{"cells": [{"cell_type": "markdown", "id": "161eabba", "metadata": {"papermill": {"duration": 0.030074, "end_time": "2021-09-16T12:36:05.411850", "exception": false, "start_time": "2021-09-16T12:36:05.381776", "status": "completed"}, "tags": []}, "source": ["\n", "# Tutorial 3: Initialization and Optimization\n", "\n", "* **Author:** Phillip Lippe\n", "* **License:** CC BY-SA\n", "* **Generated:** 2021-09-16T14:32:21.097031\n", "\n", "In this tutorial, we will review techniques for optimization and initialization of neural networks.\n", "When increasing the depth of neural networks, there are various challenges we face.\n", "Most importantly, we need to have a stable gradient flow through the network, as otherwise, we might encounter vanishing or exploding gradients.\n", "This is why we will take a closer look at the following concepts: initialization and optimization.\n", "This notebook is part of a lecture series on Deep Learning at the University of Amsterdam.\n", "The full list of tutorials can be found at https://uvadlc-notebooks.rtfd.io.\n", "\n", "\n", "---\n", "Open in [{height=\"20px\" width=\"117px\"}](https://colab.research.google.com/github/PytorchLightning/lightning-tutorials/blob/publication/.notebooks/course_UvA-DL/03-initialization-and-optimization.ipynb)\n", "\n", "Give us a \u2b50 [on Github](https://www.github.com/PytorchLightning/pytorch-lightning/)\n", "| Check out [the documentation](https://pytorch-lightning.readthedocs.io/en/latest/)\n", "| Join us [on Slack](https://join.slack.com/t/pytorch-lightning/shared_invite/zt-pw5v393p-qRaDgEk24~EjiZNBpSQFgQ)"]}, {"cell_type": "markdown", "id": "513d8cd0", "metadata": {"papermill": {"duration": 0.029401, "end_time": "2021-09-16T12:36:05.470171", "exception": false, "start_time": "2021-09-16T12:36:05.440770", "status": "completed"}, "tags": []}, "source": ["## Setup\n", "This notebook requires some packages besides pytorch-lightning."]}, {"cell_type": "code", "execution_count": 1, "id": "780d4b1e", "metadata": {"colab": {}, "colab_type": "code", "execution": {"iopub.execute_input": "2021-09-16T12:36:05.530490Z", "iopub.status.busy": "2021-09-16T12:36:05.530024Z", "iopub.status.idle": "2021-09-16T12:36:05.532156Z", "shell.execute_reply": "2021-09-16T12:36:05.532530Z"}, "id": "LfrJLKPFyhsK", "lines_to_next_cell": 0, "papermill": {"duration": 0.034212, "end_time": "2021-09-16T12:36:05.532707", "exception": false, "start_time": "2021-09-16T12:36:05.498495", "status": "completed"}, "tags": []}, "outputs": [], "source": ["# ! pip install --quiet \"seaborn\" \"torchvision\" \"torchmetrics>=0.3\" \"torch>=1.6, <1.9\" \"pytorch-lightning>=1.3\" \"matplotlib\""]}, {"cell_type": "markdown", "id": "0d6d5d66", "metadata": {"papermill": {"duration": 0.028732, "end_time": "2021-09-16T12:36:05.590890", "exception": false, "start_time": "2021-09-16T12:36:05.562158", "status": "completed"}, "tags": []}, "source": ["
\n", "In the first half of the notebook, we will review different initialization techniques, and go step by step from the simplest initialization to methods that are nowadays used in very deep networks.\n", "In the second half, we focus on optimization comparing the optimizers SGD, SGD with Momentum, and Adam.\n", "\n", "Let's start with importing our standard libraries:"]}, {"cell_type": "code", "execution_count": 2, "id": "622ad7af", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:05.655225Z", "iopub.status.busy": "2021-09-16T12:36:05.654757Z", "iopub.status.idle": "2021-09-16T12:36:07.304855Z", "shell.execute_reply": "2021-09-16T12:36:07.304422Z"}, "papermill": {"duration": 1.685542, "end_time": "2021-09-16T12:36:07.304972", "exception": false, "start_time": "2021-09-16T12:36:05.619430", "status": "completed"}, "tags": []}, "outputs": [{"name": "stderr", "output_type": "stream", "text": ["/tmp/ipykernel_879/869332958.py:24: DeprecationWarning: `set_matplotlib_formats` is deprecated since IPython 7.23, directly use `matplotlib_inline.backend_inline.set_matplotlib_formats()`\n", " set_matplotlib_formats(\"svg\", \"pdf\") # For export\n"]}], "source": ["import copy\n", "import json\n", "import math\n", "import os\n", "import urllib.request\n", "from urllib.error import HTTPError\n", "\n", "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import pytorch_lightning as pl\n", "import seaborn as sns\n", "import torch\n", "import torch.nn as nn\n", "import torch.nn.functional as F\n", "import torch.utils.data as data\n", "\n", "# %matplotlib inline\n", "from IPython.display import set_matplotlib_formats\n", "from matplotlib import cm\n", "from torchvision import transforms\n", "from torchvision.datasets import FashionMNIST\n", "from tqdm.notebook import tqdm\n", "\n", "set_matplotlib_formats(\"svg\", \"pdf\") # For export\n", "sns.set()"]}, {"cell_type": "markdown", "id": "d90abb51", "metadata": {"papermill": {"duration": 0.029212, "end_time": "2021-09-16T12:36:07.364402", "exception": false, "start_time": "2021-09-16T12:36:07.335190", "status": "completed"}, "tags": []}, "source": ["Instead of the `set_seed` function as in Tutorial 3, we can use PyTorch Lightning's build-in function `pl.seed_everything`.\n", "We will reuse the path variables `DATASET_PATH` and `CHECKPOINT_PATH` as in Tutorial 3.\n", "Adjust the paths if necessary."]}, {"cell_type": "code", "execution_count": 3, "id": "ea2ba888", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:07.429294Z", "iopub.status.busy": "2021-09-16T12:36:07.428723Z", "iopub.status.idle": "2021-09-16T12:36:07.498990Z", "shell.execute_reply": "2021-09-16T12:36:07.499373Z"}, "papermill": {"duration": 0.10601, "end_time": "2021-09-16T12:36:07.499520", "exception": false, "start_time": "2021-09-16T12:36:07.393510", "status": "completed"}, "tags": []}, "outputs": [{"name": "stderr", "output_type": "stream", "text": ["Global seed set to 42\n"]}, {"name": "stdout", "output_type": "stream", "text": ["Using device cuda:0\n"]}], "source": ["# Path to the folder where the datasets are/should be downloaded (e.g. MNIST)\n", "DATASET_PATH = os.environ.get(\"PATH_DATASETS\", \"data/\")\n", "# Path to the folder where the pretrained models are saved\n", "CHECKPOINT_PATH = os.environ.get(\"PATH_CHECKPOINT\", \"saved_models/InitOptim/\")\n", "\n", "# Seed everything\n", "pl.seed_everything(42)\n", "\n", "# Ensure that all operations are deterministic on GPU (if used) for reproducibility\n", "torch.backends.cudnn.determinstic = True\n", "torch.backends.cudnn.benchmark = False\n", "\n", "# Fetching the device that will be used throughout this notebook\n", "device = torch.device(\"cpu\") if not torch.cuda.is_available() else torch.device(\"cuda:0\")\n", "print(\"Using device\", device)"]}, {"cell_type": "markdown", "id": "7aaf0232", "metadata": {"papermill": {"duration": 0.029659, "end_time": "2021-09-16T12:36:07.559273", "exception": false, "start_time": "2021-09-16T12:36:07.529614", "status": "completed"}, "tags": []}, "source": ["In the last part of the notebook, we will train models using three different optimizers.\n", "The pretrained models for those are downloaded below."]}, {"cell_type": "code", "execution_count": 4, "id": "139dec18", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:07.623803Z", "iopub.status.busy": "2021-09-16T12:36:07.623326Z", "iopub.status.idle": "2021-09-16T12:36:08.686564Z", "shell.execute_reply": "2021-09-16T12:36:08.686141Z"}, "papermill": {"duration": 1.09754, "end_time": "2021-09-16T12:36:08.686681", "exception": false, "start_time": "2021-09-16T12:36:07.589141", "status": "completed"}, "tags": []}, "outputs": [{"name": "stdout", "output_type": "stream", "text": ["Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_SGD.config...\n"]}, {"name": "stdout", "output_type": "stream", "text": ["Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_SGD_results.json...\n", "Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_SGD.tar...\n"]}, {"name": "stdout", "output_type": "stream", "text": ["Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_SGDMom.config...\n", "Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_SGDMom_results.json...\n", "Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_SGDMom.tar...\n"]}, {"name": "stdout", "output_type": "stream", "text": ["Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_Adam.config...\n", "Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_Adam_results.json...\n", "Downloading https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/FashionMNIST_Adam.tar...\n"]}], "source": ["# Github URL where saved models are stored for this tutorial\n", "base_url = \"https://raw.githubusercontent.com/phlippe/saved_models/main/tutorial4/\"\n", "# Files to download\n", "pretrained_files = [\n", " \"FashionMNIST_SGD.config\",\n", " \"FashionMNIST_SGD_results.json\",\n", " \"FashionMNIST_SGD.tar\",\n", " \"FashionMNIST_SGDMom.config\",\n", " \"FashionMNIST_SGDMom_results.json\",\n", " \"FashionMNIST_SGDMom.tar\",\n", " \"FashionMNIST_Adam.config\",\n", " \"FashionMNIST_Adam_results.json\",\n", " \"FashionMNIST_Adam.tar\",\n", "]\n", "# Create checkpoint path if it doesn't exist yet\n", "os.makedirs(CHECKPOINT_PATH, exist_ok=True)\n", "\n", "# For each file, check whether it already exists. If not, try downloading it.\n", "for file_name in pretrained_files:\n", " file_path = os.path.join(CHECKPOINT_PATH, file_name)\n", " if not os.path.isfile(file_path):\n", " file_url = base_url + file_name\n", " print(f\"Downloading {file_url}...\")\n", " try:\n", " urllib.request.urlretrieve(file_url, file_path)\n", " except HTTPError as e:\n", " print(\n", " \"Something went wrong. Please try to download the file from the GDrive folder, or contact the author with the full output including the following error:\\n\",\n", " e,\n", " )"]}, {"cell_type": "markdown", "id": "bef9b45d", "metadata": {"papermill": {"duration": 0.030457, "end_time": "2021-09-16T12:36:08.748288", "exception": false, "start_time": "2021-09-16T12:36:08.717831", "status": "completed"}, "tags": []}, "source": ["## Preparation"]}, {"cell_type": "markdown", "id": "7cbe8831", "metadata": {"papermill": {"duration": 0.030589, "end_time": "2021-09-16T12:36:08.809491", "exception": false, "start_time": "2021-09-16T12:36:08.778902", "status": "completed"}, "tags": []}, "source": ["Throughout this notebook, we will use a deep fully connected network, similar to our previous tutorial.\n", "We will also again apply the network to FashionMNIST, so you can relate to the results of Tutorial 3.\n", "We start by loading the FashionMNIST dataset:"]}, {"cell_type": "code", "execution_count": 5, "id": "b07918f9", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:08.874871Z", "iopub.status.busy": "2021-09-16T12:36:08.874401Z", "iopub.status.idle": "2021-09-16T12:36:08.910586Z", "shell.execute_reply": "2021-09-16T12:36:08.910149Z"}, "papermill": {"duration": 0.070623, "end_time": "2021-09-16T12:36:08.910707", "exception": false, "start_time": "2021-09-16T12:36:08.840084", "status": "completed"}, "tags": []}, "outputs": [], "source": ["\n", "# Transformations applied on each image => first make them a tensor, then normalize them with mean 0 and std 1\n", "transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.2861,), (0.3530,))])\n", "\n", "# Loading the training dataset. We need to split it into a training and validation part\n", "train_dataset = FashionMNIST(root=DATASET_PATH, train=True, transform=transform, download=True)\n", "train_set, val_set = torch.utils.data.random_split(train_dataset, [50000, 10000])\n", "\n", "# Loading the test set\n", "test_set = FashionMNIST(root=DATASET_PATH, train=False, transform=transform, download=True)"]}, {"cell_type": "markdown", "id": "e1f85744", "metadata": {"papermill": {"duration": 0.0303, "end_time": "2021-09-16T12:36:08.971816", "exception": false, "start_time": "2021-09-16T12:36:08.941516", "status": "completed"}, "tags": []}, "source": ["We define a set of data loaders that we can use for various purposes later.\n", "Note that for actually training a model, we will use different data loaders\n", "with a lower batch size."]}, {"cell_type": "code", "execution_count": 6, "id": "1e446c8f", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:09.037306Z", "iopub.status.busy": "2021-09-16T12:36:09.036842Z", "iopub.status.idle": "2021-09-16T12:36:09.038587Z", "shell.execute_reply": "2021-09-16T12:36:09.038964Z"}, "papermill": {"duration": 0.036552, "end_time": "2021-09-16T12:36:09.039078", "exception": false, "start_time": "2021-09-16T12:36:09.002526", "status": "completed"}, "tags": []}, "outputs": [], "source": ["train_loader = data.DataLoader(train_set, batch_size=1024, shuffle=True, drop_last=False)\n", "val_loader = data.DataLoader(val_set, batch_size=1024, shuffle=False, drop_last=False)\n", "test_loader = data.DataLoader(test_set, batch_size=1024, shuffle=False, drop_last=False)"]}, {"cell_type": "markdown", "id": "2fa7ffc1", "metadata": {"papermill": {"duration": 0.03043, "end_time": "2021-09-16T12:36:09.100101", "exception": false, "start_time": "2021-09-16T12:36:09.069671", "status": "completed"}, "tags": []}, "source": ["In comparison to the previous tutorial, we have changed the parameters of the normalization transformation `transforms.Normalize`.\n", "The normalization is now designed to give us an expected mean of 0 and a standard deviation of 1 across pixels.\n", "This will be particularly relevant for the discussion about initialization we will look at below, and hence we change it here.\n", "It should be noted that in most classification tasks, both normalization techniques (between -1 and 1 or mean 0 and stddev 1) have shown to work well.\n", "We can calculate the normalization parameters by determining the mean and standard deviation on the original images:"]}, {"cell_type": "code", "execution_count": 7, "id": "f94c10d3", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:09.164949Z", "iopub.status.busy": "2021-09-16T12:36:09.164391Z", "iopub.status.idle": "2021-09-16T12:36:09.265798Z", "shell.execute_reply": "2021-09-16T12:36:09.265361Z"}, "papermill": {"duration": 0.135044, "end_time": "2021-09-16T12:36:09.265917", "exception": false, "start_time": "2021-09-16T12:36:09.130873", "status": "completed"}, "tags": []}, "outputs": [{"name": "stdout", "output_type": "stream", "text": ["Mean 0.28604060411453247\n", "Std 0.3530242443084717\n"]}], "source": ["print(\"Mean\", (train_dataset.data.float() / 255.0).mean().item())\n", "print(\"Std\", (train_dataset.data.float() / 255.0).std().item())"]}, {"cell_type": "markdown", "id": "53a9eb37", "metadata": {"papermill": {"duration": 0.030511, "end_time": "2021-09-16T12:36:09.328055", "exception": false, "start_time": "2021-09-16T12:36:09.297544", "status": "completed"}, "tags": []}, "source": ["We can verify the transformation by looking at the statistics of a single batch:"]}, {"cell_type": "code", "execution_count": 8, "id": "7167dac5", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:09.392969Z", "iopub.status.busy": "2021-09-16T12:36:09.392508Z", "iopub.status.idle": "2021-09-16T12:36:09.576984Z", "shell.execute_reply": "2021-09-16T12:36:09.576574Z"}, "papermill": {"duration": 0.218349, "end_time": "2021-09-16T12:36:09.577097", "exception": false, "start_time": "2021-09-16T12:36:09.358748", "status": "completed"}, "tags": []}, "outputs": [{"name": "stdout", "output_type": "stream", "text": ["Mean: 0.009\n", "Standard deviation: 1.012\n", "Maximum: 2.022\n", "Minimum: -0.810\n"]}], "source": ["imgs, _ = next(iter(train_loader))\n", "print(f\"Mean: {imgs.mean().item():5.3f}\")\n", "print(f\"Standard deviation: {imgs.std().item():5.3f}\")\n", "print(f\"Maximum: {imgs.max().item():5.3f}\")\n", "print(f\"Minimum: {imgs.min().item():5.3f}\")"]}, {"cell_type": "markdown", "id": "90f0ee3d", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.030932, "end_time": "2021-09-16T12:36:09.639639", "exception": false, "start_time": "2021-09-16T12:36:09.608707", "status": "completed"}, "tags": []}, "source": ["Note that the maximum and minimum are not 1 and -1 anymore, but shifted towards the positive values.\n", "This is because FashionMNIST contains a lot of black pixels, similar to MNIST.\n", "\n", "Next, we create a linear neural network. We use the same setup as in the previous tutorial."]}, {"cell_type": "code", "execution_count": 9, "id": "878c7079", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:09.708008Z", "iopub.status.busy": "2021-09-16T12:36:09.707536Z", "iopub.status.idle": "2021-09-16T12:36:09.709627Z", "shell.execute_reply": "2021-09-16T12:36:09.709167Z"}, "lines_to_next_cell": 2, "papermill": {"duration": 0.039177, "end_time": "2021-09-16T12:36:09.709744", "exception": false, "start_time": "2021-09-16T12:36:09.670567", "status": "completed"}, "tags": []}, "outputs": [], "source": ["class BaseNetwork(nn.Module):\n", " def __init__(self, act_fn, input_size=784, num_classes=10, hidden_sizes=[512, 256, 256, 128]):\n", " \"\"\"\n", " Args:\n", " act_fn: Object of the activation function that should be used as non-linearity in the network.\n", " input_size: Size of the input images in pixels\n", " num_classes: Number of classes we want to predict\n", " hidden_sizes: A list of integers specifying the hidden layer sizes in the NN\n", " \"\"\"\n", " super().__init__()\n", "\n", " # Create the network based on the specified hidden sizes\n", " layers = []\n", " layer_sizes = [input_size] + hidden_sizes\n", " for layer_index in range(1, len(layer_sizes)):\n", " layers += [nn.Linear(layer_sizes[layer_index - 1], layer_sizes[layer_index]), act_fn]\n", " layers += [nn.Linear(layer_sizes[-1], num_classes)]\n", " # A module list registers a list of modules as submodules (e.g. for parameters)\n", " self.layers = nn.ModuleList(layers)\n", "\n", " self.config = {\n", " \"act_fn\": act_fn.__class__.__name__,\n", " \"input_size\": input_size,\n", " \"num_classes\": num_classes,\n", " \"hidden_sizes\": hidden_sizes,\n", " }\n", "\n", " def forward(self, x):\n", " x = x.view(x.size(0), -1)\n", " for layer in self.layers:\n", " x = layer(x)\n", " return x"]}, {"cell_type": "markdown", "id": "f411d171", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.031158, "end_time": "2021-09-16T12:36:09.773277", "exception": false, "start_time": "2021-09-16T12:36:09.742119", "status": "completed"}, "tags": []}, "source": ["For the activation functions, we make use of PyTorch's `torch.nn` library instead of implementing ourselves.\n", "However, we also define an `Identity` activation function.\n", "Although this activation function would significantly limit the\n", "network's modeling capabilities, we will use it in the first steps of\n", "our discussion about initialization (for simplicity)."]}, {"cell_type": "code", "execution_count": 10, "id": "17e393bf", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:09.838964Z", "iopub.status.busy": "2021-09-16T12:36:09.838503Z", "iopub.status.idle": "2021-09-16T12:36:09.840568Z", "shell.execute_reply": "2021-09-16T12:36:09.840106Z"}, "papermill": {"duration": 0.036232, "end_time": "2021-09-16T12:36:09.840668", "exception": false, "start_time": "2021-09-16T12:36:09.804436", "status": "completed"}, "tags": []}, "outputs": [], "source": ["class Identity(nn.Module):\n", " def forward(self, x):\n", " return x\n", "\n", "\n", "act_fn_by_name = {\"tanh\": nn.Tanh, \"relu\": nn.ReLU, \"identity\": Identity}"]}, {"cell_type": "markdown", "id": "00535e64", "metadata": {"papermill": {"duration": 0.031324, "end_time": "2021-09-16T12:36:09.903283", "exception": false, "start_time": "2021-09-16T12:36:09.871959", "status": "completed"}, "tags": []}, "source": ["Finally, we define a few plotting functions that we will use for our discussions.\n", "These functions help us to (1) visualize the weight/parameter distribution inside a network, (2) visualize the gradients that the parameters at different layers receive, and (3) the activations, i.e. the output of the linear layers.\n", "The detailed code is not important, but feel free to take a closer look if interested."]}, {"cell_type": "code", "execution_count": 11, "id": "cb3680e1", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:09.980697Z", "iopub.status.busy": "2021-09-16T12:36:09.978256Z", "iopub.status.idle": "2021-09-16T12:36:09.982397Z", "shell.execute_reply": "2021-09-16T12:36:09.982770Z"}, "papermill": {"duration": 0.048335, "end_time": "2021-09-16T12:36:09.982887", "exception": false, "start_time": "2021-09-16T12:36:09.934552", "status": "completed"}, "tags": []}, "outputs": [], "source": ["##############################################################\n", "\n", "\n", "def plot_dists(val_dict, color=\"C0\", xlabel=None, stat=\"count\", use_kde=True):\n", " columns = len(val_dict)\n", " fig, ax = plt.subplots(1, columns, figsize=(columns * 3, 2.5))\n", " fig_index = 0\n", " for key in sorted(val_dict.keys()):\n", " key_ax = ax[fig_index % columns]\n", " sns.histplot(\n", " val_dict[key],\n", " ax=key_ax,\n", " color=color,\n", " bins=50,\n", " stat=stat,\n", " kde=use_kde and ((val_dict[key].max() - val_dict[key].min()) > 1e-8),\n", " ) # Only plot kde if there is variance\n", " hidden_dim_str = (\n", " r\"(%i $\\to$ %i)\" % (val_dict[key].shape[1], val_dict[key].shape[0]) if len(val_dict[key].shape) > 1 else \"\"\n", " )\n", " key_ax.set_title(f\"{key} {hidden_dim_str}\")\n", " if xlabel is not None:\n", " key_ax.set_xlabel(xlabel)\n", " fig_index += 1\n", " fig.subplots_adjust(wspace=0.4)\n", " return fig\n", "\n", "\n", "##############################################################\n", "\n", "\n", "def visualize_weight_distribution(model, color=\"C0\"):\n", " weights = {}\n", " for name, param in model.named_parameters():\n", " if name.endswith(\".bias\"):\n", " continue\n", " key_name = f\"Layer {name.split('.')[1]}\"\n", " weights[key_name] = param.detach().view(-1).cpu().numpy()\n", "\n", " # Plotting\n", " fig = plot_dists(weights, color=color, xlabel=\"Weight vals\")\n", " fig.suptitle(\"Weight distribution\", fontsize=14, y=1.05)\n", " plt.show()\n", " plt.close()\n", "\n", "\n", "##############################################################\n", "\n", "\n", "def visualize_gradients(model, color=\"C0\", print_variance=False):\n", " \"\"\"\n", " Args:\n", " net: Object of class BaseNetwork\n", " color: Color in which we want to visualize the histogram (for easier separation of activation functions)\n", " \"\"\"\n", " model.eval()\n", " small_loader = data.DataLoader(train_set, batch_size=1024, shuffle=False)\n", " imgs, labels = next(iter(small_loader))\n", " imgs, labels = imgs.to(device), labels.to(device)\n", "\n", " # Pass one batch through the network, and calculate the gradients for the weights\n", " model.zero_grad()\n", " preds = model(imgs)\n", " loss = F.cross_entropy(preds, labels) # Same as nn.CrossEntropyLoss, but as a function instead of module\n", " loss.backward()\n", " # We limit our visualization to the weight parameters and exclude the bias to reduce the number of plots\n", " grads = {\n", " name: params.grad.view(-1).cpu().clone().numpy()\n", " for name, params in model.named_parameters()\n", " if \"weight\" in name\n", " }\n", " model.zero_grad()\n", "\n", " # Plotting\n", " fig = plot_dists(grads, color=color, xlabel=\"Grad magnitude\")\n", " fig.suptitle(\"Gradient distribution\", fontsize=14, y=1.05)\n", " plt.show()\n", " plt.close()\n", "\n", " if print_variance:\n", " for key in sorted(grads.keys()):\n", " print(f\"{key} - Variance: {np.var(grads[key])}\")\n", "\n", "\n", "##############################################################\n", "\n", "\n", "def visualize_activations(model, color=\"C0\", print_variance=False):\n", " model.eval()\n", " small_loader = data.DataLoader(train_set, batch_size=1024, shuffle=False)\n", " imgs, labels = next(iter(small_loader))\n", " imgs, labels = imgs.to(device), labels.to(device)\n", "\n", " # Pass one batch through the network, and calculate the gradients for the weights\n", " feats = imgs.view(imgs.shape[0], -1)\n", " activations = {}\n", " with torch.no_grad():\n", " for layer_index, layer in enumerate(model.layers):\n", " feats = layer(feats)\n", " if isinstance(layer, nn.Linear):\n", " activations[f\"Layer {layer_index}\"] = feats.view(-1).detach().cpu().numpy()\n", "\n", " # Plotting\n", " fig = plot_dists(activations, color=color, stat=\"density\", xlabel=\"Activation vals\")\n", " fig.suptitle(\"Activation distribution\", fontsize=14, y=1.05)\n", " plt.show()\n", " plt.close()\n", "\n", " if print_variance:\n", " for key in sorted(activations.keys()):\n", " print(f\"{key} - Variance: {np.var(activations[key])}\")\n", "\n", "\n", "##############################################################"]}, {"cell_type": "markdown", "id": "e2e7f03b", "metadata": {"papermill": {"duration": 0.031273, "end_time": "2021-09-16T12:36:10.045659", "exception": false, "start_time": "2021-09-16T12:36:10.014386", "status": "completed"}, "tags": []}, "source": ["## Initialization\n", "\n", "Before starting our discussion about initialization, it should be noted that there exist many very good blog posts about the topic of neural network initialization (for example [deeplearning.ai](https://www.deeplearning.ai/ai-notes/initialization/), or a more [math-focused blog post](https://pouannes.github.io/blog/initialization/#mjx-eqn-eqfwd_K)).\n", "In case something remains unclear after this tutorial, we recommend skimming through these blog posts as well.\n", "\n", "When initializing a neural network, there are a few properties we would like to have.\n", "First, the variance of the input should be propagated through the model to the last layer, so that we have a similar standard deviation for the output neurons.\n", "If the variance would vanish the deeper we go in our model, it becomes much harder to optimize the model as the input to the next layer is basically a single constant value.\n", "Similarly, if the variance increases, it is likely to explode (i.e. head to infinity) the deeper we design our model.\n", "The second property we look out for in initialization techniques is a gradient distribution with equal variance across layers.\n", "If the first layer receives much smaller gradients than the last layer, we will have difficulties in choosing an appropriate learning rate.\n", "\n", "As a starting point for finding a good method, we will analyze different initialization based on our linear neural network with no activation function (i.e. an identity).\n", "We do this because initializations depend on the specific activation\n", "function used in the network, and we can adjust the initialization\n", "schemes later on for our specific choice."]}, {"cell_type": "code", "execution_count": 12, "id": "67fe474f", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:10.111297Z", "iopub.status.busy": "2021-09-16T12:36:10.110839Z", "iopub.status.idle": "2021-09-16T12:36:13.211461Z", "shell.execute_reply": "2021-09-16T12:36:13.211014Z"}, "papermill": {"duration": 3.134635, "end_time": "2021-09-16T12:36:13.211585", "exception": false, "start_time": "2021-09-16T12:36:10.076950", "status": "completed"}, "tags": []}, "outputs": [], "source": ["model = BaseNetwork(act_fn=Identity()).to(device)"]}, {"cell_type": "markdown", "id": "c440676a", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.031683, "end_time": "2021-09-16T12:36:13.275669", "exception": false, "start_time": "2021-09-16T12:36:13.243986", "status": "completed"}, "tags": []}, "source": ["### Constant initialization\n", "\n", "The first initialization we can consider is to initialize all weights with the same constant value.\n", "Intuitively, setting all weights to zero is not a good idea as the propagated gradient will be zero.\n", "However, what happens if we set all weights to a value slightly larger or smaller than 0?\n", "To find out, we can implement a function for setting all parameters below and visualize the gradients."]}, {"cell_type": "code", "execution_count": 13, "id": "c9298d5a", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:13.342938Z", "iopub.status.busy": "2021-09-16T12:36:13.342463Z", "iopub.status.idle": "2021-09-16T12:36:23.537805Z", "shell.execute_reply": "2021-09-16T12:36:23.538201Z"}, "papermill": {"duration": 10.231163, "end_time": "2021-09-16T12:36:23.538350", "exception": false, "start_time": "2021-09-16T12:36:13.307187", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9Bbm5vdHMgMTAgMCBSIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDkxMS41MjUgMjE2LjY2NTYyNSBdIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovVHlwZSAvUGFnZSA+PgplbmRvYmoKOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEyIDAgUiA+PgpzdHJlYW0KeJzVnUuTI8d1hff4FVjaCybz/VhKls0I7SQy7IXDC5ocScOYHgVfVvjf+1ygB5WVOLfZjX6MLyeG0X0GSNT5MisrLyrvrXD84fDl78Lxrz8f8b+jP/6Av//Az1/J7weP3+4OIwRXYsHPHy4/x1BdraXixw942e7Xvx0Ofzl4N0KrufnS+3H9JQ8fRvWtH3+SD/3q6gWXXw7Lqw+H2lzCx+ToxvkD7w5hZBdyCj1O8odZDq27/km/b2GnnY75x+NV4yFW18P9H7STihvHn94d/+P48fjl7+IZ3B/x9wf8PYE7fPmHd//z/rt3f/7q98fvfj6EmN0YPoW0P+RN3h3H4evDn44/fmrau1DQLZ9aP/361b16+PEQwO4Lj38KMbh8ai2UdIzFBS/NfXd3+P03xy//DS8Ix2/+curOb74//Ofxn/w/H//r+M0fD//6zeFPpw98BeMd9FOPIe6Nb/JLGO/eVWktlBEf5bzMzqeWenAtj4Z2/AjH4drcRty38dVP335/vPv2rx/f//Lr9++UBkNpONzefT0DRA+dverHFj61FQGsxxF8wDu2f5567UfA8EE+zN//FJJrtYaQYg/ynu/u5DO++MO7H77991+//vbjz1/cvf/468/HP/z9+KdX6/THn5y39XfxruPtqbvWy0jtM4/x+xYb0I9UR93b3eRb7WIciN06XA3Z++7Lg46Dx39v5jr47GrETL3YnvRn+oYfl3PEzF/DQ77j2/rGG3MePbXF96Y/13cKLjYfMq6DD04Y6W2N14rDasOvxjf9ucZrdB4zXxgtPWA7r7Z/PEgjX0hzWIjIsY6EhkoKe3zLPP4vf//14y/PxJfONuPoPo2ISf8gY/a0dokJayZXP61e8mmx42KNuGDlXO5Z7958nN98WN58OGBG8X3kul9LtOLSKDH33QS7qRhE+DGH00S0NbGTsf75/f9/+8Qo7DeHc8WntLd/UXtx5XQwM5NJtGGdmLw7YPXjT+uLnfVNxVrLo32sCXdAJtWEeWYT5jF8cZB+3++biqtmk9m475DMqg3zxCbM4xobeupLz1/U6uGojZH3SCbVhnliE+arGzGHlvfmLyquRzVj7Z32SCbVhnliE+aHi3GMMvbmL+pu2G9IzA17YlO+inAllZb3Pb+pNbjhKz5gj2RSTZhnNmE+uY4BHPc9v6nz2T0hsXbOM5swX7HcbCksPX9RJdpuIe6AbJoN48QijHeXS/Jj6fWLWrp8cVFG2+OYVBvmiU1EGYgAG97S990+yfOsvrVhbq6nRsU/lus11zoW/xe5YDEXcvYLllk2AoA4FQBYsjefS1kAXOR5cp8aMTfnU6cCoLnaalxPgE/q3j6lYsT+tU9xP+SbUB/X7r/ItaDHU+h1T2VSjdgnRuE/RBd762Fv/6Lue3+DYq73iU9xn10ZqYyl9zd5u8JvLRi76lOT4h1HOEbqi/VP6m7cb03YG/jEp7jH4t2XUNeev8jzFX6iYu66z4zK7UGs33FCL/Yv6jzuLw1YG/jEozjH8j3Ulupi/SLPS9uJiLUFLzUq/rGGj7HEZcqfZPkiz/fQ9lhm2QgA4lQAYBUfe/LrALjI+wFwacTgCCBOASBhFZ+y3IbeAdjk3dy3NWJv8mNOBQBW8Wn0toyATZ6/05i5WPuqgzoVAFjH51LrOgIu8u4U2BqxdwowpwIAC/mC+HUdARd5dwpsjRg8BYhTAYC1fGkxrSPgIu8BUC5GABCnAIBfYk0el7MdgE3ezwEbF3NzAHMqALCmr737ZU/bJtfuYhijL1g21Yh94lPs4whbLn3t/4u8C30nKuZCX+ZUAAwX0ERb+/8it+BSq60tXGbZCADiFABKcFk2/iwjYJP3ACgXGwCYUwEg+8fCyMsI2OQdAM7FCADiVABU50drcR0BF3kU50tsKe25zLIRAMSpAMDa3qcS1hFwkadtLFMb1ja3UJ+wj2t58wPN7O1v8nwJ3NowdwlkPsU+1vWh+J4W+xd5tk+hGLFPfIp9rOqj73Xt/Yu8+wp0omLuK1DmVABgVR9rLWv/X+RdDDBxMRcDMKcCAKv6FPP6Hfgm79aAExdza0DmFAAalvVoNC4jYJPnbT0zF2u7fahTAYB1fZad7wuAi4z3jVLw9j2XWTYCgDgVADjCPNpYR8BFxryXe2597LnMshEAxKkAwLq+lNKWDRCbvJsDJi7m5gDmVPKlJHEopLpsANjkgtXeiPiUPZdZtgGAORUAktjUEOAtAC7yvJ95asTcNmfqVABgYd/iSOsIuMjT3cCtCWt3A5lLMd8RzPYW1t6/yLN5QuRi/j4jJUqmi0Mj5yP8zfQ1lix6+Jpntd2pKad4x5NS464/9cG2PUw9NtWmR9ea5ImebplIOuZ9S3qezIdv//fdTz877/7x7v1f//bLcU6Zuc8efkq67znHeE37vcoY3qf9Yli5fr09KrXgau91uV08yTtWUyvX6b8ySi4vkD+3JlEd2NiKRVJVJaVrZ2CSXyAfNqbmYisl4xzzRc+HHQgRaxu1t7LlS7ryUH6pf1xO6esBTB6BbRtodj8CNvklAI7hSpTWch71UanU7ioD7xUZIMojOQGT/AIM0mn2RGupZJyKj2KgpVXHgqE2sBI/0bg5r/qVT80nzS23Zjj2897M89HflsH8Jggkji6YHuoewSbfjMD7GYGEKC0iRkmthoeyewtJcn0TEiO6lPBz25PY5BcigcWM1B0II5YYH6pdwNJ934RECM35eMqc3KGY9BdiIfs4O36oPsT+UBJw+2zDIuTgGpwjuN7D2PSb61mMtIOBy3lJMfjgof1W5v9nolGrS6OnuNLY9JeiUSOWLb3mIFU1HqJxNWXMCeK9zlBuTxS/ka5EVQc1MHpqVPVgSHbQF534Jfu2bC5QZdLIFFnZAEGsCYjiYq55yaNVZcrHGAhiTUBgpsHVp6XFsSJTPsZAEGsCYrheOt62OFZkyscYCGINIDDbhpp6XrpekTkfWyCYNQGRXa6jxqXrNZnyMQaCWBMQ1bVWclhHhCJTPsZAEGsCYjjfQxhr1ysy5WMMBLEGELLVqNexbFHQZM7HFghmTUAkvDK2JU9XlSkfYyCINQGBQxq9lLw4VmTKxxgIYk1AdJd8TmntekWmfIyBINYAQko0IoJdMpg1mfOxBYJZExD45VRfcgGhyJSPMRDEmoBA9BBDHeuIUGTKxxgIYk1AIHqILbe16xWZ8jEGglgTEIgeUop17XpFpnyMgSDWAGIgekjD7/N+FZXTsYWBOBMKCB1y7mnFoMiMjjEMxJlwQORQfF2+jlNUCscYhmtnQgFhQ6n56qRQZEbHGAbiTG50I2qocdkJw1UFjikMzJlQQMhQ22hlwaDIjI4xDMSZcMAhtbTkSGsqhWMMw7WzO74n5MN+T8hU5XZrhBe/tQGCb4MJ3pVe0pIxrMmcmy0QzJqAiK6PEMJyDmgy5WMMBLEmIIrDy/DaxbEiUz7GQBBrAgLxgo+1r12vyJSPMRDEmoBAxOB7bmvXKzLlYwwEsQYQESFDyLEsXa/InI8tEMyagEDUEL3PS9drMuVjDASxJiAQNsTS4zoiFJnyMQaCWBMQiBxSqEvOsSpTPsZAEGsAkRA6pJbGfteLJnM+tkAwawIC0cN1loIqUz7GQBBrAgKHlPuo64hQZMrHGAhiTUAgeii5lbXrFZnyMQaCWAOIjOih+pKWrldkzscWCGZNQCB6qBULhAWEIlM+xkAQawIC0UOLwa8jQpEpH2MgiDUB0VxurS+Zy6pM+RgDQawJiOFaT7X1xbEiUz7GQBBrAFHi9dOrPqgy52MLBLMmIDJ7YJcqUz7GQBBrAqKyR3epMuVjDASxJiAGe5CXKlM+xkAQa5LIGNhDvTSZ87EFglkTEIk94EuVKR9jIIi1O57x+kGXKZ8nFQFQskG11PHD1zxR9E5NQsc7npxxyj/9wc94SlGAKGVTQ0dvnG+aPaEqQPycVQFyjC5f7zArWDzE6xLak7yDNbXCqwJsL/D7j3p2Il2OnRW/nOQXSOjOAd0Z/Ri1D/9AQvfrVQV4RYC9nB4du6N3r70Eutacb3nIfzdWA3g97yUnlqcxyS9AoKTsWpDWsAoNz6sGkJPUZzq9/TmVAF7zdHzKfHLjkJJNCNvR31YJ4E0QdPxU4XpP4KLeXF8C59QEAEM09px9xMzen/qU7zfhIE8ZwQlQSt2TmPQXYiH59K3HXnvy6aEU56tHYL8djHqqsOVjWGBs+kvBKA0jI6SUcPXPD8CoDzwPXBZIE5Pb071vhPuma2htNZE8K6WtyayRaQ1tAgSzJiAiK6mtypSPMRDEmoAorLS2KlM+xkAQawKisRLbqkz5GANBrAmIwYptqzLlYwwEsQYQ+IWU3dZkzscWCGZNQGRWgFuVKR9jIIg1AVFZIW5VpnyMgSDWBERnBblVmfIxBoJYA4gSWGFuTeZ8bIFg1gREYgW6VZnyMQaCWBMQhRXqVmXKxxgIYk1AdFawW5UpH2MgiDWAqJ4V7tZkzscWCGZNQERWwFuVKR9jIIg1AVFYIW9VpnyMgSDWBERjRb1VmfIxBoJYExDD1Z6LX7tekSkfYyCINYBowY3hY1+6XpE5H1sgmDUBkdEGXrZ0vSZTPsZAEGsCAtGDD0vykaZSOsYwXDsTCggdfKtLnqYqMzrGMBBnckcXkUNIS1E9ReVwbGEgzoQCwoYg718wKDKjYwwDcXZH7vl/UDQKxhgCtr8BwULCS1a7F3mX3r5th2A57zYgEMPgMPC2VNf7V1zlzGxhIM6EAgIFeeDHMho0mdExhoE4Ew6IEzDpLUskRaVwjGG4diYUECQUNLAskTSZ0TGGgTgTDogRymghLIYVmeIxBoJYuzsUL5tUijzNZramyAofUyCoNQGBKKGF2JdNoZpM+RgDQawJiOpKa76uI0KRKR9jIIg1AdEd3tHL2vWKTPkYA0GsAUQILlzvO9dkzscWCGZNQCSXR85x6XpNpnyMgSDWBARiBu/jkpCpypSPMRDEmoBA3ODLGHFxrMiUjzEQxBpARLwthNaWrldkzscWCGZNQCB2CK0sJSZVmfIxBoJYExAIHmJKeR0Rikz5GANBrAkIeVzrCGntekWmfIyBINYExGBPilJlyscYCGINIFJgT4rSZM7HFghmTUBk9qQoVaZ8jIEg1gREZU+KUmXKxxgIYk1AdPakKFWmfIyBINYkdyawJ0VpMudjCwSzdsfTGD/oMuXzpBRvJc1PywM+fM0zAO/UjGK848mphPzTH/yMp6R4y5b9Xs/Jl0/J786fM78bAQQYXW0pa1J28Lp88STvSE2t8PzuywvkT370yfOIxLnqC6uyOMkvkaIrTySuFedkjuWBNOXXy+9+RYD1/hmw+5s9k/wCAGvJrsZTfYxYHpnjfJV9+XoMWgwsQWOSX4BBCxGTO1rDDI+59Vl53jVkecDZuZrE83K9X4/q0+aWG8/Mns67Wc9Hf1uu95sgmHK9JwLPzvUupcwAHpvrHVh685tw2OV6zxeU5+d6LywenetNE9/fBsac6z3DeH6u9wrjsbne6YFc7yKJNlubt+d63wj3TRfT6sqis0LEqkwamRbTNkAQawARPCtErMmcjy0QzJqAiKwQsSpTPsZAEGsCIrNCxKpM+RgDQawJiMYKEasy5WMMBLEmIAYrRKzKlI8xEMQaQGB1QwoRazLnYwsEsyYgMitErMqUjzEQxJqAqKwQsSpTPsZAEGsCorNCxKpM+RgDQawBRAqsELEmcz62QDBrAiKxQsSqTPkYA0GsCYjCChGrMuVjDASxJiA6K0SsypSPMRDEGkBkzwoRazLnYwsEsyYgIitErMqUjzEQxJqAyKxQsypTPsZAEGsCAtFDHKOMxbEiUz7GQBBrAgLRQyotr12vyJSPMRDEGkAURA85lLh0vSJzPrZAMGsCAtFDbmlJQ1JlyscYCGJNQCB6KMmPdUQoMuVjDASxJiAQPeA9fe16RaZ8jIEg1uQ+N6KHmmtdul6ROR9bIJi1O74P4IMuUz7GQCg7H4qrrcblFLiouyzv3TYJlv1tA8S1Y+HQ3OjRx7XnFZlRM4aBOAOHhjZ6W9LzFJXDsYWBOBMK0ZWRyli6XZMZHWMYiDPhkF0fYyk1qakUjjEM186EAoIGX0Jdu12RGR1jGIgz4YCYIfir0c9VCscYhmtnoNARMITaljRNTaZ0bGFgzoQD4oUYy5KlqcoUjzEQxJqAQLwQe/LrgFBkyscYCGJNQCBeuN6MrsqUjzEQxBpADMQLafS2dL0icz62QDBrAgIBQy6InxYQikz5GANBrAkIBAwl5LyOCEWmfIyBINYEBGKG0mJau16RKR9jIIi1u0Pz0mzyS5amIit8TIGg1gQE4gbMgUuWpipTPsZAEGsCQh7KmUtfR4QiUz7GQBBrAqKxR0apMuVjDASxJiAGe2SUKlM+xkAQawARAntklCZzPrZAMGsCIrNHRqky5WMMBLEmICp7ZJQqUz7GQBBrAqKzR0apMuVjDASxdsfTGj+oMufzpLxvJeVPyw8+fM2zAe/UTGPJFH9qWiH/9Ac/4yl535hVMYBGCfe3gZ6Q+l0/Z+p3a82V6x1mw8vjtfYQP2k7TNP7edL39oKw/5Dn5+vKc0wjFjj7W5iT/BL5ujVjOVliSH1g6HyGpO/XA4iJxo3RxrJhZpJfAGDHzBozWksVscltOd+viKBF51tY58xJfgkEcns4SmsptUcy0HK+sWxxI4Ui34vJg3afkfP9umfmIyeVW8/KcL+l9XTotyV8v4n/2jH6c81p8r9pN/svefaPOcrXPIpsXAsPICifh8FIrsQk39ptDDbtZga5zQz6cCmn0XCiAeyDOe+fBULAwqr7iGvERGESXwhDCFJjI44RMH89lOEdPtNgCBmDFReFUWcOm/hSHBKWez5630fsv5H2/3k4VAzYeuqkicMmvhSHiqtNDzLGemkPcShavn/LfsZxe77/jVzzmwZRPbFi3oDHqlJP8m7j19QK3xBmAwXxLCgaq0utyhScMRDEGkAMz+pSazLnYwsEsyYgIqtLPcm4GDlc5EYIe0A73RgK4llQZFaZWpUpOGMgiDUB0Vhl6klexsSlFctjgngWFIPVplZlCs4YCGINUbUPrDa1Iit8TIGg1gREZrWpJ3l3Ekyt2D05qGdBUVl1alWm4IyBINYERGfVqVWZ8jEGglgDiOBZdWpN5nxsgWDWBERi1alVmfIxBoJYExCFVadWZcrHGAhiTUA0Vp1alSkfYyCINblF4Vl1ak3mfGyBYNYERGTVuyd5v5LYWjG8kmCez7etWis5rGNCkSk4YyCINQHR5MvUMNYxociUjzEQxJqAGFIfefS16xWZ8jEGglgDiBTwytiWPF1N5nxsgWDWBERyY/RS8gJCkSkfYyCINQGB+MHnlNYRociUjzEQxJqAQPwQfIhr1ysy5WMMBLEGEFk2Ci0P0vigypyPLRDMmoBA/BBDHUvXb/J+SbW1YnhJxTwLCkQQseW2jglFpuCMgSDWBAQiiJRiXceEIlM+xkAQawBREECk4febeBSV07GFgTgTCggecu5pxXCR91PEpRHDMwSzLCQQPeDYly/lFJVSM4bh2plQQOhQar46LRSZ0TGGgTgTDrJ9JYbluqCoFI4xDNfOQKEibMBZ3pZuV2RKxxYG5kw4IGpoacla1lQKxxiGa2dCobp4nZGgyoyOMQzEmXDorvSSlkxdVaZ4jIEg1mRvuXcdK4CwnAGKzPnYAsGsCYjk8DK8dgGhyJSPMRDEmoBAvOBj7euIUGTKxxgIYk1AIGLwiBvWrr/I+3X11orhhTXzDBQdMYM8aHLpfEXm4GyBYNYEBKKG6H1exoQmUz7GQBBrAiKzh4ipMuVjDASxJiAae4iYKlM+xkAQawJisIeIqTLlYwwEsQYQI7CHiGky52MLBLMmIBJ7iJgqUz7GQBBrAqKyh4ipMuVjDASxJiA6e4jYJO+XVFsrL7qkmnOjDsGVUwbTo8D8ec2hmpNrKj4+x5HvM3S8H4iTapZbU7lF2UTcq2u5yQ27igPyNeOgz8kY+EUq9CLQLDgIf371kETKBsuQRwOGTxkdMSaZMGpzMQw/8kkODt0lFRRErmh7HM+b+kv2ctcYMrCkcGobY6yOIcUB0HbHkaZ0kovLqWU4hexT8v3cCN7aSglJ7IyMQzh/5HB5xN7RSDp/p3zeBCnd0EfGq71skPV5HM+bRDGYPCKFAju9jB7ut0w2jLSWjqWApK9x3G8gxGszSJQMl6XXkxpdlA1DaONU5Br2TnJ2pRT037HAburouZMMay3hyI8luIEmzk1H73DUpVWRY2jorJOMzolYpECWx2GmMxLZf4BxkVoXOfTaTpmycjc+lSpN48U++Prp1jS6o7UhcpEdsecXA2uAiXD6xCQE729bgnCvVQ47JdkUdrpTERzG2ohZTNYy6qmD5YZWiXGUdDw9A0BG6/3tnVj96CdQOYHwuREwHkMy+0XGEfnzq6WKYUP3CdUaeun3twSyfAk85BPbqOfhKl+Rx5DQtsghtHimXSRlHL0XxU1tyV++Ogyn2gRHeWqFfFvS7r9Lyx1LnnKUz5Y9LfH+qyXArvIcg9PzocMJlAwv9JPsK+sg6WP/9P1LHbVKzaQuBxLquWl0Tcg+n9rAoJAqleeQtObeZMslBmhp0Z/jM5xbGPD+dHwyAE6DuJ8y6mM/9yTmErg+XZbRCeX0BBK8GCduPeHr+Hj0aT21URNOlPzp0oWARzJnhsN4yv4yfeGw6rlpjzNo7Gc7vBpHW33R5X3e5YOVTpREd60oBlq+zoG/44U1pCrKk9Lo+YfqrT+luonMaRheo93f6H5CdZPOqptMTSdMaS2f6mYE9Efft5mvCxq8f/fxl+P373/+5af3//3rL+///nGXZnr4PzT1zl8KZW5kc3RyZWFtCmVuZG9iagoxMiAwIG9iago2NjkyCmVuZG9iagoxMCAwIG9iagpbIF0KZW5kb2JqCjE3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjM1ID4+CnN0cmVhbQp4nDVRSW4AMQi75xX+QKWwJ++Zquqh/f+1hlEvAwPY2CTvwUYkPsSQ7ihXfMrqNMvwO1nkxc9K4eS9iAqkKsIKaQfPclYzDJ4bmQKXM/FZZj6ZFjsWUE3EcXbkNINBiGlcR8vpMNM86Am5PhhxY6dZrmJI691Svb7X8p8qykfW3Sy3TtnUSt2iZ+xJXHZeT21pXxh1FDcFkQ4fO7wH+SLmLC46kW72mymHlaQhOC2AH4mhVM8OrxEmfmYkeMqeTu+jNLz2QdP1vXtBR24mZCq3UEYqnqw0xoyh+o1oJqnv/4Ge9b2+/gBDTVS5CmVuZHN0cmVhbQplbmRvYmoKMTggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDcgPj4Kc3RyZWFtCnicTVFJbsQwDLv7FfzAAJasxXlPikEP7f+vJR0U7cEQI0tc4u7ERBZetlDXQofjw0ZeCZuB74PWnPgaseI/2kaklT9UWyATMVEkdFE3GvdIN7wK0X6kgleq91jzEXcrzVs6drG/98G05pEqq0I85Ngc2Uha10TR8T203nNDdMoggT43IQdEaY5ehaS/9sN1bTS7tTazJ6qDR6aE8kmzGprTKWbIbKjHbSpWMgo3qoyK+1RGWg/yNs4ygJPjhDJaT3asJqL81CeXkBcTccIuOzsWYhMLG4e0H5U+sfx86834m2mtpZBxQSI0xaXfZ7zH53j/AJVPXCYKZW5kc3RyZWFtCmVuZG9iagoxOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMwNyA+PgpzdHJlYW0KeJw9kktuAzEMQ/c+hS4QwPrZnvOkKLqY3n/bJyXpihzZFkVqlrpMWVMekDSThH/p8HCxnfI7bM9mZuBaopeJ5ZTn0BVi7qJ82cxGXVknxeqEZjq36FE5Fwc2Taqfqyyl3S54Dtcmnlv2ET+80KAe1DUuCTd0V6NlKTRjqvt/0nv8jDLgakxdbFKrex88XkRV6OgHR4kiY5cX5+NBCelKwmhaiJV3RQNB7vK0ynsJ7tveasiyB6mYzjspZrDrdFIubheHIR7I8qjw5aPYa0LP+LArJfRI2IYzcifuaMbm1MjikP7ejQRLj65oIfPgr27WLmC8UzpFYmROcqxpi1VO91AU07nDvQwQ9WxFQylzkdXqX8POC2uWbBZ4SvoFHqPdJksOVtnbqE7vrTzZ0PcfWtd0HwplbmRzdHJlYW0KZW5kb2JqCjIwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ0ID4+CnN0cmVhbQp4nEWRTXIFIQiE956iL/Cq5Fc9z6RSWUzuvw3NvCQrWoXmA9MCE0fwEkPsiZUTHzJ8L+gyfLcyO/A62ZlwT7huXMNlwzNhW+A7Kss7XkN3tlI/naGq7xo53i5SNXRlZJ96oZoLzJCIrhFZdCuXdUDTlO5S4RpsW4IU9UqsJ52gNOgRyvB3lGt8dRNPr7HkVM0hWs2tExqKsGx4QdTJJBG1DYsnlnMhUfmqG6s6LmCTJeL0gNyglWZ8elJJETCDfKzJaMwCNtCTu2cXxppLHkWOVzSYsDtJNfCA9+K2vvc2cY/zF/iFd9//Kw591wI+fwBL/l0GCmVuZHN0cmVhbQplbmRvYmoKMjEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzEgPj4Kc3RyZWFtCnicNU85kgQhDMt5hT4wVRjbQL+np7Y22Pl/upKZTpDwIcnTEx2ZeJkjI7Bmx9taZCBm4FNMxb/2tA8TqvfgHiKUiwthhpFw1qzjbp6OF/92lc9YB+82+IpZXhDYwkzWVxZnLtsFY2mcxDnJboxdE7GNda2nU1hHMKEMhHS2w5Qgc1Sk9MmOMuboOJEnnovv9tssdjl+DusLNo0hFef4KnqCNoOi7HnvAhpyQf9d3fgeRbvoJSAbCRbWUWLunOWEX712dB61KBJzQppBLhMhzekqphCaUKyzo6BSUXCpPqforJ9/5V9cLQplbmRzdHJlYW0KZW5kb2JqCjIyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ5ID4+CnN0cmVhbQp4nD1QO45EIQzrOYUv8CTyI3AeRqstZu/frgOaKVBMfrYzJNARgUcMMZSv4yWtoK6Bv4tC8W7i64PCIKtDUiDOeg+IdOymNpETOh2cMz9hN2OOwEUxBpzpdKY9ByY5+8IKhHMbZexWSCeJqiKO6jOOKZ4qe594FiztyDZbJ5I95CDhUlKJyaWflMo/bcqUCjpm0QQsErngZBNNOMu7SVKMGZQy6h6mdiJ9rDzIozroZE3OrCOZ2dNP25n4HHC3X9pkTpXHdB7M+Jy0zoM5Fbr344k2B02N2ujs9xNpKi9Sux1anX51EpXdGOcYEpdnfxnfZP/5B/6HWiIKZW5kc3RyZWFtCmVuZG9iagoyMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM5NSA+PgpzdHJlYW0KeJw9UktuxUAI2+cUXKDS8JvPeVJV3bz7b2tDUqkqvIkxxjB9ypC55UtdEnGFybderls8pnwuW1qZeYi7i40lPrbcl+4htl10LrE4HUfyCzKdKkSozarRofhCloUHkE7woQvCfTn+4y+AwdewDbjhPTJBsCTmKULGblEZmhJBEWHnkRWopFCfWcLfUe7r9zIFam+MpQtjHPQJtAVCbUjEAupAAETslFStkI5nJBO/Fd1nYhxg59GyAa4ZVESWe+zHiKnOqIy8RMQ+T036KJZMLVbGblMZX/yUjNR8dAUqqTTylPLQVbPQC1iJeRL2OfxI+OfWbCGGOm7W8onlHzPFMhLOYEs5YKGX40fg21l1Ea4dubjOdIEfldZwTLTrfsj1T/5021rNdbxyCKJA5U1B8LsOrkaxxMQyPp2NKXqiLLAamrxGM8FhEBHW98PIAxr9crwQNKdrIrRYIpu1YkSNimxzPb0E1kzvxTnWwxPCbO+d1qGyMzMqIYLauoZq60B2s77zcLafPzPoom0KZW5kc3RyZWFtCmVuZG9iagoyNCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI0OSA+PgpzdHJlYW0KeJxNUUmKAzAMu+cV+kAhXpO8p0OZQ+f/18oOhTkECa+Sk5aYWAsPMYQfLD34kSFzN/0bfqLZu1l6ksnZ/5jnIlNR+FKoLmJCXYgbz6ER8D2haxJZsb3xOSyjmXO+Bx+FuAQzoQFjfUkyuajmlSETTgx1HA5apMK4a2LD4lrRPI3cbvtGZmUmhA2PZELcGICIIOsCshgslDY2EzJZzgPtDckNWmDXqRtRi4IrlNYJdKJWxKrM4LPm1nY3Qy3y4Kh98fpoVpdghdFL9Vh4X4U+mKmZdu6SQnrhTTsizB4KpDI7LSu1e8TqboH6P8tS8P3J9/gdrw/N/FycCmVuZHN0cmVhbQplbmRvYmoKMjUgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA5NCA+PgpzdHJlYW0KeJxFjcERwCAIBP9UQQkKCtpPJpOH9v+NEDJ8YOcO7oQFC7Z5Rh8FlSZeFVgHSmPcUI9AveFyLcncBQ9wJ3/a0FScltN3aZFJVSncpBJ5/w5nJpCoedFjnfcLY/sjPAplbmRzdHJlYW0KZW5kb2JqCjI2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzQxID4+CnN0cmVhbQp4nEVSS25EMQjbv1NwgUjhl5DztKq6mN5/W5tM1c3gCWBseMtTpmTKsLklIyTXlE99IkOspvw0ciQipvhJCQV2lY/Ha0usjeyRqBSf2vHjsfRGptkVWvXu0aXNolHNysg5yBChnhW6snvUDtnwelxIuu+UzSEcy/9QgSxl3XIKJUFb0HfsEd8PHa6CK4JhsGsug+1lMtT/+ocWXO9992LHLoAWrOe+wQ4AqKcTtAXIGdruNiloAFW6i0nCo/J6bnaibKNV6fkcADMOMHLAiCVbHb7R3gCWfV3oRY2K/StAUVlA/MjVdsHeMclIcBbmBo69cDzFmXBLOMYCQIq94hh68CXY5i9Xroia8Al1umQvvMKe2ubnQpMId60ADl5kw62ro6iW7ek8gvZnRXJGjNSLODohklrSOYLi0qAeWuNcN7HibSOxuVff7h/hnC9c9usXS+yExAplbmRzdHJlYW0KZW5kb2JqCjI3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTY0ID4+CnN0cmVhbQp4nEWQx3EFMQxD76oCJTCACvWsx/MP6/6vhvTTQXoYQgxiT8KwXFdxYXTDj7ctMw1/RxnuxvoyY7zVWCAn6AMMkYmr0aT6dsUZqvTk1WKuo6JcLzoiEsyS46tAI3w6sseTtrYz/XReH+wh7xP/KirnbmEBLqruQPlSH/HUj9lR6pqhjyorax5q2leEXRFK2z4upzJO3b0DWuG9las92u8/HnY68gplbmRzdHJlYW0KZW5kb2JqCjI4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNzIgPj4Kc3RyZWFtCnicMzK3UDBQsDQBEoYWJgrmZgYKKYZcQL6piblCLhdIDMTKAbMMgLQlnIKIZ4CYIG0QxSAWRLGZiRlEHZwBkcvgSgMAJdsWyQplbmRzdHJlYW0KZW5kb2JqCjI5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNDcgPj4Kc3RyZWFtCnicMzK3UDBQsDQBEoYWJgrmZgYKKYZclhBWLhdMLAfMAtGWcAoinsGVBgC5Zw0nCmVuZHN0cmVhbQplbmRvYmoKMzAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNTggPj4Kc3RyZWFtCnicRZFLcgQgCET3noIjgPzkPJNKZTG5/zYNzmQ2dpeo/YRKI6YSLOcUeTB9yfLNZLbpdzlWOxsFFEUomMlV6LECqztTxJlriWrrY2XkuNM7BsUbzl05qWRxo4x1VHUqcEzPlfVR3fl2WZR9Rw5lCtiscxxs4MptwxgnRput7g73iSBPJ1NHxe0g2fAHJ419lasrcJ1s9tFLMA4E/UITmOSLQOsMgcbNU/TkEuzj43bngWBveRFI2RDIkSEYHYJ2nVz/4tb5vf9xhjvPtRmuHO/id5jWdsdfYpIVcwGL3Cmo52suWtcZOt6TM8fkpvuGzrlgl7uDTO/5P9bP+v4DHilm+gplbmRzdHJlYW0KZW5kb2JqCjMxIDAgb2JqCjw8IC9CQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM5Ci9TdWJ0eXBlIC9Gb3JtIC9UeXBlIC9YT2JqZWN0ID4+CnN0cmVhbQp4nOMyNDBTMDY1VcjlMjc2ArNywCwjcyMgCySLYEFkM7jSABXzCnwKZW5kc3RyZWFtCmVuZG9iagozMiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2MyA+PgpzdHJlYW0KeJxFkDsSAyEMQ3tOoSP4IwM+z2YyKTb3b2PYbFLA01ggg7sTgtTagonogoe2Jd0F760EZ2P86TZuNRLkBHWAVqTjaJRSfbnFaZV08Wg2cysLrRMdZg56lKMZoBA6Fd7touRypu7O+UNw9V/1v2LdOZuJgcnKHQjN6lPc+TY7orq6yf6kx9ys134r7FVhaVlLywm3nbtmQAncUznaqz0/Hwo69gplbmRzdHJlYW0KZW5kb2JqCjMzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjE4ID4+CnN0cmVhbQp4nD1QuY0EMQzLXYUaWMB67alnFotLpv/0SPn2ItEWRVIqNZmSKS91lCVZU946fJbEDnmG5W5kNiUqRS+TsCX30ArxfYnmFPfd1ZazQzSXaDl+CzMqqhsd00s2mnAqE7qg3MMz+g1tdANWhx6xWyDQpGDXtiByxw8YDMGZE4siDEpNBv+uco+fXosbPsPxQxSRkg7mNf9Y/fJzDa9TjyeRbm++4l6cqQ4DERySmrwjXVixLhIRaTVBTc/AWi2Au7de/hu0I7oMQPaJxHGaUo6hv2twpc8v5SdT2AplbmRzdHJlYW0KZW5kb2JqCjM0IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggODMgPj4Kc3RyZWFtCnicRYy7DcAwCER7pmAEfib2PlGUwt6/DRAlbrgn3T1cHQmZKW4zw0MGngwshl1xgfSWMAtcR1COneyjYdW+6gSN9aZS8+8PlJ7srOKG6wECQhpmCmVuZHN0cmVhbQplbmRvYmoKMzUgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA1MSA+PgpzdHJlYW0KeJwzNrRQMFAwNDAHkkaGQJaRiUKKIRdIAMTM5YIJ5oBZBkAaojgHriaHK4MrDQDhtA2YCmVuZHN0cmVhbQplbmRvYmoKMzYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNjAgPj4Kc3RyZWFtCnicRZA5EgMxCARzvYInSFyC96zLtcH6/6kH1kei6QI0HLoWTcp6FGg+6bFGobrQa+gsSpJEwRaSHVCnY4g7KEhMSGOSSLYegyOaWLNdmJlUKrNS4bRpxcK/2VrVyESNcI38iekGVPxP6lyU8E2Dr5Ix+hhUvDuDjEn4XkXcWjHt/kQwsRn2CW9FJgWEibGp2b7PYIbM9wrXOMfzDUyCN+sKZW5kc3RyZWFtCmVuZG9iagozNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMzNCA+PgpzdHJlYW0KeJwtUktyxSAM23MKXaAz+AfkPOl0uni9/7aSk0VGDmD0MeWGiUp8WSC3o9bEt43MQIXhr6vMhc9I28g6iMuQi7iSLYV7RCzkMcQ8xILvq/EeHvmszMmzB8Yv2XcPK/bUhGUh48UZ2mEVx2EV5FiwdSGqe3hTpMOpJNjji/8+xXMtBC18RtCAX+Sfr47g+ZIWafeYbdOuerBMO6qksBxsT3NeJl9aZ7k6Hs8Hyfau2BFSuwIUhbkzznPhKNNWRrQWdjZIalxsb479WErQhW5cRoojkJ+pIjygpMnMJgrij5wecioDYeqarnRyG1Vxp57MNZuLtzNJZuu+SLGZwnldOLP+DFNmtXknz3Ki1KkI77FnS9DQOa6evZZZaHSbE7ykhM/GTk9Ovlcz6yE5FQmpYlpXwWkUmWIJ2xJfU1FTmnoZ/vvy7vE7fv4BLHN8cwplbmRzdHJlYW0KZW5kb2JqCjM4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNzAgPj4Kc3RyZWFtCnicMzM2UzBQsDACEqamhgrmRpYKKYZcQD6IlcsFE8sBs8wszIEsIwuQlhwuQwtjMG1ibKRgZmIGZFkgMSC6MrjSAJiaEwMKZW5kc3RyZWFtCmVuZG9iagozOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMyMCA+PgpzdHJlYW0KeJw1UktuBTEI288puECl8E/O86qqi777b2sTvRVMMGDjKS9Z0ku+1CXbpcPkWx/3JbFC3o/tmsxSxfcWsxTPLa9HzxG3LQoEURM9WJkvFSLUz/ToOqhwSp+BVwi3FBu8g0kAg2r4Bx6lMyBQ50DGu2IyUgOCJNhzaXEIiXImiX+kvJ7fJ62kofQ9WZnL35NLpdAdTU7oAcXKxUmgXUn5oJmYSkSSl+t9sUL0hsCSPD5HMcmA7DaJbaIFJucepSXMxBQ6sMcCvGaa1VXoYMIehymMVwuzqB5s8lsTlaQdreMZ2TDeyzBTYqHhsAXU5mJlgu7l4zWvwojtUZNdw3Duls13CNFo/hsWyuBjFZKAR6exEg1pOMCIwJ5eOMVe8xM5DsCIY52aLAxjaCaneo6JwNCes6VhxsceWvXzD1TpfIcKZW5kc3RyZWFtCmVuZG9iago0MCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE4ID4+CnN0cmVhbQp4nDM2tFAwgMMUQ640AB3mA1IKZW5kc3RyZWFtCmVuZG9iago0MSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEzMyA+PgpzdHJlYW0KeJxFj0sOBCEIRPecoo7Axx/ncTLphXP/7YCdbhNjPYVUgbmCoT0uawOdFR8hGbbxt6mWjkVZPlR6UlYPyeCHrMbLIdygLPCCSSqGIVCLmBqRLWVut4DbNg2yspVTpY6wi6Mwj/a0bBUeX6JbInWSP4PEKi/c47odyKXWu96ii75/pAExCQplbmRzdHJlYW0KZW5kb2JqCjQyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzQwID4+CnN0cmVhbQp4nDVSOW4EMQzr/Qp9IIBu2+/ZIEiR/L8NqdkUA3F0UpQ7WlR2y4eFVLXsdPm0ldoSN+R3ZYXECcmrEu1ShkiovFYh1e+ZMq+3NWcEyFKlwuSk5HHJgj/DpacLx/m2sa/lyB2PHlgVI6FEwDLFxOgals7usGZbfpZpwI94hJwr1i3HWAVSG9047Yr3oXktsgaIvZmWigodVokWfkHxoEeNffYYVFgg0e0cSXCMiVCRgHaB2kgMOXssdlEf9DMoMRPo2htF3EGBJZKYOcW6dPTf+NCxoP7YjDe/OirpW1pZY9I+G+2Uxiwy6XpY9HTz1seDCzTvovzn1QwSNGWNksYHrdo5hqKZUVZ4t0OTDc0xxyHzDp7DGQlK+jwUv48lEx2UyN8ODaF/Xx6jjJw23gLmoj9tFQcO4rPDXrmBFUoXa5L3AalM6IHp/6/xtb7X1x8d7YDGCmVuZHN0cmVhbQplbmRvYmoKNDMgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNTEgPj4Kc3RyZWFtCnicLVFJcgNBCLvPK/SEZqffY5crh+T/1wjKBwYNi0B0WuKgjJ8gLFe85ZGraMPfMzGC3wWHfivXbVjkQFQgSWNQNaF28Xr0HthxmAnMk9awDGasD/yMKdzoxeExGWe312XUEOxdrz2ZQcmsXMQlExdM1WEjZw4/mTIutHM9NyDnRliXYZBuVhozEo40hUghhaqbpM4EQRKMrkaNNnIU+6Uvj3SGVY2oMexzLW1fz004a9DsWKzy5JQeXXEuJxcvrBz09TYDF1FprPJASMD9bg/1c7KT33hL584W0+N7zcnywlRgxZvXbkA21eLfvIjj+4yv5+f5/ANfYFuICmVuZHN0cmVhbQplbmRvYmoKNDQgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNzQgPj4Kc3RyZWFtCnicTZBJDkMhDEP3nMIXqIQzwOc8v6q6aO+/rUMHdYH85CBwPDzQcSQudGTojI4rmxzjwLMgY+LROP/JuD7EMUHdoi1Yl3bH2cwSc8IyMQK2RsnZPKLAD8dcCBJklx++wCAiXY/5VvNZk/TPtzvdj7q0Zl89osCJ7AjFsAFXgP26x4FLwvle0+SXKiVjE4fygeoiUjY7oRC1VOxyqoqz3ZsrcBX0/NFD7u0FtSM83wplbmRzdHJlYW0KZW5kb2JqCjQ1IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggODkgPj4Kc3RyZWFtCnicNYy7DYAwDER7T+ER4r/ZByEK2L/FSXBj392TXlLiQOU6EY6mgSdB9ZleINnpAVZF4lFJzP9NvalFU8+m7atNBCczjvV1HKia03rQWihtkxbecH0AnB3tCmVuZHN0cmVhbQplbmRvYmoKNDYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNDEgPj4Kc3RyZWFtCnicPY/BDsMwCEPv+Qr/QKTYKaF8T6dqh+7/ryNLuwt6AmOMhdDQG6qaw4Zgm+PF0iVUa/gUxUAlN8iZYA6lpNIdR5F6YjgYXB60G47isej6EbuSZn3QxkK6JWiAe6xTadymcRPEHTUF6inqnKO8ELmfqWfYNJLdNLOSc7gNv3vPU9f/p6u8y/kFvXcu/gplbmRzdHJlYW0KZW5kb2JqCjQ3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjE1ID4+CnN0cmVhbQp4nDVROQ4DIQzs9xX+QCSML3hPoijN/r/NjNFWHsFchrSUIZnyUpOoIeVTPnqZLpy63NfMajTnlrQtc4C4trwvrZLAiWaIg8FpmLgBmjwBQ9fRqFFDFx7Q1KVTKLDcBD6Kt24P3WO1gZe2IeeJIGIoGSxBzalFExZtzyekNb9eixvel+3dyFOlxpYYgQYBVjgc1+jX8JU9TybRdBUy1Ks1yxgJE0UiPPmOptUT61o00jIS1MYRrGoDvDv9ME4AABNxywJkn0qUs+TEb7H0swZX+v4Bn0dUlgplbmRzdHJlYW0KZW5kb2JqCjE1IDAgb2JqCjw8IC9CYXNlRm9udCAvRGVqYVZ1U2FucyAvQ2hhclByb2NzIDE2IDAgUgovRW5jb2RpbmcgPDwKL0RpZmZlcmVuY2VzIFsgMzIgL3NwYWNlIDQ2IC9wZXJpb2QgNDggL3plcm8gL29uZSAvdHdvIC90aHJlZSAvZm91ciAvZml2ZSAvc2l4IC9zZXZlbgovZWlnaHQgNjcgL0MgNzEgL0cgOTcgL2EgL2IgMTAwIC9kIC9lIDEwMyAvZyAvaCAvaSAxMDggL2wgL20gL24gL28gMTE0IC9yCi9zIC90IC91IDExOSAvdyAxMjEgL3kgXQovVHlwZSAvRW5jb2RpbmcgPj4KL0ZpcnN0Q2hhciAwIC9Gb250QkJveCBbIC0xMDIxIC00NjMgMTc5NCAxMjMzIF0gL0ZvbnREZXNjcmlwdG9yIDE0IDAgUgovRm9udE1hdHJpeCBbIDAuMDAxIDAgMCAwLjAwMSAwIDAgXSAvTGFzdENoYXIgMjU1IC9OYW1lIC9EZWphVnVTYW5zCi9TdWJ0eXBlIC9UeXBlMyAvVHlwZSAvRm9udCAvV2lkdGhzIDEzIDAgUiA+PgplbmRvYmoKMTQgMCBvYmoKPDwgL0FzY2VudCA5MjkgL0NhcEhlaWdodCAwIC9EZXNjZW50IC0yMzYgL0ZsYWdzIDMyCi9Gb250QkJveCBbIC0xMDIxIC00NjMgMTc5NCAxMjMzIF0gL0ZvbnROYW1lIC9EZWphVnVTYW5zIC9JdGFsaWNBbmdsZSAwCi9NYXhXaWR0aCAxMzQyIC9TdGVtViAwIC9UeXBlIC9Gb250RGVzY3JpcHRvciAvWEhlaWdodCAwID4+CmVuZG9iagoxMyAwIG9iagpbIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwCjYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgMzE4IDQwMSA0NjAgODM4IDYzNgo5NTAgNzgwIDI3NSAzOTAgMzkwIDUwMCA4MzggMzE4IDM2MSAzMTggMzM3IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYKNjM2IDYzNiAzMzcgMzM3IDgzOCA4MzggODM4IDUzMSAxMDAwIDY4NCA2ODYgNjk4IDc3MCA2MzIgNTc1IDc3NSA3NTIgMjk1CjI5NSA2NTYgNTU3IDg2MyA3NDggNzg3IDYwMyA3ODcgNjk1IDYzNSA2MTEgNzMyIDY4NCA5ODkgNjg1IDYxMSA2ODUgMzkwIDMzNwozOTAgODM4IDUwMCA1MDAgNjEzIDYzNSA1NTAgNjM1IDYxNSAzNTIgNjM1IDYzNCAyNzggMjc4IDU3OSAyNzggOTc0IDYzNCA2MTIKNjM1IDYzNSA0MTEgNTIxIDM5MiA2MzQgNTkyIDgxOCA1OTIgNTkyIDUyNSA2MzYgMzM3IDYzNiA4MzggNjAwIDYzNiA2MDAgMzE4CjM1MiA1MTggMTAwMCA1MDAgNTAwIDUwMCAxMzQyIDYzNSA0MDAgMTA3MCA2MDAgNjg1IDYwMCA2MDAgMzE4IDMxOCA1MTggNTE4CjU5MCA1MDAgMTAwMCA1MDAgMTAwMCA1MjEgNDAwIDEwMjMgNjAwIDUyNSA2MTEgMzE4IDQwMSA2MzYgNjM2IDYzNiA2MzYgMzM3CjUwMCA1MDAgMTAwMCA0NzEgNjEyIDgzOCAzNjEgMTAwMCA1MDAgNTAwIDgzOCA0MDEgNDAxIDUwMCA2MzYgNjM2IDMxOCA1MDAKNDAxIDQ3MSA2MTIgOTY5IDk2OSA5NjkgNTMxIDY4NCA2ODQgNjg0IDY4NCA2ODQgNjg0IDk3NCA2OTggNjMyIDYzMiA2MzIgNjMyCjI5NSAyOTUgMjk1IDI5NSA3NzUgNzQ4IDc4NyA3ODcgNzg3IDc4NyA3ODcgODM4IDc4NyA3MzIgNzMyIDczMiA3MzIgNjExIDYwNQo2MzAgNjEzIDYxMyA2MTMgNjEzIDYxMyA2MTMgOTgyIDU1MCA2MTUgNjE1IDYxNSA2MTUgMjc4IDI3OCAyNzggMjc4IDYxMiA2MzQKNjEyIDYxMiA2MTIgNjEyIDYxMiA4MzggNjEyIDYzNCA2MzQgNjM0IDYzNCA1OTIgNjM1IDU5MiBdCmVuZG9iagoxNiAwIG9iago8PCAvQyAxNyAwIFIgL0cgMTggMCBSIC9hIDE5IDAgUiAvYiAyMCAwIFIgL2QgMjEgMCBSIC9lIDIyIDAgUgovZWlnaHQgMjMgMCBSIC9maXZlIDI0IDAgUiAvZm91ciAyNSAwIFIgL2cgMjYgMCBSIC9oIDI3IDAgUiAvaSAyOCAwIFIKL2wgMjkgMCBSIC9tIDMwIDAgUiAvbiAzMiAwIFIgL28gMzMgMCBSIC9vbmUgMzQgMCBSIC9wZXJpb2QgMzUgMCBSCi9yIDM2IDAgUiAvcyAzNyAwIFIgL3NldmVuIDM4IDAgUiAvc2l4IDM5IDAgUiAvc3BhY2UgNDAgMCBSIC90IDQxIDAgUgovdGhyZWUgNDIgMCBSIC90d28gNDMgMCBSIC91IDQ0IDAgUiAvdyA0NSAwIFIgL3kgNDYgMCBSIC96ZXJvIDQ3IDAgUiA+PgplbmRvYmoKMyAwIG9iago8PCAvRjEgMTUgMCBSID4+CmVuZG9iago0IDAgb2JqCjw8IC9BMSA8PCAvQ0EgMCAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+Ci9BMiA8PCAvQ0EgMSAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+Ci9BMyA8PCAvQ0EgMSAvVHlwZSAvRXh0R1N0YXRlIC9jYSAwLjc1ID4+Ci9BNCA8PCAvQ0EgMSAvVHlwZSAvRXh0R1N0YXRlIC9jYSAwLjUgPj4gPj4KZW5kb2JqCjUgMCBvYmoKPDwgPj4KZW5kb2JqCjYgMCBvYmoKPDwgPj4KZW5kb2JqCjcgMCBvYmoKPDwgL0YxLURlamFWdVNhbnMtbWludXMgMzEgMCBSID4+CmVuZG9iagoyIDAgb2JqCjw8IC9Db3VudCAxIC9LaWRzIFsgMTEgMCBSIF0gL1R5cGUgL1BhZ2VzID4+CmVuZG9iago0OCAwIG9iago8PCAvQ3JlYXRpb25EYXRlIChEOjIwMjEwOTE2MTQzNjE1KzAyJzAwJykKL0NyZWF0b3IgKE1hdHBsb3RsaWIgdjMuNC4zLCBodHRwczovL21hdHBsb3RsaWIub3JnKQovUHJvZHVjZXIgKE1hdHBsb3RsaWIgcGRmIGJhY2tlbmQgdjMuNC4zKSA+PgplbmRvYmoKeHJlZgowIDQ5CjAwMDAwMDAwMDAgNjU1MzUgZiAKMDAwMDAwMDAxNiAwMDAwMCBuIAowMDAwMDE3ODMxIDAwMDAwIG4gCjAwMDAwMTc1MjYgMDAwMDAgbiAKMDAwMDAxNzU1OCAwMDAwMCBuIAowMDAwMDE3NzQwIDAwMDAwIG4gCjAwMDAwMTc3NjEgMDAwMDAgbiAKMDAwMDAxNzc4MiAwMDAwMCBuIAowMDAwMDAwMDY1IDAwMDAwIG4gCjAwMDAwMDAzOTkgMDAwMDAgbiAKMDAwMDAwNzE4NyAwMDAwMCBuIAowMDAwMDAwMjA4IDAwMDAwIG4gCjAwMDAwMDcxNjYgMDAwMDAgbiAKMDAwMDAxNjExNSAwMDAwMCBuIAowMDAwMDE1OTE1IDAwMDAwIG4gCjAwMDAwMTU0NTQgMDAwMDAgbiAKMDAwMDAxNzE2OCAwMDAwMCBuIAowMDAwMDA3MjA3IDAwMDAwIG4gCjAwMDAwMDc1MTUgMDAwMDAgbiAKMDAwMDAwNzgzNSAwMDAwMCBuIAowMDAwMDA4MjE1IDAwMDAwIG4gCjAwMDAwMDg1MzIgMDAwMDAgbiAKMDAwMDAwODgzNiAwMDAwMCBuIAowMDAwMDA5MTU4IDAwMDAwIG4gCjAwMDAwMDk2MjYgMDAwMDAgbiAKMDAwMDAwOTk0OCAwMDAwMCBuIAowMDAwMDEwMTE0IDAwMDAwIG4gCjAwMDAwMTA1MjggMDAwMDAgbiAKMDAwMDAxMDc2NSAwMDAwMCBuIAowMDAwMDEwOTA5IDAwMDAwIG4gCjAwMDAwMTEwMjggMDAwMDAgbiAKMDAwMDAxMTM1OSAwMDAwMCBuIAowMDAwMDExNTMxIDAwMDAwIG4gCjAwMDAwMTE3NjcgMDAwMDAgbiAKMDAwMDAxMjA1OCAwMDAwMCBuIAowMDAwMDEyMjEzIDAwMDAwIG4gCjAwMDAwMTIzMzYgMDAwMDAgbiAKMDAwMDAxMjU2OSAwMDAwMCBuIAowMDAwMDEyOTc2IDAwMDAwIG4gCjAwMDAwMTMxMTggMDAwMDAgbiAKMDAwMDAxMzUxMSAwMDAwMCBuIAowMDAwMDEzNjAxIDAwMDAwIG4gCjAwMDAwMTM4MDcgMDAwMDAgbiAKMDAwMDAxNDIyMCAwMDAwMCBuIAowMDAwMDE0NTQ0IDAwMDAwIG4gCjAwMDAwMTQ3OTEgMDAwMDAgbiAKMDAwMDAxNDk1MiAwMDAwMCBuIAowMDAwMDE1MTY2IDAwMDAwIG4gCjAwMDAwMTc4OTEgMDAwMDAgbiAKdHJhaWxlcgo8PCAvSW5mbyA0OCAwIFIgL1Jvb3QgMSAwIFIgL1NpemUgNDkgPj4Kc3RhcnR4cmVmCjE4MDQ4CiUlRU9GCg==\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 2.0582756996154785\n", "Layer 2 - Variance: 13.489118576049805\n", "Layer 4 - Variance: 22.100566864013672\n", "Layer 6 - Variance: 36.209571838378906\n", "Layer 8 - Variance: 14.831439018249512\n"]}], "source": ["def const_init(model, fill=0.0):\n", " for name, param in model.named_parameters():\n", " param.data.fill_(fill)\n", "\n", "\n", "const_init(model, fill=0.005)\n", "visualize_gradients(model)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "ef67c266", "metadata": {"papermill": {"duration": 0.049718, "end_time": "2021-09-16T12:36:23.634899", "exception": false, "start_time": "2021-09-16T12:36:23.585181", "status": "completed"}, "tags": []}, "source": ["As we can see, only the first and the last layer have diverse gradient distributions while the other three layers have the same gradient for all weights (note that this value is unequal 0, but often very close to it).\n", "Having the same gradient for parameters that have been initialized with the same values means that we will always have the same value for those parameters.\n", "This would make our layer useless and reduce our effective number of parameters to 1.\n", "Thus, we cannot use a constant initialization to train our networks."]}, {"cell_type": "markdown", "id": "aa79959c", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.042907, "end_time": "2021-09-16T12:36:23.720988", "exception": false, "start_time": "2021-09-16T12:36:23.678081", "status": "completed"}, "tags": []}, "source": ["### Constant variance\n", "\n", "From the experiment above, we have seen that a constant value is not working.\n", "So instead, how about we initialize the parameters by randomly sampling from a distribution like a Gaussian?\n", "The most intuitive way would be to choose one variance that is used for all layers in the network.\n", "Let's implement it below, and visualize the activation distribution across layers."]}, {"cell_type": "code", "execution_count": 14, "id": "e5dc3377", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:23.810870Z", "iopub.status.busy": "2021-09-16T12:36:23.810401Z", "iopub.status.idle": "2021-09-16T12:36:31.446937Z", "shell.execute_reply": "2021-09-16T12:36:31.446513Z"}, "papermill": {"duration": 7.683221, "end_time": "2021-09-16T12:36:31.447053", "exception": false, "start_time": "2021-09-16T12:36:23.763832", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 0.07831248641014099\n", "Layer 2 - Variance: 0.004064005799591541\n", "Layer 4 - Variance: 0.00022317888215184212\n", "Layer 6 - Variance: 0.00011556116805877537\n", "Layer 8 - Variance: 8.162161248037592e-05\n"]}], "source": ["def var_init(model, std=0.01):\n", " for name, param in model.named_parameters():\n", " param.data.normal_(mean=0.0, std=std)\n", "\n", "\n", "var_init(model, std=0.01)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "8959f829", "metadata": {"papermill": {"duration": 0.05122, "end_time": "2021-09-16T12:36:31.548759", "exception": false, "start_time": "2021-09-16T12:36:31.497539", "status": "completed"}, "tags": []}, "source": ["The variance of the activation becomes smaller and smaller across layers, and almost vanishes in the last layer.\n", "Alternatively, we could use a higher standard deviation:"]}, {"cell_type": "code", "execution_count": 15, "id": "025eae6f", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:31.649936Z", "iopub.status.busy": "2021-09-16T12:36:31.649437Z", "iopub.status.idle": "2021-09-16T12:36:39.172380Z", "shell.execute_reply": "2021-09-16T12:36:39.171893Z"}, "papermill": {"duration": 7.574836, "end_time": "2021-09-16T12:36:39.172497", "exception": false, "start_time": "2021-09-16T12:36:31.597661", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9Bbm5vdHMgMTAgMCBSIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDkwMS4wMjUgMjE2LjY2NTYyNSBdIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovVHlwZSAvUGFnZSA+PgplbmRvYmoKOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEyIDAgUiA+PgpzdHJlYW0KeJzVnc2SJMl1nff1FLkkFwj49X9fAgYJZjRtQMKkhUyLETiUBtYNGjAAaHx7neMeGXHD81bNVHd2t5xjA1adyfKM+4WHh/+eK7c/vvzyV3L7Pz/e8D83d/sj/v0P/Pxb/v7i8NvHl+Zkcz7h5w/Hz17ylnPK+PEDPnb59f++vPzbi9ualByLS7Xe5l8iimnZlXr7C7/0tw8fOH55mT798pLyVvE10W9tfOHHF6lhyzFI9Ur+oGUpdat3fS/hovVr/vPtoXDxUGT/B+WEtLXbX76//Y/bn26//JUf4P4J//4R/3ZwL7/8zfd//+EP3//zb399+8OPLzVtPqLMfLniU71cxcu/vPzu9ud7wW6ThJtyL7v/+ttdffnzi4DcLxz+U0GYodYqkl27+bSJY3F/+Pjy69/ffvlf5SZy+/2/vbQNd6m0XAtv5e//9eV/3v5B3D/e/tft9//08l9+j+jd5oSlOvXTHz6yiF/85vs/fvff//Yv3/3px198/OFPf/vx9pt/v/3u5Xf9cp8PTaRtrVQEdb3Pp/wEbMC1xV5aSE5e5+YOWgrWF4s8xa24mrxcIz/lZ0QeyyjNpdD8z4lcLqHrupc3X2spNbomt7YVXYi/FvKrf7wBefP46sD/u/3DH/76w9+/++sP//6n29+/+/Djl6f789uOTwPrfS8/1K3U1EJ5szJt7ivUp73EitvkU/HlGvEpf2bE1W8uZheayzW/HXT6akGLb1usqfp6jVrpnxm24NEJLkt0OZTwZtzy9W62FL+ViPo3hX3Inxt1rlvK2UePBqS9HXW6thss5RcsDw1BYL2pABPk2vZMzcZvvv/Tjz/89T8/E18YgfpWUUu9pIKeRRx9DB/wQtzyvZcRe6dk89mH6mNMO+vLH9/0H79Mf/zykv2Waos5XJqaHLfaEt77l5bmqqLMlqleSrir6KX8+v//4I0wEXzZvMO9DtfgL+oRvIlkjeCNMBF82xJaRpnu/EVtvuXygOSurhG8EebHlyK4Po+XwyV4pQa8CdERmZAodYngrTARfNxEpo7sh0m93+NrCUvdeStMBI8Oto9S4jX4U8U9jignXpEodY3gjTARPDqBvjW8IC/BnyrCbJ797xnJXV0jeCNMjG9lcyGVeL3zSg2b5BZjnpEc6hLBW2Ei+LCFKMlf77xSQ+80pgckh7pG8EaYfWYjxxJkuvOnGtDCJzTtM5JDXSN4I0wEX7aWgmvTnT/VsNXixtN9QXKoawRvhMlpws3jL+r1zis1opErUuSKRKlLBG+FieDRX88x5+udV2raYsJdnpAodY3gjTARfGQFjildgz/VsoXsvMtXJEpdI3gjTARfNikYDF9jv4sJY+AmIVx4KHGNwB9DRNycRPHOTzf9UBNatRpbzhccSlwjdCPKjy/iZCu1VLnEfqqpbaGkMir8SUSpSwRvhcngI/qoIbU0RX/IGMKGUCTlCxStLhK+ESjjz1toLdQp/LuK/qzkyF7NBYqWFwn/MU5Gj+66S5Lnu3/IaOE93u2hXahodZHwjUC5BoUeu7i58h8q/gdNnLgyUbnoawAwImX86LRLLiFPAA65T69LdeXKRamLhG8EyvjRb/f+ukD3QcsSK6q65BAnLlpfBIERKxGg9+4rRqsTgkNGQ4dXnautTGi0vggCI1Yg8Cgj9DXPC4JTxogGNx6Pe57QaH0NBFasRICefGi1TLXglKU49HhTKX5Go/RFEBixEgG69DHlPNeCQ5aM1x7nd8qERuuLIDBiJQJ07ZNwzu6K4JAl5i1hyDf6AwqN1hdBYMRKBHizpeLDXAsOmX/ICe0YJzRaXwSBESsQBHTyc3BSrghOGS0e6n4Zg16NRutrILBiJQL09HOtbtqDdsqVwz+PkeCVjJYXAWBESgDo6hdOZ00ADrngJzR5Va5ctLwIACNSAqhbRhFlrgGHnNOGBrAlf+Wi5UUAGJECQHRbq5l/fwFwyqlsJSVXr1iUukb4VpwMP6AMNOfT/T/l5DEajL7NVJS8CAAjUgJIW2qF28muAE4Zw+HcUopXLlpeBIARKQGgm+9CkrkGnHLechBEPXFR8iIAjEi5TRVluIZirgCUjLYuJDT3ExclrwHAipQA0MXHf6thAnDKkYs/EtvERcmLADAiHRuVq3c1zzXglAOKktj3nOpClLwIACNSAkD33uec5hpwygEv/uBzeeByyIsAMCIlAHTug4/z/LiS0fCj69eXfK9cDnkRAEakAJDRt0ehfqoBSg6blBZqe+ByyGsAsCIlAPTtY3RuqgFKDpurJXp54HLIiwAwIiUA9O1jw72cAJwyIpWCQh+53OVFABiREgA69wnv8zYBOGVG6usYCk1cdnkRAEakAFDQu88Y4123BUxyDc7HiYuS1wBgRUoAbM0LOzRXABe5coevwWWXFwFgREoA6N0X38JcA16RTVyLADBCIoCy1YLX+VwDrnIVEYPLLn8qgOthmS3d/uPl1YiuOP75t7froZr5MEffuPXxBWG4Etx4bbvU+mGiGnk4Fp06itElDPHGTr+C4d3o5JaYfd/gWnkUTmoeuwKrON/VxqjS2C6Xs299qazh8RgDRYyT8FU5jJ1Uwp0FmcNH51zuiwktbjGFgMvBqBpj7FC7mrdSXLklwd9jrB3GphwMPULN3JXjWgy5ju0qrlbHaZm0BbIuQ825REm8ERiql7G66xzuShYnt+z6qzvJvuWj1Vg4tYWAS0K97nLgBcUSbmjvvJPc7lsk0OChH8SDrB5twCgjbTG7CoCln9aK+zcWFFi95FsL5FPyfZ29BdxFVATnN49b68aSZNx4kwJ0jLWSa86N9biM0mvYV2iyz2h093U6F0EqQkeFzHgv98kq4QVnNEq4aLD2RfIul5pKCH1Ni6u8o0sjjehcGct9DaWP7QA8vxpjRN3nYazQ8FPe18aSDyX78fmG+zoWjGTLfCxrLx/3yA1kHjFm9BWln8XKpcp96a1GkcCrR4ABD9nQA8psfEpwm4JDTR/fisqUauKiJMZhLRbw6Tpex7iOhs+jEkplJeeEHSphksaVbNyyGkIejAMA1ta48wPVIeM7h4xaHWtGjaj4+oixbd2nPVFNQs6sBMX1ykOZUwF4YjyrDP5zGlUmgDZuMi4A3eRavJewz51VB7ylV9KY8sAYub5WC35BPW9AfEw05VYiT3iiEN83olFGEClL7k8KyPrRGeHZ2OQrn1Y8t63cZdTC5hrnKiLuaMYD3XvvgiupfjyvqIZoUbqMINBiyHjkuQGiXzcenhhR2cdoDz396PcesCuF66F9CJD8WDfPBY+yS3gEercouTFxOjegdTx0tnw96OfZIm5oPUcj/5OHY62T8ijRPDP78dXz9viLdx28ffzWN8t2COrnnmBsCfcct5H/4OFBe72X9Prxw//23X9+/xf8ps8f7pYJ7/E4GMYKs9fBg03C1evAh75fYt5yinYLjyXeAtdFVyVfGKlSHj0PWDuOD/CfT33Jvlh1yuOxRa1P6bpeqOQnHGf3eGRTLOAY8I56nwGCT08wQPhy/NBWINwS0vVJU/Iz+DUOP1AaeiPpDQOJN4wQviCB2N9SMhM45ScQQFO8l4buxxs1yJn15mKI4LnFOUQffGPL/VxHhC/4oL6rpfnEeobuorr6T3ZJ+CoUKq7Ve3TxrxRO+UkUSkanGIOvhK5IfBuE/yYg2CdGD5Pb9y4klP4kFNwrGVtlz9Kji/Qmi/htWKC76UJEh35icerPYoHhCuqEr+ifopP4Jov8msdCC+PEySjyM7wWPhEupyF+7rj7p6ch3pzDQF8JgwIMfdq0L8fjl4jB47QvZ5IxJkhDvhRyl49piDVAGBETROTkQ5yOok9ycf3c9YzHXU9jr4HBiJcYMMRPHmPKCcMrsgltMRBGaATRtpAq/myK+CKfD8alkEUfDCNiDkZkw3C+xunWX+WGt7488rnLa4GwIiaIwLmq7KcaoWSacvjapz50IUpeDIQRMUHkzZcUZa4Rp8wJmVpKfuRzlxcDYURMEHXDsEjaXCNOOXDqysfwyOcuLwbCiBggMrpjtc8SX0AomXPXEqo88DnktUBYERNE4ApEmQ66a5kzmxhHtwc+h7wYCCNigkgcrHDT4hXEKYcNamrywOeQFwNhREwQZasuhjDXiFOOeD2Usa9FF6LkxUAYEQNEwdhBnExWAFpGZzKX2BcUrnwOeS0QVsQEgcGDcJp9AnHKaUPHYazF6UKUvBgII2KCwJ31tDSaQJxy3ppELvxd+Sh5MRBGxASB0QP+W5lrxClzATeOBUhdiJIXA2FETBAYPeA9mOcacchcuUShfflTFaLlxUAYEQNExeghNHc9Rn+qXCktOY9ehKKj5LUwGPGSApfGYw0zhkNGvyHWEqRc6Gh1MQxGwOSAkUNyeZqjO9QiW2p17GZQRWh5MQyP8ZJC7SdiHx6KQ2boLbi+/ecsQ6uLYTACBoe+j8RPG8dOteXNtZT6VhhVhJbXwmDESwp0QyytTLXhlLle4jxeC/6C5yIvBsIImSQwZihhMho4Ve5ACtnlvtNN89H6YiAeIyYHbn7ry8pXEIdMi5GG2GK5AtLyYiCMkLn/gdsNU5iO3CuZm+p92s8TaUJaXwqFGfPYChKaiEx7gU6ZWwpdk9ZnHDQirS+GwoiZKCI+yc9OKA75suvygkjpi6EwYiaKwil5JmO4ojhk7m/FqLv2NkEj0vpiKIyYiQKjB4d45lpxyNzEiQD7oEITUvJiIIyIAUIwfEDLl6Y6ccp8YZZQXZ74KHktEFbEBIEOs3cuTjXilKX/yNxAEyCtL4bCiJkoegaY6uc6cciNpwDKeF+qQrS8GAgjYoLAOAL/bTq6r2SmuBEugV/5aHkxEEbEAOExjggltKuJg5J50iFK7FvgVCFaXguEFTFB9OMq8/50JfPkSMjDzE4VouXFQBgREwTGEbEyG8UVxCGnvAl+6AsYqhAtLwbCiJggeDQkljTXiENGjwF/l7Kf+Ch5MRBGxAARuOvQpTDVCCVXmtvH5q98tLwWCCtiguAMQw5+qhFKzlsqrfUFDF2IkhcDYURMEBhAFC9urhGnzJOLpZUy8VHyYiCMiAki86hmnY7+axmx4WFIZeKj5MVAGBETRNt8RR+xTiBOmZ1JRFgmPkpeDIQRMU+qyGMyuA8XOWyhZXnEc1fXwmDFSwzByn6nZbVV6kpnzR1UZsQEka1MeFpGD0riOHl85XPIi4EwIiaIamXF0zK3U+Kv5YHPIS8GwogYIHjW4zFD3iTfdxpPhSy5AdmMmCCClS1vkquMd+e1jF1dDIMRLzEkK2/eJO9nMqYynnRSQx+Leun+GbefCeXRP2M+UyM0w+fe4kLvN8++D3cLt5T7nhcMnyW2xhPDeAOO/bQ5bjGXMCyicP1l30JHY/nsXN9UWUqSWveddSHHbs0R6a83PtyNFWqpvc9ZUondSsAX/GWg3QjzEaEjlns18iVvLsXu2dE2mnO2uss5hMTJo4ahTObWnLFPp+BBlD4CDL6Jv+9bSfTykVtCAJLptcElawCrtfnuhyqSWj+pz30NztHx40ZDqEabEaqRaUNaTfRLwIWk0i+kpo27aYvcCm8pfr6vhmNYUfD1TRBAd7ig3Lh7iPYdrXDTfu4vEVwnvr7QgYF2GzSoH2vqzfPsGfjQmII2Cvm+4MpfaLLBs5s+jlkvHh92GfWi3LqZWW5poG3oy4ABHlAuKjja/OV9CQok8Hrv7hWtttJ9GqijcjaadRRUMNojjCl5VEB8F1d0cRMjoitxX78Rj9dA7es3rgXxu8xsanlf7qnVSX+TusCuVaC3R6FHiE+l7XpFNUKHsxdfJYe2r4nQqmN4frSAzum4es5qOMeFdlpVcHNa2/WCi8/DZaOhuu+Xw7Q3PccXve5SDC2OmUU2X7Ggh0NnDdxjH8bUIupsc5XZEhA4ak/u9uBBOJ3QEh6Xxucilfv8JN6E3ncb5crnJuyzUa6issiNfiX04eivDi+bR13GABRVDDWiJyHAMIyWS0zKhNuXhDYcXUbNC336JmVuZw9R9mkNnnBgliNPcKlv4OUg3wUewGd2K9zoUamDrxjfJZf6c4TGI+6fJtHg+DCiQqLx7DUmBMFwSGj9w7c997R0lV46Tfr+UFSqeAwt0RD0edxI8nXcZdoqo+KzrQPInNJQK1oKBNrbmhLieG4DWiY8IX19AN+B+zCmsOjw0fLocKU0bguuk4+s7GYcTgaNqU2W8Ry+Ir/DjOOV89ivWTmgZPOo9sdXTSHwF+8+821/+5vf8R5zDrbFcX9Vcan+57tz+G/izhGFLjMPu34TG8fH3CBKvkBSpdjuHOcH3PWrPvtkc2RaEno9XSdzlPwEb4XIFFBF6PvEV8G73DmSe4I7xxfkV9AG4b0+nYhQ8jP4cZsGSws8YfYp5hxfDkBCgy0Or4DJNuiUnwAg8UXE0ugmVX4OgeReMeeIuLCw17HnOnN8yYf0Pa3MJ/qfoAOrrv6TnTm+CgUek6joyF8hHOqTGKBnGvgD+l/oor2JQb4JBhrdoRVsYaoNSn8SCowZKvpOaNYCelvvsyj5OijwWKfAUyJXEof8JBDM14OuXXTozqEf+yaJ8G1I5D7sw9BsQnHqz2JBlza0Egmj5/QTBiXxFYOSwPHWWeRnGJR8ItyvOtvDbpWRNSXSFPIxa8okl5zLkC+F3OVjvmcNEEbEAEFXy8f8KZN8gjD5rAXCipggopVHZZIxTrbwiCyIwYiXGLKVTUXLnDFrtW9C04UoeTEQRsQE0aysKlrmBKrLLj7yucuLgTAiBgi+0B+zq2g5bMJxR3ngc8hrgbAiJohgZVnRcmA+mdi3HF35HPJiIIyICSJb2Va03OdoS58xvvI55MVAGBETRLWyrmgZsUtO3Vb9yueQFwNhRAwQ0VnZV7QcNzQNVfzER8lrgbAiJohgZWHRcto8xrhlwnOqi2Ew4iWGZOVi0XKmv7z0BT5diJIXA2FETBDFysmiZT4Labi660KUvBgII2KCwMjhMTeLkulslsYaripDqYthMOIFhuStlCNKTmkrUuroQyg6Sl4LhBUxQUQrJYuSs3C1svRFc1WIlhcDYURMENnK2KLk3LbiSpnxnOpiGIx4iaFtUmNyc304ZO7rwDCr71dRhWh5MRBGxACBCh6b83WqD6fcuvGjdONHVYiW1wJhRUwQYSv82FQjTlkczTi4w+cK6KIvhsKImShyPx6cJxJ3teegCXXsdNKAtL4YiMeIyYE5Y0qevBiUzNw1vqSxIUcB0vJiIIyQudqNoYOEyT33VLlQkGPxfTub5qP1tUAYEZMD9wXx7ycQh8zdbpFrJuUKSMuLgTBCJgluTpwxDO3itKDZmA4Ma0CYoyUBjBoCPjJDOGT6kOSShiXLAUeLi0EwAiYH7vPL89LWofaEWDmOvYKajtYXA/EYMThUz42XIU314ZSlu6Eys9wVkJbXAmGFTBLcG0UfwyuIu8ol9lpKGLMxio/WFwPxGDE54AITCpg6UqcsiM3jW/qBegVIy4uBMEImCeYC48LlROKQW2YmxtG1PstQ6mIYjHiBoWHQkFNsU4U45cpd+aWORb6zEC2vBcKKmCB4lkN8nerDKRfZAmJvEx8tLwbCiJggMk30XZ5rxCHz/IGMQxKqDKUuhsGIlxjqhjJrmuvDIfMUTcmpZ01XhWh5MRBGxB9fkuvnhuZ9+UpOfsu4+H5qQvNR8lIgzIgJAr+0GP20wV7JbXPBh3DFo9TFMBjxEgNGDegaTp4LWs6cc6kSJjpKXgyEETFBYODgUmt+AnHK/Lsaet5fXYiSFwNhREwQGDeIlDLXiFPGABNto7SJj5IXA2FEDBCCgYOUNBlKa5kuExLbhOdU18JgxUsMkXnF6aVxxXDKgWu7fXR5hXNXF8NgxEsMTCjdJMy14ZR5NDS7fnz4SueQFwNhREwQ7TFF4oeLTPMudKHrA59DXgyEETFAeDpyPSSF1LLaOnnls+aOSjNigghWfkMtcy2POesf+dzlxUAYERNEtvI/TvJ9l/lUyJKbz82ICaJaaSEnue6uI1MhdTYjWQOEETFPzzgrLeQknzWiXtNmrlgjrIgJgr88pIWc5AOEzedzQTzTkGQ+QxMk3nfZ5+LH68610vp7P9L8GP/TPUZ8pjX0voHWSwvDliknPAJjg3HZQkHRPfsbBp1jKB7RbKJDgSvDiMMV9KzivtdQisTcjQ0qTyz0GQw6wQXHNbW8pRbCuAz8QtOfsU1P8N2+7tvTPN2qU/dXiM2NzTmRg/+UOPinF8Wwloj4WcCn3lI3PREZG1gEvRwUkrthZ/Fo4LvMLBLMI3HrSfxakrJv+ympj5kyEUQ3puFS4lF5PPzF0RC2dWeJmCodtFEhmTedDinddoE7RQLNHeptTOeWONaG6SfRhHPfrnY/j+69Hbkq5or0VQD6kdQwltpxv6JUGnQk+hL4FO8Lq9G5fviYvgQe3IeOapNo4QAdX1vLKKZ4vNwrk6/T5yOkgBq76wW407AFwb/i7uu5HnWmr1iK+HHklGpFXY+7XvNY34wbyLQ6TPhxP9p9LdThi3zrRifoYvruikEd15H7FVauCI4RWSx5KyUwDFqOxNbc2JtY6kbnjO5VTmsR/HhfPsrFOdRv4W41VJJxp3ETKi4YFUAi0/BFX3q9qGVDpa2olUyU0Mr943QDqTy0jw9IQz94zLD1NGW0XUGkSdyYkG5hS3hO8EtmsgU/rKEjfaJR50KgLW5oZcfVmE3Y0SwFF14QQSj7nJVHs4CnCncUz1/3l0G89JoueKjoOCK4kv7qcTxknmiggnrtSi7dRJLzHD7jIeoPChPUlzHqj/QwqWP6A4Rykn0yoESXxmRAtx/oKupHa9L6bmg69OxTKKiVCSGMlH14YtoYOOERS6jU/aFvaBTiPpzKjc1xP2tQJLiwjy5AiUcmaW/kg4zrQM8KTWl3PsFzh5rfH6D+VkEjU3cfkVxGz3xqenEF+XX5Ha4jr5w1f82iAiWbx9A/vmp2gb9493l2+9vf/I73uI6wVWzvchuJ38RtJFV6Kz1s4y20In9MCqPkCxxViu02cnyA/8RPej+/ckw5C5uxeTf2qT7BKiLjcUbjjubJ5/SGWcYX8xr5gvTwci7S6jTlp+Rn8MNL3/fS8Hb1n2Q28gUJtMyeZWvXHQVKfgYBXH4vDa9hdEQ+x2wk43XReq6h7pj0XL+RL4f5fe3MJ1q68EzJefU/22/kK9Y2hSF3ZzfudL9gOOUnYUAPCSNIPHoBHeifIJG+CYmGzn2KTOt+IXHKTyJRaRkVOX7hzum3zVe+TZ3A2AnDLbokXlEo/UksMO7AO40rmI2DkbdhfJtqQTtINHM1TywO+VkoAvrXOWCEVX35KSeab1Qv+qGgwpHrlcWpPwsGR04Y3PD0bvuJhtOnVxxYMGbcmnruPt2B5RPpft0JLoxGjdw/CaNRI/fPJLdezIe5kLt8THCtAcKImCCalfvnVdnEthgIIzT0KWkt+5j7R8vcuZ7ounLlo+SlQJgREwRNcB9y/2g5sMlofSbkyueQFwNhREwQ2cr9o2UuDdbajV6vfA55MRBGxATRrNw/WmYCXoyj4gOfQ14MhBExJy7Eyv2j5cCJ/GjwOeS1QFgRE0Swcv9oOWxoE1NPNHvlc8iLgTAiJohk5f7RcsBAN/SVmyueu7oYBiNeYqhW5h8txw2Bj6RPuhAlLwbCiBggvLMy/2iZ00tRukOlLkTJa4GwIiYIb+X+0XKihUCT/MDnkBcDYURMEMnK/aPlvLXocl8P1IUoeTEQRsQEUazcP1quiN3F6ic+Sl4MhBExQTQr94+SE/vUNc18tLwYCCNigAjeyv2j5L47ITSXJz5KXguEFTFBRCv7j5JpL+HyWAVXhWh5MRBGxASBsYNvLbUJxCHnuqUkI7+P5qPkxUAYERMED64k7pK5gjjkUjaB2ncYqEK0vBgII2KAiBg7cHF3qhGnzJ0s4ny3ZFKFaHktEFbEBBHonRGmY9RKbo37B1rfQKMK0fJiIIyICQKXlAJTMF1BHLJgnBUwzCxyBXTRF0NhxEwUld6u971CJ4pDZp6p0mJ0fkKk9cVQGDFzUwRGEDnmPNWKU5bIzYPV952BGpHW10JhxUwUGEPMOZ4+aJk522rB/9UJkdYXQ2HETBTMX5f99HwcqvQNraV0ZyYNSOuLgXiMmBzKhiru/FwlDvlixqMAmR49a4AwQiaJtnFHqkwg7ip357biS9/iceGj9MVAPEYMDrmnT0xtqhGnzGeASaj6wpYCpOW1QFghk0REFW+Ti/apCnfOJ8ljyKH4aH0xEI8RkwMGEY4BTSAOWaLj7nAv/gpIy4uBMEImicZNRg+Pxl0Vn1AJUit+4qP1xUA8RgwOtFeRXCb3CSXz0EbLrR9NUHyUuhYGK2BywBDC+zSZTyi5MV2d6yltVRlKXQyDES8xjITBbq4Oh1xlK9mNk0eqEC0vBsKImCD6Gaf53IKSe7LhFMdzcRai5cVAGBEDROXJr1bLVCNOOYee3Xcs+52FaHktEFbEBIGhQ0w5TzXilBMzeddxZFAVouXFQBgREwRGDklinGvEIfO4QXAh5omPkhcDYURMEBg6pOLDXCNOmWNtyX13qC5EyYuBMCImCIwccnCTAYWW8QzUFMcijypEyYuBMCLmoRXP47x1MqDQcubB0JTjxEfJa4GwIiaI2Lf616lGKDltPqVRIVQZp7oYBiPe/RBTnXJDfrjIfXmLu46vdJS8GAgjYoKoVjZMLQe8JvLei1CFKHkxEEbEH1+Kk83Ihqnl7vM48j1e+RzyUiDMiAkiWNkwtcwT+T73s+9XPoe8GAgjYoJIVj5MLYdNJKdSHvgc8mIgjIgJolr5MLVMAw+JUh/4HPJiIIyIeb7IWfkwJ7lU3y0apkLu8ueCeKYzy3S2pjquWY/N9/SQiCPBp/jhZco91cW16LppQqClRpeFB4KFfiV4O0rJMuYopTLoWrt7Q+XEbV8S9oI+dqKlF11YfBuJ03Gzts6ib8krIZTRN/eR5hY8n9kTg7ncPUe4L8vVltvoyWcZVmnZV64mBjx5SWgME0d3zjcmwXB0rsB157BPJwfPILjgltrmSttn20PAbfFxHBB1EkKVfb8LDSPSLactRVqV7Js/WqjJ+xunXGpsY2wVaIWReSoRw82ADmS975CoIflQbpW7cOO+NZmditwaJ/i5W9uV0P0vMjufrjqeCWNil9bGhhOELC4W1hR6a/jc9iU0VvoQRfr0uI8hj0lSsIi4hY3+/wTkghtrjzwh4Lr/S6ZbRwljYyzdXaQyM6PkQlONMi4npS2WElGfpXu+tPvSHT2pELt015Xo0GEsuw4+wmVe6A396H3Ns2yoqfQ77Hre/U+olxbrkHNI9+IrO19eQtcdYs67jLhTyv1qxDcX97WSBKxcZs+VVjx+XEx2vSLHEVTiTcu7jmo03GsSLYSkuwKhQm8elR03iB4weJbLOGmCOhF96/Y1icctm9vLiXQo69Yzva8cYxufZ2uYeCckFDqHpzwmtEE2htJvbu7P73g86IlThX43wk2YeKrvE5uoIYKeeOPh7UBjlTHPh0KY+anWraKeDZKFS4hJMlNktdIHMlTzhocaD8WN/z3k4Q6U8RmM/lOWG41/wF3aPk8SaExQuNGNdiBjQq16sC48Dpp4fh5NVL+QSuvG1Ggww7zQ4Fz3obXzARfOBxKPNIQu5w0tW+Zz6vrUw6jYGH+i+gZuRMYDmXigaR+N4ekMdGJgG9D2sptDK1l2g90Y2vhwE/q2yDCYwXeEcXtoUx3YRvdd76gHefRnu9GO7FYyPo70k8U5NC+5W6bgrZ5KzccrHo3JfgoL/78ftUUt2fAt9zNt+LsWjLdCqcV6WezyO0xjXjEMeM1pBCWbXgIfX/Usoc3Me00J7G9/8zveYxrDlj2yFP7zLvOY/E3MYwoeb/e4+7g2JkS+wrtrFzzq723bmPMDcv2Szz5cjScN+PFqmnrDp/wE249S0XOJTDiTMt4bX9845svxq0xxXEqZ8uIq+Qn8KnojrZcGbG/we8M45gsSwMugcjXyCuBQnxE/3mUsrDb0JH9W/K/axqAR3/gY40LZ4D/XNuYLPqY/v4X5xEeUVqPnpX+6Z8xXYUDLQJeZ5+5kcGpPYpB569HbZh65n3AGeTSM+SoYGrdT0hdRYTi1J2HAoNGhUcbXZvR03ukW81UwMHtVQHe4ag5KfBIIDsswQGsYOWLw8U6nmK8DAmM2jOmLu4A4xWeB4HA7OSdMEPUuoxjljVKk7Yetepmf4Y3yiWi/7hQT/U4fkxAVepU+JiHSMvphVXIfXepClHxMMa0BwoiYIIqVhuhV2cS2GAgjNIJoVhoiLfeps9pP2+hClLwYCCNigMCYwUhDpOXLo6H5LPpoWBETRLQSEWlZg7D5LAbCiJggspWISMsXECafxUAYERNEtRIRaTnye10/faULUfJiIIyIOUshViIiLWOo6NEqtImPktcCYUU8pmuMVERaRmxln2LVhSh5MRBGxASRrGREWk50lRsu9LoQJS8GwoiYIKqVjEjLw4q3bxXWhSh5MRBGxBgGOf7ZQzIiJWM0k13N/WS/KkTLS4EwIyYIbyUjUnKKWw4hd6NZzUfJi4EwIiaIZCUjUjIzaNTS1500nlNdDIMRLzEUKxWRknPeEl6XfdlfFaLlxUAYERNEs1IRKTn3PQp9Y8AVz11dDIMRLzAwEcRjIiIll7hhZJFDvdLR8logrIgJIvazIzLVh1PG0AIF1p5VRRWi5cVAGBETRGZWF2lzjTjkxmPctfZtCKoQLS8GwoiYIPp2i1bnGnHI4oT7U8auBFXKRV8MhREzUHjZpPkymT8oWZglpCWXHhApfS0UVsxEwWxRNaU4oTjkHoPzsR/p14i0vhgKI2aiYIq1GMJcKw5ZIlNocd/MhEjri6EwYiYKjCDw2Pu5VhxyX5CJ/JIJkdYXQ2HEzE0QjrnMrrmoPmj54nmhEZleGEugsGImCgwjvOQ21YpT5qZMqWHMSWlEWl8MhREzUXCXYIllrhWHTN8LZlUcAw+NSOmLoTBiJorCPZo+z7XikCVzs5n3fW/sBZHSF0NhxEwUGE+E5qbdcocqfftyGOkGNSCtLwbiMWJwYMqQGJmE8wLilLlBueXiW7wC0vJaIKyQSQLjieSmuf1TFW4gdyJjFKb4aH0xEI8RkwNGEynH+dE4ZdqqNe9jPxWhAGl5MRBGyCTBnYRephfHoQpCT96lVmc+Sl8MxGPE4JAwkuCWrKlGnLLwdJ6PI/upBqTktUBYIZMEBhIlTKYYpyqh56WVMvPR+mIgHiMmh7SVx536ShYMxNGh7EehFB+lLobBCJgcupNBmKwglMxZiOzLOJyn+Wh9MRRGzECBHmJoIjI9G6fc/JYR8hiRn4VoeS0QVsQEwVTR/OwE4pB5JCpn6af8VSFaXgyEETFBYPjgmM5+AnHIRdBzCqkfNtN8lLwYCCNigsDwwVWaE19BHHI/G9hSP9mqCtHyYiCMiAkCoweJPs014pDxNEQebfUTHyUvBsKIGCD6iVXn4lQjTpkOlLmFOPHR8logrIgJIj7mcPxwkdtWq3P9iLguRMmLgTAiJohspV/Ucu0+Uv1ArC5EyYuBMCImiGqlp9QyD37XnjxHl3Gqi2Ew4uV5NrFyVmqZfcmYxqy+KkTJa4GwIh4H+4yclVqOm0PAPTmlLkTJi4EwIiaIZOWs1PIFhMlnMRBGxARRrZyVWg7d+qNbLehClPy5IJ7pk3I9Z8MIYs83zVOoGBHUe7bN1v1mC60TOC3fbVKklTp2XaNnUGoZdjkhetc7SwXRJ0ERw0mqpjzshGqhiQRNKmiq4iS2UXTbUvHej6R1koCzb1L06H7VFrt7CnfmjY3eDS8hl+mJHMsmHm+htu/tDEDIXU9lq1GCG1s+0VT7GJgEho1Vk76sUlrZcqVDQ6LNRpS+3620tkWPMXJjZqzom4zm3g2vCG6O49JEkrFO4ZhZjaajt1S2hsvtjhTcQlZaqMHTlDQJWoW876hK/BndSmbUAYVRduVdSnSkaN0P3o3tNriTvqaSue8IdWzfniaed4R8cDsiDTNk35MSERsgozai2sWemwUViK6goXCDRmT+91DHFhZ07FBnuK5A9xtA61YV1buNt4rTKJnOHDX2j3vPFJuF6zGeCUi95LHWG7cmntNsdAcJeB32XaSVx+h4lpjWLJ5bsP2+Noy771vCg8DJ68a8CHVfEMUr1PdVQMSH2urbrqMhScnvbiUy3E3wMG2JFSP1VUPcyjjansBbyQe/u55EfF+vAtTxTKRuNc9laNf2RQXZfI2tZ22ALmGsz0KuOXo0YZTxLEU3FmnwkPD6Q//ajMZl7P0IsZ+74NI2LtOXgGvoet5iazy+z2UsYbb5+6oOnsjAXQGJPWxUt7LrGTe8O9o0mizFMTALNLnluUZ6s7Sa674gEhoemjzUkMdR2ErndNx/2txQx8PdZF84kBBKT9WBKoJvH1UEesGjwj08tKQJaCx6VUCD5EGG9jd9ocHT0WDo6A6zQcFjSt+fYYvD6XgXXOwpPyK944bhSmWHGXeEzjW4NHxV6Yzx1LXqu0ENcYjE8aT1hHTi8aALjYzwzI/WlNY/rSHg3j4WGrfkfWbLhdINaniEPec0qiy9j1oItdwak2nHJPdZD9zeiF94bAZDu7HOhlqXcsMNZbIrGgONx4F+9K5EfKY7AKGf16+xpytIfGTwU0DlHs93Br7s6emQ+57iYX1Ti6O5IB1nmAIj+u4WxBEF7iuukA1NpY1WvxC0yI1GL43bsl3EhbW92w3agesErDDiyuh94aFD49e6sykbuYEcMaAxd715DD61Oj7ckzYVrj4lmmfX8eGK+5xLd5aJbBEljtd45GlvtpVsqIXZyLuMgPEwu97ag7Wk+PAKDC36vRBTfocpzCt2AK/5iKDkR6eAj7YXCQ1k3mU2YH/p66W/xwgGTTHutmvjn/cYwdTbKwYQaOm33B1GpNV+gaqoaPk/lIhagAqK2qz8H/71hx//+pcf/vff+MvlRPLL/wOIoab3CmVuZHN0cmVhbQplbmRvYmoKMTIgMCBvYmoKMTEzMDQKZW5kb2JqCjEwIDAgb2JqClsgXQplbmRvYmoKMTcgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA5MSA+PgpzdHJlYW0KeJw1jLsNwDAIRHumuBH4OID3iaIU9v5tiC0X3D3pifNsYGSdhyO04xaypnBTTFJOqHcMaqU3HTvoJc39NMl6Lhr0D3H1FbabA5JRJJGHRJfLlWflX3w+DG8cYgplbmRzdHJlYW0KZW5kb2JqCjE4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTY0ID4+CnN0cmVhbQp4nD2QwRFDIQhE71axJYCAQD3JZHL4v/9rQJNcZB1g96k7gZBRhzPDZ+LJg9OxNHBvFYxrCK8j9AhNApPAxMGaeAwLAadhkWMu31WWVaeVrpqNnte9Y0HVaZc1DW3agfKtjz/CNd6j8BrsHkIHsSh0bmVaC5lYPGucO8yjzOd+Ttt3PRitptSsN3LZ1z06y9RQXlr7hM5otP0n1y+7MV4fhRQ5CAplbmRzdHJlYW0KZW5kb2JqCjE5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNjEgPj4Kc3RyZWFtCnicMzU1VzBQsLQAEqamRgrmRpYKKYZcQD6IlctlaGkOZuWAWRbGQAZIGZxhAKTBmnNgenK4MrjSAMsVEMwKZW5kc3RyZWFtCmVuZG9iagoyMCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMwNyA+PgpzdHJlYW0KeJw9kktuAzEMQ/c+hS4QwPrZnvOkKLqY3n/bJyXpihzZFkVqlrpMWVMekDSThH/p8HCxnfI7bM9mZuBaopeJ5ZTn0BVi7qJ82cxGXVknxeqEZjq36FE5Fwc2Taqfqyyl3S54Dtcmnlv2ET+80KAe1DUuCTd0V6NlKTRjqvt/0nv8jDLgakxdbFKrex88XkRV6OgHR4kiY5cX5+NBCelKwmhaiJV3RQNB7vK0ynsJ7tveasiyB6mYzjspZrDrdFIubheHIR7I8qjw5aPYa0LP+LArJfRI2IYzcifuaMbm1MjikP7ejQRLj65oIfPgr27WLmC8UzpFYmROcqxpi1VO91AU07nDvQwQ9WxFQylzkdXqX8POC2uWbBZ4SvoFHqPdJksOVtnbqE7vrTzZ0PcfWtd0HwplbmRzdHJlYW0KZW5kb2JqCjIxIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ0ID4+CnN0cmVhbQp4nEWRTXIFIQiE956iL/Cq5Fc9z6RSWUzuvw3NvCQrWoXmA9MCE0fwEkPsiZUTHzJ8L+gyfLcyO/A62ZlwT7huXMNlwzNhW+A7Kss7XkN3tlI/naGq7xo53i5SNXRlZJ96oZoLzJCIrhFZdCuXdUDTlO5S4RpsW4IU9UqsJ52gNOgRyvB3lGt8dRNPr7HkVM0hWs2tExqKsGx4QdTJJBG1DYsnlnMhUfmqG6s6LmCTJeL0gNyglWZ8elJJETCDfKzJaMwCNtCTu2cXxppLHkWOVzSYsDtJNfCA9+K2vvc2cY/zF/iFd9//Kw591wI+fwBL/l0GCmVuZHN0cmVhbQplbmRvYmoKMjIgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzIgPj4Kc3RyZWFtCnicNVFJbsQwDLv7FfzAANbuvCfFoIf2/9dSyhQIQCW2uCViYyMCLzH4OYjc+JI1oyZ+Z3JX/CxPhUfCreBJFIGX4V52gssbxmU/DjMfvJdWzqTGkwzIRTY9PBEy2CUQOjC7BnXYZtqJviHhsyNSzUaW09cS9NIqBMpTtt/pghJtq/pz+6wLbfvaE052e+pJ5ROI55aswGXjFZPFWAY9UblLMX2Q6myhJ6G8KJ+DbD5qiESXKGfgicHBKNAO7LntZ+JVIWhd3adtY6hGSsfTvw1NTZII+UQJZ7Y07hb+f8+9vtf7D04hVBEKZW5kc3RyZWFtCmVuZG9iagoyMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDIzMSA+PgpzdHJlYW0KeJw1TzmSBCEMy3mFPjBVGNtAv6entjbY+X+6kplOkPAhydMTHZl4mSMjsGbH21pkIGbgU0zFv/a0DxOq9+AeIpSLC2GGkXDWrONuno4X/3aVz1gH7zb4illeENjCTNZXFmcu2wVjaZzEOclujF0TsY11radTWEcwoQyEdLbDlCBzVKT0yY4y5ug4kSeei+/22yx2OX4O6ws2jSEV5/gqeoI2g6Lsee8CGnJB/13d+B5Fu+glIBsJFtZRYu6c5YRfvXZ0HrUoEnNCmkEuEyHN6SqmEJpQrLOjoFJRcKk+p+isn3/lX1wtCmVuZHN0cmVhbQplbmRvYmoKMjQgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDkgPj4Kc3RyZWFtCnicPVA7jkQhDOs5hS/wJPIjcB5Gqy1m79+uA5opUEx+tjMk0BGBRwwxlK/jJa2groG/i0LxbuLrg8Igq0NSIM56D4h07KY2kRM6HZwzP2E3Y47ARTEGnOl0pj0HJjn7wgqEcxtl7FZIJ4mqIo7qM44pnip7n3gWLO3INlsnkj3kIOFSUonJpZ+Uyj9typQKOmbRBCwSueBkE004y7tJUowZlDLqHqZ2In2sPMijOuhkTc6sI5nZ00/bmfgccLdf2mROlcd0Hsz4nLTOgzkVuvfjiTYHTY3a6Oz3E2kqL1K7HVqdfnUSld0Y5xgSl2d/Gd9k//kH/odaIgplbmRzdHJlYW0KZW5kb2JqCjI1IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzk1ID4+CnN0cmVhbQp4nD1SS27FQAjb5xRcoNLwm895UlXdvPtva0NSqSq8iTHGMH3KkLnlS10ScYXJt16uWzymfC5bWpl5iLuLjSU+ttyX7iG2XXQusTgdR/ILMp0qRKjNqtGh+EKWhQeQTvChC8J9Of7jL4DB17ANuOE9MkGwJOYpQsZuURmaEkERYeeRFaikUJ9Zwt9R7uv3MgVqb4ylC2Mc9Am0BUJtSMQC6kAAROyUVK2QjmckE78V3WdiHGDn0bIBrhlURJZ77MeIqc6ojLxExD5PTfoolkwtVsZuUxlf/JSM1Hx0BSqpNPKU8tBVs9ALWIl5EvY5/Ej459ZsIYY6btbyieUfM8UyEs5gSzlgoZfjR+DbWXURrh25uM50gR+V1nBMtOt+yPVP/nTbWs11vHIIokDlTUHwuw6uRrHExDI+nY0peqIssBqavEYzwWEQEdb3w8gDGv1yvBA0p2sitFgim7ViRI2KbHM9vQTWTO/FOdbDE8Js753WobIzMyohgtq6hmrrQHazvvNwtp8/M+iibQplbmRzdHJlYW0KZW5kb2JqCjI2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ5ID4+CnN0cmVhbQp4nE1RSYoDMAy75xX6QCFek7ynQ5lD5//Xyg6FOQQJr5KTlphYCw8xhB8sPfiRIXM3/Rt+otm7WXqSydn/mOciU1H4UqguYkJdiBvPoRHwPaFrElmxvfE5LKOZc74HH4W4BDOhAWN9STK5qOaVIRNODHUcDlqkwrhrYsPiWtE8jdxu+0ZmZSaEDY9kQtwYgIgg6wKyGCyUNjYTMlnOA+0NyQ1aYNepG1GLgiuU1gl0olbEqszgs+bWdjdDLfLgqH3x+mhWl2CF0Uv1WHhfhT6YqZl27pJCeuFNOyLMHgqkMjstK7V7xOpugfo/y1Lw/cn3+B2vD838XJwKZW5kc3RyZWFtCmVuZG9iagoyNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDk0ID4+CnN0cmVhbQp4nEWNwRHAIAgE/1RBCQoK2k8mk4f2/40QMnxg5w7uhAULtnlGHwWVJl4VWAdKY9xQj0C94XItydwFD3Anf9rQVJyW03dpkUlVKdykEnn/DmcmkKh50WOd9wtj+yM8CmVuZHN0cmVhbQplbmRvYmoKMjggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA3MiA+PgpzdHJlYW0KeJwzMrdQMFCwNAEShhYmCuZmBgophlxAvqmJuUIuF0gMxMoBswyAtCWcgohngJggbRDFIBZEsZmJGUQdnAGRy+BKAwAl2xbJCmVuZHN0cmVhbQplbmRvYmoKMjkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA0NyA+PgpzdHJlYW0KeJwzMrdQMFCwNAEShhYmCuZmBgophlyWEFYuF0wsB8wC0ZZwCiKewZUGALlnDScKZW5kc3RyZWFtCmVuZG9iagozMCAwIG9iago8PCAvQkJveCBbIC0xMDIxIC00NjMgMTc5NCAxMjMzIF0gL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAzOQovU3VidHlwZSAvRm9ybSAvVHlwZSAvWE9iamVjdCA+PgpzdHJlYW0KeJzjMjQwUzA2NVXI5TI3NgKzcsAsI3MjIAski2BBZDO40gAV8wp8CmVuZHN0cmVhbQplbmRvYmoKMzEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNjMgPj4Kc3RyZWFtCnicRZA7EgMhDEN7TqEj+CMDPs9mMik2929j2GxSwNNYIIO7E4LU2oKJ6IKHtiXdBe+tBGdj/Ok2bjUS5AR1gFak42iUUn25xWmVdPFoNnMrC60THWYOepSjGaAQOhXe7aLkcqbuzvlDcPVf9b9i3TmbiYHJyh0IzepT3Pk2O6K6usn+pMfcrNd+K+xVYWlZS8sJt527ZkAJ3FM52qs9Px8KOvYKZW5kc3RyZWFtCmVuZG9iagozMiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDIxOCA+PgpzdHJlYW0KeJw9ULmNBDEMy12FGljAeu2pZxaLS6b/9Ej59iLRFkVSKjWZkikvdZQlWVPeOnyWxA55huVuZDYlKkUvk7Al99AK8X2J5hT33dWWs0M0l2g5fgszKqobHdNLNppwKhO6oNzDM/oNbXQDVocesVsg0KRg17YgcscPGAzBmROLIgxKTQb/rnKPn16LGz7D8UMUkZIO5jX/WP3ycw2vU48nkW5vvuJenKkOAxEckpq8I11YsS4SEWk1QU3PwFotgLu3Xv4btCO6DED2icRxmlKOob9rcKXPL+UnU9gKZW5kc3RyZWFtCmVuZG9iagozMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDgzID4+CnN0cmVhbQp4nEWMuw3AMAhEe6ZgBH4m9j5RlMLevw0QJW64J909XB0JmSluM8NDBp4MLIZdcYH0ljALXEdQjp3so2HVvuoEjfWmUvPvD5Se7KzihusBAkIaZgplbmRzdHJlYW0KZW5kb2JqCjM0IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNTEgPj4Kc3RyZWFtCnicMza0UDBQMDQwB5JGhkCWkYlCiiEXSADEzOWCCeaAWQZAGqI4B64mhyuDKw0A4bQNmAplbmRzdHJlYW0KZW5kb2JqCjM1IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTYwID4+CnN0cmVhbQp4nEWQORIDMQgEc72CJ0hcgvesy7XB+v+pB9ZHoukCNBy6Fk3KehRoPumxRqG60GvoLEqSRMEWkh1Qp2OIOyhITEhjkki2HoMjmlizXZiZVCqzUuG0acXCv9la1chEjXCN/InpBlT8T+pclPBNg6+SMfoYVLw7g4xJ+F5F3Fox7f5EMLEZ9glvRSYFhImxqdm+z2CGzPcK1zjH8w1MgjfrCmVuZHN0cmVhbQplbmRvYmoKMzYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAzMzQgPj4Kc3RyZWFtCnicLVJLcsUgDNtzCl2gM/gH5DzpdLp4vf+2kpNFRg5g9DHlholKfFkgt6PWxLeNzECF4a+rzIXPSNvIOojLkIu4ki2Fe0Qs5DHEPMSC76vxHh75rMzJswfGL9l3Dyv21IRlIePFGdphFcdhFeRYsHUhqnt4U6TDqSTY44v/PsVzLQQtfEbQgF/kn6+O4PmSFmn3mG3TrnqwTDuqpLAcbE9zXiZfWme5Oh7PB8n2rtgRUrsCFIW5M85z4SjTVka0FnY2SGpcbG+O/VhK0IVuXEaKI5CfqSI8oKTJzCYK4o+cHnIqA2Hqmq50chtVcaeezDWbi7czSWbrvkixmcJ5XTiz/gxTZrV5J89yotSpCO+xZ0vQ0Dmunr2WWWh0mxO8pITPxk5PTr5XM+shORUJqWJaV8FpFJliCdsSX1NRU5p6Gf778u7xO37+ASxzfHMKZW5kc3RyZWFtCmVuZG9iagozNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMyMCA+PgpzdHJlYW0KeJw1UktuBTEI288puECl8E/O86qqi777b2sTvRVMMGDjKS9Z0ku+1CXbpcPkWx/3JbFC3o/tmsxSxfcWsxTPLa9HzxG3LQoEURM9WJkvFSLUz/ToOqhwSp+BVwi3FBu8g0kAg2r4Bx6lMyBQ50DGu2IyUgOCJNhzaXEIiXImiX+kvJ7fJ62kofQ9WZnL35NLpdAdTU7oAcXKxUmgXUn5oJmYSkSSl+t9sUL0hsCSPD5HMcmA7DaJbaIFJucepSXMxBQ6sMcCvGaa1VXoYMIehymMVwuzqB5s8lsTlaQdreMZ2TDeyzBTYqHhsAXU5mJlgu7l4zWvwojtUZNdw3Duls13CNFo/hsWyuBjFZKAR6exEg1pOMCIwJ5eOMVe8xM5DsCIY52aLAxjaCaneo6JwNCes6VhxsceWvXzD1TpfIcKZW5kc3RyZWFtCmVuZG9iagozOCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE4ID4+CnN0cmVhbQp4nDM2tFAwgMMUQ640AB3mA1IKZW5kc3RyZWFtCmVuZG9iagozOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEzMyA+PgpzdHJlYW0KeJxFj0sOBCEIRPecoo7Axx/ncTLphXP/7YCdbhNjPYVUgbmCoT0uawOdFR8hGbbxt6mWjkVZPlR6UlYPyeCHrMbLIdygLPCCSSqGIVCLmBqRLWVut4DbNg2yspVTpY6wi6Mwj/a0bBUeX6JbInWSP4PEKi/c47odyKXWu96ii75/pAExCQplbmRzdHJlYW0KZW5kb2JqCjQwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzQwID4+CnN0cmVhbQp4nDVSOW4EMQzr/Qp9IIBu2+/ZIEiR/L8NqdkUA3F0UpQ7WlR2y4eFVLXsdPm0ldoSN+R3ZYXECcmrEu1ShkiovFYh1e+ZMq+3NWcEyFKlwuSk5HHJgj/DpacLx/m2sa/lyB2PHlgVI6FEwDLFxOgals7usGZbfpZpwI94hJwr1i3HWAVSG9047Yr3oXktsgaIvZmWigodVokWfkHxoEeNffYYVFgg0e0cSXCMiVCRgHaB2kgMOXssdlEf9DMoMRPo2htF3EGBJZKYOcW6dPTf+NCxoP7YjDe/OirpW1pZY9I+G+2Uxiwy6XpY9HTz1seDCzTvovzn1QwSNGWNksYHrdo5hqKZUVZ4t0OTDc0xxyHzDp7DGQlK+jwUv48lEx2UyN8ODaF/Xx6jjJw23gLmoj9tFQcO4rPDXrmBFUoXa5L3AalM6IHp/6/xtb7X1x8d7YDGCmVuZHN0cmVhbQplbmRvYmoKNDEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNTEgPj4Kc3RyZWFtCnicLVFJcgNBCLvPK/SEZqffY5crh+T/1wjKBwYNi0B0WuKgjJ8gLFe85ZGraMPfMzGC3wWHfivXbVjkQFQgSWNQNaF28Xr0HthxmAnMk9awDGasD/yMKdzoxeExGWe312XUEOxdrz2ZQcmsXMQlExdM1WEjZw4/mTIutHM9NyDnRliXYZBuVhozEo40hUghhaqbpM4EQRKMrkaNNnIU+6Uvj3SGVY2oMexzLW1fz004a9DsWKzy5JQeXXEuJxcvrBz09TYDF1FprPJASMD9bg/1c7KT33hL584W0+N7zcnywlRgxZvXbkA21eLfvIjj+4yv5+f5/ANfYFuICmVuZHN0cmVhbQplbmRvYmoKNDIgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNzQgPj4Kc3RyZWFtCnicTZBJDkMhDEP3nMIXqIQzwOc8v6q6aO+/rUMHdYH85CBwPDzQcSQudGTojI4rmxzjwLMgY+LROP/JuD7EMUHdoi1Yl3bH2cwSc8IyMQK2RsnZPKLAD8dcCBJklx++wCAiXY/5VvNZk/TPtzvdj7q0Zl89osCJ7AjFsAFXgP26x4FLwvle0+SXKiVjE4fygeoiUjY7oRC1VOxyqoqz3ZsrcBX0/NFD7u0FtSM83wplbmRzdHJlYW0KZW5kb2JqCjQzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNzUgPj4Kc3RyZWFtCnicM7U0UjBQMDYAEqZmRgqmJuYKKYZcQD6IlctlaGQKZuVwGVmaKVhYABkmZuZQIZiGHC5jU3OgAUBFxqZgGqo/hyuDKw0AlZAS7wplbmRzdHJlYW0KZW5kb2JqCjQ0IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTQxID4+CnN0cmVhbQp4nD2PwQ7DMAhD7/kK/0Ck2CmhfE+naofu/68jS7sLegJjjIXQ0BuqmsOGYJvjxdIlVGv4FMVAJTfImWAOpaTSHUeRemI4GFwetBuO4rHo+hG7kmZ90MZCuiVogHusU2ncpnETxB01Beop6pyjvBC5n6ln2DSS3TSzknO4Db97z1PX/6ervMv5Bb13Lv4KZW5kc3RyZWFtCmVuZG9iago0NSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDIxNSA+PgpzdHJlYW0KeJw1UTkOAyEM7PcV/kAkjC94T6Iozf6/zYzRVh7BXIa0lCGZ8lKTqCHlUz56mS6cutzXzGo055a0LXOAuLa8L62SwIlmiIPBaZi4AZo8AUPX0ahRQxce0NSlUyiw3AQ+irduD91jtYGXtiHniSBiKBksQc2pRRMWbc8npDW/Xosb3pft3chTpcaWGIEGAVY4HNfo1/CVPU8m0XQVMtSrNcsYCRNFIjz5jqbVE+taNNIyEtTGEaxqA7w7/TBOAAATccsCZJ9KlLPkxG+x9LMGV/r+AZ9HVJYKZW5kc3RyZWFtCmVuZG9iagoxNSAwIG9iago8PCAvQmFzZUZvbnQgL0RlamFWdVNhbnMgL0NoYXJQcm9jcyAxNiAwIFIKL0VuY29kaW5nIDw8Ci9EaWZmZXJlbmNlcyBbIDMyIC9zcGFjZSA0NiAvcGVyaW9kIDQ4IC96ZXJvIC9vbmUgL3R3byAvdGhyZWUgL2ZvdXIgL2ZpdmUgL3NpeCA1NgovZWlnaHQgNjUgL0EgNjggL0QgNzYgL0wgOTcgL2EgL2IgL2MgL2QgL2UgMTA1IC9pIDEwOCAvbCAxMTAgL24gL28gMTE0IC9yCi9zIC90IC91IC92IDEyMSAveSBdCi9UeXBlIC9FbmNvZGluZyA+PgovRmlyc3RDaGFyIDAgL0ZvbnRCQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvRm9udERlc2NyaXB0b3IgMTQgMCBSCi9Gb250TWF0cml4IFsgMC4wMDEgMCAwIDAuMDAxIDAgMCBdIC9MYXN0Q2hhciAyNTUgL05hbWUgL0RlamFWdVNhbnMKL1N1YnR5cGUgL1R5cGUzIC9UeXBlIC9Gb250IC9XaWR0aHMgMTMgMCBSID4+CmVuZG9iagoxNCAwIG9iago8PCAvQXNjZW50IDkyOSAvQ2FwSGVpZ2h0IDAgL0Rlc2NlbnQgLTIzNiAvRmxhZ3MgMzIKL0ZvbnRCQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvRm9udE5hbWUgL0RlamFWdVNhbnMgL0l0YWxpY0FuZ2xlIDAKL01heFdpZHRoIDEzNDIgL1N0ZW1WIDAgL1R5cGUgL0ZvbnREZXNjcmlwdG9yIC9YSGVpZ2h0IDAgPj4KZW5kb2JqCjEzIDAgb2JqClsgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAKNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCAzMTggNDAxIDQ2MCA4MzggNjM2Cjk1MCA3ODAgMjc1IDM5MCAzOTAgNTAwIDgzOCAzMTggMzYxIDMxOCAzMzcgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNgo2MzYgNjM2IDMzNyAzMzcgODM4IDgzOCA4MzggNTMxIDEwMDAgNjg0IDY4NiA2OTggNzcwIDYzMiA1NzUgNzc1IDc1MiAyOTUKMjk1IDY1NiA1NTcgODYzIDc0OCA3ODcgNjAzIDc4NyA2OTUgNjM1IDYxMSA3MzIgNjg0IDk4OSA2ODUgNjExIDY4NSAzOTAgMzM3CjM5MCA4MzggNTAwIDUwMCA2MTMgNjM1IDU1MCA2MzUgNjE1IDM1MiA2MzUgNjM0IDI3OCAyNzggNTc5IDI3OCA5NzQgNjM0IDYxMgo2MzUgNjM1IDQxMSA1MjEgMzkyIDYzNCA1OTIgODE4IDU5MiA1OTIgNTI1IDYzNiAzMzcgNjM2IDgzOCA2MDAgNjM2IDYwMCAzMTgKMzUyIDUxOCAxMDAwIDUwMCA1MDAgNTAwIDEzNDIgNjM1IDQwMCAxMDcwIDYwMCA2ODUgNjAwIDYwMCAzMTggMzE4IDUxOCA1MTgKNTkwIDUwMCAxMDAwIDUwMCAxMDAwIDUyMSA0MDAgMTAyMyA2MDAgNTI1IDYxMSAzMTggNDAxIDYzNiA2MzYgNjM2IDYzNiAzMzcKNTAwIDUwMCAxMDAwIDQ3MSA2MTIgODM4IDM2MSAxMDAwIDUwMCA1MDAgODM4IDQwMSA0MDEgNTAwIDYzNiA2MzYgMzE4IDUwMAo0MDEgNDcxIDYxMiA5NjkgOTY5IDk2OSA1MzEgNjg0IDY4NCA2ODQgNjg0IDY4NCA2ODQgOTc0IDY5OCA2MzIgNjMyIDYzMiA2MzIKMjk1IDI5NSAyOTUgMjk1IDc3NSA3NDggNzg3IDc4NyA3ODcgNzg3IDc4NyA4MzggNzg3IDczMiA3MzIgNzMyIDczMiA2MTEgNjA1CjYzMCA2MTMgNjEzIDYxMyA2MTMgNjEzIDYxMyA5ODIgNTUwIDYxNSA2MTUgNjE1IDYxNSAyNzggMjc4IDI3OCAyNzggNjEyIDYzNAo2MTIgNjEyIDYxMiA2MTIgNjEyIDgzOCA2MTIgNjM0IDYzNCA2MzQgNjM0IDU5MiA2MzUgNTkyIF0KZW5kb2JqCjE2IDAgb2JqCjw8IC9BIDE3IDAgUiAvRCAxOCAwIFIgL0wgMTkgMCBSIC9hIDIwIDAgUiAvYiAyMSAwIFIgL2MgMjIgMCBSIC9kIDIzIDAgUgovZSAyNCAwIFIgL2VpZ2h0IDI1IDAgUiAvZml2ZSAyNiAwIFIgL2ZvdXIgMjcgMCBSIC9pIDI4IDAgUiAvbCAyOSAwIFIKL24gMzEgMCBSIC9vIDMyIDAgUiAvb25lIDMzIDAgUiAvcGVyaW9kIDM0IDAgUiAvciAzNSAwIFIgL3MgMzYgMCBSCi9zaXggMzcgMCBSIC9zcGFjZSAzOCAwIFIgL3QgMzkgMCBSIC90aHJlZSA0MCAwIFIgL3R3byA0MSAwIFIgL3UgNDIgMCBSCi92IDQzIDAgUiAveSA0NCAwIFIgL3plcm8gNDUgMCBSID4+CmVuZG9iagozIDAgb2JqCjw8IC9GMSAxNSAwIFIgPj4KZW5kb2JqCjQgMCBvYmoKPDwgL0ExIDw8IC9DQSAwIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0EyIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0EzIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDAuNSA+PiA+PgplbmRvYmoKNSAwIG9iago8PCA+PgplbmRvYmoKNiAwIG9iago8PCA+PgplbmRvYmoKNyAwIG9iago8PCAvRjEtRGVqYVZ1U2Fucy1taW51cyAzMCAwIFIgPj4KZW5kb2JqCjIgMCBvYmoKPDwgL0NvdW50IDEgL0tpZHMgWyAxMSAwIFIgXSAvVHlwZSAvUGFnZXMgPj4KZW5kb2JqCjQ2IDAgb2JqCjw8IC9DcmVhdGlvbkRhdGUgKEQ6MjAyMTA5MTYxNDM2MzgrMDInMDAnKQovQ3JlYXRvciAoTWF0cGxvdGxpYiB2My40LjMsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAoTWF0cGxvdGxpYiBwZGYgYmFja2VuZCB2My40LjMpID4+CmVuZG9iagp4cmVmCjAgNDcKMDAwMDAwMDAwMCA2NTUzNSBmIAowMDAwMDAwMDE2IDAwMDAwIG4gCjAwMDAwMjE0NDIgMDAwMDAgbiAKMDAwMDAyMTE3OSAwMDAwMCBuIAowMDAwMDIxMjExIDAwMDAwIG4gCjAwMDAwMjEzNTEgMDAwMDAgbiAKMDAwMDAyMTM3MiAwMDAwMCBuIAowMDAwMDIxMzkzIDAwMDAwIG4gCjAwMDAwMDAwNjUgMDAwMDAgbiAKMDAwMDAwMDM5OSAwMDAwMCBuIAowMDAwMDExODAwIDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAxMTc3OCAwMDAwMCBuIAowMDAwMDE5NzkyIDAwMDAwIG4gCjAwMDAwMTk1OTIgMDAwMDAgbiAKMDAwMDAxOTEzOSAwMDAwMCBuIAowMDAwMDIwODQ1IDAwMDAwIG4gCjAwMDAwMTE4MjAgMDAwMDAgbiAKMDAwMDAxMTk4MyAwMDAwMCBuIAowMDAwMDEyMjIwIDAwMDAwIG4gCjAwMDAwMTIzNTMgMDAwMDAgbiAKMDAwMDAxMjczMyAwMDAwMCBuIAowMDAwMDEzMDUwIDAwMDAwIG4gCjAwMDAwMTMzNTUgMDAwMDAgbiAKMDAwMDAxMzY1OSAwMDAwMCBuIAowMDAwMDEzOTgxIDAwMDAwIG4gCjAwMDAwMTQ0NDkgMDAwMDAgbiAKMDAwMDAxNDc3MSAwMDAwMCBuIAowMDAwMDE0OTM3IDAwMDAwIG4gCjAwMDAwMTUwODEgMDAwMDAgbiAKMDAwMDAxNTIwMCAwMDAwMCBuIAowMDAwMDE1MzcyIDAwMDAwIG4gCjAwMDAwMTU2MDggMDAwMDAgbiAKMDAwMDAxNTg5OSAwMDAwMCBuIAowMDAwMDE2MDU0IDAwMDAwIG4gCjAwMDAwMTYxNzcgMDAwMDAgbiAKMDAwMDAxNjQxMCAwMDAwMCBuIAowMDAwMDE2ODE3IDAwMDAwIG4gCjAwMDAwMTcyMTAgMDAwMDAgbiAKMDAwMDAxNzMwMCAwMDAwMCBuIAowMDAwMDE3NTA2IDAwMDAwIG4gCjAwMDAwMTc5MTkgMDAwMDAgbiAKMDAwMDAxODI0MyAwMDAwMCBuIAowMDAwMDE4NDkwIDAwMDAwIG4gCjAwMDAwMTg2MzcgMDAwMDAgbiAKMDAwMDAxODg1MSAwMDAwMCBuIAowMDAwMDIxNTAyIDAwMDAwIG4gCnRyYWlsZXIKPDwgL0luZm8gNDYgMCBSIC9Sb290IDEgMCBSIC9TaXplIDQ3ID4+CnN0YXJ0eHJlZgoyMTY1OQolJUVPRgo=\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 8.082208633422852\n", "Layer 2 - Variance: 37.87363815307617\n", "Layer 4 - Variance: 96.36101531982422\n", "Layer 6 - Variance: 237.2630615234375\n", "Layer 8 - Variance: 303.44244384765625\n"]}], "source": ["var_init(model, std=0.1)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "5a7f323d", "metadata": {"papermill": {"duration": 0.054768, "end_time": "2021-09-16T12:36:39.283572", "exception": false, "start_time": "2021-09-16T12:36:39.228804", "status": "completed"}, "tags": []}, "source": ["With a higher standard deviation, the activations are likely to explode.\n", "You can play around with the specific standard deviation values, but it will be hard to find one that gives us a good activation distribution across layers and is very specific to our model.\n", "If we would change the hidden sizes or number of layers, you would have\n", "to search all over again, which is neither efficient nor recommended."]}, {"cell_type": "markdown", "id": "06378f27", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.054595, "end_time": "2021-09-16T12:36:39.392403", "exception": false, "start_time": "2021-09-16T12:36:39.337808", "status": "completed"}, "tags": []}, "source": ["### How to find appropriate initialization values\n", "\n", "From our experiments above, we have seen that we need to sample the weights from a distribution, but are not sure which one exactly.\n", "As a next step, we will try to find the optimal initialization from the perspective of the activation distribution.\n", "For this, we state two requirements:\n", "\n", "1. The mean of the activations should be zero\n", "2. The variance of the activations should stay the same across every layer\n", "\n", "Suppose we want to design an initialization for the following layer: $y=Wx+b$ with $y\\in\\mathbb{R}^{d_y}$, $x\\in\\mathbb{R}^{d_x}$.\n", "Our goal is that the variance of each element of $y$ is the same as the input, i.e. $\\text{Var}(y_i)=\\text{Var}(x_i)=\\sigma_x^{2}$, and that the mean is zero.\n", "We assume $x$ to also have a mean of zero, because, in deep neural networks, $y$ would be the input of another layer.\n", "This requires the bias and weight to have an expectation of 0.\n", "Actually, as $b$ is a single element per output neuron and is constant across different inputs, we set it to 0 overall.\n", "\n", "Next, we need to calculate the variance with which we need to initialize the weight parameters.\n", "Along the calculation, we will need to following variance rule: given two independent variables, the variance of their product is $\\text{Var}(X\\cdot Y) = \\mathbb{E}(Y)^2\\text{Var}(X) + \\mathbb{E}(X)^2\\text{Var}(Y) + \\text{Var}(X)\\text{Var}(Y) = \\mathbb{E}(Y^2)\\mathbb{E}(X^2)-\\mathbb{E}(Y)^2\\mathbb{E}(X)^2$ ($X$ and $Y$ are not refering to $x$ and $y$, but any random variable).\n", "\n", "The needed variance of the weights, $\\text{Var}(w_{ij})$, is calculated as follows:\n", "\n", "$$\n", "\\begin{split}\n", " y_i & = \\sum_{j} w_{ij}x_{j}\\hspace{10mm}\\text{Calculation of a single output neuron without bias}\\\\\n", " \\text{Var}(y_i) = \\sigma_x^{2} & = \\text{Var}\\left(\\sum_{j} w_{ij}x_{j}\\right)\\\\\n", " & = \\sum_{j} \\text{Var}(w_{ij}x_{j}) \\hspace{10mm}\\text{Inputs and weights are independent of each other}\\\\\n", " & = \\sum_{j} \\text{Var}(w_{ij})\\cdot\\text{Var}(x_{j}) \\hspace{10mm}\\text{Variance rule (see above) with expectations being zero}\\\\\n", " & = d_x \\cdot \\text{Var}(w_{ij})\\cdot\\text{Var}(x_{j}) \\hspace{10mm}\\text{Variance equal for all $d_x$ elements}\\\\\n", " & = \\sigma_x^{2} \\cdot d_x \\cdot \\text{Var}(w_{ij})\\\\\n", " \\Rightarrow \\text{Var}(w_{ij}) = \\sigma_{W}^2 & = \\frac{1}{d_x}\\\\\n", "\\end{split}\n", "$$\n", "\n", "Thus, we should initialize the weight distribution with a variance of the inverse of the input dimension $d_x$.\n", "Let's implement it below and check whether this holds:"]}, {"cell_type": "code", "execution_count": 16, "id": "b38f6321", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:39.508740Z", "iopub.status.busy": "2021-09-16T12:36:39.508266Z", "iopub.status.idle": "2021-09-16T12:36:52.066650Z", "shell.execute_reply": "2021-09-16T12:36:52.066233Z"}, "papermill": {"duration": 12.618969, "end_time": "2021-09-16T12:36:52.066766", "exception": false, "start_time": "2021-09-16T12:36:39.447797", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9Bbm5vdHMgMTAgMCBSIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDkxMS41MjUgMjE2LjY2NTYyNSBdIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovVHlwZSAvUGFnZSA+PgplbmRvYmoKOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEyIDAgUiA+PgpzdHJlYW0KeJzVnUuTJLl1pff5K2IpLejE4+K1FIczbSabDSXaaCGbBYdqSU2rahrZpGTz7+ccwMP9OuJmqTI7qmpQbc2uPMxA+P0cDsfj4sDf/vDyy7/zt3/76Yb/ubnbH/Dvf+Lv3/HnF4efPr4077cUEv7+4fh78HnLOWX89QN+7fLjv7+8/OuL25ovWYpLtd7mH6Q537Ir9fZnful3D79w/PAy/fbLSy5bxNdI2Nr4wo8vvsnmJfoalPxBy77Urd71vYSL1q/5T7eHwn3IW/X7Pygnpq3d/vz97Z9uP95++XdhgPt7/PsH/NvBvfzy19//xw+///4fvvvV7fc/vbS6iZMU/OWKT/VyFS//+PKb25/uBbvNJ9yUe9n9x+929eVPLx7kfuHwf9W0oSwv2btyC2nzjqX9/uPLr357++X/wFf422//9aVtuEml5Vp4J3/7Ly//fPsblPW3t/99++3fv/z33yJ6tznPUp362+8/soxf/Pr7P/zuf/31H3/340+/+PjDj3/96fbrP95+8/KbfrnPh+YjapnU7OR6n0/5Cdh8CFtpLM37FF4H5xQud+L6YrFn2XILuE/X2E/5GbEnXJawNB+ifF7sXseuymooKvjUHP7BT1vRpYRrKf/0t7dUt9QiKuHffP/Dv/37X27/8bsPP315qp/farwPaHJbxcdj3Uod4X0C5VcLt8oWm9Qar+Ge8nvDxaPDcKtDo5J8SwXFfSJi7/Dnq0XtUZ1rEC/hGrbSf2bcPvgti0gSV9unbnX4uoHjvyHheb6GfVd/btA5bq5EthhV4ieCjnPQf3phMb9ggegn8Eobimop+muTMzUW/+2Pf/3xLz8TXhyBhlZdZCtV0KGQ0bUIES9CtKl750J6X2QLGeEF3Nqd9OXDN/3hl+nDLy8FUdUmOV7amZLwtKWAB0Q3M5PaSq5UryXsKjonv/r/P3gjTARftuK8i/Ea/FW9B28iWSN4I8yPL2gencu4uEvwV7VUl+QByV1dIngrTAQfNjzf2V3v/FW9h2kjWSN4I0wEj86ar7FOd/5U8Zy7gMHUFYlS1wjeCBPBZ3QIxRe5Bn+qCNOPF9KVyC6uEboRJELHKCu0lto19FPle09cyg9A7uoawRthcp5gSzEVud53pcYtSAzOz0gOdYngrTARfNyqYBx5vfNKjVuqxWWZkRzqGsEbYSL4jM5miX6686cq6NYIir8iUeoawRthjrmcFF2b7vypytZiTP4ByaGuEbwRJsYZzm0FH6nXW6/lvEnOsZUrFKUuEb4ZKONH250RSpviP+W6lYzy64RFyYsAMCIlAPTai5OUJgCHnDxudSg5X7loeREARqQEULZccpgfgLuayoZPYig+UVHyIuE/xsno29ZqcGG+/Yec4ya+TVCUuEjwRpiI3oct1FL9NfhDLQEBJHH5ikTLa4RvxMnohVO5qU33/pQrbrOMrt1ZhBIXCd4Ik9HjCluLdQr+rvZpcQTcrki0vEj4j3EyenTgXfJ5vveH3Ou7VLlS0eoi4RuBIv6APjyCmcI/VM41xyqt+YmK1tcAYETK+NGN97nEPAE4ZC9+C7G13K5gtLwIACNUEkBfPoTryu4HLfsUtyz4Fj+R0foiCIxYiQD9+VCjmyvBIXPVMvnm9kpwotH6IgiMWLkkjR59FK58XhCcssfb3icXY53QaH0NBFasRIA+fWy1TLXglBmqS665+ojm0BdBYMRKBGjVJeU814JD9rlivJ9w7yc0Wl8EgRErEaBvn7zIXAsO2Wd+SwxVZjRKXwSBESsRoG+TSohzLThkL23LMaaYJzRaXwSBESsQ4IeQo/PliuCUedGCds/PaLS+BgIrViLAqy3X6qZktFP2gQVnjoOvaLS+CAIjViLAFRZJda4Fh8yJkBxKHLNiCo3WF0FgxEoEbfMoosy14JAb3n0os8+AqkK0vAgAI1IA4Pxe5QznFcApFzz7ObuWr1y0vAYAK1ICiFtpvslUA045ty2m7HK8ctHyIgCMSAkgb66VEuYacMh48bfiOCd05aLkRQAYkRIAs3xj8nMNOOSETpBvNV6xKHWR8I04mbGKjr5rKOYa/imjzrsS4xgfKypKXgOAFSkB9JwGV+ME4JQL6nzwPetBF6LkRQAYkRIAanJwNc814JQTIm2x5omLkhcBYERKAOjih5zTXANOmVndJfZMN12IkhcBYERKAHibxyDzLLmS49ZqbWNcrApR8iIAjEgBoKB7j0LDVAOUHPs6aMkPXA55DQBWpASAzr2Ic1MNUHLc8PILD1R2cZHgjSgZPK5QWmnz3T9lxBmKb+2RyV1eBIARKQGgX59SKlN6hJLPvL8rlqWyAc04EX5lkoOPeUoOULJK+ZyoLJUJakZKAOjV54Lh3QTgIpe4A7gWcpcXAWBESgDo1pfQ4lwDXpFNXIsAMEIigLrFUtGWTZFe5HvK91TIz80Ev26d2dLtP19ejeiK4x++u1232DxszeAS7seXWrZUohsvbJdL7aOWhhGMNM5hQI0xFMkj8zHmLLn/bpFU+Hpracs11p4QGKqUOLIEq4QUMsUisfT3QKtoE3zgKIsbJXLc00rcVtFmcuxZuMbq+vYmJloFZlk2bk0LUkq8518VBIHfgdwqfu6q4GbEVnuyVsNQ7MhWSjGgPt4S3kwu5rGA48pWE74ncqSOmxCz3/Nb8OrCzbtlhCIBkPvaN4Y1kkuqt+K3iI+Nr/RhE4Rby63UDR8qZc+TCCFUn28AmwEuhD19IIlDnbkx3Oxrkn4lvuJKSkrt5j0HzbUcq814mbY+reyYZxfzWJgU1C6f9xn3WN1YiwgJ3Q6H+9T3RYYkrYV9sQ5vYSel76GqtTRfdr24jPB8cfgNxN9DCg13ELUhQvfsv7sBBnrNye0LYTE1dnDGQpjgj29Db/jbWBdA7Qmx5ND1IKiRZdezi7grvfyGBzWOSfSweT61tevCZ3hfZGoBF8+rR7AO3ep+OVE2BNg3hqFGhRbyvh7jEB+qt895w3+jtH2RouHmFRSOShAASfI+c++5gCNAiXrnUIlln8tFXWwNOu58wlPpwj7DmXNGXLeGPi43svaLRC0swUuQWylMiG2jmqIa1sotyIXVpLkxbyKbb0Vw5RgpZtTSceWcLkkZwdwSM6z8Pp1UmFOP3+d0CufX633uBQ8Wtz4mPCARFXCUwWtyzffHQlCV9yG5x9XhofdM13Reqh9yYGXIfOTyVmpIo++OyoXmwffhK3M8fdsHdbgLjTmv+FsA2fsQMKEe4NkCsYgmd2SK4E6Fvu05brW4Y6yAp1dSbzeCb2lcBXqLUnDXS+8tioQRIThyjx6n5LnKWNyIkG8WibiVo/kqsbc+czMsr4rXrYOBbSpI7q+J/3KvrbXlHiWaW3A/vrpxH5940z7ex2/9ZNkOQX32BmqX+SS7fdezr3Iv6vX9jP/zd//3+z/jJ72lcTdfeItbwrBomF0THgwXrq4JaB5Yoebk1ciGqtY8pfEo+QJJlfLonsDqcfwC/3nve/rFqlRolXlJqV27F0p+wrb4IHhFoUVkamp8q5VCeIKVwpfjF/nCxTshXB81JT+BH16iaOVZWkh4h7/PUuELMuDrGm/9+Rm4q88ggFdGxKsroe12n2kqEV4xVgiFc0B3Es9xVviCD+ibWpj37kSvI39+XP373Ba+CgLUqYhuq58QnPK7EbiiEbCvFtFclcwNpJ82IvgmIBqTXEp2E4hTfhIIDK7Q1cUrsAb3qeog3woERxnO9RzwCwmlPwkF80gwom3oz2OE8wkW+ZuxSBjvSEvoml5ZnPqzWKDjjk51xbhAEP7rLOrrnhWBQ+xLTXund8U72XIa53PnLf7raZxPzgGho4gBUMY9mLKaMGhCl9KVKatpkktfvvwwF3KXj2mcNUAYERNE4pBcpq39r8omtsVAGKERBEawGMKWOEV8kc8acSlk0RphREwQbasJXb751r8im9gWA2GEBhA5cOKI5kCXiJWMpjO42GfRdCFKXguEFTFBCDruLYfp1iuZ6zs5izzyucuLgTAiJoi8lZLEzzXilDmrXnMNj3zu8mIgjIgJoiEg79tcI045bjF432eGr3wOeTEQRsQAwUWAmtuUE6VldORrGkvfVz6HvBYIK2KCiPRzLJNxgJZl8+i0pjLxUfJiIIyICQKX1GpKMoE4ZcEbE5348MDnkBcDYURMEHWLTmKca8Qppy0jqhgmPkpeDIQRMUBUrvE5P1kqaLlsKCn15ChdiJLXAmFFTBCBq7HNTTXilDHOihmNw8RHy4uBMCImiLF01uYaccjMHq+udSMlzUfJi4EwIiYILhsWKXONOORUtxaGu4LGc6qLYTDiJQaMHWIMea4Ph5xxzXgExsvzLETLi4EwIgaIhrFDbJcVjQ9KLcL16VwnOlpeC4MRLykIMylqnDEccu3T5KFc4ChxMQhGuKSAUUNyeZqhO9TGeebkxpjzLELLi2F4jJcUMGRIWR4eiUPueSn42iYXPBd5MRBGyB9fovPsBUyJd6farUhw38e7QvHR+lIgrIjJgYvS5Uh9OUAcMpOcim/SM4cUIC0vBsIImSQSs2ono4ZTZVJclTLMCDQfrS8G4jFicqhbKI2L6lcQh8zMr1YbU/8ugLS8GAgjZKZ/MEcvxcmyQMndoKKkJnUmpPS1UFgxj0yY2rz309NxypcsTY1I64uhMGImirTh1/C7E4pDZqJpKrH2KfsLIqUvhsKImSiY3Iqu8lwrDpkJnmgnXd+vqxFpfTEURsxEgbeAw3tgrhWHzKRbujf0fVsakdYXQ2HEDBQhcBKWib8XFKfMPOgS3Ni+oBFpfS0UVsxEgZEEApKpVpwy89Nd2U8g0Yi0vhgKI2aiwFgi9LT0K4pD5oYFX10pfkKk9cVQGDETBYYTEd2lMKE4ZG4uCamV5idEWl8MhREzUESMJ7i75mqDoeTm8aKoNcYrIS2vBcKKmCBoe/iQp6/kSgeE6EaVOAvR8mIgjIgJgoev1ZbnGnHI7EXlNvbrqEK0vBgII2KC4G4cKWmuEYecZfOpllyufLS8GAgjYoAQDCayS3GqEadMu3DpB0Ze+Gh5LRBWxASBoUTO110KH7SceBhZrKNfpfgoeTEQRsQEgYFECTzR8wrilBu+JIe+QVIXouTFQBgRE0RhYnedLBS0zBnc6GTmo+TFQBgRE0TbSo251AnEKSe8MdF/ihMfJS8GwogYIFJ4PGbvw0Wm23oo3TxAF6LktUBYEROEWCcLajlutfri88RHyYuBMCImiGydMqhlnjexb1G+8jnkxUAYERNEs04c1DJ3PTdXH/kc8mIgjIgBgtvwHk8f1LLKOb7yWTMV2YyYIKJ1EqGWVc7xxGfJVGQzYoJI1qmEWtY14srnWTVC75J66X4kt8/E8uhHMu+xSfRcGLnnKbZuIYMrT+GeZI7+QEx1WA6FnEsfRQY6sBeXh+VUc7t3BJMLk2T8ufVApe6piGwwUkzdggS3J429HaWg9yndBCVvoYiMfLSCpyrgG3q+HgYtbeRfcDiPoGjy1nqr02c4mLQlrZba/R9TlrDnckVuFKRpSopb5M6JrgoPU/Do7qfE7RS57PlO9Ifo6U6+yp4xWjFGwtgx5FumZUFwfckq1IrAaLfEPZRZ+uws18Hd5qJDjeBeTVyO25NGPErsEQAN6kXp6cmNB0pHGqQ0eq/ck3Ub++CNTijdoSTIOOCPK8suFlqh0PvDee96gju+june9EKhrQZaZBl9OFfoRZH6KlzYcm6xn5bFVbiE4GgZzWtPxYXx++jqgLqL3fEDwyLv70tVKIvGHTQsyaiqY10Teq051bav2+Q2Zhw8XURSi8OABPWvO1lwmQdXFrggzLMLOKMvu55RuB/+I81X10bxTL6L6IJ33Qk9a7ouXGDPtFUprPG1pbrrXEVr4+qlxdr2JRUMcCNtVXLlzvu4F4/Qq3NcqO+nSZS99ApolTeZyw4tperjPgePCHnJ3GUo6BKP4gN+gCpc78+oS73EMTsLlJWHVYRu7cPx9piqpIEI80h6h6oOOXpu4qAPAeqb4xpxRx8DU1LxWNHXGIxa9+7gVEZplWc/MCS539iIWoSmhzMcoRvCd5OUiEoRaN6ROd7H7YvtPjuAMFiH8BDwVO5uz8OxIHjjDwfFbLn6bkpEuEmtwgzy2j2ABpaePIyuYM+eRecQj2CX0YyhIWLblsHHj0PTOapoGW2e7xadGGbuhaDiOWbV0f2lVemWL+xwolGiLQlPesRYbMxeoPvVUG9T9y3B8H2MYBEZblR3ke4mJ6hEvR3m3ny3H5XpouDvU6sduCJfRui2/Abrklc2sb/me4GSzf3tH1910MAn3rxR3v72T37HW6xM2MS2/VXGhf3PtzIJ38TKRPjEPuYIp/42eTiJRckXSKoU28rk/AV3/aqfvRFaYk889NNGCCU/wYhCeOZU8aWhtcRb+q1eJukJZiZfkCDPTpZ5geFUn8Gv0IIXbW7LaLU/08rkYTf9l0OQmJyW8Fa6PgSH+gQECUEnvvjY7n4mAl1xLmYmQldv7pl/io3JF6xab2pd3lm3mFB2Xv37bEy+CgJa46B1iP6K4JTf7ZNTRSPgi7vS0I6TEJ+g4A3Hiq8ConVXiRDLFcQpPwlEi5tvDt8QmEP1Nj+XrwKCHoqphCTxSkLpT0JBG5PI7iO69p9sdeI3Y4E2rXDA2SYWp/4sFvRYbAlDXodhwRvsbZSNSSxVI3m/jck72X7VWSAJ1TqVRui5+XgqjZb7dE/3KNVlnOoxB7QEBiteYgjW2TSvyia0xUAYoRFEss6oeVU2+SwGwgiNIIp1Vs0kozXu02FTIXd5MRBGxATRrFNrtExv/lq7u7EuRMmLgTAiBgj8YJxeo2U0ilJC33d+5XPIa4GwIiYIsU6x0XLceGpDT2++8jnkxUAYERNEtk6z0TJNOoJ0E+6Jz11eDIQRMUFU61QbLUe0ii322dErn0NeDIQRMUAkb51uo2Uu1KRYw8RHyWuBsCImiGidcqPlzCN/pc/F60KUvBgII2KCSNaJN1quWy7V9+l9XYiSFwNhREwQ1Tr5Rsm0vi8h9/UWVYiWFwNhRAwQ2Vln4Ci5r14ll/LER8lrgbAiJohgHe6iZB58KXkcvKAK0fJiIIyICSJZh98omceglZL8zEfJi4EwIiaIYp2No2Qe5iDjMApVhlIXw2DESwxty1WSm+vDIdey8ayMfgyKKkTLi4EwIgYIpk80F+pUH07ZO571kkZWsyrloq+FwoqZKIR57G5K11IyDwVyNUlPfdGItL4YCiNmosD4wflpB/Kp0q+jNB4sNQHS+mIgHiMmBwwfXMmTa4OSPQea6ETKBEjLi4EwQuZCN8YPPk4+u6fKk5IKwujZeZqP1tcCYURMDhg+eH5+AnHIPW/O+ZFepQBpeTEQRsgkwUzEGcPQmDVXSg5tRnPKiyGYY2X8teeJ1RnBIV9MOg40pnPHGhCMgMGhMScwz8tbh9ozK70bqYyajtbXAmFETA4YNKALkKb6cMrjuLboehKoBqTkxUAYIZMEhg394LoriLvqU+RBfsMiU/PR+mIgHiMmBwwbUmyTOYOSvTi+HuqYvT4BaXkxEEbIJIFhQ2rcB3clccg8sdKjAnS/L01I64uhMGL++JKc5w/SrpVCyTyRNEss3QxPI9L6UijMmIlCmC8a6pQ+e8o8etXRlX8ipOTFQBgRE0TeUikuz3XikHk8pvejQVCFaHkxEEbEBFE3fKKmuUYccimbpOqlXfloeTEQRsQA4fvmnzkzX8m58IzjcZKuKkTLa4GwIiYIDKSbSJhqxCknbjvC1bcrHy0vBsKImCAwfnAuTO4MSu7HkraxcUrzUfJiIIyICQJDCMddPhOIQ+ZmPlf2NkLxUfJiIIyIASLgY96XMtUIJfOw7RS6Q4kuRMlrgbAiJggMIXxJkw21lhNaRTfaSlXGqS6GwYiXGBIPcY8y14dTlq1KSFUmOkpeDIQRMUHwEO7m41wfTpkz1jX1jcdXPoe8GAgjYoJojwcrfrjIcculby298lHyYiCMiLlHyVtnImo5bjzTNj3yOeS1QFgRE4RYZ0ZqOeI14XPfXnLlc8iLgTAiJohsnYmoZZVTeuWzZqqpGTFBVOvMSC1fQJh8FgNhRMzNNN46SlLLFxAmn7VAWBETRLSOkpzkewL+VMiz8vKfaVsybcxpkZluTKnlgCCF7HvrllPpp3gxd7b4FodXE328+o5+ZpJKKrGMboK0NuSeOydF+lFpnnYme7ZlQ5cC1wa15nv2ROIGqOA5tKPZQcxlpCSiAxYd19cqxmoxuZGfR6PFQvcRjlySl7onq0V8ievmJLSu2Eum1YPwJBG6k5Q0enPM6EIFQSXFx0L04+QZyZ5n8YTc7UlyLqWO3I64oXccWuIsAkaMwwdWcgazHFO81W6pkcccbe5Zc7gfPNQmgtpY4MkNb8zI3XMtbEmcayNZAr3N2kDz5jE0ja7GOnTu8KUnyjgDJJfcD0PhSnoFRURBi4yY/R5+qTy5nLYl7KPUgNvT5UZzkJT2AyXoQNOLr45nWQYJYyG6ou2q+5pk8KX11RfcIAwHRioD9NIKV2sBH1zcUGWLJWcZdiCOljF9bpan+XihFwtdRSItWsquIwznhwuJ79F2vWxoOSO/lT4KHIakXa/ofQ01jREJV8okBscvpZUJrnXcvEq7Fi4O9C9lela6rydBwM/j4tue3Ng8CNRuANPvb+l+MVx1wdeKcDWGM+x4dsav0/mVPh3khXsXuqENlybwiONXwTd1t0c/Juq549t1Wxj6KccYew4MpyhdEfqw4JO4g36XeRCxoyk77iPubuzPUHK4ZaiFzMKrPHKzdd9dTuaABe4dLrxldMPGOLanHXncGaYrhnFmDOc7Gio4HriU8ZzmuheReJhnHdMguFfSl9+Tz3iCHZ8z+pV4tGr9OjxIoOqGPkRGvyfl+4CakJkxzEcy5b277De6V/C4NNYDNA33cRX4heGNiGC6T1AKqFhoVYa7UfNOxjRNROuBK6mjAxrROHROEXcHiP1omWIdvjGJTSR+yMNnqaD+1MeGGXXcfUJ+gzPJK/vSX7OxQMnmlvWPrxpi4BNv3vtuf/snv+MtziRsG9/mSCLfxJEEzwifiTnht/ApfDxmRskXOKoU25Hk+AX+I+96e7+ypxmvpS2wSbl2PpT8BDuJjJYFzUZB7NwL81ZHkicYknxBgPThChikX7uxSn4GQHokcSYArw35BMCLJcm8efvLMSiul5wnJ2MlP4FBcczhYmm9R/R5niSvWJKgNd6402A38XuKMcmXw/u2Ruadji81jk0H4+rfZ0zyVRDQlw7dwG5MohCc8rsR5KYRZPYGS0NHGyOT1yGkRweKr4LB43HwruXuS6JfLKf+fhBBg2gYeAQMcFrl0Xhvc2j5OijQfUMxuRuTaBSn/iQUvGq6BNL+MnySxTerFhhGhBTz1ECc8rNIJO5JiRyBcFfn59vVKFuShDIUkPfbkryT7Fed5ckuWcflZFet43K0TAOSWHpyiS5EyccszxogjIgBglalj4flaPkCwuSzFggrYoII1mE5Wo4cTbe+l3zic5cXA2FETBBiHZbzqmxiWwyEERpBFOuwHC3TO6yOU951IUpeDIQRMUE067AcLSNica5Prk187vJiIIyIOTz31mE5Wo4bPhb6pvorn0NeC4QVMUGIdViOliMqQY7yyOeQFwNhREwQ2TosR8txQ1cQfbYHPoe8GAgjYoKo1mE5WuZBQWk4letClLwYCCNigIjeOixHy3kLnHWf8JzqWhiseIkhWkflaJmn1braneh1IUpeDIQRMUEk66gcJSe3udR8m/hoeTEQRsQEUa2jcpTMNco89jVoPKe6GAYjXmAQZx2Uo+TuQ5LbeGOchWh5LRBWxAQRrINylJwzxlgt9WU+zUfJi4EwIiYIsQ7KUXLhkRGhr2irMpS6GAYjXmKguUjjev4VwyHXsBUuprcrHS0vBsKImCC4eSP1IzguIA651U1a9GO4dRai5cVAGBEDRMK4QXwKU404Ze8yT22JexNxlHLR10JhxUwUwnSoOO0qVrLHsLvw8IQ2IdL6YiiMmImCKS7RtblWHDIvmslI4/lQiLS+GAojZqLA8AGfqXOtOGTPVJ7GxKgJkdYXQ2HEzIV/jCAyO0hXFKfMtRhmF7aJkJLXAmFFTBCRJyTIZMSgZHqx8NS2ngymAWl9MRRGzESBHlLJYXo6DpUZhOl+5pcGpPXFQDxGTA5MaA0uzFXikC+2LBqQ5dayBggjZJDAnQ09f/UC4lCZ4NpcSqOrrfhofS0QRsTkwJPiYmpTjThlJgajQyU9kVQDUvJiIIyQSUK22tpkLH2qPuE7xA8HDs1H64uBeIyYHDCMcAn3dwJxyJ5bHcrurq4AaXkxEEbIJNGdmR4ejbvqo8fDkGOd8Ch5MQyP8YJC9TzwtUwWDErmwCKHkHrGk8Kj5bVAWCGTBAYQCGjyYFAyR5shjBw4DUjJi4EwIiYIDB9CjW6uEofcOGkbQj//WBWi5cVAGBETBEYPj+n5Sq6h7yUO/spHy4uBMCIGiMbzlVstU4045cIMAB7YfOWj5bVAWBETBAYPkpiyegVxyNxGFtw4d1kVouXFQBgREwT6h4k7jSYQh0y7eXQcRg/7LETLi4EwIiYIHoVdQpxrxCEn6R65/fgizUfJi4EwIuZuBBYb3eTCoOXGBT7p+8d0IUpeCoQZMUFg8IA2cHJh0DKfgZr6UqcuRMmLgTAiJgieri2pzjXilDM6kxhdTXhOdTEMRrzEUB4PTPxwkdMWYrxzOAtR8mIgjIjHfiXjtEMtC/d6hwc+Sl4MhBExQHhvnXao5bhJdOP8BV2IktcCYUVMEGKdBqllHnBWW+9JT3zu8mIgjIgJIluHRGpZpdhe+ayZeWtGTBDVOiRSyyrpeuKzZC62GTE3F3nrkEgtq/0IVz5P26bwTHuSaW8N/pTWc87z5pOMQz597vYMOaAvEGiVQBl3Nx1ZtSFLHT5NuIjY2p5j6hwG3p4vitjocrGnXMbgguupp4LXaHcCRFdry7GkktjPKPh1GXloGLKXlHluDHqmAf3ysKfppShcpG+by3mYUOQoiDU0XCx9URCUL3suW00+50C/hZoa3uVdLluqDYNBmjakBGCjkIbrw0PMhC8eNhPv6U9o7IW+D3FrQkOOPRuIEyvAnNFdruK6rwiTY5LgLVnoNBFwB/OQORHjpMVb5WPBrIA9g0RipdVKC5vP4tM938SXnNIwIsZFuzGMRxCuRPqn0MwEXZLWnVJyijRZaSmMg8Qzi9xXnYureAJpWsKGOvVpEe6tzpmXQNOSFiodQ6g3+jPypEKacKAxb/tqFXcV4ho4J1+3jMscqQ6ZFdt1bxVmhqGwcfkZtSI26QublS6YSequo05XH7oHP0cY/VwnrnphTO739b/oayt+19GmSqtdp4tFf89yp3lCz1SG/UltqKSyrxrR8qOO9UWuH8VdptFL9znp93KXcX+a1HFQQG1uTxTnntgUu6MLPVck+rHtCB/NPnRHF1yXlLDvuaDRSgOX1H1OKurpqGNcwEsu0KOFHjkCih1aQf11tTvApDh2K45pfR7TgErPDAHe5prGHBaqEQZljqk1eH7cAQf1KKZSuN0eHdQECN2rI1ceV9zrP0a2tcmoxBUVzZMs6x96s37UhFq3hJYKjx6+JvOXe7WsqNHRpexvBc+vG291PF9bDnjF4UFgj9jVcp9p4c5pPGPZ44ltpe8+RWF4WGgNxAkGNErD/xSVju45vJGsYqHIuBAmUAVX8JAlv4EFvr3LDdfX+OSzohZ8V9hHaCWy5nXjItRbN7plfuMuoNxbjCy57jKqaeqbxuntE0K//ezVo2Qp3T+pgfx4kTvUCjQqcXR9C12RemtO42+0Rr2prLl0f57e/ZHUF7Z6qzSqECpbTygfh6Ci8fP9gMPpRcFnVF6X32CX8spu+dc8NlCyuZH+46tuHTRYeeuOfPvbP/kdb7FLYXZuG24GPW/g821T8jexTcEIAY/9Q0pyczw4+Arvrl3wqM/bhinnL/jrl/x8rws0EqHgtTf1C0/5GV4XaEaFG6vxiKOheKthSnqCY8qXI1jR34joG037VZT8BIKVdrI18tBItO6f6ZjysGX/C0KgWxszWq656Ep+BoTClEOW5gVX+XmWKekVz5Tq0BF2XDbpDoJP8Uz5go/p57cw73xEma92Xvr7DFO+Svw8lKG05r2K/9TeHT+N6c5LL3QTdOhM0A/ik3YQ34SB9zwlsieQnRCU+CQK3mEs4YU92oC+4+sY5Bth4NS1q2wMFIZTfBYGCd3tEIMe+aRNSn7NGqTIhcb7rUHeifXrzrDQBvPxOBq0J9ZxNFrmMN/HvkSnC1HyMcOyBggjYoIo1nE0Ws7oLqFcmfgoeTEQRsQAQXPRx+NotHypEWchy9YIK2KCCNZxNFqml+nepulClLwYCCNighDrOBot60dDFbLso2FFTBDFOo5GyRlfEkLpFrOqEC0vBsKImCCadRyNklM/TGH0/VQhWl4MhBExhlLOW8fRKJkWZmLwUfJSIMyICUKsA2mUrCN+hc9iIIyICSJbB9IouU+978+AKkTLi4EwIiaIah1Io+TC2Pf3pCpEy4uBMCIGCO+sA2mUXBrz6h/5KHktEFbEBBGtA2mUzM5k3MfKqhAtLwbCiJggknUgjZJrw9/ubYTio+TFQBgRE0SxDqRRMrcV1HsfUpVy0RdDYcQMFMFZR9IomevPxR+V4kSk9bVQWDETRbAOpVEy/YdzvKNQiLS+GAojZqKQra8uzLXikJl9kO9jC41I64uhMGImisKcHt/mWnHIPD3Gp/vrQyHS+mIojJiJom1ctqtzrThkOh6E402qEGl9MRRGzFzQ88xCKpMTgpKZP4NC926mQqT1tVBYMRNF3BrPfpEJxSHTT9vH+8tUI1L6YiiMmIkCowknnIu8ojjky2Z/jcg0AVgDhREzUWA8gfdimGvFIXNPc4z7xK1GpPXFUBgxA4UwYWQ6f+iDln0/Z+1AcSLS+loorJiJIg4DzqlWnPJlwU4j0vpiKIyYiSLxqLSelXZBccjXl6lGtOzL1IqZKDjQjiHPteKQmdTq0v0NohEpfTEURsxAkTCoiM1dszhOtWfxHiN0DUjpa4EwIiYHjChEmJJ5BXHIHr3LFB4BaXkxEEbIJIEBRXLzTP+heof+pNwHYoqP1hcD8RgxOWA0kbI8PBqH3OgN8sBHq4thMAImBwwlMk8rumK4q9eZq7OMhWeujIjBYZwr28pUH07Zu4jPGYCUvBYIK2SSwDCC6bkTiLvKo2XrvYE4i9DyYhge4yWFvIXH/HQl67zAswytLobBCJgcOL+Q4uSJoOTLQuBZyLoLgVbEADE2DHk/PRenzB1L6T7qUnyUvBYIK2KC4D40/u4E4pB1eoQqZNmsCTNigsDAwYVc5xpxyDphRhWybB6NGTFBYOCAwXSZa8QhczthuT8aZyFaXgyEETG3JDge2U1/9wsIJVeaKN1rxFmIktcCYUVMEIG78pxMNULJGoTNZzEQRsQEIY9HGX64yCrfVBeyahqqGTFBFOvwRi2rNEtdyKrZl2bEBNGswxtflU1si4EwQgMIDKWNUwhfk20+a4GwQiOIaJ3S+Kps8lkMhBEaQWTr8EYt68ZSFbJsY2lFTBDVOrxRyxcQJp+fC+KZhiHXHTfiHC0b9m07tfb97jTIGpYD3IxBq4ra3wmuZPxal2nbgd6S5yZ9z2N0Rlo2Bhgu8GAaHo+QhllCAarIRY1Gt4AWXOzetj11uYWKAtH3dC21OrIVHa38EnuqYCU1dzcS5vdW7wKC5XRPkSpDJkEaJNC3ACDd6MU4mqeWHBpdDgK6ud16o7rU96am2keCVVK3C2BeYMKbnnmBAWwCIYwsuVayZNp6ZGm+5j03CFedCu07pFtCNNkTZTL+9IlqhlCd+D1rRJzkKH0CO8dIDwvqQJSd43wu/VViCHvCTeMBRYW+uOyBRVSmsaLs0EEv4mg54cEr5+56yUX3iPBoqU0zE4zu4i63iBsQ+1Kja6XFsUobNtAV+uvSuSLwFMF9wdo1KVx54YkfvqZ213MRH4adCAIPcaz1C0/mFjo3Uwee7tNAHd+BvnTX6TM2rhL1uXlpY+UzeBkGx1wRja7yzAQ6guOpGhWMy4OAVttYFwGdGPy+VpZCaCn2ZcMsaTjyVD5tFaFz4SixuqX993HHxY39gvTskOGeUvHh1ho3lnNR2iF0P74XyHFv8WrzsWwFNX9g4EUwJYy5P3RhCa4b2VTeiiRCNxR6ImSaJeyT1Cm67p7iA/rR6FcPHe1Drt0NxWUar+TuNlFTP6es25u4vjIaun8OJ/WkBhqT8NkqLoxKiFodC31ouKAex3mgnPDJqBYlcFsrmmTXPcU5/SGoRRzsNlpTjEXH4niUJr4cRbXYWFGootbgRqbGxwm0nGv7eDlLpU0JU1yKc87vo0d8H6cTEhqBNhyKKy1zcL9jt+sBM3cMvFDdG52AmHNK15+9mx1jt8ChQZDD3Qt7pxOtDf1IuDqM1mg8NGikKioEfXQTnr2AtmXvmSXgHo1uiHFcNRqpgGjQUAhLS7l7cvCl7ULrviO0bqnZyf4KQ3WW7gKCoMaJebqZR/XBwzyqlS2/we7jlY3erzlEoOTHPeAfbZcJWoO8aRu5/aWvl/4Wiw88k2jEx+Z7rvV/vsVHvb2yo1/YiqP956Z1jxtRr4WJtaUfzWzBYBbN9n1f/7/88NNf/vzD//nrX37444+X3bcv/w+G8E1VCmVuZHN0cmVhbQplbmRvYmoKMTIgMCBvYmoKMTExMDAKZW5kb2JqCjEwIDAgb2JqClsgXQplbmRvYmoKMTcgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzUgPj4Kc3RyZWFtCnicNVFJbgAxCLvnFf5ApbAn75mq6qH9/7WGUS8DA9jYJO/BRiQ+xJDuKFd8yuo0y/A7WeTFz0rh5L2ICqQqwgppB89yVjMMnhuZApcz8VlmPpkWOxZQTcRxduQ0g0GIaVxHy+kw0zzoCbk+GHFjp1muYkjr3VK9vtfynyrKR9bdLLdO2dRK3aJn7Elcdl5PbWlfGHUUNwWRDh87vAf5IuYsLjqRbvabKYeVpCE4LYAfiaFUzw6vESZ+ZiR4yp5O76M0vPZB0/W9e0FHbiZkKrdQRiqerDTGjKH6jWgmqe//gZ71vb7+AENNVLkKZW5kc3RyZWFtCmVuZG9iagoxOCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDYxID4+CnN0cmVhbQp4nDM1NVcwULC0ABKmpkYK5kaWCimGXEA+iJXLZWhpDmblgFkWxkAGSBmcYQCkwZpzYHpyuDK40gDLFRDMCmVuZHN0cmVhbQplbmRvYmoKMTkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA5MiA+PgpzdHJlYW0KeJw9jcENwDAIA/9MwQgQAsT7VFUf6f7fJhHqBx8G2RhgYbM14MHZwJfS2je9pEWT2ghWtUXdUJ67FKVYXUelTMJPmTt/UnQc7XAO29/W5ThN4+hf99D9AQ9KHgsKZW5kc3RyZWFtCmVuZG9iagoyMCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMwNyA+PgpzdHJlYW0KeJw9kktuAzEMQ/c+hS4QwPrZnvOkKLqY3n/bJyXpihzZFkVqlrpMWVMekDSThH/p8HCxnfI7bM9mZuBaopeJ5ZTn0BVi7qJ82cxGXVknxeqEZjq36FE5Fwc2Taqfqyyl3S54Dtcmnlv2ET+80KAe1DUuCTd0V6NlKTRjqvt/0nv8jDLgakxdbFKrex88XkRV6OgHR4kiY5cX5+NBCelKwmhaiJV3RQNB7vK0ynsJ7tveasiyB6mYzjspZrDrdFIubheHIR7I8qjw5aPYa0LP+LArJfRI2IYzcifuaMbm1MjikP7ejQRLj65oIfPgr27WLmC8UzpFYmROcqxpi1VO91AU07nDvQwQ9WxFQylzkdXqX8POC2uWbBZ4SvoFHqPdJksOVtnbqE7vrTzZ0PcfWtd0HwplbmRzdHJlYW0KZW5kb2JqCjIxIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ0ID4+CnN0cmVhbQp4nEWRTXIFIQiE956iL/Cq5Fc9z6RSWUzuvw3NvCQrWoXmA9MCE0fwEkPsiZUTHzJ8L+gyfLcyO/A62ZlwT7huXMNlwzNhW+A7Kss7XkN3tlI/naGq7xo53i5SNXRlZJ96oZoLzJCIrhFZdCuXdUDTlO5S4RpsW4IU9UqsJ52gNOgRyvB3lGt8dRNPr7HkVM0hWs2tExqKsGx4QdTJJBG1DYsnlnMhUfmqG6s6LmCTJeL0gNyglWZ8elJJETCDfKzJaMwCNtCTu2cXxppLHkWOVzSYsDtJNfCA9+K2vvc2cY/zF/iFd9//Kw591wI+fwBL/l0GCmVuZHN0cmVhbQplbmRvYmoKMjIgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzEgPj4Kc3RyZWFtCnicNU85kgQhDMt5hT4wVRjbQL+np7Y22Pl/upKZTpDwIcnTEx2ZeJkjI7Bmx9taZCBm4FNMxb/2tA8TqvfgHiKUiwthhpFw1qzjbp6OF/92lc9YB+82+IpZXhDYwkzWVxZnLtsFY2mcxDnJboxdE7GNda2nU1hHMKEMhHS2w5Qgc1Sk9MmOMuboOJEnnovv9tssdjl+DusLNo0hFef4KnqCNoOi7HnvAhpyQf9d3fgeRbvoJSAbCRbWUWLunOWEX712dB61KBJzQppBLhMhzekqphCaUKyzo6BSUXCpPqforJ9/5V9cLQplbmRzdHJlYW0KZW5kb2JqCjIzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ5ID4+CnN0cmVhbQp4nD1QO45EIQzrOYUv8CTyI3AeRqstZu/frgOaKVBMfrYzJNARgUcMMZSv4yWtoK6Bv4tC8W7i64PCIKtDUiDOeg+IdOymNpETOh2cMz9hN2OOwEUxBpzpdKY9ByY5+8IKhHMbZexWSCeJqiKO6jOOKZ4qe594FiztyDZbJ5I95CDhUlKJyaWflMo/bcqUCjpm0QQsErngZBNNOMu7SVKMGZQy6h6mdiJ9rDzIozroZE3OrCOZ2dNP25n4HHC3X9pkTpXHdB7M+Jy0zoM5Fbr344k2B02N2ujs9xNpKi9Sux1anX51EpXdGOcYEpdnfxnfZP/5B/6HWiIKZW5kc3RyZWFtCmVuZG9iagoyNCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM5NSA+PgpzdHJlYW0KeJw9UktuxUAI2+cUXKDS8JvPeVJV3bz7b2tDUqkqvIkxxjB9ypC55UtdEnGFybderls8pnwuW1qZeYi7i40lPrbcl+4htl10LrE4HUfyCzKdKkSozarRofhCloUHkE7woQvCfTn+4y+AwdewDbjhPTJBsCTmKULGblEZmhJBEWHnkRWopFCfWcLfUe7r9zIFam+MpQtjHPQJtAVCbUjEAupAAETslFStkI5nJBO/Fd1nYhxg59GyAa4ZVESWe+zHiKnOqIy8RMQ+T036KJZMLVbGblMZX/yUjNR8dAUqqTTylPLQVbPQC1iJeRL2OfxI+OfWbCGGOm7W8onlHzPFMhLOYEs5YKGX40fg21l1Ea4dubjOdIEfldZwTLTrfsj1T/5021rNdbxyCKJA5U1B8LsOrkaxxMQyPp2NKXqiLLAamrxGM8FhEBHW98PIAxr9crwQNKdrIrRYIpu1YkSNimxzPb0E1kzvxTnWwxPCbO+d1qGyMzMqIYLauoZq60B2s77zcLafPzPoom0KZW5kc3RyZWFtCmVuZG9iagoyNSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI0OSA+PgpzdHJlYW0KeJxNUUmKAzAMu+cV+kAhXpO8p0OZQ+f/18oOhTkECa+Sk5aYWAsPMYQfLD34kSFzN/0bfqLZu1l6ksnZ/5jnIlNR+FKoLmJCXYgbz6ER8D2haxJZsb3xOSyjmXO+Bx+FuAQzoQFjfUkyuajmlSETTgx1HA5apMK4a2LD4lrRPI3cbvtGZmUmhA2PZELcGICIIOsCshgslDY2EzJZzgPtDckNWmDXqRtRi4IrlNYJdKJWxKrM4LPm1nY3Qy3y4Kh98fpoVpdghdFL9Vh4X4U+mKmZdu6SQnrhTTsizB4KpDI7LSu1e8TqboH6P8tS8P3J9/gdrw/N/FycCmVuZHN0cmVhbQplbmRvYmoKMjYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA5NCA+PgpzdHJlYW0KeJxFjcERwCAIBP9UQQkKCtpPJpOH9v+NEDJ8YOcO7oQFC7Z5Rh8FlSZeFVgHSmPcUI9AveFyLcncBQ9wJ3/a0FScltN3aZFJVSncpBJ5/w5nJpCoedFjnfcLY/sjPAplbmRzdHJlYW0KZW5kb2JqCjI3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzQxID4+CnN0cmVhbQp4nEVSS25EMQjbv1NwgUjhl5DztKq6mN5/W5tM1c3gCWBseMtTpmTKsLklIyTXlE99IkOspvw0ciQipvhJCQV2lY/Ha0usjeyRqBSf2vHjsfRGptkVWvXu0aXNolHNysg5yBChnhW6snvUDtnwelxIuu+UzSEcy/9QgSxl3XIKJUFb0HfsEd8PHa6CK4JhsGsug+1lMtT/+ocWXO9992LHLoAWrOe+wQ4AqKcTtAXIGdruNiloAFW6i0nCo/J6bnaibKNV6fkcADMOMHLAiCVbHb7R3gCWfV3oRY2K/StAUVlA/MjVdsHeMclIcBbmBo69cDzFmXBLOMYCQIq94hh68CXY5i9Xroia8Al1umQvvMKe2ubnQpMId60ADl5kw62ro6iW7ek8gvZnRXJGjNSLODohklrSOYLi0qAeWuNcN7HibSOxuVff7h/hnC9c9usXS+yExAplbmRzdHJlYW0KZW5kb2JqCjI4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTY0ID4+CnN0cmVhbQp4nEWQx3EFMQxD76oCJTCACvWsx/MP6/6vhvTTQXoYQgxiT8KwXFdxYXTDj7ctMw1/RxnuxvoyY7zVWCAn6AMMkYmr0aT6dsUZqvTk1WKuo6JcLzoiEsyS46tAI3w6sseTtrYz/XReH+wh7xP/KirnbmEBLqruQPlSH/HUj9lR6pqhjyorax5q2leEXRFK2z4upzJO3b0DWuG9las92u8/HnY68gplbmRzdHJlYW0KZW5kb2JqCjI5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNzIgPj4Kc3RyZWFtCnicMzK3UDBQsDQBEoYWJgrmZgYKKYZcQL6piblCLhdIDMTKAbMMgLQlnIKIZ4CYIG0QxSAWRLGZiRlEHZwBkcvgSgMAJdsWyQplbmRzdHJlYW0KZW5kb2JqCjMwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNDcgPj4Kc3RyZWFtCnicMzK3UDBQsDQBEoYWJgrmZgYKKYZclhBWLhdMLAfMAtGWcAoinsGVBgC5Zw0nCmVuZHN0cmVhbQplbmRvYmoKMzEgMCBvYmoKPDwgL0JCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzkKL1N1YnR5cGUgL0Zvcm0gL1R5cGUgL1hPYmplY3QgPj4Kc3RyZWFtCnic4zI0MFMwNjVVyOUyNzYCs3LALCNzIyALJItgQWQzuNIAFfMKfAplbmRzdHJlYW0KZW5kb2JqCjMyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTYzID4+CnN0cmVhbQp4nEWQOxIDIQxDe06hI/gjAz7PZjIpNvdvY9hsUsDTWCCDuxOC1NqCieiCh7Yl3QXvrQRnY/zpNm41EuQEdYBWpONolFJ9ucVplXTxaDZzKwutEx1mDnqUoxmgEDoV3u2i5HKm7s75Q3D1X/W/Yt05m4mBycodCM3qU9z5NjuiurrJ/qTH3KzXfivsVWFpWUvLCbedu2ZACdxTOdqrPT8fCjr2CmVuZHN0cmVhbQplbmRvYmoKMzMgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTggPj4Kc3RyZWFtCnicPVC5jQQxDMtdhRpYwHrtqWcWi0um//RI+fYi0RZFUio1mZIpL3WUJVlT3jp8lsQOeYblbmQ2JSpFL5OwJffQCvF9ieYU993VlrNDNJdoOX4LMyqqGx3TSzaacCoTuqDcwzP6DW10A1aHHrFbINCkYNe2IHLHDxgMwZkTiyIMSk0G/65yj59eixs+w/FDFJGSDuY1/1j98nMNr1OPJ5Fub77iXpypDgMRHJKavCNdWLEuEhFpNUFNz8BaLYC7t17+G7QjugxA9onEcZpSjqG/a3Clzy/lJ1PYCmVuZHN0cmVhbQplbmRvYmoKMzQgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA4MyA+PgpzdHJlYW0KeJxFjLsNwDAIRHumYAR+JvY+UZTC3r8NECVuuCfdPVwdCZkpbjPDQwaeDCyGXXGB9JYwC1xHUI6d7KNh1b7qBI31plLz7w+Unuys4obrAQJCGmYKZW5kc3RyZWFtCmVuZG9iagozNSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDUxID4+CnN0cmVhbQp4nDM2tFAwUDA0MAeSRoZAlpGJQoohF0gAxMzlggnmgFkGQBqiOAeuJocrgysNAOG0DZgKZW5kc3RyZWFtCmVuZG9iagozNiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2MCA+PgpzdHJlYW0KeJxFkDkSAzEIBHO9gidIXIL3rMu1wfr/qQfWR6LpAjQcuhZNynoUaD7psUahutBr6CxKkkTBFpIdUKdjiDsoSExIY5JIth6DI5pYs12YmVQqs1LhtGnFwr/ZWtXIRI1wjfyJ6QZU/E/qXJTwTYOvkjH6GFS8O4OMSfheRdxaMe3+RDCxGfYJb0UmBYSJsanZvs9ghsz3Ctc4x/MNTII36wplbmRzdHJlYW0KZW5kb2JqCjM3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzM0ID4+CnN0cmVhbQp4nC1SS3LFIAzbcwpdoDP4B+Q86XS6eL3/tpKTRUYOYPQx5YaJSnxZILej1sS3jcxAheGvq8yFz0jbyDqIy5CLuJIthXtELOQxxDzEgu+r8R4e+azMybMHxi/Zdw8r9tSEZSHjxRnaYRXHYRXkWLB1Iap7eFOkw6kk2OOL/z7Fcy0ELXxG0IBf5J+vjuD5khZp95ht0656sEw7qqSwHGxPc14mX1pnuToezwfJ9q7YEVK7AhSFuTPOc+Eo01ZGtBZ2NkhqXGxvjv1YStCFblxGiiOQn6kiPKCkycwmCuKPnB5yKgNh6pqudHIbVXGnnsw1m4u3M0lm675IsZnCeV04s/4MU2a1eSfPcqLUqQjvsWdL0NA5rp69lllodJsTvKSEz8ZOT06+VzPrITkVCaliWlfBaRSZYgnbEl9TUVOaehn++/Lu8Tt+/gEsc3xzCmVuZHN0cmVhbQplbmRvYmoKMzggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAzMjAgPj4Kc3RyZWFtCnicNVJLbgUxCNvPKbhApfBPzvOqqou++29rE70VTDBg4ykvWdJLvtQl26XD5Fsf9yWxQt6P7ZrMUsX3FrMUzy2vR88Rty0KBFETPViZLxUi1M/06DqocEqfgVcItxQbvINJAINq+AcepTMgUOdAxrtiMlIDgiTYc2lxCIlyJol/pLye3yetpKH0PVmZy9+TS6XQHU1O6AHFysVJoF1J+aCZmEpEkpfrfbFC9IbAkjw+RzHJgOw2iW2iBSbnHqUlzMQUOrDHArxmmtVV6GDCHocpjFcLs6gebPJbE5WkHa3jGdkw3sswU2Kh4bAF1OZiZYLu5eM1r8KI7VGTXcNw7pbNdwjRaP4bFsrgYxWSgEensRINaTjAiMCeXjjFXvMTOQ7AiGOdmiwMY2gmp3qOicDQnrOlYcbHHlr18w9U6XyHCmVuZHN0cmVhbQplbmRvYmoKMzkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxOCA+PgpzdHJlYW0KeJwzNrRQMIDDFEOuNAAd5gNSCmVuZHN0cmVhbQplbmRvYmoKNDAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxMzMgPj4Kc3RyZWFtCnicRY9LDgQhCET3nKKOwMcf53Ey6YVz/+2AnW4TYz2FVIG5gqE9LmsDnRUfIRm28beplo5FWT5UelJWD8ngh6zGyyHcoCzwgkkqhiFQi5gakS1lbreA2zYNsrKVU6WOsIujMI/2tGwVHl+iWyJ1kj+DxCov3OO6Hcil1rveoou+f6QBMQkKZW5kc3RyZWFtCmVuZG9iago0MSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM0MCA+PgpzdHJlYW0KeJw1UjluBDEM6/0KfSCAbtvv2SBIkfy/DanZFANxdFKUO1pUdsuHhVS17HT5tJXaEjfkd2WFxAnJqxLtUoZIqLxWIdXvmTKvtzVnBMhSpcLkpORxyYI/w6WnC8f5trGv5cgdjx5YFSOhRMAyxcToGpbO7rBmW36WacCPeIScK9Ytx1gFUhvdOO2K96F5LbIGiL2ZlooKHVaJFn5B8aBHjX32GFRYINHtHElwjIlQkYB2gdpIDDl7LHZRH/QzKDET6NobRdxBgSWSmDnFunT03/jQsaD+2Iw3vzoq6VtaWWPSPhvtlMYsMul6WPR089bHgws076L859UMEjRljZLGB63aOYaimVFWeLdDkw3NMcch8w6ewxkJSvo8FL+PJRMdlMjfDg2hf18eo4ycNt4C5qI/bRUHDuKzw165gRVKF2uS9wGpTOiB6f+v8bW+19cfHe2AxgplbmRzdHJlYW0KZW5kb2JqCjQyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjUxID4+CnN0cmVhbQp4nC1RSXIDQQi7zyv0hGan32OXK4fk/9cIygcGDYtAdFrioIyfICxXvOWRq2jD3zMxgt8Fh34r121Y5EBUIEljUDWhdvF69B7YcZgJzJPWsAxmrA/8jCnc6MXhMRlnt9dl1BDsXa89mUHJrFzEJRMXTNVhI2cOP5kyLrRzPTcg50ZYl2GQblYaMxKONIVIIYWqm6TOBEESjK5GjTZyFPulL490hlWNqDHscy1tX89NOGvQ7Fis8uSUHl1xLicXL6wc9PU2AxdRaazyQEjA/W4P9XOyk994S+fOFtPje83J8sJUYMWb125ANtXi37yI4/uMr+fn+fwDX2BbiAplbmRzdHJlYW0KZW5kb2JqCjQzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTc0ID4+CnN0cmVhbQp4nE2QSQ5DIQxD95zCF6iEM8DnPL+qumjvv61DB3WB/OQgcDw80HEkLnRk6IyOK5sc48CzIGPi0Tj/ybg+xDFB3aItWJd2x9nMEnPCMjECtkbJ2TyiwA/HXAgSZJcfvsAgIl2P+VbzWZP0z7c73Y+6tGZfPaLAiewIxbABV4D9useBS8L5XtPklyolYxOH8oHqIlI2O6EQtVTscqqKs92bK3AV9PzRQ+7tBbUjPN8KZW5kc3RyZWFtCmVuZG9iago0NCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDc1ID4+CnN0cmVhbQp4nDO1NFIwUDA2ABKmZkYKpibmCimGXEA+iJXLZWhkCmblcBlZmilYWAAZJmbmUCGYhhwuY1NzoAFARcamYBqqP4crgysNAJWQEu8KZW5kc3RyZWFtCmVuZG9iago0NSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE0MSA+PgpzdHJlYW0KeJw9j8EOwzAIQ+/5Cv9ApNgpoXxPp2qH7v+vI0u7C3oCY4yF0NAbqprDhmCb48XSJVRr+BTFQCU3yJlgDqWk0h1HkXpiOBhcHrQbjuKx6PoRu5JmfdDGQrolaIB7rFNp3KZxE8QdNQXqKeqco7wQuZ+pZ9g0kt00s5JzuA2/e89T1/+nq7zL+QW9dy7+CmVuZHN0cmVhbQplbmRvYmoKNDYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTUgPj4Kc3RyZWFtCnicNVE5DgMhDOz3Ff5AJIwveE+iKM3+v82M0VYewVyGtJQhmfJSk6gh5VM+epkunLrc18xqNOeWtC1zgLi2vC+tksCJZoiDwWmYuAGaPAFD19GoUUMXHtDUpVMosNwEPoq3bg/dY7WBl7Yh54kgYigZLEHNqUUTFm3PJ6Q1v16LG96X7d3IU6XGlhiBBgFWOBzX6NfwlT1PJtF0FTLUqzXLGAkTRSI8+Y6m1RPrWjTSMhLUxhGsagO8O/0wTgAAE3HLAmSfSpSz5MRvsfSzBlf6/gGfR1SWCmVuZHN0cmVhbQplbmRvYmoKMTUgMCBvYmoKPDwgL0Jhc2VGb250IC9EZWphVnVTYW5zIC9DaGFyUHJvY3MgMTYgMCBSCi9FbmNvZGluZyA8PAovRGlmZmVyZW5jZXMgWyAzMiAvc3BhY2UgNDYgL3BlcmlvZCA0OCAvemVybyAvb25lIC90d28gL3RocmVlIC9mb3VyIC9maXZlIC9zaXggNTYKL2VpZ2h0IDY3IC9DIDc2IC9MIDg3IC9XIDk3IC9hIC9iIDEwMCAvZCAvZSAxMDMgL2cgL2ggL2kgMTA4IC9sIDExMCAvbiAvbwoxMTQgL3IgL3MgL3QgL3UgL3YgMTIxIC95IF0KL1R5cGUgL0VuY29kaW5nID4+Ci9GaXJzdENoYXIgMCAvRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250RGVzY3JpcHRvciAxNCAwIFIKL0ZvbnRNYXRyaXggWyAwLjAwMSAwIDAgMC4wMDEgMCAwIF0gL0xhc3RDaGFyIDI1NSAvTmFtZSAvRGVqYVZ1U2FucwovU3VidHlwZSAvVHlwZTMgL1R5cGUgL0ZvbnQgL1dpZHRocyAxMyAwIFIgPj4KZW5kb2JqCjE0IDAgb2JqCjw8IC9Bc2NlbnQgOTI5IC9DYXBIZWlnaHQgMCAvRGVzY2VudCAtMjM2IC9GbGFncyAzMgovRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250TmFtZSAvRGVqYVZ1U2FucyAvSXRhbGljQW5nbGUgMAovTWF4V2lkdGggMTM0MiAvU3RlbVYgMCAvVHlwZSAvRm9udERlc2NyaXB0b3IgL1hIZWlnaHQgMCA+PgplbmRvYmoKMTMgMCBvYmoKWyA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMAo2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDMxOCA0MDEgNDYwIDgzOCA2MzYKOTUwIDc4MCAyNzUgMzkwIDM5MCA1MDAgODM4IDMxOCAzNjEgMzE4IDMzNyA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2CjYzNiA2MzYgMzM3IDMzNyA4MzggODM4IDgzOCA1MzEgMTAwMCA2ODQgNjg2IDY5OCA3NzAgNjMyIDU3NSA3NzUgNzUyIDI5NQoyOTUgNjU2IDU1NyA4NjMgNzQ4IDc4NyA2MDMgNzg3IDY5NSA2MzUgNjExIDczMiA2ODQgOTg5IDY4NSA2MTEgNjg1IDM5MCAzMzcKMzkwIDgzOCA1MDAgNTAwIDYxMyA2MzUgNTUwIDYzNSA2MTUgMzUyIDYzNSA2MzQgMjc4IDI3OCA1NzkgMjc4IDk3NCA2MzQgNjEyCjYzNSA2MzUgNDExIDUyMSAzOTIgNjM0IDU5MiA4MTggNTkyIDU5MiA1MjUgNjM2IDMzNyA2MzYgODM4IDYwMCA2MzYgNjAwIDMxOAozNTIgNTE4IDEwMDAgNTAwIDUwMCA1MDAgMTM0MiA2MzUgNDAwIDEwNzAgNjAwIDY4NSA2MDAgNjAwIDMxOCAzMTggNTE4IDUxOAo1OTAgNTAwIDEwMDAgNTAwIDEwMDAgNTIxIDQwMCAxMDIzIDYwMCA1MjUgNjExIDMxOCA0MDEgNjM2IDYzNiA2MzYgNjM2IDMzNwo1MDAgNTAwIDEwMDAgNDcxIDYxMiA4MzggMzYxIDEwMDAgNTAwIDUwMCA4MzggNDAxIDQwMSA1MDAgNjM2IDYzNiAzMTggNTAwCjQwMSA0NzEgNjEyIDk2OSA5NjkgOTY5IDUzMSA2ODQgNjg0IDY4NCA2ODQgNjg0IDY4NCA5NzQgNjk4IDYzMiA2MzIgNjMyIDYzMgoyOTUgMjk1IDI5NSAyOTUgNzc1IDc0OCA3ODcgNzg3IDc4NyA3ODcgNzg3IDgzOCA3ODcgNzMyIDczMiA3MzIgNzMyIDYxMSA2MDUKNjMwIDYxMyA2MTMgNjEzIDYxMyA2MTMgNjEzIDk4MiA1NTAgNjE1IDYxNSA2MTUgNjE1IDI3OCAyNzggMjc4IDI3OCA2MTIgNjM0CjYxMiA2MTIgNjEyIDYxMiA2MTIgODM4IDYxMiA2MzQgNjM0IDYzNCA2MzQgNTkyIDYzNSA1OTIgXQplbmRvYmoKMTYgMCBvYmoKPDwgL0MgMTcgMCBSIC9MIDE4IDAgUiAvVyAxOSAwIFIgL2EgMjAgMCBSIC9iIDIxIDAgUiAvZCAyMiAwIFIgL2UgMjMgMCBSCi9laWdodCAyNCAwIFIgL2ZpdmUgMjUgMCBSIC9mb3VyIDI2IDAgUiAvZyAyNyAwIFIgL2ggMjggMCBSIC9pIDI5IDAgUgovbCAzMCAwIFIgL24gMzIgMCBSIC9vIDMzIDAgUiAvb25lIDM0IDAgUiAvcGVyaW9kIDM1IDAgUiAvciAzNiAwIFIKL3MgMzcgMCBSIC9zaXggMzggMCBSIC9zcGFjZSAzOSAwIFIgL3QgNDAgMCBSIC90aHJlZSA0MSAwIFIgL3R3byA0MiAwIFIKL3UgNDMgMCBSIC92IDQ0IDAgUiAveSA0NSAwIFIgL3plcm8gNDYgMCBSID4+CmVuZG9iagozIDAgb2JqCjw8IC9GMSAxNSAwIFIgPj4KZW5kb2JqCjQgMCBvYmoKPDwgL0ExIDw8IC9DQSAwIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0EyIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0EzIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDAuNSA+PiA+PgplbmRvYmoKNSAwIG9iago8PCA+PgplbmRvYmoKNiAwIG9iago8PCA+PgplbmRvYmoKNyAwIG9iago8PCAvRjEtRGVqYVZ1U2Fucy1taW51cyAzMSAwIFIgPj4KZW5kb2JqCjIgMCBvYmoKPDwgL0NvdW50IDEgL0tpZHMgWyAxMSAwIFIgXSAvVHlwZSAvUGFnZXMgPj4KZW5kb2JqCjQ3IDAgb2JqCjw8IC9DcmVhdGlvbkRhdGUgKEQ6MjAyMTA5MTYxNDM2NDQrMDInMDAnKQovQ3JlYXRvciAoTWF0cGxvdGxpYiB2My40LjMsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAoTWF0cGxvdGxpYiBwZGYgYmFja2VuZCB2My40LjMpID4+CmVuZG9iagp4cmVmCjAgNDgKMDAwMDAwMDAwMCA2NTUzNSBmIAowMDAwMDAwMDE2IDAwMDAwIG4gCjAwMDAwMjE2NzMgMDAwMDAgbiAKMDAwMDAyMTQxMCAwMDAwMCBuIAowMDAwMDIxNDQyIDAwMDAwIG4gCjAwMDAwMjE1ODIgMDAwMDAgbiAKMDAwMDAyMTYwMyAwMDAwMCBuIAowMDAwMDIxNjI0IDAwMDAwIG4gCjAwMDAwMDAwNjUgMDAwMDAgbiAKMDAwMDAwMDM5OSAwMDAwMCBuIAowMDAwMDExNTk2IDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAxMTU3NCAwMDAwMCBuIAowMDAwMDIwMDEzIDAwMDAwIG4gCjAwMDAwMTk4MTMgMDAwMDAgbiAKMDAwMDAxOTM1MyAwMDAwMCBuIAowMDAwMDIxMDY2IDAwMDAwIG4gCjAwMDAwMTE2MTYgMDAwMDAgbiAKMDAwMDAxMTkyNCAwMDAwMCBuIAowMDAwMDEyMDU3IDAwMDAwIG4gCjAwMDAwMTIyMjEgMDAwMDAgbiAKMDAwMDAxMjYwMSAwMDAwMCBuIAowMDAwMDEyOTE4IDAwMDAwIG4gCjAwMDAwMTMyMjIgMDAwMDAgbiAKMDAwMDAxMzU0NCAwMDAwMCBuIAowMDAwMDE0MDEyIDAwMDAwIG4gCjAwMDAwMTQzMzQgMDAwMDAgbiAKMDAwMDAxNDUwMCAwMDAwMCBuIAowMDAwMDE0OTE0IDAwMDAwIG4gCjAwMDAwMTUxNTEgMDAwMDAgbiAKMDAwMDAxNTI5NSAwMDAwMCBuIAowMDAwMDE1NDE0IDAwMDAwIG4gCjAwMDAwMTU1ODYgMDAwMDAgbiAKMDAwMDAxNTgyMiAwMDAwMCBuIAowMDAwMDE2MTEzIDAwMDAwIG4gCjAwMDAwMTYyNjggMDAwMDAgbiAKMDAwMDAxNjM5MSAwMDAwMCBuIAowMDAwMDE2NjI0IDAwMDAwIG4gCjAwMDAwMTcwMzEgMDAwMDAgbiAKMDAwMDAxNzQyNCAwMDAwMCBuIAowMDAwMDE3NTE0IDAwMDAwIG4gCjAwMDAwMTc3MjAgMDAwMDAgbiAKMDAwMDAxODEzMyAwMDAwMCBuIAowMDAwMDE4NDU3IDAwMDAwIG4gCjAwMDAwMTg3MDQgMDAwMDAgbiAKMDAwMDAxODg1MSAwMDAwMCBuIAowMDAwMDE5MDY1IDAwMDAwIG4gCjAwMDAwMjE3MzMgMDAwMDAgbiAKdHJhaWxlcgo8PCAvSW5mbyA0NyAwIFIgL1Jvb3QgMSAwIFIgL1NpemUgNDggPj4Kc3RhcnR4cmVmCjIxODkwCiUlRU9GCg==\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 1.0088235139846802\n", "Layer 2 - Variance: 1.0696827173233032\n", "Layer 4 - Variance: 1.125657081604004\n", "Layer 6 - Variance: 1.1308791637420654\n", "Layer 8 - Variance: 1.0503977537155151\n"]}], "source": ["def equal_var_init(model):\n", " for name, param in model.named_parameters():\n", " if name.endswith(\".bias\"):\n", " param.data.fill_(0)\n", " else:\n", " param.data.normal_(std=1.0 / math.sqrt(param.shape[1]))\n", "\n", "\n", "equal_var_init(model)\n", "visualize_weight_distribution(model)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "b1a998c1", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.065378, "end_time": "2021-09-16T12:36:52.200763", "exception": false, "start_time": "2021-09-16T12:36:52.135385", "status": "completed"}, "tags": []}, "source": ["As we expected, the variance stays indeed constant across layers.\n", "Note that our initialization does not restrict us to a normal distribution, but allows any other distribution with a mean of 0 and variance of $1/d_x$.\n", "You often see that a uniform distribution is used for initialization.\n", "A small benefit of using a uniform instead of a normal distribution is that we can exclude the chance of initializing very large or small weights.\n", "\n", "Besides the variance of the activations, another variance we would like to stabilize is the one of the gradients.\n", "This ensures a stable optimization for deep networks.\n", "It turns out that we can do the same calculation as above starting from $\\Delta x=W\\Delta y$, and come to the conclusion that we should initialize our layers with $1/d_y$ where $d_y$ is the number of output neurons.\n", "You can do the calculation as a practice, or check a thorough explanation in [this blog post](https://pouannes.github.io/blog/initialization/#mjx-eqn-eqfwd_K).\n", "As a compromise between both constraints, [Glorot and Bengio (2010)](http://proceedings.mlr.press/v9/glorot10a/glorot10a.pdf?hc_location=ufi) proposed to use the harmonic mean of both values.\n", "This leads us to the well-known Xavier initialization:\n", "\n", "$$W\\sim \\mathcal{N}\\left(0,\\frac{2}{d_x+d_y}\\right)$$\n", "\n", "If we use a uniform distribution, we would initialize the weights with:\n", "\n", "$$W\\sim U\\left[-\\frac{\\sqrt{6}}{\\sqrt{d_x+d_y}}, \\frac{\\sqrt{6}}{\\sqrt{d_x+d_y}}\\right]$$\n", "\n", "Let's shortly implement it and validate its effectiveness:"]}, {"cell_type": "code", "execution_count": 17, "id": "463d4b1e", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:36:52.335958Z", "iopub.status.busy": "2021-09-16T12:36:52.335480Z", "iopub.status.idle": "2021-09-16T12:37:05.009307Z", "shell.execute_reply": "2021-09-16T12:37:05.008916Z"}, "papermill": {"duration": 12.743484, "end_time": "2021-09-16T12:37:05.009425", "exception": false, "start_time": "2021-09-16T12:36:52.265941", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["layers.0.weight - Variance: 0.0003991015546489507\n", "layers.2.weight - Variance: 0.0007022571517154574\n", "layers.4.weight - Variance: 0.0009397325338795781\n", "layers.6.weight - Variance: 0.0014803955564275384\n", "layers.8.weight - Variance: 0.012549502775073051\n"]}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 1.2209526300430298\n", "Layer 2 - Variance: 1.5839706659317017\n", "Layer 4 - Variance: 1.5429933071136475\n", "Layer 6 - Variance: 2.021383047103882\n", "Layer 8 - Variance: 2.6867828369140625\n"]}], "source": ["def xavier_init(model):\n", " for name, param in model.named_parameters():\n", " if name.endswith(\".bias\"):\n", " param.data.fill_(0)\n", " else:\n", " bound = math.sqrt(6) / math.sqrt(param.shape[0] + param.shape[1])\n", " param.data.uniform_(-bound, bound)\n", "\n", "\n", "xavier_init(model)\n", "visualize_gradients(model, print_variance=True)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "25a50e49", "metadata": {"papermill": {"duration": 0.077482, "end_time": "2021-09-16T12:37:05.166555", "exception": false, "start_time": "2021-09-16T12:37:05.089073", "status": "completed"}, "tags": []}, "source": ["We see that the Xavier initialization balances the variance of gradients and activations.\n", "Note that the significantly higher variance for the output layer is due to the large difference of input and output dimension ($128$ vs $10$).\n", "However, we currently assumed the activation function to be linear.\n", "So what happens if we add a non-linearity?\n", "In a tanh-based network, a common assumption is that for small values during the initial steps in training, the $\\tanh$ works as a linear function such that we don't have to adjust our calculation.\n", "We can check if that is the case for us as well:"]}, {"cell_type": "code", "execution_count": 18, "id": "9121ed95", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:05.322709Z", "iopub.status.busy": "2021-09-16T12:37:05.322186Z", "iopub.status.idle": "2021-09-16T12:37:18.007787Z", "shell.execute_reply": "2021-09-16T12:37:18.007359Z"}, "papermill": {"duration": 12.765022, "end_time": "2021-09-16T12:37:18.007907", "exception": false, "start_time": "2021-09-16T12:37:05.242885", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["layers.0.weight - Variance: 2.1826384909218177e-05\n", "layers.2.weight - Variance: 3.5952674807049334e-05\n", "layers.4.weight - Variance: 4.872870340477675e-05\n", "layers.6.weight - Variance: 6.269156438065693e-05\n", "layers.8.weight - Variance: 0.0004620618128683418\n"]}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 1.2046984434127808\n", "Layer 2 - Variance: 0.5917537212371826\n", "Layer 4 - Variance: 0.2959783673286438\n", "Layer 6 - Variance: 0.24997730553150177\n", "Layer 8 - Variance: 0.2727622389793396\n"]}], "source": ["model = BaseNetwork(act_fn=nn.Tanh()).to(device)\n", "xavier_init(model)\n", "visualize_gradients(model, print_variance=True)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "6f1db629", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.089367, "end_time": "2021-09-16T12:37:18.189510", "exception": false, "start_time": "2021-09-16T12:37:18.100143", "status": "completed"}, "tags": []}, "source": ["Although the variance decreases over depth, it is apparent that the activation distribution becomes more focused on the low values.\n", "Therefore, our variance will stabilize around 0.25 if we would go even deeper.\n", "Hence, we can conclude that the Xavier initialization works well for Tanh networks.\n", "But what about ReLU networks?\n", "Here, we cannot take the previous assumption of the non-linearity becoming linear for small values.\n", "The ReLU activation function sets (in expectation) half of the inputs to 0 so that also the expectation of the input is not zero.\n", "However, as long as the expectation of $W$ is zero and $b=0$, the expectation of the output is zero.\n", "The part where the calculation of the ReLU initialization differs from the identity is when determining $\\text{Var}(w_{ij}x_{j})$:\n", "\n", "$$\\text{Var}(w_{ij}x_{j})=\\underbrace{\\mathbb{E}[w_{ij}^2]}_{=\\text{Var}(w_{ij})}\\mathbb{E}[x_{j}^2]-\\underbrace{\\mathbb{E}[w_{ij}]^2}_{=0}\\mathbb{E}[x_{j}]^2=\\text{Var}(w_{ij})\\mathbb{E}[x_{j}^2]$$\n", "\n", "If we assume now that $x$ is the output of a ReLU activation (from a previous layer, $x=max(0,\\tilde{y})$), we can calculate the expectation as follows:\n", "\n", "\n", "$$\n", "\\begin{split}\n", " \\mathbb{E}[x^2] & =\\mathbb{E}[\\max(0,\\tilde{y})^2]\\\\\n", " & =\\frac{1}{2}\\mathbb{E}[{\\tilde{y}}^2]\\hspace{2cm}\\tilde{y}\\text{ is zero-centered and symmetric}\\\\\n", " & =\\frac{1}{2}\\text{Var}(\\tilde{y})\n", "\\end{split}$$\n", "\n", "Thus, we see that we have an additional factor of 1/2 in the equation, so that our desired weight variance becomes $2/d_x$.\n", "This gives us the Kaiming initialization (see [He, K. et al.\n", "(2015)](https://arxiv.org/pdf/1502.01852.pdf)).\n", "Note that the Kaiming initialization does not use the harmonic mean between input and output size.\n", "In their paper (Section 2.2, Backward Propagation, last paragraph), they argue that using $d_x$ or $d_y$ both lead to stable gradients throughout the network, and only depend on the overall input and output size of the network.\n", "Hence, we can use here only the input $d_x$:"]}, {"cell_type": "code", "execution_count": 19, "id": "264d9108", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:18.387896Z", "iopub.status.busy": "2021-09-16T12:37:18.387392Z", "iopub.status.idle": "2021-09-16T12:37:30.823238Z", "shell.execute_reply": "2021-09-16T12:37:30.822814Z"}, "papermill": {"duration": 12.536784, "end_time": "2021-09-16T12:37:30.823357", "exception": false, "start_time": "2021-09-16T12:37:18.286573", "status": "completed"}, "tags": []}, "outputs": [{"data": {"application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9Bbm5vdHMgMTAgMCBSIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDkxMS41MjUgMjE2LjY2NTYyNSBdIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovVHlwZSAvUGFnZSA+PgplbmRvYmoKOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEyIDAgUiA+PgpzdHJlYW0KeJzVnUuTJLl1pff5K2I5s6ATwMVzKQ6lNtOOUtvMYmwWPWSLalpV09gP0fTv5xzAwx2A30hVZmVlDYxGqfIwAuHnczgcj4sLe/vL02//wd7+/PMN/+dmbn/Bf/+Of3/Dv58M/vr4VKzdggv494fj387GLcYQ8c8P+Njw578/Pf3bk9mKTdEnE3K+zX/4YmyJJuXbT/zRby4fOP54mj799BTTJvgZ77bSfvDjky1+s15sdp38oZdtylu+63sJg1av+W+3S+HWxS3b/T8oR8JWbj99f/tftx9vv/0H18D9M/77F/y3gnv67e+//48f/vj9v3zzu9sff34qskkMzuXhik91uIqnf336w+1v94LNZgNuyr3s+uc3u/r0tycLcr8x+J+SbAasgkvFupsLmzUs7o8fn3737e23/2Rv1t6+/bensuEupRJz4q389k9P//v238xmXPjvt/9z+/afn/7xWwCAYFmw6f71x48s5Te///4v3/3PX//1ux9//s3HH3789efb7/96+8PTH+oVvz03K26TkHD/xlt9ym9AzrLaGpZmQnpMzvS8jDl5fTHzyW6Cah/GWtPJb2E+hC0UlibOlU9039eWP/SlZbslX5INxRR7K1vqy3FjOd/89N2fbh+/+/OPP/zy65++//I4P73JeB3JYLaMr0veUg5F0rMM381u9ltS7J7ya+06V+1ms4VqN5VSnnHszLs8M/dWHeUEFhlH253+mb4t/r/QeDal5GeM+/c1HjPfKVfjp/65xiMKrcbleeNxNv63JxbzGxaIFpdXixegKUHs2N5M7cT/+OuvP/7ymQClGXUlGynOhoQehW99Cyd4EW7x3rvwtTOyuegkO+/DTnv48q3/8tP05acnvolz8XF8Z6WwSQnO56GpGdVkc6nqUMJdRe/kd///m1dswnzakrFGZDQ/qMXnfEVyV9cwr9j8+JTZX4i4uMF8p+Juh+DLhKRTlzCv2YR5NJO4SDPe+U6FTe9KuCK5q2uYV2zCvN+izZKnO3+qsJmilCuSu7qGecUmzMetOG+TH82fKht/sfaK5K6uYV6xCfMYZ7lSQhnNnyptWqMh2dU1zCs2OVWwBQzP/HjnO1U29CJsuiA51CXMazbroD57G9x45zsVNrNpb/QRyV1dw7xiE+YjOpxJ7HTnT1VwtzP+hxnJoa5hXrEJ83nzQTAcGM2fqmyCsZK/IDnUNcwrNjHWMGZL+Eoeb30vyxZCMOYC5VCXsK8apX+03dHHWCb/p4z6jmLLFcshLwJAcUoA6Lcn40OYAJyy38RgtDdz6eRFAChOCSBtMUU3PwCH6rdcck5XKoe8iP2rT7ovW8nOuPn2n3LE18TliUqnLmJfMQr/1m0up2xH+6daNoxpjZ+hdPIa9hWfdO+3UISt2Gj/kIPd8FXvRyq9uoh9xSj94wpLkTzZv6ucZLdS8gylkxexf/VJ9+jEm2DjfPcPORp0cozzI5VeXcS+YhT+HfrxeKgn+4ca84ZBnQkzlE5ew77ik+7Rg7FoxONk/5BT2cQWk0cqvbqIfcUo/aMv71xwU9N/ygU9nJz2p/8spJcXAaA4JQD05l0WM1eAQ2aD73OSPIEZ9EUQKF65II3+vHgueg4ITtl6Tm7mYOKEptfXQKB5JQL06PEuS1MtOGWbYNXb3BB0aHp9EQSKVyJAr96HGOdacMjWZ+jo98YJTa8vgkDxSgTo2Afr/VwLDpkvAAnBhTih6fVFECheiQC9+5CczLXgkK2xG0b8Ls5oen0RBIpXIMAfLoqxaURwylk2m2QfDp2F9PIaADSnBIA+fszZTLFop8w2zxUjE5deXgSA4pQAcIXJhzzXgEOOskWxbd6zK6SXFwGgOCWAwhl9SXMNOOSAJt+5lCYuvbwIAMUpAGBs73Pk9wcApxyEsWZiZi6dvAYAzSkByJaKLX6qAZ2MoSBGwnHi0suLAFCcEkDcTEnJzTXglGFKbJSZSycvAkBxSgDo5qMtt3MNOGV8z3tbZi6dvAgAxSkARHTyTUExI4BO9ht/I89cOnkNAJpTAkAX3waTZQJwyn5zAbf9yuWQFwGgOCUAdPCdyXGuAafs0efFT165HPIiABSnBIDuvYsxzDXglGXDfTZl5tLJiwBQnBIAevfi/Dw33sls+GUfDQ1cDnkRAIpTxuujd49C3VQDOplr4OleA3ouh7wGAM0pAaB3770xUw3oZEFrZ1u438jlkBcBoDglAFyhL6jNE4BTRp8XvxmvXA55EQCKUwJA7z6gMk/BEZ3M3UtlfwsMXA55EQCKUwDI6N1HK3EKDhjle6zzVMhSIdCqUwJAax4ThnkTgEG+R7pPhSwVAK86JQD06ZMrMteAUcavKFzu8iIAFKcEkDdJOdm5BgzyWQOGQj63BowbaLZw+/vTQ0cjjn/55jZutJm3d2DMnvdQ7xSlvbVMNsa5FucWcOsY5iObt3WPFOM/3BYZABBrLy9j3J+qHPHLuPLUxj9eZA+VcclAQ9fQRt8Gy4YLSEVy5FhRjPW41BZYgvdGYkxlW2S3VbYYV/iYuJGwbM5HLjm0MAx8L3HSyeCa0eDUy7A1NMsK1SjAXHYVP+6TMGgDTyb66nswQzZgWDhv40N2pf0i8EeTfbhxVJvEpVZI2qzljh9Oc+FFh0uocobF7Gy8Rc6A4++4L5V7wb1wt4hXYuCuo7p4xk2pkWEDEcB4RW1ZGeXB8A39Bzw3ObfPuk1EImdWU4s6yLtcrCuxbl+LsSS/L8364sQI56FLMLG0xTrcWnTEUWBxW3A1dIFygPMCqrfCGGYb6h1nPZSM+pNvjHHzBf/0u47KIlzpwM0ruNTSLrxwC6epi0Do76DnW2+C2C0asYYbriJuEwjWXxW445PJzZ/wVrxtFwk9wWZdTrOcQURd3NeYBJjxHHFwhSsIoV6OeF4bn5QWgem97OsxMdTVSF6B9SY6uy9SYDSSE13BoM3e5X3q3nIBx99Y1b05ljoS/g8qbQncxYI7Xuf4zIa6GeA7B9bTWK/Eo61HL0/qzTDcUGf3OXH4Z71Bd8DZFOrWNcqRO1QdNzyLiQ0uxwqxCLpM6D4b1E2T92llwcOFVgztBYqOra7gSQJwAWjUK5fx+faLeNB8kBhvAX1wPOCmfZqzEt6EOmMLq86kXS5ZuEsj4HvR2n1uB4+aw1X6W8CdgtfGic+MwU2ujxqasz0YAHLO3GHNB9CbXK8DX/SheA6BCmpvLq1iBb+lFLkFDxfqUyrHbFrAxYQWSsetQ5VTYICdUMUddbgX9xkW7vMFSsgwgppUxxyoSSa7XIMULR64VjVBxwCDbaPxaLyvF4L6KFFQk/bWi+0tOy2Jzwxn7tkCWuPbo5b5jJp7sChahKi0+ihcexns8rhf0bEZx8BgfzP9l5t8tY3+KFHd+/vxYboAfONFG4ivv/ps2QamPnUjZsajnsSGffUcb569pMd7KD9895/f//TzZra/f//Dn//9l1u/nXLP/PCSVA0tP8ScsuGS7WFM2eAY6XoNmxU0VzHnOIUPdfLAqivlmrqBteT4AP/z2h7Ck1a3HBvQdFno6OQ32I/vPF+RpuCVlMPL8zi8RRqHL0cQLSpeF9P937W3YFfwgsHIEe+1UJ5hN2ZymLcWf0H7XOctJU7LBJ38BhBE+CphaQ49n0+E8CifA17WuLRgG47PSujwBZ/LFzUsr938nlvAfrv61yV5eBcE6C2hZ1jQOx8QnPKrERjTI2AvE72bFJJHh+2FiR/ehQRHQ+hNYnQ0ouj0N2LB7RyQImO50Wl8YTKI94GBYtF/R5d7ZHHIb4VC2F1OBaMhZ/xL0kO8H4oUtoTOScoTi1N/KxgsnFsc0Y02zzUV+blcGUV6Jq/PmfFKuJw8+tTZkv968ujZmSd0Euucgy9TPJXDmKigWk3xVLPMaZkqj4Xs8jF5tAYIxTFBBE7g+CmlwCyntqFqLiRN+6zWAKE4Joi0BTQySSYQo3yAUPksBkJxTBBly5xknGvEICeM+eOVz11eDITimOMwt9ko2U81YpRTYmaZK59dXguE5pgg/OYjRgBTjRhlTq5Ehc8uLwZCcUwQcUspcCp7BDHI2THx3YXPXV4MhOKYIMpmsrVlrhGDfDwaYyGLPhqKY4Bg7sQcyxSPNclFcrBXPnd5LRCaY4IQfNKlKV1BL3PpHT8gE59OXgyE4pggcEmlroqMIE4ZjjGWqWtjE5+7vBgIxTFB5E2MF5lrxClzu5KkK50mLgZBcQsI2XANxU5JHHqZITk21nXckc0hrwVCc0wQXCFlGtcJxCnDMVNppiufu7wYCMUxQQQu7sYy14hTFs7eSl2bHvkc8mIgFMcEgZGDSz7NNeKUheVasRc+h7wYCMUxQWDkIOLiXCNOmZGrrq1/jnwOeTEQimOAKBg51JXlgcOpoj9tTFsd7ovo5LUwKH5JAX68zzJjOOWw+Xsoe1dGpy6GQTFMDhg1BBOnGbpTxXOAClDDg/oiOnkxDFe/pIAhQ4j+8lAcMkM+RKIf6fTqYhgUw1zVZZCYm4L9TpWhJVIDaboCTnEpBJrXtq4dY+IWnhHBIXOpxFou1A5oBnkxEIplksAlJZkSQ5xqjugyGltDtroienkxDFe/pMAgx1LjowYMh8yoM2Z/TQOdXl0Mg2IYHGw9MUCm9AidjLYwGZPtjKeT1wKhOSYIt+Vi8ahPIE45Mbyu1BDHvpBOXgyE4pggwoaP4bMTiFMOdSuYTHhOdTEMil9iQFfIuJjn+nDKfmMgfHtfdIV08mIgFMcEgbGCyRgzTiBO2dc3ZB1bj3wOeTEQimOAcBgsWM+UOAOITmZQPn4kTnw6eS0QmmOCwHjBGeOnGtHJnGQwvlzw3NXFMCh+iQHDBReym+vDKXMHgce74ULnkBcDoTgmiMLNAXFKrNDLslmf2+bhkc8hLwZCcQwQglGDJCljio1e7iamJz5Lzlerjgmibk+aA/F7mbuaTAt3GPkc8mIgFMcEgUvyucS5RpwyV2wYm3fBs6uLYVD8EgPGDcGnMNeHU+ZCZuLGqAudu7wYCMUxg8u5w81wV9kAYpRLCHUn3lTIXV4LhOa4RdnnGNE7mEAMMn6gvTwnPru8GAjFMUEEhoJbM9eIQU65tAmZsZC7vBgIxTFB1H10eUrLMMk8lahc+dzlxUAojgmibClLTHkCMcj3WMGpkCVDCFXHABHc9ei+D7N8D5Gb+awYOac6JgivnVY4ySkwYeGVzy4vBkJxTBBRO7lwlvcI87mQFQPPVccEUbRTDCf5eDTGQhZ9NBTHABGtdqLhLB81YixkyRqhOSYI0U43fCir2BYDoVgjiKCddDjL+56DuZA32orQb4F6qhlObp+I5ZrhZNo/k6yPpUYIel7vPceJzW0NO8NPSL6lrHMCw3kPAEJT6NsQW/BGSC0OJG4heNsiSdGNsDUlA3fuRlNCy0DgQootzrDg512siVI43W1yCy4qnulKIlOFcuDGMIpdjtEztYKPzLYSjd0X2T13+6HszORyqV0g5EIaqR3Cg75dkxM6/rbmacg8mK2JebPeGi/M6uBJ0u5ycoZZWNDwR+funy64g8XAMCq9y9ZKW+E1TJDibbihZgBIC4Hhwq+RxOQszHxmU8tyQhmXYU1Nip0Bp4ZIiHGc9scl1G2MwfhUE3dQRxnM5cIskrg50cZ9PVUwipMxXQjlIp65U5j1BNUkpfarnifSZHy38DxS4Y1qq5E8lRNXxg3K6Psdi5Sov6Gls2bSvjaINJF521g9YBR33Pm7jPtnUfGJJcFQHWkZdCet2XPC4NXp3L7uFyKqbiJag4qQm310ufBVIy2PhrQKwcUxlMP0+eyT4WlpUCzTSDEHB++8LXg9t08L+23wWhOtOh9K2ldUUA24adozPUlwsQ6AbNxiNsbWk8rQfkkrhOnqrU+xVmQ0KvjVKqeNe/Jln2v0plY2kbgJs3fYNtFicav24XZk0ow97ry0zOdzy4GXxjPyC7JjPNgp/SinAkpWN1F/fJidAd948W5s/def/Y2XZMtwPGUIj4Npq8gvSZfhvma6DO/Q0lzjUQMzOF1PGunkAVZXip4u4/yAGX/qs3fceiaNMRKnw5I6+Q2yHXg8b3jKmDLfoBn+GukyviDBjBe5Y0KiAeChvgW/CDIJhRk0a69OmfHlEATLfE5o+cZK1MlvACHgzZQKS5OEfsfnpczAVW3Z1+9/TrqML1irXtSwvLJaMXzpvPrXpct4FwR8yUSPt/FA4FBfnYWF2bDOa4+JmfbY+c54nT1mYLWkCO/CgSfSYmjCeOSBRKe/EQsefYzhj0voibuXJg55HxZcJ62HMY0oDvmNSDCzIOoERkY24M/HKOTroQiJeQUZnTuyOPW3ghEEgyKMeBJeRc+1E5ccKl2uDPYZOyavz5XxSrjvOh3hXdYOXvFitINXZjn70jqIYyG7fExHLAFCc0wQTjuA5aGsYlsMhGKNIIJ2EMskJytR4XOXFwOhOCaIpB3IMsknCJXPYiAUxwRRtINZJjmVlhh35rPLi4FQHHM06rQDWiY5JWs0Pru8FgjNcRuWKwe1THLyTjQ+u7wYCMUxQUTtwJZJPh6Nic+Sj4bmmCCydnBLL9dcEEX8xKeTFwOhOAaIYLUDXHqZuSByqskVJz53eS0QmmOCEO0gl14WprFu8/0Tn7u8GAjFMUEE7UCXXuYhhsW6C567uhgGxS8xZO1Yl17mAMyWC5tdXAyC4hYQotGOdull4YRHqiuaI5tDXguE5pggnHbESy/LJtaY9lQMfA55MRCKY4II2lEvvYwWAQ2BufI55MVAKI4JImlHvvSybIkbdNKFzyEvBkJxTBBl49EHZq4Rp8yjFljwxKeTFwOhOAaImjUanaKpRnSy31KIGGhd+BzyWiA0xwThUQY+NtWITubu1tCCDfpCOnkxEIpjgsC4geEqE4dDZeSNK3am08mLYbj6JQUMGkyKU3KATmYnMiZjRzq9uhgGxXBb23ZWphSup8os9ixygtPLa2FQ/JICOwL8/oThkHnNSZypMWFnIYO8GAjFMklg1OBmDE0Lhb2llmmu+3ovL4Zg9kr/aOcEH5kRHDIG2SbbmFNPptMWQ6DYBYVieBDcvKB1qomhksn5EU0vr4VB8UsKGChwonWqC50ceL5arnGsXRmduhgGxTA5YJyANm/qOZ2q30qNIJ7gdPJiGK5+SQGDhIACpo5TJ6OzxOP+0kznUBfDoBgmB4wRQknWThxOGYPJlNoBqn0hnbwYCMXxx6dgLP/wZawQvcyNB7hye+FzyEuBUB0TBKp4si5PEbKdLJt3th3jOfI55MVAKI4JIm4hJRPnGnHKsjkMqMKVzyEvBkJxTBCZcVbM3j6COGWeQedLzRQz8jnkxUAojhk2azd7DcLvZTi2tpQrnl1dC4Pmlxh4MK/Hwz5hOGUuWEk7nXikc8iLgVAcEwTPMDZuSgXQy1zLDbnmBZn43OXFQCiOCQKjBhNKcROIU+aON9v2FU187vJiIBTHAOHwNYtR01QjOnmoEQOfNWuE5pggMHKwKUz5jCe5WF+P054KucuLgVAcEwSGDk7EzzVikHl4tL/yucuLgVAcEwSPdC7czDmCGOR7TNhUyJKhYqpjgijXM/o+zPI9XHLms2IUpeoYIMRqpxI+knVsa4HQrBGE104lfCirfBYDoVgjiKidSjjLe5D9XMiKsfeqY4LI2qmEs3yAUPksBkJxzD0zVjuVcJYPECqftUBojglCtFMJZ/kOQufzuSDeMkvGtHMmMTNADZtNzJBgWhiLMdHke5gPGr3SMg/iHzHeg4KypNgyD+LlKG0lJ9YwMdNOlkL/Opd7wIhY/GKdzsaluZril+EBJqDXVad3s92TJfiEvlguKdeEBhFk9k97nkiTGc0ft2JqlgOf8InM5BlcWve4vhz3lebsmcuB+RYi0LTw75Q27zxKYT4MF/Dvsq/IGkMHAT8huH+yqyGLs3ILCW9AMdJ+sWxM1hziDWiMyTsnyMnbXMoNDEpO7fKyQSfCMoI0M5sHfmVf+MTnYXfMbuEzCgzFJXuzdJ5Dsn7XUaci/iiZgTUlyL546GwquP+kEJxt+DLP60GTFPGL4m2LxMh+4+1CvYyWO89DMwO5hEBo3BjnxDY1MhNxxHUEpkZxuWb8oBxRF3AZAXUQT3vNveEzKg6R+oJhNsjIvpzlmZdTeF8C56XatZUtMMNI4SoXSO8LgMUAXmGKEc+EEy7WXCJcBUHpTBpST51h/vQqM/tn5DGPuPqCS2plcD+1YeaNGoaVY50ZDsbgF3NNsVG/5msKEE6fordfU2zwQbYtYQgnE03yNcWG4M6Jd/eptcKcJi36EVWyDRqsw5MUrGkzLCE530bfsIaxVDlCyuvRSRxi4Fkre2rGXBuA/R2CZ+Dc6dJGZmP70p7Sh/ILcmk82Eb9KOECSlZ3WH98mLoB33jxVm3915/9jZfk0uBGIDQSdXP7SxJp+K+ZSCOUAkaXcNXErNDX8zg6eSDVlaIn0jg+wP/4V72RHmzHRTEb3oRpipzq5DfIgRC5Wo7RapLMLeqflgNh3jX75RgkTrDEaKadjZ38BgySRWNmUZoFhfJpDOyDNBB4yfDI1JZ15vNSQXw5qi97NF6ZXCNLC/RuV/+6VBDvgoA9D2c4PTEgOOVXI0AHp0OATkZAG4riC/r5jykEZdP/u4CwFj2eVN/HA4lOfzWKEHoUFm9+V5gPrs6LvDAxxvvACHU1NPsZxqm/FQzPk9CMRf8sP/t42LlmdBkQArefnWW+PgPCK+G+61AzmqAdD4Ehl3Y8xCTjN2oXdirkLh9DzTVAKI4BgnnxrsdDTPIJQuWzFgjNMUE47XiIST5A6HwWA6E4JgivHQ/Ry3WoVUe1fRmnuhgGxS8xJO14iEnG8DrFiU4nLwZCcUwQRTseope5idn7OpXTF9LJi4FQHAOEs9rxEL3MqYmav+vC5y6vBUJzTBBeOx6il3lkjrV1xmnic5cXA6E4JoioHQ/Ry3Sc2wzgzGeXFwOhOCaIrB0P0csC7zkbd+FzyIuBUBwDhFjteIhe5pjAtSNSRj6HvBYIzTFBiHY8RC9zvjtFe8Wzq4thUPwSQ9AOh+hl4SYMU9dSRjqHvBgIxTFBZO1wiF4WFCUSy4XPIS8GQnEMEN5opxz0smw5e1936I18DnktEJpjgnDaKRC97LfgOT868enkxUAojgnCa4dD9HJAdyGVunLZF9LJi4FQHBNE4nxZCWUCccqRS/6u9Su7Qjp5MRCKY4LA2EFC8nONOGVmC46+HibRF9LJi4FQHHMBC2MHb4ObasQpB0bVOisTn15eC4TmmCAwdvBJpi2NnRx4BkVpB670fDp5MRCK47ak6YOYMteIQ7ZoF7zBpecR0KAvhkLxTBQYPeA7ea4Th1wYKOPF55FQLy8GQnEMEBHDh+hjnOrEKTOlFsy3GaqzkF5eC4TmmCAwfpgPVvnQy4GxOMa00WfHp5MXA6E4JggeOxTd9GScKgMBbAsn64vo5MUwXP2SQuL+CuPm6nDKnKy1zuWRTqcuhkExDA4JZeQ07fbtVKY/qCeU9gUc4loIFK8k4LZQJJSpJnQyz2Kzvk1U9mQOdTEMimFy8FsuZcpX26kej4BtR2z1RXTyYhiufkkBD7gJNs614ZQ9Z+Gk5lIb6BzqYhgUw+RQEwFdHopD9Vy3CfEK55AXw3D1CwoZgwUb07Tju5dlS86legxjV0anroVBM0wOGCo4F6Yt373MDFKxHc864jnkxUAojgkCIwWXxcwV4pSZBiMnufI55MVAKI4JAiOFaxB1LzMvbfbJXvgc8mIgFMcAUXgAKXdajCA6uabBiDUd68jnkNcCoTkmCNxZHzBymkCccreqNeBZc1VL80sM3ARjvZ/rwyn39WGgs2p9UBwTBMYLgZt5JhCnzLwoLtZNLBOfu7wYCMXxx6dkWKyYaedyL8tmsjh/wXNXl8Kg+iUGjBlijWcYMZxyFxQ10VkyVkp1TBA8sdeHPNeHUx5AqHwWA6E4Joh0PYHtwyAzbrKkekr1xOcuLwZCcUwQRTs8rZfPANIZz4JhpapfYOAGouvRco9kHdpaIDRrBOG1E+cmGZ2nejr6VMhdXgyE4pggonbi3CTfI8xnPisGnquOCSJrJ85N8j20eipkyYhr1TG3EFntxLlJPkGofD4XxFvmQZj2z6RgXKxRo4k7vEPr/1ip2wgY88Szs1Gua+Mm5+umbsaEOXYa685waDa2CKmyCQ/IrWkQIk96r6EAYbPO2jqxHfCDbSTi4TgUm2okCX7X53uAiZccXE13kIqRXc7gnXKs6yRw59sMmC+bTTEw6WHeBBfadgAE7ggJTuru/JyltOmRwK3HjMjgP0xIXvYFel9AUpgdAUZKGzHiD5PExZq2HzfVhruM/0ZTboEjiBT2QkjHMl9EyBxh5ZR2OWNwUQKzJhTmniv7SrgUE/EmTXilpuTtrpZYSrY8dDnWVdG2VmxdZsqJIWsC9WSy4RZxy8anog7c8R0jPlFyTbLf7mJgkojI/BQpkIy0GxMKOngBj8yNh7mIOHOXo/FMWxG5+JBzs4g/LI9Kz0yagKKNv69cojhms6DZDE5NRjfSWF+TJriYW+oAru+hHMvKV7Ykqa1xRKbByDXFRUbn07bMF1wEQ+tak1wkVGq3e4yFi+fMRMF1oWiZz65NjQMSUz0wKM1JsW1ukFuOrUktehE3o8XyYRzsDK59T23gWt7xZGAGVXHPmSg1b8X8rAs+9Vh+QU6CB/t5H+1dR8nqVt+PD3fBM4vBS/cM67/+7G+8JCcBmhimU8HzXld5X5KWIH7NtAQpoc5fI0iL4YGSI8S7NmDqvq8nJDg/YMcf+ezdsfgaz+VJ07lXnfwGm/EzesdFUFrGANK9MiHBF2QQmYAn8P04MDjlt2DADcIepeGFiJ7gJzFwDxISoIMNnjZw7pZHo39GQoIvR/UFD8UrUzwwLOa89NdlI3gX/5GpcjLzR57+T+3V/vHi6/zz7LmCSlo8Txt7dvv9V4FQDJOpmlg6CKf2RhAyr9dKRt9CnsvH4L4SBPTxmFYLD3BHoRPfCIO1ZkOvyhrxHAI85iBfiwMGA8aidy09h1N8Kw6CHoSgg4l+LDr/jzn4r8Uhchq18B3TcTjFt+KAPj7T7mFgxDCZZ/OUPEhGkbzpcbw+GcUrub7veJ9p5K6nMOD9op3C0MscOQRuqv4wFNLJx3h/DRCKY4JI2ikMvTyAUPksBkJxDBDFaKcw9PIAQuWzFgjNMUE47RyGXu5B6HwWA6E4JgivncPQy7IVh0G0v/A55MVAKI4JImnnMPSyZ/LJUKdi+kI6eTEQimOCKNo5DL3MDS9R6uzpyOeQFwOhOMaAmglGr+cw9DIsmBTrHGtfSCcvBUJ1TBBeO4ehlzGYdbmtnfWFdPJiIBTHBBG1cxh6eQCh8lkMhOKYILJ2oEAnc6dg4DT8yKeXFwOhOObEo9EOFOhkLr1gpFTXSno+nbwWCM0xQYh24EInB2YrMqUm1e75dPJiIBTHBBG0cxg6OboNzWOu6zFdIb28GAjFMUEk7RyGTmZ6U977EU+nLoZB8QsMzminMHTysBTalTLoa6HQPBOF085h6OTE0BMbauhJV0gvLwZCcUwQfkspeDvXiUMO6EvjNdFen2chvbwYCMUxQSSeF27LXCMOmWc/+BLbs9Hx6eTFQCiOCaIwZ3fJc404ZXhz2Xg38unlxUAojgFCLMMj0rTnvpe70VVfyLKDLs0xQfD0lRyCn0Cccjfx0Bey6nyE6pggMHownttDRxCn3E08jHzWnI9QHRMERg/WWDfXiFMeQKh8FgOhOAYIz/iW6WySD4PMKL1Yw9b6Mk51LQyaX2JgbkfuEJ0wnLIwhiPXjGZ9IZ28GAjFMUEEHivl01wfTtlvOdejk0Y+nbwYCMUxQWDwIOLiXCNO2aNPHV17eQ58DnkxEIpjgED/0EoxY9xOpzJ8MhqTJjqdvBYGxS8pYODgfZYZwyn3DURPZ9H2QTNMDp5piOZ5/FPta8MAZ9XacPVLChg0hOgvD8Up992Hs4xlew+aYXLAmCE6O70sTnV4Js4i1n0orn4ZLWoZYl7SVBseyCqztTBozshBarx7nDDoqgpnMQxXZ6QQN3cNoX8oa3QWw6A4IwceGBpk2nffy138R1/IqmEhqmOASGbjWZ12egIeyDq2tUBo1ghCuEULn51APJBVPouBUKwRBEYLxqFHNDl+IKt8FgOhWCMIjBdMzcIzOn4gq3wWA6FYA4hseMawC9Ot7+S+sewKWbax1BwTBMYMzhg/3fpHsoptMRCKNYLw10PxPgxy37nuClm2c605JoikHQP4UFaxLQZCsUYQRTsG8KGs8lkMhGINIIrVjgHs5b6x7ApZtrHUHBOEaMcAPpRVbIuBUKwRRNQOAuzlvkZ0haxbIxTHBJG1gwB7eQCh8vlcEG+ZlmLcSWONj+m++SIbl+o8s0hs54owpD54U1duZYtSxLdw68jNZdx7yxQAwaQWZJy2FHxu+QKsly72uLica1qKUqSEuihsaD4XhuHkzbqQfFXD5jCEZw6LsiVrTFsnM2mLRmKusZtigi8tpBMtkbcJIxrOjNpcM2jVHdKOcd9U8cu2bgZkWF/AC72F9XHVLbeYJrsZXDqMBZ6x49v9hRpDMQwNDHX7X/us2xyDfmzNzIDaVhMlUM4Mrva3UDZaMvfwQjSS1qaakyK1tAVUS4o+tl2rQNc+S441wI5JJgrundsjz5L1eLRuOWzGF5/3MCwfXOBmNBM3CxoNnq0H1eZo66bFGHKMYdejtc60g6Ujvlja5wvGQYCZoDPVUHD7hQN9LPhRa3BrckitS+QMaBcHyBY3T8Ql1+IcGCYYeG2wFF1kbrcmZxdcYS6CLaPe5XsckUetk3YSAadx8y6XDFyJwAWPUGrBJBihRIbkhppHgqkrqMK1dzEKq0PAzQp+l5NDpy3eal4SJk7ZA1LQr42lZb3IqNB5D8/AAySpxqngvrckFFnQA8ajFesJO7zF7U4It4lHw+P7eEyXhJrHgiv6MXnLLCqBs7HtPPSM55dHsXBzkt8i06+0QhLulrHtlKdgwaGu8fAjNrbFcJeic21NFPWjMJVJTb6M4Ym5rxN5UmuPJPqlNSMMlwacFVPamWLZ5XZ9zJ3CVCF1FqjYZPZPwzFagdyWlVDdYvUema5FnG8ZyEqMu8xFSRNt/TQuyLcQ8mTwvDuzh1aCa83bhgvCY4RGQfZ0btJ+Ev1ItHXSZqutL7ta0JChFapuDB6HdsuG5pWb+2vilgfyCzKBPNhI/yhpBEq+7rH/qCeeYNaQF23T13/0cekvyf7BhruwXWiL6i/J/pG17B9d0VJPUok8ih3PCR/xvkx/TZjww/c//nL70w8///LTD//3119++OuPwy7Wp/8HwFjsjQplbmRzdHJlYW0KZW5kb2JqCjEyIDAgb2JqCjEwMDYzCmVuZG9iagoxMCAwIG9iagpbIF0KZW5kb2JqCjE3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjM1ID4+CnN0cmVhbQp4nDVRSW4AMQi75xX+QKWwJ++Zquqh/f+1hlEvAwPY2CTvwUYkPsSQ7ihXfMrqNMvwO1nkxc9K4eS9iAqkKsIKaQfPclYzDJ4bmQKXM/FZZj6ZFjsWUE3EcXbkNINBiGlcR8vpMNM86Am5PhhxY6dZrmJI691Svb7X8p8qykfW3Sy3TtnUSt2iZ+xJXHZeT21pXxh1FDcFkQ4fO7wH+SLmLC46kW72mymHlaQhOC2AH4mhVM8OrxEmfmYkeMqeTu+jNLz2QdP1vXtBR24mZCq3UEYqnqw0xoyh+o1oJqnv/4Ge9b2+/gBDTVS5CmVuZHN0cmVhbQplbmRvYmoKMTggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDcgPj4Kc3RyZWFtCnicTVFJbsQwDLv7FfzAAJasxXlPikEP7f+vJR0U7cEQI0tc4u7ERBZetlDXQofjw0ZeCZuB74PWnPgaseI/2kaklT9UWyATMVEkdFE3GvdIN7wK0X6kgleq91jzEXcrzVs6drG/98G05pEqq0I85Ngc2Uha10TR8T203nNDdMoggT43IQdEaY5ehaS/9sN1bTS7tTazJ6qDR6aE8kmzGprTKWbIbKjHbSpWMgo3qoyK+1RGWg/yNs4ygJPjhDJaT3asJqL81CeXkBcTccIuOzsWYhMLG4e0H5U+sfx86834m2mtpZBxQSI0xaXfZ7zH53j/AJVPXCYKZW5kc3RyZWFtCmVuZG9iagoxOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMwNyA+PgpzdHJlYW0KeJw9kktuAzEMQ/c+hS4QwPrZnvOkKLqY3n/bJyXpihzZFkVqlrpMWVMekDSThH/p8HCxnfI7bM9mZuBaopeJ5ZTn0BVi7qJ82cxGXVknxeqEZjq36FE5Fwc2Taqfqyyl3S54Dtcmnlv2ET+80KAe1DUuCTd0V6NlKTRjqvt/0nv8jDLgakxdbFKrex88XkRV6OgHR4kiY5cX5+NBCelKwmhaiJV3RQNB7vK0ynsJ7tveasiyB6mYzjspZrDrdFIubheHIR7I8qjw5aPYa0LP+LArJfRI2IYzcifuaMbm1MjikP7ejQRLj65oIfPgr27WLmC8UzpFYmROcqxpi1VO91AU07nDvQwQ9WxFQylzkdXqX8POC2uWbBZ4SvoFHqPdJksOVtnbqE7vrTzZ0PcfWtd0HwplbmRzdHJlYW0KZW5kb2JqCjIwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ0ID4+CnN0cmVhbQp4nEWRTXIFIQiE956iL/Cq5Fc9z6RSWUzuvw3NvCQrWoXmA9MCE0fwEkPsiZUTHzJ8L+gyfLcyO/A62ZlwT7huXMNlwzNhW+A7Kss7XkN3tlI/naGq7xo53i5SNXRlZJ96oZoLzJCIrhFZdCuXdUDTlO5S4RpsW4IU9UqsJ52gNOgRyvB3lGt8dRNPr7HkVM0hWs2tExqKsGx4QdTJJBG1DYsnlnMhUfmqG6s6LmCTJeL0gNyglWZ8elJJETCDfKzJaMwCNtCTu2cXxppLHkWOVzSYsDtJNfCA9+K2vvc2cY/zF/iFd9//Kw591wI+fwBL/l0GCmVuZHN0cmVhbQplbmRvYmoKMjEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzEgPj4Kc3RyZWFtCnicNU85kgQhDMt5hT4wVRjbQL+np7Y22Pl/upKZTpDwIcnTEx2ZeJkjI7Bmx9taZCBm4FNMxb/2tA8TqvfgHiKUiwthhpFw1qzjbp6OF/92lc9YB+82+IpZXhDYwkzWVxZnLtsFY2mcxDnJboxdE7GNda2nU1hHMKEMhHS2w5Qgc1Sk9MmOMuboOJEnnovv9tssdjl+DusLNo0hFef4KnqCNoOi7HnvAhpyQf9d3fgeRbvoJSAbCRbWUWLunOWEX712dB61KBJzQppBLhMhzekqphCaUKyzo6BSUXCpPqforJ9/5V9cLQplbmRzdHJlYW0KZW5kb2JqCjIyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ5ID4+CnN0cmVhbQp4nD1QO45EIQzrOYUv8CTyI3AeRqstZu/frgOaKVBMfrYzJNARgUcMMZSv4yWtoK6Bv4tC8W7i64PCIKtDUiDOeg+IdOymNpETOh2cMz9hN2OOwEUxBpzpdKY9ByY5+8IKhHMbZexWSCeJqiKO6jOOKZ4qe594FiztyDZbJ5I95CDhUlKJyaWflMo/bcqUCjpm0QQsErngZBNNOMu7SVKMGZQy6h6mdiJ9rDzIozroZE3OrCOZ2dNP25n4HHC3X9pkTpXHdB7M+Jy0zoM5Fbr344k2B02N2ujs9xNpKi9Sux1anX51EpXdGOcYEpdnfxnfZP/5B/6HWiIKZW5kc3RyZWFtCmVuZG9iagoyMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM5NSA+PgpzdHJlYW0KeJw9UktuxUAI2+cUXKDS8JvPeVJV3bz7b2tDUqkqvIkxxjB9ypC55UtdEnGFybderls8pnwuW1qZeYi7i40lPrbcl+4htl10LrE4HUfyCzKdKkSozarRofhCloUHkE7woQvCfTn+4y+AwdewDbjhPTJBsCTmKULGblEZmhJBEWHnkRWopFCfWcLfUe7r9zIFam+MpQtjHPQJtAVCbUjEAupAAETslFStkI5nJBO/Fd1nYhxg59GyAa4ZVESWe+zHiKnOqIy8RMQ+T036KJZMLVbGblMZX/yUjNR8dAUqqTTylPLQVbPQC1iJeRL2OfxI+OfWbCGGOm7W8onlHzPFMhLOYEs5YKGX40fg21l1Ea4dubjOdIEfldZwTLTrfsj1T/5021rNdbxyCKJA5U1B8LsOrkaxxMQyPp2NKXqiLLAamrxGM8FhEBHW98PIAxr9crwQNKdrIrRYIpu1YkSNimxzPb0E1kzvxTnWwxPCbO+d1qGyMzMqIYLauoZq60B2s77zcLafPzPoom0KZW5kc3RyZWFtCmVuZG9iagoyNCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI0OSA+PgpzdHJlYW0KeJxNUUmKAzAMu+cV+kAhXpO8p0OZQ+f/18oOhTkECa+Sk5aYWAsPMYQfLD34kSFzN/0bfqLZu1l6ksnZ/5jnIlNR+FKoLmJCXYgbz6ER8D2haxJZsb3xOSyjmXO+Bx+FuAQzoQFjfUkyuajmlSETTgx1HA5apMK4a2LD4lrRPI3cbvtGZmUmhA2PZELcGICIIOsCshgslDY2EzJZzgPtDckNWmDXqRtRi4IrlNYJdKJWxKrM4LPm1nY3Qy3y4Kh98fpoVpdghdFL9Vh4X4U+mKmZdu6SQnrhTTsizB4KpDI7LSu1e8TqboH6P8tS8P3J9/gdrw/N/FycCmVuZHN0cmVhbQplbmRvYmoKMjUgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA5NCA+PgpzdHJlYW0KeJxFjcERwCAIBP9UQQkKCtpPJpOH9v+NEDJ8YOcO7oQFC7Z5Rh8FlSZeFVgHSmPcUI9AveFyLcncBQ9wJ3/a0FScltN3aZFJVSncpBJ5/w5nJpCoedFjnfcLY/sjPAplbmRzdHJlYW0KZW5kb2JqCjI2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzQxID4+CnN0cmVhbQp4nEVSS25EMQjbv1NwgUjhl5DztKq6mN5/W5tM1c3gCWBseMtTpmTKsLklIyTXlE99IkOspvw0ciQipvhJCQV2lY/Ha0usjeyRqBSf2vHjsfRGptkVWvXu0aXNolHNysg5yBChnhW6snvUDtnwelxIuu+UzSEcy/9QgSxl3XIKJUFb0HfsEd8PHa6CK4JhsGsug+1lMtT/+ocWXO9992LHLoAWrOe+wQ4AqKcTtAXIGdruNiloAFW6i0nCo/J6bnaibKNV6fkcADMOMHLAiCVbHb7R3gCWfV3oRY2K/StAUVlA/MjVdsHeMclIcBbmBo69cDzFmXBLOMYCQIq94hh68CXY5i9Xroia8Al1umQvvMKe2ubnQpMId60ADl5kw62ro6iW7ek8gvZnRXJGjNSLODohklrSOYLi0qAeWuNcN7HibSOxuVff7h/hnC9c9usXS+yExAplbmRzdHJlYW0KZW5kb2JqCjI3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTY0ID4+CnN0cmVhbQp4nEWQx3EFMQxD76oCJTCACvWsx/MP6/6vhvTTQXoYQgxiT8KwXFdxYXTDj7ctMw1/RxnuxvoyY7zVWCAn6AMMkYmr0aT6dsUZqvTk1WKuo6JcLzoiEsyS46tAI3w6sseTtrYz/XReH+wh7xP/KirnbmEBLqruQPlSH/HUj9lR6pqhjyorax5q2leEXRFK2z4upzJO3b0DWuG9las92u8/HnY68gplbmRzdHJlYW0KZW5kb2JqCjI4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNzIgPj4Kc3RyZWFtCnicMzK3UDBQsDQBEoYWJgrmZgYKKYZcQL6piblCLhdIDMTKAbMMgLQlnIKIZ4CYIG0QxSAWRLGZiRlEHZwBkcvgSgMAJdsWyQplbmRzdHJlYW0KZW5kb2JqCjI5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNDcgPj4Kc3RyZWFtCnicMzK3UDBQsDQBEoYWJgrmZgYKKYZclhBWLhdMLAfMAtGWcAoinsGVBgC5Zw0nCmVuZHN0cmVhbQplbmRvYmoKMzAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNTggPj4Kc3RyZWFtCnicRZFLcgQgCET3noIjgPzkPJNKZTG5/zYNzmQ2dpeo/YRKI6YSLOcUeTB9yfLNZLbpdzlWOxsFFEUomMlV6LECqztTxJlriWrrY2XkuNM7BsUbzl05qWRxo4x1VHUqcEzPlfVR3fl2WZR9Rw5lCtiscxxs4MptwxgnRput7g73iSBPJ1NHxe0g2fAHJ419lasrcJ1s9tFLMA4E/UITmOSLQOsMgcbNU/TkEuzj43bngWBveRFI2RDIkSEYHYJ2nVz/4tb5vf9xhjvPtRmuHO/id5jWdsdfYpIVcwGL3Cmo52suWtcZOt6TM8fkpvuGzrlgl7uDTO/5P9bP+v4DHilm+gplbmRzdHJlYW0KZW5kb2JqCjMxIDAgb2JqCjw8IC9CQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM5Ci9TdWJ0eXBlIC9Gb3JtIC9UeXBlIC9YT2JqZWN0ID4+CnN0cmVhbQp4nOMyNDBTMDY1VcjlMjc2ArNywCwjcyMgCySLYEFkM7jSABXzCnwKZW5kc3RyZWFtCmVuZG9iagozMiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2MyA+PgpzdHJlYW0KeJxFkDsSAyEMQ3tOoSP4IwM+z2YyKTb3b2PYbFLA01ggg7sTgtTagonogoe2Jd0F760EZ2P86TZuNRLkBHWAVqTjaJRSfbnFaZV08Wg2cysLrRMdZg56lKMZoBA6Fd7touRypu7O+UNw9V/1v2LdOZuJgcnKHQjN6lPc+TY7orq6yf6kx9ys134r7FVhaVlLywm3nbtmQAncUznaqz0/Hwo69gplbmRzdHJlYW0KZW5kb2JqCjMzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjE4ID4+CnN0cmVhbQp4nD1QuY0EMQzLXYUaWMB67alnFotLpv/0SPn2ItEWRVIqNZmSKS91lCVZU946fJbEDnmG5W5kNiUqRS+TsCX30ArxfYnmFPfd1ZazQzSXaDl+CzMqqhsd00s2mnAqE7qg3MMz+g1tdANWhx6xWyDQpGDXtiByxw8YDMGZE4siDEpNBv+uco+fXosbPsPxQxSRkg7mNf9Y/fJzDa9TjyeRbm++4l6cqQ4DERySmrwjXVixLhIRaTVBTc/AWi2Au7de/hu0I7oMQPaJxHGaUo6hv2twpc8v5SdT2AplbmRzdHJlYW0KZW5kb2JqCjM0IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggODMgPj4Kc3RyZWFtCnicRYy7DcAwCER7pmAEfib2PlGUwt6/DRAlbrgn3T1cHQmZKW4zw0MGngwshl1xgfSWMAtcR1COneyjYdW+6gSN9aZS8+8PlJ7srOKG6wECQhpmCmVuZHN0cmVhbQplbmRvYmoKMzUgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA1MSA+PgpzdHJlYW0KeJwzNrRQMFAwNDAHkkaGQJaRiUKKIRdIAMTM5YIJ5oBZBkAaojgHriaHK4MrDQDhtA2YCmVuZHN0cmVhbQplbmRvYmoKMzYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNjAgPj4Kc3RyZWFtCnicRZA5EgMxCARzvYInSFyC96zLtcH6/6kH1kei6QI0HLoWTcp6FGg+6bFGobrQa+gsSpJEwRaSHVCnY4g7KEhMSGOSSLYegyOaWLNdmJlUKrNS4bRpxcK/2VrVyESNcI38iekGVPxP6lyU8E2Dr5Ix+hhUvDuDjEn4XkXcWjHt/kQwsRn2CW9FJgWEibGp2b7PYIbM9wrXOMfzDUyCN+sKZW5kc3RyZWFtCmVuZG9iagozNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDMzNCA+PgpzdHJlYW0KeJwtUktyxSAM23MKXaAz+AfkPOl0uni9/7aSk0VGDmD0MeWGiUp8WSC3o9bEt43MQIXhr6vMhc9I28g6iMuQi7iSLYV7RCzkMcQ8xILvq/EeHvmszMmzB8Yv2XcPK/bUhGUh48UZ2mEVx2EV5FiwdSGqe3hTpMOpJNjji/8+xXMtBC18RtCAX+Sfr47g+ZIWafeYbdOuerBMO6qksBxsT3NeJl9aZ7k6Hs8Hyfau2BFSuwIUhbkzznPhKNNWRrQWdjZIalxsb479WErQhW5cRoojkJ+pIjygpMnMJgrij5wecioDYeqarnRyG1Vxp57MNZuLtzNJZuu+SLGZwnldOLP+DFNmtXknz3Ki1KkI77FnS9DQOa6evZZZaHSbE7ykhM/GTk9Ovlcz6yE5FQmpYlpXwWkUmWIJ2xJfU1FTmnoZ/vvy7vE7fv4BLHN8cwplbmRzdHJlYW0KZW5kb2JqCjM4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzIwID4+CnN0cmVhbQp4nDVSS24FMQjbzym4QKXwT87zqqqLvvtvaxO9FUwwYOMpL1nSS77UJdulw+RbH/clsULej+2azFLF9xazFM8tr0fPEbctCgRREz1YmS8VItTP9Og6qHBKn4FXCLcUG7yDSQCDavgHHqUzIFDnQMa7YjJSA4Ik2HNpcQiJciaJf6S8nt8nraSh9D1Zmcvfk0ul0B1NTugBxcrFSaBdSfmgmZhKRJKX632xQvSGwJI8PkcxyYDsNoltogUm5x6lJczEFDqwxwK8ZprVVehgwh6HKYxXC7OoHmzyWxOVpB2t4xnZMN7LMFNioeGwBdTmYmWC7uXjNa/CiO1Rk13DcO6WzXcI0Wj+GxbK4GMVkoBHp7ESDWk4wIjAnl44xV7zEzkOwIhjnZosDGNoJqd6jonA0J6zpWHGxx5a9fMPVOl8hwplbmRzdHJlYW0KZW5kb2JqCjM5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTggPj4Kc3RyZWFtCnicMza0UDCAwxRDrjQAHeYDUgplbmRzdHJlYW0KZW5kb2JqCjQwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTMzID4+CnN0cmVhbQp4nEWPSw4EIQhE95yijsDHH+dxMumFc//tgJ1uE2M9hVSBuYKhPS5rA50VHyEZtvG3qZaORVk+VHpSVg/J4Iesxssh3KAs8IJJKoYhUIuYGpEtZW63gNs2DbKylVOljrCLozCP9rRsFR5folsidZI/g8QqL9zjuh3Ipda73qKLvn+kATEJCmVuZHN0cmVhbQplbmRvYmoKNDEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAzNDAgPj4Kc3RyZWFtCnicNVI5bgQxDOv9Cn0ggG7b79kgSJH8vw2p2RQDcXRSlDtaVHbLh4VUtex0+bSV2hI35HdlhcQJyasS7VKGSKi8ViHV75kyr7c1ZwTIUqXC5KTkccmCP8OlpwvH+baxr+XIHY8eWBUjoUTAMsXE6BqWzu6wZlt+lmnAj3iEnCvWLcdYBVIb3TjtiveheS2yBoi9mZaKCh1WiRZ+QfGgR4199hhUWCDR7RxJcIyJUJGAdoHaSAw5eyx2UR/0MygxE+jaG0XcQYElkpg5xbp09N/40LGg/tiMN786KulbWllj0j4b7ZTGLDLpelj0dPPWx4MLNO+i/OfVDBI0ZY2Sxget2jmGoplRVni3Q5MNzTHHIfMOnsMZCUr6PBS/jyUTHZTI3w4NoX9fHqOMnDbeAuaiP20VBw7is8NeuYEVShdrkvcBqUzogen/r/G1vtfXHx3tgMYKZW5kc3RyZWFtCmVuZG9iago0MiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI1MSA+PgpzdHJlYW0KeJwtUUlyA0EIu88r9IRmp99jlyuH5P/XCMoHBg2LQHRa4qCMnyAsV7zlkatow98zMYLfBYd+K9dtWORAVCBJY1A1oXbxevQe2HGYCcyT1rAMZqwP/Iwp3OjF4TEZZ7fXZdQQ7F2vPZlByaxcxCUTF0zVYSNnDj+ZMi60cz03IOdGWJdhkG5WGjMSjjSFSCGFqpukzgRBEoyuRo02chT7pS+PdIZVjagx7HMtbV/PTThr0OxYrPLklB5dcS4nFy+sHPT1NgMXUWms8kBIwP1uD/VzspPfeEvnzhbT43vNyfLCVGDFm9duQDbV4t+8iOP7jK/n5/n8A19gW4gKZW5kc3RyZWFtCmVuZG9iago0MyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE3NCA+PgpzdHJlYW0KeJxNkEkOQyEMQ/ecwheohDPA5zy/qrpo77+tQwd1gfzkIHA8PNBxJC50ZOiMjiubHOPAsyBj4tE4/8m4PsQxQd2iLViXdsfZzBJzwjIxArZGydk8osAPx1wIEmSXH77AICJdj/lW81mT9M+3O92PurRmXz2iwInsCMWwAVeA/brHgUvC+V7T5JcqJWMTh/KB6iJSNjuhELVU7HKqirPdmytwFfT80UPu7QW1IzzfCmVuZHN0cmVhbQplbmRvYmoKNDQgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA4OSA+PgpzdHJlYW0KeJw1jLsNgDAMRHtP4RHiv9kHIQrYv8VJcGPf3ZNeUuJA5ToRjqaBJ0H1mV4g2ekBVkXiUUnM/029qUVTz6btq00EJzOO9XUcqJrTetBaKG2TFt5wfQCcHe0KZW5kc3RyZWFtCmVuZG9iago0NSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE0MSA+PgpzdHJlYW0KeJw9j8EOwzAIQ+/5Cv9ApNgpoXxPp2qH7v+vI0u7C3oCY4yF0NAbqprDhmCb48XSJVRr+BTFQCU3yJlgDqWk0h1HkXpiOBhcHrQbjuKx6PoRu5JmfdDGQrolaIB7rFNp3KZxE8QdNQXqKeqco7wQuZ+pZ9g0kt00s5JzuA2/e89T1/+nq7zL+QW9dy7+CmVuZHN0cmVhbQplbmRvYmoKNDYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTUgPj4Kc3RyZWFtCnicNVE5DgMhDOz3Ff5AJIwveE+iKM3+v82M0VYewVyGtJQhmfJSk6gh5VM+epkunLrc18xqNOeWtC1zgLi2vC+tksCJZoiDwWmYuAGaPAFD19GoUUMXHtDUpVMosNwEPoq3bg/dY7WBl7Yh54kgYigZLEHNqUUTFm3PJ6Q1v16LG96X7d3IU6XGlhiBBgFWOBzX6NfwlT1PJtF0FTLUqzXLGAkTRSI8+Y6m1RPrWjTSMhLUxhGsagO8O/0wTgAAE3HLAmSfSpSz5MRvsfSzBlf6/gGfR1SWCmVuZHN0cmVhbQplbmRvYmoKMTUgMCBvYmoKPDwgL0Jhc2VGb250IC9EZWphVnVTYW5zIC9DaGFyUHJvY3MgMTYgMCBSCi9FbmNvZGluZyA8PAovRGlmZmVyZW5jZXMgWyAzMiAvc3BhY2UgNDYgL3BlcmlvZCA0OCAvemVybyAvb25lIC90d28gL3RocmVlIC9mb3VyIC9maXZlIC9zaXggNTYKL2VpZ2h0IDY3IC9DIDcxIC9HIDk3IC9hIC9iIDEwMCAvZCAvZSAxMDMgL2cgL2ggL2kgMTA4IC9sIC9tIC9uIC9vIDExNCAvcgovcyAvdCAvdSAxMTkgL3cgMTIxIC95IF0KL1R5cGUgL0VuY29kaW5nID4+Ci9GaXJzdENoYXIgMCAvRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250RGVzY3JpcHRvciAxNCAwIFIKL0ZvbnRNYXRyaXggWyAwLjAwMSAwIDAgMC4wMDEgMCAwIF0gL0xhc3RDaGFyIDI1NSAvTmFtZSAvRGVqYVZ1U2FucwovU3VidHlwZSAvVHlwZTMgL1R5cGUgL0ZvbnQgL1dpZHRocyAxMyAwIFIgPj4KZW5kb2JqCjE0IDAgb2JqCjw8IC9Bc2NlbnQgOTI5IC9DYXBIZWlnaHQgMCAvRGVzY2VudCAtMjM2IC9GbGFncyAzMgovRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250TmFtZSAvRGVqYVZ1U2FucyAvSXRhbGljQW5nbGUgMAovTWF4V2lkdGggMTM0MiAvU3RlbVYgMCAvVHlwZSAvRm9udERlc2NyaXB0b3IgL1hIZWlnaHQgMCA+PgplbmRvYmoKMTMgMCBvYmoKWyA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMAo2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDMxOCA0MDEgNDYwIDgzOCA2MzYKOTUwIDc4MCAyNzUgMzkwIDM5MCA1MDAgODM4IDMxOCAzNjEgMzE4IDMzNyA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2CjYzNiA2MzYgMzM3IDMzNyA4MzggODM4IDgzOCA1MzEgMTAwMCA2ODQgNjg2IDY5OCA3NzAgNjMyIDU3NSA3NzUgNzUyIDI5NQoyOTUgNjU2IDU1NyA4NjMgNzQ4IDc4NyA2MDMgNzg3IDY5NSA2MzUgNjExIDczMiA2ODQgOTg5IDY4NSA2MTEgNjg1IDM5MCAzMzcKMzkwIDgzOCA1MDAgNTAwIDYxMyA2MzUgNTUwIDYzNSA2MTUgMzUyIDYzNSA2MzQgMjc4IDI3OCA1NzkgMjc4IDk3NCA2MzQgNjEyCjYzNSA2MzUgNDExIDUyMSAzOTIgNjM0IDU5MiA4MTggNTkyIDU5MiA1MjUgNjM2IDMzNyA2MzYgODM4IDYwMCA2MzYgNjAwIDMxOAozNTIgNTE4IDEwMDAgNTAwIDUwMCA1MDAgMTM0MiA2MzUgNDAwIDEwNzAgNjAwIDY4NSA2MDAgNjAwIDMxOCAzMTggNTE4IDUxOAo1OTAgNTAwIDEwMDAgNTAwIDEwMDAgNTIxIDQwMCAxMDIzIDYwMCA1MjUgNjExIDMxOCA0MDEgNjM2IDYzNiA2MzYgNjM2IDMzNwo1MDAgNTAwIDEwMDAgNDcxIDYxMiA4MzggMzYxIDEwMDAgNTAwIDUwMCA4MzggNDAxIDQwMSA1MDAgNjM2IDYzNiAzMTggNTAwCjQwMSA0NzEgNjEyIDk2OSA5NjkgOTY5IDUzMSA2ODQgNjg0IDY4NCA2ODQgNjg0IDY4NCA5NzQgNjk4IDYzMiA2MzIgNjMyIDYzMgoyOTUgMjk1IDI5NSAyOTUgNzc1IDc0OCA3ODcgNzg3IDc4NyA3ODcgNzg3IDgzOCA3ODcgNzMyIDczMiA3MzIgNzMyIDYxMSA2MDUKNjMwIDYxMyA2MTMgNjEzIDYxMyA2MTMgNjEzIDk4MiA1NTAgNjE1IDYxNSA2MTUgNjE1IDI3OCAyNzggMjc4IDI3OCA2MTIgNjM0CjYxMiA2MTIgNjEyIDYxMiA2MTIgODM4IDYxMiA2MzQgNjM0IDYzNCA2MzQgNTkyIDYzNSA1OTIgXQplbmRvYmoKMTYgMCBvYmoKPDwgL0MgMTcgMCBSIC9HIDE4IDAgUiAvYSAxOSAwIFIgL2IgMjAgMCBSIC9kIDIxIDAgUiAvZSAyMiAwIFIKL2VpZ2h0IDIzIDAgUiAvZml2ZSAyNCAwIFIgL2ZvdXIgMjUgMCBSIC9nIDI2IDAgUiAvaCAyNyAwIFIgL2kgMjggMCBSCi9sIDI5IDAgUiAvbSAzMCAwIFIgL24gMzIgMCBSIC9vIDMzIDAgUiAvb25lIDM0IDAgUiAvcGVyaW9kIDM1IDAgUgovciAzNiAwIFIgL3MgMzcgMCBSIC9zaXggMzggMCBSIC9zcGFjZSAzOSAwIFIgL3QgNDAgMCBSIC90aHJlZSA0MSAwIFIKL3R3byA0MiAwIFIgL3UgNDMgMCBSIC93IDQ0IDAgUiAveSA0NSAwIFIgL3plcm8gNDYgMCBSID4+CmVuZG9iagozIDAgb2JqCjw8IC9GMSAxNSAwIFIgPj4KZW5kb2JqCjQgMCBvYmoKPDwgL0ExIDw8IC9DQSAwIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0EyIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0EzIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDAuNSA+PiA+PgplbmRvYmoKNSAwIG9iago8PCA+PgplbmRvYmoKNiAwIG9iago8PCA+PgplbmRvYmoKNyAwIG9iago8PCAvRjEtRGVqYVZ1U2Fucy1taW51cyAzMSAwIFIgPj4KZW5kb2JqCjIgMCBvYmoKPDwgL0NvdW50IDEgL0tpZHMgWyAxMSAwIFIgXSAvVHlwZSAvUGFnZXMgPj4KZW5kb2JqCjQ3IDAgb2JqCjw8IC9DcmVhdGlvbkRhdGUgKEQ6MjAyMTA5MTYxNDM3MjIrMDInMDAnKQovQ3JlYXRvciAoTWF0cGxvdGxpYiB2My40LjMsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAoTWF0cGxvdGxpYiBwZGYgYmFja2VuZCB2My40LjMpID4+CmVuZG9iagp4cmVmCjAgNDgKMDAwMDAwMDAwMCA2NTUzNSBmIAowMDAwMDAwMDE2IDAwMDAwIG4gCjAwMDAwMjEwMDEgMDAwMDAgbiAKMDAwMDAyMDczOCAwMDAwMCBuIAowMDAwMDIwNzcwIDAwMDAwIG4gCjAwMDAwMjA5MTAgMDAwMDAgbiAKMDAwMDAyMDkzMSAwMDAwMCBuIAowMDAwMDIwOTUyIDAwMDAwIG4gCjAwMDAwMDAwNjUgMDAwMDAgbiAKMDAwMDAwMDM5OSAwMDAwMCBuIAowMDAwMDEwNTU5IDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAxMDUzNyAwMDAwMCBuIAowMDAwMDE5MzQxIDAwMDAwIG4gCjAwMDAwMTkxNDEgMDAwMDAgbiAKMDAwMDAxODY4NCAwMDAwMCBuIAowMDAwMDIwMzk0IDAwMDAwIG4gCjAwMDAwMTA1NzkgMDAwMDAgbiAKMDAwMDAxMDg4NyAwMDAwMCBuIAowMDAwMDExMjA3IDAwMDAwIG4gCjAwMDAwMTE1ODcgMDAwMDAgbiAKMDAwMDAxMTkwNCAwMDAwMCBuIAowMDAwMDEyMjA4IDAwMDAwIG4gCjAwMDAwMTI1MzAgMDAwMDAgbiAKMDAwMDAxMjk5OCAwMDAwMCBuIAowMDAwMDEzMzIwIDAwMDAwIG4gCjAwMDAwMTM0ODYgMDAwMDAgbiAKMDAwMDAxMzkwMCAwMDAwMCBuIAowMDAwMDE0MTM3IDAwMDAwIG4gCjAwMDAwMTQyODEgMDAwMDAgbiAKMDAwMDAxNDQwMCAwMDAwMCBuIAowMDAwMDE0NzMxIDAwMDAwIG4gCjAwMDAwMTQ5MDMgMDAwMDAgbiAKMDAwMDAxNTEzOSAwMDAwMCBuIAowMDAwMDE1NDMwIDAwMDAwIG4gCjAwMDAwMTU1ODUgMDAwMDAgbiAKMDAwMDAxNTcwOCAwMDAwMCBuIAowMDAwMDE1OTQxIDAwMDAwIG4gCjAwMDAwMTYzNDggMDAwMDAgbiAKMDAwMDAxNjc0MSAwMDAwMCBuIAowMDAwMDE2ODMxIDAwMDAwIG4gCjAwMDAwMTcwMzcgMDAwMDAgbiAKMDAwMDAxNzQ1MCAwMDAwMCBuIAowMDAwMDE3Nzc0IDAwMDAwIG4gCjAwMDAwMTgwMjEgMDAwMDAgbiAKMDAwMDAxODE4MiAwMDAwMCBuIAowMDAwMDE4Mzk2IDAwMDAwIG4gCjAwMDAwMjEwNjEgMDAwMDAgbiAKdHJhaWxlcgo8PCAvSW5mbyA0NyAwIFIgL1Jvb3QgMSAwIFIgL1NpemUgNDggPj4Kc3RhcnR4cmVmCjIxMjE4CiUlRU9GCg==\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["layers.0.weight - Variance: 3.414905950194225e-05\n", "layers.2.weight - Variance: 3.843478407361545e-05\n", "layers.4.weight - Variance: 4.713246744358912e-05\n", "layers.6.weight - Variance: 0.00010930334246950224\n", "layers.8.weight - Variance: 0.0017839515348896384\n"]}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["Layer 0 - Variance: 1.0256913900375366\n", "Layer 2 - Variance: 1.0101124048233032\n", "Layer 4 - Variance: 1.0158814191818237\n", "Layer 6 - Variance: 1.1398581266403198\n", "Layer 8 - Variance: 0.46903371810913086\n"]}], "source": ["def kaiming_init(model):\n", " for name, param in model.named_parameters():\n", " if name.endswith(\".bias\"):\n", " param.data.fill_(0)\n", " elif name.startswith(\"layers.0\"): # The first layer does not have ReLU applied on its input\n", " param.data.normal_(0, 1 / math.sqrt(param.shape[1]))\n", " else:\n", " param.data.normal_(0, math.sqrt(2) / math.sqrt(param.shape[1]))\n", "\n", "\n", "model = BaseNetwork(act_fn=nn.ReLU()).to(device)\n", "kaiming_init(model)\n", "visualize_gradients(model, print_variance=True)\n", "visualize_activations(model, print_variance=True)"]}, {"cell_type": "markdown", "id": "fee08055", "metadata": {"papermill": {"duration": 0.100612, "end_time": "2021-09-16T12:37:31.027664", "exception": false, "start_time": "2021-09-16T12:37:30.927052", "status": "completed"}, "tags": []}, "source": ["The variance stays stable across layers.\n", "We can conclude that the Kaiming initialization indeed works well for ReLU-based networks.\n", "Note that for Leaky-ReLU etc., we have to slightly adjust the factor of $2$ in the variance as half of the values are not set to zero anymore.\n", "PyTorch provides a function to calculate this factor for many activation\n", "function, see `torch.nn.init.calculate_gain`\n", "([link](https://pytorch.org/docs/stable/nn.init.html#torch.nn.init.calculate_gain))."]}, {"cell_type": "markdown", "id": "7357c14d", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.101571, "end_time": "2021-09-16T12:37:31.230155", "exception": false, "start_time": "2021-09-16T12:37:31.128584", "status": "completed"}, "tags": []}, "source": ["## Optimization\n", "\n", "
\n", "\n", "Besides initialization, selecting a suitable optimization algorithm can be an important choice for deep neural networks.\n", "Before taking a closer look at them, we should define code for training the models.\n", "Most of the following code is copied from the previous tutorial, and only slightly altered to fit our needs."]}, {"cell_type": "code", "execution_count": 20, "id": "612da712", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:31.465358Z", "iopub.status.busy": "2021-09-16T12:37:31.463919Z", "iopub.status.idle": "2021-09-16T12:37:31.467057Z", "shell.execute_reply": "2021-09-16T12:37:31.467443Z"}, "lines_to_next_cell": 2, "papermill": {"duration": 0.136711, "end_time": "2021-09-16T12:37:31.467583", "exception": false, "start_time": "2021-09-16T12:37:31.330872", "status": "completed"}, "tags": []}, "outputs": [], "source": ["def _get_config_file(model_path, model_name):\n", " return os.path.join(model_path, model_name + \".config\")\n", "\n", "\n", "def _get_model_file(model_path, model_name):\n", " return os.path.join(model_path, model_name + \".tar\")\n", "\n", "\n", "def _get_result_file(model_path, model_name):\n", " return os.path.join(model_path, model_name + \"_results.json\")\n", "\n", "\n", "def load_model(model_path, model_name, net=None):\n", " config_file = _get_config_file(model_path, model_name)\n", " model_file = _get_model_file(model_path, model_name)\n", " assert os.path.isfile(\n", " config_file\n", " ), f'Could not find the config file \"{config_file}\". Are you sure this is the correct path and you have your model config stored here?'\n", " assert os.path.isfile(\n", " model_file\n", " ), f'Could not find the model file \"{model_file}\". Are you sure this is the correct path and you have your model stored here?'\n", " with open(config_file) as f:\n", " config_dict = json.load(f)\n", " if net is None:\n", " act_fn_name = config_dict[\"act_fn\"].pop(\"name\").lower()\n", " assert (\n", " act_fn_name in act_fn_by_name\n", " ), f'Unknown activation function \"{act_fn_name}\". Please add it to the \"act_fn_by_name\" dict.'\n", " act_fn = act_fn_by_name[act_fn_name]()\n", " net = BaseNetwork(act_fn=act_fn, **config_dict)\n", " net.load_state_dict(torch.load(model_file))\n", " return net\n", "\n", "\n", "def save_model(model, model_path, model_name):\n", " config_dict = model.config\n", " os.makedirs(model_path, exist_ok=True)\n", " config_file = _get_config_file(model_path, model_name)\n", " model_file = _get_model_file(model_path, model_name)\n", " with open(config_file, \"w\") as f:\n", " json.dump(config_dict, f)\n", " torch.save(model.state_dict(), model_file)\n", "\n", "\n", "def train_model(net, model_name, optim_func, max_epochs=50, batch_size=256, overwrite=False):\n", " \"\"\"Train a model on the training set of FashionMNIST.\n", "\n", " Args:\n", " net: Object of BaseNetwork\n", " model_name: (str) Name of the model, used for creating the checkpoint names\n", " max_epochs: Number of epochs we want to (maximally) train for\n", " patience: If the performance on the validation set has not improved for #patience epochs, we stop training early\n", " batch_size: Size of batches used in training\n", " overwrite: Determines how to handle the case when there already exists a checkpoint. If True, it will be overwritten. Otherwise, we skip training.\n", " \"\"\"\n", " file_exists = os.path.isfile(_get_model_file(CHECKPOINT_PATH, model_name))\n", " if file_exists and not overwrite:\n", " print(f'Model file of \"{model_name}\" already exists. Skipping training...')\n", " with open(_get_result_file(CHECKPOINT_PATH, model_name)) as f:\n", " results = json.load(f)\n", " else:\n", " if file_exists:\n", " print(\"Model file exists, but will be overwritten...\")\n", "\n", " # Defining optimizer, loss and data loader\n", " optimizer = optim_func(net.parameters())\n", " loss_module = nn.CrossEntropyLoss()\n", " train_loader_local = data.DataLoader(\n", " train_set, batch_size=batch_size, shuffle=True, drop_last=True, pin_memory=True\n", " )\n", "\n", " results = None\n", " val_scores = []\n", " train_losses, train_scores = [], []\n", " best_val_epoch = -1\n", " for epoch in range(max_epochs):\n", " train_acc, val_acc, epoch_losses = epoch_iteration(\n", " net, loss_module, optimizer, train_loader_local, val_loader, epoch\n", " )\n", " train_scores.append(train_acc)\n", " val_scores.append(val_acc)\n", " train_losses += epoch_losses\n", "\n", " if len(val_scores) == 1 or val_acc > val_scores[best_val_epoch]:\n", " print(\"\\t (New best performance, saving model...)\")\n", " save_model(net, CHECKPOINT_PATH, model_name)\n", " best_val_epoch = epoch\n", "\n", " if results is None:\n", " load_model(CHECKPOINT_PATH, model_name, net=net)\n", " test_acc = test_model(net, test_loader)\n", " results = {\n", " \"test_acc\": test_acc,\n", " \"val_scores\": val_scores,\n", " \"train_losses\": train_losses,\n", " \"train_scores\": train_scores,\n", " }\n", " with open(_get_result_file(CHECKPOINT_PATH, model_name), \"w\") as f:\n", " json.dump(results, f)\n", "\n", " # Plot a curve of the validation accuracy\n", " sns.set()\n", " plt.plot([i for i in range(1, len(results[\"train_scores\"]) + 1)], results[\"train_scores\"], label=\"Train\")\n", " plt.plot([i for i in range(1, len(results[\"val_scores\"]) + 1)], results[\"val_scores\"], label=\"Val\")\n", " plt.xlabel(\"Epochs\")\n", " plt.ylabel(\"Validation accuracy\")\n", " plt.ylim(min(results[\"val_scores\"]), max(results[\"train_scores\"]) * 1.01)\n", " plt.title(f\"Validation performance of {model_name}\")\n", " plt.legend()\n", " plt.show()\n", " plt.close()\n", "\n", " print((f\" Test accuracy: {results['test_acc']*100.0:4.2f}% \").center(50, \"=\") + \"\\n\")\n", " return results\n", "\n", "\n", "def epoch_iteration(net, loss_module, optimizer, train_loader_local, val_loader, epoch):\n", " ############\n", " # Training #\n", " ############\n", " net.train()\n", " true_preds, count = 0.0, 0\n", " epoch_losses = []\n", " t = tqdm(train_loader_local, leave=False)\n", " for imgs, labels in t:\n", " imgs, labels = imgs.to(device), labels.to(device)\n", " optimizer.zero_grad()\n", " preds = net(imgs)\n", " loss = loss_module(preds, labels)\n", " loss.backward()\n", " optimizer.step()\n", " # Record statistics during training\n", " true_preds += (preds.argmax(dim=-1) == labels).sum().item()\n", " count += labels.shape[0]\n", " t.set_description(f\"Epoch {epoch+1}: loss={loss.item():4.2f}\")\n", " epoch_losses.append(loss.item())\n", " train_acc = true_preds / count\n", "\n", " ##############\n", " # Validation #\n", " ##############\n", " val_acc = test_model(net, val_loader)\n", " print(\n", " f\"[Epoch {epoch+1:2i}] Training accuracy: {train_acc*100.0:05.2f}%, Validation accuracy: {val_acc*100.0:05.2f}%\"\n", " )\n", " return train_acc, val_acc, epoch_losses\n", "\n", "\n", "def test_model(net, data_loader):\n", " \"\"\"Test a model on a specified dataset.\n", "\n", " Args:\n", " net: Trained model of type BaseNetwork\n", " data_loader: DataLoader object of the dataset to test on (validation or test)\n", " \"\"\"\n", " net.eval()\n", " true_preds, count = 0.0, 0\n", " for imgs, labels in data_loader:\n", " imgs, labels = imgs.to(device), labels.to(device)\n", " with torch.no_grad():\n", " preds = net(imgs).argmax(dim=-1)\n", " true_preds += (preds == labels).sum().item()\n", " count += labels.shape[0]\n", " test_acc = true_preds / count\n", " return test_acc"]}, {"cell_type": "markdown", "id": "ed5321ba", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.101675, "end_time": "2021-09-16T12:37:31.672791", "exception": false, "start_time": "2021-09-16T12:37:31.571116", "status": "completed"}, "tags": []}, "source": ["First, we need to understand what an optimizer actually does.\n", "The optimizer is responsible to update the network's parameters given the gradients.\n", "Hence, we effectively implement a function $w^{t} = f(w^{t-1}, g^{t}, ...)$ with $w$ being the parameters, and $g^{t} = \\nabla_{w^{(t-1)}} \\mathcal{L}^{(t)}$ the gradients at time step $t$.\n", "A common, additional parameter to this function is the learning rate, here denoted by $\\eta$.\n", "Usually, the learning rate can be seen as the \"step size\" of the update.\n", "A higher learning rate means that we change the weights more in the direction of the gradients, a smaller means we take shorter steps.\n", "\n", "As most optimizers only differ in the implementation of $f$, we can define a template for an optimizer in PyTorch below.\n", "We take as input the parameters of a model and a learning rate.\n", "The function `zero_grad` sets the gradients of all parameters to zero, which we have to do before calling `loss.backward()`.\n", "Finally, the `step()` function tells the optimizer to update all weights based on their gradients.\n", "The template is setup below:"]}, {"cell_type": "code", "execution_count": 21, "id": "d0480429", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:31.880757Z", "iopub.status.busy": "2021-09-16T12:37:31.880264Z", "iopub.status.idle": "2021-09-16T12:37:31.881827Z", "shell.execute_reply": "2021-09-16T12:37:31.882297Z"}, "lines_to_next_cell": 2, "papermill": {"duration": 0.109458, "end_time": "2021-09-16T12:37:31.882433", "exception": false, "start_time": "2021-09-16T12:37:31.772975", "status": "completed"}, "tags": []}, "outputs": [], "source": ["class OptimizerTemplate:\n", " def __init__(self, params, lr):\n", " self.params = list(params)\n", " self.lr = lr\n", "\n", " def zero_grad(self):\n", " # Set gradients of all parameters to zero\n", " for p in self.params:\n", " if p.grad is not None:\n", " p.grad.detach_() # For second-order optimizers important\n", " p.grad.zero_()\n", "\n", " @torch.no_grad()\n", " def step(self):\n", " # Apply update step to all parameters\n", " for p in self.params:\n", " if p.grad is None: # We skip parameters without any gradients\n", " continue\n", " self.update_param(p)\n", "\n", " def update_param(self, p):\n", " # To be implemented in optimizer-specific classes\n", " raise NotImplementedError"]}, {"cell_type": "markdown", "id": "ea589ab3", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.100149, "end_time": "2021-09-16T12:37:32.084995", "exception": false, "start_time": "2021-09-16T12:37:31.984846", "status": "completed"}, "tags": []}, "source": ["The first optimizer we are going to implement is the standard Stochastic Gradient Descent (SGD).\n", "SGD updates the parameters using the following equation:\n", "\n", "$$\n", "\\begin{split}\n", " w^{(t)} & = w^{(t-1)} - \\eta \\cdot g^{(t)}\n", "\\end{split}\n", "$$\n", "\n", "As simple as the equation is also our implementation of SGD:"]}, {"cell_type": "code", "execution_count": 22, "id": "347cd0a8", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:32.292286Z", "iopub.status.busy": "2021-09-16T12:37:32.291809Z", "iopub.status.idle": "2021-09-16T12:37:32.293396Z", "shell.execute_reply": "2021-09-16T12:37:32.293806Z"}, "lines_to_next_cell": 2, "papermill": {"duration": 0.107225, "end_time": "2021-09-16T12:37:32.293938", "exception": false, "start_time": "2021-09-16T12:37:32.186713", "status": "completed"}, "tags": []}, "outputs": [], "source": ["class SGD(OptimizerTemplate):\n", " def __init__(self, params, lr):\n", " super().__init__(params, lr)\n", "\n", " def update_param(self, p):\n", " p_update = -self.lr * p.grad\n", " p.add_(p_update) # In-place update => saves memory and does not create computation graph"]}, {"cell_type": "markdown", "id": "83492492", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.102455, "end_time": "2021-09-16T12:37:32.496757", "exception": false, "start_time": "2021-09-16T12:37:32.394302", "status": "completed"}, "tags": []}, "source": ["In the lecture, we also have discussed the concept of momentum which replaces the gradient in the update by an exponential average of all past gradients including the current one:\n", "\n", "$$\n", "\\begin{split}\n", " m^{(t)} & = \\beta_1 m^{(t-1)} + (1 - \\beta_1)\\cdot g^{(t)}\\\\\n", " w^{(t)} & = w^{(t-1)} - \\eta \\cdot m^{(t)}\\\\\n", "\\end{split}\n", "$$\n", "\n", "Let's also implement it below:"]}, {"cell_type": "code", "execution_count": 23, "id": "28a7d8c5", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:32.707228Z", "iopub.status.busy": "2021-09-16T12:37:32.706744Z", "iopub.status.idle": "2021-09-16T12:37:32.708337Z", "shell.execute_reply": "2021-09-16T12:37:32.708746Z"}, "lines_to_next_cell": 2, "papermill": {"duration": 0.10963, "end_time": "2021-09-16T12:37:32.708874", "exception": false, "start_time": "2021-09-16T12:37:32.599244", "status": "completed"}, "tags": []}, "outputs": [], "source": ["class SGDMomentum(OptimizerTemplate):\n", " def __init__(self, params, lr, momentum=0.0):\n", " super().__init__(params, lr)\n", " self.momentum = momentum # Corresponds to beta_1 in the equation above\n", " self.param_momentum = {p: torch.zeros_like(p.data) for p in self.params} # Dict to store m_t\n", "\n", " def update_param(self, p):\n", " self.param_momentum[p] = (1 - self.momentum) * p.grad + self.momentum * self.param_momentum[p]\n", " p_update = -self.lr * self.param_momentum[p]\n", " p.add_(p_update)"]}, {"cell_type": "markdown", "id": "1dd09f27", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.103723, "end_time": "2021-09-16T12:37:32.921682", "exception": false, "start_time": "2021-09-16T12:37:32.817959", "status": "completed"}, "tags": []}, "source": ["Finally, we arrive at Adam.\n", "Adam combines the idea of momentum with an adaptive learning rate, which is based on an exponential average of the squared gradients, i.e. the gradients norm.\n", "Furthermore, we add a bias correction for the momentum and adaptive learning rate for the first iterations:\n", "\n", "$$\n", "\\begin{split}\n", " m^{(t)} & = \\beta_1 m^{(t-1)} + (1 - \\beta_1)\\cdot g^{(t)}\\\\\n", " v^{(t)} & = \\beta_2 v^{(t-1)} + (1 - \\beta_2)\\cdot \\left(g^{(t)}\\right)^2\\\\\n", " \\hat{m}^{(t)} & = \\frac{m^{(t)}}{1-\\beta^{t}_1}, \\hat{v}^{(t)} = \\frac{v^{(t)}}{1-\\beta^{t}_2}\\\\\n", " w^{(t)} & = w^{(t-1)} - \\frac{\\eta}{\\sqrt{v^{(t)}} + \\epsilon}\\circ \\hat{m}^{(t)}\\\\\n", "\\end{split}\n", "$$\n", "\n", "Epsilon is a small constant used to improve numerical stability for very small gradient norms.\n", "Remember that the adaptive learning rate does not replace the learning\n", "rate hyperparameter $\\eta$, but rather acts as an extra factor and\n", "ensures that the gradients of various parameters have a similar norm."]}, {"cell_type": "code", "execution_count": 24, "id": "5405912f", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:33.132645Z", "iopub.status.busy": "2021-09-16T12:37:33.132120Z", "iopub.status.idle": "2021-09-16T12:37:33.134264Z", "shell.execute_reply": "2021-09-16T12:37:33.133784Z"}, "papermill": {"duration": 0.111044, "end_time": "2021-09-16T12:37:33.134374", "exception": false, "start_time": "2021-09-16T12:37:33.023330", "status": "completed"}, "tags": []}, "outputs": [], "source": ["class Adam(OptimizerTemplate):\n", " def __init__(self, params, lr, beta1=0.9, beta2=0.999, eps=1e-8):\n", " super().__init__(params, lr)\n", " self.beta1 = beta1\n", " self.beta2 = beta2\n", " self.eps = eps\n", " self.param_step = {p: 0 for p in self.params} # Remembers \"t\" for each parameter for bias correction\n", " self.param_momentum = {p: torch.zeros_like(p.data) for p in self.params}\n", " self.param_2nd_momentum = {p: torch.zeros_like(p.data) for p in self.params}\n", "\n", " def update_param(self, p):\n", " self.param_step[p] += 1\n", "\n", " self.param_momentum[p] = (1 - self.beta1) * p.grad + self.beta1 * self.param_momentum[p]\n", " self.param_2nd_momentum[p] = (1 - self.beta2) * (p.grad) ** 2 + self.beta2 * self.param_2nd_momentum[p]\n", "\n", " bias_correction_1 = 1 - self.beta1 ** self.param_step[p]\n", " bias_correction_2 = 1 - self.beta2 ** self.param_step[p]\n", "\n", " p_2nd_mom = self.param_2nd_momentum[p] / bias_correction_2\n", " p_mom = self.param_momentum[p] / bias_correction_1\n", " p_lr = self.lr / (torch.sqrt(p_2nd_mom) + self.eps)\n", " p_update = -p_lr * p_mom\n", "\n", " p.add_(p_update)"]}, {"cell_type": "markdown", "id": "83ea9b45", "metadata": {"papermill": {"duration": 0.099947, "end_time": "2021-09-16T12:37:33.348929", "exception": false, "start_time": "2021-09-16T12:37:33.248982", "status": "completed"}, "tags": []}, "source": ["### Comparing optimizers on model training\n", "\n", "After we have implemented three optimizers (SGD, SGD with momentum, and Adam), we can start to analyze and compare them.\n", "First, we test them on how well they can optimize a neural network on the FashionMNIST dataset.\n", "We use again our linear network, this time with a ReLU activation and the kaiming initialization, which we have found before to work well for ReLU-based networks.\n", "Note that the model is over-parameterized for this task, and we can achieve similar performance with a much smaller network (for example `100,100,100`).\n", "However, our main interest is in how well the optimizer can train *deep*\n", "neural networks, hence the over-parameterization."]}, {"cell_type": "code", "execution_count": 25, "id": "017e170c", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:33.553073Z", "iopub.status.busy": "2021-09-16T12:37:33.552597Z", "iopub.status.idle": "2021-09-16T12:37:33.565096Z", "shell.execute_reply": "2021-09-16T12:37:33.564688Z"}, "papermill": {"duration": 0.116391, "end_time": "2021-09-16T12:37:33.565206", "exception": false, "start_time": "2021-09-16T12:37:33.448815", "status": "completed"}, "tags": []}, "outputs": [], "source": ["base_model = BaseNetwork(act_fn=nn.ReLU(), hidden_sizes=[512, 256, 256, 128])\n", "kaiming_init(base_model)"]}, {"cell_type": "markdown", "id": "fe2e1e1d", "metadata": {"papermill": {"duration": 0.116162, "end_time": "2021-09-16T12:37:33.782066", "exception": false, "start_time": "2021-09-16T12:37:33.665904", "status": "completed"}, "tags": []}, "source": ["For a fair comparison, we train the exact same model with the same seed with the three optimizers below.\n", "Feel free to change the hyperparameters if you want (however, you have to train your own model then)."]}, {"cell_type": "code", "execution_count": 26, "id": "2d013e73", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:33.992549Z", "iopub.status.busy": "2021-09-16T12:37:33.992058Z", "iopub.status.idle": "2021-09-16T12:37:34.371016Z", "shell.execute_reply": "2021-09-16T12:37:34.370547Z"}, "papermill": {"duration": 0.48688, "end_time": "2021-09-16T12:37:34.371139", "exception": false, "start_time": "2021-09-16T12:37:33.884259", "status": "completed"}, "tags": []}, "outputs": [{"name": "stdout", "output_type": "stream", "text": ["Model file of \"FashionMNIST_SGD\" already exists. Skipping training...\n"]}, {"data": {"application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9Bbm5vdHMgMTAgMCBSIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDM5OC44MjUgMjgyLjczMDYyNSBdIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovVHlwZSAvUGFnZSA+PgplbmRvYmoKOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEyIDAgUiA+PgpzdHJlYW0KeJy1WU2PGzcSvfev6KN9MIdVRVaRRxtJBhsgARIPkkMQLAbKeG1jbMfjZBf77/cVpVGTkiNrI2mMGbee2EW++ngsdtP8drp6TvO/Ps34M8f5LX7/g+tr/zxFfHo3SS2hcMb1/faaCweTqLi8x7Dh4+tpejXFUMk0WcylzLsfUo1UNVqZH3zS670B2w/TzuhpytoWkDhUzT6hL5CCDtj9FmONIcUNuLl3wNpqP857ZkVSKDOThZTmh7v55/n9fPWc1676Fr9v8bt21dVXd/9+s7r78frFtPo0aQospRYd1rigw+zTy+mH+eOj4RgoIwyPttvH6w06fZwIvnoW8RUsWDMWKc2cAzXPr95NL27mq29oJppvXrXo3fw2/TI/iU/nX+ebb6evb6Yf2nznY0tRAucKMwPdDj4D31qDNWNR8jF88wX5wgaXWHHbwHeBz8CXJG+sxZzrMYzpkiEuFAQlyDJSXuBzULa0sRZNj6N8wSgzxyDMNfJAuYPPQJlJNtZitaMo8wWjjGoNkqQUGikv8DkoZ95YIyrHUb5klLGZiaYdpV7QcxDGNtSMRZJ6FGG5YIxFLEjJReq4fy7wGSiLxI010liOonzBGIspCGmhMlJe4HNQ1rqxRs3ilymnIcp9wkR0UrgVO15/O4+3f/37h9XrT5fz2sZSthBVSErXZC3Y364JXhvHf0US9nSG+h1oXkJJF2eK/jUpQsgd0wU7kakV7OZJLEliOsxUL86UomGa5J3oQrUDT+TqjZ+WDBGVrHKYbLk8WUEbqrlSz3WLnUqVa4hFmVkspoNU6wVF/ZFqLkFVUVE91wU8lWxOIRWjpFKSHibLlydbElTILPfK1IGnkrWK7CzRIs675TDZy2sT2kMIUdHSi1MHnkiWox82SsVpgw4XbL28OrHfrzjt9+rUgadyRW+ioIpUYePDZC+vTt55sMUsPdctdipVTBJrBFHE9ZA6UYg7zYhbeeb2SHFAQS8iIaZ8oB356elsTgndLtRGbX5ye//mt9s/3nx4P9+uVn8+3K7+exZvhtyYY3dBZTJlm9C7rZ8UsWR8o4/PitL8Y+/1GV6fDDyQ7ZjrmW8IXuTJxMqMK84i7MJRLGiGZfGnDsU3RYA1h5SzH0yJcjA/vXi32D3lIE7ALbPXI76HEiVSaruNkFqVhhckMnZGmylRMBUPDXDWoOj6MkNnwQjnf/VpCS5JppYKcIQT23RbY/e0gVRChsLXNi9ujiWTw+7e1m84jN29VBFtx2wnYl5KpBkOkyx+/MYQqZrMcTgqVTPGMgvME9o9h5cDP5XiWWrUvFCq97tE8E7JoVhFE+t41VDhIgyimlB3HFsFc3RXpYxZ4WL2WxtMjKmKYUwbkTek+kM3pCobRU4NL4GwiIqkQfuSkaLFnYNECFWwOafZTaYUWdv4JMEkI55+8MbdNfvqOSOrwBwB9dxjbHIiDV8OvigpAa3UlqMWKGFvsIZDh+Fwx+HZivAjcdAQIC24ZndmezSaNflDDYRZUCXWllkjAhcVawMSCERazLvjZxMrqi1FJGogZHfWJodmJTavITShGqoU3hGkCCdUk+OM/DJLSB0WRomDV2l4DbmQJixTKrYv9nq9H86AyRsWJJcvBw16oBrbg7fkPVuKsY1H+44tg5E77lhB5FKb159NVpdX4PC+p3uD0YggECgPh6tFbqnZn8MSUtCd1knfKSJRdPPjIpH7x8vwDZJL2fLnRALOSkLYV9WVMXIZtaJ/yDxoRSoMHqNYuD5Ubnt0rxVIS63k8jBqhV8mVDmNWgHjtUatdUcqUBhIj9TSrJcKVE/MqaxruZMKP5fC4WvJWaQCxeN609KglwqYR1BT6x17qfDMwr5S6o5UAC9eX2lHKwgZUTSuFarTCvYCyJ6ho1YQRKHm1DSk0woUVVFXhVErnFV0k6NUOIxlqdgoFe5WQ4m3rOykwmeFKqa1FG2lwodTRmWNSuFRoCy6rvxFKTyygvg1Tp1StN0BMj3qBDIAemPcjiKdThDSm0SFZdSJIW8WnfDFVGS01VEnPCA4Wyfb0QnyS4X7y6gTvhzLWWIddcLtp2ht9Z1MuGugTdJEt5MJOBrxU207XicT2EUCcq9tnJ1KuGsE0xYdZcJxiHFuxdPJhK8GCb1WubFDYlcAFNlezzUvPVf/EOhzr4Ngcf910rvPv07C2CPfRi0jOwN/bTWCyLHtHjQoxrURfTTyF53al1q13+8eXn14eOptX0W1FPGf+cm72/eru/nDq/mbpxMWjGpcf4Ovbj+9xo3fff+Plzf/fHn9FXq86bHHu3ou6671+PeAb121d94G7r1IHN8GKtoZ547qFVSttRcTXpzUHNHhLp3oEpw59zhBG2QzGvIRs4H5iKLs12NXU48jte0xaAOesJu3afoZgdb16H59C7raclmw+wndXWzu7mHl1pZ3E22R7ZJgboMty79fsC3RboLFJ/teXfl70hfTF/vx+ch+vL0o1fakkyEd2BZbQaCtHKH9cjj0mtDWy0blUTr8/PUGWZ7QjS0PbZ883L553x9SGsm9XmL+P3qJR4KQ0pLzyHDB/hZFpFDcq3X63KnM0Bev/3mpD6ew6X8hh6JiCmVuZHN0cmVhbQplbmRvYmoKMTIgMCBvYmoKMTk1MwplbmRvYmoKMTAgMCBvYmoKWyBdCmVuZG9iagoxNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2NCA+PgpzdHJlYW0KeJw9kMERQyEIRO9WsSWAgEA9yWRy+L//a0CTXGQdYPepO4GQUYczw2fiyYPTsTRwbxWMawivI/QITQKTwMTBmngMCwGnYZFjLt9VllWnla6ajZ7XvWNB1WmXNQ1t2oHyrY8/wjXeo/Aa7B5CB7EodG5lWguZWDxrnDvMo8znfk7bdz0YrabUrDdy2dc9OsvUUF5a+4TOaLT9J9cvuzFeH4UUOQgKZW5kc3RyZWFtCmVuZG9iagoxOCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDgxID4+CnN0cmVhbQp4nE3Nuw3AIAwE0J4pPALg/z5RlCLZv40NEaGxn3QnnWCHCm5xWAy0Oxyt+NRTmH3oHhKSUHPdRFgzJdqEpF/6yzDDmFjItq83V65yvhbcHIsKZW5kc3RyZWFtCmVuZG9iagoxOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDc2ID4+CnN0cmVhbQp4nDM1N1UwULC0ABKmhuYK5kaWCimGXEA+iJXLBRPLAbPMTMyALENLZJaJsSGQZWJhhsQyNrGAyiJYBkAabE0OzPQcrgyuNAA1FxkFCmVuZHN0cmVhbQplbmRvYmoKMjAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDcgPj4Kc3RyZWFtCnicTVFJbsQwDLv7FfzAAJasxXlPikEP7f+vJR0U7cEQI0tc4u7ERBZetlDXQofjw0ZeCZuB74PWnPgaseI/2kaklT9UWyATMVEkdFE3GvdIN7wK0X6kgleq91jzEXcrzVs6drG/98G05pEqq0I85Ngc2Uha10TR8T203nNDdMoggT43IQdEaY5ehaS/9sN1bTS7tTazJ6qDR6aE8kmzGprTKWbIbKjHbSpWMgo3qoyK+1RGWg/yNs4ygJPjhDJaT3asJqL81CeXkBcTccIuOzsWYhMLG4e0H5U+sfx86834m2mtpZBxQSI0xaXfZ7zH53j/AJVPXCYKZW5kc3RyZWFtCmVuZG9iagoyMSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDUxID4+CnN0cmVhbQp4nDOyNFUwULC0ABKGluYK5kaWCimGXEA+iJXLBRPLAbMMgDRYaQ5MRQ5XBlcaAL+MDVYKZW5kc3RyZWFtCmVuZG9iagoyMiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDkwID4+CnN0cmVhbQp4nD2Oyw3AMAhD70zBCOFTAvtUVQ/J/teGfHrBD1vIuAkWDB+j2oWVA2+CsSd1YF1eAxVCFhlk5Ns7F4tKZha/miapE9Ikcd5EoTtNSp0PtNPb4IXnA/XpHewKZW5kc3RyZWFtCmVuZG9iagoyMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDc3ID4+CnN0cmVhbQp4nDWNwQ3AMAgD/0zBCDiFUPapqj7S/b8tRHzsMwjserJwpEwT9hF8gf6c9NI4ULTITBlo2rO+2CS5g5cjlCea0qti9edFD90fyZ4YDAplbmRzdHJlYW0KZW5kb2JqCjI0IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzQxID4+CnN0cmVhbQp4nDVSO9KbQQjrv1PoAp5Z3st5nMmk+HP/NgI7FSywQgLSAgeZeIkhqlGu+CVPMF4n8He9PI2fx7uQWvBUpB+4Nm3j/VizJgqWRiyF2ce+HyXkeGr8GwI9F2nCjExGDiQDcb/W5896kymH34A0bU4fJUkPogW7W8OOLwsySHpSw5Kd/LCuBVYXoQlzY00kI6dWpub52DNcxhNjJKiaBSTpE/epghFpxmPnrCUPMhxP9eLFr7fxWuYx9bKqQMY2wRxsJzPhFEUE4heUJDdxF00dxdHMWHO70FBS5L67h5OTXveXk6jAKyGcxVrCMUNPWeZkp0EJVK2cADOs174wTtNGCXdqur0r9vXzzCSM2xx2VkqmwTkO7mWTOYJkrzsmbMLjEPPePYKRmDe/iy2CK5c512T6sR9FG+mD4vqcqymzFSX8Q5U8seIa/5/f+/nz/P4HjCh+IwplbmRzdHJlYW0KZW5kb2JqCjI1IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNjYgPj4Kc3RyZWFtCnicMzM0VDBQ0DUCEmaGJgrmRpYKKYZcQD6IlcsFE8sBs8xMzIAsY1NTJJYBkDYyNYPTEBmgAXAGRH8GVxoAUmsUwAplbmRzdHJlYW0KZW5kb2JqCjI2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNzIgPj4Kc3RyZWFtCnicM7MwUTBQsABiM3MzBXMjS4UUQy4jCzOgQC6XBVggh8vQ0BDKMjYxUjA0NAWyTM2NoWIwjUBZS5BBOVD9OVwZXGkAdDISoQplbmRzdHJlYW0KZW5kb2JqCjI3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzA3ID4+CnN0cmVhbQp4nD2SS24DMQxD9z6FLhDA+tme86Qoupjef9snJemKHNkWRWqWukxZUx6QNJOEf+nwcLGd8jtsz2Zm4Fqil4nllOfQFWLuonzZzEZdWSfF6oRmOrfoUTkXBzZNqp+rLKXdLngO1yaeW/YRP7zQoB7UNS4JN3RXo2UpNGOq+3/Se/yMMuBqTF1sUqt7HzxeRFXo6AdHiSJjlxfn40EJ6UrCaFqIlXdFA0Hu8rTKewnu295qyLIHqZjOOylmsOt0Ui5uF4chHsjyqPDlo9hrQs/4sCsl9EjYhjNyJ+5oxubUyOKQ/t6NBEuPrmgh8+CvbtYuYLxTOkViZE5yrGmLVU73UBTTucO9DBD1bEVDKXOR1epfw84La5ZsFnhK+gUeo90mSw5W2duoTu+tPNnQ9x9a13QfCmVuZHN0cmVhbQplbmRvYmoKMjggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzIgPj4Kc3RyZWFtCnicNVFJbsQwDLv7FfzAANbuvCfFoIf2/9dSyhQIQCW2uCViYyMCLzH4OYjc+JI1oyZ+Z3JX/CxPhUfCreBJFIGX4V52gssbxmU/DjMfvJdWzqTGkwzIRTY9PBEy2CUQOjC7BnXYZtqJviHhsyNSzUaW09cS9NIqBMpTtt/pghJtq/pz+6wLbfvaE052e+pJ5ROI55aswGXjFZPFWAY9UblLMX2Q6myhJ6G8KJ+DbD5qiESXKGfgicHBKNAO7LntZ+JVIWhd3adtY6hGSsfTvw1NTZII+UQJZ7Y07hb+f8+9vtf7D04hVBEKZW5kc3RyZWFtCmVuZG9iagoyOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDIzMSA+PgpzdHJlYW0KeJw1TzmSBCEMy3mFPjBVGNtAv6entjbY+X+6kplOkPAhydMTHZl4mSMjsGbH21pkIGbgU0zFv/a0DxOq9+AeIpSLC2GGkXDWrONuno4X/3aVz1gH7zb4illeENjCTNZXFmcu2wVjaZzEOclujF0TsY11radTWEcwoQyEdLbDlCBzVKT0yY4y5ug4kSeei+/22yx2OX4O6ws2jSEV5/gqeoI2g6Lsee8CGnJB/13d+B5Fu+glIBsJFtZRYu6c5YRfvXZ0HrUoEnNCmkEuEyHN6SqmEJpQrLOjoFJRcKk+p+isn3/lX1wtCmVuZHN0cmVhbQplbmRvYmoKMzAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDkgPj4Kc3RyZWFtCnicPVA7jkQhDOs5hS/wJPIjcB5Gqy1m79+uA5opUEx+tjMk0BGBRwwxlK/jJa2groG/i0LxbuLrg8Igq0NSIM56D4h07KY2kRM6HZwzP2E3Y47ARTEGnOl0pj0HJjn7wgqEcxtl7FZIJ4mqIo7qM44pnip7n3gWLO3INlsnkj3kIOFSUonJpZ+Uyj9typQKOmbRBCwSueBkE004y7tJUowZlDLqHqZ2In2sPMijOuhkTc6sI5nZ00/bmfgccLdf2mROlcd0Hsz4nLTOgzkVuvfjiTYHTY3a6Oz3E2kqL1K7HVqdfnUSld0Y5xgSl2d/Gd9k//kH/odaIgplbmRzdHJlYW0KZW5kb2JqCjMxIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzk1ID4+CnN0cmVhbQp4nD1SS27FQAjb5xRcoNLwm895UlXdvPtva0NSqSq8iTHGMH3KkLnlS10ScYXJt16uWzymfC5bWpl5iLuLjSU+ttyX7iG2XXQusTgdR/ILMp0qRKjNqtGh+EKWhQeQTvChC8J9Of7jL4DB17ANuOE9MkGwJOYpQsZuURmaEkERYeeRFaikUJ9Zwt9R7uv3MgVqb4ylC2Mc9Am0BUJtSMQC6kAAROyUVK2QjmckE78V3WdiHGDn0bIBrhlURJZ77MeIqc6ojLxExD5PTfoolkwtVsZuUxlf/JSM1Hx0BSqpNPKU8tBVs9ALWIl5EvY5/Ej459ZsIYY6btbyieUfM8UyEs5gSzlgoZfjR+DbWXURrh25uM50gR+V1nBMtOt+yPVP/nTbWs11vHIIokDlTUHwuw6uRrHExDI+nY0peqIssBqavEYzwWEQEdb3w8gDGv1yvBA0p2sitFgim7ViRI2KbHM9vQTWTO/FOdbDE8Js753WobIzMyohgtq6hmrrQHazvvNwtp8/M+iibQplbmRzdHJlYW0KZW5kb2JqCjMyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTM2ID4+CnN0cmVhbQp4nE2PQQ4DMQgD73mFn0AgQHjPVlUP2/9fS9h20wseyYBsUQaBJYd4hxvh0dsP30U2FWfjnF9SKWIhmE9wnzBTHI0pd/Jjj4BxlGosp2h4XkvOTcMXLXcTLaWtl5MZb7jul/dHlW2RDUXPLQtC12yS+TKBB3wYmEd142mlx932bK/2/ADObDRJCmVuZHN0cmVhbQplbmRvYmoKMzMgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDkgPj4Kc3RyZWFtCnicTVFJigMwDLvnFfpAIV6TvKdDmUPn/9fKDoU5BAmvkpOWmFgLDzGEHyw9+JEhczf9G36i2btZepLJ2f+Y5yJTUfhSqC5iQl2IG8+hEfA9oWsSWbG98Tkso5lzvgcfhbgEM6EBY31JMrmo5pUhE04MdRwOWqTCuGtiw+Ja0TyN3G77RmZlJoQNj2RC3BiAiCDrArIYLJQ2NhMyWc4D7Q3JDVpg16kbUYuCK5TWCXSiVsSqzOCz5tZ2N0Mt8uCoffH6aFaXYIXRS/VYeF+FPpipmXbukkJ64U07IsweCqQyOy0rtXvE6m6B+j/LUvD9yff4Ha8PzfxcnAplbmRzdHJlYW0KZW5kb2JqCjM0IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggOTQgPj4Kc3RyZWFtCnicRY3BEcAgCAT/VEEJCgraTyaTh/b/jRAyfGDnDu6EBQu2eUYfBZUmXhVYB0pj3FCPQL3hci3J3AUPcCd/2tBUnJbTd2mRSVUp3KQSef8OZyaQqHnRY533C2P7IzwKZW5kc3RyZWFtCmVuZG9iagozNSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2NCA+PgpzdHJlYW0KeJxFkMdxBTEMQ++qAiUwgAr1rMfzD+v+r4b000F6GEIMYk/CsFxXcWF0w4+3LTMNf0cZ7sb6MmO81VggJ+gDDJGJq9Gk+nbFGar05NVirqOiXC86IhLMkuOrQCN8OrLHk7a2M/10Xh/sIe8T/yoq525hAS6q7kD5Uh/x1I/ZUeqaoY8qK2seatpXhF0RSts+LqcyTt29A1rhvZWrPdrvPx52OvIKZW5kc3RyZWFtCmVuZG9iagozNiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDcyID4+CnN0cmVhbQp4nDMyt1AwULA0ARKGFiYK5mYGCimGXEC+qYm5Qi4XSAzEygGzDIC0JZyCiGeAmCBtEMUgFkSxmYkZRB2cAZHL4EoDACXbFskKZW5kc3RyZWFtCmVuZG9iagozNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDQ3ID4+CnN0cmVhbQp4nDMyt1AwULA0ARKGFiYK5mYGCimGXJYQVi4XTCwHzALRlnAKIp7BlQYAuWcNJwplbmRzdHJlYW0KZW5kb2JqCjM4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjU4ID4+CnN0cmVhbQp4nEWRS3IEIAhE956CI4D85DyTSmUxuf82Dc5kNnaXqP2ESiOmEiznFHkwfcnyzWS26Xc5VjsbBRRFKJjJVeixAqs7U8SZa4lq62Nl5LjTOwbFG85dOalkcaOMdVR1KnBMz5X1Ud35dlmUfUcOZQrYrHMcbODKbcMYJ0abre4O94kgTydTR8XtINnwByeNfZWrK3CdbPbRSzAOBP1CE5jki0DrDIHGzVP05BLs4+N254Fgb3kRSNkQyJEhGB2Cdp1c/+LW+b3/cYY7z7UZrhzv4neY1nbHX2KSFXMBi9wpqOdrLlrXGTrekzPH5Kb7hs65YJe7g0zv+T/Wz/r+Ax4pZvoKZW5kc3RyZWFtCmVuZG9iagozOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2MyA+PgpzdHJlYW0KeJxFkDsSAyEMQ3tOoSP4IwM+z2YyKTb3b2PYbFLA01ggg7sTgtTagonogoe2Jd0F760EZ2P86TZuNRLkBHWAVqTjaJRSfbnFaZV08Wg2cysLrRMdZg56lKMZoBA6Fd7touRypu7O+UNw9V/1v2LdOZuJgcnKHQjN6lPc+TY7orq6yf6kx9ys134r7FVhaVlLywm3nbtmQAncUznaqz0/Hwo69gplbmRzdHJlYW0KZW5kb2JqCjQwIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzIyID4+CnN0cmVhbQp4nDVRu23FMAzsNQUXMCB+Jc3jIEiRt3+bO9qpSNO8H1VeMqVcLnXJKllh8qVDdYqmfJ5mpvwO9ZDjmB7ZIbpT1pZ7GBaWiXlKHbGaLPdwCza+AJoScwvx9wjwK4BRwESgbvH3D7pZEkAaFPwU6JqrllhiAg2Lha3ZFeJW3SlYuKv4diS5BwlyMVnoUw5Fiim3wHwZLNmRWpzrclkK/259AhphhTjss4tE4HnAA0wk/mSAbM8+W+zq6kU2doY46dCAi4CbzSQBQVM4qz64Yftqu+bnmSgnODnWr6Ixvg1O5ktS3le5x8+gQd74Mzxnd45QDppQCPTdAiCH3cBGhD61z8AuA7ZJu3djSvmcZCm+BDYK9qhTHcrwYuzMVm/Y/MfoymZRbJCV9dHpDsrcoBNiHm9koVuytvs3D7N9/wFfGXtkCmVuZHN0cmVhbQplbmRvYmoKNDEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTggPj4Kc3RyZWFtCnicPVC5jQQxDMtdhRpYwHrtqWcWi0um//RI+fYi0RZFUio1mZIpL3WUJVlT3jp8lsQOeYblbmQ2JSpFL5OwJffQCvF9ieYU993VlrNDNJdoOX4LMyqqGx3TSzaacCoTuqDcwzP6DW10A1aHHrFbINCkYNe2IHLHDxgMwZkTiyIMSk0G/65yj59eixs+w/FDFJGSDuY1/1j98nMNr1OPJ5Fub77iXpypDgMRHJKavCNdWLEuEhFpNUFNz8BaLYC7t17+G7QjugxA9onEcZpSjqG/a3Clzy/lJ1PYCmVuZHN0cmVhbQplbmRvYmoKNDIgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA4MyA+PgpzdHJlYW0KeJxFjLsNwDAIRHumYAR+JvY+UZTC3r8NECVuuCfdPVwdCZkpbjPDQwaeDCyGXXGB9JYwC1xHUI6d7KNh1b7qBI31plLz7w+Unuys4obrAQJCGmYKZW5kc3RyZWFtCmVuZG9iago0MyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDIzOSA+PgpzdHJlYW0KeJxNUMltBDEM+7sKNTDA6By7HgeLPLL9f0PKCZKXaEviofKUW5bKZfcjOW/JuuVDh06VafJu0M2vsf6jDAJ2/1BUEK0lsUrMXNJusTRJL9nDOI2Xa7WO56l7hFmjePDj2NMpgek9MsFms705MKs9zg6QTrjGr+rTO5UkA4m6kPNCpQrrHtQloo8r25hSnU4t5RiXn+h7fI4APcXejdzRx8sXjEa1LajRapU4DzATU9GVcauRgZQTBkNnR1c0C6XIynpCNcKNOaGZvcNwYAPLs4Skpa1SvA9lAegCXdo64zRKgo4Awt8ojPX6Bqr8XjcKZW5kc3RyZWFtCmVuZG9iago0NCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDUxID4+CnN0cmVhbQp4nDM2tFAwUDA0MAeSRoZAlpGJQoohF0gAxMzlggnmgFkGQBqiOAeuJocrgysNAOG0DZgKZW5kc3RyZWFtCmVuZG9iago0NSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE2MCA+PgpzdHJlYW0KeJxFkDkSAzEIBHO9gidIXIL3rMu1wfr/qQfWR6LpAjQcuhZNynoUaD7psUahutBr6CxKkkTBFpIdUKdjiDsoSExIY5JIth6DI5pYs12YmVQqs1LhtGnFwr/ZWtXIRI1wjfyJ6QZU/E/qXJTwTYOvkjH6GFS8O4OMSfheRdxaMe3+RDCxGfYJb0UmBYSJsanZvs9ghsz3Ctc4x/MNTII36wplbmRzdHJlYW0KZW5kb2JqCjQ2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMzM0ID4+CnN0cmVhbQp4nC1SS3LFIAzbcwpdoDP4B+Q86XS6eL3/tpKTRUYOYPQx5YaJSnxZILej1sS3jcxAheGvq8yFz0jbyDqIy5CLuJIthXtELOQxxDzEgu+r8R4e+azMybMHxi/Zdw8r9tSEZSHjxRnaYRXHYRXkWLB1Iap7eFOkw6kk2OOL/z7Fcy0ELXxG0IBf5J+vjuD5khZp95ht0656sEw7qqSwHGxPc14mX1pnuToezwfJ9q7YEVK7AhSFuTPOc+Eo01ZGtBZ2NkhqXGxvjv1YStCFblxGiiOQn6kiPKCkycwmCuKPnB5yKgNh6pqudHIbVXGnnsw1m4u3M0lm675IsZnCeV04s/4MU2a1eSfPcqLUqQjvsWdL0NA5rp69lllodJsTvKSEz8ZOT06+VzPrITkVCaliWlfBaRSZYgnbEl9TUVOaehn++/Lu8Tt+/gEsc3xzCmVuZHN0cmVhbQplbmRvYmoKNDcgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAzMjAgPj4Kc3RyZWFtCnicNVJLbgUxCNvPKbhApfBPzvOqqou++29rE70VTDBg4ykvWdJLvtQl26XD5Fsf9yWxQt6P7ZrMUsX3FrMUzy2vR88Rty0KBFETPViZLxUi1M/06DqocEqfgVcItxQbvINJAINq+AcepTMgUOdAxrtiMlIDgiTYc2lxCIlyJol/pLye3yetpKH0PVmZy9+TS6XQHU1O6AHFysVJoF1J+aCZmEpEkpfrfbFC9IbAkjw+RzHJgOw2iW2iBSbnHqUlzMQUOrDHArxmmtVV6GDCHocpjFcLs6gebPJbE5WkHa3jGdkw3sswU2Kh4bAF1OZiZYLu5eM1r8KI7VGTXcNw7pbNdwjRaP4bFsrgYxWSgEensRINaTjAiMCeXjjFXvMTOQ7AiGOdmiwMY2gmp3qOicDQnrOlYcbHHlr18w9U6XyHCmVuZHN0cmVhbQplbmRvYmoKNDggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxOCA+PgpzdHJlYW0KeJwzNrRQMIDDFEOuNAAd5gNSCmVuZHN0cmVhbQplbmRvYmoKNDkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxMzMgPj4Kc3RyZWFtCnicRY9LDgQhCET3nKKOwMcf53Ey6YVz/+2AnW4TYz2FVIG5gqE9LmsDnRUfIRm28beplo5FWT5UelJWD8ngh6zGyyHcoCzwgkkqhiFQi5gakS1lbreA2zYNsrKVU6WOsIujMI/2tGwVHl+iWyJ1kj+DxCov3OO6Hcil1rveoou+f6QBMQkKZW5kc3RyZWFtCmVuZG9iago1MCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM0MCA+PgpzdHJlYW0KeJw1UjluBDEM6/0KfSCAbtvv2SBIkfy/DanZFANxdFKUO1pUdsuHhVS17HT5tJXaEjfkd2WFxAnJqxLtUoZIqLxWIdXvmTKvtzVnBMhSpcLkpORxyYI/w6WnC8f5trGv5cgdjx5YFSOhRMAyxcToGpbO7rBmW36WacCPeIScK9Ytx1gFUhvdOO2K96F5LbIGiL2ZlooKHVaJFn5B8aBHjX32GFRYINHtHElwjIlQkYB2gdpIDDl7LHZRH/QzKDET6NobRdxBgSWSmDnFunT03/jQsaD+2Iw3vzoq6VtaWWPSPhvtlMYsMul6WPR089bHgws076L859UMEjRljZLGB63aOYaimVFWeLdDkw3NMcch8w6ewxkJSvo8FL+PJRMdlMjfDg2hf18eo4ycNt4C5qI/bRUHDuKzw165gRVKF2uS9wGpTOiB6f+v8bW+19cfHe2AxgplbmRzdHJlYW0KZW5kb2JqCjUxIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjUxID4+CnN0cmVhbQp4nC1RSXIDQQi7zyv0hGan32OXK4fk/9cIygcGDYtAdFrioIyfICxXvOWRq2jD3zMxgt8Fh34r121Y5EBUIEljUDWhdvF69B7YcZgJzJPWsAxmrA/8jCnc6MXhMRlnt9dl1BDsXa89mUHJrFzEJRMXTNVhI2cOP5kyLrRzPTcg50ZYl2GQblYaMxKONIVIIYWqm6TOBEESjK5GjTZyFPulL490hlWNqDHscy1tX89NOGvQ7Fis8uSUHl1xLicXL6wc9PU2AxdRaazyQEjA/W4P9XOyk994S+fOFtPje83J8sJUYMWb125ANtXi37yI4/uMr+fn+fwDX2BbiAplbmRzdHJlYW0KZW5kb2JqCjUyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTc0ID4+CnN0cmVhbQp4nE2QSQ5DIQxD95zCF6iEM8DnPL+qumjvv61DB3WB/OQgcDw80HEkLnRk6IyOK5sc48CzIGPi0Tj/ybg+xDFB3aItWJd2x9nMEnPCMjECtkbJ2TyiwA/HXAgSZJcfvsAgIl2P+VbzWZP0z7c73Y+6tGZfPaLAiewIxbABV4D9useBS8L5XtPklyolYxOH8oHqIlI2O6EQtVTscqqKs92bK3AV9PzRQ+7tBbUjPN8KZW5kc3RyZWFtCmVuZG9iago1MyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDU0ID4+CnN0cmVhbQp4nDM1MFAwUNC1VNA1MjZVMDUEsg3NTBVSDLng7FwIEySfwwVTCWGBpHMQKnO4MrjSAHNRD48KZW5kc3RyZWFtCmVuZG9iago1NCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE0MSA+PgpzdHJlYW0KeJw9j8EOwzAIQ+/5Cv9ApNgpoXxPp2qH7v+vI0u7C3oCY4yF0NAbqprDhmCb48XSJVRr+BTFQCU3yJlgDqWk0h1HkXpiOBhcHrQbjuKx6PoRu5JmfdDGQrolaIB7rFNp3KZxE8QdNQXqKeqco7wQuZ+pZ9g0kt00s5JzuA2/e89T1/+nq7zL+QW9dy7+CmVuZHN0cmVhbQplbmRvYmoKNTUgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTUgPj4Kc3RyZWFtCnicNVE5DgMhDOz3Ff5AJIwveE+iKM3+v82M0VYewVyGtJQhmfJSk6gh5VM+epkunLrc18xqNOeWtC1zgLi2vC+tksCJZoiDwWmYuAGaPAFD19GoUUMXHtDUpVMosNwEPoq3bg/dY7WBl7Yh54kgYigZLEHNqUUTFm3PJ6Q1v16LG96X7d3IU6XGlhiBBgFWOBzX6NfwlT1PJtF0FTLUqzXLGAkTRSI8+Y6m1RPrWjTSMhLUxhGsagO8O/0wTgAAE3HLAmSfSpSz5MRvsfSzBlf6/gGfR1SWCmVuZHN0cmVhbQplbmRvYmoKMTUgMCBvYmoKPDwgL0Jhc2VGb250IC9EZWphVnVTYW5zIC9DaGFyUHJvY3MgMTYgMCBSCi9FbmNvZGluZyA8PAovRGlmZmVyZW5jZXMgWyAzMiAvc3BhY2UgNDYgL3BlcmlvZCA0OCAvemVybyAvb25lIC90d28gL3RocmVlIC9mb3VyIC9maXZlIC9zaXggNTYKL2VpZ2h0IC9uaW5lIDY4IC9EIC9FIC9GIC9HIDczIC9JIDc3IC9NIC9OIDgzIC9TIC9UIDg2IC9WIDk1IC91bmRlcnNjb3JlIDk3Ci9hIDk5IC9jIC9kIC9lIC9mIDEwNCAvaCAvaSAxMDggL2wgL20gL24gL28gL3AgMTE0IC9yIC9zIC90IC91IDEyMSAveSBdCi9UeXBlIC9FbmNvZGluZyA+PgovRmlyc3RDaGFyIDAgL0ZvbnRCQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvRm9udERlc2NyaXB0b3IgMTQgMCBSCi9Gb250TWF0cml4IFsgMC4wMDEgMCAwIDAuMDAxIDAgMCBdIC9MYXN0Q2hhciAyNTUgL05hbWUgL0RlamFWdVNhbnMKL1N1YnR5cGUgL1R5cGUzIC9UeXBlIC9Gb250IC9XaWR0aHMgMTMgMCBSID4+CmVuZG9iagoxNCAwIG9iago8PCAvQXNjZW50IDkyOSAvQ2FwSGVpZ2h0IDAgL0Rlc2NlbnQgLTIzNiAvRmxhZ3MgMzIKL0ZvbnRCQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvRm9udE5hbWUgL0RlamFWdVNhbnMgL0l0YWxpY0FuZ2xlIDAKL01heFdpZHRoIDEzNDIgL1N0ZW1WIDAgL1R5cGUgL0ZvbnREZXNjcmlwdG9yIC9YSGVpZ2h0IDAgPj4KZW5kb2JqCjEzIDAgb2JqClsgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAKNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCAzMTggNDAxIDQ2MCA4MzggNjM2Cjk1MCA3ODAgMjc1IDM5MCAzOTAgNTAwIDgzOCAzMTggMzYxIDMxOCAzMzcgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNgo2MzYgNjM2IDMzNyAzMzcgODM4IDgzOCA4MzggNTMxIDEwMDAgNjg0IDY4NiA2OTggNzcwIDYzMiA1NzUgNzc1IDc1MiAyOTUKMjk1IDY1NiA1NTcgODYzIDc0OCA3ODcgNjAzIDc4NyA2OTUgNjM1IDYxMSA3MzIgNjg0IDk4OSA2ODUgNjExIDY4NSAzOTAgMzM3CjM5MCA4MzggNTAwIDUwMCA2MTMgNjM1IDU1MCA2MzUgNjE1IDM1MiA2MzUgNjM0IDI3OCAyNzggNTc5IDI3OCA5NzQgNjM0IDYxMgo2MzUgNjM1IDQxMSA1MjEgMzkyIDYzNCA1OTIgODE4IDU5MiA1OTIgNTI1IDYzNiAzMzcgNjM2IDgzOCA2MDAgNjM2IDYwMCAzMTgKMzUyIDUxOCAxMDAwIDUwMCA1MDAgNTAwIDEzNDIgNjM1IDQwMCAxMDcwIDYwMCA2ODUgNjAwIDYwMCAzMTggMzE4IDUxOCA1MTgKNTkwIDUwMCAxMDAwIDUwMCAxMDAwIDUyMSA0MDAgMTAyMyA2MDAgNTI1IDYxMSAzMTggNDAxIDYzNiA2MzYgNjM2IDYzNiAzMzcKNTAwIDUwMCAxMDAwIDQ3MSA2MTIgODM4IDM2MSAxMDAwIDUwMCA1MDAgODM4IDQwMSA0MDEgNTAwIDYzNiA2MzYgMzE4IDUwMAo0MDEgNDcxIDYxMiA5NjkgOTY5IDk2OSA1MzEgNjg0IDY4NCA2ODQgNjg0IDY4NCA2ODQgOTc0IDY5OCA2MzIgNjMyIDYzMiA2MzIKMjk1IDI5NSAyOTUgMjk1IDc3NSA3NDggNzg3IDc4NyA3ODcgNzg3IDc4NyA4MzggNzg3IDczMiA3MzIgNzMyIDczMiA2MTEgNjA1CjYzMCA2MTMgNjEzIDYxMyA2MTMgNjEzIDYxMyA5ODIgNTUwIDYxNSA2MTUgNjE1IDYxNSAyNzggMjc4IDI3OCAyNzggNjEyIDYzNAo2MTIgNjEyIDYxMiA2MTIgNjEyIDgzOCA2MTIgNjM0IDYzNCA2MzQgNjM0IDU5MiA2MzUgNTkyIF0KZW5kb2JqCjE2IDAgb2JqCjw8IC9EIDE3IDAgUiAvRSAxOCAwIFIgL0YgMTkgMCBSIC9HIDIwIDAgUiAvSSAyMSAwIFIgL00gMjIgMCBSIC9OIDIzIDAgUgovUyAyNCAwIFIgL1QgMjUgMCBSIC9WIDI2IDAgUiAvYSAyNyAwIFIgL2MgMjggMCBSIC9kIDI5IDAgUiAvZSAzMCAwIFIKL2VpZ2h0IDMxIDAgUiAvZiAzMiAwIFIgL2ZpdmUgMzMgMCBSIC9mb3VyIDM0IDAgUiAvaCAzNSAwIFIgL2kgMzYgMCBSCi9sIDM3IDAgUiAvbSAzOCAwIFIgL24gMzkgMCBSIC9uaW5lIDQwIDAgUiAvbyA0MSAwIFIgL29uZSA0MiAwIFIgL3AgNDMgMCBSCi9wZXJpb2QgNDQgMCBSIC9yIDQ1IDAgUiAvcyA0NiAwIFIgL3NpeCA0NyAwIFIgL3NwYWNlIDQ4IDAgUiAvdCA0OSAwIFIKL3RocmVlIDUwIDAgUiAvdHdvIDUxIDAgUiAvdSA1MiAwIFIgL3VuZGVyc2NvcmUgNTMgMCBSIC95IDU0IDAgUgovemVybyA1NSAwIFIgPj4KZW5kb2JqCjMgMCBvYmoKPDwgL0YxIDE1IDAgUiA+PgplbmRvYmoKNCAwIG9iago8PCAvQTEgPDwgL0NBIDAgL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMSA+PgovQTIgPDwgL0NBIDEgL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMSA+PgovQTMgPDwgL0NBIDAuOCAvVHlwZSAvRXh0R1N0YXRlIC9jYSAwLjggPj4gPj4KZW5kb2JqCjUgMCBvYmoKPDwgPj4KZW5kb2JqCjYgMCBvYmoKPDwgPj4KZW5kb2JqCjcgMCBvYmoKPDwgPj4KZW5kb2JqCjIgMCBvYmoKPDwgL0NvdW50IDEgL0tpZHMgWyAxMSAwIFIgXSAvVHlwZSAvUGFnZXMgPj4KZW5kb2JqCjU2IDAgb2JqCjw8IC9DcmVhdGlvbkRhdGUgKEQ6MjAyMTA5MTYxNDM3MzQrMDInMDAnKQovQ3JlYXRvciAoTWF0cGxvdGxpYiB2My40LjMsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAoTWF0cGxvdGxpYiBwZGYgYmFja2VuZCB2My40LjMpID4+CmVuZG9iagp4cmVmCjAgNTcKMDAwMDAwMDAwMCA2NTUzNSBmIAowMDAwMDAwMDE2IDAwMDAwIG4gCjAwMDAwMTQ2NjggMDAwMDAgbiAKMDAwMDAxNDQzMSAwMDAwMCBuIAowMDAwMDE0NDYzIDAwMDAwIG4gCjAwMDAwMTQ2MDUgMDAwMDAgbiAKMDAwMDAxNDYyNiAwMDAwMCBuIAowMDAwMDE0NjQ3IDAwMDAwIG4gCjAwMDAwMDAwNjUgMDAwMDAgbiAKMDAwMDAwMDM5OSAwMDAwMCBuIAowMDAwMDAyNDQ4IDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAwMjQyNyAwMDAwMCBuIAowMDAwMDEyOTIyIDAwMDAwIG4gCjAwMDAwMTI3MjIgMDAwMDAgbiAKMDAwMDAxMjIxNiAwMDAwMCBuIAowMDAwMDEzOTc1IDAwMDAwIG4gCjAwMDAwMDI0NjggMDAwMDAgbiAKMDAwMDAwMjcwNSAwMDAwMCBuIAowMDAwMDAyODU4IDAwMDAwIG4gCjAwMDAwMDMwMDYgMDAwMDAgbiAKMDAwMDAwMzMyNiAwMDAwMCBuIAowMDAwMDAzNDQ5IDAwMDAwIG4gCjAwMDAwMDM2MTEgMDAwMDAgbiAKMDAwMDAwMzc2MCAwMDAwMCBuIAowMDAwMDA0MTc0IDAwMDAwIG4gCjAwMDAwMDQzMTIgMDAwMDAgbiAKMDAwMDAwNDQ1NiAwMDAwMCBuIAowMDAwMDA0ODM2IDAwMDAwIG4gCjAwMDAwMDUxNDEgMDAwMDAgbiAKMDAwMDAwNTQ0NSAwMDAwMCBuIAowMDAwMDA1NzY3IDAwMDAwIG4gCjAwMDAwMDYyMzUgMDAwMDAgbiAKMDAwMDAwNjQ0NCAwMDAwMCBuIAowMDAwMDA2NzY2IDAwMDAwIG4gCjAwMDAwMDY5MzIgMDAwMDAgbiAKMDAwMDAwNzE2OSAwMDAwMCBuIAowMDAwMDA3MzEzIDAwMDAwIG4gCjAwMDAwMDc0MzIgMDAwMDAgbiAKMDAwMDAwNzc2MyAwMDAwMCBuIAowMDAwMDA3OTk5IDAwMDAwIG4gCjAwMDAwMDgzOTQgMDAwMDAgbiAKMDAwMDAwODY4NSAwMDAwMCBuIAowMDAwMDA4ODQwIDAwMDAwIG4gCjAwMDAwMDkxNTIgMDAwMDAgbiAKMDAwMDAwOTI3NSAwMDAwMCBuIAowMDAwMDA5NTA4IDAwMDAwIG4gCjAwMDAwMDk5MTUgMDAwMDAgbiAKMDAwMDAxMDMwOCAwMDAwMCBuIAowMDAwMDEwMzk4IDAwMDAwIG4gCjAwMDAwMTA2MDQgMDAwMDAgbiAKMDAwMDAxMTAxNyAwMDAwMCBuIAowMDAwMDExMzQxIDAwMDAwIG4gCjAwMDAwMTE1ODggMDAwMDAgbiAKMDAwMDAxMTcxNCAwMDAwMCBuIAowMDAwMDExOTI4IDAwMDAwIG4gCjAwMDAwMTQ3MjggMDAwMDAgbiAKdHJhaWxlcgo8PCAvSW5mbyA1NiAwIFIgL1Jvb3QgMSAwIFIgL1NpemUgNTcgPj4Kc3RhcnR4cmVmCjE0ODg1CiUlRU9GCg==\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["============= Test accuracy: 89.09% ==============\n", "\n"]}], "source": ["SGD_model = copy.deepcopy(base_model).to(device)\n", "SGD_results = train_model(\n", " SGD_model, \"FashionMNIST_SGD\", lambda params: SGD(params, lr=1e-1), max_epochs=40, batch_size=256\n", ")"]}, {"cell_type": "code", "execution_count": 27, "id": "448f5f88", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:34.592457Z", "iopub.status.busy": "2021-09-16T12:37:34.591978Z", "iopub.status.idle": "2021-09-16T12:37:34.958280Z", "shell.execute_reply": "2021-09-16T12:37:34.958659Z"}, "papermill": {"duration": 0.474961, "end_time": "2021-09-16T12:37:34.958809", "exception": false, "start_time": "2021-09-16T12:37:34.483848", "status": "completed"}, "tags": []}, "outputs": [{"name": "stdout", "output_type": "stream", "text": ["Model file of \"FashionMNIST_SGDMom\" already exists. Skipping training...\n"]}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["============= Test accuracy: 88.83% ==============\n", "\n"]}], "source": ["SGDMom_model = copy.deepcopy(base_model).to(device)\n", "SGDMom_results = train_model(\n", " SGDMom_model,\n", " \"FashionMNIST_SGDMom\",\n", " lambda params: SGDMomentum(params, lr=1e-1, momentum=0.9),\n", " max_epochs=40,\n", " batch_size=256,\n", ")"]}, {"cell_type": "code", "execution_count": 28, "id": "097c35c5", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:35.182816Z", "iopub.status.busy": "2021-09-16T12:37:35.182331Z", "iopub.status.idle": "2021-09-16T12:37:35.546702Z", "shell.execute_reply": "2021-09-16T12:37:35.546288Z"}, "papermill": {"duration": 0.477353, "end_time": "2021-09-16T12:37:35.546818", "exception": false, "start_time": "2021-09-16T12:37:35.069465", "status": "completed"}, "tags": []}, "outputs": [{"name": "stdout", "output_type": "stream", "text": ["Model file of \"FashionMNIST_Adam\" already exists. Skipping training...\n"]}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "\n"], "text/plain": [""]}, "metadata": {}, "output_type": "display_data"}, {"name": "stdout", "output_type": "stream", "text": ["============= Test accuracy: 89.46% ==============\n", "\n"]}], "source": ["Adam_model = copy.deepcopy(base_model).to(device)\n", "Adam_results = train_model(\n", " Adam_model, \"FashionMNIST_Adam\", lambda params: Adam(params, lr=1e-3), max_epochs=40, batch_size=256\n", ")"]}, {"cell_type": "markdown", "id": "3372cdee", "metadata": {"papermill": {"duration": 0.107962, "end_time": "2021-09-16T12:37:35.766302", "exception": false, "start_time": "2021-09-16T12:37:35.658340", "status": "completed"}, "tags": []}, "source": ["The result is that all optimizers perform similarly well with the given model.\n", "The differences are too small to find any significant conclusion.\n", "However, keep in mind that this can also be attributed to the initialization we chose.\n", "When changing the initialization to worse (e.g. constant initialization), Adam usually shows to be more robust because of its adaptive learning rate.\n", "To show the specific benefits of the optimizers, we will continue to\n", "look at some possible loss surfaces in which momentum and adaptive\n", "learning rate are crucial."]}, {"cell_type": "markdown", "id": "6afe619e", "metadata": {"lines_to_next_cell": 2, "papermill": {"duration": 0.10673, "end_time": "2021-09-16T12:37:35.980280", "exception": false, "start_time": "2021-09-16T12:37:35.873550", "status": "completed"}, "tags": []}, "source": ["### Pathological curvatures\n", "\n", "A pathological curvature is a type of surface that is similar to ravines and is particularly tricky for plain SGD optimization.\n", "In words, pathological curvatures typically have a steep gradient in one direction with an optimum at the center, while in a second direction we have a slower gradient towards a (global) optimum.\n", "Let's first create an example surface of this and visualize it:"]}, {"cell_type": "code", "execution_count": 29, "id": "5cfd764f", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:36.202397Z", "iopub.status.busy": "2021-09-16T12:37:36.201928Z", "iopub.status.idle": "2021-09-16T12:37:36.203949Z", "shell.execute_reply": "2021-09-16T12:37:36.203512Z"}, "papermill": {"duration": 0.113791, "end_time": "2021-09-16T12:37:36.204058", "exception": false, "start_time": "2021-09-16T12:37:36.090267", "status": "completed"}, "tags": []}, "outputs": [], "source": ["def pathological_curve_loss(w1, w2):\n", " # Example of a pathological curvature. There are many more possible, feel free to experiment here!\n", " x1_loss = torch.tanh(w1) ** 2 + 0.01 * torch.abs(w1)\n", " x2_loss = torch.sigmoid(w2)\n", " return x1_loss + x2_loss"]}, {"cell_type": "code", "execution_count": 30, "id": "c8827bdc", "metadata": {"execution": {"iopub.execute_input": "2021-09-16T12:37:36.439443Z", "iopub.status.busy": "2021-09-16T12:37:36.436044Z", "iopub.status.idle": "2021-09-16T12:37:37.788861Z", "shell.execute_reply": "2021-09-16T12:37:37.789291Z"}, "papermill": {"duration": 1.475969, "end_time": "2021-09-16T12:37:37.789440", "exception": false, "start_time": "2021-09-16T12:37:36.313471", "status": "completed"}, "tags": []}, "outputs": [{"name": "stderr", "output_type": "stream", "text": ["/tmp/ipykernel_879/1102210584.py:5: MatplotlibDeprecationWarning: Calling gca() with keyword arguments was deprecated in Matplotlib 3.4. Starting two minor releases later, gca() will take no keyword arguments. The gca() function should only be used to get the current axes, or if no axes exist, create new axes with default keyword arguments. To create a new axes with non-default arguments, use plt.axes() or plt.subplot().\n", " ax = fig.gca(projection=\"3d\") if plot_3d else fig.gca()\n"]}, {"data": {"application/pdf": "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\n", "image/svg+xml": ["\n", "\n", "