{ "cells": [ { "attachments": {}, "cell_type": "markdown", "metadata": {}, "source": [ "# Notebook 5: Creating a Tool for Census Transcription Workflows\n", "\n", "So far we have developed two separate image processing tools for inspecting digitized images of census forms. These tools work together to look at the demographics columns in these census forms and focus effort on the target demographics that we care about. So far this amounts to a \"W\" detector, helping us to disregard any census forms that only record white families and focus transcription efforts on other pages. In order for this \"W\" detector to become useful in a workflow, we need to provide additional code that will handle some exceptional cases and provide code that will record the results of the \"W\" detector as it is used.\n", "\n", "## Making Partial Predictions More Useful\n", "\n", "There are several steps we can take that will make our predictions more useful in filtering data for manual transcription.\n", "\n", "1. Filter out all of the known 'W' characters. There is no need to distract people with images of demographic cells that we are sure are 'W' characters.\n", "\n", "1. Filter out all the cells where 'W' was a significant probability. Even though the model may predict another character is more likely than a 'W', we often see that a 'W' is in second place, in terms of probability. A quick review of these situations indicates that in most cases, with respect to our census demographic data, these second place 'W' predictions are correct. So we can probably filter out the cells where 'W' is the second most predicted character.\n", "\n", "1. Our task is really to collect page references for further analysis by a human being, who will look at the whole page. If we can help a person focus on the few cells that may or may not contain a targetted demographic code (\"Jap\" is the demographic code for Japanese, in our case), then a person can either say yes or no to saving that page reference for further analysis. In other words, even though we may have up to 30 cell images to review, we only need one answer for the entire page, which is \"save\" or \"ignore\".\n", "\n", "1. Lastly, we could use both (or more) models in tandem, the plain neural network and convolutional one. If any model predicts a \"W\" character, we can accept that conclusion. This seems pretty safe, as not many other demographic codes resemble a W and it should increase our chances or detecting all of the \"W\"s on a page. Since employing multiple models at the same time will involve a deeper refactoring of our code, we save that step for the next notebook, where we will focus on operationalizing this tool within a transcription workflow.\n", "\n", "Let's see if we can implement something like this in our *run()* function. First we will filter out all the known 'W' cells and cells where 'W' is a significant second place prediction, say greater than 15%.\n", "\n", "## Handling Errors and Exceptional Cases\n", "\n", "There are many images which may break the computer vision based segmentation tool or yeild strange results in the machine learning model. The machine learning model will likely always produce some error rate, so a little strangeness there is something we will have to work with. However, we cannot have these tools exiting with errors each time they come upon an image that doesn't match with the expected census pages. For instance the census population schedule folders also contain images of \"cover sheets\", which are much smaller pages that catalog a whole series of population schedule forms. We will need to add code that detects such pages and skips them.\n", "\n", "We will probably also find some images that cannot be segmented with our computer vision approach, simply because they are too faded, folded, or otherwise made less legible. In those cases we probably need to let a user decide if the page is relevant for further study or not. We will need to know when the segmentation code has failed to produce a reasonable output and then take this other approach with the page image.\n", "\n", "## User Interface and Transcription Workflow\n", "\n", "This tool will be a sort of grand filter on census records for a particular city or region, allowing researchers to quickly eliminate a large number of pages from their manual transcription effort and analysis. When necessarily, it will need to prompt the user with images and ask whether a page should be saved for transcription or discarded. At the end of each run the tool will need to produce a list of saved and discarded pages, along with perhaps a code saying the reason for exclusion, i.e. was it \"human review of entire page\" or \"tool decided\" or \"decided after partial human review\" (of the non-white demographic codes). That way if someone want to revisit the decisions made, they can target their efforts. Since we are generally working in a data science environment in our workflows, we will output this information to a CSV file." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The first thing we will do is load some of our old code and trained models for use in this notebook." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "2025-01-16 12:29:31.278794: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.\n", "2025-01-16 12:29:31.279024: I external/local_tsl/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.\n", "2025-01-16 12:29:31.280914: I external/local_tsl/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.\n", "2025-01-16 12:29:31.306750: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n", "To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n", "2025-01-16 12:29:31.781372: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\n", "2025-01-16 12:29:32.302778: E external/local_xla/xla/stream_executor/cuda/cuda_driver.cc:282] failed call to cuInit: CUDA_ERROR_COMPAT_NOT_SUPPORTED_ON_DEVICE: forward compatibility was attempted on non supported HW\n", "2025-01-16 12:29:32.302798: I external/local_xla/xla/stream_executor/cuda/cuda_diagnostics.cc:134] retrieving CUDA diagnostic information for host: graeber\n", "2025-01-16 12:29:32.302801: I external/local_xla/xla/stream_executor/cuda/cuda_diagnostics.cc:141] hostname: graeber\n", "2025-01-16 12:29:32.302851: I external/local_xla/xla/stream_executor/cuda/cuda_diagnostics.cc:165] libcuda reported version is: 565.77.0\n", "2025-01-16 12:29:32.302862: I external/local_xla/xla/stream_executor/cuda/cuda_diagnostics.cc:169] kernel reported version is: 560.35.3\n", "2025-01-16 12:29:32.302864: E external/local_xla/xla/stream_executor/cuda/cuda_diagnostics.cc:251] kernel version 560.35.3 does not match DSO version 565.77.0 -- cannot find working devices in this configuration\n" ] } ], "source": [ "import tensorflow as tf\n", "import keras\n", "\n", "# Load the model we trained in the previous notebook\n", "conv_model = keras.saving.load_model(\"conv_model_balanced.keras\")\n", "\n", "# Load the segmentation code that we developed in that notebook\n", "with open(\"segmentation.py\") as f:\n", " code = f.read()\n", "exec(code)\n", "\n", "# Reusing this function to convert image pixel values to floating point between 0 and 1.\n", "def normalize_img(image, label):\n", " \"\"\"Normalizes images: `uint8` -> `float32`.\"\"\"\n", " return tf.cast(image, tf.float32) / 255., label" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now let's simplify our run function, by \"factoring out\" a function that yeilds demographic cell images. This function will encapsulate all of the special logic regarding the census template and how to use the `extract()` function. A special keyword, `yield`, is used to return a series of results." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "def get_demographic_cell_squares(image):\n", " (adjusted_img, v_lines, h_lines) = extract(image, f, debug=False)\n", " grayimage = cv2.cvtColor(adjusted_img,cv2.COLOR_BGR2GRAY)\n", " grayimage = 255 - grayimage\n", " demo_h_offset = v_lines[11] # the demographic column starts the the 12th vertical line\n", " demo_width = v_lines[12] - demo_h_offset # width calculation\n", " # Open CV rectangles calculation for each demographic cell\n", " demographic_cells = \\\n", " [ (demo_h_offset, h_lines[i], demo_width, int(h_lines[i+1]-h_lines[i])) for i in range(3, len(h_lines)-1)]\n", " for i in range(3, len(h_lines)-1):\n", " cell_img = grayimage[h_lines[i]+5:h_lines[i+1]+5, v_lines[11]:v_lines[12]]\n", " yield crop_cell(cell_img)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Here then is our `run()` function with all of that segmentation logic moved out to the new yield function. It is a lot simpler to look at." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import glob\n", "import matplotlib.pyplot as plt\n", "import re\n", "import cv2\n", "import numpy as np\n", "\n", "blank = np.zeros((28,28, 1), dtype=np.uint8)\n", "\n", "def run(path, page_range=(2, 16), model=None, debug=False, threshold=0.5):\n", " for f in sorted(glob.glob(f'{path}/*')):\n", " pagestr = re.search(r'-(\\d+).jpeg', f).group(1)\n", " if int(pagestr) not in range(page_range[0], page_range[1]):\n", " continue\n", " fig = plt.figure(figsize=(30, 10))\n", " image = cv2.imread(f)\n", " for i, cell in enumerate(get_demographic_cell_squares(image)):\n", " ax = plt.subplot(3, 10, i+1)\n", " plt.axis('off')\n", " tensor, noop = normalize_img(cell, 0)\n", " result = model.predict(np.expand_dims(tensor, 0), verbose=0)\n", " is_W = True if result >= threshold else False\n", " if is_W:\n", " plt.imshow(blank)\n", " else:\n", " ax.set(facecolor=\"orange\")\n", " plt.imshow(cell)\n", " plt.title(\"not W?\")\n", "\n", "run('pages', page_range=(14,17), model=conv_model, threshold=0.04)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "From this code output above you can probably see that we are getting fewer non-W's detected due to our use of two trained models at once to screen out the W's. The code is simpler to look at, which is good because now we have to add some more features to it. First let's stop looking at all of these blank images that are hiding the known W cells. Instead let's make plots that only show the remaining cells that need human review." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABnEAAABNCAYAAAB5c0FFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAAsTAAALEwEAmpwYAABTE0lEQVR4nO3dd5xkVZn4/8+5t27l6py7q+PEHmYI4sDgIIigCAKSzC7huz9RRNewq6sgIIIRYQ0r6CqsIroqBoJkJQwZBhic3DM9ncN07q6urnDvPb8/erqYZjLTPd09PO8XvF5TdeveOtWnbtWt85znPEprrRFCCCGEEEIIIYQQQgghhBCzijHTDRBCCCGEEEIIIYQQQgghhBC7kiCOEEIIIYQQQgghhBBCCCHELCRBHCGEEEIIIYQQQgghhBBCiFlIgjhCCCGEEEIIIYQQQgghhBCzkARxhBBCCCGEEEIIIYQQQgghZiEJ4gghhBBCCCGEEEIIIYQQQsxCEsQRQgghhBBC7NHjjz9ORUVF5vaSJUt4/PHHAdBac8kll5Cbm8vy5cun/Lk/9alP8c1vfnPKjysO3Le+9S3+9V//daabIYQQQgghxFuOBHGEEEIIIcRh57LLLuPnP//5pPsuvfRSlFJs2bIlc9/HP/5xSktLycrKYsGCBfziF7/IbHvuuec47bTTyMvLo7CwkAsvvJDOzs7M9sHBQS666CKKioooKiri2muvnfR8TU1NvOtd7yIYDLJo0SIeffTRzDatNVdddRXl5eVkZ2dz8skns27dusz2JUuWEA6HM/97PB7OOuuszPZ7772XI444gnA4zAknnMD69ev3+vdIpVJce+21zJ8/n1AoRHV1NZdeeilNTU379ffc2bp16zj55JMBeOqpp3jkkUdoa2vjhRdeOOBj7cutt97K17/+9Te1b2dnJ2effTZlZWUopXZ5rclkkksvvZSsrCxKSkq46aabMttSqRQXXHAB1dXVKKUyQasJN998M7W1tWRlZVFWVsYXvvAFbNvObH/mmWdYvnw5kUiEZcuW8dRTT2W2PfbYYyxdupScnBzy8/M599xzaW9v3+tr+e1vf0tVVRWhUIgPfOAD9Pf3Z7b19/dz7rnnEgqFqKqq4re//e1+7wvwf//3fyxevJhQKERdXR2rVq3abRu+9rWvTTo/hBBCCCGEEIeGBHGEEEIIIcRh54EHHuCMM87I3H7qqafYunXrLo/76le/SlNTE8PDw9xzzz1cddVVrF69GoCBgQE++clP0tTURHNzM5FIhEsuuSSz7xe+8AXi8ThNTU288MIL3HHHHdx+++2Z7R/5yEc4+uij6evr44YbbuCCCy6gp6cHgD/+8Y/cdtttrFq1iv7+flasWMEnPvGJzL7r1q0jFosRi8UYGRkhGo1y4YUXAtDQ0MDHPvYxbr31VgYHBznrrLM4++yzJwUR3uiCCy7gnnvu4be//S1DQ0OsWbOGt73tbfz9739/k3/hcc3NzVRXVxMKhQ543721dyoYhsHpp5/On/70p91uv/baa2loaKC5uZnHHnuM733vezz44IOZ7StXruQ3v/kNJSUlu+x79tln8/LLLzM8PMzatWtZs2YNP/rRj4DxoMpZZ53Ff/zHfzA4OMiXv/xlzjrrLAYGBgCor6/noYceYnBwkI6ODubPn8+nP/3pPb6OdevWcdlll3HHHXfQ3d1NMBjk8ssvz2z/zGc+g9frpbu7mzvvvJNPf/rTmYDgvvZ95JFH+MpXvsLtt9/OyMgITz75JLW1tQfwVxZCCCGEEEJMOy2EEEIIIcRBqKqq0t/61rf04sWLdU5Ojr744ov12NiY7u/v12eeeaYuKCjQOTk5+swzz9Stra2Z/RobG/WJJ56ow+Gwfve7360vv/xy/bGPfSyz/dlnn9UrVqzQ2dnZetmyZfqxxx7LbLv99tt1TU2NDofDurq6Wv/mN7/JbFuzZo1eunRp5nY6ndZHHXWUXrNmjQZ0Q0PDbl/Hxo0bdUlJif7973+/2+2rV6/W4XA4czs/P1+/8MILmds33HCDXrlypdZa602bNmmv16uHh4cz21euXKlvueUWrbXW3/nOd/SFF16Y2bZ27Vrt8/l2+7yPP/64DofDOhaLaa21/vGPf6zPOOOMzHbHcbTf79ePPvrobvd/5JFHtN/v1y0tLbvdrrXWt912m160aJEOh8O6pqZG33rrrZltjz32mC4vL8/crqqq0o888oj+xS9+oX0+nzYMQ4dCIX311VdrrbX++c9/ruvq6nRubq4+66yzdHt7e2ZfQP/kJz/R8+bN09XV1Zlj33jjjbqwsFCXlJTo2267LfP4iy66SF955ZVaa73P99OepNNpDeht27ZNur+0tFQ/9NBDmdtXXXWV/tCHPrTL/uXl5ZPee2/U29ur3/3ud+tPf/rTWmut7733Xl1fXz/pMfPnz9e/+MUvdtk3kUjo//zP/9SLFy/e4/G/+tWv6o985COZ21u2bNGWZenh4WEdi8W0ZVl606ZNme0f//jH9Ve+8pV97qu11itWrNhtu3bnmmuuyZyf27Zt04D+2c9+pktLS3VJSYn+/ve/n3ms4zj629/+tq6trdV5eXn6wgsv1H19fZntv/rVr3RlZaXOy8vT1113XeY9JYQQQgghhNiVZOIIIYQQQoiDduedd/LQQw+xdetWNm/ezPXXX4/rulxyySU0NzfT0tJCIBDgiiuuyOzz0Y9+lOXLl9PX18e1117LHXfckdnW3t7OmWeeyVVXXUV/fz833ngj559/Pj09PYyOjvK5z32OBx54gJGREZ555hmOOuqozL73338/Z555Zub2zTffzDvf+U6WLVu227ZffvnlmSXPSktLJ2Xw7OzJJ59kyZIlk+7TWk/699q1a4HxDIja2loikUhm+5FHHpnJkPjwhz+c+Vul02l+9atfcfrpp+/2eX/1q19x/vnnT8p2eePz7vzcb/Too4+yfPlyotHobrcDFBUVcd999zE8PMztt9/OF77wBV5++eU9Ph7g//2//8ett97KihUriMVifOMb3+Af//gHX/3qV/nDH/5AZ2cnVVVVfPjDH56031//+leef/75zBJwXV1dDA0N0d7ezi9/+Us+85nPZLJWdrav99OBGBgYoLOzkyOPPDJz3879sz9++9vfkpWVRUFBAWvWrOGyyy7LbNu5fyZu79w/LS0t5OTkEAgEuPHGG/nyl7+8x+dZt27dpHbW1dXh9XrZvHkzmzdvxuPxsGDBgt2+jr3t6zgOL730Ej09PcybN4+KigquuOIKxsbG9vtv8Nhjj9HQ0MDDDz/Md7/73cySgT/+8Y/561//yhNPPEFHRwe5ubl85jOfAWD9+vVcfvnl3HnnnXR2dmb6XgghhBBCCLF7EsQRQgghhBAH7YorriAajZKXl8eVV17J7373O/Lz8zn//PMJBoNEIhGuvPJKnnjiCWB8EPvFF1/kuuuuw+v1snLlSs4+++zM8X7zm99wxhlncMYZZ2AYBqeddhrHHnss999/PzC+VNbatWsZGxujtLR0UnDlb3/7WyYQ09rays9+9jOuu+66Pbb9pz/9KSMjI6xatYrzzjsPn8+3y2Nee+01rrvuOr7//e9n7jv99NP5zne+w8jICFu2bOG2224jHo8DEIvFyM7OnnSM7OxsRkZGACgtLWXlypUsXLiQQCDAH//4R26++eZdnjcej3PXXXdx8cUXZ+479dRTeeKJJ3j88cdJpVJ861vfIpVKZZ77jfr6+igtLd3j6wc488wzqaurQynFSSedxHve85491kbZmzvvvJNLL72UY445Bp/Px7e//W2effbZSfVovvrVr5KXl0cgEADAsiyuvvpqLMvijDPOIBwOs2nTpl2Ovbf304GKxWIAk/po5/7ZHx/96EcZHh5m8+bNfOpTn6K4uBiAFStW0NHRwe9+97tMgG7r1q2T+qeyspLBwUF6e3u5/vrrWbRo0V7buqf3UiwWIysra7fb9rVvd3c36XSau+66i1WrVvHqq6/yyiuvcP311+/33+Caa64hFAqxdOlSLrnkEn73u98B47WMbrjhBioqKvD5fFx77bXcdddd2LbNXXfdxVlnncXKlSvxer1cd911KKX2+zmFEEIIIYR4q5EgjhBCCCGEOGg7Z3lUVVXR0dFBPB7nsssuo6qqiqysLN75zncyODiI4zh0dHSQl5dHMBjc7TGam5v54x//SE5OTub/p556is7OTkKhEL///e+59dZbKS0t5cwzz2Tjxo0ADA4OsnHjRk444QQAPv/5z3P11VfvMpD9RqZpsnLlStra2rjlllsmbduyZQvve9/7+OEPf8iJJ56Yuf9HP/oRgUCA+fPnc8455/CRj3yEiooKAMLhMMPDw5OOMzw8nMnMue6663jxxRdpbW0lkUhwzTXXcMopp+wSiPnzn/9MXl4eJ510Uua+RYsW8atf/YorrriC0tJSent7qa+vn/TcE/+3tLSQn59PZ2fnXl//Aw88wPHHH09eXh45OTncf//99Pb27nWf3eno6KCqqipzOxwOk5+fPynT4o0ZQfn5+Xg8nsztYDCYCbLsbG/vp1WrVmVe8xuzpXYnHA4DTOqjnfvnQMyfP58lS5Zkas3k5+dz9913c9NNN1FcXMyDDz7IqaeemumfneXl5XHRRRdxzjnnYNv2bl/H3t5L+3qf7W37RBDts5/9LKWlpRQUFPDFL34xEyjdH7s772H8/D333HMz5+7ixYsxTZPu7m46Ojom7RcMBsnPz9/v5xRCCCGEEOKtxrO3jacZFx6qdsyIR9w/znQTpp304dx3MH1oLpxH5/c8OFpR8R9JnIbGKWzZ1JA+nPukD+e+w70Ppf8OjdbW1sy/W1paKCsr4wc/+AGbNm3i+eefp6SkhFdffZWjjz4arTWlpaX09/cTj8czgZydjxGNRvnEJz7B//zP/0zqw79/bTV//9rE7RyO0aew9v61rLh/Jceqd9GlWzHxc7o1voTX4/pe7vvz3/j/Lvpk5hj185ewkKMoUZW7vI71+iXu5+/c+/nxLJQxPcpqnqCahfz6onv49UX37LLPUk6E7XDn+rsYI81pxoWM6hE2sYl3qQ/gURYAL+nHKSHKaf99Ia/qp8ijmEsqv8Aj7h+5+OKL+fznP8/69es59thjM8f+1a9+xb/8y7/skqlwwQUXcMEFFwDjgatf/vKXvP3tbwfYJQBy6qmn8sMf/pC2trbdBhKSySTnn38+v/71rznnnHOwLIsPfOADuywJtj/Kyspobm7O3B4dHaWvr4+vrLyBgBpfDu7iBZ8jqMaDKP16O730T+rjbt3Dl0+7jnz1M9bpF/ER4PlvbaJRr6ef7SzleBYoPyN6kOd5lPd4PoihDFbwvvEDbJh83rvaBeDjNZdn2gDgxc/pxReQr8YzaLbqdcQZ2eUzo1f386V3XUOeKtrj677o1+ewdevWzO2TTjqJF198EQDbtqmtreVLX/rSbve1bZvt27czPDzMiSeeuEv/LVmyhDVr1mRuNzY2kkwmWbBgAYZhYNs2DQ0NzJ8/H4A1a9ZkAkB72zcSiVBRUTHpvbW3jJhff+MPjDHKab+9kDE9CsBp4fMIqfFMoAb9GmlSnGZciNImC3kbOapg0jEujn6eRr2eUUZ45cZtADjaoZvuTJ/PlNnyWTqd5Ptw7pM+nPukD+c26b+5T/pw7nur9qFk4ghxmDMMF1Md+CCMEEIIcSD++7//m7a2Nvr7+7nhhhv40Ic+xMjICIFAgJycHPr7+/nGN76ReXxVVRXHHnss1157LalUimeffZZ77703s/3jH/849957Lw899BBaaxzt0K+3k9BxkjrBdt2Bo20MDEw8wPjgcx9d5PP60mEn8F6O47TM/wBH8Q4KKSelE3TpVmxto7WmT3fRRSt5jA/WJ/QYL/MkUeqoUHW7vOa4jpHSSbTW9OpO2tlGDePLYoVUhDA5NLIeRzts1+3EGKKI8SBKFnl000ZSJ3BdlzvuuIN0Os28efMyx29ra+Oxxx7joosu2uW5V69ejeM49PT08MlPfpKzzz57j0tynXrqqZx22mmce+65rF69Gtu2GRkZ4dZbb+W2224jlUqRTCYpLCzE4/HwwAMP8PDDD+9Xv7/RRz7yEW6//XZeffVVkskkX/va1zjuuOMmBU/eLAcbExMPFmmdopH1+95HO2jGgzguLo52MttKqWIbG0jrFKN6mHa2UUp1Zrurnczj9Y59JwJb7XobKZ0AIKaH+fa3v8273/3uzL6vvPIK6XSa4eFh/v3f/51oNMp73/teYDy7atOmTbiuS09PD1/84hc5+uijycvL2+1r+NjHPsa9997LqlWrGB0d5eqrr+a8884jEokQCoU477zzuPrqqxkdHeXpp5/m7rvv5hOf+MQ+9wW45JJL+PGPf8z27dsZGBjg5ptv5v3vf/9+9QfANjbgaJuYHqKDZop3vL/LqWULazPBnpROsl2PZ+kUUU4vnQzqXlzt0sj+1yESQgghhBDirWivmThCiDnOUIS8aVytQNYaF0IIMY0++tGP8p73vIeOjg7OOeccrrrqKgYHB/noRz9KQUEBZWVlfOlLX+Kvf/1rZp8777yTiy++mPz8fJYvX86HPvQhHGd80DwajXL33Xfz5S9/mWd4FoUiizwWcwygaWEz63gBhSJMNos4ZjwQQxfzWZp5Dq/yT26oBgsfpjJxtE0bW9nIy2g0foIs5EgKVRkAHWxjjFEaWU+jfj1g8C51LgAjDLCJNdikCRLmCJYTVq8v27aU41jHizzB3fgJspTj8arxejtVLCRFkud5lJycHObNm8ef/vQncnJyMvvfcccdrFixgrq6XQNI//Zv/8aaNWuwLIsLL7yQm266aa/9c9ddd2WCa52dnRQUFHDaaadx9dVXE4lE+NGPfsQHP/hBkskkZ5111qT6RAfi1FNP5Zvf/Cbnn38+AwMDnHDCCfzf//0fl1R+4U0db2eVzOefPM+T3IOPAJUsoIeOve7zGH/J/PtZHhpvI+MZTHXUs4FXeIr7MTGpYiEFqiTz+Gd4iATjy9u9wlMAvIP3ESDEIL1sZS22tvHi44ozLueb3/xmZt/vfe97mWXJTj/9dP7yl9fb0d7ezpe+9CW2b99OJBLh5JNPnrT9jZYsWcKtt97Kxz72Mfr6+jj11FO5/fbbM9t/+tOfcumll1JUVER+fj633HLLpEycve379a9/nd7eXhYsWIDf7+eDH/wgV155JTCeUVdfX8/69euprNw1aw0gh0Ke5kFAU8UC8nf8/SqZv+PvtoqkTuDFRzEVFFFGWGWzUB/FP3keF4co8/Diw5D5hUIIIYQQQuyW0ntZJ+Gtmp50OJE+nPveTB8qjwezpJjEghI8V3YzZluEP2fibNoyDS08ONKHc5/04dx3uPeh9N/0q66u5he/+AWnnnrqQR3nQx/6EIsWLZqUsQP734dDup9NvMJy9e59P3gWmQ19ON3kPJz7alU9CcZYoo5lTI/yNA9wCudhqIMLvtja5gnu5gROn5KMrTfrrdCHch7OfdKHc5/04dwm/Tf3SR/OfW/VPpTpTkIchsySYhour2TbxZq6rAMvSiyEEEIcCi+++CJbt27FdV0efPBB7r77bj7wgQ8c1DFrqZ+axgkhMrTWjDJMgOCUHK9nx3KIjrZp4DVCZOOfomMLIYQQQghxuJHl1IQ4HFke0vk2BXkxPMoZX05NCCGEmGW6uro477zz6Ovro6KigltuuYWjjz76TR8vW+2+pogQ4uAcc8wxJBhjIW/+/NxZDx2s40U0mixyWcpxKFn6VwghhBBCiN2SII4QhyOlwOvi89iM2j7G0h6ydirkK4QQQkylpqamN7XfWWedxVlnnTW1jRFCTLlXXnll0tIVARXK1BZ6M+rVsdRz7FQ0TQghhBBCiMOeLKcmxGFKGRqP4eKi0FrBnstfCSGEEEIIIYQQQgghhJiFJIgjxGFImwam5RK0UhhI8EYIIYQQQgghhBBCCCHmIgniCHGYMj0OfjONoSSII4QQQgghhBBCCCGEEHORBHGEOBwZBpFgkpLACIZyZ7o1QgghhBBCCCGEEEIIId4Ez0w3QAgxPUzDxaMcwJzppgghhBBCCCGEEEIIIYR4EyQTR4jDlKsVtjZxtJrppgghhBBCCCGEEEIIIYR4E+Z8Jo6yvJjlJQA4bR1o257hFgkxOziuwtEKU2riCCGEEEIIIYQQQswY5fFglpaA1jhd3TJ+KYQ4IHM+E8csL2HDv5ey4UulmOWlM90cIWYHrRlLehlKB/AoB4/pgpKMHCGEEEIIIYQQQohDzSwtoeEzURo+U4lZXDTTzRFCzDFzPhNHey0CpTFc10Bbc/7lCDE1tEZrhasVlnKwTAeQII4QQgghhBBCCDEXTKw8oy0PKIVK2zhtneh0aqabJt4E7bNwowm0C/i8M90cIcQcM/ejHpaHmvx+bNcAKzjTrRFiVjGUJt8aJeJNginnhxBCCCGEEEIIMReY5SVs+I9SvIVxvF6bke486r+tsZtaZrpp4k3QIT/n17+C7RqsDy6Y6eYIIeaYORvEUT4fRm0l8aos+gcNUikPtenYTDdr1jGOXIxK2biNLehkcqabIw4hvaMUjmXYeAwXd2abIw6SsryY0TIAnNaOPc6+yszW8lpor4VK7zj/ZbbWrKF8Pox51QC4W5p2+WxWPh9GXRXa60GbCiNh7/ZxYnbLrHkNOJ1dsub1HKI8HsySYrA8aI+5Y9ar1F0UQojdUT4fRnUUbXlQyRQqlcZp7wRl7Ne1qxBiz7TXIlw+zLKiTlytWOOUoT3mtD+vsrwYVeUAuM3tcv4epInPyeF5WdT4eui3w4dBcQsxnaT++/SbOC+BSWNmmfEYy0TZLiRT+/U5mDmex4RkCpVMT/k4wJwN4hi1lcR/mMY0usn/fhGBrX24ze0z3axZ54jbN/JUdy25/1aBs3nrTDdHHEKuY5ByTFwtVweHAzNaxoZr8kArFn9DY29r3v3jykvY8KVSguUxjippZ31vMSVfqsBpaDzELRZ7Ysyrxn9LP7Y2sT9VibOhYfL2uiqSP0rw7qJNFFtD/K1nKYlPVeGs3zxDLRZvhllawqbPRVEaFvxIY7fJNcpcYZYUs+VTlaQKbaysFHafn8Xf19jNrTPdNCGEmHWM6ig9PzCYn9vNS61R3PYgC/9Lg+VhwzV5aFdRf92er12FEHum/RZvL2nlxJzN/LX7aBJjXpQ7Ou3Pa1SVs+naHNCw6GpXzt+DZFRH6b7R4JTyF6j29rJ2tAKZZSv2ZiILD438DpkmRnWUru+baK0o/ffXx8wmxmOOzW9h03Ax69pLWXCVxm5s2q/jzcvr5eXmSlSbnwU/ntpxgGkP4iiPB7O8FExzPDUgbeN0dR9wJGpiRqv2WeiQn9HKCF6zk1jKS+62AZwt26bpFcxtp2e/xmA6SLtVMdNNEYeS1rhaYevpn6UjptfELKjR+QUcUd3KaNqL9u66fu7ETI1EXSH5NQMcV9zM+3LX8IB1JFt9NTPQcrEnrt/D6YXrSGuT+7wrM/dPZJgOL87lvJJHuCDrNUrNAA4Gf/GePHMNFgdkYgZOvCYXT1UMO+0BrzXTzRIHQPu9pKNJqkr7KQqOsN43nt0oZs6kGXGJ9PhM/2moCZDJoJvIwrIdySA4SHvKSlSWF6O6ArwWrteDkTq4rNM9zaYU00/7PBxXvI2VWZsZTAXYmCwBj4n2e6mr6MEyHEYXlxP0WtI3h7mdP6txNCqVxt3WKn1+MJQix4qTY8axtYF21OtLbkzl0+w83ub3MVYexjTT2CkZT5gK2udhRck2zsxeg6VsAmaKREWEYKxySrMsJsZftddC+7yyKscUOlSrLEyM6yRrCvAVxUmnzfHxdDHltM/D8aVNBIwULy45llDID8DQ/AgfLn2QU0IbeTVYARxLype1y/5vjFOM1GXx9pLXqA308qpRMR0f1dMfxDHLS9n4b+Xo/BR6zIO3x2TerRq7vePAjlNawubPRNGVY5yz8DUaYkUMfzdK7pZ+3G0SkdyTh4eXsra/hNy0pN69lShX46YNxmwLRzJx5jSjqpyNV+eyrKqFayvv4aHYETzuP3aXx03M1Cio7uf7i/7EPGuYHMNDY7CLrWbtDLRc7I1X7fhMVipzn1FbydDNDu8re4JTQhskw36OmpiR/I7SV/hIsIOHe+uJ+/NmulniAOiQnw8ve4kP5KymNZ3P73k7MSt/ppv1lmbUVeH8OM5ReW081jGfntZi6r8z9TUBJn5v2MUpsnLjDPeGqL/eldoDB2Hib6o0zP/v12cjGtUVbPlmmGXlHZyQ28iTffNJ7SY7dX8Z1VF6b1I4rqL4S1FZgeBQUooszxg5ZhxDvT5i4fos3lO8mveG13HH9St4tHXhpJmu4vAzMXt5RcE2WsZyWd0Rpeor5TLh9iBNnFdJx4O2p+cXwkQGOeVjHFvVwsCQTfQH2fgbe3HaOqflOd9SDIMy3yC11jAWcH7OS6z6XB3N20pZ/L2py7Iwy0vZ+PlyPGVxjq5oY1NvkXzuThGztITNn42CZsqzKyY9T3kJ679cQqAojt+ySQz6wZW0rWmx47x8f2QN5dcPMuL48Rtpcj2jfCDcQFCZ5BnNbMst5GnrqF12N0tLaLg8ihNNcMGSVyiwRmhP5vKX1iOputXAu7UFp6t7Sps8bUGcifoNiZp8QrVDVOcO0D8WpN2Xe0AzUneOQho1oywt6+CkrI2ktUlDS45coO/Do20L6O/KJi81tW8cMQdohavVvh8nZjfLQ2nRICvyGqnw2OR6RicN/E/QPi/ZFUOcVLqFedYwxWYAA0WeJ8ZYRZjQ0NTO8hEHL8sYI1YbIewuAqUYnpfFmWWPc07Wq0QMh4n5NiEjyVhpiFBv+ZvKZBWHlvZbvKO0gQtzXyChLXK9NcR3c86K2WdiBvHQwmyWBNup9qQYcUfxm2li0oczSns9nFq8gdNC6xm1fTzjGMQXFePPCqJN8+CzOHbUQUpVFUDlGNH8IUZTFsh11EHTPgunLAkasDyZv/VYVS7HRhv5QOHLnBbopMAzzC/nnUc4UXVAWVYT5+3IwlyOLVrDYCrAoKdwel+UmEwpDKXxqzQ53jEC4SSji4pIR0yqvL1UeTQfyFk9vjqEV1aHOJxMmoUc9DG8IJtzSh7luOBW7ki+Ay2foVPGUjYRK4Hhc0Cp1+ulan1QmanK58OoqmCsOhdVPkZF4QAe5TKWsshrGZBl1KaIVoqgmSSoFJYyKDTHOKV0M/9gAYm6IvxKHdRv9cx1TGUBgaoRFhV24zVsHDkHD9rE51yquhBVOYpjm+C1Xq9F7PPiBn2ZGkdvtp7tREbxWDQbTI2dNhkb9GP1eEAm5U8LrRRBI0WxmeacyGs4KNLawFIuFgoHTcTwkG2O7WEMzsKtTHBktI33Zb1GSpvc3XYk3S155Dd3TUugb9qCOGa0jPVXFjCvupuf1NxLiTlKlxPizrwVtAYr9/845SVs+PdS8qsH+O7C+wmqJP/bvZKXWiqZHx+SZST3ofiLNiXpbpy2A8t8EocPU7kYSsu5Mle5mrRjknCtvfahDng5u3oN/5LzPHmGF4PxL5ljfK2MfnqIlq1lLL7RxW5tOzTtFntl4HJKsI3VV73AYDpIjhWnwIpxbtar5BlgYmApAwPFUb42Ov4lhWqsZN5PDzyTVRxa2mNQH+zgCG+SDjtBjhWnTQIAc4JRVwX/HeP/K32c9wZbMFGMuj4SjjUtS5eI/aeVImIkyDbSeAyHIwo78X2jjYCZYmGwi7/3LiJx2ZvP4jBLitlyWSXpyiSXLV3FxtESmq5ZSMkmuYY+aD4vy6rasbWB6w9jlhTTcHkl1MT5ZunfqbcShA0f7wg0cefnO1jfWEr9DfufZTWRpXVqwVo6ktms6ymhPP3mgnnizQsaKco8I3yk6Dnenx+g58YsIuYY7w12EVReSsw4ed5R+T48zJilJTR8JopTkeDMRWup9vdxhL+VewePpuG6eqo39Urd4ilgKYdCc4QPFL2Cqw1Svqzx8bavFYBWB/SZ+UZGVQUt3/GzrHgrdZ4U/+wrpf0b8yhv7Mdtkt+NU8aAHDNOrhEAwO9x+GLBs5yZ/Sqf+8KHGdhWyuIfvPmMnInrGLs6wdfrH2HDWBnPfX05FZt6pR8P0kRGMZVjfGbpk7w0VE2/vyiTMZNbMcRHap+j2DNEQnu5b/uy3da93ZeJjGLTGKTue9n4mofGM3DSNk739ml6dW9xBmSbcfJMH6abos9R/GN0EWltstDfQb4xyjKvg89I7z6IEw7wiSOe531Za3hg+Ege616A7/u51G/tmrbfD1MexJlYgzFRk8+86m7eV7yOemuUiOElomKU+YZoMfd98TYRSU5WF5BfPcA7ShspNIfpcbJ4paMC3RqEVO9UN/+wI2mTwlIOHuVgy4+mOUtrhcvu+29ixsbwvAh1vm4KTBMXl6S28SkPEcPluOJm/pHwon1S02EmTXw/xkpDhIwUYWVxfs5LJLRFnhknrQ3+MboAgLcHtpFjpKj0WPiVQ0neMK1DPvDIeriz1cS5OFQbpsgzTFB58aqxSUvLiNlJWV6Mmigji3K5uORJzg03E1R+Ym6SETdA3PZKEGeGTNSFG6kOk2PG8SsosGJke8ZYEdpCoTnCQssl34zxk/oPkuXOO6C13yd+b6SrCrFrElQUDrJxtIQ128soaeyXZdSmgDYV+b5RAmaaNYsqMZM5UDvKsvIOSsw4QcOHgSJiKE4qbGAo6d+vGlTK58OYV81QfQ7nF79Evb+da9vPYrgjQkUqNv0vTExiKhe/0pSbQxieQUqCbfiUQVj5cNEYjP8mkbViZ9ZE5hpaj39WvtnsxYnVUqoLoHqUo8s7OC93NZay+evg28aXztvcJ8uoTYVMpptDpdVPXbiHFxZXo1zNgpoOUq5JorYQv9Y47Z37nckxUYszXpODZY4Qt700j+TS3ZVD/eYeycCZBgYupjJwtIsHkwIzRFrHOLa4lSfGfGjrIIZnvRapsjRlhUO8OlrJy71Rwg39Mh44FTwmdp5NNH+I+b4uBkIh2mrmo+wsiqv7OaV0M5dkv0aBGSKp07vUvd2XTD3VeblYZh/xpJfctkHsxqbpe01vcZnfF5Xjvy8ARlxNq53DA9uXEE97KQtVUR4YpLbwKcw3TKfeeQxunr+bkLJ5fPt8mrcVUr+1e1p/P0x5EGeiBk6odoif1NxLvTVKrhHAVAZ5JuTtYSmgXY6z0yytnyz6C0VmjDsHjufxjvlU/MQzLWvLCXG48SiXYmuIfF+cdiU1GeY6R+tdlsibmLFxZsUqTg81Y2GxxXYBl4WWQ47h4YP5zzPmWHT6ymem4QIY/37c9LlygnVDLPV1EjRCLPOmcEnjoLlrpJI7r3o/ZsLl558aZmVZI1cVP44B+Dw2GHq/vj/FzNj5XFzp7wb8M90ksZ+MmihN3wpwctUa3hvaQlgFAUhol5djVTQP5BK1EzPcyrcmo6qczddls7xqMycEWskzfXwydzWO1gQNExNFQHk5JdjGpqtXcV/rERR+cf/roWRmrtYk+N7yP7EhUcbDX38nJetl5uqUMQwqA/2cm/0y67+zGYClvg78ysFS0OMkyTO8+JTBqZG1tBXm0mRV7fuw86oJ3trHBwsfw6/S/L53OcH/yqF+UydOu9RvOJS0UljKIagUBZbCVAYevACYysDVDmkUaW0iSwPMLKOuivSPx7Bdg8AVUZxNW97UcczyEjZ8sZSc6kF+Uv9nyj3DbE3nc2//eAZO6eY++QydQkEjRZ5hU+uJU1+0ise/M/4d93ZfOx1OkCv+7aMMbytn0Q/0fq+6YNRWMnyzg2VuJ+fGQlItDtmOS066R2rgTBOX8QDOmB6faOLDwlKKI0IdbMopBvPND8/qgI8j57XiNW1eueYYwg39uI0yEWVKaA2OwtGKcnOI87NX88qXowBcH72fKk+c7B0ZVh7M8QkLB/CbfeI3pNfsw/tf+WRvlc/P6TZRd/qo6kbe7u9gxFXcH1vCfV1LSX+3hGDzIINmIa2LFrDqO1tx3jADZaIG7mnlz7LE28H9sSPQNxdRv376M/inPIijvRa+yhgrypqo9cTINYKY6vUXbKi9X7lNmtVRE2dZRTvlnmES2mRVVx19TbkUNXVMWxEpIQ4nSmksZe/zvBNzgwPjXyA7zQjXloel+W0cEWjj3lgdcddHw1gRhtLU+XvINkdZ4O0mx4rTIQGAGaUtD5QkqS/sJrgjOyNoeElrhyE3Qa8dIbx1GGMwRmtjOU9Ry2Dh4/gV5PriGEEbDJnCOltpv8WJJZs4J+sVIoZ3ppsj9kOmNkdlDisqNnFm7quElLFj1rhiRCue66lmtCOCSo3MdHPfmnxe6su7OD3/n0SUgYFBvhHAxmHETZEEHJIYwDlZr9BXEqLBW7fPw07KwKlOUFncz4ZEGc/01RLePCAzV6fQRB2AMtNhcbh3x3KvPgbcMX4zvIQhO0i2J07ESPA2fzN51ijbdrNqwxvXnh9cEOGDhY9xnL+Jm7pP49mWamob+6asOLQ4MIZyMVH4lGfSb38xO0xknMYW5PDugrUM237WWwsymTnujmz9fdUXm8iAG63OIlw5zKL87axNRFkLvDxcyer2KDWbeiUDZ4qZysUEAspL2DA4PzSw4/4wloqxuKCbF2IB2Esmx6T6ReEAsaowPk8XoykvedsGpNb0dHNh1PXR7sR5cHQBJi6nhcaDqIZyMY0DG6+ZWOFBWx6wPIxVROgZ0STSHqKbeuU6ZiopBeP/4VMOOYbLGYX/BCDqiZNjeDBQONrFRY9PWNhLBv9EFgheC+0xiEcjeM0+4imLkq190neHwo6608flNuFXiiFX88D2JWxuLqG+4fVMxFBgCSPu6xMzJz5H4zW5nFL+IidHNnD/yDLu71hC1taBQ5LBP/VBnJCfjy54iX/JeYk8w4uLxgQc7RJ308TdvQ9smOUlbPiPUvKrBrhx4X2UewZptXO4Z+Bogv+Vw+JNHTK7Soh90IZCmS4ew8XVBq6WH1OHK+U4rOsvoWGoEONHBQTaYuCMXzRsMuczMi+L0695YoZbKQCwPBTmDbMs0k5opwGOpE7zfDKfV4ajqLSD097Joh9oRpeV8uJNlZwRbOXSkqfwqBMY9Eux5tlKGwbzA90s9hr4lEVaOzPdJLEPE1kYTm2CqwufYaE1TL8Lg+4YZR4fLyYqsX6cz+K1cu05U7Rl8racFk4MNOFTr/+GiLtp/hybT9z1Uu3tJWKMEfUMk2vF92u5pp3Xjv+Ptz3MxrFSHr76nYQ3DcjM1almQKFnJLMyA4x/7903WsOfvv5ewluGQCmGF2QRuv4PFFvDu53B+sa158usQSxlc1P3aTR8o57azVJ7Yyb5VXqPARxX0m9mnFETpeU7fo4q2UzS9bC6r5Jg2kbtqAV3etF6LOXwwPYj4NN7ruVgzKvG89Mhzi94DoBVA/O4+6unEmwZQaUdalIxOQ+ngaFcTKUyY2s7c2F86d59LN+7c12PD9WvZnOsiO3friWvoR93mwS/p5tKO7w0XM260XLWXr8MbSjWXPkyK7M2sz2dRTxt4TuApXvN8lI2fbYcXZKkKH+YvsE00ZsjFLYMSBbHVFMK5XfIC8TxKpc808d5kc2ktGbINUlohyqPg4FBrzPGgB3aaxDHqCpn8zdyqCvpoTrSx7r+Uqz/yqdEMnAOnR11p+Oul5TWrBqrZfQHFdSv695tJuLEajhmaQmbPxvFqh3h3OzVvJaM8sA1J5O1vu+Q/X6Y+po4aYeXBqrI9YwCkG/GeH+ok8COH177GkzWlgdPXoJo1gCvjVWyligbR4t5qb2SapldJcT+U0gthsOBofB5bIJGChMIGUnGysMEx+rQAS+j0QhK9TE05qd8864zNyIsJuFa5FpxRhZmE/Yt2eUplO2iUvZBrY0t9o/PdAia43/jmJvg/ngxHelqnhus5ZW2CupSI2jbxm5tI5CfxbZkIT3+FvLMGHneOAOSiTPrKMuLGS1jpCpEjhnH2GkEWWoAzHJei3Q0xYLS7aS0SaMd5LFYPT4jzSXZrzHq+gi0j8q15wzSSpHtiZNjGDsGsFzi2qbZNrlv+zLGbIuTChuo8PYT9QzjU3uvBTAx+zFRlYdTm6CyqJ+NY6W81FNJ1vo+mYk8TSbqAMD4xL4RN0V3OpvwliHc1zYCEPEsIaEtCj3DjMzLIuIsQiXGvy+138doZYSSmj7eWbKV94f/SRqDH3adytMtNdRulpn/M8oY/81hSsb3rKW9Fm8va+HU3PVsSRRjGi7x2lycgMHFJU9yXrgRQ40vefeQd8Uu+09k7AzV53Bp0RMcH9jGj7afwj87y6jZ0CPn3zRztYGzY1DY0a8HRR3tMuoabOgrQm/3Q3rX70BleTGqKxirycOsjbGguIegkWI45SfYOCjfe4eISqV5ob0S1zWoWdeD9vsYTAdIuBYJ18JxDfZnvUnl82FUVTBWnYsZjVOWN4TjGtgpE39jr9QymgbaUGBoPGryolpDrsldQ2/DVC6fzF1NWFn0uB62pyK77cqJa9D4ggKOqGjh2NxmtqcikoEzQ2zHoCcV4Q/Dy3i0ZxGhrQO7PX8cbZBjxhmeH8EqD+GtG+bo0nYM5bI9nUV469CbXpr0zZjyII67pYnUpyozhZyGF2bjv+E3nBUc3q/9leOSHrN4tSnK4G+q8HXFUI6mOjkqszqEOABqxxeNmNu0ZbIkr40VoQaChskx/hZ+9Klhem0/H6h9jYiZoDmRz9PtNbvf3zAosEY4L7KWxd/uYNT14u50+ZHSHlqS+fy9cwG5/1YhF/KHiIPmkXgxP/rah4k0DKPSDnWpkUnfcyqe5I+NR7O9IsIZOa/NYGvF3pjRMjZcnU99dStRq49uZ4xSM4ilTApNDwXWiNQymqXcSIDzl73MyVkbuLXjZNZ2lFJxm5dUlsn8b3fvUsRSzAADcsw4Wcb4UgYxneSheAl/7nkbI9+qwEy63HZpIYuiXRxX00TETOz1fDOqytl0bQ5Lo238sPwJno3N58lrVpC1oU9mIk+jtPaQ1g6WMklqm5eTOawZrkDZu55jE/WNGkaLeK6pGoDlVc0cHVzPh3NfAOCekSN5uHsx7veLqN3aJ78RZ6GdB5rFzNOWyZGRVk7wN5PWJu8pTvPPrw+TY41xRmgLYSNAXKfGlwHajYlaOheWvMDbA03cHzuCjd9cSs2GHjn/DoG46yWhx7PajDfk4rycjBK5JZuSf7bhdHbtsq9RXUHjDSGWRzdzZdHTPBVbyAPXnExko9RMOZTcba1UfWW8Pq3b3I6qryPPGyffE2PU9jGWskDvu/6iUVVB63d9HF2ymbf5Rni5P4r6fh6Lt/ZKLaNpZJgaj+FiqvFs8L/G5nPf9mUMfzeKHTBY8O1OTvS386eh43i2q4bCVHrXY+yo83hUdBsfLH6R50fqWHPt0ZRskjqMh5zrEhsN8VjTfDb8cimB5sG9fh6+N9iOccP/kdYejvK1kdYGj4zW83DnYsLJvU8gm2pTv5xaMjkp/TbiWcKw4wd2H8TJrAfo8+J6PYyWBFEeB500CWzuPiRryglxOFKMz4pzMLBlObW5SymyPAlyjAQmihzDZmXZ+CyNf8l9jhHX4t+7LmS4O0xFenSX3Y2UzaM9iwkaKY4PNGKieTVZwajrAyCtTdLaRGsFSObWtFKKoJUix4xjohh0QkS2jOCu2bD7x2vNWMKiNxnGwKXQO8KrO7KplKNRybRkT80gZXkxK0oZm1fA/GgXR+W08cDwkeR5Rjkn8hp5hoFPecYzccSsMlEbYKQ2i0pfPyaate1lsC1IYF0z3pLcHdeuYqZMnF8jFSEiRgJTjRcDHnId/tpzDC81V7KwoRdSaayWSho8RcSrPXusAah8PozqKKML8lgabWZFXiPrE+U831dNeGO/TGA4hNI4NKUL6UmEsZyd+mtHvQADODf7ZTpC2ZnlKz5c9DwlniEiyqbDCfJgVz1N24qo3yS/FWerpB4f1PCpKR9uEG+GAREjQY5hkGfGsJRDTcF2cow42YYXA4Wr9S6rpmQycJbkcmHJC5wc3MQDI0u5t30pWZv7JQPnUNCauONlRHtwtMbzhnkKo66PQMcoduvkQeCJ+kUjC3JYUbmeldkN/DMR5ZneWiIb+/e4ZJ6YHjqdmnS+KCBopABY3VtBrCu81/qLyufDqK1kdF4uS4q2sCjUzTP9tTR15bOosU8ycKaTUhiGi99M42jo0Zr7ti9j7bZyFm/YjpMXJu76SANdySxG4j4K3djru+/4HI3Nz+GI8maOym7j+ZE6nuycR+FuVlIR00+l0jjdARwNwXVtu3x+vlHY8HFuqB8AFw+N6TSPdi+mtTWf+uT2Q9HkjBm/qpqYEbe4oov3FG4grU3ubV9Ga0vBTDdNiLlLKUyPg8dw6bfDDKX8e12XU8xe2jAo9Q5RZjpYykuBYfDVosdwgIhh8ni8Gv1fRTvW7+zYZX93SxN8upLfVZ3JDy5JojWU/68Pf9eOgI/W4GjyUmmZhTzNtNfDiflbOTXYSNgI7Pd+PtNmvjXAMXkvcsS32xh0gvTaER7sqidwRfSQpu+K15kVpaz/SjElVX18uOxlXo1Vsu66ZaDhH/++kPcWrucjWeNrzEsmzuwyURvguLJ1BI0kv+9dTvR/PPg3NeN0b4eSXFyMPc5IFtNv4vyqruvmGF8HEMbG4R/xWtq/M5+F67bjtHaAdpl3i2Z0WTn/PKZiz7PIq6P03qR4V9kLvCuygcdGFvPsN5YT3tCP2yTffdPNUjaWGu+bfsfhTx3HsKWxhPpkd+YxynF4daSSJb52jvTGmOcZ4YjovbhAUCn6Xbhz6Fge7lyM9b086rfu/rpHzAwTF8+ODIG0dmhzxmchV71xxFnMGEvZhA0fx/i2Z84rSxn4lA8XzYh2x+sX7/Sb0dhRM+eikqc4PtDIQ7El/O3r7yJrg2RxHCrKdtk2ms/GcDF5RjsRwyGodgTe0Jlg9xsZ86ox/nuYi4uf4O2BJh6KLeGer58qGTizhDYMyn0DtKfz8P04j/p1nXutv2jUVtL/A80xha/RFs/hjk3LKbvFy6KmvvHrITF9DIPsyBgV/kFGtId/jC5i+HtRFq/fPp79lDcfA5e0hg0DJST7A6j04Ou711Ux9sMk7yl+klzPKKsG5tP1rToKpR7VjJmoQQzsNoNx0mMxcHdc49g4dNhJ/jL8NpI/KKV+/aG/Fj3kQZxsM06sLouwqkcrRawqxOKKVo7P24ZfpRmwQ3T0ZWP1ecCW2atCHCwHNT6rSoI4c5bPSOPfMfhhKZMiM4iNQ7eTpCWZT6B1ZI8zUSeyI4Mj5ejGStAQ/Gczdrtc7B1qE7UdIoZJXKeIOXsPrqq0jd0TYLW/gt8F30atbztH+trxKpceJ0BrTh5brd0voyemn/ZaFFf2s7K4EROX3mSI0NZBVCLFusZyXK14f3gdOeYowwuyyE4vxN3SJJlTs4TrKgZSAf7Ws5T1XSXUtvRnPhcV40tNOsh350zRXouc0mGWFzQTMhRD7hh3x6L8ofNYQlsGsBubMo+12zvwF+WScC0iRmK8btxoDcpxQSncsJ+Redm8p/wZVoY38+zoPJ7onEee1MA55NLaod/1srWrEF+nZ1L9BpVI82x7NT7DJq/wCQp31AKMa8Ujo3W0pPJ5oK2entZc6rd2SQbOLGIkbP7Ws5S0NrGUQ0JbtCTzCRopzs1+mYiyJSg+Q3bOagwZKQwUEWO8LyYCATC+TBeMZwYkyiIExmrRQR9DC7L4RPHTnBjYwgOxI7i3fRnhDf0ygehQSqV5rb0MVy9nQ3Y7pd5Bzgs3EDHGa06/8VplIgNnqD6HDxc+x9sDTTwxupAHu5YQWd8nfTfDlMeDWVrCaFmQHDPOoBMk0B7bpf7ixOO034sO+RmpjbA0fy1V/j4ea5pPujmEv2HfGQRiajiuIu56aUwXsGG0lGDTME5rO2Z5KbHS8c/XDEOTrsjH47XQIT/DdRHeUfA8i/0d/KX3mPE6uJuklt9MmqhBfKCG3BR/GD6ae9uPIHdL/4xcix6SIM5E/QWf8vD+8CZGrvMTc/wUeGIYyiWtTZ7sm0/f92sINg+zMDYIqR6cru69H1gIsXtaY9smcdu7S1q8mGMMCBlJAsqbKQqc1g49TpJb+07ggZZ6SsdS+zgIOF3dzP/JjtkG3Yc25VPssKMvAZ5K5PLCUDUqvefJCk5bB4tv1GivxarA0TwcjTB82TDz8noBWNtVSnVi1yX0xKGhA17Ojb7Iu8Lr+VHnabzaWsH8xBBOawf139SMLKvg5e+XcUawG+8Nv+E3XccTv6wKZ/3mmW76W567rZXq/ywn4c0FoDYVm1yPytH0piPjg44SxJkZHpOqnAGODjbjUwYPxUv4+ZXnk7V+DzOIDfAbaY7zN/GTy0ZoGcpFj3rA73Bq/UYWhF7klNAGHo0t4bFr3kHeeqmBc6iltUOLPcZ9w8upvN0gsKF50vWIu62Vyq9UsK5uGZ/7YilH5bfhM2xe7o+SvqmEQOsIBYk0hakuycCZZdwtTSQuq+Ru30njd+zI8k5URLj78qUcX9rMB/NemNlGvkW9MavRVGHC+CY9xt2xnHJQKY4LbuVX/3octh3m3IWvUePr4e2Bbfxl+GgevPqk8exFyeI4pNymNuZdNX7N8pwnn5EF2Xiv/wvnhJpw0OPZUzsx5lXj/ekgnyh8Cr+R5sbO99D8rUWENkvfzQZmaQkNl0dRtaMs9bXzTLxutxn7ZmkJmz8bRUXjfKT+JXzK5sHOeh5dv4j5v3CwWnZfA0lMPZW2GejJYZWq5enOWnq7s6iP92CWl7Lx8+Vk1Q6y1NeJT0Fddi9WrcO2z+ThsUJ8cMHLBI0UrwxHua9xCSU/9VPX3C+1xOYgF80j8Uruu+YUctf2ztjviEOaiWMqRY7h4bzIGgAKTJNux+X6jvextr2M+eu2Yzc2SRlZIQ6SStvobj9bffksz/Xuewcx600EcBztEnOTtNkB/tG5gKHWbMpS+x7M0LYt2TczRFlezGgZw5VhIkYCV2u60jkMJIKYexkg1rY9aVZWaChK69ZyXh31U5I3TCppoVwZYJ4pKu3wbH8tI46fF1srcdsDkOpFp1PY25oJB/38rns56aLVHONv46jsNp72yFKxs8Eb1yXfnbQ2CZpJRhZkE7aWTN7ogpFIoRIpnLZOdHrfgXRxgNI2Db0FPBZYxCJvF1sSJUQahva6hr+jDUKGyzvKttGRm8X2eIQsX4IL818gZCT5x+hiHuhcMj6LXDJwDqm09tDrjPG7oWN5oL2evJbBXa5JdDqF09BI0HZY31hMbyyEabgMbw9Tv05q38xmb6yJOyGQqKE1lk1vMoShXPI8o4zWZRFO1eFua5XPzkNAWx7CxTGOzGsnZIwPFJvKYMgd46F4CWntYamvnYiyiRiKQnOU4yrGa2v8S+6zWMrlrqFjxmcdS/bijHjjNUvYXMKIM74ss4ki2xxjZH4WYWP8WmVoXpgP5L/EIl8nv+h+Jy80V2XG2cQsYHlIF6dZWNiHiSahrcmZVDuy5xI1BXhqYiws2k7QSNGbDtPano+/xYvV2IgtAZxDJ5XG22XRn8pBOQp/rwmJJDoUwChOsKSwi6DSeJWiMtCPR7mErBRew8ZSDv12iFfby7FbQvg3S/bUXJDJmCsOEjKSOFrT446xJVFMeMvQjH4XHpIgjqMNXDQ+ZeHBpMLz+nq5j8fn0/ytRcyfWNtaCHHQnPZOFt2kGVtcwvprSme6OeIgGTuFtsd0ilWJAu7qPZbgf+WweFPHXtfPFTPPjJax4Zo8llU38XZ/BwkNL4zU0NyXS206tu8D7DBxXqdqC2m5PDKNLRb7w93SROpTlay2cqmLj6ASfZMyiN0tTcQvq+LWmgsY+uQIiZRFTXz/+1vMvPeG1xG9oZ+EtjKfwy4GPXaEv7YdSUdzEfXf1jK4PA3cba1UfaWczfOO4PNfLCaetihIpPf4eOVoutPZAPxn8d8BSGtIo+h3/Pxl6G08e+1ywlIL4JBTjqbXjvD7kSN4+Op37jMLymnrpP47Gm2N/0ytSI9K5s0cZyhNuRnjg1mv0P71HP7evICqr5TLUjKHiKH0eH0+xieDuWjuG63glisvwBN3GfhkjHeUb+NrxY+ywPLyrbIHcAC/Uvw5Np8HrjmZ3HUzN+tY7MpSNj7lwVSK88INZN3wBwadIA4GcdfLmuEov244jqKfBJi/TWqmzCba8pCTH2NRdjddTpimRAE4rwdxzIpS1n+tiOrq7dxY8xAvjdZy39WnEG4Yon60BxJJnJ7eGXwFbz1OVzfzfqphxzg2qTROTy9GVjW52aPUhXrwKkW24eVTuc8zohWvJsu4r+9IHrnynQS3DVObiKGSA5I9NUfsnDF3lK+NmHb55cBy7ms9gsK9/B45FKY9iKMcTUsqn232VvxKYwAu459TI9pDw1gxwW3DMjNgDpuIUmJ50B4TlbZlZuoMm1jj0Z8VIuEc8tJXYoqltYe0Hv/x1e/a3N13DM81VzOvsW+X9XPF7KN9FlVlfZxcsImIMhjRLqO2D9s2DyiTZuK8tnwWrpONYTqkorl4HQenvRNt2/s+iJgye5p5PGn7+s0EUzW0juSgDE26NAdvOir9NcupZJrHuhcQNJOcFlpPtpHGVDDqGrycjJJ2PTiuAXsoJiwO3sTM46Djsr6xGIDC1J5/+Kpkmoe7FhM0k1yUtZ4sw8+wm6DLgXuGjuaR1oWUrO/FaWg8VC9B7KCSaR7sqsdxjf3KgtLplARGDzMGGktBtmFyTu7LDKYD9Ppkktmh5GqFozUumgE3QXOygMjmIYyhUVoby3maGuJFCkuZlHrCDDhx/hyr5U8dxxDZKNmLs42zo1yBgUG24ef9oU4GXZuHRufRks5ndXuUtMz6n50Mhd+bJtszhokmbCaJV2cR0gvRHoNYNEJV1XaOyWtlTbyKp3rrMrWMZMWimbG3FU30jt8CjtYYGEQMD3HH5vmROl7tLqdio1x7zkkek3SBTW3+IJZy6XMUf+9aSG9rDkXJmZ1APf1BnNgYd65dzjMlteT64hhKY6CxtUFfIsS2jgIWj/bJB9IcZpaW0PCZKOmiNOHcOLHePOpvkJmps4mLDDTNVROzizudLfiV4t7YYrbcUM+8DT2yluocoS2TeVm9HOUf/0xMaIU7BYO/JbkjNH86gtNZzqKbtPxIm+UKc0do+nQEt6ucRTdLf81mbmMLwc9Gub/mFP73/zuemvx+vIbNlv4C8n4WJtA+Ql4iTb7U5ph2E1kZ4//e8996os/uWvpeSm4Y4qRAK78bXsZD3fUkbyylZEsfbpOcczPBbWwhcEU082/x1hVUXo7xjvBypJ1/eMpnujlvDVqTTHkYtgM4wJCb4M8jC/j79oX4kvZ4lvfNmtFlpbx8UwV1nj5sHO4dreT2q87Zcx0yMWOU1gzZQYbcFAVmABeXftfm98NHcs/XTyXcMESNzPqfvQyD/ECMOl839d4RornPsuWrhcTSPuZFesj1xDki0MYTwwt58OsnjWcQN8mkzVnJdRkcDrN1tJB4DsR1ig0pL38cOIGXrjmWik29cu05R2nTQPkcAp40rXYWT8UWYvxXAfXrO2d8FZxpC+IojwezpJh0URZOwqQ3FiLtmJiGi6E0tmvQPxpExz1SNHaOUpYXo7qCsapcVM0oi4t6WZC1nRcDlWivNdPNE+KwkdQeElqR1prt6SyCzcOyBMUcopI2r/aUEzDfxov+fobsIK91l6G3+yF94NkYKm3j9Pjp0Gr861O+QucEn+lQXjBIi533ejq+mJV0MomzaQvBeAWprVHWDQVQpovR4yW6Vma1Hkr7m5Ux0WdZHpPfdh3HppxSHulcRHtrPvUbpJ7KTJroG/HWo9I2erufNb4ybg8vp9gawsHgsZ6FqJRkox4KKm2T2h7kBW8l/xs4Fks5PNy9mMbWQupTvZks72BWiD90H8ugsx4Hg791LyVrff9eM47FzFDJNI90L8JvpDk1vB5XKx6KHc297UvJWtcrWVOzndakHZOE9uJTBj4TzipYQ1qbLPW1kdAe/jZ0FE931lC8QepQzWYqlcbt8rPainJ7cAXZnjhb4kU821FNxSbJwJnLVNpG9Xpp8BXyG/8JbOgvJm9r/6xYBWfagjhmSTFbPlVJqsgmkD02XU8jZpBRXcGW6yIcW9nIN0v/TokZx6fg19bbWOU/eqabJ8ThQWtito8eJ0CPk0VDrAhlS+7iXOI2tlD4xSgN3joazHmgNZXxOCo59KZmyDntnSy+UcOOegGkbZlpJ8Q0cDq7WPCjnc4125FzbZZztzRhf6qS56xlZCVtspPdki0lxAxx2jtZ9AON9lk8F1wGhgFao1K2ZHccIk5bB4u/r9Fei6cDRwHgS9rUp3px2l6fTexuaSJxWSV3+06SPprlJrIb76l9N7+67Dg0kH9LiKxtA1K3aC5IpdnaXcDj/oWcEdpMqRnknFA7ae2SRvPHkTqe+sbxFO+jfpyYeROZjNpnsTq4BJRC2S4VqaRk4MxxO/dtt6+MvLQ9a87H6VtOzWOSyncI5I1hmjLgOFPMJQshlcZtapv6GjVeiyPKOzi74FXqrQRhYzydN2ImQMnyXbOFq42ZboI4CCqRZlVXHTHHx0AqyGsdZdSmpDj6XKLTqSmdRTUxa1LMDSptw3YfLW4eAEaPF2xnhlsl9oe2bew2WbZyLtlXrSohxKEj1yszT9v2fs0cls/OuWMiuzGUiNLaWI5yIbpesoTnCuVqnKTJcNpPUoOLJqC8WMqhx0nSm44QahyWDJw5QL7jDl+zuW+l4vlhzvjvYTZ3FTL/a2XYjU1TemzX62Fl3lZODbYRNvy4uHQ7SXrtsCyRN4u4KGxXAjlzldvYQsEXorRalaA1tamY1MIRYg6ZmMmTWUJNsjmEEEIIIcRByFxfglxXziVaQ8JkIBGg3QnjUzG8SjHqav6ZKmJDrERW3RBC7NH0BXFsB2+vyZgOggEYmuFAALVjLFm74I558PaZb6omgNg/Hyx5kXvMo4j786bsmMrnw6iOMlwXpswaIGJ4MRiv19HvWPSlwiDfO4cl5fFglpaMLy2j9fgyTl3daFvO4eky1VkcQohDazbP5BFCCCGEEHOPXF/OUWkbX49Jhy+PO/NWUOIbxlIOCddiy2ghr7aXy6obQog9mrYgjtPVzbxbd8w8nVhaS6nJy2y57vggcPf26WrGW95ZoRa67Wz+YRVO2TGN6ijdNxq8L/oMpwTb8KkQADE3zf0jx/BST5S8VHrKnk/MHmZpCZs/G8XOS4Nt4O01mXeLxm6X9eaFEEIIIYQQQgghdsfp6qbuZxq8Fq3BSlrM18dHle3KqhtCiL2atiCOtm0Z2J0Fsgw/EePgatQoy4tZXoL2WuiAl+F5Wbwv+gznZq8m2/BmHudozZATYCxlgU5ORfPFmzSRMTNWGqbA7CXm+l6/v6QYLA/a8qDSNk5b5z7rJU1kX8VrcrFqR6jNHWTMtmgL5ILXOhQvSQghhBBCCCGEEGJOknFSIcTBkJo4bwGGOrj6NGZ5CRu+WEqwIsZ7qzZQG+jhvPAG8kwfPiUD+LPRRMaMisY5PtTDq4MV4/eXFLPlskpSRTahwjijfUHqb9DYTS17PZ5RHaXzex5OqVjNR/OeI89IkUbxv0UreCVSfyhekhBCCCGEEEIIIYQQQrzlSBDnLSDLGGO0Okw4VoPb3L7vrAvLi1FdAV4LbZnEKsLk1w5wXHEzZ+W8SqE5SkJDv5OkwDSwlHmIXonYX9pn4ZYlWFy6nVLvEJ2BbLYuqMBM5uDUjVFVMMiyvHbWZZUyuriYoNfC3da6y3tD+XwYdVWMLMrlPZXPc272ahZb4FNBXDSLAh08XXU84ZGq/croEUIIIYQQQgghhBBCCLH/JIjzFnBCoJX7v7KRZ7fVsuAqjd3YtNfHG9UVNHwjiyXlnRyb20ypNcjxgW04KFbF59ObjrApVkzESvDl4kcoNb34lLyVZhWfl8UVXVxQspoTA428N7yeh741nm1zXHALOUaKiNL0FHi47fqV/L1lAdEvV+A0NE46jFFXRfrHY3y49NlM9pVnx8eGgeL4QDN3fr6D9Y2l+5XRI4QQQgghhBBCCCGEEGL/ycj7Yc7GIagU789fg6sNWpYuIJgdQunJS6xppcAY//dIVZi3V23mPfnrWOTtBOC5sRq609k80TOfoaSf3oEI/kCKV3PKSHi7qfGAqRQF1ggF4VHi84sJ+LyoRBKVTON0dqFte7/arCwvZkUpaI3T3iXZHW+CNhXF/hGiVh9BBaaCE4ObMZSm1gM+5cdAETFSnJ79T7YXRRi0CnetfzQ/i/OKX+IdgS38Y6yKhGvhYJBjjnJaoJOI0pxQ0Ej/WBAttXGEEEIIIYQQQgghhBBiSkkQ5zA34qYwUZwW6OT4inYe+c48hpwgCff1AXdTuWSbcQyl8SqboJFkqbeTIdfH/2w/iRc6K8n/eYhA+whm2iHPccm3B0iV5fDly85ncbSLm2ruosK0+HjWGt4bXsf/3bCcxngBq5sr0e0BFv5QY7e171ebzYpS1n+lGDTUf0djN7dO15/nsKVNk6MirSz1DtNk++hxIvxjqB6fYfOp/GcoMAyChhfeUC5p5/pHZ9aso8I7QJW3h//tfwcvf/MYQk0xAIbnR0hffxenBVs4ItDGmqxyEp7cGXilQgghhBBCCCGEEEIIcfiSIM5h7peDR2Eol4iRGA/O+FvxKwcAVyscFAntYWOyjIQeD+wkbIsHUkfQmcrm6ZYa7JYQobVt2K1tk47tTZVjbatknV3GL7PfweJAB8f5m/Arh/dkrWV7KIKB5jV/GaPLyvAX5gCMZwFNZAIpBezIBAIwIFYaIqtkhGTKg7bkLfqmGOAz0lgoEtpixAnQnw4RMNOk9c4PM/CrNHneOG3zczBsTX7tACtKtvHBnBcAuKP/hPHl1tb1ZpZbi6gl9NgRUlqTb8aoCA6ypraGkF2329o6QgghhBBCCCGEEEIIIQ6cjJAf5h7/+LHj/1CKsfIQySv6WZbfQdQ/AMCQHeDF3ir4YSGBtlgmqILW4GhqEzFUcgCns2uXYztd3dT9TIPX4rXQIp6rXs5tn+3huMIm3hFpoMQzxDcr7iVRbvLiTVWMOAHS2iStTfrtEABBI4WlHPxGGks5BI0kI66fJ/vm88+2cpSrd3lesW/K0QzYIXpczYjrJ6EtHK12eZylTOZZw/xrwZM8dH0/hnJ5f/if+JVLtxPgT4PH8so1xxDd1Ivb9HoQT7kurYk8eoMW9d4Rigue4MfXmDzePI/q/yzH2bLtUL5cIYQQQgghhBBCCCGEOCxJEOcw567ZkPl3aLCSlsZSnk16SZV4CJkpxhyL/niAis19uxS13xdt29jtHa8fP1VHy2iQ3uww7enxpbWWWnGChocaTxsuLgntkNaaQdfARRExHAwgpMYL8iS0S4fj5YXBmjeu9CUOgEqmebCrHkcbVPl6cbSx28cZKPxKUWCmOTuyhhFt8XIySr8dZlO8hKc7aijd1LvLe0Ml0jzZOY+0NqkN9OBXaZZHGukqipDwyrJqQgghhBBCCCGEEEIIMRUkiPMW4rR1sPh7mlR1AU9dVkdlUT+lweEpf56hlJ/ftx5Ljn+MaOU9RD0pcg0/Bh58SuPikmeOP9bAm9mv1xnjd8PLaE/mELN9U96utxK3sYXAFVEeXHQSR1z1Gm8LN+3Xfg+PLOWha08ivHUYHE1pMjUpA2fn4xd+MUqDt44Gcx6jNRHq/nPDbo4ohBBCCCGEEEIIIYQQ4s2SIM5biLZt7OZWvKaJncwnaXtwd7PE1sGytUks6cXVioZUMWndS5lnDAtIA44GU4EB7IjlENfQ4wRoSuTTmwxju7vPHBH7RyeTOJu2ELIWEbO9u31MUqfpd5LcHVtIrx0h7nh5rGs+uWt7cTZv3fvx06lJjwmP1fF8WxWuq6hNxab0tQghhBBCCCGEEEIIIcRblQRxxLQZSfj4YeMpBK00S3M78Bk225MR0q6JZTgYaAJmmrQ2aB3NJel4iKctDKUJWumZbv5hLaltOmzFH4aO47Fr3kGocRhcyE2lcbe1HvDx3G2tVP9n+fi/m9unurlCCCGEEEIIIYQQQgjxliRBHDFtXK0YGfOTTHvoDYbxGTbdYxFcrfAYLobS+M00tmvQMxrGdgyU0piGxu+xZ7r5h72ENulLhQltGcZdu/GgjqXTKZwt26aoZUIIIYQQQgghhBBCCCEAlNZa6scLIYQQQgghhBBCCCGEEELMMlJ4RAghhBBCCCGEEEIIIYQQYhaSII4QQgghhBBCCCGEEEIIIcQsJEEcIYQQQgghhBBCCCGEEEKIWUiCOEIIIYQQQgghhBBCCCGEELOQBHGEEEIIIYQQQgghhBBCCCFmIQniCCGEEEIIIYQQQgghhBBCzEL/P4DESYcs9y5aAAAAAElFTkSuQmCC", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import glob\n", "import matplotlib.pyplot as plt\n", "import re\n", "import cv2\n", "import numpy as np\n", "\n", "lookup = list(\"abcdefghijklmnopqrstuvwxyz\".upper())\n", "blank = np.zeros((28,28, 1), dtype=np.uint8)\n", "plt.ioff()\n", "\n", "def run(path, page_range=(2, 16), model=None, debug=False, threshold=0.5):\n", " review_page_cells = []\n", " review_page_filenames = []\n", " for f in sorted(glob.glob(f'{path}/*')):\n", " pagestr = re.search(r'-(\\d+).jpeg', f).group(1)\n", " if int(pagestr) not in range(page_range[0], page_range[1]):\n", " continue\n", " image = cv2.imread(f)\n", " review_cells = []\n", " for i, cell in enumerate(get_demographic_cell_squares(image)):\n", " tensor, noop = normalize_img(cell, 0)\n", " result = model.predict(np.expand_dims(tensor, 0), verbose=0)\n", " is_W = True if result >= threshold else False\n", " if is_W:\n", " pass\n", " else:\n", " review_cells.append(cell)\n", " if len(review_cells) == 0:\n", " continue\n", " else:\n", " review_page_cells.append(review_cells)\n", " review_page_filenames.append(f)\n", " \n", " if len(review_page_cells) == 8:\n", " for n, page_cells in enumerate(review_page_cells):\n", " fig = plt.figure(figsize=(30, 1))\n", " fig.suptitle(review_page_filenames[n])\n", " for i, cell in enumerate(page_cells):\n", " plt.subplot(1, len(page_cells), i+1)\n", " plt.axis('off')\n", " plt.imshow(cell)\n", " display(fig)\n", " review_page_cells = []\n", " review_page_filenames = []\n", "\n", "run('pages', page_range=(4,17), model=conv_model, threshold=0.04)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Okay, this is a little easier to review. I can quickly scan these images and see that none of these are from pages that I would want to save for transcription. The next question is how do we create this interaction of saving a page for later. I think ideally one would look at a screen full of cell images in need of review. If any of the cells within a page are marked \"Jap\" for Japanese, you would click on the \"save\" button for that page. When done reviewing the whole screen you would \"hit enter\" or some equivalent to get a fresh page of cell images. This is a fairly sophisticated interaction for a notebook, but I think we can manage it. The next code block will introduce the screen by screen display feature." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Rendering a User Interface\n", "\n", "IPython is an interactive Python framework that underlies the JupyterLab notebook environment. You can use [IPython \"widgets\"](https://ipywidgets.readthedocs.io/en/stable/examples/Widget%20Basics.html) to add more complex controls to your cell outputs. We will use controls such as these to add interactivity and paging to our tool. We will need to review several pages of uncertain demographic cells at once, check a box to indicate if any pages need manual review and/or transcription, then advance to the next screen.\n", "When the user advances to the next screen, we want to save any pages that were marked for manual review or transcription to a file. In the code block below you will see some of these controls and interactions illustrated using dummy data. It is much easier to develop the controls this way before we fold them into our real data outputs." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "42920ac9fefc4fdba3b1fe8526d288fd", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Output()" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from IPython.display import display\n", "import ipywidgets as widgets\n", "\n", "boxes = {}\n", "review_file = \"to_be_reviewed.txt\"\n", "\n", "def chkbox(page_file):\n", " global boxes\n", " chkbox1 = widgets.Checkbox(value=False, description=f\"Save {page_file}\")\n", " boxes[chkbox1] = page_file\n", " return chkbox1\n", "\n", "button = widgets.Button(description=\"Next Screen\")\n", "output = widgets.Output()\n", "with output:\n", " display(chkbox('page_1.png'))\n", " display(chkbox('page_2.png'))\n", " display(button)\n", "\n", "display(output)\n", "\n", "def on_button_clicked(b):\n", " global boxes\n", " saved = []\n", " for box in boxes.keys():\n", " if box.value:\n", " saved.append(boxes[box])\n", " else:\n", " pass\n", " print(f\"Next screen clicked with {len(saved)} page files saved.\")\n", "\n", "button.on_click(on_button_clicked)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Above you can see how we used our checkbox controls as the keys in a global dictionary. This allows us to lookup the page file associated with each checkbox and act accordingly. For now we simply print out some information, but we will need to save any applicable page file references into another file. Let's add the save function now and test it." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "3126a058d20a46809e98c31d6297152f", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Output()" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from IPython.display import display\n", "import ipywidgets as widgets\n", "\n", "boxes = {}\n", "review_file = \"to_be_reviewed.txt\"\n", "\n", "def chkbox(page_file):\n", " global boxes\n", " chkbox1 = widgets.Checkbox(value=False, description=\"Save\")\n", " boxes[chkbox1] = page_file\n", " return chkbox1\n", "\n", "button = widgets.Button(description=\"Next Screen\")\n", "output = widgets.Output()\n", "with output:\n", " display(chkbox('page_1.png'))\n", " display(chkbox('page_2.png'))\n", " display(button)\n", "\n", "display(output)\n", "\n", "def on_button_clicked(b):\n", " global boxes\n", " global review_file\n", " saved = []\n", " for box in boxes.keys():\n", " if box.value:\n", " saved.append(boxes[box])\n", " else:\n", " pass\n", " with open(review_file, mode=\"a\") as wf:\n", " for f in saved:\n", " wf.write(f'{f}\\n')\n", " boxes = {}\n", " print(f\"Next screen clicked with {len(saved)} page files saved.\")\n", "\n", "button.on_click(on_button_clicked)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Great! You can verify that any checkbox selections are appended to our save file, `to_be_reviewed.txt`.\n", "\n", "The next step is to be able to clear the screen and render a new set of checkboxes.. We will be called a `clear()` function of the output widget." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "0ef76ac1cf124eed85beadcdeab7ec1b", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Output()" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from IPython.display import display\n", "import ipywidgets as widgets\n", "\n", "boxes = {}\n", "review_file = \"to_be_reviewed.txt\"\n", "output = widgets.Output()\n", "\n", "def chkbox(page_file):\n", " global boxes\n", " chkbox1 = widgets.Checkbox(value=False, description=f'Save {page_file}')\n", " boxes[chkbox1] = page_file\n", " return chkbox1\n", "\n", "def on_button_clicked(b):\n", " global boxes\n", " global review_file\n", " saved = []\n", " for box in boxes.keys():\n", " if box.value:\n", " saved.append(boxes[box])\n", " else:\n", " pass\n", " with open(review_file, mode=\"a\") as wf:\n", " for f in saved:\n", " wf.write(f'{f}\\n')\n", " boxes = {}\n", " render_page()\n", " print(f\"Next screen clicked with {len(saved)} page files saved.\")\n", "\n", "def render_page():\n", " global output\n", " if output is not None:\n", " output.clear_output()\n", " button = widgets.Button(description=\"Next Screen\")\n", " with output:\n", " vbox = widgets.VBox([chkbox('page_1.png'), chkbox('page_2.png'), button])\n", " display(vbox)\n", " button.on_click(on_button_clicked)\n", "\n", "render_page()\n", "display(output)\n", "\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Integrating the User Interface with the Image Processing\n", "\n", "Our next step is a bigger one. We want to use the screen by screen save interface to review processed demographic cell images. This means combining two sets of code, a complex task. We will be glad that we simplified our image processing code a few steps back. In addition to putting our code together, we want to design the program such that image processing can proceed while a person is performing their review task. This will mean that subsequent screens, after the first, should load faster without waiting for as many image operations to complete. We will accomplish this by adding another \"thread\" or \"worker\" to the program. Our kernel will have two threads running, one responding to input and another working away at image tasks. The will cooperate through a special structure called a queue, which is basically a special kind of list. The queue we will use is \"first in, first out\" or FIFO. This means that while new items are inserted at the head of the queue, the earliest items that were inserted are taken from the tail of the queue. Our image thread will add information about pages and lists of demographic cells to the queue, while our user interface code will remove these items and show them to users. Note that because we have two threads accessing the same structure we need to take care that the queue is \"synchronized\", which means that it is protected from failures that result from multi-threaded access and modification. We will use the queue library which provides synchronized queues." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "import glob\n", "import matplotlib.pyplot as plt\n", "import re\n", "import cv2\n", "import numpy as np\n", "from queue import Queue\n", "import threading\n", "import time\n", "\n", "q = Queue(maxsize=5)\n", "sentinel = object()\n", "\n", "blank = np.zeros((28,28, 1), dtype=np.uint8)\n", "plt.ioff()\n", "\n", "def run(path, page_range=(2, 16), model=None, debug=False, threshold=0.5):\n", " global q\n", " global image_processing_done\n", " for f in sorted(glob.glob(f'{path}/*')):\n", " pagestr = re.search(r'-(\\d*).jpeg', f).group(1)\n", " if int(pagestr) not in range(page_range[0], page_range[1]):\n", " continue\n", " image = cv2.imread(f)\n", " review_cells = []\n", " for i, cell in enumerate(get_demographic_cell_squares(image)):\n", " tensor, noop = normalize_img(cell, 0)\n", " result = model.predict(np.expand_dims(tensor, 0), verbose=0)\n", " is_W = True if result >= threshold else False\n", " if is_W:\n", " pass\n", " else:\n", " review_cells.append(cell)\n", " if len(review_cells) == 0:\n", " continue\n", " else:\n", " # Add page to the queue\n", " page = { 'file': f, 'review_cells': review_cells }\n", " q.put(page, block=True) # imaging thread will \"block\" or wait here until space is available in the queue..\n", " q.put(sentinel, block=True)\n" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Size of queue: 5\n", "First item: pages/43290879-California-101393-0004.jpeg\n", "Thread is alive? True\n", "pages/43290879-California-101393-0005.jpeg\n", "pages/43290879-California-101393-0006.jpeg\n", "pages/43290879-California-101393-0007.jpeg\n", "pages/43290879-California-101393-0008.jpeg\n", "pages/43290879-California-101393-0009.jpeg\n", "pages/43290879-California-101393-0010.jpeg\n", "pages/43290879-California-101393-0011.jpeg\n", "pages/43290879-California-101393-0012.jpeg\n", "pages/43290879-California-101393-0013.jpeg\n", "pages/43290879-California-101393-0014.jpeg\n", "pages/43290879-California-101393-0015.jpeg\n", "pages/43290879-California-101393-0016.jpeg\n" ] } ], "source": [ "def start():\n", " run('pages', page_range=(4,17), model=conv_model, threshold=0.04)\n", "\n", "thread2 = threading.Thread(target=start, daemon=True)\n", "thread2.start()\n", "time.sleep(10)\n", "print(f'Size of queue: {q.qsize()}')\n", "file_path = q.get(block=False)['file']\n", "print(f'First item: {file_path}') # We set block to False so that we can an exception, instead of waiting for an item to join the queue.\n", "print(f'Thread is alive? {thread2.is_alive()}')\n", "\n", "for p in iter(q.get, sentinel):\n", " file_path = p['file']\n", " print(file_path)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Okay, so above we have two threads; the one that is running our code in the first place and the one that we started using the Thread object. When I ran this code the queue was completely full, containing five pages, by the time our foreground thread woke up from sleep and accessed it. The speed of your computer may yield different results. Now that this seems to be working, we will go ahead and increase the queue size to 40 pages and rewrite the user interface code to pull items from the queue.\n", "\n", "In this code we are using a special for matplotlib figure integration into the notebook widget framework. It is included via a special notebook **magic** command in the first line. This allows us to embed the demographic cell figures into a more carefully crafted user interface.\n", "\n", "First I will make a few edits to the run method, so that I can gather more information and show the whole demographic cell when a W is not recognized, instead of the 28x28 image." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "# import tensorflow as tf\n", "# import keras\n", "\n", "# # Load the models we trained in the previous notebook\n", "# dense_model = keras.saving.load_model(\"dense_model_2.keras\")\n", "# conv_model = keras.saving.load_model(\"conv_model_2.keras\")\n", "\n", "# # Load the segmentation code that we developed in that notebook\n", "# with open(\"segmentation.py\") as f:\n", "# code = f.read()\n", "# exec(code)\n", "\n", "# # Reusing this function to convert image pixel values to floating point between 0 and 1.\n", "# def normalize_img(image, label):\n", "# \"\"\"Normalizes images: `uint8` -> `float32`.\"\"\"\n", "# return tf.cast(image, tf.float32) / 255., label" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [], "source": [ "def get_demographic_cell_squares(image):\n", " (adjusted_img, v_lines, h_lines) = extract(image, f, debug=False)\n", " grayimage = cv2.cvtColor(adjusted_img,cv2.COLOR_BGR2GRAY)\n", " grayimage = 255 - grayimage\n", " demo_h_offset = v_lines[11] # the demographic column starts the the 12th vertical line\n", " demo_width = v_lines[12] - demo_h_offset # width calculation\n", " # Open CV rectangles calculation for each demographic cell\n", " demographic_cells = \\\n", " [ (demo_h_offset, h_lines[i], demo_width, int(h_lines[i+1]-h_lines[i])) for i in range(3, len(h_lines)-1)]\n", " for i in range(3, len(h_lines)-1):\n", " cell_img = grayimage[h_lines[i]+5:h_lines[i+1]+5, v_lines[11]:v_lines[12]]\n", " yield crop_cell(cell_img), cell_img" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "import glob\n", "import matplotlib.pyplot as plt\n", "import re\n", "import cv2\n", "import numpy as np\n", "from queue import Queue\n", "import threading\n", "import time\n", "\n", "q = Queue(maxsize=40)\n", "sentinel = object()\n", "\n", "blank = np.zeros((28,28, 1), dtype=np.uint8)\n", "\n", "def run(path, page_range=None, model=None, debug=False, threshold=0.5):\n", " global q\n", " global image_processing_done\n", " global w_index\n", " for f in sorted(glob.glob(f'{path}/**/Schedule_Images/*.jpg', recursive=True)):\n", " pagestr = re.search(r'-(\\d*).jpg', f).group(1)\n", " if page_range is not None and int(pagestr) not in range(page_range[0], page_range[1]):\n", " continue\n", " if page_range is None and int(pagestr) == 1:\n", " continue\n", " #print(f'processing page {f}')\n", " image = cv2.imread(f)\n", " review_cells = []\n", " try:\n", " for i, (cell, cell_view) in enumerate(get_demographic_cell_squares(image)):\n", " tensor, noop = normalize_img(cell, 0)\n", " result = model.predict(np.expand_dims(tensor, 0), verbose=0)\n", " is_W = True if result >= threshold else False \n", " if is_W:\n", " pass\n", " else:\n", " review_cells.append(cell_view)\n", " page = { 'file': f, 'review_cells': review_cells }\n", " q.put(page, block=True) # imaging thread will \"block\" or wait here until space is available in the queue..\n", " except Exception as e:\n", " with open(\"image_errors.json\", mode=\"a\") as wf:\n", " data = {}\n", " data['file'] = f\n", " data['error'] = str(e)\n", " wf.write(f'{json.dumps(data)}\\n')\n", " q.put(sentinel, block=True)\n" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "45c58b475fff472385dd23b4039094c4", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Output()" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "%matplotlib widget\n", "\n", "from IPython.display import display\n", "import ipywidgets as widgets\n", "import matplotlib.pyplot as plt\n", "import json\n", "\n", "boxes = {}\n", "screen_start_time = None\n", "review_file = \"to_be_reviewed.txt\"\n", "output = widgets.Output()\n", "\n", "def chkbox(page_file, reviewed_count):\n", " global boxes\n", " chkbox1 = widgets.Checkbox(value=False, description=page_file)\n", " boxes[chkbox1] = {\"file\": page_file, \"reviewed_count\": reviewed_count}\n", " return chkbox1\n", "\n", "def on_button_clicked(b):\n", " global boxes\n", " global review_file\n", " global screen_start_time\n", " etime = time.time() - screen_start_time\n", " with open(review_file, mode=\"a\") as wf:\n", " for box in boxes.keys():\n", " data = boxes[box]\n", " data['screen_time'] = etime\n", " data['save'] = box.value\n", " wf.write(f'{json.dumps(data)}\\n')\n", " boxes = {}\n", " render_page()\n", "\n", "def render_page():\n", " global output\n", " global screen_start_time\n", " if output is not None:\n", " output.clear_output()\n", " plt.close('all')\n", " with output:\n", " page_count = 0\n", " vbox = []\n", " for p in iter(q.get, sentinel):\n", " fig = None\n", " with plt.ioff():\n", " fig = plt.figure(figsize=(16, 1.3), label='')\n", " fig.canvas.header_visible = False\n", " fig.canvas.toolbar_visible = False\n", " fig.suptitle('')\n", " # Note: review cells are populated on 2nd screen..\n", " for i, cell in enumerate(p['review_cells']):\n", " plt.subplot(1, len(p['review_cells']), i+1)\n", " plt.axis('off')\n", " plt.imshow(cell) # TODO fix display\n", " layout = widgets.AppLayout(left_sidebar=chkbox(p['file'], len(p['review_cells'])), center=fig.canvas, pane_widths=['135px', 16, 0],)\n", " # NOTE: Next four lines due to ipympl issue: https://github.com/matplotlib/ipympl/issues/290\n", " fig.canvas._handle_message(fig.canvas, {'type': 'send_image_mode'}, [])\n", " fig.canvas._handle_message(fig.canvas, {'type':'refresh'}, [])\n", " fig.canvas._handle_message(fig.canvas,{'type': 'initialized'},[])\n", " fig.canvas._handle_message(fig.canvas,{'type': 'draw'},[])\n", " vbox.append(widgets.Text(value=f\"{p['file']} {30 - len(p['review_cells'])} Ws removed\", layout=widgets.Layout(width='100%')))\n", " vbox.append(layout)\n", " page_count += 1\n", " if page_count >= 6:\n", " break # showing 6 pages per screen\n", " if page_count > 0:\n", " button = widgets.Button(description=\"Next Screen\")\n", " button.on_click(on_button_clicked)\n", " vbox.append(button)\n", " else:\n", " done = widgets.Text(value=\"Done with images (TODO stats here)\", description=\"Done\")\n", " vbox.append(done)\n", " display(widgets.VBox(vbox))\n", " screen_start_time = time.time()\n", "\n", "def start():\n", " run('NARA_downloads_1950/70-150', page_range=None, model=conv_model, threshold=0.04)\n", "\n", "page_slicer = threading.Thread(target=start, daemon=True)\n", "page_slicer.start()\n", "display(output)\n", "render_page()\n", "\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Things to do next:\n", "\n", "* Nonetype object not subscriptable - this happens when no dark vertical line matches the given kernel\n", "* Isolate the images that produce this issue and test a more forgiving kernel shape.\n", "* \"NARA_downloads_1950/70-12/Schedule_Images/43290879-California-101300-0025.jpg\" gives \"float division by zero\"\n", "\n" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "# This code blocks may help some users to convert json output into a CSV file for further analysis in Pandas, etc..\n", "import json\n", "with open('log.csv', mode='a') as out:\n", " with open('to_be_reviewed.txt', mode=\"r\") as inp:\n", " for line in inp:\n", " data = json.loads(line)\n", " out.write(f'\"{data[\"file\"]}\", \"{data[\"reviewed_count\"]}\", \"{data[\"screen_time\"]}\"\\n')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.12" } }, "nbformat": 4, "nbformat_minor": 4 }