Object Character Recognition.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load\n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "\n",
    "# Input data files are available in the read-only \"../input/\" directory\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n",
    "\n",
    "import os\n",
    "for dirname, _, filenames in os.walk('/kaggle/input'):\n",
    "    for filename in filenames:\n",
    "        print(os.path.join(dirname, filename))\n",
    "\n",
    "# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n",
    "# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T05:58:48.151374Z",
     "iopub.status.busy": "2024-11-19T05:58:48.151044Z",
     "iopub.status.idle": "2024-11-19T05:58:48.155922Z",
     "shell.execute_reply": "2024-11-19T05:58:48.155061Z",
     "shell.execute_reply.started": "2024-11-19T05:58:48.151346Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hello World\n"
     ]
    }
   ],
   "source": [
    "print(\"Hello World\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T05:58:54.615604Z",
     "iopub.status.busy": "2024-11-19T05:58:54.615275Z",
     "iopub.status.idle": "2024-11-19T05:58:55.101995Z",
     "shell.execute_reply": "2024-11-19T05:58:55.101234Z",
     "shell.execute_reply.started": "2024-11-19T05:58:54.615573Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Path to dataset files: /kaggle/input/ocr-dataset\n"
     ]
    }
   ],
   "source": [
    "# Training Dataset\n",
    "import kagglehub\n",
    "\n",
    "# Download latest version\n",
    "path = kagglehub.dataset_download(\"harieh/ocr-dataset\")\n",
    "\n",
    "print(\"Path to dataset files:\", path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T05:59:07.070472Z",
     "iopub.status.busy": "2024-11-19T05:59:07.070152Z",
     "iopub.status.idle": "2024-11-19T06:01:26.539453Z",
     "shell.execute_reply": "2024-11-19T06:01:26.538330Z",
     "shell.execute_reply.started": "2024-11-19T05:59:07.070443Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mounting files to /kaggle/input/standard-ocr-dataset...\n",
      "Path to dataset files: /kaggle/input/standard-ocr-dataset\n"
     ]
    }
   ],
   "source": [
    "# Testing Dataset\n",
    "import kagglehub\n",
    "\n",
    "# Download latest version\n",
    "path = kagglehub.dataset_download(\"preatcher/standard-ocr-dataset\")\n",
    "\n",
    "print(\"Path to dataset files:\", path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T06:47:40.125555Z",
     "iopub.status.busy": "2024-11-19T06:47:40.125192Z",
     "iopub.status.idle": "2024-11-19T06:47:40.129690Z",
     "shell.execute_reply": "2024-11-19T06:47:40.128808Z",
     "shell.execute_reply.started": "2024-11-19T06:47:40.125523Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# import os\n",
    "\n",
    "# # walk through 10 classes of data\n",
    "# for dirpath, dirnames, filenames in os.walk('/kaggle/input/ocr-dataset/dataset'):\n",
    "#     print(f\"There are {len(dirnames)} directories and {len(filenames)} images in {dirpath}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T06:48:00.718864Z",
     "iopub.status.busy": "2024-11-19T06:48:00.718494Z",
     "iopub.status.idle": "2024-11-19T06:48:00.723021Z",
     "shell.execute_reply": "2024-11-19T06:48:00.722129Z",
     "shell.execute_reply.started": "2024-11-19T06:48:00.718834Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "train_dir = \"kaggle/input/standard-ocr-dataset/data/training_data\"\n",
    "test_dir = \"kaggle/input/standard-ocr-dataset/data/testing_data\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T06:48:01.612929Z",
     "iopub.status.busy": "2024-11-19T06:48:01.612564Z",
     "iopub.status.idle": "2024-11-19T06:48:01.618046Z",
     "shell.execute_reply": "2024-11-19T06:48:01.617182Z",
     "shell.execute_reply.started": "2024-11-19T06:48:01.612901Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Lets view some images\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.image as mpimg\n",
    "import random\n",
    "import os\n",
    "\n",
    "def view_random_image(dir, type):\n",
    "    location = dir + '/' + type\n",
    "    random_image = random.sample(os.listdir(location), 1)\n",
    "    print(random_image)\n",
    "    img = mpimg.imread(location + '/' + random_image[0])\n",
    "    plt.imshow(img)\n",
    "    plt.title(type)\n",
    "    plt.axis('off');\n",
    "    return img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T06:48:33.623781Z",
     "iopub.status.busy": "2024-11-19T06:48:33.622937Z",
     "iopub.status.idle": "2024-11-19T06:48:33.734229Z",
     "shell.execute_reply": "2024-11-19T06:48:33.733386Z",
     "shell.execute_reply.started": "2024-11-19T06:48:33.623749Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['7019.png']\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "img = view_random_image('kaggle/input/standard-ocr-dataset/data/training_data', 'Z')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:43:02.109976Z",
     "iopub.status.busy": "2024-11-19T07:43:02.109272Z",
     "iopub.status.idle": "2024-11-19T07:43:02.198745Z",
     "shell.execute_reply": "2024-11-19T07:43:02.197695Z",
     "shell.execute_reply.started": "2024-11-19T07:43:02.109935Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['28813.png']\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "img2 = view_random_image('kaggle/input/standard-ocr-dataset/data/testing_data', 'Z')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T06:48:45.317094Z",
     "iopub.status.busy": "2024-11-19T06:48:45.316308Z",
     "iopub.status.idle": "2024-11-19T06:48:45.321355Z",
     "shell.execute_reply": "2024-11-19T06:48:45.320446Z",
     "shell.execute_reply.started": "2024-11-19T06:48:45.317058Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Standard import\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.layers import Dense, Conv2D, MaxPool2D, Flatten\n",
    "from tensorflow.keras import Sequential\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.losses import CategoricalCrossentropy\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T06:48:45.323112Z",
     "iopub.status.busy": "2024-11-19T06:48:45.322864Z",
     "iopub.status.idle": "2024-11-19T06:48:45.336058Z",
     "shell.execute_reply": "2024-11-19T06:48:45.335391Z",
     "shell.execute_reply.started": "2024-11-19T06:48:45.323089Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Lets preprocess the data\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:32:10.713155Z",
     "iopub.status.busy": "2024-11-19T07:32:10.712450Z",
     "iopub.status.idle": "2024-11-19T07:32:10.717127Z",
     "shell.execute_reply": "2024-11-19T07:32:10.716168Z",
     "shell.execute_reply.started": "2024-11-19T07:32:10.713123Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "train_datagen = ImageDataGenerator(rescale = 1/255. )\n",
    "test_datagen = ImageDataGenerator(rescale = 1/255. )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:32:23.493138Z",
     "iopub.status.busy": "2024-11-19T07:32:23.492801Z",
     "iopub.status.idle": "2024-11-19T07:32:23.499911Z",
     "shell.execute_reply": "2024-11-19T07:32:23.498983Z",
     "shell.execute_reply.started": "2024-11-19T07:32:23.493108Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Lets get the class names\n",
    "import pathlib\n",
    "import numpy as np\n",
    "\n",
    "data_dir = pathlib.Path(train_dir)\n",
    "class_names = np.array(sorted([item.name for item in data_dir.glob('*')]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:32:30.524129Z",
     "iopub.status.busy": "2024-11-19T07:32:30.523750Z",
     "iopub.status.idle": "2024-11-19T07:32:30.528746Z",
     "shell.execute_reply": "2024-11-19T07:32:30.527934Z",
     "shell.execute_reply.started": "2024-11-19T07:32:30.524098Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "36\n"
     ]
    }
   ],
   "source": [
    "print(len(class_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:32:33.942828Z",
     "iopub.status.busy": "2024-11-19T07:32:33.942445Z",
     "iopub.status.idle": "2024-11-19T07:33:23.723266Z",
     "shell.execute_reply": "2024-11-19T07:33:23.722369Z",
     "shell.execute_reply.started": "2024-11-19T07:32:33.942798Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 20628 images belonging to 36 classes.\n",
      "Found 1008 images belonging to 36 classes.\n"
     ]
    }
   ],
   "source": [
    "train_data = train_datagen.flow_from_directory(\n",
    "    train_dir,\n",
    "    target_size = (224, 224),\n",
    "    batch_size = 32,\n",
    "    class_mode = 'categorical'\n",
    ")\n",
    "\n",
    "test_data = test_datagen.flow_from_directory(\n",
    "    test_dir, \n",
    "    target_size = (224, 224),\n",
    "    batch_size = 32,\n",
    "    class_mode = 'categorical'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:33:23.724887Z",
     "iopub.status.busy": "2024-11-19T07:33:23.724582Z",
     "iopub.status.idle": "2024-11-19T07:33:23.729367Z",
     "shell.execute_reply": "2024-11-19T07:33:23.728504Z",
     "shell.execute_reply.started": "2024-11-19T07:33:23.724861Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['0' '1' '2' '3' '4' '5' '6' '7' '8' '9' 'A' 'B' 'C' 'D' 'E' 'F' 'G' 'H'\n",
      " 'I' 'J' 'K' 'L' 'M' 'N' 'O' 'P' 'Q' 'R' 'S' 'T' 'U' 'V' 'W' 'X' 'Y' 'Z']\n"
     ]
    }
   ],
   "source": [
    "print(class_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:33:23.730618Z",
     "iopub.status.busy": "2024-11-19T07:33:23.730367Z",
     "iopub.status.idle": "2024-11-19T07:33:23.778177Z",
     "shell.execute_reply": "2024-11-19T07:33:23.777583Z",
     "shell.execute_reply.started": "2024-11-19T07:33:23.730595Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\anura\\AppData\\Roaming\\Python\\Python312\\site-packages\\keras\\src\\layers\\convolutional\\base_conv.py:107: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "# Lets create the custom model\n",
    "customModel = Sequential([\n",
    "    Conv2D(filters= 10, kernel_size = (3,3), padding='valid', activation = 'relu', input_shape = (224, 224, 3)),\n",
    "    Conv2D(10, 3, activation = 'relu'),\n",
    "    MaxPool2D(pool_size = 2),\n",
    "    Conv2D(10, 3, activation = 'relu'),\n",
    "    Conv2D(10, 3, activation = 'relu'),\n",
    "    MaxPool2D(),\n",
    "    Flatten(),\n",
    "    Dense(36, activation = 'softmax')\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:33:23.780975Z",
     "iopub.status.busy": "2024-11-19T07:33:23.780134Z",
     "iopub.status.idle": "2024-11-19T07:33:23.787591Z",
     "shell.execute_reply": "2024-11-19T07:33:23.786946Z",
     "shell.execute_reply.started": "2024-11-19T07:33:23.780949Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "customModel.compile(\n",
    "    loss = CategoricalCrossentropy(),\n",
    "    optimizer = Adam(),\n",
    "    metrics = ['accuracy']\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:33:23.789429Z",
     "iopub.status.busy": "2024-11-19T07:33:23.788610Z",
     "iopub.status.idle": "2024-11-19T07:37:08.638536Z",
     "shell.execute_reply": "2024-11-19T07:37:08.637840Z",
     "shell.execute_reply.started": "2024-11-19T07:33:23.789395Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\anura\\AppData\\Roaming\\Python\\Python312\\site-packages\\keras\\src\\trainers\\data_adapters\\py_dataset_adapter.py:121: UserWarning: Your `PyDataset` class should call `super().__init__(**kwargs)` in its constructor. `**kwargs` can include `workers`, `use_multiprocessing`, `max_queue_size`. Do not pass these arguments to `fit()`, as they will be ignored.\n",
      "  self._warn_if_super_not_called()\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m368/645\u001b[0m \u001b[32m━━━━━━━━━━━\u001b[0m\u001b[37m━━━━━━━━━\u001b[0m \u001b[1m3:52\u001b[0m 840ms/step - accuracy: 0.6901 - loss: 1.2393"
     ]
    }
   ],
   "source": [
    "customHistory = customModel.fit(\n",
    "    train_data,\n",
    "    epochs = 5,\n",
    "    validation_data = test_data\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:01:01.830601Z",
     "iopub.status.busy": "2024-11-19T07:01:01.830340Z",
     "iopub.status.idle": "2024-11-19T07:01:01.836412Z",
     "shell.execute_reply": "2024-11-19T07:01:01.835561Z",
     "shell.execute_reply.started": "2024-11-19T07:01:01.830576Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "def plot_loss_curves(history):\n",
    "    loss = history.history['loss']\n",
    "    val_loss = history.history['val_loss']\n",
    "    epochs = range(len(loss))\n",
    "\n",
    "    accuracy = history.history['accuracy']\n",
    "    val_accuracy = history.history['val_accuracy']\n",
    "\n",
    "    plt.plot(epochs, loss, label = 'Training Loss')\n",
    "    plt.plot(epochs, val_loss, label = 'Validation Loss')\n",
    "    plt.xlabel(\"Epochs\")\n",
    "    plt.title(\"Loss\")\n",
    "    plt.legend();\n",
    "\n",
    "    plt.figure()\n",
    "    plt.plot(epochs, accuracy, label = 'Training Accuracy')\n",
    "    plt.plot(epochs, val_accuracy, label = 'Validation Accuracy')\n",
    "    plt.xlabel(\"Epochs\")\n",
    "    plt.title(\"Accuracy\")\n",
    "    plt.legend()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:37:08.639918Z",
     "iopub.status.busy": "2024-11-19T07:37:08.639681Z",
     "iopub.status.idle": "2024-11-19T07:37:09.181686Z",
     "shell.execute_reply": "2024-11-19T07:37:09.180917Z",
     "shell.execute_reply.started": "2024-11-19T07:37:08.639894Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAHHCAYAAABtF1i4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAABXQElEQVR4nO3deXxU1f3/8dfMZN8DgSRAJIQECGuUrYCyRkEtBasFFQVxR6BSalV+VtH67RdttdIK4lahdSlUq9avKBh2BRQEAgHCFgMEyMKaDbLN3N8fEwYCAZKQ5GaS9/PxmIfJnTMzn8Mwzptzzj3XYhiGgYiIiIhJrGYXICIiIk2bwoiIiIiYSmFERERETKUwIiIiIqZSGBERERFTKYyIiIiIqRRGRERExFQKIyIiImIqhRERERExlcKIiIiImEphRESuyoIFC7BYLPz4449mlyIibkphREREREylMCIiIiKmUhgRkTq3ZcsWbr75ZoKCgggICGDYsGF8//33FdqUlpbywgsvEBcXh4+PD82bN+f6668nKSnJ1SYrK4uJEyfSpk0bvL29iYyMZNSoUezfv7+eeyQitcnD7AJEpHHbsWMHN9xwA0FBQTz55JN4enry1ltvMXjwYFavXk3fvn0BeP7555k1axYPPvggffr0IS8vjx9//JHNmzdz4403AnD77bezY8cOpk6dSnR0NDk5OSQlJXHw4EGio6NN7KWIXA2LYRiG2UWIiPtasGABEydOZOPGjfTq1eui+2+77Ta++uorUlNTiYmJASAzM5OOHTty7bXXsnr1agASEhJo06YNX375ZaWvc+rUKUJDQ/nzn//ME088UXcdEpF6p2kaEakzdrudb775htGjR7uCCEBkZCR333033333HXl5eQCEhISwY8cO9u7dW+lz+fr64uXlxapVqzh58mS91C8i9UNhRETqzNGjRzl9+jQdO3a86L74+HgcDgcZGRkA/OEPf+DUqVN06NCBbt268bvf/Y5t27a52nt7e/Pyyy/z9ddfEx4ezsCBA/nTn/5EVlZWvfVHROqGwoiINAgDBw4kLS2N9957j65du/Luu+9y3XXX8e6777raTJs2jT179jBr1ix8fHx49tlniY+PZ8uWLSZWLiJXS2FEROpMixYt8PPzY/fu3Rfdt2vXLqxWK1FRUa5jzZo1Y+LEifzrX/8iIyOD7t278/zzz1d4XPv27fntb3/LN998w/bt2ykpKeHVV1+t666ISB1SGBGROmOz2bjpppv473//W+H02+zsbD766COuv/56goKCADh+/HiFxwYEBBAbG0txcTEAp0+fpqioqEKb9u3bExgY6GojIu5Jp/aKSK147733WLJkyUXHn3/+eZKSkrj++ut57LHH8PDw4K233qK4uJg//elPrnadO3dm8ODB9OzZk2bNmvHjjz/yySefMGXKFAD27NnDsGHDGDNmDJ07d8bDw4PPPvuM7Oxs7rzzznrrp4jUPp3aKyJX5eypvZeSkZHB0aNHmTFjBmvXrsXhcNC3b1/++Mc/0q9fP1e7P/7xj3zxxRfs2bOH4uJi2rZty7333svvfvc7PD09OX78ODNnzmT58uVkZGTg4eFBp06d+O1vf8uvfvWr+uiqiNQRhRERERExldaMiIiIiKkURkRERMRUCiMiIiJiKoURERERMZXCiIiIiJhKYURERERM5RabnjkcDo4cOUJgYCAWi8XsckRERKQKDMMgPz+fVq1aYbVeevzDLcLIkSNHKly/QkRERNxHRkYGbdq0ueT9bhFGAgMDAWdnzl7HQkRERBq2vLw8oqKiXN/jl+IWYeTs1ExQUJDCiIiIiJu50hILLWAVERERUymMiIiIiKkURkRERMRUbrFmREREro7dbqe0tNTsMqSR8fT0xGazXfXzKIyIiDRihmGQlZXFqVOnzC5FGqmQkBAiIiKuah8whRERkUbsbBBp2bIlfn5+2jhSao1hGJw+fZqcnBwAIiMja/xcCiMiIo2U3W53BZHmzZubXY40Qr6+vgDk5OTQsmXLGk/ZaAGriEgjdXaNiJ+fn8mVSGN29u/X1axJUhgREWnkNDUjdak2/n4pjIiIiIipFEZERKRJiI6OZvbs2VVuv2rVKiwWi85EqgcKIyIi0qBYLJbL3p5//vkaPe/GjRt5+OGHq9y+f//+ZGZmEhwcXKPXqyqFniZ+Nk2Z3cG6tOMM7NDC7FJERKRcZmam6+dFixbx3HPPsXv3btexgIAA18+GYWC32/HwuPLXWYsW1ft/vZeXFxEREdV6jNRMkx0ZKSlzMOKv3zL+vQ1sOnDS7HJERKRcRESE6xYcHIzFYnH9vmvXLgIDA/n666/p2bMn3t7efPfdd6SlpTFq1CjCw8MJCAigd+/eLFu2rMLzXjhNY7FYePfdd7ntttvw8/MjLi6OL774wnX/hSMWCxYsICQkhKVLlxIfH09AQAAjRoyoEJ7Kysr49a9/TUhICM2bN+epp55iwoQJjB49usZ/HidPnmT8+PGEhobi5+fHzTffzN69e133HzhwgJEjRxIaGoq/vz9dunThq6++cj123LhxtGjRAl9fX+Li4pg/f36Na6krTTaMeHlYue6aEADmrtxnbjEiIvXEMAxOl5SZcjMMo9b68fTTT/PSSy+RmppK9+7dKSgo4JZbbmH58uVs2bKFESNGMHLkSA4ePHjZ53nhhRcYM2YM27Zt45ZbbmHcuHGcOHHiku1Pnz7NK6+8wvvvv8+aNWs4ePAgTzzxhOv+l19+mQ8//JD58+ezdu1a8vLy+Pzzz6+qr/fddx8//vgjX3zxBevXr8cwDG655RbXqbSTJ0+muLiYNWvWkJKSwssvv+waPXr22WfZuXMnX3/9NampqcybN4+wsLCrqqcuNOlpmkmDY/lk0yFW7Mph++Fcurau23lBERGznSm10/m5paa89s4/DMfPq3a+dv7whz9w4403un5v1qwZPXr0cP3+4osv8tlnn/HFF18wZcqUSz7Pfffdx1133QXA//7v//K3v/2NDRs2MGLEiErbl5aW8uabb9K+fXsApkyZwh/+8AfX/a+//jozZszgtttuA2DOnDmuUYqa2Lt3L1988QVr166lf//+AHz44YdERUXx+eef86tf/YqDBw9y++23061bNwBiYmJcjz948CDXXnstvXr1ApyjQw1Rkx0ZAWgX5s/IHq0AjY6IiLiTs1+uZxUUFPDEE08QHx9PSEgIAQEBpKamXnFkpHv37q6f/f39CQoKcm1vXhk/Pz9XEAHnFuhn2+fm5pKdnU2fPn1c99tsNnr27Fmtvp0vNTUVDw8P+vbt6zrWvHlzOnbsSGpqKgC//vWv+Z//+R8GDBjAzJkz2bZtm6vtpEmTWLhwIQkJCTz55JOsW7euxrXUpSY9MgIweUgs/00+wtfbs9iTnU+H8ECzSxIRqTO+njZ2/mG4aa9dW/z9/Sv8/sQTT5CUlMQrr7xCbGwsvr6+3HHHHZSUlFz2eTw9PSv8brFYcDgc1Wpfm9NPNfHggw8yfPhwFi9ezDfffMOsWbN49dVXmTp1KjfffDMHDhzgq6++IikpiWHDhjF58mReeeUVU2u+UJMeGQHoEB7IiC7O1dJvaHRERBo5i8WCn5eHKbe63Al27dq13Hfffdx2221069aNiIgI9u/fX2evV5ng4GDCw8PZuHGj65jdbmfz5s01fs74+HjKysr44YcfXMeOHz/O7t276dy5s+tYVFQUjz76KJ9++im//e1veeedd1z3tWjRggkTJvDBBx8we/Zs3n777RrXU1ea/MgIwJShsSzZkcUXW48wLbED0WH+V36QiIg0GHFxcXz66aeMHDkSi8XCs88+e9kRjroydepUZs2aRWxsLJ06deL111/n5MmTVQpiKSkpBAaeG523WCz06NGDUaNG8dBDD/HWW28RGBjI008/TevWrRk1ahQA06ZN4+abb6ZDhw6cPHmSlStXEh8fD8Bzzz1Hz5496dKlC8XFxXz55Zeu+xoShRGga+tghnRswcrdR5m3Ko2X7+h+5QeJiEiD8Ze//IX777+f/v37ExYWxlNPPUVeXl691/HUU0+RlZXF+PHjsdlsPPzwwwwfPrxKV7MdOHBghd9tNhtlZWXMnz+fxx9/nJ///OeUlJQwcOBAvvrqK9eUkd1uZ/LkyRw6dIigoCBGjBjBa6+9Bjj3SpkxYwb79+/H19eXG264gYULF9Z+x6+SxTB7sqsK8vLyCA4OJjc3l6CgoDp5jU0HTnL7vHV4WC2sfnIIrUN86+R1RETqS1FREenp6bRr1w4fHx+zy2mSHA4H8fHxjBkzhhdffNHscurE5f6eVfX7u8mvGTmrZ9tQ+rdvTpnD4K3VaWaXIyIibujAgQO888477Nmzh5SUFCZNmkR6ejp333232aU1aAoj55kyNBaAhRszyMkrMrkaERFxN1arlQULFtC7d28GDBhASkoKy5Yta5DrNBoSrRk5T7+Y5vRsG8qmAyd559ufeObWzld+kIiISLmoqCjWrl1rdhluRyMj57FYLK7RkQ++P8iJwsufny4iIiJXT2HkAoM7tKBr6yDOlNp577t0s8sRERFp9BRGLmCxWJgyJA6Af6zbT+6ZUpMrEhERadwURipxU+dwOoQHkF9cxj/X7Te7HBERkUZNYaQSVquFyUOca0f+vjadwuIykysSERFpvBRGLuHn3VvRLsyfU6dL+fCHA2aXIyIi0mgpjFyCzWph0mDnZaLfXpNOUand5IpERKQ6Bg8ezLRp01y/R0dHM3v27Ms+xmKx8Pnnn1/1a9fW8zQVCiOXcdu1rWkd4suxgmIWbcwwuxwRkSZh5MiRjBgxotL7vv32WywWC9u2bav2827cuJGHH374asur4PnnnychIeGi45mZmdx88821+loXWrBgASEhIXX6GvVFYeQyPG1WHi0fHXlzdRolZfV/BUgRkabmgQceICkpiUOHDl103/z58+nVqxfdu1f/gqYtWrTAz8+vNkq8ooiICLy9vevltRoDhZEr+FXPNrQM9CYzt4hPN1/8wRARkdr185//nBYtWrBgwYIKxwsKCvj444954IEHOH78OHfddRetW7fGz8+Pbt268a9//euyz3vhNM3evXsZOHAgPj4+dO7cmaSkpIse89RTT9GhQwf8/PyIiYnh2WefpbTUueXDggULeOGFF9i6dSsWiwWLxeKq+cJpmpSUFIYOHYqvry/Nmzfn4YcfpqCgwHX/fffdx+jRo3nllVeIjIykefPmTJ482fVaNXHw4EFGjRpFQEAAQUFBjBkzhuzsbNf9W7duZciQIQQGBhIUFETPnj358ccfAec1dkaOHEloaCj+/v506dKFr776qsa1XIm2g78CH08bDw+M4X8Wp/LGqjTu6NkGD5synIi4KcOA0tPmvLanH1gsV2zm4eHB+PHjWbBgAc888wyW8sd8/PHH2O127rrrLgoKCujZsydPPfUUQUFBLF68mHvvvZf27dvTp0+fK76Gw+Hgl7/8JeHh4fzwww/k5uZWWF9yVmBgIAsWLKBVq1akpKTw0EMPERgYyJNPPsnYsWPZvn07S5YsYdmyZQAEBwdf9ByFhYUMHz6cfv36sXHjRnJycnjwwQeZMmVKhcC1cuVKIiMjWblyJfv27WPs2LEkJCTw0EMPXbE/lfXvbBBZvXo1ZWVlTJ48mbFjx7Jq1SoAxo0bx7XXXsu8efOw2WwkJyfj6ekJwOTJkykpKWHNmjX4+/uzc+dOAgICql1HVSmMVMHdfa/hjVVpHDxxmv/bdoTbrm1jdkkiIjVTehr+t5U5r/3/joCXf5Wa3n///fz5z39m9erVDB48GHBO0dx+++0EBwcTHBzME0884Wo/depUli5dyr///e8qhZFly5axa9culi5dSqtWzj+P//3f/71oncfvf/9718/R0dE88cQTLFy4kCeffBJfX18CAgLw8PAgIiLikq/10UcfUVRUxD//+U/8/Z39nzNnDiNHjuTll18mPDwcgNDQUObMmYPNZqNTp07ceuutLF++vEZhZPny5aSkpJCenk5UVBQA//znP+nSpQsbN26kd+/eHDx4kN/97nd06tQJgLi4ONfjDx48yO233063bt0AiImJqXYN1aF/4leBn5cHD1zfDoA5K/bhcBgmVyQi0rh16tSJ/v3789577wGwb98+vv32Wx544AEA7HY7L774It26daNZs2YEBASwdOlSDh48WKXnT01NJSoqyhVEAPr163dRu0WLFjFgwAAiIiIICAjg97//fZVf4/zX6tGjhyuIAAwYMACHw8Hu3btdx7p06YLNZnP9HhkZSU5OTrVe6/zXjIqKcgURgM6dOxMSEkJqaioA06dP58EHHyQxMZGXXnqJtLQ0V9tf//rX/M///A8DBgxg5syZNVowXB01GhmZO3cuf/7zn8nKyqJHjx68/vrrVUqiCxcu5K677mLUqFFud8rT+H5teWt1GmlHC1myI4tbukWaXZKISPV5+jlHKMx67Wp44IEHmDp1KnPnzmX+/Pm0b9+eQYMGAfDnP/+Zv/71r8yePZtu3brh7+/PtGnTKCmpvQucrl+/nnHjxvHCCy8wfPhwgoODWbhwIa+++mqtvcb5zk6RnGWxWHA46u7Eieeff567776bxYsX8/XXXzNz5kwWLlzIbbfdxoMPPsjw4cNZvHgx33zzDbNmzeLVV19l6tSpdVJLtUdGFi1axPTp05k5cyabN2+mR48eDB8+/Irpbf/+/TzxxBPccMMNNS7WTIE+ntw3wDk68vqKfRiGRkdExA1ZLM6pEjNuVVgvcr4xY8ZgtVr56KOP+Oc//8n999/vWj+ydu1aRo0axT333EOPHj2IiYlhz549VX7u+Ph4MjIyyMzMdB37/vvvK7RZt24dbdu25ZlnnqFXr17ExcVx4EDFTTC9vLyw2y+/D1V8fDxbt26lsLDQdWzt2rVYrVY6duxY5Zqr42z/MjLObUuxc+dOTp06RefOnV3HOnTowG9+8xu++eYbfvnLXzJ//nzXfVFRUTz66KN8+umn/Pa3v+Wdd96pk1qhBmHkL3/5Cw899BATJ06kc+fOvPnmm/j5+bmG0ipjt9td6bKu553q0sT+0fh72UjNzGPFrpoNnYmISNUEBAQwduxYZsyYQWZmJvfdd5/rvri4OJKSkli3bh2pqak88sgjFc4UuZLExEQ6dOjAhAkT2Lp1K99++y3PPPNMhTZxcXEcPHiQhQsXkpaWxt/+9jc+++yzCm2io6NJT08nOTmZY8eOUVxcfNFrjRs3Dh8fHyZMmMD27dtZuXIlU6dO5d5773WtF6kpu91OcnJyhVtqaiqJiYl069aNcePGsXnzZjZs2MD48eMZNGgQvXr14syZM0yZMoVVq1Zx4MAB1q5dy8aNG4mPjwdg2rRpLF26lPT0dDZv3szKlStd99WFaoWRkpISNm3aRGJi4rknsFpJTExk/fr1l3zcH/7wB1q2bOma67uS4uJi8vLyKtwaglB/L+7p1xbQ6IiISH144IEHOHnyJMOHD6+wvuP3v/891113HcOHD2fw4MFEREQwevToKj+v1Wrls88+48yZM/Tp04cHH3yQP/7xjxXa/OIXv+A3v/kNU6ZMISEhgXXr1vHss89WaHP77bczYsQIhgwZQosWLSo9vdjPz4+lS5dy4sQJevfuzR133MGwYcOYM2dO9f4wKlFQUMC1115b4TZy5EgsFgv//e9/CQ0NZeDAgSQmJhITE8OiRYsAsNlsHD9+nPHjx9OhQwfGjBnDzTffzAsvvAA4Q87kyZOJj49nxIgRdOjQgTfeeOOq670Ui1GNb9QjR47QunVr1q1bV2Ghz5NPPsnq1av54YcfLnrMd999x5133klycjJhYWHcd999nDp16rJrRp5//nnXH8j5cnNzCQoKqmq5deJofjHXv7yC4jIHHzzQl+vjwkytR0TkUoqKikhPT6ddu3b4+PiYXY40Upf7e5aXl0dwcPAVv7/r9Gya/Px87r33Xt555x3Cwqr+pT1jxgxyc3Ndt/PnvMzWItCbu/pcA8DrK/aaXI2IiIj7q9bZNGFhYdhstovm5bKzsys9xzotLY39+/czcuRI17GzK4M9PDzYvXs37du3v+hx3t7eDXob3UcGxfDhDwf4If0EG9JP0KddM7NLEhERcVvVGhnx8vKiZ8+eLF++3HXM4XCwfPnySs/P7tSpEykpKRUW1vziF79gyJAhJCcnVzj/2Z1EBvtyR09n7XNW7jO5GhEREfdW7X1Gpk+fzoQJE+jVqxd9+vRh9uzZFBYWMnHiRADGjx9P69atmTVrFj4+PnTt2rXC489eYfDC4+5m0qD2/PvHDNbsOcrWjFP0iAoxuyQRERG3VO0wMnbsWI4ePcpzzz1HVlYWCQkJLFmyxHV60sGDB7FaG//Grtc092NUQis+3XyYOSv38c74XmaXJCJSKZ35J3WpNv5+VetsGrNUdTVufduXU8CNr63GMODrx28gPrLh1CYiYrfb2bNnDy1btqR58+ZmlyON1PHjx8nJyaFDhw4VtrOHqn9/60J5VyG2ZQC3dItk8bZM5q7cx5y7rzO7JBERF5vNRkhIiGuHbD8/P9cOpiJXyzAMTp8+TU5ODiEhIRcFkepQGLlKU4bEsnhbJotTMvnN0QLat6i7SyyLiFTX2TMda3rBNZErCQkJuexVi6tCYeQqxUcGkRgfzrLUbN5YmcarY3qYXZKIiIvFYiEyMpKWLVtSWlpqdjnSyHh6el7ViMhZCiO1YMrQWJalZvN58mGmJcYR1ax6V6YUEalrNputVr40ROpC4z/tpR4kRIVwQ1wYdofBvNVpZpcjIiLiVhRGasnUoXEAfPLjITJzz5hcjYiIiPtQGKklfdo1o0+7ZpTYHby95iezyxEREXEbCiO1aOrQWAD+teEgR/OLTa5GRETEPSiM1KLrY8PoERVCUamDv3+XbnY5IiIibkFhpBZZLBamDnGOjry/fj+nTpeYXJGIiEjDpzBSy4bFtyQ+MojCEjvz1+43uxwREZEGT2GkllksFqaUj47MX5tOfpE2GRIREbkchZE6MKJrBO1b+JNXVMb73x8wuxwREZEGTWGkDtisFiaXj468+206p0vKTK5IRESk4VIYqSO/6NGKqGa+nCgs4V8bMswuR0REpMFSGKkjHjYrjw12jo68vSaNolK7yRWJiIg0TAojdeiX17UmMtiH7LxiPtl0yOxyREREGiSFkTrk7WHjkYExAMxblUap3WFyRSIiIg2Pwkgdu7PPNYQFeHH41Bk+33LY7HJEREQaHIWROubjaeOhG5yjI2+sSsPuMEyuSEREpGFRGKkH437WlhA/T9KPFbI4JdPsckRERBoUhZF6EODtwf0D2gEwd8U+HBodERERcVEYqScT+kcT6O3B7ux8klKzzS5HRESkwVAYqSfBvp6M798WgDkr9mEYGh0REREBhZF6df+Advh62kg5nMvqPUfNLkdERKRBUBipR80DvBnX9xoAXtfoiIiICKAwUu8eGhiDl4eVTQdO8v1PJ8wuR0RExHQKI/UsPMiHsb2iAJizcq/J1YiIiJhPYcQEjwyKwcNqYe2+42w6cNLsckREREylMGKCNqF+/PK61gDMXbnP5GpERETMpTBikkmDY7FaYMWuHLYfzjW7HBEREdMojJikXZg/I3u0AjQ6IiIiTZvCiIkmD4kF4OvtWezJzje5GhEREXMojJioQ3ggI7pEAPCGRkdERKSJUhgx2ZShztGRL7YeYf+xQpOrERERqX8KIybr2jqYIR1b4DBg3qo0s8sRERGpdwojDcCUoXEA/GfzIQ6fOmNyNSIiIvVLYaQB6Nk2lP7tm1PmMHhrtUZHRESkaVEYaSDOrh1ZuDGDnLwik6sRERGpPwojDUS/mOb0bBtKSZmDd779yexyRERE6o3CSANhsVhcoyMffH+QE4UlJlckIiJSPxRGGpDBHVrQtXUQZ0rtvPddutnliIiI1AuFkQbEYrEwZYjzzJp/rNtP7plSkysSERGpewojDcxNncPpEB5AfnEZ/1y33+xyRERE6pzCSANjtVpc16z5+9p0CovLTK5IRESkbimMNEA/796KdmH+nDpdyoc/HDC7HBERkTqlMNIA2awWJg1uD8Dba9IpKrWbXJGIiEjdURhpoG67tjWtQ3w5VlDMoo0ZZpcjIiJSZxRGGihPm5VHy0dH3lydRkmZw+SKRERE6obCSAP2q55taBnoTWZuEZ9uPmR2OSIiInVCYaQB8/G08fDAGADeWJVGmV2jIyIi0vgojDRwd/e9hmb+Xhw8cZr/23bE7HJERERqncJIA+fn5cED17cDYM6KfTgchskViYiI1C6FETcwvl9bgnw8SDtayJIdWWaXIyIiUqsURtxAoI8n9w1wjo68vmIfhqHRERERaTwURtzExP7R+HvZSM3MY8WuHLPLERERqTUKI24i1N+Le/q1BTQ6IiIijYvCiBt58PoYvD2sJGecYu2+42aXIyIiUisURtxIi0Bv7upzDQCvr9hrcjUiIiK1Q2HEzTwyKAZPm4Uf0k+wIf2E2eWIiIhcNYURNxMZ7MsdPaMAmLNyn8nViIiIXD2FETc0aVB7bFYLa/YcZWvGKbPLERERuSoKI27omuZ+jEpoBWh0RERE3J/CiJt6bHAsFgsk7cwmNTPP7HJERERqTGHETcW2DOCWbpEAzNXoiIiIuDGFETc2ZUgsAItTMkk7WmByNSIiIjWjMOLG4iODSIwPxzDgjZVpZpcjIiJSIwojbm7KUOfoyOfJh8k4cdrkakRERKqvRmFk7ty5REdH4+PjQ9++fdmwYcMl23766af06tWLkJAQ/P39SUhI4P33369xwVJRQlQIN8SFYXcYzFut0REREXE/1Q4jixYtYvr06cycOZPNmzfTo0cPhg8fTk5O5VeSbdasGc888wzr169n27ZtTJw4kYkTJ7J06dKrLl6cpg6NA+CTHw+RmXvG5GpERESqx2JU8/Kvffv2pXfv3syZMwcAh8NBVFQUU6dO5emnn67Sc1x33XXceuutvPjii1Vqn5eXR3BwMLm5uQQFBVWn3CZjzFvr2ZB+gokDopk5sovZ5YiIiFT5+7taIyMlJSVs2rSJxMTEc09gtZKYmMj69euv+HjDMFi+fDm7d+9m4MCBl2xXXFxMXl5ehZtc3tTytSP/2nCQo/nFJlcjIiJSddUKI8eOHcNutxMeHl7heHh4OFlZWZd8XG5uLgEBAXh5eXHrrbfy+uuvc+ONN16y/axZswgODnbdoqKiqlNmk3R9bBg9okIoKnXw9+/SzS5HRESkyurlbJrAwECSk5PZuHEjf/zjH5k+fTqrVq26ZPsZM2aQm5vrumVkZNRHmW7NYrEwtXzfkffX7+fU6RKTKxIREakaj+o0DgsLw2azkZ2dXeF4dnY2ERERl3yc1WolNtb5RZmQkEBqaiqzZs1i8ODBlbb39vbG29u7OqUJMCy+JfGRQaRm5jF/7X5+c2MHs0sSERG5omqNjHh5edGzZ0+WL1/uOuZwOFi+fDn9+vWr8vM4HA6Ki7WuobZZLBbXrqzz16aTX1RqckUiIiJXVu1pmunTp/POO+/wj3/8g9TUVCZNmkRhYSETJ04EYPz48cyYMcPVftasWSQlJfHTTz+RmprKq6++yvvvv88999xTe70QlxFdI2jfwp+8ojLe//6A2eWIiIhcUbWmaQDGjh3L0aNHee6558jKyiIhIYElS5a4FrUePHgQq/VcxiksLOSxxx7j0KFD+Pr60qlTJz744APGjh1be70QF5vVwuQhsUz/91be/Tad+/pH4+dV7bdZRESk3lR7nxEzaJ+R6imzOxjy6ioyTpzh2Z935oHr25ldkoiINEF1ss+IuAcPm5XHBjvXjry9Jo2iUrvJFYmIiFyawkgj9cvrWhMZ7EN2XjGfbDpkdjkiIiKXpDDSSHl72HhkYAwA81alUWp3mFyRiIhI5RRGGrE7+1xDWIAXh0+d4fMth80uR0REpFIKI42Yj6eNh25wjo68sSoNu6PBr1UWEZEmSGGkkRv3s7aE+HmSfqyQxSmZZpcjIiJyEYWRRi7A24P7BzhP7Z27Yh8OjY6IiEgDozDSBEzoH02gtwe7s/NJSs2+8gNERETqkcJIExDs68n4/m0BmLNiH26wz52IiDQhCiNNxP0D2uHraSPlcC6r9xw1uxwREREXhZEmonmAN+P6XgPA6xodERGRBkRhpAl5aGAMXh5WNh04yfc/nTC7HBEREUBhpEkJD/JhbK8oAOas3GtyNSIiIk4KI03MI4Ni8LBaWLvvOJsOnDS7HBEREYWRpqZNqB+/vK41AHNX7jO5GhEREYWRJmnS4FisFlixK4fth3PNLkdERJo4hZEmqF2YPyN7tAI0OiIiIuZTGGmiJg+JBeDr7Vnsyc43uRoREWnKFEaaqA7hgYzoEgHAGxodEREREymMNGFThjpHR77YeoT9xwpNrkZERJoqhZEmrGvrYIZ0bIHDgHmr0swuR0REmiiFkSZuytA4AP6z+RCHT50xuRoREWmKFEaauJ5tQ+nfvjllDoO3Vmt0RERE6p/CiLjWjizcmEFOXpHJ1YiISFOjMCL0i2lOz7ahlJQ5eOfbn8wuR0REmhiFEcFisbhGRz74/iAnCktMrkhERJoShREBYHCHFnRtHcSZUjvvfZdudjkiItKEKIwIUD46MsR5Zs0/1u0n90ypyRWJiEhToTAiLjd1DqdDeAD5xWX8c91+s8sREZEmQmFEXKxWi+uaNX9fm05hcZnJFYmISFOgMCIV/Lx7K9qF+XPqdCkf/nDA7HJERKQJUBiRCmxWC5MGtwfg7TXpFJXaTa5IREQaO4URucht17amdYgvxwqKWbQxw+xyRESkkVMYkYt42qw8Wj468ubqNErKHCZXJCIijZnCiFTqVz3b0DLQm8zcIj7dfMjsckREpBFTGJFK+XjaeHhgDABvrEqjzK7RERERqRsKI3JJd/e9hmb+Xhw8cZr/23bE7HJERKSRUhiRS/Lz8uCB69sBMGfFPhwOw+SKRESkMVIYkcsa368tQT4epB0tZMmOLLPLERGRRkhhRC4r0MeT+wY4R0deX7EPw9DoiIiI1C6FEbmiif2j8feykZqZx4pdOWaXIyIijYzCiFxRqL8X9/RrC2h0REREap/CiFTJg9fH4O1hJTnjFGv3HTe7HBERaUQURqRKWgR6c1efawB4fcVek6sREZHGRGFEquyRQTF42iz8kH6CDeknzC5HREQaCYURqbLIYF/u6BkFwJyV+0yuRkREGguFEamWSYPaY7NaWLPnKFszTpldjoiINAIKI1It1zT3Y1RCK0CjIyIiUjsURqTaHhsci8UCSTuzSc3MM7scERFxcwojUm2xLQO4pVskAHM1OiIiIldJYURqZMqQWAAWp2SSdrTA5GpERMSdKYxIjcRHBpEYH45hwBsr08wuR0RE3JjCiNTYlKHO0ZHPkw+TceK0ydWIiIi7UhiRGkuICuGGuDDsDoN5qzU6IiIiNaMwIldl6tA4AD758RCZuWdMrkZERNyRwohclT7tmtGnXTNK7A7eXvOT2eWIiIgbUhiRqza1fO3IvzYc5Gh+scnViIiIu1EYkat2fWwYPaJCKCp18Pfv0s0uR0RE3IzCiFw1i8XC1PJ9R95fv59Tp0tMrkhERNyJwojUimHxLYmPDKKwxM78tfvNLkdERNyIwojUCovF4tqVdf7adPKLSk2uSERE3IXCiNSaEV0jaN/Cn7yiMt7//oDZ5YiIiJtQGJFaY7NamFw+OvLut+mcLikzuSIREXEHCiNSq37RoxVRzXw5UVjCvzZkmF2OiIi4AYURqVUeNiuPDXaOjry9Jo2iUrvJFYmISEOnMCK17pfXtSYy2IfsvGI+2XTI7HJERKSBUxiRWuftYeORgTEAzFuVRqndYXJFIiLSkCmMSJ24s881hAV4cfjUGT7fctjsckREpAFTGJE64eNp46EbnKMjb6xKw+4wTK5IREQaqhqFkblz5xIdHY2Pjw99+/Zlw4YNl2z7zjvvcMMNNxAaGkpoaCiJiYmXbS+Nx7iftSXEz5P0Y4UsTsk0uxwREWmgqh1GFi1axPTp05k5cyabN2+mR48eDB8+nJycnErbr1q1irvuuouVK1eyfv16oqKiuOmmmzh8WEP3jV2Atwf3D2gHwNwV+3BodERERCphMQyjWt8Qffv2pXfv3syZMwcAh8NBVFQUU6dO5emnn77i4+12O6GhocyZM4fx48dX6TXz8vIIDg4mNzeXoKCg6pQrJss9U8r1L60gv7iMt+7tyfAuEWaXJCIi9aSq39/VGhkpKSlh06ZNJCYmnnsCq5XExETWr19fpec4ffo0paWlNGvW7JJtiouLycvLq3AT9xTs68n4/m0BmLNiH9XMviIi0gRUK4wcO3YMu91OeHh4hePh4eFkZWVV6TmeeuopWrVqVSHQXGjWrFkEBwe7blFRUdUpUxqY+we0w9fTRsrhXFbvOWp2OSIi0sDU69k0L730EgsXLuSzzz7Dx8fnku1mzJhBbm6u65aRoW3F3VnzAG/G9b0GgNc1OiIiIheoVhgJCwvDZrORnZ1d4Xh2djYREZdfC/DKK6/w0ksv8c0339C9e/fLtvX29iYoKKjCTdzbQwNj8PKwsunASb7/6YTZ5YiISANSrTDi5eVFz549Wb58ueuYw+Fg+fLl9OvX75KP+9Of/sSLL77IkiVL6NWrV82rFbcVHuTD2F7O6bY5K/eaXI2IiDQk1Z6mmT59Ou+88w7/+Mc/SE1NZdKkSRQWFjJx4kQAxo8fz4wZM1ztX375ZZ599lnee+89oqOjycrKIisri4KCgtrrhbiFRwbF4GG1sHbfcTYdOGl2OSIi0kBUO4yMHTuWV155heeee46EhASSk5NZsmSJa1HrwYMHycw8t8HVvHnzKCkp4Y477iAyMtJ1e+WVV2qvF+IW2oT68cvrWgMwd+U+k6sREZGGotr7jJhB+4w0HunHChn26iocBnw59Xq6tg42uyQREakjdbLPiMjVahfmz8gerQCNjoiIiJPCiNS7yUNiAfh6exZ7svNNrkZERMymMCL1rkN4ICPKt4V/Q6MjIiJNnsKImGLKUOfoyBdbj7D/WKHJ1YiIiJkURsQUXVsHM6RjCxwGzFuVZnY5IiJiIoURMc2UoXEA/GfzIQ6fOmNyNSIiYhaFETFNz7ah9G/fnDKHwVurNToiItJUKYyIqc6uHVm4MYOcvCKTqxERETMojIip+sU0p2fbUErKHLzz7U9mlyMiIiZQGBFTWSwW1+jIB98f5ERhickViYhIfVMYEdMN7tCCrq2DOFNq573v0s0uR0RE6pnCiJjOYrEwZYjzzJp/rNtP7plSkysSEZH6pDAiDcJNncPpEB5AfnEZ/1y33+xyRESkHimMSINgtVpc16z5+9p0CovLTK5IRETqi8KINBg/796KdmH+nDpdyoc/HDC7HBERqScKI9Jg2KwWJg1uD8Dba9IpKrWbXJGIiNQHhRFpUG67tjWtQ3w5VlDMoo0ZZpcjIiL1QGFEGhRPm5VHy0dH3lydRkmZw+SKRESkrimMSIPzq55taBnoTWZuEZ9uPmR2OSIiUscURqTB8fG08fDAGADeWJVGmV2jIyIijZnCiDRId/e9hmb+Xhw8cZr/23bE7HJERKQOKYxIg+Tn5cED17cDYM6KfTgchskViYhIXVEYkQZrfL+2BPl4kHa0kCU7sswuR0RE6ojCiDRYgT6e3DfAOTry+op9GIZGR0REGiOFEWnQJvaPxt/LRmpmHit25ZhdjoiI1AGFEWnQQv29uKdfW0CjIyIijZXCiDR4D14fg7eHleSMU6zdd9zsckREpJYpjEiD1yLQm7v6XAPA6yv2mlyNiIjUNoURcQuPDIrB02bhh/QTbEg/YXY5IiJSixRGxC1EBvtyR88oAOas3GdyNSIiUpsURsRtTBrUHpvVwpo9R9maccrsckREpJYojIjbuKa5H6MSWgEaHRERaUwURsStPDY4FosFknZmk5qZZ3Y5IiJSCxRGxK3Etgzglm6RAMzV6IiISKOgMCJuZ8qQWAAWp2SSdrTA5GpERORqKYyI24mPDCIxPhzDgDdWppldjoiIXCWFEXFLU4Y6R0c+Tz5MxonTJlcjIiJXQ2FE3FJCVAg3xIVhdxjMXalr1oiIuDOFEXFbU4fGAbBwYwZ3vv09P/yk69aIiLgjhRFxW33aNePXw+Lwsln5If0EY9/+nnHvfs+mA9ouXkTEnVgMNxjfzsvLIzg4mNzcXIKCgswuRxqYI6fOMHflPv79Ywaldudf54EdWvCbxDiuvSbU5OpERJquqn5/K4xIo5Fx4jRzV+7j402HsDucf62HdmrJbxI70K1NsMnViYg0PQoj0mQdPH6av63Yy6ebD1GeSbixczjTEuPo0kqhRESkviiMSJOXfqyQvy3fy3+TD7tCyc1dI5iW2IGOEYHmFici0gQojIiU25eTz1+X7+PLbUcwDLBY4NZukUxLjCO2pUKJiEhdURgRucCe7Hz+umwvi1MyAWcoGdWjFb8eFkdMiwCTqxMRaXwURkQuITUzj9nL9rB0RzYAVgvcdm0bfj0slrbN/U2uTkSk8VAYEbmC7Ydzmb1sD8tScwCwWS3ccV0bpgyNJaqZn8nViYi4P4URkSpKzjjF7GV7WLX7KAAeVgtjekcxeUgsrUN8Ta5ORMR9KYyIVNOmAyeZvWwP3+49BoCXzcqdfaJ4bHAsEcE+JlcnIuJ+FEZEamhD+gleS9rD+vJr3Xh5WBnX9xomDW5Py0CFEhGRqlIYEblK69KO8VrSHjbuPwmAj6eVe/q25dHB7QkL8Da5OhGRhk9hRKQWGIbB2n3HeTVpN1sOngLA19PG+P5teWRge5r5e5lboIhIA6YwIlKLDMNg9Z6jvJa0h62HcgHw97Jx34BoHrohhhA/hRIRkQspjIjUAcMwWLErh78k7WHHkTwAAr09mHh9Ox64vh3Bvp4mVygi0nAojIjUIcMw+GZnNq8l7WFXVj4AQT4ePHRDDPcNiCbQR6FERERhRKQeOBwGS3Zk8VrSHvbmFAAQ4ufpDCX9o/H39jC5QhER8yiMiNQju8NgcUoms5ft4aejhQA08/fikYEx3NuvLX5eCiUi0vQojIiYwO4w+GLrYf66bC/7j58GICzAi0cHteeen7XFx9NmcoUiIvVHYUTERGV2B59tOczfVuwl48QZAFoGevPY4Pbc2ecahRIRaRIURkQagFK7g/9sOsTrK/Zx+JQzlEQE+TB5aCxjerXB20OhREQaL4URkQakpMzBv3/MYO7KfWTmFgHQOsSXKUNjuaNnGzxtVpMrFBGpfQojIg1QUamdRRudoSQnvxiAqGa+TB0axy+vbY2HQomINCIKIyINWFGpnQ9/OMi8Vfs4VlACQHRzP349LI5RCa2xWS0mVygicvUURkTcwOmSMj74/gBvrv6JE4XOUBLTwp/Hh8Xx8+6tFEpExK0pjIi4kcLiMv6xfj9vr/mJU6dLAYhrGcC0xA7c3DUCq0KJiLghhRERN5RfVMo/1jlDSV5RGQCdIgKZltiB4V3CsVgUSkTEfSiMiLix3DOlvPddOu99l05+sTOUdGkVxG8SOzAsvqVCiYi4hap+f9do6f7cuXOJjo7Gx8eHvn37smHDhku23bFjB7fffjvR0dFYLBZmz55dk5cUaVKCfT35zY0d+PapIUwZEou/l40dR/J48J8/MmruWlbuzsEN/h0hIlIl1Q4jixYtYvr06cycOZPNmzfTo0cPhg8fTk5OTqXtT58+TUxMDC+99BIRERFXXbBIUxLi58UTwzvy7VNDeXRQe3w9bWw7lMvE+Rv55bx1fLv3qEKJiLi9ak/T9O3bl969ezNnzhwAHA4HUVFRTJ06laeffvqyj42OjmbatGlMmzatWkVqmkbE6VhBMW+tTuP97w9QVOoAoHd0KL+5sQP924eZXJ2ISEV1Mk1TUlLCpk2bSExMPPcEViuJiYmsX7++5tWKSJWEBXjzzK2dWfPkEO4f0A4vDysb95/k7nd+4M631/PDT8fNLlFEpNqqFUaOHTuG3W4nPDy8wvHw8HCysrJqraji4mLy8vIq3ETknJaBPjw3sjNrfjeECf3a4mWz8v1PJxj79vfc8+4PbDpwwuwSRUSqrEHuPT1r1iyCg4Ndt6ioKLNLEmmQIoJ9eGFUV1b9bjDj+l6Dp83Cd/uOcfu89Ux4bwPJGafMLlFE5IqqFUbCwsKw2WxkZ2dXOJ6dnV2ri1NnzJhBbm6u65aRkVFrzy3SGLUK8eWPt3VjxW8Hc2fvKGxWC6v3HGX03LU8sGAj2w/nml2iiMglVSuMeHl50bNnT5YvX+465nA4WL58Of369au1ory9vQkKCqpwqxPH9sLR3ZCfDWXFdfMaIvUoqpkfL93enRW/HcQdPdtgtcDyXTn8/PXveOifP7LjiEKJiDQ8HtV9wPTp05kwYQK9evWiT58+zJ49m8LCQiZOnAjA+PHjad26NbNmzQKci1537tzp+vnw4cMkJycTEBBAbGxsLXalBv7vcTiw9tzvHj7gEwK+IeATfImfy3+/8GevANBGVNJAtG3uzyu/6sFjg9vz+op9fJ58mKSd2STtzObmrhFMS+xAx4hAs8sUEQFquAPrnDlz+POf/0xWVhYJCQn87W9/o2/fvgAMHjyY6OhoFixYAMD+/ftp167dRc8xaNAgVq1aVaXXq7NTez+4Aw5thKJc4Cr3arDYKg8pVQ03VtvVvb7IZezLyeevy/fx5bYjGIYzN9/aLZJpiXHEtlQoEZG6oe3gq8PhgOI8ZygpOgVnTl3wc/nvlf185hQ4Sq++Bu+g6o/GnL3fw/vqX1+ahN1Z+fx1+R6+SnGe/WaxwKgerfj1sDhiWgSYXJ2INDYKI/XFMKD0TOUh5ZLh5ryfSwuvvoYqTS9dItxoeqlJ2nkkj9nL9vDNTudidKsFbru2Db8eFkvb5v4mVycijYXCiLsoK3GOyrhGXU5degTmokBTx9NLl51qCtH0UiOQciiX2cv2sHyX83IONquFO65rw5ShsUQ18zO5OhFxdwojTUGDn166wlSTp8/Vv77UiuSMU7yWtIfVe44C4GG1MKZ3FFOGxNIqxNfk6kTEXSmMyOWdP71U1Sml8wNNbU4v1WThr6aX6sSmAyd4LWkv3+07BoCXzcqdfaJ4bHAsEcEKjyJSPQojUrca0vRSi07Q407oMEKLeWvJDz8d5y9Je/gh3bmtvJeHlXF9r2HS4Pa0DFQoEZGqURiRhqvS6aVTl55Squr0km8odBsDCXdDZA+NnNSCdWnHeC1pDxv3nwTAx9PKvT9ryyOD2hMWoOAnIpenMCKN04XTS2dOwL5lsHUh5Geea9eyC1w7zhlOAlqYVW2jYBgG3+07xl+S9rDl4CkAfD1tTOgfzcMDY2jm72VugSLSYCmMSNPisEPaSkj+EHYtBnv59v5WD4gb7hwtibsJPPTFWVOGYbBqz1FeS9rDtkPObeX9vWxMHNCOB29oR4if/mxFpCKFEWm6zpyE7f+B5I/g8KZzx/2aQ/exzmAS0c28+tycYRgsT83hL0l72JmZB0Cgtwf3X9+O+69vR7Cvp8kVikhDoTAiApCzyzlasm0RFJx3temIbpBwD3T7Ffg3N68+N2YYBkt3ZDN72R52ZeUDEOTjwUM3xHDfgGgCfRRKRJo6hRGR89nLIG25M5js/hrsJc7jVk/oOAISxkFsItj0BVpdDofB19uzmL1sD3tzCgAI8fPk4YExTOgXjb93ta/HKSKNhMKIyKWcPgEpnziDSWbyueP+LaH7GGcwCe9sWnnuyu4wWJySyexle/jpqHMfmmb+Xjw6KIZ7fxaNr5d26xVpahRGRKoie4dzbcm2RVB49NzxyAS49h7oejv4NTOtPHdkdxh8sfUwf122l/3HTwMQFuDNpMHtGdf3Gnw8FUpEmgqFEZHqsJfC3iTnaMmeJeAocx63eUHHW5yjJe2Hgk1TDlVVZnfw6ZbD/G35Xg6dPANAy0BvJg+JZWzvKIUSkSZAYUSkpgqPQcrHzmCSlXLueEAE9BjrDCYtOppXn5sptTv4ZNMh5qzYx+FTzlASGezD5CGxjOkVhZeH1eQKRaSuKIyI1IbMbc5pnJR/w+nj54637ukMJV1/6dz5Va6ouMzOv388xNwV+8jKKwKgdYgvU4bGMjqhtdaUiDRCCiMitamsBPZ+Uz6NsxQMu/O4zRvif+7cuyRmCFj1hXolRaV2Fm3MYO7KfeTkOzens1ogpkUAXVoFld+C6dIqSBupibg5hRGRulKQ45zG2fIh5Ow4dzywlfOCfQl3Q1icefW5iaJSOx/+cJB3v/2JzNyiStu0DvElPjLoXEhpHUyrYB8suu6QiFtQGBGpa4YBmVudoyUpHzt3fj2rTR/ntXG63Oa8urBckmEYHM0vZseRPHYcyS3/bx4HT5yutH2InyedIyuOoMS0CMBmVUARaWgURkTqU1mx8yycLR86L9x3dhrHwxfiRzpHS9oNAqsWa1ZVXlEpqeXB5GxQ2ZdTQJnj4v9l+Xha6RQRROfzpnk6RQTqjB0RkymMiJglPwu2/ds5YnJ017njQW0g4S7ocRc0b29efW6sqNTO3uwCdmaeG0FJzczjdIn9orY2q4X2Lfzp0irYNZLSWetQROqVwoiI2QwDjmx2jpZs/wSKcs/dd00/59k4XUaDd6BpJTYGdofB/uOF7DxvBGXnkTyOF5ZU2r51iK8rmJyd5onUOhSROqEwItKQlBbB7q+coyVpK8BwOI97+kH8L5zrS9per2mcWmIYBtl5xa5gsuNIHjsyc8k4cabS9qF+nq5gcnaqp12Y1qGIXC2FEZGGKu+Ic/v5LR/C8b3njgdfc24ap1k78+prxHLPlJaHk1x2Zuax80gee3MKsFeyDsXX00anyEDXGpTOkUF01DoUkWpRGBFp6AwDDv0IyR/A9k+hOO/cfW2vdy567TwKvAPMq7EJKCq1syc7v8IUT2pmPmdKK1+HElu+H0rns6MokcEE++lqzyKVURgRcSelZ2DXYtjyAfy0Cij/WHr6O9eVJIyDtv1B6xrqhd1hkH6ssOI0z5FcTp4urbR9m1DfCqcad24VRESQ1qGIKIyIuKvcQ7B1oXN9yYmfzh0PjYYedzunckKuMa28psowDLLyithx+LyFspl5rosAXqiZv9dFC2Wjm/trHYo0KQojIu7OMCDjB2co2f4ZlOSfu6/dQOdoSfwvwMvPvBqF3NOl7MisOIKSdrSw0nUofl42OkUEusJJl1bBxIUHaB2KNFoKIyKNSUkhpH7pDCbpq88d9wp0TuNcew9E9dU0TgNRVGpnd1Z+hV1ld2XlUVTquKith9VCbMuACiMonVsFEeSjdSji/hRGRBqrUwfPTeOc3H/ueLMY56LXHndBcBvTypPKOdehFFTYUXbHkTxOXWIdSlQzX7pElo+gtHYGlZaB3lqHIm5FYUSksTMMOLAOkj+CHZ9BaWH5HRaIGVw+jfNz8PQ1s0q5DMMwyMwtqhBOdh7J4/CpytehhAV4lV84MNh18cDo5v5YtQ5FGiiFEZGmpLgAUr9wBpP935477h0EXX8JCfdAm16axnETJwtLSM2sOIKSdrSASpah4Odlq3hl4/J1KN4eWoci5lMYEWmqTqSXT+N8BLkHzx1vHlc+jXMnBLUyrz6pkTMldnZlOQPKzvKgsiszj+KyytehxIUHnnd1Y+c6lECtQ5F6pjAi0tQ5HHDgO2co2flfKD3tPG6xQvuhzmDS8Vbw9DG3TqmxMruDny7aDyWP3DOVr0Np29zPGUzOm+ppGaT3X+qOwoiInFOcDzs+dwaTg+vOHfcJhq53OK+N0+o6TeM0AoZhcPjUGdf6E+d/czmSW1Rp+7AA7wrX5OnSKpi2zfy0DkVqhcKIiFTueBps/Rck/wvyDp073qKTc7Sk+50QGG5efVInThSWVLguz44jefx0iXUoAd4exEcGuq7J07lVEB3CA/Hy0IUcpXoURkTk8hwO554lyR85F7+Wlf/L2WKD2MTyaZybwcPb3DqlzpwuKWNX+X4oO137oeRTUsk6FE+bhbiWgeetQQkmPjJQ61DkshRGRKTqinKdpwcnf+Tc9fUs31Do9ivnacKRPTSN0wSU2R2kHS10ncVzdj1KXlFZpe2jm/sRHxlEeJAPzf29aB7gTfMArwo/B3p7aH+UJkphRERq5theZyjZuhDyj5w73rJL+TTOWAhoYV59Uu8Mw+DQyTMVRlB2ZuaReYl1KBfysllp5u/lDCkB3oSV/9zM3xlWwgK8aO7vTTN/L8ICvPH10mnJjYXCiIhcHYcdflpZPo3zJdiLncetHhB3k3O0JO4m8PAyt06pPw47FOc5R9KKcsk9eYzDWVlkn8hjv6U1exxtOHrazrGCEo4XFnOioITCEnu1X8bPy1Y+uuJdPsJSPspy9mdXiPEm1M9La1kaMIUREak9Z07B9v84g8nhH88d92sO3cY4z8aJ6GZaeVJF9rLyMHHKFSjO3fIqOVZ+OxtAivMu//w2bwjvAq0SnNN6kQmcCenA8WI4XlDCicISjhUUc7ywhOOu/zqDy/EC588l9ovXq1xJkI8HYa7pIW+aBXiVj754l4/AOINLc38vQvy8dOXkeqQwIiJ14+hu53Vxti6CgqxzxyO6OUdLuv0K/MPMq68xKys5b2TiVPUDRUlB7dTh4es8LdwnGHyCnIuec3ZWHlasnhDe2RVOiExwBpZK9rcxDIOC4jJXQDlWHmCOFxSXj7aUcKI8uDjvK670bKDLsVqgmb8zoJw/wtLsvFGXsPIA0zzAmyAfrXe5GgojIlK37GXOaZwtH8Dur8Be4jxu9YQOw8uncW4Em862cCktqjjSUN1AUVb5NWuqzdP/vDBxXqi46Fj5zfuCdpWdYeVwwMl0yEyGzK1wpPy/RacubmuxQcv48nDSwzmSEt4VvPyq1Q2HwyD3TKkruBwvDyjHLhhtOV7oHIW51EUJL/tHZbO4Qsv5IywVpo3O+9nPy6Par9GYKYyISP05faJ8GudDOLLl3HH/Fs4Frwl3O/817M4MA0rPXDx1UWmouESgOLvu5mp5BV46OFwxVATVX0A0DDh1oGI4yUyG08cvbmuxQljHClM8RHQD74BaK6fU7uBkYUmF6aGzIyxnR1uOFxaXj8aUUFBc+RlEl+PraauwUPfsCEtYwHmLdv3PjcY09vUuCiMiYo7snc5Qsm0RFB49dzwyoXwa5w7wa1b/dRkGlBRWvhaiqoHCUf1/WV/M4gwE3pcKE1cKFEFgdeOzTQwD8g5XDCdHkqEwp5LGFgiLO2+KpwdEdnf+OdSDolK7c2qooIRjrpGWyta6FHOssKTS/VmuJMjHo/JRlvPWvJwNLqFuuN5FYUREzGUvhX3LIfkD2L3k3Be5zcu5mVrCOGg/DGxVHNZ2OJxrHi61uLLSUHHBfUb1z+y4iMVaMRi4gkJI1UKFVyBYG/e/hmskL/NcODk7knL+qeXnaxZTcYonors5Afc8hmFQWGJ3rW85cd4i3WMF50Zbzi7gPVFYgr2aC14sFmjm5+WaMjp3mvQFe7s0oPUuCiMi0nAUHoeUj50jJlnbzh0PCIfuYyCo9SXWSZw6FyiK88Co/r88L2L1qHy04ZKh4oJA4RWgzd/qS0HOeVM8yc6fczMqbxvS9lw4iewBkdeCf/N6LLZ6HA6DvKLS8rUu5aMtZwPM+aMu5cdO1nC9y6UW6oZduA6mjta7KIyISMOUleI8RXjbosrXDlyJzevS0xcVjoVUHig8/RQm3Fnh8XPB5Ox/T+6vvG1Qm4prUFolQEDL+qq0VpXZHZw4XVJxhOXsCIxrAe+5EZj8Gqx3+fDBvgyIrd0z4RRGRKRhKyuBfUmw87/gKKv6KIWHj8KEVHTmJGRuqzjFcyKt8raBkRWneCJ7OI81sr9TRaV2ThSet7fLeWcVXbj25VhBMcVlDpZMu4FOEbX7HaswIiIiTVdRnnNK8PwzeY7tASr5yvNvecEUTwIEt2l0AeVSDMPgdIkdbw8rHrbaXc+kMCIiInK+4gLI3l7xTJ6juypfi+TXvDyYnHcmT2h0kwkotUVhRERE5EpKTkP2jvIpnmQ4shWOpjqnDi/kE3IuoLRKcIaU0HY6O+oyFEZERERqorQIcnZUnOLJ2Xlul+HzeQc5Ty0+f4qneXv33gumFimMiIiI1JayEueIyflTPFnbK99V19PfuTnb+VM8YR2qvqdOI1LV7++m9ycjIiJSXR5e56ZozrKXOi8cef5pxlkpUFoIB9c7b67H+0JE14pn8rTopGs3ldPIiIiISG1x2OHY3oqnGWdtq/yKyTZv5zWbzj+Tp2Xnyi9E6KY0TSMiItIQOBzOfU8ytzovJJm51bkvSnHuxW2tnhDe+bwpngRnYPH0qe+qa4XCiIiISEPlcMCp/RW3uj+S7LwEwoUsNmgZX3GKJ7wrePnVY8E1ozAiIiLiTgwDTh2sGE4ykyu/bILFCmEdK07xRHQH74D6rfkKFEZERETcnWFA3uGKpxlnJkNBdiWNLdA89tweKJE9nGf1+ATXa8nnUxgRERFprPKzKoaTI8mQf6Tyts1iKk7xRHQHv2b1UqbCiIiISFNSkFMxnGRug9yDlbcNaXvB9XiuBf/mtV6SwoiIiEhTV3gcsi6Y4jm5v/K2dy2EjjfX6str0zMREZGmzr85tB/qvJ115qRz1OT8zdqO73PucWIShREREZGmxDcUYgY5b2cV5YF3oGklKYyIiIg0dT7mLoHQdY9FRETEVAojIiIiYiqFERERETFVjcLI3LlziY6OxsfHh759+7Jhw4bLtv/444/p1KkTPj4+dOvWja+++qpGxYqIiEjjU+0wsmjRIqZPn87MmTPZvHkzPXr0YPjw4eTk5FTaft26ddx111088MADbNmyhdGjRzN69Gi2b99+1cWLiIiI+6v2pmd9+/ald+/ezJkzBwCHw0FUVBRTp07l6aefvqj92LFjKSws5Msvv3Qd+9nPfkZCQgJvvvlmlV5Tm56JiIi4n6p+f1drZKSkpIRNmzaRmJh47gmsVhITE1m/fn2lj1m/fn2F9gDDhw+/ZHuA4uJi8vLyKtxERESkcapWGDl27Bh2u53w8PAKx8PDw8nKyqr0MVlZWdVqDzBr1iyCg4Ndt6ioqOqUKSIiIm6kQZ5NM2PGDHJzc123jIwMs0sSERGROlKtHVjDwsKw2WxkZ2dXOJ6dnU1ERESlj4mIiKhWewBvb2+8vb2rU5qIiIi4qWqNjHh5edGzZ0+WL1/uOuZwOFi+fDn9+vWr9DH9+vWr0B4gKSnpku1FRESkaan2tWmmT5/OhAkT6NWrF3369GH27NkUFhYyceJEAMaPH0/r1q2ZNWsWAI8//jiDBg3i1Vdf5dZbb2XhwoX8+OOPvP3227XbExEREXFL1Q4jY8eO5ejRozz33HNkZWWRkJDAkiVLXItUDx48iNV6bsClf//+fPTRR/z+97/n//2//0dcXByff/45Xbt2rb1eiIiIiNuq9j4jZsjNzSUkJISMjAztMyIiIuIm8vLyiIqK4tSpUwQHB1+yXbVHRsyQn58PoFN8RURE3FB+fv5lw4hbjIw4HA6OHDlCYGAgFoul1p73bGJrzCMujb2P6p/7a+x9VP/cX2PvY132zzAM8vPzadWqVYUlHBdyi5ERq9VKmzZt6uz5g4KCGuVfsPM19j6qf+6vsfdR/XN/jb2PddW/y42InNUgNz0TERGRpkNhREREREzVpMOIt7c3M2fObNS7vTb2Pqp/7q+x91H9c3+NvY8NoX9usYBVREREGq8mPTIiIiIi5lMYEREREVMpjIiIiIipFEZERETEVI0+jMydO5fo6Gh8fHzo27cvGzZsuGz7jz/+mE6dOuHj40O3bt346quv6qnSmqtOHxcsWIDFYqlw8/Hxqcdqq2fNmjWMHDmSVq1aYbFY+Pzzz6/4mFWrVnHdddfh7e1NbGwsCxYsqPM6a6q6/Vu1atVF75/FYiErK6t+Cq6mWbNm0bt3bwIDA2nZsiWjR49m9+7dV3ycu3wOa9I/d/sMzps3j+7du7s2xOrXrx9ff/31ZR/jLu8fVL9/7vb+Xeill17CYrEwbdq0y7ar7/ewUYeRRYsWMX36dGbOnMnmzZvp0aMHw4cPJycnp9L269at46677uKBBx5gy5YtjB49mtGjR7N9+/Z6rrzqqttHcO6yl5mZ6bodOHCgHiuunsLCQnr06MHcuXOr1D49PZ1bb72VIUOGkJyczLRp03jwwQdZunRpHVdaM9Xt31m7d++u8B62bNmyjiq8OqtXr2by5Ml8//33JCUlUVpayk033URhYeElH+NOn8Oa9A/c6zPYpk0bXnrpJTZt2sSPP/7I0KFDGTVqFDt27Ki0vTu9f1D9/oF7vX/n27hxI2+99Rbdu3e/bDtT3kOjEevTp48xefJk1+92u91o1aqVMWvWrErbjxkzxrj11lsrHOvbt6/xyCOP1GmdV6O6fZw/f74RHBxcT9XVLsD47LPPLtvmySefNLp06VLh2NixY43hw4fXYWW1oyr9W7lypQEYJ0+erJeaaltOTo4BGKtXr75kG3f8HJ5Vlf6582fwrNDQUOPdd9+t9D53fv/Oulz/3PX9y8/PN+Li4oykpCRj0KBBxuOPP37Jtma8h412ZKSkpIRNmzaRmJjoOma1WklMTGT9+vWVPmb9+vUV2gMMHz78ku3NVpM+AhQUFNC2bVuioqKu+C8Ad+Nu72FNJSQkEBkZyY033sjatWvNLqfKcnNzAWjWrNkl27jze1iV/oH7fgbtdjsLFy6ksLCQfv36VdrGnd+/qvQP3PP9mzx5MrfeeutF701lzHgPG20YOXbsGHa7nfDw8ArHw8PDLzm/npWVVa32ZqtJHzt27Mh7773Hf//7Xz744AMcDgf9+/fn0KFD9VFynbvUe5iXl8eZM2dMqqr2REZG8uabb/Kf//yH//znP0RFRTF48GA2b95sdmlX5HA4mDZtGgMGDKBr166XbOdun8Ozqto/d/wMpqSkEBAQgLe3N48++iifffYZnTt3rrStO75/1emfO75/CxcuZPPmzcyaNatK7c14D93iqr1Se/r161ch8ffv35/4+HjeeustXnzxRRMrk6ro2LEjHTt2dP3ev39/0tLSeO2113j//fdNrOzKJk+ezPbt2/nuu+/MLqVOVLV/7vgZ7NixI8nJyeTm5vLJJ58wYcIEVq9efckvbHdTnf652/uXkZHB448/TlJSUoNeaNtow0hYWBg2m43s7OwKx7Ozs4mIiKj0MREREdVqb7aa9PFCnp6eXHvttezbt68uSqx3l3oPg4KC8PX1NamqutWnT58G/wU/ZcoUvvzyS9asWUObNm0u29bdPodQvf5dyB0+g15eXsTGxgLQs2dPNm7cyF//+lfeeuuti9q64/tXnf5dqKG/f5s2bSInJ4frrrvOdcxut7NmzRrmzJlDcXExNputwmPMeA8b7TSNl5cXPXv2ZPny5a5jDoeD5cuXX3IusF+/fhXaAyQlJV127tBMNenjhex2OykpKURGRtZVmfXK3d7D2pCcnNxg3z/DMJgyZQqfffYZK1asoF27dld8jDu9hzXp34Xc8TPocDgoLi6u9D53ev8u5XL9u1BDf/+GDRtGSkoKycnJrluvXr0YN24cycnJFwURMOk9rLOlsQ3AwoULDW9vb2PBggXGzp07jYcfftgICQkxsrKyDMMwjHvvvdd4+umnXe3Xrl1reHh4GK+88oqRmppqzJw50/D09DRSUlLM6sIVVbePL7zwgrF06VIjLS3N2LRpk3HnnXcaPj4+xo4dO8zqwmXl5+cbW7ZsMbZs2WIAxl/+8hdjy5YtxoEDBwzDMIynn37auPfee13tf/rpJ8PPz8/43e9+Z6Smphpz5841bDabsWTJErO6cFnV7d9rr71mfP7558bevXuNlJQU4/HHHzesVquxbNkys7pwWZMmTTKCg4ONVatWGZmZma7b6dOnXW3c+XNYk/6522fw6aefNlavXm2kp6cb27ZtM55++mnDYrEY33zzjWEY7v3+GUb1++du719lLjybpiG8h406jBiGYbz++uvGNddcY3h5eRl9+vQxvv/+e9d9gwYNMiZMmFCh/b///W+jQ4cOhpeXl9GlSxdj8eLF9Vxx9VWnj9OmTXO1DQ8PN2655RZj8+bNJlRdNWdPZb3wdrZPEyZMMAYNGnTRYxISEgwvLy8jJibGmD9/fr3XXVXV7d/LL79stG/f3vDx8TGaNWtmDB482FixYoU5xVdBZX0DKrwn7vw5rEn/3O0zeP/99xtt27Y1vLy8jBYtWhjDhg1zfVEbhnu/f4ZR/f652/tXmQvDSEN4Dy2GYRh1N+4iIiIicnmNds2IiIiIuAeFERERETGVwoiIiIiYSmFERERETKUwIiIiIqZSGBERERFTKYyIiIiIqRRGRMQtWCwWPv/8c7PLEJE6oDAiIld03333YbFYLrqNGDHC7NJEpBFotFftFZHaNWLECObPn1/hmLe3t0nViEhjopEREakSb29vIiIiKtxCQ0MB5xTKvHnzuPnmm/H19SUmJoZPPvmkwuNTUlIYOnQovr6+NG/enIcffpiCgoIKbd577z26dOmCt7c3kZGRTJkypcL9x44d47bbbsPPz4+4uDi++OIL130nT55k3LhxtGjRAl9fX+Li4i4KTyLSMCmMiEitePbZZ7n99tvZunUr48aN48477yQ1NRWAwsJChg8fTmhoKBs3buTjjz9m2bJlFcLGvHnzmDx5Mg8//DApKSl88cUXxMbGVniNF154gTFjxrBt2zZuueUWxo0bx4kTJ1yvv3PnTr7++mtSU1OZN28eYWFh9fcHICI1V6eX4RORRmHChAmGzWYz/P39K9z++Mc/GobhvHrto48+WuExffv2NSZNmmQYhmG8/fbbRmhoqFFQUOC6f/HixYbVajWysrIMwzCMVq1aGc8888wlawCM3//+967fCwoKDMD4+uuvDcMwjJEjRxoTJ06snQ6LSL3SmhERqZIhQ4Ywb968CseaNWvm+rlfv34V7uvXrx/JyckApKam0qNHD/z9/V33DxgwAIfDwe7du7FYLBw5coRhw4Zdtobu3bu7fvb39ycoKIicnBwAJk2axO23387mzZu56aabGD16NP37969RX0WkfimMiEiV+Pv7XzRtUlt8fX2r1M7T07PC7xaLBYfDAcDNN9/MgQMH+Oqrr0hKSmLYsGFMnjyZV155pdbrFZHapTUjIlIrvv/++4t+j4+PByA+Pp6tW7dSWFjoun/t2rVYrVY6duxIYGAg0dHRLF++/KpqaNGiBRMmTOCDDz5g9uzZvP3221f1fCJSPzQyIiJVUlxcTFZWVoVjHh4erkWiH3/8Mb169eL666/nww8/ZMOGDfz9738HYNy4ccycOZMJEybw/PPPc/ToUaZOncq9995LeHg4AM8//zyPPvooLVu25OabbyY/P5+1a9cyderUKtX33HPP0bNnT7p06UJxcTFffvmlKwyJSMOmMCIiVbJkyRIiIyMrHOvYsSO7du0CnGe6LFy4kMcee4zIyEj+9a9/0blzZwD8/PxYunQpjz/+OL1798bPz4/bb7+dv/zlL67nmjBhAkVFRbz22ms88cQThIWFcccdd1S5Pi8vL2bMmMH+/fvx9fXlhhtuYOHChbXQcxGpaxbDMAyzixAR92axWPjss88YPXq02aWIiBvSmhERERExlcKIiIiImEprRkTkqmm2V0SuhkZGRERExFQKIyIiImIqhRERERExlcKIiIiImEphREREREylMCIiIiKmUhgRERERUymMiIiIiKkURkRERMRU/x9Ax1VaZdiVsAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 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_loss_curves(customHistory)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:37:09.184385Z",
     "iopub.status.busy": "2024-11-19T07:37:09.183753Z",
     "iopub.status.idle": "2024-11-19T07:37:11.557464Z",
     "shell.execute_reply": "2024-11-19T07:37:11.556703Z",
     "shell.execute_reply.started": "2024-11-19T07:37:09.184344Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m32/32\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 68ms/step - accuracy: 0.9962 - loss: 0.0132\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.012852621264755726, 0.9960317611694336]"
      ]
     },
     "execution_count": 215,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customModel.evaluate(test_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction on Custom DataSet\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": [
    "test_data2 = train_datagen2.flow_from_directory(\n",
    "    test_dir2, \n",
    "    target_size = (224, 224),\n",
    "    batch_size = 32,\n",
    "    class_mode = 'categorical'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:42:22.063910Z",
     "iopub.status.busy": "2024-11-19T07:42:22.063542Z",
     "iopub.status.idle": "2024-11-19T07:42:22.069149Z",
     "shell.execute_reply": "2024-11-19T07:42:22.068081Z",
     "shell.execute_reply.started": "2024-11-19T07:42:22.063879Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "def preprocess_image(image_path):\n",
    "    img = tf.io.read_file(image_path)\n",
    "    img = tf.image.decode_image(img, channels=3)  # Ensure RGB format\n",
    "    img = tf.image.resize(img, (224, 224))       # Match input size\n",
    "    img = img / 255.0                            # Normalize to [0, 1]\n",
    "    return tf.expand_dims(img, axis=0)           # Add batch dimension\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:43:10.754277Z",
     "iopub.status.busy": "2024-11-19T07:43:10.753335Z",
     "iopub.status.idle": "2024-11-19T07:43:10.764215Z",
     "shell.execute_reply": "2024-11-19T07:43:10.763287Z",
     "shell.execute_reply.started": "2024-11-19T07:43:10.754227Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "\n",
    "img_loc = \"/kaggle/input/standard-ocr-dataset/data/testing_data/Z/28813.png\"\n",
    "img = preprocess_image(img_loc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:43:13.548450Z",
     "iopub.status.busy": "2024-11-19T07:43:13.548107Z",
     "iopub.status.idle": "2024-11-19T07:43:13.608170Z",
     "shell.execute_reply": "2024-11-19T07:43:13.607614Z",
     "shell.execute_reply.started": "2024-11-19T07:43:13.548419Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step\n"
     ]
    }
   ],
   "source": [
    "pred = customModel.predict(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-11-19T07:43:19.979361Z",
     "iopub.status.busy": "2024-11-19T07:43:19.978504Z",
     "iopub.status.idle": "2024-11-19T07:43:19.984113Z",
     "shell.execute_reply": "2024-11-19T07:43:19.983161Z",
     "shell.execute_reply.started": "2024-11-19T07:43:19.979329Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Z\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "np.max(pred)\n",
    "idx = np.argmax(pred)\n",
    "print(class_names[idx])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kaggle": {
   "accelerator": "nvidiaTeslaT4",
   "dataSources": [
    {
     "databundleVersionId": 5927243,
     "datasetId": 3364134,
     "sourceId": 5850209,
     "sourceType": "datasetVersion"
    },
    {
     "databundleVersionId": 2729632,
     "datasetId": 1634134,
     "sourceId": 2685160,
     "sourceType": "datasetVersion"
    }
   ],
   "dockerImageVersionId": 30787,
   "isGpuEnabled": true,
   "isInternetEnabled": true,
   "language": "python",
   "sourceType": "notebook"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}