# Excalidraw Data ## Text Elements Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking ^kBlyLraV IDEA FROM? ^9VMoRVD8 STaR: Self-Taught Reasoner LMs can improve reasoning ability on QA datasets by sampling rationales to attempt to answer questions ^CQbV2u2h IDEA FROM? ^opbMBPEL Problem? ^uRwJ2Dp3 Only on reasoning task; Generalization is not Good ^KusOrxKx Process ^MnA4RcB5 1. LM learn from the diverse tasks from a large internet text corpus 2. Leverage pre-existing reasoning ability to generate rationales and train the LM on them with a REINFORCE-based reward ^2zrm4z7y Example ^hZU2X3Ws Contributions ^UZoCndsJ Contributions: 1. generalize STaR to learn reasoning from diverse unstructured text data 2. parallel sampling algorithm 3. introduce custom meta-tokens at the start and end of each thought to follow the LM to learn that it should be generating a rationale and when it should make a prediction based on that rationale. 4. mixing head 5. non-myopic loss 6. performs better ^ak8RB5ky Experiments Results on GSM8K and Commonsense QA ^hcYbtT8R algorithm ^Za6egSRx Three Main Steps ^3nda6CAa 1. Parallel rationale generation: generating r rationales of length t; insert learned <startofthought> and <endofthought> tokens to mark each rationale's start and end 2. Mixing post-rationale and base predictions; train a mixing-head - a shallow MLP producing a weight determining how much the post-rationale next-token predicted logits should be incorporated compared to the base language model predicted logits 3. Optimizing rationale generation; optimizing the rationale generation parameters(start/end tokens and LM weights) to increase the likelihood of rationales that makes future text more probable. ^7EIS8qVy Think ^rsf76pvF Talk ^eZirZJjd Learn ^izzXY0Jy Parallel Generation ^SxhOE8Ni three-layer MLP with ReLU activation ^GSnFAD0v Mixing Head ^Khc3MytL Forward Pass and Teacher Forcing ^xInF7gyf Optimize Meta Token ^HeyIssV7 Optimizing the representation of these tokens, especially the <|startofthought|> token, is crucial but challenging due to the discrete nature of the rationale tokens. We initialize the start and end token embeddings to the embedding corresponding to the em dash, ”−−−”, which often appears in text data to denote a pause or thought. ^ceTRS2EW Zero-shot performance on Quiet-STaR applied to chain-of-thought on GSM8K ^AHiYWfVn Limitations: 1. Do not support dynamically predicting when to generate or end, a rationale 2. Only works on 7B model ^QPNFZ9Dj Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking ^v4ZdxArB Let's verify step by step ---OPENAI ^uD7dsxyG Outcome Supervision / Process Supervision ^DaTq6qSi Process Supervision outperforms outcomes supervision for training models to solve problems from the challenging MATH dataset. ^NrlQ0STR PRM800K dataset, the complete dataset of 800,000 step-level human feedback labels used to train our best reward model. ^BKcogJcT Why Process Supervision? ^rmLqSUan 1. Hallucination. CoT can help the model to reason but it will cause hallucination 2. Outcome-supervised reward models only use the final results of the model's chain-of thought; they regularly use incorrect reasoning to reach the correct final answer 3. Process Supervision is much easier for humans to interpret and provides more precise feedback, since it specifies the exact location of any errors that occur ^BI7uGgh0 Data Collection Example ^iKFkSz1J Math Dataset ^l7rJS8IQ ORMs ^bHXOT5aY 1. train the ORM to predict whether each solution is correct or incorrect; usually determine correctness by automatically checking the final answer 2. automatic grading used to determine ORM targets is not perfectly reliable ^ScDjQmaE 1. train PRMs to predict the correctness of each step after the last token in each step 2. define the PRM score for a solution to be the probability that every step is correct under the PRM ^ZyqQds8V PRMs ^vV9SaMg8 Visualize ^U44qN4zT Result Comparison ^bJYR8og5 EXPERIMENT ^sUHbnn86 Large Scale ^ynatAaW3 Finetune all models from GPT-4; Train the most reliable ORM and PRM ^WvhHk7SQ Small scale ^MJVbU2YS direct c ^3d8aYzYI Generator to generate solutions in a newline delimited step-by-step format ^FHL3sY28 Explanation ^3bwc712i Data Collection ^3c6Sfn6a Specifically, we few-shot generate solutions to MATH training problems, filter to those that reach the correct final answer, and finetune the base model on this dataset for a single epoch. This step is not intended to teach the generator new skills; it is intended only to teach the generator to produce solutions in the desired format. ^CBIYRo7T To collect process supervision data, we present human data-labelers with stepby-step solutions to MATH problems sampled by the large-scale generator. Their task is to assign each step in the solution a label of positive, negative, or neutral, A positive label indicates that the step is correct and reasonable. A negative label indicates that the step is either incorrect or unreasonable. A neutral label indicates ambiguity. In practice, a step may be labelled neutral if it is subtly misleading, or if it is a poor suggestion that is technically still valid. ^kKyzNJ0z Explanation ^lREgyqX6 PRMS / ORMs ^Fw6Y9kZ6 PRM800K ^hFAX1LrA ## Embedded Files 25ce7a11dc5569ee8f8847e55db12c2e19b80351: [[Pasted Image 20240916120441_722.png]] 59f1adcc93d14e371f061928ec4226a88efca629: [[Pasted Image 20240916122541_824.png]] 8641e31d85c173584ae84712e0e2d460db07116e: [[Pasted Image 20240916122648_760.png]] b1319f1d4604b33cf9144c3c5323855dbec9dc94: [[Pasted Image 20240916124156_267.png]] 0d7783036dcde1b91b32f32148cffa02c920e09b: [[Pasted Image 20240916125309_567.png]] 235fd506448912ecadca33f3f9ec5556b3acdc86: [[Pasted Image 20240916125559_022.png]] f264a940477c0012ca1391b1f016a48e2440f18e: [[Pasted Image 20240916152402_175.png]] 473b843e3023e9ed1e5c53f7f1aafb9ff32685c1: [[Pasted Image 20240916153202_704.png]] 62a24bf685a70411806a0f9e63aac480f7b3288a: [[Pasted Image 20240916154159_459.png]] 27d75065d30725c4c9417114b54c3da9925e24f7: [[Pasted Image 20240916202817_483.png]] 20e26c21fd13c8f51c2469da23d6bffa2e14cb1b: [[Pasted Image 20240916203103_177.png]] %% ## Drawing ```compressed-json N4KAkARALgngDgUwgLgAQQQDwMYEMA2AlgCYBOuA7hADTgQBuCpAzoQPYB2KqATLZMzYBXUtiRoIACyhQ4zZAHoFAc0JRJQgEYA6bGwC2CgF7N6hbEcK4OCtptbErHALRY8RMpWdx8Q1TdIEfARcZgRmBShcZQUebQBWbQAGGjoghH0EDihmbgBtcDBQMBLoeHF0XFJSNip+UsYWdi40AE4AZgB2eshG1k4AOU4xbgBGTvaeJPiADnaAFj5CyEIO YixuCFwUnohCZgARdKgEYm4AMwIw3ZJNtgBFAwGAcXPmACEOAEkAYVJiACiAx4zAAsgBBZ6pEqQc6EfD4ADKsGCm0EHmhpWYUBqAGsEAB1EjqbhLGEQbF4hDImCoiTo267Gp+SQccK5NCjXZsOC4bBqGBjJJJXbWZR01Ai5YQTBjRZxTqtJXtYXxebxeLtXaCtDOeXaABso3mrU6PFGM1GPANBt2lLY+J+bHwbFImwAxKNTaNsDxMZBNHzccpmWs nS63RJ3edzthWnH/RAKMTJHKktpOpmZvMkiqeK15p0DfFdpIEIRlNJuPNRokkq1hdm5uN4kkudKwqdSVbCwWZp15rsQ8I4F9iBzUAUYZAAPq4mYAR0wAAkACqaACqAGlWguKEYADJQeiaA2EABqEGWAF1dudyJkx9wOEIEUzhGs2cwJ8VybBEKSywAL67JoH6AsEmTZBOeS3tKQhwMQuAnGcnJmgaJozMW7SWjMuxEBwuLPq++D4Ww2D4qhqCXPg 1wdlEpBQO8qyOBwyjcL+pTpFBUBPhImhGDApA8AMh5JAAGjAoIUPMABSPAbhQCBJAAWv6sLkUIE7OKMyTCrWVoGjwKoYfM2Y9JAyi4HA3AzNoSrxBM4xWqMXQ8DMJbLLChCYKcAAKbCrFAxEIsBuxZMQzFrKs7FoJxkDcVkvFURApDnJIXzOPQ7TIhwnS4JoBzvLJmhbjMMziepEDnJpE7pjw8Q8PMKpJPMBpFq0UyjO204QFZNloO0yRtRqSQzM ZkytOVSS2l51U+f5gXZCF+BhdK+ChFATr6PoagoQFQUrXaDFQOC1S1GWuBUS+b7ShFZ01BQl0pVUj2JnAS05PkXlTtOUrTmASReXB06/TCrnaON7lGja8yLB5bYWWADWdNooy5g1rYzGNFqdMDFlgyUOnWvZaqmlhOHxF6tZIzWdYNmNZntC2bbAze4X4GBSlUTRYSFCBhScWUAESK9tSJn0zS2Q1uyS4MwwVEanSjO1bVkqULEbKLoyJvsRzBCh FxXAgNwpV8uAAgSM6kAAVvQABKyitLi9uroQRlbhuPx+YmcIIjSEoUs6jL0VSRLECSaDqwIOIOtSKIVEHGLviyX4Tj1pQ8nyApCv9pRihKeeQLKnLWum5VU45HQGvW7Xatweoq2jXo1waprmpqg6h3H4auh6owIAPA+JoGFHDkIYbOn3UaaDw5w8JomiJsmEepm0lrJO0WOdONRrtK0s3kmWFZVmg2PzNoNbWkaBZtireEdggXZtK08Q2uN41Dsy o7jt9vVzouFc65ty7n3EeE8Z5Lzs2lPeXAj5rokRTp+dkHE5r/gqDwNa5IwITwghkJKMEvLxUgOgzYJxMDBSRnsFK4JlyEAAJoEnOOeLgXkgLQPJAhJChs0LWjMt6Bq+98KrCImgG6pF1rkUokbWiJt6JVCYixGKqDeqJWyHxdAAkhIiTEpJaSckFJKVUlVGq2AtLcHTDDL07QbQahmAWSaFk+rWVJAkYUNjiy5iSE1dqHQnFwl8sQA6y0xEkSwV xNYUVWKxVQPFDAkEkoaIgBuTQyh2jggAPK8lXLiDJBw7D0NxLJGAMwfjnBMbVBu5cmruU6PEDonRpjuVfk4/qYxDSzHauafMr827Wi1HNAJi1DqhNCjCAW5INrYm2rtGQwyQmoHEcdBRD0LohAQbdck91zpPXWZsMWdRpQfSCoQ0GXlAZIyBtOEGMJCZgGcLMcuHR3JtRrAOcqSNnBGVaJDIy41moLHNK1fGP1zk5nTMaC0uYLTNQbA/AGzhz6X0 MjfHM3UDQzDZjCG5CVOa1GftRY2/N6hC1IRIchlDpRyxaLwBYssmD9A4EMDgIw0CFjhp1TU8KNZrC1pUP0NxDjHAJbzOR5JbgSFoQwphLDfbwiRAnNEwczjHTDimUkqq44B0TgyFV0pmSVjTmMbkvJ+SwFzqKNihddgl1QFvXSjVzQq3NIsNu3LIA6lQDpJIjT7L/OZiZGxXRNWOinpGdAnpB5RpHkGcek8IwelnvPRey91VoGLAaey+8qY1hrrW IypZyyVmCmgVsdon5UVcqaCY9iLRfxHGOU55IAFLjXJuHce5DzHlPBeK82K7wPgQBopZ+rwJGtGRI7B4EAQJOgvkHFEAuHIQJeMPh9jlaCOjhAAiojFmIMkRREVxs7ycCgIiQgRgMFF2qqegAYnA+EnrPJ/iwCW9AKkmBsGcMwSQbAoCoEQGlV0+hrBiFQJwVA9whCEAQFAZwiJVy4HtqgayPgYPEFQFANgqBsCSFwKsZwbBzjOHUMIYtAAdDgEH niIlBDMLciZyAUFXK+zYH6ajft/f+wDNVSAgZZQgcDHBIPQdg/BxDyHUNEFOJh7DuH8MuCIyR39LJ/3Udo/RxMFLwREGUDSiAYhshMAlkwKA5gCA6YrPp3axBiCol2HobICmmBJPNpba2dtHbO1du7HgntvaJldBWVYBAWMULY5+zjf6ANMF4/xsDEGoMwbgwhpDKG4BoZk1hnDeGCNKdI6poTqAaN0YY6KIQWH7bhAvRUHEQgxWlAIggZcRbT6o F0g1YlgtpRkvQIEbAUQrVIHpU0Tg3BmatBG4y5lrLUA7zqY2WYNxeW2q2O0XWQqDZHtkabTYWjhKiQPBJKSMl5KKWUmpO88rtVKuTt3fE4dI68BDfHWkOrlWMbI6yFBnITXZ3NZyYUlrxQVGvba9q6ZHLCldWXcYAzySesbkNG+TVWptytP2F7vdw0QEjUPYb0pR7BlDMQbHia54LyXrsFeT26lxAzTLaUx9i22Tsq5dyVpn1YgrdwIsb9hT5i7u SceP8m2lGwFuToiJ9A8BtsQZcQhZKrntkYGcb93iKkIFVFtQD22gK7RA3tHDSiwPgUdUduDx2xLQeUcb4SAzTtnV9NAhNiHUM2AgCEzBraggyXJZ2+hvHPAwsQHguB1IkNt6LHZfbxkWTdxK9A2ANxGFwDOAEBoFxwFGPgW9vhlBwFXHUgkiII/C0Tgc2PJQJkwgTylXEXw/IHmePbUYB5mD0IBK0WSuIeDgl2vMXAXwy+9a2DHth8fznu4kEISw AJ7b2y3AuA0FADg8EIM8eY9A2AHAOM8A4I+o+VHH9OGvJQ6+bBmNnuAy4YA21vaQc82AbZfGYH5BcuB7hQFnofkWx+3oT5EJT6J4QBuZWy2wOxOwuxuwexew+xUKj6V6AHTgX4SAPBPCvAfDfB/CAjAhgiQi/4V4n7jLG6QBLo8LtboT8Ibqahbo7rm6TJSLbZ8zV4ko9ZH7QCsZTZSxRwtJUoMrNAzYVD3xzBbyqzLbrCra4DzAbb6ywbMENYrD 17vD4AwAHjkCXhXb+yKr0gfYvaPZrzPb3avaBy6qfapw/btZ/ZmqerozXoFyg42q84kzlSmTlRFhtgTD1y6j7xDRTBwqLDUy1JY5hr9xRrDygSxok5k4zwU4prU5pqoAaiowTAHyo7GTNR1KFonxvplqPwEqOTtBbydQzT1oISNp/zkgS5S4y5y4K5K4q5q4Gga6tBa5OI65togKdrgI9pQL9owKDrDr7rkgk5W4jpTq4Izr4Jzou4LrkErpUHrq c6TDCKEQMGNZME8zHowKnrnqXqkjXo1TZD3q7SqHcBc4kKsYSBJZiapb2xoAHhihCDRCCaghsDrC0SoA/DWCoAsZ8iSC/Flj6BhD4CMDMCyaoCUariSAiKoDvAIC8aCaIiIC4C4jKJMiUBhZvoQA3EpYSYPFPEvGoBvEfHgnfHCZ/G4aAkZAglgkQlQkwmERwkImuhIkololsRaavqWZ6abDBDnCUrkiNBmbuA8n6ZYY2QOanrOakBDGbKZykDBY cChZXHoC4niZIYElsTPHKCvHvFBBkk/GUkAnQk0lBB0nZYMmwnwmImoDIkhAcnsTlaVbVb7FoB1aKHbqrDNatZvodbxBdbn7sF/5j5vTcFjaDSdThlMoKxjDozMzvxYQSF8pbDxCyHCpbE7bSggGe7gje6kC+7+64iB48DB7zCh7h5aEKpva3Yhzkj2gPaJFboNkmHvZ3YjFfZW4ZyQBZw2EWrSgOEWJOFnxmSQzNQqxTSNSdyHylCI4qg/ImitD orqhtQYpZHGExERr44RGE5RHgSbm45JqU6pqrznEXyuQ1j9iNL2oDgajZEs6cjjBoyNQLBdAXEUg85Rw7zlTjA2hlGi6VGlAdHAIdpgLdqQJV4Lqm5DobKTqlCjGWFxK9aYJx6E6O7THO6ThAEAzl5kJcHAEpStDnhvH2zngHAzBV5gDsL9GcKITLqVqLECKaiTbrQiLrGQAuiHqZl0T1knRRLKJxRzRqLJR7aCQHa6InYGLnbGL+KVJoDpiLAdD GgNI4QLAdCmitIuKciJBTROQGRwz1KmgXHeSBLBJvriL24YCRJKKckTr4SbQzJ7TzLmXDFYgnSrK7JXTsVWXEAeXPT7Ix67DHIzFYVnIAxFwlBXI0W3LnIPLnkLC4zXlUy3nGVgDdSoy1gZFvlYolALpBBcwKGBlFDBmJwUomajY0rOrRlCHVg8B84+ptTJlSEGjplbbcWekgFEUkVkUUVVk3a6HtluVqqnlRwvb9XoBmEpyGqWHdkQC9k5yA72F DZDnSi2p1JDRo7dQ+pwo0zSiI79KXw8CC7qiKjvzrm8VUgHl47RqRFjzRGhGxHJpU7Sg06GGuSozqjKitReidQWjupSC+mkjvmdiVrMU1w2JJnSgi4VEu5zTAV67dHgVG7RWwiDGwVILEBjGuUO6TFO4wRzF0UUGroYRLGbqrG7rjEbFcUyI8Um67E1YHEnrHEPpnGlq7BlUSBfAHAAjgioC3r2wZKggAD86AGJzGqpoB3NvN/NgtItXJFCYpmwh mJwboI2IpFmum1mJAdmBOVR0p3pspKUuZ+ZhZskAeQeIeYegWip/gKp4WnNUtfNAtwtotA5FWbAVWrAbpmGpA9WqxPpOR7SnWrB3Wf4HB/Wg2IO5VjKYwBY1VsZUc3irc6o/1msUhnQrV8h7Vu2Eg9sXw9wNszw7wXwXwRgG454iIl6BIzANsAI+g7wBIcq2hNZA1dZQ1ccBhGqxh41ScbdkABq3234xq0o81AO7WQOA5y18lw5qA7UQ0/Y6MR1b 5jUxY3hXq+akMjkjUzkS5tS8O7doaCaUY25utpQROcapOD1EauA8QxAMwMYJ5T2iwqMKobU/Y75zObWW15aBKNYNYk0f5UN38MNoVza84raIF+uPREFpB1UaN3lCFQ9glvUvW7QllOCawUxPEYuxVOFo+ZVVCIBPw9wmg54PAQgPAqYyBteBFStKeaeGeWeOeeefghexepeCBHBSBp+k+OFIBcAnQWSzwW4Ns+AAw/w34KuGSpDygQShBAVABp+s D8xDFa6TFKxrFaxdlB60iaAoqRVpKHBBD/BFV42uY8dAm5xnhO8Jo16ad+yvV2Zm2WdNNHVKUxDpD5DlDTd1Zphehxhndo13dOhE1/jHZFhSDVhI9pqC149S1Ud09q13Abc89uYrYS9+8jOCODcBkm9dOO9GOXhG5V9uOJ9Mad1+5JT7oN9d9D9CRI1dq/Yl8+8ZoWYkw0K95X9E99Zn57W9iM0mEeRwuwDv8sN/84DuuXRYFhufRuVA6cCMFCDY 6lhlNONmDeN86uwKjYwjFNBGjkybF2jjB1Nej2x5IRxZ6DNidTNUAJxj65x7NEtdxaAiIQQxGiGhWVWoQnATAlGB4oI4JeAwmhA+gcANQjAqAgQ3zHAMUKGmg8IAoRW9w4IlG3CoQsG4JmgMAqAzAcCaGbEkLyEzQBA4QsmlGyEJwoL/62W1gzASkpAqAC49W2IzQuQYtWJmwzzdpbzzgHzxaqAXzggbIpAfzALOGPxILYLbAELULQrsLBUCLsAS LvNaLYQOQqAWLOLeLBEyghLZmnAJL4JNLcyVLEJtL9LjLzL+rHAbLPW3JmtStSUxmat5m+AitEgNmOtiYjmUQBtSSedBdRdJdZdFdVdNdddDd1tSpdt2JXLrz+A7zuAnzIQQrvzHA/zgLEroL4LgmsrnA8r8LRASriWKryE6L6rmruLoLOrerxLwQRr2GFLGQcA1LjbNrFrTL4Q1rtr5ISbLpXttWvtnpTWLWgdnI2gwdVFbBYdIZle0Z5xkNQpA h8sljnIH8EwOYaoTV+yrQmdFBoqOdSe9D6eme2eue+ebD8QJePjPdk1ATTZY1ITvdeq4T01kTs1o9th3T+cU9koM96MtYl8cwUw40Go+8dce1OTiobiGoJonSnc3QxTR9W54Rp9AYe5uCV1NT995S9TT2cQ+8lo3U401irUb8W6n9fpDUm8XoVMFoOE4wHkP9laBoW8oHgDwzDaozoDQFEznRoFBuvRkF8zZuRz8FyzkTSFR+KFIdEx6zGFODbu+ D+FfDKUPImgoI7wfkAIB4lF1Fcz8EhNCxfCU02Elo759BYnHFmxrjyyjE/Ftl1uqiTu/r+dhdxdpd5dldhI4b9djdslZidUho3iDYyoRYRY3U+YmlA0vA9OZk3iO8qs0w7c/iC0QSn0K0llEUDnMScSwlSS+2OiR2eip2hiF2FSgXFiyQnUr8hYHkxoRoOa3ZlkWlvA6YfYnUysulEX515zaXZlmXqFkyDlBgsy+0GXVnFI7lOy/lk32yj0s3/+4 sQVn0ODdyEVgMIKYV04zgBHU03UFo5oHQZHBasVnONHr8xHDHO8AZ1ysD+V+K7VBjJVeF4W872l161KNVZ8W8diUwW69jos4Ie7Chh7c1cAGnWnOnN7T7d7F1HdD7wTLdoTg1/dnZM11hsTdhwO1qSTbQ6oCQaouMF5hYpRkHuoDUF829U5jHcMNizZsch908KH+O5TxOlTyHuO2HdTL1iR71l8hlr9P15oHyTOgNUcwNvT3UDiB87K/5IDhMEA8 NUzgnMDKNcDCzcpcFaPluKz2NEAGDeC2Dmzhn3CxnJN6jdBhze68p1nJzhKWZ5z9N3tUwNzdzrNSRjz9t6AXNPNTtstrtIxmJEtvv0tztctXvp0DrEgytzrJj6tbr0f6Anr9m0oPrMpSSyeqeJ7TD57rDReV7HDI9NtIW+AHLDtfvMtLtiYfbHtrpg7ftmjAdD57WE7t3U7odpQiBgVJjMdkZs1X3Cd7W6KVoS5DY27os7wIP2d2ZKUAjQjIjYjE jzAUjMjcjfVsPYTB9hIiP8P+It7W/Ovb76cmPY92Pk9CTf7ePw/1HiwW81oDYKoNiLF2TuopoE7EO+Y3UZkio7kIRnP11Hctggw7xpmeXPW+jh0fpvVMwaMIyEZE1AZoLQjkTpm+iGjKwcIpoBYMKAxRvwMIzHUkJaDGj/ct00Nbjor2V4CdoGyNAzuc3gaTdEGP4G3H/hk4d85ORvAhIBVwYoMjGKnXqCASED2wKA8kA4HAHWxsJlGRnVRhhFM5 cpdqBzLRjb217bobOpzR3m5QUQ5cVEWyFzilADbudg2XnMNrXT84VdzE8lWAYvVbi1cqY7kRqNF1JAXwmoO8cYAlypi5hg0gyfrhN0UFZdrK0URzqs23QjcdoTldLiMkUF2dToM3PZHNzWB+UYhS3Q5OSGCqYVCY63S5FtxioAxxgGVOAYIkQEeREOCKNAa5HXRYCZoHkWGDlTAB5U8U3MVxs9xnalVeBvQZdpVXqQWNZsaoUnpmEaQT9KgPwafr Z1n6bABBQgngCIPWwb9kez7RMC2UCZGE9+rZWsi+3E4RMT+0Tf7F+3ia49yQtqdFGgIWAeQD47TbGPvQ9Q5NuohoYyNWhsZLkxC//MAYALQ4G8QBl9AAdz1w688GmdhBIPvFfhKgfUmocYELlKCUcga+A7So1CUqWhQRlkEZjgyV58dIGiNGZsJwGKa90aFuZBJE0CGG8sGHA2Yls0kE7M1GezK3goMCGcVdGDvWmhpGyB7Er0bvFmk+kj6bBUAf kGoJoEghC1GMwfb3hAA5FcieR8tKPlZkdZGZVa8fV1u62T7a1U+etJzH6zn6CM4AwjURuI3HAr8jA0jc8LI3gLF9o2ZfCWkKLsAijnSdfAdtwA9L+1R2Lff0o0K77h0EAA2Qcu91pTnCGAbQ77uPRA5wcVQ/QrYAfkFRyF92ZzDWEbQ3DKArQkgUYOeH0AbgIwygeYIqUwA+RZQ0wvxqjym7DV8Oj7GYXDzWHH9h65IT9v2V7a/swcYwZmIkC6Ak 1GkTUA+N4jXo6RGo2gMQhimS7ZgbERQ7fldTKa3V2emHKpp8KgFmNM06laFIqDbAvkUB3AV+O2KwjlQ3IZkL0KLx6YEobE0wesJhHl5kC4ayIhGtMyE6wNoKWvDGlbik5/40GQ3M+uhWN5jMaGeDHgW91oYSAtwWkDJKQEwBbhZQ4g9Xts14QW9yR5NbytSMKqycuBkeEMsYyXamMJeA/b0UPx/62IgUgYi2EMNUF0jp86AD8cwC/E/iMxMCa7Jv 2zHzDd+2/A/tmIHpdlT+2wnHo4Wv6uQugHSGuCUQDQYQZyFw3UKP0NCsdjI/YLcdaCLCPCcczwtnhfSw4QCee5IV6qSCMjtjFQJHXoa+V66QBwRpaa9CDXGwQ5YRjVIBlx0REUCoGSNWZjUJE6LN6BEnCcHiPvGEjJwBNM3lIOoLLEKRFNfXuBJn5O8GRVzNrsyNOKsi7WAojJBwFUJFY82MLAllEGYC4gAA3JCQ4DPAsgTAAgBeiJYQZ9giyaLM 8DYDvE+R4tYKaFOxYQYIpsLaKXFISlJThWqU1PNa1QCZSqM/6HKXlMj6yiDMTrKUXBIT5tSU+Lw9PsqI9xRiYxcYhMUmJTE+R0xUbW2saMKlhSSpKbfNlFNCAVTKMVUlKUQFqnNB6p4JRqcVlymrDIAtfT2r5JtFN87RX9Nvo6OglEEwyvfHgqgH/qdCMEpQxpGRzhF7AVs+yW9JhNpFuNNgDeJvC3jbwd4u8PePvAPiHww8Cxh/HMQjwaYM8qQV EvuqlHR7vs6J5Yn9pfyrGcgAUQHHCMKCmi1p1QLY+ns3FXQJcpohHUSWEVZ6DjJJI46SV8Nkl88cIH/erh0C6BYRZg/1DSa3zmAYQN0qlV+McMhHtZX6jUM0NaD3FGTDxKvKgWZKgp0CIh2IzGohSYEYJ0GdkkKq7inzKdXxqnTYKCA4Dgh5g9sbAO8DTL/iaBpQQCZQTJHLEX+jWa3lSJUG/TIhmg5BtoIwoFcxKRXY7PojOxGJLsgyOSpKDb6s dGY1oZSvWHND2DAcbfBbPzOaTKxexJuLweEIsq3iEofg6JGBOCFjdnK3lbECsmiFeVYhvlUuS9B77JDVunA9IeciirWySgdyUoWzKpgczEu3Mz5LpH5mFgrQQs2YPUmqG1CCqT3SCYYxgktCvR8E9rHeTukrtZsIIy0EUVwjoSoQIYjMsMPFQpQjZJss2RbKhlZjkZ5E+GfmKPkHSUZ6wksZnBiZn9v2h0ysf+zahoD4uWEFWDhAg6v8vUb5BID0 jsSAiugIkpDk8IHG7kKmw4j4YzLHFoAYBqRDrv2CbGfwxeY7XgJLwKJFF+ES5d6aQJlmAJ+OJktEaeKVmBCGB3lfERsyJGm96KpI4Ca5NAmTdPJW8umj5Jd6HE70LIh5kFOxKcjyIKCdlhLR4ViAh6rUpPu1MlHlVupoi3qd631rCskkAM5vK3nbyd5u8vefvIQEHzD5uQJfZUjNO4U1AhFPbfOO7WOne1Tp8g5vhdMnY14J5icCOm6Pnk0p+kT0 2OpLIahHV0Jy4H6QexGESBngq4VcKQHoA2xMA99A0DAHOCtBMAzAA8AcFIAcAjA9wQ+W2WPmM8d+p8pHufPMLFjfsmwvsotQYkrU9hYwQefZEaSrl34KsWaojlXEZgF63Y2EYqCKZLD+xqHCSfdQAFHl4i3wp7EUSGimhvkDMN+Ux2QUt80J+RStFMDMj1I7E0szgUiLwUojjxavJuRr1E7KzwmOIxgdwOvGazcaCnTgUpxfGCk+BKUHgEYD4zzA jAnQQUNQyDIGzriiIVoPQHoT2wCQr4HgD8CSBsB1CAwdoM4C3D6BgenDWdsQUgn3LTlmwLcD8A4C3oBgXwZSH5HeAZJnASQKAL7j8j4AZwq4E5U0IUbiwrZ5kqhUTV2Z0LNG7k23soPt76Nx5L3clFPOpTVh75086bEPxA6/9LQHi7Mp9NFhaKnGoY0Hr4vQDnLLl1ywUJmJSUXyT5eYzJZKuyWD0NhpY2+fRIv67DSg+w+pOeVqROpoRGAlsYWE zTzYYUDHXsM0r7FVNQFwA8BaALEldLnqzMhpszFRjFhTQDMfeDNCyZgjxebXLSVLzGhmh/62ChEfMuMmoiTx6vM8ViM2WqzcR+vchQcsoW0UnJNClyWTXJVgTXZPi7yZc1YX+T7mbNLhZsF0ioB/mqAYIFUGEz3gDAmGMsKgEcB9AEAlGcqeCSrX6AUMZaqoLqXqmSi2Q1LV9JRj0CkA4AWkSjHEBLUIBGgRJMFggFcBpiWWbESjKVIJYKsi22Lb LLqWqknBF16U3RaS2sAYYcQCmGtYJlLWcAm1QJVAMmHUDtr7YAIL4AMFvQZJ7YPwAEM4EDBhBiAi6hABQCqAXymM5fdAMWtLXlr4l1EGoG2vUCCZ61DKQTM2rA3VrcAHa0gF2qChMBe1mGV9DhldDDrwSo67QOOsnVdrp1s6/YGZgJZLrdWK6xFuuuSnkATgtbA1vWxQxrAfaR6yDSWtBBFZINbaq9QCUQ23r71j659a+vfUyZAgP6/4KKLamx9O pDQUzDKKkXyi+psilzClH8WBLgloS84OEsiXRLYl8SxJVNNL4AaIAQGzjSBsrXgbj1dawgA2swzLSW1VmxDRtGQ2CZUN8S2DBhooRYah1WkBKWOoPATqUpRGwICRvnW6sKNcLRVmuuwwbqUp9GujXWz3UsbD1qwazaeuEzcbL1agPjQKzvUPqn1L6t9eiwwzibf1NfExfX2tFDtbR3qh0bSrxUSAQWLxaOvdLLguK0IAuQSVu25WSF9kskbxeGKU KbA4AuIZ4AuCMBjaCQC4GYKuBUgDAFtfkOADMD8gbgM6EqlYXMLSULCEZWqUicjJokY88lWPZlYOUSbFKo4G9J1RLPrAuRkB5PL1CaAviZhiwT2syILiAUtKSmh6m1ryECDLQ6ZHSp4XaugXj1xoh1f5K5GmDdQlsoytrGrAnb9gZoVaH1NaB60bjK0BYIsDcLjoGTyi+48ZosqPGq9qBRK2gZiKWa69Y1lK+NQ+IcnEjk1QE1NbQXoUbKqaNIrN eoKgBmD0AiAfwU6WInBAkkDUMQPlG6jEBsAGoNuE/HvrlRCwCADUMQE0BWhfQg8VoJoDOFUw5h7gCoITGa6jAbwV03CtHlulwS++D0msB1oek7xlQHHHlH1tFhlZ+Vm8rCX9MeXPLXl7y/AJ8u+W/L/lgK4FYLt8Zyr9CFEmOIjP21/rUZiqm+VsIxkPysZ/7BpJDB7HPIzIn1f6nOVHI5gnIEwIgXMHMjAKxJlqs+m8Kuog68Ob1LCEBx9Qrzn6 /OHmXVo2qVC0mUy9+X/wmUHFfqkwZWNehwUhrZZlA0yeiPJ3rKSF1krQU6OYF7L5OdOnWc+Mnn6zIVEgSQCpA3A8BxI7QaunpwkGM67Z0grmQxy4lelKRHkzNUNqm4aCbKuXISjoM2DqaglISsJREqiUxK4lCSpJQFx53pgt4ilMfkqEaTZhFgxlZxDFyfL9NXBysTsc1DhEmVC5E6XwZFGv15zpko3UIQN0m7FzGI8QsuWzuzkVyFuCQ0MstyOS 1zHxzchuRkOuQExzkFoTNPFwmA4R69Ncd1ETAhhjR+c9/b6vxMxR3d1eD3eoa7qN16zcVLK+6bWHfKD9V2lBX8mZFY7vTAelQXThvLapMLhtK+tfRvq31GLvIzdLJaHoyVLCkZUeq+bkqVVx6ClqqxiRdvaxKh6cXQVdLpSOpjQSZ7UAXnfAbAYoTQu2pnsXraWA6OewOuIvatKBySYFmaBBaaGajFgsqjelBZ6oEC9Ma0bHOpP9X73kHZwg+ghe GtWWRrKdWyshVrNSGOTqFTO0mizvTUMLz9ag+kTmqZE7FmaAUzhS+gFEAhMA2rF4f+olptGOjUm0RTJokUKbxRHrJTTIqVFyKUo9wJ5S8reUfKvlPy0gH8oBVAqjNuikzT0erYvCjpVW90jVrOl1bLpDWqfTdJINm77p9qK3QZSYNKhU6PKyoKCEG01GcJi6OfAviXwr418G+LfDvj3zBig9xhrbbmMMI+HlhrdEwzkqibmH8lcTQpedvVXJNWOh odSvVwxzyGSZHkeyHDFbjuKpoHer7QAJL3odrV7woI09VB3Yx7IGEJ1I0owguG4db6dyBtXzBQHMwqlFWKLM5Q5g5gO8OZRkYWUQNid8skfSbmIX69SFns444BCzkG8ijinXWccrLwgENwKkNgDCvHADbCVJRklfbMESOyOKzss/dSov1YHFE/Oyfbim9kpRCuh2f2aV2krBzeopiHnaispP2JuhwI7xL91TktcYuERjwlvVci1h94zXeaKZW8GZ zwVPlD2XgaCGoGQhcyMIQskCEmmcDUaiJAQbWS4HiDSQ0oCkLW6UGG5mQigwijaiGg89DSn1LmCQUIpjQdkQynUnBRtQ6DBoIs2ADuQLAJ2S5K0AOCwE1w0dnyRk8kGZMqxWTzYFs3wdWUCGIJrAqCcbvQCwS5NM85yFbvRjPyii249CQMEePYSlTKptU8wAG0bawTgJuGTKqMOR75VtE47XfJ2HWH4TpaKmMih3Gcp6wNid8ojjVCoxWo0R7xI5 DzTA00lrS2mWAqHE2rycZJyvdWI6w/UQuRkLExR29UJGPyK6N+DbrEIkDg1fJ0NcstJ2KyKdVkqnTZLjWymTeSa0o/vuZ37MnZp+ylYwtd03NGRjNBo7cw4UFqWj2JJ0NkEVKaAKsrLfKSZs4s4hCAPF7tn0eGNJ4Opgx0UoptswKjxcKmw2qMNeOL5l8q+dfJvm3y7598qxmNpsEEvcXeLnAHQ1sEq1WjdjjfCxedL9KHGZztizYPYuWruj+zji n0crGfpNIAedxrYBkm3Nu70AoIA0MoBuWEB7YW8A4FuHoAZJzgFAOADOEkDvAV+ySzbQYbPOUSLzU1BVdfJ7LKr49WwR+dfzcvti4Y785wcpVTkQBEcmYXSLmFC4oxew9iamcfX8MgX6ZAAmMC6KVCg6Ed7UNUKjuxg1XyrvM5mL6pXQwwvE/3Xkzx0yNE65Zw+ohfhZjPimnOjWu1DPvYHazsKOy5oUvu3n7J5w9sC2biFuU8NNru1iQDAB+BGB zwXwTQJgCEBaRZIXwW9AeBmAzh9AxSZ4PQHkYm6CVSjACSSLKOW9WdLso07ImEMKn3R9SY/VIcXneIXB5oHCOhINHipnGYYp4yAVRIzADr8QI60lePMpXgTZ8kPRbghMfscrlhisYnoKtTBm4UwDBVDj7MtiiwmaGqzHMcj1W9TsM3wzTJuotWgdYk9q3GF3aQW2Uj5qtDaA5Ui9C9R8b1S1FFlIC6uKoY0JNfIFZGw1KysnSKYWvj7CLhR/ZXPq 1Pm9KLbkjNaDewkXNGL1zZi+70CnsW9Lp6Ay92zQCUZi1cW8gBtKRISYISFmyFgtMim6tW1qLWzTBtQATxKQQgAbCICyyYa0WeGgDFUAIDBB8AWratvK3wDKAgs6gNtZRiGjdrY4xACO4JkC5YY21mQKICRjjg2sUM1LMsJRhNPMaMMEUcDOcFQAhAqSBWfltlhqgIhagx60Vt7ZCCgb1AyEeqf+h/TCB8AGGTQIJjdtEtl12661iSwbuXqywwLM eypknuoAQM+IdtdOscADZmglGUTRhggzD3/0iWxjQgG0CUYL4W9iaQS2eiUZEgWUlwPoBgA8hzAZatgN+ASmZoYsQGPjJi1gwq1+LEtfS8JcMs2tnbHAV27Rpqme20s2WH25FtbU2a7NYdurJHcCAHqY7ZbOO39sTtBAU7+LSjenczuSBs7HAXO0FBqAF2wMxd6tWXdwAV38QVdke+xvrv7rW7LGojK3f+I1qvsrbaiM6BdAUB0tnGpB4Pcy14Z/ 0agHFhvansz24HZGyjQxt3XL3dka9+RxPYwzb3BMiGve+YDqnH2uNMjtRyS3w0338Nu0NMQ/fWRP38NVGV++/bgCf2XQP9yjH/Z4zAYgHcyWTZcQVr9HJLLraS+JYgDSKpS4x1TYbMCvBXQr8QcK5FeiuxX4riV7RUaIEsO2IHTthKbA+qke27SXtyRxWt9vQtYWqD6DSwEEwYPfaWD6O95tjscAx1BDhEEQ6rYkOUMZDxUlnYSnUOuL7xQuzhi0 gl2t7sGZh1hlYfgl2Htazhyxqbu8O27AJDu9IAhLd3RH4jgeyU7Puj3tHr4RR6gFnsqP21F99R1w80c7Px7ezre6iX0cAZsHRjraSY9PtmOTnFjhKbfZsewtH7HAZ+04+cBv2P72AL+x444BePYsPjjVsA7j69tTLJ0vY5ZYOPWLp2kpn6zmdaEzzFg70mG8IWsGcyAxvWlMp/l8tg8LrV1m63dYetPWXrb1j619aPMo9UlQJruueehnUTo9WV8H hYZhNWGil954fkuQF6ZNMwmoVqEhcRxGRnVZoDoPUkVBdAzIjVlnrzatWgWSTAt84B1eFs9LDCryD/mP3rBVWzhx+3mQ6mANkxf9fc6GPLdfNtR5saRzC1Nf5OTMh9hCiNaKcpVLWrxGs6U7Tvsnz6trr3XFRGM2CSBsA9CTQFAFXBY2d9/1vfcTUwhmd2TlRmM3RbdnyJ7OyBiUxaZ4hJIArQVmACFbCsRWorMVuKwlftimDtI6YGaAWDp5OQGw rkOAXHPawXwYDHYpch5BVDo6+uYZjOWEmlPZcM3yb/OegfDP68Uzlc7yvN0zNVzTduZsg1NfrnhVWzdyHVw/y9D6vlYhr7uQkFNcmhzXG6LCMPI5ijyGhRx66QG9a0Rk2uno7F9WPsT7wOUdjLy0hmJdCqpAob8N5G/Lf0vZhBN5l2ldZcHb2XZh2PdCfP6U21VxcMYHVSrcYp4BO8at1TBbHYRL4cMBG3DEaSUyFXpTZq8q9atPDBbnVkW3ah9S bxXV3h3GB/W9VDNucK6EXuykBQq2DxM151zkc1uo1tbYpifZNx9fazDbzk8o1Rf1M0WlBKbznbUctt+TrbrFz3oWokBtHAMILAhAK3CCvh1W6mUrOSxY0zIjLWQMIJBkD1B8Cp2JBT0wCU/QQVPzANT+CQ0/0Zl7Onm1np8EzIsxLvJGPsE+lGhO3Pco2S8pqieKXzrl1667dfusHmqXr196zAE+s6W9FmwUz4qWwaWfrPRWErHZ64cOewgjngzx Vv7ZwuLL1FyxdZaRed9z3TWkDLqUvc0oV5K5lUJaDz00eVgz7ovijYFVeSg3EgJIASAdD6AhA4kUEMQEwAcBVwiIZcKCBtjK6DwnQb9/8fSv3tDDAH/QyTcysgfsrXL8D5jMg8ygxgr2tGO0zR3dDV6D25wC9vbGXdYc0wE0MfpbIHkftzAP7UlHaWBG1XGrrq3VUhjWguo/+mxO9N5llxN6OaTGJdwxTy26koXDoIu1KDpGHX2FknQrIsnniVZW NGnSRcTU2yAbFFwTybaqNm3PSxc7nROAgB87c5WhSyegHqTnBRgV0bAHGHaDEBjQCABw+cGR2dQZgLo+UAaFwDlQESeAb5NrpJacD9dhus93OezOVfkmzMK3bCMmDS/bjDuyoKuFfdnXcJMKuFQiqSBIqUVaKjFVipxV42GXUq7bWHq5ugn9fl5o7VCZO23neXUHtdqOR6sjn80RRd8w3HcUZhn5/SE0BOSmjYfxJARiBQR/VdC3QdiQAWZMEagr kFsEv+k+NiKJ6RswDiQM7V9FlTBNQW4qs0x8J0CnZrLr3I266UEev1ZUpyM7x9SGnWUX85loe1/fS4ADQCAZQIiHth/i/rqy22XG5kFH7gbhpjncab4qDu8ud+vxQEsf1aadNr+/TR/ordVIEgp1ADjuMKIk0m3XoYaBwbOHGR9X92x0+nKTN9vIzA7s05m/iSWnNgAwPyBTHeA2xcANsFSOeFXDxAYAGu8ELJABAZINwk/3UBCkEkNubunKdwov 70jCgAAwAM8JpgVLh7dt/MZEjMpkLaDQMEzDAxjNx3Qg1wNAhKd08oZ3U4zncTkOuQLMl3ag1BQEUEP0LAw/JqE/NWwKPwBg+lOPye0A0KHV4NsUe7jqFpzGxTpUK/HazEMr3WsH+pb3TkFY4x+ZwVl9CXV/xUMXGeizfcVIGvzr8G/IiT649DYmyWEdtIm2SslvK8wt8bzWEyv4bDcYHf5GkIM3v5oRYHyO8poN3ymBFQL0DLh1QK70AsLVXD1L 1iTK6kI9NXB1SewD4TsyVBFgNsBtAZxI10Qs0FStDhg7+RSjtdDJAfRY9sjDWzwsx9Lj11sePFH3p1iVI20x9O/Wi2qNzbZ3nqNs1G22aMu+CWgIAM7bpwodA+eCn5FsSHIPIc21Vz30wBjEJw1ownCJzT4FLJJGhVYVeFURVkVVFXRUMkTFWxVRDILGmkTNEoLyCygi0VMUG+Ydm9IrLIOnb4mAlaz2ByvF4UZVBoO3XRdWVaQ0KIFgLGE8s5fL YE0JndVQxEClfCABP8z/C/yv8b/O/wf8n/F/z19f3Ob1Stw9PbUA9wTZb0hNQPS3zUDsZZt2zA/hamF6Q6bOtAe0N4D7xHMThA+EXpsPW73u8AdPmye8PQewPJNfUTnD7ABwJpUWCAaFBQ3RkUcjmmA5XaYHKttJUtFMgzQXeAz8wGEIPVtcLeHzTMdeAo2iD9bX1zL91DZXyaC1fDXzaDtfLoO+tEhaNxb90fONwSCk3EG278njPHx51CfCKHRJ BdUnwgAsIGsAZ9RgO+niBsAOHFmBB8BAGzAN0ZSAQAeAYgDagkgJXR9R0UF4WYAddAXwsgDdOgOF9u+WdyWD7pUyBq8iwTIj/p0Jfzh2DhA1N32Cr8UYBvw78B/CfwX8N/A/wv8H/B/dCxO4MbJ5vUMJN9rg19ieCybNb1O18rGwxtAFQLMBXpsYS0DwEjvCgIxRmwJ3wwVY5IvR5sgBGwJVc7AwPyI8tXeSRfoD4Vwg6B/6K0DnFS0UcnzAlbWr yRDMcTvUTo+cVjhBFiQ3jlJCcLOHwxEIg91249lrcvxYEz8GU1pCNrFAnlNF9QNwZCIAdoDWAa/H4HBBw8TUwZ1yLNv0P06vRINE9kg3H1799/McKzdEkK019kbTErikog5N/zDl+ZNULfkMIfNBGVeoNpBgUMwX6nsQCwRqAJlzGTwTACXKCAJnN8DaM378j/CQBYwAQdoFohEQfABf9ZIA4CIplALcAGACQcSERAZwO8PqgiiWVzhhGwJUHag1 JH03aQgAsiK61YDUM3gMfBaUygDHKWANHdKVBAOndJ3OIQndUXd6Hnc0hbAL+hl3WKmElTvbMJSY6w6OBKAOcZuCI5swI6inIbxFGjbMzuYsE7N4BLeAbFVQUSLAA1QOyGbC2wVsLq48YCc3Y94kE9yEMLQiG0cUdmG9yQlpDddCXJfEBQ2fdKoIQLRsdzFKGXCkIA0DXDKyGbweCTzMMNuDjfAEwytlAl4NUCeXOExt8SPBchrBvEGA1ah7+ZDz rBtAgsGO50w/6mu8rA4Czw9+bGELLCHA0Iz55lYSGCUkP4TnAbDUFUWT/obEC0Fx1OOfHVwUs/VjzCCKQ/IxjUiLZH2nDijLcO1NaFNNXkEKVA8Jx8GLXyVd5pPJozYssggUWhJAgV4iPVkQBADkBQHSaMkBpo4klmiTgBaJEUwnSoM89qg7z3CdRjSJ19YJjS/Gvxb8e/Efxn8V/HfxP8b/AFRDRXoIlopop+FWi0tOaI2i3aXLzMV4XAr3GDx2 YrwhVSvPrBdFI6OS2tCr3fGRXNmodyBF4/g8VGfd6ERXyr8IAKCJgjmAOCIQikIvUVQj0IzCKuCQw433kDZVRQOjDgo1bzA94wqmw0D3INGDX8pIk6lrAyeL+Q7hguNqAAMEqDxB99CTV4VsCqmCvQrD00exHshEqb8zptPtL1RQUGOYLjcEMBAsHRg55DHTGBd4I6hTkg1IIKws1bAcOFMOPYcPz9Rwz1ztxvXGIL9dpghc32DOgO9URBFwc8GO ss5KCSRiPQr0POjfQq6IDDbojkNF87lO2MXCYAfQEwB4gcEEwACQfQHthb0JIEkBnAXAHSQvgUFneBlDOcJOMqAT2NQI+sSbXPADwAkFvRvwKAC7MYI08C+B5gMsHdjuGW2OTiIAKrHPB6AKAABBpgBcDjF5gGcFGBcQRED8wcAIwCLiwVKil31tw0lV6jqLfqLIgcfcG3nCxfNoAHAavfpk5lYY+3UJcHTDWFRtBVM2ItirY8VW8jFvOQKN8WyQ KKUDzfEKJVUIPO8wijmJGmNq9MXRyAZjj9T1HEic9BqHvgmDL0E5jrAokxLCqmWEOI9VYBIFcDBJNHHfoyohqBGtfA6aErgeTPHQAp1Y/sNh8tYtZVJ8dbakJjMS/fGi6j4goG35Cu/BeOYU6jJi3SCZPd8g5pANfDT8gE7Vp2TtXnYIAOdlHZoDQBDnWFgZYSE0ll4dggNiGvUoAeKUoxVgMIEYgy1KRxkwAAHhNMiMZZygAAAPmXtuEiKH4SVM YtGESJnPTzJYOAEDFIBcQPhypJaEgAHJwSWZ0bs1gOO1BB77XVg+hsQZwFoTl7UTTudTgB5yMt4pVLWExENT5zYhnAZ6C9QUMOuzwwe7MR1BADwPyDucBnfkGXVL1X0jrV5CPjFWAvnXux6927WtX0S4MBeyS1FkV9BYcsgUxP3sUIL+1UB1WS503tp7btUHUPoOjVOAB1AwHu8D1bDHY0TEjaG1IiSfQH1Jk7QxwGwZMF0FSTmAHO3w0skMzF2h LAcjR3Ul7KhM4B4pHkFaSL0MqVrUjE7pI4BKMAhzLsGUAAAoTTBQCbtpEthxY1S1JSByJmAAAEoISVYGwBZWWDVrtQpQgHxAiAX9HeJm7cxyY1tnPRxbUKsKOy81/0SpMCAvEwMG5Er7RaOxJi1AhPdsk7U5KUdN1ChLISfk8jS+SbPFuwYTlAJhJYTIpdhP/QLNHhL4SBSCROkBhErh1ES1gcRMEcpEyuwbZrnBRKUSASVRPUSToZeybttE3RIA xv7ODCMSuHExJqTu2SxPIA0tGxN0T7E9ZEcTENH9ETte7dxM8SpWOh3lY/EnIgCSVaXaH9tUAX9DEcwkpZwiSyUwxM6TSEtkAoR4k4TGpS6ktgAaTdnDJLc0WUbDVdB6KLDVBYqgLLGKTa1UpMJIu1SpI+JEkox2VTVU3p2aSW2EFgGSOkxe1ISRk3pLtS2kwZNzYZU75Pi0tpcZMCTmAaZJOhZklLQxTl7JZN9I1kjZJZRtk6zSIADkwgCOST7F u1oSjWMxwuTqIK5PuSKULe1ZIHkgqGCB8NcoL5IESUQ2FIhjPaNqDFRI6OidIIhAGgjYI+CI3BEI5COxiMIrCPScHogUTeTCEz5OGTyEzgEoTyEgFJTSTkkFLBSEpNhNMxOEitRhSToVFNUxEUljWRTiAedMkTZMSZwhJ5ExRMWcvktRJxYCUrhyJSmnfDR0TbHPRKlSKUljSpT7nA+wsTWNelLvsz0plKugWU+R3ZS3EjxK8SeU3xOWT+WdYEFT gkh+1CSI7CVMExIk6VKdTBMOVLgx5ki1NqSMMepLUB1EhR0hcskrVNySMMPQD1TsHCEhKT0WDtXKTTUqpLgzkkxDPVYbU1ABaT7U9pIi0vUv5J9Sek8DDdSHU3VnY0+0n5Igw/UlWgDSZkuZNDSuHcNJWT1k7LE2SY09jTjSggBNP2kTkkdPOSbnS5O50s0zDTuSwMrkXzSr7AoMOlYXL6Py9hPQrwmCh47a1LS2hEpUkMrI2G1Ndl5fgKkIvIlr xd03QpGJYBzgIsDgB6Ab6WDCYZaVUJsiY/G23i0Za8z3iNvA+K28z4DUC+DJZGaA8RFgNemICEgZ8OhibQZ5Fh18TJ4WUhqAx7399bVYI2D8qef4QlccBNUDbhf4nwKFByoWEVKFew6awajQg8kKHDoEyINgTbJDqLlMHldADYBxIG2APBcQO9XoQ/gNbQJBqfQgAXAEAVcDIB24xRhIIY3buJ1MKjPqNNtBQlIJYU0g9BIyDxogJ2xJoSNin4Ul orbK4U2pfkmMzGIctPFIeQMY2rSAvGMx6DjNR6MZIiIIYJ2MfaXTJP19Mv6MmDkXQGNF93RJWBXMVYDUGcMIfRr02CCoRGMXD2szrO6yvgXrNIB+swbOGzRs/x3mgZA4mO35CYllzXiixGMPRkKbILOt8Qs1AG7EP4zMG8RQOHsU9FL4tUDd8mzEc3ch6wKMgLCowNLMDMMssC0epjyYjyXphoDyCwhukLmUGtvVBsHbF2oY0DNAiBFwU9E8Qvpj phac8fhASFeZj2qyyQwcNH16skcKiDTwucwnDQII2P48U1QT05tLOZN0PD3ZPv1v0IIvrGYBnMg0Fcz3MkOUq53/Q0FsZmTNuGTougc4RIjHyLNDRROoHCHzBncvUzgNEzICNWh+3HOQEoh3OMwLlA8ouWm5EAykJ8pUzfFTRdCfLiLwC/oKgzki7kHSAwgxyC7z8DULQsE+QmmbsPcJpgI0AJl5gPiIRQOc1qC5yjIVEw39pwBUHv5/zJ1DYlMw I9zugGAseVstmAzglYD5g2wy3QuAh6XSZaw73wJcpCbABBznjBABUhCAUgBUhZIOXDxjPMw33DCAo2bxJid4smNeCwo9QL5c4UZIDNd5QYMzkFZyJlR0o6qZ3LfMx+H30ZzmYZnNVdwLNnP5iHpHuR+oiwbGFhgPVKj3iMSs9NHC4qZWXIJ0SQhXM1j5rHWIvE9edqNn06Q2cNayy4i3PEhOgQgABAbYGcCMAFwW9BDd5gAYGcBcQAJTuz44xPK5 DDI1vx7jZsvuPmy0EiT2Gi2FRo3zVZPO20giCAIgqM8TNRDHwA2CiaLFE9og7Kktdok7MlI6g/zwR9SxHRV0sWCrgpy9LRPL1GC2QX6Nb5/o2c0tD0AsGJpR3tFc2BDAzGXLhigci+T1h7M8T2eN7YJApQK0CjAqwKcCvAoIKckFfLIk18/yM3jN8jHNJjOXcmKt9wovHIbAvzY/KXouzZ3zZRSPbxAmgGzH6gat6ciNHvydYP3xZyI0PmMcCq9R IEbFoYzUCJ5LyX+LZwQuJOidRH+T+Vo9QacpXi5EbYAvqinXGrKVytbSAsR81Zf1yjg1rAkT49EEgT3UZ9cg0ySDBotN1NNc5A/3y4jaOfIXyl8g6Q0g7csOWZh3tdUCrhCZOWMoj3w54LTlAIwbl39Q8gIQ8lh3BiN7cmImPJYjy5BPI4iVuTAIyNF3XiNwDtuGEG9QOsbGE7gRzQrOiz+IzND+z7EerkcgswOqkryduJAQnZvDEnPSKp48GEJ4 AA6KKJ5KeYFAMiR5R7lPce8k2IZUTMqOGagrdTAT5w+hCfP2QXhQwt2CHMxcIvQjAcSHoQkgYpHsLGXU828y0c2QNcLt89wt3z943HNtRyoOIHXcZoPpGJot0T1H4R2xdKleR7+YyFmp0ogAWiLH88vWyziPR80iNmoNUIZxf8sZX/z2sDwNq9TOSrMdd8FRXMgS8jAi0aziLZrMOU3xdAAPAtwAkHuBlwL4BgAFwbFSQjOgfONaBiALcAyQMkQ8 2IKOIzcLiCWikCRQSOihbM9ILbOgrzUPeHBIlpAtCtReTNgf0viUi0iQH4KqgxPjCcJSM7Iz448q7LWM/SqRxkLhg6rSeyR2RFzeySvEXznZzIwaBzAESpsFSNlbFEtFgmZOeNa81DZ411L9Sw0uNLTS00AtKrSm0rtLpA4PWRyIw1HIW8ySo/kxyAs3KzO198iKKvJrhN8jdRHDcqyfQxoMs0zBzQJMLY4QTK6n5LYip/NZzulRIpKUCOfSGLA/ qF5AzCZbeIyGhgzJ7Q8hzQTMA0oOw2enGg5XdJgVKYfIUwgKVc3WLVz9Y+osNjNS1HzIIeQigqERXSgaPdLjck8PAjs3FKGxLcS/EpXjHTUORdNXpUH3JggDLwy4l3ch6TiyI5ZUCVBZeM0FADqIiMxAiozQdypF1i8bk2KlBZiNQDWIjMzIr9i0g0OKF3HiJhBG5dj3kiazNsRahfsncuzA9ynbmMhTvZSmmVTyppQ7ztBYyN+lDMzYGa0Kvd0R Mgrdd5BeRFQeyKByBdOzIxLjCkAgXBsgKyC3BxIZQHuAZwCSHEgjAZwHeBn8ZcCgAEYjzIcKmXIJlJL2yy+VJssc7l2pKvC21C+QzQDMC6BX4DGAaR9JL+WO9AOKVzGhMBPITSjLAznnBD9UyEKyjoQ1cpCNIAMIxv47INsHZQy8rENY4yo0gMNBvwxjk/yexeWwyI/zN1FvKNYiBIfKxCtYXVKYC9a1L94C5fXQAfYv2IDig4kOLDiI4qOJji44 hfQTjSCnXMBsXSubOx9/y+RHx8RtMUM5ISfArh0KKfbUJrh5gTQCKJsASJWNB5gbAHaBJdGX0V1p7OMAl0CwPnwlA9dU0KF8oS8v0+zcypIiNAVzcP1Lz0/EssqBUwJyJoLnjOqv9jA44ONDjw4yOPBBo4uAFjjCSg30srFhLspsrDtfzJUDAshPU299hN737lI5emNbAL4oGhfovDbs0u8ao81T5LoUB/OXLBSiC1fzqYVD1p5erWsDHjo/NlBN BSYOwjmAZoGZXliCiggU5RNIjCzVjofIqvvLXXTj1VzYEl8t4AGiihSmsjlYeMIYUoREEwBJADJABAZgAYC1xHSsi26j43WQXKsDcgUPuqTTMCNNzgKzYBRiG0jGJbS0IttLvCXTF8nHJ4YY0FR1RI0AyysA8uAJwrJwvfx6L1cvos2A1KqAA0qtKnSr0qDKoyptgTKsyttzv9ZIB7AWfYsDqQP89cWFxWuJf3MC/zPMHD9ks7t2wqd/XCroiYAo ivACSK7YsoqYzFAMW5jqmuRoruI8KnTym5Jio+LCwDMAmBuhX6jlj8i84qOoCOfplrDkq1jloDC6zPNxrirGxgJr2bT5A5QyaxsBMgqa9vlyp6A4SppVDqj7NNi2AyqiKyXLNlSxCfCp4vQk2iF0Oci/LCACFqRasWolqfq3yPSUnCtJS3it84Gt3j+yhML5cdIQDjsIcIN0zhQuoGLLiBewCbHcglxIsDpRIi3HCXKoQzLOfy1y/KJ+FyoSGBuM 34VCx9y4jKUtFl6kZ1GbzCq8BJZrc/NmqfLyqpQXgTSLNH1jdvyrH0NzOi7NUk8RorBLGimCngs2B3kwh2Ts1pE500zUoIoMIae0oh1Iad1chu0wgncRQjKepA6JELzspJEeqGql6uar3qz6u+qO067IFEiGohOKx+04TGTKHs8xR+iMy0SqoqzjK90KIVzDwJryVYY/UUMtgG2GnyQCUgFTj04zOOxAc4zmANB84wuPMqiSvyJJKAa3zIPqY9Hf NCjHKwcrxzuzY+Ohqz4wIrtRs8roHcE36SYFMgALS6iqZ36yKs/roq0HWZgCOOqkyZjatSiKJMiw0Gl8EuGuEF4zVRIxXQ3zcLle0oGsAuKrWa6oujVLxQv1fLi/I2PpCRfU2KRitwEN3aBQQGACgBdOKWpQbpsg/QTd5a9or/Kla48NtqgK88PVq601GPRim0zGJQida3GK/0guNHBzBgGpcTljLQRfywqo8hAxDykDQCtVqemvxURBYVcEAOAk gOlx9rK3ZpmAMt4Ssx3hoYWYta4L4NuE99XkFuBvg3cqiIWaaIyAMIq46rYpLlY88ir2LOQg4pnCshNPMLNTi35vOLwm9sUibfqSZr3c3c+5Gpj+JPMH9MwuWSMHr+DLvMhKpgo6vHqB8lWCxdzM4QlzABwNsCwV0JbgpWB54tr0XDKm5apqa6mrer/crKqxtN8goikrLFscsGuCznKjEyBQ34QImvIswNelwjCeS8nXNfcg1Tvz0amIo/q4iw8i FKcaq0HbF/hVqBu580JC1+9pS9GDaY3IBmrqjgg7JpgbDI1UsWtRwprNgKmip0t1zkEvqowaBqrBq9LRoxgt9KBRU9NhYWsLym2zsSB1oJYnWi+QYatojzy6ljszYErT5LUQpSgdG3EDTiM4rOMMa84guJeF4yyQv8sSUj1okazLR7PkKXspQszKAYspphKZ5ZpCt0QORjkHkn3IHPwAtGtTQ2bb0LZp2bqWm4MsaIw/evJLD6uxtBq8rSmNPq0i HdzsIRc8wL8QHtFqAzAmkK0BpzG6nkpCqQFR+O5jn4zpSlb1ymBVcrnwmxgltqwkBraxHzPd08RW8r0GZtRZVI37AkqrJvKLlSkqrjylrQ1sqqWsmqtShdG8NoMbuoXOOMbo28bN+tJs7kNQaZsoTxP1+4nRnurPS72hrA0YZKK6g+kYWVmoLmVbPwb1sshGWin4ZwA2gYAJgGJJP03jRU8DwDcBQwD7egDobAy8lCg6Z1WDvg7OU7LWvUqsFDrQ 6zMDDrqlQy9AHDKdoyMr2joyw6NjLvKWNti9sO6aJg7cAODoZYCOpDuI7UOvkDI7MO+7OTapGvTMUL6tUepF9xKuYNhLYuMzIqofRFwSIjbGNRufd9AUts2BzgPyBtg5AVoEkB4gehA4AXAdfFX0vgUECgAYI6tvXj185wp8iGWxtspL7GnHKcqcmRsSA5N2engRt32z1HGhUYVsFUbTy3wgsCAm0KvIBftcKtENz6bKNCbiPd6h+9vVWGFFlavB YDg5VYzVrATtWua1ybHyqAup1EGkpuqr9g8uMrjq4+IFrjzweuMbjm438QsBH2xOOb8yCr8rfb0GxWtJbL9IaokAifcUOkDJQ3ULaZcwA0Al11gUYE0AlyWarnhJgWs3mrLgbxDjApgZSHV0dq3XS8hBfc0Ik7VCpPIHzYOSXw8hZgOpBjrAcwly4A7q1rpAIiuquJri64huKbiW4mrrMbfq4kv/c62lwp7K3CplocqnOxxv2FjQNGHQFyYMaDD8 u3biQel3+HMFliW4XFql0RW9LMxreY6dp/qnsOxDRhTXQmVfhULMqMmhquaSLhsxCVqFyrKZI0Ep592pUvAKsu0qqpDWo80w1zuahNV5r7SlgIXDnjTAC+BYVToGUAIlLquaKU1dvz3DfygeMtb1BdNxWbnOM3ORi+mzWsGbtanGPbS9mqf3laXirBXl7+5aW1DqwDfpSpg4PTIgwEV6eZstr4662pWKb9YXrVqJATTu07mAXTv07DO3bgTSVIUz vM6S2sZqn9VW44QmwtqeAVma5oOYpDMhkB5qtr7KCPJHdiKyIU+bkAtiPea5GjAJ+bizP5pwCM82KjRQyzcqCwUuc3NEhai8jqBF5GS28neLzixHuNBAfJPql1vTMAAx6Y5P+IRt+cCvLBLj3CEpMi1usyPkbKqD3sb7XLFORxCAcj6SBy2AdTokAmelnrZ6yy3QzbLrGlHI3i9657tsreykGuPrW2iKL1A34B3LLqgUU8tl416RyDiBOkPpXC5t xEduC7Us0VoFLYe7Gpna7UVmTqpvQLw2eRlYYrPlsoYnUK5QiepZRybYGvJrKqKemkKNbOok1p6qyVc1pa7Kyn9uWzajMDrtbsSW9FdAJNDDAISf7Lh2NJ4O8AdEAuuwoOM9NgBAcgGORUICmcWNOAYZYEBnxMUqeC6TR9bFzSRRqDWGqtMY7c6BAArjzu0rsu7Kum7rbiBGhMoFE0B39QwGYB7AcWd4B10HwGk2uQtq0UFcTtRaPsnMsb7aqaG2 xa4yPsxtAGY9CWEKlK10JUqUoMxAoAkgDcHcQjseYFXB2gFhGYB4gc8HuBOgdeVXjuygmLH6I9Wzr8zbGhzubaBy94NUaxyFPyhwoUdsK/lFyWVo7hhc1PU5teS/fuh7xWlcviK4e2KsSIVIocwDRfcxaqnr9ylviKJM0KpSIjjIVFAqjMicaA8rH+wU0y6X+7LpqLJOQpq5q3yz/vPboSnayRiWsGAFfxmAc8AzoGmz8tfbmm2QX+oFa1BNa7la /Cv14basPO6b1EK00IBnAVoBUh7gXDCSBngdoAbw8AMN3BAOABcCd1IK0Yvqh/uRpBOa4PN02rMVe82vubde4CMnDE6+M2Tqg84PvYiM6sPp2Kvm6iqj6i6+ioLrGK9s0cgIh9nBBCsTf3JL6bECwSSGqogsEEquIZFrr6RBrNv7yZO++EhjsIZpEB7O+wlwXAe+9AAqGqhmocs7R+6zvH6rBmxo5c3u9bxZaaShwTvr4hhGB/5xyNeh29d6QEVu 0XtXwdHaxJIJuLD8PLLOP74e4E2mAU9XSldQ6kZwQQs/8jk1Vhf9fF1qjQEpmugbsh3Vrz8cutqLy73y2IOlqkE3qqoL+q79tSDMElbOwS2RCQCoy2k14jGdfiSuyw70AVUZqxiSDUdXAtRzaL2jto31q88taXzxjKBpGPiEA1BjQdzAtBnQb0GDBowZMHxCjJwlpdRy9H1GogTUdYd+BnTNTaxOmy3+HR8ByyjpJKmxEuM5Y5/iLbCXBHPRKlBi /RAJ4WAYaGGRhsYYmHcAKYZmG5htOSRyR+jsosH7g9HJe7GW8m3e7MR5zqjhETTKjHKlYM0DPyge/mUvhwfWpHrBBJckb36/DTKOpGou4IbpHQhhpjNAdKYXPbgpIsaC8CUFTIgF5gDK0Dii1QaUofclYIgUyHs/Nj3CDch/JtqKVrBFsnCkGjIz5qjMxUxUGRs+2GbjLYPTl4YL21QfUHNB1qCdGbWF0eMHaujnu/6MfM1plGLWxgPeyARw7PN1 iApRtcIr4dGHQljLRMeXqweMQGVwLx50NbL62osaRHLB0scn7XuisYxGW28Gvkk2cFen24GxjuDXo7DDMH8I1QcJuh0FyjKKVc+xqKoHGX8k/sAMJ2RUCJ5PApVrq0kLCXMyIdIncTXHGo2rOVyyey+QQatc8UdghOen/t7jhPT9uOZ+e2gtzUbWn0uVGdR5jJozrNQIGnVMvKIDqleHSDX095k5gGoBW7O7xdErABEDXVdk7hIAAfWFIESLJ9FN YcDJzKS2SI7EyY1YKsHLEIc2IWFgLtYNQ1Kg19gLZPkJKMZUkUzBMbSaGS6MvSfw0CQDVLUATJvUY4cD0uZxDTWHSjAyBp7WzBihMU9jTSnTgaJB81AgO704A8p7LGym21JCB/QDJwAFwCQACQiWqcqmDJp6HMBJASjH4SEk1DEHtwSNLWzS0WCEnWBGpW515AtIUKYZYBE7QG1GIAL0Y9TfbdSaSg6GsKfCAfJyZwMnwgRAH5BE7MycExLJ6yfh SoAWyfXSsgBycBZanFyZEt3J1p08mCWbydwza1RwCNDAgejWCnrk+aa+T9pm1iimYpszHgdrNDRO4cik1h1bt9AdKeiQsp2tRymMpglkHUCpj6H51rpwTAyA61UIEkAqp2qeqn6pleyanm7E4GsSMsDqe7UbkhGd9HssPqb/QBppNn09XQAR1Uwxp40f0xqOs0cEL/W8gcDb2GvobTHhhyQFGHxhr4EmHNAaYdmGYvEzUmmopIZPmiCp2aa0mW7H ScWm9PZaaMm1p0yes0tpudLhTBHPafmTDpnDGOmCAVyf/R5Mc6f8BdWK6ZKmbp/yfumoM5CCenJZ8KcgzXp5gHenu1WKa+mEphREJTkphJLBngZ2GYBmgZ2FkhmVpoqbKlfJgGYRmKp1ABqm6phqZhIqSVqexmUSFgDxnupstl6msgEmd3syZ4acpni0amY+jZCwMcEH7REMd/HR8KTpHi7UYyG270qJcQ2DCXUQwgn7qpU1SR0kLJFwAckPJAKQ ikEpDKQERxCd3rkJswaBqbB9EYpisJs+EMC8i5KufolQFsQtALQJEzpxAGl4obyIwm71C67vcLsP6p2wcaTA+ee+D4kobCciQEQA4mqSJnA2GuVC94TSNxCpeVsCaROTHiYqKVS4UbyHRR4SeKGtShArAIPMSAm8wYCPzDgJXxuocXRGunqMoLJJ6graHNJkUM67RqiUOF0t4c4GIBYauGFrQeAF0Sp9I49oHOBMF1oBdENQN+Fmq+QCXSwhFuk0 OW6Dq0MedFXRRyxOrJgTgKkHuArxHA4iavQsJchAaEdSh0oTKGyhcofKEKhioUqHKhHI0wZsqvMx7o3yURhtsHn0J4edZbbIUsz/1FQWrg7HqlBuBbcd4eIZXo6DHSM5jvQX0A3mA/OMATBhS/MFbGJx6atfgyA8WLGUZodsR9QMUTqBNA+ccXKl5TOE8vdV75w9tJ7j2vWIKG9x1+bPb35uovp6TxzYHuA/IB9RUhWgA4E0ak47UrmpHgfQBeA3 gT4F+B/gIEBBAIQN0aCWPY+ru6qPx6UbAXZR7vIoX+ak6uAMavWwTqojqFTqBzdmxQcgm33MJYiWolzRru7t6zsqe6JFssfs6h5zws+7bIAcAnZrXCBux0WS6sDsgerLGA7HncziuXmLVXRbujgmiVvdAEi+kerAJgLNAPh24CUrKiMJjidsitl8Eah9VbAUZz8hRuBpFG9bN+Y/KgFhof4RpQibv3C+euUaWyFR4AaVG5PHUvtTNJ1lmgdi1fJC ykx7BCByT/0YgBgBlSXaHcAwpalNhZznLDEoxDnDOYigDJxDVoS47EKTCkKAV0FxAbPYTE6B3gHNI+Jxpg8G+Wd1eQFyd8NAFd2krPDLFdBQV8FYfQoV7FhhWCWOFdi1lHJFbWAUVr5PRWipS9WxXcVubAJWzUoIEo6xFEB2YaZLL1gY7rRvrE4WsoHKCgA8oAqCKgSoMqAqgBZv0tJWcnF20pXsMaleBW6VutQZXIV9aZIzYV1ewhJEV8DAZZkV 45y9TeVzFYFWisfFcJXRVoToEH9jIQcLmsykQ1Lm4OFc19zyOHkenipCJPLrmTulKHoB5gFSAG8zoKfjaWaW/6s6WUJgebRHpFvpfeCdIdagSBd4VsEXHaw9xt24iiB1EeRsBZYh98mfGYGwBhQfRdpG6JtZdLRkh64RmgLvNcklL4dNuEvhiwGyOx1pYsBqiMeAkNfhFGaxXh4BEQc4Ef8DgGcGXBlwchkbx4gKABmAsMBcB+BF63qDYABgehD8 hwQVcHrj9AbAEwBdRFSAwLVwJIATZH5i5efmrlgJZuXyCskSBFmVFobdKXljBLXYUiCmHAaGwG+felQOj5eYK1SUTDxJNSEtRNS9SUki+IjSHdNNJgSc0lJZLSIb1uzmSW0ntJUSJAf7pKG64iA2NSe4lA3CM8DYNJINikmg2gSWkng3sMK0iZIbSXNNQ3HSMVbpmSBv1vJRTsmVeOjJuZjpM11SLlkeJ8N4kiqTDSYjf4cYNsjcxTKNxROo37k2 jfQ2TLT6JGD85qxQzaVCyhZBjpOpcynnp66Q0xb0YY7hqXCXCCvLKjC5MZSgkge4CwLo4yQGeAVIBvAIBEQTQCXzMAUEGwAk8v2GH76WmtrEWbO1NeA95i2wZn6R59egbAN+6GBbDR+JsYqscmBGCZGIcLgwsWdF5WD0WYekLtpYIQ0Qziq7ERIB/JW4Qynvd0e/5F28lyfFpryRoeWxNUoDI5ftdRJurIEmC/IJb8W0KESbyXeQ70A8r3pZ9fab iloua4Zq5Rc3N035Gr0cWWYdCSYGl6+uZShmAJIGcACQHgAyR3gBXBm0g4kysIBZIfQC3BeSH9y2Tv7ZgHsTkIXDHaXix/fifZypssEeC0JuMMzWZ6VsAyrN3dCrhxfCXlsYNSYDmWOo/s96T8GexqiafiaRj0FHF2chST4CcBX3Pahl2t9GrQyzY6mahkiEOtSbMdP6htBj9Y5bmhXwecCgBngAYFm3ngW9E3Z9Ac8FIB4gAEDgAD8I9paiCmoJ c1z6t65YlHGmmWqcNjIOwV56v2iBav0her2WN7KgWSEwAAQIwHwAfgDcHmqfgUYHthcQASBUgOAGYFcA7w6qwADOuDlH5wUayH1a4Q/JwU+pkqLGDlidexiODzli5Zq6bVm3oc2B5gD8QBAKADcDusFwe4HzcCQCgAyQDzGYDGHbMp3gWGkeqmEmhDudfo8I7mr3v/9yIv8L4JN/RYsWanmgPo2KU6o4fD6ThiiqzqxByPqqrAWyKhuGcUK4aJgt 4Ic0LBRoJW04NaYMLNlL3qH1HQgm624fORlYjMDPiLyByHSpi+0Ha6Qa3CHcchyob4dxRh6olFMjSl8QbaBTuFvqH4q0VcRFdAxTQFrXjuysv4EDgToHHBMAKL27nzBpCZLH+57zdjCPCt4JnodIMP0XE9u+P0DqouB7Qag/6iHYmxzOV1ErWxoGtZ2BAhrGobWhx/Dmfqyzf/RF41JVELiGN4S6u3ELvXem3aV5EccZjIfCrbmgtwQgFGAbYe2H wBxIJICMB5gcb3Ehl10FPoAvgdoFuq5oUEAXAFwJIABAYALcERAxApA9XBb0a/00AfgG2ERAAQKqCR3l11HfR3Md6YGx3cd/HcJ2vF4negKxRincq3JR5yRp2pK+neknX1yT0ypDqdIgFwW4d8j/W8G0AaDLYMPdMaBCAc4GxZsQeaI1ZJD9aK9RnAZwA6CgQD6oSliVkQ/BIxDiQ/3TpDytjkOFDxQ+04BgFQ8ox6NktIELaOoQqtG2Ny7IkKWO nUvUPUATQ9kOdD5w7gB5Dgw+UOvgVQ49W85r1YLnlCuywkBwx0GInrxlrFvk6h+ZmyUlmxa6oN42wdhdwB2dzne53ed84H53Bd4XdF3xdxNfc3aWlNZn3TDHzd6WF96/gX7mYaxE1UjhOXl7bZBidhaZ5x6+H9UH43sc+3+xyVq3m4qgcAvh/9AmR4NWJoQcfMcBLszsXr8q7yl58ZJsEHaFSog5R20d5cAx2sdnHbx2Cdy9df7yeknd3HqeufVK a/VgWs2ADgFueXwFwc9CvHSmkAnG3Jt6bdm2hAebf0BFt5bdW2juunpyWS4uJZ+AcF5QFkh2gSJeIBgwGYGUBmAAYHoBJAeSHthnAABdyWxJiixYOP9wpe/GOt31Yb6et+6U0KNN2bDdQ+ETmV73vEdhYOPVwI45OOcjqzt7np9wGtn37KjCfsGn5H5EKJJyUvKmh09Xlr6RkgXeB6tpBJeeN8gLD7Ynavt6LtfzEe7NGxhImv5HLnj53Qppr00O qie02wPvS/3eoGY5IP5jsg/iAKD5Y+oOch6rYNaNShg8a2qCGE+P02t55da7ABnSW9LbbAhpVGKsLDKRIEIJgDMBGUVAFQAFAU0UMVKMREFtOglfYClgXWzYAyQrTgwBtPAMe062knTl0/ZA7SD0+DOxsGmYlEJVmjpYbLR1jZrS2djna52edvnYF2hdowBF2xdmNtsPBZ/08yAIzoM69OIMUM8EVwz905LOY6Hw7k2/DhTdkagYqhYjGylu4o73 NNxLM1VbGbE6mERtyNc2BwQCgHTYQT43cRAYAUYHBBxgVcDPAAQXAHOBmvfMdc2owxEZJODtrpdQnyxs7ZKObDZwCwEhzRmDbhCImGAnKmVamNdQEBDUFxggunuEomiw1o5on2js/e3mfhRjmL2nIGGFrCC84+ebWEFZ+nVBukMDnltuzfGQPh4duU84QuC4g7mOFj8g6WOqD1Y63G3+jY/L86ttgUaKY9gGLRbK/RcPEZ8Ae4CSAEMct1iWECm2 FkhQQG5WcAVIPJAoB4xA8AdAMkcSB+BlwG2CEWOqkguIuL29fQQBrlMQPuBCAD8RthagPyAOOBhwgAUHsl4uOrwu46nf7Al6WE4/bwFlFs62W95E6vdoo86uijBcF+pYXVsPvZkIB9vYKRjcL/C8IuJ90RbyPxFrzcKO59qko+6s1p333OOK9CtkGq68LaCLoODxHL6zypg2aPuTyLofOVlkIefOnsOWIypwOdPTfJdlreHltB5auBuNpjyC9mPS DxY8oOVjonbVL3+uBO1yoTncMllWDv/taGAB+UYWCzTzIIg6JACs5/sqzu09LPhMYQFkBwXQB0dP6r60/UTIz2q+EcRWLiwUxYWEVc+JssdEAhYpWJ5OBJ4NCDVrU9ZkFNhYIQVcGXBUWMtjVZs59goEUDFSs/auHT+q+8cmr8DELPSWGlZquHT3jHvThUvq8xTBr1TLNEaSMa+s1JrrIANniSXdeXACZ8tiWvCB0RQY30XUgYrSmZyAH6lrDiAE HPhz2SFHPxzyc86Bpz1ArnOFznsnzOVr3hSqv1rraU2vGr0a52uoAVq5xZEbiDCOurE3q/42ISc64eTIIRzWrV2NW64undWGa6evVWWDFeutM2TdTKgxmRub27FYGIcVW9u1H26J6hTp0iO4X/WxO0yAy8xLnjP3HwBMAAYFm0PE+gERBnAA4GxB0Cg0ABACyK4I23vwbbfRvYD4k9rbLLgo7sq+y5lswnZFwHEA4JunqyqWtqE87ZQ94YiamAQu aULC23twsJeF/LkJojQYwQxbyjz9t6gS5mmTuCcFpedTdiG2sKOQzAXIOkq6QchErbfh35KHQSvkdxU5guVTuC7SuaDjK+QvAY1C7vERJnY6RP9g94C3A9AT4+wAFfDi/2DZIeIGYABvIwGAPcAUYEkBwQdgHaBVwc4CEB9AZwEGEQVTqtLvyhitpnADwREHoQDQfUooB27g8G7xCAIQCjE+VNi4dKTrArqRj4gDgFkh8ACgFaAfgDXABBlwA8Hu AEATQHuABgc4BmAKAdqokuO4/Tga67l/U6eWGdpS8ROVL9Qpj9JBiI802chUQkS5sTlqiFvlBzYHzvC72SGLuzLxwu1vPN3W6n6j6g26pOmJVqHLgmJzlo171hoHptcvzBDwEQBcV7YpGnbuta/qYq4K8MIBldkuFlhQNkY7W/SaK4vLhrX8w9U47qC+SvYL1K/VPzltY8EnMr09vQuEE98dyu5Lg07aajToq9eWSr+SfNPyr9AD8h7YOjGFAtwZ 67VYDJsm4KStsKR881eHfq2oAAAuu3WiYOoLWTsNAfjGogK0InA7Vp7T4iGmik4652uGWae2xAv1dAb6u6bihpQGKrsR/6tJHmm6gAZHia7kf5CBR7UwW7ZR4ADtDuAA0fGALR9bufidV1OB9HjaEMfwSYx9wy6Uuq5EBIXbEF9trHqpNsevWk0eIHPrpjZ89pVthsoGdR+YDFuJb1cCluZbuW6gAFbpW4eNmBuNogBRH8R6SBnHha9gw3Houw8f 6NFx5OTfH4UH8fAnoh20fQnvR6DADHwjZieSpuJ7MfEn8+2/V2Bmx4DG6zhF29WAj3vKj2H7gAroXn72bC6htIkIuxP1tPs8H2UoLi54vP8fi6EBBLigGEvmHVonEuh+hCcn3VzyMPxi01lb182IHk+oijO3B3LbgswDdAh24atlGvhbFqG0bMBcXy7vOeTto8CuOjxIgJ4kquYGjv6efNDKi/Awnhvnq95KgPh5bDyDQ8EbQILS6HXBU+gvlT1U /gv0r/VufKChsnbQuea42KwuyhxcL4wDwY443BrAN8aYOue2S9p3muwq8MuY4JnZ12jetZvQBAbg8BHONwMc4nOpzmc6hu9aiFFKECwDCFg5pquYH9yza8dm93AAiP3V3iKxAxVqhXvXYkADdoQCN2Tdm4/N3CAS3et3ZIW3faB7dumlGKXTQdql2Aq5Kilc1Xz3aqsn+JSgJ7jIMWIWKXmzXYTrnm33rHc06xblD6I9og1Wfk83OtTzrh/5rj6S zRIHhexCBtw7hEKsAFRe8F8vIMpNA8c1W7DIqcwRPM23Y5OrXL4fL+RnwtsGsy9sMaHYWmXll7ZeiTlc+AfkRqy71vp+959n6nGwqPUoMUTGBUjK4Nej8ChoWYAWA1yD7zOrX633xP2j+p87iqjqXSHUofC2vUHWyo8U+h3qwOeml5UuvkcV4iX2h6Tv6HhC81O1c1h9pfdTkzi5f8rr8f/6+Xm9H4ePG0q7Wy+87EgJBJAbFkqvGkjgGqvPTxlF 5EfTiQA/ev31a4Rvqz5oEA+9sxhrjP6Ziw8ZnEzvJ9lXkkVBZOe+LgS6EuRLm581WBRED7DPwPg68g/5nxm/k2ivRTcCOmzlTdLmJsJRtpy7I7E89u9gElsOf9dw3eN3Tdi16tebdu3ZVuagNW5kddtpNZBN7nl56KOM17c9PqWwJEzVhMqBKj+K3L63Rlb+JOqhrReb8F+duy9BmVqZB+3B/nEGvO/fh0HbqXnvchT9CGoekrpU5Su1T09+8XKX 2ra2O4Cp8WyXymxcOLpOgIQGeBKwFIG7vFwi46m2ZtubZmAFtszAeO1t548kuQIr2OeNjnzoF4uzni56ufRL256OqIv8+6veD9PK/kvDTm+7+HlL48fdFGDCpfwieA+SpTI+9wzwM3lKozZ/uvgdz88+OZ3j8231bwT9yPk1nW7JPrLik5kWsRniV9RcBTUEUWYW1Koe0+5Z1UmAVP/GXZx1PrB6jAft1/PUo3fVSkFxZgP15Rej5hWLZQR+QOtl PR1xHcSuE7kl+TuGHzcbPehJ8ndvXKd+oaabMITL+4eRPXh4feTTgR9wbbWxSdM18NZcETtnJ4KeaB8NJ0FXBxWYTDLB8ANw/Y0+riEgikj7NybkdkwBEHFYhpkVK++fEndXRXdr79HWuxNGZ/+A3Vz4k4AwpBH8g1KMOEF0VfbKz3wB1PK2cEw+rvdPkw8sK2cEdLEssBgAv1ZQFfAqgfH/09Nk10AjpSnOViilsMKFnCS2n6oGBiif0vmY06WN Nlztv34s8I+MpcEnFS+HVgHg6jrgZ6rt4VyKRVpp1f9C4cpWMwHWAFf3NOnV+QfTzCeldIMAMnWAATFHs67VafEOYMVNLhn2jAbC/s8ACWeY1sWJgBqB457Z3IgzEBHK6Mu0j76R+Qsa1j++2AAH6BYRUoIFB/a1cH+ywIpHWZ2dYf5OzwAEflxN8BkfuqVR/0bgM/R/qzzH5SeINvH+xYCf2tWJ/tZgqeS9np6n8BZcsRTHp/VMRn4QBsWQIFZ+ XNDn41S/Zl38i0E/ndNkeRfl34r/k7c1il/8EsD/BI/3qM+BYFfkDKV+YMBllV+Qn9X+wx3NbX+Xs9fkgFJYVM0xJN/BMM36JxLfzZLc0x7O37hB4N0Ged+oU8iDmmW7awE9/zoH37Mc/fkQDFXTRxjfNGEP3J4oHkP419NeOPi3ZW7bj52vHD6vJYP4Igb747qcP6R/H4jA/WP5U/YjJ9/aFhJ/GH7yoeH76eDP6QA7P7HpSjJo/fa7/vQv6zPf G4l/UOy6Tcv7i/Kv7k/IFLWaWv7uTOn6ZzaQDN/Vv51+Nn6kATv5oZQf7TPMpz8/UpxC/fKbAxaiDi/Uf5dXaX4T/WX7/vLaSZSRX4psBf6dXEVLL/TFJr/e6Yb/cFhb/Q373JY377Aff5DPCiBH/a35yOOWb2/C/5O/fjqu/W/4e/VuyP/VNIj2F/4I5bYzCdb6KidZm719EMglzd0T4RFcy3kWwS3vA7q6XJIAJrA54PvEAi4gcbwmVauK4gXU KjAXFjvAKi7F0d+wn3O55PsVW5bbAT6a3Vt4ebdt6JwI7ZdgOzpSLLc575BwaVwX+S/8U/LgoMZYLBP+pGqK8670XEIYPclCrzFLYzfN26xgeMCMfOKpyxRIDvyfNZ9KdxQkPc4iPmANBnCf5C0mMY6biRmAzlLlS8jOXLynfb7EvKz5kvVO4UvM740vGnp0vRcKkXci4wASi7UXWi70XRi7MXVi6n3CbJSXKbIyXW77X3dg6M7drq86EaoEDXQy SheeAYQXAC3wQsCdAI/Yq6Wu7BmZXRM+FWDSEVnxNQHMAU+VnwkLDIwrdRFouAzqoFfWZbc3TvYwGBsxJhbE4d3QIHC3EAgbAii5UXA4A0XfQB0XXJD7Ali6Nffj47bNIE9zNt59zTYDZAk7abnefYFAmegnlNyrGQI7iaXYsruDRwb/cUvJ5oRSjTfRLYGLVoGg6HbzNIA+B0cBqieiXmRdraUJ9If4SgcNKK9Mf5BYmKSLmfA77zAlO4anWz4c 1Kl4OfS4ZHjC9x7HJrRbgW9BNxIwCjADUyQnDh56nG9507Aq4vrRnaC9QV4s7YV4A3Ic5ivYG4SvUG7SvSG7znbCLdrejgAcfsDSCJpCL+fpTCyCWToCVSiZgHV4p1PV4m5A14iUCQAhAzQBhApIARAunzRA2IFGlH5RyvCdhNQHbq56NSLeGJtx1mUvIQ0KGxscNs6x1UN47Df3rQBfYaBvUPZnDSN6fNbOrR7fMz51RN7N1c5D8g78LblVVrwl WKhigziTe5G4xb7evZGRWvoiVFm7agk6qrmBEpvwcYr08bE5/GepajbMSp6gg0FGggkEpAokF7bKfZrnQOAUgs3w9LcT40g6/jVhIDjMUYVxEcMLasldGB2QDqC96NFDq9LkFzvNqwtAoxav5fUDqLZeQGqcjh9A0WzbtWnIY4eS4I7GYHx3OYF0Paz7kvGBIsPbU4XfRg5U7I2xX3Ng7s6Dg6+SKMaCPMq5vvfY5JzcMDBAW9LCYDYw+AToyYbd AAHHX0bYQ4GJbSfCEhHdJ4VBTJ7TyL64Wjb/7MzfJ4QANEFbAjEFYgnEEMXJi74gmp52HCAAkQxDRkQ3CGoASiFbGbTILPaRpLPcj4rPbrZrPN/IbPZYJbPSnhbiZEo6XOt4EHL+7VfCQDLgXu793Qe7D3Ue7j3Se7KAae6LnHujJA5r7Egh56kg0k6FjDc4Hg/IEONBwbILH7qmQN1Rcyd6SslcBrIoOWLuCCQy1A7saYPbkFquD26g6ejitjNS hYQOnDNhYHbjYJsIeEAZjpDW8jJ+AEG57BB59QcC42yWYFHvUl7Kgxh6IXdY47jccIagr/ocvJnQsHC0F3vXl7C3dobM7H4Yi9UW7i3SW5+QaW6y3eW4zgRW7K3R3rmCRehGULnJ/xOCzemdV6xcdkr1gFeiaRWvZ3NH3rbDIN769bXZh5AipB7A4bR5N5p1gzoanDdOqNgyAB5mLAItg2Pptgmsy+oJKI2CDtz4jbuQ7wD/ismG4wziLMA59Ngx zACKHxuaKHg+T5BHCRiY1wRKHMmdUDDg4t633Ut4cENwEnVJ1BKNV4EL0YdYQjPwE25RcH9nXvqzreIAjZFuALgHgAwAZQADAWaowAdoDMAQqDrgyyFbgx54ifck763SsaG3Hr6oKOpAf8UgKR1L/gjvHCAZUaIwCSPhCuXR271A5LbrzYKE5RUKExdF1BlRMz4XlEwJqwTxAKg0CHHvcCGLAyCFI+eg4wQs44pQcu6V3TADV3VPB13Bu5sAJu4t 3Nu5Igme6chQBb3rDL5cPC4FIQq4FQLW4FN0SUKEBDXTNQBnyemBAA4LOnwK6VarOZCny4AOc4jdGMC+NDyCKhEEFTWMEE1CRs4zBFrRfZGuCS+W4Q9rbE5ZLYloVlIIEqDP2LxAIEEl0UgDhWbwAKHAkDowTAAJAxHJLnfGLmXNr4gPDr6dvcB7EwyB47nanL2QetxYKI6hRQ3lovIP1CmBZSjlw8EbMw+cwNAtmFPg0kwLvHeaLAHdw5gTIgLY D7xlROejXCbwyPIQLajAytC+gpchthIWE5Qo742fWg65dfxZsPQJb7BG2ALgG2DxAeY4HAehCdARA6SAA0ACCIwD8gfKCLLI4FPtcFRRfEAj0AWSDngW9AY7TAAqQIwB2vfADxAW+FJAAOKFPLxSd3di4mgsqHQnG95ZfHh45fGqGQLAnzQLO4GI5SUJGQXABNQTQDaaDyC4AcpTEcGaDbASJQIAVjhOw7ADxcZzJjdcqD2vCkDGhUEH7VVbolLS EHlvcYCQxfSBZUOMZ+A9+HIg7+4SAWWFV3Gu5Kwxu7N3Vu7t3QB5/VYT4T9UT42XRzpVjfpaJ0UmopyfGSAGEwIjvd5CXwUHy4RfMCP1R8FLLIIaPnb+pe3c4hNQHdx84fNYcoX3ZWLYO5xAPEYagWyI3CfMDSgglBSRaIbaXT/a7fYCE0PSz5gQhYEqg2eHbKFazUvLO4MHHO733JGL4AToCkAWSCWxfOjsvOCHMHc0E8vK0GVlWqG2g+qGs7Ca aFPJqElPFqFlPdqGdQ6p4y9HqEbsAMyqvKOpvwJtwTicGjfIV7TnwQO5lgmaGRguqFnhQ17oAFcDLgeGGrgRGHIw1GHowzGHYw7qFhyKHSgXRpD2LO8FLzJCqS7TV5vSIaHTQjXaWUPYaR5GaG1g9Or1g44ZbQ2N6XDY4r0VO6EaRFRGfmYsDqIxcipUXbg6I1sCTFfRGZMKaA/Q34ZjgiEETgjm7U1dQo+iKZRX5e1ClfPwGmQ8OGGbdGwpQDxF eInxGf6YRZ2QrOGcI9c7cIrr7nba/gi8TND2oD7zuKWsBwwMRHfdUaCNIF4rooYKqBQpqwtHSF4BXVZZKIhCSMTKaBj8IiIbsOLrTjFVq57ZYY5oSeHWIkWG2I/KGnfKCEVVBeF3rYBY3fPWGIQu3gyTR95vrZ95oQ1964JCACggZCAAkASFqscaYso69Tso2DBv/WiFlpT/4jGRD4//f670I+WGMI+u7MItWFsI3iEmaLlFso5p6iGewGerRZ7+ HaSHTBGN4D5CWw1eGrgfwDvrqNPvYtlSr5JjW5F6WD45fHH45/HAE5AnEE48AME7sIh7oWXHOF2Qj5FEwyk4fPPHIJUDMCdnL8EYKFJoKfN8ydA6KIDgZIbCSAKE3nAkzjtF27LLeFG6fQaDQcJYjsoQsC7lX+I02X7iKvXCJGUaEES5b/hw7ZSj4vfd57fECFTwk94QQhrKZXTmpOItZhSw+e5j1bC7PGeMHiQDJC3+bMZ+Iq75nApeiVQuE73v QBECvboa67GMEpnZI7pnNI6ZnTI65nDMHqgINB2LAcASGTqCLBJCos2UziAKR/ig9YmQARQN5FI22qLQqsFDIjXYjIiN7rQqN5ZmGN47Qo4p0VSKizI9yr/1NyxdHQkLeVLippovRFrmJpSctHZGN7MGzjg+lSAjDFyISTZ4VAQgJcmKY6xHPvZ5ja5FVfU1H8QZcBNoltHe1eCYT9V5EKBF1GEwrt4Fwj1G2oWpRkcZihlwe9HuNLkrdrYazowL pCQoiNFjtGFHRo+RHQvNuE/CNsCytAAzZgTlrNbFF4qtVjjYhUC64oxO65Q477NRNO50HeeGXvHK5mg84FUoqlQ0op770ol74KTT5YTTMR7GWQP7YkJ9QAsPlFMNeM5SrEI5/XZM4QAd451+C1GWlK1GAnYE6gncE6yoz0ayY4j7mWJm5SQ32HBHVTbm6SYDhHRSHCEVIwHecfJqQ/iBJANOERrFj4SAAYDMOZwCyQIwDPASQC8gARiFQSQAAgWS BCADJC2aB1EWNDIFkg5DGdfN1HdfasZzYTUB8STS7dIFIa9tZsIp6WYCaBP6jjw2RHUTV24KInB7tApd5FWQAxUwL/LjFX+JkPDb6UEe3wqRPd7TAiC7FovFFcYmeG8Y/Ib2fIoY1opz6lDBnogEREDYAaJb3AEDAEHbWHkohCGWg9rZ/QpTb33UI6loHtrtnReQr0SnidIbE6JI6GHeY9ACjY8bGTY2LE71GyE7g0B6nbakFOQmejTif+qUwXCA woVyCVwwwIIvOrztuWtDho7mzQovy6afTebUYkK4NcSGDv0CYA1ISjypopcbTiOqix3UorzKQ96dY6eFlo9mrEoyWGkoy763La76zYqqHBIx77FXCTGKjQQ5vfYtRWJazSKYiEjUpFeywYMsC2rfhzogSBzbSfgEu/V0CsJTVKcA+KRaQZ4gKzf9JMAIVLC/COhpwGQ4oYd2ggYBPhhSXDAuiR0jnqff5CA9thpsMdS18IXGf2EMBXQWFhjPbDCc 4oJJsgSjJiPezSuadVgNSP9BjJWLDAxMKSBAIgDqZcaaE4iZ7saEnHZYMnG7ISDRU4qkg04uqSOTbn4CAimZc/FnGkA9nFhSNXHc4+nFKrcM6asOXFEsJlY5YMXFTTYf4S/elhx2YPEikA5zkAPKYq4gVJc470ia4iRydqDFh043aQ8YI3HMA03FPJZTGwfD/4MzIVGMQ3671BFKC+YhQ4BYoLEhYzoBhYiLFRYmLEmYoP6mPK3Fa4m3E3pf9B24 ynE4pHFjOgWnEu4zgE2rDgER0VnFWeM1a+41PHd/APE/2IPGC4kPFmrUXEUQCPFS4yX5dXWXEL4uPGK4xPEfqZOYAZDXHW4jPG64naTRYHPEDYY3GSZM3G1nEj71nMj7WYtm7ULQ5HrotbHCEMyDYwVcxSyEDFJALcyaQyDETUfQC3+d4CaABcCr6K6AU+TACyQbMb0ALcAkMHGGpAvGGnYp54wyV1GoY91E9vZyqYuZFAYoV+j8yG+C8tJOikwe LjRRTty37BuE4ecjE/Yp4TEAJIDQImSSNrVACgXMpQqwfrqdcBqCc2Y1ybvZCxUQF1S3tHvbQ4vkyw4zjHw4sWHlo9O5zmTO7VolHFrAv8YhLCQAqQY0r3AccAzAMySRfUuLLw1eHrwzeHbw3eEq4A+GQIiE7PtC+7o43+F3fKSYGwhbEUfDCFfZcEbD5KOSTkEOHf4nyx/4lyJsYRQnKE7YLwYmYQWQhAlCfJDFubVEavPYo5Hgnc5cmekGFCWH ATkMnJmMZmBEE8+DxkHEL+NUjHvbCF4UYq6g0Eugk6fOKrSuYLjpNYgKscKcb2ibGDy2XpAncL/FTAkApZQjrHCE0tGiExHESw/jGrA9L4UotfxmExS7Y4p96oQyTFCPDCESAC3FHqep6YpMnED/XnHhnBZz/ENR7SHOc4q0WNKbQV6Z4zHdJSHOABx2dYDE/HZKCYep44sQdT7/CmaspAfF1SbLCZJQn4cAYa5UaJVjbOILSkAVw504mfGh2XlA jTWtT1Pc3H4aInGDE0nFd4m66u4gbB848YmO4uQ7TE+DriZOYmwZNLSLEuQ7+afDSrE1PHsaTYlGhXNJHXPYm+AA4nYYI4kRJNTKFsajRmOS4nXEofE8/XBCAkx4liPQvHQuYvHwfUvHqYivFogQAnxAYAmgEsQLEACAlQE+hAwEuAkt4sAGmPN4md4sxIu/EYnAxH4kt2MElTEgUgEkwTBQBeYmgk6nHgklYkIkaEmEkzjRwk+5IIk/vFIkraSH E9Yl5pDEnnErEmNAHEmAsL4n/ofEkPEjYlEkm/EWY0j4GZL9HnDDm6rfNE4YITtxgcfVFeWPvbI2Y1ENLfYL13XmYbgcSCXw2SDzAfEDvAYgDiQYgDPAIMm3oaG7pw8yF8fDcEa3RAnxY2yHoAPcG5A9NaOQuy5PyNnDJoiaB0cXsBMnAniZgMUrVufbjMg1GpkY77E8xAAQZEmMBZEgqLpgDdDNbSoQOIPnLTjJuBZgaHA+gn8JXzX+jv0NwLML cxEEvA97ZQuHG1EuxE9YhxHFQ/rHSE1xH5fOJb0Ac8CtAREC4AUEDKACijTYy+6mE/WHUojpp9o1YqUqLobbk0TwhvYZFdFEPpHohsGnolPJnFOPatggvaHQmsmj8Lrj1kzIhd1ZskIeb6g1vdsnvo0cEj1IhEHI1S4aFTmzD5XwgNQRtzf4p5G7YyOGbAGclzkhclLk+Ambgvwk+ZAImSLZMmXY1MlMSDfo3FICkMYv154Yv+iA4gb7ShYCboPK FGKuVIlUEsSQVk+gkIouJiqIw5anUUC6FE+HQqtXvTpEOXYZQixHtYqxE1E0WHDkpYFI4xokG2QTHXvYTFzYh77C3cTFdEvHGvfaTGDE8aYyUmM7ueFTFwfBM5l4gzCUklfTggT0nek88C+k/0mBk4Mmhk8MkcbARRmY00kptc0mvZX2GaomTqZNW0mKxcBrMwfsC1vdzHTeMCkoglKAaEteH74bQlhxXQn7wwgCHw2Ckxk+CnWVCQCJk6wbIU2y 58IrNYzjEIqNgH/huCDJG9teVoYhVwTO7JImfYkikafMsnUE2gmVk4Pz3DAkK6UfVyOLX+LowUxacqFwRS6XwZS8X8io6WrgcYw75DkwlGqgitHqg8cm0vSck/k/YIbgOGALgAYBXKEu5fw/xGcvPK5dohS5FLEJGdNftHRgpJDMAakm0ksAkMk0YCQE6AmwEmKojFX2oOIN5C/ZTwgpMEMxe9enDQ6F8lMGLCCzLC2r9IpZr6vO0GlIgzDRw2OF fAeOFbgROGTbFOEJAjan7NQeSVwG+JZgt+CaI+XYxcN7wVCHGBE5TmQHwXpFb+IPIDIg8n7oo8nHDMZFh7CZFno2ip7Qk4pJvHbjJEB3L1WTDwDgc8oAwIFDlUlyBgo36kfkwQx7I78nfo/8YonJ+6OY+cTn9ejisUg1EEXdhY9U+YB9UgamBUlr5a3OMlnY8kGIzHIHhUoImHgq7HX8Y7yCxPMCUwd6gqgf1GeoANAQoUyDxDTy4kYjKkUE0smT tHKmZEsKGlmJcgzKFOgMnSK7HzF/ESnD4ITYZLpgXdilVEzimNU7inNU+xE3raQnNEjHHdo6qHGFcSnH6AQ5SUgDYQAc8D7AdnEM0ID7oAb2kT4j2zEkhHICokvE5PCklBtTYAeUrQlbwnyl7w/QlHwmG4ejAUSB032n7EUykidZ7LBjZZ4ao2SHLYjxrlWYfJXnQmkxHNzGaIJIAK+Fwkr1C+FXwm+F3wh+FPwlSAvwzABvwjmlWQxDEIUiABhU wIlifFMlRUxfZYQCFDeNWgjuBAF52oAETBcLwx56QMy79ZIlBQluHkU3KmUUuNGoKOsDR3HzptwRKroouIak1aHTpUTQI+IOYAg+bkoj8Hb59kotGW0pUHcYqrYtU8QnIUEqElDel7DYq0yyQV5QzAFVKWyIanto+CGBI9cmiYzck2g6anXUwdG3UjUD3Ux6nPU5OFJAVOEZgmVydIE4R84GRGe9MOrhgyGmXUqMEgMpJC6kYsjiQJNiD4VoDMQB cCXAChgT3A0Dd9RpEumALqbsVPz3khsRzNDdHlg2aGVg+iLLQzAzhvBITw0s4aI088mx7C5BXkhPaZ5TGBH5I0Cb0+txAo85BqUI/IHcFOQOLHAkk0n8Z33KcmHI/8n0LfHK1YgaHYnQQI0IrSGaIN+n2wD+nKAQW7PIpPBRk3GHBUulrd0vmmUghyEoUgeki068H9KcYruBOfzK9RB4BVXbyo4WYChcMxFzLSNGUE7KlL0jWkxdTqAfxOwxTQMa HXkDd7/xEpTtMVPYPokdYX0yxEWfLikEok7530vjHnfe2mCU3WGtE/+lieC/Su0l97gdXonoAT2hqeSjDbQP7T7AaM76oIiFlxVTzk/SpkFJKoA1Mp47sWIgaKU0knKUyOkszCCmXw6+G3oW+H3wggBN0lult0tkmbAcplNMjgBVM1plCsczFmUu/EWk/ZFlef2GTgk0DSVKYAYnXxkQwut6eE10lLgiQA4M/QB4MoQAEMohkkM9fBCAchnt02Ml OozIG80n9D803uk8IuwboYnJi70I/LM2GqzpNS24PSecgdIU6jILaYBc3cglhVf7QRdMikegCilVkn4RiEX5BIhfr749Dd5veYBqlRPmGHnX/TyfICHkgBcCrgPyDkMrDJkAbAAS3I3adAVQjlQaG5kEAcmpMvKHpM22kf9AbGYXRcK10wZnDMxunPw1+H4AahGawl46nw0uJQAAkBGyUEAqQLcC1fDJD6AeYAmyMqAUAaKzhAQwknAl9omE4SmY 4+bEPvYULAI42FjVM5Qj7BbDFgYgBS0kXRLVWWLKwY0CaANUDLVJCB2GeGFNQZzKewvapkLQhF5fMSqzBUubKUPNo+oBF59IauZ+AuCaHMmGGVAZQDl0GcBsQEhhGAG15+QSXAZEZ4CC0O5kWM/I5PM47b7gvIF2MkmGpYnSBfMsuDeM8zgSGaImNhJpi7dJMI2MM1lghJuGQspoG44WFma0/MrHzCuC5VTFxr+BUr4swlnkQAwAksslkUAClklI GYDUsxdC0sq2lpMnjG8UholZMjqm1o54xCskVlisiVlSsmVlH3eVk6GFL5n3aS6/01VlO0rHG9o64GihM0w6szYBzda0C+gUYAILVyDYAe+hUwX0CsxJCDGQYgAGgOgmQIweBLVZXQ4PI0L8+fBFOs8EHk0q0m/kxWKfcNRlVoKZQLtbE6HA8DEmo1wnkoYVkcAUVniswRhzsg3YLspSDGWFzaRkpr6+E1r5vI3cHWMlNkRU3hHps/hHr0BmK7eR 5CoWTDzuNKcgJDRZHYshKhHIzk63nLKlq0oJl5U4jwC5PBavAyca17Wai/eFUDcHYC61k5IgVRRgyCc8+mFo5JmKgmxH0s4dniwoqGAxKtFThHU45MlonxRETEFMoUJTUxzg9DUBknMs5kXMobJXMshkUMpJFeoH/QgiexZtwR1C4Exfw/IHsCpGKGKFCRKl+7TdEYM4pGH+CJHRAENlhsgSCRs6NkLAWNk7Yh3bOmeqCe+NiSEcZsJVLBhkOcph lQ0paE1g2Glh7LhmbQs8lxvC8n8M/aHXknbgsc2DhE5XnLN9RvLcc+UC8c0fjJEBRklvRbHKMn9k/cc6rsSWWJOUiulwYgNl7YikAbgZcCaAQzrELdbZmMtDlc0h5kJYhMlYcpMmC0/ul4c94KuVOWIeVLejVocoFJESWIOUgd63wb8LFY+86lY90DVs5jnkw7kygcQxHzkEUF1aDZYuDG1zCnQxEg+JWywwLm64si2kpMwdkSc2+mMsrK7Z3Cdk gEL4AzAW9CaAeYAUAapo4LZgBIwyKCtQZ4BGAcEAaw4+F1dV44IFZhwbgbmgGgQ0oDAbxEFxAEAqQDxLggG2DnMxVmdxU4FrsylEiUgBEu0nHGtySoTpUBjgTxTmzu0qTGe0gEDiQbTh50UEBAgKul1M+x7oAUnnk80zpU80w4Ckcw5tSejpIfaw6BCIymtGMnnz4RnkDAankwuBm5mk5ZkWUy0kTIjFr5hV/EHEPnDVRfT6M02eKgct0lIxR7nP c17nvc8IBfc94A/cv7kA8xIHeEjrlwU9Dn+E5c7dLCcBboYInC0mwwPuNxDI6EyD3wBLoPaMDi6QPSKk5SZpdjeelfY0imBMmFnL0uFm9KWebzkVkx9mVHSNklviw1UO6MwJSgpyOqgxXAcFb0YTltY87lic/FFXc/iYZM3rGOIx+mLw5+lyE2qrBTdcIEgMQTf0tHEdo7l75Mo3JdFK6nhI+0HMAJrktc0Xaf3Qzk4RFsA60/7j2IfiThcgpEXU rXbRmHdGsMmLm8UVaGjIk8njIxLlTIi9GbcAFrR9IFqB8jdxCSUPlIwCPm96ePxOqVdD6RQt7glUmlfkl1kU0/1ZF0/9mbtPK7t7UNZ1vHBFeY8CnnWAvm4AIvnxs43ld0nulIUgblpswuF8ufng9IUHwcBQyhkc7HSytXNAAc+AQLc2FFLclbk41ZwK3aWsJt6TwLMYhtmCSdUBl03skicjikXc6+ndYkdmZMlYECU00FCU9Hlqs0SlY8zolu09 hT446TGPEVzR2kY0L+0iABkCrtSjY/nzyUiSxdMrJ6CoiOl+ePpmc0J7kvct7mggD7la8nXn/c0AFBlDPEUC+gU5zFMoi81VENncXlBAIgByAWzHiGbqCXGJSjvyBJl7M9zE4PC/luUzYAbgF/wHgf+7tAA4CrgNXBsAHcCIoKJadAcSDObEiQG81DlG8rrnZwx5mJYiEwW8oWmoUouFrct8jA4jdhbLQkbfUU7w405xl15IAVpEqpigCk/oSyY+ INmPzpHUHsFB3VASDLK4oUwBxZj8ZxYroGebWgL+INU1AUI4+BqtU0nbZ8w8bPHFz7PGAkDAnZcC4gKXBJKFckqs3AUbs9Vlk03fnBLd0Tf0WykwKczil5Lm6M0qfLV0sHilCyQDlCyoV38+wUYc87GWEFwWDc1/lz9IsCumKsxMGT3yx8h7R7wG8EeQ7Ta4QIime8zKmVs5bl+8sJomLbcqP8d1Ry6Helf0YokXlSFBEcBLJZC8Tk309Pk3ci95 NEhTmO08anwnPh50oiSnvLEgWe029DekbnQa4xOw4/H96oOZ4B+QVcB6geKSBKBTAS4nNJJPE3FWAJ5KUYEnFcOJ4lUC74W9qCeD6OOH6nXa67Ai0EXzAcEWW4uP5kpX2z540hKIiljTIi6D7etJgV0Q7J77RYVFMQ5D46CjcB6C5aqGC4wWmC+xDD7SwWCCiQCoi2DDoizpzJ2LEVAikEVgi34gEihAEwiq/FPJNPHL2CkVC83OYSQpwFWY8XlW UpcyPY1oVtcJMK4QcGGM0gwrMfS/mVAZwBg8jPCQ86HnhYuHk7rRHn6XExkGYQ3lBU+/khUxClm8sRaW8twV8ud+B6QX/TQ6TqCHcp3lkwbtbZgBLiYQGsTBC6FlRgMIUME2cqQwIjHaBboS57dkbh86mJtgZWJ/ZYWTTvRrF84OniwiK4Wp8m4VVFAqHMPe+nScAoW09PlnFCkAiggS+GbgHgD0IUvDVCsvlKcjHmXAyalbkw3pYMsbb181rlN8 +YYBcpHo1IbkzkwJ3yscIaFe9NBlLFXCq7kmJD98pOqD8gXpRCOLmj8hGnj8jC6J7FLmo0g6GN5JuCECCHCAonPSYVc5CNQBKqDtPbr5gTpBmhTfk19bflN7VZlNCspYH8/9FygWXiE1Eorl0uI5olA0VaCiQBVi88A1iusVDC9IHdc+Mmm8+yHm8z5ESfCKJeGDKr1gGOSOU/4Rr9G7FRGXqw5oOCxhin3kRinYUxdB6EbI3xrOQH+IG0lVr3uJ 1DwC3MVdYnIWXLJlnZM7AW5MnwHPCntEECt4VEChgrE8i077YkDBw/V9khHeTGcsdiXJ2TiUvCaiGxnEknMC8Ol0ilSkaYi7JbAY0Xg8s0WIgGHmWihHlI8yZkSAaXD/C/iWLMrOnplFUW3i+JCyCsIDus5lTD5KGD2Y30XYnHT6aC2hHoAWSCraUpDPAbAD3AZwDBk+gA/AHiy74e4Bw8wXlmQpIH2izmmAShwU9ckCWifcYUv8j5mztOyA+FaX idydqDj0muC+oL/hl1f+grfdKkIALk7e8hjm+84Jmv5KpafhVdEtgemz1Yn5CNHTdqmfebBWuetzGgepAFopPk0s6omXc/MXaxQsU1bLPntU1YGdUvfk6g9AC0+GYDZjIwD0IYfANitHl5M5TmHhX2HotGTqG045Gwg8RnzkGrlxHUBEWS3RlLhO+g9SvqUASkkHc05AlsuQo7BSyKlDc2kEl1Itlbck1SEjEcyytY0A96XARfnFLIpE+jm8nCNC RiqimuQRIDMUWXhYKCbocE71QTS7gnziA1Sz+M2lJM5AUp8siV1E3IWjszAW+uB2lrk4aWYNdBKSed4W0okAZvfRwA8/boU08kzTIygQGoyjpkwfYSU0ilgViS3pnMQ6yUbgWyX2SxyWfWFyVCANyUeSnkUJk+fKYyjSWOA7OnOAr9mpQR/EtnDm6qQ38k+iWrg9gCd7YnKyELS//EQAeYA/AIZlvqA4BvEQIAEgA4AbgYgDngA0DOAc4C4gEDkR k7yW2Ch0XDCk3nPPbzY7S3DmTCpxqP1afx3CCYBNKeS5PoPuT/1IyC77EnhoS9KUYSzKUn9LBQ/dCHGduWxixQuEoEBABpnlT8yrYo2m/UiaA+5UiUiEnilSczPkoXUsUyEst4IFW9Bb3TGH0IdyCnHe7kpQGcD0IQgCjeN1jBsrxHnjZSDPAQgAOS0YBXI5dnHAlHnKsxsV/w+76Y8m8WsysaVLmHBrcyofivwe8lbY7/HrrermGiiACxyg8Dxy xOXtcjWW+S9aVASnmlOCp4J6y95kYE6sRhZMeGLVKpSU1NfoBixsYxChejEcO2V3SqtmYS1/JnlafwTcgDp/UaJkxXdmwmBE/mJMpAXJ84WHAy0OViEjAXOImCGQy9dl0S52mFMnHHwyonk9EplG0NLDAjTdlabqQTBO41lh4zRDRsgCgBNYVFiSZWZAyYJYlvqLYFLEzq5C48aYfyimY0aH+XKkyBydTaxKxJYBWp4j4jfLCBXqPLFjfoOQ5xYZ CAh01nlqYtgXMQ0WXiywqBSywkCyy+WWKy5WWqy7nnYkBBVfy+jK5JFBXdsABUYKprABJIgDgKjDCQK/BUwKohVKo8SG34yQX348Xm1y83Rc3OwkKvC8gIC3wF1vVpY6M4WXr3SHL2wNgDg3NaXWQjaUEw7aXgSkIl8uLfYXwH8jGBcGhjQTPTnEcb5lKX6n2hD8GryqF4PS1elI6PiRw7Z+rJoxUDo9ZwJOQXqyPrfuTYvSORE5f6WnymqVX064 VoCsOUvzMdkPC6iWKciuXmEjcnGnZ+U/IHd7+KhbCDtfg7ECj2msS1epn/V1iqEBqbaAke7j2f9DWrP+VGWJtTYYKm7HXWFjDXYm4GTP2AzE42bf2RtTZAF5z9/dx7D4yPHCAlFZaJDgBrE34XqkkxLg/Z5yZSTp68Yclg4sGKCkJeaLkQSQD4aTbL4paQ5642RxGYXlBFJJtRdK71J0aCmZAKnFhokBEDMAeKRyOTKSoaTZUtTPlYlTHZXsK5CC IK7DDcpQux12fYn/yrqY3TarA4ZURW2PbiUqSgpVMrYpW6PUpVcYO5X0aSpUr/B66zXOpUEsBpU0kJpXwgFpWGpNpU1qEeyC/UDL+4wQEk/PpXL2IZUCivDL6eMZXSOCZWKouQGspWZVwzD6C4YJZUwkFZVuHNZV52CKAGpDDT8OdjSHOfZXfqQ5XyoE5U7Oc5UbK9YAn2a5XFJW5VsqthVPKsDDgqtBXWaA370yjDDfK5nmHZeiFkIFjYc85M5c 82G4Cie0j8gOEAAqvxJAqqLDlKjlacK/+XZYWpW43GFXCiOFWCA8n6AkpFVkA1FW3K24m9K6XGkAfpWyqn4V4qo1L4ZQlU1qYlXRSTzRKkq34g4VuyUqxZWAkTKQwK+lUXK/lW4ZYVXKOdlViOGKRcq05WyONBVYzaNUkAm5Usq2tQiq94neJX+WvKoywJzD5WsAL5XAYZCC2PZVHe0YwpaStVGWU/OkYtY+UwgzTZZggsAxC3TZ+AolpMfCOFfi 9ACpy9OWggTOUbgbOWvMUYb5ynSBXItWU2CwkGayvyUjCpNkvMp/ltfN0X2MmwyYwH/TJDSwSwcTmxPoN6EpyZchj5diZ1AqIoH9dmEOypjn8naYV3wbcrLiCbph8r+ihMyqmGIjvnG1V7ZS8X/ghgnFmZQ8JUoCyJXkS69YH+B+ktS7Y4Ts6OUXtdoCaACgDYADdCS1Evk6wxTljU7L4tijVlqc6cVHo6vklI0BmUKzAASymhUyyuWUKypWUqyz 0ENUWdEY4dwhJ9Sxb/UoUCA49aj/CP8w1uEAx9I3V60RaGlB9WLlrQnckbQyPb50pGl51GPobitLmz8ltwsGCHAlEV+DqRY7w/IRyl9gFpi0EOGCzIh5CXq9FA1Y0QinlJGCRcTswcBWuE4wAt7ggot67InflKMrqkF0+VyaixgzrI2GrYnB3pqK8DmdSiDVQaq0Bty/XmBwHwl2C2dXaylAm6yoxVW8kxVHxAvQIeMuoLAHslA9cPy/Io/TYwAE TXnZWnugKkaLc5ZYuKxd6tQYaC/8FCQI2H1AwCs4Wy8DnCJ8yonfqoGUhym2kjku2kCY+JVPCxDUWEjomMS4plCHeTyYAHwDWAHdTjTBTxlJRrUMC8Va4ysOlkk1gVWHTTF9qjOXggLOWyQHOWjqguUTq5hVxeOrUta61iMytMpjBFmWNCiXkydY+maizCARMpOjYnNTo9Ct9w3We2CyQKjDLgToAzgVoBbAt4F+QZcBq4eYCVgXRWd0p0WBSrzX JYr5EaBHNbkwa+Ds2eUpO81kZZoRsBbUDIgUTfxmq0teXbCx2UMEltwhFHpAYeLKrHC1ARigrCmaRJGrpai8oF6GaDOCVrG5a/tm1S7IUgyiiXq5QDXFNBraPCqGXNiirW9ooBl7k8KAG9FAy7owPoh7djUj8zjXHotAJJ5XjXxvS8mpcwRlgoS+Bg61wKjfXdpIwZ/iA4/85w6qU4+oYrmWE3vI2Y91minaXltAXcoziC5F1vdpntyntUQATNmo w3Qb0AIwAEgXADnoX9B2wO1HWw67VAPfRVcI+7VoElLH4cwyDN6RejCSNwjWK9NC1mZuDP1WDj6QOenRarmIhC8skbyp2XmgX+SBEDGCtQZ5CZFZHB56BynQ4fCWNY63XLiKkzBypqkMsorUAao/CSEuTnMsqL6ga/YIrVA0ATrUFwbhOe6DYpGICQCYQ/ABNj4AG2AwADJBC7AEBhLZ4Bv0gRjI8s/BnwlKALgXbg/KLcBbgUgA8WIQCmdAkDFg PyBs9OSV161dkBI++Xla5JVi6obGlzTUDSVaAw6hX1l1vAzmuUyyVLhbACZ684DZ6w3UcIjzVbS5wXea90WHxfLH5bL0BXwf1RIeRYUCRFcj6UHMCYwJxUBXBLVNkDEw56KKHjkEYEZazMW8VapahK6qXo6iJV5iqJVXyueGxKrAXfwzh5DSonWj6yrVwypiUsWT4V5KgSFfEERzkQiDDjTWA1CQijpta9/4iSrrUEy8hXIfVXV/KCuKa67XVSZP XVVYRj7jaiQDIG+A3CQmbWWY2tXi8iXXNCsalGS3ejblVQWM0255dqm5G2aiAAi7eCJsADcBfIOABbgAECggcuhzdeeAvw9fWOo/yXASnWWGKh7UQSvHJo6BcjES3sAp0L6VPocHzETBsQji96jJS1KW3SqF7u3XkHEeZWAh+UwI2MddBQ7Az5voPPTJ+PmW7dU7lfqr/U/qn/V/q7cbhyjO6RytqV3ihAq4gLcAwAIwBQ84A5JyvPWLhAvUHAIv XnAEvVl6ivVV6mvX7PPlmpfQfUjU2oUPyzdmGa/6FLYyXlD5f9lGgfuSctWfXuYqEZba/YJ+GgI1BG4bZeElzU+SjulG6oeWbSoDxyGs3WPakxWwwQ6jf8VQ3KhMjkOILQ1J0Bwz6066UL0uRGlhTmGv5Q85MjRsa+IdtY8w+uVbvROiBbPRHgws7l5a8+UFauPXoC//Xgy41pAGoTGpGkfUAMlJWEC6rVvfQ0ZYaVpwu/KViGKTG4QfCDBosYpV jJMWbZAeQE6PNFjsdQx4MoQjrNTG1jrRYRVyHCVUQkKm4PGy66jXdpzBAKewbTJDS6kb9A66BFZxq0gDUq8sAjTZaR04mljfgKzA4pSYl0q6Rz5qlUkQYZzQFQIhy8OfRKxTRgDUAIKZ1+IlikmkfFsgCrDu2Mk2jJY2SkpVgBkdUUkEm5OxKIN34X/aZy/y9aKsJPUnD4rhwRSdTL4aXmhsgKyAsmyjCRPIhwcm5dDWAmuw8m1ZXgkcsD24pnG3 EimYTwIU1PJEU2xJWk0EASU1sm7tT72WU0oYQGYVgaDCwAfDTM9O5z8dcwAmwJxLfG6Q4gYbFiZJKU2tODDA0mw9TsmluxnKn95WecNxhSXaDMActTRIAybu4700pq3ey5SLq5WeZQC6kFljPOEeyZSE4C4YGFih4llhw/DDoeAH5X1M0416Ac43cYMQH4A6f7PXQFUzTJ41q/Z65vG9IDxzJDpLE343SHf40mqx65E3K66gmmTCasIEmuaaE1dJ OE0Im+fL2aGKQomxthomvTB94iNXYmo1V4mkZ7J2Ik3f2Ek12msU2Umu03sq3U2kQBKSUYXmjEmlk3Tmw03mYE4Bym76ZyHXEkCAwU1+2YU2oAUU0Um7c2um3c2cmg80JTRU2t2HLTwdD3E8/dU0cATU0FpC806mz007mmU37mk03wsVn4CgS02KpcgAH2MQDcrGBVOm1DKumsE0/m92z1ScM104v01QAAM37AYM0xQUM0MscQ48qqZykpCmYxmu M0HEsxxJml0SsgQpWyHNAGZmkgBpPe1hhOD654y0SUBtcvFR0+QmhSDJB8GgQ1CGkQ1kMJIAIkKYAVfZOmdpDbJyYSg0Fm+G5tXG43CYO426qss3/oCs2vG100fG2s0/G6BV/GgtUQq2pWwqkE0dGcE2zErs38S0FWugPs1Imwc1Jm4c2sAUc2CkrE3fTTS3tqG82zm5k3B2AyaLmlk3YWhC0EAAyaUYDc1Mm+c1/m6KB3mlFXym/x43E/UnL2T8 0aZS83im4OwBWo00AW7ZwPmulVKm5804W5nFvmhlgams81am780emxC03m/817qU03AWi02oAK01gsG02QW9tTQWjjqwWtk3wWvK3azXC0+mzG7+m7FiBmzC1sQDy3NWiM0GOKM2Y3WM1dsVUmkWo1jkW1M1mrdM3J2Gi3EACtXiKiQWSQ2g06StUUATDk52E0Vw/kCmrYnBMafixfXhGyI3RG8vWaASvV+QavX0IWvV9y6dUDyvRX1GnuiP8l0U WXZdV7S75EoefnDY6IygAoXwXTC8Hp5gaGDBa2jlo1AIbDGl+IvgtoGwvLtahcKFDCScJpQ6nZhSaj+TI6ujjjvbdq1IWGpLGpw1CEuqW/6+onScqnpAaxz4ss2QkdS7dC3qNnoLgcSAtUAaVD6ztFBI+oXGFUJHAMmvk3Ung2cW/g1W5Hi2iG/i3iGoS3vUqriqNAcCN1WnbkwJV5NuDLbg0e7Ha02sCuY7vnMa3vmYMxm2gM3A3q6gg066tgDE Gg3WNIhSgOQZehBi1cj/IP/zdI6HCfmccUB7YN7RcphkHozhlLi7hkri5sH8amZHT8tcW7caYWcSDHCrib8iQtb1DhSlwSJVELnnShTWsxeyC3aIU47vJqBIwX1C/UJJrlHXEa6an2FItD9EsEHSUyKlE4PimmmcgJWzoCd+7f48CY7WxaX4AEm3Glcm2SGuLE3Ww7Z9cgWlLq1wUrqkxVSnf+rI65wZcofNknzOGCl1QLaOUiuDi5I9Vv1E9WL0 jmHGGrKVJa+VoOLX0Fpau9VvoL6US5CHA601inLG5w35a2PWScv/UxKrY2lQ4anlQwnV4CquVPGIpkMokplMo5rUNa6bVUC/e0/fWpnYyqkVF4jA09M7A3/XPa3F60vWHW462nW8633RQRomeSbUH2705iCyRpMymtVSCpa31qoEb5IyaXSGFgzXkL1nYnWuY524WVN6ngAt6tvUd6rvU96vvU2iqo2JwVzUzqweXSG4eXOi0CWuiyu1PWp7UU5Y mhVzT3xKKhT4YQDswsGmIUjQd9rkE2d5A258GjGk/pDQE6iP8dwLO7SYow20Wxe2ujh1UY4Sd8vmFKhLuEatMJWz21Y3z267nx6nHUlivG3bG1e0/w0ak02/AU9+NsWU9e2oSARW34GrXUq2tW2e3Hm325SLLlHaQRwCE5pFkqjVRwX5BAiNwTg0DIWURJjURgpznbotYpm2w8lD87Axw0q20JcnjW8Mmfms6gTXs68gIC8acTHcjh1bdMFCzAIW K8O7/z8SUXW5fIzVBHdmUhHAfJcEyt6cqDASzSvvZsLEo1IxA4DiQUJQAgOpDiQcYD4Afi1gQA8BDMjz76bZzVoOmo33MrB0NGmxlgS+Q3GKw+Jb7Y+KBodwS37T1Bb03SBFFb/jbUdb7Fkm6VbC2/UNMXpCA4mOS1mWgjDfOIWkgXxWgcGHS1wWVzSlAFArfdcwx662nrG6JWU9VBheGkDW53JGK3oCgAGgehDOwFSAU22DUzY9e11C5R2foxO3 ZtWRV/o1O2S5Q/WH69tV1vOpZK6xfVHOk51nOnsVeSqdXRkq603ayxmyG7fXNOnzWHxcGiCuTlTdIC7xr0BsS+dXpBC5CVzj2zu30OkrHxar3VRim0ATsGCwtI6sK37XmTj24z4xRF7TlExAWf6jG2Y6y+XY2zY03yqiU7GnAUgGje1IasSnPyyA2Iy6SliPREChnRTHgkWSk8uvl2yY+hr0WjJ7UizrVX2nrWSS3J35Owp3FO0p0/KCp3PAKp1z UdVXcKIV3Onfl3kNStVKi5mXaS1mUyCsS76S90R+y4B1bPXFpG1XZmM08NZQOrg0wAd4AUAZcAsk1cADALcC4AMiiFywkDPAUXZWCgsamM/uW1GjfVd0sF2jynfVV2qF2XbepDuCIVySyTyHVgezH2QHxCo9E5pkOuh3u68MXNAl7zEeS8g0cRkqsxfPoey2eR2QOrgCyczh1eEeGx0bMDLkUCYCE1HF6tbZ0J66fSyO1cVag9qVxLSQAVtIp3qE YHg+fZ4yL3Ze6r3de7mxLe473Pe4H3I+5vU0QYrs1HkBIsziTA1l3E6jI2lc4zWbdBSGCEWEHabN+AF6bE5SBL52LSzt3ggbt2kAIS2Tq6o2Buup1zqkeVW4MeV+bI26UEVyr5Y6N02CIHbuNMUq6QVjHoWSFBLka/VLc1+I41QZYbInNCOoP7I/gkfIVRAmRuQWt0VE9h63CqR33CwA3yOncLzupR2b2xbJVane01akR6OPCR6Cuhp5gYt96dMi +1MWzA0sW1SlsW2qqOu512wE113uuz10IqAkA+uo+60yup44exp7UG8ynptOtVWhAun3xTUUjmexatgKW3KK9zGquoWVcGgd0r3Ne4b3Ud273fe6H3Y+5F2k7HG695EoY/OHoE/zZICGk7TAKKHOoQXAjvd/Gc6vpQ1gQsmvhIZ1DGzF2UY/93hC+ei+NJsSTFPpDQg41xpo5sxt233Kh2xHXs+SuYq2LZ2L24rVxKpl00SxJXtEknXdFBm0Yajh pUel11uuj10HAL12Me313Ea6HA2MYz2Aoo4QcxFBkxcOsCABJ4qmyt5BTQiGmcCQGCtjAZResoojSIxkypUaqw1iRchO+SJrHCaoRzQ9DUuc+0GNQ4p6lPNqEVPDqFVPT0FcoPdwVwD1RweEW3G2qazFejPqsxQ5oXkITxEwKtzUwAPW4CUFkpRRr0sM2cXm22nWHo+nWnk7x1JcvhkbcBioBOu20lAS7aS2leR2GWnKglNGm59Wz0P8aapVUvhB L8iFD1IU71iahsBV9S71EwdLHAcZ3JYYxz2wGNKgue06luetIhvevTVb8glAekfmDgAEGBbADLDIgZdBaCaABAkROC6YKsD1ABgAwYCgDvATN244SsnzAFBGYgAzAiASFlfASljIgYikq0osKE+kX7qISlhY+9CVZuoPxo+/37E+yljfC/112ii91M+on2JIUn11G/6pU+ln0ZAMn0yGrfgC+nn0ZALRVJY8XDc+mn0ZAZuYoUsX2y+/QDgDZiXm nRX28QVn0curn3U+jX0ZALEj7ZMw7a+wX36AOH3uOxcXykdX0k+jIAAgLjXRvbraW+ylgeUVcAcEAQwE+u6YiOSqBNrP9r6uaXx1iUw1o+930IgBGKRkFRFmgYoj9yQ/WeqCABGAXKRqdCUwMAAgD1YXm34tU0BFUB30S+0cLMPAn0hgEgDYNPOAZQ3P3rRTIIF+4gAkkOtLKkJ5JteEv2bkIWDvAZ0DBtZQCBgSZJX63gCgTWeSDgMOTxAVZKJg KrDt/KoD2WRv24AZv13kWLh8ASgiTYTv3d+tP3M+pKDC+/EBUZIyzrEKBJVYe4kM2qyjqZKiBZ07ACyCiRWlACFZ6u+q5NYRm5p+uwA2wYGJoxZUhwAMv02+jf1MKLYDodEbK5SEtrx+0fAgkBA2K6367DOAwDO+v/D0Si/RQBcEAP+w0bOgJYrUUCZBgIn8DsIICBAAA=== ``` %%