{"cells":[{"metadata":{},"cell_type":"markdown","source":"# Bengali.AI Handwritten Grapheme Classification\n\nBengali is the 5th most spoken language in the world with hundreds of million of speakers. It’s the official language of Bangladesh and the second most spoken language in India. Considering its reach, there’s significant business and educational interest in developing AI that can optically recognize images of the language handwritten. This challenge hopes to improve on approaches to Bengali recognition.\n\n\n\nOptical character recognition is particularly challenging for Bengali. While Bengali has 49 letters (to be more specific 11 vowels and 38 consonants) in its alphabet, there are also 18 potential diacritics, or accents. This means that there are many more graphemes, or the smallest units in a written language. The added complexity results in ~13,000 different grapheme variations (compared to English’s 250 graphemic units).\n"},{"metadata":{},"cell_type":"markdown","source":"> <font color = blue> **Bengali is my mother tounge and I am really interested and having fun to work and improve Bengali Handwritten Grapheme. \n**"},{"metadata":{},"cell_type":"markdown","source":"Basically I am inspired from some other kernal and notebooks and learning so far."},{"metadata":{},"cell_type":"markdown","source":"![maxresdefault.jpg](attachment:maxresdefault.jpg)","attachments":{"maxresdefault.jpg":{"image/jpeg":"/9j/4AAQSkZJRgABAQEAYABgAAD/4SzARXhpZgAATU0AKgAAAAgABgALAAIAAAAmAAAIYgESAAMAAAABAAEAAAExAAIAAAAmAAAIiAEyAAIAAAAUAAAIrodpAAQAAAABAAAIwuocAAcAAAgMAAAAVgAAEUYc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAFdpbmRvd3MgUGhvdG8gRWRpdG9yIDEwLjAuMTAwMTEuMTYzODQAV2luZG93cyBQaG90byBFZGl0b3IgMTAuMC4xMDAxMS4xNjM4NAAyMDIwOjAxOjEwIDAwOjQwOjU3AAAGkAMAAgAAABQAABEckAQAAgAAABQAABEwkpEAAgAAAAMwMAAAkpIAAgAAAAMwMAAAoAEAAwAAAAEAAQAA6hwABwAACAwAAAkQAAAAABzqAAAACAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAMjAyMDowMToxMCAwMDo0MDoyMgAyMDIwOjAxOjEwIDAwOjQwOjIyAAAAAAYBAwADAAAAAQAGAAABGgAFAAAAAQAAEZQBGwAFAAAAAQAAEZwBKAADAAAAAQACAAACAQAEAAAAAQAAEaQCAgAEAAAAAQAAGxQAAAAAAAAAYAAAAAEAAABgAAAAAf/Y/9sAQwAIBgYHBgUIBwcHCQkICgwUDQwLCwwZEhMPFB0aHx4dGhwcICQuJyAiLCMcHCg3KSwwMTQ0NB8nOT04MjwuMzQy/9sAQwEJCQkMCwwYDQ0YMiEcITIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIy/8AAEQgAkQEAAwEhAAIRAQMRAf/EAB8AAAEFAQEBAQEBAAAAAAAAAAABAgMEBQYHCAkKC//EALUQAAIBAwMCBAMFBQQEAAABfQECAwAEEQUSITFBBhNRYQcicRQygZGhCCNCscEVUtHwJDNicoIJChYXGBkaJSYnKCkqNDU2Nzg5OkNERUZHSElKU1RVVldYWVpjZGVmZ2hpanN0dXZ3eHl6g4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2drh4uPk5ebn6Onq8fLz9PX29/j5+v/EAB8BAAMBAQEBAQEBAQEAAAAAAAABAgMEBQYHCAkKC//EALURAAIBAgQEAwQHBQQEAAECdwABAgMRBAUhMQYSQVEHYXETIjKBCBRCkaGxwQkjM1LwFWJy0QoWJDThJfEXGBkaJicoKSo1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoKDhIWGh4iJipKTlJWWl5iZmqKjpKWmp6ipqrKztLW2t7i5usLDxMXGx8jJytLT1NXW19jZ2uLj5OXm5+jp6vLz9PX29/j5+v/aAAwDAQACEQMRAD8A6vw1/wAipo+T/wAuMH/ota1VfP3QT79qYiQByPmb8qkVcdKAJBn1pcn1oAXJp27HU0AKGc9Bx6k04Ke5JoAeBindqAFGazp9e0201H7Bc3QhnwCPMBCnPP3ulXCEpu0QGQ+I9Jubz7JBeo85OFHIDH2YjB/CtPDMPmOPYGicJQ+IBdoFJj61IBz700k+tAEZfnAyT6Cj5z14oATYAc9/WkOfWkA0yPtKZ4PNRkn1oAjZsdMk+gqPEjdTtHt1oAPLx60vlkjrQBEQfU03ByPrQBn+F4l/4RXR2YsxNjB948D92vatqgBRTqBjwKTI7c/SgQ9VYjk4+lPCAdKAJAKXFMB4Wm3Egt7eSZkdggztQZY/QUJXYHD3XxIaG68uHSGMYPWSTDkfTHH60ni+K18RaJaa3p8u542EUy5AZA3QMOxDcfjXpUqDoyjLuIu+BrCyn00yTwrLd28uzc/OBwQQPxP5V2Vc2Lk3VafQEN4pGwOa5hkZbP3Rn+VN8sn75/AUALsCjAFGKAG0wigBhPPAzTCjMfmb8BQAoQDgCgrxSAYVprDahckBVGSScACgCn9us2g85LqF4j0kEg2/n0rPm8SaBBcrBca1ZrMeRGsoz+lVysCbwz/yKejf9eMH/ota1RUjHDA6nFP5zwPxNAhwT+8c/wAqkAAFADhTqYx3fFG4DjqfQUCOb8W+LG8PpHDBEGuZF3DcMhR0H8jXOaZ8StRSXfeWEE8PRvLyjgexJIP04+tehQwanT5m9RXNzxJLp+v6Nb63p0qSCNvLl+X5lz0DDqCDx+NWNK0aw1/QApcwX0K+S08J2sQOV3Dow6cH0rTmlCkr9GBl+D7prHXGtJsr56mNgeMOuSP6j8a7xyAeTXPjVapfuCGrvf7q49c0GMdSST71xjFoIoAaRTTgDk4oAZknoPzpPL9SSaQBtxTWFAEFzcxWdrNczttiiUsxx2FcDN8UY7W8VLnS2W2fkMsmWA+mMH866qGHdVN9gOmv9Rub3wvLqWhPGXaHzYpJF4wOox68Hr3rE8E+K5PEnnWOpGB5duUKDAcdwR0zWkMOnCT6oR5d428MPp+pXsCsY0gYtGo6PGeR+lcFtaNwfusDWtemrqS6jPqbw0wHhTRx1P2CDgf9c1rZCPsD/Kqk465NeYMciAHPU+pqXqaBDwKdjNMBduBmlUnblRnPftSAy9Rtri1uf7VskMsgULcW68ecg9P9oc49eladjdW99aJc2zh4nGQehHsR2PtWrV4qSEUdb8L2HiB4nummR412homAyPfINc9rfhLTNC8PTSRNLJM0iKhkI4yeegHbNdmGxUlanYDntLs9RtNPvLizt5ZbSU+VMiLu6AEEgc8dQe1bHgjW4YtXa03fLdKVAPXevIB/DdXbWUZQko7gP8SaLq1prLalp9q08TTCdBECxVgQcEDnGa7yHbLEkqqyh1DYYYYZ9a8/EzjOEWgJlYID70w4riGNOKaSB1NMDBHivTJNQWzR5mZm2iSOIsmfTI5/IEe9boiA7c+pq6lKULc3UA200g1mAlROwBx39KQFW9tE1CymtJciOVdrEdfwri5vBmq2lu9rZTWlxbysDIZY1L/QBlIA/HNduFrxgnGQ0ZHiO31jQ9MtdGXVFljuhhrOOIAKgPZuvJ7Y55rnJdM1XwndWGq2+I3mUTxq4yOuMH6jH4NXq0+Rx0+0BveLtS/tfR7LV7jTJrS5KBTIPnhkT2cdCDng4PNYaeBrLxZov9paLII72PEc9o3QsP4lPYEc/XNc9VKNJX6CPWvDkDQ+FNFDDDfYYCR6fu1rT714oEifSpBQBV1DVbDSIVlv7lIVY4UNyWPsBya5C8+I32i4Fro9qfMYhVkuF+8fQKD/AF/Cuqhh3V1ewHU6HLqlzZmTWLWOGfd8uwnBH+6c7T+Na4bAx6/pWNSKjNpAhwrIvNHuIrp7/Rp1t7l+ZYX/ANVMfUjsfcU6U0naWzAqt4ovrI7NS0WeFgPvxNvRvocY/WsTUdUvvE9xFbR2rpEr5SMc5PqTj/63P4120aEYy9pfQR2ukab/AGZpsdvvy2Sz46Fj/nH4VUvfCul3l9HfeW8FyjiTzIGC5YHIJGCM5rm9vJVHJdQNrNMLYrABvJo+nNIYmwk8nj0FQahaSXel3VtBIIpJY2VX9CRVRaUk2Bxug2V94dupptT0a5nYKFils1WQD143ZHbt61pS+JNYvMR6XoVyjBiC9zGenGODgA9epr0KkI1p8zegGZq+s69oUf2i61S08x8bLN0DOfXhBwB67qfp3i/WtZEb2OixvGjBZh5h59SrHAH0OaHhaco8ydkDOvKu4G47fUClCKo4GK81rUBpPFNzSA4vxT4Km1zUTfwai0blVQwyDC4How5HryDzVe+8AG/s1ibW7lpkHysVLxqf90t7etenDGxjBK2qGafhfQNQ0PTptOv9QivbVjujCoV25+8CDng9fzrTstM07TZnksbG3tmkwHMMYTdz3xXJVrOUny7MRN4cljl8I6VI5LSPZQkk8Z+QVZ254yR71zgOWRQ2Blj7VxmrePblbprXRNPF04YIHZj8zZx8oA5/OuihQdV+QF3RPGFnqDHT/EGnJYzs2wiX50ZvRtw+U59c/WqfjPw7b6bPb6lp9ulujHawhXaFcchh6f8A1q7KUZUaqg3oxbnW6JqY1jSYbrgSH5ZVHZx1/wAfoa0OlcFaPLNxGhdwHU/hTl3N0AA96zAkC455J96fmndgGfekz+NIBRx97AoIJOAPxNAhpiGcnmnbfSgYnPSgnGc9KAKmoXhsdOmu1iM3lru2KeSO+PwqCG7i17RxJZTyxxzpxLGQGQ9xyDyOlaxi0uYDzTxT4X/se8jdr2Wcyjdvbhs578nP1967O38V6ZpWhaeJlL3EkCs0VvGoO49WIGAMnJ/HpXpVYutSjyARW/j3SppljuYbmz3nCyTINn/fQNdLkEAgggjIIrz61CVJ+8BExAHJxUZYkfKPxPFYAM2FvvHPt2penFADWNQ/xD60hmRoZupPBWlLZbUnFhblGkHykhF4/Hpn3qLTPE0Vxcix1iJ9Pv8AOPKk4jf/AHW7/j+tbQp88XbdCZ1SIoG0AY6DFcJJ4H1ex1VLvRru1KxyCRBOWBGDkA4ByPyrbC1o021LqJmR430vXDcwXmptp4kmUqFtQccAdQw5PPqaZpK67beGJgsTX+nFtjxsWfy2GCCO47dOK9OLpyppisdH8OtZt1uZ7OT5WmyVQ/3l/wDrfyruHDMx52j0FeXjY2qt9xockagcCpa5RhRQAuQOD19KhvLyLT7Vrm4JSJcZIUsefYU4pydkAabqNnqkHnWcvmKODlSCD7g1anmht4WlnlSOMdWdgB+ZpyhJS5XuBmp4k0aSbyhfxhv9oFR+ZGK1ARjIIIPQinOnKHxIBjsB+PaoWXzY2R0GxgVYNzkVC0AwItOu/DYH9nQte6YMk2mcyw+pjJ+8P9n8qzbS5GjXkuo6QDdaHO268tUH720fu4Trj1Ht7V2wtNev5iNfWdGs/F9haz218FCZMUqDepBxkEZHoO/FY1h4AlgvlN3dxSWyH+AMHceh7L+BNa0sV7KHI1qMpfERrZnsLKEIPs6PuRQMIG27R7dDXQ+FxOPDGnpLncIsAn+7k7f0xU1/93i3uBqeUN2TyfU0rAV54EZ60xuaQELsq8sai3OzDauBnq3+FAyr4X/5FPRv+vCD/wBFrV2+02z1SAwXlsk6dtw5H0PUfhVwm4O6EYw0/XtCJOkXC6jYjpaXbYkQeiv/AI9Kjk8b3VoxS68PXsUnYFvlP44rrjSjW1i7MDAu313xfqCP9glSFOEUqVRB7sepr0HQ9LTR9LjtA2587pCOhY9fw7fhWuJlGnSVKLESXOkafdXcV3LaIbiJgySqSrAg5HI6/jVyvPlUlK1+gxwNPB+WpAAwJwoz9KUK5PzHHsKQEioFGAKivLGDUbOS0uAfLcdjggg5BHuCBVRk4yugOdstB1rQWl/sq5sp45fvJcqy49CNuef51mazYXiuuoeIb4JEr4RIQXyfRRgBeO5P1r0IVacp89veYHMarrum3ZisLO2S0jjcE3EnzyMcY+YjgDnoK9E8MaHqWlWix3OprPamP93Cqf6vvw/XHPTH0rTFNRpWnq2I3hGqjAH50cV5IwGKzdR0Sy1CUXBEkF2Bhbm2fZIPxHX8c1pCbg7oDkrrwhrWn3b3eh6wqTucshXyw3uQAVJ+oFPhb4grG0M4tmJ6SgR7h9MED9K71Uw9TWWjAj07wVdS3P2jWJwQzbnjDbmf6t2/Wu0CLGioihUUYVQMAD0rDFV4ztGOyAa3SozXGBGzgcdT6Co23sMfd/nQBH5YU5xk+p60gB3D60DKPhf/AJFTRf8Arwg/9FrWvIVQZ5z6AZNAhFMjdBsHvyakUYOSST6k07tDJQeaX6UCJYHwu2QAZ64OQKSULGevFIBoLH7q/iakWPIy5yfSgCUDFLQA4HimyuyQu6oXZVJCAgFj6c0IDI0zxNaajIIGjltLvOPIuRtJ/wB0/wAX4VuICTlmP0HArScHTeojyW/SKbxVeJhFRrpgc8ADdyf0rpb7xre3l4bbQ7aMxhtqyupZn57Dt+v4V61Sj7VR5tkhkX/CWa1pGpra67aQNGQCzQ5DqD39Dj0GOldsrJJGsiMGVgCCOhFefiKCppSjswGFvTmm4J+8fy4rlAXAHQAU0mmAwmmE0ARM4IwvzH2qMqx6nj2pAIFVVwoAHtSUANI5qMnaw+tAGZ4ZVm8J6NlsD7DBwP8ArmtbCgDpQA8HHWlB5wKAJORS7gO+T6CgBfnYjoo/WnPE0NnLcKu5kGQGJOaaV3YDL0DXTqs9zbTwrDNb/e2nIbkjj8v1rUu9QsrCMNd3UUOegdsE/QdTWtSi4z5FqBlnxpoCNh7t1XON7QOB+eK2La8tb23W4tbiOaF/uvGwYH8qJ0ZwV5ICcbiOBj3NKEHclvrWIEN7p9pqNuYLyBJo+wbqp9QeoPuKzPs2t6Qv+hTDU7YdILhtsyj0V+jfiPxraE01yyAwbu18K6hqMkmoSXml3kuWkiuD5QJPfLAqfwNa2l2nhzw7D9rGpxzcfLNNMjEDH8IUD9BmuuVWty+ztoBxurai3ibxC0lpE5V8RQgryQO5/MmvS7Ky+yWFvbMxfyYljzn0AH9KMX7tOEHuBPwOKQ+lecAw0xmCjJOAO5pgQly33Bn3PFJ5ZP323e3QUgDAHSmmkAw1GxC8k4oAiZ2b7ox7moigLgsSxz3oAqeGSP8AhE9G/wCvCD/0WtaoOKYD0VpHCr1NND84UFj7dPzoAkCsR8zY9l/xqaOPCkhcD6UAOFWYpTtEfG08cjNMDHuPA2ntN9rtLi9sZiCWNpNtznr1B49ulUE8G6bbPLc3El9qEmMsrMGd/wAQASfxrrjipdVqFjh9e8VT3qyaVZaZHp9ssmGXaN7FT/Efr/8ArNdJ8O9Dnt/N1C4tpoDIMo4cBZR05HU+2eK7Kq9nQfNq2B6B2p5j4684zivHAj6U3cAevPpQBHPGl2hinijkiPVZEDA/geKypfB3h+aTe2mop9I3ZB+SkCtoV5w+FgX7DStP0tSLO2SLPVhksfxPNWieaidSU3eQCUx3VepxngVAEZZj0GPc0zZzk8n1NAC4pKYDGPaoWYDjPPoKQxh3Hp8o/Wm7QOep9TQIaaZjkfWkMpaDJH/wjGkuCFU2MJHb+BavbmYfIv4sMUxEixjrI24+nYfhUykA9KAHjrVlXUQY746etAEecDk8Vyni/wAV3WhpHFp8StM6FzK6khRnAx6mtqFP2lRRYHm9x4t8Q6hIol1a7yGyDHIU5P8Au4retfHms2ely6dPKJmkXbHcE4kjHfkdfr1HrXsvB0mrJCJvBmhjWtW3zj/Rohvf/a56fjXrU00NrCZJXSKFByzEKqiuHHybmoLoMzLPxJpGo3ZtbTUIZZucJyCcemev4Vp+ecYX5vpXDOEoO0kAfO33iB7ClCBR0qAFFITQAnNNLqOM59hQAwlm/wBke3WkC45oAWmmmAwmomkAOBkn0FIBmHc5Y4HoKTAUcUDGmmNQwI6jaRQ4AOTnoKQih4XiX/hFdFbGT9gg5P8A1zWtnaapAJg5pwoAmaRFUbiBTBIzHCrgf3m/wpAOEak5bLn36flTL3SrDVYhFfWyTKOmcgj6EcirhOUHzRAyL+38N+DrFr1LGFbgA+QhYszP2xknp3PavL9E0i68S6qIIMF8F3duAPrXsYWpJwlVqMR2GlaifA+qXEWq2lxHbPHsWRIyVZgcjB6HPPNVdS8RX/jC+jt9OtHeBD8kIG7k8bmPQf0qXCM6vtm9BnY6N4PstOjjuLuCKe+OGLkbljP+zn09etdEOAK83EVXUncQ4HNLnI5NYDEyBTPMB+6M+9ACHcep/AUm0AcCgBQKQ4oAaTjmoTJuHyfMPXtQA3aScs34ClwB0FADWNMNAxpwBzVZpAT8o3e46UAMYMRyfwFIFAIwAOaQjz7TfGV5caHpWlaFas1wllCjSMm5shFB2r6D1NUtf03UNM0tr/VdbdL92Hl2xuMuRnk8Ht6CvXw9GEEubVsls0fAWta9qtwY/wC07ae2i+9FcENLt/2cfN+J4r0gh24zs+nJrlxkIQnaI0OVVHIGT6nk1IOtcYxwIqvql5d2ekXM9lam6uI0LRwj+I1UUm0mB49/ZvivxPrQ+3Wl0Hc/M80ZjRB6DPQD0FeseHtBi0GzEMOwuw+eTbyfYf55r08TWjCkqUBG0ADnd82Rg55pUVI49kaqi9lUYFeZzO1hiljQDxUgBYDqcUnmH+FfxNAChC4JPOPWlAoAMUhNADWYKMk4qPczcquB6t/hTAaUB5Y7vrRmgBpNNJoAYzADJOBURcnoD9TSAjZd/wB75vbtSHrSAYabn5hQMwNIjmi+H1lFosEcV4+nwspRAoLlFyc9Cev41wVtpjw3xuvEWn6jPbqrHYFJaR+Pvk8gHnn2Fetg5Xi31IaM6Sd01lrzTrZbD94HiSHgJ6Y9a91sp2utPtbhwA0sKSMB2JUH+tGYR92L6hHclpwryihPMUHA+YjsKeu9up2j0H+NAEqgLwKdmmAu7FODUAIXUHGefSj5m/2R+tIBQoBz39aeMUALntRmgBjOoPJ59O9Rs7PjaNn6mgBVQLz1PqaDigBhNNNMBjMFPJ59KaSxHTApAMKgc9T6mmmkA08UwnvQBC8gHA6+lRguzDsP1oAr+GmDeF9IKqFBsoCFB6fItal3f22m2Ut5dyiOGJdzMT+n1q4XbsgPGJ9Wi13xI9xNcxQC4l6uwARegz9B/KvWote0KOFI49XsNiKFA+0pwAOO9epjYy5IxXQmO5ha98RNL0lUWwMWpTMeVimARR6lgCM+wrV8M66PE+k/bvs0sChyjI3KnHcH+IdvrmuGWGlGn7RlX1NzAXhelPBrnAerkHIODS5oATzB/Dlj7UvzNwTj2FAD1UAcDFPpgLxRuApARmZf4fmPt0pMyP1OB6CgBVjA6CncY47UwGl/lxjn1ptAEZcZ4+Y+1NO4jk7R6CkA3AHQUE0AJ2qNjQBC0nZRu/lTGUup3OQR/CvekAwKFGBxS55H1oGUPDgmHgvTJooWlMenRMFX+IiIHFeV6vbeKvEuqqt/YXyK74jVoWWNPpkYH1r0MCoXcpdCWd34f+HmjWFgq31lHe3bcySTZIX2Ueg9a0pvAXha4TbJpEI90LIf0NTUxk3J8uw0iGD4ceFIJRIums2DkK88jL+W6uohiitoUggjSKFF2rHGoVVHoAKxq4idRWYWJC2ccAYFBYAZJwKwATzGJwi/iacEJ++c/wAqAJBTgaYD94pDIFHJApAN8wk/KOPU0nl7zl2J9u35UAPCgU4UwHBlwcgnjjFRE0gGbvQE0hGepOPSgA6Cmk0AJnimMwHegCPzCR8qn6txUZXP3jn+VACdKbSGMNRs4DDJxzQBV8Mz3TeFNGXlUFjBgAf9M1rUCnJYqST1J60xDwG9D+VKA+fun8qBjtrDsfypfugljg9hjk/SgQh81+FUr7kU4QnOSpJ9TQBIFb+6fyp+D6H8qYBg/wB00hyDyDn0xQAo83OQm30J60qxHOSCT6mkA/a3900Yb+6fypALhj/Caadw6gimA394c4Uj3IpPLOcupb0B7UAGG/umk2t6H8qQAQ3901E2QOQfypgNxKeilfqKb5TdSpJ9TQAhVh/CabsYn7ppAIUYfwn8qgdiDjBJ9AKBjCkrdio+nNN8oqwO0k56mgCfwt/yKWjf9eEH/ota1xTESL1pR94Uxjn++agX/XvSAnWpTQIKD0oAB1psf3m+tAEtFACjrTT1pAOqM/6z8KAHDrQ1ADaKAEqKP7zfWmBJTO9AEbfepKQAehqnB90/U0DJD0qM9R9aAP/Z/+Ex5Gh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8APD94cGFja2V0IGJlZ2luPSfvu78nIGlkPSdXNU0wTXBDZWhpSHpyZVN6TlRjemtjOWQnPz4NCjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iPjxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+PHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9InV1aWQ6ZmFmNWJkZDUtYmEzZC0xMWRhLWFkMzEtZDMzZDc1MTgyZjFiIiB4bWxuczp4bXA9Imh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8iPjx4bXA6Q3JlYXRvclRvb2w+V2luZG93cyBQaG90byBFZGl0b3IgMTAuMC4xMDAxMS4xNjM4NDwveG1wOkNyZWF0b3JUb29sPjx4bXA6Q3JlYXRlRGF0ZT4yMDIwLTAxLTEwVDAwOjQwOjIyPC94bXA6Q3JlYXRlRGF0ZT48L3JkZjpEZXNjcmlwdGlvbj48L3JkZjpSREY+PC94OnhtcG1ldGE+DQogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICA8P3hwYWNrZXQgZW5kPSd3Jz8+/9sAQwADAgIDAgIDAwMDBAMDBAUIBQUEBAUKBwcGCAwKDAwLCgsLDQ4SEA0OEQ4LCxAWEBETFBUVFQwPFxgWFBgSFBUU/9sAQwEDBAQFBAUJBQUJFA0LDRQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQUFBQU/8AAEQgCGwO9AwEiAAIRAQMRAf/EAB8AAAEFAQEBAQEBAAAAAAAAAAABAgMEBQYHCAkKC//EALUQAAIBAwMCBAMFBQQEAAABfQECAwAEEQUSITFBBhNRYQcicRQygZGhCCNCscEVUtHwJDNicoIJChYXGBkaJSYnKCkqNDU2Nzg5OkNERUZHSElKU1RVVldYWVpjZGVmZ2hpanN0dXZ3eHl6g4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2drh4uPk5ebn6Onq8fLz9PX29/j5+v/EAB8BAAMBAQEBAQEBAQEAAAAAAAABAgMEBQYHCAkKC//EALURAAIBAgQEAwQHBQQEAAECdwABAgMRBAUhMQYSQVEHYXETIjKBCBRCkaGxwQkjM1LwFWJy0QoWJDThJfEXGBkaJicoKSo1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoKDhIWGh4iJipKTlJWWl5iZmqKjpKWmp6ipqrKztLW2t7i5usLDxMXGx8jJytLT1NXW19jZ2uLj5OXm5+jp6vLz9PX29/j5+v/aAAwDAQACEQMRAD8Azv8Agl8xX9nnxHg4/wCKnuP/AEltK+tvMb+8a+Sv+CXqg/s8+Iv+xnuP/SS0r63VBWghUZvU0hyO5pyrQcmgLjl3dzUqlvWo1qZOtIkd5h9akWQ+tQttX0P41Rk1IbtsTBm77fmxQBrbjH9/+dRyXkcP35FT/easdUv7ySPc4WP+Ldw1aUOlQx/fRZP97mgCvJq003FupI/56OOKmhtp5/8AXSEf7tXto9OPTtTlAoAjjjEf3Pl+hqXc3qaXFKtABub+8akV29TSbR6U4LVASb2/vGje3qabT6Bhub+8aXcfU0lFAiTcfWje3qajyaMmlYCVXb+8aGdv7xqPcadSAk3t/eNG9v7xqPJpVoKuS72/vGnbj6mmUZNMYu9v7xqbe3940zaPSlpkAjN6mpGdsfeNR7gvSobzUILOPMjhT9efyoAtb2/vGiS42jcHwPSqC35uI90XA/vYojiMhy3JoAnkv3/hJqJbiVupJqwtuPSnrCPSkMgG9uufzqQRsasbQO1LsoEMXdtxnipNretO2+1LtpgJtPqaftPqfzpdopaABWPqaGY+tOjUHtSug7UAN3H1pRn1NGKWgB6SH1ojkPqaYPalVdtADgzf3jS729TTV5p+KADc3qaZub+9UqrzzSbBQBHhvX9aPm9T+dS7aQrinZgM3H1NNJI7mnsvpUU08UK7ncIvvTsMlBb1NDSCP+KqC6xHcNi3bzD6lSP6UNbysN0r8elSIkvNQktIWYCWdh/zyXNUrW41S6bewjgh/uvkv+PatQQJGSVHJ9zUo+Vdo4HtQBnrbc5xk+vepPLbrz+dW9o9KNo9KAKyq47mnYb1qfaPSjaPSgCvub1o3N61LtHpRtHpQBFub1p24+tLtFG0U7AG9v7xo3t/eNNoosAu4+ppu9v7xpaZSATe3qaN7f3jSUUAJvb+8aTe3rSUU7ALuPrRuPrTMmjJpAG8+v60nmH1NRzTRQrudgq+p6VT/taGQ7YyWPqFNAGgzn+/UDzFerFqqoJpP4sfQVMsJpDIJy9x90kVRtdDt45C++S5f+9M24j8TW0IQvQVNgeg/KgRRWHb0GKcI2XkZ/OrewUzbQBFvb+8aN7eppWXHQUxs0wHb2/vGje39401qazGi4AzN6muhsb9ZIcM+GrmmY01pGXocUWYGhql15k2UPy1n72/vGkLFupzTMmgBfMb+8aRpG/vGmtTWJpAO3H1NRsx9aXmmUgG7j60bj60m00bTQAMx9aj3H1p/wBQT9Kz77UYbGPLyLkdlyxP4AU7AXGkP941FNeJbpuklVPdjgVlteXV0MwgwL/ekWiHSQX3TMZT/tcj8ulFgJZNQ85tsBaQ/wB7+GmLFIepNX1tUUYCDHpT/K9qQFHy39T+dSBWAxk4znGTip9tJQBCpPYkU75h3NLtFSLGTQBX3H1NG4+pqx9lpv2c0AV9x9aRmPrU5jX0qLaPSgmxEzH1pNzetSMg9KZQUNwfWotzepqek8selAFbc3YmtnTnbyfvGs14wvatTTxiE0AfH/8AwS9/5N58Q/8AYz3H/pJaV9br96vkr/gl0uf2efEX/Yz3H/pJaV9cbNrc8H35H4ntTQx67e9G1G+6x/LNULjVraBcrJ5x/uRDcarL/ampfNCgtIP+ehO1j9BQI0ri4gtvvzCL/eIqmmsPdKwtE3MehYZA/GpIPDUO3FyTdN/03YNWvBCkCBY0WNV6KoAApAZaaTcXZ3XcoA/uQ5FXbXT7ezXEEQjHsTVtmJU55plAC7c5yOtOHy9KKKB2HUq0YpyrQFgpyrS0U0FiXFFNyadTEFPplLk0AOp20U2jJoAKKKKACnU1ak2ikx2E2mlWn7RSADsKQh1FRyTLGuScH0NRNdFvuc0FXLjuqd6ga+QfICGf/ZqDa8n+sIH41PDaovRAKZJF50kn+z+FJb6Pbh97rvb+83J/Wr6w7egqRV9qYEK2uOAcfQ1KsYXoKk2iloAKKdtFG0UAJtp2007aKKAF2ml5paXFACUU7bS0ALHxSvQlK60AMpdtB4pyjPWnYBAKdtpwWnBfaiwDFWuE+NXxOHwl8I/2kLNr2eeQwRbeFR9pILeg4r0byVVjx+tc58TfAkXxA8A6lo8qp5zRNJCzDOHAbGPTrXVh4RlU9/ZgfM/hz9t69sfKTxH4YaRRlXnsGJJ54OCQM19BeAfjF4S+JdsDo2pxrdbQz2dwVWZTjptzz1r8/wDXFghmnh+aK6id4JoQANjKSDn05rN0zWJdNaWWxubjTLxWWSGezfy2znJDFe3FfaPJaVSCcGK5+muveI9J8M28k+pX0NrFGN5DyAMV9h+BrwXxL+2l4W0fWI4LSxu9QtN5EkkMLZA7Ec818wR+INX16TzdW1vUNSZF+7cTs+evqfet2+8DzHwsutQASwRTeQ3AJwPWsoZArXkxcyPqzwn+1B4C8aXi28OrPpMkmAF1KMQkEnGBuNemJpNvPGlwbhb+JwCGRvl/nX50x+DbbUoZS/7qZBmMqwVg3rmvob9l/wASeL/D9z/YuoWOq69olwcx3iwvKLTIACseOB/WuDGZXHDx5kTzH075KquFjUfQCn7BnJGaU5p5U+lfKSNBm0U2pdopNo9KAI6KVqatAC0UUUAMop20U2gBtFFFWA0KaQg1NCy/x80xsUAMplLk0zJqbAFFFMMgVs7vl96kApTVG51SOHaFBkPdYzk1T+3ajdSZRIYLb1bPmfkeKsDQuLqO3++2Kp/2lNNxBGCP7zf0qSHS1i++Xl/66Grixj+6PyqQMxdHeZszymX69PyrRSNYxhVC/QVaX5elR7RQA3yx6Y+lO2j0paKAGbfailyaSgAptGTRQAySo2pxYnrTWoYAwqNlp9JUgRMtRyKamamMtaDI6TaKl2j0o2CkIhYCo8VKyn1qPFKwCsKj2+1PeQKuWIFZ9xqbRttt4zO3+1wvNDQFpmA77apXGpeTwi+a/otEVtNcANdPgt/DEMY/Op1s0ByFwfrUgYzW+p6pLukmWzhXkLD8zN9as2Oi29mzbIi8rdXbJP61rrHtxgDFP2j/ACaoCl5IPJyfxNTLCV6CplTHal5pXArUY9qlZB6UhUDpSAr7R6UeWGqRlFRkt60AMEI71YRAe1NDCpF4oGPCCjy19KPMApPMFAjPbpUG2rU2BUIWmgGbc+9Rsg9KshajZaYEW2l207bQy4pWAjda0rFP3VUGU7d3b6Zq7a31vBCBLPDGx6bn25/OkB8Rf8E2vE13pnwG1yxsNKuNSuZPEtw/7rAUA2tqOSfpX1gvhrVtY8ubUtQa3iwGa2tSUPrhq+df+CWIH/DOviI45/4Sm4/9JLSvrpfmU57nJoKsUdP0a10sfu7dGf1Iyf1rRYbv/rcUh+ZsnrThQKwbQhyOD9aVabmigQ5ulNpxGaKB2F206m5NOoKJFpcULT8r6UCuNoooWmguGTS7j60KuaFXNMkkopdpo2mgB1FO2ijaKVwG0U7bTZnjiXc8iov1/wAaLgOxS7gvU1QfVDNxbRl/9rtUkcMs33ztpFktzqUFucF9zf3R/jUaTXM+Si+Wh+6Wq1b2SQjDKpHoRVlVBAHYdKZBnw2gVtzkuferC2o9KuhAvQUbR6UwIY4/apFXFL0ozQOw/mlVaWnUCDaKNopdpo2mgB20UbRTcmnUAFFLtpy9qAHbaWnFfue9EahutADtg9KXyxQzEdKWPLdaAGxrTmWlRfSniPj5v72Pf8qYERQelPRKr6lq2m6LC8mpahb2SKM7riZI8d+c14948/av8L+EzJFpsTa/Mg62jqYzx65rpp0J1XaKA9+s9Pt5rWUucSKcqd3JH0qBbGZWyY2ZfpXwL4w/bM8deIPPXT4ofD9rICA0cYklT/gVcRa/tHfFVWB/4TW8MYG5leJD+HAFenHKcTJXsTzH6ZyRrGGcsNq/e5xVOO9DMVi3Ox45r5p+BX7VFp4r1Sx0TxhIkGqTN5cN0gCQztwAvLdST6V9TixVVG1QVPTHv3rnnh6mGa5h3Pgv9sL4at4R8aWniKFPL0/WgUnEafItyWbknHBYYrwFiX5HBzngV+nXx38AxePPg74g057ZZrpLR7i1B6pKqkrtz0PSvzTs9NdoVBTEiHbJnpnv+ua+6yrE+3p8vVCKlm09pdRXCM7hfvR5NfRHwI8P3PxQ0m5ghWRbaG7KSZfAVsdSOhrxvSfD93rmpR2FnAGuJjtTJ6+/Fe9/sN682i/EbxX4QuW2SyRfbY4yQNpUqhH5mvQxlSWHouUCbHt3hn9m3wtoc0U10smoTqcjzGwn5d69QsdNtNKt/Is7aK1hxgpCgUH6461YkQxSg9COlG4sOtfnGKxlWo7SYrFZqfnFNZRTj0rx5Go3bRtp1FAyMrTSuOlSUmKBEe2jbUm0UbRQBHtpm2piRUdK4uYj20baN1Jk1YxjA0hyafxVW61C2s1JlmWNjwqseS3oBQBNtNMkxCNzsFWqg1CW5jzBGR/tMDj+VQLpsjtmdjIfrj+VMBW1YZ2wqZ/UgdKb9hurwZlnGz0RcVf+yqRjaAPYYqVU+p+ppWAoxafHBjag+vepljKjAGPwq6FC9KTyx6UrgQMu7rg0bak202kAUm0UtFADKKdtFNoAZRRRQA2inbRTaAIaRql2j0o2D0oAhop/l+1JtFKwEe2mc1PtFRSMFG7BVau4DKMiq0moQRyeX5u5/Qcn9KZulmO1G2e+BQBJcX0FsqmRgmazGv7y4/1FtsT/AJ6sR/LNS22iwWszyspeR+pZiw/InAq+F5zigDKjspHbdJuY/Wr6wBVwBxVry/apBGo6ClYCmYi2c85p201Y20m32qQK+DRtNWAopdo9KAK1FTmNR2prKBRYCsy0jCpKRqQEDKajINWWpuygCBk29KeSV6GnbAetN2t60ANbPNMb61NtNRbT1oAgYFutKsdWfKB7U0IapDI1jpjx9KneeGEbnbavvxWTJrf2g7bRBMfXrQIuYHYhfqazrrVre3bZ5nmv/dUZP44qSPRbvUZN11KUT+6g21qWugQWS4WJfqOv59aLMDm/K1PUF+aX7FD/AHVHz/oa39N8N2n2f96rXD/3pfmP61daADoMew4FaNmEEfWlZgfF3/BLH/k3TxH/ANjTcf8ApJaV9dR/dr5H/wCCWA/4x08R/wDY03H/AKSWlfXkajpikWNWlXrR0oFAAOpp22mg80uTQKw9qSmqxbrQxIoGOp1Np1Arj1Jp9RrUlBIUq0lSKo9KaAVR0pypSKKetMBKKUsF6sB9eKqS6tbxvsTdPJ/djXP60rjLn0/WoZLyKHdudS3p0qu8Vzd/MT5UfoPvf4VLFpsUed377P8Ae5pAL9peT7n8qrQ6PvbfOz3C/wB1sVo/Z/apljHpxTEQw26wjCKFFTqMU/aKWixYrIG6inqo9KSpNopk2G0U7aKNooEG0elG0elP2im0FhT6ZT6CB+00bTRk0bjQAeWfSlAoVmanqtAw2ijaKKctAh7Dp7dKdGpo5apY8Uk7gJ5W7tT1hAqTj05qOa8hs4xLO6rH6k4FNajFWMVV1yzvb3Q9Sg0+7FnqEkDRwzHOEcjg8eh9Kil1hrjjT4vtI/56fw1JY6Vfy/vb67jcA4SOFCmPrV2YHyl+0L4S8TNoVjJfTXmtalAf9KFuzvCU6sQO/wCNfO1nfwzQIpgW3l5BWT5WHt1r9Sf7JtZyTJbxyFuu9Qc1zHiD4K+CfFFjNa3vhyyAl6NBEscg/wCBAV9VgMwp0NJRA/Mu5sfJMisduRuDHv7Zqi1uyyNtGdxycV7/APtEfs72nwk1nSrzRp7ifR9Sm8kRXMhkKScnaOMdBXE2fhSeaFporM4X7zOMfoa+6oVo4iCnDYyaseXzWd9bzJOoaCSNw6SqdrIw6EHseO1fbH7N37VGn3Gl2PhjxlfeRq0aiOHULh9qTcDAZmbgivAZ7WGKORpokkZP4JFBz/nmvXPEH7IOn+PvBumeI/Bdx9nvLizS48mSUlfMZcnBGcHOB+FebmdOhKCU9Lkq59mWc9nqFuHinivIZB95GDKwI7EZBFfnh8Y/AMXw/wDjJqGjhJDbajcNd24Odu2V2IAOMcEEVL4F+O3xJ/Zvuo9G8U2Et7osb4db1GMka5wdrE46fhXdfGL4leCvjh/wj2u6DqED6nZMi3EEp2yLGVYnjpwzD9a8XA0qmFr+7rFmh5g1qdG1Im0bZKowskfysPxHNP8Ah3qK+Bfi14e8QEiFru9is7mZv+ebuNw4P0PWn3m1bpiDlW+6R3rI1y2E1usu0eZC4kjccFWHQ5r63FQ9rScRH6Q3kiyNG6AMjBTuU8EHoai2iuE+A/jSPxx8K9EuQ++WzjSwnYkcvGi8nn3rvNpr8nxdNxlbsBXbvS+op8igdKjWuC1yxuTSFjTcmkZj60WAXcaMmm5NGTRYB+6m7/elpNopAR/N60Yb1pzzxQjdIQo9KzJdehmO20zOf7ycj+VImxeJHc4qpNfLEMAh3/u96rRx3lxJm4ZFT+5GpDfrV2G0jXkIM+vetCip5l3ccx4jX1xlv5U6HS4gwaVfPYHIaRRkfjWkIwOQoBpdue1AEaoOmBj0xxRt9ql2ijaKAGFcdKZtxU9RUrgFFFFSAymU+k2igBtG0UhzRzQAbaZtNSU2gCKkqXaKNo9KBkVNqTbTStAhtFFRyXUUP3zj/PpQA9qhkkWP7zKv1NQQal9obaImxj7/AEX9aa0If7w3fWgAkv0Hyx5lf+8OlVWhkuG8yRmA/uqau+Qg6KB9OKk21IFSOyjiOURVPqBzUpjDDGKteX7UzZV3AgKBuopSoHQVNspu3PajmAjp9NHWnUXGMpcZop1PlYrjdvpRtNOANHNRZiuNI5pki+1S4JpjZ70Wa6DuV9tMxU3NNZfanuMjNG2nkClC0rE3IgtOVKkEdOhUN3z9arldrhcPsXvUE0UdurF3VF6h5OOB149KtXF/bw/KGVpf7inmuF+Jnh3VPH3gzVNKtrq60l5IX8qWxm8qTJU4G706VvThzuwXLknxC8KRSG3HijRZLtfla3jv4jIG9MbuvtUx1S9uow1taFEblZJFOCPUHPSvyf1rwHf+HPEWoxz3WoQXtvO0bXDTushYHqScEH3966+1/aa+I/wk00La6zc61aY8sRalO0oQf7P/ANevb/sqXJ7ToUfpc2nw24aXUdQR89EYjj9KrW/jLwzZLtj1Wzg/3JUH9a/KTWv21PiFqs7GSeG3LdfvH+teda18fPG+tXfm/wBt3cJ/6Zysv9a82WGjER+z138U/DOn7ZJdUt2QZ3DzU/8Aiq8e8cftxeB/B979m/tS1llXhozscjnviUV+VV98VfGN7blJ/EWoMjdcXDj+RrkpppLmRpJpGlkbku7Ek/iaxdFRA/WDQ/2+vCfiGYQQ3EaynorJGGP4eaa6Cb9rbToWCxbgvUEwqQR2x89fkJZMY5Q4ZkYdGU4I/GvWtB+JlxFp8cd04mZFCqXznH510UsKqm4H3X/wSv8A+TdPEf8A2NNx/wCklpX16nWvkL/glf8A8m6eJP8Asabj/wBJLSvr5MV5RYym7jQzU0tQA8feNLUatTs0AOSh6RM1Jtz2oAKdRt9qeqiggVVFSqo9Kaq5qUoP4SPxNMBKcql/uiqzXka7lVw0g/gFV5PtFx/q3MX4CmBdmuY7dcsyj2zz+Aqk2p3NzHusbXf2zMCo+uMVNBpcUTbpQ1y/96Y5I+lXo41jHyjb9DSAzYtKe6x9rJkZuQqkjHtWjDaxQrtSNVX2FT5698880UgEwKcAPShRT0UUwDaaftHpSZNGTTHYSiiigofUtM2in0CuFFFFBIuTSUu00vl0FXG0/FPCD0pGJyaFqSH40ZptSbR6UAKgqRabipFFADBT6bPIlv8AM5VF9Wbj889aof25Bu2Qq9y/96MBl/Q0Aay7Q2WOFqtqGt2GkRNJdXcMMca7mZpFXjHfJwKrytLNaNLcOtuqruOO2OtfJnxu+MSarqMmk2EgisIziSZc7pG5yOOCP8K9DCYSeKlyoVz2Px5+0x4e8MmWK1uvtEyjcPK8twRjp9/rXmC/tnaHZy+ZceH9T1MbeknlgbvoWxXzDrXjAyboo7G2gSP7si7g7frWHcassy4U8fQ19rR4fpctpsLn27pf7fHhSST/AEvwjqlpFx/qUjb+T17r4D+Mvgz4l20U2ia3avKwUizupUjmU+hXdnuPWvyxtLj3q4bPUrW6N9peq3VheKd8U9pM0Tocg4BUg9hVVeH4Sj+7ZPMfr1GNq4YbWxnjkfjSgHOcn86/Pr4J/tqeJvAeo2ekeOGm17RdwV9Qk3vcR5xyxLYIxk8196eEPFmkePNHh1TQ72G/spF3CSGRWPPODz15FfKYrB1MK/eQrnC/tJeC18YfDG5KRNLPpr/bo1jXnciseOPQkV8Wrrc+pL5jyyKyfJt29F9OO9fpH9lRoJInUNHIpVlPQgjGK+F/iHoI8H+PNf0YwQQYnNzA0S4JjY9K+hyTEWvTZT1PPLomaPPzEL1YjI/PHvX1X+x54omvvBFzoV18lxp02YFZjuaFgu0jnnncOlfOtq1rb3Zi86a584bDFIvyBsjBH616B8EdcbwL8RNPmndksL7dZSBmwqK2NuR045r2czo+2w7fVEn1Z4w8A+HvHWly2mv6Zb6jBIpU+dEGOD2z1H4V8c/E79h7VPDzXereA7qJ4QTI9hdSlTtBJwhVOc8da+5LjCuc8hTwfXPQ1H9/O7kHg5718Th8dUw8lZ6Dufl14Z1TUIWurHVLaSO6tZWgfcuAjjqD9K6K4jEkBUjcDXoP7SHg9PBnxdu0tUbydXjF/wCW3Zyx3Yx16VwQVrZdjqyv/dav0ehWWIoKS6iPbf2QfEQ07XNU8NFgtpLC13GGznzNyr9Ohr6eavgjwd4ok8I+MLHV4XaLyZEDqOQy55H6V936bqEOsaVaX9sQ0M6LKuO6kDFfEZzhfZz5lsND5elRLUkhOKj6V8jsVciprU6kdf3e4HI/vAcfjTGNoVSexP0qpNqUUR2qRI3+z/WoJJLi7KgZhH94H5vyoAvyXkMYwzYf+6Ac1Va6uZm/crhP7zCnR2BVslzIfU9atqDtx2qbAZ62EkjZnbdVqG3EIwiqo9AOKulQ3WjaBU3J5ipgDoB+VOUYpdopa0KCiiigAooooAKiqfaKZtHpUAR0VN5a+lIyL6UAQ7RRt9qMkZxS5PrQAymU+k2igBtNp+0+lJtFADaKRpEjG5j8vq3SqjakgbbGBIfXr+ooAsnHrtqJpgvXmqzebcfeOKljgx2oAzruTUJpl8iWOCAfebblvyP41PZ6dHG+X3TS/wB9s5rTWNVVgF4br7/5zTqVgKj2+TnGab5Rq3RQBV8v2pdpqfZSbfamA2o9pqbaKMCgCMqMdKjXFWGUbsVAqigCIKKXaKcRimMxHSktSRq05aQdKenOK3sAKGpl1cW1gqS3lzBaQHrJcTLEg/FiBWD8SPiFpXws8H3viDWHVIoY28iItzcSBGZUX6la+FvGHxy8YfFKeW81GY2Gi3GTFp8M0oXaehKl2GccelenhcvrYp2SA++Y/FHh2dd0fiLSz9L6HH4HdzV63mhu41ktporuJujwyLIv/fSkjPSvyg1zWJo4ykV5exIv8KzEAfhmtb4X/H7xf8OrphpGvSsucfYryWR4n+gz14H516ssjqxHY/UZlwOKjKnvXzP8H/21LXxZqEek+MNOj0W+kISO4t1KxMxxjduc4r6ZN/Zx2sV008f2eTDI4bII/A14tbB1aPxIQjKPSvBP2gv2prf4R3LaLpemrqGt+WH3TN8iArkEgEV7A+uXOoMw0y0kdP8AntcLtj/PrXgP7WHwHn8YeFT4ssYTJrunxbrmOIn95GAcgDuR/WtMHh4VqtpAeH6L/wAFAviLa6jI2paHpN7ZmT5Y4UZX2+gOf517j8Mf20vDPjrUray12G58NXEzBNswQR54/j3fWvhqG3KSmKU7SjfMnQjHGKurosV4+MNE3bbgYr7N5JSlTvFjsfrPZ2cKRrJbmK4R8MsyPvyPUH8avIvevn39jvx/JrXhGTwvfymW/wBO+aN5GJZozgAc+lfRIjIr5Kvh/qtQk+Fv20fBcfg3xlBq9tbLHDqgDuQoCeZ82c8fSvlPXrNdYs5LcHIkUhcjpnp+Vfpn+1d8NY/H3wrv5lgMt5poa4jCYy3BB61+dVrbRbcH5ODkkcjsRX3OWyValymx85a/4buNLuGDR8DuBWOtvJH1+U/SvqrxF8PNP1mzPLbv72a8B8YaIml3Twqd+wkE/jXi47LfYSc0ZHKSSDGM8VGEFSSwGniMeleC6cpPVAQrlenFW4SSg71XZSMAVNDnZ1q6cbSsB+nn/BLE4/Zz8S/9jTcf+klpX10GHQV8h/8ABLPj9nDxKe//AAlNx/6SWlfW67s+9fOGpIymmMpqTNI3SkAiqKlUL6UiqPSnqntQK4Koz0p4WjaOx/Oo5r6G1XMrLn0Xk/kKAuTU17iOFsMyA+7VkQ6pqF/kwW/kxH7rSqQ35Vdh08OcyuXf1NArEn9qZO2GCSU/3lxtqZbWS4/1hK/7hqzDCq8KoUewq2rYXjimFjPhsIoclEAJ7nJNT7alZjmkpiD8KkUVHUi0mA6iil20gBakjplTqoXoKoCOinbRSbaCxKKKdxQAuTT8mm7adQQLzTlpKeq80AOpVp+0elLtFADCwiUtJ8qryzN0AHU1BY31nqwZrO9hudrYYQuGx7H04x+lWmVZVZXG5WBBB7gjBFfC3jPUPFH7M3xybUEuLl/DeoXfnrE7Mbd4WfJXHTcAPSuilSdTRAfc5WjNZGkeLNP17R7TUtOmW8truJZFeIhwpKg4+WtGMvN0Gys5U5QlysVxbi+gtxh5FU9vU/hVRtQvZMLbwZPZm+6alXSIFm80DzZP7zEk1oRw8YrMZjW/h97mTzru5kkk/wCeaO3l/gM1sw2kUKhY4kjX/YUD+VTeXt5Awaci07AcT8br6bRfhL4hvosrJHbv8yDnkE5r869ckY2URUEFizE/XpX6d+LtATxR4T1bSZkYi8tJY+OeqnHtX5waxopXUb7TyPLa0d4Sjfe+Unt1r7rIEuu4rHmckb3Mm/bj8KdHabejGt5bddu3tVm10lrhtowT/u19yl2I1KGl6T57bVfJxXQWmlG3XDOcda2PC/h37FdLLd7WSQbFVT0P97P+etdpP8O76+0G91myizYWofezdDtGSAeeRV6RScnYDxnWtPMkUsbrujbgq3INbPwd+NHjD4B6y11oWotcaTKwMum3OZICPlzgcEcDHFaF5pa3Cq2xgG61yl/p7RrgnAz83ygcVy18PDERtNEK6P06+Bf7QGgfHPQjPp8n2TVYEH2rT5iBJG2BlgM/dyeDXmn7Y3hN7eLSfFVnGivHOtncMwwWVs4P4cfnXyR8H/FV78HfElj4w0qXFvDMtpqMfO2WEsGYADuNor708Z69ovxu+C+oXWh3MWoJLb+fCFIcpIFJAIGcEZGRXxNTCSy/FRlH4WaHx5awvJ5cvTADDjuK0NW8SXMOmvCLeMuQSkvO5W/hP61mXWpPbskPkLGFGGyScEZzUUDXGo70fcFT5mOOAPXdivtJQ9pTsSfeHwo8bp8Qvh7o+ss4Fw8IguExgCVeGx+NdZ93GK+c/wBj/XpIbPUdElGIWQ3MDbu6th/b0/Ovok5x1r8vxlF0azRSPAv2u/Ccd54T0vxMgVLzTZliZ8nmM5OPzr5ckkNyVO4kkZ5Oa+9fiZ4fTxh4A1zS5YjK81s7IuOjBSRj3r4D0/eNPgllHkyfMrq3UFc5FfW5HW5ocgixcWxNmxiG0joRX1z+zL4sXxL8MLa2dgbvTHa1mVuDxyD9MV8jrfeTGyMflbOBXp/7JHji20v4g63ocsmw3lqs8SbsFmVh79cGu3N6PPQv1Gj63lYVDJcJGGZyI1P3dx5+tUpNWaf/AFEJx/efio4tL+1MJLh2kUdFJx/Kvy+S1sAk2sKp2xQyXX+1CP557UnkT3XEsjRD0Q8VfWxSMYVVH4CpPLHpU6l3KEenRQLtRQD69TVm3tmkwAOlTeUd2cHNXbJcZxSuK5ItmpjjJXn+KqlzAsf3RitmNeCO1UL2H5sUhlDJoyadt9qNvtU8rI5SvRU3lrSeWKssiop3lmjy2pgO2j0p3lj0qPc1G9vWgB22mgUuTTGYipsA3c3rRuNJS4osAyin7KQqu3cp496LANoqrdalbWbbWk8z/rmNx/HFQ/abm4OYl8tPVhSAsyXkdsmZjtrPOptdN/o0ZK/3+34VYXS49+6djKfcmrUcCwrtRQq+g6UAZ7Wkk3zyt/wHt+VTJCqrhVCj0UYFXFj2nIHNG32oArGPdTlqekCgdBQBHRT9tG2gCPbRtFPooAiop22jaKAGbaSpNoptADG+8PpUC9atbQ1Qsm3GKAIWqN6lxTdobrSWhIyp7dNzKp65xTVQelOgYx3Cuei9a6IgfFH7WXjKbxl42l0Hc39k6ZIFmhJyGmRm/h+jV5Rb2L6tHFa2qHoAuzACgdua779obQ5PDfxk1u3mR8aiWv4m5JKsx/wP5155DfGK72K22Sv1DKYqNJGsVpc43x94dudIt2fymuGP3xFkgD1rzJ5vLZSp2zdR/eH9e1faX7Kuk2/jfx9rNpq9o0sUNrIyeaPkfJAzXr3j79lH4FX8LSavPp3hycru86C+ihkxyTgN17104nGQpTtcfLc/NOxv7qa4V1uJXZTkMXOc9q/RL9iPx1f+MvC+raFrsj3d3pkmYGcnIhKIACT3BbrXzP8AEb4M/DTw7qSjwV44l1hIztkt7pBucj+4ygAivff2d/iL8MvhHayC8v8AUIdavow0ryxKV4CgqPb5a4MZH21DnirslxPrSK3WNSqKqj2FTLaLPG8MqgwsNsinn5T1J9qxPDfxD8JeMWH9j+INPuSRu8n7ShcA+q5z3HFeCfthfHi48I6bH4T8L6tCNVvEZLqaCQM9upAPY8dT+VfI4fB1fbaKxPKz4j+O2p6bofxk8RwaMUnsI7p9jIQedzZAIIHHFc9p3jLZcBjECv0H+NOvvD8s1rPJMTPO7b/MYksW9c1yskTQttHymv0mhGdOn7w7M+kPgx4+/wCEH8ZWGsRTskHmRxyqCTlCwyTiv0k0fVLXXNLt7+0kWW3uI1dGRtw+YAj+dfj/AOGb6S+tHXzVAAA2528195/sP/EePxL4ZvPDE900l1YHdF5nLFFVF/ma8XOMEp0/aomx9K3VnHeWs0Fyga3mQpIG53KRgivza/aA+Gq/Cfx1LaM7Pb3TtPFuTHyMx29vav0vZQpGK+aP22PhfJ4o8KWHiS1hZ5dNf9+Vyf3W1zzgHpxXg5VXlRq8jKiz49huYjbomBtNeD+PtEuIdQna4gOyRiQwB5HX0r2eGBo0VevG7PUfQEVDcaTDrEIt7naw6ZYV9viaXt6SY2j5nuNJjYYA2mseaFoZGA5FfQHjT4N6joNq12lnJLbscxzRgsmOvavGZtNdbh12bvnKnrxXxtei4Oxmc9Ip4xT4QWTtWhdafs6LgfjVaOxcL7VxKDUgP00/4JYjP7OPiYHp/wAJTP8A+klpX1vHk9a+Sv8Aglf/AMm5+Jf+xpn/APSS0r66QCvkTUbtNG3NMmvIYM7mx9aotqU9ycWsXH98/MKANVmVV3Z2r64qlNrUcfyQxvcSf7Iz+dRJp9xcSK08jZIxtjOFzXUt4Gu7PSEvRGoRu+eaCDmo1u7v/XMIvTZ3qzb2EcD79nz/AN7vVtVHPFPWgBFTp7U/vnvSqKcqj0plXHw1Mv3aZtC9OKNx6UDEfrRQeadtFMgAKfS+WKk2iluA3AqXanpTNopaYD/LX0p1JHlqHyOlAC7RTaBu9amEYoKuQ7c9aTntU+wUpiX0pBcjK0lTbfak8v2pkjakX7wpvB7AfU01phCeaALdRyXEduu6VwF/KqLXly/+rQbf71JHpqTNmZjIf9o5oASTVJJuLOMyH+8QcVzHxK+Ftl8WPC11putwxvdbGNtMBny22kA5I4HNd5HthGEG0egpxfOeevXFb0qns5cwHw38KviZrf7OPjS78K+IreefRvP2lcHMYLABlLEDAAJ/GvtrSdWsPEWlwahptxHd2c6B45I2B6gHGR9a8u/aB+C9v8TtF+3WkMcWs2ylg/TzFAOFbjPWvnH4FfG3X/gX4kPhrxQs0vhqa4MZZ877VtwUH5iABgV786EcVT547mR91BQvQYqSPrUVjdW+pWcN7aSpcWs6LLE6MGDK3TkH3FTIpr56VOUHZmlx7UsfeilWpGWVO3gEgHjGeK8o8dfsz+C/G95c3zxSabfXIy1xZhVO71wRXqgpYzt6dc5z3rroYuph3zQYuZHy/cfsG6SZV+z+Kr/Z/EJVj6e2E61Qvf2D9StbjfpHipUX+9fD/wCJWvrVQOPbpUi4r1459i47SFdHzr4N/Y9NksZ8Ra39sO7IWzA2Ee+4Zr0H4tfD22034M67p2gW0dj5FtJOgiRVDEIc8dORXqO7oe+MfhXn/wAe9en0L4Ua/JAEaWa1kgVnODkxtgj8jThmWIxVaKm+ornwfb2sl1ZQ7uGIBwD6df61l6ho6SeZuUHIxWnoLSRabCpy8i5yT7mnXUbvn+HNfptOXNBE2MXwvpJhW7gaXzotxcoRwMjA/Su/+H/jnUfhXeG5055Do0jbbnT1+6VJ+ZgvqRir3w5+DuueKvA+s+ItKeGY2lw1t9nGS77FUnHHXn9a4m61JrOR7bULeS3uMEPHIhDDPGCCPauZyoYi8LptFHV6tqOleJNTur/Ry0UdwxcRyYDLnnB7VQhurmzDwwzPCrjayqcAj0xXFSay/h+6huLZQUXJ2EDDZx1rprDxLperANKLlLkDBC7QrN/PFdfIrWRJ2nwl8dah4K+Inhp5ZdunXF2tnIpJwqyMOeuAK+8ZJAr/AHsqQCMdMHFfmT4qvF/swhdySRt5kYzjLgcGvu34LfERPHnwl8NamkgmvY7OK2venFwsa71/M18PnWHtNTQ4yPSIPmZmYMwbg4r88vixYSeGfil4g06ZVW3W5a4tNpI3o5PTjnnIr7shGo3YJuZ0hjLfKsGQ2PfI61578Zv2e9N+KmmxzLJ9i1q1BWG8XgvgcK/GcE15mVYmOHq3b0Kep8VXV6tx5ewbfLOW9xVTwpr7+DPi34X12JwEju0hm8w4XYzAHJ/Ot3x58OfE3wvu2h16wYW64Rb2FGaJuOcnHHeuF1i2tdWs5ZPtCuUXdhTjBxwa+8qyp4mi+V3FZn6lzQxyBJFZTHIu5CuMEYBzSLhV4GK8r/Zf+IC/EX4O6XcSZW808Cxn35zuRV5PPfIr1cJX5Viqbp1bMRDUoUelKYxk8Uc+lcJS1EZfapbPjbjv1ptC/JjbxikWaithMjrVO5l8w5FRm+PQDiq/mH1pWEP2j0o2j0pm4+tG4+tPUCM96KRmp3FMBKKYWPrSbj60AJSrSU4UAGKjapCaZ5sUZ+c5PpQAnl0skbRx7iCo/vY4/Oq+oXF1sC2MSPM3TeCU/HBzXnvjT4iaH8PpIz4o1eS8vnw39naTMnAPOdrkcVpCnOo7RVwO4XWIJlIt5Bct3aP5gPY4pjw3d4u1nES/7Ir5l8T/ALbV9omqNH4d8JWP9nRY3Ld2oEr9eRtfBrsfh3+2R4X8WRMniCD/AIR+6TG/zSqoe3HJOK63gMQt4ge2Wemx2j7jGu71AxVsR7RgHim6Tq2n65p8V9p95FqFpKu5ZYnBHOD29j0q3sHpXDKnKDsxXImXd1pNgqbYKd5a+lRYZBtpm2pdtGwUgINho2Gp9o9KNgoArUZp/ln0pPL9qAIqKk2Co6AAim4NOooAbTKlqOTC9KABahftT9+Kh37v4s/WgBMZphwKhvNStrNf3k21uyYyT9Kjjme4+dfkT/aFSKxZaZV71nXmpXEilbJFaRuplBC/41ZW13d/1qwilDwMH1rRSsI8j+LHwLufihbLqT3McWvW8flw7WIjdQCQhJGQMk18h658LfFHhPxNPBrdgdOiU4S9dWELc9Qx4xX6Or8vSkvrK01aAQ3tpb3cWMBJolcYxjoRX1WBzr2MVCS0NoySVj87vCnjDSPhTrj39h4sk1e4eIwvb2TKRyQf6VxHiaXUviNrD3mtaje6latIxhiuZi2xSTgD6Zr748Y/su/DTxpG5uNHfTZXzubSytuefXap/SvJ/E37Eup2N4P+EK8R262bD5Y9ZZ2kVvbanP419HHMMJXfvoD55svCtosEawwqqhdvyisvXtK+3apBawr5t5DlUt1wZDkYx1z+lfUvhT9kPxL9ugPiLV9MGnocypp7SrIx9iy4r3jwb8H/AAr4Jtlhs9OjvZFJYXWoKk02Tjndtz2FddTMsNTjaOo9+p+buofC7x/4H0u58QWvhvW9P8lDNHeIjIUAwS3BzjivObXxtcavfNda5cXV1qcpxJcXgy59ie9foP8Atr/FC38M+D7fwpbSmO91UfOIyQVtmEilgf8AeXFfC0nh+zNuJvKX5Tw7Yzj1NdODqfWF7RKwW7F+HZewDyBuX+8a5TxB4Te1vohGVZpRuEe4Zx+f613Gj6Sy27SWyMVVdxGSfXj+VfUPhn9nWPXf2bdQ1K7t428QXkRv7ScIQ4hMYZEzjOM/zr0q2IpUYpT6gfCLWcloUOCChyPrWl4V+LHij4Y+LE8QeHruSzuI23vAGJilAwdrgdQcCuguNBnvIWDDZcIxQ7htGRwa5m80M7/K2fOvWrqUo1qaXQlo/Rz9nz9srwr8ZtHtLfUriHQPEfCy2twQiSNwMqSTxk4r3vUtOtPE2k3VhP5dzY3cbQyHAKsrAr6+hr8hPBPwf1/xZdSHwu+zVbdPOWJGZWcL/dxjnivXvBf7R3xP+At+mneJLW8aJPlaz1OJ2XaOMhifavj8RlcqcueOjIRF8aPA6fC/4ha3orFkgZ2mtAOP3JYhSBjpxXmTTeZ32n617d8cvi94S+Peg6b4ggimsPE9kgtJoH2hHhO5mx9C1fPEuoNxyK+nwrm6SjU3NHbofY/7K82lfErwpe+C/EcMFzcw7ng87DOYQAoPPfJ5rx/41fsY33hTxLqN1p9u89jM3mRBSTt5PB4615X4Y+IWo/DPxZpXiXTLtleGUebEp+/FuBZcZ5ziv1R8AeMdH+K3gHRtehWK5gvrdJHjZtxViASGHY818zmilQlfoZNo/GLxb8PbnQLuSKaJ42j6xyKQeOv8jXOQ6M23BU5HWv1w+NX7OfhvxwrMkSQ3aggEIFGWz3x9fzr4u8bfsy65o+sMmmQtNbtn5lXd0P0rxKWJpzdhHsH/AAS6v47P9nXxGGBLnxTcEKB1/wBEtK+rVudRmk/dxrFH6vzXyr/wS1gjb9nXxJIUBceKLgAn/r0tK+wY4wvQY+lfHmpRj08f8tW8z61ehhEX3BtHoOlP4XoKk6dKABd0bAjqpyM81v3ni+e60WGwZQgi6vnrWDimlz07fSggevSnLTFanrQA5aetNb5duO9SxqDVASdabtqVreTyvM2sF7Nj5T9KbtpFXDaKftHpRxS5FIkKfTadTAKXHvSU5VouAq5XpTtpPWnqmKdtp7q4yPbU22k2+1OmmitYvOuZI7eDP35G2jA68nijyQhdgopRcWMlp50Go210n96GVW+vQ1hXWtXsrqljbIWJ5NwGAA/CqcWt0BtyFUUs0iqvc9h9azrjXLdX2QnzpP7sfP61WTR5LlvMupmcf3UJCHP4Z/OtG3tY7RdsUaovsKQFExXeofPcEW6+iV0Gmafb26IduSf7xLfzqpU6yFQADwKQE1wqrJhVAFQjtWxZaXFf2DXDygN6ZrEVjQA7mjmhTTqAHZboMD8B3rxj4+fs/wCm/ETw7cXthEsGrId8jDgSDB7Y617RUsLDaQAAD14r0cPiHRndbEtHxV8HfjprHwR1KPwx4ndtR8OeYsRlYMZrPJA5yQMLjkYr7R0nVLDXNPhv9NvIr2zmAaOaJgyEH3B46jivJfjh+z7p3xCs7nUNJSO11d1J2qoCyY9Rjqa+VPA/jvx98CfEDWFu7Q2qy/v9L1AMEKggEqpxgnHBr3J4enjIc1PcR+h1OWvK/hn+0J4d8exR29xJ/ZWq4w1tcMq7jj+HmvU1YbNxJ9iBwa+drUJ0XaSKuTLmhRTSzetNVz0zXNykFtalU1S84+v6VOknqanQC4pr56/ba1htM8HeHYVcqlxqSLtUcN8kn5/jXuuqa1Z6Dps1/qFxHaWcK73kmbaAPTPrXxD8c/iIfiv4wDQXDNotgyvbxjDIZELAPx2Iavdy2hOpVUkhHHWqq20KMButJq8LRwzOFwI1LM38IA7n8jVvSYbdZAJmkGF42joff2rXWzXXtX0TQ7dWP9qXsVvcYPS3Y4kb24av0yU1Tp3Yz6v/AGZfDX/CM/CTTZCoV9VddUdk6N5saHI9BWz40+Bfgj4gNJNrGjwfaiObmABJDyefrya1dLe18JeH9K0W2kL2+n2cdrGAN3CKFBOPoKlW8vLxCpKwxnjIX5q/MK2KqU6zlB2bZR8o/FT9i69skurvwprBuFVS62d98zHHO0EL7frXyk2l6rBdvHcebp97bMVkiyyEOOOnXsa/V+3slhfc0jO3XkmvOPih+zP4U+JzTXux9L1Vl4uLYBQzYJG715Jr3sFnbvaqwsfAdrHc3OnsLre7L0z3r239iH4jSaF8QNX8E6jJH9j1BZL60JON1wWjQAc9wCa4rxd8H/G/gDXbrT7rw/qWoWkcn7q7sbOSVJE/h6DgmuLvob/wX4q0XxBZxXmnX2mzR3RjvoXikIVgxGCBkHBr6DFezxVC8GZ2sfqS8Z34xxj0qRgD2x9KyvCfiSLxZ4Y0bWbf5lv7SO4IGPlLKCRx6ZrT47V+Z1F7OehWpBqum6b4isJLLVrGDULSQENDcRh1PGOhrzjVP2ZfhrqyyKnh6C0Eh3N9mG3NemE04Db04rWnja1Fe6zQ4n4b/CXRPhLHqkOgGeK0vCrtbyEFd4A5AHrgV2XC0jOduM+9N3Vz1K06r5pk2JGptJupN4rAa0G7j60nmH1pzbR0qIsKYxcmkqPcfWjcfWgCXeKN4qGigAzR5h9acqj0o8selK4Ee406kb5eoCj1Y1TutUS124jedv7sYz+BpgXgAejfnUUl2kK4blvaqirdX/UeSvr/ABVZj01IWyxLn3pXAZ5klx9wMn4Cmx6eWk33B8xvy/lV/wAsN/8ArqzawiSTDDIq4LndkB5P8cvihB8KfDoayIuNXuGURRMR8oPfGc/pXxjdG/1vUpdR1W4kvbyZ/MMkrF9ueignsK674peNLnxt8Ttdkv8Ay1t7K4ks4I9x42SMO/fpWTHGoUDPyDocYr9IynAQjD2j3Ao3Gn24UMykjGW6Yre8E/B0/ETXorHSo0S2JH2i78rKoPY468dKg0/w5feJZlsrGMtLJhc4IByRjBI6819wfDjwJZfD/wAKWmm2cQDgbmkIXcSckgkAZwSRXRmOKWHhbqMh+Hvw/wBP+HHhmDRtPDbIgJZJHA+aTaoJwOn3e1dR5a+lAyvQ4py1+aV6vtZcxFiHYKNpqXyun+1056VIIPLjSSb91E38Unyj6g/lWCTKKW2l2+1c9r3xW8C+GUkbUvF2iQGIHfGb+Lev1XdnP+FcLqH7W3wxtdv2fUbrUSPvGzRZFP0O6tlRm9kOx63to2+1cDov7Qnw81oRrF4gtbeV/nVLqZI2wexBPFd1puraZrkZew1OyulA3fuLhHOPfBpuhUirtCE20hWrXk7WII3AE84/IkVEy7m2ruDfnXPZ9QK22osZq1LbSR9d2f8AdwPzxUDLzSsBHSU+oLi4hs13TzrEvq3yj8M0gJCarPIh3ZcLjpnj+dZU3iNppvItbSZ/+nhk/d/nmn/2bLfMWuXye8Z+6PxoAkk1S3ZtsEvnN/s81UEWoXhLTKtsi9FTlm/pV+306G1XbFEsY/2RU6x7sA9KQFGDS4YW3OBM3q3Jq0sIHarTKG6iog1FgI/LC9BTtvtUjqKTikKwwr3FAbb0pxPSmdad+wxd3Sl3bcYJGOetJgelAwaE9bjASFgASSPrTmmitre4upX8u1gjaWSRuiqASxOemMdaiGOMV5b+0p8TbH4cfC3U1u/MM2rwyafDHEPmPmROOT6D0r1MJ79RIVj47/aA8XH4q/GLUriG5jnttNkm0628shU8pJZGXnPJ+c81yT6ILO5t7a+CxwMwd13KWKjqAc9TXntnqDQySCJ+JJGY+3PIqzeahdXkYMk7uR0JPSv1rA040qOiNT660Dwn4f17VPCV/oFnCbKa9tvtmwgpGgdTIHxx0yK+07dLO3s4LC0SF7OKFYFSIqy7ANuOPpX5OfCq+1axM0Frqt5b2Uj5kSOQmPA9fc10Ul74v0PWrq40XxbrFrCXLIq3bdD04PFedjsJUxTjKII6f9oTwP8A8K/+J2q2yK6WV0TdQ7gR98n2614/qOLe8kVsOH6Y7j1rf8Z+JvFnjRkl1/xBNqdzbx7VaTAYqOnb3rzm31M2s0kk43E54L17FByp01GpuDPRfhv4qvfBfxI0LW7eby7a0nDzLnAZMEFSPoTX6Wal4U8J/FXw7by6zotjqlpdQKyvNCrPtI6hu2N1fkhN4qh2yhJFSQ/d7/rX2r+w7+0EvibSbrwZrNwou7Is1u0rAbkwgx19a8LNI3XtKb1MxnxS/wCCdenapHcXfgnxHd6XITvWwmKiPJzxnHTpXxx8XfhL46+DOpG017Sbh4U4N9bRNJE/YncBgdDX61ah4ot7dlS3Zr24JwUt/nIb371k3Frc+KLWW11m1sZrFvuwmIOxHPXcDzya+foZzOn7lT7xn46WevWcflfbSbiMMC3l8uuOuAe/NfVf7D/7QkXhvWr3wNfXgttMvrpn02WbjbkooU5PHQ19La9+xz8KNevvtz+H1tJ3/wBZ9mVUQ59sVh6l+wt8Ov7Q0vVNGa80e+sblZgY5BtcLzg4XPpXRisxpYinytknusOhxtKZLmWS4k3ZO5jjOOK3LPTtOeFfNs4XxwDgduvUVQXbBb28UeWESrEHYnLbRjP860rNvNhXd82OlfG+0UJ3iUonxR/wSvQyfs6+JV7f8JRcf+klpX2Nb2rbdz8CvlL/AIJJ6aLz9nfxKWXj/hKbgd/+fS0r7I1eGK2j2Yx+NcIznD1p6k96FSl24pCuSVDUmTTBQSKq1KtJjFOxQA9v4KljqDcePaplNMC7JqMklqlqxzGn3RgDFQbqZuNKM0AOzTuKSl20gJcUu2lUZ605VNACKv4VIFpQy9SQuOTuOBiuJ8cfG7wV8P2MWra1bQ3QXItw5Zz69BWsKcpu0UB3QSo7y6g0+3NxdTR20K43PI2AOOea+Q/H37ctxeWcieBtJPnJIU+0ajt2n3A/Kvmrxn8VPHvje4kn1vXptknW3tnMcQyc8KDz+Ne3QyqvU1aFc+3vH/7X/hPwu1zZaTBc61qiqQhiVXgDdskOGwO9fM/i742+OfiFqHmXl9NpFntMQtdNmljjKnOSVLHmvH9LujZggtkE5Pv9a7rQ9Yt7pkRtglJxtHVh/ia+swuTUaWstxcxnQ3nivQ02Wni/XbaPn5Irx0HPsDXr/wb/am13wRrkGmeKbuTWtEu5Ar39w7PPBkjJOWxgf1rL0bwLdeJp1hsrWSZz/dXHXBHPT0/OvbfCH7JNlqFtFN4ktY4485e3fJaT/gQatMXg8HTheT1Mbs+hdM1a01zTYL+yuY7uzuEEiSRsGBU++fcVZrI8L+E9L8EaPFo+kQNDYRHKRGQsV6Dj8hxWsxNfnFdRjO0ToQLS03mlWsAFWR0UorEKeozRTAalXHegB+MU5RTc05aAHVJCu2m0oYjpTuBYVirBh1ByDXn/wAUvgnovxW01nliS01VVZo7yNVUng4Dccjmu+DUuTuBycj3rso4mWHleAz4C8a/CvxD8K7wnUbSdEjf91qVnu2cdcvgYOPwrc8FftTeMfA8dvb6jLFrenKON6s0vPvuAOOa+376ztdWtmtr2CO7tWUq0UyhlwevBrw7x9+yL4d8QmS68PztpV5lnFu7HyXJzgAdR17V9RSzDDYpclda9ybEfgj9srwZ4kke31RLzQpx/wAtbqNUiP0O416Ppnxk8CatBHJB4q00b+itcoCfcZNfGnjb4K694Mmm/tLQZ2t0/wCXq3jMsZ98iubj+zWkMNvOI1wMKswVXC/zrV5Ph6yvSkI/QdfG/hp7bzV8Qaf5fAEhuUwT371yni79oLwf4RjOL7+2JGHyjTnSXB9Dhq+MreWK+EEA+SPO5RvG0E+mOv40l/ocejxM6qiqx3Erwc1McgjF3ctCT0Lx98atb+KF8trd+Va6Gjb0hh3K0g54kG4isW1hhXEVqqI0gJEaAAY7jH9K53Q44rpZMlIx/CXNaEEj6DrFndJNFcyQyFtgcEN6jGe9fRYajTw0LIDR0K5XVppjaW0lwsOWkZUIUKOSSfavafgB8LxrWrSeLr20e1jijNrZQHPOGVhIufpim+AfCl/8TLqC+htDoOjRjZPD5WwzkkEkAevvX0VpenQ6TZx2dtGqRRgBFUYCgDAxXj5lj1GHJB6hYF02JZC4Ubm6+9LtYNkcVYDU1itfBTm5SbYyL5lqUfMOp49/Sm/hRzWd+xoSzYmj2Pkp1xnisHUvBfhvXDu1Dw/pt83965tI5D37ke9bbYpAa7KVaUVZMh2K2m6baaNYQ2VjbRWlpCu2OCFQqIPQAdKsrjaDjmmFjjrTd/HFYTl7TURJtFNZ/SmFyaOfSsyrjaKj3P8A5FG5/WmFyXcKiyab5h9aZuPrQFxfMNNoApdtA7iUgzUvlj0pcD0oATaPSjaPSkeVY/eo/tHmfdG2gCWVkgjZ3kVV7MTx9KpPqDOdsEZY/wB5uBUQ0eOSRXlZpivQFiB+XStRVwMAAD0xUAZbWlzdybpZBs/ur/8AXq7b28cWPkGBVnjrgH8KTaDVXAF9e9LSqopdoqQBV46VLCxjOV4NMXPpTlUjtWsJKLuB+d37T3wj1HwH8StR1OBbqTStXna7juFfCrI7EleAB2rzexvNTVlZr2VlH3tzkj9a/UbxT4T0rxpo8ul6xZR3ttIuwJIgJRv7yk9/evnzxb+w7pF2xl8OazcadO3/ACwuz5sR+nfNffYDNaKgozdgRxP7KeuadrXxFltrm4ijntbchbeZwNzsVIKg9xivtBrZ927YxH8OWJr4A8T/ALGHxF0PVILzQLl7mVBlbuymEcqt75PSua8RWnxv+HzIPE9z4n+z/wAM9pcyuq/XB/zzWOMo/XZ3jJW9Rn6N3sqaejNcOsCLyXkYD9a41fj78OLHVBZXXiO0muF+9HbTK+Prg1+f198Qbq8aODU/EPiVmxjZdzTbB9cnFZ8Z0Ncyo8Jkbq38Z/TNYU8jcvjkI+yfjz+1Jd+F762h8DW+mXkUi/vJb2N2wfba1fPerfETx14+mMl54l1C1km+9a2N5IkS9Pugmud0nT9Q8SOI7C0uLt2OVXYVz06bsV3mg/s6/EnXL+0EVsmiWjt+8uZikjKPYBv517FLL8Dho++xmTo/wp8Pmd7jxG6X0rx8mZlcFierFhWFr/hPw34R1VbbT7iO6SYb1FigOB6fKP5V9HeGP2O0jvt/iXXbjWLf/n3y0Y/8davZvDvwt8IeD7WOKz0aCOOFeZJhvIz1OTkgVM8fhqOkIplH566p4L1a4tvP0rwxqGpxlCWmgtWJH4/ga5qz8QeMdBljuNHuL7T7uN9ohSZ1JO7uAcH8a++Pi9+0d4E+F2kzW5ntrjUTG3l2tupOfqQMCvz6/wCE2m1691e5ji+yNcXLTQ46oGOcCurDyjjvdqU7IVj6Z+Ff7eMNjZrpfxD026jvIE8v7VZw5Y7cfeBfr1q34+/bB8Qarj/hANNt4rRfvz6tCwfr/D5b/Svky62yzGWZRLKeTI4yT+Ndt4QaNdKslmZd3mfMi8ErxkY/CieS4dz52Tc6qX9sf4w6TNmZ9IuYtxzFcQzMMcdP3n1r6e+Cv7RmjfFjRc3UkOl65EQktmxA3HA5QFiSMmvBb74L3vjbVLddCsGm01tqyOQquo7nJI7V6N8Pf2Obbwz4ps9futbmhns5UlEEMPl7trbgHO7kHFeRjcHhacLQZnc96fUL64/1MKxr6y53fUCopNJ+2Jm7H2mT0k5T8q2cMPr61GAy9K+EnGzsjQhhhSFdsaLGvooxUqp8uO3pUm0UnIrMBKaFC9BUm0UbRQBFVerLVDIoXpQA6SmNSFieppetADGPSmg0NTakCQGkFRhjTgaAG7qxvGHg3R/iBos2k67ZR3lrKpRSQN8ZIIDKT0IBPNaxNCsVxzXXRqulJSQz5D8e/wDBPlFjubnwbrTOc+YLe/cZJ54BCfSvnnxb8Jtf+H995Or6HfNEpCMYYmwxyRlSQK/USObZgrlSORjim3Qg1Bdt1DHcLnO2VAw/UV9Vh+IJ0lyyQ7n5+/DL4b+MvHEipovhWbSNPkUbLi8jZF4/vYzX0Fo37Htvdaeo8R6xMk7cN/ZM47f76GvoOGeO1AjiAjA4G35f5VVvtYitY/38w+hOT+VFfiCpP4NAueLr+xd8Ohu3ap4gn3dd08RP/ouvAfih/wAE19cl1Brrwf4nhXTmyQmqSt5nJ4+6navtT+0rrUI/9Ci8v/ppIAB+VOt9OlXBvZ95XsjMir+FcMc6rbT1QXPyz8Tfsf8AxW8J6kLa20UeIV/hudPVmU9u4qDwz8Gvi14D8V6R4gbwrq1g1lcpK6Rwn94qurFWx2IBr9aoZEhUqiqo9hR9oAXBCsP9obu2O9dMs6U42aZDOW8B3UWteEdK1f7B/Z99e20c80bIA6ORkhj1Brd5PPJ4xSlx6ADGMAYH5U1m9K+YrVPaT5lsVYaJPLyF4BqFpMtnv/n/AApGfmmMRWF2Ow/zOvP+f8mtfTWHkVgFq29MYfZ6SbtqUeE/8E3dHtvCf7POs28SgZ1+STAPf7NbD+le83k7aleYk+QV8rf8Ev8Axdc+I/gB4knvRuaPxLNGuP8Ar0tj29zX17b6SL51mVtgNAjOuLa2t49gk3y+tUZIynWurutEe32zL+9kH3l2iufuoZp0YxRf6s4b3oJsUKFHJoqRVFIQwNmpAc1S1NpraxuZIF3zLGzIvqwBIFS6S8t1p9pJMvlTSRK0iejYyRQBYqdeTUFTpnFXYCbaPSnCP2p8ag1OsJ9DWfWwFcLTgtTmAfNhgMdMnrVG41W3gdY0Pmyd/L5/DFUo3dkBfjjP+Rmq9zqFtYx7pZ0HOPmPP5VmzNrGrXCxWUtpZqfvea7LKPoPX61o2PgiKBt10s9/Ly2+c7x7AECtPZyFcydTabxlpdzp9vDNBFcRtG0m3YRu7g18keNf2B/GX2u41PSdbstYy/mBb64dpsE9PuelfdEdmsWdigENjA/TGamXfH9z7/8As16GFxDw0rpDPyw1P4X+N/Bl5JY3vg3WGaL5mktbGSWP8GA9q4/UpnWXyrqxvLF2YKPOgMbE9wMkV+wUnlyNtliikJG1wyg5HpXI+KvhH4K8a6bd2134X0lbiaJkW4S0QSozA8q2OCDivrcPncZe7KIcp+VNpo1zdPiMDHqTitPSdMu9J1rT7xAsgt7lJGz3UHmu98SeD5fBXiTVdClAWWyuGjXghnXsTxWf9nIx8vQelfUx5akeaJzn6L+A7jS7/wAIaXf6ctt5NxBHJ5karjO0Z/Gt1jzkcYr5Z/Y8+IrabLe+CNSuSEndrjT3kY8McAx5+gB4r6lkUr0r86zOE4VHdhYjk5qNialYdaiavnbdToQd6ctKU9qRVNSA1amVabtC9KdQA+im5NGTQBPRTFY04GgB24+tL5jetMY4o3CgCZW4x2oYZYt/FnOe9MkkRI97Nge4qnNfEnbCef7zLxVc3WwGpPKksZjnRGhPVZACpri/EXgDwDr8jnUfD9jeXDjBeOBc4+tb8dnPcjN3Iu3+5HkL+tXItPgh+5Eq/QV2U8XUpaQdgZ5DcfsreC9SO+Jb3S1xjbaSL1/Faih/ZB8F/wAer66/+9Oh/mle3R4XjGBnP41IwX0rs/tXE9ZE2PGY/wBkrwUsYQalq4UdP30f/wARW34T/Zy8C+FbtriDT2vpg+UmvSrYH0HFemfL6U5T2HQDFQ8zxD3kFh0aJbwiONQiL0UVGXPrTt4plcEqzk7ssloxUW6k3msHqBLupPMHrUWSaSpsTqSmQUgY460ipnrQcitY2JGUUUVIx3lmpQtNU5XNSr0pEldowOgpm2rTKPSoyoz0p3HcqbTTNtWintTNsa9aLgQbTS5WPO9wAvUtwPzprXKH5FT5v73NVLrTF1IxtLcyqqfejjfaH+tMqxJNrtrG/lj96/8A0yBbn0JzinL50/zhvKX+7jNTQadDbptjhVF9AKn247UDK3kGpFgHpVqip5gIgrL04ozin+YKTK+lSA7aPSjaPSlopARrS02pNoqgFVsGn+YKhpwoAezGTlhn+VMZt3J5P60gpmc81UW+oEolbj5uPTtUkz+dGUkVZUPVZFDD8jUGc07eeK2jUa20Ao3Xh3RdSR0udEsZ0b7we3Q5/Ssv/hV/gvzN/wDwiejh/wC99kTP8q6Hoc0oI9K2jjK0dpMYlrZWWnxqltYwwIvRY0AAqdrgR5bCKD/E3bHX8OKjZa+Z/wBqzxv8QNB86w0bTrq30Boty6vpruJvMO4FWIPyrgj8q6ac54iXLcD1L4iftDeDvhtotzd3GtadeXkOF+ww3kUkm7OCCu8H2xXyv8Vf2vvEvjKY2uhuml6VJCYpv3Xzkng4IY8V4f8AaLWS/kuNaaC6u5GLNJqRBYk5ydx5Jo1aGzksWdJF5XAEb/L+FfaYTKYxSqTldgUY/s0kkrzagJ3kO5zM+STWlaaPbTyeXaSRXM7Z8uK3O9mJ6cDNcs3hye6UMqBQuc/KPm9Ks6Dpeu+D9WstZ0xopryxkEsaS5KNt6ggCvp1D2cLRjoM9NuP2b/iG2peH4H0ppLPWJ40N3Ywyyi0jYjLSZT5cDPJGPevs34Y/sz+GPAmk2nn2A1G+RMSS3I8ze2BuI44yai/ZV+OF38bvBWo3Go6eml6nps32WZIwQj/ACqQwyfevZZcgYB4xivhsfmde7pvQTRTjt4bWFYbaFYI1GAsY28fhTZI2xjPtW3pVrFcRuZSAR6nFULxVjlZUGfQivj51Z1ZasixmPHjtUZjNXGXPWo2U+tY36FFYim1I360w5qbAKRTaTcfWlbNAEbVFLUrVFLSAhyaNxpdoptIBjMaBSNQKdgEpVNLgVEWPrTsAhPSgGo5JEjG4natUZNctg3lRE3EvrHz+dMDRyB1BP0qhdapa26uyO04HGIcMQfp61F5Nze/NNK0a/3YePzzU9nptvp7FreJYmzuJXrn1zSAozPqerSYIjs7fHXnzfxBGKs2uiwQPvkzcP8A3pGGattg9enp2pVagCVfkGAAKG+fqMn1zUe6nbhTuAzB9aTBp2fak3UrjsQ5pJG5pOaafm61BRXdvmNN3U9tuelR07gMra0w/uKw62tN/wBTQwPlP/glG1hH+zZ4hYF2k/4Si4LDHQ/ZLXA/lX3HpciNaqAQyr129a+C/wDglC239nXxIRwf+EpuP/SS0r7WhvVNx+5l8kFcHHQmgDo7y6EULgcu445rkrzUZlcxr8gPXgVJf3E0HkfPv9apz3DXPU4/AUCuVAtSUYpVpEibec96m8tcg45plODEvgNmqATaPSnrVG+1ux00fvZld/8Ankpy35VmrqWs6t/x4Wa2sHeS6Uq3/AQc5ppgdVJcQwgs8qxqv3iWGPrVNvFkE+77HH5wAyZGyEUetUIfDnmFftkrXUi9AwBSvCP2m/jNceGG/wCER8LiG21GWINNcD5fLyCMcHrwfyrsw2HliJ8qAk+Mn7TcHgmRbS2dtW1FGKmCxZAYsg43Z7V8seMv2mPiXrElzILj+yrd2/dx2tuEkAz3br3Hesq2sZzdNc3MjXFwzFmlY5yScnJroFtxcWrwPHHIkwwN46V+gYPJaEI+/uTzHkGofEHxncXjXT+JtYNw3VxeyD+RrpPCfx0+Knh07tN8YalK2Qdl1dSSDjpjmu/+H37MPib4syXk2hpZxWFuwQyXcrIWbAyFwrZqLxv+z74g+Dep28Wtrai2nH7q4SRnQnuu4qOeDXTHB4Ny5Xa5J9Ifs4/tp33iq4g0bx9BDaS7RFFfR5Az0+bLex/OvsDwn4i0nXoI9QtLiG+sJBlZIH3A98Z9eRX5VW+j22/p8394cV1fww+Ofij4J6shs7k6jom/MlhdOxG35R8uDjOB9K83H5JG3NSJ5j9NNQML3BeGIIG61Wwdu08rXJfC/wCLOg/F/wAOxano88fmbAZ7PzFMsRwCdwzwOa7DbXxk6M6MuWW5rzHyv+1x8M0XVrfxhbRBAEEF0wHBPOG/KvnXcvpX6KeOvCkXjbwfqmjyp5jXETIgJxh+x54r89b7Q7nSNQvbG6yJ7WZoHJBzle/A74/WvvMmxKqR5G9TJq5XtdWm8P6nZ6rZSPHc2syzoYztbK9eR619/wDgTxhb+OvB+m63aujpcwK0mOiOQCR1r8+rm2DJtVWB+le8fsc+Ojpt5qXhK9mIhlRrm1LEja2VXYB9Oa1zjB+1pOaQcrPqhu9QtU7qaQQmvzGS5XZm0Q45puRT3FRVmA5qWkbpSZNABk06kUU6Rkjj3bh/wL/GgB60owrZZsLWdJrUcbeXGonb+8vI/EjvSeXdXy4nk+zjttAOaALU17Erbd4P0qHzbmfiFdg/vOM1Pb2MUByEy/qcmr6AbemPpxQBQtdJWOTfK7O3uxx+WavLCoGNvFTUUAL1GD09KXil2ijaKAGlsHipAd1RYp9ADsinK2DUdLQAbqXimrSMcCgB+0UbRTd3vRvNABRUe4+tLuPrQBIshoaT1qPb6CnLGD1oTFdEXmN60qu3rU3kj+7UnkoO1USJGpYVZCgdBUY2x9WqKTVLYFUicyykYKqN2D74pBylryQepxVS6uIbdcvIoPoOTTZpri6xhHjVuhCY/nUUOnRo+4nzj70yyBdQluWHlRbE/vvyD9BUq2Jk/wBYzD8K0lhVVwq4Ht1oUBen+NTzDKYsox0l/SpPKOMbankYr0/lRRcRBzSU+igBlFT7R6U0oPSpAh8selHlj0qfyj6Gk8tqAK24+tG4+tP8selGwelMCOpaTaPSloAZTlptKvWqASmLUiqxYrjcR6Gm7DG+GGT3GelaKLFcFrhfi58QLr4beHotZgtVvreObZPGw+YrjjGPpXdfxcVyPxV8HxeOvAOq6Y7qkssLCE7gP3gB29etdOHpxqTswuch8Ov2nPDXj/xFa6J9kutMvLhF8qS6ZfLkkJA2DB6kk/lXr0hEbY/u8H69s1+W3iD+1vDuoCS0MtnqujXQnHl5QiWNs5x1wTmvuD4A/GdPjt4KtZ1vmtNRs1EV3aIAXYqFBYgZIBJr2MRlsaMFOGwXPXLvXLSz+WSXMv8AcAyfxxVY6lfaqrRLa25smGCtxHvJHsKktdHt7cfJGpf+8Rk1orGOm0YrxVzU3puFzkNa+EHg/wAXRzw6v4ftLov0KxKpH044rzv4lfs4+BvDPw18XahommNBqUen3Do0r+Yq4QkEAjA59q95jU8ckfQ1h/EbTzq3w58X2QP+s0e5A+vltXrYfMKynFOTLufm7pdrHJpNiXkUT+WrSLjB6cnFOkWPBG0envW/4mvrTVtE8LraQJbXVnpsVteMq43yJnJz1PaufX5jg1+p0KnPST7jPoj9hXUfs2v+KdKLH/SF+1bc9cbFJ/UV9asrb8Nz7Yyf0r4o/Yzv1t/jTqEJO3fpEpyegAmir1T9oD9o2XwndDRfC89rdX6nMtxHIsix/eG35T14FfBZjh51cVKEVuS9j2bxN448PeD7fzNc1yz0pWHS4mCH6AdexrznWv2sPhpoFu0s3iGG7AzkwsNp/SvifxXqVz40vpLzXLmbUJ3bc7M5ZAeegPTqelQWWgpqEYgFsJoh/C6AmtaPD3MrzYj7Asv20vhvfFjJJeWMUZ2tNNs2t7gZrtPB/wAefAPj35dI8S2T3G7aIJZVRj7gE1+emqeHwsht7XTJrq4I2+TaxtKx+iqCa5250XWbG8USabe6NO4+R7yB7Zm/3TgH8ia3nw/COilqM/WWMxSqriVJIm6NGwYfmKikXb2r82Phr8fPGPwT1aI3lxNqWkNhprcyyy5HopLDFfd3wf8AjNoHxo8LxalpU8aXGwGe0ZgJITxnK54HNfN4nK6mGTbRNzt2XP1puTUkylelRNnrXhysnYYP96opKlbmopKzAiplPao/ofzpMBjUlQ3F9bW+Q88Ydf4QfmP4Vny3moXkyJaQxwQt1llyWX8MVYGlPcpCmWdV984H61mtqjTcW0DuP+ej8L+HepoNDEbZmka4f/pp0/LpVplHoB9BQBltp81y++4uN3+zGAFq1HbrGMKoX6Cp9vGCMilpWATb680tSbB6UbR6UAQMvejFSbTTWXbSGRc+lPpGFNZj61Ig307io6b5h9aCxWIHQVXOaczc9aSgCKo91PbjpUVADa2NOP7msetjTf8Aj3FAHyB/wSpYp+zj4lI4I8U3H/pJaV9g7iQw7Nya+Pv+CVf/ACbd4m/7Gm4/9JLSvr+mgHMzMRkk0tYvjHUbvR/B+tahZDN7bWskkIZNy7gCRkV4h8If2pLa/wDCn2nx7qWm6dqSSMjhWWILjsQW69a2hSlP4SD6Ip0jJbqWlOxB1J4/EmuZ07xhb+KrNZNAmiu45MFLhHDLgj2z60sXhO71HD6tfNId33LdyiY9GFJ05RdmhX1sXb7xNb2+5LVW1CUdFt8Mfx5qt9l1zW2zJL/Z1t2jiP7z8TW1Z2FvYZ8mCNM9SqjP51bVTWT03GZ9j4dsbFvMMPnXH/PWUhm/OtdRtXA4pVyoxmpEQHtQtAEhjCtkcc4618G/tBSLdfGTWpULJt2q2SePmfjivvpUHp718t/tS/BuS0vJ/GGmLJLFMoF5GoJCEbjv4HA5r6PJqkY1veA+ZlX04H6VsabqENnHsnQOeev6Vzkeqhoxgx7jwfm5B/Oq+oam+co+D9BX6hHVXiYn3t+x+9mfhY0loUd2um83B53EDqfWvT/HXw80b4leHLnSdVtUlL7vLkYDMTkEBlz06mvgX9nX9qhPgjdXOnazZXF5od2/muLVdzowIyVyQORj8q91k/4KJeEpI5TZeGdedwuELRxHDdv4+lfE4rD1/rLnTQzwvxt4Fufh3fatY6jMqy2chWInhpVydrAY5BxXj+vav5ymIHcCMYrt/ip8UdY+MXiq51/VYlsy0YiiSOMIQinKh+eW55/SvMZs/aq+xw3tJUkqi1M3Gx2GoeJPHH7MHjy0u/DupQiQgGW1hmZ7e4U4yHAPHQV9AeBv+CjGryNEPE/hq18liBLJY7t0Y9eWwRXyBeQzTSF3d5n6bnbJ4rX8M2EE128d5EskLLsCt0rirZbTxG61C5+uPgL4j+HfiZpMV/4f1G2v0kRZGiVwzR9Dhlzx1r5d/ay8F/8ACO/EC01exhdLTVISsjL0+0BiQfYkGvAfh3481X4O+IrXWtADCCP5Z7PnZKpILDAPJ619Yaz8XvAH7Qnw1V59Sitr/TSL6SxlmSOeN0GGCjJOPwrxo4Gpl2ITitGVzI+b5LORAsp3YPbNQ6TfXfh3xFpuq6fI0c1ncLKdnBfackfQ4rQ1jXLDUJnOlxyLbZPl+ZjOKXSII75wo4LcCvqay9pTa7k8zP0C8L3kXibw/puqRgiO8t45QPQlQT/Otn+zD/dryr9lrWFk8P3WhO7zS2cvmx7zn93wBj2Fe6NGOwr8ex1N0qzibpnI3NqY+oxVJlxW9qqrliD8ormbrWrG3k2NOrSdkU8muAoss2KgubyC1j3yyqg9CcVltfahqHNrAtvH/wBN/vfUU+HSY/M3z7p2/wCmnzfoaAHPrT3B2WcLt/00cDb/ADot9FnkbddXJP8A0yj+5+takcYjGFUAegFTpHxjtQBWgtIoF2xosa+i8CrSru60vl+1ShMdKAFjAVtw60tO2ijaKAFzS5NJRQA7fS7vemUUAOyKfUVLuPrQAbj60bj60lFADl6UNSc0uPWgBaNtOpdp9KAI9tKKfs9qkjhB6igBu32qVY19KZcTJbrlmUexPP5VT/tRriXy7WMsf77KdtT6E2NJkKLubAH6fnVZrjHQZqt/YT3Db7m5kA/55qxx+VaEdukZGBj8askqeS83U4qXT9NtrF2eKBI3Y5LKOT+NXNoNL1oNCxdXS3Xl8fcqptA6CpSoO7/a60bR6VNwE2mosGrG01DtNIBsymjbUmzd1o8s0AReSfQ0eSfQ1P8AjTNxp2YDNtJsPvT+aTcfWiwD6MUUUgIPLPoaCntTvMb1prOfWgCFutNGfWlkpm41VgHUClX3OfrVdryJG2mRSfaqQHzr+2lF440vw3o+t+ENb1TSY7Rj9rXS5njOOxO3rjnrXkP7OP7XusaV4gGkeOdUm1LS71vKjvr2UvJFJkdSxHHWvtbUrc6xazWs0UbW0oKyiYZBBHOPXrXzJ8Tv2MILeObXfCkCXE2TNLpk8h2nq37vC9elfQ4T2FWPLPcVj6dj1Zr+1iudOb7TFKN6SBcoynGOQec1GNJlumWa8mMj8fux9xMfrXxb8Nf2qNY+GXiNfDniOxvDpaPsltb1D5tuAcZTcRwAK+0/CPjHRvHWkxaro95HcW0ihm2su5M84PPuKzrYWVCfPDYR4N+1R8ExJp8HjHQLPzJ4MJqUEKncyYYmQDHJzwa+Vvhr49u/gL8U7TX9N85vD982Lq2jHylSwznB68D8jX6friRXhmQSQuNrRuMqQeoIr4o/al/Z5Phaa91/SrfzNCvpDJJAi5Fm5DEnAAATpXr4TERqx9jU+Q3Y+zPDmuWXjHQ7LV9MdZYblQwXPTIBwffmruzHavzt/Ze/aOu/hNrUOh65ctN4fvnCEuxYwk7RuAyBjAr9DtMv7XW9Pt76wuI7q1mQOkkbArgjPX8RXkYzCSozv0JuTAEdKivLYaho+pW5HE9pJDj13KRU6qfSpo1244I7ivOp+7JNlo/L7xJENH8UavpBi2raXk1sevVZCKr98iveP2jP2f8AWNN8cX/ijQ7SW/0i9ka5mEal3hlJLPkAdOn614DLqFvJqH2WOK4tZMYK3UZTkemRX6rgMRCpRVmXc6PwN4yn8DeKJNQsXVZ57Z7R9y5OxyDx75ArP1iW2nvPkXFxI7yyvIcs7k55Nc9cLbaVqiC7ukhXcpdVI3FT1x79K7CdtHEaQaXDNJBlZWmuW3yMx7ZHavS9nFu9lcRHZ6Wt1dxpMB5XO4J+nNdJbaDeeKNSt/DXhqNpNQmAeSaMZW3j3KrOxHTBdfWsW1S9uL6Gy023a4v7pljiVUL4OcbuPTIr7U+B3wdtPhj4XW4aJZdcvF8+5mbBKuyoHRTjOzKg4ry8djlg6enxEoz/AIS/s7+Hfh1pttLfQx61rZAeW/u0V5FYgcKcZArzb9tjwpYx6H4e1ZR5FxHd+SCoAXBBwD37V9QyqZH3H72c+lfOv7cAT/hVmmtjMv8AaabfU4Vs18jhcZUrYlNyHc+Lr6BZFZWRWjPbaMVW+HPijVPg/wDEC013Sbx4rMPtvLYsQkqFgSCB1IxWw1q0tnu6msLVLEzxuuMn/wDX/jX3OIpxq0dSLH6haRrEPiLRdO1S1bdbXtuk8ePRgDVhu9eX/sv6k+qfAfw1M5LGMSW6seu1CAK9MkuY1LZONvWvybG0VTqtRLE3H1pjfN0fJ+maoTa0jsVtUNw/sDj86rzWdzeR+XcStCvpbna1eWInudUit22k+Y/dY/mI/KqW3UryXdJ5dvD2UfeNXoNMitFxCuV/vZOfzqwyBuCOKAKMOmWscolMSySD+Jhk/mKuplenH4VN5e3oMfTijyvaruBCFIbPf601k9qsyIF6Co2UUXArbDTNhqzQAPSlzAV6UNSNxSrjvSAc1RyDNPZuaYxoArtmo2qfbuqNlFSAyoalpNo9KCyFqKkO30pmBtNAELVHUrCmMKAI62NL/wCPesG4uoLSMvNMsQHUOwBP09adpvicSQf6LazTopwXIxmgD5Z/4JXzLF+zf4mLZx/wlNx9P+PS06ntX1RqPiywsI/kl+1S52+XD85z7gV8h/8ABMHRJ9Y/Z78RZvZ7e0TxPcb44mADn7Jada+yNL0Oy00b4bWPzsY8zb8350wM3bqutxsssS2tu/HzfeZT2Ir5N/a2/ZXi0+GPxf4XsFKAY1OKPheNxLqoHXnrX2s3QfTFJcW8V9Zz2lwgntp0MckUnIZSMEGu/C4hUZXa0JsfmB8MPjH4q+EN1HLpjNLZKPms7ot5ZXjIHI9vyr7n+Ef7Svhf4pWkUUksej6wFCyQzEIrHjGCW6k8VwPxK/Y7+1tc3HhaWDadzm1uMg85wFIX6V8na18O9c8F661nfWt9ouoK52GZWiyQeoPGa+z+r0MdTvDc57O9z9S1yFDpgq3zIQRhge9X7Wze4s3uNyKqe/JzXwD8Kf2pvFfwxvF0/wATSSa9ox2gyfNJNCoI+7k+metfafw/+JPh/wCJGjpeaDqUd3Cy5aFWUyxHGcMqk+or5PE5fVw8rtaGqkmdLViMfLUKr6/3sVIGPrXkFE9OuIor62kt541mglUq8bjIYHsR3qKNi3WpGYrjFbU5Sg7oDwv4gfsbeC/GUklzpTS+H7xmYlbVY/Lc44yCOPwrxXUP2FfiB58n2S80u4t/4HeYq3XuNvXFfbysVwQTn61d+0SMoUsce3FfRUM8xFFWJ5T4Ib9gT4hXWRNJpuw9dt0A36riui8N/wDBPTxB5cp1LVntXBPl/Y7mNs+m75P5V9sLM/8AeNTiQ5Jz1rp/1gr9l/XzCx8O69+wj4o0nSby6i8RxT+VE0nlzsPm2g+i183N4NuPMk3SrvRip/D/APUa/Vfx3qSab4K1q6lfYsdnMxbpj5SBX5tKvyu8w3Su5cnpySSf5mvqMrzCrjLqqTKxyVv4T/6aD862dB8F22o3jQPeLZz8eQJjtV3JAwT2rTRdvTitG3tUbULUlckTxn8dwr6LnUU3bYzsXPEPw18R+BWSLX9FdLd03x3kO54dp4+92NcBqek2rXxmtkEbsu0mI4z9SDz+Nfqfpmhafq3g/SbfUrOG8t2tIlKTKGycc9a+evi9+xu+uXi3vgOa006WRv39ndHbDnnlNqk+lfOUs5pzbp1dGh8h8c2WtTab8jp5q5zg11nhXxFEqxM82Jl6DAx+VP8AHHwO+IPgNpm1bw7NdxRjInsIpJU47ngVwOjSB7wB45YZ8Z2uu1vp+nSvUWIp1I2pu5LifZnwT8e/8Ix4w06VnQWlw4tZWJHIYjkc19eXXiCKGHCKZJM/wjJx61+WNn4m1GGPbIkyPDiRWGRyOmOa++/gn8SLP4lfDjTtRicPcQRrbXa7uRIAO/XpzXwuc4Sz9pY0izqr6e91LLTMtsh/hQCs6z0O2smyIUkb+8wyfzrUaUt1xj6UxhuNfIWNRFUIvAxTVBqUqKbSAVasx1XxUsbGgCfy/akpd59aTNAD6KRTRz60ALRRTcmgB1FJuo3UALRRS4zQAlLtpaTmgAWnqKaq1K/lxruZ9iercfqaAE2n0qRcHq22s2XXoHZo7f8AfsRhWiIZQfc5pyrc3DZkKqv+zSAnk1K3tyitIshPXYckVXM+pyD93GIF9WGWqzBY28Du0cKIzdSoxV5FC8gc0AZkOnNK26V98n94jmrM9/p3h6ESX19bWMbfKouJFQE+2TV9fvZxz9K+NPj5ZXWq/FK906/u3eLzibS1kb5FViQMD8K9LB4R4qSghn1RY/Ejwlq1xJDD4m0t3V/LMX2uNWDew3c10ohHl71YOCeNoz+J9q/L/XPCFnpupbrWFbe7jlCPIpwyHIDHPavaNB+LHjnw54k8M6RaeJIdR0qe8t7MgSNIdrMF+f8ACvbrZHUpq6kRY+2dtLsFN3H5eeoqXbXy84ODsyhOPSgCl2j0pwVpJAiJk7sD3rPkkA3bUIG6QKuee23Jrn/GHxU8J+AWca1rFvBMq7haCRTM3/Ac+xrwXxN+15e6hqDWXhbw/fpYTFg2oXFsfl54OVchRXZSwVat8KGfTob0H51IuWVgGQsvXDA/yNfCGv8AxB8beII0L+KdQgJbI+zyFMiuUXXPGemXyhfGGtbF7faz/SvcWRYhrYR+i7x7dh2NtPVuuPyqKvizwf8AtIah8M4DNrniO48VT3QGy0ecO0PTO8Eg/rXtfwz/AGqfCvjiT7PqVzbaBdNjyVu5VjEvQYBLHnP864q2WYihvG4rns+2kNOWSOTYyMsiFc7lbOc9CME8VGz9jXjSjKLs0FxSzetJuPrTck0vSoasFwqGQlaWa6itxmSSOMf7TAfrWa2srM37mJpl/vqfkpDLkkgC78Er6npWVea3DZyLGPMnkzt2xoSfrxUzafNfN5lxcEf7EOAtWLfTUgIMaKhAxkdfzq7gUW+1aoNxk+zLjdwvzfSp4dOS1OVAd/U1oBfm3YOc5ppjy2cc0ICEKBwBj6VPGzZJ3MCeDg4pnl+1OVcVcZOOqA8t+M37NPhn4wwG68hdN11fmivYEUGQ84V8jkHNfIs158SP2TfGT7oZBa/c3ASPZTR7hzkbQGwK/Q9SynI4Oai1jQdJ8Wae+n6vYW+pWTDaY5IwT0xxnp+Fe5QxzXuVNUKx4t8Hf2tPC3xGCWWpyR6NrZ4CTMFhkPGNrE9+a9u1HS7TxBpk1pexRXlndIyyKwDIykcj3FfHvxa/YtutFnu9c8A5u7ZiZn02dz5iY5xDsXp255rnPhL+0b4v+GUjaZ4qS7aCNtjWOoKyyp2IXceQOa73h4T/AHmHYjA/aW/ZL1rwLfXWveHLabUtEJaVliO6SHOWI2gDjHHFR/siftTS/DS4Xwl4llkn0aa4JjuZ2ZntmbauDuIAXINfb/gP4weEfitY7NP1C3nmdTvsZ3QSjjkFM89a8C/aM/Yps/EAutf8D2sVlrCqZJtPYDypzlj8oA4fp3xXVHEKpH2OIVmPlPqmx1C01a1iu9Puo7u1mUMkkUisrA9Dmp1kb1r82/gn+0j4m/Z31uTw94lt71tL80JLZ3iHzLfBAJTcRgEDtX3j8N/i14U+LWn/AGvw1qkd0yjdJas6+auQOCoJrxsTg50neK0IO2aQFSpClWzlccHOev51zmvfDPwb4qGNV8M6ZdnOdzWqBvzABrekjdZdh+U7sYIP604MRXFGtOk7xdik7HkHi79kL4UeLrGSG48PW9jM4IFzZqA6Z6+1edaf+wTpmmTGGPx1qq2SsWSNooyQp6DO3nFfUjOem4kVGp2sGHUfl+VdEM0xENpMs4nwP8FfC/w9EM9lp0M9/CgjF9NCglJ4yTgYHQdK7VpiO+frSE4GM4H1pjCsq2Jq1nzVNWAM5r5X/bO119WbQ/D8ADC3la4lSP5irdAT+BNfQPxH8faP8OPDd1quqXkcARdsELOA00h+6oB65wa+E/Eniq98Z+IrjV7md/NnO5VB6L2WvZynCylUUxHLQAx/un5Wsm6X/TjEu1ifT/Cti8V4ZFEg27ufwqtdM15Clvb6UFvJX8qK62HdI3PTn2Jr9BqSSgB9mfs6w3Gl/BfQ7ayjVmdpXLSdBlu1ejfYTLGv2k+bIO3T+VR+BPDKeEPBejaQh3/Z4RlmGSCa1GSvynMZp1nYZDBCtuu2NQi/3R0p7Rn0qYKMdKcyg14gip5WOgxTasU3y19KYDqKQmkyaAGy1E1OZt3U03IoAjoopuTUgRyCkWlPPWjFUAjVG1PpNooAjWo3qbb7UxlqRleinFRzTJCQH2gcdyc0FEbYpudq7mbArLuvEFvH8sRFxL/ciO8/jVSQatqSDcY7aDtx8xoAvajrVppqqZZNzHOVTk1itq2q6n8trAtrF/z3k6/98mtO10a2hYM8fnN6yYJ/lV7avoM+vegDEtfCcaPuvpGv5B08zkD6Cup0u0RLfCLsHopwKo7Ruzzn6mtjS/8AUUAfH3/BKoBv2cfE4IyP+Eon/wDSS0r7DjNfHn/BKn/k3LxN/wBjRP8A+klpX2EvSmgHMD6U5WNNBJqRVFNaCuKpPyjONpyMcVieNPh74d+IViLTXbOOdlGxJ0RfNjPqGIrY3HtRuO4nPJ9K66OInRlzRdhOzPk/4o/sc6hDDJdeEbsahCobNrenEjDBIwQvJ6fnXzzZL46+EfiMXFhPdaDfxnc1pMWjSTHUEZG5TgV+nSybehxWT4i+HPhL4iWk1t4hsBJIY2VLiMASr6YavoqebKrH2ddEcnY+ePhb+21Z6nNDpvjSwXTLyQCJb63wtue2TuY4719RaXqFlrNhHd2N3Fe27jcJoHDDH4fUV8N/Gr9jHX/Daz33hu4Ot6eCX+zMhMqADOOF5ryLwL8X/HXwXuFSxu7mNVkw+lXxkCtjGRtOOwrStldGuuei9RbH6mIvpTmbpXyd8Pf29tM1FoIPFmg3GlOwCvc2y5jVvXlulfQXhv4v+CfFkMcun+JtOlDfwPcKjD65NfP1MFWpfEh8x1ysKsrIKyodU06ZgkOpWkzt0WOZWx09DWgiBtuJIzj72HU/1rk5Jdh3LStUwb7vt96sq71bTtN2/a9Vs7Xd0Es6KT9AT1rwn4pftc6f4VklsPDtpFrF6MrJI5ZNh7H0P/1q3o4apWdkguav7U3xOttD8MP4XtmWbUb4gTRpyY4iDy3pk9PpXx1JIG6+mKTUNe1DWNUvNV1C8a4vbqdpJPMbIXc27aD6L2qG3b7VJjpX6bleF+q0tVqZvUuKorpvCXh+78SeJtN0+0VZJpJ0Yh3A+VWXcB7jNc/cKlra+bJIFK/wnv7CvqT9m34MXlj5fijWS9rI6YtbcDBAIB3HjiuvGYlUKTcgPod9QtNKt44GmAaFQAg+YkDr09aoreaneN+5UWkeMeZIDv8AwxxVq306G1ctsWRj1Y9amZl27e3pX5RUrc8nI0sVo7VmUrdytfqeq3B3qfwPFcl43+CPg7xxoM1odHsdOuGyY7qzgSGVZMHDZA9zXbeYV6UxpTyM1dHFzoO8WxNHxLr37KPxG0G+dLL7JrdjnMcyOxcjP8QIq5+zTH4q+EPxKu9F1rSb+z0fV2O7zYiY0kz8pU4wBwa+zTcSYwHPTHFOkfzlCyKkijoHQHH5ivUq5w8RS5KkQjGxV56VMOabtGelLmvnShx6U2lakqAH07pTaMmgCXfRvpi05gBQBMtWkj3R7iOaoo5qytwwXGeKAG0yn0m0UAJtNG006lAoAAKdSsU7H86oXmuWtiAGDSuwwFj+Y5/CgC9/u/NUU11HD99lT2zz+VZEdzrOoSEm1/syEdGkwzN/hV6301UbfIfOl/vN1oAVrx2O2Ecf3mFV10drybNzJIy/882bj8jWrHCq8BalWPuRn60AMt7aO3ULHGiKDkKFGKnCgLtxxS0VIC+WvpUi0wUtCAk9MV8wftafD3UE8U6R4x05Xa12CC5KqSInUswc+gwTX09H3qHUNHttas7mxvI1ns7hCjxyLkMCMEfrXq4LFfVail0GfnZqBN/M7SFXkJyzKM7m7k0R3Umjtb30Sr5lnIlwvy8ko24dOvNe6fFH9l7WdLvpNQ8IeXqFjK5Z7GZj5iDqdpA5H1rxnWNB1XR28nVNPuNMkT7wmUorcdj6V+l0cRSxdN8rA+8Ph34ws/H3g7RtZtW8w3FuhmCkZSQqCynnqCcV0yrv6Z/LNfnl4G+O3iL4M3Fw2kw2uqaZK2ZLKVyu1+zLg45FelXn7eTeKNLFpoGlHSdYcAebcyqEVv8AZXPI/WvhsVlVb2nuK6Fc+vtSvbXRdPa7vbqO1tlBLSzMFxjvXzT8Rv2ov7Sur3R/CQmWNVKyXxwoC+sbBuoryHxJr3izxl5c3inxFcOoZf8AR42KR7foKzv7PgitzFbAIn95MLn8a9LB5K4u9fYDEuLeS61b7Vc6rqOq3E2WdtRuDO0fU4Utk45NdDYwSW9i2S0MGMEBuPx7CqOn+H5pLrCxXF7OzkRpbxlycnjp2r3T4bfs763rSLceKtljpzsM2EQ3NKBg/MQeCRmvo51MPg1q7DPKfDuj6n4uvBY6HYTX8ynaz7Csaj1LdAK9k0n9kGPV7Pzdd1W8gu26LYusaD25UmvbR/wjHwp8Oo9xLbaNp9spw08gUbQM4APJOK8O8dfts6Fawz2/hW1k1aYAqLt22qrY6geua8OeYYivLlwysu4z5nutFsPDeteLLS4FvPLpd3PaIZFDPJtbGcke1c34osbTWtDDGFbeUSDypIcK0Z9RjpSanb6p4s167vJpGnu9Qne5diw+ZmOTmnnRr+PdDLbum0YcMBgY75r6ynGcqVqyMj3z9l39rDS9K8NyeFPHmsfZLvTpTFa6jesAJYQo25LNyRz+dfUfgv4heGviJaPc+HNYtdYghOJGtpFbaQO+OnUV+bU3he01JkjTTTfz/dEdvb+a/wCG0EivT/gr8J/jR4Z1qPU/B9iuj2twyieDVd0Ue0FcsygZxtHpmvk8yyujbnUtQPu661K2sY2mmkxGvVf4m+nvWFda9qeoHbplt5CuPlmu1+UfgDW/HpRNlA10kclxtG8jld2BnH45qZV2qMV8LUhyytcNzIXSnkRDeymd9uHC/c3ewParywoq7VRUX+6g2j8hVvaFzgYpuwelZXLIQir0FSRxik2n0qdcUMCPyfrUfln0qbcfWkpAQ7RSjAqTaPSjaPSncBu1cZApVUKKYDirCwvt3Y+T1q+ayAFmZcYbGP8AHNcR8TPgr4V+Kumva6tYrBdM29NQt1UTqcHncR05HWu0pSxznOPpxXRRxEqLvBisfBvxK/Zg8f8Awpc6v4TnbVbOFjKsmntIt4gyCMgDnoKk+GX7Zfi3wfeR6d4kaXVoFULJDeI4uwR7s2K+8fOCscfL68cGvPviF8A/AfxOhdtY0O0W9bIF7FGElVvXPf8AGvoKOYU6y5K8fmUeA/EbUvhD+0tpiSXjHwx4i5KXcqxxn1AducjJr5q1/wAEa78GfEDDSfEO/Y2+DUNFumHmLkkB9uOeK+ivGX7GGteH2efw7eQ63Zr80cFwcSAegPfj1rx3xd4T1/w4q6dqmg6lbyyYw3kOYxjPevp8HDDzjaM7rsLlOy+F/wC3l4i8GxR6Z4z0k+ILdFCpeWu43BHT5izgN0NfTnhH9qn4b+KrGOeTX4dGlkAzDqUiRFD37mvzy1Tw6bHHmTRxlvurkZPHYd+K5+4aBd4ltI7p8Yzt2/rUYjJKNbWO4uU/Wy28eeFNSSNrLxRo9x5n3Qt7GC2emBuraWPzod8OX9GX5gffIr8hLXWrKFObWaG448uRZmymPxrd0/44fEzw3DHDofjDVLWBPuo8xkA/OvGqcP1I/wAOVyrH6utbyqu9xtX1bgfzrzz4hfGzw54A0+4llu49S1COJmWxtXVnLDsea+FrX4xfGjxdoY/tPxo8tpIf3scTFX49+34VmRQy/MbiSSeRjktM7Ox/EmtcPkFdO9QLG748+KmtfFnVjfeIZI0sFbMFnAGwq88sCzDPJqrHPBcOGiPQcVRj04yuQyqYT/s/5xVi302d5Rb2UL3dw3SGBSzfkPoa+toYeOFhyiHakommT5hhV29R36A5r3L9nf4E3Wo65beKtZEsdpZPusbaTOGO11ZipHcMPyqx8F/2bNQvNWi1fxhaxwWCKJoLLeQxfcrLvBHQDIxX0/HDFawpDCPLijGFVSQB+FeBmWYRpxcIO7GNkf15+tVmX2qVmPem9a/PqtT2knJiGHgUtKwFMyawARqZupzGmGgZFvPrRuPrSUUCGZFJkUNTMmpAMmkooqgGUUHvTcmgAoo4b7o/rVe6voLVcySqg/X8qALTMBVSa6jt13SkIvqTgVnS6pc33/HlHsH96Rcfzqs2iyXjZvp2uM8kdl9sCkBHeeJA7FLGJ7l3+6wHyiqy6ffahGi31y0AX7qWrFf++j3NbsNpHbptiQIv+zSbdpyBikWUrTT4LPmGGOM/3lUZ/OrLLuznn27VLimUARtUdPpOKVwG1r6X/wAe9ZFa+l/8e9FwPj7/AIJU/wDJufib/saJ/wD0ktK+wVr4+/4JU/8AJuXiX/sabj/0ktK+v1NUgJBUgbmoqcppkC0m6kpKAHgip1facg4PWrWiw2Unm/bO33eSKpNgUdblXLMdy0ZBViPpXD/ED4GeEfidauuqWC2d233LyxjjSZWPQ7iK7EYHarC/MuDyOuK78PiqtB3i7Es+G/iN+xf4o8NtcXWiyJ4k0pfmW3+drrbk8bVTHSvFtR8E6z4duhDqGjajpDucj7RbyW+R7EqK/VSOVtwIOCOhqnq3h3R/E0Ri1nS7PUkIxi5gV/yyOK+lpZxF6Vo3QuU/L2z/ALW0pjJZa1fWoPBMN2y8/XNXYvGHjCFmW08W6oC3X/SnP9a+2vFn7G3w88SSGeCXVdGdjuAsrlViB+jA159J+xHpGkM5tfGN1JP/AALIQ469/kr2IY7LqnRILHzJNHr+uMs2q6jdXzL91p7h2I+nPFNt4nhkAJU4xjccnjp1r6Ysv2MNUuhk+K40t/7rRkN/KtPT/wBg3TFOb7xTeyn/AKZyBf8A2WuqOMwFPWFieVny3JceUpIwAnJ3Afhmtnwj4f1jxTd+VpGk6lqT5wWs7VpUQnpuKg47V9g+Hv2N/AGgTieabUtTlH8FxKrR8eo2c163oPhfQ/CsLwaNpNjpe/AY2cKxbsdM7QMmuXEZ5Qiv3buw5TwX4R/ssy2lxFqfjQWt08R8y3s4WZo/pIrpnNfSEax2sKRRKscartCqMAD0pvmNuDZ5ByKZuPevj8ZmU8W7SegcpO0m7rUbNTF+brS8V4tyyU4ppxQW9aWNQ3Wo1AZxTiatvprjoaga1eL79GoEVITQ3XFNo1AfRRRTAfRRRQAU5mNN2mnBc0AORalXNN6VKuO5xQAtFQy3kFq376VR7VVa9lm2iBcg9WYfyoAvPNFENzuFX3NUZNYy3lwxPK/svH50R6W0rb7iUyf7A6VfjhRY9iqAvoBQBSjtby6lH2p1RO3k9/zq7Bp8UBJjiCknOe9XFXb2z9eaXilcCNiVYY/Xmmk5bPf6VIVB60m0elADlUU6inKAaQDaKk2j0o2j0oAYOtOp/lr6Uu0elACR1ptcW/2GJdn71erVnKAtNDkqRnigCwW8sqVJBUYHNVNS0nSdcjKappdpfp6XECuP1FTbvU5o3Bu1bwquGvUXMcJrP7P3gHXlJn8PWcRPVYIlRT/47XCTfsO/DOSY3AfVraUHKtFJCPLPqMpmvdd248k/hxTmYuSWO4nkk13rMsQlZSZFz571H9jPSLpnRPGniMRYAVZblOMfSKr2m/sdeHrOWCW68UeI73yf+WLXEXlMfceXnFe8/eUcD8qfCoVgR1rp/tfEyVnNmp574h1PwF8AfDM2o3jWFk0MBZYVMSXNyB2QMy7jxXgev/tieK/iFbtbfDzSINEtWTD6p4jUwGMk/wADRuRn8K6j46fse6p8SvE82v6V4lmuppFwun6rMzRRH5uI/l+Uc9q8m1b9lP4v+HbKOO1fR57GH5/stvK+X9gMAdOOfWvVwUcPX9/ETuwPMvF3iTxJ4mZZfGHiK41UQnLK058sn/ZHpXIyeObOz3QwWgZB02IP8avfFTQfEej3VtpGt6Bd6VdzAPG0ija47gNnHY+9c3b/AA4vzEXdxGT1G8Zr7SjGnCNqKA7fS/iZ4ZkuLNDbXun3CRBWeSNfLZz1YHdwK+mfhf8AAnSPGDQT6z4otdRmwtyljpN+khMZwf3qsufbivkRvhffXEMiQ30AjI3Ml4fmHsuBivZ/2NfCvjPSPjRbXFpbySadBD5d/JMzFViYMVAPvsOO1cuYSrUqF4uwH3VoHw78M+G2Q6XoOn2cijBmhtkR/wAwK3pOmMnGMdaklOZCP73WmMor8xrVpSl7zIsQbccdqiMHsatOuKa2a4pW3QJWKpFNqZlqN/yrO6KEZR2FRfPUpppFMBKKj3H1qSgCPcfWjcfWkooAXafStFdQHk4xz9Ko7h60m5RQAiv60+od1G4r0pgPbbUIm4xk4pzEVGFDL0x7k1SdtQLCuVOQxH40XDRXcfk3lvHcJz/rFDfzqhc6pBZsEkkTzWO1Y8nJ9xVaa7uLkYx5S/8Aj3+Fa06kqbvcZyXjX4M/DbxRaiLWtEsLd1+7LbxokgJGODjNeI+If2DfB2vIZPDeu61YSSHCrdTp5I9x+7zX0lDpqsQZ910R0NxhsfTNaMaBeAMcY9q9GjmeIpfauJnw7ef8E8vHNnPmw8R6Lc2/rdXLh/0ix6VV/wCGPfHWl27wfZLO8nUbhJC8rq3oB+6+tfewbC47f5/wp32uVGyrkcg+3H/669SnxBXhvqM+DfDv7O3xC0tLqN9G5m+6rxTlR+Udbx/ZP+JWofMJNAsR6zXEwb8vKr7R+23HaRh7Z4/KmtcPJ95sn3rSXEVdqyQHy74Z/YpuJFRvFviO4UjG5dHucJ2674s+vSvcfC/wx8M+BreKHT9NiM0fBuZkUzN15LAe5rrGlLEknn2qJnLNk9a86tm1etvId0K0mOfxyRmom5pTjvSHmvGlUctWySMqPSkwBSlqbWLuwGUw1I1Rt14pgNpGpaRqAIKKRuDSZNK4DWqOpOKY1CASimtIqqzEgKFzuJwP1rIl8U2pm8i1Vr2fOCsA3Ee55ximBqlgex/CqNxrFrattaVZH/55xnc35Cq8kV5ff69vJHbyzg/jU1no9tbNuCKH/v7fm/PrSAqrcalfzZQR2tt/eUfvPy6VJDo8KT+cwaaX+9IcmtHy06eYaMD0pAMYfj9aZtA6DHepKNooAj5qNlHpU7Co3oKuV6TaKl2j0o2j0oGVtvtTNvtVhlFMVQamwrkG2tfTRiCszaK6TSbBHt84/Wiwz4s/4JV/8m4+Jf8Asarj/wBJLSvr5a+Qf+CVX/JuPiX/ALGm4/8ASS0r6/FUK4u41LCu7rzUe0VdWNVxgYoJIfsr0rQY7Vf3GmMA1HMBQ3H86N2evNPljEb46im+W/ZcmmBIOtTpgVkXWu6fpswjnuY/O7QhxvP4Z61SbXNS1Fv+JfarDF/fvEZGP0HencZ0Ml/BarvmkRB6s2BWe/iRZpDFawyTP90ELwPfNQQ+H/OYPqV095u/gdRsX8q0obSO3jCRxqiAYCgcU+ZlFAWuoXzs1xcKsDdYo0w34ntWhb6fFacsuffNTr8vQfnzQw3dearn7k2LETfLt7elPJ/2j+ZqFSBUvFRzS7hYdz680nfpSbqcjU/mIl4oGKZmlB4qOUBitxTy1N2gdBT9oqgCpI2O/HauT1jx4mkavDYHSdRuWdgvm29sXjGT1LZxiuljdvv5/CgDp7f97HluTSXUQkXDDLVSsdQEYxIeK0G/ffOjfhUgYM1uY+ar10Mlqsn3lz+dZd1ZrH91cVQFWiiigB9FFMkuI4RufCr/ALZx+tAE+w+lKCMdD+H+fpWTJ4j82bybW3mmP/PQJ8v86T7JfX37y6nMQ/55wgfrQBdudVtom2RyNK/9xF3GoV+23wB3rbjurDmpLHTorNswxiNv7w6/nWnHCzsSw3MfTrQBSt9ItoW3f62T1Zif51eVQOwx2qXaQxKg7h2pQo9KBiMu7jH5UDjpTmoX35oEO8ym5b/Ip20elOpAFFDMBQrA0gHU9KZkU/cooAfRRRQA4dKKh8xvWlWQ+tAEtRijcfWmNMkcZldtsajLM3QD1PtTSuA/mlXNcVf/ABw+H+k3KWt34r0tLpiBs+2w+/B+fj/61dravBqlst1Y3UdzauN0ciMGDD6j8KuVOcVdozH7RThio5Zo7cbpG2r6twKqPqitHmACU+qn5fxrMDTDp3+Wqd7rlpYNjfvfONi/N/KqFudTumzc+Wi7v+WQPT8RV610+GOTesShs5zSNCGO+1G8XEcYgiz1fBatC0tzuLTbpMnP7wk/zqZYdq4G4fQ4qbZu611QmkB8eftwW4/4TDwgNi+X5ZJGOQSTk5/OvCdxb6emK9u/bAmTUvirpmnTSgG3sVmHPC5ZwcivGLmGO3l2K4l91r9TymP+zpkt6kkmlzNoP21rdha+ZtE2MBzycDNfV/7Hds15put6y4j8u4WC0Xyxj/VBhgj6PXyiuoXX2FLZpmktUbesMjZUNjGcV9b/ALFFjNB8Kb+ecq/2jVLl1/3fMx/IVjncnHC27lo94aM7snrUs0Yhj8xwEXdgsxwoHrXmfxn/AGgvDnwb0dmuWj1PWX+SLT4JFZ92DhmUOGCnHXFfJPjb4/eNvibG4vL6TRdNlTaLXTJJIhjJPOWbnGOlfC0cvq4p6ID7b1j4h+FtDvPsuoeIdLs5+Dsku0B2+p5qr/wtfwGnlq3jLRQX6b7xAf51+bkekreahsfzrw4xuuHZnx9SaTXtL05rFxdQKFA3EjhsfWvZ/wBXZPaQj9ObHXdE1aONtO1vT9R3/dFvdI5PPoDVho8D5iw9zX5KWVzfaJfCfS7u+s2QZSS2nkT9Qa9d+Hv7X3jb4f6hGdWurrxFpRb94t1ud0XgHaS/Uc9eK5qmQ1aSvF3GfoUQ9L5bVzHwz+K3hb4ueHYdT8PahFdSMitNZrKrSwnuGUEkY9cV1jZHSvmalGdN8skTcrGMelR7j61cSIN1GaZJbqvQfrWQyrRUm0elNkXbSAj8wUeYKXyx6U2Z44V3SbI1/wBpwD/OmAU2SVVXeWwvduw+tZ02qtKZY7OFmlXo0gOz8xSCymulR7uY47xQFdp/MZpgWJtQRmZIGEjdgv8AjVGWG/1Bgs0q20CfejQfMfxrRt7WG1UrEgQHt1qfPy47UAVLXSIraNUA80qMAuxYj8TUxtgakLENuB5ptACeWvdc04Rj0o3UjNQAbvemSNSZFHBoAZuPrSbjSbqMigBnNJzS7qTdQAlJupKbk0AJRSbh60bh60ANk7VG1SOc4qNqAEpGpvzUjZ7tigCF6Zk1FeapaWYxJKGf+6hyfwFUzeXlx80Vv5EX96QEN+VSMvSXEcK7mKge5/yaoSX0svEK8f3qSKxO7MzeYfUirCwjpjApiMe+8OtrIX+0JHkVW3CNHZF/HB5/GtSxsIbGHyooljTGML/jVwccDpRTAi2+1DNu61JtFN2igCLHt+lFLuNJUXAbRTqKYDKa6inUjfdoGQUUUu00DuRNTEzUlAUDpQSR7Rg8Vu6ReeXbkVgSPjoa1NNcmGlcs+O/+CVP/JuPiX/sabj/ANJLSvr4dTXyD/wSp/5Nx8Sf9jTcf+klpX1960yRcmrdvIWkwTkVU4bp8tDXCwHJkVT9c0CNc9Kimk2/dx+NZUmtTSN5cEe8/wB8j5absmn5nkOf7o4/lU2Ag1rxD/Z8Tvb2sl/Mv/LGHG4/mRWVHD4i15EM2dGtXG7y1JW4HsTnFdDa26QfcXafqTVtiWGCSec1YGHp/g2w0+Yy+ULq4/573JEj/ma3YV2rgAY+lIoG70+nFOB29KBofuOCOx7UZoopFBmjNJT9ooAerdal3VHx6UuaBXFyafHUeRUq49KokWn0yjzBQMdTmNR1ItAhnlnGBwPY4p23BzipMCjaKAG1o6fdlZfmNZ1KrFTkcGgDqfMT1qhfY2Z71nfbDGpaR1VV6luB+dU7jxJA0nleaJm/uIMn8wKQEjVHdXUVmoMzrEW6bj/Ss9W1O9b5FW1T/bwW/wAKsWmiwwks4M8h7yNu/nTAi/tK4vji0hMY/vy4x+FSJovnH/SHeQenUfl0rVj+UccD07U4kn/61TzAQRWyQLtjRUHoBU/3utSeWPSjy/aqAVVCjirum3Y0+583y/NT0NVdtLt28DgUAWLm4+0XDykeWG6AUxabuyV/2elODY6UADULS0UAPopy/MeaR1weKAGMKRR81Kc0lAC+YKfkVGFBNPqQJNx9aNx9abRQAUtP2j0pQir94frQA7aPSsjxXotz4i8L6po0F21jLf2zQC5jOCm7jIq/PfRW64kdQf8AZ5P5VSkubu5I8kquP7w/lVR0A/Pb40fAm8+FviBhrdvJPpdxny9TjTduJJyCxAweKpfDz4/eMPhDOtr4e1MX+jghxaahIzxBcdFweOBX6GX3hmz161a21m2j1K3bOYpxuTkY6dK5j/hnP4ZNYz26+HliWUEbhj5SfSvraGOw0qXs6kRWMH4b/G/QvilptpLqepWum6nMdn2XzAqMf9nJz+Fet2uk2+nBVjXcOh984wcV8Z/Fv9kHxB4Tkm1bwZdSXdkrFxbKz/aIxyfkIHXpz1rO+GP7UXjH4azQaX4rxrFnGdjxzRsl7EuQByxGcAGsa+X060OegxH3Hs8xQrcrVmPg1yfgL4meH/iNYrdaPeJITyYSw8xO/K5z3rqj+7kdTu3DpxgfjmvnalKVOXJJalFpZBTlb3qBMen61Iv0pRhJMD4T/aUv3vvjtrYlYHybdYF4A4DE/wBa81aT95jgfhXpX7W2jyaH8bpZd2I9Qs0mV8Egnc+QD06AV5fHtZsl+fwr9YyqS+rKxB0Wh28cmoWouUBidhuHt3r6O034qab8D/gVbRwhZ9TuLiUW9vGMspdyQ7L124NfMNnrkdopMvBC9FIyfbmqX9rL4ot0WeTdGg+XknGOo5rpxuEWKjGL6O5aJL69h8ZalPrGqeZfalcsTLPcndtOSfk7gdKv6Xot14g1Sy02xj3z3LBFXtUGk6Gm5VicFj1z/n2rv/CSv4P1qzv4YGv3XiG3UfOzY+UDof51sqaw9L3EBhXHh2fw/MttPELrU5JxDBZ2rZmmPQhVPXmvefhb+yql9cR6z4pQPBIN8WlyrwnOcOM8nGfaun/Z/wDg7Lpa3PivxFa7tT1Cd7uG2uCZGtfnkACk/wCxs/4EDXuSsFxjt09q+Rx+c1P4VHTzK9T4U+OHwxX4eePpfs1vHHpWpATW8SjAibJDKMcdh+deX31jCylZEWRTwVYZH0r6r/bauFtfDvhebYDK9+I95GT908V8rNlpeRlK9/K8RUxFBcwmR+B9c1X4N+PLTxToUwtbHcIr6zQEJLFuBb5Rweh/Ov0Y8NeJrLxZoNjrFi++1uolkX8R3/MV+eV0qSabNG6qysuMEdMV9H/sX+J3vPDureHZHaUac6yxkdlYDI57DFeXnWDhy+1irMmx9IrNx1o37utRtGF/jCj+92/E1Qu9USE4jxI//PNTuP6V+buNnYZdYrF9+qs12jNkPhfYVDJLPefeHl02OFbcYAyPfmlYBou7qQ4SEA/3n+7+lVo9N8xs3Dtcv6OPlrQEe0YAIHsTS7T6HFMBscawhQiKgXoABipP9YRu5x07UynR/eoAbtFNqbaPSomFABnIpKSkyaBhk0jGm5NN3GgQtFJupaAIqKmwvpSbRQBXop2BTcigBtRbqcxNMoATim7qftFQtNHD/rGVf95sCgCSmFs9j9RWdNrSszR26NM46FR8p49aqNHe3yfv5vKj/ux0AaE2pRRblB82Qf8ALOPk/wCFU1kurxsyfuF9Mc1Pa2SQ5wuKtbQVwRUcwFOLTY4ZfMKhpP7x5NWKl2imbRVAQ7aNtSUjUgI6KdtFG0UcwDaZS5NJRzARUU9lAHFRtUALRTcmnVQDKRulP2imt0plEIWpdo9KZn2p3me1BJFIoXoKhkYr0NSyPVG7vIoFy7bR7/09aAGyyVraWx8g8gfWuQutWlmytlCbg92Y7FFaml2d7eWi/bLkBk4UQrgD/GoGj5a/4JVsF/Zx8SDPzf8ACVXH/pJaV9Zz6ta24wZdz/3R1r45/wCCXkF1dfs9+Io1dY7X/hKJ9zD7+77Ja9PavsWz02GI5MS7vU8mrAqfatRvOI41hT++33vyqSy0hoVxcyyXTd95PHtWoYl9Kew3HJ60CEVQq7QMD0FLRRTAlUhaTcfWkxRQA/JFSxqG681AtPDFehpgXZoRH0OarUvmk9TTcigq4tPzUTMR0pVfNIZPuX1oLDtTKVaRA/il3e9JtFG0VQDxIc9aWmAVPtHpQAtSLTdopwx2P50AOyaMmoZrmK3P7yQR+zVUS+nmbEUJYf3mGFoAuyzxW65mlWIf3mYAfnVCXWlZvLtkM7f3sEj8xUTeHY7qTzbxmuB/zzZhs/KtS1t0t12xL5a/3V4FAGBN4fu9akWS+1KeGEHItrZgqEf7RIzWxY6LZWIAt7aOMD72B/WrvQY7YxU9rA1wxWP5i3WgCNI/l29vyq29m27cACPXGBz71rr4eCxc/M/rXzz44t/EHiyW9ii127061jndI47a4aPjkY4NdNGhKu7RA9vjhE2fKwwUYIU5OfcZoaNFbGMEdQf61+fmqTeL/DutR2Ol+MNRguEPM8ly5DH357V7r4B/acvLnxhD4X1HTvtm+YQwXEeS2TxlizZNerPKa1OPM0K59IUVqaTbW8+7zxnIyvJH+e1UJtvnfIMLXiSi4uzGIqmnOtTW2z/loM06dU/gWp1Aqj60/Apu2po4TKyhedx2gYycn0oANvtRt9quf2fO2NsMh3DI+XoDR/Z88aZeKQj120WfYVyttxSMCam8tkkw6MPwqNvr+lFn0C5HTStSlaQKPSlqt0MRFHPFKq7s1JtC9KiklS3j3OQP944/KkBJ5dDFEXexwvc4OBUENxNdRs0K7gP4z938utUo9JvLpt19Ntf/AJ5wthf1pAMv/EcVvIkMCyXczdPJTco+pp+L/UDlpms4+oMPBNaNvZw2n+piWP8A3RipmYsMHke9AFKHTo4W3OfNP94jn86srGT1/wD1VIx3dacDQAxU29KftCknvnPWiihXQDlk5Bzgj3rzr4pfAHwh8VdPla8sI7LVijKl/bxqsik9MnHNeh7R6U9e5HfmuujialGV4uxNj8//ABF8Avil8H9Ye60XVrt7aJi0F1p07eZjjhgEHoK7n4O/ttXFvcSaL49huLqeCTyv7QhtzvQggESAkdOc19jtHHNuEkayBuoZQa8c+IH7JPgnx/I1/bpPoupvlmkstsYZiOrDbzzXuU8ZQxK5ays+5SPTPCvjrw54yi8zRdatL4lc+XHOhdSeQCATjit8M0b4dCD9eK/PPxp8FfH/AMF9UNzY2OoXFvG29L/SlkOADwTgjnHPpXVfCT9s7XdN1SHTPF8TX2nIR/pjoEmT1B+bHTJonl9/foyuM+ovjB8F9E+MukCC+3Wmo26s1vewqodOOmWB4zXyrqH7D/xGsr6dbC/0u+tl/wBW012yO31Ajr618C/GTwn8Q4d2j6xbyS5INvLKqyD3xmu7ZZlj8wLJgchlyAR9e/UVFLMMTgfcSA+DPD/7KfjLw/qhuvEugpqlsis0Mdi7TkScY3LtGQc1xnjP4TePdJ1ie703wHq89pNIzrHHbtlc+qiv0m+2TcfvnGBjg4py30+3b5jFcYwTmu+PENdq3KgPzg8I/Dzx/qmpW9sfBmr2k0vyh7i1ljjQnjJYr2r6r+Cf7NsPhWX+2vFbR6prTOTFbybZIrYZBGwlQd2QevbFe5tqEmCu4gHnjio/MMn3ufwrnxOeVsTD2ey8hcyJlYRKFQKoHTAFR7zzTR83XmsDxx450f4c+G7zV9du4rOOOF2hikYq0zquQqg4yTjH414aUpzsgPnD9tLxnY3mqeHvDCnfdWtyLyQjoBh1wPxrwRWBj6c1B4t+ISfETxdrWrtDIEmvZzaytjHkFyUHqOG/SiB90fNfqGVUnSo2YEjMGjeM/wAQ5r3j9jLTJ4NW8YXwj22kiRJExGRuAJIP514Ra28l00ewM8jsECL1LdMD15Ir7c+BfgR/Avw9tba4XF3dMbqYHqNwB2nFcec14qjyvcZ1AtJ7qTM8gRV/5ZRnIb8KuWulWdguLe3ji91HP51aWMR8gYP0pWU4r8wlvcRBsPpTVi2dBU1I1ZgN2ijaKKMigCFhg8UR/ep5UHrQFA6UwGbj60xqWimAw9KbTmptAxNopu0U+mUCG0UrCmscUAJuHrTvMFQNSZoAUtio93vQD6nP1qpdahBbgbnUydo1PJoAs1Wu76GyTMsip/vEDP0rMmvdXvJI/sdstrH/AB/agQ2Pb3pYfDcRk3XTfan7CQ5UfT0oAa2tXl6o/s+CP5uN9xnH1GOtPj0ttQw19+/ZuqY+Rfz5rSjtkiVQi7VXgAdBU4GM+tAFOGxit1VY41RV6AVL5I9BUtFAEDKaNtSUjVkAwgUbR6UrdKbk1fMMjYU2n0m0UriI6KVqYxIqQDaPSo6duPrTaAAjPWmOox0p9BGetAFen0jDFJk00A6mN0pdx27sHHtVG91SK14BEr90j+bFO4ydh8+P6Zqneapb2eN7fe6KOo/CqL3N3qEe1BJaJ+TUWekx2kZD753P/LWY7m/Oi4WIptXubzZ9jhMQbr5ykMvXkcY9Kr2+i/vN91d3F63YzEbRW5gL0FMk+bqKLhYqR2qRqAqAD0rX01SsJxkfjWc1a+mKDBUhY+P/APglTGG/Zy8TAjI/4Smf/wBJLSvsLaK+Qv8AglL/AMm4+Jf+xquP/SS0r69rQAPSihulNLGkIdQelNyaevPWgBecU8E/5FPCj0p+1RTAiUCinYptMBd1G6kBFIP4qAHs1IuKNop2welBVy7Y2ovJNm7b70t5Z/ZLvyg+4etNuESzldIH37e/rUfmMzbicn1pEkqqPSl2j0pqtRLNHCu6R1jX1Y4FMB20elS7QpyzACsWbXDM/l2UJnf34H50yKzvrtd93IEX/nnD/iaQ7GlLqsMbeWhEs2furz+oqtJ/aV82cR20ft96ptP02HTlxFGFHtmtEMQMUCMy20WK3k3v5k7esjZrTXG3bgbfTFLRTAKfH1pFFSRqPSgdhdo9K6DS9LaH98w/CsLiuvtrz7RCvljcD/eGKTEWYjt6V89+PtHfRfEV9CQ32eaZpI1/vbiTx9K+gA2O9cx8RPBv/CUaKscUq297B80M+AeQDhWz2ORXoYGqqdT3irHxF8WPDv8Awj+vLIHOZ8uAB371xXhm8Oi+OfDerGRY4YdQh81jwNpbk/zrV+LC6/pHiqePW7SW1SN8RyOjLGQM8hv6Vxd5NJe6a0UCebK6kqFGWPXGMV+nQaq4e0dzI/TCOTdFFLHNlHjV42UYGCAaYpOK+VPgN+1Zp+m6RH4X+IEzaNf2CrFDdTqQjoAAuST97rXpevftdfDfQ9ggvZdYkf7qWUZkXP8AtFScCvz3E4Cuqjsrj5j2SNiCxYAgds81m6/400HwmrvrWsWdiqj7sk6qxxnPFfH/AI2/ax8a+MUks/D9kvhu0U5F/HIfOIHTKnIBI5ry+78LzeKL+LUNf1291u6nJEjyZAGTk4FbYfJ6tV3noHMfUvij9sPQdOunh0HS7nW0XgXXCxEgEZB3cjNeTeIv2svid4ikmj0tNH0aAjCkQOZFBPc7uuPSuCbS4NPszuUJbwrhdwA/Wun8L+C9c8USwWui6HJeo2P9LZGjtxkdN+MfrX09LKcLRV6jHcwJPFnxKvseb491ndIf+Wd9Ki59uc4qhq3jD4meD5VkXx/q0xJ3FX1CSXAHUHBr6X8Mfsl3lxCsniDWY7J3b/VWSpJgehJHFegWf7Lfw207TJ0bw9FqM4hYm5uAfMPB54asKlTLIvlUbvyDlZ8hWf7Z3xR8MqjSzWOsRquTDNayFnx6Hd1r2L4d/wDBQLw1rV3Z2PirR7rQruZhG06LmFCeMnLZx1r5wuJLOyutQso4lMMF/cRxkgswVXKj9K5XXrG3vreeGaCJ967R8gz+fXvXpTyqhXpqUVYR+sGn3dprFlDe2N1FeWsoDrLAwdSp6YIqG51O1t2Khw8q9UVs1+WHwt174neC72w0bQfG+oW1hd3CwR28kzkJuIGOnAwP0r7Q+AHx10nUL+58G+LjHYeK7GdoBfyYWK8xhcqSeSTk9K+LxeU1cP79tB3PdnvLudsQbYk/vSj/AAos9HEb755XlPq7Bl/WtZlRdiL8yMMhh/OmrEO2fpnivAknF2YxscPl52fLn0p23HepKQsakBtFFFABRRRQAoalptANADqWkooAeKXccYqFmI6UnmN60eYFyaTzojHJ+8jIwUblSPTFeNfE79lfwt8TI5pkjOkX7DiSzAVCT/eUrXrbSe9WraUR4ycYGOtduHxlWg7xYz4N8dfsa+NfAtq19oN7caisYIzp8zpIRxngAc1x/h39oj4v/C26hsrm8uri3jPliDXopWzjggMfUj9K/Sz7UB3rmPEngnw540hSLXNGs9TRG3KJ4gWU+o9K9unmlOt7uJhfzA+ZPCf/AAUC0wxLb+JvC97FeKuXm03DQk9wNzZ/SvUdE/bE+FGuafDdf8JANOeQ4+z3mEkU9ORnpXOeKP2HfBeq+bcaPeX2kTSMXWFcNGue2COleEeNf2HPGulmWXTrDR9XgAOyRHKNtHYjHWutUMDiP4creoH2/YfErwrq0cctp4i0qaKX5lZbyM/hjdSXnxM8J6XvW68SabGy4zuukJA7nANfmhrXwZ8ceGUMZ8D6yHQ4WWztXdB7mqVxJD9is4NY0C+0m7jzma6gaLzM9Mlx2rVZNTcrqoRY+3viV+2Z4T8L2M0OgH+3NT5EZhAeLPOM4YcdK+RviJ8RvEXxd1oajrt3IqEALZ27yJBHjOPlLHnk1zkNrp8Un+iyQr/wJQa2rLT7V5EWFluH44h5avcwuV0aGrd2UVdMs8EFI9igYFbkayxwPIsbSRx/eKKWxj6c/pXSeDfh14h8ZXf2TTdEvAMf8fE0TJDkf7RAr6Q+Hn7MdtYzW974ika5nVfm09QDA3Tr3Jr0MRjsPhoay+4ZwH7N3wn1DXNWg8RazYNaaNCvm2qz5V3lDQvHIBj7pBbn2r64yI1Cp8oAwNvGKbbQ2+nW0dvaxJbwRqFSOIBQoHQDHam1+c4/GyxMnJgL14pGAphYjvSbz614TYhtNNK1NpARs5B4NN3H1ok+9TaoCaio9x9akqQIaKVVJpGBWqAjakpdppu6gBaZTmNNagBjE0xs0TTRr95wg9W4H51Rk1QTnbaESH+8elAFuTMcO9yqr/eJwP1rJvfEVrayeUm64mzt2Qjd+JNV5NHutSb/AE+8kMX/ADwDHFXrHTbexj2QRLGuMcUAUJP7QvpdzyLBD/dC/NVyHS4IGzhnPrI2TV7acY7UYP0oAjUH1/GlVQvTpUmKTaKAG0u2lYU5aAI9optSso9KZtFAEVI1S7R6U3aKjlYELUlElNyaVirCUUhNN3e9FiRGqN6Vnpm4mkAlFFFABUTO3rSyXEcKZd1X3Y8VlNq0szGOC2fcv8cqlVNAGk7BfvMF/wB6s+bWI/8AVwoZZe+Og/Gov7Je4O6ednH9zp/Kri26pwqhR6AYFUBltHqGoNi4mEMX/POAYP5mrcFlHbjEaBPcdfzq3tFBqR2IWj3daaFC9Bip6btHpQUV2pj9KlYUxlFAEDVsaX/x71jvWxpf/HvQB8h/8Epf+TcfEv8A2NVx/wCklpX17Xx//wAEp2P/AAzf4m/7Gm4/9JLSvr4t71p0JLcOny3VvJLGMrF196qDmpor6eKBolfbG3UYHNRLRYQu0UtJkUtAE4ak3H1qLcfWnbjTAdk01qMmkoAFpV71LHavLny1L7euBTTGyt0/CkAtPpjMO3HuKpzazaxtsjlEr/3U5NMDS3+jZ+tV7jU7ax/10sa+zHBP0FZym91Jski0T82+vpU1n4dtLdt7xtPL/enbeR9M9KQEX9sXt+dljZtF/wBNLlSo/CrMHh1Jk826uZJ3f7sbsCg/z71rqg247elNVivSgBkEKwJsjUIvoBxVg/N15pMClWmVcl6U4KKRVqwsaLGHY8t0FBJFtFNp9JtFADlFKpNNpwoKuO3Vr6bq4tgA/wA2KyOKcuF6Cgk3hr0X/POorvVnusZyv0/z7VlBq6LT7ezkt8niT6k0vMq5knwnonilGh1jTLa/tWXan2qIPg+vPevHPF37EPgvWtUuLzR9Z1/w9cOdwTTbtFjjPOMBoyR+dfR1rAjW8auFKbsoVq21r8uxk8pH/iVjxXfRx1ei7xkxHwdr/wDwTp1XxBLcu3ieTUXBys+oXLF5P9/EVc8P2C/iJoUPlaXq3hl19JJp934Yjr9B2WWzhmVBuiTJVu+O2awGZy2d3Pr3r1I5xXTu2RY+HLX9lP4vRIqzXPhoqowQpuef/Iddj4e/ZH8cXDRrrmr6HbWwxk2M0wmHrtLJivrNZJfm+dvm680gUbwAQQO7AYraWeV5K2wWPL/CP7NvhDwzN51zLe67J08vVXSeP8BsFem6fp9ppsYttMsYbWFTtEcCBV+vArO8SeMtJ8H6fPfarewwW0KF3kbLBQBkk7QeOtfHXxr/AG4dW1SG50nwHY/YoslH1WUlXxjGUwQRzXMqlfGSSu2M+rvHXxa8JfDOzmm8Qa7ZR3UMXmjTY7qL7VMBzhUZhk8Gvkn40ftka18QLWTTPBcc/hbSZE2y3l5G0V71bdseORlwRjqM8V8x33iC+1m8a81nUbzXNUk+/c387TOOvClieOTUcupPI2zPy/3e1fT4PJknz1XqM131K1slCQzzzljvkkl+YtITkk/jTbfVFaYZGcc8iqlnYteEkqSTzVpdBaGTkbfwr6le6uVIVj0r4a+G9S8a+J7a30C0+03iOJh5rHy4gD99iAcAZFfTui/saaLq15HrHi2+vDq3mJMU0m9KQbwQeA0eRz718ceDvF3ifwX4istT8LRzT6hDLsaBAQsoyDsJzjHHSv0n+EvibXvF3gWw1LxPpR0XVJ2ZZLIggDGMEZPf3r5TOq9SC5U9BHTRwrawxW8QykKBAXO5iAMAHFSKDStwSR1PU+tCsa/PaknJ3ZQh60lOam1mA2nLG0nQUBRnmtxYYkTKjBoAxGjI/hpDE23d2rVaNfaoJ0Cw4A4oAz6KUrSUAFOBqHcfWjcfWgBd2etNZiMYpQtJsJ7GgBSAetSo27rSME7nb+NRTXccO3b8+fQ0AWfMHrVa51Cy09d9zdRwJjG+RwOf8agNw0n3MrVSbTbe8bdPEs5znEvzDP0NVrYCZtfnulAsoRISMedLymB+NSWwuZMtcThiTuMaghM1ajRIlVUUIq9FXgU79frzVcyAljuJFUje20jBGcg/hWdq3hfw94hUJqmh6fqCDotzbJIP1FaCqFp/HpWka84bMZzcfwm8Bo2R4K0LP/YPiP8A7LViD4e+DbVg0HhPRIW/vR2Uan9Frc4pc1f1qt/MwGW8NrYQeRZ20dtDnOyFAo/SnMxYYJ/xozjoKSsZVJT1kwAqKTb70nnD1pnmt61ld9RDKKj8z6U3zqLAO3H1pMmiigBCoPWk2j0p1FACeWPSlpGYhaZuPrQAqsMYpruKb9Ka33dwHHuaYCc0ziq91q9rZjDyhpeyKeuaoNdXlxyjLbp6smfxpgac11HH1ZR9TWdeX9w/Fl5Z/wBqfO39Oc06Ox2/6z5/rU6xiP7oxQBnHT5bnm5k3/8ATPqlWo7ZIRhE2j2qcJjoKXaaAGFM9RShSKdtNG00AN20bakooAiop20UjUANanLSUUAK1R09jTKACkbFM3H1prSH1oHYjkplKxpmTUsoRjTDThz1pGZSucAD3NBNiPYaTbjvmo5dRhhbYP3zei/49KpXC6hfScL9nj9WFTYdizcahbWp/ezKp7L3P0qh/ad5eZ+z2rWyrw32pSGPuuOoq1a6bDb/AHl80+sign86sN2POR05p8rCxjw6P84aaSSct/z2fIH4VoiJQoUKNo6CpiOuOM0m0UrCsNKgcUNStUbMaoQ2mtTqbUFiUUVHuPrQAxqa1SYprCgCvIta2l/8e9Zrc5zWnpoxCcetAHyB/wAEqMf8M3+JR3PiqfHP/TpaV9dV8t/8ExdPt7f4B+I0slYQf8JNOf3nLZNnaV9WXkAgkwDkVZJEvSilWmZNWINxp2TTTSbjWYDg/vT1aodw7CnKfTmgZPRVPUNWtNLjDXVxHCW4XcwGT7Cs6TVr2/LLpqIpPRrpCq/hV3Eel+G9QsLGFvPZYmbq0hH69q5PxJ4ms7rVJYtMjadlO0Mq5T65rFtfDchuDNf6hcXYf/l3ZgYB9BjJ/Gtu1t4rVVWGNYlXgKgAFIDAOlarqT77q9FtBnIggUZx7sRWxZ6PbWK7YIlPueT+Zq4qjoeRjGDT446AERAq4UYHtU38We9N2t6U7B9KAF3H1p+0elRLTlc+tMCSpreLzZMdqhq/psLSTAp90feoAuQ6XukwaS6tym0AY29K0TMkJz3psyiaPPU0gOfyaMmpJlVXwBioqYFhYxt3dqjyalEg8jb3qGgCRactMAqRRSvrYB6/WtXQwy3CsRuAGW/wrM24ZR039Mcn8BVyHWo9FG6aeGBM5czuqKePVsVfK5bAdda/vIUVh5cynO0CrssjTKVCE5H3s/0r5x+In7a3gD4Y3LwzXE2vankgxaL5NwF68HMq/wCRXEw/tial4s8G6tqdhPpmkCNvNt1mSWOcDOAGAlPJ/Lmuunga9TaOgrn181ubqT5leJUABzwGA/nWDqtqLOb5cKvfdnA/Gvy6T9sn4zw6st23iaKTDcW0xkMeP90MDj8a7TQ/2+PHK30TeKNO0rUtOVh5i2CzpIRkZxvlIzjNej/Y2LSvYLn3vca5FCuIo/tEnoCT/IVXtodQ1OPFw62q/wB2Jc/zry74Q/tUfDL4pTR21pdroOqOQkdnqLRwtKf9n5278Yr3fytrB8FUYkq3Yjtj2rzKmHqUXaasFzn/APhDdImt5Le9sIr+GVSk0c2WDgjBBBPvXzl+0t8EfhX8OPAVzq+n6Aljf3JeCFYZ2DKzqWBAJIAB9q+m/E3iPTfB+g32q6ncx2trbqXZpHCBuOBkn2r84fi18WtZ+M2uNf3+6DTA5FpaBSqoo6MQW6nNe5k+HqzqqUXoh3PKo/DaHGXHH3Timx6Tb7s75M/SugkjWPGB0+9zUsaDbuwMf3cV+nxWlmA3QbWWOTiJ2ixndtr1n4P/AAfu/i9rSkRFdKhdluLhc43DBC8HPIPbmsnwH4FvPFU1jp1u0n2u8uUiEagsYoCQDcMB/CCa+9/hn8OdO+F/hG00jToVRo0BmlUYMkmBucn1ODXz+aY9YSCjF+8xjvBngHQfh7pq2WjWQtoh1+Zm+b15JroJLrf1Ge3SnMVyx5JJycnPNRMw54r82q4iVWXNJ3AGpFoYmot7L3rg3ETGmUnmH1phl56/pQBYt9rPgjIrRVsx1joxVsg4NaMMnyZbkUARZdpcEHZVuQBlwelNVh60yaYL0oArzQ7TwahMdSs4brRuFAFUqOeKNo9KJJoY1y0ir9TVaa8K/wCr2v7/AMNAFpmRepwPUnH5k8VmXHiC3t22IJLh/wDpku4fiR0qJ7aW6O6eaRx/cDYX9avW9pHAu1I1Qeijg/WmBWYzXXzO3lj+7/F+NOitwuMDp71ba3LHJbn6Uoj29BQBEqn0p/l+1TKi8cU7avpSuAypNo9KNq0tADt5o3U2igB2RRuqHcfWjcfWgBfMb1o8xvWmbh603LUAGG/yaTcfWly1N2mgBPL9qPLHpT6KAG0UrUIp70AJRTmUCoLm8isoy8rKNoyRnkn0FAElQuwXqfx/hrPTWjcf6qF/+BrgUx/Ok+8c+3GKYFm6uisb+Um6TbhPTPvWLYx65qQLX32e1jZjshgYs2B0yfetSOMj881bUnaF7AcUwKFvYx2+cIM+p5P61L5QxjHFWDzRtFAEOwijYasbRRtFAFfaaNpqQ0UAR7TRtNSVHuPrQA2ipNo9KjoAbTWp1NHPWgBKKCRTN3vQA5qbRu96RmoAgL1HuNObHNV5riO2XM0ix+5OBSLJGY5qJpgoLHAC9c8frVOTUmcsIYy+ejnp+FNSzN2yvdOzAfwL9z8jSADrKb/LgjNy/qPu/nULWt1fn9/J5cf92PitSOCOJNqIFX0Ap+B1xn60AU7WzitV2omF9Mk1IYx/k1I1JVARlS3Xmh1FOZiOlRFiepoASkalprVLAGqJqkY1BuNK5ArU2hmpuTUlXHVFT8mk2igLjaa1OqNs0BcRu9aen/6k/WslmNaunf6k0Bc+Yf8AgmBIT8BfFIBUoviicK6HriztM19Tz2pK+aSCPTNfJX/BKO8F5+zv4pZgsRPim56dObS0r6raQsuCcj61pEQLUdLuI70y4ure1j3zzJEP70h2j9apCHtUbTJ2P51lf8JNDdN5dmrXDf3lG4fmKQafdXsnmT3Bji/uRKB+eRSsBYu9atbXC+Zvkb7qDrVLzta1L5bcx2Kf35EDsfpzitKDS7eDcUiXJ7nk/ma0I49vApaAY1n4XtkWN7jfeyqdwady+D7Z6VvRx7QAFAA6ACki+VgR1qzdlI7srGMR+lAAqjb0H5UmMUm/FN8ymBIuKlSSqysaeOKANAXCUjzDtVHJpxYmgB9SIopiipI6QD1rZ0ORUBAGGbrWMtWLWZoWyh2mmBvzW4k6CmzMbeLA4pI74LHljzVa9vhIMZ4pAZbOW5J5oXmm09fWmMdilWkpy0CHLT1YRq7yFUVVydzYAx1NR+dHGPnKgf3mbAH1ryX9pD4mP4T+GOoPo8yy6hdZtkZWDKhZT1wevSt6FJ1ppMDh/jR+1wvh7UptB8FmK6ulUGXVJE3RxnJBC8A5FfJXxA+KnjnxNqHm6n4w1KZZP+WNrcyRRjtyM4qS2tprq3a4mYF2JJK8Ek9TmuM1EtJqD7xx71+kYLK6VOmm1e4rlWzi+z85JY9XYlmP1J5NT/aiqlQ3ynqOx+tRTzCL3ql5u6TBNfSxpQilFIRoKyzHJ5PSr1vCJMBl3AdAelZtkG8zGeK6TR9JudSk8q0hknfqcLnA9TitqcY7JGepU/smD7RDcxg293C/mRzQjayN6g9q+l/2fP2wtT+HMw0fxve3mq6RKFEN48hd4CcDBJONvX3rwy80GXTbiNZ5Y9zrn5ZN4B9DjvWTqAQgKVUgeozXJi8vpYhcs0CufTf7SP7SFp8TooNE8Pz/AGjQFKvOzjImbduBHsAP1rwr+0CibVVQuMYB6D0rmNIuNo+zNjbH7DmtZWDdaMLg6WFjy00WWDMWzk/e61u+HNOhuLgyXb7oFOCq/fZuMKuO5rEtbc3DbUXzD7e/SvoD9mX4bS+KviBbQ3tlImmWCi8kkZcjzQRhT6U8TWjh6TqT6Gh7r+y/8Ih4R0W48Q6tbKNYvtwt3bPy2kixyJGR2IIb8q9ybPqfzpdwiVVQBQoCgDsB0FRsxNfkGNxEsRUc5MCJ2HrUZYd6ewNRMprzVqIkao3XvRJJ5EJaYpCFPzPI4UAY6kk4rjfFXxp8EeCVRtY8Q6eiu/lbYbuJ2BHUkbq2jRnLWOornWFj601iTXMab8XPAeuMosfGWhStJjYn9pQg/q/WuotZre+h8+2uLe5gP3ZLeZZUPsCp96Uqco/EhjkYnNT+c23bu4qFRgmkaZFXcTge4wPzNYjJFmb1oa5bHzDI9TxWXNrMKsFj3Su3QRKW2/U1A8d3qPEknkxf3FA3UCL93q8VouC67+ynkn8qqrqV9df6q38hfWQAsfpz0pbTSLe05QeYfVuTWhGKAM6PT/NfdK25vcVeWADtVjaPSigCJYtq4C8emKkXNSBaNvtQA2iiigAopjMRQrE0APooooAMml3VFuPrRuPrQA2iiigA2iiiigAopdpoK0AJRSsw7gL+NVJdSijbaDvb/Z6UAXmVOxz9aryX8MIwZFD/AN2s+SW4vPmYrGvZV71GLWPOdnPvTsASalczcRjyh/eYA01dP3SB5f3jA7hn1+lWhGB0GKkClelMCHy1/uil2j0p+0UbRQA3ZjtT1FO2ijaKAIqKVqXaKAFooooAibrRTsU2gAqGpDmmNQAeYfWo99Jk0lAD6atIrGigBrdaZStIPQj6VQutWtrcYEnmv/dTDH9KALpaqV5qlvbr80iq393v+FUftF/ecxOsEf8AtL835U2HR4FbMo8xv7zEk0gI31K+vpMWsHkRf35afDpMci7rnM7/AO2SR+XStJLb3/CnhdvSgsiEIXAC4C9AOgp7R7uozU20VGxoAYVprVIxqFmNIBGpKKKoCJmNJtFSbB6Uu0elAEFNansKRh1pMCKmlR6U496ibdWZIyiiigQUUUoFADRtqNqV2HaoZrhFXd90d/b60ANdq1dPkXyePmrnLjU7eH+PzD/dXk/pWlp8lxcQ7owsa+jdaAPlH/glWf8AjG7xN/2NVx/6SWlfWFxqlvZ/6+RUPoeK+O/+CXbX7fAHxBFasqwnxROX3AZ/49LWvruPQbdn82dGnf8AuyHd/OtIjZVXXrnUJCtlbvGq9ZLlMK30xUUfhQ38hl1G7mu436WjlXiX6Ajn8a6CONYYwiDao6AcVLHGFwccjoepqhENtp9tYptt4I4E/uxqFH6VOoGelOooATb6VJGaj5p1IADFelKzFmyT83rTaTmkBOpz1p+BUKsaNzDvRcCRakqFSakyaYDqlWMmoq0o419KAIPJYLu28UqrWvJJF9n2heay9poARaevFJilWgCYzMwwW4qPJopCaAFpy1UvNTttPj33DrEO249fwrDl17U9Sl8nS7Qwj/n6ugPL/AZzQB0st3BaqzTyBIx0ZjjNYlx4wDSNFp1q15KvR2UiH8WB/lTLfw3NdzRzapcyTSx/ejViI/wU9fxrft7eOGNEjjWNV6BQBSAwoNN1XVzHJqskNu38UNkx2H65Ga5j41/C2HxV8PZ4LCL99ZubhIY1IabCnjp1r0tY1UcDGfzqZOMY7c11UKipVFNgfmrNCsayWjwtaTwsVeOQbXBHqD071xmuaSkszyI3Pqtfo347/Z78HfEW9a+vbOWzv24a5tpPLDf7wA5rxzWv2AZ7q+J0zxakFgf4bkOz/mMV+jYXOcNyJSYrHxE9iwfa2R7d60dN8FXF5ubg8DjJ6n3xivsq1/4J5o7BNR8T29xC3/PISI4/HNdFpf8AwT98M2MCIfEOo7icuqXb7Sfxr0I5xglvIR8Wx/D3VIVTFriWT5UG7lj2xxjJr2rRf2X/ABNo/wAMdR8XahqL6PNbw+bDbxyALKoGcP8ALnn/AHq+uPA/7LPg/wAGX1jqXm315e2Z3R/ar15Ic+u05H6Vn/tjeKrfQfg3cW8LrFc3UyQxx84ZecnjtxXnTzlVK0IYdeorHwJdunlpgc8Mx9Tjr+dczeS5kxnitWW6Dw4Jy2KxrhBI2R/OvsdZK4itY3TCYeq9K6KDc0eScmsKxs3N5t7fSuw0PSZb6ZR5TvEhDSOo+ULnoe+alJ7FWOk8JW8Fnai4x5jybWGf4QOcV9hfsg263Gl+Kr04ImuYkX2whz9eTXylFYxWcaohC7egr6g/Yq1SG58O+LLUMxkg1EceoKLg5+oNfP57/urXp+ZaPomRhzTd1Ei+1R9q/KJRblqIfzXnnxa+Mui/Cmzja+aSS/nQmC0ij3s55A+XIOCeOK2fiL8RdK+Gvh271DVLhEmjiPk2+QGlfBwBzzXwzqnxOm8efESDXPEDRrbfaA+1lJEMYfcAuckcelexgMteInzW0Fc0vHfx28c/E6Z4G1W68N2LIUW10+eWLcCTgupc8gYrxzWvBbXoze3d1dvkkyTSbjz15Ir0PV5rO68Yao2nnfZSPujKkiq+sRx2unmWUqihsFvTvX39DA0KEdIknkq+EUtpA0bypIOmxsN/Kvbf2bf2mD8I/FkfhfxHfXF1pGpSJFbKG8xkmZlUYyw4x2Favwz/AGZ/EHxKuI7+93aRoRVHDXELCScHBJUqw28evrX0t4L/AGXfh94HvIr230c399EwdLi9l8/YwIIK7uhyBzXlZhWwii49Q1O+/wCEilvpXjsIHGDt8y4Qqn1zVgaRJcTebd30si/88VICfXkZq2ijnAxxjirOAw5GfrX55Us5XQ7kUVrHEhVI0RT12qBmpFUdcc+tKzlcU5MVgUKFp4UCjIpaACiiigC9bwxyxNnhh71WpofbnBxSKxoARhURY+tSVGwoAV6Ep20GjaBQAtFFFAEagGhgBT9oFNYUAMqTaPSmbTTywXrQAbR6UbR6VDNfRRpksoPuaq/2k0h2xjn1oAtzXSQ/fZR+OKy5ddYnEKLIf7x6fzpZYVuOZQH/AN4E09bdYxhVCj2p2AiWO4mO65kAH92L/wCvTkgWMYUYFWSfMXDc/hQsee1MCLH4U+lZR6UlADtopaRaXafSgBlFSeVR5VACUUUUAM2+1FLk0zJoAdRTd3vSF/egAptN833pnmH1oAcz1Ezmms9MzQAZNGTUU1xFAu53Cr71nz615h22kbTn+8i5FAGqzD58HGPb+dVLrUktxtjBuH/2RkfifSqK2d7qEiNdSqiL91YiVz/vVoQ2UcKbFQBPTqKAMtYb7VvnuH+zx9vJOGP51Pb6fb25zCig+oHNajKPQZpjJ7UgINpbg07YPSpto9KNopgRKooan4plBSBqipzMajyaTGIzGo2p70xqEAlFFFMAoopVoAgamt3pWzTGNJgRsetR7xTm700Kvp+tRYkZRQPb5qa1xBCmZXVPqaQh1MmuUjXcSFX1J4rNutVmzttoS3/TViNv86q/2fJeNmdjI391DtSgB99rwB22sDyn+8BuX8cVVFrd6go+2zmGP7wW2YrkehyK1orWKAYjQIPQVJKBIPmGf0oAzrbT7e3/ANVBHF7qgBP410OnKDDyKysVsaaB5FA7Hx//AMEqOP2efErDqPFE/wD6SWtfYQXbyK+QP+CVK/8AGOfiU9/+EpuP/SO0r7CUbhWkQIQtSp0pdlKq1oINtJtqcx0mygZBtNLtqXy6QqaFtcRDt9qKmKUxlqLDG0vWnbaTGKLCJVUelKRSI1LzQAq81pQNu61nKMc1ZjdlHBoAvMwbrUO72FRs7LnmmNNHH9+RUH95jxQBKWFKZFC5ztX1PT8T2FYUviW1aR4rRvtcqnGIzkE+5qp9j1XVGRp7w2dv94xWuACPRs0AaWoeLNP0tkWR2uGJ5W3G9h+VUbPV9T17LQW/2CBf45Fyzfgau2Oh2mns/kwR+Y33nxya042I79KQFWz8Ow2snmOGuH9Zm3fzrR+6oAAA9AMU0TOP4qM9KYDxzn3p60xactAEqtUq1AKkVqALKmrMchxjtVSNvSp0akkr7ATx55569easKx5759earpj0qePCq0kpWKFcMzvwAO5PoK3TkybMu26t82PkRBlj7DqTX56fth/F2D4lePItM0m5aTTNH3QyMnyhpQxDAAemK9Y/ag/amj0+xn8KeB70SXsmYb7UY2yiId2RGysPmBAr4ovP3KSgndKxLFu5JOSSe55r7jJcucn7aotAGtI3XdTljPrVL7USev6Vo6DYz6xfJaxgF8bnb+6PU+lfoSajER0XgnwreeK9Ta3sIZJ7nBwsaFugyeg9xzXouk6DqccRs4bYF4iyOq9dynnOP619P/srfCjSvC/hGPWjZRfab+JCrNHhhjcO3Y8V6brnwi0W+W4v9Ls4rXUWQlljGEdxz8wx3r5ernlKjVdLl0XU0sfnot8x1JIrrMSluT6CvU/2SPiJa+EfjVe6Fc3IjsdcthFBn7jXCsNoGcdcmuJ+Lfw18XeHfFF553hXUo7VR811Fbu0GPXdXh+p68bG5g1DTdUax1ezkDwTQybXSRe/scjFbYqUMdQ5YO9wP2MmuIYo2lmnhhhVcszyqoHuST0ryX4iftKeEPAdu0UV9DrGp7WK2tnKknQcZIb6cV8AX37S3jjxnZxWmv8Aiq8hijTyWjgmKK46EtzyawPtkLSGRXTezBi3Gcjv09q+Zp5G3LmmxXPVPiT8S9S+JutT3+r3LyqTiC3yTHEvXgE9cEVxaXHl7eMEdCKr2+pBkO91TPsMfge1Lo8zeL737D4bsrnX71WxJHp6iUqPU4NfX0KdLCw5VoSXjqy2rCXzlt1wdzHAA9Ae1e8/s7/BnUviJqUOt6zbvH4ct2WS1jlU7p5VZTuIOcoy5xVn4MfsgX2sXUWr+O4/s9pEwlh02JVO7ocShlPYdq+0NI0Wy0PT4rOyt47a3iUKkcS7VVRjAAHbivncyzSNJezpsRl29nHY20VvboIoY1Coi8BQBgAfhS7c9RWncwBegxVHaK+Cq1nVd2OxFtAPFSR96j3DPAqQFRXOywk7U5KRqbuK9KkRJvFG8VDmnUATZNLk0lFADdz1JuYdBmm1btZo06igCqSaOD2pWxTaAF3UbqSigBd1G40lRzXCRLuZlRfUtxQBLk01n9Onc+n1rMm1xcYiUM/oPm/PFQrHc3sgaaQxJ2VQOfrTsBfm1KGEEB1kdeoTmoFmub4kjbBH27tT7fS4bMsVUAt1wSKsLtUAAYxSAprpsKAgnzfYmnhARjHFSbie9S7AO1MCtkjvThmntH6Cnqo9KdwGNHjoKGO3gVM+PSoyoPWkBEOetOVQTyKVo8dBSopoARo8dBUm00x2OetPyaQAtDU1mKqcUxpDQAlFM303zDVAOplLk0jUANprMKT5qhLevyn3oAU4pN1U5NVt4TskljEv90Nn9KqySajeH/R1WCL++/LflQBbuL2GFcvIqn8/0qkuqvccW8TOP7zjFSQ6VGozN/pMv99qveWi9FUfQUAZa6asjZuJDKavxwxxrhEVR7DFT/gPypu0UAOH3qXd6UzdSZNADiRSNSUUAFFMLGjfQAVFk0/JqLJoGhGNCqPSm05WqR3CRPaoWqeR81BuFMY2iiimAU7pSbgDUZlpCuRyN6VAzGnOSo3Hgd89vfNZk2uW8Z28yN/sfNSGXCxI6Ee5/r6VSuNWtbZM+Z5jf3YyGP5Cs2aHU9SbbNN9nt+3lYyf97NaFnpUFm2UVWk/vY5pXIKzXWoXZUwxpCh/vD5vrU9ro6rH5ssrTv6sTj8qt7cgDnj3pf4dvapAieMbcYGKa3p2qR6ZigBtI1P2imNQOxHWvpv+orIrX03/AFNBR8if8Ep/+TcfE3t4puP/AEktK+wgCK+P/wDglJ/ybf4n/wCxquP/AEjtK+xNuW46HoSM/pWsYu1yWIq5p6xs33Vzjrjk1xfjb4weFfh3DC+sajGhbI2wMHbK4yCM8d6+Rfi1+134j8dNPpnhZG0LSxIyfaIiwllXkBshuCeK9LD4KtiHaKI5j7G8afFjwl8OreafXtatYGiUn7L56iViOcAZ69a8gb9vT4fJwugeIZR/z0Xydv8A6FXxRJp+oXk0l1f3N1fSOctNdTtK2fqxOKcNPSOTbIFQ/kPw45r6jD8Oxkv3j1J5j7N/4b++HQZVOgeICR97/U//ABVdL4f/AGyPhp4nu47aH+1NPmlOE+1xpt+mQ1fnjqtva3l1m2DjcuH3ADn1A6461Pp8dxYsJbe+Nuy9OM4rs/1fp20YuY/WrTtSstYto5rK5FxE3OU5bHFTtGfSvyt8J/H74o/DHWJL3StaXVIfuyW127lWA/2QePwr6K8C/t+zyPCnirQo0hYhZLizVtydMnG7B714OIyavTd4xHzn2PspGQelcr4B+MHhD4m26voOqxzTHBNvKNsi8DjGefrXXthevzV8/UpTpO01YtO4wR4p3FSLx1GfrWbqWs2ekx753L/7MQ3H8hXKM0W2r0FRS6lbWsbPNKsYAwMnGTXMw6hr2uXeLe2j02y5/fT8sfTArXtfDNviGS8na7lQ5LSD5aYC3WvTXUbJp1sZJD92WT/V/pzVf/hHLm+O7UrjcP7kBwtboURqAgCBem0Y/lQrFRgcD0pAVrPToLGPy4YI4x0xGoH8qs7evTmrFpJFG2XTcfrUTUwEqRVqOpqYCU+mVIOtAD1p1N4pcigB9aOm2qXMmH6fWssMasQ3DRHKNg/SgDQ1C1SzkQL8wPvUMbVF9paRgWO4jpWH428c6N8OfDdzr2u3n2Owt8bi4xnOTgepwDxVQi5u0QOiv9Ss9HsZb2+u4rKzhXMk87hEXA5OTXx98fP2pLzxRHd6F4Scw6KQ0dzdthZJmyfuMjfdNeOfHb9rC++L2opY6S76d4bt5GIijkZGn6jLgtjB9OlePN4y8vEccZVP7ueP519xlmTxk/a1hXNqXWFSR8H585JPJJrKuLoyyGTrms4XRmfdnLGt/QNDudamjiiiZ13bXPA/Gvv6ahCPLEV0aXg3wFqvi60uL2CBotOh+9dzDEfqTkGvoL4J/A99W8TW9law/bNOYRz3upRMSowwPlZ9euR7isvwhot9rUNr4R0GCTddlY5GhX93BGflLMQePWvtn4ZeCbT4a+E7LSre3jjby1lnYd5Nihst35Xv614ea5jHC0+SDvIR1FrBFpttHBbIsMMY2oiDAA9KfFcFeSefXvWHeeLNPhnEEVwLidvlCQguQfQkDFRRzanefORFbx/3V5b9eK/LJ1HOTlcvc3tU1W1WHyb0wzxOeI5PmP5V5l4g+EHgHx5HPBN4J0+1EgJ86O1jjkJP8QIBrtV09I5FeRVldejNyRV+MFvbHpWlLFVKOsJWGfKvjX/gnh4b1G2WXwrrN3p2pFiT/aMxe3IP+6gOa88b/gnj49hkynirQmj9W83/AOJr713Y/T36Uec3rXoxzjERW5Nj5X8F/wDBP/RbGz83xbrl7f3O7pp920cOMDsR619B+A/hP4Q+G1glr4f0mC3EI2m4MSCZx/tOBk10nnMv8XbHrTdztkkH/ermrZhWrfEwsSySgZ/lU9rc7V2ucisxmzUmc9a8ypd7snY0bi6EnQ1Qkk7Co/NI6UzJrIsMmpAeaiY0uTVAWKa1GTSVIDtopMmlyKTigB+4+tG4+tM5o5oAd5h9actN8s+lSbQKAFopGI7nbVSfVoLfp+9b0WgC5UM1ykPVsVR+3TTfdPk/gDQturdRn6k07AJ/a0sjbIoWB/vNVf7GZTveRmb+4T8tX+Nu3tQMA5FMCOOFEOVRVPsKlX5TkcGnR/ep6ovpQAUUUVIDPLNO+f1py09gBQBE27tTlpKdQAx91PGPSk5qL5h/+qgCRmFCsKizmjOKdgHSfep9Mk+9TDIfWgBWY461Gxpc5xRimBFk0u0UrKPSo5J406yAfXigB+6mvIqnORt9zWXJrG1tsMbzv/soSOPeo2spr5cyuUHp0oAdP4hggm8tG+0P3jjHI+tVdupal8rH7LD/AHUYhvzrQgsYYH3RxKrf3scmpdozmgDNtdCtrZtwSSSX+/K28/nWnGuxcLwPapDwcgnP1pmTQAg44FGKXimbqAEzTaXikoAKKKKACio2Y+tLvoARqbSOxpGYjpQA6oqXcfWkqeYBlItDUmcUAOdqipSSaQsF60yrhSZFV7q+it1yW59BVZb2ST7gx/vCgLlm6vIbZS0jrEq9STwa56TxW143k2FnNK3/AD84/d1fazW4lHnr5mOgPT8qnjt44hhEVB/sjFTckxzo9zrC7767kjZf4LVjGj/WtO3tIbdcRRIg9lqUMRgDpSUXKuJ/DjHFL/FnvRRUkjKbk06mNxQAklMU0rMTUe40AThRUT0qyN61GzGgsbWvpn+pNY5Na+l/6igD4y/4Jj+KNH8I/sw+KL7WdUtdMtV8U3JMlxIEziztTgZrR+Ln7aV7ql1PpHw/hxB9x9SnC7y2SMrg9MYNfEfwOtLrUvh/e232u5FiNRkb7KsjCMv5UXzFc4JwBz7V6Nay2ehwl52ESLwCeST9Otfb5RlMMRT9tN6GEpalq4sdT1q7a51F5L64ZizFmJ5Of8TWpawxWcfzhYI1AyWwMH0PatjwLZ+IPH1zHaeG9Cur6Rmx5zQsIwCeCT6V9OeAf2L4DCl542u2mvRJ5sVvp0oES9OHV4zk19dKphsFHQwuz5r8M6Bq/j66NjoGg3moMjbHulhLWyZ6FivPpX1L8Lv2P9I0JYdQ8XpHq2qS4f7OvzQRg9hxnPSvedD0DQ/Cdi8WmadZadFtG828CRk4GAW2gelee/Er9pzwV8NbVrq/1NdQYbo1hsWSSV3XqoDMOeK8StmdSq7Uikj5C/bAk0GHx7a6VominSjZKYJ51t/LEvLY24HPavAlmyu5SR75xXrXx5/bIuvi9cGz03wfptvoqHP2jVtP/wBPGM9HVyvp2rwebWvtFxvZTEPYcV9DgqkpUvf3BnRW8LXMyxqSM8lVPf0ruPCnwr1zxpsh0nQ9Q1IP8hktYGkRCfVlBrznw/4ng0a+tLyRftCROJdp/iweQfqK/QD9nn9rDwNq0MOjWugWvhS9l2okjwpEkh47g9anF4iVGleCuxW7lL4L/sX3ng/ULXxFrOs3GjXUQV/IsJGRZF4JEmVHoa+jLzWraFglrvuWYYwvOD9as3Gl3WrSJNc3U3l4B228h2svX8RUsNhFaqVjiVVJzjGf51+Z4/EOtO8lZlxZiKup6lJ5s0qW8HZVHzfjV+30uG3k3qiu/wDePWtGTJye/wBKhxXim47+LdgZ+lO3Dmmnim5NA7E/Wm1H5mKduNAWHodp4paj3Gl3e9MQ+pahWpA1MB1SLUW4Uvme9AEm4Uu6q/mN61NQBItOWmU+gCK/uPsFlc3YjeaOGMv5ceNzYHQZ71+Xn7S3xt1343eOruGWSfTtC0+Q28WmMSp3IzrudScbvmxX6l7F7rntzzXyr8cv2J7rx94xn17wneabp5uowZorwtGDKC2WAVCOeO9e5lkqKqXmB8DrCsabFGF9KEZic45+le7a5+xj8VNGuFjGjW9+ZGCqbNndV9SSQOKbH+x/8UFPz6HED6ZcfzFfokcVRtZMVjxy1liVlQzqjt90MwUnPAHNfWfwD+CfibxZax3Mul/2VpbHDXGpRtGXQhTvj4+bjJGeORXe/s4/sj6f4TtZ7/xhpVrqusPMdi3Vuk0KR5GzAdSQc5/Ovpy302OGCKBUW0gjAVIoBhAB0GBxj2rysdnHsI8tPcixh+E/DWg/Dew+x6HYR3GoEZkvXjHzt/tMOcVZutD1zxHJHJqeqG0gQ5aHTyVDD/ayP5Vvxosa4CgCplcFdpAIxjmvz/EYiVaXNJlWKOk+H7HQ0aO2hRHY7i2MEn1zWirccflSSNuPIzTN1chS0HZxU6tgcVV3VOrGlqBJuNJv96ZuPrRuo1Ak3Ve8xP7N2gfP61n76PMb1o1ATil3U2imDsxMChqM02gAp9MpcmgCeiolc+tJ5jetKwEnmCjzBSUUieYlyKMiq095BbjdJLs9uKqNrHmLiCPc395gQKCjUacD+LbWfL4ggX5EzJJ/s81XWGeb553yv91QKsR20cP+riRP91QKdgIJZLy8O8N5S/3cCltbNYvvLVsDb0pW+brTAQL7VIqntSVZs4vOfHUUAQC2k27tvFMK4rorry7a32bQW27v/rVh7Qe1K4DY/vVItNCgdKWkA7aKNopaKAFVRTWY1JwO1R8UAG0VF5jetPZiFqKmgJPMqPzD60mRTN1MCSikyKTJoAVmNRtTtwDZY4WqN5qlvacPJ8/ZVxk0AWzJioJtQihXMkix/wC8cfl61mfbb26B8iKONT0LZ3UyHRU3brgteN2MrBlH50AOm1u4uF/0KFnb+8yHbRb6FNPzf3BkPdR0FdDpc0EMEiyRqvTaMce9VpeaAIkt1jXaqgL6Y4qTaM5xTfMb1o3H1oAYeOlN2ilLUxmNAEjGot1G6mtQAuTSUU3JoAKTIoaomfFAD99G+mUUALSUUUANkokppYnrQWJ61PMAzJoyaDTZJVt13SYVfVjgVAA1RlsdQT9KzZNcFwzJaRtIR0fG5f0qL7Pe3nzSyAL/AHYzj8/equBbn1W2hGBLvf0Xn+VVjNe3pyrrCnrtzU9np0drzsUH2FSt8vA4ouMr/Y41bJO9/U1Jj2oBwc96Mmi4hO+abT6TaKkBm0elG0elOYUxmIoAZRRRQAymP1pzUxqAI6RhS0UAItMY07NRNQUJWzpf/HvWNWzpf/HvQM/M39hj9mXVPjX8MNU1iHxNHommW+tS2jwiAPIziG3YsD6YkA/CvtDwL+x34C8GxwzX8TeIdQUhmkuOFJHT5a8o/wCCUsm39nHxLg9fFVx/6R2lfYTS/KRng19Fhszq0KKpQMpLUh0vRtO0KHy9PsreyTaFxDEq8AYHQe1cd8XPjh4c+D2gXF5rN4n2uOJmt7BXJknYD5QB2zjrXZXN4lpY3V3L/qraN5mHqF5I/nX5a/FDxVffEHxjquvahI0iS3LiKORi3loHO1RzxXdhac8fV/ePQz5bHonjL9rTxz8WMaZCB4c0yU7WFsDv69S4Ga87+J2qab4g1W1tLRmntLCBYw8h3M0v8bknjNN0+4bQfCF1qZsw32iUWsJ2Y5YZDD1/+vWBkbXPlhi2GYbepPNfdUMDRw8bWuyHoVJNOtBbN5zJGnqx21qeF/2ffHHxKhuLnw1osl3ZRdZZMIemeAeo6V9Ufs0/se23iXTbHxZ4veUwTxpPZ6flTGyED/WZXrnPSvsLULzw78N9FhlvLnTfD+mp8kbXcyQI4UZ2jdj+EH34rkxOYwpP2cNWVHzPx+uvhncaXcXNnfzmG6t5DG8IGGQjqCMVs29kujxwLDMzSJ/GpwR9MdK7X48+ONG8dfGHXdS8OweTpMYWFpPLAEsqs25wVJyDkY+tcKJCo4NerRcasOZrQk+nv2cv2stY0fVoPDnie8+16Y2Fium5ePOBgknoOa+37eSG9tYLqGXfbyxLKsgwQ6tjaRX5Exx7Y/MDEegzX3r+xX8RpPGfw3udHvZS99pExRC3JMWAF79q+PzjLYRXtIko975Y01lFO6U3JxXwEo8rsdlhGprChmbnmmbj61JQtPprCjJoAdSbqTJpKCCVSaVWpv0pVHWgB+aM0ynbRTAWpI3LdTUdThAvQUwJKcKawMce+Qqo9zx+dZN34ltrc7Yg103/AE7Lv/P3oA3Nw3ZJ2rVW61y0sFJklyx/hHJ/IVjodRvZEM8i28JG4IoBJ9jx1rRtdNgjXcIlaX+8etLW2gAuvarfqiWqtZw85Z87v0PNWoYHMm+SWSdvVjT41WNQFG0DpzU27b0OKpSkv+HAmEmcg8g9c801cY6UxW96M0ry3bGS/rQMdajVzTlNMQ7caN9MLUlIB+c1OpNQ7R6U8MRQBJRUPnGneZ6UDsT0UzdSbjQIkppagmkoAKKKKACilUUrFUXcRx3/ALo+tADVzSjGeTis+4122hP7sNMfRMN+eKr273+oN++VbZewXlj9aANWS+hj+86j9f5VSa9ubgkRL5QP3WYVJDYpD98Bv0qyqjgY4XpSJsZ8OntuzOBN/ugf1q8kW1doGF9KkqVVG3pTKIdhqTb7VLRSuBDtFLt9qk2j0p+0UXAZtHpU1nJ9l5bk1HxS7RSAs3dyZGJByT92qlK38Pt0pKACiil4oAdRS7l9KZk0AG4+tRbj60bj60lMAZjjrSLuPbdRV7TZI45X3jKAUwKLA7tu3aaj4qbVNQg+0bl+Qe9Y02pORiDBP+3/APWoA1X/AHa7mIC/3ieKx5vE1q0xt7YtPOvZR1+prN/sK41Jd1/dvs/55RHata9rpdvYKFihVQPxoApxyX+oAGZhDG38OBkVeg02NY8sis/94jJq3tHPA5p/FAELR5KnHK9KQIV6cVLuooAYu5en+c01mPc05mpvy0AN3UmacVpgzQAbRTWFLk03dQAFajbNO8w03NABk03Jp1MoARmNNZR6UrUNQAlFFFABRSNJ6KaguNRhtV+b9639yLk0AOaobi+gtc+ZIFz90Hv9Kh864mk4CxQ/m/8AhQlpGZA7oHYdN3OKzsBBJqM0vMMWPdsbf55qBtOlu2zeyGdvRflX8q1qbtVegxRYCvHbxQqVjjVF9FGKf04oLUlIYN83Wo2+8acxIHFM60CGUVJtHpTdop2AbRRSbqQA1RPTvMFJuXvQAyiimlqAGNTWqRsUx8U7ARUhNHXio2cDpSAVaY1NWbPSqV1q1var8029v7iDLUDLkjbelaVjfRwxYZsVzUd3PdfdXYv95q2tOsvMh55oKPlL/glY239nHxH/ANjVcf8ApHaV9fXF1FawpNM8cMZ7yPivi/8A4Jf3N0v7PXiO3t1Vd3iedvOcZCn7JaivqmHwQl3P5+qXsl+edsYchPb5c4roiyGTal4gtNesbzTbFpL3z4GhMkK5ClhjJP41+fni34N6j8NfFeoWXimSZNNZ3nhul3BX3NkLkgAECv0kt7W3s1KwQpCp6iNQM1W8UeE9H8eaTcaXrFlb3cEy8BoxuRsYyDXvZbj1hZ67CaPy11V7/wAQXFtYabFcXFvAB5VvDlumcEeh6VHb6bcWN3C17aXFukEyyuDHjAVgcc9eM19qR/sXt4dvrm58J67FYrL80a3il3XrxkDpyKzfE37JvxG8VaGLLUPFXhwgPu3Rae6ybfrnrX3scxw1RXc9TFxbIrn9ujQ9F8I2umeEtHvLrVYbdI1+2QqLaIquPm2vkHPpXzj4x+JPjT4vagZvF2pSXVoJzNb6es0jW8AwQNqOTg4J5r0m+/4J3/EKLzpLDxNpscgGVRFZdzZzliRispv2T/jRpciwPp1vqSjpcwsSfy/xrkhHBSlzyevyGkeP6nbLalI02gJ1VRhfyqgPqTXpHxE+DHiH4cz6cuvywJc30e/yYsll9zx7dq5qLTYrb7qh6+ihZq8NiGZNjZtPNsLEe9fUv7Dsx0f4k3llE7Nb3dk7tgn7ykc14NZaYG6ALX0B+x7prW/xXeZc+UtjICOwyRivOzKN8OyEfYbNTN1KymmBTX49V+I7x7bfSo/l9KVhUPNZgTbhTcimqd3Wn7BSFcNtJUm2koJCnK1JtoTC/eJ/LNAD+tLVO61S20//AI+JVi4xhjzn6VkXGuajqJ8vTbfyl/5+JsEfgM0wOgnmjtIy8zrGq9Sxx+dZk3iN7iMDTo1nkb+NjlB+VQQeHWuNsmozteTD3wv4jvWtFbxxAhI1QeijAoAyI9L1LUv3moXjRDtDZkpH+OefyrUs7OKzXEEaxD/ZGKuRqDnIpYwKAJFBIx75qYfN1qJalpgO4oyKbRQBIrcUbqSNS7Ko6mrM2nzWiq0y7Q1AFdaetR7sUqyGgCSkY0pNNoAnqHe3rS+Y3rTaADJp+TSbRRQWSbj60oY0gpcUEEu6jIpm8VDcahBax75ZY0HT5j3+nWgC1TJZEhXc8gVf9qsSfVrq4yLJVHo8gwKWKzeVN104mk9Y87aALVxrYZtlmguj/fTlfx5pvl3l6v72Zok/55w8CrcKiJdqKqL/ALIxUobI5oAhhs4YfuRKv0FWPL+XHOPrS8UuRQAFQetKooooAdtFO3EUlFAC7j60bj60KMnmhhg8UASU+mUuTUgJRv8AcU1nx0qNadgJ2pKKKQDfMFM8w+tGD6UMoB4pgLuPrUm4e1Q5ofC9HBoAn8tfSo32jvtrMuvEEdv8qHz5P7ijn8aqx3V1fHMv7lPp81AGleala2fLzDP/ADzXr/n/ABrKXWry6YiG38pTxucZ/rUiWarzhSfUjJ/Wpwo27QMD/Z4pgQLatJ/rG3fWpI4RuzgA+3FT7Se1Kq0AKsY/z0qVeOlA4peaAEwaTFO3UcUAN2in7RTCwpDJ70AI4HpTML/k0u4Go8igCRZN3Wo91MDYpm6gB2TSZpuTRQAlLRRQAUyl5o2mgBjUNUdxdRQrl3CH0NZra00jbIYXkP8Af24X9aANRpEXq4FUJNZjDbYkM7f9M+ar/ZpZublwT/dWrsVuiDCqAPakBXZbq/8Avk23+5z/ADqW3s47VtxTLeuTmrSYXpxQzDvQBFs9qU0MxpjMaQCbqSiigCKilYUlLlYDZPu1HTzz1plHKwH0yl3U1qoBKjLUtMaswGUUrYHSm5NAC5FRtin8VBI4FOwA0h9aYz7u9Q3F1DCu6RxGuM/McfhWc2sfaOLVN4/v4JH4U7gabSY3EEgDox6Gs261JYJMLmc/3Y1z+dR/ZJrzZ9pdm29BH8oq0tpEOQgFSBR33Vxh5BsVugA6Usen/Nuwpb+8QCa0ghGfegRkUDsRiMDtWvpynyeKz9o9K1dN/wBSaCj5D/4JZQMv7OniRmGwHxPcfeGMj7Ha19fggBQBjbjHHpXzD/wTh8eXPjH9nnV/tFpa2xt/EE1uPs0YXcq2tscn3ya+nN1aRYmPqxG3T061V6Gpomp9bskuKT5YTqvXB/z7VIMnmmx4PWplZPSrTdwsPi3YUZOAcgZqwrP53DED0zVZWrnvil48t/hn4C1PxHcqjyQxOtrbl/8AXzBchc/ga7cPKU5qKYnornx9+1V4zi8UfFCZLZla00uBbYdCRMGYOCPxFeGtIJJFA6E4NM1LVLrWdQ1DUrpmSa/u5buQf9dG3YHt0qK1B8xPrX65g4+zoqJzM6SGQfL719Xfsc6Ru07W9aZdhZhbISOwAJ5r5Ks4nuCiIGLkjGBxz61+hvwd8I/8IP8ADnS9PKYuZIxPOv8AtsoOa8/OK/JQaYROwNItNyfWhSK/Jp3k7o7Lgy1DtqZmqI5rLqMZjFT0xVP+9/8AXrhfiJ8evA/wluHtfEurR2+oBPM+xKwMuzJGcHHf+VVFOexFz0DbUcjgdP5Z/P0rxD/htz4Q3UIeHxF5bt/yylKBlPoea7Dwj4/0j4lMG0fxJZSnbuMFrMpkOcEAj8R+dbewqbtE8x1194gtbNc72mb+7Au4/lVKSbV9YYCLFhA3R14kH4HirFvo8NnISsfzr/Gw5P41fVeCBwD1rHlZRmaboNrY/NM8t5NnPmXGGNaqqVGB8v04q3DZPIuQOPaozH5b7X6UgGhT35NSA0xqTJoAnjpY6ZHTx8tMB61Pmqof3qTcfWgdiXdRupmTRk0CJopijBgcMOhq/qWsHUI1Qn7vSsujaN2e9AEgFOxTPMo8yp5gJMmjJpmTRk1QD8mjJpKdtFAC5NKtJJtCliwRV79jVP8AtOOX/Unf/uHNBVzS47H86qXWqRQjC5Z/RBmqri6uceZKIfTaBk0trp8dvy6hz780EkX2vULqTCLDBF/f5LfkadBo8fmb3JmbOcsM81obR6UD5enFAAqdOOnT2pwUr0qRVHpTtooATaaVc08KOKGG3pQAUUUUAFFO4oXHpQAtFSbR6UbR6UrgNTrQ/WnhQOlBUHrQAuaM1FuPrRuPrRYAYGmqTS7896TimA7cfWjcfWm7gv3mxVK61e0tFw9wintuOCfpQBf8w+tNuJ4YRukmSMercCsWTV55v+PeMIPWX+mKrx6X53zXTtMf7rcrQBcm8QRyt5dpHI7+rLlfzFQyWtxdbWmchT2i4/nVm3tobZNkSBF/ur0qwG6UAUodPjh+6mD655/OrAXjFWNo9KTYvpQAxY6esdSqo9KdQA3b7UbfanU1mNABRu96bk03cKALlvYy3CsyLuFVpPlYL0I+9Vu01Y2qFRyDWfLMWcsD8x60ANZqazUzJooANx9aMmkpVoASmVOyj0qJhQAlFFH3WwQSPUUAFH+781U59Vtrdtpl8xv7qcmqS3WpXRyqpBE/3SR84/DpQBozX0Fu37yVUP8Ac6n8PWsz+1L64k228GF/vSKR+VJDpMcMm9/3z/3pOSPpV7y9xyaQGeumGRt1zK0p9zirqxrt24GPpU2KTgUANK57fpTlOKZuNLQgHAimM1HShlFADGems1I1JQAuTRk0u0UbRQAxqSh6KYDKZTm61Hk0mAjMaRmNK1MY0gFpj0m4+tNeQbd3QfTNQANimbqzptctfMEcMgnkH3ljOcfUULcTTd9n4CgY3VPEVlpceZZC79oYRuc/QVkW+uXutbmtLKS2i/vXSlG/Kt6z8JpIftYtlef/AJ6EAsO9MktzAxUDaR96quIxV0uSU5unNwc5+fmr9vF5f3BgVZEYHal2gVI7Ap/yeaSl2ijaKAsOzRn/ADikooKCtPTf9TWRuPrWrprfuaAPkH/glmf+MdvEn/Y0z/8ApJa19fA18gf8Es/+Td/Ef/Y0z/8ApJa19fVURMlznnvUsfeoV5HNSxtzVdBF2NqkVqrRkbwN33un/wBYVzXi74seEfANuH13WYrVyufLRlaTIzkbc5zW8acpu0Vcm52FxdQWNrLczyrFBCrSSStwqqoJP6A1+ev7SP7QNx8T/FU+nWkrR+GdPk2wxdPOkTcPMJ9CDSftGftsX3jiY6D4NMlhosbASXW+SKWU/OrAjOCpDD86+ZjrLx/6wRnnPAxX2uUZb/y8qImTuj0Jb9GCjO8LwM9v84q7BcQxuGkOwZ9z/KvKpfEUy9HUVVXxddKEAuAuRgg8DPsa+654x0RhY+z/ANmvwfY+MvH0L3tykNpYKt3tdgFkIZQFb2+9+VfdCX1o2Ejv7RvLAA2zIc4/Gvxm034lXWn2ohj1WS1ZvvmFypP4itMfFW7j2uPEmqb164u3H9a+fzLBTxj0loVHQ/Yn7HI/KNu9xg1EIZMcoSfavySg/aS8Y2MezT/F2sxL/tXLEfzrQj/bQ+JOnKIIvE12Qn3i0pbP618nPJakXob3P1d8uT+4ar3V3b2MRkuLiO3VeSZXA4/OvymuP24vi+0eIteMZ/vuZC35bq5TUfjh428UTNLrXifVLvcNpK3DqmPpms4ZPVk9iuZH6NfGH9qXw78M9PnttNuE1XWmVljW1kSVI2IIDMM9Aw5r87PF2raz8RvEeoa94jnF1qF5KXYljsQHoqg9APSucTW23OzSSuzHLNIxbP51OuueZjaQec19Vgcrhh9Z6sxbGN4dhViwjjA/D/CtHwp4w1X4V+LLLVtB1Ca2kiAJtlc7XYYCqQDjlsVm3mueVGESN5ZWbakUQ3O59AAMk/hX0f8Asnfso3/xM8RHxP4x065tNFtG/cWV1E8RkcbWViGXkV6GJjQhDVK5Fz7a+C/jG/8AiB8MfD2u6nb+RfXUIM6KCo3Dqdp55wa7eNRUlvZ2ljHFb28EcEEQwkcahVUewFad9aWkMeYl2n/eJ/rX5bjJQ9q+TY3iVIdSZE2rwPoKhmkMpy3JqL7vTijktgkbfWuAY9qSjk5pVFAEkdPqIMR0pdx9aoBafTcijcKCrku6lzUeTRk0Ek+RRkVEppd1AEm0UbRSbhRux1BP0paASZozVObUoLVtsjZb0Xr+VUl1K8u32w2rRx/35B835UwNe5vIbVMySoo/2jis9tca4OLWIyDOC46D3qGPQ0lk3XDNcv6PyK0YYUh4RQgxjC8CgDPaxuLqQPcSlkH8KcCrtrZw2v8Aqo1T6VPtHXr9aVRQA/72cj/GlwDjNLihe1AEm00bTUoK0ArQAq0tJS0AOoPPWil49KAEooooAULmjaaTzCKTefWgCXzD60eYfWod3vRu96AJvMPrR5h9ag30u73oAdvpd4qtPcxwn55Qn16Vjy+ITNIY7KF55V+8CPlP0NAG60m1sscLVKfWoYm2r+8b/ZBrLWG8ul33M21f+ea1bt7NIVwq4H1JoAZ9sudQk+YeVH9OadDpUMbZZBcHsZPmx9M9KubR6c05VFADWUegH0GKckfOaWpulAECx1L5Y9Kk2j0p+0UAN2+1Hl+1OqXK+lAEWKRjS1GzHNADt3vUzWb/AGfzf4aq7hU/9ozeT5W/936YFAEBYVFRxSbqADO3OKZQzUmaACiiigApTx0o2mmTzxwLmVvLPoaAJN1RM59FH41m3eqzzNt0+HzG/wBvhfzqK1024fBvpjvbqiH5V/GgCW41yOD5Y4pLmT+4nWoWiv79/wB7KIIe0aD5j9TWjHaxQ8pGqn1HWpGORjtQBTtLGK2bcsahvU8mrhAI55/GmFiaWgAfB6jNN47Co2Y+tG4+tACUlFSMo9KAIadTTTqSAa3Wk3e9DGm0gG0lFFUA+pNo9KjpwY+tSA2SMelRHjNSSPVdpD60wGsT1qPJpztVaW7ii6uB65I4pATM1RTTJHyzbV9TWUuq3F7xbwFR/ffp+FTpalv9dIZfrSuAx9WDcQoZT/exxVdobq//ANfP5MX92PI/UVprCI+FUD8KXYetSOxUtbGCzJaOGNWPVto3H8etPdQ3UVYkUVDigLG5petLZWvlkAj6CsS4kE88rDgNTBx0pCM0BYiopWpKCgooooAKKKaxIoAZWrpuPJNY5atXTSfJNAHyJ/wSz/5N38R/9jTP/wCklrX18OlfH/8AwS1J/wCGd/Ef/Y0Tf+klrX15uPrVREyUMRU6sFZidqKinduPIwMkn2GDk1VVt1eCftdfFifwL4XttC0q6e31bUz+8kXqsGGVlyDwWP512YejLET5ES9Cj+0R+1pH4LDaH4LnttS1cj99qEUiyRwE5+UFW+8O4r4q8WeLb3Xppr3Wb2a/v3ZmaSf5mOfTNY15eNC0rO5nlJ3F3OWJ9SfWsdppJ5iZJGYnA5PpX6Ngcthhldq7MrkJYsQWBPpmoLm1uJpMcsK27XTzdsqp8x5z/Su40PwDdalNbQWdpNdXlzKI4oowSzMfurjHHQ819Ha0ddEI4Lw38Odc8Y6xa6bptvJcXd0yqkcalyAWALYUHgZ5+lfeXgH/AIJreANN0uJ/E1/q2oarIgMwFwnlK+OdgKDAzXqH7Pn7PejfB7SU1vUR53iG4iBlkk2bYAyqSicDHOa9Om8VRzSOLRGnwdrFOmR718XmWaKnLkpD5TxqH9gv4PWseF065cf32MZP4/LWfffsX/BOykPlWt3eXDf8srZ4ifxyOK9pbT7zVNpvbp4V7x252fnWq1jZw3PmQ28anu23JP1NfPf2xif5jRRPnCT9hD4aall2tdStEY7VWOVFI9z8lYHib/gm74D1Czf+wNS1Kxvdv3r2dSmfXASvrUtuAz24z0pVkK5A78571SzrEXvIrlPgW+/4Jl+JXdBY+JNHRf4mlklP8krn9U/4JpfEi3aJLXxH4bkRvvvK84x+Udfo55zKxYHBNI0zEk7uT1rWOeVeqJ5T837f/gm18Wf4vEXhE9P+Wl1+P/LOtvRf+CafjWSZP7W8SaGsWfn+xSzbtvtujxmv0FWZ/U0/eecnrVf25U7BY8N+GP7E/wAOPhheW2oeTca5qEJQhtUdZlV1OdyjaMHNe7SyBYVjj+WNRgL2FVgdpJHB6UtedXzKtX+IXKJk9c81K0jOMMciot1Lu4ryXrqUhF561vyW9muj7wy+bXP7gtDSFl254oAcSaVWNR7jSqxoAkyaMmkGacwpALmjNJtoUUwJ6Kj3H1pryiBSZJFG3qexpgTZoaRedp+71zzWLN4jWT5LKP7VJ6pkr+NImn39/hru68lR/DB8n/fVAFyXXLeNwiMJnHDIh3EH8KrSyXuok/MbWM9OPn/wqzDp8VqzERJubqQoBNXEywAJzikBUtdLit23tumP99+v51ojB570YIo20wJOKKbRQBLtFGKbk0u6gCSik3UbhQBMtC00N6GgN6UAWKKi833o833oAl3e9R+Z70zcfWoyetAFvd70bveq/mH1o8w+tAEnme9R+YfWkyKY0iqu7PHvQBL5wpwkVvas2TUoVk2L87f7NRrNNcNnPlL9KAL1xqENqu6WREH90sAf51gXPiTVdQ40nT/JOcb9QRlT6jBzWlHpsOcuDOf+moz/ADq6o4wOOMUAc/Y+H7ybE+p3r3UzZ3Rg/ul9MA8/nW1b2628YSNdijoBU5j3ZyPT9Kk8v2oAhCkdMflTlWpdpo2GgBFTPWnKlP6UUAJtHpT9optPoAKdTaMmgB1Sw28k/wB01AWrT0m6ijHzDmgCjdQPbttPWqrMa0tXulmYMv3qyGY880ALkU3NFFABTKdzTaAE2mjbSeYvfiopLxIxgfM/pQMnzTJJkhAZ2+90FVxcn0zTtu4qSMlelAiq2pXNwdsMRh95l/lSR6YZG3XMhmPvV7aMYAAHsMU+gBkeEXCgAfSnFi2cnrTeKMigBaZS5NNagBdopuTSeZ70m4UAMPNJRRQAVK1RUu4+tADG60ZNJzSc1IDWY0mTT2x6VGTTAbk0ZNFFMB+TShjUTTJD99h+JxVFtcjY7YVMrf8ATPmkBddx3BP0rK1DXLWxV9z+Y6jPlRcuarzR6leyZmMcUXYRkqxqW10+C0bAhUyY/wBYRk/nSArW+oX+oR+YsH2aP/pqpDc1LHpyr98B/qK0G+YYNGMUAQnJ4xx6AYFCHb0p+0VHUDsSeYKPMFRUUFDZGNRsakPNJtHpQAzaKNop+0U2gCLbRspS1Ju96AGUUU3NAC7qjem7j601pDQA2tfTf9Saxi1a2myYhoA+Rf8Aglr/AMm8eI/+xom/9JLWvsWx0+XUJdkK7jXyj/wST0B9e+AfiOFW2j/hJ7j5vT/RLSvu610u10SFljkSJo/vyA5P61URMwLXwHdS7cyAZ618o/tmfs5+JPG/jC31rRII7uzs7MQGLzmU7tzEkDp3r7D1HxAYZPKtJNzJ/H61lXTR3kplmUyOerHkH8BXfg8Q8NUU2S9T8jda+A/xCtY5LiXwnqUsS9fKgZmb6YFc7Z/DnxZdElPCWuFlJU5sXX8eRX68ap4osdOuBGbtZJu1tHINx/DNZsmt3mpRsqWsdvbsMFpE2tX1seIY31iQ4n5j/C/wHrmua82mrpd1DdNt3maBkVRnG45XpX3v8KvhZpHgGxhuLGyF9rjLl7+5jA2eyjtjnmu6sfD9jZ3T3MdtCZ3Xa0gjAYj0zWsrCPO0Afh/n1rlxWfSrw5IRsVymO+h3GpzmTV7yS/3NuCMxUKPTitqG1hhjCRRiNAMBV4FNIXvz9eal4FfJyk5O7LsPVOAPTpzUhIqLzNvSjfWYyYNxSq3vUAel3e9MCbdRupu4UwMKALGBRuqLfRuNKxNh+acTUdLvFFhC0LUe4+tOosApzS7faipAM0AR0q0lAagCVWNOY1GCPU/XGapX2sWlruDTDzF6xqQzce1IC+ZKgutQgs13Sv5afmfwGKyFvtRvh+5gWCP+9OpVj9BVuDR4i26QNO3q3JpgL/a017Gv2OH5T/y0cEfpUf9jNesHvXLsOiKSB+QrUVfLACjaB6U/b3pARw2cVv/AKqNU/3RVpAahyamjY1QDse1OCbegopcmgB1FFFABRRRQA+im5NLmgB26jdUe73o3e9AEu+jfTM0ZoAduPrRvNJRQBKGoYirUlg0dqshI56+/wBKzJtQgh/1jrH7M2DQBNu96bLcogyWCD/a4H5ms17+a6i/0MDd/ekBA/DimLov2g/6W4uP9lwCPyoAl/txZJjHChc92IOB+NONvPeL++fC/wBxaurbIi4VMD0HSnhMcAVPMBWh06KH7iY/E1ZVflxjin0UcwBtp2BSbT6Uc+9HMBNtHpS1Gue9EeW6mjmAkopNoo2iqAWjNRMxB4pY/m61PMMkzTsio9y+lLuFUIXd70b6j3CjcKAHeYKarEttXrUJPWpbOQJcIz8igYSF1+8CKj3Ve1a4jm+6B+FZuTU8wh+40ZNMkljhXdJKkaf3mOBWRca5JLL5djC057sFyoo5kBsSN5a7jgL6t0qnJqC52xfO30qtbw3Nwwe4kyG/gXAA/Sr0cKxrhRgVQFNVurqTdIyxJ/dUZP41MtqkPIGT9TVjceuefWkDEVN0Am32p9R7j607Jo5gDJp+TUOTT8mjmANpo2mjJoyaoBmTTXY06mSUAMoozSbqAFpuTRu96bk0gHZNGTTcmkLGkA7Ipu+k3UxmH0oAczVFu3e1Mkuo069f7o61lXF3f3fFrGtuM7d06nd9QKYGvJOkK7nPH+e1YMviaa9maHTrSRnQkO8qFVHuM9au29iyrmWXz5P7xGD+lW1ZlVVB4XgVNwMhdFuNQ5vrozHOdqjC4+taVrbR2i4iQIParH49sUxulUAnBOT/APWpu0UmTRk1NwEoqPcfWjcfWi4DqhpS59aatSWLRRRQA2iim80AOplDOBUbPQAjYqPdTsimNTsAlFN3ik3MWUAEgfeIosAxjUTGie4ig4d1D/3c5P5VmzXdzISYV2xn7pcZ/lSAsXN0lvHvchR+Q/DPWprHxNbmH90kko7sq8ViQ6KTN51xI8hdvkSVgUUfhXV6ZCi2+AigewoA+YP+CU2vTaT+zn4mSAbHfxRcHzB1H+iWnFfWQ1SSRS0sxwf7xwtfE/8AwTK1Oe3+AviG3trfz5j4kuHGThf+PS1H9K+rF0u81Ft1/eeZCf8AljEoQVSEzd/4Sy2jmWG0cXU5+6kfzNSx3Gr37ETSRW1ufuJACH/Emq2n2cOnxrHbwpGi9PlyfzNXjM21RnIHQY4qiSOHSLKKbzjDHNN/z0miBf8AMjNXiwYc/wAqgVsjnmplx6VOoD0xS7RUW4D2p+4etO5Y7dT6Z8vpSKxPU0APLUm40vFJQA5WFKWqLIo3igCcNQKYpp2aYDtxo3H1oakpgSeYKWoqf5g9KCB1O59akt4ftF0IwPvdKsX9hLp7bZRg/wD6qAKxp9V5bhIV3SMsa+rHArPk19RIREPPz029PzqQNbKrGSzcr1NYV54u063lMUUv2q5XrbwkM/GfUio7uxutahaO4mKwN1hj+X9Rz+tX9L0ez0eIR29tGij/AGck/Unk/jTArwS6jqiIGj/s+Jv4OPM4+hxVmz0e2s5BJs8yQdJJSGb86v7RThzSAaoXdnHNTI2OhxUWKfQA+nUw8UqsaYDtoqZVFQbqkWT3pjJaKTcKN1AiSikyKb5g9aAH0VH5lHmUAPyKTJpmTRk0AOopFpaAH0USDy+rBf8AerNuteit8pFE1zIfuhcjP49KANNW3NjoPU0++uLOwGXmMny/dj5OfeucSHVNTYSXE6W0DdIVGcf8CrVt7NbePYAJD1y3J/OgCCXV9U1KMJAEt7cdN/LDtTbbQ0hk3zO103c3DeZitGMBeF4HtTlUf5NACIgVNgUbfTFSKAKSn4oAfRTcmjJrIB1FNyadQA/zBR5gqLJpcigCVZKRZKj4o4oAl3H1p2TUG80vmGtEwFJJ60BivSmZNGTWYC7zSeYaY1FagO3UbqbRz2BP0oAMml3A49qqXGoRQjAO5/ReahWW9uDmNFRPWgC5JOtQSTSyRkQsoc/dZqFhDff5qQKNq8dOlZAZ0WmySNm5czezdPyroNN0+GZQI0VPpVHGKkguJLb/AFbbfwzQBYvIRC7qDkfWq2aa0zuxLNkmm7j61fMAtFNZjQrGoAKMmkyKMigAanLTWxTlquVgLRSNSM2KsApr4pjOfWmM9ABSU3d70b6kBGpMmgmmSMEXcSFHqx4pgP3e9I0g9QB6n9ay7jWUVtsCm4b/AGPu/nUDW9xqC4nlEaf880zn86ALtxqsUDbNwd/7qcmqslxd3jfIVRP9ofN/hVm1063tl2pFgepJJ/M1NtG3HaiwFSO32/Kwyf72easLCq9FxxipNopaAGn5WOOKTpTdxNGTWYDsmmtRk0zcaq4CUVHuPrRuPrUgJRRRQAjKKFUelPIB60nA6UFkdFOwKjoAPMpm6gmoS/vQAM1Rs1K1QzsqJlpFT3PA/WnYCXdUby7Tnjb71SbUlb5IyJH/ALy/dqq9tc3wzPIYk/ugVQDr7XLeyyFLTucbUQcnnmqxbUtUUhFNlEevmY3Y/A9a1YLOK2UBI1LDocZPPvT2jXpz+ZpXAzYtKjhl81szyf3pOv51dEYAUAYC9MVL93p+vNJUgR7dowOB6VracB5NZLtWhYyMIeDQK58n/wDBLTTzd/s8+I3A+YeKLgZH/XpaV9frorjOM18sf8Emef2dfEn/AGNVx/6SWlfbCxJ/dFUgZyq6K46Zo/seT0NdmIk/uijyU/uimScf/ZMi9M1Kuly11iwp/dFSLEn90UAccdHf0NH9kye/5V2awp/dFOEEZ/hFBZxy6VLmnro0ldkbeMdEFOWFMfdFAHH/ANjSUf2PJ6V3X2aL+4KPssX9wVQHCf2LJ6Un9iP6V3Hkp/dFN8tf7ooFc46PQZW7GpD4emHQGu8toUYjKirtxbxLHkIAc0DPM30eRfWmf2XJXdvbx9NgqJreMdEFAHF/2TJ6Gl/seT3ru4bWF1JKA81VulEaTFQAVBI4oION8xNIlW4klVAvQtx+Q71R1TWr7XnzZQSM39+4Tan4GtjRbKC9m33EYnfsZPmx9M11S2MCfdiVfpSHY83tvCt3P895O857r/APwNaMfh0RY2RhQOgA4/Ku+WGNWyEGfcZqX7HD/wA8xTCxwQ0OUHIBBqddGl44P5V3K28f9wU828Q/gFAjif7Dk9DSrocnPBrvra0hkkwyAimSQRrKwCjFIDgv7Ek9DT/7Dk9DXbrCn90VJ5Mf90U7AcK2iSehoXRJPQ13ghjPVRS+RH/cFAzg/wCxZPQ0n9iyehrvfJj/ALoo8mP+6KBHCf2PL70f2RN713v2eP8AuCm+TH/dFAHDf2XL6Go/7Jm9DXfNbx/3BSLaxf3BQBwf9jy+hpf7Hk9DXe/Z4v7gpv2eP+4KAOJ/seT0NL/Y0noa9AktYlXIQVV1aNba23xqFb1xmgDg5bNLfmSQRj+8xwKzLiS5kPl2dvI0v9+RcRf99Cu302zh1DJuUE/s/I/KtmKzghH7uJU/3RigDzKHwvqFxh72bcW6rEcIv0OMmtGLwyY8bUwB04r0YW8YBGwc0vkR/wBwflSA4H+xZeeMZ9qG0R92cEGvQPs8X9wUw28f9wUXA4JdFk5qRdFk9DXefZYh/AKPs8f9wUXA4T+xZPQ07+xZPQ13n2eL+4KPs8f9wUXA4X+w5Pej+w5Peu58lP7opPKT+6KzA4f+w5Penf2HJ713Pkx/3RTxDH/cH5UAcB/Yr+hpv9iS/wCRXe/Z4/7gqHy0/uigDiP7El/yKP7Fl9/yrt/LT+6KPLT+6KAOI/sKb0pv9iTelehfZ4/7gpPs8X9wU07AcF/YclL/AGHJ713i28f9wVD5a7vuj8qQHDNosnoaj/sh1+8cD+92/GvQLi3jVMhADWBc/vbrawBX+7jAq+YDkpoiocQo1xIv/PPlarf2Hq+ofMfKt4+yozbvx4xXpa2NvGcrEo+lSLDGnRABRzAeeWvhc2xzjNWl0NlGBwPrXcfZ4x/AKXyY/wC6KOYDhP7FkNH9hvXd+TH/AHRR5Mf90VAHB/2K9M/seWu/8mP+6Kja3jHRBQBwP9jy0f2PJ/kV3zWsQUfIKGtYgOEFAHBNosmOhoXRZPQ13TQp/dFCwp/dFAHB/wBjSe9M/siX0NehLaxEcoKGtIQOEFAHANosvoaeuiy56Gu58tP7o/Kjy0H8I/KtQOE/smSo/wCyJW613v2eL+4KabeP+4KAOCGjSnrmmNosnoa9AjgjO75BxSvaxYPyCgDzttFk9DSPpgjXczqq+5xXW6pIbe13x4VvXANZ2n2sV42ZkEh/2qQHH3XnzNss4JA3d5Fwo+lMj8M3l02Z2Z/boPyFepLYwKu0RLt9MU4W0S9EAoA85Xw2VXCx7R/s8VJ/YUoGMHHpXoH2eP8AuCj7PH/cFMDgf7EkPY0HQ5Bng13nkR/3RR9njP8AAKAOB/sWT0NH9iyehru/s8f9wVD5a/3R+VAHD/2G/oaP7Df0rv8A7LF/cFRNBGCMIKzsB5+2jyKe9RtpMoruZI1/uio/JRuqikBw/wDY8npSf2RJ6Gu5+zx/3BR9nj/uCgDhv7Ik9KP7Il9K7XyU/uipPs8f9wUAcR/Y8nvR/Y0nvXctBH/cFRtGgXIUUFnDf2TL70z+x5PQ13XkR/J8g560rWsW/GwYoA8+k0l4+oNU7i1S1Yec6p6KzYz9K7HVG2WszjAZY9wOO9cv4fhS+XzblftD9jKd2PpnpQBgzS3d4zR6dbMGVtrPdIwX6jFQxeDbm8Yy3k7S8ZKqSE+gBHNenR20UShUjVVXoAKeLePgbAec81QHCR+HvK+4gX8Kl/sd67Xy0/uik8lP7oouBxi6S6tkDn6U1tIf0Ndo0KcfKKY0af3RSsBxbaS/oab/AGS/oa7Bo03D5RR5Kf3RSHY4iTS39DV+x0xvJ5zXQSQpn7opYVAXgUC5T//Z"}}},{"metadata":{},"cell_type":"markdown","source":"# Import the required Packages"},{"metadata":{"_uuid":"58a24e74-628c-4ca5-a46d-d9d17a6baca4","_cell_guid":"e4adcafe-84b6-44db-bfe6-8b100a1329c7","trusted":true},"cell_type":"code","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 in \n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the \"../input/\" directory.\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n\n# DEEP LEARNING PACKAGES\n\n\nfrom tqdm.auto import tqdm\nfrom glob import glob\nimport time, gc\nimport random\nimport cv2\n\nfrom tensorflow import keras\nimport matplotlib.image as mpimg\nfrom keras.preprocessing.image import ImageDataGenerator\nfrom keras.models import Model\nfrom keras.models import clone_model\nfrom keras.layers import Dense,Conv2D,Flatten,MaxPool2D,Dropout,BatchNormalization, Input\nfrom keras.optimizers import Adam\nfrom keras.callbacks import ReduceLROnPlateau\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import confusion_matrix\nimport PIL.Image as Image, PIL.ImageDraw as ImageDraw, PIL.ImageFont as ImageFont\nfrom matplotlib import pyplot as plt\nimport seaborn as sns\n\n\n\n# set the matplotlib backend so figures can be saved in the background\nimport matplotlib\nmatplotlib.use(\"Agg\")\nimport matplotlib.image as mpimg\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import confusion_matrix\nfrom matplotlib import pyplot as plt\nfrom PIL import Image\n\n# import the necessary keras and sklearn packages\n\nfrom sklearn.preprocessing import LabelBinarizer\nfrom sklearn.model_selection import train_test_split\n","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Calling the dataframe"},{"metadata":{"_uuid":"21aa2d72-4b19-4cb1-8ad5-d49e996ae87e","_cell_guid":"4dfc158c-eef7-4da4-966f-953b340b4bb1","trusted":true},"cell_type":"code","source":"class_map = pd.read_csv(\"../input/bengaliai-cv19/class_map.csv\")\nsample_submission = pd.read_csv(\"../input/bengaliai-cv19/sample_submission.csv\")\ntest = pd.read_csv(\"../input/bengaliai-cv19/test.csv\")\ntrain = pd.read_csv(\"../input/bengaliai-cv19/train.csv\")","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# <font color = RED**>  EDA"},{"metadata":{"_uuid":"667fdbd0-0ff9-4d9c-8ac3-1b0f066541fa","_cell_guid":"32ea432e-99bb-441c-8b39-b7a2fc96465a","trusted":true},"cell_type":"code","source":"train.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"3c70f5c5-af66-457a-8fbd-c1257bd3c8a2","_cell_guid":"4b6c6d21-2cd1-40c9-bbbe-8a991b4dad9d","trusted":true},"cell_type":"code","source":"train.tail()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"13840e7a-4ac1-4cf4-a93c-40b6cc12c758","_cell_guid":"5a93966c-9c92-4064-b9eb-f73e260e5003","trusted":true},"cell_type":"code","source":"print(len(train))\nprint(len(test))","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"10894421-3a0f-4792-b42f-d734b2dc4498","_cell_guid":"2978714d-7521-4dd4-bca2-5c4ab7fa9aab","trusted":true},"cell_type":"code","source":"train.describe()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"1666936d-254c-4e1c-94b1-abe169b3227f","_cell_guid":"820b9ded-1fb3-4c30-861f-cb5bb7ef789f","trusted":true},"cell_type":"code","source":"class_map.head()\n","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"8251d553-8340-4f0a-a7ce-c61f86f8d9b2","_cell_guid":"11520c9e-35b0-40e2-b77b-993157248191","trusted":true},"cell_type":"code","source":"class_map.describe()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d31b0949-7482-4f1a-8b71-33bc8186c916","_cell_guid":"7d98a285-92cd-4fd1-87b0-fe54357ecce0","trusted":true},"cell_type":"code","source":"len(class_map.label)\n","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"61cfd997-89f0-494b-bc03-999db138fa1e","_cell_guid":"82d74e5d-3e1c-427f-b2f2-0d49989f197b","trusted":true},"cell_type":"code","source":"len(class_map.label.unique())","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"cf6c90a0-a347-4a8e-9438-c643dafc27fb","_cell_guid":"cb9bb35f-fb38-4c9a-9653-1abccf1924ae","trusted":true},"cell_type":"code","source":"class_map.component_type.value_counts()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"f851d70f-3015-49a1-9518-e31fb709f490","_cell_guid":"0dfee528-128b-426f-9979-ffa4a734b57d","trusted":true},"cell_type":"code","source":"print(f'Size of training data: {train.shape}')\nprint(f'Size of test data: {test.shape}')\nprint(f'Size of class map: {class_map.shape}')","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"23c631c3-ab07-499c-850a-5f127fdbb4b8","_cell_guid":"9a25500b-2862-4fc5-b06e-9fc235377cc5","trusted":true},"cell_type":"markdown","source":"Top 10 Grapheme Roots"},{"metadata":{"_uuid":"f445cf94-d950-461b-95db-86a3e5d157c2","_cell_guid":"d4195ac0-a21a-4461-8c67-315ce17dce52","trusted":true},"cell_type":"code","source":"train_df_groot = train.groupby(['grapheme_root']).size().reset_index()\ntrain_df_groot=train_df_groot.rename(columns={0:'count'})","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"55735643-6014-471f-b11a-234cc1bc2236","_cell_guid":"a4264b2a-b0eb-4b79-8c53-684061258587","trusted":true},"cell_type":"code","source":"class_map_df_groot = class_map[class_map.component_type=='grapheme_root']\ngroot_merged = pd.merge(train_df_groot,class_map_df_groot[['label','component']],left_on='grapheme_root',right_on='label',how='inner')\ngroot_merged.sort_values(by=\"count\",ascending=False)[:10]","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"0b2d6bd0-e90d-4ae6-ab3e-7cd41628a42c","_cell_guid":"70034932-2e6e-4cc6-83dc-5d7ab320aada","trusted":true},"cell_type":"code","source":"groot_merged.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"22b98e93-0ce0-4689-aed5-78c295b809cd","_cell_guid":"8a9ba3d1-09b8-493c-9ae9-72881194b9f4","trusted":true},"cell_type":"markdown","source":"Top 10 Vowel Diacritic in taining data (There are only 11)"},{"metadata":{"_uuid":"a44e4fab-83a7-47ac-b4ca-1143483327ea","_cell_guid":"2e4d4301-c718-45fb-bd4e-eab8bff83f95","trusted":true},"cell_type":"code","source":"train_df_vd = train.groupby(['vowel_diacritic']).size().reset_index()\ntrain_df_vd=train_df_vd.rename(columns={0:'count'})\nclass_map_df_vd = class_map[class_map.component_type=='vowel_diacritic']\nvd_merged = pd.merge(train_df_vd,class_map_df_vd[['label','component']],left_on='vowel_diacritic',right_on='label',how='inner')\nvd_merged.sort_values(by=\"count\",ascending=False)[:10]","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"e0bf0d9a-9c7b-49dc-bb27-410f8f5f0d93","_cell_guid":"31a393f2-5171-4ab7-83ad-3f81041dc5b6","trusted":true},"cell_type":"markdown","source":"Top 5 Consonant Diacritic in training data"},{"metadata":{"_uuid":"f54b9eff-7103-404d-896c-0961df6e8434","_cell_guid":"624e7eff-298a-40db-a6bb-5e3758439fc3","trusted":true},"cell_type":"code","source":"train_df_cd = train.groupby(['consonant_diacritic']).size().reset_index()\ntrain_df_cd=train_df_cd.rename(columns={0:'count'})\nclass_map_df_cd = class_map[class_map.component_type=='consonant_diacritic']\ncd_merged = pd.merge(train_df_cd,class_map_df_cd[['label','component']],left_on='consonant_diacritic',right_on='label',how='inner')\ncd_merged.sort_values(by=\"count\",ascending=False)[:5]","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"e4a556c6-2a0a-4e07-9aee-a22345a8c4b0","_cell_guid":"ede06ddd-47b2-4354-aa79-2ff7be59f783","trusted":true},"cell_type":"code","source":"train.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"b0b7b299-c277-4a8a-add0-528f69b75e35","_cell_guid":"39467ab3-27b5-4959-9832-5e2b0fd11ba2","trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"87608b59-5985-44dc-95aa-7d6ebe17f46b","_cell_guid":"0890cf6a-f4cd-42b5-ba17-54ff1f75ce26","trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"02b554fd-0cc3-464c-b91d-7a8269f314c0","_cell_guid":"02ce247c-931b-4170-afb3-b0e1d3683055","trusted":true},"cell_type":"code","source":"train[['grapheme_root', 'vowel_diacritic', 'consonant_diacritic']] = train[['grapheme_root', 'vowel_diacritic', 'consonant_diacritic']].astype('uint8')","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Unique values\nWe look here to the distribution of grapheme roots, vowel diacritics and consonant diacritics."},{"metadata":{"trusted":true},"cell_type":"code","source":"print(f\"Train: unique grapheme roots: {train.grapheme_root.nunique()}\")\nprint(f\"Train: unique vowel diacritics: {train.vowel_diacritic.nunique()}\")\nprint(f\"Train: unique consonant diacritics: {train.consonant_diacritic.nunique()}\")\nprint(f\"Train: total unique elements: {train.grapheme_root.nunique() + train.vowel_diacritic.nunique() + train.consonant_diacritic.nunique()}\")\nprint(f\"Class map: unique elements: \\n{class_map.component_type.value_counts()}\")\nprint(f\"Total combinations: {pd.DataFrame(train.groupby(['grapheme_root', 'vowel_diacritic', 'consonant_diacritic'])).shape[0]}\")","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Heatmap\n\n**Heatmap showing the distribution of couple of features**"},{"metadata":{},"cell_type":"markdown","source":"Following code is a function to create a Heatmap.\nThanks to Gabriel Preda\n\n- feature1 - ex: vowel_diacritic\n- feature2 - ex: consonant_diacritic"},{"metadata":{"trusted":true},"cell_type":"code","source":"def plot_count_heatmap(feature1, feature2, df, size=1):  \n \n    tmp = train.groupby([feature1, feature2])['grapheme'].count()\n    df = tmp.reset_index()\n    df\n    df_m = df.pivot(feature1, feature2, \"grapheme\")\n    f, ax = plt.subplots(figsize=(9, size * 4))\n    sns.heatmap(df_m, annot=True, fmt='3.0f', linewidths=.5, ax=ax)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plot_count_heatmap('vowel_diacritic','consonant_diacritic', train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train = train.drop(['grapheme'], axis=1, inplace=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"34a99a40-d49f-459b-8f89-bd025a81413d","_cell_guid":"507f7e72-577a-426f-8ddb-640643c93c6b","trusted":true},"cell_type":"markdown","source":"Read the image file data from the first parquet file"},{"metadata":{"_uuid":"1effb1cd-106c-4041-a2a8-23623d6d615f","_cell_guid":"ab2f22dd-427e-4490-a993-3663d4c2e1f8","trusted":true},"cell_type":"code","source":"img_0=pd.read_parquet(f'/kaggle/input/bengaliai-cv19/train_image_data_0.parquet')\nimg_0.iloc[:,1:]=img_0.iloc[:,1:].astype('uint8')","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"c67377bd-987b-4c30-8af5-3a5fea4d564e","_cell_guid":"d25893c7-7495-4eae-800f-f1849a713438","trusted":true},"cell_type":"code","source":"train_df_0 = pd.merge(train,img_0,on='image_id',how='inner')","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d86ca853-7ace-4b5d-ba2c-39b830eb3d25","_cell_guid":"9f7d8329-f001-46c8-b74f-9dfd6fa9e917","trusted":true},"cell_type":"code","source":"train_df_0.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"0f09c6e5-57fe-4c8f-8937-59eb94ac220a","_cell_guid":"3556b5be-d3f4-4413-a0e7-7e1556f4a2d5","trusted":true},"cell_type":"code","source":"train_df_0.shape","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"fdf9e3de-6058-42e4-9ad8-d87fdea07fec","_cell_guid":"d2d256fc-b604-4d95-8027-2a69f1b67ecc","trusted":true},"cell_type":"code","source":"train_df_0.vowel_diacritic.value_counts()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"02b6cfc2-5186-4a3b-8364-3270941b606b","_cell_guid":"e3b1ca1c-299c-4bb8-96a6-1276c388ee66","trusted":true},"cell_type":"code","source":"train_df_0.consonant_diacritic.value_counts()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"80cd6c56-3699-4ce3-af0f-056fdb566c31","_cell_guid":"10007e2a-267a-48a1-b9f6-b744a0c6bf27","trusted":true},"cell_type":"code","source":"train_df_0.grapheme_root.value_counts()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d59f197e-31b2-4135-8560-144cb065b1de","_cell_guid":"e52d31d7-4ec0-4b53-ae07-7b8af04c7a5f","trusted":true},"cell_type":"code","source":"img = img_0.iloc[0,1:]\nimg=img.astype(int)\nimg = np.array(img).reshape(137,236)\nplt.imshow(img);","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"c605bf47-8e64-4137-a142-b727f805f3e7","_cell_guid":"32a86bf7-640a-4636-bf78-b53fe406e740","trusted":true},"cell_type":"code","source":"img = img_0.iloc[10,1:]\nimg=img.astype('float32')\nimg = np.array(img).reshape(137,236)\n# Construct image object from array, needed for resizing\nimg = Image.fromarray(img)\nplt.imshow(img);","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"2738c5b0-3740-4971-9dd1-3d45e2c17fc4","_cell_guid":"8519e1ef-62da-4941-9f17-d0356e338de7","trusted":true},"cell_type":"markdown","source":"# *Resize an image using openCV and check (resize to (96,96) input shape for CNN)*"},{"metadata":{"_uuid":"a358da06-2f00-4c9b-afa6-d81cb813a4e2","_cell_guid":"a4628bc1-6a2a-4f5a-90fb-4c8912827d07","trusted":true},"cell_type":"code","source":"img_resized = img.resize((96,96))\nplt.imshow(img_resized);","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"fe97130e-0c90-4911-8767-ba45f5c47c3a","_cell_guid":"6951a811-c40d-410d-bdb7-7f570e2f8b94","trusted":true},"cell_type":"code","source":"img_resized=np.array(img_resized).reshape(96,96,1)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"1f90499d-afd2-4761-bf94-af6c01e95404","_cell_guid":"eee4dc22-38fd-4717-b158-ef6ae8cd562e","trusted":true},"cell_type":"markdown","source":"Define Multi Channel CNN, one channel each for Graphemes, vowels and consonants"},{"metadata":{"_uuid":"c86ea822-9a4e-4533-8b25-25df3d3e16e7","_cell_guid":"147fa342-bbd7-4874-b775-e01b6262fbd8","trusted":true},"cell_type":"code","source":"class BengaliNet:\n    @staticmethod\n    def build_grapheme_branch(inputs, numGraphemes,finalAct=\"sigmoid\", chanDim=-1):\n \n        x = Conv2D(32, (3, 3), padding=\"same\")(inputs)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(3, 3))(x)\n        x = Dropout(0.25)(x)\n        \n        x = Conv2D(64, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = Conv2D(64, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(2, 2))(x)\n        x = Dropout(0.25)(x)\n\n        x = Conv2D(128, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = Conv2D(128, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(2, 2))(x)\n        x = Dropout(0.25)(x)\n\n        x = Flatten()(x)\n        x = Dense(256)(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization()(x)\n        x = Dropout(0.5)(x)\n        x = Dense(numGraphemes)(x)\n        x = Activation(finalAct, name=\"grapheme_output\")(x)\n \n        # return the Grapheme prediction sub-network\n        return x\n    \n    @staticmethod\n    def build_vowel_branch(inputs, numVowels, finalAct=\"sigmoid\",chanDim=-1):\n\n        x = Conv2D(16, (3, 3), padding=\"same\")(inputs)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(3, 3))(x)\n        x = Dropout(0.25)(x)\n\n        x = Conv2D(32, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(2, 2))(x)\n        x = Dropout(0.25)(x)\n\n        x = Conv2D(64, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = Conv2D(64, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(2, 2))(x)\n        x = Dropout(0.25)(x)\n\n        x = Flatten()(x)\n        x = Dense(128)(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization()(x)\n        x = Dropout(0.5)(x)\n        x = Dense(numVowels)(x)\n        x = Activation(finalAct, name=\"vowel_output\")(x)\n\n        # return the vowel prediction sub-network\n        return x\n    \n    @staticmethod\n    def build_consonant_branch(inputs, numConsonants, finalAct=\"sigmoid\",chanDim=-1):\n\n        x = Conv2D(16, (3, 3), padding=\"same\")(inputs)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(3, 3))(x)\n        x = Dropout(0.25)(x)\n\n        x = Conv2D(32, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(2, 2))(x)\n        x = Dropout(0.25)(x)\n\n        x = Conv2D(64, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = Conv2D(64, (3, 3), padding=\"same\")(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization(axis=chanDim)(x)\n        x = MaxPooling2D(pool_size=(2, 2))(x)\n        x = Dropout(0.25)(x)\n\n        x = Flatten()(x)\n        x = Dense(128)(x)\n        x = Activation(\"relu\")(x)\n        x = BatchNormalization()(x)\n        x = Dropout(0.5)(x)\n        x = Dense(numConsonants)(x)\n        x = Activation(finalAct, name=\"consonant_output\")(x)\n\n        # return the consonant prediction sub-network\n        return x\n    \n    @staticmethod\n    def build(width, height, numGraphemes, numVowels, numConsonants, finalAct=\"sigmoid\"):\n        # initialize the input shape and channel dimension (this code\n        # assumes you are using TensorFlow which utilizes channels\n        # last ordering)\n        inputShape = (height, width,1)\n        chanDim = -1\n\n        # construct both the \"grapheme\" , \"vowel\", and \"consonant\" sub-networks\n        inputs = Input(shape=inputShape)\n        graphemeBranch = BengaliNet.build_grapheme_branch(inputs,\n            numGraphemes, finalAct=finalAct, chanDim=chanDim)\n        vowelBranch = BengaliNet.build_vowel_branch(inputs,\n            numVowels, finalAct=finalAct, chanDim=chanDim)\n        consonantBranch = BengaliNet.build_consonant_branch(inputs,\n            numConsonants, finalAct=finalAct, chanDim=chanDim)\n\n        # create the model using our input (the batch of images) and\n        # three separate outputs -- one for the grapheme\n        # branch, the vowel branch, and consonant branch respectively\n        model = Model(\n            inputs=inputs,\n            outputs=[graphemeBranch, vowelBranch, consonantBranch],\n            name=\"Bengalinet\")\n\n        # return the constructed network architecture\n        return model","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"b1b4576c-8220-44f7-a647-43c0efa9b9ef","_cell_guid":"1568fc63-a966-4785-be04-6b7ff1f63f4d","trusted":true},"cell_type":"code","source":"EPOCHS = 50\nINIT_LR = 1e-3\nBS = 32\nIMAGE_DIMS = (96, 96, 1)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"fc866829-640e-4b2d-a8b0-f0a545877d21","_cell_guid":"81002a58-fda0-4dce-b167-4a9e63bd9ddd","trusted":true},"cell_type":"code","source":"data = []\ngraphemeLabels = []\nvowelLabels = []\nconsonantLabels = []","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"18fffdab-31ce-4c0c-9738-0ccb529ffc4c","_cell_guid":"8f7c819a-1660-468d-a401-0e5fad3bc266","trusted":true},"cell_type":"code","source":"del train\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"68fbc419-8cef-45c0-92b1-c1c327594956","_cell_guid":"64e6f59c-733f-4e7c-a09d-d667dd608c05","trusted":true},"cell_type":"code","source":"train_df_0.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"2ee291f3-d830-4bff-b6e6-0566a72b1c21","_cell_guid":"b4e0ef5f-a872-44a5-bff0-ffcde7259a02","trusted":true},"cell_type":"code","source":"graphemeLabels = train_df_0.grapheme_root\ntrain_df_0=train_df_0.drop(\"grapheme_root\",axis=1)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"eacefe12-9718-4f44-8136-b073b938f0aa","_cell_guid":"2ffed88e-c436-4874-8c80-423fcd58585b","trusted":true},"cell_type":"code","source":"vowelLabels = train_df_0.vowel_diacritic\ntrain_df_0=train_df_0.drop(\"vowel_diacritic\",axis=1)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d544cb0d-ae10-482e-879c-f441dc16cad3","_cell_guid":"de3964e4-7db0-4f45-828f-92b6014e6c68","trusted":true},"cell_type":"code","source":"train_df_0.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"fbe42040-1b33-4d43-84dc-0d0c4308a6ec","_cell_guid":"e436842d-7c0e-4761-b250-883dbe9d8914","trusted":true},"cell_type":"code","source":"consonantLabels = train_df_0.consonant_diacritic\ntrain_df_0=train_df_0.drop(\"consonant_diacritic\",axis=1)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"fc4c3124-2a64-4183-9084-789d08c495c1","_cell_guid":"331683af-52ff-40ad-8211-7273ceed8ecb","trusted":true},"cell_type":"code","source":"train_df_0.head()","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"deb898ff-08ce-49ba-a90f-d35e111af873","_cell_guid":"ed4a9c16-7291-4d74-b2aa-a6f048d41fef","trusted":true},"cell_type":"markdown","source":"### Create the X and y from the Images and labels (Just taking 50 values to check the process)"},{"metadata":{"_uuid":"edde373c-12f4-4682-9f65-785c28d08860","_cell_guid":"d35ac341-70ca-4a71-8b76-00e834681fbe","trusted":true},"cell_type":"code","source":"for i in range(len(train_df_0)):\n    img = train_df_0.iloc[i,1:]\n    img=img.astype('float32')\n    img = np.array(img).reshape(137,236)\n    img = Image.fromarray(img)\n    img_resized = img.resize((96,96))\n    img_resized = np.array(img_resized).reshape(96,96,1)\n    data.append(img_resized)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"5fb7bb27-33ff-4169-9e5c-dc872ec1ebbd","_cell_guid":"bb4cba8f-4ffa-4a23-ab4a-34a2576f76e8","trusted":true},"cell_type":"markdown","source":"del train_df_0\ngc.collect()"},{"metadata":{"_uuid":"aed66430-1d01-451e-b9be-7bb9907c505e","_cell_guid":"dfa17886-9564-4698-9475-a283251921ac","trusted":true},"cell_type":"code","source":"graphemeLabels[:10]","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"c20cdced-da1d-42fc-aa22-92cadd19fca3","_cell_guid":"b2b56703-613a-4ca7-8e58-47e5e6e9e9ce","trusted":true},"cell_type":"code","source":"data[0].shape","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"69fbd492-3c8e-4c9a-890d-7d238df99179","_cell_guid":"60671d72-b106-4be7-8325-f398f6af4c3d","trusted":true},"cell_type":"markdown","source":"# Binarize the Labels"},{"metadata":{"_uuid":"18e4a322-93d7-42dc-adec-291f9e4e9741","_cell_guid":"7a074326-870c-4a98-a91f-d98b52c0020c","trusted":true},"cell_type":"code","source":"import sklearn\nfrom sklearn import preprocessing\nsklearn.preprocessing.label_binarize\n\n\ndata = np.array(data, dtype=\"float\") / 255.0\n \n# convert the label lists to NumPy arrays prior to binarization\ngraphemeLabels = np.array(graphemeLabels)\nvowelLabels = np.array(vowelLabels)\nconsonantLabels = np.array(consonantLabels)\n \n# binarize all three sets of labels\nprint(\"[INFO] binarizing labels...\")\ngraphemeLB = preprocessing.LabelBinarizer()\nvowelLB = preprocessing.LabelBinarizer()\nconsonantLB = preprocessing.LabelBinarizer()\ngraphemeLabels = graphemeLB.fit_transform(graphemeLabels)\nvowelLabels = vowelLB.fit_transform(vowelLabels)\nconsonantLabels = consonantLB.fit_transform(consonantLabels)\n \nprint(graphemeLabels.shape)\nprint(vowelLabels.shape)\nprint(consonantLabels.shape)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":""},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df_0=train_df_0.drop('image_id',axis=1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df_0.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def preprocess(image):\n    resized_image = tf.image.resize(image,[96,96])\n    final_image = tf.keras.applications.xception.preprocess_input(resized_image)\n    return final_image\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# partition the data into training and testing splits using 90% of\n# the data for training and the remaining 10% for testing\n(trainX, testX, trainGraphemeY, testGraphemeY,trainVowelY, testVowelY,trainConsonantY,testConsonantY) = train_test_split(train_df_0, graphemeLabels, vowelLabels,consonantLabels,test_size=0.1, random_state=42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"trainX=np.array(trainX).reshape(trainX.shape[0],137,236,1)\n#trainX=tf.convert_to_tensor(trainX)\ntrainX.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"testX=np.array(testX).reshape(testX.shape[0],137,236,1)\n#testX=tf.convert_to_tensor(testX)\ntestX.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"EPOCHS = 1\nINIT_LR = 1e-3\nBS = 32\nIMAGE_DIMS = (96, 96, 1)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"> Create TensorFlow Dataset from images, after resizing¶"},{"metadata":{"trusted":true},"cell_type":"code","source":"import tensorflow as tf","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_set =  tf.data.Dataset.from_tensor_slices((trainX))\ntrain_set=train_set.map(preprocess).batch(BS).prefetch(1)\nval_set =  tf.data.Dataset.from_tensor_slices((testX))\nval_set=val_set.map(preprocess).batch(BS).prefetch(1)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Create TensorFlow Dataset from labels"},{"metadata":{"trusted":true},"cell_type":"code","source":"train_Y=tf.data.Dataset.from_tensor_slices((trainGraphemeY, trainVowelY, trainConsonantY))\ntrain_Y = train_Y.batch(BS).prefetch(1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test_Y=tf.data.Dataset.from_tensor_slices((testGraphemeY, testVowelY, testConsonantY))\ntest_Y = test_Y.batch(BS).prefetch(1)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Merge the images and labels to create train dataset and validation dataset"},{"metadata":{"trusted":true},"cell_type":"code","source":"dataset  = tf.data.Dataset.zip((train_set, train_Y))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"val_dataset = tf.data.Dataset.zip((val_set, test_Y))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import keras\nclass MultiOutputDataGenerator(tf.keras.preprocessing.image.ImageDataGenerator):\n\n    def flow(self,\n             x,\n             y=None,\n             batch_size=32,\n             shuffle=True,\n             sample_weight=None,\n             seed=None,\n             save_to_dir=None,\n             save_prefix='',\n             save_format='png',\n             subset=None):\n\n        targets = None\n        target_lengths = {}\n        ordered_outputs = []\n        for output, target in y.items():\n            if targets is None:\n                targets = target\n            else:\n                targets = np.concatenate((targets, target), axis=1)\n            target_lengths[output] = target.shape[1]\n            ordered_outputs.append(output)\n\n\n        for flowx, flowy in super().flow(x, targets, batch_size=batch_size,\n                                         shuffle=shuffle):\n            target_dict = {}\n            i = 0\n            for output in ordered_outputs:\n                target_length = target_lengths[output]\n                target_dict[output] = flowy[:, i: i + target_length]\n                i += target_length\n\n            yield flowx, target_dict","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Data augmentation for creating more training data\ndatagen = MultiOutputDataGenerator(\n    featurewise_center=False,  # set input mean to 0 over the dataset\n    samplewise_center=False,  # set each sample mean to 0\n    featurewise_std_normalization=False,  # divide inputs by std of the dataset\n    samplewise_std_normalization=False,  # divide each input by its std\n    zca_whitening=False,  # apply ZCA whitening\n    rotation_range=10,  # randomly rotate images in the range (degrees, 0 to 180)\n    zoom_range = 0.20, # Randomly zoom image \n    width_shift_range=0.20,  # randomly shift images horizontally (fraction of total width)\n    height_shift_range=0.20,  # randomly shift images vertically (fraction of total height)\n    horizontal_flip=False,  # randomly flip images\n    vertical_flip=False)  # randomly flip images\n#datagen.fit(trainX)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Build and train the model"},{"metadata":{"trusted":true},"cell_type":"code","source":"from keras.layers import Activation, Dense","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#strategy = tf.distribute.MirroredStrategy()\n\n#with strategy.scope():\n# initialize our BengaliNet multi-output network\nmodel = BengaliNet.build(96, 96,numGraphemes=len(graphemeLB.classes_),numVowels=len(vowelLB.classes_),numConsonants=len(consonantLB.classes_),finalAct=\"softmax\")\n\n# define two dictionaries: one that specifies the loss method for\n# each output of the network along with a second dictionary that\n# specifies the weight per loss\nlosses = {\n    \"grapheme_output\": \"categorical_crossentropy\",\n    \"vowel_output\": \"categorical_crossentropy\",\n    \"consonant_output\": \"categorical_crossentropy\"\n}\nlossWeights = {\"grapheme_output\": 1.0, \"vowel_output\": 1.0, \"consonant_output\":1.0}\n\n# initialize the optimizer and compile the model\nprint(\"[INFO] compiling model...\")\nopt = tf.keras.optimizers.Adam(lr=INIT_LR, decay=INIT_LR / EPOCHS)\nmodel.compile(optimizer=opt, loss=losses, loss_weights=lossWeights,metrics=[\"accuracy\"])","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":1}