tc#i zX7QmV;1CIvC6Z@G)4o&GtSC}DUcKrF$?j^5O#2CD*0%>=ka$?6eliEDSD8~($_I&b zMDgJS1HDTSyNz-4$vh#|Rk^#oM-rC;!X#DBhll~Th%JjpWhWTOjvwjpSy#(7tMM~h&BE&DHh(Jz~|Cwb1Tu0pyIDcvcub7fk$qr zYnEk^b^G)%(rv$QZLdmeHj-szxO5TW-BlA;Ss0=!))RqjqqOQhd3SpcELc_|YZrm^ z+vORP8oyA5IX4F_as*;Netbbf_alXGz`@x=HZf&M@kJU9W$@Z}{k^M$g*3x2g4n0x z6F&vn-%31hU-36bh2=K`CVRxQ|APmw$Pyuen^l=cL=^v53bnrnKrg{gsX@p}DQ|R! z>zztQWb&6I7gO3U%E!|ybbnow6*SF943-8Rcp0)0L*uG~!OvvMD zIXrUa{t@JW{F-S-R-;tI&IJ&F!(TK~y>1#wY(EtlU1X=GrgoKf3%Fm8n6YK4w>o87 zajzb^t&1^eGCe2upQ0EHvwDhSz6gsL3m2Q}5}K?^`S%L6__xh84+U~@*2IUJx?7W&amk>GS%IHM^F#7r zyS&BS!fCGFqZf~RO_es@+=YK}3jP7k($Z?K{#}G*Zs?r;ya?4!Ai*FQ^D@M=o+KDO z(yA*(9$w>n$*^@Lk_>u0mKdNJeJ#$2&+OK;UurUgCzcyX7sgF=i`AFMr!(+jdSaJD z&RmKG!%&jYYVTa*CJK`feMvXSsF4e2u2~lsZvb_i7gyAE&uJL5O3l1c-auuk{MQgZ zSyQH0voGyCMW=i(5rhQyhC|-{_9t&QqAwd@Z-R@#;qP~Q(?+CH1Tjje8Tr6HolJz% z0C*s6w=1C-JAN{=gwLk8F&>4&WXz=`lH1aI@hr_K3vLnjfrMK%xS(&6z*H^7T2j}X zNaCL&o+(fO(I*O@4AiiY+asle z^ZQnsi~w#Tkyi}P(wf4vU7hMiNM#yKzO{#Lqp5N92wUw6^X>IC=KgOLnH4$qL;qk= z@moU>N$}v0eTm8PV%gqPCfqoP7R%cTK4;__M3Yb4)3)xm!&zQTM(kg;YMD=4e!F&Y zA---`$GR=N6UNOFw&c=lv?ak7&~obHl}>E?}zgQDZTb`e1z zaewW!rN?Syx3|xQI`gN{9R`d#j&heV{!1AX|7F9J1DI%D`-dYfAcgz=LUuNkCrUYl z2~u_E8Qy$Zzf&L)Ld$V}Ozl4Wh$Ie7DDA|Gql5+`b;IUpZ`+kC-=X9^daLktu=6!B z5-26bB4?M;s3k1Ll3etZ6=eO+$sZ5;-70Bg;2Sl12QW27zfdy$q@Thj;VnH=!z4qlv2Y2gXOZwy8WWi=4eAs`t z&Tqhfwhrt8-s^%aYotn$FdTiOa5`2X)$g}6+F{{;c9?p2X#>9D=`90Q%r>7BALOyO znx0#X5Kx*Trj`O(g3}v2ld(PyQq zYY}Qf7`XL}n5qinb-P@dtQeEQ1Z=LM0K+IUdr$Z~C5R-2vX2$3WEa^0+XPqON~hs6*1(S?$zY(wveFKueS!sR5h zU?ppLhxfRRVxc}`lxYl$pZgI`cWRot5#;*4u+L0!-@7}__>(pES(^ORc~zFod%Uo& z)L}oP{~S9NY82t{vB>W3LhkC9#WVejYV+N2!t=j#GcGFEW+AIIym_$sVQ>Z^=q1B7VYGrUYGXdF2my7 zM>E`_DA5z`7TlrAY|7DFXdOd=Ov#rGz{U!spqgOyiGyC$_7Dq`z;7B$RWVb50H!@$ zJndr3{sf-b{X})Iuxt33i8r!ulrNDe3-wj}*t7XXw^v6AY@fItp6}Jc8-Yyb97-T9-u6DLhP9S4-S2?Hb0u zF;H%B>P29YyU*YXfD2aNY7b9Y)=f2xKs_nD>!Wj91KjoM6w<=P%#V`9X7`E5u^4W)djBV>e|1$3y4@e-u)XoSKX>NMMI~AasX6F#j>{8 zfr?1%2^?fXOAWm^;T}htg<<7GRrv;f1+C0XC{3j{KvdmS6`hz0FD%jWovhr)g5)Aa zAY4g0uz_w`eUr(4LaWAXQcF#tC=xB+*&?WnAs`BQ>EXt6T{VXKE%hQ}MG{W*_p#-q z?IM;)(EZ0lNQQSoPB6~ZN_oaFhf8%>*u|jSMaoc&;kXWwwBInMTmPn6OJfzLYU9`J z3(FneOOoFJ)0Nc9yI@r5b}wp7Mj|AkI4>GF7*4#8*P(>q1RiVW>lD@}_kZvN?@zz{ zAc{T^pd4!@e$DfAWwXV-09rVad@(S|$K3Ds^eIM3?@iS;jqCJH_@8B-=$+y*SSyFt zcAmIr;IZKV?Pts=oPi z(r`@C1AWV7t-mV()r&h!y)1NM#-3%NKDlNJjp;b#l>c5V!)rYmmXg@IVEEsmfr<9Z zyf5%}JL4j_5~&aEat04@I!@oiUNtv1-nYZopE5T=)iCSeTuCjlt@AdLy=Jv~H&aPp z17m4HI%vNu>=5AzDbjL*aLhBjsUp37ZM~IJH3!33E5ae^30^F@-37sWl~?-QbExaO zg+4?hv&FFoi;uPvj@#o>0v@xHGo{b@TytU5yV}0ot4nWNa&0P43l|D+m^&_H*Uc=Y zS(1AEgJQre?y)$vD`UeWUVm*_*Fg5^I;rXcmT;#@Qrp*~%F-v#&@@HO6O-XLdXPS@!APNZ($=U1kqX0nbM6;0W{|ZsJmk}70CV~5I_XLsSpN}|84pK3 zz?doY@^WCv{fX*XIYwF+-RU+;&(lCQIEwKV_PF@D(pV|;rI zQg>LJ6#pug>yjozqyVli^xC;X_=#D;SOsw=_H}*?>`bRo1Qf@2gw{YHu-ZB?;ByqC zRdq+b7zM#GwX{o++d4}S-S{-&uRNMAO(Goxh)xWrmK>=~7@-DA(*tSAtZjp7o)eNF zDC!4}XaWgGLLrN&6UjG$4a}8pCn~gciF~RilHDKI9hf23D$iFLLC8Ue`uifp!M?}% z!aFsSkqoQVNZ#}e2xV8ooj>9R&S}2Yfy+1t53=st1g_<;@QuCAk}H3;OfT8&5UDre z?&|GPu4!bgkiC$c@wO8di>6Qx#l=>porxBDW5Xk>A@z|h&j>K4;rtH1ZgLVXL}|?$ z&&VqkIl(@v-hx`7O+la@tnWqWAo4Xclb?9n+KV4x}p@5Yd7d_TP z@6rx%Ef6y~-TPg9^^?7oW)21z385t4C7%Zn42jT0ij#z1OF9wetq&9Q7KHYX7|JB~ zo(B#NNq&Rv8yhT;i-T?1!u&i*b(Z4{tCBl;0E=CrR`CzyB)dEzy3~MqfoOgVzB;p| zQC1uUT$!!wL@K)*e|WJMA9FHpG&QLw!O={M&^IQLh-?LRUFH!^-tPI9#B@ZqA|1w) zyC_*Cs~)#M1bqmD>6}dC#<{KzYO@<6)JE~|uikkXMdBilQWB4a>@V~&Z6Fm-K-$$T zDcb-Y2ZJWxrEx02l>tnQQC4J05na0d@DO~MJPH5@X*ba{cucoDd$AcN7{MD+J?G+n zmH7Nq%0fiy+5G$EQ65fz&PSLuX6gOJJ`3-!NV((>98Nnz&aT3kxO-5#wczz*;G2LV_UIn}bGtChu^9cMqa9J|BKb)mesf26$?Lbcbhq*A^y_52sXBC|_nm~l^Y3tf z79{UcRh-@Pp#RjButI`&!7QGym>xJ9H580P`M|>YvNZj|-Frqj4~$O;e8YZRuhQBQ z&L(cm4ySah|839#*u}9CgBE~PJb&g!32yYKU5x!D%N?;3t0bk}`f*2TLu9KJ`blYP z_zLK@hyyz}286`5)r3R2rv@Ur+yEt>yVK6l<^AtdV-ESx7H6gm=O;RksTHEG>x&+J zBSJ>REF`6{OD@mg)mG!t+HK;koQv*>lZS_NTvzg+fuPp>??Av8TQ|aa8>Z*O6_zZS zFn38uX0zoDU2JtclsrCZwp)mLY$?R4(5Ct|KQt)}kIXMV@Gr9Uq{3SQ2<4)?dCS6Lt1J^{Gra;1no7n$eioE?#duJRZY5a_}|^b`WU=kGK=d_Q#bXzMy)l{ z)@?+g!;!3gAAg9Mu|*y$V*$D`zW2~|B+Wh}G$opo0y+;MuVwhI_mgGT!@Q@D6L19_ zf)k(x>7TS9*l@0v9ULHMG@mgFx@~Jm(H9zBIuU;~EPaVVJ>8W6X%E>6r`;=unLKi6 zxDGx);EDSjLO1rsDD|$9zL}Zq#qT?t^__XJzl0~%ZTygD&Ft`8&~`tlX@>}oBr!oK zZTMW0=zc};921j$^Fru$aeB67b+!m5I!l)>UP;_M+fj_s5}~dn@lxt}XGu7=s8ROr z@UlfR2|kG5p~E-nE431OhL%cs>X+0lyzthZKRNs`kjzNrwyON!5rO|ZA}}GpP8b`p z+X@~>fiv}*vkI|OeC1ue1hitwpdPwqmRj;v6PaufLEX%S|452HK^{F2@IcZcIdy+C zhss6sRoXBt9j!*^MRH1E%gexl32t}+U|C5dZHQPe*h;)l7cf*L%adqBrPW}%5&X9A zNtz=7m=orxQ%DM5qoRa|#rMu_yHlEmgHSBfzlg zUGcrK#4KlJ&nlcJDP$g|7J3g4PtK-B)?pwh_5*cIvBYnpN!hC3&ggb?-MQ>yy&e>2 zXT%op)WXhzO{3>5aQRDeVvTFxW?^@)E>WqrECsmjQ4#L7*Oy(5j&R%)1KCT=_)(Fs z_jEf75R0z^An|xs6#Ee3k!a-EI7=zZbuY|>l~-l5OY9K{}Qtmbs7lhH8wSV#RY2am{h;523*(1 zyC>qO4+6bDddaq&t}=4lt{9VD6>+LGe=`uL{j`W~nQVH%c*mxnx|K*Mvz(8cgNvikJ+G+JK()` zKDzwNv@i9w>`$6JBF|ALn?(=WkUpc2F!bu3k_v|k`>6RLO{9wRA!T7#K>%FQA6Zm{ z86e!7ZrGqnraIYqm`2_^%Utzb9-LB9KM-;~Q$cd0W>B!6r$sS>H1(f0xFz04lVl zX8fhVxR9G*>eb6heV#&=wxQ1n8p95;OTB7%XM%4C_pG$G%hL z@-p=4^Tm9jKY1rZ=64{m@AckgxDT{wqR%PZnH+tow+KR{S*|X|&|gCpWB|*TqsnX? zAK2-vrK_rmx~6ccJT>WneXd~t5G_Ri(j$u#;N&>!rHPR{ByZS6?)m{Ua%%JE$Fd)?rt8E z97QX}w0MtBmdGq%{gE!+ovP8ks9*D7u@{m_Q9cK1HzAN|r8+5hF7iBS&HN4mc63w< z-nT!y+^BZunXg#LJ)&#K(~yhDj=K}N{`VKhatgg&b~D;EDYQx6RLYofvKK`&{Lg;-G&vnZ zBnh;puL@aH>J_O>wiLTK8RE+I{HH3OZ%7+J%u#Z@5?VU$ggJYm{#wwM zkQ3d^4?W7i@nwFN%a57?w*rzeersH{wFfiNDVEYrlv=bk48(kuU~X&B0bT zQUI2th$WAeVjrCqCISW1ma0B`^prQH|H>GF8J98}KG@GMICg1jC}UH}oFs}d3;=_e zBpBHc-Pn#F{EoW6MV5eRR*-$Ctnz_o$`d^JXt2#2(P<=2=z=bN&%f zQ4d?7PII8YWqEp525m;fp0YOUO-@oBFCYU$GsMlR(k7nf9rXwsG3W zGWTR6{W~R0T~*j|eTQ{F<9~*AKX33f>kNx)A$pm(Tc#i{zrY^=!G_bs@+%wk!O9r9 zE)C+;2Cr57SCJhfL0Wz9a*ngw-*B9EoYXic689&_POI1CI|sj~=Cus! zlcC6qk9GnbY@z7BKGse9?Mb>dQ5!=%+~NmC5xK*$GqzMyPI5U6D^J+IfPEIT5?C!L z-vy%F#T)EVrV)gZsZuPF2-WyJJ4>clk}QIHRh&W}Nm9-EjgRv$$XF0FmF9ZUja4#l zTcPGvRC48N6mN@pWMden*ZOcxhvw_CY?3~`LS7(~8$5?A=Yd=PRW?Mj{6PKe>Ug;G z-Ni31GLpp0@5f8lX1wDg6DuKv?J_v!WI1!YjXqNML;95l09?<@PO`=CTT%!))*r;C ziUaBq3l6CJPxIMfG-6`m$^CvRaiUTdrJ4S0O6ZCK)6ZA$Fh0uMNFUo(J>_~@w}8`e z3gkhhGj`2dSM_AsXpK>1bL0O|Af}b#vyT5fuw&`}{Oe(j8_6uMr!mZ0o|Y95cr>fi zG2lvwUf2pGnhoWMry=o4;D$82)$ziGxiYHFaH&wv%J z2N~uFvO-+d=rNR&A#*Z5gK;!(f#I-VmGC4);CafA(WaAq)v5rRE~^Qt{6%+FN<ceT_E$$U` zV-Bvve3KGelJyOy4E%}E=ACtelv%*R*y<|47c#A?S}4WyAKZyPpqtTtE#j=^nN61o!yh z^7__u9ZiT%t6-%jXfd@a{c&Izz+#cM|0U{rOK=mn-(yVSFRU~x@xQTH@TM3b(+bao z0S;oSLYsq>^HGSt-;Blg=`OS7%Kzzf=8$>io>YuFfJsCioyA6qp8oJ>v)r_J7bdEg z2a(~%(WL}l4hA#p)1iC|BCyKq#Gi>ZXE5+(Kpax1Li$w=+}k<%K@%Xlsg9m@HFpD5 z$_Io3ow|aTH5{!k@=XFUzpwB;6Wl}9u2~~<6Faj%q*4;?yqK|f6Jf_XgUS?sX5AEa z5M}(mwH#WYNOv{Qii#v+w_h3p(!w+y4$PU506Da42 zN-QG8gFVBt-%K-r?fi&W%CO2{*b2|{2KS`A|Lc8oWPt%221FsCkhIOiQ8eqP$C(1svS4r-LqTM1u8dxA_TIz0N@|Ry!@iq7BCXpeQ>W3&9 zPlHntBoYv$BqvRe&A|eL4>^SPB6R~xhSEAQs_=))hXF4v^oZhZ$(vnaCYD%n{Lj9a zpO}@Ia<#M~*JyKO>C|Y1ThKH#LoX%EN>0Mr-PA*j1CtnkXOapkeTB&U4&fp`0|{Qd zeBs}r2UATxs#^PR5#Bt`5%6cc0#r43cxSmXUgpWay~|$e2}*?Wt-b4KmGS1p%6a&# z=1KJKx}nT@{#=c1m2gi?pgFRuYliPK8X`Wkh^bEkWE0L0i!;PoyLb5$b4&VM|A&K4tN#jwGBHZ8%jhycW^uw;hSIj!s}U zl&7`e3m&ZYcz!wPM!;p^Xq1JLuNDnjSE7~rWJ23Gla%~=!$-8pu9Mo>oi%%Ve9SAuf9_5wfej;QK`H?ryuMOS}{^}6tIda$aiOSG& z8_I<>&*A+x9p^aeoP>8)Vo7P7@;;u8;q7oJjjFklNRe4nb=nKj`+gQ@cd@v`8d2Jv zLk=^^5$dRuFTej_+Sm+Bd_c}Va_L{=CU0;RBOFr)U;51*0jxbB5%W?`Jw#D(%ZPU$ z{+R_tFZ3Ske;A3AbBRBpzHu;tw|RZ+YD>FzP3eq-u^#Y)*hrI=C!&#}29WwzsC}UT z5LrgMmJvFcwJ(Ya`9W|)qo*?9k56T?1TaF6uYp7kJ7V1l(f6z57*y>?YV zZx`CAc;>OoM^*pEmNp`vgXfw45~!tKzgeyK8bG>5jn6upheJ4Z;38qS|9=~ zU&4m#q>^$B5}dRA!Eza7P-`wJUiXa}lgODrkyIE_*$rM?VEBC7FDvvTzx*Hj&y|(2 zsZ};jp=y{UN1K~;X1*P(W{ATgJN{0vBtO(tps-|sYMa468fFi=&&_)-14m1CD6vi? zk3(7;^{ca&^s5yR-sWL<t6cOfT0l6g3`)ogJ_izMLl2ESRC+20Y%D2s@wZY6 zH6oIOr2P&eM z>PEAh9Y(i?w2ami;0>k9yjeasJlsUbz$tFg6|o9)+A)HDK~>d2Kf9Dr9{ef|a26V! zWR|SPDH}(EiomPk9D84%;rXDv@`zCkgGpuaHxpg=Ek#J1n1LTZlAHGQ)JVc7;ztP8 z9dmRunDHAGzrSDR05eLJx=qh|5mVG?x;F~Mey3k&eCQ!PXj?ww_YJL<>XbOLteC{M ztDAml6LB3>L^2#?5u-|ffl!7pN6l_F1YWM&cX&(&I88AV``eZMnk3-5`pNa&D}|(P zqyh~FqJ~lZx$O%{wJj&ih~EFoU!MK;Q6#p#>mB%KOeC~1`DMqmP$OsTbywqxcD=XF z{?ChfCKll9Ac6y`He=X4Qe?34=@IYhzf??0cKD(Bds%YhCA_FpbZ3!NdtXGTA zW?W500i=7)FY$Wl`-r~`EXnO^l{C)emJEKE1Tgn;Sm`0C@4~)(07SMP!(;P=sM|>p z4&ofzc}dx-DJKSWp9o}XeoB-o2v)~_`)155uJ@Dmn-rUh(;weV`M{VArOb7&7U0wl zIT3OLI3gZNA`+hVo>p1nO~Exa->2LBKq2qT$GD`xGiW!2;x1fF3aVDYx!*`FLnJWX z$;cgDRsN)sIw?a^hpKAXM;p}1+l!2$ZBsy%et!-e4FB_;x`d?$!qu9a^wNjF)i2SL zQGf{IfQj4LL@x971(B>do_uh6US&_m8d~o}I1*~2xLyVL4?`CUj32fEe78r|P1(M=pk&1tP?3{YH7(e2`Q>Oq#7rzq0QZ-UUU(wY7%|S7f~?Qdo6*!$6a@a-8b; zniWU7JjljgWojN~aF%jQ3^x6)=&2--!l{2O&-5lRA>%412sGB1GJO4mq#RsbEb;WZ z0hk}Uphvh_rWUddWLau5(6AWbxT1eaaHc?LCh)|%bXU@L--lDKQ+$#s&C>(bG`8)_ zy8rEOM{VIZj!#WrQZ$p8BieuK0AmnFCe^;S>{R8alSo>fexZ8?oZXk>=jCYGH6!TY zrCt_%m%wxN9~JDQTjmR&vdn5$8r1hBY@BEkb=w*uH$#0?r?$=Up>M3_)%>Ens1C03 z7XB+e$GU34zhe7G4uvTF?@lNnmXOHm&DyG0XbHyb^p<0OxojvC49768@H$i%NQ=v$ z=dH3D9_8ofBAvnq`9x!+Vu3$azsrJ=$B+YwY3a1f!QiDs3O}m4gPhrc#uQ9c+?X_d zM3uS2FmV5pGZK@!CmN%L@)@B|hVaJX0Js3HDuXlE}>9J*WeZ>vkCBHoTR!ju? z=mtTR*ZVR!zt}`{b;HlqXx)5p<=kb@VmDt8AnFOR92hou-mLGS3(NflYbtNSdk5>2 z%ai>A{sKEp+FP>yU~QVnRj6bfJ{GQ+H@LP1FBO}9B7V8Mx;AJKy*q0>@OD_NYXI4E zfg1-zy{}poMn>?IWZ1r&CIYIlTFD6`$ftI0aYWd^Y&Jf*@Aw?1vn&qQihp$oS#K8j zC<%hR!)zGfJ8IJ2B*ba*aTqw-Q20i)Qdi?~Yu-(#gLOZ9QPbve7ADNyVm=s zZ2*p$Gb&DpN*G?5yKr|nlg*_3ICA`p?(wVT&{%~1xe2wG*+2% zdz=aLl(#5CqgaGRn!T8qO*#2!vK%6Sixv^siS}`-n_lSBeQ~Gt$mwx{O?sgv?x0+; zRG!zfBlw&`eXGZdOllhWKtB{`&bz?(P$YlW#9~s3)$r5EXNGBYr#cW`6&~30CBly3 zgVVJttAh@_-TbG)@l=K&`6R>Tv~gJABt34<?y5Oi$1~90l@7~@002YgOeu$IOt)n>WIXmHGe?QC!d=JvJeAFN!kLRQurV7cKCm=APC7_TBLGLb zgfUDyjG=P~k$mZBJy?2u98eoULa|BS+keL4qAxHDt`n1c25QV^BL)cLR$VVmHgDpv zq#?Zea4|EPcn6-ut(SVzjh^Lt|C#7V0Ra!~Aq^$&@w7mI%8??7f)wLq zTou1WwYC!$nHSs1)K|Du;I7=)Qy0kuM{#P8DEBn5H)ameidPGW06APKa$I6xxA0lNOldwJCvT#r`JYBc|?sjw3jbU;%VY| zl>G=C{T<59F`?;|N`AqU&8C;)jJNJ;rNJ!@0?^*K9u_l1Hm925 z{7Ao_Mb@1bVra26TY|?88QxI(R?a}|-9%F}{A}O+6!_cA zfP+%~I7|9)WoG+5#`X2u8%Yt%hQ!czX#%aLALMfM*h|&X>&-G^l~b}!j|){g&GJ=t zS)U2(w2eucJ=CkCt$LK4BTG=!Uv-QC;lx=hlQq^q*T3 z%u*3_obwbJwa!i{%NQb=e}m|NIcUra7xDC2s2wtf@ymoxi*o$v8xU zBdbB}R|%kbKIoFZ$EAJusfcKm8nX-~Qv3Vx^Mx4K0IuHZ$JB>(kM$#j_VdkOYnxNU zwz{0~X~Xxny&29!jM{SWis(7k>pHD^`JU}igk@LCR`6I7xVsGzg-D;_e3foIp?67P zqmMnIfX_F+<#La;%uSUmP1Xx^H>WaA(W+`?&JMG;Vk{wr^&avp5bVqPNj;zfP zds^Zc7MVkx(rigOBoqW4X&H1yT3aR0&Lb2DMO0L7u2N$E-4nsj?IFT1{>j-8Q~A@W zX@ac&CS*68D1(+SN_&9?7MPjQXcxRXd=gdR@*ANc5>64l%(*l2F~uwyrkx5`{gi?B zLwIdYQgLJ}Qqe~P^{BFs$|kdO?Cdl9w}#*oq0IRG4utl6yo+CKpAo#v**}tDhXrKZz@5~NUu>s$59h)#I5re52KKb?`SH(UF_^Q3>$9Od;V1z z9iGzDYnex6W7IlV7K4{)(+*9JACL*sx%A;xg*JHROSdKplOwgVzMbxxgfYeysUn?- zm_oAkKe)*e)U@pMn%4b9gkZNTJs|c1cEXN+jw;K?J8EYrHz`MoQN3*$1x#uA^Pa*U{lmX-lEHu9%`e50>2}yY(q=_Z2GFCJ z?lO6nm)p;R{Vvq*v!AE3XUmhh8$ZYva`A9-+Se8-EzI5!dZ5}ji5twHIF$ap-`CP8 zVWVAfQ}TLncSOD#O*`x~30!=N*{MpA?cDLSWBEA^X{f8SER}04*KWmu3k$`$2Bqm& zj~cXFWz&gF6?Ph}wTD$R#cFj&4c*F0|>F3m&M z_BJ!)=yFHJo6l1-$py-(d7(NyI5n-NoLBy`;_Oo(02gj~N|+xD`bME7D>MM{Vo7n-}?#mzzSJ5}cn* zPLSxCKqfMAv6LTS6?)DfCjw-3G?dQJ45;;-w$yxO((*WB7nKMYhJS}&YU}e9b@j9V zS#G^w@w2B;nQ_iYJ1hY1FiM`XDPC7Cvpb!Mx9K>>j}0MIwl`B}v25-z2C^|+V=wTN zrZX}-GgB%nja*?4?X>J-;ZU(!sg?Q2GpuhK1K85gWaO~2Kf>MrnrN!$6W8BogYW@k z=OcHeg$W|f zsN;$2HmJg=}ir@fbQE^7WXm>DL#LwcMRY^|2!SASpGjxhv-DU zxlT?!44osh02gSi>f$Jq&+L#xyT$YRb-TH!-#K~qG_pFh3L@aNS2z4Kt4f`2^lRLu zGL;oGj4K7?$4`ms+%%HEvIep}kzn2E-dvr6rq7y=QieniJm01qYdN{}94)ulkd|Jv zP6v&$E6gG}L9#@Bm)dVW)3(UB&LXk7-X3U~x-OctJ%+(z;J4_t_-E@QR`CErPY+dP zG96srbfdTl9cfSJ_b=CkV(%T6-eY*djenqD5fZ$~y94T{wtC_<>c+M`d9JrJJh*%~ z8f@&#C!KA$t4NuxOGFO&Vq>>npwf5l<8iL&ie&aS>6p`H6H$HFfZ7c6y4|6%^6}U7 zQN?;$2H2D)G78GY!PytofGYcoO#kJrF{FjT`zo)>z^4MMW-ZGUfC= zNSay$ZmPT<5>Cpu*Dia@hQ;UVj<7%dnGjAi5*t44@DRYlh+)x_M&JHMZJ-xVcyDsDy|7bi$9Dwwh~ClJOKD8= zIHVpnz(+4qSwufKl)4~VPw#Haq+Qo+`-&HN zA~s2xvKoX9TZb&^q4c;5xHtvkjBA9;_h_dDut*F-Le?@rqT&UbM7q=qOXxO8af(0$FwGmuCt=ddJ$-@>Q+6{(x? zoI0jPCA@m|D!Hb}C^sxWVBxpWv`s(|97w#=(#qdi)T}jXy)c~N51=F@KTL@LDMV9z zKr(Cgb!+&6Fdt`c-?Vb;yYzN76rtRs7I}MBHQCdl$uCGL){fXo0~14_p=`he9!MiKQNYUYXGr4DhP_UBYs}+h<#T<;j80gfODJHYGnS}~U(b_!8y~8d;;!Lbam8(a`BbmAl=qG@l zGZB@*XT_xkVx^VFQN4vq)6^aw#in8Rk(PnoUR7`-f=3QiR( z30h^1!BVd5Y7MWx>xY`!aQ}+z@ZnR0^i-kG=^{;|dXOJNyccN>u*IiGCz%!_XK`CY z9!|Q#QmUE4gZ7p9g!In>v{)d!5_~%3RdpTwV~|Tcr5Nh~f-LBw!P`Qg@5c;5PpJE< zkr)8k%r!^#J8!%Y-v?!9-i0!5OulokrQu-0-A^ijVdIzFutxU)2n~^O?`xZsMEA#e z&BNuG#6&6d{Uo2{x)7AXr@u*T^qz` zf(L1{&(7a{FSku8$DYTNE>56Lc1dAywOrm7*WziCKmznkN;AM^5Ld|l16C%|}m1*?c@2_Y1lIA18 z+&pX9Vbe>5*?YSL{6iab?<;?)&ig{Y6A#g$qQv@`v4N-cqswzZhA8EeP@(sJhO7cEr zFe~csXQ7Wr)Ym2g%2co}wwQqJSL#$t0#m=E4=y#j%n4rY!V0)|FjBaa(v&wNjmNpY zug=}CsGqCN`ihJg*C_V^i#EaCXf)Pc!6sIl8j@OS+NeX;#yZ*wp%#H$dtYarX#>S< zg$)&iW0d1sj<4=Q4sdq{glT@bQARb`o!&q?%wh&P74R@4@AEOj)02-wMVhU@a=jM* z-gE!GK&>;0PgE`jTfG`gWyP-P(+hvDDDV-GG*fUW(e+Aw>vcRw&~f#A!T~j$tP(#T z{xP}R^(SYgPwSbRc-->-0^{%Am`FQ$yp;s@2E`CmMqcfQ0$c8C9cCxWb*Gu$v8F}o z!c0xOF4=XpnonGEvHmE$@lCooT6g;4v5&BDq^zIRNtT&V{NJw@pUe^xcCzFJDeE)_ zk|VoZeG6<01Ufe0O`Rk;Q=M`)plM|an+aWHPvWRc;B1zG-iZ#df6p@P0T6fz026pc zFFjmhIqWWZOLf+nVlT1`=V}hk9XH%NNYfK@o_I3~ zHIPO`wm@~`>kSa=9oc0(ZPJ>6Vo>_tWTrTr1^IDGYrKK7XY;S9H)D#0817V>;yEb! zRO}Ox+9|W@ktzT)NOa1WZ7a59@1w%cU_2ON2$*4UZR8DsQX>wZ0G z{+HBz!}qhv;9F`%H_Cyi!`iYI%T0Kx@sMQ4WrTdl6o%ytQFVTIITe0%CHZTHAAxdo zh#$GVNT4I=zloxk$_KPq)>Q6VaO*lou@6jyf|Y5IAR-TH@wmKcG1R?VEWW+nj^x)L zMVr3Ias!TcIsAag%8c=6<-o=C8gTsM+x`t3hpw;J?Y#NCqOHf>;Xjnsy23XelPz<- zepcfo7(P6~r@=Ohy*m0icH-YSu}O!wu#6sxLqC?#isa!)-;FKs0r>g1ocrapFmur3 zJ111)uEEEPxpSUyOtI5-8Dom6HOV0@P4p&+amubQ&R%+MHK~iY+6{pV=N&vCvNo6LAVaknT^Nt^y>SSr9^lpHG zru{KuP3?!u{KyM~{p918cDGrBiqzYuvqwcE){u^iF1e7&8HR zTEDF91W?mdH4Gk^TW)J9=eEgXNZZL@lzx$&O)K*65jk5Ig(ldAIoH**ZeIvEmR-XR zFjESH7|+!S*O6^{4;*MCmX?mxw}U&*!OEY=(&P*7>=Qo@4T3x`x?Qi{>lW>5)4=B) zvD?im-ITMZz(1WLt!b!xyKTPIr~K;xja|bebm#wvF_4D24`D1asdB*%WmR>c;li^B zgYJg4PeNob%8Z4yhebgTPG8GZ-E;&9weHh}tEYZ-J_dnIM2N>obN=f6ka zDMm4&mf${Z_IC`-slBNCLi#9la-XYvuFr)fi8v&a zN3SsiaSD1w*h0jnL_XC?anW!tB(Yf`CgtSU}uxN(wCw|RGL zM<~~zA`u0OQVq}a$Z|>4zQ4G&N+NuBeofzw{FB-LX`Y;J2k|#)aOTgV1Cw)t6a`H6 z8MQ@4CnEjly}h7*Qtz+OW|*A1QM%R_hTkpJp~c09{T&dsvnDv;o=eGcjb;VwY{!HD zQ|GeRaf%(RmD+d--@>1V3p}VeoL&7HPYjE&B>lt)YkQc94S|^=hQCUq{C~K53%97=sB3uW5|Bnx=|;L?1nGtu zx{>Y%X%Lm}kd|iX?(Xi8?(S~B(ckmD*Z02H{U4ljpMBO|d+oJ?OTv3%M-kM5SMdoE zq-G*I#fTC_-Nns4QyFwHAn?y_WIA5C!mLbr+Qu{dHFaxcFgfKLA>}Sxe81t^PNgD% zWP_GYfqC9uo>%l9XHKHMU9o4?-qR0fsIEE=hb5-4nIQ_BWD}d#^ae`Akq848RaYo` zaXlO8EBqLz-SsE>&k@Y6j|nqd520D{*3*+5PTaoRrwjW-uYURb_+DV-MOKq4#RXgMa+n+h6}Q+D1TzNS6|BsD zcX&ADYFAj`We`=*qN9fE?H8+h{OEF~Lk4EfK?)>gT$p$3j$@xqq|!+d2oOnvxlXhpH9u8M zM@&|Efzs)O;9tt$8jlD&eo+%<4+OtmjP{dT`!ZLF3s;E2fMR~$qEl3`dhjs?GfeD* zFxk<&wzIv2o21V@s0FxaxZq|6FrgnuBlwozezW0;uN{|Z^lRj8!8My71=H_^yFx*H{;?HQAxFCD6j|k7LFko;hxh@!G^EqVSq$^HeS~X2dVWi88ecIzUB9 z7?nap{Hf72!W#Yie)?DmTi;1sts$CK+ z_v5uTnSlU2<8|h{eM$73h3kQv=8R9g&KQ{)SbRuGv2IpqRxUm7-5Za9F4mn z{3d3jsbYMg0T(*$gSNRL>Tny1;6FnTYD=zbS-)6OS20F%F;F}DXn4sxY}VYl(dy6* z;FWTlJ>6`fGs-?l5GvVQcD0|0NBd9TwL#wNS2m&SuSoddDcAj9tF^z(e7iP+HSOun ze`*NtLy9xYhvwY6N^b9y*D+z1P_@13Lu=MQEYdfBkctlK(S>?(mFUOvszNDFtJz}K zjzQLTqfR#ni~AjNMaj4>PSE{>ArZX-S(XiYpYM9wrbDzfHMP}U)zxnJmg`PhcGHhvY*(Co+b*A( zU$i%F7vpm^6Q6vVDw|q%9!!QW=;+F2ZguCX-qD_ZhFANP$d4%!z3xG+epFTGYd%!> z8|?xluw1FaZDXeQRGa2=?#a!6?P3n9{?cb}Q`g1&W*b&%<~5&cdxCyxTDMy_&F^5j z8S$B1{1l&{V(rPm6iktu2N8?sER(^yazO$dY_I5sEIrj6XQw#b3@nQ?4yXpCha@e1 zP=qNUx-)q z+0;Xk9c9heT{Opg>Yucqp8A`^Bey#IjDR!C0^D6{pZF-yvnrucg;{Wr(m!=$%MxtG zCXCZ#hRQDc+rCDsWj+Ke4k2D_-vQuYk0cp-wKZ50bvH!1+@H$MNJ>OB;74hU{ObB1i>ot>Kl0G?IIF0XC!c>DSnzV{j26sim`YG==*qM==B zF@7b>G+4zM(%%vVBpO6?&fLd~;qIjC6uuB<;X*O0h#Y;Ct}qrf_8VC1H5DFTLFRqf ziyabug174Y{PD|A`S{Ux^$cR#o4HX|V?^1xM<>Rcx(uOmgyKS-g|2C5Li5UHLO7)w zQkTvep#Oiz9~wJ`I{X8zK-4|XfUzTlie@;F4I^{^3xi_vXNK(yfo&H)8EiIP69t5$ z&kdG#P9bIz22o6YGH7$fjn2d)pP2Fr;zVOB{p|G!$?K-+fOw%f%5Ru+gkiue?d)M0 z?F{Oa*w&Qy(;e(ztmLd;UcPF93`xHe$KU^vyt_qu!Zs`OG?~z-i~d%WP-{7)sb$le z^1S0LRb#`E)n^OsG4{*MavLv(DXi+8UB4lO(Q~>eVkiA_ha)Xi9~)Jco3(jBLU-So zc#hrmC2u0_GNdE$gFFjSKbLM9Tq0C+=l43<`poNxM(TSS=31UE3y|2pE(r`gsdVsL zU=cv5@kzMM8WWGtrN=XpVG0f~N$V`A>t7bX^R7>-LMgV3Wrt2unoOD12~@_6$e-Zr z)0d3X#f+ym5kSNTGH|Vwu$MiJ-7c%JU4CALsf<}_rJuVp{WcHOE)KgKLyFo2y~owY z5-)yCI-&BT<4Kju@b=H=5Jvvd*f=m@pXcq#!tk&zBWofP`gnCJ{h}-Y+~&RG^my}2 zU^GkSMx`N}7P{_h7F+hxnv188k{ECdUs`9PjDSf-uWe$Hawm%dq`tqvZX0%ErWV%4 zfM2ZAz+J9#-Da_UIdKrgJoR&fHxsAciEE_VjsA23HxTmjGk*@n^L`g+gA&n=$AVR^ z>d=lS{G>8$VdxVYupU31HrdZ6_)9de>J+bY(#G!y2W?06r1JRO6xY`q0oTVt=2$bQ zimj38gl*c4I)c@^MAxG$9)DP63R#qb&oo%?zSALOLfSA@H%gcE+W<-~xOKjPuppGr zzK^{W4;oU*&pxXUvYn*bIN?*xq52-SMniX6m02o+r}{eH26d(|SKBO)QLDER-}aTE z!Lvd?u`6K_{yP{*{|J9;AYM$yV+C@-TeuzU)GdBR>NU}d&~EbzWr>g9OPh)n@Qz8^ zI}jS4=ke4L`>E)1{fXOVH2okZ@R|X%w1x-R@~~2790}yi8ehHMl#wHQivCW;!wYS4 z{ICJPEoCn+CB==>Ohl@cO!WIiGN1$z(q)?|^T%9u7GjcuP+&=UB|rg0b}NF6`6nqf zAUGe3%NI6`48*Udw(v=Y8+b@Go#z~A=aCV3esRG3OrNPap)_29->dqIqgwI`;MT5{-v{aq{YSGd9$C)Sqa&tRQ1=xjgU|p|z z&P3R-f^HVM6N z2ob9nhR|PDN3Fx*ZC6ICvG2p~>wwS0C{jZEvJ2_QY}Bz7*06;hGy)rI*f|F;vD6xo z=HfdMO%<;3&)8CUnz<6IIOj(p$_!*I83XfD7)3LAxD{HXn;l_(5KoTU#zrH}o#$NV z9z7!M@>M=3xXgo6I0!uTc0^hwf!@8CzzaJ*_?;0G7Gpxc))m-cP#Z;46^4@igLGe z=WwF^^lIxdv0OSHs?%f>n#wChnlSCjaH{|jCwaM6KFC17{?jhpx14;* zHL~?vqZmIIMf6Y}XNE(Bv^oYI{XDK(1X1Ww2XnTk^db+sGDJSyH~v*!!rxjTM22S-vY<@-@7e(zcuuu4$T?0 zt$cRg7qW^8kn5?`>zQM7P*aVjcZ5`)7!rmT_lVacfK+)FZfyRbbaE6Nzd+81Y` zM)wW4~(%LbDqer6s!=4Ii$A(P{uVE!)S75%Rul)+_b zuK}}RQkBCoclW}@pSf|AY~3T=7I&iIBLy|V<-l=a*@2AX@dqpF1qHU9ympS6sQ|H( z0|`b16-ffTv9NVVlW*KEW#^xB_u7i4_G66k(Ez`3bp+x^DG1ltskH5FsAB5_rJ*#B z|A`82D8ITGzgk}ompWZzm4cqLiTMj6biXZ9%|o{hkCeuqK|`npTwv#R+v zhW++^9w%H8TNzVnob~VF=3GA0C;)>ZClV-p^Wk)Qw)EAYhWnA!#C>qdi4LVit_y+6}n)s=*j`3v0&35A1bV`h|+JRmx6eunYB% zviVcoF;G_u7FdT@$#?!B4vt&v`1xWB+-fxy4Ny4iAG?fJ!#0Or94FRl*jLS~5y{_7{@P7+wM9TC_e*#+P_F(4=1spcAGObaDNc7JO0y{+ln1(r1-QAi|$#%B6YX0AsKtBB>P8`?UX z_jsg$+8>JRI4wOsTx5O_A7ysSd<;yt6Z4MrjLzIgn2W9Rrmx?y|NH=eLlU{Yu@#5HXCfinvC zKUEvi|F4@Wxp3yexep3PF%)G)EvKdK=)0LlI?XsAeJV2A4mwoRZTV66M(4HTY+uU< zx^{*&^kGlh3G^w!5GhS9kEv{&G8~^xU^3-^SC-+1#UoS3^P6q{ITUl&%h}@G1p=9l zr|>%)AI5k1CFiHT>pcxujpR;XN2p=|yvk-@3Z3?VT*tY)s8iWu+{==Zmop{>VWov1 z(GI6=_}x6ORB3@X?4m+w^|m9Qjx&Sf8PhA|42kB*D+yKe)4C0hhIjexhZ~R`@KrSr z))Nrq=cL}u+Jj{`RU#1Yt@8k&SM*w3mUI`|Fz0=Vdj0E40kX zr&<8;TIyvVrh8e|Gpa1@p_JY3=Bq)mb*eC*;p+VcAATR_Tq&)(JR8L4ZhwamoUFpx zmWAGm;x}c9Qh~xP&-LUNK~BG5*Xr|ih!E_1KZ@@8E$=?B@jx+@ObA~XDs|J#UwKdl zKxQqlA~61uHhuIS(4ync<~~Zq_8VgVh37=1()Tgn*Fwa17bitGh_%U;4-TID2%+nuyB;Ba~ zQol+1%Io9RrxKx;(#qOYP-~2JQZedQ8$Ya7%Ny*oByIaHlv4*lr7o7Q)EwDpOViRYdqL6Z`Ip<=Q5^~&At=;0$CGSsjqJz)j= z6D7efL2nyLroFSKRx0rg@*|&TsJ!=lAC$k~jCdPi{N8O#E^@ z4|lihP+bB5LiWn{J=q`j?-`@|3sT6`GA&sjXj=Rrde=P~0{!rzkc%gmqBx37na;Ao z-4GA>D-@<8K_%*aoOS%$4^}|1oGsY%5Qvw$k!*r@P1OR2`c6&!J0pgyL^2l^-Spvg zcC*!@Q>tZ)fXjtZ;w%?1yAq2T9%Ulwvg&kkyaf;^eV6A`q*bvUgw$PX$kwqezF2Ag zB|J>Oe0a|TA0?zUYw=Oprhi&GpdLn{oPZks{T>Y{nOU#!5CqQ1h4vQDLN+j9Ny*94 zEQUn9_f>5-WOcRQ&NFm=^^HhB^*z^@To;l#h2z5|27lt}GRjsh!uRJyB6Swc^x2%O z#`A;^m`&9lHnXu5U8qkXFDPC00a|01B=y9M;!S_$t8DR$p(g zjh9iUL|?CcWVj3VDqh?p=~`}$g?w-WQ|T@sAjdD>zi)W$MUhPd8a(sBz1+eW%E0-W zNWd({lxv&G07crGptKJXc=%w}S2A=sIxI2n;F{LBf~DIxXG4~u@C%JWv4K#g#Rmg; z(|*B)JFvv_IonXF`}a2ND$J61EVbGaJ$Jbs&4AP94(t=e>3sLjL$VMs!Ppw z(<$?FiSVKUNjqjm! zo9d0osK$XNd!%^Xx7BrYubrFw<@YXy-TxW|CC3Ybux1x+Xz9OeQ7Eqsr+@V$FXEcK}%INerzH zpYAR6oKgL?!*YT;W0u+AV%N~^t3GM8?)FSmPf#<%fM(!91J~rDl?hyzLB((oTzzR~ zsx)yt*fKZ>!qw@rkbNClKUOv|q^^fXX@rc?vYJcY$qG>;nv zo)s~L4p8C{kSfLZua!n^dAiaHy?kcTsL%A_RO9JZU@k%GBbs+k5n3RgE&f}m2TK3C zbn|_ynrCcn`+lp@q+&JCYRd$tcT}-~!;rq=Z3~ih_5S4{SMS z%7yP7_3TH>v%9~3{i7@U)kS@sYqhz8Q$yxU6Y)!J?iN3o3`yqwL}Y^UQcNFU?9x{I z@=p`*zp9+xvZ`Od`5(*S{ZAxmhF;*fCM)L}f?qtBD~%~ z?YJ%IY*!(ypvEzct>NGV)285E-9DxbGW_K4m?pzhN&Nu2PHJ0zow~-nk#Oxt1#gtx zLCHP4sy#?BdtxaBiA5@5Cw%%EGUr77xlQ`(&xqCC{4+n54RVD;;xqrREu6k3w2jRe~r+K+_(Xt9YG`S|x3G+yswse7qWw?_F z4g_B078bTiT)l&1Y`#!j`_#uAl&geFP#%mb0~^0+t*%kfL(23%9)?$1UiKt+`;))` zp^l`ZN73l~!dWYGk1liSm#Dsht>&6H{WJE)u7VAlW$V!}TO?wKItVHtzeJ)FD+hY&1s~Pdy2hn2Lgz%L9#uAqj13@buq$vjRi3i;v4z-vi}#QhbMHA(*KQj zl(fiO8K;-wh4I69rAb03a7ts#GLY_ruHRBAstbLPa54x$(k&98F%%D8l{w}tg#txZc{`O!6VPuwe zE4|Q)qiJxuXqb(0A-K-&SA@-~KWPLDVZf^uHXjoBy;%nzS7Ry7*c3Hd%3j`1*yGKM zGdh3BWDtN=;tDuyD`tXJzyB?f9fac6c0$&#%1>_&3uWex+1y(}7k;0KD&kp0995ytxK z{7*_d{g<-{=X;U&0glm zfN>6I5Iwzp;xZDh!>vYCd$p;>Wegbhm|jU&wbnUJSP4vzPBR*8SHGsat9IKlCu-Fb z37RCfJ^fH~zxE__A*D^nV|li80~C~JG5A^@529}rtss3C?HE3fZ-g__#9C+gU{CRO)y63$RPv8JP|C;{GE7rs_f>ZRF1G&Ec)c;BLIIzHD~hB|40_}AiP!4)~QC^aLT-%mBa9jp9Q>OII%tS2^Z^3w3da>$v<1{%o zW+hstrMm61ogsqM<^kTxJlo4ZZnLs_ClnI{lMANxceN`DWi>BP&OM-20S8Y^A`|%U zwEw4J?ZN;_8(a;*DF~09B7j`T50`cytkq}t{~3=>fnOVb8v^G}Pb z`?Y-YV%Np@eM^!H!;nZb7-!EsN50oAVIy6n!t%1}bqh9(xNOzbSUGjQkU^TW3 zx9uj31PNODF#ndmK*#5NjrUb(m$jmTShv_03s@B^EvrH+_0L#M88vW9Gm4fc4JA-m z`s>SY(Tsoor>UQzQ?VFze|}$frV|;gp&r5+qk?*I9aWz7U0m4Q&WsdH{G;rmy5lK* zZ_`unvuNEWkHrLlU<@4_-dZd;9XWqZ6rC}Lvgir@h{?uQFj52bL8nlsQ;)F8MqdCb zsrZ_Z1ZKB^kBm^{)85dvvweXuH0wA5~xc@jK$>1CUJ2g#H+z3qZp_yB-O~xfZ zc7K##y2@KeE`V^VMO&@?QI;qwE}dUmVoqszWlYIP>vQ@%b zYuPvUT1qWR;}KAAG)%!mh1M@>2`A3|g`w}#>Wx?&bh4ZAqj1iMGqEx;FEoL+*5TH3 zB}32ER=jPLKhWc{H>)a2u$UsRnTqcF4TV~ilZMLrWu(9qRXGRxgId68XTv}i=>VGh z(vl@QPjSXFCC&TQ2yfa2a>g%5!|witiPvGJFF8`@f4WST$k+)^e>Xv2&Evo~#Q!sZ z2>BNb#Hh60LH(@eykvFRmYs)Fo$csg&)`%1kbXZ%bEu}@5-PDR2CodYk`hZI9uUU| z!$RIp$C!HreCgoacqj|Tr=>`P%cC0O@0A%3N2EIomg}RArL>|+%ny}e$UTHi_eAeI zx~W7bR0mH($2iWf5ZTAktY8G#`^?06HCdQA)$HohPs*)e(`MN)^7RbQ*rt}8I!KH` zMsyHf4*-&|b0Q$G>5}pJeY`|5T;YY8Mj##C;&zF)j0KzM`}h;(*Lz7JOeOJ!tkUGB z&68{mQ|JA64^ZCvS?CD5Am)sO=SB6r(R+atsl2_yyo1^R(iFG%zT3l6#X7B{;?3Yx zG`Dm(t-f$!H$cfeXnXQZv`!@iFLPu_3=P*xONFm?ATmFXBFKxN>mFBrub>@Yw{3rKTq95qxjZih?mXUqQ;R!8SO{mGhp@(h z!e~v^E|^pm?Z*&&5&O1fsGTC|VQMlnQ`tgS147fQdBSruv!k9ku^vgq&ZcP#{209t zUxK;V<)fC$W4D!jds!`qyuiXk`zV!67P4DzR8n{2^B5|l`rEY?p z{Z{7!=x?*_N{g!|&qnR*ltsKB7?q?eduqvB6!atJ5-g@L3WI2av?z_Q7gFOPk;Ndr za+yNtykfWE>{})n_5DYbBy~ZD(Fpwq-4x-=$Akg>4NbG+Uy#a@%9l4v2pZH@T=czE zFsD=mZ&d$PE+q7RsRW zYSKrN{k>iyTdK!T1WSn;^Vu0=6Zag9yuVQpm(NXA(bQ5|aCq^j;l@R`?|h4)kKOCt zjblRDP@`n;I=9RBS#x$SSRg!*+V}C@{j$hLw*{Y_1*4LS)4;;DM{jfc-CJImS|;T$ za0Qd1Jjd0@mCc;5QGnUKZix0AwUNgT?nruf&`-DLIt+#V1`nWY-$Ng8=`yTzK>WQ4 z32D}%qsbGg%9Ex%Ou@oxEP;fRGgWv|G)2IODSGpBP3z+VD$(;)(dFpO#EI2$NzO_x zH;+y)*91vu09OMby7yM?G?FHp^R(W`xjv|*1ogy*#G5``Vrx#H&xX=tC;gqul1!zlRf3{@Xeuyv1p5bjSakzvEN)|&vYNfbS*7WA<4=W9iY2SI2=Ab{?jJ^+< z2Y2SPdLlxItrh$7r;4;q=Hj-mE4|5rk|S0B>jhw06JqVt?^8s7cg>KX>Q?bJ&TDgL zC9n}mNVGlMuEn*N1@M0V-ZlSzJH-2Sx|wepPj&xZfxY2(MRig7focRm4n8C~Xf#g&!n=3Bo^Ij$@9s@lFyR z<+Od#XU#Z(U23|w6>7s^6dP^cJB7V5YmNt}2eKbz0k=bqY*(Kqx= z{Z?Z!oEUkno@c9ds*rvVwY;+j2@Q>zuxC3+QcY4@O(_W%lrl*t4^T15`}3tj+yP|> zxe7L;nJfP0C3(zg8&q4Bi>s4jtUpR>eJxpcphh?g)v_G&{dlAUfFC`5u(x_I>7nzx z@?Bb=_tGN!}S6MUOmZG23t z4k$SZzck2a^vj*1?K>9G=k43b)jroD4m8E+`FtNs?%j(!f2c?Ifq*Klr>7>Qy71ms zmE!eoBH(#v$j{Ud{J`BEEA^DI(qh3;1=~99$y8X-UVh?L38#dYt;+ibcMqB%v(>a& z2OW_>ek$VAW&0*Cbn{sG;E!$RscJrFBEN;vdWfT2y>uE5PiFc9zxP=J#GB>%r1yit z)-rr&ZvJ;ufrUCg(B#40`I+^2o%VH>!EMz!VepUZY(%)ipCAJLnu@uV%=ei?2PA`E z+))ZxD^UtWQHp}t)%v~`mDuGy&JI{2n;^T?Qz{{lha=3#-}+kcl<9UMxCfQs^;qD+ z++`#_GVX1(a5^S{$k}lD#;NmGD{$!o1^Du?kPmHdFzW|CBqQ8BmvBIJlc?Hk)Fg0{j zvbXzZ>-J%^#Kdg5FyqBg%3?(k=2IkoqBTnBOnUSAR4bOyI2;^=d&udb_K6oNGI&&f z`+)X8^KD+KfAj5mKgiwXDI(?5#&4t~kXmOJQsz@bsoUCiQFpj112K5-xkr&{`&dbX3}QE#NK7~aeG#lcIsCo-OZ zb$94$p(vd@-V7?o8&Sr`nKbqT?P@d&jYPSm2|nX#1j&l|n`LVHsrxyO2-Nd5_wgPU zSCDP<2QV?UY^c=ZRk^4fab@P%TKUBK*lAhoPVe{J1=U1+tfP;rP6WHnM155UI~Hw4)sHr94jo&b=A#=;G{8InE}I}v@Mtz(J=qaSMd>MI zb#|M!IlW%EiWDVq5Fb}&Sjg9z)sUBE6L19i_*FDiJ$E5p-1HR=j2&h-lHKvAXj#vL z6N=F8RMGnUVN?Ef3QsJU#7Z;Y%Vu~aQ;pGDCp?M9Z{;df3@FyiQdmq;(-5n=CgOO_XBU)$s0+AI&ZcL>!# zy0aWvZ!?>Xl6jy%gE_O;vbIYW)bIKKMP z%`>X-QKL=Cg1@({C~S4({@fTPa|6wUe#TO$8sv#@#b5*)2t*FkqDicI+T$e##uM_G zV6!X`q1rMGG3=6Lc%f`n=hT4X_f5GUI|B-ouN55wCECChg_Bcvw<4kC;r65L)CLl6 z)NbY)$`u+>=0ikbjm$cE(`HdZ>M*+!gr(3NR9n2;$ePZlB}(|4ei^y<8Y$w(&=?>g zHoeJi>~RsoL*BdUnTqfGWfrHA+2VHvq@*H6^nc6nMW;R*pfE8{;z)|ED{Js&VNvn8 z=vJ1yO2;Ks>(QdGMSGZ9i%$Eqdz!Y^j;=|>gv^9b>|VAju7RFGe9GsWMe~e9xEmoJ zAw(sYuV$H+ZMtX$$$)~d4g^4O^-5zA8B|Qj$Q~p*_VFKZ7a8qB1c`?g)yrb4oxAw) zTh6BOdV|W4GB_TEidYiDqBe7b-O(F(s|PQ zdTI}I_xmw{h8Tr$>w?>Ry_(eX9&acI25kB60j>Zt^t`GYS4MVmj0BAd88|>9m=8vA z*Dvjw<4;)D-0e&~lsKapAC%Kz_eD)t(?hW52u%?>v`f_`+B&vxv#N6YE%{uZd ztR~9u;0BySR~EX6RCr&k-H~ z!ef23ci0`zXR~d(H7XE_7R(Kdr{0<+8LDuF@l&~FA+m3Cj#CoU_AP7e`Y;dW4qh>p z+j#s-2qGdgp8_B6;f#8>(=Q&nVwUxO#)b0R?Y!8Xp;W)`8Qq?V5sFHLpRjt1G(_V5 zvGg@Sd7R=iXi+fe#;Ge9XQQ~I@C6r1_a+r8)*%1qP+ zIzp}lW1DJaoO*Du4g$rnYN>M3!p4&qJV?2!6r*O|@8hZr*>N2B^3VXxdS;%82NM{V0JHD~sPmKjU>%ZZO7w=5rh z>3?2Oz*@r*q#nkOuJl$JTeP@^=u?kdMyUWG>337eY>;%`U@;_t|A;_la+XX6;fp?z z8^Joe-voBMWj(hW^XpkCI5n14b1#9*k6V#8+v|_BI)z!|tCbtwKXu*rsZ;oGxvrQ1 z%4c7q9x<=jdoL2Gfqy)cL6};MlPcxc)J=-bhh&V!8FB zGmlrdaWO$S{Eo+1vhkp3dn#q|1<;c(gM(8UqRrB?A(LlsgIpi`4T)fU>%EX5OLcI~ zjCF^dlMYpDY-!y)ZM4u_L~oubu7gSixVM(H4&6w+aYn@!#1$r-$4()aT%Kb)WN%q{ zLLGnM_TMG;hcC?k5~)59b{JBjHF7c6<{2g}JLOZuXP!N;j_NV-_~85Io#|i*wEYWv z+r%9RCJNnplM9#Bk)bLq18Jc#alZKO`BzgbqW}C-2`FMsC9KF6he?|#vwvkiEkW=% zAz#+6&@6^rK`@Obbw0E|evmPz?nx&7=3skWpA^}8{qSeGr{G|wSM7GTh-na>KuP8C z;Bu)oZP~r z@r=4Ari`@8|7uacR(1Tsqo{B zzf(*Q<=^=6Bd5T}>?k_RBXoao<1kv@UFK;Cb>hN0Hwu#j;)xd56;uxM!9&?m*31TZ+t((*tC+vN0G? z8ZT7K1J|a9K-XzxgkGR^05$kfL;#7u`B>rv@I7f~@J`=8;SXH2(lOEAq0+{BNpP>i z81A%!dn)4w&(-ny*Th!YiZ*1A+zu?~YuMYw#e^Oeu(FdpWZM-oHP~-c=W<_53)u8V zyu&fs5{;76l?@p%tsuJn$_2=ow)}&C;$_4m{Me6g6)moszF&yS#sser{&{z!gp2G&TAX>{&VSZRb!+v9R5k=iaASh0Hi|tT`XJ_rcj$Sh8OfKG<(& zjiLP3kO9a^<0sT4KDo*!BxJ5wGEtIKwlBzVc77py*I+qw5E4JcSdSKM@~e1;$})fa zg+wgnk}ja}DT1<`GEHDTjl1a+L)%9zB*+qlppHp%1Z%#aMbSf!6ZCco+s)rV36WQv z60Qa0K5+q0rm@p1R@Mb$bQ#??`BVr$QbopjgiS7>lNhjkS>>dm{GqvYgj_)F*cfI- zpyfy48iu_(^f>G&f2(%K+!p=JBgevFKyCwe3Wz`dd&IbSgdQM`s z%)!g;bI%zEp0>Qp6Tunqg{SXuoE=_iIB(_w>+wpyFA)s5NaLPElAsp`!lGZh&!LZy z*$F0X{t6EgYR^JF1zb~)uJrFY3gSIm?{tI0%retoXVqd23xbOcdOv-*|U$Z+tW86su9}y0fVvG_riM zh&2>SE|C#c>Tt0*WHNbk$#{FL?iQ>~7tl8eHS_PtVz@ErH2mK&&gj3(UJh+nMyowO zN_$fKlC{+?SG4UM7NTe5dpna{Q@%b*Lw&sCRexY1#v>tj~#4ffLwi^pN(N^YV4q1a&C!x=&xOE{b`3adsM zSuGa3$Iy`37xS5ZqgVPBxzq+&^V$2CvBlS{=s}IeO#N%E9)@k`Dk(On5_L@qJfsW7 z=&`_1Y(9-E%?K zbdtK&uh(HGzJ9%TT8(lBTqpE&!+`nu3Jn?LTN8d!Y#fxgx!^=yKS7#2P6e#_PK(yd=ZC zTAt!*9yGGRGB7#W!L;6~5~-OylAL95@q<(Zq_sT-r13uEEgc)a`<4hm0>#5XWB}kA zUL|pTI#s^Rke>BDfgM7_#!0@k!$(9#EBd8ZtV>85e}Gl1;4vf>Unbj#p~*+F5r;*} zQ+^dzM-yj{VUSY4R(iAxmZf^Tr=31{Jt&h|5wVhCZIMD_=(;_tM39UZfI_yYrJ|%+ z@zroz1+SRd*2h3yFk9 zpi1TCxviG6wN6i0lNn3!o#roCan{q9;FU7`B-58ab$WZl3dy_Nc`UW=o8qHiN63?T z5J1lWb?;DW z4tFJse$3%g@zcNhU9WBg+)P1Aj@xCf?C^Zb3N2x^<2j?aAR&?9_=RM_S4)nldbP2a(pLWgcQvinj>QTsHf_?E$K#ofC`P|Mc3Kb9SJ)QD6KKK1RX&KT-zCmHs@~QMNn-AO6j8PES?{o~=A2`ow*WLB;zvnJ*ySs!(u&c04-O7ik({GGg z;X**xzcYb#tfd+@x|lQIVhFW`)T=*RyemHuYPuNkto?_`ub?1CTv@r!R@LMMvB8Q(b?zYqKXk2&%O@ex(l$ zVpuCZo|(#qQs#14k9O{u%zJnx8^R}Do<$n0q^aWZ%@KRgWuU{_?%IGN5&3r~EvsG6 zf;CxT#~G*v8}DuyEm$BR%bP&VQv>B1tbzwFKS<&zneoZK{*QNYgFb(C{X(td3jbT9 zGoif>!j2&Tf@{}{)=2hq(oRCuQD^+xAa>Q$fI+W!?$*SMcp>z6T=^_Wj^SGks7bVZ z1EIRJoMn>mEqfY1g^E6+RiLPFq{t3d^OtpY%nuqTy;O*Ksl8DPfrpeUu~=t(GU@Md z@*zArzUlway7!JlYmS7TpiE~Nx+P4Zr?l!mA!oK}rQ7wk|k?3tfxK(vO^9N&W zCF&tI+6iMeaM!lJ1@uzrW0S6*V%kKvf4w2-1P$=To^w;npVW z(D=L$JT@|46*={F+I%y_xi?hm`HWR7I{L@2a?VMq47dp6<}c}>DBUb32z89`KT@O9 z5CO7rJMT?$KBZjhe)3fph~q^q{m3h|9l02%>IK0@8$7oD5 zAe#6H+U>Hm7{t=Hy;0(Vy=BNm^{EZv!Yh3xQOb2!G+;K`-+s{og0nhc3-2^_?L+%M zqA?(~9c_T9VM1cWFW~t-5I9QN(rdul-f%sckk;Z4(|LYOr1>CBvL)|6u;hCZdqdJ-zKk`srTJ~>^^9V&D1 zKvw|?3?DKGbY8DB@_zr1^62+}B@_{41D`-eTdED(*Yb)591ZosF!(aP1GmSCBnGKC zvly!Q|EMVLyw>Ng`6gi$UlK&J;41UvSeW`9Wei5m#L*874P|>;ZX#5|#%Qs^La9`^ zRDmtkBS)zqU5{#|I0&ITQXT9r}maw5{N^t~w zk@5BUPRp79kF2lYiZbll9a>O8=>|n=C;{nK>5dsX1nH7)7%Vyj29TC!hVBNHPU&t& zq`T`p>ieGco$svY2h3V?*WOp{h33i?bfv4$zi;ABmsY;N8|WLzFlKJfc%LrDS=rBy zvqA9o=f)U+3m^}C^fVP4&+RXdcg`tsUkWaySYEP{u_{(4S@PY)$9Nu1%Nowg6>7$7 zyXUn}pbjY!+_tW9L|p}%-}k)zPQ{*KkSyjeDAy^s%^@{DR>+M}*DXUYCQ?$5rCm*u ztw;vZ)!S$U=kl_}2*1sH+{b+Wx{J8db14D_?z}$~BX7!|+LomNHFgOtzB*9IuvYD; z1TPv5*TY!+o2p$?q994c0>GGa#l%bbXR>Uq3MC;8vKkM<*}H=88}=%{OH5fJiBO?iEF}w6nCMDwhoop(*q*Cm`U5~8 zK0-09;kv4JuYc2U63%+C@wf3?7!py>Jwxi1&SYB@)%TL_+jElb zFW80mi=C;hv5cc@NdX-)iL+o00Gmf1-?RGX&Im&PgZccu2YrH6j6tX0OC_WC#MMzp zHSAj@07YY`1tYM=M>oeeXu=)HKdUO_(#p8D#|3Oo;;-!HA*fJ3f`S>-IKM2n{>t{nF@tnjTa{ zTFuCH9Bdnnh49ryGfqwo2I87u76n3rv_px41!;luwilxJ`LomYMre_7HHM1I-PwWo zJ7FEPI`91Fgb+oGlrL%PMr#(X;qo{l?Oj`QDj>QtpC;yXxIS8e4>aAaZm28wuhm)q zfv7qufPcA)GKB4=oGTlR8k@!gH!MaWz%mM$$d%kF+Q&_Qg3#aeCs^J}R*vBJ#pDk~ z0qL-Hw<#hQ*z10)2COl;z~F9;as4#30)6a|4G<(MCDL%3X;liYv-E<8!z}Eis$K9% zg*bce`wOM(irYHz;Hbo<=p?+|o?K=>i6T+Y;5yhg%$|WxIA|oPi6W=`r=n>gT%NQ@ zOau!Ctx~=(VMIrUID1>|k0Vn~nrzP~NF)78NCsuD`+?!xd1bY-Tl|2o^2mo zT2oRZ%DUDNCg48>OvwOI7CiVh@gI`)^>0{XFm1s|bYGA>?N>oAlD&1^9iPWDdyM4o ziArvmziK;zwqOvLc6*T`8Z`HzVr8s6o18*0zCs*NTAwcVnXU~E)Z(>T?s7;O>o4th zUopZAhlcB~lv;=hrbB<|P^K|9xzr@Q304RZkBTLJ<{sIelsb{G%4BalT7F^R$J6 zXccu%n1P&kW^dx~zE}tf4dF#;1YAW0ZeF8+JsYp{_7C|YE_)q@f&K1SRa_r2Bge0yKR&29v+Um;u%=&R?&x~3_k8%bP@Zq4#@M7%zkR`7=vo|Lpeq-gK!%`w5VDp;<>W z$s+ify7oHsZlWU2q+C1Qr<**MZ-liw7xxXS4IBhrlM$UM4_>oV5v7#k1IfTbZI_^R5RF25PP{9RX3!p6Hk)20&H74hJ5Fd}YymPL=HBP`Q>JlO+Ksxnb*={+84t4&l7vu~f#DZ3T#xd>YRWjx zk8t3}mBp`@Pfo|LJa#T0ghQJcjpaoPq)!@6%&by{oX1>X$F*d7&*D?!UV~OHU*!Yw zTE6omtrs}0n{e7gbgXo$za=K|-ZM1Fid@J2@byE3Yh$iPnru&$VdK|7*s2(TI4a#Q zrmVEo<6%f_OSQa*6TT%Wm8nfcKC^4UZcNtFgxj`i=ol(z;mVOmCe=Y(7ENIjr9e;+ zMo{~PPD~P^3$-m`YYX!#D`rBnK#LQ;y+C-)j1afJtSdBGZ5(=Gf68&QL}$1_tjY0a zQumdG9sF)^lYOz)CwP}Kz1_>>9(Wx4#hcqx=XL();|-B$g!<1UfqdSh;acvDD!#7r^D*>SI{NM7 z7a6UAT|se1iF1P4@2*Cy2P2Iv2*B1QyNc-TT%$vF>(P~;Dk!(d1b>rf3`}v7plRtQ zcg77W!s2%I+PU9~{dFHT(5e1kICv`kuLs&PQ&Ei_-KE|2Ra-oHY;|HL^;sxAH?bsC zR3v)+!b4!KkPtp9SWrzi3_ViGSkQxP|KUgJ@(hPpv=}_}P zKWbVzq2dvYiL)!-75|;@LzeyTPfGq#`K?6Kux9oMdHgNX6F25&>wLfJ zIiOzSM*Qg)ATJh1|BmZ$zN8h%RFwL>KI(H|Jl|9$#X=}i=t$#rc(JmTyr0!lH1it* zy+*pNRXTVLZ#F`??|u^HZVB{0vC2xQTKWAU&3`8=t?n34M2q-?cNXdtJ#$c8>~!9I z*lRdQ8h52D%=4pBEpvE6VLX{hH3Jx?80wTbJJ-{>YIKlt2`HbRR4*kuyt{b<2u7hD zq+UMP_)4@_j7ns!LBphqMimR_)^_IZ;|a(Y9OXnRbLh~8ieJ5Vj;IUIqL{ijg)|cn z*hmgzqwCTurbRt>W?1$xzC(&bop$vfOJ8M+W!zCQYy+y(cfQcieXZSQ;P)=6l4nxd z0^t=jXQ(0Z7*wB6z});0Szr@KUY`imw>f~`KwFYz`cE-pPDGe4qql(tLu7XVT9hTp z#m44WMY=G=fNpn-gCgA~`=+UQeFB5MbK~=fvQSH`EYujx&2m!Ei zdX3LygkW@lo6@DDsWm&_a#$o>cy9}3b`0UZfz9BG)Fkc z6l3y?aY$pOSYK;63TeW0XHtEoSdL$RWV*XhjOy83shXqU$BQiAg|XJNjh!2y=bP@X z^_D86o2d#fL>(02cOQL!KARa{zbwA&|jkU+E64NSh zW8Pn-uUD7;0rx%lj);Q0?H*O1br~3iuU`qoldBypz7Pvc0x-z&*8GhX?KG*BPzi!71J&Et?WBB zC1?wiY3w3iy5jwg;q>MY)zJ=E(ElgZA!-j$d2fa$-H(hLCP+L{%yDbi_xYSHy%CaTF(*0gxLsZjUQ=nr?jqN9 z7;S5+4;CnwSoIrT&-C_umA8I=2dFYT|AHSReSq;}C z71iovIzzIyYK=uJ`PAWy=mj%e4ob=QyCZf#zu%(_)Z z7GOnX;4x><|9M*i0B;LX;LIzw`%#XOEJtTgZl~E~%E^a{R)xBA%ro*UoHKjZkr9XP zB`PSt^i!8{7CMryKwWKrP}>n-O7}u_M9~ujXWzPu$I3O9=-FW+2ZsE6Su%QBGInZR zNb%sxi)eAjk~CiUx#|I8gy+3ru=7dn6M-+BgVgO&=D}N}4N(P)Q2$fhj&OLR#L8q% zZiuGWFp=CXPK$TxavpaudmE(hho~ruyQ7j9#&3v~mm`UY1e;~o4da|>8PeJ z4#)WLgWt5$l1~r_vG=z>?1Zcuv+wg2Kk6vI^tT&P69w3HAi)R(H;nX@0wl1n2bf?0 z*0Rz_o5_SFEX6^n?H!{uDQ-RPUOUup^T*CtKq~rs`;l1(9rtln%Z3{FZ0(a4Uk}QT z!eC@~Q}>Tq^2GxL2wV>~GSn^mh_&##zHY%O`@oo=h0TL>&nl2Q%JVR(i^>4Vp@{Qp zMgniBpwSD*{{p{c|Cner2vNYQ!jg>sedeYB$-hlrwf0gJNt{429WQ*Tg{zK!qdz}jNammoe8JnS? z#>&ARncBj*ldHWJPaQ2%f0q{(PJG6EwzHFose|2xOlJWgd)QkQjyx*+sds3mkI~qF z@+u3_RIsn>|9txtwyGr=IxER@6Vh1qBpl4Oa{3cPH9!sDbHL6KVR-hd4=fq5y=A+| zz9D?Ec6r9~vH~C#?&xxAt*?78LNTJXPb;+KE}UjfBMVR6|Y$88<=`n>lQGUKCVP<@U#+Wge4>ilSc2h?sn z;A8lX=eF)aO|BB_hu4&F#)Yl+&^V7})FeluB)TM+0b3?SV4|@at0<1Ft*{JX%fIBH zmx>)@Q<<3>g9X3acLU{BTjd;yvmA`BwDO~G-p4ej;q2((FsvE-d^d~+!G|GylIv<% z-}e+D2mJXd?Rq;oBO#Cq2`h))%KAeLj!rUbT=w$kj+k0XtX^W@vo>8DPp0|%+JY0y zO}Ui5*dM2&x?0UqW5P>fE5^8{X3ho&g<`t7%>v>t3AJtZYQ2Or@|)Y(J0>ILnQq0O zC|}o>S!vQ{;QuM^fvn1VouReqqly%drJ&xP@hb+f-v!%8+zbxzI*b_-|g`B@#N5G+hol+4Fu!ZJGCP+F}lVOs#$B%6G-_<$c|gAj^H=A5XOm-{&vHT9CucUjH;Y^OaywH6ZR6kMeY$qVb*t!G zJx4e7g{VPxrQ%wjLZ4w2+!Rj^5hvQdt8X7tZ#EdQLri(07cI1Qe4j0P_nnVhD%|QZ zR_3OjSMM_2*~`Dm^kX4=DTv>;$f;i?PTrq&0%~7Hn?-zW@3)E$=c_{Ms9F8ii5egL zw}sVaJC`c#d6MNlClqD&~4RMEC4_}iXzv`pw1Z2c$6wV~mMG?8YSaEm-VPd&sI%qs zbebM{FMQmw2@SGfeME>%_a6__GO2Vio7q%3rF2B;R(61(HT=hg;Gb@zMAJB#YG4{^C0qVtX8+wu{(GaE*r2X} zqsuIKlBkFz%4kBH#~5@zI}T@>)6KTdl>zlsH73)a{!}T=dh(@xy&qPGG3EuR;I|aS zJ&b|tGR5ty**_<0<=;qgz%0X5?=&xKZ6v>WVt2gs@s!T*kO4lV1yH5j#+~$6mz*D_ zt>+;yV)R^7Rx6qDP#q>DIR6V9JNScgC%YgUIs#MI`*u7?cUWe^T!})hZ7nud_c0Tw zbPfFK337MBK_kD4;-GEVCYT63Vl#yso(gtHgvjq2av~UN)C^zFj@m@_PS{1Hhru-N z6@rmiX17Q^pr2l2?x=qp!4;I~{F5?rlze9la#zy8RjKg%1l|5x2uxmGXwKBS$Hg+G za#ySJK;X7fU2k!r=7eClJ9f43{3DXtYJ;vB_TAZ!UTin_CB78M`RqNex z{@QaPdS~-NKfj6yMQp%Pb>->Ym(_kgyUOm8Pg4vvJvSGzH}m5-(87@z5c2;<<2o$@ zC%8paH6AeVc6@tE$lDRb5~NI57M3_(KIBy1J!}0cGXW>pPj4}ycVf$h(i(s_RR@IG z^BkuSRv^?$5*0KzPs{2QQDFk^w*_e;S28%zYHFA^XAz@1rew*{91h5^-z5F#m55EuZ4de=+gd%DIOhC4!;V0r+# zfzSjF?)0pFFFTmq752Enm4c8<(;aC`jhCBs8cXtZIvw*+Z@lC`xSb5qa*q~*G^*@+ z4xS@Vu(@u@KnO*gbu?*?nmM$U?{Tdk^E5z#$5|;&0AKs<&Rf^4z|wN^4V5lJ6?{bbVP*b+4AFomkJd@-+{tDEFS? zP7oW5+Y@a$-|;ww}Obecxes!WTBBIn9{naGgV6 z&ym$3U>2h`Wik-ab8@FK0c1H{=4Q7?h6e=R>@6SYlbHG`nU0GJYP3%%oqO$wjq4{T z=svr;GLdjvSEVy22>VsoMfs98=q%g7^>o-_CG__K&h~}aptu7B08kAynAdRdUj5yx zxAug5>P!pc7>bW#NP<*)iA{=|Vfl4ik|r(GZHFNEqTmHV$Q|_Jq{hO#;|Y%xB$h&j zu_svI^CeI~Lf~EF0``1OPMxwWl^&d?z}nx@ncwHGTaDH=2hJ)|bnu0SMLto@$k z6h5@$c%D)5Dj@l14SJDv(>IFpmM70hr>LYL7X)@cT_VFt^{Ub3tL+Kt3@d#J_jGM? z?S;>+g{g{aj%o}8hoMBRjsu}QByVNU-DFGZ#6c*-v*nf{Gz}P3q4ReTc&8(-OB&&o z3N|9n%vEXf@*AzT3X%lssZx{6g`?w!$aSsL!yb*dWNd1?&5;3PF~9XNt^>BnGEWHJ z|J*Yuxn-|zfGaT>=wfo{P?5+8wyj}c6_pkEpyqf|IDYvyTp7z=Y~O6P&uOK$+%S{i z=QHDZIJ-pr61WRkOI!INS}9;+hTkzX+xXiRm|+5_7Q%7J!u}Nln0Q5hd#u;ki0im; z7{kF=YH2nj9pf`*wAu9&^m4yOpipfpx$vW9Z}r_f%JQovQu-#3p@7Z;uls7TWDn|=iBr~%As?RRz&@v|7OcCEQrxw%df2q z-j?F?`pu3YaGF!1_Sqk+=jL5pDdTpN5TuHjSl??S@aoD&U9Snm(&ZVJ0yY6q!J2ovo?BlYW-P8^uuuTz0|9(H={*=+?b_Tw}N*|I?WX? zjWgS{vi~w{EBF}dC}|c#Rv^T%sSKbT4>RG10`Js5cfdC2GNm-__jxKwVKn_TLf z-hFeP!7H|)R`4^miV9SWXW@teJCLQ9BdFA-*2Z$u#GDs8qgC_*+=v8{&ygI8f|P2^ zOnX=yW&(=sFmTZQ`Cxo9sT>x@0HF%|uycn&qk#2BX1>DtoHywSwq>Vb!nqxKL6E8t z2qbMvw`a}*a4T?p_N9A|&`Krt4RbUg)${D1I$)sPap4uXqqj5!{1*Y=uWlL1oH5&04nrSNb9O(mTvUl0{g`rg48Z0uN&NYxqv(y zQxX2$e;fH;Nq8CIe^j}D?ByD!0c0?+_*~@DPm%l_KiuyJymZzh{H+&T2d?lfcqwlS zz79;XH#~7ZiQ5wSL!;h|rjooM{`VTv$KZD7k-Vp3o6^|($8LtnhPa4}ds!}eBSh{! zCsGq4KWK51R=U6o{=1atyMAd`b^&`^SFSto>xH!KeJgGYPwUN}Mu3wn&gTv8)yzlS zH-m#$XE_bt5gu4j&Qj>kP{1i$wf7#*+lbHV$yI^pC*%%ANn7{jNm<}k4o>qam8}hj zA@(Y7`iy`ZZ$%LZAPcj3)?WHKxhT`83y9UOc^>bfPO5ksy%rR!2(D z+ZH!qYqCiaXaZaBf^%K5(Gp%!6pxqCAp)Mfgr`ZI;yQv7ITQ>yIZ8CKnw*zc9PQo4 zRMW-nNG8H19}oPZFj`Nq04~QH z+QV&kv-r}Q_yw|Z`>fKZyLTjn2}xD%diuF)Jl)|FgXuuFKAIR+xd)@&37)xa59hM~ zAUUW2)EL4KG2)yZxe*l*bp7!wG<>vssXJwlk$rRmwrGN&>p*NGq1=9>iWjjcXK~g{ zaK}3z8dw-d-Wv$1p1Q>dC60DsgqJ6wYR6gcWWvBN4#EP`_0^kyw;Vj+{7TJib!5Mi zJUV(Rl<{Vr`i%F5_cO1MtyZ<&M1j|8mkzbM^twiz(~O4K!VmREW#t0ql_b@Qi$d7% zYdXC?855DXro{*5ufh;N?tJk)|3u>f;dt7A{bH5q+79Pm+q#53{nGk{mU1|Na_Acq z?rWkXfcKU>^&LOyngM(yH~c1Cv2LgZ%x}6K9iXnatdsBT>aZ8*I7#{kQ(UA);KFbXkHJ}7XOkD z{ejpVkCQ>KkeeA)%S-HX;`g!Y4>&(1ZR88*_8Y)_zl2BFV1q6E9Ol||nBua>49RImDJ6pz{ zG;5mO(d6VQOa*@qC7c-!5)d(F4{vPsafKf}D$0HfpG27Jo%{^@vL%qmDQacD(Lkz^ zSDL8tg3)q(s=P@>S!H(WKnqHfQSsvS_lgRUnD1l`Nc=9-aE%Eol?8lE`JNL2DAZxa zpn3+q+|BqqYKbT#eeg_fk>Oyxgf6hklMG9q0;GkDqxy}7g~AT~SovOZ8zN`U&E4>= zQa%T%|LiP&ky3Kq+7HHeeyrL^WKD&L6QR}KGyv8^!>HMvcXqOSbA} zPcqieGO&!rL6B?cRs3;ss<396k`|Voy~W4T{1y*eS)$Vv!Gt7(CuoMP7D0V|>_~ye ze!us9Drd%`ls@?r*h>w+9!K&!zUBx0pgMm1;O_>LzN`Llfb};T>;WJ{`aB@jhXxenx!m$~?WDOAY zB-p;1sDQBR=6Vl;cM^k!2lW`;F=G^Opyhc_(n+M?I7yiD-p|14Ec*&vH)0gIcFuN) zVi>sNEBr9U%Wd4B%KxJ+L1)WRgb3VUq(M;1Uo<>>sy$nooxD>we`ZI+5Z`W#KMen8Ytj7YK}77* zk3i8O+&`bDlj9)9@=7{NyP_ci0d(A3N7>C3UH8U9mkO+hbr|4lCWO#>-w@upWsg8p zg-J4}Ni?uU=ya4zds%~v3loJa%Ws?PW=W3$_YWF@b9VGYaZ6Nc^9~nw>14Nad;GoD zFp2f)1)R*y9`QSH$-5Rm%l@bc%hU&FYd$k3eQXPbF4dNMCSS^DqZ5iTtyi(P=kB&k zcWCX1S@rf-@7>q8!Zp3l?aSv3mMTqptj5QKUtW?}=JRsKfHwg#TOsy@$^iDsg#gnmK0^uLnX#nvw}wj^pqI=m zJj(>n_+o3nPd#Ffpl9(Qz4~59Di;nD^1Cy5n+}uJ=+-?rTlsN9$Pxg93ZY|^#^PWe z)EALZ z##+&S#3G=!{iGQOCvBmn+`5(y^dPZ3(~Qnld9UocXugRNo2m`(*}~G|`-tWPa>2ip zw}3>-+kc6%4_W^VAox>j1tE63XmB(H4BtB^gixu`=^pqR9d!ll9bJ3-?M3^4u`Ue| zQSje<)nDGH%&?1Bvy3TK=quJT^QoCV>}?VMn`rXk^yG-X*j>ZP#yLtawC4+XLvdue zol7dHL`)ruyr8_o}4Y!q6=7z+Xu!bdOIzk!=#ZwUFlN z(2v-gX_mR#UMAYr%M7Ahkk#cto^e<(;Rpsf~r#-!f!UU>_o zwbtOoNw368sLiI9(!ylRNMK#q;xv?~C5Ds*UKPq4Oo3`9|43}m{ql?oU8sc?AxQt% zw%2J1?Z87%Y_t4jaGW0_mK1*n?){%^Yn%g~oGI*6W5dKyj);d{;Bfchdj{WVMali^ zE=KcZ7Mr?zd()m-qPZAqWIys$v8JfgL4)k~gsGu)-$udLU;J>19>C^7l#LZkQb;;C zyBQ-BimEE?+57B4+L3s86mzZAqCen_QfSslKn6iU9qC-NUbN6)J+P*LqsgS}rD{0~ z(rW*mW!Bq~uaA>90l5A{fEqrCHE=r!OsLOh(iF^5I$?OXk$#fl*D4+IRTa3_7gtHI zl2HEM0TtF=T@|}GncOK;svr-8nj<zs`d@6VXuH|T?;xv@0o&t))~LK@&~rfKy5Y6GHUv03$>vk$|G%Rb+#ZreS13F z0HeyiK*i4c*R7oL!Ju$bL-=ofKP?Rv90R66BiBy?LB>B#Khzm%t;&cB2$<#P%kHrH z=H{gs)YxuspzNmZBz>LGb0?kWFXk7>kI^&H+{{fvKQh$CF|f(Md#WG*yuwe+dAmQg zU189Uwd;S^0O7z965{1b+&k{W!YcZ7 zP^S4)!{q6O`5Q`3rk-c2gp!t z^ZSZ=>46SXqV}6sgaCAzn;HnuCBm%xjQM>_e6&Hq1aY zwQ3a+wp%2l2(2btj|}EE`qPx;`-(q~u|U*=R)IcwBtoO<>?Rx^3+T)MYsC*<3En4+ zjF$$*2X(vte7WhvqpR7;f{$16+9Q#{Wf*nGwZ$(2`@r#Xe)apORH=n=>4gGto0O5c ztc|Ucvl7)_gu6mpvQ1$Tw6)k}-U(so4W)%ONc_fJ`fx^EzGs1YuN&ZfRFD;QRG>*Y z4xKCz`Fb;EdMQ$#HWZq7PYqoiJ})h^xA65U)zL&Jk)rvqpQRRc+;MuZt;f;xzLJV` zp`2%J?>BL3TKeowi2E|MG_^jT}2K9X|Guwhd! z(#&4RKH=;UBFoiWDm`bB&&5=fPHV5nF23~=&(@>N;~SPj_CF^fL!g*YnaM9l1)Y_A z309oH>Z5Bcm2DeR778b4qD04;BU(DDk`;7qFN9R$x?DEeH^@faK?oPgWccOlb%l8( zUUV*;=$rzpHT&{2GCxrF5~s_(^+bb`g6=cjNfp(o9r^FLt%OCv1^&|(jmD<0`V*TE zjyf+zY$i~jlSvEoM|Jc2Dzt_72M6fXGEK7%l0TQ+YSd5jv(hi@rtP?*e~B!QfBHkH zGr19HRKbNGs{9l)DNYX&h#SgwQxrxw+@MebbVjEv9Xl=m2*^fFz z3hjJe1y*8{bnEA)YrFI2nU3f9FsZ1Xv6u1ZB*pkxscgkUN0o)@i+#@VZAxfAk2g3$ zUe@~ep(w7aMs(XnJuzJa4npohiXoH_YAlYq@VsxENZwS&@2kv;$L^TgZ zOx*lRfXf-n&UDZ1u_|GWkgwp*Q-_;t2Gu&BWKvOUnc9++*}kyg`|3AB@1P?&F%(fi zSK`{)YZs54;T`6zbKBzhI9tWNI+NJ*g&9?=(V1uP#>*_+c=e|E$`<^Vd7M~NkFTJQ zRuvf>#&xE%P_&q;!!CYODZv*^hPCR%lKYAWTqYI+j#3{)v6U5zk;(u;FfUbd{ZJP-jvMx^YF;{N zO;ca{LJ3z^ep$iiykn0K0IlrY4!H=;(}%!UCWcFr{msx;YW#2))^BT`lg~HzAl^BC zo0?~$1*1^@1mb?@cPi>?$pMm_XWvjgiS9UaF^tt2>BG%ehy`@^E|OFhm$?PG9{L3K zMo1;Rfp1;)PD1xNg=oT}?G^1J^rl4&TwdYte0EMSOCHg+a@fKtXA{8AjA_yDJz+O0 z<|5bb2xd>y(94&#O{1zDPlje+>F=K|`}3dK_M&IhKMhKxT%E|LgD#5t_6;ZxS@=$8 zrU9D>5(u0&Met+J3?f$9=Gnx>Fm%MRT#)}4O*F_c8@f%F#Wf9UF6<#J9!)jJS#=u& z0!6uO=Jr;aUl{SI++3eaIk?kj4V z$_27a>nPe;)7%6!Lag=#a+61Pm(gKu?k1XI}%iwIGe&HF#00w#_|5meWH87*In`FrUH!H2~M$6T=EP! zL;!w3@IgeW;)Rv0TH!~(+H5T=3AOHUwY~UGi2J29>tpBlD$lSc;Vhi9{Nf#1SEKElH_9y+r!qz!3a#&c^U5gfD#t!p)y4p# zDYP|lZ7$-acsA!-uAxNg9ZvU5m5)=t?R~ncK4@GGu*$+EP>u+Jv56rbL{vP-D67kB z3eqDRjT(2(Bm*VVdA+o8M!k1P)qSEKRAysjNf;B}LlH-cCji+zNqVQ&KvHr`5MRE% z*q`a6H9vln-jLIhB?lcpR`MUHQ~mj~IkU4fckZi_qR@=8VxY(cX+!|7LY4|w^|W=U zBv#VyNqWCt^uu6Tn{+a_oa5*|i_ciaOU$2bUVQOU{`{<%mu4b^ZJtxsyL7Y}IWzXs zC!9;IPJ_8O*#hAL*B)b2&-3lkMg(i2l&eQTIhtlO@_HRPTq;6^rdl5z1+|b;koN&r z6M!AW?VA)<>4OknH86}>iSJo2h?U)ZL!u3CTLeedFCB;1vtN3ui5MS-NdUu=PJ^9Z zpv^TGBj2oU&b-=~qn*b%?48YEjrC~7d3~B2^D>Mv&gFPxzQWmm^j+iGXLr}W{;1HOmk9E0G z|KRLVn=h9bNj<@>epx8rD(hg4NaU7mb&~N~dt(A-Pccj5(mK~+8C3Eu+zu_~eZE43HH6^R4K>*4A5mFy;b^^?{%X$k53>0-@SiR50wHJT%a2O-0ku z@SeA{Ez*}xmj394ee&b9;A%QOwP>WF;Q@m$R3(}{>G>1LXhX)B_D6JyRo)`0iS~$H zr(9oI_>i(fbteSR3}S7Z>+^uzWON)Fh?4j6DOX1s__PUuKx=<{M45j&d^~8l*x}8`eji zVrvn+tMxJewI=%6e(sjh;pueBaI9C)YfipLYE(Hc3Qsv?za$Xu5gAKcQJ$) zS`S$D^=mP?`}U!%Z8;BwjdBd73tZa3Seg4UOMYG|HmzifHMS>^uNQ-}VHSSX5x1TU z5U{B&ojjJx&3sW|p@z?+vNsEm_5@3F2c8?wjz5=nw8)S~B-hL4(qd_*r4z zI7AEF^}ugW&r;lw)vwZb6_p#@1RgFzb{+xNt0xB&jJHv0_ zL2`oYG37$Jc&_gJ1q}$E<~Ue)JhKo_-KL)>TMAz1hV{X1qvg=$V-<-|n6Cik)Ef}d zBxVUCHye!YBujIMh5bkHFV<-P>fN*t$SN_Ge%;s;8ly} z+4c!t{fsc{Ji@%OOcT_|Wf>Gvs<{A{|_N|L5bJ#H=0qcNF8rnp_^dnO4fwElRnI1e8O zEssU}Xcp2zu?Q4+BeX)FfFOlAL?ef6L1tr({uiX8Ko-QslnoB-A${PDUBXG_^OZU| z2&P1#Wc&pGe(0raxQIuLZ2It0YU!eum-fIfJtHskC%yeuK4J%rnXM;|JuTBmyyL{Z z2Td+t@B%hkd+;PCG#T6(a2&EV9~a|c70zKPovS4xrp3zP_-?$ zt{1~d1?6nhfWo4GC3Ff02SM7gtslkd=;dDNZHRm-p-b0Fjyj7+(}gZrs}H_QGoVqc z-O>!lP=}{H!ZFax_LbflZbsN@2P)mUq4I7%792j(%#JXWuv-t|GYU@1K~b8<-VJk$L{P@+yF6=l^-- z0n)OI8`Rm9MEcCddEG91t_jlrSr3dEBRdm~0;@nqe1tRFtx+I(jAg#$MkM}ZG(T_L zPEopW!v^bvr}8fez0D&c`Fmrmeg& z6xoToEH{YeRe32_m^@a2XNg8d`FjmRAz@I?6(b8w;%6qwfg`oiWy5an>|Tv-QSQcW z$8#(c^AW2Am(0MY;kjh%6u{xa8y{JgBHpJ^S?nObt1*O~jih5SzesCvGw`{R8;g>E zsQzOW*%x2xh=}X*aEKqEPBtRHUKd%&&oySm5GxY#!CoUiDQTGDdm6`xrwfszOIONG zY(G>O0bB;~cG0K%B!Byq7}Gy}3UKx?7vBa|GX&}3VrJUf)V|~6aK}EGI?p*iKI!ew zq)NCPmTkqgpgaDd_fn-=>};RR#=! zj@vDo@MdIveGMh{r`#X_8gRJw3w;SD^$4F@?p93lUj4Af zX?|E67Y2f(<}zSAk$)y2nO(FaKxbR7(Cq@pA1g_hDkiY9Fu7Q2BnSP~MCfy##Yihe z6*L!JTR)j8_@uNGrbHEou;Ee@Y%SCtLh>G-Qxvtzkzt59J1)hg87TYEL{x~KrhWGf zGd)>FZLoYZ^ZXi^xg306ouQ)$&@sL9)@lgB)yq$|)6D^JGpVD!*1772=e^q}W22HB zo&K;hRY!ETUtSBA|%W@3=*56y`JW(9hn?wa_-{}AtQeb0*jpZ|ZP&hsU=P7Nl|eNCTFETA+jtMH#4%{%RdEd3hrM)EClc_*ifgrnuPU&l1xwfyJSYMS)@(S67h=O`V$?ZC4YghN@-25w|R#3 zC42@K>Sc3nb%&n`+VHs9gN->8e}6xT1D0*&`Tu%WfFTL*>5l~oeA-J@m`|^g!OQ~* zx7lJ2x}s}9ETH{{RBI}{+@++b4D z8&rgMqfX~-BV)li1W7Meqk5cfr9?es*~klZ+L z77b3LNp5b0Jy(fNJ@Oe`xV>)*6;UZVGF<3SzG$zA1eKK3%hD~@D+p8-#eTY^I$K!1 zfe$UMWjYYBnYzD3YmC#;Chz)jpSzNj;$PK_wICIQ`d3xG05GmURWs>fl%~Hxq@@iu z>b(n&QT9$(>PzbPlW9X#O@uS51&Q>)jRtN`MJvc-X30@QLo6@9rlcIPxIM3nQxWwv zQzY26k9*l@0Dqq8$T#TW^(P6XMiYK9m zw)1_8_rENt*_*<->5n}Uj*doI<;s#=4>wE@QpRHKeR*KwQmwE_;F#PLM6*9U+K#d; zG(z*GKSCxK~eBOQQ zP&_iyq-g8@j%>*P`eGo}%R@1lMz%A5t%xh!YuVBSbE^2+9^&c7D)2*t&6EXy5LHav z`2W1-kk?KoCf^LZg>vRzTm*2l+PO{V_`G@3PRsG-yM2FN9Z_LxXkc?fmCeJ{pKXkilA=oCW^0B(_ z_adhMoKXaB*y#dFc2Bab?irICc%Fm3OyDtd~%{QN36!2 z`R%cdl2l3IY3YlO(;II9*^H|5S+Y)Laf&-D-_O_I@0-oZg>XPWV+Z;1Ok^qH z2t=#?&>|LQAMjN>t>RL}FnN{JQ1I9+J+(dMEna~^*(oNNU&l@>>HVHE7`%&`EpBq< zea-#-{1cC?cp0adNz*j*mMd&%qpAW4_kKJC%2l$Rr(YI%@P6}_wXi@Rv&X)+Nuj~j zgxOOF`XzM9-@rQJ{_R+fzfUePQPdzcy|-#`$GkNsz^$&G?Vr4o5T*Jl^j~Q89-IH$ ze?`Zj{kDfby9+}vwfhO50Mw;z@Yge1O3Ik!WiyIeA_81-hUsR{UF%cibOr2%vQ!Ss z1t#j8c_?PfncwnS_7wx+y{vkjW zzG=a;vi)s+%SBN!xT0_IVbih!0P13OwYO~b&wGio0w8o%Sn@YtfqII)uTh*B=Kauv zqwQQ4rn61S9nL#^@LA#MvXV5w)+R8UAX7ODS2k73d-a<~NvbJI>tqO>x_N)mRIGClw?^Okw^Ax}!Kt-v&nT)y^p^qGU-PE- z4)-rfZFhb_(l(Z$;GX6EBMkj|cy zP|mfehWd6pY}u)>xmgwISPEm=CCK=+74U-A$63L>$Gl=ESNMKGgr|6fP6M~!(|;%@ z>)(tzL^Yr-^6%f0jg0|;xE=rmfj{$tdas@U6L)lScW0rxXm8#ioa1aiL|X9aQ`pBu z5R{vn8xv(-#^67Hg@llu^V(_#2WP?1UK(nHlEdi?}IWp5p3lS#%Vbc^!JN#wK%urGJ+naO*HLzHfB_SqJ3xmFv_lI}^ z(00KezA+GY7EEEd`1PUNNWLcSF48{f&+=p6zhmU4j{c#wrXCCKcf}HMo6Gc-Z}u>g zU>|5!FYuTh#y5^^Z`6pJ^4B92{h<%?xCP$)2lH&L$W{NB$NTla>FVrE_yqw^K-J*H z=Y2r2I~@yERiqNxvYwQb)JWbJ4ldvKO4A=7QqvgPIA84BLc~d*!?C|$%zd=Cb<0C) zc%B@vl=SM)_&aZ1lg9B+vsK?`gvyrGGG)nB8J)caTR?f}@qe%1Z#agx`W}$)@~~Ll z?7Vp}X&a;CZ1--BV@`{9k5MwUG{c1!^6WhLAQQJwYE`M71?ReFX|r!Gx9c)uw#}&1|S9a1*Kc`%yHK&FQXlB8k!|q z@`JULF1sTX@7XxEjjy!!ToRPd-V$X(#V9Eu2(U~O_hFzr+kCt@o-G03yKQ!wEykhU&jWYnp@Ut<}r(l>P4u zMITEwtxwuEr<0dv_1KiM{MfG_;(ZZI>a32VAKGs>S{OE7$HEx`)NJkQp4#Tc=7|EJ zVkB=FsQ)dr9zBck{~xHoxFV9$5JOPK@5*7%g`@W)4^dh=L@O^Z?_%YEtVNrv5&G+u zze6vYG~@@~jOtmi=v82n%}3Pe(Ec4CjSBfgxZZ7G+r(U6FXyY1ZvoMo>O<}2+^ApO z3sWOz&3&`EKiJ1Xdn%_RN;4Zbrw7*W^^V__Or+#)ItJL0zM%vJcgK! znPlqdoIghRHzF%@m39n2dZ<9!(De1Ee$aq_Vu4pL>1&vxXLfAto5q#BpUD*k`>eSI z+UUN*&4IwQjcm;aghOxbllVJIt(#K0HD=`=*=D#{L1UqQGCAxnql0#2{6!Q~iyIw9?$UHys z;dfESdz(7OGAb&HXN@Q?AGXKGUdSV;a&&WZ`x^A>Rov?{a5P6`*dzUq|Gbm=-xms( zM`zg_D?5i0@X6RevQ0}%rE0q+=Lt_GIhnTk;nXuigGJzd?FFZ>qjZZd&7EKKGcwZQ=KG>qyhu3)5gYXF=r@*^gpmm!|M{)qdWUqp z^COshX&){OE9?K4QhLb#WDLgVPeo4~*WUGpYD^`*xlSoxg!`qj0gum4id4-TXL&#= z`L|afR8?^EA@Sh7KWc652@&mlZZHSzJq4Y3ia@jE@(7 z*uVBm|GN2FU3wp+etLR3aCB59!B%d3F85|Yx|1X8tNF}4<3o#x+Y$i6@TyYIjz=e2 z3XxswcZzJfGip<}a*9KFxkwtNKQAwcuWcyPY~_}b$+Yw2k80|8;1zvYXXlaIf1AO* zSHRU67#%Ocjhl^(eP|xLp5%9k0Y(WN%ltX3FTD*Ki5~9iKIVM9H*WKgni{#$&%9t= z&tveT&FWkyK(hPvmgI%+=~;9?HN9CJFwwtkr1SP5NH*T4$^{p`%csi8W zJ_Ui^JB2z55blJn_#sE|AQv#GAZ>vAvDt?*2!J2}!u3$%$i{2Haz zUwUZzZ6s$LedXG!0<&#Jhyuzd0D{MkbM2Z0+E6q7=N-4YSGP_WGG#v&jQ)?%`T_xU z+z!5>${b6%Qy~2Y)LVKbPI1*qk4Ly>!<{u7c%yk9J4W5%Lp#L()I@%$HFM3`_|oSs zjC-k)Vl*$ZUPM^FSovK`vc0y$xP~7Z)UIFv=Rtpa9 zz(Qop-+J!;9tJ}5jTX|MOmO4b?;A;Ah%1vNB5O@CLj9GBa#yeYt?T{XAsuP6-}YRj z*PiT2UwC$o-SZ0wckJ^yvsC;rOO>LXZxHEiLJU&fam3;7?h{JEmhN)0bB`ObKe?#Z zU+fL4W9xt77ya_w5c&#=X=7P>_Ig2|lieW4wD~-U6Zp{wY^mP?5oMo10oM zD=Tx1W5E~Vc3%o)o7+n7_$z@LngG}2 z!Zd@&eUh$3{BS7tpr=Aeqs(^>`cgVx!m{+(Ofp(S2*+84<@0?i7n1m$n-BQ_D3t6J z^;!mHHm`|cqsr~&6#pymcT`usJG@`w_DV#k0x++NsX#zi@xmvAn6wIp1>$M|VtQr1 zZoT$H%t{vbt3h2$APHLFsKq|lwScResuC{1D?7U)thD}aYD7iF2oUP?vdf%h zPd9{P{@9_}!;WaN@|kgyk4&yK&+yfYps2#0Bl#-fPeycn3tijPOFq|$bi8&uV`sd~EsIf%Pn&1U- z2V*ks8Q=Rj(n=G*nNJaRMdo(qnJvFr=vTIxWbNjWIz*t!YYUe|b~Am-KY343Yfl!e z13lW(RbOaN8@8DRou0VHFKbZwrL*JTET79T-&*Z|UVEPQ3j+hvCWz|ZWB&_HZFm;? z%{b#^GuV9%3{EKxw5TbV$mFcRdpj;F^ywvyIK3Up1x0XNXj(Q{BEF-_s!JPV%n!Q$6vt@gnEJ zapIxd>OFIR|CZ|*aFnWNSfx3Xms4E$`IU9A{ z_|KDdk`df1mU&K>UKw8!vnXV=pfY4pm2z3E|3VyTk&iCBpJg;#aGte7+`4bIcSjUTWUa zJcrOwyou0Bb!!c3lqwDaZI@F}pwphctViZ2^Ngz4_;n*RXs0$fLpc_=P*n+ZkFHP3 z_rFjb_rgGl9{1j=4iugvzwuG2x`x(b3U3hCVO6QI%Eg z-e{$at~w))rum3Tn&)N5lcPre!e?LR!| zEDqS(KZ#|J=!)Q2(g-oF@PAxAI&tPX?Cy5$XH?WVo&VI`YqX%$_}7l4L4x?f{?e_x zS=0h&Mj$ebeBk-aM|)XJkbjWaa5Ev`ZxeO!Mp2nxyS`v2=4O%5oVU1NZe131j~veW zwi|dzQOY(PQVsK%D7 z&7X6-{Kcg>8`cwy3~Y4z%?92)&i}+K=ZDuDThRdm`=Yr2Ctj_zo(Q1co$o5Jz@Gu2 zw$FGJs>DV33P5k`5-r_y!%;uSyEZbETh1oMM&m1vWwUO?6J4B*q-Wx-%rT?!&Wb^G zOc@3DFWwI45iooRh4ZN&`Q~XBLJ~^%Zu!h|9)lB=npb$RLg-Yg{DS7#*-p(lwOkP5 zK|$$q2-Gn@*nPioCyTl~ejo82eeh19U(2ZXfh5EPv+=`B!H&nXKb!(~mf;dpCNhkr zY>flW367hbSj-;su5gaWaBT%CbaIC31CFY`XK-QC&S=c8fI|S$aDyzswkV zIKWpTf~?_Z&avU%&f2e~{ob%r75f!$PNK4;PuP0InR3iIUxI?Ky0a2aO#Ad7G^*JN z0HA6~Mh#BMnOkV#saT^x;br1Y{^|X(SYW{{^V!wuuKUSfji$ishRIzpjVMDnG3_^v zG~wZ$iQ|SJo|yFoaKw}q$($rv^XS`3H1Z3B*vt+z%Vm5(dQ_?vsxLCVE18@7u|??K^SWuf9N;aPx0&idhj-1hmFV)(56 z^{I+g^uH(eqn~9}2WoE4OD*C8Q@f9~J zw>e(EJ4#)AJEP&fJ16MPhfgZo?t(k~wp&O@sK)=qajv73>{H`2+YkE48t*4h;@l5Z zTTI%P*u49CGrY$`oxivKIKVM;=zdbuCkz*l)8&qzy9~7PSq2b zH&mw@*-zyUYV)DHRKtH~{F_d2yCrzPar2}${wz_q0W$&QYS=rgz3Z{>ZujoQvChPHmQ0+|GMdQ_*BbEK@V>JAK*ci#bt>hWomJ)6U9*0|VuUB>i#N?STj98&RD1 ztXk!p9hkau)BNwH{o}y9wVntRYKRxJp+vY8z3ymbl_z<3vF?6x^k%NwN=e!f-`QFP zgZ5H+`yHl@BwAW*e3aW90ALiAt!L@*F7?I^)&gl6A~OvTuO+Nn%`Mbz<6~{2nc2oT zvoIZ6p?RKoO12PyuV7#wLx59e!A{7CIkrFL!D>N!tFDFnI{)X`zM9_9Z{A;Dfd4^y zAOrvlU?|GU#!BBGjKZ&J>@ z@egrd*5zWo*N$=IR-0m1?jaGXt&D4OVWBF%T9fZ!SMt8!qpAG!OgZ}ziO)W&Bk?f( z{#H&fBS{!j`%v)l&VJhQJwY3l%ZmHaH#72G3QaegR2Q8UGON#Fp6j4fWgBaeR+d+@ z)D=rZ&!mYi*=6p6a%-MhGwNgx)-#zT$lxjn&bT7+NNrJ_wROV?@ zFuL~B&7JVZCYE}S2~7rZEf8~azaM*2_qK~gpZhGusT86A1MJ?d0#!~ zF-wl+RTRbeVsKrVi92Z`I~|wY1QNPbT7twsvZ3cTl3C8}pGap<*EA?N?mC)MPcdgS z*EJbtb1{q`9UakZ1CzVdRpG2~p&78y`yJ}QgGtY}^2(q2UG#YF`C3Of5X)Wx!5v=M z!1}!P*F8T!zeV7Yf+00KMNv`p>zKs+nl`TOls3hqliHoH zIY=fZlo!kAkA+FVB{eu`X)~RISvyQJX7Z>%xwT{n(6E4deBzq1`N`R4s%I9Ij*;v; z6UoJj)j+@Xc2RCunaoZFAGTwodrv~u_T7(vo7$NA1AI)rScyxslyg5s-7sGP3o2|6 zmbRnqFmWNv51ln=mDY}|r5*{zM)8DzU4oVFqF;5yoO_*2TL2I!mL2qd^0F%a$i8vC zt;ppU7+MxhmS(Rw|6VruipLx?UiK?(-1)Nq=uXp^#ZHG-Ao3N47=@Cs_Oj*cvXxrD zwlj%ojJbr6u#;yyTJB9YJ0m9~FO0@rRSmExr?2btYUK+(vx-~&bqbye!PIa{8|s7y z{(Mfc zakV0YzDE80ORu)z^?$-h#c4b5efQ@q4i4bP(_;Shk;oVHFIH(FfNV=Qw=tk1?m>(u zAfT%G=Tr!5*9aet7(f?b`2rhATqenf{seqdRuLjr#;hfcE^f3Pe(K=lh50Sx$V@ zDQ{qqgoOaCJ*ablK;`8_v8-bx@T>sJp)~64ceCTe+yZE6^z0UEvGXFDR27xWLyGC3u$!^IWDJ?yE#zr8PH%_4ESe5z3S zvZ0|C`;V~F8T~JSRjBJf1a>=G)gUM<-I@6Hg>iVgWUvxiMGs5#^u%^ELO64*eOI8S zsa)mE#fW!zM!1>fzKJ64>G!Rq^@YBHAF*NRfz`TR|iTY8CCd_{8Qe~ZrQ z2jmxoMd$LJcn2Vu!4QjCLk-i>MOISzc=#I%Ad&0Q{@~yn*S`hja)Kmqgq-s2ov_Ae z-K^@YYvtrMdP{6_huOY7iP!X-ftha)_igxB{Bq13E#CvNr99+Lk@X2H4g&T*agOu1 z?LQK}B@n?;B*S@|w-_BUG&d*fcpX_1IJJ}~yf9Xb^g^uIXnVgy8gT%5k!|fk&TW!; z7Sm5n9jP(|KkTUhsP`*)Chtb(b}?+Mcu-E__3L7_0M0R2Y{&fF+X0aXB3Iq=hLRFR z{l;|TPkzg&yOG-amOaC3*!&p=H0TR5GXE!tz*t_q`2D=&`j#03=uJS-2OKWwl%s&E z&Q?bN0eiG1(RNn|P$@FDU#uF8(55q{c~?a+N`ZVzZ|gQM1i&#<+TY(N{s1R;21veL z{66kGVrbM+!7O^s2EE7l_DTEVo)hcokTaCLp1Bmx{v?Q{m^~`HvMmvQEJw6$1NnTK z0J8_^y<62$kU{~AF+^*zEN=f2o<|w24=e554=m4519FgqPa$uc9Y3X!gTW)7PRB4NhH78HcK~5re*=`pG{wgJl1{it!4n@U`l}+(p6Fwh z9fQ~Jvdul+rzKARY8ez%?NkZ%eqo<>H#{$!Z*3pnDK04fV(Y>1i1Vy>cm(LXyC+ni z)~DczZc)v}Dg9OLozPy%-0ArEGe&k&YAlN#NPQAP`b+C9I1mjrGF?b)sK)qx$Ta%@ zX&`t&!fuLh@%^TeSEhpJL_DupS*{|41G5#g7JL)1hxZJa{Ufa(0=V7=A6cj&fXbA! z@OSoG0OG%o<=+BuE-6C+luk1ho))sGYyO?i@+}Q-Y+z_e@!OXzy z!HL2`GVJjJQost)qY&N=%j^fe)j;|F(4bCEyg(LkyGmndGx{L|@g*b_^;4g|>Ge}s z4%Rk^`=Yz*W3Ugia8R+lv?pLWM!-`gya5NfAmkYvek^Z3P(%!&==Fb<=f^TV(O%bK2in{iBv(j9_^^RN_D-A(;Cq|5%m1R#%*Dm1)&D#pfioqUMPHPZ za=tiIpL%-8(fNd|=nXER@d4KotiZW1$R^w@sVedw@{aj@yPws&S)gjZELa zRNilwnx}ViBF2r`S2kW!i3Hjlcn9PdHw}jtnMA*K@sCOM!>}tY2Ei)JZOfH$4hJ1Z zlp)wfk89T0E(Q_!Zfq{q>1kL0(j2?qVm8LsgkGOq8bvN>!s|F#_V>^NMbE3=8h z72PjEzY-aVe~^;Epy`8cXd~ksScFm0R|wqN(HmHii}C+E3qZI@8m-#ZW3Hpm(s9}| zIp}w_#SMVs&=+?8YuJr}@UtB8E?=ZBQb}g1|D}VM27sVd;lNN-#PJrE80Dvie40+6 zMF>@t?NPGqM@EO$kntbc*gqMx>8d(AIkuBg`vkx4fRBExvnN5wAQ8uy4NtjRG`$j= zK)oXrh={RMrdG5BOet68myY!%`JMAEc6yFE8cg+u;Yk_Yrn0;yE{ef>g2+0jY=7lL z0N?M2*&*bTSTCGstGWhKQ9-MzUaMhOM_9AnLe4SZ{;8>{vVbX74di1P4ntLCTttUcpdAK<*GQvwfEF7U@uL8EGRPQ`(7#m2tZT=fwFWTvn1) z|M1ks^bRU1(>&r=|MV}wkfy(yh(XN?G#ZBEZNA&4W+B7X8Vtw-6!e}ZPe~Hv~gL7490|->%@Romh1Qw3c0)j<1lQi+wV+ zx<{TzXNG@OphcTbj5R9DvxO1k?WBwe&i@=^pnT}fcip!&4RagJSaRK6UN?Jk*AuSl*R#S2RffZTHU2{-xg( zafU9yz!m8qrawp}vUX9ox6q(@c4){u9CfNG^7{E6TZqH&9km597l6HV(2SB0j>S$6f5g^NeLacT*b@$7^jO zFuHo>2ub)HH5aq~Bn~y+dQ=e>rC!hdS|0G5VL0~x6!g*DP zV(6iA;BU^;<+3S%y6UDa^+N-mNe!mSOR3h-4B4AEw7=f0i^Ycu!xjYtKS^H@h zuM(A0Xg~%2)Mvc)f$g4?SEiHFUC8`gwf6cUU&)xQnUm3V8+vrJAso>s90cX@CK`%r z*ICje^;cmNdVVxp)S%JvUe9WKZd;7gboBRtzOxhcM(5XuHO}i#aY~sa?q&NrNRh{j zdOOy2PK}3ybDXs9MM?&mWB6_gk9}lL%5zL)ISKu=_Ex62w2JAnumf+ZqhvT_PG@EK z4s?4lCYJCow$P+S|rhk9K_REFq>nE+Y9 z=r5m8yDLyJ=(`mzC7+T>uL*T=s5(IZ!5ChD2zf6*p1X+Olz?&KUyhOALh#t+fbHXP zI8`p?F|Lo1sr{c}Y;gY^gaZZqOz8yY_o?y9ap>y(b+MM3QZe=(Z^Mc|SL+BM<@3b{ z5WwL9XJ;V8tLtB1&Kbcld$hB*`}5&syP?OHk})ndHj&%(A$~Odq|(_L!`cW489!{j zVBM!N_~nTC;kVt4Kz@RlTs3Um)t@ax7q&monZFvnusH~*FhqwuZ$?Y=Q#~j+Kc``^ zLH|W2Uf}0WvfMiEB5ZcAZ08z1Gs+wg9%lAucgBbh-z+fp;Y&6D`;Y+JnbG_~0tZ=; zZ&D|?;wI&0x@l#=zva^Y1wJHUMaw@aWki1Y7ZJTrp6<5lw)nOc&9TMl#nRVZo#UE2 zp=O%8sEX;R6D#plbCGGWD^|hd3Er8{PExmP*44&qD(Z7R7;R!14xirozFrbRp%CdGliekAz#iCw5=S@r38)Z%Gg=1=2s9L zb+PTc|Iuu8QF}q9q01@Ub<|S8$+!&;Z9ih$C=K^k$wQv1jP;q%U!QN*w%m=pCjmpI z->>1P{AXb*;0z(ms2sX?ulX+?^ULGqz4wqbx#$OPcM9#m$>6GRK3hunBqQl9{gsKm zr+JT30hB*Yo@5V9xctN0Z^esPk#rNqgoyhPq*Q-8PT`E^01$PGMw<9Aj#0cn?MW;N zW!uH`1?Zu%7~Nlb6nv9{U6|LWGKkUPm?qqN5LrG!D~gvOUk=O*A*|p zf!dlQYlc+cZk?509J*g=U(SwxfEII!t*~kHYyz}ppm03Za7>&NmzsY=NwT1kzNW7 zkT&)-qXO6j+o~{Wnxv+h?%`h;6@R+6p#qUXu0k8y>W#ddO*qqj@0^q$e%?F?jKS^} zO{lt$E4q7mR&ftQl*N!gi}A)cK?bSaZDeGP{Cg2gAY_3ewA#U5LNA*+?6#G+B?>uv zPFVm9s6-O`2&|=FNVyp_r`k|FtKZY`5Y7@Lb>LvEmJuschD@-TS>P%3ayu{Dy z=G~;C@6B=F7N!eM%LnDDxjmlLBZ(N3DGwfd5%lTy(N@N;PpWxP$(Ha)5PkNa6M3rn zI)3ukVosAj<{Z)bqn3FTu6?NSAtS^Hc*CgOe3Z8~{lpT4_};_Q03IArYqa(Qf&DtE z=@vJY*>gOXesae)6#Dt1bD}01DhA3~ANl_x84Ng>TQe%x<3&hdx<9Vs7Tk-1MI8mt z0yZ<_Bpb2=6Y)gbW1iyYpl=qk+c0@TS9hloqW)t z;c-Cc0G@Wwl4UOzU*`aZ3Hl2lB_ik~OmxArKN@m(>3z&MV;kVEv1;|FYUDxf=902N z_mk!^>|{PIYUCSOiW1qi?7N5iCK#dc#U*AY?BQ<%pTcx*r3cz!MU=o$%DZy2pb9hK5u=^udY8@5J45Pcuc^nd(`82}Pp~`LAyhu0tPrx0D^~OM z5OLf@pzb)E{dM>>i*TkG8)Zj(q_T(+gjyOMcrFqg=;sLo|IEAh$0hNn=BSgzu~1=%p|X;)#y#8I)-Z5#1tx2K+&dbc8hLnw7`;ips4yRY8{%!1Ft_XLrD&KO+=AvQls0?n?udMo(ZbPNfTKy%_ zhyA{$2U;R!n_ZJuEN8a(lDkr)B;n>kwv}bNt{ieC=wox~t5m_62AV>5YdHE)XL`tj2f zmF`L0RDcDgs%^GK8Q_;2K>1Oz8ZDf(H823cyaFK`tWCA^kOac3^ymV{*{P zdX6!?qRvG}h)>}t%l`IpFv=vu#dQw&*~6)<$GlCuNf7kF011oR3YTWC$rnS_a4 zBUT>6ikF+Cv=MCKy9^=kRae30{wmEMD@o(SO$0NsF_@1lNb=Izls(zyB~=kc&tM*} z(b15xF=Dn7iCXdj~`4h$5Xw15RlD+~CIZ9u$hJe%+g!AM|nGE6m!OS0{r1>TI6(`3NmXnGYILr%x8v51~F1(OB69 zo($;*l|lY@a%$gsgtNa?tk@qW&1eOc zIc@ItUHL~JkmT2ZA~c7;2P%LM7q~EfE+;%6wzAG3u236);fft3scTk80bFwEW{2?E znY6j$lYw$o`taV(NiC^ouZ`3GUU1b(FPlsjHlLSz+~iB(W5%Kld_e!#n8N_Pl+n7) z#mp-2uH+o+Rv79)Xy#j3`?&S1zX~NZX3q7vRHa zXRnRR%e0cmq1lI|))*4G8mp^jeuzb}Oetb-)%t`IJdaasnIZ|QTeTWS3QaUrJ1^fE zZoH<}PfT7dQhGk_9wg$j=w^}39jiuk;mm$+a$dh@;pS1%SfxMcpz z5WHs5YD@qU%U_3M*oPOjsuLFfH~g)y-lblw#QEyc><5#!Ip!js#V@U;u%+8?e9eI& zA5Jz^09g7-SX~jw>b5lNmXX%-%BFK!ci*3fm8SOH7eQZN(k)$xY{>ZRfnWU_zTT^> z-e0a7Y;Ft(8&sWnOuOmAIXZ#?+UnU%KTnxp?2~c*@<~jrF`CliRLz+?p*#Y^^vKov zm%jGzhJhM+yNLPljp7o)DP1=C=tt4V*(jM`-@v>SJSH(SX8yyzdOBd~vE59cHGJRW z9*U)J)A>(|tsv1D4d|aV}k2jdfv`c@8!K#_5yvggaL+)6|@2>}{W7R3@R4>?5VD z#aID~Lw3W0vxT-dzx22`8y9Ee%gkT#hSSgSXp}ml>CBo;jX1WJ0_9N{BHg`F%fO{% z;NfhmUM5O6b7pSr1I^+RE4b8vmJ?>kUMC$?yw8bhq`SB4)ttxJqW(X5?>+Wc^V^jT z-qE}6+($2YLO8HVMIbMh1F{E+NZ=fjfdHxrD(d`ebQurdMmOKAZS7C(b(>VPsoo-k?Or@L*Xkzq|Gn=oS{#WAjq%{lJ6lLCM-9>6##3oZ?iQ zVFg9{Y=Y&t!F;`waE`>gZh@~gb*}5bd>$#Y(PtBl4c~YEA>cyw`D8~&(q7vDKaMAE z{t?2n|7b@R64cGP{tMWF-8=@wke6?S5(;?bp4{Wo7#yG3M9-_23vw8Gdc;BiG<+Wt6OeF{U=3hqt|^ zgRkX637jd5ERRa@g5=F+@8}EvEdg-iIwfX0cveJDoN%0E3?8y!wAb^8fwROx2EKUn z{*R>aGQwd&-s1&hE93Py?qh<1&oyO&G^98Q?Lp4lxy&RtH$F>Y6*(%Vu$NceLN^O< zJO`L#wB|gvM)1|{Go<{zG>NwpMkP-oThSA0n$ZQs?%9&))ud$D!~+r6Ws|Wyg6zS0 zFEVuW@vr)0HTnnrxGeyq$;6)YCuQja5mza=7cZ8bf0YOJhBWlERTdWZ9^2MAjLw8Q zFLKGm&0akWbdp&Hn=c<(1-#3Q){cabxzkU7HapB{VN7z}ZLbu%jJ1*O<;)ZIX;3V* z9x7z@bzjjhsVq(AQmmh_Zy8mCXhCz8%|6U2j8-4Gi^B|WJ&VdW+{%CV> zbOxmC%I0GxJOUj7VtR?#iFjVOV;IedmBGSaHUxTaw6Qj{>BOGIh~e@;Fe-^v7p2Y9 zNIWEfv^t*NYD5W$61}kaVK|8k#PVXy{$fNvet;lT~V&ik^QCNdml(UtnL8Z zgpIiQJzHB*>aDGr-c)!9T2d<)zM9=k!rzhwojiLXUb5<5FX`=;R`z-+n;s|-%sZS{ z`!hn0y0>KZwuLu9URX#for5zU2cJ|Olp&wAk=1s$`My4@^dn}NM2cDH@~>wMtRgwD z`Q|=(z#L(8Fi`*%j0ffdqidZR_x%-@@mVIHdbnrfqZ!VH|ZZM$t4Hdb)| zYp$^d4NbZG57JkxbcbC2!_Nu^IZl^#O+%#PTmuWT8eNv$m-T%1C$&&1H+8}HL5(-` zoiTHi>g|^ANnjdD`NTDrzR2~ymG~!Y%julaU;9xZT2iO+yUc+h=EyN&)L*x_D%c_A zS#u4)I^0r@exG40)e$GsvIE+liR37}lux6n`;14m|87iJTtHkBn2~X|%@$}~w!@3( zTi5r-Pt?lQOFy-6i)G=W8*iV#>uadTM2vvV11~TGUT=5em95EsdzzBu5Jk@eJw~R3 z(DCYbzK3P+q5FYM(=539IR`uL@La#u-Ue2&?;d%WiaU-Vuwe|x&b5PUMdOmOuU4QW z5Gz~!bU@=)5ro-8A9vA9mhb4C$cCXN6yh9jbubqzv*dsl&JjHIPs7o_U#BF#`l$L& z??^bb=yRfAT2j*w9({i8%?M2WDF4I_+iECMnMjq03;vjha|AygQR(S0AGKyJz<@()$lZ8>hViX=PzwX+-5VPEocTe;%TwOg z0(bUmF!Y`5huxnl0NM@jj^LNLeJvz?3+|MIEZSke01%>h2)&{fli|~?1`&?9hOq@A zE z?i6Juw)vlr?Cfo$KZriaeJ1o0f@_EEbG;f0-k2(RMO^jXdp_*IDq4!whadR1^9d?v z<;V@gpy7Q5#8@aj>Sl(JHFFAMchb~^($53Y;yy>&k~c@)#G;H*L2}oF37jJWyKU#o zX=znBQF%(y%{JK8YWCr08jXKxVsOPeMfNA@Zm6~EgO>c}nw?9rn%^ei#DO7?1DP)1 zjFOMDd&Ktt9QU9;ps8QD!>sib74xL!L*3*1Tyg^;!kvs4%k9oB>s_p2jLd_nW}{+V z_7^v?c!BfQH$Yq<`Jc~}DQ=_V-s@P#d#1QXy&$vgmJmCz6gmd$n9^V#b*;Jd_Jbu3 zBIaT9X#UrMYxGk;XG@Glr%S=)$XcQ2W;7B+eY|2{CIrk@`ePns)TDZP{9bCPb z|AXX0E=evL4{zTG%-&^Oe&#%KQ~*M2cmJjV zHzKI>9_2k{#!$)qMP*<;KCBq{aVX=5H!o>dKJ`AbVJ%7!k{^Ob-l`PTSw4H?1DvpE zee<0p8UM3RI3w9%Xnn|=w{U7tmA=&<5F6G!s)7>aB<{{sS7-Nj-?lu@7=y)P7gjI7 z%{t5^hcvh>%U0b=IqoCPpJ6n0M)<$cA=rp#W+q&Cte+p9( z1%7d#hED623rqp|07;nvD3!Tp&(g`9@nQS4hOzOnQk78fb^v-@2{r5*G^rt(D>V;5 zU*2rV90C{y-)XZ!tL`w?Kc9fpX@kw#YAa1}{Z)(C{x{x$KZGL!1g5DlpWH)C=hvTb zdhsO^Q!?7|3t_4sSiVQfro^qjM=^c}OkLvjS+x@Z3peoSn`^aEf_W+NpiN0JV3nvo}Cy?pMV2PQ}wwCB`-ZCO>DXUv+HNou(3;vkpCKg&B|mOH<;92bCzw7_Ae6Fza$Y^^%txMY<9?W=Ncj;z!(u>)i+(D2tE_(eyX8zP7ts{}H%2Nu4FHgIR)u zt|$mcl!U1;4z|q`{6`^QKk)R5`Hf?EAsX#2Gj;U}%Wo7R_zK&Oa4~fkOMa4O0f+QF z|7tY91y(2U$v1%IMTSBL)(0`zcFf*fZC}XKv3zIZ!AIR#t3R&~J9t;Fm(OXqC;IW= zmy27{Tb89mcyoF{WBp!K4)6%uFpN%>3WKXPeNqS;C-%f4H*Y;2Hc5W0;}8X!_#3wJSx*d}`v z$ZZXxr%8{C{srt$wmjut(v?JZO^MR29YkSCFyW=9Qo|CGFV(PBCvn-=p?+yn`ta7A&FsKkz%HScU?OdNUc;VEi}SwF-X3DT6cdiA&`-x#+wS2e zhud)ncYhvivny1E&J2A3=_7{?3jEf+rm1FxD&4cmu8pRVW`)q^$l~SV8}AX-#iXVO zi=(YLLhiwUd?iow<(%<%RC>`R743{o^96D8;ZwbDbj=dl!R%?&Uhd;Pdt=zvQ+)Og zWl80jrlvm`)Qi{P=wD%SjHkyPm;Ex|iW+bH39ChtFe4eobUwuWZ)DaZ*`{_*3J%y4 zD5=rco5xJ))RPeCe#T=R$w;KOl1ETLyiUZ7A4=SfXZ6WK@ky`2QSleGlRKwZtpcaT zKeiiKtY0AT$2u>1zxS+vt~UI_=pYFp?4Fj6w-6N=*|3^j^HrqACfWz+iRqBuXyw2; zc71TpKdBF;sL{Ib{5IX8kFt$S+EAB&qFVBE#Er}WQ5tKS_Pf6e+b&))|MZMc?WR00L6S29l?G zP=P)#b3_Do@Yl++VL~5$gTJed<*&(ymeqV>feVX8;Y3uq*pPaQqm;sxJ^D)G_-S zB*-K!Bc$&UoXUd620wAs2S5@_ECth_hXjX_cRqHvdJFIZ)Kx*p>~dr+xQ^a}rm!%K zNzL!gF5VFLIx8!8g^*f|CjCI|@q_`)Zg5OG(+Etj4K9~zp#_hzx|uV8TQk4gz4d+B znq`ltjjxS5?xXjHJa$*0XG`HpeRW2r4Xdhn{}*&3BDkL770#Abt@0GZqjJNHU1ynIM`w zZwH#@_VkAQsac|X5z-f?STBf|h)GHK%*#3n#vA6!UqWFeg-1!4)a!*+QbUM0zHX|( zroe^yiV;Ou{aG|z+ctM7Ea+TZSb*d7mFbL&rxd;mM!EiC;qy3A(s$K;sjkh0x&@GA8--ka_Y>oM5X{+8;>SWwE!NV&YL#)Q$;9 zC31;_;O2Ib_yoNPlAv|}tP?}ImR{btUns_5^`f!jse~o-_#V3cQz7!wYI>mOBa{bZ-^zrEHCdJ%fHq!eT z)E!3dG8Nz&I$>$Us?YE_%78qTJ?h;U#Ox9%;w7^nu4rPrbl2&xl%X0TOCD6X9GDa; zX)lh#{lh$r)l^SV_7|$|K)g%2)>(n99!sOwMTOzXQF62yhWzu?mkXx3*d1iXBGk!q z`^Nrd)1+E$Wnaq5bgWT6kWJ>&>-`DR`{KFwt8X$!hkMRda_zXg=jF+jGDcV5^CUFW zf5q0tXc8=5tkBP9sg%NZN;mO{hFZh`psO``N7bgeR#6Ih=TTw8*&O`s$+uSgKb ztU|(uV~}i7r&q3+X(6X>C}S+@aOZ2CAP$f}ngb>htHy*pDdTD)v+GZJXqQCqkwg5* zc*7}fZOVW*6a9H1c&~V>{BZUVXk=d+9Sv64?;G|p&ki%P)=%3Ne?SLd?9YdI)O)We zJf4hhyJQ2obw4QKktl%x#w`z(+oHRHlDrSEQxg18bgxG@JqZ~i2p>?$y>w91zm09t zqky>s&STZVT^Sf{)Rn*+-?w<5@|+SPAmIV>{!(}-Y+_~O-Gj1o=wT=V9h;bsOLN1Y zFDlS@PTQW0)sZY)Y5cF^AMKaW(JayJUJ4Vpd)rR%V0fxT7tjOMSk%j6y~*M$%7FJq zG@D@{iX6L!6lSJu_nu1}2nCQ(>#dUjZY@F2MKi_h=_dd zud0q)JT%r)1ZxCZ3aJQ4qJWb5Q0y`mE=r<4KN9ws+q?{IWX*{|0kyeSfOhh(_>p-1ePB<$Zz~ zg+876om8$v;;XsLZmh>3tPUa8mn0~%=O_Rki!mNA*kQBGFkCwZX-bI|ADcfs@~~+) zv^z~!ISQ2P=q^n{{jP+Y@a)ofMxLfox!{n)Uy)-#uQJ#5Cq3Dns@~}8T>04=mQ?A> zWbUV;Iz%ecVG*Qvw>%g&x#?bTwGt-6IwM~R%#qvAVn)BpWjb1HRH*+`@gwdq<8p;~ zrHso0ClL21xV>?gV(Y>;fXhN130bR4WIF$Y29wEB{cG~KyH0*q@^}n5Qe+VXePk*o zYqvdrFqWb3m$9vh%)FdWOe{cV0_*R4Uu%G9&@c%xDHRwaeqkJ#f78NZj=!7c-NZ4C zz3{3M+}uaJ!K7or#iWZU#JINVii(+aX`e&(C#mN33V_ZmV>Z+`Tfx)Gs6I z!w36_vc4sE>+Sk_>xIR7h5pnV1sJ60-|2r-^0b}cG#nAJ@?Xbh&f@I3_2Z$lEJrWk z`;C}n)HfKaPF`RdK{z7JbC@vT3?0k(>^_np+!&MSVw*c~!T+qX98eb(M0OV*5ROpr zM^{oRkhq)Q*Dl^Z_FVN?d*(xhnu~GObs(5ha^gi(8DYIY8SMk-PN(AJ%5YM$d1r1Qskn zFcb&;GV0|3GZ4RpLSPNqJE{^ss;|KKUy);3C#bjZ2riuYiAvJeOGIxOn{R&o&WC`% zk2YV=$EL18{Z`itIq1cc$(%E8mSI8}J{f-;ZNEp>mnYW$Q26zJ?9cYzZTMvWQ<+Af z;%wz&+PN)q6|+d7_ts2)xJh=z%DAclU+01yfnt6a}OOB+gjf@jf1!ga~YjHW_r?Gg%XR;LAQN z<8@+Lx{veE-ffC<4a46E;EbaIOfQBd%I0Jo1JYxWIK;wIZ zGG*QW$+sd1q1>Mm|0CY6t{C+OJW~pVEkfSBst19QEi}0nPYsW^Ju5-W?3&t5)-xW& z(Y!h+s}uoq+!0ou1u2pZPvvQMiw37ch|z1-y?a)*Q25S}k?6T$#iKTo_aEm7%kYfQ z9Q3JcBIs8Q2dD^HPhb@K95l5x@g`y=zCB>S%zMV;5Q+^d9Mzv-3Ro- zjdZ`ePFNsH1&Z~u)3U+d#U(!dx6{^uMtU)L&c9TysdZ;FOE($ESElXMg`Ez^ z?g4)?k}JxiaFKA4AY=l0K|Y6!5c)1`MfXYr))}ECZ@4)sNx@c z7_S=&wf-q>Hx_y{x-3=|s#LU{B&A8=+bLz66k??Y^iZ52%%CvABXntz+!&ueif!N1 z>x0*SuHT{n=y_fz#8J)F1+VBA9Va?{g?Y0VDw4s%+0fB=C)X3*`ue4m6Dz3()Kk@S z6U^)%nWCF7$3RY&AbYg{%3gc4co2x*1v#Qy_-Q;5IWvKcxtVeA;OwR9=#D=SnQ}US zeXobH*^+ysrVwY*cp#wzKI(CHNQ$^iU$}8m_=ESMga{e`I*N$m5{EUtcrE@tMga*& z22udX_=;lKvFj7N$;ftH7INsNh;S11k_}CSG+PFRHwZsTnoyRln|asHVm-yUKK5V(HVPJFsG%>0y+|}ml>}+!jn1sYi z)iZ_j=+$iKG{aX3QeC$Oh}j6m0&=ssq`JKnxdxv5ZW}=Fei***z+L%qS13z*%6?51 z$_DT7jS`o4oubTfpUJVy8-vlJw@6pORmqj}0y@O=O4Of*0(-Sh){36}Q3i0MzwvM*rZ9j_`!MKp?0?JjSZZaon z)CugN489E&Q*95Wm;RmNsaZd8^T2`Uc_v8(ErA}&bja9(?>%I=SpOwCRsfI|han*3 z`a3RzN_&zwxfN1{27N?+Xk`k$meHO4s&gqbP$OM>-wfZN#-K!{=E>44;0f@)R=&aK zv_?VOLGcn^Q>0;Rhg28~I;?JWFa~k$C8`HiE@-FF|6Cc~rT#>jh#Yh8d>4`t)OT&Fq1kc~%-S9-N53_rs%UqRK^&{WMXppRDB3|vnABB@WOpRZWX&&D-@NYb z!wNh|T|M;dLy7gGZx*5#GHbAzXzA z)B3t2=yj26dA0zOmvKLjkW&g_sN+9Tz4em^} z;0gzNLlnm&ws1A`%XM(*8d&mjpHwUzGWsUl{H*H>`Lg0x!PJzbS7!l^&vb(?fobe@ z$rvsvSPT|*u^}7i$){;Fe4e+ejU;(dsSQcNU=8*-z9x^WbOT4&RtuYzv*|dJfH_a? z=I8O!c{{Ok!$#38yg4bv!{n6Rl9$LCDep$8kHc}{UqgsP$bf3B+-x&<@a78kfZ&{5 z_HqJuVCDX(e}}j_=ezwm24S*0%pXMiOciT|jH6-w=KP3Q2}Y zg}8nP=$Mt53fy99aZKb9xtboCZkaKlAMyBpvm3^GhT6Esvl%1CpUq3loAj}045Tz{ zGFPKMTVL)TgnOJC9ENJ8V8pRHuq%rqfciPt?qOW0Lm1BIqks+>@j?$mjY6%fx#j_- z`A6AMKG3M`5j9@l_F%te(X@IDAw-gH!&oRbL8TxR*0hg9^YG!V1dW^{9+6oS>(Gt2 zHnmQJ#jmBZy*qC=yPJZoP)qnele3P7yJQY$pV_MIi|nB{7Y-UwI4XJWO4ZSt@<3c) zUjMa4R#Uz}yH6hmYY82G%&2*_TRT}>S*yNo*FoTK-g~b$?NmM3L_QrMK*$(1WRUkm zC0n5Ss0ZY`DFMF)r=Aqq?~K*0L%F?TFI>90Hh&aL>^-ABb1(5~%XGEtJji}VlI$E) z+&r4Qvc(343ND%mYJBeL`;>amljZ#%MWGVSKvZf>o%Q$O5M0g!%WkAKR>(MI&e(s1 zv)x9l7`=2YvL466VsjMzpwZ_bQ!+4MRM6C%>;e5O7d*vnLQuDk<@WSSuc93E^m#_?}etUf&$wl^}~ zMSl0RYWT;gr%Js=Ke9l3oJ1#n3tL8iA!U{eO_KTCltgv+9G<$* zbD@?~oTXmefzHpVnuZo0@Du$q^!jDtH=d?`4=xXLHg&(~Q=x?)4h!F{2ET7RJzk&N zo!Zk4Do!>XJWw1R)!rB%8lB!97aU(@KN;H*+lfDncTarMs(X>WarI(Majr{A7@=`b zO!bjFmf~XXkhI^FFfZyQ`FS|s*0yNBfPd2Y=`$I!f*hpOl5KmL%9S{! z(}%vHK4fTYY7Qj*iO?}nerb7?=j#kRewj6&bGE66=7hLK7#(Ht#cilwVDFDHAK{f3 zS){@FFRE_Ow67npLNoY^DQOZ_@?x&D5AnOS_6)pZ!b*igHUF@_(DBYmfbebLXctXj z-5?5YblcjKkI<)bHfbOE!$OfWolMlxZm=HZGVHO>r!99nu9yAQ@-ijnBTPNN?eTQW z@sszb1K}Hwx^M=~_w+8|5P@ThbcrGbx4amM@`g~@)_ITjP^ZP!!5g!?_BRu}yZfhI zOxk%v4mHIDVWl=X{(>ZaL{UnYk;J8x$Kgr<*A2!Zdcy`sBLNr=B&i?E*vDdP`OTam z>2@5*@(egs%e#aMeOaVrS=w*Z;vs*NBeFloe|MA@Ngtyyo&fxy8W<%!)NEVD*rXX= zB5&XA8ONE;uJLu2_TC20_s86@go)nt_nR<|le84@677Kvj&53l%A+^9MyMWGr`JG- z`)HLk5<8arx13$?Bk9^0LbcMU*0$NG+SyuoZ6VQs>GlL1XnQ+%T$*-6*!-TP6f;g0 z19|-9a{X{FFSDy9{mqlJPTs!%N)#hqg3NpqR;7{t_LshzT>>K-iw%&3EdsCsG`l3= z8UW-$RW5lisWiVwhKbuZ!q=H%OqkcY2zxkIC%Y+bg2ez*_5dS$Y#j)3-4rX#H&;Wm z4~#m0>FOA20W*)R!Pc#d^|*Vp*sb7w6VCT*bg7p<^Ipd`4_8uYnOjasc#1XEHO`}x z4D#S59W?m%yKB*F+I6XqCqnhYiXj-P*k5lUETq>gIxl`b-)T$==V0p-x{IS5sej+u zz&r?*cv$NCjV3p1Za<3MjbUSHgina7UPR=|NXA1r3jPpjrJlq?K9xe#NsNfNSahte zE2+x)$w~)e*ZSC5i*UAnN>FU-50f$OmzGsfr)!ek&0PK_&LKi4zHEHD9Dc>Y!5qA| zavG3gHdpN8u@R^xQ(eG%*Vx|$GvL4M;+BX8=N2?u$HyS%>(LtwO;9t zz&>pLkAKs3Y=jKoj`hX;!av1~NST6tgDn7o2I7MduU-DUQzP-cAIc@v0(@jdsKi9U zFxImRp{H2!_IuwKSb3CT93dajXeqd{gNVwA=>6Qt%t?0#Wu?OAgH1cLYYf?lqZqXq zv8#*4c1uD>NIjGcWdYKZGV|;F^^GUnM6yIWhQb#$(<%WE(uU?2rX!De1UI5DbzDXq zyw^tG9`P~75-Uk$VF-GuOor1f?x;5Qkvjnc$h1mg8itp)%*lf5p+XN4ua3{}w~y9N z?FYe{;oq=;LoQBDW7xV;hIKi7EH=^-IJ(5&L!Q`7I$bJgh8D4Hy8QOXAk8c3IFPhO zWPaF-xtWw4fBG4-J1`ZabKB>S1#xuU{430hlYI$e9l~_K3}__H2{|9T-BR4t0WsMf zPYc(QM^E-ovH3Ty3|J~^CHY|ac8;_?*OP@>HWDuVrD%#orAq@U`7otkk#`VQl1JgY zl5)G=5?Tq141|eb5#rSivF>Fj)%oST1y~-9=_b;ZCtETQA`V{*P51_II^||N{R~U3 znLnFHm;OFs^y|HrP3ZA7o@I~7sW0?C|0Vn(`(4Sq$=Ay+689T^4IHvLx0=JszOcBq z*%dN{sqoWrjEPcbN*z~TV(bkaFQSiEdmJqnF-?O0L&$>k(-?C!CFAx83>AFZyzbaG zq12jU(ZR>bz>N;Vj*_}15uU9cgTVf_nK^$gEimR_!R6HHjLuCNL1zWV$w|qvr3mqf z6G{`Ja!KlBi)W;N4;3?}pFlrnN>}FsMQx;G*<^!7s=Ilk53yyi*+KIz#Nih#_KDEL z+{~00a9Oc)Q_y>TxXK1KpGQ@_UD_|88-a+9m%^= z*QM-rKm9QK2J0hoz(=kPlGcd_J1pG$`ha&KQU-dlj}65qb+x;rNn!2#nXg~wakM5( z_;R&r%ICO3A^r)Pe~_`8cwLF@rtV1HMzNqYwJRMW@o0g3@`*3dIIk;=&gHC0KsXmD{j!{G|N}p+t1q_-e(I-#fkrzUJ>f%Vg0*u6^pVT1gt(_Vjp~ zqL~cPXXy{$i@gY;AU@bJigcLDCO_OG2qE(q16t{_*l#Ih<1lE3pDGx+CmX$WisAg6 zZr$qmDJo?GHUq4r)V0cK-_)Dc98QV|)j*c#G1V9UE6+zb4{OO|i-(Z`@3m-=&)M5~ zlR2D-fzmtFIZR)c=WtLS7tK8670EVE7;(um<_7*o_%0hzn!4XOJ}9Q=t$h{aH^whK zU$KCGZqL0p>RU5&Y&85qXXniyZDM{2Rw*-8x#r=TtkC-&#IudAb6;6t1+J}kQQV4a z7E1JR3ra?EBIC7cvPm5_sBzv8hl?G!Nan|)KlJ;%o#SO>c55`6=-Y8%+hz-`3f`nW z@C4)29Dgd1WZ9nUT5YhCCm?rqz%8-WGZ_+MBA_I{p9h)o>H^bVen@nR?F za~u8Z2Bd!KIMGqj2)bUv2XjPVswID$$KHf}`m3nlGJyUjNA^x3X#9vM;X>aErw6NF zdpCt*$c%Hx#il$wqYP&ZUPdFkCKBl5g<8lHeCV8>puFIR(YNpSQSdR~{w$`lxTVmQ zNs|u{0J}>$+V+>^q6$om|B<9gmOXq~yLZTvh^MQOm2(G1ko&XWeNsI>k|BF!#?fsu zDTv7)w5x2lZin}B=O;L>YT(m7xBHR~{4L)%?Ki0Z(%WRSpG(f=X`h1sYkY!TIsXnX z*=@0UmNl~JqDLYjFZs)+Z~iXEXCf;z4wdaQ_ATHr{gzj)mPB_GK7(aZ5QbZ%PdQ*- z0bEgitKGj;Gi*JdIx#cbP2GXj1%Mo;8=)@fT19GGyB_HT?|KVkG_7Lnan24)m8BsrJ)1a;yQg~k>+Zzw8zxF| zdWMx#-3vy+X08#i=te=#Y%0r1>lhPmE7b)*)OC*WlEm_zz3YEz9N0YONZK+VMs@6& z0VlFK5F8lxCb2rp!55u(+o?;u*y@x#sxBswV~F?P6dOGDp?OTT3($BBFkiZGsjP}V z_M#fXPhtn>AXeN0Uxqp@6T+?el}mM!uLh;nj5jNGzTXN?M=~F? z)Ec#xgIDcWV@h#%U2Ma{FLt=pjn+Hxao6ovJnFma#g2CgBM$o6C*Mb-r3Q(Fmmw>V z>t)$a==4Q_C9UM?o)U5Cp2hJ>%ba1!#!prf|NMxE2-^e6*69xrcy-!nHJ_Z7wF=~Xq3YzE7URn<61wQaMJWN~1wVMy9C2PL1bmL%0 zV`4ztR$Hr_4Vd?6sCY$%-OHde@mYRPs!-Q!-#tNM1F)KDMWgeP3r(z?Hpc_c9AR)&=be^4=4T~4+h#L8lZBB37p{gTEpCHJWTbU~ke{PY@@lG067cZa7g5!6( zt+Dvy2b;>|jTUL?-eJ$Z;VOkp$zq#tgZS-MV&9^?>gkccWzlEI_78*OJ{qT}BD-tY zmYKk4Wx!P#o;Jqh+!VF$e)~|>s7^^2nl>-w0XBTS5zDP`xZ0H!AYi^O3cXb~fcSY| ze$R!1FaH>e9p0orTu!CcC)~IB-(T;g`bM73%+K3;Lj4;!Kh?1x__`X)%*e-wBarZR zXlX1q$0k(j&+Zw{lK)=gv8Ws0RY(fpm7+k+_2q@El7!8jmtCgB#Ex2qX>uMkus*1u zWFzQe0%*ds&w3BXhsNrO?v4Hbz+DQ!P<1@udcES_Zp6ofMD0weFT7+*U=5t&&gMWu z(5))$PtdN=KwZZv!S)ypEWThgGD_QV z1@bN2*zg-Yrp?pXBTHWBiw}(t0I2wije{pPW9t(Vu<`}x0u{S|PXJ#Py&yf}QZD@2fj zgw~O&?E@x;=3(B3CFVn40;54Me>k-q*W2Dn=8&gBvcgIiKHs;w^qr5L(StN+CMV7R zd5Xl(dq@*(<(GrM^ix*{ za(S?OKb`0N!|g@E14-$37Cc|W*D_?!y?$47M3EJXh#Yq$cr4XCDWw*PgAg=y)=YyB zS^ZsTy2t3G41K2u)zOIuN3f6D4%c<(g7uDW@ZT(do%x&xTa8B)10Osqo|evJR`I*P zq=c24t!tYRBau716WULK($ar`J$oyP10tPi!^{KQ+(S~Tn^%^Wh}OoCz(TK&tFNsD z$XqX(b%+aHvi5n1HvD?xQrdL<9hYuKf3|=pH zwGHV8T{^kA5bou;i(orhS!_ca(#F}P#_B(2*AzuNt>5f=0tXaLGs8+Ibd;Gy=y3p$ z_7m;;+X2!Sb4lhzrF*rinZgqv%Jt0%#m=NI`I2uww?5(OOZE@zN6{M=n1tDO3XFov zZ;{IFXDmY6y7(!&iMCvmFz54l&(`h_<%VW6y1#z?s1DZ*pYJT8H9pBdRejyDsO-7w zUCjBhhQ;Un6~de%r+P2~|7Q5!hIW&}j*XX-jtSR8IR!)Vr0K@E<)~(uNxJvm-f)xu zICl(U(NExuxZPz==pP5V?Bt$O;x0AA} zRv9>rJW}hzL-mym=>0DzY_mB0tnQp1BDoJEh!ceshrGTD25^2cfPZ`|XZ2n*Z8p<4 z61%T$Q?&CL#^0D=6pt16IV!XlI5eP=Pqo%+0yE6s)ru?2o4(!RX#$_MdZwEQwy>@n zuaI_N0{gi|74Ej@YPITAiyla;u%cjeQ!AaLlkmNy5A0z5Id^(ZFN@CG=O-^~N$Tp| z=Q|nne=p6@M$-uTIM)_7iXX>@N$I6d@1Why^VjB23>GT5-gnVo$+cIQRo4ui06;>p+UOe)M zPS%D~!9BU~Pt5$tcQLaum5q$WcbwKrqxbKGPB?P+$3zAgaGJ#U&C1wILC0J3us%Nq z6AngBEETx|z=VemgDhjBrPs&YSp9v-3ux#+it<77!o}&l6;0X=59x_VYJ(@L~C{ zD?t}OBY|G1RG>D+5V(r4i2yXUeh0sTHee(=sd_5!%H>+((2Cs_J%VSzfG3s1&QD0l zCcl<;^ym^O!R*1Xde%kO1?Z3BD_8E-aR!TD>rZ1@4>{)4 z-=}ZS%^!O%JF29-#`lc;W2~+1&M(6L(t~FMmob|7Epnr5oV_-fZAi<+nG^maeIh8( z=A8SDk-gIvjSQ(dnZL`dP5v|O3C14AOkNlk^rfP=APl)G00Y%hk448}vF;^_=>Wge zYOB(GC-(-~5L08-%}0`fe%`;0^)_SKAvn75F7^i5NP>m2fa_XY5e2zMoxkSNO3i{K zf62Lx{SeFY1XU68(}B@AUng=tz=bmHqO=cZf7epXLQMM&z>3FoM*NG^8Pvy+`;*wB zr%lmtAPYsQpyM*7f1t32SdYku8)#slwTv#$jBmxw!(%D7%vy!g!#iAEY%35Y^A7Be znJwZLh@;yLNR@WE8@B0&jt^=Z)gP|$`dauBm3KwGDkJD}{w%)8Z+nbAA|sjOUeXoy z>lOKERZ&`^{lomk2deBBy<%CO75AGtt>=MDt%;M3$ii~85wdXvHrDJ0EB z<9~Dd`~xw>&zdxQm%QjHchTvkpJl}2!AUd%R>pcOFBU>`2l{Sn0Ozf}gG>2p#%CRwg_tOCDsiJ$PydmYJo$OP9WL2i^VOFh-T(tmAO7xuA zhqN8;C^r>y)rQ+z)`c*aKk~pv>&d$HtwDFw%u7z1i%m9%1ARbDcn9?NQ$zj!8Fy@N z7UilikG)eEXmxqV;+@Biok|bwgFXeXl11<9>O~mhp)fhSSS-G#Mv8^EI zkAJ@CvV`zHzOhik)xt0tvd8zI%mE#tCoNAzt|bMzMRrf^rh(A8O@@JE9v5XU;)xd%*@Of znaZr3C^Cj~_;pj56BX@RWR=ct{{CR3O_wFxV`fQ#j7u!H{0(i;usAWSmJOjmqopb6 zt-`SZD8rmjiKrycf(hlAFE`K%kQDGoxIb@v$vds}`)RtccsdwwC{c`#Sv+7wmOfI3ECIbC^@xO2iY_-Sk&7EYp?l)tLm(AMUR~l+U zq-1nI@C-#^BacTCckm;daYMV|1p+0v(O2i+`CwtxVce+N^wmjt4@^_meP|}do>v%3-t6|7ly{EpIkG+?HUm+%x0&UJQgV&<< zccWi?9G>e`%Dc7nLnmD>^i2o^Tmx;n1=?SSS^0DCgRns;Xq?3!3v=1N><=GiwfTO( z#LUCxjX(&=kjbdOv(0zPpr+e;tQFR|Yr1rzO6ze4;1!)&l)feL7X{D)=!-n;zI=%} z(xUJ;?;%hyW5nh4U+xUWVqX78zn(U3(s>O3Zjj?e-Bo(Hr)mx{%Msb(=OwS-qvh+0 z5u5Ie;t&MBm#3F|PEmT6A(3@+Xy}F{zbPllXFWGW<|yB9YcEcu*x-tDYusx~RIRGN zo8N5wrv239C8yfOUi6O~|5O}vzrjR)?7b@r{p`B)7dfe_c--kUG)_Ryzq1*0hb@udWFvuI%zZ!|SDB)2YMF{3`%ZsgGmj3gL0P%p) zi@q^$NaxXtdd)MB$4_T`VFMK5N~5G5yx(I!^TJh`a<>&xW>FZNZ(xSZ*sV#G=MR1m zoq(ev6HZ2#g> zuh31zm3SAS1G5Vkf=ANd!6A92%B?~zN5|xAuw6yFauY6S>bC9cweRtx-s9(H6%jTx z=7^4hjHkcZOIKOnKWkKD>w-$mOWRsbnqT?k#RQbyIPKk^E!>wyQ3VjJo6p_;f~DRO z68BV2<=Cv088|k*_|y``XV@(mf}=h0uj@tJ<{%l<(-Upmb|Ga9Ww3ggH+i z8gs2lo>5*LR@J?z?#JpaO6-;y8MyO_QBh}|aA~2o&CLHVxj}n?|8zF8X_61vlw@8` zJxLJc-DbL>z=WSLCA;>KNALe_uwNjFo6Psmj&dLpsqRe*_~rYRH$G8wyAZdX`%T_Z z^n}f7O+&Asqc{fGt7E5%%$QjgjHlsNN#XD82#X*kD5cxm%eg;@`cpudFq zIwyZ#=-j$66{L@PFNDe|?V_Th;&oUi&ZlhPRr@U{OX?encpBHXU8~Vx4LV)>2p6yQ zkNZpbRlNx2S8JlVyE-9w9J~Igvdqjt+r7A1*o-As{48YQ*R{4s z)vc+()irX-9aGL;3Ujk&{QLJH2J_$LcjpvyB{HdTHrs z;Yc;%!p1_81xNHiqID#;CM{wxB#oAwM%7O3SeVm`+Ac`vlG8H%WXf|HrWG(?M%Y z93S6J$PE0{+zx&xhRtls)M(a2982yCBi3zQ-@WzM$abr5wSQ>NiA}k3-M@Ae3g9$a z@RA78*5VLWQ*)6qRg_X$O7+YU1K0tM0n6CBeyn@02Kxum_v<<0?LfAh$Avt@z`5^& zd~ZD@Yd-TQ`Xu+OB99J&ylUVqx}LxQdB6N1Ek-it2_gnQajXjXPHpO0JbjLz1jksX z=j5H-QjBF`n=V{vn{L_A2Dqz7e+X3V9}+h0A@*`PB|HWSJr;kr)#Jt1z0@Oi+bbB& ze9Q0i`5AF|Tp9|I08vfr)4SC~HVx_!0yMIz8%Fa+rO?JVfrtlNfteXS6S`=zB(@j# zKUEo#rYdaq`A$K9RY7Bzxf=l!_S7}w2ba-fB8_jrya=z5#*kloly7?KLgtGkHt?>I zm5S1!#!`8j+cOVGv=0C2tzr{04w7+^$oeeYZhEYQafq~xVO%&di7W9gftN)irDsnNt}JKLrF?bQ5qhGWp+-JZ4gYzmccn3f{xt42_y5uD0W z#v#tAh2=q%*FnaM8-GILMt8R+0(l&OpJ0(5c2kNq2lZj45y4Q?`(4kJ*_m_o9LSGpRJq6oU6RE~NONWj?;n-MGXMD90c#g?#o;tY@!H zc`sJ5;v=rg2tFCWgV}wqd$M!nRT#wx(=h%>JWw`mz|v-g?Y^>Tp0{crcjn|ZmFh6 z^v+(UL_b^V@(Z+qHdD?d5aZUw+^&cRoGs(!4Q`VH*>MMj zM)|o<&9Nm*b`$$R+h%q*4zW`J5k+)o<4q<75Ikbt$qh@T@iJtnh5Y z0<5f9w5-}|9t{3^1p0k=B)k5V5lAaHbEF>5p2zV7#h>{f>&pD2Vxr(7FrqU4Q&Beqq^DmGYip+UfG_jf6*-;O&Po#G0PNQE9c zsM&rW`6}DUnF2_Kry-oJEZ0$4;uL@c`_A}xPk_w zW98TID5*bEG|=EEN|%&kb9ara3%B1LlugjPfL*h5?mHticq}(Xs?{yoXuH$yN7Aj! zB_nwDZ>U6(486*WKKNN2>r;$`BFN%EeTgm*Y=RdkiFP84fa;cxuOQL&Xu zINDE9J8uTBr|MY&PzN)Sj`&d`zuEv@uOH9F&Z7vlA7T&Qbj&^!$9UdBanYo-t z-t5|ABK5I0m(|GQxe9pTq|xMCV#;$sz@OvxCDNs*nv$iDwn!|=RS-{E-Q%OF3RQT@ zoZ!d0239e1Dl!qHcZNVXZoS{fLbyUO9!-0wd-D4F6bxFuT%$QJbJ?iVbQ&V;6 z7YM2Y!{Gl!mHFi+Gx6kaKv8>^gZ>;#(dy8F5Pb{^g~W*fn0HQKZwy(2X{Zs&?s;M} zRa=jnZx6wActF@FLX0hCpq~Z*;)Keevb{SGayuX;`S_I;`=YL0yJn_{mHg_JtFyPLY$HN zoT7_STMU+U4MXl@MeKsJTYXZ=V+!C(q@}q-p5|`)m0zHFoAsY}iAASZ&1_mi8jZ46 zKI+G866{RGo*0+W8hxJXpIJ!0C|=CmC;k1ZTqSv=y-cj$x}&sa0od+<=syk~Zs#%e zmu+MJC#dMC|8thfF>nS~WNTmlu>sbR8jXCeq&TpY2`!`O z%Z9=Rw_?s^ZghpNNc`Az?;9)rG7~7D%Z-Yo>z|g?O_dFNqf1u*QflCGJ zAL7)EC#$iOFoMlCu&G35HY>!Fu-TvKLGk}b)`%L(I6VO4(j426X4U2jFEZE}V(zk;ib>eNy zv86$DTXXk7B9iQ;FsAL%v$sf@kCd}Ve|4>jhb?|7_Eadh8yaugtSMJCNvKz&1bwNv}U1PbF-gg)>1K#JyeC!PbV6cZ8C z00X5tv;M_PQEb2O(jO}H*C`Cdq+nwJ89)l)XOu4BLo(TVTszMjYzdO`>UB}n46IRr znzx-Ct_aG;M!F-#O zxG@5VVl4owE!row?_B0-(@IshWx=Xwvd;M3(O8*NaMVzFU%{FA)fK8D!i;7mx7IPJ zb{+Qyh|SD9s*(4}(t*1U>F_cML#{~Y-N2p*3Mh+ZxSS(K+iRI(D*_%qn%uQ;5!^O) zG86?IFnZ@u-d`sCO1z6AX3AsBWDazH1`)P8_kX;*y8g>QCMQESS2gn7 z{hz*3M^Czex7+%d1_x}`dM|XrY01h-XSG!LnMb~ksRn3@mnG99E#d=bG zdtz23;Y(+{dUGy22+MHj024{gem0pH6Vaxb^KzDm%fh3tx)S)Q721?_Lyt-!b2KI zdIY-DUl#0o&2qc%GI+ZEPOKz z91c6HoW(jlQ%_hPVA~&S+IMeh(B`VIEbIsWDhoR8$kB*eyIqDRN~YDCeD>D=t6-WL zKV4b`UIcqm%0Ff*FA{3&|s85bR0Jq1T`4 z2IvFe>$W?X=5eaw&nk1oY3esr&(f}PHlqEu&IPN$Yz>1=TYtPDaGOf#Snno_8ybHIfNO&+lU#P6T8OsFPf6aie*Dq(EV%fR!+#)&fh!5)* zMG$7|evRz%sD#nVxYd2K|n1d>Vjl8&Hd(cHLjKn_}=vm;|Xw>&HXsOXkvObKM z(7yP{lD+dc&i=Ch)EG|H;yckkmr*5xJ#~du}P6K=%X=}Y1j)RYXhIDuZ9XNOZ+Xy7D zn;-OjiIqscX(qF1LHZ|Lc?8=3v`u-uiM)k-0&%o6!uH``dFf0A`U8I&EeOx-5I4rF zsEtjjT$T#hqb*>A64n2fn|#S>zU}{@`#!6+(1Ms<1+Cof;+P_E*n>VWZ*S2AS>d0G zc&u{3C9sz(!l(W#mXd;c48Qt?4)!VoK}x;gkySQNV)Ve5V~d*q#v~k;>N!Pr9>`QG zt8T<#*|yi1#-NLOqf*X!=IUblA^9f)wjQ>IO@uG)H<71nievUu53HJP!G%O%aQv2` z-|23I`IH6EO?Oj|g9UVnOcMB{O0SXd%_}=&*KF*`FzT|^I0tys5NXr;m>_Q7cY-c) zcZ)cBcGVhIn)m&c{ab66wHDX9E4OKz)#l0$CP7G3|Zu`24V9Ce@ zUk`~Ie#N*7APujd#gLsvb*fVBpPt)3+nxQ3<$p$haQ-xvR-w|YVn%riSn6xSUcEV` zqf+_PY{!eBaQ>yI8&+twMa8KZJ+KfkzoJ;LH>FA7bg1QYsH+R^p8I81dU@)N7SAnk z{nT(Z%|}0L1gVV#hic$+qx@*Q#oiy&D2bw-;=aup$l0%H@E21$XF zQ^o~OosyThm0VPqTaK@p6qIutZYJlT)J<8O7U1Xx-@SqS?3(gih0|ua1a6Y4b)+nc z43$&sL-;j#^LVz*v{mm`(p8U4J#Ks!Qr{cv$PeMjc-#Nz)0Q6fo#3E_p>pD|SI;1r zRg}%zl|&oEWVh(o6Oc7Cm((q0g;#g;B7M6HX&(|AbI{lPY7zU2V6}SA( zqH`#j^f_VoqSjq}CU(1Xa|2e9@pW-(8>d`k6g2VI^S{L8h0f+G<^hOWfRV`UR^~S#O@At~44LA7q!CSg@WyVZ@f`435C2 z%ZcP95yiQtQz<82rcQZDeYtV`yiB~x#CmOgk5u>#k+uC><~ z5hOMa#7^a8WEF-Hv7JXx84s!HNO?dshom+tq{(yFZi%1+F)cGXR72fFdvN&L zo#iEyT7`Ja%uF)*U^PXzy%+1&tr$_YMYQ}qZBrY5F%c=gV~<0EO-QnX+z6|!4l79W zn-GLm;FwYtU4~E(btXC0KSa-^bXr%k`&r3cQG=RUwlKicI*9eB+Xn@IO|M2u|B;^> zms5pI{mO{w>~g!x$>S?2(~gyqGdXz*U&}Rmws7;yB_j1-OR*KAy^S|$& z)za^{B$U;+hHz24@EGU4|g>=_u-F zjmIeU+(;q;aNHenPS!0=kbI+N1WC3M*TSmLf9kgjopzyHAK=f!T~=Wx3AR^W>%~+-oKeBQvk_FE1|-yUoJel_KBc zUc0Pa129fmphc#|ZNG8_4zy`aSke{M7m@Co_c{>n0!T>q%wCKiuDb{+0xGfnVzPPZh5}O-Q!@LRk02sBn56fmsrD?dyvBgcppOd z_C3t785(XXljYL>T0jWjvmLSY%Qj^oez-<_9?9<#aB0zPqIPQ8t6pztb}>-OI;71? zZ^vH_Zp%vl-kxiW87BTQC0paedyfstVIpkkEGSvM7Oo4-si|K0An$|Q`^^I%2oL_T zbl$p^=I4XhuJqd;1J=DxcG|9VeG|Ip{t5}~oxsiW4+M!&2P0vy?~Y*Fmk#xx%Ah_oZxwhpt6C=`N@^=|oRy^e0K)?HDKyXUS@`(K2mfncX<>(%H_<7sH1tq=7| zXPEHI2^d2n_Vb8i8m#$;@0Gwy-c7Wxsswj(T;7b-+|Sf?W(yc&Pt*0`Z^GDwp@h6w z7UJYd-)2#Q6Xl!CC_!w7~GOC%y~5rSOvUAy8KXDul25 z8W_8m&{8(p#T5bHC2BgVow0%R)PBuB%X3GDR_X_?&s9V`Jjii_vf@Y=#HTh2e{(VLf`$$(vOKzBu$_p`tWrngB1`hd5Ew;wcl9_Q{3FF*4{ zJ3d)WX!D5#q%s|HZFl3|)&rP{e4g^w*J)wQ`z&2^G`vJLfQVflZ_DX!|0w$R&KlrVRn|PywPFHjOv)Wsn3pRQW^37Y&BSl_u*7xKYweATdRG_u!Og% zYbmPCD;wUlx}?((=#%uj=yP}Y6ykd$mDCv&O4ubEiJp?RkS=cGw2aq}=P#BLAv@rF z-yosh+rcg*{2b6arxYvsGKEj*I~Y)v>Ht%d5ZVC@vqI>)=ys9MLh7}@YR}xarlP%R z(>a#2;9?x@Eujb0i%i{nNpy0y?fvv#lZ<*JNe<>NV z^3`^wt12gVGro$no0r&%d}aLc8SgFHxgbKfbS(`jh-9Mt2ui*X&+xfN=<3uMv-u z(ZJrg7+l+h$bF(0<*A`y`y%JZG_ns3W+38f9v1OxOPbAs(X^P}UZ(OeCfkoRO*9md z(~7_?-(1Kc(AQ){vJJx z^Il5g@|e?+9nLv+StguEJ1nF&KeHb5yC>QH=G3i{$e+^u(0D09Pv=aRH+2sOGqB!X zqls0$oD*;K%b&$bnvb}m_254 z;dDs?i$?!Z1{E8fw-M8t_az0Ds9UHX*T?lg4f<5-w5nmkDk5gQ`7wAOEoL1kc}_7b zSDOQC@e_qame8+0jDsb$ozYIAoKGVK=YHZMXq)twa+GqtK0|phw6{tlal={-l zR>TdZI}-Ukg)(QyK9G9dop%~&V{U3dn><1}_aPg=ZF?bVjfVN`6%oSUVP5Q8G+@-U zsN6G1%08sY3|2be-ElJds(p1@T2bkCQfDCu7_}Y~)aD04MyJm_mMPl6mR6N2<-^Uy zArv$(%lQaX2;tRI=QYKu>y!Lw#;hS6ZUNLb%M-1G0Tu$!O```&1jfn;IYOM!F5`PF z+l^KwqbO9C>IFETZ}d4R6GAcn8I|O$*0e3ee|37=LTw18^j8WHmGAW-gG=c!L#S`e z$B$@S%x;?-E_2+6xXNDED|8pa;cag}{?#-uzW{x;?_B;p$0xcODx?0*)Tj+FHGbue z2ceCC`-5IUU=J^EjGXj4xp_qc$~%+ieh^?hd-O>-h>4uu-9IiE%qa6vRtdgj&fAkQ zdXw=A3OIoA=|<+6p=ft(|`A z)H*FdmA-LZxWgdE^+L=F@9f5#;6KV)1~3@x!ezyq~+cG_?YYH_h%^W)FG z(&ha`)Acg0HqwwQ#zV>!T3LnCBR)$eHE#HapXlW>Bu?_DMXlwPt`#1Q^~7Wt~e&S9`W! zdd&NGbjFipyRd%{f4+u%gz9oL(Eu)1Vcw4*#|P@!tgIH>oeRjp|5lbnJhKm@W8izH z$1Kf87cfup>H$3;5Ydq*A=C<-{9DEAxOOdGP0tY8+?dgN+IaE-?c>_Pc*hEU0@D zp5m{d2rbv`#~rqLyGe0uTr_>%`V6nZk{w( zm_%f)L+K+MuV3{1Il)1!^Uf$n7q%!mJA$P?W&gy+>WPh0&;ww9n@6Nma-<^&F~8^4 zuY>(|trSMq9!%A8Ql{io1~sH#p7Seb4OIXomV4*6R2V7JMRSz8v51ifx3!wPv#30)cPYj&>NDlFS)F>lA1)pO68OAiKnP0G8uw-CmX zJ5BvSs$Wy2<$xDkcs(n3MYmHDL)y%g8BVoDbsPi`fVQXEL&JwmiRJ^l;wf@}O8HiF z{s38Jc3Mzz!iQ`=ET{@{#ZZ8#x}&;G>Jp^`*|TBC@9@fTwIg$;^2<3rDxQc6x=r=U zDfe+H=|`&ibQ#?2d3=G-%g^rb=C;ROFX10J(08*LRsN-KGSfQ+IBQz4+pM>_(PH z=u=cy!`1ux`|)W?BRM`z6b*6Adl=GCdnI@6P+-$irp%r%49h^*h;2tf zqpB6I!h>hen~5IA9s4?-ptJAu=vk2%FeA9PQ_0Y@eNh9_*62xX6cLt!ubJ0_Z{|lt z*l9ZeT|>~aG4EUaav;D>Ud9n}rZP>>ETG%a!TkW#IHG@|^Pw;~Hrqc=bBEqWqh_*w7JosCQIWI1y!6h4 zFD+iHFIHIv_)Jw~Baa01tXW8fRG(r-eZSqS$aF~!1a4$=!Oz5YiX3{}MK$2d9$wzj zN}4UYFLCw7#Ueo5sx$Ys4&+#3zrYpZyMQN;186#JX!Xx>Ia!h<{KM-=D3S{ZT~f!B z?Sbs`a8uEz4%A!q`%W}h zg20)xpie`L@~g|z52D9VpK`qetjC)*(Tkvc=m&JX1an8+UAK}(ya!?V)%PE?vjB@q zsL-b)dr%Zi*dTb_?t{L>@YRHz>3_ENR7T^+l9LOsL`YbG4&khqGzF(#GNX(bsu z+J3#Dy6n=sb*b&p-LQGjlB62^7C5lnxUb7Mujv>H7)2+UBg4sr0#^H{XwQ^Sa9jG& zG6X2|7(#gF-@cl9I`t1D%%JjF5Lb7Rk}z2vEM8E4%CL)#Lr{=XDp9RpbM--On`DmM z-u+v!~Xd8 z|7=LAV53ZkeE06lk-BNTiRIc;ruM5E{m*I(T;A++W8)r65v~Eg>4m!oFR(e_Pw;C0 zs`2p+`N5-!Ax(~YCJ9&bqy#|O6h&orYXP7IH&Jd%J0?z#txk*dZ(nLSa0Vjd8ebzH z4(=v~j3T1VO;YI{+eK)grj$a`2kURWVbj*jI=}2@T`^KH7uk2k4_tEJTM*EAS2AW+ zBd8-=YD0(?_*F7|nu6(zVCWM;{HfhPB=?k9l~+A3(8NRWiMmet;v(L0U2Jc+)K3xM z2%qho%e#GWsT-8I7!{b_Q{;^cCIao#CWp*){Nf{Z?}27LToy9#cz13d`D8}wdDW8M zbS#k$_nly*@SR#DNa%02kf1uNYvu6JlF06wT4^iM`j;9Iuh$J>E(apqXhTizSMK?< z6XT||>Pwx{LyyQM9GVsNK}I%*c%RADfO+&LcNvhDPQ-{yF5o93HK#k?y=P&ou&$Jp zF+X1=Jt%$o=D~yWNi$^|Vrd?tzd#@d81kk6qCTqA*t+-gh!4cB(IX2x4BXPkfer@5-_yJ}8{d=&8-)-}) zb&%4uHtSRBGOEw(&m7!!QrqTDSBCGC5sCGqK9f-a1*j~<@R8g)@U0&Yc_fLODj~n_ z^wx;ESvq>uOVuwI3O(d9^?OV}=<^xWJ{PGYS(2ppglk-WKk3EjD`)>miSs527&&kuPaTjk zii$Dt1!+Qd8eId9L-uHNjH*#F3z$W{Wx#&Cm-u|YmE&}3<8t4LEY{Yf zNgMq0=qshOqLq60|Bh#=>gq0FsvcST?y- zO4WK;G=I+^CIJRF@ASe1BOG!OhsYUVLT}l?*DjdLctewh7Kz_TRt|Qq7k`gL7!-i+ zG~}q_rs%#j>N8S)Ex-?>TYb2!Jp#-IZ~zO~lJ)Jx$zLyo4+z2-dLfi18)i2CK3=Zq z?gv?23uozB@ruaX@$GOWvZ@Nr67>8ttmry}b*|9}3DBEU!f|ck0TCldJrwsLhJ5;J zR2O5N1I_UPv)C{>mmj$VikGbkf8}=Y>AVFlG+Eo}Q3@(ET>I7E#ACxn+eamrG0gmU zdIZylGfh|j;c}oaQ(1CRxzU*TWXZkzDd!O@DeP22TFB+{zWY`c!r&r|p&Cp!pP(!4 zU1t{`a;ufMbf>C9FC`ik9q<4eMsqU2cs;d}DC zp3zAiqg}waQcWn`2TTK6uu?A_zhoHCxrl|e=`F^}#ptCex^uzI3bMB=)96Ee)(3AX z2NWtlGlR{;Jsz%ZK4uCcKMGjVNy_+aR6%Lh5Xh=U_4LP=%OU(9_e1qQRwsvFI&%Jw zbnC^Kw`oJoJ4DVox31C+|J+yLe(a12XSN0k(zX|{$T}aDd0b7*(>UAFcZJ!Ti;co{2FCYIi!Lz^cgXRCb z*Xf(Ut+u`h5HA1ojsNvIUY-DM5aEdF-zwy9kLn-0^{;PM$p9R2mNasq?EgBx93?w; zkMf28%R2tY#i>(m0lYwe;`8o*z1hEeKM@MHO-0@KP+k7-OAG>$`fR5b?Emu+1ae95 zJWn3&|GhZI(0Y(?TXV=dw8a1WL-_IVWC}C9goYiO+kdw*^8eMXDM^r;N0~?p(Z>@< Pz~7Clx3mg1Y##pyej2g| diff --git a/quickstart/IntroNotebooks/images/tf_trt.png b/quickstart/IntroNotebooks/images/tf_trt.png deleted file mode 100644 index e2800821ddf430a0467847288508f2e6edfe7d58..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 364204 zcmeFXWmFv77B1SjJ3&H_Ai;tKcXxMpcXxsY2=2k%-Q7bVxCD0yPUG&phGg$^#(O`` z`+LW@J*um#tCq|)^_z25h@6Zl68u|u004j_E+(V^06>=k0N_us(4aeW+RQQl0HUS=t=Eo=}PwEWVl0mKovd)^EN3p*6HvRz+>%e$LCP9j+a=M~{x%!^w^^ zCP(cMzB)*{ABIFP!6>4!1~G^Q)HZ(oQqL0wz))mCXlC_bQy8V)+?WB3*PdM6J@JW3 z?%%K`RQg;!t1_U5VuAyr-TKu90T-hdNyKPeqpKn&7lgCc@4)u^u(X)u_;w_-NY zo_wm&Uv>!#A1M70+{pn(Z!_t}!1x+*_tk~{-|dQCe=u?_mLrDZi$Y|VBicrIFDgKw zB1f;@JC4j|{CSIf$CjrS`t#W>dqCNMFwLf1h`_h8%qO338$Oc!9S(ZPFmq-G?neGp zrfQ9;?s2(FQUOaeBcTxiRqGJe7QCuGYraQA9?lV{d$-LHkcXX-D0$%Zn}p^F9jbXbz1NC<`|MU%M$aRTL7tI3 zM2Jxt0z(Hr;fo*uhCrb|1Rlpv8(rtj>fEE7l3~{c;}8uY)FoYtixX3F6!YDZm;%~| z4;`#`tqIk+5*CyHafiQrdmp-Hx?Ol5&3%nw6C4%lY~42BBi_lb0rLdbg7i5_8(=sR zEIoZKiw|RwQ~!3$SNGeBFBwBPjzl)xq-*Cpk9)m3zNAEZcm%#2Wz6+iJcJV#>lACbtdPKY3+1yam<`FEK z7n+otYoDYu0&u;Q9n0~1)BP2K|$ckC>w zMP145sQ4^|@;oSu2p+fihvPDDw>Jm~jWzZUPpno$AdCrGh1d!~KsDdtoN$ql}LO6hu zUPEO=_6FPOCbL6Y3IY~laKUh|(d-JMe)%$v25k_k_WtTC^>4qb7(P4^%>5#H$hUCpU*Cz25+C-{Z1VgHY!l$g!I2P*kVBXP!}qI^M}8al zAsb`-ofEu9__v&}ajH73(OocNps8y_2#t0MV2O^lckI(`U`9cuxaV+i9{gb>S+){viH z4}S#z&_NF|l5ik{ivBI;rSQHuye_3K&MDq0c}mrT(Ws!T!vm!U0&^JPXT8NcR6<*FK_Q^ zFH0{`_E*P`ne}kOnZ7bo(5snM^{kb3D+c&D1eCjIc!R6a%8My>iy1u z;R~Y>lTmkc(EG&THbz_vbSvUdl&tgC4kMenltWCzOk+;P=5c%4j7!74R%P=_wazAm z!!5yYkuC|Ta5Q+Fj>lJrM${V;{R9qNb7!UUky=x2SY7cv_N)KcgJ00zu#GG%svj^6Tb3;hRs^_X#P9rgv zDf`Vs70V};O>-C**~i?=&&jZ-dWVD$6b>CG-Z)>nBs5hxTN8@lY;#I+l)20yv8Let zEblL0IefC)KG|I->q!mj#L;IZ{`}(?TA`?y{k6lG>*|5GXWn{01Bs+aL%7%0`tSEA zMORJx;@61RX&dtgw+`_@iislUTpJIDHiI_r6N}5cgToD)Ynzj!v+DDKi{jJfZWe=| zwb(UdH|So1UOP7gXOC+;hpM}cv&XVNtj}D}_)oH;oRL@&D_nQ_zS00*JwrZGB5Ha@ zJ3&3sXlGR3B9>kLD(fnGH_H=@NJjF++PHJ#AKXF&4n&*yvr(L3o*yw0B4BAytPvOy zC6P5TT3wE3$aj|(BgSM?_lcZ1PnCMiz7uINV*vphBh!+85ML%`BRwHGAXY5tp7=JUKH0?iL}t)~Ic zD8QYKbF%YAEH_%U_BF}ost2t=PhsQU!a)DP9q)IyhSO1xVKvzGG*>!pHgNVKP6Lg?uQ^MmpUKvj8FN>yhiVO8eJ2F-S(%R=Vg%(Eqe zC9bncvmcAx8&#UF&RW)-k$^Nas%6hj{59*_eP1IY>XGXcQop8#vT(1Y)!bfSU43gQ zH_@bS@Lu__JX6iBe?P#H+0uf)&51RedpUSlk&eqrcL={cy|gw{WzXv9+DlL++PpI_ zo6M`didWz0Oz5fG<9+|g41)tRh!@kh7r`z4yC8DAZu=_05V@h@rt#g4`_0%QvN~Z2 zkCm6(DcXL_FOo;mu&677ID8;ahU0@iFmGm&yf3}U!{K0{K~cVN^a1^8C}c8iQgjS5 zjooSH_@LYIz)^aqeT-G-q{VJcwNe+Q&b^UY=c0MeYv5Wz@_zSac$FeiS>v20G91ov0kTr8efRv}{QSRjAmZUdVHa$OTFN!)-jl0-$|MWsm)onK4bR*N5Cz1OYNPFM27L|f6 zLwEIz`>+$3Vb7gTw?%iaOXFpI=dqY7Z>!TI9?Pr=(%G?+423%gev(k_^^7JtCnpg`7KI!Ouu7?8T zVF7rC?(gqsuC9Gn5Chtwx~eHt&hsU{o>*1!J9~r8Hqs^`bj0|8o#A=bW3u=D0_7K> z#%kgw($WAb&^0Uo66`Gi3Umbq`tpF`{d+9}Mgf3$y$=om1e*gO|F!|8Ca<69zyJUI zLc|6EU_f`!L0`8l@PArEmt{fxa}ECV(vM$JP+T1Jt7zz8Y;5CbX6y7@_&p=&2ArLk zx+4I9MfUOq6IUQT0ns#TuB7IqCN0HbXlqTYZ)9s=OzUQC_c9KE%Z&qcX>IJJPvmB8 zW#h=<#!dX%f&+B@a+{8r=(UNHB{#8}v>cJ3t%ETUGc6M>JuweF5fKrWgOLe`f{@7H z?x26%#AZ%Tb{uqcuCA`Mu8g#{4yJSr?Ck7x^q=THeWC%ipmB7!ang6Av2i5%Gsr*V z2pKyXI+)uzncLbBy^O1GVC(F}O-%eU(Z8QRbQ-&v|7Ru}$G^)0Es*Zz2^|A1J>9=! zgSv9P+~trncQdw96Ee33@eDKv4+ASb8`o=x|IefUO!=p)vZJwspsh8irxVYAxc;~E ze?RhbQZ%;0|m!f;le6?y{_#!d#Yzz++8pvN_VzzmYrpR(VnOKP$q5y*8vp(_CF7tGCP$2O)CI)E&+h3;?f&6$p1e{1wiUmedqrFK`4=) zDmcB`{2N{>nEy>Es&6;^_5WllsDZCG9Eh!|4@K|(7f)YS0%9t}|KupB!T)9I|DjAx zOXri8MtY+?Sg`c#JYLg5rgmqC?2WsUR)%E0)mRjD$zV}Qfj?C#by3M(?DpKebMV6- zefWlUa(42bZp>rzaNz1-?&nYxxYzgo%txrZnb~(R(_wjcPtrx+e2OC(nVxQLYAR?7 z0WaDYhECBDFY0c*vu9vWwN>+3{x)#Ih1J$k*TyxmhK1GAw$`Q5WNcwgh?!Ob(&CS& z_zLG{wy>yEwtP6Uu&A6N})?0U|9k5Dn|Ti=|jz-H+gT+QO9*M9f=H>!py z$jC|n4l~|ZUSz#`$Vb>p*Dkz9QbA3$X?R{a;O}aXCAtG?P#23WJ zT?w#2os*c{jx;$u5@~xJSy-nx8UQH1$APAPla(X($KC9Z%V{g-Gf1TmEK8f7yC@F! zF30O;96f~cL;PuwNQAn-_|@QhtKxap(NvDNt90SQWX}+pQvf=M_ea9Mo*(;Z*`-a@ zJnyY^(DSRVTwMlmul)hmf1J;|6uvJ$;W~@0{Cbvd{BnDt-Fo@xYM2PGu5rx^@@(kS z-c2B+#HdUb27798S*vRu)k6)u6K*Agp+EvBT6+h1 zgIKbO5E7pjmu2qbn7AN{Z!puhQJBSTX|>E}@VT{B6v zvQsQ_-m`GDiQj|E1ej4fA0Nm%Q zPGu)rA6`x6l7#_g|C>K_67g;FXH1diVK`f{#d5H|D_2KB&?+MbdG^bLXpBEX7l%7> zE=0~N&wIr6Va&YwnXZKJ6||05&%2HG6*)AQw(NYV`4nfwLv4FYfa-0J16LZe4vSUNV;bI zw`>|*Yh13Y4;FCqe{~;fT6s z`TawNI-lq0Grux(Vw#|t7~J13E<|fv>w>2dm4*B?N4dLeY@A|iH_G4jwr*x&(fwc7 zp~gpaD?nV%d5fFBNv0l+ApUQ(HTdo^UNsXX57I&3R~kORRzLXO&X|1aojNYS2o?NKly8FcQ)FK>cRqWEQ=;|TLICd1NZ$N{ZA`+sQ%~`1n^?b_C$nXaQzMONvvZb#dpgZVL!7nD=rT@?yZ` zp+g$Vu;P}2U;KIgIRB950m;SPRxDc!$Ha972Y&+R%nda%4hv1qNc0}=kAiKW7L`?G zjabR=T$1w^N-C2@lHibn{lzWVALPcUeS;b@gdpa?Css0eXv-*-Aiz+^(TggI3$%B< zv!{QWD%$r7Thm|toe}^}g>iC?oL`z8cEmGcOD>B-mtg5w_FKQop$rIS{zsDGsAA1c zso(HD5^6$a---wRbv7Ul)x!!wH58GV(ojYJ0?4^!fGY-DQST#D z*@)5E=~;a()4nQ7K*C+PJjms}*eMdEjEQZ&E=rr&?G`Reynorh?2y2Ky|fl^fgOm zvb3il6>=Y;Y379gb}G3TYqjM;_12UWXB+&x2PcjN`ay&Nz-`2cN_Rz1;O@q_7egI8Adl(7S^${-Yu8=P zuuLdGnT_A3a!o!MJLpS@=K58iAkbs~aC8da@3T3z*=Fv-4V5@F3qVm&xaFWx^t6MU|kmS*(j|@h|?Di}%fRg>^)`d0!v)Hpi^_ ziWU)?+EW*73gy)|WxbbE))x06y!UO$`v{Hpm&pJy@(<&@Cd3go_uwFZ%vp_viVvxk z_yVV_xtODx0cq0AN{20>0*wEtnL9Yau<)AUi}1IJ$0y0^0=oyvXi4H<>Dc)orE4>Ai6-`EZjivT-lJm7#2xLj1p9v`;fDnIZ9mv}Wh@44U%K%a1j~0~RA0gW z`cNW01iz;K(j?C#+|6SQ14#zt7;I?j;3h;=@E6lxO5{_SfYWFKz8|EPbkv0%|0f5~ zisnnwHEe0z@oK10#>1XMrt#0vl|0aue&$?UN7ul`~ITBxyRQs1vP^L*fZ!Qf73nRd-RiYi!2SgQTf82yWqQ^lO z<7fH!6{8ARTccIvpdizNyQ<3~E;c#}cNLUF&S8=M z6F16;*ADkYH(H0YB9TPZ5>{7|^{Q&F-|#)YE+q&Oee7dSXJiL3C`HitA2I5J&o+1_ zLpL=2{0zqDh*|~lqJ_JO5OjLcN&jVpRoGxCXL+fd?b6(-D+Y9EWfdS%luo=S`D5wK zkTaP#;VPL)hSDmV|2RriI|p!*1*z<8YXV0H8X_K8HW=?^C(-uA(N>q=|j~X`p-#I8Ir7*4Iye zT{pTf9$&5->OaxnMjUViwtQA2Xr2IfAoZB=BHC&HWM!&ck<-6cy`}_NVfmb**aY*X zd$8Ds8uk^@j*1e+G?Ux{DMZ{jRucmL;#i_A9zV;)b}0zWrCe{}A~47$->#0jHy7tm zJ5?qLcv_Z4|KOZ&7$;dTej-EKwwlWlf7`da1X(}#70%hQf{X8bo3Mq^bz-g0ru&!k zP>?oz7d)3{9<@;H(V!^N75%DThZ#SMKyi~Ee&;X2ueQ6!pTU$3e78Skm)C>t*FC6D zhXe@Al+oABRrbmub`C3*Y?(>uP<*>K#pFyomL3|ULi(QzD7T? z=uNI&LS-*}nT7E$E&h;g+>g86D`7?MYN}9@$`uqKE*pl#{^9b|3+(pqm~-JwP{#8A z6RKa4R1)dh4!%^>)niDG|3Z>f7$k^fRoKmh{K*~7pYr=y;4xq{$rBcEP&Z#n)FXP0d-RMMJ{kS- z2+<@jE?D=lDOlmmWq^#I?H>jJVEfU(!U2oz<$rJO(WFA7ffgdf%*=eTd^F&+`TgzM zza9kuRtoz=1Tn|cW0XV|MYi;;GzJcunz<3n1dxHDU z{HsIo=qFjNt)E2>bAu_>3ND4a^^CbWH(wz?>P-~Lp^M)+4XKqB(L!MSXY{IM->5=D z?iC+OP)gO%{1FqRL_$QB9|N{+I$|WK&a}j7cK;JJM8UmRy2}pR!->ID=H)|<HY& zmIQ#k(%%e9X;jB|G;MP)zLSiW!{+^4jk2QvkMLid%=7MyZY+!v9Dvg2>(YTNu`JmG z;$Q#TTROZ#2fO@(++XyB!p|gIdoQMgx9YO2?V#lSe+p(@w{Q>67Z4I$=Mm@B;7EUT z7j%9X+0})0Jz0OKW5?D^gbF(s(AB4SZa#gAdsXz-`9C9r`K_0ePoB(?ACyr3*Qm{K zX~QcLy2mY1l=2%(p?n_y@&|@&jUVMgi?5vIp1sHml;ZCP@o99wzq10el~fd$rNYZ2 z{G(vLi)bfAbRKzQ*HdlYtgU;<0SU|Fxv!a-oelU_LN62Ny2c>gn4OI<5%<5H1Ygmb zEYe+=S33m0h72qWxCPo8ZPRUwvU)H1UqJdNl==Dw@iqB4rhKwvzjeEv4}=A+pgU6H z1p>YB#r{Tm!TSnkM-}qz-}WF3#pTB{a-4<7*|vV-Ai#$z7tocm8D+jaAUjR**7#q( z`1*e7%KC=#Yqw!c{5y~s0A1RW70&T0)Lz&;ttDyQqox3JC{-HUSA(SR~y zY_six7^mWrAN804EdQFJ41ZRJDuA2uMi>6ddJXeTKn%6&8=txG`w@e~qu-%nGgyAF z_y^?lL5Cg&81FIa%$1I`-{p5Js-LmL^lIt?2hk|x{fjN-;sx_a-c#}=ZEBQ zy$VDg{Ke_ZXfJK&EQCy-8R2g>eWjghUXPff^+3G6{}PJvgsq{LA)XuZbml}TwC_ru zIXo$0$Mpp{9i7kF)ft&AifxHe&{<6=7gg!U@rQ((4S!H^G$m7r@;11rs!a`PBCTY& z8ebIhJ|Aph@Q1ac_zgYs z4t)F@avv-GexaXu+b{OJ?f)|SSRd6PGbxi$np@HmJGyO?pNp+BQH_AdASZ?#FT<9B zi;u(Rqy}yuqdL*fy)zx+MXxpuXQK4#@JFTBltx|dEOvx;RK64MaTmExTAB$h0f%L6 zn}pe&F*1iFb4E;gO#4b9m-XHS_)empTwUal-_hkl^IKmU(HaqkY?DTI!4Pv(4W=<6 z+X%$Ll+_kjwZ$rNd}i=tDfALH58Sb)8d7KM(c3Ch9sSfycu=3DA$(E&5kn!sHvmJ7 zj!hubsySJF@2)mP2e)Cmpn;_`{$b-Lw<Jha=DpkJ)yV8`$t`A3yk!%lvH3Ks+5q3K)tB- z+5=?Ak_${e{YDt4mAs$4ljF}lkYAHUb7n+9nM@mb$c11*wc`QDe|wSD>hVB?>D9ZA zZ8kK&8fXt&m*I;JO})GjrK={5Cp(fo**wtFP3wagkT9fc5aG|D&%GPd@q3lxng~Fn z9nltEieq!Pg{FTYe{PL_;cFG52|UTK6`w0zQt-}{ZcJu-zdW6@pWgG z(tWH{%ioNot{IO(PaG_%R7o+oi*LZ)Oe(6nVm2_$@K-VkcH8-!`Sq(5@ENFHam>z# zXs`D%Sft!*OA<+xpMfVXOw~VShoKTX5YQQG5#q?IPEGmqBz$k)n?7h2c$xxCv zUKT|H6*e!>ONnMab9WTXK|?50;0pWmEn|p{tjcEECRB8)UdK*JM@Zn6HBGq>3m;v3 z9_d#!D*{E}3M;n;TN{OM^iq=T>ZO4w*@wFqtY1XHY|1>EuF*r`Z3>G7mWtXa_Ry#* zOXTX7TBxPwcLIkZObd=G2rLwWx&+%ha=)Rq$5TobVB(Y+brFGzm7s9A_6f{2S$n#~ zTQgFaAwuErbKdfYaMQs@knwkdjz!*;NTf)b;LTvB- zgj)$fZ{7jZc6Jc zs1I9MEm=SLYhTNo8#ll0EmbJc7(>9u8VjdNL<}qu%c7l$3G9~WMfjTY8Vu$^h04R+ zvBn5T7iOu!RjxX+?48*JasguM;! zo1OV78a=XT?{mnG%-1y!dx8v9gm}uS^4<2xNmSXK{29_i+1VyX_l~>)gWjY{Vq{qy zYHLqPU0J}J=W^awRuKNmvVZ3JD=Db9R!gOu`aoOG-P~VR=bW(Ht#dxlArwG1-K|H) z?+?HQkGp%48u?z)=3^ZD9tRS%!vKPvDd}uM1G{Pg@Kawtm{tu^vrA%!Mk(6gLYf65Ud{5IcWijlVVPJgW_-$gV*Avi^r@{I9hcei0-6VCWMBdkdi$fnUQ}i&xC_OzV7zLczL;n?ZeQNV z1_zTHW4_DcMd6h-Js)EWxRD{xP3xv=yg~e&L@!Yk1>-I^bXPr9Ixn=}ueImyIfGHK z;kT6#$n?1f5d3z!q)!yk1R38BrS&OC@q*b&l(mV%c_0iZB0%okIUD;EAQ0q9?Cowy^SRBh!z&=c8X{XNzT)|gA>lM$yHSzro{I4!Gbq$rv3 z`P^Dqq#Gvwox{Sy!tC<2(QIRn5Eo`hr1dOD!Rzpis)QMHd$wPU3YZH5o*DQKufS6hL(P3t zdm-(&noVvz^a^nq68Ds8;6eYePn2{n(y&_K>$cW@sBbxR>5zlR_|2%Tf>+$BK=rxg z4Gar1uFV&CMe`-88sh!sA zWj-FyBG2qC0Q*i z)0Mk{V5V`{+FMmWcBGIeJK-n(b0f-F*vZ-Z@sXDj5D};Z#I8)yrI!z{g82&ON2RlB z?$FyCZHOFk6ONZclcJwG2r?C{0!jmmV--5{SCQLk!Uc(5)S^h34>#ZqVfLOl%Eetu zv*bPpdTTAt`9fQ<%^Or*9^3 z*iPcs8chQ(;7z4&g>66KvC(W8R zwF{17NxHDy+!TIZZ4adMVRYJg*Rv=_wUYRY0T3D31NnjgSh*}&w)a~Nd8+$#Xt!n4C^PZ{-4(eXS z0xM?FF|ypejk~>wUyo_}+EVJhHa@)e+| zbp($KX2+$|e*o`>^?@EnSf&FA-wkdar|2m}&hIzIW{B2kWvELeimN`xDImno6E6-n zRVZwb_})TOqZAl2Ihsz$oe!U){7eq~NzuOBHn1~7!4)7@pCL^sYimJ(HRR2zQF|@G zl0L%ugU18=?bH!Z;YQ1FRz*amD`8-Eu_P*6O1&S}vG)%$r)nj*d_)xzja6AeL)~bl zvDs>m$JJ34AG1VEwH>hi`tE7B>zOwMM=cfhVv_L26-8d{{_3gRQIA`4-t$m{pC%e6 z2Kyr@(7>dbjBng-6>Y@fH7jorJ%h2#<0v{@IyRJB z(p-n1!6zPkj=nM`g#|W{Mz>B{;w*iw2QFAFeW98+b?MvbO~2E$Xsi`f=k>exTWU|I z@gN}oF_t|VQGq5rj|r+~O5a4|zUE7DT@QyCFMAHNN=Llq6}Hz*nyvft*^!(I9R;gV zQrLad!bG-K*G7aXoWiSJCZdQ;90pqg6AT>RF6qw6pYR#=w$w>in5&}{yfzjs&eH~# zB_yvpH9QHl=%S&2ER~C6?620bXkGi_jedf^n&kbtT5t|k+#(;lkr@0Tw$inV)T<|C zukPvMI@Zn+{KK`T0FwX09$9%san|jkZC>5?bC2~Pau8OL>I|9U+1Ch#wJ^ld_S6BD z!gkK%cFksZ>$opjSpQOFb!2Z<+4-j$s1_j;zwS~ zI$jz4RETGEHm7`S85gRV#3vq*prx*rXB%fzNP>MOmREX=E%7;ZRlGW;hmxZycf4qf z@l;)7zV|03vnw`!`w0sE@WOjaNm^eYQcw-0W?`3a75e+7XcJH%XMe#d2qx1Cw}(eA z0TC0GRvE{tC?+@wE{Cw0( zLqJiTyZ&~{Cg42Oe;;Z*ucLFeMFz9|-mT^PnN*3%G5Ul`Z=0Q0+ZJ@YE3wb4w28H- zEQ{~=ITtoZC>mZWmak^#xDNA)R|t=B4}mNxZmlL_K*}jvi@F0VjTINRZ~q(p9Nac` zB9o8Kwx9SX$uG_>CQK$5O|kLGkBFUs^S1MRU0UsI6CohHS}8eT+9!9_vTuv46W4Ns z#5ncl&`OZ_irp52$M@)%#`}4J7!6USVO&t>HBILSpSN zcu!icE6tXxS)37vT&gY;+Jfshq6!_`sse+yA0@nwab^l>HVjV9Q)1{fSTb4C*E4)b z-O8RVdh}XA==|7rmwL$0H7Tgx!x#4>Ih?kl82l96T6zObA#ZMdAr^Y4>hSz#IJ8EO znU#=eOFIyJ{jo^%sk%(L&g?p_g`cTzE1jw6Qn3tzAP3uKpS19X6twTl3>-cOIvXQg zBIRHQeb9!BX6G2>f&94g7%x@EG?#E}S=`VN*uPA-w05Hp{%prHm3&52N$6+6-`f?@IEi)Pet7mu$V*6CTQ)MDmGeq5A)_p;WmAH6bZ%?F(36= zex^8-(WDD^Jmf=Z(0iZ?onrXi+|V&T`tt6RZq^=NP!)@Cer>5i!h-~E*&P#}AgfCn zS&Uz!bz%7)Del4fgI>ziuJQ(8x9lYtqQ8ieJ^b8K-e9y`5}fxuWhYu|NF zi9;5)O4V}L%2;brLE41rEhfhez(`r4e)&;OwzZBP#O)7gI2z5vjGbjShY#hP&1#H2 zk$B8M>z;l6=ZJ6}Symb~G9P%Xxz2Ovt-Y5WfGwTbnvMBCy@)2x1zq54t!0yCecmO% z53gmi6L`3{${HVuwOSIpnMqmsxKzWl8Rhu_wC5%MlE*6)XaSCEk2qR=pST%jcZr1R zGq7nBW5>cPGLEKJ1voXY2N&<>0xp3wHLg+^!cczysgrrhyKdDi9dW3;5~QWrmfWe;R@~Tl1GWOvnkfQ9NX92HQ}l$*E>naQH$ zbRVjEOFNm9b>r8>wk7UdqSRmvmv(GBGe-6~X{i)g4hpWu6pxc&NKZi##ei{>C;rPh zVC8-4+#{X^&M$QwTN^!zBbGb;F+-DPp?F39a3%nNQtjrrJd+H{a(6pHwZP`Pr%&}xiV>(Si| zl!(wV$Kq`(Zp<^48M5J}^Ku=%=~-KFw29A1wzEEL=EM&-V+{N}rIUyFur_gjm0&jc&Z(3R!3 zb}e~r2?!0DZ40$FYwmL>ysvJd6+rss-)&q0W?f0e{s$AQ^ki|>H=`4u%>^2TzmsN? zBJ114P6pb4gLVly5^Z$Vq!_-MC;)THCe_3tZ=w4FM+3X5BY}0$BI%Nq@RD%L515LS z)^J`nj*rghIRBiF3Qr$2>6BRQZV-bFNT9g@RWn~orqNw#KY+>4HbaOoqe0oWwZPq= zNX9?|pAKiz{Ov;6jVZ*o<{pHp(%Z*SrS8vFV~ad9c_=Q;eY~#RrjcO6GIPDHeUlRh z?caArJ?Jna{TD~H-fH|dh^$;Z`&Cu`Jz2PReu0bl`ErSQ^|9TX+0!Ur4lJF|hVRB{ zF`|ZcpiX=X5w*~PMPYJ0LurgtE>+28jJIaYV64DpQ{T!>w54#CjzznW7lpe=OT1#d zRB;bEzIAEV1J#-n;$-Qs+GquH*R{ z!)e_RaZp5#@Ms==e=ESON_=uVb6U1cX~WnYo8izxVO-^r)J$U0y@ZZyvjjEwowPW+ z2MtL@%Y}Jy?T?^KPLEjH$jjKDy|1;MC?vTc2CBF`FIf?{sx}`Mntqz=QGcdG=IUg& zmNmSDzK8xLDd>t2_b_Zyb1c0|5v~aApr9DbqHEBhW6~(TY_o$(@G^08e>Lk=UHhMl zZ>!1PX5ice=lFwOZ={gHeBE^%l_m;n<^~le`kE8y1@BI^p9)X4ZIa^0kdqn;TJ@{; zclxlJw-Ov@9n}MaDVc4_l~oeubE9bw=rYP--jGr+iZ>h$&bEH2cB{68%9vOJiMiPR zTe-Q4+BcDIEBg&-{zJ++R*8nAsW$D4+Xl|S(^Djv8fNH@Z$Ip>jw_B08+zgj6MK?6_+`wQH1BXm|k_->cSmCjJloVU-m4mIMd_$*?5hn zl)t8d2`JI((ovV|(P=W>acR;wai_we7m7DLj62pbUydz&jH+SQNU5#Y*ld_uN_UEh z_<&?ETP$y;+R&i`=J*~aU zs}ow0y*$?7Lt(9Gq=q1F+tpQ)$By#t;ux!M5P#4$u99eC>5l0eL_yV`qJfov3DUGF zD%b0D&!K6&2sdFVOS6=5^jU2}3w>_D`j%f?UzTD%bBNv?Ha!>}-;(6bc@cS~RHF(6 zxe&3@@<bH_b+xC; zzjRAfSPyKi+8IDQJ~Hd(gyz02LxfVUK_b}QkiEXt>R(w;G`(bvJ!Oejc8f2W^Qg*+ zsI+GG*%4~L;&7E8ia;eGAjtVxufN?C$8x$L~7vJ%Ei?Eu<;mwe#RjsK{Pm)%_?@h!GQ(7AnU<)N< zmI5_|>kyqrCF|)F`NiyM^Ma9wr{fUZFYLOXClYU>14oDHZ&HkrM?sdPFeT_|GSs8x(U~>eju}oIr;~Z zu&Mxqao4Ar>OCTP$~fG>ZL1D-^a>L`B|Qj_wDKNO4c~9Go@|cnj~;8r10GUd_!Sw_ zA{Hj01Gf&WOb$5v`X#4hvzEU*R$bFHRqpjQHAPzcE1akujmbZKIblL@T^U@O#UKAr zx$IeuvN47-PjrC!)bPw8A)%?6Iy2V|s=elXLkl1k?~qRE!lIoeZ2qavJ#I-*AByu) zd<^2|_b+Oc&aRpBd0Gap)6v2HIZCUm#n0c{xa61=#aXLCD$I(+yJGiSaKX+aW<}KZ z3aLdts=)Qv*u8EE(oSC>YsCspn`u+G^+1PsY`qIH1hJXWLG7^ZgPL@``R%>g(Tsf% zF}(r-rrcO13s;d2Y-({;@s*aR-hOUC>+-Ayt;E6B5w;|f$`by_CfGB4ru-9w#g)n9 zJp1xnB>f}G&NAifI^rl6P<3c85N1!{RK2Ait7uyU6;%EUg|`wCoD*A$M6xZCwskw% z`&PQ=#b2Udb@BkJcm|H`;FWSp%;2M85y93yU5^AnH#S9x{Aj1uA`f7^F*-+GNBqQ` zDJ}S>`S=OVT9*~Ef_tYEijZtjH*?_&tp|PErO{i0`!8V@F!fr2T?Lc6pp37~QE{6J z(TZ7(8TP9YFu4L9`S&JW@rwto%5c!P8 zQuX-mI*P|OM@LtghPGlRoc4_yBnFvPW? zP%9ZN3JrFi*Q`cRj3fdVTqjodlRo*GSM3HOaNFcDjFE@z*AR~mC(`=-eMe!S5{O3X z12RyMw(z&)9WceANABltHdDzV(>!pH`T^K4X{EiI|GlN&PlAW8O^PhzuNt@G;XC>; zI}UJ5@h2{UUj5p34pMsAZqdUAC#;Rq#62KMEVZRWS|(*mQrpN~Qc-brz5lu(s!RRR z1eiQ#-aEk_i>BKWqB(6lG?S`9`am&)#6W1ZhTyIWG%4n0nYoc|Zc4xvFOaiQ_qO)v zf#8w+9pYCS=zPrij6uUrlreAb47@i}I|vKvMI&5i@!5fhwd?VPeZFIIK2aGEli<0t zH~DS`S9uf>%;0L8iHDWTzM6i=UCwsZ@q@@LiH$agarxJl;`(RBE)$dCIi|FcLG-8= z5dJa)t}(WH^WKnxkaQ)|rhniZW+6&z6qIdnY18N=6yFPZ}su zoA~XC!G|$ZW( zkz}qWnepjr%vi(ndpK;UqEj@!qm4e$HWQQW&a%nKQcLmU`4p#!g%EE-ZAf#JyI78d zHcx7sgLmc9s7Z^3cJLWJ&o8$(c%N+sY~jatLKhJPS>%f{+rr3Ir*JvgQT))iy>HU% zKhG24on{g5O#EO0fm@ZOUPtdeca|RB#KBVdfm!M#+}&KJqs8wJDHhPGIHk>1Tw(Q; z!PBEMtWv-tXzKdKwU&3FYW~?-xjCoBR;;J1;}{^*wl%$Zlkur->ijHwHs}~_G@8Y= zVLYipcHH+{Ts`shQAG1k$1hJJ*zSyaFW*+pb}TwsgqSd8n(Mp}ie zMV5+tqBrTLo!Q30yr`Q^5AyD3zRus-jutj^e*nWHcNvH0KIXAY8`zKvr;^Re#IMp1 zry$33uFeKMl(9GV>1`b!N=En_n(>8+zZFcVb|$@n-GXW*s1*S*iD%<F_^a7EBYAyFt{Z^Q_;H{&GZX5Jo|S%LRv_ocGOEoJ zf6p&L-%hz>S^t0d`pU4Xx~^S8O6l(I2I)pRL_nmwyBnlK>68uy0f9|-H%Lo&!=}5X z&f54q-}|2LI@kHdg_}Lsnxn@(?=fvhvTcaADzE%OY;H@aGbkG6&dR!wR-3Cv)8Kx6 ze>gL?yJ%wjCikxYAxX)&*0*@=%5hzFC_yIFuzWKE73A$`5l@xjUFu=q;^9z0dBH;k zosQpbZqvs9DVwdPA|R#{$qZ%y+3g)GZ~Gs!b@$pupVPl=4xZfULD~9W z?2=f7f7G`2l3m<9b&=zG)HkvI5OTvcUYtg_q^CWltih@{TjZP3yOg9y69+lfvE*I`RnGxvw?D5o2YR{>shP88{4N18{3~;^D2^8xkjY< z#abei^n2iy>^RMj>ZwOrN-IA}W-Ry_R`Y|EuuKZWs+-=%K-YS)c$U$F?^{r)| z=Nw$c{0VWF%}TomQXIRc>e{^?LCUS_Kf#@ub>pWQFFSejUfxC`2&XAi=vnLckLZ(3edRQgRk)o(nU(mq|sFOI_>C3?8W9G`Lb@U$vun;b4<1WhG8_~nL@XIh*H>cIgIqAU^f=*55cp#CVb_vrLR*W}xGu>GRI5 z@{3f*?MIyXxg3ER$Ikm^ag7D^ZSH#b)0SD1Mozi+$=TF*`I_wzz&OU_3wxhh zguowkUJ5hH`Ei(f>`z>1HC9YW|4E(2&!(>4Y$uFA5he8l=NK1s3ngbna2S(5HtfH( zt_xt!{s!>r;t$G7tyoSwu4S}k{8R&HJtxzvDS<(|n+vgmd;0^`rble`` z=7#vulia>J=W{RKIq-(T$EF9>`8DVGGVV zL%d+&pwgWEhaDThR0Aniir1uvjl3N=itp0WlOpGR=&Fy;K+B{zKTs>>z*mO^P0)+ z%=uc+S`F(gv_jx1^Yb`IcVkDRc2+0{01uM~$IW{)@3bwK<)sq^+NEY3Js+qqtzO=M%5gv2 z#nmtV{K=5#K67alVWV!5V0HDt`Y801-WzMn$7@m?MIW>Csx@qwRUZ5 ze);L&>#qH-CNZWZIg(A;)mlV+M~$#am6A&U$t;aK`(cD>#-`Epwjz(1ZgD0QhFK)J zNuP;z*D1RCzU2b(&K7gHGLf-LR)&M1UpTRwJ68YmdzLc0n`YaYN-l}_dvLE}rfv|v zO&+H&uJ1d+F_s$W3Z{$2A|#^Kmp)ieib(DKk&Z`$B2)SdvuL`c;T$`F4?;07{T4nz zac26G+9Q2wVImnh!(w=zIxgPmHnOJ7O~?Lzn(C49zEjZ6J-sfsBZ#1&$x^C0=npghN z%$ig~8|5kqQ>h3N-nS0op66zCqT3bk^2Z-9zOob~XNXbw9Sa_YVv>%x(YD49=3CtO zgLx)WP{vW(cQYi?+RmYh_b?g;H(*Olb2p}6OZCxD-6W?67`!*v9AOUrg^@z%IlLn|$8^i67QV>(-U29(kEoV^C^$xA~LXXsmm4($|Dx44Sm zY$Dl)Rks&9T7Fgmk{qQ>hoWn(+*J0r4T}#(va+A@!EvCk)f}$u)Z0X6nmyK?_(}L+ zTruPM7MlB5gOuqEXLl}*sh%oJ6u%$j0Dlx|i9`gEPPF|y!oAxnwH$xj18EI_4TABH81>1j1*RL<#B&?tS~u%W=Q~2=x)- zQXuJW)fc0^>nY@Mq9@NC{P`pxs*b#TZtIeOwr=^kt>cPra<%hdR@h;<%rP$P1}M~W zNW0IhIoMa9%6$>EM`@Q}xiE__B%bsOm1WFjHvtam#V|{uqOGhLEHqpV0N>PkYkwGz z8~S>2m{kZlWrvV(|I~U&#JijqDq3uvHpDyR5j@qpBoV?$j86pd$eReHH)XcY3RiG9 zYg!;F>|NU*He4BOt>Mlcd^ep}X?geyr+%0T^a{h`?`INx=gb|kWHo?K=Jw;0LK|tS zT`mlXi|f9(rA+$v&>8md($c;20oQzH?J~%V`B`&_#bXjq*NCGoY#XSrFT%(~xL;5A zO~f!156*&Gp@;QlZzbhe`DH9&;b(_>IdJf4Jm)%E11>C%^FE4{!-}YyII9OunR@IA zr-$p1@z7CC)}gnsxdahfyvEK9YX?qY>6$hpIl@3`pd=v5qVc$U#CdVGDbmF(zF<~1 z!UIj-y|A}jP`>^{H^x(F{p?dNd|F=jN@Zm~O}oitYmuh&?>z=f;vKAC)_-3cM!%yz zTJrSlg^Olz)ZZES~0s&6^xvY=;~kEq`ue7OvY{JrP|-dn&~b#R10*v+!QNZ%ZgY%WY=W6pGw^ zOUQkN>+_Ayywpf|_P63rXpA|*7#NCR9_%UGAh-$~7aP{s1OQl=an(5(v1$;XERkJ9 zx3tnwNgU{#-_%4AP$zb<{hpkEJ%5qp>(cI1LT&z_cjVj8EG$k+lj82fTq4xo7eli5 zKq(17Sa#VZT}CRSFU3b)FsJ_-o5MYUycK=&%hvALCW{r;yyS-xZ$c2uhSL6AV!M<- zQvmK~Dy1*_i8g{1qTg6vVd)R#x=&vg4l)oovH(El<8tf8#kU%)%>~gM+-!wM*>$!? zrFiBr4kX}C7dXZ)Z?Wq|jAVb;xfR~DOP&&5ZfW(8lF?;R>A-ZG5IKFsbzy&o zK0Abm4RnYcWoSVkoiZ64dh0$!`DvSJIAU*pYO@#QOHiX8r35fnFV7wBvrCle36F(T z*4@$Ev(l_hscd}hN%b=5koTKNXmSLg>YTGw3-ubW_<89l^v0pF{Ea(gTC-mSbJe>{ zTpDYw)A3clz)9y%%ksvuzJkhq)vd2t1vUN0m71Q`A*DPG`Np%KqoD#(iNTLqf6Edc z>*;a~U7x&c1!ZCcOw9VaK6R{__7*ts7g@4NbcBq7x%c{q&2NGh(M>7~kdg>M^&nkB zUCE90?N>dm2U_XZ){&^fK@jn;Az$c!)5(^r4uzzi$X#Owmew^gf)&*RoR!(e_xS@{ zk)&A8j8S&1Zg)IB;;aLVjBGRuK7v7!x9{x%96i{Eth z>_Y4YEo^;9lgGO=;`XmL5|Ix6VSviU_38}DVi3aoYT!?_{sfd5byEgg9r+g`1*e(I z?-XG_2eh~%uphrLz)2^EHPa2@ar+mnb9XNe z2V;F>)rwB-i#vdOCdQqhwe79ag8LxtVeGV`O2i42mt!9*$c*D9?S2S;7NCLM{E{P+ zQv#9Dt#oe|bAP1?gS@ZI+|86x#*{QK_d~3MF0Q;sA7n&GGxBlsy^RWi61bshd8UzR zSSYDUfAWon?u^^(OMCjvsZ@QhXyGL6tzO^8OeykZi%c?}Sfi(7geQvPhsmvZo2aq)knp&-j>z*!Jo&T7Tb z^+?~y4JJRJ^eCiyPe{h80bCyBKC@fLQ?t8uWlP!g8zAY%Nu)cqfPkVsiMxIg-a~u2 zg(x`BR$A9X&Y^Ih3$!`ks>7!Sabs?A@ z6{sDLuLLP0Hv%Btd8bdyr=2Kt_?iN*yAT2U?;CvUkV=llAE@LOg+4z{e;o{SyTZn~ z;thXhG-n&COaT{Y1*4`u*9^C>K4o~kG%c@@U9?TOlNscr%45l2K`9jY!`036MUNt`>$5_3*(y8dzc!d*?M@Dx@T+T%zmWK zb7*A>YEAmp)U`dYs1MeATj0ZTueQ9LVaMZnNVJ|#>z2=1q zyM=%8=ykjM=eZ+W@Vy3lZYzAyrZA3^@hN~Cn9=`8>KI7z##0HTzr4#&>IfgSAFQ&J z^Xt|p{|#XKLVUjgjp>Y$NJmSpRx#I&H$b2JY7;BCac#zV!>7K{aMTQ~Nt03UYoWlh z)}J2*;5F+`UJ|-D8UU$oZ2!qEXD`ia?o%jwsYQ9^*3#gu#=P64X;F7mQz^1q^!Uo} zEsNn(hbuwkWX~*&_apgj33*v>nF`H6B|j;D+HmY-S~+fwKrs!wvk?r!xDKgZPrBUL z`7S&NQ4K(1GPSM_M_vLt1nul+^&tJx_DWW`f;qI&sd6Q{uR#ccX#Q+e_e0WjDgOoQ2JbwKA zOnwJgNw^*YxQVr)EgPzps(1%3Xdnr>nwGEgyZ*y0`FI>PMO>I8cW^cgU?OENu4~Y$ z<^lkuodAxpRd^AAd}U3F0i=RwraQgmSFILY@n~K2kI8$iuuotZ0Q{i2Uih+)yK+7%#~{y-JfVq4*JW$#TuwXOSR#^;dIxaa^N-=Hlh4kVFNvzerw+vZ9iD zIdA%|XGMXQfTAIZ=|Dr;-Yh$+Q<6>SHdXH02z4^jj44Hl|v znYnjLz(Mdt9WW8ckQ$e^Au&W6aBsn zSRW|c3#iuwMGJe|00MNY2E*}C%DftpZ$NFv!_vn$3ip?XS^2)qB-r^6EW7XdKN%*o zz3&(0dnuty0J@RRe2U^{`e^;WlWc%#1fy4TBB0ZUS;7(^NIc|(cF5TUn@M%N{o^sn zY&K|9+_Co$>7DvAmB5j){~IQx)18N7w9Zx_R;8chNy&ju=eUl-ovX^QyvU8Q6L#nK zwX~psz2kPE{rbK5CFE+ezG=(d4gdH#MJhm_I(urD3*$jP-5hkc-p*HG_D)wk`xw9E zM$Y)w?bc;S%FJeuNgFcM>qn{e+B9=FfSXKktXJB^WhEhHpizMlddr(PmKgRpv?`<)DG?5c{^EN~y7mvZ#X?(9l4hkIaNDyH-UeH0LwWQ|8YVGD@TE8sRbV*%BP2@+vV0QSoOs8fGH7ky9y`y z^C8BQwvUH|F1Z4YKJt25!qvBvwAE)fI)=W+nFs$ch7r#3;ZW?@vq~+mlMKtj_f0