Skip to content

GitLab

Commit e3f7f7b3 authored by chenzk's avatar chenzk
Browse files

v1.0

parents
Pipeline #956 failed with stages
in 0 seconds
Hide whitespace changes
Inline Side-by-side
Showing with 2539 additions and 0 deletions
nbs/examples/MLflow_and_neuralforecast.ipynb 0 → 100644
View file @ e3f7f7b3
{
"cells": [
{
"cell_type": "markdown",
"id": "0065a29e-b528-40cb-9b14-5eeb94c98e84",
"metadata": {},
"source": [
"# MLflow and neuralforecast\n",
"> Log your neuralforecast experiments to MLflow"
]
},
{
"cell_type": "markdown",
"id": "e4fb1958",
"metadata": {},
"source": [
"## Installing dependencies"
]
},
{
"cell_type": "markdown",
"id": "963311a3-2427-4694-981b-438fce1c2981",
"metadata": {},
"source": [
"To install Neuralforecast refer to https://nixtlaverse.nixtla.io/neuralforecast/examples/installation.html.\n",
"\n",
"To install mlflow: `pip install mlflow`"
]
},
{
"cell_type": "markdown",
"id": "1125a0bc",
"metadata": {},
"source": [
"## Imports"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "687f6677",
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import warnings\n",
"\n",
"import matplotlib.pyplot as plt\n",
"import mlflow\n",
"import mlflow.data\n",
"import numpy as np\n",
"import pandas as pd\n",
"from mlflow.client import MlflowClient\n",
"from mlflow.data.pandas_dataset import PandasDataset\n",
"from utilsforecast.plotting import plot_series\n",
"\n",
"from neuralforecast.core import NeuralForecast\n",
"from neuralforecast.models import NBEATSx\n",
"from neuralforecast.utils import AirPassengersDF\n",
"from neuralforecast.losses.pytorch import MAE"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "dbafb920-4af8-49d0-abfa-2ff79f2cfd00",
"metadata": {},
"outputs": [],
"source": [
"os.environ['NIXTLA_ID_AS_COL'] = '1'\n",
"logging.getLogger(\"mlflow\").setLevel(logging.ERROR)\n",
"logging.getLogger(\"pytorch_lightning\").setLevel(logging.ERROR)\n",
"warnings.filterwarnings(\"ignore\")"
]
},
{
"cell_type": "markdown",
"id": "2570db3c-d8f7-4b07-bd43-d79d730b07cb",
"metadata": {},
"source": [
"## Splitting the data"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "0d556dc4-8965-4dfd-809a-7b87d348d5ea",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>unique_id</th>\n",
" <th>ds</th>\n",
" <th>y</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>139</th>\n",
" <td>1.0</td>\n",
" <td>1960-08-31</td>\n",
" <td>606.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>140</th>\n",
" <td>1.0</td>\n",
" <td>1960-09-30</td>\n",
" <td>508.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>141</th>\n",
" <td>1.0</td>\n",
" <td>1960-10-31</td>\n",
" <td>461.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>142</th>\n",
" <td>1.0</td>\n",
" <td>1960-11-30</td>\n",
" <td>390.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>143</th>\n",
" <td>1.0</td>\n",
" <td>1960-12-31</td>\n",
" <td>432.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" unique_id ds y\n",
"139 1.0 1960-08-31 606.0\n",
"140 1.0 1960-09-30 508.0\n",
"141 1.0 1960-10-31 461.0\n",
"142 1.0 1960-11-30 390.0\n",
"143 1.0 1960-12-31 432.0"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Split data and declare panel dataset\n",
"Y_df = AirPassengersDF\n",
"Y_train_df = Y_df[Y_df.ds<='1959-12-31'] # 132 train\n",
"Y_test_df = Y_df[Y_df.ds>'1959-12-31'] # 12 test\n",
"Y_df.tail()"
]
},
{
"cell_type": "markdown",
"id": "919711ee-6fe1-442b-a11f-6031cd4b4999",
"metadata": {},
"source": [
"## MLflow UI\n",
"Run the following command from the terminal to start the UI: `mlflow ui`. You can then go to the printed URL to visualize the experiments."
]
},
{
"cell_type": "markdown",
"id": "00768993-bcf7-43d5-9ac3-2f916e52acc6",
"metadata": {},
"source": [
"## Model training"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "a4467763",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Seed set to 42\n"
]
}
],
"source": [
"mlflow.pytorch.autolog(checkpoint=False)\n",
"\n",
"with mlflow.start_run() as run:\n",
" # Log the dataset to the MLflow Run. Specify the \"training\" context to indicate that the\n",
" # dataset is used for model training\n",
" dataset: PandasDataset = mlflow.data.from_pandas(Y_df, source=\"AirPassengersDF\")\n",
" mlflow.log_input(dataset, context=\"training\")\n",
"\n",
" # Define and log parameters\n",
" horizon = len(Y_test_df)\n",
" model_params = dict(\n",
" input_size=1 * horizon,\n",
" h=horizon,\n",
" max_steps=300, \n",
" loss=MAE(),\n",
" valid_loss=MAE(), \n",
" activation='ReLU',\n",
" scaler_type='robust',\n",
" random_seed=42,\n",
" enable_progress_bar=False,\n",
" )\n",
" mlflow.log_params(model_params)\n",
"\n",
" # Fit NBEATSx model\n",
" models = [NBEATSx(**model_params)]\n",
" nf = NeuralForecast(models=models, freq='M') \n",
" train = nf.fit(df=Y_train_df, val_size=horizon)\n",
" \n",
" # Save conda environment used to run the model\n",
" mlflow.pytorch.get_default_conda_env()\n",
" \n",
" # Save pip requirements\n",
" mlflow.pytorch.get_default_pip_requirements()\n",
"\n",
"mlflow.pytorch.autolog(disable=True)\n",
"\n",
"# Save the neural forecast model\n",
"nf.save(path='./checkpoints/test_run_1/',\n",
" model_index=None, \n",
" overwrite=True,\n",
" save_dataset=True)"
]
},
{
"cell_type": "markdown",
"id": "0f72c67d-8d93-4a93-aa15-632c1ca3ae00",
"metadata": {},
"source": [
"## Forecasting the future"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "62bec1b2-c07f-4ffd-8e2f-d2581ec3fcf5",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAACHgAAAFhCAYAAAABCY9wAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAADhOUlEQVR4nOzdd3hUddrG8e9kMpl0QmhJIPTeexdQmgLSpMMCVmwoYtfdFV0F9VVRQUEUAQXEgkhHepNepdfQE3p6n5n3j0h2Iy0JM3NCcn+uK5fm5Jzf756UQ2CeeR6Tw+FwICIiIiIiIiIiIiIiIiIiIiJ5lofRAURERERERERERERERERERETk1lTgISIiIiIiIiIiIiIiIiIiIpLHqcBDREREREREREREREREREREJI9TgYeIiIiIiIiIiIiIiIiIiIhIHqcCDxEREREREREREREREREREZE8TgUeIiIiIiIiIiIiIiIiIiIiInmcCjxERERERERERERERERERERE8jgVeIiIiIiIiIiIiIiIiIiIiIjkcSrwEBEREREREREREREREREREcnjVOAhIiIiIiIi+dbUqVMxmUycOHHC6CjZZjKZGDVq1G3Pc+ZjW7BgAYMHD6ZWrVpYLBZMJlOO15g1axZ169bF29ubsLAwRowYQXx8/B1nExERERERERGRDCrwEBERERERkXyrc+fObNy4kdDQUKOjZNvGjRt57LHH3LrnnDlz2LRpE9WrV6dOnTo5vn7GjBn079+fRo0asXjxYt566y2mTp1Kz549XZBWRERERERERKRgMjkcDofRIUREREREREQkZ6ZOncrDDz9MREQEZcuWvaO17HY7Hh4ZrwF59tln+eKLL8juPxfYbDbCw8OpVasWv//+e+bxmTNnMnDgQBYtWsQDDzxwR/lEREREREREREQdPERERERERCQPGTp06A2LFUaNGpVlbIjJZOLZZ5/l+++/p1q1avj6+lKnTh0WLFiQ5bobjTFxOBx8+OGHlClTBm9vb+rXr8/ixYtp06YNbdq0ueW1AKtXr8ZkMrF69eosx5cvX07btm0JDAzE19eXFi1asGLFihx/Dm40omXTpk20aNEic/zJ66+/TlpaWo7XvplrxR25sWnTJiIjI3n44YezHO/duzf+/v7MmTPnTuOJiIiIiIiIiAgq8BAREREREZG71MKFCxk/fjzvvPMOs2fPJjg4mB49enD8+PFbXvf222/z6quv0r59e3777TeeeuopHn/8cQ4dOpTrLNOnT6dDhw4EBgYybdo0fvrpJ4KDg+nYsWOuijz+1/79+2nbti3R0dFMnTqViRMnsnPnTt59993rzrXb7aSnp9/2zWaz3VGm/7V3714AateuneW4xWKhatWqmR8XEREREREREZE742l0ABEREREREZHcSEpKYvny5QQEBABQv359wsLC+Omnn3jttddueE10dDQffPABPXr04Jtvvsk8XqNGDVq0aEGVKlVynCMxMZHnn3+eLl26ZOlW0alTJ+rXr88bb7zB5s2bc7zuNe+88w4Oh4OVK1dSokQJADp37kzNmjWvO/eRRx5h2rRpt12zdevW13Ugya3Lly8DEBwcfN3HgoODr+uAIiIiIiIiIiIiuaMCDxEREREREbkr3XvvvZnFHQAlSpSgePHinDx58qbXbNy4keTkZAYOHJjlePPmzSlTpkyucmzYsIErV64wZMgQ0tPTs3zs/vvv58MPPyQhIQE/P79crb9q1Sratm2bWdwBYDab6du3L2+//XaWc0eNGsWzzz572zX/9/PmLP87Qic7x0VEREREREREJGdU4CEiIiIiIiJ3pSJFilx3zGq1kpSUdNNrrnWbCAkJue5jNzqWHefPnwegV69eNz3nypUruS7wuHz5crbzli5dmlKlSt12TWcWXVz7Oly+fDlLEQpkPO4bdfYQEREREREREZGcU4GHiIiIiIiI5Bne3t6kpKRcd/zSpUtOWf9aMUJUVNR1H4uKiqJs2bJZsgDX5fl7lqJFiwIwbtw4mjZtesN9/174kNPMN8v7d0aMaKlVqxYAe/bsoXr16pnH09PTOXjwIP3793fKPiIiIiIiIiIiBZ0KPERERERERCTPKFu2LBcuXOD8+fOZRRGpqan8/vvvTlm/adOmeHt7M2PGDB566KHM4xs2bODkyZNZCjyu/f+ff/5JlSpVMo/Pmzcvy5otWrQgKCiI/fv3Z2s8Sk7de++9zJs3L8vnxGaz8eOPP153rhEjWpo0aUJoaChTp06lb9++mcd/+eUX4uPj6dmzp9P2EhEREREREREpyFTgISIiIiIiInlG3759+fe//02/fv14+eWXSU5O5vPPP8dmszll/cKFC/PSSy/x7rvv8thjj9G7d29Onz7NqFGjrht50qhRI6pUqcJLL71Eeno6hQsXZs6cOaxfvz7Lef7+/owbN44hQ4Zw5coVevXqRfHixbl48SK7d+/m4sWLTJgwIdeZ//nPfzJv3jzuu+8+/v3vf+Pr68sXX3xBQkLCdeeWLVs2S5FKdp08eZKtW7cCcOzYMSCjQOPamg0bNsw8r0KFCgwZMoTJkycDYDab+fDDD/nHP/7BsGHD6N+/P0eOHOGVV16hffv23H///bl52CIiIiIiIiIi8jceRgcQERERERERuaZcuXLMnTuX6OhoevXqxcsvv0zv3r0ZPHiw0/Z45513GDNmDEuXLqVr166MGzeOiRMnZunSARmFC/Pnz6dq1ao8+eSTDB48GKvVyvjx469bc9CgQaxatYr4+HiGDRtGu3bteP7559mxYwdt27a9o7w1a9Zk+fLlBAYGMmTIEJ544glq167Nv/71rzta93+tWrWK3r1707t3b5YsWQKQ+f7/Pl6Hw4HNZruu4GbQoEHMnDmTTZs20bFjR/79738zePBgfv31V6dlFBEREREREREp6EwOh8NhdAgRERERERERo7Vp0waA1atXG5pDRERERERERETkRtTBQ0RERERERERERERERERERCSP8zQ6gIiIiIiIiEhBkJ6efsuPe3h44OGh12GIiIiIiIiIiMiNaUSLiIiIiIiIiIudOHGCcuXK3fKct956i1GjRrknkIiIiIiIiIiI3HXUwUNERERERETExcLCwti6dettzxEREREREREREbkZdfAQERERERERERERERERERERyeM03FdEREREREREREREREREREQkj9OIFsBut3Pu3DkCAgIwmUxGxxERERERERERERERERERcRqHw0FcXBxhYWF4eKgHgMjdSgUewLlz5wgPDzc6hoiIiIiIiIiIiIiIiIiIy5w+fZpSpUoZHUNEckkFHkBAQACQcUMLDAw0OI1I/pGWlsbSpUvp0KEDFovF6Dgikk/pXiMi7qL7jYi4i+43IuIuut+IiLvofiNivNjYWMLDwzOfFxWRu5MKPCBzLEtgYKAKPEScKC0tDV9fXwIDA/VLu4i4jO41IuIuut+IiLvofiMi7qL7jYi4i+43InnHtedFReTupAFLIiIiIiIiIiIiIiIiIiIiInmcCjxERERERERERERERERERERE8jgVeIiIiIiIiIiIiIiIiIiIiIjkcSrwEBEREREREREREREREREREcnjVOAhIiIiIiIiIiIiIiIiIiIiksepwENEREREREREREREREREREQkj1OBh4iIiIiIiIiIiIiIiIiIiEgepwIPERERERERERERERERERERkTzO0+gAIiIiIiIiIiIiIiIiIpI7DpsDW3I6tmRbxn+TbNj/+q+1hA9+pQOMjigiIk6iAg8RERERERERERERERGRPC7+WAzJFxKxJWUUctiTbdiS0rGn2m9+0X4oM6AyHhaz+4KKiIjLqMBDREREREREREREREREJA9LuZTExbXnbn6CCczeZszennj4eGL2NpN0NgF7io3kC0n4lvR3X1gREXEZFXiIiIiIiIiIiIiIiIiI5GHxR2MAsBbzwb98YEYRh09GQYfZ24yH1YzJZMpyzcV154g/GkNyVKIKPERE8gkVeIiIiIiIiIiIiIiIiIjkUQ6bnfjjsQAE1Sua7WIN7xDfzAIPERHJHzyMDiAiIiIiIiIiIiIiIiIiN5Z4Oh57ig2zryc+oX7Zvs47xBfIGO9iT7e7Kp6IiLiRCjxERERERERERERERERE8qhr41n8KxTC5GG6zdn/5elvwezrCXZIuZDkqngiIuJGKvAQERERERERERERERERyYPSE9NJPBMPgH/FQjm61mQyZXbxSD6vMS0iIvmBCjxERERERERERERERERE8qCE4zHgAGsxH7yCrDm+3rvEXwUeUSrwEBHJD1TgISIiIiIiIiIiIiIiIpLHOBwO4o78NZ4lh907rrnWwSPlYhIOm91p2URExBgq8BARERERERERERERERHJY1IvJ5MWnYLJbMKvXGCu1rAU8sLD24zD5iDlUrKTE4qIiLupwENEREREREREREREREQkj4k/mtG9w7d0AGarOVdrmEymzC4eGtMiInL3U4GHiIiIiIiIiIiIiIiISB7isNmJPx4LgH+l3I1nuca7hAo8RETyCxV4iIiIiIiIiIiIiIiIiOQhiafjsafYMPt64hPqd0dr+Vzr4HEhEYfd4Yx4IiJiEMMLPM6ePcugQYMoUqQIvr6+1K1bl+3bt2d+3OFwMGrUKMLCwvDx8aFNmzbs27cvyxopKSkMHz6cokWL4ufnR9euXTlz5oy7H4qIiIiIiIiIiIiIiIjIHYv7azyLf4VCmDxMd7SWpbAVDy8PHOkOUi8nOyOeiIgYxNACj6tXr9KiRQssFguLFy9m//79fPzxxwQFBWWe8+GHH/LJJ58wfvx4tm7dSkhICO3btycuLi7znBEjRjBnzhxmzZrF+vXriY+Pp0uXLthsNgMelYiIiIiIiIiIiIiIiEjupCemk3QmHgD/inc2ngXAZDJljmlJ0pgWEZG7mqeRm3/wwQeEh4czZcqUzGNly5bN/H+Hw8Gnn37Km2++Sc+ePQGYNm0aJUqUYObMmQwbNoyYmBgmT57M999/T7t27QCYPn064eHhLF++nI4dO7r1MYmIiIiIiIiIiIiIiIjkVsLxGHCAtZgPXkFWp6zpHeJL4ul4ks8nQq0iTllTRETcz9ACj3nz5tGxY0d69+7NmjVrKFmyJE8//TSPP/44ABEREURFRdGhQ4fMa6xWK61bt2bDhg0MGzaM7du3k5aWluWcsLAwatasyYYNG25Y4JGSkkJKSkrm+7GxsQCkpaWRlpbmqocrUuBc+3nSz5WIuJLuNSLiLrrfiIi76H4jIu6i+42IuIvuN9nncDiIPRINgE85f6d9zjyLegGQHJVIakrqHY99kbuPfv5E8gdDCzyOHz/OhAkTGDlyJG+88QZbtmzhueeew2q1MnjwYKKiogAoUaJElutKlCjByZMnAYiKisLLy4vChQtfd8616/9uzJgxvP3229cdX7p0Kb6+vs54aCLyP5YtW2Z0BBEpAHSvERF30f1GRNxF9xsRcRfdb0TEXXS/uT2fdCuVYstgx876wxuxH7U7Z2EH1KAC5jRYtWAlyZ4pt79G8pXERI3nEckPDC3wsNvtNGzYkNGjRwNQr1499u3bx4QJExg8eHDmeSZT1ipCh8Nx3bG/u9U5r7/+OiNHjsx8PzY2lvDwcDp06EBgYGBuH46I/E1aWhrLli2jffv2WCwWo+OISD6le42IuIvuNyLiLrrfiIi76H4jIu6i+032Xd16kYTYGPzKBHL/Pfc7de2LK8+Rci6RRpUbEFA1yKlrS953baKBiNzdDC3wCA0NpXr16lmOVatWjdmzZwMQEhICZHTpCA0NzTznwoULmV09QkJCSE1N5erVq1m6eFy4cIHmzZvfcF+r1YrVev3MMovFol8sRFxAP1si4g6614iIu+h+IyLuovuNiLiL7jci4i6639yaw2Yn6UQ8AIFVCjv9c+Ub6kfKuUTSLqZgqaWvQ0Gjnz2R/MHDyM1btGjBoUOHshw7fPgwZcqUAaBcuXKEhIRkadmVmprKmjVrMos3GjRogMViyXJOZGQke/fuvWmBh4iIiIiIiIiIiIiIiEhekng6HnuKDbOvJz6hfk5f37uELwDJUYk4HA6nry8iIq5naAePF154gebNmzN69Gj69OnDli1bmDRpEpMmTQIyRrOMGDGC0aNHU6lSJSpVqsTo0aPx9fVlwIABABQqVIhHH32UF198kSJFihAcHMxLL71ErVq1aNeunZEPT0RERERERERERERERCRb4o7GAOBfoRAmD5PT17cW9cFkNmFPsZEWnYpX4eu73YuISN5maIFHo0aNmDNnDq+//jrvvPMO5cqV49NPP2XgwIGZ57zyyiskJSXx9NNPc/XqVZo0acLSpUsJCAjIPGfs2LF4enrSp08fkpKSaNu2LVOnTsVsNhvxsERERERERERERERERESyLT0xnaQzGeNZ/CsWcskeJrMJa3EfkiMTST6fqAIPEZG7kKEFHgBdunShS5cuN/24yWRi1KhRjBo16qbneHt7M27cOMaNG+eChCIiIiIiIiIiIiIiIiKuk3A8BhxgLeaDV5DrCi+8Q3wzCjyiEgmsWthl+4iIiGt4GB1AREREREREREREREREpKByOBzEHflrPIuLundc4x3iC0ByVCIOh8Ole4mIiPOpwENERERERERERERERETEIKmXk0mLTsFkNuFXLtCle1mL+oCHCVtSOumxaS7dS0REnE8FHiIiIiIiIiIiIiIiIiIGiT+a0b3Dt0wAZqvZpXt5eHpgLeYNQPL5RJfuJSIizqcCDxEREREREREREREREREDOGx24o/HAq4fz3KNd4n/jmkREZG7iwo8RERERERERERERERERAyQeDoee4oNs68nPqF+btnTJ+SvAg918BARueuowENERERERERERERERETEAHFHMsaz+FcshMnD5JY9rcV9wQTp8Wmkxae5ZU8REXEOFXiIiIiIiIiIiIiIiIiIuFl6YhpJZ+MB941nAfCweGAt6g1AclSC2/YVEZE7pwIPERERERERERERERERETdLOBYLDrAW88GrkNWte3uXyBgHkxylMS0iIncTFXiIiIiIiIiIiIiIiIiIuJHD4SDu6F/jWSq5r3vHNd4hPoAKPERE7jYq8BARERERERERERERERFxo7SrKaRFp2Aym/ArG+j2/b1L+AKQHpdGemKa2/cXEZHcUYGHiIiIiIiIiIiIiIiIiBslnk0AwDvUF7PV7Pb9PbzMeAVnjIVRFw8RkbuHCjxERERERERERERERETyoCNHIhnzwa90fOA/fPHlEqPjiBMlnY0HwKekv2EZvEP8AEg+rwIPEZG7hafRAURERERERERERERERCRDWlo6q9fs49dfN7H7z5OZx+f8tplhT7TH09P93R7EuezpdpLPJwHgU9LPsBzeIT7E7lcHDxGRu4kKPERERERERERERERERAx28VIsc+duYe68rVy+HAeA2exBmzY12LLlKHFxSRw4eJZaNUsbnFTuVHJUItgdmP08sQR6GZbDu4QvAGnRqdiS0zF762lDEZG8TndqERERERERERERERERAzgcDnbtPsGvv25i9Zp92Gx2AIoWCaBbt8Z069qIokUDefOfM1m1ei/bth1VgUc+kHQ2AQDfkv6YTCbDcpi9PbEEWUmLTiE5KhG/soGGZRERkexRgYeIiIiIiIiIiIiIiIgbJSWl8vvSXcz+dRPHjkVlHq9Tpyy9ejaldesaWUaxNGxY4a8Cj2M8PPQ+IyKLEyWdjQeMHc9yjXeIrwo8RETuIirwEBERERERERERERERcZOzZ68w7KmJXLmS8SS/t7eFjh3q8lDPplSsGHrDaxo1rAjAnr2nSEpKxcfHuLEecmfS49NIi0kFE3iH5o0Cj7iDV0k+n2h0FBERyQYVeIiIiIiIiIiIiIiIiLjJr3M2ceVKPCVKBNGvbws6PVCfgACfW15TsmQwJUoEcf58NLv/PEHTJpXdlFacLelcxngWa1EfzFbzbc52Pe8SvgCkXknBlmLLE5lEROTmPIwOICIiIiIiIiIiIiIiUhDY7XZWrNwDwIjnu9C3T4vbFncAmEwmGjWsAMC2bcdcmlFcKy+NZwHw9PXEEpjRESblgrp4iIjkdSrwEBERERERERERERERcYM9e05x4UIMfn5WmjaplKNrGzb4q8Bjuwo87lYOuyOzg0deKfAAsIZkdPFIjlKBh4hIXqcCDxERERERERERERERETdYvuJPAFq3qoHVasnRtQ3/6uBx+PA5oqMTnJ5NXC/lcjL2VDseXh5Yi96+c4u7+KjAQ0TkrqECDxERERERERERERERERdLT7exclXGeJZ27Wrn+Prg4ADKly8BwI4dx52aTdzj2ngW71A/TB4mg9P8l3eJjAKPlMvJ2NNsBqcREZFbUYGHiIiIiIiIiIiIiIiIi+3cGcHVqwkEBflmjlvJqUYNKwKwddtRZ0YTN0k6m/fGswB4+lvw9LeAA5IvJBkdR0REbkEFHiIiIiIiIiIiIiIiIi52bTzLvW1q4ulpztUa1wpDtm0/5rRc4h72VBspFzOKJ3zC8laBB4C3xrSIiNwVVOAhIiIiIiIiIiIiIiLiQqmp6axevReAdm1zPp7lmrr1ymE2e3D27BUiI686K564QVJkAjjAEuiFJcDL6DjX8Q7NKDpJOBGHw+EwOI2IiNyMCjxERERERERERERERERcaMuWI8TFJ1O0aCC1a5fN9Tp+vlZqVA8H1MXjbpNXx7Nc41cmAJOnB+mxqSSf15gWEZG8SgUeIiIiIiIiIiIiIiIiLnRtPEvb+2phNt/ZUzMNG2aMadm69egd5xL3cDgceb7Aw8PigV+5AADiD0cbG0ZERG5KBR4iIiIiIiIiIiIiIiIukpycyrr1B4A7G89yzbUCj23bj2G32+94PXG99Ng00uPTwAO8Q/JmgQdAQOUgABJOxGJPtRkbRkREbkgFHiIiIiIiIiIiIiIiIi6yYcMhkpJSCQ0tTPXqpe54vRrVw/H2thAdncDx4+edkFBcLelsPADexX3xsOTdp+asxXywFPLCYXMQHxFrdBwREbmBvPuniIiIiIiIiIiIiIiIyF3u2niWdm1rYzKZ7ng9i8WTenXLARldPCTvSzqXt8ezXGMymQioFARoTIuISF6lAg8REREREREREREREREXSEhIZsPGQ4BzxrNc06BBxpiWrdtU4JHXOWwOkiITAfAp6W9wmtvzr1gITJByKZnUq8lGxxERkb9RgYeIiIiIiIiIiIiIiIgLrF13gNTUdMqWLUbFiiFOW7dRw4oA7NoVQXq6zWnrivMlX0zEkW7Hw9uMV7DV6Di3ZfbxxDc8AIC4wzEGpxERkb9TgYeIiIiIiIiIiIiIiIgLLF++G3DeeJZrKlQoQVCQL0lJqezbd9pp64rzJZ39azxLmJ9TvwdcKaByIQDij8XgsNkNTiMiIv9LBR4iIiIiIiIiIiIiIiJOFh2dwJatRwFo68TxLAAeHh6ZY1q2bdeYlrwss8DjLhjPco1PSX/Mvp7YU2wknoo3Oo6IiPwPFXiIiIiIiIiIiIiIiIg42eo1+7DZ7FSuHEaZ0sWcvn7DBhljWrZtO+r0tcU5bMnppF5OBjI6eNwtTB4m/CtmdPGIOxJtbBgREcnC0AKPUaNGYTKZsryFhPx3Bp3D4WDUqFGEhYXh4+NDmzZt2LdvX5Y1UlJSGD58OEWLFsXPz4+uXbty5swZdz8UEREREREREREREZHbSkpKJSUlzegY4gbLV/wJZIxncYVGDTM6eOzdd5rExBSX7CF35lr3Dq9gK56+nganyZmASkFAxmNIj9c9S0QkrzC8g0eNGjWIjIzMfNuzZ0/mxz788EM++eQTxo8fz9atWwkJCaF9+/bExcVlnjNixAjmzJnDrFmzWL9+PfHx8XTp0gWbzWbEwxERERERERERERERuU50dAKfjVvI/Z3+wxNPTsThcBgdSVzo0qVYdu6MAKDtfbVcskdYWDBhoYWx2ezs2n3CJXvInUk699d4lruoe8c1lkAvvEN8AYg7Gm1sGBERyWR4gYenpychISGZb8WKZbQpczgcfPrpp7z55pv07NmTmjVrMm3aNBITE5k5cyYAMTExTJ48mY8//ph27dpRr149pk+fzp49e1i+fLmRD0tEREREREREREREhKSkVKZMXUnvvh/x449/kJZm48iRSCIiLhgdTVxo5aq9OBwOatYsTWhoYZft07DhtTEtx1y2h+SOw+HI7ODhU9Lf4DS54/9XF4/4IzEqShMRySMM7wd15MgRwsLCsFqtNGnShNGjR1O+fHkiIiKIioqiQ4cOmedarVZat27Nhg0bGDZsGNu3byctLS3LOWFhYdSsWZMNGzbQsWPHG+6ZkpJCSsp/25XFxsYCkJaWRlqa2kyJOMu1nyf9XImIK+leIyLuovuNiLiL7jci4i6637hWerqN+Qu28913a7hyNR6ASpVCSU+zEXHiAtu2HSE8PNjglOIqy5bvBuDeNjVc+jNWr25Z5s3fytatR/L0z3JBvN+kXU3BlpSOyWzCHOx5Vz52r5LemCwepMenEX86Fu9QX6MjyR24G78HReR6hhZ4NGnShO+++47KlStz/vx53n33XZo3b86+ffuIiooCoESJElmuKVGiBCdPngQgKioKLy8vChcufN05166/kTFjxvD2229fd3zp0qX4+uoPJxFnW7ZsmdERRKQA0L1GRNxF9xsRcRfdb0TEXXS/cS6Hw8GBg5dZt/4M0dEZLzQMCrJyT8twqlUNZuOmc0ScgMVLNmG1XjY2rLhETEwK+/adxmQCuy2KRYsWuWyvxMSMJ2yPHT/Pzz/Pxc/P4rK9nKEg3W+KJhUmjGLEeMSz+/clRsfJtZIexSlCEEfWHeS0/82fe5O8LzEx0egIIuIEhhZ4PPDAA5n/X6tWLZo1a0aFChWYNm0aTZs2BcBkMmW5xuFwXHfs7253zuuvv87IkSMz34+NjSU8PJwOHToQGBiYm4ciIjeQlpbGsmXLaN++PRZL3v6LhYjcvXSvERF30f1GRNxF9xsRcRfdb5zL4XCwdesxJn2znCNHIgEoXNiPIf9oTZcuDbBYMv45vkyZU6xbP5mo88k88MADt/33brn7zPxhPQD16pajT5/uLt9v8ZJzHD0WRXCR8rS9r5bL98uNgni/ubj8LClJSYTXKUP1qnWMjpNrqZeTubD4DIXTC1GjbR08rGajI0kuXZtoICJ3N8NHtPwvPz8/atWqxZEjR+jevTuQ0aUjNDQ085wLFy5kdvUICQkhNTWVq1evZuniceHCBZo3b37TfaxWK1ar9brjFoulwPxiIeJO+tkSEXfQvUZE3EX3GxFxF91vRMRddL+5c/v3n+bLib+zY8dxAHx9rQwccA99+7TA1zfrv0XXrFkGb28LMTGJnDlzlfLlS9xoSbmLrVy1F4D27eu65WerUaOKHD0Wxc6dJ7i/Y32X73cnCsr9xp5uJ/ViMgD+4YF39WP2LOGJV2ErqVdTSDmVSGB1jZa6W93N34ci8l8eRgf4XykpKRw4cIDQ0FDKlStHSEhIlnZdqamprFmzJrN4o0GDBlgsliznREZGsnfv3lsWeIiIiIiIiIiIiIiI3Kn0dBv/efdnHntiAjt2HMdiMdO3bwt++eklHh5633XFHQAWiye1apUBYMfO4+6OLC524uQFjhyJxGz2oE3rGm7Zs2GDCgBs3XYUh8Phlj3l1pKjEnHYHJj9PLEU8jI6zh0xmUz4Vw4CIO5ItKFZRETE4AKPl156iTVr1hAREcHmzZvp1asXsbGxDBkyBJPJxIgRIxg9ejRz5sxh7969DB06FF9fXwYMGABAoUKFePTRR3nxxRdZsWIFO3fuZNCgQdSqVYt27doZ+dBEREREREREREREJJ9buWovi5fsxGQy0emB+sz6YSTPD+9MUJDfLa+rX688oAKP/GjFij0ANG5ciUKFfN2yZ506ZfH0NBMVFc3Zc1fcsqfcWtK5BAB8wvzyxRgm//KB4GEi9UoKKZeSjI4jIlKgGTqi5cyZM/Tv359Lly5RrFgxmjZtyqZNmyhTJqN6+ZVXXiEpKYmnn36aq1ev0qRJE5YuXUpAQEDmGmPHjsXT05M+ffqQlJRE27ZtmTp1KmazZoCJiIiIiIiIiIiIiGs4HA6mT18DwGOPtuXhofdl+9r69coBsGtXBHa7HQ+PPNVsW3LJ4XCwfMWfALRrW9tt+/r6WqlRI5zdu0+wbdsxSpUs4ra95caSzsYD4FvS3+AkzmH29sSvtD8JJ+KIOxKDtaiP0ZFERAosQws8Zs2adcuPm0wmRo0axahRo256jre3N+PGjWPcuHFOTiciIiIiIiIiIiIicmObNh3m6LEofH28eKhnsxxdW7VqSby9LURHJxJx4gIVyoe4KKW409GjUZw8eREvL09a3VPNrXs3aljhrwKPo3Tv1tite0tW6fFppEWnggm8w27dzedu4l85iIQTcSQcjyG4UXE8PFWYJiJiBN19RURERERERERERERyaPqMtQB069aYwMCcvZrdYvGkdu2yAOzcGeHsaGKQa907mjergp+ft1v3btigIgDbdxzHbre7dW/J6tp4FmtRb8zW/NNt3ifMD7OfJ/ZUO4kn44yOIyJSYKnAQ0REREREREREREQkB/buPcXOXRF4eprp17dFrta4NqZlx47jzowmBjFqPMs11auXwtfHi5iYRI4cjXL7/vJf18az+OST8SzXmEwmAioFARB3JNrQLCIiBZkKPEREREREREREREREcuBa9477O9alWLFCuVqjXr3yAOzcFaGOC/nA/v1niIy8io+PF82bV3H7/p6eZur+VTS0fdsxt+8vGRx2B0mRiUBGx4v8xv+vAo/kyETS4lKNDSMiUkCpwENEREREREREREREJJtOnLjA2nX7MZlMDBhwT67XqVa1JN7eFmJiEomIuODEhGKEZct3A3BPy2p4e3sZkqFRw4wxLVu3HTVkf4HUy8nYU2yYLB5Yi+VsdNPdwOJvwfuvwpV4dfEQETGECjxERERERERERERERLJp+syM7h2t7qlG2TLFc72Op6eZ2rXLArBjp8a03M1sNjsrV+4BoH27OoblaNiwAgC7/zxBamq6YTkKssRr41lC/TB5mAxO4xoBlTK6FsUdicFhdxicRkSk4FGBh4iIiIiIiIiIiIhINpw/H83SpRmdGgYNbH3H69X/a6TGzp0Rd7yWGGfnrgguXY4jwN+bxo0rGpajfLkSBAf7k5ycxr59pwzLkZ85HA5syekkX0wi/lgMV3dd5OLac5xbeIJTsw4TvfMSAD4l8994lmt8Swfg4eWBLTGdpHMJRscRESlwPI0OICIiIiIiIiIiIiJyN5j10x+kp9uoX68cNWqE3/F69euVBzIKBOx2Ox4eek3m3WjmzHUAtGtXG4vFuKddTCYTDRtUYOmy3Wzbfox6f31/yc057A7saXYcaXbsmW+2LO/bktJJj00lLS6N9LhU7Kn2W67pGeiFX9kANz0C9/Pw9MC/QiFiD1wl7nA0vqX8jY4kIlKgqMBDRERERERERERE5AZSUtKIiUkkOiaB2JhEomMSiYlJJCYmgeiYRGJjEgkJKcwTj7fTE/MFQGxsIvPmbQVg0KA7794BULVqSXx8vIiJSeT48fNUrBjqlHXFffYfOMOmzYcxmz0Y0L+V0XEyCzy2bjvG44+1NzpOnpN8PpFLG6Kwp6RnFHak527EiNnXE0uAF56Blr/+64UlwIJngBdmq9nJqfMe/8pBxB64SuLpOGzJ6Zi99XSjiIi76I4rIiIiIiIiIiIiAkz+dgXr1h/ILOJITk7L1nU1a4TTsmU1F6cTo83+dRNJSalUqhRKk8aVnLKmp6eZ2rXKsHnLEXbsjFCBx11o2nerAOjQvg4lSwYbnAYaNswYEXPgwBkSEpLx8/M2OFHecnXHRdKiU67/gAd4WMx4WDzwsHhg+uu/HhYPPLzNGUUcAV5YAjOKODw8C3ZRnzXYG68i3qReTib+WAyFahQxOpKISIGhAg8REREREREREREp8DZuPMTkb1dcd9xs9qBQId/Mt6BCfgQG+hIU5MuRI5Fs3HSY3+ZtUYFHPpecnMrPv2wAYNDAVphMJqetXa9eeTZvOcLOncfp07u509YV1zt6NJJ16w5gMpkYPLiN0XEACAkJolSpIpw5c5kdOyO4R/emTKlXkkmOSgQThNxfBk8/z/8WdJgLdsFGbgRUDuLyxijij6jAQ0TEnVTgISIiIiIiIiIiIgVaSkoan3w6H4AHuzSkW9dGFCrkR6FCvvj5WW/6ZP6pU5fYuOkTNm06TFRUNCEhQW5MLe60YOF2oqMTCQsL5t42NZ26dv165QDYuSsCu92ucT93kanTMrp33HdfTcqULmZwmv9q2rQyv/yykXnztqrA43/EHrwKgG/pAHxCfA1Oc/fzKxfI5c3nSb2aQsqVZKzB6hYjIuIO+k1RRERERERERERECrTpM9Zy9uwVihYN5LnnOlO9ejglSwbj7+99y04NpUsXpX798tjtDhYs3ObGxOJO6ek2Zv6wDoAB/Vvi6Wl26vpVq5bE18eL2Ngkjh8/79S1xXVOnLjAqtX7ABg6+F6D02TVq2czTCYTf2w4yLHjUUbHyRPsqTbij8UAEFitsMFp8gez1YxvKX8A4o/GGJxGRKTgUIGHiIiIiIiIiIiIFFhnzlzm++lrABjxXGf8fK05ur57t8YAzJu/jfR0m9PzifGWr/iTqKhoChf2o3OnBk5f39PTTO3aZQHYsTPC6euLa0z7fjUOh4PWrapToUKI0XGyKF26KG1a1wBgxox1BqfJG+KOxuBId2AJsuKt7h1O41+xEAAJx2Nx2B0GpxERKRhU4CEiIiIiIiIiIiIFksPh4ONP5pGamk7jxpW4996cj95o3ao6QUF+XLoUy4aNh1yQUozkcDiYPmMtAH37tMBqtbhkn3p/jWnZseO4S9YX5zpz5jLLlu0GYEge695xzaBBrQBYtnw3kVFXDU5jLIfDQdyBjM9BYLXCt+zMJDnjW8ofD6sZW1I6SZEJRscRESkQVOAhIiIiIiIiIiIiBdLqNfvYvOUIFouZF194MFdP+lksnnTuVB+AuXO3ODuiGGzDxkMcP34eX18rPbo3cdk+9euVB2DnrgjsdrvL9hHn+H76Gux2B82aVaFq1ZJGx7mhalVL0bBBBWw2O7NmrTc6jqGSzyWQFpuKyeKBf4VAo+PkKyazCb9yAQCZI3BERMS1VOAhIiIiIiIiIiIiBU5iYgqffrYAgEEDWxEeXjTXa3XrmjGmZdPmI0RGFuxXyuc30/8a39Oje2MCAnxctk+VKmH4+ngRF5fEsWPnXbaP3LnIqKssWrwDgKFD8mb3jmv+Mag1kDFCKjq64HZXiP2re0dAxUJ4WMwGp8l//CtkjGlJPBmHPU0FaiIirqYCDxERERERERERESlwpkxdycWLsYSFFmbwP9rc0VqlShWhYYMKOBwO5s3f6pyAYrg//zzJ7j9PYrGY6dOnhUv38vQ0U7t2WQB27NSYlrxs+vS12Gx2GjaoQK2apY2Oc0sNG1agSuUwUlLS+GX2RqPjGCItLpXEM/EABFQtbHCa/MlazAfPAAuOdAcJJ2Odtm7yhUSi91xW0YiIyN+owENEREREREREREQKlGPHo5j14x8AjBzZFavVcsdrdu+W0cVjwcLtpKfb7ng9Md70GRndO+6/vx7Firp+rEP9+n+NadkZ4fK9JHcuXoplwcJtADw8NG937wAwmUyZXTx+mb2RpKRUgxO5X9yhaHCAd5gfXkFWo+PkSyaTCf+KGV084o85r8Dj6vaLXN12gas7LzptTRGR/EAFHiIiIiIiIiIiIlJgOBwOPvp4HjabnVb3VKd5sypOWfeee6pRuLAfly/Hsf6PA05ZU4xz/Ph51v9xEJPJxMD+rdyyZ/165QDYuSsCu12vWM+LZs5cR1qajTp1ylKvXnmj42RL69Y1KFWqCLGxSQWuw5A93U7c4WgAAtW9w6WujWlJPpdAekLaHa+XFJlAclQieJgoVD34jtcTEclPVOAhIiIiIiIiIiIiBcaSJTvZvfsE3t4Wnn++s9PWtVg86dK5IQBz5xasJ1Hzoxkz1wLQunV1Spcu6pY9K1cOw9fHi7i4JI4ei3LLnpJ9V67G89vcLQAMHZL3u3dcYzZ7MKD/PQDMmrW+QHUYSoiIxZ5iw+zniW+4v9Fx8jVLgBfW4j4AxB+/sy4eDocjs2tHQOUgPP3vvMuWiEh+ogIPERERERERERERKRBiY5MY/+ViAB4eeh+hIc59RXfXBxsBsGXrUc6du+LUtcV9oqKiWbpsNwCDBrZ2276enmbq1CkLaExLXjRr1npSUtKoXq0UjRtVNDpOjjxwfz2KFAng/IUYlv31vV0QxB68CmR07zB5mAxOk/9ljmk5GoPD4cj1OsmRiaScT8JkNhFUu4iz4omI5Bsq8BAREREREREREZECYdLXS7l6NYGyZYvRr28Lp69fsmQwjRtXwuFwFLhRCPnJrB/XY7PZadCgPNWrlXLr3tfGfuzYedyt+8qtxcQk8uuvm4CM4jCT6e4qFrBaLfTp3RyA6TPWFogRQCkXk0i9lAweJgIqBxkdp0DwKxsIHibSolNIvZKSqzUcDgdXd/zVvaNKEJ5+6t4hIvJ3ntk5qX79+jla1GQyMW/ePEqWLJmrUCIiIiIiIiIiIiLOdODgGeb8ljFe4aWR3bBYsvVPoznWvVsjtmw5woKF23ns0XZ4eppdso+4RmxsYmZxzj8Gua97xzX165UDYNeuE9jtdjw89BrNvOCnn/8gMSmVSpVCad68itFxcqVH9yZ89/1qIk5c4I8Nh7inZTWjI7lU7IGM7h3+5QIxe7vmfi9Zma1mfMP9STwZR/yxGKxFvHO8RtLZBFIuZnTvKFTLPeOxRETuNtn6U23Xrl28+OKL+PvffkaZw+Hg/fffJyUld9V5IiIiIiIiIiIiIs5ks9n5v4/m4nA46NChLvXrl3fZXi1bVKNIkQAuX45j7br93HdvLZftJc63YuUekpPTqFAhhEYN3T+Go3LlMHx9rcTFJXH0WBSVK4W5PYNkFR+fzM+/bARg6JB777ruHdf4+3vTo3sTps9Yy/fT19CyRdW79rHcji05nfiIWAACqzl3FJfcmn+FQiSejCPheAzBDYvnaDSOw+Hg6s6/undULYynrwpzRERuJNt3x5dffpnixYtn69yPP/4414FEREREREREREREnGnuvC0cPHgWPz8rw595wKV7eXqa6dypAd99v5q5c7eqwOMus3TZbgAe6FjPkCe/PT3N1KlTlo0bD7Fjx3EVeOQBv8zeSHx8MuXKFqd1q+pGx7kjfXo356efN7B37yl2/3mCunXKGR3phhwOBw6bAw/P3HWwiTscDXYHXkW9sRbzcW44uSXfUv54WM3YkmwkRSbgW/L2Lxy/JulMPKmXkjF5mgiqVcSFKUVE7m7Z+tMxIiKCYsWKZXvR/fv3U6ZMmVyHEhEREREREREREXGGK1fj+eqrpQAMe6IDRYoEuHzPbl0bYTKZ2LrtKGfOXHb5fuIcUVHR7N59ApPJRLt2tQ3LcW1My46dEYZlkAyJiSn8+NMfAAwZ3OauH5lTtGggD9xfD4DpM9YanObmYv68TOTik9iS03N8rcPuIPZgxniWwKrq3uFuJrMJv3KBAMQfjcn2dQ6Hg6s7Mrp3BFYNxuyj7h0iIjeTrd9GypQpk6Nq5fDwcMxmzZYUERERERERERERY3355RLi4pOpXDmMHt2buGXP0NDCNGlcCYC587a6ZU+5c8tX/AlA3TplKV68kGE56tXLGCG0e1cENpvdsBwCv/22hZiYRMJLFaFtW+OKfpxp4IBWeHiY2LDhEEePRhod5zq25HRi9l0h9VIykYtPkp6QlqPrE0/HY0tIx8Nqziw0EPfyr5hx/0w8GYc9zZataxJPxZN6JQWTpweFagW7Mp6IyF0vx+WmZcuW5Z133uHUqVOuyCMiIiIiIiIiIiLiFDt3RbBo8Q5MJhMvv9gNs9l9r77v1q0xAAsXbSctLeevQhf3W/bXeJb27esYmqNypVB8fa3ExSdz9FiUoVkKspSUNGbOWgfA4MFt3Hr/cKVSpYrQpnUNAKbPXGdwmuuZvT0J7VQGs68nadGpRC46SVpcaravv9a9I6ByUK5HvMidsRb1xjPQC4fNQcKJuNue73A4iN75V/eO6oUxe6t7h4jIreT4T7cXX3yRuXPnUr58edq3b8+sWbNISUlxRTYRERERERERERGRXLHb7Xwydj4AXR9sSI0a4W7dv0XzKhQtGkh0dAJr1u53696Sc8ePn+fI0Ug8Pc3c26amoVk8Pc3UrVsWgJ07jxuapSCbO28rV67EExpamI4d6hodx6kGDWwNwIoVfxIZedXgNNfzCrIS2qkMngEW0uPTiFx4ktSrt38eKjU6heRzCQAEVAlycUq5GZPJhH+Fv8a0HLv9mJbEk3GkXk3BZPGgUM0iro4nInLXy3GBx/Dhw9m+fTvbt2+nevXqPPfcc4SGhvLss8+yY8cOV2QUERERERERERERyZHdu09w7FgUvr5WnhzW0e37e3qaebBLAwDmztvi9v0lZ5Ytz+je0bRJJQoV8jU4DdSvmzGmZccOFXgY4czZy3w/fQ0A/xjYCk/P/DWSvmrVkjRqVBGbzc4Ps/JeFw8AS4AXoZ3KYAmyYktKJ3LxSVIuJt3ymri/unf4hvtjCfByR0y5Cf8KGWNakiMTbzlmx2F3cPWv7h2FagRjtuavnzUREVfIdX+qOnXq8Nlnn3H27FneeustvvnmGxo1akSdOnX49ttvcTgczswpIiIiIiIiIiIikm0LF2W8GK1t21qGPWH/YJdGeHiY2L79OKdPXzIkg9yew+HIM+NZrqlXvxwAu3afwGazG5ymYDl85BxPPvUVly/HUbp0UTp1amB0JJf4x8BWAMxfsJ2rV+MNTnNjnr4WQjuVwVrUG3uKjcglp0iKTLjhufY0G3FHM7pFBFQr7M6YcgOWAC+sJXyAW3fxSDgRS1p0Kh5eHgRWD3ZXPBGRu1quCzzS0tL46aef6Nq1Ky+++CINGzbkm2++oU+fPrz55psMHDjQmTlFREREREREREREsiUxMYVVq/cC0MXAJ2dDQoJo2rQykDHuQfKmfftOcy7yKj4+XrRsUc3oOABUqhiKn5+V+Phkjh6NNDpOgbFzVwTPPPs1V67EU6liKF+MexwvL0+jY7lEgwYVqFq1JCkpafwye6PRcW7KbDUTcn9pvEN9caTbOb/sNImn4647L/5YLI40O5ZAL3zC/AxIKn93rYtH/LHYG74o3GF3EL0ro/gxUN07RESyLccFHjt27GD48OGEhoYyfPhwatSowd69e1m/fj0PP/wwb775JvPmzWPOnDmuyCsiIiIiIiIiInJbc37bzDPPfs03k5dz6PA5dZstYFau2ktSUiqlw4tSs2ZpQ7N079oYgEWLt5Oamm5oFrmxpX+NZ7nnnur4+OSNsQ6enmbq1CkLwI6dEcaGKSDWrT/ACyOnkJCQQp06ZRk/7jGKFAkwOpbLmEwmBv3VxeOX2ZtITEwxONHNeVjMlGgXjm+4Pw6bg/MrzhB//L9dIRwOB7EHMsazBFQtjMlkMiqq/A+/soGYzCbSolNIvXL991dCRCxpMRndOwqpe4eISLbluMCjUaNGHDlyhAkTJnDmzBk++ugjqlatmuWc6tWr069fP6eFFBERERERERERya79B87wydj57NwVwbdTVvLwI+Pp+dCHfPzJPDZvOUJamp5kz+8WLtoOQKdO9Q1/oq9p08oUL16I6OhEVq/ZZ2gWuV56uo0VK/4EoGMeGc9yTf265QHYufO4wUnyv4WLtvPGmzNITU2nZctqfPrJwwQE+Bgdy+Vat6pBeKkixMUlMW9+3u4y5OHpQfH7SuFXPhAccHHNOWIPZhR1JEclkhadgsnThH/FQgYnlWvMVjO+4f7A9WNaHHYHV3deBKBQrSJ4eKl7h4hIduW4wOP48eMsWbKE3r17Y7FYbniOn58fU6ZMueNwIiIiIiIiIiIiOZGSksZ77/2CzWanQYPytLqnOt7eFs5fiGH2r5t4YeQUHuj8Hv/69w/8vnQXsbGJRkcWJztz5jK7d5/Aw8PEA/fXMzoOnp5mHuzSEIC5c7cYnEb+bvuO41y9mkBQkC+NGlU0Ok4W9etnFHjs2n0Cm81ucJr8a8bMtbw3ejY2m53OnRow+t0BWK03fu4jvzGbPRgwIKOLxw+z1uf5AkiTh4lircIIqBoEwOWNUUTvuZxZ6OFfoZDGfOQx18a0JByPwWH/bze1+GMxpMel4WE1E1itsFHxRETuSjku8ChTpowrcjBmzBhMJhMjRozIPOZwOBg1ahRhYWH4+PjQpk0b9u3LWuWekpLC8OHDKVq0KH5+fnTt2pUzZ864JKOIiIiIiIiIiORt305ZScSJCwQH+/PuOwN4f8wgFi/8J//3wWC6PtiIIkUCSExMYcXKPbz9zk90fnA0zw7/mh9/+oPIyKtGxxcnWLR4BwCNG1WiWLG88UruB7s0xMPDxM5dEZw4ecHoOPI/li3LGM9y37218PTMW08MV6oUip+flfj4ZI4ciTQ6Tp5z6PA5Jn29jH37TudqDJfD4eCLLxfzxZdLABg44B7eeL1nnvs+cLUH7q9H0SIBXLwYy+IlO12yx9GjkZw5e9kpa5lMJoo0DaFQ7SIAXN12gcQTcQAEVlWhQF7jU8ofD6sZW5KNpHMJQEb3juhdl4C/undYCtbPnIjIncpWgUdwcDCXLl3K9qKlS5fm5MmT2T5/69atTJo0idq1a2c5/uGHH/LJJ58wfvx4tm7dSkhICO3btycuLi7znBEjRjBnzhxmzZrF+vXriY+Pp0uXLthstmzvLyIiIiIiIiIid799+04zY+ZaAF55uTuFCvkCYLVaaNGiKq+92oO5c17lm0lPMfgfbShfvgQ2m50dOyP47POF9B84lp27Iox8CHKHbDY7i/8q8OjUqb7Baf6rePFCNG9WBYC58/L2GISCJCUlLXNsTof2dY0NcwNmswd165QDYMcOjWn5Xw6Hg7femsXUaat4fNgEhgwdxy+zNxIXl5St69PTbYwe8yszZq4D4Jmn7+eZpx8wfKSTEby8POnf/x4AvvlmOUlJqU5dPz3dxn/e/YWHH/6SI0euOGVNk8lEcIPiFG5QLPOYdwlfvIK9nbK+OI/Jw4R/+UDgv2Na4o9Gkx6fhoe3WUU5IiK54Jmdk6Kjo1m8eDGFCmWv4v3y5cvZLrCIj49n4MCBfP3117z77ruZxx0OB59++ilvvvkmPXv2BGDatGmUKFGCmTNnMmzYMGJiYpg8eTLff/897dq1A2D69OmEh4ezfPlyOnbsmK0MIiIiIiIiIiJyd0tJSePd0b9gtzvo0KEure6pfsPzPDw8qF49nOrVw3lyWAfOnr3C+j8OsHjJTg4fPseEib/z1YRhBfJJvvxg+45jnL8QQ4C/N/e0rGZ0nCy6d2/C+j8OsnDhdh57tB1+vlajIxV4f2w4SGJiCiEhQdSsGW50nBtq2LACf2w4yG/zttCrVzO8vLL1T/r53vYdxzl1+lLm5+PosSg+GTuf8V8s5r57a9G1a0Pq1C57w3t5Skoa/x41i3XrDmA2e/DqKz3o0rmBux9CnvJQz6bMnr2Rc5FXmTFzLY892s5pa8/+dRNHjkYSEOBDyZIBTlsXIKh2UczeZmL2XiGofrHbXyCG8K9QiNgDV0k8GYctOZ2rf3XvCKpdBA9LjgcNiIgUeNn+bXDIkCEuCfDMM8/QuXNn2rVrl6XAIyIigqioKDp06JB5zGq10rp1azZs2MCwYcPYvn07aWlpWc4JCwujZs2abNiw4aYFHikpKaSkpGS+HxsbC0BaWhppaWnOfogiBda1nyf9XImIK+leIyLuovuNiLiL7je589Wk3zl58iLBwf4Mf6Zjtj9/xYsH0LNHY1q3rkb/AZ+xd+8p/thwgCaNK7k4sbjC/PnbAGjbthYeHnnr56hB/bKULl2UU6cuMWfOJvr2aW50pAJ/v/l96S4A2t5XE5vNlie7Qt/fsQ7fT1/DmTOXmfXjOvr3a2l0pDxh9q8bgYzxIo8/1paly3azYOEOjh8/z5Lfd7Lk952ULl2ULp0b0LFjHYIK+QEQH5/MG2/OZPefJ/GyeDLqrd60aFG1wP4MXGMywbBh7Xlr1E/MmLmOTg/Uo1ixwDte99KlWL7+ZjkAjz5yL1avy07/XHuX88e7nD9QcO9leZ2pkBnPQAvpsWlErTiNLSEdD28zPuX99TVzM32+RfKHbBV42O12l2w+a9YsduzYwdat17cljIqKAqBEiRJZjpcoUSJz/EtUVBReXl4ULlz4unOuXX8jY8aM4e23377u+NKlS/H19c3x4xCRW1u2bJnREUSkANC9RkTcRfcbEXEX3W+y78zZOGb9uB+Ae1uHsX796lytU7tWEbZui2Ls2Dn8Y1ANdfG4yyQnp7Nm7V4AAgMTWbRokcGJrle9WgCnTl3i++mr8PW5gtmcN165XBDvN8nJ6WzYcBAAq1dMnvx+uaZpk+IsWhzPt1NWYPa4hL+/l9GRDBUfn8q6dQcACC6czLp1q/Dxhl49yxAZWZTdf17gwMHLnDp1iS8n/M5Xk5ZSqVJhqlcryvo/znDhQiJeXmZ69axETMxxFi3S+BvI6KpesqQ/Z8/GM+rtaXTuVOGO15w3/yiJiSmEhvrhZbkEmArk/UageFowIRQl9UIyAKc9Itm19IDBqQqexMREoyOIiBMY1s/t9OnTPP/88yxduhRv75vPRfv7X6QdDsdt/3J9u3Nef/11Ro4cmfl+bGws4eHhdOjQgcDAO69KFZEMaWlpLFu2jPbt22OxWIyOIyL5lO41IuIuut+IiLvofpMzycmpPPr4RAA6dqjDc8/1zPVaTZvG03/gp0RGJRAcXIlmzSo7K6a4wbx5W0lPd1CuXHEeebh3nizQadcunS1bx3LlSjxe1nA6dqhjaJ6CfL9ZuGgHNtt2ypUrzpAhvYyOc0v3328nIuIbDhw8y7EIB6+/2snoSIb67vs12O0OatYIZ+jQG3/tEhKSWbFiDwsW7uDQ4XMcPHiFgwevABBc2J//+/AfVKwY4s7Yd4Xy5evw5NNfs3ffJZ4b/hBVqoTleq3t249z4OBmPDxMvDNqEOXKFSuw9xuB9Pg0on7LePG2h4+Zxt2aYsojRY4FybWJBlKw2Gw2dW+5y1gsFsxm800/bliBx/bt27lw4QINGvx3tp3NZmPt2rWMHz+eQ4cOARldOkJDQzPPuXDhQmZXj5CQEFJTU7l69WqWLh4XLlygefObtzi0Wq1YrdfPuLRYLPrFQsQF9LMlIu6ge42IuIvuNyLiLrrfZM+XE5dy5sxlihYN5IURXe/oc1aiRGEe6tmUGTPXMWXqKu65p3qeLBKQG1vy+24AOndqgJdX3uxwYLFY6NO7BRO/+p0ff9pA504N8sT3WEG836xcmdHtpWOHunfFYx/5QlceHzaBJUt20atnM6pXDzc6kiFsNjsLFmwHoGfPpjf92gUFWXjooeY89FBzDh0+x7x5W1m6bBdFggP46KMhlCpZxJ2x7xq1a5ejQ4e6LF26iwkTlzJ+3GO5ukelpqbz6ecLAejZowk1apTJfHKxIN5vBCyFLXiH+ZF8LoHCdYvh5X39c3TievrZK1gcDgdRUVFER0cbHUVyISgoiJCQkBv+OWxYgUfbtm3Zs2dPlmMPP/wwVatW5dVXX6V8+fKEhISwbNky6tWrB0Bqaipr1qzhgw8+AKBBgwZYLBaWLVtGnz59AIiMjGTv3r18+OGH7n1AIiIiIiIiIiLiVrt3n+CnnzYA8NqrPQgM9LnjNQf0v4df52zm0OFzrFt/gFb3VL/jNcX1Tpy4wL79pzGbPbi/Y12j49xSj+6N+e67VRw7FsXmzUdo2lSdYtzt0qVYtu/IGMvRrm1tg9NkT40a4Txwfz0WL9nJ2E8X8NXEYXh4FLxXv2/ceIjzF2IoVMiXe9vUzNY1VSqH8fJL3XhhRBcAPD1v/opYgSef6MDq1XvZuSuCtWv307p1jRyv8cOs9Zw6dYngYH8ef6y9C1LK3ah4qzBSLiXjU8rP6CgiBcK14o7ixYvj6+ubJ4qK5fYcDgeJiYlcuHABIEsjjGsMK/AICAigZs2sv4D5+flRpEiRzOMjRoxg9OjRVKpUiUqVKjF69Gh8fX0ZMGAAAIUKFeLRRx/lxRdfpEiRIgQHB/PSSy9Rq1Yt2rVr5/bHJCIiIiIiIiIi7pGUlMp7Y2bjcDjo0rkBzZtVccq6hQv707tXM777fg2TJ6+gZYuqBfJJ1LvNwkU7AGjWrArBwQEGp7m1gAAfunZtzKwf1zPjh7Uq8DDAipV7cDgc1KpVmrCwYKPjZNtTT3ZkzZp97Nt/mt9/38UDD9Q3OpLbzfltM5DRqcdqzdkr0VXYkT0hIUH079eSad+tZvyXi2nevAoWS/afSoqMvMrUaasAGP5sJwIC7rz4UvIHs48nvuH+RscQKRBsNltmcUeRIupadbfx8cn4s/PChQsUL178unEtefpvp6+88gojRozg6aefpmHDhpw9e5alS5cSEPDfv6SNHTuW7t2706dPH1q0aIGvry/z58+/5VwaERERERERERG5u0386nfOnLlM8eKFeG54Z6eu3b/fPfj6WjlyNJK16/Y7dW1xvvR0G0t+3wlkPOl7N+jbpzlmswfbtx/n4MGzRscpcJYuyxjn06FdHYOT5EzRooEMGXIvAF9O/J2ExBSDE7nXuXNX2LT5CADduzU2OE3+NmhQa4KD/Tl79gqzf92Uo2vHfraAlJQ06tcrR4f2d9fPmIhIfnFtLJavr6/BSSS3rn3trn0t/1eOCzzMZnNmS5D/dfny5Tsuqli9ejWffvpp5vsmk4lRo0YRGRlJcnIya9asua7rh7e3N+PGjePy5cskJiYyf/58wsML5vxBEREREREREZGCYOfO4/z8y0YgYzSLv7+3U9cvVMiXPr2bA/DN5BXY7Xanri/OtXnLES5fjiMoyJfmze6ObhglSgTR/q/ighk/rDU4TcFy+vQlDhw4g9nswb331TI6To717dOCUqWKcPlyHNP+6pJQUMydtxWHw0GjRhUpVUqvRnYlP18rTzyeMVplypSVxMQkZuu6desPsH79AcxmD14c2VXjAEREDKb78N3rVl+7HBd4OByOGx5PSUnBy8srp8uJiIiIiIiIiIhkW2JiCu+N+RWArg82omkT1zyh369vS/z8rBw/fp7Vq/e5ZA9xjoWLtgPQsUO9HI0RMNqA/i0BWLVqL2fPXjE4TcGxbHlG945GDSsSXPjuGxXg5eXJc892AuDHn/7gzJnLBidyj9TUdOYv2AZAz+5NDE5TMHTu1IAKFUKIi0/m2ykrbnt+cnIqYz+dD0D/fi0pV66EqyOKiIgUSNku8Pj888/5/PPPMZlMfPPNN5nvf/7554wdO5ZnnnmGqlWrujKriIiIiIiIiIgUcBMm/s65c1coUSKI4c8+4LJ9AgN96NunBQCTv12BzaYuHnlRTEwi69cfBKBzp/oGp8mZihVDadK4Ena7gx9/Wm90nALB4XBkjmdpfxePjmjRoiqNG1ciLc3G5+MXGR3HLdas3Ud0dAJFiwbSooWeh3AHs9kjs5jo1zmbOXnq4i3PnzptNVFR0ZQoXoiHh97njogiIiIFUrYLPMaOHcvYsWNxOBxMnDgx8/2xY8cyceJEEhMTmThxoiuzioiIiIiIiIhIAbZt+zFm/7oJgDde64mfn3NHs/xd3z4tCPD3JuLEBVas3OPSvSR3li7bRXq6jSqVw6hYMdToODk2cGArAOYv2E50dILBafK/w4fPcerUJby8PGnVqrrRcXLNZDIx4rnOmM0erF9/gM1bjhgdyeXm/LYZgK4PNsTT885GxUv2NWpUkebNq2Cz2fniyyU3Pe/kqYvM/GEdACOe74KPj7q9i4iIuEq2CzwiIiKIiIigdevW7N69O/P9iIgIDh06xO+//06TJmqNJiIiIiIiIiIizpeQmMKY9zNGs/To3phGjSq6fM+AAB/69csYo/HtFHXxyIsWLtoBQKe7rHvHNQ3ql6dK5TBSUtL4dc4mo+Pke9e6d9zTshp+vlaD09yZsmWL0+uhZgB89vkC0tNtBidynePHz7Nr1wnMZg+6dW1kdJwC59mnH8gsJtq2/dh1H3c4HHz88TzS0200a1blri6eEhERuRtku8DjmlWrVlG4cGFSU1M5dOgQ6enprsglIiIiIiIiIiKSaeLE34mMvEpoaGGeftp1o1n+rk/v5gQG+nDq1CWWLd/ttn3l9o4cieTw4XNYLGY6tK9rdJxcMZlMmV08fpm9keTkVIMT5V82m51ly/8E7u7xLP/rkYfvIyjIjxMnLmZ2N8qPfpu3BYAWzatSrFghg9MUPGXLFqd7t8YAjBu/6Lpix+Ur/mTb9mN4eXkycsSDmEwmI2KKiEg+8N1331GkSBFSUlKyHH/ooYcYPHiwQanynhwXeCQlJfHoo4/i6+tLjRo1OHXqFADPPfcc77//vtMDioiIiIiIiIhIwXbxYgxz520F4PVXe7j1lfd+ft4M6H8PAN9OWZmvXyV/t1m0OKN7R8sW1ShUyNfgNLnXpnUNQkMLEx2dmPmYxPl27z7BpUuxBPh707RJZaPjOEVAgA/DnugAwORvV3D1arzBiZwvKSmVxX/9XPTooQ7iRnnk4fvw9/fmyJFIFi/ZmXk8ISGZceMWATBkcBtKlgw2KqKIiNyGw+EgKSnVkDeHw5GtjL1798ZmszFv3rzMY5cuXWLBggU8/PDDrvrU3HU8c3rBa6+9xu7du1m9ejX3339/5vF27drx1ltv8dprrzk1oIiIiIiIiIiIFGw//7KR9HQbdeqUpWFD149m+buHHmrGD7PWc+bMZZYu202nB+7OcSD5SVpaOr8vzXiSsXPnBganuTOenmb692vJJ2Pn88Os9XTr2hizOcevy5PbuDaepc29NfHyyvE/i+dZXTo3YM5vmzl8+BxfTVrGa6/2MDqSUy1bvpuEhBRKlgymUcMKRscpsAoX9mfo4HsZ/+ViJk1ayn331sTX18o3k1dw6XIcpUoVySyGFBGRvCk5OY227UcZsveKZaPw8fG67Xk+Pj4MGDCAKVOm0Lt3bwBmzJhBqVKlaNOmjYtT3j1y/DeF3377jfHjx9OyZcssrbaqV6/OsWPXz18TERERERERERHJrYTEFH6bm9Ge36gnj/x8rQwckDFGQ1088oYNGw4RHZ1I0SIBNG7k/qIfZ+vcqQGBgT6cPXuFNWv3GR0n30lNTWfV6r0AdMgn41muMZs9GPF8ZwDmL9jGocPnDE7kXL/9lnH/796tMR4eKnwyUq9ezQgLC+bS5Thm/rCOw0fO8fMvGwB4cWRXrFaLwQlFRCQ/ePzxx1m6dClnz54FYMqUKQwdOlQjwP5HjkuVL168SPHixa87npCQoE+siIiIiIiIiDjd1avxpNvsFCsaaHQUMcCCBduIj0+mdHhRWjSvYliOh3o2ZeYP6zh37gqLl+zkwS4NDcsisHDRdgA6dqyHp6fZ4DR3zsfHi14PNePbKSuZMXMd97apqX9rdaLNW44QF5dE0aKB1K1Tzug4Tle3Tjnata3N8hV/8umn8/nyiyfyxffP/gNnOHjoLF5ennTudHd36skPvLw8eebp+3nznzOZMXMd69YfwG53cN+9NWnSuJLR8URE5Da8vS2sWDbKsL2zq169etSpU4fvvvuOjh07smfPHubPn+/CdHefHJe8NmrUiIULF2a+f+0Xxa+//ppmzZo5L5mIiIiIiIiIFHjx8ckMeXg8PR/6kC8nLCElJc3oSOJG6ek2fvzpDwD69Wtp6Ku3fXy8GDQwo4vH1GmrSEtLNyxLQXflShwbNx0GoHPn/DMu56GeTfHy8uTAgTPs2hVhdBy3Sk+3sX3HMZKSUp2+dkxMIj/+uB6Adm1r59vxN888fT9Wq4Xdf55k+Yo/jY7jFL/9thmAe++tSVCQn8FpBKBN6xrUqV2GlJQ0jhyJxNfHi+ef62x0LBERyQaTyYSPj5chbzktPH3ssceYMmUK3377Le3atSM8PNxFn5W7U45/mx0zZgxvvvkmTz31FOnp6Xz22We0b9+eqVOn8t5777kio4iIiIiIiIgUUDNmruXSpVhsNjvTZ6xlyNBx7NpdsJ74LMhWr9lHVFQ0QUF+PHB/PaPj0LNHE4KD/YmMvMrCRTuMjlNgLfl9FzabnRo1wilb5vpOw3erwoX9M7sUTJ+5zuA07vXxJ/MY/txk+vb/hAULtmGz2e94zfR0Gz/9vIE+/T5mx84IzGYPOj1g/H3EVUqUCGLwP1oD8MWXS1xSLJMbV67EsWLFnzku0IyNTWLZ8oxClR7dmrgimuSCyWRi+PD/FnQ88khbihUrZGAiERHJjwYOHMjZs2f5+uuveeSRR4yOk+fkuMCjefPm/PHHHyQmJlKhQgWWLl1KiRIl2LhxIw0aqE2aiIiIiIiIiDjHpUuxmd0b+vdrSdGigZw6fYmnn/majz6eS0JiisEJxZUcDgcz/3qS+6GeTbFas9/W11W8vb34x6CMJ1Cnfbea1FR18XA3h8OROZ4lP45s6N+vJSaTiY0bD3HseJTRcdxi//7TzJ23Fci4749+/1eGPjKeTZsP53rNTZsOM3jo53z62QLi4pKoUCGEz8Y+QsWKoc6KnScN6H8PISFBXLgQw4yZa42Ow9GjkTz86Bf8661ZPPHkRM6cuZztaxcv2UFKShoVKoRQq1ZpF6aUnKperRQvjOhC3z4t6NO7udFxREQkHwoMDOShhx7C39+f7t27Gx0nz8lVP7patWoxbdo09u7dy/79+5k+fTq1atVydjYRERERERERKcCmTltFcnIaNWqE8+wzDzDj++fp+mAjAH6ds5lBgz5l48ZDBqcUV9m1K4KDh87i5eVJzx5559Xb3bs1pmiRAM6fj2bBwm1Gx8lzUlLSmDptFQsWbic6OsHp6x84eJaIiAt4eXnSrm1tp69vtFKlitCmdQ0AfvhhvcFpXM9ut/PRJ/MA6NC+Ds8N70RAgA/HjkUx8sWpjBg5haNHI7O93omTF3jx5WmMfGkqJ05cJCjIl1de6saUyc9Qv355Vz2MPMNqtfDs0w8AMGvWeq5ejTcsy9atR3ny6UlcvBgLwJEjkTz86HhWr9l722sdDge/zd0CZNxzc9rWXVyvd6/mPP9cZzw9zUZHERGRfCoyMpKBAwditVqNjpLn5LjAIzY29oZvcXFxpKbmjbZvIiIiIiIiInJ3O3PmcuYrup9+siMmk4mAAB9ee7UHn3/2KGFhwZy/EMOLL0/jnf/8TExMosGJxdl+mJXx5HbnTvUpXNjf4DT/ZbVaGDy4DZDRxSOnYwfyu++nr2HS18sYPWY2D3Ybw7PDv+annzcQFRWd6zUjI68yd95W/vXvHxj54hQA2rSugb+/t5NS5y0DB9wDwNJlu7lwIcbgNK61YOF2Dh48i5+fleHPdqJf35b8/ONL9O/XEovFzJYtRxjy8HjeHf3LLT8XsbFJfPrZAv4x+HM2bjyE2exBv74t+fGHF+nevUmBehL63ntrUqVyGIlJqcwwaNTP4sU7GPnSVBITU6hXtxzTv3ueOrXLkJCQwhtvzuSzzxeSlnbzDkg7d0Zw8uRFfHy8uL9jXfcFFxEREcNduXKFWbNmsXLlSp555hmj4+RJOS7wCAoKonDhwte9BQUF4ePjQ5kyZXjrrbew2+98TqKIiIiIiIiIFEyTvl6GzWanaZPK1KuX9VXXDRtU4Ptpz9Gvb0s8PEws+X0nAwaNZeXKPTgcDoMSizOdOHmB9X8cxGQy0bdPS6PjXOfBLg0pUbwQFy/GMnv2JqPj5BkJiSn8/MtGAMLCgrHZ7OzYGcGnny2gZ68PeeTRL5g6bRXHj5+/5c9qbGwiq1bv5f8++o3efT/iod7/xwcfzmHFyj3ExiYREOBD374t3PWw3K569XDq1S1HerqNn37eYHQcl4mNTWTCxN8BeOzRdhQpEgBAYKAPw5/txA8zXqBd29o4HA4WLdpB3/6f8NWkpVnGc6Wn25j96yb69v+Yn37egM1mp2WLqsz4fkRmN5CCxmQy8cTj7QH4ZfZGLl6KddveDoeDKVNX8p/3fsFms9OubW3GfvIw5cuXYNznj2UWL/340x88M/wbzp+PvuE6c+ZuBqBjh7r4+eXPQi4RERG5sfr16zNs2DA++OADqlSpYnScPMkzpxdMnTqVN998k6FDh9K4cWMcDgdbt25l2rRp/POf/+TixYt89NFHWK1W3njjDVdkFhEREREREZF87NDhcyxf8ScATz7Z4Ybn+Ph48dzwTrS9rxajx8wm4sQF/vnvH2h1T3VeerErRYsGujOyONmsWX8A0LJlVUqXLmpwmutZrRYefbQdo8fMZtr3q3nwwYYF8onkv5v72xbi4pIoHV6UGdNHcP58NGvX7WfN2v38+edJDh46y8FDZ5n09TLCSxWhVasatG5VnUqVQtm77xRbtx5l67ZjHDx4NksBiNnsQY3q4TRsWIHGjSpRvXqpfN+RYcCAe9i5K4Lf5m5h6JB782W3kklfLyMmJpFy5YrzUM+m1308LCyYd97uR98+LRj/xSJ2/3mSad+tZt78rTz6SFvCwoIZ/8Vijh8/D0C5csV5bnhnmjSu5O6Hkuc0bVqZWrVKs2fPKb77bjUvjuzq8j3T023838dzmT8/Y3TVwAGteOrJDnh4ZLzG1NPTzDNPP0Dt2mV5992f2bv3FEMfGc9b/+pD06aVM9e5ciWONWv2A9C9e2OX5xYREZG85cSJE0ZHyPNyXOAxbdo0Pv74Y/r06ZN5rGvXrtSqVYuvvvqKFStWULp0ad577z0VeIiIiIiIiIhIjk386xXdHdrXoXKlsFueW6NGOFO+fZbvvl/NtO9Ws3bdfnbsPM4Lz3fhgQfquyGtONuVq/Es+X0nAAP63WNwmpu7v2NdZv6wlhMnLjJj5jqeHHbjYqSCIiUljVk/ZozVGTSoNWazB2FhwfTr25J+fVty5Wo869cfYM3a/WzbdpTTZy4zY+ZaZsxci8lkuq6jR9myxWjYsCKNG1akXr1yBe5V/M2aVqZcueJERFzgt7lbGDSwldGRnOrQ4XP8NncLACNfePCWBTs1aoTz5RdPsHbdAb78cjGnz1zmo4/nZX68UCFfHnu0Hd26Nsr3hT/Zda2Lx/DnJjN33lYGDLiH0JDCLtsvMTGFf/7rBzZtPoyHh4kXnu/CQw81u+G597SsxpRvn+Wf//qBg4fO8uLL0xgyuA2PPtIWs9mD+Qu2k55uo0aN8Nv+DiAiIiJSEOV4RMvGjRupV6/edcfr1avHxo0ZLRhbtmzJqVOn7jydiIiIiIiIiBQo23ccY/OWI5jNHjz2WLtsXePl5cljj7Zj6rfPUrVqSeLjk/nPe7+wZs0+F6cVV/j1102kpqZTo3o4tWuXMTrOTXl6mhn2REcgY9zAJTeOQciLFi/ewaXLcZQoXoiOHepc9/Hgwv50fbARH//fEBYteJN33u5Hu7a18fW14nA4CA72p0OHuvzzzV789uurzJz+AiNHPEjLltUKXHEHgIeHBwP6ZxQ4zfxhHXFxSQYnch6Hw8EnY+dhtzto27YWDepXuO01JpOJ1q2qM2P6CEa+8CBBQb6YzR706d2cH394kYd6NlVxx980qF+BBg3Kk55uY+rUVS7b5/LlOJ559ms2bT6M1WphzOhBNy3uuCYsLJgJXz5Bj+5NcDgcTJ22ihEjv+XSpVjmzsso/OnZvYnLMouIiIjczXJc4FGqVCkmT5583fHJkycTHh4OwOXLlylc2HUVwSIiIiIiIiKS/zgcDiZMyOje0b1bY0qVLJKj6ytUCGHSxCczW/2P+eBXLlyIcXpOcZ3k5FRm/7oJgP79WmIymQxOdGut7qlGjRrhpKSkMWXqSqPjGCY93cb0GWsB6N//HiyWWzcN9vPzpl3b2rzzdj8WLXiT2T+/zPy5rzPq333o9EB9ihcv5I7YeV7HDnUpW7YY0dEJTPp6mdFxnGbJ77vYs+cUPj5eDH+mU46u9fQ00+uhZvz6yyvM/e01RjzfhcBAjUe6mScez+gstGjxDk6fvuT09U+cuMDjwyZw6PA5goL8GP/5Y9zTslq2rrVaLbz8UjdG/bsPPj5ebN9+nH79PyEqKpqAAB/uu6+W0/OKiIiI5Ac5LvD46KOPGDt2LHXq1OGxxx7j8ccfp27dunz66ad8/PHHAGzdupW+ffs6PayIiIiIiIiI5F9r1u5j/4Ez+Ph48fDQe3O1hqenmeeGd6Jq1ZLExibx9js/YbPZnZxUXGXR4h3ExCQSFlqYVq2qGx3ntkwmE08/mdHFY978bS55AvVusGLlHs5FXiUoyJeuDzbM0bVeXp6EhhbO88U8RvD0NDPyha4AzPltM4cOnTU0T2xsImfOXL5unE5OxMcn88WXiwEYOuTeXBfzeHt7EVzYP9c5CopaNUvTvHkVbDY7k79d4dS1d+2OYNhTXxEVFU2pUkWY9NWT1KgRnuN1OnSoy+Svn6Zc2eIkJqUC0LlTA6xWi1PzioiIiOQXOS7w6Nq1K4cPH6ZTp05cuXKFS5cu8cADD3Dw4EG6dOkCwFNPPcUnn3zi9LAiIiIiIiIikj+lp9uY+NVSAPr2aUFwcECu17JYPHlnVD98fbzYuSuC775f7ZyQ4lI2m50ff/wDgL59W9w14xbq1StPs6aVsdnsTPom/3RZyC673Z75M9a3T0u8vb0MzZPfNGxQgXZta2O3O/jok3nY7a4tWEtPt3Hy1EXWrT/A9BlrGT1mNk8+9RWdurzL/Z3epU+/j3nple9ITk7N1fqTp6zgypV4SocXpW+fFk5OLzfy+GPtAVi2/E+OHY9yyprLV/zJ8yO+JS4uiRo1wvlqwrAcd936X2XLFuebr5+me7fGVK9Win599b0hIiIicjO37pf4N2lpaXTo0IGvvvqKMWPGuCqTiIiIiIiIiBQwixbv4NSpSxQq5MvAAffc8XqlShXhxZFd+c97v/DtlJU0bFiRWjVLOyGpuMr6Pw5y+sxlAvy96dypgdFxcuTJYR3ZtPkIK1bsYWD/VlStWtLoSG7zx4ZDRERcwM/PSs8eTYyOky8NH96JDRsOsm/faRYu2sGDXXLWJeVmUlPTWbZ8N6tWn2LdHzM5c/oyZ89duWXXIw8PExs3HuKFkVP48IPBBARkfzzK8ePn+eWXjQCMGNEFL68c/dO05FKVymHc26Ymq1bv5ZtvljNm9KA7Wu/nXzYw9tMFALRuVZ23/t3HKYVdPj5evPJy9zteR0RERCS/y1EHD4vFwt69e9UyUUREREREREScJiUlLbN1/JDBbfDz83bKuvffX48O7etgs9kZ9faPxMcnO2VdcY0fflgHQI8eTfD1tRqcJmcqVQqlffs6AEyctNTgNO7jcDiYNm0VAD17NM3Rk/2SfcWKBvLoo+0A+HLCEmJjE+94TZvNzj//NZP3Rv/Klq2RbNhwiFOnL2Gz2fH2tlC5chjt2tbmkYfvY9RbfZny7bMsX/oWX45/An9/b3b/eZJnn/uGK1fisrWfw+Hgk0/nY7PZad2qOk2bVL7jxyDZ99ijbTGZTKxZu5+DB3M/6ufXOZsyizt69WrGu/8ZoK49IiIiIm6W4xEtgwcPZvLkya7IIiIiIiIiIiIF0C+zN3LxYiwlSgTRo7vzOgCYTCZefqkbYaGFiYy8ygf/NweHw+G09cV59u49xZ97TuLpaabXQ82MjpMrjz/aDk9PM1u2HGHb9mNGx3GL7TuOs//AGby8POnbp7nRcfK13r2aUa5ccWJiEvnKCUVEX05Ywvo/DuJl8aR+vRKMeK4Tn419hN9+fZUVy0Yx9dtneeftfjz2aDs6tK9Dlcph+PpaqV27DF+Me5zgYH+OHInkqacnERl19bb7rVy5hx07juPl5clzwzvfcX7JmXLlStChQ0YR2te5HCW1YME2Pvp4HgCDBrbihee7YDbn+OkFERERkVsaOnQoJpOJ999/P8vx3377LbMJxerVqzGZTJlvPj4+1KhRg0mTJt1wrb+/3X///dftO3r0aMxmc5Z9y5Yte8Prr721adMGgJ07d9KlSxeKFy+Ot7c3ZcuWpW/fvly6dMnJn50MOf4NLDU1lQkTJtCgQQOGDRvGyJEjs7yJiIiIiIiIiGRXXFwS332/Bsh4hbHVanHq+n5+3rw9qh9mswcrVuxh4aIdTl1fnOOHWesB6NihLkWLBhqcJndKlgyme7dGAEyY+HuBKCb67rvVAHR9sCHBwQHGhsnnPD3NvPRiNwB+m7uV/QfO5HqtufO2Zv7Mvf5aD9q3K0uPHk1o1KgixYsXum335kqVQpnwxROEhARx+sxlnnp6EidOXrjp+YmJKXw+fhEAg//RmtDQwrnOLrn36MNtMZs92LjpMH/+eTJH1y5duosxH8wBoE/v5jz1ZEd1+RYRERGX8fb25oMPPuDq1VsXEh86dIjIyEj279/PsGHDeOqpp1ixYkWWc+6//34iIyOzvP3www/XrTVlyhReeeUVvv3228xjW7duzbxm9uzZWfaMjIzk119/5cKFC7Rr146iRYvy+++/c+DAAb799ltCQ0NJTLzzzns3kuMCj71791K/fn0CAwM5fPgwO3fuzHzbtWuXCyKKiIiIiIiISH41Y+Za4uKSKFeuOPd3rOeSPWrUCOfxxzLGG4z9dD6nTrnmVTSSO2fOXmbN2n0A9OvXwuA0d2bokHvx8fHiwIEzrF6zz+g4LrV//2m2bT+G2exB//73GB2nQKhXtxwdO9bF4XDw0cdzsdnsOV5j27ajfPTxXCCjqO6++2rmKkt4eFG+mjCMsmWLceFCDE8/M4kDB29cdDLtu9VcvBhLWGhhBg5olav95M6VKlWEzp0aADDp66XZLkJbuWoP/3nvFxwOBz26N+b55zqruENERERcql27doSEhDBmzJhbnle8eHFCQkIoV64czz33HGXLlmXHjqwv6rBarYSEhGR5K1w4a8HxmjVrSEpK4p133iEhIYG1a9cCUKxYscxrgoODs+x57diGDRuIjY3lm2++oV69epQrV4777ruPTz/9lNKlSwPwzjvvEBYWxuXLlzP37Nq1K61atcJuz/nv9Dku8Fi1atVN31auXJnjACIiIiIiIiJSMF28FMuPP20A4MlhHV3a6n3ggFbUr1+epKRU3ho1i9TUdJftJTnz009/YLc7aNqkMhXKhxgd544EBwfQr29GkcpXk5aSnm4zOJHrXOu807FjXUJD1JHBXZ59+gH8/KwcPHiW+Qu25ejak6cu8uY/Z2Kz2WnfrjYPD73vjrIUK1aIL8c/QdWqJYmOTmT48G/YseN4lnNOnbqU2S1kxPNdnN6lSXJm6JB7sVjM7NgZwfZsjJJav/4Ab436EZvNTudODXhxZFcVd4iIiNylHA4H9jS7IW857W5oNpsZPXo048aN48yZ23euczgcLFmyhNOnT9OkSc7Hvk6ePJn+/ftjsVjo378/kydPzva1ISEhpKenM2fOzUfCvvnmm5QtW5bHHnsMgIkTJ7J27Vq+//57PDxy/u8gnjm+QkRERERERETECaZMWUlKShq1apWmZYuqLt3LbPbgrX/1ZvDQcRw6fI6vJi1l+LOdXLqn3F5sbCILFm4HYED/lgancY4B/e9hzm+bOXXqEgsX7aBb10ZGR3K648fPs3bdfkwmE4MGqiODOxUpEsDjj7Xn088WMPGr32nTugZBQX63vS4mJpGXX/mOuPhkatYszRuvP+SUJ+qDgvwY99mjvPra9+zYGcHIl6byn3f6c0/LajgcDsZ+Op/0dBvNmlWhhYvv83J7ISFBdOvWmF9+2chXXy+jQYMKN/0+2LT5MG/+K6MgqEP7Orz2ao9cPQEhIiIieYMj3cHJ6YcM2bvMoCqYLDn73bNHjx7UrVuXt95666YFF6VKlQIgJSUFu93OO++8Q6tWWf9+smDBAvz9/bMce/XVV/nXv/4FQGxsLLNnz2bDhowXnwwaNIgWLVowbtw4AgNvPz60adOmvPHGGwwYMIAnn3ySxo0bc9999zF48GBKlCgBZBSsTJ8+nbp16/Laa68xbtw4Jk2aRJkyZXL0ObkmV7+Rbd26lVdeeYV+/frRs2fPLG8iIiIiIiIiIrdz+vSlzFefP/VkR7e8IrhYsUK88fpDAPwwaz2bNh92+Z5ya3N+20JychqVKoXSoEEFo+M4hZ+fN0MG3wvA5G9XkJKSZnAi5/t+Rkb3jtatq1O2THGD0xQ8PXs0oWKFEGJjk5jw1e+3PT8tLZ033pzBmTOXCQkJ4v3RA53aScPPz5uPPxpKy5bVSE3N2Gvxkp2sXXeAzVuOYLGYGaGxHnnGkH+0wWq1sG/faTZsvPGTPNt3HOO116eTlmajTZsa/PPNXi7tsiUiIiJyIx988AHTpk1j//79N/z4unXr2LVrF7t27eKbb75h9OjRTJgwIcs59957b+Y5196eeeaZzI/PnDmT8uXLU6dOHQDq1q1L+fLlmTVrVrZzvvfee0RFRTFx4kSqV6/OxIkTqVq1Knv27Mk8p3z58nz00Ud88MEHPPjggwwcODAnn4osctzBY9asWQwePJgOHTqwbNkyOnTowJEjR4iKiqJHjx65DiIiIiIiIiIiBcekr5dhs9lp3rwKdeuUc9u+97SsxkM9mzL71038592f+X7acwQHB7htf/mv1NR0fv4l41VS/fu1zFdP/nbv1phZP/7B+fPR/PzLxnzV5eLcuSssX/4nAIMHtTE2TAHl6WnmpRe78eTTXzF//jYe7NyQmjVL3/Bch8PB/300l527IvD1tfJ/Hw52yT3ParUw+t0BjB7zK0t+38l/3v2ZgAAfAPr3u4fw8KJO31Nyp0iRAHo91IwZM9cy6etlNGtaOUtnjt27T/DyK9+RmppOyxZVefutvnh6mg1MLCIiIs5g8jRRZlAVw/bOjVatWtGxY0feeOMNhg4det3Hy5UrR1BQEAA1atRg8+bNvPfeezz11FOZ5/j5+VGxYsWb7vHtt9+yb98+PD3/WzZht9uZPHkyTzzxRLazFilShN69e9O7d2/GjBlDvXr1+Oijj5g2bVrmOWvXrsVsNnPixAnS09Oz7JkTOS67HT16NGPHjmXBggV4eXnx2WefceDAAfr06UPp0jf+i4SIiIiIiIiIyDUHD55lxco9mEwmnnyig9v3f/aZByhfvgRXrybw7nuzsdvtbs9wN7DZ7EydtorxXyxm/oJt7NlzktjYxFytZbfbOX36EitW/MmEib/zwsgp9HjoA65ciadYsUData3t5PTGslotPP5YOwC+/341sbFJxgZyohkz12Gz2WnSuBJVq5Y0Ok6BVbt2GTp1qg/Ax5/Mw2a78X1s5g/rWLBwOx4eJt55ux8Vyoe4LJOnp5l/vvkQvXo1AyAuLokSxQsxZHAbl+0puTNoYCt8fa0cORLJ6jX7Mo/v33+aF1+eRnJyGk0aV+Ld/wzAYtGUdxERkfzAZDLhYfEw5O1Oivnff/995s+fnzlC5VbMZjNJSdn/u9eePXvYtm0bq1evztLhY+3atWzdupW9e/fmKrOXlxcVKlQgISEh89iPP/7Ir7/+yurVqzl9+jT/+c9/crU25KKDx7Fjx+jcuTMAVquVhIQETCYTL7zwAvfddx9vv/12rsOIiIiIiIiISP43YWLGSIEOHepQsWKo2/e3Wi28M6ofjzz2BZs2H+annzfQr29Lt+fI6376eQOTvl523fHChf0oU6YYZUoXo2zZ4pQpXYwyZYtRonghPDw8SEtL53jEBY4cPsfhI+c4fDiSo0cjSUxKvW4ts9mDJx5vny9fHd6xQ11mzFxLRMQFZsxcy1NPdjQ6EgBJSan8seEgW7cepXLlMLp1bZTtz/+lS7EsXLQdgMH/aO3KmJINTz91P2vX7ufQ4XP8NncLD/VsmuXja9bu58sJGffb54Z3pnkz179i08PDgxee70KR4ADmztvCq6/0wMfHy+X7Ss4UKuRLv74t+HbKSr6ZvJzWrWpw9FgUL4ycQmJiCvXrl2fM6IF4eam4Q0RERIxVq1YtBg4cyLhx46772IULF0hOTiYlJYUtW7bw/fff06tXryznpKSkEBUVleWYp6cnRYsWZfLkyTRu3JhWra7vuNisWTMmT57M2LFjb5lvwYIFzJo1i379+lG5cmUcDgfz589n0aJFTJkyBYAzZ87w1FNP8cEHH9CyZUumTp1K586deeCBB2jatOkt17+RHP+GFhwcTFxcHAAlS5Zk79691KpVi+joaBITc/cqDhEREREREREpGDZtOszWbUfx9DTz+KPtDMtRvnwJnhveiY8+nseXE36nXt1yVKmibgTXnDlzObO4o03rGiQkpnDq5EXOX4jh6tUErl5NYNeuE1musVotFC8WSGRUNOnptuvW9PLypEKFECpXCqVy5TAqVwqjQoUSeHvnzyd/zWYPnhzWkVdf+56fft5Ar17NKFY00JAsqanpbNp8mOXL/2T9HwdITk7L/NivczYx8oUHaVC/wm3X+fGnP0hNTadWrdLUreu+0UpyY8GF/Rn2eHs+HjufSZOWcu+9NQku7A/AocPnePudH3E4HPTs0YTef3XVcAeTycSQwW3UuSOP69e3JT//spETJy7y9TfLmTtvC3HxydSuVYYP3/9Hvr03i4iIyN3nP//5Dz/99NN1x6tUyShg9vT0JDw8nGHDhjFq1Kgs5yxZsoTQ0NDrrvvzzz+ZPn06r7766g33fOihhxgzZgwffPABXl43/72oevXq+Pr68uKLL3L69GmsViuVKlXim2++4R//+AcOh4OhQ4fSuHFjnn32WQDat2/Ps88+y6BBg9i1axf+/v45+XRkv8DjkUce4bPPPuOee+5h2bJl1KpViz59+vD888+zcuVKli1bRtu2bXO0uYiIiIiIiIgUHOnpNsaNXwRA717NCAsLNjRPj+5N2LLlKGvX7WfUOz8x9dtnsVothmbKC+x2O2Pe/5WUlDQaNqjAe+8OyGypm5iYwqlTlzh56iInTlzg5KmLnDxxkdNnLpOSksbpM5cBCPD3plLlMCpVCqVypTAqVw6lTOli+bJTx620bFGV2rXK8Oeek3z77QpefaWH2/ZOT7exbfsxlq/4k7Vr9xMfn5z5sbCwYJo0rsTKVXuIiLjA8Ocm07ZtLYY/04nixQvdcL3Y2ETmzNkMwOB/tLmjNsviPN27N2H+wu0cPnyOLycs4Z9v9OLipVheefU7kpPTaNy4EiOe76Kvl1zH39+bgQNaMfGr3/nu+9UAVK9Wio8/GoKvr9XQbCIiIlJwTZ069bpjZcqUITn5v3+fadOmDQ6HI1tr3Wi9ay5dunTTj40cOZKRI0feds/y5cszadKkm65jMplYvnz5dcc/+eQTPvnkk5tedyvZLvCYNm0a77//PuPHj8/8BL7++utYLBbWr19Pz549+de//pWrECIiIiIiIiKS/81fsI2IExcoVMiXoUPuNToOJpOJ11/ryf79pzl58iITv1rK8891NjqW4ebO28rOXRF4e1t47dUeWZ4Y9vW1UrVqSapWzdrtJD3dRmTkVc6fj6ZkySKEhATpCWUyvseeerIjTz0ziQULt9OvX0vKlC7msv3sdju7d59g+Yo/WbV6L9HR/+22W6xYIG3vq027drWpVrUkJpOJYU+0Z9LXy/ht7hZWrNjDH38cZOiQe+nXt+V1oxl+mb2RxKRUKlYIccuoD8kes9mDl0Z25YknJ7Jo0Q46tK/DxIlLuXgxlrJli/Gft/sVuMIqyb7evZrx40/ruXo1gUqVQvnk44fx8/M2OpaIiIiI3IJHdk+8VpESHBxMWFhYxsUeHrzyyivMmzePTz75hMKFC+do8wkTJlC7dm0CAwMJDAykWbNmLF68OMueo0aNIiwsDB8fH9q0acO+ffuyrJGSksLw4cMpWrQofn5+dO3alTNnzuQoh4iIiIiIiIi4Vnx8Ml9/k/GqlUcfaUtAgI/BiTIUKuTL66/1BDLGT2zfcczgRMaKiormiy8y/m3myWEds91lxdPTTHh4URo2rEhoaGEVd/yPOnXK0rx5FWw2O5MmLXPJHikpaXw5YQnde37IM8O/Yc5vW4iOTiQoyI+ePZrw5RePM2f2Kzw3vBPVq5XK/PoEBvry0ovd+HbyM9SuVYbk5DQmfrWUQYM/Y8PGQ5nrJyam8NPPGwAYPFjdO/KamjVL82CXhgC8+NI0Dh46S6FCvvzfB0PyzL1W8iYfHy/efqsvvXo147OxjxAYqO8XERERkbwu2wUegNP/8laqVCnef/99tm3bxrZt27jvvvvo1q1bZhHHhx9+yCeffML48ePZunUrISEhtG/fnri4uMw1RowYwZw5c5g1axbr168nPj6eLl26YLNdP+tVRERERERERIzx3feriY5OoHTponTv1tjoOFk0a1YlM9O7783OMsqiIHE4HHz4f7+RmJRKrVql6fVQU6Mj5RtPDeuIyWRi1eq9rFm73+nrf/TxPKbPWMulS7EE+HvTpXMDPh37MPN+e42XXuxG3Trl8PC4+T8DVq4UxoQvn+Df/+pNkSIBnDlzmZdensYrr37HmbOXmTd/K7GxSZQqVYR729R0en65c0892ZGAAB9sNjsWi5n3Rw+iZEljx2DJ3aFhw4qMHPEgQUF+RkcRERERkWzI9ogWgMqVK9+2yOPKlSvZXu/BBx/M8v57773HhAkT2LRpE9WrV+fTTz/lzTffpGfPjFfSTJs2jRIlSjBz5kyGDRtGTEwMkydP5vvvv6ddu3YATJ8+nfDwcJYvX07Hjh1z8vBERERERETExVas+JN1fxykedPKtGxZTTPeC4hz567w409/ADD8mU55clzAs888wJatRzl37gqffb6QN994yOhIbrd4yU42bT6Ml5cnr7/W85YFAZIzFSqE0L9fS2b+sI4x78+mWtWSFC9eyClrL168g4WLtuPhYeLNN3rR9r5a141XyQ6TycT9HetxT8tqfDtlJT/9vIH1fxxky9ajmesNGtgKs1nfF3lRUJAfL7/UjU8/W8BzwztTp05ZoyOJiIiIiIgL5Ohve2+//TaFCjnnL59/Z7PZ+Pnnn0lISKBZs2ZEREQQFRVFhw4dMs+xWq20bt2aDRs2MGzYMLZv305aWlqWc8LCwqhZsyYbNmy4aYFHSkoKKSkpme/HxsYCkJaWRlpamksen0hBdO3nST9XIuJKuteIiLvofnPnjh2L4u3//Ex6uo2lS3dhtVpo1rQybe+rSZMmlbBaLUZHFBf5csIS0tJsNKhfnkaNyufJnyOLxYPXX+vOc89PYeGi7TRvVpmWLasaksWI+83lK3F89vlCAIYOaUPJsMJ58ut0N3vk4TZs336MQ4fPMeqdH/nkoyF3XCxx4uRF/u/juQAMGdyGdm1rAo47+tp5eZl5clh77r+/Lp+PW8T27cdJTU2nWNFA2t5XU98XeVjrVtVo3aoakP37h36/ERF30f1GxHj6+RPJH3JU4NGvXz+KFy/u1AB79uyhWbNmJCcn4+/vz5w5c6hevTobNmTM9SxRokSW80uUKMHJkycBiIqKwsvLi8KFC193TlRU1E33HDNmDG+//fZ1x5cuXYqvr++dPiQR+Ztly1wzY1hE5H/pXiMi7qL7Te6kp9v5bvo+0tNtlCjhS2qqnatXk1m9Zh+r1+zDy8uDShWDqVatCGXLBOoV4vnI2bNxrFyVMZKidi1fFi9ebHCiW2vcKITNWyJ5b8zPPDK0Nn5+xhUeuet+43A4+G3uEeLikihRwpdCgbEsWrTILXsXNK1aFeN4RBS7dp1g1NuTada0ZK7XSkuz8d30fSQnp1GmdCBFiyQ4/evW9t6ihJc0s/vPCzRsGMLy5Uudur7kHfr9RkTcRfcbEeMkJiYaHUHczG63Gx1BculWX7tsF3jcbjRLblWpUoVdu3YRHR3N7NmzGTJkCGvWrLnpvg6H47ZZbnfO66+/zsiRIzPfj42NJTw8nA4dOhAYGJjLRyIif5eWlsayZcto3749FotejSkirqF7jYi4i+43d2bS18u4eDGRoCA/vprwNEFBfhw58v/t3Xd8zef///HHyd6JiCQiQhB709qjNqVUa5Tae2vR1qYtarS2VmvXLEpbe+9NETtmkBAjEmQn5/eHNt+f0k+N5JwknvfbzY3zPu9zXa8rybmcnPfzXFcoW7edYtv2U4SFRXD6zF1On7mLi4s9lSsVpHq1whQrllNhj3TMaDTSvccsAN6tV5K2bRuauaL/VqNGAp27zuTKlTBOBEbz1cj3Uu09kX9j6vlm+47TXAg6hKWlBaO/bkOePN6p3uebLHNmf74Zu5q9e0No1qwOhQtlf6V2xk34jbt3o3HP5MTEiZ3J7O6cwpU+8e67qdKspBF6fSMipqL5RsT8/t7RQDI+GxsbLCwsCAkJIUuWLNjY2Jj891p5NUajkbi4OO7cuYOFhQU2NjbPnPPCAQ+j0Ziixf3NxsaGPHnyAFC6dGkOHz7M5MmT+fzzz4Enq3RkzZo1+fywsLDkVT28vb2Ji4sjPDz8qVU8wsLCKF++/L/2aWtri63ts/s8W1tb64WFSCrQc0tETEFzjYiYiuabl3fy5DWWLN0LwOefvY+n55Pf3woVykGhQjno2aMup09fZ/PWk2zbFsj9+49Ys/Yoa9YeJXNmZ2rXKk6njjW0hUs6tHnLCc6cvYG9vQ1dOtdKF88da2trhg9rSsdO37N791m2bjtN3TolzFZLan/NHjx4nLw1S+tWVShQ4NXCBvLiGtR/i6NHL7N5y0m+HrWS+XN74eRk91JtbNx0nLVrj2EwGBgxvCneXu6pVK28KfT6RkRMRfONiPnouffmsLCwwN/fn9DQUEJCQsxdjrwCBwcH/Pz8sLB49kNPLxzwMNUSLkajkdjYWPz9/fH29mbz5s2UKPHkjZS4uDh27tzJ2LFjAShVqhTW1tZs3ryZpk2bAhAaGsqpU6cYN26cSeoVERERERGRfxcVFcuXXy8nKclIvXolqVK54DPnWFhYUKRIDooUyUGfXu/y5/ErbNlyku07TnHv3kMWL9lNZGQUgwZ+YIYRyKuKjY1nxvcbAfi4ZWU8PNLPipl5A3zo0L46M3/cxHcTf6dkCX+8vNzMXVaqmDRlLeHhj/H396RN63fMXc4bwWAwMKB/I06dvk5oaDjjJ6xmxPBmL/yJumvBdxg3fjUA7dq+Q+nSeVKxWhERERERSY9sbGzw8/MjISGBxMREc5cjL8HS0hIrK6t//R3xhQMeqWHQoEHUrVuX7Nmz8/DhQ5YuXcqOHTvYsGEDBoOBvn37Mnr0aAICAggICGD06NE4ODjQokULAFxdXenQoQP9+vUjc+bMuLu7079/f4oUKUKNGjXMOTQREREREREBpk1fT0jIfby83Ojbu/5/nm9paUHpUrkpXSo3/T5twNZtgXz19QrWrD1KyZK5qFPbPCspyMtb9ss+bt9+gKenKx81r2jucl5ayxaV2LP3LKdPX2fU6JVMmtjuuZ+cSc/27j3Hpk3HsbAwMHjgB9jYmPVtojeKk5MdI4c3o1uPH9m85SRlyuSlXt2S//m42Nh4hg5dQnR0HCVL+NOubTUTVCsiIiIiIumRwWDQykkZkFnfmbh9+zatWrUiX758VK9enYMHD7JhwwZq1qwJwGeffUbfvn3p3r07pUuX5ubNm2zatAln5//bU3TixIk0atSIpk2bUqFCBRwcHPjjjz+wtLQ017BEREREREQE2L//PKt/OwTAkMEfvPQWBNbWVtSpXYJ2bZ+sKjB+wm9cC76T4nVKyrt//yELft4BQNcutbCze3bP2LTOysqSYUOaYGtrzZGjl/h11UFzl5SiHj2KYdyE1QA0b1aRggW1NYupFS7sR4f21QH49rvfCQ6++5+PmTxlLRcv3SJTJkdGDG+GpWXGCh2JiIiIiIjI/2bW3wJnz57N1atXiY2NJSwsjC1btiSHO+BJqmjEiBGEhoYSExPDzp07KVy48FNt2NnZMXXqVO7du0dUVBR//PEH2bPrTQkRERERERFzioiIYvQ3vwLQrGkFSpXM/cpttWtbjZIlcxEdHcfQoUuIjY1PqTIllcyavZWoqFjy589GrZrFzF3OK8ue3YOe3esAMH3GhjQXMIqLSyAw8Bq//3GYM2euv9T2utOmr+fOnUh8fTPTsUP1VKxS/pdWH1ehZAl/oqPjGDFyGfHxCf967uYtJ1j92yEMBgPDhzVNV9seiYiIiIiISMrQ2psiIiIiIiKSooxGI+MnrObevYfkzJmFrl1qvVZ7lpYWjBjWlNZtp3Lx0i2mTF3LgP6NUqZYSXGXLt/i9z8OA9C7V710v63J+++XYdeesxw+fJGvvlrOD993wcrKPKuGhoc/IvBUMIGB1zgZeI1z524SH/9/eym7uTlQpkxeypXJy9tvB+Dm5vjcdo4cuZj8PRr4ReN0ucJKRmFpacGwoU1p3XYK587fZOaPm+nZo+4z512/fpexY1cB0KZ1Vd5+K8DUpYqIiIiIiEgaoICHiIiIiIiIpKjNW06ybfupJxcuhzTF1vb193r18HBh+LCmfNpvHqtWH6JkiVxUr140BaqVlDZt2nqSkoxUrVqI4sX8zV3Oa7OwsGDwwA/4uPVkzpy9wc8Ld9KubbVU79doNHL1ahhnzoYQeOoagSevcf3GvWfOc3NzJJe/J+fOh/DgQRQbNx5n48bjGAwGChbwpWzZvJQrl4/8+XywsLAgKiqWb/4KCjR+vwwliqf/71F65+npysAvPmDgoIUsXrKbt97KQ5m3/y/AERsbz5BhS4iKjqN48Zy0b5f6P38iIiIiIiKSNingISIiIiIiIinmzp0Ivv32NwDatX2H/PmzpVjbZd4OoNXHVVjw8w7GjF1FvnzZ8PXNnGLty+s7cOACBw8FYWVlSfdudcxdTorx9HTl008a8OVXy5kzdxvly+cnX16fVOnrxs17TJmyliNHg4iJOfTM/f7+nhQpnIOiRXNQtEgOsmVzx2AwkJCQSGDgNfYfuMD+Axe4dOkWp89c5/SZ68yesxU3N0fKlAkgNiaekNBwvLzc6JaBvkfpXZXKBXm/URlWrT7IV18vZ8H83rhncgJgytR1BAWF4ubmwMjhzcy2goyIiIiIiIiYnwIeIiIiIiIikiKMRiOjxvzKw0cxFCjgS+tWVVO8j44dqnPixBVOnLzG0GFLmPlDV2xs9KttWpCQkMjUaesAaPJhOXyzZazwTe1axdm1+ww7dpzmy69+Yc6sHimyOs3/LyEhkcFDFhMUFAqAra01hQr6UqRIDooUyUHhQn64uNg/97FWVpaUKJGLEiVy0b1bHcLCIjhw8AIHDlzg0OGLPHjwmI0bjyef//lnjXB0sE3R+uX19O5Vj+MnrnDlShijRq9kwrjWbNsWyKrVBwEYNqQpWbK4mrlKERERERERMSe9CyYiIiIiIiIp4tdVBzl0KAgbGyuGDvkwVT5lbmVlycgRzWnTbirnL4QwbcZ6Pu3bIMX7kZf3x5ojXLkahqurA23bvGPuclKcwWBgQL+GnDhxlStXwvhp1hZ69qibon0sXbaXoKBQnJ3tadjAn3btPsTe3u6V2vL0dOW9Bm/xXoO3iI9PIDAwmP0HLvDnn5cpWzYvZcvkTdHa5fXZ2lrz5YjmdOg0g/37zzN12np+/+MwAK1bVaFsWX3PRERERERE3nQW5i5AREREREQytsjIKDZs/JPBQxbT/KPvOHQ4yNwlSSq4fv0u06avB6BH9zrkzOGZan15eroyZPCHAKxYsZ+dO0+nWl/yYh49iuGnWVsA6NC+Os7Oz19lIr3LlMmJLz5vDMCSpXs4fuJKirV94+Y9Zs1+8jXs0b022bI5p1hIytraipIlc9Gjex1m/dSdjh1qpEi7kvJy5/am11/BoaXL9hAVFUuxojn0PRMRERERERFAK3iIiIiIiEgquHHzHrt3n2XPnrOcDLxGYmJS8n3DRyxjwfzeZPFwMWOFkpISEhL58uvlxMbGU7pUbj5oXDbV+6xQPj8tPqrE4iW7GT1mJXnz+pA1a6ZU7zejSUxM4uHDaB5EPObBg8dEPIjiQUQUERGPiYyMJiYmjpjYeGJj44mNiScmJv7p27HxT86JiScuLgE/Pw8aNXzb3MNKVZUqFqD+u6VYs/YoX3+9gvnze7/2VidGo5Fx41YTF5dA6VK5qVO7OOvXh6ZQxZLeNG5cloOHL7Jnz1lcXR0YOaJ5qqyIJCIiIiIiIumPAh4iIiIiIvLakpKSOHPmBrv3PAl1XLka9tT9uXN7U7FCfvYfuMCFCyF8+dUvTPquPZaWWlQwI1i4aBenT1/HycmOwYM+wMLCNN/Xrl1qceLkVU6fvs7QYUv4fkZnrK31a+7zPHwYzcJFuwgOvsODB4+fCnEkJRlTpA9LSws+6dvgjbgQ3af3uxw5eomQ0HCmTl3HF5+//1rtrVt/jCNHL2FjY8VnAxphMBhSqFJJjwwGA0MGfcjPC3fwTtXCeHq6mrskERERERERSSP0zpeIiIiIiLyShIREDhy4wO49Z9m77xz37z9Kvs/S0oLixXNSqWJBKlbIj4+POwB16pSgXftpHD16mUWLd9O6VRVzlS8p5PyFEGbP2QrAJ30b4OXlZrK+raws+XJEc9q2m8qZszf4YeYmevWsZ7L+04ukpCRGjFzG/gMX/vUcZyc7XN0ccXV1IJObI66ujjg722Nvb42trQ12dtZP/thaY2tng62t1V+3n9xna2eNi7M9jo52JhyZ+Tg62jFk0If07D2L3/84TKVKBahQPv8rtXX//kOmTF0HQMcONfD1zUx8fHxKlivpkIuLPT261zV3GSIiIiIiIpLGKOAhIiIiIiIv7d69hwwZupgTJ68lH3N0tKVc2XxUrFiAsmXy4uJi/8zjcvhl4dO+DRj9za/8NGszpUvlomDB7KYsXVJQZGQ0w4cvJTExiapVClGndnGT15A1ayYGDfqQgYMWsmTpHkqWyEWFCq92oT2jmr9gB/sPXMDGxopuXWuTJYsLbq6OuLo5PPnb1eGNWHUjpZUsmYtmzSqwbNlevhm7ioUL+uDq6vDS7UyavJaHD6MJCMhK82YVUqFSEREREREREckoFPAQEREREZGXcubMdQYOXsSdO5E4OtpSt25JKlUoQPHiOV9oe4x33y3FwcNBbN0ayLARy5g/t+cb86n/jCQhIZEhQxcTfP0unp6uDOjf0GzbSlSpXJCmTcrzy/J9fDVqOfPn9jLpSiJp2eHDF5k1+8kKK/37NaT+u6XMXFHG0rVzLQ4evMDVq3cYP2E1X3350Us9D/btP8+WrSexsDAw8PPGCtqIiIiIiIiIyP+kDa9FREREROSFrVl7lO49f+LOnUhy5MjCrJ+682nfBrz1Vp4XCncAGAwGPuvfCG9vN0JC7jPhu99TuWpJaUajkW+/+50jRy9hb2/D+LGtyZTJyaw1de9Wh/z5shEZGc3Q4UuJjdUWF2FhEQwfuRSj0UiD+qUV7kgFtrbWDB3SBEtLC7ZtP8XmLSdf+LGPo2IZP341AM2aViB//mypVKWIiIiIiIiIZBQKeIiIiIiIyH9KSEjku4m/M3rMSuLiEqhUqQCzfuxGDr8sr9Ses7M9I4c3w9LSgo0bj7N+w58pXLGkpl+W7+O33w9jMBgYObwZAQFZzV0SNjZWfPXlRzg52XHqVDBfj15BUlKSucsym/j4BIYMW8KDB1EEBGTl008amLukDKtAfl/atnkHgG+/+507dyJe6HE//riJ22ERZM2aiY4daqRmiSIiIiIiIiKSQSjgISIiIiIi/9P98Ef07jubFSsPANChfXXGjGr52tuqFCmSg/btqgHw7be/cePGvdeuVVLf3r3nmDJ1HQA9e9SlYsUCZq7o/2TL5s7or1tgaWnB1q2B/DBzk7lLMpvpMzZw6lQwTk52jP66Jba21uYuKUNr07oq+fNn4+HDaEaP+RWj0fg/zz99+nrynPrZgEbY29uYokwRERERERERSecU8BARERERkX917txNOnSYzvHjV3FwsGXsN63o0L46FhYp86tE61ZVKV48J1HRcQwfuYz4+IQUaVdSx8WLoQwf8WTLj/cavEXzZhXMXdIzSpfOw8AvGgOwcNEuVq8+aOaKTG/r1pP8snwfAEOHNCFbNnczV5TxWVlZMmxIE2xsrDh4KIjVvx3613Pj4xP4ZtwqjEYjdWqXoMzbASasVERERERERETSMwU8RERERETkudavP0bX7jO5HRaBX3YPZv3UjUopvFqDpaUFw4c2xdnZnrNnb/DTrC0p2r6knHv3HjLg85+Jio6jVKlc9O/3HgaDwdxlPVe9uiXp2KE6ABO++519+8+buSLTuXotjDHf/ArAxy0rp/hzVv5dzpyedOtaG4Cp09b966pEixbv5tKlW7i5OdC7Vz1TligiIiIiIiIi6ZwCHiIiIiIi8pSEhEQmTV7DV6NWEBeXQIXy+Zn1U3dy5vBMlf68vNwY9NeKC4sW7+bw4Yup0o+8utjYeL4YuJDbtx/gl92DUV+1xMrK0txl/U/t2lajXr2SJCUZGTpsCefO3TR3SakuOjqOwUMWExUdR8kS/nTuVNPcJb1xmnxYjpIl/ImJieerUctJTEx66v7g4LvMm78dgD696+Pm5miOMkVEREREREQknVLAQ0REREREkoWHP6Lvp3OTt3do1/Ydxn7zMU5Odqnab5UqhWjU8G2MRiNffr2c8PBHqdqfvDij0cjXo1dw+sx1nJ3tGT+uNS4u9uYu6z8ZDAa++Ox93iqdh+joOAZ8voDQW+HmLivVGI1Gxo5fzZUrYXhkdmbkiOZpPoSTEVlYWDB48Ic4ONgSGBjM4iW7k+9LSkpi7LhVxMUlUObtAGrVLGbGSkVEREREREQkPVLAQ0REREReSWxsPEaj0dxlSAo6ceIqHTrN4NixyzjY2zBmVEs6dayJhYVpfm3o3ase/jk9uXfvIaPGrNTPVxoxe85Wtm4NxNLSgjGjWpI9u4e5S3phVlaWjPq6Bblze3Pv3kP695/Pw4fR5i7rf0pKSiIpKem/T/yHVasPsmnTcSwtLfjyy+ZkzuycCtXJi8jqnYm+feoD8NOsLVy8GArAH2uO8ufxK9jZWfPZgEZpdosjEREREREREUm7FPAQERERkZcSeCqYLwYupFqNEYyf8JsuwmcA16/fZdDgRXTr8SO3bj3A1zczP/3YjSpVCpm0Djs7G0aOaIaNjRX79p1nxcr9Ju1fnrVp03HmzN0GwGcDGlGyZC4zV/TynJzsmDCuNR4eLly5GsbAwYuIj08wd1nPuHHzHqNGr6TKO8P4sOkEZs3eQmjoi604cubsDSZPWQtAt661KV7MPzVLlRfwbr2SVKxYgISERL78ajmht8KZPmM9AJ071SRr1kxmrlBERERERERE0iMFPERERETkPyUlJbF7z1m6dptJl64/sGv3GYxGI6t/O8Tvfxwxd3nyih48eMx3k/6gxceT2LHzNBYWBt5r8BazfuyOv7+XWWrKkycrPXvUBWD6jA3Jn3wX0wsMvMbob34FoGWLSjSoX9rMFb06Ly83JoxvjYO9DceOXWbMN6vSTDjtxo17fD16BR+1mMjadUdJTEzi1q0HzJm7jQ+ajKdXn1ls2PgnMTFxz318REQUg4csJj4+kapVCvFR84omHoE8z5Mtghrh5ubAxUu36NBxOo8exZA/fzaafFje3OWJiIiIiIiISDplZe4CRERERCTtiotLYOOm4yxesptr1+4AT7Y8qFO7OM7O9ixZuoeJk/4gfz4f8uXLZuZq06er18K4ciUMO1tr7OyssbWzSf63nf2Tf9vaWmNpmXLZ7NjYeJav2M/8Bdt5/DgWgHLl8tGjWx1y5TJPsOP/90Hjshw6FMSevecYNmIpP8zoiouLvbnLeqOEhobz+cCFxMUlUKlSAbp2qW3ukl5b3gAfvv66BQM+W8CGjX+SNasbnTrWNFs916/fZd787WzafILExCdbspQrm5dWrapw504ka9Ye5ciRSxw9epmjRy/z7Xe/U6N6Ueq/W5qCBX0xGAwkJSXx5Ve/cPv2k5V3Bg38QNt+pCHu7s58NqARgwYv5sGDKCwtLRj4eeMUnc9FRERERERE5M2igIeIiIiIPCMyMprVvx1k+Yr93Lv3EHiyzUGjhm/TpEl5sni4kJSUxPXrd9mz9xyDhyxmzuyeugj/km7cvEeHjjOIjn7+J/P/fzY2VtjaWmNvb4N/Tk+KFPGjaJEcFCyYHQcH2xfqLykpic1bTvLDzE3cvv0AgICArPTqUZfSpfO8zlBSlMFgYNDAD2jddipXr96h7ydzmDSxvX6+/vLwYTRDhy8FoHKlAlSuVBAPD5cUadtoNHLp0i2Gj1zGgwePCQjIyvChTTPMBemyZfIyoH9Dvhm7irnztpPVOxP1TbwySXDw38GO4yQlPVlFpFy5fHRoV42CBbMnn1ezRjFu3XrA+g3HWLv2KCGh4fz2+2F++/0wOXNm4d16pYiIiGL/gQvY2loz+usWODnZmXQs8t+qVilMvbolWbf+GB+3rExAQFZzlyQiIiIiIiIi6ZgCHiIiIiKS7NatByxbvpc/fj9M1F+hA09PV5o2KU/D997C0fH/Lh5aWFgwZHAT2neYRkhoOF+PWs43Yz7GwiJjXAhObUajkW/GriI6Og5PT1cyuTkSExtPTEw8sbFxxMQ8+fff4uISiItL4OHDaMLCIjh4KAgAS0sL8uT2pkgRP4oUyUHRIjnw8nJ7pr+jxy4xbdp6zl8IAZ58X7t0rkXtWsXS5PfMzc2Rid+2pVefWZw7f5M+fWczeVJ7XFwczF2a2c1fsINDf33/Dx0KYsK3v1OoUHaqVC5ElcoFyZ7d46XaCwuL4NDhixw5cpHDRy4SHv4YAI/Mzowf2/qFA0TpxXsN3iI0NJz5C3YwdvxqPL1cefutgFTv9+q1MObN38GWLSeSgx0VyuenXbtqFCzg+9zHeHu70a5tNdq0rsqfx6+wdu0xtu84xdWrd5g+Y0PyeQP6NyRPHgUH0qovPn+fxo3LUiC/VroSERERERERkdejgIeIiIiIcPFiKIuW7GbLlpPJWwXkyuVFyxaVqVG9CNbWz3/Z6OJiz6ivW9Cl20z27D3HosW7afVxFVOWnm799vthjh27jJ2dNdOndiJbNvdnzklKSiIuLiE57BETG8ejhzGcP3+Tk4HXOBkYzO3bDzh/IYTzF0JYsfIA8CS88fcKH35+WVixYj97950DwMHBltatqtKsaXlsba1NOuaXlTu3N1OndKRX79mcvxBC775zmPKGhzxCQu6zfMU+AN5vVIYLF0I4feY6p08/+TPj+w34+3smhz3y5vV5ZsuOR49iOPbnZY4cucihwxcJDr771P12dtaUKJGL7l1r4+nparKxmVLnTjUJvfWATZuOM2jwYn6Y0TnVAhJXr4Yxb/52Nm85idH4JNhRsWIB2retRv4XvOBvYWFBqZK5KVUyN59+0oCt2wJZs/YIp09f58MPylKvbslUqV1ShpWV5b+GeEREREREREREXoYCHiIiIiJvKKPRyNFjl1m0aFfyahAAJUvm4uMWlSlTJuCZC8PPky9fNvr2qc+48auZ+eMmChXMTsmSuVKz9HTv9u0HTJu+HoAunWs9N9wBTy7q2tnZYGdn89TxwoX9+OCDcsCT1RcCA69xMvAagYHBBF0MJSwsgq1bA9m6NTD5MZaWFjRq+Dbt21UjUyanVBpZysudy5tpUzrSq88sLlwIoVef2UyZ1AFX1zcz5PH9zI3ExyfyVuk89O/3HgaDgTt3Iti1+yy7dp3h2J+XuXIljCtXnoQKvL3dqFypICWK+xN0MZRDhy9y9uyN5CAXgIWFgQL5fXnrrTy8VTo3hQv7/WuoK6MwGAwM+qIxd+9EcOzPK3zafz4zpnfCN1vmFO1n7bqjfDN2VfLXu1KlJ8GOfPlefSUHJyc7Gr73Fg3fe4tHj2JwdMxYK6yIiIiIiIiIiMi/y9jv2omIiIjIMxISEtmx8zSLFu1K3q7DwsLAO1UL06JFJQrkf/lPGTd87y0CA6+xfsOfDBuxlHlzeuLh4ZLSpWcIRqORcRN+IyoqlsKF/fjwr6DGq/L0dKV69aJUr14UgKioWM6eu8HJk08CHxcvhlK4sB9dutQih1+WlBiCyeXK5fVXyGM2QUGh9O7zZLsWNzdHc5dmUqdOBbN1ayAGg4GePeomB7CyZHHlg8Zl+aBxWSIjo9m77xy7d5/hwMEL3Lr1gF+W7+OX5fueaiu7b+a/Ah15KFkyF87O9uYYklnZ2FgxetTHdOsxkytXwujVezYzpnciq3emFGl/zZojjBm7CqPRSPny+ejcqSZ5A3xSpO2/OTnZ/fdJIiIiIiIiIiKSYSjgISIiIvKGiI6OY83aIyxdtpfQ0HAAbG2taVC/FM2aVvzXVSRehMFgYED/hpy/EMLly7cZNmIpUyZ1wMrKMqXKzzA2bDzO/v3nsba2ZNAXjbG0tEjR9h0cbJO3cshI/P2fhDx69p5F0MUnIY8pkzu8MSEPo9HI1OnrAKhXtyQBAc/fTsTFxZ66dUpQt04JYmLiOHT4Irt2neHM2evkyZ2Vt97KQ+nSuVMsxJDeubjYM2VSB3r0/Ing63fp1WsWM6Z3fu2taf5Yc4Qx3/wKwIcflOWTvg1eaEUkERERERERERGR/0UBDxEREZEM7n74I1au3M/KXw8QGRkNgJubAx9+UI7G75dNsQvkdnY2jP66Je07Tuf48av8+NNmunerkyJtZxT37j1k0uQ1AHRoX52cOT3NXFH6kjOnJ9OmdqRX79lcvHTrr+1a2qerLWde1fYdpwgMDMbOzprOnWq80GPs7GyoXKkglSsVTOXq0rfMmZ2ZOqUD3Xv+xM2b9+nVexbTp3V65VWIfvv9MGPHrQKgyYfl6NunvsIdIiIiIiIiIiKSIlL244IiIiIikmbcuHGP8RNW0/iDccydt53IyGiyZXOnf7/3+HXFZ7RvVz3FVz/w8/Ng0MDGACxctIvde86maPvp3bcTf+fhw2jy5vWhxUeVzF1OupQzx5OQh0dmZy79FfK4H/7I3GWlqvj4BGZ8vxGAFh9VIkuW11tdQp6VJYsrUyd3xNvbjes37tG776v9XK3+7VByuKNpk/IKd4iIiIiIiIiISIpSwENEREQkA9q0+QTNW3zHqtWHiItLoGABX77+6iOWLv6Uxu+Xxc7OJtX6rvZOEZo1rQDAV18v5+bN+6nWV3qybXsgO3acxtLSgkEDG2v7mteQwy8LU/8KeVy+fJtevWdl6JDHyl8PEBJyn8yZnRUMSkXe3m5MndwRT09Xrl69Q5++c4iIiHrhx69afZBx41cD0KxZBfr0flfhDhERERERERERSVEKeIiIiIhkMMHBdxk7bhVJSUbefjuA6dM68dOP3aj2ThEsLU3z8q97t9oULuzHo0cxDB66mNjYeJP0m1ZFRETx7Xe/A9Dq4yrkDfAxc0XpXw6/LEyb+mQbjStXwujVaxb37z80d1kpLjIyirnztgPQqWMNHBxszVxRxpYtmztTJ3dIXiGmzydzkre2+l9+XXWA8RN+A6B5s4r07llP4Q4REREREREREUlxCniIiIiIZCDx8QmMGLmM6Og4Spbw59vxbShR3N/kFxqtra34+suPcHNz4MKFECZOWmPS/tOayVPWEh7+GP+cnrRt8465y8kw/Pw8mD61I1myuHDlahg9e8/i3r2MFfKYO387Dx9GkyuXF+/WK2Xuct4I2bN7MGVyBzJlcuTChRA+7TeXx49j/vX8lSv3M+HbJwGuFh9VolfPugp3iIiIiIiIiIhIqlDAQ0RERCQDmfnjZs6dv4mLiz3DhzU12Yodz+Pp6cqI4c0wGAz8/sdh1q0/ZrZazGnf/vNs2PgnFhYGBg38ABsbK3OXlKFkz+7B9KmdkrfV6NVn1kttq5GW3bh5j5UrDwDQs0ddsz6f3zQ5c3oyZVIHXF0dOHP2Bp/2n0dUVOwz561YuZ9vJ/4BQMsWlejRvY7CHSIiIiIiIiIikmr0DqGIiIhIBnHwUBCLl+wGYNDAD8iSxdXMFcHbbwXQsUN1AMZP+I2dO0+buSLTevQohnHjVwPQtEkFChXKbt6CMihf38xMm9oxOeQxdNgSEhISzV3Wa/v++40kJCRS5u0AypbJa+5y3ji5c3szaWJ7nJ3sCAwMZsBnC4iJiUu+/5fl+/jur3DHxy0r072bwh0iIiIiIiIiIpK6zBrwGDNmDG+99RbOzs54enrSqFEjzp8//9Q5RqORESNG4OPjg729PVWrVuX06acvDMTGxtKrVy88PDxwdHTkvffe48aNG6YcioiIiIhZ3Q9/xFdfLwfg/UZlqFypoJkr+j9tWlelXLl8xMbGM3DwIr4etYJHj/59u4OMZPqM9YSFRZAtmzudO9UwdzkZmm+2zHw7vg329jYcOXqJqdPWmbuk1xIYeI3tO05hYWGgZ4+65i7njZUvrw8TJ7bH0dGWP49f4fMvFhIbG8+yX/YyafKTradat6pCt661Fe4QEREREREREZFUZ9aAx86dO+nRowcHDhxg8+bNJCQkUKtWLR4/fpx8zrhx4/juu++YNm0ahw8fxtvbm5o1a/Lw4f/trd23b19WrVrF0qVL2bNnD48ePaJ+/fokJqb/T+2JiIiI/Bej0cio0Su5f/8R/v6e9O5Vz9wlPcXCwoIxo1ryccvKWFgYWLf+GK1aT+bI0UvmLi1VHTl6id9+PwzAwM8bY2dnY+aKMr7cub0ZNqQJAMtX7GfNmiMm6zshIZHbtx8QGHiNLVtPsmjxLiZNXsOatUdJSkp6qbaMRiNT/gqovPtuKXLn9k6NkuUFFSzgy3cT2mJvb8PhIxdp12Eak6esBaB1q6p06VxL4Q4RERERERERETEJs24AvmHDhqduz507F09PT44ePUrlypUxGo1MmjSJwYMH07hxYwDmz5+Pl5cXixcvpkuXLkRERDB79mx+/vlnatR48qnIhQsXkj17drZs2ULt2rVNPi4RERERU/pl+T727z+PjY0VX45ojq2ttblLeoaNjRXdu9WhYoUCfPn1ckJC7tO7z2yaNilPt66102TNryM6Oo5vvvkVgPcbvU3JkrnMXNGbo0qVQnTsUJ1Zs7cybsJv5MiRhSJFcqRI27dvPyAo6D6/rjrI3bsPuX07grCwCG7ffsDdew9JTHx+kGPDhmMMGvgBPj7uL9TPtm2BnD59HXt7Gzp10MovaUGRIjmYML4Nn/abx9WrdwBo2+YdOnWsoXCHiIiIiIiIiIiYjFkDHv8UEREBgLv7kzc+r1y5wq1bt6hVq1byOba2tlSpUoV9+/bRpUsXjh49Snx8/FPn+Pj4ULhwYfbt2/fcgEdsbCyxsbHJtyMjIwGIj48nPj4+VcYm8ib6+/mk55WIacXGxnPixDXi4xNe+DG2dtYUL5YTKyvLVKwsdbzpc82FoFBmfP8kNNu9ay38/DKn6a9FgQI+zJ7Vle+/38Tvfxzhl+X7OHDwAoMHNSZ/vmzmLi/F/DBzAyGh4Xh6utKpY/U0/T3JiFq2qEhQUCg7d51h4KBFzPyhM56erq/cntFoZNXqQ0ybvuGvEEfQc8+ztLQgSxYXPD1d8czigrOzPRs2HOfYn1do1WYK3bvVpkH9Uv8zEBAXl8CMHzYC0LxZBVxd7fXzk0YULuTLN6NbMGXaemrWKEqLjyqSkPDi/9eKvIw3/fWNiJiO5hsRMRXNNyLmp+efSMZgMBqNRnMXAU/eNG3YsCHh4eHs3r0bgH379lGhQgVu3ryJj49P8rmdO3fm2rVrbNy4kcWLF9OuXbunAhsAtWrVwt/fn5kzZz7T14gRIxg5cuQzxxcvXoyDg0MKj0xERMR0wsNj+HX1Be7ejX7px2bObEeNajnJmfPVL4KKacXFJTL/51Pcvx9DnjxuNG6UN119kvzS5Qds2HCZR4/jMRigfLlslCvrg6WlWXcRfG03bj5k0eIzADT5MB+5/N3MW9AbKi4ukYWLz3DnThTeXo60+KgA1tYvH2JLSEhi4+YrnDp1F4AsHvZkymSHs4stLs42uLjY4Oxsg4uLLY4O1lhYPP0cDA+PYd2Gy9y48WSLSf+crtSp44+Ls+1z+zt0OJTtO4JxcrKmU4di2Nikv+CdiIiIiIiIiKQ9UVFRtGjRgoiICFxcXMxdjoi8ojSzgkfPnj05efIke/bseea+f16oMBqN/3nx4n+dM3DgQD799NPk25GRkWTPnp1atWppQhNJQfHx8WzevJmaNWtibZ2xlt4XSYsOHgpi+vcrePQoBldXB7Jle7GtAABu3rjPvXtRLFt+jipVCtKjW228vNxSr9gU9CbPNeMm/Mb9+zF4eDjz7YTOuLk6mrukl9a6VRQTJ61h+47T7N13k7v3jAwe1JgcfllStV+j0UhMTDwREVFERkYRERFFRGQ0ERGP/zoW/eRYRBQRf90fG/tin3KIiX5yXu1axejZo3FqDkP+Q5myFenS9Udu3X7M8ZMxDBvy4UuFoMLuRDB02DLOnbuLhYWBTh2r4+oSSa1atV5qvvnooyRWrDzAT7O2cuVqBD//fI5evepSu1axp+qJiIhi+veTAejR/V3q1S3x4oMVkQzlTX59IyKmpflGRExF842I+f29o4GIpG9pIuDRq1cvfv/9d3bt2oWvr2/ycW9vbwBu3bpF1qxZk4+HhYXh5eWVfE5cXBzh4eFkypTpqXPKly//3P5sbW2xtX32E3PW1tZ6YSGSCvTcEkldRqORnxfuYuaPmzAajRQu7Mfor1vg4fHiocWHD6OZNXsLK389wM6dZzhwIIjWrarQ4qNK2Nqmj+evueeapKQk9u07z+2wCKq9U5hMmZxStb9t2wNZu/YYBoOB4cOaksXDLVX7Sy0eHq6M+rolm7ecYMKE3zh/PoSOnX6gW9faNPmwHBYWKb+ax61bDxg6fAmnT19P8bb/5uXpyid9G+j/PzPzy+7J6K9b0LvvHLZtO0XegGy0blXlhR574sRVBg1ZRHj4Y1xc7Plq5EcUL56DdevWvdJ807JFFSpUKMDXX6/gzNkbjPlmFbv3nOPzAY3InNkZgIWLdvPoUQx5cntT/93S6X41GxF5feZ+fSMibw7NNyJiKppvRMxHzz2RjMGsAQ+j0UivXr1YtWoVO3bswN/f/6n7/f398fb2ZvPmzZQo8eTTa3FxcezcuZOxY8cCUKpUKaytrdm8eTNNmzYFIDQ0lFOnTjFu3DjTDkhERMTEoqPjGD1mJVu3BQLQ8L23+KRvA2xsXu6/eGdnez7p24AG9Uvz3aQ/OH78Kj/N2sLadcf4pE99KlTInxrlZwhJSUns3HWGuXO3cfHSLQCmTV/Pew1K81HzSnh7u6V4n7duPWDs2FUAtPq4MqVK5k7xPkytZo1iFC+Wk1FjfuXQoSAmT1nLnr1nGTakCVmypNy2QadPX+fzgT9z//4jAKytLXF1dcTN1QEXVwfcXB1wdXXA1dXxr78dcHVxwNXNEXs76xde/cHT0xUHh+dvwSGmVaJELj7pW58J3/7OzB83kTuX1/+c04xGI6tWH2TipDUkJiaRJ7c334z5GB8f99feqzZnDk9++L4Li5fsZtbsrezZc5bAwGv0//Q98ub1YeWvBwDo1bOewh0iIiIiIiIiIiLyDLMGPHr06MHixYv57bffcHZ25tatJxdFXF1dsbe3x2Aw0LdvX0aPHk1AQAABAQGMHj0aBwcHWrRokXxuhw4d6NevH5kzZ8bd3Z3+/ftTpEgRatSoYc7hiYiIpKqbN+/zxaCFXLp0CysrSz7tW59Gjcq8Vpt58mRl+tRObN5ykmnT1xMScp8Bny+gXLl89O39Ltmze6RQ9elfUlISO3aeZu687Vz6K9jh4GBLNh93gi6GsnzFfn5ddZDatYvzccvK5MzhmSL9JiQkMvLLZTx8FEOhgtnp2CHjvN7JksWVid+2ZdXqg0ybvp6jRy/Tpt1Uhg1pStmyeV+7/c1bTjBq9Eri4hLIk9ubMaM/xscn00tt2SHpU+P3y3Lx4i1W/3aI4SOXMevHbuTM+exzMi4ugW+/+50/1hwBoHr1Igz64gPs7W1SrBYrK0tat6pK+XL5+GrUCoKCQhk6fCmZMjmSmJhEubJ5eeutPCnWn4iIiIiIiIiIiGQcZg14fP/99wBUrVr1qeNz586lbdu2AHz22WdER0fTvXt3wsPDKVOmDJs2bcLZ2Tn5/IkTJ2JlZUXTpk2Jjo6mevXqzJs3D0tLS1MNRURExKQOHQ5i6LClPHwYTebMzoz6qgVFi+ZIkbYNBgO1ahajQoX8zJu3nWW/7GX//vMcOXKRj5pXok3rqil6sTO9SUpKYvv2U8ydv53Ll28D4OhoS9Mm5WnWtALOzvYcOXKJBQt3cPToZdatO8b69X9SpUpBWn9clfz5s71W//MX7ODEyWs4ONgyYngzrKwy1usdg8FA4/fL8lbpPAwdvpQLF0L4tP88WreqSscO1V9pvEajkdlztjJn7jYAKpTPz4gRzXDUChtvlE/61ufK1TBOnLjKZ1/8zKwfu+PiYp98/527kQwavIjTp69jYWGga5fatGxRKdUCQHnyZGXWj92Yv2AH8xfsIDz8MRYWBnp0r5sq/YmIiIiIiIiIiEj6ZzAajUZzF2FukZGRuLq6EhERgYuLi7nLEckw4uPjWbduHfXq1dPebiIpxGg0smTJHmb8sIGkJCOFCmZn9KgWKbqFxT9dC77DxElrOHQoCHiy9UTvXvWo9k6RVOvzZZhqrklMTGL7jlPMnbuNK1fDAHBysqNpk/I0bVLhqQvFfzt9+joLFu5g9+6zycfefjuA1h9XoUQJ/5e+cHz8xBV69ppFUpKR4cOaUrtW8dcaU1oXGxvP1Gnr+HXVQQCKFcvJlyOavdTPe2xsPKNGr2TL1pMAfNS8It271dH2F2+o++GP6NBxBrdvP+DttwOYMK41VlaWBAZeY9CQxdy79xBnJztGjmxO2TLPrhqTWvPN2XM3mDV7K2XeDqBpk/Ip1q6IpF/6XUpETEXzjYiYiuYbEfPT9VCRjMGsK3iIiIjIi4uJiWPMN7+yecuTC9X13y1F/34NsbFJ3f/Oc/hlYeK3bdm1+yxTpq4lNDScIUOX8FHz62/EhfLExCS2bgtk3vxtXL16B3gS7GjWtAJNm5TH2fnZYMffChXKztgxrbh8+TYLF+1i85YTHDoUxKFDQRQqlJ3WrapStEgOIiKiiIh4TEREFA8iooiMiOJBxGMiIqOIePDXfZFRhIY+ICnJSJ3aJTJ8uAPA1taa/v0aUqK4P2PGruLEiatPtmwZ2vS5F9//6d69h3wxcCGnz1zH0tKCzwY0okH90iaoXNIq90xOjP3mY7p2m8mhQ0HM+H4jfn4efDfxDxISEsmVy4tvxnyMb7bMJq2rQH5fvh3fxqR9ioiIiIiIiIiISPqjgIeIiEg6EBoazheDFhIUFIqlpQV9e79L48ZlU23rgH8yGAxUqVyQsmUCmDN3Gz8v3MmSpXsIDr6bYbe6ePw4hm3bT7F4yW6uXXsS7HB2sqNZswo0+fB/Bzv+KVcuL4YNbULHDtVZvGQ3a9Ye5fTp63z+xc8vXVee3N706/feSz8uPatevSj58mVjyLAlT7Zs6fffW7YEBYXy2ecLuB0WgYuLPaO/bknJkrlMXLmkRXkDfBgy6EOGDFvC0mV7ko+/U7Uwgwd9gEMGnM9EREREREREREQkY1DAQ0RE5B+MRiMrfz1AbGw8jd8vi729jVnqiItL4Ny5mxw/cZUlS3cTERFFpkyOfP1VC0oU9zdLTba21nTrWps8ebwZNXole/edo2u3mYwb24qs3pnMUlNKSkhI5MiRS6zfcIydu84QF5cAgLOzPc3/CnY4Odm9cvs+Pu7079eQdm2rseyXvaxafZDHj2NxdLTF1cUBVzdH3FwdcHF1wM3VERcXe9zcHHF1dXjyx8WBnDk9/zXUkJH5+mZm5vddkrdsWfDzDk6cvPrcLVt27znLiJHLiI6Owy+7B+PHtSZ7dg/zFC5pUrVqRWh76Rbz5m/HYDDQpXNNWn1cxWShOREREREREREREZFXoYCHiIjIP/z402bmL9gBwIoV++nVqx7vVC2c6hf+Hj+OITAwmOMnrnIy8CpnztxIDhgA5M+fjTGjWuLl5ZaqdbyImjWK4ZPVnc8H/sylS7fo2GkGY8e0onBhP3OX9kouXbrFuvXH2LT5BPfuPUw+njNnFurVLcX7jd7G0fHVgx3/lDmzM9271aFzp5oYjUasrfWS7EX815YtRqORJUv2MP37DRiNRkqXys3XX7XAxeXFV1uRN0fHDtXx9c2Mb7bMFC2aw9zliIiIiIiIiIiIiPwnXU0QERH5/yxevDs53OHu7sTtsAiGDF1C6VK5+aRvffz9vVKsr7t3Izl58hrHT17lxImrXLp0i6Qk41PnuLk5UqxoDkqVyk2D+qWxtbVOsf5fV6FC2Zn1Y3c+/+Jngi6G0rP3LAYN/IBaNYuZu7QXcv/+QzZtPsH6DX8SFBSafNzV1YGaNYtRt04J8ufLlqrBnjdxJY6U8PeWLYOHLiYoKPSvLVuqEB7+mD/WHAGgUcO3+fSTBvoay7+ysLCgXt2S5i5DRERERERERERE5IUp4CEiIvKX3/84zLQZ6wHo2qU2TZuUY+GiXSxctIsjRy/Ruu1UPvywHB3aVX+lbTqMRiOXLt1iy9aT7NhxmuDrd585x8fHneLFclK0aA6KFcuJX3aPNL1lgLe3G9/P6MyIL39hz1/bYgQH36FD++ppru6EhESCg+9y7vxNtm4L5NChIBITkwCwtrakQoX81K1TkrJlArSiRjrg65uZH3/oypSp61i1+iALft4JgIWFgV4969G0Sfk09zMoIiIiIiIiIiIiIvI6dPVCRCQNW7P2KD/+tJnHj2Ne+DEWFgbKls1Lq4+rkDfAJxWry1i2bQtk7LjVALRsUZnWraoA0LFDDerWKcmUaWvZvfssy5btZfPmE/ToVofatYtjYWHxn20HB99ly9YTbNl6kqtX7yQfNxgM5MnjTbGiOShWNCdFi+Uki4dLqowvNTk42DJmVEt+mLmRRYt3M2fuNq5du8OQwR+abcWRmJg4Ll26zYWgEC5cCOFCUCiXLt16assbeLIKSd06JahRvSguLg5mqVVena2tNQP6N6RkiSdbtgB8NbI55crlM3NlIiIiIiIiIiIiIiIpTwEPEZE0KCEhkclT1rLy1wOv9PitWwPZujWQcuXy0frjKhQrljNlC8xgDhy4wIgvf8FoNNLwvbfo3q32U/dny+bO2DGtOHDgAhMn/cH1G/f4atQKVv12iH6fNCBfvmzPtBl6K5ytWwPZsvUkFy6EJB+3sbGiXNm8VK9elLJl8r7SSiBpkaWlBT261yWHXxbGjl/N1m2BhIaG882Yj/FI5dDK48cxXLsWwbJf9nLxUhhBF0K4Fnznme1uABzsbQgIyEqJErmoU7sEfn4eqVqbmEb16kV5++0AjEZwcbE3dzkiIiIiIiIiIiIiIqlCAQ8RkTTmfvgjhgxdzPHjVwHo0L46deuUeOHHh4c/Ztkve9m2PZD9+8+zf/95ihXNQetWVSlbNq+2LPiHkyevMXDwIhISEqlevQj9+zX8169R2bJ5+XlBH5b9spd587dz6lQw7TvOoFHDt+jcqRYJCYls2/4k1BEYGJz8OEtLC94qnYcaNYpSuVLBDBPqeJ769UuTLZs7Awcv4szZG3Ts/D3jxrZKtdVkNm06ztjxq4mOjgPOPXVfpkyO5A3wIW9eH/IGZCVvXh+yZXN/oVVXJP1xdlawQ0REREREREREREQyNgU8RETSkLPnbjBw0CLCwiJwcLBl+LCmVKpY4KXa8PFx58uRzencqSaLFu9i3fpjnDh5jX4D5hMQkJVWH1fhnaqFsbTURe4LQSH0/2w+sbHxlCubl2FDmvzn18XGxopWH1ehTu3iTJu+ns1bTrJq9SE2bjxOTGx88qoRBoOBEsVzUqN6UapWLYybm6MphpQmlCiRi1k/dqf/Z/MJDr5Lt+4/MnxYUypXKphifcTFJTB5ylpWrT4IgIuLDcWL5SZfvmxPAh15ffDI7KxAk4iIiIiIiIiIiIiIZBgKeIiIpBHr1x9j7PjVxMUl4Jfdg2+++ZicOTxfuT1f38x8/tn7tG9fnaVL97D6t0MEBYUybPhSsvtmpmXLytSpXQIbmzfzv4Lg4Lt88ulcHj2KoVixnIz6ugXW1i/+tciSxZWRI5rTqOHbfDvxDy5fvg1AoULZqVG9KNWqFSFLKm9Nkpb5+mbmxx+6MXTYEg4fucgXAxdS/91S9OxRFxcXh9dqOyTkPkOGLuHc+ZsYDAZafVwZL89o6td/F2tr6xQagYiIiIiIiIiIiIiISNryZl7VExFJQxISEpk2fT2/LN8HQIXy+Rk+rGmKbeORxcOFXj3r0aZ1VVas3M8vy/dx/cY9vhm7itlzttLio0p8+EG5N2pFj9u3H9Cn72zCwx+TN68P48e2xs7O5pXaKlEiF/Pm9CTwVDBenq74+LincLXpl4uLPd9OaJP8871m7VH27jtHn971qVmj6CutrrFnz1m++no5Dx/F4OrqwPChTSlVyp9169alwghERERERERERERERETSjjfnap6ISBoUHv6Ivp/MSQ53tGv7DmO/+TjFwh3/PxcXB9q3q86vKz6jd696ZMniwp07kUyespYRI5eRkJCY4n2mRffDH9HnkzncDovAL7sH333b9rW/3lZWlpQo7q9wx3NYWVnSt099vp/emZw5sxAe/pgRI5fRr/98QkLuv3A7CQmJzPh+A5998TMPH8VQqFB25s3pSdmyeVOxehERERERERERERERkbRDAQ8RETM5f/4m7TvO4NifV3Cwt2HMqJZ06lgTC4vUnZodHGxp3qwiy5f1p9+n72FlZcnWbYEMGryI2Nj4VO3b3B49iuHTfvMIDr6Ll6crkya2xz2Tk7nLeiMUK5aTeXN60aljDaytLTlw8AIft57M4sW7/zNcdPduJL37zmbhol0ANG1SnhnTOuHl5WaCykVERERERERERERERNIGBTxERMxg46bjdOk2k9u3H+Drm5mffuxGlSqFTFqDjY0VHzQuy7hvWmFjY8Wevef47POfiY6OM2kdphITE8eAzxdw4UIIbm6OTJ7UAW9vN3OX9UaxsbGiXdtqLJjfmxLF/YmJiWfajPV07Pw9Z8/deO5jjh67RNv20zh+/CoODrZ8/eVH9O1TH2tr7TInIiIiIiIiIiIiIiJvFgU8RERMKCEhkSlT1zHyy1+Ii0ugXLl8zP6pO/7+XmarqWzZvHw3oS329jYcPnKRTz6dy6NHMWarJyU9jorl6LFL/LxwJz17zeLEias4Otoy8bt2+Pl5mLu8N1YOvyxMm9qRQV80xtnZngsXQujU+XsmT11LVFQsAElJSSz4eQd9+s7h/v1H5M7tzZxZPahWrYh5ixcRERERERERERERETETffxVxASSkpLYtfsMK1bs50FE1Es9tlDB7LRoUYkcfllSqToxlbi4BIaPWMrOXWcAaNO6Kh071MDS0vxZu5IlczF5Yns+7T+Pk4HX6N13NhO/bYerq4O5S3thCQmJXLp0izNnb3DmzA3OnL3O1at3MBqNyefY2lozYVwb8uX1MWOlAmAwGKhfvzTly+dj8pS1bN5ykmXL9rJjx2l6dq/D+o1/sm/feQDq1S1J/37vYWdnY+aqRUREREREREREREREzEcBD5FUlJSUxI6dp5k7bzuXLt16pTYuX77NmrVHqVqlEK1bVSFfvmwpXKWYQkxMHF8MWsShQ0HY2FgxbEiTNLcSQeHCfkyd0pFPPp3DuXM36dHrJyZPbE/mzM7mLu0ZCQmJBAff5UJQCOfPh3D6zHUuXAghLi7hmXO9vd0oWMCXggWyU7FiAa3ckca4uzszckRz6tYpyfhvfyM0NJyhw5cCT7Z06ffpe9R/txQGg8HMlYqIiIiIiIiIiIiIiJiXAh4iqSApKYkdO04zZ942Ll++DYCDgy1NPixHqZK54AUvVMZEx/H7miPs2XOW7TtOsX3HKd5+O4A2rapQvLi/LnimE48exTDgs/mcOHkNOztrxn3TitKl85i7rOfKl9eH6VM70afvHC5fvk2Pnj8xeVJ7vLzczFZTbGw8ly7d4sKFEC4EhXLhQggXL916bpjD2cmOAgWz/xXo8KVAAd80GVCRZ5Utm5eFC/owe85Wlv2yF29vN0Z93YK8AVptRUREREREREREREREBBTwEElRiYlJbN9xirnztnHlShgAjo62NG1SnmZNK+Di8vLbXVSsWIBLl27x86JdbN16kkOHgjh0KIjChf1o06oq5cvnU9Djf0hKSmLL1kAuXbpF/vzZKFY0B+7uprvgHxERxSefzuXc+Zs4Odnx7fg2FCmSw2T9vwp/fy9mzOhM7z6zCb5+l249fmTK5A74Zsuc6n3HxMT9tRpHKBeCQrhwIYTg4LskJiY9c669vQ0BebKSN59P8god2bNn1vMhHbO3t6Fnj7q0bFEJJyc7rK31MkVERERERERERERERORvunIikgISE5PYtj2QufO2cfXqHQCcnOxo2qQ8TZtUwMXF/rXaz53bmxHDmtKpQw0WL9nF2nXHOHUqmAGfLyB3bm9afVyFau8UxsrKMiWGk2GcPHmNSVPWcO7czaeO+2X3oGjRHBQrlpNixXKSzcc9VUIBd+9G0ueTOVy5EoabmwMTv2tPvrzpYzUC32yZmTGtM336zub6jXt07/ETUya1J2dOz1Tr81rwHfr0nUNYWMQz97m5OZI3rw95A7L+9bcPvr7uWFhYpFo9Yj6ZMjmZuwQREREREREREREREZE0RwEPkdeQmJjE1q0nmTt/O9euPQl2ODvZ0bRpBZo2KY+z8+sFO/4pWzZ3BvRvRLu21Vj2yz5WrTrApUu3GDFyGT/+tJmPW1Ti3XdLvfGfer916wEzvt/Alq0ngSfb41SuVJCgi6Fcvnyb4Ot3Cb5+lzVrjwLgkdk5OfBRtGhO8uT2xtLy9YIDobfC6d1nNjdv3sfDwyXVwxGpwdvbjRnTO9O772yuXAmje8+fmDSxXapsmXEt+A69es3i7r2HZMrkSNEiOZKDHHnzZsXDw0Urc4iIiIiIiIiIiIiIiMgb7c2+CizyioxGI7t2n+WHmRufCnY0b16RJh+Wx8nJLlX79/BwoUf3OrT6uAorf93P8hX7CAm5z7gJv/HrqoMMHNiYAvl9U7WGtCg6Oo6Fi3ayaPFu4uISMBgM1K9fii6daiZvyxIZGU3gqWucOHGVEyevcvbsTe7ee8i27afYtv0U8GRbnTJvB/BuvVK8/XbAS4c9goPv0rvvbMLCIvDJmonJkzqQLZt7io/XFDJndmb61E588ulczl8IoVevWXz3bTsKFcqeYn1cvRZGr96zuXfvIblzezNlUnut4CAiIiIiIiIiIiIiIiLyDwp4SLoSHHyXw4eDcHNzxMvLDW9vN9zdnUy6TcPp09eZNn0dJ05eA8DZ2Z7mzSqYJNjxTy4u9rRrW43mzSry+x+HmTd/Oxcv3aJT5+/5qHlFOnaoga2ttUlrMoekpCQ2bT7B9z9s5M6dSACKF89Jn971n9kSxcXFngrl81OhfH4AYmPjOXP2xpPAx4mrBJ4K5vHj2OTAR5YsLtStU5L675bC1zfzf9Zy8WIofT6ZQ3j4Y3LkyMKUSe3JksU15QdtQm5ujkyd0pF+A+YRGBhMn76zGTq0KVUqF3zttq9eDaNn71ncv/+IPLm9maxwh4iIiIiIiIiIiIiIiMhzKeAh6cKdOxHMnrONteuOkpiY9NR9VlaWeHq64u3threXG15ernh7Z3ryt1cmvL3dsLF5/R/1kJD7fD9zI1u3BgJgY2PFR80r8nHLyjg6mjbY8U/29jY0a1qBWjWLMXHSGrZsPcmixbvZuesMAz9/nxIlcpm1vtR0+vR1Jk1ew+kz1wHImjUTPXvUpWqVQi+0pYetrTUlivtTorg/8GTbnfMXQti48U82bjrOnTuRLPh5Bwt+3kHx4jmp/25p3qlaGHt7m+fW8mm/uTx8FENAQFYmfdcuw4QVnJzsmPhtO74YuJAjRy8xcNBCmjWrQPeutV95S6ArV27Ts/cswsMfE5AnK5MntcfNzTGFKxcRERERERERERERERHJGBTwkDQtMjKahYt28svyfcTFJQBPVmZISjRy+/YD7tyNJCEhkZCQ+4SE3H9uG3Z21pR5O4CKFQtQvly+l77gHhkZxbz5O1j5637i4xMxGAzUrVuCzh1r4umZtlZmyJTJiS9HNqdmzWKMn/AbN27co0evWbzf6G26d6tj9iBKSgoLi+D7mRvZuPE4AA72NrRuXZVmTSu81qollpYWFCzgS8ECvvToXpc9e86yZu1RDh0O4vjxqxw/fpXvvvud6jWKUr9eKQoX9sNgMHDs2GU++3wBUdFxFCnix4RxbXB2tk+h0aYNDg62fDuhDT/M3MSSpXtYtmwvgSev8dWXH5E1a6aXauvy5dv06vNXuCMgK1MmdcDV1SGVKhcRERERERERERERERFJ/xTwkDQpNjae5Sv28/PPO3j4KAaAokVy0K1rbYoVy5l8XkJCInfvPuTW7XBu347g9u0H3L79gFu3HnDr9gNu33pAVHQcO3edYeeuM1hYGChS2I+KFQtQsWIBcvhl+dca4uISWPnrAebN387Dh9EAvFU6Dz161CFvgM+/Pi4tqFSxACWK+zNt+np+/+Mwq1YfYu++83w2oBHly+Uzd3mvJTz8Eb8s38eyX/YSExOPwWCgXt2SdOlcEw8PlxTty8bGimrVilCtWhHCwiJYv+FP1qw9ws2b9/njjyP88ccR/Pw8qFihACtW7icuLoHSpXIz9ptWz13hIyOwtraiV896FC/uz9ejVnDm7A3atpvK4MEfUrnSi23ZcunyLXr1ns2DB4/Jm9eHyRPbK9whIiIiIiIiIiIiIiIi8h8U8JA0JSEhkbXrjjF7zlbu3o0EwN/fk25dalOhQv5nttywsrJ8sjWLt9tz2zMajVy4EMLuPWfZs/ccFy6EcOLkNU6cvMb0GRvw8/OgUsUCVKxQgMKF/bC0tMBoNLJ1WyA//LCRkNBwAHLl8qJn97qUKRPwQtt+pAVOTnZ88fn71KhRlG/GriIk5D79B8ynVq3i9O39brrbCiMk5D5Llu7hjzVHkldzKVokB3371Cd//myp3r+npyttWleldasqHD9xlbVrj7JteyDBwXdZHLwbgIoVC/DVyOavtYJIelGpYgHmze3JsGFLOX3mOl8MfLEtWy5dukWvPk/CHfny+jB5UntcXBTuEBEREREREREREREREfkvCnhImmA0Gtmx8zQzf9xEcPBdALy83OjUsQa1axXH0tLildo1GAzky5eNfPmy0bFDDW7desDevWfZvfccx45dJjj4LosW72bR4t24uTlQvnx+rl29w+kz1wHwyOxMp041qVe35CvXYG6lS+Vm4YLe/PjTFn5ZvpdNm45z6FAQn37SgOrViqT5wEpQUCgLF+1k2/ZTJCYmAVCggC+tPq5MlcqFTF6/wWCgRHF/ShT355NPGrBtWyCbNh8nZw5P+vR+FysrS5PWY05ZvTMxY3onvv9hE0uXPdmy5VRgMF9+2Zys3s9u2XLxYii9+87mwYMo8ufLxqSJ7RTuEBEREREREREREREREXlBCniI2R09donvv9/ImbM3AHB1daBN66q836hMiq+E4O3txgcflOODD8rx+HEMBw5cYPfec+zfd44HD6JYt+4YAPb2NrRsUYmPmlfKEFtt2NnZ0LtXPapXK8Lob1Zy5UoYw4YvZdPmE/T/9D08PV1TpV+j0fjKjzv252UWLtzFwUNBycfffjuAj1tWplTJXGkimOLoYEuD+qVpUL+0uUsxG2trK3r3qkfx4jkZNWoFp89cp23bqQwZ0oRKFQsknxcU9CTcERERRf782Zj0XXtcXOzNWLmIiIiIiIiIiIiIiIhI+qKAh5jc/fBH/HnsMsf+vMzRv1bRALCzs6Z5s4q0+KgSTk52qV6Ho6Md1asXpXr1oiQkJHLy5DX27juHlZUlTT4sh4eHS6rXYGqFCmVn3pyezF+wgwU/72TPnrMcOXKRtm3eoXmzitjYpMyUcPzEFWb+uJkzZ65jb2/Jug2heHu54enphreXK55ebnh6uuLt5YaLi31yWCMxMYnde87w88JdnP0r8GNhYaDaO0Vo2bIy+fL6pEh9kvIqVypIwNxeDB22hDNnb/D5Fz/TvFlFunWtxZWrYfTpO4eIiCgKFPBl0nftcHZWuENERERERERERERERETkZSjgIakuIiKKP49f5tixJ4GOK1fCnrrf0tKChu+9Rbu21cic2dksNVpZWVKyZC5Klsxllv5Nydraio4davBO1cKMn/AbJwOv8cPMTaxZe5S+fepTvly+V2770qVbfD9zI/v2nU8+Fh+fSGBgMIGBwc99jK2tNV5ernh5unH79gOCrz8J/NjYWFH/3VJ81LwS2bK5v3JNYjpZs2bi+xmdk7dsWbpsDydOXuXmzXtERkZTsIAvExXuEBEREREREREREREREXklCnhIinv4MJrjJ65y7NiTUMfFS7ee2aojT27v5EBF8WL+2qrBDHLn9ub7GZ3ZuOk402ds4MaNe/QfMJ+KFfLTu/e7+GbL/MJthYaG89OsLWzcdByj0YilpQUN6pfm/UZvsXnLdnLnLsS9e48IC4vg9u0H3P7r7/Dwx8TGxhMcfDd5JRdnJzsaNy5Lkyblcc/klFrDl1Tyzy1b/l6JpVDB7Ez8rp1JVucRERERERERERERERERyYgU8JAU9+VXy9m779xTx/xzeiYHOkoU98fNzdFM1cn/z2AwUKd2CSpVLMDcedtZ9ste9uw9x6HDF2nZohKtPq6CnZ3Nvz4+PPwR8xfsYNXqg8THJwJQ7Z3CdO5UCz8/D+Lj48nm40y1dwpjbW39zONjY+O5czeS27cfEHY7AiNQpUohHB1sU2vIYiJ/b9kyfsJvWNtYMnRwE4U7RERERERERERERERERF6DAh6S4kqU8Cc4+A6lSuV+Euoo4Y+7u3m2XpEX4+hoR88edan/bikmTlrD4SMXmTtvO+vW/0nvXvWoWqUQBoMh+fyoqFiWLtvD4iV7iIqKBaB0qdx061abAvl9X7hfW1trfLNlfqnVQiT9yJo1E99929bcZYiIiIiIiIiIiIiIiIhkCBbm7HzXrl00aNAAHx8fDAYDq1evfup+o9HIiBEj8PHxwd7enqpVq3L69OmnzomNjaVXr154eHjg6OjIe++9x40bN0w4Cvmn5s0qsGxpPz4b0Iga1Ysq3JGO5MzpyaSJ7Rg9qgVeXm7cvv2AwUMW0/eTuVy9GkZ8fAIrVu6nSbMJzJq9laioWPLl9WHSxHZMmdzhpcIdIiIiIiIiIiIiIiIiIiLy4swa8Hj8+DHFihVj2rRpz71/3LhxfPfdd0ybNo3Dhw/j7e1NzZo1efjwYfI5ffv2ZdWqVSxdupQ9e/bw6NEj6tevT2JioqmGIf9gYWHWHyt5TQaDgapVCrNkUV/at6uGjY0Vh49cpFWbKTRp9i3fTfyD8PDH+Ppm5suRzZk9qztvvxVg7rJFRERERERERERERERERDI0s27RUrduXerWrfvc+4xGI5MmTWLw4ME0btwYgPnz5+Pl5cXixYvp0qULERERzJ49m59//pkaNWoAsHDhQrJnz86WLVuoXbu2ycYiktHY2dnQsUMN6tYpyeSpa9mz5yxhYRFkzuxMu7bv8F6Dt7CysjR3mSIiIiIiIiIiIiIiIiIibwSzBjz+lytXrnDr1i1q1aqVfMzW1pYqVaqwb98+unTpwtGjR4mPj3/qHB8fHwoXLsy+ffv+NeARGxtLbGxs8u3IyEgA4uPjiY+PT6URiaRPnp7OjPqqOUeOXuLGjXvUrlUce3sbjMYk4uOT/udj/34+6XklIqlJc42ImIrmGxExFc03ImIqmm9ExFQ034iYn55/IhlDmg143Lp1CwAvL6+njnt5eXHt2rXkc2xsbMiUKdMz5/z9+OcZM2YMI0eOfOb4pk2bcHBweN3SRTIsG2vYvn3LSz9u8+bNqVCNiMjTNNeIiKlovhERU9F8IyKmovlGRExF842I+URFRZm7BBFJAWk24PE3g8Hw1G2j0fjMsX/6r3MGDhzIp59+mnw7MjKS7NmzU6tWLVxcXF6vYBFJFh8fz+bNm6lZsybW1tbmLkdEMijNNSJiKppvRMRUNN+IiKlovhERU9F8I2J+f+9oICLpW5oNeHh7ewNPVunImjVr8vGwsLDkVT28vb2Ji4sjPDz8qVU8wsLCKF++/L+2bWtri62t7TPHra2t9cJCJBXouSUipqC5RkRMRfONiJiK5hsRMRXNNyJiKppvRMxHzz2RjMHC3AX8G39/f7y9vZ9arisuLo6dO3cmhzdKlSqFtbX1U+eEhoZy6tSp/xnwEBEREREREREREREREREREUlPzLqCx6NHj7h48WLy7StXrnD8+HHc3d3x8/Ojb9++jB49moCAAAICAhg9ejQODg60aNECAFdXVzp06EC/fv3InDkz7u7u9O/fnyJFilCjRg1zDUtEREREREREREREREREREQkRZk14HHkyBHeeeed5NuffvopAG3atGHevHl89tlnREdH0717d8LDwylTpgybNm3C2dk5+TETJ07EysqKpk2bEh0dTfXq1Zk3bx6WlpYmH4+IiIiIiIiIiIiIiIiIiIhIajBrwKNq1aoYjcZ/vd9gMDBixAhGjBjxr+fY2dkxdepUpk6dmgoVioiIiIiIiIiIiIiIiIiIiJifhbkLEBEREREREREREREREREREZH/TQEPERERERERERERERERERERkTROAQ8RERERERERERERERERERGRNM7K3AWkBUajEYDIyEgzVyKSscTHxxMVFUVkZCTW1tbmLkdEMijNNSJiKppvRMRUNN+IiKlovhERU9F8I2J+f18H/fu6qIikTwp4AA8fPgQge/bsZq5ERERERERERERERERERCR1PHz4EFdXV3OXISKvyGBUTIukpCRCQkJwdnbGYDCYuxyRDCMyMpLs2bNz/fp1XFxczF2OiGRQmmtExFQ034iIqWi+ERFT0XwjIqai+UbE/IxGIw8fPsTHxwcLCwtzlyMir0greAAWFhb4+vqauwyRDMvFxUUv2kUk1WmuERFT0XwjIqai+UZETEXzjYiYiuYbEfPSyh0i6Z/iWSIiIiIiIiIiIiIiIiIiIiJpnAIeIiIiIiIiIiIiIiIiIiIiImmcAh4ikmpsbW0ZPnw4tra25i5FRDIwzTUiYiqab0TEVDTfiIipaL4REVPRfCMiIpIyDEaj0WjuIkRERERERERERERERERERETk32kFDxEREREREREREREREREREZE0TgEPERERERERERERERERERERkTROAQ8RERERERERERERERERERGRNE4BDxEREREREREREREREREREZE0TgEPEflXu3btokGDBvj4+GAwGFi9evVT99++fZu2bdvi4+ODg4MDderUISgo6KlzqlatisFgeOpP8+bNnzonPDycVq1a4erqiqurK61ateLBgwepPDoRSUtMMd9cvXqVDh064O/vj729Pblz52b48OHExcWZYogikkaY6vXN32JjYylevDgGg4Hjx4+n0qhEJC0y5Xyzdu1aypQpg729PR4eHjRu3Dg1hyYiaYyp5psLFy7QsGFDPDw8cHFxoUKFCmzfvj21hyciaURKzDUA+/fvp1q1ajg6OuLm5kbVqlWJjo5Ovl/vFYuIiPxvCniIyL96/PgxxYoVY9q0ac/cZzQaadSoEZcvX+a3337jzz//JEeOHNSoUYPHjx8/dW6nTp0IDQ1N/jNz5syn7m/RogXHjx9nw4YNbNiwgePHj9OqVatUHZuIpC2mmG/OnTtHUlISM2fO5PTp00ycOJEffviBQYMGpfr4RCTtMNXrm7999tln+Pj4pMpYRCRtM9V8s3LlSlq1akW7du04ceIEe/fupUWLFqk6NhFJW0w137z77rskJCSwbds2jh49SvHixalfvz63bt1K1fGJSNqQEnPN/v37qVOnDrVq1eLQoUMcPnyYnj17YmHxf5eq9F6xiIjIfzCKiLwAwLhq1ark2+fPnzcCxlOnTiUfS0hIMLq7uxt/+umn5GNVqlQx9unT51/bPXPmjBEwHjhwIPnY/v37jYDx3LlzKToGEUkfUmu+eZ5x48YZ/f39X7dkEUmnUnu+WbdunTF//vzG06dPGwHjn3/+mYLVi0h6klrzTXx8vDFbtmzGWbNmpUbZIpIOpdZ8c+fOHSNg3LVrV/KxyMhII2DcsmVLio5BRNK+V51rypQpYxwyZMi/tqv3ikVERP6bVvAQkVcSGxsLgJ2dXfIxS0tLbGxs2LNnz1PnLlq0CA8PDwoVKkT//v15+PBh8n379+/H1dWVMmXKJB8rW7Ysrq6u7Nu3L5VHISLpQUrNN88TERGBu7t7yhctIulSSs43t2/fplOnTvz88884ODikfvEikq6k1Hxz7Ngxbt68iYWFBSVKlCBr1qzUrVuX06dPm2YgIpLmpdR8kzlzZgoUKMCCBQt4/PgxCQkJzJw5Ey8vL0qVKmWawYhImvUic01YWBgHDx7E09OT8uXL4+XlRZUqVZ6ai/ResYiIyH9TwENEXkn+/PnJkSMHAwcOJDw8nLi4OL755htu3bpFaGho8nktW7ZkyZIl7Nixg6FDh7Jy5cqn9oO+desWnp6ez7Tv6empJT5FBEi5+eafLl26xNSpU+natasphiEi6UBKzTdGo5G2bdvStWtXSpcubY6hiEgal1LzzeXLlwEYMWIEQ4YMYc2aNWTKlIkqVapw//59k49LRNKelJpvDAYDmzdv5s8//8TZ2Rk7OzsmTpzIhg0bcHNzM8PIRCQteZG55v9/3dKpUyc2bNhAyZIlqV69OkFBQYDeKxYREXkRVuYuQETSJ2tra1auXEmHDh1wd3fH0tKSGjVqULdu3afO69SpU/K/CxcuTEBAAKVLl+bYsWOULFkSePImwT8ZjcbnHheRN09Kzjd/CwkJoU6dOjRp0oSOHTuaZBwikval1HwzdepUIiMjGThwoKmHICLpRErNN0lJSQAMHjyYDz74AIC5c+fi6+vL8uXL6dKli+kGJSJpUkrNN0ajke7du+Pp6cnu3buxt7dn1qxZ1K9fn8OHD5M1a1ZTD01E0pAXmWv+ft3SpUsX2rVrB0CJEiXYunUrc+bMYcyYMYDeKxYREfkvWsFDRF5ZqVKlOH78OA8ePCA0NJQNGzZw7949/P39//UxJUuWxNraOjmV7e3tze3bt585786dO3h5eaVa7SKSvqTEfPO3kJAQ3nnnHcqVK8ePP/6Y2qWLSDqTEvPNtm3bOHDgALa2tlhZWZEnTx4ASpcuTZs2bUwyDhFJ+1Jivvn7gmrBggWTz7G1tSVXrlwEBwen7gBEJN1Iqdc3a9asYenSpVSoUIGSJUsyY8YM7O3tmT9/vqmGIiJp2H/NNc973QJQoECB5Ncteq9YRETkvyngISKvzdXVlSxZshAUFMSRI0do2LDhv557+vRp4uPjk1/QlytXjoiICA4dOpR8zsGDB4mIiKB8+fKpXruIpC+vM98A3Lx5k6pVq1KyZEnmzp2LhYVeConI873OfDNlyhROnDjB8ePHOX78OOvWrQNg2bJljBo1yiT1i0j68TrzTalSpbC1teX8+fPJ58THx3P16lVy5MiR6rWLSPryOvNNVFQUwDO/Q1lYWCR/Kl9EBP59rsmZMyc+Pj5PvW4BuHDhQvLrFr1XLCIi8t+0RYuI/KtHjx5x8eLF5NtXrlzh+PHjuLu74+fnx/Lly8mSJQt+fn4EBgbSp08fGjVqRK1atQC4dOkSixYtol69enh4eHDmzBn69etHiRIlqFChAvAkoV2nTh06derEzJkzAejcuTP169cnX758ph+0iJiFKeabkJAQqlatip+fHxMmTODOnTvJ/Xl7e5t2wCJiNqaYb/z8/J7q08nJCYDcuXPj6+tropGKiLmZYr5xcXGha9euDB8+nOzZs5MjRw7Gjx8PQJMmTUw/aBExC1PMN+XKlSNTpky0adOGYcOGYW9vz08//cSVK1d49913zTJuETGt151rDAYDAwYMYPjw4RQrVozixYszf/58zp07x4oVKwC9VywiIvJCjCIi/2L79u1G4Jk/bdq0MRqNRuPkyZONvr6+Rmtra6Ofn59xyJAhxtjY2OTHBwcHGytXrmx0d3c32tjYGHPnzm3s3bu38d69e0/1c+/ePWPLli2Nzs7ORmdnZ2PLli2N4eHhJhypiJibKeabuXPnPrcPvRwSebOY6vXN/+/KlStGwPjnn3+m8uhEJC0x1XwTFxdn7Nevn9HT09Po7OxsrFGjhvHUqVOmHKqImJmp5pvDhw8ba9WqZXR3dzc6Ozsby5Yta1y3bp0phyoiZvS6c83fxowZY/T19TU6ODgYy5UrZ9y9e/dT9+u9YhERkf/NYDQajamaIBERERERERERERERERERERGR16KN50VERERERERERERERERERETSOAU8RERERERERERERERERERERNI4BTxERERERERERERERERERERE0jgFPERERERERERERERERERERETSOAU8RERERERERERERERERERERNI4BTxERERERERERERERERERERE0jgFPERERERERERERERERERERETSOAU8RERERERERJ5jxIgRFC9e3OT97tixA4PBgMFgoFGjRv/z3KpVq9K3b98Xardt27bJ7a5evfq16xQREREREREREdNSwENERERERETeOH8HHf7tT9u2benfvz9bt241W43nz59n3rx5Kdbe5MmTCQ0NTbH2RERERERERETEtKzMXYCIiIiIiIiIqf3/QYdly5YxbNgwzp8/n3zM3t4eJycnnJyczFEeAJ6enri5uaVYe66urri6uqZYeyIiIiIiIiIiYlpawUNERERERETeON7e3sl/XF1dMRgMzxz75xYtbdu2pVGjRowePRovLy/c3NwYOXIkCQkJDBgwAHd3d3x9fZkzZ85Tfd28eZNmzZqRKVMmMmfOTMOGDbl69epL1/z48WNat26Nk5MTWbNm5dtvv33mnBkzZhAQEICdnR1eXl58+OGHL92PiIiIiIiIiIikTQp4iIiIiIiIiLygbdu2ERISwq5du/juu+8YMWIE9evXJ1OmTBw8eJCuXbvStWtXrl+/DkBUVBTvvPMOTk5O7Nq1iz179uDk5ESdOnWIi4t7qb4HDBjA9u3bWbVqFZs2bWLHjh0cPXo0+f4jR47Qu3dvvvzyS86fP8+GDRuoXLlyio5fRERERERERETMR1u0iIiIiIiIiLwgd3d3pkyZgoWFBfny5WPcuHFERUUxaNAgAAYOHMg333zD3r17ad68OUuXLsXCwoJZs2ZhMBgAmDt3Lm5ubuzYsYNatWq9UL+PHj1i9uzZLFiwgJo1awIwf/58fH19k88JDg7G0dGR+vXr4+zsTI4cOShRokQKfwVERERERERERMRcFPAQEREREREReUGFChXCwuL/FsP08vKicOHCybctLS3JnDkzYWFhABw9epSLFy/i7Oz8VDsxMTFcunTphfu9dOkScXFxlCtXLvmYu7s7+fLlS75ds2ZNcuTIQa5cuahTpw516tTh/fffx8HB4aXHKSIiIiIiIiIiaY8CHiIiIiIiIiIvyNra+qnbBoPhuceSkpIASEpKolSpUixatOiZtrJkyfLC/RqNxv88x9nZmWPHjrFjxw42bdrEsGHDGDFiBIcPH8bNze2F+xIRERERERERkbTJ4r9PEREREREREZFXUbJkSYKCgvD09CRPnjxP/XF1dX3hdvLkyYO1tTUHDhxIPhYeHs6FCxeeOs/KyooaNWowbtw4Tp48ydWrV9m2bVuKjUdERERERERERMxHAQ8RERERERGRVNKyZUs8PDxo2LAhu3fv5sqVK+zcuZM+ffpw48aNF27HycmJDh06MGDAALZu3cqpU6do27btU9vFrFmzhilTpnD8+HGuXbvGggULSEpKemobFxERERERERERSb+0RYuIiIiIiIhIKnFwcGDXrl18/vnnNG7cmIcPH5ItWzaqV6+Oi4vLS7U1fvx4Hj16xHvvvYezszP9+vUjIiIi+X43Nzd+/fVXRowYQUxMDAEBASxZsoRChQql9LBERERERERERMQMDMYX2chXRERERERERExix44dvPPOO4SHh+Pm5pbi7RsMBlatWkWjRo1SvG0REREREREREUk92qJFREREREREJA3y9fXlo48+SrH2unbtipOTU4q1JyIiIiIiIiIipqUVPERERERERETSkOjoaG7evAmAk5MT3t7eKdJuWFgYkZGRAGTNmhVHR8cUaVdERERERERERExDAQ8RERERERERERERERERERGRNE5btIiIiIiIiIiIiIiIiIiIiIikcQp4iIiIiIiIiIiIiIiIiIiIiKRxCniIiIiIiIiIiIiIiIiIiIiIpHEKeIiIiIiIiIiIiIiIiIiIiIikcQp4iIiIiIiIiIiIiIiIiIiIiKRxCniIiIiIiIiIiIiIiIiIiIiIpHEKeIiIiIiIiIiIiIiIiIiIiIikcQp4iIiIiIiIiIiIiIiIiIiIiKRxCniIiIiIiIiIiIiIiIiIiIiIpHH/D5eeLH6mPvKBAAAAAElFTkSuQmCC",
"text/plain": [
"<Figure size 2400x350 with 1 Axes>"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Y_hat_df = nf.predict(futr_df=Y_test_df)\n",
"plot_series(Y_train_df, Y_hat_df, palette='tab20b')"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "python3",
"language": "python",
"name": "python3"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
nbs/examples/Multivariate_with_TSMixer.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/Neuralforecast_Map.ipynb 0 → 100644
View file @ e3f7f7b3
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# NeuralForecast map\n",
"> Modules of the NeuralForecast library"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"The `neuralforecast` library provides a comprehensive set of state-of-the-art deep learning models designed to power-up time series forecasting pipelines.\n",
"\n",
"The library is constructed using a modular approach, where different responsibilities are isolated within specific modules. These modules include the user interface functions (`core`), data processing and loading (`tsdataset`), scalers, losses, and base classes for models.\n",
"\n",
"This tutorial aims to explain the library's structure and to describe how the different modules interact with each other."
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## I. Map"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"The following diagram presents the modules of the `neuralforecast` library and their relations."
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"![Neuralforecast map](../imgs_indx/nf_map.png)"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## II. Modules"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1. Core (`core.py`)\n",
"\n",
"The `core` module acts as the primary interaction point for users of the `neuralforecast` library. It houses the `NeuralForecast` class, which incorporates a range of key user interface functions designed to simplify the process of training and forecasting models. Functions include `fit`, `predict`, `cross_validation`, and `predict_insample`, each one constructed to be intuitive and user-friendly. The design of the `NeuralForecast` class is centered around enabling users to streamline their forecasting pipelines and to comfortably train and evaluate models.\n",
"\n",
"### 2. Dataset and Loader (`tsdataset.py`)\n",
"\n",
"The `TimeSeriesDataset` class, located within the `tsdataset` module, is responsible for the storage and preprocessing of the input time series dataset. Once the `TimeSeriesDataset` class has prepared the data, it's then consumed by the `TimeSeriesLoader` class, which samples batches (or subsets) of the time series during the training and inference stages.\n",
"\n",
"### 3. Base Model (`common`)\n",
"\n",
"The `common` module contains three `BaseModel` classes, which serve as the foundation for all the model structures provided in the library. These base classes allow for a level of abstraction and code-reusability in the design of the models. We currently support three type of models:\n",
"\n",
" * `BaseWindows`: designed for window-based models like `NBEATS` and `Transformers`.\n",
" * `BaseRecurrent`: designed for recurrent models like `RNN` and `LSTM`.\n",
" * `BaseMultivariate`: caters to multivariate models like `StemGNN`.\n",
"\n",
"### 4. Model (`models`)\n",
"\n",
"The `models` module encompasses all the specific model classes available for use in the library. These include a variety of both simple and complex models such as `RNN`, `NHITS`, `LSTM`, `StemGNN`, and `TFT`. Each model in this module extends from one of the `BaseModel` classes in the `common` module.\n",
"\n",
"### 5. Losses (`losses`)\n",
"\n",
"The `losses` module includes both `numpy` and `pytorch` losses, used for evalaution and training respectively. The module contains a wide range of losses, including `MAE`, `MSE`, `MAPE`, `HuberLoss`, among many others. \n",
"\n",
"### 6. Scalers (`_scalers.py`)\n",
"\n",
"The `_scalers.py` module houses the `TemporalNorm` class. This class is responsible for the scaling (normalization) and de-scaling (reversing the normalization) of time series data. This step is crucial because it ensures all data fed to the model have a similar range, leading to more stable and efficient training processes."
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## III. Flow"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"The `user` first instantiates a model and the `NeuralForecast` core class. When they call the `fit` method, the following flow is executed:\n",
"\n",
"1. The `fit` method instantiates a `TimeSeriesDataset` object to store and pre-process the input time series dataset, and the `TimeSeriesLoader` object to sample batches.\n",
"2. The `fit` method calls the model's `fit` method (in the `BaseModel` class).\n",
"3. The model's `fit` method instantiates a Pytorch-Lightning `Trainer` object, in charge of training the model. \n",
"4. The `Trainer` method samples a batch from the `TimeSeriesLoader` object, and calls the model's `training_step` method (in the `BaseModel` class).\n",
"5. The model's `training_step`:\n",
" * Samples windows from the original batch.\n",
" * Normalizes the windows with the `scaler` module.\n",
" * Calls the model's `forward` method.\n",
" * Computes the loss using the `losses` module.\n",
" * Returns the loss.\n",
"6. The `Trainer` object repeats step 4 and 5 until `max_steps` iterations are completed.\n",
"7. The model is fitted, and can be used for forecasting future values (with the `predict` method) or recover insample predictions (using the `predict_insample` method)."
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## IV. Next Steps: add your own model"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"Congratulations! You now know the internal details of the `neuralforecast` library.\n",
"\n",
"With this knowledge you can easily add new models to the library, by just creating a `model` class which only requires the `init` and `forward` methods.\n",
"\n",
"Check our detailed guide on how to add new models!\n"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "python3",
"language": "python",
"name": "python3"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
nbs/examples/PredictInsample.ipynb 0 → 100644
View file @ e3f7f7b3
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# Predict Insample\n",
"> Tutorial on how to produce insample predictions."
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"This tutorial provides and example on how to use the `predict_insample` function of the `core` class to produce forecasts of the train and validation sets. In this example we will train the `NHITS` model on the AirPassengers data, and show how to recover the insample predictions after model is fitted.\n",
"\n",
"*Predict Insample*: The process of producing forecasts of the train and validation sets.\n",
"\n",
"*Use Cases*: \n",
"* Debugging: producing insample predictions is useful for debugging purposes. For example, to check if the model is able to fit the train set.\n",
"* Training convergence: check if the the model has converged.\n",
"* Anomaly detection: insample predictions can be used to detect anomalous behavior in the train set (e.g. outliers). (Note: if a model is too flexible it might be able to perfectly forecast outliers)"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"You can run these experiments using GPU with Google Colab.\n",
"\n",
"<a href=\"https://colab.research.google.com/github/Nixtla/neuralforecast/blob/main/nbs/examples/PredictInsample.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Installing NeuralForecast"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"!pip install neuralforecast"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Loading AirPassengers Data\n",
"\n",
"The `core.NeuralForecast` class contains shared, `fit`, `predict` and other methods that take as inputs pandas DataFrames with columns `['unique_id', 'ds', 'y']`, where `unique_id` identifies individual time series from the dataset, `ds` is the date, and `y` is the target variable. \n",
"\n",
"In this example dataset consists of a set of a single series, but you can easily fit your model to larger datasets in long format."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"from neuralforecast.utils import AirPassengersDF"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>unique_id</th>\n",
" <th>ds</th>\n",
" <th>y</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1.0</td>\n",
" <td>1949-01-31</td>\n",
" <td>112.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>1.0</td>\n",
" <td>1949-02-28</td>\n",
" <td>118.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>1.0</td>\n",
" <td>1949-03-31</td>\n",
" <td>132.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>1.0</td>\n",
" <td>1949-04-30</td>\n",
" <td>129.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>1.0</td>\n",
" <td>1949-05-31</td>\n",
" <td>121.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" unique_id ds y\n",
"0 1.0 1949-01-31 112.0\n",
"1 1.0 1949-02-28 118.0\n",
"2 1.0 1949-03-31 132.0\n",
"3 1.0 1949-04-30 129.0\n",
"4 1.0 1949-05-31 121.0"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Y_df = AirPassengersDF # Defined in neuralforecast.utils\n",
"Y_df.head()"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Model Training"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"First, we train the `NHITS` models on the AirPassengers data. We will use the `fit` method of the `core` class to train the models."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"\n",
"from neuralforecast import NeuralForecast\n",
"from neuralforecast.models import NHITS"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"horizon = 12\n",
"\n",
"# Try different hyperparmeters to improve accuracy.\n",
"models = [NHITS(h=horizon, # Forecast horizon\n",
" input_size=2 * horizon, # Length of input sequence\n",
" max_steps=1000, # Number of steps to train\n",
" n_freq_downsample=[2, 1, 1], # Downsampling factors for each stack output\n",
" mlp_units = 3 * [[1024, 1024]]) # Number of units in each block.\n",
" ]\n",
"nf = NeuralForecast(models=models, freq='M')\n",
"nf.fit(df=Y_df, val_size=horizon)"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. Predict Insample"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"Using the `NeuralForecast.predict_insample` method you can obtain the forecasts for the train and validation sets after the models are fitted. The function will always take the last dataset used for training in either the `fit` or `cross_validation` methods.\n",
"\n",
"With the `step_size` parameter you can specify the step size between consecutive windows to produce the forecasts. In this example we will set `step_size=horizon` to produce non-overlapping forecasts.\n",
"\n",
"The following diagram shows how the forecasts are produced based on the `step_size` parameter and `h` (horizon) of the model. In the diagram we set `step_size=2` and `h=4`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![](../imgs_indx/predict_insample.png)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 37.76it/s]\n"
]
}
],
"source": [
"Y_hat_insample = nf.predict_insample(step_size=horizon)"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"The `predict_insample` function returns a pandas DataFrame with the following columns:\n",
"* `unique_id`: the unique identifier of the time series.\n",
"* `ds`: the datestamp of the forecast for each row.\n",
"* `cutoff`: the datestamp at which the forecast was made.\n",
"* `y`: the actual value of the target variable.\n",
"* `model_name`: the forecasted values for the models. In this case, `NHITS`. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>unique_id</th>\n",
" <th>ds</th>\n",
" <th>cutoff</th>\n",
" <th>NHITS</th>\n",
" <th>y</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1.0</td>\n",
" <td>1949-01-31</td>\n",
" <td>1948-12-31</td>\n",
" <td>0.204289</td>\n",
" <td>112.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>1.0</td>\n",
" <td>1949-02-28</td>\n",
" <td>1948-12-31</td>\n",
" <td>0.302111</td>\n",
" <td>118.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>1.0</td>\n",
" <td>1949-03-31</td>\n",
" <td>1948-12-31</td>\n",
" <td>0.399522</td>\n",
" <td>132.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>1.0</td>\n",
" <td>1949-04-30</td>\n",
" <td>1948-12-31</td>\n",
" <td>0.429369</td>\n",
" <td>129.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>1.0</td>\n",
" <td>1949-05-31</td>\n",
" <td>1948-12-31</td>\n",
" <td>0.518200</td>\n",
" <td>121.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" unique_id ds cutoff NHITS y\n",
"0 1.0 1949-01-31 1948-12-31 0.204289 112.0\n",
"1 1.0 1949-02-28 1948-12-31 0.302111 118.0\n",
"2 1.0 1949-03-31 1948-12-31 0.399522 132.0\n",
"3 1.0 1949-04-30 1948-12-31 0.429369 129.0\n",
"4 1.0 1949-05-31 1948-12-31 0.518200 121.0"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Y_hat_insample.head()"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
":::{.callout-important}\n",
"The function will produce forecasts from the first timestamp of the time series. For these initial timestamps, the forecasts might not be accurate given that models have very limited input information to produce forecasts.\n",
":::"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. Plot Predictions"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"Finally, we plot the forecasts for the train and validation sets."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import matplotlib.pyplot as plt"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.legend.Legend>"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 1000x500 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(10, 5))\n",
"plt.plot(Y_hat_insample['ds'], Y_hat_insample['y'], label='True')\n",
"plt.plot(Y_hat_insample['ds'], Y_hat_insample['NHITS'], label='Forecast')\n",
"plt.axvline(Y_hat_insample['ds'].iloc[-12], color='black', linestyle='--', label='Train-Test Split')\n",
"plt.xlabel('Timestamp [t]')\n",
"plt.ylabel('Monthly Passengers')\n",
"plt.grid()\n",
"plt.legend()"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
":::{.callout-important}\n",
"Note how the forecasts for the train set are very accurate, while the forecast in the validation set (last 12 timetamps), are less precise. This is because the model was trained on the train set, and deep learning models such as the `NHITS` can easily overfit the train set.\n",
":::"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## References\n",
"- [Cristian Challu, Kin G. Olivares, Boris N. Oreshkin, Federico Garza, Max Mergenthaler-Canseco, Artur Dubrawski (2021). NHITS: Neural Hierarchical Interpolation for Time Series Forecasting. Accepted at AAAI 2023.](https://arxiv.org/abs/2201.12886)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "python3",
"language": "python",
"name": "python3"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
nbs/examples/Predictive_Maintenance.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/Robust_Regression.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/Save_Load_models.ipynb 0 → 100644
View file @ e3f7f7b3
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# Save and Load Models"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Saving and loading trained Deep Learning models has multiple valuable uses. These models are often costly to train; storing a pre-trained model can help reduce costs as it can be loaded and reused to forecast multiple times. Moreover, it enables Transfer learning capabilities, consisting of pre-training a flexible model on a large dataset and using it later on other data with little to no training. It is one of the most outstanding 🚀 achievements in Machine Learning 🧠 and has many practical applications.\n",
"\n",
"In this notebook we show an example on how to save and load `NeuralForecast` models.\n",
"\n",
"The two methods to consider are:<br>\n",
"1. `NeuralForecast.save`: Saves models into disk, allows save dataset and config.<br>\n",
"2. `NeuralForecast.load`: Loads models from a given path.<br>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
":::{.callout-important}\n",
"This Guide assumes basic knowledge on the NeuralForecast library. For a minimal example visit the [Getting Started](./Getting_Started.ipynb) guide.\n",
":::"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can run these experiments using GPU with Google Colab.\n",
"\n",
"<a href=\"https://colab.research.google.com/github/Nixtla/neuralforecast/blob/main/nbs/examples/Save_Load_models.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Installing NeuralForecast"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"!pip install neuralforecast"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Loading AirPassengers Data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For this example we will use the classical [AirPassenger Data set](https://www.kaggle.com/datasets/rakannimer/air-passengers). Import the pre-processed AirPassenger from `utils`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/cchallu/opt/anaconda3/envs/neuralforecast/lib/python3.10/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
" from .autonotebook import tqdm as notebook_tqdm\n"
]
}
],
"source": [
"from neuralforecast.utils import AirPassengersDF\n",
"\n",
"Y_df = AirPassengersDF\n",
"Y_df = Y_df.reset_index(drop=True)\n",
"Y_df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Model Training"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Next, we instantiate and train three models: `NBEATS`, `NHITS`, and `AutoMLP`. The models with their hyperparameters are defined in the `models` list."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from ray import tune\n",
"\n",
"from neuralforecast.core import NeuralForecast\n",
"from neuralforecast.auto import AutoMLP\n",
"from neuralforecast.models import NBEATS, NHITS"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"horizon = 12\n",
"models = [NBEATS(input_size=2 * horizon, h=horizon, max_steps=50),\n",
" NHITS(input_size=2 * horizon, h=horizon, max_steps=50),\n",
" AutoMLP(# Ray tune explore config\n",
" config=dict(max_steps=100, # Operates with steps not epochs\n",
" input_size=tune.choice([3*horizon]),\n",
" learning_rate=tune.choice([1e-3])),\n",
" h=horizon,\n",
" num_samples=1, cpus=1)]"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"nf = NeuralForecast(models=models, freq='M')\n",
"nf.fit(df=Y_df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Produce the forecasts with the `predict` method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 98.79it/s] \n",
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 123.41it/s]\n",
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 161.79it/s]\n"
]
},
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>unique_id</th>\n",
" <th>ds</th>\n",
" <th>NBEATS</th>\n",
" <th>NHITS</th>\n",
" <th>AutoMLP</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1.0</td>\n",
" <td>1961-01-31</td>\n",
" <td>428.410553</td>\n",
" <td>445.268158</td>\n",
" <td>452.550446</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>1.0</td>\n",
" <td>1961-02-28</td>\n",
" <td>425.958557</td>\n",
" <td>469.293945</td>\n",
" <td>442.683807</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>1.0</td>\n",
" <td>1961-03-31</td>\n",
" <td>477.748016</td>\n",
" <td>462.920807</td>\n",
" <td>474.043457</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>1.0</td>\n",
" <td>1961-04-30</td>\n",
" <td>477.548798</td>\n",
" <td>489.986633</td>\n",
" <td>503.836334</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>1.0</td>\n",
" <td>1961-05-31</td>\n",
" <td>495.973541</td>\n",
" <td>518.612610</td>\n",
" <td>531.347900</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" unique_id ds NBEATS NHITS AutoMLP\n",
"0 1.0 1961-01-31 428.410553 445.268158 452.550446\n",
"1 1.0 1961-02-28 425.958557 469.293945 442.683807\n",
"2 1.0 1961-03-31 477.748016 462.920807 474.043457\n",
"3 1.0 1961-04-30 477.548798 489.986633 503.836334\n",
"4 1.0 1961-05-31 495.973541 518.612610 531.347900"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Y_hat_df = nf.predict().reset_index()\n",
"Y_hat_df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We plot the forecasts for each model. Note how the two `NBEATS` models are differentiated with a numerical suffix."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import matplotlib.pyplot as plt"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.legend.Legend>"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"text/plain": [
"<Figure size 1200x300 with 0 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plot_df = pd.concat([Y_df, Y_hat_df]).set_index('ds') # Concatenate the train and forecast dataframes\n",
"\n",
"plt.figure(figsize = (12, 3))\n",
"plot_df[['y', 'NBEATS', 'NHITS', 'AutoMLP']].plot(linewidth=2)\n",
"\n",
"plt.title('AirPassengers Forecast', fontsize=10)\n",
"plt.ylabel('Monthly Passengers', fontsize=10)\n",
"plt.xlabel('Timestamp [t]', fontsize=10)\n",
"plt.axvline(x=plot_df.index[-horizon], color='k', linestyle='--', linewidth=2)\n",
"plt.legend(prop={'size': 10})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. Save models"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To save all the trained models use the `save` method. This method will save both the hyperparameters and the learnable weights (parameters).\n",
"\n",
"The `save` method has the following inputs:\n",
"\n",
"* `path`: directory where models will be saved.\n",
"* `model_index`: optional list to specify which models to save. For example, to only save the `NHITS` model use `model_index=[2]`.\n",
"* `overwrite`: boolean to overwrite existing files in `path`. When True, the method will only overwrite models with conflicting names.\n",
"* `save_dataset`: boolean to save `Dataset` object with the dataset."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"nf.save(path='./checkpoints/test_run/',\n",
" model_index=None, \n",
" overwrite=True,\n",
" save_dataset=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For each model, two files are created and stored:\n",
"\n",
"* `[model_name]_[suffix].ckpt`: Pytorch Lightning checkpoint file with the model parameters and hyperparameters.\n",
"* `[model_name]_[suffix].pkl`: Dictionary with configuration attributes.\n",
"\n",
"Where `model_name` corresponds to the name of the model in lowercase (eg. `nhits`). We use a numerical suffix to distinguish multiple models of each class. In this example the names will be `automlp_0`, `nbeats_0`, and `nhits_0`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
":::{.callout-important}\n",
"The `Auto` models will be stored as their base model. For example, the `AutoMLP` trained above is stored as an `MLP` model, with the best hyparparameters found during tuning.\n",
":::"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. Load models"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Load the saved models with the `load` method, specifying the `path`, and use the new `nf2` object to produce forecasts."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 153.75it/s]\n",
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 142.04it/s]\n",
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 105.82it/s]\n"
]
},
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>unique_id</th>\n",
" <th>ds</th>\n",
" <th>MLP</th>\n",
" <th>NHITS</th>\n",
" <th>NBEATS</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1.0</td>\n",
" <td>1961-01-31</td>\n",
" <td>452.550446</td>\n",
" <td>445.268158</td>\n",
" <td>428.410553</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>1.0</td>\n",
" <td>1961-02-28</td>\n",
" <td>442.683807</td>\n",
" <td>469.293945</td>\n",
" <td>425.958557</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>1.0</td>\n",
" <td>1961-03-31</td>\n",
" <td>474.043457</td>\n",
" <td>462.920807</td>\n",
" <td>477.748016</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>1.0</td>\n",
" <td>1961-04-30</td>\n",
" <td>503.836334</td>\n",
" <td>489.986633</td>\n",
" <td>477.548798</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>1.0</td>\n",
" <td>1961-05-31</td>\n",
" <td>531.347900</td>\n",
" <td>518.612610</td>\n",
" <td>495.973541</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" unique_id ds MLP NHITS NBEATS\n",
"0 1.0 1961-01-31 452.550446 445.268158 428.410553\n",
"1 1.0 1961-02-28 442.683807 469.293945 425.958557\n",
"2 1.0 1961-03-31 474.043457 462.920807 477.748016\n",
"3 1.0 1961-04-30 503.836334 489.986633 477.548798\n",
"4 1.0 1961-05-31 531.347900 518.612610 495.973541"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"nf2 = NeuralForecast.load(path='./checkpoints/test_run/')\n",
"Y_hat_df = nf2.predict().reset_index()\n",
"Y_hat_df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Finally, plot the forecasts to confirm they are identical to the original forecasts."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<Figure size 1200x300 with 0 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plot_df = pd.concat([Y_df, Y_hat_df]).set_index('ds') # Concatenate the train and forecast dataframes\n",
"\n",
"plt.figure(figsize = (12, 3))\n",
"plot_df[['y', 'NBEATS', 'NHITS', 'MLP']].plot(linewidth=2)\n",
"\n",
"plt.title('AirPassengers Forecast', fontsize=10)\n",
"plt.ylabel('Monthly Passengers', fontsize=10)\n",
"plt.xlabel('Timestamp [t]', fontsize=10)\n",
"plt.axvline(x=plot_df.index[-horizon], color='k', linestyle='--', linewidth=2)\n",
"plt.legend(prop={'size': 10})\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## References\n",
"\n",
"https://pytorch-lightning.readthedocs.io/en/stable/common/checkpointing_basic.html\n",
"\n",
"[Oreshkin, B. N., Carpov, D., Chapados, N., & Bengio, Y. (2019). N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. ICLR 2020](https://arxiv.org/abs/1905.10437)\n",
"\n",
"[Cristian Challu, Kin G. Olivares, Boris N. Oreshkin, Federico Garza, Max Mergenthaler-Canseco, Artur Dubrawski (2021). N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting. Accepted at AAAI 2023.](https://arxiv.org/abs/2201.12886)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "python3",
"language": "python",
"name": "python3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
nbs/examples/Signal_Decomposition.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/StatsMlNeuralMethods.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/Temporal_Classifiers.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/Time_Series_Scaling.ipynb 0 → 100644
View file @ e3f7f7b3
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Time Series Scaling"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Scaling time series data is an important preprocessing step when using neural forecasting methods for several reasons:\n",
"\n",
"1. **Convergence speed**: Neural forecasting models tend to converge faster when the features are on a similar scale.\n",
"2. **Avoiding vanishing or exploding gradients**: some architectures, such as recurrent neural networks (RNNs), are sensitive to the scale of input data. If the input values are too large, it could lead to exploding gradients, where the gradients become too large and the model becomes unstable. Conversely, very small input values could lead to vanishing gradients, where weight updates during training are negligible and the training fails to converge.\n",
"3. **Ensuring consistent scale**: Neural forecasting models have shared global parameters for the all time series of the task. In cases where time series have different scale, scaling ensures that no particular time series dominates the learning process.\n",
"4. **Improving generalization**: time series with consistent scale can lead to smoother loss surfaces. Moreover, scaling helps to homogenize the distribution of the input data, which can also improve generalization by avoiding out-of-range values.\n",
"\n",
"The `Neuralforecast` library integrates two types of temporal scaling:\n",
"\n",
"* **Time Series Scaling**: scaling each time series using all its data on the train set before start training the model. This is done by using the `local_scaler_type` parameter of the `Neuralforecast` core class.\n",
"* **Window scaling (TemporalNorm)**: scaling each input window separetly for each element of the batch at every training iteration. This is done by using the `scaler_type` parameter of each model class.\n",
"\n",
"In this notebook, we will demonstrate how to scale the time series data with both methods on an Eletricity Price Forecasting (EPF) task."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can run these experiments using GPU with Google Colab.\n",
"\n",
"<a href=\"https://colab.research.google.com/github/Nixtla/neuralforecast/blob/main/nbs/examples/Time_Series_Scaling.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Install `Neuralforecast`"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%capture\n",
"!pip install neuralforecast\n",
"!pip install hyperopt"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Load Data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The `df` dataframe contains the target and exogenous variables past information to train the model. The `unique_id` column identifies the markets, `ds` contains the datestamps, and `y` the electricity price. For future variables, we include a forecast of how much electricity will be produced (`gen_forecast`), system load (`system_laod`), and day of week (`week_day`). Both the electricity system demand and offer impact the price significantly, including these variables to the model greatly improve performance, as we demonstrate in Olivares et al. (2022).\n",
"\n",
"The `futr_df` dataframe includes the information of the future exogenous variables for the period we want to forecast (in this case, 24 hours after the end of the train dataset `df`)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import matplotlib.pyplot as plt"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>unique_id</th>\n",
" <th>ds</th>\n",
" <th>y</th>\n",
" <th>gen_forecast</th>\n",
" <th>system_load</th>\n",
" <th>week_day</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>FR</td>\n",
" <td>2015-01-01 00:00:00</td>\n",
" <td>53.48</td>\n",
" <td>76905.0</td>\n",
" <td>74812.0</td>\n",
" <td>3</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>FR</td>\n",
" <td>2015-01-01 01:00:00</td>\n",
" <td>51.93</td>\n",
" <td>75492.0</td>\n",
" <td>71469.0</td>\n",
" <td>3</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>FR</td>\n",
" <td>2015-01-01 02:00:00</td>\n",
" <td>48.76</td>\n",
" <td>74394.0</td>\n",
" <td>69642.0</td>\n",
" <td>3</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>FR</td>\n",
" <td>2015-01-01 03:00:00</td>\n",
" <td>42.27</td>\n",
" <td>72639.0</td>\n",
" <td>66704.0</td>\n",
" <td>3</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>FR</td>\n",
" <td>2015-01-01 04:00:00</td>\n",
" <td>38.41</td>\n",
" <td>69347.0</td>\n",
" <td>65051.0</td>\n",
" <td>3</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" unique_id ds y gen_forecast system_load week_day\n",
"0 FR 2015-01-01 00:00:00 53.48 76905.0 74812.0 3\n",
"1 FR 2015-01-01 01:00:00 51.93 75492.0 71469.0 3\n",
"2 FR 2015-01-01 02:00:00 48.76 74394.0 69642.0 3\n",
"3 FR 2015-01-01 03:00:00 42.27 72639.0 66704.0 3\n",
"4 FR 2015-01-01 04:00:00 38.41 69347.0 65051.0 3"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df = pd.read_csv('https://datasets-nixtla.s3.amazonaws.com/EPF_FR_BE.csv')\n",
"df['ds'] = pd.to_datetime(df['ds'])\n",
"\n",
"futr_df = pd.read_csv('https://datasets-nixtla.s3.amazonaws.com/EPF_FR_BE_futr.csv')\n",
"futr_df['ds'] = pd.to_datetime(futr_df['ds'])\n",
"\n",
"df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can see that `y` and the exogenous variables are on largely different scales. Next, we show two methods to scale the data."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Time Series Scaling with `Neuralforecast` class"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"One of the most widely used approches for scaling time series is to treat it as a pre-processing step, where each time series and temporal exogenous variables are scaled based on their entire information in the train set. Models are then trained on the scaled data.\n",
"\n",
"To simplify pipelines, we added a scaling functionality to the `Neuralforecast` class. Each time series will be scaled before training the model with either `fit` or `cross_validation`, and scaling statistics are stored. The class then uses the stored statistics to scale the forecasts back to the original scale before returning the forecasts."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.a. Instantiate model and `Neuralforecast` class"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In this example we will use the `TimesNet` model, recently proposed in [Wu, Haixu, et al. (2022)](https://arxiv.org/abs/2210.02186). First instantiate the model with the desired parameters."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from neuralforecast.models import TimesNet\n",
"from neuralforecast.core import NeuralForecast\n",
"\n",
"import logging\n",
"logging.getLogger(\"pytorch_lightning\").setLevel(logging.WARNING)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Global seed set to 1\n"
]
}
],
"source": [
"horizon = 24 # day-ahead daily forecast\n",
"model = TimesNet(h = horizon, # Horizon\n",
" input_size = 5*horizon, # Length of input window\n",
" max_steps = 100, # Training iterations\n",
" top_k = 3, # Number of periods (for FFT).\n",
" num_kernels = 3, # Number of kernels for Inception module\n",
" batch_size = 2, # Number of time series per batch\n",
" windows_batch_size = 32, # Number of windows per batch\n",
" learning_rate = 0.001, # Learning rate\n",
" futr_exog_list = ['gen_forecast', 'system_load','week_day'], # Future exogenous variables\n",
" scaler_type = None) # We use the Core scaling method"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Fit the model by instantiating a `NeuralForecast` object and using the `fit` method. The `local_scaler_type` parameter is used to specify the type of scaling to be used. In this case, we will use `standard`, which scales the data to have zero mean and unit variance.Other supported scalers are `minmax`, `robust`, `robust-iqr`, `minmax`, and `boxcox`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 99: 100%|██████████| 1/1 [00:00<00:00, 1.13it/s, v_num=181, train_loss_step=0.413, train_loss_epoch=0.413]\n"
]
}
],
"source": [
"nf = NeuralForecast(models=[model], freq='H', local_scaler_type='standard')\n",
"nf.fit(df=df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.b Forecast and plots"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Finally, use the `predict` method to forecast the day-ahead prices. The `Neuralforecast` class handles the inverse normalization, forecasts are returned in the original scale."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 26.56it/s]\n"
]
},
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>ds</th>\n",
" <th>TimesNet</th>\n",
" </tr>\n",
" <tr>\n",
" <th>unique_id</th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 00:00:00</td>\n",
" <td>33.748502</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 01:00:00</td>\n",
" <td>32.393269</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 02:00:00</td>\n",
" <td>29.000997</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 03:00:00</td>\n",
" <td>26.264737</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 04:00:00</td>\n",
" <td>28.841827</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" ds TimesNet\n",
"unique_id \n",
"BE 2016-11-01 00:00:00 33.748502\n",
"BE 2016-11-01 01:00:00 32.393269\n",
"BE 2016-11-01 02:00:00 29.000997\n",
"BE 2016-11-01 03:00:00 26.264737\n",
"BE 2016-11-01 04:00:00 28.841827"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Y_hat_df = nf.predict(futr_df=futr_df)\n",
"Y_hat_df.head()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"\n",
"plot_df = df[df['unique_id']=='FR'].tail(24*5).reset_index(drop=True)\n",
"Y_hat_df = Y_hat_df.reset_index(drop=False)\n",
"Y_hat_df = Y_hat_df[Y_hat_df['unique_id']=='FR']\n",
"\n",
"plot_df = pd.concat([plot_df, Y_hat_df ]).set_index('ds') # Concatenate the train and forecast dataframes\n",
"\n",
"plot_df[['y', 'TimesNet']].plot(linewidth=2)\n",
"plt.axvline('2016-11-01', color='red')\n",
"plt.ylabel('Price [EUR/MWh]', fontsize=12)\n",
"plt.xlabel('Date', fontsize=12)\n",
"plt.grid()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
":::{.callout-important}\n",
"The inverse scaling is performed by the `Neuralforecast` class before returning the final forecasts. Therefore, the hyperparmater selection with `Auto` models and validation loss for early stopping or model selection are performed on the scaled data. Different types of scaling with the `Neuralforecast` class can't be automatically compared with `Auto` models.\n",
":::"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. Temporal Window normalization during training"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Temporal normalization scales each instance of the batch separately at the window level. It is performed at each training iteration for each window of the batch, for both target variable and temporal exogenous covariates. For more details, see [Olivares et al. (2023)](https://arxiv.org/abs/2305.07089) and https://nixtla.github.io/neuralforecast/common.scalers.html."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 4.a. Instantiate model and `Neuralforecast` class"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Temporal normalization is specified by the `scaler_type` argument. Currently, it is only supported for Windows-based models (`NHITS`, `NBEATS`, `MLP`, `TimesNet`, and all Transformers). In this example, we use the `TimesNet` model and `robust` scaler, recently proposed by Wu, Haixu, et al. (2022). First instantiate the model with the desired parameters.\n",
"\n",
"Visit https://nixtla.github.io/neuralforecast/common.scalers.html for a complete list of supported scalers."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Global seed set to 1\n"
]
}
],
"source": [
"horizon = 24 # day-ahead daily forecast\n",
"model = TimesNet(h = horizon, # Horizon\n",
" input_size = 5*horizon, # Length of input window\n",
" max_steps = 100, # Training iterations\n",
" top_k = 3, # Number of periods (for FFT).\n",
" num_kernels = 3, # Number of kernels for Inception module\n",
" batch_size = 2, # Number of time series per batch\n",
" windows_batch_size = 32, # Number of windows per batch\n",
" learning_rate = 0.001, # Learning rate\n",
" futr_exog_list = ['gen_forecast', 'system_load','week_day'], # Future exogenous variables\n",
" scaler_type = 'robust') # Robust scaling"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Fit the model by instantiating a `NeuralForecast` object and using the `fit` method. Note that `local_scaler_type` has `None` as default to avoid scaling the data before training."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 99: 100%|██████████| 1/1 [00:00<00:00, 1.73it/s, v_num=183, train_loss_step=0.977, train_loss_epoch=0.977]\n"
]
}
],
"source": [
"nf = NeuralForecast(models=[model], freq='H')\n",
"nf.fit(df=df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 4.b Forecast and plots"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Finally, use the `predict` method to forecast the day-ahead prices. The forecasts are returned in the original scale."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Predicting DataLoader 0: 100%|██████████| 1/1 [00:00<00:00, 20.26it/s]\n"
]
},
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>ds</th>\n",
" <th>TimesNet</th>\n",
" </tr>\n",
" <tr>\n",
" <th>unique_id</th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 00:00:00</td>\n",
" <td>40.024895</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 01:00:00</td>\n",
" <td>35.253803</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 02:00:00</td>\n",
" <td>33.185341</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 03:00:00</td>\n",
" <td>33.572426</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BE</th>\n",
" <td>2016-11-01 04:00:00</td>\n",
" <td>37.039207</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" ds TimesNet\n",
"unique_id \n",
"BE 2016-11-01 00:00:00 40.024895\n",
"BE 2016-11-01 01:00:00 35.253803\n",
"BE 2016-11-01 02:00:00 33.185341\n",
"BE 2016-11-01 03:00:00 33.572426\n",
"BE 2016-11-01 04:00:00 37.039207"
]
},
"execution_count": null,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Y_hat_df = nf.predict(futr_df=futr_df)\n",
"Y_hat_df.head()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"\n",
"plot_df = df[df['unique_id']=='FR'].tail(24*5).reset_index(drop=True)\n",
"Y_hat_df = Y_hat_df.reset_index(drop=False)\n",
"Y_hat_df = Y_hat_df[Y_hat_df['unique_id']=='FR']\n",
"\n",
"plot_df = pd.concat([plot_df, Y_hat_df ]).set_index('ds') # Concatenate the train and forecast dataframes\n",
"\n",
"plot_df[['y', 'TimesNet']].plot(linewidth=2)\n",
"plt.axvline('2016-11-01', color='red')\n",
"plt.ylabel('Price [EUR/MWh]', fontsize=12)\n",
"plt.xlabel('Date', fontsize=12)\n",
"plt.grid()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
":::{.callout-important}\n",
"For most applications, models with temporal normalization (section 4) produced more accurate forecasts than time series scaling (section 3). However, with temporal normalization models lose the information of the relative level between different windows. In some cases this global information within time series is crucial, for instance when an exogenous variables contains the dosage of a medication. In these cases, time series scaling (section 3) is preferred.\n",
":::"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## References\n",
"\n",
"- [Kin G. Olivares, David Luo, Cristian Challu, Stefania La Vattiata, Max Mergenthaler, Artur Dubrawski (2023). \"HINT: Hierarchical Mixture Networks For Coherent Probabilistic Forecasting\". International Conference on Machine Learning (ICML). Workshop on Structured Probabilistic Inference & Generative Modeling. Available at https://arxiv.org/abs/2305.07089.](https://arxiv.org/abs/2305.07089)\n",
"- [Wu, Haixu, Tengge Hu, Yong Liu, Hang Zhou, Jianmin Wang, and Mingsheng Long. \"Timesnet: Temporal 2d-variation modeling for general time series analysis.\", ICLR 2023](https://openreview.net/forum?id=ju_Uqw384Oq)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "python3",
"language": "python",
"name": "python3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
nbs/examples/Transfer_Learning.ipynb 0 → 100644
View file @ e3f7f7b3
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# Transfer Learning"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Transfer learning refers to the process of pre-training a flexible model on a large dataset and using it later on other data with little to no training. It is one of the most outstanding 🚀 achievements in Machine Learning 🧠 and has many practical applications.\n",
"\n",
"For time series forecasting, the technique allows you to get lightning-fast predictions ⚡ bypassing the tradeoff between accuracy and speed (more than 30 times faster than our alreadsy fast [autoARIMA](https://github.com/Nixtla/statsforecast) for a similar accuracy).\n",
"\n",
"This notebook shows how to generate a pre-trained model and store it in a checkpoint to make it available to forecast new time series never seen by the model. \n",
"\n",
"Table of Contents<br>\n",
"1. Installing NeuralForecast/DatasetsForecast<br>\n",
"2. Load M4 Data<br>\n",
"3. Instantiate NeuralForecast core, Fit, and save<br>\n",
"4. Load pre-trained model and predict on AirPassengers<br>\n",
"5. Evaluate Results<br>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can run these experiments using GPU with Google Colab.\n",
"\n",
"<a href=\"https://colab.research.google.com/github/Nixtla/neuralforecast/blob/main/nbs/examples/Transfer_Learning.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Installing Libraries"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# %%capture\n",
"# !pip install git+https://github.com/Nixtla/datasetsforecast.git@main"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# %%capture\n",
"# !pip install neuralforecast "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"import pandas as pd\n",
"import torch\n",
"from IPython.display import display, Markdown\n",
"\n",
"import matplotlib.pyplot as plt\n",
"\n",
"from datasetsforecast.m4 import M4\n",
"from neuralforecast.core import NeuralForecast\n",
"from neuralforecast.models import NHITS\n",
"from neuralforecast.utils import AirPassengersDF\n",
"from neuralforecast.losses.numpy import mae, mse"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"logging.getLogger(\"pytorch_lightning\").setLevel(logging.WARNING)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This example will automatically run on GPUs if available. **Make sure** cuda is available. (If you need help to put this into production send us an email or join or community, we also offer a fully hosted solution)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"torch.cuda.is_available()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Load M4 Data\n",
"\n",
"The `M4` class will automatically download the complete M4 dataset and process it.\n",
"\n",
"It return three Dataframes: `Y_df` contains the values for the target variables, `X_df` contains exogenous calendar features and `S_df` contains static features for each time-series (none for M4). For this example we will only use `Y_df`.\n",
"\n",
"If you want to use your own data just replace `Y_df`. Be sure to use a long format and have a simmilar structure than our data set."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"Y_df, _, _ = M4.load(directory='./', group='Monthly', cache=True)\n",
"Y_df['ds'] = pd.to_datetime(Y_df['ds'])\n",
"Y_df"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Model Train and Save\n",
"\n",
"Using the `NeuralForecast.fit` method you can train a set of models to your dataset. You just have to define the `input_size` and `horizon` of your model. The `input_size` is the number of historic observations (lags) that the model will use to learn to predict `h` steps in the future. Also, you can modify the hyperparameters of the model to get a better accuracy."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"horizon = 12\n",
"stacks = 3\n",
"models = [NHITS(input_size=5 * horizon,\n",
" h=horizon,\n",
" max_steps=100,\n",
" stack_types = stacks*['identity'],\n",
" n_blocks = stacks*[1],\n",
" mlp_units = [[256,256] for _ in range(stacks)],\n",
" n_pool_kernel_size = stacks*[1],\n",
" batch_size = 32,\n",
" scaler_type='standard',\n",
" n_freq_downsample=[12,4,1])]\n",
"nf = NeuralForecast(models=models, freq='M')\n",
"nf.fit(df=Y_df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Save model with `core.NeuralForecast.save` method. This method uses PytorchLightning `save_checkpoint` function. We set `save_dataset=False` to only save the model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"nf.save(path='./results/transfer/', model_index=None, overwrite=True, save_dataset=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. Transfer M4 to AirPassengers\n",
"\n",
"We load the stored model with the `core.NeuralForecast.load` method, and forecast `AirPassenger` with the `core.NeuralForecast.predict` function."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"fcst2 = NeuralForecast.load(path='./results/transfer/')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# We define the train df. \n",
"Y_df = AirPassengersDF.copy()\n",
"mean = Y_df[Y_df.ds<='1959-12-31']['y'].mean()\n",
"std = Y_df[Y_df.ds<='1959-12-31']['y'].std()\n",
"\n",
"Y_train_df = Y_df[Y_df.ds<='1959-12-31'] # 132 train\n",
"Y_test_df = Y_df[Y_df.ds>'1959-12-31'] # 12 test"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"Y_hat_df = fcst2.predict(df=Y_train_df).reset_index()\n",
"Y_hat_df.head()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"fig, ax = plt.subplots(1, 1, figsize = (20, 7))\n",
"Y_hat_df = Y_test_df.merge(Y_hat_df, how='left', on=['unique_id', 'ds'])\n",
"plot_df = pd.concat([Y_train_df, Y_hat_df]).set_index('ds')\n",
"\n",
"plot_df[['y', 'NHITS']].plot(ax=ax, linewidth=2)\n",
"\n",
"ax.set_title('AirPassengers Forecast', fontsize=22)\n",
"ax.set_ylabel('Monthly Passengers', fontsize=20)\n",
"ax.set_xlabel('Timestamp [t]', fontsize=20)\n",
"ax.legend(prop={'size': 15})\n",
"ax.grid()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. Evaluate Results\n",
"\n",
"\n",
"We evaluate the forecasts of the pre-trained model with the Mean Absolute Error (`mae`).\n",
"\n",
"$$\n",
"\\qquad MAE = \\frac{1}{Horizon} \\sum_{\\tau} |y_{\\tau} - \\hat{y}_{\\tau}|\\qquad\n",
"$$"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_true = Y_test_df.y.values\n",
"y_hat = Y_hat_df['NHITS'].values"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print('NHITS MAE: %0.3f' % mae(y_hat, y_true))\n",
"print('ETS MAE: 16.222')\n",
"print('AutoARIMA MAE: 18.551')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "python3",
"language": "python",
"name": "python3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
nbs/examples/UncertaintyIntervals.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/distributed/distributed_neuralforecast.ipynb 0 → 100644
View file @ e3f7f7b3
This source diff could not be displayed because it is too large. You can view the blob instead.
nbs/examples/models_intro.qmd 0 → 100644
View file @ e3f7f7b3
---
title: "NeuralForecast's Contents"
---
## Automatic Forecasting Models
Automatic forecasting tools optimize the hyperparameters of a given model class and select the best-performing model for a validation set. The optimization methods include grid search, random search, and Bayesian optimization.
| MLP-Based |RNN-Based | Transformers | CNN-Based | Multivariate |
|:------------------------------------------|:------------------------------------- |:---------------------------------------------|:--------------------------------------------| :-------------------------------------------|
|[`AutoMLP`](../models.html#automlp) |[`AutoRNN`](../models.html#autornn) |[`AutoTFT`](../models.html#autotft) |[`AutoTimesNet`](../models.html#autotimesnet)| [`AutoStemGNN`](../models.html#autostemgnn) |
|[`AutoNBEATS`](../models.html#autonbeats) |[`AutoLSTM`](../models.html#autolstm) |[`AutoInformer`](../models.html#autoinformer) | | [`AutoHINT`](../models.html#autohint) |
|[`AutoNBEATSx`](../models.html#autonbeatsx)|[`AutoGRU`](../models.html#autogru) |[`Autoformer`](../models.html#autoautoformer) | | |
|[`AutoNHITS`](../models.html#autonhits) |[`AutoTCN`](../models.html#autotcn) |[`AutoPatchTST`](../models.html#autopatchtst) | | |
| |[`AutoDeepAR`](../models.html#autodeepar)| | | |
: {tbl-colwidths="[25,25]"}
## Optimization Objectives
NeuralForecast is a highly modular framework capable of augmenting a wide variety of robust neural network architectures with different point or probability outputs as defined by their optimization objectives.
| Scale-Dependent | Percentage-Errors | Scale-Independent | Robust |
|:-------------------------------------------------------------|:----------------------------------------------------------------------|:---------------------------------------------------------------|:-------------------------------------------------------|
|[`MAE`](../losses.pytorch.html#mean-absolute-error-mae) |[`MAPE`](../losses.pytorch.html#mean-absolute-percentage-error-mape) |[`MASE`](../losses.pytorch.html#mean-absolute-scaled-error-mase)|[`Huber`](../losses.pytorch.html#huber-loss) |
|[`MSE`](../losses.pytorch.html#mean-squared-error-mse) |[`sMAPE`](../losses.pytorch.html#symmetric-mape-smape) | |[`Tukey`](../losses.pytorch.html#tukey-loss) |
|[`RMSE`](../losses.pytorch.html#root-mean-squared-error-rmse) | | |[`HuberMQLoss`](../losses.pytorch.html#huberized-mqloss)|
: {tbl-colwidths="[25,25]"}
|Parametric Probabilities | Non-Parametric Probabilities |
|:-------------------------------------------------------------|:-------------------------------------------------------------|
|[`Normal`](../losses.pytorch.html#distributionloss) |[`QuantileLoss`](../losses.pytorch.html#quantile-loss) |
|[`StudenT`](../losses.pytorch.html#distributionloss) |[`MQLoss`](../losses.pytorch.html#multi-quantile-loss-mqloss) |
|[`Poisson`](../losses.pytorch.html#distributionloss) |[`HuberQLoss`](../losses.pytorch.html#huberized-quantile-loss)|
|[`Negative Binomial`](../losses.pytorch.html#distributionloss)|[`HuberMQLoss`](../losses.pytorch.html#huberized-mqloss) |
|[`Tweedie`](../losses.pytorch.html#distributionloss) | |
|[`PMM`](../losses.pytorch.html#poisson-mixture-mesh-pmm) /[`GMM`](../losses.pytorch.html#gaussian-mixture-mesh-gmm) | |
: {tbl-colwidths="[25,25]"}
## MLP-Based Model Family
The MLP-based family operates like a classic autoencoder. Its initial layers encode raw autoregressive window into a representation, and the decoder produces the desired output based on the horizon, probability output, or point objective. Recent architectures include modifications like residual learning techniques and task-specific changes.
|Model | Point Forecast | Probabilistic Forecast | Insample fitted values | Probabilistic fitted values |
|:-----------------------------------------|:--------------:|:----------------------:|:----------------------:|:----------------------------:|
|[`MLP`](../models.mlp.html) |✅ |✅ |✅ |✅ |
|[`NBEATS`](../models.nbeats.html) |✅ |✅ |✅ |✅ |
|[`NBEATSx`](../models.nbeatsx.html) |✅ |✅ |✅ |✅ |
|[`NHITS`](../models.nhits.html) |✅ |✅ |✅ |✅ |
: {tbl-colwidths="[25,25]"}
## RNN-Based Model Family
The RNN-based family attempts to leverage the data's temporal structure while reducing MLPs over parametrization. Recurrent networks are dynamic and can handle sequences of varying lengths through a mechanism for updating internal states that considers the entire sequence history. Modern state modifications help diminish vanishing and exploding gradients.
|Model | Point Forecast | Probabilistic Forecast | Insample fitted values | Probabilistic fitted values |
|:-------------------------------------------|:----------------:|:----------------------:|:----------------------:|:----------------------------:|
|[`RNN`](../models.rnn.html) |✅ |✅ |✅ |✅ |
|[`GRU`](../models.gru.html) |✅ |✅ |✅ |✅ |
|[`LSTM`](../models.lstm.html) |✅ |✅ |✅ |✅ |
|[`TCN`](../models.tcn.html) |✅ |✅ |✅ |✅ |
|[`DeepAR`](../models.deepar.html) |✅ |✅ |✅ |✅ |
|[`DilatedRNN`](../models.dilated_rnn.html) |✅ |✅ |✅ |✅ |
: {tbl-colwidths="[25,25]"}
## Transformers Model Family
Transformer architectures are an alternative to recurrent networks. These networks build on the self-attention mechanism that directly allows modeling the relationship between different sequence parts without sequential processing. Attention makes Transformers more parallelizable than RNNs.
|Model | Point Forecast | Probabilistic Forecast | Insample fitted values | Probabilistic fitted values |
|:----------------------------------------------------------|:----------------:|:----------------------:|:----------------------:|:----------------------------:|
|[`TFT`](../models.tft.html) |✅ |✅ |✅ |✅ |
|[`Informer`](../models.informer.html) |✅ |✅ |✅ |✅ |
|[`Autoformer`](../models.autoformer.html) |✅ |✅ |✅ |✅ |
|[`PatchTST`](../models.patchtst.html) |✅ |✅ |✅ |✅ |
|[`VanillaTransformer`](../models.vanillatransformer.html) |✅ |✅ |✅ |✅ |
: {tbl-colwidths="[25,25]"}
## CNN-Based Model Family
Convolutional Neural Networks (CNNs), originally celebrated for their accomplishments in image processing and computer vision, have also revealed substantial prowess in time series forecasting. Navigating through temporal data, CNNs utilize their convolutional layers to automatically and adaptively learn temporal patterns from the input data, offering an approach to uncovering subtle, underlying patterns embedded within a series of values.
|Model | Point Forecast | Probabilistic Forecast | Insample fitted values | Probabilistic fitted values |
|:----------------------------------------------------------|:----------------:|:----------------------:|:----------------------:|:----------------------------:|
|[`TimesNet`](../models.timesnet.html) |✅ |✅ |✅ |✅ |
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment