From b63dcfcc9b80ac85415d2cda6a013600d5cfbbf7 Mon Sep 17 00:00:00 2001 From: Alex Blank <38751347+blankinator@users.noreply.github.com> Date: Mon, 4 Aug 2025 15:05:48 +0000 Subject: [PATCH] further work on introduction --- main.bib | 4104 ++++++++--------- thesis/main.tex | 6 +- ...se_case_contraception_decision_diagram.png | Bin 125554 -> 240656 bytes ...use_case_fertility_decision_diagram.drawio | 106 +- ...use_case_pregnancy_decision_diagram.drawio | 100 +- ...gy_use_case_pregnancy_decision_diagram.png | Bin 205433 -> 251452 bytes .../results/model_performance_overview.png | Bin 265522 -> 265660 bytes .../transformer_results_by_input_length.png | Bin 230786 -> 230786 bytes thesis/sections/background.tex | 201 +- thesis/sections/discussion.tex | 9 + thesis/sections/introduction.tex | 88 +- thesis/sections/methodology.tex | 147 +- thesis/sections/related_work.tex | 29 +- thesis/sections/results.tex | 28 +- 14 files changed, 2377 insertions(+), 2441 deletions(-) diff --git a/main.bib b/main.bib index 6ce423f..a3e8129 100644 --- a/main.bib +++ b/main.bib @@ -1,1261 +1,989 @@ -@misc{taylor_forecasting_2017, - title = {Forecasting at scale}, - copyright = {http://creativecommons.org/licenses/by/4.0/}, - url = {https://peerj.com/preprints/3190v2}, - doi = {10.7287/peerj.preprints.3190v2}, - abstract = {Forecasting is a common data science task that helps organizations with capacity planning, goal setting, and anomaly detection. Despite its importance, there are serious challenges associated with producing reliable and high quality forecasts –especially when there are a variety of time series and analysts with expertise in time series modeling are relatively rare. To address these challenges, we describe a practical approach to forecasting “at scale” that combines configurable models with analyst-in-the-loop performance analysis. We propose a modular regression model with interpretable parameters that can be intuitively adjusted by analysts with domain knowledge about the time series. We describe performance analyses to compare and evaluate forecasting procedures, and automatically flag forecasts for manual review and adjustment. Tools that help analysts to use their expertise most effectively enable reliable, practical forecasting of business time series.}, - language = {en}, - urldate = {2024-10-14}, - publisher = {PeerJ Preprints}, - author = {Taylor, Sean J and Letham, Benjamin}, - month = sep, - year = {2017}, - file = {PDF:/home/alex/Zotero/storage/GK5AIG2V/Taylor and Letham - 2017 - Forecasting at scale.pdf:application/pdf}, +@article{gaskins_predictors_2018, + title = {Predictors of sexual intercourse frequency among couples trying to conceive}, + volume = {15}, + issn = {1743-6095}, + url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5882561/}, + doi = {10.1016/j.jsxm.2018.02.005}, + abstract = {Background +Little is known about the predictors of sexual intercourse frequency ({SIF}) among couples trying to conceive despite the well-established link between {SIF} and fecundity. + +Aim +To evaluate the male and female demographic, occupational, and lifestyle predictors of {SIF} among couples. + +Methods +469 couples without a history of infertility participating in the Longitudinal Investigation of Fertility and the Environment Study (2005–2009) were followed for ≤1 year while trying to conceive. At enrollment, both partners were interviewed about demographic, occupational, lifestyle, and psychological characteristics using standardized questionnaires. Multivariable generalized linear mixed models with Poisson distribution was used to estimate the adjusted percent difference in {SIF} across exposure categories. + +Outcomes +{SIF} was recorded in daily journals and summarized as average {SIF} per month. + +Results +The median (interquartile range) {SIF} during follow-up was 6 (4–9) acts per month. For every year increase in female and male age, {SIF} decreased by −0.8\% (95\% {CI} −2.5, 1.0\%) and −1.7\% (95\% {CI} −3.1, −0.3\%). Women with high school education or less and those of non-White race had 34.4\% and 16.0\% higher {SIF}, respectively. A similar trend was seen for male education and race. Only couples where both partners (but not just one partner) worked rotating shifts had −39.1\% (95\% {CI} −61.0, −5.0\%) lower {SIF} compared to couples where neither partner worked rotating shifts. Male (but not female) exercise was associated with 13.2\% (95\% {CI} 1.7, 26.0\%) higher {SIF}. Diagnosis of a mood or anxiety disorder in the male (but not female) was associated with a 26.0\% (95\% {CI} −42.7, −4.4\%) lower {SIF}. Household income, smoking status, {BMI}, night work, alcohol intake, psychosocial stress were not associated with {SIF}. + +Clinical Implications +Even among couples trying to conceive, there was substantial variation in {SIF}. Both partners’ age, education, race, and rotating shift work as well as male exercise and mental health play an important role in determining {SIF}. + +Strengths \& Limitations +As this was a secondary analysis of an existing study, we lacked information on many pertinent psychological and relationship quality variables and the hormonal status of participants, which could have affected {SIF}. The unique population-based couple design, however, captured both partners’ demographics, occupational characteristics, lifestyle behaviors in advance of their daily, prospective reporting of {SIF}, which was a major strength. + +Conclusion +Important predictors of {SIF} among couples attempting to conceive include male exercise and mental health and both partners’ age, education, race, and rotating shift work.}, + pages = {519--528}, + number = {4}, + journaltitle = {J Sex Med}, + author = {Gaskins, Audrey J. and Sundaram, Rajeshwari and Buck Louis, Germaine M. and Chavarro, Jorge E.}, + urldate = {2025-07-30}, + date = {2018-04}, + pmid = {29523477}, + pmcid = {PMC5882561}, + file = {Full Text PDF:/home/alex/Zotero/storage/S3JV4TU2/Gaskins et al. - 2018 - Predictors of sexual intercourse frequency among couples trying to conceive.pdf:application/pdf}, } -@book{hutter_machine_2021, - address = {Cham}, - series = {Lecture {Notes} in {Computer} {Science}}, - title = {Machine {Learning} and {Knowledge} {Discovery} in {Databases}: {European} {Conference}, {ECML} {PKDD} 2020, {Ghent}, {Belgium}, {September} 14–18, 2020, {Proceedings}, {Part} {III}}, - volume = {12459}, - copyright = {https://www.springernature.com/gp/researchers/text-and-data-mining}, - isbn = {978-3-030-67663-6 978-3-030-67664-3}, - shorttitle = {Machine {Learning} and {Knowledge} {Discovery} in {Databases}}, - url = {https://link.springer.com/10.1007/978-3-030-67664-3}, - language = {en}, - urldate = {2024-10-14}, - publisher = {Springer International Publishing}, - editor = {Hutter, Frank and Kersting, Kristian and Lijffijt, Jefrey and Valera, Isabel}, - year = {2021}, - doi = {10.1007/978-3-030-67664-3}, - file = {Submitted Version:/home/alex/Zotero/storage/JMVJMLJ5/Hutter et al. - 2021 - Machine Learning and Knowledge Discovery in Databases European Conference, ECML PKDD 2020, Ghent, B.pdf:application/pdf}, +@article{pearl_factors_1933, + title = {{FACTORS} {IN} {HUMAN} {FERTILITY} {AND} {THEIR} {STATISTICAL} {EVALUATION}}, + volume = {222}, + rights = {https://www.elsevier.com/tdm/userlicense/1.0/}, + issn = {01406736}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0140673601186484}, + doi = {10.1016/S0140-6736(01)18648-4}, + pages = {607--611}, + number = {5741}, + journaltitle = {The Lancet}, + author = {Pearl, Raymond}, + urldate = {2025-07-30}, + date = {1933-09}, + langid = {english}, } -@incollection{hutter_general_2021, - address = {Cham}, - title = {A {General} {Machine} {Learning} {Framework} for {Survival} {Analysis}}, - volume = {12459}, - isbn = {978-3-030-67663-6 978-3-030-67664-3}, - url = {https://link.springer.com/10.1007/978-3-030-67664-3_10}, - language = {en}, - urldate = {2024-10-14}, - booktitle = {Machine {Learning} and {Knowledge} {Discovery} in {Databases}}, - publisher = {Springer International Publishing}, - author = {Bender, Andreas and Rügamer, David and Scheipl, Fabian and Bischl, Bernd}, - editor = {Hutter, Frank and Kersting, Kristian and Lijffijt, Jefrey and Valera, Isabel}, - year = {2021}, - doi = {10.1007/978-3-030-67664-3_10}, - note = {Series Title: Lecture Notes in Computer Science}, - pages = {158--173}, - file = {Submitted Version:/home/alex/Zotero/storage/WQIHZ7IP/Bender et al. - 2021 - A General Machine Learning Framework for Survival Analysis.pdf:application/pdf}, +@misc{goyal_accurate_2018, + title = {Accurate, Large Minibatch {SGD}: Training {ImageNet} in 1 Hour}, + url = {http://arxiv.org/abs/1706.02677}, + doi = {10.48550/arXiv.1706.02677}, + shorttitle = {Accurate, Large Minibatch {SGD}}, + abstract = {Deep learning thrives with large neural networks and large datasets. However, larger networks and larger datasets result in longer training times that impede research and development progress. Distributed synchronous {SGD} offers a potential solution to this problem by dividing {SGD} minibatches over a pool of parallel workers. Yet to make this scheme efficient, the per-worker workload must be large, which implies nontrivial growth in the {SGD} minibatch size. In this paper, we empirically show that on the {ImageNet} dataset large minibatches cause optimization difficulties, but when these are addressed the trained networks exhibit good generalization. Specifically, we show no loss of accuracy when training with large minibatch sizes up to 8192 images. To achieve this result, we adopt a hyper-parameter-free linear scaling rule for adjusting learning rates as a function of minibatch size and develop a new warmup scheme that overcomes optimization challenges early in training. With these simple techniques, our Caffe2-based system trains {ResNet}-50 with a minibatch size of 8192 on 256 {GPUs} in one hour, while matching small minibatch accuracy. Using commodity hardware, our implementation achieves {\textasciitilde}90\% scaling efficiency when moving from 8 to 256 {GPUs}. Our findings enable training visual recognition models on internet-scale data with high efficiency.}, + number = {{arXiv}:1706.02677}, + publisher = {{arXiv}}, + author = {Goyal, Priya and Dollár, Piotr and Girshick, Ross and Noordhuis, Pieter and Wesolowski, Lukasz and Kyrola, Aapo and Tulloch, Andrew and Jia, Yangqing and He, Kaiming}, + urldate = {2025-07-22}, + date = {2018-04-30}, + eprinttype = {arxiv}, + eprint = {1706.02677 [cs]}, + keywords = {Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Distributed, Parallel, and Cluster Computing}, + file = {Full Text PDF:/home/alex/Zotero/storage/5MKPLWI7/Goyal et al. - 2018 - Accurate, Large Minibatch SGD Training ImageNet in 1 Hour.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/TRYJFLDW/1706.html:text/html}, } -@misc{lightningai_pytorch_2024, - title = {{PyTorch} {Lightning}}, - url = {https://www.pytorchlightning.ai}, - urldate = {2024-10-14}, - author = {lightning.ai}, - year = {2024}, +@article{wen_time_2023, + title = {Time Series Prediction Based on {LSTM}-Attention-{LSTM} Model}, + volume = {11}, + issn = {2169-3536}, + url = {https://ieeexplore.ieee.org/document/10124729/}, + doi = {10.1109/ACCESS.2023.3276628}, + abstract = {Time series forecasting uses data from the past periods of time to predict future information, which is of great significance in many applications. Existing time series forecasting methods still have problems such as low accuracy when dealing with some non-stationary multivariate time series data forecasting. Aiming at the shortcomings of existing methods, in this paper we propose a new time series forecasting model {LSTM}-attention-{LSTM}. The model uses two {LSTM} models as the encoder and decoder, and introduces an attention mechanism between the encoder and decoder. The model has two distinctive features: first, by using the attention mechanism to calculate the interrelationship between sequence data, it overcomes the disadvantage of the coder-and-decoder model in that the decoder cannot obtain sufficiently long input sequences; second, it is suitable for sequence forecasting with long time steps. In this paper we validate the proposed model based on several real data sets, and the results show that the {LSTM}-attention-{LSTM} model is more accurate than some currently dominant models in prediction. The experiment also assessed the effect of the attention mechanism at different time steps by varying the time step.}, + pages = {48322--48331}, + journaltitle = {{IEEE} Access}, + author = {Wen, Xianyun and Li, Weibang}, + urldate = {2025-07-22}, + date = {2023}, + keywords = {attention mechanisms, Autoregressive processes, Data models, Decoding, encoder and decoder model, Forecasting, Logic gates, long short-term memory networks, Predictive models, Time series analysis, Time series forecasting}, + file = {Full Text PDF:/home/alex/Zotero/storage/3M54PVSE/Wen and Li - 2023 - Time Series Prediction Based on LSTM-Attention-LSTM Model.pdf:application/pdf}, } -@misc{alexandrov_gluonts_2019, - title = {{GluonTS}: {Probabilistic} {Time} {Series} {Models} in {Python}}, - shorttitle = {{GluonTS}}, - url = {http://arxiv.org/abs/1906.05264}, - abstract = {We introduce Gluon Time Series (GluonTS, available at https://gluon-ts.mxnet.io), a library for deep-learning-based time series modeling. GluonTS simplifies the development of and experimentation with time series models for common tasks such as forecasting or anomaly detection. It provides all necessary components and tools that scientists need for quickly building new models, for efficiently running and analyzing experiments and for evaluating model accuracy.}, - urldate = {2024-10-14}, - publisher = {arXiv}, - author = {Alexandrov, Alexander and Benidis, Konstantinos and Bohlke-Schneider, Michael and Flunkert, Valentin and Gasthaus, Jan and Januschowski, Tim and Maddix, Danielle C. and Rangapuram, Syama and Salinas, David and Schulz, Jasper and Stella, Lorenzo and Türkmen, Ali Caner and Wang, Yuyang}, - month = jun, - year = {2019}, - note = {arXiv:1906.05264}, - keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, - file = {Preprint PDF:/home/alex/Zotero/storage/JP9K74A8/Alexandrov et al. - 2019 - GluonTS Probabilistic Time Series Models in Python.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/RJYSBT29/1906.html:text/html}, +@misc{pham_dropout_2014, + title = {Dropout improves Recurrent Neural Networks for Handwriting Recognition}, + url = {http://arxiv.org/abs/1312.4569}, + doi = {10.48550/arXiv.1312.4569}, + abstract = {Recurrent neural networks ({RNNs}) with Long Short-Term memory cells currently hold the best known results in unconstrained handwriting recognition. We show that their performance can be greatly improved using dropout - a recently proposed regularization method for deep architectures. While previous works showed that dropout gave superior performance in the context of convolutional networks, it had never been applied to {RNNs}. In our approach, dropout is carefully used in the network so that it does not affect the recurrent connections, hence the power of {RNNs} in modeling sequence is preserved. Extensive experiments on a broad range of handwritten databases confirm the effectiveness of dropout on deep architectures even when the network mainly consists of recurrent and shared connections.}, + number = {{arXiv}:1312.4569}, + publisher = {{arXiv}}, + author = {Pham, Vu and Bluche, Théodore and Kermorvant, Christopher and Louradour, Jérôme}, + urldate = {2025-07-22}, + date = {2014-03-10}, + eprinttype = {arxiv}, + eprint = {1312.4569 [cs]}, + keywords = {Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Neural and Evolutionary Computing}, + file = {Full Text PDF:/home/alex/Zotero/storage/IA52LNE8/Pham et al. - 2014 - Dropout improves Recurrent Neural Networks for Handwriting Recognition.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/LJ8EJ5PS/1312.html:text/html}, } -@misc{cho_learning_2014, - title = {Learning {Phrase} {Representations} using {RNN} {Encoder}-{Decoder} for {Statistical} {Machine} {Translation}}, - url = {http://arxiv.org/abs/1406.1078}, - abstract = {In this paper, we propose a novel neural network model called RNN Encoder-Decoder that consists of two recurrent neural networks (RNN). One RNN encodes a sequence of symbols into a fixed-length vector representation, and the other decodes the representation into another sequence of symbols. The encoder and decoder of the proposed model are jointly trained to maximize the conditional probability of a target sequence given a source sequence. The performance of a statistical machine translation system is empirically found to improve by using the conditional probabilities of phrase pairs computed by the RNN Encoder-Decoder as an additional feature in the existing log-linear model. Qualitatively, we show that the proposed model learns a semantically and syntactically meaningful representation of linguistic phrases.}, - urldate = {2024-10-10}, - publisher = {arXiv}, - author = {Cho, Kyunghyun and Merrienboer, Bart van and Gulcehre, Caglar and Bahdanau, Dzmitry and Bougares, Fethi and Schwenk, Holger and Bengio, Yoshua}, - month = sep, - year = {2014}, - note = {arXiv:1406.1078}, - keywords = {Computer Science - Computation and Language, Computer Science - Machine Learning, Statistics - Machine Learning, Computer Science - Neural and Evolutionary Computing}, - file = {Preprint PDF:/home/alex/Zotero/storage/E8WMK2IN/Cho et al. - 2014 - Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/6PTCL8LW/1406.html:text/html}, +@article{wallach_prediction_1980, + title = {Prediction and Detection of Ovulation}, + volume = {34}, + issn = {00150282}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216448880}, + doi = {10.1016/S0015-0282(16)44888-0}, + pages = {89--98}, + number = {2}, + journaltitle = {Fertility and Sterility}, + author = {Wallach, Edward and Moghissi, Kamran S.}, + urldate = {2025-07-15}, + date = {1980-08}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/JREACGRA/Wallach and Moghissi - 1980 - Prediction and Detection of Ovulation.pdf:application/pdf}, +} + +@article{guida_efficacy_1999, + title = {Efficacy of methods for determining ovulation in a natural family planning program}, + volume = {72}, + issn = {00150282}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028299003659}, + doi = {10.1016/S0015-0282(99)00365-9}, + abstract = {Objective: To evaluate the efficacy in ovulation detection of methods used in natural family planning in comparison with pelvic ultrasonography. Design: Prospective analysis of ovulation detection by natural family planning methods and ultrasonography. Setting: Natural family planning clinic, Department of Obstetrics and Gynecology, University of Naples “Federico {II}”. Patient(s): Forty healthy women who were highly motivated to use natural family planning. Intervention(s): None. Main Outcome Measure(s): Transvaginal ultrasonographic findings, urinary {LH} levels, salivary b-glucuronidase activity, salivary ferning levels and characteristics of cervical mucus, and {BBT}. +Result(s): Urinary {LH} level determination yielded a 100\% correlation with the simultaneous ultrasonographic diagnosis of ovulation. Mucus sensations and characteristics yielded a 48.3\% correlation when simultaneously evaluated with ovulation. b-Glucuronidase levels yielded a 27.7\% correlation. The salivary ferning test had a 36.8\% ovulation-detection rate the day of ovulation, but 58.7\% of results were uninterpretable. Body temperature measurements yielded a 30.4\% correlation with the simultaneous ultrasonographic diagnosis of ovulation. +Conclusion(s): Measuring urinary {LH} levels is an excellent method for determining ovulation. Although variations in mucus characteristics and basal body temperature correlate somewhat with ovulation, the length of the fertile period is overestimated with these methods. The salivary ferning test and measurement of b-glucuronidase levels are not good methods for home ovulation testing. (Fertil Sterilt 1999;72:900 – 4. ©1999 by American Society for Reproductive Medicine.)}, + pages = {900--904}, + number = {5}, + journaltitle = {Fertility and Sterility}, + author = {Guida, Maurizio and Tommaselli, Giovanni A and Palomba, Stefano and Pellicano, Massimiliano and Moccia, Gianfranco and Di Carlo, Costantino and Nappi, Carmine}, + urldate = {2025-07-15}, + date = {1999-11}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/UAS9AGTU/Guida et al. - 1999 - Efficacy of methods for determining ovulation in a natural family planning program.pdf:application/pdf}, +} + +@article{thigpen_oura_2025, + title = {Oura Ring as a Tool for Ovulation Detection: Validation Analysis}, + volume = {27}, + issn = {1438-8871}, + url = {https://www.jmir.org/2025/1/e60667}, + doi = {10.2196/60667}, + shorttitle = {Oura Ring as a Tool for Ovulation Detection}, + abstract = {Background: Oura Ring is a wearable device that estimates ovulation dates using physiology data recorded from the finger. Estimating the ovulation date can aid fertility management for conception or nonhormonal contraception and provides insights into follicular and luteal phase lengths. Across the reproductive lifespan, changes in these phase lengths can serve as a biomarker for reproductive health. +Objective: We assessed the strengths, weaknesses, and limitations of using physiology from the Oura Ring to estimate the ovulation date. We compared performance across cycle length, cycle variability, and participant age. In each subgroup, we compared the algorithm’s performance with the traditional calendar method, which estimates the ovulation date based on an individual’s last period start date and average menstrual cycle length. +Methods: The study sample contained 1155 ovulatory menstrual cycles from 964 participants recruited from the Oura Ring commercial database. Ovulation prediction kits served as a benchmark to evaluate the performance. The Fisher test was used to determine an odds ratio to assess if ovulation detection rate significantly differed between methods or subgroups. The Mann-Whitney U test was used to determine if the accuracy of the estimated ovulation date differed between the estimated and reference ovulation dates. +Results: The physiology method detected 1113 (96.4\%) of 1155 ovulations with an average error of 1.26 days, which was significantly lower (U=904942.0, P{\textless}.001) than the calendar method’s average error of 3.44 days. The physiology method had significantly better accuracy across all cycle lengths, cycle variability groups, and age groups compared with the calendar method (P{\textless}.001). The physiology method detected fewer ovulations in short cycles (odds ratio 3.56, 95\% {CI} 1.65-8.06; P=.008) but did not differ between typical and long or abnormally long cycles. Abnormally long cycle lengths were associated with decreased accuracy (U=22,383, P=.03), with a mean absolute error of 1.7 ({SEM} .09) days compared with 1.18 ({SEM} .02) days. The physiology method was not associated with differences in accuracy across age or typical cycle variability, while the calendar method performed significantly worse in participants with irregular cycles (U=21,643, P{\textless}.001). +Conclusions: The physiology method demonstrated superior accuracy over the calendar method, with approximately 3-fold improvement. Calendar-based fertility tracking could be used as a backup in cases of insufficient physiology data but should be used with caution, particularly for individuals with irregular menstrual cycles. Our analyses suggest the physiology method can reliably estimate ovulation dates for adults aged 18-52 years, across a variety of cycle lengths, and in users with regular or irregular cycles. This method may be used as a tool to improve fertile window estimation, which can aid in conceiving or preventing pregnancies. This method also offers a low-effort solution for follicular and luteal phase length tracking, which are key biomarkers for reproductive health.}, + pages = {e60667}, + journaltitle = {J Med Internet Res}, + author = {Thigpen, Nina and Patel, Shyamal and Zhang, Xi}, + urldate = {2025-07-15}, + date = {2025-01-31}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/TECKQLN7/Thigpen et al. - 2025 - Oura Ring as a Tool for Ovulation Detection Validation Analysis.pdf:application/pdf}, +} + +@online{noauthor_pytorch_nodate, + title = {{PyTorch}}, + url = {https://pytorch.org/}, + abstract = {{PyTorch} Foundation is the deep learning community home for the open source {PyTorch} framework and ecosystem.}, + titleaddon = {{PyTorch}}, + urldate = {2025-07-11}, + langid = {american}, + file = {Snapshot:/home/alex/Zotero/storage/K7CTLE77/pytorch.org.html:text/html}, +} + +@article{li_menstrual_2023, + title = {Menstrual cycle length variation by demographic characteristics from the Apple Women’s Health Study}, + volume = {6}, + issn = {2398-6352}, + url = {https://www.nature.com/articles/s41746-023-00848-1}, + doi = {10.1038/s41746-023-00848-1}, + abstract = {Abstract + + Menstrual characteristics are important signs of women’s health. Here we examine the variation of menstrual cycle length by age, ethnicity, and body weight using 165,668 cycles from 12,608 participants in the {US} using mobile menstrual tracking apps. After adjusting for all covariates, mean menstrual cycle length is shorter with older age across all age groups until age 50 and then became longer for those age 50 and older. Menstrual cycles are on average 1.6 (95\%{CI}: 1.2, 2.0) days longer for Asian and 0.7 (95\%{CI}: 0.4, 1.0) days longer for Hispanic participants compared to white non-Hispanic participants. Participants with {BMI} ≥ 40 kg/m + 2 + have 1.5 (95\%{CI}: 1.2, 1.8) days longer cycles compared to those with {BMI} between 18.5 and 25 kg/m + 2 + . Cycle variability is the lowest among participants aged 35–39 but are considerably higher by 46\% (95\%{CI}: 43\%, 48\%) and 45\% (95\%{CI}: 41\%, 49\%) among those aged under 20 and between 45–49. Cycle variability increase by 200\% (95\%{CI}: 191\%, 210\%) among those aged above 50 compared to those in the 35–39 age group. Compared to white participants, those who are Asian and Hispanic have larger cycle variability. Participants with obesity also have higher cycle variability. Here we confirm previous observations of changes in menstrual cycle pattern with age across reproductive life span and report new evidence on the differences of menstrual variation by ethnicity and obesity status. Future studies should explore the underlying determinants of the variation in menstrual characteristics.}, + pages = {100}, + number = {1}, + journaltitle = {npj Digit. Med.}, + author = {Li, Huichu and Gibson, Elizabeth A. and Jukic, Anne Marie Z. and Baird, Donna D. and Wilcox, Allen J. and Curry, Christine L. and Fischer-Colbrie, Tyler and Onnela, Jukka-Pekka and Williams, Michelle A. and Hauser, Russ and Coull, Brent A. and Mahalingaiah, Shruthi}, + urldate = {2025-07-04}, + date = {2023-05-29}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/9J5N6YIW/Li et al. - 2023 - Menstrual cycle length variation by demographic characteristics from the Apple Women’s Health Study.pdf:application/pdf}, +} + +@article{ecochard_menstrual_2024, + title = {The menstrual cycle is influenced by weekly and lunar rhythms}, + volume = {121}, + issn = {00150282}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028223020769}, + doi = {10.1016/j.fertnstert.2023.12.009}, + abstract = {Objective: To study whether the menstrual cycle has a circaseptan (7 days) rhythm and whether it is associated with the lunar cycle (also defined as the synodic month, it is the cycle of the phases of the Moon as seen from Earth, averaging 29.5 days in length). Design: Cross-sectional study. Subjects: A total of 35,940 European and North American women aged 18–40 years. Exposure: Data were collected in real-life conditions. Intervention: No intervention was performed. Main Outcome Measure: The onset of menstruation was assessed in prospectively measured menstrual cycles (311,064 cycles) over 3 full years (2019–2021). Associations were calculated between the onset of menstruation and the day of the week, and between the onset of menstruation and the lunar phase. +Results: In this large data set, a circaseptan (7-day) rhythmicity of menstruation was observed, with a peak (acrophase) of menstrual onset on Thursdays and Fridays. This circaseptan rhythm was observed in every age group, in every phase of the lunar cycle, and in all seasons. This feature was most pronounced for cycle durations between 27 and 29 days. In winter, the circaseptan rhythm was found in cycles of 27–29 days, but not in other cycle lengths. A circalunar rhythm was also statistically significant, but not as clearly defined as the circaseptan rhythm. The peak (acrophase) of the circalunar rhythm of menstrual onset varied according to the season. In addition, there was a small but statistically significant interaction between the circaseptan rhythm and the lunar cycle. +Conclusion: Although relatively small in amplitude, the weekly rhythm of menstruation was statistically significant. Menstruation occurs more often on Thursdays and Fridays than on other days of the week. This is particularly true for women whose cycles last between 27 and 29 days. Circalunar rhythmicity was also statistically significant. However, it is less pronounced than the weekly rhythm. (Fertil {SterilÒ} 2024;121:651-9. Ó2023 by American Society for Reproductive Medicine.)}, + pages = {651--659}, + number = {4}, + journaltitle = {Fertility and Sterility}, + author = {Ecochard, René and Leiva, Rene and Bouchard, Thomas P. and Van Lamsweerde, Agathe and Pearson, Jack T. and Stanford, Joseph B. and Gronfier, Claude}, + urldate = {2025-07-03}, + date = {2024-04}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/T6DNB4M9/Ecochard et al. - 2024 - The menstrual cycle is influenced by weekly and lunar rhythms.pdf:application/pdf}, +} + +@online{noauthor_kegg_nodate, + title = {kegg® fertility monitor \& kegel ball {\textbar} Track Key Fertility Metric}, + url = {https://kegg.tech/}, + abstract = {Plan your pregnancy with confidence ... kegg® is a medical-grade fertility device that gives you accurate and personalized fertility tracking through cervical mucus.}, + titleaddon = {kegg}, + urldate = {2025-07-02}, + langid = {english}, + file = {Snapshot:/home/alex/Zotero/storage/HNFFJPSN/kegg.tech.html:text/html}, +} + +@article{moreno_temporal_1988, + title = {{TEMPORAL} {RELATION} {OF} {OVDLATION} {TO} {SALIVARY} {AND} {VAGINAL} {ELECTRICAL} {RESISTANCE} {PATTERNS}: {IMPLICATIONS} {FOR} {NATURAL} {FAMILY} {PLANNING}}, + abstract = {An independent assessment of the {CUETM} Monitor (Zetek, Aurora, Colorado) as an ovulation predictor was made with emphasis on its potential role in "natural family planning". The device provides a digital measurement of the electrical resistance of saliva and vaginal secretions. Twenty-nine menstrual cycles from 11 regularly cycling women were monitored with basal temperatures, urinary {LH}, pelvic ultrasound and the {CUE} monitor. Patterns of peak salivary electrical resistance were able to predict ovulation on average 5.3 (51.9 {SD}) days in advance. Despite variations in total length of the follicular phase from cycle to cycle, the within-subject variation of this predictive interval was quite small. Nadirs in the electrical resistance of vaginal secretions occurred within 2 days of ovulation in all but one patient. Variation in this interval from cycle-tocycle was small as well. We propose an algorithm for the use of these intervals in "natural family planning" that could safely reduce the monthly abstinence period of present methods. The simplicity, objectivity and consistency of this device could result in their greater general acceptance.}, + author = {Moreno, Jorge E and Doody, Michael C and Besch, Paige}, + date = {1988-10}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/FFW8KBZQ/Moreno et al. - TEMPORAL RELATION OF OVDLATION TO SALIVARY AND VAGINAL ELECTRICAL RESISTANCE PATTERNS IMPLICATIONS.pdf:application/pdf}, +} + +@online{noauthor_trackle_nodate, + title = {trackle - einfach hormonfrei verhüten}, + url = {https://trackle.de/}, + abstract = {Das trackle Sensorsystem hilft Dir, einfach, sicher und hormonfrei zu verhüten. Jetzt informieren und symptothermale Methode nutzen!}, + titleaddon = {trackle}, + urldate = {2025-07-02}, + langid = {german}, + file = {Snapshot:/home/alex/Zotero/storage/KNFY8X2K/trackle.de.html:text/html}, +} + +@article{zhu_accuracy_2021, + title = {The Accuracy of Wrist Skin Temperature in Detecting Ovulation Compared to Basal Body Temperature: Prospective Comparative Diagnostic Accuracy Study}, + volume = {23}, + issn = {1438-8871}, + url = {https://www.jmir.org/2021/6/e20710}, + doi = {10.2196/20710}, + shorttitle = {The Accuracy of Wrist Skin Temperature in Detecting Ovulation Compared to Basal Body Temperature}, + abstract = {Background: As a daily point measurement, basal body temperature ({BBT}) might not be able to capture the temperature shift in the menstrual cycle because a single temperature measurement is present on the sliding scale of the circadian rhythm. Wrist skin temperature measured continuously during sleep has the potential to overcome this limitation. +Objective: This study compares the diagnostic accuracy of these two temperatures for detecting ovulation and to investigate the correlation and agreement between these two temperatures in describing thermal changes in menstrual cycles. +Methods: This prospective study included 193 cycles (170 ovulatory and 23 anovulatory) collected from 57 healthy women. Participants wore a wearable device (Ava Fertility Tracker bracelet 2.0) that continuously measured the wrist skin temperature during sleep. Daily {BBT} was measured orally and immediately upon waking up using a computerized fertility tracker with a digital thermometer (Lady-Comp). An at-home luteinizing hormone test was used as the reference standard for ovulation. The diagnostic accuracy of using at least one temperature shift detected by the two temperatures in detecting ovulation was evaluated. For ovulatory cycles, repeated measures correlation was used to examine the correlation between the two temperatures, and mixed effect models were used to determine the agreement between the two temperature curves at different menstrual phases. +Results: Wrist skin temperature was more sensitive than {BBT} (sensitivity 0.62 vs 0.23; P{\textless}.001) and had a higher true-positive rate (54.9\% vs 20.2\%) for detecting ovulation; however, it also had a higher false-positive rate (8.8\% vs 3.6\%), resulting in lower specificity (0.26 vs 0.70; P=.002). The probability that ovulation occurred when at least one temperature shift was detected was 86.2\% for wrist skin temperature and 84.8\% for {BBT}. Both temperatures had low negative predictive values (8.8\% for wrist skin temperature and 10.9\% for {BBT}). Significant positive correlation between the two temperatures was only found in the follicular phase (rmcorr correlation coefficient=0.294; P=.001). Both temperatures increased during the postovulatory phase with a greater increase in the wrist skin temperature (range of increase: 0.50 °C vs 0.20 °C). During the menstrual phase, the wrist skin temperature exhibited a greater and more rapid decrease (from 36.13 °C to 35.80 °C) than {BBT} (from 36.31 °C to 36.27 °C). During the preovulatory phase, there were minimal changes in both temperatures and small variations in the estimated daily difference between the two temperatures, indicating an agreement between the two curves. +Conclusions: For women interested in maximizing the chances of pregnancy, wrist skin temperature continuously measured during sleep is more sensitive than {BBT} for detecting ovulation. The difference in the diagnostic accuracy of these methods was likely attributed to the greater temperature increase in the postovulatory phase and greater temperature decrease during the menstrual phase for the wrist skin temperatures.}, + pages = {e20710}, + number = {6}, + journaltitle = {J Med Internet Res}, + author = {Zhu, Tracy Y and Rothenbühler, Martina and Hamvas, Györgyi and Hofmann, Anja and Welter, {JoEllen} and Kahr, Maike and Kimmich, Nina and Shilaih, Mohaned and Leeners, Brigitte}, + urldate = {2025-07-02}, + date = {2021-06-08}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/PY7HCR3K/Zhu et al. - 2021 - The Accuracy of Wrist Skin Temperature in Detecting Ovulation Compared to Basal Body Temperature Pr.pdf:application/pdf}, +} + +@article{shilaih_modern_2018, + title = {Modern fertility awareness methods: wrist wearables capture the changes in temperature associated with the menstrual cycle}, + volume = {38}, + issn = {0144-8463}, + url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6265623/}, + doi = {10.1042/BSR20171279}, + shorttitle = {Modern fertility awareness methods}, + abstract = {Core and peripheral body temperatures are affected by changes in reproductive hormones during the menstrual cycle. Women worldwide use the basal body temperature ({BBT}) method to aid and prevent conception. However, prior research suggests that taking one’s daily temperature can prove inconvenient and subject to environmental factors. We investigate whether a more automatic, non-invasive temperature measurement system can detect changes in temperature across the menstrual cycle. We examined how wrist skin temperature ({WST}), measured with wearable sensors, correlates with urinary tests of ovulation and may serve as a new method of fertility tracking. One hundred and thirty-six eumenorrheic, non-pregnant women participated in an observational study. Participants wore {WST} biosensors during sleep and reported their daily activities. An at-home luteinizing hormone ({LH}) test was used to confirm ovulation. {WST} was recorded across 437 cycles (mean cycles/participant = 3.21, S.D. = 2.25). We tested the relationship between the fertile window and {WST} temperature shifts, using the {BBT} three-over-six rule. A sustained 3-day temperature shift was observed in 357/437 cycles (82\%), with the lowest cycle temperature occurring in the fertile window 41\% of the time. Most temporal shifts (307/357, 86\%) occurred on ovulation day ({OV}) or later. The average early-luteal phase temperature was 0.33°C higher than in the fertile window. Menstrual cycle changes in {WST} were impervious to lifestyle factors, like having sex, alcohol, or eating prior to bed, that, in prior work, have been shown to obfuscate {BBT} readings. Although currently costlier than {BBT}, the present study suggests that {WST} could be a promising, convenient parameter for future multiparameter fertility awareness methods.}, + pages = {BSR20171279}, + number = {6}, + journaltitle = {Biosci Rep}, + author = {Shilaih, Mohaned and Goodale, Brianna M. and Falco, Lisa and Kübler, Florian and De Clerck, Valerie and Leeners, Brigitte}, + urldate = {2025-07-02}, + date = {2018-11-30}, + pmid = {29175999}, + pmcid = {PMC6265623}, + file = {Full Text PDF:/home/alex/Zotero/storage/CXNSXAFC/Shilaih et al. - 2018 - Modern fertility awareness methods wrist wearables capture the changes in temperature associated wi.pdf:application/pdf}, +} + +@online{sl_ava_nodate, + title = {Ava Fertility Tracker}, + url = {https://www.avawomen.com/}, + abstract = {See your 5 best days to conceive in real-time. Go beyond ovulation day, and make use of your full fertile window to increase your chances of pregnancy}, + titleaddon = {{AvaWomen}}, + author = {S.L, Ava Women}, + urldate = {2025-07-02}, + langid = {english}, + file = {Snapshot:/home/alex/Zotero/storage/SMIE7YIR/www.avawomen.com.html:text/html}, +} + +@online{noauthor_fact_sheet_studie_210621_2025, + title = {fact\_sheet\_studie\_210621}, + url = {https://dfxyyqidohkoi.cloudfront.net/media/filer_public/ff/86/ff8646d2-8d33-445c-9148-5979b60abbae/fact_sheet_studie_210621.pdf}, + shorttitle = {daysy\_fact\_sheet}, + urldate = {2025-07-02}, + date = {2025-07-02}, + file = {PDF:/home/alex/Zotero/storage/8VDBFS2G/fact_sheet_studie_210621.pdf:application/pdf}, +} + +@online{electronics_zykluscomputer_nodate, + title = {Zykluscomputer Daysy - 100 \% natürlich und sehr genau!}, + url = {https://de.daysy.me/}, + abstract = {Daysy Zykluscomputer - einfach, hormonfrei \& über 99\% genau. ✓ Medizinprodukt zur Berechnung Deiner fruchtbaren Tage ✓ Natürliche Familienplanung ✓ Erhöhe Deine Lebensqualität!}, + author = {Electronics, Valley}, + urldate = {2025-07-02}, + langid = {german}, + file = {Snapshot:/home/alex/Zotero/storage/ZGIZ669S/de.daysy.me.html:text/html}, +} + +@online{noauthor_natural_nodate, + title = {Natural Cycles: Natural Birth Control {\textbar} No Hormones or Side Effects}, + url = {https://www.naturalcycles.com}, + shorttitle = {Natural Cycles}, + abstract = {Natural Cycles birth control is 93\% effective with typical use and 98\% effective with perfect use. Learn more about hormone-free birth control today.}, + titleaddon = {Natural Cycles}, + urldate = {2025-07-02}, + langid = {american}, + file = {Snapshot:/home/alex/Zotero/storage/MCUDFIHC/www.naturalcycles.com.html:text/html}, +} + +@article{weiss_confirmation_2022, + title = {Confirmation of human ovulation in assisted reproduction using an adhesive axillary thermometer ({femSense}®)}, + volume = {4}, + issn = {2673-253X}, + url = {https://www.frontiersin.org/articles/10.3389/fdgth.2022.930010/full}, + doi = {10.3389/fdgth.2022.930010}, + abstract = {Objective + Timing for sexual intercourse is important in achieving pregnancy in natural menstrual cycles. Different methods of detecting the fertile window have been invented, among them luteinization hormone ({LH}) to predict ovulation and biphasic body basal temperature ({BBT}) to confirm ovulation retrospectively. The gold standard to detect ovulation in gynecology practice remains transvaginal ultrasonography in combination with serum progesterone. In this study we evaluated a wearable temperature sensing patch ({femSense}®) using continuous body temperature measurement to confirm ovulation and determine the end of the fertile window. + + + Methods + 96 participants received the {femSense}® system consisting of an adhesive axillary thermometer patch and a smartphone application, where patients were asked to document information about their previous 3 cycles. Based on the participants data, the app predicted the cycle length and the estimated day of ovulation. From these predictions, the most probable fertile window and the day for applying the patch were derived. Participants applied and activated the {femSense}® patch on the calculated date, from which the patch continuously recorded their body temperature throughout a period of up to 7 days to confirm ovulation. Patients documented their daily urinary {LH} test positivity, and a transvaginal ultrasound was performed on day cycle day 7, 10, 12 and 14/15 to investigate the growth of one dominant follicle. If a follicle reached 15 mm in diameter, an ultrasound examination was carried out every day consecutively until ovulation. On the day ovulation was detected, serum progesterone was measured to confirm the results of the ultrasound. The performance of {femSense}® was evaluated by comparing the day of ovulation confirmation with the results of ovulation prediction ({LH} test) and detection (transvaginal ultrasound). + + + Results + + The {femSense}® system confirmed ovulation occurrence in 60 cases (81.1\%) compared to 48 predicted cases (64.9\%) with the {LH} test ( + p +  = 0.041). Subgroup analysis revealed a positive trend for the {femSense}® system of specific ovulation confirmation within the fertile window of 24 h after ovulation in 42 of 74 cases (56.8\%). Cycle length, therapy method or infertility reason of the patient did not influence accuracy of the {femSense}® system. + + + + Conclusions + The {femSense}® system poses a promising alternative to the traditional {BBT} method and is a valuable surrogate marker to transvaginal ultrasound for confirmation of ovulation.}, + pages = {930010}, + journaltitle = {Front. Digit. Health}, + author = {Weiss, Gregor and Strohmayer, Karl and Koele, Werner and Reinschissler, Nina and Schenk, Michael}, + urldate = {2025-07-02}, + date = {2022-09-19}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/C3IU2JVF/Weiss et al. - 2022 - Confirmation of human ovulation in assisted reproduction using an adhesive axillary thermometer (fem.pdf:application/pdf}, +} + +@article{moghissi_accuracy_1976, + title = {Accuracy of Basal Body Temperature for Ovulation Detection}, + volume = {27}, + issn = {00150282}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216422570}, + doi = {10.1016/S0015-0282(16)42257-0}, + pages = {1415--1421}, + number = {12}, + journaltitle = {Fertility and Sterility}, + author = {Moghissi, Kamran S.}, + urldate = {2025-07-02}, + date = {1976-12}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/GKF98BHR/Moghissi - 1976 - Accuracy of Basal Body Temperature for Ovulation Detection.pdf:application/pdf}, +} + +@article{bauman_basal_1981, + title = {Basal Body Temperature: Unreliable Method of Ovulation Detection}, + volume = {36}, + issn = {00150282}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216459169}, + doi = {10.1016/S0015-0282(16)45916-9}, + shorttitle = {Basal Body Temperature}, + pages = {729--733}, + number = {6}, + journaltitle = {Fertility and Sterility}, + author = {Bauman, Joan E.}, + urldate = {2025-07-01}, + date = {1981-12}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/9T7V8ACT/Bauman - 1981 - Basal Body Temperature Unreliable Method of Ovulation Detection.pdf:application/pdf}, +} + +@article{luo_detection_2020, + title = {Detection and Prediction of Ovulation From Body Temperature Measured by an In-Ear Wearable Thermometer}, + volume = {67}, + rights = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/{IEEE}.html}, + issn = {0018-9294, 1558-2531}, + url = {https://ieeexplore.ieee.org/document/8715448/}, + doi = {10.1109/TBME.2019.2916823}, + abstract = {Objective: We present a non-invasive wearable device for fertility monitoring and propose an effective and flexible statistical learning algorithm to detect and predict ovulation using data captured by this device. Methods: The system consists of an earpiece, which measures the ear canal temperature every 5 minutes during night sleep hours, and a base station that transmits data to a smartphone application for analysis. We establish a data-cleaning protocol for data preprocessing and then fit a Hidden Markov Model ({HMM}) with two hidden states of high and low temperature to identify the more probable state of each time point via the predicted probabilities. Finally, a post-processing procedure is developed to incorporate biorhythm information to form a time-course biphasic profile for each subject. Results: The performance of the proposed algorithms applied to data collected by the device are compared with traditional methods in terms of match rate with self-reported ovulation days confirmed with an Ovulation Test Kit. Empirical study results from a group of 34 users yielded significant improvements over the traditional methods in terms of detection accuracy (with sensitivity 92.31\%) and prediction power (23.0731.55\% higher). Conclusion: We demonstrated the feasibility for reliable ovulation detection and prediction with high-frequency temperature data collected by a non-invasive wearable device. Significance: Traditional fertility monitoring methods are often either inaccurate or inconvenient. The wearable device and learning algorithm presented in this paper provides a userfriendly and reliable platform for tracking ovulation, which may have a broad impact on both fertility research and real-world family planning.}, + pages = {512--522}, + number = {2}, + journaltitle = {{IEEE} Trans. Biomed. Eng.}, + author = {Luo, Lan and She, Xichen and Cao, Jiexuan and Zhang, Yunlong and Li, Yijiang and Song, Peter X. K.}, + urldate = {2025-07-01}, + date = {2020-02}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/YLIR4YNX/Luo et al. - 2020 - Detection and Prediction of Ovulation From Body Temperature Measured by an In-Ear Wearable Thermomet.pdf:application/pdf}, +} + +@article{alliende_cervicovaginal_2005, + title = {Cervicovaginal fluid changes to detect ovulation accurately}, + volume = {193}, + rights = {https://www.elsevier.com/tdm/userlicense/1.0/}, + issn = {00029378}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0002937804018770}, + doi = {10.1016/j.ajog.2004.11.006}, + abstract = {Objective: The purpose of this study was to evaluate changes in cervicovaginal fluid characteristics to identify ovulation. Study design: Several ovulation indicators were studied in a university-based natural family planning center. Fifteen parous women during 29 ovulatory cycles detected cervicovaginal fluid at the vulva. They self-aspirated their upper vaginal fluid, described it, and kept it for later checking. They also took basal body temperature, collected timed first morning urine samples for estrone and pregnanediol glucuronide enzyme immunoassays, and submitted to serial ovarian transvaginal ultrasound scans. +Results: Considering a G 1-day period since ultrasound ovulation detection or allowing an extra day (ÿ1 to C2), women perceived ovulation from cervicovaginal fluid at the vulva in 76\% or 97\% of cycles, on the basis of their visual description of vaginally extracted fluid in 76\% or 90\%, which rose to 90\% or 97\% for the instructor’s description, and in 76\% or 86\% with a rapid drop in glucuronide ratio. Basal body temperature was less precise (71\% or 79\%). +Conclusion: Evaluation of cervicovaginal fluid changes is an accurate ovulation indicator. Ó 2005 Elsevier Inc. All rights reserved.}, + pages = {71--75}, + number = {1}, + journaltitle = {American Journal of Obstetrics and Gynecology}, + author = {Alliende, María Elena and Cabezón, Carlos and Figueroa, Horacio and Kottmann, Cristián}, + urldate = {2025-07-01}, + date = {2005-07}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/NN29EZ6G/Alliende et al. - 2005 - Cervicovaginal fluid changes to detect ovulation accurately.pdf:application/pdf}, +} + +@online{noauthor_ringpng_nodate, + title = {ring.png (800×800)}, + url = {https://ovularing.com/wp-content/uploads/2021/07/ring.png}, + urldate = {2025-06-25}, +} + +@online{noauthor_ovularing_nodate, + title = {{OvulaRing} Startseite}, + url = {https://ovularing.com/}, + abstract = {Erfahre hier mehr zu {OvulaRing} Startseite}, + titleaddon = {{OvulaRing}}, + urldate = {2025-06-25}, + langid = {german}, + file = {Snapshot:/home/alex/Zotero/storage/PD5DBIS4/ovularing.com.html:text/html}, +} + +@article{wu_deep_nodate, + title = {Deep Transformer Models for Time Series Forecasting:The Influenza Prevalence Case}, + abstract = {In this paper, we present a new approach to time series forecasting. Time series data are prevalent in many scientific and engineering disciplines. Time series forecasting is a crucial task in modeling time series data, and is an important area of machine learning. In this work we developed a novel method that employs Transformer-based machine learning models to forecast time series data. This approach works by leveraging selfattention mechanisms to learn complex patterns and dynamics from time series data. Moreover, it is a generic framework and can be applied to univariate and multivariate time series data, as well as time series embeddings. Using influenzalike illness ({ILI}) forecasting as a case study, we show that the forecasting results produced by our approach are favorably comparable to the stateof-the-art.}, + author = {Wu, Neo and Green, Bradley and Ben, Xue and O'Banion, Shawn}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/GHT5UMNX/Wu et al. - Deep Transformer Models for Time Series ForecastingThe Influenza Prevalence Case.pdf:application/pdf}, +} + +@article{vaswani_attention_nodate, + title = {Attention Is All You Need}, + abstract = {The dominant sequence transduction models are based on complex recurrent or convolutional neural networks that include an encoder and a decoder. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 {BLEU} on the {WMT} 2014 Englishto-German translation task, improving over the existing best results, including ensembles, by over 2 {BLEU}. On the {WMT} 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art {BLEU} score of 41.8 after training for 3.5 days on eight {GPUs}, a small fraction of the training costs of the best models from the literature. We show that the Transformer generalizes well to other tasks by applying it successfully to English constituency parsing both with large and limited training data.}, + author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, Łukasz and Polosukhin, Illia}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/TES5P5PX/Vaswani et al. - Attention Is All You Need.pdf:application/pdf}, +} + +@article{twenge_declines_2017, + title = {Declines in Sexual Frequency among American Adults, 1989–2014}, + volume = {46}, + issn = {0004-0002, 1573-2800}, + url = {http://link.springer.com/10.1007/s10508-017-0953-1}, + doi = {10.1007/s10508-017-0953-1}, + pages = {2389--2401}, + number = {8}, + journaltitle = {Arch Sex Behav}, + author = {Twenge, Jean M. and Sherman, Ryne A. and Wells, Brooke E.}, + urldate = {2025-06-18}, + date = {2017-11}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/7FLDCF3U/Twenge et al. - 2017 - Declines in Sexual Frequency among American Adults, 1989–2014.pdf:application/pdf}, +} + +@artwork{chevalier_english_2018, + title = {English: Schematic of the Long-Short Term Memory cell, a component of recurrent neural networks}, + url = {https://commons.wikimedia.org/wiki/File:LSTM_Cell.svg}, + shorttitle = {English}, + author = {Chevalier, Guillaume}, + urldate = {2025-06-11}, + date = {2018-05-16}, + file = {Wikimedia Snapshot:/home/alex/Zotero/storage/NMYA4ZA3/FileLSTM_Cell.html:text/html}, +} + +@artwork{fdeloche_english_2017, + title = {English: A diagram for a one-unit recurrent neural network ({RNN}). From bottom to top : input state, hidden state, output state. U, V, W are the weights of the network. Compressed diagram on the left and the unfold version of it on the right.}, + url = {https://commons.wikimedia.org/wiki/File:Recurrent_neural_network_unfold.svg}, + shorttitle = {English}, + author = {{fdeloche}}, + urldate = {2025-06-11}, + date = {2017-06-19}, + file = {Wikimedia Snapshot:/home/alex/Zotero/storage/RU2XSRZN/FileRecurrent_neural_network_unfold.html:text/html}, +} + +@article{hochreiter_vanishing_1998, + title = {The Vanishing Gradient Problem During Learning Recurrent Neural Nets and Problem Solutions}, + volume = {06}, + issn = {0218-4885, 1793-6411}, + url = {https://www.worldscientific.com/doi/abs/10.1142/S0218488598000094}, + doi = {10.1142/S0218488598000094}, + abstract = {Recurrent nets are in principle capable to store past inputs to produce the currently desired output. Because of this property recurrent nets are used in time series prediction and process control. Practical applications involve temporal dependencies spanning many time steps, e.g. between relevant inputs and desired outputs. In this case, however, gradient based learning methods take too much time. The extremely increased learning time arises because the error vanishes as it gets propagated back. In this article the de-caying error flow is theoretically analyzed. Then methods trying to overcome vanishing gradients are briefly discussed. Finally, experiments comparing conventional algorithms and alternative methods are presented. With advanced methods long time lag problems can be solved in reasonable time.}, + pages = {107--116}, + number = {2}, + journaltitle = {Int. J. Unc. Fuzz. Knowl. Based Syst.}, + author = {Hochreiter, Sepp}, + urldate = {2025-06-11}, + date = {1998-04}, + langid = {english}, +} + +@book{medsker_recurrent_1999, + title = {Recurrent Neural Networks: Design and Applications}, + isbn = {978-1-4200-4917-6}, + shorttitle = {Recurrent Neural Networks}, + abstract = {With existent uses ranging from motion detection to music synthesis to financial forecasting, recurrent neural networks have generated widespread attention. The tremendous interest in these networks drives Recurrent Neural Networks: Design and Applications, a summary of the design, applications, current research, and challenges of this subfield of artificial neural networks.This overview incorporates every aspect of recurrent neural networks. It outlines the wide variety of complex learning techniques and associated research projects. Each chapter addresses architectures, from fully connected to partially connected, including recurrent multilayer feedforward. It presents problems involving trajectories, control systems, and robotics, as well as {RNN} use in chaotic systems. The authors also share their expert knowledge of ideas for alternate designs and advances in theoretical aspects.The dynamical behavior of recurrent neural networks is useful for solving problems in science, engineering, and business. This approach will yield huge advances in the coming years. Recurrent Neural Networks illuminates the opportunities and provides you with a broad view of the current events in this rich field.}, + pagetotal = {414}, + publisher = {{CRC} Press}, + author = {Medsker, Larry and Jain, Lakhmi C.}, + date = {1999-12-20}, + langid = {english}, + note = {Google-Books-{ID}: {ME}1SAkN0PyMC}, + keywords = {Computers / Computer Engineering, Computers / General, Computers / Software Development \& Engineering / Systems Analysis \& Design, Technology \& Engineering / Electronics / General}, } @article{hochreiter_long_1997, - title = {Long {Short}-{Term} {Memory}}, + title = {Long Short-Term Memory}, volume = {9}, issn = {0899-7667, 1530-888X}, url = {https://direct.mit.edu/neco/article/9/8/1735-1780/6109}, doi = {10.1162/neco.1997.9.8.1735}, - abstract = {Learningtostoreinformationoverextendedtimeintervalsviarecurrentbackpropagation takesaverylongtime,mostlyduetoinsu cient,decayingerrorbackow.Webrieyreview Hochreiter's1991analysisofthisproblem,thenaddressitbyintroducinganovel,e cient, gradient-basedmethodcalled{\textbackslash}LongShort-TermMemory"(LSTM).Truncatingthegradient wherethisdoesnotdoharm,LSTMcanlearntobridgeminimaltimelagsinexcessof1000 discretetimestepsbyenforcingconstanterrorowthrough{\textbackslash}constanterrorcarrousels"within specialunits.Multiplicativegateunitslearntoopenandcloseaccesstotheconstanterror ow.LSTMislocalinspaceandtime;itscomputationalcomplexitypertimestepandweight isO(1).Ourexperimentswitharticialdatainvolvelocal,distributed,real-valued,andnoisy patternrepresentations.IncomparisonswithRTRL,BPTT,RecurrentCascade-Correlation, Elmannets,andNeuralSequenceChunking,LSTMleadstomanymoresuccessfulruns,and learnsmuchfaster.LSTMalsosolvescomplex,articiallongtimelagtasksthathavenever beensolvedbypreviousrecurrentnetworkalgorithms.}, - language = {en}, - number = {8}, - urldate = {2024-10-10}, - journal = {Neural Computation}, - author = {Hochreiter, Sepp and Schmidhuber, Jürgen}, - month = nov, - year = {1997}, + abstract = {Learning to store information over extended time intervals by recurrent backpropagation takes a very long time, mostly because of insufficient, decaying error backflow. We briefly review Hochreiter's (1991) analysis of this problem, then address it by introducing a novel, efficient, gradient based method called long short-term memory ({LSTM}). Truncating the gradient where this does not do harm, {LSTM} can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units. Multiplicative gate units learn to open and close access to the constant error flow. {LSTM} is local in space and time; its computational complexity per time step and weight is O. 1. Our experiments with artificial data involve local, distributed, real-valued, and noisy pattern representations. In comparisons with real-time recurrent learning, back propagation through time, recurrent cascade correlation, Elman nets, and neural sequence chunking, {LSTM} leads to many more successful runs, and learns much faster. {LSTM} also solves complex, artificial long-time-lag tasks that have never been solved by previous recurrent network algorithms.}, pages = {1735--1780}, - file = {PDF:/home/alex/Zotero/storage/CZSV2ASE/Hochreiter and Schmidhuber - 1997 - Long Short-Term Memory.pdf:application/pdf}, + number = {8}, + journaltitle = {Neural Computation}, + author = {Hochreiter, Sepp and Schmidhuber, Jürgen}, + urldate = {2025-06-11}, + date = {1997-11-01}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/5IE5G9KY/Hochreiter and Schmidhuber - 1997 - Long Short-Term Memory.pdf:application/pdf}, } -@misc{lim_temporal_2020, - title = {Temporal {Fusion} {Transformers} for {Interpretable} {Multi}-horizon {Time} {Series} {Forecasting}}, - url = {http://arxiv.org/abs/1912.09363}, - abstract = {Multi-horizon forecasting problems often contain a complex mix of inputs -- including static (i.e. time-invariant) covariates, known future inputs, and other exogenous time series that are only observed historically -- without any prior information on how they interact with the target. While several deep learning models have been proposed for multi-step prediction, they typically comprise black-box models which do not account for the full range of inputs present in common scenarios. In this paper, we introduce the Temporal Fusion Transformer (TFT) -- a novel attention-based architecture which combines high-performance multi-horizon forecasting with interpretable insights into temporal dynamics. To learn temporal relationships at different scales, the TFT utilizes recurrent layers for local processing and interpretable self-attention layers for learning long-term dependencies. The TFT also uses specialized components for the judicious selection of relevant features and a series of gating layers to suppress unnecessary components, enabling high performance in a wide range of regimes. On a variety of real-world datasets, we demonstrate significant performance improvements over existing benchmarks, and showcase three practical interpretability use-cases of TFT.}, - urldate = {2024-10-10}, - publisher = {arXiv}, - author = {Lim, Bryan and Arik, Sercan O. and Loeff, Nicolas and Pfister, Tomas}, - month = sep, - year = {2020}, - note = {arXiv:1912.09363}, - keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, - file = {Preprint PDF:/home/alex/Zotero/storage/2R2H34KB/Lim et al. - 2020 - Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/ETDAYW36/1912.html:text/html}, +@article{salles_softed_2024, + title = {{SoftED}: Metrics for soft evaluation of time series event detection}, + volume = {198}, + issn = {03608352}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0360835224008507}, + doi = {10.1016/j.cie.2024.110728}, + shorttitle = {{SoftED}}, + abstract = {Time series event detectors are evaluated mainly by standard classification metrics, focusing solely on detection accuracy. However, inaccuracy in detecting an event can often result from its preceding or delayed effects reflected in neighboring detections. These detections are valuable to trigger necessary actions or help mitigate unwelcome consequences. In this context, current metrics are insufficient and inadequate for the context of event detection. There is a demand for metrics that incorporate both the concept of time and temporal tolerance for neighboring detections. Inspired by fuzzy sets, this paper introduces {SoftED} metrics, a new set designed for soft evaluating event detectors. They enable the evaluation of the detection accuracy and the degree to which their detections represent events. A new general protocol inspired by competency questions is also introduced to evaluate temporal tolerant metrics for event detection. The {SoftED} metrics can improve event detection evaluations by associating events and their representative detections, incorporating temporal tolerance in over 36\% of the overall detector evaluations compared to the usual classification metrics. Following the proposed evaluation protocol, {SoftED} metrics were evaluated by domain specialists who indicated their contribution to detection evaluation and method selection.}, + pages = {110728}, + journaltitle = {Computers \& Industrial Engineering}, + author = {Salles, Rebecca and Lima, Janio and Reis, Michel and Coutinho, Rafaelli and Pacitti, Esther and Masseglia, Florent and Akbarinia, Reza and Chen, Chao and Garibaldi, Jonathan and Porto, Fabio and Ogasawara, Eduardo}, + urldate = {2025-06-03}, + date = {2024-12}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/6RYA5HIQ/Salles et al. - 2024 - SoftED Metrics for soft evaluation of time series event detection.pdf:application/pdf}, } -@misc{nie_time_2023, - title = {A {Time} {Series} is {Worth} 64 {Words}: {Long}-term {Forecasting} with {Transformers}}, - shorttitle = {A {Time} {Series} is {Worth} 64 {Words}}, - url = {http://arxiv.org/abs/2211.14730}, - abstract = {We propose an efficient design of Transformer-based models for multivariate time series forecasting and self-supervised representation learning. It is based on two key components: (i) segmentation of time series into subseries-level patches which are served as input tokens to Transformer; (ii) channel-independence where each channel contains a single univariate time series that shares the same embedding and Transformer weights across all the series. Patching design naturally has three-fold benefit: local semantic information is retained in the embedding; computation and memory usage of the attention maps are quadratically reduced given the same look-back window; and the model can attend longer history. Our channel-independent patch time series Transformer (PatchTST) can improve the long-term forecasting accuracy significantly when compared with that of SOTA Transformer-based models. We also apply our model to self-supervised pre-training tasks and attain excellent fine-tuning performance, which outperforms supervised training on large datasets. Transferring of masked pre-trained representation on one dataset to others also produces SOTA forecasting accuracy. Code is available at: https://github.com/yuqinie98/PatchTST.}, - urldate = {2024-10-10}, - publisher = {arXiv}, - author = {Nie, Yuqi and Nguyen, Nam H. and Sinthong, Phanwadee and Kalagnanam, Jayant}, - month = mar, - year = {2023}, - note = {arXiv:2211.14730}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - file = {Preprint PDF:/home/alex/Zotero/storage/DG4ZJCWV/Nie et al. - 2023 - A Time Series is Worth 64 Words Long-term Forecasting with Transformers.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/H6XGVBY6/2211.html:text/html}, -} - -@misc{shao_exploring_2023, - title = {Exploring {Progress} in {Multivariate} {Time} {Series} {Forecasting}: {Comprehensive} {Benchmarking} and {Heterogeneity} {Analysis}}, - shorttitle = {Exploring {Progress} in {Multivariate} {Time} {Series} {Forecasting}}, - url = {http://arxiv.org/abs/2310.06119}, - abstract = {Multivariate Time Series (MTS) widely exists in real-word complex systems, such as traffic and energy systems, making their forecasting crucial for understanding and influencing these systems. Recently, deep learning-based approaches have gained much popularity for effectively modeling temporal and spatial dependencies in MTS, specifically in Long-term Time Series Forecasting (LTSF) and Spatial-Temporal Forecasting (STF). However, the fair benchmarking issue and the choice of technical approaches have been hotly debated in related work. Such controversies significantly hinder our understanding of progress in this field. Thus, this paper aims to address these controversies to present insights into advancements achieved. To resolve benchmarking issues, we introduce BasicTS, a benchmark designed for fair comparisons in MTS forecasting. BasicTS establishes a unified training pipeline and reasonable evaluation settings, enabling an unbiased evaluation of over 30 popular MTS forecasting models on more than 18 datasets. Furthermore, we highlight the heterogeneity among MTS datasets and classify them based on temporal and spatial characteristics. We further prove that neglecting heterogeneity is the primary reason for generating controversies in technical approaches. Moreover, based on the proposed BasicTS and rich heterogeneous MTS datasets, we conduct an exhaustive and reproducible performance and efficiency comparison of popular models, providing insights for researchers in selecting and designing MTS forecasting models.}, - urldate = {2024-10-10}, - publisher = {arXiv}, - author = {Shao, Zezhi and Wang, Fei and Xu, Yongjun and Wei, Wei and Yu, Chengqing and Zhang, Zhao and Yao, Di and Jin, Guangyin and Cao, Xin and Cong, Gao and Jensen, Christian S. and Cheng, Xueqi}, - month = oct, - year = {2023}, - note = {arXiv:2310.06119}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - file = {Preprint PDF:/home/alex/Zotero/storage/7EFZ5IT6/Shao et al. - 2023 - Exploring Progress in Multivariate Time Series Forecasting Comprehensive Benchmarking and Heterogen.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/W6RNWBLM/2310.html:text/html}, -} - -@article{zhang_crossformer_2023, - title = {{CROSSFORMER}: {TRANSFORMER} {UTILIZING} {CROSS}- {DIMENSION} {DEPENDENCY} {FOR} {MULTIVARIATE} {TIME} {SERIES} {FORECASTING}}, - abstract = {Recently many deep models have been proposed for multivariate time series (MTS) forecasting. In particular, Transformer-based models have shown great potential because they can capture long-term dependency. However, existing Transformerbased models mainly focus on modeling the temporal dependency (cross-time dependency) yet often omit the dependency among different variables (crossdimension dependency), which is critical for MTS forecasting. To fill the gap, we propose Crossformer, a Transformer-based model utilizing cross-dimension dependency for MTS forecasting. In Crossformer, the input MTS is embedded into a 2D vector array through the Dimension-Segment-Wise (DSW) embedding to preserve time and dimension information. Then the Two-Stage Attention (TSA) layer is proposed to efficiently capture the cross-time and cross-dimension dependency. Utilizing DSW embedding and TSA layer, Crossformer establishes a Hierarchical Encoder-Decoder (HED) to use the information at different scales for the final forecasting. Extensive experimental results on six real-world datasets show the effectiveness of Crossformer against previous state-of-the-arts.}, - language = {en}, - author = {Zhang, Yunhao and Yan, Junchi}, - year = {2023}, - file = {PDF:/home/alex/Zotero/storage/NM9CETJS/Zhang and Yan - 2023 - CROSSFORMER TRANSFORMER UTILIZING CROSS- DIMENSION DEPENDENCY FOR MULTIVARIATE TIME SERIES FORECAST.pdf:application/pdf}, -} - -@misc{noauthor_pytorch-forecasting_2024, - title = {Pytorch-{Forecasting}}, - url = {https://pytorch-forecasting.readthedocs.io/en/stable/}, - urldate = {2024-10-15}, - year = {2024}, -} - -@misc{liang_foundation_2024, - title = {Foundation {Models} for {Time} {Series} {Analysis}: {A} {Tutorial} and {Survey}}, - shorttitle = {Foundation {Models} for {Time} {Series} {Analysis}}, - url = {http://arxiv.org/abs/2403.14735}, - abstract = {Time series analysis stands as a focal point within the data mining community, serving as a cornerstone for extracting valuable insights crucial to a myriad of real-world applications. Recent advances in Foundation Models (FMs) have fundamentally reshaped the paradigm of model design for time series analysis, boosting various downstream tasks in practice. These innovative approaches often leverage pre-trained or fine-tuned FMs to harness generalized knowledge tailored for time series analysis. This survey aims to furnish a comprehensive and up-to-date overview of FMs for time series analysis. While prior surveys have predominantly focused on either application or pipeline aspects of FMs in time series analysis, they have often lacked an in-depth understanding of the underlying mechanisms that elucidate why and how FMs benefit time series analysis. To address this gap, our survey adopts a methodology-centric classification, delineating various pivotal elements of time-series FMs, including model architectures, pre-training techniques, adaptation methods, and data modalities. Overall, this survey serves to consolidate the latest advancements in FMs pertinent to time series analysis, accentuating their theoretical underpinnings, recent strides in development, and avenues for future exploration.}, - urldate = {2024-10-16}, - publisher = {arXiv}, - author = {Liang, Yuxuan and Wen, Haomin and Nie, Yuqi and Jiang, Yushan and Jin, Ming and Song, Dongjin and Pan, Shirui and Wen, Qingsong}, - month = jun, - year = {2024}, - note = {arXiv:2403.14735}, - keywords = {Computer Science - Machine Learning}, - file = {Preprint PDF:/home/alex/Zotero/storage/YEA28FMJ/Liang et al. - 2024 - Foundation Models for Time Series Analysis A Tutorial and Survey.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/9SGH3AIN/2403.html:text/html}, -} - -@misc{goswami_moment_2024, - title = {{MOMENT}: {A} {Family} of {Open} {Time}-series {Foundation} {Models}}, - shorttitle = {{MOMENT}}, - url = {http://arxiv.org/abs/2402.03885}, - abstract = {We introduce MOMENT, a family of open-source foundation models for general-purpose time series analysis. Pre-training large models on time series data is challenging due to (1) the absence of a large and cohesive public time series repository, and (2) diverse time series characteristics which make multi-dataset training onerous. Additionally, (3) experimental benchmarks to evaluate these models, especially in scenarios with limited resources, time, and supervision, are still in their nascent stages. To address these challenges, we compile a large and diverse collection of public time series, called the Time series Pile, and systematically tackle time series-specific challenges to unlock large-scale multi-dataset pre-training. Finally, we build on recent work to design a benchmark to evaluate time series foundation models on diverse tasks and datasets in limited supervision settings. Experiments on this benchmark demonstrate the effectiveness of our pre-trained models with minimal data and task-specific fine-tuning. Finally, we present several interesting empirical observations about large pre-trained time series models. Pre-trained models (AutonLab/MOMENT-1-large) and Time Series Pile (AutonLab/Timeseries-PILE) are available on Huggingface.}, - urldate = {2024-10-16}, - publisher = {arXiv}, - author = {Goswami, Mononito and Szafer, Konrad and Choudhry, Arjun and Cai, Yifu and Li, Shuo and Dubrawski, Artur}, - month = oct, - year = {2024}, - note = {arXiv:2402.03885}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - file = {Preprint PDF:/home/alex/Zotero/storage/QF6E6J8W/Goswami et al. - 2024 - MOMENT A Family of Open Time-series Foundation Models.pdf:application/pdf}, -} - -@misc{noauthor_decoder-only_nodate, - title = {A decoder-only foundation model for time-series forecasting}, - url = {http://research.google/blog/a-decoder-only-foundation-model-for-time-series-forecasting/}, - abstract = {Posted by Rajat Sen and Yichen Zhou, Google Research Time-series forecasting is ubiquitous in various domains, such as retail, finance, manufacturi...}, - language = {en}, - urldate = {2024-10-16}, - file = {Snapshot:/home/alex/Zotero/storage/JV9JIF73/a-decoder-only-foundation-model-for-time-series-forecasting.html:text/html}, -} - -@misc{shi_time-moe_2024, - title = {Time-{MoE}: {Billion}-{Scale} {Time} {Series} {Foundation} {Models} with {Mixture} of {Experts}}, - shorttitle = {Time-{MoE}}, - url = {http://arxiv.org/abs/2409.16040}, - abstract = {Deep learning for time series forecasting has seen significant advancements over the past decades. However, despite the success of large-scale pre-training in language and vision domains, pre-trained time series models remain limited in scale and operate at a high cost, hindering the development of larger capable forecasting models in real-world applications. In response, we introduce Time-MoE, a scalable and unified architecture designed to pre-train larger, more capable forecasting foundation models while reducing inference costs. By leveraging a sparse mixture-of-experts (MoE) design, Time-MoE enhances computational efficiency by activating only a subset of networks for each prediction, reducing computational load while maintaining high model capacity. This allows Time-MoE to scale effectively without a corresponding increase in inference costs. Time-MoE comprises a family of decoder-only transformer models that operate in an auto-regressive manner and support flexible forecasting horizons with varying input context lengths. We pre-trained these models on our newly introduced large-scale data Time-300B, which spans over 9 domains and encompassing over 300 billion time points. For the first time, we scaled a time series foundation model up to 2.4 billion parameters, achieving significantly improved forecasting precision. Our results validate the applicability of scaling laws for training tokens and model size in the context of time series forecasting. Compared to dense models with the same number of activated parameters or equivalent computation budgets, our models consistently outperform them by large margin. These advancements position Time-MoE as a state-of-the-art solution for tackling real-world time series forecasting challenges with superior capability, efficiency, and flexibility.}, - urldate = {2024-10-16}, - publisher = {arXiv}, - author = {Shi, Xiaoming and Wang, Shiyu and Nie, Yuqi and Li, Dianqi and Ye, Zhou and Wen, Qingsong and Jin, Ming}, - month = oct, - year = {2024}, - note = {arXiv:2409.16040}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - file = {Preprint PDF:/home/alex/Zotero/storage/49N63CMZ/Shi et al. - 2024 - Time-MoE Billion-Scale Time Series Foundation Models with Mixture of Experts.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/HPHN7WNJ/2409.html:text/html}, -} - -@article{coninck_dianne_2018, - title = {{DIANNE}: a modular framework for designing, training and deploying deep neural networks on heterogeneous distributed infrastructure}, - volume = {141}, - issn = {01641212}, - shorttitle = {{DIANNE}}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0164121218300487}, - doi = {10.1016/j.jss.2018.03.032}, - language = {en}, - urldate = {2024-10-23}, - journal = {Journal of Systems and Software}, - author = {Coninck, Elias De and Bohez, Steven and Leroux, Sam and Verbelen, Tim and Vankeirsbilck, Bert and Simoens, Pieter and Dhoedt, Bart}, - month = jul, - year = {2018}, - pages = {52--65}, - file = {Full Text:/home/alex/Zotero/storage/SFH656QN/Coninck et al. - 2018 - DIANNE a modular framework for designing, training and deploying deep neural networks on heterogene.pdf:application/pdf}, -} - -@misc{noauthor_keras_nodate, - title = {Keras: {Deep} {Learning} for humans}, - url = {https://keras.io/}, - urldate = {2024-10-23}, - file = {Keras\: Deep Learning for humans:/home/alex/Zotero/storage/MS4QLPRC/keras.io.html:text/html}, -} - -@article{masuda_machine_2025, - title = {Machine learning model for menstrual cycle phase classification and ovulation day detection based on sleeping heart rate under free-living conditions}, - volume = {187}, - issn = {0010-4825}, - url = {https://www.sciencedirect.com/science/article/pii/S0010482525000551}, - doi = {10.1016/j.compbiomed.2025.109705}, - abstract = {The accurate classification of menstrual cycle phases and detection of ovulation is critical for women's health management, particularly in addressing infertility, alleviating premenstrual syndrome, and preventing hormone-related disorders. However, traditional basal body temperature (BBT) measurement methods are susceptible to disruptions in sleep timing and environmental conditions, limiting practical application. This study is aimed to overcome these limitations by introducing a novel feature, heart rate at the circadian rhythm nadir (minHR), for classifying menstrual cycle phases and predicting ovulation. A machine learning model was developed using XGBoost, and data were collected under free-living conditions from 40 healthy women (18–34 years) over a maximum of three menstrual cycles. Three feature combinations— “day,” “day + minHR,” and “day + BBT”—were evaluated, and model performance was assessed using nested leave-one-group-out cross-validation. The feature “day” represents the number of days elapsed since the onset of menstruation. Participants were stratified into groups depending on high variability and low variability in sleep timing. Results demonstrated that adding minHR significantly improved luteal phase classification and ovulation day detection performance compared to “day” only. Furthermore, in participants with high variability in sleep timing, the minHR-based model outperformed the BBT-based model, significantly improving luteal phase recall and reducing ovulation day detection absolute errors by 2 d (p {\textless} 0.05). These findings highlight the robustness and practicality of the minHR-based model for menstrual cycle tracking, particularly in individuals with high variability in sleep timing. The proposed model holds great promise for personalized health management and large-scale epidemiological research.}, - urldate = {2025-02-11}, - journal = {Computers in Biology and Medicine}, - author = {Masuda, Hazuki and Okada, Shima and Shiozawa, Naruhiro and Sakaue, Yusuke and Manno, Masanobu and Makikawa, Masaaki and Isaka, Tadao}, - month = mar, - year = {2025}, - keywords = {Circadian rhythm, Heart rate, Machine learning, Menstrual cycle tracking, Ovulation day detection, Wearable sensor, XGBoost}, - pages = {109705}, - file = {PDF:/home/alex/Zotero/storage/87H6TB3Q/Masuda et al. - 2025 - Machine learning model for menstrual cycle phase classification and ovulation day detection based on.pdf:application/pdf;ScienceDirect Snapshot:/home/alex/Zotero/storage/PC6FSQIA/S0010482525000551.html:text/html}, -} - -@article{maman_prediction_2023, - title = {Prediction of ovulation: new insight into an old challenge}, - volume = {13}, - issn = {2045-2322}, - shorttitle = {Prediction of ovulation}, - url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10651856/}, - doi = {10.1038/s41598-023-47241-2}, - abstract = {Ultrasound monitoring and hormonal blood testing are considered by many as an accurate method to predict ovulation time. However, uniform and validated algorithms for predicting ovulation have yet to be defined. Daily hormonal tests and transvaginal ultrasounds were recorded to develop an algorithm for ovulation prediction. The rupture of the leading ovarian follicle was a marker for ovulation day. The model was validated retrospectively on natural cycles frozen embryo transfer cycles with documented ovulation. Circulating levels of LH or its relative variation failed, by themselves, to reliably predict ovulation. Any decrease in estrogen was 100\% associated with ovulation emergence the same day or the next day. Progesterone levels {\textgreater} 2 nmol/L had low specificity to predict ovulation the next day (62.7\%), yet its sensitivity was high (91.5\%). A model for ovulation prediction, combining the three hormone levels and ultrasound was created with an accuracy of 95\% to 100\% depending on the combination of the hormone levels. Model validation showed correct ovulation prediction in 97\% of these cycles. We present an accurate ovulation prediction algorithm. The algorithm is simple and user-friendly so both reproductive endocrinologists and general practitioners can use it to benefit their patients.}, - urldate = {2025-02-11}, - journal = {Scientific Reports}, - author = {Maman, Ettie and Adashi, Eli Y. and Baum, Micha and Hourvitz, Ariel}, - month = nov, - year = {2023}, - pmid = {37968377}, - pmcid = {PMC10651856}, - pages = {20003}, - annote = {also not “really” relevant, as domain is ultrasound and hormone levels - -}, - file = {PubMed Central Full Text PDF:/home/alex/Zotero/storage/MTHDPZ5B/Maman et al. - 2023 - Prediction of ovulation new insight into an old challenge.pdf:application/pdf}, -} - -@article{luo_prediction_2025, - title = {Prediction of the fertile window and menstrual cycles with a wearable device via machine-learning algorithms}, - issn = {1472-6483}, - url = {https://www.sciencedirect.com/science/article/pii/S1472648325000021}, - doi = {10.1016/j.rbmo.2025.104795}, - abstract = {Research question -We aimed to develop fertile window and menstruation prediction algorithms through machine learning based on women's physiological parameters data collected by Huawei Band 6 pro from both regular and irregular menstruators. -Design -This was a prospective observational cohort study conducted at Obstetrics and Gynecology Hospital of Fudan University. Participants were recruited from November 2021 to September 2022. Each participant wore Huawei Band 6 pro to record wrist skin temperature (WST), heart rate (HR), heart rate variability, and respiratory rate. Algorithms were developed to predict the fertile window and menstrual cycle based on WST and HR. -Results -We included data from 270 and 84 qualified cycles with confirmed ovulations from 136 regular and 47 irregular menstruators. For regular menstruators, the prediction algorithm based on WST and HR for the fertile window had an accuracy of 85.47\%, a sensitivity of 70.07\%, a specificity of 89.77\%, and AUC of 0.869. The algorithms for menstrual first day labelling and onset within 3 days gained an accuracy of 83.6\% and 75.0\%. For irregular menstruators, the accuracy, sensitivity, specificity and AUC were 79.85\%, 42.79\%, 87.28\%, and 0.763 respectively, for fertile window prediction. The accuracy of menses labelling and prediction were 61.2\%, and 50.8\% respectively. -Conclusions -Based on WST and HR data from the wearable device, the algorithms demonstrated reliable performance in predicting the fertile window and menstruation day among regular menstruators. These algorithms also showed potential for assisting irregular menstruators in managing their cycles and planning for conception.}, - urldate = {2025-02-11}, - journal = {Reproductive BioMedicine Online}, - author = {Luo, Chuan and Su, Yun-Fei and Ren, Yun-Yun and Zhang, Qin and Li, Ran and Zhang, Qi and Li, Cheng and Hao, Yan-Hui and Zhang, An-Qi and Zhang, Hao and Huang, He-Feng and Wu, Yan-Ting}, - month = jan, - year = {2025}, - keywords = {Machine learning, Fertile window, Menstrual cycle, Natural cycle, Non-invasive wearable device, Wrist skin temperature}, - pages = {104795}, - annote = { - -need closer look, how do they determine the fertile phase? what do they test against? what model do they use? - - -}, - file = {PDF:/home/alex/Zotero/storage/YIV6T8MS/Luo et al. - 2025 - Prediction of the fertile window and menstrual cycles with a wearable device via machine-learning al.pdf:application/pdf;ScienceDirect Snapshot:/home/alex/Zotero/storage/3IJMUH82/S1472648325000021.html:text/html}, -} - -@article{yu_tracking_2022, - title = {Tracking of menstrual cycles and prediction of the fertile window via measurements of basal body temperature and heart rate as well as machine-learning algorithms}, - volume = {20}, - issn = {1477-7827}, - url = {https://doi.org/10.1186/s12958-022-00993-4}, - doi = {10.1186/s12958-022-00993-4}, - abstract = {Fertility awareness and menses prediction are important for improving fecundability and health management. Previous studies have used physiological parameters, such as basal body temperature (BBT) and heart rate (HR), to predict the fertile window and menses. However, their accuracy is far from satisfactory. Additionally, few researchers have examined irregular menstruators. Thus, we aimed to develop fertile window and menstruation prediction algorithms for both regular and irregular menstruators.}, - number = {1}, - urldate = {2025-02-11}, - journal = {Reproductive Biology and Endocrinology}, - author = {Yu, Jia-Le and Su, Yun-Fei and Zhang, Chen and Jin, Li and Lin, Xian-Hua and Chen, Lu-Ting and Huang, He-Feng and Wu, Yan-Ting}, - month = aug, - year = {2022}, - keywords = {Heart rate, Machine learning, Fertile window, Menstrual cycle, Basal body temperature, Wearable device}, - pages = {118}, - annote = { - -potentially interesting - - -also use temperature plus hormone levels and ultrasounds - - -}, - file = {Full Text PDF:/home/alex/Zotero/storage/7Z8P97UF/Yu et al. - 2022 - Tracking of menstrual cycles and prediction of the fertile window via measurements of basal body tem.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/SMRBT6EG/s12958-022-00993-4.html:text/html}, -} - -@article{luz_improved_2024, - title = {Improved clinical pregnancy rates in natural frozen-thawed embryo transfer cycles with machine learning ovulation prediction: insights from a retrospective cohort study}, - volume = {14}, - copyright = {2024 The Author(s)}, - issn = {2045-2322}, - shorttitle = {Improved clinical pregnancy rates in natural frozen-thawed embryo transfer cycles with machine learning ovulation prediction}, - url = {https://www.nature.com/articles/s41598-024-80356-8}, - doi = {10.1038/s41598-024-80356-8}, - abstract = {This study aims to develop physician support software for determining ovulation time and assess its impact on pregnancy outcomes in natural cycle frozen embryo transfers (NC-FET). To develop, assess, and validate an ovulation prediction model, three datasets were used: REI Ovulation Determination dataset (500 cycles) split into training (309), validation (90), and test (101) sets; the Documented Ovulation dataset (101 cycles) with confirmed ovulation (documented follicular rupture and LH surge); and the Clinical Pregnancy Rates dataset (515 NC-FET cycles), categorized into “Matched” and “Mismatched” based on alignment with the model’s ovulation determination. Pregnancy outcomes were compared between the groups. The ovulation prediction model exhibited 93.85\% and 92.89\% matching rates with the REI Ovulation Determination and Documented Ovulation datasets, respectively. In the Clinical Pregnancy Rates dataset, the Matched group (282 cycles) showed significantly higher clinical pregnancy rates than the Mismatched group (34.6\% vs. 25.9\%, p = 0.04) and similar results for patients under 37 (41.1\% vs. 30.7\%, p = 0.04). Logistic regression indicated lower pregnancy rates in Mismatched cases (odds ratio 0.67 for the general population, 0.63 for patients under 37). In conclusion, we introduce a highly accurate AI ovulation prediction model. Treatment cycles aligning with the model’s recommendations had significantly increased clinical pregnancy rates.}, - language = {en}, - number = {1}, - urldate = {2025-02-11}, - journal = {Scientific Reports}, - author = {Luz, Almog and Hourvitz, Ariel and Moran, Eden and Itzhak, Nevo and Reuvenny, Shachar and Hourvitz, Rohi and Youngster, Michal and Baum, Micha and Maman, Ettie}, - month = nov, - year = {2024}, - note = {Publisher: Nature Publishing Group}, - keywords = {Infertility, Outcomes research}, - pages = {29451}, - file = {Full Text PDF:/home/alex/Zotero/storage/BSHNTIFD/Luz et al. - 2024 - Improved clinical pregnancy rates in natural frozen-thawed embryo transfer cycles with machine learn.pdf:application/pdf}, -} - -@article{pratikno_pdf_2024, - title = {({PDF}) {A} novel women's ovulation prediction through salivary ferning using the box counting and deep learning}, - url = {https://www.researchgate.net/publication/379467622_A_novel_women's_ovulation_prediction_through_salivary_ferning_using_the_box_counting_and_deep_learning}, - doi = {10.11591/eei.v13i2.5847}, - abstract = {PDF {\textbar} There are several methods to predict a woman's ovulation time, including using a calendar system, basal body temperature, ovulation prediction... {\textbar} Find, read and cite all the research you need on ResearchGate}, - language = {en}, - urldate = {2025-02-11}, - journal = {ResearchGate}, - author = {Pratikno and Ibrahim and Jusak}, - month = dec, - year = {2024}, - annote = {Not really relevant, as different data domain -{\textgreater} saliva -}, - file = {Full Text:/home/alex/Zotero/storage/MVSBPZQG/2024 - (PDF) A novel women's ovulation prediction through salivary ferning using the box counting and deep.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/8FYLC8YZ/379467622_A_novel_women's_ovulation_prediction_through_salivary_ferning_using_the_box_counting_.html:text/html}, -} - -@article{shkodzik_innovative_2024, - title = {Innovative {Approaches} to {Digital} {Health} in {Ovulation} {Detection}: {A} {Review} of {Current} {Methods} and {Emerging} {Technologies}}, - volume = {42}, - issn = {1526-4564}, - shorttitle = {Innovative {Approaches} to {Digital} {Health} in {Ovulation} {Detection}}, - doi = {10.1055/s-0044-1793829}, - abstract = {Ovulation is a vital sign, as significant as body temperature, heart rate, respiratory rate, and blood pressure, in assessing overall health and identifying potential health issues. Ovulation is a key event of the menstrual cycle that provides insights into the hormonal and reproductive health aspects. Affected by the orchestra of hormones, namely thyroid, prolactin, and androgens, disruptions in ovulation can indicate endocrinological conditions and lead to gynecological problems, such as heavy menstrual bleeding, irregular periods, amenorrhea, dysmenorrhea, and difficulties in getting pregnant. Monitoring ovulation and detecting disruptions can aid in the early detection of health issues, extending beyond reproductive health concerns. It can help identify underlying causes of symptoms like excessive fatigue and abnormal hair growth. The integration of digital health technologies, such as mobile apps using machine learning algorithms, wearables tracking temperature, heart rate, breath rate, and sleep patterns, and devices measuring reproductive hormones in urine or saliva samples, offers a wealth of opportunities in family planning, early health issue diagnosis, treatment adjustment, and tracking menstrual cycles during assisted reproductive techniques. These advancements provide a comprehensive approach to health monitoring, addressing both reproductive and overall health concerns.}, - language = {eng}, - number = {2}, - journal = {Seminars in Reproductive Medicine}, - author = {Shkodzik, Katerina}, - month = jun, - year = {2024}, - pmid = {39572028}, - keywords = {Humans, Digital Health, Female, Mobile Applications, Ovulation, Ovulation Detection, Telemedicine, Wearable Electronic Devices}, - pages = {81--89}, -} - -@article{lin_transformer_2023, - title = {Transformer neural network to predict and interpret pregnancy loss from activity data in {Holstein} dairy cows}, - volume = {205}, - issn = {0168-1699}, - url = {https://www.sciencedirect.com/science/article/pii/S0168169923000261}, - doi = {10.1016/j.compag.2023.107638}, - abstract = {Predicting/detecting pregnancy loss of dairy cows offers the opportunity to shorten the time interval between artificial inseminations. Although several methods of pregnancy detection are being practiced, models with accurate, timely and interpretable detection of pregnancy are still lacking. This study proposed a transformer neural network to predict the probability of pregnancy loss based on continuous activity data, which were collected from activity-monitoring tags attached to 185 Holstein cows from a commercial dairy farm in Cayuga County, NY, USA. Our best model achieved an average accuracy of 0.87, F1 score of 0.87, recall of 0.87 and specificity of 0.90 using 14-day time-series activity windows (90\% overlap) using 5-fold cross-validation, outperforming commonly used classic statistical learning and deep learning models for time-series data. The results indicated that our predictive model gave high probabilities of correctly detecting pregnancy loss prior to the increased activities and veterinary confirmation by transrectal ultrasound. In addition, our model interpretation aligned with the changes in the temporal activity levels, revealing that drastic fluctuations in time-series activity data contributed heavily to the final prediction. To the best of our knowledge, this is the first work on developing transformer models for the prediction of pregnancy loss in dairy cows. In addition to facilitating the development of future precision management on modern farms, our work potentiates an increase in the reproductive efficiency and profitability of dairy farms.}, - urldate = {2025-02-11}, - journal = {Computers and Electronics in Agriculture}, - author = {Lin, Dan and Kenéz, Ákos and McArt, Jessica A. A. and Li, Jun}, - month = feb, - year = {2023}, - keywords = {Dairy cow, Precision livestock farming, Pregnancy loss prediction, Time-series activity}, - pages = {107638}, -} - -@article{luz_p-656_2023, - title = {P-656 {Machine} learning algorithm automatically manages and accurately predicts ovulation in natural frozen-thawed embryo transfer cycles.}, - volume = {38}, - issn = {0268-1161}, - url = {https://doi.org/10.1093/humrep/dead093.983}, - doi = {10.1093/humrep/dead093.983}, - abstract = {Can an Artificial Intelligence (AI) algorithm automatically manage frozen-thawed embryo transfer (NC-FET) treatment cycles and give an accurate prediction of ovulation day.An AI algorithm automatically managed and predicted the ovulation of NC-FET treatment cycles with 94.8\% accuracy using an average of 3.01 test days.Today the preferred method for frozen embryo transfer is natural cycle based on ovulation detection. Currently, there is no software capable of managing the treatment cycle automatically and identifying the time of ovulation to support doctor decisions. The aim of this study is to develop a physician support AI software for determining ovulation time reliably with high accuracy.2083 NC-FET cycles from September 2018 to June 2021 were used to develop the ovulation detection and treatment management algorithms.Each cycle had data from at least 2 visits including: hormonal levels (Estrogen/Progesterone/LH) and follicle sizes.The dataset was divided into a train set and two test sets. In the 1st test set ovulation was determined by experts’ opinions and the 2nd test set included cycles in which follicle rupture was documented in consecutive ultrasounds.Two algorithms were developed, an ovulation prediction algorithm based on an NGBoost model and a treatment management algorithm that used the model to determine if and when to call for a new test or declare the ovulation day.Both algorithms were jointly tuned to reach the highest success rate, defined as providing the correct day of ovulation using the available cycle data, with as few test days as possible.On the first test set, which consisted of 176 cycles in which ovulation was determined through the majority decision of 2 independent experts and the attending physician, the treatment management algorithm required on average 3.01 tests to reach a prediction and successfully predicted the ovulation day in 94.8\% of cycles.In the second test set, which consisted of 29 cycles in which ovulation was determined through the follicular rupture in two consecutive ultrasounds, only the ovulation prediction model was tested. To ensure that the model provides a reliable answer and does not rely solely on the follicular disappearances, examined cycles were tested twice: Once using the ovulation day without the day prior to it, and again using only the day prior to ovulation without the ovulation day itself. The algorithm accurately predicted ovulation in 28 out of 29 instances (96.6\%) using the day of ovulation and in 28 out of 29 instances (96.6\%) using the day before ovulation.The main drawback is this being a retrospective study: while the algorithm was trained to maximize accuracy when it selects the test days, the dataset test days were selected by the attending physicians. Statistical methods were used to overcome this, however further prospective trials are needed to validate the results.This is the first AI algorithm designed to automatically manage NC-FET IVF treatment cycles and predict ovulation. The high accuracy and low average tests count might improve treatment outcomes, reduce the patients’ life disruption, and allow physicians to spend less time monitoring their patients’ treatments.not applicable}, - number = {Supplement\_1}, - urldate = {2025-02-11}, - journal = {Human Reproduction}, - author = {Luz, A and Hourvitz, R and Reuvenny, S and Youngster, M and Baum, M and Hourvitz, A and Maman, E}, - month = jun, - year = {2023}, - pages = {dead093.983}, - file = {Full Text PDF:/home/alex/Zotero/storage/C8YM765Y/Luz et al. - 2023 - P-656 Machine learning algorithm automatically manages and accurately predicts ovulation in natural.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/WANIZX3H/7202977.html:text/html}, -} - -@inproceedings{azaria_semi-supervised_2019, - title = {Semi-{Supervised} {Ovulation} {Detection} {Based} on {Multiple} {Properties}}, - url = {https://ieeexplore.ieee.org/document/8995235}, - doi = {10.1109/ICTAI.2019.00039}, - abstract = {Despite being a well-researched problem, ovulation detection in human female remains a difficult task. Most current methods for ovulation detection rely on measurements of a single property (e.g. morning body temperature) or at most on two properties (e.g. both salivary and vaginal electrical resistance). In this paper we present a machine learning based method for detecting the day in which ovulation occurs. Our method considered measurements of five different properties. We crawled a data-set from the web and showed that our method outperforms current state-of-the-art methods for ovulation detection. Our method performs well also when considering measurements of fewer properties. We show that our method's performance can be further improved by using unlabeled data, that is, mensuration cycles without a know ovulation date. Our resulted machine learning model can be very useful for women trying to conceive that have trouble in recognizing their ovulation period, especially when some measurements are missing.}, - urldate = {2025-02-11}, - booktitle = {2019 {IEEE} 31st {International} {Conference} on {Tools} with {Artificial} {Intelligence} ({ICTAI})}, - author = {Azaria, Amos and Azaria, Seagal}, - month = nov, - year = {2019}, - note = {ISSN: 2375-0197}, - keywords = {ovulation detection, semi supervised learning}, - pages = {222--228}, - annote = { - -datapoints - - -basal body temperature - - -salivary electrical resistance - - -vaginal electric resistance - - -ovulation prediction kit -{\textgreater} measures for LH hormone - - -markers such as breast tenderness, cervical mucus and custom - - - - -models - - -cnn - - -LSTM - - -conditional random fields approach? - - -“semi supervised” idea is nuts, using the model to create labels that feed back into the training process (?) - - - - -Use leave one out cross validation, -{\textgreater} questionable results - - -}, - file = {IEEE Xplore Abstract Record:/home/alex/Zotero/storage/JSAEWRGB/8995235.html:text/html;PDF:/home/alex/Zotero/storage/P74SEG9S/Azaria and Azaria - 2019 - Semi-Supervised Ovulation Detection Based on Multiple Properties.pdf:application/pdf}, -} - -@article{fanton_interpretable_2022, - title = {An interpretable machine learning model for predicting the optimal day of trigger during ovarian stimulation}, - volume = {118}, - issn = {0015-0282, 1556-5653}, - url = {https://www.fertstert.org/article/S0015-0282%2822%2900244-8/fulltext}, - doi = {10.1016/j.fertnstert.2022.04.003}, - language = {English}, - number = {1}, - urldate = {2025-02-11}, - journal = {Fertility and Sterility}, - author = {Fanton, Michael and Nutting, Veronica and Solano, Funmi and Maeder-York, Paxton and Hariton, Eduardo and Barash, Oleksii and Weckstein, Louis and Sakkas, Denny and Copperman, Alan B. and Loewke, Kevin}, - month = jul, - year = {2022}, - note = {Publisher: Elsevier}, - keywords = {Artificial intelligence, in vitro fertilization, machine learning, ovarian stimulation, trigger}, - pages = {101--108}, - file = {Full Text PDF:/home/alex/Zotero/storage/UNX7ASLP/Fanton et al. - 2022 - An interpretable machine learning model for predicting the optimal day of trigger during ovarian sti.pdf:application/pdf}, -} - -@article{braude_machine_2024, - title = {Machine learning for predicting elective fertility preservation outcomes}, - volume = {14}, - copyright = {2024 The Author(s)}, - issn = {2045-2322}, - url = {https://www.nature.com/articles/s41598-024-60671-w}, - doi = {10.1038/s41598-024-60671-w}, - abstract = {This retrospective study applied machine-learning models to predict treatment outcomes of women undergoing elective fertility preservation. Two-hundred-fifty women who underwent elective fertility preservation at a tertiary center, 2019–2022 were included. Primary outcome was the number of metaphase II oocytes retrieved. Outcome class was based on oocyte count (OC): Low (≤ 8), Medium (9–15) or High (≥ 16). Machine-learning models and statistical regression were used to predict outcome class, first based on pre-treatment parameters, and then using post-treatment data from ovulation-triggering day. OC was 136 Low, 80 Medium, and 34 High. Random Forest Classifier (RFC) was the most accurate model (pre-treatment receiver operating characteristic (ROC) area under the curve (AUC) was 77\%, and post-treatment ROC AUC was 87\%), followed by XGBoost Classifier (pre-treatment ROC AUC 74\%, post-treatment ROC AUC 86\%). The most important pre-treatment parameters for RFC were basal FSH (22.6\%), basal LH (19.1\%), AFC (18.2\%), and basal estradiol (15.6\%). Post-treatment parameters were estradiol levels on trigger-day (17.7\%), basal FSH (11\%), basal LH (9\%), and AFC (8\%). Machine-learning models trained with clinical data appear to predict fertility preservation treatment outcomes with relatively high accuracy.}, - language = {en}, - number = {1}, - urldate = {2025-02-11}, - journal = {Scientific Reports}, - author = {Braude, Itai and Haikin Herzberger, Einat and Semo, Mor and Soifer, Kim and Goren Gepstein, Nitzan and Wiser, Amir and Miller, Netanella}, - month = may, - year = {2024}, - note = {Publisher: Nature Publishing Group}, - keywords = {Outcomes research, Computational models}, - pages = {10158}, - file = {Full Text PDF:/home/alex/Zotero/storage/URDGBHLV/Braude et al. - 2024 - Machine learning for predicting elective fertility preservation outcomes.pdf:application/pdf}, -} - -@article{noauthor_temporal_2021, - title = {Temporal {Fusion} {Transformers} for interpretable multi-horizon time series forecasting}, - volume = {37}, - issn = {0169-2070}, - url = {https://www.sciencedirect.com/science/article/pii/S0169207021000637}, - doi = {10.1016/j.ijforecast.2021.03.012}, - abstract = {Multi-horizon forecasting often contains a complex mix of inputs – including static (i.e. time-invariant) covariates, known future inputs, and other e…}, - language = {en-US}, - number = {4}, - urldate = {2025-02-24}, - journal = {International Journal of Forecasting}, - month = oct, - year = {2021}, - note = {Publisher: Elsevier}, - pages = {1748--1764}, - file = {Snapshot:/home/alex/Zotero/storage/SFYESIWK/S0169207021000637.html:text/html;Submitted Version:/home/alex/Zotero/storage/A9AYS5UI/2021 - Temporal Fusion Transformers for interpretable multi-horizon time series forecasting.pdf:application/pdf}, -} - -@inproceedings{vaswani_attention_2017, - title = {Attention is {All} you {Need}}, - volume = {30}, - url = {https://proceedings.neurips.cc/paper/2017/hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html}, - abstract = {The dominant sequence transduction models are based on complex recurrent orconvolutional neural networks in an encoder and decoder configuration. The best performing such models also connect the encoder and decoder through an attentionm echanisms. We propose a novel, simple network architecture based solely onan attention mechanism, dispensing with recurrence and convolutions entirely.Experiments on two machine translation tasks show these models to be superiorin quality while being more parallelizable and requiring significantly less timeto train. Our single model with 165 million parameters, achieves 27.5 BLEU onEnglish-to-German translation, improving over the existing best ensemble result by over 1 BLEU. On English-to-French translation, we outperform the previoussingle state-of-the-art with model by 0.7 BLEU, achieving a BLEU score of 41.1.}, - urldate = {2025-02-24}, - booktitle = {Advances in {Neural} {Information} {Processing} {Systems}}, - publisher = {Curran Associates, Inc.}, - author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, Ł ukasz and Polosukhin, Illia}, - year = {2017}, - file = {Full Text PDF:/home/alex/Zotero/storage/MU7NU9LR/Vaswani et al. - 2017 - Attention is All you Need.pdf:application/pdf}, -} - -@misc{noauthor_vivosens_nodate, - title = {vivosens medical gmbh}, - url = {https://www.vivosensmedical.com/}, - urldate = {2025-02-24}, - file = {vivosensmedical.com:/home/alex/Zotero/storage/KSM7HHBJ/www.vivosensmedical.com.html:text/html}, -} - -@inproceedings{rigotti_attention-based_2021, - title = {Attention-based {Interpretability} with {Concept} {Transformers}}, - url = {https://openreview.net/forum?id=kAa9eDS0RdO}, - abstract = {Attention is a mechanism that has been instrumental in driving remarkable performance gains of deep neural network models in a host of visual, NLP and multimodal tasks. One additional notable aspect of attention is that it conveniently exposes the ``reasoning'' behind each particular output generated by the model. Specifically, attention scores over input regions or intermediate features have been interpreted as a measure of the contribution of the attended element to the model inference. While the debate in regard to the interpretability of attention is still not settled, researchers have pointed out the existence of architectures and scenarios that afford a meaningful interpretation of the attention mechanism. Here we propose the generalization of attention from low-level input features to high-level concepts as a mechanism to ensure the interpretability of attention scores within a given application domain. In particular, we design the ConceptTransformer, a deep learning module that exposes explanations of the output of a model in which it is embedded in terms of attention over user-defined high-level concepts. Such explanations are {\textbackslash}emph\{plausible\} (i.e.{\textbackslash} convincing to the human user) and {\textbackslash}emph\{faithful\} (i.e.{\textbackslash} truly reflective of the reasoning process of the model). Plausibility of such explanations is obtained by construction by training the attention heads to conform with known relations between inputs, concepts and outputs dictated by domain knowledge. Faithfulness is achieved by design by enforcing a linear relation between the transformer value vectors that represent the concepts and their contribution to the classification log-probabilities. We validate our ConceptTransformer module on established explainability benchmarks and show how it can be used to infuse domain knowledge into classifiers to improve accuracy, and conversely to extract concept-based explanations of classification outputs. Code to reproduce our results is available at: {\textbackslash}url\{https://github.com/ibm/concept\_transformer\}.}, - language = {en}, - urldate = {2025-02-21}, - author = {Rigotti, Mattia and Miksovic, Christoph and Giurgiu, Ioana and Gschwind, Thomas and Scotton, Paolo}, - month = oct, - year = {2021}, - file = {Full Text PDF:/home/alex/Zotero/storage/U2FUGVF6/Rigotti et al. - 2021 - Attention-based Interpretability with Concept Transformers.pdf:application/pdf}, -} - -@article{kitada_attention_2021, - title = {Attention {Meets} {Perturbations}: {Robust} and {Interpretable} {Attention} {With} {Adversarial} {Training}}, - volume = {9}, - issn = {2169-3536}, - shorttitle = {Attention {Meets} {Perturbations}}, - url = {https://ieeexplore.ieee.org/abstract/document/9467291}, - doi = {10.1109/ACCESS.2021.3093456}, - abstract = {Although attention mechanisms have been applied to a variety of deep learning models and have been shown to improve the prediction performance, it has been reported to be vulnerable to perturbations to the mechanism. To overcome the vulnerability to perturbations in the mechanism, we are inspired by adversarial training (AT), which is a powerful regularization technique for enhancing the robustness of the models. In this paper, we propose a general training technique for natural language processing tasks, including AT for attention (Attention AT) and more interpretable AT for attention (Attention iAT). The proposed techniques improved the prediction performance and the model interpretability by exploiting the mechanisms with AT. In particular, Attention iAT boosts those advantages by introducing adversarial perturbation, which enhances the difference in the attention of the sentences. Evaluation experiments with ten open datasets revealed that AT for attention mechanisms, especially Attention iAT, demonstrated (1) the best performance in nine out of ten tasks and (2) more interpretable attention (i.e., the resulting attention correlated more strongly with gradient-based word importance) for all tasks. Additionally, the proposed techniques are (3) much less dependent on perturbation size in AT.}, - urldate = {2025-02-21}, - journal = {IEEE Access}, - author = {Kitada, Shunsuke and Iyatomi, Hitoshi}, - year = {2021}, - note = {Conference Name: IEEE Access}, - keywords = {adversarial training, attention mechanism, binary classification, interpretability, Knowledge discovery, natural language inference, Natural language processing, Perturbation methods, Predictive models, question answering, Robustness, Solid modeling, Task analysis, Training}, - pages = {92974--92985}, - file = {Full Text PDF:/home/alex/Zotero/storage/BUXFCXRU/Kitada and Iyatomi - 2021 - Attention Meets Perturbations Robust and Interpretable Attention With Adversarial Training.pdf:application/pdf;IEEE Xplore Abstract Record:/home/alex/Zotero/storage/ZNQD3FRW/9467291.html:text/html}, -} - -@inproceedings{choi_retain_2016, - title = {{RETAIN}: {An} {Interpretable} {Predictive} {Model} for {Healthcare} using {Reverse} {Time} {Attention} {Mechanism}}, - volume = {29}, - shorttitle = {{RETAIN}}, - url = {https://proceedings.neurips.cc/paper/2016/hash/231141b34c82aa95e48810a9d1b33a79-Abstract.html}, - abstract = {Accuracy and interpretability are two dominant features of successful predictive models. Typically, a choice must be made in favor of complex black box models such as recurrent neural networks (RNN) for accuracy versus less accurate but more interpretable traditional models such as logistic regression. This tradeoff poses challenges in medicine where both accuracy and interpretability are important. We addressed this challenge by developing the REverse Time AttentIoN model (RETAIN) for application to Electronic Health Records (EHR) data. RETAIN achieves high accuracy while remaining clinically interpretable and is based on a two-level neural attention model that detects influential past visits and significant clinical variables within those visits (e.g. key diagnoses). RETAIN mimics physician practice by attending the EHR data in a reverse time order so that recent clinical visits are likely to receive higher attention. RETAIN was tested on a large health system EHR dataset with 14 million visits completed by 263K patients over an 8 year period and demonstrated predictive accuracy and computational scalability comparable to state-of-the-art methods such as RNN, and ease of interpretability comparable to traditional models.}, - urldate = {2025-02-21}, - booktitle = {Advances in {Neural} {Information} {Processing} {Systems}}, - publisher = {Curran Associates, Inc.}, - author = {Choi, Edward and Bahadori, Mohammad Taha and Sun, Jimeng and Kulas, Joshua and Schuetz, Andy and Stewart, Walter}, - year = {2016}, - file = {Full Text PDF:/home/alex/Zotero/storage/XQLMYHUU/Choi et al. - 2016 - RETAIN An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism.pdf:application/pdf}, -} - -@misc{serrano_is_2019, - title = {Is {Attention} {Interpretable}?}, - url = {http://arxiv.org/abs/1906.03731}, - doi = {10.48550/arXiv.1906.03731}, - abstract = {Attention mechanisms have recently boosted performance on a range of NLP tasks. Because attention layers explicitly weight input components' representations, it is also often assumed that attention can be used to identify information that models found important (e.g., specific contextualized word tokens). We test whether that assumption holds by manipulating attention weights in already-trained text classification models and analyzing the resulting differences in their predictions. While we observe some ways in which higher attention weights correlate with greater impact on model predictions, we also find many ways in which this does not hold, i.e., where gradient-based rankings of attention weights better predict their effects than their magnitudes. We conclude that while attention noisily predicts input components' overall importance to a model, it is by no means a fail-safe indicator.}, - urldate = {2025-02-21}, - publisher = {arXiv}, - author = {Serrano, Sofia and Smith, Noah A.}, - month = jun, - year = {2019}, - note = {arXiv:1906.03731 [cs]}, - keywords = {Computer Science - Computation and Language}, - annote = {Comment: To appear at ACL 2019}, - file = {Preprint PDF:/home/alex/Zotero/storage/DFZ28RG8/Serrano and Smith - 2019 - Is Attention Interpretable.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/63B9XS2Y/1906.html:text/html}, -} - -@article{lyzwinski_innovative_2024, - title = {Innovative {Approaches} to {Menstruation} and {Fertility} {Tracking} {Using} {Wearable} {Reproductive} {Health} {Technology}: {Systematic} {Review}}, - volume = {26}, - shorttitle = {Innovative {Approaches} to {Menstruation} and {Fertility} {Tracking} {Using} {Wearable} {Reproductive} {Health} {Technology}}, - url = {https://www.jmir.org/2024/1/e45139}, - doi = {10.2196/45139}, - abstract = {Background: Emerging digital health technology has moved into the reproductive health market for female individuals. In the past, mobile health apps have been used to monitor the menstrual cycle using manual entry. New technological trends involve the use of wearable devices to track fertility by assessing physiological changes such as temperature, heart rate, and respiratory rate. -Objective: The primary aims of this study are to review the types of wearables that have been developed and evaluated for menstrual cycle tracking and to examine whether they may detect changes in the menstrual cycle in female individuals. Another aim is to review whether these devices are effective for tracking various stages in the menstrual cycle including ovulation and menstruation. Finally, the secondary aim is to assess whether the studies have validated their findings by reporting accuracy and sensitivity. -Methods: A review of PubMed or MEDLINE was undertaken to evaluate wearable devices for their effectiveness in predicting fertility and differentiating between the different stages of the menstrual cycle. -Results: Fertility cycle–tracking wearables include devices that can be worn on the wrists, on the fingers, intravaginally, and inside the ear. Wearable devices hold promise for predicting different stages of the menstrual cycle including the fertile window and may be used by female individuals as part of their reproductive health. Most devices had high accuracy for detecting fertility and were able to differentiate between the luteal phase (early and late), fertile window, and menstruation by assessing changes in heart rate, heart rate variability, temperature, and respiratory rate. -Conclusions: More research is needed to evaluate consumer perspectives on reproductive technology for monitoring fertility, and ethical issues around the privacy of digital data need to be addressed. Additionally, there is also a need for more studies to validate and confirm this research, given its scarcity, especially in relation to changes in respiratory rate as a proxy for reproductive cycle staging.}, - language = {EN}, - number = {1}, - urldate = {2025-02-21}, - journal = {Journal of Medical Internet Research}, - author = {Lyzwinski, Lynnette and Elgendi, Mohamed and Menon, Carlo}, - month = feb, - year = {2024}, - note = {Company: Journal of Medical Internet Research -Distributor: Journal of Medical Internet Research -Institution: Journal of Medical Internet Research -Label: Journal of Medical Internet Research -Publisher: JMIR Publications Inc., Toronto, Canada}, - pages = {e45139}, - file = {Full Text:/home/alex/Zotero/storage/IP23WZLE/Lyzwinski et al. - 2024 - Innovative Approaches to Menstruation and Fertility Tracking Using Wearable Reproductive Health Tech.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/TJN3WV3I/e45139.html:text/html}, -} - -@misc{noauthor_zyklus-apps_nodate, - title = {Zyklus-{Apps} zur {Verhütung} – sicher oder {Gesellschaftsspiel}? - {ProQuest}}, - shorttitle = {Zyklus-{Apps} zur {Verhütung} – sicher oder {Gesellschaftsspiel}?}, - url = {https://www.proquest.com/openview/739071fff0941b30f3a5d33b56259c60/1?pq-origsite=gscholar&cbl=6629261}, - abstract = {Explore millions of resources from scholarly journals, books, newspapers, videos and more, on the ProQuest Platform.}, - language = {en}, - urldate = {2025-02-21}, - file = {Snapshot:/home/alex/Zotero/storage/QHYUJU9G/1.html:text/html}, -} - -@article{goeckenjan_continuous_2020, - title = {Continuous {Body} {Temperature} {Monitoring} to {Improve} the {Diagnosis} of {Female} {Infertility}}, - volume = {80}, - copyright = {Georg Thieme Verlag KG Stuttgart · New York}, - issn = {0016-5751}, - url = {https://www.thieme-connect.com/products/ejournals/html/10.1055/a-1191-7888}, - doi = {10.1055/a-1191-7888}, - abstract = {Introduction Ovulatory dysfunction is a major cause of female infertility. We evaluated the use of continuous body temperature monitoring with a vaginal biosensor to improve - standard diagnostic procedures for determining ovulatory dysfunction. - -Material and Methods This prospective interventional study was performed in a reproductive medicine department of a university hospital. The menstrual cycles of 51 women with - infertility were monitored and analysed using three different strategies: sonographic and hormonal assessment (standard approach), continuous core body temperature measurement and analysis - using the algorithm of OvulaRing, and lowest daily body temperature measurement monitored with a vaginal biosensor and analysed based on the body temperature curves used in natural family - planning. - -Results Statistically significant differences were found in the temperature curves of women with luteal phase deficiency and polycystic ovary syndrome compared to women with normal - menstrual cycles. The analysis of individual cyclofertilograms can be used to detect cycle phases and estimate the date of ovulation. - -Conclusions Continuous body temperature monitoring with a vaginal biosensor can improve the standard diagnostic procedures used to determine ovulatory dysfunction, especially if - dysfunction is due to luteal phase deficiency and polycystic ovary syndrome. Analysis of the lowest daily body temperature combined with the basal body temperature measurements used in - fertility awareness methods may be equieffective to continuous body temperature measurements with OvulaRing. The results of this study show that a revised diagnostic approach using fewer - hormonal assessments combined with continuous body temperature monitoring can reduce the number of appointments in an infertility clinic as well as the costs.}, - language = {en}, - urldate = {2025-02-21}, - journal = {Geburtshilfe und Frauenheilkunde}, - author = {Goeckenjan, Maren and Schiwek, Esther and Wimberger, Pauline}, - month = jul, - year = {2020}, - note = {Publisher: Georg Thieme Verlag KG}, - keywords = {infertility, Key words - fertility awareness, luteal phase deficiency, polycystic ovary syndrome, vaginal biosensor}, - pages = {702--712}, - file = {Full Text PDF:/home/alex/Zotero/storage/QKPIJD23/Goeckenjan et al. - 2020 - Continuous Body Temperature Monitoring to Improve the Diagnosis of Female Infertility.pdf:application/pdf}, -} - -@article{regidor_identification_2018, - title = {Identification and prediction of the fertile window with a new web-based medical device using a vaginal biosensor for measuring the circadian and circamensual core body temperature}, - volume = {34}, - issn = {0951-3590}, - url = {https://doi.org/10.1080/09513590.2017.1390737}, - doi = {10.1080/09513590.2017.1390737}, - abstract = {Fertility awareness-based (FAB) methods represent a term that includes all family planning methods that are based on the identification of the fertile window. They are based on the woman’s observation of physiological signs of the fertile and infertile phases of the menstrual cycle. The first approach consists basically in symptothermal methods accompanied by cervical mucus measurements and clinical menstrual cycling data recording. The second most often used methods are the urinary measurement of E3G and luteinizing hormone (LH) with a personalized computer system. Hence these systems lack the efficacy of the continuous circadian and circamensual measurement of the core body temperature. Only this approach enables the accurate detection of the ovulation during the fertile window. A new medical device called OvulaRing has been developed to fill this gap. In the present study, the system and its first clinical results are presented. OvulaRing is a medical device used just like a tampon. The device is a vaginal ring of evatane that contains an integrated biosensor. This sensor measures continuously every 5 min the core body temperature throughout the entire cycle. This device allows a circadian and circamensual intravaginal exact measurement. With this system, 288 measurements are created per day. The system can detect retrospectively and predict prospectively the fertile window of the users. One hundred and fifty eight women aged between 18 and 45 years used this medical device in an open non-randomized clinical study for 15 months. A total of 470 cycles could be recorded and were able for analysis. By the same time in a subgroup of patients, hormonal assessments of LH, follicle-stimulating hormone, estradiol and progesterone as well as vaginal ultrasound were performed in parallel between the 9th and the 36th day of the cycle. The validation error due to software errors was 0.89\% for the retrospective analysis; that means that the accuracy for the detection of the ovulation was 99.11\%. Accuracy of 88.8\% for a window of 3 days before ovulation, the day of ovulation and the 3 days after ovulation was achieved for the prospective analysis. In the subgroup of woman with recorded pregnancies, it could be shown that after 3.79 months of use (median) pregnancies were observed. In 67.72\% in up to 3 months, in 16.36\% between 3 and 6 months of use, in 7.27\% between 7 and 9 months, in 5.45\% between 10 and 12 months and in 1.82\% between 13 and 15 months of use of the system. With this new web-based system, a precise determination of the fertile window even in women with ultralong cycles ({\textgreater}35 days) could be detected independently of their personal live circumstances. Exact determination of the fertile window is herewith possible so that OvulaRing represents an evolution in the FAB method for the cycle diagnosis of women with regular, irregular or anovulatory menstrual cycles.}, - number = {3}, - urldate = {2025-02-21}, - journal = {Gynecological Endocrinology}, - author = {Regidor, Pedro-Antonio and Kaczmarczyk, Marta and Schiweck, Esther and Goeckenjan-Festag, Maren and Alexander, Henry}, - month = mar, - year = {2018}, - pmid = {29082805}, - note = {Publisher: Taylor \& Francis -\_eprint: https://doi.org/10.1080/09513590.2017.1390737}, - keywords = {Infertility, central nervous system, circadian rhythm, circamensual rhythm, core body temperature, fertile window, vagina}, - pages = {256--260}, - file = {Full Text PDF:/home/alex/Zotero/storage/ITD68HTW/Regidor et al. - 2018 - Identification and prediction of the fertile window with a new web-based medical device using a vagi.pdf:application/pdf}, -} - -@article{alexander_fertilitatsmonitoring_2014, - title = {Fertilitätsmonitoring mit vaginalem {Biosensor} ({OvulaRing}©)}, - volume = {74}, - issn = {0016-5751}, - url = {https://www.thieme-connect.com/products/ejournals/abstract/10.1055/s-0034-1388603}, - doi = {10.1055/s-0034-1388603}, - abstract = {Thieme E-Books \& E-Journals}, - language = {de}, - urldate = {2025-02-21}, - journal = {Geburtshilfe und Frauenheilkunde}, - author = {Alexander, H. and Kaczmarczyk, M. and Pretzsch, G. and Kersken, T. and Puschmann, D. and Schiwek, E. and Goeckenjan, M.}, - month = sep, - year = {2014}, - keywords = {60. Kongress der Deutschen Gesellschaft für Gynäkologie und Geburtshilfe}, - pages = {FV\_08\_05}, - file = {Snapshot:/home/alex/Zotero/storage/HPL6XYJW/s-0034-1388603.html:text/html}, -} - -@inproceedings{regidor_identifizierung_2018, - title = {Identifizierung und {Vorhersage} des fertilen {Fensters} des weiblichen {Zyklus} mit einem neuen web basierten {Medizinprodukt} ({OvulaRing}®).}, - volume = {78}, - copyright = {Georg Thieme Verlag KG Stuttgart · New York}, - url = {https://www.thieme-connect.com/products/ejournals/html/10.1055/s-0038-1671278}, - doi = {10.1055/s-0038-1671278}, - abstract = {Thieme E-Books \& E-Journals}, - language = {de}, - urldate = {2025-02-21}, - booktitle = {Geburtshilfe und {Frauenheilkunde}}, - publisher = {Georg Thieme Verlag KG}, - author = {Regidor, P. A. and Alexander, H.}, - month = sep, - year = {2018}, - note = {ISSN: 0016-5751}, - keywords = {Präsidentin der DGGG e.V.: Prof. Dr. Birgit Seelbach-Göbel{\textless}/conf-president{\textgreater}{\textless}/conference{\textgreater}}, - pages = {P 23}, - file = {Snapshot:/home/alex/Zotero/storage/DW6578ZM/s-0038-1671278.html:text/html}, -} - -@article{noauthor_prediction_1985, - title = {The prediction of ovulation: a comparison of the basal body temperature graph, cervical mucus score, and real-time pelvic ultrasonography}, - volume = {43}, - issn = {0015-0282}, - shorttitle = {The prediction of ovulation}, - url = {https://www.sciencedirect.com/science/article/pii/S0015028216484360}, - doi = {10.1016/S0015-0282(16)48436-0}, - abstract = {Ninety-five menstrual cycles were studied in 20 women undergoing donor artificial insemination (AID). In 49 cycles basal body temperature (BBT) change…}, - language = {en-US}, - number = {3}, - urldate = {2025-02-20}, - journal = {Fertility and Sterility}, - month = mar, - year = {1985}, - note = {Publisher: Elsevier}, - pages = {385--388}, - file = {Snapshot:/home/alex/Zotero/storage/FYI9GPUM/S0015028216484360.html:text/html}, -} - -@article{sato_novel_2024, - title = {Novel {Methodology} for {Identifying} the {Occurrence} of {Ovulation} by {Estimating} {Core} {Body} {Temperature} {During} {Sleeping}: {Validity} and {Effectiveness} {Study}}, - volume = {8}, - copyright = {Unless stated otherwise, all articles are open-access distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work ("first published in the Journal of Medical Internet Research...") is properly cited with original URL and bibliographic citation information. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.}, - shorttitle = {Novel {Methodology} for {Identifying} the {Occurrence} of {Ovulation} by {Estimating} {Core} {Body} {Temperature} {During} {Sleeping}}, - url = {https://formative.jmir.org/2024/1/e55834}, - doi = {10.2196/55834}, - abstract = {Background: Body temperature is the most-used noninvasive biomarker to determine menstrual cycle and ovulation. However, issues related to its low accuracy are still under discussion. Objective: This study aimed to improve the accuracy of identifying the presence or absence of ovulation within a menstrual cycle. We investigated whether core body temperature (CBT) estimation can improve the accuracy of temperature biphasic shift discrimination in the menstrual cycle. The study consisted of 2 parts: experiment 1 assessed the validity of the CBT estimation method, while experiment 2 focused on the effectiveness of the method in discriminating biphasic temperature shifts. Methods: In experiment 1, healthy women aged between 18 and 40 years had their true CBT measured using an ingestible thermometer and their CBT estimated from skin temperature and ambient temperature measured during sleep in both the follicular and luteal phases of their menstrual cycles. This study analyzed the differences between these 2 measurements, the variations in temperature between the 2 phases, and the repeated measures correlation between the true and estimated CBT. Experiment 2 followed a similar methodology, but focused on evaluating the diagnostic accuracy of these 2 temperature measurement approaches (estimated CBT and traditional oral basal body temperature [BBT]) for identifying ovulatory cycles. This was performed using urine luteinizing hormone (LH) as the reference standard. Menstrual cycles were categorized based on the results of the LH tests, and a temperature shift was identified using a specific criterion called the “three-over-six rule.” This rule and the nested design of the study facilitated the assessment of diagnostic measures, such as sensitivity and specificity. Results: The main findings showed that CBT estimated from skin temperature and ambient temperature during sleep was consistently lower than directly measured CBT in both the follicular and luteal phases of the menstrual cycle. Despite this, the pattern of temperature variation between these phases was comparable for both the estimated and true CBT measurements, suggesting that the estimated CBT accurately reflected the cyclical variations in the true CBT. Significantly, the CBT estimation method showed higher sensitivity and specificity for detecting the occurrence of ovulation than traditional oral BBT measurements, highlighting its potential as an effective tool for reproductive health monitoring. The current method for estimating the CBT provides a practical and noninvasive method for monitoring CBT, which is essential for identifying biphasic shifts in the BBT throughout the menstrual cycle. Conclusions: This study demonstrated that the estimated CBT derived from skin temperature and ambient temperature during sleep accurately captures variations in true CBT and is more accurate in determining the presence or absence of ovulation than traditional oral BBT measurements. This method holds promise for improving reproductive health monitoring and understanding of menstrual cycle dynamics.}, - language = {EN}, - number = {1}, - urldate = {2025-02-20}, - journal = {JMIR Formative Research}, - author = {Sato, Daisuke and Ikarashi, Koyuki and Nakajima, Fumiko and Fujimoto, Tomomi}, - month = jul, - year = {2024}, - note = {Company: JMIR Formative Research -Distributor: JMIR Formative Research -Institution: JMIR Formative Research -Label: JMIR Formative Research -Publisher: JMIR Publications Inc., Toronto, Canada}, - pages = {e55834}, - file = {PubMed Central Full Text PDF:/home/alex/Zotero/storage/NPBA84BU/Sato et al. - 2024 - Novel Methodology for Identifying the Occurrence of Ovulation by Estimating Core Body Temperature Du.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/BAJUVPFE/e55834.html:text/html}, -} - -@article{royston_identifying_1991, - title = {Identifying the fertile phase of the human menstrual cycle}, +@article{b_s_novel_2022, + title = {Novel Technique for Confirmation of the Day of Ovulation and Prediction of Ovulation in Subsequent Cycles Using a Skin-Worn Sensor in a Population With Ovulatory Dysfunction: A Side-by-Side Comparison With Existing Basal Body Temperature Algorithm and Vaginal Core Body Temperature Algorithm}, volume = {10}, - copyright = {Copyright © 1991 John Wiley \& Sons, Ltd.}, - issn = {1097-0258}, - url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/sim.4780100207}, - doi = {10.1002/sim.4780100207}, - abstract = {The identification of the human fertile phase as the time during which a woman or a couple may conceive is elusive. The fertile time depends on many factors in each individual menstrual cycle and may be said to be more of a statistical than a physiological entity. This paper reviews the application of statistical methods to three areas related to conception and the fertile phase. The first is the prediction and detection of ovulation from serial measurements, such as hormones, basal body temperature and cervical mucus, throughout the menstrual cycle. Typically, such variables increase from some baseline level to a peak around ovulation (the most fertile time), then subside to low levels in the postovulatory phase. The statistical challenge is to detect the rise (signalling the onset of potential fertility) and subsequent fall. Analytic methods considered include thresholds, Bayesian change-point models and particularly the cumulative sum (cusum) technique which is both simple to apply and understand, and effective. The second area comprises appropriate methods of analysing and interpreting data from clinical studies of the fertile phase, especially in so-alled natural family planning (NFP) where it is usual for women to observe several indices of potential fertility. Such studies usually try to establish the temporal relationships between markers of the fertile phase and examine the success of different combinations of markers in delineating the fertile time in comparison with a standard ‘defined’ phase, for example, the interval from three days before to two days after the peak of luteinizing hormone. The third area is the assessment of the probability of conception on certain days of the cycle, which is vital to the understanding of the fertile phase and its application to NFP. Direct estimation of such probabilities is impractical; instead, resort must be made to estimation by maximum likelihood of the parameters of specially constructed models. Suitable models are described. Finally, the need for a new prospective study of the probability of conception in relation to the markers of the fertile phase used in the symptothermal method of NFP is discussed.}, - language = {en}, - number = {2}, - urldate = {2025-02-20}, - journal = {Statistics in Medicine}, - author = {Royston, Patrick}, - year = {1991}, - note = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/sim.4780100207}, - pages = {221--240}, - file = {Snapshot:/home/alex/Zotero/storage/9XL95LUJ/sim.html:text/html}, + issn = {2296-4185}, + url = {https://www.frontiersin.org/articles/10.3389/fbioe.2022.807139/full}, + doi = {10.3389/fbioe.2022.807139}, + shorttitle = {Novel Technique for Confirmation of the Day of Ovulation and Prediction of Ovulation in Subsequent Cycles Using a Skin-Worn Sensor in a Population With Ovulatory Dysfunction}, + abstract = {Objective: Determine the accuracy of a novel technique for confirmation of the day of ovulation and prediction of ovulation in subsequent cycles for the purpose of conception using a skin-worn sensor in a population with ovulatory dysfunction. +Methods: A total of 80 participants recorded consecutive overnight temperatures using a skin-worn sensor at the same time as a commercially available vaginal sensor for a total of 205 reproductive cycles. The vaginal sensor and its associated algorithm were used to determine the day of ovulation, and the ovulation results obtained using the skin-worn sensor and its associated algorithm were assessed for comparative accuracy alongside a number of other statistical techniques, with a further assessment of the same skin-derived data by means of the “three over six” rule. A number of parameters were used to divide the data into separate comparative groups, and further secondary statistical analyses were performed. +Results: The skin-worn sensor and its associated algorithm (together labeled “{SWS}”) were 66\% accurate for determining the day of ovulation (±1 day) or the absence of ovulation and 90\% accurate for determining the fertile window (ovulation day ±3 days) in the total study population in comparison to the results obtained from the vaginal sensor and its associated algorithm (together labeled “{VS}”). +Conclusion: {SWS} is a useful tool for confirming the fertile window and absence of ovulation (anovulation) in a population with ovulatory dysfunction, both known and Edited by:}, + pages = {807139}, + journaltitle = {Front. Bioeng. Biotechnol.}, + author = {B. S., Hurst and K., Davies and R. C., Milnes and T. G., Knowles and A., Pirrie}, + urldate = {2025-05-27}, + date = {2022-03-04}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/U8UPIT4Q/B. S. et al. - 2022 - Novel Technique for Confirmation of the Day of Ovulation and Prediction of Ovulation in Subsequent C.pdf:application/pdf}, } -@article{su_detection_2017, - title = {Detection of ovulation, a review of currently available methods}, - volume = {2}, - copyright = {© 2017 The Authors. Bioengineering \& Translational Medicine is published by Wiley Periodicals, Inc. on behalf of The American Institute of Chemical Engineers}, - issn = {2380-6761}, - url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/btm2.10058}, - doi = {10.1002/btm2.10058}, - abstract = {The ability to identify the precise time of ovulation is important for women who want to plan conception or practice contraception. Here, we review the current literature on various methods for detecting ovulation including a review of point-of-care device technology. We incorporate an examination of methods to detect ovulation that have been developed and practiced for decades and analyze the indications and limitations of each—transvaginal ultrasonography, urinary luteinizing hormone detection, serum progesterone and urinary pregnanediol 3-glucuronide detection, urinary follicular stimulating hormone detection, basal body temperature monitoring, and cervical mucus and salivary ferning analysis. Some point-of-care ovulation detection devices have been developed and commercialized based on these methods, however previous research was limited by small sample size and an inconsistent standard reference to true ovulation.}, - language = {en}, - number = {3}, - urldate = {2025-02-20}, - journal = {Bioengineering \& Translational Medicine}, - author = {Su, Hsiu-Wei and Yi, Yu-Chiao and Wei, Ting-Yen and Chang, Ting-Chang and Cheng, Chao-Min}, - year = {2017}, - note = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/btm2.10058}, - keywords = {ovulation detection, family planning, fertility window}, - pages = {238--246}, - file = {Full Text PDF:/home/alex/Zotero/storage/ZEACCGE5/Su et al. - 2017 - Detection of ovulation, a review of currently available methods.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/RDDQD8EA/btm2.html:text/html}, -} - -@article{owen_physiological_2013, - title = {Physiological {Signs} of {Ovulation} and {Fertility} {Readily} {Observable} by {Women}}, - volume = {80}, - issn = {0024-3639}, - url = {https://doi.org/10.1179/0024363912Z.0000000005}, - doi = {10.1179/0024363912Z.0000000005}, - abstract = {IntroductionConfirmation of ovulation can be difficult in clinical practice, as gold standard methods including serial transvaginal ultrasonography, serum luteinizing hormone (LH) measurements, or laparoscopic follicular observation are impractical. Numerous surrogate markers have been proposed and evaluated in relation to these gold standards that have more practical clinical applications.PurposeTo review the evidence on physiological signs of ovulation timing and fertility in order to determine valid markers that can be easily identified by women.MethodsA literature review of primary resources in Ovid Medline was undertaken to identify studies examining physiological signs as they relate to gold standard assessment of ovulation. Studies examining the efficacy/effectiveness of different types of natural family planning were excluded.ResultsThe most commonly encountered physiological signs were urine LH, cervical mucus, and basal body temperature (BBT). Urine LH as assessed by home monitoring systems indicated ovulation 91 percent of the time during the 2 days of peak fertility on the monitor and 97 percent during the 2 peak days plus 1. Cervical mucus peak characteristics were identified 78 percent of the time ±1 day, and 91 percent of the time ±2 days of LH surge indicating ovulation. Further research supports the importance of cervical mucus in overall fertility, as conception rates were more closely related to mucus quality than to timing of intercourse related to ovulation. As a lone indicator of ovulation, BBT is at best a retrospective marker, and functions best in conjunction with other signs of ovulation. Additionally, salivary ferning, salivary and vaginal fluid electrical potential, finger–finger electrical potential, and differential skin temperature were postulated as possible indicators, but were not found to be temporally related to ovulation. The research on differential skin temperature is promising, but minimal thus far in number, and has not been evaluated as an adjunct to BBT as yet.ConclusionHome urinary LH monitors are becoming more widely available and less expensive giving women the potential to assess the ovulatory status of their cycle in real time. Cervical mucus observation is an effective and cost-efficient method, but requires some teaching to increase the confidence of users. In conjunction, LH monitors and cervical mucus can give the best indication of fertility and ovulation timing.}, - language = {en}, - number = {1}, - urldate = {2025-02-20}, - journal = {The Linacre Quarterly}, - author = {Owen, Martin}, - month = jan, - year = {2013}, - note = {Publisher: SAGE Publications Inc}, - pages = {17--23}, - file = {Full Text:/home/alex/Zotero/storage/IBVUICCU/Owen - 2013 - Physiological Signs of Ovulation and Fertility Readily Observable by Women.pdf:application/pdf}, -} - -@article{soumpasis_real-life_2020, - title = {Real-life insights on menstrual cycles and ovulation using big data}, - volume = {2020}, - issn = {2399-3529}, - url = {https://doi.org/10.1093/hropen/hoaa011}, - doi = {10.1093/hropen/hoaa011}, - abstract = {What variations underlie the menstrual cycle length and ovulation day of women trying to conceive?Big data from a connected ovulation test revealed the extent of variation in menstrual cycle length and ovulation day in women trying to conceive.Timing intercourse to coincide with the fertile period of a woman maximises the chances of conception. The day of ovulation varies on an inter- and intra-individual level.A total of 32 595 women who had purchased a connected ovulation test system contributed 75 981 cycles for analysis. Day of ovulation was determined from the fertility test results. The connected home ovulation test system enables users to identify their fertile phase. The app benefits users by enabling them to understand their personal fertility information. During each menstrual cycle, users input their perceived cycle length into an accessory application, and data on hormone levels from the tests are uploaded to the application and stored in an anonymised cloud database. This study compared users’ perceived cycle characteristics with actual cycle characteristics. The perceived and actual cycle length information was analysed to provide population ranges.This study analysed data from the at-home use of a commercially available connected home ovulation test by women across the USA and UK.Overall, 25.3\% of users selected a 28-day cycle as their perceived cycle length; however, only 12.4\% of users actually had a 28-day cycle. Most women (87\%) had actual menstrual cycle lengths between 23 and 35 days, with a normal distribution centred on day 28, and over half of the users (52\%) had cycles that varied by 5 days or more. There was a 10-day spread of observed ovulation days for a 28-day cycle, with the most common day of ovulation being Day 15. Similar variation was observed for all cycle lengths examined. For users who conducted a test on every day requested by the app, a luteinising hormone (LH) surge was detected in 97.9\% of cycles.Data were from a self-selected population of women who were prepared to purchase a commercially available product to aid conception and so may not fully represent the wider population. No corresponding demographic data were collected with the cycle information.Using big data has provided more personalised insights into women’s fertility; this could enable women trying to conceive to better time intercourse, increasing the likelihood of conception.The study was funded by SPD Development Company Ltd (Bedford, UK), a fully owned subsidiary of SPD Swiss Precision Diagnostics GmbH (Geneva, Switzerland). I.S., B.G. and S.J. are employees of the SPD Development Company Ltd.}, - number = {2}, - urldate = {2025-02-20}, - journal = {Human Reproduction Open}, - author = {Soumpasis, I and Grace, B and Johnson, S}, - month = feb, - year = {2020}, - pages = {hoaa011}, - annote = { - -not really interesting, as they only look at cycle length - - -they use “perceived” cycle length, which is hard to defend, when there are intermediate bleedings etc. - - - - - -}, - file = {Full Text PDF:/home/alex/Zotero/storage/PS9UC298/Soumpasis et al. - 2020 - Real-life insights on menstrual cycles and ovulation using big data.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/LTFATQIC/5820371.html:text/html}, -} - -@article{brewis_demographic_2005, - title = {Demographic {Evidence} {That} {Human} {Ovulation} {Is} {Undetectable} ({At} {Least} in {Pair} {Bonds})}, - volume = {46}, - issn = {0011-3204}, - url = {https://www.journals.uchicago.edu/doi/abs/10.1086/430016}, - doi = {10.1086/430016}, - number = {3}, - urldate = {2025-02-20}, - journal = {Current Anthropology}, - author = {Brewis, Alexandra and Meyer, Mary}, - month = jun, - year = {2005}, - note = {Publisher: The University of Chicago Press}, - pages = {465--471}, -} - -@article{noauthor_monitoring_1987, - title = {Monitoring techniques to predict and detect ovulation}, - volume = {47}, +@article{garcia_prediction_1981, + title = {Prediction of the Time of Ovulation*}, + volume = {36}, issn = {0015-0282}, - url = {https://www.sciencedirect.com/science/article/pii/S0015028216500028}, - doi = {10.1016/S0015-0282(16)50002-8}, - abstract = {This study was designed to evaluate the accuracy of various methods in predicting and detecting ovulation in 14 spontaneous and 17 clomiphene citrate …}, - language = {en-US}, - number = {2}, - urldate = {2025-02-20}, - journal = {Fertility and Sterility}, - month = feb, - year = {1987}, - note = {Publisher: Elsevier}, - pages = {259--264}, - file = {PDF:/home/alex/Zotero/storage/RLCZVH5V/1987 - Monitoring techniques to predict and detect ovulation.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/M5P9EZ67/S0015028216500028.html:text/html}, + url = {https://www.sciencedirect.com/science/article/pii/S0015028216457304}, + doi = {10.1016/S0015-0282(16)45730-4}, + abstract = {Prediction of ovulation was established by correlation of clinical parameters, follicular development by ultrasound, and estradiol, progesterone, and luteinizing hormone ({LH}) determination in 71 menstrual cycles. Laparoscopic follicular aspiration was accomplished in 41 of those cycles. A 28-hour interval from the ascending limb of the {LH} seems to be the “ideal time” for retrieval of a preovulatory oocyte. The variability in the amount of {LH} to which the follicle is exposed during the {LH} surge seems to indicate that there is a relatively low specific value necessary for ovulation. Ovulation occurs approximately 10 ± 5 hours from the {LH} peak. Progesterone occurs in relation to the {LH} surge and is helpful for the retrospective analysis of the menstrual cycle.}, + pages = {308--315}, + number = {3}, + journaltitle = {Fertility and Sterility}, + author = {Garcia, Jairo E. and Seegar Jones, Georgeanna and Wright, George L.}, + urldate = {2025-05-27}, + date = {1981-09-01}, + file = {PDF:/home/alex/Zotero/storage/DB8PW3QR/Garcia et al. - 1981 - Prediction of the Time of Ovulation.pdf:application/pdf;ScienceDirect Snapshot:/home/alex/Zotero/storage/QLFLZFDJ/S0015028216457304.html:text/html}, } -@article{noauthor_physiological_2016, - title = {Physiological predictors of ovulation and pregnancy risk in a fixed-time artificial insemination program}, - volume = {99}, - issn = {0022-0302}, - url = {https://www.sciencedirect.com/science/article/pii/S0022030216306725}, - doi = {10.3168/jds.2016-11247}, - abstract = {The objective of this study was to determine the relative importance and contribution of several physiological factors as predictors of pregnancy risk…}, - language = {en-US}, - number = {12}, - urldate = {2025-02-20}, - journal = {Journal of Dairy Science}, - month = dec, - year = {2016}, - note = {Publisher: Elsevier}, - pages = {10077--10092}, - file = {Snapshot:/home/alex/Zotero/storage/EBK9WJP4/S0022030216306725.html:text/html}, +@misc{wang_timexer_2024, + title = {{TimeXer}: Empowering Transformers for Time Series Forecasting with Exogenous Variables}, + url = {http://arxiv.org/abs/2402.19072}, + doi = {10.48550/arXiv.2402.19072}, + shorttitle = {{TimeXer}}, + abstract = {Deep models have demonstrated remarkable performance in time series forecasting. However, due to the partially-observed nature of real-world applications, solely focusing on the target of interest, so-called endogenous variables, is usually insufficient to guarantee accurate forecasting. Notably, a system is often recorded into multiple variables, where the exogenous variables can provide valuable external information for endogenous variables. Thus, unlike well-established multivariate or univariate forecasting paradigms that either treat all the variables equally or ignore exogenous information, this paper focuses on a more practical setting: time series forecasting with exogenous variables. We propose a novel approach, {TimeXer}, to ingest external information to enhance the forecasting of endogenous variables. With deftly designed embedding layers, {TimeXer} empowers the canonical Transformer with the ability to reconcile endogenous and exogenous information, where patch-wise self-attention and variate-wise cross-attention are used simultaneously. Moreover, global endogenous tokens are learned to effectively bridge the causal information underlying exogenous series into endogenous temporal patches. Experimentally, {TimeXer} achieves consistent state-of-the-art performance on twelve real-world forecasting benchmarks and exhibits notable generality and scalability. Code is available at this repository: https://github.com/thuml/{TimeXer}.}, + number = {{arXiv}:2402.19072}, + publisher = {{arXiv}}, + author = {Wang, Yuxuan and Wu, Haixu and Dong, Jiaxiang and Qin, Guo and Zhang, Haoran and Liu, Yong and Qiu, Yunzhong and Wang, Jianmin and Long, Mingsheng}, + urldate = {2025-05-09}, + date = {2024-11-11}, + eprinttype = {arxiv}, + eprint = {2402.19072 [cs]}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Full Text PDF:/home/alex/Zotero/storage/76BQWVIW/Wang et al. - 2024 - TimeXer Empowering Transformers for Time Series Forecasting with Exogenous Variables.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/JL64E9YL/2402.html:text/html}, } -@article{albertson_prediction_1987, - title = {The prediction of ovulation and monitoring of the fertile period}, - volume = {3}, - issn = {1573-7195}, - url = {https://doi.org/10.1007/BF01849284}, - doi = {10.1007/BF01849284}, - abstract = {Simple and reliable methods have been sought for both predicting and confirming ovulation. Application of these methods could include management of infertile couples to aid in conception and for increasing the reliability of natural family planning (NFP) as a method of birth control. With the advent of specific hormone assays, serial measurements of estrogens, progesterone (and metabolites), and luteinizing hormone have been the gold standard of monitoring ovarian function in women, However, newer and simpler methodologies have been described and are currently either in use or being tested. These include the measurement of basal body temperature (BBT), the evaluation of the volume, consistency and electro-conductivity of cervicovaginal fluid, salivary steroid content and cellular enzymatic activity, the use of enzyme-linked immunosorbent assays applied to solid-phase formats, and the investigation of new hormonal molecules as markers of reproductive state and function. These new technologies are described herein and their potential for monitoring ovarian function is discussed.}, - language = {en}, - number = {4}, - urldate = {2025-02-20}, - journal = {Advances in Contraception}, - author = {Albertson, B. D. and Zinaman, M. J.}, - month = dec, - year = {1987}, - keywords = {Birth Control, Estrogen, Luteinizing Hormone, Progesterone, Serial Measurement}, - pages = {263--290}, +@misc{zeng_are_2022, + title = {Are Transformers Effective for Time Series Forecasting?}, + url = {http://arxiv.org/abs/2205.13504}, + doi = {10.48550/arXiv.2205.13504}, + abstract = {Recently, there has been a surge of Transformer-based solutions for the long-term time series forecasting ({LTSF}) task. Despite the growing performance over the past few years, we question the validity of this line of research in this work. Specifically, Transformers is arguably the most successful solution to extract the semantic correlations among the elements in a long sequence. However, in time series modeling, we are to extract the temporal relations in an ordered set of continuous points. While employing positional encoding and using tokens to embed sub-series in Transformers facilitate preserving some ordering information, the nature of the {\textbackslash}emph\{permutation-invariant\} self-attention mechanism inevitably results in temporal information loss. To validate our claim, we introduce a set of embarrassingly simple one-layer linear models named {LTSF}-Linear for comparison. Experimental results on nine real-life datasets show that {LTSF}-Linear surprisingly outperforms existing sophisticated Transformer-based {LTSF} models in all cases, and often by a large margin. Moreover, we conduct comprehensive empirical studies to explore the impacts of various design elements of {LTSF} models on their temporal relation extraction capability. We hope this surprising finding opens up new research directions for the {LTSF} task. We also advocate revisiting the validity of Transformer-based solutions for other time series analysis tasks (e.g., anomaly detection) in the future. Code is available at: {\textbackslash}url\{https://github.com/cure-lab/{LTSF}-Linear\}.}, + number = {{arXiv}:2205.13504}, + publisher = {{arXiv}}, + author = {Zeng, Ailing and Chen, Muxi and Zhang, Lei and Xu, Qiang}, + urldate = {2025-05-09}, + date = {2022-08-17}, + eprinttype = {arxiv}, + eprint = {2205.13504 [cs]}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Full Text PDF:/home/alex/Zotero/storage/V9E95F7E/Zeng et al. - 2022 - Are Transformers Effective for Time Series Forecasting.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/HNVIZG98/2205.html:text/html}, } -@inproceedings{serrano_is_2019-1, - address = {Florence, Italy}, - title = {Is {Attention} {Interpretable}?}, - url = {https://aclanthology.org/P19-1282/}, - doi = {10.18653/v1/P19-1282}, - abstract = {Attention mechanisms have recently boosted performance on a range of NLP tasks. Because attention layers explicitly weight input components' representations, it is also often assumed that attention can be used to identify information that models found important (e.g., specific contextualized word tokens). We test whether that assumption holds by manipulating attention weights in already-trained text classification models and analyzing the resulting differences in their predictions. While we observe some ways in which higher attention weights correlate with greater impact on model predictions, we also find many ways in which this does not hold, i.e., where gradient-based rankings of attention weights better predict their effects than their magnitudes. We conclude that while attention noisily predicts input components' overall importance to a model, it is by no means a fail-safe indicator.}, - urldate = {2025-02-19}, - booktitle = {Proceedings of the 57th {Annual} {Meeting} of the {Association} for {Computational} {Linguistics}}, - publisher = {Association for Computational Linguistics}, - author = {Serrano, Sofia and Smith, Noah A.}, - editor = {Korhonen, Anna and Traum, David and Màrquez, Lluís}, - month = jul, - year = {2019}, - pages = {2931--2951}, - file = {Full Text PDF:/home/alex/Zotero/storage/I6J6YP3C/Serrano and Smith - 2019 - Is Attention Interpretable.pdf:application/pdf}, +@article{barrera-animas_rainfall_2022, + title = {Rainfall prediction: A comparative analysis of modern machine learning algorithms for time-series forecasting}, + volume = {7}, + issn = {26668270}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S266682702100102X}, + doi = {10.1016/j.mlwa.2021.100204}, + shorttitle = {Rainfall prediction}, + abstract = {Rainfall forecasting has gained utmost research relevance in recent times due to its complexities and persistent applications such as flood forecasting and monitoring of pollutant concentration levels, among others. Existing models use complex statistical models that are often too costly, both computationally and budgetary, or are not applied to downstream applications. Therefore, approaches that use Machine Learning algorithms in conjunction with time-series data are being explored as an alternative to overcome these drawbacks. To this end, this study presents a comparative analysis using simplified rainfall estimation models based on conventional Machine Learning algorithms and Deep Learning architectures that are efficient for these downstream applications. Models based on {LSTM}, Stacked-{LSTM}, Bidirectional-{LSTM} Networks, {XGBoost}, and an ensemble of Gradient Boosting Regressor, Linear Support Vector Regression, and an Extra-trees Regressor were compared in the task of forecasting hourly rainfall volumes using time-series data. Climate data from 2000 to 2020 from five major cities in the United Kingdom were used. The evaluation metrics of Loss, Root Mean Squared Error, Mean Absolute Error, and Root Mean Squared Logarithmic Error were used to evaluate the models’ performance. Results show that a Bidirectional-{LSTM} Network can be used as a rainfall forecast model with comparable performance to Stacked-{LSTM} Networks. Among all the models tested, the {StackedLSTM} Network with two hidden layers and the Bidirectional-{LSTM} Network performed best. This suggests that models based on {LSTM}-Networks with fewer hidden layers perform better for this approach; denoting its ability to be applied as an approach for budget-wise rainfall forecast applications.}, + pages = {100204}, + journaltitle = {Machine Learning with Applications}, + author = {Barrera-Animas, Ari Yair and Oyedele, Lukumon O. and Bilal, Muhammad and Akinosho, Taofeek Dolapo and Delgado, Juan Manuel Davila and Akanbi, Lukman Adewale}, + urldate = {2025-05-08}, + date = {2022-03}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/FSXMH5N9/Barrera-Animas et al. - 2022 - Rainfall prediction A comparative analysis of modern machine learning algorithms for time-series fo.pdf:application/pdf}, } -@article{noauthor_attention-based_2023, - title = {An attention-based deep learning model for multi-horizon time series forecasting by considering periodic characteristic}, - volume = {185}, - issn = {0360-8352}, - url = {https://www.sciencedirect.com/science/article/abs/pii/S0360835223006915}, - doi = {10.1016/j.cie.2023.109667}, - abstract = {Recently, transformer-based models have exhibited great performance in multi-horizon time series forecasting tasks. However, the core module of these …}, - language = {en-US}, - urldate = {2025-02-19}, - journal = {Computers \& Industrial Engineering}, - month = nov, - year = {2023}, - note = {Publisher: Pergamon}, - pages = {109667}, - file = {Snapshot:/home/alex/Zotero/storage/TJ634T5V/S0360835223006915.html:text/html}, +@article{ahmed_empirical_2010, + title = {An Empirical Comparison of Machine Learning Models for Time Series Forecasting}, + volume = {29}, + issn = {0747-4938, 1532-4168}, + url = {http://www.tandfonline.com/doi/abs/10.1080/07474938.2010.481556}, + doi = {10.1080/07474938.2010.481556}, + pages = {594--621}, + number = {5}, + journaltitle = {Econometric Reviews}, + author = {Ahmed, Nesreen K. and Atiya, Amir F. and Gayar, Neamat El and El-Shishiny, Hisham}, + urldate = {2025-05-08}, + date = {2010-08-30}, + langid = {english}, } -@article{hu_pattern-oriented_2025, - title = {Pattern-oriented {Attention} {Mechanism} for {Multivariate} {Time} {Series} {Forecasting}}, - volume = {19}, - issn = {1556-4681}, - url = {https://doi.org/10.1145/3712606}, - doi = {10.1145/3712606}, - abstract = {Multivariate time series forecasting is applied in many domains, such as finance, transportation, and industry. The main challenge of precise forecasting lies in accurately capturing latent dependencies. Recent studies develop various frameworks to reduce computational complexity or to enhance the learning of intricate relationships, while lacking interpretability and generality. In this article, we aim to elucidate the capture of dependencies as the recognition of patterns. We believe that patterns can be formally described from two aspects: the shapes of segments that frequently repeat and the corresponding forms of repetitions. Drawing upon this idea, we design a multivariate time series forecasting model named PRformer,1 which incorporates a pattern-oriented attention mechanism and a pattern-based projector. The attention mechanism can perceive different forms of repetitions by embedded with various similarity evaluation metrics between segments, and filter out noise from segments to extract potential patterns with a statistical-driven weighting scheme. The pattern-based projector is employed to form the forecasting results by deriving the representative patterns from the set of potential ones. By incorporating explicit definitions of patterns, PRformer is interpretable and general to various time series scenarios. Experimental results on seven datasets demonstrate that PRformer outperforms six state-of-the-art models by about 10.7\% in forecasting accuracy.}, - number = {2}, - urldate = {2025-02-19}, - journal = {ACM Trans. Knowl. Discov. Data}, - author = {Hu, Hanwen and Han, Zhangchi and Qian, Shiyou and Yang, Dingyu and Cao, Jian and Xue, Guangtao}, - month = feb, - year = {2025}, - pages = {38:1--38:26}, +@misc{saluja_towards_2021, + title = {Towards a Rigorous Evaluation of Explainability for Multivariate Time Series}, + url = {http://arxiv.org/abs/2104.04075}, + doi = {10.48550/arXiv.2104.04075}, + abstract = {Machine learning-based systems are rapidly gaining popularity and in-line with that there has been a huge research surge in the field of explainability to ensure that machine learning models are reliable, fair, and can be held liable for their decision-making process. Explainable Artificial Intelligence ({XAI}) methods are typically deployed to debug black-box machine learning models but in comparison to tabular, text, and image data, explainability in time series is still relatively unexplored. The aim of this study was to achieve and evaluate model agnostic explainability in a time series forecasting problem. This work focused on proving a solution for a digital consultancy company aiming to find a data-driven approach in order to understand the effect of their sales related activities on the sales deals closed. The solution involved framing the problem as a time series forecasting problem to predict the sales deals and the explainability was achieved using two novel model agnostic explainability techniques, Local explainable model-agnostic explanations ({LIME}) and Shapley additive explanations ({SHAP}) which were evaluated using human evaluation of explainability. The results clearly indicate that the explanations produced by {LIME} and {SHAP} greatly helped lay humans in understanding the predictions made by the machine learning model. The presented work can easily be extended to any time}, + number = {{arXiv}:2104.04075}, + publisher = {{arXiv}}, + author = {Saluja, Rohit and Malhi, Avleen and Knapič, Samanta and Främling, Kary and Cavdar, Cicek}, + urldate = {2025-05-06}, + date = {2021-04-06}, + eprinttype = {arxiv}, + eprint = {2104.04075 [cs]}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Preprint PDF:/home/alex/Zotero/storage/6XB73Y2F/Saluja et al. - 2021 - Towards a Rigorous Evaluation of Explainability for Multivariate Time Series.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/CDCGX8EZ/2104.html:text/html}, } -@misc{helbling_conceptattention_2025, - title = {{ConceptAttention}: {Diffusion} {Transformers} {Learn} {Highly} {Interpretable} {Features}}, - shorttitle = {{ConceptAttention}}, - url = {http://arxiv.org/abs/2502.04320}, - doi = {10.48550/arXiv.2502.04320}, - abstract = {Do the rich representations of multi-modal diffusion transformers (DiTs) exhibit unique properties that enhance their interpretability? We introduce ConceptAttention, a novel method that leverages the expressive power of DiT attention layers to generate high-quality saliency maps that precisely locate textual concepts within images. Without requiring additional training, ConceptAttention repurposes the parameters of DiT attention layers to produce highly contextualized concept embeddings, contributing the major discovery that performing linear projections in the output space of DiT attention layers yields significantly sharper saliency maps compared to commonly used cross-attention mechanisms. Remarkably, ConceptAttention even achieves state-of-the-art performance on zero-shot image segmentation benchmarks, outperforming 11 other zero-shot interpretability methods on the ImageNet-Segmentation dataset and on a single-class subset of PascalVOC. Our work contributes the first evidence that the representations of multi-modal DiT models like Flux are highly transferable to vision tasks like segmentation, even outperforming multi-modal foundation models like CLIP.}, - urldate = {2025-02-25}, - publisher = {arXiv}, - author = {Helbling, Alec and Meral, Tuna Han Salih and Hoover, Ben and Yanardag, Pinar and Chau, Duen Horng}, - month = feb, - year = {2025}, - note = {arXiv:2502.04320 [cs]}, - keywords = {Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition}, - file = {Preprint PDF:/home/alex/Zotero/storage/AEEDM4ZW/Helbling et al. - 2025 - ConceptAttention Diffusion Transformers Learn Highly Interpretable Features.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/SAHGAINH/2502.html:text/html}, +@inproceedings{hsieh_explainable_2021, + location = {Virtual Event Israel}, + title = {Explainable Multivariate Time Series Classification: A Deep Neural Network Which Learns to Attend to Important Variables As Well As Time Intervals}, + isbn = {978-1-4503-8297-7}, + url = {https://dl.acm.org/doi/10.1145/3437963.3441815}, + doi = {10.1145/3437963.3441815}, + shorttitle = {Explainable Multivariate Time Series Classification}, + eventtitle = {{WSDM} '21: The Fourteenth {ACM} International Conference on Web Search and Data Mining}, + pages = {607--615}, + booktitle = {Proceedings of the 14th {ACM} International Conference on Web Search and Data Mining}, + publisher = {{ACM}}, + author = {Hsieh, Tsung-Yu and Wang, Suhang and Sun, Yiwei and Honavar, Vasant}, + urldate = {2025-05-06}, + date = {2021-03-08}, + langid = {english}, } -@misc{chefer_transformer_2021, - title = {Transformer {Interpretability} {Beyond} {Attention} {Visualization}}, - url = {http://arxiv.org/abs/2012.09838}, - doi = {10.48550/arXiv.2012.09838}, - abstract = {Self-attention techniques, and specifically Transformers, are dominating the field of text processing and are becoming increasingly popular in computer vision classification tasks. In order to visualize the parts of the image that led to a certain classification, existing methods either rely on the obtained attention maps or employ heuristic propagation along the attention graph. In this work, we propose a novel way to compute relevancy for Transformer networks. The method assigns local relevance based on the Deep Taylor Decomposition principle and then propagates these relevancy scores through the layers. This propagation involves attention layers and skip connections, which challenge existing methods. Our solution is based on a specific formulation that is shown to maintain the total relevancy across layers. We benchmark our method on very recent visual Transformer networks, as well as on a text classification problem, and demonstrate a clear advantage over the existing explainability methods.}, - urldate = {2025-02-25}, - publisher = {arXiv}, - author = {Chefer, Hila and Gur, Shir and Wolf, Lior}, - month = apr, - year = {2021}, - note = {arXiv:2012.09838 [cs]}, - keywords = {Computer Science - Computer Vision and Pattern Recognition}, - file = {Preprint PDF:/home/alex/Zotero/storage/3FRISAP7/Chefer et al. - 2021 - Transformer Interpretability Beyond Attention Visualization.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/QUPBJPC9/2012.html:text/html}, +@article{leon-lopez_anomaly_2022, + title = {Anomaly Detection and Classification in Multispectral Time Series Based on Hidden Markov Models}, + volume = {60}, + issn = {1558-0644}, + url = {https://ieeexplore.ieee.org/abstract/document/9509347}, + doi = {10.1109/TGRS.2021.3101127}, + abstract = {Monitoring agriculture from satellite remote sensing data, such as multispectral images, has become a powerful tool since it has demonstrated a great potential for providing timely and accurate knowledge of crops. Detecting anomalies in time series of multispectral remote sensing images for crop monitoring is generally performed using a large sample of historical data at a pixel level. Conversely, this article presents a framework for anomaly detection ({AD}), localization, and classification that exploits the temporal information contained in a given season at a parcel level to detect and localize outliers using hidden Markov models ({HMMs}). Specifically, the {AD} part is based on the learning of {HMM} parameters associated with unlabeled normal data that are used in a second step to detect abnormal crop parcels referred to as anomalies. The learned {HMM} can also be used in time segments to temporally localize the anomalies affecting the crop parcels. The detected and localized anomalies are finally classified using a supervised classifier, e.g., based on support vector machines. The proposed framework is applicable to images partially covered by clouds and can handle a set of crop parcels acquired in the same season bypassing problems due to crop rotations. Numerical experiments are conducted on synthetic and real data, where the real data correspond to vegetation indices extracted from several multitemporal Sentinel-2 images of rapeseed crops. The proposed approach is compared to standard {AD} methods yielding better detection rates with the advantage of allowing anomalies to be localized and characterized.}, + pages = {1--11}, + journaltitle = {{IEEE} Transactions on Geoscience and Remote Sensing}, + author = {León-López, Kareth M. and Mouret, Florian and Arguello, Henry and Tourneret, Jean-Yves}, + urldate = {2025-05-06}, + date = {2022}, + keywords = {Time series analysis, Agricultural monitoring, Agriculture, anomaly classification, Anomaly detection, anomaly detection ({AD}), Feature extraction, Hidden Markov models, hidden Markov models ({HMMs}), Monitoring, remote sensing, time series, Vegetation mapping}, + file = {Snapshot:/home/alex/Zotero/storage/HIWT8MDH/9509347.html:text/html;Submitted Version:/home/alex/Zotero/storage/S4K2XXHN/León-López et al. - 2022 - Anomaly Detection and Classification in Multispectral Time Series Based on Hidden Markov Models.pdf:application/pdf}, } -@misc{sprang_enforcing_2024, - title = {Enforcing {Interpretability} in {Time} {Series} {Transformers}: {A} {Concept} {Bottleneck} {Framework}}, - shorttitle = {Enforcing {Interpretability} in {Time} {Series} {Transformers}}, - url = {http://arxiv.org/abs/2410.06070}, - doi = {10.48550/arXiv.2410.06070}, - abstract = {There has been a recent push of research on Transformer-based models for long-term time series forecasting, even though they are inherently difficult to interpret and explain. While there is a large body of work on interpretability methods for various domains and architectures, the interpretability of Transformer-based forecasting models remains largely unexplored. To address this gap, we develop a framework based on Concept Bottleneck Models to enforce interpretability of time series Transformers. We modify the training objective to encourage a model to develop representations similar to predefined interpretable concepts. In our experiments, we enforce similarity using Centered Kernel Alignment, and the predefined concepts include time features and an interpretable, autoregressive surrogate model (AR). We apply the framework to the Autoformer model, and present an in-depth analysis for a variety of benchmark tasks. We find that the model performance remains mostly unaffected, while the model shows much improved interpretability. Additionally, interpretable concepts become local, which makes the trained model easily intervenable. As a proof of concept, we demonstrate a successful intervention in the scenario of a time shift in the data, which eliminates the need to retrain.}, - urldate = {2025-02-25}, - publisher = {arXiv}, - author = {Sprang, Angela van and Acar, Erman and Zuidema, Willem}, - month = oct, - year = {2024}, - note = {arXiv:2410.06070 [cs]}, +@article{wang_systematic_2022, + title = {A Systematic Review of Time Series Classification Techniques Used in Biomedical Applications}, + volume = {22}, + rights = {https://creativecommons.org/licenses/by/4.0/}, + issn = {1424-8220}, + url = {https://www.mdpi.com/1424-8220/22/20/8016}, + doi = {10.3390/s22208016}, + abstract = {Background: Digital clinical measures collected via various digital sensing technologies such as smartphones, smartwatches, wearables, and ingestible and implantable sensors are increasingly used by individuals and clinicians to capture the health outcomes or behavioral and physiological characteristics of individuals. Time series classification ({TSC}) is very commonly used for modeling digital clinical measures. While deep learning models for {TSC} are very common and powerful, there exist some fundamental challenges. This review presents the non-deep learning models that are commonly used for time series classification in biomedical applications that can achieve high performance. Objective: We performed a systematic review to characterize the techniques that are used in time series classification of digital clinical measures throughout all the stages of data processing and model building. Methods: We conducted a literature search on {PubMed}, as well as the Institute of Electrical and Electronics Engineers ({IEEE}), Web of Science, and {SCOPUS} databases using a range of search terms to retrieve peer-reviewed articles that report on the academic research about digital clinical measures from a five-year period between June 2016 and June 2021. We identified and categorized the research studies based on the types of classification algorithms and sensor input types. Results: We found 452 papers in total from four different databases: {PubMed}, {IEEE}, Web of Science Database, and {SCOPUS}. After removing duplicates and irrelevant papers, 135 articles remained for detailed review and data extraction. Among these, engineered features using time series methods that were subsequently fed into widely used machine learning classifiers were the most commonly used technique, and also most frequently achieved the best performance metrics (77 out of 135 articles). Statistical modeling (24 out of 135 articles) algorithms were the second most common and also the second-best classification technique. Conclusions: In this review paper, summaries of the time series classification models and interpretation methods for biomedical applications are summarized and categorized. While high time series classification performance has been achieved in digital clinical, physiological, or biomedical measures, no standard benchmark datasets, modeling methods, or reporting methodology exist. There is no single widely used method for time series model development or feature interpretation, however many different methods have proven successful.}, + pages = {8016}, + number = {20}, + journaltitle = {Sensors}, + author = {Wang, Will Ke and Chen, Ina and Hershkovich, Leeor and Yang, Jiamu and Shetty, Ayush and Singh, Geetika and Jiang, Yihang and Kotla, Aditya and Shang, Jason Zisheng and Yerrabelli, Rushil and Roghanizad, Ali R. and Shandhi, Md Mobashir Hasan and Dunn, Jessilyn}, + urldate = {2025-05-06}, + date = {2022-10-20}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/H5LLUB5K/Wang et al. - 2022 - A Systematic Review of Time Series Classification Techniques Used in Biomedical Applications.pdf:application/pdf}, +} + +@article{masini_machine_2023, + title = {Machine learning advances for time series forecasting}, + volume = {37}, + issn = {0950-0804, 1467-6419}, + url = {https://onlinelibrary.wiley.com/doi/10.1111/joes.12429}, + doi = {10.1111/joes.12429}, + abstract = {Abstract + In this paper, we survey the most recent advances in supervised machine learning ({ML}) and high‐dimensional models for time‐series forecasting. We consider both linear and nonlinear alternatives. Among the linear methods, we pay special attention to penalized regressions and ensemble of models. The nonlinear methods considered in the paper include shallow and deep neural networks, in their feedforward and recurrent versions, and tree‐based methods, such as random forests and boosted trees. We also consider ensemble and hybrid models by combining ingredients from different alternatives. Tests for superior predictive ability are briefly reviewed. Finally, we discuss application of {ML} in economics and finance and provide an illustration with high‐frequency financial data.}, + pages = {76--111}, + number = {1}, + journaltitle = {Journal of Economic Surveys}, + author = {Masini, Ricardo P. and Medeiros, Marcelo C. and Mendes, Eduardo F.}, + urldate = {2025-05-06}, + date = {2023-02}, + langid = {english}, + file = {Submitted Version:/home/alex/Zotero/storage/4TZJJUSU/Masini et al. - 2023 - Machine learning advances for time series forecasting.pdf:application/pdf}, +} + +@article{gharehbaghi_deep_2018, + title = {A Deep Machine Learning Method for Classifying Cyclic Time Series of Biological Signals Using Time-Growing Neural Network}, + volume = {29}, + rights = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/{IEEE}.html}, + issn = {2162-237X, 2162-2388}, + url = {https://ieeexplore.ieee.org/document/8066455/}, + doi = {10.1109/TNNLS.2017.2754294}, + pages = {4102--4115}, + number = {9}, + journaltitle = {{IEEE} Trans. Neural Netw. Learning Syst.}, + author = {Gharehbaghi, Arash and Linden, Maria}, + urldate = {2025-05-06}, + date = {2018-09}, +} + +@article{wang_systematic_2022-1, + title = {A Systematic Review of Time Series Classification Techniques Used in Biomedical Applications}, + volume = {22}, + issn = {1424-8220}, + url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9611376/}, + doi = {10.3390/s22208016}, + abstract = {Background: Digital clinical measures collected via various digital sensing technologies such as smartphones, smartwatches, wearables, and ingestible and implantable sensors are increasingly used by individuals and clinicians to capture the health outcomes or behavioral and physiological characteristics of individuals. Time series classification ({TSC}) is very commonly used for modeling digital clinical measures. While deep learning models for {TSC} are very common and powerful, there exist some fundamental challenges. This review presents the non-deep learning models that are commonly used for time series classification in biomedical applications that can achieve high performance. Objective: We performed a systematic review to characterize the techniques that are used in time series classification of digital clinical measures throughout all the stages of data processing and model building. Methods: We conducted a literature search on {PubMed}, as well as the Institute of Electrical and Electronics Engineers ({IEEE}), Web of Science, and {SCOPUS} databases using a range of search terms to retrieve peer-reviewed articles that report on the academic research about digital clinical measures from a five-year period between June 2016 and June 2021. We identified and categorized the research studies based on the types of classification algorithms and sensor input types. Results: We found 452 papers in total from four different databases: {PubMed}, {IEEE}, Web of Science Database, and {SCOPUS}. After removing duplicates and irrelevant papers, 135 articles remained for detailed review and data extraction. Among these, engineered features using time series methods that were subsequently fed into widely used machine learning classifiers were the most commonly used technique, and also most frequently achieved the best performance metrics (77 out of 135 articles). Statistical modeling (24 out of 135 articles) algorithms were the second most common and also the second-best classification technique. Conclusions: In this review paper, summaries of the time series classification models and interpretation methods for biomedical applications are summarized and categorized. While high time series classification performance has been achieved in digital clinical, physiological, or biomedical measures, no standard benchmark datasets, modeling methods, or reporting methodology exist. There is no single widely used method for time series model development or feature interpretation, however many different methods have proven successful.}, + pages = {8016}, + number = {20}, + journaltitle = {Sensors (Basel)}, + author = {Wang, Will Ke and Chen, Ina and Hershkovich, Leeor and Yang, Jiamu and Shetty, Ayush and Singh, Geetika and Jiang, Yihang and Kotla, Aditya and Shang, Jason Zisheng and Yerrabelli, Rushil and Roghanizad, Ali R. and Shandhi, Md Mobashir Hasan and Dunn, Jessilyn}, + urldate = {2025-05-06}, + date = {2022-10-20}, + pmid = {36298367}, + pmcid = {PMC9611376}, + file = {Full Text PDF:/home/alex/Zotero/storage/2E8EIVER/Wang et al. - 2022 - A Systematic Review of Time Series Classification Techniques Used in Biomedical Applications.pdf:application/pdf}, +} + +@misc{dauphin_language_2017, + title = {Language Modeling with Gated Convolutional Networks}, + url = {http://arxiv.org/abs/1612.08083}, + doi = {10.48550/arXiv.1612.08083}, + abstract = {The pre-dominant approach to language modeling to date is based on recurrent neural networks. Their success on this task is often linked to their ability to capture unbounded context. In this paper we develop a finite context approach through stacked convolutions, which can be more efficient since they allow parallelization over sequential tokens. We propose a novel simplified gating mechanism that outperforms Oord et al (2016) and investigate the impact of key architectural decisions. The proposed approach achieves state-of-the-art on the {WikiText}-103 benchmark, even though it features long-term dependencies, as well as competitive results on the Google Billion Words benchmark. Our model reduces the latency to score a sentence by an order of magnitude compared to a recurrent baseline. To our knowledge, this is the first time a non-recurrent approach is competitive with strong recurrent models on these large scale language tasks.}, + number = {{arXiv}:1612.08083}, + publisher = {{arXiv}}, + author = {Dauphin, Yann N. and Fan, Angela and Auli, Michael and Grangier, David}, + urldate = {2025-03-26}, + date = {2017-09-08}, + eprinttype = {arxiv}, + eprint = {1612.08083 [cs]}, + keywords = {Computer Science - Computation and Language}, + file = {Full Text PDF:/home/alex/Zotero/storage/4SBUNZ4A/Dauphin et al. - 2017 - Language Modeling with Gated Convolutional Networks.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/TQBL4EZ7/1612.html:text/html}, +} + +@misc{ba_layer_2016, + title = {Layer Normalization}, + url = {http://arxiv.org/abs/1607.06450}, + doi = {10.48550/arXiv.1607.06450}, + abstract = {Training state-of-the-art, deep neural networks is computationally expensive. One way to reduce the training time is to normalize the activities of the neurons. A recently introduced technique called batch normalization uses the distribution of the summed input to a neuron over a mini-batch of training cases to compute a mean and variance which are then used to normalize the summed input to that neuron on each training case. This significantly reduces the training time in feed-forward neural networks. However, the effect of batch normalization is dependent on the mini-batch size and it is not obvious how to apply it to recurrent neural networks. In this paper, we transpose batch normalization into layer normalization by computing the mean and variance used for normalization from all of the summed inputs to the neurons in a layer on a single training case. Like batch normalization, we also give each neuron its own adaptive bias and gain which are applied after the normalization but before the non-linearity. Unlike batch normalization, layer normalization performs exactly the same computation at training and test times. It is also straightforward to apply to recurrent neural networks by computing the normalization statistics separately at each time step. Layer normalization is very effective at stabilizing the hidden state dynamics in recurrent networks. Empirically, we show that layer normalization can substantially reduce the training time compared with previously published techniques.}, + number = {{arXiv}:1607.06450}, + publisher = {{arXiv}}, + author = {Ba, Jimmy Lei and Kiros, Jamie Ryan and Hinton, Geoffrey E.}, + urldate = {2025-03-26}, + date = {2016-07-21}, + eprinttype = {arxiv}, + eprint = {1607.06450 [stat]}, + keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, + file = {Full Text PDF:/home/alex/Zotero/storage/MJWRDPWE/Ba et al. - 2016 - Layer Normalization.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/F9WSU957/1607.html:text/html}, +} + +@misc{clevert_fast_2016, + title = {Fast and Accurate Deep Network Learning by Exponential Linear Units ({ELUs})}, + url = {http://arxiv.org/abs/1511.07289}, + doi = {10.48550/arXiv.1511.07289}, + abstract = {We introduce the "exponential linear unit" ({ELU}) which speeds up learning in deep neural networks and leads to higher classification accuracies. Like rectified linear units ({ReLUs}), leaky {ReLUs} ({LReLUs}) and parametrized {ReLUs} ({PReLUs}), {ELUs} alleviate the vanishing gradient problem via the identity for positive values. However, {ELUs} have improved learning characteristics compared to the units with other activation functions. In contrast to {ReLUs}, {ELUs} have negative values which allows them to push mean unit activations closer to zero like batch normalization but with lower computational complexity. Mean shifts toward zero speed up learning by bringing the normal gradient closer to the unit natural gradient because of a reduced bias shift effect. While {LReLUs} and {PReLUs} have negative values, too, they do not ensure a noise-robust deactivation state. {ELUs} saturate to a negative value with smaller inputs and thereby decrease the forward propagated variation and information. Therefore, {ELUs} code the degree of presence of particular phenomena in the input, while they do not quantitatively model the degree of their absence. In experiments, {ELUs} lead not only to faster learning, but also to significantly better generalization performance than {ReLUs} and {LReLUs} on networks with more than 5 layers. On {CIFAR}-100 {ELUs} networks significantly outperform {ReLU} networks with batch normalization while batch normalization does not improve {ELU} networks. {ELU} networks are among the top 10 reported {CIFAR}-10 results and yield the best published result on {CIFAR}-100, without resorting to multi-view evaluation or model averaging. On {ImageNet}, {ELU} networks considerably speed up learning compared to a {ReLU} network with the same architecture, obtaining less than 10\% classification error for a single crop, single model network.}, + number = {{arXiv}:1511.07289}, + publisher = {{arXiv}}, + author = {Clevert, Djork-Arné and Unterthiner, Thomas and Hochreiter, Sepp}, + urldate = {2025-03-26}, + date = {2016-02-22}, + eprinttype = {arxiv}, + eprint = {1511.07289 [cs]}, keywords = {Computer Science - Machine Learning}, - file = {Preprint PDF:/home/alex/Zotero/storage/HVUXXRXJ/Sprang et al. - 2024 - Enforcing Interpretability in Time Series Transformers A Concept Bottleneck Framework.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/6QS78DZP/2410.html:text/html}, + file = {Full Text PDF:/home/alex/Zotero/storage/PU3ZGP4G/Clevert et al. - 2016 - Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs).pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/3MAW2IWE/1511.html:text/html}, } -@article{yuan_dcfa-itimenet_2024, - title = {{DCFA}-{iTimeNet}: {Dynamic} cross-fusion attention network for interpretable time series prediction}, - volume = {55}, - issn = {1573-7497}, - shorttitle = {{DCFA}-{iTimeNet}}, - url = {https://doi.org/10.1007/s10489-024-05973-2}, - doi = {10.1007/s10489-024-05973-2}, - abstract = {Although time series prediction research among engineering and technology has made breakthrough progress in performance, challenges remain in modeling complex dynamic interactions between variables and interpretability. To address these two problems, a novel two-stage strategy framework called DCFA-iTimeNet is introduced. In the first stage, this paper innovatively proposes a dynamic cross-fusion attention mechanism (DCFA) . This module facilitates the model to exchange information between different patches of the time series, thereby capturing the complex interactions between variables across time. In the second stage, we exploit a decomposition-based linear explainable Bidirectional Gated Recurrent Unit (DeLEBiGRU), which consists mainly of standard BiGRU and tensorized BiGRU. It is proposed to analyze each variable’s historical long-term, instantaneous, and future impacts. Such design is crucial for understanding how each variable impacts the overall prediction over time. Extensive experimental results demonstrate that the proposed model can effectively model and interpret complex dynamic relationships of multivariate time series and understand the model’s decision-making process. Moreover, the performance outperforms the state-of-the-art methods.}, - language = {en}, - number = {2}, - urldate = {2025-02-25}, - journal = {Applied Intelligence}, - author = {Yuan, Jianjun and Wu, Fujun and Zhao, Luoming and Pan, Dongbo and Yu, Xinyue}, - month = dec, - year = {2024}, - keywords = {Artificial Intelligence, Dynamic cross-fusion attention, Dynamic interaction, Interpretability, Time series prediction}, - pages = {86}, - file = {Full Text PDF:/home/alex/Zotero/storage/JEXYN7BN/Yuan et al. - 2024 - DCFA-iTimeNet Dynamic cross-fusion attention network for interpretable time series prediction.pdf:application/pdf}, +@online{noauthor_temporal_nodate, + title = {Temporal Fusion Transformer ({TFT}) — darts documentation}, + url = {https://unit8co.github.io/darts/generated_api/darts.models.forecasting.tft_model.html}, + urldate = {2025-03-19}, + file = {Temporal Fusion Transformer (TFT) — darts documentation:/home/alex/Zotero/storage/5QNI6WSL/darts.models.forecasting.tft_model.html:text/html}, } -@inproceedings{guo_exploring_2019, - title = {Exploring interpretable {LSTM} neural networks over multi-variable data}, - url = {https://proceedings.mlr.press/v97/guo19b.html}, - abstract = {For recurrent neural networks trained on time series with target and exogenous variables, in addition to accurate prediction, it is also desired to provide interpretable insights into the data. In this paper, we explore the structure of LSTM recurrent neural networks to learn variable-wise hidden states, with the aim to capture different dynamics in multi-variable time series and distinguish the contribution of variables to the prediction. With these variable-wise hidden states, a mixture attention mechanism is proposed to model the generative process of the target. Then we develop associated training methods to jointly learn network parameters, variable and temporal importance w.r.t the prediction of the target variable. Extensive experiments on real datasets demonstrate enhanced prediction performance by capturing the dynamics of different variables. Meanwhile, we evaluate the interpretation results both qualitatively and quantitatively. It exhibits the prospect as an end-to-end framework for both forecasting and knowledge extraction over multi-variable data.}, - language = {en}, - urldate = {2025-02-25}, - booktitle = {Proceedings of the 36th {International} {Conference} on {Machine} {Learning}}, - publisher = {PMLR}, - author = {Guo, Tian and Lin, Tao and Antulov-Fantulin, Nino}, - month = may, - year = {2019}, - note = {ISSN: 2640-3498}, - pages = {2494--2504}, - file = {Full Text PDF:/home/alex/Zotero/storage/VV3I2T4E/Guo et al. - 2019 - Exploring interpretable LSTM neural networks over multi-variable data.pdf:application/pdf;Supplementary PDF:/home/alex/Zotero/storage/57IK29PA/Guo et al. - 2019 - Exploring interpretable LSTM neural networks over multi-variable data.pdf:application/pdf}, +@software{sherar_mattsherartemporal_fusion_transform_2025, + title = {mattsherar/Temporal\_Fusion\_Transform}, + url = {https://github.com/mattsherar/Temporal_Fusion_Transform}, + abstract = {Pytorch Implementation of Google's {TFT}}, + author = {Sherar, Matthew}, + urldate = {2025-03-19}, + date = {2025-03-07}, + note = {original-date: 2020-01-11T17:54:01Z}, } -@incollection{iliadis_temporal_2023, - address = {Cham}, - title = {Temporal {Attention} {Signatures} for {Interpretable} {Time}-{Series} {Prediction}}, - volume = {14259}, - isbn = {978-3-031-44222-3 978-3-031-44223-0}, - url = {https://link.springer.com/10.1007/978-3-031-44223-0_22}, - abstract = {Deep neural networks have become a staple in time-series prediction due to their remarkable accuracy. However, their internal workings often remain elusive. Significant advancements have been made in the interpretability of these networks, with attention mechanisms and feature maps being notably effective for image classification by highlighting the crucial data points. While human observers can readily confirm the significance of features in image classification, the interpretability of time-series data and its modeling remains challenging. To address this, we put forth an innovative approach that unifies temporal attention and visualization as a blend of recurrent neural networks, self-attention, and general attention. This synergy results in the generation of temporal attention signatures, akin to image attention heat maps. Temporal attention not only enhances prediction accuracy beyond that of recurrent networks alone but also demonstrates that varying label classes yield distinct attention signatures. This observation indicates that neural networks focus on different sections of time-series sequences contingent on the prediction target. We conclude with a discussion on the practical implications of this novel approach, including its applicability to model interpretation, sequence length selection, and model validation. This leads to more accurate, robust, and interpretable models, instilling greater confidence in their results.}, - language = {en}, - urldate = {2025-02-25}, - booktitle = {Artificial {Neural} {Networks} and {Machine} {Learning} – {ICANN} 2023}, - publisher = {Springer Nature Switzerland}, - author = {Katrompas, Alexander and Metsis, Vangelis}, - editor = {Iliadis, Lazaros and Papaleonidas, Antonios and Angelov, Plamen and Jayne, Chrisina}, - year = {2023}, - doi = {10.1007/978-3-031-44223-0_22}, - note = {Series Title: Lecture Notes in Computer Science}, - pages = {268--280}, - file = {PDF:/home/alex/Zotero/storage/LV7IVKZK/Katrompas and Metsis - 2023 - Temporal Attention Signatures for Interpretable Time-Series Prediction.pdf:application/pdf}, +@software{noauthor_playtikaosstft-torch_2025, + title = {{PlaytikaOSS}/tft-torch}, + rights = {{MIT}}, + url = {https://github.com/PlaytikaOSS/tft-torch}, + abstract = {A Python library that implements ״Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting״}, + publisher = {Playtika}, + urldate = {2025-03-19}, + date = {2025-01-27}, + note = {original-date: 2021-11-28T07:08:32Z}, } -@inproceedings{schwenke_constructing_2021, - title = {Constructing {Global} {Coherence} {Representations}: {Identifying} {Interpretability} and {Coherences} of {Transformer} {Attention} in {Time} {Series} {Data}}, - shorttitle = {Constructing {Global} {Coherence} {Representations}}, - url = {https://ieeexplore.ieee.org/document/9564126/?arnumber=9564126}, - doi = {10.1109/DSAA53316.2021.9564126}, - abstract = {Transformer models have shown significant advances recently based on the general concept of Attention — to focus on specifically important and relevant parts of the input data. However, methods for enhancing their interpretability and explainability are still lacking. This is the problem which we tackle in this paper, to make Multi-Headed Attention more interpretable and explainable for time series classification. We present a method for constructing global coherence representations from Multi-Headed Attention of Transformer architectures. Accordingly, we present abstraction and interpretation methods, leading to intuitive visualizations of the respective attention patterns. We evaluate our proposed approach and the presented methods on several datasets demonstrating their efficacy.}, - urldate = {2025-02-25}, - booktitle = {2021 {IEEE} 8th {International} {Conference} on {Data} {Science} and {Advanced} {Analytics} ({DSAA})}, - author = {Schwenke, Leonid and Atzmueller, Martin}, - month = oct, - year = {2021}, - keywords = {Interpretability, Attention, Coherence, Comprehensibility, Conferences, Data science, Data visualization, Deep Learning, Explainability, Global Class Representation, Scalability, Time series analysis, Time Series Classification, Transformer, Transformers, Visualization}, - pages = {1--12}, - file = {Full Text PDF:/home/alex/Zotero/storage/VRGIDFY8/Schwenke and Atzmueller - 2021 - Constructing Global Coherence Representations Identifying Interpretability and Coherences of Transf.pdf:application/pdf;IEEE Xplore Abstract Record:/home/alex/Zotero/storage/Z45R3XWF/9564126.html:text/html}, +@online{noauthor_create_nodate, + title = {Create baseline model - {ValueError}: too many values to unpack (expected 2) · Issue \#230 · sktime/pytorch-forecasting}, + url = {https://github.com/sktime/pytorch-forecasting/issues/230}, + shorttitle = {Create baseline model - {ValueError}}, + abstract = {{PyTorch}-Forecasting version: 0.7.1 {PyTorch} version: 1.7.1 Python version: 3.7 Operating System: {MAC} {OS} Big Sur: Version 11.1 Expected behavior I executed code actuals = torch.cat([y for x, (y, weig...}, + titleaddon = {{GitHub}}, + urldate = {2025-03-19}, + langid = {english}, } -@article{schwenke_show_nodate, - title = {Show {Me} {What} {You}’re {Looking} {For}: {Visualizing} {Abstracted} {Transformer} {Attention} for {Enhancing} {Their} {Local} {Interpretability} on {Time} {Series} {Data}}, - abstract = {While Transformers have shown their advantages considering their learning performance, their lack of explainability and interpretability is still a major problem. This specifically relates to the processing of time series, as a specific form of complex data. In this paper, we propose an approach for visualizing abstracted information in order to enable computational sensemaking and local interpretability on the respective Transformer model. Our results demonstrate the efficacy of the proposed abstraction method and visualization, utilizing both synthetic and real world data for evaluation.}, - language = {en}, - author = {Schwenke, Leonid and Atzmueller, Martin}, - file = {PDF:/home/alex/Zotero/storage/SLVVAAXA/Schwenke and Atzmueller - Show Me What You’re Looking For Visualizing Abstracted Transformer Attention for Enhancing Their Lo.pdf:application/pdf}, +@collection{pfannstiel_entrepreneurship_2018, + location = {Wiesbaden}, + title = {Entrepreneurship im Gesundheitswesen {II}}, + rights = {http://www.springer.com/tdm}, + isbn = {978-3-658-14780-8 978-3-658-14781-5}, + url = {http://link.springer.com/10.1007/978-3-658-14781-5}, + publisher = {Springer Fachmedien Wiesbaden}, + editor = {Pfannstiel, Mario A. and Da-Cruz, Patrick and Rasche, Christoph}, + urldate = {2025-03-17}, + date = {2018}, + langid = {german}, + doi = {10.1007/978-3-658-14781-5}, + file = {PDF:/home/alex/Zotero/storage/DTVM5NBP/Pfannstiel et al. - 2018 - Entrepreneurship im Gesundheitswesen II.pdf:application/pdf}, } -@article{wu_interpretable_2022, - title = {Interpretable wind speed prediction with multivariate time series and temporal fusion transformers}, - volume = {252}, - issn = {03605442}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0360544222008933}, - doi = {10.1016/j.energy.2022.123990}, - abstract = {Wind power has been utilized well in power systems, so steady and successful wind speed forecasting is crucial to security management power grid market economy. To date, most researchers have often discounted the interpretability of prediction models, leading to obscure forecasts. This study puts forward a unique forecasting methodology that incorporates notable decomposition techniques, multifactor interpretable forecasting models, and optimization algorithms. In the proposed model, variational mode decomposition is employed to break down the raw wind speed sequence into a set of intrinsic mode functions. Adaptive differential evolution is then used for optimizing several parameters of temporal fusion transformers (TFT) to achieve satisfactory forecasting performance. TFT is a new attention-based deep learning model that puts together high-performance multi-horizon prediction and interpretable insights into temporal dynamics. Empirical studies using eight real-world 1-h wind speed data sets in Albert, Canada, and Five Points, USA demonstrate that the system using the proposed model outperforms those employing other comparable models in nearly all performance metrics. Examples of TFT's interpretable outputs are the importance ranking of the decomposed wind speed sub-sequences and meteorological data and attention analysis of different step lengths. The findings signify substantial progress for wind speed prediction and aid policymakers.}, - language = {en}, - urldate = {2025-02-25}, - journal = {Energy}, - author = {Wu, Binrong and Wang, Lin and Zeng, Yu-Rong}, - month = aug, - year = {2022}, - pages = {123990}, - file = {PDF:/home/alex/Zotero/storage/MHAFY926/Wu et al. - 2022 - Interpretable wind speed prediction with multivariate time series and temporal fusion transformers.pdf:application/pdf}, +@article{murray_diagnosis_2005, + title = {Diagnosis and treatment of ectopic pregnancy}, + volume = {173}, + issn = {0820-3946, 1488-2329}, + url = {http://www.cmaj.ca/cgi/doi/10.1503/cmaj.050222}, + doi = {10.1503/cmaj.050222}, + abstract = {{ECTOPIC} {PREGNANCY} {IS} A {LIFE}- {AND} {FERTILITY}-threatening condition that is commonly seen in Canadian emergency departments. Increases in the availability and use of hormonal markers, coupled with advances in formal and emergency ultrasonography have changed the diagnostic approach to the patient in the emergency department with first-trimester bleeding or pain. Ultrasonography should be the initial investigation for symptomatic women in their first trimester; when the results are indeterminate, the serum β human chorionic gonadotropin (β-{hCG}) concentration should be measured. Serial measurement of β-{hCG} and progesterone concentrations may be useful when the diagnosis remains unclear. Advances in surgical and medical therapy for ectopic pregnancy have allowed the proliferation of minimally invasive or noninvasive treatment. Guidelines for laparoscopy and for methotrexate therapy are provided.}, + pages = {905--912}, + number = {8}, + journaltitle = {Canadian Medical Association Journal}, + author = {Murray, H.}, + urldate = {2025-03-17}, + date = {2005-10-11}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/Y73KE57K/Murray - 2005 - Diagnosis and treatment of ectopic pregnancy.pdf:application/pdf}, } -@inproceedings{habib_n-beats_2025, - address = {Cham}, - title = {N-{BEATS} \& {Temporal} {Fusion} {Transformer} {Based} {Surface} {Temperature} {Prediction} and {Forecasting} for {Realizing} {Global} {Warming} {Trends}}, - isbn = {978-3-031-75167-7}, - doi = {10.1007/978-3-031-75167-7_3}, - abstract = {At the pinnacle of civilization, where the impacts of climate change have been increasingly felt, weather prediction plays a critical role in mitigating the potential disasters that may arise. Moreover, with the gradual change on climate, surface temperature of the earth is increasing. This increasing rate of the surface temperature causing global warming which is a matter of intimidation. To leave off this global warming, weather forecasting can be used as an arsenal. Selecting the appropriate tools and models for weather prediction is a crucial step in ensuring accurate forecasts. In this research paper, the focus was on studying the versatility of three specific architectures for weather prediction: LSTM, Temporal Fusion Transformer, and N-BEATS. To assess these architectures’ performance, we conducted a number of experiments. With the lowest Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) of the three, NBEATS stood out. This shows that when compared to the other models, the N-BEATS architecture had greater prediction accuracy. It's vital to remember, too, that the trials also showed that the Temporal Fusion Transformer and LSTM performed well. The only distinction was that these models required larger sizes in terms of parameters and computational complexity to achieve their performance levels. Consequently, considering both performance and model size, the researchers determined that N-BEATS was the most optimal and versatile architecture for weather prediction. Its ability to achieve excellent results with a smaller model size makes it a favorable choice for practical applications.}, - language = {en}, - booktitle = {Artificial {Intelligence} and {Speech} {Technology}}, - publisher = {Springer Nature Switzerland}, - author = {Habib, Adria Binte and Ashraf, Faisal Bin and Hossain, Muhammad Iqbal and Alam, Golam Rabiul}, - editor = {Dev, Amita and Sharma, Arun and Agrawal, S. S. and Rani, Ritu}, - year = {2025}, - keywords = {LSTM, N-BEATS, Temporal Fusion Transformer, Time Series Analysis, Weather Prediction}, - pages = {30--41}, - annote = { - -shows that n beats has better performance on smaller datasets - - -we have a relatively large dataset, thus tft will be focused on - - -}, +@article{rosenfield_adolescent_2013, + title = {Adolescent Anovulation: Maturational Mechanisms and Implications}, + volume = {98}, + issn = {0021-972X, 1945-7197}, + url = {https://academic.oup.com/jcem/article-lookup/doi/10.1210/jc.2013-1770}, + doi = {10.1210/jc.2013-1770}, + shorttitle = {Adolescent Anovulation}, + pages = {3572--3583}, + number = {9}, + journaltitle = {The Journal of Clinical Endocrinology \& Metabolism}, + author = {Rosenfield, Robert L.}, + urldate = {2025-03-17}, + date = {2013-09}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/BEF7PQT6/Rosenfield - 2013 - Adolescent Anovulation Maturational Mechanisms and Implications.pdf:application/pdf}, } -@article{papacharalampous_predictability_2018, - title = {Predictability of monthly temperature and precipitation using automatic time series forecasting methods}, - volume = {66}, - issn = {1895-7455}, - url = {https://doi.org/10.1007/s11600-018-0120-7}, - doi = {10.1007/s11600-018-0120-7}, - abstract = {We investigate the predictability of monthly temperature and precipitation by applying automatic univariate time series forecasting methods to a sample of 985 40-year-long monthly temperature and 1552 40-year-long monthly precipitation time series. The methods include a naïve one based on the monthly values of the last year, as well as the random walk (with drift), AutoRegressive Fractionally Integrated Moving Average (ARFIMA), exponential smoothing state-space model with Box–Cox transformation, ARMA errors, Trend and Seasonal components (BATS), simple exponential smoothing, Theta and Prophet methods. Prophet is a recently introduced model inspired by the nature of time series forecasted at Facebook and has not been applied to hydrometeorological time series before, while the use of random walk, BATS, simple exponential smoothing and Theta is rare in hydrology. The methods are tested in performing multi-step ahead forecasts for the last 48 months of the data. We further investigate how different choices of handling the seasonality and non-normality affect the performance of the models. The results indicate that: (a) all the examined methods apart from the naïve and random walk ones are accurate enough to be used in long-term applications; (b) monthly temperature and precipitation can be forecasted to a level of accuracy which can barely be improved using other methods; (c) the externally applied classical seasonal decomposition results mostly in better forecasts compared to the automatic seasonal decomposition used by the BATS and Prophet methods; and (d) Prophet is competitive, especially when it is combined with externally applied classical seasonal decomposition.}, - language = {en}, - number = {4}, - urldate = {2025-03-04}, - journal = {Acta Geophysica}, - author = {Papacharalampous, Georgia and Tyralis, Hristos and Koutsoyiannis, Demetris}, - month = aug, - year = {2018}, - keywords = {ARFIMA, Multi-step ahead forecasting, Precipitation forecasting, Prophet, Temperature forecasting, Time series forecasting}, - pages = {807--831}, +@book{cryer_time_2008, + location = {New York}, + edition = {2nd ed}, + title = {Time series analysis: with applications in R}, + isbn = {978-0-387-75958-6 978-0-387-75959-3}, + series = {Springer texts in statistics}, + shorttitle = {Time series analysis}, + pagetotal = {491}, + publisher = {Springer}, + author = {Cryer, Jonathan D. and Chan, Kung-sik}, + date = {2008}, + langid = {english}, + note = {{OCLC}: ocn191760003}, + keywords = {Data processing, R (Computer program language), Time-series analysis}, + file = {PDF:/home/alex/Zotero/storage/CIYMBUEW/Cryer and Chan - 2008 - Time series analysis with applications in R.pdf:application/pdf}, } -@article{dunson_day-specific_1999, - title = {Day-specific probabilities of clinical pregnancy based on two studies with imperfect measures of ovulation}, - volume = {14}, - issn = {1460-2350, 0268-1161}, - url = {https://academic.oup.com/humrep/article-lookup/doi/10.1093/humrep/14.7.1835}, - doi = {10.1093/humrep/14.7.1835}, - language = {en}, - number = {7}, - urldate = {2025-03-05}, - journal = {Human Reproduction}, - author = {Dunson, D.B. and Baird, D.D. and Wilcox, A.J. and Weinberg, C.R.}, - month = jul, - year = {1999}, - pages = {1835--1839}, - file = {PDF:/home/alex/Zotero/storage/8DE3ZLPJ/Dunson et al. - 1999 - Day-specific probabilities of clinical pregnancy based on two studies with imperfect measures of ovu.pdf:application/pdf}, +@book{hamilton_time_1994, + location = {Princeton (N.J.)}, + title = {Time series analysis}, + isbn = {978-0-691-04289-3}, + publisher = {Princeton university press}, + author = {Hamilton, James Douglas}, + date = {1994}, + file = {PDF:/home/alex/Zotero/storage/J8GVKKHG/Hamilton - 1994 - Time series analysis.pdf:application/pdf}, } -@misc{wikimedia_commons_basic_2019, - title = {Basic {Female} {Reproductive} {System}}, +@artwork{wikimedia_commons_basic_2019, + title = {Basic Female Reproductive System}, url = {https://en.wikipedia.org/wiki/File:Basic_Female_Reproductive_System_(English).svg}, author = {Wikimedia Commons}, - year = {2019}, + date = {2019}, file = {background_female_reproductive_organs:/home/alex/Zotero/storage/5MAJ4CG5/background_female_reproductive_organs.png:image/png}, } @article{silberstein_physiology_2000, - title = {Physiology of the {Menstrual} {Cycle}}, + title = {Physiology of the Menstrual Cycle}, volume = {20}, - copyright = {https://journals.sagepub.com/page/policies/text-and-data-mining-license}, + rights = {https://journals.sagepub.com/page/policies/text-and-data-mining-license}, issn = {0333-1024, 1468-2982}, url = {https://journals.sagepub.com/doi/10.1046/j.1468-2982.2000.00034.x}, doi = {10.1046/j.1468-2982.2000.00034.x}, - abstract = {The normal female life cycle is associated with a number of hormonal milestones: menarche, pregnancy, contraceptive use, menopause, and the use of replacement sex hormones. All these events and interventions alter the levels and cycling of sex hormones and may cause a change in the prevalence or intensity of headache. The menstrual cycle is the result of a carefully orchestrated sequence of interactions among the hypothalamus, pituitary, ovary, and endometrium, with the sex hormones acting as modulators and effectors at each level. Oestrogen and progestins have potent effects on central serotonergic and opioid neurons, modulating both neuronal activity and receptor density. The primary trigger of menstrual migraine appears to be the withdrawal of oestrogen rather than the maintenance of sustained high or low oestrogen levels. However, changes in the sustained oestrogen levels with pregnancy (increased) and menopause (decreased) appear to affect headaches. Headaches that occur with premenstrual syndrome appear to be centrally generated, involving the inherent rhythm of CNS neurons, including perhaps the serotonergic pain-modulating systems.}, - language = {en}, - number = {3}, - urldate = {2025-03-10}, - journal = {Cephalalgia}, - author = {Silberstein, S D and Merriam, G R}, - month = apr, - year = {2000}, + abstract = {The normal female life cycle is associated with a number of hormonal milestones: menarche, pregnancy, contraceptive use, menopause, and the use of replacement sex hormones. All these events and interventions alter the levels and cycling of sex hormones and may cause a change in the prevalence or intensity of headache. The menstrual cycle is the result of a carefully orchestrated sequence of interactions among the hypothalamus, pituitary, ovary, and endometrium, with the sex hormones acting as modulators and effectors at each level. Oestrogen and progestins have potent effects on central serotonergic and opioid neurons, modulating both neuronal activity and receptor density. The primary trigger of menstrual migraine appears to be the withdrawal of oestrogen rather than the maintenance of sustained high or low oestrogen levels. However, changes in the sustained oestrogen levels with pregnancy (increased) and menopause (decreased) appear to affect headaches. Headaches that occur with premenstrual syndrome appear to be centrally generated, involving the inherent rhythm of {CNS} neurons, including perhaps the serotonergic pain-modulating systems.}, pages = {148--154}, + number = {3}, + journaltitle = {Cephalalgia}, + author = {Silberstein, S D and Merriam, G R}, + urldate = {2025-03-10}, + date = {2000-04}, + langid = {english}, file = {Full Text:/home/alex/Zotero/storage/EL585H2P/Silberstein and Merriam - 2000 - Physiology of the Menstrual Cycle.pdf:application/pdf}, } -@misc{pedroso_menstrual_2022, - title = {The {Menstrual} {Cycle}}, +@online{pedroso_menstrual_2022, + title = {The Menstrual Cycle}, url = {https://kindbody.com/the-menstrual-cycle/}, abstract = {Fertility, gynecology, and wellness services in modern, tech-enabled clinics. Best-in-class care, accessible pricing, and a seamless patient experience.}, - urldate = {2025-03-10}, - journal = {Kindbody}, + titleaddon = {Kindbody}, author = {Pedroso, Dr Jasmine}, - month = jun, - year = {2022}, + urldate = {2025-03-10}, + date = {2022-06-03}, file = {Snapshot:/home/alex/Zotero/storage/738JEKW7/the-menstrual-cycle.html:text/html}, } @article{munster_length_1992, - title = {Length and variation in the menstrual cycle—a cross‐sectional study from a {Danish} county}, + title = {Length and variation in the menstrual cycle—a cross‐sectional study from a Danish county}, volume = {99}, issn = {1470-0328, 1471-0528}, url = {https://obgyn.onlinelibrary.wiley.com/doi/10.1111/j.1471-0528.1992.tb13762.x}, doi = {10.1111/j.1471-0528.1992.tb13762.x}, - abstract = {ABSTRACT + abstract = {{ABSTRACT} Objective To investigate the current epidemiology of menstrual patters among women of fertile age. @@ -1279,50 +1007,60 @@ we have a relatively large dataset, thus tft will be focused on Conclusion The study confirmed the normally used definitions of polymenorrhoea (cycle length {\textless}21 days) and oligomenorrhoea (cycle length between 36 and 90 days), as these very short or long menstrual cycle lengths were very seldom recorded for a longer period. However, the high frequency in a normal population of large menstrual cycle length variation challenges the view that an intra‐individual variation of {\textgreater}5 days should be regarded as a sign of disease in the woman.}, - language = {en}, - number = {5}, - urldate = {2025-03-07}, - journal = {BJOG: An International Journal of Obstetrics \& Gynaecology}, - author = {Münster, Kirstine and Schmidt, Lone and Helm, Peter}, - month = may, - year = {1992}, pages = {422--429}, + number = {5}, + journaltitle = {{BJOG}}, + author = {Münster, Kirstine and Schmidt, Lone and Helm, Peter}, + urldate = {2025-03-07}, + date = {1992-05}, + langid = {english}, file = {PDF:/home/alex/Zotero/storage/JWE75VL3/Münster et al. - 1992 - Length and variation in the menstrual cycle—a cross‐sectional study from a Danish county.pdf:application/pdf}, } @article{bull_real-world_2019, title = {Real-world menstrual cycle characteristics of more than 600,000 menstrual cycles}, volume = {2}, - copyright = {2019 The Author(s)}, + rights = {2019 The Author(s)}, issn = {2398-6352}, url = {https://www.nature.com/articles/s41746-019-0152-7}, doi = {10.1038/s41746-019-0152-7}, - abstract = {The use of apps that record detailed menstrual cycle data presents a new opportunity to study the menstrual cycle. The aim of this study is to describe menstrual cycle characteristics observed from a large database of cycles collected through an app and investigate associations of menstrual cycle characteristics with cycle length, age and body mass index (BMI). Menstrual cycle parameters, including menstruation, basal body temperature (BBT) and luteinising hormone (LH) tests as well as age and BMI were collected anonymously from real-world users of the Natural Cycles app. We analysed 612,613 ovulatory cycles with a mean length of 29.3 days from 124,648 users. The mean follicular phase length was 16.9 days (95\% CI: 10–30) and mean luteal phase length was 12.4 days (95\% CI: 7–17). Mean cycle length decreased by 0.18 days (95\% CI: 0.17–0.18, R2 = 0.99) and mean follicular phase length decreased by 0.19 days (95\% CI: 0.19–0.20, R2 = 0.99) per year of age from 25 to 45 years. Mean variation of cycle length per woman was 0.4 days or 14\% higher in women with a BMI of over 35 relative to women with a BMI of 18.5–25. This analysis details variations in menstrual cycle characteristics that are not widely known yet have significant implications for health and well-being. Clinically, women who wish to plan a pregnancy need to have intercourse on their fertile days. In order to identify the fertile period it is important to track physiological parameters such as basal body temperature and not just cycle length.}, - language = {en}, + abstract = {The use of apps that record detailed menstrual cycle data presents a new opportunity to study the menstrual cycle. The aim of this study is to describe menstrual cycle characteristics observed from a large database of cycles collected through an app and investigate associations of menstrual cycle characteristics with cycle length, age and body mass index ({BMI}). Menstrual cycle parameters, including menstruation, basal body temperature ({BBT}) and luteinising hormone ({LH}) tests as well as age and {BMI} were collected anonymously from real-world users of the Natural Cycles app. We analysed 612,613 ovulatory cycles with a mean length of 29.3 days from 124,648 users. The mean follicular phase length was 16.9 days (95\% {CI}: 10–30) and mean luteal phase length was 12.4 days (95\% {CI}: 7–17). Mean cycle length decreased by 0.18 days (95\% {CI}: 0.17–0.18, R2 = 0.99) and mean follicular phase length decreased by 0.19 days (95\% {CI}: 0.19–0.20, R2 = 0.99) per year of age from 25 to 45 years. Mean variation of cycle length per woman was 0.4 days or 14\% higher in women with a {BMI} of over 35 relative to women with a {BMI} of 18.5–25. This analysis details variations in menstrual cycle characteristics that are not widely known yet have significant implications for health and well-being. Clinically, women who wish to plan a pregnancy need to have intercourse on their fertile days. In order to identify the fertile period it is important to track physiological parameters such as basal body temperature and not just cycle length.}, + pages = {1--8}, number = {1}, - urldate = {2025-03-07}, - journal = {npj Digital Medicine}, + journaltitle = {npj Digit. Med.}, author = {Bull, Jonathan R. and Rowland, Simon P. and Scherwitzl, Elina Berglund and Scherwitzl, Raoul and Danielsson, Kristina Gemzell and Harper, Joyce}, - month = aug, - year = {2019}, + urldate = {2025-03-07}, + date = {2019-08-27}, + langid = {english}, note = {Publisher: Nature Publishing Group}, keywords = {Preclinical research, Reproductive biology}, - pages = {1--8}, file = {Full Text PDF:/home/alex/Zotero/storage/MUPKFK2K/Bull et al. - 2019 - Real-world menstrual cycle characteristics of more than 600,000 menstrual cycles.pdf:application/pdf}, } +@article{pratikno_pdf_2024, + title = {({PDF}) A novel women's ovulation prediction through salivary ferning using the box counting and deep learning}, + url = {https://www.researchgate.net/publication/379467622_A_novel_women's_ovulation_prediction_through_salivary_ferning_using_the_box_counting_and_deep_learning}, + doi = {10.11591/eei.v13i2.5847}, + abstract = {{PDF} {\textbar} There are several methods to predict a woman's ovulation time, including using a calendar system, basal body temperature, ovulation prediction... {\textbar} Find, read and cite all the research you need on {ResearchGate}}, + journaltitle = {{ResearchGate}}, + author = {Pratikno and Ibrahim and Jusak}, + urldate = {2025-02-11}, + date = {2024-12-09}, + langid = {english}, + file = {Full Text:/home/alex/Zotero/storage/MVSBPZQG/2024 - (PDF) A novel women's ovulation prediction through salivary ferning using the box counting and deep.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/8FYLC8YZ/379467622_A_novel_women's_ovulation_prediction_through_salivary_ferning_using_the_box_counting_.html:text/html}, +} + @incollection{holesh_physiology_2025, - address = {Treasure Island (FL)}, - title = {Physiology, {Ovulation}}, - copyright = {Copyright © 2025, StatPearls Publishing LLC.}, + location = {Treasure Island ({FL})}, + title = {Physiology, Ovulation}, + rights = {Copyright © 2025, {StatPearls} Publishing {LLC}.}, url = {http://www.ncbi.nlm.nih.gov/books/NBK441996/}, - abstract = {Ovulation is a physiologic process defined by the rupture of the dominant follicle of the ovary. This releases an egg into the abdominal cavity. It then is taken up by the fimbriae of the fallopian tube where it has the potential to become fertilized. The ovulation process is regulated by fluxing gonadotropic hormone (FSH/LH) levels. Ovulation is the third phase within the larger uterine cycle (ie, menstrual cycle). The follicular release follows the Follicular phase (ie, dominant follicle development) and precedes the luteal phase (ie, maintenance of corpus luteum) that progresses to either endometrial shedding or implantation. Follicular release occurs around 14 days prior to menstruation in a cyclic pattern if the hypothalamic-pituitary-ovarian axis function is well regulated.  Structure Genotypic females (XX) develop two ovaries that sit adjacent to the uterine horns. Each ovary is anchored to the uterus at the medial pole by the utero-ovarian ligament. The lateral ovarian pole is anchored to the pelvic sidewall by the infundibulopelvic ligament (i.e,. suspensory ligament of the ovary), which carries the ovarian artery and vein. Each ovary contains 1 to 2 million primordial follicles that each contain primary oocytes (ie, eggs) that can supply that female with enough follicles until she reaches her fourth or fifth decades of life. These primordial follicles are arrested in prophase I of meiosis until the onset of puberty. At the onset of pubescence, the gonadotropic hormones began to induce the maturation of the primordial follicle, allowing for the completion of meiosis I, forming a secondary follicle. The secondary follicle begins meiosis II, but this phase will not be completed unless that follicle is fertilized. With each ovulatory cycle, the number of follicles decreases, eventually leading to the onset of Menopause or the cessation of ovulatory function. Per each ovulation cycle, the average ovary loses 1,000 follicles to the process of selecting a dominant follicle that will be released. This process accelerates in an age-dependent manner as well. It is also a common thought that the right and left ovaries alternate follicular releases each month. Ovulation is regulated by the fluctuation between the following hormones. Tight regulation and controlled changes between the following hormones are imperative for the development and release of an oocyte into the adnexal uterine structures.   Hormones involved in ovulation include: Gonadotropin-releasing hormone (GnRH) is a tropic peptide hormone made and secreted by the hypothalamus. It is a releasing hormone that stimulates the release of FSH and LH from the anterior pituitary gland through variations in GnRH pulse frequency. Low-frequency GnRH pulses are responsible for FSH secretion, whereas high-frequency pulses are responsible for LH secretion. During the Follicular phase of the Uterine cycle, estrogen secretion causes the Granulosa cells to autonomously increase their own production of estrogen, contributing to elevation in estrogen serum levels. This elevation is communicated to the hypothalamus and contributes to the increase in GnRH pulse frequency, eventually stimulating the LH surge that eventually induces the follicular rupture and release from the corpus luteum and luteinization of the granulosa cells, enabling the synthesis of progesterone in place of estrogen. Finally, the low levels of LH following the surge restart the FSH production by the slow-pulsation frequency of GnRH release. . Gonadotropin hormones are heterodimeric glycoproteins with alpha/beta subunits. The alpha subunit is common to all glycoproteins, including TSH (thyroid-stimulating hormone) and HCG (human chorionic gonadotropin hormone).  The relationship between FSH and LH hormones is responsible for the process that induces follicular development, rupture, release, and endometrial reception or shedding. Disruption in the hormonal communication between the gonadotropin-releasing hormones, gonadotropic hormones, and their receptors can lead to anovulation or amenorrhea, leading to various pathologic sequelae as a consequence. Follicle-Stimulating Hormone (FSH) is a gonadotropin synthesized and secreted from the anterior pituitary gland in response to slow-frequency pulsatile GnRH. FSH stimulates the growth and maturation of immature oocytes into mature (Graafian) secondary follicles before ovulation. FSH Receptors are G-protein coupled receptors and are found in the Granulosa cells that surround developing ovarian follicles. The granulosa cells initially produce the estrogen needed to maturate the developing dominant follicle. After 2 days of sustained elevation of estrogen levels, the LH surge causes luteinization of the granulosa cells into LH receptive cells. This transition enables granulosa cells to respond to LH levels and produce progesterone. : Estrogen is a steroid hormone that is responsible for the growth and regulation of the female reproductive system and secondary sex characteristics. Estrogen is produced by the granulosa cells of the developing follicle and exerts negative feedback on LH production in the early part of the menstrual cycle. However, once estrogen levels reach a critical level as oocytes mature within the ovary in preparation for ovulation, estrogen begins to exert positive feedback on LH production, leading to the LH surge through its effects on GnRH pulse frequency. Estrogen also has many other effects that are important for bone health and cardiovascular health in premenopausal patients, which will be discussed in another article. Luteinizing Hormone (LH) is a gonadotropin synthesized and secreted by the anterior pituitary gland in response to high-frequency GnRH release. LH is responsible for inducing ovulation, preparation for fertilized oocyte uterine implantation, and the ovarian production of progesterone through stimulation of theca cells and luteinized granulosa cells. Prior to the LH surge, LH interacts with Theca cells that are adjacent to granulosa cells in the ovary. These cells produce androgens, which diffuse into the granulosa cells and convert to estrogen for follicular development. The LH surge creates the environment for follicular eruption by increasing the activity of the proteolytic enzymes that weaken the ovarian wall, allowing for the passage of the oocyte. After the oocyte is released, the follicular remnants are theca and luteinized granulosa cells. Their function is now to produce progesterone, which is the hormone responsible for maintaining the uterine environment that can accept a fertilized embryo. Progesterone is a steroid hormone that is responsible for preparing the endometrium for the uterine implantation of the fertilized egg and maintenance of pregnancy. If a fertilized egg implants, the corpus luteum secretes progesterone in early pregnancy until the placenta develops and takes over progesterone production for the remainder of the pregnancy.}, - language = {eng}, - urldate = {2025-03-07}, + abstract = {Ovulation is a physiologic process defined by the rupture of the dominant follicle of the ovary. This releases an egg into the abdominal cavity. It then is taken up by the fimbriae of the fallopian tube where it has the potential to become fertilized. The ovulation process is regulated by fluxing gonadotropic hormone ({FSH}/{LH}) levels. Ovulation is the third phase within the larger uterine cycle (ie, menstrual cycle). The follicular release follows the Follicular phase (ie, dominant follicle development) and precedes the luteal phase (ie, maintenance of corpus luteum) that progresses to either endometrial shedding or implantation. Follicular release occurs around 14 days prior to menstruation in a cyclic pattern if the hypothalamic-pituitary-ovarian axis function is well regulated.  Structure Genotypic females ({XX}) develop two ovaries that sit adjacent to the uterine horns. Each ovary is anchored to the uterus at the medial pole by the utero-ovarian ligament. The lateral ovarian pole is anchored to the pelvic sidewall by the infundibulopelvic ligament (i.e,. suspensory ligament of the ovary), which carries the ovarian artery and vein. Each ovary contains 1 to 2 million primordial follicles that each contain primary oocytes (ie, eggs) that can supply that female with enough follicles until she reaches her fourth or fifth decades of life. These primordial follicles are arrested in prophase I of meiosis until the onset of puberty. At the onset of pubescence, the gonadotropic hormones began to induce the maturation of the primordial follicle, allowing for the completion of meiosis I, forming a secondary follicle. The secondary follicle begins meiosis {II}, but this phase will not be completed unless that follicle is fertilized. With each ovulatory cycle, the number of follicles decreases, eventually leading to the onset of Menopause or the cessation of ovulatory function. Per each ovulation cycle, the average ovary loses 1,000 follicles to the process of selecting a dominant follicle that will be released. This process accelerates in an age-dependent manner as well. It is also a common thought that the right and left ovaries alternate follicular releases each month. Ovulation is regulated by the fluctuation between the following hormones. Tight regulation and controlled changes between the following hormones are imperative for the development and release of an oocyte into the adnexal uterine structures.   Hormones involved in ovulation include: Gonadotropin-releasing hormone ({GnRH}) is a tropic peptide hormone made and secreted by the hypothalamus. It is a releasing hormone that stimulates the release of {FSH} and {LH} from the anterior pituitary gland through variations in {GnRH} pulse frequency. Low-frequency {GnRH} pulses are responsible for {FSH} secretion, whereas high-frequency pulses are responsible for {LH} secretion. During the Follicular phase of the Uterine cycle, estrogen secretion causes the Granulosa cells to autonomously increase their own production of estrogen, contributing to elevation in estrogen serum levels. This elevation is communicated to the hypothalamus and contributes to the increase in {GnRH} pulse frequency, eventually stimulating the {LH} surge that eventually induces the follicular rupture and release from the corpus luteum and luteinization of the granulosa cells, enabling the synthesis of progesterone in place of estrogen. Finally, the low levels of {LH} following the surge restart the {FSH} production by the slow-pulsation frequency of {GnRH} release. . Gonadotropin hormones are heterodimeric glycoproteins with alpha/beta subunits. The alpha subunit is common to all glycoproteins, including {TSH} (thyroid-stimulating hormone) and {HCG} (human chorionic gonadotropin hormone).  The relationship between {FSH} and {LH} hormones is responsible for the process that induces follicular development, rupture, release, and endometrial reception or shedding. Disruption in the hormonal communication between the gonadotropin-releasing hormones, gonadotropic hormones, and their receptors can lead to anovulation or amenorrhea, leading to various pathologic sequelae as a consequence. Follicle-Stimulating Hormone ({FSH}) is a gonadotropin synthesized and secreted from the anterior pituitary gland in response to slow-frequency pulsatile {GnRH}. {FSH} stimulates the growth and maturation of immature oocytes into mature (Graafian) secondary follicles before ovulation. {FSH} Receptors are G-protein coupled receptors and are found in the Granulosa cells that surround developing ovarian follicles. The granulosa cells initially produce the estrogen needed to maturate the developing dominant follicle. After 2 days of sustained elevation of estrogen levels, the {LH} surge causes luteinization of the granulosa cells into {LH} receptive cells. This transition enables granulosa cells to respond to {LH} levels and produce progesterone. : Estrogen is a steroid hormone that is responsible for the growth and regulation of the female reproductive system and secondary sex characteristics. Estrogen is produced by the granulosa cells of the developing follicle and exerts negative feedback on {LH} production in the early part of the menstrual cycle. However, once estrogen levels reach a critical level as oocytes mature within the ovary in preparation for ovulation, estrogen begins to exert positive feedback on {LH} production, leading to the {LH} surge through its effects on {GnRH} pulse frequency. Estrogen also has many other effects that are important for bone health and cardiovascular health in premenopausal patients, which will be discussed in another article. Luteinizing Hormone ({LH}) is a gonadotropin synthesized and secreted by the anterior pituitary gland in response to high-frequency {GnRH} release. {LH} is responsible for inducing ovulation, preparation for fertilized oocyte uterine implantation, and the ovarian production of progesterone through stimulation of theca cells and luteinized granulosa cells. Prior to the {LH} surge, {LH} interacts with Theca cells that are adjacent to granulosa cells in the ovary. These cells produce androgens, which diffuse into the granulosa cells and convert to estrogen for follicular development. The {LH} surge creates the environment for follicular eruption by increasing the activity of the proteolytic enzymes that weaken the ovarian wall, allowing for the passage of the oocyte. After the oocyte is released, the follicular remnants are theca and luteinized granulosa cells. Their function is now to produce progesterone, which is the hormone responsible for maintaining the uterine environment that can accept a fertilized embryo. Progesterone is a steroid hormone that is responsible for preparing the endometrium for the uterine implantation of the fertilized egg and maintenance of pregnancy. If a fertilized egg implants, the corpus luteum secretes progesterone in early pregnancy until the placenta develops and takes over progesterone production for the remainder of the pregnancy.}, booktitle = {{StatPearls}}, - publisher = {StatPearls Publishing}, + publisher = {{StatPearls} Publishing}, author = {Holesh, Julie E. and Bass, Autumn N. and Lord, Megan}, - year = {2025}, + urldate = {2025-03-07}, + date = {2025}, pmid = {28723025}, file = {Printable HTML:/home/alex/Zotero/storage/D6LTGP64/NBK441996.html:text/html}, } @@ -1333,1031 +1071,1153 @@ we have a relatively large dataset, thus tft will be focused on issn = {00150282}, url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028215002034}, doi = {10.1016/j.fertnstert.2015.03.004}, - language = {en}, - number = {5}, - urldate = {2025-03-07}, - journal = {Fertility and Sterility}, - author = {Sauer, Mark V.}, - month = may, - year = {2015}, pages = {1136--1143}, + number = {5}, + journaltitle = {Fertility and Sterility}, + author = {Sauer, Mark V.}, + urldate = {2025-03-07}, + date = {2015-05}, + langid = {english}, file = {PDF:/home/alex/Zotero/storage/W6LCEFP6/Sauer - 2015 - Reproduction at an advanced maternal age and maternal health.pdf:application/pdf}, } -@techreport{mckinsey_health_institute_blueprint_nodate, - title = {Blueprint to {Close} the {Women}’s {Health} {Gap}: {How} to {Improve} {Lives} and {Economies} for {All}}, - shorttitle = {Blueprint to {Close} the {Women}’s {Health} {Gap}}, +@report{mckinsey_health_institute_blueprint_nodate, + title = {Blueprint to Close the Women’s Health Gap: How to Improve Lives and Economies for All}, + shorttitle = {Blueprint to Close the Women’s Health Gap}, + institution = {{McKinsey} Health Institute, World Economic Forum}, + author = {{McKinsey} Health Institute and World Economic Forum}, urldate = {2025-03-07}, - institution = {McKinsey Health Institute, World Economic Forum}, - author = {McKinsey Health Institute and World Economic Forum}, file = {PDF:/home/alex/Zotero/storage/NF6LAC2S/WEF_Blueprint_to_Close_the_Women’s_Health_Gap_2025.pdf:application/pdf}, } -@techreport{mckinsey_health_institute_closing_2024, - title = {Closing the {Women}’s {Health} {Gap}: {A} \$1 {Trillion} {Opportunity} to {Improve} {Lives} and {Economies}}, - shorttitle = {Closing the {Women}’s {Health} {Gap}}, +@report{mckinsey_health_institute_closing_2024, + title = {Closing the Women’s Health Gap: A \$1 Trillion Opportunity to Improve Lives and Economies}, + shorttitle = {Closing the Women’s Health Gap}, + institution = {{McKinsey} Health Institute, World Economic Forum}, + author = {{McKinsey} Health Institute and World Economic Forum}, urldate = {2025-03-07}, - institution = {McKinsey Health Institute, World Economic Forum}, - author = {McKinsey Health Institute and World Economic Forum}, - month = jan, - year = {2024}, + date = {2024-01}, file = {PDF:/home/alex/Zotero/storage/US6TWL7D/closing-the-womens-health-gap-report.pdf:application/pdf}, } @misc{global_burden_of_disease_collaborative_network_global_2020, - title = {Global {Burden} of {Disease} {Study} 2019 ({GBD} 2019) {Disability} {Weights}}, + title = {Global Burden of Disease Study 2019 ({GBD} 2019) Disability Weights}, url = {http://ghdx.healthdata.org/record/ihme-data/gbd-2019-disability-weights}, doi = {10.6069/1W19-VX76}, - abstract = {"The Global Burden of Disease Study 2019 (GBD 2019), coordinated by the Institute for Health Metrics and Evaluation (IHME), estimated the burden of diseases, injuries, and risk factors for 204 countries and territories and selected subnational locations. + abstract = {"The Global Burden of Disease Study 2019 ({GBD} 2019), coordinated by the Institute for Health Metrics and Evaluation ({IHME}), estimated the burden of diseases, injuries, and risk factors for 204 countries and territories and selected subnational locations. -Disability weights, which represent the magnitude of health loss associated with specific health outcomes, are used to calculate years lived with disability (YLD) for these outcomes in a given population. The weights are measured on a scale from 0 to 1, where 0 equals a state of full health and 1 equals death. This table provides disability weights for the 440 health states (including combined health states) used to estimate nonfatal health outcomes for the GBD 2019 study. +Disability weights, which represent the magnitude of health loss associated with specific health outcomes, are used to calculate years lived with disability ({YLD}) for these outcomes in a given population. The weights are measured on a scale from 0 to 1, where 0 equals a state of full health and 1 equals death. This table provides disability weights for the 440 health states (including combined health states) used to estimate nonfatal health outcomes for the {GBD} 2019 study. -For additional GBD results and resources, visit the GBD 2019 Data Resources page."}, - urldate = {2025-03-07}, - publisher = {Institute for Health Metrics and Evaluation (IHME)}, +For additional {GBD} results and resources, visit the {GBD} 2019 Data Resources page."}, + publisher = {Institute for Health Metrics and Evaluation ({IHME})}, author = {{Global Burden of Disease Collaborative Network}}, - year = {2020}, + urldate = {2025-03-07}, + date = {2020}, } @book{noauthor_research_2021, - title = {Research {Funding} for {Women}'s {Health}: {A} {Modeling} {Study} of {Societal} {Impact}: {Findings} for {Alzheimer}'s {Disease} and {Alzheimer}'s {Disease} {Related} {Dementia} {Model}}, - shorttitle = {Research {Funding} for {Women}'s {Health}}, + title = {Research Funding for Women's Health: A Modeling Study of Societal Impact: Findings for Alzheimer's Disease and Alzheimer's Disease Related Dementia Model}, url = {https://www.rand.org/pubs/working_papers/WRA708-1.html}, - language = {en}, + shorttitle = {Research Funding for Women's Health}, + publisher = {{RAND} Corporation}, urldate = {2025-03-07}, - publisher = {RAND Corporation}, - year = {2021}, + date = {2021}, + langid = {english}, doi = {10.7249/WRA708-1}, file = {PDF:/home/alex/Zotero/storage/93NC3JH3/2021 - Research Funding for Women's Health A Modeling Study of Societal Impact Findings for Alzheimer's D.pdf:application/pdf}, } -@misc{noauthor_create_nodate, - title = {Create baseline model - {ValueError}: too many values to unpack (expected 2) · {Issue} \#230 · sktime/pytorch-forecasting}, - shorttitle = {Create baseline model - {ValueError}}, - url = {https://github.com/sktime/pytorch-forecasting/issues/230}, - abstract = {PyTorch-Forecasting version: 0.7.1 PyTorch version: 1.7.1 Python version: 3.7 Operating System: MAC OS Big Sur: Version 11.1 Expected behavior I executed code actuals = torch.cat([y for x, (y, weig...}, - language = {en}, - urldate = {2025-03-19}, - journal = {GitHub}, +@article{dunson_day-specific_1999, + title = {Day-specific probabilities of clinical pregnancy based on two studies with imperfect measures of ovulation}, + volume = {14}, + issn = {1460-2350, 0268-1161}, + url = {https://academic.oup.com/humrep/article-lookup/doi/10.1093/humrep/14.7.1835}, + doi = {10.1093/humrep/14.7.1835}, + pages = {1835--1839}, + number = {7}, + journaltitle = {Human Reproduction}, + author = {Dunson, D.B. and Baird, D.D. and Wilcox, A.J. and Weinberg, C.R.}, + urldate = {2025-03-05}, + date = {1999-07}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/8DE3ZLPJ/Dunson et al. - 1999 - Day-specific probabilities of clinical pregnancy based on two studies with imperfect measures of ovu.pdf:application/pdf}, } -@book{pfannstiel_entrepreneurship_2018, - address = {Wiesbaden}, - title = {Entrepreneurship im {Gesundheitswesen} {II}}, - copyright = {http://www.springer.com/tdm}, - isbn = {978-3-658-14780-8 978-3-658-14781-5}, - url = {http://link.springer.com/10.1007/978-3-658-14781-5}, - language = {de}, - urldate = {2025-03-17}, - publisher = {Springer Fachmedien Wiesbaden}, - editor = {Pfannstiel, Mario A. and Da-Cruz, Patrick and Rasche, Christoph}, - year = {2018}, - doi = {10.1007/978-3-658-14781-5}, - file = {PDF:/home/alex/Zotero/storage/DTVM5NBP/Pfannstiel et al. - 2018 - Entrepreneurship im Gesundheitswesen II.pdf:application/pdf}, +@article{papacharalampous_predictability_2018, + title = {Predictability of monthly temperature and precipitation using automatic time series forecasting methods}, + volume = {66}, + issn = {1895-7455}, + url = {https://doi.org/10.1007/s11600-018-0120-7}, + doi = {10.1007/s11600-018-0120-7}, + abstract = {We investigate the predictability of monthly temperature and precipitation by applying automatic univariate time series forecasting methods to a sample of 985 40-year-long monthly temperature and 1552 40-year-long monthly precipitation time series. The methods include a naïve one based on the monthly values of the last year, as well as the random walk (with drift), {AutoRegressive} Fractionally Integrated Moving Average ({ARFIMA}), exponential smoothing state-space model with Box–Cox transformation, {ARMA} errors, Trend and Seasonal components ({BATS}), simple exponential smoothing, Theta and Prophet methods. Prophet is a recently introduced model inspired by the nature of time series forecasted at Facebook and has not been applied to hydrometeorological time series before, while the use of random walk, {BATS}, simple exponential smoothing and Theta is rare in hydrology. The methods are tested in performing multi-step ahead forecasts for the last 48 months of the data. We further investigate how different choices of handling the seasonality and non-normality affect the performance of the models. The results indicate that: (a) all the examined methods apart from the naïve and random walk ones are accurate enough to be used in long-term applications; (b) monthly temperature and precipitation can be forecasted to a level of accuracy which can barely be improved using other methods; (c) the externally applied classical seasonal decomposition results mostly in better forecasts compared to the automatic seasonal decomposition used by the {BATS} and Prophet methods; and (d) Prophet is competitive, especially when it is combined with externally applied classical seasonal decomposition.}, + pages = {807--831}, + number = {4}, + journaltitle = {Acta Geophys.}, + author = {Papacharalampous, Georgia and Tyralis, Hristos and Koutsoyiannis, Demetris}, + urldate = {2025-03-04}, + date = {2018-08-01}, + langid = {english}, + keywords = {Time series forecasting, {ARFIMA}, Multi-step ahead forecasting, Precipitation forecasting, Prophet, Temperature forecasting}, } -@article{murray_diagnosis_2005, - title = {Diagnosis and treatment of ectopic pregnancy}, - volume = {173}, - issn = {0820-3946, 1488-2329}, - url = {http://www.cmaj.ca/cgi/doi/10.1503/cmaj.050222}, - doi = {10.1503/cmaj.050222}, - abstract = {ECTOPIC PREGNANCY IS A LIFE- AND FERTILITY-threatening condition that is commonly seen in Canadian emergency departments. Increases in the availability and use of hormonal markers, coupled with advances in formal and emergency ultrasonography have changed the diagnostic approach to the patient in the emergency department with first-trimester bleeding or pain. Ultrasonography should be the initial investigation for symptomatic women in their first trimester; when the results are indeterminate, the serum β human chorionic gonadotropin (β-hCG) concentration should be measured. Serial measurement of β-hCG and progesterone concentrations may be useful when the diagnosis remains unclear. Advances in surgical and medical therapy for ectopic pregnancy have allowed the proliferation of minimally invasive or noninvasive treatment. Guidelines for laparoscopy and for methotrexate therapy are provided.}, - language = {en}, - number = {8}, - urldate = {2025-03-17}, - journal = {Canadian Medical Association Journal}, - author = {Murray, H.}, - month = oct, - year = {2005}, - pages = {905--912}, - file = {PDF:/home/alex/Zotero/storage/Y73KE57K/Murray - 2005 - Diagnosis and treatment of ectopic pregnancy.pdf:application/pdf}, +@inproceedings{habib_n-beats_2025, + location = {Cham}, + title = {N-{BEATS} \& Temporal Fusion Transformer Based Surface Temperature Prediction and Forecasting for Realizing Global Warming Trends}, + isbn = {978-3-031-75167-7}, + doi = {10.1007/978-3-031-75167-7_3}, + abstract = {At the pinnacle of civilization, where the impacts of climate change have been increasingly felt, weather prediction plays a critical role in mitigating the potential disasters that may arise. Moreover, with the gradual change on climate, surface temperature of the earth is increasing. This increasing rate of the surface temperature causing global warming which is a matter of intimidation. To leave off this global warming, weather forecasting can be used as an arsenal. Selecting the appropriate tools and models for weather prediction is a crucial step in ensuring accurate forecasts. In this research paper, the focus was on studying the versatility of three specific architectures for weather prediction: {LSTM}, Temporal Fusion Transformer, and N-{BEATS}. To assess these architectures’ performance, we conducted a number of experiments. With the lowest Mean Absolute Error ({MAE}) and Root Mean Square Error ({RMSE}) of the three, {NBEATS} stood out. This shows that when compared to the other models, the N-{BEATS} architecture had greater prediction accuracy. It's vital to remember, too, that the trials also showed that the Temporal Fusion Transformer and {LSTM} performed well. The only distinction was that these models required larger sizes in terms of parameters and computational complexity to achieve their performance levels. Consequently, considering both performance and model size, the researchers determined that N-{BEATS} was the most optimal and versatile architecture for weather prediction. Its ability to achieve excellent results with a smaller model size makes it a favorable choice for practical applications.}, + pages = {30--41}, + booktitle = {Artificial Intelligence and Speech Technology}, + publisher = {Springer Nature Switzerland}, + author = {Habib, Adria Binte and Ashraf, Faisal Bin and Hossain, Muhammad Iqbal and Alam, Golam Rabiul}, + editor = {Dev, Amita and Sharma, Arun and Agrawal, S. S. and Rani, Ritu}, + date = {2025}, + langid = {english}, + keywords = {{LSTM}, N-{BEATS}, Temporal Fusion Transformer, Time Series Analysis, Weather Prediction}, } -@article{rosenfield_adolescent_2013, - title = {Adolescent {Anovulation}: {Maturational} {Mechanisms} and {Implications}}, - volume = {98}, - issn = {0021-972X, 1945-7197}, - shorttitle = {Adolescent {Anovulation}}, - url = {https://academic.oup.com/jcem/article-lookup/doi/10.1210/jc.2013-1770}, - doi = {10.1210/jc.2013-1770}, - language = {en}, - number = {9}, - urldate = {2025-03-17}, - journal = {The Journal of Clinical Endocrinology \& Metabolism}, - author = {Rosenfield, Robert L.}, - month = sep, - year = {2013}, - pages = {3572--3583}, - file = {PDF:/home/alex/Zotero/storage/BEF7PQT6/Rosenfield - 2013 - Adolescent Anovulation Maturational Mechanisms and Implications.pdf:application/pdf}, +@article{wu_interpretable_2022, + title = {Interpretable wind speed prediction with multivariate time series and temporal fusion transformers}, + volume = {252}, + issn = {03605442}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0360544222008933}, + doi = {10.1016/j.energy.2022.123990}, + abstract = {Wind power has been utilized well in power systems, so steady and successful wind speed forecasting is crucial to security management power grid market economy. To date, most researchers have often discounted the interpretability of prediction models, leading to obscure forecasts. This study puts forward a unique forecasting methodology that incorporates notable decomposition techniques, multifactor interpretable forecasting models, and optimization algorithms. In the proposed model, variational mode decomposition is employed to break down the raw wind speed sequence into a set of intrinsic mode functions. Adaptive differential evolution is then used for optimizing several parameters of temporal fusion transformers ({TFT}) to achieve satisfactory forecasting performance. {TFT} is a new attention-based deep learning model that puts together high-performance multi-horizon prediction and interpretable insights into temporal dynamics. Empirical studies using eight real-world 1-h wind speed data sets in Albert, Canada, and Five Points, {USA} demonstrate that the system using the proposed model outperforms those employing other comparable models in nearly all performance metrics. Examples of {TFT}'s interpretable outputs are the importance ranking of the decomposed wind speed sub-sequences and meteorological data and attention analysis of different step lengths. The findings signify substantial progress for wind speed prediction and aid policymakers.}, + pages = {123990}, + journaltitle = {Energy}, + author = {Wu, Binrong and Wang, Lin and Zeng, Yu-Rong}, + urldate = {2025-02-25}, + date = {2022-08}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/MHAFY926/Wu et al. - 2022 - Interpretable wind speed prediction with multivariate time series and temporal fusion transformers.pdf:application/pdf}, } -@book{cryer_time_2008, - address = {New York}, - edition = {2nd ed}, - series = {Springer texts in statistics}, - title = {Time series analysis: with applications in {R}}, - isbn = {978-0-387-75958-6 978-0-387-75959-3}, - shorttitle = {Time series analysis}, - language = {en}, - publisher = {Springer}, - author = {Cryer, Jonathan D. and Chan, Kung-sik}, - year = {2008}, - note = {OCLC: ocn191760003}, - keywords = {Data processing, R (Computer program language), Time-series analysis}, - file = {PDF:/home/alex/Zotero/storage/CIYMBUEW/Cryer and Chan - 2008 - Time series analysis with applications in R.pdf:application/pdf}, +@article{schwenke_show_nodate, + title = {Show Me What You’re Looking For: Visualizing Abstracted Transformer Attention for Enhancing Their Local Interpretability on Time Series Data}, + abstract = {While Transformers have shown their advantages considering their learning performance, their lack of explainability and interpretability is still a major problem. This specifically relates to the processing of time series, as a specific form of complex data. In this paper, we propose an approach for visualizing abstracted information in order to enable computational sensemaking and local interpretability on the respective Transformer model. Our results demonstrate the efficacy of the proposed abstraction method and visualization, utilizing both synthetic and real world data for evaluation.}, + author = {Schwenke, Leonid and Atzmueller, Martin}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/SLVVAAXA/Schwenke and Atzmueller - Show Me What You’re Looking For Visualizing Abstracted Transformer Attention for Enhancing Their Lo.pdf:application/pdf}, } -@book{hamilton_time_1994, - address = {Princeton (N.J.)}, - title = {Time series analysis}, - isbn = {978-0-691-04289-3}, - language = {eng}, - publisher = {Princeton university press}, - author = {Hamilton, James Douglas}, - year = {1994}, - file = {PDF:/home/alex/Zotero/storage/J8GVKKHG/Hamilton - 1994 - Time series analysis.pdf:application/pdf}, +@inproceedings{schwenke_constructing_2021, + title = {Constructing Global Coherence Representations: Identifying Interpretability and Coherences of Transformer Attention in Time Series Data}, + url = {https://ieeexplore.ieee.org/document/9564126/?arnumber=9564126}, + doi = {10.1109/DSAA53316.2021.9564126}, + shorttitle = {Constructing Global Coherence Representations}, + abstract = {Transformer models have shown significant advances recently based on the general concept of Attention — to focus on specifically important and relevant parts of the input data. However, methods for enhancing their interpretability and explainability are still lacking. This is the problem which we tackle in this paper, to make Multi-Headed Attention more interpretable and explainable for time series classification. We present a method for constructing global coherence representations from Multi-Headed Attention of Transformer architectures. Accordingly, we present abstraction and interpretation methods, leading to intuitive visualizations of the respective attention patterns. We evaluate our proposed approach and the presented methods on several datasets demonstrating their efficacy.}, + eventtitle = {2021 {IEEE} 8th International Conference on Data Science and Advanced Analytics ({DSAA})}, + pages = {1--12}, + booktitle = {2021 {IEEE} 8th International Conference on Data Science and Advanced Analytics ({DSAA})}, + author = {Schwenke, Leonid and Atzmueller, Martin}, + urldate = {2025-02-25}, + date = {2021-10}, + keywords = {Time series analysis, Attention, Coherence, Comprehensibility, Conferences, Data science, Data visualization, Deep Learning, Explainability, Global Class Representation, Interpretability, Scalability, Time Series Classification, Transformer, Transformers, Visualization}, + file = {Full Text PDF:/home/alex/Zotero/storage/VRGIDFY8/Schwenke and Atzmueller - 2021 - Constructing Global Coherence Representations Identifying Interpretability and Coherences of Transf.pdf:application/pdf;IEEE Xplore Abstract Record:/home/alex/Zotero/storage/Z45R3XWF/9564126.html:text/html}, } -@misc{noauthor_playtikaosstft-torch_2025, - title = {{PlaytikaOSS}/tft-torch}, - copyright = {MIT}, - url = {https://github.com/PlaytikaOSS/tft-torch}, - abstract = {A Python library that implements ״Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting״}, - urldate = {2025-03-19}, - publisher = {Playtika}, - month = jan, - year = {2025}, - note = {original-date: 2021-11-28T07:08:32Z}, +@incollection{iliadis_temporal_2023, + location = {Cham}, + title = {Temporal Attention Signatures for Interpretable Time-Series Prediction}, + volume = {14259}, + isbn = {978-3-031-44222-3 978-3-031-44223-0}, + url = {https://link.springer.com/10.1007/978-3-031-44223-0_22}, + abstract = {Deep neural networks have become a staple in time-series prediction due to their remarkable accuracy. However, their internal workings often remain elusive. Significant advancements have been made in the interpretability of these networks, with attention mechanisms and feature maps being notably effective for image classification by highlighting the crucial data points. While human observers can readily confirm the significance of features in image classification, the interpretability of time-series data and its modeling remains challenging. To address this, we put forth an innovative approach that unifies temporal attention and visualization as a blend of recurrent neural networks, self-attention, and general attention. This synergy results in the generation of temporal attention signatures, akin to image attention heat maps. Temporal attention not only enhances prediction accuracy beyond that of recurrent networks alone but also demonstrates that varying label classes yield distinct attention signatures. This observation indicates that neural networks focus on different sections of time-series sequences contingent on the prediction target. We conclude with a discussion on the practical implications of this novel approach, including its applicability to model interpretation, sequence length selection, and model validation. This leads to more accurate, robust, and interpretable models, instilling greater confidence in their results.}, + pages = {268--280}, + booktitle = {Artificial Neural Networks and Machine Learning – {ICANN} 2023}, + publisher = {Springer Nature Switzerland}, + author = {Katrompas, Alexander and Metsis, Vangelis}, + editor = {Iliadis, Lazaros and Papaleonidas, Antonios and Angelov, Plamen and Jayne, Chrisina}, + urldate = {2025-02-25}, + date = {2023}, + langid = {english}, + doi = {10.1007/978-3-031-44223-0_22}, + note = {Series Title: Lecture Notes in Computer Science}, + file = {PDF:/home/alex/Zotero/storage/LV7IVKZK/Katrompas and Metsis - 2023 - Temporal Attention Signatures for Interpretable Time-Series Prediction.pdf:application/pdf}, } -@misc{sherar_mattsherartemporal_fusion_transform_2025, - title = {mattsherar/{Temporal}\_Fusion\_Transform}, - url = {https://github.com/mattsherar/Temporal_Fusion_Transform}, - abstract = {Pytorch Implementation of Google's TFT}, - urldate = {2025-03-19}, - author = {Sherar, Matthew}, - month = mar, - year = {2025}, - note = {original-date: 2020-01-11T17:54:01Z}, +@inproceedings{guo_exploring_2019, + title = {Exploring interpretable {LSTM} neural networks over multi-variable data}, + url = {https://proceedings.mlr.press/v97/guo19b.html}, + abstract = {For recurrent neural networks trained on time series with target and exogenous variables, in addition to accurate prediction, it is also desired to provide interpretable insights into the data. In this paper, we explore the structure of {LSTM} recurrent neural networks to learn variable-wise hidden states, with the aim to capture different dynamics in multi-variable time series and distinguish the contribution of variables to the prediction. With these variable-wise hidden states, a mixture attention mechanism is proposed to model the generative process of the target. Then we develop associated training methods to jointly learn network parameters, variable and temporal importance w.r.t the prediction of the target variable. Extensive experiments on real datasets demonstrate enhanced prediction performance by capturing the dynamics of different variables. Meanwhile, we evaluate the interpretation results both qualitatively and quantitatively. It exhibits the prospect as an end-to-end framework for both forecasting and knowledge extraction over multi-variable data.}, + eventtitle = {International Conference on Machine Learning}, + pages = {2494--2504}, + booktitle = {Proceedings of the 36th International Conference on Machine Learning}, + publisher = {{PMLR}}, + author = {Guo, Tian and Lin, Tao and Antulov-Fantulin, Nino}, + urldate = {2025-02-25}, + date = {2019-05-24}, + langid = {english}, + note = {{ISSN}: 2640-3498}, + file = {Full Text PDF:/home/alex/Zotero/storage/VV3I2T4E/Guo et al. - 2019 - Exploring interpretable LSTM neural networks over multi-variable data.pdf:application/pdf;Supplementary PDF:/home/alex/Zotero/storage/57IK29PA/Guo et al. - 2019 - Exploring interpretable LSTM neural networks over multi-variable data.pdf:application/pdf}, } -@misc{noauthor_temporal_nodate, - title = {Temporal {Fusion} {Transformer} ({TFT}) — darts documentation}, - url = {https://unit8co.github.io/darts/generated_api/darts.models.forecasting.tft_model.html}, - urldate = {2025-03-19}, - file = {Temporal Fusion Transformer (TFT) — darts documentation:/home/alex/Zotero/storage/5QNI6WSL/darts.models.forecasting.tft_model.html:text/html}, +@article{yuan_dcfa-itimenet_2024, + title = {{DCFA}-{iTimeNet}: Dynamic cross-fusion attention network for interpretable time series prediction}, + volume = {55}, + issn = {1573-7497}, + url = {https://doi.org/10.1007/s10489-024-05973-2}, + doi = {10.1007/s10489-024-05973-2}, + shorttitle = {{DCFA}-{iTimeNet}}, + abstract = {Although time series prediction research among engineering and technology has made breakthrough progress in performance, challenges remain in modeling complex dynamic interactions between variables and interpretability. To address these two problems, a novel two-stage strategy framework called {DCFA}-{iTimeNet} is introduced. In the first stage, this paper innovatively proposes a dynamic cross-fusion attention mechanism ({DCFA}) . This module facilitates the model to exchange information between different patches of the time series, thereby capturing the complex interactions between variables across time. In the second stage, we exploit a decomposition-based linear explainable Bidirectional Gated Recurrent Unit ({DeLEBiGRU}), which consists mainly of standard {BiGRU} and tensorized {BiGRU}. It is proposed to analyze each variable’s historical long-term, instantaneous, and future impacts. Such design is crucial for understanding how each variable impacts the overall prediction over time. Extensive experimental results demonstrate that the proposed model can effectively model and interpret complex dynamic relationships of multivariate time series and understand the model’s decision-making process. Moreover, the performance outperforms the state-of-the-art methods.}, + pages = {86}, + number = {2}, + journaltitle = {Appl Intell}, + author = {Yuan, Jianjun and Wu, Fujun and Zhao, Luoming and Pan, Dongbo and Yu, Xinyue}, + urldate = {2025-02-25}, + date = {2024-12-06}, + langid = {english}, + keywords = {Interpretability, Artificial Intelligence, Dynamic cross-fusion attention, Dynamic interaction, Time series prediction}, + file = {Full Text PDF:/home/alex/Zotero/storage/JEXYN7BN/Yuan et al. - 2024 - DCFA-iTimeNet Dynamic cross-fusion attention network for interpretable time series prediction.pdf:application/pdf}, } -@misc{dauphin_language_2017, - title = {Language {Modeling} with {Gated} {Convolutional} {Networks}}, - url = {http://arxiv.org/abs/1612.08083}, - doi = {10.48550/arXiv.1612.08083}, - abstract = {The pre-dominant approach to language modeling to date is based on recurrent neural networks. Their success on this task is often linked to their ability to capture unbounded context. In this paper we develop a finite context approach through stacked convolutions, which can be more efficient since they allow parallelization over sequential tokens. We propose a novel simplified gating mechanism that outperforms Oord et al (2016) and investigate the impact of key architectural decisions. The proposed approach achieves state-of-the-art on the WikiText-103 benchmark, even though it features long-term dependencies, as well as competitive results on the Google Billion Words benchmark. Our model reduces the latency to score a sentence by an order of magnitude compared to a recurrent baseline. To our knowledge, this is the first time a non-recurrent approach is competitive with strong recurrent models on these large scale language tasks.}, - urldate = {2025-03-26}, - publisher = {arXiv}, - author = {Dauphin, Yann N. and Fan, Angela and Auli, Michael and Grangier, David}, - month = sep, - year = {2017}, - note = {arXiv:1612.08083 [cs]}, - keywords = {Computer Science - Computation and Language}, - file = {Full Text PDF:/home/alex/Zotero/storage/4SBUNZ4A/Dauphin et al. - 2017 - Language Modeling with Gated Convolutional Networks.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/TQBL4EZ7/1612.html:text/html}, -} - -@misc{ba_layer_2016, - title = {Layer {Normalization}}, - url = {http://arxiv.org/abs/1607.06450}, - doi = {10.48550/arXiv.1607.06450}, - abstract = {Training state-of-the-art, deep neural networks is computationally expensive. One way to reduce the training time is to normalize the activities of the neurons. A recently introduced technique called batch normalization uses the distribution of the summed input to a neuron over a mini-batch of training cases to compute a mean and variance which are then used to normalize the summed input to that neuron on each training case. This significantly reduces the training time in feed-forward neural networks. However, the effect of batch normalization is dependent on the mini-batch size and it is not obvious how to apply it to recurrent neural networks. In this paper, we transpose batch normalization into layer normalization by computing the mean and variance used for normalization from all of the summed inputs to the neurons in a layer on a single training case. Like batch normalization, we also give each neuron its own adaptive bias and gain which are applied after the normalization but before the non-linearity. Unlike batch normalization, layer normalization performs exactly the same computation at training and test times. It is also straightforward to apply to recurrent neural networks by computing the normalization statistics separately at each time step. Layer normalization is very effective at stabilizing the hidden state dynamics in recurrent networks. Empirically, we show that layer normalization can substantially reduce the training time compared with previously published techniques.}, - urldate = {2025-03-26}, - publisher = {arXiv}, - author = {Ba, Jimmy Lei and Kiros, Jamie Ryan and Hinton, Geoffrey E.}, - month = jul, - year = {2016}, - note = {arXiv:1607.06450 [stat]}, - keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, - file = {Full Text PDF:/home/alex/Zotero/storage/MJWRDPWE/Ba et al. - 2016 - Layer Normalization.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/F9WSU957/1607.html:text/html}, -} - -@misc{clevert_fast_2016, - title = {Fast and {Accurate} {Deep} {Network} {Learning} by {Exponential} {Linear} {Units} ({ELUs})}, - url = {http://arxiv.org/abs/1511.07289}, - doi = {10.48550/arXiv.1511.07289}, - abstract = {We introduce the "exponential linear unit" (ELU) which speeds up learning in deep neural networks and leads to higher classification accuracies. Like rectified linear units (ReLUs), leaky ReLUs (LReLUs) and parametrized ReLUs (PReLUs), ELUs alleviate the vanishing gradient problem via the identity for positive values. However, ELUs have improved learning characteristics compared to the units with other activation functions. In contrast to ReLUs, ELUs have negative values which allows them to push mean unit activations closer to zero like batch normalization but with lower computational complexity. Mean shifts toward zero speed up learning by bringing the normal gradient closer to the unit natural gradient because of a reduced bias shift effect. While LReLUs and PReLUs have negative values, too, they do not ensure a noise-robust deactivation state. ELUs saturate to a negative value with smaller inputs and thereby decrease the forward propagated variation and information. Therefore, ELUs code the degree of presence of particular phenomena in the input, while they do not quantitatively model the degree of their absence. In experiments, ELUs lead not only to faster learning, but also to significantly better generalization performance than ReLUs and LReLUs on networks with more than 5 layers. On CIFAR-100 ELUs networks significantly outperform ReLU networks with batch normalization while batch normalization does not improve ELU networks. ELU networks are among the top 10 reported CIFAR-10 results and yield the best published result on CIFAR-100, without resorting to multi-view evaluation or model averaging. On ImageNet, ELU networks considerably speed up learning compared to a ReLU network with the same architecture, obtaining less than 10\% classification error for a single crop, single model network.}, - urldate = {2025-03-26}, - publisher = {arXiv}, - author = {Clevert, Djork-Arné and Unterthiner, Thomas and Hochreiter, Sepp}, - month = feb, - year = {2016}, - note = {arXiv:1511.07289 [cs]}, +@misc{sprang_enforcing_2024, + title = {Enforcing Interpretability in Time Series Transformers: A Concept Bottleneck Framework}, + url = {http://arxiv.org/abs/2410.06070}, + doi = {10.48550/arXiv.2410.06070}, + shorttitle = {Enforcing Interpretability in Time Series Transformers}, + abstract = {There has been a recent push of research on Transformer-based models for long-term time series forecasting, even though they are inherently difficult to interpret and explain. While there is a large body of work on interpretability methods for various domains and architectures, the interpretability of Transformer-based forecasting models remains largely unexplored. To address this gap, we develop a framework based on Concept Bottleneck Models to enforce interpretability of time series Transformers. We modify the training objective to encourage a model to develop representations similar to predefined interpretable concepts. In our experiments, we enforce similarity using Centered Kernel Alignment, and the predefined concepts include time features and an interpretable, autoregressive surrogate model ({AR}). We apply the framework to the Autoformer model, and present an in-depth analysis for a variety of benchmark tasks. We find that the model performance remains mostly unaffected, while the model shows much improved interpretability. Additionally, interpretable concepts become local, which makes the trained model easily intervenable. As a proof of concept, we demonstrate a successful intervention in the scenario of a time shift in the data, which eliminates the need to retrain.}, + number = {{arXiv}:2410.06070}, + publisher = {{arXiv}}, + author = {Sprang, Angela van and Acar, Erman and Zuidema, Willem}, + urldate = {2025-02-25}, + date = {2024-10-08}, + eprinttype = {arxiv}, + eprint = {2410.06070 [cs]}, keywords = {Computer Science - Machine Learning}, - annote = {Comment: Published as a conference paper at ICLR 2016}, - file = {Full Text PDF:/home/alex/Zotero/storage/PU3ZGP4G/Clevert et al. - 2016 - Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs).pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/3MAW2IWE/1511.html:text/html}, + file = {Preprint PDF:/home/alex/Zotero/storage/HVUXXRXJ/Sprang et al. - 2024 - Enforcing Interpretability in Time Series Transformers A Concept Bottleneck Framework.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/6QS78DZP/2410.html:text/html}, } -@article{wang_systematic_2022, - title = {A {Systematic} {Review} of {Time} {Series} {Classification} {Techniques} {Used} in {Biomedical} {Applications}}, - volume = {22}, - issn = {1424-8220}, - url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9611376/}, - doi = {10.3390/s22208016}, - abstract = {Background: Digital clinical measures collected via various digital sensing technologies such as smartphones, smartwatches, wearables, and ingestible and implantable sensors are increasingly used by individuals and clinicians to capture the health outcomes or behavioral and physiological characteristics of individuals. Time series classification (TSC) is very commonly used for modeling digital clinical measures. While deep learning models for TSC are very common and powerful, there exist some fundamental challenges. This review presents the non-deep learning models that are commonly used for time series classification in biomedical applications that can achieve high performance. Objective: We performed a systematic review to characterize the techniques that are used in time series classification of digital clinical measures throughout all the stages of data processing and model building. Methods: We conducted a literature search on PubMed, as well as the Institute of Electrical and Electronics Engineers (IEEE), Web of Science, and SCOPUS databases using a range of search terms to retrieve peer-reviewed articles that report on the academic research about digital clinical measures from a five-year period between June 2016 and June 2021. We identified and categorized the research studies based on the types of classification algorithms and sensor input types. Results: We found 452 papers in total from four different databases: PubMed, IEEE, Web of Science Database, and SCOPUS. After removing duplicates and irrelevant papers, 135 articles remained for detailed review and data extraction. Among these, engineered features using time series methods that were subsequently fed into widely used machine learning classifiers were the most commonly used technique, and also most frequently achieved the best performance metrics (77 out of 135 articles). Statistical modeling (24 out of 135 articles) algorithms were the second most common and also the second-best classification technique. Conclusions: In this review paper, summaries of the time series classification models and interpretation methods for biomedical applications are summarized and categorized. While high time series classification performance has been achieved in digital clinical, physiological, or biomedical measures, no standard benchmark datasets, modeling methods, or reporting methodology exist. There is no single widely used method for time series model development or feature interpretation, however many different methods have proven successful.}, - number = {20}, - urldate = {2025-05-06}, - journal = {Sensors (Basel, Switzerland)}, - author = {Wang, Will Ke and Chen, Ina and Hershkovich, Leeor and Yang, Jiamu and Shetty, Ayush and Singh, Geetika and Jiang, Yihang and Kotla, Aditya and Shang, Jason Zisheng and Yerrabelli, Rushil and Roghanizad, Ali R. and Shandhi, Md Mobashir Hasan and Dunn, Jessilyn}, - month = oct, - year = {2022}, - pmid = {36298367}, - pmcid = {PMC9611376}, - pages = {8016}, - file = {Full Text PDF:/home/alex/Zotero/storage/2E8EIVER/Wang et al. - 2022 - A Systematic Review of Time Series Classification Techniques Used in Biomedical Applications.pdf:application/pdf}, +@misc{chefer_transformer_2021, + title = {Transformer Interpretability Beyond Attention Visualization}, + url = {http://arxiv.org/abs/2012.09838}, + doi = {10.48550/arXiv.2012.09838}, + abstract = {Self-attention techniques, and specifically Transformers, are dominating the field of text processing and are becoming increasingly popular in computer vision classification tasks. In order to visualize the parts of the image that led to a certain classification, existing methods either rely on the obtained attention maps or employ heuristic propagation along the attention graph. In this work, we propose a novel way to compute relevancy for Transformer networks. The method assigns local relevance based on the Deep Taylor Decomposition principle and then propagates these relevancy scores through the layers. This propagation involves attention layers and skip connections, which challenge existing methods. Our solution is based on a specific formulation that is shown to maintain the total relevancy across layers. We benchmark our method on very recent visual Transformer networks, as well as on a text classification problem, and demonstrate a clear advantage over the existing explainability methods.}, + number = {{arXiv}:2012.09838}, + publisher = {{arXiv}}, + author = {Chefer, Hila and Gur, Shir and Wolf, Lior}, + urldate = {2025-02-25}, + date = {2021-04-05}, + eprinttype = {arxiv}, + eprint = {2012.09838 [cs]}, + keywords = {Computer Science - Computer Vision and Pattern Recognition}, + file = {Preprint PDF:/home/alex/Zotero/storage/3FRISAP7/Chefer et al. - 2021 - Transformer Interpretability Beyond Attention Visualization.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/QUPBJPC9/2012.html:text/html}, } -@article{gharehbaghi_deep_2018, - title = {A {Deep} {Machine} {Learning} {Method} for {Classifying} {Cyclic} {Time} {Series} of {Biological} {Signals} {Using} {Time}-{Growing} {Neural} {Network}}, - volume = {29}, - copyright = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html}, - issn = {2162-237X, 2162-2388}, - url = {https://ieeexplore.ieee.org/document/8066455/}, - doi = {10.1109/TNNLS.2017.2754294}, - number = {9}, - urldate = {2025-05-06}, - journal = {IEEE Transactions on Neural Networks and Learning Systems}, - author = {Gharehbaghi, Arash and Linden, Maria}, - month = sep, - year = {2018}, - pages = {4102--4115}, +@misc{helbling_conceptattention_2025, + title = {{ConceptAttention}: Diffusion Transformers Learn Highly Interpretable Features}, + url = {http://arxiv.org/abs/2502.04320}, + doi = {10.48550/arXiv.2502.04320}, + shorttitle = {{ConceptAttention}}, + abstract = {Do the rich representations of multi-modal diffusion transformers ({DiTs}) exhibit unique properties that enhance their interpretability? We introduce {ConceptAttention}, a novel method that leverages the expressive power of {DiT} attention layers to generate high-quality saliency maps that precisely locate textual concepts within images. Without requiring additional training, {ConceptAttention} repurposes the parameters of {DiT} attention layers to produce highly contextualized concept embeddings, contributing the major discovery that performing linear projections in the output space of {DiT} attention layers yields significantly sharper saliency maps compared to commonly used cross-attention mechanisms. Remarkably, {ConceptAttention} even achieves state-of-the-art performance on zero-shot image segmentation benchmarks, outperforming 11 other zero-shot interpretability methods on the {ImageNet}-Segmentation dataset and on a single-class subset of {PascalVOC}. Our work contributes the first evidence that the representations of multi-modal {DiT} models like Flux are highly transferable to vision tasks like segmentation, even outperforming multi-modal foundation models like {CLIP}.}, + number = {{arXiv}:2502.04320}, + publisher = {{arXiv}}, + author = {Helbling, Alec and Meral, Tuna Han Salih and Hoover, Ben and Yanardag, Pinar and Chau, Duen Horng}, + urldate = {2025-02-25}, + date = {2025-02-06}, + eprinttype = {arxiv}, + eprint = {2502.04320 [cs]}, + keywords = {Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition}, + file = {Preprint PDF:/home/alex/Zotero/storage/AEEDM4ZW/Helbling et al. - 2025 - ConceptAttention Diffusion Transformers Learn Highly Interpretable Features.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/SAHGAINH/2502.html:text/html}, } -@article{masini_machine_2023, - title = {Machine learning advances for time series forecasting}, +@article{noauthor_temporal_2021, + title = {Temporal Fusion Transformers for interpretable multi-horizon time series forecasting}, volume = {37}, - issn = {0950-0804, 1467-6419}, - url = {https://onlinelibrary.wiley.com/doi/10.1111/joes.12429}, - doi = {10.1111/joes.12429}, - abstract = {Abstract - In this paper, we survey the most recent advances in supervised machine learning (ML) and high‐dimensional models for time‐series forecasting. We consider both linear and nonlinear alternatives. Among the linear methods, we pay special attention to penalized regressions and ensemble of models. The nonlinear methods considered in the paper include shallow and deep neural networks, in their feedforward and recurrent versions, and tree‐based methods, such as random forests and boosted trees. We also consider ensemble and hybrid models by combining ingredients from different alternatives. Tests for superior predictive ability are briefly reviewed. Finally, we discuss application of ML in economics and finance and provide an illustration with high‐frequency financial data.}, - language = {en}, - number = {1}, - urldate = {2025-05-06}, - journal = {Journal of Economic Surveys}, - author = {Masini, Ricardo P. and Medeiros, Marcelo C. and Mendes, Eduardo F.}, - month = feb, - year = {2023}, - pages = {76--111}, - file = {Submitted Version:/home/alex/Zotero/storage/4TZJJUSU/Masini et al. - 2023 - Machine learning advances for time series forecasting.pdf:application/pdf}, + issn = {0169-2070}, + url = {https://www.sciencedirect.com/science/article/pii/S0169207021000637}, + doi = {10.1016/j.ijforecast.2021.03.012}, + abstract = {Multi-horizon forecasting often contains a complex mix of inputs – including static (i.e. time-invariant) covariates, known future inputs, and other e…}, + pages = {1748--1764}, + number = {4}, + journaltitle = {International Journal of Forecasting}, + urldate = {2025-02-24}, + date = {2021-10-01}, + langid = {american}, + note = {Publisher: Elsevier}, + file = {Snapshot:/home/alex/Zotero/storage/SFYESIWK/S0169207021000637.html:text/html;Submitted Version:/home/alex/Zotero/storage/A9AYS5UI/2021 - Temporal Fusion Transformers for interpretable multi-horizon time series forecasting.pdf:application/pdf}, } -@article{wang_systematic_2022-1, - title = {A {Systematic} {Review} of {Time} {Series} {Classification} {Techniques} {Used} in {Biomedical} {Applications}}, - volume = {22}, - copyright = {https://creativecommons.org/licenses/by/4.0/}, - issn = {1424-8220}, - url = {https://www.mdpi.com/1424-8220/22/20/8016}, - doi = {10.3390/s22208016}, - abstract = {Background: Digital clinical measures collected via various digital sensing technologies such as smartphones, smartwatches, wearables, and ingestible and implantable sensors are increasingly used by individuals and clinicians to capture the health outcomes or behavioral and physiological characteristics of individuals. Time series classification (TSC) is very commonly used for modeling digital clinical measures. While deep learning models for TSC are very common and powerful, there exist some fundamental challenges. This review presents the non-deep learning models that are commonly used for time series classification in biomedical applications that can achieve high performance. Objective: We performed a systematic review to characterize the techniques that are used in time series classification of digital clinical measures throughout all the stages of data processing and model building. Methods: We conducted a literature search on PubMed, as well as the Institute of Electrical and Electronics Engineers (IEEE), Web of Science, and SCOPUS databases using a range of search terms to retrieve peer-reviewed articles that report on the academic research about digital clinical measures from a five-year period between June 2016 and June 2021. We identified and categorized the research studies based on the types of classification algorithms and sensor input types. Results: We found 452 papers in total from four different databases: PubMed, IEEE, Web of Science Database, and SCOPUS. After removing duplicates and irrelevant papers, 135 articles remained for detailed review and data extraction. Among these, engineered features using time series methods that were subsequently fed into widely used machine learning classifiers were the most commonly used technique, and also most frequently achieved the best performance metrics (77 out of 135 articles). Statistical modeling (24 out of 135 articles) algorithms were the second most common and also the second-best classification technique. Conclusions: In this review paper, summaries of the time series classification models and interpretation methods for biomedical applications are summarized and categorized. While high time series classification performance has been achieved in digital clinical, physiological, or biomedical measures, no standard benchmark datasets, modeling methods, or reporting methodology exist. There is no single widely used method for time series model development or feature interpretation, however many different methods have proven successful.}, - language = {en}, - number = {20}, - urldate = {2025-05-06}, - journal = {Sensors}, - author = {Wang, Will Ke and Chen, Ina and Hershkovich, Leeor and Yang, Jiamu and Shetty, Ayush and Singh, Geetika and Jiang, Yihang and Kotla, Aditya and Shang, Jason Zisheng and Yerrabelli, Rushil and Roghanizad, Ali R. and Shandhi, Md Mobashir Hasan and Dunn, Jessilyn}, - month = oct, - year = {2022}, - pages = {8016}, - file = {PDF:/home/alex/Zotero/storage/H5LLUB5K/Wang et al. - 2022 - A Systematic Review of Time Series Classification Techniques Used in Biomedical Applications.pdf:application/pdf}, +@inproceedings{vaswani_attention_2017, + title = {Attention is All you Need}, + volume = {30}, + url = {https://proceedings.neurips.cc/paper/2017/hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html}, + abstract = {The dominant sequence transduction models are based on complex recurrent orconvolutional neural networks in an encoder and decoder configuration. The best performing such models also connect the encoder and decoder through an attentionm echanisms. We propose a novel, simple network architecture based solely onan attention mechanism, dispensing with recurrence and convolutions entirely.Experiments on two machine translation tasks show these models to be superiorin quality while being more parallelizable and requiring significantly less timeto train. Our single model with 165 million parameters, achieves 27.5 {BLEU} {onEnglish}-to-German translation, improving over the existing best ensemble result by over 1 {BLEU}. On English-to-French translation, we outperform the previoussingle state-of-the-art with model by 0.7 {BLEU}, achieving a {BLEU} score of 41.1.}, + booktitle = {Advances in Neural Information Processing Systems}, + publisher = {Curran Associates, Inc.}, + author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, Ł ukasz and Polosukhin, Illia}, + urldate = {2025-02-24}, + date = {2017}, + file = {Full Text PDF:/home/alex/Zotero/storage/MU7NU9LR/Vaswani et al. - 2017 - Attention is All you Need.pdf:application/pdf}, } -@article{leon-lopez_anomaly_2022, - title = {Anomaly {Detection} and {Classification} in {Multispectral} {Time} {Series} {Based} on {Hidden} {Markov} {Models}}, - volume = {60}, - issn = {1558-0644}, - url = {https://ieeexplore.ieee.org/abstract/document/9509347}, - doi = {10.1109/TGRS.2021.3101127}, - abstract = {Monitoring agriculture from satellite remote sensing data, such as multispectral images, has become a powerful tool since it has demonstrated a great potential for providing timely and accurate knowledge of crops. Detecting anomalies in time series of multispectral remote sensing images for crop monitoring is generally performed using a large sample of historical data at a pixel level. Conversely, this article presents a framework for anomaly detection (AD), localization, and classification that exploits the temporal information contained in a given season at a parcel level to detect and localize outliers using hidden Markov models (HMMs). Specifically, the AD part is based on the learning of HMM parameters associated with unlabeled normal data that are used in a second step to detect abnormal crop parcels referred to as anomalies. The learned HMM can also be used in time segments to temporally localize the anomalies affecting the crop parcels. The detected and localized anomalies are finally classified using a supervised classifier, e.g., based on support vector machines. The proposed framework is applicable to images partially covered by clouds and can handle a set of crop parcels acquired in the same season bypassing problems due to crop rotations. Numerical experiments are conducted on synthetic and real data, where the real data correspond to vegetation indices extracted from several multitemporal Sentinel-2 images of rapeseed crops. The proposed approach is compared to standard AD methods yielding better detection rates with the advantage of allowing anomalies to be localized and characterized.}, - urldate = {2025-05-06}, - journal = {IEEE Transactions on Geoscience and Remote Sensing}, - author = {León-López, Kareth M. and Mouret, Florian and Arguello, Henry and Tourneret, Jean-Yves}, - year = {2022}, - keywords = {Hidden Markov models, Time series analysis, Agricultural monitoring, Agriculture, anomaly classification, Anomaly detection, anomaly detection (AD), Feature extraction, hidden Markov models (HMMs), Monitoring, remote sensing, time series, Vegetation mapping}, - pages = {1--11}, - file = {Snapshot:/home/alex/Zotero/storage/HIWT8MDH/9509347.html:text/html;Submitted Version:/home/alex/Zotero/storage/S4K2XXHN/León-López et al. - 2022 - Anomaly Detection and Classification in Multispectral Time Series Based on Hidden Markov Models.pdf:application/pdf}, +@online{noauthor_vivosens_nodate, + title = {vivosens medical gmbh}, + url = {https://www.vivosensmedical.com/}, + urldate = {2025-02-24}, + file = {vivosensmedical.com:/home/alex/Zotero/storage/KSM7HHBJ/www.vivosensmedical.com.html:text/html}, } -@inproceedings{hsieh_explainable_2021, - address = {Virtual Event Israel}, - title = {Explainable {Multivariate} {Time} {Series} {Classification}: {A} {Deep} {Neural} {Network} {Which} {Learns} to {Attend} to {Important} {Variables} {As} {Well} {As} {Time} {Intervals}}, - isbn = {978-1-4503-8297-7}, - shorttitle = {Explainable {Multivariate} {Time} {Series} {Classification}}, - url = {https://dl.acm.org/doi/10.1145/3437963.3441815}, - doi = {10.1145/3437963.3441815}, - language = {en}, - urldate = {2025-05-06}, - booktitle = {Proceedings of the 14th {ACM} {International} {Conference} on {Web} {Search} and {Data} {Mining}}, - publisher = {ACM}, - author = {Hsieh, Tsung-Yu and Wang, Suhang and Sun, Yiwei and Honavar, Vasant}, - month = mar, - year = {2021}, - pages = {607--615}, +@inproceedings{rigotti_attention-based_2021, + title = {Attention-based Interpretability with Concept Transformers}, + url = {https://openreview.net/forum?id=kAa9eDS0RdO}, + abstract = {Attention is a mechanism that has been instrumental in driving remarkable performance gains of deep neural network models in a host of visual, {NLP} and multimodal tasks. One additional notable aspect of attention is that it conveniently exposes the ``reasoning'' behind each particular output generated by the model. Specifically, attention scores over input regions or intermediate features have been interpreted as a measure of the contribution of the attended element to the model inference. While the debate in regard to the interpretability of attention is still not settled, researchers have pointed out the existence of architectures and scenarios that afford a meaningful interpretation of the attention mechanism. Here we propose the generalization of attention from low-level input features to high-level concepts as a mechanism to ensure the interpretability of attention scores within a given application domain. In particular, we design the {ConceptTransformer}, a deep learning module that exposes explanations of the output of a model in which it is embedded in terms of attention over user-defined high-level concepts. Such explanations are {\textbackslash}emph\{plausible\} (i.e.{\textbackslash} convincing to the human user) and {\textbackslash}emph\{faithful\} (i.e.{\textbackslash} truly reflective of the reasoning process of the model). Plausibility of such explanations is obtained by construction by training the attention heads to conform with known relations between inputs, concepts and outputs dictated by domain knowledge. Faithfulness is achieved by design by enforcing a linear relation between the transformer value vectors that represent the concepts and their contribution to the classification log-probabilities. We validate our {ConceptTransformer} module on established explainability benchmarks and show how it can be used to infuse domain knowledge into classifiers to improve accuracy, and conversely to extract concept-based explanations of classification outputs. Code to reproduce our results is available at: {\textbackslash}url\{https://github.com/ibm/concept\_transformer\}.}, + eventtitle = {International Conference on Learning Representations}, + author = {Rigotti, Mattia and Miksovic, Christoph and Giurgiu, Ioana and Gschwind, Thomas and Scotton, Paolo}, + urldate = {2025-02-21}, + date = {2021-10-06}, + langid = {english}, + file = {Full Text PDF:/home/alex/Zotero/storage/U2FUGVF6/Rigotti et al. - 2021 - Attention-based Interpretability with Concept Transformers.pdf:application/pdf}, } -@misc{saluja_towards_2021, - title = {Towards a {Rigorous} {Evaluation} of {Explainability} for {Multivariate} {Time} {Series}}, - url = {http://arxiv.org/abs/2104.04075}, - doi = {10.48550/arXiv.2104.04075}, - abstract = {Machine learning-based systems are rapidly gaining popularity and in-line with that there has been a huge research surge in the field of explainability to ensure that machine learning models are reliable, fair, and can be held liable for their decision-making process. Explainable Artificial Intelligence (XAI) methods are typically deployed to debug black-box machine learning models but in comparison to tabular, text, and image data, explainability in time series is still relatively unexplored. The aim of this study was to achieve and evaluate model agnostic explainability in a time series forecasting problem. This work focused on proving a solution for a digital consultancy company aiming to find a data-driven approach in order to understand the effect of their sales related activities on the sales deals closed. The solution involved framing the problem as a time series forecasting problem to predict the sales deals and the explainability was achieved using two novel model agnostic explainability techniques, Local explainable model-agnostic explanations (LIME) and Shapley additive explanations (SHAP) which were evaluated using human evaluation of explainability. The results clearly indicate that the explanations produced by LIME and SHAP greatly helped lay humans in understanding the predictions made by the machine learning model. The presented work can easily be extended to any time}, - urldate = {2025-05-06}, - publisher = {arXiv}, - author = {Saluja, Rohit and Malhi, Avleen and Knapič, Samanta and Främling, Kary and Cavdar, Cicek}, - month = apr, - year = {2021}, - note = {arXiv:2104.04075 [cs]}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - annote = {Comment: Journal}, - file = {Preprint PDF:/home/alex/Zotero/storage/6XB73Y2F/Saluja et al. - 2021 - Towards a Rigorous Evaluation of Explainability for Multivariate Time Series.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/CDCGX8EZ/2104.html:text/html}, +@article{kitada_attention_2021, + title = {Attention Meets Perturbations: Robust and Interpretable Attention With Adversarial Training}, + volume = {9}, + issn = {2169-3536}, + url = {https://ieeexplore.ieee.org/abstract/document/9467291}, + doi = {10.1109/ACCESS.2021.3093456}, + shorttitle = {Attention Meets Perturbations}, + abstract = {Although attention mechanisms have been applied to a variety of deep learning models and have been shown to improve the prediction performance, it has been reported to be vulnerable to perturbations to the mechanism. To overcome the vulnerability to perturbations in the mechanism, we are inspired by adversarial training ({AT}), which is a powerful regularization technique for enhancing the robustness of the models. In this paper, we propose a general training technique for natural language processing tasks, including {AT} for attention (Attention {AT}) and more interpretable {AT} for attention (Attention {iAT}). The proposed techniques improved the prediction performance and the model interpretability by exploiting the mechanisms with {AT}. In particular, Attention {iAT} boosts those advantages by introducing adversarial perturbation, which enhances the difference in the attention of the sentences. Evaluation experiments with ten open datasets revealed that {AT} for attention mechanisms, especially Attention {iAT}, demonstrated (1) the best performance in nine out of ten tasks and (2) more interpretable attention (i.e., the resulting attention correlated more strongly with gradient-based word importance) for all tasks. Additionally, the proposed techniques are (3) much less dependent on perturbation size in {AT}.}, + pages = {92974--92985}, + journaltitle = {{IEEE} Access}, + author = {Kitada, Shunsuke and Iyatomi, Hitoshi}, + urldate = {2025-02-21}, + date = {2021}, + note = {Conference Name: {IEEE} Access}, + keywords = {Predictive models, adversarial training, attention mechanism, binary classification, interpretability, Knowledge discovery, natural language inference, Natural language processing, Perturbation methods, question answering, Robustness, Solid modeling, Task analysis, Training}, + file = {Full Text PDF:/home/alex/Zotero/storage/BUXFCXRU/Kitada and Iyatomi - 2021 - Attention Meets Perturbations Robust and Interpretable Attention With Adversarial Training.pdf:application/pdf;IEEE Xplore Abstract Record:/home/alex/Zotero/storage/ZNQD3FRW/9467291.html:text/html}, } -@misc{wang_timexer_2024, - title = {{TimeXer}: {Empowering} {Transformers} for {Time} {Series} {Forecasting} with {Exogenous} {Variables}}, - shorttitle = {{TimeXer}}, - url = {http://arxiv.org/abs/2402.19072}, - doi = {10.48550/arXiv.2402.19072}, - abstract = {Deep models have demonstrated remarkable performance in time series forecasting. However, due to the partially-observed nature of real-world applications, solely focusing on the target of interest, so-called endogenous variables, is usually insufficient to guarantee accurate forecasting. Notably, a system is often recorded into multiple variables, where the exogenous variables can provide valuable external information for endogenous variables. Thus, unlike well-established multivariate or univariate forecasting paradigms that either treat all the variables equally or ignore exogenous information, this paper focuses on a more practical setting: time series forecasting with exogenous variables. We propose a novel approach, TimeXer, to ingest external information to enhance the forecasting of endogenous variables. With deftly designed embedding layers, TimeXer empowers the canonical Transformer with the ability to reconcile endogenous and exogenous information, where patch-wise self-attention and variate-wise cross-attention are used simultaneously. Moreover, global endogenous tokens are learned to effectively bridge the causal information underlying exogenous series into endogenous temporal patches. Experimentally, TimeXer achieves consistent state-of-the-art performance on twelve real-world forecasting benchmarks and exhibits notable generality and scalability. Code is available at this repository: https://github.com/thuml/TimeXer.}, - urldate = {2025-05-09}, - publisher = {arXiv}, - author = {Wang, Yuxuan and Wu, Haixu and Dong, Jiaxiang and Qin, Guo and Zhang, Haoran and Liu, Yong and Qiu, Yunzhong and Wang, Jianmin and Long, Mingsheng}, - month = nov, - year = {2024}, - note = {arXiv:2402.19072 [cs]}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - file = {Full Text PDF:/home/alex/Zotero/storage/76BQWVIW/Wang et al. - 2024 - TimeXer Empowering Transformers for Time Series Forecasting with Exogenous Variables.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/JL64E9YL/2402.html:text/html}, -} - -@misc{zeng_are_2022, - title = {Are {Transformers} {Effective} for {Time} {Series} {Forecasting}?}, - url = {http://arxiv.org/abs/2205.13504}, - doi = {10.48550/arXiv.2205.13504}, - abstract = {Recently, there has been a surge of Transformer-based solutions for the long-term time series forecasting (LTSF) task. Despite the growing performance over the past few years, we question the validity of this line of research in this work. Specifically, Transformers is arguably the most successful solution to extract the semantic correlations among the elements in a long sequence. However, in time series modeling, we are to extract the temporal relations in an ordered set of continuous points. While employing positional encoding and using tokens to embed sub-series in Transformers facilitate preserving some ordering information, the nature of the {\textbackslash}emph\{permutation-invariant\} self-attention mechanism inevitably results in temporal information loss. To validate our claim, we introduce a set of embarrassingly simple one-layer linear models named LTSF-Linear for comparison. Experimental results on nine real-life datasets show that LTSF-Linear surprisingly outperforms existing sophisticated Transformer-based LTSF models in all cases, and often by a large margin. Moreover, we conduct comprehensive empirical studies to explore the impacts of various design elements of LTSF models on their temporal relation extraction capability. We hope this surprising finding opens up new research directions for the LTSF task. We also advocate revisiting the validity of Transformer-based solutions for other time series analysis tasks (e.g., anomaly detection) in the future. Code is available at: {\textbackslash}url\{https://github.com/cure-lab/LTSF-Linear\}.}, - urldate = {2025-05-09}, - publisher = {arXiv}, - author = {Zeng, Ailing and Chen, Muxi and Zhang, Lei and Xu, Qiang}, - month = aug, - year = {2022}, - note = {arXiv:2205.13504 [cs]}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, - annote = {Comment: Code is available at https://github.com/cure-lab/LTSF-Linear}, - file = {Full Text PDF:/home/alex/Zotero/storage/V9E95F7E/Zeng et al. - 2022 - Are Transformers Effective for Time Series Forecasting.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/HNVIZG98/2205.html:text/html}, -} - -@article{barrera-animas_rainfall_2022, - title = {Rainfall prediction: {A} comparative analysis of modern machine learning algorithms for time-series forecasting}, - volume = {7}, - issn = {26668270}, - shorttitle = {Rainfall prediction}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S266682702100102X}, - doi = {10.1016/j.mlwa.2021.100204}, - abstract = {Rainfall forecasting has gained utmost research relevance in recent times due to its complexities and persistent applications such as flood forecasting and monitoring of pollutant concentration levels, among others. Existing models use complex statistical models that are often too costly, both computationally and budgetary, or are not applied to downstream applications. Therefore, approaches that use Machine Learning algorithms in conjunction with time-series data are being explored as an alternative to overcome these drawbacks. To this end, this study presents a comparative analysis using simplified rainfall estimation models based on conventional Machine Learning algorithms and Deep Learning architectures that are efficient for these downstream applications. Models based on LSTM, Stacked-LSTM, Bidirectional-LSTM Networks, XGBoost, and an ensemble of Gradient Boosting Regressor, Linear Support Vector Regression, and an Extra-trees Regressor were compared in the task of forecasting hourly rainfall volumes using time-series data. Climate data from 2000 to 2020 from five major cities in the United Kingdom were used. The evaluation metrics of Loss, Root Mean Squared Error, Mean Absolute Error, and Root Mean Squared Logarithmic Error were used to evaluate the models’ performance. Results show that a Bidirectional-LSTM Network can be used as a rainfall forecast model with comparable performance to Stacked-LSTM Networks. Among all the models tested, the StackedLSTM Network with two hidden layers and the Bidirectional-LSTM Network performed best. This suggests that models based on LSTM-Networks with fewer hidden layers perform better for this approach; denoting its ability to be applied as an approach for budget-wise rainfall forecast applications.}, - language = {en}, - urldate = {2025-05-08}, - journal = {Machine Learning with Applications}, - author = {Barrera-Animas, Ari Yair and Oyedele, Lukumon O. and Bilal, Muhammad and Akinosho, Taofeek Dolapo and Delgado, Juan Manuel Davila and Akanbi, Lukman Adewale}, - month = mar, - year = {2022}, - pages = {100204}, - file = {PDF:/home/alex/Zotero/storage/FSXMH5N9/Barrera-Animas et al. - 2022 - Rainfall prediction A comparative analysis of modern machine learning algorithms for time-series fo.pdf:application/pdf}, -} - -@article{ahmed_empirical_2010, - title = {An {Empirical} {Comparison} of {Machine} {Learning} {Models} for {Time} {Series} {Forecasting}}, +@inproceedings{choi_retain_2016, + title = {{RETAIN}: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism}, volume = {29}, - issn = {0747-4938, 1532-4168}, - url = {http://www.tandfonline.com/doi/abs/10.1080/07474938.2010.481556}, - doi = {10.1080/07474938.2010.481556}, - language = {en}, - number = {5-6}, - urldate = {2025-05-08}, - journal = {Econometric Reviews}, - author = {Ahmed, Nesreen K. and Atiya, Amir F. and Gayar, Neamat El and El-Shishiny, Hisham}, - month = aug, - year = {2010}, - pages = {594--621}, + url = {https://proceedings.neurips.cc/paper/2016/hash/231141b34c82aa95e48810a9d1b33a79-Abstract.html}, + shorttitle = {{RETAIN}}, + abstract = {Accuracy and interpretability are two dominant features of successful predictive models. Typically, a choice must be made in favor of complex black box models such as recurrent neural networks ({RNN}) for accuracy versus less accurate but more interpretable traditional models such as logistic regression. This tradeoff poses challenges in medicine where both accuracy and interpretability are important. We addressed this challenge by developing the {REverse} Time {AttentIoN} model ({RETAIN}) for application to Electronic Health Records ({EHR}) data. {RETAIN} achieves high accuracy while remaining clinically interpretable and is based on a two-level neural attention model that detects influential past visits and significant clinical variables within those visits (e.g. key diagnoses). {RETAIN} mimics physician practice by attending the {EHR} data in a reverse time order so that recent clinical visits are likely to receive higher attention. {RETAIN} was tested on a large health system {EHR} dataset with 14 million visits completed by 263K patients over an 8 year period and demonstrated predictive accuracy and computational scalability comparable to state-of-the-art methods such as {RNN}, and ease of interpretability comparable to traditional models.}, + booktitle = {Advances in Neural Information Processing Systems}, + publisher = {Curran Associates, Inc.}, + author = {Choi, Edward and Bahadori, Mohammad Taha and Sun, Jimeng and Kulas, Joshua and Schuetz, Andy and Stewart, Walter}, + urldate = {2025-02-21}, + date = {2016}, + file = {Full Text PDF:/home/alex/Zotero/storage/XQLMYHUU/Choi et al. - 2016 - RETAIN An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism.pdf:application/pdf}, } -@article{garcia_prediction_1981, - title = {Prediction of the {Time} of {Ovulation}*}, - volume = {36}, - issn = {0015-0282}, - url = {https://www.sciencedirect.com/science/article/pii/S0015028216457304}, - doi = {10.1016/S0015-0282(16)45730-4}, - abstract = {Prediction of ovulation was established by correlation of clinical parameters, follicular development by ultrasound, and estradiol, progesterone, and luteinizing hormone (LH) determination in 71 menstrual cycles. Laparoscopic follicular aspiration was accomplished in 41 of those cycles. A 28-hour interval from the ascending limb of the LH seems to be the “ideal time” for retrieval of a preovulatory oocyte. The variability in the amount of LH to which the follicle is exposed during the LH surge seems to indicate that there is a relatively low specific value necessary for ovulation. Ovulation occurs approximately 10 ± 5 hours from the LH peak. Progesterone occurs in relation to the LH surge and is helpful for the retrospective analysis of the menstrual cycle.}, +@misc{serrano_is_2019, + title = {Is Attention Interpretable?}, + url = {http://arxiv.org/abs/1906.03731}, + doi = {10.48550/arXiv.1906.03731}, + abstract = {Attention mechanisms have recently boosted performance on a range of {NLP} tasks. Because attention layers explicitly weight input components' representations, it is also often assumed that attention can be used to identify information that models found important (e.g., specific contextualized word tokens). We test whether that assumption holds by manipulating attention weights in already-trained text classification models and analyzing the resulting differences in their predictions. While we observe some ways in which higher attention weights correlate with greater impact on model predictions, we also find many ways in which this does not hold, i.e., where gradient-based rankings of attention weights better predict their effects than their magnitudes. We conclude that while attention noisily predicts input components' overall importance to a model, it is by no means a fail-safe indicator.}, + number = {{arXiv}:1906.03731}, + publisher = {{arXiv}}, + author = {Serrano, Sofia and Smith, Noah A.}, + urldate = {2025-02-21}, + date = {2019-06-09}, + eprinttype = {arxiv}, + eprint = {1906.03731 [cs]}, + keywords = {Computer Science - Computation and Language}, + file = {Preprint PDF:/home/alex/Zotero/storage/DFZ28RG8/Serrano and Smith - 2019 - Is Attention Interpretable.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/63B9XS2Y/1906.html:text/html}, +} + +@article{lyzwinski_innovative_2024, + title = {Innovative Approaches to Menstruation and Fertility Tracking Using Wearable Reproductive Health Technology: Systematic Review}, + volume = {26}, + url = {https://www.jmir.org/2024/1/e45139}, + doi = {10.2196/45139}, + shorttitle = {Innovative Approaches to Menstruation and Fertility Tracking Using Wearable Reproductive Health Technology}, + abstract = {Background: Emerging digital health technology has moved into the reproductive health market for female individuals. In the past, mobile health apps have been used to monitor the menstrual cycle using manual entry. New technological trends involve the use of wearable devices to track fertility by assessing physiological changes such as temperature, heart rate, and respiratory rate. +Objective: The primary aims of this study are to review the types of wearables that have been developed and evaluated for menstrual cycle tracking and to examine whether they may detect changes in the menstrual cycle in female individuals. Another aim is to review whether these devices are effective for tracking various stages in the menstrual cycle including ovulation and menstruation. Finally, the secondary aim is to assess whether the studies have validated their findings by reporting accuracy and sensitivity. +Methods: A review of {PubMed} or {MEDLINE} was undertaken to evaluate wearable devices for their effectiveness in predicting fertility and differentiating between the different stages of the menstrual cycle. +Results: Fertility cycle–tracking wearables include devices that can be worn on the wrists, on the fingers, intravaginally, and inside the ear. Wearable devices hold promise for predicting different stages of the menstrual cycle including the fertile window and may be used by female individuals as part of their reproductive health. Most devices had high accuracy for detecting fertility and were able to differentiate between the luteal phase (early and late), fertile window, and menstruation by assessing changes in heart rate, heart rate variability, temperature, and respiratory rate. +Conclusions: More research is needed to evaluate consumer perspectives on reproductive technology for monitoring fertility, and ethical issues around the privacy of digital data need to be addressed. Additionally, there is also a need for more studies to validate and confirm this research, given its scarcity, especially in relation to changes in respiratory rate as a proxy for reproductive cycle staging.}, + pages = {e45139}, + number = {1}, + journaltitle = {Journal of Medical Internet Research}, + author = {Lyzwinski, Lynnette and Elgendi, Mohamed and Menon, Carlo}, + urldate = {2025-02-21}, + date = {2024-02-15}, + note = {Company: Journal of Medical Internet Research +Distributor: Journal of Medical Internet Research +Institution: Journal of Medical Internet Research +Label: Journal of Medical Internet Research +Publisher: {JMIR} Publications Inc., Toronto, Canada}, + file = {Full Text:/home/alex/Zotero/storage/IP23WZLE/Lyzwinski et al. - 2024 - Innovative Approaches to Menstruation and Fertility Tracking Using Wearable Reproductive Health Tech.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/TJN3WV3I/e45139.html:text/html}, +} + +@online{noauthor_zyklus-apps_nodate, + title = {Zyklus-Apps zur Verhütung – sicher oder Gesellschaftsspiel? - {ProQuest}}, + url = {https://www.proquest.com/openview/739071fff0941b30f3a5d33b56259c60/1?pq-origsite=gscholar&cbl=6629261}, + shorttitle = {Zyklus-Apps zur Verhütung – sicher oder Gesellschaftsspiel?}, + abstract = {Explore millions of resources from scholarly journals, books, newspapers, videos and more, on the {ProQuest} Platform.}, + urldate = {2025-02-21}, + langid = {english}, + file = {Snapshot:/home/alex/Zotero/storage/QHYUJU9G/1.html:text/html}, +} + +@article{goeckenjan_continuous_2020, + title = {Continuous Body Temperature Monitoring to Improve the Diagnosis of Female Infertility}, + volume = {80}, + rights = {Georg Thieme Verlag {KG} Stuttgart · New York}, + issn = {0016-5751}, + url = {https://www.thieme-connect.com/products/ejournals/html/10.1055/a-1191-7888}, + doi = {10.1055/a-1191-7888}, + abstract = {Introduction Ovulatory dysfunction is a major cause of female infertility. We evaluated the use of continuous body temperature monitoring with a vaginal biosensor to improve + standard diagnostic procedures for determining ovulatory dysfunction. + +Material and Methods This prospective interventional study was performed in a reproductive medicine department of a university hospital. The menstrual cycles of 51 women with + infertility were monitored and analysed using three different strategies: sonographic and hormonal assessment (standard approach), continuous core body temperature measurement and analysis + using the algorithm of {OvulaRing}, and lowest daily body temperature measurement monitored with a vaginal biosensor and analysed based on the body temperature curves used in natural family + planning. + +Results Statistically significant differences were found in the temperature curves of women with luteal phase deficiency and polycystic ovary syndrome compared to women with normal + menstrual cycles. The analysis of individual cyclofertilograms can be used to detect cycle phases and estimate the date of ovulation. + +Conclusions Continuous body temperature monitoring with a vaginal biosensor can improve the standard diagnostic procedures used to determine ovulatory dysfunction, especially if + dysfunction is due to luteal phase deficiency and polycystic ovary syndrome. Analysis of the lowest daily body temperature combined with the basal body temperature measurements used in + fertility awareness methods may be equieffective to continuous body temperature measurements with {OvulaRing}. The results of this study show that a revised diagnostic approach using fewer + hormonal assessments combined with continuous body temperature monitoring can reduce the number of appointments in an infertility clinic as well as the costs.}, + pages = {702--712}, + journaltitle = {Geburtshilfe und Frauenheilkunde}, + author = {Goeckenjan, Maren and Schiwek, Esther and Wimberger, Pauline}, + urldate = {2025-02-21}, + date = {2020-07-14}, + langid = {english}, + note = {Publisher: Georg Thieme Verlag {KG}}, + keywords = {infertility, Key words + fertility awareness, luteal phase deficiency, polycystic ovary syndrome, vaginal biosensor}, + file = {Full Text PDF:/home/alex/Zotero/storage/QKPIJD23/Goeckenjan et al. - 2020 - Continuous Body Temperature Monitoring to Improve the Diagnosis of Female Infertility.pdf:application/pdf}, +} + +@article{regidor_identification_2018, + title = {Identification and prediction of the fertile window with a new web-based medical device using a vaginal biosensor for measuring the circadian and circamensual core body temperature}, + volume = {34}, + issn = {0951-3590}, + url = {https://doi.org/10.1080/09513590.2017.1390737}, + doi = {10.1080/09513590.2017.1390737}, + abstract = {Fertility awareness-based ({FAB}) methods represent a term that includes all family planning methods that are based on the identification of the fertile window. They are based on the woman’s observation of physiological signs of the fertile and infertile phases of the menstrual cycle. The first approach consists basically in symptothermal methods accompanied by cervical mucus measurements and clinical menstrual cycling data recording. The second most often used methods are the urinary measurement of E3G and luteinizing hormone ({LH}) with a personalized computer system. Hence these systems lack the efficacy of the continuous circadian and circamensual measurement of the core body temperature. Only this approach enables the accurate detection of the ovulation during the fertile window. A new medical device called {OvulaRing} has been developed to fill this gap. In the present study, the system and its first clinical results are presented. {OvulaRing} is a medical device used just like a tampon. The device is a vaginal ring of evatane that contains an integrated biosensor. This sensor measures continuously every 5 min the core body temperature throughout the entire cycle. This device allows a circadian and circamensual intravaginal exact measurement. With this system, 288 measurements are created per day. The system can detect retrospectively and predict prospectively the fertile window of the users. One hundred and fifty eight women aged between 18 and 45 years used this medical device in an open non-randomized clinical study for 15 months. A total of 470 cycles could be recorded and were able for analysis. By the same time in a subgroup of patients, hormonal assessments of {LH}, follicle-stimulating hormone, estradiol and progesterone as well as vaginal ultrasound were performed in parallel between the 9th and the 36th day of the cycle. The validation error due to software errors was 0.89\% for the retrospective analysis; that means that the accuracy for the detection of the ovulation was 99.11\%. Accuracy of 88.8\% for a window of 3 days before ovulation, the day of ovulation and the 3 days after ovulation was achieved for the prospective analysis. In the subgroup of woman with recorded pregnancies, it could be shown that after 3.79 months of use (median) pregnancies were observed. In 67.72\% in up to 3 months, in 16.36\% between 3 and 6 months of use, in 7.27\% between 7 and 9 months, in 5.45\% between 10 and 12 months and in 1.82\% between 13 and 15 months of use of the system. With this new web-based system, a precise determination of the fertile window even in women with ultralong cycles ({\textgreater}35 days) could be detected independently of their personal live circumstances. Exact determination of the fertile window is herewith possible so that {OvulaRing} represents an evolution in the {FAB} method for the cycle diagnosis of women with regular, irregular or anovulatory menstrual cycles.}, + pages = {256--260}, number = {3}, - urldate = {2025-05-27}, - journal = {Fertility and Sterility}, - author = {Garcia, Jairo E. and Seegar Jones, Georgeanna and Wright, George L.}, - month = sep, - year = {1981}, - pages = {308--315}, - file = {PDF:/home/alex/Zotero/storage/DB8PW3QR/Garcia et al. - 1981 - Prediction of the Time of Ovulation.pdf:application/pdf;ScienceDirect Snapshot:/home/alex/Zotero/storage/QLFLZFDJ/S0015028216457304.html:text/html}, + journaltitle = {Gynecological Endocrinology}, + author = {Regidor, Pedro-Antonio and Kaczmarczyk, Marta and Schiweck, Esther and Goeckenjan-Festag, Maren and Alexander, Henry}, + urldate = {2025-02-21}, + date = {2018-03-04}, + pmid = {29082805}, + note = {Publisher: Taylor \& Francis +\_eprint: https://doi.org/10.1080/09513590.2017.1390737}, + keywords = {central nervous system, circadian rhythm, circamensual rhythm, core body temperature, fertile window, Infertility, vagina}, + file = {Full Text PDF:/home/alex/Zotero/storage/ITD68HTW/Regidor et al. - 2018 - Identification and prediction of the fertile window with a new web-based medical device using a vagi.pdf:application/pdf}, } -@article{b_s_novel_2022, - title = {Novel {Technique} for {Confirmation} of the {Day} of {Ovulation} and {Prediction} of {Ovulation} in {Subsequent} {Cycles} {Using} a {Skin}-{Worn} {Sensor} in a {Population} {With} {Ovulatory} {Dysfunction}: {A} {Side}-by-{Side} {Comparison} {With} {Existing} {Basal} {Body} {Temperature} {Algorithm} and {Vaginal} {Core} {Body} {Temperature} {Algorithm}}, +@article{alexander_fertilitatsmonitoring_2014, + title = {Fertilitätsmonitoring mit vaginalem Biosensor ({OvulaRing}©)}, + volume = {74}, + issn = {0016-5751}, + url = {https://www.thieme-connect.com/products/ejournals/abstract/10.1055/s-0034-1388603}, + doi = {10.1055/s-0034-1388603}, + abstract = {Thieme E-Books \& E-Journals}, + pages = {FV\_08\_05}, + journaltitle = {Geburtshilfe und Frauenheilkunde}, + author = {Alexander, H. and Kaczmarczyk, M. and Pretzsch, G. and Kersken, T. and Puschmann, D. and Schiwek, E. and Goeckenjan, M.}, + urldate = {2025-02-21}, + date = {2014-09-05}, + langid = {german}, + keywords = {60. Kongress der Deutschen Gesellschaft für Gynäkologie und Geburtshilfe}, + file = {Snapshot:/home/alex/Zotero/storage/HPL6XYJW/s-0034-1388603.html:text/html}, +} + +@inproceedings{regidor_identifizierung_2018, + title = {Identifizierung und Vorhersage des fertilen Fensters des weiblichen Zyklus mit einem neuen web basierten Medizinprodukt ({OvulaRing}®).}, + volume = {78}, + rights = {Georg Thieme Verlag {KG} Stuttgart · New York}, + url = {https://www.thieme-connect.com/products/ejournals/html/10.1055/s-0038-1671278}, + doi = {10.1055/s-0038-1671278}, + abstract = {Thieme E-Books \& E-Journals}, + pages = {P 23}, + booktitle = {Geburtshilfe und Frauenheilkunde}, + publisher = {Georg Thieme Verlag {KG}}, + author = {Regidor, P. A. and Alexander, H.}, + urldate = {2025-02-21}, + date = {2018-09-20}, + langid = {german}, + note = {{ISSN}: 0016-5751}, + keywords = {Präsidentin der {DGGG} e.V.: Prof. Dr. Birgit Seelbach-Göbel{\textless}/conf-president{\textgreater}{\textless}/conference{\textgreater}}, + file = {Snapshot:/home/alex/Zotero/storage/DW6578ZM/s-0038-1671278.html:text/html}, +} + +@article{sato_novel_2024, + title = {Novel Methodology for Identifying the Occurrence of Ovulation by Estimating Core Body Temperature During Sleeping: Validity and Effectiveness Study}, + volume = {8}, + rights = {Unless stated otherwise, all articles are open-access distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work ("first published in the Journal of Medical Internet Research...") is properly cited with original {URL} and bibliographic citation information. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.}, + url = {https://formative.jmir.org/2024/1/e55834}, + doi = {10.2196/55834}, + shorttitle = {Novel Methodology for Identifying the Occurrence of Ovulation by Estimating Core Body Temperature During Sleeping}, + abstract = {Background: Body temperature is the most-used noninvasive biomarker to determine menstrual cycle and ovulation. However, issues related to its low accuracy are still under discussion. Objective: This study aimed to improve the accuracy of identifying the presence or absence of ovulation within a menstrual cycle. We investigated whether core body temperature ({CBT}) estimation can improve the accuracy of temperature biphasic shift discrimination in the menstrual cycle. The study consisted of 2 parts: experiment 1 assessed the validity of the {CBT} estimation method, while experiment 2 focused on the effectiveness of the method in discriminating biphasic temperature shifts. Methods: In experiment 1, healthy women aged between 18 and 40 years had their true {CBT} measured using an ingestible thermometer and their {CBT} estimated from skin temperature and ambient temperature measured during sleep in both the follicular and luteal phases of their menstrual cycles. This study analyzed the differences between these 2 measurements, the variations in temperature between the 2 phases, and the repeated measures correlation between the true and estimated {CBT}. Experiment 2 followed a similar methodology, but focused on evaluating the diagnostic accuracy of these 2 temperature measurement approaches (estimated {CBT} and traditional oral basal body temperature [{BBT}]) for identifying ovulatory cycles. This was performed using urine luteinizing hormone ({LH}) as the reference standard. Menstrual cycles were categorized based on the results of the {LH} tests, and a temperature shift was identified using a specific criterion called the “three-over-six rule.” This rule and the nested design of the study facilitated the assessment of diagnostic measures, such as sensitivity and specificity. Results: The main findings showed that {CBT} estimated from skin temperature and ambient temperature during sleep was consistently lower than directly measured {CBT} in both the follicular and luteal phases of the menstrual cycle. Despite this, the pattern of temperature variation between these phases was comparable for both the estimated and true {CBT} measurements, suggesting that the estimated {CBT} accurately reflected the cyclical variations in the true {CBT}. Significantly, the {CBT} estimation method showed higher sensitivity and specificity for detecting the occurrence of ovulation than traditional oral {BBT} measurements, highlighting its potential as an effective tool for reproductive health monitoring. The current method for estimating the {CBT} provides a practical and noninvasive method for monitoring {CBT}, which is essential for identifying biphasic shifts in the {BBT} throughout the menstrual cycle. Conclusions: This study demonstrated that the estimated {CBT} derived from skin temperature and ambient temperature during sleep accurately captures variations in true {CBT} and is more accurate in determining the presence or absence of ovulation than traditional oral {BBT} measurements. This method holds promise for improving reproductive health monitoring and understanding of menstrual cycle dynamics.}, + pages = {e55834}, + number = {1}, + journaltitle = {{JMIR} Formative Research}, + author = {Sato, Daisuke and Ikarashi, Koyuki and Nakajima, Fumiko and Fujimoto, Tomomi}, + urldate = {2025-02-20}, + date = {2024-07-05}, + note = {Company: {JMIR} Formative Research +Distributor: {JMIR} Formative Research +Institution: {JMIR} Formative Research +Label: {JMIR} Formative Research +Publisher: {JMIR} Publications Inc., Toronto, Canada}, + file = {PubMed Central Full Text PDF:/home/alex/Zotero/storage/NPBA84BU/Sato et al. - 2024 - Novel Methodology for Identifying the Occurrence of Ovulation by Estimating Core Body Temperature Du.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/BAJUVPFE/e55834.html:text/html}, +} + +@article{royston_identifying_1991, + title = {Identifying the fertile phase of the human menstrual cycle}, volume = {10}, - issn = {2296-4185}, - shorttitle = {Novel {Technique} for {Confirmation} of the {Day} of {Ovulation} and {Prediction} of {Ovulation} in {Subsequent} {Cycles} {Using} a {Skin}-{Worn} {Sensor} in a {Population} {With} {Ovulatory} {Dysfunction}}, - url = {https://www.frontiersin.org/articles/10.3389/fbioe.2022.807139/full}, - doi = {10.3389/fbioe.2022.807139}, - abstract = {Objective: Determine the accuracy of a novel technique for confirmation of the day of ovulation and prediction of ovulation in subsequent cycles for the purpose of conception using a skin-worn sensor in a population with ovulatory dysfunction. -Methods: A total of 80 participants recorded consecutive overnight temperatures using a skin-worn sensor at the same time as a commercially available vaginal sensor for a total of 205 reproductive cycles. The vaginal sensor and its associated algorithm were used to determine the day of ovulation, and the ovulation results obtained using the skin-worn sensor and its associated algorithm were assessed for comparative accuracy alongside a number of other statistical techniques, with a further assessment of the same skin-derived data by means of the “three over six” rule. A number of parameters were used to divide the data into separate comparative groups, and further secondary statistical analyses were performed. -Results: The skin-worn sensor and its associated algorithm (together labeled “SWS”) were 66\% accurate for determining the day of ovulation (±1 day) or the absence of ovulation and 90\% accurate for determining the fertile window (ovulation day ±3 days) in the total study population in comparison to the results obtained from the vaginal sensor and its associated algorithm (together labeled “VS”). -Conclusion: SWS is a useful tool for confirming the fertile window and absence of ovulation (anovulation) in a population with ovulatory dysfunction, both known and Edited by:}, - language = {en}, - urldate = {2025-05-27}, - journal = {Frontiers in Bioengineering and Biotechnology}, - author = {B. S., Hurst and K., Davies and R. C., Milnes and T. G., Knowles and A., Pirrie}, - month = mar, - year = {2022}, - pages = {807139}, - file = {PDF:/home/alex/Zotero/storage/U8UPIT4Q/B. S. et al. - 2022 - Novel Technique for Confirmation of the Day of Ovulation and Prediction of Ovulation in Subsequent C.pdf:application/pdf}, + rights = {Copyright © 1991 John Wiley \& Sons, Ltd.}, + issn = {1097-0258}, + url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/sim.4780100207}, + doi = {10.1002/sim.4780100207}, + abstract = {The identification of the human fertile phase as the time during which a woman or a couple may conceive is elusive. The fertile time depends on many factors in each individual menstrual cycle and may be said to be more of a statistical than a physiological entity. This paper reviews the application of statistical methods to three areas related to conception and the fertile phase. The first is the prediction and detection of ovulation from serial measurements, such as hormones, basal body temperature and cervical mucus, throughout the menstrual cycle. Typically, such variables increase from some baseline level to a peak around ovulation (the most fertile time), then subside to low levels in the postovulatory phase. The statistical challenge is to detect the rise (signalling the onset of potential fertility) and subsequent fall. Analytic methods considered include thresholds, Bayesian change-point models and particularly the cumulative sum (cusum) technique which is both simple to apply and understand, and effective. The second area comprises appropriate methods of analysing and interpreting data from clinical studies of the fertile phase, especially in so-alled natural family planning ({NFP}) where it is usual for women to observe several indices of potential fertility. Such studies usually try to establish the temporal relationships between markers of the fertile phase and examine the success of different combinations of markers in delineating the fertile time in comparison with a standard ‘defined’ phase, for example, the interval from three days before to two days after the peak of luteinizing hormone. The third area is the assessment of the probability of conception on certain days of the cycle, which is vital to the understanding of the fertile phase and its application to {NFP}. Direct estimation of such probabilities is impractical; instead, resort must be made to estimation by maximum likelihood of the parameters of specially constructed models. Suitable models are described. Finally, the need for a new prospective study of the probability of conception in relation to the markers of the fertile phase used in the symptothermal method of {NFP} is discussed.}, + pages = {221--240}, + number = {2}, + journaltitle = {Statistics in Medicine}, + author = {Royston, Patrick}, + urldate = {2025-02-20}, + date = {1991}, + langid = {english}, + note = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/sim.4780100207}, + file = {Snapshot:/home/alex/Zotero/storage/9XL95LUJ/sim.html:text/html}, } -@article{salles_softed_2024, - title = {{SoftED}: {Metrics} for soft evaluation of time series event detection}, - volume = {198}, - issn = {03608352}, - shorttitle = {{SoftED}}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0360835224008507}, - doi = {10.1016/j.cie.2024.110728}, - abstract = {Time series event detectors are evaluated mainly by standard classification metrics, focusing solely on detection accuracy. However, inaccuracy in detecting an event can often result from its preceding or delayed effects reflected in neighboring detections. These detections are valuable to trigger necessary actions or help mitigate unwelcome consequences. In this context, current metrics are insufficient and inadequate for the context of event detection. There is a demand for metrics that incorporate both the concept of time and temporal tolerance for neighboring detections. Inspired by fuzzy sets, this paper introduces SoftED metrics, a new set designed for soft evaluating event detectors. They enable the evaluation of the detection accuracy and the degree to which their detections represent events. A new general protocol inspired by competency questions is also introduced to evaluate temporal tolerant metrics for event detection. The SoftED metrics can improve event detection evaluations by associating events and their representative detections, incorporating temporal tolerance in over 36\% of the overall detector evaluations compared to the usual classification metrics. Following the proposed evaluation protocol, SoftED metrics were evaluated by domain specialists who indicated their contribution to detection evaluation and method selection.}, - language = {en}, - urldate = {2025-06-03}, - journal = {Computers \& Industrial Engineering}, - author = {Salles, Rebecca and Lima, Janio and Reis, Michel and Coutinho, Rafaelli and Pacitti, Esther and Masseglia, Florent and Akbarinia, Reza and Chen, Chao and Garibaldi, Jonathan and Porto, Fabio and Ogasawara, Eduardo}, - month = dec, - year = {2024}, - pages = {110728}, - file = {PDF:/home/alex/Zotero/storage/6RYA5HIQ/Salles et al. - 2024 - SoftED Metrics for soft evaluation of time series event detection.pdf:application/pdf}, +@article{su_detection_2017, + title = {Detection of ovulation, a review of currently available methods}, + volume = {2}, + rights = {© 2017 The Authors. Bioengineering \& Translational Medicine is published by Wiley Periodicals, Inc. on behalf of The American Institute of Chemical Engineers}, + issn = {2380-6761}, + url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/btm2.10058}, + doi = {10.1002/btm2.10058}, + abstract = {The ability to identify the precise time of ovulation is important for women who want to plan conception or practice contraception. Here, we review the current literature on various methods for detecting ovulation including a review of point-of-care device technology. We incorporate an examination of methods to detect ovulation that have been developed and practiced for decades and analyze the indications and limitations of each—transvaginal ultrasonography, urinary luteinizing hormone detection, serum progesterone and urinary pregnanediol 3-glucuronide detection, urinary follicular stimulating hormone detection, basal body temperature monitoring, and cervical mucus and salivary ferning analysis. Some point-of-care ovulation detection devices have been developed and commercialized based on these methods, however previous research was limited by small sample size and an inconsistent standard reference to true ovulation.}, + pages = {238--246}, + number = {3}, + journaltitle = {Bioengineering \& Translational Medicine}, + author = {Su, Hsiu-Wei and Yi, Yu-Chiao and Wei, Ting-Yen and Chang, Ting-Chang and Cheng, Chao-Min}, + urldate = {2025-02-20}, + date = {2017}, + langid = {english}, + note = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/btm2.10058}, + keywords = {family planning, fertility window, ovulation detection}, + file = {Full Text PDF:/home/alex/Zotero/storage/ZEACCGE5/Su et al. - 2017 - Detection of ovulation, a review of currently available methods.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/RDDQD8EA/btm2.html:text/html}, +} + +@article{owen_physiological_2013, + title = {Physiological Signs of Ovulation and Fertility Readily Observable by Women}, + volume = {80}, + issn = {0024-3639}, + url = {https://doi.org/10.1179/0024363912Z.0000000005}, + doi = {10.1179/0024363912Z.0000000005}, + abstract = {{IntroductionConfirmation} of ovulation can be difficult in clinical practice, as gold standard methods including serial transvaginal ultrasonography, serum luteinizing hormone ({LH}) measurements, or laparoscopic follicular observation are impractical. Numerous surrogate markers have been proposed and evaluated in relation to these gold standards that have more practical clinical applications.{PurposeTo} review the evidence on physiological signs of ovulation timing and fertility in order to determine valid markers that can be easily identified by women.{MethodsA} literature review of primary resources in Ovid Medline was undertaken to identify studies examining physiological signs as they relate to gold standard assessment of ovulation. Studies examining the efficacy/effectiveness of different types of natural family planning were excluded.{ResultsThe} most commonly encountered physiological signs were urine {LH}, cervical mucus, and basal body temperature ({BBT}). Urine {LH} as assessed by home monitoring systems indicated ovulation 91 percent of the time during the 2 days of peak fertility on the monitor and 97 percent during the 2 peak days plus 1. Cervical mucus peak characteristics were identified 78 percent of the time ±1 day, and 91 percent of the time ±2 days of {LH} surge indicating ovulation. Further research supports the importance of cervical mucus in overall fertility, as conception rates were more closely related to mucus quality than to timing of intercourse related to ovulation. As a lone indicator of ovulation, {BBT} is at best a retrospective marker, and functions best in conjunction with other signs of ovulation. Additionally, salivary ferning, salivary and vaginal fluid electrical potential, finger–finger electrical potential, and differential skin temperature were postulated as possible indicators, but were not found to be temporally related to ovulation. The research on differential skin temperature is promising, but minimal thus far in number, and has not been evaluated as an adjunct to {BBT} as yet.{ConclusionHome} urinary {LH} monitors are becoming more widely available and less expensive giving women the potential to assess the ovulatory status of their cycle in real time. Cervical mucus observation is an effective and cost-efficient method, but requires some teaching to increase the confidence of users. In conjunction, {LH} monitors and cervical mucus can give the best indication of fertility and ovulation timing.}, + pages = {17--23}, + number = {1}, + journaltitle = {Linacre Q}, + author = {Owen, Martin}, + urldate = {2025-02-20}, + date = {2013-01-01}, + langid = {english}, + note = {Publisher: {SAGE} Publications Inc}, + file = {Full Text:/home/alex/Zotero/storage/IBVUICCU/Owen - 2013 - Physiological Signs of Ovulation and Fertility Readily Observable by Women.pdf:application/pdf}, +} + +@article{soumpasis_real-life_2020, + title = {Real-life insights on menstrual cycles and ovulation using big data}, + volume = {2020}, + issn = {2399-3529}, + url = {https://doi.org/10.1093/hropen/hoaa011}, + doi = {10.1093/hropen/hoaa011}, + abstract = {What variations underlie the menstrual cycle length and ovulation day of women trying to conceive?Big data from a connected ovulation test revealed the extent of variation in menstrual cycle length and ovulation day in women trying to conceive.Timing intercourse to coincide with the fertile period of a woman maximises the chances of conception. The day of ovulation varies on an inter- and intra-individual level.A total of 32 595 women who had purchased a connected ovulation test system contributed 75 981 cycles for analysis. Day of ovulation was determined from the fertility test results. The connected home ovulation test system enables users to identify their fertile phase. The app benefits users by enabling them to understand their personal fertility information. During each menstrual cycle, users input their perceived cycle length into an accessory application, and data on hormone levels from the tests are uploaded to the application and stored in an anonymised cloud database. This study compared users’ perceived cycle characteristics with actual cycle characteristics. The perceived and actual cycle length information was analysed to provide population ranges.This study analysed data from the at-home use of a commercially available connected home ovulation test by women across the {USA} and {UK}.Overall, 25.3\% of users selected a 28-day cycle as their perceived cycle length; however, only 12.4\% of users actually had a 28-day cycle. Most women (87\%) had actual menstrual cycle lengths between 23 and 35 days, with a normal distribution centred on day 28, and over half of the users (52\%) had cycles that varied by 5 days or more. There was a 10-day spread of observed ovulation days for a 28-day cycle, with the most common day of ovulation being Day 15. Similar variation was observed for all cycle lengths examined. For users who conducted a test on every day requested by the app, a luteinising hormone ({LH}) surge was detected in 97.9\% of cycles.Data were from a self-selected population of women who were prepared to purchase a commercially available product to aid conception and so may not fully represent the wider population. No corresponding demographic data were collected with the cycle information.Using big data has provided more personalised insights into women’s fertility; this could enable women trying to conceive to better time intercourse, increasing the likelihood of conception.The study was funded by {SPD} Development Company Ltd (Bedford, {UK}), a fully owned subsidiary of {SPD} Swiss Precision Diagnostics {GmbH} (Geneva, Switzerland). I.S., B.G. and S.J. are employees of the {SPD} Development Company Ltd.}, + pages = {hoaa011}, + number = {2}, + journaltitle = {Human Reproduction Open}, + author = {Soumpasis, I and Grace, B and Johnson, S}, + urldate = {2025-02-20}, + date = {2020-02-01}, + file = {Full Text PDF:/home/alex/Zotero/storage/PS9UC298/Soumpasis et al. - 2020 - Real-life insights on menstrual cycles and ovulation using big data.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/LTFATQIC/5820371.html:text/html}, +} + +@article{brewis_demographic_2005, + title = {Demographic Evidence That Human Ovulation Is Undetectable (At Least in Pair Bonds)}, + volume = {46}, + issn = {0011-3204}, + url = {https://www.journals.uchicago.edu/doi/abs/10.1086/430016}, + doi = {10.1086/430016}, + pages = {465--471}, + number = {3}, + journaltitle = {Current Anthropology}, + author = {Brewis, Alexandra and Meyer, Mary}, + urldate = {2025-02-20}, + date = {2005-06}, + note = {Publisher: The University of Chicago Press}, +} + +@article{noauthor_monitoring_1987, + title = {Monitoring techniques to predict and detect ovulation}, + volume = {47}, + issn = {0015-0282}, + url = {https://www.sciencedirect.com/science/article/pii/S0015028216500028}, + doi = {10.1016/S0015-0282(16)50002-8}, + abstract = {This study was designed to evaluate the accuracy of various methods in predicting and detecting ovulation in 14 spontaneous and 17 clomiphene citrate …}, + pages = {259--264}, + number = {2}, + journaltitle = {Fertility and Sterility}, + urldate = {2025-02-20}, + date = {1987-02-01}, + langid = {american}, + note = {Publisher: Elsevier}, + file = {PDF:/home/alex/Zotero/storage/RLCZVH5V/1987 - Monitoring techniques to predict and detect ovulation.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/M5P9EZ67/S0015028216500028.html:text/html}, +} + +@article{noauthor_physiological_2016, + title = {Physiological predictors of ovulation and pregnancy risk in a fixed-time artificial insemination program}, + volume = {99}, + issn = {0022-0302}, + url = {https://www.sciencedirect.com/science/article/pii/S0022030216306725}, + doi = {10.3168/jds.2016-11247}, + abstract = {The objective of this study was to determine the relative importance and contribution of several physiological factors as predictors of pregnancy risk…}, + pages = {10077--10092}, + number = {12}, + journaltitle = {Journal of Dairy Science}, + urldate = {2025-02-20}, + date = {2016-12-01}, + langid = {american}, + note = {Publisher: Elsevier}, + file = {Snapshot:/home/alex/Zotero/storage/EBK9WJP4/S0022030216306725.html:text/html}, +} + +@inproceedings{serrano_is_2019-1, + location = {Florence, Italy}, + title = {Is Attention Interpretable?}, + url = {https://aclanthology.org/P19-1282/}, + doi = {10.18653/v1/P19-1282}, + abstract = {Attention mechanisms have recently boosted performance on a range of {NLP} tasks. Because attention layers explicitly weight input components' representations, it is also often assumed that attention can be used to identify information that models found important (e.g., specific contextualized word tokens). We test whether that assumption holds by manipulating attention weights in already-trained text classification models and analyzing the resulting differences in their predictions. While we observe some ways in which higher attention weights correlate with greater impact on model predictions, we also find many ways in which this does not hold, i.e., where gradient-based rankings of attention weights better predict their effects than their magnitudes. We conclude that while attention noisily predicts input components' overall importance to a model, it is by no means a fail-safe indicator.}, + eventtitle = {{ACL} 2019}, + pages = {2931--2951}, + booktitle = {Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics}, + publisher = {Association for Computational Linguistics}, + author = {Serrano, Sofia and Smith, Noah A.}, + editor = {Korhonen, Anna and Traum, David and Màrquez, Lluís}, + urldate = {2025-02-19}, + date = {2019-07}, + file = {Full Text PDF:/home/alex/Zotero/storage/I6J6YP3C/Serrano and Smith - 2019 - Is Attention Interpretable.pdf:application/pdf}, +} + +@article{noauthor_attention-based_2023, + title = {An attention-based deep learning model for multi-horizon time series forecasting by considering periodic characteristic}, + volume = {185}, + issn = {0360-8352}, + url = {https://www.sciencedirect.com/science/article/abs/pii/S0360835223006915}, + doi = {10.1016/j.cie.2023.109667}, + abstract = {Recently, transformer-based models have exhibited great performance in multi-horizon time series forecasting tasks. However, the core module of these …}, + pages = {109667}, + journaltitle = {Computers \& Industrial Engineering}, + urldate = {2025-02-19}, + date = {2023-11-01}, + langid = {american}, + note = {Publisher: Pergamon}, + file = {Snapshot:/home/alex/Zotero/storage/TJ634T5V/S0360835223006915.html:text/html}, +} + +@article{hu_pattern-oriented_2025, + title = {Pattern-oriented Attention Mechanism for Multivariate Time Series Forecasting}, + volume = {19}, + issn = {1556-4681}, + url = {https://doi.org/10.1145/3712606}, + doi = {10.1145/3712606}, + abstract = {Multivariate time series forecasting is applied in many domains, such as finance, transportation, and industry. The main challenge of precise forecasting lies in accurately capturing latent dependencies. Recent studies develop various frameworks to reduce computational complexity or to enhance the learning of intricate relationships, while lacking interpretability and generality. In this article, we aim to elucidate the capture of dependencies as the recognition of patterns. We believe that patterns can be formally described from two aspects: the shapes of segments that frequently repeat and the corresponding forms of repetitions. Drawing upon this idea, we design a multivariate time series forecasting model named {PRformer},1 which incorporates a pattern-oriented attention mechanism and a pattern-based projector. The attention mechanism can perceive different forms of repetitions by embedded with various similarity evaluation metrics between segments, and filter out noise from segments to extract potential patterns with a statistical-driven weighting scheme. The pattern-based projector is employed to form the forecasting results by deriving the representative patterns from the set of potential ones. By incorporating explicit definitions of patterns, {PRformer} is interpretable and general to various time series scenarios. Experimental results on seven datasets demonstrate that {PRformer} outperforms six state-of-the-art models by about 10.7\% in forecasting accuracy.}, + pages = {38:1--38:26}, + number = {2}, + journaltitle = {{ACM} Trans. Knowl. Discov. Data}, + author = {Hu, Hanwen and Han, Zhangchi and Qian, Shiyou and Yang, Dingyu and Cao, Jian and Xue, Guangtao}, + urldate = {2025-02-19}, + date = {2025-02-06}, +} + +@article{braude_machine_2024, + title = {Machine learning for predicting elective fertility preservation outcomes}, + volume = {14}, + rights = {2024 The Author(s)}, + issn = {2045-2322}, + url = {https://www.nature.com/articles/s41598-024-60671-w}, + doi = {10.1038/s41598-024-60671-w}, + abstract = {This retrospective study applied machine-learning models to predict treatment outcomes of women undergoing elective fertility preservation. Two-hundred-fifty women who underwent elective fertility preservation at a tertiary center, 2019–2022 were included. Primary outcome was the number of metaphase {II} oocytes retrieved. Outcome class was based on oocyte count ({OC}): Low (≤ 8), Medium (9–15) or High (≥ 16). Machine-learning models and statistical regression were used to predict outcome class, first based on pre-treatment parameters, and then using post-treatment data from ovulation-triggering day. {OC} was 136 Low, 80 Medium, and 34 High. Random Forest Classifier ({RFC}) was the most accurate model (pre-treatment receiver operating characteristic ({ROC}) area under the curve ({AUC}) was 77\%, and post-treatment {ROC} {AUC} was 87\%), followed by {XGBoost} Classifier (pre-treatment {ROC} {AUC} 74\%, post-treatment {ROC} {AUC} 86\%). The most important pre-treatment parameters for {RFC} were basal {FSH} (22.6\%), basal {LH} (19.1\%), {AFC} (18.2\%), and basal estradiol (15.6\%). Post-treatment parameters were estradiol levels on trigger-day (17.7\%), basal {FSH} (11\%), basal {LH} (9\%), and {AFC} (8\%). Machine-learning models trained with clinical data appear to predict fertility preservation treatment outcomes with relatively high accuracy.}, + pages = {10158}, + number = {1}, + journaltitle = {Sci Rep}, + author = {Braude, Itai and Haikin Herzberger, Einat and Semo, Mor and Soifer, Kim and Goren Gepstein, Nitzan and Wiser, Amir and Miller, Netanella}, + urldate = {2025-02-11}, + date = {2024-05-02}, + langid = {english}, + note = {Publisher: Nature Publishing Group}, + keywords = {Computational models, Outcomes research}, + file = {Full Text PDF:/home/alex/Zotero/storage/URDGBHLV/Braude et al. - 2024 - Machine learning for predicting elective fertility preservation outcomes.pdf:application/pdf}, +} + +@article{fanton_interpretable_2022, + title = {An interpretable machine learning model for predicting the optimal day of trigger during ovarian stimulation}, + volume = {118}, + issn = {0015-0282, 1556-5653}, + url = {https://www.fertstert.org/article/S0015-0282%2822%2900244-8/fulltext}, + doi = {10.1016/j.fertnstert.2022.04.003}, + pages = {101--108}, + number = {1}, + journaltitle = {Fertility and Sterility}, + author = {Fanton, Michael and Nutting, Veronica and Solano, Funmi and Maeder-York, Paxton and Hariton, Eduardo and Barash, Oleksii and Weckstein, Louis and Sakkas, Denny and Copperman, Alan B. and Loewke, Kevin}, + urldate = {2025-02-11}, + date = {2022-07-01}, + note = {Publisher: Elsevier}, + keywords = {Artificial intelligence, in vitro fertilization, machine learning, ovarian stimulation, trigger}, + file = {Full Text PDF:/home/alex/Zotero/storage/UNX7ASLP/Fanton et al. - 2022 - An interpretable machine learning model for predicting the optimal day of trigger during ovarian sti.pdf:application/pdf}, +} + +@inproceedings{azaria_semi-supervised_2019, + title = {Semi-Supervised Ovulation Detection Based on Multiple Properties}, + url = {https://ieeexplore.ieee.org/document/8995235}, + doi = {10.1109/ICTAI.2019.00039}, + abstract = {Despite being a well-researched problem, ovulation detection in human female remains a difficult task. Most current methods for ovulation detection rely on measurements of a single property (e.g. morning body temperature) or at most on two properties (e.g. both salivary and vaginal electrical resistance). In this paper we present a machine learning based method for detecting the day in which ovulation occurs. Our method considered measurements of five different properties. We crawled a data-set from the web and showed that our method outperforms current state-of-the-art methods for ovulation detection. Our method performs well also when considering measurements of fewer properties. We show that our method's performance can be further improved by using unlabeled data, that is, mensuration cycles without a know ovulation date. Our resulted machine learning model can be very useful for women trying to conceive that have trouble in recognizing their ovulation period, especially when some measurements are missing.}, + eventtitle = {2019 {IEEE} 31st International Conference on Tools with Artificial Intelligence ({ICTAI})}, + pages = {222--228}, + booktitle = {2019 {IEEE} 31st International Conference on Tools with Artificial Intelligence ({ICTAI})}, + author = {Azaria, Amos and Azaria, Seagal}, + urldate = {2025-02-11}, + date = {2019-11}, + note = {{ISSN}: 2375-0197}, + keywords = {ovulation detection, semi supervised learning}, + file = {IEEE Xplore Abstract Record:/home/alex/Zotero/storage/JSAEWRGB/8995235.html:text/html;PDF:/home/alex/Zotero/storage/P74SEG9S/Azaria and Azaria - 2019 - Semi-Supervised Ovulation Detection Based on Multiple Properties.pdf:application/pdf}, +} + +@article{luz_p-656_2023, + title = {P-656 Machine learning algorithm automatically manages and accurately predicts ovulation in natural frozen-thawed embryo transfer cycles.}, + volume = {38}, + issn = {0268-1161}, + url = {https://doi.org/10.1093/humrep/dead093.983}, + doi = {10.1093/humrep/dead093.983}, + abstract = {Can an Artificial Intelligence ({AI}) algorithm automatically manage frozen-thawed embryo transfer ({NC}-{FET}) treatment cycles and give an accurate prediction of ovulation day.An {AI} algorithm automatically managed and predicted the ovulation of {NC}-{FET} treatment cycles with 94.8\% accuracy using an average of 3.01 test days.Today the preferred method for frozen embryo transfer is natural cycle based on ovulation detection. Currently, there is no software capable of managing the treatment cycle automatically and identifying the time of ovulation to support doctor decisions. The aim of this study is to develop a physician support {AI} software for determining ovulation time reliably with high accuracy.2083 {NC}-{FET} cycles from September 2018 to June 2021 were used to develop the ovulation detection and treatment management algorithms.Each cycle had data from at least 2 visits including: hormonal levels (Estrogen/Progesterone/{LH}) and follicle sizes.The dataset was divided into a train set and two test sets. In the 1st test set ovulation was determined by experts’ opinions and the 2nd test set included cycles in which follicle rupture was documented in consecutive ultrasounds.Two algorithms were developed, an ovulation prediction algorithm based on an {NGBoost} model and a treatment management algorithm that used the model to determine if and when to call for a new test or declare the ovulation day.Both algorithms were jointly tuned to reach the highest success rate, defined as providing the correct day of ovulation using the available cycle data, with as few test days as possible.On the first test set, which consisted of 176 cycles in which ovulation was determined through the majority decision of 2 independent experts and the attending physician, the treatment management algorithm required on average 3.01 tests to reach a prediction and successfully predicted the ovulation day in 94.8\% of cycles.In the second test set, which consisted of 29 cycles in which ovulation was determined through the follicular rupture in two consecutive ultrasounds, only the ovulation prediction model was tested. To ensure that the model provides a reliable answer and does not rely solely on the follicular disappearances, examined cycles were tested twice: Once using the ovulation day without the day prior to it, and again using only the day prior to ovulation without the ovulation day itself. The algorithm accurately predicted ovulation in 28 out of 29 instances (96.6\%) using the day of ovulation and in 28 out of 29 instances (96.6\%) using the day before ovulation.The main drawback is this being a retrospective study: while the algorithm was trained to maximize accuracy when it selects the test days, the dataset test days were selected by the attending physicians. Statistical methods were used to overcome this, however further prospective trials are needed to validate the results.This is the first {AI} algorithm designed to automatically manage {NC}-{FET} {IVF} treatment cycles and predict ovulation. The high accuracy and low average tests count might improve treatment outcomes, reduce the patients’ life disruption, and allow physicians to spend less time monitoring their patients’ treatments.not applicable}, + pages = {dead093.983}, + issue = {Supplement\_1}, + journaltitle = {Human Reproduction}, + author = {Luz, A and Hourvitz, R and Reuvenny, S and Youngster, M and Baum, M and Hourvitz, A and Maman, E}, + urldate = {2025-02-11}, + date = {2023-06-01}, + file = {Full Text PDF:/home/alex/Zotero/storage/C8YM765Y/Luz et al. - 2023 - P-656 Machine learning algorithm automatically manages and accurately predicts ovulation in natural.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/WANIZX3H/7202977.html:text/html}, +} + +@article{lin_transformer_2023, + title = {Transformer neural network to predict and interpret pregnancy loss from activity data in Holstein dairy cows}, + volume = {205}, + issn = {0168-1699}, + url = {https://www.sciencedirect.com/science/article/pii/S0168169923000261}, + doi = {10.1016/j.compag.2023.107638}, + abstract = {Predicting/detecting pregnancy loss of dairy cows offers the opportunity to shorten the time interval between artificial inseminations. Although several methods of pregnancy detection are being practiced, models with accurate, timely and interpretable detection of pregnancy are still lacking. This study proposed a transformer neural network to predict the probability of pregnancy loss based on continuous activity data, which were collected from activity-monitoring tags attached to 185 Holstein cows from a commercial dairy farm in Cayuga County, {NY}, {USA}. Our best model achieved an average accuracy of 0.87, F1 score of 0.87, recall of 0.87 and specificity of 0.90 using 14-day time-series activity windows (90\% overlap) using 5-fold cross-validation, outperforming commonly used classic statistical learning and deep learning models for time-series data. The results indicated that our predictive model gave high probabilities of correctly detecting pregnancy loss prior to the increased activities and veterinary confirmation by transrectal ultrasound. In addition, our model interpretation aligned with the changes in the temporal activity levels, revealing that drastic fluctuations in time-series activity data contributed heavily to the final prediction. To the best of our knowledge, this is the first work on developing transformer models for the prediction of pregnancy loss in dairy cows. In addition to facilitating the development of future precision management on modern farms, our work potentiates an increase in the reproductive efficiency and profitability of dairy farms.}, + pages = {107638}, + journaltitle = {Computers and Electronics in Agriculture}, + author = {Lin, Dan and Kenéz, Ákos and {McArt}, Jessica A. A. and Li, Jun}, + urldate = {2025-02-11}, + date = {2023-02-01}, + keywords = {Dairy cow, Precision livestock farming, Pregnancy loss prediction, Time-series activity}, +} + +@article{shkodzik_innovative_2024, + title = {Innovative Approaches to Digital Health in Ovulation Detection: A Review of Current Methods and Emerging Technologies}, + volume = {42}, + issn = {1526-4564}, + doi = {10.1055/s-0044-1793829}, + shorttitle = {Innovative Approaches to Digital Health in Ovulation Detection}, + abstract = {Ovulation is a vital sign, as significant as body temperature, heart rate, respiratory rate, and blood pressure, in assessing overall health and identifying potential health issues. Ovulation is a key event of the menstrual cycle that provides insights into the hormonal and reproductive health aspects. Affected by the orchestra of hormones, namely thyroid, prolactin, and androgens, disruptions in ovulation can indicate endocrinological conditions and lead to gynecological problems, such as heavy menstrual bleeding, irregular periods, amenorrhea, dysmenorrhea, and difficulties in getting pregnant. Monitoring ovulation and detecting disruptions can aid in the early detection of health issues, extending beyond reproductive health concerns. It can help identify underlying causes of symptoms like excessive fatigue and abnormal hair growth. The integration of digital health technologies, such as mobile apps using machine learning algorithms, wearables tracking temperature, heart rate, breath rate, and sleep patterns, and devices measuring reproductive hormones in urine or saliva samples, offers a wealth of opportunities in family planning, early health issue diagnosis, treatment adjustment, and tracking menstrual cycles during assisted reproductive techniques. These advancements provide a comprehensive approach to health monitoring, addressing both reproductive and overall health concerns.}, + pages = {81--89}, + number = {2}, + journaltitle = {Semin Reprod Med}, + author = {Shkodzik, Katerina}, + date = {2024-06}, + pmid = {39572028}, + keywords = {Digital Health, Female, Humans, Mobile Applications, Ovulation, Ovulation Detection, Telemedicine, Wearable Electronic Devices}, +} + +@article{luz_improved_2024, + title = {Improved clinical pregnancy rates in natural frozen-thawed embryo transfer cycles with machine learning ovulation prediction: insights from a retrospective cohort study}, + volume = {14}, + rights = {2024 The Author(s)}, + issn = {2045-2322}, + url = {https://www.nature.com/articles/s41598-024-80356-8}, + doi = {10.1038/s41598-024-80356-8}, + shorttitle = {Improved clinical pregnancy rates in natural frozen-thawed embryo transfer cycles with machine learning ovulation prediction}, + abstract = {This study aims to develop physician support software for determining ovulation time and assess its impact on pregnancy outcomes in natural cycle frozen embryo transfers ({NC}-{FET}). To develop, assess, and validate an ovulation prediction model, three datasets were used: {REI} Ovulation Determination dataset (500 cycles) split into training (309), validation (90), and test (101) sets; the Documented Ovulation dataset (101 cycles) with confirmed ovulation (documented follicular rupture and {LH} surge); and the Clinical Pregnancy Rates dataset (515 {NC}-{FET} cycles), categorized into “Matched” and “Mismatched” based on alignment with the model’s ovulation determination. Pregnancy outcomes were compared between the groups. The ovulation prediction model exhibited 93.85\% and 92.89\% matching rates with the {REI} Ovulation Determination and Documented Ovulation datasets, respectively. In the Clinical Pregnancy Rates dataset, the Matched group (282 cycles) showed significantly higher clinical pregnancy rates than the Mismatched group (34.6\% vs. 25.9\%, p = 0.04) and similar results for patients under 37 (41.1\% vs. 30.7\%, p = 0.04). Logistic regression indicated lower pregnancy rates in Mismatched cases (odds ratio 0.67 for the general population, 0.63 for patients under 37). In conclusion, we introduce a highly accurate {AI} ovulation prediction model. Treatment cycles aligning with the model’s recommendations had significantly increased clinical pregnancy rates.}, + pages = {29451}, + number = {1}, + journaltitle = {Sci Rep}, + author = {Luz, Almog and Hourvitz, Ariel and Moran, Eden and Itzhak, Nevo and Reuvenny, Shachar and Hourvitz, Rohi and Youngster, Michal and Baum, Micha and Maman, Ettie}, + urldate = {2025-02-11}, + date = {2024-11-27}, + langid = {english}, + note = {Publisher: Nature Publishing Group}, + keywords = {Infertility, Outcomes research}, + file = {Full Text PDF:/home/alex/Zotero/storage/BSHNTIFD/Luz et al. - 2024 - Improved clinical pregnancy rates in natural frozen-thawed embryo transfer cycles with machine learn.pdf:application/pdf}, +} + +@article{yu_tracking_2022, + title = {Tracking of menstrual cycles and prediction of the fertile window via measurements of basal body temperature and heart rate as well as machine-learning algorithms}, + volume = {20}, + issn = {1477-7827}, + url = {https://doi.org/10.1186/s12958-022-00993-4}, + doi = {10.1186/s12958-022-00993-4}, + abstract = {Fertility awareness and menses prediction are important for improving fecundability and health management. Previous studies have used physiological parameters, such as basal body temperature ({BBT}) and heart rate ({HR}), to predict the fertile window and menses. However, their accuracy is far from satisfactory. Additionally, few researchers have examined irregular menstruators. Thus, we aimed to develop fertile window and menstruation prediction algorithms for both regular and irregular menstruators.}, + pages = {118}, + number = {1}, + journaltitle = {Reproductive Biology and Endocrinology}, + author = {Yu, Jia-Le and Su, Yun-Fei and Zhang, Chen and Jin, Li and Lin, Xian-Hua and Chen, Lu-Ting and Huang, He-Feng and Wu, Yan-Ting}, + urldate = {2025-02-11}, + date = {2022-08-13}, + keywords = {Basal body temperature, Fertile window, Heart rate, Machine learning, Menstrual cycle, Wearable device}, + file = {Full Text PDF:/home/alex/Zotero/storage/7Z8P97UF/Yu et al. - 2022 - Tracking of menstrual cycles and prediction of the fertile window via measurements of basal body tem.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/SMRBT6EG/s12958-022-00993-4.html:text/html}, +} + +@article{luo_prediction_2025, + title = {Prediction of the fertile window and menstrual cycles with a wearable device via machine-learning algorithms}, + issn = {1472-6483}, + url = {https://www.sciencedirect.com/science/article/pii/S1472648325000021}, + doi = {10.1016/j.rbmo.2025.104795}, + abstract = {Research question +We aimed to develop fertile window and menstruation prediction algorithms through machine learning based on women's physiological parameters data collected by Huawei Band 6 pro from both regular and irregular menstruators. +Design +This was a prospective observational cohort study conducted at Obstetrics and Gynecology Hospital of Fudan University. Participants were recruited from November 2021 to September 2022. Each participant wore Huawei Band 6 pro to record wrist skin temperature ({WST}), heart rate ({HR}), heart rate variability, and respiratory rate. Algorithms were developed to predict the fertile window and menstrual cycle based on {WST} and {HR}. +Results +We included data from 270 and 84 qualified cycles with confirmed ovulations from 136 regular and 47 irregular menstruators. For regular menstruators, the prediction algorithm based on {WST} and {HR} for the fertile window had an accuracy of 85.47\%, a sensitivity of 70.07\%, a specificity of 89.77\%, and {AUC} of 0.869. The algorithms for menstrual first day labelling and onset within 3 days gained an accuracy of 83.6\% and 75.0\%. For irregular menstruators, the accuracy, sensitivity, specificity and {AUC} were 79.85\%, 42.79\%, 87.28\%, and 0.763 respectively, for fertile window prediction. The accuracy of menses labelling and prediction were 61.2\%, and 50.8\% respectively. +Conclusions +Based on {WST} and {HR} data from the wearable device, the algorithms demonstrated reliable performance in predicting the fertile window and menstruation day among regular menstruators. These algorithms also showed potential for assisting irregular menstruators in managing their cycles and planning for conception.}, + pages = {104795}, + journaltitle = {Reproductive {BioMedicine} Online}, + author = {Luo, Chuan and Su, Yun-Fei and Ren, Yun-Yun and Zhang, Qin and Li, Ran and Zhang, Qi and Li, Cheng and Hao, Yan-Hui and Zhang, An-Qi and Zhang, Hao and Huang, He-Feng and Wu, Yan-Ting}, + urldate = {2025-02-11}, + date = {2025-01-07}, + keywords = {Fertile window, Machine learning, Menstrual cycle, Natural cycle, Non-invasive wearable device, Wrist skin temperature}, + file = {PDF:/home/alex/Zotero/storage/YIV6T8MS/Luo et al. - 2025 - Prediction of the fertile window and menstrual cycles with a wearable device via machine-learning al.pdf:application/pdf;ScienceDirect Snapshot:/home/alex/Zotero/storage/3IJMUH82/S1472648325000021.html:text/html}, +} + +@article{maman_prediction_2023, + title = {Prediction of ovulation: new insight into an old challenge}, + volume = {13}, + issn = {2045-2322}, + url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10651856/}, + doi = {10.1038/s41598-023-47241-2}, + shorttitle = {Prediction of ovulation}, + abstract = {Ultrasound monitoring and hormonal blood testing are considered by many as an accurate method to predict ovulation time. However, uniform and validated algorithms for predicting ovulation have yet to be defined. Daily hormonal tests and transvaginal ultrasounds were recorded to develop an algorithm for ovulation prediction. The rupture of the leading ovarian follicle was a marker for ovulation day. The model was validated retrospectively on natural cycles frozen embryo transfer cycles with documented ovulation. Circulating levels of {LH} or its relative variation failed, by themselves, to reliably predict ovulation. Any decrease in estrogen was 100\% associated with ovulation emergence the same day or the next day. Progesterone levels {\textgreater} 2 nmol/L had low specificity to predict ovulation the next day (62.7\%), yet its sensitivity was high (91.5\%). A model for ovulation prediction, combining the three hormone levels and ultrasound was created with an accuracy of 95\% to 100\% depending on the combination of the hormone levels. Model validation showed correct ovulation prediction in 97\% of these cycles. We present an accurate ovulation prediction algorithm. The algorithm is simple and user-friendly so both reproductive endocrinologists and general practitioners can use it to benefit their patients.}, + pages = {20003}, + journaltitle = {Sci Rep}, + author = {Maman, Ettie and Adashi, Eli Y. and Baum, Micha and Hourvitz, Ariel}, + urldate = {2025-02-11}, + date = {2023-11-15}, + pmid = {37968377}, + pmcid = {PMC10651856}, + file = {PubMed Central Full Text PDF:/home/alex/Zotero/storage/MTHDPZ5B/Maman et al. - 2023 - Prediction of ovulation new insight into an old challenge.pdf:application/pdf}, +} + +@article{masuda_machine_2025, + title = {Machine learning model for menstrual cycle phase classification and ovulation day detection based on sleeping heart rate under free-living conditions}, + volume = {187}, + issn = {0010-4825}, + url = {https://www.sciencedirect.com/science/article/pii/S0010482525000551}, + doi = {10.1016/j.compbiomed.2025.109705}, + abstract = {The accurate classification of menstrual cycle phases and detection of ovulation is critical for women's health management, particularly in addressing infertility, alleviating premenstrual syndrome, and preventing hormone-related disorders. However, traditional basal body temperature ({BBT}) measurement methods are susceptible to disruptions in sleep timing and environmental conditions, limiting practical application. This study is aimed to overcome these limitations by introducing a novel feature, heart rate at the circadian rhythm nadir ({minHR}), for classifying menstrual cycle phases and predicting ovulation. A machine learning model was developed using {XGBoost}, and data were collected under free-living conditions from 40 healthy women (18–34 years) over a maximum of three menstrual cycles. Three feature combinations— “day,” “day + {minHR},” and “day + {BBT}”—were evaluated, and model performance was assessed using nested leave-one-group-out cross-validation. The feature “day” represents the number of days elapsed since the onset of menstruation. Participants were stratified into groups depending on high variability and low variability in sleep timing. Results demonstrated that adding {minHR} significantly improved luteal phase classification and ovulation day detection performance compared to “day” only. Furthermore, in participants with high variability in sleep timing, the {minHR}-based model outperformed the {BBT}-based model, significantly improving luteal phase recall and reducing ovulation day detection absolute errors by 2 d (p {\textless} 0.05). These findings highlight the robustness and practicality of the {minHR}-based model for menstrual cycle tracking, particularly in individuals with high variability in sleep timing. The proposed model holds great promise for personalized health management and large-scale epidemiological research.}, + pages = {109705}, + journaltitle = {Computers in Biology and Medicine}, + author = {Masuda, Hazuki and Okada, Shima and Shiozawa, Naruhiro and Sakaue, Yusuke and Manno, Masanobu and Makikawa, Masaaki and Isaka, Tadao}, + urldate = {2025-02-11}, + date = {2025-03-01}, + keywords = {Heart rate, Machine learning, Circadian rhythm, Menstrual cycle tracking, Ovulation day detection, Wearable sensor, {XGBoost}}, + file = {PDF:/home/alex/Zotero/storage/87H6TB3Q/Masuda et al. - 2025 - Machine learning model for menstrual cycle phase classification and ovulation day detection based on.pdf:application/pdf;ScienceDirect Snapshot:/home/alex/Zotero/storage/PC6FSQIA/S0010482525000551.html:text/html}, +} + +@online{noauthor_keras_nodate, + title = {Keras: Deep Learning for humans}, + url = {https://keras.io/}, + urldate = {2024-10-23}, + file = {Keras\: Deep Learning for humans:/home/alex/Zotero/storage/MS4QLPRC/keras.io.html:text/html}, +} + +@article{coninck_dianne_2018, + title = {{DIANNE}: a modular framework for designing, training and deploying deep neural networks on heterogeneous distributed infrastructure}, + volume = {141}, + issn = {01641212}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0164121218300487}, + doi = {10.1016/j.jss.2018.03.032}, + shorttitle = {{DIANNE}}, + pages = {52--65}, + journaltitle = {Journal of Systems and Software}, + author = {Coninck, Elias De and Bohez, Steven and Leroux, Sam and Verbelen, Tim and Vankeirsbilck, Bert and Simoens, Pieter and Dhoedt, Bart}, + urldate = {2024-10-23}, + date = {2018-07}, + langid = {english}, + file = {Full Text:/home/alex/Zotero/storage/SFH656QN/Coninck et al. - 2018 - DIANNE a modular framework for designing, training and deploying deep neural networks on heterogene.pdf:application/pdf}, +} + +@misc{shi_time-moe_2024, + title = {Time-{MoE}: Billion-Scale Time Series Foundation Models with Mixture of Experts}, + url = {http://arxiv.org/abs/2409.16040}, + shorttitle = {Time-{MoE}}, + abstract = {Deep learning for time series forecasting has seen significant advancements over the past decades. However, despite the success of large-scale pre-training in language and vision domains, pre-trained time series models remain limited in scale and operate at a high cost, hindering the development of larger capable forecasting models in real-world applications. In response, we introduce Time-{MoE}, a scalable and unified architecture designed to pre-train larger, more capable forecasting foundation models while reducing inference costs. By leveraging a sparse mixture-of-experts ({MoE}) design, Time-{MoE} enhances computational efficiency by activating only a subset of networks for each prediction, reducing computational load while maintaining high model capacity. This allows Time-{MoE} to scale effectively without a corresponding increase in inference costs. Time-{MoE} comprises a family of decoder-only transformer models that operate in an auto-regressive manner and support flexible forecasting horizons with varying input context lengths. We pre-trained these models on our newly introduced large-scale data Time-300B, which spans over 9 domains and encompassing over 300 billion time points. For the first time, we scaled a time series foundation model up to 2.4 billion parameters, achieving significantly improved forecasting precision. Our results validate the applicability of scaling laws for training tokens and model size in the context of time series forecasting. Compared to dense models with the same number of activated parameters or equivalent computation budgets, our models consistently outperform them by large margin. These advancements position Time-{MoE} as a state-of-the-art solution for tackling real-world time series forecasting challenges with superior capability, efficiency, and flexibility.}, + number = {{arXiv}:2409.16040}, + publisher = {{arXiv}}, + author = {Shi, Xiaoming and Wang, Shiyu and Nie, Yuqi and Li, Dianqi and Ye, Zhou and Wen, Qingsong and Jin, Ming}, + urldate = {2024-10-16}, + date = {2024-10-02}, + eprinttype = {arxiv}, + eprint = {2409.16040}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Preprint PDF:/home/alex/Zotero/storage/49N63CMZ/Shi et al. - 2024 - Time-MoE Billion-Scale Time Series Foundation Models with Mixture of Experts.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/HPHN7WNJ/2409.html:text/html}, +} + +@online{noauthor_decoder-only_nodate, + title = {A decoder-only foundation model for time-series forecasting}, + url = {http://research.google/blog/a-decoder-only-foundation-model-for-time-series-forecasting/}, + abstract = {Posted by Rajat Sen and Yichen Zhou, Google Research Time-series forecasting is ubiquitous in various domains, such as retail, finance, manufacturi...}, + urldate = {2024-10-16}, + langid = {english}, + file = {Snapshot:/home/alex/Zotero/storage/JV9JIF73/a-decoder-only-foundation-model-for-time-series-forecasting.html:text/html}, +} + +@misc{goswami_moment_2024, + title = {{MOMENT}: A Family of Open Time-series Foundation Models}, + url = {http://arxiv.org/abs/2402.03885}, + shorttitle = {{MOMENT}}, + abstract = {We introduce {MOMENT}, a family of open-source foundation models for general-purpose time series analysis. Pre-training large models on time series data is challenging due to (1) the absence of a large and cohesive public time series repository, and (2) diverse time series characteristics which make multi-dataset training onerous. Additionally, (3) experimental benchmarks to evaluate these models, especially in scenarios with limited resources, time, and supervision, are still in their nascent stages. To address these challenges, we compile a large and diverse collection of public time series, called the Time series Pile, and systematically tackle time series-specific challenges to unlock large-scale multi-dataset pre-training. Finally, we build on recent work to design a benchmark to evaluate time series foundation models on diverse tasks and datasets in limited supervision settings. Experiments on this benchmark demonstrate the effectiveness of our pre-trained models with minimal data and task-specific fine-tuning. Finally, we present several interesting empirical observations about large pre-trained time series models. Pre-trained models ({AutonLab}/{MOMENT}-1-large) and Time Series Pile ({AutonLab}/Timeseries-{PILE}) are available on Huggingface.}, + number = {{arXiv}:2402.03885}, + publisher = {{arXiv}}, + author = {Goswami, Mononito and Szafer, Konrad and Choudhry, Arjun and Cai, Yifu and Li, Shuo and Dubrawski, Artur}, + urldate = {2024-10-16}, + date = {2024-10-10}, + eprinttype = {arxiv}, + eprint = {2402.03885}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Preprint PDF:/home/alex/Zotero/storage/QF6E6J8W/Goswami et al. - 2024 - MOMENT A Family of Open Time-series Foundation Models.pdf:application/pdf}, +} + +@misc{liang_foundation_2024, + title = {Foundation Models for Time Series Analysis: A Tutorial and Survey}, + url = {http://arxiv.org/abs/2403.14735}, + shorttitle = {Foundation Models for Time Series Analysis}, + abstract = {Time series analysis stands as a focal point within the data mining community, serving as a cornerstone for extracting valuable insights crucial to a myriad of real-world applications. Recent advances in Foundation Models ({FMs}) have fundamentally reshaped the paradigm of model design for time series analysis, boosting various downstream tasks in practice. These innovative approaches often leverage pre-trained or fine-tuned {FMs} to harness generalized knowledge tailored for time series analysis. This survey aims to furnish a comprehensive and up-to-date overview of {FMs} for time series analysis. While prior surveys have predominantly focused on either application or pipeline aspects of {FMs} in time series analysis, they have often lacked an in-depth understanding of the underlying mechanisms that elucidate why and how {FMs} benefit time series analysis. To address this gap, our survey adopts a methodology-centric classification, delineating various pivotal elements of time-series {FMs}, including model architectures, pre-training techniques, adaptation methods, and data modalities. Overall, this survey serves to consolidate the latest advancements in {FMs} pertinent to time series analysis, accentuating their theoretical underpinnings, recent strides in development, and avenues for future exploration.}, + number = {{arXiv}:2403.14735}, + publisher = {{arXiv}}, + author = {Liang, Yuxuan and Wen, Haomin and Nie, Yuqi and Jiang, Yushan and Jin, Ming and Song, Dongjin and Pan, Shirui and Wen, Qingsong}, + urldate = {2024-10-16}, + date = {2024-06-18}, + eprinttype = {arxiv}, + eprint = {2403.14735}, + keywords = {Computer Science - Machine Learning}, + file = {Preprint PDF:/home/alex/Zotero/storage/YEA28FMJ/Liang et al. - 2024 - Foundation Models for Time Series Analysis A Tutorial and Survey.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/9SGH3AIN/2403.html:text/html}, +} + +@online{noauthor_pytorch-forecasting_2024, + title = {Pytorch-Forecasting}, + url = {https://pytorch-forecasting.readthedocs.io/en/stable/}, + urldate = {2024-10-15}, + date = {2024}, +} + +@misc{taylor_forecasting_2017, + title = {Forecasting at scale}, + rights = {http://creativecommons.org/licenses/by/4.0/}, + url = {https://peerj.com/preprints/3190v2}, + doi = {10.7287/peerj.preprints.3190v2}, + abstract = {Forecasting is a common data science task that helps organizations with capacity planning, goal setting, and anomaly detection. Despite its importance, there are serious challenges associated with producing reliable and high quality forecasts –especially when there are a variety of time series and analysts with expertise in time series modeling are relatively rare. To address these challenges, we describe a practical approach to forecasting “at scale” that combines configurable models with analyst-in-the-loop performance analysis. We propose a modular regression model with interpretable parameters that can be intuitively adjusted by analysts with domain knowledge about the time series. We describe performance analyses to compare and evaluate forecasting procedures, and automatically flag forecasts for manual review and adjustment. Tools that help analysts to use their expertise most effectively enable reliable, practical forecasting of business time series.}, + publisher = {{PeerJ} Preprints}, + author = {Taylor, Sean J and Letham, Benjamin}, + urldate = {2024-10-14}, + date = {2017-09-27}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/GK5AIG2V/Taylor and Letham - 2017 - Forecasting at scale.pdf:application/pdf}, +} + +@collection{hutter_machine_2021, + location = {Cham}, + title = {Machine Learning and Knowledge Discovery in Databases: European Conference, {ECML} {PKDD} 2020, Ghent, Belgium, September 14–18, 2020, Proceedings, Part {III}}, + volume = {12459}, + rights = {https://www.springernature.com/gp/researchers/text-and-data-mining}, + isbn = {978-3-030-67663-6 978-3-030-67664-3}, + url = {https://link.springer.com/10.1007/978-3-030-67664-3}, + series = {Lecture Notes in Computer Science}, + shorttitle = {Machine Learning and Knowledge Discovery in Databases}, + publisher = {Springer International Publishing}, + editor = {Hutter, Frank and Kersting, Kristian and Lijffijt, Jefrey and Valera, Isabel}, + urldate = {2024-10-14}, + date = {2021}, + langid = {english}, + doi = {10.1007/978-3-030-67664-3}, + file = {Submitted Version:/home/alex/Zotero/storage/JMVJMLJ5/Hutter et al. - 2021 - Machine Learning and Knowledge Discovery in Databases European Conference, ECML PKDD 2020, Ghent, B.pdf:application/pdf}, +} + +@incollection{hutter_general_2021, + location = {Cham}, + title = {A General Machine Learning Framework for Survival Analysis}, + volume = {12459}, + isbn = {978-3-030-67663-6 978-3-030-67664-3}, + url = {https://link.springer.com/10.1007/978-3-030-67664-3_10}, + pages = {158--173}, + booktitle = {Machine Learning and Knowledge Discovery in Databases}, + publisher = {Springer International Publishing}, + author = {Bender, Andreas and Rügamer, David and Scheipl, Fabian and Bischl, Bernd}, + editor = {Hutter, Frank and Kersting, Kristian and Lijffijt, Jefrey and Valera, Isabel}, + urldate = {2024-10-14}, + date = {2021}, + langid = {english}, + doi = {10.1007/978-3-030-67664-3_10}, + note = {Series Title: Lecture Notes in Computer Science}, + file = {Submitted Version:/home/alex/Zotero/storage/WQIHZ7IP/Bender et al. - 2021 - A General Machine Learning Framework for Survival Analysis.pdf:application/pdf}, +} + +@online{lightningai_pytorch_2024, + title = {{PyTorch} Lightning}, + url = {https://www.pytorchlightning.ai}, + author = {lightning.ai}, + urldate = {2024-10-14}, + date = {2024}, +} + +@misc{alexandrov_gluonts_2019, + title = {{GluonTS}: Probabilistic Time Series Models in Python}, + url = {http://arxiv.org/abs/1906.05264}, + shorttitle = {{GluonTS}}, + abstract = {We introduce Gluon Time Series ({GluonTS}, available at https://gluon-ts.mxnet.io), a library for deep-learning-based time series modeling. {GluonTS} simplifies the development of and experimentation with time series models for common tasks such as forecasting or anomaly detection. It provides all necessary components and tools that scientists need for quickly building new models, for efficiently running and analyzing experiments and for evaluating model accuracy.}, + number = {{arXiv}:1906.05264}, + publisher = {{arXiv}}, + author = {Alexandrov, Alexander and Benidis, Konstantinos and Bohlke-Schneider, Michael and Flunkert, Valentin and Gasthaus, Jan and Januschowski, Tim and Maddix, Danielle C. and Rangapuram, Syama and Salinas, David and Schulz, Jasper and Stella, Lorenzo and Türkmen, Ali Caner and Wang, Yuyang}, + urldate = {2024-10-14}, + date = {2019-06-14}, + eprinttype = {arxiv}, + eprint = {1906.05264}, + keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, + file = {Preprint PDF:/home/alex/Zotero/storage/JP9K74A8/Alexandrov et al. - 2019 - GluonTS Probabilistic Time Series Models in Python.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/RJYSBT29/1906.html:text/html}, +} + +@misc{cho_learning_2014, + title = {Learning Phrase Representations using {RNN} Encoder-Decoder for Statistical Machine Translation}, + url = {http://arxiv.org/abs/1406.1078}, + abstract = {In this paper, we propose a novel neural network model called {RNN} Encoder-Decoder that consists of two recurrent neural networks ({RNN}). One {RNN} encodes a sequence of symbols into a fixed-length vector representation, and the other decodes the representation into another sequence of symbols. The encoder and decoder of the proposed model are jointly trained to maximize the conditional probability of a target sequence given a source sequence. The performance of a statistical machine translation system is empirically found to improve by using the conditional probabilities of phrase pairs computed by the {RNN} Encoder-Decoder as an additional feature in the existing log-linear model. Qualitatively, we show that the proposed model learns a semantically and syntactically meaningful representation of linguistic phrases.}, + number = {{arXiv}:1406.1078}, + publisher = {{arXiv}}, + author = {Cho, Kyunghyun and Merrienboer, Bart van and Gulcehre, Caglar and Bahdanau, Dzmitry and Bougares, Fethi and Schwenk, Holger and Bengio, Yoshua}, + urldate = {2024-10-10}, + date = {2014-09-03}, + eprinttype = {arxiv}, + eprint = {1406.1078}, + keywords = {Computer Science - Machine Learning, Statistics - Machine Learning, Computer Science - Neural and Evolutionary Computing, Computer Science - Computation and Language}, + file = {Preprint PDF:/home/alex/Zotero/storage/E8WMK2IN/Cho et al. - 2014 - Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/6PTCL8LW/1406.html:text/html}, } @article{hochreiter_long_1997-1, - title = {Long {Short}-{Term} {Memory}}, + title = {Long Short-Term Memory}, volume = {9}, issn = {0899-7667, 1530-888X}, url = {https://direct.mit.edu/neco/article/9/8/1735-1780/6109}, doi = {10.1162/neco.1997.9.8.1735}, - abstract = {Learning to store information over extended time intervals by recurrent backpropagation takes a very long time, mostly because of insufficient, decaying error backflow. We briefly review Hochreiter's (1991) analysis of this problem, then address it by introducing a novel, efficient, gradient based method called long short-term memory (LSTM). Truncating the gradient where this does not do harm, LSTM can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units. Multiplicative gate units learn to open and close access to the constant error flow. LSTM is local in space and time; its computational complexity per time step and weight is O. 1. Our experiments with artificial data involve local, distributed, real-valued, and noisy pattern representations. In comparisons with real-time recurrent learning, back propagation through time, recurrent cascade correlation, Elman nets, and neural sequence chunking, LSTM leads to many more successful runs, and learns much faster. LSTM also solves complex, artificial long-time-lag tasks that have never been solved by previous recurrent network algorithms.}, - language = {en}, - number = {8}, - urldate = {2025-06-11}, - journal = {Neural Computation}, - author = {Hochreiter, Sepp and Schmidhuber, Jürgen}, - month = nov, - year = {1997}, + abstract = {Learningtostoreinformationoverextendedtimeintervalsviarecurrentbackpropagation takesaverylongtime,mostlyduetoinsu cient,decayingerrorbackow.Webrieyreview Hochreiter's1991analysisofthisproblem,thenaddressitbyintroducinganovel,e cient, gradient-basedmethodcalled{\textbackslash}{LongShort}-{TermMemory}"({LSTM}).Truncatingthegradient wherethisdoesnotdoharm,{LSTMcanlearntobridgeminimaltimelagsinexcessof}1000 discretetimestepsbyenforcingconstanterrorowthrough{\textbackslash}constanterrorcarrousels"within specialunits.Multiplicativegateunitslearntoopenandcloseaccesstotheconstanterror ow.{LSTMislocalinspaceandtime};itscomputationalcomplexitypertimestepandweight {isO}(1).Ourexperimentswitharticialdatainvolvelocal,distributed,real-valued,andnoisy patternrepresentations.{IncomparisonswithRTRL},{BPTT},{RecurrentCascade}-Correlation, Elmannets,{andNeuralSequenceChunking},{LSTMleadstomanymoresuccessfulruns},and learnsmuchfaster.{LSTMalsosolvescomplex},articiallongtimelagtasksthathavenever beensolvedbypreviousrecurrentnetworkalgorithms.}, pages = {1735--1780}, - file = {PDF:/home/alex/Zotero/storage/5IE5G9KY/Hochreiter and Schmidhuber - 1997 - Long Short-Term Memory.pdf:application/pdf}, -} - -@book{medsker_recurrent_1999, - title = {Recurrent {Neural} {Networks}: {Design} and {Applications}}, - isbn = {978-1-4200-4917-6}, - shorttitle = {Recurrent {Neural} {Networks}}, - abstract = {With existent uses ranging from motion detection to music synthesis to financial forecasting, recurrent neural networks have generated widespread attention. The tremendous interest in these networks drives Recurrent Neural Networks: Design and Applications, a summary of the design, applications, current research, and challenges of this subfield of artificial neural networks.This overview incorporates every aspect of recurrent neural networks. It outlines the wide variety of complex learning techniques and associated research projects. Each chapter addresses architectures, from fully connected to partially connected, including recurrent multilayer feedforward. It presents problems involving trajectories, control systems, and robotics, as well as RNN use in chaotic systems. The authors also share their expert knowledge of ideas for alternate designs and advances in theoretical aspects.The dynamical behavior of recurrent neural networks is useful for solving problems in science, engineering, and business. This approach will yield huge advances in the coming years. Recurrent Neural Networks illuminates the opportunities and provides you with a broad view of the current events in this rich field.}, - language = {en}, - publisher = {CRC Press}, - author = {Medsker, Larry and Jain, Lakhmi C.}, - month = dec, - year = {1999}, - note = {Google-Books-ID: ME1SAkN0PyMC}, - keywords = {Computers / Computer Engineering, Computers / General, Computers / Software Development \& Engineering / Systems Analysis \& Design, Technology \& Engineering / Electronics / General}, -} - -@article{hochreiter_vanishing_1998, - title = {The {Vanishing} {Gradient} {Problem} {During} {Learning} {Recurrent} {Neural} {Nets} and {Problem} {Solutions}}, - volume = {06}, - issn = {0218-4885, 1793-6411}, - url = {https://www.worldscientific.com/doi/abs/10.1142/S0218488598000094}, - doi = {10.1142/S0218488598000094}, - abstract = {Recurrent nets are in principle capable to store past inputs to produce the currently desired output. Because of this property recurrent nets are used in time series prediction and process control. Practical applications involve temporal dependencies spanning many time steps, e.g. between relevant inputs and desired outputs. In this case, however, gradient based learning methods take too much time. The extremely increased learning time arises because the error vanishes as it gets propagated back. In this article the de-caying error flow is theoretically analyzed. Then methods trying to overcome vanishing gradients are briefly discussed. Finally, experiments comparing conventional algorithms and alternative methods are presented. With advanced methods long time lag problems can be solved in reasonable time.}, - language = {en}, - number = {02}, - urldate = {2025-06-11}, - journal = {International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems}, - author = {Hochreiter, Sepp}, - month = apr, - year = {1998}, - pages = {107--116}, -} - -@misc{fdeloche_english_2017, - title = {English: {A} diagram for a one-unit recurrent neural network ({RNN}). {From} bottom to top : input state, hidden state, output state. {U}, {V}, {W} are the weights of the network. {Compressed} diagram on the left and the unfold version of it on the right.}, - shorttitle = {English}, - url = {https://commons.wikimedia.org/wiki/File:Recurrent_neural_network_unfold.svg}, - urldate = {2025-06-11}, - author = {{fdeloche}}, - month = jun, - year = {2017}, - file = {Wikimedia Snapshot:/home/alex/Zotero/storage/RU2XSRZN/FileRecurrent_neural_network_unfold.html:text/html}, -} - -@misc{chevalier_english_2018, - title = {English: {Schematic} of the {Long}-{Short} {Term} {Memory} cell, a component of recurrent neural networks}, - shorttitle = {English}, - url = {https://commons.wikimedia.org/wiki/File:LSTM_Cell.svg}, - urldate = {2025-06-11}, - author = {Chevalier, Guillaume}, - month = may, - year = {2018}, - file = {Wikimedia Snapshot:/home/alex/Zotero/storage/NMYA4ZA3/FileLSTM_Cell.html:text/html}, -} - -@article{twenge_declines_2017, - title = {Declines in {Sexual} {Frequency} among {American} {Adults}, 1989–2014}, - volume = {46}, - issn = {0004-0002, 1573-2800}, - url = {http://link.springer.com/10.1007/s10508-017-0953-1}, - doi = {10.1007/s10508-017-0953-1}, - language = {en}, number = {8}, - urldate = {2025-06-18}, - journal = {Archives of Sexual Behavior}, - author = {Twenge, Jean M. and Sherman, Ryne A. and Wells, Brooke E.}, - month = nov, - year = {2017}, - pages = {2389--2401}, - file = {PDF:/home/alex/Zotero/storage/7FLDCF3U/Twenge et al. - 2017 - Declines in Sexual Frequency among American Adults, 1989–2014.pdf:application/pdf}, + journaltitle = {Neural Computation}, + author = {Hochreiter, Sepp and Schmidhuber, Jürgen}, + urldate = {2024-10-10}, + date = {1997-11-01}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/CZSV2ASE/Hochreiter and Schmidhuber - 1997 - Long Short-Term Memory.pdf:application/pdf}, } -@misc{noauthor_ringpng_nodate, - title = {ring.png (800×800)}, - url = {https://ovularing.com/wp-content/uploads/2021/07/ring.png}, - urldate = {2025-06-25}, +@misc{lim_temporal_2020, + title = {Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting}, + url = {http://arxiv.org/abs/1912.09363}, + abstract = {Multi-horizon forecasting problems often contain a complex mix of inputs -- including static (i.e. time-invariant) covariates, known future inputs, and other exogenous time series that are only observed historically -- without any prior information on how they interact with the target. While several deep learning models have been proposed for multi-step prediction, they typically comprise black-box models which do not account for the full range of inputs present in common scenarios. In this paper, we introduce the Temporal Fusion Transformer ({TFT}) -- a novel attention-based architecture which combines high-performance multi-horizon forecasting with interpretable insights into temporal dynamics. To learn temporal relationships at different scales, the {TFT} utilizes recurrent layers for local processing and interpretable self-attention layers for learning long-term dependencies. The {TFT} also uses specialized components for the judicious selection of relevant features and a series of gating layers to suppress unnecessary components, enabling high performance in a wide range of regimes. On a variety of real-world datasets, we demonstrate significant performance improvements over existing benchmarks, and showcase three practical interpretability use-cases of {TFT}.}, + number = {{arXiv}:1912.09363}, + publisher = {{arXiv}}, + author = {Lim, Bryan and Arik, Sercan O. and Loeff, Nicolas and Pfister, Tomas}, + urldate = {2024-10-10}, + date = {2020-09-27}, + eprinttype = {arxiv}, + eprint = {1912.09363}, + keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, + file = {Preprint PDF:/home/alex/Zotero/storage/2R2H34KB/Lim et al. - 2020 - Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/ETDAYW36/1912.html:text/html}, } -@misc{noauthor_ovularing_nodate, - title = {{OvulaRing} {Startseite}}, - url = {https://ovularing.com/}, - abstract = {Erfahre hier mehr zu OvulaRing Startseite}, - language = {de-DE}, - urldate = {2025-06-25}, - journal = {OvulaRing}, - file = {Snapshot:/home/alex/Zotero/storage/PD5DBIS4/ovularing.com.html:text/html}, +@misc{nie_time_2023, + title = {A Time Series is Worth 64 Words: Long-term Forecasting with Transformers}, + url = {http://arxiv.org/abs/2211.14730}, + shorttitle = {A Time Series is Worth 64 Words}, + abstract = {We propose an efficient design of Transformer-based models for multivariate time series forecasting and self-supervised representation learning. It is based on two key components: (i) segmentation of time series into subseries-level patches which are served as input tokens to Transformer; (ii) channel-independence where each channel contains a single univariate time series that shares the same embedding and Transformer weights across all the series. Patching design naturally has three-fold benefit: local semantic information is retained in the embedding; computation and memory usage of the attention maps are quadratically reduced given the same look-back window; and the model can attend longer history. Our channel-independent patch time series Transformer ({PatchTST}) can improve the long-term forecasting accuracy significantly when compared with that of {SOTA} Transformer-based models. We also apply our model to self-supervised pre-training tasks and attain excellent fine-tuning performance, which outperforms supervised training on large datasets. Transferring of masked pre-trained representation on one dataset to others also produces {SOTA} forecasting accuracy. Code is available at: https://github.com/yuqinie98/{PatchTST}.}, + number = {{arXiv}:2211.14730}, + publisher = {{arXiv}}, + author = {Nie, Yuqi and Nguyen, Nam H. and Sinthong, Phanwadee and Kalagnanam, Jayant}, + urldate = {2024-10-10}, + date = {2023-03-05}, + eprinttype = {arxiv}, + eprint = {2211.14730}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Preprint PDF:/home/alex/Zotero/storage/DG4ZJCWV/Nie et al. - 2023 - A Time Series is Worth 64 Words Long-term Forecasting with Transformers.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/H6XGVBY6/2211.html:text/html}, } -@article{wu_deep_nodate, - title = {Deep {Transformer} {Models} for {Time} {Series} {Forecasting}:{The} {Influenza} {Prevalence} {Case}}, - abstract = {In this paper, we present a new approach to time series forecasting. Time series data are prevalent in many scientific and engineering disciplines. Time series forecasting is a crucial task in modeling time series data, and is an important area of machine learning. In this work we developed a novel method that employs Transformer-based machine learning models to forecast time series data. This approach works by leveraging selfattention mechanisms to learn complex patterns and dynamics from time series data. Moreover, it is a generic framework and can be applied to univariate and multivariate time series data, as well as time series embeddings. Using influenzalike illness (ILI) forecasting as a case study, we show that the forecasting results produced by our approach are favorably comparable to the stateof-the-art.}, - language = {en}, - author = {Wu, Neo and Green, Bradley and Ben, Xue and O'Banion, Shawn}, - file = {PDF:/home/alex/Zotero/storage/GHT5UMNX/Wu et al. - Deep Transformer Models for Time Series ForecastingThe Influenza Prevalence Case.pdf:application/pdf}, +@misc{shao_exploring_2023, + title = {Exploring Progress in Multivariate Time Series Forecasting: Comprehensive Benchmarking and Heterogeneity Analysis}, + url = {http://arxiv.org/abs/2310.06119}, + shorttitle = {Exploring Progress in Multivariate Time Series Forecasting}, + abstract = {Multivariate Time Series ({MTS}) widely exists in real-word complex systems, such as traffic and energy systems, making their forecasting crucial for understanding and influencing these systems. Recently, deep learning-based approaches have gained much popularity for effectively modeling temporal and spatial dependencies in {MTS}, specifically in Long-term Time Series Forecasting ({LTSF}) and Spatial-Temporal Forecasting ({STF}). However, the fair benchmarking issue and the choice of technical approaches have been hotly debated in related work. Such controversies significantly hinder our understanding of progress in this field. Thus, this paper aims to address these controversies to present insights into advancements achieved. To resolve benchmarking issues, we introduce {BasicTS}, a benchmark designed for fair comparisons in {MTS} forecasting. {BasicTS} establishes a unified training pipeline and reasonable evaluation settings, enabling an unbiased evaluation of over 30 popular {MTS} forecasting models on more than 18 datasets. Furthermore, we highlight the heterogeneity among {MTS} datasets and classify them based on temporal and spatial characteristics. We further prove that neglecting heterogeneity is the primary reason for generating controversies in technical approaches. Moreover, based on the proposed {BasicTS} and rich heterogeneous {MTS} datasets, we conduct an exhaustive and reproducible performance and efficiency comparison of popular models, providing insights for researchers in selecting and designing {MTS} forecasting models.}, + number = {{arXiv}:2310.06119}, + publisher = {{arXiv}}, + author = {Shao, Zezhi and Wang, Fei and Xu, Yongjun and Wei, Wei and Yu, Chengqing and Zhang, Zhao and Yao, Di and Jin, Guangyin and Cao, Xin and Cong, Gao and Jensen, Christian S. and Cheng, Xueqi}, + urldate = {2024-10-10}, + date = {2023-10-09}, + eprinttype = {arxiv}, + eprint = {2310.06119}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence}, + file = {Preprint PDF:/home/alex/Zotero/storage/7EFZ5IT6/Shao et al. - 2023 - Exploring Progress in Multivariate Time Series Forecasting Comprehensive Benchmarking and Heterogen.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/W6RNWBLM/2310.html:text/html}, } -@article{vaswani_attention_nodate, - title = {Attention {Is} {All} {You} {Need}}, - abstract = {The dominant sequence transduction models are based on complex recurrent or convolutional neural networks that include an encoder and a decoder. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 BLEU on the WMT 2014 Englishto-German translation task, improving over the existing best results, including ensembles, by over 2 BLEU. On the WMT 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art BLEU score of 41.8 after training for 3.5 days on eight GPUs, a small fraction of the training costs of the best models from the literature. We show that the Transformer generalizes well to other tasks by applying it successfully to English constituency parsing both with large and limited training data.}, - language = {en}, - author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, Łukasz and Polosukhin, Illia}, - file = {PDF:/home/alex/Zotero/storage/TES5P5PX/Vaswani et al. - Attention Is All You Need.pdf:application/pdf}, +@article{zhang_crossformer_2023, + title = {{CROSSFORMER}: {TRANSFORMER} {UTILIZING} {CROSS}- {DIMENSION} {DEPENDENCY} {FOR} {MULTIVARIATE} {TIME} {SERIES} {FORECASTING}}, + abstract = {Recently many deep models have been proposed for multivariate time series ({MTS}) forecasting. In particular, Transformer-based models have shown great potential because they can capture long-term dependency. However, existing Transformerbased models mainly focus on modeling the temporal dependency (cross-time dependency) yet often omit the dependency among different variables (crossdimension dependency), which is critical for {MTS} forecasting. To fill the gap, we propose Crossformer, a Transformer-based model utilizing cross-dimension dependency for {MTS} forecasting. In Crossformer, the input {MTS} is embedded into a 2D vector array through the Dimension-Segment-Wise ({DSW}) embedding to preserve time and dimension information. Then the Two-Stage Attention ({TSA}) layer is proposed to efficiently capture the cross-time and cross-dimension dependency. Utilizing {DSW} embedding and {TSA} layer, Crossformer establishes a Hierarchical Encoder-Decoder ({HED}) to use the information at different scales for the final forecasting. Extensive experimental results on six real-world datasets show the effectiveness of Crossformer against previous state-of-the-arts.}, + author = {Zhang, Yunhao and Yan, Junchi}, + date = {2023}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/NM9CETJS/Zhang and Yan - 2023 - CROSSFORMER TRANSFORMER UTILIZING CROSS- DIMENSION DEPENDENCY FOR MULTIVARIATE TIME SERIES FORECAST.pdf:application/pdf}, } -@article{alliende_cervicovaginal_2005, - title = {Cervicovaginal fluid changes to detect ovulation accurately}, - volume = {193}, - copyright = {https://www.elsevier.com/tdm/userlicense/1.0/}, - issn = {00029378}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0002937804018770}, - doi = {10.1016/j.ajog.2004.11.006}, - abstract = {Objective: The purpose of this study was to evaluate changes in cervicovaginal fluid characteristics to identify ovulation. Study design: Several ovulation indicators were studied in a university-based natural family planning center. Fifteen parous women during 29 ovulatory cycles detected cervicovaginal fluid at the vulva. They self-aspirated their upper vaginal fluid, described it, and kept it for later checking. They also took basal body temperature, collected timed first morning urine samples for estrone and pregnanediol glucuronide enzyme immunoassays, and submitted to serial ovarian transvaginal ultrasound scans. -Results: Considering a G 1-day period since ultrasound ovulation detection or allowing an extra day (ÿ1 to C2), women perceived ovulation from cervicovaginal fluid at the vulva in 76\% or 97\% of cycles, on the basis of their visual description of vaginally extracted fluid in 76\% or 90\%, which rose to 90\% or 97\% for the instructor’s description, and in 76\% or 86\% with a rapid drop in glucuronide ratio. Basal body temperature was less precise (71\% or 79\%). -Conclusion: Evaluation of cervicovaginal fluid changes is an accurate ovulation indicator. Ó 2005 Elsevier Inc. All rights reserved.}, - language = {en}, - number = {1}, - urldate = {2025-07-01}, - journal = {American Journal of Obstetrics and Gynecology}, - author = {Alliende, María Elena and Cabezón, Carlos and Figueroa, Horacio and Kottmann, Cristián}, - month = jul, - year = {2005}, - pages = {71--75}, - file = {PDF:/home/alex/Zotero/storage/NN29EZ6G/Alliende et al. - 2005 - Cervicovaginal fluid changes to detect ovulation accurately.pdf:application/pdf}, -} - -@article{luo_detection_2020, - title = {Detection and {Prediction} of {Ovulation} {From} {Body} {Temperature} {Measured} by an {In}-{Ear} {Wearable} {Thermometer}}, - volume = {67}, - copyright = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html}, - issn = {0018-9294, 1558-2531}, - url = {https://ieeexplore.ieee.org/document/8715448/}, - doi = {10.1109/TBME.2019.2916823}, - abstract = {Objective: We present a non-invasive wearable device for fertility monitoring and propose an effective and flexible statistical learning algorithm to detect and predict ovulation using data captured by this device. Methods: The system consists of an earpiece, which measures the ear canal temperature every 5 minutes during night sleep hours, and a base station that transmits data to a smartphone application for analysis. We establish a data-cleaning protocol for data preprocessing and then fit a Hidden Markov Model (HMM) with two hidden states of high and low temperature to identify the more probable state of each time point via the predicted probabilities. Finally, a post-processing procedure is developed to incorporate biorhythm information to form a time-course biphasic profile for each subject. Results: The performance of the proposed algorithms applied to data collected by the device are compared with traditional methods in terms of match rate with self-reported ovulation days confirmed with an Ovulation Test Kit. Empirical study results from a group of 34 users yielded significant improvements over the traditional methods in terms of detection accuracy (with sensitivity 92.31\%) and prediction power (23.0731.55\% higher). Conclusion: We demonstrated the feasibility for reliable ovulation detection and prediction with high-frequency temperature data collected by a non-invasive wearable device. Significance: Traditional fertility monitoring methods are often either inaccurate or inconvenient. The wearable device and learning algorithm presented in this paper provides a userfriendly and reliable platform for tracking ovulation, which may have a broad impact on both fertility research and real-world family planning.}, - language = {en}, - number = {2}, - urldate = {2025-07-01}, - journal = {IEEE Transactions on Biomedical Engineering}, - author = {Luo, Lan and She, Xichen and Cao, Jiexuan and Zhang, Yunlong and Li, Yijiang and Song, Peter X. K.}, - month = feb, - year = {2020}, - pages = {512--522}, - file = {PDF:/home/alex/Zotero/storage/YLIR4YNX/Luo et al. - 2020 - Detection and Prediction of Ovulation From Body Temperature Measured by an In-Ear Wearable Thermomet.pdf:application/pdf}, -} - -@article{bauman_basal_1981, - title = {Basal {Body} {Temperature}: {Unreliable} {Method} of {Ovulation} {Detection}}, - volume = {36}, +@article{leader_prediction_1985, + title = {The prediction of ovulation: a comparison of the basal body temperature graph, cervical mucus score, and real-time pelvic ultrasonography}, + volume = {43}, issn = {00150282}, - shorttitle = {Basal {Body} {Temperature}}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216459169}, - doi = {10.1016/S0015-0282(16)45916-9}, - language = {en}, - number = {6}, - urldate = {2025-07-01}, - journal = {Fertility and Sterility}, - author = {Bauman, Joan E.}, - month = dec, - year = {1981}, - pages = {729--733}, - file = {PDF:/home/alex/Zotero/storage/9T7V8ACT/Bauman - 1981 - Basal Body Temperature Unreliable Method of Ovulation Detection.pdf:application/pdf}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216484360}, + doi = {10.1016/S0015-0282(16)48436-0}, + shorttitle = {The prediction of ovulation}, + pages = {385--388}, + number = {3}, + journaltitle = {Fertility and Sterility}, + author = {Leader, Arthur and Wiseman, David and Taylor, Patrick J.}, + urldate = {2025-08-01}, + date = {1985-03}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/2GQKZ2YH/Leader et al. - 1985 - The prediction of ovulation a comparison of the basal body temperature graph, cervical mucus score,.pdf:application/pdf}, } -@misc{noauthor_pytorch_nodate, - title = {{PyTorch}}, - url = {https://pytorch.org/}, - abstract = {PyTorch Foundation is the deep learning community home for the open source PyTorch framework and ecosystem.}, - language = {en-US}, - urldate = {2025-07-11}, - journal = {PyTorch}, - file = {Snapshot:/home/alex/Zotero/storage/K7CTLE77/pytorch.org.html:text/html}, -} - -@article{li_menstrual_2023, - title = {Menstrual cycle length variation by demographic characteristics from the {Apple} {Women}’s {Health} {Study}}, - volume = {6}, - issn = {2398-6352}, - url = {https://www.nature.com/articles/s41746-023-00848-1}, - doi = {10.1038/s41746-023-00848-1}, - abstract = {Abstract - - Menstrual characteristics are important signs of women’s health. Here we examine the variation of menstrual cycle length by age, ethnicity, and body weight using 165,668 cycles from 12,608 participants in the US using mobile menstrual tracking apps. After adjusting for all covariates, mean menstrual cycle length is shorter with older age across all age groups until age 50 and then became longer for those age 50 and older. Menstrual cycles are on average 1.6 (95\%CI: 1.2, 2.0) days longer for Asian and 0.7 (95\%CI: 0.4, 1.0) days longer for Hispanic participants compared to white non-Hispanic participants. Participants with BMI ≥ 40 kg/m - 2 - have 1.5 (95\%CI: 1.2, 1.8) days longer cycles compared to those with BMI between 18.5 and 25 kg/m - 2 - . Cycle variability is the lowest among participants aged 35–39 but are considerably higher by 46\% (95\%CI: 43\%, 48\%) and 45\% (95\%CI: 41\%, 49\%) among those aged under 20 and between 45–49. Cycle variability increase by 200\% (95\%CI: 191\%, 210\%) among those aged above 50 compared to those in the 35–39 age group. Compared to white participants, those who are Asian and Hispanic have larger cycle variability. Participants with obesity also have higher cycle variability. Here we confirm previous observations of changes in menstrual cycle pattern with age across reproductive life span and report new evidence on the differences of menstrual variation by ethnicity and obesity status. Future studies should explore the underlying determinants of the variation in menstrual characteristics.}, - language = {en}, - number = {1}, - urldate = {2025-07-04}, - journal = {npj Digital Medicine}, - author = {Li, Huichu and Gibson, Elizabeth A. and Jukic, Anne Marie Z. and Baird, Donna D. and Wilcox, Allen J. and Curry, Christine L. and Fischer-Colbrie, Tyler and Onnela, Jukka-Pekka and Williams, Michelle A. and Hauser, Russ and Coull, Brent A. and Mahalingaiah, Shruthi}, - month = may, - year = {2023}, - pages = {100}, - file = {PDF:/home/alex/Zotero/storage/9J5N6YIW/Li et al. - 2023 - Menstrual cycle length variation by demographic characteristics from the Apple Women’s Health Study.pdf:application/pdf}, -} - -@article{ecochard_menstrual_2024, - title = {The menstrual cycle is influenced by weekly and lunar rhythms}, - volume = {121}, - issn = {00150282}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028223020769}, - doi = {10.1016/j.fertnstert.2023.12.009}, - abstract = {Objective: To study whether the menstrual cycle has a circaseptan (7 days) rhythm and whether it is associated with the lunar cycle (also defined as the synodic month, it is the cycle of the phases of the Moon as seen from Earth, averaging 29.5 days in length). Design: Cross-sectional study. Subjects: A total of 35,940 European and North American women aged 18–40 years. Exposure: Data were collected in real-life conditions. Intervention: No intervention was performed. Main Outcome Measure: The onset of menstruation was assessed in prospectively measured menstrual cycles (311,064 cycles) over 3 full years (2019–2021). Associations were calculated between the onset of menstruation and the day of the week, and between the onset of menstruation and the lunar phase. -Results: In this large data set, a circaseptan (7-day) rhythmicity of menstruation was observed, with a peak (acrophase) of menstrual onset on Thursdays and Fridays. This circaseptan rhythm was observed in every age group, in every phase of the lunar cycle, and in all seasons. This feature was most pronounced for cycle durations between 27 and 29 days. In winter, the circaseptan rhythm was found in cycles of 27–29 days, but not in other cycle lengths. A circalunar rhythm was also statistically significant, but not as clearly defined as the circaseptan rhythm. The peak (acrophase) of the circalunar rhythm of menstrual onset varied according to the season. In addition, there was a small but statistically significant interaction between the circaseptan rhythm and the lunar cycle. -Conclusion: Although relatively small in amplitude, the weekly rhythm of menstruation was statistically significant. Menstruation occurs more often on Thursdays and Fridays than on other days of the week. This is particularly true for women whose cycles last between 27 and 29 days. Circalunar rhythmicity was also statistically significant. However, it is less pronounced than the weekly rhythm. (Fertil SterilÒ 2024;121:651-9. Ó2023 by American Society for Reproductive Medicine.)}, - language = {en}, +@article{albertson_prediction_1987, + title = {The prediction of ovulation and monitoring of the fertile period}, + volume = {3}, + rights = {http://www.springer.com/tdm}, + issn = {0267-4874, 1573-7195}, + url = {http://link.springer.com/10.1007/BF01849284}, + doi = {10.1007/BF01849284}, + abstract = {Simple and reliable methods have been sought for both predicting and confirming ovulation. Application of these methods could include management of infertile couples to aid in conception and for increasing the reliability of natural family planning ({NFCF}) as a method of birth control. With the advent of specific hormone assays, serial measurements of estrogens, progesterone (and metabolites), and luteinizing hormone have been the gold standard of monitoring ovarian function in women. However, newer and simpler methodologies have been described and are currently either in use or being tested. These include the measurement of basal body temperature ({BBT}), the evaluation of the volume, consistency and electro-conductivity of cervicovaginal fluid, salivary steroid content and cellular enzymatic activity, the use of enzymelinked immunosorbent assays applied to solid-phase formats, and the investigation of new hormonal molecules as markers of reproductive state and function. These new technologies are described herein and their potential for monitoring ovarian function is discussed.}, + pages = {263--290}, number = {4}, - urldate = {2025-07-03}, - journal = {Fertility and Sterility}, - author = {Ecochard, René and Leiva, Rene and Bouchard, Thomas P. and Van Lamsweerde, Agathe and Pearson, Jack T. and Stanford, Joseph B. and Gronfier, Claude}, - month = apr, - year = {2024}, - pages = {651--659}, - file = {PDF:/home/alex/Zotero/storage/T6DNB4M9/Ecochard et al. - 2024 - The menstrual cycle is influenced by weekly and lunar rhythms.pdf:application/pdf}, + journaltitle = {Adv Contracept}, + author = {Albertson, B. D. and Zinaman, M. J.}, + urldate = {2025-08-01}, + date = {1987-12}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/SQCGXH4T/Albertson and Zinaman - 1987 - The prediction of ovulation and monitoring of the fertile period.pdf:application/pdf}, } -@misc{noauthor_kegg_nodate, - title = {kegg® fertility monitor \& kegel ball {\textbar} {Track} {Key} {Fertility} {Metric}}, - url = {https://kegg.tech/}, - abstract = {Plan your pregnancy with confidence ... kegg® is a medical-grade fertility device that gives you accurate and personalized fertility tracking through cervical mucus.}, - language = {en}, - urldate = {2025-07-02}, - journal = {kegg}, - file = {Snapshot:/home/alex/Zotero/storage/HNFFJPSN/kegg.tech.html:text/html}, -} - -@article{moreno_temporal_1988, - title = {{TEMPORAL} {RELATION} {OF} {OVDLATION} {TO} {SALIVARY} {AND} {VAGINAL} {ELECTRICAL} {RESISTANCE} {PATTERNS}: {IMPLICATIONS} {FOR} {NATURAL} {FAMILY} {PLANNING}}, - abstract = {An independent assessment of the CUETM Monitor (Zetek, Aurora, Colorado) as an ovulation predictor was made with emphasis on its potential role in "natural family planning". The device provides a digital measurement of the electrical resistance of saliva and vaginal secretions. Twenty-nine menstrual cycles from 11 regularly cycling women were monitored with basal temperatures, urinary LH, pelvic ultrasound and the CUE monitor. Patterns of peak salivary electrical resistance were able to predict ovulation on average 5.3 (51.9 SD) days in advance. Despite variations in total length of the follicular phase from cycle to cycle, the within-subject variation of this predictive interval was quite small. Nadirs in the electrical resistance of vaginal secretions occurred within 2 days of ovulation in all but one patient. Variation in this interval from cycle-tocycle was small as well. We propose an algorithm for the use of these intervals in "natural family planning" that could safely reduce the monthly abstinence period of present methods. The simplicity, objectivity and consistency of this device could result in their greater general acceptance.}, - language = {en}, - author = {Moreno, Jorge E and Doody, Michael C and Besch, Paige}, - month = oct, - year = {1988}, - file = {PDF:/home/alex/Zotero/storage/FFW8KBZQ/Moreno et al. - TEMPORAL RELATION OF OVDLATION TO SALIVARY AND VAGINAL ELECTRICAL RESISTANCE PATTERNS IMPLICATIONS.pdf:application/pdf}, -} - -@misc{noauthor_trackle_nodate, - title = {trackle - einfach hormonfrei verhüten}, - url = {https://trackle.de/}, - abstract = {Das trackle Sensorsystem hilft Dir, einfach, sicher und hormonfrei zu verhüten. Jetzt informieren und symptothermale Methode nutzen!}, - language = {de}, - urldate = {2025-07-02}, - journal = {trackle}, - file = {Snapshot:/home/alex/Zotero/storage/KNFY8X2K/trackle.de.html:text/html}, -} - -@article{zhu_accuracy_2021, - title = {The {Accuracy} of {Wrist} {Skin} {Temperature} in {Detecting} {Ovulation} {Compared} to {Basal} {Body} {Temperature}: {Prospective} {Comparative} {Diagnostic} {Accuracy} {Study}}, - volume = {23}, - issn = {1438-8871}, - shorttitle = {The {Accuracy} of {Wrist} {Skin} {Temperature} in {Detecting} {Ovulation} {Compared} to {Basal} {Body} {Temperature}}, - url = {https://www.jmir.org/2021/6/e20710}, - doi = {10.2196/20710}, - abstract = {Background: As a daily point measurement, basal body temperature (BBT) might not be able to capture the temperature shift in the menstrual cycle because a single temperature measurement is present on the sliding scale of the circadian rhythm. Wrist skin temperature measured continuously during sleep has the potential to overcome this limitation. -Objective: This study compares the diagnostic accuracy of these two temperatures for detecting ovulation and to investigate the correlation and agreement between these two temperatures in describing thermal changes in menstrual cycles. -Methods: This prospective study included 193 cycles (170 ovulatory and 23 anovulatory) collected from 57 healthy women. Participants wore a wearable device (Ava Fertility Tracker bracelet 2.0) that continuously measured the wrist skin temperature during sleep. Daily BBT was measured orally and immediately upon waking up using a computerized fertility tracker with a digital thermometer (Lady-Comp). An at-home luteinizing hormone test was used as the reference standard for ovulation. The diagnostic accuracy of using at least one temperature shift detected by the two temperatures in detecting ovulation was evaluated. For ovulatory cycles, repeated measures correlation was used to examine the correlation between the two temperatures, and mixed effect models were used to determine the agreement between the two temperature curves at different menstrual phases. -Results: Wrist skin temperature was more sensitive than BBT (sensitivity 0.62 vs 0.23; P{\textless}.001) and had a higher true-positive rate (54.9\% vs 20.2\%) for detecting ovulation; however, it also had a higher false-positive rate (8.8\% vs 3.6\%), resulting in lower specificity (0.26 vs 0.70; P=.002). The probability that ovulation occurred when at least one temperature shift was detected was 86.2\% for wrist skin temperature and 84.8\% for BBT. Both temperatures had low negative predictive values (8.8\% for wrist skin temperature and 10.9\% for BBT). Significant positive correlation between the two temperatures was only found in the follicular phase (rmcorr correlation coefficient=0.294; P=.001). Both temperatures increased during the postovulatory phase with a greater increase in the wrist skin temperature (range of increase: 0.50 °C vs 0.20 °C). During the menstrual phase, the wrist skin temperature exhibited a greater and more rapid decrease (from 36.13 °C to 35.80 °C) than BBT (from 36.31 °C to 36.27 °C). During the preovulatory phase, there were minimal changes in both temperatures and small variations in the estimated daily difference between the two temperatures, indicating an agreement between the two curves. -Conclusions: For women interested in maximizing the chances of pregnancy, wrist skin temperature continuously measured during sleep is more sensitive than BBT for detecting ovulation. The difference in the diagnostic accuracy of these methods was likely attributed to the greater temperature increase in the postovulatory phase and greater temperature decrease during the menstrual phase for the wrist skin temperatures.}, - language = {en}, - number = {6}, - urldate = {2025-07-02}, - journal = {Journal of Medical Internet Research}, - author = {Zhu, Tracy Y and Rothenbühler, Martina and Hamvas, Györgyi and Hofmann, Anja and Welter, JoEllen and Kahr, Maike and Kimmich, Nina and Shilaih, Mohaned and Leeners, Brigitte}, - month = jun, - year = {2021}, - pages = {e20710}, - file = {PDF:/home/alex/Zotero/storage/PY7HCR3K/Zhu et al. - 2021 - The Accuracy of Wrist Skin Temperature in Detecting Ovulation Compared to Basal Body Temperature Pr.pdf:application/pdf}, -} - -@article{shilaih_modern_2018, - title = {Modern fertility awareness methods: wrist wearables capture the changes in temperature associated with the menstrual cycle}, - volume = {38}, - issn = {0144-8463}, - shorttitle = {Modern fertility awareness methods}, - url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6265623/}, - doi = {10.1042/BSR20171279}, - abstract = {Core and peripheral body temperatures are affected by changes in reproductive hormones during the menstrual cycle. Women worldwide use the basal body temperature (BBT) method to aid and prevent conception. However, prior research suggests that taking one’s daily temperature can prove inconvenient and subject to environmental factors. We investigate whether a more automatic, non-invasive temperature measurement system can detect changes in temperature across the menstrual cycle. We examined how wrist skin temperature (WST), measured with wearable sensors, correlates with urinary tests of ovulation and may serve as a new method of fertility tracking. One hundred and thirty-six eumenorrheic, non-pregnant women participated in an observational study. Participants wore WST biosensors during sleep and reported their daily activities. An at-home luteinizing hormone (LH) test was used to confirm ovulation. WST was recorded across 437 cycles (mean cycles/participant = 3.21, S.D. = 2.25). We tested the relationship between the fertile window and WST temperature shifts, using the BBT three-over-six rule. A sustained 3-day temperature shift was observed in 357/437 cycles (82\%), with the lowest cycle temperature occurring in the fertile window 41\% of the time. Most temporal shifts (307/357, 86\%) occurred on ovulation day (OV) or later. The average early-luteal phase temperature was 0.33°C higher than in the fertile window. Menstrual cycle changes in WST were impervious to lifestyle factors, like having sex, alcohol, or eating prior to bed, that, in prior work, have been shown to obfuscate BBT readings. Although currently costlier than BBT, the present study suggests that WST could be a promising, convenient parameter for future multiparameter fertility awareness methods.}, - number = {6}, - urldate = {2025-07-02}, - journal = {Bioscience Reports}, - author = {Shilaih, Mohaned and Goodale, Brianna M. and Falco, Lisa and Kübler, Florian and De Clerck, Valerie and Leeners, Brigitte}, - month = nov, - year = {2018}, - pmid = {29175999}, - pmcid = {PMC6265623}, - pages = {BSR20171279}, - file = {Full Text PDF:/home/alex/Zotero/storage/CXNSXAFC/Shilaih et al. - 2018 - Modern fertility awareness methods wrist wearables capture the changes in temperature associated wi.pdf:application/pdf}, -} - -@misc{sl_ava_nodate, - title = {Ava {Fertility} {Tracker}}, - url = {https://www.avawomen.com/}, - abstract = {See your 5 best days to conceive in real-time. Go beyond ovulation day, and make use of your full fertile window to increase your chances of pregnancy}, - language = {en}, - urldate = {2025-07-02}, - journal = {AvaWomen}, - author = {S.L, Ava Women}, - file = {Snapshot:/home/alex/Zotero/storage/SMIE7YIR/www.avawomen.com.html:text/html}, -} - -@misc{noauthor_fact_sheet_studie_210621_2025, - title = {fact\_sheet\_studie\_210621}, - shorttitle = {daysy\_fact\_sheet}, - url = {https://dfxyyqidohkoi.cloudfront.net/media/filer_public/ff/86/ff8646d2-8d33-445c-9148-5979b60abbae/fact_sheet_studie_210621.pdf}, - urldate = {2025-07-02}, - month = jul, - year = {2025}, - file = {PDF:/home/alex/Zotero/storage/8VDBFS2G/fact_sheet_studie_210621.pdf:application/pdf}, -} - -@misc{electronics_zykluscomputer_nodate, - title = {Zykluscomputer {Daysy} - 100 \% natürlich und sehr genau!}, - url = {https://de.daysy.me/}, - abstract = {Daysy Zykluscomputer - einfach, hormonfrei \& über 99\% genau. ✓ Medizinprodukt zur Berechnung Deiner fruchtbaren Tage ✓ Natürliche Familienplanung ✓ Erhöhe Deine Lebensqualität!}, - language = {de}, - urldate = {2025-07-02}, - author = {Electronics, Valley}, - file = {Snapshot:/home/alex/Zotero/storage/ZGIZ669S/de.daysy.me.html:text/html}, -} - -@misc{noauthor_natural_nodate, - title = {Natural {Cycles}: {Natural} {Birth} {Control} {\textbar} {No} {Hormones} or {Side} {Effects}}, - shorttitle = {Natural {Cycles}}, - url = {https://www.naturalcycles.com}, - abstract = {Natural Cycles birth control is 93\% effective with typical use and 98\% effective with perfect use. Learn more about hormone-free birth control today.}, - language = {en-US}, - urldate = {2025-07-02}, - journal = {Natural Cycles}, - file = {Snapshot:/home/alex/Zotero/storage/MCUDFIHC/www.naturalcycles.com.html:text/html}, -} - -@article{weiss_confirmation_2022, - title = {Confirmation of human ovulation in assisted reproduction using an adhesive axillary thermometer ({femSense}®)}, - volume = {4}, - issn = {2673-253X}, - url = {https://www.frontiersin.org/articles/10.3389/fdgth.2022.930010/full}, - doi = {10.3389/fdgth.2022.930010}, - abstract = {Objective - Timing for sexual intercourse is important in achieving pregnancy in natural menstrual cycles. Different methods of detecting the fertile window have been invented, among them luteinization hormone (LH) to predict ovulation and biphasic body basal temperature (BBT) to confirm ovulation retrospectively. The gold standard to detect ovulation in gynecology practice remains transvaginal ultrasonography in combination with serum progesterone. In this study we evaluated a wearable temperature sensing patch (femSense®) using continuous body temperature measurement to confirm ovulation and determine the end of the fertile window. - - - Methods - 96 participants received the femSense® system consisting of an adhesive axillary thermometer patch and a smartphone application, where patients were asked to document information about their previous 3 cycles. Based on the participants data, the app predicted the cycle length and the estimated day of ovulation. From these predictions, the most probable fertile window and the day for applying the patch were derived. Participants applied and activated the femSense® patch on the calculated date, from which the patch continuously recorded their body temperature throughout a period of up to 7 days to confirm ovulation. Patients documented their daily urinary LH test positivity, and a transvaginal ultrasound was performed on day cycle day 7, 10, 12 and 14/15 to investigate the growth of one dominant follicle. If a follicle reached 15 mm in diameter, an ultrasound examination was carried out every day consecutively until ovulation. On the day ovulation was detected, serum progesterone was measured to confirm the results of the ultrasound. The performance of femSense® was evaluated by comparing the day of ovulation confirmation with the results of ovulation prediction (LH test) and detection (transvaginal ultrasound). - - - Results - - The femSense® system confirmed ovulation occurrence in 60 cases (81.1\%) compared to 48 predicted cases (64.9\%) with the LH test ( - p -  = 0.041). Subgroup analysis revealed a positive trend for the femSense® system of specific ovulation confirmation within the fertile window of 24 h after ovulation in 42 of 74 cases (56.8\%). Cycle length, therapy method or infertility reason of the patient did not influence accuracy of the femSense® system. - - - - Conclusions - The femSense® system poses a promising alternative to the traditional BBT method and is a valuable surrogate marker to transvaginal ultrasound for confirmation of ovulation.}, - language = {en}, - urldate = {2025-07-02}, - journal = {Frontiers in Digital Health}, - author = {Weiss, Gregor and Strohmayer, Karl and Koele, Werner and Reinschissler, Nina and Schenk, Michael}, - month = sep, - year = {2022}, - pages = {930010}, - file = {PDF:/home/alex/Zotero/storage/C3IU2JVF/Weiss et al. - 2022 - Confirmation of human ovulation in assisted reproduction using an adhesive axillary thermometer (fem.pdf:application/pdf}, -} - -@article{moghissi_accuracy_1976, - title = {Accuracy of {Basal} {Body} {Temperature} for {Ovulation} {Detection}}, - volume = {27}, - issn = {00150282}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216422570}, - doi = {10.1016/S0015-0282(16)42257-0}, - language = {en}, - number = {12}, - urldate = {2025-07-02}, - journal = {Fertility and Sterility}, - author = {Moghissi, Kamran S.}, - month = dec, - year = {1976}, - pages = {1415--1421}, - file = {PDF:/home/alex/Zotero/storage/GKF98BHR/Moghissi - 1976 - Accuracy of Basal Body Temperature for Ovulation Detection.pdf:application/pdf}, -} - -@article{thigpen_oura_2025, - title = {Oura {Ring} as a {Tool} for {Ovulation} {Detection}: {Validation} {Analysis}}, - volume = {27}, - issn = {1438-8871}, - shorttitle = {Oura {Ring} as a {Tool} for {Ovulation} {Detection}}, - url = {https://www.jmir.org/2025/1/e60667}, - doi = {10.2196/60667}, - abstract = {Background: Oura Ring is a wearable device that estimates ovulation dates using physiology data recorded from the finger. Estimating the ovulation date can aid fertility management for conception or nonhormonal contraception and provides insights into follicular and luteal phase lengths. Across the reproductive lifespan, changes in these phase lengths can serve as a biomarker for reproductive health. -Objective: We assessed the strengths, weaknesses, and limitations of using physiology from the Oura Ring to estimate the ovulation date. We compared performance across cycle length, cycle variability, and participant age. In each subgroup, we compared the algorithm’s performance with the traditional calendar method, which estimates the ovulation date based on an individual’s last period start date and average menstrual cycle length. -Methods: The study sample contained 1155 ovulatory menstrual cycles from 964 participants recruited from the Oura Ring commercial database. Ovulation prediction kits served as a benchmark to evaluate the performance. The Fisher test was used to determine an odds ratio to assess if ovulation detection rate significantly differed between methods or subgroups. The Mann-Whitney U test was used to determine if the accuracy of the estimated ovulation date differed between the estimated and reference ovulation dates. -Results: The physiology method detected 1113 (96.4\%) of 1155 ovulations with an average error of 1.26 days, which was significantly lower (U=904942.0, P{\textless}.001) than the calendar method’s average error of 3.44 days. The physiology method had significantly better accuracy across all cycle lengths, cycle variability groups, and age groups compared with the calendar method (P{\textless}.001). The physiology method detected fewer ovulations in short cycles (odds ratio 3.56, 95\% CI 1.65-8.06; P=.008) but did not differ between typical and long or abnormally long cycles. Abnormally long cycle lengths were associated with decreased accuracy (U=22,383, P=.03), with a mean absolute error of 1.7 (SEM .09) days compared with 1.18 (SEM .02) days. The physiology method was not associated with differences in accuracy across age or typical cycle variability, while the calendar method performed significantly worse in participants with irregular cycles (U=21,643, P{\textless}.001). -Conclusions: The physiology method demonstrated superior accuracy over the calendar method, with approximately 3-fold improvement. Calendar-based fertility tracking could be used as a backup in cases of insufficient physiology data but should be used with caution, particularly for individuals with irregular menstrual cycles. Our analyses suggest the physiology method can reliably estimate ovulation dates for adults aged 18-52 years, across a variety of cycle lengths, and in users with regular or irregular cycles. This method may be used as a tool to improve fertile window estimation, which can aid in conceiving or preventing pregnancies. This method also offers a low-effort solution for follicular and luteal phase length tracking, which are key biomarkers for reproductive health.}, - language = {en}, - urldate = {2025-07-15}, - journal = {Journal of Medical Internet Research}, - author = {Thigpen, Nina and Patel, Shyamal and Zhang, Xi}, - month = jan, - year = {2025}, - pages = {e60667}, - file = {PDF:/home/alex/Zotero/storage/TECKQLN7/Thigpen et al. - 2025 - Oura Ring as a Tool for Ovulation Detection Validation Analysis.pdf:application/pdf}, -} - -@article{guida_efficacy_1999, - title = {Efficacy of methods for determining ovulation in a natural family planning program}, - volume = {72}, - issn = {00150282}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028299003659}, - doi = {10.1016/S0015-0282(99)00365-9}, - abstract = {Objective: To evaluate the efficacy in ovulation detection of methods used in natural family planning in comparison with pelvic ultrasonography. Design: Prospective analysis of ovulation detection by natural family planning methods and ultrasonography. Setting: Natural family planning clinic, Department of Obstetrics and Gynecology, University of Naples “Federico II”. Patient(s): Forty healthy women who were highly motivated to use natural family planning. Intervention(s): None. Main Outcome Measure(s): Transvaginal ultrasonographic findings, urinary LH levels, salivary b-glucuronidase activity, salivary ferning levels and characteristics of cervical mucus, and BBT. -Result(s): Urinary LH level determination yielded a 100\% correlation with the simultaneous ultrasonographic diagnosis of ovulation. Mucus sensations and characteristics yielded a 48.3\% correlation when simultaneously evaluated with ovulation. b-Glucuronidase levels yielded a 27.7\% correlation. The salivary ferning test had a 36.8\% ovulation-detection rate the day of ovulation, but 58.7\% of results were uninterpretable. Body temperature measurements yielded a 30.4\% correlation with the simultaneous ultrasonographic diagnosis of ovulation. -Conclusion(s): Measuring urinary LH levels is an excellent method for determining ovulation. Although variations in mucus characteristics and basal body temperature correlate somewhat with ovulation, the length of the fertile period is overestimated with these methods. The salivary ferning test and measurement of b-glucuronidase levels are not good methods for home ovulation testing. (Fertil Sterilt 1999;72:900 – 4. ©1999 by American Society for Reproductive Medicine.)}, - language = {en}, - number = {5}, - urldate = {2025-07-15}, - journal = {Fertility and Sterility}, - author = {Guida, Maurizio and Tommaselli, Giovanni A and Palomba, Stefano and Pellicano, Massimiliano and Moccia, Gianfranco and Di Carlo, Costantino and Nappi, Carmine}, - month = nov, - year = {1999}, - pages = {900--904}, - file = {PDF:/home/alex/Zotero/storage/UAS9AGTU/Guida et al. - 1999 - Efficacy of methods for determining ovulation in a natural family planning program.pdf:application/pdf}, -} - -@article{wallach_prediction_1980, - title = {Prediction and {Detection} of {Ovulation}}, - volume = {34}, - issn = {00150282}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0015028216448880}, - doi = {10.1016/S0015-0282(16)44888-0}, - language = {en}, - number = {2}, - urldate = {2025-07-15}, - journal = {Fertility and Sterility}, - author = {Wallach, Edward and Moghissi, Kamran S.}, - month = aug, - year = {1980}, - pages = {89--98}, - file = {PDF:/home/alex/Zotero/storage/JREACGRA/Wallach and Moghissi - 1980 - Prediction and Detection of Ovulation.pdf:application/pdf}, -} - -@misc{pham_dropout_2014, - title = {Dropout improves {Recurrent} {Neural} {Networks} for {Handwriting} {Recognition}}, - url = {http://arxiv.org/abs/1312.4569}, - doi = {10.48550/arXiv.1312.4569}, - abstract = {Recurrent neural networks (RNNs) with Long Short-Term memory cells currently hold the best known results in unconstrained handwriting recognition. We show that their performance can be greatly improved using dropout - a recently proposed regularization method for deep architectures. While previous works showed that dropout gave superior performance in the context of convolutional networks, it had never been applied to RNNs. In our approach, dropout is carefully used in the network so that it does not affect the recurrent connections, hence the power of RNNs in modeling sequence is preserved. Extensive experiments on a broad range of handwritten databases confirm the effectiveness of dropout on deep architectures even when the network mainly consists of recurrent and shared connections.}, - urldate = {2025-07-22}, - publisher = {arXiv}, - author = {Pham, Vu and Bluche, Théodore and Kermorvant, Christopher and Louradour, Jérôme}, - month = mar, - year = {2014}, - note = {arXiv:1312.4569 [cs]}, - keywords = {Computer Science - Machine Learning, Computer Science - Neural and Evolutionary Computing, Computer Science - Computer Vision and Pattern Recognition}, - file = {Full Text PDF:/home/alex/Zotero/storage/IA52LNE8/Pham et al. - 2014 - Dropout improves Recurrent Neural Networks for Handwriting Recognition.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/LJ8EJ5PS/1312.html:text/html}, -} - -@article{wen_time_2023, - title = {Time {Series} {Prediction} {Based} on {LSTM}-{Attention}-{LSTM} {Model}}, - volume = {11}, - issn = {2169-3536}, - url = {https://ieeexplore.ieee.org/document/10124729/}, - doi = {10.1109/ACCESS.2023.3276628}, - abstract = {Time series forecasting uses data from the past periods of time to predict future information, which is of great significance in many applications. Existing time series forecasting methods still have problems such as low accuracy when dealing with some non-stationary multivariate time series data forecasting. Aiming at the shortcomings of existing methods, in this paper we propose a new time series forecasting model LSTM-attention-LSTM. The model uses two LSTM models as the encoder and decoder, and introduces an attention mechanism between the encoder and decoder. The model has two distinctive features: first, by using the attention mechanism to calculate the interrelationship between sequence data, it overcomes the disadvantage of the coder-and-decoder model in that the decoder cannot obtain sufficiently long input sequences; second, it is suitable for sequence forecasting with long time steps. In this paper we validate the proposed model based on several real data sets, and the results show that the LSTM-attention-LSTM model is more accurate than some currently dominant models in prediction. The experiment also assessed the effect of the attention mechanism at different time steps by varying the time step.}, - urldate = {2025-07-22}, - journal = {IEEE Access}, - author = {Wen, Xianyun and Li, Weibang}, - year = {2023}, - keywords = {Predictive models, Time series analysis, Time series forecasting, attention mechanisms, Autoregressive processes, Data models, Decoding, encoder and decoder model, Forecasting, Logic gates, long short-term memory networks}, - pages = {48322--48331}, - file = {Full Text PDF:/home/alex/Zotero/storage/3M54PVSE/Wen and Li - 2023 - Time Series Prediction Based on LSTM-Attention-LSTM Model.pdf:application/pdf}, -} - -@misc{goyal_accurate_2018, - title = {Accurate, {Large} {Minibatch} {SGD}: {Training} {ImageNet} in 1 {Hour}}, - shorttitle = {Accurate, {Large} {Minibatch} {SGD}}, - url = {http://arxiv.org/abs/1706.02677}, - doi = {10.48550/arXiv.1706.02677}, - abstract = {Deep learning thrives with large neural networks and large datasets. However, larger networks and larger datasets result in longer training times that impede research and development progress. Distributed synchronous SGD offers a potential solution to this problem by dividing SGD minibatches over a pool of parallel workers. Yet to make this scheme efficient, the per-worker workload must be large, which implies nontrivial growth in the SGD minibatch size. In this paper, we empirically show that on the ImageNet dataset large minibatches cause optimization difficulties, but when these are addressed the trained networks exhibit good generalization. Specifically, we show no loss of accuracy when training with large minibatch sizes up to 8192 images. To achieve this result, we adopt a hyper-parameter-free linear scaling rule for adjusting learning rates as a function of minibatch size and develop a new warmup scheme that overcomes optimization challenges early in training. With these simple techniques, our Caffe2-based system trains ResNet-50 with a minibatch size of 8192 on 256 GPUs in one hour, while matching small minibatch accuracy. Using commodity hardware, our implementation achieves {\textasciitilde}90\% scaling efficiency when moving from 8 to 256 GPUs. Our findings enable training visual recognition models on internet-scale data with high efficiency.}, - urldate = {2025-07-22}, - publisher = {arXiv}, - author = {Goyal, Priya and Dollár, Piotr and Girshick, Ross and Noordhuis, Pieter and Wesolowski, Lukasz and Kyrola, Aapo and Tulloch, Andrew and Jia, Yangqing and He, Kaiming}, - month = apr, - year = {2018}, - note = {arXiv:1706.02677 [cs]}, - keywords = {Computer Science - Machine Learning, Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Computer Vision and Pattern Recognition}, - annote = {Comment: Tech report (v2: correct typos)}, - file = {Full Text PDF:/home/alex/Zotero/storage/5MKPLWI7/Goyal et al. - 2018 - Accurate, Large Minibatch SGD Training ImageNet in 1 Hour.pdf:application/pdf;Snapshot:/home/alex/Zotero/storage/TRYJFLDW/1706.html:text/html}, -} - -@article{pearl_factors_1933, - title = {{FACTORS} {IN} {HUMAN} {FERTILITY} {AND} {THEIR} {STATISTICAL} {EVALUATION}}, - volume = {222}, - copyright = {https://www.elsevier.com/tdm/userlicense/1.0/}, - issn = {01406736}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0140673601186484}, - doi = {10.1016/S0140-6736(01)18648-4}, - language = {en}, - number = {5741}, - urldate = {2025-07-30}, - journal = {The Lancet}, - author = {Pearl, Raymond}, - month = sep, - year = {1933}, - pages = {607--611}, -} - -@article{gaskins_predictors_2018, - title = {Predictors of sexual intercourse frequency among couples trying to conceive}, - volume = {15}, - issn = {1743-6095}, - url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5882561/}, - doi = {10.1016/j.jsxm.2018.02.005}, - abstract = {Background -Little is known about the predictors of sexual intercourse frequency (SIF) among couples trying to conceive despite the well-established link between SIF and fecundity. - -Aim -To evaluate the male and female demographic, occupational, and lifestyle predictors of SIF among couples. - -Methods -469 couples without a history of infertility participating in the Longitudinal Investigation of Fertility and the Environment Study (2005–2009) were followed for ≤1 year while trying to conceive. At enrollment, both partners were interviewed about demographic, occupational, lifestyle, and psychological characteristics using standardized questionnaires. Multivariable generalized linear mixed models with Poisson distribution was used to estimate the adjusted percent difference in SIF across exposure categories. - -Outcomes -SIF was recorded in daily journals and summarized as average SIF per month. - -Results -The median (interquartile range) SIF during follow-up was 6 (4–9) acts per month. For every year increase in female and male age, SIF decreased by −0.8\% (95\% CI −2.5, 1.0\%) and −1.7\% (95\% CI −3.1, −0.3\%). Women with high school education or less and those of non-White race had 34.4\% and 16.0\% higher SIF, respectively. A similar trend was seen for male education and race. Only couples where both partners (but not just one partner) worked rotating shifts had −39.1\% (95\% CI −61.0, −5.0\%) lower SIF compared to couples where neither partner worked rotating shifts. Male (but not female) exercise was associated with 13.2\% (95\% CI 1.7, 26.0\%) higher SIF. Diagnosis of a mood or anxiety disorder in the male (but not female) was associated with a 26.0\% (95\% CI −42.7, −4.4\%) lower SIF. Household income, smoking status, BMI, night work, alcohol intake, psychosocial stress were not associated with SIF. - -Clinical Implications -Even among couples trying to conceive, there was substantial variation in SIF. Both partners’ age, education, race, and rotating shift work as well as male exercise and mental health play an important role in determining SIF. - -Strengths \& Limitations -As this was a secondary analysis of an existing study, we lacked information on many pertinent psychological and relationship quality variables and the hormonal status of participants, which could have affected SIF. The unique population-based couple design, however, captured both partners’ demographics, occupational characteristics, lifestyle behaviors in advance of their daily, prospective reporting of SIF, which was a major strength. - -Conclusion -Important predictors of SIF among couples attempting to conceive include male exercise and mental health and both partners’ age, education, race, and rotating shift work.}, - number = {4}, - urldate = {2025-07-30}, - journal = {The journal of sexual medicine}, - author = {Gaskins, Audrey J. and Sundaram, Rajeshwari and Buck Louis, Germaine M. and Chavarro, Jorge E.}, - month = apr, - year = {2018}, - pmid = {29523477}, - pmcid = {PMC5882561}, - pages = {519--528}, - file = {Full Text PDF:/home/alex/Zotero/storage/S3JV4TU2/Gaskins et al. - 2018 - Predictors of sexual intercourse frequency among couples trying to conceive.pdf:application/pdf}, +@article{owen_physiology_nodate, + title = {Physiology of the menstrual cycle}, + abstract = {Modern techniques of bioassay have permitted correlation of hormonal secretion with genital tissue changes during the normal menstrual cycle. During the follicular phase, estrogen secretion rises while other hormone levels are low. At ovulation luteinizing hormone and follicle-stimulating hormone surges are associated with falling estrogen levels. Secretions of progesterone and estrogen again are characteristic of the lutea! phase ending with menstruation. Gonadotrophin-releasing hormones are detectable just before the luteinizing hormone and follicle-stimulating hormone surges. Basal body temperature rises with ovulation and is still the most reliable clinical indicator, although ferning and spinnbarkeit (when present) are also quite helpful. Vaginal smears are probably less useful except in the hands of experienced observers. Am. J. Clin. Nutr. 28: 333-338, 1975.}, + author = {Owen, A}, + langid = {english}, + file = {PDF:/home/alex/Zotero/storage/J9ITWN5R/Owen - Physiology of the menstrual cycle.pdf:application/pdf}, } diff --git a/thesis/main.tex b/thesis/main.tex index b4553c9..cc7f4b5 100644 --- a/thesis/main.tex +++ b/thesis/main.tex @@ -13,11 +13,12 @@ \graphicspath{{resources/figures/}} \usepackage{blindtext} -%\usepackage[style=ieee, backend=biber]{biblatex} \usepackage[style=ieee, backend=biber]{biblatex} \addbibresource{../main.bib} \usepackage{booktabs} \usepackage{amsfonts} +\usepackage{pdflscape} +\usepackage{adjustbox} % Document \begin{document} @@ -28,7 +29,7 @@ \includegraphics[width=7cm]{leipzig_university_logo}\\[1cm] % Adjust size as needed - {\huge \textbf{Body-Core Temperature based Ovulation Prediction with Machine Learning}}\\[1.5cm] + {\huge \textbf{Body-Core Temperature based Fertility Prediction with Machine Learning}}\\[1.5cm] \textbf{Master’s Thesis}\\[1cm] @@ -68,6 +69,7 @@ \include{sections/conclusion} + \section*{Declaration of Use of AI-Assisted Writing Tools} Parts of this thesis were prepared with the assistance of generative AI tools, including OpenAI’s ChatGPT . diff --git a/thesis/resources/figures/methodology/methodology_use_case_contraception_decision_diagram.png b/thesis/resources/figures/methodology/methodology_use_case_contraception_decision_diagram.png index 329dde4770ff049db94e23c1fb867b0f385832d4..cd4b8ba3451b2f8ffd3a5e24d6296945cee07f99 100644 GIT binary patch literal 240656 zcmeEv2Ut|uvMwShW)MM988Co>AkbtLTM`a#xw68=gfQeJNM2sJFHdVuUb{LYVGmXP*Ykkcllff z28IQ@ckSHAz`#-h|1ZTd3$E}78;ZlfW;pCq+QE=iy0V>tL9K(RpiQ)PF(X)+GVtN# zDZlu{L@aC_hon5MUfd37nKLP7--#&#wKd zDtuz{@Y{-T)D-@uWNLiW7I{V9!rs;zt|*G)q=iM0|G`B~Lo-8r0%iz=CMQ#S2ZF5) zg={h8VJQl+4#tL7rj*M#$`vD9dlOT8$|V?2j1Q;4C$b&JK>iOWPvsZ>urY+W$#T#D zmW4zquMD_V)!O3172!a3lM{x}*srnQTby`i0js;!Bs6%1_RLK#9v zgi54|EA@gX@Pe>oZjVkwofWyDNjOfWN0dxAJmyF+F?FE4OMQ@NYimU$*iF4?Y-?j< zYE0D`^+7{>dt2wJkD1w8QJJGYV+VXs86CNxNs}n%F)e}#(Sq`trFF99cx4MVd@133quoIXUhBJOIW2)#E1&h-qsez!{`iq8-2*yMc&j3ku@qu>O26O zfB)S{Hti>Y|9>-_kc_D2F`~Juk&~FAx{#Kuyt5@iNK}HNc?Y5^idRz;5H|HsTYI8~ zt+}m@p_Kyd>JEEbM;jAUL}!599kgduY;En}B4Ub`rbMDE6*-2EL|eFILA17_-lfcG zXGmj=%BzE|qrEXz_jKu{l1ntSHwV8!wi3F}AqqhdjW9!bj6B1?)*nraN~Yidi1w~9 zg1xDgA(7yO9R++56~X3{&rP*w%1D3oPsHhI*b=~-fuo{QCr%}pYLXH%)W6Z0C>&GY zLOA<#NMWbP$#1u}H$?bB2ed;7a6qK=GeVg<)^zew3!{>cL!SRDR03BNLVt!R95`=9 zaB`>;{`EW5r^yETXHQPEW4a&{g@puR(#g=uk%|vBTgs&$>`BGY2wdg%;EtY|YFZ&`9HFP3!N*pBU;jf2s2}bs-S70l0{pM}8lI-y*^i z)W3EjI6_d0+}ep~8u^1FYb+p9(*LI%zNj>1j$i2T5xoETvQBnU|AS6Pj2cg3oX#N# z#>j^CkMKGuYJZK_k(3anM!SED*O8#&{D;Z^$zBItG-!*SmaQX1@M4ly5RvULAvnQL zbL6LmA%soT%U}kWJLpx+B~4@nQbce3_&Bm`V9;%EVQX#V2tn8mXA1(+RMXCo95p&a zx`WL76OraL^!@d`;{~CPn?*nUnV zu=rDi;i#_u9|N%;QciKyZJ`kR%PHsI4l!zM|IZDPs5UZfkrEU64aJ^vSkyFV-&(x@mgC=f@Z7c2_UDEV&zfswqX zs)@Lxyr$G%mwjd`QZ9Q8kE2^F7-9diSoen@fv)NPX)#=kT7Uf0(1fC*zhL=4U7AB? z_#0{=(+jZjw)XbOo|Bl!PE#9#A?(Hcx7(8_8h26MsuY?PVhlJ&ra|5CR?FX~{Pg>N^jE$fI%B%m`y`3LZQ3=$M|GC9e=)?Y#&eoS0n%YwF{};z6 zCmsKsy(6Rv0r5+Ft1~iWOC+06W@oDSgFAII>Bq@Iy>HM^##(_;CX7#_0T1@OOI#P5|nPvx*`QHCu zj+$!j7)On!Zq!VXmOr4GB633rHtk{e9_jEBm9&Gs4agtS`|yjLB9b@mDBC8mnEuN& z5CQamxQl>dIau?gALLPe$1m{w&@~_`E<#lR6?LdL{x!Z|l*%l1a_RpEx&}~RiUIU? zW1=IHDp3np1mcv!6?t0^ZvIafuB6E}fp}{y=BP{iG<5yU5>CHvC5AQzQk6LA$NntA zn#{V;rT&-Jt>~64sLLD8<%s>Vb^C`Um9EzQd?BYY@aGv9xsLc3l>XCx0a@pMeaQk1 zP%zfF+s61GSG2(WkhJ}Six%Rj$DuVeXxN*Y+ZfsygBwKL6UN&9`>L^M5guKk5LhEb z|Cw_`v-rQox6utO#Q$MC!|1Z|r@jnD*bg4`Kk3|lX6!!<^d(WwsF;#OE0ic!{;|ym z;?(#J<9*j_gRwwBPT8cR#5%~XkB9~w1^L1He?btABmrnEuN2(?>0jE;w6N(vQH~Lj z5TUlyPyzCHmSd!-HD?r?|HN{PI5ow=EHu9?tfLcq|7jG5`iDRFeE;2b^JygxDN#7= zg%cN-7L$^Ic6Ici6cy%^K@N(wf2v_H_kJlX7mf3AX1qS6Q!;u zztjt&*!xHNIr^I(BLARxdomWHDxF^Y!(`pyzk9b9_2L*OfL1msjVhSL2d!L^E1PD9 z)&wiyAE|BbGPQCtMb0geV-=AfE1zQ2$|v;(rRNiQ&&Jl?8cI$`Rg`ki2~O71pTr8|@}wZ|H{DGs~oNL)9G}RX{ejs0>U94t7>> z8j&nFf{hgcZitDjB3RqOVNw7AnG_UqXtA7bkrKQi8oJvUIUsFzD1@Q7L$BS6eByvE zu2^}B3MzX@ERCAV@}IVOHF^ecN9UfJhivc#MI2tk(}~cAqP}<7#bfnCs(-82dLnu zI!#KyDRR<}GCiHtVOr?*GldBHW&0mjS_sE~uCg_`{Gy@KFD={CP0uB$Nh)Tk)r5ir z9O)GkIp6?h^<$;szpl>s<0guz^iOKzZ>=-Z-9th5FKPAOY2yE7+4m16!N0btf_9^0 z({Z|-lTr8=Eb^!KRDNdcKP)>+p$1M3ttHVt2Gouz{QZS9Pw15Y(QhC~q2yA@7f0jk zKi_7e(1%h9&;NWM8~W0;QidejgZ(2ps5Jix#S6M9o_+xJzo~O>+6wqb71*h^PM1@% zr}`P2qQ8P7A6<|Tqt+Jx!#0ef>c6}!L08Pws>I(>ef@X+KVAB%!3Ea+(~m;POZH#D z(@$du@kamAJ@tRZKPI?BgXSp_Go{h$4@(=EC?0AM7DOQvL;Iua?ahtWBkL>joiN(J zVQUDvfZP;ALO0|O?2io_DA}~pWJFELEGa1^xi)|X&q%^LdBBL$7WPwzhNiIg!#>qt zh#u*LJR|=`k&DkPd>27n`YxxX`S9}ha*x#bjBwskP9tk+JU#bc|viKV6G#tar# zUc48>3`QoF;x!_oqL-T@-`VDFRaH~tOl&F>hRgf68xV=9L5ibRqpfTFq*n{+W~R=q z!p(+TldqybI9Zpmg`~($-Uwk{z3jp9{tjaQYpr>O+RK?(ULA~LyXnoKI%>Wuq*j=j zeI0djbOaXGTlc)mRnA;EYDpcRq{g&b5GN-myuNBq$U_)Gk@qC-&cVdR5ecgKs@KS* z)86LAM=fOBfpffdUuH3tF;3R=_lk=G1TNH>wq!QRA2W_WQ&7R@bmU$POy-0CSd%gI zcF6oj_``yN0w3p_g9ZlxIv2n(Y;C%rv4dzg1eax=C zUXY#V)Zkn5$W)_zpZcvD>OlMY%lPay>-hP7qB~mVawxi(4S#LUuTVuqiY1pU`=SlJ z;+mzIToQ;L;BJ%vX^1u3>g|2Uk@Ji4zO1TIYD{veQ}M#*Gp)w6KMAf0xHD{TqM1L} zT6X$qY3!^a3A!&-gD9f8m4!bs#LdGK`Gj@z%ey*h%Gz-~7oLvJVb#_!Ncybuk!AB{ zheKPY55rWn4v5ML;S!}1Wl+J$cAX$+7L%T}RweeDBIi~~DUeH)AR}9ZcU@+a`A47d zOeXW9ecL~-ou*aZ&u%kra0!s8>e65*4dtyU6q86|DZZWi}CUz#`|8^UgLa&nz{}f3#YfeO3NY`)ZUE)MPn45ORwfLtWtxoFBROdznm^aGE7Uu12JL5k7q7(lz89hJSPjk@+A8!&`3r)S;vN8J&XL*CZu@nU=@p4Gp>IJU;GN zQEP*FZ((xrJ*Pacp775Ygn>F@>D+^|M&6wqQ#DRs|IO=j{MC-podYu!GUh?6!ue>KO)7 zPp_)V;vpqmtK22`RCF;7+%q`=(;|-wQ6fk9?ty`d%g7ue<~nuP zf+%Z~mD&UG1&rg7<6KMQughN;v`22JSAo?lar>f_m8XHks!z*;TJeL759|-` zz|~#tKz$^}O!2u)r_DzA?&wpg%ku(oj@$0(PDmKC>_prGX(dy3q{FR~qF9GXwGC}8 zyhXun1Wu#@<}jCvoPg_0XL%c=yI9M&UI8Ilv$$+aC&kQ}4hqsLz{LC+b!*VIZjANw z-iJK5N5H`Y#r0)%E9jED4I%f&ZFQ7fR;ClzxEV8*kmX{*jNVqsOX4WXF!YKrfeEjQ za$Ad{mxocVXh8lzBtW}OFYkMXoBUSHk-I<1(7VFB9@``hlBTpu*_iOVE^oqq`h7B z+%$wmiGh^`Y>4%ylH<4(#6lV5`GLIeGSx-PrkfbP4m?~(4r4LmOn{(UzCb5s5Diff ziVw#tBGwQn{B2wVt1yKggb5Aj89b()dxj}n+QaCzuSxQpZR-&$7Vrmh|=dA83Uzs=1$H$AdowSX|N>dLN}D1|a$HbP3_zE>DX zBXeA~RVTNYsh~`wQ}F`#H-hb7y;zSDFT{M}nl**uv#T4&wk&a4#5ei$zodj3Ku&T6#=8=5Pthu6Af35G>)%l*p0AMi-F1D}Z_3js#7D(~ zIQFbwfGVRI@gfvebgXL}JCYj`iUygqmm&H-ZB|6qD&=D85_k)txP>)HdBep^tID+i zSUF!i!lr^BvB|uf07p*MRz*~ldZ2)o1B;@;?cn9}zQ5JMa77+Ilk;=#SVxG)Y^IjQ ztRX4FD6Jf9Fe@e8+lbJbaV3P6iUka<3P9_|eY>jAQHRzaT>x}IaMs(33C^%Do;JIS z{)txrMJMiBrFx2iR0N9FN{Zp%NgJKY9;{&<5Y4pW^AWs+-!_tn*21LSRDnOQdls#B}fIz|=bh$>Eg$(wG zh#h>)8)Uq%OFny)J`h$U$Lv$x3@n(-QC;renWyxYBKKQiUV{$J%9cB(K8=vW!3J3u zS%nleG)CW*02?7~0Vn+Wg(sMn$1^@TOT)b46L~RAxy>qDa z0pd?_f=^0))Jh|AeK3CZ<^_RPW@q=1g?)c1Ysd@o6o}+lSS{2+C<9@s5PRfB!JM;& zdR#FSQrAVMTF6BSWxEAXGaLXjG}B)_6FhYs=u1KWRgiCbHP)mevVT`K|1MjFFgh~IKU@{N2TG0tO~6$ zh~dr>0#9vmtZdEX)001d{c;sIH62tr_aWGjq$pB?13We1gK8*MQj_oCc_G~n?%HGG zg<@m?T#PkE6Z3VzQ7kZ7E zfIYXvpC*PQCw8|+xY=a(WOi0^JHNloziwTXb$Z=($r^)Gsb30K#G7yJJk%~z=Bg&! z7ImufsQ&lb)DLG@3^*}JvPW(sxul~Z?C&&XQ614c$`CO4MjIm_dzuSajDJ??#{Bq zpRQK5$_-}pHJ)1Z@O8W53hQfXvh6X%V$s6X*Lu+n^H@r>=QzT$p0p>n{V=D}Y=zLQp3FNeCyI0A zhMp{nTXdpvX5FF~{;ez}uVTvMg0mxq7^w)t&gBY0(WgU8qtPM(Q4myDXo9nG`E*&S zGIH!~WUN5aBcqXCf4Px1B3z5Bb9!eTUTUlq@^Vw|mPGB8)%l(_M~+S8eD8TzXJBcc zUaq1#*E$5!N$fNzo>X9g6%+C`L{TXVD@^--E3@w}g|h^?T^dPMo~{L4<}v9x{fR=G zAA`;+@0-4?bRVqKO1&tTGupf)mkZ-^TD>IZYu+f3raos=6U#s}Z^e7>fsw!7E0%*|&Y z)cWC}ktd5hG*&i$($UUS!yVkEiv(@^2;ZicsMd#qoxR;?mW{IL3MSKSLs;oC^r8K| zLz$4QYgK6dwbCu$iXU$F{jAg_{H?%eUt+&G6)nRV{1n;Ukg z*Et=oC040c4ZO3BtpDQY95V*L6g{_%fAOy>lp9ECidd%P`(1tflX_)n!)S|)f3#1i zZASC;)b5AxILEBuS9?-c&&Kg@AreO`w~c??7T4=Bm_@L+7u)(_j{2_6J&%klBwM}~ zdX5%4l*U$-7ngaCeN#$Xn4(XP@%v53D=t(D=noe~oCz>0jH+>8 zpM;cu6ZIZvcK4TCxV+k{6#Jl5_Ri498M9Yv_`2tG-oBHh969uCU8TCm*Q-XVr&V8V z`d!PjPs#I|-`0VeQ~*Y*#95p0rs0d_)!!l~Mv0w?>2V1O3B-~1_@$klBb~|WWgY_$ zE=aUkszbJiaj!H}Wtp#ow=lyvuePg*F+EY#Xz6-TCM+xS>5_aU^U@>by5}6;yMMeW z6?bZ(?E3t&GfAqFZ)0luTZTURM?QFK>+vnL@6)<%8AElr$l>B8tGvIJZcb)#-ZIh_ z?NIN~kq{r3J8~y?qsL(ECie}GdmNMZ^yEJx4#bciVb?m(9I!sn{X% z`}?Ne>ch5=`<}1yk6~AnUgh2w%@gh{D;nrhzen%u6~|iQ;t=nsy-Z5Ww})r^wrtha zWo}GSnU802gbcq<_Bh}>cx8_;LFQ9+c8i2U&F#J4Gi=(vPB7xSj;Dq(AGAS+_MoTDjxQrW1y2C*L}hZN0#4yZVTIK8IE@Y-8ZrS6vLu9c|9H z5vgzjJnunV=*UQQZUl^-({5gO;#&y;>@xUE*O0q!+OcrY@!|IVWKZ`mg>3>(?r-g$ zE~t{IKBCzrTlFYnb}ZH*M)UMY#x0$+jRaG%M605gs%yfVIg(uz z*z$(zdOZS#b4R~6X-5}uJh>Y9xXV<5g)hn2Pln%pq$)Dv%lpW&F7y1$+b;_E^?2tr zWgJ_n`Hh1Verl1N^w`(HT*2{?P2)|Q%udTU_14ZQIngYjzrjz}N1$bNV!UBukLOsm z=VLXQ)h_mZ6-x22G(`>RGVN*F6_c-bilyY)GX4JBS{(AeBKx`|*|*<*tF_WoLQYO^ zWi)7AVoPYUo~*)<<*5ad>ofgzt(#nu3)B?YzIx~FTlqHi>B7x|*VCN_UAI*yed4E+ zz%B;oeq~9K2yPCH(SSZ6jmB9*?exa zFMgZQ_~?yA?&}{Dg724CBe96l zxhn;)H;R9&G4|*9Xw{Xvm~`{1x}2-tfVz8!WbFV{*AznQyE?u&QTSmEU(h1||Zj#p>Z zr&*L`Hnfc$8RRe+pN_u_WOdi!`Z&3|uZPwBoWVD7&xY^yfpT9j`xg(tt7$iRd*3ww zi&9GTo|WuK_>|t?=+=`FFMWK&$@Ry!RJOIXWe-a5Ja8EOQW&p(qKUmT)jG+!Z|Uah zVn+qMS!=j+#P|VFaMqP#$NbC67aLW(&Q{ep##cBwS(r7r4+{Bth4hASbB1sqCxGu7 zZQT}G|M{#GcqqFyUCRq+mGl?a^z;8N2YU}f9(5%)I0QO2qDys&dqv-Xr%11PZIL73 zv~=XdB2R6m+aWWBDsp(#N;;3e6iOHL^RcY2yOpSrT#@QE_#mGa%T9)=)Al<25U&-3DILLN0qht1oRZ5RG7 zv~z6Gb7HXlR!Qfw}#IeOr#qHXV7J*b_STefw#%5dSWp=#-}CcWZXXRoy}OzVmwFws606zwXD8o!X8L)04LK zrywTbx^3c%#w3!6ZOlw~$}d$%)RgVb{#FSAWa#Q$>^H6qCnY|t-I)JH?c7THUGLPr zql1jJuDu0_OUefa&M<& zR(rcj*|k^n zqecci1U6Ji>IU^yT18Fp}hbUxPO9L;8k;7UnD(8lwY0#aSr!YM|XK9yxQjO)V>6D z_l;XeKR;FQgfL1k6XKF8$#+(`=m|(31oA~?!_P{Cn}2WrQiuQ#L^#Wp99`q|kc*o; z{DF1+jbIRfkkFy~Cff@)o%k>>pG|W?sM2q)CJ3j+klID`1P?o?{ucC}0HQdbK`F4SpuH%7Ph#umdf* znAlh=go{l>F_6gI9LicC_5L0UKD$j})5hN=fFEllbdG-)Nvz!6{5X-0#qSZ$i23Nz6cn2bOJ~0OIc_Iy*}?Y?t=%7;bHD3; z@s;-Jy#G>o4>*nTFl&FAmY}|Nb$tOs2?9lUuwPDg>8l}vYI`B9A3S*n0-|Y%bEopvOSL= z&O6;yYj}>ka*@a9Q+_<$q+2`_LxuV74^Hi$=g7Hv{280UhZTjS{L6YJoBP5X-yCX( za5t>yCXd<<9Cx7GlNH&KZk-8IbW+pDP*H6Ui19#KmgoGu*Xa?nc80yFPKuF&==pHx zdKq1}?zF0zKE}bqj~Y)lm35z9=Lh!L_qNRAp+%&F1;5iy@4Ee*nZK{KY6{`zx8w+s z0r1B=o68K;7+0;^HvEFbG5ktXT@{IvFSO==*>!C6;I~}k#93bY7B-|-eM2{$k-~u*EP@q%*6*y-bt{~Yz)GvJs660Gqzhy6D91R*MQQm(iv*)Z}Zy%NbU z+H-RHVmHOssNSl(iZG4B#AvAQQ_g!r=_@6e3Y>1s-O!5^Ah_rchI>-F)^ zize)3mpOcWu}-~r-{SDmCp@FuGa(iA8P6HY-^95+XgEo_DM2RevV%oSx%?v4)=1Cs zYzJ?KawNYsiof>g_hD;Ww|B(v2X6#JI`ytSf0M@-;&&&%QzI=>xxwSGNJ}9V5&}e) z7FlNZ zW1-_Y)*tWL4=LN)t%Dh5;p5+s>^QYOF0}MfKP)*OkVJ&HbZzppmHnJI^Fo6)qz)Te z1_6u?nbxpgiJ!<9(A(R^xJ0X~#bMyBdD-SN=^gFlv`rCKrid#!-|Kog4de3l_ExoA zQtvv>t|W<634+pJLY*LbS?d8UU5Pq=Ctlg6^J~oBvpe`?D3Q?T|Zsk zaB?HkctCN4EIm8Vn>lH$Iagi`g}eP-EsL{Cx?oMmwXkoGtcK9M9rB8e6obl;O5~o7fHTJP* zJ=-oGMN(a))xjyx6|yn8u|c-$-UhG0>LQkWvpn*xyMWzYDXKLt)MjIdVpi^OXl}5> zx#;gbS;^Yx4B51pE+ZAOvY^As;it_qb!0o=ca6P2to|6SuA!b2B7wF52nmzd_RCYl z;$`=d0?5+F$no|_AG;kZ(_fuXkY4Rq;O*%#{8B%TMf91d%&})`kP7EAL$W|Y5P-W|R1X{h&Af%s@T@$djRx0V~qHWnM zQ9H}|ot}6LfG88Lt_B1+wlTf#y>7VqIU}!=mQ zi1&)$`;l@_-($9*x8OE>bX|>Rhf2@da8h~Eme2SKe(-hpsxIAQoDX&wIV1V%qgz#7 z0#04t;Vw<*V-=-uWvCWQy?Ipy>$VIJx6*EfQ2mLK>U?s^rAa&C*4XvKHo-kmwrtW) zc5zx$oCb-{(kD^(BHa5c0h6M4Ovl~+Jxoe5pAX-rwr#AhZ_&i(MOB3oX&tVS?gP7f zPqc2nz>N>-EI-lEFy7t3sWeu+@RNA$vtv6_PVMev&ngn=S z4!I8!4M@#o15y-H?R6>K{Dx3zW5I9ZTX%eQ6#M2;L81an^T~O_LC5CrzcZ3H;@&ma zp(WTnPE$>WukLtfVq!?V^~1OP?w?Qb2s-%_92FS&ckE?*X6e`sHJkR2JQFUprh&RG z*0lobZUi5E-PSc&Ul#64nB0F5;Uy&uZ3>)u{CXi8e~Ce{_C>15S#;;CZ9fz<)bcrP zc(SfPOA?iOZzrD#`qNyQ%f9X7qZr65ZV#>Vv-7drbDFC*0~jAqWC`B6&AHyy{CEK_ zVlPq?@Ry$30gLqjQ7Yx>i!X&dk9&W(gTY2Pif%+}afL|nTBYUeLMWljGF64; zeT1#B>x|fq6Ut<;c?dmmn9W032%H`oi%AjJ!v^50pcbazkW^v`N*bnY@CTt}#tnH` z;1)g+U-$VDBwevTpL7*3;08euS>j|!g|Ocz#7rm7`-3eAn>G#DEt3~1nX^S0vVaXJ z3Nd8=K^EX3e-7%H+$RZYjmKv6n(v|48WjOgxW381g^tDzTXFA1gICBx<3>BHiUXrR zY@{BZi;)dF|5te81!n)ZO&hGh@5Or%3Wj z%iiQ^kayqZxk=}yZEfC1?CzoXEVTcyO$BBw%}#~HKQ9V4CT8<>%V1_g36+Ii)b`>% zl)VsWV0Z`sGCRC{<{Erd3$lx<&k(1Mfh+|`ml6SX4`W3|P}rFxuxp7$dW>iB!EPg_ zn2AmVER_etYvy`7DDC$-V7KvI<+5cn7$pE~NyCgH?$cCHvI3yq7rID7*DG1}=U=0{ z_A3)&9=KY7CnDXRGrx1Ji{nUIfQj664j)6SU3=QCaa5%8i)s+8Anr8p~nh zuJT$D9<2}3-nZn3ZN&#CLxm|UYv@9>cIWV690=w;2&OqHvI6A;9U9sJpdBtQGfe_V zVI#^6Ts75q6z~pZ4uiQc*fm90sJ^Y=I0i>b=mW0D>qS1(a1-tce1A7>#5ju_KztFEGi0~@jmI1YXhhKj8LG@P)~3l0ohuD? z_7lbyGfa*Fkfcz1D+Ib7UT?Idqp$-=w}UcY$tenbw;n?blBUnd0Np72p&O;@_hp>Z z+!?+bKw_E8#6Ttd0Mc1Qq4qXEbk?N#ge;o=(gDavZj%`&%hM9sknRp-U|?htgU#yxK{NP*TTfSz~HGq0fv ziu4v_sKX9_O>48ky!lUUNf9|r7Qz_}D*&BWNC)!4X30ZT7!`pC4rXMwD0<<80kAKR zY&E8HY~wx%F5u2l8>-BaB7Wk54Atulq{Sffz+2j9r|(l~AnBG)z&jI)X_Gc0rqSFH zvSJ3KDYT>rJM{ZbY45$jVK>@r)chVg7yfWv249L_3MQi4m_esM8qWhD965gn^`&{d zV94At56!`IMY)5K3?!yb-x}r>*#(4dw{$~M4c+C?AHb$n4cxZvfX&KzThh^9E7Y)% zThSmCo@0(^EVe8L?Bvv|0YQ*#DpbVYBdxfY=cy2$1+_oiF6%P1*x$DoY*MbNBLwuM z3k#Dosr4X5H>i&x8?QUE;!!%@$$y`MNX6+=Z3NOyGysZwmtKjq`~EN?NZ_>Dtc31E z)axWUq^D2mh%R`}sCgg};m_}dTQ5)vDOUo=_MM^+BGRnO9C3KMpJ(aY3y_4M3$2jP-v<(*>Z2BxC7TB2JwGxA zGn@j6a8Z;?medHcYeLZnNf>;6R-S{vG=VUH1XxP?D9S}N0#F-TznP*B{C9v>_hACs zY)o@Hc#yyj996l9ivy^^?gaG&!WodrQWu-)3kxqu;6^Z2xrhn?YCNX1Df+O`2WYtq z2+@(qeCXdaqbQfm0YJ_FK2;x?te-u47|RlW&@>|Q5hQY?D3>geSsS1aoT`s*SjhfG z5Zov41|GB0-*2+hI!NE(#T7p2F1T-TkJ1tEL)NPpw6?v7VmV7)Qm_k^0d%W!y+j+} zHP}d|FR#aXdwUy6!s(gaaEd1Jd9ezzsx)()|G(&_FJnHKtQTHTSa`vsu%Mvw_JKRc z$LDj?!6WZIX%K&YMg|JVOX2tv*Dm@$IF_&Uc1W7iTmJz`0je9S=%_((Ew8w-JIdk4 z)%V~4SgL(ID>ymLylnNW`~cg{04T*^n4BTnMCGz^i$B>U#x z)q5phT#AlqBgC>|JMegpBl+B$nc5ZZg8#3qHKc)BJPk zt}Q>Yv#PDl;>+bXY^od^?45NSCap?DSbJdS(<2o_iMDA#nZ%`0Cek3%zeo*i@7 zki-#B_366gEl&bvX6>s*V`}BK27#v*bG+1;!+Psl-m_;reacln-hZhQTXa-xx}|d< zrvNdB7s^>J>f7NYr+EIkca)Q4q;Zzwg@S$;FUB6YbFjbF@HidkgpWE3=8-hV5jExK zn!IxLigx@S1stT{w4>E6GTKPT8J?NxG>3r3PECx=sB6$pa` zPw1u@j87L^EG&MuY~@SwyUH%q7#^Jn83R6oB$s4G z12ZEdqY*_?aOwOsbk)Kbmjn(Uv5;B@!UOU zr-m79>*JRDBc`TF*VLLJ9^asz=!+hQEoTR_kWh&Kk|O12b-z?K2y|k%B2inhMX2w(fYKUd!hhX=CxA80$HSPs)6fw$= z@uKsm?(_z|YyorqFCIr8n5l9Tijf0nKj;u|iuBlPDzl@r391?7d{kWiSi&jFn^koU zo_*UH0%vPC9I%BWb6t-(wyl^5g8~rFsmb!UW2a4_cZdP*>TtBQo3Q=TQ0|KpKQV7}ZBw7yxn78N5Z56a7UDd4cKEdynM~v`gVIffJu7FPWb5b>vffkt1(}rV^;F4gLFbQJld0 zyqm$=x|EIZbmxJt{&JP6UHwJSg04BfJOn-VGu2S+k!JYEI|^tG)KO=FHwnQ^WObeT zEFyy+&z2(c)Pb(yRLM#-62iQ7`7`@7v=;z*G83u(l$OZ|AU-m}o7Yi6#^nk%I4TMhr`9A3N=?h6l*&YjyejsT!o>;8GUItA|5Oj)n>e z1S*CVs&woYUi$GQC8Anr0#sQlvLxF2#}u&B6TX9&nj5c}Len2cu0$cG&@=iTO-TW-cphMC#OpS_3<{i3h?)X;N$H1Bv#SLZ3cVU;JKh=x@90&GC=U3^j3YOv7mtzsh4xsU8C!% zL=52Wj8kqaF#<^hzo)Z;u$?-?a~-6vaiyPSfgh?v+S|3XyVv%B z1dsHK>Cm0x8j=e89>AH=?s6a{JQtEG`eB2qD&YbjYofsW#ZSsY1S;G^yfc+Wtyp()S9;XgQ!J>~FU+Y!VCR1RCaXUSW{ z*`~^3))I`sU^5FGI(9Sb*W&4wJa0N~Rp$+xsdYd&B2?}hi!o5V6L9!8;F;GHzEMUr zWWb3O1-v>-wb0;tYw6E*3|2|hRgMZ)<# z$1uJw75oIjI)<6zhLJswCp4!Jb&zp9xe+Pt&IU!}7)<;zkch$oxs+|MsbrDqBbCBj zvQiG{M=TH~-4UOJyVry9Ntp4l#R`NP6L29{b$8D{PiOQQi~&eyd}OyO?Jj;8W-QW| z$6>}qMGL;H2SlHt^?ZmX+*ink=2-Z^CbU<^=5Omha$!!1V2ez* zufhttM6%wRS}Q_Gy}$T;c`-DtEnl%>4|1ID2hthAcC`0sPc@^`Iz+^!{v|Lt8h!( z7jK*%+~U?cuS{@BNU(q)yL+vvndpII)Ax%IHf9MTVC4FYYjq8q@B3fL)Oakv`fN%eFknJNa-O`vz((9H4u`UFNLRo3OHte`LMQ{I8Qj?Loa z*|ocDaNXTvIF5+`a|?%W(@S<_nLIxX`9tr>GVVzUk(B2Nk}2GO{7?`!_@vQ0Yc=>f zRO%iNNUcZ1lvF^s4}8dP5EJ9h}0G_8<7w&5;NcV>YAO^5ne>6FbFOQt%Ym| z&qFrZiP4fM8+d)D+B)k3VVwJi>J5{$V}Dp89~*mlKU943A}O2tj_d0Ziz&3B9!$dw z{Fj(kSq;qr@81!yMc`{2zNXm4Qj+l-8Gu&GCcJdtfp4#rtC%ViR@3g@UM1p&;BS9) zk{MU09yMy=fBXVzLk;JRcF>ErumK8aD+A#Muc!Yv0Nq6z>#u_o_r&(iFrBM z4!@l3p=%usq%+`k?y_Qx<39T`9v*eJFfQ=SP#i3>IWdcF!U_8(({e5N=+Skd!`Go5 zE3P#azPp5cqfWV`!URJl#+Mrm!CLe*nZ!oplK1nYyskWznTmD=i~mS(ki+%q?DxdT zKv3Dh$-oR2Bd4%K*J48H&lbfExN^FdiWwGGGWedFv9a+6r@%vSo?9owzPax~w{%st zSR0)$eSm(_puIeSyw!|?5AVU~z-$Jcq6g*^?d|v3@#kLQqSpPVIxnV;$iZ}Pu@H3o z1u$qoS8z!Xp?H^FTv(cfB}NG}7U+ae3e2^0ipArEr8O%Q40JGZpviIBP1yTn{j!x- zPs8gn%fef}Oy(V#I>mHKi@@C&ly=+w@95Dik2OfiK$H}$sC5!|ee$CpG@K$r=tH*f zwXTd{7GH_=MLo>NVa4{{VHfxPcpHg<^oCptc~g=) zeTW}>CxDQmwdF7KRE+CS;!JgB6wuj526>Odq+joNvrR_!`}Lyq+?MvO5vkH%eG(Ti z5tz?}VV*W`xfFmWr=kP%of?wIEVJdACvk7MP$(dB;bgs4Jz?{7ANnd0>WWotfdv>X zr3#3g2SNo;;<^tiY^JMpRtscBorig;?}))AQgA^dv-xp7MY_9-_^3)SrvTyP7n9tv zF3ZjnVZ>cLG`DnqPJIM*TZo1dgh)0R1q4TN=r_4=b>sCD85AekYSkB-tpOb&y}J{?}MhE~Rct>j>MP31{& z*YyDUiL1eSm`vVqP=m%^GyAO_25~pQk>vE7Kf|Dn29Y+T!k4VjPfIYOPBI>tTP$9* zMq2a98B1ip0>wJ*M;jh|!c0L`8VX+Rjq(ANaWk8DJi1vzzjm7o>q!dsuv(B-R1tta zhS_{{9tM0gHIt5mBwjt}F{6PuDi+Vph+Dw)B>4Kv?esHJI#r9hd9OBzRFy9dYxk%F z*Mk8K!e2Rjvrl?w>^9xb00t6LO@Ea+%1_(6HP9u=g1yW=i5tH-T^4W!?+&=?&^(Hx z!LEE?R4FNVmS3aFOVgX!pph)ksQrus2^$V%%)g$8hxK_@Q8y-1{mQ`_G@M(@Qu--N z?IPu?p8F5voeR5t@J%*G8P5+vRN30G3>uuMIwH-O1D`;hCnSxF%5N3LDz5Fa7lV;M zuOlVlvIL){8Eq80k@jW=6O&ns{!-vD9P8(gFdD6Cz8zRWMfX!kow=Ntw@|-22b|}z zGBcU!e4JNGfByjlZ3^Z54aTp1X^_A#g2`%?mV^=LciKRI0@8da7;t+`+Iqtwa(g40 z(zXPzJeAh4leqBE40Qi|Dr{ts`tK2LW18WV&z?aF&X1m3Dp+3{jui=%uaz0cTk8+p zj>~`q(;SRkR;%ItB%&?X@rqQrzhE}clzK(^kjRg!w?|r2ZpeBOmRz3!ljJZj344()sfe~}PX!XD z;S)TGH*oTGQ$!>7Bd}QBnL3b^t1L4VZI(m&C78^LGp2O$O4?1bjFBz3S`n!Y`3Y+xBjJ z@cV()TsS`YS9k39ytuUUjE~t14Kdv-McO_wyH}qU*eG`7##(VUF|BNqxm;_`4CQ)s z)<5W#DRU@wINKWbAinJQ!wv=n_^Z80L^EE7eNp&|Vx}k2fgr ze-KeC${w~maZ}7ynfD5Zsk>gazNg)aSjsE%ad^W+cP}RMsdB)~%)ltj0R~#>9sQ2>W-0RK zsLWU8MRQ7BO!>KHPf|5$`^XS6ngA);G?}H7x4mo;Q!~5Gjng=Z^9!CUdveMPlZp~% zgE_Bw-wdVA*~12NMr3@|QHMFL6}c8o4mxFsn0+wkMn}SmN!8^!gP=}GHT$g)dBRnw zI)@vBHkxV}W}E{s9&3A6E<`h}!8xo!#~Q+1(+Zi)vt=Ln&YK(*`y$ho4G_8phosUx zV&h=|Qcn2s!i8atzQ*xq{Z>s8Daw_XJeZEO$?CRQlgxOzF|!mua_-5lcQ}dDi!>}a zH-)`P3BgJxewBNxAI-8pKZf>}27N}q+2F{AHrzag&N6iXgR9^GLmOp;Xj6Viru1Am zCdPOY=MhXO@SgexO8OTT)(~%{3z?X4N4322<_Dg9$HY?l$hx3y)8v>_Mu^b?o!mY{ zpjmUC1L$OnRBR3x>%Vjm%RoNQymoTXRKf%w_%nON`Cb@rSI~{yH9@l^tjX~n0<>S@ z+r61aD;4bnB@A;)0fAiW>CpC!910F`wmv)s2&!rji->+nomyju8`9cSjg`0(`H zMbol~t)}Hs+eI_R`>XZ$-80zUS(D@by~u0EOe5KWniS(J+dMsFT))O7Y%_Vz>I&bF zGm?E%s(jjPJYL60u1fFYR)alt&1aVCiS!F+wyl1f;b8vCI81|7uF`S{Ec3L$Rk=wG z(s+Ykcv(PujC`;I?e?P`dGknIXX;ao&s`HdR4Ep@#r1xwO`UAV$|LYikozuQnl2sZ z@ff-xZe1(hks@*Q%{2~zgF(la*Kpy?RH(BKk!QX0hqK-W{Ke=54bTK~WL0B8o(e-2 z{O4(I9$rguZ6wY9R3g<-Z2ubwfge5vkd&aA>*0R%aGbDi=5zRn=(Xk)x51u(n+g|n z9EnG7L3KI9sD4h#&ItaKxTHs?+i5m21;{4y;1?u6OVdK2ycLXk^NQQ<_O(^WpIyjp z$$tx2H0uw(TT~p3OXM&;yI=+Xrrgo{Z;cX^12*0V0tNkF2Ux*(8HF{K(uuXXo}Mc2 zw#$6zSS5b7Ul8o|_XEm2G*x1M@g4w8@8jE(MbkkN>t`ba#ScpbwRsn;kW6hc7#r!| zmvdtHyu(d_BHv}Azh~#d_XWA+PCPpC`N`~-R(+3QJ`=uR--}!N%H`iEOTw1})6$ex zA6+Hn=8SX*9Llhd9sUOY&7c>tfqEMQkJx(yXcbd=&dlZTsZrMzRVs^DXtk-hsc=Zo z+_=zB;++Daf`ceGG5YP&$d}V*53S$v>v%$Y^H=yrO%ePTbgV~Cz~_~Ag-Tk-z<>Yd zYlRZe?gwTi#WEY=+X}D)eJtN|_94eBjlP}nmuAo)g#CqQf*|(DavH}u9G@)>_(g8O z*GFeEK;jT9)a?Dr()S5c3#yAyg;nJ~M{HC_;zR zZ6taZEfc|y40MMkBqb4}KLiYaYc+R$93XgTAJHv7)3NhxqPdoZS$(!E|FY;iZKnk? z(_ROO8n3mj&#sKw&90d*aD$6`9xMRkwpU1Wy+9)Y2#$!*n?+j%S$F4klDN3jlU!~J zL|PE)voj9y@f~@a<(!;$rC3#}ZmmOWX#ltbpQPo{&Gl}*0dN$5U02Gs(a)!jz^8!T zxMj}W6E5qxqr*}o4o=t{^ioQ^vhJ*a76VBbWe6C+VU0$kxV()M6eS0-u?*X0`J>eSn z^qGQn7yU-x^PfJCi=ZLgb?~jivuKXDVb0Jhk?53p+iced7R%OYoXmYP02mF?C{o`xW>W6~pVcv~xo5bl zJ9e`;&Y?wxP@9_2ek#f9f%|u-4F~UP*G$|yyNE|vean~A{Siy$akA1UaigAp5YA_> zhHzd0_d=Q#NuTcS4%X4vQGyk(a&2RLo{Cpr<(^Gx2Y0tO#GV*b#%NIzS&tB zVBP%r!&Cm1Hx*_IsBi7RHaw`M=c@1@cvOsam3blVL7|6^1pG?`Y`Lp zLije)<3eT(;Kl>DgQ{&ZnmLb}a*Yjxs|tW`idJSj!?&1T^tJyV z&b~XI>c0O!qEI0-l9`d6O~xrBGNZ^$=olfhq^#;nMH1PgY%AeJc*Q~d~E$QU)J|8?c5^%mX z9ErGZFsN-@BO(>abiy%PMsO(deopp|kUh(% z6&?vXDE>|;?JnVgM`tSsyJ8$@n1s0D|D`O2cY&(*8fiLFi%gHPccnSM9%|iS7jR&- zVp)~(G7`+fN3?%T_OW)!X7njgjvJ&ZoBM3Z#ppOXjpA_`Q4+^hC;CCS?%f|>@Ddth z@zAOLd64&Et|L6+*07#loJ`DmokCnQrald{3y0EPACk+!twzBD%M4ovmmg*1m`{wD&)|b97W?7*(-koP(JCu8vFG2ov z>M!#K=}Vio(J20a7AMirt*5W%Of!Gnu|5sWq<6z&SvD8GX^!<2Ci!YBsY|m{;Q}|m zys5UdqlCtw$`kJ*18v!Eo_j?HGD`4Cf|Hrd)IW;d;9wIl3KV>?c#`Xesi6GB8Be`o z(C_07si36`GxlHw7EC4@q$r&<4OYA`@9{Rqp==2{APGZ~H_6tq0k zuBN}^Gqf3zSCg=+GD2OtA{DpTJQqma(9CRR#Z!GREEcQ^gN4~K`|_l=4!!9gM|wlx zI(3Wfo|n{!`)g@AwApRKM zgOo?2!6p*7&_faF%;Qu0%}$*I5z@K^tmCMu$IF9%et-Y;D$9MKdUVLL+=bsHTd&!N z%$TU^bBcPHbTv-Okgsxi=3NoeBD zeRA z3zNhrk(WWr246z%p_3GzhUimZyuhcptTcX4tlD4q?jL9jjVx|9WxofrWpV-8p9a9pXMuhFvmv^OEo*x@WKPjL4 zek{LJBk2=p9ZQ~MupPB$Mq3NjKWuBIcys^w}=D;d=jxH|Mtx((o?Q7C@WZ&Vgwrz3b7LNee~}>f3w=BecPSE2;4U&fWw05+&mW$BGZ! zGrRA;$iZ>(;g8-^pSldyT1#pJ7Z;kL%jv1{XFC(&xWcKaQd9$aIk?=n+zS^B`fyFK zWn97H&yGZ#Cu6cFewky8g}Cm&31bHaZkYzn5aMUU5O0^hmGd3i@kj2pZ;Yq0uJ*YV zXXCE}V#?MuVvkKtmAQ_9G26Z~1wQzWCpCp4dH6FrNW3L0cs{MEKpO3Ex<+m*Sou1!;w5>wqPZDnb zxgVkRFVbOccih)NE=OI}F(J|A52)j~_B9aSkrv5pIp{wjhI{Q;w!)dHFIVbpBiV@Se|lgDrdPm^WH!>BP{Woat5pfc zPH=g1oYh}Qo&WJAax%Bf5YDeEoL%-1IPLaNH7%%G|2Q6^YJ0XcBqaMEZmAY|;S;Uz zNHiG*IH`bY%Zi%#F}c$#(p|6&ST=W?K4@(4|B530&&`a+Y}`z%{k8vLsj^L_j7v{J z0*6{)e>ciMwL14H#u$)+UtoE*{LmS?Fmr=hZqKm?cQ4Gp5ABiDW83-Ws2Asj>HQNY z3-9yn+*Zb+df{zZMt-+b6?UZQmOyTX=A{F=v7*$@?v3GWo#~oeW)e9cg9EB(kCW6x zs_>2MT@zp|P~h44SUsq5kE=cZ^Q9)MwL`@n7qCzBEWtt&5oKb2`w@_MzY7=HIR8ZSp8ox}&=9=ZD z-(JwEw;3MakXmAtb|BqSag4=)y$5HUc7PH$nIW{iom-(HhJXo@*l6V^14h7Y%h=M? zNS_~u#cS}>SnjXfE^JYj`=zh4-9HE@0d*dTXYZbFhQAA{8$#*gimV?s`V^Qdd5v7H z0izr18|&6r(F`tsCMJYjpxZb@Yg;LAc+oI}uA=Yw?X5fI^q>nWU#W2`d~sx?IT{&j z4`BmYXQu|!_k(Lw(FBgSOJcY81!yA5)B1uv5qHUla;G)f!OHSZHE`SZx=mD=P4<>Y z6Eu!mxfS$88+Cm7e8S)U9#U}{7SjnIC*+F>lbSNO)K`qi58W1Ids2dsKAO8OR z&VScd_u!0jHI88Fqsb{8l@$xO(){CBU!V}cm+p?egoCKR10wa0;#4?$NH7Y!{ihR( zgA@82jvA5+Urc8>Oh$H`iaz<}*VsUI-DRL{F(713F4pRp-5oO;$TNI~|0Gf*pVg(-1~MgC^)h%MVG< zCA1Hr!@k4E#v8_EPPk)9%_TOE0~bdCGu)Op1NN8h?;n%VovJt1S>P*zBpBm@Am_#y z)9ur>n5vHO^gk2Fm!m**3Vcy_5;&nms{wcE3TI_664W9nD8net%)E;iLN7dw`{n69 zLV`D8XI1Aw_L`ms-+%|~^}verj{%_0gIPFE3W7%N)nft)`_8&o^kx71p=cMv4Rk?uQErE`%d+{d99;d@N{bI41xX&Ge7)$e-m;6S_MQIi||w5AOQ3VMq1E+tJx{kP{)rlrzQcvqapQA*hg?nQ~eL6?au$Tg*6r- zyZ~qIoJ&6P;5Z-P z0AisvNmM^X?l>Fm6~rj%PeDHqxgl&t6f3jcaht6I zSo^uZ;c|o#w3G*UG~RWn&bsc;_g9fdZ2+u@P4-pNIp+s0FMV&3ZjXO6>!%`ML2Wbl z6}*;8a+#2iUu%Ebu&D*syw(FRPqRUh)y6OFOK-WZoLld1r6P>;>w7VP1@iYgpBsYI zBS#KNY9BJP&@IUJF7RUEpm{BE23cTH`_6ui_g8~Z)kA1cePA5uX5u^aK7a0`{O>4Q zJsW7a-!XUoO5PAdOHkTnQ3ngwXN2_ z6qs9b{-^u9$3c&OMe0=^?BvkDFDzlGsd|ZZogLNtPhWVu4@}HsQ#XC$AHvNC}%F}nK>&Lkk1M)xP0c3BQ>i28gZ%v>{0(I7K54U%Om#Urm{?zyR z4*JOOG`7XY3is1S&{W*~(V0U2%puj|HGZBgcJH8k(!Au^1?~#y1Rmd>)2D;4?DDnT zyn<3;dO%wD2mR&n$JW5i4uy-}hTLG2*5dRVCYz`Q6j%pv^TIp6{OpUV1=BFP3I7_<&U@qp1>n(bmbT( zy{04&;= zo3rYbw7Zm^b-&V3mqwsgKMeaz=X&|vl7>Z82nDPFc9$Ad&qIO!d9C33aQSQ{;Bl_* zu|v|guDc&+s6N#p6g2(?pi4_=%)(hVu>!e| zMW+A=dNdj~$Q$V%;?XDLjPZil<90JY>a&LVCw5p{iy7m%^W0~CKH1FmOhsec2wugz zvGDFw+02kGfV1{|^~RG`YrY3zPu;dV^99n}>$D!ZI}3O|qwg}N5-6ZQJVWWMB!4YA z@J-q5)R5XpCY!o6m~#(Qk9B>zGnZy}mr3|Y(}gVtm0H(L8K_(Fp_cjx$@0n2y3um! z3DRgo%S&frePQ^`!3$z%tnclsvhsLBIXVZPTxd_oTcc^y)9DycG0C@I#uV9RTDOvzMNLGP@#^VJSK`bxLfZ z&~Lznw_pClrAlmPq4C*adcDCIQONI8zBa*R&Q;t+yA#?tI`3 z@p!`ab2c@VHyCq<;^{y>)>WckPXB(*_v^|qlq12Q(2{2SrQ?(QDtv|B78cPaubo+7 zbzKj<#aERKO`tjp%%w|Pql<#GEjOKd1P*x$^?e7?(~lP`%1#fd^1etqUFTfEv{k&i zJk9-E=6l1%lyd;lCW&Rup1oJ}?4Bj-=AdGYO2H)O{1dNNqTN6Fo7f!gF4t00YP0ltv8b+|CXUFIuIy09(n;KX9(wTwT;O}d7ir794#0YIGm%Yrm^*> z+nu&^2?l@Sq5)5g#-9A0!)vwqNg=D=xs?Q)E08r$ThB8H3;I1KK!b|+ntid(fS9EBVJKN^5568cf2)ElRV}CvAE)mlTZX}H$XzNpu0(C-vv1^h(Bf-< zk=`GdEWYL331k?EI`)EIh+W|I?-|VdRQcCtdnWlJ*}-Yg^UOL*I_lHlh{M4m;br&B%kM+}ZV-Hk!xI#=b_R>{(nXa{X-Rta4Ha?1z$0{uRVPKb_D zT31{;B^~|b!)^r;a;ho~?jE{aD=CiSzS0UZj2wP9F4J47!JwkY%ZsN=mBK3W(*6Qr z={Ux|M<(NO*WB4hu`n30+4W9*F%I$Wc9FwBev*vw#YLIuAT0`54dRrRgTX@!eQ!9vmPu_!hvm> zX;JQC8WV2G&5)=wJN{9D;tK>a?eD8Ot_x&>i_Lbnm9_Q;FFi-%`_YcUSNtD;=)G6< zJGB-y9nWOU?J4Ved|mbFSO~F!o}4E^C7CX2bA5Q; z&AnoF+@!hOx3Ig1E>HOo4Edo9wY7Waz6T=Tl}4SDksr1tXecz z>t2*JOb=>69-tvdCx_|+()*56uny2l_0Tai=w}87v;AoedSg&=Qz>SJc~&8aT1KIGAXh0(o=Z{)9l7r# z&21QpL(`#QAUk6qTrAHRw#jazrN;QUXSJr5vgV7lpyO#(>Yxtyjb5|i9cBs%<@9gr z6*l?}DCXXGRh}6rzTx_wNoe7bEu^*?gG3g6l=>}<*R8kg7B!>5Gi9sVpURL>qdsn! zq9og8+OS{&3cS!CWlEi{`z0_s19Q?TYIy3s51($~%OJrA+K<6ColWv?WTNr#jz7?^ zkQ4#HW70_6A&$*nUSF zF4b{$?NzHN7d*$G$G10cZKXKvoy#@v>?e>vbcc{Y#*Jl8JO`-z2F2V;v@bW%<~Gx zn?8lmcS+)T7+jkPX*%X|_w&=4MUP32C=7hnsSLQDgRd=i57{Av;@I!LOI9IJIlWJn zGESq7+3<00*amWE28g37+U;X)*R=OeUmxXSgP^d7zTctXB=ygFjs0>|0) ze!)Q0SZ53zMxvaXu(V@aS+CQZqLgqhHC6XaUd`+Ale@?@)hM=JIL3O`e#Oc4vjP|D zk)T3C8rwjuPGxyVx}~lYhCiR-&*W0U0qM0I)&vt>8@^u?45me3PZ!yI7s}TCY!sN~N1HRZ#@{0*z z;SI73Kg_NC9ET$zN0|lyppCmazuQFSPV0ub>GATIy0m|lszUeO@W6RyAqcvp`%LtR znM}ZOtfXeWDJt&;!#g^hLs80ow{_YmD2$Sb2}no4voJ6r9u1jB7VG(wyQO7JimVh% zpC-kk8DugVjCoF1O_w>rxR5j9&+iDY*8j=RS^+`)uyHI%&1{=bPP++#K@j`={u+i8s2}82Aq;u^2jK!%VLd z`xmqmev6S`Aix%O{w_P-yxb+mw?E^isxV}8y7W(aAj=Vw@h?h$O`dePf)%fwHAco> z$UKY-)1wzKM&T+M=gw#T`z)NWPcKuai!?c(`%?U(<;+c#xmR~?`8n%x?OOl2r>O>& zAq4oPs#C?pt*T0G|9>c#~6wy;x`4Bz~$1 zCNA|9U^0Gwg?pX8vY}nGcJIw_5;D%c2wrWcGv^D=vav6*ac$3A)4Z|LgCIQVkD~mU z2H9QLzRnf>ey}+XiVwFxqeY$9RE6R+siK<>P>7)YhCoEn1p!I1QDM}wk!XbP~8C_swh z>11WQjjKFPl)9u8c`(+fXjW;>_Wb}OP5Shh3;DtmY-W1U&s+fWz%<$aeY90%DDZdT zZHeQ90nrneI2q%=J=ra-LD4BBXFLdlf5I z8o?*SD9^JQFvOb>mZ>`RJ(?|%ZLi&B@LQj6xGjA2kZFePYA-6{RDo!SpyTadN45B@j&VE^FS1>r(oKtW46jLUqGM4 zynly1z;r^@R=zd-e&?@!DPx>Ex8i;JX@X!rkCg(!ym;n3#Njp7%DU1Y#{qUDx^Vue znMc=jGk|HW=NI$?7xxAC(A|av$wP41`W*uOrn+Ej#-UuDBN6~Jl^^}9h{_uT^g+!x zcUv-bq6_>N-rGaGt@WNfN}5y;rELaBiVvi{jIU2Qn{|qoSy`HbE-3uc(MyBe!^|OH zu3%@kaDzmSd!Dx6sP;~|y=)vZy(uuhCo2G!)3Umy|JX0Rj|IG^9+9pl1^Ve2;JN_W zfvqe=*o_$U|FOlx+i97D!avf<3A?B;w!hol)hh8eeX9Q>!*DMX7~*{9e)+pC=@JXj!{@xB97CEJk)wxkW@ zvDu#I8UuOxOahF)`Bnvzu-QNcR&;|eq(LFdNVx|2gA@zH29+*U;xMr*!OFW;b~J04 zZg%#wc3J|UzRge};2?`KyAh)H@j#~=8^9Agyjl<#f`&v{O&YCa8(=OEhtJ%<=n#ta zc~<>At!Q^C=%;s(qmHk9>s!vYEnOL=?qOr%F}My_u(maWrXX+Kcy@}By&8$Ah7V}+ z;Hdgz3`&I@xh9x;b&LP=>&#Loo>IK4_RWVZkLOl72-rN5YqTFh!w4@#NvWRYI-h1C zT|?)lwB!Ez`{eJowYv6K9ex=n!SysJzXoaU2%gukyHO_YsYeZ%7qhPftF0N4eBlTA zg4_pE7}6sGJc`dPlqxzGH^_(TVDyX`0oih5mEmUO@Yn+>1Lm><;5&~tR6uarQso9y z;yD-L9ZtdXVt#KR4sdj_yswc(d z4CN#x3{CD%T7BY>?uvK07b4}}uTZ)xE&6&l<2do5Ng#ZeY@m z5opLQ4_~5;vvE`1cife}Z2g@$d-6{Jy2Sz(mms!VXa%Ra2GKK}mKz!Lm|-=*@2VIe z!nBLL8MY3|t?f+%5y*Cij4xn}cC*&vV z)-5lcM@1?S^Xc(d0k+#HU!8JMTyd8WfIZajO7`TxcTv!)3|H`<%{@8^S%WmxxZT&n zu*mWlf(3sYxGil<5e1;Qz@Bg!u-5ic?dFpywS7Vk`CuoS-k}`)(nZ^XEP&up?n-H4 z^b(3wjI<{@}e2{uoC3}|sHMK>R%D4?9ShjLNd6Hz3>+)Y4%yX)@$jkq^)VwTu#a-j#&y2Qy#_O9_Ccx_@*fVhw z!Ox6Q6kwqWa>w!)K&%NvSrEp=v!wVGR_Q#4jY00Pz1QI{l$+L&v zCMo&4EYK@5A`5OKm?o<64p**X7$9wNH`y%#peykZhs)U6%Zvhgkr4U%EBO(mZb?B< zAn#bQ-CNZ$m9Jp+e*(ug4pg6|Z|b>GwlPRT%(hSU`#_CqFXpO?s+6Ei0T*NHJ<_z@ zvmf%qCf|+styz?s-OW5G7U6{AcBy?b^ynSrz%&cnj;6sin*Bv1h*1s zL5TWAC{)1){W}WGOIsSAhG#+UyfMt+efXv1Ri?-Lr=w5qiaRbs#Nj3A;sf_4)b*{k z_CeUKZGW~P9Axl8h)Mc6@Um?p-vfo9(Y(hJH8+HaQbsW+;-Ve;m$Z*}F`q!yNDl!& zYvrna(=BlxKNRMZp9#e~JefgIt|0fbpt@JD(F@bbiMGSI^5`PbZ_tsG`!Y<@^C!hk=qIh-b%X|95Lbg!C^*+_fc(r=HZ@E3RG^oNM6*Jx5a#}mFa(BXYV54<15(ak()0T zM;tc0hx=420VAT1(WR}#;xWmf_&!i%D3Pn8n#|~H*mx=hBW~o%k?UNNX4lrqwDNq; zj6InMQ<=4ttr{c&B$7@`(zsrPzsHjZHZ#AM!V6gJMhaMLu+H6eNiRjH8vLbmM%OK% zMW43vTL3IXKnc(I5AFAd0J~E`^)j8c=g&fM%}G6@g>7{&IM3^$#Nj$hB#XfRE8Os8 zHK^>)2&=1L%Xm!Xr;An>qZqJIRs$Zqln>shv{zSD)9mU)*lX6oo17=i;+x+#qC`R% ztT?L6Rm@^-oef!hqg7HTUq18SF$elS1o=M0nH~HDIk17u3`!zBpUdEyk5K~a92K6M zi`tcl=#VoKFF6YRr>=qOrHQq$yV9cWCd{2=uP+MnTv0-O3bYV z=HAdaBJ|-R*r^^Wa5j*e^Ii!)c|-|Mqz?Z)|LvpH*{@cmi}wT5)Hlj32m;cEP0EHi z-`fJAt_LU7A1;jNGupj{g!>v$BiJq3V3Hj5nKP~Z`MNN^Y!UFlDPsuVFh-4aq+NPf z6c6(9D=T=nzduKACqeaZ5fsdMwk?qtj0O`5-O^P$$~+uAtjAug$V^TPHSZtmb<%X# z5SIVwtchzdN! z|Fi+yx~vp`l$r1%p)^w`IH^Hqzc>5psdqx=kHEI3WOB02s5rO7p@UEy>X8D1+IoGy z5T>37QM}Z4A3jC2k{7p3SG`t`>mYA)S|}D#!3V|qdimi1w35FSS`%|5+!%&OI|Td; z03(5QV`$*2GU?;8w}XZVSxpOAi>1BU0~M(-D>07J)wxsZUM-f*ef65l?yweP1#YsTJ?-!faBP`Jvy8I!G5scx8ZFR zKe26ZejoEhl6Lh`QQ+d?M*KV3NrVDT@ml8{ivso8$eJC@| zdk`dRsE9f6QH24E0mb9pqGX_utkLasd!si$gk8bhQq6JMntS z3Wrff14_oH-sASJL=(J!G0=^-$th{HzkPU~1*T4C$w^&+kPXL`qO)V^I25?=*ZxQ> zfJ&vn`S+V=KEpKHNDym>z%$C)e`%A~8|$+=5gel4xA3>YE+^nVCV8W2awPkz`k%iQq)|$(OfpL7|w?>PxM z!65kWH>Ika*BgKsax!Q6d7N;%FkrRD<9sff_Ph( zi#UB3F|$J6R%#M6iy>xy+M?6R56oPhskqh?p$(Dz+ZseU{y)FE(bj#qF5>X|-qF4C zGHZ2EdB>r+4l0dqynER^_6f$a$ASNx=pHL>-CMxn{e+U$;oT(QWdk&f&fE^5aSgb# z-4^*I?RTmO5*m`+q1iq05RX{gME%N2z0FFzci0eEaw++2dxVlXP&uWY?drA9V?cxa ziSDm(vrJI%OjmpW&c774B|d&9O4l&A&W1W7F9a2+d<6(a>Qs*?DxbXo6YQbN*993U z?-=sHbO3**Qe*1?N)YY@C)=BLfY%w7>E;@xUH#k-7@_&DQMe9C-^QC<5&$(wFAv^3 zD)Q^iUd>RCwcD{R`S8sdOG@bek>uI9D&RMe z8vxpp%ZLqGSFc@l7`SJr;|WEOQEKM?o#;yu%i(G1>5Lq?Z4d zqa2(B(zeRE0)OfV==Kbho4P{aLn!w#;2z`%og9Ny0!56Ofq9KzF}Nr4KQ`Ir=8>^l zkMSFoEz&RxHvz6F1Xd`+w!0KhB`pMv{XW0o3IZVQUemXaFZ4|y;IOUl4G4rp$#9l| zGtO}SU~_!i;m`gM6>jbZ?kpCD#QI}k)z-L{k7Zq=Zt-)LVsHCBP-`QE^aFG61ENk}^i3}^(HyR%F{PR~K*!>EXSD3n? z8dldEl&_V*VqLMNvslGc{AVLUIv;d(gn9(*7C|wXF&6vjRh;9!-@vK~uTP>pyqUqf z`!5wphAsUWi8f3WOD%KgkKOZ??9b0&$Ys(7njeARmA#LuF~IYQck)hEil`6I$&P-q zW&;N6A`jy4?OrM%akV&r29J>fD_4)0M?1WW>>#CM+f{8^$KD@fT?aSOa_p$^u09Zc zg%3&=`QT+zX)%z?r?smIu)j$hkEU>~xWa)Tg*a z*;9#bi3n&$S$hC>zXae(Le`}_5>&E)n`Jft;ytdc z#uf1if7omX+A+gMd6KO1OiQr*s-G(!KZGt|kHn2tqj1EF&peR?Vbl0(P+sAC<{9&(s|#;m@k$L`a4vPZ!2&G{HY!5}j(hiBzgqGT zw)^dp9|?83l?btNlhz7a)1qJsC!@ORJ<|KtqIf|)8SijLPzSJx;T8a!T>ak^S_-1n zCy0bML#KgOIc;`PYH5VMIq4wZqu0y^O3Eb5biUv0IZ))EXD5c!P;^W)eHLN@}V^EUgg)`!2r-u_0QYF6ZG{$=j4b&WjWoX>zxFIG?S z&ej75<2J5yjZjh-=sb@SMB^FV%O*k z7BPz^$n)L)=Gx3iGoPXAFNbILK4ICANcR$W-t9HcHq@Ou1`MLAwyYT6o%fJO=r!KG z$Gc$DB~I94Ro}M{WE9vMh5T@WiBZD~=HvaA-5-g9jb6JF0^>em-gUQ8V4PTMjvD&F z;<5m_y9IVdLZi&FwEc+~=Z?D}<+a9V%z~-k^sRLlXtAso>q2p@r2sHxJ3;+cZJq~1 z4%tn15%~~n0_n^9Rj$Sl+$xDSVS7TqCJ97hx~0v>2qrdUI-t%=DZR-epO7(|9M8DM z(Tczr@9uOurk>MU_&If&9n8VR0;`9eE@i;DQl1|)W`OciJ^`T9wnV+$chQ|P>+?Q< zxtPSQj;TV~ULF*^Qnw|a|N5{53Q6}Iy?4aw^8|!&sr#J(jJkt=`^N_QSS`}n{vn8) zO-S6Fh{f}f#0?{IgFQ|0CQp&`srMpe^>^nj>f4_rfY{K7IEo;UPwYg|BLf>qys5hX z7X)n#0flgkcr;nj=K$uQ>CS+K*;{~{Cf-5ydJt5Nc(mxT8N|9ydk~n4@kRy!vA6PE zbL}cS(Xa{fk8dj0K{l>&y2iE4**AqZz>1t`!JS^SWv7$Cid>ZY7|u%6OvGME5%l2S zZtubfuAZ+)R`Bsl($)8_7ZpKXy(2>AAuCU4R8bj}5()ucb}xfURZvS0B9z+ zJw}^u@u6xzT#ySA$4K}>nI=P*CXm}fRK9)mEMzKTpC=A?LS21nBFMHFEFfUp`<5V& zDHFtGx=7tfYm!;r;txy4m(=1=n~`B&u60Qk@<^i~dqf-lE)a&IZ^j>Q%u289IwNGG!Z$k{u%kgSXSp&KL2Pp3Q`T;}rnrR(9_#i%OyvMpGQDhW!y_vX$ZEd|Mqw?NK_j6_1p z)vNM1Hm7QP5Yq3KCey)SMGHe2xcSW$&< z6L{e2^;z|nr| z{tX?;WK*_oT4Rh0yJ1UQ`f}=|h>`X^kbRk6o$&*(rCybd$!iVGrl{ukDwH~R39!k! z++K;hp3%o)XA@5q-GajB=EBNBc3*zhZAJK(V%5g_rk^0WDil~$6H>^zj6oo|A*h*g z5QiH5Tq&@PF4lMZh>E!-UR3Kx-%*!_N;2!o8e$VL1v3PMr0n#+vZh18N_Is+NGg{< zYYteKo)-@j49^lK;0CSzKIN>|vRJsj^!+gq!hH9qRQnb$Vvrf=egk=0x84cwNiDH^ zPLm*c;-c$7_0bB~cf2l83dyPqnF1sC@{6SUG-8Af^SVZZE~yo@C~K-=oe-9UzB8^r znO@212X1SHD(cU|*=zM-ZS;^tgp{M1T7>f`tR?)ra*CCYudKTu)LT zGEmvLicfH#BoExymbvxzm0c)q?rCx$lpyOvs^B5)@@x>yC+^x7xX6QNB_GL9U+~hJ zPres45r;Ih&+cY`_M)~P@+PSKV0{oK<L=no39v{n?VFQIflDNi0AxDs%R4zH0V*ydb_3iT@Ijvo@ z>vS6%$=0fYD!ay=GfhHEBK562aRlL$sSfUi(rv|q!VrJSB!WHghTm|UF3TJp8wsKo zf{v*A@pBDsf(ppn%X#gfG1U;;$ka`o*$s!GSF+oAGeL8Bq{1>% zy_6P`%9qm#VBc;?E6@WIa5uCc#&!m`>>=DM*-o^Ozco7bD2KqTnITt7yAbT<4aJ%u z1tWNP#R#r}M|@z;P$ddjRU}|fx8e5^J_!z%MgrEJ6R#PC1T4euR>E$qfI6Ux%!?fb z+qWl~5EHxPP{2s*g26f{!v%=6C{Y8h} zk~SuuXvhevJ1i0`d?#Apnnr3_GW`(X=R3zY(6XxnaIC+{dX;j)WbOnxij(2?rXfCq z%n+pvkCZK(HR|VBIQx}NadndhKwlqkCrqDK14D%LG1WX04ofNLwMc=lNEO(}yP1zp zp#WatA5f>Qbw!*9)FRwaaFDD)K=R}IFih|^#3*DbCL~7jI{=I#n0;$M1)S%I1wqW` z0CwQ{Qg1$o3(92IH|uQ@0_jOSGW6>VfcZto?>wk@1P(LQVC%rDSy!I}7ji-Ao7~nU zgc>KWLBJ3bjRfl2CBx>klI665tk)k!I$ze>M=lYh9PY9op`j#H{CYrtilu#A&^WkN zk&l2;QAok1-afT1C4$(jmX;Zeois-E%2h%nT*)2|RbW?Qtusg}UPN3@C3s;Hab+Ol za#6zTodJGrDdK*yhS6j0Ayzrm^LUaW{@YZ)v=`wPxe?Dlgy84OHqam1df?I7xpRCn z0zTQbu1`=nkcRkW2Z(Lx#O|IX3cEoBMahreI!ye8D)TrzVxd*ho(8!IJmUIOtydP#Z)5(w|cr=B7@-d!A4R*Cw9N< zEZK&1VsP7@HJQZC1130i@vjALXoQ*?xQVE}9QMF7))mwKN6Od`@&HJqO41IM>gF>+ z1a9s0f)J_Cz;GDhDQ+5RtQApqCf_)KG6ERRA8 zq4Vkgg3hCkwWwd{M^2%TgsNr*!VQX|68w#Hcc@@3f{d+_q_sqe!&=5NWt@;IJ~Qs^ zzIOgi5rD9N!JnHs22HV(&9sQ!Vlxi_%jOwnt;?xY1K;#pgXnLkaA10;61Fws<3y$1 zPjvJ>v?;I*mp@7pZWfpbWY5?$X;;E08$w(Gv}u;(*`@&TGxc>%JVP`Pd8CgP+OIYz z-A176NLo0O?7y|2^h0PTtK;hz_8&$&UPQNdtnP9em2K1MW0Tc+BpC&Fye92E98nlO z@uvO{I{yHXFGl%<2|9pn7WO!7AZWv2Hza#*q%xA?UR74GidTY>Uqiv=Y+@#~*0F)B z{h0Nx#b5MaWRLb=unLp3LrH{ZZXgdtN;=g12W|%_X`F-fk=1j$fJ&KvGwRzm{`PL6 zB5VT`@3fW7>!)Qzn&loo3L?}d#r#f0Sd#i31LcNQ2~+qF8()IhayBl2A@C#l02_SJuwz6a+x)hZeSf9&BSxVg|V0eF4A)q$x;eWUwT zFU_iHz-5Y45x!r7zTYAa-*4gPBPx+(WQ+^8kOvSL283w_s8snj_J9baQdtST*ZizO z1N?T3uMhuQ6xaa89#;aj1cQjG3R&}(*z19fy8+A5(>p>~Z5$XWo>LSI@t#xtIBx>^ z)i=W}`hNe?Ay4>Z1DPNYq$hk-Sq1VcJSlwR+UeA1!6Ej`dqUZuyyGmvqoDT)gF_sU zGMOb+)7Lx0F(k|J%?etAu)&7iytH;@^)PMHv0wTGJVM%Rme@`db`6hEI5kV^kyB5N z9wE79md6<$A#fw=3Skw{(8w`a?J&hkojoCb23ecv5pePZRRehd;y?>$n!N{aqkLxu zQ+9%b5Q>iCE5-JB&#(O*(u9rI>muIlbd@6YfDiFOc0rEZ=R*%cMC=h9rC6iXVIvTl zZAm-G;SWZCRDr%C?){E}fB@=iB#I2~NMxJjv1BinAsEKg-GJZ=&_c8JkIB?Lk{^>F z-bJ}7D-Q|gLiksxITjj*5biV&cp{qSXm6Q5&{Q8@kzhS3XaG?xGVZr+TvEKdneh8i z-Uv^qWOJ+`e6oSO8G_o;qB^~vmth+b@tc$gtSJX4QGA-5s^zU?djf&2$kb3+yuGyj zE%6hI5=b`?*yqnNp@8JJJiiaYl!SP>LUf=XWC{GO<^D2Yqe3Lt8nyzNYl$dxjYAkr zEOo&v?rR2Juo&ED=L-W{fqnXeCPvqDex3)f#T|?_DB7Y~ePo=u%OFfk{`h?b`An zZxukbtG|Y_A$e-(HVf(87=S~azlon@EEMriu$U;=jghhWfHl^;n>P(r!^&6Dx1C_`oatIr*FgHsqx zJw9|J0-Qp9#f@uc7D5531<&XX+6n|pN04M1%vJy?`?1BGbeT1%G5>k*@rmu8-uQ_z zM`k7ED}-Et>9x1e=fH;EK7o!(?wTZx8#0Sz{vzaC5QI1edJ$j7)D~Wz$R`-{2d*%bCiHeE-etUeIo^>$`rF0W9lY$jG@_MD<^GDd6x=VWzLn>hav9HM z`t)r#p3&P0IQeW{T$&KMU*s?Nv)|-%@;#+TXBJ(=woLUr_&L@0vo@e^p>2=`%d%6q zh!+#QI=KaZ9Ai@!v~B`UlR5^c$LZM}=0>i{!A?W5I_D(hIVDrupXUZoT0VSiidp** zhW2e~3j1;T>owvB=ct4doB1OAi}mY6)zw4Rro%i)4t|=DZX{XSzr%|(8)ljOd-`J1 zd0xV(91JJ-TWVDy&SB-IBh!>V_w2Al1@*j#lMrDJ zSugVr_4nZy6VAK{e?**9Gb_O@(fjzJAaNku2+v(7ftw2B&*jLtXtzn+=B;kX$ zSO1cx!7i}vOg*qh*4!PBgKM0YM@e%BBSVAJ6O=#Xvi4D2P>^9FjPF^b-kRs3^+Ye| z& z)M2J1mg-JT3MwW|9toEn1cHF^&UzSJi=BJXp(FL74H_FW%vT8#nk?)=x83ZIA0v+2 zt(PM^cbLdOFw4vic!R~;`)E3Fr2zYm)-?t$Vbv~1%m2|}04a29>`_786h`D+SSt&o zkxXfPQa%Io^m|EMy%Wpo`Xy;XZX@~*w`&^PO}Op<_QhI!kp~c&;kL@}B2EN*E5I$` zLS%vexa0a<3R837O6(DJeaHWK)k+~p(V?>qNOqkdQidDI65_o;3g2^}6w%Sqdm5A% z6!c^y6E#=qBHZ|HJCY?C9Ck{?v3(EpF*e;qaYN7jARZ{?dB0UmE?oAtO!TRaBuWgBPGQtSiRLY> zN3YS$d&E20%7m>+EF&h(B*YzC_Ye+?j67(2B&#etiQJjqS};A&R^@v8zLcre=juey z;uh!tyiQP!Zg3_O>+&K7VW4Uys0tbS_#DI`8;x=ZZ=wn}#n||irM>Ya z97za2CD02s{(Bh+2R#u^3-vWQ63^d}i8d~(sYaYoz2gyV{O!TuC8CrQ6~60)J*@bK zf6s8kiKr7ep|$7MUm@R#a@yAR z%g1=^zDG8%k)(9Vle>?Ix8)5|qV4%72h#ISfsEODr2vkgML1b$?Fhn+Y-`cw@o`X9 z>k*`@WE>oUm6d@2K?ZX_5V-5qXxnrM?ylfz9w4^p<@MT><9|tK2*c zFe?hih;-Ep;#oQHfiaR~G_Fkrh78Bw_Ox~clAx7v#hO+@#B1zj0H%%2{Jon8D4*d( z2Nj8$NU!o|+M#9BjU7>;rM^jAzLU^hr)-|vT~~M{!|6gB9&{i2r>|b&;g(PeS8(Ph z9;QDWW@p$n7BP6^dj>yoqC6s=2TWFrge$nTHAxF`DCP&bsrf6_XG-yZ!F`K9&1H-M zx!y=n;slZgY6JVC2J6Ck@EmT`v~4$$w2fGeWbBn4_+f{p@4MOlX_@R)O<{5vFB-xs zH-z|t0@}f14mM%@Q3c;CBpM?9$|hLkw#ejlWQ=6k6-`r+U|42@b;C|0i-WfbO3@Y- zt=v_}AlOCc1%F})D+$qqmpL$K56qXs&b0nvXJq8|NMKq%n~Ni?)~$!Q4vI#wX_VGW2aF|9U~iPqo!v-R z@NQ+_zm9|mkHFl7%T=*EPn;@^F@GMgU)lNHZp~NPQ)44^&UWkTLn!mdNJ_IP1*+QX7v)Dj{)*Iha>3+Sdq3nk?~F8n%P7NIvEa#h6^QJGmAA`lsy`0EMhmgD@=~ zi5$56aT{3S*BdGX_jDauZ3roYZ*^S{?IGUL)`MWSd)Q@VZG$s$4IW`4upQy9(YvFV z=YLc_w&s9&A2#-C*`^?vX5K1evJ zkoUmNY&I=rejAa5+r0kd2#-{x(kqfPnd)-**5ZE-=lopibwGhsbb}CXGw~@Wk&z~X zU&sVX+s#aSR%ZcApMB-^r;({@lZvQycw@Jsg$3=SP$7^Bf znd^$gq14=tZ~%%2&t48Nk&%9ScS7plk0KVcdEIQs2D5u{+T-zMv z)k_>sj&LgLAHcHR=~Oveev4@ORN_$@SF!7MUBWwu->t41^dLwex^GoVxlUNCZ#|a( z@C1AM)dz&1D8mzxhk9lB;wa9`{X6^}MR#ix5ms171P_AiOY&Qfoy3y)#1TJskN}I~ zH~r20OV_cD%usl~j>gx>Op+z!q*+q&%a&tW*8_*$zLUJm6p>M9nLF`gNd`DrKI++H)p<@>(OW%+@Jv^Uj;E&6|1yfYo|{r&g$TJD za&g4l5&@A~%IkM?QtSWG_0>^PZ{ORBdKH^SNdb`*5v0o?R6;~L1t}?!ZZT;vkZuK~ z5d>+}0hAgL2}v2GB}N(s=G`A=2Jii?@B5FeHOo13&OSSy{p@|t0h1HWow@R_d6L8- z7fbdWStuvmAy4zK4*{cw1Z(j_z_BBH;Tzpr3Ky`u*&f%xUfUdIsSNo0ZKSk3+8263acin42M?>U!@#0nAp`tew+p74GDgL`5*a9MEW&euCT#<%sS#(-=T?c^x__dR+!cF>*B=y+jj z_;Nq-(u%Dp|+{jhlLfZ4xJBMrD# zp|Xgv{+m!0sMnq)Zru>MbIe?xFm=m82Lr2Z?X6jZty6$oB8YSM_XUKrFg0z>gjfcB z_cZC(o%Sujug9j{u?VXX+g4!K%OE+9Z5U<`ckt5l8IRAOq(-#p#_#-24W%YXJ; zUS~SgL!?6C@vN&gIFFYRNhGt|cXXWS>B`tZ3cUB1bx1p>VnPEukKJ#1P=0%^DWJ7S z_)UARk~T8pIp|A#>RW+FsmC6WKFrti*W|=yCGy7WPSejhZ-09fR^FTw|JRo%AAsL9 zd{Z{o0K1$ z2|82>@$K;`lTE3LXltS}Mfl(IQ54!|T4D&kG$*fo8OBJryc13YkP5tKUOn`q;Q%&{ zDOW<3xH~$f_g3fs%MWz|3u|U_>m6A@0Is)L{adNV{mhr}u! zJP)?FPyL$$X*~!j50V%DKa$Tk^%vy8@tQso;$^mFR!csej|68Wg6EzC5gZW+6T<`y z?NQzKRJ#=rHTph;&YOpFYVccTqLpM(&})3ErqmKEj4c=?c}ssr7z58(`}v5r*z?cs zoup)|CK~a<+&%2$S*bA;yLIV^!fe@77d@AEji$v}rb3Ej;}HpTM)rrZx!&!bJ z7Y^rb>8XD?v3gtd4#jnXZ%F`TJGpf zk>aF@Gwh23-Jc4{Y%9*$l}RST_O>S6dx-b2g9V8W^pQg78Rwnzmh)pN;WwvAT6h$= zwbWzmCrJDkaeaxqIdB;ldE+Onk9rcBdpQ$1(Ckt4uH(6}+mP`LRMkUHJo_(9kvl?0 zxmoy>rlm>bE=iG0_~6Th>~@<{ya>SsV{*g{1$7v<6{!;uw^Ibx5&Yyj#?Mhwn}mxVLujpV z{{==`;`Eojm4z9sIWE0O1oc|jJx%h-CNP7jBMB({pT{#K_)Z@LTJ`7YSwEwJRwlws zVIE}=U)vLMB=h9JOYCozenwaveaO9seqJfhRvggZ5=j5~?fz)l61rpaV@rJfe+&4A z;H0KK%X&%d#uiH5y^n(zr08qKvDqu?LLx5YcwjT86+huheKg!irh8P!2!0wS8W910 z{^9);ZlV;bsE&BiUDAA0$2I$7XE%N!s=Aa)rFmzY2_z_FH?yTl3vP|P zx9X(*ZmcqI)yc0j@e;{sp@stb%<+Hp`6TFbRE_@J^+5cdAp?>XD#^hr@T}3Y%m426 z7(+tEjsNSglEI$t-{Cjl#14|B>XVk?39(}4SjCdq{8+~iuWi<&CJ7XYo}9$O5KB&c ztgJ)TJMex;QR zQ8A9qb|^TITnhqM7wYb>I(Z7Vu@T*O4L#y6Pc*Pkrh&WMu4?+9gE=_weSq>SD`-1Q z_`fCid132;iZYvUR!Kw{&FI8}TwQ*fo=v&Q{$d<*?7l2Z)eT-UBlU2|oR!&0{ zA@FSI#K~r`u=A5t@)rk$tPHEkt0`7>@S4`v7T0mG4#~z_UBw*%PL}A0e({1)$9Kod z{_C@&^ktyNDT>$DOTe_u=Acs#9$w#*QH{$0MSyqv+3}2`@>Y^D!H!FxS&>EehZn zjthzt)s{XV)ac%WlAH-j{O9l#X5y?qBQsFVPJjO$=dmLAeHT&uat-k1>tSKV%|rx6 z{O`LcVG-BzErc~OvPn9V#GEA}7c8QzAg0886#Z{*F8M78BK>*`!$bIc{^4bI;=K33 zTIl4i;wS_0s9mo-h$|$dJyk9Y4x91D%8LXKeUvo<0e{q~$%k@~vPbB*P@<20xO+`$ z#Q_*OjSD`^S2Z{`>nW`_i2JujJ#MzyXXz(3^?T_gd|&+yZQ-HXiWRS zPuz>`=0CYfbs%7rxY!3^V$9Af$IlJmN-3WIZDu)+6a)8-6x{@PdQ2g{i0Cn`IKkgF zaX%^K*mMd=l_3t21%nv5>|XP#`j5BxVhA=iL&2Qx*Asfg-NNHFe7x8uaSzD&{)Ls&8k z7}|!I*A+Dot>^tEdV&TeQ^63bI(t1d0`cb`s}Pe*|3vUOUjL}AIF6#p_bRrN;z}Td z&&yJBv7$+^*Q*ur#7>e|?5smcs(jb215tiFPg@TYhGvxqe?gu0u16$BHm`HAodiHH zg7*~q{>6H^3}$kW*bZP405RlZh@Pz7%az&s9~wmExrwA&yhc97ow!mpJ2a_=`s;7^ z7_jQZvfmkDSW1d~Hh$>}?ytL+E8FrqoRKeMWHTtqr%7S{AV7(xGniMAhroJ;AKoRj zF|kHG1kXcxll~Gr$L3xN&QPMFN+4dGUM#IQPS%oI*^uN?osnF_%%o&i8?>Z)Adwh0 z_)dcx|In|K_hN;`=GS5zAyV%HXfVX zHazbLs%_E=a(!^vm@*Wc>EDvw0*A?FAC+@mxZr;h87s&pl@A2iKMjLfIh&@OB$~Vj zpp?*{;@yA|o;kA+x7Qq2@*+(ryEq2fBE2ok@Hn8uBs9BnxWElk0s3=xiO?TX_me2| z>olP7=gxk%i$dY=m}QI*2e*ZBekim=t3hHQBVbALk9CNDtjF-7atO~Qt7hXC)w9l zzxrtd@fE{aZUj*y0M;yUVRGGivto_jYA4b1SZToW-d?&OsRwyZ`~I{zqQ~I}XI64= zoYsk435VWpM`0C&C7{U8-XD#Y4QKsQM(P`~0>qQP^O?qDq(C56@a_86E<_Z9y0)~b zlsi!3t?L!1h-R~hC`gDV7*F1n0Vk5-{eRMQohx#m{J3T4N?2L{ja(|>+sg{m=QSWt zIt1q(w}!6y0M-&Bx%p$!s7=}RM8}D(&(Cmy<(+@TJ5_fDXUSxgKmZXx_C?EDQZ;;Z zL*YE;PJAay<6g}t&{NuUr<+0W_j*W>$QHctcLZsKv$tHFe2u~*bwYwjy+NeI!5)*+ zB7UmU7kWt4+5e_%&fJA*JUw#nVYvV{>CPSD(5=l1(E+_eeN?NM8bt^0RhlP=XeSvH zM1_eR9sJs-K%g#5iG=Q$Iv&uk&M!wY1|j?gM!wwAFMmN;O8AL}&ogPO^ZCOWizMan z0BJPwj^bXyUk)bbP~yS?&05lZvz6q)CfQlJgltXYAmCw-YeUCf%$k!FnS2w7>6HRx zHWs&bqASaTEbfNqZxV#z41u>j@XZz$aLHxl}IoO)>Z)4F=77wRmMoD@$@3NROpL% zhzEq};CFI!J~Sthu)-IhRT@E$i}~0gSqe$CA?c_!5@94|^h6QA_}7w!A>cYFkD&bb zDdCQ6S3>@QOiQY@#(L=nxYh>`h|`EbzA7~=(PP9Nd(3X!C4TF3XV}<5?!36XYp|9_ zE6jw@MPza=j`tva11{~?wggP#8S}5eLbARu|L%nSEmEC)Oi;=NkZT$X*^gaHI7=2& zCPIj)1n;p791JgzvPL`06iL~j1VZtj$kktlf?oP|ySexQrIj&8L!YbzN;n20UnIu> z%5KM1R(}*^^W^ABI}*qhMl!#que!UC+JA(cHRqNo(cc5P?zsOdFlRqx3k?$%gvmL> zAZZ6y$0cnyrPfaDASOLukW_)w|9hw`$XAe(s30PsX1&*(!%x5nM+9&85KKS|Xnn+g zeY_3x&>o)IW=up7DZHI9;)S-L!aql?2=b<{fO-79_-5Bt7)5}5{4T*IoCO_yy2ro` z4I6BGjp*`rKSp3`H@1C^93=ag*btKsDJ=f1XKwRlIgBiT`2$<)8&^dVNPRcvmyiwT zV_uW02L6EBj}8Bqj*!cPEvdZg)x85k)ZU{UDg@E-0vBR*qDt`;cPkp_f%GRo`(3R- zN@pG;y^C+2w$Hf0y6A}|o*Fh7DfmsP#X+(SfU2x-2~D1?0cGuhkue_I0N@Z3aR`wV1?JGjq zjdNE-uOQKPzk2IG1PL;NHyc7-xvI^}imlf3QzLkcx8RBp9HV3i>CoanYf@|(Hoj|s=8h{)i znJ3Q%M%UJWq$A&;AM&}Ycagm+=!nyOaRPA~q6|aa+S%3L0$VJZ=3=%bq=*;b&8`o% zM(+flbVSTOmKe+bTbd@--{&F5pwmVS$O>UCmJH$qk8u##H;*rE?Mu|KKN{NsDMw!2 z_7j8@D5SFND0MMzQ?{q*AYh-?Fo?R6VKWy@uFGKtqY3cF102BEG3`n{aIo`}X32X< za=i>g)LkNVL2a$lUkfZx|E4n)nFh?{5CJepM8R5&u;1er!0!H)cMl`|p))I%U0zTV zdvxQ;mWtI}2Nk=DEsd5l+Zl+jyP-vi1|IwFeDgPk!$F zj{Ar>5E19&TheEzgLDkhm`DVF{Hvbvze3Cbh?N4UQ8uz1KtUx66AD{WOQr*JGm4ge zSo0sxp05X@%vxMtc!_u-50w)s6N2#nhS@)*U=IK6 z@&C^N5R}aBL7jXQZCxi%?1}KNfT{mt8f;qxUU>XM?q|Y!Bj!Z+?R&)!^9RM>Dm&I8 zRSw84vH3_epb^L+vFZ> z3xnO0M9L*15w>u_G*$j8?5Qppn$QxoEyOS*h>r^wcWgQLYy>;i?woTca{}!~laG?h zST*+%5ckkYHx~XP1%HCXiXl(vrFD@d+kiN|Jp7wL5Sm;EK}NHlQHGsBO2fpR_+QBS zmjGmvej$rz96+l{$Qou$?sbElGW~$&GQp6n!Bu2X?9E|qeF-iL znkxwmT_@^0~tT6H z{#ayR6u3b*8uF(W0zfwuscj93y5Wf=7zYZUEI<(}UBDVmNK;`iR<`y+3-qgca-X{l z;=&P{JN_!i;)9wxn2!Rn<|CIZ4JJ0NTc$yVh;-k5PJ+LBiHP*RGXvQ!iii~!rDWkP zYW*1$@>E5Hk1GI+Fh8cMLu#R;Mx$Zj;I{eAjwxp_t;qhKdEfj*YpuE z$i4fo-|m(G@HIn+qa)8IA7B-0Y@V$CsjD3LgNPNk+UJynC3!!{Mp9l~N>4)0AMh%cdqqh*z zHIset&5HrgL@0HTLe=DNX(AK^3p^$Yc+89mL?W$F(Qx7ovAQ0+9v@Ka&8|JgEL)24qPz# z$DnHrK{H!Lr;DIr%oko1J(WW`7?3t4qQZ6{ZX}`dlpB(#@48V!a3lP%l9JsMBD?p! zAfv?pv&B^X0$q#udbk)aR}DjK=Si$*7-2t=oKpm;$L{z_be?-4vhd0+E0M8-7hyRc z;evFJv4e6jTmN*4ldFd@n0A#*Z^1IwGfFT9(*Li9P};Tx<6VG_-NF-J{td7nLJyE5 z2qg*;%4C@+aezxOz{QYQ5jCX7afdKK5i$VM8T>^P5IT`mBn2oze@&2P-i3Z4q9kr0 zigEndv>wv+cER>E0dyvS7jpYFnZA3UGPn@Yt^#296QCFSUC@Y({YAhBLJ|XQbE6=XMD3>4E$@3zE)2%F8!{QN z{>sV%^|8wZu~;PqPzIo;X z=2(yuSaLzIr`!tiaa*<~jnI4dclcAod^7^?ix7v%LfUC~JkZm`Vhus%j)Pu^n%;#H ztu)}CWnw)rKOe00lk;f~x#V4-?T4@RlZt#KVm=O`w0TCP2;L_b97lN~l#%5mVJ2_} zY%6mrg8Z$o>a8RtYkmcc_qVY4f0`yeDYg#3872NY0g|By;uFvM>DANS3IL;Ids^W# z1Ejl~=;80lsZ&GPB63>LvM(=W7VJK{p(*^_Wn` zB+OjlD0;BNKNJKLO6+blQHIT_qn@T%&}?hkcRxEA*|P|wlzBF#^(fiy;}E$;iN{Ku zAF^1Vn-1NH%#PfzSZ|sWFjO=mQZK-s=NC?z)l=NWz;G@C(a;DFt*MJ zy=6B4vyPVR1_)x{t8~N;|HI(NBvCFl$fUUdIUIu6vAFowp41C)7pkQ#(3VYX)A$ZH zNPpx%)kShpxl_~!RZhqyp9BBi$<@DAxgquP+=X1ZMKE)XSSAIQ;GcI$XFcf*4<(ev z)Hsexiwq~>SCfL_meY6M7v<*k2z6`(LzePnQ+jn!fD4H-iK^xhb-X zN-OwL(Q0m2czYG&w>jq5x#rhnok7!8IpdC=i2xA!_o__ z7;&^m%eUo;^?wS3j4Z^_WmU7Azi?84`1rIe2N2%4w#mNvF<*HpxT2F01i5|U&L0c5 zbLGLY=_*W4-hC?N-9Kfd{X75j z;aW51`RK@^n?Ihk@Qhk{;^TuYVo@GiZH;_mx?PHlu|;Xt6?MLeC_M4#q=|=r3N&-h zzLkXK|3oLrheXmLQl6+WtJ;Wb90t5lWSrwJTXur+*M!a8!WdB?DG0DLnUUIp98&h;i zc-$CGGqnd@a}mXu2*5on#o1~O2B|r>>UBeGY*$5kH?-CUO&Kkh@!G_q@WcbZ52)yS za2|JRjw?g827KMW%17yloP_@)Cr?^>y-#>Q{IXv#uMsw=Hrx+4e0_4-!OyF75#4O> zT6nH>)@RnMcI32M=>W&@BYdrZFBSv;3+n74!4^G1^Y7pk+xxT<;oi5z6$6GNE*|C%llt?8y1Y$4&cb7{Gs0iis`G=MYoZaHP-&?-hHm$s zAM0bwl-*K-|CtcC`XhaF$c_iE_l?QCt~16j8q*2JT4&n!lXf9~y-N*0^>Fx-IOc#$FX8tHw0+y!0dr3OCiVV@66 zX1mu3B?>O$xTrm`!NvyO7^);L-#^cjUhLT?`8)rjW3!D>;fJ1%T&=#P;*x|?`Rn?Z zF;&Hu67Z;WFdm1-^QlTs7o9ai?hY@2D~{>lc4K|VVTQpEA8PsS3sPp&m>1t5?vMw) z8jfEH$DBXndChS&#`2}vsk-ygbxBE1E#gkK9$L@5U(Odz#f6nDS1ni8!png=OH-T6 z)qdEruFBlD*$X~=_g%)CWtmhSvQI#bMIeERaSZytcjf?gIA5jr$(9W(n6+nRjJ zU!H1H`kce0S?bC#=&14PJkmmM(<>}JZLP+8Kppd8ts~FKdojz{uJ60job<}aBd29> zLr2sqIkgJKadV%89~_Z!pN>zLKWIqTEQJkJ<4atw7>nfXw{`BD)GFZ#zO*?ryJ_Fw z-gAZ0%}{v0J!!Kjpjc~G3bcT_{CmB3f^$oVUCml}L`C_dp>ve}SzmBDdB(@71U6ceS+KQsx zh=4wzH%{=HsLzkSWw-;qwrYJw{}P#an2rjk7I*39VxBRoQ?TIKWvLPga4*`_;NwkJ{mfoJ zcVERP%z9_=oedbKjQW6b&3e{1DpJC|iDkhT;B@JDo4+5B#T-L?LU4eC*v8NiKdve` zHb*<_yIkJY!@`S3W%AKg%{@Y!wR2pSqAD7mj!_jKdBJ9MT=NmXx(UTbXt$!D*+@!& z?t33?oYekvOMOo0dEt%AYM$3u|8R$K-CJA8kTtA;F$ZQU0&J*nHdBq6KgYJ`I!Fwp z&Ydbgl@($-mG(sH*9nHlGsAGpa1X7e$^pCBkVj%;aEY>Asp)mw^0~-7Z}alTPfN)U z4)bWQmUh>k?p%0TnIKw(@+)$K=ZvPC>mNXjI>YI2K8Ai?yB|C2LT#SshFj~ZKHVa- z{u=}DMQ`?0h_(dNSeeSBs=@cLNcrf?v)aj3fY|xQ*rTd(BaNlEf)w?Q5TjvSf;TkY zqc6R6e@{`*ne?Q`U6seeu*l46e(dzYK4PmOJyJMFNPFa5k~QhbWWI{YkB4}k6K1cY zkZ^5ImFtu+MQx13(3RMzX1}8BEk{uZ3}sz-X>AvViF2R0R{14-+B*zFCFXim+4r84 z*iU%GD;sT0YDf7k;Wr2Iyqi-hqj3pfv3HEMxu%*c)R*HzlZ z+%$zX-dA3hoYYsFhO;ZR$QjK;7qtgttrPhZM9Qoyeh1TdW85-$Ofy@;Y0{Xn0gBXa zA3ie$!CM;Vj%o2qTJ8(PXNIbu5#jCReXw8xrg2awIdpZNZ9WYV(ul!HE$ggep^%)5 zX?I5RHYfA`;nRB|KdN5DsMtiF`oUGR{vyj-D_c-Vy^b)|{D_X7;FGjaF_BfN?{!Xh z=LX&!xd^b6dDjMkqSze^sk-&(uWWzVqhlm4OY3y2!Y4n?gaEQA9Tg5Omq7L)nbegu zU({_{)P4P%RLSCg$2!>^+K65H{h)ik1&6DnQ%1xy*< z@WVyj30?KTaA8U59cf44r15(5SSWmy&=6ep4K?+km-aLAB&$~D7^ga!(|+u z2lw}1RSrH=`oSw&HjI)pAIx-r7d4(P&-SvWmaNT`=y3O_mmE+>xvDI(o4Nx^eBtN( zh}R&cDER&Z>t);HW7I(w81tT)^4Y4+ej&6}VSxC&XZ^2EgH*B#4y`*SZVbxMyF zwX5=)+E!sh9r%3l>w|d-RQhM+&#O`|DEG424n1-8zL^<_|LR)163WKCbzrsVRBI>B zBW6|o{X6-Nti{oD(~BRHBwybUrs7rh-JDO-7H(G?m-M>A;As2D<4x$8^W*%M(CVf4 zd6B&g^_CxX^^1#t z6CM!G#ol8PP#0_|J*7J$ltw=UNX<~P#9C0kS6)axVs0_duQK3%!1%fkSi;-sa9<>r z7_P$xZqC1(d*v&9nNl=x=B&V@^nTg>QKk$J0484@;id@JVKPr+uM%V&+#PCG*@ct-*TGHySOys|C>5kMSOpG=6>! zMEytOL-xDTR9c+lDl$j=9j+AI*YRurqjGH|5Qk~^4-sU38p2)seIbtnqQr|nB z)-DC{TRAwVG0+s)*)`P<7hdss|Lt_|vv0W4M&B|wC;#p;3D^^sg0=DTfyLZWQQfg7 zolv{G2PRia+6m3o7GNYXCx4ASfFw}p%^hn#gdfV5OKyIM-ydr%^9R7MHWsq8Qruuc zM=M~M5B3_&e%ZH`cQrporH??_#X&cPkjnZGoFQ&$HKc3htyYlg|Jd?_B{FOFkYFf=K!CLRt5TO2Ia%-$}B(+T~3}}oqso< zwkGnr;Pv>er?-7<>bg^fi+MsnO`eh(A~!EkX?l<;Ym5d+sCIg+xxlm!0jS6m8|Eo1 z_T|%BsHJFZQ@Wx7e#Wi3dKlv3Pz_NXW8EIbmvS6lA9wnF#QWzim5b`iE$wDDC0+ogM`%%^A3h#B zK4{rSr7i7eRMek4rC?H-8smQbm*Io?&j>$kn^#)Mkqv=bWzWj({d++z&v#&Q{5*8v91L!B_s${Z5}<!af zH?M|EIipkaQezM+tu3iq%wek{*OkS{%X(w{;c{olpI`GWnx0*P$mY2?Er} zfi#{`>|MJ*kGjC!Jy%dM(^jh+Tkyw9^;p08Z-B*ANou!ux^Z?&^>1umN7dYfSYG{NbC&7~5$JfCTsuKg}75b0iIb##R5 z&ZkWFt%#o0LI;#$iiW>MRXJ10EAD0Sp6lVBr#*jzE`73N!OixEqu7+V-KsDX`Xc<= zr+X`Bl^6>&uTxK>d(K+lt*TPT_IhH;({y5Zv~DK!rHYP5sM8nER;`Py*`DLwZ1=4S zTAfd|b1^I-YY(zuJ4I#uBG4{_P8{wINFCiOU+x3iuI~K5OVM({WQI`Mu{9dj@6c`v zm^Nf9r58BRDarQM9myjCPzH?H>7H&a>pyR*-2K#}rw#n!Uwy<}K>!n7LTG0y4b}_W) zSD`F1>&``$UW$a$-u@Y__&cA-&8smCV^448v9PRn*9TQD^YE%H-BM?kAS>KWC`u?3er|%O>8musr&GWIvb71y9xl@WJdTq*5 z8#j-5>0sYQRobRJ26jE5Sst-+h{N<>kYXFD%faema@IpX2D`Xr<^q1T$ZpB;Olx)?f$M!b(r)_KjY`V#HoW7rG!?dcU)PM0a; zMEPHmTV^#iN$Mp)i>?0znPE#(6u*}HT6>a_)aqBY(+qyoHpb-vISNyXIKGy^;ns^s z_>BN$(rK4AXg1iGN=E5;2n4lE@z!4Qziewwt(bju8G*+EktUOh)J=@;KmBdwKakLI z>W<2)GBfI#y2*2kc27UC8Z{^is%7&Wby^Uk<&F*Hk1;*WqVHNqku|{tZ&TI2KJD`M zy@mWub=&h)ytXKcq`jss*9S39i5{8_Mk;=JkTWt!XQ(83)H}6{wZvMYdF&mA#TRoj zVw;91^@W-XU$+dGQ3BQ5!+m*lVaYMX8osVIyGkPrF`4cO6Lqbac+!i5XUCDt5XhM$_?wT zmZ|EeS4@%@D)@HUmLk7NJzSf^X?V|R47|vpKZ_jZdtwz#jHEGrq9I+ zBQa|7`Dn#@(d^*>I&ExomwdQRm~0h4LqVormgEebKt`%D0P%sDksqn(D}hwpa4~NVi+-^!H*#3;Pq?XmJCkd7 zN`N0bLMprbn0)je!Py?O)?*6TujleSn%|3mmK?x_u@&27vMf$Jo4$_H(X{B%igl}; zG7SEp=)!khcpjl?;+M7NL%8ncp}n%ap7g6Yc(%X4kdz@c!Y!|(so3tJISKnI`tlg; zXPVvielDvEQ&lqnK3_jR7Gs&H+w!$blC#7jC{o_Rt-SZ#q2;D|%nz*jvz3r*gJs-W zVIiN^SH67;wRcX7S&kWdSUiTFm+CFeq6;Wssov#?@ZP+BbTUA%cG7f&O4fON99=x`j(2KsIZ4VDfwy* zaxm((HLU{@c%KuQx-tgIzM-&zO;>1&6{mzg-gUAch=aR}F=CTfE9+-3a18Z|ZLZYI zzC;ITMwse9TyuI9+SNU^hO`YzK{I+%E%VL{?qzCebFo*Yk%|eR(AK z$(o2w!Rw)}s-^eSS}%3dBeCzqH)pb0p7&PrIr()_>O{)1KtKTPrQX02-uy!z?1x_paz-wnm*9=M|wv70k| zLWeVujdpL58WwMHpsW{zv*0mSiqs!9v*elJ&1IWC4$tcrSJ%veXTp>ZZrE?}pw=H4 zlIVEVnc#xMkwFGPxY{MGdsGXioH}(aX$r_1i`acPOZ?VLY-E`ss8lTE2)xFF>v+{|osmFG-qmx?4E&$NB)tD- zZ(eN{{E!9rb8jueVc+e2g%po)^DD)!)JE@pP>t&=OOr?~W12FhKpKT)X73Jlh}p0I z@ELGuL*Y{)v=3&VZMS^!1e?8*5Ntl1rCFQW0R*fA_GL+h`NUyVbsX01GWFPvdOHWN zUX1jt2m0Hx>n^e^>T#RR3Z;SzO^?!8IW{YEjkpws8hD|cGamt}ft|I*d9gflnqpe? zxP_C_ig}AJ>@4ThGOqyovI>IJewU=pg^UvYP&T9A^|wXa!*rzCi|?-0iB_p6o^^n9 zg8_B5Q!Ld)YP3}STK<=|Tge1icG?$F z`hT5{@de)cg>i9!x1!4>-{Ea^6F~q?-W_d9uGw z9N}g;Mvhf#QT&wpl|tZ{>B(TFgY{S_embF8Me zqHRC*;InXtw`%tN)B^o1i*FR)mworv-sm2$n&;Sz;(z&e^Kek=vAzD6{a2RRv~Ij% z0ry)g6wHjJoQ6Ggq!6gfVmGR*ua<7Kohr8SGEhwB_m=^-&TZ;jThHxBi*c9Q>@=I01!%HT``o|!0!>pO zb3pN%PF<3v%=naT}8)QTyX<9odY}6$7!-SYq2RueBNx;8kpuG&siKSl^+&CyA)uQvwz;O zd)?{Urd(v6xH>OoXDsPPx{jHNT$#2x-e>8@NK_nV5ap>lY$00y!$XU||IOt_>Sk1s zs$i#?&0KnY|BP$N-IG6c_hhKdj%{@qK-W-e44hiMyBF%(5^^qA5gnm_A(upsm$_-m zJ-(8z#t5i6A;=AiI!@A?ACLInaa?FqjOs#(`m4%rEKKKyJM5Fce>SW+PBiN3bu~$h z#9rqN+oJr*PsQgCxd-?v4t3Z;dNHl_Ad92E4DKS$)cB<39hLyA|~A^KpX)z87VxzG};wj%1WnPTSRL=6QDP znTfAESxAYu185-)w7_vewE~&~O63aofJpXCb|aqp;DDcfdAy}=zNG)f@Ol`j7Xb6xYte&K#K!P zPtty!)hVu&z?ge-(2j{udezjsJPinxF3K8y^<0%_=FStz7kE*v-jyaxnZg;$8f>a; zkZPtwx9kYF8aF)Q_s%tm>vv#MJw=ww_vPW-64BzCAAa}^Af)M{pR!5-{aS4wHMxh) z@zbp~W$C~&cyy8$YgcRS)s$B>0=h@A2iAg)8fKXnXpwX^__A<)0rAi7n^JUJ; z?+ovr5tGqZ{CJU0G49cbVG&xIj!MTNOhOkp;^Uh7_}c2L*yYPEDAy+S1t#9;W(mlS z#IY!$9*2i*Uup23AP($;iQwUj@wA5T7Er=0Crk}o)^QBh>`&K7rTkloRk ze|{Z&az%hcqqZW)Kf}E%NPz(~3G-aXk)85XM`7v?&Sz`b&^0mBW7FNE>4_i6=RKd@ z>e+v^0OiYJm>Cs~6#p8{?@cLEkK-DvaaSGTb?uj4b=m?aPX$}FfKvT2zrvwCyn=wE zKV@qa>kkxDd5rijT@ZPhD?9Q+qd^0M=F~nE9m!!Ykz2+S8#uN_R6(it$P1M~M~X*E zL&6@Y61Frs?x0JU0J<;F2Y1AkSfyLXN2%07JBzg>r;6#pOLwb$QSM=bIVy2lCH5W= z3ZXDJtK+icU~dSo>&0kH7wTmoQalV~7Bbk?Ub?Ry6F6_lv`BMej6e2m3I4Wrx!&qor?#}q6exhjOVz=S-Non%A|iI^ zl=A8vODRsIvJ1@VN`13YFT3e@tjhC*9z0iw_5?+6Ye=f~uFRS~I_zX_yTMRhU+_tJ zBVegE^IE~T4)fDaG3lReQ|6J|Z?3MTa>@${j+)lCiD70lx@WlRynrx1Q_tkq*W}xx zFHF_DuWr9;$E>l`Pd2(zrP8%I1QotTXW{kr6P%yN^}RBXf1j9lwj}=8*;82N*#?mBQrhnBfi9NV&bV>NYpC!taRAd`%gWN(=}d^1qPK@AkdDR zP5f@k0Yj^4@H}Nc%l&S4KPFbrH}RXUCviyxqO2mbLy~ez2ED1vE>Q}$gtHrQ)jh+i z>riD?zvD_;gWwL)26CpHu>lNMqBV9P8G34!v|v4uAnI`wz*?LU}(tAIKyIAfpkx|^(T62`nn6bel5 zd52%4wi-B~Y0CE3G2ZJcRbuY7yha5}?<(6={<-svj$zln>-8$X;T8bUHg`nZ7z=vz zxV0hzqDC7T7pr+q2Xy_&#iy$TEq%Jm2Q)X`Av5dt9j$3$3(6xngEfc1U z#nVEw?qTzBD8ZJvbLhlHk2g7EH?W>fX3QEDbB_f0@UuuZ`krigE$FRa=dh4XuVVJ4 zRiMJM`Hb?a0@MNClZVkvnjSvtr?}v#+>*5Jo^I*Pdj@b@QF;}bMn{>>>No;+yE_tdIh1DE&l}o3BMp#rwPuGTy<{Mg{Tq*`Es9 zGLK@LtCXw;hMAXjL_6DGo(!j>EI_-s#QUyIj9d-+in_ItuHHTOShONV*+gp)=Ytz& zF_v2YfgYIoq0UIvd%i)+>>)D41)>*b= z#(u8Ug0;oF+Td{c%BXX;)(yMBa~?BN0#TVZaVU=;w{Y20Y5XcZ1xE2^jEeUwBusy! zN>&-C$!&|wjpDAN)%4pkD-%Cx941?w?#tX%wigu3N!m&`9X8=L$zEsu^cAxVeRbsY#G{WeMdZ`N@+OB8K{BZSWVJ7 ze>#z(aiiX#=6o`xPA#r^tOJeN3ymU-b9#kM6K^a>rGNV7JC035)}Cqg3H>?Wu~b{gXEyRPT00SdxlKmh zQa#vZuchJS*zSM$e@g#Gj@A*bHig#_qFhL~uj|N}(7^pPu|HOxN%jpLwi!$ho_5Fd z1~)PyokYRixu?Qb9)J7AJPq;k9_821Txy-=n=g%1y)suG4_(uWeY?h- zm4g!fY-*Fkl^PgHZL0Zb${4>k(y-raw$e>Eo;s{O%;Pz>!^fhQpk#5yW|h8OCDOg% zNE_*T!j^W&a^iZq5+`gVpDsD~S!8!NUh@f)AvTKX!9q6f%#@CCN=YV0=N>fP`RkE)qlT%$v!YV_#yfLYSO9v z_q)dI_6=_iG9PqOIhWR(C$|$lR4TXzZ$4q_JRR<^RL?^f%q(~$*zQ>-`~GiP7w#?2 zc(jO*z38>AlM)F7ri_!{GUvh^+$fqJ@20KTM3qGhgLSz`7*ty@PsB2gi3Xa*i`pdi zyMagHmZ?TnH#$~~p3CE|W;~j7a(J+9rDPc?Z8V`%R4^>O>2=;eQB-PuGYec6q5RRN zjm4Tch9AzGeXqh>E*mFb2CcGH%w@aF&_R*4S9ZA+ozU~qBgi84r^FxSMRu(-!F?+2 z?0>`;1_Lnz$&&ZYVxlJA%XWYZbs*Z^`a@9E{rTdK)s1}%;(jx!Iw`rr4o}Z- z#Ni6~t!;-tz_Tbe>Wq*@4?ptV2#7RKO+LN2}lg{Det-4(c+cxOe9_0_?0(<(2@~8$0Nq zQ4zWGbannZ{b{38x#vLZ{ZdzC(v+H157)DCpiV>Kob}F+Nt&3X77Rr5*j5)7Y*{AXtl)n1zME7*x^sJ z4-jcJm$6l?=#uG=<&~w|e7k0KjF1kZ)LJ39l3T{rtn{;IuW)T+fV-fm2kO}(+>n`v zq3k{D16^)?bLJOpC*}}hth~3j#NMoQtsByZonY6#S_7iRX>W{Ny95I0{&eiOnW`Oy zq*rrQWo3;*K*Ch>z0Oux!Pls~fX^)vlMbQM3Z*PP>0R z5Oxx`CVnv)F4oen-LJVk${)K54ng;K=~$wTL_q^)E2Gj|{PbK$Ua*)Vhh?jEA=^?K zih1tG{2I73ve(@dvezhkpVSTR>L^fZYr1?Y;Rb5*h2y26eoLA8n7fhnU)i-zg^X!t zO{v(cYvb1iwIZ_}v^A==5n{cq{P+{Q^(M*0xH`ulsWHWg8sXZu;k3N*b=8{c^hfdb zTD>S#btO`WZlBBbGO9c(vS}CwTWXav%OB0jQ1Tn+!A(t2@ut=B zlnOpU5)!w>CA0=8O6n~1ZFJ30@GjKxsLeA+l{T3d_pXU;8pa!#yAOCtB*FDd1E13w zne6o8{?H~k!mgrPW8C@m(IuqyOwQyHna{~h?QY(09Ve{l0kqeIBs`*U1jy3lHi^ab>`HAb9$7hCt zYX-x|yL6(&iZ#R`<7;$S(y9-WoYC3DC%E=l_uIG*c#L5(pi19^W`?4>Ly7RFCtZSL z-Lcv&a>wTFcS8420hv@aQvEh9TZcDBk$R(1&Q5ZPp1ncw5QF1_#j^ZkB) zzrXJLp|{idI?v-gj_2__R)ag=FH?`U`)_VgCCd3JQ`UC{0~3(-)8$k**<&aBYInc> z9gBIXnsZrS9xmnz><@e9yUf&_ko3SwV?T7LzjWw`ocl>x=5|Gy1FXKjn$l!)CmwyK z5v1h&Ktx^kF;NkF;DFqIvyU`vT^eCS<2x04Z9l$?uMKssidsYtyE8K0o*S~xUGWf! zt<#-1;10U&&%np@`W*QXlP0?|pORo+;wjRMz(K*04Ki^D*)4k(y)Z>zEQ7=d@q3OY zzG_dGBdzC~cslP|&}Wr>;mQ))@y_(grS8k>^ocq2hjDzeb9Upq75#>4OyjiZ=qMGI z#~!0=ja?HCm0=&C%IS<+~GZOdG=3H@42oI&UGh#30bk}LI`>L!bWR@ z?UO)q1)KVuhodSYN`fIWgE91sqwN>hO9#kz7@ymlCwoF(@m;oH7rn@7lCVoPVs}p_ zr36+IE5=O9wMaguTFB$QYbk_p4IcAY@^T>{)}<}Cy?E4Wi`TNn3uG^2 zrdM&JEhFXRSdqNvI?1(*>b-^@HUp!Kkmm0+rlxVGlkrK4k|bo66p)#ZksGh(L*%N)vV!E{hW}dN$+zc zNv4LMT^f(t+s|Amfb_PM8$32)IJH_{qP$Or^nH$Pl50Xf)Z+Setv9n-U6QbaL?i1Q zu;z_^_@{!}m~SnKD^YKlJe=3nm(C1-lJF|?N)Qh;Ennl>W^n3IV!!>0f;D<#<<;V% z|7bhYhk~N9qCKZ~f&#~i4$$bo>+Cg4>^*vCQ{t<9mV}6(ZK(OkYedgP?VoH6ttwZF zaf7msNY)hLFChwy(V&XjoCX_c_3CL0G}+c&Z}x_l`N{Ylc>9U^;>DPrNZ*v`D%^K{ zp}nW4lc0Apdi9CKL=WR}e3wrb`R(yrCCZbbghUve%^YAEkvKqmvvy8(VML#z7UqTA z>TLxLd(;l(RIN|_aS3_<&jZp9$li-*>pv(RWQ#qp^v1i%D~b7~$t?-;6){efm?;L? z)3K5A=bOKmIub0}#&EDp$ zTDt$Bs5hfo{$&o=ShKIsFf8FSMZJcT>(jXj=WnR+Sr510+Tr+f^Oyy}zvBz}E z_STjB2vy-#g6um(!LxkzoaqVP>rqVutRh0Z+u;9&%qEfL#lC&}?&IM}0ND=yDe>4Z z;O@oDdJ0$FEg5u~%x~hQTWXM7nOaULyj1`pv;{ewnQcD}A-`<$F z%h50_J-c|=98My3z72+X1&=4#6t>Gxg8zN9ntqO*KcObEzU`sfwO_1_iS2*BNiqke zJrlUhBKaNaWjLw1M84nqM5&NnH-`z4Avb-l_66nPQYyTw+o_*pUW~&InvdY(qXjL} zZoC7=Ot5qBA^I0OG}wz!XVDXt3HMZJxCxo3W=aAfW+%$6>E-yvlw3kzF^WuDD&!bM;h8hFuuSfUG^B6O@)b3rt`H?8boTSb6S2+8R zspQ`R&wSz|cSfptwtCfwoGG7~QjO3)>LIT_%Sd{4G1g6ao+$a%LIM|hN|g@MNY>{O<`?7wT33`Z{xzs0)5y`M~ry-L5) zrR(l?ZEz&Jx{`!edTo3hhMT+7JWE-_x!Q&KZ1wu6hK%WkpZ~JxsBXxB!-RR*A?URI2UCa zBTxHPz3%Dk62pG)joE_XFhld`+v^$K*>cef3BO`(OHR&Ph!91>(+$I(i=%I%Y}@$B z=Mz0E+pcS}m)XVjS6VI8Cy2}TRNRiWDOQq(B)tZ0G7>RJT8Pazus_A>lJyN-vCbYn zySS}`4x5#iq2JBuLFGOH*J|E()eCH|_KF->5*R?%A@6CJQ z+R;;e)*a0|+w(i-;=Yogp;S|G^8?5;?;EbhjXYnPeRAPu2WE54e&%fthkflZimKA&X4R(J0DcK%>HYeb0 z4&WygXej$Jso{cmat3VD!g6~)pJHf8s$Jw5RY`9+U+0xq!@o_N3fARr)$p=SZzFfk zdnPZS%j^uPl|({31*>Gw8?zix)27zH_krgUzjqZfL}2^6XBnp__V*vAl`dg*fNED! z;@Zm_0QbYYDHRmsAd^$HOKhFX|DrczrE;a-SL+T;MoWw^cO_78bo(xPtj-jIGBEeK zVAEeHARWtNH9ilOEFCBZc)TdquF7qFG1%w3j`7Hinr}9FMVrGr8$O+=5+P^GUcm1> z$&IJyvoBjS$KZbF|4U2k?UYNUuzx5CklH7GK3`p55(i<6BAOqeC?^+>%`+Z);**Js@uT~~@BiBNoDrFswH)EBd-3cg+@O|7HWJ(qNfdY36B51D z8en2%R1>?Naa2YtP8A2mM(;ZtoIMSAPD$--F1<87O-hGsX))`LcB5e6VaRBRn|W4t z9EyaNYzBwzll5;{gZS${HJTtfg82vhaZNEue(QkfY`CQ9u-BR@R9ddq#EZQ3ZxZrm zj8IyA<(9|rMI+(H`==|hLXJQ}n_)TCC0Dcy?0{vzdm%n>=U5q3jm7v#^+mZPnJNRs z`CbU1%z37$O==_MeYz&ot&K+Z1WYclqhdLq9`FCcX3S80m#lM0?Ep)Q6W7nOtHih) zkxZCV_%?_K$oUtI5{)gPz0YGR{vSr)zFOW;BifnR*-Th&gG#lQerm(}Kkk04?sTX> zX<>%tc*p^kBcY!$ZyEC4QkPoA`Me*Y>^p%uz)vXfVIw@SrIEX zdPqBHOGUXIO5_SoL+;breT%P9&OUrLyndGNZ{LsjPqr(ZwlOS+5M7XHz7^LzEjJ%9 zFQ()r%spYj=RN_fKjZw+C;qZhuN8T4FgF)b^qC08+1ibKUtDuV{26IY+waRjz8 zTa7=a0ni$2rTU`Ju~R0j)vC|4n&p!LZ~Ds5Relsp&v@aB-8ppF}pQmuL9V;jcB z^n6J;t^vgi}8l)borFifEX zlUT>AI{H%Eq#+ulGsj9yr=uC}FVy7F$5O|M3%Nd!R{D|`VChUkBUYK^C%wbxvy&(h zKkD>6HSZ;wjfp;2MTNr1gw1A5TS?Gfi+=uO!}&ly2=jP!^jqh~ z8AASedTN9&*FgLDaUa~3vc-D3PULc_2_czEm(#Dh0W{*xvm`A;=*bAMD=Yy{UdwG&+ya2SABSYt*A%h zr8y&{hTipbg}Fa!pSyXj*lfk-E0*%@t!J1zZLnk0;L}Q+EBWp)I5g!>luG&TnBJFn zR!)bALLnu?LCI~u=i8>;p$mQ^hof+Owp0uEr2pmu+${{)A#(L*zxFG`sW1)6<6UUG8X_C0_?hn0KB_o-$h4;ymIz zXG-*Yft-}YO0B{jD4>rVyuI#2)dg5>`W?b* zt}JJ)(wNz3!sS46ctSeV(DUYah$KgrGu5gyO_**S2u9vYRCcBC>CpEtOy#wUto8v) z$=^Y0tfFs6Ox_&JYhYA)&R2DRX&aBSKr>4AUhkz-#q|pRUS}Y3V`%{(mzk9lGu|BC zDI1s)`7LJ~l2%pqexVu*tEO`$cMWNtQ-H&v7bE56zSI`)KJ}(>#Gm^D4gW}qJ99`{ zr`^y}n{-0aA}Xqhb(R9KW`$*qupE%0Iq^#t^gab^FF;T#xbjTdtI_~Rzm~%7wPx$z z!CA#c?mJYC4I_B;TnZvGEse~Lie+VE6$Qgzomuz6GvzgR;7*ElHsWBdb`p3h(8F9vN=67i^Z+=BTTfs z+^H^-D+&eO9jI*b&)lDH4N?(^w-`@QAa6mQrptHx1_5l!s9nILkCu8f&pJ8U(_o~R z$m8JPgApq5Ir@tox`5>=GCtwl43#H;8fi|pG#Vh)8v~z}DW~mo3La7g>HlyN4>K8$ zR9DLYSJRf;Th7|@%_^*R`vP40|p`okmF73IG6Ru{zJ=`iw0Y zq$TOe4~I@;M27C0mH}MuB;~943gysH;ac`9>3N6CfX(&cn(03^@FWk(XPFS~KYYe` zQ{VTDCF&<6&jP-L>;j|sJ-DBJ?5n=F4&!bd-|YaEs0IepBY0Jyf&OJq_4c|F_x<0E z?43tNzJJ~PV;*aB;Rn;L(FZNCvu3*?8d1>{*P`utp8V%==IgoZ{Q9gLz?zBI8H5gq{CV(RK2)3p7@pe! z%*629P1JH>RSy>z!NmU5ssHIE{p(1W9zMIj<{YyyKh(_7O@%fWhG+{XAeK`gdzrM) zmM<>s+gWpXC6g#`H>@!^u*?>efD_ccr-44LcsW&IZQ>dYuiLB7D=WKLw_&Y&^0I>F z7@ZB$b=~s5HI`s)6kucsddEnCG09${J9zetmHCOxDf(#N;FZrX=UwL90Srg)`I(~m zLhR0_5!}F^Soy$EbNV)+?%3NOefW32zbUF9a$|21Vu^=ftrKLOr^ecrac)5fs+rt! zBo`;I@Zi_-WM=`xZA6r%pj0%;y+c5CO+pph=OANuNFeGQ4VI2)^jdD8;YiEyNe#cI z%>DM=

r>fc0r$;xf}G-eaK&YJFSfm(t)peoWxS(r}*Yhi))_?ch`}Kx3-o;7>Ih zdv!~h0ar3}(76ei3129i3=j_Yi{x{6YAN72Cf?NB~S7d-EnIgcNrpSYrPvd_|`)8UJE} zQQ{INrW7D&!ElNXcpLe{WymMg+U$e&6tNe8;+Ub$*mvyWnQTA<6fXaum(avhfLSfj zxptEk)t@UajklTAl zI=iR?@Momt4vIDxY`Z+TZsg@G1_?)>G}BV;9< z{c?M&EWqb8kadx~mCGGYpi)y3wq`6QMM}2W&Xs@9=d4d!>_tHWD5Diu?Ju@v+d`GU zsdnB=v6E%fIZfM#M|G|n2v-8|Rf9}_kAB!yCzQNIP#bwJjxQy7k+I+Q-b=APbQqZk z+HkN$6TiWI^yow-a;!b5v;2dAx>nCH1J~BJ5a^dc#*(R42RV?aNaK0=7Di! ze~e9uL?y9S;047k$0*{k2?`(8%Yo-BPeI*2CQ;w z-lqK%inKK%mXTlB;(9av>f*ZTq9c4n6hiA^R$SCZgV2zqEC;9z6OY-Mi3OR+t> zjU_a_p)G`yviJz7$MMSs7l>+WL$y|Yo7;|7niQO5m=REZe@rtDQIH;yusSH=_up#J zS4J891|QPjm>6uFv582eB6&2}$9{nF9Vjc=>C;gWv+`;Q@D5uJHsr7ZOVSs03~_3B z483X*BH=h&)SCuOp=+}+1#3(5D^bA#Y?|hUL8s2b@fhngfF-pBNaK%5EYyrUhnkQ36k%{{JK0z9!^g&$jW@p!Ek>CJ zDfJrpY^BghK~ZZTIO`^u0oqS$PFgd~G)nb>SI5R)tcpD>Px{Oo7m!dMlGlmxI5#@6 zDvu^6$uIyF-Xk!z@~a>v+P${)LgjK^j%tQ8ewOlK7Gy?^$D)OnV0fftOT47`z{zoN zKp8sUDPHUYvlfe`wx30>WjqEsAf9K+Ar^5MgB_PYRj`l=5!h1^>Hwo5!UL_|XdxOZhY9 zFB^8zd=2*uq&-b!&Nlm;RM1J7Y0(VF{9^#J0MPwz{Jxydhzx zvEt)-<(aM}s1EmT7LLQL1LLCyFBcQ^Gzd@`I- zb77?G|x10%bzJ98Dfxvh>J(o=y9MyDDXM)jx^ zNRfwOxoy80VLQ#skr%6MP5v3S6?IH>>;5g;yrm{t5~l1mxywrQKPZ1@ zL#KhL;S)j9;!wiZfl|U|CPFO9S5&X%079WGZu8LcUCOZx zj!B-71{%vRkkRf74G<*j7BQM5CGpIq>@KNRVJ=F7$0e6FfFKCJIQ`tB!I$4bfIPn(DHKxhr z6%i|TWcD2tUVMZau7;!b2QekhFgtrib>08v)te?L? zQC{Z>(Uey2)j&H>X4KN+Q$k!m#5s-2c4y8(I4*Ds|#Gyms>|dG6 zf@fD({=0G|FURwlGb-WUf0!4Z-Nv4uqLZjwTB;B3sZUfP8RCCBtPPi9#@pC+S>Q+M?YL?0#9uwoJ%i#(9jkpE5(@DrVpM*G2qf6*yY9adW zi!oe5zEf!lf)=gKE?d}q0Q@Q7RaEx{!e^pRO6yPw@WSA<_&yYS)*^r9HXhw)uM@~* zCI1n_x6}OoDcYwz(m$0R@uL$^P$igM+<#{h+6a8E*SD9j{0}%N++Lry7=0_;bHnuG z6Df#MB3G%=yBL9Mg=l$)HRJ{a9zfj4%(r%Aa6fLG2Je$6WD~ih{tWve5^0o<{BbZd z?tM$6v|aLBvneR%<^*uWTxq^Nht{#)_Nl!OOH4yhFdc+wQo0L+5;mMq?!2H)sL#s* z=UZhR2w7F$dnUmlnJN5&H z=p2u=iwA3-$0It3`*T~AL~wSSWC`EHjTj8~(8J8biZO5Qm|6Gpot}<4Xjp=@MWO~^ z_Ht5%_?O#}M&520en{NJl5F-En3|YKv5o1eOq=~d^%k7$xx-Lvo{q|4`etW?c&u*(vGUgV4<6n4$php^X=NKp$R?ujxFP?&gwA&KZaLVobZhx zDG{3%#%eH}pEpRrmgQ0dS%N*KC4JVZHf!*e zQu98I-hb0$KsIUtI3>X7`i)zU@AgLjQka?P;0Q50T|Q8omzXg8jT5w5!Mich zN$GZbl~KKR)sAMAGbtb*6_Cqa>>ivw>~Ri!Qg>ATF z3cHGNntwkUL}A$3-{x%)q3_T*ZdYG#BFCd!$PHxCkEm&r$OL&Sk;&7fyo-8GnqqE4 zpN>nvvZ);Aa8`{8;qRQwUdF#&jxq#A$-X#l(u=dV;$cl%9-T=tRmi=3D0sX;ac-m^ zxaT_<<^;IS@m7vJ50x23+W!JI^@Y~Cl`czFicDD9IF3BhF_iRlcSgLbHVXIf^ro>O z2@AOaJ}8opE2zJJ5f5gV`aC_c2+DAUW+%%If*U>^)ufBwg=`(*7`1+=m5k`j04UwZ(LUjlExS|LY8>=HL@L3Y^ zc{x|PW;i+lxP4~JE=+MeM%enA#m<>B{2=K&oZ#$5+{%?pXKm?XYF?R433m1}$xg>R z8ppu?&W?snr1(@fo^9{94ON8>F}=dfHO7EcP&+N;_W$zUbO($NB z2|Lr_Zp0E+OQxo2Qd0MAZ+cPfvA{en(=gOG2=tuU&oA7$%|Q-8iB5^uePWP~Hx>VYsYpMj0jA;; z5Sg>nK9F@}#vrjv_pk8Gr`C$UxxD1ia(<+FSqjoJ@4|JW~Ek!#$LggN%zq}5p#v>!pehe zTdd7T#!ice)=ookpW6xePISzbhSmS60zO}}^i^E0w7NY;M;y_Nl~G}B6})bam-ujJ@LNlmZ3rCaoRry6;(u~LE7dK>MzD$|;& z_D8hoMXAa~)UIuVdNxeofyILm8 zcX8=neH)|z=&nBn7^if46>|GHDWVc7Z*tir# zDzDCb>EJ*eOthBIenCNffTp2$*`p1Wp@Wh$Hu=p6aRx`+s-TM))_@@y8oe=2ZZ5Ui zxg-zb+#qQD*(E)I{Qt&oCYv*gX@=X;_Q%19PH})H2-3_|NB~X1+*&l58tiOyIXfAz z2@s=CaVWc4Nl<~&&8)cFJq$8Q(@Vv_@Ef*)&f?z49+1bbS^o^bZoBP;q!8Dx7=~pD z1m#_3LsQac9y{&~uh|mOHRAEh5Ut+oghw4`AT9O7Jn2w_U3a}*Z;wrf(MQZc@7DwP zzv%!L{zGsAA8_P zQNH`b8)?Z%Ir0_=4fI#A0@;K3fsK|yd8@V9jI|*TyNH^l4i&!JCAA+Fr?+O=worGr zRNT1{jy!`MJU0bt5G0}_Cv!apr}NLvi$ei%@A*yrj~JA-h)(#$QxHbuhBkrKfc=4Q zl&9b48+0_Ip0#&hiQmM^F7=qbupJIT zzuc$?MyGJ$BdP0mFu@>(^*Jvn7*fSFujmFl={mK`6$so*E~Jx2a7f*C$j!pVjlHcL z^hN$4S75>{f9I_9!yM4$GJ}YB*gBeFyTiJx6@#76eah*?OT&4PK+}EK##5qyHKO3> zh=^u#E%k&zpX5#OL@YWE(%pT*+EAOah17Rp{V8j+NCgJX>tQ#JykG6NZrHpwj}v1S zU=5W^;wfl8-!c|m?+kBSHV$Z^n`PjS6ik_3j|6Do+2LKEBa|nu8V2#X^T2O-G#A0) zBD7P8c%DW%OPdl9-laq9}v+$DnQ=eF=3%d2$|@_rZL}B^G2Zyk=|*%7+%c zDToO*inUi_H?#-1>q(T0Lb5-$HgU_ZyMD1_Lzq>7ssy6Zu`1f?! zc1Y(80;cLfQM*Cv=X(}66{)SxwLekRModIK&kJogq%kI?)}TK#07_Ay$g7Xs9A9)~ z@lyEh8b+%J5V_2t@@yXI4?IYjcS9g$(V0bm=PyLTAqI1u08#*l>$yIX0Aw<*@;g3b zA4-z!M}gGSau2p%s!R1%yajDKnHs5v*bo;lmhc9jmlz!gNxZyN}f&gA$KM0W=-u*Lh|)-rJec>nwK3RHSHWB{1kdM z4VtQfqmRv7V!}-hg>&Q$U}kta*9YPjC+wX%!7jz3(BDki0qPrP28N{A3c;29GeevN zmP~vgbQ@(eZ>Do!w&N>`!5&$X zwsy8+)Eiyd6}-z|20je$4$ck8W=qE0F59(_w!^_a)246fcjCwz(&fq5sIKdWj9l!J zkN!oREYZ+r&!Qb*k&Jkhmu%~W#eYzm={ibD*=mQs9yP1n#{uRZWyY;T!9|?GHt^4^ zUSrIsqeJaB1qyUjV+6eCCK8y|@Bl>?NjJ@Zq&xfcbooTH(vDG6U!gE-yA7UX)35~X z(XAMRjhq31*$RI3%)9e_&uOq{^5ix~p7TllU=7(*04;pXv1ezMdBi-S?Y43V=6n^n zCc_!x-X3en(GIfkXNhT4X2HppGg+A@fqYimw}F-j-8%RiFVlv1IpG&2Y|jbH{Iqy`Co5sC-?l#1@IqL83?~H-tWaqpRR+wiB!K9!!`h+Q^hnY3jusd8cLY?=cP1at1J0{8YkEue`GaJn3+|uzF zE5MCvsAqPuph`Kpcv|g+IeQjOjIlJZI+I{aOm#tBbb>;O=rQz|T}DCWjp0URkj&8p zvhtSNp0)Nil)Ko+E4Vn29a2G|H`4}@`+>a{`RIICs=mds2YLD%+#eA!8J+)h>!s?2 zMv2jq`^kvgYZ;4yPEX7m8<}7G1O*{JS)UgeQpSSD)=xFlYK z`gkW_^6-tPvt+F{k7*$J33tNESOgF$WCBJY7nqn*A#sjkrw;ol;si&8FUgKc-4*<7 zoWzSn&WzHLqLZGRc{28%SLq>>2RjD+*dTE_%@HV)JBFtk&dGiJ3Jh2bzy=Y_`;8Ht z`zUdnA})wrNhl;>kOP6&#n*avmCgloE=rtvS|%(DLN1P*&_qLMd>eQ@2$S@7Tsp)M zCYft5G4p_w_u^S=iBacV4+BU>N-nfEvfpmB8z%o78Q9}D1SWdbcB0jU6l4X)n&qf~ ztnhx&#}6mlf0#vM0=q->_>vYer95QSpkOfGcx(8T1tm1S5~9fMK(?C~LAq>z(+i{T zTL9P9y#H{J&a6=YLge?>QQuwQ1p{Y(XE0Cs|Asbg=L{fLFi8bzbSchqCdw?&Wp6^N zYz$=i(V^Inm)fzr2}zoGQ4dUy_tm$j84_v7l1u9HZQ zWW1I!VBfZ;y<@<}pXy8E!$+8~^$U|2G&+&b-XiPJ1lg9|M;4&Z&r8%RWf|;>JLW6m zE(;N%ijvnhmY^-VWFOhOzG~_@cj1=nd4@ zaK-}u^(Nv->h*ge1Bn+3aCS{;$v2bdx7&gRZd+q7$R>55+<1n9ehwW)q*-TTC%Mqa z{!)kFn~(P#Zw@dwBlYr4?k5)V=JvLnLpLguXiVh%3deVYSmj z2pc2rYFh>zLacOfZG!xWiuEX@xtIQULU;Z-d;I+*w`qtChq?W4=8fJ!r^jSdzvr_R z8lUuq9lRn>fzA;aQQ6`&K9(7ZayI{+R@+b9Wiv8Nddb+t&y6!R(r2{fm@?bu^oiSE z_%O&N;x4OA{l_WM_YbK-Bo5q|`eWQLD``TA9e|@x0bkbhqYBerb3k=1BchD#QLGuVmnn3xWw|nc_b?hs& zPCj3BJei!<0++^-3zGf2Uacve`ZK3hn)l06!?z}oS3)Dmt4kPq<4%^*S?j|mHFs7+ z7^2xa&oj1>My+{wVCuYgj&cX4;&k3dBlIxJNMjIrrfXBwjnCs7g64AOJK{G_%64EzZvxPvKS#9{ zU>af)8NJe@$EPxM6L4W?z)8V;hI?$>aiGVx^2F8Y|4^1Me1}S^zTu~z*kbR7x}lWJ zKGSPPUMuDfEzl<-WXr41f6=jzDXkKHaW(ql294uv=!*~X`I*4w(ZI+{LObN&}T&qJr{rVct=h$ogYhu8a zq>@G-RA>Fibao4_986O;HRL1D;-(C|_G56pe+R22qg`ZIr&B=MBu@;`=5nmu7Hb{2mD-MrP)WD_<_~(XJIAEsd&*NaNeIZRs&u`B=~q> zKfbcmSLPNf27sMjHU_$7+wcfFkTV)QaC3N25J1E4y{d*kZ%tIJCValXI~TU(ePGat z1I+N{$5%|z$@sFz@GoSoOuPdobd{z>I)3D-DWytRiCB7342q9!O8f&$@HQ?7@OlhKR;r~93;+C zMw{@JnS*00eVO?}Mjif-WL_y2fukp=hUoJ@j3PpCURRP=6K}!$J$|K6kg!x9ka;vP zSNysc99MU69wVOOD}vXX1tU^96JSe!(D+2yQWk8fOulM;3zC_)XnQlCO7#>SE@?%- zu*Xjc*6|P}4gkoqWBFZ@Fe!W)-tYI^#0`>VHWagsEk}P7MjiItQh@Q<4*|;w-ZxdJ zO@hNZktd1g(upbc;jljJOK}UQfWs<%F?9kVyeD>yjq*Pvy7Fu2t zUNKd2raGk7)qDJ30+0h6a{Q@%($DPUunJz&?9V(0uLV@i-~wWs?dGs(>o1VGg0#JB z^lBM*NNNGU@J3vpicF$7 zEHYQ2kDqg2M*~^oo3vK9>>%0I-O=&-C8#$eJw@9>n9T^G#YLm9HI~P z!ld0NaMnJSU5wg5=54dWbZ5f*JB#!ZNx!KwWJl;DJAv-S33O#KDNkgCGVJLM3`?5@ ztvf*S2OsYRgBI= z@I9{$cB{0iyj-(3ueaPMuf-7#>IzsVY0!II$m~8iq_{+ouS@Xjq~ydcw54DZK3_@P zi&o;_4=Y`fS$!4)&cR8N@~6kZJ#g2^pCSzFU4h(4ql6pNleK79C(&oYJ7~UY*CgNq3q!AqN(Zm7g zU7ukEl{fGXnBlxnORiDhg!2wbPPQ=s8E{n{C(QJfF+L0{83vcZmU2~6RtTeUj({3* zdslUbF9?Lm-sCka{){Bxj6M~troiCpn3JFxj^2PXs;S8&2(mCGoYAMm-nUI4-qiO7 z2`9IY8ng(ckt^_mc>X?NiI=Sa#Tv1hXa7jL`~wnxP;F-KC(4C9#P=? z)!$C{Xi;oCKYzbX5T+lX(T+FtX+8N{3WVPC2t3_$YCMH3_-tPHOq*IbCFZ~5cl{2N z!bdkAs_oM{RHq?Bsvs z?*Q2F>t~(`&p-Gtq=PKtri9tO;3fk=U#?fF@D@)A^jn0`y_i8%U5TEt68Nl`_Ks@= z88C#K`x$63sAv1Jdd|*wo3N*LIH%zHL9Goy6?4NHNAVAZ0j{nan;IVgby}l=|D&iw zCh;pMZc>;-A1hPysJiz7{)$e}ro!{!SdlFMx1TVq7rI8#jx{bWvHRd z;cWL=!hR4J_xMqnJnK!bhIJ^W zb(gyu%BXMkC=u2{*veW-q|0w~NaDp08AxqP}(19zBxYcWg{`4;mP?t|N zc)8Kdp?b!20I92Z@(DgZ8S40?p*SoK1IOe_Kv5?+^5KPqBx|)5pOo$uoO8!(@Hk{| zzkdn;2_-aIIrpt>==Pqr&fT5b&94lfm)x~BgItrLL-#$I%L*`pBEZbmM{?4tAMtbd+B#sZ3 zG(3R+Z@)%^$!&e5uOkJ*{quP=wK^1N&aNPn2!F``3|LBH%?t}V@CW{W!~QSuV)a*v zBTBZQ4k|D%0!bJ22z~qM7>sj%3I}a|u{HiY3&{4_OCCqie)0CE)&~R?DhEDFL*)Se zJ?|yDf}jLf$ne~2vJ^t(Z~(lQB90bigbYE}khdml^WuGkQi_f)_1=SM!X zMtk=Veu*4DDV*4pQ|#8JzCXVe3;L^&*CclL15uJ_Db%Ak240nu13CjqoFDqO-!#b9 zMfeomq4$S;%-8(A>u3Ml`~sl!&e2yS?L9j>@o~2mOd<;sCvuLK8C^Jd@OI#mgPoD; zBZc1I_SN;2%3Ebosm)V(p=u6<51ER8qDQa2I?}~ zdMs<$PD!vk*go6+m#IP4y=na5p%Wb0aMT@7Y{!m@{E+^u^x-s!Kn0K`ULOmoZ=)hf zvBm$zv!E@hnU0(i*aJJ-C!_Q87r+4pi+MbLfgQQRj>?xDoY4s#*xh9uoX|A=*3i1` zzX0C1#tbg4lI}n8g8w$Wp{Uf(>L2;9cPX~vs}%_kDDfVM(_>PlTXbfd`cRcii^M7H z28a{(Vfv&NSh>Xq|AQv{>w`NQ#R>fwrb&q1J!yZyIoEj-_x%c5T-u|dgRlDCe^nXw zbMXITKVX;m-1Z#$=1H-=uU1Ex@Om_K;e6oO{y8e*&3G~I5dM4u5yriUqna;EiS|GA z-ga;)FGsI-B@ziY^t1=)?rai$T^)n_H}O~|kpM{U_;{W7o@;9xBuFi)(RR7)ZFmjN z88IU29H0W~c2&R^9uT4C_NJ0qw#}s#ZMxkJsltZt)E7k>o|Ilo8vi%1+4&nXYmBha zPndx20`PrsZzv3$MT_}eM&IUi|4GM1fbX|fVLmkey)(gCQ<0dR9t5kjuPvMKLsEDi zD3wDSlz=i7Z=el3ET#6DFlBj<-+2H6d{@E_cFI$1!z<$6-Ia|Wq!0otXVuM?VKX0) zfVdSrk|jex9tMyflLH{H5hZDM18n%G&{cZE9$--PEZg>>EzgzXMgZj(S|rhQcGyV< zf;rA6eOBrL3F6@n4EP^@MwE_J&>hZrFa`*Lt|OZC@78@~0Df-*{N88%f@H_9J(GZ% zPh^C^QeF>iHuV86a-xMO(K-l2{Q-e$?A5YirUtNMzefVkK|n3B0o>tS%HyoBk9>Yy zPP&9Y8DAMJa=_^;j>UqELq$E|dH(-XAN&20%VSB_m@5glSfEWoI4_w*#90iwe!!KM z<#h|959X)>VHjypB=rDppc8!H!(33X4^9O&;a9$vzyq!#A-(QLvh!Cf!qpx|a?TOu zGYL%GibWH6A(F0-N_9IC0pISZFKhe_ zgyqS(Ik$aOzkZrZem(dGPr<-eUZC3FSC(gpx`U zyY1t4=RDtGg=4Pxap8y2&^N1Hw?)Gc{gwJwTxmsKrvv`4Je8TQwuY88PWi_MVt6`F zyoW@~ne$pfy|0V2!-TX3zFsdFjq`k))#q5bYYwK+{dtm!{4lDDvJxO%g$4s!xDHSX zkkwF42$z>2V_Fj36QLdYcpT>{68hL4|Z^EnO0x!uEt++tvZ*LZ;^;bYUqPf}kYvn(0Fl#shbPhM+YNrq+? zg34ZPtpf!qzru4k!78CiQ{AOChZfi9Mw%a8du-*Zr3vYv9-LK8o#RXg-VAqOhzOfTlU^ zJNJ4|O8;x16Y2aaTw?Q@Fd|bq=N^DwLUoP~@P=Q~rTXJl#2!C#(7&e*761yg4!tB8 zsJ%p$sBD9)cgL%~ZHq1wkBo@NIA1i_3c5r$3NlRrH<%KAi^HS-E0C&3{t}nq6`{qv zeWAz@K;PfR2xL{Cn6M1!Z|jIk;xPJ3!MBQTYVq(we|Ks3enzP0Z+#(XLPhw_ zg$K7j`rPk{-yK~K6AA#XCYnv;_Tt?B{c#!9FN>0s<;_l6v+GPKdJS;ytO(efQq9a~ zyE^+BJoXQVIv(u+V?+2LKp2Qq{h&f8eSGK}1wmy>6IIS#FpZ4%?iv(Vh7}O<{to*Z zQ_H-(tD&q>h{wucSokv?(5q^_u;krteBKF&532cF`e!UK;gCg^IsO-3q|5M?w|u+J z)<&FR9fg$9o7tQD5N~+TVRpv4tk>06^UT7QE4kQTS9={~;%qC8gj81H-b`##q zLDsLc=n771vvi67hdT!M`qc2*hj;$w0u&wE)tao}V<#AW1!HgVhK;a$Ul~oX6id$o z79G7K?hs`A6ZtQ5bbLp+x#9(h;iDgF=uCWun+^C6It&HdIxnrkwz(mqjju36Ldye2 zdQ0I8uGHs>Fu|H@5<8U*1Q?jhlQFb4!|TGt$4s3n1z`z!Dzope#G$1m6`>ra#F2gX zK_tF~r79IbAG(UW5*&CYVyAqxl>Z4%n_mjBl(hTYxmU2%e_5CZ0MzEC;Gvd_IO4~d zqFr&pQiBSh0X+m=3A;)nDIBn==M|KD9<{m*WON#h@ERp@l5br}zP0=2_aa)05%LHA zl~J$-Iv^jd%+`M^GC_f~A4u!RFBKfW>Wr3xn))zQ@cb{+>WgpYQ*!(f$qu284ljUR za-iwFQUVhO=t>DLbC{BGKK`Mnaom|kh^FFnIC22r+?(2UkN8JNP_Q)F5fX)ZvXv2ZlF*i_;(eV3O=*@m;CSC+)Hyfh4oO zVO@MyANorHYBjjNV}Z#Y8xqI}X^{$TGdM#nsdCU_K4=#QKY4Ohmanxte3!`s;R#ez zpL&rFt4auW6XLAIT?UD26g%YRQ`^PU(tn=`DP z^zg*9+UT1F=7BwtY;|QDI4aoVCkzm4kD3k=Ci`s25 z$Uai&5(@?}P?KXH5-aD-n)JFb7mCqtmci(tn&I`5rPw$BxsXJ*koqpD9T#Xo%&pYa3FO!rKEtAH!r&HzWA zX-^`B1)TVLQKr!ZoH*aO`>u64VwBj7PW;JE)qN&Fz=+f0R&y-G^X+$l&DD?9nC5^F zdBVl-9mGtInj{EzUAT`t)!)=n9QNj4hoPgXn}VhnGcbFCSG|w#J|pOpQ;+o_5fpa6 zj>YZ;o!F@-i#`v#ecCwP#>J!Y1Z|Vbd#`gK{>ETy2upbNmAob^sw`KJIrAZz1*)2~%?XMq55r^`h*r zE$4U-pcjSl;8r;@WHvXLdp?|!`-r);FBLL`#<)%lS6vL{4Lz7FJAz>$t=}2f0i3}4 zCjK&lAhNVVyVNLT3z4yJvo?1v$aUQ&NQkeD65)Ft_Wu5RI@8WGO|P(Fl@u!HlLVs(oR1W47~QgAsON@jYMK3Tgtc*%kzI# zQ5q<9M1-rK<>yNk3WPA`DC||ty!BNyB&XIPY*j>IxkWu*I`Dtkdke6tw)JmR5R?!_ zDFp*X5F|udgd(ymR7w=2%V5zR3Q8+WA1HE{v_-`luPY>9zzc(cXl z1;jh7OPtZV-k5NgQ$+wKM8mu2v3DIb{g4rSODw(A_w+?bUa>AYk+tfOl&g2*<~na^ z<7JH-pt4Og&&m?SHq`|}JE%H@o&P)Jz|B*!R)P$cmv$_V688imBu~)Vbil~?uZ;TL z7jA9u3UPt!86##JEMuTofkzEY_qz=b6fhM60WPkpE#$7`CpPL}dq_Kw3;8+BA=x*Q z`2{X&9OJQ2xV&)%VS}R*V)7P7eGZmExd+e>)Y2Areb}U0wU|CZp7J)l8N;HjKkQ!= zY)idzle%;;Uy!I9QD*hki$z$ZYWV*Rt7S?yXvG$$d5g=zhJ^#CyCp2hk-O@TO8obV&EE}#Kdj@!+a$r zf_C`boLFc78JUqO&s10_>Kuszito9+A1jX`1$>fWoP@;1j26(b$AyBpph`Bp;#WHG z%xeD=StQhfeN76`_ZXNa^&(tlQ@l7>1EqRyf3$TW&MyRDO5b2=%mN=l z?GUcs3y1Cl7c%P(CSa*Nq8-Bto7HV^gb`@~BeJJ!WNW=LxrbEYx1UqeNBD$R2G%wK z%aF{$Z$eW2jG8BcB{EDy?WkW38BR}zB=`4NM-P4!6m71o^8q!6CHz(^ELy0-x>->2 zJ>qZWOZt?)4fFyhsL@o=9YNREgo%Y2mNn$$Fj;$JX8ojV@K-BS@WyS2#)0fyEhL41}U0L2Ne)5*42U5Vz73K!s2 zCuhF4P1_Dy!~jxDlw1Nbq39&BGw~wUy<8=a^FVDJ8f(xi69pb@f`3~niY?PuHt{G> zXET@jY(`1U)KKSBdmlpS!Ill(GAJ2!J!%zJ)k#?IUw^_u8aH$v@6OmCtdQN%IUF8{ zlgeHh@6JHat)t{KaGSPYI;NHNYzvOvrTz?WgtQ=6Jf%^MWg~hKF*gcH+3REcso$ca zNQOlZ9v@-VHC*M&u=7537yh^KDW7(5{1{`{FaF1yu(d%K(og**YxOTEAhzDijggV? znmCyKoaJ+F_O^zY1e*Jgs>`2tq|zce+7W4(qeJPliTehQk^yBARt1}$^o1ASovqL< zAYNx#S$_(2` z0c;hVO8lwC?JryoA~O4@BN;g)wSBRY>qD-z>}|81J>XXeni>WNM|u45D1|Y%v3!2A z(y%1o8kOmHET>2Vhb=^(g&&pwh&~q(MPLih_~%gpAI=C=NJ)>(2YqBqVbz4Gd$lR2 zxpk~~oYqi1;LV$Xv)&c~^Fm)_$q844* z<1YI+8cBbQ!`dxo9|j4vr*{`$G&8Adq<8y)=4box2zNU>-~k<5TTff zGPo$`FaCt}?M73_=-DbeEEQn2Tw5uiXMY?)LR!`0Ma57 za$?6PR2dg%{TV4F=v<-g;qwOIf7XY}zwI9V6U1xi1=R!AsB`TD3$}uqpbv~@E7G!$ z##mTV%RP8g=&;&h|39lkXuOiJ9>m8_u9xvQzCi@23Ps|Rr_f>qKMChk$X|Tx3AfvN zU>@jhqNQS3kY;ahx8R-k0!ar%z{RZ^C}UW7mEq^Ox-xmykp&8Fl5l*BJ|$3kJmBBb zK>AU+gF^QWSXy~=HBSrmc&)^dMo(VudT#+ynH4)k&7S%wukiHjqXi*-v?>Br@^Y<8 zj-}2gLIu+01O1jM$d4VPEJMVk|H&B^x*MX#%cxv~muB$oOIUs&k4974`FL<3A ziuk*jYSDi^s(T>;T;xLRq_on)tBJ4+mes4({-lN_kQy%J>!pMJtpufnw5R|}#H*#C z1>c$)zNaaYm6<+{EosFH$aTPuf@(~J?;uL-?+rkWI{b$zlJL{_dae>EGPBR?s}|7q z00}ME9LWSMw*KuHgyYc*1i$7JP$x+;Wd*kJMjAqJ0%q5PGHIj~;*m@lbffK@_6ApO zbOIHZ$JJ3W0wZJm$l0S$TlEa%XPfERP|dnnwayz-pdJ;TC}*R!ZiKKvSNNPJ!% z*+QfGAl82LiMp2zObS|@v1V{HswGLX_kkKa7}Fd7(Zuk-PV)5PCJz|^qbFUXokTdG(P6ELG9%`=4};Jp|xtoo)E4a(%V z%tkG;X(i9=UFQ*#zw^mTUzE5!MkPUl{Xw+}Gsox{Vm$5G5}wk=514CNl6gt z!nhCLa*DXJ0sW;0=4=HDmLY(UH45ht2n?FXu>&wNkIQi_0$0So!@P}XH0+=UuI{ZH{Cx1bo?D|?4L}cseSi zeYAS-m%>2$O2wIT9}w6U%(H2FW(WC9O9Z8jfK<+nmfXbY7`vP8fDLO-&M4&|JY8-B zo86%>LjlvQBF7q}Gs8hIKqCv~W+GVME1+7pe~kJq;K8J{&=d<(RQqc6If)202lDvF zS}R(wZ;+>l&aXJ@2_ynDHz?i%N)(|!QZBS@Tk_4XnQaO&=Cq67S#z? zqPLsYfIV*zW>oJ+`GkAl5Z#oK29vRlHl)*mmLzCW`$I@wYm-7Wn3;h8E53F&#o5{H;7p(?k!E?m6S4 z@cr{~rqwQ0yL(`!Az>sAfb}=+-=G@X34@UiA!ITSfUpch_o*_o6vn?i6miQOLzl4a z055I`9uY!-u6mCv5D;r-i(Av706TQg(WUmt0kcW~g%GR;7-IjaZ>Kt80B%7w0J4+O z1eN|f0IVu5e_45b6(t1s9fj*N+&cfO*p_O*{lW22(z7KZL^JoKYIXUqfzv3a=3g@n z^Stigl>2zq1+LGj>3eT{4pW@(AA-xsP3P*%LS?rg2=EqkTdZ;ffCXH7<73tJ>j4pf zoB|h@`@0d=J`&!=)m!ccQNv58g2noZ>Y&qgs7c&E>vWpzv|=vN*PemRHww{aW6&LA z>j8RJG5-xYo#OWHDk3pkAf*PiY-9V@>3z%7LPxNqx<`| z`D0x@B=dW3dUFB(K**+VKaWvhD}XAP*NXw%L=OTD(YMR#bov_b7KQmJiji(zocHSn zH~^HKBeLKB)B!lgO>7twys3i#NYaJ2WHy}W0nmuUk2FiQUafQeP=Iczeg*KD#!fW0 z;xR+`-lf}50HT19t{;kvt&7_M)2rPp}rx8Fig1(s;gbbe)x-8v;)L&@5 z0${x~K2imsf%55oQIj!c1f?Kvo6=`R)kx*WVF{<}KPtZ$Nb;tfD&7bj$leu}Q?#fW zb7VjrqKZeNk`7Dt4Y0$k)oaREM&PYCDEFO|X*?Cse=sWibnA9;!1guDYKJj)11G)2 z`{;d4Vh!JV+&1$Syi=Bw#EA2|!^u`AzPiq3sZCGsYX6JnX5OEbjh-=>ZGj1Pk|L&w zhUcTi3APQ}&o4R0{G4Qe1i$_4{fs2e>#;40p6Zp`!`j@vMm${ zM+fu9J-rP_gD2dF-W{lcYxt9B9*ULgHw=UJjbL!nn(otH-KVt+Uc^tJ`Onn|w@I<< z1B_&vK|@E}azTmec#j~=@GH0umro#k?-LaVtw|oSNjZE@p6XU)OS`_8OB;&rHjHsm zy3Y8kZu)$4{rXuyYF4E{j%3ZT<+NS^bZ~_y8J!fC5SlI*-qkTCr2vSbz@nzS^<98L zt($kdORv(Su2*M+CqRS^rbuEy<)BHm>lFaHQhnXn>jTvs_r70zGV02gw#Xq{g zhxttQip}BYso5`H?(cai9iLfe(?3+}^Lj*brI|cE!nVl;eEaI@sfp8U;{e#8zQud2 z6k+!lWG?S|!Ma5ip{`pc0iL$35nMifdb3Ho(WP9br83`hq8hYRmq6=aT~@yI)Qc^X zFM2C;Wh*D2ubbTM0aNb-cLCTy>U5+1WQl#$eSkV{aNWg|As%5F^L6rum!HoS(ac3u zuaW;U9P!ql!W1x4svnD#dI&z{%JU+^X`OO!}HwE5H=5{!u-Uj03p0ir%l% zEtLTD7~DFXayohX8!Bexf$XJhXRO;5fKRDBWKSlE9FHL#C@i|`9#&2>+#&jfqq*^y z{0`4EM9M4^ZL~;-S(ZW&41Ug=gVR^4)1jQk}+U?kiWh= zDu#4fMvXPnBfEM}P%))xg(-u-{j3qI9=E`(7f$&@nmT`dk?9MDr98g?_T|+O;0SvT z7a$TRDta}4S{pn7h;Tz0u?k;gPgNqk%Yof)?_;IhS-~_4wh*}oh>mBNlQ^;g z!>2~Y_~)Gq-p$g9E^4Apim_o%%HxyWiUs3M>!v%_sc*V}&S^{8rkx$kJusZ1{KMH? zlXotFOBV~ud!k+1Jj{{e+PMwP7zgBLix3U(2apNm4>CV43_A#V-P}(*l=W~ZeD2&> z=l&mS;deMiG+cIJY=s=TP}KOx^&xZb5%WZq+{<^-5a`8QynoysMaeMBWgZ-#Jy@HmT;(I(RSG>ubGN^69je z)WaTVz;ABq+4M4SJ{%l;MN^+`LnuXw1v4q%##7L2tKqdX2W3}JTayq(?9fBS@rONR z>y$$Xfkh|lm$aH*T+y;Sea=T~RDUH`Yx}HC$+&C(jvCQB!jatwHMl)LCVtBL*<(Qj+jJNJc_n5wx+p4+oKy_*0vqD|wE=+*>J&Kjq~D*i+0hkpvf5;_*d)q+X2f3c z^EkqdH!6Q8U3adf-qR=-iChh+qzm=U;1(p7kWwOJtr^A(7w_`Vb>5@MaC0NJ6W=K= zWl(h*J)E?=BWb`ZY@1BJ%;OdO+hMRCHbiJT5e)a99_xVtbl%!T1ZC8N)C?xym{!^P z97gdLB?^ULLCr1L_p)gOLiDpXkFLu)zsq`QPB*m498~-%py>4demG!h^dMf=VH214 z1_b>rQ960O$9f8h3lqI3AMoS}Pr;D9>GkzCwzkP$jJ1FS^klpnfx^~QPL@^fB={+g zIx7aFK@dsP?)6Ny#o2>1>rR+!xp-3SW6#5Z4<8wK8j9@N&rQ`5>b^Q;v55ik`t#Ld<~QHZ<+;-t_lnGm;8`A(3dU4wxM8 zhueONw>iB&gF@UI}!>92BBK>Lq~234f?a1v`cQO3~e2X8A2TDq~5-jK4;e^ z>GB4{NQRw;%y-qFQ`wD;^_OwI!_P~!sEoF+Gc~>nRVwWO^{ndy7t?+P;g2|CxX0e3AMU8;tWIHq5;%IW(}Y*Uojj(L-Ejn4PEJPE{X_q zOvk%mpc~+t)e#bG@2x3vOv|>$f&12EOWDy4ir0~`rv*;`+}J16w=@~I!XW7wgby5V ztu@x%y(??H52C|hb-9jv7=HrAFdjL5@S&r7 zcD2RFOW5r*2BwP~-}594l^N$V*$_K0S@x?yR2{${qfk=ouG@_kHIT%Xi<8t|c6`KA zy0*UF*8#_9sZZJpYM4MX?kw;1g%F?ENzwXxyMnd-kH8CjLoU-Q@nXX5XIswyilMb( z(R30w(n)82;Kh(P!@>M{JggQcb>#e#r@lb4((F)Lr;@(`Y`*zH2`ybjd){gV+CWT5 zGWx~;J5mN4@~8N)q<`we;&0O%h(s&wPbquA#C}x-`CY1Oj)2kNw!tlqt)GAF z*2s!0K`N6NM+a7O87&Fx0 ztpqRFc7m6JA;8X`76Iy;^mYNM!z;i?2{d> z&VjcBo^!gvJ3qf4cL+UCK9PxjacDXA{g_+1);D;*J4FBD)Z?e@A8#rE&&!-}LN#F; zyJ82Ug`SVt`2rO<-lEl)5Lg7;JCx&GSSXMT>=wFkc9#RTxXgVL&)-yJ8VK*Wx;es9 z$cj)bjSSY2b9OYY<`e*zZ)upC00}!sAqpt^Y4->?m0Kk^icf|1tGKkrxyqqtM|2iv{@_S7(Z7A#zJ1D%eq>?Dl$_esS7^S;{%3G#BAm0nCnrVS4&A?v-k(88Maup8@ z=l}sU3e_^Tg+V#&iPn^)uqD0-DYme4k@G(;0`Ve6!YJNt?V4%}!?h_m@|ok*U+n<3 z#2S2*ZoZGMzyK<7U~9Arh?vG*%4d&7x12c50o#^c?@be_IHZqv{5@bc#XYHcH z={p_*tDCx0qPn2UXs%PRZ$W{WNSZvcvivg6;?IwtU#TcYI@x$9J1fe^?4z6;K#vzm z2()*+-8Mj=x#OQ9CnAv*_k;F6n&^_z22w-_*2osGqkMln?rG*3IJxVGZ!^+tfXoV@ z8_X*F?Eo}F3bmCj4O=FIgv@M_>zniFGMcLq^3W8BDxvqAN}Bz9#Y%r+wHp}H9)nV7t+?9_i&*Sde|OE!ir~+ z6e4g5wcV*_J(GQca#aJq4%l4AJOZ~K&^U(1k(Zf~GhyR~wUC^noWR@4CtM@!u=2%0 zh_aGdJ|?{iS=nPuTH>H<-}j`SnXBobH5K@y0VYmS3L8BV$xK@h=4HtdX9R$a zwjZ$*_VBj!8rN^H;llB_);R2OvZH6j}=@e*N|2M$_H2 z-Ed+K!R_aIobaV>OxXDctY39m7$tI?oQ`AA8So{5$jYzxd*_Vd#L|ucR;KY3ZQzN# zF1jvIM={VRZ(RcF32NkM!!z3-Mm4hj$S(kk{a?S}|0RAw;Yo(fqsnKq zfxxD8z;WDo+4%_}4PQj?Vun-efiYi08V_mgZJSCQN?FciKTaxWq-mNQP<^F>mekA3 zv))C767nvreOt)zi)q&=QWt9RG=>*6N5Y}ImuQ;f=H~XZ;cWBubq1C=EWLbU6P`bl z^6;b4L(cCYfG~W^NG+E$fU%IUIB|(~-Gs*MHXGY*PGQQZD$NH{g8^HHhUxS14Qr^H zfw&Twy!iD);rZI`IK6qetM(3u_QKw-sfp28NPjvv;IZjQc;dj|qlvJ7uiVO41I|ug zZpoG}x$A6l8|0Uh#C?MnQVgNovMh;j{+sdiWCFdQrZ5B3DV}TqkF0@IvG%8dmZ!1^ zf%-ys6;3Lo!cP#+_UH-iBw8D7l?h1mG+}_YF`7VrWO7+{;pvqZEr&-Xv(F0mV?fCH z8zL-A)cMntz|l9LwFLZQN537#u?WGrw+rPlUA@0i-;$R8$f4gHqP>P=fB0e=4=`nZ zjWr!E^gUn(RVn`P<8R%h7~RL2R?9i?yB0ivN&_GYS5J+$wROAIBQGtZuwwAjBl}O) zI1Z1*44$wnqF*}=UsC9SR|OV4(RECVN*<~%^eH6JBOLGEiUXP6Bj4KYSHSSRayvIJ z`1kXy{f&b&GnYo|v`0|#&uX_c#) z^7RbZBfJu_163J?FnzmVd|0)mvNi9l&iz{_y>R8efGOyRC>*?kl}6zu$kMTghW?qt z^6QqIwkP$p%6vj6nGC5+sg~}OqX7>iHELA*(*|{L;ZT)Lyi(+~Kj7x913<5pTNZvANcDo>XE$p9Uh+Ke>}U)C*-98RkO_~E@WwHK zwo-n5o^s6;7_S7qu(b*;;di0LuO<#eI@SY6J(=j?B@Dn&KRLSV$P2b+h_U0X#hvW? z7j31(K4N@ZrVpx3MKkx?g!I$}4c2e7eU)GtlaXXliOyQ-M-KJwIlqQ4xVyzsFZRq` z4%~=lNP*G38TM`Y7eo;bZH-y5k#-e}IO@}OXUMJ09gwo0Fu#78WM`u$<^Izwa1SKP zL^duDqIChS6oK5xL44MHU-&p_CxR^|Q*gEs+MsZg ztwXkhGJejEME`wvla?Zx5OYU*U;_%uU)asI?YV&hc^trt-Q@r6B$zf?1A%0-URU#oIZhy1GW0y%^ui?sTGc}V&Q>?H)_d>$nx*O7?wf1QSddBj4ZqSsC2bSUdP7@P0H{vc zRmnuekf0QigZzbwakg9SyDo8n=sfCnc_6h((Wppuv}9GelpCa;qght>`02IicY9d} z#gLAkzUs<`BB$Te*v)(cRDHS6e!BAY4br791vf3PRc=s| zQu)R`Xbw~51w~_@!{KP;6;|YGcuLY!q>78B8Nf7dtaj5nX!hp!kG|eqr4ZGcWp4sT z8eiwg2G@0oVf4;3n_~o$jcQaYe0i-M>C~id#1_({w+4dwe^7lKyPHuJ0Y5AXkT~_5 zZ+?Bn7qWS732A_B8L!24Wxrx}3vJU0oszFDPqVJhG$_}orWK^74g zsFPBLT$?b47Rw*zlglm~Z#@b7TZ@91~UQW@7Y3@*7)_J+f3yiq(B{9w?JHQ`dIqcs# z%;6x{R9X5SK6X_iLSkQK+As9kh<$X!Suc}xY1k}OURt>cwjXx?X(?9~PSegV$;QX0 zsF7k|!I5-b@8EO^yFXpba%)YHG_4mPuAn%kU6Xmh8prxhN(=+!6+VKW65lFXH!nwA!fZia)!jYd1ltJ=Y1mVmnZcLqIUMY=2LQy8u#@e6(L zo}nc-u>77fr-rvm*>&l;N((OkkqISZIZq8?3!PmhZx=Dig{`6BoiT80@X&A4u|Txb zNb(ssu(1-OX3A-_T*6K9^)9gj8|>U2RDZaz!$xt74k%jwTwnu*qK}AX#1)LSZ7b4G zpsVz>N8t~qVQ^lKfQtVAZu0*hCwb4KwB*3*du!{TQjuh+XZQe4{o1__w@)>=8oPnL zN&^lur}P<7Tk>=I-WR|VnBRkb_(uDI=8i5Po&av+1exL!0WsJZ(Nwy^LC!#XR}xCl zssPifm={hWa#cO>8E-mKS4<64sPuaWPw%UAc>2Q<)qmCtoR;I;mV6#W9%RAwxMVmV z=A^(6!>4IKzVSyA=Xj9D&BTy&i_vS!gB_R&(NUg&o%KwfOO z_MM<5f3a@^QdGtKrb9S4m&JrA5Jjs?$Hl5S}$D}CtAh~OzKT}z`NWI z2!pUjR*S5J2Yo@@lsI@PYaQa&@ID)mLjzf%zJ+rO$T%e6+wt)5!j5L`7*JW;22UJ= zE$b@djdYb;pP}7XJqDUNjTF-|auj|!f@oHrl#MHISQrWWzN{Q;iq@`=H(IG?LFw7A zGQdxLodqhTotwrp5Y&&Fr-%#T6-zEU8<0RxmzLc)hE81s2rlA2hK~)wHCn!zR?72Pw%q<~Y_7N~=(a34SWPt=ShN zBJKNlQ!Ci8CM%z}2>e&&P3Y+raBcx89)HWdix#I?3Z1*TB^ zV=%@o4it0a(BPRZ!tn#aL)&Vv&Re=A;xf>GnqZkp%5~W{z$hABY(f2j!rFKvVHs&| z%ryWXrH$sJ%o;YFCxf-^;Vii44OXt)JLY4l;g|9%)0 z+kcPxe}K7u5sF%sMZQ|fcn~=n_Th~h_PvqDWPv{wM6<3b`lyy!(v zvMLx7v##O<(Ee$GKDlWiF-ZGIm_f|Y*?(hhTP>|p2u%o!_kh#Fb{A{LG0gzSgIu)$ zTX#ej*xt}Gwz&cwaRyXMaqMgQ#mMns>dxH82#yDdClR?!N5Szp`m+m_uZTe))IsT6 zP-$gWqs&&rqb*C;G6}Xj$WY1@#CPzZ#h!%}b2Vg_24&j7^FL($u_NCHnwUd6{bM!m^0Y6b{i#GCbqZzp=C%}x4$KQexHilx7<0^LyT$`F=N zn-3m>gZb{_B8s-mgVS!MjaHd`yck6@&_E`q0ihLMF6*vHyF+H#GhLLd4O*eW13my& zQpykmCo#|w$9Hllz)N~^QO--;}SSA z5@8E!zyL-_(z!;aUDREQ{4L!2h-sPIgxz!MZupa6&UsbN?CpR%Si+6s7keEtFNxhi zc;KOsZwjTs+J0W4;q82g7c1cY@%cajh^oR2bq^WC8r&89GSeXSaf!p!13t$o1mR#1 zdnr08q7a;#P_-BpZ=M=WT0kqP>PcpMVM8@3Ho>QtG}csP`kQTq^_e<@9|92C*iwmx z;6q6q`to_l%jaR-hZo@P@~Hdtpl~h(!a3bNV%SNM*)_aF52o|aNnJ#2g~Rd_d=}{-U5*Rv z=icKg*F?EfL@Wn4s65$_vQ#JLxPqLGe)IZl-Y$p4xVl8r##C6tvwKPgx`iG6FPJ3} z5ETnV(JU%gJJCtuIpEYdEV-%R-&AKECjPU=r%h93@bP51H90B=Gct5kxiOlNauCTYGq{e4=A^8N>4R!!jX_@uBZB>2SE2Vu%k zjVZ3+kIMp>vjl2B1lC-#qE#dNf0z9h;%R`K2KukW@*v7BXHJ!|EY=mKbI~z#Zb*k^ZWRNEowEYLrb#w z?#GbaoY!r0x{|)WW-~IvOQ2@8s3wOp(*=I;UQGL8R@fj{ZyD22tJ`o8EUExqi*XQ0 z`s1I+QI<~^N>92w7Iq=KP2>hk%AGDjm}8Ehj&CXCakbbxuEa& z7)#)f2db`-Z_7`2bQ5a&N|=az((N<2+s*4$kO-4-zw&-8|Q!Q-HyWAwv&`zbRC|GNH+HzB4(Mo*mn z_4R%W*<1!*+9;N;Fi=;g-4>m5uP_OgO6Zhntc9h-PXE_J|~y!~jps%%VF*~_-Q%Qw{GIM&}dAt&7KP)9&<7%`l_ z_xu#abL-;K@a~qCorglPa)ZOKCW6~d$1{Uvx5(%J;&&){vWSopL6&e3Nh3(X^?e1*JRt{UFAsGwzMI%@iyIWsdo6+bbN z(nDyul~smIzm%{{t44KLgtJf`j&TI5@jwRw#Xr;4E-VNZ@3Wstnu+wJH=IPVB6ct!7!-cIxit<4eB+SE!f`Wbr>^h!Gw z*p?U&$Sump7NuCmyv@)X+JW|gbwK=6b|};(8rp~c2qHZqIm|Qz@w>QX+oBMkc1p{^ zR}vsJPf#H-Vtl#)h>u@!X?FeCAIjnWO_MJiWR`2fGTpy>kdj)a{Y`{tW`wwB6G2d! zNsr3UNoSP{D_6I|qdh|rPAh!Z4a}x1gH2cFFmIfDhR+HcvdYhtLTDVPyg*IrSWNkY zdZ?y7kwph&8VeO}MaoJ2b%ooBd5I@l{^z$oj$NqQ4VRUKxS+}{lKLFtD~)hSix|BX z!=}U0g9A@@N0@w;9%eYVi%(fc7nKqV+o640iG<3X%IjrQKY>0yaI(^g1EkaGb(L37K+z9eb*d)!5_z!%-1oivH82?3sd`DkPM}0OJNhwm` z5Q3!WN6;z*AzEQ_H>_U<6Xbs`09y2Ya4$hzO7j}u>3?Vchyc2y7ywtY99Pk>;RuBE``$#!X(Lg`;hH0*e=hJz+!Py0<|Fk?b7hGuaP2Bjn3IO_ zDO5*`7#U~@&CMz`^qv~-`B}O}@l3pIMT6`Upl`gwg*xx6K1F(b*fkC}$l+0J4*Oqs ztMSJc-sU9$I9~31PPp{rQ$h-BDA6~BV)93kG@mfop0gD%Lhl`hM9*-y1_Kffjr$P5 zm{cTDhg#AE)`oR;@^Qihub3!MP(y9dCi@!@;zjL z=LJta=Y0&ywp_`QAB6HEy9Upms3Yvy3P`X5y%%DXIj0m2Ov1XB2={7S7~FWi)JUc) zkV{A{Mn*^S!SeldK6Oa}ELB>`-sR zZz90Tw8S!`(Kx1r#JGa+7I`p_op^05K?hLVPK8w<3qxrMrK7h6C2sp!58Ya(9Sh_s z_nN8}m1V-=t~?w0UKpA|$zu_Y&9Opon45j5^*(~ zR8#4DIjNH$Rqs-hVdCQ84)iPuL4l+hXi~GzA}jRW2sWsv`0czgx1Q3={kl;yT|Ck; zW<~@o0mz~(Iv7XNUL?Y4N!F1 zUNFp%P@mUOU&B)NPBMKBv4mvlY&moe$nsNmEa6^Kdw!M18_6T{MiWN?r| z4?V%p5vGU@#O)x^dbkQeTje!QKsOfGZFRYlzC3+1^nfxgf$ZGnrX1?bdoB&{D{pg9 ze|j`a;K+?LDz1+usGGv)(5v9eY^*K9L{6LNHtK+KXkh+(jVqF*!aG3V-Iie_vAjhU z{GjI+K3M-28sZF~tbr~x2{Og8B8k8VEDZ#H;(Oya_6Lqt<`RxPPH0m4q#FP9W$xc^ zo%J`0h=NNw?YN3lhtuh#z}md=vAF64QWFFfC7>pr88WTRdTEQ$zAe-sT`T_S;~a4z zRx>Eugfz-j%V~bFd$jiN_m!ODfH_+P`GbB^jn%d=_MA-E#?Sj;HINk*^AK6j16lWQ zCt+^5J_n?#OS8}XLDoP4ED==UlytGMpI#cM^3WoT! zs=a8|t1_4s+O0*rHTaX(t9a(#=vNi5Isv92zJwHiI`{`5W&O?e_B8E?PS^vdpVaia zQ_8@Qn!UVd-9S9j147X?y!~LN_s72ah_nQM95~dtM=xkMXfcaKY-GfhO)BP;kQ%d0 zq&T3yE%e@&0qWaSdl3pAL#T4jO?Vf~%>elt)Z?Yy>!mRD2d*l~!G>mQfGeI^7vpPK zQx=$8au34F{{GKGptAAM>bK`z&m%dcC72+rT%?wihSu|kHIt!Zlj@5HUdUt%B^|6$ zm5R3$(72LLHJlAvP~O*i>x$V!e(*IyE2|G;KhWM#bnXu!+*+1&c^!0<{QAG8o7h+_LRQ^B z=O^`v&0y{B8r9OqWhlnI+OLXeL&~L$ZQ#>@-kY11=>RtReE3lbX);FKH_kHV zl9(*#!=lE^cf_^e%xypzZG1i>xrBBBDD0IlvBLOamJ66RMI0yO23m!YDr9ld^BCEixC=3)gN>XjPIA^LGXg#`o$GTi%9jSABsmxwN}c3R{dzBZm86u( zJzPbZxzU6wnP62JF}!XAB#|zoU-+jzR#78Qg8(rC8;@k+g38}8sAFNjlXD@c| ze3T<$Y<1u>dhnFZUw!a^yax3;%tZ)m_r&Trpu^VyC17B2LTv7Y83GTSfc)48X6cH+ zfE(4?58#%wuP)qjzLNg%S2KbVbAVN32eM$dS~o+(hUmdNf^#+cXOJMFO#=$xg0-Gf zYldXuHCz@)R8gawlQ_B|Y_kD);K64~&E8)Jc%aV`hcN_A6T}_H7pdlKuaSd0jY0aP z|5bIBU*i$%!||us+yGw*Nr9T&*aBX2p-m}Pa1_U)O9ABvnqETv6;UPG0QNz>PQeri zy*@#=7bjBLfFcUSF*waK`&}+deo}ioKeohL3K^@1-hAC*Q`5tPQe$fcAj!dr>7Tz$ zauoiAg9xhJ-Nt1R)+5S2=%FUHb?}3Y)>89OwY)-bO&sp-630}J}@KMV7ToeaPb=J%}$eo{Fb`&J*QQ4M!b@0CWuDRx`v zp9QCd+t3yq@1M~2lhTZ<+Zq!?LZC=|kSJ(je`^M7uz-Gdi;eGIAeWr@0IVzRb}x;V z{F1jpCtow=S*`u4?EEiQ^<+!J?sJ;mJP_SdZmq=WL3F$QY#pXlo`6nzLa507!*c`4 zHPr;A7^?jKba)AA+)2i*818$rV+#g+Z0-Z!Q^y8di5yK1tFI*GlQ_c1N{T)^Q zu!>)A>f|^SU&#y9;%c<>iw^WT$eg>*2$xTY#=s3*BzUka8Bq)jG`G@)IfemGgp*)V zeY+1j%A}Ut5R0tHB9bCOI75(62W(U||ARsUK;i|U%w@@YCVK8q_n@(kN=r0VF&2ps zrkgR=*4iOPMUSKM+BT$a8T8&s@|tf}uDcp-MP5bhuKS66>tGJi-z;>5I6Pmcoxh)U zgCEl=+(6(A|4xBdagIV>1JdO$D;+hLLoeRumh->)-w}B9?cabnD8yNWghZ963mE+P zQTLa4T=~lid*^rC!)zrw=}~H%s#%~fu3+sWT4UVovB1SyNvRNtW zy8i0pI5-$+1FYmsC?Sy|qz*r{6q1VT0A+< zqPOj*ye&6|6uYxiq=)1FuYDdLIqu5|%fW^ChUR$}-6`h+R*;wQB08hN2Uo!vN`55< zO>zo3kE_{$U{v>pb8{NhMGOXtc}&7uM>4b6^mAfJyKU@{NE4MA*LwZx80t4yly8B> zFBXu4vDh^*)1k*@^2aOlYrS_=Bo65qIu@@#IW_QY+HS|phr-CWIgt_S}^V9pJk;{ zM9{UOp%GV8?hZ3UqBRdINKYYX=jVck?wxmg0F?h^s-PvghdN`FX+u>Wa~FV|6UIMp zQgoMmfH)>`QSc@}(7*8D*?4bajkWP0dGCPdEBV&j#={9ZY`m3fb2%KY0kdJ(fHi?= zp~~gn;V~q*0Vx*=kt?$!8$j_AuK}LPu&T+g7d+JnD?4_eAh2y3#sr_NZaLi+p8rWc z|Kz9lpDoBaFhihvvEr}iL&4r-!SNF(CYH^7U)IwG^H8d@y!x_Y004TI!c@=U!?EFR z`W_vuCL!wEI|KddyY^s=BpHn#U6ch_)!z=^+BN>=xg`Qt9?HzhOCHU*KhfN+EjE?%Cc9&oooipnmwH0ArVQ|o?-|dGV*v3dKaClWR|m%+ zv~bn7vGBH>9|{tboS!E1vjNvs6??D2&}FNpE9Aea#@_exNEI=e{KyjvP~V;H3N5K- zx!w29R%_lLuQmRZJueqswjT85QZwfz%OT~z&-TMvxR~2#Jw!q^-X(TB;9O2|%008} znx6P}cvHB`QN&}~z^E1X5wNt8!*fQy@?9C0T`s{j^w=`#)!_JSO^f*-m%}N2fnETTX}mrG-HE&h90g{yX8IkNtWDK!e_7A4dY7I z*Bis=ik|nWUc*YGPv?Qgl!?Hqq0Y-jS#>U2wfTg1e+%<7wf8eKKW;7R2;H)W>4$&y zKEY^v-krQl1O=5JKC%~Kv(Vaxi>;0#=3f%JTNSKyv3jS4w#d%F$fJ?C_+N^<`bHHe z&Q|1gH#?7PKdXE8TfN85op1Bb>h_H_j!JYlj~d+&y~2;T5MmhOx69Gnx0%sVV1T%Q z{y0k@TlqN{8B-yH7JtuojeK2?arh3`vR7)7DsupPGeDxFSdZ*xl=A86?K=Q7B zu%ERN(6P;Ws-u6x@QzU1P?w3Ap_l?<>N@+nB)xozpY9Co_yu$f=eVl2_qJbHR+0Vc zdqFof4R1p{vu%~{__eZejofn zdb7P!>pX{UkF5`vnTiaxI~_Q^P~M2K3q~MRG#O+?lI-)xN+y-2Pq%q?8$Q*4Vzuq~ zi;GSLX7{}Y0fJqBXt94!ve8qwpHQzUg{f7yv+l3{@*h`{3;Vu>*jx z)Vp{tg(*e4^HOd*nJ*Ih;}+(9Kj>5Cd%)jt5u6dz?uZ9fzHa8dwei&t58U#-2M4Nh zbLY?3ENA;ZBJ}chMvZXQeV69{s6y}_(r-4zb)6EZ)c6sJnd$E1~!tHC8b^fhe zr|!UbQ|GZn?eaz**+`P&^sgctt{;G$ zCtqI^Q4uJj$-YG~5iL1cO zdtP?eA!fctf9WD&=!(gBQ-|WrWV&OQ$mOD*<|XK)w#US9;}c*%)Q09|6HzivVPY$ z*WTJU9_lT*u2S?V!NwkzCC#Bvjt_M{XAozp$`}T(kT7 z#XV-hbJF*FOnam{8lBD8F))??@_8$`9KV3*!@$B^ZlNl-i`C&(diPa8Kg%xs zNvz!-DPYCqC44evIX9>4a;K|Ojz%jpik^e11D#%yaRqNu3X%`+#R$%CLuBKVIhd%O ze5&!WnYzx+UE>z+pc=yqynnvgL@-t}A?;tdwP(=2(a{30bwRf0xpvERPjs;Hows*I zb>#Z(3#RIl*aN%<+Mi!AD+}4$Sy3bFKjdnXVCmBZ!(>FS4g)s?!_c%>+cP~UMz-0v zK09C0$v=0ybQ1vM+@QnZp+$!6R!?462(40m

Hey&!UxXz89<5DUPdV__?|X?>=* z6i^&F0uM<~(Sq4N81?x zcCa#cH2b1%2DYoF-Q9%aNRmB0s0Z4%ve0lX(oshW)hv{rZgu&hk<)RRFKq2GQ1aa3 z`lm84AG8N%_2ZKb;u6)Ci#2g9=5i{RuFqckWRZsp>s2jz2);f;8Q;lsjXE-|?D)bh zbZqhcFx2pY{fj8_A}vBWS+K7z*?E^+Uvk#lh}{=9_m=<#qTIfISk>L*b~yKIKxmA? z9vEJ6wnAj#q9y(?|76pr6=2n3zGNQmNe(%i)5G&>zOoSjWphuziUWvM@vAS?%XN>` zb3Nx+z!xZI@F^2Vr+c5F`7*(4;;U>Ugh0qCP`@>F#d-Wf`cRh;Zm`mOTGMNK%)EPu z&;HrglZjXNT=9Cn{>tm~jI3crEz_Bve{hEK@-rZpqDwM4ABzXHdLuIlmXF}FU*}R2 zC?6x&!#deHFjOtv*@ZQJyK~nm!@Uc4pb;T@T$|pzOz#@5)$B|wS{kr)&5o>}a3W0wRSGdr1?iQI&KcNbGOTe$@f zq4J)smt0UQC^WI2CYfyJ^4cUl&AL>RZ-xU2i79tr8d3GIMbi`OO{sY!WaqV!tLiU? zVXB(lAK#YmY4-8uwax+UFPW+yAHm$bwS71Q(+UsW-F(c@@8cyN6xDQ zt>|+q_YVxEkMn&y`Rb;tLUp*?2a%JEVJA{Qjh9e%;^U)2$9ALlt+=R8N|~JV(3f{k zKZ<#HT%JJdDt>(Qn@QCpyQK*_srSwemdolLemAC~p8{Tn|V zQD&Lxl0?}uvPbqPBH3hCR%TWq#Z^R7lszIlS(VJJ7TFgiJ7ismWK?84$LFH^+kHRd z_5A+n)pDMnam@GeK8{mG6uHx{5W6oR$6bxcHMRrg?i1f&h+&y?NtY2uu9Y#d7u@YQ ze2(SPz1fb8kd;~AGVK6aE8}vF5}2oHUReF(Y+2lh&z6!B)8b;f)Wi4d;>;5YRN!Mr zImz%*p_HG7li$EoicWfinKY|_FTb@UTj~06d1JwK-QGAb0@j1?r`yhU*>=+Z+{bZR zHsMk7BTToku$RrSlMeL?W`!&_cBCcH;3_yF035!r_$&=my&gnPpJ`Ql9V3^Z5KzP7 zhfkZh5W{CXcQ2V=%yYoKp)p|dx7DP~?nC}htJW5p3)SJK6f|f1N<`aG05}F8z`5jo zaF6fWjA~`6CrixJ5>Om`Yww{s7UOC^?~bAK-7yy^ zT-uK>hI>tQa*oaiQ!+Kl|Ng3v=akw*kZ7zTsU~%+R?2Zx$Y&XH+RdDw1^2Q(`>N@t zM5Sq*k_2%IO2!u;C~N}EB6hiS)$JxA4X4m|D%mRE$^AP>Br%6r!JWK*N=-@ONc*MnueZHTk*II$G#1B)#S)@CiB|BbR)Hkj1uUPyJn}8P? zf$ztM--jsp2u6XdMY09Ha;m+QY=8p26_;!4M2EKp`T$zL4Q2Rs(27n-O>iME9Xbvd zKM60?%;E3MEPJz8!5s1@dcQ+ce_JCsO~q_D?i4Z-zJ0LNS$buMdNZBEea}H^A}Q-; zQCLGq0?)=56uq32serF0CR-YSae#FlnfhG)lHm6mJcI4TE`4~cd&KnR-aYq#oFVDyjmd8!gv7k<$D5z`M!n&j_=>Opo@#bcWL z$Y>IuJtFf$u>_A^-5p$uRlvHfQIf{q_beXo+0zdNz?XNEQ^qM;@kGp(9rM}{n$uNY zQ(^Dtk6BO4uv94vk%AoWj={)w=(olMmy{uwub*=dzUQ^~v_1FfWFD9w{G)bJaBd!E zH8-bRo#Qbxs+V7#w%kGSO)c#o~!Cuavg2h@rH3v0Nz)o`U#KNA}C(skdzgqqn>LDn=3ktNmCDH zue8aCP>)kFi=CVUgB0GL3SS^w03BsYd&LVE3inIGJ*?)Ct&=KmcOcUMTyYX)a{}c` z3_dvkNExzTThg7xgA>#htQdPh+&&@vCv4bWcrAo}zc`z{M+FrxWiL_CC-M$P@|$od zIv{t&h)iNupfMPdSxBhH=U#qm>%AGKEb7{MDU$iLA^3pi{kXV z8>Or8LiTU9o7d|{Dz)Dwxz+G?k8GLYg87NBPnyI`t>--FDfa$Pg7Wp#O`^#>)tQU3eB zoBnASIgJ2Hi?SG;q0NAVMeYP7k8!)Gnaky3Mz^dlnK=krcw2$&?2D+QW~D7twYMz2 z@VhwQe7>+8$L7e79}pKtNZo@Y2bb zRV!<$0UH-re}O@eZ2-seQ}5;F*&A(>{|(K*LA;6#N)#{L&#y|jpU?M11;DyoIH!xJ zgPojt6c`8hGY;Oq#qE*36UxD=Dm*8f06RWP$s~BxfnjxL4FpNXh2>u;7H=MBM)1H6 zVqJnDj{4)%LNbkuEHNsdV!FLfe}<3mdbBl0aqGm2wn?d=0BbTHK#@j(Qkoxnjs!C` z#UHzn9N`pd2N|4QaJgm--}sniwO`3pQekM6V;{GL_`Uxad767WunCt2DxCb`HW-Y} zlttfFxW{H8r4XPRM2Rm2&AbdK#n;ZPJ1`nrbbQXz!9iW;;{|(&E?0L}V?^zwY6lC6 z5)NMOd9~JYh|F;_LS~xBD315hC-8Z&AL}?Iqz4UqXCsy_FmrTCdvse|5fAX1c@0@o zV%AOXyiB2*Ux#;Qnikt}wWNEx^p?5~$`pHl04E;zJ(o*6@qHtA>e1C9NL~`UT)_Qu z0;Kz}SN(f#PD+!BTp?^zbuJ>e&DKrtk@O;DZ7$5<-|go4>>xADqxOyaiOsO59ZWsKN6@|kDH`GOrk94cCVw$ z7G%`E&FOQM@2g37w!fbV9{CD{QOE?78E}9|&QIRNX^j}MV2B`!qUE8pP~9>tSO-Gz z+rdg)ePRbzi>OFNG&gS(KKi}@@n|D52_TzdM9k~@xq6(%+WH@|wOhkP;QNA^=!4>F zF}5oU<09ABe2PgwKu?m;Om~r_8Se&Y%fu>VT^VA|J-8qn2nL0`XFdr`%P0Q_ggR3R zGS^96K}_rl5jR6Nm`GvndtssA>OvX9T>q1BY+vLbZ^@w^Sf_PfvsDo+hE!B7Taewt z0e96P$Uambn|Lu2qKN;uU)X-Y@tn6hzssQla_KNiRfqb8-SKl3V0+L-{U?n3Pbfk2 z$A2&p!71B&T%}EP|9RB?#W>OxA)3MexW<2n=rkvTV9%=22}dOtBd5VZ-3D0Ow^IJaum;nGsAy>&v^+x zFWFe>J=2o2o0^$J-02eu<$oUY&mZe9z+zZ$dPtKH)3HgZIB-2Wakrm@6F_X9i*q;O z)Kpahec744)8qc1kCVm%fu98jbDpzq?}-bBr+1;Bk3L#gbw-4p z>A$D2?T$%2U_T=xifpXJ8hhGk5^pz%w(xA!?oePE*CqVONhWK{Vc)HvX7eE`Jwf$3 zS@z%Jd=5USK(6xe>fzy3MZKG%L3V3bv&%imh}SNOk$Ab@hDDgbBCOrr@dQ_(=u8?P zaKY^Cu76o7G9Xsj`Ru0&%~7O7&-k)(l>4$lm2^^1r|2cP61Kb@i^O|(ZP|ubl){_4 z+}aNQOS-XaV5&%RTdQfvEZHu;!axqMaABb11kvZ@W!kt?M?iXL8LTZstvm1{=rM|8 zu#O{E%Det2R_J$Qka@m&gH;qEdBBj&E*2GTFLNwothM)TbRrh*o0MEwh~e#9rLFt$ z;j3Glf$2d1)|7n6mM&=X;>n1yWE`C6S|b~7Cy{X0B^4`acpgIAVJ2>&?08IHO)y`f zaXS{>s-_zq{;i*kuoBNeMG?{mG-xF)_OhiJ6Ro{Hv9VZoScw{1$-8=ut(8P`)UBD( z9pXm&{GY6RBBzOeR>jd)7sz)dHA0+0z?zR4qbOX6E5|w zO5lnltG5lq{~iIq3q(R94MM;e>UrrFj;h@!pTx&g!f4z^82(kU5%$7K@3Ae^=5RTc zT%#oa(xW;}kW-`B;M43hiHhdvf|ajkmOnd0TqQjd!q<&XA)UXjie|8{!xQ}=Afm#X z>SdRgSpHEJB(W!9J0qs>&(6f|R3NS+rV%ka+}i4&BvT=rpT{)NH_b*_aO(&Horor^ zx^1?T{6D8+!U=Zgi3bitl20FM2D=IOj@W;YPc2Z={#3`PIl7zcu%GXFyhjKa`BSUF zmF^AZ4wZlHhb<0eHq`J+RkqVe)`AU^FC0Ds7kr| zPAEqGb*`%VVD`mk)Qf1zPO%B1hn$mB8t7(10vjB22i?bgMi5x%DZCUxV1tmrnjf#c z_xEFNgOo%Olthy*R0Z~_R0mj}FcYe_AYqK<#> zC$tbuIsVhVlq1$R6+fcuarOzr-?T^9gW+UC2XUSquIG-32pQpeLdoHJ@~yp?{=ObM z0Yz)Bj&b{WG9NCQ@nNboOgrmNL7SU!Lw#pRSJAEX905xJRVi}9hCE2(1ETbu)WiPX zP{%$tVfxwCj;kcj^r4REf?Ray)_FD11$`N{K^JrxF6cw??z4Zm0MGb&cM>iuG!PVt zButg*Z_B_24#tjdlWa#|pb)Vx5}WViy2I(`$mb`VO%?{w7M|({`o3zb#PJh0=K~U4 z|FR~0D}UeKVLHV4T&v1)Idhy+X=KW&)PM z>Eb#3zQ3=WjqNa4IV#u!8J36C9^l3As+S__kTMB7x+0&astop?&Ne-YASlP^tvBwP zDq&CxfgVcu5SXTq^%QVI*rAKyAgnktZ!`H2DAzYm0l2cB`uo|uTlnc?X%*i3yrcd$I80BlC> zgZp7vGR-O2%8+~D<9z4335w9CcMe@ z=snfHc9oBiR9jO#1GqMtr^p%IIw58RihUr@Lu5bv+#?&<)C2O0ZSg7uBeN~g|41C$ znde|1w*F5Ln>E48kOfc1Zh#8q2b7v)U@4*LI~J80^3+oas+TTUw{~EeqBGzxfujjc zgr__<4Bt2$s934kgDl?hh-Y-et-U5OQLoUUaKar$YrQAklF~~8;9L{Kx2qf5f{RTD zH_xEW)N~ucrXO?l8!CQE4*4Hhp6;T|64I^7trT~_)nQ&IiaUvzFG^3!Dl|#x&R&g2 zZ_~Zzh`(Y4_I3Of;c5Q~ZYGAHICJE0(d5u)3$1H`Ty%Uoujq2%)g?T;ezukdSNZOY zHl~S7o_FBmX9nsKtg{o*`3)Oui|x;WSpwY0cZ{2T3}FD!|T#*_YOasF*9Z8}}- zRtf_j?_8jyCp!W{ko~mpl7T8RJconRli;O_ieOZUlspe?-B>5s3}2T2Sqt>L0MZ*| zEqJmN-w6y9Ku%S>f?^!1RM7TgA?GO;(ya6(rWB&b+Dd*YCJ_1*YyG19T~`azNa6${)=NgQE3<0WRCG%&G%Bi zo)YZAcJb(Buy8V@%I^EOZi*Em6Ti^Ltbd&iRt(FwQxxU!K+Q0ba2;^)n}=_=H!1Bt z$N_0?&dP5DQMvS&QooRg)H4dc-IsKgF7dXTL;arxV6S7}&f6&=ukdz$L}MWm?xwNJ z{wrF6FVO^P3dPpV++ecap#Dz~uDC{egR13mw(Bde4mBoB4;1m=0~JF6NG=cH^YO3z zzb1RO-G}~I{xiP_<$~G`KTjQH(eKo)KDzco%A%G_M{q)s@t==|BGVZ!e_Z3>KDRcm zl7;XCaPsd3>#FLawxSosd2j@;k~~lROS8)qaac^(0H8TG9qDSjew6mRMo|iuE3*0w z6HN`()~=YqQpNzykL5FbQvTWsaA>F=zj0p9R7!@LveA#vSKWF6lpY(4(usvCQQS*Q zfZT1Nj4Az1zxMhyQ$T!H*5E&Iy__vg^2{#`FYYoj8$bRNb`uSRq~4V^cucG(qLw_! z)z4`HQ2MCM9*ev7P>Ki8+Vf&ZdT?*)G(dQ72HP_}WFBIYdlp?p(O)8{s~QB}JB-a&q-I*YW& ze{ut5_>$c=F|Wuu%X_H31+ZbU?y65}-k8qSz2tzmul{UA$2x#*@bX!2vhg zZ57b!+jjp1EAMDSL>yFBWf~ah&dZP9jr_^b6QRN=7CD(MyUT6v&IUw=+aDV zD!l)O3*Qw4&sc8n5;I84qkDLLb;@*a$J3}!I%a$W(VyMc%WEQNN^s%wZcw0m00o=( z;VwN-&v`6&8BYqN%J;*UGh?MsHx#g^BY;n?=qr(>T!f#+dgZ6pi!?k|A7(qm#3m$; z#(k>{_W#_DWK@~2?$~`=@Ez}?!|owZj3pBmB6;1Sy``7$X{ zUI8e)-7@uTZDGrIpw3BM`*7i75md*ap6fazbiVm($>qNiZ=!X++!-d zVmqR0-kGK;B|U#^DIeiC3!t@bwpPNI0Xr!7iMC$`zFBK=K_1mLL?@V(z5La|=un^E zOu^#y*{jzAt&s7xG!B~y=7WGIZNU@eVRbTKcxE(}LkQ7 zR}l|P%QL~33io-D9-wZ_l%$>lkiWs-0cxKf6AAXQFCQLCOMrO?O+X}dnad)I>@BuG z1*o;i+&buNyrGwC?fT`#??6&Z5IzHZ57B#flpWOHXjjfYC>1E}Q7){QV8+OJIY>KQ zlkn<_26|s&`^Cs~AV6RmTn5T@OMC4OEFLRZ2i#oi_Vk75{Gx?4d{iZ^+{On*VJUJ~P|?mA zpoyo?_?F0#qhp%`8|z*qJo+zKp_cSDMGL+hiX!sd;pqy)-37j##?d&JSgMg@tcy;)$^+R|2( zsr(*wL#=E?-FN^O7$h-xDTAW2@mMpNLmy_Vp_0J?uIF5v(>^`>1InT@ zJFZzSS9E=P312ytazEeZp5@k>nU=Q{<43Wc@@0APar*+$Je*vMyg&I(I#x1A7m8$z za8sWG-lI$K<^ILdDuhpcuTnJ^)`AF*T0iwFf;xTx2!g0Y1iX+YfM{aqS9d(^R%8zU z49+t!fcV@lJ1>LX2fZRuq17%B2)Bg;0D_P2%6oNxhhGVNP`(ax4*0jp!nW=_q zLvmi0o0FK&yfqXMwCdAJL+n_6$7qY2A|$BG@$vfo{I&;*Z>2+7*S>=>ZW`H;mcE@S@8w$dQG6#oSc>5(A%mNTV;m(3RQ@d3ppO9ctY-Bg~}I@a1escU>c zQO5hp)U84?@9Er}iLVcKO`wV?J=Lkb%8a7BgxbD~pycIF*d=Nw2=G27>fvaSE>nOm z^EWASZ|O1@=L6UNjXT`Ely)F<(dqPag_Qwp67 z*>mu@3|>MjLSdtPCAsjl`?u0+hb;+q>I>TP0*q>(z^c*(Z#1lh>eHIbI)EBd#xzQT5(;56=%@|KGD^wotgc zQusMr`#{A)6aP3wC#C1b6O_%X$UrxERTKYDA*gctr|iNR{C4c$^C3=O>Q zFt01EL%hJpa*9Dk6r8-sD1YhH4@=5@N8d}y$%!o@#_+4%hs;ODVXxY03BXD@0nvzp zy;h|~LXG5u&oY<4@)X;(iX7?`cvyMmo8{4Go+(W28J1uBoQh)3y}4z(lmi8;&Qce% zmG0UNP9WzzGIf6f)}QldZxLlQGqS42w&kuo$f{DvN+J+0AMmrMrD9%)I;+T54_5V< zprFdXtZF-{GbA;SybiDrYHIWYr>b2lVHD0dLh3 zh$k213;Q>gzV;S7yitD5a6QP*sL-6Z(nYsd_+$-M^_ZA`k)>R*_nBv?)VyzSg1EnT?*+V0v!P#Q0+T^cz%VeWDCB zfX2Q8U|^ogIsLGuNfn5Q0;MxGDaM)z9xKWk`JzM#i6!pKz+TwsQ!@Ys=2mftV6+44 z1L;bAabY+jP5qX4{^0zd2dnZDj zmGN2x?9DR7PQvZ;(StHxmr4h6&z3wP-}4pdp6NvQEiU(}O@Atw3D+>jR{*1@b?7El zD_)b^t$Nw0qcoTP!%S9Q8Y1u&7Us$`S$OHzOHZXG=qRikTpsl9rYJH!`Ot;3bpQ&# zA`UnLew(K5*xv4%#_Renegk0qhy~DMzhp>Fff9yGWSF(vS&iqTZ zs%eBa1S`f7h_<#Uz64HP69WIcA%(bfGv13Iih@Db8C|Af<9SD5KvG=Q=EiTX(5!PZ zv$d`xv5ruqtX+Z=z`OtQJ%0>^Zr6>1oOTOB=!T8*L|Ap5BbfWVW|c|8t)B-p^1z4J zB1iuc0-AZPy6n2}J(^`v7@#?0tN=n$Grr6FVruUvIP0sPzqU!fn4A=kCSDQW1)IfZ z5G0uZ;9-H!YCdQTjS~=eF}en4uHyKf@lfqNQcPG^vj=2?Qa(zO&17D zOMn}*fHtyoA2v~1=n5fp`b><>F>)|^P7p8J05#vPr&2EJy2Zu#Y#;z|T}tQwir19Q z5-Un=Q2w1hB4Xj|0IB6fG#Htt=eErha&q&d11Mi7u{`XG&q2UtZgU9&F7Xf(s215R z?I&pcDrkL25&MyUgdZKqCW$j;3;wdus*EcXe(i9~tK~o$rQC%xH1zUjbO9FkeZVqo zgA4?#m-M)KJb|^)w6?OyU-bjpzYaTn%0JeC-#f9a^MsKu5#^>|!85f<1AjK-;}b*v zNx2WkbGVyLx9q@^vfp76l9*Ij#b>0a1)CS-+k$Rn3?R*wy#-RI91GP`sAW;mPQnps~|1@ znJYqoi3O-77tH{RUf(ydiV%YrB=Jn?hC+UKPz*$~7vdtHkN0)A(9i9-=qGV0shw94 zfgV3M(2MyB24KW7E8)4BtagIPyZrL|j-HckF_1|Qto(*H6hvl_c89j86Du%vvvo2@ zJ82?_X6YL+K*Y-R3eJkJ0DrMnTr8*<=&O@eR&jzBV+r;t?($X=Vj2BL0v44Y{jWo| zWw#t}?y}WzV?6mmm#WoaeRH`eK>7l1))zcNEk*5q%WKPq^b(x~;%*P7NVue5(faFTofamMz1u zbzZ|+fp@W(gI24TgnPT=k^7DN{cUah=v20ab%Pk`6P!x#s1pObvOIp z>CYuKt#UIUp|$JJf&QsOZReQhIwB(^2DvnLlJ_RM-pt7JcX$cWs7+lUk!b|~0h$gZ zGk+Z6Bp?c{9`Fjg>Q7+%waZnFD>X`v+>=rU*6u_1wY>^!)<$C*;?!&(^Nia|DpDeA z!XiGK?G&_p;WY{4=9+fyJs!FEocXllU}Xupm4k^!YAQt20;4|A=c7EG@5(%UQrzMp z&!Bj!XU*brW$;qQ(Qx@gxdKg)pK;hbK~qQf4%1u+IgiN+U%@HuZUnTG1L%utZGwd+ z&y<-ZU7h8N+MW51>rcn8$=60AjI%bcZh0po#AAvT6&N^P>*^hgH!H7&}W z;*;(F4v{O#tS>K(DEER*5-H3&7kYhtHNicCMOx3R zna>5pv6{Rib2sfmu=K$2Z^p!qc~zKO`>+tP=ehh}0NiWmrBhbZX?>D!KWd>5+m+fR zK{(&!Z3I~6mh!6Z<$MU0Q^;v!TmV;*#6HtY0--Hz$|TA$sMB%+MZaPf6rr~c5UHD5 zt?P0xE>?iWc)DlpxPle(Y^kf}vX@ZSlbn12$*B?GYmN?;-q(3 zP5gO-Qc$?R6JILJOoy_lm4Q&&Bbg9iV)}IuhUD>@oIC!}-5-*GBU9H{rD0}o?y^Nm zRgNrkyfmbpEsipv>w_{0!sl)v(i=;L60Jo=CO}o=w}vIe^iMuq47OZ=DuC5)-IEH> zyYgTKcFh?EA0~C)=mRsuOaF6zgWvjbN0R~w(kCjqO956)a4B6%=q|drp7N+)bE-In zBDSm0qMsxSc>Adj8tBJJ`6wXVCVx&h(0m0IYahlRK&O764vMt~75Jk_xD~YkK);}O z@)iL#rQyXctw`z}%qd3y^qK^$;DsBzN?*hZm^A^G;asWjI3(tu!Or6H^6T?E#eQmz z`Kvn-wrP&~nJ7Yv)^P=KUNN!HwRi69vtJ*%(wnXDv0K1SPy40-GRSRLFOsS56hfH` zxnCx+f?vdvK-v;9xU?cu?1?uDy@|W-K|*Uwe3E%dW9SpNcU7T$M#C7TXSuJ$8W>;& zFGKo4VkhLxFP~Cjs_Z)h5&U!3$$y`q9-JVxNviC%4Lxn+O9_w!C3PD4bbdhib7?{p4*FLDw0F;NM?d?&l{D4=K>FYW&FPQt<^3!1dr zrWPmL<5c#_y%-qZ0FrDIH2qvS>SY!W4PK4;CTRhviMUE3ygNo{7u9|?0EtT>0Xt+D z6Ax_h=BMZ0*fbox(o&xK`P9urK+F{8zm{nf)!|~d^y_1jm>=rCY^h)_<$1aO;JSIW z7bnIdOT8#3?-r!Q%=_DRp!A-hJXYLm#i01Jr|fN7Ziq1^zz=uS_`Ur@F7N;+`fJ#%FHc_4{CNV!Y$j z-XY*a`58_H+(Okx7)c9O;{T)i301hANG!!j#JrAqjJ)b{0m1U%(1#5>i$IJf)bnlFj(Yhr^1SQ|1)J{odp%>B@mj$`l2?@ zy{L_oDTGzY0|Sp?0gitOt8fv(iO=|YEiuschV zdl406m(Roaa9$J6%TmMRAUZErIIr)Ik7R6}S1g>@i)l4C0y6RiD)9#*WJI8-6j}>s zgitfC6V|V8kS+iny&2E|0Ui03%zUfb&zLo6tr z^>a?gu*&&GA6Nxf&#!a*dk~QSrEeeo6kdU=3%?BQJ>s?*M=Z(azu+j<3XvUs3$;LY z`4Srr!aF)>;h0Vr;CTu5)*tNcsQxK+;O92d0F4sZ4^2Llfa4egm>>wbRTQ{5&HhrC zEMOmdeAGSh?&>h@{)b7B0EOVRNWwIkD4Ae4KrKa#H4Y_s6S<*XIN?pDaQA&iw5y0v z#}-KfyWucMI0ia00A$a%1?l0t0l=lXbRfP%o<8wen#a;BQ6ftgKlR37FmG5BC``Of z52pfR0o=)XkQ`e6{euK1e`qk=$UnrK>TS=l_uk>U_R!wL|> zNJbHhuTB{l%lq~P3$h0WE)!=#fD#B|DjT63nHuw0N%~M z{|A5<2zq2~92srl(3f3BJAW7oI{*M_%aud>LzUEvOd2ml$#|*?SjYoq95A(D!>aHs zt|tRj`_N6)(E&Z|6orx`S|p7BZu1Fz&k6VNgV>VPjv)-E_4QDk4Rs__NaGX940Kxy zkXDY-e_7l*E*~KvCsgwqZ}E@WQsFA(+2tc({i>%xXSWoM&4fs3r)Zxia1lR6yj(d7 zn@q)4AYKg3G0^`|#_+kVv_l#6f1j-l4dL0ei1lOg3XjU)X97Vvi6f<;08Ym}>}{=1 zMBd6VcP#8@ps)M>Z>u6QXeR{_W1Ahjw?)l?FR4uo5va(7hq*$@d@#(N@J?zt{VmcV z)|zCEti_?7yp=>>VU8|xhtQs#Hzp+DClpmdDh})MR3bHCqv3~#{b6U2sj9*<|L_H9 zL0-cY@AU8@t z>V#EU17EVpVwu2;5A23@sG>9-Bj@L(^n&@l!{@3B{8~Ab&fW(S?Ho(pT(ei2yyYI* zoyJhqFzmg#5Nrw5h}(;vUza`OuFEOx6!klV@RJCHpWJy<*I2<0ue2_KL+u(*^CX<& zF;K3%K>H=IxdOXDzK!BS+EBoqLdGE{HOl3DW}mq+FO-`eCqwtM9yAT>OplDKRqe_0 z>6NUG?Ovzk6BE7I$*}@Nl1xaeI|0KjG7YP=Oimq5SEP_?{01fui_;EnMfOE~kO+N1kueIsP-rud6B ztW1-WJ$uW6ebVO!hhq@obc+iP2ZygXh7KnS4kyMoO@HfffRS+57mFcu0|dsP!$C|> zIGh?X4sod{Rta2PiO!rUJ(_{Wfv$b11y8= zytddHX?@IzxBET6{-|iW;(rF*&hE4X|E}T1`^boGdHx+lLtqj8xhHlJRzemHsxzMU ze>Y+@*@#yBF39eMIut{bpuVRUd(`y2W+}jG&Cnfyv0G^xl*OM3>wmR-Pyq-Le@Gf> z>T+WAF_S&Bo`xOYm3G;^I0Ox0`K;X(7T{;o;sW;||7`@8!NhQn%$9$k0hI}2NIXZ- zY!gnv8x5uLRB&oEp^BY1MBhQX_kI9CU%+6gW7&dcE+sLLxfnsYr(bc};Jx;ZJy{H$ zIZGFA%%w6xVeT9rz!u3jp}?)9Xn)o}$=Cf|Y4_GLp0IX5Q-)<2MPpB+WqblhW1>$X zLugTnWP@dpFeasKE#n9*gTP3JWjt1Va6>dJ$PU_M_vXS-jy1TIYGt-*x3G;PEX_#F8Z4VxL`TXv)-xLhKGb#K)-gy7+0e(6ah$3gmt(5Hb*kIY42k zhoxw-cw?&PS%w#MN=z2(N<=RpBxa)u;&&0-CI(+8uNX|Xz>hoft9Ru0+e zi(qYd5i~5lub>$^(l~dsTnhAry7^3W(K3;ZbfUeIt?iRN+6vBc65-&~ zpucbV_9H7batJp#`|Z{RhZk7(Ad7mK1H!CDgq&iO&K|JzI_(R zJsjV6Qu9)F^i%AEnSI*Kb{&_YBQUH`^?@VvbKKKl32HS5GOqROtL_h-ii7P?k?6f` zoCO#2(&NVuJ=>&3>{|8%*5%T4&?-*%lXOW)CFK%&PJ?;<*Z7)|8 z9F`ZQ+hlh%?iU34W`ai-niqslF8qac@`KstllbeG!Ro{Qo)O?+d{brEETGaDu^VRHzLh_65AiJ=lgZ5{6z?oCtQI11_ z1q~RX?Xe^~h;M_h&~~1}Y>~D@K}6WUcse0JO@}StZLdjwU=4s$uA}$+jO-W28a-gV zjdTneic(Ch1&#A#KqHjouQmr>37W;w1f>C2BgT#{M?Gy73}h@K0bMRpUL_dnFU}R? zu^64VFuLYkVa88f_9)K!gdhI_%?^R|7b%tEA1~%SjCIey<$6zH!>9P&gHl0@%J)FA zXqF4!xqDAE9A&@1d4v4bAI=w{skYv|(k14{Bmhv7hl&jIV*Te|?d~jhzhL!y(%ylq z>-&+2<9YzvUf7U<*+ZyDU9>*aOnDY??*$+YB}C-4)1F|tznEuQX4g6V;8xZP#>y=H zUdq)K7PEzd_ui7P_p~^KK&zVI0q!0Y2PNgH456a~N*~jWe<*#zn+sHgjEL=TitbZg zwl)y9*SnY>KsK+R{9vl|I1F~kS-e86oEc)lEbhpGh7VB+RS7#KhzDixm-@pfo{7b6 zcuQ>Zhki(t;tPPn-sP@G@hYFB%K=-9H zfVr#PWbgBuVYk3=ntc$CxTjW>HXCc^dLL%_*sx4{Z#t8;;&-4%;)G)tZeg6heXd5s z5Tu>rYs8~&MF52-CcTFA*9r?$S6Zabb11WtK-0r@n)2#0h)9)fgFi{VMW2hs%^;o5Nq$_Q- zqpb?B-&ZL1d3?4~k?m^$(qx}0_OKVh}9 zTb*5QdIiorYQ_Z}T~2(C1Of;o&l8a_P{G9V`pS6KwfEk2!1HZ&6zlVH1~=wYpJw^9 zJr<|c%joM-Fc*a`HB&6#06b!P)jI5J+JKC^mm$7m1okz;K8(1~UvMxK0o`2>8xBZ6 z7dta=L+kzXM{^9KLU_t1sA?sQ@9&3y!CfP86+rr>P^&>jWjLYLo=uSWUDd|j<{r#$ z!Vz4?mXq3dzS*@+o)xm6LUS96^_l^|5WQCcVDUKEigrIsKLXOYYX$#0IE! zYt_~PD{^YU4E8RNPS|^y)ijWHgCl-G^_d$;6xq_*K+fW#iY}Tm(t>GBqWn*QgEf!; zq4F26p?WSA@e;;JnatS$Xbi!8gBV`WbE1W}0p$DxAZbD%Uacn{E-qeIXi&Ryw2s~d z&>I}&6~A^|;@>0%I0Ytmpn&W(Lp+Rxz$-lV;?nJgw@t#*Q*1wlgwrQ2X#2`=RE1{~TIh(pgRDh+*oAOPITSB}} zC{F}0J=JnF zUz}*Y5R)K$r==Pt_0Wj~sXOY@LV4*9EEQJSa#NWAokVJ3%V7Xk5H(D|;1oFfA`-@z zXm}aMp{Y|F-y;GHe(lKoAF05T)hg%o!4)?-(>^ za*xL9C#_1(DA(3ZGcm?~q~f#2G?Tr|@g2`#>tD!+S+SrB!rT2UL0plmE_`F zRI6E?=D!Ghp5SgYRbo6d>a<9M0n8|63{IF`87mB4rmvz3W@Ey6d;@A&2Cn5JyF3TBEs}Q|?}0=#*bH%`3g;6m@8Y1c)N+hjGhX&ep{Kv! zCd{XxtAGfRzY16i4>=)>E!~p&QjqzVCn8u+YkOVgTJ93o&QQPG^2yltriL$N%IJ>T z)Y0?5$Gd%Xfi=`a-ui2qClU7fp`&( z*vp$ILUU0N{?C%Zjv3tG?fOG72eFjC+F z6vQw)2h*ouz?h`jF>npO6BjC(+aX!d+(=*kEiG7alhu-q531ZEX_&o~?OCr882U^6 z(QcFloFQJh-aqJ~U}9Teu)l%OValNk;KOZ0`xb)h)}q(-mw zpebMhD+>@IdQwyZl@gd#*V~ATpmLljcJYhz<|4%ZcF8~v@I(V1#~PzNEx!S!CJMaU zoekrpTe7?Mu;5v0AtGx8$JIKNqm$}-L%5<3Lf9;)?!vYlO7hm|K7|92TO?{hyk-j> z1NF*fL!E|xXwcwBVan>a0=*;<%-r?G=`|i8JG4OZ{^a<-vt}&0&r1Is=$=V$IW0l= z-lEXMQn3=DP^jnYGt*sLQw%gvu?^=be!Q`*doP4xY*7?U-fR7MvP_{fP^~b#U-IH+ z$w3%;lwKS8;4@gAB~QqPY7qU&R4CU>cIvWnAFyxv5Y_^uU73cONQV>+j${hKdU%Q=LGvOiy59fL&`w#1^B~eRn$Sqd z=C+|=02hzc{Z2<6^=fsZ*w$4Q<@Un*;B-Ektl{vXb34m`{K%|PcCyn5rCwwAPb9OQ zT3wuneEJj^T^pzyPP_s4nd~+Fk$3bys9rCycD47b8z*(RJnZaGmi8PAqCXu8vl9~b zrf$!jvt{uFMq@L~aBFGs3Qj_Eli-Kq+*<}JJX`mjK^YuGPkJzF$yFZWp8h17R$%9i zuzd60yg}BAn3kxyMD!L)Oe}pyNHvPwON)A;$p#5<;EpoL2MH^Pi4)UzgPMp_Nd9JBXMAw@>so6Isao)Ln~IpFO7aN@9567yR|Y77H+UDn#CX|@r7|nF&%wk2 zE~I5utsD4Ah#8aIZkVh0E-19qc!763uj|`nd2~&TpQ{0wYnaKGWt6&D0I*t`;rJs_ zr1EUUSsSIYc`YP*T&@H>iAEICk$A@Fk*$K2W#CV>Kh5(&e9%F6_J? zqM;&I(Fa@swz2__I}Q$9W*=r_0BUrQ&*HDbdkew*LeD{SZR4_zf;bqCT$LX(0OL_a zE)CZuuOp6nIJ3CsEh^0UxrCaBy3;o*SHoFH33u-xg(BDV_RGE5uhMSrk>SbfM>#OD zdTuN~-ry@i7oM&2Em!gEw)6T5aojaxsUbe_y=S*EIufokO>QrXo3yNLzo-*#_+O2- zikYN7R$`sw83^hRWF#)#jVv?+eWC{AX^`kHGbr100VnfVUdvY?Hy|jt7oe$3=qx}} zb&|a;E4^|}N!61Ic`_!Qt^y8Z(-h4+QR^3e%gl>28b%?WM5FVdTdZKdw`0+ZGXc>h z5b4s9*Dnr!Lx{JbvdaB|Zk;bjxCVTR(S5}<<2p4apEb7YF_UscS zq%Ny;9ZFy)>X@+S1CX8Y=iSXjs8O8^1v{Zk(JD*`RS&a+ow(0K{}9zZ@|mFN_ktYT z!SJfUTPPxe;ZliR0ZUHj^9oAC*2>qI6o9}^XL>47z0NvV@fbnl zjH2%muuC#k6VSNV?=ofx_W@m@*E;^x9nxS<2YF0tQpnLqFf0%TPiHe#ipRL>sx=GJ z#MOo0yN9FF@r8*5?)$e}TI*s0`-d*u$6$=BSO~dvprR2Jdsv|$j7;BG@K(CA(~3gy zCu}(xly6UQVnPmJAu=`v*K`=1i+El$KGvwxVqZa(VluI)j0_aX$6-1 z66~q@q6cdJ2ES(tZo^lY@CJa#KtJ9T5y_$Vec^&?95N;=Bz`~V1fap}k+jxk0H6gc zr!~eVNB6~d9HvXu7|UQ8Qd;~XTvK9e0}oVWCk`g4AVYfNF~SST7=W~I7Ega-i|8|) zv*X*HKb*%gI3o?$*(k9EG}#u7Uc$&;&KLdqhiS`kj6v*^6cpTsgX}2y6n3krfp5^w zWmre|xJJ)2RyJsATdxp6$Y0qa0R$Yf`SehiU<{#Y|9AS2k2g}#F5T&2kZD<%2rHw> zQ)}#T7+$sjWYlJu6pmomE7pQl^U4(~6G>RUBQ)J$ZliEi5m*1doJTxZf5O$PJ@eS+ zfiXgba3OFT_bG&3-P)Q~fGtP0^{c^Ak~LsfHGwDEM7;(*EHS_YQ|2=+Ogq0TZ1 zZ%O*+EjQ1eu=&FU_|xI+fDI$uRhZrQcJC4&1JntKwiC?b;Nhv;Q)JO*HDJt9RJ11d zg=EBigaoe{0nk8ZLB#W^uDm!F?~l5TJ1!k-D&K7veFGPMXZH=2SqPTW*2U99(6!MJ zR}fxy12@S^xHb)h2W|)Nfq~Qz{W+>I;tAZt04T_PFFJNamCb^U9D*+)XcGPnO~PM< z|IF6VAnqkRP4n{&N%S* zpE&WST9k=*J<`JT^VEsu`?UoeHM@IV-pA63pxX2TW&ONes`s_S%0 zwSln0^E{9eyh#_~AZ&OBO!LQWTg4U&rN8s0M!RCHu^fQRLPA8hhHG%gS4o9mFvRGa zlPraBui68_H+?hmQ+7ZO@DD^t_-?~uGG$Jb?D9thOc z(=?YLP!EG!6EM2B2%*zG6tFiu5#fRrz~B}V`*dAyzuSuP*vwsaAA@mPH90riPaxjK zeLkj#cPkX94>bu@FVT&}$9E@=z&l;B2BGRZ#a1_i9DL5M0m`*(77t;DBE$c3JWI;? zn#bMg=7<}mSa%mH?x+|ZcnzlVv^a>9wf6kf>dc@66K6uqnsa#ge&EJvIee*|#P;~Y zmH|_XBbJ4Ct0043nc(gwgbH zhTGF!*iOS4elgIle*j;75#2PlRTkCo>7y4@w`AdZ+(XsH4C-oq<{gKe0{P*kKYt)3O(P zYV%^qU<4xcK_1v4QU=fFCjoKXI%@DGn?j10 zZk-?^dI=*6Y(9nH!Ux;=t37!E?MB=P`0!}ud1dhYqJ@w6o1qU9`&5g(Fd5G2JJb6g z*-DO-vPKhOFo^IV`=VR!jpPW$P~s%->LHqMsqkB!mANHMgk0rT84hKfTNLOGqIL;If|Rn^$~_s;#QVUC$5?bn$YvaYT; z0IUu(@b90OZN11i3}Fj@ib_3jWc4s!=Zy_ak!rNCR~Bvj#eA}b2euHeCyI*%l;qac zRQYQym$e9!GqGPtpI5ur{p|iCH|oQ+=*E)Q6c1YI4F2jVf89E*Vx1G*`|{c78G;4# z0Si`tO1m&P@G6KJV-gHF+thNu^$t-)%B8#n7Knj9DA^yH$*h38>#u=pa!Wj`K<31fO z;O%s`(G%ME+XgF$g{rEhSUWjwbeuUBm;DMu%{v`%u@M^Lw2i7U0>~JH6c;L7ff}b9 zlHvm=3JHP!Fwns;1adqd2-W-~%v|1@w+GvKbP^bX#8XjEHgK0=e*FCdI}f7f)pU+W zZY=npmg_|lufueoq+GdXyRqE%fFE5!Fr3-G^tTYfSQo=~{vm~(F$x@}Y9}GyK;vvD zPdU^-WIAC6{0o(4O4Xl-xzAiSBeR$dbWuKRtS$q-gO-8kWfR#@z!8(Tbd@b@X2=Y} z(>dYkv4IV_U;-;q++ay81Ahw{ogQfXDd9aQ9x1|Mkt$~$1cF~R#vUe--BG!7N2U8P zZ>0RQA5$MraB0QUG`iz1BHJEDJT-LF)P*aI@n}tZOYOUUcP@C=nI{kcBYlKVi6B5~ zduEX0BL?ucS$KCr;&C*R`*KF*8*J?*oX&y` zA;eNO1*;LAcVftehD})iEF)QK^h85%EUI5Tzkz#rlG5ObLF1G{UW;mFv=N(C|u4~`V`|S6*?t6dU|K5MD&)(PGeShb$&b8LL&f_>%-Ib6z zzuYS2<3ArGT=6Nb-AaK>neqNsYe9w17VduU+U3W~R3NJJh=SnKKm;S=?9kfMf#YEG z?%PPo1mI+!b?#pm;BX02T!)<@E({Ak7!Ic{2JPFhn<_uO22hmlUAFMHvLDv8LJ83DVz5_BV+L^+B!=qE_X*@$tV^k~SW|8cq12k2TEa?TZVoQAd(28>z`1 zQlF-dnUSjMOqd!sHx2xnD(~^gFw{;X+*36o0c$?I0Bhz=!;x+)UVTfQw-GCpHZUsX zM{Mi_iG3yp>H{TY*h-LZ{@D0pCns(ek7;|qU1ipsKc{^{esyFVQQ8U%QTZ8}3HU>J zsr)W9%Kl3wjvzblP>`Lx!sO!#+ERPS$05z{5C=i6(LKTSQHo_dqlKgm!dM@vs*~gd z!eo`E)>vRk=jR(ccJ-j2g+ms_OP&&qnY-oxAdDp|DF5MZ_~`@Fw$eCKf76Cy*x^)l zwR0&M9wePYr3fhQDUrmwvv-RI3}5}6cQG;i%G}F`SXhVORF@4OzTwgVHnD>vCE<9_ zkW;pEXXch}^BLbGAFS@!)>7bYM2HFN3|QA);!^+8Wa3?k-?%I9&GidMYsQ3sr_|hQ zuR&&V$D`g_9d&Wh?wCiVIm_^1a*%g(&|F~KmzFLc7onc4a83z?f}=0u+Yr&VA;Ie3 ztDzpEvJW0kS~$@@9bQ!U0&Jh zokSpL=DTg4D!id&RZOMp zEMt(Jgr;m$LRlm~{a;AJbxWzI^UprynnxyI@mL99Qdw}i$)8J zW*P=7RMHnk0eBN3=kYZUbRpv<&mkzy7Oi+>S__9r`L5S%Rr>KfkF2Jr)<|o2QU?Y_ z>tBr+7MT_!*)$^iBAS`GH4E9H9*L`L>f#Fpv>li~%cfQPpCPUa-CPPC+oRv+&)@nL zoRsx=mHY`cwY$L^we?M+t(Q~ zGKtOf44aiVgzav_K`{*W-Ac}Z+}bVweH)#5OAf+-i^ zt)EF>B%%g1Hy`zrNTGY$mnr^gIYW$mzjf<2H%WS_Z!{S>+{X^?*u9Ky+{wTTSEEeGsjvLm*>-PGU&YTl6G+Lxk6O6TVxSz=m^9sjWrmYmCrCf#OH>&hSo!s%Y zxKZFqCx=Iz{ilefCE>#L%{DqJ;uqW-su)C2mMt3qs`|dzT8H`aaHHZ7x!JUx@F~*gP&+lfL3BtTj-F0Atx^!7x|Iv)KM+=O2P=YactHH zq!lc*ziC)sExq?y@5qQS^;Vl(5~hy%)Nu<4ASB|GNTv8-|Hql z%7QQAGky&Z4U59cgteMTkbVL~x-gxC{9hA#cWnc%!j!XLf1HZdq5J&hk4n5Y4P$h1 z(ua-&I__7cIVLMNubFv0g;sm|T;;+Rt&_nPoo5k?;>8xxY}%ui zd5R2j)&FRayoy**+YL#jQ>RXG%kinTN3HhEO5`95x}H6MeQPlJB&ieq_vg)PK|U#( zA^ZKzDeeo|2A#5T_TAweVkIiU$@gIqv>}o>>fVtg82Mr8-f(jnlr>00bmaZ)D9!Z+EWjq2; zYPbwA%t&rAcMNlv`wE&q#?8WO-o>zzKNp}`c`07MGGC|9MQ@WM5eJQtsGoWA8^2Zx zQ0u66jw7YE6s(J>ce0;ztvHR;}MhKt8ZRk zO-nsNo5CJ_Sd(b-sj%YDwFSO|0Kuu9jc z2n#np8!X@pH!&?K_ebYnzGCSejIp2&nZe#XlFOkX-i#jJOu+E7aQsXc8>2ai5(A8z zmA$va8boo<>sW=m7Vz#xaG6!=4Re06n-2v=HC4vsq!=+a!#7>uQ9qD-Gm(1)vn|Cry>Zq1)?lQXE&VIS9ls2qi6L#|E<%A^JTAbOxtUx; zd-!aGc675x3OO~f%O1X#CvYh7ezVOcWpQ<0X6t<^tI485C3WKUvzioVjtLn-Ed^q< zX4v=Z83}Ogjl|WZiwRO(-Haes)oLQT!-CbmZ}_z@*1+%K$(Ik_;on@4{B_oxu#gEMN|^Bn&}ggpJKP_-qR5f z|EPz7{$y^ehuX&)^|n4snh2QFlZ(WocM9VR4wgEV5i@CVIYiLDz8H&Orl-AYQOE%H zO7~e(9K}9oNpx$eF==;UwHMjz^>GchKZ2`7jGt^j-G8?Ig4lk`2;;48f;K`tLlrLT zorK&xM|T_IY^wSTceBAKE742$pzJ;bqrb>#ma~NIou*|vsgLFe;b5(j;eek<0bcD8 zwU+{M%DCpzXr{k^`ItUtb9~RSket*{4KJHBCcK92I!7vO@I|JR`$)cLt%(g{&dGhq zZ1pc|aQJeVY7w)CIu548PPJYCHC}SNBebI=?n`;7C?6;Ke z>^w&CKRfF&95A43tc^c^^V$gAxGMHSD$ag3(m(ImlV+Q|`@4tsv0Cqw7j(Qr0wW)m zPJ_n#lTY7A(|SkTr;hU1&wSD?7}-eu4C&EmAl(RA%HZIWaGUZgDVJ*p@+yGIRK+(5&B=8A3iVllsGYu}6dwI!-!aU48{XNfP)w>cN< zs)0;w56K$&gM+zXLa&!gO+57r9O1!3!Y0mG!MM@~9h2Obl$f^lsMqhn9`P(&rAum4 z%#Rb`;>OOzpW;4p$l(giCIWaT^T<(%nL=h-hZ4VnR+r^hQm^#585H^ZS;;GSe?x zx1*g6e!Sm1@19wi)ITmLBKIK|+tKPs4D?{4&u*KBG|}DHN-~^2J>(|2!ouoMx43}w1aWe?*XL&yub7Rqe6yJK|u|?n1E}TLFdvI4I()aZY z6?apxoDCa=A0HwvbGh<NtDX=;dyaxC`0XN{Xvw zW?+Z9iE#dZKb2cV#fFoaqF-FPjMe^%_tz1TIQzhes-?#V2U}0_c02?Ui5<@`3?xUo z0vA+t(E9^l-$xu-y!RbNwD@FJKO#=OoE&zKaYqVrI#>>$%vtBmYTqt*sb+b?Q8IjHGoO@bYKy^c`#dwW$S!0p?E z3?M_fl~DFpSfx)Vu5MoJ!l~4ObJq-+Hn56+xmM-9jV|3{}gM-`@| z$9>+A(>$mEQ6PTqZAZ(ukum9U(zDtV9b7^ptUysW%SB!|qaG4igJ{onp1@rR!rSz; z(WIK?RR%(O0RM$RU1j`MAM~xSe<<*DrHZsZP5*T&)yT+k!P^2)O#|8Tx}`H-*x&Ux zzA*hn;%V}Pg2E1MyfP4dAl=<*0VKvF>DqwUt%^bqN@OC(SwgXU&Ky?r?1AVOg(`^NS9gFcPB z^Bktfr(X;bkzS%M1_cJ@NWtw+c~BB+N3*lgNnl|FuZJO`2|hVF6OMIE+|m2|S(QoQV3KoP0kDU2<~yU_(wH&^3Jzr2J|^*@*5h zf!Ul)R3)g;&`ZC8Fi&9NS0}~ld)p5K(Z6lt0b4r=5{7*vhXm{HJu!DF;qdA4X^#?m zLI_7{X47&vr<;3813zeY2>pPdF~hSC4u>YUuLFl}MohoXm*~rSkUpUhk#ub6C#w}w z&Kq*^j$2wr1q?gHIw-U9t~YtX>T2jgy+fN=J z8r*zxNAB)IWWSc_fq?X01DdQETE-{YlF%@bm#Q!2P_*#sY@)jL4C-~`H<{$R^$CN* zEMO$J`9L#$8|~V_@f}2bIh3(O2d#TXAd(>jxVKS^^faVbPj4q=G>DB!1mthKagPlD zRMLMkeAt-I^h<#!hKC;|&U5fOol#tGfIws~lJ1kIjs)b@*;geg0Do?%iwj#@$J8-1 z(SNceMMO#b&a4-q{oL!ot9pCrd|AbsXQC1w*q}x=uKn!krbRiamZEn6T((xshm!M> zpd3-A5e8|+2`h& z^f4Y)H{3c;;clo^u#Y_ZzypoI;j{I?`wZJ#?zIZe;U0_7L5iGG=+U0K3PIj--2pZU zpj21RshFg+mH1_?Rs2-GtK6sqJ#0k?@o);wL2X&mTra2Ng@dj$@*MC>a+UO z=dz35PC*8EU>byqSb)lwY_|ehY!E^Zi4P;Wnq^*D@&oOa2*42U#xH;lGCm6rIMEY9 zSRnpWZ5no9k5L$vK&t0JzNWp_+hL@uvWWDrw>c)_`3vk2wd(5jlIV*0>kci;n?L}G zTywNHFUo?Wx_ zdu_)g%U|j+$7Kdh<3tR4>GW|7A?q_yL=GLe2#}df*!j9q2KhRMeOGs}H2{RkGe?+~ zU~Wcc4T&Qh5WHqBdXTI37kqyS-VWjKT|*CUEvOoXa-sPpb=IXDa8)UUrOlr|h8nO2c?R)}}@qt$#I*qE|j zr=;7g!Kx!#5wb@=gPgd@A>H%LaI!EH50PRffYc;2`_zDZr^zs(*x}gSoQ-;Ot=nI1 zkxG_ko1UFzklAr?y!Y;VXOV|pH%G4&MU++%?Ks)a4b!*5e^q;PMRYb{s~;qLxMiFT zna_9IxU@%A9fllUJj8O>J6G`&Xk$_~{c2-IKR=mIQy zsQjps(f5L>$Ql5}Hy7-h?`geEFIU%850&1C4E1&$Po;*T$M;JOxMo=2X~)^?oVTf6 zz5bb0-q_Oz*V7z)yIYj=zIbQ6Xk_|vJF2CTo)rFaz_oWYZ+}i9FbAtlB7;bf6SB#> z$>LxruF|eEXq-Mm>d4D!!iCI9;I#?!QA;`}+qH4_zJ+_O+T3LF9RXI2{g?aE3E8{+ zp3Yb+O(ZbqwBd?GHuNSm+h`kwFSyq^#YF9uW6>~SZYv3%YPF5yvI^W{`ksIOu^tp_ zPL!_Me^J$>w2=qW*UOAW+l;iJch_uf7Y{tC;Oq|~E`>JrdVB92{QW9XtKxjxOm@*L z(!uY2WnJ94Z!Uw0Cd=ns@s2eKCE2B-g$0Rgah^LCI9@T+O9KkBYvW`{Z6HO#?bCS zQLVHsN@qrRSiW<|-zYk2{A#}V-u1e=zjoo=@Bg=5*h+T64>z`p$6E8vHPa@9yVg!6 zLhC40hl>pb=$aVQ>Xb$+9yOCst%-_p@{??$J}*}6cl!W=x_Ho7uVvOjgK}%J`$xzC4# zNC)-nYo8Lrn8v!}6-069Zi|oTNhDOS6eZr7zu6WrHEg}QXm%Hy2NB1zZeFJZwfPv{ z^@8)QMveeH*Y=+GaVdf1?*oS9@SIctRTVp_g^xJitmvM-U*RzK&aP*b4qFt#D=I*Y zoniE%f_L6~VaT%74T%P7*pz3iYl+bH~0A1{Q^&uy%@5td7E_cf6Mz8d? zOn-?J+Er5fA&qt@R@z`T$u2N@+T``m$m8LM1+}DBZHhmXlH@STegFE;pi7ZQjUUZ_ zIy9*~WOlK8^B4I2bY4`fRLl`*^8|I3{sdYQsg9#2)P7?=@Je_N_ob}QJFNT_rg$Iw zY#|p$#&C!x1}eZ7Pi}YZuDLp#Jg2oLXm&!QobPPuldMMo@|#Li37UK-a=H+;T$`3T z4vziOH**8af9uZ7+Fp=(X}W9Awh2`Q@iO?&oeRLeR;d!W(z z1W2FOj5)ySD##DD@0KX3p4LCH=}m((R&G7v(keQ50@0(9KKAifV7{n`lN{Q z?GtS^&$fPijBtB$-77eN>Z9$dtV%>V9df@$jg!(7s%+9ef;_tH?Jo>`EHb-o4#6?I z#p}w(f>BEIlt9x5!L6lx3Fll;Z4S56 z-a~m75+k=VMPF4QTn^kyy2F$>$Fo|&%l2+932=6sZ2k^#X*sDihf)e1mbEx%yb#*{ zYrNRcA^n?Opx*|*3U+jJMqRb(7AbR~2V zdE6Dnm(AIn<>7RnCrzs#BY|S8RrF~V=xNYA52mYpr+QX*?4_fbZ&vo;Ow%S0Ji6Ac zz1jCpjqs5<%xxoat~uXEWld^NFs+f-_o+1B?UzDVSV0Ff;N8*RKMZEoT zG~cn#rBNDn53%jK?!GOu&HgpNJAxCdml>uSSFYo;!V;qA?mG!30nkqxBs~yauCCyP@w@U>kkuohEOnjp) zVp3SznOzhyy!S|MBGG$>$S>FICv=Do)-}@Y$rGXKCOl41zMEJuuekn0;;l_CVM4QsCm9*+B`rkB#qy zc2-NL0ibqQ=Ev1Q(&YlTW9RZG8;3p`bNj~oG|9?;IVD!zy^jRKaXq3@ihCQ=To456 zvWl)59_vh#j=!|h;7Be{M}~9qx)${G1s`I+NOVwoPAYzPtErmimlemuEh-%Mo}7NE z%iA6gr?RJE#`1GPZ-ejc2i5=PARKZnUA;k9`UAP7y5im9`OC+x;tO}YCk!(Qu!pGE z$IiBEGsZ~Y?};VPAyOQBwi@tLS>3KEN-!}wu7KOx{SM!U=a5|+2lzi6szmE@P4?DZ z$#X<-qWQqD^Fg!Oy*KOPXMS9LE+H(}?Qx%C3ah{}%XeM9`(in98{oj-TUS=qxUn1&ta(LW&4eQ$!I&N1iO9}}}gnw)rqhZAK&PWBK?$a}X8L*?l&)W&$-qnJE_TP0TQJ>od-ugW z!rh5$iB)y10z?7sZ4X2iWE&|~wfV%^3ilp!>Gc}tiFgbAf^pUIo>7hJ*|zLX5qZxz zp{C3zlGqciA21qN^brXQOOg|PiLLmd%{2Q+a1HxB(&_*2Q%9C{k_vW zu77mjVb^8wF^p}#KuapEypnMAAos~NrPqJ#0gHwvNTF(MIWPAg*>47E^ZPOTiMj?Z zY_#)wNG5}PS6&jW_L|VK>a&Q<;ijDb`D5R{7Dvy&KI>4a$WJcziY%9NQHj?^;dP9n z-UXTh>tinzSEWw$G|NsU{aa%vC7|uUNA%vWTMpt(>CH*){v(<#8A^8dPSNkEO+FCd zv}31MVG||a7|DqbUFYI;53)84d(Op6?-h5lM|+3mChWD4CK;^g*b`}&Wuf7;b6>(- z@|D(xoKzy|J@vMchvc-V&?wIak<*?&qXADKz}w@cy`o1%C*FZb`cG!agQ|Bw(N8n2 zaJFoDzxSZAc*e-?K$lqQLnIkNBJM6D!*=QTQR5p@`+i9(-(X`g2NXCLb(z7pl%kAE z{q~SeLeFA0_rE*ty6e2w|5+9NxODsW%QiTyNuWn85p_^i`p|Mof@s!4iFU0Q{-F)l z_kBJei`77TohJ?2Nn7@74=?$DNR>SJAguQ$_)8YJmxYA7^a`qev=6Vpt5V}#R zSk0WGAH!Fa_x`kEcwCi}3dv;}E!}{3!&f0y)CeZ2Rcj#Tf0Q!+#ydeWA=-CzT!-bM zbZrNj*}K95PZNF@t<9*Pv8+vuboj~o&K$+}fQE6T3vQFwSF|`EQFpdNg`IEbZWfhH z$T?xGdC?W>YIb3b>mK1899dv<;#|I%8ars1XcA#jx_X2s^=HN?a% zsj42m#>_)EJty2tOA21H%JS&@(>so%P<53^Jlnl_(@Ov2_dAI4*v`&OFeT+%)r}@; z1R$cz&TKf+nItV0IOD-B>uTD0n{r!5;u|`b`tA9>{s-Hg+n=Db&IXcx{LFcCW`wIb zcR9GZmtG-n@%LcRl1iL^@h?vxhFcm*q~^a|XJ{Fn`M(>5Rmdg*ViaF~yXfMhkHe@#ac{O(F`a=`*5hY z%nyzDV6R7T^dg|b`87`ZG;iwJy8Hanb%FX&yJgF8F+hpTWJE-gNXxh0Mr#4u@I60* zTBcfbk2RbNdY+Stq@fYpz&Ukw?ACs{Ff@sBEbrPkhMZw?>AyQbm5nXR+lqR)3sD5< z38xbSKzkqZc(22xbb|-P>fq*ZDS;<5$G>#79z(G{jM{Cp5P(Ztv>7H%foABt`|8w+ zo;=~?oGgbCuKDU7zYEe(kGiX9kIv;@1F;%sUUBYbZKWnsRsJCHgdYSCm<@2GNSmkQAp^yO!^``-Aj! zqGuDL3GyRp_?z#~#ipRvusq&22~9VMthycY7e>A2!zK!kwCAo2m2Jul-;u^^*mzmA ziIm`?ZoJ(9nU+$=dwN%Kf1(^c($9SmvTgKrzOU_)8 z8Y`ZySh+Uv>jRqbzFqE&5cTsOY9~^0LdOU}P2LeGr54|uY_g#$gOiR?8Gp=w{!~yN zC=!xQ45G#QWqiNCLP&4J%vaKo0)%y86pKoZWRxE8>a;oA2$_n|OG>yr?N~j!*7J5) zp@7qehg??6jQ2G1)D5dLnns6U57K_v49*-;e@W zD;CY#((4{*&*9fiXzNG%myAnLer5r5*7Yrk)W;)wNW-k#D? zbfRZ^W`zchp#o{p={qeg@Ra?|?vxGLcc$LC zUVl0Ot!efiWhcmN(dG|d;YT$oE;so=ougn~#5tjN1IC|PmBbi$t^Q>YOHRf5zq~5)T?U!D1B8pD3?vwu&tyB z!E2p>8c0C#xQC@;V90 zuBYYgefo{rbLVfI-lavh0~_NbR7MNvdi{(X~NZr62^O~F|CpI(rYcx|t%OhnuUaMg8-(H485 zX(C*B%#m8-Q8dqeyTY$QrSxl1fwQaK%nq@5C!zM!eAfuQoCpQuUBmrha@u4RB~9z9 zUF}}8gS>OZeJ^oumO=1RASsY){_N)hFZrL}wOiIZ<;{G)m+Z$eFXJ_-y0AF6gmtzZ zs#rsAXjq_W$^uQk*UCo3RzYs&PU!~?FYuVFTmd%`?lGe`Gd220vKUcAI{EDcs!%)b5tZMQ z90z@-d3yY`Or8`@wThMaDDsQxT@0Fezm&8t9dAKnPIh@|N{chXIOmlmMRmf``DHDd zPcq0vS$_Gl_8e}sWC2xY3}tmv;fam2n_O*oPioIBsepXDgPQJLZAlW;%e&&-t@4VV zI^Kh4=MjX89m9=63s;h7G@M^2fv9)G6;MKcG8oatK6P~vfS`As<>;9OczDa!pakw>lZtT2HKq9o}T>p~)|Out$A7rWcf zh~PK)bRExjbBY{Cuu401r4Ft_LLLe}!)ZS2p;1%pGO zSVSSW?20}7PEWh%lfYa01iMM9*4Pfqmn`? z$Qb6w#VV{MFxh_dYxH>p%!SX<-yhdVoSl1^zV-s9-g#44kl-$b0A|qL*k!R~qEb$_ z0+Q9UoYsNg6z8VaQYGVqzQX1e9F&(Fs$(=dWTc1mEShZnyMZ1XB7odI zr^40wf`iXeKtWsofP(010CWB1{>f1RV%h*lSUeb!SO}dTXmp5W9ui-hknlg6Ij*t*OJ?w4_q$_%A+24d(H!V3j~ri~Cpa1C)fk zE%tk}Uip@w75F)vR*b?vf{E6y1JH;AsGa@c?4Jo4AOH10h%>8$X3AB$jEZ6}%vq$m zPiHCOyAQ9XbXx#ES+atK3IAU`iiscnM-z`@35z4OGs!KP&;P$M@kP}068c(X*4^S} ztcZiV+j1oC0b)Y+A3;nOQOVJ)(k2Ctb1U1Mns5pWyCbA{fDUgB;cZj>2&Y!1tSZe6 z;u$RZnO|@CKR~pYH4kK(_URxHsLgLXg3MWOg;4VCzpwykV0AI@Pytm44LR@3$yps- zIPb^3ou3d4&+ABGB|r*XaXKCNqjLjx1YYN{3II}Z3;?kks};?ogAcyywcoWHboP2r;zSmk=Kc65I2c{d z2K`OMEltf93j2q+3OuEt+#Y2u`GwC`Y*(*nJpn`b4178aVLQbTj{Kt`v~#pR;QoCx zR*_TUrNpoT^W!rr0FPf|VGOV#cpgw9-rvFsynRQ#2yQGFM=EJagZ}lfvh|t~#9)^J z9JqQ(t~?RD{rO!vm1}zDL)L;~2VIUp{ja^+t%mV+@F#`w1~zfTpA_zX4aLXHkCHVz z^x%Yb;e;#0)LbMc&V6|Jnv}H7@gq)J1}ng%Bf#WR-d7X^G5!|?G;M35pYS%pCV9$u zRoo~bE_5YaXzwn)z+KwJF~%`l2kzPVB11B~OzCQ-Ao<{JV7I=5qVP9%tB_9MR)vAc z#Z2O1>y`1Bz@O?~i`O>n`JhpcPo{uA55CUJo z)LMw>cLhc63K3ax3VjkU*L2i3*QwR_8RjF4@)WU@t|EbB@YIX~Sf(?r25-Zp$osZA zjPun_rCVFW4+p$_PMV*pvRLCu;bARmW)S07b1Bs8Sx6UVpx6?-&hy*VWPo2*PA~hO zntXrR(&W-yA&l^OK_L+x3=^oISrgt|8y{Zgy*~RcF3;gWOm7IuyFh1)EU#@DO{_SH zSb!MUF8ZR8v^4C4kr>{EU-(!$W@CJ^>ZfIE@x8UL_|%Uf3~`Y#q>=WIVQ7l>=yq+y z+YT1E5FQ6-EiJ&L;bVn5r~`))J4S$Nmfr+V?lCUnxOJjrQ81b1Hmhl2G7Zp=Y)dOSbnkR9Yygo zc@Da|#A?_v;*I9%IM9{o#9NW{^EpNGI5>D6_TvYO!@6Ty?NR6Vws>6fS9s##F=FnF z5o(=Fiy*t9i4BgeNmopr`2%0hzyOc2I@X1&EI~ATs=_=%A^;XJiE;ddt$(oNffv+( z;Oag)Eej@y;~_)Gco%%nQ|w>`=x2f0%4dE#|Lo)b|7CpQ6riu=*dn(wRh&BQ1o=m5 zICcL!ym6+L?SIHAA%1=Q!~k*y|NGb!kF&PgFhZTv@z0)X!enGZ^UYygnNcvAs>3Y* z>|(ike{V!6(a}Jf=96A1D2O!8AT%%S6CJymWb6v1*YpxfI}l!wJUWPG_#WH!kKyGf zOfdKGo^vB~G1BqTQkC$u%{Ii-I#DT4*>1h*gG_G?@+Uv}qa4x~XN-d%2osZLLXPDU zya(3@{X|?JjIMf8&$>DYueE8zgBF&J5^xHW>gnXY@i{m& zwVb0PFSG(XuIEw2Pa?1g>|)PQX)|ZoEMdO;_$UQptx~@NUUfF?`uSS|P7RL=9BNvC z>yoeW5tG3PbJTU!$an;B#566RUDOG@$jHrvKh(njDO?EN#fYc&ykS0sWrB0u47YG? zn~HRqGxqmNUjNS3B<5`qsc0nMtKr!nF#An6!e-5M)a0x#@JWq)xd9J*{Zf(S+ zJf!WZaQJ!zQje%WD0^p5n(XKZKa3|gA>9_u~%ssG!4vc*8AX+Wz=#FgN~K zD!M9ZXQ11;WuFl>-XZoMzq5FFJM|c$O6@5oPp9OArNpeA2ON(mU=>Lq%_k`&Ldwg1 zSVhxSE7Q=o=PiNz|Kj{7B)JOjaZI{hF2BJCaY^e@s`)-r;R6q$v0~A>07;$$q!%4< z$jXXqZZf3p2^3Z7V<*RE30$L$O1?InE}nAQu5{Kch}?p_KsSDZohs7cBGSi(LXWmi zFYHSaTi*~}0txW`{Ld&iA8^eq8KmXxb{>flym7E#0r8gT!3GR4 zgC@>my=knVgtirM_O-p1uniGBU(~_hQ7^ zju5S&dye+FZRVo}6s7+~)F-eX0~+baKcT#*tgdwb_=myG_wVMq6w;4>-zeiOLJ|}{ z`tg0MNS@$7_jjZlpPd$H^xa&b|Fap#BV$+j${vubs!IP2m_6n5sOoS^}KA=;IGeD5L%sFL zU-ec2lI^tLORy5faytv^-3kYFCI!5M0HI8)rqO}&wLf!Q<5>I8tv5%J%p4yGKo5Du2PFSy&l;N7WD)R$-g4vfEHhZ0PTGxcO1wsqMD< z+rHOL9SY;Z& zQdHfK(}u*JVpmuscDJO}mpOZ}{B&!RH zsidgu+N0(gZmcbKv~Eb)M;6LkG(pmKfdwz-c$y8-z%tNTo!`damkqhVED40Xz+N#%+ zzx#ez{J8_=&>a0^_%Rh&Z7oIvB4+5n1w8W_rIKHeb#U$$t}0KW>o}-nMTu z{!gu$^#on+HtDsr=*f)lefTiZw^L9R_{7rt;|~Q{6?6{iI&LfqvH`5zU&SX7x}Vmt ziJF^-z`{9rJsGH3XeWa@t^w$kWz{`;M}acOIF^D7o7>8)wiAJB$;7=^DsePRn3$MQ zGK(VC=x9B{NxhcTruLSPh%5|T2B9!decbRu4OTmx&0t3wNmUAU_ur-)9=LkKGk^Z& zE;f5u#@O!fBZLqgQ@5y*wJ_eoY<-KB=>4792yplDJ0$9zL1cRp89*)}3*%=ZRh9M1 zMjJ%n%Cc^cv&rBd8B2-RzkjpRH6>t{~M)uHGLXPb?duq7Z|C`yeD~ z94lr`a~#Q>)53R*vG&ikGXmH^rrykA%%H~0-YSp^2+rAgy}o7OH{pV+o;k+lhjSm_ zmaTC?IDKC_`RW1_(o_UzzrDV7)4$6`Vx-yrZ8q|Nt<{ZB+DRx#@Dw?yZ_WFssX8m&Cl+3bDPXntoUFZdftD%}JksLZ&>{&Z^7 zr@(EILuKdp=C@D|eG7UK3WIs>$_#Ht@PG?zc8vtAs{W7OU*Px^esX_OlsL=1Ol1Ex z(W2;AMIUEvx*e~kwUFCL_nG=x_I=Z5$n{uS{SR(L+&^SNoH%3RjJcP zl(1)Ul$$5-wU=P%|M5{RWki6QZY6%rg)Z(Oh7kZK@u#+lAE_smwDcA0c{Z*GVZ2&B z5Bcj8L-~;?erhJ&oi3A4Kr>lcS~M_By)(!Cul4KdSj(Mvnx?R*r48dO%e zN1wX!_^X_abp(aAcy!L3XYDFv$ik=`r%B4cSU5esm+_-(m-NJo$4mEXWgZ_9k~n)? zHSI*#rcAb1I6BD)&TqPPxlHz=COf7d?&@DQti5Il>#vVi@*;}dHx!Li6d%sf%P1>hN7uF(+b_(e> z_QioYWXJ8E@BrT=)KGjgY*CAq2DO?L{CG-zqR9baV2K!;6oII-!#lpDu`KPM3lX7y zgV4Xj7#^n4*;eT@mzGqkXiI_V^;j&d@lA|w=1TP_Dv!fYxSFC<%?h<~&EPvjE zS6>tnqB1TYdW~Rk3?2TPgFx|F_y@(6?es`>$v+P)wo;83H{S0+9p3IR7t0_&*|D<7-4Za*RCSBSNoV z&VEn%#&Tp2zVb(%!azb;;q^RR5CU^ji`8FtoYxCd#k7Tl_+PX5H{R3QOO|*~d_SF|x3<@H*pbq#3YE%jK|IUX>Qcz<|N4ST<|!Cd z_bM~6wuFjerz@;p^|9JImu2LzpG5El`&q zxKdPlE$g`SJ&BM*(F*DW#a!M&+7)dB+?go0-{BqplVn1_6Mkmrzw9{vI&D};8{3o}598nMGjWEA@6y1$%(PxALcx|eE_t_vnbP{_E;MsxJL13L4Ydre} z<|EUcb-|Bdt_8e|QJetYNq?X&zFP4!^?4WEMGoKoVQ(Y|5RwkqQ)1w+#RAA;Rp<9v zB%rJq{jWI^zXMyU4aHhJbi8~2gY+@N)!=55 z4n@-473O-CUignuJ;qk?$9p2yBI^s=_UqGi();Q(C-|3HH>Y1(wBsq9`VwepONT}x z*8MmuRO#roJH7w%*|l-N4Bp5LLnE9e;fl@TH(b-RYOAzEhKeNPExx9QAk_l<-ACvg~L4PtB3o@cUUslf&v2=Hp%v6S|W1YuAx`R!?G zD@pUxK!1kNUk8DWD`R4)TyUxS{P01}Gt^MQ+WjN7mK+15XcL3Btn%8fte7-OC zLLj(eFN?n3W0YX;k&MXp!z=Am$1)ev)KiX$%DA2+rV8IaA%} z+0tztpyXZqyh`39lXM=W9b#IiZ21g0lxwd(T+-h}m5-h%ljtnCF`d9L){~tXZ8y|J6U6(2kZ$X46Yh zki@t-^oi@HN~-x7rF1j!2E1Rg+Lc-0)}|YuuGuvUSBar5VUGZMrP{uBaohSH94bau zd7p;~ZdI?Wdma7EumMf1o}s&8Hh;hubONyG_5O#P7xpN#6m~O4PmW`VHZ3qO z5~K&ny^a-%pEg5g-ZeM*?9hJDiM_g<@;NMzkUmmiQ@*q|dst=Hg~b{K8acB2jQRe} z=t58nGN2tQFOtAoW^z04XxAADBOE?LT;@v83^%V!G^ou6|9!H4@rqOVIKL;nzTEnr zbR+80LaG)+j{!@vsBhh?idutvqz?UNg$tnMQd-2LKymtSAcl%9vWwWp8shTC_hpWW zNJWeTD6DmMIS=2lO$Tn80&=y%f2?-so{KL>46tKn0pC8ERop!IMzVx}{e4FbH&n^C zmzW_PGhY`>djO1Iur<0P4wJeBl0{ zc=(cJL;w6#k??fW|3^lo)H|z@Hs(CHz@p1)UiDLAHJHeZ;b3ku<6c^;2Ui^@>#cxz$ak zW`KsffBO;`fRceT2espcbVn~{Yw@5LSM^=v7ZT1OcD7f#m}L73BuW#e+GBL3({FtV-?$I_ zKW+*|6X}ddahNsYcAVWL92VvCWfzVF8@Z9OE~DchNHY&L_>A?hB)=dh)0^Y^+nz|iCF_{OdJoZA+UUKP?pim{1wnmYOU z!qmnpF#~h}QLT;?sp8Z9w8Y3PScucfs|B2U=GkN5uc_^BKf$z0S`)W9xyE_EqOnHs zwi99bWlb*){Zie}R|3ao;^%eRK@f99f<&4EnYKm%(R~0Ez8lxXjsaC=I^AcxwPsi!NjWq6N~yJhS3@`8THUhE8Hqca#ews^8G?3p_D zY}H^6R7h2LG~7NtG|jYR*pDtKwsItrpDlYgno_!Lr5HM z@o4$M-D8f=3u14Ab>E(?alX^{!bWSrN+k&))HfZ>*AMYFB#&7LEf+!fy~9q3GHQp}a-$*SApB zLSlpYaSG0N6m7y7Wqah;BWVxc+@?vUmq%TveYW&vtm4$XuMCI5qXc0CThRA}HR+h_(in(c~VCv<9C z)nG(iYAs5ed75;k`^V%$M$2L7kSv~JINeQj;MXR}H@Nf3% zlYWx}ATzI+FFy8tY!X2%rJ(eRB&WFL#`D`IHOpz^-Q5gFQ5Zh?*RhExmF(|`eeg{O zJr!dzl#X$39z&d`4YwSOl4+wrko7upR^O^lGIuWYN}lxo*LpXT_1?rQy#XMY>l3pz zgu3}S4eMR3mM}+I?;2jL_vwvoB)`w}qfT3M4Qzf+>0V^1T)NSl{r|Ys{J6%AX@Oay z?Zr^R5GpDuC@#A*XAU)tyxvPt4oJSK{${&lEAMN!>K@y0!K8wMPsJB)w>qh_CHin= z%^KN1{{)mOY0X)4@HPxn3oPpkp_0#K=k!z*RTpSJH9!-odn#hC-otV2hrV2)VjyxKGR7GtlQ{G zNw^*;ht79@K$ym@+?VsOM+D@rf7eshVIl4;WbvV$h-)N-35mn`uh&RR)|JYEHRM>k zUK0{Q*TQw-zL!D2m_*yHSEBP6f6}+lZ#Ec1o7@VW_+Ep4I{Lc?D#hsP7tA}71(Auu z>gJaW{X$PSnUryOVW8)86{KspPxhoN($5ux&C+2M}!A&Y*~USqB;ZIA1;qzD-kQ#k;!iOp?d+42mu zqZluQKBlYGMC|k4c>n(OZaxW#UB!;}Nbx$&XY&0~S2`hbEKbT1WlGb}4Tzji;gwB) zL1-}Uzh8rZUQO5igI+yLdsJM0&9*esQI@y27%7dzE{*#n6*i>#IsI`1ia1OkyTFaK z#pSBuV}|naI-iktHKP6#XNOwSX|E5d4y(#;T&Dhry`T5>O(kkLPztCueABnn&Ntp( zr_%o+hsiQl%V_1F!&f4oCq_)hdqr;@-AKe9UJQD!UvRx7`n%z<6`oV96oTQp?kr~~ zNHTd5TqD}*+gvFNpAQGUl}C9MfxQZdm&2ACL2zY#d3`X+q6y>%*P93SYuTfTkWDSQ zLi<4RDAWD-d#tqJ1=nlg6F+xzm^vTJBTiQO@-S(?R{7FgwDMc$f!L88?)_$(Yzc^= zYs{dKaLD%#yo)3n)|N*h{z!B#iC73L#02WaaK`?~A`WbVEFu{`Zosre?~@ovI|{59 zZY#);)J2?rYV^f!lChLGqQUfyZQnW%so%(?XF(^k9XnFVHli)Ou57o*E(izt&0Iex z_tx}4&~INeMDTM}mkn*|>{>}{wD=8r;%4uN;sf!=pR3-8C(6FI)Jg}3cI*noIkGH* zDHN|+v@ Q(*rEB>XXKY9`zi_>-^^V^Y7syF2^|1n|g@e~$Zt0gUMA2Y7Rw6<5s z$k(u<_@Q6!NGZRo*rPURFO=XzR46r7RoJ+Ee3QFv0(_$>J_v7>W}vQ`YRpwBKnLNH z;WWd;*Lj7yg`5zM1w-r7?o_%d_(VU5Z=77GA7CO>+~G+&%fx$ zy1y-qVpR2dA)rL?n+on>JL-MbU8@#Fp&9xi{c4gC-jR%2gg%n|j8MJ^!TW(|IyhbB z)IN}gywb$j((|FbZo$tkP=S;@NNae8+pf(Lu(PxZJ6P!*2}vF5k@0@ zbf3rj7G>d5qC9HHlyTU*J<9#_|6=d0!>Zice$gd@0tPK8q9`dL9nwljW6&)qrGN<1 zX(6SO(x7xpi6EsCN_UrtgorfK?0d|KYq{R_p7ZTsaZ9OE8$$8Rh{ zO!080o5BkaEDk4Zr68XQ<;#`D*2eCIPVXZ+!3F>(H+sBKoUE>6$zWYNFSBY2*Hs=+ ze~Fd9rI&tZgYuh%GUEMrr`5?bbi2=HQ{R{G&xyt6v?2kg2cu-TV zf;gHL;!z35A8;fmTw~iqO$7&tfeiu#Tk8n%!lSc@ecRDe0EhgnmO_MOBhN6AZ%wR6jbTzPnHyitmPm^+HIGU)FYgo1vH)&2 z`)V!HZXS$q9n7=nwIrbn8rxua^|l)6)(flIlnH)|j@0xYH`{W$QsmAR8^fN_WnVAT z7j(;0Ax6LAHUt@_`^`F$;d(z>7PDc_ZmvN4){a(oz-DOy%6ZZJ&*i*ELMtKr$PR;$ z*LqC14(aOMU})lUTL*=oqXgHdmwa;L2-9h`*y>uyPSkswK8l~( zdL$Q%EWxRX0*EiMGui_%G<6E&JVKTcu4TQP2jDxSG6{3*D2Y}fT5i_*a&A93^5xMG zyqK`Q1UzeW0(fHJdiKI})_bXNK(5;#cfRknrB1=jT@8lTU^~7y72C@>fg9;Q+)#A( z;lUWycRK=0#k6;%pXeG2nKO3J(Af8!>KV^Fg|7Nj`=uQU?I_+sE~A!2v?2<67l-GY zb!Q`@8+B!;h(QGqL$>__g7?KJ2WX>=t=Mf48wsfV7)U2w(ehCo0qK>9p69EDq^h*X zfmQJHh&Z~hKk^}Cl7|s*YPP%{ex;1eJo#5&pSXX%UJovc$H#=HV#hXD763ZtYJrwg zBridb+fq@au>E&4g?8hj4_z}6pxo%#bzZYg#?@TZxPr|C%_&F}={ke&^R7pyOB1Kkh4KOfB3pZq1^fJ1E z=Y0O&0-)6(G(cwOa`&N*=4UCk#Ly@ImqBk^GM z_urQZ#^&Ass;kmIORWSS?kqt1olTY0xSY0O8p?aYgHFd|T0Oy9}^dF%B~OilDL>aj}~-TMxC%fx2QKo;)Xc>q$R%VW2p!q-VbviKA< zQQ31p70MnMGEu?*)_Q)~C&}-=0NNWP01xU>% zSQH&qD6j57=AYJ3TS9}ZRZaTNMLwPR$u4)Ww71T;v&b=a;T~~YnK(SSeKGhEGAr$a zA)8r=OKu*Oo_pSj^Jj}0bDPVYnk`*ahM{&zLdCm$l6sx5UKm5#BNS?q-YJuHk0Eiv zcY>ObmkILt=M?1gBp{+eQ36tGez*hfXHl0ITU#RPMTNzvGLEb)7UhZEIf8gxLS)ZY z)*yH>4<6ov@DA=Ts*-((a0u?et+nL3ws4pECy;fLwCI6PDx%9F2>lYT+kU!>eKlZ) zIipKet@wrCQc>k{y}%afbOTQ_ud0QmU+&$cqk2aridsI>z{Gm-B>z!tv0)qWNOGU0 zBT8Mi&YZTbg;S3Wt+SVCg zz&i<-E$bw=fH1ZH6(l|CAzOHq-xMlM%suXDn!Ph8}((C`Qc-=(_VP98z7VhUici1Jf2@g`mjPZE$+DxbZI=RdLbu( zaBRk$S<6NM=8Qs*;?ne4`^j{ck_AP$y6$GIS+rD}f6XYsSrNT0f47giqyJcbxWZsF zXG!`?p*uZifE%z^MsS|fL03k=K953;@mXtS8(cyNypaKM&2BXuU-J~F)qr5H`^LB_1XyIm%aGE`9IiQSa5& zjX}HLfuZH;!3^i8w_y6d^KMWE`XStCvKDq{RKa*P2ve`d&&ICIfO4^ZoUNO0`0m~- z{K0*5Rb(n`%y(cQJQRLRe3NHzVF3mGVE8n2?n-xWL4x282ne3eg|9gUVIIhcutPCG zRjJ5I{i@?tuUMH<(C04c7OrmcU?;`Rd%~nje9A6n8GLLF{BSZVT^61ODKD++#kxv5HUVxV*3c3R(@*J7kB^lg3)WIgoqRXnq-}CC<>!UiN-CZw0 zg5$o>EQlGWtGwZbss;dfm-Om@&FjYv6fZO8zFPD2Cd!euQdr*UKWW}p2W2YVN_D?P zrJ0`)(NNLPjxnu`9ZUYyP!cdsz!d##y-NuyXUCzEb!YHy_ro86^!p6dYgm~ZAOCtW zp98PefF|$xsOI1sQxk!oW3Z z3_2`0@<%abDL#B#I&f-BX{w~3c&sSphxOg*=Fx2vhurNGfd6#|)DXqyBmy>FeNWHM zG(4TlWiC^M69v^Cs$*MOJ06=~PL_N7ZX=ow;hVM1*0JWWMK(kU+a-m=m9Dkinx&4p z-yxs+lAyF$QvBpp-Py?Kw4JP-CK`xTwBOl3&t=IO9fg1gkF^<il(_|h@xd0#I`gzO$5#mGgLv~br3f7~ z^K-rwFh`+0Gf(ysw4xHHA}%PrqYF(TNFV;yxhYY}?&=4KS1 z^MpV#RzwS`!P+g~EeO?UB0=A^c=yasI`D7&{3miU`0 zUPj`eJrD0^?XUp zS3plC%Vb2DWX{zV{d`l`0hcom0=&p{ezsb>HVAf&6Fqje@n()ga^~6Q)lW8%yLrqg z`t`8VL2e}crfGJo*pz&9yE`dW*m7XdNROME>Uc~CmfQra~w7RJrf2UNofz?0i;k{k&GPp)WG z`X-V(dfXl{L4wXl^<#gfaSnWkXmI1!&PSaS|1-a*MrTTeTZht)GO@6Dkk_9de z;YC>VxOMoM5QHtXUOf5Gbn3NK$?~(QgxzG)JtGLQ9h?A$NKqqH(egoo`{f@MLqYEj@<=ffyYwR#jn9Z1qCvhwLx!oyc( znS)?uf4?l7?i}4$M_ZBgRgZp$#sgo_buJdK{UcEucf*^+yr?699N6GF$ zVpo=qa7vnw3>b!(0XXd!25oXYteLOwMLT~>yOCvgBN}`~yq!sT?&`yq({MHYsd~12 z#omQGObwBI8Ty&L?sv(|SLs+SK6ozT8x~>5p zg1xaqGpx%i5!bHQS+88o0jj5xfWvk}%OWi)C_U&7$z%H4P>|i%uQ+yCJ;JA|_7hSm zCb&?sbT9`@=<^=>5Ccs0=BMO}LG?A~{K6~7CURkWKvFRb73+Sz+Il?%)!|UrG@^G) zVIm|m@ubPDW&Ej5!Ck1ns$-+oe)pU6OGcG{v^1BN@KJc{p<)LGkv7>?&riMKp1LTl zdR-BeS`Ci81J?c$L$O0D+zfjBXbnRvA&}*0m5M<5om<_nu6?);WefGf;F{GlbC3+c zGqC!q9%0^BVE(M55b%jt2`=OyYw;A%q#^v2$9j7lAui-o$jrO5W4J=S_U`w!T7#?w zaLm1#_@!nJ1QHE!7XF2aOEao*+qZxMxYSe)cB|Fj#)Xj5v zg1dAZcra}5VZr!s3vW1oHBtby^En+l$8?kehub`h4b2VBuwH92_%6Si6J!-f5FTW? z;3t}Oo#M#w_)omY@jK)Vo1)LNWScvY*SKE1Xo_1zBZm$e^vpiZVQ5k z#lzf?d9P0AhRlVl@6N1x?8d-EG6Ei&fi$9VMNL!;FET30!<*v*#^(L6gd?_&wLNtm zqnjGJPjtfG;Weuqs3N&n_QEuKg7e#L98_^|@m19-onE8|Ugm(mi9c#pt_e&`T2Sx7 z!L6fs+&zpCY^T7BbKj6anKhd)(x*7hy?&NXh)zAGdt!EyUtQD}H_}`5`?cU~-!CMz z5}p#$xThl=5(4-mFGmgUZ2vR|WF4BcM>dBFAgE5+SEbI`)RaX#vHgw;=>JC zRyrFz)*c)E?Oe6TdD?PobLYViyh2}(RMZsY;qO{R(ZQdpghoi$Z2AprYaVHiC7StYTA^AXF9f3 zOS4bKjz859UV!o&c*g)*n(yLuNm;gSUd0`{2_njv6`y7wt3Rgi#~~Uno8Gu zNHgkh0i@Rys0y>$QJ%Op#Xc~KZ-3l@BVr9w(`FzQ0#2+efR@-pIuE^Gzl9WQ`-Irf z;!zd)gLhe>P`OrgwdSuDgUxGFQovE;ye|!X6*3Sr(*6=G_%FwC0a8AU#o+EbxCPGkyc9BNzlg z=4y|7?XjN3;dTaztK_+|r6dj|@Nw;3`k^W-yfwxoK2>a?SQ<3Nz~S`C-}nT8GEI5z zJxrB;V$sqLZmmf9a9EhFn(J+-?1+z=&3i*I@zfjh`+Kj!nh^}~Qb&r|T^X`bwM6}& zg0iYii&}C#!8Z0G1Rw=5#04MV;NuzLvL!x(5DQWysp>dQ6=#O|wLntH2 zQtH`V-e5x7&3Qt2_}r^jkj>zs`6mtyJwK?2S7rJD$Zew!eE*ry^3`gN*MJ4=EiR%5 zeyKU&W_`c8;VzEYIfcXZ^8p^Q?N{ESY%`D-nX=v3Dm zzkMxh`FoYLFleE*sFt*pA?G|Nd^*07fSC4Kgy-S~lRcfs5sgzHdM;YCzfTPMi;-c~ zoDV7?E`N>yqgKKKOYeZENmP)65XDi_dkSzQxIX_ zvd#w!o+#U)32WaNQpq8*XKbt@JY(%tuhRzAyeHQEMGU%&gxrTGYax+IG)*>H^G2xJ`jCq29JuN#;5{t+TMjK} zJ0wAE^-la&!_UvMkv`oAr57$1xI`12GtL;>sYpm?Xe?bR%uzx)=C=`Fo;1GMiYyhE z<57Cz{>W0`T-3Eod_ExlbohcC{+^~+F|x*)2nyoaZGJk8W}w4A4zgkn)rIO4qZA}h zojVd_dZLuzB#;3s#vOIt2(J}k6rHQwJM(lC*jBFDb)0A8g3Gkdu#oYMNN-k2nn-miFI3(7sXlB z(~BYJ*0)~(=e$$EGM~1DG;e|ibJCI*kt?m+7~RQ;+?;0jm3;F!x0R#y+taUt zSr{6nNUtsht4}%by%Z9h(7Y(eb55r4oh7&6t=n^aF>qvhAUJjhu>9x*Wcm~(eA-Kn zkWJwXd-&KM8oxih%5kGTHjEZdVZX!g1rX2NdeG}IvnBl=Wr9RrZ}nex!zs-vND-(3|*w5V0hR%d3%8FHKZ9ofV(jmE+5$*XPz3yS6=|lR)CHe&~ zAp)x7U1zkv$%C*g;6+o`3E1OeSFBUHrV)p=zQL5Yw_{3+>5v}&!>jib$;pkL+BPIK zoQIAGLNdDzR6+X@O~C8)Q!_z+<-QEniYY)CI7!EMT%vO1_8)8~LET8~1*#n*zOF=% z-cv^)xTAiSo<$0s?jhu=N<8C3t|~4I(`_itq(F;iiCzvkVjL*8o%&4V;rR7i`h{Q; z;6s`=C_DFy^^jaV9tfG__)o_NEN+;cd<8)KoHfRWj2^DW|B*MMm_C zZk{Zbd!=0r_u%0Ti3Jj%8!xi3=ihZ9UzIr3AGH6C z3olM19&MooF4>2ta}^y|{1u$}Z!yq(%*(D_}_J zYk{F`^eyVbSWQ0Q!VHrAg_x5wA&driVsl8Rn+v!Dki7!3o7!MWq6pe5)n}nGb_&G! zX3^%>tcS-s2ydT+k2Akko1*BK{35jlO7WW3|L|EM0giIb03E}eqy z5#mXzq3!K~?GfUhr@(B_2DTS)pm-nV-0)y~ti-r#nC;0N7yu8@9Ikv(Kut0IpP@XZ z!r`%2PdyXl^5x62E_Cn6fi->Wn=JOj=Ba`>_`dC+eS%%u^DW|F1n7Ml*FNuc#DVyB zgn}HCyPA3delb|UK?Z5q9fS2R0$iq=ZV1B~VQDt;@`-~#)yk1qFTF0a!&emLz|hFg9acbJ2{;0>WVgzP zF)hjTpfwpBy+QooqeG;SptX5h7_&1>i=))23Qlgam!(Di!NXCg z8Yt>lt~o4>Kas`!ZnVcTW--8Fw@|-D0Z}gC+~yIP?%*;q!itPK4)YSgp~*1|?}IOX zqZ=u`6sMVX20b)JH?kR}!d8gcCVaA~^Xn7h)C};UN@SGFTnJ<%rx?b57LsxbW~#kW z4TCi#9T4OQQU3l1BD{p*xjhpGG57w>(jwvalg&x2m0cJbd6~Kb%+lUnqkEqj2#-zb zfIwQaoJ1OD4~{X&JPmac4GR50L*s#lPF^^Vp}2&hEci!~?|U;B;Qm3Dlcy<;OM&7d zW{*RPYa2Y#wOY*s7~ypWe(3b`Ife-!;ggDApLkXlKrE*yjPt|*d4_q{XB=2?_IY2u z=5PsqLscD?lK0{=7y7-An&5qZ8L>Eq$OE6O75(~z_E`Wtd+S2lseD}e#z|5w(9k9n z$^j6NSAK+L(2@1+K@SmA!7CgLxv8<-7zmrW$XKu!eF8`14V{I0^ct|GVwu-3X25Wn zjqpVQ32ecYn#72)gyFwu>^(R)4nWI^PS17<1Dvb&!V?`|&v)0(WZEMm(FcOj!EeWc z(V=|6X<0UNs>mYZTCl>o_P3-7Vpf-mcG34hUV1+{Xl2^)ry4rIu6^^Dg<*w?+Q6=D zqffAg{mPx~&I1_!t{j71Fj@b2jNODeoJ8c2`SAjcHbq;b!N+08j`N>xI3|ZVJE7(S zZq{S>G2is8?7}{ zdew*c(}EQ~d;Q|74Ii}N(=PX#|QRU8I=Ly<{^y>z&-2^rftNY?9`3Fl^E4SLCx_4}GjEiW z0v&dQ(_wIA#B|6G5Fz)!F>mA%yro@L_#8f3BcO5|qrA$FbrG@TljsHV*B^Q9fzFdw zqH>^6>>1KIx^(?5_D7HN^Ug&o0YmxY9u(mwOiso~)9%U*FTkTrxHFuXC*$A2qN$H> zHbk8@XI%V4Aj@f=x02tR71$E3I1ql*C|w%+li0O0>%i4tz3*4B?lXS)+phh0C}70x z=5^}6v-dwd+NYln#C;a%>2*=&T?aXKdDufzl&vhCd=O8AQbFKUXjTr~y?;N>sr?`R z;20B-%RsU4G4?01CgKLL3p*hQKkCNT9 zr@~O7#c+>p_o2h?<4VtYgPlC1xq zruP;@2ILc8-XY89@2*Mi&FO`M`6vr~nP6x^xVv1;z8UF(t1q-1#^g(JkFfz)zkAnQ z3cf2h(e3s~)7w1-k8g6o6ANTOE{h@9v3nNZbff(@D)%UG<#lpJ zB`o`5zgzhLx4};tzwa7`3KIA68uSDNS0e=U+7!q4SZObUK|W$LgG`S*kZY}4)$ez5 zU3v{No%e7ClODt;=5)v+8e@TqNB2km0=<3ShYppRfO|ws2VCiYFYV0lTp9ih9wj(0 zd=Ntgvsa$Q)BE;|FD5HQeiErvx*7ZKTKnnjGDI#FAXi1W#jY~l72_iy#;;#(VWP!7 zNQk}uBg9xCuEl?+8+JvT^Zg1x*4M&PT|xJ}TV>4TL;Qd4-P;D@{CtNV1uq`&S0{o zjLe23T<33j|4scdf)3*$NHoC;I3b91_U+rMmh@me=;nLhSb+IZ^LxCGKCX-hUc2wR zCW7IOU;bt|+$1i)R5Y5kagwHjs(Y|^YlwG^f70ur&n#fk=U-j+*QQm(to*~pcO#9 z>=Xd$6cvVS6i7G)*&cQ(BRyw$50*|=WzYOO>8_s!pqXF^3*?W2El0dgxc_43FYBDa z2ytTJ<5VK_@j(`#U21sqE@xv84}h@T;_GngugD)r*FAq1Wty`T- zWFBbIA=Oq0s;%pDr{M2f2!SmW>{1whLI^3h!~b2%jp+49odMe?*{}c7>fy<7fac$X zk;T9#HhTmlG^r}~$qM0ZuSzu?FuDK1|9*x!`sgvNUJ@rO0!ulpUZ+Zcr@4&!C@@dc zwy1&ieG>Q|>-+!S`X=+uX@a(Sb~5GfCYS`IotnfX02}XIr3mjDSXQN;83`bM#@qgw zQy?X9pLe5AVHk2+G{8LQ(q-uVeqzij;IufsUfJ~)H?`ozV*leW0ssETUqUkde~iDx z$n8*u3W)DMMv75@ErTGtB@YsD^+AOONL7W0R7&kT5p3cAW!I2kxx7*e`CSis&tv_r z1mt%KZE6YVw<&ggw$%Rueix-JJ91O4z@QXMtGfJd&MuXMK3fQ&@xrJ#UOlACT_nn9 zz)iKjMcvB{x*QvQJqrbGsQ$SF>(63+CwxUIumn=!G?*+1KCd=Bo$w*Q@ltJje#^u}sYtReo2y!zoD670g;(Zb92kM+CD@vPOn4rGTQ=z0%d zBs>8L1IO#wgu$QB|2J+s->etzp>89KoqJH=I6qq6$I~z}2Sou{=-}1tgL&UJ$jNth z+qhz5lqEXiiEzm53eh-EzA%SE#&^?T-iAtaKah$|lWyL1IfL4{Q@@));#B1ATGU>g z!ESvJw%#pz@0++X9Q&7!f}Gp~RZ5j_df!0u^mJO1#zF(F98cEu2r?Cs+&6(v&<5{c zwr!71VvUNB1;^N&on~?}d6pQVo~IbyZZurh6#VkRW3F;MgF!cp;?yHr+tFJf6Q^FY{oPipDAaUZA#y5=jclW!PO+Dy;Jc+xf9Xys6CQ#b!zg~$k2t>nU!84VpT52Vt=_AS^sg^fK{iZgkfl6Be| zbU$%tZNaTEU4xQecjHUUS#hR-uM~sL8xk{L1=fxzS<$0}6T0i8`e9n@D*{0oy*wV^q@PBEl^E| zxXz+#TxJZJ+&sJPVr42$8TZQm!3s5#eH4pMzrU;lnbR$!d?+U#qTx4LELzj)1sza&BS78kLqsli`9;jgyEDlq)jH+ zw@MhmTA=v$q5-YtNJ}!WX%BWSC$MWNuY84Bi~Ft4XaD6|cGHmleZmcxwWRJfLpcOk z%Pq`WPB?b%W*#t-UHKSJ>mqK!IqV)eb6`U>iGRIh67v$~Rg`)0QpDZY2qy&z$8Jo* z@mJ90kFssJB2l~fGB<+ow?F(od2^FM0!fZGs?Yy?Z2GG@#j7qp_7Jhem*Ctt`L1&# zxg=MW=P$}aqV)=x9Qs12N20~>s(@TB8V!vdQHV6Zv&3WOQlX-zd4uu3;Y~zbZi-0( zbkGs@e>DCcLRJ#PhP);633fO14Ai058`n2Aa7Gr#MbvikCB)HtWUMhwW9fuTQMuNk z5$Q6Q%CM?#_cUA@tkkKb{IF# z#(VwAC-4WPyh|r5uk-IxX&*TV0;HC^uYjs#|B&34k2mguSa){sbg)>Wwf9hDK(-K-y%flHwVNuLgQdyBB&f2lSTsCNX>=k1{+4QTDrJ_TR~( z1!Va%`3!wvkPz}4V}pdy#3cVK%NX>IWG@pMxP{Y`@YCCN$wN!FH1 zfyCY*lVhZl$-Ma(qC}F~%zzStq*Ru_TSmherOw}Tty;h-zdSHtO9sdR8hliXFMyl4 zWuJ<}HJ8S|4f)LT4u4DU?w(324%E%q>`8Vkw|1bNC=hUhGA=tv;NYLRix|MOkC91OiKHu}8(_7+>BHi0;nZL#Hi2IW{TI65) z;6|=`$nmy6mHWX5c#$p-3ZMSr<&G-#mluxWd^qrCh-#=Q3>B;{buHI?l;fFaD1&>A zeYn?__}G8X-(lqGIvsJa1yl6xdt(u|PHV?Jf-Jo<8{8S<^$nG_IT$TeRrYjKFZ)Mn zlhJY>^|H~kJ{ZGGX3&N_qudh3B2tQ5qOI-PVzt4d!BfYd{GN%$u){~a_+_EFRqN?y zS613tD0o-vhOVUzxX7%{@Nj6C=hDG(>OoogTa8TP0_p9tKn;`xLE-dqD2&vkmKR{{0f!mCytUG;`(fucgj0E`Oxni zj(*BZ1lQ;SJE^Hr;aC%QHbeB{r8Zx8c>7Ai7!Q1RH6Pj+VkHa(fm$Dmdwo&kssEyL z!fawhxS?i&S?j@hO)bVD0(PzB2%5_oSVQsKKJF4v@<((z7rvngo+f@b)zw-$X}3ia z^M}_u6#Tmc>@CEc;LX?)?gBAY?Xrvk^vJNa7e{BM*epy+X#d^&gB6(tNhl^)v(#n% zOt{!@jCaF_N9h~h{ld=DNke@xR4%ZBvfnzM*BYL#_2emdfJ)CWWI7-g$Pq|f?^P&I z-@vAocYR5_fA4SW+$D<_!2!D{S)O6O&$Jz5x8X(^Qsz#2gec!sQHbBJ%32I@jbS=p zc+)#QYekHj`gql_XF9+s*u=z?AI;lWPlx}lUVu~#9PC|+$=9y=PB|qq`O(f%lnpbbY& z2wdWQ84p$dA&TG(ypfoY(k@LtqptO&m@_E?`8P)$vKarYNmhLW2~WjJQz_nh*^nj_ zUBWVY4euZ76CyQP_`$4=Egt^1V6!}lhKJQ{t2T?L<>{NMJ{Zp%Bq4OFE%F(oA^{xE* zi)MWP$_DXrIQenwd>mbglPRO8-oNqbWEmaKT)hbn>aWfe75&QPn%&Idsg`N5$pA^%_brnjfCqPWRra+{WyN6qu=ti84 zxtl;~NywXP$(!%->BU*pvn~iWxi0WcaF^46d!rI12=ZkifM-0g4_25U5Kd_+6I~dD zs+*uJg~cocN`Qo|hh|W4oj}-dx2#n;QtQ>oJ3BgX((Awu{Oe1-ga{q`%AQ=N&T0aO zYcU_uxaJhJI~(nv0S`32(s*3soPC0-KHoaN~Xnkc?Ez>>>w!Qt(F%4hbDmz zD+q^3S6d{Blwd5FF(eJlv*O+iElBQO0WyYNyE-KXWL)wZql4vyKa(h$4mfR|YYQt>k2{Z z!UNNEA-7GDJot}b!iRs)gJ8^ns#d)1r*XhYYY-D|{E$J!Y29@|} zYGJ7R?LNv7)Echxs`oO>Y%~t@x$nhx96!AE9zTnBRu-AS2#yM=qkc{;T(QWJPaIOk8_+_OO4X`fRPMtUUy_{Yj-um$Ex$X`BUFiwBZ124_$FE?2RK=uN zWTCeX4-}c)8TOm!*mSzKhxfIafiC*ThWFzEE?mpDpHu#(^P|;qOs$-jXwPu7Q`KeG zZMJyvX##jAHCbZfr>*}cL!&B>vW4`n(?E4!Lid}fW4`G@;R0;yBXET=?!eeb;f~QL z2iL+U=rC5@v#q~(xH|3G(#k<(f>xEyQ|2#l+EH~1RmrJmcF8nE2>NQjTp zk5U%?9!p{5jyUxhJWD|+L!d@7OnPZ12>#cYcy3E%pVV7u8r;^6xRi?8@=fq%BU?}X zlOw>xza30Bd-im>Q?14{99Iskd%Z@v@i}En+%h5JsH+eW@qKxz{;v$M2Zw_6c5uRl zcg!zXMs;w~^&r4BJcv^sSL0UTX<5K=cnxc`QOPT*MG85@+Y62P{2 zEK|&F>4G$`E#R!@>m%v2KoO9MZig>+uYaAlMEnJKR+X8GK~>sSL84Fr7V3Y#^5-HI z@yttZoORpXUdixNwB|C1&HwcUEs~W=PfFiDULB_~$fA>oFNbv_9c7>|;n9UWn@dWt z>P~+@?ccK}*w?7a!n1^U2@ldATN9+(qaGqI{8jf8dtd6Q%1C%*D>AF1yetpOiDBE} z|NisN_-{T1I#7}fV_UmBG=S6Gj8n$z4@@>Y;{l!0Avu^FX>*zNpF&4S@aEIy7lt91 zLdRKv_{>loh1GENY~){wXLd^ziJ%dzUY$RTbf?zJ`{DdN9v8n{6eDYv(NkF9sjHtWbnae405#jkivre*(8FaJ?5 ze=^Mf5$fgtOTv9g#zP+p+<%~+Roq`48~^mzBWW+yC_Q0()0q@7s`FenjS0L&vbn8+ zg2xks`djCI*(h)Vz((z6WPdX((1tZI7EZhra+t3C>OBWd&;d(eh%^1h=;OsF$dxQl zNzgKZO*tohRp%et0au;`OeuL+-FH>jzs%-vvyhZB7&GP8jw3j@cc4=y_$qJVKfUC5 za?q*r*RmRvtBd~hbl|5Y|66}rk^qgUP+2HYgNckkgKRmC`Tvhy82%S`QFl*d%O%TO z(cneZJQSJUH{Gn|t?>o?s$8%v$M&mq{*&(*=|&i9uScicC7+K8EB%7JD0eVS&a(gI z(JN|I2LHHv*z?14Mleq11))8;tLmo@vLM&#D;TF~@^b`}%~*!}h(@5XjD+PruQ8J> zOoKe=g`+t@#}@_GrTXwBkhnASc6NU3lPnu&(lRb_!eyQ}M8IXof|##R9! zpCnxrUvHhd{M1=h8zu$YnuKSg?E&+q0N^`R@|>Ui3OPQigV%OTQioU3HG^chV&iRv z!pWW0+KqHYf0tO8=1V!m-$N_97}2Z2{2{+a(8g_iePJIZ%X@X~8QC!#h-3;0!|^9|EJTVi(Smn4OBn znAt=C7D}F-{0et(t|!kv+o1KzMdQ7!DKWcVZt#)GFy6(G6cGkJyH@F`YeX)Z2Ayfn zZ4Gc?0bsq&grI5OCNmZK>P#5CHHY;n&qtuK`>Bfd49T6W&>9U!We4oKpm8Nl}E6`NoG{RS8sMP|%Z6ws{| z8?nP#3T^(=QG{FfRVi?2kjsUgb&-hg!ij-QJMU+}(a|Yja(A!Xvwx}cH49VYX2s6W zowl0>8+~>Fpj^5ZqzUhV>j_+-GNwXy=*O zypIa`#-PW}R(p!ra@y#QF+~`{rK=_#ERzonHYs1uvTh9j9`J5dXJ{czUr8S3k}aKk z>-EE|0Ap)3A0cY_`WXLdF(552aPE<9M!ou|PJry8U|ur5oi8VM!eh6-o2J06HRWa> zO`mN7-k$Sr>!(KMX_e9~NS!fvY8|cXFVtQeD>%kQn;S zRv|j*L}E|=1B_xQBpnOn#Fm~$`*+9yp!{<;h}Av zpAy;G>95$*5dwroU-8REU*EW&z-{0%uUS^Y+abxg@*0>c?@-nwjC^>R;M-;Ml6~Uf zaG&>UUyR$?9F_aPYzm{b>lqd68BSdLlp@z=tUaO-iiI8EmryEr8o*pFMov^5-;?=@ z!<@4cT)+`iL+S~DR03`=$$DmG888yk%F>m^g<&$U|0iF$XjuV+-IisckpYU}tB z2ITte&i3q5aNH)nC_pE01l{Ea6zl?+jBkjr1EuB4=r*$VB=eJbG@ieY>@vXE~*(zFn$)sPZB%KB1QjLx$*4by)@%Y8TdS<9Nynak^@e4)KWf)nU zB{$AjUmM=p8eWl4RB;_8^HVbqW7i22Tl=Jv3^U{&ayyISJHcIg4uea62AeN$@))@; z2Uxq(=>ZN;wIC;?5p&xZqi$XC-?O1L#cO|}Rt!+7{8A>qG$Uma9J`cg)Br=>Wf`JB zoNIVuFZ&B5RIPr8t(d&Ef{9a|`P)(<4-Zf#Gbn_S@9ZEbgDqf%n`r2*9WZbg?a=6Q z>acCx#sGQFdVt#s3z5j2#QHalib%j^OGJ{L{K`)VtydYcY}cHdY-58#O;tG6y58vG z_;zyHcfX1IQo_lklba*^&(utDSq0c57B%>Bw)YGmEvr62fpM{#*!u;%LlKxFjx6`w9uVCZ!`fJN?p(QP+)6=862jG+8A++s?qFA%M7i zF?DOVkfM4%-Zm*B;fWAS8N5pw;W@9@dTiFdbPi;@11OA12#3U0-2wWL!d?BPV!{QZ za;jKE5FkfNC?H+XCs%CsS6GTGk?bwhR7&wD4X!{Yy(nr&X>MCW92Np_VJ=v&|$IF5&IA5VuPjBi=yePE(7F#xhfO$qZ2m4nL2HX z^v_Da(9FsLmeykQrQCGWQe`8A*-gnqYIXXZNp>z8WxbpkNk(*3WWDw2S@i~=g?fW+ zZiFa4<90Eq1@Pi@H$`9ja9PmXRlS)9B#3X?-e?A2T6nZ(RF@&h(dgv-@R~gR_Xo3J zYI@-AyiSgq*m>>o$9mHy7Da@&K-#8&eRKmab+90g))Jv5WjL{Jmi0}@K-l=C$D)A8 zRag4;i}WIQ-3?T~7$^@F&TU0IxJGpefl;jzZmyj#+vqAPFI;`Hy-~5VVRTo{kbX$aqs8v9Vsi+_==dFUTbM=XhWVxw|zH4pQuhnt7Wp7u%U88xp6 z7xZHq5MmO8lK7X=PetA^iz-<(2TD%?P5S4$v2mUF#2y8fzhz_k^=Ck9k29KTSNGYf zZNEPsG6rL_S43iZ|?ydj3;CYd#$QZ%Rp;dl;sh9bkH0Ql#sOI^6Ti4Fwoz#ASOw`KkYh%e3 zlJ`$AVhDz-V&0_g04ivh(TxyUoD#qS@O@gN$rI$J&EP&%;zYF*K>xg_8?iPy>r@p> zW}qFJaj_=N1VHInT2m8+Rn3|{hnd~8mH_TkE%-{!h@qQmu_o7JIm3Wr!>_rpjv3Bk}B9;G^#O%$@=WhX(n?<^$R{M<2LWCOSNyk?^Ra;CiND#?eQsw%Tm{j}whwD5rdxpK|48|B?6mt8VtahwOI^}DS! z=qF3TBA;*=~JsEwc?#xWC@^Qf+ z8EqBcT0(YE3)A~_Gn0>gdG1Jc+NuSr;id)pWz=6PHJx~Byz&)FuL!8qnP(A_5eZkx zN9Wh0{!t&MSj|nr6SR-VB4z6)*>WxYZ_w>w)X(i;$qzOeoru}8hq-Q6QZFOCt~b3s zT1J}`71miN$TD8pF+ML4qwcR>6_nSkapXE7JHZiZ)(-7NiQGV$fTnF3ZQ379`lqQt;$lqD-Y{+ z8PsQeqS9=5Smi5n>>_Qw-MK;_u6WUNOZ77H&<%x}cWm5!Z0$lEscZnVj(uC4RNk1c z*lHqb)eTTgY{VX`)g@FpiK<8WP*j^X&(^r!pr0%tRZei z4;=3!!vwTHZ605?Vv!C~%skx3GJh;q(>zhlXtUkv>UW!`F1lRV!jtj$OS;&^PO`QZ z=>P1Pexx;E=*OiN0x^RK@^a#Ghh;|}Yv2{0c>GvHtXeZ$=N$Oo{X|{`?!&C+jAB0# zDo9heWOWAf=v~Sz(oY-+P-|m0@D@2yq?D0dh$Fj9W@cMfl~R?aTQZ-_lkO|kl1K9E zzjspwT&bYCS;O)N@^Y>oXRH|ve9)0)H*`H9bdS;3Xvc6y`2p8};fuY{mfR<&IJnZj zW|1s=;+yW})Tckb(;1V3Z@?(~HthlO4GzL+YgLNDdfW44xK8VB&+YQg0cZFxC;BG6 z+zYxPPr(h*BfF-eT61ohJ^o;kJZ0qgJ@pVQD)-eZtkK*#9;|8kd z`eCXX(~^N~hZg}s!*gC+fFC0Nbk>IJ*jIm^Cg!lRSUvO;cy7}K-kN+E*e8e3=bDE% zrc^wfUATF_IYMmYPfx1CTSIn#P+sIHiNeiVU*387R89VZ!p$Yu{Q)Vbz#tEg(#;nP z61mUaCzlhm4Snzas;1ePyX$6z%(uh*yX`^A zw2<2%Z((wjosXei5%E5@l6=LVD|WpjC7EikjL-aKHZSqFb~Wc}B<4$AP^sKyjS;Xv z{carWU(Cp(395=L5$D)PZe4fDI5BUL;ON!@;1?a) z+Bwj-#as+A*!V3K|MpmL!U})>>>u8?FZSjbrJA?!HM^NvNs6}6)U8efI1<%l+lx+5 zwJRK@)sOSY`!gkE+m;cY+cY|OjP*eEp(ZMY+P!p_A?7V8fD{GAP6dPzM>c$R>{(-S zOf1UICLY_l5~+4zxVh2S;YU_y+}n$e<%toRUs;3nv~%u1-A`}qZeu?c3gJQ+LXgYV zYM8BYaxC{gJ`1=JqiX&Lb~9J0baZxb*8S%ty^iI^xL)1iL|%+FO3t`S0QI$>_#|Hv5kH=vUC7oDP_O- z&RX-&f#mM}M%9syeH=px{TWqcq4P`99d6lD;f{54;@e-u_gQGr_e$j{g*c|q0<_(v zcBrO2Uzk7uDOmv5=i@4mbOGtVy4i@9T4njy^ru$|RIn~Tmp!DRGEdWFIiPdM_Uak^ zYHB&^d18qW*~HcPrykKLr^2C8)+G2-``kF#cQ+SB%AAmZq(zrd$GexNZMp1yQ#y2# z7UkB8Dz;|Ms3%9{RMoM^5GgOoM@HN%eBQ-1s#cuUsXl6&E8l>k1%0K+#hUwfB_em;9qlpEqgnR3E>PNlu=lpw0-< ztkc4{n&4=jBc2`w3-`DXWOL3>K*J8Z5yP}&Jf}RpXN{Bl%BB@ClAuQ6IPZpNjsxT4 z8YgD;Lh9r<gScpiFa@?v~at$I{+9lxI?* zou<*&P_Ms|M^2V|d3350;0jYrr%%f<4*5|tsK*?g;BMk7KB`h%{l`24KIZP2V41hm zqpF@xGkLz+3m3`>rYID-)l`kaZB|z{S@aK4^v;2l-YAANOd(UYu_5HcQjVM)1b65P zn{@WCsO}poQ^maGbJVMLxSh}ada^W)Jj16-{<&=^)&6REbhT-%|5GU(kDEy?z5Vtz z=T6X0@+(_dlCfrig6*HCIGkmPE=D@-#Uw8x(t7TP695|zyFWH8*EmqPLLsnXbi9+L zEXYI|uJK!1cwXVEoCB9xHEKR&R8Q)KYnhxMap_kVT>lhgBpud7DscGAvh?NAH#VDV zr5CddN{0ZSWy{LuF~Uf43#gy4&@Gk82rVGr*aG;IvVn`j293PJVWO?BhMB02TxV{2 z^+-$H+x_vtjut+j{82c_S+N6(Q(JKX;nm4?!$mCf#}fCSa5SvRHsV~x zf*g*jsb&3qM`SyI%+uC{$M!^Y)FUnWk5%+HGph2s10@Oo*pxnYXecGGz z*f-P~-K=8tb6(>7|7h>a!>L}sx1)(Vm5dc3Q=v#`lev&t!nO?|^O#iFW=7Ok~W#cyB?>$=X5&P^}F6b-ap>!_w|>{wSD&Ud8W12z1F(# zF*cR+ozhXey`19UC4lwb*YETBbUATz$3@uT^AMsK=ig-Ux@Kb8#BJ5Ar~wW|)|-j$ zmLPA5p~!O}%^Fp5ZUp?cXFC~9X`|8j-B?H87uuFL!QoAz9H+g0nP)luK&2|XC0Y|; z#RT5*9A<#CW)H$qbq?exZ%&QI*YTfhyz1R74$Y`xQ1&vUGHhn=<~O1+Z5WWQ9v+lI?@mYCh&BD~-WL&9x0%c)HMY;4O0@K}{Q-hSP?xKI!^MO0l=)y@h-j`-EvU zb6#)Vqd#tmNwjuX%KPj$;MNo^U{3smgmA8tJ_CCzS#^rn9_X#~K>esTCAm&C3GrgF z>a&(<$-D2IN#(e6k*nSD$Kup(NN zYq?$K)C!#ad)XG>X61%(&p}w(&ohTz3rewzcP?8JM8YG%dWXpsob+zfdB^Y$>P>s# zj|&svvpR#!q-M+-#1$#6#%U&Kc`D$`lZ_m~sX)B#C{K}of><^;hBtiL*5Hy~V1hZ` zYeW7WyYewWx*dhg2t|szu%9zp2<(WIlDUaC3hknUAfTu^a+BM9s*1LYtdSkoChF_} z(5oo7R7Bf^?|T3&&iRvRxV|A|uda-qDD>dAJ0B=12>@f$zZvmvwa{ml_lDwVYmGuLWnAQ*+Q%G7XuhI_fNmS;|x)|Fu4UhN-v zQmlYlmzUiPB#sn9%)`l2LSVtE%RmfYi3=3c+9_I#Xb#<9K|wAevZLTo8pJXGsiZSY z!;jvE-%w8Kpv!c(I>7dMB7H_ZHrEgIh!zHXvgOf6D8TWfqZZxS>0L+bL*$j514w z=`5b5eeS^dN-nh7%4=!BCKxrJ$x-c;aO|yv4l)GmoH21sUf$4d>fS~B`sdKTo z)8oi?4*AI4+B5KjAF}?5<=j(_vSF1)h3LA`?jpjz7}Z^b*}qIUx_w^2ww9*HJC2bzVOH0Jcw5-gWV;~qB{S;t${tSJ=eQWbm!OCM z$q7uhp`Pqr8Ow+P+k82fFcbd(xHk@Xvk)A^H@fpB4x#9)C|u=TL?l~S>*G+ zVdIRf?Q6?F;O>$5arZ0`?h{+IlZ0E|r2TmZgfcSxc2kO_GmNna`AYXp_hSM_?#8o= zHme@ei9Mq!nc9$0ZT8ASH$wB-8~?#s0CH{CV4TA2(r^x6fiGR$Qjm=$@8WG7Tds7G zjGXV@QC58&)v&!g#ED;?s`v~{PZe9e`g&hKP^9{^7eO2NbS)&VMN*`7ZSRz^d{4Xl zj$%_L|4mn!2ZS?xWbNY4h>+jjowiny4>=2T&XJc8o3U_ez~l?EBBk?aHFy3s&Y%LC znVX?^TO%~8CKQ4Z$PA(#39qP>F$pt~b$|F|!#L;dTfj&5i4#O9?NNKgaAZTLiB;-q zibfJOsw_8((MAUm5KNAyvgy*k94Cg~msVoh_N)YZW4G|!VQyr05y?dYiY9q4Ldz(( zO`+t?Etz$$w+MOyZvY2~~*)FQE+LsR_XZBUg` zBws!si%b0wE?T$`>poImM?XWO z@L3ll@9I~@CD=uDDX4|n=Og>UvBL903gkmhEkGOr5N}$?HRFIEO$y4pQLW{%*eisr zxvd_U<^y>b43TJ;FjmAp4!A`3BjVes!x(Hw1e>1iH;g~AV?O*U@L?9?TCygtj+ zTg240nh8n=nZTl|#!q!!yZaGlo)_fs3u*|js4GD}(s^qVP8 zPo0)-*d{XZ>A=}p+f@!a1Acr&jZg$nxiqocjJ)Pgi z!Z9ro(`#yTQo_xasRROW_FIwqL6f(7dBSof`8|c0{rI-;!_G@EcALlqerK0Tcm!3e zhFg_iT;&+J8M^o`5Kfv}oGlj~a7>O;E9ldSpmlx{BJXg5ERLAvFpGY5!?8`MK%kfB z`nm*-_fB#1XKI$HO3SuK<_0f1a%FykmC8U89l}w*b=!|;Lz?oQYsZfNV(qhg@w^;}m531Q{$KuyyP5a&JUi@UC;(+`X(qi94 zIL)z~dEBjli%B%Q%+jE9LS$W_w>u9fDm7)mA1+8i=TYDpoX62aMa&kiK^W}s&IPa8 z2gW_cl26&B&46fzSP&OsPlW1=N-g4shtmgbl|)mTx**j5X4){_Gzog=9uKCZf&cSJ z(TN}T5JTu`@D>l!Y9`SulehC%QUSw0uuSUJlB7+65|-pRggfIc!D!WJcBk@2X{wf= zvPwxH7wlEve)i&FniAkv*NxADbV7|yJJ5}ia4+T3-jB@~o#Mg>Sy82p4C1k2xd&|4 zC!myxpDo`~BVp*=M!Z?ey1#C(zMN1%K#+rBT3@DFDDzd}77*K`_ zii`s?V5Lzyas;Sp9S!2~kCqWnCPL60I}UKCT%$P)tT|vwtiUx%Y30}Lnod^e=kjT5 zL`s55D$2ZfEJPfz37Ph~{)hG^}l( zX974JhK>~5;csWk;~XecMU%LEEfzsV9)d2inbj6|mzM@ssuP0Un#qyezgI1NHlta_NSsf}<8Htb28)i<&gD@#~W ziMI~3x8kYUmnJojQn?O=YyT#K5=1lSG716Kyvq+rC(P~&<`9n z96{wXT`r$s0|<*Xn+vAZ`a_n%g!n)h8(RsD8rOY##DOt(PD7r&Jv~~GET=8NT$*+z zqSIas3}zwW>;h!8NH!}5UHfRUPr#Ry+!W*f!(G3-833RoG^1_2HHIHNc!(~TxJ;Xv zeJq$xAj@~4c$F6QXhp=^q@eDYZ2W6pEdlGgQK>#GL*&(sM;a7k?3k&AxB|)gH~{$N zQ<(CgQA1DO?zM>h-ln>QL)?wmJnds~+odcgI0^3IY^wZ>`$jngu~v~+wT|$-e+!v& z!w7|}-86yVSWozm3VCc@*yTJS(wa4#DVnM(MkbrfwRY6M{kF3OuTB#geB&Yu&KljY z`M|=Pb1&qJ{rw|6;oVlMmZ)F0O&nJkHRznX+WN(8F6az?1C zL*+H(cDEQQ$-+q}G$3l!hV9*LMsOW1ok+MBVWLl7XG4gy;QK=`jnz#7ys8@Ni|1g6 zgDPJ`=l=hMCBG-6cN$@Yz-?+8W&nfU89#I zCg&`)^#X)-Y_{(!>#Jxz_jg=uXnL-J9SI0D&dV%X&>8b0Q=#C=_L8}Lz9(}0drMK* z6s421Xp1yoEV-umi$??siJ1*wAEXDS-g}a_i_L4y)-!j2esOC zjOGcTz|np;-m<4~38*g$wvk;VX?0J4`8)yGK2PQ+On!1ys!HexcvGAfXOnqh=Fl4| zB`kLbb>s^EsW`Lo8SpA7BpNFfrPV0<#HY^Spd9F*Rn$OYrQz8rPA1>xQB@i@1XCRP zZf=iW2$7k)B&$NlzK{r;|fsC-R@ z1DZ9=8{VWsUm1B+4C;hhQ0(mGrhZ_Qi!d?=bov$@(%9li6CL5s?NIWSio5sXBmKe@ z;-B`Q>%!c+o<5fRBww)Ux1kU1e?9$2R&)z1sx8a74&4C-T4R*X$25^UbHzwU0@ZIc zKd7Fqp3uF@6NINNf_gv!KK0h2pirbTGs38kKl&KzC&N4KDRD-CM>WrFvx`2q8TwIx zQGksV+HiKT8t*c#oo;%5$#x(VMxhC#IOFi6?dp3m2g)N|f##AJpIS2X<6!T2(fsb> zr@>S2H#Hgt4MlyF#prrKj*M!6BD__Vgo*zsv}|g6Ubq!>JB~w&tH!8Aba@3t8ygEF zy*zGA3_m`8Vo46#9G3QAU?w^v?!g$Se9jJMCCz^Tp+DvU`onclW6oMJ=tn9SDUcfs zS=)w6-!hM&!oee;bwx7}PBNsL3sU`CVhhaKB3zracl$M`?4g zAEUFXf1Mw@{jeYFpG?2^>3>py8w&z>K9>YFa(eL2eMD(-pgtOWiO-pj;1^l^qwDEZ>dgzZZaSvDk)cHZOp_RW8Um2`>#{NWxepUv!NgertV^ zLITSc?1#!(pY(KVA;)zn0VTV*$E!`UFy~47ZBf8HuZ>9*0XbuWx@4dsn@S;H`O!G#gR#4h}d*ZUHWp!VX zu`gofUQM*~(Q}_;J-tQp-nT27mzCjS}^!uVl?Xjpb*JDAw&4g~~K_i5Wov*HBnLnYdkcZkW>sVh)u}F$%{-=KW zKQ@m+U~tdz3_Df^wV=%zyHKO&&UF^)lDxb4B0~JO-r5ujW6ao_;jFIm>mFs=vHi=9 zta0X;fwC!AZ|H6n9bB;R_^hhPdZOh4>OPZLFV$4nc>Uhrj@I`trf0-I-@b3>`lB%e z^!Kj7;Zfc|*ERgxwJQw#8EMb5kMhFgd}MS=GWfSvNA-2`@13et-J}oy$;VnCf*}>Y*pzm#=xS2ki0pCAl=MSO1*JW^wCVR_5;D%G?%?VlbXe8Q2x~L}` zkIK4!*nv3Py4OiH*=0!~E~=mJ9OXPX`6Wr+!BaQ?H(XY=ALti>$S-~cFRC=M=eTHM zKrr`YuXj)rFd+D?w@4prN_~Ene$>>MDWR%XXGP+aN{@7t(G7t;zq+dLuyFtMF2pNV zeY?$mSP^EnX7Av_Hw6rd%}`mP;m_M0?8BD-+K2ziK4f%B6IqYe5$oeCSfH4*43>i4>w(26MB1{pRQzH zrwgg?%(kgsx5Irdr7q5eL%+~;@hT6#qIV{J>LOTO9nfC1cOGbLy{-I`5ZGFF&gC%J zZSlcn^Wd#b8RFK}cleMg(;p0|E&`W9T$Artm!YDd_SGwGpWxtN2AgV2pO%)E@-5H1 zi#Ox1#ie$Hz`?69t6iFPIl6Wo%fo(#+f08T7h>%3)O0<3e4^bdhaVtzcY2whPw1)x z*l1dNtO}5aq|8s(%hDYsfptoZZg$}Dq+pLH-B_e2zY^*~yuW^fuH?3D+ZJ%dFcJ?*+EHrEGv#Qf7z8?vL2O2(t-($Wg>7 z5llZk7D;v}?&cw-`&L?bSM_aW|CcA4ZhqI-=LL(najl2rnPOHwQbT`@Q?l$~AXXS< zIlEFI?#Q=nQR{c?GQ2l$TJ5}_ZBg&|%!qh&gG2EMHqV&&TH>3954ay!repVGujsquSFcNj6X#x3Jt_vU5;0Nh|l++&{-gN8h@x+3F6dcl>dQ)sT~6 zEzG8`^xtN)24<6bPE-kTV%((XDZp&il^m(dC2TM!bX~3g_aQQ-1dTRwCJx8=E@nB6S-9=3(D zS@GmER$mY8Rf6Wv(33ECckLoMHEaF&Padv28dA%(UeVlxDBUEG&@OfyYVd9ECm=!G zpLV%*4ZR}W{#ni1nvVfvwT+BS6_K#A?(b5){ODRvlbq6OMD1zh0UDw#ol_nxo>nI~ zt78beem#T2E)~5_xNLOf1L4{(T;weuD?$^{w5i zTV0bh*QR*C-5>u-%$+G)Y4^_qTXmAnSY?7kiYy6L!^6Yuxh{I?3~xD%x#5Ee*xDr@ zGY`k>d*HP2Z_9+cNRd819|Om5B))CE+)`px$HC&>@u4h-oL`??%>3nx4R&Pt?4udA zP9zPL!*+Hh^-XWL4Fk@j?)k-i@CDJef@6^??K+go>|O`O2nSGd`p4Y8eo#BI!7M3@ zVDBQ`9^dq;=A*q{5c0LVHdU-xihi7RWq;UzPuAaNFX*y{YIku?9B#@Et059E%+Ahw zoDag|KXUW%xE^BqUS0Y3a5;3_`nGmX_8pCDq(L9CM?7)b^ajuYM0Zq|PsdC4V^Qi`R8uX#bZ`0N$%)Anx$% zvfgp?w|SkT4+lQ*o@IR~{(v1e*$ore-h|xN9;S!=PW;ghDj~{Tp%aA|MfW7lcI~yf zIf?C<=IkqpUtc7b+!_3jC;8VeCoU?x#A*&W2l?Tj>1;`?(v6r=^Pqmthxl{v;Bv2} zUemtxkZE;AZT{`N2R%Yv-*9H>N*?c-GP!h@q@TQLL^PS`<X>}c23i+HKAPeA0|l12a-K1 zy#cJi8_qjzdw}iy_ss$R)wJeA6m4C}7axjm<$RjT!h50^ZoZ~&7JmUSP#WD#5Tv)R(<;E! zypuYAY5t=$reP&GV6Co8rtIO@*-Fyws-g#IKi{XYF-BQrY8Yjge7{Z1P7G9I1k#&D zq+ewdY#(8atqQZ$+^nR`)a#Cg)W*V|G<1>cN?D(e8A=H}zgO9Mbh^Gg*&*t_7Tv(W zfKTr@cINNH|2|)rn7oj7s-2nIEY$TOZ0XmgD4sY!+SFi zi$1CI5z?nh45>N`s``stq(-U^wa1^5{q>=lj9jFn0o9iN#X2SLrFOIsXVf@#GK9DD z-q3|dbKVVQ_-$Q2h80l{Ei=aAJnUSPWOds{L_OOF(n*TPkN$Pr1_))SnFfBJ6kGaX ztBuRlhUjNQrVA}9gIG8`9)c4@C{`C7T$YO~nPQs5fpzr)oaAwB$5X;z?>+bSfB&CzKKIza-}=4D_S?GA9zlPyqHBlHQ$P$v$}JRJAi9;Ww3`_dhU zb9Q$;IQ5|o&ehe{hQ4U@9+XsrJoO6iXd<2>AK=&gR20Q&sht=#NEQ&9KK@3uoLG5= zr~hV*L?e*?722zREK-cYpe3jR#!UZ6gsH8mQv63Pe0uA~L7svhkk>F~&*&cR0Vr5MgVkDIZ z5r3(xq@XFPunC^n;auJ65(#rc=7Ksc=+Kt22((9ZwMFj?p3#@$pw;9?{C!fX!uZeG z##j`l75O0q64H|L%8D>6G%>m$NYH=&Jz^+7b#VGGh798D><%-`@vkC&vJ6G?w<Q*z*h zfrk1Q>Ks`)8Z()6jtMv{)P41nrTgWT4o$%S!Wz<*9w;3}1sS?W_#aU^3iK`d)8N0d z)}ebPO!8-h^8hbbR?!)p6wUpPp75tV`qRM*TnqYR*al2*7*ClV?LwAHoAKi3=E!O= zNw=E=&c)gTOrWNhgQL6cu7g%oN6HIg0yOT=-Wam1KN&)1gGXJXzb<96ic|d?Sta^! zuSo~~hnw!e?7TC4cZN7HyIlVn?IJ_hE=;`s#e8Nu0e|Q;|5bnHPn(B7BT$nb(@#FL z93vj3D-KyXx=b-dg%&*gf14j5;otA&01445wWTpR5!sPy{Uiq6u?5D6yo}soF zBKFIMWwPu4KXQ?%+v(5f8iYswh?JS*GHLnsQ@xbGCoE%#*{>LR$;t`}2wP5GURhR20pJDcslL7fvaiT>WXRfIR%r}Ft>NZog^)8whjb|B zPez)wUKoA3c>iSdl^D~d8+~~O>icH_2}aM*o;J}xw{Fm!Nj3cQZ~pZGtpZ)$Fa;F< zQ$aME+W6-TW~?F8iZCl{c?DT2u<>sPZy6H)E2?LPc4HDL12^0KjyCRq?8{1PBcK(I zbkIhIfYZta1&!2*6&~z%0Jzlw9ozg2sr>1n^Du;?j;{8*X(tA><2$5<{=H$SJOfsz zYY#cbZ{{!*5saoXB@AUeWGwlAJ`82dh-m>C$N+-6fdCtvGY&v=I8#+Ty{Es$YLS-e-N z4+m)`Cj0e2)u*Ltg}<;@=w~$2k}|S#bijr#BXl(4-?Ev0BcO~dB>LEjbnlvB27g8O zSMFCwC;qgemV-_9?>*%GGlI=fe!sM${t+(x|D~wu+nq^KGdzEKpfTy|GXf3tLIN^! zATSlHZ2#uddb&aS?MdigP6?u^nLo2Z z{vq@%BQH%~0J_vMwDI4g`(@~gl0LXH-35KS7kNr`=E5iU87|Z>ihdpTKY>L~s#ntAQi+SZ3 zht>2Y`GqkE`QL6Q12MV5q4-OdFty++Q%Oe~-;n z{3&fffk7BUjsCKkDe`pl&ty8qjlnX4KSw(vpqWa9IZGQrG0jiA|8MZ&kPpd7^H%z2 z_Vq9ObhK^u7liLeG%u|nP0tXdYybabG_OR5EE%%-D@OD3bPstl5Rv`GQu63V=#Mqu ze--Td8>HwDat#=^mLY4utR7_KEdN72$T%wf&nv-yPpu1W$NhCB2x;eV4oSBrf2$T` z$lfoh=YAhEm;QzPZAHf6$*&*yFuXb@CBOj2XvgGChu{q0iwca{S-Chm1N{gZ+idIX zX^U#wsJ@Ey&%l{19XO-Epd|#O?_6uG^6Y7(#0Cj`PI4gG#NYJ45{}0>Q zSfO&=-|fnao`8>D%ERF}br5GL>9lo|peJtp_Dh`GK?f^WdY1q=qlrO9fx85~h!1u4 zqmGa9&JhYpQCT;7Da|02`jxvIbY=%cGDdPmWq!~MFbir0G5Ul)j29GFPo3D0z9hVC zt(_d*C8&KKc6Y-$*-Fq+k11nPvl6I%&>>4^zEMBkZwn{`l!D@1CEOhxZJbbni*bQAWVr6G)`wA&4?~0jkNv%54{ABe?+Uh?jqQF3Dj5=1 z`ww`aa(+b3rCDy!tC`^-!g75p}nAgfNpr9)fTsk9qyYpO=#Md zO&YtsEjm8VjIhTu$2TupxujO%!c~^bD}@4tw=7}P(Bd-UV!O8D zM7!jgiz|z!Z7tD|Jbe-7bOx-XW4>n!IFn&!_z z7AYe6v_v0e#jj!dh%4dIF(KmHcM8<5S`)BES%NB_9gA92Dc@iB5|r!~Dvj5c2d`i1 zzkoC6M2bTY?r~1K*)o#*xfq=ULH}`Xo#i_kc6{5w!Zec^yj)Egn$(G5IoT9z_g}gm zyW`<|R)aG_(r2+YwTT@4d@KZ2`GsFIoF}Kxm5>NCNxY&UiMnb%i)9PoZpg=oWQRx} zDvFTdU!Q)7OXxdBuD?;>JkQ|goO274W4d24hzBmjy0IRoyuDI&4D}hiiIRRig4dn~CetbRPn^Zkb8pX%#udPN|6q>3d1RXz<3r>|6*mj971&sUf1aBP z`DR-g`@6&eiAz{sb$VEHytDaKx6 zJ^pgl4QK6GA=V(s#?Q8OY(_nKEO0352nx5$JMRmE?ZzP=R+XMdp)2h>dvDT=uwpkHq^4^8{g5Z z?5p%5FI(@)^5`kttK}y~-LBLwftF=-S~%9FKaM)+~WHb*Vp8p*8uyO&zw##n5`KtqPO;1gsG%kUC)77l>v(TU3Z+-R|}7r?Vfj zeIUMWcqGHbZT3Z6O}r1qpKn5|n!8YZKUb?0DL2EQFCmomR3?FQb=)pCj%^a3C!Ea} zqhUdeW!Tt;a}!QjP=CB+XY*&53hm-vvMGD4y=PKfu(2J$xkp9@6I4!%02hL#Rc868 zaXcl|zxozS?QGwB)~rTZXIT~ZE!j`%M7`trcf)$JB9*|xs>i{>8ijo{%m4W7rvy&Z zB_ElVCDsvGQMp@jB~l(6I5}9`_xjmLWgHt#?^s+#VB!SZ@O>J!WhajbJ)E^Nic35J zTE(6!mtblY$-~v8q-VPkJxX{C4NKlB4Lss}N+`6@82bmp_P>)u!u3CbbFmc~<4u<< z#1&Ri1y6#Di!E(h`3jD0qGeYx$DBC4!n6jZ35518fY%~LPR>^*l10^2EA7ygCi^i87aS6@gCETQydqkVo*m_npiI5 zKv};1TI?T)izSk*CbeF|1(=3*uHtSqjU~^H2_0()R-5H+mOWPAIa`@J zA37H*K@-L7(HKQa4=cgO=Knpu8H-X~KOPdR)@hF!HL3HcF|T~P_QAETXD!!PY_ItK zVtk~%a-{tMrfERY!6NqUAg*_I8d53rm5D7v#=>YjiK@bESPkd0TKbZ6Q^|RJ@uS^D zj9JP0mTT)X3B%3dm9OVxu6;Vvo=oZOOskiS_NX>#d(rKO%P*?;5Z1tF><>KpA@LJ2 zcX5pzS_Qg@m`Vs0;GQHjqP4<`vL`NMb4{zYYx{QZfxhMNTce%{9vJWX$P1vX4Lr}H z$8QCXcBHsxv&_8RO_&&sd+=Wkien&e#Br z&l~5fT@EvIeJ0FN`FJW&eCM5!rP{|s`TQr2s1JQUvA(an-r`7G zoO|9<``e`Pf%LX=9YH<9cyG+%&%{@=*WUIU{HWyDBqDg#I$W+)RW4|$+$ukEpyf8q zNO?%b4juk0_g@UuWuB(2xtDaYfsgeJzo+q$(xaa)VTer^bt{t7KSh2T%o>afQmhmA z>pn`-?JZWaES1@=KHQW#_~~f8xo55DSU!QGGcu4)jCOtc<+OWCwAa@sZ8o*B4+vxJ zJ`>t}YYGUT{byEe$M;-vhtFG@Pwg>vj7zOHb&4AsBw-}d-dpQgdL^5RU<4ld4rY+_ z$k^r(0>wssLbh<-CB+X1Uw1sM+Ns-b7Ty@FBKgt1Qtw$BT9wL3r~6m)g85b6jf|3A z0_XdFp8m`({iCOuKudGD<@K)-HoqjL^AkR z%ph0jJ9M`t%JsRUWF9!&hg;RI@wB(8nFV{7JfeyCxJA8zg?XjzKUt3V zgtra1#FR>Xe9;iF_DFCIMtY2;X|wz%+; z(YZWL6MgX*z0uBeOMEC-i-p@8(caPNl37Cy@};W1w~?M2Wg)?=yajHl8C}#h;(gv2|%I%owv=45DsdWeLH4{S42=90- z609tYS?8Fl?3wxA`l{>Ph`KjBo+@!KU_e)}C@(y**;o5R|U+_w8fR$ht?1ujE%4^b1MbIpoh7<#=T2(|l$1CgG-_ z{%Wr}$M{vZ$8nd8*G^ouzy8LlIRBz_`7(|1M}?A6)*r_{b%K6+TbQ>gV{D+la+y<- zOm0ieaFKma$B6MIjDY$kC$`}+ekt{Czas52l4Yo%Zi#$#c}SdlgN8P#+L%Y&u9F0U zw9zzGxF7_lkaOY`N1w>P`T2QI-doPB2HB*{4ICQK@a4RL3&iQaz^87GJI7TVe*_j8;XO^eVH)Ez?emeEzWaH~&OxMt&tPpbcv zk^8(m<>WTigF_YJS(h;f*oLR`OQoHg$<4hdf7m)>{5Zw?u8ZoNv!1Ot&J3s;wEB%N zF+2G!rDOYfkA&&~AD%V)Zw-i9N}#c^iM&>|@yz5U`uirJBOnS^m#w z_>);{FKt(U*3z61ZtiJPD!V~OMcwXA{<6WN-7mZ>EY-vE9E%b|ve(-k8{}=mJ$kcB zO)96em~^n~#q#Zqy)TbbLhyaJeE3JlO*0=VR-X#t{hFtHXq2-@M7&#SiC}BR*|-zS zoXGQ2^+XCnc&AN#nNL{KwII@(Xc|AIgQRFrSZ2?~%m@ z4bI6)pWRT;UYzQ^#kK0j;CV`oB|)uNzH(b}hee{xw<8l_iJLd*GEo-GdfbTKesh7!0${Zd852t-|DDi8g(?F4Fw^@lX?*wcu zqw&)(XYe_f;X(!F%57H1I!sVnFujF@_$Omum55tnC{1F$8LXdtdy6&Z^Xm2|WsQ9E ziSlmcl4~Dzni$H*RA}*Ym+L4 zh4mwdoELG%5vo%q%9UUmD)_P|m)G)|r8J)tsZ^n6?}pmbDwQ-Fjd6IYVaah* z4Kf)DI*}7-3ugNhB_=xJg5V0uuI*|H@0JJ;xn?E}f=uYjs1S)uhcAh3FOe&M-F)MW zz}CWbkBj6jugDzDI&=+5u+NM!Y-fQ;$hI?rcyGL;%XE|&oz-EjBs`v5G{0*(Kiwa+E!hZU;beJr(g7rH`4(Nli<=65zo`~qo z0p;?5B>NCV>9V3EdG>;jXxje~q*IUQ8sJ%S@JVnMP?=}QV^V=T;dFQiM;2^9MA0E-PfwT0$5fZm}QzBw-=HTxvnEG#+A=H#X-Us%QBftkhF#w)GHh zn)o~u^Zv1!s_(hpTrxxoUntTxFWGsE>Q8M{{eHG~MCJr0Qix#OoXQL(u{{?SD(Ie_ zacKdHW@_0Imo@3|3;2G2=??MM*V+McMU8!opJK6cHy-hMyI^7QD9`F`SRAV!X0;Y+`af&X(s!27em{Go=NG`8Us zMT!(*xKMTa7(_L1{P9Ie$VgX$fxPk+?DBP8*M~wF#ei&e*1m{0g8n`yzpPD8-8oWZ zuYQSNyJ8$vUkg5Gda$v|1azZ2udnfoPM;t0n^zDB5(2R~Zp!81#scC3*D@!@x=Zsc zN9$^OHil(M*!AkTxMmH1?ckm-)38Egom^`B{fq(CDz2uG+mQn_@QP}Fj$vCsTB4P_ zw5if^1AFBJWkJ}j_`(!BLea&LtG(olMnPBtYnNpWJCm8V1)UlG;`j(P3g_qMV)H!d z^InX*X?|2vn95UvlZodE8c;#CDfPvn7g`c`e&v(}; zPeKyAjUX!gShs8`8ZOlU%k^Hfh(};+C`q$uw&2?_vo8(Hmi?HCe3%21-p8I*^kHmKgw{|e%SsHpq1oX@kQS4v}1IS;QJ8O|X zK%W;cdXK0rP#B$T><|vXg!!Xjs*Ss{fQxNU;HA1v-t$+GmkBRFjd${hlDhEtTZ5xuP)!D z@Lyo}%skyi3cjR5#g^&QkO1T*eM!|9YTHO?d!YJV!nYNl$CGuUgQ!s$dUFhW1y<7V z{{0kcs9p#_hQuv*Ay$6ggA<}j=L1n{-<&Tq zh+<_+Tn8A6n@3O(GAX93ggBg+#!csLeG2GgW0O7=Wv4a92Ex!1Gdz?LOB3z@iWKRg zkaz@e)b<_rwjk;Wd>O z1we4$CEeTcZX8}ntaKwpkfjFSCL@wL1S?F`| z=Ri~I;tCAwbbwsV`6nj|U>nHm!0c~SNXdN@h!gkM;A=TeQNhx@-&S^!G_P#5|3Maj zTmFDwu{n5B&y+@B!tF+fMeYX!OE)}`Uogo5kHuU~HBWl-sHY;YLGb1U7Uls^{&`#i z0p*B_11mpAg(|E`yRks1%oC`qKNn4M<33RLe9xz+4C>f8cI&`oqfM91 zLQF~B0h86eTx$v=6C>(U(J{ll?SmK zK#;#xRcuC|#Rh?JDmHA8_g4fi)z9hp&cvrw;1kKyS`00T{~7?!?-^*>B-{hS?_W`J zoh1AnH|(SMFm37qmkS7I8h6c&S*a$lkL>oHVM4lNJ7MUwg(_6rzgq*a#ZzmRZ8Qg2 z@<`4+GznNF1j9lynCoy07BcARJiXXrV2xN>U^X6y7HVk({nAH44`ryf^` zpM?*EPL<7N72!HK5j?F}l8L58TY;SkZmn|=I~N(k6sq2~@js$hLegK&^0?^bes{IPn1O$`j9XL=(<=x=4(39xA)fv);LOCFMZz zph+kx^NIvo=GZ(~f@QpT<6u((>8S)SMp%XUg6C&F*G$X7WYn1b&Es^q$J0q13 zS=V_t21V88DgqT)p|1L;gvP9(Q6sde5M|U4mcrMzNs;PQm^2@LPISz7F|mlNH5Is7*-%1NHgSuT8_tSUJU!G{@#T!*^#+mxKu9a#BuFQ^ zZ&H&@gX35@6#;M+hAuma6`7N&szX07iwF`N#1-X*m<%uT51#GaxL4;{^m+O$e^M|0 z$WwJhU<}}_`Ut8h3~tAQ+#AJ`Z`MD>HIBcdlE(O28}`uBdv{8yqb~N(+b|qf(>9xU zX~B}TW40cr_3UO&N-5iHbbjSvE4zNvu1KFLA>b5K0XZ`lkz)PRnqk#g9{BF$wRqDu!{v2#JFFHB;lG8sb z+Ra>5w!lBnnWFr@=?(WyM`q57KLE~VKB=U7#I47%BFoR449gVOMAn;&AK$fLiDoQW zKk$8#vu`*jV-oUok zsVZ)-klm=Z!IFvPu;3PUIwIKEC9FYQM34D(NiOE+-+r6)6a0^W)VG@xZlPW4&po@( z;(6ou*YmgQEI*XP5wc$ijEX4N;1`gFo-ga}R<1G3XYx#9Igp9mts#i!5qp{Sh2}igZQ<=WusbEnv{S=@1!Ru$rYgq3MP^BRr-B1(Kcn6zW5bvx2>lzHKC!nM9Pre zBdvq=(ki{Ty5?GeN2SA_2RtDoIxkyflVo%)HK)_8MKT!j^qrL{)N#XDxxeOjQ1UJc zu#TBp%$V^}!#_{W3w*b)UWS``t?TFUWZz|)1K36(BEvQsEm|sYyi}vkn7YRTur_z=^0v}8{5Dap`NmVzFEmG}Y6O^I2#dYPVkVP401@K>Z_w;m zD;G_6u|B|&=H%$>lePdPNfN|r(&O~ULmSCC%#)@Xdm`1?Q~QZcMs}uW1NE2!sY2?M zQ?W1s)=g*Vzu&F}b%5*an}RHOfGMkqtH|3z`=k%rcv=1EfvQ!@VR)?vq2p-LpkEiT z`sr6{UQSwUB>e0!0Un_ZY&0$!Sb(7GaQX|z(G`ag&DXy57BNSB%sXp_ObWr4OO@@u| zM3VS&%3fUr!m;!67BSj(@b?6zSTEZCBkZCMShfa)K$)SnrP2`|itQe!K)H*QlxU~;XFi+>$6iO2kv*AEKA#q;G({a?_ z&dL4^wzCi#@7aGW+;BhxZ?ZO?M5j>~G0MuF=GV|!z=b?gBJxOS8)uPaY_JBt+)iL| zGA5UJpg)vG7V6ypIndv}=TM_4?b|zC2{xhKOfe2>;|6UC4W(Yt+nBB^a}E`(4-!y& zo@?(#jn5|MDGja}!9m1~zPzG5XQ!crCr5HFFV#lx7FoKcAf!|bk`1r>_PmkDQ-&7k zwob|7J)EK(2yzG;_DHB0YHgJaJ|)$Yl71jBr_q$pG0zUmW$D59_zwdT=G_l zkLV(LaNiVSmH;t?h<%E5Yxq{GtSHI%F}cvvUSx8$!Ra>y0W`dO`w)G>^*HA!tR*>9 z?yPoU1oN9wG^C8)W6fYkx-@$nqz|6JdZHK&o?uuh#XNY`E)amy=X?s(#Y=dRjiu5u z7RSP530PQ9Ed=|0f8_=$JOao4TVshP!*-2cTS-22)gL^|c=V1oM%TGfCa-ziWPgfu&#RZe2-LV7D1|GXqh`7Irku;JB_i3X&- zp>3;3uu@7tlGS1_vIYrZ#+S#j0A{U;_XXX<}su{RIu2%VUse3LdeGR%J^}DrA z;U`BQxIq~!>8rez^Zd#j-mgX@7p13Yqs#iRNY5{}Nz)f8DLWN$((Z~2-Q%M(xf6k5 zX79;#lW?CSRT#SSigVYe`>ynMey}!@RO|+v3^&%~IY7ac3^?awo?9FUW%&L58@ACT zV9u(tV4$Eoa_atf1)n72ZubZJs_U}XEp6cJH8Wi^b80&)j?0P#G1qo3Q#kKmcbPU$ zLK;>~)7J|BD+Y2f9}duRLw`u5HW)X- zyVu6(L*JjhXayR&f5AaFR?)@ZMJX3~^wSvk8=&o#7Wa7OjTHA-9iDfsv>A ze}Tcz+_E*JtBZD4EyQwj;yU*!+v^IfpV9No8MIYR6g6lBtyc#PGS1A3F3s`ebtfPUmp45l+7b$9Hx6- zwu@Ev;Q@0NLo~&A&sPO8<&eYiB7|06x}SXL{9|EHTm-|Lx6&x1OQST3F&KGrfO+in zUvnB(ILtzL9^~8on5KUnSa&BQx9K!+J$G?}+gYh1sV^tg3Hp&k#%PVk2x-DgEEQS z*(Q$0tJ)QNq@8m%d1ltdQ^Wn4g+Y;%3}s>62pm1Lpl3a8W`Ier7%#j0igF{!nTfEU zT`%(mTn$R|rHXK$-i9$VKH&r~ex9Ks^)N-K%oaNP&M8zRm~~ZcigX)-bQ`X9+DrWw zvgS1@7kD)zFm5VGnZ3E#tsoR@3@eRMa@>(_QpfGCGJe34Rx*j13(A^jrM`9MVanQ%j|Fkgu(D)FfH-p zfUr=Glp*O_H{pDD1W@}|8dZD+}`tiY2wC5xo+L48cVKV-&$9-ePhG~mMf zy@yPPy|q^gUHYl7UqR1Q-p>20k8n>f7UX~baro{k(}D($sG1rHst>M z*36)BO*#QMlgWEl`Uk4WV0a&B?F<3>wqs#GIn+#4J~`a{;08B?Dd-(ULG(fgHVE|_ zj9OlCefmE=sQW>5$=ZAvn0du;Ol!MNIP7`nC)}6Jjg=))iM?;&Ddc?@`Eo2bzn13O`sMk zQ-i4vc0^yLcDYD1wtmGp$_|~U((y1gZ<*8WdK}RYdr58ZgB&%v0kXyPl}Nt5$Lc#F zH?OVZajAaH=!cdHlzb4pGzPMA9U-AaA5P5;pF^@llzKPHj zAulv8Uy$Hi(D^kAm(|x0D0K(&8r3@CT-)VPdv~nuoF1Fy;L4f{PjQf^6!QofxcUp9 z99p3a6>6#9Uvb8_Tt6zQ7a_f(^gwq#I#*An^c2QJ(wx{>{{w;^&}C8QS?>~G*9AF- zMRKFJ)Ow5LhMz_su$|ijyZpK4HF^|JlRFW>{6apIT&D>^Dy!BkGT*U`jz_xaoS_af zd3Uij_|-$gYh!Au`si&OzBj7vy-kQ-MrWGY2H@_Ux-Q5r#gF7EWnEAodO81&8ScfC zTm7pKhe{Q7x5hQ>XWp&=@(U5(@ApJQ`+s5$Tez0gZ;;A`VMA$&sUJP-lC>itld2MkzJhGB;wSG$ipjd#-joo(b-9mJ3MJC_*6lkn zx93?Gabmm=vOiY|j*4N1x8Os|9FphspoDM=>j&=}F4t59Hf4gGP}9d_Myld_j8U@0 z{mhOuV!dVrsZWZNZ;A!90mU7>Lxom{(0U)cV^GA=c)(LV5LA5jimpXmEsx&oW+;Zp@Z_gv8a+C%V9~Wl2 z1{DLec@am=N4}rIs2V=82^&WV^TFyPt?u7n&TJ_)Tik>qMaQeQVOwy-G0%Xj!@82s znijQ$46^2ovfPBlB0sz#t`}i>d28!TX2rO3K~wtoH3i6cpZD$k{wlxZN$ zW)&OSlP=jsSe@=mCyq%J!>?M$j~+F8pvDf5ycQe}4YR}#FApdM)(<~^?+OX^MH5*l zNgXn0B%%-HEmiyC%{C&l5Pw07r_D=m?0vVB<*4a~@u9{_9rd0InIH60_plE`=8Y$0 zw?$V#hFG)!B-z~kCZIbTDipd4F;b9}=LbfvZ;uBp2BL?sbHo8sr@edh>lk83r#kY__6bq_-x!cW{_ zwp<8~y!$rpF->Xt_}|z_Y48hQ7k2ervmI245c_~((qpisrMk^Ir8al`Zbf@Hob&Oc zoS3&kTq@)6ms7lVT%g1vF?;-JW3VpDHm(E(epM0_AD)aKcpMXF?)6cDxKGI#U44VPYM(0J3y;NOMriErF>FFA} z3v9x$Bj`=td-tCo027c=(UsAC9Y0bTbLVl%hP9nA6}6+EAR*r;%D7j(I%@(Y)9R~q zrY>r4%Qqbj;*q)<6ke<-CxXdh;Ucj)tSacn<>C7dfy%!woQg~G$CRC?4Do1YUPhVF z67HaZrM`~cG-A@gRVX(h#*}Y0lFO=1u}LfNm&K|xW9+Z{xW!qYyXH;q1Z5j#X`1o* zIH&%Om$^J<>rm#EPmHnP#LJn2*VfDlG3nxSeQFipf^z=*kFw{SAC6d08sfn^FKYjm zKQbbA?V#rMVN$wZ>qfz?N*~>K1>k0MfNmacgf!qnDWdAkQSz&L?;Eb9YHuNKj8xBg ziZ(I%s^uk1Vdl^YAV9X^Q)|!7qprLE1SAR)Nq2Bx@cu-RI9{AsJOkw5z-#Fw_H11vb$cPx{|d2l3oSMl(mBdz3VQ) zU;~jJibrM&45o`31~PsdD{Ha_o=|#WVOA7Nzn(Mv7?K{>LOqUa+s#19tq(GL-iB)6 zJ9=>w1Gu(uBE>dpJD5q8HBx#PK8et1d}I zV#U_AadQ>l3!P~LxZ3{tx5RrD;E`P)@5M|!wpYY2|9m(8`rF5YmQW~^3feaw*~i?D z5wMK1QsjmC=fcVJo}}t8Ut7Xm*q%8y5S*VR%*X|(rfTo!lZ(=043v%TQEX%>9oZS;5Vq4MJ=W-OP;~fvOj;)0X=L(DODxF)JUX{7Zuj81AVZoa}bW)G5+vc{4xkd#|t0WZzNP`>v;ldsJLNYTEp z+K9=T%hk;i0ci~LD2=4Ts*}UEiD6mQTdAju$=c2j>3iA#E%>2pEv=~uHISgSTO#Ly zTXNzCMXSk}mxgI9+5$WN_JLMfdcu;^5ZNv{Oq%KBh!!!fEbyX*8vY9)cGNXa z4e&CFuq3nc2>=VXuGtrNmw{RB0|lHFNINRLybZ#hN5|hKWW1s*TU>9zB>t@lAUC%@ z=+oepJ9^l&7#IA2D-p&X%PXOq7c@3@3^sfJy=UAgV!>{KKtNPt@7xTDq6b#xb7@XL zZHiNlE@KrS2&q=yu(gHlw8EAzVHe<=# zH9|#FPR)H$lL5Q`O#mnY_i&HVK)n}iCx;}@JmTc&{i*JEO1i``EQVFnH(bjQ24}+i z(E^o8fPG4DiVmj^F$bbjFQ46d*a*uf0n3=--DrRbymr<6IY+^_L@sh^treaWB~o3ojiaL^M@`;Y|CJdL4v)CvwEpFU$(t>-NSS ztW`WXe95aXqFRD6C@?6gM_J1#ZgV|=;rl`Kk7E$XC>WVB455_Y^(#cERGYVYRe|8e zf*G)Il?UfSP|5NjHe+@#suo|CmJM1myaWP_Ovsh)495?>7M$mE9!lv}BNO$uUgBXk zYa3LNZKM`ieJGv0E}|bi%$oE8ZT)>~{Ix@gBj4_|OvBXGIX$43681sqp_hRS{BY&jW?zAG@t8xT)uT%VC!qB zql}&hC6?D+kaOR6DTcE0Bh+iOLB>~!u4~ehYSZE6fl#|xb|UEu4{S`H7soe_w_8p` zqRLNdAUc@YqSiJ@obWS+Da>5n25wU{!5$n3KcA@H5)~;I@KhFGmRb2@Q|$qW-t2Ch zG22$jGL9aR%a%~DSAl9`RBtjufTiy4$0-+iEeA7J0pt~&#od&@fah_k`qAeQkQ_d! z2z42^QN=8BvWHtx6c7nM(b0L!bPYQoP#-4N7~2FP8_*Ivj%pa7NMWxjL`u7g&PO8+ zuL!4}dJ9)=EBG=B)rc6wmhtgHVoU8~+>8D_7Bxq}^?g%M0k@=B?vO}MpnBc{zVZiw zQd!EyP+U435!9>{$#=*vZNb-?_-aFeHWbC6>LS$~@3XuMh3|k!;LrG0?a+xsK?ETU z3LN4CC7&m}mCySD5SIzVt_u|6KaE|?*1(=F6Da@c&QO4NBfB$qaIR4nuep0U%b8j@ z4wgEt(suh&1w=NNER*;q0u7u|Ny|i_g0;M8udrdZFQpM`J!jqOA2+Yc$S1cXT`{s)kIIi7;k>zcKII6;*V*|f zD$kgCEM*MEM-8Rh$t{nD#vVZqmL7RD33DN$Uy!sZ08^UVosF3oE**xUefL_Q6RogU z@4cE@tXO{=jO4SYu}6OJmXVT#4Z(^OOn~x<6!slKQM+(#2oGWXSTPPS7rRGJ1J8B{ zCE>d%D;fmWI?0SqKxI!~xNePI92kz*YjF@aytofFON(?n`dl%9NmQ`RTpo&QXco!Y z+{)d#o-F*9BSknTTu*JL; zKFAL_1)tZlKGd#hT53B{b8AI1k+Fk6wH)ku6b}%jrq} zHW!zJNY8=m|MhB|Up>Ghp*0a*W1~*um3bW_Lu!NfEXNY*@6v zP^`V{n?jCto#M49c-&zV=4=x*7Gc8o(wM|kI4!rRCgO=;d!N&E&c1oQ>$$}rVY@My z{x18*!L5{}?(SQa`YIBBZMv3kqw-%Dc#GGj^MgJc<`*hgr+k?o-|gF?qE>h|H+>1J z>81$DnEI|o=uSj~nl@S5?ZjNg%~0@a^J!HWPatMjGSqYVv^_BHX+d@V<8tPy5nlvJ z&z;vzZz?s@%KDc0F@5N<0p;puItn=hos$iHga3K@RARX)K?6@=e<)+N&hC-6bU)k> zaGRsd+S7>en0R`2kIf#ZMyV^@-}1MmWJy>oNJj_-L^UcCwsIEBYa`x#l?szs_k82x ztwQ4!8xA#WBKYv%ToWWTu~Bfu01vT;Vp$8M3YuP@D@c-th;RI|AkRQ_^RmKo3IfJ? zhir-|%G=jHSU4(QhoH7a^H((bYlyG~+DbLlZ!4dt)>pQ?XV8T_O~|f2KSRg*jV~fv(eRgI;;62R3s&-ko9cg1pe*n_}ie%)pEQ2y&vclRfpha<~ol- z9N`&Ep%DBP z6mwh;_sH+g2X+?6q(ArG@s3h;oDyj{?KHJGg_d5urbRnQh~!-AR{enQg=2 ztnCgI83gg@ZjrY;-rv=@PXoX1P4AoScv+kPK_)0pa>fDjtAZg;TqKN#KWZD01_yUFijT1mBWw)iP9K&QEHuwasxlUQo+98ukqlTm=y!sm{B@_U#1#Po zg`C-qAe8No(E+}~cc|{S9Hi)j5vxDlZxLdgK7vq=?=vhfB}*t7Uec`D7jZ};f8|+k zhb5%<=eBT>WjlsE^;5mYm*^hpcj3M>0p-(Pc%dbubNY(whCcGNH|u;~I-jGF1?#&} z+_WeSw{EmQW@0#|?QzVQGqFMDFvhMWe_UqA>+ZN^dn4HSt@pMZDy>o~5PZ3-`*`)v zbmZg3jses*z~|LD=Q!nldH6sMTsuS#f8ovILOCU6%4U+1=hUre3`Lo_*4n125#~IF z#&fD5znaijG!#O-oTtPh&uX}s>#LGw?FFcNDsj&|EW-LEB9}B+ze(nj!mCt>uu+kG z5TBA{t}HK#JC1z28{e~`0^N&XQ5VmBBzMk4USKxpW5OL1yQ^`-8nut$*rMAyAu@ZG z*tO6s09?OXC8y*q^O<$cfsB<7t7=o}Rnwfv>ka|OCAp<^@APLm&Z9$jwx@dQ%&&!p(za?iJI?=YBu%UEc5W?%?*x{-HgLe^2c`P+TRxP<#&Td?J zV|H)UTFU0*lssKKo23C9Uq{n8YK5KKr3wg8S==MLggh!~m#3I@b-vfk)!rg18Y?)IJ@d{oPs;Uk zCIjUS(aTx@vx${&%nvCg$p%!9#yz8Y8i~m%=Z#bgJ~%yi(NnRW+_E+Z%Y62Ry#VJg zTK7xj2lPTQLB|w6_(1`hazkziItmB_lfDsMZSn4wym2TI-GTs7s{(G63j3i#*l&5; zQoRkhIsv#Wdc02o?wsTLI^X;wtT^afn+<>r^qe1wFd=1(0Zdj}*tzMwLfJ)2xbSRU z^1|-+vrov4V*=@LbHZ-jHfQ44fKM@Hykud^`C}N1lYHGX+Pe8gHil$9H{&4RGcyy> z9nK1=9U{j{nP%n;fPZ*|D330CIBNH9evRa~Di1-@Db={g35;TiR7&T^Jr+LAl%QkA zZq5KmZzJ#15P|z+Al9MZ9|KomY2dFWDimeS?m08du2=O^-vu3kEz^@V_$^0+Kd-G0 z0=TQKwVsXeBdgM14$eC>q+2)s)E0he85Hs!#oT?&t;-kX5nI>VXE9)E8XO(VgT5|w~AU$t);bm0!m?S2LO zx0|c&#VU_XXC({YrMjazb??BHOzyE&O-h+XK9g(PcyacoP?B+^^!WLAkgc`ERCUhz zGsi_7Q#KhVWsov*mBekXMk~8!77f4+C(#YP#u)^tLNAue>>Qh0c;OmJo$U@B=3*p{ z450H;Y8K(zy@xdLrvmHm+?cz#Dc@0KPGIf5@@iP@z6{cssToc#YLZS3(ve>02%Mngnd202bp&Dtn!5{d)Aq77y9Y} zy5ohdw-c+SFF4Cq8u`kIen$0-$8lG^!Q zR~3Zl=hY>1>uZgchu|-VUwdzLhQ9)ek)sIJN(Cf@NW*15(U=VI{i#>+v>Jmw>RQCI zw zV|342w)V8{4P=j;+Y=eU z?KkF`=IyWGxhg}EwKMN{iC0GcTPXeP2X zU5E&Gv~61y~eR4V|=;jz7pJ?#p zToUD8z;0RZOT7vR-QR)E9Oc33Z|Q7kt}`f%Dni}G;0F7T(hv%Z3jwC&6& z2M5`Ffd>zwLO;cG431~R-&m6tZda8leHv2ROK!RE?dY=YtT(dyis5_Rwv4>bSFhga z3qj%?2)Wi_<`~1(ES2cofz;1EpTn95ue$CjDB<2iC}znl2Ay&l0l6E(-@MO&gP(5U zEb6^P*ftROxV0v||A(*lj;Ff)|G-PsL4)jMhmv^{Wn@%TNK00B8Cl7m$0>@;?g}N* zqJ?Z($6Z!gMcJHk?2(Q=j^p>b-m3fa`F*~>-yiqmejL5W^?qHiJ)RdICj!%{TV>^T zjT2bj3H9g2XSd4Cz;_Phm#--RbwS-;OE9gkh~>(x8h_)<`b;}S{S;&t}hM32wTP6d7f#1Y3%ArG~@8Q>#NI+^g9gts?QI&s1x~FC9rkY){ zaWE>8zm|4h!y_^*mSO;c-rEW(Ppgybt;?C8NtD?@ zX>W?0p{a;)8i$+KJ(Xtab2{8|Y>HPE@|)M;Dgn-fXq10V9Rx>{$r5nkn7G#^xo~`LbSgQK`PiU7+sC$NRAM%QO&tFD1)X4#!H!KQ`Asqv< z&?R^2LP_Ts9{5R`0j(vG6RIqE!yQ<@byN(>@|UX^7CEFdAKg8z;b8nR=Wbey7j7D* z|4@IuCo2D?KlGPSTOQAKgi5Mjq-4%fQ}-&47TDJZiqJTsaOv!P-y+JWB44@((0bp- zLJrBp)bsO^76kr_LNb)qyE}M)$IB;L)JLI#*c8TpwYn{+1B-@bZT$5KABDe-9MRcp zI3T+Ufmpru0;)ghqe{=tULqt|QS$BtaxiI&SjrN@8I)rzd@|>r1230H$Y|pux}>f3 z5Oa@!CjGD?9l+N23nPGF1I?k4r#6*gN!cs#qwZ8X>+I<`bzwynwBcU>xNcera85jD zzZ#`y1rF?8#6xn-*B}`oy&waWhP)`Xxpv!;kPTapLdt1g#k26vkTTw&43H*!TmXm1Z%-1uL8hs5@@(Wc z+MP2ej-fgaiZ;@eClF2;D#c)6X3%7FAOi;Sl1EK-n?x$z_M~4!Up@!i-Dyd;4zaow zH=Q~2R7QbpltlF9GnzEq_i@C%Ot`aq>QUyS8H#2}_C+LwVRvo_N}lzP@c0SBCjO)9 ziysWG#PQcrD#sTfzq)q<8ZjK9QDm%R09=(1>M5?pBU7gH2z!S8M&!ZQkndTXUYOrc z!fNe7GoL*HATS+rvX4pAZh-8yI1Si^A`@V3o|SwIUHGL7Qhy@3OgOp;lVc6ESmp~S zJ2F8Q%uF{x;zHvLogw42t@$FTYK2(8zRCzc#-#5BDYTh%N1(s%q+e4&v-m+uL>@ro z4bINV{5K3+CvpW*<~Z;Bb>1|HUJ$ zO6_P0j<Z@znN7Fsv**r_L#XK7{0?U#426;C7V9**DrI9LO z9lUuZTB-Laak*!*8G+wT69H|=wBBDGfW)A^hH|>=VuHEiD)rN$ZTj-{vhUAck?T^o z8UeUu>KzCiTN8^DJC~#bmFX6D^m8USD z+L(-7tN~caU776lsF_KF8GcArN_IZ_(*m2iOlQb69pFTpaKr10rvzS627IQeSQ%*X z$b){HG}Jdkd{EFa2Mr@bUgH6U`+xKXOeW)zg<#jEZ2sEE@D?qlY!x+Ix#WG;J~VMU zu?Y<5p0F5XV}GNRVuBh5Qp;(>IdKvuEe#LjLDQVvOn7;H@zF9$X{mofjXHUtID2lzY67Yh$5C-% zse9;P0;(ceC+C{+GE+)KCre=&&(WZKdA<^ch{O#8WB3KTXB~(Bw1^iApNyBqH|%TE z^XYciLFMp(_{NcKelSCD1vO2FWO?9{(|+mI3MCy#N_|TEAIBJIGW4REUnkQ#_1DTVrlBQMw+D zCc+8@W$(bMb_=DfwZ*Q&T#C`cB`o=S9hm@5#ydjwu^p4hxbRazrQa(DXe(sv%5((n zdcYbIn>6vIcNDL#BBa6w$UgWbsY909v5Lc}UT9?`J>}uS<&LJgao*f<1Ynwm+A)q? zR`(unr<#<`c!PGS&Hag57k3Wolt8f5g7A$HMp0>HBLF{kyNTMAAP-gwdI}t&nB{mX zzte+TC^+Ce9|o~|9rD%NTrM!9Z5RB6Fa{P(|6~13u>L4mOW~fLN#8ZGHZJZ z%tnrYq#B?uo`?EaPf1hYpu8VZJMxpd-HHx39&%KJ6-1x0ve2eI=@0()hWyNnMx*P3 zmf@tnq*@Hw2Y|+}Qmt=y5YYNAA>^IQlr={Yxd~ff0A{*NcHER7yy{;QZ%k-n<@_;E zY@~?^?vT0co|fS^of=&qzaa7lO!13DKB_#|&m1FMK9Q^b~Eeh8<_uhFCW7 zblz(bEo%HJWfu4XT;5Tz9l}R%puPoyDH!Ly=k5OWmqWnm-%!h^9L7RC^yviFA`c)p z{o4@84N_1Z$UBu;_wNmHaKoWbE~%8<_1SBQ|cksAHa z$^Wx@U>LKn&hf|m`wmTb2g3~O!_(pFTKh{d6Ql#mo^QY#>r&heK-R>5hMlhgY#$4D zzGz@{-%@Q-AjC8sX(5z^h`;g+TMhq3P2%`m#{6NdH=T2Nq(4Q6*;b^na{?Y;}- z-9NCsj|(Bqpw#$Unk;7xY6(g6j(?^3Dnv&mmN=o{d8gD(Xz{WD1Gk=Pf_ddm{9Pf?0Ogu zVHU2OEPX4F8Br%#bH=PEvg$0)5kSwoS&l#P9UT*HTE&Q6=+8V!Cy0hyN`?r=8neI} zUtc*kgbh(XV$Dw@U0gkYLkc{EYzC;bw+I-7=Rd23Q znasE`$T80ou-^xrgo2ly`LNC(Wf{mq{9d6StZ*QsCr*v4f7`JO8?cSM4P>2t1D!xg z%ED-CJz-LuWRABN|GC;;!gkp#CdZ zl$G!Q{AKX6U*MMq7CnZOZ2@f-#LP_O0h=K4AI(4K#IWe(uAiGb5S8Zca}LnaSy=|> zYa}Zto0@MSwiiUh7YFk{ZAYI_3kZuT_-IAfM}l};WdVfbKzImkMO}xOD1}O%^w>4v zGp5*^=iNlAYYMi1$RxqSRJ3+7?6W3%pq1`=YuaFl7PVm_5JUNy0;Mx#DI$eEEfBV6 z@yp}+c-&NgxEI#P*ABu z{`^jTP&5_w&=I9$u}+J6Z}sHxL7<5pWUErFKYL+JqX?GLO)p_}6T{+S+fim%qO3hE zC+sU%U(=$vCPH$T1^5nHpi|ZmTF)Egumt7W92BaRNjtj(IZ++J?`O96HO)=6xx*Uj zL0_mZ>#jrO$Gry5wcPM3&m4kSL%jg>Gy}qc8;e?0KZ|L9Y}}m-w6QrOdbUPt35ZgO zp?d(+ROH@!8&SfifX-v{bl5^}S+^I4_)!iXw$C287GC@V==9@U)x@4!nam42P`1Z} zxqeSQtbFNon;#+vn2C+(!66l$aBcpFkBjhv2d0i#SMPa2*&DK!;ggrvMuxd?iMJ+? zqX)y7BTQ_c2Vs9r{`vhe9g-?kEY z;3g|qSkqg>^LTWN6))_ZqW?YoxwsTRfZJLDG>E&O}&BaLI|;_F<` zrID`tb8&?OUA)oG=*|e7C$O7OnuMDUfX3`=vB6l@EHxBugT&e-SPy)1&x1APO%EVS zmJ~1amLlw^2Oj4BorBC2`#|fpy}ORPk&TiK*$8_$PLN7(8DWPJL!56 zOTg{c6|8|5ShntanvZrtwhUN!X@|oaOe>nfNtGkwAL-1CxJuO@2*d#Vjand>vv-uCZh$M46V2WbRvf99u~oN_6$5TE$p)-B1bKt2&7yzS#rE1NEFb~ z!5kJULmCdaZ_$#=6dT9(gJ(`>#WhkiE8sn z6+Lq|qT}Iv3Yfb43#^;{>2RU@LbU=~U*)g`*c0U5fi-M23Ms&zc?_QU_seG)7`ec< z<42h``vz;$LVh8GbTDn^fqlr8D!GnbPk`&`BGW!#*WZKd|62NPAXzKW+s=cCCTKtq zLeGDg-ztHv2cc@ZPUK+{U5)0d+zxi~Ff2a7hHKcEck1oo+vm>v;GjC;;Cr|)c=oF> zb|0vJ%it}u8242}SQ2Zmwsenv@yRcU$_rwjbkKLdetsXL{Z~*eul77cL1;(?LhxdZ zOIHpP#7nFVYZ4yN7cc@{4;Y>{!}b9r#kXg*RIm@|XzF9QN!I+{_4e!NMBjZh)JR+p-6DpuWrb=I>nNit7AIj}rG zX@J_aI}6<`l`ns*2$k}@ga|Bnk_ZZtwqp!K{G0(jcKU-SnLAsSZi^7Ty{j4Z!ZO zgF70&eg(uexC6m96gp_vu=U%Y-K>w@;mpt}tG>hGK(ce)Q%tv>1l>w}_jnXL_G&Xo zmq3*F(gx=BR!VG0@Fw##k&;PXpoW}=V4 zNQ;(!85Z)L;Y%R8iyNS0)FtMN#M$~|$0jgL&_7pxf^8%9d|Vs-gB_&o`=gT9-iTiG z4qg<-b#fY0yk}t_o_CWU>cy#N7Up7BiGZJKz=t@;zl+00jLa7 z&zz?}O~2SRy!Z=B!O-gWntoUB&-PXo78~N3!R))#cS@!iG)FhKqC{O+8vVR-p{Q-n)d!9YXaeKBA8K$*8-K`0$ z*1DT_3;nzTif-t1#QIK9bnCtfA@cpZ^<1#^GErNx=t3U(`{x}1h^*@Ce?bkhBZR5a zko#Eg))R==bO!k6F97g&nN;pyd2h(r`qwz%B+Baac8+2|!86DcCn+R}Fiqv|3A_HE zFAk~L3KIdkYZC!$w3Bc(&nt&AgnCUr$323sEh`RbvLXS zYSN%fA-)lcGM?Wm2=xIk{m`Q|tb9QxepcU3){IlQ%d*crzo-_!Ou&{GQc+DeG2dLZA&(F&nZ&lr zFaN`G>hFY|?pt#yN_O*_NJl(B3G8Kks!N*je_5P`P-U`cp0v}N*s)-qzc$C@fih3$ zFa95lCH%(u9PMRVQZ(`Sa~gi0YQM6t6$S-b!KP&m8Gh-#N`3tCS@zn()>!-bIjdxZ zQDhXrGmIH|XK<}abTad?Oj90NadZF46juCnWAy-!gy(MywZPmqf6y;>!uj|H6a9*J3}fs;aaJmS5}|@>#P3e;6ag37tfV6FbGbwS~UAMKs%)w{DY(UAf~E+No2k5+OUGN}C=c_yFs?dHxQHO0Qu21KsSXCP`OLeU)AZT<-6V!Q z{*#1|*R04vOWO8q0>cA2iEs+zk4ul-_&>BMr||IvS zk#Sml_w_Rv=Qz_x5p(zC$Dkh?O|Zs?njXr~ZIXV;0h^TS-}OfRaCp8SnV9c1%QWrr zGQI$Di-r_kHOW-s1*&fgqJDKe)9kIrJ+%UiosQVLcM~oOX>~zj<<)fHnHIfuu$osZ8y&p_%l^dE$}g`54{M27QN4s9 z080*W7c2SO;nOAfvP(~F=D`}8DR057;d=VwP~1D}`fq%g3v4H4MNRfv^um&f{>Tjr zzt;r?k++72F%Jc%SiIm>w6lg+EHJs(hozYvnZ@t`(%^c77wSLXGAwS`pn@6-aO^^F zj)e=>%RJ88ff-dEW^TN^bW9p%_iq8sgzH)}K3}D}p1@8ABlt5EFtUa&&RTG-_4L1o zeYr8~h4EQpI`Cv)G4me{Q*oSd5H04ooeCy~g!ix`(rfsJvQW~%bGJ#xNYpkx-n;{2 zJ~rG8k1!54FcsTCja#r0!gXn%(MFsMhlj^$xisr((!?$u#GK@3%$r%^dR?MK9%k&X zbIQ`T{=yxR^#SgxGD0i*kUK;Ve2)3Y zxbwj0UZQIsO@XV)Cb7X{nzZ3xXzy$|J+yG~-Fa&tdJaDH^~XnY7{_fu3qQgM>i4`k z_Bw5=kFbZ_h z-al_YaulsCE>rT z#K?lO;4k3UeL25%Pi@n)?VT9IbTQpwhmS4HC}=`PSwk9La9-lk8mxRpo$IUejR4e_ zU^Im8=Ot{19ill)8}LDFtdS4bUea9yK|J|9gl92~GgCL0VbT1~_^;^q&Q5r?M|n;j zR_d4o+7ur|UmcRCDcjZ@#eaYY zu6;g%x6-7|?E6HAU7(km2HQ&P*!K<2U4`|PhfsV1J?zmguctGIoi%^P)NK%FY4uqF z%XgiQ2GWzlu*}n25YykLv!WZW<*;1;M~gl`n>q+jwj4R~$_$?TTmw{qCaV`#5P6Aw zk@Gfx?E2Z=vNzDR-@{&^vu1anhgE0&WV(d1ZMb@bUF7}@swZn)e)3+$1Ej|;a%KGh zI;2{`?w6w@KhlDhLl+q=tM7vgOz1#03hDemf+w*((H1C^HB8sa0S`q6(vbfeWj=JU zOqCS?t?EjG^HHA41GJOTd9Qz$CM-LImiFPuYhJXplNAd|_gPj$+2Bwk7{MpE$b{_STI zNz+FLF&wYUp&^vEg|QmU+x-YZM1on#T`8{qtX(-95NY%T1rN9LnJ23M!!UwmD=`9H zuX0=0TnW7OlNrF=oy`_+21#BJJq!-0ZQ8o?c~Lsb1g)oghQqO*Gdi9~?5xGk=X&!H zc0NA={=0?2Tm9dOS*_3?R_{W8CuYkSqKG|NEMzyIy%^S2#R9M-r9nsKi6CC#k~kJW za?ojo&dRV_zgx09h>6P$;F>m~a1?PYX%PcKjsoXepvLd-Tu^-0KBqCZ6oL<<{>>uh z>^gLt{t_oq_1z4RS{e{Fic@YSQ=UWK5ps(0EZ`*dO0s@X&gp zcfLnh#iHyd^x&Db<`oMn7GY6DvWW_Yo4FIEDHmS;QNbE^Kr{-oupT$5n+M()mjui} zy~&NQ`E`~|5x_xNQtti<+X+$d6)hZx5@nOiz7(yv@`#ceAPwnab#A+veS(bC* z;W10E+Xya;O0e?S|5{M}hJ*tWsZll-;eN9-)>`HHpHbD%d9M|<=iKF@CSLrFT(1js z^xn84A!W16vA=r({uv~^nLvx(he2oHA4L&=6$nlu zK))HLAD@H&Xc;8PiR*&ipJ8HCT;rSzyRE1VZ+4d`d{{6$k&%b|OHt4>y01-Z6q%<~ zv^p?z0F$z;=eA!gye2CIo*Syvwt|b0sjs-WDT;VIL1P1a|1UL=21fT+78qu{aUwSJ zffCkQP1w0H1Xd7a>@Q@uwZE*Vi$)vBtO?)d?uFhZ!iZzlzRLeo;a%u@cRqx=w*dre z(^Wpon`#nkb}}j#{l60dQSe#Z>;0ad!Leyy_1*#WoB@|)V!hn$L*U@;3AIDoaXWY# zqG9J-xBQ{TX3&G&+`d=eGqZ_l_#i{}dXE*=ob7c;Pc)H%{TLgPiT}eB(+c&^T5qjY zmj4yby!{9=9B=u)$KQ&YQCNWdM1mZcDBDZ1fznXk{B{EJc?2flVh(oqBV3`3vu(!P z7+H23fSIyrXm!b^_?XLKrwCV^K(0_*=pX&Q5%l}!kozM-VT|`UwR3^lpMkU-yr@KL zOmFo^6&)ckw1=MCUD4AF77P1?@aWlr61!86swtgH{Dg`X&rQEs9UpoAGGM>qP!#dZ z2Lm4qke|6mWKWu5na_+a$Iktco>ZOnLRztjVSL9EPa6NDkg$&M5`0+0yB-;hpiBF5 zFr2)=Ge|_A{VSqS*u$x5at3l+YyDHz0?zSGkWWst;@%yu)QRY30y6hd;0UgI_Q%cB z*amulpJnfXP}xU+YmEYu+!648B6qj!j#i+>yc^#I?ED<>2aW=0bANNyQ5yJM8#W;-( zP{y^SVeM&n&b3XcXtv6s!5cFWRzPG1y4oydGmD_ z2)J=W2oC%CQMl`rocQQXS!-e6bWx2|2e|9=K;2_XHA(VgDl1%53n1tzh6k2hq6CSS z20GxPYdFH*HA}?11%T>n)vkkUfSHpl^^d#Uq7;7l=!ihQuR|VZ8~J|KDTYOpJDMVh z1po}vdi&f<0|~w7#ihzRAsPmtnjZ@C4aR{pQHA+|$dTxC`d1eOfbVdgA3BiPu4Cv! zm%zPmJz1r?3_2@U+7z|oW+PdPjP*UJNegd3f&QWg?0Ap;ai1NIO#(|KzsUx%fhzPk zP;vry5v8-wVihqsOzc`d*R=%4?zOUr&jjHQM{~krrEk2T4`n>_K}BXg-2o5-kLT}HyMR^fduEaPgGFW` z=~>Hpg7k%dJDq5%ktplM@E~O!VGl^?ndd4)kAaG_Kt;kL>N3xK`{xy>k89B9qKqx46iEL-GO4|SzwK^7Xz$ntiNI<%(>~Xr0ZCZW6Qxx3J{P`eAb=4 z(fzNdpSl~Cb!xygD_eBaV()aBMaG1FNJS}p(ez87V`Us-cg{+=h-z4wgUK$onG?_j z=O8|B8u?DYi}7Xsg7}l%n;;&7w2D^f)kvk=Y0PYPryY8B$;!iR^-LC4KA?*^^>2eN zev<}4k^Y;%y}9}mut$efTmo|s<@kOfLA}~1TTL|)hDkLkcqiXQU%34zhRH(spTH69 z5VA}vhB81E3vRy9b%Y48Ol({gBW;*n8{7b_zZ)qay%()+&!LizwG?Q?z#@0PdaQ2_ z8-|5W8kz9A^Lw9|zk+Re6-reB-v0=$eS%3Lvsqb;|yM z0QkiOe|XF6m#OfFf9}bLdq$@ki#^~y36?{Q7i2)Z$`9`-dIVd3XuklGc{^akdZuQ3 z5BkNROUmec`)XEDjO2027Ld-gTpv4dS-=t&DX!*MWG3MaeEiOE#Q&TAOR;8})wRXyi@MzpPRM4^(nt!{EsmdT+UZP(g zanP-VGA1HUNQR}UsHaG;nbABHQT?~_0LvAMGrW7ozLkYV>^0#cbY1u(Z~Me(NEk3a z^PW3)NC6`(=swJ|Lf!}A65RCl&S@P*b_fYDCFu-CqUUDQjlH0H&Ye7L4)=JY&(f9t z{aGmRkofqVO$~=*^T!UfP=?4uH~Ln6z{mazF66_s2Rf5dIGKu~x4K;#6~fri2RK0gNeWc2%b$O*^3_YK+U z8KtoO_@#xTz{R&xlMw{98SVmizuRYC0X1k{o4kq=02Z*;($Poz7MXBtQyJ0mth-O8 z#!}imbGPr!Y8ZoMw{hg6gY2qN`DsjTv;&p2R8WOwj(u3+Tw7?lZYq(Yq~n9u8cC-ZL|cXK<7FuS32bi+}N#JE#(3Sk-qSRVRb6WF=j z-Pg3?d%oe2PUH&vt>&u0RNUNi_qv?~Xo5W1`$q&lfck?B&86cY*&PW-H!!j!qVim+mcb$9dItW@x0{~`tPI~3 zP=nc)=46;8yt|yQq6%=11~7GVX=r7+ld`%(TumsSRw2p<)fEQnE_Jq*dk?>xZO zAnpLac3g_BVP3>K!dsL>Q=PgMtJmaPEx(l&AQc4E*~jO<)9$yGimkNAt{5-9A}+l` zea3U{wfV8xj&b6*1=s4o$ANzn&SmfG+zVYoSgY(grO3bK7S$0Jv>6 zUK)-$O%LcM^WR`tEa7uY=H)u@#r{k}W(u2_=iMAQ+LY<_4*cp0X;sa9sX3vSW9Xw% z`IX}8eVMn=)W8fck-|H2+^D#@KW@UO7(k)=MyLQ*`{f}T4Fz;L6MYr@NJ6mVl()B| zMxIh9pTWm6O0?Vg#ao86oqi$QX5_Rv_O2X%>Ll!53&Pd>aV4STsZ#0pEqjyhPP6$` zlkN`WqIm6p>Qd93#vH<%pH1I1DivPoH=tcu5?CeU71q;S+yc5h^Iw@ut@Myq%x}u9 z202rzRN`N<_Pb>++5=bVy985f@TK#J>Y;j)l$onsrGn;A!qPMFeaEI)(>ZY4B$N@= zpl5scZQ7+_#Y~VUPuj=^zB22PFrUjwP#ceOk1`0Jx;=u1I$C@JYkZ!nFWy!^Mw$;I z6&`t2usZw+P$|T@lvZtpD?W88!U6976V8)TZhmz$0LC?T&D&Nf%j+t&=V5Byf!b|f zB&$i9udvgT{Qm*_xD3SrDQ(9`4~w_TiDKRDdrYP=1pH+#;>}uaq2jf zc5%ut+L|w*-u@9_cOUs>S*=LR#k-crvg!4lC%Ed8Ykcc%`ENg5s1#4LR~s`KvECNY zp80*onl_Jd?nU}-fpD1bEJiB03 zqD@PapCZCeK6OCj$E`Ev^JQ6#>F)s*Hl@Bk_McS7l|vhj;Dy6%k#p=U&f z)+cj+mD6cSN5IEbPIq;3l{TqlA%f3ZIt&95gG%W&&`>^tU&&g{pZUvS%m)UFpB9!) zxbS5bi1;ivtx}t2YVg#W5twoh_9$=f3v}-lQnY@@0!Im?o+o6&9>uL9kgHGo3)`m+ z0{htZRNn{eOQoJ+Z6)QI61+q*luXMz*_0C0=dxoj3iMB&%PQYl%x}+^)nnJ6K&4Nn z6pr1sQ3nIK8KtVK_eDN|!ZfDuWZ4LrLho31PUyPY+I7Vmz=DQH6p7kE(z9Ir zXmtN1tqbOKTaYeDhp}6CvY@)%I`kPXjT}nbg1>DF7DS-8F8u1T!GV+W!EXL-CB1j0 zy?3E&{a#aaX9ePHU2vf0joMK zU3FlOet1u@b~rjpzKHR<@Xr`tsnuybttrit;^w~Myp+>bHacyvI%CjqTG!oUm)$*n zuP?V+YTZ1Nx}K`9POBTp_2LoZ=AZ>w2zF6kop!$=r%{e}e z`wE-b<`-k_^dDJ`?Wq$m{MuyuI+E0aCZQCUdfL-_0X(uhG6hU2x4DVv>fOb=j&K^hMdhw8n`aKEwe3o-vDrJ~@i9?&@Q%GKbvNy+C-e ziWr$WO2@170}>57WM9@dljlwCZ}Ti}+QzSkpf8i|4XbSEFS^B|#7JApT2)?YoTQ$( zdAYQwdh-;SdGeLT?90IqA?a!pjw8ZcZ;!S_WdS}+!hjvYjMTfOjg z`&>V4kJ zU8Hiq(JLLC1`5^r%@=pNxd~R%M=Yoe4h8=hx<5cu?5N&}S8_~gLMHnhnCy(ZD}y_{ zDLR&c9p~qE&$<=nXBp?jRmWD zzhgD>d#y3(`WtN4y6p%j!%gdMmDIZ^-tKQ38=BSSRES%$1ZY2Rq{AG_6Th2+b~Bcg zn?|bR9=4kn-KXXTwNzK$$cOq)-A?JISBX?2c$a94yontXx;j4W;p}evjVt&&*L@%P z%IMTR1&`i`R;5sIm%Mu)+EcDCRf}07}G%y9Bf$N{AUEgDN;#tgF%aRmaf(P$-2o^w;e$1 zlqIE?_l0H1l~>Le8b495evkX7WaA?10_}^z%2{kuE7PyRNzpo>J0!NerY=zAJeo2i zm9i^7+`c>k&fK8m%S3h-v%D41Qx?*WlNYSK8+#rF+ z;Sz5jk%Y?%nsC}SQNHE^zP=Y-bSYA1#hpzxUi9WbgiZX#*FHq&NRz(>7Hf6xn=DWK zb-*P~z=^6Oo9mGz3@~egc2}#N%W$`C-llj*u9GO2lupVZ=1$1HS8&y<91_`eQ(Uxs zD4^V)V`TU0!NZAf7N&c=Z3+c1-Z`NRM_$~ZSLDODF;3xUbnpr*p9NM<-JGrO;(A`) zKv@V9Ffgxbs7q8MatBuj*q4u$l)rlW;pZm&&rJ>6ohLX~rCbYYCT~^oPP!jD7=JD@ zX;dXeD>8}QUF18n<$_i^*rw>3Y&6BDDBnk(yw=`lF8Mh&ks=@y8ZFLXv#h>4w5kvk zk@Q(5p*3RQn@vOsjD?OiH2l5awWO-#q|Jj!chM!b$+n0I@aG+J^(?!&eQQ*DKS!xe za4H??6VuMS7$u0mJ(u-z(!#Oz?k~{sSU#|wTSd;zATzCxQ+~x}(6Yp3N59XACT zPq@q5WX(=3I*4iXQz-1pVaq{K?enJC%0S_KY=g}Iy z0bB>!&b9kXwA8!BIfg0L%<8D=s*IqId+ge3#(yN6@qVdhv!q-&?>w<=WAh=&_=odQ ziC*8liz?%ni`u~@>~(!-=v{bi3h>nO*lbmrSU7Rp98((E?RsK<+4P`RK&fJpMd@c& z-TcyT3^ER`bJ|4C!o)p0<)KG`U!<1wbmi5k3N+~Vo;OxJS7@-!o3AEerLxM40{H8t zet(J3q{Alw#@MZT!Pw?aU2EGy*TDJwAGz|iu_^P34y;Pm=Ump>dBCj0Nw1@R#%{mc z9TLx4C9ynex?Ur>jDoiYA(7|i4;*_Qit`g7aw?Cis&nd+;_r#Q_X7or_Y*Q;f8Z31 zR-5{`SQ7&41D%tsLe+*@{$JDfP=+|vMG^VMp>FSbAeUW(Q?7TzufAzS;lNt+Znw2Ti*Y}^3Nvyz^$1lpyWg!Trp0oVsYa693;`nvvJsGFS zK2#ObVq>GsaF0}CjA)3>NKt#=WcWn`%PQps_xHkf6E(^c*;OsL?49P4Z`s&3db>`v zxF*pr78ZZa*c$MbniDA+PmUzIDY~q_Tsi?Lstr$lC4_suU=n)hsY)*;=0LrR zPRrR1oCY^5x2*bZA1EO%mz+`k`pK)yjymBVbAR6H8r7Gi34qmnFml^)+W1wxG<$zr z;KgBHf(X_Y!0|hIrYH8sCH#~9`!2zPIyu{hIh<%`ez%BK_<&QNuGi~;-iuudwU%># zbsOs(EDzD%&tiGnyEoL~+RzKIvZn+r?Q9#qC0ciQ)$N0Q;dXzr>*3XKr~P)#HTAW| zW9k5*-^()E+=Sy09uH5Ih(8fNp;-7<-ygttJ0(T3i$c#NkY~%9iml(h1aw^h1Sc6P zQ+-*BtG1LGFYPHXE)eN$`Ro-ra9G&$lDdEr9obRT%$zPLiWmuMJbB@5>@%s(Z)$UH z*Lo90c^~~~l(N&V*?u6qccJgJLgaFl9NEV`@L&v^wQp!&iPSR+jPVjpMxW}&3Fy8L zJEQucz++12Mc?*>{e5J25n%vZByDu}P%3z3K5?mG%gvtPjkb;{ zXAIQ5p71Qb8s%8+J*f7^aoOmr&rLz~rLVo~4j{-^K%#2C-p5N?&}REt)9@&CDU~s%z{~iRB`{#?vlLZqQBhz3^9tBeMsz{C5RZV)!wCJ9ty|q~*$3Jn= zy@FkO6@9Y@rl_4q6dpjbo?!{cx-{C^mvca+?R$@3jPY$Hi%*<{pg^7|DZ9QJ^3R~x zw@BB2u=Z)2dAky4bctUc1W-<`&fTK=3R%3qSK<*VvqTxV_)?H)&M)*X(fI1~v3RFA zfeE(;^06DZUfaQ8gaX(*S5W9R(aXAAC0W0a!($LonRAqKM8WQGH5+HP_taUl1ldQr zA*2ETO8dT-=O0U6)oSK__7a`PePnSC_rjf-`5zB zkNmoinif1g!V68XpXRg&ZAiA90o{qCIaQOr3w))cjBkx^iUk*Xg9Jf{0!Y}3b#s+d zYOe}S7r1yuR=%^6xTd>Pn+n!$-T+eBy(=i zQ+0(+H>m2gafu5!=?gvVN+o?8^{EySs}UqW_ZB&tcy*H(xU(m@f_BBhfspDoFK51p zq3qLEXbd8_HU8C!8XHB$?YfV=Q5xWkRjZkjF?##a@P7UPnVIHlRsx9tIj1)DXq^G_ zYAnxr3WJ=*#*(6{)2fv1ZMZyg-FfGdhFZ6-E6AH(w0Y|h`KBMG!>j@IwqNd*q@7r` zWM!3JNW9B=kfi^l4WJSTdz|dNZ@NGX8urVMmTN`p5W2 zskxDAt33^+q3(*u4!pHd8)_Gh3$J(*L>c_Jt?LnvP_9KCc5d4@<#Wx1V@7*HpzT}Z zeDgR%PbPub6-C7jKx(JUn`w%?rx0O=?};ZzBs|T8*QAjrZz}GO+V`dFS=c$gtRTv$ ztNc{gG5SNB7Ly9_hC%5}Q}5L>2}PWp7b81G?T5AN_YDr zDi?UIlk^jbUG>JR=pgHk}4b1UCm?cEg6E5GtxfM%Vm9US=9wb0k^2ih%m`Eky+YL^XHlPy1W z-E*THQJ;(D<(uG?$M-FuG{RQqsV5DTFyYn@vC~zOx$QOuCm$@MWSxJ5lZ8inX9;D} z7EUF~oU=O=m>n(iG4i|B@mNvhV_hs1JjC+;jy5Esqxpu4tV#iFxI)AcJ@!MhVy9#Vv z(>3Rz@;HQJ{i5&jvhKqI%C&Bm8Uiwb(f5YT3w+a$cPqrMkUdli zai-)Kx255TSHW1XFD2JYN*G!w9z5hJz;Po*@lXzHBuG6x8xB~M!ppvn)W(eDU9z5W zM8{>1ty5ju>NHwp1Sb^#qab&%`d7kRy@V-+<;qo$3S*F7Y$jZ5HhCcVn@=1!dKR(9v(!I~D&v{#7JNoKhlD z-S+%Ux8ThWeC``03yNTr@MsB7)5j7mwPU8xdf=EXi08|AH^bxyWHR5r9C%&&$; zC#6BoBgusQyHow0$eKs$kWf+3Y<-zweopkEM7Yeg>?#32N)7%skvH~RO~M?mx-->d z$a!(Wxt=~!CRWatCZ_cEm_KA@dhhRh!rY7D}u!|G0^d@Id{S4PhnuN5h!U_!` z>Cd$VySRC;!!~rC1=pr=%};% zdns4iKn~-!eC6OQ!Ue z*DF?saLfC+y0<90rusk(v|NZU@v}Up+OwVfrH_+*rF5@o>zml53v#VL_PtlFCgQdw zs2$*{i4IYB?L`QUH+nro}<_O~5J2HJIb@iIV zKgd!FeN=^tr;|^AZUfRSZy#@0(pp)#*Ymb&%ZzW1Vtjri?+E3BLBNJ4-*X>1ysT9i z76-mMCv!4SwSAUww$u!^W@v#kfO0r%_teMgSlGHO;^Iit59d9v1dN%HI;jMH`Erp| zS0b*G>s=zBfVJ1oy|jS9w~-yntO}B@pdIKIjoXy?Kg95;OpO1N*u&@4Q^OT%>>+PQ zIr4^6*S6AmdjEVQPrux4wIx=i?KYez>o>>hREappe3Q1@p!S{7z=3~5Ooepoi~u<9 zUFCXJ_-NAK&*ujp31uKi8tkmF|4ofrA+n2i&D&PP$1uE}>)B)F|o zw^Zb0R7!fX88tZK)84t*Z*mWpB9d-UroEw{<2ulM+C5j?j;rQxb&5Qv`XSa!hXr@& z-mbc>Psur5$FkYSWcC1UD_?)jEO(nxX`q6|c3#`=zUb$;6Xaw0RY{|)FRN$SbD!7- zy1AVQbEvP6PWr~0;{^_bxv2XgHNcAe&{aNjfbRYNA8aWLm4gwl4e(1X`o>&sp7rly z6ssHBxFb8I;&C^hUG1B^npnm+;ZV51D;U2nf_FjgfZ!6F*>b2w_k)=9T+Y=PUsRKc z8gbu%UyE{?rsk(EiR)F?5Be&dbf&~4<38F7Yb>is(Ux5eK26M3nveC;wy@XHswN+M z)ld=%e7AS=1JMH5lfC=9gAxW4E8$zH01+Gpt+ol?}nip#GPRi7hAk*r>`g2)0 z;kZMM62sDFw$$y3T7nO8;WCl#^1@e}Ozlm)7+H%qF3CB~p_4S?2|3XsEJ|~I*KN85 zODjdCKbLmd4mI<$zAKdB?Q!9`g|c2IoqOLk^c3Lw$kTOp2ded(65nzw^~`k&--%53 z5Ejhc0w;7YsCdQ+l*K|s6l$7xKx?|_1z)Zs*+2T*KG%0OTe_hO=r3jT@%_PlfF`NIcyJghlKMo zQ95ubwvnG?dyQECSwYD!`k(GmM&(m?8=|u>Dk%dfP;fP;5{IzyP2QapkDW?NHYR0q zFuL4Hl&B@uIVB6?3&n@wP_b<1Q}kuoRk!!JzjsJ>4~!BYD2^h&r^fGIehvw2$}W9s zV|aRtO=N_U98JnK0R9wua@n&UXTGg$EX5~EP`#i0 zA=dcaE{hp{gOVz}`69v&;x*fI6Mgmb1*fy|k|k{Nd&tw-c-2Ro2DwLk`OQ>?nk~}n zu9D}B<;<2H4FXI(XvPvEu1Bk#$92?3Pc=CIghc6gC}1E!!UG2N>tgPaCLr*jmR7b% zxxX$tz^c^QCuS`9J=cSoj9VpFANe9lTP+frtG7QWj~oT3ho2%7UR1XPy((}ST)U+RFAR=yeX^h zy6eaadFV}$(i!Tq1?c1x8SOH?=`1o~B&W!v7|LM-=dsdZ|7p{u3n`G7IJj?GEnZNa z)al%l0~#*qJ*oA=R@4}9f-WeOG5G{c)~hj9)=~P7Tw~o|?MzMVG3hGM4}W8n>UNH% zVDz!CzBlm{cPv$>f`t$W5c;wTc8mOg<9B1?m01jW9mG>z#hrWhlyjcjh~QCluXcQ##dWccJ^|hmk$0 zF4ZLJzzw(W@Hr3E9$tt@_pph4CT835*1KzOt@@85^8F;w>?$@MqtZT(`Efl}KQ<*M zxI)UKxcYjQ)}lRBnl0Z;tHHGDDGARM_{>K=&CZZUzkqstV0b25OVAF70`YAJ$`@*! zGZPtja)q3_mY7(*12j`i7Ri=!ry>*8cIuMPe`hjsh%bs5I}kq`K7pL=>E4)Rv(ORz z!VP?Ty>v`dC0Ex)vr>^=@oW(qlC2M7c-Lh=fT}`0q)eb5hF^U%`wv$vhy1*)H;B32 zU($`E0+!DF_6oGdCyoAoPc>98vK#P91>B?Du?UJLo|&^A^~i#g6?R4??azL2TCXzP zUHl4UN}XfmBM3Qb6UDP_~_Kv{TtOR^jo@Gq!4dfWMY|ocI7AZgU_8Liv2h zKBxoF@6+8}$!UE^@LTr44F5; z;Ud)~#+_Woh*~t(=R=IkdE43T>s6In0`?1 zDU{NV_5{Yjyn8xlSs}obtGw6%-8j(GVSP37Dh$gX3SEd)GgU2z-E32~E3Mf9Zlq5M z>H#=`${7h8x65r%x9h0!2DCtxvObQi_T10z9%8NKu`ZNnZ&~Ay?Bh@D``qVRvu4fAnl*DATrZBh!0lr>9f za=rurtBI)HLaE(CIi8x4Y=dR;ML2UQ&mw1?N3nN@z5pIotVXV?E=SYotsL)mLTGY) z!bZ-h4qz}ZAMwx0`{0-a65P*m$^Q=Z2GCsE)OWR_i^*&p;ZkgkH*pPdv79bg0A`rg zmpbe77X^CHL=*=NEdaNt!ObOt4#;}uQ>)>EjdHG9O}hC_J5LqL!jc+>bs8zkT1>g0 zR>5_MFwgmLzH|C(m8(}?FW+|f5fArN=`*AK(?3r`=54Yp{|rxk)!rn}Xv>R8=gJ}; zyHT2nM&-xjQ|%e~0%(bt?;&X}#-Xt$r9AEoL$%u%p^e7QR_iM}-Zra-Rcpzyl|I}; zq|&VfYOAb$rc-e*LY1twx8^TJ<|ez?&jVJj-at1q3Av+dgKyLGN4&DitIIyQvl{k}mBEG~0iaz<`eT|saqoRHLPD=quPMJ#ch za(?b$CK7Ey@vTvGYx$XO#gVCTh65LebA`)6nX4U&loKRe&l>?#&b8YpN|Rs(-^fuN zjByqe$YnuykWV<2q|xSUEWgZ=L1QkAwQq{{c11e+v;3rfu-lz3*;8>+xFn9r6&k0A zT}2!9q59M@gp^tV3f97`+`AniDRqZ)c4GplfAa#lh*nI3kle~Fu)L`~N^^1=GCt~6 zp|st7Q%#ksy#G_+LS8BflG^YWD))opi2vx=)7FxX%gc$|C;Z6;i@SO1t56e3nl_76 zZ&4(dzP5W|;;^+=1u)?bpRe8Op9Vu^TWF9KirN@3PbcyE<#%4+dRIVdRiNApjpRB% zfLV!y04?|R(D0&DYk^paB~+4i?TER1BA(Cc!~n@Q3R=Fn-s!J`jTC%l%}q zg36rzB{hiG*%J=l_ZH_wLt)B%8turhz}Bq2tpl4$Xt8L{9u`GIS->w($5_XjiNUmA_>O#H1!5(}@Ku}$rk+Q>vaJC{1zgmE<7Pb(L; zmyfLHJY4HkG`A8d4eSv`w`E7{h#~c)qVDEwc_HAND2i7kS(cpEJbDhzX>)ca2VTMk^}rxPULZ4Yw`%XnRt^|$T2s_IUM^v1uI zn?JgJKbT;Sr30a@sN(1Z`1kEK_v%geS#Ia|&1Ab3YiT^H$bE6;6JcB8`ta5vNLF)f z7T1;G2t`zn9}^wA18V1xUE$??s()cAIwzUISV{&>nz2oHGb-6Y)y?WLW3}*Ad}xBJ z>BVs%pg{jAm$S`6&*h>o&;gOQ@-Tokai_n>-$n-a`Oe1R*(Q=i?P)YZ*oP!3646t3 zXtw=2sXaNQysoq5`2i#Wf3UqhD6!iu(W5ar@go*$sagj&leb&Uzaq69W}|UF48rV7 z_aoh314^!!a_nNfvdx1fbDQy8_AzlX;kEYX%gz=?jqJa>pT8m9&)LipC5nJO^F|J8 zRa5YN25bG3phJRY%cUkN<~4V&OQ(I_6Lkxs-I%XMNlmmDE;}pfHL~$$!P01t z-ihEwT4IkB$wf(FWym2#yvjZawf>U6d)la-Q@Jje-Jo14X#{FkgsrBKDkv2a*Cx4y z;u)$^1OX$qVuH2xXTRNkubnD_x7lT|_sKx;V1eChjGcJlG*6BP#P&m+`xFB9b`Mi& z({#U$QO=S@y2!7NeM!=-__ZSJH7=Y$Q+hqQKL<)~gq!cm$I;K}QKK$XhT_aJw zEiOM)-kRTzdyxoj_%wp)b%ZMp8G%TD|94qr*E*U{H1dlhiGwab`n41`!$@gS{lHzw z7Qg$r^Au_a&1OTLLid?dP+xAoctU@$)hKM_<&Q?F5O3nn=Sx6wry3ON)qTCp{owJ# z((++0I)@g%)=(}32=~&EawgRohJ(ss)9e1~~|JNdwM z2lCZVs>a2tSfHlPPI+`)Z?wnbHF#fY@PU&zaOaloMpw8-zdmQTm9IxdQW-&sla|E8 z7f9LREXdM%2dTbWJ0kV#$SfE)Q&E6Z?s6YleK8eSw+{Ao%B*42YiW~yr^QxAc1^;T zvYyDVZW>eE79_d$a!Q13N^0?bB*%tnsVr{wd1#S^PM&$wRlS_tOGmdGM!B^DJqacN zsI0wQ(zEzl)M)J8z(%qObM2F=$lemhyLX(T_rwazUe(v=oM&w!<6&ziQX3}WElBW- zDQuaATV9P1;Q!;`R`a=Bp9=0;yIXHzLe0LJaJf$5;`-PIrtq!SWOy!)^&n|4ww0i}vh%IS7=N=-1*;23hQlM(m z`z6}$!u$iOZR}QW)k1LJw+6-utxg+*M|eM##a7>6@7o!IAg9puBM4jflvyoO^E?&e z^W-9v?3JFcnmCgXo#j5T*jtu=&OSC~3@vcGyDH6uoZ#1{?@J&JC;(1=h((}5&`rsf zMLVHLX0ovbkiuTt5+?Bp1Xbe!&oIY^nNfNO>4je&q(M=k*-=kU)55aL>IblG&SQ$+ z1e#IPPKdJrO7Tp>einiX^+PuqQp!hXcgLgKeQ)S5=2p<`)E8$PF`_zrnXxJj?R3&zE49 z=5S$c+q1V?z<7XEd3`0)|K+=OQ)>NY>sL`IC*v*qc*3@$qPwsNPyEpzt)?uo1D!WZm5V1P2MMdS^!SDS|JargUS;t z@R^{2s?_D3O^ZV00#X@R*tXxcfYcNxbV&AAM(X^^|Kzrp@2|$`PB8>^<-?q56g-3B zR1=He1{n{K0uFt_p~+E}{z&hG$B`a)-5^FNSsSbb|1**`e(+XNBE!|k*1FkuugnJu zxmaC{&%XepjkN{wBNtm!F1LJCl-7E=j`{=hLU2p6G* z)gsr(_hYPn#LvC-w;8jCup5p#Fe$gi{9A2fB^R1de~OvuH{uD#V)2&B_0HvA+`kMJ zs#8EAx!o>Y_#Mz-zM1JHY?byi1C!}uzKU9fxG>auRJYpLc5!|#bkL<-J7|jBQ`#Aa zaB1x|yTm(JzNLa}y$#@_+zm4qPaQDTfpC8dDlzx_rc5i405DRdJ5+z|)x(3`(duAC z3x{+0=$O@C*W5wYN6-F!+j5A=xZUIG48eq&tAMd`_070(9KxQbhHO;PUTZ5{4-5H;nRm zrHbJDSV9j{S6QSdpkefnvAj?NMh_>2^p#sEg-;zPacBvioncdf6KSZviP-^io0i3d zkRPIsKz#B5Osk*|6o7dxufZSrIye?^S#DES)d%cZP|b?*2y1g%)uO{-GQUCn20OO# zMX!n?q_%L0%rnf?2k|SWbcjc-5~PFV88)2X&cp6qWFHHwKC(AvuY^MbJ#+#W8oX7> zCP;$C)?Y$tuvX1BQU)9^{m!WNnT(fGGJu57LNMp*hWJdc?O4<{!5+cZ_~kO>EcuA_ z9`p0a*r^8P2TNeBxVSVcNO@ymCs0*It4#ch*XGK3{tC|nh<3%2+kW=!cAtDHar8p zZX-)?bxMko*vGibE;^u?+qT5omPpN+c+((iy1KvI#?{2o?zSdCxT47JEDHhc8k@Q? z^bkfJ0DYO@Fj6coyF*p=E1;FgYG?uB&HGSCx7bmUvH?=QkJJF{Kq=L?iV37c;!fen zKD3hz*UfFhG5O-~IQ^y)Qh3~jFZs)7YV>0KV}|6=l6rHhzM9AbG{eC%!+HHc+UD(_ zd!?pU6QxTy!RbegH_%Co57XsrB4Q^b;c7<-aspa~$5#RfD)-&&8F6rrLW_`oS|@Zo zY@5ndA@ibk@IvPGo#%KEThwEyRY)~H-hUE7yS&M8J4y=E4~dO4MEH!7O@(eE$O(4~ z30iL8ZZ-mMB!z_ml)+f^mPdkrxG?}pvsHB?Xg!#IjWit08dH^PKbo`|OLNW_Z%2IP z8iW*?I-2aVXrwtn7jGLtEi9@gszHlX*`M3Xei&%A?ai7zXmhR-@C1vj0-OR!=swjT zTMkxrd~_^OKI;dA?l+HMsX@3KAFFMY+{kaez4yKqqFn3E@|tuALA0RfN1`e6iwiUN zD58Gj=tcXBI5)uxrFSzdeFKz>EJdlOjHrhX*a0dys(|T_c>Yol>%8xH5@Q(Ho|N6kvj1Y8^yUv5mM+awv$qj zLLl}u>!R7d@9lDbQ2+w_sRrq+AcjI}llMtTTg5EUnm+gdc_NhdkQ32L-6dPi`uk%$ z!3R;Qx9QClyj{)n6-U*-CQd@d_&I3s zCmVVPsF5lu{>#Eaj8ezGhm?sUzd^h)U*kh~7m4>ih*t_hyhm&5g%5a`S`=JJymt|Y z)t_)^t=)Vc8XdcDDoZd)GYU1weR^WLT)nw46x#;2eF`;1FaaPLtq6#$p*YxDm)@H% zu0%<>3{SQ+KDLz~;6r<)nNGGD6;*uGhIZur#ah;yD* z1})Bg1fm4TKRxn>z0fnK6Du{}MnqP?9oay}n1{6gD@!vls6g(7I{OI4?*`S9d(fcD zQB@(uiG;4V5xfB#$Z`Qbak8rQD;`zdW@48B0IuWk(~)$gb4mE-z~tumHTXFe@Dja=;aOs2^`QH zRHy+a4S@E^?ob0tiVi$7J*PJ0Jib(fZo9^)2&MHA8ne}^P-`mGuKzQcuN@E@e`Bi} zLed!N!FeC+P5N;iI$W`vzr8=&09}+32>jbsfcA@3Pg@wWP(W`44$#X)CEHjEA{fek zfJYD6tPWNj=#Rz&p>Tu2`Ea@cS}G%07sPSDAU$tdg{rqXn?~icquK$k?yzYTP!fy) zC4sp37PMm5A|N9G-DRNlsRPnNF9Yb3r3YZ91@nEy>p+To18Q9P+g7guBgT0mSKxLsfX|+H!9bpTAf=fkyh5p2UqA6TWXe+(L>!|O0ofYGV{flL&@ zbmpO99BKlbp(fycvF^YF2r`%v)H6d$L}{P7MqN8bb`+B-Gy_#ny7Cf#22>r|$Iq%G z>;@}1s6uzRkXqfIE3}f z!!eLHKBu96QAl&WN%B7;-t}W)Q9A?_bCn(?3-P2zx-!Yo&|b5M$g0s6{_F#Miy<)S zrKdQ@SAF;R=Zda#;BC|9ZzDh|W^k<)cR9Xq?lq|q7dJvnVaNjRT)|oc)wHGS=;{~& zsNBALop0GC9y6kDppDeHn;3fgI$_kmyU92DyqIJU5rlj+KnFEsSGQs7c-+fQ$Hfq= z{7(toTYz=&Hsd+1qe5^VSt1G2ILraB)%4Kt=-<-4OylS)eGy)3eU;Bb4$!<-uDRPI z`1x%JO2!RPrl}?($HhPE z;BG(76GOnP#+}p3Db%pr`O*e6Ui$83 zZ>^w#X?SVJf)1X2Ktakjq51g7Ad5UTv=AfFSPoc3z5EB8!sr-XF)sF~L)`ZDE=d_L zsR$kV%LqcZAcFp(Lu7rL4FThdc5D~SIYr7`MhPv}QOPw>V(ul;`Dz%nCFca-R=~N9 zF0%^K{j>B_3Dxt_a-2`ck2pR2h<(l1ANjVBD zd_v>SF#dnxKT5~Ih&h{?>v_^SZzK3N;jk*S`Kl{SOduX0f|9-pzBRJ>Km8T}Apv{| z0)!<0KLR0jBUm5rKTrX8^}pR6?@M@0cJEX=B>*weaskLeO$lrs8@!<1ie+aJT}3wO zlth?f;>}ri2rzQL(=Pxug^b} z=btmBw{###W%&`aJazkEJhHpWVSp9KXi$xXPHcZMbLRjv_XbDbOX!6qR&v7Ni0gyG zR6gLAh$A&}F)>FF2P}~r77nvLIt(yTz%=O8fN6=V5O&G14vO0V#&}Pei}@N;G#*gMwC>^R`4`{ zq6aVB!n`pZ-ndUBADmYeuLfqry9S*1=z1=}_!0&oujUuda=GPbm1p^jw^mWrb0T${ z^)w~WoMe|8yKRoQ9Z#jSip)>|f}KrYBC-#<0~ku<4FE`Ss+I*nzU;vMRRmm$iTGz6 z;JhLLxS~5Vsa?QTRUp`5aB3nMShi;omNf(qP}-zQ4^b8OziM_8pc=IGL@)p@m;t)M z0DJV|KZOl&N1{HT!j>kUs!1|;jB+(aVjZk?so9^!j#q((q_3E5pnUJ3*b(U~pwgDv@6&F}sc*OU+_{jO?lGjqNUxbfyWy0w5d z8;?a!rgZ4zHiC%r+J7`-N&2vM=kTQWsb5G3@9anJ`!FXa3>=HIw;98VAMX$KZR+l zGc`wc34t!6T{9bSojy(1A*;yi>&?p~Bw#)^sj(cFB~_qUHuD^V(wLCXUYoKt%D3?x z2z!+cUT6zMFPBcu&CvpAZA|MIgC5<&Wi7(!)g|A(2ltAdl!3kH>2}RGTqHzF&`n>D z1z1l9VhVqPZN13bfISEZ6V#|-`Q+x}oMha46x!yRRUxvD8o=+RB|$PD{wJ9WkTY>6 zag6o3Kbd>!5WxVZQt>x=pV(|Rinz)eQ3T>Scm4s{M4twLrGHxQMjBwdxp~uhWr7WM zd$Le4sdI|p4Sr+IWlP5*K+k>7m5@rzoP9|j|M(9OT=cOPCKb{#Ew|-41aIaT=pE%9 zaaDF{!x$Vrvq=CAA5CLw%FTTP_RcB;LxSQb-D>1ZiXzs^lZ3sJq5_*U-mwc96Q(%w z`m#J=fhxFRfqIK)qiO!ae7=_shli?HLNzx*;ILlHrM}E_IM_IaWc^kMTC}77QQ-NR zzW~-a=L3T8_MV{oBpbo<^n;!`1Wx}D0ZvW=bXF3l88b4weZ1}U-rg`68A8uu&4a-( z-w3+z?}K-bfWm+4h{4%I1lvkDD*q-hI*x#i!~6Wy&F`!KrjUv0rD^E74%#ejc*o~S z9WE2PtnbkwPXP*c3!0NI1k94$G0rvwMqFujHj)debnq*NlLT@^MD~3X4r?Hx?A~J@ z@^lQs?g$vk%BP}?FmyMnL-6dS3LQH2|HApg^0CCbdJM+erO(l(|H%7t+tqi^oZ7(8 zi7oBSOJD=HTs>T7dwK#^1OMr)Xj;gb2^({a%OMf&>u5xKJM@4uMkaW)^2H&+5!Sc6n)pk3A4&CAg@S}yEA-OUIQOhUTQWpSRvfM$#k(QMwIC8$98K4^!GzLRYKfvFD*MAkjYc* zof82%Gl^r=x^DGZX!VBCIpr)r#3JPV+*Q0*olL~YSrGF7L5_Za_h8FJBiLYZJ#`Y1 zL|#1jLIGh&)_F@a;JAApH2v6gVDS95K<$w122NvTE@7+RcQ1tK`x<83KdO0#$~v~J zcs~CrVj97kJesTQG9q! zVGvNIV3BuV=WKAzxxf$i>lr33G2Gnq(k%Z0KCX^d&FQEN`tI5_aGuyW6a-=0jdP+X zc{S-D&yZnA8w!HAvFN@g?_CpycX1odH?=fP6KmF;L-dDY;eKTj1|jkd74>*zk19I& zg}wq2LH9Y{O^q8?S??sH0S1k$_yNMgM2IuwaX>x3=|C7*Ty#CkK$$gm4I06I<38d# zouXce?6E-yYhq~d7}X15!4QjvjdPQ5ZF%3MX|(<3`rnbpW1$%A110sG&G z-M+9h@Tk#JkeflNAH=PKw#1A|N1S`nX?^@44HC!WW(>4onS>_5CL!> zM#S~pjDj&LpeAZmYG-TzIo?+lS};&$zmO14*S)WzJZIzBJcG}5ypnAb$0n&)ikIyV zr^;IZ+Huz&wr9#QAHFJ-_3G~XV-vn&z};2ml9>QLj&%ZI_>wl{#!1}E7j0nZbXe}g zkR|~P`Lgp1$382FO(>{WcN^KNlZEG_)z$%db?K%+p$XxN07vi?qC_>XapvLSDTdCc zuS70kWJRqFyO}qyjsoh_YEF(y^=|Lq!Fzz01fEd6e)z**j7Mx7TEd4g=tokogy+JL z(;}EbS=h~oq*rnBhF_K#^t`0he8*lZgJPXBphhN*p{8QY%dgQ#WnJ4=Tx+0PkwKP- zQ613aO1+sa!JDQMP5j!Ii<5F25SB#P;y&z2ooi~ZRybr(>LTFWH4(a90C+A4GE|MhPR8n9{jP8lo17zWUVrr5$X&UmW1S=$jB-^~h zq>gYeK$1xFmWm#eG?c}hk@GM@51Kxo0D6`y{f(F20TG6&c zjqF_`52<7Lb_n+q;`CD!&e9hgsXofTi|(>T76bq*T`s*TalkArA{x@TSm_X^Dqg8} zt6K<5pPOa0;VY^JSBCnrrPZA35eftX`T<)&uPSCerBkG_4XzS#Cb03q~&fR7XS11e9O`*ZCYVT#tVy;6Sh(uuY+NXH-0!5q0}Orky*E`dQycjLU&o z-IO(;8}bU6Z^IBN2wM&(@Hp)m+<^^{c$s~yKlC~oK+)-it>> z%|pcf;td9%{cwQ+{1e6Gq5a&xmhe1!6QGv9f-@f4v(v2xU+g7OD4Jl%{2Snu;O+&( ziWUS_?f19A{}$)dCXf~-XjXEgqVJ%kMf88N+3pgP;bKWr^o>cQ6mhtV{od;Y8Sd}_3L=8#QXXJ{bgx6;ywRM6Vowa2f1@W80PJ!_OWwHz4C*&# z8ivN$OkOBFAk{QE4_q32&8Y-z1)QM@h%y8GVncJwu~(OVPuPbB!@MF?=QTZO%H(wr zO*Um`2Yn(VyhAMT6jyj4<|zdn$MD!H9J7a(md7x&Mt;f%M|SVWBes^&HqFC?aSlgZ zQ^M_WWEu#@z3InK&F^Oy-%Z@1`(v)4=7@BR1Z709DlAM1mB6w4?06K1ka{rnBk;`c zO#W}hMW)`!0aF*Sl0loQ@J@INM0o0oq+8TS5hD*^bUq~`;A$B_)JYb zh#xXXjk5vU?tPHVUMenkY{sR~r4Yr8xkVZ@)sJ}|NlG9C0yk@y(^&Oi?$7?S3Q(Q8 z%K#yrUY6%YMc&Tz-(2=UI??=xbmCCg;Z}p9`gtaBFH~s4Sup2EvF?R9?Zg&*%zPoP zY^I^qi$%nTk(wDR?-J<&a;zVl<7$yAxa+$&#y!4xf>PyS>1n%mew1s(^f-d`PHg}q ztvck{Jrp;!vf&65TKh}gbHzARStcVYZk`hm%-9MJ5gWawrw}bBdVNSUg?Ss`Y>wFx z#CA6NZ&OjqMJ0Cz3^?D!rSJ?DrAGwN4Z9>OKVwJUTe=5_{y{bdHyF#O3fF(j^U5{U z86z~F5Rkl}xQKgkKGqu+5o5yGQ?PkCF1*QsFr&&ls%r1_9Ks;UPf9vwuI~RLk+Fqp zV$-jNEJ3V4+mIzihR9tMd?8X)e1ad@s$=nB^h^ z67c!N!#m71?T7zbxVM_p!3H*#B#zc?MpbZHp4a8Q{>Rt3QSj<-D9lR3i?$91#7OKj zAI;m5Y>en;rSzNh^`8i4b0Q=Q@0WgM9QsJ_;9}m(>-FUIIoCgGW4Tl13wLYhFOzYc zcJG_ne=!>8>9ANi{jG5GQh7^@*R-ogl)HI*)_RUnlBbvZ#%Tg+EF4mAEZk!(-oXac zJKxy3Gvgwol6P-3+sXQe>}d;7;4)%i6KpAaw=1Jef+k6SC14eQuOwz@kV78y*2N`V z9)6@q-}~%(i(*r|iRmdp{91XL@9~tFpRHi7nNHLCQLFvuJ;;w0F~Pe=@UCCvV_=f{ zUL3(oW)I}pnD5_UshKvtcaDYt^9lJfjPE_Ubbb4&(x*?OufRK;9}vDVIaw<&dWHrM z^Q~Goc%~m96aRUK4=*9(SP)I0LdAtpn9gX^+^>n#{)B2c0rALW8P(5WOY#tXYM;P- zI(!SMk_r}?6kox2o^yWi48t#eg*XC{va+L?F}x%DzREB3>^Lsw)8S)2E$|lei2fl$ zS*lgCl}f#Y^%IJ5Z`*_Ek$?Zn(qSAn`((sE>=`Y*%|WhJ{@-u2c!b65b#v?L{rgSE zlCoYk%G5^8-@lKY#)$j1f_Hm!w>m#5=F{&b;)f+lrWyQxlvo4AKw|q2-Ol76;_Vk| zh56m6)9{v?a-XID{gx(wcnh1hAHASFM%b*OyiKp>BL36@{67Vs5{Ap4!~ZDQpy6_K zMElE-ZG!FDEFyMAnK*lXPHH@iAR<64U-P{UiT^!Nv@#Z1a#{zcPgK2!|DKtaUtJ%I zipZHWDt;h(ENpXvU#_z%=Xi1dy~f#~M>@wX$p|jP+|;Zti--`{`uoXzJd0UW`v|P$ z*tP&%zuy{xO!4Y9EVAkeBE}<|a-<-A0d4A>;~~61uAa?1g8AV)pcZM(ogXj$Rg0*j zK4GC{=|;>ZMW9jUe0ML`I3rTYjQ&wmMnFM~i%o&~{Tf#s4}DPv=HvdD@nW7oM~E%I z_XPKYw6G^7yp8CNh>lt~?_(jOh9KliKBO=y8rhS}*#Atb#u+Dgt4gthVb}8%JhHE2 z62f&kPAM#m2=kT{f84DKPNRSoGFG0#@8dj~Ih5>z_Y2*Z3vhtf1h?fpY`!e+1!we|bz6QSQq zj%YuYk1+w`7*$=8&VgbLJ4P>}L;tr;m=C(%sBpP4jBtw?hGt$t`sfV3(JQ5$mJ^tl zRFQ%v#3~uF|Dy@g=U`FgLhg!qZ85?lL4_ltF24WUdmQ0CeV65fdEr*k-qJZ~o5*SPU^B!*=c#mhutxl&BYLgEqjs2UE9mS-~oP33hZRJM`CuT={ z7l!?@AHmRzF(!{1|6^E!?@ z85?=cD-O(Xp2dKkrk#+8_}4ooq`|Do@(#O_1(gG%ksh7;EZD4lh&ip8?an`}-TVEld#Ti!$Q zh@Z|03VRuN&`!i-5ehngc!lL1D%m6C!DYT0oWs#!^vE9cD2PH?+)_Ek0;5N#LjK?D zn=%1QrCfz|^}E+&W3QXCl6EccD@DoJzD1DqVwBvQ7L@!U3$gZZ)kX9bR!-iddOT=- zneZg*lrZZ=qjTho%cRnA=X^ySWHD~Y4xU=$GVUk($2@;O4Nvtk`%36goRftKwToup zMMUymXY_azh*^C*#15&xp!g~r+zQjpVG|9nyHeCa{^XZKK zpH4%X5Z)8Z*i_a}>XQWuyBAnH)_9qK!mFN;g9)(+DcWG_D4UZ9FrN-Xfjw}~OKMqp zMbtPt$gcQQFBq#s#`fT2%h|(?bO|<+1PSBRzm;4Tl$@$Xg&Zz(5hcf>ROF%}{hw{u zAOD04TS=;;W=zuzt}_t}PNcPbghk~f!x2u5mBQX6ga!Jme}_g>1jjOXE4Lc`LC5y} zWEKS!li9WU6H+{KIF}l~IcUY>@R>C>4tiR9&On#4qxt*!obnb*UF++m2Hij4Vu{}m zpCn>fDm%iiVmVO9X~4AAy0F!(3cWMmYm~X|K42Of4=lC`HYEcM-ySq`S;l`~5U~se zA&hVb>;$r?beqeQ{9eZ7*sOn)4P<$XJt{Cb_c4jYnz{oth?QRBcU_s#G?F^8VbT~O zec=Qf!GVMXP-6vK1NI84}W|!`(uG>=x$Z)c@Ws)ILS>drQ>^Bb>$E zH>QBUmB0LFLFhKLD%uiD?b$m(chu{tWk3|La<1JyH-VHKPZhRHDu-WAJ>t2$P->R* zeu}@+ebbg@qVF`%XFnpUv95Emf;q~{pml02rz0YKcz;+J@)BY8&2uEeyZNJr`XBAa z#wU2qI@TNwQ%m&P--)*=9qv80m1tcMqI380S0eYN}TtrN0d<01@h zGfcFJbWZ(Dj!dHgTY}fvk~q2$rRWL?c$&(nB0@tM=8M_2jmWOmmgdF(XV<2v<5)i0 zdN2WvdP$Ixa1WbH-&`6U%_{hEVJaI z*YkLO(?^~&9mQ5!g_AM5339}j>Y*=6+~$pHft((dg3t=>@Z$r(exw;dje2gzLkq)T zi}-3wOtYr0jp;_eNhGm0{&8xcvZ|+VKwhaQBP=5~- zl~BGn%6G=-*X81VozspxG<5-B>I!(*RF40n<=;WIF7M4m4PBawtMmBv5#J0^^szmo zAUZR==^@+l>I}frg(&UDwRMzP{oV2GZ*C z*q)5ld}iSCBcSytt+k=s`dm+5RPKsY`BGJRk}?U&#X$>})gCQ__#wOM;hp@!dztZ# z45)7~cFydL>GM!oJL0~PZIxa86T4P4Skt&PZoUFb!k={W?Lz6Om)S(q<0-fW zE**2N*I-6Nk-ge#%@LBgM7yd1NsC#iovu`=?nFCW@e zda$X(p@fu}ZH-ZMf(bQ78!`iWi_HorF{_|5C$G$X*uy0_v+Ys`)_8Y%+U-V5hHS+g znz>5jKKkJg-QEF>t6V@0JND^d>*5sW=-Ge7TSz;thU32bRW$r(lbxj{pECFR*TW$r z`qI6AxDsbw)_b&C@u=Gq^Jg?(w?!^Vdv3!))?*DoFMn5%05Y# zch+{*`z7IWlX59-XpOMbLSK%b7HrysUoRoiS?f4=?@I{+(5@y@REq z&_N_vuE{YJK&5Z7G3;!xTcbo0@zpvUv4EBeAM^7`+0GCVOks{i-cJz=`1d@55oQ#N z52ySd18T4H(b@&OS1UZxrm98byMeT}d9sDL(_bii?!`ih>0CFlr=W?XwEO^szSlcC$vTmFCSrt?bm;R__Ms=D z(F?_W`U1^7)fL0hIds!RoO#~+>tJN86~d^EMwvT^*IcKQN?RRWiMA8O-O5r*8P<-< zm_9AbA-@t$;v>4t{*rD@O=A6-&cn{j;rCXwN&v=7DtlgO{<^u~+ga7X5DIop&!`l9 zMR`0}1kWf^ERI38wV+EVUM32_`r}}*4#MoZ@M-Y5o3piqe@sx+Y`~h23Krj#0 zts^MtL@!!zDE6x?c=#o|bmSgMbu@K$(t2ma13l-7$BKh`Tlewq=skTCWk7-|TGb^X zYlvhR9jv_0gbJiF%bV1jD-PxU(DZP=w>5_KhAVSre2Uv<&tOJo&OPNii$1SgZ0zJK z8si46zJsOA{vOdIokQ4zj=A{k7d9z`q6NRwQua8__nBxcY-U6sES4_~u2SBTRa7ls zE7mpetr|Jg*`67_WqG$sr0*}lL+IuNE_T`K9r}NB){C{1I z`VAbNkgeKxr?&v!Gf3j0rdTFNqLY}icF&?k$QSb1R5t7b7$QGtSJ|nx@uH#tkLb6Z z#P=(OQic!6y&Zp9nGTsH+E{NVzZJSL(YUPUdA23fI?O3Iv$wo{dSlY!RsI_3<$cz+ zO!2VWXD8;Dc3gKg{iAqjdv(m|bQ|^4tib>>SjAZP)djcBO>2#_{V=e*GNV@Mm?FBj zFk})bVb-f_C$Vc1ef6tROM-zaNzARQib1^1IpVj;k$^ce_@rRa5+qFe;|>!hSCNzd z6TCCd{|d-BeK;?BEEE`+#hd;VCzt1NzOA@r;5N(3F;SW;Gg9tsSKhS`?MlT#*$t4l z-5y0jV8UN!+mgu@R{pIL81UYpIs%lC~nuA0Z-Xz)QkTbF}KdDr>N}~L-mcOzILgd85Z5KD?$4iCm(zo{$08+X7bvul z_~B}ADw=?T9*^keVF1dD1fYR0LjE=1W6V)$!3O8+8}6$K4LRmn-fGsvkIC=|m%r8X zF|zCQsg+E+z)aB<@=C4SmaMBYsgX`D1Z0Hmmto&t&mEo@eXij+F@8}iz(bO)_7Xm)UXvM%p^DYzs& z;*vVsm9D0m>4=r%B&(bkYoGfjX$$EoxgfN{(y4c`DQ06Km?a|5A@kXzFP(4KAzF#) z>13)Hd^cQmf?mA;ID=$G9^G~)Q(pN(Ng|ib$ zZ;moFOAw-)xy!Q_GKDX^(%b6%1a|NKyx*%f|Gh7t7eCB%yIU+mJB5Yh_5_7)qtv{n z8s;p7^S_>jQm)|Wgl_SxnHtJHpuz8q9`wWBxZVexIT6&Br#L)HSoX0iE@SFZT0GI_C#*m$zZPm7!LlaX_9E8l}H=zf1$EvW2IaGTYG5dl@kh_q3dpg54 z=k)gAO9mmj7v;lFLj$5y3Y>^X>S0!NKqbFR`^6kuZ9ZfsU=>htfcJDSAK__6%P z-QGleH(n+q-HlcVF?$=_WYI!>cTAjy9knGB->J<&?9Y2-(R%2Dzjkl}nr+W65 zyh(9UQQRr5JDo~V2H<)Y;;}`gUJlAVx<_<%UdEhyMkcg(1cHm9K~)mZTjmBN{%GJZ z84Qw!+w|h)TPOCRTe(Z+xk$boAwyCZW0h4_eqg+y<&B&Je=8(U_&$(n^fM{j%4Dnc zVR*xieBQd?A-*`6jbPC1rF04X>g1W74dPre98Slt`|& z2$-dYUo0^GP(MgW!PY-oI?9!61!SzpcGacDrnK}NH5|Ye>xkXTxLmxbVdV)0dl#kl z*Pco=X_oIh{>mg)$o}@2htCC$g+(M(XSwG;`pX+Ge2s0&e!ng)ZPMHzvC&^=UAf-N z2H}q#?$3f|?wBC|8LrZAUw8JW675bCt_)7*29df2lNGzkR^=>=RF*GP{RpH@&@oTy ziJ`l3g(f=NQlTmLRPvYN>+h}fRZg{^RZWWuJWgXw@@DvcBi%w>Qmld3z~YDMrK;VZ zgE#g)vfCj{8kVfTZpr+~U|Hi5LnYa-Iufb4>15PM4q2rt^J5f*71xZfUhvFBr1>tz z(Sg*TP6G$#;7pq-pw=QH(1|Z zmG)Zs%x*%%{#YZ;?(9kBlo*2SH@Bw;>3H&Qo6Js=XF6`pEI}LCMU(2)HZjd84OK_F zgp4PgR0baT3#r4y8IHL}R18r2jvU?h($`_hO`H?07B|G2il6Fv&B*wEIbvzPHR8Tx z)4Z*;*2BH^QZ6Xr-Ocm_sK6ZPU2DJD6l>rXW#4(3YT${Co5d?(qK8UZaN@gjeR2qY z&Ooh6i-C@xll#IzD%~OyUV5r`WJQhYc)XIF?R{QcvC$%IhI+=>ylv1CF zTu6E?R0N0hM9Z!>@xU9^_;#am?47X48k=8l82wkJioX3adfXYAh@8$u8fBi!hsWZ-6(Gx2p$!2tKHHnFss&z7m4cfzF>OOuHg#642?3GhQI51ES>=gqYmdH`kUy)eb zU#jDb_;3d)q+85Y%dSe}cPzres>DCiX*~0yXS{B<02k@X5&cu&Ebz|Ux#K5o2ru`+ zJLwEM(~sMP%D>)82EQ1GwVm-Di&p~o2D5iuGPzQqWr8>HD+xIEw!neQrkz{871fiu zG*D>XWwI+SZH|8vvXEy6L`xh}Phl=>d|;w@lvjLZ41L98!WntZkk+^)UimUv}xLW0LhE+Ba~5w^>h!+lt9C z<^)5+PNk`U&!0Bix{UtV zjOCLDKM+J`P~i|MeS-|r`<2vh?vV-Kem8cPmq40P4#a)@2frYsNd>W2lDmF<6oEzX zg-0i`e10L1GMI=z^pj>`mXxic&4;1@+ai9s<{GdhUtq%LkC?*uYiMyMr0iq^UXq@Z zfT_#^SL9V$EWYZqJ`Jt41cDdFAQ4Yl;K4Sfz`@Ygf`d&UsOIg_BQ6KEt+;zU2Gd(SdqR8Z?s1qs4H&UxNr z#JD(YAZO&V&F2EH2l;BG#D=!R+`DwmaL5z2Ss+P`-BG5XGsb}XAT>e01et}eYY;S9 zLv%X@sk5esZAb(C6p{%!9wE6iqwSyvp^6z4nnHMA!dLq3L1?AfT=%;-R(&}e2fDW( z{#@cw^=e8sstb%~i4bA#-LiBXy!%wOMUF(`mSfp~Sz;?^em%L_`^gvqU`0yMfjmZ6 z!o{ce6HYCg-inR+(sb9D&5!c3H*NN-CIKE$foE9T7tZtT!2W+p%6Je!er3;FdJ6>n zqmi0~spEw=kFnR;37(2ywPw*8nwH}YaPTzIclm+0IvHy)n7jnZiYW*Jn#B7}Nmqc< z$r=qH{fUv1JLd#mJdUcaJ;u==#238)ZbC4(awB0Ij!-Qs0LaOcR8|)K$O$W=346Sz@ zT1}w~IYKsI?_4nCRP#yh-#R+8(#+E?SaRj4+Rb{f^s*y$Ro^UJsRdO6rLVyBuby`K z1{r_S2JIuTzl3qi3|5iA-A9F>jB~2-)Jm#2n>4uQ9OMknm1!EI(@?SAt2+hxxBN6? z+p-@=AqZ)YgLK;{XF(HlNjs8$ab6jP&=yH$O^qI)OLZWNhXC-=I(+xRPY9sA{r1GB zQqw48kXu!ShwomyGF$XX@PTpOE1)MNl9Ufobtw*ZS$`-gm_;QXfRdxJr1c zwv5>hjy@qyC<)Hpe)l+8@EHC2v_f;Lq0t4nT=n+QSi$q&B|XOnjwi$~a-Q@T?A9-Z z-bUQG{EXi-;uy4=4j`IuL7?tFzr}5a|F=CjfPlEzbv7-w1rsj(%qe&@JC&c?sYxieE;3INr3m5Ew(c1 zG9KQGkM}q#_p1Q;99)IC6;3=+9&9e^Lcw8Cpg@s%5B8%qD(5zOE>-vlH=L6SA)j5+ zwhB->xB2vID~RWNURGWY_4jDKsa|*GPc1<4c!XHVv7bBpNY-NdD=lupc{imFkkI|& zdA*83qG`9LJkY_RCt_~Dcp-yOYV@u^cqP-k{l&&d*-A>SCMo#K1sa06T;bD?tRXG4 zLdaCz)MO?uEPR{hh?xucDk1PyhgA*VX~0*RW7pjwc!q=_^&cA@35r<0Cm=%U!!Fe_ z@GL&nTbw9)Z*V(DZ-qRC0d8|Hj$D1HHJbzp#a{8fUv*}6Cxn0I(HAMBRwh1|c(m~~ zo-si&5N_|F2%E{{APrQUnN#jmRK343r^1ed$8=Hh z-kUb+;HAM~WtFbFVEO$%EshWYT#14l*>oG@jgfW>AQ!;$N^SwGAQqNdAg*gyS-k(A_+1Yzq0X`G3)f%YMlU!w zswG~;xV@4Jwc_EO9;}CK6Pf-|Z5$$S@9^^obtPSjR4n+_51fTFQN_VLiDWBJxC4|=y7xy{y_eXIsn)UCP z+bQtwd6VWBbk9y0xGxlI7Ufza_boq(UFh!Z)>Ek6N*aTWDyVqozQ2UFW}0<3Gn0Yi zR`F2v?XEUg35znge6jX)E`~#WgMMb6;lX+?YCz7lC%xbu)Yd`5?f{S|_#oh%;r&0{ zy?H#9ZM#3rYQ<7AESbk;O6JT{tjuGn%w*Ak$WWn3h7giyK&FV2$ds{UC}XC~Q%RXJ zM$0_E$EEw(`+4?$_TD}3yZ?B9zklwJTdv_e&tpEm$M@=np5TR9)6;`Kvx}cjHJCDU zQY?E{KM&cQ2-{^*%*P_uOHJV%8xSdwTIVptx@v{|TKD*h#+}VYT6zB2TW)^85 zL^dWu{VE*w5cDdWp<8!)=`FOi2K3`3YBQRT&wVYA(5(S`Uma6kx=HdfO=<44VPizW6lyyUlCIUI%4jdah3vT$JPzpCzuoYvBa$I!N{4% z@Xg)XoF%Ap$4AJ1?faRo0v^%%fM0ZI9E_=be#}0 z0R;(BS&MsUlq5*#TL^`!&o(3uiF(-F+m2~2eeVWwEc-d+v(Q2n`!c8%q2cq(8OkE;Y;Vna0OiXq z`Cm13idJ;&7M8GvB`5=Ug-kE<2Yl6x3e6c@b2N|XX!{?CP%*oihRCujN;CBlG5=wI@l{bfW@w0M#9%2E>iZONhHL?;HJif0 zo1DI@?|<_41{5KE19OjV}A`v#I90A4eBW7BCPm zR7K&`X4ELy$zB!gS&VrPmgU0U?a~%wTqOqC^q-l(3OQvho{&)$;0#qkUa(~Bu~6*Q z^WUt8X(O}`4qrj!JJm-KaF*gA4%zZ>J0oH`;#Q*i9S)WIUK_ahA`U((TF3Ev9ckr# zhz#=SzTA7P0BPM!d(6FCxYzHyG28M?=i#fqn_n*1_fNMfZ(I`c&Wd>#w4L!}2g)ig zM#c8fxhq*(G_`b=UVvVn2MtlWw_@X`6ZK*Qtafp-vmS|)#skPX^3G*R3yn6Bmp>P; zrj%sr9_%H3JDc#;rY-Oe!Z@^iEOd1c;;%4pg08yHJ!uQDR#=emtMWG;7_toy0M-UKao+qt80||FD$y#JBpg>kt{e8yt zp5@V@ega0g`C7e`@Jrs55XOPvN`QYn-Y@p9%~DP@6zUdVMU$`JhEhr-T{&2zVT!2S zJCURe;7DpAf20R+9;C{npx7VdF7HJ*8l-5@m(P9<8Q{xNDN?lZ7+5+a;)4=ATnN4X z{t`x`aV=NeYM&MDFIwVFtD_W`TBkLqW4m?{mu+$Z@1Qlua0O}-BmtM8B=!1xV+I7K z>u~o;$qVY3N5N33G3OraIT5{jQirp<#grMxnV^SM3?hXD0$RQ}@cC<)!YU0xvQ>Xc z%7f$Mg0vuDK8^)~9Cd%6XpHHPr2Yf3-1@hPaI*@r`;M*0*)hnuCO~L12rfr)ndXbm zL;g42-`a{5Dj_--n*=;2=EmIFv{V@K%SuWg^MydgH|oWtYNu zbi){IkPB?ko2Pc%Ai+nVKvIvoB@e$fK#dA9kV7uZ0IyT1EpdDRZN&il-g&S1+{op$ zvESenC@?x9Jy^UV$bLRA6axvcIG?bCo~l*UP#^KFd8u#mRutZtyn^LT`#z8duPDb)DDATZq$YC>S-8}MQ`$ei59anql6`Jjc;II&d)6BfHF|KZWa9sWzDYI zbN2=f1~ytS#b$UPYz+J;Mk}fSwUE6~5~9Z>zr-1$;Enk(ug{6D^TV1Ii~T!fLRhm4 z!z4K10(z(ln4D*3ggSN-R0WnWww!1#8YnO3Gr;@4f+Mmwc>GI9rIQR$GyM9GZQgp$ zI;QK#4ew191^=m#{eXds)TIzbDJXi33lUQMo^VUp`!isF--b@WF3KT#xf`Vh&|%b? zW4i|U=C#+RpF}GD;u*lKZReD|jk%ikJ9s74f|FvSK-+;bzJbE#*u6vwP)0IPQ5JOl z1-S27lx^r**x>jrHq5H<@P2AlU#jqD8HY>T1@9;=03s17J3PRq0Y;T(ALxm|{Tgaa z5R}a5eeeFFL5V^g^4eh*+rHs4kU7peg;$e%Wi0~8nO3-HXkaI1ZHTAr@cY?KVa5{7 z$^`&-{kvKDPJ?fzJ`|&I1A7ZDCM*gzUiXEH9$E7|yq6;g?8~R;xj-n~eGbm(i&AbWl|0 zJj)j3zW`#IM15b9>^7p;iUsff$BJuI2kRq|y2qkz2ynYi4%D-*KYFauPy=ecRi@$u z$~}LuRuc38Y_a{ZE^eCF$nmVc1!FQ#GQxpIaf4C{JT8T7j+Yz1trryD3X2GUJ>UDm z-Vw~>9sqeVXxu`akA8Z+1@Jm>;>e=fQjlovg)n9aR&~PzDa=wFa9XNIV>DWYIr3 zwK2GmP*ECdfT+Ju1h}v3{+Fpyq(~>=puQD!0|s0ba3|DvNoml6aF9G< z{bTNeX~>q=bY#z>|6bzp?`!BGoq*y!f!oR*igo)S5EuO2YzN@^c)jMBbN%Z(yI!;Z zx7R%VqFlfv9;JR8A@mx}`VDpu(g88Dv--6Ug#*JA%w+7-c$dUN)y`pr^6q2KkO#>( z8--zuH{RHtvuuelxn$%HlClqK&w_F!`qUusnqc#oy!qh`uy6xDwC1&4uNS%&JpAcA z@ilU;al%5^kOVNv%!UG_Odc8zLICvrXAOtO<0MB^66yUzcl3nI3JVJ-f;lbx(WULb zQw&Ivb5RF`U7%~>0on2tcj_Lk7Q1I1c*9IHs> zaBvZgfFxL{xU|SQ8Ua~2vNt1U5AV#SZ9u^Fym7W7d3BNrH&#gsk$b#Lv2!7&tNIgqPC_zyZv-UH0=(fnH zW|pnH6R7pNPY6U*$i@1j^V2Da6?@SkzNc+-tHALRXXO$#6r1MV|{V zd~uMDh$00ik_ftV1=S)2p~Y1QEq)J^1h9be3Lk*I=7q=p)S#oS^wp;!PD>Jt$z?6R z-lnR?87hEubp*48fIjR%il2Y*?_D_w>M|mpe9<>m)Hko+1>n+U;%i1(3+vkl(r|{l zaOtVr>@Sc@dq`$`vJ)l;in5@hy-y!*f|1kn-;lOf3$z6*LI=h!!9TX+dFbk|W89wI&rGJ4QV1`(J5-Re2ygzJfFeQ4*bSy%gZ8Pxt%KlFI+R|%BDb%^W6ave z>t5XntRlG|g7awvq~V}mR!1)E{zTaQY5oozRF<2qF!)`>A*k9McGs#oK}CyGk(YGM z`KH;FU}!6V)&+NEyv|2Lf`OriwcPf$5}2)!@;=fFa(4iPN17x;4;GONV)BDgRDH9~ z3|bF$E#X;E2Ibagg5blDCmxwPfdb!(6Vm!rLwu6Zu21sEtHDdNeu$aiRtFIuaMx^b zO-x!ag-Vi7s5s0UtcrnJ>HQadJgzn-#I%>zzS&2sLiK#$CJixRzv$I+wJE_nyE9QiVOjsN_?^JYPGer+|X-uK;i|yW?Ld2)6bKrU*l!tA_03B?R#JkhzQsQK*^^kaf2SZ(&_EI- z6@CjZd<>k`aPor7UKY( z%L;{r3*X=N7TS+irqfD5*boJoy7BRpI0&TYZGl%=tkjjp5Y&5!nDMNv83rsJ99e*M+V=Dv~=cKy_94ICxjubV3JXdaDr9S zK%k%+%B_q~VS}w03zf8n{fVFzNP8R^I@D)zaa<5a@+RbcMW7Hyj#+>!OyH0i!1Wly z50$SqIVy4AIN|X8j#Qt8!y}+BU>9hAEu67zzXY=&Zl5lgxD@`L1;OuwCmhC~$AF~h zLAWVaeY>4TfDiWXaP_@TkQpHC1q}Z$I?+1^%Xmg#U^<&O`f#~Q&dp(wK~PIJfhiDonkA2*3|`r z*ikhi15H&>4;X{`JN59_L#rZ79W*})KC6b?Ll2xZ`{6yX_socsM#tEKtp5Nfok&ef z2EW~bCcxRluzVVayZ)cfDs%guJMw7?&gNBvf;bc@e-9%Dn86O`i?>rikunOBs;ZP> zEE)xu?vu(Zg$c}2)hewcoU|&j&+j{6gTRt*PBllw5Pp5Iz&whBl{UB6U&A~~0Hyr$ ztNsRN6(Ym2(}j8TkZ0ZZKXJBVf!i0L|KZ^$TYLC}HZQF)C@9EU1P|L^frBT7b3~cB z2a}+0L9n}DFi~z#Knm>gEh@2A*m`-|`D4x3tZP0UHW57@7&Uj+=+x^l*6c@k2})2R zh>DEIL*rn1Zjbz9*K(gJrm&q+eY`i|3+8%U?v=Fa-C5XSIyZATsYuRxZg1{U#$!ic z$iHcCq>$=_eKCMM;J520JptRMbb!rH4MrxxMo}t=LS`99>xP{2C*dzI56o0VeY%%? zI}k#;-!qX7$l{N>DQ{k}zNJqd@BwtY3!17e19 z5F35RRgu7E^22`U8@eNddr?vN`^b5Ju%Z9Lj;s7ea$f$Ztc7{U1wcA;L2V&R>Y@Y# z1cG3Pl6NbH$oXaNo*$me5~YDWxxY>-ydAnPDN_xP!R!5wef&vhuc63T)CS9 zBT^%{D>Yo-4hAUI!g61za(-qC+U(2NoNkDS+0}9+am(CI+@M}i&jQx?->R~Zdg^mg zgNzLg(cg-ibDf#8fIQ=jx;y+48^m2#cxfI(emV-Zl~~0q{18sFfOyLE{F!eQvKFM|jJ;emOz^%oq_ym@zcTR5|CBsDiowHd2e+?T6B$nB)%D_E zw^_4yr?U~5i{LKIg^LExQcOlr0W82?5O8`nl60_Gx&|q<$V1=p1x9gN>suZj{?lNf zgH9W!afd_RSkVxx?`AFC`O`;o%@eY7p~!BsFbyH87r+)~lA{mMs(g8VD~i+Q5-oAh zJ&~MaTv8~XBa|Ykt4K5gb~;?|jxT6OJE)SK`mS9lc&u<+1~*`FEgNcB+PA5Eg_ zn{j_@?mOe)TaS^8R#ivpq(Q#?#cn@U$EtwieFkBcPFOo20|X!@&lDgwBjc+n$^oKh ze7&xc;3}1cl5cPU=C1sYF?X;X3ON5L!o9J8C{DLTC?xEHs;ZDe!hV6Dh~|JXJ)4%w z1F9wt0{x!U#v22_hP)x4g!^w;faiogFbyq+l+^xcC;efmlfCetDm_3%{h82n!E3Hd z`vcJe_S+wdJ1K^1G}iY39P&FPi~i%K{Vy%2vC|(JFu(4oXs1?13jvQ7Du5Y;3#yD+ zzpIS+6misiAk5OQ>zMv{A0d9HAvWbJ>EcFXJ*v56Eyh;0?#9w! z0>FX<2@FG&2C*KI@Sh63crFgopv~)vK6A$Va6>Cx(vB*!77-6d6u?-y!t1>YlZ1-u zH^@+uh3;2Tfs}E;;r}%|j=u~K4L_IsqZza2)zRuD3IQHjt2?Z$8WjCN+K4K|I#-c; z;-1ho8Ucvxjl*t5WI-l<4K6I-Z>~lLxnEFAWH{u0Jjcx-_j5}pN2=d|==#~)$a+I5 z#NQQeRiQz71FV1q=(w0xA!1{oXZB~t6`w)}GCF3)0diR%_bY$Ri}?GokA6W2MhEK8jlbyyXe3BIM?=^Kch+wIPrBqt^9|=^ ztrzGM&3JJfieOw(iJ%Nu?3AqoqUO$ZogHY7F!}BN_kkG^R6;X>sn>HRm1Fkx%_UC#DWj_<-mjINZ9sO-LzX6i{ksEqOcHfb;bkg%+zq0@p!B3h)q-XXOCf{LILYRAA4|X_v#aHwcof#b7&L+y_7gU zsws4+O&@%QlmWm*rBumaSxVqO7Np!B9|%AyYtiBprBGnLu110zN-&hZAfO6dC|@cb zvx=0A1bWa2>p5o*R;mID4{O-NGZ1DpYJ-7fI|*e-xz;~~ z4ui=(j#!X$UWl8p|6`^6XT}XrvnFWA{Ip6n1aU79(YG4>3?#lm+zic~Q6ZJ#plJBD zJVhQ)CEMkH@{Uq36_wCBx_7ssa=NoyImLei1=;NZKv9K)YxD+tMPD5t%QVLbS}ui~ z;mizbNKS9Mo70nkGWLL&ecKO-n%Mmh$3da*B9?>JBWQE^GL?hr{~Oa095r_Xt(c$u z$J*6K0Y>MFFy!?9BOu{_==eeTw^+b1UZ?nTA5TNnt0Y4(?Wlix))GyAq)8hGS^1G5 zL!aV&Os+&x{Cgxkc@#+U+(Q@pBPNDnaJ&=6TNPpQAA?0eEp3|aqL%)?@k5iLiU?TK zE%txO;Qaq)Oa3#=A2c;+vm+P&;~z`%-`ZV7r$O6jGLHGPfX!vuwYg%q{~JsFw=3Xm zmyolS+l09B-^i0d6=(o>4a;VttHA#ET-3Noc z;ED=@x}lCkJ2VB6|3zLYg$0qfSNnJ6jj?(+&iLEE!R!9rkK%QRo-^<2IsEnCIFkNa z*Wdjiyl~6z3wOaHkud)s{=}XFfi&B}41dvblK0fhab(!A$AnVS4q(!Dfc#^q1@7{* z_vFW!@{kRK+NvV4e4lA`9rvj0-PW!Bw8X(kS_X*r|L)~%X|Ds)rv(A7`fmQx+GP*` z@J&mM(sh8m&wTO&wCi-9Z$-C>Wv!?0{I~aZtMlKBcL3z6gzJ%E zbdNI@09$TBfEqF8cFYCK15OFsTN?@Hc5%xx zvn&taWjgYK4LAUn=aA`AFL;#}?r`B#u^j;0cJN-Du-#~p>eL?x(6%g0hiMK@tIpDx zHn`4J25kYxzN-#@>*j?G7_e$auqx+XSE2Ts)p?WPkqr6U9?~Em&eX1V5x{sM>F$XR z^7&mQ|4R&zSH%0H5nKxb&z_F+^Z*3lIQzRVE@iXLFuHZtuL2PK{l51}P9OM)kPm2V zz+BP*AsKdU9KhH6qpW>m1z{Fge7!r@xBL5WvhVsweE>fy8zGrcGcvC=@B`qUl4`dp zC*2~xrklaYJUHwKxQnoh7@ z@rBvo7T_FVY3gkhD~C39e$G=!SpO0*w=jQCH$d;M_dElLc-I%;crb1O1ns9?S_&S!XZV8)L1mgxw6c>C3a#ELsnD(Fxvc_$pG|_u&iB zCM@TA3tR!(k!x-m&pk{CH$6744h~fSG|3LY(P9G^#v!!fYCF8^X=o75zYEbA!WTz{ zbr2i<1YcG4G~EV#kX5fm(B|(j{Gc{oQ6BjKU;@mQ)mxm*veFcxt6ZNFDj2>aRBWBR z=--4ogLG`U_yZ8g+YF347GFzNzS#^IM6;Dfz+rvvaWXwE;yEa5%we6yKA-jKweIxO z93Y%mOt#zf1UnuZ>a7i?2zold@yemAUL>1q`z zc~`_C_P-9^9CrUxyj#EsjMVG`BUDiU>O1Q#Wnj|C{#+eGEen$XZZ9wpwB)JI)1tQN zCVF!TSq+|T$H?Qr%Du=r9W7nJXnoz*GSwEeJL?8JvW0$_iEQ?*vY1umJ_!ik|0n}1S$=w1_N-fCBFHd9#{TwW_4@f?OgbrXMq4)ed0t@ol?V!xZ`BRnNgyGDYd&aZ; z;f>l8^p`12*CmXyYUznf8f(ZHFF|4{Sa(N<$r1*TX8i^x-?l1DuimBo47UuS&`#`Z z)n{Lp8+7lC)8g=|$>8_QUQXWGtmoLbHXJTCAc~;k`hhcoeVL{c;#Ez|U1rJjrJ-V~ z^l}(F#U40qE=Hh68MSJ@83s~=JZiGV^t8sRXM<0Ch&n=Fn|x>WBL}d0!nWJMBar}B z`!h1LBK^YHg%t*TPTNi5-mC5xd361*=Cx7Nm>%`44b90d$GCX&4zCzFAt*hB(q<;td*$a~ z;kW2~CrT*HOKID{)VLHF;Wy>pK;Tl0Ht6t#k9fqxpjK-21ygIU=lwTV1|L$H7VN;- zl#{HL6`Y`^E2VKU$5HRT!4+?&rv>+{T!3Lo;hRawO6H_(7)M@@WAqvr!4lb#x`?aH z6vaq+5*X$oA0THt|5u3M7KJym(2}+_W;Mxv{nN=_@#Hkcn!*kcj6en9oI>)ya#2D7T4 z>#vHZ7Xhl4uc+LOj4Ui0>Xo-YE4gcQ)pG4sLHd~;>^g7rr1)24(8ztVqgJ@M@2n?z zE?T}j?+JjU)$ozjG;R<4ThSdk-bDyq%B9lpw-59qyiQ+TZiiZ2XxY3?cIn!EVmR>w zLO8UZfK0I|3q<*Y<;!|;#6a{MFFny?qVy+ybV2tcI{GXa6E8?FkM3yOE8eXOrjh8uI=gh*I{I!OkB>W%(z#BtC>RCEFB83@^4nWR9Zbn z&5%VwL?%MDIByoVQ}tUIsJ_d|jZqu|;Bf+Vw56g`WA!zl_AtNlk*`zp{dK5096@II zvRx@Fe$}x$Q=K+IX~S3hI8`c^uoN@ihOs7{o@sxdMx`C632=8~o(&ELx3>Y_PKBJDx$Q$4#`;Dz) zMRf=Hc(mC@g|J--omF)EhPMnax;d@iV5;*3VYspe_Jxu`PfG(L;~Dq+gD2BX<~%Nb zfN~)|g&CEm8R00Bd~%KdL7maZecdr8pJ9Nb1*EvwHWQhgOT1jiy!>OJJS$Ly@72X? zuUdCjKM!s9#ET&l?vfWfp2>vRZv*;SllpjCNFsKVw;}r-v2ocpU3WX9)ze5uRdLh9 zS|?iq-;(zYdOt~3$fs|p-x_yX%0)bSE)Q{qGsXL5gOIi<$`mP!3{_vOOyhlo<*KD+ zDV_r@V6U0K1@`A|t8XeO#u$G}i1&1EWyR4FXE1ift}6(8)!6Nhl$C*l1w9r{aV1~9 zn<~OG4Q_nN@}#1}Rlctsuf70l54Ac|$Rx34|;N1bB3 z5k{Xfr`Y($r7p+yTlBLQNdW8~VEe*Abx{e*CRVtV2ZZ$xPt%p2lZOqwKYYxZpwGMN zP_lhWPo%qUwDGl~H~!@plUcncibIJupEeC2=H>b@Nk@;i$7&Vrp^~dg1b=Uwp9o(O zLWTtMD?9u72ELyF!_xuZJDmm$YA(BBS&h)OoApT=A+1yIt=vnqi60?qv@=o=YN>W{ zitd^0%}s9M$;vLauzI+E&h!mWrQ|uI0e$&5H&%Ww`x;T3KMYm(emDtnSn+9XhsNQy z?;mW>LWB|(>fiVToUw<@4;64t%gQIHH{A)>VNB?A{&9s#-m8EvuV+~S%USrm#R#SE z>OHmj0KnHgJw+PscM9vguTB>n5oCOM!^C+i%Va8_C~3$_vSgEBl7N&|v-UvC7-5HH zG zSx=`bd!N=_X>+`!G^vz2$1(6^~Mzpe-Ch4AZ6oSg12cp>sKEzn*l zLA`K-G>vGySbR!^wMcIn2@xtQKoldWs*O7P6d2VFTm>W86?ZqcpxJ8%=$pw%k%{)qgz5P>1Uz@z2xkgAIrEN5ruK( zaR|H%2Fr8E{fpu`EE>n{;=fvS#PR3=bxQ?D>%^Bm(w?J-BCuOH8V1s&lBXp%U;sj~ zEI^g@@PKTR&a)Dc?nRk4AQ5x6gAv|%u2!W5{i`*!T75&)701b+lvG}Gk2w20Dwn(; ze6yD^-I7;^sJo4KeF_4_zCf#!!^?B zDog2d<`glFf!U$B`IxVok76cgXA+#QO(K5PFhb9n%`e=KY7BD20Z)D?>~KdERdTHgiJ zeozfQqr9C>9#x6Zx@6?`2+HOi0D?1tUQ;I)wWaS%Yfoh#7BXkwzae*XrCZJ`>HEpU zqtj1J?8UpC1m(N3OvOGMB*%I0ImT)sXgSt@T|sXO40o>VrNe#zlXUBY5;@X;0b0_f zAfGs^H)^(l2W0Yc!<_2h!_<4<+q7dRMuZZkCQl^FwQp;a<4x%51UGVl_t2!uG`t{i z=-8vjKv;cU^BY2gn0s7r^kgTl7-EO(#$tt?f&<3MT?XYJpUK9jrk}CCF_W(qB4*ea z=$+A@rbS&)%V24gtrhR?vJN$YB8m*9n2H-OFRnaoB=2%+WEE5tAl2eyazF75t1z2T zvxK_#|j(9tNn83L=j89dH-prFG`-HS?cGETJ>3V z7OMimEIoSYUqP7K*3JH^NjXS86(Lx@^ep{`O82MSOWxDELm|oDQ-?Q3phA-Eameq~ z2V|<_rqgNiw5&@JGS|-cv(OSRe1fqZho zI4?qGAmeo4j|5lc8*NrRxg{^uZk>d5WqPF=V``I&1w<+Q>TqB zjW2jg&wt>6Vs4ZjIc?S*TF;*DSaW#SpM8FlSA+uU{ zjcNEoPYE&PYsNY_@8;BSQAg@pg!Rwz`vGQ&8stMA%X90>ujXnaWGL1I?=wh8Pn^71 z2w&Kyewmov@ij*?coPO*n_R?CK8ygTb%H!rXk)@zoPGPalODwV-{_Gp;_qBLR0wz` zf;=fQ(r1t%eAgg%p0q)N<81KeNTkYFN_K^=?VN%$IqT()lP?cO#y=TXfsl~u!3QAl z=2Ayp4&H1u4L1%%-t+i@5YNadz5AYBQ{jfsCwSWp6+cWxAJS`KBD;kT!(?n48}Y0{ zIw>^})MgF?*Ml=piya4wUiuXzC0%|RNN0SDOYk|pb5O&cd=3W37cYKx@}51wD$^Vm z$h1I1tdHb$g>=IC$nmSsd^Z<8`&(Vp`5bn;xm$_(<{_Baj}rwIcka&PTw~{Kmo1gyXGnguuy<-tHFVE{^Wm zRCkS>#73tz4`n8s&e=4;RQ|_A@QI#m?3fpU-yM+PlgI^ivSc6_HTRx;o~LtoZH3cDvu*AJBo#@afcaW!7}Xn(Vn!OKJeF6r!xiA7)*A7*}z%&UwJt7oqX_xqh0Dw&;!O8rgeE`U6uYQIztLS=3cme2Xw6< zWbZ{;RcrF?JOqe~se!!vxg$bDy0p0DdCJIwN9fk+7src2hBFV(P37ssVxmMzXKy2Q z+}V^<=?WWy-PvpW`E;RIFL1+R%#=$$fPh`aO_y+i$$c50_l7q4I22YsgjF_y$2< zx=BIcG|5A*C~{JYZ|$s)-bQw88GT8=YPiU5OddAv>>gd#o?z+N_=B9y4gcjScBykK znS*WO`{hNMSh6@_>q^%Q@p~dl+!PfzUFZN4U@h#Md>t((g(GiU~RLMVR5PXca3}uhKJ=`(vxqkND zknctuLPeS17cut@1$6o7<6JpWVw9QpbHxrjD&(1jX!kfO%Ay6k}&A%C!NL|4(* z*wvUk_vI_C>}Bxw{bwo(zH841gP#u?0_70bO_Tx~eVXXsumFYPf|Dx(|%0Z`#`=)CzVfh1e4Nr@uEdGE1t*bfG7%HTH;b&lw+XdwQV2< zb!eJy1_A;0`3%wGnrmY!?1cih7lA0ba7zZJOs~$BuM9k312z_imZa3^CG%)j4?{jA z;7#^-q|0K+A4uL_e9B`4G^@AoOTK=j*!bT)jbcsFRMEZ5-^2~O29-MrlSvKb3md)|*r4K`k3aYWJvEJLs$Ua^my{=iPh2mqKnF&9~1Z*`uEQ z8ZSrAO>^l3hF+`A!=wrC@b*cUfYD-E^w*9Q6n z=5?Is%T-h>=4fYXeE+f0Bd5XjGD>WOuS>73*Qu^3e(L2t;1isnpIM#-Dv;t_jb0~j zIdPXSa#Z`X$nOlP0pDS?Yt;-?{WbkwtZAhXQ#9C1`^?uQcHZp*VJS$%K4PQ0jaW{u zSnI^=re;I!zL+gFUYF1Dw1qpT+3ZXS@5rIatdK`|LZWgbJiQZ+mCZw0 zZ`O?86=v77e<)nE_ggeke-*xGE(qMK|6f7gUyT=l>1hV}$ft;TA_cIc(`vagS!7w( zhk9DQW_unO=2lBONtegDL#)y_u!7LmtbGflzGOOEh4A>YNDazU6gil4=`a-YtXJ$5gmm8UH>D4(3ckDw)Ej+Y7^IJ;81~l!m)+6!#pvfA z$V}ecxvU~+J=InmR)Bco5DOvcTAwYs)7~sY$-nhoN=5F&r;g_yeYIhMq2{eP&s)ndQ+_ioAEp@^ z9{GQL+-2_OsH2-+j!eQ*J_B2lG1a<(un$_nc)q9AY?lgfU7h;HBbc0ahKN+8*iS1C zEZt7m4Uyh(lcxrDLQMk*^9ObiMr;@fKT5BRqXe*?>OrSf*wlafgbv~rkvDaVsha;@ z?fel{ghUx^ew)8q2kzL`Ate$`KJhMupb&Bhx!t~Ho~x$17}d+bisQ}+gdV+&yd=d^Up|4LeL z`9~|n6Slt)`+GC+8-%kFP8NsPaZqTr`$rZ1hYj6}B8~0plxN2NFg~rUP*^NE3M?(o zH?L=v6u7|bdgh>UMV3}1$MQ7=)=5~t^Mqk96eubUJ|@2zk8eTGW?>m;iJnP>b;lKZQUe_kcqbEP4BeP-)VGFnx*6TPB42_Z#mKM$s z_!NDpY}I?&O{F*^j<8Mk6S5Lb-K(6iA)a5o>?MJ+Wh3R-@D@@=8V5EcUNzpSgrcaO z?;)J=Q1I#-C#H=#n_2%nCo`s)({pXFgDU-`ze zPLYY1O|ft8!YOb@3&))|;LbBo2Y^iM*6cOlT8f93pX7ZNs4+?a6YM2}GVB3{w5vU; z6|Gcl%qYiUU!W_k$;(ta1AA~!`Y&PrJAmdtJe~xNFr1c2TFt`uuf-2NxqenL*Y~B6 zk(bT#F>?0vt-$MK_EoPY_OaMWgpE|@A-{nH!nF9xDSYBlQB5bv{jbe&wT-|a~PVd^QxM45^gzr5{4Mv;_n9L!P? zUs$?9g56v9N$jV*2aBo=)Q8(EK zl7K>rUGLL2fY6}7L;8_G&wApj}`CJ=WNN3QhSZI1EFk z0EEZ>6*v&jZKPR`@P~_Asyp zK|MzXVw@)~v!TAi=!h4i1SJ5CA?L3`3>ib4?zKOMz9!L3bUY8=a@Nz^pNQh$^%DOI zB=~;`H~ZgW-~Z2p0smjUoDxL@;#}&?-E%C(9Du`i6jzQS!Z(9)pCmR#XzPW$=>9RT znJdeu_G;d(Jxz;X_G!RK24c>gIRx{y+{6+!6BszhzDE#~#(?2s|CwHIOgM1uP3n1RL6v|jbgMH5~GOPN=OR=+tvx@u&7LR{A;gh1Y3#N1m+6=n9{L&Zr zizov1kWp}Cq~JrFJTeYq{CgaPD@qZ7Dqrq1O5orNWPJ;9k65#PnY@5V*S&lK}Me>*~6rOJO1vbsab- zdN6Ln0NMKIw%!XtwDR=)Yq0455v7>oxVHUf$q;-}^IyoszZ?6R-}ww5@zi_p+IQtKNdc31fzv&KJKX7m#^PsN zFWVonVK7oZ_?Wp$SkQ(csWynY_O7k{5nB=s+JbO-)Y;aRop&vjKbI3f|47XFsqi9e z@cbj$xQcUcSr4dyrv~BEA>sfIJzK!s4ky6ZXr9^S2c?pc> z_O((Knq0EV;tz`d44rq=A!{yEJTS?ZcdObV}YuG?NM0Sv9`UTDA zklj~mB7zBc$lqvMvuJuOzOeg|Kf0BsZKq#=mn+~oO&dX^_EgMi6gE`5N|Jca>q%Yyb~wRFV-LJh)b!=ti{Htp|El?q+3JzVH--Y zkHVx|-DX$59)b<|Hjh3?r2bTT`;fqBuTAit==T9V0afwYXkNoa|LshF$q+8Ju@WfDE{0?I` zvBwhC#FJ8K@dWr6(Y!Pv{;Ix|mO}s9vDuK525&O8sQrr@^S^O$1}Guq?c{82;c`j%Z8O>YFSrD@y}ATeMK4QAZXqO-WBjnH5i^lHB{8AXDfMGY@WJ!@RgNI~ zCQcSY`~lsFWVb#u(MDm-Xh~H(%ec?|w8)c}m$$8OvA`jBkjgZEP^C*Tk!#{;mcqU9SI>H2)MAdpW+eD;=elf|DS(7{qSl2h1JQ6JZwe3+r-TQ5ryIm{3SPfsb1=4}3FjfD zxWt!IIWQzgY2E*+L5Md`9rTFLqD4XvhC{cE^~)@sKrEPH#j&<8UOqQ!NY>38-e7ky zzQh&|kG0ak`&|fx9ZAKu?PwIeTZ*Y7!W+lfOX>-eQq)E*4j6+96OGQAIeE7)vJ;fDMdn&!6W&r*eJXPSTK6#>Qpd#S|C4) zf-anJK>A!D>Li~m5wpt{N>UM5Xs|OW(n9}1v!uen4|jKq{JrcL+*ZKdN+@7^+* zAD@y{I-jmktWEkhdaM!0I%*|V5KH<3?k__|q0J>oxEXZeb13}|`@t)3Z0q8T9d|;nbW>lqHS_kc9^Q*=|U<)>-SpN*4 zi^7jscR)9-_^Pq%_fX*qFkTcOczn0>n7}NzgBIPp_k|o>0T1k>=B))8DlX9xPZM_aCU7D3}9jez^Ry`)8Y_W0|8S&o#({C zJ#2Ub5;J_#erc)ZEXo9L>K>3xHh-vc^KaK6mJW-crzC8XdS|KUVJpM z{@a2%v)OOp;AMYbK3#mlW%3~LYM8OlRvYqsA(>Q-hj<#en(KCPJgkKowXVV%3>$k8 ze)rt0@e0TXEd^y*qb!TxO%ZT47i8G6(za6Djdxu0TIw=yR46eF(%Vael z^DELVus95BC4m)?Xj^lWQWzb>`xyr_3kM-~)*4loN{px@*r%OrAUr*DO~ZgGqEa=` zl8-_d9FXJtx`GK<051rH&gwhZN;z)uE3~#70*G8QqdhgoSR;~2uN;eoc%Z>dE&Q#B z;|GV$r)Z~lt))L%vd_k)gJCeRcc6m1NWfxZF3ReO1cbs`XtQ0C;BcSQOoSu*aJZ*Q z?1ieYa)HK|c1C0`xb^LEd^F%E1QS1I=b?G$ci{3%``|vO5R-lOaMNK@3U!_~M%ajL zZ_5DHZC&RwX5>oU1I|Ufc_`RLdZuEB|Gl@kqc zF{)@^t-s(m^8!3w!YXZYY$`u|`7?f;IM$@)^E&bYVh^?Oe$n+!S+LqA5asOpm>zYM z(;;y6Wx?6WAQtCQ>SGUShz8Sq;<^n#_O?0#SGuLE^Uj5(dR{QjhhVHD%=yGK1o5@C zGdWj%mDqR&cCeV5LLcrD-|BXcZPD+l;Yyqn4_MNB8cU}2-)Bzu7=Wp8*=RLZ+*+_2 zz4$7-~me( zdON1M&cgi)uz`*JC;g1?J(t%>0Xy30c_hYv+e=t}?qQWU>fF`>7p`({{}Q99m4jYP zYFKAw^_^*BY!r!@=Ba^}4?klHJA*4bU(zm$S-%r_MG0p>X=W`L5CG5EbF9TDZrc?C z;8kOA%ItNo*0ZYJpf3aemOo3#TwedNtAizU0+Y!r`_=}u6mWAuYl}aaq!CC z?w8>T7@=jFPA=KEc3%6?I05ZbQ?u~n6KAfbvJUD!g?pvKan8{|z9-2;LjsGY?-j$l zlaoLrl<6jR#?Q??QZqcG^<@2rM_;>za4i0w-%ontRSAKB z$!8biukuMqP(uPxHNL+)&9jwdwGT1@gXSsi6ef7zBaZ?lJAQ}uasqH$-*@X-TX)~7 zqUsTAMK*TfsM^ju;Tz@yOIW6<(iAAyaxi0ak;b{)32{@j_3hl`(zEE|X# zJmNJf$-rV6XLe_`JwALXJBj4S=%FJ!Fw>D$UaECCd3#FW@y$^5YjA6??_U^7*=f=o zKY*3n2d``1yq=XNb(E7k;1cY@rSn;3$9CQ+JhXTkP184)oQox%KEyu|yHIKXHuWgl zEUpY{!2<&Bj05q@tQ>i5gzipFkuWqC42$jOTibJaJ}qZZr$OR!(qB}@cf`LlMTJ1n zryx8uaJ5&Z*9F(a7*{2_<0CkVlufg>wYdj+BONteBznO1xsN;LmKTd~Xw9f^krZub z_AofU;dN6)?rjmcJtuC}VvOPgLt?Fc+Rym4aa=5#N)7Ml!IV?4vYmgCl(0T0tJwsxDode3)fCYMaj z;$!gok7%Y2#~@tZWU?hLna5+I`F`XSrW5sQLJ;#T*0pBg`;uvv!v_;?nt+Dw%)+=O zzvp4`$!ID)uGc%A1omR*xH7aZUFb}%Zo#XgDqB});oCIW()g-$KbPRmt3RwuJ!kq? zWf&=h1Mvz%;j_}}|4(~g9uM{V_A8_jVJH!@lTZoS%80RLixwnHlzmCEWf(%po;7Qc zR1&gfX-2jo%Vf#Ukg+dgXDl=4p8B5WdCu?moZtC9uixvO*Xz7qkN+(5ao^YdzOU>0 zT=(^U-&2~qZ$n9roG!ga*Z8264^~qD{ZP*Zb<+y$6?98DVGYQ4YeFTmX+Md8F=EY8 zh^pw8ncSCfd^(6c4_LH@ozC6;^5br;k2_~*mdW=r*7xW%NcXVt!~4|GgIf-byXbCs zhXFN)pQVdDR}|Azfp|_CF-R_?|9!->qCYmVGo9HmTvj}xM3FM!NUC24c{(X+TB@jF zs+6A^l7+OaLE*p#n}69LMvX6rVuO4a;&J*!5tc5?BwG`|fZdrP$@@g19|o@$vQMnJ z2(y};2@*H{B=hmRx}qp92&2kX(8OfGmjl-5@m~|4WC@=IYlLwyQ&y+Wqla%8=W+IB zWs|DB7)+JvH`r!(XFU2rs~7#ul8-W~L!G)!;KvowND>#nIYp74boOap;p7m6@;R4x zWO2RTeW8ckVD@0Y$-A`WYrT@83|yvRXuan*^hJXLsLjbArmRK#z>*U9U&oBpVQ**x`Ks?#kE5(YJAjWz{3k27RSnMEgOPYE=S$ z`^;$}mO`>d0U^Vs%}(c)%%MFuhaAtTeE)^dVR`W?OhLU!tQdaAYW)sG}cWo=fv&!q#k!qH2zxMBUv2?0*1+Oy;6-DI? zs((HkEMrr^84`Hlo%R}hI2(LepSo+;t!%e*umVv5*1wpXEON&2K#R#S^x~kWDP@aPg-CBHhmhP{Zrvg3P0oz6rY5MAk+=#`9u<<=l3>3tqVpp9 zlCrl7+fIXZR{+NF05O*KtT4-xI(H!|>5>w#kszN~W@~Mb@P_UmimQcFm$Us@)S=<*sRR`!0AHrU(*Eym zr(lHSx(wQ!O`QE^W~mF&8F&IKvZP+wr~NWF^Q~eTFzZ)w*=lW0!q>%u?9czQ-z+Q} zG?k!_60jokgf4A}{L97=!y7bAkFvgdsv<)^S%|tNbCH2RNEO&ARONP>Ub4SI+FJ!< z2X%nw-(^1g#|5>@)TZZuowy0R<8N$rv4HPT=&Y&>c)NkT8Kfk)e>TdGH;6a1Od540 zzYArz?tkLoLi=Uq3wg$k(4^p*3z4?HNqD}gX|CF z?)c9*n<<+CcQ%03VJj4eEt#cUk%$Jl(@kOg%<&N?kjM_x!~LL|1w}6A)K0)0nYMcW zvN8S}8UJCM{6l0!Zx!}lVAlb4r89WuK)f*PyUahnuWyV5!tEpAcRlPgzux|*+&9=J z+;=R*gRcEqQMbH~POy89AKD|Fan6@(bvuQY#rb#3bX3V2KRoOi`P+dV*6Tg1Rkx}U zcj*<9$=bf4>KgE%1u_Lm&o%zIzua=Wg&J4jBptSra54d7l0gde@C9_&yprAU+90^v zsg7ipW^Yb(z*R5aa}bJC12wd;;SAbZmJJN^as_t{mcL&2{Yf|7!+&V!VCG{CUjl;uRUM2WO|7n&st2LX06$M^4~E3f93%HSC}XIwPFuOu(G^s z{Jh;`Tk*Wa_>Rfb+g~zGYk!)3y|tc?+|BQB+i!N0DLNP`%G~?1xFCkBM6^|wVfTm8 z2gB1a@;WTt7`bWeX0cFiQ9ul2-U*xzR-O$`QyQ6I8YDhxKu`j2|8*RCw*c+v-%y%; zus9t{-ib}8&?bDRj02`Bcv0(0Xb*~(fNatBlOj3G+&e7^$oc$$55BS6KC#ky{(~w0 zgULA1-7#Fdynu4`An)POzM1{LnS)NAqEbcX`p4V^Lpivs10c@JZhN@gs8Qx2(^5 zL2jhxgZIAyWNEzQ{Ik5FN+pzUuH2K{X*uM&oVTN~!?Gu!g3Ez>yC=UnH|`(IAW1WR z>&Vs=M5|W}na*v6uHz(f^`xBJ?pwDf#g9_m8G+L{QtV;eBD$Bo6OnB{w&jBJT1bcy z#*=a61IzUY@Y9Zjtc^Dn65h3wJj1QrF;k}4C%4xLWD;Qs=Pm(Tp@IE){`~C&7L)l% zABoYt{%s}nx>7oJza8tGwKp>JQAOoIH)rT@1BkOGPv^j}wPoyh@ z2Naw40+1lut~Tkj-4{WDdo~W{u8{Y!mQ)oY; zqrdipJaXF`*}wbbyS_@3oPUoT&tt=TqmunY;JkuNwZunHgiHX)AaS*=>9yJoxQYl( z4C@+tu?_Vxn%Y0i+b3*V2Zh%;mC+tq+Y(tu8$P>#0G4H%0X5E$MtLi8TfgMi!z9#0 zI8t{%`7pi2rut>}Z!Z?wR3;5$pWHOtj3uVVI-K5kb)-Df?tKyeG_pYsp*F0k4KHUM zJ;gtP41dKfh)0jKeZ9L**<{A|cERIuUh%9!vn8?q&U?d3Uwf2f!KI@*>pAClb0Rey z;iQ}nx%Vrwd$7=b?YE;(q=VQcc34Z4G+M~zmV4#TLIYi`PjbW)n2|dTW`^WjmYcWI zEf0QLx(GHWXGOIXVV=tF71$VFQw_k$3nvBn3M_3IknjdWNT{zaYic@3v|Smvw}?9B zZmd9a^n}H0{)Y_FOO?pI%3-RXD2i?C!m+y_v5GemfwMt6dYemOMdOl*=b*bH^id#L zggqO|@_PZA7I{e%9UfWh8Pg61es^lF?F1gV{A#fh(rGk#51mdREfZYFR(Wu%R3Yu3WrFfh;h0Azj{eXj0$`w2LQQaBhk)~`;qHFq zLBCw~npk!>De6o8uc09}P5gxkNnyOV@4;>n`N_I6VO_Zcx8H%&RqlIBr}X>8)%W{# z3R+>h=~e}o%qjr3U|&+>vadZiz{|H-?u1&}kF8(9l^*+L^2F%hjCD!?ei~9|Mxq1y`Eircw}CQG;!$4UfxB$7MY2N^yj)-54D3?>zVT_lbH7DE|oXSydFPISWTR8;Jj)7g6oFk*tshLTN&KRUm5;9%X;a@(^IDat#4 zm3H6nYro&ph~fcotC35l%Nc~{4>c7ylmeRV5E@*C>ee~mO&q16d01$~u#xwIuO&QS zTiNUWqYTdNm1E2?psjRd*MA>_MC?h5jNQ0U7Yl|)egtCly3&%|%K^^K7ER^Pqle?V zKjnJ*Ag`a3|MqlN2?edMr@1tI`5BxaeR|ut6s^dfm*IYGEeS^npNihrBlSC>Qq+bA zz=2r)99P;I4)Y%f%e)<2LW7DeVfVN~T?(Oo9qT;ounrKmwB-H)j~R$Jhau2mQe!o> zZ8b7-BM;S-EQF$YwjZ9K6i3wG$XZM6>!0X_b{5r--%GX~C#Ne}a7kO} za)-4u|Iqa%EGh;>DA4a}ZTou{FY6dKmHDP%>fUxa4u+{h#;v*(3^7tHG!CME~pNWV-NdY&M$rm#$B}^4S^mqD-eX{!6pf0N z5xm#F5sLY?DH%!Oc^lRyGMN&Vw@m0nZl=$$UHqc@D)*zVGN=3skbof*UAnphBk2W+ ze0@@{RH^J9Zf&`dxBOuBsnS=cU)H`+P-f(992d?T*}Zo}`MP(1%~&~c5ldS1S^@od z?9is}N-dM(xpUR7Y-8*16U$S|O5I8&+dX>wdtspLY*rU(Gzk}qK9)~@r zvVS@u*kG^Bz-f=46c}-`UcrWI7s?Pi+PL#nU7`^Q=WRT*PoKcbUY+XWSt-oMQ70dQ zTXowpT^6o;Dk#npIeTNS)~8u|!<2oBh4^xouWr%|n!J6WZ(2o1{*Dg~)#G6KSZ5N~ z#2`@g&ije&2j2SQlTu#xtCvIk5d0CO*z8|!i4kisVyl5G#vG1K*a=_w6l*;VA>zZB zF|mdLS()9byZ3CIukvuVS&PSb;>{0usn8HXKs!yJexokhon&{a&_u5$5u=RAHT-wYSKmVvasku-oS-ZyfEzm|s`J zt)g0Ugc$;(7!;HvNnx{CTup@5?U?yr!yN7X96A@h_>o!BEe)5C@#>vYT7HcaOcmrS znQi`(WOan3B+2>%ySgdnkKINXa6zu}(L^Rod^9qqmNH4b-~J2<-=9ARVdC+A8%j%U z^Cc|wTSud{6mQ=BMMiYE!%cED#PPMK#RvbMoo5Q?Ny7d%(NY2du}vt~ArU-m^Syz~ z0V64}^Q^ly zr|lT-NFA|5T;t@h`}`#WD;T|rwrO^EpWUa*T+LOU&%MKoz06lm{B*@vbXEUe z?GyCgs9X5^f-f2%ZL`$^q9gBLKD=;`|7=$4$h67_gO-^)A|Z?xsUsKo^&;rcL(Q-O zTczHo2TM(5Zx5xK#M{4N+_?m6EID?BfeJoSs%BNnv>y#&3d!T$ea-85#9mBF?0!<@ zPs7_v+VEbfS2lH8v@IFD#!7)iyVghV&;KeZPByXJr)IFywV183VT*hSw~-pR;;Z8m z5zbk-0au0^IPf*_(XJkWhL=u2jjFeo_A*J>w(KT|K&8)Wej|?K_ywrIsdBDcp@AJj zFD?qst~iV6F|u%k^Es~?>GAeyaMpl(i0^<{y3)vxodup^&T@WG8H>6BS|HOk6pUjhAy%tSL(IcWhAg~ ztjI+jGP;I>B131hXib~-7qxuH*0cIAD5yLZkUFPIg<-xpr`jki1-0o`AE#Ej|6;tu zP51r^T0(@cy;$(fLFSp8&HVe!xZW#I;`Y1_^iM~*`J6T^l7PPnoe>RzLD*gpdv9U|>y)Ha-HAdxl@PlhM^B8I=RiR4{f}ZW> z)85ExkWcZ3vRN8s$6@t_y{wAuYZbm(U*+gxBexD*;!)SMI7-~ysY*NeS#svsiM2tZ zDocY$_TGrsNZn3QBO+<*$&%r1~0h6KVrNO+J#N z82J|pm%8#09?!S0UQ?LV_o`u~>WPLgRnU)#N*rI|9^-{39C(l3w*26q5En}pc=$9? zpokNBL3#15GJp2!;azTq=Mr}XY(-Dgny{z@;`p^9iCb|pH(O?gy`!SPWVy4*gRUdDULBtxc7Ny(%oOYuGEmrrhvueyKC! zn->RrV{B_tDu(w^Vl%FoDei8j{lulvD~it#_Y|qom5^ggXr^TIu1bOlll+Fn$2S{= zm5aaeg~NRZ`t%=jM;?Z!Yzb#@brnE-D2Xv*n$?SLoEDl<-5GOBZ$6MJ^o9z42F>~S zvsuu$JO{@izVsmPTwlUOibiu;q_*`np7geNsi}h401k({R9>XzWZfbD;{-+1s zs`y)SOM5&Mk++`Yu)o+D{R*SMAaU+tvb0#{T$Ft5Y%Q_GjgDD@nl&_Apf8pmKUomI zYkjeOJX+>$fP-q=fxq|;T$!(W(gyn4Xr67YL>yW%>8 zc>~D@I0CKU_%?^qgtuI!l}GL28nFmfVMW;Cb9_IbdcD!rdRSdU4Twkh!#6e%;?DcW z`*cuOiLIDI%d}MfGrt(*VCQ0(n@>uge&DPNM)wlvN4;^TedVk%D0NOf#Kt{2bTRSx z+QO>m?-#67XYp*evUh1*lzDiPciB8Ny;_QUs?t|{B{*Eq*}KFjvD89Q+Tok+dV4t? zrdoAqf#eN(V%6kn_QT;h7;J*2oRpm2q}tdav1;XL9aD~`N`VjFH|}urb-y6uQ{C;3 zTF*uH>Z06?kI$40VKcBLJlkcBK@PUhb*--A-Tckzvk!ZN)VontXK}}^UOxrk zn%TjXOqNScHjNOk-eI(Je-?;$y>_aO)!PE}d(>%e(GxGzl{^nR6n|un@hoXf6fZ4x zf34xU`6Kz!oQtj2!$J_|emA`ZH}Dj-fM~;?vAQ(ByptYr z3r-7X?J^XrPbO`KI~Q2`3V&VKhJ)pd-VYT+*iyRq2M>*On)qz zeN-*84MKB;rp2UkRR2thfErFkxye6wly=wfV6E$3e(efdHsc=MRyv3;y*_xa*ZZ-+ z11-$aJag)B8|^q-fl9IKo8>SQu=u!{HiLc@k%Zh?Xvm!m~ zv1xBR3Y0v4Lup7DFL+=2yxL9cjFzXQ>H3@V2L6$W?MHg#1gdMcsy%fuLgRMOFIo<( ziQ`cyLZXy`8XrysKNID*JSscg7ATKxiBzRLRz1EV@y?`Clg&at5gT6!slDDP7 z;3Y~L-JZ!?mr13Z7V69RKYIbFecp*EU zOVw06s_)k7Ku5t-EGL`Xe2=BDhSN03ET9OTS!ZO~XyfvwaaVcSJ=V#)@L?jgA&rd@ ztH&iQ`|@7hDgLnGdS^nsC0Zn;Z>q;?`e1v>WlH?%7;qs9~*~~m^ zhjeHU_~*dYAc_I++8*j2RMVW>pPv5oZkdp=m)(gO6TwT@38cleAm}*ZZ>7EwL}CW( zDuhD!`6V4Q`+FD3VE53nyqV4ZD4~85@qVuv(1^zqv1`BnBn)tePS$?ORM{DMQH$-j$xXpCkP? zU?f|_7UERPb?L_L*BSm~TA0MQwcVe1N$H~_+}o8eAre-IpxFB#uUZ=>1fYC{tgBOU#RjIz z^jff-+i{{b7*6Es`3D1xLX&B(y-uf61O_e?jp)v8FQ4m|n(Qny8H`?>2^SnY+}^-h zw0`b+8mORD>>DlB=(UIr@pL$@w|oV_tCOe3@1n4YAG2cSPbsG8Q6LeFQ$4A2SAULW zvV&5IS@O=-Y8R99)?D+}AGYGJGHAiBqVck}zays-?3FXZ?|J6VXy&V^4*Cky`A+fB zeg7*ijfCM4#Aj=9NZ03ZE$1T`Dt}ewt{lb}sT;1t-xHQ?SO?{#Iwci5W3^?1wdlMb zf81ETrpUgZTb)4I_ge7K0>#*40{%rUD0od+PFF- zwYk9bg(ayPlV^zK=Fj#xI(VYam7Z2CXW8VP9+bv;Z!-KCn1SS(W`gV<8jKnP+c{sT8VaPE&MbVt+ zr#qS7d7sksJ}LS`VX&#BH`DkiaqkH}9Oq3>ls_%SM&Fxj>1oR2Hzt)QoGxf*Sy24uSaW=vAIKQk#<~DG;2(5h#8Yi-F|P+|i{UqY*E`USyZS0ThBF zGUpRePgz@>H(0!`{S=C08oWH?YnJ6;=wSvA82W0W+A{Oy7GkMllCx*P?L!>Qc7#4# z2$ff&pjoti7?MXs6#h3sC}1h0#|GdXRP}vts!lLaY*F*^oY=#Pmkq;NR6SqLRPKQ%(tu31@TE( zl@6c066j{~y0YG3I6@G!5KS~NYQfr7Y{!_ZD$B9OKeKAU$a0ijxT-EhI7&o>gd+xq zTJVn2{Lk_W17zwp4$r%U+L#ISm)QR}*wMsoI!g;!Kb_6r5Oz^iu+z+I3j4~Uc#_56 zK=DkP_+9aw7f^B$r<m2i&afV|YPL8k2xi-`(+|9gk>K!{hB@dSd~s`U{Akc((DiV* zGhdHa{&ov*_u6#TTU(^3gh1HwwHlI;|FE^z*KW=RT6=frdD~xpuG|>uTP=j6+O*j& zrwcAU3WGInFYmrB|OrV zom8G4<#yNlp>Et_K(g-VB+hX;E}Gmo?m|7*;zzdoeB9F$**D$IGOs*6r=)))bpPLy z&<}2B1(g)Kgi8o7C^;x$Azm;5kO>mN9hR&c5s%dy!Gc0;;XYFDtO8434t(#J7l^Nd zIPbB}h_KRMwRsTNT`AD-g3pj0I^dI>nG$3X5c$O@#}+J2>LiRx+nQb;f9Whbl`9B| z%V5hAh&25wttVRCdvbLk?R+_EuShlE)^G{A(`DsZjc zPPz0-e_?v+kCN*cj7znq$?v7@91qi?ifQD-p-9+e{s4pcdvi2O67s)}tQ;Cm6rkcr z9beE2i)yihhMa_nuOjSA{A&ACcc&nMdz~$I_Yy)63+N9<|J0MHe`}*&{qeFO^5eD% z$sfdm7hruqY2m`|M@R^Ke&&%pCm3!t_k^0CW5)UI<62X1GZ%f>GAOxaH+YxGw|#2p zfIp-i{-~fYFvv6!C3V9x#8I&ztJdV*I8ODRX2l#-aQM!=PTPZWq;0}y#n<8cj@(${aW!xNN^YWKd}{VBO3 zR(VURgNFCvRtuy#4S@-EZ@$T&Z8u3C5DR=)RyBDc;zpu0@ezrPa(BRS5^b6@FGprF zhHtk&fbNdBa%5JyZQLiDdo!4P@CP}hiKKe?Byt(l0LpCOF^oEgKDSz4NXDCSq_4X- zYqVF81{{3}EM1|p_<21hpeyp!Xj6I;* zuARA3G^!e&+KC>GwHR+?VF$wcn92l0x~#ciXU2J!)Pd*q>_VMggrgq&ay&uwa+#5h z@roHzT39CRoYM6Du64KV2LVJ`1;&rQAPqrB{;=-K61l} zeXRQ_Cp0WJU?bPSFfWMyJRkc=b(gAxY9q!&UQRaeHCv90-9#5T!nscK`lTGfngf`+ zD^LU1GnTx}`n7KG9kXwX+K{B5SjkRH7cArFcRm6C9VM5f;jOFfsIAX#PooCZI0ue} zykc_7K~>ug#lE+Uq51;6z|3jap;MO@2cBNRobo)$cYF==Bz6i1B(lZD@PZkoKTnjL zwoE}CZhR^{bLY{PVi1kQQ)ourrZc~7RGGrPiBs&S6h7qdnf->HgPy}$J+VD1co{LH zl)d^D#M37RHN{!0QFdj#3W8sPGGiUWgc+iDI*!kNqZSi840GOFF|jPrXEP62P&&ee z4kI__m-^IIs=lujkRI;;40HSxwBjM<7rW1V)*{a z8Ar+ya+zDk=OXUCUA9B_xS`=%Sne1F>uWp5;S$t`XKkW-w8V--0t|Ss*=>7V@qkZ% z=H64h1*&G4%PA!T(^3xYhXzYJd{}uKk)uc2yqeT)O7)+gv?4Y@MB9tc~7kv>FzZ@5yZhP&lSL zM73^-xG<^Dns&OH=kChpAIIiTs6@G)o$6#Z!ZyW9<=Y0mB+_7(#tC~*A1418KC#AZ zmB+8D96L1-sj0jcpD#QM0J{1;N!;bQ1e;#PF^_%f%xG=i=$J!vW&(1#Totz>A_SSl z{mWjITDK;|x`7G@QC*eeK3_Nlr^mhe&DUDu=s3&Eo2BYu-}oO??~M}1p2s>gUHr)` z`rWr+^wBRZ4&94Z0!CgfXJ-!SmZwSl4rqaS{V6FtPnFrYS>vy5@b1d`$2T@UJA45i z<@#%?$u&b%c{RdCM&Or{KL`0+WANmM*|1UOYfPCepd9Y2OwByZ?7O3KW5b&?2h^~cO}n?p`g77V@zbdvJ*kn+>>F&291}Lj|TA? zK36YkypYf9#RxAGFjf8ZG8#lEmLQI_<*!KAa`i zvaWwZ@TI809cPVHX;-WEuxx&*{6@FcXlS#0J8Vw#Mkr!UWM>i@>jSSD|L!WZueZC* zI?21MT$3=+6+sr+p|Nc@^-!*8DSd>=Y<)V)Q@V4!IYw=!zv0fei$$Nw=XzY#biUr) zV}h>T3)E;z70c7(=S6;Fa(~K`iiBEyW8$GkaoAm4R_)jMwXOUqjC~Cgh550>fxNh^ z=4M6GzfMn}qn6d8UDH3YhKcEKjyj2HXMV9%T~qfchCFJwrx0M=0;Ae|t*SgjkW!h5 zy4n3U4vS;ySIA|7o(`)Qg06C1eSWp?li;J>4;lg7Z4LNTZAk_iu{%!@FGFyYK7eac z3)5;cSEjC}=IZJ)x<3jh=~nc>x_3k7=MlpcaUpJSWC zs)@`ie3zl-dY16I=}GYoQHxJCbWF~}@N>&od#yifJ@)l33ECfg-?*)?@>*A8Rh?V9LOhTD z5l$9ib?g~q`-+@4WF*;A|KJgA9Y}kouEK394;wnnah;*GXAHN>8463x{;=Svwf6SX z0rmS$zBUx7-y7d)!5Hnxb(S^~imz~3hFv&#b>zyd?uL9n+qc+kYikW5_V`QAnYN!b z?%(nrN$iQO1SL2&{+Ff1)tRX1+4UWbn2+D&-lD1p*HZ5b)J5~27;Kn~V$h2*egg-=1{(6#0paJH!lW+= z<($se3#87N#Vuz1f`mu`tF?1_d-=X$=DoOOogdg5zF%G$l(#zr(6gn(TQp5}1FrBG z8Y3r0uWOU_T|m@EgPn@Q4i9(Yy+d)h_|dv$-{#o%@~#CAQ7B7&ZTjki?O>FeXK)9P z;h%%2_m<#2ush?aFYV+|9--E##j+->B7h%4Y3J#a2MwFa3ESGQG$^f{x8wh|mGdmo znf;jQ^^yIz*IVs+6Hi`sk$A##S2OV(L#9Y580(>f^&!bxbikba=jATjJ;I&k6N?Xp zIbuT1Vv5bzD20is3N@W`K{x(%eLCGpsY?S*nfazS{_$Ww5;M~!rwdN@H-kIb;X4w6!0@5Q1XR9l&u z{(A-eaJL`)c)?k3G#dtLee%48b8Oa>{cB;(eCB)gpJ-wR@eu2rRc{_K1tD}*kzq^r zZ@S@O$nQ35w;&N~cGh>#D69yvzas)X%I=U8rnIP)*E2zPD6Mkoc2^cmmtO85G$#QM zaY=5lFD<+sj|gb=>s*>KyAw?mn3(8oHln_G3-wRfPAm*EG!O zUuly8vqP!F>+jlitFY51hjV1JH{OVrYl`zZ{_d57skFH(Qu-GjmEP^$T$ES&k&^{~ z`#H5_E4=qE%^=bvLUQ5BQjLUPwB!;4Xr0?S^D_VPl-QDH7m-gsfHby^^7gB3=UILj zR$^+5r1xiRw~D zt54yUJPSRjWW?O@=t7m>X?~<_1p{n=-^|mWf01V9xn3PkRG|Z;8M=>II{%T7^>UZ} zrpSy83Y%P44@sOuGZZ|2^~({b(p>SOKoBUDPza#mf0#^i*69_U3&Iqdh}r+0{`q{? zhcg(Ft9yjrOSiTUhAMySi7X1digVkGt2_GFemk1>?iLVGdHnj)E3!Ckx3ka%nm=?G zJ9{H#BkCi9R^%%jLuW7cp>CyqQ`G_-h|qn*fm&^9Pv{>8QT3g4AeACRy=6(%6C599 zH97Z(PLMKuQL>7*^LIn;Eqk2juy=I12d;PN_N^968-_~#iNCz~7dn?vV?cdS>ZtuY z7x|y^ZTnn zaoH%qf};Xm+BAH=5wwn?f7o2*m9REZz?s{^ao@2k`u~L90Vq^-fGNmJ)v6v-sc1 zkJ&%9ut8JcY_S}WE|2pShm`03=BG*ii0V+K{t0HJFCF^-Y(D>hyOXX;l5qy?pR*A7 zIueix)dfpgSkC~;nHl7M-k?41+A$kIaRdRDMA4=Q8v2Waw6ug>(*}gd1G%uiL8t%d zb^iA>Q27&Ul#3t}D6Q#~rg4vh&HfquL^%+Qh@zdXq;zBsyUn4YteZbA)r-qK?{fX7 z1Rwk7DZzl{I;<;HtV=CNQ8TAdK!Es++c`^AZTq$R$wNV!A{Wl5s0h;v>=ty-{udr= z{^j8W9%wRjgfg5=i#&P$yDm^1fdqxqNFowRj{L`Zt*^ZTgk|*EQL2qU9PWS2U%sHy zf<_S6=ncQd$IKMp)3Q1S=FeCW-Yj|+rg|+{=inBtlNm-9N+D7Y8<36D-VwFO<-xFF z2S441k`3YkL6*Y5epWt{{}(47D7KJ@BS81Y>k-hUxda%^dLBU1y!FpWH(oRBN*>6m zKI!zx>X|JPFd0((6s-bB-Wq!LU!ZX7gRgD{)6+#Z3GsCoA@H%Scj!ABT;++ZTE7Y9 zg@ESxd#J_WTQJmn!^u^Ej!Yr9%e4JFKe~Y4d3WY^i-9Ft_AG8%0PxKD&Sw7H0Cyeb zU11kD>b6w>T;XX>^P3ZY@RvRRb51oBpJ#6KhpM2U<6k4N-=jBNahfR@$`PFTD~K ztYrc|Cy^Yr3;RB6P_z)DrZ$mb6ONx)&uFWVw^h@23tkj{uN8X80R$Kt9u{6P{)^6U zKXgMrycyR^)(=GlXjy*S4ChIYqMa9t>nxp==Ly+w0izz=?enJxau8vqFGz&nnq0y_ ze8DyU9r|onPP-a)7!V~>dP{TPA1iG2+ z{a#<$F!n6d(cupQeGvr>6$}Tsd&IOF+P6UY#9Yq+DVrzjTLeT1farn=s(j|tZ-AY@ z7-UW&p5=!{*oZwQ1CCBdQ{+j(($Zx>jiowR0ei#6RDkz=sYBHAZNY_(iP+br{e^QM^1Ud>JTF(2AQj+tY3lXQWk?1FOU4 z{A>ftL~Q(>1b!T_L<13*<7!Ox zFK0xLDMivOly(}dT3Rfc+9m?o6szW#2kQ^~C!q@x5XOw&w&p zbOU;_2;5(@^3x1Nyt&G!I2g1z8?@od>-@ zo-Yp0GdT(V&%MR>XhlY2k7%Gab7Gj3=U9Y1r{0YKnj1kcNWO7#SRkB8R{J4pJpiO^ zxS!P#*Jp;+FOp=V0jf@6I%+5jhfLC0T)+f-W5r%$D>E6z?ZhVKKzra)zkk0q>|u7 diff --git a/thesis/resources/figures/methodology/methodology_use_case_fertility_decision_diagram.drawio b/thesis/resources/figures/methodology/methodology_use_case_fertility_decision_diagram.drawio index 698541c..2e3811f 100644 --- a/thesis/resources/figures/methodology/methodology_use_case_fertility_decision_diagram.drawio +++ b/thesis/resources/figures/methodology/methodology_use_case_fertility_decision_diagram.drawio @@ -1,68 +1,61 @@ - + - + + + + - + - + - + - + - + - + - + - - - - - - - - - - - - - - + - - + + - + + + + - + @@ -70,12 +63,12 @@ - + - + @@ -83,49 +76,72 @@ - + - + - + + + + - - + + - + - + - + - + - + - + - - + + - - + + - + + + + + + + + + + + + + + + + + + + + + diff --git a/thesis/resources/figures/methodology/methodology_use_case_pregnancy_decision_diagram.drawio b/thesis/resources/figures/methodology/methodology_use_case_pregnancy_decision_diagram.drawio index 1c63f92..382c778 100644 --- a/thesis/resources/figures/methodology/methodology_use_case_pregnancy_decision_diagram.drawio +++ b/thesis/resources/figures/methodology/methodology_use_case_pregnancy_decision_diagram.drawio @@ -1,109 +1,135 @@ - + - + - + - + - + - + - + - + - + - + - + - + - + - + + + + + + + + + - + + + + + + + + - - + + - + - + - + - + - + - + - + - + - - + + - - + + - - + + + + + + + - + - + - + - + - - + + + + + + + + diff --git a/thesis/resources/figures/methodology/methodology_use_case_pregnancy_decision_diagram.png b/thesis/resources/figures/methodology/methodology_use_case_pregnancy_decision_diagram.png index 550dd270236ccd5dcc96a570056c0b54fd9b3828..6aac0f518b3399be53fd19d9b80faa2852034ae4 100644 GIT binary patch literal 251452 zcmeEv2|ShC`f#L@aw4LXQc_a1%}HgaP%?&ONM^PVj*0Y{@T5BKC(Ng2&5a1X$ZXD;v z4eK|L8#j68xN+<)>=WUMV(Fq3_=m+~vzqd_)Ms;=#*I_grK)VDI`6clIN6L7kzPmt zBqBw!clDr(NUs->lCp4fldz^(*tuIcdq}w0P~j1LZ((QSg1?~Y>PvBQvJjC{m6agD zSE5@bB}Jr_;gbf%#d9b8yK{xyb~%|P@M*1^o0E;9jpZf^%%dtREiWN0hlXt0ps%SR zBDD@aJ5d~M;2$*`D@Ru}WSzact1~=Nm6To~K|=q*Lp=*y3wO$}DG*FvHtrr2R~I_i zQs`xQIxWUFp%#I%D{Xprk*9;~foGS2uWwMA5;9O7+1R$HJ593Xklm&QAEZbe3_m z_^~g1s~)bN?p8SI8G?(yN40Rb1HFJr3B&h5IFWp!z0lvn2L3wre$Y_O2IQaW?gKNp z+c;TJDPF^80ab+auiemlXtCczV|WK=kza=e>BUi;#BiTDKD;cPJaG{9ESy2%0XPO8 zpl&cM3?uFj(jh8&+f%4EdTtgN%Dq8X{7y1~+N0VK+61lf>ziSSSbVsY49?PoD#HKt z{vp?rXgor&{VoDCFS7n zu_R`!LU6tvBh*qP+uX=zdOL4K&$D$AgjO~%9!obNH!$7~T z(+N8O0ril_Pa+&xX<0hNIG0E&;GXGl4^|p~Hr#`iBpwpB=|3M$NI_-kS5x#u?^{VtSqdP5)$t{%`uC1Otwd&p+N5B}vFh zN#m~Q|1>&tjs72rYX5wd@PEdK#4XP-mGF;>PNn|3=ybS7_-kE0+?Bw!5#bN~6Ju4} zA`j!XwN_M53zQN6HSdoe)L2uzh#+K`H%R{uf|Oxjk8xAnf3J_oRI`jRIsd*(C`S|` za4wRTl7i$M=2;DQ3XxZdEb|DbPzryMV1mEYDYTSS_OPL7im!Ke*{ZjFt%k>D{SD$c zK>rD!aFoNw(3Ss9ClOa5zv(2>$?Apx2w{~%Hz5UXFfSZX9V9hZ6 znTDY{`wxb=e~3E69oK)DI{Q;~i#(BXAyni43EhG-!Z25B-9J#X$d58>{|wE-F!@O& zr++AW9d{c$7Yi3Fu&RG|4KFpyXO)vzkdVR6*WX&nGb9C}0R9tB>%Wu8l*FCv-_+_j zkBzb;>7vKs6IL9O_r9CSs^7ay8^r#+?EanQ2#ZG z5YJjNRfI%=_dl-%|COdFT#Wt#EeK6fl2VemVaM6yFI9swct}iWLHy5uv=$tjhelNo z(m%){mmz#YLSy{19dcaV4bu`t`w?_)xBzK>7%ZYy3DuCrZ7rNBPJjZ`%(TJA$;$?H zM`GR#3HyFntCJMo>V$tmZ^K37TwL9qAha6EvTN*cS4_^>3dcd z=wJiGsl4%?Wb~yBq-3#0oS^&N#$6omM;SZP)!ohB!Udla(yQp?3N%zx#qoAfG}%x` zDDfQydhXEy81yBH*eW)P>JF3H!X}B8Tu{R(yas_AcGdz7!I$xdV<;m58=6bp+s4w7 zLKVm6^q{)CI@*Ziq2Gu#v957!4!p^1INq?q);5sjgtmEC7jddR#mdpe#sl9h#f3tp z5VRQmF3tjIv0r`9%GJpgdS%DFXY1r@fg{MpmB4{D#ly`B^4j?8P89e;iZq|%>;?xH zfC(^65iCcPpBNUcKp_%MOD>iksCS8AVaQ3-xtQKtM?A&~Rty>2-O9#V9NUJtg|&kx zI{1Wk^8@Pt@jliK4LHG=*IV1zT6khXu@h>8SGKTnw8QEg#0xm{k7(jUM;hoM8iTCL7~kyoq$<$C znqg@WqVg^?GEX5xWW#%3m70633(QwWaZ*7@>{-i5Wbk0hc>7%u%!)+&UTHH;}Yq9dnPaN^zrB4WI1 zH7c+j76pt96;Lcp`orUe5xzgD_P@sYmmlf-OA~(PaNnQkQW(*zNjxO@;&0&l<1X?r zDr0$1JvgmD+Nk{Pn#@*=q=Wnld^5x7Fcg;!r^DZ`$Bb~!{}U#BlzIKlp4neB)xQ~% z(R0-QM%?|o3%q~23IDbI4-=z5NP-xc&M}#de*>LF8iPn%5bm%nE>gopNRnXAzgmPa zM~lfNIz%_ekL!wtBLVxCxFCuC1wH^>B1P{`)^c68cicFkaU0jK)!%8>TI+M%#$flP zkEQtwPQSUoNYv(-h2%oAj$EjX>axS!GUmL?)@O20$=yumk&!3}Ry8tJpK56tFzZd; z^zpF)M>qc^)syXg=K00zkHWm$GaCH- zEBn;itYmH2*1StZLkHfrAomWv)BkVsBzF44;fVV**=n@>W@oMc4|f+1FpS_wzGNN}9Cp z%yc20?KJjTDO;~87bV>lxhaJ|3Ws&a)S6((cDafy2GERE0>ZUUig(2FdjNk;w$$GHc zaAThqv|E}g`$M%(kxi#B->Ue!b?&^v-6YqCR^}(z*x&IUzN))F*z(q`FCSi{MFeDW zV_04^MI+&qYWP(1wh{2^c(bs?Wg5QdMT^EZ_#a;k5OZZ%`dTE;{03h)?b4!bM8-AV9%SC zlN+7Fr(g91qjc` zH=Ez-n3Ur*U$~?E;F_tYH5oA)ZQ>O6nXl6c1nSl zb$E|sPgLr{p|yv9pfyispJ{N#)>Z?X-FC*|Z9$ejv0FMn)$CIS0$5IXW1sf^=^1U# z=G$qOyN_bCj@{Dx8C)bWSLcFYjNjhzot-6Nd}rql<7{8g_7+dHG|i8-ZGPMeF>$XW zLEfoMreSMdCSa*P)NH`ntgGZ@qr)#U9X^pWb+NIpPn6kD=PeAmbkJi zcVn=UEMtxr(lJCN-%^o@kzA5>$1>x-L+wY79C>uPRvC~^nhJ80n1lCC0b2KThT9?+f|hVFGk*$-c^u7`BW>Ojp88n!Sri04X%6j6 zQKD7JLA~+bB8-tTz`;JVyRxzo+ic8o))2LpDURJOzyQ% zfw3sXo+O=w-9#o+#L*N*?6fbDQhs?Z>z|1KJdVb}o;L5~P2cWjUPU@Y41kD{0JM5u zyx@uCpKQf;THNIH>E+v_QshHagM>?D5L?eoCHH0dG?m0J8A!$S+gMm=nLyt1_@67* zBJv*Po;h!x_O&40GN(OnBo=>HcxilFaqM{Xgm*az-JOfvlX(}D!Zr&m zaakeEEu=Gzz8j;yg-ck%?lj$A2Mb)FM5aG2u?MhhEbLMpDhfI)jIcl)Uv9jWd|s zeGO=vh4@GT#ymCRzd;w)(SXukfzWK`@36zWiLI=Gr25(BqosQt>h+X(}eJVQhXm z5_+-)2Vl%So{#=tWX6oc&_vz_kTx3|wIs&`l&Mb$vdw2ZxByV2^P=_?3>AcJDo$@= zAXW&l8DNZ(?dvtpA;7|yz8>%8;pqAv*v2HMFXLsmq#|RsX*7df?ou?wG!E@N``3Wg zyk7E%Vby?wA?sH%0aXP+6*FE@ia9JFIKBqdzN|J;0Q>+>CA@bRgdi}+5??CoV3%nJYBfPQ7U=?*!Y6)WFb1)cF7Qhr-?KENMq@HynEb6TTeJemh+Up;n|A{V z{n!O*?Eq!wEe{dwtC@j7`gIoGdH@&|$k@K&!uYXc&=Y-qKuYWZX=k(r`&1ZKn35gJ zBc$VkxVrLj>)4H%;@}&badSQ5u}qbMzSQu%QElkq5JI#K$J$&*ZnU~ zY?#D|$Jn`&lE6%k-%j3EsBQ1L2%GMDJBfF6&G7Ar5t zOkI=>Q1*D?KI%H|k1V@HTV2#adxrp+f=8%hl{}`~1Hp_Nq`4(vu5savk1Sak>9^cj z+499-_nWr&o)x4cezZhT!Qho;M=r%YksKY+Npm@}&Vk&=!~vLU3j~=kbu1udhIj-8 z#c5=Z34S>)E^R3h<7n6p89G)&Oa@Mg?sLPc>v!eapvi!o&H;{y^Fy=g}-vLBG}TSONVlQbg!;N%Aq z)=vv~4+_Q@6imtKM-Q3MLJbtmT$XKE*uyqa9Be~OZtXF^|2MWf(VYrSbccP+;BsU( z%?b<7B1#EJfNgj%Z`@OmhctjPUi|10Ch~9$NWYpjGXf*FiFH0GNv%>pwhV0+5uz=XzNWT0z6E#^;~4A*{EPeKwxI3f!f{2 z=dA>PyaS{8tlf&_|AuudK8c8fJeRhH>s0w=5cEAAmz|$}D2yZLLP`Fnc2+ zG`P?9pd}`WxBJ5%d5d57M?BxS9Q$Il7;z;<0nNJ|Y#*UvMQ$)`_Q1HUYgkz0XMpK* zra+2=A)LOvVEVX=z$Ak`Q=>gORkDG?|6#lrWs{X9hNiv@Iundp5k&M6P(oHBH69U) zyc&cs)oxo#;5b?{2%+JzN)68b<;Gj_!K&pEIQ}3 zxs^?TWE?hL_0Sym4ViBOmtGf}v%9O90!ghYsd?8P&pr2SbYj%bcnvn6j4vc43DV-~)Zu$NVZ<;N-6X+cbWGtXveenvxj*-2cmsrt^p9=F2uF zoO21+THRen{r>53L-Uo=K_H9I^za2>_)e*tFpk{ALGsG(gqvWJr;jA>)s(NV_f1RK z-BT;y|M8GggxprP!#AsJ&$WI{ZS?4=^|q@&8`d^{-dD2kU|(Hb-+BS5y#c!GIGn?t z^i{26!d0Hg)pWE;!NlJIxtgx=fD^US=G)b-L_AgLb-;Mva~kVX`A7~M&Wr2iH6ANUmwXQcsx^B+!Mfk>d3Yrv#wbF_y(oHZ=Gg6U*8@}+$52) zFsm}tv!!JGv{lFF`Zk_VtM#gh7xaE9IYmDA*~W9uH>)DvrcLsIh_m$J!}S$C+Xvon zFZtFG9euaI%{5W6{kE|4wOl?*Vx*V4>z=-vQre`sE1d7l>u*$RiksJO@13L<93B63`Mp)!_Vjl*DGj_;D&drWdHkqgr+@vF?PtwC z1sd}5wFX8eSsGrlSDUw+YMkM2)825t{L>4oW10<7yIj6hBB46hXWAC-n7VP^;Md2c z$M$@G+Zx7xE@?Yk`M~RG145B*_a{>Tu;;Re%p%_mqy)U|jOw{V?Q6T0_~gkGYK4)V z&GX8m@~_TDbzg>$XHRSA^eIH2G-JR8Kq#HnU|*Zd3xcsQqaMPHo-gCpCpF&beD-vr z!uO94N>%u`mDL2?Fx}l*D4)1)+G?5H=+5*)`MOm#noSo@YK5oQT-np#(sg0+>Lq<` zv!3r%4J1FAz;Akdg2mijt=EjMD7+1pF4_J)cf0ky-JgQCYh2mg=2+_L-+8XP*X4Ho z!6vJDeOrBNYF^!Z<;pGU`^+QiWPjhp*-ppjf_zawKUti3%t({2jCC z18l(yPR*EqZc7s7=4@4=#UJKcTr;bQVO72N{r&PA-^{$KYy=a(HyhqtIVB<8L8)ef z%8J#nlcpx^J-%{&Y)9I&dtP;=dS@@ql`eCDyMP>Ep6Lo*%RIWrDl08=x)7rt?89!RqZm|Zw^N~ z-tA6twaH>jH*WO@74i78_fvLm6)w?g)!ELc1RK8W<-a2|Wop}v=WD98)VUvKG$&t5 zs9o+Q7f62k)a=W3r!(gpvw%SocC}|L2K;T`)Kj--uq@jpySpL{7Szu2{yx{kIep6V zJEZJ(&nr&{1gvshl+QU`-b%$*tl(pwvT#dZ}X+fnpsO+b5M4&Mj^3$!DU`W zX=j+G>LrqIr^N|dmM##IOA63s%U(11?YVa0?qjRJI_wa<-z25RG*J+k^G4Y$uBnTf{C6z0&D*NUwz+hABjpxq}9{H%s z=4N` zwEtw{SQ&i=c*JPy`&;+=oijh@FTD|$Yb>t3a{2D2O*eA1(%v3ZTC=-ZFFM?-E_$FV zvFh5k6u!rHxm$PCt(cyYt#Z5R>Vuy)-TRum{Q91O8%ljQC~L#N-de>Hw|vjHr&Yyv z#g>J8x()Y)4t!od=$`GqenZ3Uw;`LRt@&#H#bm$Ro2}PfYJ^x-B_A8!{Ty1wD|~xK z)=ZT~C)*(5#z&7I--v6C%{6{-BYQ>@@7i<_;1#picx?(4{$i|Xt|BCMf7=eHYOvxj z#a-W~cKW-RJ`upd$p{hOGUH-9TI2fOB@_ z5#vff(0h3=joU&L+c#Y6)V$=$F?036*SA%xOccMa|Js~Zs&(#U|HU=Y5`YXieAo)>qe3x7L}qUE1F75Lg?u;3Lo4S1yZVT5K3JQEJE{FimriP} zT(bN7+j-lsxbDify7PjYa+A*@7EF|LbK;9!QLn79l-z5!DaSXrUN5UIyjbpSo6{ID z{-?YUi*uI>r*0{n$`07r-R=>+3F*%p^}_F_d!@u>*=E_bCh-1}1Qm?d1~V^7f%Q3} zUh^~G{k-XVMk;u7HR6t`d|6xiUr$oXyKmQ4AKh3tE%8f(BS88j#L=-%c4_6hE2XCb zdTtzI61yXhLZ5JI%eH{zy-^1P?=N4|z1Zt$Q<>$VNclVUoK51cK2Pt|pV4}oHaItn zJy$}l_|1mopyhXWWG$5#zqqwQ(2we#NY*oWovfWv{^mYbwr|dV^1KtLjLZjqZgwUCj^`IWLCxpNo0z) z%&R}gUQcyU6Z9<8zbAk1TcKW0JGHi;ZW1@iK}l|TNMevxi{+u%12_Fwq$_QY^a?R< z^gH*Wt}o+=!7W=xo({XrI*65KdEUgr@}v_GE9LTc0=ygfP2JXJ4)nI@?GITHAZ=I! zl+&AY&FgFZp@ygpa}M!kwH3_ixj$vOMl`s|*LMtbX%EWe>RL&g?8?6PIrPc}a7)}O z`kt+7Z*IPHMT^@O=YUkUF(Uz6rFBFo>Z6^j>R$h3qnsS$~<*;yeY;L<^%G+me zIwC;_cooPSN0Icpg5n-nRwml-SY_gPDCVWgx51Eo>@)kpb?-}i``!!WwQHc~o>KOk z8CSNto{9bT0V8%x$)@F+hrlILp4{zu3A@X+#oJG*X0gY|-`yo$XK(;-^e*3lcr|6%gXT+;=2qzxVS$hqA8 z_8|@Y+-KF@5C?3U4FpEXd`Qsnd!&H5_k4+Mp`1#3-AI7de_~aY}e{JZ(n3*f2m%o7`WVPcQpuJ zDd>lH3YEpc<#r!hwh#JNgK^Drrxp)RF6-=HHEXG!B}2NSxHv$U9`(cSo8j&8UKuT1 z7GcuQS4r;%rKfRh)#nXzRh)7!glZuqqJRq(E%z;}civ5&iNu+b)pqk7 z&!zsV!-6M4`t91XiUNwk4L!jpZY<+^0$EAB$9K}5ki&9u0I*@=oNnOznl>rZpFMWs ztyjK|hL!x>q`(-XIUD)gAT-fBv)%vOWsk{mD;(Ot)VM}>_Z*e07VZBuZ*T{n>8|%n zw`Z6|Blec&zFY&|YbLhWol3i=kJ0s*(vL|6mFAgp*%MP~&Y;rfno+J~d zeEsgEwpL=ae{Wgc;>C+^2p6PgHYc~%$yHgOT(<15}m*}+Q+yw;J~l6rX%lyh@hOKO%x zQeORcy7&6?)m#6lGL-O0To)wMWsk^;+oD{g@g&j&eLL5 zog&e5hkvIMe+vbx1&+z}aETrRKK602ScFkq=nQDEjI+HJ6zJ@rR9&~J`U?DMf)rU| z`wX?XZPa3ws=7`#O|ZGQyTYFPEo#~RAwah^jVG(Cpz{7gr`W=&W?*?h!RqQ|4MadR z(Zmz|b$wm4yvJ+lXDfuWTefxN4=99)1_^(x9+a=3EVwt&eWmndRM&-zE<1wFmfDv$ zRfpY!97SQ(Pg*;=%+gl8kqF6jO!r>_USFDeXHSp6L`A9nkV7)!gU(!Vq(7J4*T8t0 zECIPB56QynrqbvEYA<3ODkzSGGZ4wP2(Asad-v(nr!@EX`IlQDC_62Y@$Tn~PrY95 zcsmu(gC1uSu4icJMK6WGfiN~jiF)?ORf(!xvf1;^&x zn(y_~>}tKMY|c<0aLT`1qB4K?-EW1i3H;#D=~-d1$+b-euPW|38kq7Z3gv2TsJ;1O z@)EE04(Xr_6B^d^WK?bMt6ykd<{WQ(?^$2QR_!<@SwtV`9m&_5B28}=$bdA|vs@X+ z#4lENqCTHWw9b0q1d23ck{{$bN;?JzI3RsyIAXLdAzPKyKWOo)5yKzzx(mP zG8q}eAghY%eanC^1$R8;ZUU1O;UUg_8j>fGn_F{u#3Sl@D{g6ZgE+#!cSBJit=awD zV_^-luU$%mok|qoan0GTcXqT3Jgbm&s1EN_X4Sfu+IF)>Vs%&1x~i+Y+R_aJ)6;IP z_`ED6s1Ib}_<{hrx)VMrEjM5JwoqZX&7Pw8K0euWgqfHzzdZ~FjDOQ-k_unUb=IjQe5^sk>wlDnE6tw-Jgnka0;9cgXm{h za!0x0_Q%I+&bqX!FLHB!CTOC_ng1p7;+w-MRp`ajaxdK;T#WVyPc5-wRfB@woL+4P?N6w%q-E=OgKaUN4TBDvwdfP}RP<*z!ygs6*Fho{uDLm+$;# z%JtXJkZI}yw{!6nf6toOgf?2z3(X}Vn!E1vM3EYfd%yCrlY8!%nqix)-Rk5C=_2i{ zMHRe`K{ubT>AT=~CEM%O4x~O?XC1v+zQsgy!cpUWd~JZl8 zx8Er?_mdvvG|rWP54Of}X9IXrntXmogzrh*{rcwOj=GNrYvc8Td9u{gtD|x8%R&cE(` zudjAVMY)4V+h<-YwXaR!zdgxRy?1Y*AGBo3od`@YSDrk@De)w%?nyJ%jI;?+JqhOhY{I!s4oX%Z<}SQHrLD+puz* z_cp%$8Qm2oH|kDHYzuGm;^{YOONtB0zrT`OJ+L-3Hgvu@_0cT53n#VLD4f&YZv)=O zEhq0t6c99J)%CZ_?7QOkerZ$A_-PSeTzyzocdWhb4-vz;Ou73i;2L?~tkG)?=2bhP zlGO6cqgSg>(*XZ6#wwWacvcuSAytTx>in@(R5j#{8(NJ|oea=TbXeBv-O+Hsr)g|UUXc*($ycSF z^7h%M)ng6<8_6z97Qq~_Y!p-8ynAi7R;wztR8v0Qv!x<_1?$p*VBx1X!JlkZ+7WEb zMqBXpL}Gu-y#YtQydyh61EqF#B`U41$rV+W+%c!-1|;t;T38~;S1w60y0NHc$&!!{ zfdM6JCkdYFUa}5iLhEX& z9?@Ymwqj$R5D;*}j$UifN2#O08JG8#&LAb57i(njq5_}jzKu*mqIqDNdjlLcU;q|r zga6TY(Xo{a>JxLg8nvJ!Ls8z4iB57ae$uWmXCbqP>U>Jz8gY|bUk|9rGVG%yYPHTsnW|oJ?42C&- z_EXy-lKn0#wGM-wq?t3xY1;aGFki2@)AvlM{8gr3uV7LSWmS!$PpTscOD^4Vl|#!g zCuy&W^!v?F89Ng@brIHBHo{lK$`iW*HO=Fb=hG9PO2FLhiK*8hrq1&q$0$PWw*93k zlQrK%)kkd4(SqYx_D_T7BTLPAScT?}ij+4bY?%zph6-M{i#;2CmpTCMr2IBpfGiKH z{SoKX)tKUZ)4%|#(C(!aJhOafyTAQ-Aq{ka6Yd9_IkYNLE!M_wT; z>G%S{H*H_QvydEf9`?v7e#eG!CXWTS$SHbV1knIm7opT^xH{$FI9e|-bY|dv91T-R*HeHS-u)aN4PwLfGhW=u%UV4-H|=EFh-P{xze3iZLLBe3-KlTE|5mS`Wymi|eKSN;r%Qv@Gb?_K)BN6I%EhGugylPXelsP>I>8L3-~4(GbdIh zsH8gG@rDZz4-bV(O)C}gM?HBo7C0!tIgWl{WUNF&12U^B-VqEgNp=ouAAw_V*IBk) zF^l9|ZuYX@9Ba>{DYN}6-TTNU%BX_*k&M5o1O1^3uR_*NX3d(Y!kzdz808!C28>x# zhy*7_PAuM!*v4$=~SoYpsReDF_fho&O zZY~qwes|Su?;j^Ba@;0VKZt%+_#ov#7DI0$fcyy%lY35puBN&S%nQS1mxOE#l3Tel ze9@vsy}nA=0Y9=GO(lpg9uk7Kf^8C&Rhy;SQ_wala&4vzbpszsfT26|4@kBH{QmXe^+eE<5SwnMZ%vut% zAqYFecZTUXsl|ZAl?J88Xl*#O(KhdfRZr}TT>S6|`}*6Wubvj~Uvud9AEs%0eBo$F zi0!;I!IKsI4*S5lFgK%xJ>H_Fp>bj7a*TANcZ1f}0z`WzC!}K=iesxjw$WpucAUEF zHR#;m5VN~yfnt7W5-EX+`*r|Y7-!^DNEq5lD_F8-az3k$L|k$XJ1})yJ?n|j7PQGx zh(t?=prtP-%Oa;4Jicmg$?G1fR{1XHVd?cbTJ9oC_EQ*+M3{#-ORE*3946Ys>h-)a zI^_iD6jC&HjFc_dxUgL%ik}HP+yHGKdCAck4M(Sl1ZAL({Skg=ZkY1xK6YBA(n{3ST;ia&^4eS$Q+uO_wvsX(NWTQE&j}E`iuj+J`ZI z*#;q2m8WNxM;o6wu6s1laS0pyR0DI{+x8$XRqZOwPG{Ng1-$YkI~~RFI+F(H4p~*( z6rF+*U;gq|8%s~~E#m4ZXXl4BifyxUee8i*&13U97`K!|z?$sOnL>wDgXi$EdMG>B zJrTXHvUJPk4WD#PQ5!EbI&+H|U$=ge_RPx4s;W<)ac?gRtLAj*#Mqpk5{h;xAVTWA zown7_Z{{(p(pKBDB(+~GCv$C$$#Kg1C6|kPSzu2Rlcrs^5zTrs>OyONe%wC$lieki z+oE1l(`6jW;>od)~c!SM;I&(a`g; zA3RV$Gj`Gr+r&W#&}C$N<;R@iq%P+g++%sKZ*X>t{~Y?+CKlFE=mkpFxTt~k0wM5~ zKz;O{8AAhme|Jm8gzxoXb1@^!6zOMeQ$0LYp#p)#!dk+ zvUe!%XNemg_w$MaRB1_{h66HkH@U!XGh7|k6V#(cz#=O#>p%oU^=MMWCI29gBS+@Ba6ev9Lk?M z5y%_OYcyN%6J&3+Zy}%(JpfQG$v2-q`c2G~A?(q7kc ztM|ilX)*QyRjkU6-ZP6HkVt0b*vBxZyFg&VB%oJMfW^&Yge`^C3TU z_!z{<27G`A_wIs70uMW2omVC^h`5?C=n>6#)0{%1vWUV38=6jELLf z1U3!`55UD;H_ZD^aIh@G$c`Q2ObXH9;v5CcI;&F%7grlZDs=OoYY;5oymoL!f7FvIQm&AoMlI<8#sM{?kt&Z zBefCNIB?bnjszPSUSjNm!WNprEzrN>Ys`snCuTG7U#QrbCqwFehet9hx&=LSj+5X8_{h1Ak;Q%NMl73boaN~fwKTGc$ z5Anii@N|%BG7c@ZB=%m8bnlkzg43hlBq(VQG}o1GaVaOD+6-(H%DkimubqXMJ#YL_ zh)0rvHP07kVTX8zFgpAL)MEm4hu32Q0o`K8WjjY;lz<2cBJB-gdU@Fe1p3KZut)9@gYM`sye%BZ2N9 z0QEcU4g#GtKzDKMqhVG*MjcCJifO(u9X5~sWq8qQlWRQ=HLV^vlJnj~+u@`bxu?d3 zP;=Xu*DHz3O&ZDqi0f-G_c3!`!;U7pw*M%v48mdH zcesyM#M8)2Zj z@yyb(3an9`Ty}M-96!>xc&*u?Sh484I?X>r*X}iY%f6_arE>RA`S7#$lM#N373bAVrkq$(l%JjJG}=%lbHX zIw0G9$=!G=zV{eoKwf$BnJU6QjZvJ1aERE!wl;-;N{0iW+I;Glg=5|v0<06XgAaM= z`~TVQ|JiN^!fWt<$99jB-w`zhq@3JAVBYU9VH!gTkx;T?{+ldWLTsBrY&G|=UPc)( zOzvX7WA4Fc?b4})v7tY|0&nhqFS`PNFCfPc9z-!In;^URzuXJKtpCfs3^`dNY&6EQ z;Aj3u?uB4Yocl%5y-z#!goiB6JFqPIljE^F2*=EitOu0nAi?&~+Yu)S8TAf@53u3B z1OtalX9CEHnEsPFgbc`{v-}hbaVGbMfd3x~4oONrMH07OGzai?4KA74(%Zd;jM$G4 zGJ3F&Fwa{K7*skNvHXaLolA)ajP#NF@#5$`jqJyRT#5+*TYb9XJ{l6E&uUp6FLvn3 z=uvE9DBt=^Y?u@w>BDcd8--&!N<2|c#=|p(&?h?Xu+Y!^d0iN6WRVH2FZ7>W4?}mY z4TpB*&1x`nH=4(LcM6l6wE%gCkT=(CPh7r9QwK$6aNTJUk@XuxA{lhW8C)_)G-Hs% zz}?H&=ky@ly)oNoFMuAwcTa|9`Xh5YH2=xyUXWhd+dD9Zfs^wPI+zKS5;2FqiaGS` zQl9)QLV;p71xo3aX<3tJqK@FB$Bre!AyLzwq9Y86K-a$VfcaP7O^5JNv@t?^F$@xsRkHz=4_^ zlvxB`q$U^K3esKPnYK%Jxo;}-Dh3tZTM~Kh5`hV}mF(2&T>J9@UdxS>&JHl!KHSs_ z`w!e9!$Ht0XDTfg>e#zZs4=;P2UZ}>feqK+&O)VGv|tQ3t4f=tbx*A7Dv6=Y@EAm$ zLU+PlI3Oubs6Ror!0Oz*b`{T^0+*Q_E(V!Q28u-7F)~g{Whin57&qmON7mLTw5C@RK1qx(~_ZbIuOFP@E zyBFK%eVx8UoxQ(c;r@bCT;04UeC95gB_IS=O2cI>)?91wj+TC8&5Ru*EbEWc9W#?`0w)9HhE1v#6FSC^F2qwpLDg7c$ zo`ftM2HHyY;9H2Lnn_uYhAE?wink1PR=7wztRhAQvIoCH@-icAMJ-c89fVsTut7?vojLbv$z3VPQ!Baxxs^T?S#w0F{E+Tge7!Hr7ao@j>~a?H|XEfyJ0uo08s zb%bF(X{sE-Kk>`nU21JTR4^UOCHd%5wYv(%!Z^6(lnJVbf?jo}dRIM3s*rYZW40i< z9L#9)nti>5948&mX>==w6R~G4D`U}(MuP5n^9xVBc#@=iGWinIY1&XWLn(J6aW$*x zp`5MtvFOCIQNuvbo^UBf@* ztp|$~Y2HpMjyirKoyj2&*qSJE-V{EpT~1Iwom+H)Y1x*`xl@T=fE+HVo>lm5Y~_Ml2zqkkigH%st>Xgf}=)?=#ac{&4l7h@#_Ziw4v3x8dYv5 zu?So12Z>m64WWM22QqdG?8+UPghwTMlDzzZ%F!JhvqCA4y3u{N*oYW*+JKAg_Vvfo zM~d(RD8I;DmUexK5D*t|@ABy#Te^;rO=$jr=twRp2Sq`#A3%bV>3ZYP}=vzA%a@^k~CKt_QK=QmJ2#%gXCk@eM4psj`FHYlrxnnRTRVBGlxf+_8^?=dZ0mM2UCsg(&Tr0=$2!bl3*IexA4RawGKw5}IQj$a75Mz!6^27lR&75`i zgGm;N;N3fa5bSVcDkr^R zmI0)qG?1*xi`McJylumiO@nx)(HJDwB^w~MxzeMT;O)i}WWoC?oC5RX z=88`*9EzJquec6!x-$Bu&R6k?X-6hBgPsf4&&x|5atB7UGCSP?j=n^&Q95^$)$Jrn z1!dKAqlAe=r9nc+5aXwk!Ca1ObQ$VrI7g0rpu*`mKkjPbo3s zRKlgwfMB|)>YJF)1TxH;E|vVnxO^9(stJ@3;%rU=ZrQiP+Yo;NqL712Ov}dcONr%{ zlrpWv#VEtD8R5&e5pe+T5>hZQS@dz;6H7OmFeOAS5Tbd_gGWRo$;@@;XJL~+%@$MUxw@)26WV_j_~f51x2;j^`M;BTx)i2nXm+ zShSI!Ac@LkXe(U`TvOIu^3!OP5Yy0qfZ^HY9Yhm>DvUaMZ{hynJEjExjos*5NZ%L< z5)3n+i%Y^m>4~yx+tD(D(4)DG?I7em>{m3@Cegv8&7aOCk*`#xtokjbjPS0$8zK|7 z@<$U7eFgGSK#IevjDv&?GEO7I*;C8%qyT>chvScJQhE=m3-U&^U3Ab zGCug-;iE?&jC6zUORkF$;Ksicn#}Kck9~h|rhFOUzY~Zf(2lO)XNmJIh|?vu9L{=Q zC0Bgn{J1yE(wTH6Lf-_G(33Jf5*r94QrPwf1#ah6hAwsoq8>f82JwMjZvL>zm7wuk z4tD--mmTc0x}R-%I+jj2{;-^^lA15!34tSNT6Q|Ag+p0&8;=9NBhdpG3&uAd(IR$2 zFhDudn$Jv|h}$U5HCXCn2gIk&AMUS=Ma*0pZ9=f$mgH4IWNrFXv}OeFp7#C0y~B&U zqX36fwiEeC0wg>CYp zzrNbfAUmOcd#Qvch^^LvR_l1#h|YKtSRe6$e)sl4V?w^VOv$&C8Xj(|Bj+VKDszse z*w9!J$Y6>!J!~WzMq=g*nd(QT3t8(&W|W|dD^H2oSGviKYOxDT%6&an`b!!6lLWykN?#ayyBm%#7<)=;SXZJ->Sm zcXs1>fYi08sI~Lw$E8k|9kU3r+Ej3+>~nZtf5Ep-I|#(yUAgzT?(T+)JSOPKA)$QP zgJrYvPDwZ%TOV7Zty?Mh>uAgg05ncy>$o{=vEq-HC(Mtl4|5bIjL~RBdZI?3QYQgC zkD;Rgi{)z?RSNR`I?bX%x%N(N_ZqQSDq1XZmoPI2zyQ)O$!{(o*pIo%OP{$qifVL^?=-)O9{4NE}BKL*i+1x(gR;=m>Pvbxyc`6@D3LhU&%n%y<}F zJGew-Z}~D@q|DO-+)l_!Mg>4vGt#TafkE$`G6J<+a3ju@{*3GlTG>H5boK1R$913E z%QAJgsnGjXCKB_M5STU?S3UWg|2lGB-jK|WrWjj3%FRDADyFx{lfgf{uaX{?f0uK8 zUXp8>3sd8Pmp*X6lEhBCxFo)5;*PhLgCbq9j;C8se@TU=OZZassTiGvSx0$*CSqM@f^Z$--iXi+R9w9IJD zEJQK_W%(pOgQx|NKz>fx=%xi6hd!lhB418SJ;VwtL3=gpYwX5poU;NDbXxTa+x}o* zkxTU7Ibk5OLg>n^xgws#G>C}HPMA#rD`~v7j|ox&a&`h%DkkFG60-?7T7qR;lnFTx zIz0oKjUv-v)C5Zh{O46>u7P3o=Q`IO+nQ%zB>j3=B9MRvo5}nu2*N?vXGifKQ_&5e z^n^APQkQ|Pk_?+N;^kr3q)EWI%p4Eg5|i!#LLMmg*v!f}u4`ngti)s-xgWWtN{>}> z4}cd@oRsuok+_hx3}q!h)8If5WY04oUJS<97rgT|gI{m??;vKbjh=v&@g%tQDk1IQyeW zNK;4EONUGNhtPE8t+x1L6VEj!Chz&kD8R>ta@=(8qp2!6V@@9c>_k_f{Ox&xJFT*jQX_mS@D(d1yGx7ryM;l-aAJA~J`ayL!j8;6zG4l>I=W(*ex+VfGK0vFUI{ zNO&Lg(G26B}royp^MONATMNaD8MVI$g@JQMeDIKK|li zU{;ky>heFBbvWb!qYP)`6;L&MvTKi$f7NArfRa_Foz7{a76U7xtFl<@7L&zt7J@`A zqJ6+K4n6$ylmDVt7om~AvCdy(RsB9Uko=XmWXDEqK9wDDScP`M3x!X^xN7>rD3i9;R zF1XVl_S3YBU^b6y-LV)->W@VJ$l6Gi zjNg{XjAIQ05B|Uy*l=PrgUsn4^O2z^hID4bTGblyG_E{QMa!5S8PwSg-f8pm^W1d3 zKoEL=4m+^N6e90FdP#j$Gp+7S>>0_6^GOK^z@ZRB=MK_Dd1HTAX<%>hVfU=`Ix|2;OuOmI&QQqW87D!YFC03(?bINO1Adz;M=NVBPa_Ig>XK>?Gv;u14PmIo6WQBs;1hUiCV?zK*=^Q) z^5&tM;3#5rkwN+=C56*@;K4$0?KEyUUuK&KoGSBGp=dl~dO(245?6oDLwFH}J2)D{ zPfud3L^~X}iPS zSqnvQxnp4+&wn<2bGXav?4l8RfC>kp%fil7)8)7kG=B39w(q)GYdjPRgt?^zK2*tj zF;@nsgwY!P(t4dlj8rI__8!Nd1*s?Ffmc4MKg@h?9;h zuxIrPz@E`o0;J^>xuZ;&dm{3(M16YbsYVf`gzGOfPG?~S>6k^A4#N8zPK24LQFQ+H zaR=#*8v>z^Cp!_n07#AjpU_(>mE74Xz@%%p3q|vl_p9m7EZjXNk({%zX~AzUycsPV zx$^;Cs5oJl;MCD9rh~3ZsFK^lgws&pgXqRKy3yx=%7lSi2{mjSEfs8A=Atqt>S8Kf z{Z)Ki^L9O*{3wwHjWszWM32ta*-WlEqCzNGcJXX4hOd~3fDYNS>(@a$m?IcO=9%Yp zVC-cU@;INRkAncxS{&egeGm9^3$qhS_UMN*Uu~e*8h^MzLOL;^2x(`-iJJLHd7v2s zr7emrbW-yHTW(TvMy6J5oh94kE{e% z1J9?6Y)0|VuS+)l_i0glH8m)w^SM&=79*QbDqMFhduUaS(~Wras~ehOQt?~|_x-d| zbT|BN(9GfrxTpBE*DId}(n;6=tqFA!Wv7P_zEx~rz91b$nq=-~@==kfu5vlp_@&@X^&G2M|1 z-P5W@-nE{k@BUcFMvb>*CfMe3;K|I3oo2Y;Ml|62N}GD3a#7jkBBqOKRa&ZNyIj7PD6o#% z#`7s2M8p0C3XyZV>e1ag25w*39kt`hh$xD3H{%~f)F z+DW7mQ%TmFWtxmOzc9~fQI|y^Y0+{OlNVv)wOf?Uo8{wOuf~YyLI2ob-NZ?(Y~0H{Yzy#miEYJ$eCDwp3v8Jq<((Eok&9B@Y455$ zYN)s^Vya6PJ6PoecFGhvs!2d^;27JK>xiT15uDy0%m0IOR|7@As?R&>VRTj%g&+ML zbO0Bzd)*X(e48GxxF5R2-M$*9O%>FEYVL#W8& z3t9RjxWZ2fc8Gz+SqU=UO(QhC{`>Q7`ezXyW(@CXxo(8MDSNt>x5+QPde<%TEiCxQ zTQS`gCp@IC>)1l(;~VNZhZ{jSS0O3V>Mq02?q-I)gVFFEYf^N54p6L${qkTB6xzo1 z!b|lI8&tDAymB8Q4w2m5fC>Hy@Ae6wbPKL}cN6oGBk#&#o+!I2>eX8H;q`#spU4|t zppofp<@erEao1$+3*?@ZriF8nZO1XMH0)33>Avtj`Eu4XE+k-pf-_*)d@u0nS!+Or z3!l8{-Ho4~mC zbyuxrzgac_KX7nl*Po})!SVQb?%B(Hh1}@LWoXA}Y9~va>BVC~0AsAkX$-H|kaKIB zCq_1M)o z@S4b{MVW*_Ku~P}!#7e%sZY-Rayo$Bf}x0%ilQZyIjJAeBusMM`kvT-e(W36Z`S1O ztX0Kq{7(*V*tg7iJ1)P?*oivCk|@aaluqX#LB2S30=s}Yha8%FZ~rJ!PHAe_T*Xlwk!5>0lOj_HW0ulf^(dM>k(V+1suMo z-9`XHsj%ndYw@mADT@1j&fN(-|2o=R>5A9%+g*5|$mpO9J$qOrErcL)red+3=4f8L z{C<&X@lLPJ^(+JBVpCduMoTF#0Cr43uwWytIySWgcQ*m_s);@2RinFHmFVn)d@;ST z8Jk+zSj6+okMAS+o+Q9l6h^Q`9hR?k%P{8;u*SOIzKJZpNjW+{(z^33d~)1mnrfWY%%gXxx6v%DUZVW z4@FN09(|S{%;+w(pyZ~RS}7hIL13;Ab+KAi`2e_>l=WHY1m%d>>UemF=tBFeNo~v2 z$n~I)IyBI4zb>eE>U{Z3p|;ES$XWiTJlFG59ylz%ikT{|Okz83xYZa-9Wn+fdcha} zj7|Dw4xn01ywm9Og9HIcjsO;_4fZ61U=`Up0A$lSK5pe6#RJ#~5lFeL!$rkTn`JvM zd4U>CL}-cY07&t{cjmLxo6{Zu=yefboix){>_S!K&#;Vj0nWkGu_%q=ehPEIcDRI4 zNTYxU>~(RzB^C|i<8_RWGrhUay%gsq*k*VJ?WI-JW zL3sCxS7E+mG$yaS8vr-)DYL2_Ak!)c7T6u@dXe-r%GdWeJ0J~82m$VGRqBuD0wrv^ z?lYAmpHA0VF|;ksgnEgr3dHAR@43l4Dn)e!Yf=Cx2vsFRboQU_5oI&0wqsm&9Kr37 zGMTSgtr^GC=N2g_iMh1dw-88W1l9wc#GIe^i z_LO*|j{DtSddG8R$wxUO|F#c)je(rJLr6@1gu7MA92JABj0>IW`FsxNbk5IH0>f0k zeISxj&-sYkBO0HZ^0I5s_!#^`#6qlJVYQu{hUdg&IdW7B7dkC{gM@rjF0S+34!rrY z%b*&7YkT(|uef#RTC4{aK>yB*l1=&_?0f87QpJ~Q;=p|bOW6rX&e2B_h|k_`V-7rd zzdvUc@N&;|wG6{9bXqUFpVj+_@EWV0tP`)ekLmPUcyiqowaeBS*CXG5vHyq+ANqlWrPN#$mMDVKLsuP*LECk@NncS1ZOq2bATu+ID`}w|b zPDePUI_fCAPu4lthKx}0$3)sYq#Ej0CyTf^ipDcL+}5PdjK6+&eD7o^pBT*XsVjW* zcQD8&f-vY61p3%dpeM4cP!{o~*X_?w5R^%re`ZYgm^?P~$uK4hbPJEJbD!_gsK9Ow zgoiesoI^~rwm#)m*BQL+yvQ`8Zs$bnfufh=0MBuhQwNZ>vW!{{cO|sPU~cZPlTal* z-hXuh!E4N|&DIH|14EiLf-Uew50Gutuwstx+CG`L1LLoC?-cIe{tf|1S%uAX`p3M} zmUZVEmFTc3U2h~|eG=C1ctAh&XyjARl@Y5UTr0gFbI_q`E(!l#G`*`mi^}8N&Dx+V z)6)?o^6CvjGTOH6Rs;u|$)tp{Dja#h&HHI43;CXZ6e00 z=W#dBhG^l}n|m^i4$!hN)K{ERBvTDKD4bnU$_IbZIP0?lK!Bc9rC?bh-;@f{s2Pu3 zobR0wdNM2~x#CdxboR4RIt;jYD6(!Qy1{`x5MU(A%AUHtQ$9qEHodl;jIG=yqe+X- zgepBz%3tv-09i$p?AJ7V=KF&)A)5JBW)6$_Q|w7v1{(Su+nzBm2~}&NRyW zxz)oyV$~rINl=?k-N z0RS$MwhtW4c?e8)Q%Y_YkzkAG8fCS%C6hh($F27}Hftm8I|mX|Y_*YK&&e>o*{C+- zx4ttWn?=L%fsM$nYzb}C%of$_CLXPj*U#uuh(N$9mG{h36X!3cw4cLZPtZ)gPb8+2 zq7CMG?A-v?ZzL9=d|vqz?_HmCDy5wz_J6Ho6wDX(EEzdVzG}E{B)28HGu^&>1UqEy zV=a_!5Q|A0DxTH6Q8Z`gKI$_3yE8)v$I&Lw;OR}gvm(`5yx8{#a;=~iU*ac-x3>kA%S zXzkV1(v6joA0N4FHWq*TRb;{&63!UxRjTxgQy0mU19VkrSg>OLXG1;<4g?)Ki2SQs zEQ3V}e@LpoQLNNM=jrwp1O5$6m*D3Y-69CMv4hc;mZeVqAOLo+-yn|QxI)g}O(-li z)kxvB#^`am1Hr42_dS*}7u{zy_ex^;c&q1811AuGN$o5q@du^Pczip`_~9k=aAB~5 zfkv_Y;k8PD+?$?2Oc$(yP!zL4WAT@Eo0koPrT2-!utMpeQomWgGZl}X{|o|@`^3;G z@%m^vdGLMKN6JL$#p%V@Tl<_BF`cZFO&?rCR$K;8jRFP~`MO_hQE;0+p)+k?SwmH7 zHLg~8su<{RMI{Pk!mPa0r0dM=0@&2p051Uq7DU!2Db96M3c)ieOvWUIpOfG1K_UOVn59c+uDp*Q5Sg+vW1gURi?e0xfpA;{YqVmYA2N?zK~F?2 z9-RNn{q-)ZhfogbL^G`*z-e3o?{-9V%B9Y4JBSk(a|G>~;dY+R?U2{7RRl4AMefKU z`e6Rn43h?_u{6>{BvRr>LR>5&h^2)#&5Q#Krw+duRm127fI~aJ_{|+uf=tG| z``Rgi4a)tBtTY;vchu+}iBrFm^f~p|pc)`L!+ozI2t)EBVoCq7x;6kIG|KGV*djY`yGE`%sQLfgQh3`6iH7+25S?gl0r-zuZZ~xn7z6te?U435b07`Jg5B9xia4~Ct}-v)wukQ33H=iljrKZR}V2M z`jw>>O-3=3H)HHn8v%Sg27v0PjQ4NSvXl$jkvGH#atif%2(%mA8b4x!vjX@^wh-PL z9nfSTTO1|kr#u&n%u`ZxqXsSQuBhT(75}Ii0ek7Q8|}Qsam7|!CGJpgu=kyxG4IWg z5ux(irf_>y4-}_qq;q%W3d0G(W}5eYM{7I*d#?FU&hRQ9cYkdv@Xi*@4daW$Rd-p9oM>1EpJZ79cA_FL1ser?z*JY z>ZZ@ItuDDXh&7o0@h%Mr*3Y-!P#_J{rsrnu3NDT4O_#YDv()s051(~d+bVB;Zl``_ z_oOcl>;|VJIC)b=e5oHM3m}-42#3Mb9C`I|#*?Fbb7oZ>S220sF2Sh?kw;Hn^{K(s zqHSQCeCqgo2j{hTy<~B?`}lxcO8Ohug>wlar5f3^cc=aij`xts%g<_8yE@6c%R+K-P~w(;|Yf;Tuq^u`h-ocq9bQ6g~0^JTO*efcQk3W?NjXW&;PL8q9$5a%OU@I6$jMfwNH4ze#XsP+o+Y;p&uN;XXx@-B!eCo9qMk>-L$TT+-I|?|`@NRlv4yz#p_5`Y z5#Vu_<$z!%*#Al06IJV;PYq-!ihmG*4$+xvpy6i!OH8pCkb86hpji`&I%G^Dv{Psu zr9Blq<2{Lxw-<7jmg%Zp7g8!Nd~P>k8{}tDl|n#eD@Tl*)78F;;R6e@KK>}MfS{hO zGb#@sEj&;foeprnhp^C+(}1x56+ zxp^5VI8k$5ad^7TN8N`Ju#FGh@%}*$N;9{_m6hfEF1yRJxQd38 zGu~M_5|0qRwQO#eGOmL2#0*JYJ83^?M4Nz3^-&JR1zV+Ir-S2K&`PQ(*W3rUiMCL8Kc$r_Snsx7x+Hcq@aKT9a+aA3iW7w={d%3PJ8@D zYnp`O40v#LFCA&PksI6uUUWmldexW${<1q=1m5(15PFha+Y0mc#O!+ITkV_cPKPR` zUk0Ywbr-=oW*LzVZ1T&D&6wF)nGIabDYox6FKIu;*0ZbPLY1?3hqmbxyQ(?B#}%o3 zSIW93c!Oy&*#mU!A^1_M4`+_)8d||AA z|9xDZN@%xyKjyi7jhcpTB z$}2|gbG|*p2GXy(0Q5eeyS-Ihs^}=mfMA8-;Zri>{D)(&OlDVbUJ8DF&hqOuf0hRq z?Xs1N0P3zN!Ny}uM=M8uhg9miKDd@m=>4&Iw2P}QH`&ytYxdL!I%;Ms1SN*g)6LI* zGHO1dQicgR8+6w-_~<3=3vOf!{$YK$_KQ7|v!=Vp1CfyLF@qtamG3GMKy>ApAePgD zqoK6j`XrXs`5kRfy;4z0+DYSP%c~Q3w}gx62iUAbUTgc-;PnNH4*CobGlXggkx=Q{ z*Yb7jx}hyQev@WucJ_gwe9pYn!cxpxth6VjuNnuhIH#HK<$|rNo?F~Y0F@!wK#cfDAI+Jk^KhM6idnn2>&D6j!o*%%w z-;q~SOO5avkna~`7NlY+4+VhI@r0t-pH|DZ63x4ks}-6;T6Ici>qMkVc!+}H&muVQ z0Gu?=D!0S5!w#clm=}kfyVXsj)auXYB_CEZxSi}`;SLe-?+(}gXpxfoKy0NU)bwiK zl0f_`$@j^xPUH0PGC`h6p$yY$M*5@LOWKw-ggbDcqMAL1FlZlKfCJdA{L%j1Dg(oP zfI6C{yBOUY8gu*Ml^Y+(<{@@HnEj*X#L)E%TXG!SmmwDZj<>5y=$W|4utv?%1JaZe z`2!*0qsSeI+9$U3m8`Am`fitx!)I>}mZa&H2=cRI-Ko;okX0#GOTAx~bvF%Pnm)i* zH%W{AxM#)s(9F%jzxx<6KC#_D*0+;X-1@$xkE?LqT*xaGjqAxSwdT4>pXxMmEbP?E zIdMYLK?w0J+@eaVLJlICbbbl=2_qEg&Xu<(a>P{8WeY#%fM3=w45&0_ed76;u=4A^ zZ%tJSo$O?-YW&AUJ#PTIlu!GkOR{-_3tC*j`RSbN(f-VbDPS!=q zGbgGRrYAzq1fISA)c*6C5n$?%p61TW-ygDCiQbeVJEjfcbm`_jr5E;{Id-j1l{TRa zfgf+chxcXfv7YJ{&X6DBd1NdAftXlH-4^RZwR9$OVG_PD9gskLHC0pPi6UDdJK z<;MrsCb+!t{_PhdKw{uH(%-C%W&)-b!;2j#ih+`*K5;-*+Z!7-;})+qXnnuayXgm_=xX9UXPjK$AHDrt`J4v^uhM zC5NlyB-^o*1w)ey{5uN7&P$Ow*pSBTOE`L625-bw9r*PgW}M~o@Eq%}Yr!r8@MFP3 zhlAWGThR#Nx*jV-6li6AVtc;Yo=TXp8zQH*V+gN-+)Kr;-)S`DjP9>=(41mi_F9^m(?h*6$f{ccAZ^1 z*DB%ex_n=j^Ksabo32SQSZ!^vEHkGsSo+fk9@aK{HrNo*@UV5@u#|v%4O>z|sK>Zb zEteAwcfg)D2%V1e(I<9Ks%2!gd={R1(h|&FjUaFbOJ;Yf)J#S?kuF-$A07;{iuhd% zFmlLf#6ff@rkZDG5c6CXSb-)gR3Q!3Rb+PnMq4GoNV>kuvGHGBdyJ#cXuMV}XXJ(DIM!H6)tD!=#(m3>XdujU+Q)Ex_s11fGrQf6{jg;R(O;+euE!@)} zQ8M*rwK2M9?z!DVEz66nAtHXb{h)(fTbB}U5K>-;KoPI{G%xm$HpC8;=}22J;UrY+ z(|6Ba9!oFvY?Tl(@b~_DQ1;|0xlN=^$326MSoR4G$B|hM0y1>n8%7qazqu&S%bW~| zG^{T^(?C2T-)7(AGVWN0u{oEn$2NbCBl6J7@D7zBoUGZuX<~RJKkcdW`I^rqzTVXm+h#% zTItsJk*n0bf?Jq`sxsqM#syoP`-rT*3}E`|S6d;B?G=c*^b~W3V1sd*xTTmvBYDbS zi7KK!6}_Z1l4Z#TS87q9uq*-9@>UMU3+>JpySN~g+IjkmHCV(X+BZJ1NHbN)A4Lw4 z?2cJTsu}OstOH-1%p-uFq5nN^A`QXJt}&P(QgRSdnnc1hr(4XKj;6UGeEErK3YL(Q zfxEUF`BJ}G?Vfs0k8MA)M_j_6Tj92!_D5Z(?#Kt)7B2=L=6<>H@l}aN$*5XwdS!!I zM!1ZKw_n-S+fnhCj~kGmz1H8<7_6&h3K>?bq=#l_v2uY2uRd$19)$Rf)Iu-9Ajguo zM6jtuc?96*4{I6U8}yPE&z&dEX+8A-g0d%aACA4g_~*+=)oI-J>`yLlJvhdPr5MKE zeGfU|RmPS2yN!_lRPQjVuhJec@(pC(`vQMuKAC$QfHh|-Ru`O0TeJ-qzgp*ZJYP-( z+c6_>N<~4mrFfzK!hX5f)h|Or&viWJTbj5$X95QuQ@7(nXzhQOSj8D^uS?y%vjtx| zqYggF=dtvTand+>bf0D%JKX(FVZ+@ZfGYUN$hB=s=B0|&jVR+F1)otsTaVYc z!futHB{Oa{XtJD_gmC=TSsk(!e+6DH06^~uktA~wlH74KQ0%L`HMge z>wdT%6gD+!|LCZhppZ0nW=bSiM*~503r*$^@O(~2Vb72_YbTE$&!DI2g7LP13{5`+ zY}8v?a6+?ZnLb5t{PKZwrPSA+WJH_FIHlSujXamQi$qD>4xbN%LGjB=F)Ju{(!(Zp z%(j}|DZ#myo+adXDE%ItkIszS0vBs%id1ny6f1S76WWP#9IqvQ$PFLT8~w!ughCY4 zUtX%5nn%IBxHhnlN)ESrnh+K%rF>G&nJ|MyuweYL5N68=XWJFy$AL>^*m2I8BNSi9-5nEWiFemy=VJXs*O_9!SgvvZ~y63hrHB4 zHG@`UyV~~%6?tl<65TMI(lWxP&lQ#(dv~v4 zF$Qe9M))_;%-i_X&x|C5H$tjctTwL|xW}QALL+F&RjcI3Ms)p{qh>^%ajX4P7puIq zLI`b%d>g4AUUu#CrBr^^)QS@xxLDUvw0$sUsIcu2=gZZ-w2Tms8Rg>c@K$p7V--6v zjl7?G+M!&YQ0CCM_w*-j0^4Hm28coSej(-M^jmBWHqsV;>?3HJ&o|A;$Seqj*{i{)ueghwBW#ZLhVCuCwmRVKQzobv+{{lpD2?*d_P6SgX_Tvx&sYxLM9p zFN9Zx`a=`#$h(xJ=_85u7*fWk=6146iYJF>LJBqs67Quhm&W~mfVY$$e!Q8(U%caj zRLMT<<>O%uj|$rzzu!J2|DJ(cLi`&PwaC?pE?&WB%faV;NKbwEYIMW?6IIbGNTKTj zcd~8Rj9Z*u{9=Eth^q45J~A;+#KmA)kXILC_PUkq z!!b%OUB-&$MOS0TUP88>DtoM2b9VlNf+x?O8A*R8z44{Qf9k@##pwx;({cx7RNKXY zjGZOE(_R2^D-=p-s7?u?G7p0*!IN^{@J0bhY1Lrz?4{KvY*XNqF zhy^^QYQey!U{IRTrQ~GqQ4K_f5VxSu^&L_`dJ2d8{r7kC)*zq20V)KpG1mC&UZOQA z_EsA2DcV>W3l(32oOL*T?2|lvgfh)M0*ZE8Mjl^#dI`Ly6F>N$G5ks<#3d0t)tQgb z7{Gsl1md8V;tCR;67(duT+<8OP38N3_~`q7%Z#Ra*Rj-x^ykPo0~KV%fPp)duGwAD zqN{7xMr!lq(&<(*X$7%mW2}2;YF>{}f0VN#NW3o7{Z>g(XCV681Jw~hB}?9T`(u=R z6Rt{{-~S9nMzDJ$ls%yE65sPdb=(JvjYR03zxs594N&DBQV_9wbf*{ls$B9{CxTA^ zi33o29|P1?jd<_aXBOgKMSblu(sxY|*b?i%b$Wm(NNS3hj)G1i3~Se zy0={zGI?52^D#QM~}M8SzRmGug#_3iY3O3s_WY?JG=V-XUeT zSL=2VbMfiNnP6&|2s3bi#WkSE@b+1YhQ8MK4eGC@U@=TGbEIxT|u-6&g#yEIACpegX7p!7Tvy1 z;5vK~yEqG8UOdSey)q2DggDzxnc#y{lAcgK@Lj0iYOa?wJ{Soj5ONmpPWemy>N-_D zUuK^ob2(fd3SkYB!6>_9QAmnq-fq)8nk5#nm&&&WY`!NGRnt(?aHdNVqu?KN7Ymhr zbC4Qw{S*@@l}EPU3|ze_hwqUZgXUr>hr04Wj(qNLgaD)2*(6lz1xxq+wI1eQ_cQ|EsFag`?x8&CuB2&n1J zNyTgw9)nD#rShVkk|R`)zC5krDq}!J2qhY^R!Q+-@r$`(+A&$R{5mTWHeS)|Pn4Ap za%S$v8yuerB#DB`hX)upomWORN`*MvplBt37`f*mQnb3L;;3O}7vaXyufb7?@Sj?!2U6(t ziX?Qjqk2jd=?I}(LS-N**pdyRVL#)0s-uTFGxy?mH%TD5GpPRpA7o7(T%ky=%Gi>U zJ0S4$nV~0hZ})q0uNufFv0%jmVr~zC?2zYD`(1YkJt%dluD>um>z_fr|6PNZZ5l{T zd``VbmV66{h2FG*ida?O0L(KvILIaJ+7OrovGiwu+zWL-hALE7Zk;e7gl=Ib3w&Fb z440yD8D^&&WNs1!uZXEh*i18c?69LTH|OZ!=+zvL6E*o?nJ*>-%sg|Cj~ADe82 z^J{@x?}zR%#Mn@oPo>5Dzl-TW7bA9IMH*ikPWihu{Elqgu(#5%^)Ju+EpM0$F#Of# zl9kbvrh=iT(QiPgT&oV2_UCQ@UWi`j3b_!tk)GG%ishh5GqL>D_;?@D<&91U1gii& z4&126;R-9F%lJf?19=NozDE?VBF=P&g5rObZ)3FKP2(kV@SWM+kpFNzF}%`A?2J?Y z*){=n&M()=X_){39Og8=tT!I{Jp$lY0xLNTq}Z?7SY=^wcG3?Lw5D#r{GGHf&UuTw z0M*u#-vkUXwM)P{cX=e%SI0T1+Zep(64`&$LCN?OQq(jSQ-#opdPHQ z7ku1M2!i@Dd@eO^2@@(_1%hA~aVdlzG@mGx!*MVqO|ge92npCj&(JM61kD=VFgP^O zj&|jL?P$MqaN=>jNE7LNS;tc*Y9SFm(#uE#f8mblOVS8&nBoy+A&|IJ!du)Q`cPy< z&J_M0LG`QN%iU<jAz zmHz2mRQXO(|2s9~%m%#vBpe+q<&rGX2fBaDuU{m00iLSmfL!h~V2U_q_yomcO!s&R z=-m=L9{UOqT(iCn%;0ntm^^M^ybGt7D$*CP_rcO&}}5o$i0 z*E}EJ;tl`GK7OC4sQqp7Yx7CB3+0j?tWLis4*Wy01`)tC5Af3|BG9~AtJCpkcO06Q zvNo85!Lj0@@Y=3>pJXkgV%^l&<-2P;L6&J~8~U@{(2G%qj}+``xD{Z6CF34_dwcQO zrM`GErXt{#KN*v$UL9yER{}rDQvCOLpznw?i|ujOfLWo%t-0iC9tRQ|f(!EmQXv;l z+%O@bqa91%}fbwJ1^sListi!spU%=xaNe=KT>2 zkZ$nc%^M^s<)Bo&v=LP`au$5MPp-eZSh^Xb4ldU?CN~D#$;JV{HCk9hWq3flG7BY= zhqaEOK&?4c-Gdh^FLUjQom3%&Zkiyd6~h3A;;Hma=w7(T2?cFPnFPd|-{eoK+|0HtX3?Ivm8NU^U2=M=ntm}kkRlBKlp$bo{2qnkz<;e~RtqkLuc8N{<*o{dDvcF;MOU9ba)(R+~ z8cJ8NK1JnPT4Q^=3$yt*!1FpiQM$RI_0>H7M@Fc}$vqb#0+$~3T)W4+0jk#@!Z)+# zR(m%+zvy_NDvw?C4vJAw^cd|(7}S~i#J!i2vt}q^>wuzqeIU=3v5D)50qPONM9Dpx z3lQY$47;3t^&nFiG|J6;9U-UY4xqnd_%M#R`DF2Qxj7ikmr;zn338r~CfMh|1LfUC zRn+$ivJyvku6b=PdYM0eY+Cc?@HreRj5Y^9kXGp$)Y_tq{)BBa2I@7mA)J`lh_V_C zP!ceA_w3q{E{`N7!=V@yB)mM1VhUXu@Qb9U?HOi}E>)9bi`xat;bUi7L}BrMrJW zA)Vc}GFznc&}<16OoNerHP>DVCt<*jRnms+SC)7ilwI>O8_ehCm2j-1qHxHL&icLi z4sqB?1u|O~2y8{QRmm&j>ul32OD5l}PL8{w6acEukD4b8s?yX}8TeAXmTuD*zou>h z((^CcSx1t)v}Y2p7|~b((Eq27A%=8{7j}Q3+3hoW#^)x;rQJKNo37m3j>~D)KC%9d z_3Q^MT_-T8^p`W`XP}tPE&5qRvKw4%Q1UFQ$8`GQuj{kWZWQ_|6*VVFPP3lBU}x3I z^YLU;F(-ilnMJ^l0F7d1h|*piIUT$Fi5C}AUk`OOoqIn?V9aij@>3~sy~%BmqBYSE zq>nJ9#Hm5gQdBN;_95vKWQKPv#!n!}N}>c$ViNCwv(!7U-ldNk2+|GOCpuTp3)~f& zKZQy>U3^U<+P9CsKN?D>#Hvs>e2a0-KnP-=S!0mqdEfwczUge6Hom9~9MI>e#byLj zl@^||WywWtuWgl_Jkx&mCjL**x}E?nV&jp6@1T!?LI@aV84$G_v`CXb-$??(4vW!;m zlMbf$t=ip`tw-4xEhxoNgW5T)t|hkKfPCz81YZW_Ks+QVzl@!wMvSnsJ(xENsLs!B78sd?u47Oh=sIhfBEHn*) zI)xU(h+di6D5ODuj$x^yfV0>2V|WbaP$hiVTeISW+aI$$(2QT6$he6~J-avv37&Dy z%=X7-uYN$#iPaj)qB`=JgB&25#v7_ho8n362-TVy&a$hVng7mG)!${fp2wVd{u4 zAXM-eRjFNp;tc(j#)rz|VdN&LbRNWke$p9&wE00h@*ape&5{h|oJ}d{bRy$?xv(~} z1)YQHFbJMICeMS>(8-ox^EJ&w!R#74^?H!x044|AnW>~C5V$CgG6w-khcEJHk-&FM zchj9+1+CKNSEl5Sn;+6EqWAB2P6MgIWyw`8PPW@Z17zxc-Z7xX>ulIn^N3KdwE?!O z1K+!+Je^j+%Ll*hDGJRm#G?91`|9|mDR7$4luTILSoYXJC_x@2HX0%6 zb#{QHeBSPMzRsuV<%D{beau1dz1mH%Nj5E2 ztBh<|XjU)LAp%#Ix#Hl}2mUf+bvfAWxY+gaYvvFPrOGP+>57%GeXf$|R6Ia`G-M%Q z3~5DFF%e>B0aZvBQ(k4tXgjv3sob{dff!a2c*AEF1&bX0joMN9*0?8LNV^QhbnVNT z*UW)hKt9psZ{f9;aGt}E0(6)vSsZ-Riz;LeL%E;nDh$?0O42%)k136e`kxrWI|Jfp zARf+MEi#NZ1=nCU?RyKRX-05_P7(o&u6dysdvzOL%!uBFzYywCf)K%wL6O>Z%bEzb z5!6VLa1}8eV!+Gx9(2+R_>y9al!mFS@L6Wl*hnfDz|UO zSKPA&1In@M6eMhykB(MODg`%k8t|elHnJxc*HMz27uuQlx+u9VghjQ8`r_T?Es(SD zQ=a_%;rl^QMOS8@382%3-S@tyCjRv-XdHP z_aZwg`#p0Y?-(&vP7Fg|{DOqV-y{ZA@kqHQ9KxYl(^Q6F?ed2+#-5A4?nn8ao3xuY z=W3L-5sNs?I}B4|NL+J#Nyn`X#$~kh6OZ57)eRpf#aW2-T<)8L) zGPYIOxXcVC^g^TpGa-@w9Ae)et*fb}i|N6((kcOy^dz!9@!WwrQt{0*^Qdt&2{^>L zd{q~o+#D+B%v;I0X<}vblP4P0who5E6oA@diDg19iDd^+jdfoF?T>auVhEIl3T8zn zKZ;cP?%jaO3_=9^{q`Dmm^`$eEK~{lRFV3{sn-Kvk)a+ipI?7|58q}Hi1(zk>m#Pg z7y@>PJLz^Ec{3zBW!~Ba60Ye~HQi78`N!N?!M(d=)a?u{4GI~zRB^;q!J?uNrGA)- zr*NZlCiO|kQ=h0}_AZvo=7-z@lEyzGtIPDvhU{wG$2p&_bTZ{?icc<498IBIh3Hum~-eM<%y)tV^w> zo_)^{B~egXet*C;w7)sdb)W)^KKPhoK6G=7G#y=-@zr1Q?d!Bmw-$Nh`(nHf-J%p)>#NkJ%mU2=Q! zyt|xHDbkpB6F#mbWPn551Oz<-u#4RAY@j|V_dFP?&IDV><5RwD$FT%j6dBg7v z+yUp}YHMqayjey1Ncvpr-1PD*Ij-9=)IdSe+i8y89~^xx>7Mh$y=Nhtx{nJaQs~qf zr^zoEtGU*^H;!{t{UwlOcFC``zHYyL5;~>cOoM;oNq#<{4aYE_Je~QlWW0E_b01a7 zQYF{EP?t|eQXUmew>^iPJXu8+GD1rXtV$Q%N(FYWp{jnm^_~-YGXc|n(jxuEtA~k# z`fGa^ta}f8E;@NR%@rVL`^W<0EuL@=*5#*c@Lo1S|w1z$KH9?Szg8i;zzs~LmSg98~(8O&EfbG~2@g?``R>ca|t0U8bg>7gm>2RUDQ5Iy<{zx)rg zV^o0IA3O4$fP46M63ln|`JfWD?fWfKZ&(Cv^C<^zf52r+0v9u;zwuq4p@iSc{YIUL zKUI_qnhkN?F-51gnB|V|)1?a5kn_BOX?I1(ul9k#VCwfadL(7yP>|s&_|Th_p8c$z zhI0YgRDw}XmE5Xy8hfc~3!_lsF7UBHu)LfrC?m1F^d*@>=YR%E_qUX{_q%`Hfm`h# z8|XjLYVigpAEv!2*iv?$Ni?n#sjHs_Vq81~DEy3i?IU5b>9|&s6e%#hHG#ql`X3zR z4#)~g8odI2z6?^6dV3D* zFU~`K_E~o*ciHDu)YV-MCimiWqW(%^Uv%kmv~=m`qCybcs{u(=m2E|pE>#-gl}&;D z{?ez0K|N9Wrf;sKGse8)ow9%up4iYNvC6iz%BC1S2w{b8oDmXPdb^*&1(_$f>yvh& zZgjiuZ2*vy%UoTK9?Xtb?OH5;ye zI}zl-AgdO2yV;*3gZ@zrj=S6QUS>_20w#%11lY*RX(1W&>|f8NYFv*YV^VwshRanV zK6(Ow$Put;Q@z_77uE(*)FSj^C*F^cgZ=Yu_b8?VLi^s_co{UV6g(A`Q^~<(NeFH~ zkcnp9!omk&`;e|e_Fxo*_bn(+J$*|kl1wU&6Mhbk6vqx1|FXt9 zla&NME`~lH3?COr(`He82_JtE$;Y+zo~ z!u9psPbbL<^kz38D2QJI>SIxB+KM>h<46e9ha1X5o2&PQ2ZZ49C*`tR+7C=ln-LK5 zzysDKYu=GD*ai={z20%_Jmh0140_o+8mXv&EM7uk$;;fi<6<1}c{E22@(l2K+A$xH z0WfPCWWe9Nv(1MSsK@a|iZwc%rX-9Cqu`5WL!ZBA!Uy#n2O&Zu^BQwU!ay8-e#Q}~ z{4e_Zzkl8fKEJLt<}Y3YpJ&skZvX4~fMJfn=Sg3r+n~=sf{|Rtx(D+0%5I^&>oy-x zkg+TCt+YPPKzqt5elvf*J=zBwNqe5m3z{l?&#e-8bV5wFC18(1UTHhQIwY3nZPKI( z*kB$~q9T#eKcfBD56Xvu_HH-^p>&fY%B^WRWj#FtH`4`%_fdZaum&9Gyjn zW(ShcIUu219`sZs5s4&g9fG@{QU|Wr^eQOk z;GS$zX%g*F~LGqL_WE2^_};dw}R^^hJ)P@ z_0l>G_WB;7VucD(o9qX^cNIso3<+s~@%e7wp!KPMq~bxYu)eL1><5sVGP%4vA*SkM z0aEiNM~S}zV*KN<(Xa==d)o8dGz1?(XmXqw^xVRG-euQ-ZGO&t`wyu{{0V|lvNy}-k;#SP5D75rG6L%n6oQZ(@RIG%jhiXJ#@!xhBOT&{g=g=yYnIs# zYjP7;mt>1eqqhM(eZ{ z(2nX6O!4FJBK8BID$d=Ly9ZEUpeh2`68()i;nrQYlm4!1KzqGS_}TY>JPR+wg(Kny zW*|5(Xozp3zrzhxw}2A)Kc+8+zw+bd!-fZt&sNowtKB-%^8fyJ%aF|Yu5tP(=LKM! zAfc$OdB3nUpj!t^BV+h>vKkzrwI+Iz2wgs!rVI39jOay3AB3H|<+X2r-aUG4G0+%4 zUPPj)<8d%84X4d+{d5oTEXT1MXj0vMAUj0V>W@HW;FbR0e}G8~PX6IRRT&&y5n;RJ zBnr%!1Ld@$bmtEkC|6VDDHZz$R?+OT^IS)2YQI6Gc_M;rk=Ns7LfnI2X< zSX@1}b+st0{zVl`l(6~=yUCLn;MlPzn#lSXmIJaEu%M@a*H?A+k^7DP^P7e*ssrg} zCZ2*~2Vzyo!K+%2$hm8SsON(Y9iL2i^_rJS9dj@Oxw*7}VT+pDPGAIQ|2eHP0X**r zZONc(+;DNX8paU3L<~H;qf|;CSmlAcxn7U_OWN!KY4b*6C(MWP83nd^d7^-%4*We+ zYhdEI%OERx#w^sk!N7C8mSVEy-o3?D;NxL|0?P+D5NsJ64ku|tH)?9wnk)_>K_2j<{rzzBMw=^=W|$I z#Eiq|P4v$rtH10#?B7@7da^3e{4WpwPni?;3f3FT8X3SVB?Asoa%W@oQZ}ebOJsF9 z!j#k&rE6JM@NUgtYW@Xd=<^07EgYB;-HI6+@u*tkQ3>yU;!VYIbe~4sj~os3DRYF~ ztK&=5{wHU_AeV!yGjFX-Miid~G+Wyuus93msn!9s8(BDdnrrlgFIv2nSH+ohVev?7 z6#^epOo7FzkpfeD&|);sI9UM}zln;T`9FNw8hyE~dKfLn!Qx;wfuYBU)Azzjnt{_* zuXS5K&9OyWa@z9bK;WzD1*$cI@REuMYDF}MzX>lUP%&#Ws@t)du#7JFPpgzk11(nn zoR1c#!Qz4(zAP3bB}fL*AX)O`uw3(;+z`RDJ}nkU9C&@gbB!U0FGqgE>GfIf)hIGo{}>Wh0TvGFZ05VWYGBM)cE;Y>3sBd;3E8WcEu1l z`wa{8%F&pu%adXGFUj{F;H)qM{nV~S zu)OmpLZu8DEKj^Y|LH=0@>A{T-&k+IZyp{CzHVu03SWN@$^p|>L+vyM%e7fsnBY?Z zX!-MqN(G60q~Ou7{(Lw^e^@TF+ZtsaAcRzg!5LCp!+WJwzuz`AcWiRuneHbo6ygn?bZ3oxP=oN@EI^ zyRx>Soo1rtxek@2#pe0LSBU#m?Xm=p{JjIvj|HG3YEn}lgZu?903C7S+Lu(26_zo0 zNLUQQ#~J=%daB@gU;N4AMqz)Q0M7wKU}I!5p@T8;!5?U5Hx#)AE*Wi*9zsHS4`wl@ ztI_1FZD0JjEVxI1-T-2bLttBJ;lUOE%PcfT#+Bo8CS9?u(ogs5@AYNMu>S}BH}H^O zbAW2zyd#tI2Us`{q+^0mf|m`MhMJK))yJ-=n(Eg6Z{9ilKeQKa4ujqrFaNcVI0g_P^Ua-?L4bq> zf&k%+QSP98HRM_umyAXWM{xb$=5u{ssw&_cDWldH@@P?6)!!Fi;=t z@)?8W_#_ci{T(p#XahPQfmaM&2!BW({PKBtfS2imrZ8sOrx}Ga(!*xv;S8yu+sUY| zFg?q&AiaKdTQ{f;#{q4l6qxd>-u+~Y#@kMC1fCIJtNUNih*1VQc%Qi-#0JG@1&&(^ z4v8I@zoxsT)mZdesGW%g_Xx%n!(|rUZ@G>Ig3LVecspDslM86QZ?X~N+k69m88v@j zgDV-tRrX}bgEK`!d`p|S7XC)mnSvAokVX{b9ptcOh)RS(RN}D{+dl>?HSto$)n74w z{MGo30!gWr5^h5zc?L-Gn~b6O1w@iEdZqB@7=1X)8ciPGg}(UTO8QT-e6Ir%v1k6Y zD?D5T0z)%Yt2JkVJ)Fo`K7go(7hG%$Vzn9Ta!Crx-sXSZqzcgF$^TP;h43ZFq?gZ| zWffHO&CGci{}IH$e;@~nd^!9h3bnXp0z5)Zrq#}FUy$dVsyX;jLKE07trd9SxPR~* ze$mnYi=N|z;0Z=$#GJ~1PQL#0*ogE_0vZ3L?*Fpc)PR@Nv=(NMb6hbd-AaXo#Gej9 zgdUk^-v1GyN7i!~SkL>*CFfgvKK<*?tDO2O0l@aPoc`A*b`OfuD@} zd+E=MA0uwh)0VEghIR(fRNHZI-H#LD+E0Eotw&S?QN83x%AaHWY?YSz__x+TdW-A8 ztN-@9vmr3j{2&3FNeL(LAi;-VIv~1XqF-Ksz#a2)O}Dbpf#LH&b`VNJ0XK*-5eBt3 z%yz_s?L?s_ACBS^sNSvP!S%^9-@uxb7t~uVW*Bb`?343)0iD1ALED*Be$jsgCoq(? zG=vkpmj_W?bibA!N`KDe@cwH!zj!R1V5Pa05>Mk${b3RR?I}BKKuyUs9Cjtnk2Vlt zF)8jF_#^!p^}i_K&cW9nnqKv@{hzbQmq99WEUlc&d zNe=XbhI{`f{fK`hgcn?UaMxDJlkc=4kjY=)2a;Pz{T*?FPxFcY`DutKM_8C}(mk|Q z63gMi&+7j9>OcPq1iN4{M1hf?B@Z_zGTymuC{ZF8o zU-|=2LW@FoDBRnK`o(-qcPG^rEq?LbSEQPE?>KeG z3OK+-g?$GEtHt?HbSa`!f)b>b&arQc=(3oI;QZt-Z8xZH2R5eoquT1f#pewOOGmm9 z1tTy1J2)MY8Y-u?l_0npU@L~Fwh|OylNZYax%VIL;x?cAu!BfdIeSEG)UX4S=7y8n zaEw-hcogFij}w3$^dt{qk3ro!Y0%$ahkpQZSC16s6P5^Y?md)iV8AMM`3O9y{kvfo z{N3g=Ghg=I4!o1q_6w%-f2WAs3F3!OwSx|h8-{=y^yYaw_aQC%tLJ?fRa~F$EFj%O ziH{R*r2%VrMvIjEW#D%vgq|~JX?$B0u5gS*TY%7lOpx-tpxQ^{T6;4%#|AriPq8^q z|0W~2I=(&adV`PZe%Auzy)0N6M8*6F1iSH7pb{y>?5wvr=pvzF=#zWe4C1w1$NSX`+KwIX96{`3BdSWgp;LUg1&e|0Cc}NSuw}9T}2s zm41b({uZoeoS=?=@oO9-T&m}j`THnpx;Mbs5AClWGE>Q?TE5Ww2DbheB6S(!Om+KN ze6Qg!0*-)5@YldT;mEp44Q(NW#3$hZ+822jclO1%Ot^0?%L8HN0T=Mw&+Y^l$D&Qe zk{XQEkNM7b@a~X3%I)dXl}mn%UdGp<|1BGeUx3VP! zT2S!v>}OY0`|mRdj8&WZMJpe!uzZh|E4so1w}S?`2>O`4#Q#N&gvuaPBbE7nV#~*a zWOMvevNba?89Q8!^bNGoTprvqOBjma}^DghL)E09x9}9m97Tuz0qDuLpzrgEp07HThk~F8fd9d zT2z{(A?>00JKvjr-uK1(m$$mleV+52^_=rL=M;1Qi)IK_h(RH1|2!g0|D>@nLBG1B z3YmtTc;*5eTb4{f$^S0hf8R#vR2?AN&)i=G@}{mR5q_Kk!o0Yz@Fb$WX^c`wBO;yJ z2AkjE{eEl}NHbY@B3t^CGvSn011TJC%w3`m8WD1V3HuW^|9#hjbh5(^_)pmeB4;sx zYqJg0kSj;nrt(@nZW>4)aR()Q+`uh%<+oWZW_sA3%_C{PdFQ*6`VsvLVz`Sn?;n9* zNWZ5B!U@#h@b>i~fAKGc=m}nVQ=SP1CaXav4n`~GT)yEFnv~&8;cR8aY46}5My~he z)YQz3ko~1gLFwt~7~isOfJX+O+T}*D7GGkm>~RXp>__~vQu8G3*#g1hX}$FC3DX038~+I z?`sr>?}7X+PH;;?(RaW zIIkXMudA1OP+dC`b0FNuqb3U^ux@3~a6x$GU$n)FA1#)e}-t8amsWz{A82j@RWDRYY(l0I&lM*xNM+bzum&;pQnX< zM$L9UmkDW!sJ~P-h5jRgkDMEBI19c0v1e%L{D+<% zTgeyNyD^XMtt^GF_-_}Jwq}83CMEVL>7-6+h9;ZD*gx2h)LcOZR^fy>n^DL4A+i4^ z5XqPyK1Hwp&|d#xu&(^=+%-(|p3n;qflS_%d=kfx>v=z?+rF=MP<36W#ZV4+FMM7| zvyBVAg`Bu%AC^)M1J$wrh2sG+CQ!n7l>XsVMKmY_N>M-?(CrNwem6A?5SFB*qz5iE zNXG?2b9j`+oNwFbDTaNjB7E7a&3027$G=|)Gk&sR!tZS8iXFdaahFomfKn*&z#;i( zG<=u6)W`E0*MUkx$;eQ}B`-#_*(ZbyqUbhP+*?bLID-rrTj{b54dpgL{~_fyhq13U zLpd%bP>Ci|3nfzV(UpE#IHHV;qgTM?PiW}&z9!1|4vuUOiQ;{*3-Q=$uA<}+mf6Gi zyzqCH(K!^qS>9zX7kw&y{A!!xi?k1c?IAh>w@7iUA_vwtP|cC|j9DkLNv*eZLO3Vh zTGi|(qKCW}+kYSDg04o*s|Y6^mfuHL2OVLH#{E;oE@V?L8lI^{yj={eNtb8xBUx9f zu5nyGR`75;&Lt^F`6AooLy&X-5x+hH6^T0Q*zyyVNa`79)GUC#5K<0mFxC~h$WX$Xz;MyO$?4cpQF;v-;(wNQndRqy&D z?ty%8Ofj_?Iijk32TnEpYfFHL*k&=4rc4|`2w9-)bx$X;DDF(10tYE@XQINdK;)!A zjgSf1^}p{^!63B&CjqVr`;iByJb2K^%S#`VI?jzz5j4PbR^mv(93YOcug>rMfPR3Q z#2sHlH_#J^3!(H~=?GLyiil4Fp{^8_EDi>&mW-xH$TA#vgXlLmp71Q;Hp}pF;!ds! z3QpC*mz@*jo;91AT~}l`&UT05cM-v9EVQiU{)2CUZR+W7^GLBsIgWQ~f;!K3mNk-1)K&@m!Y3q++|CKzwhsU$#Nb|J;^|_ohX^chgo=xc%dRhv*xA|f zew&&q`H+|OA9+C7D1(1{=}hA@=?FUDHBg&fx4SQn{6l~L0bz6K_C=YLgp>E0_oyeb zm=O$EjnRdFvmlO`viOH^>Qa{2fet9qs0HsRNDjC^L$#CI%VMThZMTbU&yutdBjn=z z1Gz}H*-fnIfD15w%?#}3Rd)^Ho_E3w^l?heg-cWQ(y1gUMLVO5uJwVKqXN=bdVuc zY0N1V8UJsN{vRhQsWC6J-p69DwHPJj2}EaO1#5ejdtU}G*6pSIh%gbYn>ND~Vvul` z@Y_BJPh<6m(jJ9IM2g{1K63DK*l)aR;|fnJrp#Ir4ZV|k~cyH}0VQ;mBMLAf~pXR1-kFr(E zxymr~hh#b)_B+1ijSvKyJvur{kWB{%(cO!iNd_-u6h=oO5@j3^Qm-b9kr$wi-gaz% zZRr0yBU})0(tz)b-`Mcwq+K+jWx#ng#YoUeJsEvxC`~vnLfyH!KH%3!c22ePlS-1a z({^F10*Q=``4MexZR*n_owyeiBr44yQ5ll#$pq^pN2o_CH)ia7>i5l55{mAp4%^FW z-l44Tilep*5QZ=80{-eg>aaFQJL*ew;C}x5XW=Jf=gdAjwRxC9Sx~IP8RSod5JZ8%& zL2IClr|gJm*|jaNBeAeH;12marG2m67RkB-<8ir|znuV51}q519z^;!=aZ5Ws->kR zgQPy!M(a<}kp}effyj5JasgS=n)DmYH4adRA-U-W15xF_?x#i()jpI4uy!XWr&fu? z2=&PfI^+R}#X*sd@M0nGaZ!yWJwGn)#oy)9fGYmLhAWjo@b#yHownnmxRBscgk3`H z@h=lohp9`oSUh>YFXk^|4c=TL+K1ghveMeejaJ~XV5h-HQK+2Rh2liIj4qMn0#FE7l*guWgSW*88TsScUJ3;2kc2 zyAaIpKuf{PVkBA^(486kh~o_ozxqc85K=~X?Q>O!d875ki>%2fT57g&1ulPBQW9IcvW>)@gLr*YwoYeg z@n;c#LWN7I1`VL`|c@bTkf*hm9A#kkD3r#dN-btpt~T`H5r zc0~2RKGB%jF`1Z;T8bEU6H$m8jrzlJRZB9KLOD(TjfWL^+KaQk5 z^=}*6h0;VhU;4T*-rlg~4%hdYyd!~11m2~D&Ih)=6Q?Lp)r-Pq=r^EZHz!!qUo>0c z{Y@7D))D{m0qS$#mCscRoNLTUWVcU)ztscOxwVA^lRco;#*PLo>--5GnTvAdp#j+65s(}Y+?~<@2^LVz!BOFGI&@4Z-{Op-BYro8Z(98ODNM_>C zl<#0xyHI9rRC-?yBbLwP0%`@#4HuWGC6ew5Ku&?Uh_(x?SN^gYShC^}*9jxhr@aFM zOn2XkR)M#SDe?=cuE0rF4WJqt4aT$q|8Sx-LfwV;%HP#f1zCK5Z_`GJ^?WFJ-QnTk z70jJ5f(KsvNEFIu=olA2$HHlxCV|8Oh}ILL3H~owTQvM-{!2s zB3KSoySL#_d-(cHNZHqKAD^7Ra%I^=!1;Sy4bB~N18=C)cfNL2iw$lO<4m^us2l7H(h0`SeI60sfF?uYkl`~SFPESFGaX~~l(6c<{vx|ACc z)Z{UsrX_s6KJ8t=SG&EGu+?frV)6fWl>fmdP$$uwXbi(HB(@~u5y`cLH?cl7z(ert z?k-&0b0f6kSl8d?{W7Zs)c~7*7#NUA`PS8?^P#sl@8aBk%M%RH#K}B+k2wx=9f_FM zIJYuIBI-peLGwQ*WPXY!R)io67X}&{987I_T`0^8c|Xd;}{E8NIMluE@hT}I56l8*C_p8cLY!BOo! zB)x5`IpF!4WE%+=#tm$kX>|8=1BX$J^?8m!cJmm&uuxpfvOf`Q?2MU&5v6MH2?>db z?D?18x4bo+r-i@_TGN}}d0JAUz0!h1uPy?AcIwWTSn?SaA2KasIwUd?YiRYa>;H$a zps&Gof1KX( zuPo+VOdlXVvClzD=L$_PrBe$`gjFnt2Za7aqy7Ig|NKWp85q7uO}7euaO_H}43HOc zD@cW$F7x%x8zCZG@2|1I7EI-i?r4SIlyM?at#4#|wkp%W^GA;kLK5p^fp<5&D_*=f zbyLg<2{57iVMWjGqWLxwYB!ZW!@c0~t55#*5lEu|ZNZpVZaLn-g;@|ixQ&P)vC%Bc ziI7#1YMq*%4*K-z(}@2w&@pdQqzwDbe*=F1f5SOS$o+9LslNNz*O%YKa^v+|Vy_dg zxo$MkAnHR0p1KB~3ddB5sbdXIIB8=I4UZxqqCbgn332Vl?Ww6L@cBH>LjSaHw$M+l z%XAtp+1e1|ISy^89tr`T)SDz(;8y#Ez4;qcbsS+figC=m5&FI$ED~t+6EQ(t4OqV! zlSX`9C%orHajez6_px;<-7@emB&~oYz z!avak`C_3mDA^ih*R#cw#V9+MBO2^ANCdlX)aG6Ystxi`M2a)#5WgZolk7zwR~zWDXO$yO(rsaSPefO=0@0}!JF-lkf*mA9>U zjkIir=+7E9=In|x{HHAK+z!Hc*nWbnMPc=C35>yTpTgb`nB6Zt;zYq87{L@nT?97PL?qUjGQ->%5OCfd^ zabJdQbZ}lflvmYQ>usrcgHnwtZK}T!n#=c0@h`3mwjcihDXCoF2%V`4=wRuD@L@kK z$x5EW8?#_(&U)?(z{3WEmL}$}a#Fv4mSap*yOC42Me84? zYeYZmW@Bh*Xzca?*G@d16*!>vJwVNJ;uFrwH*!VG(&D`7|MT{*e<(l|0&Ip&G+%$` zHJH%KwI3Ot@CD?RO5-9Frw}I$i3Ws<@FlYRjpq2^(2#S#+GWvB%zZpTdsVsy)L zt{$Oq49Vp6>3u1-#5!*rK%Mz9o?`?TubQ{kkZ{hZG9 z6cp?ci2flJgs2bbF$%RM+FIYjvv|xq4(40JgX|$1iN&)YF@om^56#mEbpm;D`M;?3 zhu3Oc%ecf3s@?;WXLMhT>4CnPj@lbVj93vN1m#mnfor5=48H@e9g3Gr5Mb2LisC=0 zS9BN3dD_g|`{ZFpM#gu*Xp>O~q&?u_s%!=iSER8;jQ$Jo6?paPUQvnUmaSy{O{HyP9&ru02;o@8s=qLXz|H0Kv9l9;oq*v5z34zCI=e6Ed$sf`yy26hlRD} zEe{vr1k@T*)j3klDuw{ZfHH0dz>4m`wzlc{uVumOSzx@=oyKnR@NCdAxF%QEm~s1k zNJnER=A+5fzcY_vH;OqGxo`ILE*4gj0Qki}7icuZWoh~~g%Ff3n7-|5WvVOO=336P z=kDTRO(3%6b>5_s0tYG*;Rnc9@x*!2e=hQmeL$V-@MmjjSwMKIg8F4u+ef#+E*)#8TdgrM+u`yiE zZV@)sH9z^S9Xx|JTMskX6qY2_`(H=<|C*3Jg|f1svP4G&pcUT~279*#&E658TMb|< zsPS&!DS7=5b#|6;w_y-zT@8P(3h%4t`!FCV5*0w`KYDWd}lUnyS-c|qO8#9gDHzulD>cHF6L zbky?r@#EG7zWWbaL+xwMetC$$s67MM3&jii410v;Wh!j!`rKslFX~|T;j%y$#~tPV z?{sh}n@JS27-{MFDkp|j=0xS%2mk;Dd=1f<^~DWh>itG|pEy|uUkD^<(EQEy(&HjC zy>&=ckj@k_q6;99rC9Gva-okqMr$B6)#`0CXwvgPvnJJBIL zCi@apq`u47wU-VXFq+C@5)QRLBqD>bOihmN_elu~0XjV6JNh8|7TN;W|D1ip!kB{) z9&T5H_T&S!4)*o|N3Gr-f1Pu!CkgO|d9bkp$I|zZHPg8;f1el}5gxt@?mrs}sA+sD zk5p9ZAgI5CZySD=xHY^7Q0L8pyd$8Sg1OQaf7z!UZbp8B9>s-2dzv8}I{HM!2gOob znNt=-kC?#TZ4>B$P}fWe4wuai@c@cqM{ul=8YB)}u$qV{8DeEh){+lM**As6HV2-; zh`SLci8_CG#G0SOuDtp-OIp&fW%Ry3|J?kr#Q6S@H<+nirRreyx5b zQjiTryRAr{<||Jd_~8&&^Fz{r1i=DDuW^PPiV5rAF+(GrPAvvFBnOu6>-4ii=q)>^e857A}z!G4lw@S2Fngm37++zYT#FF4 zE`IW4B&feRwc`J^{bpHEc6~$r87wx%TeD6?77yWxju$=9KLSrX8XxQ((x77daQxWu zk^Y9M`g5yg`qP)POV|q_W7tC(zc_IEk*|N=^K%vsJD%4BB{VXSc-feRof zT;KrEG<`|92vxi^!O6|ixb92~gR`$V&}RIw*OHaOHK;8hxtYlP5K5MC9EjGSG16-14;$2U~m|>YGKF-cO&TbY?1wo2UcE zchryyM6@1D5{iOxv32~S0=We+V+QGkFXin9e{f+B0H;WxJK1m-$#oh67RBY|CnTDB zK>DEs6OaZBs@g<->d}8P6 z#4Lwh@xghhktEa>dZ*=pYBt!FPt>L)w7cVGgIUsP@ZXEs+bb4yswQly&xh2#sbD9WkDh~ku+hnrCt3EeL;jZ! zir$K!C#b@%N?~2{MTo$onhWsBfs>U|2?Muz3P347fme(cYwB8|8|ihTvwqZoE_*04 zw&RdMRKvFr@pS*+ng(r#;E7+=4c*UUiRmR|LkdguPVM_Sb`LNeUg|z#20rqMNHDd8 z=y+Yu)8Hgz$_Qc^24%4jrFd9f?Y~aHNgl2(n<|AJkBE$Z(etDe&~r8rFcHfMa#G*T zR*F<*HX>D-bhyussnv`lDic`p8$QRenB)sF#y|G$Z1Z6pIQ)(Cg9gdcXF(}NMBadb z1N%YYk>v6jX}x0F?UD}@A*MX|m#H68zsHOXXi`^%8 zp+12jemToWF3y6JVg8Rf`|gA&-+(+d9AyK;ZpYQh)@Z)oXAz2Wc&jk|7Vy$87$W;55y zAg4$fI{XQ6hJLoB+J&+psb$S+m@dCPltv?hd%8lcesuP8qVon+J6Cua2?eMu@Mk}T z9i#h#g4bGSyFKb;l4b?HS3I@*-Nd|5C19s+2(4<|^Vyrl^ssZa;l*=1C<@LknH$An zdNYYVHdMF!B=wNR_jv>toe3#Y{H5ML^FxUA%0_6F4;CT=D|d#UKdxBsNn*ji$Im6D zA)=0bn(-zG+mH*g=Ln^2bBAHiz-fz&HY~k1|91H}CVAe0)Tals_^w0bBe~NsD z#t||8``<|1`X zJDe}gs%BiHP}VpHrv4WEl!ASETRcSx?31&Rb~b2chFt^E;~g&*cG$F;`0PC+N*N|q z86>RmlzSPJc4|$&W@ECssM#r3h;OoDb5Q{53+oCP4g#@HhHe~z`K$l*Zlr}w?M zDQXJJiAzRusBrhwnte5F9vA$CwpCS}F8PVVKh16lSjf&QR7fc#jlf5B32A$^E!#Cbuufjik9`n9g~mFC_u449j8hvCo*QrOS|{8GtBdAW<7!f_ z)*RVI|1~6F9p=@P0MF(B>WUa~y=K9ryT9KAAcalm$LbZX)+C~KAX){rGHbIZk&HC< zK0z=E2+ID-j0M$Sq6qvJf!wL&q-6)VeVl{dk7ewnb{roa(3J-AZ>$ilW{zmJKmZNE ze5WQ-Y8(*3Vk%)T_8aov5WK%F8uvjoNci(@4g!ei>h$X_WDLwU%UK(nc@^0~Nla}j zHUtm53U1g$UO6p=f^lm6RrMc3X&rZ>W)J3Qg6(`;v5gt|-i9zkIk%QOr$>Sa@e=ZF-t%YZ9^<4@*n;B7FnWHCq6+belcd-clVRPBG{u zu=cw_{*49q11t$F>)z+u9X_}cD8gc15C&spo{d@xY;4Zf7PHNA=#*a}{1_{Sutc;e zYAJQQBxWA}@jl=WeW-!X;2Yc}_+^3y0hvHO87#K#%}f%4Wc@6iMs^0Off}b8N{DKu zav)1cRxMZG{r0WO&FuS#c?79&-<<#MEtJ!DfK9J2e;2$}n7KheBR(^cw-bXTUNE#7`1P4&g-~pb2 z4jJ(LcmUfyn@L{(Q{#7Vfn~nf`6dCq1`(-$q`OrMAh^qc|Ck6lekt^tf@woB7du07 z$N(s|bxy$*+hPkre%<-=eR0h5FaxPfd%!L;vIGzA5Zr@X0qKE(H2ec38aW{izu)CU z0zB>;2{c$9;PzauCyGdN1}zJGL%|b>Jm)gtdf1vTsQtX2q3xhYBLqE$o`_2!sGFPK zYR-aEkOE)%M2^dWce-`=fFY5^;-fH}c_h@@WjyB+Aafy$Avn|-gRnL9vZDXG-OEbh zbj`B)ksCrtV;%TSDjhrpY?U0+@b?^q7dt?yNub>3dgz&K@W^;0T)|4^^xd$s#J5G< zlt&W?2!9t!4p@^(*7zP!hHcPWB-@r2SK$TQe+E&K?^L3GC{HjFU(`9%*&?2yA0rpj zcPDNZOcrzzY6K}oR{2Eo;rMjF9x1HC535YC(*7Ju5OEuq7?+=Wa4(X)5qec|l?Kxa zV%PDb5Ptkp8W9?gR8z_#)s)-tCY^&1ovQ%m30`nC5b$TSJNO#t2srz-CR_EM^mjji zgu&!5?fo0%Ip8@ojtven*G@zCAQXNVm4W~mrea4$-xS1N0&*%RN*H%Fg>W@41@RT+!8)=t2rP(`d|_vK9znhthi)A51}T1G@@Aq1;sw^TuzB7kBLM~HkCeGFE4 z=G+N&PhWrq`M}xy3C+#m6k-qvW(~~_0u(~lF+G4~r0QYyPo3d}&OMG4tgT6C5rlgE zwuP(!Foy<{N;x!vxDVICB?xB4aDdG`1N!0QwNAcmEv8n9UiX~-2^4#Zq;>UupnI*q z$6J?_qjc--DNDspKfc`gdLj~ghG8lNrW1r?W*4CE_Xk(WeEGK_(lRp(QPY6L{XtI4 z3=f_tLZJE^(_>pRq+4X)p-nX7z)Y$CTIo<)h(MF$eOjSiD5Rfu{sA0IwH6|#T&e{6 zPcwQtq*kIx_OMW}~sAGJh{6YpabA{UVGblyiR zMBP%Ep2K6nn-t}6<7B?VCJkO@2u^2_n7&Ff<- z8WAJPww?Cei;~qeeI8v;Qp9fBER@YP@xFo<;PW%PWoC9?4?X`$g_^)L@u}(MnjFQu z)C2*CP6)309Z53VmZFs=3dK^4J4`bE}w!CdvV?V7>Mf=_Fl!cW` zJ#E*=qwlSMuDk9A-lH8jMFn+VU*TkPP_L%Gybe)~(XdYgmTiG8$7z|-(q~UJ06kDx zWg6lRGlv87imFy6H+uU<#nYuy=s%5NDs?RmdsIO4=EIwdrduV;o4qg_eJ0a}seM;E z(!6;H>3ZFMn`vPALxZ%qdav@W2|82j8^fduR&W#fFui`L3A~c;M6`!7JyX`6SQS2$i0e++VT5VT4 zb@fW!kTCS=woZkz>q!XvI3RxO{Ty7aL5_Y+z2|l_u|$wa`?SxD+X@j0xZHmw?;J2o zWsXe#K6BmoeRPUw){zj!9V`GbjnsU?yTcpvJAJ|JXHOecjWkz%YkoPf{$>E4UXOQU z96HLwuBy|ISkK?Ck=baMxgoPTB6DD1mSNzaqjQL1@kdu)2f!*KBX>KA7Z4sca4gtQ zCA9j8Y3Q01E3%R9c=D3lgl`H90Y zc?~VzRYUtRv(Wz8zhhe2WGf5Uu6-8?mPSD-j(Xn@Sm&f|rmC8xuS4jVIr|*5`GkMjkc|!z( z9;a3s{MQ@E*8HIv`H3fC$RDM%Z6br&XS)42x`&VjbEc0ET-xZmw4cPpYzUG!#zofsd}OEF(R_kM%(_{bXC8pW%@*`Bj7cBPa$ZT^nW>pKC+*oC1eWdWPd zj~#vK-G;N~!3yCTyF-*D5d^BU&nFn^G2)}zrD=EcXRfB=paP2#rPb0)>MEz=t%F4? zE?Z^PGy3-y^@m_`?Aok}`g-$cnK@RO$CS0mH_RY2s*Ir7`cTp4MA1unX#DTgaI#<6 z2%3El-6!=O8_UQ@`t;56YYFZ05x<$e{xh|Hh=B;4UM-xa`!rw6DJ{IdbZKj8Y-{WF zz4h1gFAqGSyR~NET*L)Inb%t_@77!TsxIVA==y!peTg*w55o07-$+XuSRYOIeI&R! z49)XBYFQ6eW!ZlkZ1!;StvoV?j1VfD$j!s7xjZu2nwQz4U|dMu8@5z0P`U7aW(G!) z)%N&IFZ*38Tc2w_v%EffY148A*+!e6k&(r^T`4Kzj9p(+W$N!Ky7I0pm3S*pSAd4yd*E3^Je=CI!^ zwg2qO63tA)0h&584WHw;mi@k6WJ>DY()C}|4Km$YG-bN+aMZYZmcClpOE36Ta&+Me z-r*_7%h5?jPKLcuF>5ru+s-w7dd}FjPIss-Xyvt9`+`+Z&UBjd%Em`}u8YTGfbd=0=C=Mh90h z1=8W)WC(P%^2ku#J;er1-C^nW=e#LVlwUWb2Ii$2N=5hkxH7<;dZvxG8RhB)<#dV| zm5SjM^?e8p1o;<%9^c(yD=a_Ta3ec4d)9yM^*zbtpv6)etdo%&u=WqZ)~O0 z%lX|DjFws$6P#(d+O^UoY(`0H?Xf<>B$FuXW2`S}Zq{bkJX|$m{QdGB^&B3Pks9sx z#0D~*eKmDb4J%u!Cg8lLM;P}dg?N|9LB12$xpt0Xj2-drR$p2}4}DU_V-;;-n7g&6 z7i-v4`sN3=HU?Ny@E0P2B-wm6zxr=-3Dc`mufjw)Ps!Bos-=?j0VY!!U`nRpTPKBI ztijv_hlL{KMrKzQ?jiF#SftI*nnXXCU75Z}E|FC;?G$t}0GSU}8y#~&$;YQ|i{!$g6X9x{f0D z!n3xEbBUcQ`%c%j`y%oj#9#MvndrlDQ;3Qnp%Qp+{p#@8 z8je<4GoJlgj|>cu?HjnM>%;nLC5&M99JRB*wGRU$V;~HKwr7#;Q@!;fh$i@!u4E?H z_aC*TTF+rj7Q<9gtzScT(MC_vjcjl$()~j@$6x?m5z(vzg>b~dqIYZ0-<_N;PAa25 zxU4hWSx1D4W=Z&Prq3=|EJVe5nzu@Q`1a=5?eGSs<;TM|G9#4_6jKkQFwO%!-y_51 z(86aazS7$~a>3&%^S8HFZU-UCmSN-KzK}YpI^=F;F1@L( zsITb}ar-`jpPn_zADA)R|KY>fk9YolD{rUQ-%@Px)mFC8+U-dSjItD4K_;8IPILj+ zKA&IuJX|@;RzM>fe7;H^n|;IIsr&hkxF{si6S2M{8%?1-qy~caHGP*Lo^;vb(eBrjPk+Zx9%n6Z<7Tn`K zocC^+D*7qnB`{o&@L7e?9V-zqM&Wkl{ME$7o}QVuyc|9CZ)s+`d2?=FI29LMdBiXBXD7a)Xr+%AJia1ezd_dH1GfLFzG)8f5 zj|&zLpBWFE0fCVeS5jdaLUd{y{&7|Dp539Oy9m*ED6*?)I-)VZ)g1Noz6X;-c6DrY z7ER0<#{8JEYSC+lb6b7m@^qyufp*%Jt{jh+CJlFDm3!g~c zjRm231G|DR?!3a$7p^~!)iHP-Y!s~Dfnu$z7EQSrOu$|BMexfK3~TFMiP4OdOUqCe zD9?CJ+0HB!wW9o`Ia$hS`O_6c;n;3UdQG3V4?G@?XC+lMc=X=&98D0f=cE@hyUH(| zt>Nuu>@3eNrgr%4>#LL<@0^K3lgsxe1eLBQi4by2>%4vbrR`;mHg9LP$Pw;Cc`a(> zWhc_mCA#85cMWT8VzqT_I>u`)&(n*T-#qEZdi06dKGbeGJQS9&mPbg|h_Z371-&+_ zkVa7YncMC+j&1I89lEkI6!c6muQqYQ-;4idje_abCZW&fFS$9M&MrL&IcV(B6l`3R zBU6-mhWeV)xPGqXG!(aelYip!=;Vc3!ZmiImhaVDex)ge0vwserr-@YXfCFIlWrY3OzlKrWHysK?&k1doD;JXV<<%dT=6>jpk#vSz>~|{w~B|K%Q{F+O>mG~>zKyF zbk{8tecR-{H`ya*b?>Q-Qr=26Z;D7OYlOL4MwjuXuYVxlcP>f%P0O(XpL0FDS&zTO z8CD#jioGyTOmU4+y2A#!$zyE9QQR(3_@^_w@;stM8jd?`SmZBwy(hrHkit7_tsWtV z`0ex0lHR30O7v=hton)R_h-}6%ctD3D}BadR_!`-V=XvYR31oQUp+yDN{6XH<%6fz zRi%H_=3Gz`p1*UZ{rbwu?`0Psl3HHXp)<_7Be$?|lX%QiyIB-O$0+RJ% zG0#M9VWnXnXqpY%-Eg~+bQ?D(5w1H~v_|EF2U*XiG69~r1*7HKNof^H1Vc)JMkSzieGBq{lGSFuLR#JFWAKM;-AEXdF!zX}uk8Wi|O z<4f0#ANk?#i}j7^Bh4V)@^6y;bJoF{elRxf+_2TqyCdgrt6{h-&=g-`>{)3l7T<@Q z&TYD1D(;c;n|mvolqs2{gy#3qd??mh`8J1STYOBc0Gq4uxcPM3SvQt1stGiw%`!rs zIFHnStW&BzBCO0lZz8lb)!W_0u$Z{aVU-!Gz;i>~Z{5SY?D6!Nav7cG53$T2`<@(l zu*vQeWy5%}A>1t}+|W$xR!GyZvG-JoyR_2zT-T5$HOj8xnM)r|v{oJ8B{27acAL78 zlpsugB8QOKDaCw8jzE662o;Wx6eLIBJ4bevz-YO|-b>Y=1k}G&EugOmpZxalRo{DS zud!7aFU`}V@{)u|$ef&SeCX0G-rPsqgg!sYBOJx3aOLcmFUgB1akOnZ!ejL{{xP*@ za|hMMAyw%an9%oO?>xgiaVVwcd$|o|zDK7O^G|IgQ(_}8WUIspS_rPf)sQfL7aYB! ztyjMAUip=ML90V7A=+88ZjtTqLJx(iZczg%~qrI%B-l(uefVO$!!Z-VkTlpJ9-2<+5f1S|{d#$Yn?sMkRh4?@LP4Xfot% z51+3(Mw4Oi$oqS_)oZsgWw+tlTT8y3ZEi8$=Zz;mJ%It9nddmC~VXwlGyX(9=W*LnF_;h@#x9Im(eR%oCjH!7qSSFEJ_! z>?$zqeO0*`MWaBQq@=xjHoWx1g%Dbcw~^-(MC$Czh62C$nqXmcu!Fe?Jl!5^?w|OY zK^O{?kx%j_y7&~XWHG*9POO$pyr*7LMdFn#LJL#GZ7}TL%)`iNY>eveX^Ff`O2{|r z%1)h3XTWEU*pG@N4t>uipcJx}dywYbU5U;t44+kh(ydP7(C6d$c|+xOhYV>l=@t^N zn5fEZ6c>ybHf6kxQ!>mFEH|n2(muN1kn!w<<42g~Y<}Rx=ZN5qn7nuNr8PO^)aN#V zu19z8J;0~eprOofp&6U&K53ccU>SpY;yFLc-c1+e&wDNN6c_4ds4NuW;^_MMCLVVF z@QI{MSv-{X!9huVD)9r?wp9$8`3(O0%j67mfhdya33@T-F$tDR+QJcL7do|dORnwu z2z+jAGolQhtgdFAD6y<7tSJ*`|*t{9`-ER z&9jK+PuWld%}?CzKGhR@SB(;t4x?7P6n+>NP(4#ve5-6{3c5hqy}CuDnT9*uqGCl` zsns^$qB(9fh{=zz^R~uquZDK{ii^%9p*+scqxTR*H*z}Du^u3v?zW0_E zm4C08Y^rGIJ**F-MZ=UX{&@1qBTqbM%cea`yHfrdvFb1x&8@oe?0b%uE%66!H*F|z z1KUPn8utDYjQ_>0>udeKmuq(IjR&u>#C{7D1sgUpH+NyI?Ny;&r2L^r!;%|rElDL4 z5a4iD&h5mnXn7ILuS9sv{)5mwtU^_+wNF)%?m3n6aN?Ni?Y27Bd&_4+GniQlz{C zY5N`WNG9N_dNxm5%and{PdYJEvwKRo36-%B9bU&Tlv`Uza{g`d37xQl$DX9pE2)u} ztK%8mI@CuK?FzX)cn@spd|^J2tAH1&E`M4SXY#Pg;K!sK4SHfkh-Mbu^)~u`r1MWc z#Y3H7iEwppb9pCarNxkT=-Jlh+8dbvZO~dbGxX|BWn0ygnFj6~L%=OBkx|a)*6agQ@?9#g?%uC|mJo(^cLj9#hUH{tvwBW*eB8 zDr79Ukx}EQ^y>#C8v8Cc^INR#oA}B`6#Ds6ES_r=r-!@_`C73}d%e%SD+J_X5~(As4jK|t_n?b`>nJjiv-K06fg$$!0jCFSH9Df`ME+r08&G9vQ@bh(UXv`Y=V z3bzt|e7!d?R>bE34&#;NIZJXS#fKVsUrw2^VS=Z)>p3U;E1#+BuqHev;1=L$w>p(@ zBTZb}qI{{x?#=#VeP&NMJ2o~oG3A(P%`Se=}b0~(C5@H8fQouB#JGW&j*MqwHl&Yr z)gx)Yt#sSiW9Y^?0rbK(Qz23BC42At22EOWC5EMr-+SRPu|8s#(j>X~_T==3(sh|Q za38b}45!xMlhG1jMQLPq2}v&__iE1o{ud9T!FURFR6OMj8jG#O2oO1 zGcZnDZ?&oQ37GJ`a{;;Z3a444Jx+bpf8MZqQ5~S(q9Fnm)~Qt+VPGkcN6skB9Tzy6 za*Sh~nx+yw)~Bw?XHgG;v--H6sB*Wj!Bo%3YwkVxn9buc0lfT%PxG5h-=6Y3NCT-3 zJIh!5j}Dw|7sT5i{c3 zproJ;JAC53>r{_e-0T=7icCQOS}J00hD?lf)Dwk1J*p}>TV|XQQ+(#I^XX`Ud(QQ= zqwDkUN{c}9S4V+Wa%h!kux*e_P}S8c?$j})H7<@j`lv%b?(7A+4bJ(cbl(Ee#8`V6 z-tF8wo;Cu^dWE#Du928Ge{gfHfyrN#-=LW5W9AA^=lzR|X6Z7lSk>fb4gQ>zvgo-*{qMmC*Woak(>Dt#+4NkbIgl%RBR5g=I(QR3Lh z(>x@C&mW%jw-{0vPch&VM6n8MSVRFF%D||}H8Wrjc`|oZ_PZDGuDW=`wxlOoRobTT z`JB>x&2Cak*Ua~JO_;350jow?b!4=eb($>{s9AET?F}f)xu!WvcX9Mqy>l1^joZyPwdoVap$QONEG z?=29g{39sOsV&t+(bx9CTHl>3u5vBkC=7AG<&@l!P=TIdERrbDkb0LW*1W{cgyhcH zzm`A{BR+`HtI@am5H@n60i2>|Pc7|wpE{gt2xs7RJiUSU177nECyByiFj|Me`A`Ym z6G?_PrzFBiY{wiZk(k4UuflE+a}a>L>LNIL{pYilIRb^MtCd%L({#VA>B&23y9Zz? z3r9iGkjk&L%8`n(VnQD_aP__ebxRyR;NI&HHq0OrD@`7xk+lq-5!u?)S~^I~<@So^ z)Qnvt5qK!el7P=x+V5redKQMD6Qq-Cg#RGG7~JXN)HJH)^_#nr`ypIUdTV|DyCGOv z8;DfBhsUYo_sZq|$7`o1;}jg$rb-1H`#&F!;e<<-;Y@bAW?XTc(WOP$@%0I!cggP* zb4cT}XJJgWC2XgZ;}$KYoc7bS-K3I2&K9lL3aKL1)ijE)quq%D5JjImZD}1-qp&Ei z?V`J*=qZiMN{urGjfos%+|^JtL{6?R%3pNCX7-fElSya&BanX3@J{79L*aW7q;@Up>exg-*Nv|Tf zt5TIY&;jRDye1w*i3N(q)oK(NlswinaQ|pwd_VGNL(YcbfU0WM89DP9#w~w1^-|xM zq;Gntm6zAc*N@{;Vujk-$Xl-zRq@8I>WUBe%qM`bCU=x5y6UKmW=zTxmFy*DJUkgp zl-v+}?8)6pqCVk^^&jZ3R?y~5K0qGx(lqxwR*~>dn`Ypu@qNlsb;&pJ`tf{xv%mx# z{qkqYAn1gPN=cBliuosrJM=}m$_P9dP6KD5Mq2Az{4&c#YEFvhC~bSDGIeSBFh!F6 z<6CRf;J1DF25#`uK+|wqP!rYUOz(jImIHx=go!)M`c7bx3oLuXcJ2L|$cE!wT)^B? z=4Ev$@1r0 zAsKwfhl_WqR<2{SV2hJ$fc1!oQq|VimV}E(Q^)r{ya)lld%hLBuA{Z^rPpWV);@Ou z9@ybd)BEgc=B;H7t&uiTfDD7m=-9tEQ{c9;_36Q6Uz)%gT~l8R0T^SSp45BPvW3uf z>_P~?fwt$^kw?GY#r(x)SvQP>!b7(5==K2i~K*{*VQ{uu=X(H7zgY9h|->b zfW7_4qmEihIX!5iP${`N|4wEg86JTI${Ky2V>DGBrBU9JPTr^Nhi!vrkxku&Ts4;V z>%$Un?PN9+@ zRH&TBoKkE%fcnTvriYje;kTDS=TPTXn7N&$^`peU^WMNl5E(bWEOjgR`2${dfgacP za}MRA58BX4OcmF%q@rH6NCGQ7jf*G!K$6a z?q|8B+2iI+J@S|=CTE|<4q}KcZ@Fm6o~3@KF>{h*_U(s97egFFm;1eom`_GiHrFVrYDc44CysCQD8F}*<`z`NF`>eAzIghhhc6`;9SvhRz zGg~)xd+9Pe%k})8q-!5+1hJzaX8)t=x&yK7qjsJstKyMO*?X^S%E;DZuk0O4Hd#$O zdv6hs;4$&O`q7)|8HN%lbz4w9;>m z%hCN`;Eom?M2+o~?{rn&_FrPT+f+jXhUE&<>I7d7uEhrX)4`6o-huk9%t8#eOkwY( zPeef8V!d8-eFFw-M+}{D)3kACa`IRO(ASsyj-N2(5?^dQ-)qb+%KQ2QU z!PTSJ*@@L6D-XNVw?2&=38p~L{br@B@4~?&_Ry8*QSYE#)Pa%2u>$^1;r>kB3cKdi3ijKb^jK=Vc$$ z3a#q7?xoD6fJf|{J1Z}mT`qnbl)`>G_3*wDn#rdxuvyzMo^)a84J%8bK9Mh6!9Y5Uo$e-H_`lNu?V}gmOG<=&4}Dn z!!}E+mFW)UC6{z{jA8E^T!-{89zU`PmXUf=lBXkRFgM_CVcZ+9EdLs-%Juw4+tkt( zQ}WL_JOs&H>5u;!0vONN5LJkmy`<96T51L@jA&dA=4azyn}LU{pS0V|luvgw?$cHg zH+`cRxgceyYQ02?c~>VL#jBBR6pUfd+>?HHhHUytk+tJfu6t2S9rYVujq|O1E}}5# zTjdl!`w>Qv%_LPG+tlz*KN*iSSdVa8XI84kT7uBS5 zo=b~`e(Z;?7@rL`hn>X(p_!&9vg2n0BVV8`IB@bJB=x|-*^DxDV&eB2P|VvChAcK61b-r6lb-E1!m?oMO$?u06wq-U(Rrf?m;^3gO0`G znIqfte6$Dpc0!M>DX@B0`aj*6gAr1XZi8)}bG^}M`h~uOAQwWzeNJ)3#vWG7EPZe85hL!K~0$dztExoeQ zbY`BI+AiZ$0?sz&oAPloJ;tNbVHGmXE%vYG5;@aQbrU9=*uy}cIVXzW#5b#w`d0g8 zRadgQs{$z#dEpfsoMA% z-@COY?c#V>#|`ahYrY0I@v7vvyb!uMGr$A9-{clIClD(LTZMeT*wM|wCrJu#A0K!7 z0eefVY11l(&k>5l?{PKI-apQ!{+J05jfj6u2>+^FA-fQXiR$!TN8}0KaB7aL0pIO- zwMuV|5uUR|lNU9l48IpI=v6GHHth=7U%N!zszXN85r44wfZUX?FID)NySLoOBsvuv zr}#_*%0L}2=~A)>EuDt-0qI?j0DiKy{CH=4bF$?&J15KS{ZIGK^`cd|V)ROSCGxR3 z;JA1=O9n$)d{izeWIW}@422Lgt`>4wNVzCJHF^^kib+|JSy{G!XulGCPjJ0GDh1IxI*c&BuKq|xPKoH>AXWSAa_j= zlheu?TE$DXmpOK8UkB;7`djUIJh4U6bx-#CLZIFzg;hUa!{(7~L$tgbm<4Z>x(&?OK{ah&jl=FG5<&zt^6;4D@EYHZm zcP;+q7dS{{O>|O-x#^-Wp@8A%9}1CZ>{U&X-6A@13JdpZQVYUT zt%pmb9407xgV?Z(gC6PLl3j_Wm-{foX=4@TK;;hqv{kl+6ydakd9H)NM5H7N$$86d z5c0mY$h5n*m#h@!4}i-Pd;J#b%w-}|_KX_}kLCM%!QCKP%2AS+eML_q##c+$NqRD)G2nK2=b2*Hk8g*7!zg)vETT zgc495=M(fzUDMca`Ty)<)fbQ0p(hdxehKGF!iFIE3$M}y-sk97e!2cc{`GXdSIuNl zqOr@X!IkkHqy6nQ=Tr14-8scvj_lMTPRZPPEF)WYRg(9dLB9~WF$5&8)5p3`cL{ny z4qFx(5i_xQXC1FduvY$o7BqV=9=q3nQ~U>FS(A8)qzZG~c&5STz`(R3CCdAa%{5yu z=csQsSm>Gh*etfmg5*JXZ3&ISqccaSR`e&lg5ykg85riV(+&6e?J_(PB@Ccam(Ud& zY2U)U9)Qga_1BN(w)-R+Y>T%coJG85e4g$bJ<}^)eavFxpmdjAE=GW>kT6w%oK!I3 zKF0R6lzKT**c@&{#=$|4>Qu~tNJhyY(G7xAk>W4mag#3^u0{EY0>Gd6x3TXBej2M| zwD^FYz)id! z#Oaf`XJVi8K;;6v$^Gj1zD<;#UUJ>VQwmh-EfRBR6UNYE{A+LRaV_NY0$Jz z$4tR}gOfsVzE-DzDM393)IwPA!c02TiNJ>GL;a#x|cDFX@co|nD&IS(+UA^dxAT+Ni;N*oq`tMwC6T86+ zwECFyc!{Cl2ITuxs4Wvs^G84Q%jBqJ>9DYAa$+vwY`{^H@Cg<~yQ4TRc>*JK8hyke zh#AO~Qg048Pzigu{V_05(B*?@y^}xr>D|nq0a;7?nsaL5DLeatAiGuAXalM2wgII4 z`+4-*1=m($<*WU_snkvcWA4$meE$=ea7z_~PA^-@sdf-4aR*E7d>FAE7(duJODmpC z>-Ayi(8UzWkam56kUC8aCbgl>gdVDoif zDw6N`5lq>Jx4W1lyYod02<0KfUDZ#r+c%A9NTgi7;M2Nxbv74xo3TW*k)xFtRwqvp z^5Lt9bPIiiwskPfqQPyu|Ei5WVqCjd-Q7vl`jdxm3e@c0)=(0H?G|8tIdb>+*PHuh z!+Bi-jt;zP){qWqC!R1Eqi|nH2@eZfk=<42N&!1o33eUi0AV!k0SsU8hza^@7PU@6 zx_Z*=D-8IpjuuZVJt5Qco2s`t*5R=y|A^9Ni1-dEj^k zP1Z@8o`m(T=$TIgQklW~mG!Ml;<3^A`hm1=(95fjFSsOi^KF1xfh@kD&{HzujTRpN z%@aDp`MMl)(5HCS>rxrus8CyQ+!F!QSQ_SXzDd|5OPZ5qPd59Pw?J=GIHvlLQxxgd zULqL9I!$A)lgb*xzVV>@yJ$hn_FR&g%TZ_x8}8*PrpObN5wY_zn*$5*(?m#E)cBx$ z*7-o8Ufp@7&sRb^a*u{-4Pd{QKuBb3*DaayJ>vN8JebND<7zY4Mv%V-OoI|Xm3g_ zeWzCJ_xHCaxqpOlKIcO}5i1D#1<^>nXh3zDjsw-U?K;s!YZ2qZ)PtM;h*@br8yoas zPPe%P{Zxy$4^s>)Y$xC9ywkp$CH&6d?V6&)J7$*W??%TVOs{Q!7wi{(%O(+P)vZn>yHM9+b8ug69pd zGPaQF`0FAT$n5K<68L?9Jem&fm)bGO%QlY}x%Elb4P2F^Yvl&E57;`kcedb0Q761BA4!?@ zQ(z&HD`r@#{Wl#^L=Y)O5j(lb>!2Zg#8jGyOUG0E_|p5T8+Onk__q_r#P@;RN}sQGZl-va7Z^px6x}EUu?$f}&e%0VJv_F`(A(@dSH{fut+Y(Qmi( zx2~rXUF%g%kZ7>+VM|IRS*TEhzT`7D5xMXO9JLduGU8j{{lkjEIHfO9AfI{``P5me zMwRv!VCndL*88syi*XpQ0dT7#Do^I{S6=o;dX`(jv~JAgd{@St&nnY)^75gxv^B)?XWg5R4*iTdt{$<=Vfa^u@*YtN+C~_f2jg&TF9`ht z9WHJ4%v#dR6cN-pB{|$*xcB=^b(9Uhx@&$)FV#DT!q*eE*%g^EKPl+3tm8Ft@0|2R zW}FSkIt`TK?)KK&T4ExRh){uu)oJjbKd>3+$}4h&??J0W39B>cu0@?#P~QIQC^&)g zq03HAQt8pml$^ZI!;mNGor%OlBvZEA^KZLiiu1XOzW2)LiKK#2!7+7Lt_4%_!G^`$%LDTt|dGLBUPUW~9L;@QR0$74CgHA{IZ) zHHZPp;0>avPKS||=k@)PKRX%qYKsW{kikfMG`^0grW~ilN9cd%wsW)!q24zg?7&Ih zL^9;y0l7qr8I}5%9guzTlQkrN$ot{u4>MfO*wql#w4G_=`cK0;0?3s-nxYA+YMdQXc7EVz&WB1?lrEgx- zQCS7)yXr*z7Luk5zz`K2tdxuhk%qysY77H`hZ|Yne(*N^%3T6>e19G8*Q6~KmmQjp zlO(5{IH6n73+5v0Qr6+u=G;+s!h0<4vHaB*qs6Xn=O0Y2@YV#JW+|Lr6$XC`={IP@ zyAQ}tAV+W-iGUL+^mNfDV7q~0Dmg;{*lISVFs@szNRB$4OW>9FD4}u zb7Q%6;ptBKP)$>~ah0RZrhfJAd|Dd#h29rd+7pfV{G7V+!Xr%@GNcGe>^M`QZX#1k zIqlMZSv)&+&U5k^fPfvv(-UC|jgF^vovL)enZ(*Uc(HyuG>a)! ztE-0%r|)v@Ad3`27HlOd8x5qiMu92X+0&R=X$w6?vzzj7BXbbL9 zV{c6;;6h2Em_g*`V8~k)OD8>c5(j|_)D*Z}^fbbeP@({>*DXNwz#m)#B7?sG3cs30 z+VdHj7-s;G^9c0#h|D8I2(WS0e!7bxG^;*`qjIx;O+@g{uUni>9Jbx(bAp};aItl! zo!%gbfW}E|V?FSwK1X%^U1=J|P^bf}GLmipUN5<85N7(L^2^D`IJj% zfN<)QAVu~fLT$Y=LCVtn}35d$L{woqhfK&g$ey1dUj^a*E24 zaLeV>wAMPQ>a!PLAbw>LxAcvSMzA4JEP^k~+95AC~0cp&kCI^*)@6FPdoB0v2i*uq`fRC3Cg7hzo-M<*{ zE9cWkM`X!f-5Mut)_IrSlPFV&a;ln(kV(#rfFOJq52ovQ)$fK|*cF-={$9qW^7vhl z=HhrV6ximhHH`(g&@gvXH#N%(3F<{p8^6CHsJ{Om@S95o&oA|jHIO94mJ9PDx*P9T zf!N;ziBH1f9KmSQ^?FZRXa~=ZiYVnV;Ya9{UriC9(Kg>pwu-_lrX_uE7Bb z-zX`2nDNXlUH&Qs_HPL7a@K!{sgn~Zg2~N>ojtu23lGh6f;i{WnIcv54NMG;EE-2~Djd;1C2UIdSK8@_$g$RU zv_{`B*!5%gfB9)DvhRO%uoKxW7j;2uiaSNmAs^kkz5emzDp+7h1+kTBgv?t`( zsimCmh@{rPC8x9I5tV3eaD{d9iE>=Gv_+}(rW8k>juErw{Eq+s!)LOeWcYalUkM3N zw$%%6p2)iJJ*4~h&MFZFDQlWFPlBbZ#ST6y?`5mwO*1u<)>oCD&o`0bAY3NCVTPYP z0{Dzh-zBb3-qes-6G9}1me<3+F6o6hl9&OhUgVsxk zR>CuaN|k57=jQ?+&m#t&uyCjE0igJnZJd|^yG1HH9qWE!|7^+YRkPt#`9 z-T%dH8w;oIc=mb7Ui@gvKwycscdDdM!z5uf!V5YTRRf7pWS=C?s+lf2)a138^T;U2 zg?S@UIUkPLX!lOmhPCuTfAK%^_18sU85l!%Z}_tjwYTBX@j3SP%bLfhJ90-(V*Q9L zb}K|Xix#-C&gTB|9udTjm8W~zq7>I((b>uAbgk~URM z&&|8%|K#^&SZXEFtem6{qeSI#v>fda9mPrc8qQXL@~Quk{pyu$!eD9tJ%ix(Pp|J; z&4Gxi*Wi8@3)ed6#6lOLXbkh;oJx015U#c`BjJbF*_ZG2f@)}1egLy=wul>R$6tyq zXa8=#QxA8I5JGY7d&WKx3Z`!x6yu zfGpwykJgWg(Vw7mjVagE5wNilJ=h%SgI72&qFa=D@~mj~aE^z6^9(BUW%#q5-?_Dg zLnHn8nO#H_54OCQwha$y6t6DHNC8=yYVUkdTJ_Fo9-Lop5;4h&xyg=dh9nJ!f$_SZ zOKWC|lOvUEhN?mTyG#F~ZkYoNU>!Pwk)4l+)UiWvK{xFSAbK?SzzZ0jLZJ_UkZE^+ zZ7uUTFb^|*M*s#vB5zD(xrqf)NNP}A8S`>L_}0}=4Hx1mPeR8i2pWgO_64`U?UXE? z3aSk?XDtvGd+TM>X3P6i)J~s7^lT@AuEUX-NWMEfX7Tupq!K>vfQuWm@N-DD!9t1JiHJJ%xK(lnGEt+tm9#umZGE&|@QOAv5LDZ)4O zUcUsHfY0lL>(#~r^scBf3B^(uzZCJjaKctv;NdQ#nFQ^X>`X8xkH?8v0d{dWVvO?I(rX0D1TnCOI3t;^B*VHcw* z@Z@IX702+YTUBgscusrh*Za+<-IUJL5)S7oSF-ffZp4waAU9p#RiNk7nlC2cscqpQ z4aOKePIAeD0-+P%X>y*l%YJ`p{6$t@@0aBJ(wAnutqL#C1nfH-XKKIE?Lcg$8?d;6 z^Og)^;DvhGw7WT?IclbBuS{T!VjfaUjWpTY{3O$e1mrA54&t6=(U1K!tDa1FqDnf@(=p^;F*C;yU2}`R(^Ny(HB3 zOha7eHsd{D2&Ceh1?(oo9aRB+t`uABKNVh)B>7pn0kNj#Xdy;K;*nLUaZPo8;3j!^ zl59LG4^DlsJv-oFOW#ih!%x{*V32X4T8CX^*9Q_4M_r+jm1o!-+sbcPX6n4{J^fbK zsT5>0&)*qdGn0{cdbvh7R@?lh%{a-AOm0M~>xH%XbnF8OxGMD-koLPn*7iD?=cjLb z1iC@Gtt{&x<2I^&6WKo64ViIz`2s<;%#D@-P2#Ihc772;#exm$~5wPJ#gun+; zmdjzN0s;TYCDj3mOw*(j`N=IxhO`<|TRNPJY<5>^J>p4&xL&rP)KkG>ySxH)k_h>A zcT5)#x!E08>Q!#mTRwnFOI|7m_N25*&AI~ZO6E+q^_z;DFE%iNeiktxtto@h0Q_b9r?!TvM|af^P@5Ru{Ne5vpa*6u1=J26S^5et1|)apdCb zMd(c{yWUr|u5Pw8?j1nESM^msXSZUZ*`}vrs8dG1?h#mO(%=@!AVR5Q4rYvtDZqKv z^H*8+x({rXRP+mkT^E1Y!d6EDJ8t9+D;Fn*Nw9~NvTdOpW}J?}l)gD-vy6F2KMB2# zL%;OClUH*uc(}PL2d`{@L{mwDA4{a-@%V2PBTkX7r+P{_&4@y1d8<-GZg@eGsc+FI z2E|EU_u-o}4L+AY24D?v*L;{YonSMLiPHXvjHdy33-c@KUP5fA1vkc@t8X{8uMg*k z$0T=@sf$duo*H4!)hbAX*IxE%_8W9G0*`2JE5$YCW}2m=@(_~EmI5ue@W$h@)HWY9 z^f}j)e%}H3l?G<6RzeECmL@VCt|vh+ynF73#s22Y+YH~Efe6MI2vy?dxyUIIeuwKp z6Jv?CpUp_@Z73NFFP866E-bt(#$*0U*Q$W$D*wP?G5r_N*R`b&&(b;uI_U}E{rYXV z*f08p(Cyo+1wHY~ulsv5Lq1DT^gP~|0P5V{SoZJETgDQ7$rKEuJxp2%0si6D@}8{b z(RmWaFIu=yA|XRp`b6vkrL~0&c+9j?p8J=fg@_?T2i{a*=hM<9vcZeZR-An5Zsys= zoK!g`e!k*XvAlgk_rnSx49K}^QyejkWoEg9h{9ZUtpt@02nu#Xb1Rh zX29uYf|({gS1CECHs8R=#8*@;>u^wTj^|(b`Vi0 zY=}WKbLOgVFx+E#20nCjNjP$k+2n%mGxsa>fHu_!-iC#5d5C!UTGGZ#3@jc_r*eQx zY{M6I$I5jQ?!Nmq8Ji1nM<4015ZLg?+Am!^`gQg7Aq~i&Yges0E6;|SH_VfNFH}qW zT||(B$o$!**eY81oYapH7*KYXxEUFjm6XYQCqa7mhN+h>F`WB)u1(xSSKC$tr>&du_=uv=JsfqpB0qdv3z^(a z!ofu0yqWvOHYkWaiwu_pv`2sh^dt@7t^5AtN!Si>j7NxM0N(x(+>Rp3Z&4~s?22Q3 zFPheT`{9L%8Td_2Fkw?1(qF0^Ej?9E(P84$_$8!he5V|q3I`Yd3gOTd-SH!}!=(4! zz_kB;(KLTvG%{8KD#z5UA&8TGkl-B-7KEjef$rT;LsUaBw~~J|#*J`1B4p$#&Y+h< zWCjKj3OtQO1q25$v59{QI{D!sZ`h!?W;t1&(bQs4g|pOJr%Ca0rzOklSh((MToQ5cx?h5Tuv?@ey%hvwb5_*6w6~$dWVJR|_bVw$M z7r5yY2z2;#jmKI2>2KA}?;lb($lq9Bm~hZiJ5SkV?rw?4q-_4=pwNHGV6eof7L9U% zwa}7hOf91zKQkguI41vgglR=yU`C6HRj;IjnSA!{ip6EA?73utQJ|It2Frx-uhOXm z$0)u#x!EtbcbHPxFb&f^n&LZ4 zG$DvK9in5-lx|2$smtjUT_|VM-l|Zb#WaH!>nGa`=;Ex1yxXrOMqWY8s95W8sN<+o zL~2~^WO)11K)TpSf%$KUX=mz$1+5gGj^C&YRSDfSqrd7x25FiPwvFE_l!Y_>CA<@u>*w#A{Er^*fe=V`rNl zb#a!M836sJ*X_WuMup4PlUcqdRfh+mePc0Gj(!ln_pG1#8;S*92va)TsQVHj zOUPj$EN3(Y;gIM9gZv-=MSdcoxhS95id`d#y5_ zy}cS%HcX&h1OmzO*sOSGJVv{jmnW>5gJF#Fy7jY~P93Z{pe6XjLK6iN-fXm+5mV?M zO+tqtB@uNWd{Xi}eCSH9u!v$}5lNsWMGj_Rq|S0DmWQoCJ zw2=ezsFgOMmva{5BVwa_yBCn~l%kCShg}DaTf{3EWOa&zYFSQaZtB2uxHBNW!j(3s zR8SRpQKl(dd2~ZHtfeyxQ>gQf?WT{=bxGxkDzITV(?S5U5K@<)_6oH?Qdbq(fos-Xlc5)z6%Bv`F3K0iIB$5lQ_W(C-d zLAc)n2hTmkO#6>ogUPF-3n5zO?rT-!sbJztNZ;gb{;d_QdkA-xn}}wqg^yn%VSx&q z)wI0!z25*@d0P4T?{5ni%g-9b0~ir^oI`s!mB+T&gKI<>8Qg8ljnFRM#Pt8RIRl5e z^J=Rj>(m)4YxlFQ$WIl;wj`z~+9(N)Xn1og~IjbJ#@a{BU?_&md*-?z#S`K^(XSJ*QMBY=_#D7n#ku=)* zHa%T(Il=q1f3s1ulIS?J!CuKxv=XTxa*w>De9o+jdF9gq0o`2^vj#d|Y=Dr}{sUTH zv{g$G){wgfFARB`4?j^>d~>Y>;k(lsl{~unajhQ!Dk$-e^#H!Z6$B1Sa>Km`e^QMs z-7~Q7w5?1+BtMq?88QjAS5?D+l)QmdQpfZeKzE$Ax2>x>ZF1)lyEnf#U6_4g4gawU z58wK*R`dU#JSg05JsmhjtMoY&kW$uiubS&Pkc;@kA`Nd_8mwSRWa|l2vLTw%=fN3^ zOWjDHr!%6;?I~m(fyksa$p~13>L2UK)q*r_QJ`P${2}YD5(}jj z>zL_>xu+by5#JeCjh1{GKc41L-36v1Nj;7F^LKX~tO~H=AVX#!m|a0xN{!PH#}gpx zgzfhWDYsd}VVKa84yb`sh3+@4L(uKwJFt(tE7HEo>%h?Ksdwp>~*_RliVnEJv0S7%Qm#@#$0gH zhX$;uyiXFtygbWORnIVFcL-4yxY4V0gr)8^k=MnC5uA4OQ#v@-Cm*&u&lp zn21$`y8xV8BANt1vF}|#0()979P9jrFGs@%a7+=Xs5vK?UbBUxu23g~P82p$sMWk` zRqc-Y6u)l-B4J6bbzY>1K;5-iZwH7l&-Z4~OU?szQ*#9Sfsj)tp(yIjVB{i7gBhlQ z()AS>psuS|IcZp`utvTvi8P0Z1Zg~@A?%tFE`4P3SbUicxV~|ndTf%Ng_p~rZFlHq zbCa1HW}U9@3eGg}c9P=eOxxfyE^4`u#tD!UlPt z3yuA$^e2M(fDC$=qhS0`q@w#SgE9>9VjjwZg#o(gdA7`L5I+n`{xfd0hUH{HH>L2n zNO&8~vR#N5PhzS-n^wZaXN34vVeZ%N6|%)(KKtHTjvf!tdKF1+XynY#J5lc`#D#?4 zbX~r)_?3}^q2XQO(1F6)bsSF z|8&(ZP~%5>!MV?Uw|~OxO9eWhVR=J$G=N<=25|5DQ!nP80k^8ig+YX!Gb6l zKUR2A&H{H>1*&D<6J@i1)Z*7j*{Ltrl{fWMaTGyEbbaQ7Eb6p{3Q$J(oaOj9m6J#) z{v7A>V9q@sFs*R)fMaOVk$lJMUWqWjU4h5N<=$&XNwoV zGn~oYRvi>$V~o?;N0>e-dEs06OfTAJ4{7uSpNc-djHqAMo272QfH+&$mt1gq=ydi(cx+&Q%Eq-!kEM60Fi7JBbC~f#$Z2f(!s_lR(HjUYm5x_{%zD zpI@hfJ!+Vs8U8l4_%ql9nsj}zJL8Wg*3PpBlIM~ zTmts-Rv65-M~bw*hFbagVo-Gl+9_z|%yj~^VRm{}u7jQ^D2Vx*esIhwwP&4de-3or z|!Mr+v$Oudjofbj=a?OoXC! zj-x~687(Bq!zh3F8o-dm;O-}l&}3E9q~>+%B|ZfQ{X6%muqj>(mb}~7I36z~_n(9< z1S-iqPVWJr<*=Urf)x^Oc|R@t`POJx^?R2ah;7tMk3)FzfIdEZNgyE}Jt$@{Nsw{7 z0R4g^E@%0z6;UUwC0{Vy;#z^7JGl%{!Z>0#PyVwPX5uYM*OUjUfrq9IHXPK^#@GfA zQQD<{w#9>i!L8{MiLn@5FP;1e)#QmhMIKhZFmVw>W%JQz`n-a^oAv~)S)w@<`P!ON zRdk=>ixh_aloz<#vS}FIFtr1;rS#EWC=YT9JZ~xmMOOmqH+Q=JoBp3E-DEDBdVzKm zWv|$@V4>FXrYD$n=V$1FSHA2n zOuU#?2SgRjwI}EPH+5!tC+S;1@$*G1Mt8A& z_@-3R(rK2u0vxFb;#4&}0j{j{$TDZJ>)Qt7kllK(Nb843;y23FjTiB!a82KOd#2## z$EL)-D+JuuVCAw%1xX|R*=!aJ(+fbakgy++KjBbIo0{=1B)t5Ksvas>k=Py<1cs}p z9mneg?=EDv50cP&dw$d@UMfaArZ%fFV;UTN8vzU~319E6f*-Khq~jx;uz9;_bn8JQ zGzqoNjSd95dkatEG&rUP5N{iA`6lyvvVf1|;jb?^Yk8SwQrL>fY8GUv5^A@@utZcW zKZYsycK0WoJ}uoPCUm}8rAR5J8dgnBta8qEuPS{l04u+Mt>~376}Ro?i#rQ^t9NGY zAJ?*-E#CYCseF)SbT=>9HA3AtfL!?&o%d%;9dg&%)Io))-IJW*F)Gh|*1!eHhcN!S_&sOE7wPhNU0z&=_Iw^}^DlW} zZJ~fG*Q>9wIL<5@>GsG9i4AaJ88aUl+GW^$G=qv+@~M+kFVr_+i->~OcQNN1jy2fu zd-EPk8)g6I@%-rnebuT${>O*sF1nnR`Tae0Kp(TGkiX{s`k9uC0AB8M&qS0D9Ps$ zUPeTu{eYrjOLP!d&4=4A64y63SH=epjTvIkh)lHDW?)|NuopEJgwxs3{+t2@)lu)d zhi=V>Bo#keSlw35l(;w8HjHo==Z0{qhaVI-TEQP)lJYCZGYPwco8wW7Qc6feN?PUF zgenqhyl3w~)Xu$VG`65S$=^X{a_cf@?N8f~;_ZiB6q>~H{77 zoB5cfU^g2sBj2Tbf^-U-Rc5Evve*{3?g?w6;2}T(=3lOb-4uIQW_6W1Tx<0f*b~0p z`_A#XpteXrzUzFR5eQusKd_2iNkhUNu{%FQgkS|fgyG3FAsALB6S;p7OI{=9HlK_% zwTVh2NxtlH6JKEQY(+5)`|=v*Q^vEqMeiXBfOB)cnqi#IA!7RDp85(7$4ijZ-Mf6) z*S^3!j~dW=YsoX4F&dA$9iNC`23}l1FRG;6f2dK@?N(3 zXtk6u_l8x-6qi8Aesu6m1dU>Us{t5ejeHCC2I_e7u87*O?E%DCBY#!?-Ja&UXwlcC zOE+S>vopMBLlw>3p_V5tjTXg`vN=P8;#_+$ZX1qaM-TT0yMCm8mqP**QbgnJ<~kO-68`Df;KWf2e@;A|b@Jpt>?@m%0lMM=YZfWYUOz7phbtQ% zImzCXe~^7!=btI2S9*Wd>RF6VIh!S~h1FxgpZyJ>HJ&1+aFDd|r;-V1z<53d;b-X3 zn7v{K}td%qsu)Yakoki`pQsvWn#+YlnVFKP3f+s z24O4Ky-b(#tJc+95zHC3G7!O|#32k_VU#9`} z-T0{EBbpHUwp^%=|OL`oNfTk#_96UAWmM9_lIGg^RYv(z)RAx)dY7>4WUX)v$6Ttx=Lq_WBx zj!}sc8N7?GpQkH;&4|4UYVprIZzBFBmQI#n%isV3N}G}K5`2_0-6xp^dkyggWHlGw zh}BuR{HJi4Kwo?w5+U%i-rx!ws?m2#02a5mg8o(D@ z0X%~4HW{ioC>Q;4KLl;D|AR-Lz-G{D3Uzn+vg|@CxOoCm$!*?MaS$KICh^h#`;h_~03q6Gdwwv3VY#5pcE#;898>j(tr{9OmHhfL?s)fpxQ*MShY{EMo9qt zD)s_xh_vwmD7*roSCliQ-3X9iQ8E|$YuOn^diKZ@Vl*hCP9&wJc<~b;=5`=NW$AK| z0#Trul{q*mzzu&sgZ-PItvfgW^^IO))Dm+vesIip*ow2xVl}Uhrv`4bRq#670`0!N zbRSVv0UV7W>wZ6VQ-C9t(v7X>&nhh)BIsrDFYzlw9u0;@OoT{U?w>Lh?FT&i?Pi=4 zCxSa+`XG#unHD}Q7TyzBDzyb;KOb|F+|@xvfaN8ieV$y zjp$PFC|(3MHe~4ccMnID%%AhyW(sLY-g8qliNE}`oaxU^B88h&belX5*jFl9xJf6o z1^-y@kpj+kJl(v}=u3Vbl;5#C;8 zROJNrPxvxbpQhh^j=;lwM|c>!cH)1@MwX5T&UcnUwV9~%0#+cQ<6!n>#f50ghfyqk zhjt1?N~qvb{0QuGF&K0;ksCaW200jlzXxNDyf{J`J^~1aSMX$(@k_mb?h8ci9b{*L zW^;uj?tOY$LJmUiDY(3@#7b2yO4oo;t`c4(hYunbL{QIHY5#XrHzWyoXz$%&2{_mQ z0F5K-a(YPU9pCzGT%bkg1mdi3Zk6D|VO-6I@8_}w4~k3NOru{~3@kV}10Ps+8t;jz zIg=&GFHJWoV3OfGk*rCpC;aY66f?nB25wmI-kU^HFC)Z;kF9Fh^{5=`0 z=GsCq1=s>t9)v8Fh3T6VSZ7MZkoR!);;9F+VEgC!Yx_r@8Vygy7JtBw#3B-Lm~Z9+ z|2?&ZS+*p+5B{1r5WA1y{4T<8nTKbsXwDRE;w|he!7mLj3Et?RmxNq3DM;E+|9Ix@ z3E>__e(o>)?|F2%SeDl>vkjd>O|oYHV`*0R(@C$ul@ws+L*L?Q7=F2y%!oa#2~XYg z@5KLef$QO^cH=cA@Qy+rLLl3V4kNbvKL;EV9%OkP;+P*zAu3`&NN0TI48(0Y@`P>w zUVQp!JyZ$_MHg{M#Bd#L$|daK%kY21f46cSJ4fl7;n@r0J0*}ga1|kN`7|p3E#o~N zo-s%&PeSN`j8hDGioW;D=&xNw?e%h=oufzzUw)f`frv537GUQa9ww&5d4B@=$8Vs> zoR1%F-$HirO2>++G_HO8LV8NbSAk+F!A<5`kCX51sEBD`_C?Pm* z=?AIc16gjkRAW~jTfr^zCGG@cA`vm=*bPqpv)JY8a1zxAJ>*k`{a2=!`HJhGn;dD(YKIvZM) z*kC>9N1{r|!l`Ho-(UpOzo?QW78EcjUJiRmo27wEx-}gE(8N zJo}I?>?DM&Bi)^cpdnu&>3qB-1bkmc>HmZ{ffu{V)PwkeWEq9YSYJLdhDMXbowYpG3pr1 zOpWBqBUwJkKzChp{~#$@x==W6)*;SbJctT3M8$vK3UM6OpYdQ5C=9iBRk}=z-?5_s ztUj6%B>ffvR2^e>6Rck98&DttbAxjc}3mT^Jjyhdc*aOsW9X?esR{ZaMo&d z2?}t%iJ=%%(yP_|Z*ZSs0EX-GCrUDuFuWigM%lAaOsQ1pl2DAqSFJ|}{X0GpDRDu? z3-vfd@)vnwvt+ZjxHs~qN&Aay-a5r{71poP@j^p3*EUiD0e<|FgG)%*Saj&(Fz*`s z_lDJ4j#9>qHdEH7s&1qvwZUGPU>f4i(iSCYm_*IX=+^>L5yJXJwEf>5xJp=9>t5oQ z9JD_172v|>PH6es6&KkbCotXm&IkPu0>ax=;zfddi&k)e9LryCAI6q9(|Y;88PDA~ z_tFU0I#~8Oj^b@PhCD7Ot9S54-VjE|UX4F-5PJniXtg9?OTz1BLL9l%{LdkQSvxLhM_iNB352dby`pBA92yiw9^a7} z1CL(>3;Q0gw!GrdO&;I-vQ+S5RqiE0Zv1muX@X&0bQfXi8eS~|Z3$}8?3^1a;qjbLP% ztP0EHPttiBYpi^Hr$J5A@wTAw#Q?hmh&n!G?7A!^-NxF7pgjVNcz=qNES)+OqWmu7 zUf~tEK{nHGOU|K?{qmg)Zt(DZqL;H!T}}Q}S0snQ!>^BU27RN@d^q=Z9f6RzisqR< zhdUo&4&V1I)&F_fomB9$4L8_Ulo=r;9X6@lARllL%bM%lCt}D`dH-;SKH3G2Z+K0{ z_HUbud!=RHU-MxUU8_A1ykCQZE)9+X)Jc1d`+uLyf^~EER%!G~v(sLRE^}Ov%-Oca zBI~_SL%1}s;fIgoHDL_qS6?eJk&LnoBXayOV~%_$_~LK2 zFQ>d_yiTtC=O;f0$1?Z2r-a`DK{$Pky|8j9Bo>+}0U=^ag;!&X1gJ5<;y*PY+s}BN`14x@R}v>fC_HUf_G>;;#xY}5LvFv8W=mRQ zNUuwg7c0TPcX|SqgA0K5L3le+Vb}@R!s3^%L;!OHP`D8&Q8=0tN+mU1klkvk87Z^#py@noAGURGv!$y8@^ zTQLwsA2j}KLPWC|jZ~T@T=)R2z|-$vRY(YdT)uj&T-xQKf`izL_m8rF{1@%@C+OM8 zv{)~QyoGVv{XdqCr6(H3XsG;e-R;EUiI(r8V&mJIkBz~LJsY zSUd;c$KgT-k4@A#RfO1kUoov(A%IM~+KJ!Ta94PfAY+$0?KbyVG^!`)Ua@l6ru8eX z68xoD4_kZT3=_ALP@?Y<10Fs@;-32asvk2VcVE!8}&P{o*!bB|(Hj;gt zW96zS-%9?+TFqo!B?m?{Gr%vuIs3xHJo_V-MEgWt>fbl+e36vMpTKO*-d~xft-T-b z5R>*?n`KQ77+U@=Fcv-U`QTIv{=)&dtOh7WxO{~`ghT=ArQFtXzchidVs=SZ}sQZNsV;_c1==?ptBGDYv;!wzQUSHxZQogDYhfvlyoK zGq$OIRz*Cd6oEZO|H;PW=j&O^oWZ9d{JAH-zni^5m%STe)WzYgr*z0s>Oa(z0B-Co z2lwa@wh$!*p`wsn8ZjVr^tl2z(6+hU;?)w|dtQ;D$r(}bg$_Qv{f&&S&{3DdMs%rcPE=APZ zrtwJd-^*2e=GM@JBoUtMwNP$5h~rOJHi}$#mFy!zI%SUzN~auF1JWh@q}m26x}*)H zYghsUDu19d5=##d{fdXUusyr+;nQZP^lWeK*P%DLZ$o|C9n@U;*7^M= zM)B|%i2$gsMvdF;5pa&UCPU-w5USq*jKdrUmkyKpLYPJkl={JDqO6>~gss37Gx_B6*LE)=CC!E$m61R_*%IE4aT|FilH$@7Rnd3nk9ItzNM>pre z3jm3xFuwnIP%0V!6!zdl4O(Ab;P#i&CuDI}kr6!ScpjyeRJIfQ>w;Qtk9e$c3Cf4X zeX&gi((FO#`*ENCiL=dl=kIe%umMwC>k^~@FIV^b5v4{4e5n2*Y7U427UH+n%+tBD z2WJ;qe>}|!k`t>9#if82r+{x!@LcQ#=h^tJjk@mKBA|Ko?XHK3BVzg8kzf6MNDA@| ztFyr1@VjRCUE9zLkc-aGQrwV1I|p;8{V+cvw>5jb+|Miw1pycCS&Ne}t$7WAdhibg zFS0fPap89g@Vk;K6SyB`PR0woqG4c8ZrrQba z-Kr2vBX|SOJ`QJ5o_&N>v7^g_5>TOoESWATAenbUl&?)pJ+?Umqf@XPXu(*OQxW9o2F@(iY!{SWd3+24JCB@_V=M_6*R~;Jw zM7p)qUeG35$s0Lnb?hqV#0i<;6`1HoxA+*$tDrkd&>lBa2f$swF77)ZtqqM9mm?fp z_PAl@_FkexZY2`W8?3`wLcwP$U^|fa#*6KKr|FW+?=F{8i7XNqB(ca_1&0Pi8s&6X zW?Ln}IOO7t?c}!Y+eaPdXWVz!t_@BvYtCr~5NN~Ml>MJ1Ed*u!a(bhD@2REGX`L>V71^vG zLc!~OyI{I!KUua>kQgYDC4h~W=#OtOEsOK`2T@5~Uh zI*+*)tXQdA1rju<)tCMJkwvN6kf zI<&=K`G<;{MpfdzdenW>y<^_}@PoVq{NMZwMd_4R4?KmoDB_4yyqdATsDDh|4n5k84T%xgFk6h{qO;jBC z?xe57*_q{vS&3tzKQqU&#m5*auqqTQxUuBzv{h`iQrCy2ZGxrQSZNDnb8;3rYjMhq zTRkzjj5dotcC@CV5`@wP#kyc*g<{qgRx2r6q5Br8caOvG?w3bI80%ZcBpT~8_TdgO zas_5K=j@AE{Y>s+L`2sfTW=BKcKNaV0k}p@9XP!}9}p05wPRqWh4rkL$4E=cpV8rd z78wbB%bhU4l?@i3*I204oY&acdhAo3_EpZ_QqrE>SefP0&UMt2_Dn!?%DCDW33B}bFEvI7OM?)>XKf)p!b!!@4XYN*h1 ztDM8YIjF6{dpOM+IH0ZWnTqzamX;=7!D^$X4mbOQ5mROv&tR(c8ASSInovjf(r z5+}e$VB=_`;%#iJ($4VpwNwU9)IoN2L7kDS$ClC5=%WjN2!Lgy!!^f&@=DT2U#o% zd9KDIkbRY2NiHYCx9iFQZxV?E$- z?h70lD1Wfr3Aw|1I1nP@)$q261M}`eY}o}byf*q(8zT3FL-2P^ofNhcIctWp+LQbD z(653CjUFenoCLR70C^(1W186JJT}*Nhc4A=zrFL=nvK760|>?O?!!684BTxEV?#HW zhgDC{%hF0QlK1-o$d=$UVo3fA4IL8$9e1FYI@6iB-g^#PhIxFE++cZtbcefdP8rj2 zn;Tv~y3SChBmp~DbcMKEetnQJsn*K+TVM7DX1BZPMqnQ!dfC;U(~)&i zWVw9phONtb;+E)`#k4_8*>w`gf8p;ugd0k{m~?c?h6(zZ8Lw^k`DWB|S#hf1-fL$5 z#0xOxuCh&DBAPnR@)f$}#7D6aRiNr0VSG55hlGv%*vxu(NMTt=lP z>SX~x_Naj0wlb=;{fLgwF`nq%cK+rUi|RIlo+6mYv;jhTo+>kR#Np?d>kRjku*0AI zer<|_`57(eI_+OQuj1*Ln~~Vx3puaO3%v}cf!}n1b14H+B+f|y=L#@Z)0CXk!7o|t zx_v%lPh2k7tjti>Nlz~k8oG~?ppQOB_qIRRf3+{G{Vl(3cg3Y^azy#Yt_OxF@k75S z@h|`2+UXkp6eaL(kHH-HY&X$&Z{wumCP2WbxPD{NY#~6g`>@fceNl62+6@9&Tf`WKcTnNmd zap*os+oEX3JX3TWt8C&BXp${lnH100F3CQA(Q#8nMOR7aeoGt#$IUK2mV)MsG*R&{@;^Hb zZOzrYZMsDenKVUU&C9&NkK6*;iwx4X$qoi>XNjNipMn+6c)3K8b_fC~c;Bkoqcv6f zkhix?hOWr2B&*02Nz1vt++Weo(g-_ag74%LD^G$yv8xBnWNxY($xL?2!2dkn)9OJ% ze~((AXMhsBjAy%j2TYcSY_;bQQ$Lr2XE;lAT)%_XHX{3(J?;T%LcQESQmm@f((yrN zTvB{N(b)aqFVboE?5ivL&zBKQXBmiePu;+hVN`toL9L;zx!(3>(%e(b%yKQ+=iPP_ zAq+rn0mw97XkBXm$lLB!WuRBB=}^}1#o0Z;eO2&p#t*dL+9;4c$6->QcS>oS6Olu> zMfjgnh56hh9)N}A1N0mB61Fv~e!vGTOFRv0XE68Z!oXrorC>4LE&Ire0a$2xT;sNv zVlls|0q8T9D6+|krI$L@6X5FOWT2SGv5 z@)aI_LSponjMC8El0+PYj){2|E&klfZmO)r;mq7MRte|bOy+J%loFXD3 zE-H~eMapyxDBY4D?Q6W_ybTB=_8L&I->it{sRUf)Y8(~!B0u^f*$`lm8;Yn%8Y3|k zAG4mft|@_cA=2jY@uRQ6!?lUrAEa0xYSJacQ;B}d?*sByHF2=@OB8s`_7B;HvPQie zX~@E}f)#%a3%lw%2D?gYZ!?7-Ht&XAoskfcTPKh`8VeA7y*i!5G`e`eBgONVLMcZ$ z00Q0RG78H;0Nm9{Fz0VU*2Z;icwN1nXe|AlQI$oeV9<6IA`y{AIKT%rd;W*KG|&Vi z@HW)ELQOvGsEC6f?Eq6fV*7UOsECW%mP)RQ-*=E)u5!qc6SitSqXwS%>R*7yHgFZ$SF6GeQNoj+JbdW?nF#>MMi(eY zCK?BVRmf94DBU6ResEofciyVC;n#r0E_!!dZT0|MhJ$({6G`z=aUur=zd;0m?E@lKT1>7TlXOws zQRCb7(Jh_4yhP+Bd2-0}k$Fx-7cbg*JMp4#SCCho9nj)aG5qFo;UIn>I=*QymkiHf z%MCJ#OaQp_DpZ?Djv#ClH$pr7&SEjZZqci1SXXB<;4rj3w8eO<*s_n6gGD_u>b@0? zZ~#FR=>JtLNV;OGZUgDB*gYJQ;*?wvG=HyTE6z8}BJRXe46W{hKun zG2RmxQIKG>HxdlL{=IcWTWc-J^!X?Z7SU3@Eruu}utsl&j~E43nekB@WiQzMe(<&w zi>{BkaWX6PUp!FmG3t7IMbZhJdj^^yY{|NY!kvu3n^EtGj^||L8u6ESY1TN)jD8;v zNX4W3F-8KywWV8)RufdYJnzqat?twZV*p9mnf}$jmlGDTDJ(3eMNPZ3wG1HFv=mu_j zt}?G8ZfTgr1Sxtx&t^XT%UU{h+MDg?b4bx;Cq3w+Ve)GsGO{s$x`KDweWTk*frl?Z z%6ii@kb9?v`b#9oxWllC9OLQz!qj$5Fok(>vLYlWR=3124ujolIV4BPHZg-t1A%On zUXb``s2I+T$=aIrn{*A{o0e|OMyZE7|Nzq z3h|MgDL_oD3FOVtR*?fEq0i3a)nKuNM%M^}8yo`I3|<;g^$@8)whfy?NbUG5^W} zg^uzk)&T0~eMyRjWDwVIhEb@cn(_AZ%!w70as_jjlAmAHPh*0RpywgeF$#0%hbuOc z_YX1tAgn~0f&sBTVBYMZV*63UHviU19)iL%pN_cb+G&GhN6=AZXSkLG9au+CT(P12 zlq@b+2Dk37wN+s`7HsNo3Wwj_AI6PAw6~E4RUeJ^ElQ;$c(4kMRxTmzZ0!C;JTpPDricD!~ zy}btQ?g~IqF?%tK$5xQ;dNAgu3`0#YY(SD|2XZ)_chd2J4-J7ZMPJq}*MW666l( zA?Lxo-b#-pPY`LFUcMqJK7%qUPM07Q$Mx}16hrgNUBnLb#V#05-kXdSj|sB40$s54a+Tes+l$u2xuG4^D_iM7^>3v$k$>8$adfoq2dC zfo!<=A$oqxZ8|O`8=f1;l;I;FY8Xi@!g8Pu9Jw5Kj;!-y295)!@z?Fmcq(H48Y%=O zgaJ9}dAECi>1cbVT_jtA#4@u7HeMu?@#F%l*f!X>qTZVh3UNJ4{UP1y z2+Ueq@Sr0#;)b%`d|mA zB3r3hWK0kWumB#Ka*rOFI;J8Uu+z)Cn>vp;>)X(~(Q3);F2C2u7Z8^om#lk)G7 z2i3m;+s)$S=sG|-L1!Y~aOa=H&W9G}%DdjLz(^Z|6dmUE8lQ)lkbCxym(XfwXFJ!B z*?OHeRkg7>JUA_&2Qsy!zf)0)9KSKyg!l$LmB!cGKmkITN$7r%9>yS zi1xnO5aPGTmqgSA26yO{tHkJ#+><~CsEedsyT&-#7uvU)h#lXQT0mAT zkHjCrk5d-fZY9fScMPq@#J83M{HYTM15k!YNh@?tw_eejQd0ms3D&4CJfz^7vVpqy@bUSCaOm*R``zpTs&SU_NIuN$faK19s9ZPH|20Yq&%7f2L% z$a?I=9j?<>rxD2hk{U@*$@WY|ss)SN#-4<RsYVqav*or%LNa z&o$d5?qcTNI78`jf=B=(jEdruf$nzTw~g!FZVR{tE4aAF8#I*24M+RTcwggz3t&0VqvHMT)JOlK#xvlD z!RUxLQq(+q;Ae=ldg&ezh`%alG(60Cg0|h)QWK5W(Dn+G28MsIvfcMdvD(da1`!T? zEN!f#Jw7h&HkpzUcqi(>n7EKXjNI07l%)9d5Ej2uY2#1K0tJgTm(S6msmfETyK9OT zw89QuMnx8xaU31DaJ3@Sc_M!hqf!`NJt|1H{D6K+O#VGHg~g`_DoK(Dc*&s1Ei;|~ z`qV!j{MP_qIq&=QeEZFfe6>~obyP70WFh~R8rjd}*T=DgG+IF==<*d^N@;|e@PY*= z9iqqoAq*ze;?Z#-a+*52(kzro9^`Y)6-kKp66AJ2lpz!auDqG%_>qG1IK$fkV z#o}(ev62W2|G4+|h%Ezfi3cQKpw?4&GD z(xh<#1y8PzC+H0)c_!*#B8g zooCXbf2A+5hk<~4+1-S+MY0ei?O1QQ^~=83F8ZUSNbp=&Xplr#x&y>FH?d2I3j&_0mnoRQ-;SZ;on(q~$3;*e9o7mmGGmpgcSOa7G(ut7%{q<1Z{h&IVFgla z(TMvDK>x~OF-FetIZt+8xmO#-k+{7-$2eY`bwdUudU zg?~)3fbtnhw;`{n0?>5CM0fg%H9oHYc@z(c`}~Qe*!H0;zN$py7y6y=Nn1xsgEa6# z{s6&7D1jHyLsN8Z2H5vk8lGRS2+IWuz$=h_*z(*MYa5A6=2Q+oINpo-f%dYu>8YNuD|AU5h+y^ewC5RZ`YqyzD zY$5~8Rj*ww!N5K8D4ks70lEJMfIel^o#S`Dh9ueiWk{FO)#KD9U5o9?_-|( zkl?pE!ilPEzvO|J^uWOzVkNpH7;_PU&rj`!`u3w~AxbSM02DVbyFZ84QDCm0@<(Emr)q0e{d3D3e9g87$Blu0Vwd`RPS&T zey=TE=%2ei{KH*+w{6x96C)al`!Afp=$^8s{-XYVsOBnrN6 zE|}sQ{lE2W;PV$F@+#!n8R@us9{mIL#gg8?5{0R(_kCQCB6eauA+*dlZwTUHgRNt# z&y$5jN|N8!Y?+!x;qST;7$IQ4tNMTY8ic2NV3=89wa7|B2KaXUfBRMfzFn+?Z)?`- zQ1BeP9jF#q@(tM}Tr&7#CU?(2=#c=T|J7+^Wq~ahyFt4Zs(RA#qyY4PMjpHsh>Wv* zPe@XtanZov&;yHS`7eu~p@mAK9h0Wc=ia7f3gBrr0!K1TNl6gcQwoL4yssRJOE18C zO3}#|AhL`Q+i@8*T1;qZ>Xi9Uke!Nua2o(Q#(gIZZUFz!f#nvq!PFNx0MGx7hLAge1o-hfbIWQ zQqY8-9s|pp=f$yoP#Ae0n@HkLGmhlF?{VzAsQ7{dJt8D^fq!TM={f#y(yNL=4cVwG zI({nu_Eq%LRek$!MvpM6Ixu}!rSCN~5r8gHgb=4HZWNU%>>v1ie0!fD+Kd!5gSbRPdhSOSr@+Ifk!^SA$&^b0 z%r~z~Gs^~>lTz0`PFG z@v2dWar#Xq1xe*fk24^GJIBpN#G>O1x_ii#EL4wv^=PFH?jPK7km0ftZyh@vSPq5* zN~D}Zr!{P(JvHpiIS61F$-)Iz0%EzI)nX*h6Y!&8wkISGM7S)^Duz@}N-Tx>UyK}f z@zvBxxIcm16rv7sqNIS|Bok7>t1!Q-jn&IETmB3#rNNB!s^8=)r~in3^(EMPv*v-d0RMGnl-j~j?frJ@F-&SlLDGQpi7IkJoVcbZDzLP2ZK zcGNgT?As}z0_#oLGsVvcM3ic=nh*jfc%+xIyH^_Ezo#yXTLIRp0Ey<7DtylN8eqa? z(Ey1F*i2D=TNi04ak`)qCxtM7It|O0F?+oUEuYw(Bgy_F%R=C84r!ak?94JP#dpdK z^_u_PM!FSbSz03XCO8_If*?ej#Y+#L=lqeS+~2&J%8Z0sNXvu@p)FJh`&4xWvTZ5?DfO8VzAww%7Xv;x4$M0y z@Xq_3$EA=NDkNv%@E$t?W;|A5}E)@K$ou-x_`ON5M(Tw zV(LsOiB>|Xq) z$mj68BA3zf=_?qk(*Q)ryZ7xKE|wozZq{VHw6cZ1x|Y0y(A(+l_lI)nX|VKe2`4qQ zOaQX=1R!%0Cx80N7(OlV(1TCKga42`K|n3UboK&ACiJNsYT#4-HWJfDQNoEQsRmPnDKPy?F(%~_2HhNE^`Pw09p zNI6K6UZ;aFbx;Dg6+?}5Qwq9jiuviDgSwTluW$tYISQ7}aZ3VPxIJCtQQyZZC;vAm z70(XBYML2*%L4kh{v~dauy>w@J|)8B60r9}|K9sCYrvY;9-lsFmO!?N1f&4eUMAaIptKw#d!*ElH#mAwuZ}pI4F7ex^ON%@e?grX2D zz(nfWe^1VZCOgZV2gbHxATFQ!1<__F0FjJ3?>)00qLSEXfL@G5T zUI#PdFD##V36j}*YkO9koCsb^0zKH7c>AYl#QRZTu5bU_Tql;m*X=z@k9JcD2vWud za3})s?SOUwa_;RBzV*C3;s*Gpye6HTibT<^UZvVvOVnCQo0mqk*06z03y>M2>)8b2 z&a>v3^jjF9H$n9pF_oC<1tMctSpI7cdg>@UdTIKRb!qF_b{xp@5p{*S+qHQz0NX&m zNxnG$VZg^*tR~ezKw$f=l$=4pu8;`aL@VEb!6RZ?u< zQ3STXZ8DgU9F~CX7yM(M^cBogZ+j&-K?@O+@tRaaTP089hrzyBtR*TE=>1WMta@@h zxh3Imm8i`Gh&F-&7pDp9&q(|o_61N+_wv*Ctjy43n@JZ#a#dkuE7(AjlRqI|>jtDW z6MT601(>(g}%3Ey)dOD;_l$9*5AyU7bklrxcauDmsy1a*(HkNPeVBms{DSJj%6CWnn#VYO+I z$bxQ^pFk)ab{qU3V>?^`)hU*gXu~S0Cq6Oxi!O@n{6S{ zH%)!*-8|yMBY)IHM_39J%IdV+eov@t!1sSQ2lmUrJ|IFv0s2Q#5M$_6_RvFOTo+6V z%0N|yvH-7p1$D;XTUCzQxpRRzJf6CxpsgJCNDv=5_CS0I|)tXIx$Kmsb zmYwtD&p!egB{FQ%=wYe;kRxlkeQ8v_1lRz_T1dij;3n`91tD_${4Op#5Xn~w^goze zFjlt6SV>yKBWTz=L#%0GA9fY*G;O0L)-Bm_dSDyGn{S1fwfG-h^{r z7gPQNMp3v+a2Qt3hEYtNcMzt2JNz5v{&2=1$=OkhRoh^0RZ~wkqKy@j1}Yv4X1a^c zd2E)c$tR~tJ&**fVhhd{Oc1=O{;#FyK&pVy0|s%S3&e$F1|GN!Y5=N7FvZ}*W59>6 z`2rE2tAi685v_=mpNA@SbF&lqYi#BQ0I@)^34%u|f|{M+2=z+ zS%Z(;T6{-@Sk59(e-UUWcy53sy?`9a0D?JD(-@?wW>B^><1UJMBLe38D91K?_!}Y> zoa}*9CUzRxqLA}YDCVT(ohHzpqo5yAirCM0!~%GS+E|EWCK`dgj!q!E=bouiE`Sbp zkotcXiaNzX`3+Ovxa<@HkM*Y0HMk9(NE#ln!9=rm1@hw_TV=M7p%8CuJdGGw^RI{R z2^v#?+z&kqxSNf*v`SG>olvFi$jQGY1F91e9e+<5RHw{%f|yP2)J5~BzP+^*P7ZcI z$-0K(v6T^U3^2cWM1d~)!xgCddWh`WI3@)1=X$0Qtq{L7d|x_#NRg%h`cDgNtQN@g zPeGL^B|M=m5g2$BogzKh82?BC=1=MJBvws3kqtMzJos^^6wE&~d<}K_t+-VL}?&sA%!ij_fbEv0_q1Q)3F#5Cn?(82B+~~ zsHY-Jcs*|_f=PWLrqq2Lh(f|Sp*-i-21hSJ0osW=5c>z8}iiLVSp_Nz@ z;tz&@#~;woD1-VLX(|PQ7dLFK{ngI;j}z4CY54ngA~(TirT|ybuBq zy(AX7o7p1)JfgpH&@?^9hr89Zz@7T*%mv;b^myfL z@BgpbmRSZixJdd2j%^>DCKNjd=gG;85CUID*qSpEh>gt!MylF{C2av;%#DzJCK$W@ zTfr4LmkK?uRuFbJ_fx=-qdX{aFQGL5|Eu;wd+JbDZ_4usdH_=wgybX_3!Yqq#q%8M zBhXSx5MrO;f5$#>7{hkuFvb-#i1bb5e-il#GK}oUyIG~3K=@MBx{+YMZB!4W(7djI z(2s>6%mOxVo$}I!PDm7{x7B4LYB&79{~mftp&ZL8r~sS+rfqKdP2PKA4cL#>vs9l@ z!+oEGca1(EaES^@QQ#zsOtCTug(;|?;n7p1u!wn|DJ(rLP+kI-4rUdP%<4~CWOnHY z+JI2C5hMor?P=h`dTQEbP>3rkauPG|ODNVC{x{v2z_gs zzcA9Wf@4!Das!(ybKa+JuO9~k{uOw z0!j{6(X1ty5iOWKmF8cw!@|LP!?vJ>uOwksO-?e}wXIPRVgWz}wGKdq02@HWRrVK( zM249{PRM|MOdbv%{6~`Fzv^hG9SHij-gnRGK^B{lERxPKqw~SgIRz(5&@K;DgS5;x zg?OBUt*5fL$Pqm~VU*uxDuC`;PZq$rXlqe2?f(1ps4VcUXCR*bBUsC1Q>8X?(<8)E z3l&}oD=H}IAy~@&;MFVMxR$E`c#?X^*x(!iPEfp! z-8_`VRrQ{BC|dU9p>UeLoBZ~ud@X(1?V^V>esWni?bNq`3|>RE$p&3M>zQ{M%Zp=& z9=$z&t>T6%FNTAeXGCrMm+Dy7Jq?Jitk;u?G=(~ERu4QLg}kX}C~ZQmA~2cQ9tCTJ z(mzb@b+$No&hc?#NEPUfduYo&_X8d~YUZ?&&xZ}s;!9REB7pB{H%5L&+7APyY?$^XU^rL>G z>~D6tmhFyaIyYLOKUMy`O?afstE-8|G1^MPBuH%sH5|8nW`I1bXK?aP*kS9MJNjU$ zA&J}VVzr;x3qWL2x8wo$a?qT?r7WNIV5+0i8-ko`=2?jodgFT*krf9`xqTJ7wFTq3A{zhd<+%XL8lQ9v+hR z3D#GEhn3?T-oQI@bw6u7$;k!ay@wTYU5b1lgHw*By)dFsiz zH-k&Z%NwP=yOw{!>-7`DW5?j-vZ$9v8tT_W#`6W*1=FWyUS=#Q4lIst-?(m*FLv^C zxv0jkXG1#ME4wcJNnttpS;_vX=!>waq>QMa)!$DIF*?j^&JT0dVrmvgWY4^L`|Y@P z+pzmI-{SjqXE-D+q;?E?bT9mZSJb!2RDR0359dC7U+9WP!~4Uxt(+f(4gP9BQ704# z@87l7!>gokzyAbW66X;442*0p)&YR*Abo! zH#}Qgn_JOhbs_Rk(}Sw1lzUQP4o@q!>*2ljl01jSvE7UASSDU&KYDJZn|FeEEjvI03&g3)?qxB zDmkdX2T*)o+vnVq`&P9*z(JxN3ex z3$}9g%_XJJ+#0S?`Hm^)S==_zymr(_%uBs#bol#(RV^lsC$Q0oy(!}DttWNHU(Zhm z2%w%6aw3~w9mIL{AbV#Z827vP{+a%c(lMkXjlbxiZ80_^{O5^qu6H0QT!|Ng8U*V? z-%Y3Q8h_v6g!v@6b{p?}RVwUQ3K#PXN8NlE`1A3g`oO62BXPs^l`}8&>$lhb zYWs1{=w#ONwBQ9bg{tWr-v}i%yvyKVoq?B^|pS&)+5*p@Ex^C8%|bPK}$^*8;t{ zPBq?tzE|USm+ytp0q5O=wST_)#J#*~Kb_XcUKHcyjjDR7Y4a5Gz9w9 zF1+=57H-(sk=@9a;q+d{{zI6>(eft>ea?4Y*YnNY&x)I0*p=?YULR232M=!6!PPsl zMr)e9<}(=e4J%J?f!k14WBV0{p98hkGF!exiZ7P}EoEj$_*#s?Zk#t)TJ6k$e7!+* zT-J2>;?wY%$er>`EE*S-f=qN5yI&&|8x{$#LTAjB;hq?mb6_|$*fTCw{-|S&3w<}0 z`iOmZbmLukif5Y|*ge^>V{a$yAF0TH3h8@nzuaq#yAb63nkqZU-$d=1aE$IShejla zQM&b)*h6VSQ>~AH@WW3hUmofRYTCv*kQRa$i*XpH*Q#doa`KQ!|5VaF+uv97a6aCq z>KhCgV-)`TC8Kx3*FUc_{#Y;f4U5jDFU{nhao;t4RsZ1*!yBe_m`3`BS)RV5Rq%8g zo3hWls*hD`H7;IVezTsSbxpG+Cd2%KT!yfmf*s}I_ z&xZ|M$<+=x><>g+ zJ-vTrKM+kp_Jli1bA1?2TkT{>tj4|^%EE~3fM?u)w@)gj1xd91;HVdxWY2y#JG3!e zWng*fr$gzkNmGyIKZDD5i&E(!Q^*>PtDa|5%irp#_$DquHGM1&-?EF`-@LBvg1`*#CNHHz|dpXXNb*= zZm-?334v!-Q*K z^LYnMl?j+f~#sIupGlwwL_bDuHUe-C=PQ*)8s zqKa?w?B1xKRb%RmVfN;D51x+s=UscZY+jnyST0!(QoH%tdCP*;Ip@Zr`!>$ zPKefrQNaDLDrx2^YVoMJVO^=oX>-8h3k-PVfDUCW|H`=BuxW9|T5VEj?XF?b&X{}f z?xLJ6yj$BS3o5+>J^njRod6=B;K**|8Jj!Z(6OiSK)q6*ZOIqgevDit;JL$dpMJ?O zr|$y+9dD5>g6UW?wGQxftPU(%cZTVr%X>J+#`S!%Q)}aT zFy(t=9pW;~cP{>@&lUcrVc8TCpHK}x&<4@#zICT32R`X+XE?LTe_uJcf1KaKem>?B zy=wQfvhPOlb8{uARj1rJPIRhh=$=F9;CWn2nji)hK!}kL&!R>M9c*mXv_w~;d{yy8 z&2q82r$)hz9akH@;0d4&gMYrAsE4PhN?Z)WeziKpRL}KGE`r}F_AMA>b4&VEa_$%3 zW$<}F_%?D_ve)uyR0qPdh=z7!fEea9@iOq@AG5}0%TuzA;@D~g1d}BmUE9M%j`kNm zpMAw!e>nGs#dV87cz9k~^^TX_gv}7u#vrWUa1XLhhtJr54uC7draVz6;YxUm?}msn zCuf=fsAj0IpXH9*8-ULu@|882kE!#c;+$ytW-`}BZw&hpY04#-e%v2khyD6v5yet9 z%YVLe)$$psht#p9%LQ)vKx5Rh0dsV)Mjc_q_wYGSc(+|zlk-GHtDPp(%KN3c&6SwGnd94Z$5PF8Y{QB2`aMT=GThZCtQpg}7q_+a z)h^BY79M=-`C9pDet;)cc3;(~($Ht4@DTZX%FH> z9xj7jj6#7-qx+wCu%t7EcPl-+{F?X=F|p?a(R1In;&3LVg+EmSU?49OxQ`71i9p^b z4=dD>nX9~sg2Tpei(Mx6+EnRuT z?U!g+yOjOp&pP#Qei!l$TJ10BumAY7cKOu&i_0};QN2z-xNWAO*K$G{Q6S5Sd$m~7X^l&+(|$lVnPdo8 z*T|0hRZrN-HoQFBRP97J1m2sm-vT6icipFKsV>giULEtDRBEizH382O{=InK0^0 zlNUDTriXaSFA*YDh-+rMXUElFMMoH4pJ3U$iN#Rk&x}+5B-`Q1pXN+_Zz0%`uG=yB zF(>HsODW6O6Mh2^8u;ziy2k4mWu^G8!g)@(-iY_-WdCml=dTl z!vC+&+L`_qCmct#wW;uTEuSBbi<^8J@tk*XzUjdzt50n1Y=dd&a2dQx8Ko~72D3Klx!4{tr-Z|gvRdE7U_88$m_ zc+AyOJbMcz!7>FusNb;j&imL5-373s)+aZ7-LgEn1svd7BSV#;sH-Xuldb2N&j+m$Kqy7~m1oVIo^JLAQAm%UFsJ{Z1XlFfSE=3|}3 zUJF227nbLsrjLOE)%DeO^q5A|*g43Jg<-!zy5X4Lcn`eEckoI>Abb7E5?j~1PgULp zWxYTyD08HK>#5)OwvWzpm$%B+R?WY`sKkEI{l%*A*S$8K9H2p~DXL6$z>u%`DHknJ z7n74?8!-u!!@f`6mOWQt{@Lxv10ym+0buePww=VU;bYzf?mAb8`pnIZ>)q3XM^7sw zGj;~^@NC~8I9oq81e6E5-qXwVRkR1j8r}QT=&-20{16iri^~gYG)igw(d1;g$ZoEt zE@BVyEN`9s(@iCxPSt~3|I!!#iDX5`9G0do6CCuohk?(8FSGBy9(3B|dnFh*u3cdB zWy+VIYApHry)gYAdnf_YC;N}^ZNaE*mYCaIk9D<;Y)6NceS%vLrXEpVTk#Fpz_mLC z^Cfm-toKpP?CAy3^_T?Hu}_@!GS!mlAxfVJ>D>o!{hH*N*!uQ<--bgG;zCAWTm^wz*E4 zW$#rVH~w>+kgH7Vt8&|YUnKoBo6+Jq=h!}{j3iY^jV&Cy9F%6vq_nLZB^E`1 z$4)F{{-7%^N0jr+T4>A)6>QcM{K@^)`7v?Bx9n#NETi4kK5#cK4u9Tu**@g(D|mib zG~H*h+k}aIwr)4k|LyY`OY^!*Klw2QGxOO_{urCluGkFOg%-=27aGE1OKRzhZ1r08 zx%mxAht?V~R{|wiAd$ZiB*K16to5v! zE%M>eI9MBY3f-u2+6YfDJQq%W6pTAyvgtb{$Asq1Y2$}6PI;e>`oQ&JD)P3+`X0Y4 zQ@JhvdvPXLV?9cLSlS1>`WZ3xG_d*2fuFlMNA57*b^!(C(v_(+`@{_fXH~Ij6uA)E4>NDJ*zFqsA9E&S{&jY z$Bk+~@Zi96v460K+woX1SLb4P-^N(Q_CYSVbtzx0dd^8@X#T^6jk2MyTT{PoJueux zB$d9j?PQ~`_ND+P1||pje^@l8yB|BzeFC5L>!nkT`29@l-h#^|ttnsaYxc$YN6C=~ zIK-pIXNK}(Wv@8ut>MxQsX7~ZcBT2lMeqKA;stde(k_{=1!WmNykm0PF-v&RytqD= zDkUhkXT7&?r1tk~ijIngH}=v->Fl@=vdhrrx}TKtQ2Dpv-xGF|E}18)K5R}(KK?qi z)%L=JYz$}Y?(!hLuDl84ML`+Fp2ecbMrpiPC>-t$aIt10U3$qM0C>IH=+=oYY~bW< zODu%;YKw$I3!=}L!~k|%sUgDnJGu!1JcnzgS6jX zJk`3HEf=P1Atha{6|f9pA6vWX#GSgWVQHTs;biT{s<$`@G2p9;SG)sYUd=E%D!6K)=y@%Mec5tKKM{-Udr+RvG?V1HLvZv8l;hwMnxeF zilRbM%M>CN4WvLzMt#5?z%6Izs8KsW`dg7tQJf?G{yo$>;4a>`?|H(khUng zsg^U3!{}qm13cNAN}-Fb3Q@n(Uuz9dTXAvltBR>O>&*=0qWElaL0SXcpS`7ZfC+~!Tvo{{?2I6oH9>8>eUeimw71r z;B5HoN`0WvJTvVS31FJTRtc5*F@v-J2oVb|u^70-(^>g=h??K8YM&8PmB2wCgyIjH zrhe7J)AP1TV{;opAR!D{O6$FZMf?v)2vQ_xNQy*()t>WL$4+gY*T|z6AeyRl{nl>v z$Mb&G?LEdiKWf8Q(C#~OEIU2=2FEYA0&~lCGJ=|w&A@!@bpvAbYkX=qeB(Z0eD<-dnSG2xa_hfuBH`JJPl@lD*qZyFc*Zo9xvZ$ zHR;ASkvIa3sUnZs!fz>>|3)wbh+u~XncoW7k zSM*-$pEhEth)no?g1FxA&J-7VW`PN$KMh0TOau~VO2kd1e@hh}-ja%qlgSU10iP)E z^@w^&cn%H|)u3Zael2D{MsaoC11>=#eBnk^Z<(*SVB+{qgFRg|LMOS1mSF<;Vf>0H#4*JGHDRw$=Jf7|z~ zpw!(&piunB1c)Kp2rvMhHbN62U;t1ZCGhH8yLICpW;P?82)zs0FI)%k2T(&vTxN?$ zN6)TH*x<7g5~y)b!>zYjRP7wW2}>k7PG+b_|P9%wju4AeclV*I5sE2a{jICfk7l@ zuQt)0;@H^|s4Q1=<}W~Ho2_gYF%Ek+&{fk4NMuCi42*w?T0THHIK$+Qx!4}By1D?{ z(4}*K<^37>Q|y^}Dzap9ci8e1sZSRxNQLdsm;2&w-k&{vxBmS^T?-dx*ah5_33h%) zt{F-1`8|iF0!@B(0=Wg0ZaFc)R98h>J)b$Avo7x5aq32c#KHG_o1W_*-FN?O#+?iA z%@j2AxAe2Wm|=zdFW!!2yz$1zH^sY=tuVVtgEnJQ|J+xRtG#q-54gASO>H zypXwME^sHiEotHa9^OxLxgsHOe!G_49K# za#DNWKUK(CfLWP%#m$g`vZWs@gA}ZVXHFmW;1`D}(V6}rZZ`U&vDvYLu@)Nh%Bs$k zTPx2F2+|s^c^zubhS$^+1wBt^&df#nwZH2K-~_@mhTlC6_0D`Xy=`*$MvbiOr=~jj za#7~#4M&HCw&$EQt)hlN1erJvO?vwR6u#C)757>U^DIhB<)=E zaq!(!EnyM?^@@~C@fj>6V5*N@0}VcZKBmWY)@7YIzez*>3h}{^=i((k;2@A0X{{f# zfk-Ntq}A=$eIdnL=8DhnpG|-9%$NISg%nm#9$t!7B|P5QY?%MDX~OH~OIDV_U$#UK z84(Co!42(^uRAtsc0JmE|3l{9qg>LTIi(NA2Ce))d(E$P3W59=N@S+_uZ%xXF|T^ob)$2EGcUNU zJhl3o`97|n{^z)_Puu(ZkJ{Cd^Z~8mg+G=p`tdU2@auMWOAW7-ZE0_ZAHI57u`Ig! zC9q18AI)cawqCpYMt7dSvB@@>TWPN}nyz+v@7%ikUXk;3t1As}HZ1s{S&#P#eqk%` zp8AFxJfeCTe`WNOL555Fkdh}T)NB!QNdz&g;;0#BnWryC1+c5vZ zoF~c~&)zO^H@B0}bD3veSy>_~GIk67p6Kmbm3MVH2@gQQM$rfJzK9s#Db8Go9k{|2 zl{Jqm$9>jvr=HqNg<5r=Ii9iBxn=3CXU6;cNoQ-^>B&7@i)U}=jelaT`+kS-9F72; znG1d|<1m=ZEqVmqn9}9*XLuEF(Z&qqR=c>7O57dpn)|cGSzXI|+VIfeQ|%dA{r67m z)iz_9CGa+NMV@Vb*86=Mgm(x2x#}F$6xDj~&+lRVcegrv_%yAjwLk2y-F0I6p8L{~ zU1SruRT{3<8$SIjY|ADtl8%2n|IP`~A@Pwvc?q43fAUgOO5Cq}8PwYYJ*v;+Y2|+Z z{m2=om5uv=w4JJ>@@%b}vRXEBdu9THQ+RgmjG8SZsWer1tF>Q8mx^OA96x2R)Yq*p z-uZG`Vsg12N?H#b?Y8eYC6juCyQsSa2}!zRg=}}$z_3W(U{Mp+l@I8K#_-&ylMazX z-c23}cL_$DOdAL~@~3sJHaA|sUV9C{!iJ>|LFDe#=jQX|X83#xSv*VU!zB;y!5DDVB7_)w%UZ)P^F%NQmji%`g4^!eeXcS=642ot zUJAY9PcQG^+Vy-Tz+yh{4mKP~ljW2=)acg)+{&u0NBXMLZd5(FAr-P)B_F7(hju<8 zDf5H$RRXVFzoFH4(^l90{obecBTsCfD7~5;V6Ef&iEUYrn#;YYp1$3fsUP8)Dlxry z)K~W^-S$4S-m~59_s=h{GnQfBq!65xz4qz-c}R5^229_xB*DVl-CW`|@zguM1kKNO zq+m!FEj(&;pPJg|JK8lyy5=4ZSKDRMU;P@$%O+T{;qawJ)v3#IdrT$|A*s=8qfH-u z6gvNVQBca-An6oQ#h~r=)kz^;cQJz>md4)96PfS)H)1YrDQhkPOD!NbqyZf)<*93) zKPlr0kK-WIX<=cnl~^b~CwwZ~8du<=JX3qBw&LP+=;~5YU8o_dy1(V+U0g{1Q+W7}jk;bk29|Lppo=R0Me4!|~WUDpc>aViW(qsbI~j z`M!nY{FfE^114%HV@Wp5G@^j>Y^ei}14fTi~Ll=htq$ zAuZn}X!TmiszNO_WJ>pw%mIXM@lQ{)+*ndfK$Y7>1eXhaJ(xFa(I0phSOq1og|ENA zSfKrNkX(-)PV4v9`TQ9v#G8<2f=kV(UApDEMK@;YWpw5Y)Hgj5)PEWAsKb5%4|yL?kSw(0o2=P!w?lsI z-k|-O?YGj3Rh7KF^|(YseBM;p>4%7uwUrBcd}odDw)h2I7@O5G@dC2Aw1aTZnLgqw zf$p5DGmVlLNt{4RJ)|z-d^xpy#W;zsldLX0Vj+_q6d!dLG_65;BHlk2hCWSp!t%k! zW45*4EsLVLuw%Q94~P7gfOYfH+>N~P`F?E;2^Hh$IL#%$2S0f|4Nch@l;SkKrg+k! zkMqjNahDhQe)C3@c>VAGteq}z&34^ZT%z_t*y@A_cBs96n^;cK-7Ve|x4tJ>tmFG% zE#3LD3qJ4`R-L`)y6`|z2n%(o^k7=4IuG)u5kY%<+siG7YpVkYD6zs&e`pwkZeUB^ zyS6D-{GQ}!_RW{Z#*Vwxm%-Ez2Jeh8_5VI-(W#DGo;e;+j&t-U(0~Avwvy(UKDSQC zCp1)_El_`A#P033A!-GsiyN+tc&6T{4{kWvyQ^vkt?k;08EtA}GaFCoTh>}1<4d^G zatX_SxBbpT2N#D|S1#wgaf$6Du99rD_da3A!1ku+ve8#k8L1!URk09SvlX-??yjIDZ_+#*%%o#Q0!#pl54a|jh0 zkTsUV^I0nYe;J)p86+RiS{LI{Ox$Ff{e`<91Lp+{rpMq-L6#(Ss>66}d%v_s?#Y(F z1?ckmOgJ<}h2%-0W55cw6c2qJ%AK*pK;Wn(eCVj@`SA8neetAVV%54~3)z-m?Y0827t>7_w>4#~2e>S@iAm7I` zp0`%V6;;O;kw4@69vWkqlBSZwPzk@*Av(Ho*&}2z%oaSBW0+%5O|HQh2P#KF#l68I zkmJWH&d#0D2|(Wbj564wKO`rQyiZ&F$by{SH_`!iAN#K_I3iE*^A(}ZyR3%I-r-25 zdf?8t9z>G#ZQV}q{MJ9eyZ#j5S2XRT$#WpNAJf!%l+q0bVYc4vnpwCZ#k+Z)SIV%N zwco^o8D1Sz$+P-#xa>=L7AW0&B9PNRbEI``q4Ws}yejJFJC^mRDOpdm+Mk81!T!f- zb>d5jZkqy=lv!=*85CGgtI2x0fI`ts&>jIQ8ZAzMr+`~Yk!jRuCe;m{thQj$a1hi zWK-=2!?lt^lr0)K%&K!NG{p(;q(-r#;?oZwc~8#GU7U6M8QXW?TXQnya!}oA#OrsP z0;X@;-hNZuC?|B1xcM4Q9uZ^qTdz4F(AT8WB^m%O#9$xtfTF!Ggc#7zeFPaTp zs83xLcBI(!up~V5ZIPml_dlQ8;MuwMbXUx=$=pblKTYV^^?LaLZ_oPWdtV4Diy!dT z1M2v2-Z%G@_NqC#vy_I48_Qo_UX)h-I%LXTz3tuW$xyISjj>UUWs=8n<-d}lT%}h4 z%SItXQlI5ZI)u=dnJoKXu90eQd8Z*=pnOhTq?* z_nkeML()o+P3KFL*PeHK9(FD~HW-n<-8DzeOR^TwqVpq)TFqbCW)9}(e+f;h?#X9U zyvNoQE1=>&iuI>S*|0Qhm!Tt-XdAHQ%%E$eFu(Db+1^E7%&<9o?wK9W65FhG&ia;< zY=HWQ0~PhDKRrt=K4$J$mq+&fSdUN{A zg)fc+Su2%xG9^Dq-zOW``n0^YuSK7TA4v9;oT8DN+Ao6Qo13*qw-Ngrk?tHwj#0v* z=|pnOMMy0=7Y&$1`bt*1!DR3le|RR{BCKlpnfFFILk_W6)a zv-K|qN++GIZfriHYaAuGuBv^>YtbYGQAQbeNzD){y@^Dy42)%Zn|(Tl17Rg!Ly@H2 zp%2)-4PF$*GhnN*BAn8w0af)lI2K*^OdX6!yi?{7>@`|DDs)Pdg}?sx%e?NzONfujl- z47+kxc4AQ>N_iVdAq~FpFgRi{XqWJ?A*8WYoAzG z;gNf3llD)fPBq`V74qdYl;Y8ETcdNYU@XT{Q(}0<>SBhuMfemPeZ~wgBusLQHR-xj z>%i%R#+#VeO<1c7B{!KQMw%@9W`N5jU&@Iq2l9M6B^$*_;_Wok_@exByE=lo>b|Le zMBxc);%@gYszI1?endfQsa*dscCL2&!BizJHTm4%yMZgceD^s4#oxtF(9^gxQN&!? zyLHt}ZbHBDcv+q@ZTeSbttz=pd1{C|&N$xAcv}+nM(t>jbaIG#t>en9PobLgd2B;= zpZ6~qTCPGF;p!L|;eQNE2*dK)waWnqU>%s|a#bZ_2fvbGZAx0j7(Ip+F@vmgISh)O z&669VMK@SKV8NO~HNQQJvh4l!MO*t(xBmiOxZ9iFdA8p!+!3Ut{#7Fze1NnYJVi=c zH|4&~%$k?-XyFxjQZ+udk)rH@`tRUi85e8^LLLsmBkPh3v4J3cxoo)1gsXFYxxeg; zFUXMD_m`wxzi5_}<>Sp)OL2xCcBFRW9sC5zLTsabHBRv~_UKsn)O0PNnTi~7^*x~n zkCO!LuOHEWUOK0e%oNEAH!THp@3;q^*=Z;J?t4BnC7zqYsy<4~tK^^|bQ39F|U=MxY3f;)j(U zw{&=r*Jy=I1>?JuJmrX))Cv{{$y9o~U&xy8#o+v@5G@*SZeF+*O&%Z`0fP4G&zE5xn+@NL z7BGuBi3ysj{N zCUKWhlSbonAb=U>&7}wR-In_=2wgtAhk?Y4wEe8P~C$xnT4j1 z28W*c&5X+eC-H(tmvCony<^A&I2)`OCO$Y`(@=~jsBOVEbnqOJC14zob402sb|c4n zYiqn=`f<54PZ#+R=i>pBi=(>7&DBcg< z)4LPKfU+^H=szeu8_0r+61fm5EzPX;NJgm}SC}*RaVegPok@kvd?^gzSZmE_rE`qB z2fHMdk}r&L!T|J=PxkQYp@D1Vn#zoGGx`LX1%1E7=85Xdnfr%@brrU96@Qr|Rq-zY zy9^>c{|KBJDgjSt8C=B+oO`Axb6|+l?8*E;7-}K&??C21ZN82gJ0>u|_uCo9fg#nh z#Yna6quS5|N+}osH2~27>^_yg8tLO= zMIPl5G#RvV;cnY_bxxAgOQl5l{`DmMAn&kNtP=#FTcmr-I)Et$f{1zraf0bJ%xN!( z6KrW+@nQj-;G^@DMm??)5j@y87@cNM;>SDaO5<_52b{t1H!rAB=0FeOxid;l>~zE- z;tQ$w8*QKk<`8205fBcl>=+W2WULNmPMw?+y((1^%m{-5P2dRL2J=`9MMljI8>W$O zVm#6C?>RY0rm(Pbl{e!C@BR-B5ff*GPM?$Zygp0-+<*1 zrgSs@<#y%|M*+oWsZwRUI1LVr_b0|q*Tct&HXNZ2|HNvRrktJnb`B;|uWa(DWjh0X zNx8)r0-*PXi9Bw6V|0WN+#;5B@*F*~^!ULO4X`sP z{=2Yv8jug0Z$+puELjCtaBf$pkWz({`@ld?jrgK$%_EBlN#QGNoRy)1wUI{3?IItMiE|D#%%3pe?PE?Dfy8-yK$)Qs@Xx81Xdlg z4K^t>u67+mjKmWGh|L+n8P9Qo^&sHQcq%`}eZOmYV$2A2QfPZwsvAB!^sj>yfLw@% zm^~lsre0>Qn#=|5CJyrTeQmdnp%qERYdO4Stf}_OC%JJs zE6}R!BhZM`+fpnd-KbmOFt-hHEB`r)DH^(vMu`6y=$UiA!}aovqCpT#xRd8{GyZTe zsi`&>X3;G+YfSlbGpw?1#=hd^iZ};0&ovpMng&^DV)XXmeB4*90M$Lz7KPB7|HLb; zJm+i109q?B6h4K3YpM|BN-Q50aF4NL8OYhdQAIZq&!iiNBH7lTp99NuPBC(YsdVFN zEB~a(ka8m?7C?(RE@+-Yp<_$kz0FY3)xGCj%mCN(&T7mZkn-%OACnQT`@j3YIHa^d zYCtVdM`!DKU%?~ek0^>Lok0;L4gIYU3Tyf(#t?WaBV#*KQA1bb>s7*PFTN@{9Wqsw z-K=7kM=;5K+cX83hK-hi;Z+T_EhiEnz0M9#7<*o>1UpAw0+!xPMEEEP_lJdp`C zqKYJ)%H3>vm3jMeRzhUTXwm&bHE6ph-n*`%OgDQm9GIhBG!MEijq;}azUp)TdZL@i zM57$K@q8%TFTKac*amzeZVeBcCN5k?+H}0`+wMB=h4W@GtyKt` zaJiMVv>8^bBrcw5Jw4EkI!l`$)KkJ-?)$&h?t zTdCzU;w;F{0qD=*9xtOF7HSq9=_)xTNjHWsWg3Te8Kx5%uG<&XQVOs>{?9wc>_j+L zdB?;>?+b8}#{L2IHIH;%WQaD#oW&ssQ4Csf0lL3Ntg-WJvHKcAH6o}>f=nYAU@LV@ z)@{c@s)evT131Jn)vdvcfB7&Kp8Ne~H}@IdpJw_4TQ~xB!*8w6Kb1Sz@x-VnCz=go zT(-e_0z+$tNGG9j=udp}%1Q{EOH~cz!nU1jq_|3n(9BM5_@vIWq&(|Eb_;`K4hWVd zZl(y7Xobk7g3}Dfb`wGyrq`0_dI`!IzU?VOZ%kmiGfmz;;zko9bJIra%h8Ce=P9Tr z9CJf`B*vA}J@E#_d&u<~KgZ`~tyYe1R{qviaQ3FCK2%+a-3klD zkWGR!|IKUhBO1Nl-0f*@BQEya^c0Q*EF#7qJ{x9Ip0Oq2h7)bfkHO=9hK7e z^PKp8;7)a2IhhwIpSg`>Os*xMpSL9urZz0HLj`soblXO@@>#5GP+-mXo19Qj%}-#k z^}`eza}VCU!YFL;0E@C;#VITXd-bn1i#|KEUujL=TSfFs>EEG&x%gU{lssE#ew%Z1 zDML-C)fC&~Xmba>%^?xDk3x}m^2tACCNSo- zbro9CD(j$^gXF+)B>Bs}h2Xi$dAAeqtvzTZyGW9RKJ<-1jK1FOjR&8i8uo|UNJecP zE-0nl9Ojh95c(mEnnF`NW;Kx3j>#= zAP6h4&`gqkk*uF+?xPf%31>C~LcYdZu~U^VOP=Eoms}Zs_Q^UvMr~|IQjz|L^P7yR zTq5qbKejO3$uK2MaZN0x9m&F4!E^ujCfE2i5%7h`^)rv(EjDF>@g@2MV{j0A5tFeh zu!Ede*zT{q3|^(<84Qcu3)E8Y#5!qiX?vw-l%Kd?W^vPVtkZ2i#v-8&0@2(i<_;KY zmD0CEjJ4a2-LhwweJKS~icrMF0D@5`g?=D@$TEf{xIIn~ku!`K*@cPU@Lo6{?}d!L z*o-^XBI&J{TeO2YGk#4Z|L`=Ybgj^Q=dp)v z(F9CxVdX(e-Vf-}c30DR4Zjcbh-Ty!JA`k^KcyubMsiAw!?cofrOw3_+h8H>*SAs5 z7ye0^T!bcc6PRd)3+(9rwYS#7%+A=k7NU!%wESj?%^E#NaX`U}8;8!BG}}DPu7;PY zAdZ~XI?T)QK>I}UU1j2!+_zzGR>MaZg^%9CxuksixWZh4SYx5ZX;1TDWpA-FMo8-; zmd&+;lj^PYBUm=I@_%L7Cinpf{A{Hh#9xh!HhM9e@Iey;UmO-te^)-~+uj8)!G#oI zKMy6s4DKc6)g_(Qqtl@Tp+o0wE9PAZCHk}3%t$a8C#dcmea8B*=ij=1JJ|=Xw(3)d z92bTnYcOXO4Fp$@Q;j|5z0ommQR#)HBysnh{+8V&Sk3* zR&Jx6)l-SaPSKLROo>Q@|VdxxBa2+ZNFb}P9v-3$qkg;eybHa z_q(%$Ugcm&2u9EJ)log7R;Z@G;3WXMtD}=e%7GDHVtV~?%07X*msfcT?Mx$QpjCg_N$|V z*3V{ce->7M>^ttg8AJx!?h#da?CPL7J zt4O?h?tsF02p7SVJ+;DwoV%#>$UYlX-I#s=J`~mN}IqOYo zGmNd$g1cWC(GHRxq)LKev#6tT!5X%n8YZCI;3itR@V$+G32c2*d zy)@|Km00uBSC^ERdOwg!x8Nv1b0@DaQga6(koGOn)MZ6axTZblAB`5wi1y^3_5ML` zZeKminCcJk+#YIXs8i_$$57bE{iqt(jQj9H`^~K52w?>0M_n6T7fL4nwE z{r+JZnqUg!i{61zfvL{XfSqpJ-FGD%rw?A|C>+Pn1bK<;TQ)Vt9&omgXN&_2dz-$2 z4D!&0k(SV{r?J15zY0C)i_f#JS-6_H(fv$d9M$q3LC_oN2Fl^YCv|x4ce&8TLj8uY z%i9+peBy~{ux+nL7&~qqLAi6%%-`iqhvS&|Kt^~zm1bi&9m0pB<^GZlZsFH=G=(VL-m5Z!?XMK1V{r=&zbLZ>O zKLc+^eOWlxchO{kjP}4@|DcF1JhPe3@-(6Di8~EdI;mC6Z^RlOo8ETqGFmHzrY_%# z{8L9cTPM>k6)|zv+}J5i3#K&1HZke|ku^HTiJ<#r)Dfx`nh#nTc~Fx)E@`vcN-({U zc?+VXGL*)RCM17OC|DzNtc|Ias?1L8EF;%fb{KZ{)dIh_tni&O)90EP;S48+_cNE2 zmA#GVM?O8=Yj`}UZSg_o3ex_!9=c!g#9)UL{0`~Hb&tfA5Oj=t3Gk&oMT(?t`-Flu z;{whYXLKf6_^mdzal*GFBusKlrhQ`7aq(A2u)CyuF@jyHf+FcsnG5Ent$l&HQO>>D z%z?I%t-HzPDLst&BMdiN;9t7gQU(ukmAQW=yo@&HF;-S2QzFxKpZX4aOv!xHybt?D zl$}wh;ttuoEQ)%gCytlA!@h|K=y~MvL*bA*1cI}}_&$pMR`H47} z6;8xwqUr+EjoVkaITuXlp#{Z0X1O7N+|dGZ!IVW3AuUk#D)LFaigqs2!f3h)sV*Eq zKFbA7v5)!s=z$JwS}JLQk))w*#|f%KbSI#k%JI)c2xmO3NTp9H5-C|}vh+hslM=!# zY`Gb>mom+f#5AMwj+hdz8PtFI$2GH#=+GYea6Bk5vW6Z7r>~{1cdrE7xm%DOa-l=` z|NTMc4_}Ss;-5DN(Os|0F#F<8`ofT0rUhcF%bS#j|G-UyM~(zR7{$L;FHGMWvq3uU zXtXBMq#u1IBruO;I1N$y3yRjo`R}51oL*G|z8A9KEm*UrD4j4xNN?g7AUt3I6X3(- z_btwW(xetFqxU8=cX*DuY-`llj9Ru$SlUtx7?n+mN6s@HC|vZpS!?KCQyZM%+YRrh zwj(E*^t8rM00D+Ry+jdn@HY@y`@|d%`fXd@mE62&YJ6&O6OIfdc{f|X|Tw0j`Kw^t>j(o6E?v%`uL{d zz$k5WI6<$Z9%f^*0J*Ah#|zw6SFJ`YCn&l&v=HIzVHRSglSKoY+p{J~#2AcVL z{Fjz7Cf5ihFRtVKUmATHM%2)tWXhggXTozp-foV0W!v@L`=}4XrCABjMC9Wey54!% zEF@&OKUrVBe@ZL7VU8D%C}f-~rr{pkGL~A!C!Z}S^C#u%fA_^OHy#SeAz-;y(n1|< zJwZhMqzJVldXzec6xR?u@?A!iu~1|JE4b5_qBE*4Kh>Cy%YjYs`CeUkf1H?%7#3_^ zKDoc7jet5b;!57MWg6CdEX_xL@_RTC&hE4ehQKsO4mj zP~(Ek&FIe<%J~J?#!M3bGo`WwGL&zryO<`a^$M1mPGiH(#)k7e-GF^n;%dY^0wz}z zW%dRu@h&6!y0A@nT-H*-}_$k)-!cOJ4zO+uHpL1r70P2lo-LcIrq& zl>7R6xSqzW{QbH0J2y~a0`p6@+jlZY^g#$CdYG|zKN*o|6g-UX4#@{*m{{i{eNku7 z#LA+DM$B`>!YacZPxg$$UY1rSYP4(ztipD4^DR{7?my^8AS8Q88)g>ts~79_KkbGP zXFNZeNQ&vsW}fEstyo`Sk?Kk^G$c!K4~IFB&_-Lc_J$4ygk(xm4Qy2Dozvya1A}9< zD&{#kMj9r*w&YZh(^w0`aPT3cR?e9KEVDSIbcr}uQE5l{W~5Xu3cpQfP9c!r(;Rm^ zTX#Rvda4@JKJh<8K0lM6`kjC8K5|yhDT^l?MQ_bPQO0D`cu3}fsj4iFY-Zh{fS7eO zdU2$n<~GR@cP8Hy8ykt}>KHCEk`)S$V-UOljrT`$XNjxnsfp~{SYXH*MsnZ*v07t; zN#9@{PMYz!EtAIdlz3M)o}3-pD(TVtXGZZ>iSuNVP&*p_*Srhg>X#++h`PUi6D)`* zPWO>-!00;;EoPRV&GF#lK?({9qk>Hg+dv)G)(zP93)D3;=;H27sYS8~&775DWRu)* z@MQLaxjfVYFT;xBo(`jz(Ny&w+oD1|l~c*yxPLNxbTybpfNYHg_+;+8j!Lp8)J>|c zj14_(o`F4C$MVNg*vLh4B1zfP`$-@=vnq?$-+ z)7TisCh5PK+$wHf0Ww9I>YikZ{3<6B2QglY>(lrdqwT_Y&9>+YiW%WSo$1fK~QeqcUe~jZSe+*=$tE zJilWd$E0A;b}Ddh@E`ec|Db%57KLL{E>*yIfcg83S#QO{fS~Z6B!ATI%;6ZDW~rrq zH_2o@zfj>jwrpcfGrmgn=Tl`{o!2}1I|d89@OC-Yejv}rX$#H3v0e4+=QBXg9{Lk~ z%p%P!dc=$6qAg@4;J>XtBYlN^)m9DDe-M7#o(Ihr?*c)4ap7qo!xf2sqV13dgWP1P z`-cI-a1lAw6J#ZE=Rr$^O2vEU|DR!`^8TxMZbvDqIn$iG(+z)r1fb3P( zK1)mTYw>hf-WQ5#8}mNY(sR`ZHd&|oQZV9nA`@t)e9lETQlsEXJuBu|MhYkVD`ADEblc)Nzq&U zK@~_rZEJ@4oHi|7!Ft-mA&{K~EWe+-G_3=8kT^U}tfS#U7eO#H8^JxgP5tg*QpV-b z`NOw1=Ce)VCHUE=zfz+YC}FMJy8)lfs@C^btHmyz3AFLjUi$RVR`}s3+E1k1-V2*K zUgQ6xwd?II%LRn4bszfk1*GHC2pm;xGdbJuyxmX(>!5B+d?EqU-x`zBG??+ zTiy4Qv>55xOQmiwECYa#m3TCF`lqizlC5%ZT;gl0VRM`u`Br!X=DL;|_Qt>I0>Dnw;c*_!I#?y8!bKfH{ zU4!1uH~DaddEy0meu1nxnTY2Uqx3=HWFA7CPLS~e!fTax;7ac`wa+dNR5?O&YF&S2 z9GDOazrJ6};|1v0wf~EgzL(inhs*W_0Jj$c;wYkyJ=_h-ghp_#l~!lIj@o^qDowF@9)pnTkFwbA1D?A-7z+9}s3I^K6j{5iPS-CK zKy|&KJJ9seEUzO;A&|9Tg*;cgwiu?tPIp{UboO)Ls_Bj^<&$V`O_WfG3cu&m%RrN- zEl{f7Y8xVddIW^!MNe)Bh3xK8o4*)n*{~EHukX)g2?UQnbZJX?$A;(2>cI2(?$&nm zMgZY-=>On9JvHxSi0#9k;_3T4e}vswCer~q=qAEp-s*hw`o7LMIR+D&;)(($)@Ie*L?c~tei&^VeBT~&NTaWM0Jh< z?xsZKMsAZAuQ9m}Y=V3P?8E>@>YChg#atVElhO&txT$1b%9Y;dD|()E%Q0kGbmZ^xmp0F(YQPqB?mqzeE-Wn{j0@clxZ?HL5F z&neb-eJ3RsEw_pc40vR2LhI?T%g*)OY2zE>ykdddWil* z)UFPc)6h<@?2CgD8H4Qcg9l>$MJ$(i-L4}t2mH~`_^k#O4sECy z)V)N&wl5bqc!#7!vQVjtVRn$70MNARv-AhNUyLnhb{&dwL|H8U)c!o7Ziq{iNmIC- z+^&@Mej!iuozxNqg3#S`A4tf;+gbGyIT}hw6GU^AQlHJLc}>`6n-;ImlQr0i3BX$- zGoLh$3$DRVJeiG$q*&rKen~RfK4+q6-a}&-O_e^DLw`nA$MRw;L73od5c zE>nm|4yRV%)64jPvo4)mEc437j_9SGTYmd7J>StLEDGXK^$pP{A$3rhKWDMn*1*Fq zs5<%3hloA%#Fd9ex_qMjeTXt@Z?+}w-$H_EmXDS@1Xj2|e1_dG-PCdJ6w9Vtt#_Vn zkbn#Da>S6X8E%2E-^G@VS$`&OlbL4`O``pJv#&FwdUB7k;_Nz;?Jh#_DY|-pAFa?^ z?BDlxyX#9qpB;xBA!%{-KnFDCty0gtArv;t$U*K*VrFe1Eqis}D+n6+^nboqKKzYT zRC#B6`s%Jc;zQW@3z@WzEDH$H@sZc4^1A~j_g!2R{Sn}J&n4V&N$2GfR2Dv4UfDQb zE|5rOazpK|y}0p-!8cq1S~6S9ealhu5eMJuKRV^0p;vuxOPE4v$9fb@n&ptvYU#4l`XV8jw$vIn?WRr>UIy4T|#& z2r8UV_UiAr`Q7f0KJ63og2_1fqM@_97rufT@y%?0HeY)E`w_TQSNB;y(0ohBLr>!m zpN|W?mVn_w$NgZL($i<#W9-zUpUX-}?-ge!G}3s%6=e{m}o+xfZwq(03Ie37T=+CcEP zcrM@Ys|L{ice@4(L~@nQ!Pb3nI*wl7&G9cMKN|-&Z7A_yCe0r_oZQhhA1o=+i zZ}Y5vmZ(@9A!3n4oPD1lVVl`nZqfMj>&J;hdNXJ8zxr@3fPF63D^(FGQ2b{M3kKsx zudjB5Y3^!g^XL8A4qD+#@8);U`U4g}l73R1vJ4<_NN>XcPbNaCVS=aim5JCyafgW2 z17WKi6T6@kdr~yV^Wl@~J6F8FKex>u1h-lIV{)~Q>xe}1v-|0E2$wz2XLQ?k2CS#8 zrJ?vS%^ohlo`|L8SfK;Gx1VOHtSBAsKcEURyps<-a zC||?B=b%CW0(Ks!ZC57eN2Y2^fbVzY;WxG7zTsPN=rK?6%g%(zJta}u26jREn>65P zl1|?*jZoXQQAFisda?K}JN@nAH7Fdk^Eg;;^1HLYCe)l4{n+3_ZU30(tcTfj1aqQ2XEXi@fkap%M7)T^HhRzxr%KB&>rt~(Q*VWS^sXI;K`lJE4D#DwJTKIc z`|IPb&4If`EDtq>SKoT#yO~!Zq=D2*x%-CcCGK+!)V|tn>8U$=;*r#X}RMdx7_Le_38ZS zZ*VQ@j}nQFLK0&XjJ)&9w?m)ba`@3-hDI$U3#&0MRbwWE31b^(C;B5Dr)W6kUE$A< zfFC)96rZ@m<-_FS-^pAft^`(66PWiS#~(u3oQv{Dh%9NmMX_%MA-t*A*q!fquG90v zoE|jdjGXl~lyH62BDw#iO3t+f>18XMVDBa`+WAUi(bZWjafngWLEK-SGXnR&tQ%@* z_m%Z|wJi`m=8V+$-8txwz~tWlXkj+_kt!w<7?AB;5v%9t4r+p3PHHSgW*z?KfgA00 z_zkt-h8_@TrVys>g>3nM5ISf9;u^p33xLLit#1{Eu1cSQeMEoabL>t;C{>LV3Bq~40@<NPjuRqQ-h2C3=a$!9-Wet4fjg)EGzh@L+w9KrN82q)Ez0j9L~Ey~TDdE{ z%^Km&)(qeCHMCi&xsl`|PW{)Egn>W0E|fIHbjH%RSeGGa+4n=w)pN^~FY@|_bq7%a zU}@M5_7SK_yv+5@0inErsJFX;ce8P8`r!4pDpM;-6U)QR9>P_(sO(lb>j+qD$C)ge z2aB`77u9%wphC|ElJccTg;Vl+gs#P9J&)48U%)-@IJu2;5{SUi$rLH2?Awr7u#LJ^ z>8icvbn+2qeX^@Mc+VfLPlnt_3^$CIf>9sy&UCzyAHxzy1O#l_Z)eRowjZ|j#03*G zr&eotcKiF&V@b|KTI;K0Z)S}{q(2G_@Er-}paQ)Qf~||Kp}0CCe+yz*(j3lfGY5xW z4`!F$Uk)D4{)yhQAPLLjI^_SP*!!bF`zR%^KUxd^!V~-ykUmQ?sA9h4#av}Q#t|G_0w;LJ z$1OrSnWVgySICC*nS%%XQIug3cB5pD%~c{7MZ|2+2RKo+70X!sRSPMb5yDIgM2|x?+7jL1-}Kcg5lcf(c+aWThqF#a?KlJnFn~}iT}5iA zhMORhXEuwENw#M84bhP&e2ca|4^>9a3ndoq*@G#`5M#U9b@(5VTnJb!4z<5=)q|TH zJI_x;ME;66O2r`+zSO?`K!ykgb*AEzAF!7>liZ(7U3 zP(bZQmY26s9jRp29og_}t`~Rt^!n}qezuisGD+Y^SXr%|0(s^i_PtOFI&e*ix=KX1 zk|QpWzgFw}S+~|}C;E#QLb3|KCS)xka{thd+1XL zvYx~kPH#6V4;D3mDDhb(4zCpGUzn#Xnf~XebpKw^(2w(}zl%|om7IE#T)!;;{`%FN zSi&i~=JKLfOJ);^3EMM>Q8L^h{c=L>?N-~Bj|Q)QA$FQnehFAR^GJGi3(>S+(s8%2 zrIsk|>o0+vu9x;{eM_R!yVCn(LF;hmr-Q$c;kp!}q~+vf7i23wa(%aE@-7pyWdMy_ zh>WHs)?2@YyZ4E#*m5+G^CGYyrs)zRLC@TP_?|%^Y|d6O@?RiBn8w`lHC$Fd2tIstkt|;h zSa_33;fwyA*ZhS`a-RjKXauFqdpOcx_msDfR3{MrbqX_|2ghJlO?w5>2-zxmX2>KL zkg5j!e;y*A?4Ylg-IY6BK*BV)Dq~ZB_!5h~Vhf+C6F-|M0TBNA;mhfC8s&AaFd^kY zHu#)5L_=2f>$?Q62bH)uqbIsQbwxVcW$Uin83mB;e{}bWzy`bOmchU`hb(fy9DJ&J z+VA6_bLF9stf_K_V*q>PX9UD{=qIE3a)ZXq4s z6y^2pvYUAYWWG01w>_pqV{UNd!E-%a>v}FElA;dv7U{DM!1^QUKo5XAG<1SIJ(&D_ZnG8}Sw;rcDaYrQP6cuK9TW1hIJCuuiT8vLYc}pO7BqLTLn_ zet(C0C9Mocgl4b%wi>(u8{+yRt4;-^7fRnR6?70*(V;RCPr)jOa@FsS@ zDoM`Q%-izu(gl^>)BZde{)k|bN-xF761xBH^m3ge+ulRjy$M2c&n>IJL*7m0$&Dq; zphZ+8D7xbhDO*xN80QlMvF%TH2rW>jPxx-XR5qQxT9CYOCxuZpt1F2s9(u2nwi8zvYz=~RhvW^}Zx>yYr2>*E&pA;jG4hHazpV7M)Y_|e5t+6lszP)KCd zUxW6oO~})|)G>l4?>}K%7}Pk}7Ga_)|BZ=y7_axQnA{{eQ#^E=>*uHuVEWPw<#W3NokLPmA6Aam}O% z3u;pGeSv9UW3A1(hwma(-ztFltg`NpqJQxC6o9o@+_w87)6#)Q)%D>?lBgh=;B*U& zqf_YHw_kXn+3vMyrYz_(>-nEa^UwU&7k%PL^Sc$JUkzV&*XTyUyXrTuDI4mk>5O@S zZdlwo`tZwq>&kCK5f&`KL$Gd*U?yzHhQcjD5mT9qDFY0t1;FRj-)WNpD3VwgF-IQO z4>100ZZt413c<2t`!I+?_{9GZ2MEv&+Ktp@K%22e&Jj41n9)IMQlw4*&H-ZVIb>&a zQ+IP%5N6r9_zl|^QPL<-1C^hC+Kz(Am66bRyQ7RY0muIyZMcZ~!vf)Q{a+i3YfhN3&fRE+QX3$~=-K znx&MV0Ho`-#D|7FKEVz-Y6Yuz|uCGtSCB0+mj9!xUto{YM1IG14O^ zNy>3BSB*UYIG6z@#5 zD9H6QzuQQuU;O%W*uLv8mc}F#4$t4|-$)NoO6HX5C=>wmZR__?iV$Q0JY46#zPjZH zS>imZt2?+sh7ierGoHKt9_>E3sXA)%X3lzCHH1WCj0Q%@ zY;;q_9(-$bFn%_m>?~J2{FH*`&j%I5^xdgi}zBDy~S2sUV{(^Go zHFA6Au+!5<)b@?JfZcyvF?l@`MsbeZo~IUy&7rQXjWB!m6VuCD zRD`*11#V$~1uSJaV6AK_K$~SJuf>j9&F9cSoh2*6r99YU$*0GbNy^N(XT- zXQ~aSZ-!5B!Cpi~4 zpxP@S%DkZ;Mx;UaH7pvk&JrPb`gaD4MGRh1Ft1oRQcILO;U-hWx zFkxqA&qwP zpnD2z*bS3$x^;-81h%^A4J^jh+R!nnNysoE{v|R@7#C>)Wcu8L1c*+$RyGl*e7@0x z$o)Y5JA5ehWCiMI1Y61NFD-(tNU7;i*N#rpHi!eFuGK{aVR_m2ni(ul3PuG~h- zt0+x7IS~DOzjl(ppsEZjySP*%&nu2;k16ytG?(t=U7dG$=UheW%FCN0TMwW=6=CiV3(=wZs|;sSzY!A-!>O|1b_GVHd&5M>>>g&cD1RI{_{^<* zYyEKiYi2jcC@ooPUL78*xPVKlRzM0?l1P= zAh2g$ps&z>z%E&Vt@`mM z$OgF-K1PVYnsP3$V! z*=A5%Ik(*luV9|F?#U#u!tIZ_5>Z)e7rHpUGs8=DUB`%)B#@xLcgroN5(gf64t=e(d~H8lQZ9 zXDaCfI^q|>)BXNo%cKdVYsn81SNs0X67cXa`)dN1j*L_q&tmo^_j zcp$9O7R3D7;OzYUu0A`~Z=go@ern7gcUhB4)kw6lH>IXH{BSBuYPvS-mDbU|(!V`_ zE{j0$?m!NAs{iC?Z8De@&!Wh;v&h81lcxRIwN@f=&2M~Tvula4q!4Dsu`yAdIRp5u zi#%tbBH)`igZZ<@dzh8wljrNLA|twMf?268Gq^Wn;yUHFsb{K2rlJjfMa(pFIVZ5H zV`lBPs6q{m@w3J|`2a45+);XZdZue}sAOc4^oL~wGbAKbKNW{8TIBF@UB(pQmDCjG z@5J&)$8D%f#S(`Z2RG*kwYKQI*v>0HZr&%bJ|LT8y3BO?hww?~1?%}q_(Ey!#?F@^R+j(hK^mep}4 zuXE~Tg#GJmdSi!XY)_XUzY>%4zu0^4aIE{kf4p3jk?fsOD0_99Wt5y`?~JU<9u*<7 z%gTz3oC-PZ9g;24GU7BtB1NZF_R9D@-_&(opZh+J`+NWK`{(;Re*bjnOz-pce(v>p zJdx)e2G-v^!a)h-YiU9rvriZJNeO78e~)kY!X6vok7>a!2io8&l}C0cmrqsrv>f&1 zE6bks9L|>*>vKQ9r7PNd5xjuHZX1=Z=fP~1f?re$zc@h#r5OJq9(*w|d~{#iI_1xI zDby3mL}heE9KkZIWW24}`?!IuYj=~)YA6X_R40y7z9b5y{W*y^xzeBGz=QPRL8=or zwtqj!nJ^Bm^NzF$uVnn%h;8pJqY0K9-6Lh3kAi%T8!3#*BIlv~AJ21uz$+%>C0^;6 z-u{ER;4CLup51E%T5YhHnk4E35Mm=O1@?LUn`U{0Dk7?{Th9ME%K-v2;&0;j39t+JC5FKqh&R<+ntk3(5#jxNOrK?1 zAcvvTGU>SsgpnP4#`iDQ5m(T0G=<&$>yIQTq^ks)E9+P;XPm$`SMJ#+A2`d3cv(fk zL-_B<*oo^-A%aTq?}z+1fl}b0#ka;lGD{B?w#tT8?Y~ zt%1O-z!UHh-}waRJ;Sy4tQMjgR;55(gjDoCGx6qNTko9dzoiF{`OkB^1Rm_ug>j!Q>7L+jK9dgyoq<=crr~!mDFrGM{Z;Pu_cw!qLeyX5 zFRk*weFZ6D_ugUX2t&>G-t#s;9LCws-e`|W)c-psr6(v1>CiIPhkZS~cj~|`utiJd z*VH@^OYZ5DYZeZtN(k#H-ly9hnR8>|fa0VQ8h@*o0Hq2uLS6x4YJ(` zU~E2d>c~HgQJy@;JE*d&^-5>zc<#v2DSt-> zSLVnA1cHS3r~NOAjNefFKb!84{OJrS5J7O}z57h5a~k3;Z3W@8r!+7ZGIWAP=iLXE z8ynDx9><~1K;X5~2tf}OdF7X2ppW^zSoA*@io@~=eAXX)k91!GwZj6;osJ$kLWre* z&gY%W9z8~-Y5h`#E%;tg)QJEEB7!fyF`K8iN_vQ0uD*>X{}BH8Fh6mkZEfc6iaL?A^O}n{LGxQHvzlq5QViEjLl51jZmR<#x{m2#9+%(sL0;)UaNxcu@ooJ$CGW z(9c=QE;YB6l@2B0ww3P343i!$;U>UtRs!$AJTA_tReE7Bjo~ZG{CP@DAmG+KHtl$h z0WrsehYAG@PrjhyBG(gX%HF0t-CdEYLXjUNB&B=` z@)<=@L3|!z3KhWy08j$*9WrWFQY)_*b8u>*|f&TuLm>|T67;bu_$>PH zgXz6T_nIr{9X`S$A>hhUox2)JcmOmZJ>Ww%=BQdYH=nAQ?)jXZ%w9DxW9AFl&*2}7 z6owG#cKF)zSr9^ncNv!2bbq^=gM??1&#;!=w8&?HCZsLDT%h3+5SjW{oxnqTlS0ut0X9`jtm{tqx`oGQaYTVC|c=W%gi4d zcV1WNMd1Bs=&qYULbpoHh1i`zhzWF`0WIhb2aN2|f9<8=8IX<-_q_%IX0nt zG|Y+MAfYZ#(d_*hq@(gL3<~wCu!+yioDXBlI7^wvs<&DFUpbJLX71iZ+Au3 zDd4R4_*LtD0zF1(lXpYMc>k}=phn{XK&3K5jIvf&uN~$@)Sh%{M=NymkWTal)yE4e z?3z|e{Zc#-*iD3ZCtlx@JkO7Lph7(2oMf9q;4^Pj)2h>C}(_K#8Pc z2F}l?`gl$T0@7Hw6j4jJ$L-ZIdCH8C|90+9oJ?*xVgRAf=S$a>N*ky zu;0w2Bd+2~$B8ZX6}NtTc9onnt%g>1`u+ie29lN%A|zqm{PoM%@x+{$R=`fuUT9dV zoNfRHAXawa=}9`W`HW}d2TJek6IzuveN>n##cW>*9w5QGG)EZ;A>}(R$~0<)_e!c_ z8GDD(=cLq&z4Q8LlE#q{EdT6M40U|??r!%%@Tj%D|B}5Tg{@C#{kCpn8}+uRAwS_ytvGa$aEsFJ2O`v*#T6$G1NSBa-ksck z-i-?ZextDpI=|!1zhm`QT1H8i$2cBRK2;&zI4UAUqtZm z%fAfXms{+MNDa-HwW6m>A&_>21k7Q3J4|XDdA*5b$}TA>sU?55#|{M-0vX|%t-_3@ zvJaiKjo|XP@1H7ykWiiw^UH!dax09#=-s&hhNykwL}^5wYo8bw`~a_W8zz)N+o4F|#JU(Ww3BHOX%HX2(`1lCqj`So5>_D_nIs}k@l__q&#!0~cp@TcOv^-Wd_ zlh6YMKad2$Wj5CbP=rU=GbZG==jh z72-=kHN~C}GQi;> zYlUbzXhHIOqu&K4TO$106LMC%{bm6YK#P-~?LY&t-46!7Alnykc3{VM+eQki@flAa zdsgV4@I%y6`wTltM0(?kE{9OcehcKK7#OlPg7fbCnyP|SCKREc(n<@esqVbz$1@qm zwjO_Oea#(S5}||r031FHgqWtZ9(rU90?~f05QG7~GN2q@@k2=T{r#w2A50YHP^t~5 zb}R@4T~oO!A`s{HY&G%!Z>#^mZL9xZF=4;W)>!^ar>pH}iw#q~IUmFv*!Gi)cG`>8 zkhBEKy=s_k5I!&zV^YZ1&Vt={vc0^RPq)ya7B`1jsts^C@g=8Bmo5a5F=!E7JpEhR zA-G5*2v^$wTPftfBuZAWZT^(@O=L}W&xBu$cX-_RhR*Nd%2GlK*j75LEYHw{U{`o4}1l*W6-=j zbu{dH*48Xy+U>v!Q5{CARQo1l&u;7+kJDi4q-ps0?gucyBbHA@q9CI^i}-LW?W-Cm z{%c(DA5E40$MtFgy@!GyAL7)`J0L8@KVkbFF@&aj_W9UHA&-aYKPaB0EFg?sJzaSjJ8N_P6^n-#;h$&!5tR;dP`15HURiBpqf- zs&i2CS&F-M26?{*Al$^v`~H0?Ml|5R{B)LTj40<+U?vmz%qQR#ex4ud;`ncV;Xe-i z4_^)t5RuLYm4hFf%iUAM=>8`TzqlZ@&9OV`O!i;xxZ9sEPd{hbUubNzKL2GB?t$Om z*#?N1U_c5ndvHD-6$qEi`e6$)8*azY%-!I$^AM^W?u5r19`=Be)}NCon1dj+|EE4-;%^e3{VYcY~V7SxL%HCl!+VMG)@VpIpy6YgBioT zJa?G&+6TZ9${m6Wn+er(mBZ}_&V-5HKaKnVg$^X;S^6tr64JKkpr=LT8bWlO4b>uL z5J4z#c!J+^J!O6O?-Oniz^3EJ8*l-QWzL9`X~chb+h%9ud(6%pfb67hz(RGLb~Tdr z0d_?F@tL~ep`WlId~&f(_`1>dJI)vsvV8nS%C&knLb;uVm=px$qCijp_KVwOB^Zjq z?$I)1fSBrl`?*fH3X$dQd6jE`ERz!BI}3e)x9l;A@_@e44)weI6=ULmvEG#Mc9Xpt z7g99m=;A>#5ze6hl8LC6Y`|SSC)m5nxwN&^dH9z(-12Q-p6Ws5e#+e{^fej&^;oVB zwjJ5ZIP27~F?kx$*c|6wWA|!6mPr90I)U6~`gs5#_6fPOXVW1*;FOD>9W1{)fr?YF z<2k4ju23Hj7k)n=5KZJM2!vh`S57d!Np6xxxl0m3Js0Ei;R(Q*a5)NaV&wpn!PT$& zUWrpV8g;#4!=|(lE!_Pq2N#Ld*8Ye_)rps6nwL%}$wGwWf>_gNssAE}2_{Rhm3hDk z76?xQCQR7suLz5+SIC9ktp2-C68PtP^|X?yIMI(JKi}s3xNx7p*e+Wm#Rj>m?#ln_ z^PU0?ks=@=O)8=#b&IMM?cd%%?06)4)uea{P=b=lB|mlX+>{a^6Rx}Q096?4P{RxG zQlssFoMVeDQ!i=&kN`s8khn;mK4RKfOWMk7fLTPIt#AiWW=5agG(q=d1qVBZJ46$$ zLe1#+%XQU*qSwB-iWDl+RlrUtO&c~kkDK})$vuehQ58H?bc8+uf_TJ|Ex9o?&|vcc zZi(BCAoy96G->M>L{!fip`M4B8)iRI#hNK;o|!|rZ7sZ&+`2v95yJuMT5xMW`Ky%65D{Ju1&w8x2ICUU(2P)%tVZFK%b zyV9qlB)7ZKnSx;SvGT71O4ksAl}6si^K=<*zD`YV?F0+^SiO@2bV7Zs{V-1-32m}v zrlGxlGlJfD^Ikq6t8`oj*pTFC+;s0p$Y=lNZ4{{dx3K@wjprnpI;JjOAmIbWNJ@r_ z_bfwa+$P&CFWkFP!Loe=;1tLc`{Ef#;*kreDq%VVdPWE;3sXH~k?E8~Q|=SJL|#o4 zU2BZrGS*HHgI4aD(%xfuKHcqavC8f9Uz)M1b||JOyG(oahpm?yHx1mHh!{`;5+Wj& z2-r=g0gK;fKI6QB9eF(>qpBhWw;&uy*04yoK`ov8$O^#Bq%fVfc?7$nyDz)<8Jt@* zE?9(>z}$@QbMm2nftqP}Oo z765I~4k&Gmf><1t`}VoL%W)8P+^VDiMB|pYYcI<|LLbOTW(DQ;UqBX zy3uz)NjeQ!2O8OP>t#3Gf@0BmQj?FaQm3o|hUX)B5cDRpZwP9PVA6mc8hULJ#(+vN z?D7+kAtYd9z45YJy^rr?vTZv8NMLlBe$NBHvw*VxiIo@L?|X$vstm}H#p#w6P3tOw z2z+DEwQ;b3JkPp|veCZ)V#Vo};F#_Ppii)J)!K2T0<@g#yOmHi=kNjXPC`!SHQ4%n z8Ua(ht5Y_fHDz-%5*D`&zTZE>aW{;ck7o9We?^O`x%D- zg*XtShxgW{30#Qr8>{OJU+;eZ1{etW#()Ub4v3TvN6*ZPNHS|%3hs{X{_-bh=V$~# zNE<*&wGj1K*WW(eJ)(6A8Bqcom@Odu@4MU5UPf08$CPb@R1d5(JlfWr0MP|F!P~y#AzGisrE%XZzlB68A z07}FOLBsnj`tf`bQNu#i^+yP(7ArZj=a@rOsk3;9S@G-V@8Ii~$g?_W}S(DEl*pT%S`R%_Qm$(;fHcS{+vugLLSN zFoMHpy>S5fW`?CP$7G}!XvFQ+nSB;=_?!V146qV9o1G-+7{_{Q&MJN9IWUi+%odM=?TltCGjHwIdRU=IVa;9yM_)my++dt z;EH0v9GqEGX~}WZ>;DNypBR&8Q5J7K-K6p$$-=LC@OtAmvw9;S-=rXTT-fpkDW+q` zA`_>6{-hAH+>fIsNbD*y0ANjE=nK86lU26nlOcB^ly^)6pAP^7Vx94C^ew6@nSD#W zc1?IS*hk%o+>CTc+G|L9SHh{En$Vg3d<-~zO?T1ks#Khe@2vrJ6LrbL*Zr^9>|wp2 z+q(9}J4e^OCXnoz>BPEJ?J4Jwl0Ri^Tmr+zJf;p z!PbNTMl`woG*>;_WEakzNGvB3%>ae1Uvul_!RwtXH-FwEp$W2o-zos8WKN9dE^;pU zEj~;WC@1az+18izR46sANS;w3-}1?L!FMO<$dlNF*$b%&K99$}_1moaS^YirG!j-e zpI{dK0XiZHXv^VBxTSKw<-9jtIVahNi&yjB3#QZ(axz-u@^NJy6uh-hNDI6+>3s)_ zlM?ugDs|*0g{-Vsm|WoLmIceY2!~ zl@#r33g|7&eyf1P+l`mbu<<2n1BdIh-0lHG1XF>=Vh<1aP9GE6?ETjKrcX}1+S)A% zh6W=HH4^KA*6n~l$o$;Uf*zk_weXsuQy`tUYPn2Zc#k@ zG5&K?2v}_0Xazq$_zO;%oprX>02dY7Fx#ba$UrMC@ojVKqy!&5qBB_k{<@F;>&@)c z8yE|t)SGXZ0y*~^gSP{xdU7Z(UciyPnYS_)7S}#s!Y1=3f?BOvhujqGRCS}cwP`_Y zw0EAG@C+Y`nq(g{m5%>3f2*O=obLGrz!{r5KES;&A+9_6{#4p0wFK&cK3aDq@)&nD zrUyph@6l+Aq?7eFk8srwo4r>aKm-P!|IQvL~k>(`5%(E@8&uU zvkAI*aoL*Br*rXl#}SyK`0bhsnoAZTHMjjdrmAvlirvFXQm0%18*1{2=Vy{Cacl&& zhPwD+X;tuNMgWu*neie_i!PbuH!K7VlZYr$04@|SG1uO6=a zQq<9h6P$GjdL%jf?rBDan~n0zd2UWqdJkhnAvXJ#*kKk#h!rpqH}(LyK9uLDo65Z= zBlmm&;CnW2B`3|#yIJDaUr%G>UNPKJo!@}WakMlBrtB~Pr=O9tZDVFU*WZyic_ylH zo%98E)~Y_f7sY39p=wj4SUa)gXOy>X_fd(wkuedfIsHIQ{%`|Br*TC=A;6VwYI9CO zm#8=qk|8NwG^yP(LlglG{k9o)`I`X*!N<#SXcloNP0o!d7Re8LY2$8j$+h@Lm)&vr zJD5Wn!sCFY9hKN;3`o5}7rp>2-bXu^MS);KELu<&e;bNM8UF2=rb%Gxo#Bpxde;t< zaXoWS%!#5A5BJM+U~_RXD|NpB4|CQi%KYp~`)1?r_Vu@O><}HIh4D+#SNp&6%q%W! zI`mFLr1}Ojb9vFt`ku<=gZ&uCV*7q$Th{U~AuNZmPw`jNJYMg+%JQAheO_K5T%NJa z8gc^6)hc=_HiDSE3q4s=ObY=2#M=JueAX0LSOa^B8bp794+xYvTqZz`)< zF!G+tV#%h)5nrgW!e`;LHF{!NEW-HCDRUm08aLQ2k#Ek3Wr~m2H_&-gibmDoGa}jn z^J}1t*;_jMgN0-%0!2+);y^`$?hsz@K1x;ag*nOS3JLbeOz9SY?2%qk-%AB~G!+yh z3euv}{QE54D5F}h{_L=5v`dls`IcSVV@X6brY3k;+ago=WUqyz_cISA3fBb2Ehs@8 zUq4@_nH9#{z5Ei|<~rn&-s$X(Z`8zG*aY=RRAnmjQ==vFKJU%ul`28D`5hf(-(kw0 z03k}-b^eukB_xiwr;fbkU)^<{jL51xmqhhjSLgxn{=py3CXey*)hm&>4>5jk_2`l!QYZS>^x5lSWm1&O~Z+Qx6 zb-GxDOj5lqe5=zxp?E6lbK&Tq2UaSyZL0Z7R0Gas*%GFJr=D#&ghp&%A0oydE1_7A z2dCUHAATw|Y;&Sw80X52QefYm*)Yig?v6Ys@tuB*y)Wnf<9bqwuS4R}(Y@5Jv`{GZ z>tN9)x6Ye8wGmqaT#z<8bXmP=jP7C}*d#D-t_@YPD8aYUv)o6EcH5y_xuRpHX-iTX zd~jjJnDZLKz7c#%T#VwLkq`l27ui(AK#imp6)BOP@=%k4q1)-9^_^=kRf6-ujJXhk`jCChL|ks-~q0aOZrC ztn#Vvu)aJuvo?qd(3Yg~61Z~hIW5kb7PYfYptWB-P&E(H7f2)I+PmzuAjFLG`V!0# z6#EoVO-GgFN1<5W-s!hmo)~(iBpNAi+>O11y7Hh*et}2AH?Te&j*ccUYTL zhV9s1t6+6`i#xa}Iw4i&UVpH8D}24_&ny7M9!Gwd8r(X@cU(|F#$_3&{`qFR_c3cT z^A!Up@xLxe)yM5PKh%&AohfPb>w>|=m6E2P*KZ%ef8R{sgevMd<#p;vv0a(c^CSd9 ztBv-?DDPCiOU>OsJWtQ9$_Of_4Gr&0RD@jLuAF$=M#!EoGBQ-OcHjyqwUep`49F8N z5@#l2%_H*_9Su-K$#D?~jNzx%a+S zEmlRj2mxmK>I3g@^}B|{dh+E{*(rwqp8QDZl;A z#p9nPG@fNFg{E$>B5|*Z>Ni|xvySevk7OzjtxzFX^o?t5QPC*-?o7*?#M?S?9Z-9t z8gZOv)4X_Ufm-W4u08}UDg$_bp^+y_?p)J2MRQV~_aIuB;H#26nX-iLlUo4nt+)i( zi$b@f3+#JPWwomP6VIcwZ;<8CUh-`O*!)KTIPLN-a21MnpL{0xEku#XaQ?JLRyHp- zMa?|=zMj;0iClW0a#?-+S?b8;NVA@A!Ac9=9$&eyYeo4L67nZ#a;|@`9^JDwee~O@ zsgU)ZtBprx$5cr|E?VsP3Aj3Q)}ggR)_1(aU`TCPuvFldfYsBuuF(0z+zrD^&1aA zbtpNy=za+P%viu*t|JzoS$fn&t!WC1F`;z^xZbPGuKJ1%lLm|GBHhgg>TmU`O_kzj z*vN?rl`E}mbu^;n=eP2_Tbo`pAGwF!mqO0b(7uGem%M7HPS zjQ3M7Z(}o->7npFnd?HV%lCCirM`c?XWt?HkmErKA_%n?MdF5_|25<>EhI}8<^L$R zg~<0C75zxc^j{aETGL%2ccIv*61}_%d3+YEGLfU%GU()TH;3SElzq!=7?*pLc|mys z+7-^pg)}ZR02i*a{46L=00~P&D3~!n<&h4W8(`Kl$~rn5)EJwySKaIRRH5p6|L3Wl zFAwC>aj4#fxFl?){ns|_5nBhowD}!fPC8yHL4n229=&0WB{9^`Z4b9&uYW##7_Aqj z{Ty;}LOsqryL)Q&p7sXUibwP2%|f#A5&lwz?kXa!FGejLH;Ah~?q)lC`RJRhWKHWu z@GR$TTu$lA`5-GZrMEq&SN`}e{APy1UF!ie&k_8jMn}ijwh%)n_Fo_0wnJdg?3Y<` z)}?L%Z5!|YFug?m7R?F25wM|xO*J}?RJhtw|=%3Lr z%9Uir8h5-a@*8COFSYC^Umqa(`i$4xsF*Al-+KF4X5wb&=VQLv;@<)0mz6-MVQd!M z$@_B{(%KQvkDag;zNPD9)w zF56o-d^BR34c!wV2=S>iMF`TIdK5AXHPuU{?M$|hMXzm^+&;20tE;E%UX#{Qwl$Gt z!4{cWkdaC0n@OelzCXX)d$6|O#8FAlPcr68d=hrm`6|>QvhrTI7v>=iIP`#|)%)Q+ zrdNb*+ek<2cNVtIc&f7pOg-u$z6$quoc*=N^*&4ub{pG1=_?f=L3=Dq_Gak3e7yrG z)n7Yri(}Ou>8E&XSG3`NEAv_yR!h8T#YIZ6Sj*R8Xi0TRyzOlHibSBUW8h?xEC_%Z z%E!)Jq^8Lr2nTs%U+Kk>2M%HSBa}6Uy&H~Fy1o-M4Fdc!O)3^Ec(z_g3O?#rdP5$9(sCU75^ zI7DE8hh$E*`%(Hn~V~8I$x>X;!lBX8cN( z+U66xoee<3rS-jBIQnefQ~WYm+QJM0i;w^a^-LIi1(-A`@?#DjK76>R7-KKJ+3QoR zf!F)a<^jd6^`RUa>1q)u_6l|;^_5~4tepEk0ZMNh6k7RSB!UMVwtW3Uu@1@ z@yig(RP~ee+9$UT=zLdn_+Oez_vCBcLxG5>rI>AFPQ?R^UtSRX) z*jn^x19piS8M+saYC*NO#<1(IWt4Z-=5)bY@6IGnz*Kf@*8XrIj0+Rr<^#9;KG51^ zNg55_*Qc2iPvD|n0xT(g^bMFd2FhUo1k)# z&bWOD5g+A^XS*70w*YS!11q+8jC>&AvcG*a#e>B@5v$eledr#9<2ZDhL(l~e_FuiI z@9{U7pt@t{)le7`vi zxpB&x(i8C^hwS*Yxrs}l<&SP23sq9Yk0Sjuz<=h;ZqhLAt1yGweQ2_rulMja0JfV? zLM5+k5E|fWc@I0zOmR1;r9RbJ5w?j5v@(z5b!Wee9*C#fDUuJ&s7)EyDL(x}AddGoex zIgZCj=IF|&%ney3w+Db;FkNb$MM`*i*1t#W2{bd$>E^e>+Qch6P}s^g6Yp{T+ZWE_ z^;rCge5b)3UPj#E#70>SuM=Caf}4>eU-8*iWOd!Q8OkW1dFWCyj;P#I60TS-)0Nbd zppSZ$SxqN?MW@!@J`pMir$s~_Op~QlIB&sPAh~NU-zsLNiJpQDODZszm_Df~9DYZT zx#Pr6+@+JD$6QrPQmlQUg?k`Sf8l+Mkgwtwy!E*D-CO+adK! z#$UOyk;^xhVK0uvbdo8ZS&!W&%9qL5kU$?d_n~Hck`USYu~+0VmehNhL1d8ipLNRdNOEai>G~k1N|0O`NH7e(T_I-azQa{KcjXoOjyM@TCpCHv4!A?#3@grp_1 zzq3t;rNW(%GfIr^!;N%XJjMZf&37DXp~yj_R3!aB+c~@_K=MuP*63H$vxBlx-OA>FZD%8zDDN#?Uf4Bs9 zU#M>WJa+fG!E!A1hvF^?GUG%9FfQs8+vC0_K0kikknZF4BgbhB*BC2r zKMUbB7(G_v>0>%AOd&NEz%i&o@?wP6m%Kn{^S%J4SUMmAlbwttcE z&cST)pF4Q)bkZK+>9$2%Fe~&o7Km)=NYJpP4ud;wN$2~xdUxlSr~8;4ZQuN$jHEQeG}gJ4 z6lKHKVZC1}OJWdk(0hI24cA=5eIYiGiC&&n?4*mY$fk%&f++!f6|!M8-QwX{|sRAw?n4b)6=kIv;oyF za_Wd>?#M|I7a=l%QUwvS2zx{=%08wT4kU z+H6Cm?dRDD=14z_2ls_ESomZ%Alz|Zscybpx+&Ld-K+k{I&9L63%QRJN&<&SA-h=Z zI$b2)gRzx&F#0R-DH3lMo`4uOI=IunQ|fDENM^FY+di1s%D`##xY)kBM>3g1oMY_rn4$ z{7+D?=6&@#5p#yyCb;~(or!|1URUnGTZlL#ZQB zGUsux-w<-LvcM#nQ|~QnGuVyrSxK)Hv4`oE1m3qZLshJFERdFOmMw)ZoecCsLgsE^ zHU<1Rsfw?u1%+krXkRIf%lZuQO`-(Mp&g-)LNPz2j)@~BL0>#S=wIxZ16yR2_~} zwV}CVP?$$!U*#ilIV>zHw*Tk>{9%4Cw`Fu9<`<^!7?GuEP^{c%gVAN!>Ueko_pVa) zP;^qA?DIL4uwul=>#=Xmw$?6ObXIkl(QkkOBSKDv0+_??%L`A~3otrf(P}w8h5@O& zZdf#Xo5}K)8_cVSLT;fpU#HjO5E?8eo`zI*tChKZ^F<` ziMNl(xI{FV48lH<4jAibw4;HQRWTW}o0D@AG41JJbXPJ&G#8Mhtym8RY-Bcb>>^MZ%QauCf-8L^U=@fj|c|TSTf?Jj`H+iC|0aY+&VQ0 zP&r~qqth_fayja{52q?$_tF^LkB%1AV-+_`yf1g)qy?hw}4fdGZNNeFZ`?2qHc~7#um`qyyBF_Fms_#p% z=;=_Rw*;0e$U|vu3>pU!aB(*iM?Rwgw5cE(phj> zRwp$dGj8odi;9JdZ))TTsjq|Fz)9%%P^fL{`@<~GDRv|@?6WkkzZ{{Z=K%w7$*>9L zX+dHm#Lj5F7k0UgrAVh`5}6{ctsgMI)8PXRP6=iW-Z>dFgJSzEWGida{N2r=%s#ur z%r^SMx)jpi9$1<@;=Aumf<}^H#gi54@kbKfQy2II22UZ>%zp_PqIEBOOT>q_(>XP~ z%+Ibv3NMK)0<5T^mPg6vHt@~rn`k6dLZ|K#^yAv?{H~j}f-!e=srqsVOK2Mv`7vJr z3_ugHk2hrzuc-*N zkO=@x<(aL3bDZNX1J|R!z>EY{UBH}v8COUydpA$D)y1WzA5gTghP}gr*r_gz{CY>5 z&dM|1CKcfhaaHa0*j@;h?`D=(J)eI3BBwHW(%O#dSBL8{YgtX$KTD1*Eh7wt%B&ef zq-t+5J~<36K|)4?UUn%)iY7MCwqZ*PozF7UNiv6#LOti{W>w#_Oo@|A2mkOj#%Z6g z--UnkHOgon*^A)-dvnd1FRy#*A@hvM$O0`~J`z+tkWN7{$C~drLl#F~yVXSDXL%I- zluJJD<`bazFupC^L-<(2Z03~RwNj`&d~etlV5tynQdznxNinbvv%f4+GD`cc7Ib=_ zz76veWs1Za-c48SZOb}4TBfLz53?&Ud>RxY!z%^$mB!!1kPdl}4yiEobSist>>2&Z zn|~@kgY@+#vP_&M%pWENQ;?#yTxk#7F;++yftVvrt*v5m>gefgajowyy2M;Cg%ebM zpO8~JnpwOR*(`T(Dyd;P%T6`h<||a6i=Yi(zPu?nf^uK%1!G0(!+b9TI@VklazT*w5QhoaN z3px!6l)0Ye7un-AF{t;)U_1#|QxtcY&i_wh{ur`zq*%7h@Vle7AT8hbiG=1PVe^c& z;P=OF(YVJw3mNJg$O^^2?4uCig|LW53CwpXQZSKkC)?_LFGz}9$guB}TT1REbmN)6 zGMsn?oQBJyyA`!C#-^m*{;N|s`3f?U*h;MKFhSSd0z8jUDTv6}B3L06NItbUd8Y zJ2^ioU6oJQUfiqCT_JZx~CLig~9i0hZSKbYuo@?u>TQmK3^VGn6eH41W8?}Zk-@gs&|#MSLwix-Xt0 zk{<{iQ%TEj-J@44>|TF3WhtmQ9^{zN)6=#YaUu=og@IDHT4_t-d8yI3aBl~|cK zXIx+-?@Ox0zfLoEh3W2c93LiM>7rmwUVA|cO(beDTBu6=@fa+SOB@h#9gYSQ-JvKv zDf;%di;)v@2Vh=bL9tPCO$Qn5kp3EZTI~5c_)KLH!jh0eiiphtxMjcrP%yNz9DOpI zd_n@MNBWzKGOfOs&*p?#13(??SS z=k?dFEnGNUlrx9Va_||1M%o~@lEgk7LVYhj1en%$qnYvEU$i8Cb%To#0 zw$gFMAtOI$PYuYHeQ0BL9!I8o=y5z~=IzElyIn7739a(tOg0VnHIRTqolWB2{)yCD zBHi05L;aVRsUF5%qGg{uWgc&r{U~qlBDYp&g|MPQ&e=UV7={>g1Q}#FI8kcV97NF| z{%Qqj_Hh*l_XtI?%YTk!I-PCRdfr!{61Ke`x=1D?1lS)kCpNt3kQUg8)Kr}pY!W|2 z@?wGF+}CaV$rPt6%V#fjhV-SiM4-cZ!sh1Wc5;4~bw21Qr%hjnv}H>PnY?5cxC0|D zhZz#B;O>Dz&{2M8+~cR4CgzgQU~S?YvT{wN-5E1Se)K2?CPeeCbm|{$LJ*c_v!dU| z)1;#3%bTY%_u;e`uK|&mz%|hQ1_hvJK?eOd)|4YC0pxB$uc`vYI4a-5=+<2?h=y&% zK0+gGucH{^j@%De>bUNAy2s;H@QSuRh5y&R&SlrSF6!~;le}b0=Fc0v}jGQok;omM+RK4BU+1thXM<8;i zC-87#C2jq9qlIvXDg%F0t{JZwjRsvIy>c;& zBx9T8kcjTL?Eaq$VXJhWNlEqdubgaCuMC<$Pdi4r+z;94CGxGyb?bXj>_uxPRZuTP zn*t>+(GxFVmg6UtS?%#Kv(l_+gpg_kss#77zGM-y`5qiRyX;*nVJ~yqH;-Oeb-ru9 zro$r2=BB;40fy}mk~JDjG?z(}etcw^nX)kX1tYELdPF7#FdKftH25b%i%GE}uf?Yy zpk9g=2vjiHV@I+jZOt@MeD*7h=+zKpl>Zv)NY+Tzn@Cvh@y8^#?Ct| zXMAZ6Jv?*0>Ptvdujs{2&j`)suosFw(H0(q*5`azjz`4Z+1%U=qI$d>_#pd1kBYjz zT1m8U-F(mnjBqeV8f`&5f2PzF_Az%rG;W7PXPHIl(#u1HoUau*HKw#WRQJ}5sS30X zpPp&co|BJ}O1$|guW^7_2}5$8{l23Kx9@xrmqdMmF|6$4)YMQz?rAXleuQW!oVm*bFHI2&sEbSG4w$Lflx8drWnVQ!}fsUtP1T-(H7z-~@?x%2}! zkw&${Dk0~lKF#IYH#uYr={blYD z>cCghna>LwG`r#j>oub?N{CPdx;{T}NJ|`9t!@7HY4uGBZbor-PG)m)x$j;@1d3OY z+wgHjCQ&DnY%r?iyz<`5c37*fuM@DJ_T0R<6&vzen_{{o>PSV5t=5u-i4fjO25wul zBRdnHUqL2*8(O>}PfmJjy>Gja%$lK8VqIdSoGjeFqexai_t9PCCWwfgJ|Ipw97nFU ziO~EfRDqe5_`un;1>O#*yWGT5Q7qdQWVEDpNDQ}sf~W!)1Cz#9OAtgQ!(zl!N!||c z`e9Ha%gU7D@{Pca#iYi^ZL4MtS9ufelP$p@kpv5P#%L=XEK8c2ur+b0~$Rs?0QMF|o+`sX3DTC4gHo z&+_}_I~;G>QZebl#@X9y1{Z2;Sk5@WHDEu$Y&qHBN7e=43koka;fm&LWCY!F2Td;N z?d6^hq_Lwkl6yg%KrdTe6xkP8u|#+2vM6J{9qc78%0554H@Nvsp+F}xp5vIj*1)6C zY}1qTaOdJPk|w+4(qS2l>A4V?ix0%Ws3%X$c$>n2QD@P)3AF@uFFgP~1ui#l(T{DZ z`)t#v12#oKz!MgSeEG9Spb@UdV=s>5L?PC zsPYz772U|w3w46&>GneJEa@h&1>gcdgP2dz^fo}6nxn+v5IL(!iL6RO&qY&L1c01H z|Apel*|}*JIL!$WQs?JCV$f)2-ke_$Mx`^*+I3v;_eK62R0F`+7r!+=&`X7hrqU=~ z5e+>`jd?-(2n91y9h0km$P+HCngZx}>{DKXX07OnWH^TwijA!e{A7ndLW?ba|=i7~^ zruKr8w^px1X3*%vMqeNY26`KAM3^r?AEO4-r+|H^1@UJFmIQBUFOuk5Sb1h|a9V9C zqy?KHe`jRN6z%k=8m4aBi9w{Ap&4R%-rzdImo6278$>aU{HWKUJ`d$xVFG@-whPhz z2B^#uNwJ$_csuQy#XcVLr=%Y+(Ed%yf_}`9@f7EHDZb$E@jZ=>CtD&CVl2h z;+;K*;qt*};SN;Yw8VN&W}YU2c>w}yErTvAA5#T19_cv(C|EPJVTopNoXse&=5QZ4 zhvrh-EhMQioA8!~4cH}zPU;*3MC5Jwc+}2y>FBHsEOp5L@eUKz_Ns+Zl}2kiNa|{* z|A(%2Bm3rXxiMB(Xh&234b$FvnjuMZ)}Eu+L>hx`rJKizRyWG^z%+Z}W^s zz*wbxS3C7Iopk>MNq<&9GrWbEa5BkEnuO6^9g$Fyq!QYT{y6m9s-Jw~yA8h4mgfR2n zQ=vlk)-7dk;#Ji7OYzhA&o1eNj-N?$xfIa{PTNcEs^Hx(ca8TviI~jP0{c9IMN1^$ z{#^G=`Sv|Ws?DKmkh%M*(27`_;%cgT$0xU1NW9E?bzDAxUiBI*d8Cfh;GSpQ1Fs`a zbG1VM_74->Wx!wZ-~n-oIbCw)pHNf&uG{_#+-={5WrxlQuG7)g z`M&E8MZlBlzs(@RXoAa5XEeC|(_}S5M8?0S0#XceE5E7$Ej!%XF*gTqHi?(^La<*i z0F|jF>HE-&+MbFiJzFpj_kkpf0g|~*0&T;?2Xg5)H$!61AzTCa^q)Vdp$>b#CNOt^ ztRvcNAG6T+=U_pUb_r?2L&bq7yy)snLb3md>F{@0Pk;Y7`~V1oqPu>!Y~#x@!~6F? z+&ozAdbAaz=?eOu9pqr>?u!e@{uQ3a->nXqVuDu>P z5DPw|@X(3c!sWv2!#PCOnD^-fIA2(pW<}OQ0HJeG0+w#+)vpCxN06wDSzlk*vmM>E zIzVQ+)vE4i2F9(7N!&izZwjNxe_iQzCyY<eBGSOndO*)RD@*~hNA<*F+Kkv1Hc7w013079nKwGh1i6j zAzGHj+?6*^zCbu|SS#>i<`&O8#8Y1VI&uE9RtDHY$}OH@228&5dJ-Ab z7-?ESYucfj76^Iy6hQdkfA{qDr&?KQCK}|Qr6RyJAz1}iH9)`Ff=xieI$`1GQ{{fF zWQzp>*bVSlNx*J?zbV3yFoQ4E(t4wwFW;1I_ziC#LV}CP2k~F=d?Bq1EJEOokV3a# zphOq}_eLmp}Y`aJwt(K1DL+kcd#AD0T)Xjr0mf-EsM(|G0{O3$iV4? zT@`u~6IX`hLHI#Z{FW#zh~cV+%el<{V8ByAi!J29>UB&5k{9cA*rUSeL!;aPXUm6I z5JXxdpl&65EA>V511OIN{fX!;d{G=V;M6dl5FwSh`1z^99!(f`W!=W@Mz~_C5tjj& z`-mW@*Y^dfc~S&z?Yi5r23#hh#fRn=bGk1zbx%&{>7q>{WZ=Z%*}(|WEzts;Gb;5h zoU<=Jcw0o$0t47%oXh)c!&+JKlcdtqZ-98CR~dh$nE`;s6sdTdw|0f~2OJ}6Ugj;3 zvTAPFqX_kmv#N=4gWF8AP@0{W=aOSiGlO0 zplO?3c4}lMtr8)fktc5t_484opby^Kqcy^NXQsH5g+09;u*K|Y&m$A8gDmCYE1AOD-UbGRSXjg9EcjgN zwIWF);0qT%{mwtIB!>W~%uDt>&DOy_76e*sVVwb36Dy*`|9!9~dl8AxtdhE+pLILv zaMkKSxGN#lZbz%FV0~+$qM}fi{R%*G$~mf1`Wc3Kxk#fnpym0tp7hLg*EKi=k-fBC zsCXOXOas_s|Dis{Ei&>UY&y*0a$rrsairLTE-wCp5J)@h7^UeH`62}#_ceg+YYlz4 zo8u(^TAY4w0R0ufpw!>r1K#!r1M`PY&cFd=6~zSK%tPYS zt$_~sT`(Z#FW~B#VaL{>FnD=_N zCE7Y=>bA>h{j;}kRVBTauJm{yRH-u}5sN<0M<&Wq?q8Wl~ z9O3U4htg#zOSu!nS!%*}U2KYaD;1na9_GpI=eLYG+3Nac=dU;Tg{rMF9e@xCAa8)D zX_>N(B~5Meh2z$%gIEH>@6ryNWk){wxw>u$`z?W4K}6OUJJH|qo&ZE&{+W-pSWDy~ zF93K?c_bB6eiMv%+!-}q*}DaMEKjC=D|@~uNIkOKl`Y5$pyqw*9(9vauFdPSibX4r zKZZAcg1j~qa6XIJ{f7xJgApC-1nN45(CFXmIZ8(pJAJiJ>bFX1IAmZKd7| z;BMo+WdYi)z!XOn;u(=wvr;>GT2Bt*Fl3lFcEm2&w8~pG>IeOwUrg>eQhtgJtx}W&%L{+#&g=W!pKE}x?gGC zS+(E_stcY_-6G)5Q-+ILu)C^}lMWa0TF=;)xnGv1XIBF8!ckWN2z1wXd)?jh$Gu zlqC)FL8!42ky}vgZ?5-+%_DX{aFmV2PJXyMxcp>RLUqvgE$n{$HHtabx`6iJ525?K zrbuqh-V-g(Qm6bJG2_0G(uWF3h~}DWVrKUNpsWesnl=e!a=SbA++?O@Xu;Xg;c5IxF_S}CEMod6<0)RWcDC9{a6xe0$4PSt;@ZBj=*wQwd z^o{aCbXEr=8Cm}(ChT}8^+Prb+g10 z5sn2WKiSkysEHkaB#$CHzyZg8Br*Pdp4&3qrvVt>Cn-O_b%O8JMhZ#T8F|`b&1>eP zZDJjVh1TS^zxj1Gh$t@8Xo_|iAdz;!&_a0ulwm7{}SW5O2J4)r^S(IdPJy}km zxg?$Cl4ie-gY^E1x@yN0AqC`r=Q6RSj;teb(Rv}E0rR5t*vY+_Ps~f_xBg6bM2u{L zX>N5UvgU}~R9HT?>2S$Bg9%2B@JE!OIRaI^9Z6zHpVNnRkrV^%)#KR9yyTq5a`Y#% zNaGJqv*S}LV~`*SfFR`1+39%iSMhlRm3ptze9oY7^{R#SK&gxRcz?mKochIpF$3S# zVekEY=z6V92df3VN!LIJbNK)){`r;|`)$H_D8wwAwz@PVuW5jB0(O+lQnt^)J`_?S zngVRBtO3!g4aP&*?k^8q#oF%=%p06!$_Toa4DaQ^eX0mMZ>z4Yr^}O$fR#qw}dr_~SB#BJp z!NmD2OQ$`1_NWVoaVp6l$=9c)gn15fC@zE~_UoT89^r%Ro-57g!2a*w!+CZ~Xt8&Y zo%5!PBnwyD@&Yq(iCrYNQpNW2VJMGdHWMm-Z!1p`+Pl+7B8aIlF~F^7M`7*l+l6m! zbfuQ=U2vB+v?1#hV7}4S8_{bU_#p{f$ZsD)(W}N-t}Y@FL_^2DgQ<4)Rl(+X)6Dvt z-(3pJI&$Mp&hhtm-l%$J;4|B|Y(7@0z~EfpFd((Tn`~^(7<^zfas)=OQPM`8!ZI_} z{gxWN|6A%&-)@4)?^)VprxtR`ms}DbH?pfULIdn1wa6&cP3|=(k_MsT9Pul(DUyi?c{s$VxdKq6Aug=iAB-fht z?0EHtLf}JJm`Pi3Ix%qF2>;-Tn)Q7@438Ebjf(AO>SaFKn>r}=Qn|2tLQs9faG>_Q z4lm9R{#ZdBqEgq`5$w6!zlDl6il5f`Q^RIez$Pb0VX1)f`xDO}d_U^94If9r?1j-* zd}r__=kaUI((r4W7Ji&{BSw3siaO-pn}aw;Co4;dSTYfm7 z_7pS4<`ofH+dN?ElZcpnXNB&}4%yPT18l(#jtjc8+aKT3ARr{hkGtiFtVCp;Y22MtQS(xi~Q;>b+L8lm_jqbsVGA>)Vt?)j=qhS`FSq-S*LJ=f&gJ8JPLaIqS7W zxS}iXU>K@*v^l5e%@~#)}bKLy)fI9TrUQe#kSMDtICpxfD zaPC)8F7u4$DIM#rjbPnhr>~sPgn`rN@$FOGqXyoCv+Q%46BlQehMi&;kNV+OF^2B~QR%h_o4Z^1d@ZUGO8z%xe2-|hm8dPEVT$CZyqOy+tjHpdZt;)`R()rP@ zw;lUqSPI-_owL+1ix?mp0c2W(8*At(X=ffphBgH-|+-5@=##7GvrS_=kPU2XSmPrC#OsiDCmrQ zLU2yJGW4d8=bOPUQ&rm+?ktR_47hNKlE7t`I_0@>*6J|v8@OE#J)1V_C`d0wwAKsS z(@@JEwd(~9Szsm2gsA~0Cx9`>oNNiw$nMq)Kj4B_m<(R}S-HFbxcyv3mRo*^xks?|e|fuM8}GD3D3NqL;?cfQ-{EO@LwkAr0~1@CJA zd8CUWZIO^imvEaloan#Yoon_Ur~r;JkY7)*M}|_dI&FGz`>xl5K=e#t9Rnw%uW!SX zpHh-2GhqobTT#YYj-)iS$nr34#DpZ}LoRx3L~J#Y<+00ELMEGK@Kf`sT>;zh81gX{ zmy3h1-MX>`NLsC&q}9kU9YBwCS63!7OfwA(t$&lryhRU9pvRNDYl=poX3ErCO4(T= zCfNDAuw)p2nk)S;1#_&be4L&J1yk=W{Q*=duU>n@P7`AD{K1{-6wK1h$<$aF?v1}l zK0Vlc#NK}!z7OWH$2Yfj0b7EJVI%uJD`lVeUKIoi)MLF4Iq#}Ylp9FCDrWTNJm5hC z0OS<`yBlf9L`YHCRr=}7Oscn!(jCt;u0ILnRE}atzC;^1qM0vgaL@!Rtm@HqQ4t-!LjOFvuwsX zd?Y)$o2u8B{dVW6YZ&T~hctm?i+tGjWxjkX404QpF*R^!&nn|Q6#JPmBoa8vgvNvZ zX!qBt3%{0TEPLHFVOtW~lNh5K{_uaix;GvBoi-OP-OHj5sY*HyfpJ>^wI)3p?5OPsA?PH+Y@UX zUwi}+??w+ZWHzqwbUOlf`PmGwoCZxRo~94lno*vMyiwjJ4}U+*^egMhsqH<4g7V>V zh4{o7;Nge|z#_@k`-JeffwRQ~q7bK&VfrY0eT%4`;JZOzondJD@tyj_!uR^_CSW*q zEHiVRxS&QuiDMxVSuZloE$R^8dyCO8m3baq9A27{8nrn?PPxkrSZ10Sql8z6C(j(S zNI)_paUDda%?MXEZ)APO=}L1>n%{9NcmaG~2V2P|T6vL=)3Q#o^jC}o877b#DijXZ zrgPQ4%NAzXtsnmGh5K?qz5_7E-<0}Zka*BUegja;kD|wvThjihfD#w;z7Dc0~!#Uk2rkVPp*zxa&kZh<_D;z zj^IE#Ispq<>nENoK8O$zNq+0Hbs4M%48SPSr#sc@F!ZW$W0kpeINU-VHA8G z{3M;Y4~6&1zI@-wiqGW+;SmHnti*9LuVs?YyvlTSxZKLOf&o5QEkQ9|^)BQ$?Zw<` zJ4!v7hCon9fl;x?{i^*a4D&5H5VE+Y^;CE6xvs3U+q)%T;gl&{wDZ@RS_uI+<;E`U z!~;8gHbs2w3B*{iu{jPINqEBoWy}GR>JR|caY(4X^Da^yjKz9);rNcTYTw?5XQue} zp4mlO8%QPy>y_ghz=;*P@8xYvJL!irgs0`yTquAU%D<-Q zazE=FxPaBMMT33;q0Z!r>mq3VxynH2GRZ_J z?PIp#>B+ynr%$rm1Byfds}jlDb#d&xqtqyb!{n4wqF5DK(JUnl`BR1n*POj*RoJ!) zzx7vMu1Ys}z>#FXRB;}QZNEgzdyE&CXhe4Ng^^Dj@PPG6vX#COW6kRR0vW_)X?|<5 z7tKS0-~#lTPUZt?arCGYCe<4;V0avv!LAfdl$8j8GlG0NGK?r?<0H{T`-iaf-IR^3 zyRppucZ>DxthIx;;ioCSjZYu5!;rtAk8piK9r>1|@3=tU9mQN`qalY_f|2|N<%xx- z7@*@*mI(e_T6kTGAXX)PUnHaY1;-Rx3R}ddM`4j!aAaC!a_L%*x9CVr{u8rgCH(vh zED2dcdzupq#qZ7G!_`12JjK9uD%wVMoKMTaGjV%6`zK<}@54@_PT+jv1KytmlJR24 zrN}e)kw^uGX~S&$$4nC{0f`Sn<~u%vk{M3RD?xDqE?(sCxybRNPPe*bm?`3q4&aZD zq<+u@a%6H1t7uhZm_eFe-*VX=TUij(*Dl>!J3vf#s2)C!A@^j8;P?98gM8Gdz(=BC zspL7k9gxYmOx@UexxOxg^05$h{4tx!JU3D#k3l*?ZT4OSPe%$IwqW{*iD2qgTsMu+8SVwlTZ^j_#Kblvu*1mG2(JVTH-$__MF>6G2SM$ZfqT${UeZ zh{NF@sHs9Id)!&@e+RR_GR8X+e6I$_4NCAIRM{&z^Qfu6KkE2TV z2zGpz-0GMP$ryln`u(uS`7;o#pSjyd1_&ne{P1ZY7(Ec|c8@4*z>pRQ_TaxG!Tzls z{;eJU-=H0|`6E6br7Ta8jGfjFc_VU(Jc6VD(<3QW04+jtDu@}5UW{Uj3Y3n1DbgVOufOhD^&;r=UH z=k(7iLZRH}qrD}f8=~4f@rS{DJfYJlvcpwx+DZEXoH%(9a|D0zp8WR^#}tAYbNOY; z`MS9MaKRX;Nr_ayHATMIk6;2caV^h+n$?hr#91E%@A8zln2s?7WHoOOy>x;&6=n>m z`{8^rbLV{yN!nn@?{Q&O_VToJG=1uiQg4=qFWOn*e4K`j_mWe>&RPoD(4)}3*wXM! z^I~r(*ltMYJj6qkq#Cj7!8W`lI471XZlLb)d=%d<1=|XeIlSvf-`R384hcbtR-+s# zqJu|aFMwj zW|k!)p=j@OuzRq@_h81f&>8rHXHG+va#;kDA?NjId%3XfX4J3;V@q88%cFnVe0|m@ z#Se}$h9D1=1{RqOE}9s{w^*^_MK$ys@)9s?xp(y;1^5IkRF){Vz2~4S85w3fLj?a$ z=BpiiI4nEaid@WTH|5g~$*$Kpz}#DNldat3KFAM|a)e)O2W`s5EmG}<4el06zsQ$Z zEdVC!N(^vX%DSdF92>f?M$jy$Djo$2po zKA`S&4pG?z|8;Eq{4pDA#5t}(B+h{&NrGcMzTMVXl14xx`q}C2sqnzmU&)p~BEc{AkyTSj6k6RJWQxWP9FsIQ z2LX8UztQae|IUiEpR<1DPkZA>frtfZKY5OVS#-bpkl>>TtB$=k&lw zWZv;-I^;XWx0Awm;e59IEd4?7_jU!&(OxGc^36zCO>;V`C*@ zQ*J*r@pVZ9o}KZL)rn*Wi{Y{N%EB+vAveP}xAKDT$-l+EkB!|@tJY#JO1O%! zIX;??9cIYAZ%$g1K&%dTTN$iLu51t0E>m2~zUxHocqAj9^aYzsu1h=A(otSfe)*@#8>57HEI=9KawPV>xY)`o@q4n?8ry}PYu9oES9;3t85Uj!t9iM zk6Z?bUBOiAsHooxZR~0=K(Dw{+E=dOc?xyBQZP)Ob4mcA5mTx z!>YLJD@0h6$U!{-qxr4tWPxzgLI5;ssJxeQZ~zzbB4Jh2awqqnFUAKiLCJ1{Fn zp^3}|*AX4XlV|Ijqjw6#I}(+T|3dA-ic4)|w; zP_Zz)@pHIlC$#BKCV14Q|E^14%f0>>t}}{iH1(EUG0m`#t$VpA;2%5*jV0yu$fKEo ziAA$2&LAeHry8BG047!<-vK6d1bitrvgp+6iOuz@O%^~+dj%sL)}BWAG5J7Co>&k8AdMBh z7m^R`w))-Qlhspr3nZ5ERcW1Q4U9aSDu4+g_o~)ESH(H_@-_D^vk(2cH+hntkbuAz z2PV{IsX>mE-a#KCbDm46!x`;?>5H;7FfmGfad>#O`&A8e#Lt(d`7RYD_yb5*^6FyN z+3Xs)uE^fp7+S7m83L$>`>oz%l3!74D{*kv$0&4G=2e-t@hc~bt})F9Bqz$mw|nn_ zUJ!#@TRE=&xq?SzA7||x(#T+$R)u5>|zfhkq zbhfEv+R%-|FO?4uES0<(ynKZaXy{)0j$_P6y#D6mWS6ksUDoANG4-iQ=+@f<$!rsX zJiin_unNgJuOFe~@V%`!Fl^J(4;}wCXy%#PazmYIrkV8u430bEy*7GaFPaDObOXLVQ}X9+E|Tyz&`ejtn%QnvSc?8hXypyrJjpb_0~3B)0gDjT+rQQ3UCeLQ49~vZtx$iP6@9)auKU6pByD{!tvzX^zQKn_1+Vx$`y==0F z5N%A!5NtKVQpy4tY)Ot*llqZaV{}zVEGGjfbPGd?v){vv2N-W5U^g{g2M6y$foW}L zPXng%_Q>zSyT5QQmuh%?{yb#JWEC1#wVGW;jc|9ett!7zh*OH+e6M9_PAm-NSNd_X z>!!=C+K~c+ccIO8ohTFV5T6zF?#KzPJEl-cYzp;6$X_>I z5G(rDpJ8CVu=J`*sWvIi*!cATp)*~%xZhXItM6mW%Y(NJdn;#u=M*el+Lyp|ah0dL zv2Mq* zl2GHJ&C^W5z zH7r@_V(SRrqI?M;uXBBrH;M4S12djI<@XRo|BS}(*(%?~Zz;XYXl6%Ng~!l@M}xu; z9UN1Gs9km_p^jg;uIEc#`uUJB;_*~KwasU<^3DR9E@qY`)bXFHJEC#l0}WOCD%W~b zXL0xuGtD5tvh6{gq}dJ!MKw*Oh(50N5V+XC9ua>j!@QS3v}8;8Bg&aivWr^m;3;)J`WeCrk)sC=RHQHHB4!i z3ChC699ouE{e%l(kXOSElUY8rhF-G|*@kC=U`7kMqyIXX@y(kzsgzm+yoNJOw4s(4 z+yFoAu7=_WovUL5CZ^qhVdRl-{Ki6h>#et|1^1P5gxG>~(*v!l!@`Uk=@Voa!-)$C z2w>^!%~r1=Ob*N&T<TvsG093t{Ac zqR{WcT0oC7fWPGh`HwiXs5h*^RoCy)sU?_y>vi_je00y7wAI5m*8;3&>J!4_H5n6T zBrO9Z%<>~ zI-`=%cv>b~54fP8W*qHx7FqIy2~_44-J+cBhbs?@aYCQIYUxsOzLL>Prg0XFr}85P z&qpG^6Fz?MIJIh8%*VcPClfO4nY}w)9)P)xYFPdhl|Aai5>7OI4}IMmUA_%RyNR_` zb=Tb2R(L7Y54}=eDh$u<5B56m9i%5Ry&hae4368e1YoO`LKoSXUe^;UMws%WJVRG= zXtRt;R&xr7NFdWTlDj7S=UkIx*&q!}0cl|RtwjF6N&_cMc|R53ZL8Y;cIN@b9z$;? zL+{(o69H=6qgI}XB)gGuc?S^1l@s4aR|r%Z$wqm(3;uR%80ktM5l&ny064SeY~FHvc;@t&%=DX;QWHX{P+7f3QyaC z_ZauvO{D2{i(8%n)7M&_N*#XX!b^obJ}37l*h_iRt9wfifO#zW(z%jPA9KOG&`(5} zZWLMJu2%TjLQ7M^tWaNdJY4sEfbRKz=;O^Y7>%4Fj1!E@Z~QcW+hg-Hq>59N8K$3P z3Wcg1ddKR&rWkTWS5nKt)}=U=I%K488Acb>@GBDnCu_dPm%f6PHobQ_5THv4Zun@q z^g@m1ss<05S+svd_|pg-jd3erb!PdwM!&n{|I=Ps_^!leQAjIw$6BEMINM4Yhfyv) zwxHB6*7gEQ$85vTfJPjQnJQ!ljW~E@if=o3ZFjKyxKW8#f)69?`2uKC7h(`&@kZf&P)j|$)t zcCukU!DayzsN#N?TDq>*yqnDqhWSP@n~(47l@k_i2uzlM7c^sTB+SHGHxdJlxKmm! z{617~0GqO@c==JnS_)eOk?wau{>wnuP~D|K3Oy~3(flxlg|=ha=@ug6N>juUhe?|R zLV&JskS#~WoRd#gyEwB}9lv3nc)_RvOt~8#A%L~UDlRe(`50?~``sANiVJF+I#+KTO8IF?Fb*| ziR4+?=HG=37Jnr5fr{x{)2!_8H&%T85sp~c5-gZu{#uBJh+?SM#U|I$22D^dvkg6W zy8Cz{cU4s2ZYsNT+5@Xfmf{PK>D(2?7tKec_$q``5Cx4M8 zdsc{%_uMqDzhfVgcmsNMFF~s{H4NxDCYeCxUiQ)AsiJJp9jWdKE4QT^)16Bqf|myN8||H`H!Q4Ul%^d zrBvH>69$3fcA>~4%Y{^CWKmu+L;p-+Q`lpgI1LSzrz(m)Cti7T7I_DKzP>7t(e?-I$LEv;b6s9SGDvh}-*{bL|Ge^8I2?ImT z*HyrH$)?G6Ye-O)>6!HdFX!TY#a z3ZS)FE_FsdN^4*D=1bSF@Q1TRR1oXL^DOsC$=6$jGg}2S_q<)}+1cD#7VH3_Y;oxh zw-?|el=2Ho1Q~Hc-sIf;So2w)y*C?=jalAg_N~yzx1-TK;P@tVeRyA3dPC8@fqzQd zm6-GEl5_&o4Kay1lXKB3dOyv1R9=l{oyCwdv%od{uD=?iH&jW5(6~P0o^+w-BVgGU4uZ*CVPY{#6;p5f=z86bWGTl zR}1*Gdx2l)I^%ZEWeuhmj=lq6VRoCsP6%f@XvZbwEy3HqkESlJUSOe6A0-?CG{<+g zFp|ZT5aoN#JGE6ZVWnyE)#$)Thl2W)E$BOhCGUZ9VFp%W@qgyL>yi~NgSGZ5?48@| zOFY(f+FR64L_P>9JIGuZ9RXCsY*+r~_p{(<)*z)l8BOej;+fSR15w=srj4etqM8;8 z_5X_9=((wB#Wn=7UYys%RwO0AV#gIP9ZVtckDZ#6e@2{qMx>5zZ@+z`3feidTC68lMr{@nxe13}5n>$tz}kg_OE<04xr<7&jj6hf>pnF3FmL?A0% zGcI7Y-~NWUY4&R0M@yLymUEmito7?ea`;eYk-`um>?AjmWj6}*k?K!t6m@m97w{O8N)qXy!qOY7Rrhv2E8O^y)^c? z4(Wqeo=F;OFO-yjd}z;y3}3$QhYBXj7f8UM(;-zhL^L%A1vD z=RLYGCn0;-ET`JeyERIRUj>+lCf}n@I7{ZhLgLQm@58ShnlAaP1h3>2xZ)qQQoC(R z6}wli4aDa@rJ^+!z>XU;k<#@ry!N;A@|fqeIwb66TRh+6a7nG$6ZcLa6ydq^`D?Cm z=?kgn)^0Ce5ce{XE+H;FclI7-m*i4{z-A9}oV(P$D~!BTo33fB9hFu7dG7j43+tE5 zLtghmO%+XwRn{N$x%sA(&MLhGN{P$FQ_2M-*sR!h?!SubeR1LA3m>H9-oez-({V-i zU0IU}D-BhfzpL(}lvQSvOc$1UFno>`(h| zgFZSx#mHL6Y(uyx(iEWK?PlWbX6hY;POfxv86tGHS=;7l>93IYv2=Mh_|wT9hOGdg zS+3x!=QoHTSFgL+hQm_g9*)`J1cx z9VDt_~b8>b)x>nyaNGt=hGn;ig63wi4 ztr`R=40XDm{0qE@g2`FrODvvAN)1C1{MXrynb|R)0X4kd+MK+PoRWMC%rdzOw)g7$ zLzbX166S+-Jy@COL#=2N^smQFnP1AI6|9`$SFZShG6XuooP`npz-LZ)y(`}Vj)@xL z?>x2;-BvG@uHalMxPP|FGN)P~o;`mbNAOR*(PRaH=Xtl=>FS-b9{u?1pH3=RSY3(( z2DA8h$)C1zW07PHGv@g-2VK99-?&w*<5%4Ro&?ih^(2szs#hHZeL$5!tmV%l5o6O&BJ z%o!xCp6MD#X~V0;ulL2}dmu-aFZ42KbRP$@_3u!8O)(z6Sr#Nlcw+mJV|2WA^OJ<7 zJNU7OYPl0V6u(mGMtP}yUXy7^ICUMI88HOtwA<*-2ouXs`4-m&#!ym1AwMu?A#nfY zwfBNvjB~Digt!Ygz$f$U0H#c5u1jC6tNG|r%fBl$K8a=i`~b_?0Z@irYQ^U=YkBr;W>pjK(p@8TkgNiOq)_zdY<&+=FgTl;<9V z--UFLb9%?fY+d@#+Tn0rJvI(U7FOg`(V%!cduGJoAjxIRw)knU5Gty?4_kSdpIOVX zw}9)-gqnDkk599MWs%U4*PW?i#T_f!@MVDROz=ulpB7L)#}lSt8Wu)m%!I^~@fpVO zCzh_vG|z0#^{$ONI(wO&t^GCu*AX)+J2JznA!Im=i21ahM$8?+!V%nD<)n2XKPx0b zb?*o3DgaOePPr%0LJcW;BYq& z&>ak{(}slI=0F+Hc^%+Dvn=Iqyrp0{bF1Fyj?(D9R!TzGWx|tlh1No0;+z1ccphaB zbn*qfnrt~4IKem>&Ijlg2WWIqsG~u;0lKF0gjxpo!y#5@hEd8MlGm#P22mqag!h^# zgU7uAuLz#Nz<`)|4XFtFejh=!GI5<^@r(rmR zg3@!brGhgk{XjVBo18ve=Q<0rEOwCKufO(nh!*#9gB%8EQLYFw_zmV-KN>@9`;tT`A5` zJvmD`0j(~bP!}S_{dgqup58%aKK{zp3kwVWsV`^1YckC5eLZqdv|k6EDEYdsr-1`fpC85gA$xH)e>zat$n=ND`jnLwlR;?M!OCq`nYt;VrABrk+q=ATh_1C9f*g)87|l0%*|&&O zS@??Yr^Ht3U}&Mei;E0#5xqxK89S%r!l$B*(9OdE(q&FXP3|dndz) zFYkkl-1-U=g)JAq#S+h?#0-2p*p+>uJKySn4Oc5f7W;1u-&!skLyo+t#|0C13dQue z8>WkQli+@aKh|retjoZA#300maK`&WZ*iU*fSK3zL~y{+z%GYjGjd|9_eK#&Kg-@_njC3T9sO&u8eYF(qvFu5KEY}f(n32fg&+I|8a)k; zv+HZm_ujHw)Rfkk$5gF-7bm#~R{@{hP**evf_b>2jF~+VYyhsBK?jrX^I|>kfSq!x z2~3}8%3G-yZMju?WL@F9w*!3jj?ZPNG#YNqgsnH@3|rhXp{Phx>1ch=GpI1$R~!cQ zx@U0lO;T6)?N+E95vlZ<0swk7=}C0k{uK}=*ODXytj=428${B>i}MURN4Oy0AKqwJ zva-t!`W?SQZr4cq3voC`#__|YU)%R(t|>;~bPs7e*iuVH@`}PaMSmrjJohpGHokf> z52{%_;o8Vk!3b4jZ^K`PBx%QoGBI?uovG1ClxK*TP)P2kBf~s{BIw} z^XP~fIuq)NDgn?4&zd+G&=Yr&1bU15RJ>O{TIKop0%j=-8;v>D(Pbdu-@ z$|CA4@NstHlmXP(Kej&~lI;q7Ofl!kb=Shw&gc>361gFlSfgC3BM$NZ)t>SpX>^Mo z7xNpj)`$vnZr*`wQj!=of#N@umD+KstNyO8(qiY;0hb1U#EW0IA0So_bfnl=q50At zi`iC#>aY4^^*|ZY6@2-a8+!YW!k5gIoHplQw+!v-t$dv7^@QPQ55i^W*f)`_Xg>vt z?oVnvrb73oK&l<9Rz6&Y=C8@~2NKYIQZ3hW6Ze*-M?J8^b%r(Unm4r z{6e-@;QZLS?&zHZYt|l($2ebDW6@{c!TnSv8mzdaa)QSg>i*0awWy5QyVWy+a0@IZ z2~5$2kqI#_onUbuLAB8%Ik{jG+erypZ#wE5mtx-J7ZCzoKeyIt8vMw`mtB97GntgvBQ&hLZ|kj~rB346GY z?$T>9d4MOWqKIx!s4MqFhehj;`n1^xChWH~aGQ?Q#GJOYnNNt`MXucfcRn$ z4(mV*HI$M=k-0d?A~YH#Fa~CjV|U`TuaE$$#I9SNt1_k~)@!G+Bz9hby93 z!4#82LFg3@5k6~2V|FlyCk&vbO;Su<4eIjc(otPLaxqdwpw&)JzFsC6TrQIZbF;_L zh<4_PHnegxLrWC%X=$j*$M1kEs~xm7jy3%PQUz&hez+xYNtd{!purXlhl&43YpPSc z5fl7xBzFF-7GPL6&Td+VTY?d4FOy!8>g(Wo3ev>~FEC(;Yo3;sx|-exG3dECPn; zV!NxO9vC9MuTs!$W(rllX1Fc{g1`nnP3$<4msAePZPn4tuxW59`|h_vpCcc3{5+G` z9n>g|KMKW%{^Xyen4!S^g#hI=^t+2vz4EoY0H=ixGdM!Ee&kiZ%SIuff7P`&OP6{tZ#u`#_3IE z*upt-+?f)yfXjN+i4H~DsC=pRtpLzf0hyZ9Zq5-D`w;kAidw{>pIY&k|;|{>P z@PdPcrk8E_9AGa}>nXhObO7u#0vBD6fn_mM{JVr4#QonrEC0{-to)nV|J^TvnzH{3 z4O;&J^I+mU_*y#0wAkJnlLp;k_yuws%dizW&O_o2YtvW>fNCQ$pnEZ3Z_gx35DoM)zQ)pHBUv&vN!>=V?WBy zGkbM_R{`-c_y01$06;b1&PyY;^HsxO$C}l}sqE~YTr-w1xv?lgvuh@S02@G`Y_9|T zSYp<*ElRiRb+IadwlN?4j1D0l;5HyFwPtI#BN*~9FAo2`F@?Er_X+^ObutB}NgS@V zQb3nviB@D*`b$#@p0i6cojm8gp{XALrIczjVkGfC6o69EV>OEmqQoiijU+~u@0!5H z+|kF)F5F~+KBT@k32Quo{j23W$uIualYn~!=r@*(^R9{`9l?-k{I8)tpqjOJlU>=_ zoRAJ>DsdX@w{ea>h)(JxdA-WJ`ZxRMi|$mP0H`5RAQDL)@5DK)kUa!6v}@ z=tJ*G8Y`2n{ZUB6G6eMt90ha99dZ!7hmRPjD5o+jxsvJG_vWT+fbJj2DJqyzs4C^@ zEpaqc+1~E(35cHZ&h{!oJOJV#2M&K6AUt$a~AQ^#gcLy-vE3%ydJdUe>8;aUYYAVbo_AlF@zs#gp-tnU(ql3e076n8@w9%h=cH`aKXse z%xg~@^%3RxKDS>R*u@_F#Xf*z=)2ca z4b6q8u|5wtXhPJj>>AQ9)TUoNLgE)QvV7wo-3sGF(}Qvvt-SysaI$*HLSL%cA7=@o zG|nWROmgZDgS&H~wp|3qgh;#)`jZb&9hC5~9WV$vDCr6gA@o!L63%U#)-<<|4y{8L zmZzEl(`XIgX;$i*0;nEy1MG<411UsG+QyRr(qQm1q&N)Re|gguQByca`8NkO5V%tO z-8O7m8e&v`Kbr@FZ99AdE4;qYR4>3rEC3EadQg~;hz68ciPoO3%5bFyNBv1-f8sLRB_JNnklEg~QQIxV3MC`>x!^$B-r70^5!04-ye z|6#@tBx6m&H5IbUI>pAUjy1=;10PVa++-MxV@NUM?V*9FY1H+Ihx8Iy@pTc|No!n6sjED3%R@{DdWQwk~F z(}P>w)s_yhNV@CHt3wD$(EoS0L8 zpxfMnd$~w%lV?%kxoB=n$W%MjBTspX2M%GYRXKy-dNyzT`xS&)2(+@W%0+r`8@>)0 zuv_h)f&ntj6Wp7doC#TUn;hW)!kCcr^`ts8C4#PD9}1;_DSCTt!r+J;Y=1h~MZqzg zs~^k(CQ=6`0x&LEwmD$dyw8m#^?#E`gXGKbWyW=)o*-tr(I48-6BKIM0Du%&iB0E; zdhzO>JPY;_2V;PG1MZCbn_72&eMGrTK>2N>s)-k23d2umNr9;>fl<`eGGng#0~);}*5lMNRWtZ6tAksvGm|Iyx; zht=G@{c0;BLur%{X&%v_da7+Gq*;^FAVae>C`!XNgho@Lanne18fehwu@h;YOQO&y z(kRKf*LPPM_VfPU>%8wd@43$Pd;aPv>$}#v*FCR$eQqd+e|#2d%LT+ZKRa33*N^nq zTS=+;kQM=~7KS}mvIlmdhC6--tqLIQe&qt($s53`<0&POR>1^1Xo5ZjTGoAFpvfc~ z{_hm6#fCsLt>T6xcz6IHc^@|~B}jv)+9%rFL@A zs`74N`*<8eR$Tw5z%z{LicDk2K^f_jZyz7Luskqqpbm1o$CrzYd^#4bG*aTnt}tU!2nN}1-^9X4k`4gUvA^W! zBsWZQog&dV|5nhB(SU=)7vo|I&cKtCz9G)60YO2L*E>jbD_Y-~z>8f79#!BdD9NSBm){{6Gc(l+X>BL`*R3Gj?JA1ab-ppQ|20 z`WU?VK?bUked}L<3EF3Vqv%(7<*AjIK&PcJys)@G*1YeMa>~J8b5-EeTc^Z$@I*sG z^=Lo3obU^by2`{64+s8tK>i`j0V#qL@%lCVE(&zih+TFB+yg})BCUXIRv%qEdng#& zWdl+>(Or(0;*=>@(X{qYyZjbR<(At!F5|a>)E(P(1YkH_1kAtoGa4HJEC8|u!pWTI zIB0OCfIuaBZQ)COT29t0W=6$FZ#2+LHD|y}Q0~f1QkVb!G)3FBMuU+IjrsD}+vWAZ ztMcdTSJL1j0S2yh*&_}3r@-MFfVjhOe*xkS%!tkgn|kJlW#D5(o;tMx0|WG9P5;fd zuKAg!8lmd_?L|J<0euF!Bv=M#?ohx(1CNtM-5EWjARGoQ^<>ZM9lfAv9}0h-uz)LK zi`}*sGtp(^S7H%6wDs+X9ejIFct$v z`;RK?umzu`w`17UcjKj2AmwUrY9GAFsd92D!#1n~8;<8(+YttD#;ZViiK~Dy=DWp* zvDd;C41D$K``%gfE5qJ+qJ9VsQ5ZO2}iC zProrU1Sgy+zRt{dgsL3SNwug2A^<@utM}!WsPp`~iFAht`?HlcRBAwg;Q?}A3%6tk z0dVP{pKk_rP9-J7^c6yl<-?QxPV%WnAbbCTzxapMFebtlW6lJ+T9g4>**~LoQ9EwJ zRq)lK?!3N=eTTReIOXUqPE|GF6bqWI5tG2{lYPT#)Q0}&0WYZ#jDAP}6ix709;TX} zfu32qZoHj8m6SNf82&yMx8_%@PC1Gd7m}O61uC}uUp;I9_PCZp%3*vi))e#(-0WFo zBk}^ZjL+9$cvg&tDv(1yyD*15YrxDZxR1a{=2b!WFq@bGqG7RHe(b8C1n0z>MG|nx zDJY*hd#;!x&OZ7u&2PK?@u?-;^@0T>T^JVMLK z4BnT@gL^MP1@f{n{@tw!-dBZb1(hk`W0eJ<0?Wc}u4DU21{c?4e@TTix^lPu7?AG> zMjb=(q;|UaV336=q(w%KWOxvqeox%19gOtqjAa@^0KqBr=`ZRuF4+l7MinAxRa#-- zYGqvstJnv4DcOGjgJzEM&kz6a8;W^kCcOX}X{vn!BsfIvJFwHOsw;Wu~W5uVQ=?0R zLNeVSq*5wVs;+@4W39p%(1Jb!_FNgSPVpfFP){aXJK2S~tJv2WF;mqOByJRi9&1z& z!D1OfTi2)Yq?P$)w0!N`rBR;TW9u`9UDC%EC zD2AN|YTF2G1xde3_u|wzBrJYDYS>P7D7~aO=)}QzBMSVxFw`Iztoc587(DtG6XSi ztPJCzy459u5Ylu1eX<&Drt%SRaOa|~r&2(rlq#DYGH_*+du1)Eu|S`m$HyZ4@1N!l z|Es#`h2tP9%zy>XVgHOE%W%uPHCsy|$&4jz5N zq6{HNl7Hh`bSGRr3N)0gb?X43>xku~oZS3X6E*-V_A{9^pg%|$g3OukHk)|C%E7wP zf;mDLe?MCH_1XD@D4V}_0hTPg=M@K96;Tps3@~ZY*CBqi{S!aVIo9utsG|UH@I6jm z^ob!zp17q=D4*`H**kS{2%?a)BKgNVQ0BubX&tp-<7+ot{6&2M68^9Nfa*jr`4G23 zE+G=aziw20rlg|M2o87~Q`Si;$As3)3^bEJQojt_OC!+Cb|*DMOFTwXQ79&x_gQ6z z+~|kJOdv|!1Zj0d@qMb0mD1g@B7P?@34$kp;#bA5V`~*Cbehj3eImdltp8?`e_#^I zdHrXD)deC5T)YpCcIjHzo0$I=1-3w8DSk8sM9C&l|L-RdBs@`mb7r0>McE88PuBuW z9UB0ijuv7mT+jpmh%-BfEvERC;<;*Oi(X7Gec?6LGNT3p?gVXT^;MvRoUYv1>bjY` z%I$eFGvc6yyXe{n9zziEG^GFl>FUtk|69=ts7&ds3( zivKrJnGzou>X-wjH@hP3KbJ}%ROk2JmujwD(4I?g!by~T0aQKgAj}VgDo~)N3KkOW zzuqM;z{(#HF+N=32q2&k%MlhqO#ddN_6! zIs_$JUt|~MIjDmtY@5Yd4(ad;QHx>>sB_`OrH~r2hpr_+g#Q4{V&OP2_+UpzU+7M$ z*FbPd*TvLoS%7+7!*J!p5I=NUuuQY0x>rUkQ&6v;)w3xvq&{TuQ@4EVe1fhzq@0Rd zAA{?#9hT720&}s0#>sXz= zGQh#hxF-d$X;en9(1D%q^lgpByfjPThYFfXPg|gpQB2*blQ8ro({@v;Lji}8P~q7R(>-L~%*2)F3s$5$C%G(pI+&@hcLPzyVkCP{3|!7ElU52mu+T!4K>w z{RRZuyqlCN5He-z_3iz~=@4Rn`iZizQ>bSIJn>PFl~TXomU;YuVUd?-BitM8@!WyB zHBlpn3$H{O^hPL-2RvW+Iuup?%7ywo0LQb1ls3WvA}b~7r-UU|CTiT@$jV0E6>s<= z3$D#A*8f8ig!Vg2qH~*k#ad2`Ay8exWH{6nB(4Q-5O5HAcM=RI9-%k1*Hd(@B=_In zC@#V&RRH)Bw3ZJ0sI3uP7#G5RNc**xOkgYsAnYS;P*Zl>52O)Y)uzw~L|I6_8`HU@ zQp!6{UhZ!Z9%Wjq+9(MfL#XY0CH>g>!QZ)V;`mVA;a{ZHLOQYci@!1I`Zf9O*&@Wb zf{C?}PKr~xH22E;6}PtTy84~bOxg9e;rHd=uUV0Ix$5I9@uqwh_jc?S3}f7c*J2E5 zlET@r?z*};Ti|D!nX%iX+X>6h@8ok69HRR*4eUkbYjP`Oqs=^8o^;B0-ty^MgKg6s zPgBuG^DH7N(8otxImSnwyj>i~ZfL_Ab7YCb@cko01(XP5!J_nyNVJ`* z4Ec)KN*nb_gAFECx->K~=qsP;hm6QpT4G7;D}_FvqsokoiSPl-x^_Ux&9Z?rvP=I0;vvsNpiPg~vh>2zRDvQgW_ZT?ZwDiD7t ziQ`aVqEv1GfvSDswxwZ?ICNXH6Y&@KLD;w&LqdxJ9c_U78rge-xZ^1n$A<66zhx?> z#3}O>3*iLAo%eRGo`FG#-FgJy|1hq|>#F;P!>RneE^n6#?(bAIKEN*}i7JCzJ@Hvq z9@@6Z7kkmb!}M)p{jz)&VMa!Yhl>s|4>NJNvhpOc*Cm9y|HEa%0hr(AGIk(@9=gB0 zlVrpb$SHzxUtT}a-JQ1CyteW=XMH=a)P!8pqueh|ff<$!>CFtzxKh>N?qf2y2gEBSGyInQm^rQA@SluTPu@kuJ zM%X%&cY+H{x(_zw34W+XNF&Qe`9`n3Fl;Bt7DP{*RRPY30HA?pQHHW}spi%Ptueo=RK9 z34!rq=<0zA2?!tP)K0B(=t5}U@a#(qIC&>IBHAC6f>=m(1uL_m9r{2D#}kHsur_pX z0>b!utoI5OB!Er{ozW?t3?eyb`{R5r6ic>HgJ=O4tn|Ns0L@={2Bt5*{WyCui-e7R z_sRk<BA;@?NlS?7Z;(Ft>9WaT2@C3j*@X>Z?cpWMY))+51X1ko`^(nw-K9DEU8yeKT#%5Fk{_GjtR;%(ESv5dp8cHHo65Y!DTdl!>a1SZ8JCteK-{+5hHLLGHj}L?VUK)_Cx)C`L<%On8dK~ z{lZje=EQcTVnfH3B6;snBU#UL=L*I|syl;a4NTgWNg4wyJnFS|LBq5(ffd~AmD7CL zOZd#QbkhaAe3ZuCU#J21#FyQX`dy66vKN*&Yieo)pQ|wZYClRFSf6TQY#03*wuXTq!Hz*bSd{*cwlZq7r&J!1oOZba{ zn~$+1CA)3RQD3Jll?6SB#6fxsFP@9W{_QPsP~=Y^2-oi>j?l!69r3iVun5vmHe~UL zU|_53;H#!w{fvc{BG+YYk(c>@NRH+Y(v`;B;L_9+h z0Cm6jLE4uPmPC6%0}=>Msgm1!(f`DW6Z+N9v#Gl566vH}n|PuvSH0!0!F#^1_Ww=*Br$^+c~vNMxn6=g|1;##FXRTClp#V=p_i&kMZJ9BpoD^83d?J|!4 zarjYCNHst13vg*us{1=0IpP3A*H|rJV*Yfw#9_>-?=;$CveVi&lm`A&hVwJy|67&L`(&8tXOEf{C zy`*Z6q8*f&frbCg$@%O#j%PJTnOUJgj>;qUCoNccCZ<;Yh#P^A-jJNQOiywN?#FUT z=kw0Cs@D1E{O_YK4PYb)CltNcnMp71+im`l28`3NHgCcoYX8w=a7p~)#h$Q;2$69I z))>3e#17g_7Kvwfk2kqsS1H@rzfF`C_~aEOl1dhDSR<4syTus$xaPn6c-uT&^ZHp8c zr;N&CXzJ+&hcM?T41D{Rx6L`8$9Pt9G;K)vk{dlh#bQ`p#*a!{A3b{1`oxL&^^0Xx zakMHz=I5PbCpeLEA%w!&=VW~o-G~=#O({~ov0{{fhSg6 z845;rH<3LdsRFqV?>A|n+{H4$spV}C6OzdEsR9u2YA8Hk-f;KsT}f%_dflTPl($v$ z1P?!O*6N%X=iYN0IHUdL(7tW}BN*s3P?Rv~Iat~LfcWL~98%et#S|qmmsBm=4k;3cv~f=9>x@oF)cY z@VRTIoRAd}1}|Z&yGCDYNb=AK^pCN{WZMuqfT20ViCeu*xccb>pMNE5`bDu0ZH z499lIU#hVO5P~xmud+_%VXOu%KqpU_C`Ov?!jnqAQ0E9gS~Y21+7x(cx18py?8zzT zd7kr4jW6+}V?`do+^V2z^EYYNuf1`_$pu%vb!rJBR zN-XfR5dwdq=6P3rB?NxEne(V{FYxm?46cm6`3URX>Y*VUlu8NemY}e2Pw5lcW%L}M zbVWU>IrtE#-du^rsmJ)=EAZ#v=$3#l_VA}mjZ8c=%k6o~Q_&`8u(!<@d}dCS3R~(Y z?E{BJlO6#GnU%mMzI-keDU$#>{7O1zdhIGi3w)QVvCF(a>X%}lzq3m;UGrWU=!4gQ zu}%*|p8|^#X=##HhWPB5;Qh<{vW&|mcVP0mM`G?&nS3#wnzUqkqVFDJF?_-1Fqa7m z_X)+Io1}BHQa0Z4(e8DiSu{_&Fk0&JnrGeOpY!AVi9)UAEze30--sHsJ!BD|Hw zZ+frrGLW*YhhQ92-W~R@qo?Q4>qus$q84fC@ASDr0HpN*w6p3H0F**>T#jPsUDGi% zHO)tKg|*rktXaRg(hB4+P%P~kR&=Efo~f-7M=Pfho@8aUi+w&TVXMb`!{BjeCwTer ziT+H5;L9pK?Dr&YP?s01WM%k4?L<6`OQXgQ$3)FDH}!t+{)^B@ThMvTVB31k#aRLp zEKR)Pn@;rcxxS79Vme>^S(n|hOQbh~r>!iAmFWM^vJyd!Hm7)R<=r7YnaivaO%ln( z1^~~w0$i~(D}oZty$10Yh28h?bDQfHg>|7P;0w-oB@C{FTn-xqbZ06aSRbL}5jaJ` zJ?Q$L<9-UKk|I|0jEZZ3ywg{{n0M+k%7tVBzMg0beC+Nls?P&+20lVvhvqXe)WE4p zKVAm^RtNYnq#4@8Bep1F2HD~Nx9t3(*#Ez>!#X^uJY5U$XTMXd@_$#3fNTG)lK-ue z=WE9Q8SAb!eC%%s45}L&4ZS1L)4f3E#BqfLN1r7W=R9=N`ycr$B*kBW>nh;CtsCJ# zAo(G>19t!W#oft|%ERo(uJ8oyyKU;;jbHx*8m2^P$aTlIOYujYv5*72+$HXwA^g7q zLQ$yLtyDBO7D1~D@>%pl(lG1qRr0ZsuFZnKK5U#k;tojop{R#MoJ?W2x-|eXCTDs~ z>1vqN>Iz=(=E{NJmi+wuHK!X)ZbQC=^S&hqT-XX})p?W?e>qVuZLBiE!MK^!(x0)6 zBRhXm1-p=d4;bp2dyxb*3ZPF~&8Ev-)wxEiXkq|7k3Z6}cy0HM17iDE|Md-TcGRPt zIr`ynIBqxo6I9#=>pAu=8#x^>qgrw zny&Lz_KUuhJ9bp*Stv4%eE{yfdP?$4F36XqYZZsWWTMcyhke2sFp!D6) z!^*8rpNg z`tXysd6Pt+hlc#E0LsIhTaYL*PEAOvgDay%#q7%Iaxw_5^9&@KA3Gk+u&H?G@ilS- z06gB0ZZP$*mjj@)(7r(;@blL~8EBPeCWMm*I{l6W^s}6-mXrB$dP(}M-jnU(lbaE2 ztT4G7&R7kE*4i=UXpk(QiA$BTr`agr&I#*_%Dwkly!@u=%L?zw!6DQLuGOVRy&E;D zsus{`_2JD`c1T|77n+cyZq=(jt_#H7UOuTPl;|oZdrJfb9%&qb$ zc^w->l9uYO#%m}l#ks?SKkQtI4{)u*9G>Gf(Hem3CtrgOJ-+Bic#)tl*RwEsoINbi zf6aJOg2nTqpV-r}_I6H(v!%v5$<IBPd%zi_ns2SA^KO=OdRd4Pw56^X4N)mh34lFNJ*%PxTnw`F>x-FB9EF4_!-0| zr9E}$_5Bn!!*iu994?thu9<5m9MBgxI^kFlfL~0v#IwSho3IJX$`W2ffFaD^V6F@s zW8G;>X|@RtYdGGG)b*S}Tn*vTW(bu#M-4^{0lXfOqx_I%WXQYUlcVBtSzb7mV8wAO zxxVzj{X+cZ_R^j-{dS%`7Qwi811_)EI0C?yrHMpPcjVYtUw^&(DjGgqFj#(oKjPG` zf$^6L{f^zBh!3FVUZI|amrOa15Dj-GI=>hm1-Q~IXrGjTG$aE44qv&7EbC7f`CMMa z3xHSIZ|6;p_pqp+RhW<@<@&|GuTZ zIA^M&sXEab_v>k=r2X4sN;RYk+&9asdfKPO`uJ3-_T))V9WeJ%LcdSdkIv|5gz{%0 znIYAPnt!wxAyjI%O`*`^?&!1YBg_(cA8#Buk^9XZ;CBxuTevh^C&ayk)=18&X92aB zDqa+?G5OS z+4#!}fP59R%1VC^Z2Y|yx>0;pMap`huu=63;JXe>cSd@SG+at$%d$I*x?7EYW?~T3dPKY+Wm?8HP zbw{y##CAEQYKhog=qgKzx`g0%HuFfMx%LWDk8((e_HJk~3;AvJ2QIhJ4mh_!lt;AS ze9c8e!y`kvpoNj7^SA?a0mvoT%)w$DPj-YhlJUD&$sP;u9a_#|a&Ca!edlP9plL4P zV6(UQegO{l*Z+thM>7f=cdu5^jbNs^M_<3f`P@gIje_UZ8P;cfuY?=4YMP~7*DQ>4 z=*Rl9YBUtW^ZPPOtASJ_?F9&Lyc@c4h~TvS*E4N^-l>JvZ{DamzP3o>@B|P}(VQUv zlxho)$MXG;V+<8~ynl~+*MO0^z2WRfCgYglX{q z6M)IGO1MJ{sR#@I^c@zqnFquRvn=)C`l<#%lKlrQF5N2?FDmG?b&q6h3G@V%c#Vd` z&gy9g{X?#zHV{LouUS=PCG_EnN`2h!M<_lGg;nK2zmYyTm|UM#>G=7sWqIo1A%x%g ztG8&<)_uU=EaEAElCvdXz^TwMa6BohFu|jezAG&E!bdF+axkb*X963=4w@Gr9LqjeDNZ*t?-aM*oA>!j$HlQA4pqAY()4%A7RE%H5vcOno}u<1$z# zlK{3jg?_sNG}v_(bE{HjFFQF{|Dzb%61=^4UK6!ttVZo%mGjr1K-RYgplME2oPEw5 zrIiJZXFSB*9-Ho+cK);oKCA&k({&C=0107`8mo(C%+vORM517T*mSLpRTN%{CWd1e z{N5YjH!t^*c{&jI$T*5z+Qk;Xo7j)>U=Tv`L+wbO z91ZzTw+@E3#(6E7(%d+4>k^}z2^gEUimRcj{P7XH2{ZP_vk{# zr6#7mT-p)rNqh*wz2`?*rsi04!_&}J3PY7mr7G$CI7L{b=1+iik$%Dg@Gcb{xV>*2;0r=#39{L|SfA@(~PC3;+l0mLc@ zp~CLoeQ-`~?z0@iX9l<}DJoW(zM(PhT4xb`ILbpXJRLq1CbC*(ys!5%NA}{ z0;KpF=oz;?L7eHv4=}BI=f3nksxaNATXXgUaDnk*;UG|i;+{WS+Dz-md_AjdO$Wn} zqn7bl?V`tKms(lWo3=@`T;JmOz3t^Ycs!w#mQa{~O4SBvhtggoUhwU?3mAo!@{(yY zH}-}*31pl~B;VzgoQKF(bpD;#YblCWhv+K<7|P$@p=(+o;%JvF;{ZbLwvP53*$)jX z#KeA71$#-7*-RmNNvHnmP_k5~A2bsdX+Dp{F5cQiH7Ez_O-E7YMvKRX5xYX$x z44*4gjYx;v@+_o7sxh`#muZ!3d2@a5hPz^gL)GnVKo^givz?I!M(h^yIaUs(p`zBe zOn zR^LU;j1o@#(*}%QL_lI6NkHt>k59J~atpuMld~;)GaKc;K zh4B6%yg7Wm$!spr&XT1ieWm-JZ`}aS@016Qum6{~2&UVs@<8m>t5*@P7o7B}6nw6S zV5t4Rl1ChK-#V3&8qB)r26I*XM^M2H3r&3bmB00^}b2{kvM8$74XdXlAD{^@J!FR~gd&KepbUx`#(*zBc=3jTk z!sE%MprW7oMZ)^-ejV>Gtk!@^7_#@?Z&naTMw=*ipJx!ev)-)j!cGx8-Q{(w@y*Ed49E$IWsQ8v5>oUd?+4 z+_wnx)jXcCAXZGcCg#h&~qQBpdM_)bm9UL7%j(U0^V>R>x8*}MR$&I&E9K8D|;q94@K;;sF=bB7mc|Sgf z*KGV%wQ*Q#1OH5chAw~lTOaL&%fFTcCR^y5sAvBP^g+2xgL2Ql_qupSllTH2rAn0= z?Nzy!P-VswAQZ~I3@PT|=$dOLyRz(!@NPPOI}WM(mZ34!vk6>-m=kGk!*vD;2wS|A zqv@;Z_X;>ja*rzXopA8;e))PIilqGYZlk_ScFwIFdue?koT!NFI?#DV5j7N1_`Egv z?qC?edW)r)7P^nvI(YwVIcQhL_942z1EPVm?!CrsVHhyJnb#Lvv+&7JW|POuWuXy* z`qK3-TH^q#u6^zoIAjOF_34rJ{A+ToqXzZUZ0}_X8wWK1s677k(7;d7v3W`=J~~rv}*=*qU=IdaQSPs$idB$Ak=p^fHaFW-X8O!jYlO*sE;CBOWc>o9YNio zmIR*fM*)MQYO)2(rg#(ArsL(qs0;nj)6v4TF{8$JPNkv@Cu2rXM-LD_Q)RP`D-FtB zf2szwHWD&HDKm+(N@n_Ld-kLv{HK|nmY6^eWpsIu(zLR;L>szJ!mr=%?c=QwqE>V_ zL9no+8$kK(Gd645U6)qstMR;!Itq6~Yi~jQ3BlemKxn@$2KHbGthlRdX_je2(3d8Q zznCueWbe|tc+R3Fw_H`j-{dE1>66#)CtMB9KR?PQ*u5)|Rp^wh5InEpI25+U&z8$; z>;w5pj%u&o{*{?d6+ybqr~oqB3a*#drHeoR7*BoOBH>VWBu)S}7p zCQ}hsiJl0SBo6Y3IjEs-XuW;>E?VCjZ9>@-Ask-l6_&&7lFE?P+4 zhxYXf%88wOmBzIPMpN_q!yrfuZr8MnVM`JvyB@#5CG+mgQ4}!xIYI*)6G1+2i%h`x zAK8yG9Hw@4Y{Jr_E?kgY>4sR-XkaEkq7WXGu47p$WgIvB(P`Q*Ch8DEmUEn5vQ^2d zR+wno{eJ4sar&5=KtmrlmRmnflRu{i$vYWB$SRK0f648PF*K1p-#@RWMj`-$VF`Cr zs?Tf)G_((4-n`mF3eG-aa`S40?BF)+WBd%G%>Ds_Prao9LY$!|rk9AF2k`?q?kmO9 zKg=kjSu54r2w~`W?Jmu^L%iX)lhc@4bN&R{>K=H~zU_k+Q>vO`pE$0=mae>)qP8Xc zPZ-f`;LJm*Uzy?(X82{8mh#crTbTCOxfx&qBClg_&K3%82~8-o5dEhz3)%po^YtsH z1cqWYZwaNlkgQM{hfYg);hl}K2E}!H)XEML(|jtFdyUsA%w#sust9DdWb=>x)E4Li z#GaGaq1Z81Rz?Y-kQ(wll2H}~t$=;5=g+o-ljG zmF_^9`ZnAmYEvX4D~gpMOnY~*cKluV%yFAM1mCxCP$<;#J9!|E00V7&P_;#``Ua&y z9V)VrS&FUnr9h|69ZQ1hOu!mktT2Z9yTZk zaio!)&gRelAl?IU&OuwA+4Uy)n}~CC)sM~ue6xlS>tVOqV!`EexSg~aetVKqxIc?D z7`y3if%|2qta(C%UWUx6{-0(|#998}`9Hz%IHw~vJojE0z?BqVL50Tj#{{{}Z%tPM zPSY`k9m1f6bw`+R5+1@9AKs{6HRsH3!Dn>i<89|`ZF*hi%OLTRyL8Tgj{-37(Ym3# z)Bk8@xxm~aa8#L7wcpc!Ck_LhQi|vn%~BCF^wJ7}NI|c=Yh6UCpE!Er zGBWFUU|ud5;FKJuI2~zC(*qIiS{zIfdb|b$shJ*@WeQw9oV#cdwJ9D#F!@l`NPZ64 zp$gYW8>(=rmH43UvG8bQ!<+%oeaJ^UNUw(K3&cz$qnUPclxcQN63U-m!)JsJ)@0A#@64d|i=Z1?{=0R)m}Z}8_5#>;?9rI^oJqRisGQ`faTZ(@ z!1P#I(Ckme!Mq`_yDjD%)ilYVaU`jvb&fI6d~PYaXwmZh`}geB#mU0pPli&Svu)1w z$GH!{bRwnkHM75LRS1jz=i)99y^t)|IfqA7gU@V}GNJxl<^X8VCkk^0{MVI4B_jd7 zoZ{TZhE-2u(ryJs^4doa*qYDTO&0tzrqmWda2Ha%c0WjzMqMTLe0W;S7`qJMw`TtU zy<@%)_PU9tFk+g9*jcM=hN8_iu7NXYpjmTOs=x?o{nk6P&DXUeSk|PJByJ98Of6v> z6b8OV6{6F8gq^iB5?4lQ+h9!G2P@ueRW@r@uw&>(G|nlHI${hPXk90_xdVb=0D86sNj=__^>cP8a1d1^!&O}zJX>Ww z3e~#%p%VIdP$_kFpd}BqYYy3SdHSQ+_KB+r*XdBh@tg4p3I1XlPyh`Nc}A;Ds8*!o zxitx6e@QB`2l6Pc4y;gKKbM5uOJNGGhq1wPe@!9%F>wh9f#A`O-*k!jF(tmmGn|5c zAh^;8h?uh|?R&Yp*73#a>Vd`CbJf@M($&EyvavZHI7dp#*f|c7tN^#w8kJ{%gBtaN z6D@Oud76Bvr22PtT*V%&qTU)kEMy}x34ZPF814) z3}yQh;pOyH_&aMq@uzbo$0%Ho_*Lwl&m!!E>q`P~4?Myp4=nC*zMZJS{531cL=s znhO1{x4nN5+s`@2>zG;x5l18Z*uW=w+3dYD8akq9C=+ZIehr5f&mzjq?7AWn3JN9@ zPRBdb-L=Vdl~~#A%wKQup|>Xi@jqqG0`+tz($`}TXl5PHl-OSjCry|=|SCfb~3U5oTFOMH`hOm1!zaDcv+&z!|` z$XKBtI=C+!>BQc(ik)m|7Z~cHJUnyfjtQ_0tAwAEnubO^WGI_nC!6F;rx@pI#4Y>= z{6@=zFlM)1v7?XcxpKDn-BojRIesH9n}I_)nnx=Oy?W|rSUaq}pQ^f%&>{TzaDQN6 z06oCJKC-`K?uv;k2u4c+ERwS5fv@8$Ag7*j>(m4W7rlU~@FukFgKy~B;ca7utMzbC z9l7)Fya>9I7JUvU4NuD8osg;ulbB!I*wu~Y^Ei0QkZ@$|5+L;ty=N158RL2~eb)A(Cr#%MZ$N>m}BOnh|SeWLZpBj`gQE1cve18)Kw*OXZX z$9?Yrku3d ziUr=9sc$|KWBO@?s_G2oVN7BrKkV-4=+GB2T1A`-KLDlaO-lpvhVyTW2BlS)RqyCk z?fz>4gMk1G3spAth)f14K~mY$&s=-LaYg}PKMlnjx|l#pmpDng_p zX);d@q=^2{yY^b!xX+aP^ZVa>?&s68)_ULfnWy)8*FK`9sWx-k{ArUWO`5rJ z!+M=blbE?CO`1BLi4pG9HE7(2f9O1Q)RZS(d9tu|(j=v5FO{ucE<0@XOk({qad8WGcM)p`3p)=B7f%sa8!xy8&n@h1T=5q)-25Dzoh_D$t4fK8 z!XtsLVq(iAl;KLw{JT?Ddb_lwAY83=cXzfiw6Wam0Q0CyNyvyuNTVT}H|S}o zFB4w}*Uk=3Ht>&{jg^xd8nVvb!_5WmsESF*iio2B;HIvHt%Zlfh$#?EJ~ken4sNa# zu*K2KG8AAvtt^~vD7PgjcP!mJtZh6fw_rZ;WfCgOMAyO`=>HPyaD3qpR|{Ahqk|Tp zEG$MD?qE&1r*F7UN?F!bR!iAcU);}bqn(t|wz0DjbGqzQ^RRHY*Ko78afXSlcT%Ph z6BCoAOlrN0az{o22hz@iSO;Gg-OzRL!vPY*02ddv^LDVd@x*7NOy%X~=IrI*KKiDW zo2#ph70xt#Mhg!QH{a2(*}6I7*x~QE1LC7*M>lkbNjl;+LkDXwd&;}AQaE_%?F}{# zcJ{=|VxssH7YpKL{D!B!g|(Y6WjuCkBv&YOB&g}(<_7bPVA{yX#7iza*V#BDViR=4 z*8#ct*T+)P^q@-q|7AME%_^3g4BYkfz1PX=+IhNemEPeYj5FTTYZoD}Hr7CE{7*Lz zFMBsTH&+X1mEpU}9&X;Q);5S`Ai2uJ@2I=Exx-DQijFp3Ub}F?vGDeCgIo4qF3$K< zipZdaD3qT7QT&#to41D*{?>%>~K_dr`mtK4|*4f9aV2K3*{V;Atz z!^YXd%fV+fpcc6N+EL#_#^6_Y4Da9~@-Iu$glOUD`&c-8;~?ru!Nj*cWWYL~H^Z=Ve2f4$6e?7Q+&WLGb&?Oh{OKMsZ18q;XRu zi~s36tk+^_JR-3D2%VP3@BNKfL74vks49VF@YuB#1YcA{3b=>;xefhE{IMPpdiXCy zX7m?@qW=~Z#8gXM^bb`~60P|sB;{|ef)Y6XBXm&HjY8fLiiaqzZeaOE&CJld6V3l`N#H}m;WK@r2mxxtb`N=Vq8kZ{V$jVOm%*R zzhlJdmu2U_V-g4_Gr}bNznYzj|5>X(fkpVwhI}{xcw$V%1OLQa6?ecRq;0L0m$wC~ zi2q*nM@ed|9eju+WJEMbc?L7UOKyihPgN3=X_>|8BetsvL^yDNscQ<6t9R#hA4vIsx3*_Uc#+W_I2@5NJ0L% zzDg{HRavR|-0y?=7Zl^i*JK5@V({bjEUY{WJ34Sr3z|vBDF1(W8%XB&ieP_!8whix zzfXCBTN9_=qEe!xZ4OS=UrwfeW$|GU#z?2I1nxkI_zVwm2;uxE2Hs*IXcNfVKc{r} z%a%_J%Q1dM0RC*I{ws6S6fQ|A=ogcal#~^hk&*>hipv%*YEnce5+Q5^Z|$oh$HVI8G<_r_siNu&srQgEQcdcCa_t zIQ!V3mKZFQ5yhU5*xwe%_qXu}l-?^e&ehGs1ln2ikWnU~*`zvGDSS%|^=n zG~ctbK;5hpPUVX?qM?V9V9~KfoT25>#zPq2gd9K8&BNW^!WEwq8lh0TGVDuw3FAFL zXfkRW5b?|b8ZA-RGkPdWpo)$1@_@-~0ZC#dSJWj0uR%EjYp_K_@MV0V_joj5)VYLx zZ7iJ}yo9kiJ-s~KoNR>gg2<>fu~tEB4!pBqB;ByV);3TTf?i2CS79%E2P-F68&4cr z2UiC#2Z9%4-^E1$E%v+bS-Cm8L4(w|_iUZrEN}w3x)C(6cJOp}w!kIG!PVIT9*B!B zc5rcrZdDKhj8cTi5jDpNE0tgkiJfX!OHZ_6O^7h$ArvX5G*l3`u`1<+g6&~tV=auK zA#7po=#9GF5jMk6|4;U@7C4{@CM5rpeL|mT$-brAPGO>hohnLyDCb%pZk867L?eb4 zA-Mmp3OkbGbrfY?Z*601;f-a)&ZzfM*}}@n4l4{3FW|yIs&fPNu~L$mzqU*-F{b_~ zLDb&=s31&jNG3w1|9+8P9FH7_9qK=m6ZP3RVBRA_lIVM zQhxdq()Y_HOJtP(^F74k<1^yXX60{Hul{Q)H54iS6TBU^ED{o<>v;cB!!qI0|75d5 zIQtR4ZX-JXVg;xDJRwuVw^GI=TqF8FMu)H{M-%fq$)*{WmVxu$mCTuZ zA@InwQeLt7@#`fsrYY|0m`^9gIHh&(;FX_!d+Mtm`wyPZ_t#{X`r1=q_x;Be8Nu!| zo~q0F`T0+(22P@*XJC>vGcSvh5**r=<7#S@@8iF|uysnJktdV1AmD-O3}H_2x%@CJ_^e>yytVNO zo6p!jNOG7ER)M2M)|(tIuC5Y0Sv#GH>)EY)92*cC8XSj@Hf=t=a@$efHK!k_dGeG- z|M1$XKu2FSWr;J};`kSxX_<1;WtqwZ&Ny#yU^O^dL>>F*nh&BywT#K%HmrgF+4!eh zRR*!IUhB@jD>vy=Zb$$9z`#V1LzK^3J4cvu%mq$KuWHie(%9Y+4*VD zmY7J2iD?gb8}q_cjEhBeOGu|?eStbFletIyIX z?9Oz~h}IIx?&k+~=bfakj9+MNCqE2-y21a9oel8Fh*XqZ*_g8v_{EZZlGI_umd}|T zb+o}do=r3dqD{;chLAQBQzt$|8&#%zm5RX}!j4zd~ncncwS*lQV22cTBiE zON1J!sDzKAWR9Z!! z?5Dh25CxD_F~}d?xPY5mV~2AB?jy$Wqird#xn7c^(R4IgQaR2uX$DWm0&S1rrKC7| z`j`h7)65Fwi#1i}Ps~l0`Ed!&_ z8?JNY&uwP`fa!C#X4+~5^?7E=E)ovgST!*n14*X9kv5%;?U>@-XIg2LC6H#CFKyg? z@80_P+}tjuwZh>roelfpK98HR@x{C4UK0q3MkBM&FnbJ@v=)QHOk;nnhp7_@#Anwm zB)4ne_UC8!LMDbJZZB|jFtd6tHXBz^QAtVnPmP9$55LLmmBNUzFR;&ZUW)mw2~~tw zlnJ@Z=ng0jsyJ}3yWxqYE0<~#B!#rDSL{(ilR|x>0g&OUJl+iv4tab z1Fi)N+Q`_b8TtV0d5;Ci4&+X`8SAN=dklLpPHu&kgWRV77Ab?5XuUCcvH(!v~yt*XASQ3rZ+XhPt8J53dW=YNjvc{x3xeSjS zfu;R*WO>@%deKv@wRAk!R&15j1*5?V2!;w|s*KfhQ|1?umsW^y&IKaeLotv@7{$z! zBN>1=CVvy+5GWQ*ua-3bu(JF$&rC%`Zqhbjp@@@BY49X+;`E(LGK0+50_bh#ek=?P zPg~NoZmf+=&LC+_R@E0dk_!j}FLIFt>>q<0&{3Au5(6a;VlVZ=Ob0-C4&4ve zEgHx@>n*kdgJ3LtEG?!0pzk`2f0M}Pr-AP3x$XnaJrv&rBuz7!5hOr4DyCLYs$jC413<`fp*cBxNALf1xc+&hqXks2Dp94G5 z>u7r)k*mTThAkMHv~>*~eF8hkiHn0eGH#i|HVl&O6>C$&3ZbT*Gfj!#(tZWTqHONGu02BOioe^&h7eeP(76 zNP}Sxo2@J-!VTyi4@@{3=*o{t=Uoo4CokUJhj5&<284Q#<`o)-p$+9mS8Z+85ld!r zex%EPJ3Fv(F0Ymm410R|{yiF0Q3EC{qN}HvwS&T-3)hN6)d1u@pbOv6nv8Ks#4OK{ zE|`^-Mj(-xFY;!x&t0BLk0C40f?)~5p+{&?oYT$OHU8>@ng^(JrwHuNl93tVfHs**v!3?R>n>uKSgf(y5w>iL8&6- zJiSaUK4=*4e?swE1kEKu(OY-W;{wBwHy2p1u*wj`)qE@P;_eA)C@mm(#!_g8aa>Yh zX)2mL4$u44xYO@76QB-9l|?UkMg zjf^=6N$6$sL`wD!E_Mde%Ku)FwTTb2k!8;6`$vn8*D^3L?1tHP(6XQk_$GkZ$y)<`H0 zye!u|w8+oiuguBpr1VSclS7(ubOop6%)Lq$AG)~W%8ezdcPiC*VhpaBy!2KHQ~dhE zt1GwJwvT(y$J?!+AFt}pn=Mnl4lLaw@2qJP$iq5Lu2TMl^Q-O#_Huw~`}(JaA*whE zfNJjz@+(*DyI7^EAngA&Ggf1^Vvp7pO__SHDzCOnNr~@|Dy@0mE7Bhm7E%9lK$jAJmjKoEMhZrL>!A0?9^lbp;HKciLrdWQRyD zVC}oovY?1)-=C>8km>bI-!NN2zU#*Or(zElp8WdCv%FXSXiV6Zo1Y(LIls(c`NB|o^*#UEs=6`vBadh!H}Mf@}dBU*d9<_>}dTe(30YtV`-8xAe>*pPoR zvwV$0L8Gu+WKLDIYo4I+wv46qdw&kT%&vTJ=2lO_u5M2<-_zBT%UC}KxN4@Qp-gT3 zIvQYb9gT|9u;4>z6>!CE+3MbE-KyQMznr^#($u$J%D;Pe!{hq5tx2)xUJrKp+C6v6 zZ9LsN(C{ctvpkxdC{i5$Q0A=dCCrom=C7dR7|m41HreBJl${n4b&BDqT!IMwHQx00tW+JJ@0oLY~1Ewbpo^L3^JqrF>0;{Chw1x5AfUt3uT76qn`H`xw_+ z9OBLhxc<^M+N-bL*Y4-8-nM70u|ut~PknEFiOKTmtnJy)zj|?J#$%C}+3Ba1hTbck z+aX-f7!m-|obY&6;cC;pgT1YVL+yo6XUe=f*311R;O+G75oVu)410uMuq$X6Uoof* zLnUaB#z(Kp*R|iVI;togv%4$z<*)iJf)~@T(PdD<% zNWQlz9O}JVb@E)##SY){o%bbl7&eLAc=OvzrY|q7j+X5#pJ#XW_)y!`A)$eK&cbiS zJ+;qoNe?#36mILT1@egccUbXkcH3Q7S2JMB5OnU#*=Fm#g|`_OTvX%nmyFr-`C%&` z&-aqckB?;E+qo*>Q}FZr#OOfn{TWNFtENenNt!cwSaz{;v}ey;Vczq2`;d5ax6K(- z>F)6DgKdU(x3dg7?efm^XwFvp;qKR86?@K`XYGaE-32lQ?WWzEKF!Fky`U1z5Ym+P zqI=cHfG+yS3FDOoGqQr9l$%aHwx(1tHmx!5`^vWG-P~CFseVgVdv<0&f5H{CtwkVx zpgdrp^9=CkmJcWF5g1-q(AY0)*&OBO-){U;WNjKpe#1h~QnB`#^A=xVqF1T2j}Ogw z|H|y+Z=BZT^Zs97^4Dz?{7u%-)caZZCri6{rmUMCHme@{%nNer1H0mdEaz4KPN&wP zfo`{>-tB2QiJdtO0gqmM8i;U8HQo94K&$uF-KJlydf4-sT{0dnZ_R7J^4LehF*|8> zqhM=W>Qys0yNfsJ;$cU?x%sR?@HL%|(yZPZV>{2n#?!_bF%d>-Jap+6Q$m)s^nNej zdrPqoj6dJ*<>v-F?Ci?dchvMW{2XlPY~b;#So>0QLt4KtNAh;fqpLq}FnD0kDjD=s z%Ojd>L||r?1%+6tRvsLF6_l0_i6i*h29Zo{B?yuzNSll4En#{D5zTkK}`5 z$fre2qrz5yE@8|6*i-wuza}RU*}nGWcNVa=?0UG_Ogy3PMWldFkldG2i8{~2Rz=tA zed+|C7w^z#WpT_d4(3ck22vqWn0{Rv$Ah@yZlnC6n?D*?mERF61Pf7h%Fry5Te!xN z_35tG&D*1<_*~%L`{PrqZft*b_>R3GEVd7PvW*tTvild`*s;{eF4(;aOv7oA!R^+2 z2RZkC;?(_3@{{i_de!YipU-5*mtQE3`q|_*w0y94`_Ep##>2;+?0CR^W)b+z4qw(g zN30%l)E;CN4U)U{`P9|YM6TOuFNy_}?W&9yu9aD{<{Y;Tr*O^ub1`YLnyh+tJKu+9 zc3I_Il(oE?J=uWa!jf0)4-z|Xet9yz(dCstR9aT#PQj9hv9yg7178&?n2pt2p_Yd{ z;)xyAh6}$-6?B4FZrj>)m@h0MW8aq7tnUTVUmS1ZSNguu=kZjfZ?}7H^%c0?uj6~g zs_dLR_ksl3o~waPJdt7!|h=7MZPgyH)FF^ z--Q>&$(b6nxr)2NA`V#<#2hi|+XIHL$=a!BOjlE zZI)g!0P8>O;L5fKwe;-=!w{LXzTH$NBv57~CnNbRi??B5L<_@YZGtXSdGy}4Bi4d0(k*)2V&+TEO}L|3zALc!}tI|(`04X~hF z3gXRtr31*{iaF~?SfpE(+2{qt~ zosy=8%#wW**ui!EWo?i7v!1|;J#_Kq{#~A}y;t}8*ZA)@xpT$a>$$qseb-H_AyE+# zB?0x1MO5!HaTSZKKV#ytQTn-ls)0?uO<9fQ^Aw&RZ>Dbhc3I-b+~>3JT>A38);8KD z<9)5XZ`q^Xiu`0Ks&rV4i_%e)G-fj;82tm)li%+ZQx^QyqBz zWlD%yU*7Wsa5Vz?YwpOf-if@*?Ome|d)uB{O?7f9b>F2w`E@rhH>i)otn*lNGkUv56-onrg%LTgJT~XO`EB~s-sjt6 zmw2(!oqbsgtc!kkI2)i|-M9XoY}N}GJHJ9svz-?G5Hz3c{nEU5s97*H}gG`+)o8t_k7l{U$p0w<{m}_*IFRUks zBfmQ z`qw{)`jQN9zGH9i)RBDtW9r-L?|V%BS`%6;*3IrbUD3-8G4diBJ9){ekY1Uta5BZ5 zm@$~ysoQILhCUu@-9FT~U1y;J$J$8Q3?cf}F`Z{UpKNs7^L~L{jQW1ErCRI_-_F|X z_Ijnk`>W~VBR22%4QN=dyNR!pMg}S@!u?;Cl&O)R#5!@aLYMwZp3%nTM%<3`9plRHp~@L zg}zLF8kNyDA;|TgKQb~*r`Ij)<51v?hbo2wDq?Ma%Jlf`uBOP=ld(Un{XQI5%V2iC zcAK#kG9~pox5B}2=6YB_Ff5}@`r`?HbFVs=oCKuZ27yoP_a~$HN(~13ztgK4yJYM8y&+HKFnN)DJ1ysRUfjR{pc%Jq;YQBR zH~VI5p4lGoJ;akaUe>YeYn@x{r@jR)S=#+oN`qA^Z+v}Zc*~=}i4pi+y8fCg#IZ)l z;-dqg4A!ynmo@vwQ!~#psn)>Bv+TFL`EYWJW>Rc`-#{~jiXBXusaNuZ2J$teUoM>< zeDmXhRi(deQ0Ta}uGTVGxb}IhU)!Zt#qM2OzIsLimXJ&a#lLNGNb1 z^;ah)BAh!PV*>44?``Nam#RLyrr8JigN9V7l;pGSkrd-Op0ZPKo4q zWLz0I6942DxMI6L?p+^_ql`84>!-SgA?XEtVR!T21+93Bl<r zpqk;d0)4ZRhtiv}u> z287qkD0)d7+NGb=TX>iA){dpy7p6YX{Za+05FIAh6yd|U&wY+)Tiz?r*KY=2ZQoKr zqu#X(s(_Dlr%OV3}>P)Vb;^S zWv~2Td*P5%9)zV;o45Br?G1~2RQ2WCx1C109yUI$ClKHNIKXMQ-q_ zys^K|@!*m$+N?Cr5y`&@ylYixEzL?~dUj7lwbHUaeF$-^=XKvL8UJs0d7g+gK%#(U zXi04TxlJGmwtf)Wz3OTY-u}Zf$Ta2i^(UsO@6NrT6ZBCwvJZ|qW~p}3z!(ZDeqEli z(Sx!G%L4|EO#?S$2ih;mOps=SYziC&1H8-8VmT(Kyl0^Lmm6VwGx-bGRTPoG>=H$} zRM#vX2t9ia?frJBlcf^EqMFx(%jLiCQ~I{g>jAf~^`Jirq?FfXX3N^9NtEa`1y-`E zEj%uA^W&Xu&*xZNit(uxES|Nx^3L}(H-9J0^fW-g{1j!?Uwe0jSr6QoxpjO9M1B>D z0Uh3l$~-Z>8z~YJb-n4eMv-7mHiXUzP!SSn_qkf|byNE3Cmd?$ig)aQpm4ShLq(Ik z_*L%)%zVS}jt5K&i#^$0STv#hswm%H-|;UC?Ryo-^vkO-k6UiP3c~rP31`fruD|kl zY!aht-_(&=?z(+C!R;6drz~1T}L1IM$O$F>1ug> zuXooK?tS^#?DVa#*;mti{p||e@;k4utPnSDzbSo3cgMXN^-ijvGE{=Ry2=wW&9CK3 zgRNVtIEKl_@D68CgY~8Gf0XYwsM_=69AaU+f$k%M}SnuwfDg4p+ zvP-AZ@=P*>I&L-I_r+{i?Yuob9RKKIgRj94zLc2n0yRL#si8(O zU+Xp&*`ME1R@%2qsqeY28x2dh2rS*_>O1OK>53!*uFX@d%ul;xC6;Nclq@$|sMMP% zQy|D4@a?U{LEnHc(GAwO&&Ymme;2&UwQt~{%+1AVThlE!ZT(OxQLfH^Zl`5_0~nDt zieDnj&x)U6x`+~1L1o3qr-e5so%PuCb%03@)XgtBVFq2ihD^O=-7*0mG5I%vlM^rV zD{EifRpx}u_u0Y0<7$f{4bqIH8lxbDV0yjM?B)&shW710``OfLi#2s)I_;DOwgqvY zGP1aASkR{53MG~H*7JwCBQC_g&`)`WDrs4JzJGSMOF$O7v+$_7_Wlx7jY`^Tm|AxI zv2G9d+azPx+-3QaR|e_|2kYkfKFRwzG?3XWy7GdGeLFH9;n}@F$K)h6yMpL~e^pVN zZnnh!{2rUhtF~}g&C4Yl<~-0XR)QkNDjkM3?~+qwn^H1WzAf`f-357wY-9J{?~k`9 zHiOS>MLE^iKInGZ#DCUZ+2`If;O&_y7On-oH66a2&xn-yiOfWz50=IY+T_SXD{!7jFaSNAloXf2vN`|M|wbnaMp zBLEDMQ^SI{7f^Uq*D2_u8Wdl$q`DU(3)|;l^;n0ZZq+?iK6mUzcI&mfvmUoe78Gl0 zrQg+E=vgyO$J=afd#GL8r7iw#A?qIOaNSXMHJ~@Oqi=!pwY5H3#?=8(HQfkRhb=x? zW+C;{Zn;2BY};X>%Skx_^2}Vq9~>eF7j-ue`L$u)EtvR?3cVlkLFjePD1y=_RI&^T zfFg;zRt0g}Jdtm|+;jBjx1*hWRZaO0KL`7^bnkqikP%}e-(In+p<%GT!RCRB#_Uzu zSB~%%SlXR!;rCh5w)%65w?>B}+w=7+))fiRE3+P+cJ*bg>H?SWHJ^_ky_f?@eY@!i z8FMz7`yyJbwm~uRO^&QDa$7D0CoU|O-agng;NKzb^+c}uV$~KU#d@gId@((D9Xxms zdG;ROx>NSII4sUH9aFJ)xrOc~9EH-91BwZHp9aV=bCnr}b_H|~zKS`MI$@eYijzls zf|pK3V#R8*IykhUHB9LXG&@uwY>`RuZm&s@rC%`=!uJ?eDmr3iRfT^evHmEj8X9c) zsW|<^dLVP&dLOSbb7vSFRu%KZ%TqhA2bCp9;r=1ZSVLGigu;-LC6HO;Ua{JqP4 zt5-%b-fI@jojX_1=XKYHCTcU#Zbxww^k{wy zG#!}3d5c6Be`m8fJv;12U$K9y9z9BUsWUXOvxr8WDM4FVP>wf&Nb*eTBJZ=If!f@u zjv9AaJ_am0m?1`eebhx9>Ofjj9l8;Vc>Gu()1JCIr)dt0=%eBfOX%BS+iOkvOCij&RbOc9E+Nm&K}r1?b( zPaNBQfM$MJcCAHrwrjMlQ*Tul)MPl+33pCU~#Uvz`ie&c6e)0 zc{%Fh0Mwt@AGx_>vsM&9JXjx{i%|MAq6CU#9ti3GbQOs6gCDm*(E{5J6UcFxMT+k(WGj=!8B{LvRGM_efJVhTw zSs;cbUu&Z@lFZOFe|Tb<101k>IVsifs%9|fsztBN{jjz~6Z%hPVQ{)&=c)$-+xRLL z(ivZsPO34W?C_3M$qGG~GST2QcDg1m1LCh!!ZAmnE^kFXwZbqF4m~&@u(GmpC=Ozm zGD}OKQ4Ae*@aoa@ds92U9pe^WWB%ZfGk)9x;=!7R78+=61sHDkw^4crdrZyYl!^nK zjM`QR*}2B2Ol*6wc$(KirGn_-v;&niQof%6{GiM;K8#je$wg68PC5%ua=3VTNouI8 z&+u9ugQCu&AM~gc;Pm`TnMhXts~=`p#+gw{)a8h}+EUGIDP3(DvtN8#od3+=s#y%j zDl_$iNRpG%o`t_5)9iAEG;d4y5J0Uu3`eUqSa2;yAH$Q0?f7oM zj&@lxEHPp0ZnRg9Ev_y~{uP4HSReN#BCE2pvSZ(MtEY_0X}TbarGlb^FCHB^a)hnh zhDJ?+sd^*We$8iD97uHS!%59z4q+c0^jd{aPa@A@_UbaTR?D5keqV2-ybYV@xU0Y~ z|HkTh6n>oxGz~du=du*7y;Rh{?}F@=r>-97*wNhMiI9w%X;IRLp28WQC#E4CXDcHz zkmp#&dhMji#%UJwCmhKu>W-9X?gEMP&@TyLCF^y#~!^= z6P>4b8b*})iz{;X-G{TRK}TPGO2XuHA8W*-u7#BIS6rp$ew2iHJSfXhinNwhow8Tf zvxJ2wkI$NIuWGP+C7kVnqe)mlSQ86NN6*d)2mmA*ecaJ#+6#^N(>B=x&}J=AF;66@ zRc?OTeiiX$6BDon^?EjJ^rm#g2eEF2a{hcC9=F-Iyjq{4PC3ivR(deBT8sB}`ll-{ zQg1KJ=XiNyB3?k=qe@4|H=7Y-Wb$zmLlEqOyf!<2Ogl8lNsxgFde$k27M=T+(2^y| zwIEB(`qxzvxb`8Wg{L^^;-7F!tE#DG9({w;wx0RkZmByePkTu36WBgCKlE^PKX{gE zXc!Q~F|3KBLpncJX!?hSSPGuX^qJh>tBjPLp#N!M_8rwvNpG|d1UaXGmI_tWJ`^5u z9`u9~)0UTS)4nt@aTFi7e@$LQ;}`<(u@iyElPIf6s`IY7eDvtiBq^sFN$1AXU}eeQ zrz|`qU3h%O#6lLg3z%A#n#+#W4#;rgS``keaK{FzzR|efgJNNMNlOh^aTC1&V+eypngk%xf0{}7G5b8Q5LL00 z+9}1K3D8P`&|2?+P$b3U!VHPPAt~X@HZ-Z14du2JeK}#weFFSi!my^}v&YU}HRHL8<~MqlOr^D$!ViS8ED@ zC8YOoAwe4{sD7q6SO^D?oR*3vTqK_YxYHgvh~YQMqeGZ6q($pWqJ%NhJI9fLN)Sp- zo3t5Ej`z64ol-cd_<@cZEafz2TOk zM!Z2j3xXwg-h{^C3_w>v3&0@lSx0~oB?8G#4}VcVLA;T4La301P(V&O1AuWL`4SLM zW4hC}v2POmA*BIr9Q8GW3EcHI!9$|im7&wV+1iXKiahZcS za+7=z){<(oK23A2Oe7tqds7MB8&$xVkpH;=orhTNVd9}V3(#zz)-!%87-KKu3gL~B zJpm+6xM6CZJ8_>nf>rAn0KWF)L26L5VPjE=E-g<2Y(WR0zdKnU#1SReB1&A(;-JaN z>J30t)#5BRLL#35UMU7*U#LqGf=JLf3Y{9emtu?}I3(h6O96{zg8Lna7sksyL{*}n zhodSsba0oefJ{qP(7|6I?udiGaSe!c9M32R+{Z$MK`|;Dfj3B+pw4Htyl*;-CrIi% zIIxX!lzRNw(O^6(|I5BF=1*mci zOz|$3VScc!5S{mP@2+%RQTb8p>5%{-dKGaMLI_N!w9OO!d?fKod8#*S^2I6O?dnQ<8NBg(J#8#CSBk)&T@AkrB^v13-E_Q$nAmnusZ7#h6AV!{At#E@i z0)pGnvAT&zRqTZ~hN9b!6L0V)fi!6foT2HGA&Uht@bgQ^FeZY%mckBJPrL*vZW@54 zpL{}&pp7Hi*QmN<#6b%xtrY;3(Wxp+f;P`k9AHpTdYS-ptf#e41h6IAU7z7Fkaa)_ zPFoW@Cicr5~aS;=H-)}|E!Gg$O6PTcd3ARVoAQJX!@}J~Tg)_Vc~FGcbhVeqQ>Vr~?t5w}8qL^cf5f_IpYoS_eIP(IW^@tG|^>(8sHu#UU<1*e>R4EtuJ@)s$6f@9?S zOfx!u>T_>>5GtrKms9BPl5{6u51eeAVycvh*Gk$}fC4ns8WB~`xE^3`!m8uUR9PHH z5GYF@imRp{(m0@V=D`Py6{K<)r8u{3d^}6s8NH7LbDB$oUX@dJK_*V>B+@rY^cktvErGp+og+v>)g^uOu{Y&Yr9|qh zr3=K1;;6|O_a;FbBT&a@DQ7u|?}jCxhBdR>Y_3!19CJaE01Fef%{P^RDjYQ)ve}3- z6G}2lF%f@?Q0a0J3-O$Mi|I0azYl#WLKwb22%oBKZ1Y@97Z1B27h$g*K2&<#X$B{& zeHNyEmGm4F9-O74_XlL#CSMSt;ugSv4ZdKo@`r@HeAEj2YB=-_AC2YJ-5&vmkq%Ne z)+Z1b5LzaHZyiaRa2Iic^sGbMvh1~Y=YE+I`k5g_exFKO7(c(2F$b*;EhPF}Fw0#n zX+rqp*3ze$hRB~;ZCqi0C9qQc_Jx+JE|r;n5i3vO{2iCQjxFqy`9Q%ritE@o!6y*Ct1>+eZ<;o%DPnQgL2PN7aDs}!2?n|B-=)$KP2-gQlYO{#_r>46BHTn z(OMxOxME$ZM2!?anz=on>t#VQ>(StG-vFedybLN>LJk({#D0R-RB}M4rFGv0`PDOP z#_@+BgdTXCg}K8goJNMCxX|dj(MipP^xX9^qwgzf7EEoT?F|G&ex;%~axfQPC8Ryx`;Y!Hj0} zCr_vYm_;AeTaNAwvb6Yc^5t{(t~2Es0~oOqWRjyk+Xplv%_6&Eh|SVFLwm7U3L{*lc!keGzzC7j`Y`w1o+gYZp zCl#$bG2^!naZDR-F#Yyn;nEFrSe6PUCoiR=d$^WOW#hDIEo;RO(`hZ;pgBj0?byY@ z!)66nOdZ~GJpRP-w{}NHL1j#E~?ptzE7aDbO zY&%5ug{PvZH>XU)zJk?X<`$sqqFs_qGSQ$NaqhD=BSJ6*I|41QT}D-$>s* z3kd99i9SYT`TPRafU`u4)6RR=;INo@BcG-tewB!ex? zSVR?tTA^e7CVH)b$Fz-;l;2ZnIrf5_2=GQ(dQ!h(O;90>B`HJ*Gk&#} zv)~(=FV`Qz5f{~>FWVn1^dT`kgy;nzDXUJ-)|)}I7%z|uQ#|C~&_}>rAW*UCMzXls z;&`*OyZnh@XdK7-(FYcJ?4Css<4oBEmH?`^j%jo9KTV_IIswjZsOW7Q?twhO%r;LX z#cNYZ^0n4V+IBwg`k~tfAwavo~B!XmDU?VrXxTE8vVU&m8 z$7itL1UwLw<|MS3(Ks&p?S-J5$12UQu8k)g>p0fQqgHD^VT1cSnVN!{Yc` zTXh(z$->y>!cAbc4|6Z>CWtA{94=_0w}Vw&sA^gzCrw?%_7-;-gc3qp!AKTA7%X&` zP2BPBekaQ3o5#Tjtoe8ac03ZOP-q#(m4XH~9|&pFj31VtvAh_oO6Kkiev3`?B4;j+ zzrF2@D$RXXFHOMCv!;s>uogM8*)&dl_X8n*u92-+SNcRiYI%TMp4f7%i{L@tvWj5f z!Kq`~IR>EvG-IYgw6T3O(@~uuCdU*%ITO8kdB?PagNqu+Vns;UB(g7fyOqOkYcYf} zA}KRAl$5-qsHz;p^2G4_jmb&VsAJGYPhjBo&wl17ypEH$fCBb@$ZEwtM-4c|Dbf$J ziKH+w&&#$;v!@xub1ei+Z$~=u5rJrkV8v0;z`z3`D+6dTKalhV3~5OXYyK5-1YlIZpzcL67cm?;KgU-OGFW{WdV8Z zEKdF9;9A{IXOanViJB7;p)>z^^(hfT;qWK>(1y!SPuF`Cu!;hwGjR_3(lO1O`7QYN zM*@SfFq2JyqPY=z#AH2(0^KVwei5*)=5eZ{amW#3*{={>WtST~#-}6eFf(z6uHjZz zO)yH4AK_t04;dB;Zp~n>npjXqq+jm?3iGfQs{ zggh)#&n9C-$rki?RKB0te@r`Ig%nNvZbLP0{@Gk<9SQ*Elcwwso*(PAIR1W_2~FeF zgW9_GmS3&J^>Fe^UX$t{AEwWfa%-lXj-(or;fuJR;2!$jG((BTnq&KejkYr=tEMe* zr16;>=;WRadN35B7EiciJ%BIBc!siS@is@o#tpB5o9Y%*81*_vL6NY^>PvuzKQ6F{ z3%%A09?N4wbxCf(wceUNt;F$9b3MAZ7rqM5S{(oNrDG%!?u>OPl+K*y4YLUvdjp7V zbFxj=mL%IK&m;tMoM0=$9$|I-+@v}}NP+_0wVP{xJ2o=L7>63=%NtF~TZs8)f*eb! z*LBa9$3_{*fM=pad0FQZp!>yjxco0pn@F z&1kuEF5Tkz)67!ipd*&+ISc}FR6X345Mm*a#l~&Zl~o%frz_GlD7DB}t}zxSDpS)q zSww`+W;HS%2r*h-O@(Usf_!HRGQjMwQy78zqBlTU_bK$L;vb;zi2#c3;#V(8zUdin zN0aUCFlvMhR`?*p())-%bbmX!@Hkv7KNl&)jkBL|Bk4wRpPg|@a$j==jqk&P z9s3EPS;ia!Pf)2xBz2Vg=w; zAgw6RaUG^$R*?+JV!UM=VQ>gZrz(0Uw2C|CT>DC~ z?PM|tsHw1jSwN2J$7;c(TOLz+cBr1VE134SDbvI zuw7{FLQH1F#vj%?OVZXB6-7e`|0D; z=bf@y-a}0s$4t!qx;0KT;70<0fxSAzURW0Xc)2 zY2GOp)gFSpV=Y$?>TLa4c8L|(3(+&G-yDN2D^)P+^Q4}8DFgH7KI)*|dzeMF2r zcI-Yr)I;j&a*;AEX)Y*(MqKViqz- z)mQrxRr;}ZsX__B9F#btikGR_cmd2w7nW~Y#%ryY(poW(7J4UPoT8kIJnpcv)&cXl zQ)*Z4109-YyOw7p=|}LNgi1QSU9Vc@aLtPY(@u_v5=ww^f=Mb#w8m`#jqSZB*}9Z6 zu0j(Gr%#(>zB=G%6E)MHR#a91?5kGrOi04%vH+x|HZbL(&UBDbKR++*8~76}8mLYq#P_*X7YPfF z(5S}D*}4%GLXr&f%;&q#;m{gF%#rt~yJ;D+qLtPyqZ9REQ58ssRgdtiJ;zsl2jhZd zo>{s9mc3tpv?|h2-hg#>!W|unu%rsi~4yZb?!A{A$mqXi(;j(*r3@h?*56m=5e8hJPHiBujQlUZj;x#q9&}=Dz2H zH@U}o0-Q0KK1oBs{kHO5QnU`XRiF(x3CAnAG;Tu4Z(D0+?jqn{+xdr2sZA7PFsy)s zlkw8;9MKP<0V*vwfe+1P_34FRefxI@gwm4I50L*pioZzt{f`zE%#;?q_|!D8(xm!k z%nIfZppUixu$+uC8sys$9?b9fF5`!B1IjITRQx4QG3w|*egsiDkFV1t&kX<7(FZA& zZ6lVCiE+c81tWQ!ECBv)Wj0<4^Fo37=8vul?x2>FWf~cYUd&Mo1CHX52298E<`xIg z<#XrO#B{HPg%_QbG^G`S3WM>;W@_=krDrz)T%YON40wnS<1xb8P^ihZ8eklbY{62R zV3tyGB?!UYlS%Eggl8>`W4x`0r{Q_4Dqg~wTv`3=1N&+w(=S7Y;O)hhIb&FfEuX;o z@x%EZ-HZG9z-(>9<5AKUW-n=zYF$bB&2fL=tSYo&=`fFcymzGnCD$wbt}Lrq0~Gq5 zbUqw8{~4Kal#~U2HV`PAw_b>)B?a`$@OwSwgG%G&Aq#Ehrrn&6qY5975fM_>UcgYh z3P==S#r$;7>~k$fBch=*f7@ush__V0Mh&d15~<1k=ny&Li3J zBoqktk1P<3Wn~;Un1N2MXZS8(MHZ}AeGgxzUd`cZoJ_zg0f2V_2UF?%! zoPu|@M%-@XjWd1iXyB)X%rR?FElzGD4zQ{yj{`emjqHdVjd7xhQ2w5R`w=fw(7|@y zxXBFiOMrW-}<3J;?EbXh~@Sij5Y0xp;R(> z4m+5+H!PI$3z%ML+=>p@O(=N|f#i86QQ#Q^W(jeleE{>aMErIJ zKVyVm^OkNJ==Gs-duN4GBmtVcZX~l+#m!~tH-K?&Ni{{Z)&sT2Hcl`p4@ZjFA#rgp zLq8=fN{>|&L;LWtQ|+BcGlH_vjLT3wD$*<bm9yPOdO9gZsH*v}l3%t;_?j9%|%M#F9ZEGNA3qkFQmpYpDb9knu8o1DctoI} zOSDq=&rpeP^wgnrJ^)kQDNJqI2wvmdS?QRG!yEDmFvq9%{N_bD)#s%P%SHEhX#|oo zVY#wLxwL{nsCk%M+u-EngM4Efbb0h-fX81|e(u)9#Uao=_-r}SJxugAp|tO}0&l5e zq8DIe1#BGt0d09veU+lilgNv~4RApB;2@&xmQ0706G=GuUUbsm^bM}h=su|jQ$0|I3`AZfrkQ!8TcZ*%}a_JL8^##;v{ z5eQmq!hb`ptfXJT4hK(6UizRzsXtuD{)92E70bYI3q4%lLg~&G zKyHmm1=Tanp^$u>oras(2L*D;mK8Y=GYl^v&**cp_`{{}J}x z@m%li{}DxHW@V3~Y_hW(C?Z?7gzP=CX~}M&QX(I+bL>6JOd)#|Lb6v8nZN7xsm^Jf z-~IjLemu@CZ}0bOT-Wuy=F=PK&pOwZ#`Yg-X9=diQ?(S~c))rFCey1Tf(Mp4+iMVF zfY`tj36F;+YcHrfLXJGXuRQ>qB^|(;C-dP-Tfp|`sc@mzrmCWas*R^G*t_&|JWo5r zj8{xrX3D%$rF$Vv&C%`E?~*x5?XFs=ynZyNbf9(X)$ooC#FGd*leo;#C4tn5y{Fd>dOX^!$ z6EUkY5m~jl15%TTS36*m{=?OxlyqH3BSIQ?hYWCQF3qX_c~8+#VE8Z&#^v=@sFQpX zSo3@@)1#`dEsn&cthfqf!Zt-`ABi8$F6dC20;a~3g&})F9C|mza>9`}^#?mO}TP> z37CVGdw|5F&erMUgG-Fg_kmDB-}_7drB0zXsk=kKtY;J7*FOo=2g=Gkxu?p#28R6A zgC4L-tPd{uiy@u@&TWQGpt93Ab@zI)dbBRvj8oQeyQ;4r(J*p;-$&WP=j)_;_S1|{ znUHNYxhD9hvzd4C3t7{n4S$m*H!I)^Y=9)CzA$WFS=Z26K3~PMcMG^59OEqGtZx*M zF&+4)x#Rn(1n#@YJ<7fuWSom~U{JdhwhLd}9{3$y5*B9ZCSl8PLa|8!ZHl>}-f5MMl zk#*2*T!r@Fgze>2>oDFTt1?Mk1(nXyZ~nkPl3kzz7R|nZ7MSPBr`bYzkT+7J#LY+kp3S?Z?@;H^g zxe78AEBelDoe z@reL+(GcrHz5-qr%!y+nSWGJYmw+{BUn8vdJGs(cf`6MDg0Ol_-<4K|n}4 z-0#4F=!4vy!uYE$W2sw#a4p`p2J!9;xV+*G?%HbgtkNvQxVjK2q*PH)>XdMp8W7Ud ztMB2?sc;G1Wg2~zL`EF(XGu$&SinD2UqysU9R`k-?516{+!<+Jb|R))TC$;!LS=#T ztUvO4{%b*HD#6gPvtvOPhr%GJen~z7$p`Et252v;TIL1md-u^4`g|<-4TsW#PO^FT zniA%{o_>!QA~Un{nr|^OxHPd$E!L9cynUr>;wK`{3=;-3qbn3K@_nox(xH13fU?8U z8@4}hN9EXfv^5Onf6eHRsDZ3gVsw;%KFEPq%)xMQJ6vm|w|d*XYq~R3zjAY zcq^W+Rs-kh4&3Lbvy52cxc7Ty`?{nJyXbHS4^Ur+=?|EaZ6L@}QKWZ1bwj#i#tso} zAP%K3TZl#wQRz*A!aj}V?laY>P~|pZvTpE)6G$E2{!%>ktbD%t#0?-R3Fz3zou%f- z&>;OewnWBwcxmAkFge9FL@_1EEI%PhDeTebo&$D;{+VaSH5?~MFFkTr@OTB(QDTU_ z5;L?N?>Te^cEMe=Ra}N|Ubd0m)f}gJZIyTa7-)Us((Sr~=U%T1=!9%603Q&W4(CJN zp;BvihL4r!GG)AaUFlo;o1}USbXRZ{Fh&WU(ng5S2{?G$Xo_L{Kbo~a!72^p8;FfF zx*Pp#euAx+GP|i(I#VOJf`GWOF36+rkfnqUsJ@F`4%ql9I~fhPDTR@p<-L&-)ta=w z4rbAP*9AZPZYWJ?a)CQAfe*}7)o#t4salO>Burx^iS@vk@;bLZmGfe8lF|cE44r7T zAKR_S@Wi*4J3A2z?)G|b^~XHmzv}K-1y;d=Hz)RwXE3db16N~y6(X!HefGdE?_~$# zlu;EQxh4!IY`jb%Wr%Fjql6<_pUA_|tzF(45%<0s28Pq*fXy)`J?0#ZDq!ERoDoyP zyC@3Urr@>UEAyZ-r9+&i%N@%6DFoVvPS5c#s#?*%^gvXo7oOxcBeQ|nVQ%0t6(?|~(Fiax6+dwfK+ixX*>u-%udrsWg8{Rw} zv)bJ^Es^D!kmUR6*@W?aAmuxO=)D?f8e{6)o_jaO))QO0wn=-6yq45ab4vb7i0NCJ zMSQuTJJm&5RqNAwHy!d@M3t9Z|K^cz`WE{@3Y##9U`6q?jMAnm!_F3{{XFlPHMMiT z1g4>!Pp#5G*AY3V{}$SYV4LZ;U|V85E*nG4m)s~no_b;m%pAS2ui~{y6FfXt4gq~H zu$NCg(Qaw_G6&5~^kOYViI@J<^cljcw#2+ecpLg2)J(eoF zWi4`_l!O`x)&htKY2?$6ypUFBl{%QyUh=f)0LkIH`4fmBMC2O9Nbn+p4Vo3kJD=`V zo0rp8F22~WC$zM6dCMOTi^;ixZSwO~rp?6lN`fzS?GDgMB}Eak_Xt^dX#90EF$`jQ z&U|ZEGl?iluaI4!WB1?;Wq1h+yL1_tUo8Bh5>KRbF%Hx0RMp#VrKdh__V5T5tUVxf zpr7axFs`aDg?v)t%-0>dXPOBna)V;(uBSAAUK1-(%f-s~`uUPH<3{Jm*29d}xt>=O zw9Uh7E~*X#P>9w6detn|A+?UvL{*?!rr>QYxFxl<#rydyoT!9Ddz~t4G=~a%Y@j?z z2j2**^D0@@=mbJ#WWV%Xv<9g4dT2qn8Tf6ES>6mt&fQ0|(!B4@xnBS@4RoJij>p+w zWPj6$4vDIfx4yP-ggsI<^rE&6lUl>lnI+Cwxsifzfo`pX{hpy6I=&sj;g^3-Ydooha`%Rg}_N1G#*xVpw5pR{Zw>`EZzoIs`-6hpL}hEE zVfh#mZg$4zx}$1js^R341$EW@!24)b*y;$xL_AuQ?CTUeeBnCHvoWibgVPf9Gcxk) z!ex&?w_YsARiL~lwUs8l&FeuLijivMDtywf5UEX<-O@+K`NAu=Q;*()CMQKHMD0Cl zN!glBxb)cT6JKUg_d!7$y6hWRr{WZDZPaO>ujBXft8qSRicb>gFCrCdsU6BGN zee)T80QcG$)6RYT)r+h3Yw&;7DYFN4u`T z%=ifiq}*!MT(hp$K(43xji;QT>35UrHBh=7j-dOQ%VxlJIG@gN-3Vau)=7>5z@qi zlv@Wgcr?Y`ta*c|hDLMAE)+L&Won&NN2KKYGL>E)BnT<9SZ&^*b+llJbdO5%qcv*t zn#cu)sLS?ss`xRVz{(|m1%z;FXrvJX@~4f6cD074o*D4ZGh7%&n7?Vs+IiD#bAwd& zD!|tIvblnMsPTkk#+C^-+>aU~-cSaO;9!6?QnY@dF!%F?xnOTI#oE3nLXOt#rW8zd zUx767j{j=;@OY)s7fE;eSNe1x+x&7ve>HBVauAZ*W1d8G7kVS=u1*jfmHPVOZuiKu z;vw%T;95`HUfoS1bt~4$qn|}``_yF)4|AIAy4$D8G(s7)xY#|8hsZ{%igasT?2y^n z{M_Gqvaq*-Ow<{#YmuCU#g?g~fyshfQL_7CHL#^gZtQkQ`OvSOWc;vdSUM>0df17i z7}LhZ_|j35?b;|#EHxXelqP3kV-Qh72*F^Tch0qGWD_aMlf$(~IZ}Zvh&NNvmm1U8 z%@gStdQM33Yfo`i46Bo{e(9G1+fHCp(%kaE;9%hExqY0-8Wm54U`9r9nylCxR-vm( z$Vctq^GE_sEv3h%O8l3{{R#{7h!x8f+hv>+6nb2{+JvIibKs;V6*sf8E4En+3<&oU zi#?NOvWA|TZCj4M}V1JSEZLIy}Hb=GxtXt(d+8tmoq;>i0B)ucmlO&E>CjV0etRCvL*>%pk?UR`k50$H zrG8SZM8JO~H}~$)X<0|GLODAe*l@#ABa?}vg#NmJh3HscL0^jiBXLix_#XGQjjOQ^ zJ@4zGcI!$yHasKECLm=>abdVw-HPm!G(fb>iy`s8W0kt#ZDNQD9t5^84ySY5Hxh^D zVyr8I6M%|dbM^{NHB()?wevN7`;g(|qch-T2eiEoiu_T1^-T!|sUpu&X}}Kxr1xtO zT`enx7N@fvmH`g2^;sMbR01+MMXD%0RPy9#W{3A?2_w;jjwcD`~d($=bDr)9j3UP6Zyh;^tWO z)v{2jqQ^`nih@3a$h?ls0@yAa(}eWt{t6e+z0NR#Xo4+(CQM zyvGm%_6qj~zp4_9nH~q9$5^OKEsLGW*4MW59m;ZfbfI{+ zLc1pmM!`}I+qX4|)nEjj{6M6vw>nm7;7iKvFP0YFXT%Y?{7jdYh^F3WO}wGPgZe$< z$9_BRw6-u*xXM`d+1%H0JVv)w=4SMSXSny+q8}D-s8s~A-3b%xX8i`-Z$Zf#M^3ai zgyqBK@VrM!!j=FvkCKD}x1o_YJfC4AIhB031L)7hK9-gm^%Ms$yk>g7eJMIu7?y&+p8pKq%-J|_0TfT= zY=Jqm&53_j`Y}%Z$Prxc!V49amPb9r(&=42fc5t&gXg=WwufcX>C1DwxD$X+UXy-_ zG^IUIeKw=0Mf9yXsL)VEBwzV9Vtq^Md9?9j^Zqx@5?Pgs_BCl%2_@&)W{yoyJZMbQ zjRmSQi@PeKXAT}d+jJhNX(!ykNRvEhxXe*JV3TQK5|;4NIr>tnl2l)~QLLH6*ob+w zAYOP96J-4L@#nF5h2GG z9lT~nzI{syR~R(U9+eNhc-O2k37_gbvrnQLA#J#*Phn6v;?nLO3{uEoYUHVN-k%+f zx>5Z;3MIUZ69-cC{IQ2gb?vsf`FfS)BXydp$JjG_tZF6;I#rc7X7kUTT&2sF^}51Y z`ChxN{L_{~&&YZEo;~5AUqUwiz3cmEx@CA!nCl^fEzM4`hE0@W80G}GV#avSw^0QldlE@yR|e^0B&&B?|vgl!GaNdeMJUEXynv!!F_d37%9h9Op}dN3NyN{ z$Q^+f4yI)%x$n?2aDQfqQdeI!^j)eiy>X~ItdZtw2nG z1NQr>U}wV_XvIM^brdXDM%WkHt(=WEkGrgSOy_<+L;tGlG%}I-g9cpexJqPNw8my) zH~T2flF3O<%dVuKvlxy#r_m$RWS`7G9jly3{X(8aRH~26)*TZe%5@-L^p^Sp%?okd zqmd^7tA^L7#?bNRDr>OYsnM$WIe-t~XDMcew!_|uBqkk1XsU+Aq)wGCQoTs9Z@(P|07-5on=GAl|YN$&f8HsP;VtXjC&|ZIpn^ z@Ga=fsWzUjcc*tRXNBc1~yJJl;45j`zFla;Tt`{2dG(&u~VlEq&LC;_0k+(&ldIS1M0O$Y*YyRCl4+DU#n$({@C zGDl=(#`?b&)Ne|DY}o>s#{yPWqarROm4nf`*=OI$!m{2;cP0v z$3BqPp~Rii<~Q%ECnvrg!C?!{qBE~2vEAXqthbJsQp|PVSRboqD>Y_$Yn}u%#t{SC_R|5xrmbOUVqkXZdWFRr)A7&7x;e)5ER^N4;X022 zde3$~T6*u6m8d?xj=V&~LEit2ul>s+xqmQ||K&=Dum05d9$@vBYnitH4wI(Y=$Rw% z=o-qsE!VIpyRG=f6av`R1M?tC^e)%lf@c_i_@Q0i(YKj?w>qoD-P z`1;fJgd_Jm_8WWc9gcKAUvMrw<@O5w%IDV?ySNQ%CLt`AAw%&9kqGG)$0(L&6!H3PA%X4ua{U=6wF_Lrk`$1ZF-sB;r z)M#Tm>8h%9FQWx=KT|J{B+PaUejY4BcUe-5yy>RebHKp5l~NumA%mK6yeNVjifQH= zkAkpOT0(N9L{7ygr8saUHKmqs!lql5j3Y+a9xyoSWo8>Jc!1=`~q1eD$W zeIKtKRBiH*r#FU!;LWWgM>&}s(Od}GrfQy(F1+xr@9QLB&m7~`asGf~^}`fNYw=|l zRH?s*;-Q+K2CDh7^T@F5Nc9A5#S^ZJ~*nP=!5g**6;&ruo3oo4w_ z3*Za8-GRV4^5If0zo9L60Dt(AbAuZI#+2(+`f2O-0SD|?XA6e>xo0X?%6lEL8l_y6 zw&^^aY-fzF@dFXEg!7LDF1)yZ^`_HM|D}ZZ9w1oe*;F$C^9im%s`n$a=>i`@{r}XMOq2vcq zz8~279pn*dYkJi2S3+M#8ZADSDtzl`h2au902C+0i8t?Gj7;~C#eGR4vt}A+`>gcF zf%Myf?O-E@PbkZw-c>L!Xie;M{qb+`{>{;}X(?m1XU z{Hr$DVh%`w3!oJE76-(!J}G1eSeA4%gtOL=lAN_wp7kJ4u$%P=k$utNKt?8$HrOD& zz5=cm>lQ@zW<96}0}ce;j;!Z&g;tm6k^9s!LE!`geh}o#*@q2?L0pwayFxViU?Y8< z^IkAE5}Cdd$FA%Lht!sq^D=EYVRl|K7iZ*+jA%<-1;9qy`i_+=d51@zC|$#)21|No zeFq}J7OR=`?9NRr!3U1bF9IlQtzu&L2uYU~%qYsu^BLc2Ofs7J+bTh!*LBFt&a;NY zQPyUH+kd-m_j9qK%WT+80))?0QL03T-JiI=4Wpx!wbgw4lJkK9_;Zi(H@&B53UV7{ zr3|(Bo%%vkY_UZXkjDMQcw46wed9Bd>iIGftYt^Jlvy#tA}d7OqZwD=z?|+S2y1m!rli-eEaLuRh6#N=D5Q(~QqG zE`AMtbBD1P1XVZ+32adi+zeQ#_ z&L9L>PkMX??qsvQnA=L%LitKR+?g(BT|Ss#28i!^K0_GmBq_0NcA-wFsobnLMaq}S zYt8kLT2SZrI644YSfj$9+XC{quOfZJ-kZ`tTmc6|yIM@6Vk9e zkms5(#G^=#O=rJ0_S&(xnvme>NNrzvC-EQ?Z0hc03325!9RDVK4Sz0syi zuU-O-^#{I!%x;cPKQr~3@J#m3r8|zd^)${VLnn@Nun@{;ACfbmGKrwW_z>Ew&mQ%h z1}gg;@7}c@&Gz02zcwPaGyd{&p8q}P1P zw>AJ~Ip(Oq2_3KKE>!m$qzZX+ca67|O1$qQ^!1)zf~30Pg?LU@fVwet>gF-zBTM>0?ASDk*i&v(&eTEa|_6m|z}PZe}yqYxS;VR!sTCw6oS| z=NUV&m$i05a^jKe)4Qq?obtPDl3kz#A{8Q;K6f0m^IIjR0H`v4TFLW<&{%;F1J!uv zYwuQU*#L;_{a~pr8W|H5X$W=K_c6RgKX%M_Z%xJLcMB`fzUg-(B1%p*$Sh#)rI%(m z0R{gXh;Yf&9b_WD45ln2K0~^B4WOYaIB@m1h@+?k^u4E2G9|Az;4f{LxtQCsu|={4 z+t%(Z1zo{JN@UzV-1VbQkb>)dsrHjcw5Z-J2p0mD`{cS#1rS`u^e}xE_oa}>3-<@U zc}vOIu2|%w*E4B+j=j7Vuu!>bshC;6qXQvSmH?gzOkbqz_8N~lX*sEV>3A$B?um%# z?4kj#E$~sYjmgU~zSEisHpztUl z3dskMNT#1b#M00^jep!`VGA^^^nOx#X12(*P)siv-Dk5+MdD9caL+FNeKE@g za*fF-Nn7p@af4RIo~+6*18nm>S_(N8{FNR#wUdfuGA++0yg#3JG@>LE*mu=?i^jUD zYVkV1IZG7>SzQ=Ib@Umi70ud*i|4H>N18G5G^>-rLDZCD3xbyc_mN>|F^V*j(uEA( ztZi+SJ_E050aO5`zTzv-8ffLp%(`QCq*Fnu;z$$;dsnP@hIphksQ4z9_Y?BOe<5Z6 zrUdSN#KTf-A=&_Hq%$_jUcD74W+N@EN90zBG6`ftMM_b7S7XR=o`4lUyq~s=9sVtFZDR2EaW&Gn2-F_nJfTQt*Q9B-dUJ9NWu|Rb*8ZAYl)d%Q zS~-56Fk0HYOZt!}lsL@d?xlB`A>uMZG5TXJV9jV672w2SUe=#nu>$<8etq~BEkr0n6KX1E*h|@kcL{gIp=P!pbA`>%MW#@t5;a*CEKJvCqu;r z@BX2NwiF1G6)oLZU)wj5B=gwtBuEPsa2C)Ccz?L9m`8ffck-zQ^0VB)LRvC4hcGB{ zTfo|9hswe15IqQC+L~1CS8w)u1^VdTXq=K)o6pc)6!;Z}q}I6o3F+@q9=Q?R#|A$- z7_3kWy}aDL)hO5iXlX2DuqT$i!c#>hNt6xcK>g0Ama5m1KcJ!?26rk*6055e zT0pkG?TRcLgXc0-(wH}{WO4i70XnqVEb&*B;u2sR@{;b zZh#t=AToU4^@bPBYWu+KDjg+~z?Sccz%7W~4U>N=)MwgK)XZx*B3Yn#xqmjR~dFw&$=!?>_#yeiS6oez>ZssNbd)0naqQ4i&2EH z-+ez{iRt`|77yy}E}c+ZAUIG~LnQX@7Nm@$mf?RE6yPQJNx(1fiBFO6A7NF-cBSn( zo(S4Pd=R|5yAZJ5E$mpaq!8nr71)jzSMcGEvZdrcU8UJ=AApV#>o|`H1i+y9qsl@W zL2byETdco4=PyXmc`5um?tclQ%Jmg*iV@(^9vF#Wwd=^>Ek!DXLUk3|aWJ7B2vVWa zz*J;824TH=eH`1=GaQjh4?*_x8R(H`y1Wz>P{H{w^|3b>Bqa6vk6Q6pE^BzhW1XnN z%P&-D0Hq2#AGx+Zzz+qKbWa$9IHK-M)G?^(0Xwi`j|!Bg5I1>nl|k|^e{WEX;8?iO zQ+VpgMt^ehUxwybzQmeD0t!ljeNc`u!0iY5Q^pvFEO5u%AWeHo5ktgJXJ4SMZx+Jd z@nS#!qpcx&Np@+d?&@C{fNIUvThl0k{KdzgKHx`>C|()#F>4oJs@XyS-?Ri;mPC%f zH^Oo3LNltGxtPthdLFd%<6b@TH$?BAqUyb-8b}s@bVGsa6>vosxN{hmH;%aV!QqD+&ZKWOm`eea)ieDO){fjX1fY=5 zrU-heSlEAaO?0?aERLFQI>M_%8R;Yr)x;)l&bTo{?qK|YB-*4bm#cT5KArA z*cFdRu@Ck7Lg@6I$=0-P=K-PDw+yQ`K+ndzNN9XyyeBNc>Wrvu9O4!MYllLJ!(3m!uHj6uy z7UOqgIcirB6o=;#!`G_GK%w%dmJr5Y%jScOsbK-Vm(+IPgx>^!D%uiDRp?3veXQ~V zxN%_KCcz>1NAHM`=E#8A9QD7K92cCdam_@Xd8i)vN1urwJ0MpRmDJ~?lRi-HUzK3C z`!Bh~AI1DX#yn#$1teI~;BciSoXenlB}fKQC62umDQLi}P`7=p40ZfvsH1V~7}WAZ z6jiR|r$GYfHefaM5j?5l>1%mD*#E-@tB9~lf7=VYu(GmDOCMB43B9~;(?tY;U=^lK z{5y*B`xn5{kAN~;d3lN!6mO?ZJOONBS?2UJ{kqXOL_BCsm(c1!ubw(FOQp6F|Xu0-H1(4eF8ec5Ge# zjY(j+434H!=IS~KV$e1kn^=8JB(DI~_8;9+ z6x^_Y+{gM@=-va>A^KunClU~JKXT9-I%dpa0r4gT26nmfWuV{pl`o+>I6gp1Ha-@_ z9TcLIhiSR;!(goXLH#`CYhkF>>8D|6=*0G%hU<$HXLl0Z1-U3Rv7qVHl;s9mVRfz~ zTns(9zpdW;Y216tUBqzMBxwq5#|fzmm37Mh+2?Qn0ZmeE*`GMlzoo^XziT?V$R|c0 zbd<)qs8%xOH_32dcb!xWJpe*IMK%dz{1wDM6--ac`W6J-j;RQ>2xEFoEMl}fpj2LC z%m2YuVHGDzk24R9W=*yFggq-*Mv@}$mSabI93Vp-5RL!z?U~kqxeCR9U%3=mec(XC zuAn;*Hi@x2;g;$Hg7Cd6IS0PM$M18IKZLtdEn@Ya>Tf{=T`7zw@AS|sO?X}D-u4@z z79L~*Ful<;@ytjPF?0DpRCC|2xSB5P8R}6+DXCyM7KsD%9^)uFQ-K|ORlo-VTXrFn z7dm;k#6LmGD)Kb9iv=MyWqLaAcai-^S}0-H^xIvrBhX+ZMGYEBF?$9~jJL~xTimW7 zm=zLETVr!l8@lZi8%9{hwtak0W6DnyrYi2(cM~hjQSlY|eQ$&N@khS}vjaA!n_n?olObkbUPdKCBay!bdKSiY)Q(5_~N`M+q zFBR@=PRLq=%l9hM>@A?=4wR=D4Kx8AjqJ&+56b9!Z!R@<)`cAGgqDTJtq`X~h3-$0 z7bgwwHk1^K*Df#*@5bp2{XBomtv4yk+yVa}vjEDFl68%}*ju2C zt9wI!CrMUm270G{8dLT|S90jU>0`SNVY{&t&v$Z#s0R z6fG@4{r419fVxm!rp=s%VqiTvn`sE-CF>6bwnOYOR&>y<(+v<$gm)&t@5p}SN}UmU zEpW0hGEluT=I+7Ip6xFYM4#LBg}Sc z7Ru@c-{!&(QXlu_TA?y<51~-!y2#@(TGn*jgfX>y`!2vk{LB@%7& z;LC_6I9t!Bq|8@pPBmEufGrp=OJ==2>M6Qc){qQl=qQ$gaO)N)E+2pOrpwAeAE6BZ z*S0FERn*s$zkurctnV3lr1u;Yu-S5lAr@s%es2^C`a*Eop_@@R8`Zr07#W(^23t3i zPbfl>X#n$15n=_ye#zMSi?uF zIB^oh$45lq;FCm1j0BvnTpK{0I)1z!YN1BqI_EE3KQj>eMwNc$3&ayGhQo@)hJ8P? ztbE_Jgzv?BX^}!6M^UP=17+DhQ~|2D`A8Y7Znih=V&pN))A6ufZN49crofeX_3R<4 z>V|CNofZ>&Y>k=dTf4ickPw@JsD?&@0cEX{G=Lbt_k7M#0xqkp(nsSRf*L@_(u%pt zC+q{$mwM3lJPfu%GmuL0H2_e1Ge{C@UlFKPBy05XgScR}|5sQbQ(n31Y=< zL1ilUMroeiR_cm^DRI(jne|DslUgyvk01dA%?V-Tk|*Rz zxjZB@q5;`M0^w*2=t549=uq;JBynPgWEY4Ar$B>;M^iD3#17cnNBz$PPc=Slb>jH+ zouTgt2yY+C0x+L$!5SJY*O9tSS}$^xR}OUrKtCejIvbs5xig?@AQorSPpp86A!y=( z;^l6~D2h|*Tcu*CPoVKmed{T<7!;$hE}tuURYiFrS&`8z*}HDw^a@*AK2+6PD3-_{ zstQw|J>drNM$6%Hq*owCEIEDtu`pa=GNyr zLf^-$G`|gSV{S8$tOBm#&itJ1=*R816lRjulS3^a0j{g&>@5n5rw=TZP4DzFmir?`b;|eFUy>4`Q*P ze_aCr*jA^$dCep!50CYC6Eb2psU|*ih0D+`kh8uO%;;@(`aMUL`i5)HP~<6Rsa|<% zqv%5F3yB2y-Q38LE*xU?nZ0dRoS@VX216s9rJ2??>o#ieu?tSJRGJglcMlmG8Ko&w zXeb3=-m?6Re82ZhcIS0!|ExGl21J~30u5xY*q|NC5Ke1y z^o9J70N)R<36$z&Bp}N{XHQ*(2WTL(-!zRzLW|kNLL)5nFPe7^QG&NP2{Fx=kwkRa zb_WXH@9Y5&gN{xa#y6z3$>djA6GbTIUD0QarF2ySAK{|tCCST#mT~yXH2cD-p zbv@X5o$O-pQ!k-fRxV&T!#=AtTy$&gLnGaSuCEID3NF|evf!-U!BG3SmiXuIzG=TGk7;4Sx zfWXy5pGVncbw}+`&eNkB1%FY8j(f1aPu2i(U5z1xG%W&%3>8j$FSjeY4ljUdR05OW zSm=j?DeKhChKBoLC6(WC8a?uAE;24-Bel>*Yzl3}mvw9CM6l79og;8!k7MA(tdiN< zZmgC;rWNUB^GlCL>?4?DWl$VWrIrU8Sf-%d|G|fK@yJ7k2d+mJ43ij2ynz`^9+dlDZu@JaAnEaxYQdBXySBY`(3iXhi_ z-n!5&K7mEYuF+HXoVR3NH7Nd=;}Bt7eQ&n{(q0Jva$G@~g%3 z?jmO4asoKfr^S-wC{7)h|Ine1gHJ(5Mtb|}ME0;wr>5qhd&cym&qI&z-n!t~>c!u- z8sA_1aqBhbH~D0C4f@!YEtnHE0&$&0G-`(`)1lQB->!@20Q|;lY6yH>vWTV^SijBm z_4eXj*{^pG-rOto9;eZ9a7yvxtj^HaR?iy{1fE`dTEe&h)gnwgt_e7p38{%=jZczj z)Vv_zUFj5R@-@c9%Xc{c@k*8x#B1qg^Epi2MqXTBd|3VU;oXf(wP!v{_|Lm0(2de1 zKfM>&zWIrOn%A(%;TMGnCTt(6lY_S0?E@$T6aoIJh>laIv{8*h%QTX}De&iKaw6~S zR>(JLp{#x}t6aUT+;-|~zU)T+A_$I@)Ti4~2gJ(zTOt=COuW7n6<6|BtYkY9q29RK zlZ>BkT`%!0>x_R@q@|Y$oq?cjdDR?ZgeK|-5oR1d%olpEEIX;6khg%wp@ufHV z!J^qMad(XxM2)Ucnk-R{EG{nQ`B3ABZ;#1tv%MSV^L&7pJGye`Ud4yJ968DUJ_H{h zee)`()#c4yRFQf;tU2=mc0|L^PIl^k@!~w}2z_QTxO?xo%%oo8*>&A%pSiy) z2*#g=>+tJL;b3U+2Hw4s+H<*C^D(*q*LwlS-c9QimmL^C$+WJv7VG#mb}>V@(&p0w zN;C(fW}U`W!!kE(a#3#xA<2C8Tp9voBKmILqqrP*6**IvKiy4`B4nwbagOpayxjlF z*rVUiZ*REq<|uXW5vdWvyE3D1ula8j2S`dSg`BG}v#Z`Hua+d-^o8dhZNB3+UJ6~j z-zfDlZ-WXNb6bYL*bh_rPk8!Y`t%Nf+YISO)bR&*b?HwYa37p=@Z7)P-CG_Rpj0y{ zr1U~Qb7$xS46*qM%YZRIhXd*Rp87QS>6<)=!oKPjTYUvP!Gkof8^jE!KMMQk!d8eU z&~SxbxHm;vD{u0}!bJ3IS;3U9@7&zn^)<`Km+OmZ>cyxAp=GT)oXy{Fpx3vh_Zq_I zcLM^(=!P$U(zaPGX_S5e*~*ERy@8d`M?sw9H)mU$DzlRQQo6Wu&i=@gl)ciE$5Xdg zC9953LFakf1AK|E!yBl0ZzPDs_r1kxV-9PW^f#Ffa$7=XXT915tg(yZF^L` zQVXv0r?G@6mUQdWljawVa1{X2&K#ja=sAih!i0`m+N@?;#nHRdnTFf{x&gg`?@6tS zB+bqBk{Wx8uDQI>kI_KSS_psxpo?E@x67sU{=HQ2!tggWEFhg-06CoNZQCm6hp0ikQvROZZ zl?P$@4;xYPD>%jqQ!7pWty&<3pZLYr!IkG1Z zFSfz7TEZNSpgT}Ep_CDuOUz}85#`8fAS6wL`E6+(FXE_zUPFn1ApflqG7jw;lS90P zE)D;Vp-NN&!O)KFP6okHD}W7a9~WAa0rRxp)dlllup?fy>|NP!8OxP$_z{Kw1fSuW zfM;M_Q`)`Q;En7^&^UZV>}MCxdR$Q(O6uM&_-7atZslagK*~&d0>%p+_(dq;2_3vg z289~dpCOemcMQ@3<6X_C_fq|>PDgY8GqBJ?VnNF>rmBc6~`;A4z(rbARc_ z--GSaOe{JlesN$bY*!F1Y^F_G63TEV@`3Iyo3<9UCFo1b4w_I$$N^1!;UOSB4|BGy z7^K@!=;mk5p8_ApRe7ZWB?i{vPfcg0AIj!1#T2nVf{X;S(+wC;Gk!d<2F*6gO!LC# zi2vDF2oj#DDb-f#7Dl{6U5NTok)b7NFAiNHBdG{vn1YT05{QMk^DCG<&x;v1Mt9+- zFG$pZYgMSzQvp9!10o^y(n6t6kKtN*e=BBF1wqs!2F;7cBqe=d{QYr!c$`DO1o$F+fAqIB}} zFs#Rlkf-?fol`+#mGt^B600QGxlpcq&xtuftdgGL zwt4tR=Jv>j+5&_nM*Ro|vS$t;b!9WQ9&>1Nw1JP%?|MY%pp!#k+(Yc7Y*RFTcf_vH z@IHL~1$-eG7nK2?Sw>O8;hr=Yt%gA-@avhI^1|zFm%OmaxjpFhyR`3G?1R_qiD96M zNfll%F=L}M2iLVw*H!2@J&w0d274cc(LtM+C=0LO<|))Ff!B}V3&U`>41gyY&yo*! zA@-7s_+m!Cy*^PFUSH{AgCzjpH(iA7_A9QwPpo z$16w$&K~A&hllb%A=HAi*TVkj=ZE3CXoCfU%ZklG5`z|Jxb?2%27Eo8yC@V8G{Y=I zu5y0`lN|W^%0dyL-!_jqBG5Hcc-|W+caR{60TF|I^`&9uivtar43{`$AqlcR*EMkV zSIDs|XbJ3B#HDQPaNS9i4{mvBL~&+1Q!GaQnlJ8{wv%KLbXTU6h1uQU%6~i765NS|R}3 z@eG95LLEt!oCEAYsdnAvJA{2{?3i9Oo>vF9K>tDORlFW@Qa=5#w%T8BFRc4IKqo5l zQmntMfF@Xj0sIOlIBr14h6>H}K->514RQuE+OX0C>fo2Z@a8FOq3Dc55B#)A585(8 zWxP!T&4Y27L3gqevIi^8Fqt?N4N}C8Z~FNP1UaKk9LqwchIA4pqje27$H(mNOVc(s z?F6UH8#ZxT+3rdRyq%mHrp2O*^6wRr(}a(Y<0^g$dOVQ;ZZB)I`K^m^1`$uPNZ?N% z05dL}mDBSUxHD{uzgS{Bg@5`R$#DH*8jNz#kDLH9-tIpBeh<#~u$ep9DI`TrU}ZTH z^l1oxUwR}g-NVNIDLl=AW(`Iw9h!I#OTT(>27Mw_#vh7M$i_6ow{!9TJ@xy^_3sle zZY~nugNn~A;qt}uZXpzI!Q>J^2!DlL>m?0bxi>~|4a;LB317oZ_`H9Ksef$56l`Eh zey6Z3X47wB%tSQadfmoGVGN}I$1eZ8e>K*ye_2Fm^P6C{6i0C=%{f6Z z^ybSFpr>jQ;pjoQD^36g+81V{^wZ8^@0&9`9m@Rx2S5;<6b|;JjKETtDV3r!_;FPH zrYppz$_7!(@4Nid;q4021Le^JGcI8Y;UDG_5i_!2)bG1%psCc)>xJCVSN`$>u+nrD z#Whv4u4=%}E2T#}BS{Q^liOA)ypQH|KXyO*$8`UitUy<8nBWPg; zs*HqFx^*3Z5|xF(O&k!M@1XD3lfMKfFa5perw3G){zJ9hg~;Yl|iU>$ipVaI=00C-fO0rXFQ*n@614Jc15 zbEm;^ZqWUX_Tfl*a>6lcq|M}?gLTNjJokSUAKE&2MilVxaVc0iPXz^_N0$QI7^A^C_PgyT zc*OhuUGqL{SH(NP+Af(lm;O5Vx4N=I4&pnbfREQY^&pNBNF+PQ#k|mhe>#OkLAV4l zlwEE}{DVOJ@#TfD`y=tkj(Eni0mGM3T4vFv0z;$M5BI( zoc}}mAP{g10s-7Ei$Hi2+miT)+Ej*@>xaH-L0ckl1)K$frZSF`H452)mDrI)!&Khp zTrT&>GbL%QWpV#-jDNebfHME5;!8tpMWKS0(u5;PS+6esR=x+f)EFEcETDk*Pa=AT zy{A;VSU~8&Nktlk6Ck6;bnjrXlltI_h~#g#Ch&i&9<>-)y!QBOGrF3=4+q zmBe)NwJ?S8oBTqFrNfhls;yDmJP>b3Hpz$Ddb@^YVap!K~W)cgz*X zTN%;@U^m6+;Jp(B6CK3I7vRijzJ=i|SpO2D89@HksYYD9CVh7a~ zTjOFo2x`;4w|5pk?bcnmMylPXC8i7{E7daFawee8c1(jJ z{C}Us{(Bgln|@Z=F$5VM;qD@6!UmTw;^v^PF1mi}HfJT_p(bF9;K=dL5~LkT((aP_ z!)h~=qF~Qn^ymK!_J9gZ1>6x+E(~N4+3~mGWxr@JJQon5n3I&@SceaFGwPAxIiM4y zK%dL32=Ysx>GI`1JpLa^B>-&jI0Uw^WOU#$S74|uHO*EuP;9a zfyw4li3a5IhyvUg$ST)sBZGHCFcaR5rq4gh{Jhg}*f2UEBK)`Mw{U2oaQqrv7LQ=* zR*G=^@%j&^aFM@;Ewpt3%-vjk0ty5nmiz18!>+J{>gLL|X@EJCL1pmn>1R3pykIzr z03abf!TbO|UT|Ggl8Uc6=di*LGV*`6kHZL}qL5NRgxg&}O{-<1(qbhK&JyK_6m|wK z2Et?!s{S^?@Bv*fPXF-%B49W((&Ih?><*GZQCU!M%f;k|I6rCuz-{Z=YFK8dFbrD|+>*$~PM%V#= z_s3Oua|NEK%CJLJz8uQ%e3tj2xj=Y+_jl5)J!oKh_?NC5*TWjbdfEOEB@qy%Z0&cx zuyeN%)#0{?;EUUUz7($XU}wEAqbecEsdS=!)IVY*|Gph$hf+ZC<07r`Q|K~_CzsZL? z6_EHse(S&ZsT0fL5axB}(VvAt2wr%DbB{6(90ESYODtneB6|E%hNdG0pZ>7gekAb! zoFMi;G7%PQY0*UgBJc>Xm{XBHA9zyPe_uJoht7-#?5X}~w`n2>^uP7`!;_OU^@DSq z`A)gL4yI_jSJM_B;xCkjQYVq!3ep6f8f!xTLjKi5FRaJ*%SR-d!1G()HrTq`9tc06 z?~DN;tgUgn6MbwDHz1-ierR0$zR+K95{)WYQut^T6%FeCe|)_MIF|kUKmK%QZ?Z#1 zMaT*v}3JLNX&n$OxH4z-NsYT=llO3N5|*r z)5CSWuXDW4@jeHTK>NnLW&n_Yf$>um@}{r4V2bC=Aa`|bgzCSDicszUFMR~hE z<)#t(xwEq54+3)X!r7&@YIKqJa3VlR{09(X2|)t|GN5L*y5abl|GpWm2lCsH9^Hln z0E8gFh%P07{K|-y;>!;Bp|@xCJZ(TQzi#Vu__)76J53$1BtS_58TUjFSuahuJ(N{) zfF+Rp#RJa@U`YUEO*UmD6#nPk8Hny=y>>@*#~9x2WoK@|13U|*$j1A-DnhN!^#B|z zKac(M-(UU*a|}qP$AL)9|7ko(3H2#hNSO{9IF=zy`TN}C2cd*w-@`bBVkW2I)wFpx z9~Fc9)&EOW26+{;8a6c|(y`5PQsu+{it9Rv>qN^XRph_4h68HIh3UUUgSg(|j>Y>f znS!kfaB`b(-v;IX_vDda_IAEemk-?Eh>B_(Z%Li+uG*JYVd0yiao$l{lt_(7`;Wp#R+0fE(knTxYQfqaVl z{-X9fO6|8^SMWuj+mBxuN5Q3zh?&hjhNu6%768S|L-6cDCpGeB7X(=Nzg3zMj#j)I zAF+5Yp8`b{!#=ki9chN->ij;n3poh$-#W`Ih5WQ2m%179-213<{MY@(FHTiBr{D)O zb&w$&eK1_#e-?|b6!e=FhWQpVWu=ViA#%fa7W)ig_gR4tx_CzYCC6$=(1Q;lcEe{s z5Mzi11w}PAHMvDapWN?c5QUh@&Gnub z+BCom{Z3@^C<{H{Ao_DL`I@Rd;<9pzVA0K7JcG%5$aAhD=ln|$F-3$!NyhKwjDvmd zSE#nyeD>9#V6=B|xOhj2n3&iuTxV+mxe55a^|3a}*qmZ!cTbIDs=o< zr5&X_57PGvq|ea)7vjM&f_UkthZO$jsCKOI|uct4)K+ z5jHiVP^-Pl)`Y=k`{67@wjx&Wvr&ftg|#a}UJ+xo29WPQ`E@HIP2ZKD5?BH&S#&*yWggjngqG_AsuV5lrse~>1u~^C$5BCs;n4C z;DKLc9c%`C0U^p9IjP|P*p`1V9|@a58u@R`FYfyPAHY5x6&T&%CwP22II@^YOXJo< z{A2bI)*@m`sf)e<_FxAew*L}xxb@pnN#ShIpEPe$?!XJOm0SPbtDf zo-pYnR3$z)x#HdiLf*3FPi2Y?~mAw>>OUA3j$s);p2&b5Ju;GY!l}NPz_u2}) z#uC)z*2%M?`{5<6NX5{k0CsLg*aIQ3L?!w>Uhg~Uba??b3mGf!{x6h}(dz!fi?j-c z5^1tkB0=)OSEMfONeMW8J(}7^;yZtYP_307W=r`&Z;ul4BYrvnyZG!*zD0vME zeE68YPf7r2qiI-r;eTRvv2V<$Ign9`^S30TmS$UtXAb{jF?x(e~IkzKW>(M5uux@}J=ql|?uTPbB`fK400SDpqXSnC^PRt#MP0553pXLaQl%EZbD~O>=GJ?w6{dQXw z&4P6Ij5S#ID&xx~{-f|&m0pcl?5bbTO1n@BEsW`GW4omzS5K1JQp$4gcA!^f60JBj&jJtviQHz&S`6x{6`k2a|z2@7>hA4L`W}Dk#4{mx?6o z|3dCvV+)W{famn{28Vq3p?WY~e2_osR!3O}1V6u>%^_AsuOM~2qqXRN5a-`dYxTt~ z$lBewam>wAV=v)xC3qb@^onXirE;DeMrbhkszb)zkk$P;kN-c8ft;sjU-!TOb#HGk zm-pVf&Q5L5f2_ca>_L5oD|bsBy1lIDOrAKlxlkx5D{3kJw|jW^gWxao1VbwwEAwN# zaz%N0qV9egYi)ngXvqeZ;UE0mZ=V=z)_Ix#_9^~*Lh+-HVg}*?&?Dt>Hbxdsm=D{; z8lKY=iACXL29v+XK>1pF*scGCoZUwx<6b+qX)4#(v*PoR9Uy!!1+bg)gsUV>7l`>R z#BE-R|M$ZCUq0Drb0#W+cX1cI7;`^_r&VzvXH7!X`vAAmjPNuh10F~-e#}6eJt?A% z=F!K)|EH`V2f`X6lHup)ms?uu{_aYse@KrZFLK?E**z9PAaX2z3vreA;#;frcE9A0 zzu2gK&2(&|H}~-whzp6m!g3ff_G_S-KQr#rW*rf+D6x?d=SqECT>CT^I|CpQbvaHR}R|N6st!0)Y-HG0?f_4Zzw$iNNnzwPSk8cxhm zr7Zhua7zjmL5V3}W=v~>C*)rHP@9 zDIN>QBrl&^e&}=Ma((^BzX*x>Lb$py;lP#hMc_Ju_-RNsD%(R+A}${I;>MNo{4o|O z@9;%4pcLe^qS_9oGf3IBILsQ_f4+dPdfJg5Ei#Im*@rANzd4H@0S;RnfeJiJJqp`M zoO)cY()z6xcVhOPn+t?r=&#o>LZ0594jF&ju5&aHJtn@s@491)`U6kx7yJ2mpx;IU zADr`oI11h!A}C`Mf0h{K7uLbhil8n29O(C`4^u1C6{^{Vd4Kw2SZ@q$J0Bgjl9G}d zzdZyCpf?XVT{u=~jbkELx?PKpws~eg({P~IQ3X$v zo>nVqe6${3?Rv-p^x-T5>6;)MuPVUt6H_MaKQOKwSBf!_*acSgRGbWa?JR$^|Jgrp zqOrP|T<);g7_ZFXWwXz&~LF4x=FaOTREtH2Ymo} z8Xdp=_Cy|#2!wscoJu7&>8x`5XJ=*tr2@YfL1aGE8?_Y}nXrRPb8|uVL=gG_F=YzH zLWmlh%d>3Xja`^?W*ol(%uTqF8j8;;?J)NjIU5H@;Ig!u^c zI8!lQ*fSasGE^LR?fUgFvuAi`It>~B5GJF2ELfD9^-_xTe;!Mtk~dboq^5E8_Fv%U zjiIYnYUhVSsw$4AKbh2wIzq)NLwy~)pWTImAOjz@BYeUL5?-1C{V6b@Bl@%60Ry>% zfM91Ifrto=ybnz{5E!Tv@x|APL=jWI{Kfva9M#g)n}QsjR8HM0t=sv$Rl=dM5A$O! z2QIppoAZot2*py}GO8c{`jzvobjcPC^?-(*XPhx?RIHbH{zE>BBN6Su9nbo6c5zz* zza?D1$fMYVVztlIr)}J{4TP65gL$zb(XzkASpOaSbRJ7w)+S$v&W+uzBUj;%7=|aE z`2L_(dIvm^Ck8PfzS{k_#hXBU#^W#C<&QXs@n?q+A%;)`d_%Yy{(Kh0229YYmp=m< zznh)FUc|&Ntu6N7*4DrBr|5uAdWU_1jM&|~ck3P=l`s%bfV}ehWG4ksvGoN|F)6lZ zLNH=pUUBR!g-eQG*q^u|l8z=tG+iUwzj;@z;V{{@G%}Q+K|)ncA1g_ zKQczJJb&UvZW#J3@|2tW)4RMzKD$F${RUW%iuLvN_q|V^ch%O`j${lBg>SjHB9J)B zXH;h3`;Il$jFC?GkJM#ys}CXEcy!M5uNroHl1A_5qw98UY8gFSR)4<|QV$9h0|z1b zcK`i_+aytLt*v#hyRO#A*U5M!?y`mX3U=gVRmDpG^{tqJ_2;4Q&3*XIm?3} z!#RMms8z~ehCk&MC&d19G8fu3MEv?vKhFGV?a=#CS>X6;F^Jik-;P77CA7io(^qOb zrb^o3*qYnnPh9d@zV4s2*xjp!@eA1Ja3EKQ4+{8*XUFFAX`3wGm;o&6?&;ybb?O7v z+R(dTQ&-oe7Ahl)L}SpQ#FT!fH?MWepE<)N-0JfP7orYrA;k{dS{xGj zE$wP;&HEO6AtO3MBjeF=2jIb0ks=Xo+6vZ`Cif8Q&DOl_8#p{l6YaCXD{lACHjzK| zB7dLF;Efsnaj5Lt>R7l5Ncu^y>sdkmp8bgf@^#ydDEQxzGQ_3=(Ta^p|KwqsF%l$B z&CY`I+t~?BO;3N4n+rP@PsviWVf7;H+I1v`y9wI=>5=%fzrKu(iU9kazsx)K0)`RM zF#GJA1%DW1Eakt(-NyW)iwIiC9b;_v4j#XQD8FY9Ij zh#;x|2Z$5m{d8!ZWjo@Fq(U3K_|ljMg^H0WK#c?hCdZX2=}d$mTCJ&<(y!(#yX zT3qYr?x{U@953$Mtk9>uXMCS=z(~BpKbTRc|4D}#k1N7yD>z(i#)*$#*u+B!ZYv*` zbop*)Laa3=%hf{+lwQdU$j{L@>3I^EQ9U$ZSvPUx~^mqwL)W)50K2aU)_%7KY)#r&8nfe_Ye@>ERML7bZS?M!(Jv zP52j@G1^dkx05B3zrTAfNTlCn2dP!haqcmrK0xm0F242>xn!Dk&cEf+!{B^SZT3KOM zR#y7xRlFZSbH_JQD>us5FEF*_LH`t%=SbDo9RwWW9a;Chu(6Ei?6)S_t+d59x$_9% z`ZPL`Nvx;0mpm>muEsT+7~uso5Yt9{;^ztHR&^svc9$rq(%q5gXY{#a#)hEJH&0yr z*Ou2Sg;qv0Vy^_@)aKeWX|P$RT*||6`@=Br%!5}89E?#7)}O4 z)Fqr=v9ubfzXJ@2D8S9he6>=4dVBQD+ARome|?exmRv$EGP2Qo)=${*VO7Wj$FV7E ztg+DSgMXZf${x>+4VUSP-vu2A19xahpv`bEsi`?h&Sz~IvPf`OsCA=s3_KVrDk_Cc z3{r}Tv;v4gD}Y~kzkJQtw<$NjV$H>E*>|Kmh%R?*w#s0!f19hjkDe#YUNbr|n7m-8 zRKL>#r-uNN5#s2WaR!;LD$y756>sB?6KOZ87MKpCHj;w+=05 zX=7mU(3j!f8<2UwVpeaKJ%7ij-y%xsDg}4*XS39uS^(ZXb$Q6$qzt_5Cfat6k(lzA zcAxy2Ke`wIEf&{4*ks+xsaUeTp7>{~0Xnkw`1HefL%m zg=>6w`7I=gh$){iSojY9&9Xb7qjdt0E0)|kJ^pwo78!--WhZPAwXG*xDQWGMb)bXr zkNe-H$8BD8TJ41PQb$Z$5BHhY8s%P!%A*)RliVn@}#f@ z;vtWjga#pO9|a?!2c=HLl+!{4#O%|Z{fI$B{LEMImF-=<{U`ozY{Zm-$~2_Eh(OxT z^g+fNktghfZZEWBt0^u1cj37c&`H(sqH?kSKb0PQSD;X~ag~&1tNX_n4Y!H5UhGmf zHeP$;V=DcqwzGJEFfV*h`gq?gT5A3S#Sq2NLlq?5Tv}YLhAfrdLmRh3CSfBABx#4r zyOkG+_lxSU00U=NDP&~8Apd0iT;dnT{lThrN9Qhs&dnRL!QuX{gMnWwfnRfBPl(Xa zIkE&~qEx4|~ z`P73+jP)FTa^E;IOUI`@^Ge}LLz%1|WwZPl?QbH?+ZvCy1{i48{QSjcHsQh%ia+i~ zG3NgL;!#^+m~K*6{jt$RD?)7*j&PX+Y4yuU}2N)rwA)L1Faxhzt+~_Yph> z)rXH(0Gh=nAJ?R@yNsw4n(u&}AS1x(wW#0hi~XfK^~<}M1t5IIMA8K^JhupICtIwC z<}!Z)eqgi!PPHEij7t+!5;6$1KkezT)Mu{Q(L*TIY%y-Rl)IGu;;|LJ>|~4SI1kjE zK-*An@bhTLHgIQUr;j<2YKej2)d%29f4z#=(R=L$FZa`4(b*9c#2QEA_(!xoXBEgS zW48YpVKT*GP6Kgf`$fvSmbV~#T-liIKt2;6$VVWF+^kJ(#R(hcsN(mv>@MPZw;pa= z5Ha}TmJ%sDs-}L=3+V7ad`tCA9Nf*{OLBb?eVmEmO2Z>!4!M1G#X9ebL@Hn$!RP`f zyc;TiL?%MTast`JN3_CYm4~q@&uwGLw?t1pAH?YWr%B0Mp0n50W_HMeUr@2rj55zr{@6y-DIj5iaGTBE!X=JS z{WE`%kU~&I*kB|}d~XU<|LX{it**u(xKQGM?N7x@CTyZqSMS z*+P`rb5^jygyxS&WpdHTb;Zm1Y?L4k`Vs{ zsk%{U4ky@oLh5(KUxU!#(%*{q?<3dO&@HRy+^g2Ci_Gy|s!N=rXKyL&cA zNJKbo^n!ARrw3AbWeDlCGCs>&;UGC=W>PvNQscRNJ9scj+C85$*7dP{vHz%6ufYx) z!Nv4Z8o8=#TY387`_M^Y*xr%CBz6v;)M8UyTD) zGm!?Y{r39HI?Eo&lWyHW;faYc0KeKf>-@>#o`Hd+*X(v4_SdeRuk;Y!7)O*6vjHEw zTEJF#2(eJMUTApePH7<|34zB|>CT6rgjBfF-AdRoH$aPAxM7^n&QU6`2&93FHyIfj z)?@*D_yMdQ2p{Sk*cM-ImZ2voDj7joMvSK(?lXL;wu&(dG9PDc@pY+NRK=VEl~`5p zP}YG`uIx)Q5Z>WzL47kyLl>j`0^8DVis$-5vI8Troz|om0Cc_fJ)ga#!u3_EC`=^Y zVb2AinS|ENwZj4yQFxn!Ap@F;38ni5h6J122#`vR>)vnL4$#r>;6Mya$LBhxD)#-= zw8+X7v;MlgxR?p76RVEsdLDt0(QD>gkf|>ZwiC{l$1i8ggyPb|!~v6w0Inh0HTMbG zTn-BP&TMPiHNb4}!;ALjEoDPnvZR3H5r0RjZFmItK?u~gR5N+niG-!d(5HYGZNdzc zGd1Y4{Ul$b@5Ld6#>t=Aqa?OnAOQ`$o*VM+65a~9{$6J=?Ib38+Yun{9(W?IuD5O} zPqe)Pq~fqPsqFHt&Ol`cdcg)|C+#q z4n&=rwbG?mj3xNuWf&WP$cQ&0BTiGP4EuA8m?bVn051>$+#)(!7>HOQVrxOT$8#2L z_x~_I89$r*H~5-%K8?fmsWtx^i^nvG&f8c%$$GSd%tHy3Ai zI!^VcptVEEkY9iS64pE>=F$JUG2;-&R#~ezWgRt{t&n_5v?G9UN)pqbTX}lblKT95 zL)_Tl7Xb^nG8ut_vZ!1IDptCoNg35W`^&e>qrd~fu_7Sw)!hZSJPTZgtp1iAMfx=d zlP6RSrU{0}2K)N@&Z4h-Ov7E*vZN4R2;+~!Lu~@2Y*YwxdZ&PV7RKCrw_QLP2sd5* zov(<&L$^9bx1XwBxMknn@%G)Y`7rSTIqJG%hj&G?VBcd6$PQod#XUgWr8~cz%KZXQ z(HChH0M}zP?5^Y4fTaX!l$+S7h+Nc)w##AfIT6&&n@4 zySQ|C%CR04d4*n+W!seHh17h>tKqIxC}#*gNxzQx+DMmZ7=|P)5P2R8Al(Gl0 zTO~zl+Q9UlkCMkkJ!h@X*(d*d%Ew68HZ)rik~QvHb(FPvJ^Tc-8pQBYy~B^$#Y22E zgVTXWnubOX0px%iS|r``IvUCkL1vCiLabFYGc%KSW-Trbf{Bqo8Vb;U02h>u6>Uez zStRt`+7F{)oC1uAD|uY?S-$3{RU)UzMV$QIo7-815q^`FBs2v|O^ z`p=dWLZ%{7F76U@JsJ$c-Z!?7cSPP7JqjQDgsJ|=R&l&8bWqSkHIz7dkWHMzH~lP% zmw?Sh7bcS+{6*UW&*5foQx_g%es=O-yMpFG?7UR^^5SFgv1EyC=n0hRVwei&52r-_ zxT)xieg)!qGP+h?u>~8;;~3=YN2R(w8c^jB(C?g4)k}h@ip%dOqj!%6)ksdWRdD5u z+`d8w0&6jm$A_&p5N4h6oh9o&OrhX=_h94%{-ui@sBy9l;{n7yXNb~_bU0^Z<2 zx{V>Ozx)@2$;Pf27D4a+Cw;y72~`sxAD^GMa=s(A4=WS6f2rFOQ=a6s zy#tU#CDVzU$8UoGMx6!GhKgZUn%t%9*!t(YYeQS(aZ}bt#=)Im=tKsoQ1c4PkOE^* zbU`fH1ds?Vda-fc^(Cc1ImXNw<<^`pMo@ ztkBmcfS~~TBSvqFN&(=gGHHv{G%4fyPk@?{Y#S8N_REA`qeW6A3&=`{s+WwUH1$Yl zin;x@G8u)^P_z}04sjW_{KpW`|6@X7PZ>b}4NpIz*zHZgYV25d+24@=%q=eNN(;QY zZ(PSj(qg2^WPKHy4KQBdI>>ZX%PSHeET*Z~1^wTeH3OuJxDSwz82ur_sfu729yYUD9wz%$h;|`7b8cf4it%vz{DgMWp;>O^Hf6bK%P_5b=NKJ-~k8o z;?bsq!hoNzCrrh6nQBk=Hs^G(gK(@-wUT}I-U5z`IO{=tN&w9YyuJK|UHq9|gvTL~ zLf@vo*6NF@;BG$!D=7R)t(96j{*Z`qsLeUQCt(Z5KL z>|0TMa~Gu0CbHm(AYb}b5c3&d23X=ac;`5a9?Tl02QS1W*^7!b$mj#nMU$(5R=ZIz zy)Yygtw@Doj7sR|rwhBW0=oj z2A>d}wIY@(A$vQ!jNB*8NVkmwjR=|P&$!k1v$&9|r?b=3$apGoZg@|mLL5GT7e5N` zaap-=iU-~k5@gD^OWsf5j%Zyzx9_DUR4wu)Dmh%gt}SEU4kL85F{68xsaS0ddq$NY z93y7`;9k0Q_Z)Q)r0G&oND<;9Z#Q#oDXBSxEBHYalSu#5*ce2AKWS`aOh&L=;Ywhy zoM|0fBUKNDY?-A6UN}`!>SL0HBPO$bFP^``ix6T&^r8Au=%;p+wvL~BG}6fxYyCyT@%%7ErgW@hN^LQMz{%~JbVOOo8oHMoU= zvzg~Gp(aVdr$J)(WxH;;JAy)>SPFP*{cSzg0X_Uh7!4)(+2#<(Z@CN_h?{NNNeYGX zQx#o@JGRXn*y=4WrY|akjz%m!J^rvF_6TMa1yqaGA2_UPVhL0eXm{VLdp*dMJ9g0s zwU-kvQ!17&8C>7F_@J>dK%6=0q27y-o8KQTdK-0{jAFBEeT;(b~Rt1V1s<06P9 zgF`Z#kzRd>u7Gc!$jW+xd?);$#kF6ZosAP9h}q?pR=d7YYdHGLafj^)o+u!AHSZG{ zW~*X;r50z5ER zF|^+jwd?SdzW6UnSQk*Oj{-JpyNa3tRdzE_Wk-JH%9RYsF=-f^OIPAHXV%@}OcHPXMsS%kic(CuY()COrG7hpiB4F1BOWCXA#$=Wa^2&WDwRD5r{>jiS^07nf3 z%ncB8B*uzhC5a1D8b7G*Q_sOZkd;*sxX$ApTiF?1h#GwKM3?rZDFXV3$;1-wpg+zG z^79q={5&P3#R`%6-$qSMG3Xi8Cq&jpDmy~L(Fut3ZfhyjIW4Z?XQYN4GMPR3xX^nS z11^y+e~AO1b$6hSGFA!0k5ZO2I4Wd#<8VYc$ixn0s^@@A;IF7YkaT+r|1ld7hE(pP z<|DR?sLoy%0XZdt+GG~@g3stysEsjsik?HLMs7C+mD9=yV^Lc(k0Af!xnZU1Fa%dR z5Vga8p>2RJLI4`{wdTzmdHZ&8n7z{pnF@Wx31hrZr$H2NjNiaJULBd?uXTZb-5 zSfjv*e|-c-9C6n$Uf>qZXN5-6)}ICc>V+-yZe|Cr`=ANarZoB=%uDnd+w1c3{X?@{ z?RH3kMe);8pod5y+6g(OqA=++%p1H5jDZLnAG&2h$X<+05zUPS(EIbYP=(jg&5e1; zdf@xFCZ+GTPpzR*`fbabxoPXUr3c@koYL%2=o7R%N{J8zC#o115(B-vR5IJWyS}(| zl-}$LVuZ6~j}e5i>5l@tX(7^Mydptp(LE`d)U3)! z9;v|{zXmvR_qC2$AzsW*tVA#Yp+j~`nALU+N|UAsPaK_y-H+LE(lMQcK_r06Hn*R| z%3<;$5{`n3e8fe922r@obm}C>^nlH<%gJ5Yu}Ka1I;-MPHg@($C%AZc(VXO%WofC6 zQvEeiqq(KWP*w*Tt{2k6iPWTW{zg!%j4}Q*jg1x@a=Vs=W}F*Lr_%eANRJ!1>HOdD zom7Ysh6Xqomj}#~}sn%xA!BXS1%@mUiflgPX^#URY?i729eH7G$iK1duRwnu8m8;1!k3 zPU}9-5%e10`HT^9Ke@KC;f>(b=)Ii#Rc>?Ke~jKt0b|VMo>9-YAj8#)UxsBV?D)~> z>m`%6u)Fb=h!(w9ZRjSY(LD$_85xm;r!FH<_X>1j(WwZV%qCP4Det&-Hd>rBN?`AO zWY_Vj`i==(5eL*T>5j$sj6j^wNiqv2P=nm-XHQ=AMU%q0&!s&OMeGQIjca&p{Evd- zxqkdNbroBYHpRon`%nF5lyDHsheb)cI-GF`%WtZFcL&=Fk(q|sr;t+0Jp+5b zC?_;IBSAslyI*H~ac3bjK8ZnR1CEqx*WzmkP24dhV><`E zuDw@JoC~q`N2X@qHDx#+XtQ*O>9vX2jPcDJ%Ic1<4-ptEJ)(u z)VoMA6BpS08sM|~!iNl*l8qk$@1>I$Nt=eGpkZ7SenN(n7UUct0*7vF&)N8NQ^9Tg zgb@i)rk?zj+<3YA2eMR2tdaRk39M0m@Xwj>$(*WFUw%F=ra;c0WI-``tBh~?W6^r- zUS-3H*L<9X_3mx+c~pkAZAC^!{l!&Pa_;3l3y?O9@w#b_Z`i&x*(e*D8)C(G8r{aGz*F~t67H-!yeE0!A zJ6!cWK+xFb3fD`N@?4XWE6lFu^`jS=&=zo|9ajYE9ti5egOC9z=#7{DPGB4^LZ+h( z`A58$Fs{gXvt%6RsoS~X_Hu|kX+TD+J0cU^Uk|g-#;ncHuJ_LdFrR1n9prO$UB~B_ zj-RhajF)}i`y3U|@2bYE954<(-0*??>SRjuZOG#@*Kwj*B>63(Of@L(qfJp~5%;?+ z=Yb}y8md};H%NNXt=a;Mf*qRQ)KItK?rU68V?V;{2z|)c;Y&qfwvgwn(7Bo2ib<&p za7ex|0dK<3FbBNlV&HP&i<97hRf>rQB+O|gy!F-#OKK$-jf>P9vyR1aP%era+vF%5;_TsfZdkA zw5)&m@$%yBbGWo2Uk&4%M#oJ$y5Qi1KLd}y=8~Nv|W= zgyeJQB8FxS=axi}d_5y+2rGi5X>2+hUa;b!YV&v1#SI>Y96(=@^~%QLs`cqXSZG4+ z^OM(y4!>8Tq%Gt~|MGhFoE-g9`M`SixJc!xc|<-hvX8URy4C$aRu$av*}(4s@RM8i zJIlY`B!| zG8gLLHL;#22{Y6`&9&4uwOacuS|8Y)ml4YG{*_UKY%p618}3#!5SViiK0PM9nk-C) z>;%KZlbXt*L5A3}(CI1Yj56pbHQ1p33et3SHR&@-UHHXCfTkofGE;Fm|77+7f@ zgpDp#;V@X9*WnzpR>DayZccEB*!wGS^VArju$%lc&~4`ZNo2Bqjl+7U!-b8_@b&I+ zL&jS%T`IC&LDouhK|W_Uc)f>s)usgk)4k`Id_>p2Cv99^hGi8}(!ymhSYKM2>36m8 zS~ap*!D_H8Dr>*j6RPW(pv!%eModQYv0dkz$IL3=WE;H55d4Jr(37x6o*LQ8(URG; z`xl@bzI=`koZLT_{|FnXRG7oO0wiGbZ*WT(m#|6H|l@98dY`+U71`S^nDl09L! zkOe2%C0Vn+ZTvz`HLMB4t}8-iLpp=+E`VWgpE`EC!?jInZOsd8@YXQzoh;2GWcr(@ z!&X;061`x95yO#?kU&E{aoIqWZ(mi=5u<^n+XJp^T36*@{l_WN>9>=xB;ul5Z*PC6 zyPK0S+dvsz%M-Qop(_uDURRBIR*o${&c=j4-K#9OpeRT0Ju~Kg{U=dsLAIL}{ay)V z@yb~LE3k|VF5`d+>%CP0Qk_?ICkvf!*Btk%83~yxWZJX#m3rgroX+ICW5uu}ZSG+e z?!aFC_Kdf_8eZ}>4`I9Cd-|cSNAIF31Xp|V-Dcd1%lqu<(u2Ch9X}O~iuyC`b)0Jy z-PnBC1)kJF3#OLQAwByE)p|Yz(jkSRf!G0UV`f-Y;q=huVcZvsk7w6Ji42C~JYf&a zhQQn7d+>`>l9t?)j-PF;EhfK>><{5+g@t!L?^`bBByD_8qM8dmA^PSw;;8)0LDsR% zElXxaw}zSbF=YxbJNs%U!Ujf zOhvh?YgLOh$J%sr!io7OpZiMZy~f;E7xK6dCG%Z~mlAagzjv;KoO&R6LY_v+B}q$L z?_EBuEM+=Jp!sEZuI|7r_G62_i6$jeR4eSV@@#zvu+^?C$W6h|%N#>|+R9m3^DzCF z>m1)iD?j;=xumkO$L{6kM8TgXG}>;%iaB-jRa^L-hKwZd)vNDpQkp)MuNE<0mtB%F zAO55i{jEr<*{Q25hqvq@54Nay^bW~A-JK4`0nmEhka{JVv|Gvg2)`?d)PYVWKoP$kV0#A;G2;3(9d}}M;&^= zsKE2q%jwRGZtClyga^gaX|X~~7)c&zMo&`-&az_bJ&592{S|Dy#y!sQbX8Zx!0`>C zRd=rGp}xb_G_yQ+-l*IBMzQI{bRD?8pz(W*LC3(9O6sexiWiBe%I(0Xg(ue++9@Y~ zs4AF z6*>2T@D*R1=xOcKqTCc#-85&A-Cnoy!bO$+h$#7wJpQ&UIXsr5V*fZLMJ46wt1W)} zlXjRlwY*nVGm$ax8zN7a;QvnYmelxT2P0&YM)j)%i4U??ynOs>+6na(k(d|u;SJV` zXuH{KQt6DeG<&xH*hrwJUKqos6!~H)uh5Ih8p;G*Ys}2VsnU+qakFDweUJ97N7)H0 z99k20s9Jw@RN>sw#PapMy%v$2afFe2B5Mk4Rz{`)40mj4kL6wTJ^HHKOnZdRQja|3 zhmVt-_41t`4gr@-D>haqJLSHKIdh(qS^9G0gm@kcf%3g7730xST!L%=)pweegZpdb z#>9*=D;9g+NO+e<@|@8>aNt&>Xk9n`jb}Z>H`jt>3x@CBppAGCcEbq8)3tbT`bT>e ztvt!5?xFeil+?2s+SzK{+=Ib~#OP7SdC8v1w&9kqW7;363Vlw?*|$Gn8n%y8&J5Jo zU!(NH?hp1QImjK6flK2E3p`zMZAjr0Nyzw2cb@s)!3!n71=kh_d^W#YUdoKzY%d03 z^8V+($O!k~>HGF;GI}AoKi(!Yigb*Td+3tTeJn6&?gw3cFBIEn=Gqt|x_CyaVFozJ znIsf^R`O@L%XhQjOVn+*?-lw#Z4IM4Tk4^~T27Og)n z#9NyhpNAcM`%ka}V~qDqZ)RQbeYwDZt7UBL0W4$^rwLA@2~Oj9)Wap@Ch61$q)o+( zXFFa#q_mqp6Z~=vG6vH%)NeAGgOm%0B}t9c+u21e*k+#i66Dpvh*}>gcG(fS2@(j&{+;H*s#5|p5+x7?teC_m^t2ZLM+5F zCalm8*>R5Y`T^C*aSpwzbeJ&5>OM|(HJjN1SxocF)g7jR3ioZAqw#Msf((}mCU0#~$%gV1X zpt|Ja)M!Rad|8uObw-I*T7i>}i&T`REy2J|(TD3jwba!n6Z*P%3Dv?Nh!iu-aF9u|L}g3hUb&w@{Q1h4IK< zaqV}vzR_!6{P{X;-<}wSE!26N7sW3q#C7#{8#tS091`JYVCE7UvM>?QWgLzV!0`p& zDy()BPp8)RJRbCVMgZ{uw*I{1^mVLliB~6ZqP18Xv!EJr#GYt?-N03x60CQ#le~Vu zuXiR9rZF)gc`UUST9VAF$}js~-z4gJ*1brnh?$TD=j_dyB5v}KdO-!56V|7Nf3|lh z<&h8{ScdiB({H@2D{&-Abaz+BV3~Qo?!&Mf8U)JAYKBq5$xIg#WQs-{eL6}Wnp$?G zX8rXIvPgr8bX`|LIxdbvyQ1EB{|c0gh_17?vkq7;{w7#4VoIoXy# z8B9s^XV{ATVQ8?fA)M@VVM>36XLQs>ZV~RW30M+tl@olqpuAz7<`JKR;WPH7N?{Rh zohjcvdm9%04tzTM@bKp)JEy7T8;jNC!r~)lp~tG3%A5tV3NK+jM$29kDmxj#TK7Zk zFD@ow$?Y=p0-GYsVZHW?OEP1kl}-k~yq!_{BZsV0A8e8G^S-pP*iBemUd|Z ztu%SFUSUe)jVy+ltRh!|kmS`JR>51olhSJ$EQI^4)tut>S?Y6rb(zg~l+z_@g`1|o z&ZjNqOgNR&T#*#lUKRn^d17}qfcF<2b6yVGDikrO&s3M+-237`sPiUz=!y|4#kTb0 z-2qrrx9IHinr1I;-b({XnoVz5N*m7QOjSFW^-256QMXOm2O;boO5ti87Ox2C=^MttrPlw^R$@EOvy0Ck*Qo|l8_SN7TmMOd7g(_L90=P zKK#6Z1FQsM49uU}+ge$dhdnhyapRwKjEkPq3Oo@Wd8utOa(7?I?AbaxsX!8m@wkC( zSjycnI(O7Mtw7_d`C2_Ci`cVRtL_DpwDLT=?9y7;qkbyymu5ptTrrL>j676sIOJI4 zecnXfkKu$&2CXs!^?hYCWs|H5Ca}ji-tu>B-E!?6EsQ3?$hb;?YJd{hHZ*}4xB9F^ zl}~j@(<#Yya3{#;f7DR-ap5oEnc1B!9*4U( z#qsU5WnRL173%Z{Q=fNFghhJ1=9`*euc>BUI-W1I30!G!`dkO>e+LNDBw-zv2UZuS z?Tpi1JFX?)oBM2aEx|2U16SE9CFW@~FYmP6|&pyv<2fblWykt7B!fi>d z^~+viqWg=lX+ADcV$^$keag%|leF*lCM3HoHLAlamN=BtZQ<??gvF*`f>w%@z;QREKh$>^UDVmG|u6DcL*X%PgutZO@c?)Nu1~+t5x8 zMqO5W>3Z2kj+O?4qOCt|(m`@Y5yL#%69wo`^5(q)J&zJWScteJENgdJc3K4ZR5-G^ zAw{_O3L`4Tw&{X z^v$T#32QANEUcV{^s4BvFx}JqTR0f82cw+1@>%7xV-2PKK&;u-J7*41$Xk0?te~eRxd{%BjAf)F!z&bh!F@mFzRpOUGC0qs#ZW6 z664@KqLCbve}v7taT+#cw=$*-kYuk}hm}nB{nl$bGckOlX}Prh*`prZJTI!eu)6x$ zdb*~X-K-#AI@Q1N5T6hxIXF1+_M%>#SNXTwfjU3QZgl8G2$}KK&|#F{h&WfG`!)Tw{Fay-Op( zI$juBPr<1o&3r(HgBaT3ChrI2iC=>=yM>2VV*Cu4H(FqUy!g@1)8DE#H$0Esn$q>l z4X;J$pw4VN9I26V2ED|D8KO>C&l#8lK0G9rdaaQds9Gw|nKK0! zGuL5Nwvytw>c^mxcgMQ?+iSqxk+?{s1-twJ8`!q6!JW>gBi5&CeUZFc3ryIew)k|T z54EWJ@-rQJYVNGu`wibktWU3*DpIzQYB;TV+>4lp4briMjJh~RIau92y?Xz`E7yVB zU6sn^28BGv=%ia0mis+tKT}?`*058UdC%H(fB1+hnL%<{t2cM0?B?3Md~b56TlR(( zE%S6ym(e1P@3GFA)6(A}aJ)%gX2&b~j_t9$Cw9)IJ(t!fe(ntot7!J%UOF|CD69DS z5R<4vc3JXL>%3=XTyjk&mozTLTzxzrS?1bvl~<8o+Zqhl^Mu}O<6_14jlNy7xE$Ez zn5@6P+Evv&KPjAlJ*#3e%-}fvPGm%bDWf3?4!+kp_lEdD$G85o4tIWhMHYMO%jUDu z8mtu^4^X*V^zHGLo}aQOv}7Xoxz2;5T!`YSk|nl9Hkf0Nu)Q*}dSZ!kB0;ev9;%6q z{X#fJ{X_6gl`cr}kJ*w1E0R>2%nsXW*w8^+dD6?P8;99z?{Cv6H⪼X5d8nqpb48 zv^vBsEd}hqL()BSa`D7H<)#8rbNz$EEHM@G(YBGZ${$nnta%`!vynqdzzxGU8=DuV z{Q0k4tMh1xbHUHK7kwJR!ZJDnlZ;mK#;KtSx~q;}hK@Ekl=bNgl697aQptMDMA6-EHykLW4KGO1Sb{-r~qa z@htt4YQZrc^U;rz^QB{_g@;OWPl69(1(uL1o@j73AjtS1ma3OUuC-{bgq?JpFNt^z z`q;f-H?`5+chSb)SFbwR+$p+!f4!P{{>GV@(>Ik)`*9g@$^`q^cQZ-Ybqcnpsz#?P zr?O(&>LYoYuD(dnp7Y+=IIU~))|6mLwr*bREYS_eV-@{ z%1J_0%4ev8BA@!#;}CB#^q=Xq_^aHOyrxZqe6VF@GI*&wsT<<%tX!Lfo1JGbY1K5G zz8})9Guf46T+{OmNr;FCqCcThl%-i&{a!k`9?i#%`|nc|u}*Y<9-ye}%?7J%Suv4G zAF&Ptm7VW3-@$*=Sd7Q~SAgzXMy8d7Bd$SK{bgW>dcbmi^08w;EW~ zj{HBW-a0Jmwfi1cL;FPAK|o4rknTo! z@9{k6dw%bAaru`pGoO9md#}CL+IwWEL_WW5MP3H0_f|^jIe%lce{w)NCsg9y;*RF* zIyR`m{|~Em^;U}h#}&|jv}tF}z9$^%`Q;MG=yEk3jecEcRa4XMy**?jKT)J+C~nzL z%BF>U)YkdMam7x64z@3-lYpsSK}-r24pZcODc;vb6K9o;!-E@VqPJ<5i{~;TXmp;9 z9SLbE)8Bzjlv#S9-Q-KeP>YLTVzq!TrpJIP8|iOcj+gp9FjVf@2QLMOu97!Qlz=;*<}Kun ztvbfs%=R_RNRm=DV;l>ACD=WO1?HmB-wex44iStqXD%6TY?ZYlU7v_yn~C9Hj;QBs zET(*$BeWI>fWngMS5w)3lOsHlT~8vaBU&-#&6((wIpuTN(8y_p+FQEU<^fj6&Zhm* zc2wR?U5U{uq_sqGDGH}wN=cR#7z$>Ceb&AU)(i8 zsBCbwP3k^6xI_P*t|>)Sdh4w^J@@^!?<*ZVYM-3D7H?^bxC06$_4b{6lQ(NbTk;h4 zM1YFU9>o`laEA-+bZ}v#-*~4R31us`zvc-^M{-)Fm-x*GV7#e)0VORTk+kR~N~=tA zr2!qD{8sMj?T(LAtyprT2=Lpi_Nm8W$-2=0Zy8DkuFrwsW)=Wd@%m+v2k*k zer!+(y7+GS`~sr#Z|WSx)rhCA0VAq{Kh2TU!R29(v-+Y)EFiIQ&jvqoA;X_758>jW z@7ZZ^8cMh^k*#H8Ms~%}IH-$1a~Tq^&0xEtu<-M)0xZ^KS%FEEg)mK++G^VS5PP4cIGFXC)5&oSf7-Bj4?V0pgSl8|)7ajv%G10n3I&ZCPc#i%Mr^7;9iUX00H*i6`eeM9#8^EN?bJT|$4v=8azO$G+ba z#ZPJZobCI)eP*@~XL_!CN8<2(M_2a5SJiqU8IcKiPOM%BHc__r{AgEd2Vdmh85cdt zQxh17YU{6=R(yLBdQ8~1q8#TzF8zMwg=&wgOHdqCl4`BA8# z(Ce=*Pb=R|{2F+qX`A<7)k!2UrK%m%RJqGl(XFfHgg6tPl7wft8Ui%G9^cM;#(Xu_ z;g@QgUFVojX`(PGlo%)#y2&$f3A%e*4IC)Z?sH0_{U8wGhd*W{7nPqS~5MPpm5k!e0lQlhWun$KA)d#QYs+axHd?)xW2q^2)l zI9muM2GjJU_`eq0AU7eBFZ8H*ah9xOn&_Ht=JAd3?c`G`_enFV?BOn_=r0mBqF2q2 zB;UEfU$0Ji0u@LwcfJdN6ND^rfIFuiLX3jz1ArHivU#oGZl>XbD5D zUxIZ!@=@-5e1aPr26HR{ai!)~&#A0b$7~$Y*_)RVo^;tQhpP8g#6>;9P+-+jQ_avZ z?ZB77BeW9TZLGuAkL`T_9?7z$~&BRg)u@V(%?CK$lNnm8)9mP~~2$x&e*&*axm4 z&mfZLNVyEp9&ew|mqfn{ZLxYr5FGWtH<6LA+ap4?t$-#HBOXRe7*K=>NeOzQvs~Ye z%F1d|8m$Md+A!;dzTW(E}PJFQ<2yFc^_9O$szHk(GXufHGo>z(>7TM74I= z=GuT#lUiK+s{w)tk9QoI7R6#rg8qw^t@Y6)yYJz!8vaOAGDH^}6-}7H zp=Y}t8pb=-?u(qxdKeFMCMyMj^mVd)5I%3!Xw#@)fcDwVf z@2B5_@bA!xAw{IR>C|BBb#_q1kl&eVw8VuqoAQEix18R6?8s)&Hn9=A40TZwt%=N` z(4S&y^bqQ7!Br(62X~z8mtX>6PnN6IsQK%+E>;p?q!oiO*W;k|hqIsR>B5 zVg=gr)_lz9kcoDhxg2?{NWJ_DQ#bh_8p=)fAR(+TqsnY|cNcp0Wp!bki$TI4t~K)f z_d7{+j)xX;=`S$i+!Y0(HK0Qv&J1!n5**V1P0$-#Wgc|Kpj*aE`2hK`4clRZ&LGBw z_$%&x%MM{7JiV>#d5;iJjBeG$U@B#n#3^oOEfxVl%&@6tw^WsT?EXsT=t2`uzox?_ zLDyz~__N>^(M>w>ysKnHj=IEo$0z)SsyBrE9Ch<=+@ZbgvXd2*b7y3w1p0)rIrO60 z*Ro>OA?i=(HI^#7+EaKn_x4^JvHCubC%#yCW8~G5*dB_n?p)uir9>;Xs`ObpKb#i1 zI}6IruWRZ9iVcP51zIHdS65v^^i-^EHO|a+@ii(d;VP{rR)PlBT0#7hae8N0LQX<+ z0o~NE?51uXJ;^FxefG2K)M4@p;M9^YgZCQ>s~Or7#O}%$SwW8zn8N40YssFLNDi=Y z8(1W1+uaoMt2tHtIvy~2a@e$?HkSNFA>N5Iz70;~cOY1XpwUtt5M?ong z9lQM06F(_`%60>`}&DcgZ8M? zNuMt$N|vBdy_iOwVM@j^Ery)jBP})b)VETw00fj0cOas!CRtF~v;*8s97#O0=Q+Rd zzw}%RHRw;)_eiXGN9}?3_GLNU!yNXg*e;`?#eloVK)8Df6%Y^mmc{#qy5Q1IK&8&<_x360WsgY!# z?$g&`bh@8Y>@SQdP z78vsAZcO#520t8{uVbne&D4@%7gF|tz$)?Z$>rYrZMA|+aJE2U z+?TtMvEt);$|%E;b_F0sl-+%pFzc3%cP<^7t8=1C-BA+DO)Oiti%49k$E;I#k7evaz+-%~ z=i);PFRZqF4-AAJmiqq$(#n>ZLlDoIQCWV~53xhX*yZioInIcX)3D~C*ER>s-$L4t z&18o=rCWp+R5Y)>?nheRc~ffS@pJqnFW%{O;2*iS$()7)#jVKAgh-ckDVAuL*<+;{ zDF0yp^IF}G%YLegpq_Zb*d7*Pwj`ne9Xbdb_rvg<58AKuwiR;F<*O1D1%r0z4aFB2 z!zxhxuYsGe8C!c-c0wo10;VuWU(EeM9)TTUH)RoS7@BrPl?-_cC(&*r3|N>WQd+QW>4B*)sj1k~bqEJ(nFnl9ng~_R|At9_ z9kH=QdLxdu1e*0=hZ-^|@e}1y8fmPR^rqNNi4y zcZm@49M6UIGenUfDc?|H!SRS}hthqI(H=J&)*&ZmExN>q>#y)p;;}(#FC%?yb!*Sh zBx`TGa4mA#=Jv3$R#pxxeYsS{iC0O5f&qJ}Ge295WpZG}=fYlV4g#O3u$wEfx=3GUN0U9(j0So?$(2FjJ5=$akv>n8)fbC%`@5O{N*(!Wznd zbnWTno$TgG9vCBqSW&%vb-U{_pE5BbcNcI0xxUOquE>^{eEs#3dME z8xU>&^osOXPQ$lMh?;RSE4feLkgT#3g+19Sx<_`VD|k!gd0o;GD3K3S5gsoc8B7!8 zd+UVq%1Tx!lAOO+LUcs%N(nGqHs9gg&i%qj_1-PPU!Lvvqks9H8XbaTJ1u@EGfxQ- z8v9u@=m@Bdvyy*p4f>do@m2I2yi2<`$usJOE0q(RKa|bKIO&W1!dq_$@>#ea#^*@i zVfx!&X;6U7Qn=JzkPN5|rn+|+YLda`TE)vcs|!lt~Jr zs!b2mBBk_9#{fNpS3i?uEAzhZ6ZSJ*=XwRQk;uu$f+{GXgj}%Wt=nACJ#zpD9)`!w z^@#<`%NZOVJ`E-n;at(3#W9tQQsittpId`B<3y-0ZXC_2!^D)>xbMHUq94Y8GNhy` z;Ko9sU9n-$pVo3^XUscpJuQiCrIfBAmS)- zi^GhY2mrzhxIp0rDi^hevuD)k)akdvL^FUNWbU1W?*(c0u<%mQAj(!oclTL6F&*bu zvK`T>G!9kbyqGrV)M%8{bCuA$^0}skL?yz7)#=2rXXqoESB_s2sU&>r#221?6H1bw zR20h~tO? zSJnVf7Za*B@Ky?;q|l@iO_7J_vSXXLrbfmaA1a~z$F66 zb9=vChLAcWACC|+13HDIoO3!k{$VmR7773AB!3jJmicPXb!S$1=OCmxhBn-lDTqLx zQY!k)vKk17CA(sZJRv54;>U^xu4eUBFneI64@baiQM5-t5U~crwYN*8XgR~<{B4Lq zl`KA2s;>f_q&^*Uh*Ze;=SthT32Kkv+g}x>75R8fhTBhT2_GD9&-`fp62C9L6BMv_ z^5i)aU~qWGn_ivH6C*j5D)k4iklcw_*pc2xzs=6Bo};uf>3PIp8+#hU)U^$Z`5ky1 z-`(S^wcsBM$EW2?;ojwdH_|iARie-S)k$JvhbMldaY>#5b2r4D_wiYNvLO$!jAx8(5}-A8NvXbm>0ldzu2Af4X+9 z(il`)i(1T##H?zL1jOSgb5BY0g>7KqtZ-}Cd)j~ze}%12X)hmr*=>IH4Z`Gf5c0z^ zFWp%gvhXqgPy7nJ&guPo`tQM)wFq3OYQ zUjdB3ufo;jI`!4ySs3FRzEG66gm!ZN4KSj?CDMH^cfRayi9^_n>lVW;cDo(g!L5Zf z(h!rziA3{-O@hj8kwcKg>-VDOr~AzOwj~iM&o#;wHH@KYPB&X@J|uBj^uUr1l4)Nv zB|>CR5&%Vn6C1gXNYRE$rR3{!u23Fg@|bWWulVFP!L3Uw) zv?P>rBomefU^6vUwd%jv<^4F8qT$#EebIH8ls6i@o6%O_)pUl4TLr@TPaymHi?Ih1 zx5Fa&XyEUwu@b@K|7~q;f3@+KE$52{`cl$VXby>XTI&709r4py!r3CAX@~Gh2z6!K zEp~)ig4Dw0h5cc4_=7Bkp#Hf;GWu6dEWR1uN zvV~b?C-GBs2WS%#sTlrDt0x9N5T#7m+0&v={sDpbapl+h5!fAmeoA`$!bj(<-*m>V zGOGUc*DIAmJ&^qD`Go9%?uoC74BlN*c%1?mHv5fg6tT(@0=X{@xQs^~6bSi;51X0G zmFHH>cyL$Rfm4v;EUHyLirfNOL2LnRDz!3{k{_Rsc?!e(Ox1Pz(#snU%|!_oi$T3iDN9E=@S~`k18i5c(c2m!UWlQq-4l4|nYnc&5mN z4%$y&KN#@}!KcYQ-rm$PNoE~2%prS4{VZeQYBlI_`tkbr?#82073p$7v;LrT8$&p~ zUH`+x7%KpZ?tswWxMF%jaO$ijb6tRfq_+WW1pUgJF*8ge2~Zp2`7Gf40HJ5R@&0Ss z0$Nfs<_qMTb`LK5wrp6`Hn_h5lfl;V$kevEbwd^9hZk??1iX(%6zt8(<{)juOqyaS zU3Ima#=pt|l&~t4Kc#jTaRQDKKQ?=rl#&sq0x(9<%s2=*D^i5gy_9oYju2Q3;38w;s&+B-o&&Da-dHWx=IywlzB_sQQFs2o>ank`3a2Sr|xw zCK>^WjXa2v5E@GVjSHXjaKae!jJcDZYt9^A&Hf6FfXre@i$7cnim4ZXKv3bz-O3d} zNitHOL6Of{!d!anuBQL1(`zeRUW=Coxhw@o>4J21S8uR6F)2HDI;lKnM<$!B?b~ z7NFJYS@Q*~Ai65bv3&xig7B-2SnVDD<9^HsX(Z2h@uHF zMmpHG<{E4OquE})2#C!9DBC1sfl~23Y?!|3bPaR|(DlAwz?&1l`wAX_3~d|H8kD@b zipP`zxh-?bIwz`M*TmQ2t^`t!N9N@v!W9`3WC~eeMR6*;W^L$fYK!1+pVqjU&iNch zr96)oe&c>1!uk^#8yL5^QPz%ES@1#8H~4+m$9%E#<%EG)I$34S*8Qe_4~A@pX$n9& z-2u!S)Lu#p9wQLY*ESmU* zOM}M$Z$X5=$HffF^7oQ|<{$?E@Ng#jB-=b~w1e0aWB)(Lmf#7zRA+J9sD{hCp%u>{ zrKqs3vQ1k2B$qlIKS3rdny46OFvMM zo<4c1_+jZsz>`EpCAQMWisIANGzS4-gn0T*TqvUjr0EIQId)F@$8dDK(f0%C!j-Z~ zLbt<+rLFbT_mgliPNM{Wy_2gH8I$e^1e&ykjt4awJ*>jed;Oy{McCkc2U-$2b@bv$ zy^ED>2j1HX)~ol)Wa!#sN9gIpNMWsIe#tp7uKGQTeb*pM0rHSeuPxcc?;P)qVh0wu zsEu?4dW<<#J|2XBnC*~k*1RoQX0o@WQ<<%hujON?jZSnhEg=~8jr(0XL`yJcCwxf` ziU)>%=PLpggfayzvgvYXHl(!S2TxwqA7EOu5`u{Go*xpCj^Jgl#D)x#Y{*yR2i!*# z^C@yYKu5E3Py8oh3%)J7kIok(xwSMKObbP^#=Af6?)PWBHPR z?18i17|XxnU;r_l#&SD9#~Zz$kbqD0+@`I!zMQsYiiLie$cc zNOG}T4NaKlkLO9c-ennO2+>NM%?Cz0eAxI052Xg&ipHaE%}M__Fnnz={TnEt!;qg} zJ$?}cAQiEyN!KN7rfu0Mw7-%K+GsA#`z+s_{~(LjqMV6(H2n&czo$HrbCxaQ5$qMn zsaLyA0{{k@KcQeOcPAmGqooawlf*z85LjWtkgJO@QH^VGyUSeeCjh_M`(LM6MD_!c zUI*qBh;IGjMdj?Xqu(R+`S@lJ+D6KLg7kgj(s76{qnup@#R|T02NTD73usVnh)#U- z9s1mBB@>+4Wc~@zuWjub&4L0nF)*=KFivKG5 zXWrm&yQ}y4cTxStFeci+6OIkQ1Dttss_s2X0f!#sgPG0xleV*zGRH5{?dv`_nPMJeQ03T zMIJ$)eeKVQSzyqShQjaU4t!(cdlg0O$*oPQSP#fP3 zo89KilKS7-5&Zz4(0_kV#{heFYeZxMU|(;6Vzbj;5jpZr4(d-nUx2^Nz9Wu7zLy!@ zW)SkZX}^m?iBmg;i5|kQJL>)wD`HZwGclT(Lpq5BKpJl6Wkr{Z^}Lzry+(C!F?}8| zB4HHiq`~&jWxD37y)BGq4jYz1Kf0Qc2`nf9LE{q-DEs|nG5ZsZrkfOcKbIfyKV((U zk3Rm3uTB_gF%;G(9l(49^I?qnzg~bARvTr{xoE6kBrxHGG8wD zyUN9@?|WUfo_DHFUhd!4IsNC8f`9KzOAWOIliuT(0ibXID+nKAnJW&o#ws+7A{y}Sp`|P z-2B-kj9CJ9bsO*F4Q)-oi=*{VuTFQuHZ{I366OndBN)x%=3W$^M@V9f$v}JFw1Nt3 zrzI14b1EBWg=pExkA6MzYuxF(HtKZ!^P_=#JJUaGdA&3^iL;nI{vX5VC}C_Ji#V?l z9om2lEua8Qg(*`lCdbrJcQ!xcH6`T;4$5|P2)a^qhhJba+7gJArBrpg?WngJ*!mlo z^f?M@LF2Ltx^mQzqr;;nlaIYV0f@JpM*Tp8vkk$eu+%a3T4|{MkS_>St9cgyExQFy zX>QI!!MvTvmOq>&yh1wgUJ8Xn<48ZSAC@22qXFsGCba&m4^~7pfo}erV1`)mp*6vW z)-OKC7Cy2~#59*NohZ^5I9LSljdDy77$^c?8%*e-f$4(emL1N2KQ<`;p4W^wl4w3_2RK`g<03Y?jXekgydf-5QYTqq%8dY{)4vC~iy$_TKJ^B~ z!r^hB06+)O@s$j&_teAWGX3+<}F zY4d2)rBa88nABn%AFgbb|q@i~p#AJ80-Tn)4 z>>DXPbFIxo*YKyG^HJ%wit2V)#PrLa@dW{fVQEwJJHnLHj!gM|;8*NrrSPH;WdDpP z(4+u+Rd#MGD0WuK`j#pEtKFdG#I^a6PPe@Oq+_V zhA`vA75V&$t4?zNYI*cRM#pX{4^r4tuswiIyuc~)6l34D`gFu?96CJxMdW1zXwd!z zb`=b?afoC-?YF3p1_viskyjRdI#4RN6wvxf=Imj7Afq$auT1!0v_^x?Sv{QN@EAq7aj#`W)cF?5f{!`VvoT;>3lb~^-~$a}}J zXrcRY{E`86zvcArNB9^R917mdBJkm+qq$0!p;Yc=tj6ahdM}Er`!iqd*g&xR$W_3E zV%Ncspa0AP{LL*_SA7?eYo?Layz8UDs8uR(_qdX0mBfD9!|f?ZJukqh8g#Cl#%E7@ zq88x`VtJQV1LUWfWk;4)%mOUbd5DFA{In=70cvT8PwSwfnOs90`xdG_{#;ZMp2G87 zUgIZgUive^M<~N@zt}+!{nuA080!Uu+5s@uklDG-HX%J-15(hGAsF`5nIy8n=BSDt z&_^*jso>6uoeDfyxMvnH0*hdO1_n4<_^WKvUN#iaaimuA4hpqX`d{ti+9$ls#0_vB zi8vr2J};g~l}zK&;y}H@dBWl8r-#!21vgyb_1aun&u{Q~Ii~R0jX#gWP)UeI0G!}( zaS7WDOmo%u*{#U5Zws;f2aDqhW8uKcE@VwQ_#{1FuDoalMbxCuU9AdApGw?ZxdblF zP5}mDD87YtuJ~!~{GYnbh&I)2$n7kU)xNE^gJfgoeY#_$c?QDC2e6q(#r@vI3oy^T zz1j%~I>JP~SKHMC-_<4UMS#ECWa$eD0y-x>2swWhvx)@=Aho;K zrJvwY3Jp_&sf|y-FKF{wFxSZoB+DlzZioZ$){UVRj2;3qZh`KHV`~%=+DXVU> zBRC(0QaUSUU*~H$VZr6N3Pa0!n6vr)oz_APP~D3UIw(`kJWnG2?|@8)0#dJ7cmgac z@z3E=O?lK=rq06qsmE~9$H}HUCcpV`NYdX?f5a!l=JGrH3L~d&a-6299IYr?IU4;- z#?Zj>_2;N@6!s`QWfTn5{I4U1& z3D|!C1I1~eEBe?}WOoj`ANelg`zKZG>Bs)-wMF$+h7}cP1|C=oyLP55_J`9u%Sn7@$+Ui>HZG~amJlb@9@Zdp;N;(&(y$2(=vP#j( z0b&s*7yLQnvn?bbn^2K~JDYHO$5D~q4ijHw#35RAN62MZ8iUcjb;4!QoMZl5!?u1c z7tot)^TWRUI=%tAwjf{zK1@0HaSV{ZD2n4xt|=--Jt*9KO3L+~YrvME^k=quBw%8q z3MP;S@LW(m*CeGMyft!9d5|Tqe`fXtRCiPf<722E>l}R zBuEQ_xk_tqR89N?Dg58mJu3ObX_d{707n1eGh7x0`LqEnj|GkvI(le1i{>xB*=b;F zLgzl9Pdj$~`qUwJmyD8M7TuXCR?ieJIz6>8k4>=(hL|e`Eh&32HBgfemV@c~k9^>s zb-=OZLOpskZpAULH20=gckGL=6IxF0pY{bcKBDvkpB78l5>d;hOe{+4=Hz5c{(kp?jWNH`R%ISyJc-=!z zZ}50<7xT-XcPDc@YY%`S0D;s&RPXb>NiGFlUiDjJG#qr7=0)Wi6w*Xmh1 z7(}!L^05UQD(C|sXOJ|ufPq%B6Y{>PdlYu2Y9i~o^V{H*9nhr~O~72f^^&Y9vb;f5 z2*=5>hDB7o&liEe_rl5rviX3!o6uR?J(xaZ1U85~a=hB_P6HK-gGKow4PbiyozaG7 z6NlUC(}Uw*eD&FFk%mE)5YT0ABj0k(^=7qivHWpR1$9o{HTZ!JxBq#c$(#F-@3i3s zT}abag+Fu+Ls{2Hd6U-gT)7#*1F5|98j&MOk;M}R^K0gQxcFC?mRv1Y&Zw#-KQUKo zA}MQI7bP&wa(MaUb6fUDMcMgkF#qw^l}4HuQDQ!z^xDm>n3eQ!!|O}Q`Qc)LH^@c? zY2eZoi#PwnXoHfBlzKskbWT(;*Sg_ZgZu>a4Ed~QZzAg_tD(#=3P8oHE~+hSl=T=^ z>~)^^FmU&pp~-2HT@U7B&ZV5)F$Xc&mZk+rf_?k9W4`?4tPjI~sJ*0J?}PG&a86Z?0c@C88g8ikrxC~W@tfi+rIzdW>A z(6$>^fAoNSgp@J>1ms7?PX~~*Dbhpi4=6zS_lE@sNCM=NL+6pc_Vg} zv&$tx*Dqa7Kw_b!=N2aA0={;{bSxS#)HWEm4EA6UpyKCN@I18)oOOp)03F);W4*CG zN8KBLZTHET!j}tCbIR`8V6vXxxY>uzAFnc-0YhP%0bk@dEk#NTw9m1#0tP@%Oq_b9#oDk(f2(bw`W-ZyFi#r@{I2);ooPzcCk-x zWqEqhPb0a)bhreFtDY04LREAb^yz)m0H(iOI0NJRYLCC}FjRH_`_st`;M((UMYJ5B zGZj;Y0E7;&%3^krVj(ru+7C?|ICH&4g(+>J0zQCs`K0oea~04DwHZ7y;wO4az3)&x z&?nI3oiw2a7Ue}RK$AIuQpW)4?|lf~ei~}-t`X#iINlux|I0M5y{{MpsFn#eFPr8U zdZ5Blo%kn)PZS0EM4z^!QKxrQVK58CG55SuN%I3&CN)XN;M*swuM5p6R>}_zJe-@! zd}5S1>L&DEw~dGslqZ4v3Lt^t3|@hW&a#)=IsT|7=gp-rXnHL_1QXxCjwU44Zt!b9 zU*+Ys8f2d#=~_V{P1FJ+jSwe`^u@p9;Z(YUF-)qd=HE@VpX*0P{BI;L4Gb9fMDyT~ z7<-zD#`uBNR#PmTN~w>iI#@0Xd+OMbpFdSfwD_E2z~B^BFQ`%VVo0yx;`ESY3$E^BI5;eyqNm{u! zNM~~LSa8zx&gQFh|LgO>K5H5Ye*JB92F}8a51?M_z%&wCF(KQFj`?zvx0pwB>8`$l zp(7mGAPOr?O~UE(q* z+zou%{wz;d8V?6wQ1v9CbE2|pQCXNj8`&bYC#6#dG_JGarAf|R`A9AVGis5+%pB+{ zN3O0WIN~ZQe>hN=71$=BI~j( z=7tVk^xSgpJAa_0oWe!NhastYp$5UY@C~{) zhyaR(0vgrb!xf-L@w)W$g#MxUFaq4{HuRTg#9+0?8Phg-Tts>~3!O5~;sat;R4Mod z3qg|mtPcS!%jgz*S4?o$QMRm%&H(xYar#7E+yameN%IDpM`Bld?VkX{Pqiv>ItL~; zai39Jffg*;ZJ0hAHaY;{vPnhV7r@Q*gXI|CKa>>Byv=L<4Vs`DXo%@=;u64dH0j${ zkK477uH{6F^rpYssYe<5%S#y_U@yQ!1e3KAi9%9V>NHN(M7OmFGSI7qoc zHaz}e!O9I|!$iD+j6s6Ko4`VAy;jRju+zT+K5US^P*nTNn-RCP%Sz!cXo=nqC@YAUeq%M0Pe;+jY?^?jBxD&rYWTQ(TM=e~YulJvOEhAfbqzc>-&_7{TWq^J|Pe-B$ zo5cUv3v8gpI3xYmy!aR`XG1dPItkRhc0@H&iVZc>iyo`vAv6ZFqusdGMo?=_CipU) zOw7}fJH`*x@K$L>*S8Dpp);5=3N*e|4;;&G(HY<(V$8?J zE}-7mQ11kc;KR9_g9SH~zIS<6K&AoVsL~75kYRlsgkR);bFIToP-E5}k^)Q!W6h7} zXuup4UIbqoVw9RB<&J`IL@=t~{c$qh9w%iU2VQpUWC?Ws+^E4lr7Xq15UE=a#P4OP z3QXm0d0?xV!;Lw!l}D>=y`r&~n!)Om>8EvL4#=Liu5`Lt7S;pqHa$FYzI;0qwx+m$ z&ci%74+msH;LwhzJNL3muV9J*;2GJjp*L9r zwd}rJ&8_)5oK-w3bLIpABY>zxQP>=roJI}r_LsV6_79^1&rxk)3K&f72ig^7BVzz! zJIgFA!qzIE9RB_RmN^qQni`uZ3ML_($lG#!ZB)RUVf1(YlVejBs;I>Dafks zex6P6`pe`Xyn(s z^i5$|?%k@=Hwu|O0zwrCVTGx8JkQ^7`{$;@susaGiI%&-P%3|4i0Kv3SMA6~UE)6V z>3im6R^x4#0*IA4B78nExTtBRBW(9+d6LzQsLpUvJv z1o4=WdN|uafv+hAiJ~@u>_$E%i+UTxLI92|?5dz7$%fSbW<{aDDWC}T3akU_yXM>p znh>00_i?hfe&5X)dM%}mcLZ*kBjH1i2vi*#24WC7z zwq`M^oz>L@%}K83CfFs?M>hWiDhZ<;KvrNY&jM0rC0qV{eMC=;ZYvcG{f$0TGm>~| zc*W;QVP6&J1rU~9P_`Jy?vPog{ZPeAP)2k0n@hl8MG=D=UEG|ax}F-QUY5=}=eDBM zvs!n`}hL4jpXr5lUYlv_!Gi< zYhcExl8boAEyJ*;V8;HV)H0=8sf$sSthnj>u%`aMzzLvaRbU}mku1@_AIaHtq)8!M zmyVLB7~?kb9;ax+gcck4Kt_cA1gZJNUCGz|)mFK}{0x1DNQz?m|33o z#jWSW2Y)A5Hr4vMCUAuV2nS}0<~tTBAU=<|4CBr|&~XuB5MNN5LyQpI?pe`rD?=e6 zUZ1nQo~gB)o;kqPTeiNwZDCb#vtnGJ*764B1gsNWpF%P%u{R=Uk`DQ2-+eA==%sXQ zy?p{2T;r+xd5e**)+`n0h*RA1gr}r-V8y}eKhsm2Cl7!H%STh%9OC>S_?WNxxA-%DEZ~Rocn$NI^BcCwafNBWAW9kNRb98)BYXl92NEiK#u}RO$l*rHP5Xs3h~h z+O>%?=+E&CYMOQ|Ac0G_(Jt^?-MrP_Jh#T~uO^k=y#1-T&cOCU($G`k4jL=}t0fxy z*J$ipr{6|T7f*9^IQ1`?mNM&sfculy*D$`H8naW!(5drQ6S}-pW!ciD_dT_t$GS)< zQGy;-Q}livXQ8v`pejNOa5n^&T~{LN)6XRja+m2IY6#O0i~$U3KgUUO%;Zx{(-5nL zhccyVFf=06ml^%)*rq^5M44#ppfVEiqa3}|z{>ZzQGU3?hgKWTr~>0Ito8+nuL)2x zj5hqp=n7y_O`?59CkEPqOM{P;l{Tvu#7k{|6v^ZyOy`IV9i=NL3|^#wiP-KzdS^@& zLbgfOt+4|s>@j)Db4G*N70lYl<3XdxEnk%A!k*V@B4;jYH9hPs(9-N{QULr>c*7YD z4C<(f@ZwpE<6`uMZD`3k7Iv3d&q%a^? zpVQ->Ngkkh5f!bPg_OEs8Ei5?Po6f;9nc4sCmf^pcCv=b%%Ez{RpphV?xPlobo$)O z1ACy!hLpwnHGiL!Cw~r#i>Y%Ze7Q@&28 z5JvkYKAP|b3zRAc=3~(|*(SEr$G(8wl-qQ7U@3IUF^>d>sDX||VU$Q1gK`tud7*Dz zg*rBA=8o;T+}^aD<^^t+_S9XJ)M3Av%RD43##`1!A8(QFG&~Qe@iMyksiT}yN^e?U zz9F?H@8b%(E$HXBm7s-MbGiNEG=Qf~dfUI34$Jk6E)?e`E5_99!4 zcgXC3I`#m+5XZ2GbkOz(3px`FMEEq;G+;*S63QivR_M*vjN z{%)u&2Su6EC*xEaFPL>1S@eSSB)|sG9h`vqc4oGebi}+FJgxcEy_g*^@NDh)Os4{{ z(&IqSoFx9L^ayOmdM~5~fpONkn!qnw_sdXUYe1|l1S2-~TzD_W>@7SM!3ZTCpKcbT z!YX+JdvPSeU=uq}lr7n8#Zu3&+y$7;Z~`7}rsEH0WX z3JS$0q~T0+yFBpWRD^Lq^ZC<$4}R~@pIv_%QQSXP=&N_Kxo*530X`(Ff@{btefe6B z6Xb(91qFMstmS{iXh0XC^;@es?*vY;xKr(2pyW0v5j8!EaRawg`9kow^K+uZ8*_E$ zRLJM(aq}4(dqFB-<_2oW=W}o!INk!T_v;p5avzlY0T%G2ESXo1TUh|RJzFU(BHrl< zE7aNw%?(O=yKaiasQP7Vmudx{5fl*IYHM#{h6)0O+ss7fx2!Fv)k4$-YVj6> zy{^^_G>p=@hwh{pNOdD5tNA<`7`Lgm_ zzEMVKA5e~8A>T&HTAruFvGdAtB}vKxY5|U#?1QwqP`MFOXYzmM^~2XcEXR5vu!Hh5 z-os$<%yMlA(oulR6dV#|OPy#+{2bE)@KnZCTOiG|^o z#H9QEn-yM;Qr+T9AJJmpJMuTbOBH02WzIDJ=7N> zZ$y{hPY6%6hI2VHIU&S&lfe1n{`3-!-c;}~`|||uWZk#qRfWv4=RtQmP{Y)(3WWw8 zA6I;f0%n}yPt1;h%x%=*>VJNSI*ZJqyQ z`^;$cZrZ3;45!%5Kdq#Gn`=LxD^lpI>212)DyL)| zwem-}=##3KF+GMtrt9|~1E<^hh@3arVI@-Nm?=Qwe!bT%x?S}|!+x&WcYBweH4J#K zaL`k|05o#o*o3N4!`Ph`H-R*I7a?7LiZ6KuffK{17J=bp+aPO>wwVNA6gi6G5Ee!7 z_3?GcPd@69PGt23>gxpJcgn~V#Aom1(CnPIKERPAaXg6GXRx(B%cVtQ%w$t^2|3=K zn?w2Fs2~KvhodT7)urE06{x*(1FbX1oT@^5GD%0!rrV=_^8IE&j}S^bC9-EMNmB06QE2wfPnHJ9@v~S>l^>K3Ez@ zz4_dgDFCse#OQ>0KqF!VLxY)OjcJ>9;HHf)X!PB)i?0W<{^g*eH~VDA7$uR?qTLtm zY^OL@g7wpqqNJnkZ9T2nHv{^>tMkUEym9byC9=vWeGG*vIO5@^S}!ic=M^yzX`4uG z=(*KGZ;ng-tJjCqVQDPt>1rjN71VF}?CCm**7wu*#mkoCRX!E8B@bM>ylapaHU{!$ zwBzR@@@^VD$!sMdddA3`VNHwihR5h=Xg($cHp3sO<*b#*q@C*l`c1*)+XV1x$zXyW zGaO|MofmLuI!@NN`Iz*kB}{%9JMavTHfbkLftqZJ4_yV0_G}voU5^aBy8^~bIeBvx z+aZ@__fDUXgl1v0=-O;Yy|)IdSvk(}P~yOw7N{o5xG1f7sQkhpNc|{yS9%Me8*Erv z!~IIocV%J&lq33g%i&8ru;Q{q8cMAqDNa738s=;ewIUqz*+xC56zmBcp?`n*|IkA& z1oKmGf77Uo?ZbzGtW`+S2W~Bjzm#5ff?<8h7afmN^w3VqiI%oo^Eh@UM*3LD=4+xRVC?8*O`|Nq!~^Khu!zhOAj$WSqsLaD|sMY2W6 zWT$Mcwu+K0SxN|5CKW=VvTq?I+UylZdm&3AglR)bitO9_IcJ#B?|zT>IiCA@pZEUV z?{jrrf4FAmJiq7p+4pZ%(DFBzPmVw-MgkhW`D+slJHGPjjhm!f^vW>zir{1}n0H)M zd`IFEHr4h0_#h03X_ZNk&a*2LzcTG>t1Mj@TX`izdm zmC+s<2bDo-aTeXqbGPoy!W<#Fb9P^SvBHlZLsrdE*!`7O+CEICj>WNLkK+mY)F~>< zIq$%y{H*ey0j)ibEy8)J_~<9h)QD|SO2DDnvcQ*$nfML7RmWk1wjI=y6~)fN3^qlW zb{lGLfBpPx&&9umobH1Q{5*y=+mc@e(_5NvE^Fjm2~Bb?i{)k_WYyIy#*{a{n({b0nDWA*t+Dq$ zWO{tCrvm2nF!;4D7@H{0I#rx7&ntU1Xzg^^r z^vkx<=;xy^r#C>5f5h-PYxwQbPi%6hdQGaMGD+J}%s-Y6zMx!$(?edglC5#5xxxmU!Ei{HKYdP+1O)Bi%R;M1*d z-csL1WWvf0miuTcrv7ov{Bu(M`;>JZwAF1lS3fTJ?cD3+(v)p}4-VM=79LoZz1V^2 z3<`<4i5!OtwQiV4X&}Gt_8qsA-(v8lW%8%>Gxj{IE!z0vd_{xqvK~Ip1Clp>R&D;e z%L;$~NvPbXHUWQiu(aAnzL*|lzoeRzORa^(qtg~cEXU|w9t@Dtl0o8;*{K%}@fOw4 zta16K&Z~7}^Ux-O^D2iG_n;3=3is;Z&2<$a;u}A3+<6#!L2FN@0UFj6w?n$@=}Ma; zJT|sPo{jqA{OJNVf6#1Zshg60Bkw+!5%j-*`DGJy(llf0k3H@+-VM6{2h@Sh*ST1ra-gEiN97=)T0X!nl)HlE>}&quW&16i6?{W{q0*;y*Qce?20uRlC@&xy5$MB>4ilH217W-P+9Be;T>V4E0eEqQ;ft<1`4t-pKPqQu9I|%XwgM)b zJU5;`u$4V&H>x`P66!5ffyCQ(i<{8zC#J^~RlM6ffcp$8QTj zEJg3z#z#NJ+U#d%rZ*fe`<{ENhKWVMx;pRC;OiwH_ayy3J_y4SuL{UhR;8$$O&->V z*1@(4A5~rv7%@{q>`I)m!a+(P++ZjY?kx!@%sr2(#NJSqO@1CzP{J1+?y8kqbo#j9}KnAY~pY9xSlx@p^Y~UJ~Ii`CN;n~&Zt0da&XrU zsiRwJ^h34RY>d&13cN=avy%9l+z)kN`<|YYU{yFpjlh%iwEVw_O>DPNc~8iD9M669 zOV4n!DhT0+9*_3;$|~2L_$V@Cm*~7Q^F84jM|Ly+j$KFbtrweI!pASG&)6qwII49o zbI)B>J?U;{n0Wtz4*Ml)4Fd8!NJ3opF~HTHGd?2v-j&PxGC`m}cw!fK=y`xNVHLjb z*2xig-PD^$;J1(TpEPxS+%nBF7Bd^jfB5skiigTDyYykoi_V~WIes8K{aO;Kp_FW{-I&**AbPmref301Oxn+yR9eM=64yC+!S9mi8 z=Aj)hZhJtjY^4RT3OPs;Q4P>mXep%Xux`~ipEkg^PgoA_UYOidZ4|xtYj$5>5y$Zp z&&Bd4)U5E=W}UFFi(DHa@9@r|bEve< z^WEpIFm_&kLY%jKlH}f!%ME)`Y!X zEWC@eml0FmUC;6VmQ$I4@=yCfe=RyvzB<&rB(5u=;Ys zIl0dv!k6w!S|o8n0oFZ?sL?lCGG$SajAL*8^=_!mk%e&^-&T=2)?c#aFVw<2++qo}_fywK^khOSv64&9hKf{>%QsIbAIvNA zy5zREh!K?vArc`I^XJO-y;z+;NjY1g0wHL)^<&rJZTQU>7Hgp*gI4IgJ&vU)RX-k$ zw!2Ik0Toc@+44Pcr=-j$Xx=(JOX@YMQ=fe_y?R~K*1o(W-Vb`Ke<&yka>ef%dzdU` zV3nEU*sm(wC!D9 znrm$mPb|1IKD5#-?1vA`i$n&h<(7gqPH!cUR0A z7OA49T~*FwKch+|T)*cYLbd(Zu{s1-&tOg8ovVjsH(@fjE5VzJoQE@m-PvOv{mk}R ztqW}o&+oM))5@b)8K=d!OnEPrYlQi*YC3uzKcGQKu?ISaeo#d|kpWU=LH^ zE>RNhl6&C=D_bC@=kaW-{#Wg$UE#yEnfh1G4*E~xI1jC7mg8LmuYJ0pr)EZ#pqYJLqfMug(O{pqgrIWAx&s2 z;)K#AN32K3ZfsjH@gr6z&DAbgb8_dJVR@ zjMdJxcGqz{su6vJ>f@erOv)23mz~3)gpAPeFbqxFxVSNMn&#os22hx$=#^RB`l@!> z$@x%Bi&KBHsNuW$pbcWt$Qa_)j6+f3jl1@EgD%a)KJHSsdf{aRf!64K*H5CKm_!i; zp<}$%d<}botB{ElNl^1b9MtT#RgcfYuBZGL*V?@p!es@R=~sZ>~oS}B)gv-c*e^+jc7>~9zM<+A$wkB8a>w|>Hf(^ z9XB(}jn$ue7p-ff;d27;Ss<1>K}RP1kt+y-U7_hz1hvxhOv?`_*64=%2%Pk_QVA*T zYfwcOB5Yzt;#}^k6RauVR=x>igDJ49l@(8NJL~aMDN*+5MZ*AslN9EpTVeC*+$}VT z2k&bJAEWbexIA>NTbBaLro`E{DwiCV{zhbDk`LgPu zp8j}V`+g2m1I*@dSnJb!UF-AbtxLsu&Dl7w?FfpMH`p6(Y}YxYq`!I)MQuCJ?WOi@ z&L`PcK@#-F^&9u${0=Y-<7!OYYlsi**^yQc6<+U-L;Wg?MoSSdlM$f{UZc{DED5N_ zNBQq%izp^!CgJ}80xtV3DhOacVB?jjtYv*4$_mP$;tkVqS=|5%k#3Rm@ zKE5Cx&1aH74zEUwQvP~%ca6SyvTfG@#FkeGOb^CZE_;0p>IZP)Xc(;=Okr*h2Jfc{ z6_q_-_c_P75o|F#sF*8acRc_FXLgW%sr&LIc2xCEW|YF7^dhh2zfRp;s4`rzFD;0t zIFG?wPFGV`dy&D@TEBlgpr* zD9>H5c`Nn{#XkAo9%C9+4hq+Agu1Wu{4fkJQ$^U|QW(!uoO3~~q8(W3a;R@-yO!p; zG?sN|WcVJPy!j03+zO(3oKX7#)mr!1kazd2xyQS`h9xc?#~<7VYpBdt7VjLAqgjEf zJMAE;$=>J})z$~K({h)}sPc=6EpL_DJn50d5Q9#;LPxkg^ElD19J~SuR%3gAwLa2zzq9>hsO^Ev!sNj_!>~ z++9gVsLAR_=wNx)|+4Lp@S8IbhC6a6JpjAtw1lrZzX?APSeG?)x|p8EC>!BC2QpQHWaJolgmd z>uFnEHcU}%{Q$2r8aecufE9Xd1YyzRr)Pmr%0KPt-YUdnn5bBL209fK**Hd9j~CP) ze|-mPutiE!R9thN7IdI7;0$nED)gOS7` z=mX3de_^7AC)tVA@4*N!;_diYAAW`C4C6pvra-e-tK8lQntbht*8$Fq@H-95CoBT- z@?jc8fnqLCYN%22R@VLeQ1L+NcW-niv_nB`uy+&$g*1BfB_^cmxOIh4_9xxP=RWnQrz5q$mOQgbLVE z8xgaqu1-IiN(Lc29gxBbpL1^37IBS1<8dw6a1Ll3Y{r}waS3|Jnvruy|IeNomJ~gf z$mI|+S%>-9W4)_!XpIoc$B}$E?JY>0fnVHmTI3u2;AGKw=!0bpxh0)~s~Er7aU(Gm zbseK(K4;xAsx}{2xu(fcIP!)mH<$7L2sj@f>Up?uaLYBC5vBDVGq_Kfw+5V zI3K^VfQ65LuDejZJht$zQ^NIKnl#HPAu_Oe%23JP@uVYNb(W2N3JY?A0-M7=BR}FF z+}>9SJv;l>*$f*qK-hK#nqV_uiu5aUG&ymG2;GHQL$Lxxs0S#o0e5^_0L!}`>Vq~!pNH=*fj@3_#&n3}ZhL0hdVAgUM-rHHjK z@T{yY*K!an6U&ok@76Q$if6hUnBuFj2GDi}W7BZ`sdmCTxNn45y(~tU==kwn@3+j! zFrqJKWKc%=bIbA-RGN@j!UV14o7o+;6|pED2Lq~EP|J55dNG$BOo@;M<(iFNns>i! z`_gxy)0QsJdI0$rnvh2Dj6#py4v!_cLH2?_omcrvsooH3q=&*D18b~_)T4_s`AthW z?foz(g-_l6IA06kRyf0VE10&^Gto3?e0-rJ#Z6L0uWDr3`)>WhhuC6NN1-tX-qS_- z0fvsEmRuCTvJj?hg;@lIztiNj--nSBmpM%M;%49cEyptq`e@*y%Fp%lY4m9CjIkIh z85zsE8YBHtn3S<;YTd**4ei)mT|Jd+FD43Y)oe#AQ5!8WM~+ z*oWDR)^ICnk{7?=&+Qh0b1`9@VH;-w7z%JPuua84>?!s2;^@|RZR_&U+a7}xb;AWS z(G3$ljceDgO+P2Wj={3y&K#Gvm|YX#=cguZgji(s`uMBhqB5_U7bMUVWwK`&@1AnFwD#b z@_hN;!xWQWr`lynOE7DdRuK};qRnupxJ-t;&K55tUBNbWidQAXP(OuDtyeJ_>b%Wt z)WJ>h)LVTa;0ck2X*=QGd>rhpm$FZ@hOhWV-rU^v9gD+qowkhmx;yRkIStAtjHCaH zAfgddk=L6?p$YT5rBw!sWz2N%6LGshzR@jIY2bOG5ISz`*PG;4*`FVx)DMoug^?e@ z$+~OvlR1G#$Ws`{nQ`~y!S1X%oKJH!i@S#e;i4Ayyo0;{r*Vs@` z-M~=Z>`zDOtu>Aq3{393>Lg=fsHSvE$vlP-X+~bG`{@g7GWAGPaBUSf>`7csx%(IW zt8-+@g!zq?=YLim8iG5{S#MRS{fZBRO=r7$T*l(c$11_^Q9Qh)dvM0*6PC1X3VBS1 zN}A+(Nh27?fsrgb8V$g1IihYiO{+HXzuUUlGjRvyb7g_kpT5e1(b3T|MIj3e_BGp8 zXBi8*pHVfKu2QHC-g#QjTyG@TNk10}9XUCfDpFLZb2 z;lyxX%UAg4SK6H}b@Dk7n10$RKo5?CJ?c`187`a~(DSDAHFZ(u?HraZ28r9OQ4Fg3u(VLF?G3G*2%&l^!tBFlv7XO`!K zYBx&#j9YvZaS7V>zK9NWdA$y?w(EJrMy$&U6W=b$9$(s?2vxBAE zC2h**1MgF{RnE6&BI2#L{=;F@TEab_DA=@VvE_%?w>9PH0#BkoSi0-j%I zw&}2<%aH?+bYLb`0Gca0_X$&Loux<{;6%VGM!^i+~PRg-gfzxY-BP*XR+V z16eNVwc?jjVZB2SwN&%F%^xLRoIEs zPv>Sxywac`II9NdN(gI9Oo&=UQi2oCL@(uFKf#`>hbzSYA~$F>2v9e&mIf-Z8ZS&# zf+H%G;aTapbp}~bG~_z!$Golce5K!HaUShZ)+?Eo>+F~PR%H{D;j-=>oN13b(}*z@ zz%v`uy1lmI3IU@f!}6==_gX~Cz$AoUVw>1T>cjke+!7HEcdLZEEm!_Sxy$qlk0CMh zc#y6L%Tob_*CN)g1ElJ0N*ggs)mTkKE*v**59P>}s6+eWv?%a)qUy`*)H_@J zX&}7t^sp&xkhDIn%0A;ZOK^nA1x}n8g(PgIaZZzZJ`Bjw#p1g5>w(YKH;U-d@$neW z$0pv@Y|K6O-s)69K&7g>dTMb>C{TjTE@DsvyB`O;%H3t6cp^xYKoJ~?Dp5qZO@_pG z#JN2Ik@6%{gFr(K`f5gk@}|!XZ@=N!JqKfVQ_DjIoCLfvZ1g5cqQdhmzmat$(8=~+ zF``t&apQ8v#r4vqP66P&z83Q_0H=xh`6zhoOg#A^c|yFQ`Ov0`=?1GWZg_&pcGzvX z`Mxt9f!!OhTiH%idPAJoQ!gZQGcX~b@oC#WcaX^LySPEQP;U>cbilfNXxp}J(oR87 z2kJ^(yJCi0%+BEOrP66-y$_-|vA7xlUhnfqK5nI)Z(49k8!4pEw$^XC(GDaHkj@&f zir*yy^tRj^=VF`^kLgPW)qV|B`{-$T{_}bfgtJTYX5ZwEe(YLTX~m;z^H`Xeb`vnZ zIBBf_J7rvw#Eo_FrXU^imi1m-G!>lqI89=$y6)Qk^m4LPk=LX|jo#e35j%2{E*t82 zd3jaGZaOp--N3o0E7<~gu<2WoG6t)NvzKWL{8J(byJIG{l5~*1+n#WNw(r>Gz=W>r zgkP>qMJ3#H?M$>@3nV`DrP1gOm0!W5b)(V@H1dGI{g(0;?4lP*dTI%HYHVTNoVW zD^{LQ!Jv#KSSK<|@a}G~A&g|6P=(DRQXZSDT3ODi3FnkWBo1KfRcl!}*q3iM)V2-+f?BS3VNPWl!ghwrhB11?{$vSr4H}uD zJ>4=uE3f`|#@PZ)g|}JF&jI_UKuQsA0TT%A3WEN!y=Gd1<-Lr2Fm)?Ff4BObHaHDgobLz8TMtu0okgi3F+;`M}K z=gd9cZc41pgUNLl81xI-{k6D2UwlQG*w6haDm(QqVAB__%xVWljGSL%>fhmmmAZ{B-}@y8U+$v zLHe(G1_svC;wB#8k1V|mL)xCZV9#COQXC5$2W~bx$T7fYM8)sKgwJaWG0}RkjQl{G z_ldf$0mByV`aMYjw0V_O>^_jKZB496RHnVkYcdq8j`_Bzsq@<|G3$9|W5PTE<7Ih9 z$uCw=Z$?A~Q6|jprS>wbA`fz~FRqu0H-JrYPq8AK_8W6&%+c|y{=ABKLjaqU@Tc)B zfo<)rba4n4?;YTV^25({!QSQByy37Ak?;Vd)OO@B;5zSr1Ts5k9JOeUM2(L|e)?SSLv%_R0DfJPf`@bd4v^~S zUHZI-lFE6S5^{~BVUBBG0N1`)Uh8%#ux(Oce*r^q=ZK`e%Gp)i(|W5r-TS12I2QMZ zx8Ay`YwFMjA2+_FMb*K`5E5H~5u}OwRiNlr2POQx3meH?hpJKfoySDMeldM1^6U4u zss8qoqiRpKAxzyh1Zq*t-mMG#b*<$KE0D2>UNnC0;+k-A%`+7rOo;s*__uo)s6~!~ z?H7VA-fi8RNYTw)1R&5L)5vAsZn$Wi%_&}!dJpju!ki@014#lv9RTAjn5IB!gj0as zrpP~jG_D}C);Ej{uN;g8tMXC6@W+of-abALj!JgHt%wdlc%;UGDnS$3bB)@7WnR0m zxoeTYQX@s1ZsaNQx_7Q!nt>N4V2UMamV>$E7v1TTlapV^$H%{qE~9F*Ss*<&Pf1lJ z?2ZsOBe0xLtjZ6ayDtUqOJPlrQ7sxhr(Yz>AU-DM4_=GKZk*Ozm)I%KyFwjbY#d0o zfs+NZwi~A{a&mQiZU#(o_^YNo4M@ZQkWB$Yxf^oJZ`pNL(Dp|M4Ci#~S_7uxEL&ax zm-F;+de^#IZ6joUT>^66>T>$^GPb+B%?wxlWpUtff@$3xQ5S9onfMwfA53IerTBf; zXtR`NAdwZWU-kA! zmEh^K=Ss{$so6;`o`d6Ifa77-fx8EayjW~^Fo=-gwihh9hK`Po!LQZ4mA0q$z6rEs zmgnuDxdHFRz=BJBTW`8n;KHEgBN9mIB_UaJB%HON78 zrbRP`>u#v2s%kCUzmt`Dn<8)iu^+?33LlNm?F4J+J#h2ko}NPu-fntVH(sBCP{l{^ zm}&es!kRNa>dikr3$&iSkk(&fViUr_+>Zfjd}$LVMGEXr$LezlF=5_<31bjrz`_2C zYoC%8hrszOu6?xQk_AC?3RXlP+UFa)^PwTt7^Bj-o_N|Ti36`6v*eXYoVkeA3nbgZ zp}54Ff;1mzu2|Z-SeL;L^;Y0+C7o(xeb-6T`tuTD=lZl)_OA=sTbq!Reo;YW#Y$0P zS|mup%vU3nNcj`Yk4MO+gjdCL#@H!Nh#uG&m4~#ccuZhvc0VGcv^??+v5cOcp07WD z-p&h+7~nRmtM92x?W!zHSyQV$xhw*rIAkd}po{6!!8IK0+qne8y};67Up&XKY59Q( zmH9n83vQMld!1E24@>Z@;hs4M-&ppWxtng#AA&Eek@qldcfA5!^gUY*SU!_7s5TLG zCnZy2ZGezVI zaXf-32tw6ixzR*5w}*LSc3oUE)an`4Bh9K+y;q% zXLapXjea#JyMAPC&eQr{$_wz40+PE&dw%}>*^^pfG5c+|CDGlqf2tiWb`K1qXRlv; zv{zLU5RjaZAd!|b?)4`6WpV)-Gy}7>u|tI zDX(qX#(_i^5C~;MUYmV5*tx|N-p8oAArU(#YEA;@ZC$_Lx8+q)GDE)++{RT84R{&`64?P>%7dpQfcL@G^ImB0 z5edbC3F@ByBU;%bof$_r&fk=%L@GIIzU_RN0WlQz6E*5wem&q3cfjO`vw2-0gkZ3^ zr+AR}!W`MUAFDo|`sBb1p&=9zrJW=W9Z&tbl8|Q>|ST(nRC=R+!yq7!1)HgNI(O|(gS)34({Sxb(rQcr{ zaRw~haWE<;vmp8zlbxMirs%N6k9jKnARy=(XT9|vv3IlWIV{14S=H|z;1FQ>b>`X} zyktzkWWdipzwxef%j5A?^S6g@Z-A_?jsJPMRvY4L)k8x=Nr^=p-~=LI7F|^(m7E)> z`na;lqU$%=bVHZtoYzt#p4MG=Wp|NRidE><1t__P`T5|;be-Ask+*T(>2g7eB#SrnrD|CQ zl=qJ6x;Lnn6>(#*-7D&?jIZO&Rpt*Oyl$K*jz7$IbM&+{dcxV@zW&M%3Jt zmd+m#V%k%`2XjX*#d34biwM{{a))(!-|b5Op~1nywWVSi0?JnI#hqaR0d})r(-B2| z)r&woyX^&YaBuoJ%W1bb^%pIky`;|xk+UgtFpIPVb6Py$?|>0}R|ZOwAC$V(p^7uV zcj%DbTI82WxJjM^H;I%8Vy5G%VppuHQcdjJe3e1+y}hrM=4b{)hQjRF)Xnl2Jw)sR zXx4zSDq*}$f%ogHYkKIy;=n3zo<9s^3HAxDTC7{1)V`TU6ee&ab?I5hffFP9P_|%> zRIFV{!?z!HmNhT*&1FWP!_>kf?hOeUEiEltDtM}cp9lC(3JPF7?Uv(CP<3lsAfIN| zr*#hmtZnL?uFYQ{0*O*c1cGd+o}TuT3JHAp&WcB7y@n8|xd_($H$P`b zx=BZ71lk4XPDODAg!}hs_IK2^87#PN)5>~Vo=vj}VV}NgA)-nbi&q4BQE5#UaYVZ2}!~=Hg*&ZwPDL{iaR*Kr+-Bl#svfWbo}3m8$6WjRN*)*AO|#S~6rfN7z)PKtS?U@6L|d+WfNKGe;nIK_r}oU2x`lJV7kMs;>sg zQR4fB=A#x`kfs=@yU^xa8Snbnv9iga#{Fj=ycwti9;;LDD4KcO69@L=5&sH9;IN%Q zbFxP_=;f^yI4kmLlMQ&5sluG|I#vkmgOX#q{pB66>0wWV>2Mwte3_WoDx0~!dFJ${ zkz);6dbf~EPe~74BFyqxjERGN%RP_rXqHRTH7n;$o}HW0x+_x-7ma!d3oj5^kdc6h zOZ#zqX~#YXMc7^oxw`0jF6k87@e*k!Ozj2N2@OJBcJk_GD>K0x{!p&bs4&XJb z{?KtC-uOZJ0+NFluw7|zZf&+Ym6+%XpIa{hktAoknY>n2DSXFSmP-o+&>leLrm(h$ zjiDv`7Z8Lb4$_OUUoVe-g5kGPx2{vwbrY!TJ;HSdJik?4UpiE|@;AAm>%+x4;Q%1H z=Hy@G_4SJDzyf(uvk^>pOV5t55@mkKaFF7`AgN92VDh~+MX67;6l}5l;6wlUE#k9; zL;d|f814fc9t<}_9sl^Au5l`@S+%oiJ(gehs3KDP3YvlxBlf+NU5*_oo}U zeJH(BE^71lAqBAQ$sK>!`DWWfx#0*`e2~;bSI_Pdnl`cMD8DuVxY(br$6Q-0V8k?4 zow-)YC(QElyzrt#6VKC6N%W=~IoI|tqOzAZR`P@UXx8A7x7WJU%g;WCi$L~bxsLl+ zxh5B$!~JhRk}}F~B3Vb18ZNvrk9xmvm%sEGY7%uGv*yfp6+J&|Uoqn8h5O-U3K08- z*rk0%(uFqDn;Jj9%!WyG2wcBZY(2jY>(5$aHgv0ejqCi~(T$r{m<^O&zI?g5&3b5Z z2t`z;6;dU^m)|SvoC-I7k2Xz>ZPItaiueaG>CPEr8c&6j{GYA+tpR@RC8hdYrvoQL~eS zz1;m9DWjxxn0{^-&baFc5YgT!kTMVIKuC|ROAH5>AO;fWJEwQ(B!qRVR-;rIX&U$E_oRT`~EGhr@;e9k>UB(N_z?JwRc9g4co`jqIGjHU(YP5!0#Y!;H|}sB+_lz z=D|tJjWAev;D`6$WU~^(k0?5`;I>+6%BEZ{`or#V#_=AJm)Gj8wim5XVrxyjqe)6-XE0%@m0Wt(rk;61w0jGn_q zO)b)b>D>LFPUa+pHD=c@0a(>T1k7}c&_F4$L-FfuYYZ49v27X3)JI&KO z!nd^%GD+&{>Z&_B4kRqUx^4(@|u6yb~j-c z+BV+XiGF0-wl#)S`nW=+kF|hU9|_CPxx`GkWtHMus8!gU^9@dLcQ?h^JjPh1@`w`R zjvu@s5XQYzJ{{mQ$#MHHeMaHH#IEv4pN(#K5!aHRzqzZctD0teluClxBN@e1julc> z`qS1o;PG1B?az{uweu2&Y(k1c$hH42Wi$^EAWhPGW?2RP{CiM8L=CCeo$pRxfBA+X z!izlMCA%SEY$Yj*gT3Ub?-!M|N|$=a>9IoMJ`yKvsnX}bxzL@&ep+g4FWYFvGb<1p z3G~Ace0KEZu6umCa5AwGa>GIi8-?|01`^SW=cQ$T=j2c~4=tq5GBFoYf6*LpX~tf9 zRWef_l3{D%XhUC4<|h9Ege>p*n|FoS{^Wpk<&@hbhQVg7q$bRoR0<*GMP+aDI%a7F z{_=_ncHYU;TR?YT$ktkfXlDb^j`R0!jz=$mcC`OSGTxk0144F<^3Y*AM?6~ce z7__TqcmIOeAQFCm9A!^#Ng0}YzXCQAVzrmKcy;-`<6Na*t8CsrUCy%p)2hP>2scz{ zr6rM5V4DL!Prqm6BH+75)-%Ut9jB``P%O%SYV7yV!MUV_I~VnbPgjCbrlOuc8AMKD z3GNK}TL)X~kfacIJ+Em;0RSqUp+|P}DBXJ13vSc&^z`7=l&W|-l6Zk+6QEzsaVdW- z>*}B(@6a+%&(y?=dBFmce`}XiX2Yq+bEN;Pl@T4!? z;Jj}-CNvXF=707+pS)Jd<=PP4(qO^_tSsF#)CK-!U%<@d=j^Uz!h+-B@GlAVg9n_Z zuh{q9;hhEsd+mmK3ho*S8G0vsjQ>;*4%yg9rK73~HH+E)HmCZ8u-mNT19Xc4_d(k2 zX7{XGVt0D`0l&YD2JFPrJg*PKhE9I%6Vt7ew{Mi?1a#(6ox$q#s+a5IG^}lr9$0@V zl9mobki+uAd7Y%$Qsh5UNuG3pT9D$XlzNlKy?UhKOE6-n9_e^K%f#^JfjrmUs2I-P zRxBTBPzyM^<5%DB9hZ<4k{^z`KhOD#NY)sq+S&1;p3$7n&gN{Lb$`Kv5NV7e>=QlY z$}{8l&p`9-<-`W7tDEYCP!txq3U&};JvHs_kPs#??}h`$rkUXdy|Z1?D7n-%Z+_CK zpx(&Z7+ulxryY_3q;_Dzj_z(t60^@#G<;5GA!ytCH>qeRU?F~j)|^8r2U)4JbdiBD zPg>2~4B7Gyhv}XP^MF_OaD={8=Gp0SNR?mm_VX)0p?!$PDJ4Mv4{B>m&Mr-AMFQpY z;CZ_KlPLnH;O1K=cLsgp@Dh-Up09>GB+q+Il}ix;F%oX8k)>@Sg?)pd*d!#^Juj;u z(t&Dh+WY@D5wV}xAm=^-$%(WQt9Q)9Su$mt>j5S9edBpf~j83!6GpSnFDM~MkTay$FQ0{kn2%x&9wlrX;|6&oTe%({xU z1Dk}k#wbA;MrIKzt5QcB%Enz&m7qoHChLUG=Ie`9v@|DQ} z!s%+{U4jmh=TjEAfgOvfs~N==!DiKw z#7oq>q6ZN*nPwL|Y3|%jS2xs8&6q~J?%ss?MyMjsUPdy}XLV(h?I|y>lDygRfx0J8 zo(%r^gIyRvCtzk+nhBwJMh1a;a7oIHujHLqNP19^6SyU9{Ix1d@AwZGLDE<9xIEIWY;CwRL$J4cK^EhriOUX6$1`{fyQqoD9t!8Ayx^dnXJCtn0%`aG1ciUuja}nkCDiR7A2|OG{_Pj8f|1XHbZ?BtxIu$#RmJ$8oau48zldG7=Bpg_a;9 z3^EP~gY$bcs;DJZg781N4F^TLIm&E|lYLs6)Bd?f?;PV3IH8*rd3v<;IzPHQBoUWn z6sOqj*qCOrBT4l--TJ!4 zzA9}0Bg%9>!JSbVN2q|a`d`vo5bTEYyE@(lXE}6>= z3j%o0?+i=R!AUgfSDwETr1?rE zWevLDLin|V*zr^1uZ925&(;&Av_$W^vEwcvLqZ&MLF~!Z3q?9nA09hbfp-phB~X9Y zw?56J@1`@oafAs|0%Yv!KD6Ky*zTn$WhQaMA>H)hfIq%kbzZFaUPu||WEU4{fP~b@ z)^$S!KqC!*ww(4g*uO|Bhd`OJ6s#UW4uTbO5Wemx6larYGJ2|@zyP%lP%$djWH285 zydJ3d`deXSHD3h_caLnpqMttTdbid5pBEC!lAsltp!#=gLqw0qa%TFNvJt# zhnf>AbJ0Il23mCQHbgQZ&Kz=jHLXP@2Y{juSS_K;Gd!PAn#;P!E~t2#afMz`TJ8KF zPNL6f=|G{yxxhNPwr$(!oC1w0?bODU1q%sx_mZc*Ze2yS4}|(1z^$7c`p)s1O5#dq_r1KY21X?z7Noe&eMQGpPT&vaP$Kq3Lj z@!mf;#5khB=GUN1<`B{xyo?+DJ7^BPgcOw*8!?KhW?*;G5Kq#FBO4}|wYP0bGigVC zNA%120DOHH#K)-s(*ZbLn$u8Ipq!(Iwo_sQHEK=gs2x`04-q@os_P2Tn^UMX_Zi3t z$_uH)>q(Gk-8G3h@H3eLz-fGelj}n;eIAxB6C)CG$?f+yrlv0u zm!^`Qn-#Ice@@KPISzja07cV9Bu8MH+@l}-XF=1sE775Z`^P~29~r3Fl^zc@=`f}K zu>wqw4YsXo0X30`OK+bE9nk_~IhzNRN}x(fWkk4!Jw)cWm3gt^e;pl{6SkTd93F|Cqx6VhSU50gZp7 zT}j^DNNw7I4wrZL>~{uvZRYS(##N(vEi39Tl zIdS@6Ac?+pc^eCo?9KnW69*4C3Wns-UCtQ{fPujTPqf@Rt%%I&`52ZkJW;qQpAk>7 zP7QqE2l>T?3ki|CfkBUTYJu9EO9~(GZ3m~m{n=4Qq?jZKfED}!)t(E~FzF2d2#WEk zF_`_JBMyklrd2|+ z5XtE-Ui2J0_00cKr#S=jWMdM1zI6wf;x+I;B3Ny%*1&(%@irPC8+rsoNMs3xBKsdWe-6ZcS?p;VDZ zE&$`^m|HapW0Cj5(*2>!a)i=MZbdsx()BxZXt>lRsoi4`@q1yasiAgACWgERkh=TE zTRXUz%y~HW`bK+hDWD`No?3>}2fc(IlYjLRB8K6ErJDx{Y>27G0^sRJi%EwtYzn%J z%mwibKx&GXAY+i>B>>ghE%MYbLm}{XQ}f&7p}?Wk_T;&rkz|1cWp5VGS)S0c8sPW+ zxXlpUf`TH9g4+&J1C-oAPRW6euaiJF%t;y=j-=qp`Dfat7J@n#Opblj78oMoHe9x} zjMMxW(BT{iX|OKc4yMN9ggZ!p%|kzsxD_NB^y?rV$%4WaRCw!l3nbZ`F-!aDbFi~t zv|7f~m!Ur^8E}`V+d@aDgWjNsf7fgg$ZXXau{;ubU{fMY~r93mA z`yhD6+q5xs_9NT{2NU|w7a-^#2a_R-{&6r7Vf=p(nndF4AA4+`5C5}4b&#!6>$-l-bH`o-S(G~`rB^pwUd^^OuA%b#nYSH_@iZL%jylACuP;1Z!a24MaqtC`Q6U=c(LIh+BqtKBh0kDng8F%WnR^a}A-e zkeim}irWKm%goTC{AqaRv|RYBX`N9X{&O!2ZdGgvSFrm_r*8~CoZ#%$;=%LLcqAIQbw|oeiHGdVJ zD3$D=2#IK-9@uPxZ-sI=fqdq&i?q|hT$_%HzXsy~QNCOcX$3jgbnrXKH4xjfdvm+I zA419fpJd}`mN&8q@!%@qJ;NA2AJV(Wz)*Zci9zWS(k)nen~UE>n|OVn=-UyxzaD8$ z>_}1eJxtG#BfydJfx|3sgi}`@gq1$}mg0{fF+gqVoPVmdX^G0h<`^^s1hQKiG5x3l zxC|)$>+4v1V;+;7oV+8VMZX#ag?m^s*bUiUwSmL&d>Dc!Fo_a%Z7u zj42X-O6IC31`qI|6 z4IKWz(uLu8lak8|Q^ttC{-BD3xkC9;R1pXbR2AV6Nm&?NC)|Tb%EI&y)x}H%!e+w^8D8Pqw&iiB(}_)3ppF}`U;B^fZ)EGWbY*U zsNMonQ%gchKw3XMX-Vy@rp=+U)mskIIzqJ_hSSiV2@S3uzCWiz=R0VrVjkw)LAi&!EVZC} z0+o9lbG|_VD}}*|WD)ggU)3#7oca8HS2gKMU5N5=qEm9p?~5$^Y#WZcK(U;ASi&k z`0oAInLQy8%JGqitBa+%nlo`SGCR$=-X5k3OX16#l?T zx{XY4=`B7_?K~wR*Tv{k{RGq_Pxx3y(C8Q^omMS$`t0`$*)DeyeN`4j5Lx*}L*{D% z0*RtvNE);jqlpt=V5r1ZVdL~41SqKvJkC1jjlg~^P-f4<|s+pKavM-tO>p`Ox8KO^1QuXoPM_Qyf(3ldr_kvlF#mr7q3I)A}ZM00eP>4l31N9`l^Xyvp)E46JSmn+hG&$U7y z$4wO8qBf4GCG^F3J%3j(uD`r-vJ#}`)&!e2Q#&_A2~`qzemAxWvW;mwhm^9@r3Nug zgsHM57W+*G?#5sx6DN3X(b}~UaB8v{o1V<7YMYjCudq|;-H#UhDgf5D1cVpP7vagg z!MN`>aDUEE>|(`B7{bLJuCi0Lj3}#~j3jHL%^)v8ZRSW0wf&1~kk|75^7S2hVaD;- z#R=v@AM~l;7jeU_9||_4nX;ZX-7M)nOuaR!l~v%xQ_h0+hJ#p~t(4oBDBa@rfU%Fk z$2xc;#NiyD#*{8sRM3(0N0?H+SKXi(dAuY^z{1{}`L`{Wj%nvNKgpoB6Tq|t(wp}Y zbG_v+cUb@YU`Etoc*sO+Nv47PBx(h{$rdJ79go$V{rvHPhnjSB?i{9~TxJ7&TyVI6 z*4qv*9|AI47DR1)p~~4NGqUfkRUDVHAX{a<6Alsajodf#2;*<+zK;rg(j)BFBPB

-?Tie?EG?!wJDRjUNN=C*Lo(HJumx+=TMyt+DHigQ#kdA9tZqcY^nFzs?kyiETi1JQEC*)r2 z)>w&a*76Tdn7EH)xWZp=@p!tES+%$36h%G}kfOjDx7w+$IaBI$?K)8&J3v^T5(Ah+vYlK&CVfr0P)1!$*z zw=|^kUE(cTlrW_74Jixw$?^Nk@B3|-cqbm)deVL-yaNAF_w}8VRCt(}5r0VUv}2(G z=WgJbip=%Jw5L4Ad~&ssRe={?RY_bSHP(DwT3XN1WwxA#_nl1l9_}=JL2HYD!i#?x z?3zdI@g&r5Ak1tPr!qNl81@+McFfZOI4ityd_^Rm-xjpJQH%PxdO?Dx${N}a4+0zx ze!fwJHm_gC^t*)ra}KrfB{DW}EV1RZgYzr^b)w|{(I?DSy{os8A5d#_Be?MoS8vJO zN@e}!a0>c2MLhTbC*_T+sKy9<69C^F`Jl$Cb3}FQL8lZ&V*p;@fC$T*HKfBkBDR?( z=acWImJrnIIHUL_g_%*V?%vWC#9lC!3H0hW#$qIabAS>$#Did5p`Up()WXnF< zF1;%N2HZ5%`@kMP?h5;a2~|0(juqJwC>p6%1P7S^3*YzR|1b95Gped?X%kh%fT$oK zK_o~3I(3iGwqMb7MfCpMpLI*?+=rulT4e zv(s)>QCRE?$_)l4|0~StNjjrZtnX?Bi3C$?&|o_>JcUvGxuSFYd3Dh_^5EZBs!!5= z;ag0=PvN8W(qAHn6@s&YVBY>=ivGPq{=Zuxpc3*I=v_iz8Ggs<&Ii!2=r){w7#SVJ zxEftPid%Ve?wQ5cNV9&F|ed!Vnc9EZJzRgMgL=oZp-XA zU;Eo)(Fz8yaPM_rxF0N${}oq-wi9Sn=RuwXI*SoO0L&gNvsF9O)~8_gM7$n2W81r( zWuj(FL#-0;d;Jbtg4jF=T-g)$deBtrNfOlP_QSbSx8}(h6|Uyz z|IxY8@7=P?!5u3E_VZ}$1z2707&8r@zzQXNIfyG&54s^UPs?82f`u?Lcfbb2V#dx; zO${D;spj!1ob@^mt^jP*3xAjhl_N?5&v?FEmi96c!c7zZ;rc~>AsSqNh!ciz%Rlp+ z`(R`4HBKPE+)3V9SfhOxGno&86FG52kremrXW)07aDVvE7F0$6zkb{-D8&GIl8gbE zf5fWq5-RY{j<-af1ew}x9GJsqh6Jz4%-fEoDP6^%w zrP925tNJ)d^BIcf`M?ixE9%SUiz950~q%aXNCtjJI zGOT&xv(b1D_x7e>c}?$$kt1UzrmTWn{+VZkgO~wPoY;T46B8luV(;y5DIWoYM|C2c z1&1+D9n2e^!Ab6)^~2BX*z@Xk%Bz*glMtIWj=-yJ3UEZ#5nR;VUoMaTIM|PEE@big z!ROJ^^L^*Bwzd>L{JsJ~$cAp~oo=>YS2l)!9HgCCROCH7Q9H0|dyvAz93MM&Fv{~C zV3cv4Po9ro?dL?|nvg|BY?4TXzEZ&1srG;eqoB@N&^9MU<9L)24yK?~-(!*k2 z`~x@+fifN<6~oB}eA6B|OTiTumrK^V!y$IT`Z0F@Bv;=WJR99*P9lgWKzY2@-zS$dvVL?jiWD(1REda&W(alJFs0_=>fKUjtAI++tmRs@euM2nWSI769+axJGyRwQmN8t z?_S{ia@p{YWqgZgysof$6St!p7V{_$aDYUPfHEFsaXNsVAwX|CTqZ7hw6IS$+JIaBnP-o~03~o@|GilHjw$fIZ`9_Z0!(7gs}TDO zX5-h}rsE*M-}EE)JeW9>bE{Z^fX@0NfPKLbmf8NTV(U_{yLCsk*0HZ_4-0GJ>5XUD z4`1QKpDU4I9=M*#rTn^rwyQhp7KCE0wFo+DIDi7X7k?T%_P+rMdH{__k#9WsjV^Iw z2kl6D2TBzymcMmV3#9qSGAe*FhW*UMb$NuvjQV3vPk=I>Qkh1&&4FwH{px}QVqhX# z?j{L3Zh$!d03bF^0Lq%ZdDth{{D52jndcQiC4kV8oBUoZnSX+%`)@jpUI^kaxzn-d zqt|dy`)~RoctvK%rt#t|HnISn-J1lPHe3zMY%kk-@IJ`&Dba^T?B?O``1QB{sB*gbJZ+^qdG)c!Zn{5Q}<^!7hA6qBht zJel@tQm`VlT%;0U!U-I=@$c~T|Hbh1=hz)vCCh7L`duDin1WCMOBT%^>|w)GTd={l zjA@lv!Q+ENBS^_0LO#5RgIIO=IUGKZ#M%-0`V&a3EnP2$^#v!9SX(%$99x7NjKtc% zuLvQr_C>Z>Y!nVxeBp}8@Fj?~#aG0}I3U*k1u6d-DINgK^S=>mgX{a3NTTMxNs03Z zLZatb0BppG|CHLd>EU(O-`M!yf$o1Y6;!HUH}b0I_3=5)GYC`^)Bp(o?ZYe%wF9qY z`{T8<0xbBC zCDI=c3j~#p;#gB}h>N^62YCgMxbkH5BrYiUr-}OW*sas>Jv!!L><#@K0O>hM(qv1) zB+Uwt>uB6CLTr6^y!GGvvi)QHjaO<2J1~^yl~`5r099E%OO(y71~2L!8S*GMqo|vze z4D@>;kM&c}AmPjb?K8MZsBt@9u^hjoLbp+KbSOSFCMc(DGPbY=kFaomG1??V(av{! z81=wlC7B|%l^ezB?DL&hw|3NZTr%FikxCy$=k~>K=Ces?{KEMKkq0%^dvn!kbbCLZ z6^Q+X4&+IQ)U_=u-1*s>-gnzjGP`8VgS{idTyjp^n`Wu|DxE5iTPtZ4vAqv< zr#J=e}X0daJkMyx>K?mR!7o^0S@>Cvog(6hK< z9LN=!gIX*vY`Ku>bf+y|yrf9Kjj30g;UF7{-x2nyL|-3|%~r!P=jT&W>t)$Y=}*XY zEVizQ4<3D!=rDeXBq&X6nkCLUd~MJ?u@I!)@p)2#d&nXsxTlUtdSpo0j3>a6$4o%u zp9c18`TqVNO~|x4*#6dWCGygV5OE!Wh7pk70%M}j){JGS1JEhLcb24O}89He6eGmDrc0idZfO~V846bU&lFuj>u*IL+5`F6Zbzd3&7{~ z`8*f=l|lTdY80jKqBxb{X^R@n^;ymArKtMx#1e&B&GnVEK`j*P4`)^HZOycBlCnO74|3z*_BPU%YQ*s zhuW^fzWnP`Y8_M);er;L)!^0zG0 zxOHQ|3{f(%`44qwLq&5_FPS-PO20mx*&g5bu~OtTb;7G8kXPm^x-p<~TeE0B*LT;d ze0OcQ7nujAx$KfO#}3i-2rpr6_uja%PqEDwwBY6zktl3ck?Sews2;mJJ>sf+?rH&Ee(4(a`!wT`3b?#qZRmQHMm zeNNn4NiQ7pm`hFE-sb2pP6%S*ISx7Qr>fHIbq=1sSs~+Lodv&hxeIYebX|2 zSC$%OAl8)D**iJNLaExni7av&kLfqr%P9r2`CVKkBjE|V^X6(+YmFjReBN8ViMDpj z?{8OLh>bt+&i)o@8-&@*1?~d;+jQ0o)==JWQ@(qlt_IH=mo}F`0i0UC=ef z3bd2cbC27A9rk;AY^B41Y@po@qj`6-I$Dm~j5;=2o5iAE%Ov>L!>-GOT!JZ|?kcCm zJG8LRoV0VR3)=3td{n*johSQlRJ6Z&LA$Ur-OiVzbDt`-OH$kLQ!MIt*T)n4^<78% zIkn?c*Vl`i&z|$Q<}>LgRg>u6%F_-qlgx4f$5Qk0@JMy2*Rk;J_3ltV^52Hx_d2YD z{F$KMxOFJ408jmi6Qs-}H;lC-4wGvoWvW&0t>*R)S>=u;8bxX;P`3uS9B|S`5vPcY^j3$_zAAxhTy~{SQNz99i{JbzRRmU zLS6XsX{461e}5K%W_wka_^|5AzT}d-n#Ve7TtAv2JbylM9SPzV3k{0EW{Cx zZ6Cf0jeiXi`5fR!UKh!~z#5~A0A8L>#ICBF3>GYTFL^FV)msx_lB~-~T611|%)VPt zHCBwy%&iiW+~4fvxr+&LtuRTTI*$H63*J_H`l9M!>H5L%$${Llr^Kd{s8BFhyWG2I zifVtoV1NrSc|`jFg`9&(3LNoD_yS3z^=-#*J^H*(a;E7p+`Ze?gmXGWv^P=nBeR=5AzvUl;2rRbVpd$7Pxu|??uN%W#5fDIu=QXww;y!k}9FfNuk`BmGu`2 z5Bvu=o@z_|@FRM=+OM1Y>B7?pqm>-jN)EZ8MNSHRCv}e*WkWHO?eV>}l5~?0(X`6Z z;~ZHpe71+jJ4zEzB6h~85WLaVxp7_8_PR&#v;KdzLEpjmUKkggV_fsurR3>Hr)o+w zhX)$F7l6BbTj}0r?bO1a-d5__W<+fGxhYGWl-W$W{AUYC??XcFixYct!q0^p0^2Ls z$HW%^?Xa8azPDnZIQU(592}=%k=ZMcFMdbM%c%wAvjARdlBle8ZT0Kgsj7Jn_0G1W z>4T zd1~aN)G1;`t*~RxJ&LSFdnE#IN*&jyZ!^_Y&71NlJ5IL6{QUTg&d1G4vHwowownqQ zS@xO%)i+CR&-a>kb^hcGnpbc_kb3jB zevwmW$M=zh%{KG2{hiKz>)m$~S@Nb*NhdxEbX;Uy09DaXLzcT{n+LDHKgGB4%?{V0 zIHbag-C!PzZGO1=5qVEvnfdaF$kvgw;>Tav^@*&~o6Uzv6D%;)zRnm{YaIO&|D9=cmA34CS@JqG=ymG z*BqQF>-4rY>{H^MHH)?OFZ6l&nAT_GYvH$WON|6=Pnbq(1LxUrq6HIFQh*q+8RuM2efB0-Y*Q@y zE+@aS(90>+u+D7CfBZ#wbapExR7$LKeg|_t=Q6!HKz`9sgJzqh-7WiKWWF1>PfjntbC`By*R$Z&c$3BMEr&4Sx zn>e=*nx2!6l-wL$@)^jvmziBW3#d8Yr$L1(UOB5?^?uv)yW}YsI`Eu<^m`=JwntJ> zqw;EFDLOSh%c1V1NTJ!AIZpAV3P%%oX`DSi6L!1zcg&2M z)fY3fSDCPHRullmXXMY#Cot%rc}=udSLO?Qug#1&cG6I?+&c4o>To<=P`%@b7QgN< zbGSts%~{K~CNcHeQN!!A=rBXAMX|$LW&s4>VxiZb*r34rq0Gddo7O>qCjGX7TsP}J zhjH4&Xv_1F=i}JI>Ph@%E0yC%0GT1aa8}99A3qf!e)avz-Nhd%RN00p+Y5 ze7b5C^Fb6PDF#wDlmzm8-1zcnp{_|IbwJNim8!3OmJ))_aUI1*K8`z_ zJi<5qKkxsshf{NXLKB4cnz3C!5ABr7Nd<{GdnF|jq23C)f_ zF}qii2uRfE94zGh@eGu2;*KE#R4e$W;V3F+ii_V~T%fl{R^FQry8}3KJ_lrC&iM$-0eJs3ujvZv=}h+~U}#`l zW2S+}$DZT7MKQC1Z6ifQyy3NT;D)kO)x?yS9B4b;8U-{sr%o$ZohXp_`H^v^O`y}d z@&JT?R;<~x+MOP^JSM2eFZO)zeHvy|xKy>BrsjnvU65)DXIqC5FlZ!?b13X8?ckVg zm*d^W<2DGjrGn28#**V7&up;MpsE>qZ_hu(1qmwzSNRisqd1o{k}o2Z-EYw-hRZ-r@ld?Y&-BQ;Zap#ka-0r ztE8ab+r90}?-qQ!genn6WLL2ShCge<=lxo-;#Wq39Kzy#dSFT0k7m4|Y#>Z^%M(;b zTLrjQOwb&LAXs9fILc_N=5kGT7%B&I~OAE$tE z$hBKkAkp=hNktWhKI_Z#ivKgEi+kQ_0zJFiDwlnL+wki?R)4jO+NMz51MSZ`bTWED zxfSv4+2(zDu{E-Tz;KkChG9cjsvlNj{V7oYmy}(tEDS=e&N@vWJJu-$bX{ zv{^XKzDrLly%d*aUG_aPo5CFl;XQg@CJ3&ionowI+aeYqQF%rAjBbCUNnnlHOyN2E z0r98N`<9dMyCfX0@t+4$L- z@$Ip7odWW;NW@DB5gIXNaH8h(BT6!J&YX%hB5Upf&d5q^d%mFeDBvG5Oj$|&l0lxV zP8dn`(X~BW$4}~4Q==OVvG#N8SgG&+wpOWc;@)c~z}BFgFu{HXeveLxv3UcuTHKl& ztSFrK7A96sK`h6N;of({q+h}6;lqgAB3o)q;Vhp{X5T|OGW!+v>tq&r)a)#G8cu&C z2|8{+coA3X@`*Rjeaw9-HQsfsOkb|ef>eO9Ek?ln%e@J!jVXvwX29QiZ)@3k4iD=X z_8Ef67penOB@tT6Z;oT)00zx%CJ>+tdO)UNY2oijOxur^8Z;O|jK>vp$4v8HdYpWUtN zBZDzG9YgF{_MrC!;~RgH^=^zh5_^8D%BosPoWD0aLIFgyXVa_TAkPwXp$a_W&ZN@8C}BEhwe3n zF8@?}W6|(@1qWeOl5R5W&4Gp!-1%j^-(8Bs?0zXNRF@Mm;==rhKtJcAalEzPxHo8P_ zoqld!K9^hV159)ZXOmup%*S|I{X9s<6^G9GPG?om#m{-X41argwkXcI+=gp4Uh` zT8wvHnmO%I(A~Ta66N=8Yi0$BVm64|DJO)P!l6MC$Cs+YinsOb4?;;g+0VrxR88#fcqFJ~J(3Po>k`BM9P(XVT)*T6wA~mds9s z8<8zhqD|0CVC=jHN`PA~9&5_YYdq_{WO4;h4%pl?XRN-gg;{Pw*j_POpmmlra{mHq zrK&Js!Vci@49qJh>iruU9@|&O{vS%U%OSd&ddC_W8_0gI^@1rAWy1g`-R|^NP0mbG zs%!uk@>T1VOADwnjH$B9*I{im$18BdRZAwH5z1JV)U7+v;FGf^U(jb^?PS~fHyixt z&AL$Ai>To&KGX;>|M5VuzfsDV$n=kQ8G5%$@AUp>%k8TCRg_$eVrhhv?Xd{tr3{;8 z+k$g7(d#9Y`;aOS%Lc+XwNNnpx%dEi=zbN1AWEXekFcScDA=KGSEv2Oy~_Q9hDaoO zG%(6f?GYZiU$^pzZpWx4RCFRGWn&|wKo;GFpa6F9ACRR;EoBA9beYhVKRgQkXn5@V zXBP46Z60(5Q|)Q>j=^@`bGqR{W{=FVr|q%HC9TKUWU&x;>ahk^%CU4r+dS+pn@Wf? z*qyzdu0>v9;ZNkt1HOW6>H8VHI$|L4PV}{fP-%us{ir##aI2|ZzOC`P+2)Lz?@;-R zhG0!|cZ5AMKD+Fu9;xg9HVarnJ|yh7+CC{>%Vs-^rEA`y1KV}&DvMTG&DdhIWOb{5 zJmFcXEOtFN0nQ`VJj-~;<7J4MbrKTQchUPP{Cuwch(&3zpsJoMBhIeMeLy26Uir41 zvDH#t)o))BaiK{?KJ4u|5o!AgL2dt-2;$D_NRPJfZqIknd4i%EY2m$WKem$>>dm=; zTB=&Uu^Cmple!wFIK_D!b6U-diGOtdC%ALvIjPEm`dv4P{m|5YbeXN!tFuj&2zT!< zRmp6&<~urPAi{iwGkX-J=gK)i(wfz9LIHSaYaK64gKlg2?xaYyW&75-PBAgV*v4>f zX<#S`B{OOJOD|$icnwOFSrQmVsLX}bF5{$6CWJT_6ChnK5?ujfqPmIb=Daqwc^d_to_Ky>ZP3&&O6L^duDNCPMXDd{_ zYKePuT_U6JarMh^XVy0e*!nOuTjY~iU2Yt}lyMcDp;$b|hIo~LqKhdf#A`FM`23kt zqe(>4-f(G5cEra`Ci?E>M2*;!Qn%KzBgW&lpIg%Uf}JmFbeG@U+uxMS#AC(nfQY*B z(DL(lft>jY+e-|!8LmevEi~J5ZaDkj{p40kY3Sfoj4GhP+(+^--W-D##RXp)+Putx zhQE%Y$m5TE!%^@_v^Ha1&j-?A{4+tprXClurMD|Uo>ls0j&Y4v@9Wot@H)?XLP8aU z$G3gmt6oeUeW({SN3h%rv`{gTE96&6L7Y3b-tpOdnol zr!}LL9E0sEAY@9>)@8eH*II$q=1wG;uCj-V@9g zUd-+Y31xeig(K|SxjX@qVD!^;g&FTsox9^cB{Yr7rDCj<4L?2NiCniU)fWePT@W0^ z8w~*|+GlZmw3ioM*j-u%~P183r?!Q>UF``3lA$ zZ#_K5Bo#ip^B{95>^e2V$p;nGI1$KybpRw=lwjf-sn8M}8l27{M;f#??z_Lz9qLo) z{ZZjmfa%bd_9Ew4LzDPs2TA9YE2O&ZX?48m9>N+qpET+8F?vnV_L`XTh*y0SiRmjX zGsY^}X--U!P@4Ec#Yf5S>}ra)4ZWQM-d&&5VmRQnR*-1ZuJuXuq>Z_*NqHG#sD#51 zH7ePAJ9mF8SDVM4RN%5Ke-{9{SO(rk6?H!6u_R_Qhf>Il4nK>rj9Xg*Rm$l-si%wJVu4zbK$7+AfR1ZzH0=c)L+cbII)MG!Mk}Vy%rDK`ep9nVe z`(UEL(=^!JCrtHF7O%#Ojt?9BX-zGw(4O%cJ?@KqfjEy%x zT%10A^;{VGiP0-9my~5Q#_&L-0)Wg)W+1ixh7;zpV5GbMzO=S~F6SQ9IS3%6^o&kU zT9DKV3uQxKHP9#Prsq8Ro1R;bM0ejwKIi;NedfiLQ!%TUc^JOm$l+6xD_u?imP=5l zl@%N3B_F-oaNt9?J|dF(ETy7Sca|Wbx1m{qO>DVM0QEknSH+R9SEK%u<~32~h9Li! z5vBeCrRBi9tBLM<%tHnA3jymfZ&uwSLTUZpfG%&^%7w@yK(+w37fVUTZ?C(117h1s z9FAVND3U5@-nNy6!k=aRktS04u4rUK)_qB|YxiG{N`{*r^JuZDf;k!gp=-l9+jkq3`ahE*%92ehVpadP{*56C5T$ zGR<1cHs;wRdn#UF2gi}U9eJk|hM(eQPP^10-iqayfx2Ipyq_1KI>n!g;XZHdVfXO< z_|d}ieu_||9=vIZ0rr6{yP!i%|G51fuUXARu@`Mf)L#h9K{_@OG*M5Mg+SZ@id=OD z^=*T4n%}hmEa3qNr&n0wzZx%|?n9k%`q2D2Lq}QTZqcdNy+W(=_2*5@o8HF2ePxnGXObNl$x~&GlBRlL0-zR|4a!{2Zt|7C=m z#aSYDlm4{wj{6s>Gs?Nc$TdIn6twZ{>jz+F;%~VMdlwTSmgp5^2fZDxt}kgO8ZM5v zZk9~3A)xLuy}=$*Ai*AaE{=__&Td4YN#Lo!1LSs651qPbG{n+y;s}N_$+4P{r-HYk z$4X8q%Zpfm?b@qS>7ea;uvOgXGGFr;ReiZ3Q(brrl%W^PNE67bn^%x7=p_r1xS|sZ zU5k?Z*Ul*N1`yDQNv{RnMI-R5tIa3u^*Oo=U=(+P@r|CyXhT-6CA@9rBA==bXG)*8 zasT7!5T_SE7hS8xD?vI|bPObs=sOlF6a(65!k`V0&0Xb~hIt8&+=`!B zHO9r&S`GyT^}%v&EWK(m*6#{8JiOn_CisMiw%OQLZBF*zj!jSSn0?inPUG0dHp8Y_KjZP9QC`mCGN5`D0Mme+OTWIfe-*t~S?x;?VA?SssZ^Y0 z_<^8{A*LjsNcq63`uV0xI=0=FbZMLR^6PKw+Y~WbUdae0q%SbsG=7y|b40n3u-!i& z!rav8JMMv8#TM=L#QV55(SBltD{)4GgFUHvB$t>l5;KZwvFU+ev-T#W_ZYpua~N9E zO<=hQS0cIAUMO8~LPAy7?N8}aeIB{l#B9pI3~GnR&`u$B7#=@gCXV!NX{B*iImFrc zYX>4BReyVfY|J9z2H51(h|dD*k%m5-P7n#FX$_PGkQ7MP2eyOfya`ssJs@x?M0``2R2%{0PBv!+@EcY9oVhSXlQZdw3FxP+;re2&wMXe*OKsE=^%u=N8M--up8`2> zyM(*RNCDNA_;-24uY<81iV!Mi+jV_OZjZgxmnRvkmY=Q|L>4L?ZoNv`JX> zZbpHgQ@30%cbSy-5>M3zI>+=D!zFf-2#;38LG(K=H!Ee~e}JS_p${zw-idOD%aZbn z9Efa_Eb(?(DCMdd$>lY=O+qF6%{D$>l?oqQD=ulT(_pG)Gs9FvJUP@9z3^U|Z@C*_ z#h6rw_!uT+1yY{XmRHiG8KG8E@bXUijY{B7ow+p0MW_)4GViN^ur62V&6g=;_W-pZ4;9Eo?86VWvFkONl;m zF>NT3ZMj{|;&RF)5($TtWHs0wq zDKLX9h8dbb)1oDSPl}=2NTRk9s^wlgn4v#gzIj5f#N+jqP?s*hzD*v(e;b#FaVBPi zXm%kudfku6D`|8Es-Oy2yXTSAL0>!rfjlPU{Fso+E65QQS`{Vv#G`4hRG{M7fa6r# zs%MqzURo?>nMx)3Os!`B7%to}=S~kzR z6*8%#H%+rl${=gdkw(?KxC}9HQqfA=4dn*ee9bIYnjNxk^M0rh^ ziI~MB(*a~xn4Tw++9JT_?S@+AT98n)YNLV2Wr4(yyq@An0sHAtMsC+gEvro^`{ER& z?y;Wbwn{`oMRv1GJL$e#>HVs@q)Phcya9t8pvyzT7{FSu(5@`ldfZt!&9l-|Vx(Hl z+o#+&cc6OI-m)Dj^0lAI@*$y2u0d*(oS=1tYacP)#ELGuYzyLufu%`rI}wQ2UEBM+ zP!r&u;-yyJ5EYrRB+Ne=8(WyD%g_?~qH5=%Dcs4v)tHa%a(|ckx?t@zBuXq*6Ha&L zvbNlPX-)>l-L(gE`lqZv)$T;yaIz^7da8&g^(=wAN_svd<=t7%5{DD~1kAP>kd{+#sHGP|` zJt~_k!@itu4QW*2FW`E~6|shRm4-ScsP03I)zNB9gsQoHLeS*Sbq+AoW{`wQ&{K|M%&$s>S!t-i4-m!wA6>?I&}hhxx+TnFF|=@9yflOLXs|4RN?+4 z{!&O~8sVx$e_RmHIHQSUvMfU+q|a)?iSy}olS zzs|G|saM7}=)y;oc68UX+WhVIx=*S#)bIIi5uAW%l}E&N*rvp#30;?pREIIfFi#_e z9+eM}ijl;s)oeTG?6+O)bUZh60ZFd_i;VI)lq`q9k1jXAK_<$>EkF~88Ji~g=fjXw z-OQ^XUthR)0d2>5dPqAhHUdeunr(sg868{UvU67sU{(l>bW86NE$Fch)hsj^4n^n* z4AmUFr^jO1R2t&`c9+lG(t2YR+!XVC5AbVAzj6rSq^y77ZKecHB+0cfwz|CJH6?Bd z#&Bs1RZ_WNGXptTg*Ou_CHBDi-S`^DM!L%_;Qp>qes4ksjI#&0=$H-wkS-#vv~l^n z=a8M=Hwh&6HYO~anWN^FNM@(bQW;_(qsB8En=d$|H_877rneZtbRkJ}K;z7Ni8wtq z8qy$y*;@HZR7QEzVZSDyPiMP|VJ2f*NHe5x~QM4t#u4WaLBc zuBS-A_d$dU&Z$ON${r*$p-q$PnqK<(s;?uPvxki6VC!-qcR5=nBt+W%G^i1X=j0c? zwK8AOX*)n#K+n|c%NP4BfqR&Pw8;ru;KiL3tF%ShH9aB?a#kIBJdg4CJcog$;J zn?U)a)$YxuZNGOUK`x)~JadAIojg{w$v}}*zjABPDzn> zC!744d2>q90WLft5L_36E$#fAskLVcNtgMd#jYzvsz~0)Y{d2bNrudyNY(GwV5d+2 z7*iyF;M;V6_U~riKiXWfdJnZIoE^UVQg!HI-kK}^d!+6q=G{Gn|9i?vtnAUa)fmql znXm-~3hQ5QOlROL5hH|wCa2=C_Llxw?&Db+?1o&k^AKo|Tb!8(~|b&ZJ4voj2+O%`uR`E?H6Mp|D(nA0}vf@z>AAwofIUWwVN=CMyV z>3yojz?}}a2flNgNOLTYDVZIjSb!}THKmX`&7D#}!H-f4Boc5SV`HqffVM=|TyhX2 zGk)C%!t@oRz(EzN+wXJmM$ zK=fTu=3k!B^Iq-KSB~W>{%kyhRG|j!$7ji+)!KUQw+jsv`kteDB{FjVN-*l~+FIWoM)EdXRHRAJ>2^`fScRFv=NS{sMU^%Q>s= z4&yb3siW}Xvj|(7VDb%qOeEigrSMTbDF{CW!Mp5{MoXL+aoGx)g^~)e@~YTcT;5qL!6{V2gJo~~`Vz`?*45f_93JE2^&yZzzrvUJY& z`cBhzBE%{{a=M~I&;5G@wjy5Gg6Wty6s<)6OhI4U)S6`W8INYe*kjs7#z$hG+(}-k zIu>l+5at9PbDJHuJZNPGW8N&wA~3ZFo<7*Qh;0T2Cr}f-3Ns>m=~&MxAKxsSSMR># z?G-Jdo~Fbbka+c>2Ga%AR|@mN%~D(1=&beiXBXwOa)>iI!N@?EawE-bvMGRy9N9GE z!V_uWF|*QYxW_LE)gbK^=CRTl51tJ?IzE(mz|Wo}+L}BC*_~p!yE;I}dc-@stj4Bf zG%{iN^^_CWRchr_q`C~*WPsIYSQDZSc;s+8wGz1a$a#ij^*-=$r zu|iNhTX$MVRQ_h@TUxAT;l_jfo17tp{Vnb!9>y#wNY$N9zzjsqkkwAZ))pYyUpu}{milY*U98O99n8D* zgrP`Y&#;$>1>Gbl_3gB4dpc^ozEnr<;suIwgYo*Uhw6pJt_1lTsA$V+Qyx!k9ayuv zzU2B&-Ti@WTC?-79Xk{;mz2!i`wJ5Kly>~JgFIRxJ=Nvm;sv3H)9y4|3~1qKwGIy`k#VhIw`^a(^RqU!fDSY`qAdY#jdLc{@au^?k5?B+qJDfh{Ceov{(f zJaSN^VvEV9?GD$fm%`&-%b-fIcd5U@GC%HF8|wY0_wkBv%sh(7J|oXLt=r;NDoCY( zjL|(mB=N)^9)RH3HTkRO>}ADve$?firYw_Daju5)#A3QhZlqWfzjO*>Vn(~TBUnlA zBRR~Hvr&+i32jxFg9eDI6l?;uY?0>!nNk$SVpO%EO`NU(t>pG#XM)4J!Z~b}w$k-j zRsR?ohOZWJHpMaRi)%QGa)=PJ=POVcnufTn-L;bH_(FXFX3B>L8gJuK7<$>GD3CG;t9wGGJNOK9Lj98oY)Hjg4`gCv z(San`FfNVa;!?$O59r^>7#rYArVs@QjblMd^i|}+ac#hyqDCK=-U;SW#$t$%YBjr_ zJ|@ChgmVlLos911yuAA=+8uE=DhOG!4Fw!9_liep z-Hhg*m{l_;@KrB5=UOe40Ipm?#Vz1`@C=?QkQ@^Xq9T!Oaak)bjmWkE2Ad+}m(qZt zueafDaPwt^({zG_NwFEn4KQf1x!}WH0V!Kf-_?f{74!`&Y3kVR1r1iI5w}`OZ^s(k zKsFcIFjF{R%X9^d`&Mi1Dz>(04G)ki0#!; zPOD$uQxFc2hKglNh)z29P12Ix_E60%abz|wR@+c~#fYsy|E>eO5Ua=vE+QqG*;JJ1 z4WyFzHBtp~>vH!Tl<)MD@L+da5K|V(K(R}ABILNZp;X8)F5L<^p@xt=n~FF|*vHs| zxTRYv!6ePHP*&hIr^fk=d?EC^TDy+mtbnD&Y?P58?Sx)w6orA?9W1C?^Rd4ZpiN<1 zu3;*Hzc%K%x1VDUEC93Tq$qUpgx}s_ zbYPU$tW$EmwN98dvd3`6msl6s+bOdERTcxb3K#5NY_GR8-^AuU5EzQv7TXA6dAHUn zPv7`OZ-O(L5l9X4?t*n@)?ttIMPLEjbujDR2B<9-0|1+tM{YG6_B-QMrZ z^(k6C{OEGh1wWt?=PV(7v$;)S7-RsoJ=wW>ND&k9M?+B8J(G*Wv`|2n{d)J$({aD} zC4-v{2ng<8)e} zHU_Q>$m|c}(K?Ay{r#~uKk(tHo&8}5yZX1G*MU(Nw7%2SDPQxb!}_Vx73^QgDA(|F zT*Jr80ea1B|7pzmx9`Ry5g-x%-uW+=(l!4FEt8=6&##tR|MXCLWPj>rNPJ`z=a)<7 z%}J+ld;Xq6(?KGt&tGwmZ-yQde=hygt7KdUw)@HF_wP%PMw_&B*+|E~gLsuln@?PG zr+?|$^r6D-f6|-Ht!waipR50x9kAG#cp*KP-bIV-n1NQg&#NbRTeh$(-xj(jE>R!?a3Gym*5wz-i}8#MaiqrN|i?an$6jedS#p}zth>gq1_-o-9w=ycZZ zbIjT3ILPGM$^VPHgg%)>8v0(yvTMr>Hxis=eXzKn2#wwf{-;{`yLT0KDgC?A+rJyV zVH*Md*J|{3y$IXr<2}bsIzLNjyT+>U?+&j2haFt?70j%D*8+gk_)l*rb;I>dBGPclIL^;if@Ssy4eiH<(gM(!!z?7z0Q&*1j3Zat4FJ~7FU$|} zB7NLWW-Makq6zq3BT?QdEH11ATmR<@IdmZ*uX`JY*t9NQXf1N*(0g$k#JnH*zu3J0 z|Fj7-?>l61pA$xeb=?md4js(DfJHkp{s4gdqfg-))^0(AtnJ77R9IZhrVdRTPxrpL z2fgw$Zq>cvOCf_!x2aX-IE}K0c9W$slGwY{uz{MW&x(Ao9bXr`f8E*tfZrW}5Ovp` zRj|m87kYEfov{2rY{)6YjK}|sB5;-)I0WqPH~< zVB`8A8R#c}u>1mIt&AWr8PAhGe3lSQS>BaQ6FF!Y7#J{!J0djb7JMzcdI)FN)WNQ0 zL@!`J)O>+v)nA^5;3C-o=$|UrV4WBUagd(8lK3-dY%J|$dHqj1)~FtzHR`(<#?k*X?bbgQC zD^8PCJPpx6i3{GUaVn0a0hu>GO`@;A3Ox`~E zbp=y`T=9j@;NMjKLwo=Ir%cGdH1#&t@hi*gCqOoO60_+7IMV%unT2hU8QhF-Xc&=W z05*v&u(k(|i-MCa(D{5gC4q$_`SW}i)?iTrE#i9OsD2dIE`0<0^6{C>p{Go+Ur*t~ zsW{*gynRg48j$(oh4VSq-9oF!u-#C70I2=G@}e3Vpzc%79z^c{jTIU(i};XN!pUE_U2M&>XD$tAh##MiH?t2_+J_*i;gtR*@rDFt1XtdPvc z{or9M+Xbd$=!d?SYMg7feSN&s)wU8mI~HC=MoA1 z-Vt^}`AgX0qTyZr314G=K-Y3DucuOR2S_Dpe~tHf;DJ1G?#d73xUtF1Hy-`<&U*yJ zzPkeHS8*?S07F5i5OIJ&{@sPv(kz%AUB(H#U(=Bb(@}{|cf@^sAau~UM2MgJ*9#W1 zKr`3i7e{{G6`A+`kEQe^c#z&@ATx7Od*EN?*|>s80S0?t5xPN#v)Et_rg31~mv^_%*y-ySBQ3J)X*nKVjPs?11S-;hW!> zOl7c=dvWogu`3!+KmvQ8W8rWg1#RBrzXhYjc9?DLHy`x%rV?pW%OV$ z_A^h4aLRYkSVZdA`~Fs!UxR*Ly&>}+Ux8QY=3rY>k;{S(?6}yDPF0SIZ3v_gN#<*o zt@c>GS8%tPq`<5_R^IzM0ZwetDFL3~cI1EKb-7^L%D*emNs?%lIhcm+2nY%Fv2}lB zsm)N~)R0jp>4nWZ9_QHgui&P$haU9n%2j!s+H4ZTsTwIlj8Cv@=|1RbT{oUGZpP^F z>c_nb&E779)*d5LjC>I8OU|u#kvj8=ZI5qSI8V|GaC_qNQhg?>E^ z?II*pKYTpz_tm=RUT>Yhd^SEOp0%Dzg%8`9U*64&O_gk{#Tx>P*tU6agGZ!s@-XCST($%#NaPe;E>jd!Qz~n939L&$t%@X}K2m+_{K- zkgN)<%RIN}Q&?ejgF%z$pm{1z4`%hybHUR`58fVpJ8wYtY)p#)`xh7(NK@{4Mx5!a z0EsO3#C7A|k`jh>XZpbz>|SN}x^cM*HvBY%lgG$>OuLvzV?YKbLL1fwO`EVEEZZsZL;T25R&<9@OYG_AMVW z%IuuHMsfoCP`?wfyakA3&UY}#JUo3WjK)u~Kyk2=P5kzq8l2pYf!ydX8~w6AN$26? zmFqVi;YNIs39>H7UZ(w0KWORfOZ_le=l6seU#SsBrr&y6gW}f=CxC)UYU(`3egD6) z_ulbXxBuTbQc=muN>)PIBUu@dO_D8JvPsz)iDaIlL1cvNy)sXmitLq-(O)Kv5p>ba4Rn!f@0C+8faU~Z!tnNf zcH*SePPj5pj=elY0IoabqGouid#`C0$@lB%qfUYaRJwFv-=2`rfCaQ<wYk?Hv;Sk&+hHKone6AFD)?viSQ%5EAmS~{Ep)_IT+yoe)&YH z#1%Jvp7l1`e(#7G&UO;Zhfa`^Wz#>30G z-F+f?iYIOj_XAvlIg1J!#I+5d4x^X^`3At|R1oZ+#aIkVxrh zm>-f8KNTdiq@t6iQ6`THH;;JJfi(xUZ1`gZO+m-+=MoS6jxz#5$Nx33_@xSrAZ`b* zutGliW}d&S1wRl8MFyWJ`)`l!$pjI~ezEx9f3iZ9nLtfP_1B*}@&Ep(oD>Kd!NrD) zlK5}Xvc5LP8T}S%dsxW`*(cB6+iMK6QCmGn^y^?5A>`TlEzBPh!4OH_@?4W)e1)IV zpRGf?3zzI|!}xbJ*L)A5Jvc_7qN+;K*fl;?ODwr_qu8oD`B-2pH9rf~Z8Y5jiO?oh z!6p@(l2r+iKl>=nC;;ZT9Y4aVylFhBsR8W7pSai^Hg_FA)&ZVsM6ijey>N>9Ot6BU z7Xr#rJT>3EGguqYg>5bHD%)4;`DI}WpfleUio5(2mxx8*@a;RoH0FS5EY>*?6 zO0vZpb#qgV`F8&w5#6ZR-JST@_=&fBgAxFc-E|V&ic|-@t4KD(=b_-S6`T80@GrOFWxpM< zHsuF(dtL5CfzSBh063oGqAkLQCLF`EOa<^6KZM-*ytyD)Z-pf^;^{!Bn#FeS_;om~ z55jp%Y2u%ipXW4Tp$%86(@gH}d7&IS*?XGg3uFvSp=WpDwfW%ZTQl^x&yiR44#tpz z&^-jXR_nlU%&>vn${?8n=fBe|RtX2V6vDX#A2~xZ%l^YtnYTxI6|THnVkkZJ@}j4! zdDZGwX&_$y5F+IvSKxX(n^I?Fu%hUpYY&L@rK`qrkXe+2=(x5$9AISzDG#I70qyaC zCCp^rrb=P`;d6lL?meO++x?u@)P4LlK%*II>$@ve>TFBCIhp%KMr>;ex}MIpiMaDg#=NE{vP zUA$gMqNyz}xJlRZaooT7P=d_#82i1Su>O53(? zz@ysz2crP#}tNrg=Uj>gzobEjcuTKje&>>_bB6`!dGmHbtM;|Fr zOOQ@%G#Ps#%W~&7-ovKxOJrdswvj~&Z4GUX8)m&H;hRFZbm3)QTVHYUG0JTt3*hql9iw=e_I)Q<@&WcVvtK>VIK=H!z5{@zzzf+64+Ba~}= zxRWd!+McDAZDx8u9(^{M>2my{i)O5AaPKo^zdI;|=i-}3zAPpTr|>qvgMnZ)B+&UN z0{Z}|^a$~ct|!xvmgHluARL9@>{{}Yu6_@$b@1ppj;xOper$^|9M7<{^~Z*L`{ES1 z8!pJ4s_Qmq%DSuikfiHbw*7_DI&Jl+NHWoNq5XKOuxoKq$tZ{L`v!WwMALH@qKN9f z?U4k7?3*L=Ie3${QhEqTMajc(=t&}J&^C1dPA{1cK00`$^5;3L1A$J?kiEo(kr@-*=3PlH4yksFSy za(Fl_Y|~aIv=V)A7~ZeUGDiAv)A%vXYezsVRPbY)j4fd&ul89}5n<+oUmF%s;E@D9 zuDd}gx$2mswEZ$}XbDOZNOedc_w?aCgKM^t`l%e%Bc#N@PPy)w)y8hpR^1%}3)!Ru z7Bc&$!y@aC%8v>h=tu1c`!ZvNRCM2L;4gw*NEC7pC z#v-@376&-Dv9Cw+OS|P2$$T3%J`AjL{<#ZqC=?D8k^n&s>rKfU{ZgFx{B(>%=s{=*#!CH zj86y2q+iZC<$6CE-xYR51dR7g4t5uu$hY=#6qnwy1#ug^*ytFsNsE=!B>-`Q`cfRW zrIGp6kLh#~W_W%BdVbj@-W~^O7mw-)i&;wjw zFnT|9H^#_UUOMidMag)bC|=5h;{txl{D6WI_mwD8FuEWr+f0`Yy>6-KOCE3P;wTh5 z@VEfts8{#k^r>Eqgg+b_!!YO|UH(<*xjtl1>f1PQaN*DCbjNMRMGSxY7F@m5iV z(WB=Nk!4)pdbne)%(J`Qw7e#GJpi`J3$|&+db-!_AH|j=8aRJ}OMHVTf3)P#m%*`Q zLOPq4wOHvy`qzt5W6(woVR$7D)|tWPdgEp5U6JIiL!>{`9*;u8pF{ulF-BR z++zc-krs-b5-dOhQHLzSc7oK7sqjbnLM zbdr@z^ffOv2qYeL1tMM=8b&@a0^P z<|GjCO5VmB(-4{+A$PAtX5_LA{H{IKh4GSg3ujnZHPk{vbi_FKv>*uQ@QPeaQBHv4 zwKHP=V6ILw*6p0|QxJa|m`0nBKU?}!8IaP(#78zjL&=?7wzN(zHSg3_l?kxA8ENhIGsUqbQCtqjX>yY+DiDV z{Y_pVYTrb|LKWF@LTKTkUm}LJe#LppjTW+|c;}vb z;#%V}?Kr_zh8(Dfd}XQId6xgpg}GH|o?zuR+R$AR6QAY#RryBUgH+brXlW)O5brQL zcBS1y=b8(kECTkEP_E+iTVb$*c7z2ao!Z;Oev`DNK^aI7lBi&5H6{_s8eYtYp7N5x z=O3VBPfl?h-sTsDwHv_)R%`~FZm!u&2k+ITpxjyi3bIF3zWDGM=de?LdCL5e@gODc zFgQ~Fld#Ijp(irULY%)UUw7Ex5eZToeMH;eexF(&tozc4Y?qL|Hw_&q01ssC6a7cq z(21TsStnghn#RfTkyh_bM-nU>s`HZaj91iHCuqg}(yyWyul~?Jd@!h1udVX~eBci3 z_(4w0zTU1s<+Uzcc%tYXrH5#zrQwO^PYE^@?H?ylCeA_IbQ!dyg{p_G%>@N-y+C=p;cRzkgiL z)>&JZjrBoG8zMFFDVwCYYQGk)%C`u|a1(k+>fA~Km*5Dmf@&5&#=3skl;b+d4nION zdk>bzGn6uWY&=MsH4ILY{}dcs+3ajw88!@JyP#SC^eD%2hG?Tua@1jSID#-ZBg<1|Q({y8;4Sa+{0 z>6+@Vl(0z|N*V5@jW^MYhZ7W#I0l?)Q`(e!QY2sn%f^EZB@OC(7`!Q`_F-I`7z?#;q@n+LF{ zR?DaYf1A!vykTE+k&SXi=T#26y!gb@D)+ggRfq{O5)2x=MUwzF^s5Tbz)Usmci=nk zm9qal_g=Wg1iy0hqul$mSkj6+F%W8>f~AS*S2TTb|6#(=-iaGJS$B5L1t*df?XfeelpRg#IfZuj0f-l|Vt9dFy--8XK3nwvAHi|2 z+;f8C_u7ZF>c4dS0-yyAyhoee}mca_;`?kR6A%v6tdYl zCAa94)YR0{#dT37M2hhw)OCWIV(T>5;&F~gqa*(UYh@;WFt_Kj62YfzC`v8@p+>=a zJu0F}QX}LTp z;BO{i?I-eGRXOFySs!KW*~H(K+P%K65jc~L5V5R|FX0)V8gyZm%&m#J~sYCWLv3yy>J;_u9L3V5KwmWt>H=9mZ4e zL@2>0h+iOky1O;X!K|v5uo-AbLrSN9E)53zjc^=mKYTva=w2&?~@G0Is#8r-7%y!-d zV1!>GBa9s$3_mfR!QAv2ZRi@rOBD>uSWj7FO8Dt(4GQpi33Xa9VEd<{<`d-5CA)RQreaGh*!#Q5vjw4{X0mq&jHVSALYSb*rA*~HIV#E z4ZxFD51Wdh-ID{=6l!>6!Q#cN{Ep@yhT!7->x*6~Y+szwRg9!S_>p1m|4RaA zMu0AWld2;|HnB;DF>XtQUh$cJ2=63!?7`H>=~HKHq&Vdz{;z=RV1&zjKY zy1zw>(=+a+wq_#R=bm5OOUZvgm-y4~GSwHQJNL?UM%7&c9WSy_wSeEFf#9e^!LIH3 z1nrZ$4Ac~^BZos*OS(OO!233y+DoL0HlP>jODF&4^8jURuN;qqf1DJ8G6Ib!kH3U- zT3B|0nBl`qF_UDnk*|a&)b;K)1&wZ8U)t9+o-<|n1wK#*AJ|9P>SsN>@K^FiC`Tvt zLJ?UJCU_!!)N#{ycO3(Mm5)hI6RrtLssW6^HG)_vV=U>}V_~QbN`eCdhI4VY3x92i zgj*G5hE^BA2<)#W;FcXnjUZ5AWl(vGUpW%^Zt3%ErAsX#jzcTS=Rjz2jbSa`k(hi2 zFJ_6i1t?ARM-nRfzd?-OJVSfp{lgb2{8j|l6d18vA~KPl$NB%yHTJ`L(I}*c1BO5> zFiG0&T#}=koiz(vzQbe3tP17z)lZz#oT&hB4o9{6LEY zvH`rxi8SuK9BE^nQ+r*VJ6G`*rf2qF_uNi&piM7*4@J=1(K!2%CWW1X{yJta*mMGy zM^07V{N+nH<(G2Q;gGry0w5MZ>(RR@_X*DoE@M341fjb2g6lxwp4I=#bOWGG;?P|o z7|ibhWGP=>7zq#e02FL9BY=3C-e=mb@N@j*a*A16@xX)6z8d?XFjml zSTJsvHMBO9%s#^5b2%jvWsY$d!hI7FYzLblx+qcbSNwz6rI$~BKZ$WuFz_gnN z?ik05+OK5+)~Gi${@vO=O(nQSnb|_3v9QZI5Th?y@wo`n!Yc& z^IQ*r1dewf60e_ap>;fqIBmln-ek)XUaZ9J4mIrG0&qC{v!+L)0Vlcn8GuuT>%yft zL(*6FOz$aLkBu}7HC!$|t|s*E8u@C$eQ~=~Qy_$u3@^*jZr=hxhO^lnz_en6xIXXm zF(PTwCpYnN2&qlGySButNhLl(o@j z1J$sBXu88#?8QzlpPfmDID`ky{Jo+#e)$?g%%EtE+Ts)qPO9`rh!Yg6<*ni!>4fzg z6bo%#)H5?5;+Yut-Rjw>UhNP-!0>?G_k&;F4^cwzwAJo}BAPZ!z;F<~1ymj#St)dr z5`bwK2dFRtjO#sT@5R`J^m+$nu6}pR;p?B)fMzQZ*L=sx%Qeq04*;#vO2otlAjQRc zz;oXRX}=|479zanCLWBBzf-~ony$BnRdTA%N4=CfxP^~7#3zKhsLU0Kj&Hogoofn4Oin8>OF4&Aiq$g4$atM)bDNfX2GH zwzR4f0Y{n@8tF)1;zhXGjtSd+=|=;T81tE}=HDQxp9qwF*TilkgPVxwhb8z_MDjV* zv3+-f)-*%twX%%fkHOj>{pSHV_neNQixwXVOD&KXjT4z))&-YP0!i5DA>$sZN!pYt z)!p}xN~L=kY!5^efDgDBa6%w#Z};pI?L;fpe7-)MPVzs>p7mXYH*0tU7?{R@M(p!x zqrinzcZza!3mZ70eLQzrzyEu?rG)uW;SYC`iNxi*36DH_hi~Rhb9{0q?6V--1Ai=l zo;>$xAWtv>QtdLlcEit}gC}u|Ust^Xa5-P0e6U&vr4V0wEMjC?AryG~I4O;QM8k6D zPK^FfkD!0qxOth5pFckTJU85+=Bruaf7kU&1_u)xbv!U3HZPN~BYq6hm^exnqdBIR zZ-9?deY9j&Snv@3Hsgav>1yJizHtzYarDP#HoQG78i=6M-e3icHBy?IrirJiDY@k#ADj{1xfD+tSMLcyKx zdtUO`xsu{;Kr`tVIc5|%z>L{5^0#V^_njt8&CAEg06g1Ln5(x&x4P9Zom}g4L9XG= z=bp)7ZwUodzbCziF30?lgJw3nY$yN4lx2;RiGx20HM|Xk!Rhrh&yBwi54HnLrWlXl z2tM;zkX4){%c23Ghc`d%5H;4E5ucCX_Xe8v-ZXTf zwZ^5)&u%pfhRxjfw{HOy4uw#6gzwH|sBju}TC(Gam5>xgvqy7IktWrQhK2@QXclbF z&#V5bwECRb`xU@Tj-RH#%-(Xw2Z7ykVM#O3ExduqLTxoz%_})*7of(%NyXlkgJp^X z4^yacZs6s8goRzG>SCQJ`q)t4*ecE=*T3Xy^1z|vGsw&%3Jbk5Yx!(&EFb+76=&vY z-c24jIW|T0!gy{(bISFr%ZaTTTF9jxgwuu&Vfpm?`#!7caTI7<6+JP{?fRV4xiiZ1 zGw*hU`Zz10-1D*8u-?L@Rg;nLW+U5;u2t0$3D+D31>Imz-XfgWLI$F-&%xRD*2S#q z#+8ow688r(2@tu?j9>~;=Q^M>-7pr*BNvZ2t#&6GTtUSjXT8XIz4!G ze@Dl#c^Mu5+omUdvucC-@ezSY#mNsz)Fcw3o(&}{#-aN{zcU1}cz)QxH6<*2D0Y1fcnLB#^UG8R^mFqv!jZ?672LWX)tI#^JGQ=q6$e}V z?TCgw?l^U>-TX#c4#o8e3EO)swlo8Wc00G*J9llnUhLAKkrvHQ^!li2X=@%k>5MvUTO!TkGeU z2gWGPo+Va=>13A17hz*3*W2^F(z1b9slj4vv{$Z;)}g%KuB2NIMV)Qc(WDw|x671N z;+;6m<|`v~#>DPIaR0Vi=1-L1{VFq_1VNCQwc^yw_ATSGyU@H;B@27x_Y*MaONS~l zI4ie?yf?-|adrUIwPBWad+J+OR&gu+?PEghF1Ridiin_r*obj{=qUo0Y7SM4mVGTX zd?A_B#&}!>zEEmtmKu6w)*A12@42u zTt!X|A9^75=I{!Rle`5PmNd}7-uSqUB^}FJ7K-wwA8A~oIiljEH-d>?#hN{x*J;bZ zXVzI!+ZjJE;H~{=M~k4P#$b~#;uoCW4S>y`tpxvFdR&Z!Ri*>oo!7P zySWFgZ59A$jc@q*oO*<(W@OS@{AaK9A&M`E)$zkH;MHbq?%wDYl@6|)a&6k)O}M?R zxA-g-i{QPJ#rrwhe0K+ZpCU}VTf_4WSoHi>go8&TfGQ0$n^T9JM zK1Y0d=5=}VO|7u)bBer&R(`x#BaH6!xBnb!zYu*6jE;lY%i1|rNA~5HrRJACcy0qH z+Y#5A{#$qFBSlpuL+uAo&0D%tF)wvw&+Ifd?^p9CROUpuq{NrlySz)VNd}Z-R~HUf zr*s1wHo#-gcsVaR_|Ir(~A7EOFUBRq?L+dYViXi4_%d z&G{OLO@!MLklr?b^P6hB_%1!zwYqJE-bS~(Mnv5_ z&%d=n)9Ef3*_SnDeN5=hgs249?>j|XUi|Mi0Xt+l=~PjE2WNkX9Wve7#g|(b}Z#jiZ(>0I2{?`^(m)p?XA&V z>N+N50AFd5rWEUBpNO&JaN~UxIQ*hOg_1j5lmF|5gLIm9^?4_s=u2kKURAjY*z1he z%9G^LPO??fXCki>Gib)rXH&MMpP}P`HeGkmx&ovz7xzLU-dhaq`lf68fu(mBp1Kx& zi%Ddgq-@jLvqu@B-?PhZU6~{2PCPjBo@!(u`fc9l6ho&z?=SnQ3@)4WtDoNk`1f8M z&QW}WL1ar4F`tWOw`zuSDF@F>u0HLgX3>RoP=1Rg*OZ$3d+DAejr~KCUT%|IcXcA$ zX6Lb~H~5&fTJ4iP~@*HjkHQU(*>1WUwIN+mC~oae-gnK@g^;n_;nQY zS7bOEla1YRO+QqDPLQJuuj*=E?EyD(ENHKlC`KWv9@_#GM>+7y*GCvPbq`=dHF}mVYS5%4#t4Mt;~YLNK=(@d zTGyic<&`(9sUpX7Gp|_pDVs4R1Rkpo$Aeru>KyUxqsz_!`%H8oIi;Y)UsdmzV-{fr zL~7pcz$VO(sv(#Y)eD=c0aW%YSCo@^*aL zkABiT3-#m&Ww*FO$%;!JCPv-RboPpIU8K=kD^RnIx*09k+(&*%*z{8W$K0`zq4c3V zxW{^zQJ3? zyu^cQH;qSUyh}JRJpxV9&nsUJs7p+BG!JJg7^j{IU6>)WdHK5S^5CxPl3QM)Wz#nQ z*_JJlM?;kUbY~W_Eg$6H9$Pp>vzBvZVTbz=U;o=;Ewg#i&!^G`YMCK97;cAa|&mt*T>m#*Hy9Xj;f=&U1M0TTLHw6zdD9c@}MQR`zMXsZx5TIBKFQdXBG zlzwQpp>O|FFyBKpa_DT9)y#9a(%iaOI%w3a&Z6|?9Y7-5h`-o2(d4;Zs%q;I@HQi~ zlmHHou02-JHpDn$q*;N7vGhfb(Wm*TWLNvH5Hq-ORzLSVATW9`B|Otj zI2hpQen29vN)epUrNK+LZ3pwX^aoaqMMcfaHC+ZSog03AX~kC4i7xTcq9*`CQW)A7 zHdD%;V0iCWP$EN2-Ix!m;%uZ18#CT_vV6TT4xpyQ;P@!e&&dvyKD5+%9i zW`(Zb)TR5%uZ$({ZcauB>9@AGR^@t&%4vCLS#7Yr6LqLz^ig`!gV9pT8_2sPx~)33 zS2H+qQ(@&ZLVcDM+rF)tNgtt2??>Ch#=--3V6|uDdI|Uz{XO3j(;Z)Wm#+~;CFcaJ zUb;L@teP5eTi}})1G9X3g9k+=6+BOB-}Sk9Gc4IS3<^uD@J1OTLTM+S{A3=cxi$q0 zr?(nQ=R9bJWoys*Tw8F<`ux>HE{8|P{8{36j5XhSViDFeZxbqvH4kJu-P%+s4Hjfn zTvqquQi0&iGDl=hDA> zkvPagJ-10oN1B$L^18&{#4yvRptSLUtDDtT=g}CW2N+R(i*oxVl{Ls*I;(-KgZ?|& z7r;~J!UPr4L}TR7wE6-$GR@M!*{tfVw~q9S25Tkh;<@F1a5D%2zN(KkyO!H~+^jLi zJHyko8>Nyj1Tm>9n}_vguA_IhZnY6g&KMs#Ckt0WuF*-OsWnqcw`$OuK931RWH8~b zpW^bZ~XC<^>ZfhFz5^V#vZ=$q(l{QqIxJvFoRJK^Pey?RNEh z3FgfAE~iXHXvJ3Ky%`riiBZg@k4O`Yb4|Kp#@{Rn?Tw%9=5wSFPHOIjlLK=d`ljE& zfJ+M(3$5M~)^|2l4BNg0Br9)rm8l>iSPq?%Ik#>&E?*TpB8a(69#&GE=lY$?K1bq~ z+glCqvws~}*)7AdKVE*sCi>RyqxF8Vi*$x*MKFaB2h${OCk6i~?8d3IShJ17t@or& zP_|@c`cnw0yk9;nI}2HXiId9HHyM*}=i%14*ku~s@fD0)43v?|wC|zx&)-v&Rjz8I zoc+y7#LZ=g{ki*@mW6Jvd8Rx$-oL~@3KK-jJ)j$d- zpC%+ElvYJznYJF>KcA}YZvQUNaNqsO>Zm6!0Wl%lT0a|Kya8Q=T6AK?mxjZk+{3H{ z3mN_V^~y^|H|qD_vdgEBI5)1JLdi(+?^Q(8_+`^PinMk1O5Uha>7ri8rMj3zTT)p= zCOL052}f0Hp4L*6PQSgHIam$r4-W z{58sWt`mFD%7iYdQZ67@wDfT+u@6oV^@V(f;%)+6i2@O`NDK&x~3g0Yj*Iizk1gtZ}VtB>b5XOwvn zrhn)du7y7)o)$CH=zUg^*T!}nD1bhgc|R|@Ii-~1=y6G(yJoFN@3pOHhqio=RvUT3 zmw*c=iw15Sq;u{{9dj~NhU}jUm(@!Ivj5RNPnpvjU)r#E&3fR_N0rU?&W~2z_gMbb zS`z=VmNy-Up>ni-Z|2LT-8>JaPF5b7s@cJo2)o@p(ZLG$1kzuXN$v^FEc>*a&$dRZ zv1S?vC~CxZfueXcIIKz2ty4QVQ-B+iwOw_(bqwP`jEw+cy zH8Jc@Ea86R)6*Zu(I$u+kJg(+%Qh+dm7d5&xc6W74LQx}Kw5_a{ey3X zg68-2LRJ5Ta7P(u|75o_XY9b4M)?ta+~k`aa7ejdSiJe*klz38know&A#ME^4k;tS z@|C>S6CPuu6M|PDgza~D82#LE&3010%qM-D`%f@a5i0!HCJus`jUb-e@7fOZAV_h+ z*!u=ldkWRmM+4zZ$B#!*CfC~U&QU(4KBKml$F`<9Y`aae0V#`KA#FV>dQ!>^cAF@P zUIBwM49RdfUiO=JmWlSx+}4KO-?E0UYrSY|AKzN+a$+5Zw^dqHzpJL_>=lbCSMOfN zL6y<5g6q3JfDnHcBA>QaqZIQVRR1^vuWu#dEg}0yrE$sDa@_1vL3{Y;RW@) z8cFT-^ZWR?)e#Gk^*)WaKqjWSGYH5AS`ldrNn6kIw!cW5_a4t(i#GYUO`=2n+Y;Tv!|@_(Jq%(lk?xEgqoc-T&v zv1N0_xm{F$3+M`HHGd+0WcsaF^5<%*oy4Kk;Z|s$m`kQ%If{v~QQ50{*h~gnSmmlf z;X`D;W7z2ePYy~_3pI__vvi2a>`Hp?V2+KL_!EVU`MlvAgTT8U)%S~sFGsZ_#()V+ z=%Nf)D+*5CbG$t=5;lCL`>-?}VUN7WaBsqO;3n&&FyyF^SCqGaIvBmc(3$6<@%P@Z zr3pPny*XJo(c?={X*h#&VNYv?*P*DuQT@nSQs1jlyjM~G$lQzP3!9o_eCe;%(}!2o zu5LVev4GDk{;JP)k$u=6G57KU@cDOL8Z>xGjMbZCu2*P%H;3#0R@6#@p-)*)T=#T@ zFUU(*4ZfnHo7Gj9$9eQNDqpu=tskGBF?gKMAcWW!`nUvH7^DFZ5H9m~H|`zKc1ge{zoytv22}hbUghywIh-=t-!>7OK5YM$ zza-SBYE5jgQ{Yz@VN~r@Xz}2qb*PirzC{U^A8#xbDin!Pg~ZG+zhmcfX%P?3Z0UnC zC8Pr?QvF*`Ar!pXSE|Ij4Gpj>CtlwKW|L!|5VaUJ-N*HWjq27DfU(}(DF-p(D>Az9 zw67sL)eQhP@0iJw-*lX1DC0cz#e@28o&Fd$K|fNOFUoo$cNLP5exjalHQ1r2z+Mm zhl=A?MS2JOJxVdzT!V6=CspNNoLQL#Nl_~tNhPZFE=KCEok_Y$6|bL$GkdQ~cGHy}I@Ns&1Gyyi&%s*N|3 zZfCld^BX_c(R^gjT=z6)Kj4w|aKs zQ0kP!g;oY%FQ+4M_UaS8K+H`{K;uVQ z+q--ec4u%iy#5QFgjmH{*t8EP0>0VAQtsd3HWL0ZKT62fvP>)1%{7 zJ9HVZ-WlLa9}aQFCYRoEJ;|v&NGJzP0p_F&TFy;b4wbDG_LA5e9~&+dl|YAiC1s(Iw`BF2ozN`;E~+YSP@fMie%P{Z?w~ zCCp!TcT2>4#WoP*7d&4^d>eu^yBlco$u212>K=~V87 zgKke}KW+nSGkdm@SK@iNrcD1Wl0{TMNs3@O_S8S)hWQ9kkd z6U*5#Ow60HEad2yW?9LL9Xk(Y8qT|EZ=h1z-!EuEvz}aN(+6p(&SHlK{qA!|Db0LG zc2mIu+rdN_T8sfEf$F&_*`LJjb^8wRaFwP+xj&u~8{fjbv(2wL@bl;U4Xh>ow}Sg` zM}EA-u5+zI|FU!_T%olsoE`NzLgN!Ov+Az|%L7S4s|gj0z3f%2g5CMk>~hMC>-XGC z8t_eov7~nh{A$iR(;rb&-{9Z z9g}a$6~N@sfo_vDJ4ynz4Mi1MIEUyr&2fD8P*4sQ6%MCLPH3P$u7IpsXh6g#w*S4yDCK}-fjJh`rph;kytX5#rpcBNYAAAX-=xcM29WS1*gq5dElC0-?` z+{?QLh1$H$n7|y0MP}U9V)5dbyr5rZaje!EeAhRLo;2*j&@8B-ITNU%@~tJZRN<5& z=pj!iH#Bw&j^EWV{L>IWw(tBS zejJg%8`&oEdVTf7xeaSi`fNI;X%244&2HulYi$_RytcIL5e#-^*ROV8`Jh}MXIG@S zvZ!wyMWbb~YWJ02)3tp0B{Xby|72&X61II$nV_*gtScq|;&B|S)#&!!Nw_X?A*JFb zw#cmc(5+9A@<;RNHY^m(=9$B=CYe3cdv208ICK3ofmwLwxx)K&zb_&vLZ;?7ugt~L zFuQ2+SY3>IZboUb+d-Z^D?ddWQLT&E%w8oC1+4GF8!De2%;z>)Y0aV>`48*)j5`*nr+Y$ zkxz;ColfkyY|IMntlw(Eu7)3fZT7f-mZUf*kE*6Q>WF4AH? zmsKy2g-O7z7h>1jnzj1(UpW+VhSJN;GTn^|(VWMK^tLIoa1%$PR%o;S{T`^a%={1} z=*I&G%)eJ7<89vdD{YuX>4a(f5Db3Ic0HzmO1aM6$m~Dj0R1rosI*MDLxu7Pwl6E5 zsslJsv^(!YzlF-+#s#OJD5>)KveI{W6&25$i|3r01v}GtWDGa{$t~xZX@x9JCp)gk zTW3z940vKUElNM(>!b8`PovC@JKM>r({KeR^g&2Gur;lm&|yrg^w?|Xalws)z21xJ zi9eUgV%C_Jf#2}TVvfexekq~|W!@{(d3UWU8t=?YfVmVzh3XLVHlOb!^!EF_;MyVeQf110kK1s+$Rr4E2%X`ce^m(m&gZil^8Vq{KT+V;2+_q_ye3H)lLbd>hTd%BEQLiB!o?C?V9SzzWV z?r%-2WbIyc5^tpE-Y&{p=31{d+SLrjo=&mcOu>HPAH}rIkf|C6Rt^;Y-Hd26RJL=> zBP9&QB1}Ndo>32RHqzXMu z5ud3)Mq^zrUg`u?RFsm_86~1Dy_??LhgZ(_eHA<8)yVVNspj5h&*ieRaJBxGd11}w zN_C;vS?Pv&kFTvkQSRD+kbUJP-Yes%Y8s9?NB0uj)t%ZF&BHTM=`fo;^6`mzU!Gi6 zO3yUzL`HO=jmD_6*Qj-R{JH#I8~YcfR~|AOj&F4-j%EJ23((<4`C8M$>G3nVmUMO> zsFhC+7G#rtpiOzbclFP&RC;bJhoUmpaO7p%&S-+^D=vubj0xqqKN}lJ5IfvT95`G5 z3NoY6?@}z!XDe2QOKoSon$}*#%8SpdYctf$$2QGNo3{g};k6XYedv}Bt?~SPMZnv( z%h96HHTkR1uyXKDh~9}v?Z;c7Zw`gGJl8`5RxA9X&PPaW&x(B3mp7DAvHuAbbkRw@ zroQc8rSpP6E!VDTro7`lL}8e<(`{aaau#&^wU<~6uD?Bgt6xipLTqK$;uBD2?Ol8| z%dP{m^c?aRC7};onR`@UM0r9F(wM?`f#p41qR`msK76Idz0b+w+*DB;w}5D<{$%C$ ze0FFYHZwER%o5nf7BbE)LAPb`<+zVr>&pzQP;7CBXm*(`3&&&mJSR?l$*el*Sz^9(xPlKbA(C3)yQk5Sx)TOnY@KKso`e%D(W*v^REifja~nbv(!bv zJBAhhrYc4?uPelI+B%x)G4~>e*dWofvZz7N3n!YDt1irSLeob(=YA~Z(r{|iBLAGS zeCKJtSh53P`gxrRx(|WrX9v@Nsao4=wyCTzv=qeImD$ma_clTgb!)Jpb%9+vGgEKuLuJr00CwyH1D zca1|~C2sFJ_G8^ovFS@~m)Ac}cm(A#yc=-a`V%7j((SDLgspgw>8H)ryO4=?&#rF4 z^@0j_Oji65!IB!vISB=wWUQc^j@#8j`}qQYS1T$nbxFnpD{ zZ5q#unKqMVmKraIVyEJO?S8R=_D;kcV)q?dP zBZYU{N{=!R(eV42@ZRoXNB2hB5T;poA^Un2!uwFT64np6PKYIqJvlTPhhw|nw@Hs< zsPOu~j>Mv=#-9It!ec{X7P6Ew;SObAxR+%KQd)uH>DWfMdA=>}^gx(+A zjf$>rss^G-?(H!k7j6e}^H|N07q@vrw|93wibc7dCEDQ1ju+6E@}h=i%}A{0ec+la z_IPV+WZnTBw)L0cl51it6v)o3KbIGhn4XzEwR)Ww`;kbH&f`>_em(PIY@RNMO5Xm} z)lZ9NC>$Vs6RH{1>iqNkNz((AP z3g#~VfLA>3^WEd6VDnL)%Gr}>QXvL(a_pJgdk_Dy)UyNV*6to{ttU&1@OdTw;$>vF8Ifj$T=K203Eaio-DjYx}T5oWM%LPYYc>|oMgQ~sx@*I$g4 z@Lj)P9N)_WOD2bDM3GE;A#JZlK9qvAw-h${Ub|J5-V+Ts6XI%T=dCbJW zZMb(hv=$|T#fT|+At>9Dv>e)8Z{gC(H3;okQ}FYkgVuZgvqvWf8TRx?G;XgGhvjM% z+Sb1FqY6&=xpL*?$o1S0fnvX|-QLAMXZJS*NnA#f_zma(^U(OePn`ZcKnBc<|H2!? z!Oe#WzuV1pT9%t~nm0{3VJAr8@WPtPPOxK53B%_1#x`MUziKit7PD+qftN6BPeE{qteJ<$wgRt?Bz5fZ=X@75IRO(}|cs z>`}%Zd^CA@6ya3jYR1A;CnE_NJdr7r^a7UrmxF$KDr)O<% z)EEQTuj%rKz&P54m?D8Z3?sNV)nlDtR_YyqBqZdnmF{a5vLRL9!{Ok8rQNQlmrdW4 zJH~GcBi|3u#C(aY{9ni-5TNzv1y9Y61dzF)g>SOKWW4Y68bHfP)p5X%8ac`L(_l#- z8&Ja-=QT9t!d%F8Kg3ZNF9jgwaB$V&H^lvxGileB{x-AqtSt1O)ad`wfX80|n@$R- z7Z|D31eGQs&)-vr?)6VgiAj8mpCs192XJA2GJ}TIX|LCadp463GR~rf4 z&hC`1mi+d7Bg`=%kK`K+mRG>Wc4%#FJ5AEAhh^+Z4KX~(`|`#h+xE7(sv6QTI~Z>r z=?cl`-fl^!LHI%*FnWHPIG%(iCn7h@_}J$KLd} zAoqVMj?4HD`N8pJp4|w#!;D=R@Fq)}P5_RFDstFlH@E!af%A=oj9PEG`07LN&bT*Hay-g`+vA{)iI(xlZTWVXT!Am0kH@Z&vS z(RgRII)?kRP`J8MU^g8w0%ZJ;-p)N9>a36Bh#E!4PG|wuDMn7qTvqR=L!q%SI$j%8qE|TI_Bbg?&C}W_sFvUa$S*c|8|@ zcnxOG@0{=Xe!u5?j^FS7(JNX^un_aq<(_g5<~T$~bZS4_8pj|miyA}A?*9iLyIfW2 zLg-gATq-D3bk{RB+KpFi$}cOSd#9CAU$r-Oe?~dLjGY);wLzb<*E?a_$MnO@PVJTV z;@n7o1+|p9n&2Yyz9eQ^rtuS<`d7I4L@ASW_Ztv z)>BtLiml`VRISGj-d7?Jv=b;=Gh0r@4E`D6M2qtfF)1c&q7A~miUyiTr~hGNGG^^R zl;NFHg;J^n0PldaR8l^y5r4J6{Rci_%!8`D~Wcr=F@VsC}9|Mqry zEK7~_YdQE(Zqr=1V$`wrESU1^T&8zKPUTBY&axpU2P77+T-JgO>4Uk;m` z*(PsYWOz;_8$=S8)wDa|;u>ci;{RRl+z6qFDR|2UTkSA7RxhH;zSp*{a+4D5d&$|5 z-P7pThpMVr%>=V0YlHFp%UM*Wf#yQribkLY$~(((?g=@Qv>Eg|F+hletL#rd=hgi| zrUnfA%CM9VgmQEIEXIcM{{U*>2M~4~PxN7{6Xrv4d``D6M*OLAauyZ}(z-rvoHF^1WHk0ig!XAM7vCp|$4HrLRvyd%h92 zTE-`wR1sL>dSCes4DO%NYECHixY$B)C9z7mg~?L%Hwq&(B%xWs~KO<_~9VU*eta zk?(neku>|~J82dq7f|AB;uHpEKFG|*69Vd#0$|hYv`G;BM?_gx&r8SxF}1!k89df> zJytB8#I}{or0bN#B-@qWztY0rA>{Xw{?RHI4!(ukBTJl9+@EhHsO%9UPW*^10ISay=elS{f+|E#LtNY zL(;>cx;2A4r10wsU~?nZSPqsGO8WlYu$e#HV64}9yngjICMb&s*0J5kL$cWJ9fdoB zgD(nhB?Laav1jT69Et#yd@Z^ck7gwIY@SZko}{7-p94tz^%=U7C$&<1WJuNE?i}^v zC214ss`k&E#zUMA50JD}fuZcaX_4-kP+e?%5IAxBe)+mv!d6A5oEI)P(shgR9#lLI zMs8}!+01}Zi&5rairbRy2#)obp_$%IGQw+Wl~YDF*22F^P}(*y`)lnJtH-uBBoO;x z-M!hIyDcf9+D|PnnmF)lTqVFrk@>~Z)*1F<7^NRgvq85dEN#%R&0s|I#PGc~O&7@M zDL&P<1x68XR%Y3fP7}9iXx2`s{;I#>Jlfw$+HbmX&VV0q>T=%nwfvplH$U2>{glvg!T9bnIIJct7U+H{=44{W#I{%k48ID_%rTQ&@us0o zMHBR^MinGvyQlQtw0A*7UdJC+R;ZDcb8MJj8Fd zE4G<)sR$Gr7SS15_-S-XykO|4RWt#j5ebH@D@y^^Zavimp9i6Pa`d+Cf$YTUkI%;{ zqYEkW{axRe%Ino$3R>js?5uVDd8<$w_3<^sx_dDxL68IJ2xsQ60e;`;lll8mcFMs{ zGrcoeYxN>lLnaIdMt9)^MtYm`xzre+XWbwX%g9N@#Ex78N-noR30o1$ar>j9fyK2f zo#2u-p^2mEXQe^TH3{ZKKLeBiZ}#m=0hCs?O1i5o;*gt(0RM7*<;h}NW#zLBpM1ih zysDfrvWmS59AU*;g9xpWr4}o-z|N$ufZ>K|Us)U$4JcS(Uwb5o2n z=^j*E%&TrtOa7w&UfTS69K%obl+wKK?^E|BDN#mFPmz<- z;_EZ}NYc(@<(Tjg#a#s2%^f2FL~X-w3KJiSarTe#I+AZpWJJW2!Yc^Pj$e3XySqXQ z?tQ)#!<->|QuJg^T&n#iXE>l%PYKf3y47VXz&d)sg&lb$K#CXd4GIj9M`KcHf@65; zc5n8$G=4%dT0H$2a{$mH0kk}F7FE;s65RQvy~vI9?`Ar2O|<8 z5yZ_?GMgKsN=9fgM%{HDq>VbFK%ni6fRYkWxjj-5g}dQe-+^83!gMr!=%W4iPh2Ef zE=c~Edv2*&{=uxIjN-HDIuIYwF$*TST$5;xgo+>nG%Dt`av#CE8@vQVq6DeZ0BV=6 z(5m)zi*|Lr(tlJz#CQ<65^0F4 z`2l8%%vDB#W$)rAP4V#i<)f;J%xBgbv(dx^O*1XgEJ>hH7~r>S0L1r`UCBTTa{w~$ zh@aC&E*~KVn{q!Qc65FGKr^v=-?C<4wWY!(@#YXNkrd8w6G(da#HDsNY^k8}Mt|D- zO7~fUA8%6_V-R`VmuCFp&bGpYvq=@t8Aji&8NOQxVwy&PWVk)1HQ3kXxMXTWg=rajb~k?v!`$e zrc%nG&rmSNy#TV2gbNUc3nM=iB0qfTksXFfhfV5K-okN*Bs^^?5G>mY>F!aM8{T0l zA9Y*kPfKfn87>j*(>ns zr>0r5U9Tcb+IjAkN7m14oFyAoFu&wwyTRlSfoq&fG0hSc!44sj!*QRvPOa)l^5y;0 z?lLP7$BE3nIawzIS z5$PSIi6}_#z5V9)YT~`)-ap_Q@8=lzM&Xp_dG^|C%{kXx``UTAvkY7JZl$B6V~{-e zn*tpj<2^dM&2?XI!cSVV?mfVNgdI++IVjo~J5coOjOb+b9BeIZ94yWB_c)dmW#$#rB-KJslm}Q}Vwrtm_=b>As+& zll<)$CFh8dE{d~~(bD>4?SpTmwmztne*edzVkt%8hLnA$e|tR3sn(J^dgRFUlyrt| zyIhqE9lkns)|LCmZ-ftg@!hZAX#^~;UU_>qgh$f2|G}w)M z?4rx{b2ZYw#3lTfKcX|Do!bB9=j2EGOzwTb`rm)I@7gAZ|NfiDJ*Gcz{_np!cmIEV z+=Rg1Iy{X3_9x(fe%#8<|8)(Q5B$%M^NRgn*YN+}(!U=6|J_yEC%Ko;{O-|RH>ztB zFT{3g@6EbWKi;jMnNK}RH-EdoI-FVOY)h(o-&n(~FG@Dh&2MF3;1k^}A0gN~)*w?( z+4w1Z=Qsb)@C!J=P1*hFb#!(miHSd*K7G2tXw2^2OAg9dr*$YZ`R&BTO9MT5q00Au z{$eMmN>!K9r?;@_^WNL=@gJAzJ(frFT<5zeT_@+$wW&(6XBhIT80Z+o_$*oyPv6+Z zG}an>ma#fYOhnouNhK|8ZoDVD*Yn$Nzr}@DH)zo9OEB*)Dria94)66`o*!2YJffVH z>L+C1pJ-AW#WW^h`}$|uFkX(zA-by9HKjH@H|)23{OZ@%|6B$golw)Y9f#?e<$pbO ziqEv}o@K^nMt^)g$Du=KHg5g?MZii&o=vUN^x{?X`LQmIhZnEN7M;1dcc>vw`cdnK z*J}bUQ-&TZljZ6`(Pz{?-RNbXZ4B=}{e$l3?T&BCw$PfH!uc%DY!J?By2&iSYteF4 zY`)W8Jz!nW#;~*4Ew8sg43~^&FKF~C$80Z;v9G9@m~5>_@6(NGYMD=GivIq*7X|;G z_4W;fin`bZy)AZ`jyJA}2t8Qv#cMsswH24uxegok0Jo)?K`AMzmjRvsd_w5I<)yZ^ zETbFS*~KkgyK(1(Q|~I{2R2^z>3Se}C+mQXirB5M{(h&>?tgCV@j+QS~VqJ;98H|%f@CcuwHg;hjvZv>UST%?s3<@U1p+jbr^r9Sz|oM!Gq6ogzn$J zpX0u07w(*hXP+thZ=cyz>_%B!T6$GhZE4K=QaRgL7S~XnBSiOF?_#DJ@5RS-;S2Gy z;WMqeUZE$;@1MTmF!|5W8)PdOKV9hMh#Q~Rhs#)#1U50(wz7fR=#2b)!Qkof zSGm$|^JBT!vX_MD1WjH)zLKSOl>bsT8;kzkgo{spH}5Ga=2-l@F5R-Pzr6j^r_<5h zGAv5A($~LlTpKlNHzs-KP{I#d#%{-zg-fhLcE9cB(mF3Cb)72xsZ<+&x$ghXr!TzB zeERHJ3p~VErCz1mS%n`+tgk2rb7?DRX+^y0>w7mkdQ4YWcPEF!Z$_0t%2?2*`=__@ zS+<>zii#TJTEF5j-7jm^Rj7=;*cx-n$Kie550PSSDwi%@x_;xvon~|8C{dU1zx(cf zz(MI`*V!TSwudp;M6oN2(^TI1_H2`#J9gYDWc~JAy?otYfBhY)Wc>Wt^_`qAJ32bz zma4wly*mNlYqq*H`-)1*`0>Xdzh`Dnv1m=NC@(*|wzh_hqpYSD!san^A=joy=g)1s zoh+(@xpiAIO4pLHxMt(s#YxdV?u%1Nc<@KG@*nOMuujEocFqqNXB$_4zjtr)j>B>f zuE|G4RbQQ)i@ff$2i8#E!_xkn4glS<;At7~Ue#^R7xfZz|%TCxMGUTD+ zdKDBD?sDsjlgIO}GFX-7&CvYFO0zlf0&Q>+7kl^Cty_g2b0g}sReb{kR`I+Pm9!xT z?K6+8I^ywqi{-h|rRiFUojZ31E_o>B+v+u^X{zF~PCCk&Jb3WH6#G-7)U%kpyszAU zc%(i1y1##Gb+|wvzI1VUImNd3X|h7x^<<@_`v}}*>#oA7iQ@z6tG938*2vWV(`oco zoHo@dPKO!0`~KUvhVQDwRb8h0m9KocQ7Ku81DRP0g+S@*CI+Pxm1Bzw3n2+-8f~`Cz+Whmf@Lmr91iMrZ$E}lkPhZ*UYZ?toOu09q~N7r5g&z4qhV8TD@Z7$ zI1DxgeYSXD(PBD49xuRCwghZm)`t=dk%2rb{bes)hRAS@ruVsSJu|n zb{X}F+hKE5hw({Q1|4hcE_U0wdv~z0D1yNC?d`Ko>DnR;EQ06bqyk^IEIqRBN}iao zB{6>1pRE}o8+hy1JFJ0(!Em%k#AXi|M+h*45DxrM(e|#VRGgyLZ1IetwfVjk{#w zey&EYxyp@a`t!xI;|b?uer@quU+!ta(+r(k8d;6%aT#FsnWASAtc#ofX+N^Q_o8&* zk#`O?NQr6^%OgA%2)R@&x#JM!&AWF-qxG>qj<{*LR=x3Jp1#Ae_RZawjN7sIb9GBS zTRIEuMU?Uq5)yI|HZ6|BEukvX#<~_tN_S@$2cq-(EAAX_oDJeod@f8G(`q#5({y&u z?|NmRub+xkw1lG8;j4ah*Knpmc|;UtZDrBhz9w2if>v0y5+&rdRy3=-z7(8`yVxFm zDv(q4kz5#WMcCcDZzqbd7L{4{B4T&|yrRZPnPzj+l9HL^1sYff`hjg zjB2-EnlT%eSYMR4=zaQUq-*Jy8#bgNMP#R|XFng@8^o<^Jz5uYwYjpgaCJWIXmt^%k@ph?{e67~qg~WI$J!{d_-6gLFK+K|8SA2IiiwHwXlurvxtWS3 z%BoPtex$8duWZ-KF}AFw;iLG`xrFlr{7&6gbG1nZj zNw(}bBpA4%0g0-q$i+VJsJa4o=}MO8@~FVULx+R|IXO5a?fa`PU?KTPDGpp?Q`xd* z%ia_Ee1 zw4U_noa`%aM!|O+YO09ovkUlQ!=|>>DbrjHKBJf4C)kt7I<$+0g%kt9$i{C4Y`UA? z*T*Wy%P<9~z54efU41iyjphiCO$>V<;F;u8S*WOzvl(8i z(nc@u)+a=J4$b7g1||p-_b3b#ut`VF5OEk;oE~TfX3n+hP>~KgdO<-UK#}iUp2VHf z)wytQdZzpMB9X|$RJAmfVn9Rlx_ehtQq_9vGRg0UTGCPkdfXK&Dk`?KideQCr!9~~ zx-g;VV`ubyymUPswV|V|xTt6b5gslc&E)RkfkGk>zJTq#lbwBBsiuocu|W9oUviu8 zHooEK*K$W*JPBagtp5J#?a>wj)}3{=66>kpDW(e(z3O)IbI*3`CYm=VhlxTI%D)^HH>c6Ta=# zlIl>N@(CL`Rg!E~GjwvVZQi(%N#_(xuEU@L;Dj>rl*4^1(b)A~yds#DtF zcvn-BqB7OVDq$}R+X3WMa;wk%SS?<zzrK?OCof^=GZ#D8|cN8j4{eAg$0L>0cmiZq6zB%-INW4-N2*7O%Bw zc15y$xHHt8l7LqfQtvv9b{A{%nK#uPj&|2Kr}ZPjo-DROA=%4s z8FFxbca6C3Au13-y?^q_wE5q-1@_34^8rdlMhF-xSg``0)N0wj&5pyX2sm zm4!Zk?@2#akz`oU#@OyFZ=Gh?;=X%n8bL?Dl4V25$%se?{^yOOhOSd14ZQ; z0I677TOL==&=CW+Ec;`~+AAp9V@|zzra`(8ET_?>WPQyISWLWr)tu_F7Yq z6rreKnHq)k_4S)_EmT)0XpCIVb~ZK0FvA4{Q9N-;>us3^uGq7)&8oTuIxHWPR_Er= zD9U^qwn7Tx&)Xn9waMZD00}CA01GpccaptO-|~2AtHjzQBPI9>l|{@|3DK&EqDPg! zbb|n|`m^kO9;e@J+s*Ypvz)Ebn}J!T+N~TkAx%}+qxr6P9X7$uJtv<2k(?D>eg}3ejuhjbP;*WD|E+n;mX##Z}8zhlhkDihC@DH0&GXLNF?HJhF<^krbt- zY85D8y`2)(J1qfXRe%I#y#A_MeS;foM&gmJ$Th6^ZuWgWE#}NzFU^PfIX1+I!iI@r zE}f!8hx`LXJd6SPC!fHpgis?{YeowOKk zNh3VHCPK)-yqlnYkn(zbs&>t+zt;*a$A>*tBUcJZt}4ste9_8OCCS7z&KD+c^LjiL z!}+b~RU6OrySTazwVOoqv1GR%9ZU6Co@@8Ge5ZKtNef`ty@Iy)IHI!wXUGn_yPSdt zn0Tr0xzoBgMa9l8>x>^ulDzn0S<@#5eQSqH0rgYSp`%l!-*f!2ZD0c`p>QpPL{J%gfpJ$Uue%VQcjR##lAm#P&!D7pj${B2M-_SgjQfu6H(9Z zwGx9Y&iUhyPp#9DGIfAik85q*;kSjJUJ03myqhGhD_5?(1P~*zlv6eR{hMbukiUcZ zE!*y4@j9$ngn@8D+GFuqmHL^6(TKH1k_as&#dCE@eNjOsI3VCY{yGS_ZuSP`k1)9?=Qmq$2fyI56$sDU zu_N}B&vvtC*LMQ^D?Vy2u{5I3esd zmr6KJnr0qgv4>pq_SzFc)k=i$X)W(OP3V$D=D^`DV<5Kii~Iveb{_%7Iied!_g zQSsW+aHw_naC>%wj_cq-?Lr3**%hofS*9+ZZ-vZX-sSLS7zO%Nz*dnB-Jha%bQi;( z<3Goq^_SbBi%KCEC7PrVCuJnuis!VGmDOrT?a*>1N=q{K!5}~1K*gQ2xHu)eUAf5l z(#Fl3Igxl0?pk(AT`dM}XfPVJ~%tpYpm+?z~1 z&vS&#Thpxs%PlgBg9cfEu_eXQ!5UlSC03hUhnkXP*a0UkwF%_L#&z;yhqCxkr^vZJ zG}v^0eg*YP@Gensyj>2-MZ8zip1=OrU&pI-fz^HNJ*FxSCzbC$qV!@3v_Z2qP16KL zx5B}%bZx0p*m>e2wRDGM3qieF;mzi`;4aMl7A>ivii-^f8BQh+NXlLFvCu?YkQX0> z?+1!sYLMnuKep>emne2_c6a_IT28A*j%ht_hn}8Z3m&pj$T;Mv768DGYE4jfAB(CZ zbK;5tsn|wlbEB;p(beLfo}Mj#{PBlOZJtRj8?68!rBSB!UDbRQ5LzlU+?;S8LBX~q zWKYe$H*Zc>If6}Dtgo$(_JiEt<-=vfEa7fVLo9IP+e=kU`X}e-^R79LwABOb*VEf$ zt2d0Pp*n$v{%T^Fl^P2 zMMz(26}vmL;@drYOs738GPq3A4J0KcTOd!wFS{(3ORVV?BiIUE?#IS5lv2mm%sL*W z(Pq^_^8n1#fFK`*=RrwtIi~pt>m(%9E)1X>Zt*ZWI+}0azQ;M^kJYn{Qmp1|36gM(%~o>OlnZ|>!L*&L9qtfi$DHkwhgB!#V< zg3J@%n+af1m7JxZzOJ)6)4eled}4aK5f44rw)a;;-5Xs_ojNu9oZ066>nB&k#M~Yc zEvWZ|{Xk71_Wm_hR7psrs#tJVoz*;C{!*>i+J|a6`Q{m+3arB7B3;EAZZW_9nPX7C zoxsTnt$fW$VMp@M*^%}}+fMV`A;>VQc6N60LwkVydzXu#jsqTuDhf4Z>(MsR0__k0 ze8@e)(XIWZ6QZ{%P!ChO%4lzI3}}T0 z;!-j6q>(}&fOa$V9hUTp7eJ+a?hCfG6Mz{W=qYJr8D87VYkEtM>&7#JC(Q7sWNqrB zUU8d}#dwhU*9z5)B5tJ_*ijrvMekOKCIAK?TlDjlFNYvsNf(S4M=wBY;`3Ztkm=Ix zTC{9m4!?+E1Z_8h#s>r~2eib;vTfV8Mr;~B*BR44|M>7jE@o^y1(q| zfrAH6XB5wcbxi^X{|4}qdP+zW0FnsMRtK2TYU(dOoA5`$V!%e;4Q-zqi< z3Tb5HB8n4=Kn%)=Vg5J*jG<3rWdgbX_lWdfIaXF-hdQFm1y31nM%&<+c41S@G8>j& z7TpS9P5~C=oQv_{v7AbY_jVoAXe&!oNkO`Z9t;~KjRjkvHbKCsj&bt_?lOmJ7ixalv58#gv**uWm607|Qo};6t{LP2V)tBMr671~ zj7NZfmOzsuId*zkv}eWjx-qk}XMlecM)Y*gjwU5UowQ7LYGJl1iW!J_>nHex!tr$C`X0faoSbZp>>o+mRreA=a_feCmPDZ}_Ln zds$j4nCsY$X8{K#sFd*ZMv^wkn8{mnS1=i&@*#=7pNNXvIE?jka%wc zqPdAUr7F7I{GMA!OJa5APM7!m%v}z}GYBL;mgUv?9_?etI$WHciQ2$FN!XdVhcK-V zDo8RK&*Ay8t#W6Jm&d5yjJ-dpXFb5Tji0~sx_E8U-zyP2lYbIQ{vaqf$Qu>l=IL0$ zJ1(eOuE203ezoAK?=}|*z5$ov&;0T9^+aTXc9m0d$BnD@5I)n}=kD&#Yw(<&TK(r7 z+b+E_oxTy$o^+j%0|G_c%u2BT{VB!HG(z_o;sD2MnkLUM?jcY;D9WCMYwu81!>ENd z6tT#cBKl=tL&NE^9YG=6p7#4%>H2Ez@v-6I6s)xHY_UR#y9;TEZ)XpVLN~-5GKjYU z0P$@?eiH1LmzRO*)Td@P5w8`Uc}4kT>@{H>nPLSGUz?o-|Ex^_Fjtd0c@~BWo-+Nk~W(W50VF z%&D0D*5ZacVhFcKj2weZ>wP`TzgXGS!`;0>`x-A#x976ePQY4Q%Qt1$;*0Y@cvLQ4 zyw_MhKipb-+!HMv9~Ijw4qugGA~GQv@rbr0-p(ZXxhd(+mOv&E5W4=VPy@j^C?^F) z`nd-@fCRlcFB(63!jte&?P$D!M3^1`1U5oQZ`(sXS`)CCWm>tZ(J~=a77Sx51$IUW zDzYW9=WhS#r23!>9STsypd?WsIg;3{-k{&wiv1jsZUaV6nn|)nYMJ`mGm1v7zxd({ zZ^nv+cfq>!x_w>HToOP>+RL(+Vt&|hNancj=FOWsw3)=R0gV;k)kJpKR&|S27Nb>> zfvz00U-9H~X7A{VDpGBd7_%3UCGQ3V=tIZy0(-HjQ2p6g5J68ePU`p^z}Gt=>`RE}jYK8gHrU*3~Zc>>8t8*6~_( z3`9R=aA|B|zH~jU$YuIf87M(oS=n#F(;qDvKg*rkApD8zIPaSQ7MNhadKP_KP%#8O8;Wc<#8;tL(M!|+9#mNqjx~z$P-Zlf`MxC@R~^zJ=%2# zY2J0vu<~=hbU70*6H1zk;m(L|t+lT1Z+Go7Zu`oE^q% zf3mS*bhEnq#FNeKn!bh$53fL%Gj?Mgo55p#iQ*W?Li%34FFPAX``=eh%+0j`Q8ARG zu5M1S$j*B@fq!b7Cpzqo^`1n({g~RreO-ONHDdDe@?x`1iiebq37H{PySHr{chTzx z^Jf9*^P11|WMAq#L-Ud8vKZ?u@RscYQ0LQPXumM0;_1``t$@$IUxt=R>G5dIshxt% zY`TZ??uT#kS0De0b`7(ikede;3TM|cACIa;25@L4qfbSoNM4lNm=G-!@8>*nt`<6-a z+_{I;u;}8Qt=XC=R-o|uJ0xshUuI5z!U}N1@7v?g@FH{;2^1}p3frpG{@fCe<(D+t z!ulJnT=T%nMNlepDSHQGONPF~auK#dc_MrGc-VyO4s>nSt_nGE3f+z>B`%(0TKRRy zyA$NuX=n6Ja?P7x9%qBw;}whgu>^PjE^v%+jRL#4UbPIJ;?AkpyJLkI*+din`s**t zjAyJ!!Xw?fLQh*;RiSb4ve>^}5_5Qeo;H33duW*C(pDoIM?EyL#HOowwhvTpflDC? z8u_cTMmO7YXo1AbM_cc(LGSJQ_Z<4>9+YfKiL^qg=c-I7ty}CbFkYH)`7-<)t23*?ibQTaj_7nqP-)eHw*|b; z>J97yLukX%7DrADe<3k9){N$T4qDmt{&^I}wm9kCRK}k#J-=$p_^GK7dit|ncd^YP zUK0?x*W8pT6q_DhAFDTHZP93?EoNQ6&?tpfSIl7Qf7p7?p_G50$#clW^nFCe1M86u zjYV=IougCJB2T-oZeTI8$)&_Tf4F?5Sh;dEP07k86`7B88m5uJ=2QuU&Js2)Zz#{Io^8FR^=QJ6*ldjDu9x zG7$n!&2jQ%!y$3>zp^W(Y26eGsEE=dOLfJ5Q%HX>WwC~>TLaYa`p!d<4$M-36r2pQ z#x;=HTGay^1j(%LRw&K@3VJke{z+g1sQEQHE^KEYHD5rouzUmJQ4d=KYmLeHQVTpT zifmD~KI-#ox**=+8BF_G^Ft!zt4ZyogjxHuCB?X+yg|4Om z0W$kqPyVG-khQaE8P(k{2R)W%E}#?4QWLC+`|rEJS-3nJ}#Hqm-TP=%aS zk5SG5c(R1`d^;DW!}4x?+-6UyE^{6BbVg02NJp91+M-DpRGC91UhAHtodu1DqMB2p z2@l0{<-Y`=>r*O!Qh!E!zoZW%JHKO8>KlB<{4ZO^3}*^%AK>QBIO^g2iz{}5*&f|u z*IZ&^0M>kIo@;Ec+>Ty=aQr*MhV(>13^s3PdEvn3mhzQ+$nhr=doRknyBEPlW+pfX zlx{36FK6iRp#bNYHPX)uXs~v5=l00Vl~B=}HW?C~OJ`^55{VF@$N`s~WZ{|aJiec_ z9~k=lMW_C%EEsnz_1fHT2+InOZ->R9yu&~K_$wOGCvy><;63P-PO3YrtdUD_f6?M=3kKnt z&^OvvH?N){3Sp1edY5h0$f1{i{q^-KFjLaac{DPVw6r49b&8Ww4$$({lzUe_Vt)T2 zFv%u@(F6^)IJ zO$8t`^nG(GyKBs_IS$--_Gav)Wt+uuJRIUzxl`TT3BBZJ(x{a}re^9ur;b>fK!!+% zi9jHtIW7%UM^;&OZew5wEG;eV&7DAlHD~$qZPFamIVg3v4s^P)3@)xXIB-Z)Ec7K0 ztcOr)JIbI3rXm)Mj+~X-xocNEd{bu7h|S&r%eD|D^N`cxlo1r(7l+P*_R4AtE$e^` z2z##N_o}IMitG{d9m2QC%nQ2p0Sd*htr|RcF&-DG8`=Eu-26J_q8vXO>_q?O_p-3; z%)bEi@X{fBeqpkYf!$95olk0o*KzCdabJ-Uk2Mbu4~%3b$hIy`VWt0`)MwWCSdjIfkC;Sk#H?41A}Tk{kCn2Fr7uF zkNcs4;sJe+G`%6m#G;g$qeMZDxMaR#3AH5BT0Hmx+=EoD0h%$8F#)v7z$t)Xa_dX2 zrM!+qN<^=MJ4YE2&{#dtaCKQEQzK0nj+5Sy(3Bsg@|d+ph(SoYVB7aUy>6JKv{{r% z%E|E`EEUN??*_K9N7s6&=GGsJQnoG}GNb7s=0SMid!IA12=0ReXGk7)B!+>kJOyjH zI}%==l;Df+yS?9tpcx%h9k}9!V&FPmEzsCC0P|VHW_I^`z+OaiAf(ub-7^sv+&$mA zihLCogzixy_KCNFpH?Z6?9yKkzc}(_ZjKewNbk_2dFEKRH>8zD^%DsA8$%Ewqk>iJ9 zo;eRag1-Eun}@WVTnp$DLwS|fs#d^DNN>j$A-h~`JMZOF?DlwJ!MRal9qMZQqE?{? zqx-v;I|pdH&X?etUvh%0&KNHPFd}V@0uWDO;t*XR(^H*Q(dCpqw z$;foF?4hJPJQ3cl_uO!ut;UYFgHCsf+8M&e8VF1p&vj#Gu^WHnKz&k(nx?K2es^tL zG@(-jT6KM|d(8_xPH9H57h8I~8dj%SLmK~uMBq3<)z}Y;d{kY3A(XsoX>N4b&V$0U z?DxWY#{5kuUJ=PWnNJzG1=#GymjA;(?X&R~;NS$$f?D#=bL08lQ=_;w?&LP>X_VS({ zUF4qXr!TFcpPC2N-u^aL`49X11Z`2jIyJ|();qSd1EP;7Dj&%hQxmo$(dt^Ip4z4B zYgA&*P)=35fDGM=>=AhP?%mVPVZ>zDnFYt2mVGV&WZw;Sgc34-Ai z<9$!%zEwiUb*L>fR@NPSF{w2f?PgObg8a*d<)==K`M7O%ie=E&#g!}+a<(Lx;0y;{VwuGhdlsEkaV4WPO$U8Wz_O?4RQ2LjwYyw|LdoFQwZQhQx zIaYm@XN(`v+GqGhh40>q@e&ngzG1h2$uH{KuRJb|U(7g@J$y&Y?f5yl#kI<+t0lLD z=F)BE!YaoF!Njroj0J;swhCmAk2i?qHBgOhKlzP+Zke#7{Ov@hGu{SoB_2NmGnH^U z_^jK+_(1{DPT*z9&dYK&_GJ-LK(xwIqS35t&d}9Gr63-itvioAKxeoTY(PgxXKL!? zTB%c%=aLz%ne%oDVCEoPJ@OQG8ylPEbZt$jhf1VppJRFfO;3c2JOosDmjj@h2w8*O zFM$h3D#r~7o?&#)r`0Q4@dVtub&*1J>iC+Cnh4|uJjRAH_=PioDTlD5k6Y56)DRNm z6f8W+XGbJQnlmU0p1UqYebQ_tVj^O`;rET(99}>E!e^QVK3=E^^-#14L4Vg)YXfAF z#M$|Z>MdfwY7NPc*w&za#9w2_YkPZpT@fnu{^2|(e#o;{ZvZneZg7L<_|^~=NEy)R zdnTsPL*ke`K;Qm!V}g91P(K_air2n7m{s)1cZ?O-LMFVj9dXSU=3%oxPS@a!-O%iY zn&FPTjJ3sDuOnPsT(Y8~qFS&cD&i3__=f|n_g>8<7C7L5P@M$MXwChD9z6+^bi03+ z@w164(Oc-;o3>pRhu`letB4BesEAqv5?{Z5J?q-?(h>pba?M$AGT&sE=pJ*1mNA2# zPBK80cXbt0g@)C>%$;w+OSOD-67bQJTrBP z%di->Bn8qhs=l5{FbasH)f_h@1|o_d%Wc=>%f_{P%Lcu08Cl(aGwjcLPq&X$tWhsH6u${`JK?<6XWb;oZmk8958dp zPS&_zW_*Q7z*_5(2jlLeYR5JEUG}`K3??$d-4dWX-b*RX$#9NbAZlMVq!13L^?vuc zc9ZM}a@n`8k#lGx+9 zEKB-bEW@vIQ}DJQA+aY+OIViP0rhopbHfwV@3 zvFxx+$i;l~xRh3bvu&Ek=Ft#?=i2;H^A`i{7CKdftxdS3;fll2O)v*?NESK_2H^7& z0F{T~bZRDWRcsAzbya)~T<#qsKehe<5OPE#C(&#XxLbb@Wo`+*W_jYw^{*~|f-iN% z00nk6ssfc!O9_4scdLK!;6U-j({23AON)y_tB((!jhh##{w1hKA{-=Vd#TY9sgFKZ*~o(!Fv)Z!U6&$7hklr&^RTa8FOAl z&k%INB!uDsk)gUy@Pr@G>$&&KRZIOd)dRb39Moj4gALSdk5^kw`EqD+PcJ}kR>60z zUv1n-T<)ZuibfL=UDoIdVS?EGiOy1IBjE&7WU^^Kb8|L_H#?DaIr$6 z>L|16E;4g2oo|$vVCpwRn*g4Q_eu;x+R*aIuuFJ;y(1OcqFmaUabJMXcCo`*!`(l4 zt*s2EV`tf$joXi;(tH@H4g>;$;}vb%(sN}#13r`_q(a3E~9%tLAUinZVzT2@wJN?r>U{m=VGzJyNPlE9tA=<(k+F|dp<@qO} z<0j*g%QZhluao9AL*#d2txJlaNTVjBn^EafWtid8kyu~BG@ouhPLFC_0yr|r|D#b0 zoW9yMmr9e)$NaA6jvjA(jRChVNyguNb8$%PI404kGHEKE#B_W_`3Uoo!eej>9QQ=y zm8JA|e(MxWQ0Td12LexP@T(d`LAXJe#cT<;3Dxrlng`l{>4d%58+~AVW(9H)f-%sZ zXeU&|N9w7d^!Nu{-p&qvavKh7Pyn};-!DK52N zvyC$|8EwxlkLv^WS4M+@mPtG#ahWT9_K1*Fh+YiO}Tl)uX1GIki-U%(ggiZ;?; ze08=_+t)M4-}gpKaBd{ z3}JhuTX6-wk9MsK&yhZiHPWFQ?s*Pu2{C%ap&Mvpk9g7Rx|maDCH|0S65qc!4PPZ< zjp_xjTQkG0^5|v^j=@mN#AnuUApA&Sb$2lhDM~i7yxq+Q%HomZ$Lr8#u{QvO;Tm{2 z&zg?~4YsFwfDa|1iy6Lv&5Dhc)HDiz(T8lgOk95uRco@AFkY4ek*#s6U1ax3VO4nf zza@QgTi?GQ_vgBk^F3I>9v}C#$##Jaw<})HP3R=8d+4ok+SZ62;ms?ns|bbLAQmzkCO47`}p-Il$r4!T>u0pv$Y5=^!(9% zm8V3bmn4XEY|_3w*8RvT9G)tfPy;+JCwrs*?kC;i=*VA6N*4*ZbB7Op=-_G5wS_+K zNz4}z!z9;$kK0=)e-lxh(6Hd1mG4>`Oo*oU_ac=u^^r)G@NURZCiA^kOkymw^78Vr zg0RN1FFt1I*Pg$A7Ss2LqCYk9aJBR|GRp*Z@WA@x;G1zNJotn|%hlRsgx2{s*V=UE z-;ST$m+&>_XMM-P-rjYRIjC&oo6x9GslmHH`C2#2`ZZ;gC$oJv3+1w5cj67@17-lam!e z(@SI&b4BFU1u|{sJAi#b+vHHOCiWWmMAq$_y&n^dsO9H0GzZH&3^>cf5{|LfA+H|l zk4voSYq6_v7}6FHMp`d9Npa4iJ}nf154G`l31wU-Qo^e{uCvEuKC+VYtB&PJP~{7G z)?IcS^Dy+vYR9$eq@gJBS#8gLTN|y5PANYNo@N+QY_Gx^A&;#QdxMl5?(Pr@&$uPBY{9tDSLYf#NWmoV4#F6WXZXT_# zkTPyP?@5z^d-T!rJ)f%U*RIv)U+SYTC(`5P%a`fxXGS|(C$8uNoh-p4ubup9V)fta zZ&dM4c6>%Ud}O5OXRFLhsFLxavv3UVdF}=8L$Jzq5lqkNxQ+59_}5BdtmTB|cjpjS zxSq%ww~@9J9@xq0cMnOv#Q0C|ktEa-GB+Vz08iU}ELgBzC!AdR-kfY_&Ym535okxo zm-e7i6zaIV+Zk-vsq0&RYadI}Ia;R(I|#hWh}=01FQ9Ah#qwj);=#p^ID zlvVJf0dao&vSXGNH`lHuTZ0)QX5yT-g?M=hEh-~^RYUREK=Xr1ruFyRwIuslM4jub z4K5l-NOcP)E1hxLB}VDdNhB-Tk6Xx7*HR8ckyxoKRn5L0bg1dm@DFwMyoTI;6Ra4fptC zQ82wAqaL?+OmBQ?4q)1p^coG?T|C_SZVHSCJCs zPG)fF%cD!`$sd1xpwLpzguZA2y@$D1ew|*2DSu$fdC@;dwcKo zJEl|A)>Kechjt|ROueloBo6{eBVSUTC-ps+1{1su(3NZ?tyLu*(Gc;a_xowh<^UB* z1cX|xIi7ZK$iWyWPcf#bjWEe!u%i_79?+v~jLqg*85#W@!sV7{hZ7L-hSJ9fzyV)O zKtMLyMWJ2sVK4&yL__AEJAuv5Sex*LYJ&N+=AYW z$G1S_M59tBoV|t#Ud68EQEcvpLAMmL)FjQZa9+6o2NFW?pw=AQ{T=Y~5ZVfFx@~?5 z;Ft9JLb(KsIF}f@+~<*?=*vARTKfhYbaMq3Iww`L+;X@Is|9+RhBHd7 z&zSK8Lm2Myf)xi=-lU_Ujp}IQEFlG3P~0)P=)}HoQ658|uJdD7^E}h=JN|?|cT9xy zLorN1I)!U+<3ix&c%^-;=>jy2Mlw|c)GY-qq;}>K5L~iif?sbNXGW1eWoMnwX~6F0 zEF)=J=8fIl$tY`1;rTo?^IUVd(vbyE7cQfG=#UDGkzbMtiC8%(6gLyD4Aa>R+*#DS z5}at75POp_M6!7Qf=hkgOY`Fy5SjM)sU|B`+;vN%Vl4P*2b1YR{u#Rg* zjL)M~6lT(##AWm@plfh(;rtJ7aBf)6+uZ)+kFS8Rw18V=d2n0qQe{I4%T0dlG+q*$4 zAj9lueD}n_d1;Ed!&liR{$2LVzrbVaOl1tthqoA)mH;L>gz56o8j^B}|DPCqh>aYN z)o3Umo;p+5Dv1RT`hE&}|9jE}!=u1=lVJuTGhw|8kO9y?eRsZQud5T416q9Bq0I< z5Kof8n1lTV*y?0H74F)0tKY2}<8PiyiSC)i5E)KQD7Ur5#250WiZ`RQ6jl%xU!`JK zM`9x88ygrPE(#xulM2woiI#(tE4=%#HB~U;GlOay35~U;;&< z*mc$}KOSgQ2U6?_7UHn^aBIktM73fy4+^#xh=ea8k7%zX^gUpf4qXcWG^sqMDm1=c z{I|hX?jkw!LS=z!K><;-nMe z^&*`GXkdVETGx>~v1uUerNhLzm;YMPVt+V6Ip}rTNTnH1PE9qSyc$vt&fxdhGg3pH z1=+72S*1YAII?fw&*Tsa^c8bFS6ztrINV@V{AYb+IATgC7l_fD#NV!cM~WC`#vBmHX*$IkusF!) z)mNb%MEur}^RkAFUW*5Ds**212Zr2IAA6QzULAfQa?ApzHa|@6bmnKGD;Ey}u(-H5 z2#R_A=1npP1ZjpF{s=`7tp+?t*`!Huz!n^TL?7PD4J$eZSj{mO!PH;z{)o^%CBZzf#Q$S`Gds*PDS%p$a>K>PNp&?Hc2I=5%`&D9XA(qq((DWqd$u~He8@FxJOh9sO)15VB;r7oc7;Zyvm9U}-qsZ)_(c2-2H8-O z5lnZppsUu+*7hNprE<^)Bxr@imW~HOvvtP@WB-Md$v~j~ zWbnu%fNPL{c7F5Cy|S{hCi{I@I1*1t!VRK=pA08#oBpYNmwkT6G$`DE@ch4*86_OM zv48v5U)MpvjvO<;K^6|3`1j|#(LEX(&FiTN*u=$y(=*PKYx!UnkAlVI)r(*zpq?;c z$7D!TmQo&QF~pE_3aFf!hYuVuMRzX@1UTM))22-X`4K-B2OnGpX5_>T1Wb8OggSuj z3OR!%jNi((7H;J)EM-Ma4eV^;-G z_Q`oHrtlcW*W66yabK9AXRgbhUg{qlBwIC^lnypF41wh$JkoG@M{figfngF}CDlVi zelI{TNv^?e=M)v~9P_+z;X)W%Z@?>kGg=>^&&0lI+s`_f?FX}E-yxjINkuY|R0x-@ zC>^qutv~_f2m%bQ#=ooDN|t-Szvvq~Hg0Zik~)Df1!9srNIe4xtw00&t~A z)2A{OlsqKTVNlZJVm2yd5|zvZ!yXh8mWLqMfRzhp*XWyr4_)7NESYQteDxUDv$~Xw zl4J~3Q>nfzSZF7L0i=SC4vs0v?Ee1y2XI1ZV}b1`9D_?e>uV;jPMs$(0~pn4a8(*G zrum`XDQX|cA91~8TY3tD;kbT~v~DJxK?=WzuL`mApMxYaj10>kb0i7ZA9m8{Iz(TN z#W6m(+mESdmBq!oy*-Lf(0OzsNBn|e=0|U=>iKo+p6PXvyRA4~MMxE)Os0gzmfpv5 z4NPIa2{d&U6NyIcLYqE5ZbspMrV}2YB7|m;|3H^s|Ih%%4Ve$ee+c3fA;CymW1}MG zr;>w?s^3SjhpB0|e7$AM1uzFt{SgO0`1(!J-XxPiE`P3$qFmP#kBId-;(&aF$yD&P zaHY^R(snF`(SW?r$UYoRd&Tp5I$$yeDWNc-C~qfL^%xifkfh9kTgV9tqM}vHCfjj7!a0hlprByo z-kxPJ0!TuM;BM*xJu(OzQlN=jJE6x5pX9YVtV>R+2t-M^!=jvH;5v1;|DWO0d))k16aJpus6I3u)j5&bS&B#Y{m@NLin<5^554NjlpRRSWJV;ozEVlwfn1Oi?cb%^ab0 zNCPE?5`QcPg~Nj8x1&#p6865t?fjmP>u9*^v*u*XNzDH_l%p4cj(+LUa4(4euSzgb zH#ahPe#c)vK0Xk+?&B5)@#t$NIiq{h;Poe=OvFV9@5W2{BoaVr8c>KsyUjE*UEC5rqeT5ue?`o2S#K zFzf_5lbk03i-|l@s^2@KprZ7ufEbXjf~oEAjpXvsxUxWIAmvATKpOaz5OGt4G>$5Y z0NhIf6&H!jK<*+Zospb*!r*x#{+W#7N>X)jLd3uedl4#54$%bzXvCe&!f5%jr>7@L z6cv^@Y;w>vN1i&=lyeo!hyOy=C#qbHd@Lj zr3kdxZEoZO@>2GLN2t=~VE9DS#yKA24Yd6vdlEPez*1=2`#Y+9NF3u2pDaz3Eckd+ zs5d@FARV3a9*_<#0G)b-sRO1;zyKsMU`mW2n5;#bOhOp(S-nyM2+8Y6bV5%SN6oaA z*(1dJ!gdkK2@Ht1yGTxe^&$x^BWxmBSiK2&Ai$BJE6`w^>frULB^U>9pu!!2+W;rH zT!34U9M6E2_;9Wm@)MpPakP=}A{GXx9<`U=Dq zh`YrEz+DM1&%k4vDj4@WsNE7_hX=iK78f zqKG^V#X|=Fb1-vk!#$Y+#F8m~3_#z9hlRjiwCizzoGM%{Q~lK`V0uLE#&yH`;f&~}|3oVb*Pt<+X#(UcfNc(9w8 ztt_ryw|@WAEvs2_G92)CGl>US zbjYz3r~^c1hHvJ3IAQUx6s#tpuaBCIFzk4p-=;g2U<^be872a^jfDh_=^AZp=OB~> zJXoa%=e{Kqc+jB~(7OcNi#@8F%Q{F*H^flAV;m+P;Hv0@_I^NHs1oV2E0L7$dc4r` z4dDVPgt_|9wxVZyw%nhcFf()u--0az6=PMUk?WLk1=*u|^pvva&&f)Ykc2N?JW|v} z0c|L(v=Ss99L)q*IX6z>Ag3na0pjSs0PL4oyrsY>(BFstUMGG&Cnu+Zq~8$lgSYtZoCT0V$v!4Q;Le+xYhQ>JKKzT>f!+;X4G) z(ERqU9stl#Mh59R|1DjJ0XgrYdVn0Dg8oY!jwnfF{T*cF^>m`gNnRlU44TUr z*G4)Sa{dTdEhkP%=!*OB8vGxd?LU-_{Ot7=2=`(HF8~fuC>(Sb2@W6_iinX^=})06 z`c=_?94h|$C*XII#BkaUF_5Ebo$eF5y!m$=!~<_i27pyE0*%;$$N`s>;7^AsJO8By zi*j?<=}*`6AOG#FrFDIJ?SK4=&Ti8`82CSa_kZhD@#;!EBikD600Ph?6@WO3iUcD+ zE!V$2n=uO7gizIj4iK>cxw)ogaZqjJm?FFyq||k4f7*Kg_BOh#7tn&0QGzWa8~1CP zd>)>kz3{CX1`g0V{qi_Q9=l}j3{=d6DK0jXl zpL^2 zP++b_z)nn@ladNJfDBVJ0DKM)oCZRQ54K zmL$d&St61oq&;R1BUu`Wj7p*?TS5z33@s)Sm7>*@q@q+(Np;dW|9^dpZPkUqcp z3eN*D%LOY}uH5jd)IP42;-5|JpWk_FxwteRJ2pU<#=_v2A3o1sCZ)}m7PRqb!5KN8 z<;GtJKtO9QT08OYV*AdUefspdyCGzz8}-|gTCo@mqygh!`syOYq&VJm>8``yu7ve< zh6##y)YaBr4;J_I=Toh7d*?Wo)tvmK!1(=v!|#y-=Qoa{x7l<-NDb{r+11Y0x#K1d z_}l1GigV@=-HJDFe%jOOwI*LbL&gkPW;Z6*^ImMYAKzp6??3tD|K02{Zq8HU@tzHDFm!`}u>1%>}}kL};ROoe9?$81O)uw=;5R3C%~H*fk4vBu{>Bk1Ys z;FnBBl@2ZO6-Gsnd@-06R}Yvd*0pzH*-PfaH(UR#y>R{d^~bB3=hf|^ngR-;5x>9d z=s|ak;oaO{9E0q;zHHpiR<0g5|4_kxy!=#?L1e%UcYo>yU1CMSqoDk^hkK3h!*q}E zo4fWT?{ys;e)AvCeBZ#*p9AB^hqH8O-%PNGY1mKbX18nG*755_9ai97C*efSOeX^SM1hroQRVRd;FizpYgA{h2|JyCBkhO@)(tKg| zw0DN6l#job$qt87Kh!yE?u{3%x7>EHcBn%2u)hxt3JRr@kp9oOnhjX0`CSWz;mZHh zlZi`OD8hlP1&Ru{X8%v?LhwEF^xfOH+vfLGP#D$BlJd{jIC&|HkqXr!Em>@A>_A9A zSY`2mr^72Nt++B9Qc9pqX}j3 z23mj-v&`lg87)aywdJ}Ab4X=;0QBiwnJo}J5r6aH_g=*Oy8uQ_KmzHT0Z;n&>4TnQ zh77G!lffxH8sa&Q0U%7-7+}~#=i)CtJRem|{ODct^e>^;5ie1}B~V4K4PGpoX<^N1 zE)(C>0JWRZ8VjLHBX9&SMxo?nzMJ20+>IZe{N}jG65jzhAQ1N3=Bn7X?PFGVei)xUfHSG< zCT8wn73=Awfq#!C27}4Yn4gbj7;9FfQ8#t!R4!OZE&afunuoz=a!l!_lPBOKMDRs} z(nYlldISAOF;sv)5Yg)ghYx|^7UwoLgb6If8|zj&KV-vq;KPjuFj0NAtL?7dj`Z=i zp8%h3N6*zZzO=zHQ@KZvi%1j~kRKm1#{8_WuIh@4imA*a zMIVal8)R-k3vuo{Hyr?KaFgNiH|Qq;Gc?C z#EMynn?bRlV=}2IhDrRc%IxYA&o4TkWa(uDGw^z8e-5f09?E}Mffw<+Px5h4-bp(%M9uo|Bv7HUC-DaN*?v?Jx4a*S$Z6ibKrg()n$NcOO|B ziG`J(=mAqMOg-Vv4|{yW%8U&)EG52A|NbRV4v$EJqa62BtRnq8$v}v9>X1mTc@cp^ zwnAVe@wfwL{`SvQ$|l7 z@4sBWV-QAT=-qd4?6}ncR2xtiAT9rLkLv^ksaWllJ{)iseM$^}dxkt+HMel>UKC{khM8A6AL;gy(lj_#bJk zx%(A!!{16I2h86_@Z7y;PYLi=Is{-}T|v^>yM+g@EFZ&EFMoG-lUzI_WRI9RfRYnK zUqW99fJz|0;`bv&g18~O$$+F8bgL{0?{P56mco4=0X3DQc!>u+qrF}J9`!Hah0kx} zIkt}>jm?$Ia2XLz8+w5Ou(9B{V;yQPzj)w%p3nvUZ^*GcEP|`~MnUIM?uKY&BDh|D zykiEYgEGL=Mb)gsTp&auF-iTnh9#6x721E!lpx*s0Q&(Mi|4L*N3kbPBPKap8cp#k z)*{EXx1PBZWQyD{BiMM4UuI!8K$lpH;}S8D5qGg{C5jnr#Z~Sb%K4*xgaYA{%})Fd z$xhYlOwH_CfyDQiZQH{+5IK~aA5i0ZLS&$^5Vq*olP5Hk=33+4oXY!vm7O>m1X?Q+ zfknH=G$K`ZzMSN@w+h%1twOtk5h-MW$2^1Qk=rF8s7x)0&xly%QkYxekRYb%_~t=` zU3I5SE_!z^ABkBCAh+*yRofWf+D5FoM4J)#Ww{L(e`w%MC+rBA{h`>^k6@@fkX^XL zW!qMP+XB+6Sa#`AQ zXe>N!knc{0zzK&!#*>?+Im$Z(xupnO~KG??YP#5|Zo? zHatcVEr@OM>wa$$#}xCJOnyAr*4BFA@-Qq-(V|wdmzw&7iD(C!=4r2&dbJU=4`gMG z;3fY4DHvrC->`m8%mI0poMfwsFh}%z%m$8jD!c(|uoVN)DX=Mzi5~Ogtr>nvH?mhx zyzbq_B|wEO{T@)UX8imLu8V23qdhs(t$ElX2NJOx+p1< z_o7;#E;B>b;V^XsLs(HH6GaC7jDD?O%h4X=le_MO?KJ$*=KmOPXKPy|^P?o~<3|dR zJ~&D}Qq(Z?fID0WnLw^3>OQvJ&s>2tN62_+h?xhZ4Qn_0q-!_Lp^4imcNU;~O4+b2 z6$P=|i>x7_5NOR)E;MYKOKov&!{b4SzvdDk-TXw;O|sZr)2p@HTYzirh*J@VBnESQ z<*K^3BVf=CsF#mFk@!peI%B^KqF2WMDW4~WD6Mc4`Q`5c2wS4EmVI3JeS}w;m}TysSoriAX?^b1NKPZ9qe<}-_V3dVo3gLJuSJMLe2dTs<~)p_pIB(kHsTPOyP9y{cunggriL`JEnY!@or^6tY%R zESrXMC{JU-wdTDIt&6DWO+1d6(|v60dgE%Q$kK_cv&(X+r8P2vhNTY6t|yOdOgY(r zHExP;S2WB)XzR7N&DOI9{UfDM}0UPtY4kS)!|Kh|20)ZMg%B9Y1c{7hWlJ zTpB(FB(${~V(oQBL&3C<`;FI>*hr3kHZnG0L+erJ1!l^vEsSh*e=v&7k95UvQ`yt2 z{Zr=DT8q^IJ}AFkFYDO3^UF8IB`8O9EA0>YNrB_rUBonm13H9yr~6-ArM&orCi|o8 z`i1r+W#gpU=ZD*ACXu}|e>;=y)nt9UFlhUp`($nPg=^6- zJ@(b+P8YeeZqLDAix^@(|NQy$RCDkevP)v%GGT+RM#5eeCMn@reD|Kpn%Y|N z5HdM(IL2q;((;r%QvWiAjI^Ht>TR33y7%{V) zbN|ti;IaVPvhxMuI+Ui_%YQ(c*kfWpRK2F5z$crmbMM~0tC0ztxy@$cqJ#GWS%;%0 zbugR2w|ezibqYUHat}h^mJ;$@K8hdVxR_0R{0Vbh$rE1Wlq1I1a*WitLoNloO^IR1 z;`Q5gHq;iTThK-fObIzwTWh7^L2*kVkQ{@)p)@s%4|jOa^o*m0ob$EsZ5|j`{(U$7 zbV$E`as3uDG<0^0cFI(5<*b`-;p1v*YSx%G*e{xH{cd#OuIhx@9Krj^AEp(0O!zXx z-hPq|-A>BJVHhb;uWMv!q9Onco~^vKuvmPSc$_9S{P5wF7_a&!4-X%o6~~`YKaX-8 zUl?5bw~)`yjRLTFE~divN+@_`S^5IiU}^g+S&`>YGmCgHaQp*R4=k-$mh89pT%A%m z+@amp;}kd2js|9OZpA@r+b~AFug*+R5;c;^ zO-9;;N1S~xX~SLZTq!I}`o2ZKy`Cb3461iP>`M2WKs@Oa=B(LXq&a_*3QplM#P}!C3K0OGkJ>5T-D6%b(9pQ z6YEg5qIX*ynr;`=^l*rZB$%{)9PWx#g0YR13r|X4Rm%lci7m&n!xV`jZ6ke=23=T#Uv%GuJ(u@W%~`^_SGC6E^XX={0ZfmS3(G;dkBF1_I8O( zx;ycIIIe>cXCgUEi9s3iYV!%iAbYw!C|wi0IBWTo(7Wf-Yr$a7K5a*D#1$I@b(u8!NiMiG z`9RNq&uwgX0Sk@{n5Q-eQy;5M=i(wouxMgW+XqtHuJCEN<>@)Ea!UGN`RJC`KbKPB z**~ktq>D1)x9-s&D)zxD6+Il$t5+}6&>rR=vXSB;8|Xgv#j3}|Lyb&vz-I2Sdjh_E z<+nzk1X9fCH-<4`%>{W~*;*Qz2T$y)-gm*yTi^sz!HIs`*FCPZUo>eLswl)-76}n1 z>xmo3UyJwxaSKoENPEUT6Ea_FyLqw+5R!J{cGwcJC-B(tZ?oSI^3KNi9LMQPmUuh1 zjB52y6NRF$XDSa|xN&ETPJV%%jr$I8>G+S)p-{an$NvfA{JUWcJc2z1-yU^642t?GC!T4%#)Jvz>zpllz0_l74y8Dz76)I%?+F7JAbazj zw!4?+A6aCymw?bGbxeKrp^I}Bnq6{!@(X->M{R$yYR|5Wz8=fkvo&U!HiV=bb)yke ze;d_G)n?`G8tT0L);YHQ{jUSCF7}=Krs3jE`d{TZnggK3m)_|KfWYetQzMjifLR+@pdzPrk);8hq~%-&FEq3-GQ0Hf(X$d~oi z%zAFUKSBn8BFfPXon~7FBb<$p3m5ZHJ`O!pKXWE+d0xQjgEIpVkPcu#YTgAa#xwLU z)Edgwg?V4@sQGq^OldlcB6eUR1jIL9_NCVqL(t;%tIt|fqCC0Twe161%8HF0VNF1K z#Ao4R-=*JIG9217eYy4WLvz(875N=LMVwa*p-7xS@%h~|udJ*rn`hs*%0tmPB>6=2 zTKv!Ewyp;k9NTWf+lR-rELTTurOHxS=Vduw zt;h0($>v9IeBZLww2%#t+iyS_Z%NlV;gF}N;j;*e*?vgqRndyh4p_1Im(O*FO;}kn zYgkEhErgRuom4=Jw?$MW47LrS35#PEs|U;AN;r!fdh=C#`nv;w0`aJ2^S^JU>daSb zsHr8VNG+^dPM3ZrVjYV<*0D z7dww!Gs6K*6N^cJk-o-l%Bkm4k&i&g|B_!cWa+g5K}^-vtcS;7I!pBt(z5`{ZuX(t z89nc=GLY0}1T&ZexzeONE`A%9IPNmo%F&I-v~)E#kG3J`l*RySbpYwzFCT`7?~Fa{ zUW&I^kT#h$9>FNZv5{+M&gRx|BkqDpj2iE&Ar~{!YPM$$O^jZ1==Y%wi|@6kd;4*( zE`OVRPgNEcS$}TE#ZLDq6nZkpiO z(HM?2)|m3T)@Twaw@Z%0KoRr_!N*vMlD4+CQhCa~u~2;-uy5N-lQWfsu^P?wUOvcF z6BCm)YuCndqVmCf4njc0v+a7IncGLM)tMT@>oPdnGjPWCr)xbrkGI*3waKucQ0md} zm#6mSv5fnu;m66EaORv*2E?&gd9*rvnO)IB3hI41jcJrc&ppoTU-`$wW~EKSL&Ci( zUuQG>*zjIKp@gV#klj2?s-M2f0(FZegS75A+0~h0=LRJ_yJp`0^(t`mK72*TT#9^` z(V~#vWAq8evc-^mc|g5W33J0;WSO^>;hLxT90|_0&o8IW4BY(skWazcXsAR_!vn}W%WP* z{B!0qg$e5F>YHv9oq5Syo8kzfIfAOdq3{*aePFAHFe^RTz%Il&jiQW8H%q%pN$ikf zsLSrmLwfGQi|7-vl3?A@68Juzk%$#PW}8o7F!%t3Rt%R&rSF=bG^ysb9fp@;Rx4lIqV z&Iu4?>cWbLNmqy_VGG_3)@^#}oREXyZ>ulL@s0RbyWf*?TsWs-#s}VRlz9d6v&@TGhGCH## zv^j>AR}mO1c*iA=5P30-pQvhPZT?oatY>5xhwl*&a}0MLvB;Un_Pv#lJ>`Tf0nH&AI94Cf@={q@(qyoV{??bR~IuFh{Sy;{dYInl62eC?d(aO^w<-wA2Xx?~0Cizlv#0=m%Y7P95+i5-9+1}Rnn_r)uxEQ(u)L=FHHTP9fiMuxsG;Y-g!IbszPyV3*A2c9SRB+VGKLk{I*-A{RYs^VW&5p+?GXA zrI=6U(9qB5nnJ-5g=P|I1^4-jmRagB+dsx)8BSy``Q%-;|D&()tOv+b{ZofVytzHH z_@B;SH7}2L8fcPZc+~?_y(l+EP;nj7ugflTQXwA={^h~RjrCkMd;rO1;Q>Bjoy1Ch z@V`zr6Gl!RSeM_kv9s*%tnHcR@V5ICbLuD=bf}r~EOyixPMOfVcW(){2Cv$Oo>N980#?xK5`Qim5s2ZDYz7BQUI82u$F5^l5H(Et+ zVg`RdbzMgr#+fGIgH#$IOyByx z=wTV7R89h%K)bZ!*#txm!>SbC^9%d!)C&w`$A5vfLA!}H&(wnEW$b7*a@45LAJ!)R z3zRh#)ebAXQ-@~#{ ze=HD>suV8t1>+w84-5akB`~2yz-SmD8e_@?_;%BnO3t)I{cI%B+R5$7+opyHk zbT!9F^p^TF&3y9D71jzf^>$y`)chqY?5y~vc~;=0eY#Z@6jt{BY}+T_sP;)V;rkAJ zveF6)&j0$$_l;@a{2VAKe69AE?|U-i|9>cj|GkwO5ZOG2?OYM^{pfo){+)Z~_mr~J z+)cT8<@o>hYdQV)lU=s+hqde0#WM@pR3?+Xbf}y)aQ16A{>`oR7(2~08?r2XI7jT? za@$+)apT4@D5ET?iizOqv{NqG`6QQoL?-*ClL8Lz3O2c)I$%1ZauS=Q37bkn+ZQ?4 zi6o=U+}kP+Uw>3-o#jgr0maF|Y1Mu6WDUyL1@F|=8AmsklRAP)rNv04K;5TOIGv_= zmdZx@nWPAyNSmmR)J9CdjG3vMFflg=N=A%JUyX^(3E#M>t$P zL9W^jbeikbFs>R;j#xfI@BaP!%lAhU=}&Pw{X>UcuKFaVyC33H#{(5t#7U>j(E4yS zH6~X&VT!jO{&1||Q5u@m&9dP$L4H}>rLUPan z9SGI3rT~@>MH<}*7OJ*knR88J?qQl6Ej=)TDfNz-SxJ7Pwf=tw3SGJMT(PtR`BW{H^sPIW);GL!0zJv&~qhYd6N zm2G{df9TL>$4lQmO-&?o(}VnTU>3SlEjV_!&ay1WAjD#XgQ}7gQ6jwJ(&xc6$12d9 zGfaa%%;t(mrr$aX1|;rSp8UkA;#6!&%@QZQQcs6B-yEt|SEn33NTZ=~aBOivQ)AQO zPNCAIdTVcLRBqe;T6S({+|fv> z7Awni)_eQX{XNEo-tuXq)Yxdh1&lq8_SPdg_mO&S;iSecjCWb@?>s^+_l@?O61C=u zPygWf4{z=~3bV*f{O6$^rbt<&!HVHqRFo!t4uVwwHoWq|xZawYn*O5&BBB}CN31a~ z&jI?BMr>xui*+<8k0|AijkVg~TPTJemJh#a6$>Q1;^lRg*jY{;46pBk!e# zjyJT3{O;xT6-I z{XbL}cx(;2{^GdS>uV`vy1l?;d~J8Tm%fT2bH05Zh~1xUaW$3D=ueuw|^C^`&%`bTP&rWF*3aokks@SIsiny1&Me`FR=gtOIZ$e^mN2u zdA!*&KgifvPD5e2Q0EHK0D$gZsqwEO3(v=+>kdHkoxP)!b55j9?X!EIjr zye^LbJL^U@*V`SKR9`gLGy5$4ief&mD2}kV=%o`@BR_Apv!T=SvW)BE=?%U7Q|I9; zCmYn<E|{PxROsLpV+PEmb<;0 zz6rcS3C8oFV0~zf^3WBz_OQ+hepu47VE)og=YlSOmT&}QWh#)Tn%Zi(L$Qe*S^D3h zr5PhHn=G-OY&&YR?GkGy5?f0S0n<8!;G@I4A?+K(zIpm&no}c!hOPS%T?85tECtXx ztY_5VFUz9Je%+NJ$Td2#t)#~o(6??PXV5Uu(8#T&R(1slj6+LbxdHgQT9<+C8G4PQ*k4aOHpD1GD@f$hP5X`=wgm~rR zaj&n!{JElhu?rFl4(d%D435s+6P- zO7p;Z1QR(;1@QR;dn7^meuwpD+8z3Z*wBsHX@{t-JX1%3IEJifXxwpW+1A9-mUr4} zb~slkyPGfZ(wxXNWd^sHq6g=R2YLdmCN^wzt|{8xR%&_azk||X z%xuVBn&W~_`sjA)(q#l&$!}e|=(T+t#^n!lxB4+v@ny3VR4{`yY|P;?{Nh+UkdA>0 zI2oo-*VEGzod{kChAV5vtDPb(%P$Pj|E)3pKGj1vh=J!O)!Ebvc|UgBG+=>z5oV|Y z9rbtok>bX}vJ^!7=vNbi=vl-O%kT>l4eQoVVnqO;3|EeV+mTdLJ%_EL?bL1UV~t}d zVwmOp&b+f_aZVe-B-aKnj=)L*2;ym!G2uu1xUlp16EWlDRXbi3p4tELyxoNsT z487*m=IzGG?zQ(;G_mif@1a|1dS*zW+(^LVOTJ^AS;t?q#Sta}F)Bb!wT)rHpW zDzm<&!B3dBFVb2P`$GqMqwe}C_%3{S@5|61IdS7xEkhU<{_4uT)**2)bU(jwbauAk-L#H~=2ZssI{KsF>OzxC@Y}^n;$TGOxu!{(N1eC?f#gO|NQ{E z$yty~&noI`cOOOWI4@v==JI@?*#W5^oSivBO~eObFc?JXZbE{~t5>fa<~{)(bXAVy za_U%<^c;-^*+hgHvt|w6<|;U+gpQkKQJ#zJNu*Lq>^3^D1(*{F_QuL=&oJF>1Z_L9 z{OqLqfxTvbcrdQ4gUz{j@&~zJ9k#i{dT{mT?$-A`uB~X$-0!SK zI64hbxqInpHm@AGiL!%Qq-QT6Navl}H#DJ}Xw-vItI+zCniNieca(pNuw1${!=&X0 zsRYS;{Za+!PAr}G6GDxt6dbF`AWBMypeBWVBT-NCuNB!p3j{uwFYG=uI?f0K=bps7fINvV2&7 z97t)KYztjtnuVL5@g!h(muqPRXZNor`_R4J2Ytii<#<^=>0ur0t*@T&sYQU&>*L&GpY&Di#*QnF{eX1{WKgywo#l<4R~&&&X(YheVW!avENug%b4$b$ zemwU;qyHSA6;or_RLlKBD(sisd2wPDK_idGZ$^N*-PjLj@HX#(=Tf7GK5b;KNudMU ze%%lXVU=3@Ng&a?LRwcw;=pd1P~9S0wZ$h1Y3j6T(}L3jP_`*J01gH96>W-7Y)jS^RNOuNi5H#C(Ys@J3xdDP za@C3WGgq!&)x5dgNw{;2Lv*HGb`YurY{SnU3vajKIaB}wqz_vi*0*4ODn9-o5H^P0x z#LK==%WIr;9!bgQ-=gNsot6tPdBTvPqc=VoE-^7VYtEhLLT`iY7apMT=^_PfHwu+$v}y)RZ*pWI6b|WUrbSoZ z3e{;I`**4*l)0B8KUdjsfCm!Jzs=~v%Xiw%>MXwN!Uj&cIPXOFZ4bXWboA&^eR~?) zHjxidqYG+g`g(zR*ON}tVFC;7<&{jV8-gp@Pn+vS!LrTOe{D*Vh_|M1-@cvoU^4Op zsDAdV8=h!i;sw~NjJtmQx}tCWoVinx{;o?s_uXLs5Ki;))O&Q9J?H~wU7xfw?~{6G z0xI&Dw96)G9!;?OaWoAbm5B&?fy;1g&y7#SaCHd02KQ z{z^B%-);&Qj}X)I=o@AQaL?d)%HOY~c#;?XTgI918`ZneaNf504|S@tq3zC7pTO)Z z_fc|5)1w%&N7ZhC|8;`9D-!8g#w*%HTKq-)^D@NA6iry{@fF>7yif6IAIG!oC2+)n zxz0bn11u<(_SZvy95n(2Q||&*c-(wOUgkDEArbq?HnqAX>%5O_+hD(BEXVs422hG2 z0_jbp2y%ZB4O(LfLDw&pu&njTR zJ{GZ~mIW-GB|S-e%P+_Oq@N42nRmO*up!%edj2z@5cEly$0=Vjbm~kPJ^#?m;|z>$ zI6*~ugRG>mJhW#wpI#^V?>&8R?4T`MwzMynt;JAn_}v02W~7P^$llRdFRFpQsTfTg zzEs!&KAOd!%|iknk--@f*^ed1qMvt8lr{pl{fu!Bqt=#?42qMuKXE%zZ%V0 z&{5Jfd)bBA;&@@)JCEM5z<~}lm~3G}&w6ORFlQ|kGrgMyh{-HL-$XV;1E*ULbNPR32Z2|@C82)*s~UK;W6vM)qCp<Pw!#1KrphPC`$S8^0;pKc=>1Lb%p9Lmx`AD3>;o`XW-2bOoBWXVXVG# z4U7${2aiSfp#1^f<&_&OJxzHj`B&xo9#KIwygfTfhj1H}P6MO_w>+fcy-*G^7f%1} zRYT^hEXK(2+bn8TIqaQm&3FEw^;t8$)7kIkq9 z5qBkY#Dc&rzhEWg396X^XWRnf0 zd-kzP8faZ-&Xz@R-DghqpPZ=y?&mMv7$g-r$7Q;ozkf#?kz?o4zB&uaPhyy-W-l`c z03JjW1? z!Fu4&@sLe~pA}OX<=7Xw0|rFk&RQc-uar|LhGKJ0YjAN&FJb;`fDT2Z^$6xaF(;Qd z8UE~tHce?1VBh3wZtjAP+h#g&kn_lxj=|}!&QEyv`LgpwqRMoF5MH@?UKC>@vRTIa z8o}ui^x?yAU+x%i$Z6+_uxrHQtSg&Z42WokN)Bv1^@oj^!bY8zW=E}Bxofuc@UTKH z>bH_SDO87WP(Zdp+p1xP$vebJCe zhboF0Ap`0EMI95Ix#^}39oVc$SGUWg)Fzl1`G=5!0HDH#t(SyMZNzsiN>e{-Mr9z< zA32nQDoA4|tP;n9z4J;5$Y*Ftr&;$Un2}c2N=OjUnQxP7Tc|>rNZX!mmEK}*SSyzx zjUDTTlzd_Oeqnaok1NZaD~`~D$6ChsSafEB26t+qOffn$!QCC*p$>wDdp8c4FC@ZK z(o7V&&J*4}Iej5h!D;u*#>XQ*{)znaM&(a-sQqPLJ|{mb|LMZfX?wTBYj zP;4cUC-{+_U1#7_23eqB$6|> zQ~-6zE3nUC@~=fz6rH%Tl(-IL8#1qpsL-Tf7Zy@DgVYO4ww|y4?`06bp9p_l1;=hO z%P6G}Q+*>TW)?kj6fv6dLG6A+Z@fGuwO#1yJ)W9;l?|0jK7{XnMD~LTXTh=BqQeCV zj6pBgjXHUKVAWAzEfN2U7ehI-r37}U9%1q<3HPyz{k^_2H8u55Q*M4JH>#Tw8?LX- z{zssU*JUBIwnQ@;N04hhx|&LUAKLk|_w`<%K%_XH=~$6QiheLSjXF0nk5+oW%$4a| zPqKyD&7-k#g&yd%!3dt!7~D9RSm?2SlHo||<@L~-Q$KVl29+sanBDV_Cz$ocqep>kC+|7AwXop;6ts35^M|TcF?WYo8)o8SYuBE8d zyJ11nctrhfQx;jhI=+YXkk_GY@XfFbhO&znkq7L5lMomLt+MUUAl|)1KPKg!V7;*! z)nI=IX|smrn(^-F4fe?jigr>cGMKs^|y@yA}0YW}naAn%{ye^R%dmphmVN6~xm2h*$6Z^FjP<$^61-NP(=lRfvo9LN5}$hbcTMN;wVi^vTpxaw5X?8cAaJ)3N6?wrYRNk zCAgwlUf3^EVdI>R2{%3)Q$~M$fTXG(7{&372WC$=jLH|0EgB=#G+3cWeBC zZYqLO1y)7$o!;GYs;KFusgOT=3!0bv#x`c|*;8MlL15ayJ$YeMRCN|*SUG~4jyGIo ziWTXI&$iok=s6;I>E%obj~d8c6q(BJI*LnjlZy0TDmPUA*zQgLpWs2oUZ%K%<(-$N z{@KQkEnXI6L1i^*Vx(bEZ=lA}Ax>$I3!|wiC)37}ntsNY)hy@$v8(-StF6h&$$8e% zb<0IsFLtrKlZLM#IQ49uc&4T@Tll5r+{s}jea-IDA{N;N2($0wo`eQ1#zMlh8sXNY z{Tm$EKw~WCwa~@Qo>t-H1;vvd##~%CPxgUsRdoGp3#nfGuQjnBbhmovhEZz>6T7~asgcjftS<9ita|Fvw;oto>O7{5=un3!U%@ff->q*FBpS?L8GX3o}jZhTr9&x z31-cnVDdTHh(s#e9lz&O#hN2ZlZVAuv~fyOul#HvkD0){T_o;cco|N zXOq)Jp&f_7>V*CdnH31yQEzrAZOC9K?1V5Q-zgi)=1X|Y<0#EO$BkSU?@mDBp$;bC z*@QF*-S_#sW}hQaI%*a=9;28ZGRVOjkhs;Ii7BD}-P1&&2IteNoXUL~eHlbUe{j35 z7|t9NQ=Y{^f;~uME$LOT0`~MYev`3$L}3tRW^}bZmpZD+Wg?=t7jhmLS!}B%O|J(u z6tinla<8r{SVuEoPG=U2i^ys&4ntd@N~=J{`lL75gvXW_ff2Q=uS^CsX+2;z<*5YB zk)Z{&VcJJU8z%8XCK07X7u1BAo2mS|)%>IZyp!$nkN1mHHdhpMr|Y}dZ3TGx2rgq_B!ER0jo%5U z_eG2GGD*&&%WC3-XgCK9KlvW6CYB2*ddl4}WZ#KJ5*NMrYEkm&kQ`k;Xn!o$t5sF4 zy^5?jo(^1@$y?_p$kmlo))|$_erD-N9V*>RIiX^K|h_z`KehHU%^!0sL zIZC_%#=m{|ttCQ6QC(Dc%jM|dw8eQ0sy)z5^`^QK00{-wBq6oRn(!ZyvReY$b||?j zVoK)l4$5@8$VZ31iQ?YE7_JfNA}Nn+G3>6IJwQDq&Gn~Ric=5E zYA=Y1AbX~z07V{et8WseM-WIN%KxM8%UW#_eGPs_?lcqi*+h8$#WD|Hq+kT{DkL;i`*s+tI!szP+k0!$B1nx$|tK ze{JoAOkqhynP{I-_H^1*k)-GteOxW>%)RuHknp~ygKGKwhgC9SDts`>`0!?P1dNu* zP~xT50#DcwC8~&as~kR35}teLsQ1664os+%MHTu&>(%8uGBCE6Phv5K-3zl)%@1^X6pp9#*ne~;h&a~nlcXRyLa{4WKoH)+pB2k^J z20m_jW>IZ%))domJ?J1E{4yjYpZ}QDyJ>Pzpc#<^=@eVZheSg&l}y1_PiuMtUEp%nVR*~kt%(`B=ux)3RIGN)W zeHj`6zww1a%L#*EF&>8YC4U$7^x_Xqpd0mZhjZ!FHCh~_y#LfVQxe5NQ;K0rqMklD*g}*I#kp}#5#s52`V_-?q*Wx+l~JndSW>T*h|fmg;v)#AJNP- zF4#CtERkVe(Mm{)61-Ogg%zcSL`AMIpCm$A(L;*OJ+rD%B3hKi`9R$VmRo0!CM9(d ztg&hM?w)#BpXVzW{_ZBN2|DwZhOf3Fl(Iw*(y!nodqG8z%56xNh$kInqD3$r85kB2 zTgnLA&(i)7Ap*;}@|OA+tV;#0lVfv=rxlsXG@`kvDg=&~{LE}s99%{R*`~Kei};OeV;2toGA_X8n{$ED);m9pYKs1`*S2TNnh zJr)*{Ta?P@Pb+3Lv1wXF#@#6^T;y9JEPDdkt(&s=JuUfuQN{HN ziIewFOchEs!UEx^u zowUBTNlj{K&eSv%IF(VeM7SzP80dDzt9qyHzxMTrnNXc|ctI|#mav-A$Lo-$z@`4%Hdq? z&Aj|rS!K1Ln<0XGlW=>8sooUP^-9klOZ#{}al*Xc?S@sVyb|CKrwRMXO>N10;|7{%y{3x>-EfYMyFQ~jWPC=nPRJ1yki;yX@ zS6)@U>l(F4ahuMnIoifEmxTS~2LDxPKge1n8a&7W6y4WuT`%<7swZg`_XB2mHxU;E z07XKkDfC8*mRJ85!c6M6M|8QnhHs$m%NLk|_uWZrn3_~iP4arg1Qi=7j9i3uk$K5y z5tz_4DdT~7w!Z=!@yyQmkgk)0T~}*Dy?_(q3Sj7Ek772COJxohgXmgPcg08%LX%zl zpdK(tF0*XPy-P*6h-~Qh!GvL^5Yh+0U7p*EIC38j315pdYFU+#yrN!Md}WQUkvSER zi%eYzpCTfHEHb*~BOR5FHOm;EQ<4Kc&ZIi)?N$*vKB8t|V+Q^ZHH$|scir)!_t40<$?jjSKj6Hc zg}ZHSM$u}?XWX>?0k7Qct*xl|-aPx|5yn73RCy+a{KG_1GE2jK-QiFR_uXy{-sK(% zW4}HAer})8<0~3kewa1dtW#NIFR`1D3e@j?XhqtgeM4^ykS5oXVa+q$CpLn*9;UyJ z-9Zo)kF;|g3!gsVj9=YBp|rwld$VwQmg@_aL&`<~Z$UVy{a%^O zHM>Oh^92-gubLZF7c#V~d9zHUoNdV^BiaFbklj zmd7bvw~S=s={Qt|1!VZacZ>i~$=WyJWR@7rh~xnTDodn(aI#xLkMu1FVaf^L^1{tVUpzs59%6uV`l`Uo}tm`_-he`{rAp#hQK$KkVF6 zJZwkpEQVjU@2g(BX3bP?+^LLwNAG%e+j?V~JRw^ZL+IgDY#NCj`5fsvO@15F52=?{ zT@`W_)mpXR=$W%>5J)S{DI390*8*$-aqM8$lDp?Wc~9Q9P2sJeB^3`ffG15``H{0Xj{k$5_d zW10;>XH61W-?Ra`t$J+Jmp5U!Hi%AJv*=kCV&~q%v|YS1>GEqJf;}?1XhGnLd}7vE zF=xUgNz*LZ9M1Bb_x@^mXA6a2d|_W<^Ue5XZF9m^oWf#vs~8;jnplv?LQ9RGixcKcv>8%9#V#M1Z3vemAm-s>@xVx^>w(5ZGWM>F7bosY^y_bQd4mj`Lutarc|){nl^ZM~s}{-$DHy$F>sDjLVInCK z%9%b*2SGzrms7yAefCEKp-rjI;-t9iooVzZk&apw0YN{d%rL{H!qFDuwK7f!xJdhI z_d=Z9r&E9%!@}BzBsvz-D|)=im78BhK!Lv9r-xwQv^XygyHcW%WX4AJFTsD0A8Y)gb^&X&y{IFYMr?OTr+q^-@e{}&7vG9 zKXzti4llX}$WS?n`|c>}{ASn}dMwFsyFD@8S=<)-?)TvCwU?u38d^pV6O9S0+R^O5 z4^dkgBzm~Rl&Ph?5(o6_>MsQ3UUy5Qf`ZV<8f0Ifas>mvdkdd-d$Jwe?pr$L+BAAX z!&f1dhNn-~CnS^mWoSUAi~Y>driuSR_>2;}(2qZn`#LKszEp(6i}&4>HliZZg4f5S zuBZMuYwMd!lh!_ znB~T_A+l47vW1-#{@E_+gAAnu-R$eWe%QH-wl+nig!XZ=Rg>>^*qPhuv$tCUs9bB) zu3ov~61m+8h3LG~PfCvZkp%@dIoz;npRx2pnb-O#GJ1c1a zvd>cv(ag2kKITI^G>{giic<%e!r()ZE?Z6(1XW#5$aA`s{64CFqnSfl=ynCi6y;~L zQqBz9E|9F>*?mV_crIcX|LY|bC6yOZz+G5ea>*!iY_8_>4%b#JkpV!N=^H`8;b$i8 zxHRg677Tcw9tyu(+b~i;Yn|c4t_4L;KlXi|Y}uh8_08?Y7>RzL4j<*wVPX=@6vI~W z6iZDq7nyMvvR8Ny;E-LDD2CM^=M0>%wl;FGy5T5=D0cp?^0aoDEi?NC1rFFFzD~cR zgu!X*>z%%-GdGu}gPL+Q7NRA7W>PRq`_Q-I%MuxumrXh4Vt7*M$Z{2)*bw5mE#1(_I*+K;zb-wKi^F~ zxT8c@#eAIG?R_WrlFl&j9pLiOZ~ET~ovIM&CD(u&G`aE+2vuwo1$e)fly(J#p$W}l zXi@h@8ZufN(1z2z)xJ>{-}uzht_`j@4j9FO7xH6Y#2**RH3cSj%{@J~FTPzv3RF+~ z<)=VVliub6%!J*?MFBMoxcK*nV^&c&1daNpBQn2&Dlq+0_Mp0>el)mKIYR9>e2I1l z!MX4Lu`NPBr>t7Qdpo>A#D>hkT%1u-PfDvjk?fs2(eRE;n)0u%%q{d=K2H2^;H_K) zbxFvP>JSXL&tsaz$7(@8=Vsx4|NCt4$E5GhiZisDius7GMA|gVDv90DZ$B`?@ro<} z^;kNDMae8H_eAt*QfGF98fF-Pjd2NC(`_4IEB~lSV%2hE8zCm7 z*AOS^1u_h??^+Hr;*L-yiXkAhH8|+6@T+p^4N)kE2>0$8H_~KslXKN1%MYn<^FFI- zmY6$N9+rR~tZSNQc3$Y32P?w59D4u*iMh^vSmUcN!h^ zris!GLA*W@#sBgED~gDGI^Q|xn1s0#KCi~ev$zwSsYdA1LgK8>>j;EKl*nJ?R*?3!=qdxx0T z6k4FK!meffBjFGR20MRf3qiT*dcje_Q1!+6Eso`nc0OIo^~%{EtJ~A(a+M$n(VPOI*$b7y&F41GiI*_#DCj z0Nhn-tR@8*({bVkCC&oEF5$m)Z|Kbq;ki6U56$dt^P9#nrCEsw;pwcheA_$;KiFG- zB}NJNy|hOO6)I4aF|#HBa7fX#D6z$U$-id70CS-uX9H+}3>;om$FGuhM~0bSf@3wY zR0LE>_)?h2fHJeZuA(5|kMrPd4$_qYNjWX`r6vGUeV5l~BVJV2%VVnQ?&R}{E>h)nsi)c|q=D6$damLHDCElTWfXwWGc&CGS>=IfB_Wd3U`RI1qfnlC!6xWevwCtW6A)35PF<_a@iY z>*v*X@?)e|(c)fubiQlc>_3pNq=8Kcyz}y|8NC08&$;q7+kV{Hrgg{B8(0*qN z8yBqsy^xWe8kX+xrFxMii%R3tA2;YW1)3*u-APR&xD2R9{E*~&lYJ3-M4Z#qbDoif zqT3ddMbIdImY!)gX(o=W*sZBZ?>6Sb-THY>a&ss=M=&7!)+P-y#zq}%FLJBZRA8D= zn_OeY<_Urah?-G%?T@HXOz%frF<81cv4Ik+quoZQIm-SIyXfaoh;3z4?(6mKk1@bG zKKg+Cb3~Ucq8cqwyi+`#KwyG-lPNUhmkZaHT$H7x}^+5v540}-_$ZpyF{{*=QeBTUdfoOsWFFL4EDsn76r%zdv5iuH-yAietH;d9NT zudRWVF`~)2s+SLLX39HW- zNy3de1FZj}8XNnWW49&=JWH1c(Ld}>;&{U*XAkN|i|Fy@r}!Zv0GrM>7t&Z~>1jak zav)=ikT=?YZaf7A_eEH=xl$y16swBiA^_`^1=TAvh3I(+TDBxCIPu*4vj?N+PX3`a zs>8s#6rY9;fJd7E+(Pv$g+l2(V)ay7!`p__T0f+XtzjNHkKTDms_ARdWJm^=0o&?+ ziXZ=G1%)RIwfP_^P!LU<4H?%y`Z$dyV+;_LWJeJLiw2|#OxEr1z2MceT4Vq&qbF%< zhAplP`;iVTmL1Pbr86Vtx6-AK9j;Hl#X62?Z`PcT7 zbEjEhy(wyMB_(-TOijnq_Vs33ly6ZH!+|c=;jc4jU-A7Zbrw(#pU@uSmO~Bvhe}fD z2Wj+6#(Z2mC@7qaQ9Rk$T6C1W*6-5HKMY!HVj`}x25IKa7v$E$d4Dm-Q#k3cxy2{$ zci+opC9+Hg+4=aqs6VdytfuM{9q`VPZzw!t;BFS4vqzDuzgR?N&3~UlLaV=6h5f~1 zEFlI9avWa57BnhMUiE*b{Pw4l01DJz9MzGf|3GtBT$tN&M?-db`X?p?PLx+ZYMS&L zf-1)0&ZnOv4plfOd?IVUg-bZ|kPX&=eqFY=P?&7{_Fk@%+|2!>w|l$h%UZLWGW2=9 zh5YH55FRC>vHo6ej~(jGCFIXkWL%8+$&-@JU^a*5AB46Y2A-5p|FPfMnck9HN2YbL zUQqGr_adJY&k5UnqT#=MTPl3@^yyRahl#MHQtr(#^B3a=_4w3q&|OnfC2C3W>S6je zj>)tp)e#Mb1T0XZ0tJN+on#I83K`%jdkzaBMIm)s!sV;U`I^n-HczHfhe@0h8kkBL z4|N&HBa3tvl@}hkIR0OF40!F2X6hoCx;!z_; zjVKn7A}R<7C{@aUiLn5}AOa&r#zG&8l%bdZZ%s^@P4YeGf1mH(@1A?(laoUPX7By( zx2&~(Wl`pi0O>S3P4wuEj{mJCK^?;{GCGb(@z>eE3hkfWWwPhft*XXa*Cs!V z{o2g&?9;`v)3h6Na&OxhIq!&EY#ri?%0K*t=lTZ>vZ6#pfqS2&a-o;|2WK zR)(t`PBbaexCKAtFM6QjZ?v|HIt;*rR$|0_qp$-ColR#l;+E^&BI%WwWvR^vt+#VWn(Fx0s7%ozfxr?C6IywdaagqdJF zi|>5lBr7X>1#`?AyQ1(5t7W6>r`?E%Sj6kE>UIDeYKihh>%Aec5jiLVT_ho4=s@*i;0oLua4oR? zvf&X512O7fIf{NzSo6nW2HycNssWsnk7|8}8b=ed9EgH;w`30w5xM4A!z*Ax`OA$xJZ{<;H_+S;S@2KG(_XM;dAv-6 z`#T%XlNTdKQ5~2TzY)rW96(*|r|mr#pvWE94NT3k{iLSnd_v8mV`7ZGy0%Hf`4oME z>`#T_sB}EH5Bzy{SI^sl@O%gx-klp|-2oYh}Q;peAVu04njD&RlSY zbHTr)rpL)m_cb~(hBfe`u+o|bv<%Wqf0QX@?r{PCkS6HyY{wl?s0IUP%B~vZ4RxiG z`I!Td*3|M!4L!+#1iUu$>L=%BLMcF~9BFD62RmsS*3LxH6=;#}MYibBg;h!_`woHG zSOdr{p9tJHykB^3ys3TgdbeB>P*ETt1yOu__G$n(o=yAb9lCo({xJ+bYd}qpGmV9> zKpN`1v&2HPy8L*4eznh?~AqOD-3y{11%9m!=o zM2GD@PcOj3x0dW3h>i+SVKGMR!$;!%G9^UrKI>V%K=4x|8W`ChYv7fkOca9NS$EbC zg6!?0{O+4cGErN=X>x!(MdsE@NaBb9o@iM)8=N_kJzkfP`*?0Gf7GGKKTJ+A1>IAr2hZ^@&kZH?4KSI96b%~*aHg*7o6wxg*4TWO4jNw&} zXP*a|oNEGayhUU`Z)7tRLD)9la@<&fbitYhC}~wRd6RR|(Q+_)+EF~Wd*;a)$kNF9 zW9rHM6DIsdE00v6%0BRFoXQg-EX^CN-?A1A?gYFv3()Awnub6h5uNBj}bN)eUH_L z4auyVy~K}uM0r#wLHdVDk;IT8%3mL6H-)ww-fZ zguKCGmB&x4nY-7K1QjH6-gZ;lFMMm1ZZ2TeL@eXv9#a5)v99RU{4Eg!Kw+?LW6zQX zsy3GR9f^iY;u9JO?k3+GvN*61RF&TbO@TP5R9)BA)4K_Hx&iu+puuj@tNS|Kx+Qdr zOeC4nwwvkY*ot_P9^~J?-WHGxnS;eNrCtN(USR;8XCHB#aKV3@1jqzdJ^v4BRC4dUWL&8}fZF$*2M{PG*z(Wx(4^ z#jHtc5z{1aftXzbe~!At|2e3{bw7u5MLN>ZD|HbyjRS%&E*82thtTNZzWCu>q6K4P zs-6c*okxuv$Gp;@lKf^q9ZN2)k4D$qXM3kc zMMvkNzti~zXV%Q0hqaangwGOut*tmz60vYdw?y-Xtjn;`D)pj?5Z5Xc#ENRzY(3$4 zcO_!_$a=)V5m7EfbndWh&iIWFgWp$|AP(A_3Tu!2%d4(eIqT`^0j=d(_uE?GWIkQ0 z;AA!A6Nl4Nj@UElWnRmkVX-&@QLe?Lc?vyVMSUO+nwtP{8{mvbgdWeDWo9|R$mk5I zgGSi)jyEJGY0wqZXr*@bh>l9(8%asuEO0kUGfd%lK;E>(50J{@LiD2B+Q)A7vZpiB zGxQOi4$s`TH}Q6##k-ZuuwMwkt7y8CJYaRw#q<_k%DRAwkpAN!Cdz8s(NNy!-tMhx zn6#Y4kEpC$m(CTv{+(+WiL;=U!d~jujKVBzn>1UJ>y<%l%Ow{-?WhakOSEwk(vq&9 z8hXP*TEKxpfjfl>&dDgfPQ~9l$0-;LMZYgYYvjQlRk|7oC9;Q$DV>1Guykp zLHoc;uudW$0Qs2KM+5glebHgOVri8Fj-dy`3dv2ZXL>uIOuK$!*t58}~w+atiW76h=~`tL-OUkIzoM zCe|9O(%|{vJm*yS6H*z$27zzew^Mu;LdK)z;8T{|4nf}yD=l%HQP@Q?ECgPu*h5S} zKPV2X!a=fn2V=7~j@vh;UcCq;s)6=P*-y#n8w8!=2|@P%*zA)b)@pksdx@V7j<{Zs zS3)Qzux+0xo`{e{?v@gs3>K(m)4Qo_Mo`kEAWNKgr-!Z`?ar)IGQTX^jg^(MM{Qgo zw?Ee)%b(_JZ)KhGC4?8UYkEERf74Ojx;O;cixLGxpil`rZBTp~2L0PiAp z!e#j`+EsTd35qKn@}#by8DLy`W3)|TzFduWfJ6@DSBcm;ZIc+ zBM%#?#>d0thw9yVeV**CZnzy1XY;`((R;X4gFMssVoJVTMK%0$bHP6hkf7uOui^su z$d*`>4Y&{(!~r--B49gd9^&$xTtkQ)C$J{^ACT4{I%d)|cTJ^TPrP^%BtxK~g_SvG z86&QOE~A4s$fF+f`d>UQ!qL5Y#~(TIxF=(~7?3gcL#28FaH%3Efk0mWIkkJ;s*U zadV|t)Os&S9G{|@US;K{%Zfl34yMqBMh|=yI-B#|hctcqbS9#^=N=H5Pgf1UnUQ@N zOth4rZ;1|nVFac5awK2OqpS@h&W-VDKOg!OfkFq8#C_K}YlsK}b1kwsCL468wV)kRrs)Z9*aBpmm#{6Z zpn#90nPlHE6L!NmQJ#m@Q9wts)mYPOAxO>8kLT#PN#2iT)ppsq@bHv%Ac}HeSdz7H zf5}=fJ*ANgf_NHb=p{NpVS7-_xR0?G?(@BmreqT(3^@Vi&P+9&fWvY>{`Ut#$TuB zlCZOuGGUrMWWSG%?+LcR0Bi~?sBx6Fwe>8oILz@fWH`bc3V94?i%Q=!$h=}sM~Xg$ zZ>i4LqdKw6-_Gw|9~yxa;;-Xl^TxKZ&2ITKCr0(`>Ho@Jz=?T5B1La>M~nur@rWrvaHS`+Z+nw4MdbFVfLO! zgl|4H(Ax`gVgeT7O@o1`4sL9 zmwoyoeR4q@q11umj(l3=c+V+n-HcneijU4!QuQcAX<-kvUv41RCgOyaP^UKcM5@fj zyv_i?%dWXvEqAQz~v~E;qY6B*zHxj&v zC4R^psiZ?yd362hHIQZ|Vy-Sy_X~fr(!1>xS|!9hWki>(q!pt&Pp`ph z)S0UGQh=~|*CR&s#4-wH^+B(_YQ8#xX@3bSM&zy)T~sX9m?^5f?`T_S@j&{Mm7y@$Dor+$dSuG@4NUV;=ZN zT?u|R%=V)&`8U~U3GIk881dar;TXZH=T7B*#;qhtVl2t0jVfq}khw9C+`cvs{H+($ zFW07zVW@QV0U&yG^`^T(D;(NJmkvE6um{&ih3<116+i_<@3F$>mN9M)wB}9 z=j@)f`#`0_~BHIN-^ z(K*JtZBX>Gtnwhg31RW6%3oG#SIt#i={KKZI~>24aJVJan=*d7d6zK$yQv+erozVw zAHD>X+Y?krKz-ONx91|`=WQeBYpU8eJm+x9K4B-c51x%Wp}D9}J(mOpO5{(H!L%8w zDhS$)Z)l#FIHw3PDIH`aQzOMR5s9(aTb~;e`7IVJAmYvFdI|{AwlKc)LK1~Q#^15Z zM{*DUi@|1L0fOw=(zaOngCH7Hxqxw#Nzn=83OV4fW1TvjZ9DTv@aWP9GkaK=lIok* z>(a0J#+sCksmIIM)$*xa^k<{n8K0G3VmG?JS+@FR$k8^&x1^#>J}5kv$(Y}{@lB&T zkxylexUC;tCaY${Iw4Pqv1PE-v5PT2oYj2v|20-fV2W?4u8z)m0u3N+G>LvHt6A^i z*g|`FI-_A*+=|bucE5t2zv1+U`N52+gq0bDOmGlBSvjZ%hK4U`JoYp9#1jPs)nsji z8`~bG8*QA5Dn|qYU~Hf{FudEW3S3^-EcLPa@bP}VK8_J zT?E&VMvjb2VAwQ)f~ATk=8lwm4eOaIv~a`HS?QTYISOo;3bF_CvJb$KPNVTTUMCqc z^gi|(cuYNfU>x`qXP>UC584c_MNudbKGFJa40Z~=rqT7oZS#hAQ6U&>3uS>hv~dNN zbff-FJFO$m2%L{o)y|Qf4#nRT*jCPOt-;_7aAXd{z0=@}pk0K9>+`GW|c zE68qXCOVb9{FqOkWYno{dIRHr3PbdCS=2Ah5E2J=IrIZYPG@1zQmZoDp+YCcw=HB3 z(a{(apHi>Pcqz>qY9AJc!tueKix3TY*T%qs0Yr{LNp=+6QL8k}gOCiop`c90g;f_u zqpXT9T#Y7^X0Db%_Ijwj1HE=F&U4S8j`9AaMiLw}uw3-Azx*Zy`TVR)+iu$6NrU$A z@HX7JsGC{0ngjwK z_GRaA7%NjZ5W}6Vq?LtuN+XW?d9e8?L(96wqZ%M{OA7U}KmVO>l*K8gpgO=JF4>T( z?Yf~WxlrAr;A&Yln+#3HCWD47(<`Dd!mheaXV<7s6j8iTE9Hb{=LjC0m=sTsMwQn^ ztpdnY5d+$>&zDkl0ZP{C-hVW+!|nQZ-Q0+Z&>oT$4-N8$m2!r!S`EPi z%Hcd8#kZeGBB#c zko;gTmZS6xJJ&f0LzrI@C`|1&F)ol)wx2ueDS=JMj8@}EWPXc^%r+pO+Nmi^{3N0# zgvL(__d%<_0>~38bL|<`A1za$$9o-}q3p-P_cYm^xx{aebBSJ%XKh4~T`raQ+(&f? z)_01j@cH8F;EzgJcWOc2@2^bwRnmfKuA|-kJs~opMAF!Gu&c1Aq8UZeD=0}&!vO-^ zo%!-zm3?7-S1d!WkoD-Ce#V5i5tCf)s|67`7Q~#^Foa zQMUL30=GQ6BK0eUFZi9}X|pIoQrcyP1bz8YOQsc#YitL@Od1bpovdP(2djTs+@ zW#fT_MH6%wUZ~Fg+`Z__?@h6SkVn=6m)Q!B8A3)Zkw(}2gOSYbjVVNgMpGMV;?>aZ z4WJ>LYnC9^-W-{o*iJRgqF1_|*ko%_vdE+Q3s!gyl`>3GWm?Ts5`ijgFf z*IVF9mf^I@$S;eK*U!Hn>e6|J3N3j}#sjvY%$r<>(8_Wtie<~*R){_zOZT|TOjTYX z^~E7O99nwh{7naNkm58YN)REBB~+i(j`VR^eI@a=I9H_}FJpezKV9ajn#DuDw-yyC zN^2|uzfq;-A_5Tk2Hj3L_$@Rran>W`o~nR^y+b~1`Y4rg1Y7w~bSFgy$~91}+YR-G z4zDFZmMqc+P@#DOjQom(^Wtmq!}97TTM15}T8R+Hx;}5viiGAQB()v>?AN=zuYa(R z+alHdQPKj(^a~!pOp_Oi(@^x;k9~#OYZKjqhCsO4VG7$H)p^E^H<`B;sS{yG1hC9@ zF5Wd0ybJOZ!vc%^#GY~Qh03~4EU%5U*esPht`KtX2FCzL*qfOXJTZB**}P3)csO^H z+oETk0xyS%3*^CtBiH!lLD3FoELQ7uVv1~>*K>cI0}5jY5Q(LsBd0WA29(zt<~*n; z$`5v+f_41cO|V_-eQK$E9)*V6-`g`5L!nDbE~+jRRN%Pj0J~s0wKJCG?l})H2)U2K z7au$Ai>^@!oxeeqHC37E>zG=A!#px)KG9=b!Yrs<{|c>G4&8R zwn{@+J;x&S#kS%A5x8ADCY{J1-`E`eX{vCj@+RjN0vp^=w)DS| z6B$edo0%~8-3iGrUR$I_Eu(O6rAd4r?1<(-9;pBf1MfimS#+Wx6-$!ZW5cgy2Mpq> zETkEy8!{kqRUZJpt1+5FS`#|e`Dc&I6Q2+qpKWGh5{%m9I*>5(DI7sukW%(kkw<5 z2d^=DYyxl40xs5eEaQdUI@}E48?HR$HZQzycmdIUd!lC#d6=yR526Vhy{+_@qL;p} ze**`v3gR5XIk$pjaeiYR%)SdCP`!ZezSU4Em)GYo77{Bn(PfJ2a7nzkUaPy#j~!sQ z96+6SEt+yAU|A;RFepDP7`shT)s&mFj)%g&YWaPYBlQFg4wfuS@KE_CsA1OQ{ z_oh8&yvHNlgE;8uyg&l8tot8~>$MtDM5n~0XZg0{cx0>tR&ES}SdriNo!+g<&QluI zod2pD2Z68Os}GpJ^j};x`gc`LNX*Gqx_|L9in^c~W|x>47#P&y)UY{Mg_V;BNJK9_ z1xZE(uq>&f-`bh{@-o8MU~ai%s-ID8e*z#OEKUw5+VlxwWLrc1x(H@ZFDGLbpFysV z?TFf7vmVte_aM_LfCUpu9iVFYd%hLtS=}8X#uy=k=|gI}s}l5gjdgQ+ug;0{o(=OP3TnBZ>S`^&sGeJWg?1wRCn2 z3$lj{PPN`eI9Uq=gSd(X>fQhdl6x4(0;^g5{RT&LtC%_P&x=jqqP~EL=BsWXdN=gC z80*1mCDQAm;UV5|jO$Battgn@*6S)qw#hdchLgs@056eS1C#9((rO}TRS94zqitgH zQqqHsLoVKxjYOMFF%AW=nZI)szq8UTE3?G|S%h4L&2#Jb&ULDZq3PDBir~#?uh*LD;Tc>EK_Ws*tm)_Oc8{|G$$j8 z9GeUsZ*sXe<4$Yzk?w+*c(Y0X=iG|`{UY?b0uckT`6!`}8Z$O2*Oc)q&FeAfXu@6< z<^rr2-^cV=`Ad<6O3!~j7}Ohpkk|7AoRig?k2Y2aSDEoNRFigD(w@HjHi$Lmb?GY^CS5^@9|wo@!>i?ZUaJ@q57*8PJdV zkoWjLG5N|LS`jMdw*Xpeq)?my2Y%O9e>`e6c$l^>BU3rd?N;o^?RU?5-ENuMDfprj z-8*DxWuf6&rkWQ~49uyXgz7#2r#fNrq~?%WzY&qMniz=;tNc2!gvB{jsJnqmqBEDj zk(SO=OfCm>s-ecf?vIrLP!|w2gl>_FZ|!JEQ2Jp7_XGzKYpZ9ibm+lkSH{y?trQ0P z)I_1u>A6NPa$!>g<9Qp*P&p%}Lk)n_T|`y@Z(IRw5VoREViBNB0=ZyAoKLJ>a<2n7 zNZsFkadbR4)t)zihl@2}ypR`0dPJ%rY=_txW<5X6Xtd@ZeZzlHP5TwbLEOuGFfdgm zW5|>oUt`P$C5bKnI)`z~<&n`2lL}`OL`*#msN!rImI>yNmtVM5XGfPa;U|!N*8t^y zioUlM1Hp{hn0qrE_Ymj{NrILEP(A%(gdA=Hxh6WAouq>*<|1?NgFu0BJ zDYkJq9j(-X%xF?>MZ{MamvC^>pCdFI^&1W*YXZPIvS@;3PCi11OOw3nH?D=v3-zYF z;K4XNeG@98N7wW93=Av*$5B#i(6WnhS;@`VyBU8M=Edm!`kyTDqqyK!T%4U2N-e~c zJcjoR8n~Wf$(vdfivL=p8e&FxLGOi0XvP z*Q5JtLsqxwGOG<24Z4H}VoV%kXu=NW&izGU6ILVESI#D$p~cQi4w}D>pz6X~lHI@k z*ZVJ$CjBd_FMSBn?qC9<0)@~t#sF-=x2tgyuhwATsCNYK>B$sVYab~aqiB}&TKw5u?q)(OVoe)9N$l@L=PHhPc)~ZRk_C#Wzl7Q@O5Lb1 zk*7FK^v+akVi5h15&;Ih?;{POGXb<9*~H44xgy4UK(WJdCMc;>nPLx|y_~=3L<=%m zTDQb1EiNv;;K8T~`}*mgLrqv%9weOvbqYYC1BM&tgnFShhxZDJGu-SUqK4_VtYYkP zlACF2q#gfcrgp14?43Z#L_CCtdSB&pgg@0Hb)t8HJrYS}=}b$Ug|PGG-A-eCR(S@F z4+=+C*n|vduJL3GcwF9ge`?t01L{&>a@}zznrjma4o6m1-{ii&JH5xh>Ad*K_xAs= zU@VOnfqFJ9rLR|u7!ZM6zCHR0D_w`VZyT9D{IJjr2$jQG(bvJcqenUdGSK;5fBF7L zQnn%4VEnjm*ZTjMVp*Mv5;k#km_VhT z%#&?^^)i%eFYh;6=&KB0*RdWW^b3C>DPsDHM_6*nPmHOTEll_$0xf*b zKUHyep5Bed25df#f`Ac=HRTK6tWROk=)mSfpEYtZ3%^et@%~R7jeTI~T!G0ICh0G# z3pk?XDlxzl)tE-&QgkAy$MiETyKPB8PbO3k`Q@%a1`OeNDkR3i^=TMoL*m~>;$)63 zL$Dl_yycKPig6ui<7SDrHiT%x0M;M$6lEY6#1$?=DwraBLYAQF;Ne%v?2N7bzc)wx4$ zUV+pxA+4`5rsI6*fCMUVFY40?8zk}1jI4LZAZUvgTWph#k_+|Zpc=kV&W9-B3QQ?+ zJJ0quqwe(s3?q{p6-jh3)H*zvO{E&@_h{#20!EfY@vbd31iDa5Z2e$HDY<+D?sc)) z1D+H+YpI-KFQON^yF5Zr-t^m?#I-#UKITzj9l|~K+dfolm8e3AN>_>}@foRXh?SrE z;|+0n1Ye7e<>$Qy3;d};c3%l-?$vX+)aZty6F7WcAdH7h2Z@X~R!4LnR^?l{j_SPO z%^GN~n!AQqTwW{bH@_9`h>l1YQp)EB?T7GiEzIE@VGwWktOFjcPYAy!FGY|L?Orzo zT7cC`c$#5X)djkD;xhsyxCfyjx{)o|a^@+*#~{49ly{7HrBF^t)7k~_QWMe%N0cqh zQBO8kS*55>*7)eZXD{Pfg?ZtLgXAgT>JWr>H$j^{9fy6j&H>d(Of)^;Od-ah4id!1#8g*$Lw5IGA0I2AE*|p$sCS_9C`V_92|FTU5eXc# zKzz5r2nj-n08G&|iiq}|qFZWV`&h5owxjdhx&xoztenCEbYU>6^XSBNO9(n##MRX+ zQAwg!A_Y+M7&+cf@IWj~^=+A3V9Bo`868rKS(SW9ujmp4a6xBQ0BWXL>ySMT4LSzb z4i9t~+qOhV25c(>xpg&~$y2*nV(SA~2q7kSNjGJaSvXVWZvirvC(@lXer{}cbaXV; zPHPJ{PRfV=@)8QP+zb`WWTHX$q6-YqXuoJUs2XEsFv@9GH!|+EIIUw0_?Z8AhOEsG z>g>(rq4*>%D%sVuJs__Ax&U$?Ii;wKyLWm~&CqM@?2@2Deb7@OtVc9kwh^|4A%6By z%(HjB$BC>FWS71avVnXMDt5P45mx8f;us2t*QTO3jAED99@qx3V#~qd+-)b z$YC^{k7Kq5f4LJJ0;2<^5k`t^3PT3!VZfc_Zzb2-FJpL8 zARFpo^?{vDA|(=rsLwiH0G7kc?uZYHG}krUAUZCke7}SlXw(q^)HzbE&U^7+XdIX2 z00yNMGFDJKl!A1|i^5(ZK{Pwa+gM~wS(3&`ZmrG`*ieS*)N1I)r>gCm@ekg+#5Bm) zE}ivi+V9NvuLg{Cu;Cf+*$Aa$%4Q@p)-sj6N}kJpr7DeVDbWiA#O~8n4F*iriD zZGCro9_?JqyZWV&bbz~K$MJwCUPTUe6T+7KtLY6W`BYl5)1-9`3?iY%rYMQs2T$U_ zl2resQu`nY6qc;DDvEe6un`T)%m#`f)yRPS#C4EQVA|CY>W#uJd#~HBsYHY#>BztX zm4S^U)gTo!s-IQ2hp4|863JaNisGW8a)7-#z#ure-oqvrC52hB0A#M0vBf8ZWe>U^ zDlN}$uCT@8K^3U>45rP1Pgjd#d1HFjaL=I!YoDI?oB!>oZl6Tu2GkE}btce=67Ia;rTxH_U= zlm#}&nKD*daHlVA>ZLW@)Am-Jt>J};&NgNE5h*wxcBH454hN4C2n6PqsEFAKDhqyc zWejG3bGm#O2b&b7KeHaz!TR9|T9{uYYHpbq(Al%pT~WLx5tB|Xh`nmotXYyFuE@el z5|`zZ+&iYuWqNQ;h6#F$hDQvj(XWe%xop^ivIZIRMjMMzp#4 z6A*Q}r30@jrC~t+d2z+PiuqIHWv&+mWytH{gQ1?ZI>Z%C<9sa|w3|N+q-g94+*HRO z+u-+bAO1}ogG_Whw(sy@d)H#uNPzlPE?ZM4-L=)U%@b;~*btM0n{~P?ruN#2eq>m} z&S_yhq`e@i7|aUxdNU=|>XHXM+ryfgKzOvj!N527%O}OP|xc+%r{l zQD_-PjxW5l+_0#=jzsr}_-rnD2wHjBO-<;ahp+V#9!@vbRiv_FOf5E>7(;l$ZsytR z6r1vFb)>7d@fgc2*chJ(xA6wC;XO;H5?jP2!vNcUUICFaTrzlkTR^MRIH6!m>gU@w z_C>U$E-ioso35DoXj9QnGV`e+>tR)P$T556gHh_%Xo z+0yM^VUwuwupcJ9`(pzuiUpV&Rwb1MH&E=H2~X)Vw?o+#U@y)mMu&pWGB3tJIN1eD zs`}U-HeEiQgy2aPY*P5Uq?B0%C07iZ#%nD(0QU#;Iw~#3)ROD^G|raOvr3}Vg7Z@)!c--_5eNC!TjksSVT`%kF$WZs;zPlzZ&Jg6v^`*tdt)M&3sEk>Pt!}A;@twWv zV=M|l_G_|%o5xHP8e(c)&_1B6*^0%me-hSkYuc^48HFj=#`yxq!vf7T%&SO`Zmr0BHvozr&euG zxAB^ZQ~@qx>Jdh`;&cNOF1sVRbqCIvS-~2i%Iy5sL@@bC*R^!&_@&cN;onreE+0^~ zw?L#~Zi)8@uaKRmYNbrlllI*)YryqLBZM&DypBTVQ?;>hW5D4dg%{$@d?BWIA@T9? z91!_3zZk!A`l&Ret3+d!V{=GV1@ri{XJv3#D39oV@ehh}Q8Ugf@FDQXs!GqmBw*j& z7xlOC;(*6X1#+)I1HBTVgHM^;_UsCAyL(@t|H4Xz_(wkSqHCg18&}r!hpbh$ua~k9 z!YSYg5us5_5O8m4YF5C8Yja#+LF8n!AtNIrk#k%=Hk71HMK{-(m5`?m=`+_EV>=aw?uQC+ezaq=x)|&O>E6rw znFv2LuW4CzT0(x2Uw-}KY{#do%0P3-D;TD5+R%|Cj2N41O_^(%E4%pHoioB*(c75V zD%isOhv!|Jaq1zlB&hEiV%bXvuxaH7_F@55Qw~k4Jfcf(d|`^M0QM!($A6|tuME1t zrQl%_utLVi-X%>9-bs>eUPs~v>P}G__u`-cT21PhQPxY0ICx&{y>3C2Jft~G6?oP$ zu_X9Nao-~P`s%j^ufXdQR@=cW%$wf{rq{Yq@I@0#^o?sUnuJ+6!vFk&2{k9Lg)7XH zRItm4fMD-YQQ-hxzXdjH@i!g&?q2Te>jU1u0LRjtKF|1Wgx%60$GDaClBGT5(d}aK z!QhKbgyxAD+~%PsPS4czJQ6)(*C6#ygdfSJ;85pM!!&%8_pY3zheR%=;5dO(pfWr_ z!9jkPXr0&hjFv(3henNXQCqG>2<8=WDHR$&oz=@X8!mt20egTq8a|eT*__kVEo~2S?lQ2t69F@AhJ(u6 z?}>SiZp98%x~9y8{5$4CWJH@pZ-b^!Agx)OT2jMJ*}3Mc>#m}2&e!PK!7-f}aD)!o zIl#sOks2M@s5n&qUFOpGVVgl&zDf*s1QIu5btXTkm2yNGp$5gkJZwECU1f+~rO{;W zbcignO6vY+n`Rm*hx#)mr&ko+98j9z)+Yi%Hb=xJu5^e5tXL!E?rNg9Qa{-wnyq>W zj1?(J_%$6$^bE|wM>+jdyTD$}r@7)8uPG5jPhtYnt!O-~<9oH8GJA@d^&CndX7JjfTla@bp^M)0++pIIj<2*8gHS=>_yX=Q1&`h+s#K#za9&20AkJFx(Pm+thjdjC3|o~tqrOuB^9hQg31Okks!}Ev_4uN(Z66{jGlu8K1+7jx??YG72QB|KZ1z2HMleU*EpHQ&7Fj$3J zKZ!hRLDCVbpBmqJ0_u%T_hwVyA<)|noj0_#g_FrsTDqi5xT5UkZkO5iXaKX(y7SUZ z3G$UnM2J}j)+C3PBQ#z52ltDPrM(IsH(b+dmJO6(UP;bbc1l>h!Sd-b>C)el#0~frrG&+sN)>?97_wopw>G1fN7Al%3vz&eioKR>MgNToo zI%(^O@lnoa+`Uavx~|7J+e)(q10-ppC#CXND|m9Gp{~Y}6x;5?_@$tTz>f>jGWQ}@ zJ;fE7yO&!ak)eQ|Oa|b^wnkVur#y07&>zzye2`5b#>H-E-0%L zCJ!dB>LV)NRh%~zc?`W(8CsBy6yteep;q%_#v|2Vu$Bb%7?ZjexPLJ2TnQbmd`L2D zgU(MHkLppb&!>ViHLFA*Z0mqrt`_B&_%6mGV>wC6&J4>gB`P|NBcXR4Sf=ZU7(}## zsiz)FEcIB{hb(6`T1Qa}dm>Y=$m)4Rr>36wc+s2lEjQ>G)ki?*T7=piFp-d>(n=~G z22JQDYGh4g?_OSu!bUF9g=zG$Q|!S;p`zEnO7x>x*7}hj7n)Ivd!iF2AW0{Qe(%qV zmCvF4D>L(YHqsiZ(NKR^5>=WFHX3w*LaICRqqaqUf{!Bp2kL3Uk_|ChFeIT82!YkL zyYxFr)K5Bh@>_y-$^w-X%F?2H8FE;b?6v3P>iHDp5F5*EsQnUY#K_l;HjqU;Kjb2g zXir*4LMlj`rGZ>ot1zaPs-ErYQGJKVqkstHiRf%PfY!}VF|&gSbb45yGwh?&2NKH# z6gKqS4v7wqZ)3T{m`(w*Q96_8!D;KGDic_jtMEuAri|z`tk$N8E{M!%Ci>)1l^ObC z>v+am)N+*T{mcV9I_=uvDVtTp&8ULXD`X_m_n2B=+cMk+F&ELwBbSOEc8`yL0nGVG8buvTt$a> z*PblVhi4_#l5b3 zO_#2t*0PRRLzd{aScaod!TT%GU-sR3qjpRu7+i~BG-pVSBMH?}uX<-&heL>1-gIq> z3>#4J9dYf#nI#8&voJsQ_zVh!V+Z{L`jT~tUGI<}>K6GY%!Wx2rJ*B#yR1a_7s@VB zJGwqn<-ls*H*PoUJ)!@~!2E%AeV?qN2Wwt9NiMesmUrg#`#4N z*)aH7?9&Bi{l`1cR7TF~7A;q>>aApFU{~z>2RkaXBIq3>#5%GpDvf2*$yi@nb7ZH5 zxps;dFs55iDh?oEgz@?wjH~#j%KXVGzM_Q5-Pucm~ZP@3kr+a};B2VPVcd8d=*$T>t-GuruPL|D{LU z#3-GUoij6zS~%i4{6~!1k`Y1M4r*IhPj%3mz% zq%Z!JU1#K3^>4Zl|HodCHY%%sSRP@EykdQl<-^GaiZ#E%l;n;*3wGEM> z1!Qy$ZIU^PF>a3)s?H&!lQl{2QD87^!q^`bDqN#VO8BW2xzh~xoFiMHUJsH|0Pzdt zc(piE{b-!s3q=q{RNB$i;BzV+08gu^ZneGpa5~vQI>Pnk0187k0xxKoLV&3yl{lIn zLMw2pN(2=!L6F@8EzRHxvBc7y^2lX<}>b$v< zn7IeL1mueduFw^dw9QF5lm$#xj%4*&fPo{gUAtz;m6MS<0Mb(#jN7VYj2PT*xeyk6 zij|cWS?^qdCOhj5RRoCV=K#hcd3geF(tODC03_rHa|Tk|*l?VA#U!tS>!Cd1ln*&T zQLX?gT0o2>B4y&WUVqd2lFIOaJW(4j3*n2FqV5}SMqKn)ePS&U!;hF#r{l;qP8oP;Ef{Z(8J7WJ*g@BE1$vl|zDGyn@>rE*Z6fhgKhX_iDOXHYQ_A&Ro$!*Ag4uIV%<9 z5Wm~76=wOp&Y}MVhacQ-I^e)Zo z&{{oT-`F;Jmozb&VDxb@n7Xn*g>b&2e41!E2>l3gxO<2@I2gS16g#}dBIdq zQo#c=9{1XvWR+cO1kN46ZDW_AC`~WPV;im+CZg)8+;&p)LNFv2i=kyZ)OJR6y1dh0 ztQxmH|)lVw0`^QP%Ut&5gR|v&4KUFKsC5Ox7$B!p;komD%2gywm67d(6 za5Rz%WGv8+3J;Z1u{R-+HbNmaJCkrG)c%($#CPhqcWE`aQ6UPZCLl6^&QAp0gpM=r zjLH0B>c@qLiopjlbdf`~wwTLJJyHweZ(afT)gjY4ht6!{(1xX-f4ftWs%Pkz*~<#*nQ%T0d{5h;r=oAb_w+*5`1LrVZyl+EZMZ%s zZfaQeP+dwXJ`Y_T{Oa<(LO@$TKu|ak&_sL&S+3+_DF?uU#u88D2B)45_t!CI03GBf0*VzmlmH7(o zi0pfMVGer*dV`ccWzo^SgA;MAtc;AVB6Xc0Hr!o|o<7HGL9;+n!zBAI#}T&g$@N*Z zKa)cV?Uv+;T~K$6J@g2v;Z^-->}F&Lx;&@Z|0L_dR79*BEH^K@U>xUPTx?$ck^HP0}#PJkf7s4bQc5V)B>?~i}Nl0T@I zz4g~jK6{IIuYRQHsLahnQ4?BbGEWCbaTzwdxGn}lS-2cyEsK3OArX6El65V_Xn6&g zAc1gzPqA(ZB0e0VuH1r(w{F;`Qm2kThR=$_u|ORL%0tW>E{UY4$;l5R=pfhZK8OhEEW$=w-X0xq5|DDMn9~VO=iwCwI zNyd94{$Sb5NG08L4)9n9;Q?7i_U*LPuD3j|!V2fJGp{4>72M{tH*+gCw!O#{eVnWlg))8@atmAyp{ihW z{*n4?Cmg`XzZ6Uhjo>k(jD3lje3RDPFHX$Xb{UGb_gV(*w_06<6 z34rWM1sq{-Nx+)*IvE=WQ8TZC40x?6k-X+j(wzpfhp20iXG}lb*nWiZb~o5@;D8b* zh)gBdwSs`P0*Jpw^IaT~L`0(mgW^R}x{>>~kl}+zF{9>kI0ouy^P#PsJfz^vU`ow( zKy8Z1**ioXkRZLwv}hI$Y*@+R*nr+W8ar|JWwLJ|PW|-`%GD6E=i`@!DF&hV2;Dm% z^EZdAQJCK$EJ)a`piUf!*@(F32tGPZ{abJhWSvJSo%{^spgfXt6LFFdNYJlbZ~6=m zkz;8KUJ=^4d>pn2~DEMKv?4BFxS!K z97#`g6{nwhXg;)z;5t{;8L1vFkK&3*XZn=;N{q@w=_wD@7c{#N$KKXnW5v zkMdZdg}>}uYV3e^4$m#js7s3_VAt4mSAXI&3ELSWlC7O!*7I$TD9q`*Q`5KVq;L@+ zMj|VB8GbtC=cF(pT_(uxRXZzW32Q|Msh7ewgWwPHZXcozj+C5a7=chdv2`#<(|a|- zs3Sv2O=?LFLzx53Vj^fEzV93U42%(Bt_x-;Z3V<;SOMEYBoFCj3Zm;?RVo?<85s@) z`3{JC%?qxsrFBXi9kQe}&z6VXBsCOR+j#4`!_!rdlh!GkAzzJ&v(H5XgCXLJQZ|b% z?h2{Tg{1XXTTo|&9+YIiNvVBMrY#xAcvQjdhwyyy0?JqD>?PhM>2Q&hEgbCc<{)1# z%J%WGUn`n1>^;9owe<#dM1j_Bu{ufgkoq2ghfD)hmNdsPoCa5q$~8}yALynSOdnF zen+j|kEp8cWG4OW-xSIG-&E=xS+f0K1q6(E)&H4ni2u0-^S@`2=&|zmEMnB?{-Z?X zzfoh6iS=h(?Z1cU?;-kq1pb>ZG~eOJRbE@R=r!BU_OV!Wey;ne*D(yhToB z)o+-$FyzYfgcYuJsomaPznENsJjf)mydNE3pAj}`zmD5T-=1H;(=)O$&aWT$_l0OA z{vJ9Sg}-MVjl$oPK{yJ3PlmrIgKz}sU%nDKi(}@oW}^J9++3Eph2S?`-3tSQg8{%Z z^=@gchs5(%lZ%JG%D!kf#kqqJF_x#=hr*{m5)k0hFTd0RwxA1tU;*rwa;({2{6({s zJz-?gT!dQCv}w~8nwgo^BqpVx9!hKnJl7r-D(a{2P*Wvw;nBN%Ix zT!dm+3<{t;f6KHZZQ+kji$_AjxpT{z+3H;$LqkK>g<;=-Ix`2=rq+v!jOx<+J2|ZI zGAs7RxuuG8&yM}%A7M~p9)#V&(&Li!;6@x6tsfhWt|aQ`&YeqbP}fQ5-qAYr@PJK}4KXv-_2t0p`_RVdLrLjT#_U&6OaDz?P4Zs=E@SCGgTe@zY5e#9p z0kf3c+O+^~#gaOqe=#yIF*57p%(CO~8TIghDTYq!K><=|BxqS82 zSJAt!AI}Q!)bQOnCsp1RY|O|0lb)Qe0%ih`GG1l&DD+E zIVTuGT7Bgf4K+2jD6|JIPcYx^A3E-X4{mJ>kVHSTILOLZ_4fBK?q%l($;SYsPKQIy z@sLQ=cF`nW+ds6-MotN~+|fFbXQF{>!4(ZZ3tS)yKZ~hncKk19&lAZH3R@?Gf29X* zQTW?5hngbz{yDw=>BHvAG#!9CT!@gS&Yc@K&u3s4lx7-E1w5bOSfB%6<*V6zvLF6tzD|Bs+pY8W5x*3mdqMncfyH@6+}|HN~~ ziaoobaod&U*{iLnf?l}eF)OcaIQE$ruo!ta#Sj@K4&bWWfY5$cut+wNb#2+B7p1-M zp`(|lC%qF)MEPfyu!boHx?a@DXiD_{-cO>Nuaie_z#W zP{2KT=ER8;hp07%_oJlC^7bJU=PF*V@4+PctXU@H>bD@!CmH@Rex(~FGK_8k9(0&ve3uTwV9o%-(MbRKQg}+sA-sZhb zLZFOJM84ZDEIAj)4_LBm1G+cubW)UUJzt!-RU0(6=Yr)q!5$gjDSsHu-@jC~_#DG( zTb8i!*G7AQtFtvhvy^tqzfombkz(86ei24p(b#OR z^ex8$wQJL+P3iFrbo2cMDtBV0+{GEQaGidX_RUXBI>0gA=CpkN{0(IPUFdB8sKKcb zhaz5CXQtaZ)Gc<+e(DfY6{&S?!@=}WiDSo(Eu7iHsZW$}J#9SrS2uM^Q(uMUKYi}p zkm@!NDlO{M<$^TUuU}86Pv%>kn={i}@++geb6_Rv|5(R8)EKW@Nz>gY1zvzV+ByJj zUiMLTtlg2`GOBW7T;7btzE6M1zv?7m`oRa?a+dd`W9E%2dz$Y(>8EBj<1_!X$~!mN zcI{{3QXFNg$LA(DulX!|&Xx^HZ_6eK{}x}?Z2c}p(c_xbmOV*tw|?g``{QS4ELYN( zDfsf@f-lP3n~weZrNhr^3TK*@+u0S@IK1rYpQ6f%2oDeUSDZV0_SM&i7Ee>v?BQK$ z)@N&KX=(j(^Mi>Rd5Je}z>6qqsD-fUI1iSp{?@_ajMr6Pe~yT+Y)6~jfXatI1h>gg zJ&Efnmpgpk|HgjtXV>R!@l1NF+VzO1C9&D~OpjIGH%+IqgSUR~GP~{BnK5PJJ{!;d z!*=bP>r&$qVKA3t+%YQ>Z} zj%#Mm%(B~NSJE#bSfQQU>XJ@rW8At9-OIaIH4*!qdQr?6>q`%a>vSBs2(J5uMY>;#GaEN z#arXg>!Y6=rs_k&NT%x3v2h}*(t+I2H-Gy)XHZdFcpJu$a|e0XQFkA1@+el>7Fy`rkz4)>s(A;&i-@w3tzhcYbHLE?FZ^WIe z99*3j>mFKNpxNNnk`gNQ`&RRZuYNH6D9lpli!Z)VpPTJav+a0%aIq|?nCD*eBdReS z9O@!Fm0c@144;qpAMj*3tt}kMi`*3F&!4}Md-}{7{T(}YM7Lftq}Sx?y{*+TCS+mS zp~}YpN2R<#*PA_r4E#6laDzG)-oL9r=2Q_fsxMVnESY(v`BEAfsjQQ+0 zb?JN+mF!K}Zd5|9@fW!cA?^vi#;@Apz~%UL-!+Pr+!eklLM0FEjq0`|e|gBeY+xw` zCG1`<39d)@Tab>Q7CKLwv+hE58#Q#H4G0!v%hs$}^Q_IdO&KF6x%I<+Z^ z+Z~wB+V6)dVZrG7vKQ!Z6n zz+5-KyIWgvWmIT0NDb8jcWHs4r>CbbHg6|0hLqs_HSPuDUG4`3M8nAj%r3w0zu(a< zD{u4Shv@#w@#-%vab7k|^zvYvqTS$9y>9h<*A-@2w?Q!M{clWisv+)ld%laxn$hVHc3}?02B-OB3m1$OJ&2%sE*+tPMv2`V$b2mIAKbzy2 z@zMy4k@PrM5ZxtQ8+JA-r*6>hPCNYi+Z20VNE{zAsm6?zraMZ!K8uF+7@qF4IcA_Z zq=!6gaIz-s-~G?$-RBu}R8np!JYqf3t+ye)3dTrNn&s^7?(T%;SOnkwAVxy!*gyX% zu`Z;ln+U7n`|-BG@hjEfK0uy=VC>`>ACQUYXb6mRMymp^{^nGK zN;YNMh#8{NH*2d@NExSC!BXL%`*d}LKs&;^S5n?wuZov1yTJt;0}WJcQCyGk8thX zyLY-KDr+$bcHzZLnbV3?HZa5-<4yeWzf6uMVsbgE`AsEP(!lzj3rlrmz#`E#c?;VP zWT?jBZ;=LSDtX}ON3J%{#f$(aDaf=H6n#o1&tO}rPh6SRv*E`de>}+S|NDNTe7@{c z)$B3Vd6zWdcY@##c|DRL9eD629e%w5+0;(ERO3%S{S@}K$>^9ZrWb3ZN8y2!9n72b zQ1hs0>p6J=oi9Pg#jhhegWKG5a3sT!Z!T3-+=E|7sL==W_2LK(aapUv&vU*UgdetU z+@3TUl~bXiq2zeAWOA#(y@c)Y|FQPwaXF`N|9{43?CV&Q-IZh~$MYM4^dDLX)H&GbB-@6)hxM6{!&7`#duDszl8azR{NJkg zjDcn$S-u&qcce2Z<4>8Em0u?IP9F*!cpaaI0~mnJigkRuXzwqY)okOyve9|2RR#tp zM-4MA>KiLVr*s@Vc!pA*`&dFjV#+3p*`|Y*-7pzVvPs$W@7w!-|DWxbZnY|QI>aDt zYpSNJPuG>)xXZA#3xv;M$@x`FS77>dVMZ4p&5V3;l&Wt##s;5@yVZF4^5xMXkN^C& zA;;{%yZ`MHh<|@h#Ukqcp?JDZvwZ{mFSr;`5=(3^!%CMzEbK`(Sz7I|);x642#b4d zD0HVVF+`ab#lPpSS#yE8KFt>oJT`0dY3iyq;BgMGna{%q++{KCKxco(5pkpj%=^EQ z4b3`+WR)eC?w<6SFkq`;*`G)_n8}8;cF!zkmN3jxna2-@lqmVjFH#^nYGh3E%&l;p%@@{*YpE*ysQA zv#F^)ZTSDW!|&g}_x~FsJ?Xkr-RDiCPZ=sz+2s#k2G&1HPI8kU&adrb`BmA!rT;x= z>cz3<&6}s;&5$ey$24Qwal$9oOwQ%7-ShtxlVt92km@)JQMl^AKh9~|4}C%21Tgy{qxU1ud4a)+ZzuB0r^L4`LNF5_|RV|`n(7hOv;`O z5i%7v${4ssNmBofoq8P`^TXlbbbtb0qAJ7SE`9s<#g1HAEC&Q5p97)o(zgG;TxR%A z-j6~U5@wZVLZw)QFG((dFG?f10=llg|8bR!llH$NX3y*@mpD}bzS&~ZN-Cw3McC6< z1qEmM`&JMVWTc?l|F*bS%r^@$rtnf#)6=(xPLiujE>k9w?_3z%23mc-du7Z} z)W4iVwXRqHXFnOetaRpi92Io zw>S6x(S|$RSGRt>Uo2EcMwfwl+QX2%d$(Sxw{UNb^SFq*r_?9Ux{h_61*d*A!|JG~ zr)R8Lh-5eBQW`gF(ZXVZ(~tck|Gs6q3Pxy2taB)Th@U30kMTrpCe*t!b2`3}^I5TE zTH`OY_XV*JtqElt$DDd!QzIPpCZaBl4|zu52SMoZvjOHtSGRs~EPoO@S@mlDsQKq+ zbnxgg(Tw_-)oF35q1qvX&2eUnnLPYlxfLb5qU*KiphyG(&r1uf&rEE$q@h~REgPat z-dFCO7(S82M7&U}8iOKckq^S<2-}+3tP{a=&>V<2&Ww$Z&$>$Dt zsir@vR6mD*A3$B!VL-2j7%gU*ALE352k3=J3@xpS;nx~^BL*N z4>|pzi`w>eNo!(2O2P8`eCILSzD7z7^VeTPI}9;1K}4FziltvM9V?a8eY!>z{}l6O z_o}gak-<}@PF+i_&}HPh;|IHs@6Wm9AHu3S>~$mU^-tc;7pzpx1nJ^eiC!+i z_SZ#4-Wef3soff{RdLIO^vdm%KW(L3tTZjq^QWJFN_snZh3{R$TOM5l>CWtE;q+Veva6_JJ`?|!JH7;?B z>)H@t)K|@#wJdby&b+N3O`t)^)_(g^t4)EPot^pP_pH`N2kZXu5_-ZFJld}MgNngrgm@F z6|=aDR58I+LC^kL@kdm@lexCu4_>R{R~{ZVa0MAp4I0{8CQm&#$* zU4Bc=CNK8sx?rk?*Bhf0iR-e?JxP4qYWM9+8O()W%a54}X&Q;H7507sTdybCqE&FC@m9g@CqW zePStVOP_@Yci;S~EeU9KWwx5y`g&xvK4E`$kQvhmK#2YG+&_q}hcq~kC6tAR16{-J zU&695Ex1>3MwgJ|-T}1bUZ5LMNQ`4~bWVIcgNc3(XY_qll^5Ck!E2*QleRbO&_95b zJVQk@LOY}vz5Rae)zlQ?>Zf^QDP4t+iA$@(p1vIP=~(;V469Z*wv6kJ5GTiyGm}Qw z`9D)QgtlRvRjG!%OxzYri)?S-2aHFtCsRY-U=Y5z48p2&hh=X z)SKVyOwD(WTu{o+NjekATm)s))~&PXlo>$-@%)Tejbd^Khh$mrN)_Qtv^@BURVzlR zO}d`%l_6b-hvwc@TqjDR0xT%J6G4;=U-MX>+HQFB4W0>S8DKD5N?@i$ka@J|Sz9Kl z@KssafZezKxxQigFTKd-wtXV9FZZ6hZ@P-@ZxNBLseJtSv0NIWZkQcz?X7N9Rq1ct zkf8MfZ2cM7K{MK#ZsT3jTO30T0&dUdoG;|d^0p=MQN=E1#CF(8!@x6V&nEx9N%@&M zi8X%Tzy2oLFni(Zhj_N&?mKtxtbs-Q7s$P}&`Km7N9Ee;{8d=#%sk4z&;%s}1xgSv z@aE(uq>1cTYo2ga5RXoBTIaoSa0A@v zu1!_QdsQX1iN0e9u*1{en{wo5~gnn(P8E0jFTw>QZ*OdtjV(2w(ZR<`Y6LZni!<0g96e_H^pCRu%qQ^iRdZLCkfEhsRrNHL;u zC)_4`A?%wGN@x>*sJZ6q;UB3R(_@_MxcKM!rNqk%gGChQim4^Ck2r!aPg$qG7ragl z4@1x)#6Vs#P~t4w)`{i`KarYm7gQeK^a))WDR&}^c4u9hhg(dpt<(+_2=k{;Z=%a! zp~XoBm9obqHMZ(HYI*&RnLWPd2+eU9C?d6*seb$IH)}WqpP?&$cYK>?Gx$jbv3c8v z59Qgxco6PGxOa@}rluAuCD-H+|bEirs8KndS`!u zh0pneRC}H`=~KRW{rbOXY!0OCkz--m(stHC4IRq^Qez>f-tp~fjLINrlqBER9W*;= zW%?_~AgGJbk?Wo;Z$WF13EPSmEhjL_>*DSFp;~Dz)rU`! z$UyG1ee{4x?HQI`-etJO9)SwV(6OPN+r2z+eGB)BkJ<2YN~}3W)raBFieQzNu{Vw- zTk3dHEhe?_ORx#H8M!oeOm11t#VLD#5jG0xbw&668b)U?qnl(7%MAX`Aqlrk?jq-W zW$3-<&zI4UraC{{?+SHwBF|k=IbDTSsovJW*ccc3KG_Jp1Jl+rb0udqcYv5SY}_iH z!<%fnj1sGGSm|5nc&ppHT3Q-Z)VS0@HT$&7EjZ{6h^ZQP)-+^M=oy08pG>y76MKD_ zA$E>Ux=`>!K}GBc?R-)p8vu+DgquZke|&G?C;uiTV2}_<`x$aB`7L`t`@x0FF`LTH zduZtOy~y?2?;j|vH}-M0y_Ub=;{3VP?=^&(8T-mZ^sL_FO$i2#xj$;4w)kK@);+84 zCFi+sME0P-G;JJIN4e@n_40Dw=_8R|g|uY{GH7Cd8!2kG=h|k$;kMhE;xY6YSYJp* zWo4(M)Ta?yeD`)YbUi;(E&>(9~x)|g5<~X?+D*?5nVl7dO z6q!-(!a1%OJ?&H+h2hTKyM1H+8dq1HCTq6n+U#oGgJp7D^}66Ewe_=`w`^%J!tfjI&uicen>A}TV}Zdofi9$Ish8W|>J4y;AM0L^7H&8`N{)X`(_VMMS&H#ivhFzj1(E($b$OnZl+Th(qNf3^i+m!E$p)Daw2eez+N9p(wh3nJXSyFr!f_N9<7|t+UFKj z-h1?D$m5T;yXrU8JfmI?+z@lUj+;DoK5aX@{%NHE767a`v8f6GU?;%PEJQE<)T47x zeyzd%u@<2~wX!h@=H_7S%0FFl{?q{4j{_I*D&fdSP~dbjo5P(JUTnZ~xE{Fg!n;1n z`8vm$31f-=aXCOkcQe5wdFL~`%w4ESN>qi6KY#s0G1rR{_EbQ7eWxgs!bP>1asEw; z69@F@L{fOTx1CnfhK+zgsoI?MWZ0l^d8k99~%ge2RD2mDh|d4W=`urj^BQ9zS=*w~IR0gk+PZ@NCD^hHij!T{)(2K5mI~_=sUDnhkrO4A5F=AENxG0_!Uo_#EMe680w5>bRpo3!kPmV%EU?zn1@3~7r?Sh?<=dfv+?wY#QB%7U4*S9VYr}|NU^3;} z;Ka0c6DCqdFaFL-jU74wzRbcJ3sffa%><`f@p20{n6HFb%*H+CDAEXIk~d|#SPan; z&;$C#>rWUh%yQ(`10(fxz0!Z#mE!U1ChrHDo(Vq|`Obx~vuDmI;#*yqX#}!!0LnhU z`Yvk2I^Zwn%fp{Pln?09qsKE~we3sGx>u-4;EaDsP59rmts%ZUQ&uM(m*5WMXr~WJ z)e!zh5r4gk0)E@JZT;t8IHb_bzTQ{&@%YQSwY616!aWDESwXtTM%&Kb0@H9QvMyt= z_e>IYCNLO~G68@)^|#;3xW8!*_^^f<1wqsV+L>?U``p{syH6iYYTSlD{gk@EOC!H3 zJ@XaV{xMlUglI4FMYQJIMrxDvJ7d$IgZ!_#rwr=2AEVem`@m~=4i9bO`0`FuiPFNMz+AQO-1!3aEdb#H znz!ZBtzl~F^)deR+||qCTo$Byn9d6^jX95lRGaGwl&8H9@2yDn`1>=IV=2xUseFMX z14)pM^1gna_WU!y6wVEI?18z6koQlGjrayHT?XtoCXfQsM+$1?|DDHmq-k4kroV9? zyjOy!dYks`70UOOm50+xj(}a6d7#i7|8#sGK!%wP1Ae5Sha|%sFrT?YBFMiSdiQ+e zmR*EI@5hZD6=a6S9uALmI11`YGuy81HE+%!;o0-4>F+fat<5^ACtIt4z1ys%b^h2?d8C z()Yi?b1kT*G{t{%w6W>U#Z7VnToiR(#yS7R?N`0R{y?(G>Ac8qnE5UwvjEb4zv07E zTbxiDO`pC;urKnID5AWRnF|)3l=rdt!G+MbF>fa*LuHJTV*Pl^0i}(ruO9KS-@4j zsBLx8QeDFDECM2r^!)Pq^L@&rg3M`tp8*}$?28PE6AO{LQ?O*7WoxyOZQ=Xi=m2lB zQxxy&OJ2K&^B!FC`ASV>Irb(56k&>6GQ4=A7stZ#z78NFM|kpL6~d?&BzR^q27xtC z%1y#*8^po|^QL8A0i}4lXxp?I_n4E|BxWJDTg@aE-Y3gQubzJV(%7!>3whT{(r!AB>r@ z3}QlcWM#IEdkRU;5yg5I`z6;dQ=Uow_eYy04Z*k3XmWRhaH8*LzD9{Ne3$x9q$g;0q=DENPn~ z%swOm<3mZvW8&X!7NOeuy?iFt3DwF@r%lP)QPWY?B7SK=t~zwN^B?(AFMO!TweDC1 zM=K;5Dr5MzIv1Oq$hqvnOjwQ@E_3PEv*#qEb^bQ)85G}t5_x?G+_zlVq>fwcM><)o z{lTOT(Ej#{H(s$o(ht-?_5A@MYl!`*6FfNuqfWL1If4xdJ!HQ;_w`2p;sOr!AJWUf z9U5ilZS?Bn<3V0AChMf@S&Q0Nh$cv?8pL1S9G5q%sUJQgBBGf1iL=dGx7Ksp4gGk6 zH?(n!8C3q*>CVoPH|jnt6)_IPG16%T9xc&GECyFtUHkj)&Dp4=1T~BM_5pp}zd_qU zj`p8dA({CW!5Un`hpC>vwS@@#T)p=4qELO6>XM||k1lcOf_UUpDP}_uK7`>We~F?2 zbLFg&*O@Rl2(6XakSY8C+|KcslK>CCH2;YFfzUONucwGVXp@lJmNK&xC@m`E2!dgl z)Q(2!=bjw; z1nz+P93?;GS=j2?=Xa-geCP@=IS{{cE?*wh+%U{E0=6D`Tiy{>W*G>L0&e=QMj0Mk z$mt&4#x88!s_UU#TRo@!L6k|5+je!$0XdEL?%n&C@TSX8;3Dc-3xhLGezKjlN9WsN z+gZfmFaID=1qwxn*-~{R|L0s)Gk*!CyhWT`?>~MtUcY|*UD+vH_B=l5J;2ydg)l7S zsZ^cNQ>p;Az7?TKy_w(4;kVa4cdhaeO6lNjmXUs#+btd=#D*{|%?AM?ctgS65|*cM zRQT=uIZ&8F9Z3fSMPHy$G?H0FvymOc|32?J2Vv(^T;o3l;qDYlkr+hsm-bL>Ar;X@negrQ$ldM=~zy`zS%Mp2+8Xni z8IXAymln@UP2T_ERp*cU_Y1iY>h+YdIE{r~AxY~x%^~PX-lwSD%*&XJ-^Yjubd9J{ z2P(vJb^tp71agM+-*VnXlx){`RtilG+tRLQmpEMH*mfDcF(BZ{^5_SzhmWnBSXVty zV1nvM@kf-@Lypt7j&mrUBbPgwWiMq2k(BVF50U>h&)F4KMP`O93lX-2LbdQGT}g;A zp{ny`l16o&6~Q@P#4GoiADQ%R9Sl0ItKwzfKYRB#zH&(xne2phEDnEzX72HRtJACG z{9@myFwr^JLvEGy8NO;C^XRPds<_HGofoQjd>RC>5HiL1k47zVDCpY%60N|rviRp_ z?Nhk8vA1?zyWEzv>CAygyGFep);9!qmrNY`@nFO5p40<0WG!6N=~p3{i8C*dRUo)^&q z&=@h$z5B;;&726J57+o{Es`R&exbx1@uPf_?%y;mhR#f+ueQ&VS(tJj3I74Xd5K?DK z;i5lzGL3QnQHF+J%Pt=|xtO-4Yy6t#s)s8t{ZFC5MQ0L7ea3IO9uUyUjot!nB)yYJ zD*@zBr7CHv`OEuj47yX&>g734#&M-Pkpq7mIrIF&%S#^bYN%GB+>k9=OfmeY7gS@y zy=6jt_LyLTZ8P!1GK~b-i7Q!VTMQ@$bh9gAV8iMrIKTT-Mi*Q7X`X8nja(yb}Y5$xO4fzqy$_Um$xXFm@Q*UmC&vpM&TU!F+SjaEWzq~{jCjz{Y5S_p}0!kOW zdJ99$Y)*Th4tIXpu}&c0$KhuoTOCCvExcHPn8Vg~seSFOiz%qSD=li8h6#5)#xeU$ zxEe}u>qd8}o^S;_ynFF0l8FhYCVFj*O+>b?^SkBtFLJO-BqaO>0)#nX9J%KRy04PH z|0={T&JQft?&ygDDA@oC`|V!{?J5Oyr2QQBOv&2r^XnDJ`=(N%B<|}t=win~OC&1TgbuSC9A@<%g#w;erb!F) zqw1{HSS_b}wHi$Dz_w0-N03CVc{{2A+<8>e5AxSjuoeS~jvV>i`zi2-@EZAAaM^nE zl}{y!acE6P`X=Z69ZTP%xI?J0yt*#(%x=`UQKM2A9o?}B=bo%TF|6kV*H7zTBV;XT zepAswU$0RvN#`=Van=18GI@k@kw)P8g(J6h`oyT>bJXJ}ny1~d3F{_YFT>u%emJiC zBZI~n1!gglG?3B)o|rC!8lgrxhm&#+jzItO)Id8Z%*C#tTKOd(0Dc)!R4{9I&xA{l z3Mwp44=^4a$8Ev_)&rh=7;D>PD*^9B7yTiU$hK6BR?03uw9;vkLFpsyVWs(CoKljR zcl+!QOD`|1(m9)K7P;Qv%sGFjbMss0@2v6NQd4@48+vCLtuug?ntx_WQ+!pcA!>T) zj%Hmu;-Z3I_bEr4m_yhef$j%88({TYDz8~w9E9#8?PF^9Jr0~e&!AK%smWi;+U-ga zQ`pggi;axdefkRaJ*22&^xlhI2hv}K#C`A)Z!i37q_^ z@TlvT4pYGb*~Yl%SD`M=rTj`IBrFiVDAfNNdxo)?-DgzQG!+?7&dqPgHR5QX_~vP) zVLDmwxmXDEUVU&s(5eRP*RuP#MIhQ^?KOJqHOeF~2Jwbw%rI^9aFiQB$px<~Fi8DWzt%YYd*;80LO3H;L^X?fY#3;%zl5oe)zEntC++yD@}cfRCpRv6q?@w= zXg=PziQk-@0?K+9?!MCn04PKMCF6jEw{mfDar>Lre=*G85>-mWu9tm}Ey|GO&J@g=oe{bz=2GV9F6gr&u_APl3 z8e)glnO;~_6o|XzH?$|i1)s(N|G@~-YneD(7Sz`MYHR=D!2>8^in!es&DGTPq zV!qt@DCqA7YST?tFT?su(uI;X4AjY6o#=5i;c_B4Z|&e0KJzDjg+o@HyV3V1m=nfH zWnnOZS^??Br+&q2bLtEG6AzmBMX-~|A!4vawO>HnbKSe!t%XPy{vhn$?AeVY+vB5%IC(K-efzrEzEebGb1W9?hv+;jhRCzVPAS?7buzBL zU?i+5lA7pJ4kCW*j5s}i>AJ3V)No=rl|BS`f7=M(;7F0#O7r4kICwt`0OGhS9d6A1 zB3dsG{~1?+5J*a7%A3-HPP?qQ+0e08T=bb$H*B>hRs*e=+$*phzR!b9W3InP zkG=RgXN}iV`oQVy9Mx4YIo_ThaX#a(6_a=F+U3*B!D^kFTI$0!-Z#I0+aiMq z*eW_&zS0@YBh3Zpf`ZSs&t89a@L-ZYRl!B-8Ec3U%WON7Flx;)uAly*6?QV2@XosD z9*#NnCzs7r%{h|_S@kmO6-9a)oS#xOfk5?|D#EhRGEr-IwLIR@fcd*0afNPQ?Qfko zyZlqssE@mC_>!9E-@j#zGt6$OcB|#IvE!LZ5Y|_((7O|cI040cUJzOC{F7~6{3N6> z=R5^o5^x&rTz-Wj8|rRTaY`>5tWCjYN0KdPUhgZ^uGkdBWLN)SfwQw1Q2%QT4^}?V z?D`iBE{p#pkbj2Zz{9rs%BRBhqa=;8v=C(c07G&8(%BvTwqX6J`V-=g-7e`R0+HCt zOR=xgBcrWd-t8j{O)y+P$Oe|pevoN2O0X%G2tt70HgKeLKb_BtWXprbF9J7bM2D^n zJ`##{wL$Bt&am!k+ZSPR6?=yOVq9y-V=gM@`|JPAq;Pv77ckHr7b@k@mQU~U#n!Kx z+c{RR<8hhYKoXEZBh!APnJb%YZ33@~nk`ZyM50B-L(73nI<;x@C&%w1W{wjd{~#8( z>JV`QiuQ`Qp1~?oF$rvx$5n>5E0&J@(z3fYWxQP?>k_3(-6(9)KwyltPh&##DMq%q zl`Q^F5k#Z{mgi1rSMmEXoTD;HqMq4we$fETe$LsH3Wnu5zA3!@QjvoJJW3-{_5yfX znU{dKLC6A4W1{@YCIYZ(Ur%O^jiE9zE3{du_9+} z#Xie7&XzcEF;cTg)FpB;{#YGJ3dE=fJd0E}rqRx9nvtb)u+B2OB3RLCca1`AEzX}6a^ zU@>pu89EH3z}k!Hq(g^QA8J7I+B6E$Q6Z5sgLxNE%RgjSd;NqXdOsbN(Ba{$9&`*V zcLlp(IXlzd`W>ht()^WPb3cs)tRvEfAy(1@I{=rBu#{LDe8U}jM?-7&A}@^uu?ohq zw*;v{M;}?K(XCy(vv{+t@jMj%vwqsmD00op+vQZ_3%Ei8uBcwK!IFBz^@NDNgN8NZ zYG2#2P?z}F*APu*N(ez>)vlD5lql4a-ZUOvBxJ3+5SH@;jvi8=y#1h^XjUUSD))7J z@||07CMQ5|202hdqk{Z>Qc+anB70F;IeAUfRntCwWdo&PJYHv*G_%62(K9DCHFV*d zwS^kexv;xcHx&>F7F{bS!N%C*?+-%vk8zkzA)1A4?<`yIxPL_CfnS{=YZ3YeC=^q$ z5k&zv%y*?8SR$1d-V%4l#2hXjGc7W?pi;z9@vqR=jTYYS9PPre57^>6_V;;6W-#m zzCdwjeuFYs9AfgLXgm4L%j~;%`e-x;$VgA5krnvoaq<%3~;#^QSe<8W) z2BI0+J%(}6Qs}SxWX1t)j_yH8N?EkWxmy_mrRvBV_m6^wyNwxfd#sJ#)DMUU_9i32^>^zi}nyUQaKamd&dlk;Ou?-%Tg8S_=cF4mvU z2>E$5QH*Gu<*}nCGT+Hu71P!oeozeTeAFczThctya|hK#DP|f3JP}f!E~KY?+0yIl z4!4id$mZGw&2wAljy&sUrc&p;x%myJkc)K<9gbK z-H>2CLI&$STggtwH!FS>6zB{}#h2k3yCQ#mPl;;H?vTR!6*)e$V1rgHM)t75A8TrK zwVvU@m0W<)ZFr|6vz<}QD>kuZ=R-ROXcRo#kX_+EyvqbVS^(=Uq=1<~X`rJiSK*2e z#~{+=iiPz?^SK!uj2T8}LO1dpG)0067K;TLeP!j~((c}sG24e3Z^bi;t7QM^P7^AA zFP~^xRl}9#Uf*IqqZEu#eQSEwi5gU`ZuDHPsk-05Gn3R@XQT3 zMtlKbhJ=JDf?&sFudv0AysCS-D+Mm@f(Z5;92Y9IBU-b&hg}z@*d?ye>b+rLT8l8G z`sw1Q#QTF&v$nAKkb5t4 zbHgllkTHgLEUFW9p4h&*h7%%r_2B@bN)S_o3a_k(=-1KWa+B?gFfGyO+2W$Q&3h;@ zaMI@~&u4d@(7H*J>Gq#;E|g`_%B1Z$_Y+4!f_XZXgW~pu#33LV`nF1mkNflt&46?s z>N?Xfg{HHeq*tQm?ek@juZR5nmc{yvBkMD%=dmcGR4m710uxqUD}ag;dR9m<)X5S! z)rD+PlflNZJ~^F-U~mIg1kpStmMDDSvq9hYMh-MPHd#DslX8ycOAi9&gp&5$GXYP| z{*I&p$h@r4&k_Rs0CBAzxiVOU&^4HP${24rL9{Qmh-u#A^%>dXC*IZ3j8h zRl|v5E?CmK*-*(LysE^61TC|E<7mM}vwV>9uxQ$-%ZQI%#=D4IN=H{lxV7V$jcA)Jd(cw&F5q$@I<{h8Gi45?h>vurD@WTn-6 zt|TzS*FTp-{>xT3_K>hLQsDELC-z6sn1Z{ej*CE2k+T*+o>*jmaxbrBj8yXS9LyQq zVbY6UQvqf)%s%Flt06L#OD0Dp+h0f|HReeJ3kB}+aK?0`QJqy6`r7%FueW^TYU}C< zI{3Fcnu5nGp&uK+kk*fn4>M2tdDY8jv75u>ArV-M z5lIUlzGhpK6{LfEnrC-0GihdbvVG%V`{btEO>rThFc!nENML{<>(IV^0Dq>wZ$^yh zg7OPT3y<*TcQFM{4;K}TKQHwPtL{~2mh)~iRUx>nx3n}#>5_b~D&U6S{3f|DqvR|$ zpR{zG=8}|I&HcZF@r0ZeaYRTx+>(1w!<{RV%W9;77D_&YUr#oDC`vzxGz^|6C`IZQU%^pyibVio^$QIVEp-;ai)gc6GDXS=0y#CVTvORX0E8eS4VGd?EKq2^2I z_7Ne1lBJaKKgwMxzj<^)wdkLF&13e(!Gt-gdV`ve4b&#NQvVxMlW8V)9WaEKEbtbY zx)jetz4G`^(JpkqQihqO5n$!XamB$YiV?N4qXW^tQn1vc@gGmp}E}T)ZO^yy6U;i5XmQmo*mK5}j8*3KtU5K<5a37>Dc59q3vqdKud#Bw#jXc}O zXLdr`DwNM|Y>k=lxwsF+L*Fq8#qO<}HhoN8C3YKH2YqCg0ID7_a?s^vH}kq4f7ArV zTZW)XueneR^$rD0Gjs91u#x(eOw;HB6)9>c6@_ZJmg`_8#z*4I0{raK)Kyzsi(fg2 zaUm{mLKAFX2jgo^_#(<#{8HR+WaPEG9(io{37FDgJ&Qx@{ao(SbW1i(Eijz(9coEe zobu5Q+VNjB8ot_KT}xDYeedvO_6JaOM&!c5)Ro9W#Dww{%x?sy^0A5&&F@c7t$H*u zMB0Q%V5sU8$b~mG*OE8Ha?OCP*s9dt6`N@_OaZs|E4lmj=eu!IWrTX;J0Hwoz>%^Z zKUy4lp?X@;T9*gEvhVw+8gH86mUb&%D=&*0yoQB6d zKBo1RMVg+69&1ryWK|vLH*T|$pk@ZGT|iq*Z6SjK1ID^mP+QSP$l0F0?{YkTLo$cH zu>ZbKSUV!>K<(>~>Y!zuuIg^iSrUmC%_{A@heQZo@d^89h#rs~*HH3A45387XHayF z=hnoZ>FajBZ(}tJr)#ssatvFh3&$qBvP5-M4XLcOn)xTRw9`W{Ap?68z@WbPK+#(* zoTkpycCT<4libFI+@z!Ho_A@n4i8he>`W12k+#m5ctIJ6OGMbP$*v?E3F$awhHJVG zj(5CH3*6Fwbzxj$+2sqF{8-HzKfjpuAGH?;+eIvN#6}TNUvb&Y-rr%sc`};sfM_kf z?G!0VN_#Qj%P>CwHql4kFCX2ogaon!d;a&6M&BKCZj@481T#fMD9opxgLB8Z+s-;U zcKwU(fwk4I#@qT$$um0(j{Xt6ux+^7A2&%^gy zp-3vm%BCHScS|xLyIHb>und?`-~zi0;VaJYUa93bXwhGj$g%tb_UzAO3(uAe(eANk2>_n+w~r}QV7lbD2oO`q;s+znY}t6AO*Cl{_q8~5`tcG} zsXZ`~y)f(0wyi%DwiN@cR7cjaVDD(W2OGZ;t2}5B7eLGj$c4 zPI_KJ?e;VEi_u`xF;)&y7S@nQv+i6@{P^an`QTUVi(2M&ywTnv`79CmuBE9#Cou-8IpfbRY_*#F^@_^r2B zPr+N?5dCo%d?X#HNqp4k&Z-3T={cYed`rn zN>26bu6HRjK-o@w6#PKEDc%m|4X(r4N>5RS&FY*spDFr;0Xx(R_Qhw+ri)BD)D`3l zQ1d!6%AH3PnIsF;2 zXvn_oU3!!ir2BNXjAr}?PYssz0)~VULo)Tn)w%RRS<8-U4>=+^Upt>&0lxi zAAfSSmJ)t%q4>VPJo+bL73U)rCP zs?o;=DY|Bix8?14(&PQt z8>P-r?s~*nt`K=Bm-tcj#G3knR<-2JQsw}8DeNcZehNcxZ=8wu_;BgF%{g;6bE>N; z1n6){_YT0gVMP6?W31u;*scg&lxk9Nm;MRK$V1F7@mem>x)iXDp)_MN@%3fs&l7YF)nOXX{#0dACDQhL$xBV^(^*XcUJ^TF};PShu| z$yrIfM$$pezp>rYA;d|AkKWWb5&?FaUh^G+AI7R=n20nSrg}VY`?A#)p<3*2)Sg$I zh`Dfn8cC{N9n^C)8iL}^XevGM=HQG+IEx*#FuF(w2=@2u0o6MrPj4<%CoUzK{29-V zp+FJWmUJmIE}p|w$jY7 z)OlQY?&Dp;CSC{+vp8#p5tgs?DErlM+|X9}aIeOM2%yZUu41{9&SKUj@Y9!HbGmGn z7Fp31Y;*cy#L6sHRYMnnIL$u8T=oghm|@eBes1afB@-^-TQYwL2JHc_o<{-ZNK=P2 zq=O;pFVqFTTRYtxLtp`4lkv& z0%!1Y3{DK6n(FaxGb!mBB{{0ZfO&3=;jjp&1iBz%4=u|KOcHU21W3WOZ|~m23RzsB zcp8fjy%b2{84Sa#e(|U|*obf@<|Ldg!;7xr_5`BJNC2sb0rv4H9oevd)HhNErMst` zdva%u&CY(jB?783^&MT56PAJ^`9`DvD55C7rVLv4N58p!``Un_qb z|NdSdm!eZnjQ9Obxwg4KKK{kL1h4jEdkx|tST!mDBEHCr&f^$c7bRWL6+1RbtO`g* zYu2UDK-vL~zJ9n$^PpIFGMFc>G$Ozw7IhioYIbqHHUxG4`F#6^a`aOR@uZE&E}<*y z0#fhV8+WOIWCD})&vs21`78$=)P9}tW|ql*ZR_dRn{e5HxfwlN-|dhB5eDJT7kEJ# z1MUKYELsvghxSwJc$?K23qyF}C77%V50>&B_(#}1tO5T6s&Rtty|M{UwD`J&$R>?y zl}!6IZq%qnq{~wR;rC9Y8-m{7IOw>sJI9) zh|gI4=|tKu`hmdUoR;1j#5-T68RV--DNVRWxpQ_KdG?F zd7K1l!4EPi2&}J@y^($sd{;7^SY$^Tcf|B9i;xZxVPB=0nrTCd#FN#QEiM@|-cYij z6eY0Fip`giW!BQMCSB^#r@c$=-f~&K{7({q^z)|u+FC8r8gbKzjOx{XsWHvuiy7#C zWpuCcgKYA6b>gIjR2^g%nH+gNtIXh|#plTS2Ck9F`j6o@beRvWtjvBuX1T~Z3fG+$ zSG8Hce!)(OOMtw(FJkf%m$E`Jnn}YMUzl-6NmG{;RIK|f25Ybywj%ZJziF}gCq1Jb zvP^<8wXmGyu(%J1+L)Js0%e6LMIDf>1*C)PeBbC6To0oPU4Z<=MDf0Oi^}ePq;0I4 zDa6uvBQcEsbSO>NmRTE2mQE2w=9AmGk5s1Uis{#FmKNS_pjOdenemE*wKZD57z!a1 z>w`4@D`#Bo`<-Ga5XC@gw99&<^s(g3{IRe)`(Cgq>!|-xP{WeuoPY&Z$ibIR6j)H5 z@;rGH_r{5zoOdKzE3*X?TLjeN<}IkalN}XHh$?*&dY0XG1*K}{jL}V2zcHDBxTg#q za7V){4-TItKl_X3jQUv=RQ)1Z5&PpdH%6nZIxLoPO=wKoc-%MTY8hN%RN;*pJ>!hKz&Iur6#-Q@#eiFSBVO#jtq z?RJT~{8T^Nx>#0)OnScsW!&$L8#f*RH8XjvU!a4XL<9xdVU%PNMnV<3JDBe{j>SrR z?m^LpIU7pI#%qO#V-(5eoZ~*oTlHEgHdyBS%szi$HOcp~NtBWBwbcACcc#?QgUo8m z-x4G?T<&|XXe8RB93KRTVyw%Rzpoaj)J~7+CyOjhRU(6+|MZQy=aRF;g_V$*>(%^u z8DAu9rJIY<%Y#|`ES-Md2-XIt#z>RLnT4$F0#RK_F}~%YVsrFPIG}!(P;#zx%m(W+ zC|IF-7`(q%M*@v=4V{tKB!2}a)0Q(Hvvv#PFXf~r5p}l zal27i0ansGohWm_2Mdva9uxdgoRi&vJ48dsiKvDTP-Mk}D0yG#wE&HVFO-RB>@>Lr z^VMdmM*HVRT27hb4vz5WMO`G#U7X|pJLN!#Tvsm5-olM?7H95j8KRzFB|nW5q=oho z87%jsif>RM?V2P+j_3#!oRj#{uvLgj)KzNjGh!@NC2g}&inLd2 z3a`b;*-7ks1E)_pkr8uah~QO{6v@YyncpeRRH8U;t&Hy^ZQ}(7OtE@l%So-9EkBbt zRXdxks?eq_EF){L#&A1)%u)$)Gwc1XA&e2`2^lRSMgt=U2M3rO5Q}t@!`vOg+SiBP zLPjKE${_@xBF&(RcqkWoM{0~ZWmxYH)-p_&%C&Msfs>l~!=?XtEIJU4bvRE7PL$Xq zuWAZ1Q~9f@5VYyQ4oIilIS!ya=&TdJj^zRHtWx*?R61=%5{XA1u*8xQ5)1RC*o!XbM;}n z&)_iXCKNK~-+@+qxvjw9Eu(xK<6Ci#dnU_-G=A~Un>L;rQ}i4Snclry&IUPdK)-(d z(kf2*oJ!%-giV6~TJ>xzHM7PPj7*}-6jUP>9jFB1f%H4i1GtDc@!k&M}~i>o;%7Fp!AZxhzPc`{78nq&?ZBEn0L zMCl2Em3;0o_Tv?d%`(BFn(`|cAH<382tYmetLut?k(y-${=Kk+26hJ9!cL`l_H?(^pn+zE^JHI~I}GfYem}ZR zbiJW?9zvaX+t_r2(kdv#P9=3-^gM1$gqcf;ri|MnHdr>fj9U^fj9dWVkBZJD|EY|& zF`Jar??bEP8@f?s} zlqg4;h(9{7P36>f)~$wuXHLiDQ{y%dH#03KHGu;~AmE0MM%2;>J}#U1hR^!VV9d0QTRJvB+CuJ8(w&I`PphF5!4 zyAL*ty@02xm|x4_35fMU&VG%5{uwpyH>vITkQxbhKgNowyU`ANRr_pTvWW{pQ5Vvz zMb;~39(YVGQpja8GMJ{`(^jQlj7dW*teb#l&h}4&P^6s3{3JoafH)pR&Rqj0*%)Oa z@JcX(sPDwTOevsO88pc)eq#rRBja}0GW7d}Ei;Pd!GhoGx(M+-gQv3N2@OekdX2O2 z<^&=C*JC~m9~jz`cA6bfmVhUL^vMd%+_pU;=6$F8d<@+L6Yv&u3(hoQLa#phL}y^{HvqqQQw2da6S*)y|`vm=`9*ISb^i4QgsfTrvI z4Y%fW)s5Oh-zfw?8ryA`e#xAtw=Kfl5oF+0;M64vpBE0nbRM$`y z$4H>mmYW8c@zLuB;17oek@=9Ct1k7OEf3UJ!e0jSMA($3Muj_5C|`qFF)N6{5|?dk z4+gPI;D_`GWF@TNMrCqAHLpV(L$H{L(Rhw%_Ns=RlB;emb%C%LUP!G4)c-qYMLcX5 zI3UPh3D9mLl{ISIQrF>ykrd-L>J)R0oHANfXfecsZj zo$Z4IYWy{)eYEQf1S33ie~sm;Cgc0$Z{}4GsDIU_h%@~#fPwrUScT)&E2ip_lNY9N z--atmSW;ZcQ)5AWJu|6_E+QnPuk*#TwOLHjmH!keqg-pRO7#GPB<&{MJ+;i5Y=s`P zdUh=+;K-C&Wf3V-RSAFf&Tcfhi>AhaQLAg5d+s>dmaC{s}f!C;G9l zt}LCf{Ib>H4v5Gy_+I)-xwO8MvQjJTc}1J(4R6`SRzN@XXa>PKMMvuX&!EbGi~FE4 z69))k{bJhEubegM4^{|oRkxwsHg$);yBzKpYHCSN!jIy-Oen)9D>@Z-VsNw7iR3Dfnsq=+j$2+K6>0}q0 zT^`Mw5_2D8A}FN=7o@jGh7 z-Urf)}}Oo}7!(O5-!Zou(LgSQXCFHlTQ@=~BbhZTN<2GFIB6nC8}HS>d(mTh@=r;Vq0 zM@;f0Bd*)^Xz>H8#K0IRhk$DlYd2WG=X^6LVZN%G*y&sTKKbj!El%An?Q6ep_`3P} z;$g?$Sbu1CH1+pio`%i-FsS#@Iof+aoN+C;c%Bt%r<{I$VAI2zGZ*XhYGHZc#OSiO zp0!&iJY2N8^Wwsq$SYSEs}gp~t)lA2+b^4*f3vZL%W;OQ+l(Ydb8}(mF}FAcB!br1 z-lASRnvwkn4=W54M#%Lt>D1!U>AGn1SBI{Bt4I%O2btOz3Z=0svm7VJ3Vi)2R}4vGhTCWNFoiWjVAeyv^hR*^Tw+W^PeB1f$=zIX>S@T)f7RA9#fal zWT9r_R^hv6&-5kdp61X2pmI+r7^bx+;qKk{bR>o{m9DDJ+Mqj`ok23v&!({}g^Jdu z;?QZM4I4HT-t>65@isxz6Ag63--gLmA<)fTHM?6dqP2!ZAx|%--ZL3 zW{HlyOxL|T*X3Ml^Eu<8UDvbQefI3h@x(QFo2uNlbkW&UmrP@5Pe|%DbZDFH)Ll9j z_YTc(?f;$NB_rGx=oj;t8#T`!dw4{<9okO-5wp51%M+5}Jp)7@qk)a>sQ>m^_yD-w zGuV=59Ut0Y`WrXn%Y%FO{&qn`=JAZ!Nt6KU4dpWdid)UIIz(Y?~aPPr`h8z5@KQ_!_w4+i+ ze=#!OxV=6_YOSaCc0QKJNPNFcuXg%=(ib+ln4y;W zv;p?zk0;i!_^&=g%};`%uKhS~_V}OI=Pt;lNh7y~KO+UlWlWYA@G+SEIuxhByg5L> zU1?W>@Q6QNG;P|{GDfaC*qOoyGeDjJM#w^L$F8a7;#&GxHProEI zQ|YKL<44+NAZ6jomJb6F*xB}XZ{^)_b;P29%F>NLK5Ou_rDZ)v=~-pZIx)P#Uh@Lz zq`O5`=hpnXss8tk&J!{YbkW^M5P3}3vu`thqQOdY^Bn`Me|*BXrCoG8qJ%MYLUY1| z)0E38*rRAE`}r1JYDrF%4!#j(=4jjY`V;9C{qbh}(X?GT$St~$d(t9u?Jp-~QCh`t>=oma9xC#I=SFJ%rGnM&U>o^zQb4f!y`v zN^9c?yS%QqS{(SJelIx%Kgce4U~Nu+__yc9&qq$ES_U*MKwYeT%ggZncxIbZ+qO~O zxD-4gK^Po6YQxL_;88eBL95-&U!%N$t;?VWkEat!wDqeWwZC&+*0=q4JL_J~u!^?| z=eD|zRdEnrh*U|Bv=U@D(V6bkt6G}fWXG*tLOMU8Yu#~S@KFk;VbF_$%-ZuP$?3-i zo8BjE9mp=fyE2rc+o67kyzD`PdYtbV>?oODjoi`KsuMTmQCLya!FC_ur;<3**F7QgHaQT zs@nGG;WqD7K{J02xZ)Ae@F9pS6riu6-mcwb_K;C~eZ6NXXyL!p9_!0uHc)EC45&Xu zPROiDenR8JjxV!aZ=UK^&3A08p<$i*{_f~2n}*1a_CeylieIW{=0bCcrU77)(tDccMk5x;)z(E&K#y>VDemRehkr z@6f~GfYuP{O((5c(=RkMG+?&hX-eri0n-ns*9~W#D^_r*ED?$M&}M;YN=&5pjt$b_ zv`P2nSK_05Z4$B<5ZyO=7cJ3F*1-(@rJebfZA1YNlwA%%qlVXECY!B6Qf?L@BK3FJ zN03Q7XtzJutVLD%>qo60b4q?bez`dmX6^l);Nx3#9d(E#)o|}0AEw1DMQ)Jqi}!TL z%o^v70~h_ZowaATz%8kX!JuV0@h-_ncmH31{iO)!J?=4ij>9fkGPMe`By&AmsV#}HO?GeE&i1pM>e%I%EF|C$L<1P-ikj`$6 zkB{2WTMDSm@t>DDK?t?zJajt?-B1zr98KXp<<~qk>Pa?H>}s24fAVeuHUME~k>ntY z!rZ=Zv(K_;%YcnHlYo#?yHGcc^i(Q)ojTdZh1XI1G)#-?LI&tfS{Z4>a?2jY%hRAu z*RHcnOuF)9%c+XSeRyK7?K7VPyKy4;5eN7#{9VV{sP7OJwKh)#l5iPmyAB=V!SkGe zS*ORHU@dYSiX!Wyt?}v>zmt7@WcyQr401mZL<6oS9 z15Whp)oTX~!Z2FB2xoq8?9Vg1hN9DxmTH+LSoyP_GkbP7J)YKAEDG~ z-#f`!xY5i%*pDBec+A#0PKYH~px%GuQ23fZx!WoTjGH)&xXboD$qt{#k@K}R_CmWV z56^NgyBxv5Ux7&PfSNjx6mSj+Y>E0P9M!s?$(JW2gz=^a=ojO(qfh@RpNMDo(NL~Vr7rjzcTs~T&6@QlNwU4e zxFV_h*6O(iuASN*9`TkvG0-gbIEmvgcxq-o?>28%+-&h>==Er`O`QUdG_-=1dCAE zdV7;eEa|q5+`WP5wsDzz3QD0N_6=uB@dci2Q8jSnDhi(#JtmyQ^i3;9VO1@#%hE4) z*wnn@vPPe%z!5lh*VbmM>${P*)rT%uql+}EwlOFoDxvFmmm?U9pLgn2&DTu>O6W-1 zQFW|GUaD#F2?;uQ5txt>kC8fb(ZElx6EO2`W5<)VN8@mYO3+W`4~sKHj={b9455J-9v8c|s38+paw5D~4n8s_ob3bRMVoZDgZ| zmlqCTCt=KE=HDd#yj@9mRm}h5>P^6M&cn98+xxlh=6T-t|92eE@ytBWsQdo?zTay(&+|I3 z_CEpTR%YFJKV;!V9>gLCeUsw4x`(3PiYsxN!A>prx~dg?oEdI0lKnza>l6Kx0m(9B zJdFM~wa^k^V6u%!H=Ulx@mj`#+l!jr;qb5wAQ)t_F=8L%#fy8X@7qpj@@M(ij{oP@ zQm-s0B^fXZa_&oOVsK6n#g!f95ukJUg>l6RwbT{?r2BGF3VzDy&Hb^Y&q_I}#{g=T zH|jd7`c&fpBP(_6^Aq#SG`bFI3Tod345;yRBbg0t)^JIQw8Meo7P#4vwl}5Kb}*50 zl{_y&Mfdwc`RyNX3_j0I>Xu*TqC~cSgk<%;g^FXT_4V8LC@Lk#a!WjQrl`QE%6eNd{AQGL(YbgVL8Y)XTgwhcjR1!nY{2 zDP?%!)8L9KH>4kqi#K!J3V-{g`|#zu9*cP{ctds#%U-)^sPr138aZFJC9h1}~8kM)#v-N5U2)O{LM#fqVXc#^w?U;3rY`f z4yN*<4`l^s=Uh>jQk3R`01TS!y2h*VuN~wQN2CyJxN)r*PSvT-YIJmRfA};j1W0?j z>-j;p9A5i9^`fdJCOV|r6gmi-V545WlsuVXd4_yLySP)AE<4hStQcO1LZ39pJB_Vg zNt9Xg<9(|a0rTMv(+}A|BWCwUB%T`5sU?%LLn*<(mF5PDWWV>#kHc*?d-7?^A*2r{ zcDZS247^1hHTQbyllt4d&^M4sMVP`;#%k)Vz%5Ae(?_NjpTS#YB>NX$b7rmeKR84T zSQIrHd`1t5aVbw)*3{Cv_CPYY>{HHMi&{*!UXvx7n!Wh!d7%_ZSu;r8ROlKl4~i2L!o|qUE2~{{9JcT9n&~2P|97IM-V7ep}2O1}Tj-Mh7&Ck{6Z1jvWN zxsRUicOI+l`{>KJ+TXs`?&9nXygmjwg~+wE8YUsMFd6c*{N_F!5& z4Rat^_{pLPcKg9kc0=tk*dlzt%v(nuQ`>#Z2#pOnY&!(6^0~0~KSAW{!AQoX*T^`xV}4`7JzCmey>PDUA!bVR#oNUpy}vpfG)*{E^vx4K^yLmu@z zSd-CveSNR2OWu!V`5IL{VG5CP_&@FK zzpR}%bp!RhBtW(hD((DTcVFTb3J2%qTPNy=qYBNj)n$ZcDUnKYkXgX`lWjKAa#5@m zEo^U6&*3h9aY+qnZWpe!r;P*OHaF6FMvVJr#O$p!ZUup#tdM;uzy8{9)gYf+T$%vD z!pGS417sf3uzE+;Lo7QiQvv6?nbgBt;WUT85+DB5ERXuyu&yTie0(OgGfBgzva1Oh zaUxw(iPJO0>6|hD^S!^3W1c`uF~~B|(bp-xKpu-C13T=CNbfW1x8HV=8LlzDxn>m3 zRNIl~0FpP6b+Q+muRG*X$lHK9b$oO0SN;Pe!My+hGR7vp>+wUXWUBs)i|zz~n(k*> zjq5Ug>eM#qxpi~zN)Z~vHp}LuqCUNQH!?e#f^}g_6cA?2@9E98U4IdZjVFo95~}&iHDANZ*pj_!VaRd zy`1%V?TrgshjtTZp3yL}=wZ`@X0?P0_nR@&&0XI##5)=aNGNP$9U6iiiW^n2gShtP z#R9t}H|N`}dW(wVVM_8RmF>zQX^GO6N$Go{9G@I7>B1Rs5S|*l&9#CUnagZ^H0l}( zaF5)*N*ki(c>F~rTS?WB+9L(oaw?a+dF$49Y!=wwxP<%IYh<#E={IT!cyqiT8!_ZH zGp~8Ajfdi7h4u9={p?iYUFz|f$LMz!p$w^8iEf$>^qm8CZWF;hH>B?Tth=&Q{O+ zav3q-_mowNX@18>h&1qq`r~cb2m| z4EPBD2KR{-ZmusaB&KD_DWN88~qvY(MY2@%Ctu3?P>Ok0V-%MtMeK^lg2H;PzQi=UhK`)#6l@g0wJ&Wd`){9X}-q6W@p!<-Fcp4<<= z@O_J=;k8=R&F?@uq)LUJdCEEFkxw!)`P=m}wbBT5li~pOg#(&Y&vMNEG#CZc2fe^y z!Y#BuIq7%Wy_NWLi28`mVJLX*^3uy2$JXWlz9afl6^qe4x?6vqRio6#QtmxO9{e{g=+e+ketv9GIpy-qQ2WWN)>V^x8Xp?- zF?9LE1jgQfVX!suTEK-TzA-em0?JD8!YDj5(=KNs&#=mMcl?WtTZ{Ln)_^9d#L5{U{kJRckhs}28gb$+nSL#K!0IoeLmh;Q@gVHcKHOb7(?7H*9(D|1@p5X11Kn3&wK(JcuMFb|k z{sWM`p8<;z%aY(Cif-~Kf?+7y^yrZa1h_ZuO)tm*X<)*>9p!3N4C?ExVsWkgZ9cZ& zo%Xlu!)rlalu{rc=jXj&|C9sax$JX8A6T`|S4^!S6b^!tvDjotsA&@h)_&u44jdDv z_;Lx8i%lh=1@>O=$rJ7WUU9geFY$itHOR?C*QD6^y0b})fp=+4``aFq>L`+xG&VIn zt?}9qC2}t%{VavV`lU&e#yfkwp*}mtZ`0+oJsF724;U!-)QC#zu3IWJTS}U#sp$y1 z8Wab~;3p|xkix~ku|FL0{j&@5a~6J!8O_Fu(zTP5CMK(_=1Z^or#^NgvL3WADj2Kw z5(!6E9c86*@`Bj-m1lY)#6iLJJGB9jVD{SV=D;1}FPBOM8Kd^_o&NFaE8RUV=nIIO za{*tjDahq3*;4Wn51UCLMWkSmx|XpQvzf~>;oMV=XIxCHb)JOxO?N}L!_t15tbT3kUln1EjHK8mZayPYnDoz{Mz)@mW)vK$wAC)uu>$!K;FNm-?@~XWSvjH# z+#bsEHwC)gRDs)n%{xbj2fcZJ>9t{y77TL61*#lZi#({dc9h1C{1wT5?{4d#==zyX z#)rr}^|FN?jY%TZvFH6^k!?3R612x|Iw&0&sN+?}dNqo|UsVk~Q|CgIT|J`#8e97O zD#~;#mbOt3QwQwWx$`&=aZOIRFUoefimt{nN9-FQHs@dk(pI73Gg{tvy!0466tDX` z*P5=|JgE0KEYTYuRxVZ|p)A8Uy9UNnNx&3J!>L>nX*PMoO3hley8q5WI_Zsf4A2;U z`XvNxmOI7n)ANQ{PTTnkK+l$jLbVj9~r~t=O@_td@ z*K~LM-w-I)p}za~$MO~{(TimQ(oAxfT?P(gokw_MdH5x}+FruUF}7ad1=C35hjV`o z4M@On`!|gR`~{i^N1_%{&aqI+9UZ24*3K`2S{x5sn|+X)E$2s3dycQ=T(|Q(go4}1 z6v7!w3*)6PQw}GAajnb=A5KRVXLAz+@s!6|-*2qWc$D*gFs=QWXU&b-S3xNwut?&^ z@9a=sUh1R{@?c`pc>m~~Ab1~y!79|-Y#0*p@B!epG=X> z3Y>-;06hpo0K%`EVNWtLKHDPwAY6yw;SbK~mTsQ5^6jW)X`I34W6R zHT*YdQJylvYeE;qNORPpp{Py=>Y9s88lH=I3)5k=P*w2tj@0^4duShU^33(;ndo6i(S?L^fCsL8{@^+8k*^a1{8 z+BGj2y{Zp0rjtB^>8$HtU~FY&WlCIcRkU^`KMaM49DDEVId*HvD}ag~S^c24%FPhf z$Pr-YA(qdI%C9l#jPT+CkIC8f`;Hnl z%G}7ELx2lKKe{>HqSV~8`*O}EFVyNx>b3%tD70%)AK*q8Ew?hPwo0hUzw|=hJ}X+4PgtGHyu}0JdNNvDs?x%}gQ&>+YieRkz2({aWyB^W zU7e&$K?>4YrT+WmVoXmiJ_{jMZ%vypz94!wIEWBWfL|%M(u-9su}9tIWCFi@005E= zpB;eNsK~5CuVFFFDWBlxX5bXr!|dpObv3nzID^Ci)UMr@rmU=-y71YZMMF^l7DN;U zTVeDn9W;Ms8h2Df|BB$gxV@0G{#DeMYaR5RO(=#9ME+crefRKqOGtnDSQ@F3-0JPB zJ(~zY2AGI04)E#()85vW5qB#qz<=iPLx)%z>FFPr#Wd!M+~oWGpQAyK@mU+Ue*L#0 zCEj?=r)b~&1sGrcOtI4JdB}b*)#T|-gU+5T?JqgZ*yE;

?p8d^1+xWZ;nt=?+?$uz$2V5-u;^BCb-;b^hip8%)`dXcuI61@6z%rVh$KQrQ`Tz z%L<0%C4M(4HVoNEl$D2#w13^h3}KHiOq|@T7Pv6teVdIWsCALKd_9oWWm?Nl3Yb-p zFgQ9e=(MzZ&iBEdSHafpfBD|_;(mU||0hoSdB?Rj3{fR6W_(L9dJOd`;+(RnygI7YjSemtEJZtXE1=>2%zL&Y2Aw43`6vG8X{6) z9Ys0%enb^1xi2hPIKr>@53uBvm<25Twk0o#x+pa_oyWC1?oDioQ+~OE^d_~P#?W5i znR!LudWK=ZZRivkL^M}EpUk~3Vavl(0n)o_Z-c^UJj^xzI(q7-GW5)>`g?Tq^ z{>hwUo3guQixx{c&*WgOSx%Z<08X|K6(G!UH^<)Hc(ib{M*#hXNcUghZuLA~_xwKp z1^(!N^7iyJ3PwPTl7WvmtSA}iJdiB>6g-T6X8*+q2M}t92>cDk02(bNgEpXa1=E$N z&x)qmg&&GMH8>l=63`jcib zC4Dg(y*L^cT6nW~I}?7`D<>R5gi-jvLcNBFtd5*93fFIyO zEVG-~3{y|5oAcL=2@>F(FJpL9UBp1a&aT;NQ2Fq_Aa+c7utzJ&zkgcC#D!OGMd$YZ zhFqZxeR1P%GX<$Gnmw-AG~0+AOqmc5P@QVH@#I>JhNx8?r$;GND{L?tj|~D2uA^XW zn}nZG&8zfj)3vKP28rNuQ4upHqMs$$F1g{=;Dx(`}ZrwBARAHNk{v0wNa3uT?R4pwgeUH zbFhS2rK!!1OXj;c|7HtM;g-o3ud{YPmeNM&P(}p%gYM^-*+X4YDVdRQ5l$35Jd{@R z*Pr8xrRm}3p6g^%5wk+j&OXU!l;nbq{p}_VP0iuw2w-*PZ#nGn7YKgKZ-niEGGP;NEEhJ8?^n13i) zee;~Dz6bMwQnvwq+3BIQy*Y9k$>8bv=P#QJB_Nn1@;IJp%G3abbm=XdSeviqbDUt> z6#FJGI;|r`;V(Wb5uj*D2QWN?)}XN9HnjzX@48e&55C+D8=jlMZG6Z!?c2xk(o=5h zA;?-0?7EHVt{rgTa;OGJmZ88KxK|l0IPhbrP*=^0WCsFsTymo0RH7&5?lI> z+t$i_FAKL%Ffee@BI4<%L98MHn!&B@2&5#7Jg1+}In&mJ9f~@bYA&IdEgXBkmeR=F zhJynuwFN}Uo-LmI{y#Y1TsloYm7sG1gcHX70CB~`Y1TM40zJ=ZK32!1Sc1xU+u#N$ z_38qw)Ix5~9Xw~q9T)Qf+EL4c z9Zmuh7s`rjp`BNrcMW2W9qY-5hxY}Fm$1ko-v@#5I!ieRpjDFFPA%2r)l(Ktv~8A4 zO$xlr>rdpXQ`?yMRg|;WnQ~hMRCIL1h@vrXHh0yj0=phSdyVnmNXwpZU3i#hO)|X3 zS+9*Raca-y1xA>AYfV>q|C{(~A!*y0m30v049hl6n#I#JddasR`Ir<#n;wS;NmdaM zq8WyoLCmzU-$E1+{0nqVduX@#qAOtlCv$?cdau8d@lb}yeqFAUH;JIsj`N0Pj6lxw z_??tZqw*qWD|YG8Zv4)qj|)cGezV4@=^w8^5c=}vOMH+yxFg0D)A!h1kXP9ES4D-o z?p#}wm|1mOtEH}lzHDSax;qRpPvsA=dl=~2wGq*L&pY>u4M+(Rh@riXc;KQ4s*yy0 zI+5?#Tty}NZRN+UkEuz|$xzd_$N23ggZ-}LxE_;Qa0iF3CD(Qve)GF{-$&ZNlkJ^5 zr3!>qevxk)*!nS2XTfCX$Gp_GKc*aJ! zDuLV+`+z5KQc8p^$kBXokOK1FDXvgh$!#7%MS!e`kpX~D>j~t?5d(@E4NPaDbDQd~ zC|H(-J#Tf~(*09OT@R5kT-mW}7k}?>Xtgi=B$2LUt{ENlXtlaktIKNb-ze>3WKkr! zOiz2Q4!fz4H|7|g>u3@aT=1eWA)>A_UsVbQG^)$A=Fefk9!HU**SAY&uNnki!+90j zk^Nhn((yRG>!tDg`Ynn1joMYe$1~&)rUL9-7B=W&6Qv?X$ddXqk}pa;w9s|44fhwR zYFqP*$rS7N0N}37^H|4+&lnw(j_UQ#l1d3J|62{tT6>H0 zxSVF5rmm0$k#A9}Ds^CV2QqIz3Xf%JD;f~BM<8fRXqh-iUv>o;K=23_5(Sa8?7@NN zgc_pYyc;XqLLx~0L0<|G#;rk_iu!|U=?bdjH)@M48@~P@jYJq;o#er$jJ;FL%ofwi zixyfcQYn3CW^r8*K^}eTxx*a&vT}~aS43ZjJ%l5_Xc-fXZsgS2)R_vzFNR7NG1TDs zH+HCx-FTIr*zvp5?syY!wTN6vr<&lUbGqaGXV03VM*u;dr+Tl)iG=Q;%V_)88$S2e z>-~*84ItKVnOiJWH<4_aYun3J|9D=D`eO?asC5~$et`YCD|fh@Qq4k@6=5fRN_SxU-ZZJgDeL_J7lj82mazgH*Sx~RM|*0tPq)P8NS0lo zT6?1Q>LJY;yHSsEYu(`6Ab$IPk7>r^+O$K_Cn+-`Mj88>>J1{H9jfl^l7F#QS^xRf z*!c37HRFl}VW;n$)B5-yVCrctZTrSSJ=*J zR{J&m70`}IL;0KU_#I9Cwxw~bCR1v4aB9y9PW4!vdzCef)Rp64rAGz&P?bC(+6T_> zWvVzA*=Z}s@k6jj>m)9=dPjyp1V^#88#Vi3fZI@~QFWj5qA|bnB-K4*Rg_X_!YIxA zvcwG$(AV1)4t&)Afu1<|qQ4CLJWJ53AGs`nAL)?ih(7 z?41NoXZ@3!%#}8#&s)YfweHcqT{{yHj^ex+Bi^mHk;s%OWIipcL*X-=c>;LFzaelK z!*2Nq^R&OI0orxyElWT`cJEZ=`$W)B#SkCnqgdoRV6%MUQ zyUx}N?`+f|!^fOkt-#40s=ID*TEBiI6**+T$usI(B zK99zOe^1l2TsVT1`I z?vO2F`O>1Q?@v;s=xuarw42J#5<8krRItssdI|3wmV{;V;E*L1`HqTXPh;Z6AfI~iVy+!B^K2Wy+BfJzV*0qLh_SuetKHKc z8nK~7lxJdmrxU#{Sd8Qq1a9C|nE3zj8WCbzXoY%ghKl)C{A7!efDdj04bZNUX?a+I zgYFfHW|u_NpdGP(ULZ^+;3=_;Pfyd&jMc-V#kPNq)ZJSaQfzx8Qh24Omw)bCp1~`f z_r}o^e(EV)ydc1Ea^7rwOOpkF3hZoBdC$J=71I*Ji^47?NdFMXMWdZ-?fFmx=ieBh z8@WAzN>0zwn%FV{U(b}-K!g`O)R!;k;v4&bHfqMlG7{p`4&a$YOmiFYuI?X+O+typ z9WsRFI$KYK*}w)&baGcrhJkT>&oX;87UT|!%jmOVh{M|zhnMHIxM zgc6R6niv3RT)sReB4G0EGY`iV!?e(;!G#IDx+ksufDS_@H^1&oKJfL*zqPA!d(&5CfhzQ;N||gZjvE@mk=Ws^Zc=;8!(|_ z8M|QdmEm#y0c)+$CuJwOpRq$4Jl=L6TZkGzzPwN$;E73X&a<|+iYTB%wd*$aP(nTj zswqN&Uc(Qd^0SZ3g=G-Sfnc|<<|6Xp#*iVyMjFFpJ6zFj@@Lfh9Z{JN&y%d5Tj*97DU0gq$i z@5Ees1XbWw<(dAxHJPOvyzq911~%-Cy}%s&SnDSDFd@;K?LLTBk+-*lhmZ32O0)k- zxeKEIm1lF^y>=rxqOry!yN6^Os=^^MwBR(!(EfGo>b?GXYbFKj+AR{M7}X?P!$_Gk zQ~c%}l};how(=3cw8gKRj!h}0%@~bN9~o7$=5Nfs^g_+)7wQwSN#rsLl++;GhQJsH z4A)#$wLg!z#SH+WnixTDP*A{zA`GTlB+OFb6#%p-9T>9UpeRw(JK3TWT1+*I25GZl zPY(6W^D~oM9bbJH$((P&&OLhyNPO?0N3AO_a`6S$C4s;CQu$mZq*~jiiZUD3lj_Q; z;B^`|Kd}wH)$wXKL)^Xf^rEpXWdKZ>#WTJll*cE27-kb9oB9!|R-}S_95_?hBB!wj z<;uoE=&7kqq?E~wa@FR;kUu1WO2eE@ z@JQ1`Q*F`qhYiFcm@6l=y}J5dM*5hchEClEFY^V@NMFt$Ft2EG`qW>6S7WMuYh|U2 zqrKW|m|NqwHlKt-`t7$Cp3mp@eDt#UW-eL5qad|I54S=+^7ZtOJHLOr75MfyH!uvp zzvwdWCnPjJd_0skRYTwH6%Q)Ao#FQTdF>5D(pt%G;nwFlUZ>hw_5X%vSij7ht!W|u zCqZg{w1DmJD>wvgm=GHj1F(Yo3$*tcd~F7|8`|(;)4)^KkI{3h9OfDEcq3`)bHEq9 z#TY41ST4D01nMggsjdbt!@vakA?yxGTX8E4a^Olv%UH15M!bukUxOYG!ndx;D`-2q zC?+GOLavodQRcALbdtJU@$<=QBUjpIu;C2Nvx1KJG0lWTrqMb)z0K{i+qt$Ya>CaM zh7AfXgC}@@ubO;d#(sfA{3vV%F8(hsXImZ_m7%T7fpKGI2KmkT?`f)KcxUq<|V~-AS)=79&6gJ{)@X;Sx^>c@`=+8sHEx=Jm z#lYsri%@$)I|W8Y;r7%pXVn4TgD&tIB{23T>mqxa^SOc@04e~w$DwtvLeih<`Elrn zfNt$`fks?Yf6cjKYW0qXuygZsB0N7(n}HcT3v5{s?$~s_=aRT>WdQ4EvTj}hvQ`YN z^;*|EFj0Ne;PlG_d3Z+pE?;^{oV|Y7h8y9jmQfHkX0kgq_{Y8RmXk&Dm{f8yb0xoH z)%}|$04+N4y}!}N3+4Lyr+*1B(9_$fSbTXT3-JQA+^~UOpgX60)|E{++iaxGfv4r9 zw&<+2X7W?X$``booMWSblEw?@=|j{Kl7?PqdD&S_t+jR{{C|XRjl+&4Rc2)#Op!_6 zf$MIzxjpS!e_iiI@NBO(ejI|e|0GAR(HtN88ELzi0Buc>tRJdQTT>7W*S{4h5Vkft zm9mdI&dOHD&7Fcoq?a-+;waz;vSuZ2W>o2L(@E~y`32jMeo3<)%CyMGCb>M6f|%oiW1Dqz zkPm*J5ALp4a+$BF7{9v$0^1%UWBHJ+ic>Uwj z3yWw_SbIdBPuXfL(UI*`e?pno?M`e%>r>Q4=Q>lN{ewQ|#q|abBjQ)S9Id>OH8YF`E|cXDG-b5M6B;`HzdE7jHn5Rn;#1O9*N`;l z36Vkbw|t<7U&^6W611uR;2hE?XFky=n6mX_9D%DG^bQ9V23aG#I9bNCOJ9Id3N>jJf*2t)_K49>cG45d&{_OM9h8|YJ zAIF>rp9x6#!SJwj(>u&zH%P2d1mPk$ z?ls)a*88=b#h=vt6X^SBpsbD8K&#Cid{DzbcZSu}+%Dm`sSo13C$l=3> zv&%iXbAD^Ub<@x{G9}i?)*usRL*o+Ndk^YM1Tc;c)ffpF902Q(Nl>;wo&`-8#U{gwL0if*i+W$V`Q)}j5* zB_j)H%Bt|RK{j8R#6S_5rxZ3@{H#Sy8gS(^%pQAR{;i7>QuF#bKj9)y$`lqh#a&G* zJ`dnT`vy95>Do5QWAqE8s(`~_{=LmcKm}ThMzQRoQl!Afss>+_6coa8m5VG9%F7x4 ztJb9TG5`F&!RqS*W|I)Tsc?4av6N7VBEdGJ^sy_lg4r?~K}&7>JTn>cqG9}K6p?7| z+%6)iu|Jnt3zp0$Mlr4d4{;nh_A81QWDX{T?@Y9g=AW_jY5s!M~L6r@6w;Gz zy)qx#?7D@ziEgJuTMld&f}r37_eEv<7KHaohf6Lc=D!JYezk;SsKi*T?kpPz078lw zS>4Kr3Kc885$u}*(L5$-A0OCZ%GWBljdw>}noRTZEk6sgXTp{|I2`*%i z(0J)2_d#zQ!)N-F9;(sQQ$E`qo+b)45c|>B!!g2c7F0>Vy~yh z?nLPZ2vm7r$2-SCxQPdY3_>rxhN=Ufwsy>pn>~?fqk@=g*eD}yqKi?Y*5P+NeAoCG zy$~08H=TUmo5&7?=1mK))Muad;wDWY4a9LH8o_oDc1}}@0co`h{~jQ#++K4}meJp` zlMa)mYilw`P{$)QoF7i1{q*R8Z|41rLQOZZ0K=RkHkTb0sIO>sQHIm}P@;>37PG&J z_V)HR2HWKLOmMLI%4MB_xQFcV3*{AN{NostuLL;rT5u)Vxg=`>L8l7HB{d+5fJRd2 zj-R5U^iok!^5X?W^#nM;!l#u|$5Kh;`~93|G=#86@78`};2TjzU+U+HFUkN8QbLS+ z#Udc@*bB_7-NO)?LwwZT<5cM-wn&a+NA{5yrx z=WlDdo;lMFWfWj_@s6=_rOQTddN5lq#>@2PR;8ZgFw~eD47qrM^utOw%dwcmvIti& zGyGkyS1qCxs$_@EDy)q=c?&q19OLuD#Xtn2Nl9i9M}I z+KZGUzQ~e8Eb)&k+W;|W)|H4ce;Qh6v;ndbp@O@*Wlz2al)}=bz%@}ZgnMK(VAg*#Y#NhbaRP4@2ta#MQ zVpEw@!~0OHTU*8v5(EdW?kOOGJza)34qrFP9ZLipOVdOhv-QpEA1?|xiLKN|?cvp) zeyR%v)<<_@BL%^<6m1@DU)!lNrFcJ&CrgGNL{ol`iPnxCf7NKhiaA{r>;15Y6u3<2 z#14b!4$Ca6LxzB~(N{r3hGhy@%0*ALe>qNd?081G8Cqzzt7AQkmdOx@3xki0T5LDTM~HszMQfjn4ld3QV+ zJ@=Z?`@^p;5Csmfvwi-;>3GtPgLNM5Sl9{*H6-oODJ?0iP&wRVnB&@9zdQO_>5A20 z*KE>%V`Ie$C%*U7Q!o?A!hRR(Ji_606}^>iXY22($GYy0Qvw7Q`CwZX6h@y>=;ngB zp%M!zo-{^`XlJ(O8PwMX3ed(m;hH@Gf#o2I4aBA7KlXHy?UU0_Yo=kEBD+f!I9KHO zuk$wu8m!%rZ12BwSjUhi+OzxeeWlo%3|GuiBY}{wD4l67XhQw5ERy*r;h70@JNE7! zmG;i4*gfLEmcB+8&vC8`W)*2qFVvKC2>Lo|aHg&!&goS#NDc~at}PjBAT?0Vloqk6AOi32D- z1~o|^XJR5+8z!!3G;9PDDNq*x1J(Q_sw0MYM#dG{h|R^2iQq20H_#6s7Qs2-;YUhl znur$;3h4qsu5>wbQ8)h=RLv%ccZdTz*>mz^h;IOc3BY6i&p#e2&(o{r$jZ5#^Spq$ z4RI7|{!{gw249&oj7#O{?%D&DL)Gw~wBiqXC-vt;qBvtmMDv9-}ZY2LMcBGNr@49yYItzP>**Mz7)C@4sKKnYm#I8XM8qFonnjM1)3SRai3e zQa40&0=c1YYHYsxE6S#6eD9L$be*f7^5DAAm7_-xVVC*O=dBaMBaEzmpKaC1c^o8Q z7w#e0V|by?f6s@txYp=cXhwqoTSvg}B6ui8_h`;CzWIk2q zC#Wap^q8yLrO|piN2l{@%5RipCf6>itO*0SO~aSCI_(73e z2YUin$MH+yr{_Ysxr&9Bt$0%Z(k-Ik`-N!{zHW(=KQmwI9yRM0Ve$xwjY2`tP8dV6 zt6^w&O)!%qVwct9DJWSi;L0T5J`*mxmK1DzjJcBuC$+(9YcI$ITHm|K9=U~%?;cDb zB)T#qA+>L+mEQt?ne-9Xh?+aVF@vCII(W#)kq5-D3L{`YTFtgyx-`?+>o21OGUdX5 z_?29*(2_Emt00A`Fn3g9_+|}>@_*$LQvF+Kk+uJ#l^B5E#iA>n4?Tmv}f3idQ`4S^zN?LJ;K~TZ&tgDpegl#hN7Lo{*56OPaUtR5cX2ml2W-T# z3r}XxK0DQm5?Tz=kpB5~S3}&-?U%gtGJIWMmd`g8Eq-}bO77#;>F47S3nc$DC`0G- zaMTP`H%|bBKo!r8T0o5cO`aixLg+F)j@J18p}quVb_awy9M49{^(C+CBYJw&_=xq=;AO^jY%a5+&lT>5Ef1s&;1ZNMGa`5x#m9^ zTwBor5s3p0W9q#t2vOSDauOC35yh^4k%bjTPTSm=eLdT)cD#@ta7r_c`CJ)`KwuH} z;|o96nVx>T3tA4Z1I)A!UyY;@)+_Vs5@m%5@DFec5|rHdUz#=z4AIG&YxODXWxvhT z_-KK)!o@2yQAM|B-BY$?9vQRYH*#M<%S)Dnym6m{R|TyR0g24^{+TC}HnsVbo|612 z3=o*{eS-KFubz&MsJGjL7tvODh)|Z;a3WUNUR|29RXZWRmMz8YyVM65AXkyU$~nxt zCmdK($ti(4_v*Exvw`Ms zlwxkD#2ay=CXn?S^C~vd)s5blm@fiHna&!xQK9cW@?;dQy?;Bua^23W3*3^$U&0(q z8_Ns2N9WY!wYrS^93*v@5qKQgwAO3gmmH6;L11!7E%Z;R`G4ax|3N4aFQfspFTR{q zyI0Z=ejB8%p^WU*I<7d`wzr;11s2me_mBc50gSKVIc-nZ9^s6w>7 z$)#Mke|eQM=R}QtL5SZ?vnfOn6}TZ97Kw@OhP@InWew}@n*=!IiD}>+B6zf|F3H=L zkdW{c?H3M(YOf!(*~qo%is@{}#HvUG?>bs1Sgw}FfPqx)_a==!Dev#W$}?cT53yZ~ zr%JNT9v$IPy_t%5_LW7c5^H#WscBt`bQ6+>bb8oP$5liGh{3qO!x8dM6J@YN=NTfK zmp`Q~Mf62Kg(qn7ziCgPk)afE@f4xf`Y~bvN!@UJN&l#Ntutt>ms9D~0MZ#h4SGyb zpF-R>LC52_VM<-^Z%kMa1Z3QU=c-$)7&BI@UpO04OA2~~)bw;jj3O37kE_AfgjxM> zs}KHu1ZNh`jRJTYXlaL_zT{!CRuiddelA@dA{6)D(2%<6?qJ1%_(L@_>prMy zO4--{3?UX)R06)U9_AQ)2*tEcay6fIRgg#ETnwjV|J{oM&^V%7?vH>P27BBXEm@Nn z9fYJs(={!|uxV0=kr7P`Xs4DiLfr6k3#HzD*xm&~MY1tiuMM>#J7x*i`jV&|{`U zS3Ef};V6D=$i-+K$*CuJ-U3Q8048HBi`K!G$sgZMv*>q=Gp<I^6HNw9(1|FE zC~$vu*x1g^ow8h}ypg2)(P@U-MN>-3`-ftXK?ycgDqhA_u%;7b?TkIv+gG!C|9Xn;seIsmc52Gvbx;(qy1>Wuai~mpzf7-p<)OJWg?&X zu+Hj(5&ABaWTFOFxWc&-OUd1Tc-~K+!}@j6me06xKGLJ98!3%>vAgfri))|u;8LeZ z5wj!uJOM+TY-0y)dY@xivO&#auDXvy=l`C`xXkwXYWr@tzhTqD{uC~SluSi zeQsrb{O*8>3{{0f)jmJ3^3%IU-x)V(;)%WiErZDPX`3T+-Q$l%ubXoI4r!i01&f@% zy#AfNT6Qmvx^??@Wuz@d%N9;i@SwRTlNP&dy1v=HPJtbIM;l{t8=sPDToguft-% zOXH_r-$~5M1J&kM*p>4gTu+=h0j)fkLU+=Ls|q7cCf*-lw{)@m1?Ld}M5*DMrE;(F z9~ip-Wwa<=DAGp6?@lK1^V{PE^CN*}-dx=o-7NSBQkJ{gX$IraPEa10KiX7vO7~

vO`iB0r=Ee&3(Ifi<6&NP$RuUsnUo5#N+M*<|Ud` z0J0~oPFUL2xyd%*NhjO9=pNxC zn*{Y9*M#|Yr)L#Kc?nc!zRU;-q7B6t5%Ex<9^lY%Oal0CzI>sUOVj5I;rl9!lgdA1 zNhQ0V=^&1t2rvK{rA@y^(35JCS5LVuT5C*`#LYC={pTuBTd$AgiwuQdmV@21HNZ^1 zf%Y6#mcu*}dj@KT)C`}659RmkzXyiJX6RQ5*MkQS&Wx`Kk|pe*6n9bg5oyPAfkkia z{PPyK!B4!fsEdYm_N1im1%x3oW%7c`}J?g!@WZrY?tJ9vd<%QSe{r}-~#*yaWh zgDt}z{rK$2NV4c{3ZjPr`yigSEu9LzObMCHwRV}|_>?-`zh>>t`7_qqF5sn#{{vWW ze=RK)MwoML`G>vK)HCQaN^GQR@!xZMBOq+-=>P*z5Gs#$W{V~XuTPCIR6ZEXS4OkW zY@RM*JXyK^lCJk zFyhh3WAbfV?&(%T9BAaHr78NhkLs&it+#`PfF3)_c%YnZVtEq{m8WVU!`&^Jbd^dH z!1osomEcWxUN!gv@D2MoYok$1%wIy?qGa9Er9o9%yy4^Wldl%Bnh(8sZwVUm>><6( zE>>V~IQdL_KGKg&**YjABH%zJql|(mKy&+18bOg0MSoSIVJ1vk;W@ORZZf3Ib0-Equ&nGOQ!0HQ?@27cdiWz(5*?Uvuj^{6Rp>N#(V z3;`1f`(Myxy1a)>;-C(b&Ykns=#Y)Wp@+*~6yOX6N+Q%1L2g4}2x!ds(oW~IxEIsK zH(%}xe@&=1Z5P}1d+{%Gd>c__=Lv&60!B$*8QpVf7G$xZ?wCsOSkL?RIqxpy3WYR~ z?s)}cckq#pU*{2>yBfNUrY;XZmwOkscnS-Py0&VHC6ESCUf6SMHUH0qLEI!6cr+-i z;-BTa(COWNUrydP>w{yZg~idTFbG{>LU=;U;a5xRof&2jz<>`59r1Vj{*$Gdt8y!B zQMP|{tON&0dn3_eh(lFsMMIh$|FjPU@6@ULtHRQX@~K*Q^NXept*!_Q0r#O&m=2PP z6eET1kn~je=xkkx&IgB+*8Av^ISuBPxH2Wx+H#W`Jh8T`sHhY|%O0m#)4gdG!ObH* zF+uRu;0$j51Qx0s&js_{(wY8j*6bMXLAueH*1Sgy%EI3}D!$`M&5A2B3~I;RZQjQh z1!xNPsE=h?^M_&XHI>WKDSmr^=wQN1mD33UY^7UtcKv=nk;5OWX49IQjc~oQNL+g< zcl5yNBJ$5S7*7C>k?|)W0ra{*Yw1{ZRnTuvbh^zuz^v1JYT&Sf%Po&vsvDBWH0cCM zR#%xrCVo3_|2$QuJJBo1jJVhRy=6>?e1VW5 zSCA_Bxs7)AC^u~yHQYsLE?rziVDwWq=B(Vr5(QPjbun4XFN5Kgmu6fIwjn0RE>OYZ z0)lyW_rLh@)~^-mzOinC$`I3LvXGR1=fhruu!eY>l-q|>W4zQOl84P_2l5KkKmT)R zN{eAor_8-4F?Vmyk#jvdEg!7_2?4W^LDW&hS8qV$$2d1yC9JsUhR*1JjpC!$_d~1wW2Jf0Ibmr$lj&$WJxoom^wlO;nLxdcq+*oR(m z2T$N+y_%jxF|2V3|00wDxoZ;9KXJRQ@cgg{w{L__He~FkQAi;Rc)9{t!{p%nE+v7C zm?ot0FZ>qqMFbHVS~&TGyzu&lvF#>%CYK_0N2uxX<8z)}sS8Z(Ku|P(bw00lIU|HM zR1L7*el#p?yLGdfyqE_{#YJQ{W+JilJliks74eyuvDunR>kC7Td2jR~(9PHQ!%;ot z@_63&+TaP*F%=$1L;qrBWYncQHFk^ual;3)JGm-2*R}|5MAXE+X3bumiX0b!MLQ(E z%f&p_Q0uJ{nf3HXHaFXK>-K;qqUB;0H(mspB>}3ROA4KiolY#Bx2&xwhVQ}|rD-qB za7ko|`FB5l|BAWxQ--Z%wv4}%D9jE2X-YYPNrtRcdW`6t&`^7|HDA3ax}3@0X0hEn zvuaBmgo&}(QkWB1sPGEGTTh~Zm-dVXNFrFhT3IQcnzS$f*<(X&Jax%EUv?EFr8AzzaP%$5c$_$qPjg*!s=#F(bMkyE?)N=x$zU|lLaP3h!8LU*^2yG^_1`G^5NW>O0@2y*N-^ipv!H-QC z5m#a1nX>L3h4fY0PIZcYjr_bis{#y;mw>v@1eS*4^y7UFdK9@LODqAdw zMV`oS5Ehy*!YK>XaU~nN;$;2kKL8v;Ye(1F{51nakRo-064I{`+in2MOtGa(#xWt8 z_ec`a+47cyx1172gc!+tPbXw)AwN0;@>4@JF<^{{UJ9fexeS>IR))nCH9c~hFBJV+ zLJxdY!g z<{Mc}y~BqT4h58kp-gLDrPss_?l1_98h1nVkNTVN4<1R$5m*GdBY(_S!;wRp6@V{l z(vUIbOe&?iMU?Lkhb<1xSR4hP+BVAODSu(B6M%j5E)t3C=TS(J`k&(^b3#{V6 z-D^o{6ImD#!**|;m1r*(m(CPz1z1{Ri*EHlm~aw5-_Ap&PmwtNbmGu;d>xGZxvFxwFpf@PtoCItnY%J1Q~?HNHEaN`ZUvIjaL_ zmDW=u!r^I|Cyj+7WZGK)dKVw5+3p?QV~7C@!^xQ%{uz5l>jyn%W6;lB9Pb~=j9s+t9%U+uy6sbV^Fdu;eiABu*^+$6mqZ0Eaz!H!!G*+`yyz=UOA+qVGZndIQ&~m)_C4)A?^S;U# zexnHJXqOJbn*ZR&UEzAY2#36c)Xnsm(p@!{kU{JGbav4__@UT;_YkhONw6P-z zi9sJnFf3aSA7d!-LfW*bw>jcZ+`8LHCBHiLN&=_&{o^Zf=Y934CU2>>cEr6gyKdsf zi^G=5?D{d@@{Ut`hJWeNuH8L;{(rq8`V@)?4A}yfXo1@y z?b)8XPNupxE~smn5c|LHc!&&>FzhGX30y_Lc{@4|TTKr&0SeL;(Exp`3R3gvr4H3M zyXc?lw)qpdC{upf6;pZn&c0-wqp%8DNB3Fp*%^a5J~e&TnsJsJ-=EfaQ{Eu%r$TXb zVXyY(jimyav))VD%%UK^%R@tCM`Y$V+{&i5S~7F?OM zLnp`5u@&qeS*rEi)*H*pqm5WBk9 zIGa{e)u&`se9lwVaM~gIcS-ua=4}lf-un)dyv`H;yQCRo!8CwY9D8>`M8k9LLVE`Z zJY4^tFy!Do2r>#wPw}MRQF|XpaB2Q%RVxSQ56^e9rLs|M_ZtkKNmYsv0_oR=85L_j z#nR}^#O}p!{fp|)w+SSDbkkn`&f3BCR1Mw?%1om{79(ITh*`}LQ6>Sq1IW1DV?ji7 z`#5t)AmsariBM?48X+hH!;9IKY;BG>*&$pg@OuXhtRZR#$MT#LsRbet zWf*7mHS&^SaYg?jxJ)7}u3{;DX5AddsQ5Y8@!1xKEXO4QYB%Ev zkzfPh$f}D=&o4SRy&vAxn-+vH6EnWlpxVcXFW@&yTJw%!6PDt0dD2MbQS{i|}of3eVU$wqx&a7N1vGLBUkEMS% z5f^0_1KR6yzLMB|(q!!L@ex%EZaRcr*$Oo(a7gmvpFUaZ_a(*w@as8!>Y}ne&cfg_ zq)XV>!#p4PrvV7y6 zMKU^oa%SnaH@qFE$sIX%aVz%${j9~F;~`Lkf69hP@#w+GL@&CeCv$=h8sNZD*7di` zHu6ivElpZqzNvQFXSxRrl{hsTW5xtf=LaTQe6H{+J2J(v>VB(ZR1Vw*5qqOO`tPJ@ zT#A@KDcFME_BaPkk}gBoTnKf7!ryUkj!`W$+ZcF0x%1$;v4X;%(i}SdR8s2LV^TED zP^gM}UmHDYOo5EBQX4iB;U1M2N0ah^Hl5t@%0RQj4z*Jo8uHlb(vN2=7m)Cz6d|+y z`QrQK_>~{2K~wqQ>nOp$hait&;OWu<14c&WPtw978=Z)2k>{Cg?tgaN*Mww!lG-N zJV^xN&dCw^pM6Vks!!P@#PzYw>ny`hsu52{N5mAokkacUK%k6-pU;?RK1rP&S7eaU z&ZS4^Ju^c3#>D#f^A&=*RCHFTp~FY+yu&SKThq%=UtU3Fa4K}?tZ$k!`Rf7iwARkQ zd0$pEb3^z1s1SV`m;wWiUr9TG%9+6EyvJXtZ7zS14ejio{~eOd_T|5vct%o1TP*rm zEV9HsTKOY?L5A4j9z-|JIb8nA-y?7>i;OH}8zyvkW3AQ~(RGFpX-;P?)s_!BqTcB_ zOx(urHtzMc&Vmg=ov@^WFV;?7eKJQW>8M`$hNjp2wPS<)zUajYzd&`Mq*15Iii5Wt zfTv4G$i3xJO+h**mTe5%-G?{Ngxujnadd;FGsFN?_5>)a8Z1`dbZKLysnz>|my(4D zgA@LRS^m`xDS_ua@kfKFDHQpdp7i^B;+Kl+I>Gbx(>_B1QNdM3@Z*p1Gz~V~ zypgW}mF@At<9Gh4qE~%D4OewJppHzF6Anb62YxgV;|4Ie6XbpC9&TgLE%cY|(xQws ze*3v?WbS0}nw0ou4I|HagAMyml0nLZ9tJM^!O}wk-jy225hVfeFxR8-MtyOD;wWtW zkdtw*r3CBaEhFy=S)TqZlVV}al4^g0SwQqMf+2r^i6aGLt?b1vRI(Fp5gMDyx}3}% z3wGV{x0wQ?6fTZVwbpL>l1cO=*)fAP2DS{v9fRKU1jW*a8HZt5e!PAfyOGYGe&7vx z{jV1fZt(qAO<)C!utTu=cRo~0--Vsd2?jYvvK`v3)~RVMI~|l^4O&{9zaXdxwYUsh z*k*6|5AdxU#6y!A4<^p)&VL3nVx#pGv?AgREiNIp9+O^OTA)h;Ys#=jPYiuz)UQQ{ z0ruIKr7sH`QFv$}v(*5U0wc+Mx~U2lfJr!Z$>f7)LCL`uhqAsS=om(Wjtl4#Y=Uv~)yLEGNW0ivk$qpGI0fzGo&L+RFoY4CnE@?oY{RuENE0otE>F z=WNbnSpdR($se=fS-&z!RPYj39p^Rxt7sDrTX~B!YwcG9vi%E(t>1P9E|X#8VD2 z@G%hvyFA1Fo(OE1Z^1VEAQo#w!ANwqR@ zB3f?Ypb!*2ajPXAJt?*3{!vku;{Y%X1^Uc8J<*Esp8~|ypVz4p3hwRB{)&rys;xo} zxc|n=I@txo0#LyN!}~0o&2$6Z@?T&`cB@u6n7x`@OuMxNvRP&*CKSz9v|uYR86g8p z(9!On&~+_Smwn*^LFvlb$x`;xYG3Yi9S*FR95A2I4;m|c@z0!ox?Rs6M#xPt?5NZXAObt$rh)7C(k#$&md@6j$1Iiq#_wb2AC|Z4$_rbjRHo|_QeLp3WfiB)^+=l&n6-aMe@{C)pFS&Qsj z_LyXhElWa5!YC$;HAzZjiBKvj3dvw>Wkgi6FUgWrLJ_LbVo;$HIm%Y4q%84$+$Y|j z->*MD%lkc3=e%Ce=e=C_bzS#r_EuVAfyopFqxK#4qYJDJK$o-MUVF_3^}Z(_r0|#0 zsqWHZSM3_CRW#Cg`N14Z>$bA>K_F!E&44!7>q`+Jh7y!m8hn!lM3F-4!F*dJ&(I)? zPkkTZdSqflS+&FMo$$|>vU4fXjKlfG`+sh((mAW5k12nPXlrUv#{rzN*6+W`(WlCf zcil}5X6&X4K2)o+>HfHA+K*SaY&k?)Zm~hnJD(z#?m|gmiu^k|9&F=!p*NR=kPtfENi`5sKJle2j7u>DYncBU^bv= z19#C3p*C!er#Z7W+$ah!+==L=pfY>$&rF-=@JtnEW%E{k3;3uI#GZdZ8ej@PMZYGe z0TKkCgdwrJ!`$^hDz45e#SM1i!DPVYPcV}L9xCr_cr98_6dw%6(ELy=>3zArJ)@4u z86eVZqT@SSZOrO$hJ_hCIWdTMgi*z=OS)1q_qDoY=8Rn1%xWru2#bNO%y8IkOj^Xy zQatMc5E@J|@O8p}Oq*_waCZ@$0_o1DLhn&XLu}cZ>3FK{abeIH&6>q#-NVfg; zh;CQ?C9R>OPnv?OMP3&VWec*T0Ap&h`K0}LgCRrfzG6o@qxGJEZN0AVyqoix>?!Op z5`=i5u2^94l(u;G_d0}gWyoQbEvvpYve#8imTj(A7c^GEnxzlb!@iXII{k5cId(Pv zZRRf@BZ1hwH!AagA#(U@hC|2LPcYDy>j(;Vd)}vI z{Fk9n>LDxSU2^+G_`q0gK3$c`8{d2VzrSO=B(xBIp^~dE0Mj2A%|@6oj#&*Z|N357 zT_H}vn>?fAy5}u?~bAxluPkK4?_*-3`SH zzyI72y0`$`mfpF2X|wD-UtU}YL)U(6?-Q%7DNf)#vC&&iUnv+`8`7}aVemg zpd+IGLgKhTEOHON9JGX~5JC`|4rIADet%DM%`;=xQSU%EA3{?ye)~BYw1m7InUz~x z(2bMhz{E?8u2=xVG9T*jkT=~v-1TlHb~#XPh33U~-Lhqi@KY&Z_Jd&IHMjTQN%T<8 zrF8bP=N9)vHlCwMld2Y&m{Hr?pd{=tie#oxTRtPpIvAM+K?=I)O&W|{@4h@{oCr!R z-h8sahN@^zlx`>!e7r0Q&U&lyM4Tu680U(|$+-@)=nODy400QvhmFHJ_P&i5Tn96o zYMdL_)BF394(l`r;bcG|o(Q|Y_!FzC6;AN_RO`1ntgKs}%76+n6(=8wW*Byo=+bva#G&5fMLzo5Q7*6)fEeMe@C0U@F~QsAByw9xnwNZD^#P% z=jdKHPy+)p?#$hB?nj}nygca53qpd3e;8BQW<3DH6TeVqhr1~+&Sc$(_kgo@xinMY zfmGr$eU3lX{4;+@rnmcWEAtuV@$_Vh?8XFCm6y+hn%NbDSatI+S9wNh!Z4+C^8aum ziABLo$Xk&QkJ_l`R3nZ*F!y+$o<^Q+05v}t|u~c8T^OnV0X`t`BVKGryx3a!19<8 zOG6I6kF^5~lNEpxQ*mYDI9G!d@M=Ge{8%z#V$z%MFRI_#ER1&O?3wkhyzlO5<6DP? zhfOgX+wIg}LwZd}?&taO*gvKV^8fAk*nRBfJH~^m*3^94m+!jYPA~CMen!%fafKm{ z!QSaL!-}qEwAh{FZ>J@z45g})c*e)J>iizmRJiG!<)IetrX#<9>-E}l((95lJc{VZ zTg47k7}ap$&OUiut6`>Wn?!cIxa`C%M-dl*<7lCRonOhE$*IYu6AnFBCF%(amvi2o zZ97;Mw$N;#QLm1?d75&W?TOV?W+nc{@{_!MSB$PV*q&tj@XuH&-orKRI;@AgSxRVu z&p{?4mr#TU&Tr`I@^yrMw-4ATZIo6Pfh{%b#>D5^&DAGUzjqkC$b@zoy$*aQp81%T z(b?@fMsk7%V~;&PNVS%$`>BAKv+D}?FBBQ-S+{H~=;-EKZx1+eNw<_Ei)!w7!W&6I z$sJe-VH!-tN@Z9jOgn<9%1ExopCSd&IhNi`j3_wVyFPaNaK>)pL0QPizx=T2GH&?n zL7E$ZlyKOvi@N4uh}!5b;d!ma=vReJn~wCKr4WN8CiAFxbZJ^A1N!hy#4Ud_nsP>kcoU}!8#bU(lQ8o#3|(+`Oc7E z&Cl-HPx15RdTxw*5QCQlc)tTerJ2?NJ-=1`<9CmwpfWjtv`0ArS@}LeR5ly3#!^Ku zesM5gSGc)Y({3TKAAd6(q~kCju~sBsPYwk&1|J?jY-m(SAwo|FjTkZFAZ@Qe!R$g- zoN5%-Ga`}(Gg(p_OfpzjV(B49izK>M?EMxagfW6f?#{`+z160oqW9~e@gbPb%qXau z!>1#Ui?4ll@iz%m#a}NEhWPXT{rlFzj+eJ(_f}br2S$U6+n|=FKwJ`)#C@MfZ*+7* zV2OY4D|43~_5_F$A`pv%75-VTPc;eaPZNWBUOGSsg06kP)#JCzPc|DGuEU3t-^RYu zcHpk-;zP+{pqv+bN2D;#>A}#JD#)I(`o1=|Z{sE9ynM1Gj;j-XrN5O~lpjzf?Qol6%RkiRVp%sXa{mi?S9@r0a0{`7+=IJq zeE!ay;)$7WmJ)o9Fceh$u}@D81F_=smw9fEH@`3!_?yrg(YXTjYfT{Ks7;_J>^|$> zTMVvN2WCe1i7?-LnCSuC5uMqRA5?v+!OO#x)#g_?Ns9q7G%JeHmRTDT(a2QuD%oVUH9`^AzgTWCF zp-+$k4Y+dh>1n;*UftvSUt04kc3KD#TCf4M_`KU7UvZCdYUt3@jd9_A$_??4RzSq(x)C$159+t2}DHsn^RN)ugdu;abk-j|EbC8%<`ZP9uMgMZ? zdMoa=*%OTh8-;U%x3+dzlXT2r;MkfCjLn^|)c~aw=*JWU8%*miZhXUt{9-zAny2i6 z=bKz^@Nb%ivxE9IU{%Eo^_nS{5)McfW(?G?&Uyv@SP53v*C>sJ9lIiYRsaLl9gq_C zGO1w4MZmj@S=d)`*|yMomsZu)!yi?g&^O%B_Or8$ix_otezl<9L_wPq28eKDE7?f0 zfq^elr^XbC8y>vA^?ZfpwmAMkIi#wo=$5B#%CI0jiGK*Qzq*! z!>_S3;;&q%+RXAV@kfed=@OyO&lm@dCA!fvvwdc4$`NE<_$^eU0w|ez^t_^SA6Z;V zH!ejs==0woe=v^2o34x;WCTn^Q8oN__B7SCx^t1F2K7lqj9Is0=nGep)pb3uvhNOT zO{969hObgbE4H@T>XO(Z))F@h56Lplf*yvKMHRb|I8dP_;UV}hu-7P_u#Zbk!70^sO0IN=_8|7v zZ;Q}+@fDltc!u+>XaQvfE5TAPWvB;;OTgKbjW@JG08S7_IP3)Y0h+&(hHK($X@ z|M&@;CA74WoJ=m40tj7Jg)IIgQyBL?iF%~wLXJ=mbox?mf&=A$&ifR%okO-y81f*V zaNAHIh`6hqQZj(0zdA`+Yjcf&xI`~6{6?Iy*|!@$e1w*!RX6+@ekTx;?@m9c?R}Ix zq}oLez%i^%C1-Fk_wPgZaM}MGz0nI;>&50e?&O?Er4oNy%2izd44WI_v7NNZoUIL~y9O zA@gWwowM|WSf~)b=p~{GQ|i}IWq{5*st5e8UoFpT12XD1SJ#>Kyswt#MI1%M#S%)L;;2*44?=sv zQ%S+dJ!;KY;zs+SD-O-r+=jpLH_T8$BpH_3yB_A1K-^Ji06R=<~E3JRP;L;oz}q%k|V|m30Avw!3rfB zH-vueU!~q)BcEFsciLf3n>B76`3*8yodhZeS^a8BQ(_hWlMP^T79MXxY)Y~2AU$lO zO&qJHGd2aSi7&n0Q8s?LfmW zt9{tntF{Y6i$-nxBxATP02)O=gO*zzHz(R{dW@TT-ry!-0$&KlTS&SuRU0<$r{+zi zn7k#LwoMsL5@oNmJW(1Ie{x*o;ZM)`sj{EJolmWB7(aBDLn2fqG6vUGPZN8>TZqyY z)p!jkti)vQY=+?G;?}Qlth&rYmZXmdmVeuGJ$Uu<_U{hp8QpUTeh>U%#4UvzN`dub zmRdONIKiso+a269*fh_lFN}yQxv8f~Crk%daINa;5h$nG@ABkuI^3|mw`5_Q>)Sgp z`Di*haJ(i#hAF^)fjwZN?epOERfi0J`Z1b<`il|?3%f7vv9^Fuw1O&RUZ(Pf?JnA6 zA=Dm=(aqTX%jV3z?S7r?BMeLGOIxi5P&VOQ{l`LY#|uNbgS+=k-iZV#x17(!tonwE z`aM8%BR!)m4`Qz3UoR^)VHTE_XoO@(rDS`P`x9l66y9Y4y4_Z}WX7PVsi2%ZUY0{i zpBqHBAaaveE5Lp*!*y^Bd;nr4K3%m4uMPn)r)-`6l#ZqdyXFKvWkKE;|7aXbwe7+> zFq>rNN=T4}Sk#@dIYGV3e6KP?lHdrVMN?)Y;&vsTJ$#Go!XEtl^5Zm$Z6%rw*ux;+ zm7qe%G}b&08nN=%zRlH}^IxpFCe9&ZhRd9aOLT{3$DCWs`m-yAr5MXQ4;q-0iq+wH zq7V$(HfG*jV3{{CE}SvB#yM5bs=#B}mbyy<8=V4-TpteVUz~6L(CuG(-e|IF!+!Dm2!Y(isoL>IqtT zK=I#`tbN5E8o2PebF3B>BYjHHNlXh>HeZXfBZAwQ#^-L0AY3~{4p{dUI8tR4Oed|# z))^!Im1hRz+=PB20oAc`1~HEbDm4ttXIq@=U}-S=rNeCRShp`4v4d`p><7##_X|=2^sU+0n4xrST1{?03j)>*U@cj-QE0*`NDpX+rgP zdJ?@{EI=tv56NaGO6~{l{LAtaJ6+4~hIC!D+JuCLb<|3(yzNIDx$p1*=ZbyuTdal} z>+1U1>It4Qj_3t8+nFJRO8en{l{uAgJshob-d0Wtw;TW_v%J`AUF^3Nk;VGw#I8xO zGsc&A!ZXp)`FqRUk~I!UUO?yQr!^C`>9&M!K>j~|+mXX+Hc&GWa#P(I7EYIok&bD@ zQ9X4!rCFB|w6-j|7E>N-B9ci@a66F=b2U!kJJ4eHG}YwOsaJ~YUk6&<#gd%n?TGjH zZ&~R@y)?oph2B%A*3s!`>J5!$eecF$--%SPq^+J;+{%_N;z3`acyN&>ad|uVV@U0p znaKFvvD^MVZify8m!c41u8Z=G`!a0g%At9yVx?02MN3n-8^8~Ncf8eju;%Zm}Vm3eQlrL7Pdq57>*wRNE=~OVBm(vgRkba-LQLKRZg= zfa32Hdr`K~&b%m76uC|20}f?6Z2q4<167uE3n3?~L?-80d;iTSomkJ+wV9!W~L??%H*_L zdopa?BK5qtl{WuPd}wAcj!I{lQ>c_tmkV7>-r0 z_#R1VjJ9AKxD{hrQ^ueqou*tA$?7xqef>J*>y&P?IvNkU6#CEKLB}J~y2_r>w*Emk zD)u}~--2q0wvz{(b+ihQRdMvya6h~EdG7faP-4`UpagL7m_&nNwl%c2;vS=Pa9L-; z#$^9K>1$$N^nXc7Mw>q*Wg&1t_vY$~UL8GUXQp5daE{~iGf1D~;LL=7tDVJ7T-e}W zQb--!O8d>ezCtSY*_PgSLT&^zYG2eR@1d)RfR<|9p>q*Cguc{>*}qWVxsOa=|2h20 z$wA!+-id^1#K(?3Wl&%^$65dK;|Sevxg+}N>wn^TFnC_&|AMw_Zf~;?|JWjWoC4MT zv!P0b$?O#MUG>0||C!Q=veQXpOJdF>qsrDa$;W9L#2YB3(Q=u&0|Q(+>}ma*n2}Fm zMVMjJ)b}3a4HZPyl6qu9$v~a9 zR?SBfZl>%E1bXiO#E(;Gpt+GjE4WUE=4I$@&6&qW~;gg}p__Ut1x>ycQVuFq^E6q+9Z*3h$bY7`$Y43I`hulC3gFni82y0g>M9Z?T-WP2!*k? zdvxi^YTEEQaeFypf%LcJ-SGe#^z}LjFS0s)+>@pBvjbtEfA3*z{B-zolCG-1zUkD? zvmez2GlVp5Rd_|(bv2bJ-#KcJnC>13vD2EdI*$Gn7mbcSbb8O6V_{fjU7%t#dE-87 z(42n0-}EH% z+E#DTT5kYc;lx1a!2x?EBmu{~?_WO-U#VULJgwyS9tWailsj6*D?5ACA*%9L>hKs^ zNBO#p%)z1^if44`(4mnDi{yGuIjWAly&Llf#;>0)E<0tXC0jZ~B*)W#Ya`=jWcM(c z-XG3bYEwlS7}++4nhRN~xbxd&O0bikTumZ1QCH?^ti_I8T~^R|2kW*Oriha6TUGpb zVbWQJ&NZ0~^ZPgrM)YP*mNs~ldT3v-m6l%-fXUuELgb6G)zU3CJwZ_m*W5fAB>}T9(QXhH zWKtif8G1k(if^htT_@L~$rd6Sj3H>Cr5o%iq4_M(bv8?B3bKtb=IqDndG;@<5g95t zgC&*cERR>ysK}@8x+P{-D0M{vxIpDzsivX7S`;d1WWz(JtE|Q{@4cfSiG96)QzN4smyGys zVYr$T@36_G(EJ^t`B@Y33!n|X&*C@&K9m&5P|qRteW{s| z)jV1ecfS4nGmQL8tcj@31@bE+c?dT~yeQ~!eS8yQ$?rc4eIgNtPUE@u9x`AC@Dgu& z{pUu*9D;42OCtNt%x-!f8jTDuct7*|mMrPgRlOdd;xS5enQ@qDIW`o9CQxhxRdu^B z_3}_so+8WgBmJS6_XTvSYia*lm>kR@UJ#YVjXG=IyY(iDL_KaGrGV^KLcdpU*U{3a zwFND^7*^yxolbgmxJX^q+qjBTq|#vmM1H_{i>|M(hJNhbmltK0ngsEoJyB0kC%|J{ z&L$gTEF0g_NeYhy&6zjn?g%$-o)-lNs6cMeIj40qfxRKKh*PxV#JE`jfmF4y2^1=F zi}KCmzaOR#MnE2`1{gL{h5^*7H!%GOZ~BpbET{Og+o5I|C)ju2G1kG-Xi5Xm z(k4KsP1AxMvy{{W!YrKeSD%#@P7Us)Wu&d4GolxFK!Oe3KFQ0D&zi zKDk7z0ZNcVpqEktLP(+39fz})Ydhg!&#iCv`A48?Fo=LnJsN(2q< zl#-X&Mz<0k2sko2?s5{aj60-2Q%$41FayR> z^RJgF9@UQBW=bt{8;V^D5!Mp&5DonJ%vc8tWBN5(ti5dL8fv}no5KHGQ>NH5O98SpFTAnQs0 zcVGOwa4ljNNQFu7Bw`>jy{6Qko^YVPGmp8N*|(7FgIcErHj>Fb)yi!Z4a=)Q2njmU zCiCRBM|sBDL9@N*!}u|w(KakLgdU4rNz|y|HEQWi!VjE@Ym4g&m4pz)S|E+)(F-I7 z(qGr#?_NB03;Z9K(NARs^6!hVO2&v*7f6{HSWdc}!b^Ns9`%lQBe#*t#2%YM?gu?~EXPyAj*qgP%eHOXjsOJ7?gY3A0xwI?EQ+iA@88DJky6PP z0P-Ik?%K#q{fZb1z~13SB{SMMUH=7DA3sVz)>gNmfU7GmOX9r8o+7T=K&RZHaFn*S z&Qv;h$HsPCQ*isa=socMJ4r)#v8X{EHQGG*l`lAs;-rGqW!Gf)qU7B|?3CLt_PG|P zCs_w@#Dak`WVv69VLKmpji*LqB6gGS?{hS)9Q;T^F-?%97BLx)Iuz&8$pnR?ikpYy zDGIG87I_azFobY*omBUVFEvslNb|tWn&6VlH`xJ;-k#Xp$OOQNS?~C*qhAqJ?I2LF zFQaPflJZ-IA@cbc&t_QxS2@&tCH22&{^SBktmBfWkV`{2W7;le&&+6&Kd&ta`U>Rn zoW!CU5C1$~$K-fHCLciX;+rxpk8W@^?fWF0MRE8)`q-#{ThqoCl zxNim6ET=W-@p`p5$^e}UJOq_jtXsG_w}~ldr{}h6rs95zFEEj97ZD`@1>4)tbkt!y zP0FuU3KdKD7?|td2$F2|P{YP_E0v&m?P0{613XEqVq31PTxiP6>+{v~I_=`Gi`~e4 zjwTZkQkDLneR)LK54C3_4*!nqBc^nK(8c5}Ja~F)@FU(bq&M-I6A#FCxjA2}n5~Er zCFSpt$`M=K7Y}GynwC+sF21rx8@3109x~0KzO8>)D;P|J)-KyGpn7mZNz7tk%o7jP z3_GCz31ndr+Gm=I@lyg8-rc*MoHiNtGN)}P!*Zh4JKV%!XIfVuu=L3E-q1HS_MZqnJaW$$KT9Hq?@=};Gq`|BW@2Qk!)ei~J`hR^qWvddhn7`IEGp%cf6Q0Rz%Xz+IhBK);a{@)X8B<32mi+p@8s zZ~&P5VUYY0%pc5m=x=mVNW}pPbEl}nPB0EhFtoYO+n*v%Q5If6q9yK6n3$c-o0NL~ z<1y1$h4FND4EYv7u*otNy&y+ERN8$!1_0>xDnmrh0@FlHtpTKFhaTvCDW}+t&1Qe_ zBG~QlU;U~f4L>mPy3@G-Rgn<0_O@ZV#ZrU)8S|rappA)NE!us;C|K2%-@T{v5ct`B z*AQhWzxf3@=$h^QPkbDQ?Mt;9KsVUero*75M5Y3xOJ{gh+d&V>F|=B?X2|4w zGiN6~t@!|$UNr79CU#KW$(k<#Je@C5}0S7WVv~7C_#Std& zJARy}l&ayoOcy}2g19&JX!Gj*a&WAHC>6A!tH9luW`LP zxb5F-6zXUp7~h^jGi`}5ePCTi5C0d!EOt}03TJOQeKTyXj^0{Yei{0JAbImkgtEK-B)enW=d)6m@c^*m?%at3*~C+4j#tGu_G8#*Fc8sTUOOG%f2&lgD47 zIl@GUOeU=$A1i(Z?UB`U7Z=k`Vsf;IZzE(DAtKUj)@8gnYvUIz%1$uGEXG#bETt|pV%;vQ65kq@JNohRZaf7 z_^$YKaq1}@*7*nj_lLvwT*l;W`&jxgByozKt^bL?CL$!7@T_YsOE=7*GzIrgqRdGN zHjJLW^FY-OS_F}bl93eVpVKX*p`oXq&jPmAzfRUpsdKK-+hL^X4?(h!6X7ytj&YZp zpqF%wZ6efd$4Lpmu5A+DBr=AZ5g(UdTEwvzlQKL0O(ENHCPOoF&5gQuuanRycmIoe zDN)}gG&WAuOFm%GAB0J)J!YgW&AKg%cxbakz1r@eh^vrs$YIJy8iH_ROc`x^@?!3IGdsD->@{? ze|&z=4+(-f+TU%Q+A3p!X@bMjs`oVg8|#p8Frk% zMiFd8N3n(DK6~4_<43!Wp#5Pu5y-b>qbboGN_p~=Me-}$9I{TJAEI+>z!`tB;R@~1 zO_Zd49|zxiOP&FzaLiN!?K!|sX39Xi7|%#Xz)08?piy~-4B_Mr3y^nl?^6^Pkk8xD z3lsqc>KGe$;Yh>tq`y+`UlHi)gF%B2W~*?MSnI? zT>iIht9uA94hFXycrzzrKI-?p)s>k9#1-kLdwXHa4oHX?{%_DrI+TzJHi@iy>&GXC z?+TJCC*iHTRdYb*lEJ-}1@zDSJp{KV$1HJ4Qf;8XV!``{U%y^!FH<4XQ*>m)rUEPm z3#-w6&mP}(3@e-NPgn+hn2f6?*nLW49dTS@OLi4&>Q)6*$wo`H}Z zWl+HEOeZHjbZ!8&6VONbsz z&34nhv>VHICMAlV$KUrJXRe+JPC^`M9k=jO+w&eg*mOC+tNZwv(!u0x&v9C|R8IqxdMZFrtSi z5M720=jN{XdscwsVN?~li&MY7B9;I^HlU120{)kE*5C$WD@RQKxM5w1GQ#ePaPnd% zExB1#XoNpjI%i-Qdw*e6m%sF&cv0K*ka&=8o23LLku}`3lMWWwp+}z4H5e@`JI%E0 zLCl5m0LU@tPYYS_EiCFTtZxBfWD(%a(kiL}S!&H^Rss{l0_gMDF(KA14)DHRrz0w6 zc|^k=h5RYMfsK`n69&M=7T&62_^MY5UQz#-&fPw6pX)4!-`Yz&F5EBP+NgvEKz!w( z4`pEqA6d2|l4LXAD6c;`NTo%;*;NnHsl1ime2rV;kSBaGH0JeaVcCdotPOCQ@Ovb(}mB zO)yLcfn%Hr`R`m)KYssO+M+O{QAjQPot*Y&yy+|AdskqZ)9wP(XuImidR%R9Dzud! zdi<;iN9+)!#E|F)H+kE~8*`U*k5f_nTS42R|GkMqSgd^>!SeLa2iIX8_HA*x;#{t- ze!iJT?o^|Q9(&nHZ9;D=f|d2YrFD@FfKjw)+0ws2%wTex_G@F>tX`cN>eo2D;t#+e zZY6HYkhS!))&kwA&9n)Cc9nEGWhcfdeKp&`rG_tdiz@C9@?R}mD#jGVJ zq8P){9d*g^#?mAT;SAx%n7~&d$_rVCh}hdl1-=d&>1%UxayQ36o@j>QGnQwim!7+w zamH@6q2dlZUEhN|35LNR_T!zfFDE=LLlPn#N_Y`4aEsXXuV;%XaC7p>eA+iTbAZsZ z&z!Bpn7RR=1CRGEL1U^{+O2LMC?!CV zd%^2CMtJ96YA`hV#iwJBmR{~{<&pR9@6v~-S|M##XrmTL)FV>sz!|Y{mJKe!x=dVo zA#(@&8ZUo3;;HxdwQ!JpklYhj8`IttP3tirQ{G_Klh-{6j-p{^JM;g&uc_ll zzY^*e9FQ#E)sXHt-9EQxSkt+P(jcD8+EYDE=KoZ~vPp_ZcWDhB>+YN^T1qWu7kp^3 zci=Seg3zSL=Qlezdwyna4_CgH6qcZ64EkY+&g?YLaBjg{LEL#S=%FTP+x;yC6+=a- zFPTW+SB6NMIZSeOXSX0}SDd4XQRt;2ufkFd@ z9LW}G4o9a?Vt1)oqua*E403LGdx2M1Wp>PJOmPJ?b2XJtBq6X5zyhsk;p$;imwK z$ksG?fy92aLw9(yy=H_znx6MSGgyRa)O&~=?bo(riYg^_F4NYpc?oEHegBH#<8o`d4&784Vr6L%g1jA&*nJuOP)vl$M@ z92C;?5f%agR^>O^{>kfl;~o0#7P!V*^D#LK>rH$B!Urt22rc9GIBFoKfBfHT*Qlryt_mP&F6rg)xY=y>(X^A!ldzyRt}6!0MhlqFP0 ztT~uVZoI0q?fyRfPpESwtUXY6k*~SfA}56x?dmFAYwSscx)kejqRqE*xWEYVWP6M( zG+J5RNt=S=i3ai-Wd}J$xN3{9=ZtDpx|VUF9sTSmV!RK;o#|tB=`UL94d`MmJAZaC zTzf1vb5YH@kmyxN7o=>piy6YBNnVc)1D==GC>({nXM``GTCZx`Y`47juFm*c!Z zvQAA{fear}t^2wMHKFb|*m2ZY|u6JdB+>hkPxm8^5#8=!X5DvMJTpV%p6?8T7 zi#VZW+jurHJ6Q&+NP$?hVayw2PkDUX9}|~jU=;hEhjr2Y#CrQ{(PK2I{ zCL$!N*H--CUazvUFQj$gKOByn`pedk9%b|1(`XP;h4ap++jCU)2pCb|zN*^5KlzxO zsexlT&&g8%6H?lJOlgAh;z@(D!wbf*7pj}Ah4KvyI@WUDV_o2MQKyqo)P*&wyMMEv zUpYEV8H4dQy>EETY}sq=(@u$}aY2xe05C`m{^`=${YC zGK>($(7((N#KKA;NDH0TjJTx}*TL%G>&tqs)6-7*>n>YNDb0T;yDX$=*&Iy|Fd5bE zThGn(-25aL|GH9rEw!@rr5LWxHVivZ8MBy~X;%EyxYeHX3S$e?mpQ!ZY%7>yB(*~F zu)?8k%vC7fA~Sc$0}-(elN;fJ^O?e3Kp}53AF8HbJ`2X$?mAVu@FRge!#7#=Pg!Cg%!U2V0-S^pC?>jUNW zLao&xv_41mKXH}97ZSBv#6Eu2Q3P);h2zglO(p#VW$uQcyx#hT zsX6Dq_Uinm`!*s@ zsBUEH>2ql;lE+#v?fF5uq@3bl5Y_hJ*q$a7QJk*xPnNTjZt2o@)8`5cR%!KT>4qOK zill1Kz(C5Z1=*aAd)!*0__-xj54M?iGR@fAH>ZBL4>W$Yr0n(%Cq9e5j@J?1w2Zle z&Fbg%5%q(g_8JWyKpG&=LYYhDpZbHEO6(t&Td*4_ z0X?QSPM69%9Ra=K;Oxm!U`;X)MkI#ZFf(x7x2w^Jdj>gX^&t(}MMci+U^++;NxO{V%?6YEe)q?=wyiJ4f0vy$t%aU#s*`@je9_CL*cX%`JSi( zS$x3U7MA=sM`tnK%ef6sT3e0KND0faG(UgI-RIUPQ#+Y>V;dt&l0gvFB8+_@l{xbb z$5ELG4oJvJEOwh6{GQ%~?nCx|oSu1ZoyE7NMC9lrN}{*>NQ0hn-f*1%2pQ2{+kC|c z{nXM%OaZ&GN+mvjd1XVOf3-%3fAmVr8BC(&a22vHU7wmVGjh1R-HSqo!;$(47Hkc z{GG9-B0Oy_6k8gVmjc-W?gNkfy3B?aM%i=)bk_XRhPFVkslkSU)z#H!u7vUdsEo4e z40Nz?D)}oy)#W<g+=tE{C^)ax7Favz=*wOm{?^3u0z+ zGvP|?)CufJ72us;X$0~%?Zj28%ircKCCLx$lpwCH6evRQJy7fgI;L_xwu^c!ZfyN> zd;G2nQ^YB^Q=)6-GRx4Nk+_&spQ|=@sHx!>t7%k5&&nQ&`?V-H*UGv$_YQ;RJ2(#y zrdk$ebM2@F=DP9NioB&m4rK0?f0_*V91$-FFVSa!=f|a3;K(h4?niATHY9R{- z!&J#J5I)qgj^FY!h03WwG?mZ^o?RJ1rc6#N`O0zyz+NsKrKeaF0<}~5_ImyGI^j@u zVVoN-?T?!c`)R3-&A_2xlGdxECZbWnYB@z?IP<*UH+zUKv2|cR%1&I_#HK;kt$@10 z^rj&rMztwj1^Pl=E2&%zJMwV?GANebSbmgkoq6ExCmXK$@T*XQARS*X@B~-n_HE97 z8>Rc8R*b@>=zHrk9hrDOYuj1pu<~3blmO&yXZFd zKj5EFVz@;$ek~|lKJSq^M9|CG5HE|!_6tZnGrgMOTuxPtCc%73IW^(Y%riXUzm5(| zyMrl!V+r4fSKLD*K!53@_kOKYYjZF91(uPIBeGUA?ojk&0}9VLSi3CYGmb%&PK~8J zxK&hvEC-0{C=i*+o3{QbB)%ZudorY@a*zgDW}-4=XICeSB>uvaO3kVT8b;+3NLTM~ zwVI^!SR=2KCCCOPc306)ihW~3ht-p|K*OgSeRO?ooc^w^nhj1(JJphWc5UzfgfCO$ zrl}h>SBH=(K$1BAAjyigwTr%soL5|_p+eWuuz$hv6zMRD$+2v_geu=#_DW&(N+H99 ziCDFtka4DOzdsMxBZ-@59u61p2}TALd?*oi`Ai!L#o#aNS5T-`QkePF4};kd$I$G= zn(H164oOs)rl(wS(9=>%yunBF%J0iXq~lyU_!p9EMEl4iUVMtxBSLh}L@nS6#|-HF z4AI-w@_&0vaR(qZkSn8Yr}unKu`GOb9yET$c^jO$B$eLT(fKYdZ}P!mkz78^N!dhD zwe?BarAA>g=%*PJ<**=ci@LQZ&|hi|36{qXI-0iL6^bimvInY_@cMbFx99qQS+}b? zk^!=ef|;M&R(8bDVFzwGM}JG^Ps@+9;gx9ubH3mKgA#}q3y~;+dSq5?;F%N}WMLvo zg}ke|Ut_8`8ZDY!)snRx3wG=t`AHcdtXYJ$T)S6Dut_ApU<*a&Vn4LPKXv_tQSM2m_i; zc}}%lC4;`=(W(p@V0dLx@bLONzBi#;0PaA3RTwzc&w|J$LMm>0Q=KNclG?XuCZ2zC zlZAqQ!ld5dySMphcU{S>uJtkRw|($EN_6yPYo+#m^L zApu^F0(u{-sQ6*so*o)J#;8SKfD5UjRRjt=`#i5UI~%DA~% z*TS>WPYcAv5zs3k_Wi5^J?6ZLFUro7agZD>t*oJpZ>iZp_`P6Nw~F0*U!|FIzr;>S z5l=G=ZmFC3- zit-b9QZ|i0yRWL-u^QSoJpeX7KoJ(Bzw+{vUOLxGPY&!&yEkPR=VLoLjaEoXHQuaP zfNM42vQfcR2s1dz`t8D))-FBd%dV$87xEZ5I2>ZaCgcoZ3}OVW1L0gwuWf?jhnS~5Mp;cm)%$f6pNtAFg3r!0JyOze3C1R%2w~gH4kj7mn?sGh@5ri{DHc! zXrYc$pvb-A8|!@9u>!pczu+b~iZ~bU(mPpFwwjwXh_UZch|Q;Ejh+0o9FTOyz4*yk z26VVmY@EImZ)nE?PbLep`du`utfap1xr+03TZkokwh9Kk7j4P|>S#80lD@ zVu#}OCbd^0_t;0-FiV&qAbopu)dr{p@J>x6BWtfI$g=^ZO4t8$cdYF+TbD7=D)ZI#-rg58vJ?Cl^shd5N{XBo=j<0=gP7GH? zo=*@1H=#nO``am~4<*nskLN7No7ZJswcVPJ?^0+|o76T(!yJ-U)4%Y2q-=ec6WuJ~ z&CrGn`dADQiU?~vRp4>+j-Ij2KMP*ZC-46%K^mO5cVCq0QCLym@7D2csx#}g(J*cV zlq5YZY_+#a?GMy}%(+Y7fd~lY$5{g+W?l2PEA-zwyG((RiIR+K%&=H?Q&G3-VQmY5 zrZNXojsPZ4`~L+>+XBIR)&g*VP zUJ#8|sPS4wYuwjb#A>2#=WD^{q|g#F&~C<3+F`#Km7P8dS7;tj-sJv@&;QU-wvwU> z7LEoTWV`5`CJ92nrkA6uogyZm$_9-I(sAlYg4wv`igP33o~`M#s&UZq_DtuvB!d+E}ZP^ z-#$D3e3&}zq1GKo)BNzv%w`FV{Pq`@P{M?-EIt!l<&b{t4C@hCv&Sg($&%1Cj}_2( ze|*@1IbXOCgCN#%?^cbn2(e_@tbYdakOxcs_aF9A5Yi9TKGnMH3V1SjS z)jBP}qj!Mh2d%*IGId9dTtJK`XYurYr`?u?QQ=mqAeDTq zM|3Ml@=DL*cZc+!CtY;wVwbGB-Bu&W5@H(i5AyV!ZT zOV3q4{7pQYnwj7O9{xUd_%i{0uKwTj`gv}ZI+9pg5DyHM&@ZIcFFLUC{gNLh7GtLL zE1URA4537p0f9Qj*B#v)R#svi2TcE=ezxA1N2;Nobx>@p;R(QJ(D?~|1SJNnV9EFk zPNWhKAM%+A%NBRjuCVf<7Rris`sO8dOC(fOrsr!I(FbP~+4EVv2$EuHbU1s8xHol= zZdqx@L(^XXO2>6#)IyQi6j!M>&&8LYdkB+mmx=a@%qmD-@%xyJ?wWiHEM7V)a9h_* zmDDugU{_Ob1eU^w+NBp$3~{BwMwwv(`cTc4433L#Z~7h7R&f%WXbMENmb&Su>PnMd zXzO>6PU8;_)d0=9s;PoF2$6fmA!HCBedF}$o<9rL1*zY@fR~9}fZx&{Wzlp;;?p;F zL5k{;3+-oQ8TS`ZbLgMq@)$cT(NLTh~;?!wMn-^{~!gqi7?f6A4p-a z2zRK*sg&S}(w>KLmDr#7+v5r;9-Q<)Hz59Ij|GL5#s}U%Xq@YvbF6_Aa953n7nxkUgljZ~_cI z|0!qEgUk)i?`Nm-hGq2$U1c-twd#4(a5VYx&Hn=~dwr#N&F`WZwff-V4zhZ;(JD83-FjHPa0bMkizJ2^3Rm-!W^{5&jL z%6#&$l<`EP+*h*nPkv(i?yspLZ-`$Acc{yW2V3$>Pt@{w%YGkun8&@$9)iqOL`O41 zqle8v!ith%B-M=?y>s2xog^u>+E-iynk1z=1DW%8QfepZ#bZ>$hzP(z6HrnH>X91? zX`Ttz5%(A9{&0M4lgRNU0q;$LMNot+0VkC*Jg1OQyxLUB(hr%v6C9B5w({@LnXt55GIeT>v1-)ZJrI_-AO_XksMjDR zhkja+?{-Xk7oU%VtB@+;f%~W^5`%(AwslH3K*0>YN4)9h@%5#uVCpWPp1G26Z8=n&1?-G);O z5~_F{;#Mw598fLGG+Saain8J)&;+QT4nqr^?{qRl#*FxfKlnZDXmPGXP$1uO{|LY{8M!y{`Wt2e=owxfeo=jvV3&K z^-QIx@=B{_j~q8{oC}-tlLsbKJT*CDb#DYXjl=G=&3n7}B z^B>B=B7>QQD;X}DXo~@xI0V5tRGH>0zstrUVHi?m*nHSWdG;svG9m1k=ZjAnT+J{_ z{>=aV45gtuiexQNy35o-8I!4prsYi)pBvRr zY+f%p?*_RbX0I_Vl;R8d?-LCKCQO9_Iys^S)muZ&31TTgtgzB$nL}AMr_rDdWDoeC zuWNT~Ul1#4$`#(4$_p&K%~K|MlQ8Q?UMdtO+4VWp(dg zXe%!r3imQ6j*Xb^iB6PD5_m22e}7s2IzP@wzQVwgVFJ@y*)@OwOy8KX^bYnqBi6fx zL#28>UeGY+y%+Cfa13C;Fewo~8|;Kkh#5Y6HVdiZ24m(7>eKZ{yL~{aRtCGxs@sP> zHdDza@}qP(Bx8=;Y2^2j!h(a3`S_Zr0nv{AlZaQL{_u}Y1{Nd3jNv~%`~ILT!*D)f z$nqRHZ-hKCQKLhnB9sZ!K2a(Wz3^b3TT7$l!DH0)ri6ni^}2*>tXN-Ptg^6Bufd={ zOiWQfPx@)KUHcTeN5{(Sma8b;9fdP<6l4grKkzhU6<#hcMku>k!}y&G0Luhgq+PYF zNaNJNW~9iM;kWqvX}cPR0X0;DAw77lgA8AGs&S6zZj5IFr}{~bI-+_M59Lvi4dKm7 z{l(5aK(e#>y;pO#{q@cp)7J7IBtn*z4^-^-b=^qH-shqvBdWg?i4;&L#&o->;QqT{ zKaNRu*WO2+Lg=jBM3vmZz)39c`+0tc_`(MA)zlc`i1Dj+%=@Ts0Y5&@&rh{1S#uMF zl?*wrA~nXd15F2mICQ>TmxXBq1|>zN-F_e-2@h%Uqw`^|5kF~tBGpB?dkjR@yu^JN z%_o0P%xuLKheD~KL(LB?hT@`$3eY?5a%$q2J#Me4-*bEMu*E5W36&kMJ}co0%eY`# zN$G}<_}wS3bH84z_0SWVFDwG_x>GTOBp_i8vl>ameH~vE^W*8r70lO8z~IdOuw0J$ znc?2M_4{7=$&*;CW6<>Mg~d|+ApuIDjRu$ka=eYi8;3uJbl7mLL4Ie-JW~UUtR4F8 zp>?FPtKJ}Fb%91|uZIo?mhXrC6l^%{w-D=<_&Y3 zbu~`K*y9fP-!MA51))iu9>TwRNsKQ9jh25?k#)M{XEfH?ZL7CWxiHeGKIl)ouRkXj z_x&CxISprQ!A&k8z+8S)Hhb;TC#6;Suf{R&H=pV-w9_H#P{HeKr`McokQ>f*=={W@ zEg5*Dt#4!8OhZF0J?aaD59qx7*RmA=2;suIQ2(0iqE>;9p7?dsIkZork|eHjhia7dahZnll#TuFKLVTCSo2+ z=*@dI?nmY^&w|2CX6?-@6FOD9cH<0wI48mOIzAc6=M7k*)0l-oX;3LjHF}XWrfJ+cP84L+we0D_VJXrsf^n<8U zUQozgSACxD&eJ9F%TMgsO1KyhkelIz)+Br+Y4>luSdx>ICo~NgUKb?8dW?=p{e&xp zxa`1am$3BlOw%xTbVk%J5U}d~DtusnOFoC(We`dR=2@R}t+jM72%^@q$M}2VnT@ge zwXXV^VDgEGif;&`70Q(d)4b&$egVUmA4PN4-@(0?^@ze%8N2f?ymE16gF`0pM|PRU zISL6+@zci#t}cz;ai;#+m5$7`@pJ2wM zxktmSR+}?Wv}T{1+ci2TjV3Qu5dGC$vA`%p%%GA!Tk$N z1=5Yc^vxqX8gK@gas@yin_V~aRPZ~uv+_YEtEaB>FbdsYeU9pdwwvM}FIoBElfDOOC5HZrj=UMz*!3He z7!EsE%DQh^KwWc?J_t_`Q(CBi&c7Jsyptz=c`3tiFH`>Wee7EmjV&cjmb)+U3dc7N zq4~)a@9O6bG1JJPH486W;c~Up`ESdzUZDrW6Jn>dpb&s^lau1Z2`LQnrq(Xs8q1V! z`TGZ|r{2ErDL#wb`NjhpUcl`vk%z1hN9}+A*~X;JW@YtPKrC@`5E?1n2&72x$x9Kg zG#AEfxr4}>3Ztg~mxG0h_3e!fYW+7_JYz)q2R0snm_jh8J5}lS{IcIdfVB*BQRzt& zUavxCF2(-vRev%5z=NvmdrG6a8u`tf_PYL_(RG@E`#~u^Sv}74ul=YH=;5#rgjXtt z7`?ra{X1l3JB++KYa4!MV&*9;abW2!swr_eb=#sSTge;?$KE|K$7qXJZVW}*p?p*P z^x6ScKfb7lj|UUI002}iW>K}5#LaUyH5=Z0fQwaAxTc`^f$%{Qw@YNWJj>ZP4F2c) zgPrM(4oBkK$2aC6=cclO`v&#e>!;V^+89UwsE$nhFK4o!k#CDD%rF-}1Y`pl_VB+P z=9D@(ZpLm{*0TtS@so?Zkm&CJyAT7AyB@>sMhMvANJvM+QK{Rnlcho8Ph;(C+Fsw# zE8b^<#ZUzOpMj`?L?gj2o(EQ$Mp4~9-0FSZvC^GH?z^P~a_I}0(z44a>yYg2A^<7x z2JRGS5!!-i*l7Y7aCPS}X!+q^o`|^%d{2fncX$p7Jh?i#odWXGyH3Nk>jlNE=;uXr z1EAIMvzL4aj#Uwc8)9A|GfT8TOv-Azy}K{Ff`q;5c+G?lExvEuJ}uDtG5sxjVoEDU7a|uJB(oRU(+_(m}NAN`caRnJpRxWqhk;QS=_b>OpOOkF4N?j|0ov@~V6TJb6H zydI`~9o8MocD5zGFdyROi$IX50uo?8(P}?D zJ^ig;ziR*7Lu3iY#^4pzM-(2#&wleZL#qQey`ufBO+QN^E3GK3=n`9>sf zf|((W7LKBz+zE}sSW(GY8AJN2Md;_8cl^A4#i<}K@IR4JDKD?2vbl%P5e)}d=`drk z5qMLlbYUr`6b6DH{$`5EM{+VFAtRA3P&FSz*NVfZ{G!mo_@Tb@BUS zOUn7DRR$8fXNRQ4%-UO{%Iqz6w|0b*CtF2DEPNDHvYn9|MYec$U=Gn(Ws*?8WgR8R zls3hr`3?ia?_$L-Ed*F==B(vS!-b2$E1D#6diPy{q&qbwd5`8)&bqcnA6x(|Ot^56 z#Dujg3f^H3aR*lWf906B#{vk_a1xW;cq+(lJPc`^St4KMclE`w2APZ~PyfV}VxOlP zyW%gv><%avuk$|U7EW7n@_>CO=pb!|nq{|DKqD*{P86$6&YHsMk*~Yd-p+`NWxV4Qj?jiFKyR^0#>t*qLYc^pzyM? zTQ%iWy23J{yZdcs#KAkusS2FGh?_W=KX*~Y8HG*jGc%%-*!XQN@w2l&-|92tnW6BMo?JaIpl9;)6^eVbfxW$)E=OgI zsL?>a4@Q4&hsImi8 zU|?se3U8snic1GfpIM;?o-g+yzu#x+4V{HU+;#5vy02y9#7<+#XNu+`I`!E39T%%e z1p=KwrF2RA&d{DVk&}CbU&8=ScbSch1cyXnF-~3NkSW?~dJ;jo#ANnt${?bwl$G?o zXWeqbQ`Q=HM#~?ydf|BewmEz7W z)}(vA*_yCB=MG8r6V@5wqhRt(fQKN=18!|ro&G5U65j|cL`MKQqK-!Is{}s ziF0@3bdLY;zO><4_`yjK(sjK$h_rX(J<+Ot`~B?q{PL)C(}sjHUbRTdDI1e`AYi8p z{547)j5IZBN(z0e-d=bc^nWjlY2a2=`Vbe?pb&|A?N{VME1dqD!JQ-RsN2>o9hg%I z1|TlYZ;MDk!ZNqtY2=o&Q19GK6km>`9>WU8c`==G0Nx%0EYJ&v4fsdReJ`D|b2`^d zSYXowSWml=@~)J~H-Em4_}np2XYB0DaD|OxC-9aN(-9VoUzHCDy={zRb2R-a9s+j zta;%=bR#yu1nBO%i;3v+>z@j*l=?=V=U_La-q z|NpgwS2=o`=Wd@qD}n?A>xFIL8b~ReL5RRo&!7F$JhMjhMI*`xwZ25 z$xd6l+KRNM)QQ-pKr68pr)tr7;H;Zf{V&(Fp)B1Kq zWAen;pLw#Oz2_`k*)BzkoVR0+fdRMkFJ;kN7Eq@X)dTeEOxFNK`r;qX?o# z;_HhX*W37qbj~zB;nBm%o?)ECAcr5JlX++I*vErCA{|uLiacD_N z(LW&+X>w0MT{jIHP!6l&l`p#RV`|E$7gi6T4`o?hm5L1x@$#z2s^amx?3+0mv+I#n_grsM=&Oqu%46$@<%m+V@~(TF-P@v&kSbo!MqnxPFz;Gb%0ZrtIn zt2URoOgkN`SwQ7PNcYE(Pas6#IDu?`KL`Z*D&sGGPtXy;q;c#zK^-c!w!Smr7Afq! z|J;}}Iy%we$JGPG3WsvB<;UJr{{L}lVz}VC{9Nz3ZtrCfB_$z&^8iLs)KeZ&R5D=v zfH9pv_N9D2IhZpj{tL0z6hc#2)NGJaU0MqJopRp9{#-MY( z!o(QjV9CL|vv$(kj$WCrp~l<+RQ>}Z7k_v9xkVR#ivp6UZIL+op!AE!A9U|4W)Vx0 zUrFDLkb!xQK#26|zY)+_NikolfgkK?pF%wAzEy5`p_wQ~BWg3cVXIgDdNkA`k6PY-4*it8H7-J|fQ z#{UN^H1b{5N3(Va24IXTwmck|_;5rx2P?s9` zM#-HCXl=JAXVIsFz7B=0Bk)&NA*;}$ccS7tS@%j#WMSKi0Z>VkSV#~>mSrqeM|zEf zU=#?GK$-snv|!woI$o+q8nt0Gvb>$bswlTF|L-v$j*o5WEMBVu@3Ax`)u7oNm4+u` zz8vvM#((#R>Ctf>7^VU?9SVPM2Wgw6<9Mt?3?#0iAf+D#Vg-SmHmd9AZ|x*Xm)z?2 z&pU$p2)IgBXt}buw&i5Qb3lBZZlO%No;2Wazu|yxO-UmBLfp8lyqiU;>8%$NPtj4z z57TA^zy24g%vGC6%|8;j@yN4B<)jfINuoW(c(eG zk$D^_H0p;(ZHwqLWlpjBD_v@bqa|Dp5hI~Bq5Kgo@nc4%#$l>J!3$@H*ra=MYG5g| z`ER2-Ei=NSp=e*to$nv&EK}8IB+DW~s z{OW5ago;Ks^=Eh_G!$74

PNPxS8tdaNiM`=>mv1d4-TG70#NxF;jl{$B%u zvWyriL$Ii8m?!wbTX96q7JZkbGw+~YY4-Nuab9LRJj2!We&Dp5c&(e}GAZ2`l zB1{c)USk|d$R4IpGFF`>c3%P__2lp)ygz&U=02mI6AL7AM~obqGRLCkN+Ud`6rK6(*3THL(H4Qo zbd#-Pi3hv86anEp!?urZ2R|-g5ji9pfg2h*P*C$ zpEg-1U1k}Do+l(RH=&S`KV7s-(IG2;YT%0HqDPJQn36%RM$mQh0|wxnaHW&6Mj7>M z?GT@jo<0A5)HZ=T7leJWt#xO0`RTqKN9cX%COOtf9HqeTr<;^4$KX+9qlImwP+c<- zFna9atcIgWcQ?qV_Fu8}2Lw=OvHzBZAw-JwAb_8^c?|F!H){yo6rB;`q>Ro36Z`A+ z-_5ip)cv^CSG+mPYhA}fg!T+&(p;o7pg8O)V8)%qjn-1*b2cN-2w_bN;t8f}@4dZV z-mFZ`$&ZcBFCP}sKHU&ULMl=PS@49L4Fx>dT4TTR%!e1wAdLlOhM}N<8`nc{WRVT< zYz>+10ek^`%B)*# zfhOl9>j4z1$5q8Dl%%6=Q}MA7cOfd!vsY@#!coloKKl2+t%6j@SQBDTqvJrSce|g6 z^!~rLYzg*>Ukx^C+iHvta}NO!%p5rk$K&^u4SeBPohA*fE$l) zdPHY4{E=DHgd8rXQ<;$&hf5msdCrCJb?_aNzjSf-qE+UwK|eHlVU06>`q_Z_!ell>b$cDvJZG$aINER&eNXA)188Rym)-b~Gw zVjJ??wX zR_~<%>{G7d6E+8Pn8I4%WD(e3rpkLK=Jl4ej)urd@_xk?5qM!)1*d>gE1>cFk|%*_ zIRO#f+r2URDDd%ye8L?4%2pS7Bg(P;WiC)WDdGDPz@H|jg$^{?RFXJ`$+FAiXbHYw zRamSWdqe8ax8dyUK;0v5f~WK2AZkU?c*2}6yZk4QoRh2_&D>wYJIYYbd6yT5=3Hoe zfl>Xptmb#%RCt@C=7l6{xf*ED43iI$Q;5kY=oZoh)xQsplA&U6ZZM+BFg>&4;s7iq z8O)gMo4d1p`!xQ&(4weRwBRgt(`%4$;BeE6*usHK6ofRRLT`}f->vQP=P`>Ya?Rv#@PKV+-kmiNw3OR?AS2@3=c@yh#Zwyc4A#V|)y1NNF zzVU&^2y3IJ1&SU3;95hCdF_2 z*Zd_#u!Pg0t96S1das$K)&0r`KYi2L00EkbIgQ}MVl(YzU;LepYr@p*L!S-R zjCnNQjlA?dEP9^OT>P^1H}MsVeJuJej!_b)Gh)f^ti3t!P;1KhNXbmZ#(dLC?eFkE zv7?h=|B`N+pEfwq?$BVMcl}27`o*Y2TXmk&^cz}MxXfAhsu}n@oNzi|Z2- zIwzs|TfI1GLBR^7=>SNCiI6iRPNyICoZ6C(4E{Jp7|YVSvbyHGX5ZZY6rApOU4^>Uf0GnM+N5nrk2k&m?~Iw(aTLnUcg%r8K5Ix zs2*+Vs-BAwJRx;CqECu2u}NB@lSDBMqwrn@9oAYBU@3*YgZoNU3hahDEV1eQ)u%WQ zAJAjCr`Q33v~mu9iVgN8 zA-^{}>9j&4%dxhRTmOFk*XlNz(4&3&v|qo!xzqf$I&$(EKL-a_0FCX`r^tjJ`j>$) zd)#wgkwgRQcoIoLEiV&u-n?MM;3UP142|xnuG#9w+7vDwtm|Tfy|VO?Cd~y}56R^u z%QZs6fvrjUK))@@q$rD%^0wpJ(^=uF2Nq0i=zGel*0bwQx1Bt^&0Lq(hih{{Vf5k0 zL)52TA@2P@#YeACJ=TkiUB)BXtL!1Qs%MoZ%w`us7f9sb$-4K>PCU6TC*(G%;tFmx zWW63hE$@8lZj620HcwjEDC?xLpyB40kIi_d?y90;t}R{!LQUpWCE}|cEg||j+ldp0 znzIsL{*tkX0u(0hI@6iEJ2l9RjnC!kEeb<)rX_NIo*eOtltb&=TIO2hr!?l47t*@( zNT(l1d>nQL)LRY<+q#fR#Dhn_{nvID;m^a7Gmn^!4K_`%%Mu6QxEsqn|Fy?AxG&O0 z>5A*csY5Gj@Iaalo)G;!EUbC{?`x#Fb+Y?kbrXK}x1QcB$9g;Ua)-KBtQhw2qW!ZTE7z4f#yXbCQ2>Pdkj6Dd znJ;@@er2Fg+&!FMTa(h)EGtjmazb2ksyiNf=T04Rz$DC9 z29#+DMvNRea;eUR`F&@7>DoncDIBTU#69+0Nx>zuSX;dQ5%!W7vdny7!Z?h}G>GAMo`Gh5gloYYLF` zEO2&?M^7>0>NqjAEsU{RL=`r2`MFasQy%UQGje-P8B-WLc^eZ z$du+$impeSoc$z)2e`3h3ERFf#l;?lj*x)^j+Dc%CB0|l1HgKDb`E%nbQWAhV zx&y2apRhwT6em_Oo;IQ?eJKzJ)ubg8_x?`F6JQwo~yd;Wxwuw)>;MdxhJVv>wo-ctU*ni#whRR{fEjK zX%C#XBNeQ?TNuyoGqWRl-;!A`%KMx?^w{*F57L)%wwGK0GLMDSqnu@>tJHD&`QM>j zfMRp=u-*KjQ0R!K)$Nd})`KTIp@6x#cEB|b)|@auX0k)xOo0isEaPmfP>?SGkdNQ~ zdgd@rp==x*aoQ>&f-*fC8gChaqxTer!t=pA?_5Mro6LfzSi&y-$V*P^_WSAq&5wFs z9p@TTVQv3ONXrUdC5n;=*ygH+9q-9CY@W2oWEfDOEfr}2l_oBHwq*DExK8?dBreY7!Tz;V6G>?#yAEr-- zNGZloB9dIv;q6kIDaShx%Ze{J_RV{!5`LFUwqItTQN(j=DynjDa8Sh;Rq@3f3OZFJ zaqI7$iz1c$8&bJ8XT5x@qV}_Qnask{Rc{1lju9@$dQY*FII zG=040ak}fi@2X1gvr9(bJbk+_EjSb{{Gpremn?uw|LuSZ!c+Djras%Jh4#3*IW*0@ zYscD(7HV$?D%*IRZ zZod3MLvGklUTw`|eleG)^g+*eT;(ba_F~uMm=~^;qoB`$08&o!JfcAN8%mPS3^qyH z`0@{VXqDzM^fU`NM<8eXhsEl3-F&^NcWR3 zy{V8X@E8_T9kk`GqM6g=D-|ym5|TK_BO2X)J21}Z1Xd19i2wL zxqz2Ef0>{C{l4a^S35IamP}WP0{Zs>Aq1htW)fDURUIsBHd-NCag!Y{b*;K4^@ALW>Ly-3?1XzZycFLJlB$=fGB%+SyZ zU@w;lFeWQL{^7ml;d4Hjiu>!JajSM~)#H|(3*l|9yS-V0;-$MDRqviBYBVANNClD5 z4mS4Q2}9G$(J$V^o`Ooh3J=abd16KtRobDAC3EM_RXGHXWMB5Uy1>S!bYZdNaK)vR z5mTqy!W&3DmETHsL>wY?-Z1$oxj z*9VSYv-dvSEYA^?!UoBI2qf|;M8#_jISSH}g;&H29iij^){e5FN*uY4acZWsws*64 z1vu^*0CRH^->ujZDM}`NGMzsZyDR6+otp<{-pWAY9E*V7ULZDw={vS<{O8XKf1}XP ze^H0h`8Bn*K;GxSzw^cK11HSFjiP{a?_ri$a-gf`i zu%`%Q)-xC%+0Voy^)#C7GVCP+L9vds>%5~FT5h7rZl;o7{XAt;nG9-+2!<$D_lO5 z)tMba>H=vn=j4MBm7kxsopCuOMjT-IDR!5R@#}dQHp*v?_+D?*(;~7JiJ`n_!nd<# zMFXrzwZah_2L_IC@p(hsx5PDZ^8Vp=`_F7M4zcMLmbAWbE(M&g7B9jCp};{c6DRiK zuPvpfW@eVmh06Avd@SI_8+7-Q!`$HVYZ_cs@?=sPXX<+{UhFSc4oGL`BN@<%xq&xk z7)aCI9g~a>YT2(DRR7eZj-jcajUJhsQAGGA`8btzCUW9{jTAg3-)fV)AGMj@{AYEc z)jt6tA#p&p-(xNksWI)azukg#_Ny1IQ6aK$j>Mwv=MYw99_&s zs-`97=jZFAx|P5P!D{%ChpcEKYV~N>UG&luaZ@B8O`eD^#*E3^ovfB5Ua6(jb<&xn zw3-sX181i7C5f@3%V=^;7;P$Hk&Ectr_TvuR@z5pGs1d*RQ8Udwx7=SAp-|WkrIVP zPV=@Sp%;19;+q!twGOpa41WXHWkOvtEp&e3#;l2E>I{7ULQ1b8#v{mn?lzR*nY9H~ z6ca{HW<&b)amfoT?G|dJ?^%;$5S)emag1O#^s2Xt&Ko+GJEgx^Cxs`uo(I-V-;nzn za^gK+qe8J(tmaPuwydP4a$T9F(eCrxZ~1IbHz1PQxs#ozNETbC_5ZS1D?u{FuYv26 zO5Y;AW$bC#lhI=8dKi?>G)Cr9M5tnY70 z`(wx`_46nwSgg!LRUZSaF-lwe@2Srg8-*EFKE$afdfD0I;rsVAXMf?vuIPUxEu4m9 z85E)WY#GkAe32fT=HZ zi>r_#3emX0TcRrnGRP0HJH<#Zt85myHd=u&pVY(xtkbEMNS*-2woj|A-Cgf=8$~SslTu9P~A22%A;PG z<=@((<7yvN2AO+x^=GN?DY8w?pTCk?gpz;;XG_6^@fy9d*iR~nlr+Ls?QTIaeY|kt zk^K--EN`5;rIlT^7*vw$p80)47rAwfp zO9ZH398FYO*+Tj!rwjkhLU=ugBJEtl+KuD;KLx$TwpPk z|M)H|k~gouNLpkid>8QF)_n=NJ80E1dxj#rZIrGLglHl}9pAL!z;)%WN$lMW!V;vhmcEFJ z(!=unMr#Ugzxo?DZVWPVj{!~JvSdbGgu{B4q8B3A^MHo=?bM{*2i01=0v#L09B$80 zA_^k*sZXwd*^bz(bWgL-qPw|!Bmnc1d6;MkC96~s=Bcu{vE%3{SLEF}ddV=@3?w6# z^Rr`cTPsw3)&1gNlQ*>8E^1uCf-!B@*)fpKxvHDM0z{#DDu4yx zX?%2#tJnQ{rxrI_vK`Qk(#^ElypgL_QA_S1MXJJ~T9puvX`!tXB z&h{Iv;p7T&uvQLid3t0e>5-R z5eobHCUq)T2AUcJtep&FAHAzMHuQ#ZSjRobGxgj}fI}eN4B4BUSSXbq5qGmeP?J-w zj)|1jtlu@zY`KJ-laQpe9y3(oh{Tp0wsF3(Fx2hJ1`!kt!LbS8gTQ;mw(fQAjF0O9 z8kd5i#cO=21VqoHYS7VLvwf;YuRc!cnB1Mh(zh z6K6llTTV1bsWgX$C$XvQ(Txi?MpK(bgPW!TIxeR5r}Unq3ZY83jIyXcl{3W3Pe(_` zKI#TlPQ_><`uU9rITzCImAL@hbat2HuDb86SEqiT@2mpmP|{E zR}5j+gIUPp(pK)wSjmG`EvBkPGiljhWmAX?`?P6O(%B;t(F`2AWN0?sGUFYJhU@CpvjpFPks$gzy^j0BX7wp z#Mnwz`U@B2OJ~~C3!0q0HPYj|3ot_MQq8g3C!}sMbvKcrE-5`kmV9W$!hiSMs4Ex6 z3-pq?Y>+r=!6+;xGGEQ}@=9eorVwoz15PD7I;1OxC_4!=IHbiO^0WD%YP`TR+gg5gl8Hxm>9wrL0`OJP>F*_OHKY zUc_rTv*Q-0G8Dl2FfW(6qa(4zDPO*a3atH)Ov9Wi6z*n|%Uu0f8#-C>rM>Ep{Pa+E z8NEZl{{0WXoZO9VYcpe{dX1Z3IWm#+XNqWpv~L7o8oS~`viRpH^Ux|E2f*5!+tyT4 z+nmgSOe;N&L-Kp27{2#p@9I-jj)U~p{>+2=;+MzNJef~exip!U&E-hg02+5)yIN~; z?a>EW@qH=lF-w`F8f?Pcreu`$>!Z22+0WJO@8q`y?jM|O{9&&; zG~Bt&)s(fTg1xOTPrn$K+7MkCQqt`rxW$ceJ(BVR%O5Qr(bvL2!`&!pCpDrMaZDdo z?~9Jt6LwXSY{q53N-A!2n>^1$d}4EhCvVD+_+2a5qU#RW4lxaq$d{JU&8NrFKGQx_ z41Fm&EWF(dA!@%uqyg-93$JzOn({{aS34L>l-5tFsGJ8u-Q$f|YzxLgdqoZUu zp=qPPA;R+C%*ZWaxPbZeO~H2zx3NlthRm{+-8pn{(5fVdz}aT2jD|f$qpIYI9Io1% zyX|S+u#&_w<@5Xs?_BMa3YV<-uA!7Zzt~yLH8ss=oP<|%9ic|*i<+a_^)d6-em>A- z7pKQAM#Rnzr68D!I=!7~E3F$l($#h0f?T^^+wF(Ac)XS)?;0hRlPeqBs4s0nIo$i7 z`~1Mhl99Y_Y?R05PpoeBa$|c$6D?GF*Lqy3_!C$vj5;p$qkrE~OwZPYM+S*2sW0dj zW@;f74b+=P@HSMaUOak;^7#%3lvS>I<*cg*RY8wacz~}+@A<;spB$b+E8lwQy^q&6 zTG=3`Od8ISW6{=0ae$}Vb_G?58&6XLN;O*Yi6oX6ed^zrXLkJq`}jT_;9YVkARq#6 zHSNs6y$7o&JZu1vSz`Wp8wM?^@C75ZK>OHe(L7a8=P^_|4R;eFdtA*04m=?NjF{_u z=D_W>@{tN((W30YZmK!i0|bEa^k_s@G)h^CO((8hpWx!d>{t9~(*V0Dv{B@f9|@^p z;Af`+cD?9V3c{6sqd$5i&m|a`jWSwYfj;o0S*GmcskvWKKZd7R6f~Vu&);}qe;RM; zXY`0h%VSX>5FR|jT+W|5WwB^cLO~UnxP5gxSGRfnoDp)B)5qSxu?)bZtXl`8wgr9W z8;9X&+{j$C(TU=Z71O^-z>%@3G$p^MI|VEKdex2dz*sPdQ$g=!HB98#YNdCMdZS<8 zz@!(@U&=tl zG=JdUVRUw5ozAe_wCqmredcwM2YFbWx$Y(gS0`h=ycmPY1Pr9Z>gwtycyPBt`<(N- zO&v4F40+gsfCp9})I5-<5E>0wrK+@G=SG0#zyN!p_v{w^j$-p(Y6+#qEKN-cz`zJ0 zR@%<@IlIxC_y9YaFR`pR`1}^UT8gAz;@*7isnISO{@6>6XZvQE7j@e~tEJQAosOb@ zD(daPDQ?kPx|1d?5n_N(ITRS^^~|XYn9VX**K@{0<_G-pm?}|u*ZR+pKG{D$6V4+b zWzo^A6BByf7_#ZhL9&t*PmCfzbAP>SbY}GL^VhLsN$>%7git#8qL#acE4euelVdcg z$!l65+*4leBmznd+Pqg@`sx8zQ^LMx`rHd`NNbr#xl+YR$qe+YgH&~Ll9Mz9M-^rY zRdX1GVAb6n?uLlA>yYP|ol$iOU(`@M%V9+B=eQIa#%~o#` zgTsdB;}(6yN|l!jy|94X3wj{54IiDV&q+Sn&xUVcMBuRh!c4Xu;g`JJcHYmnSbBX( z{^aseN`?leoj(eaVCa6q9Jl|qM8m`uz&5cE_-ca!olDAbRjRMA*=Fhq>zz2!2w1X zH%yL&{n0aPNiOax3-40Ob*s(O=zU{XTwKmz^g-EZFp1d*9j514k*w+1sQ(Ds` z9KdsX9x}~5Q>{r6;ittvntUmMrb(JEid3VT@$`7bt+>^*2h&ylXRp7GJ{@O43~{Nx zPb7Y(G0+5;4dLTpygK<5{>P_%`UsENHbBY5<=px6i@)t}8!n!BQnVZ~IR6NG^Bkjn zw~tJiO&oVJYi{+TS`%Y~L&47_n>8*D@Q+yhfACY)tpktgI;^*Tz zuO{#n)lU&0r9~>EBrK%K+%?FfhxGR|JNx@=af;EL;LR?u3yh%FJxz}yeq4(zEL3up z%)Th!G&gXh&S}8k-fW2JZ_Hhde3Dj`O*9iAO~7zj<4R|p>vRRtoa45H;JYQx#xLRM zOO!z8S>#RS87s+&i^dI*aM6t|%uo0ITrzEC1455tP6?B*E7Ng3*Y@FNA=Lpk%C#Z@ZP#V9_3I+cJrNz6D!!E#7m zgyl*$x;!cJSu^wZ9N;=k`Q&8>u^uq$uvYM6BVJK+mWCKXUO=gLoHot@(!6_@pLj0- zyI|pEGnd@lx(+dHwYm!aXKKtFbN}c-y;^4WZTVt}XX4+9A>3rPLPQe&NTb%rLLy7yD+L`((nU!}4 zW@Abc6TAz!!OhJN`m0OQpw&SBfWyH;>6xn1PNZ?{>{jEsRPe&UI9JbYDob0MmywnV z@W$DB3dNh1HIHAA7@`p)`3Gj2HcdM@($YhF}f#;JU;t}QMtLoX;M0jQvSMs z|Kg#&7f_-Wg3uCI1?uwsP;Xx@aWsaAtejlB`KR#x<98{WYbH{c(O-H|Y))hrK;4i0 zTTmsP3sa%AGrN@SNy9xRc~3CQkKJ{(;?JE7_)Nu@rQzQB({jq0a$wF1otZnsN5v(Z z@UJf#CErEPY%BY#Yd^+3BGDkNnUOet3;YwkUR`;40vxMb2|HTD)DT=E8 zN~JPT&+b>j+F+_}ha8yAe=t)^&io@nmx0JuuYRPr<_O3g_skmyszwKt^GRH947~Re z7?sNZl3SVI^p;A7NPz$NcY#dxAjbH>p>>x5EVxh@IqPF@1({Z;u7+A;zp_8 zizrjVk+VAze5YVZ&+*pL^nI4>n*>7fq z`9hEJ{pk5(DY!IS2Gn|X@J6CoXYC}^vd3FT^i|POK|I{!t&8k~W(DBp_UQhPww+fKaM~yt(SBn@9@EQGE4r9wFS!(q@ zM$PgNjE5XREH1ExRHXMVZc7?Glu?4FW6vbNV7y|`?6LRiYFR~$y$8yoZqqPd`t(b~ z*5riTsTA%E=vz`r4O~Op7;4}gvmJ<}`yLa>NDK5TmN2r$7bxzYASJAlD#t0K!<~5n zU})m-dK4=1FQLa@zPP3T0@&<8+V~(XDjcD93CXFT@O4g4ol2gKcXa+x4Vzz{@ivis zVQ;a6-7Jq!9~KGz@0-i379f=<>^ZsgSY(LT+@-c|hmbjCs~@#E(KnQ(C2?4%ky6l= zQ%rg5^zdb|=d8~?_5&v!)bG6=tb9Rquog@;r#5S$&eK~tb$G}$05tmfpM7wVrQiic zRlwwa8}L6r+267*;TA2~CxWnhTuv!&R9^z1N)+<3kv7Sj11ceHPE!sEO(;Sny8(L# zf5C){Dx5mg3do{iOXej~AnoM)NsX!b-;`N3tSw0#JU6G*VJadYh8z|U@e*un-&5>P z$UHLP>!O<24RzhoT|w2t#2ja9tl*$HkI+zdq=XpktD^B3mDOwIQ$lruW-ppK<_1a$ z_r~U1`kpZOPl0@gw50)VMBcMy-u+i1|9(p+h@O@Y}uCDa|Z zD0cv9tNz>Cx7WVvmKUWDQdSRe>YoJJza9}!Du_kS!>@a*zt#Tdgct2}EX36>W0#%1 z5}U;7@9_myy&^x$xp03p8Vw;T;NX0${4)N#HrNZpKA&xxRMdeJ?0uwtq^7vZefRBt zqqZgGato{cgZ?(h81kwEurm7TjiA#8N54Oz-m;#5#gArph{BEEbH7@feasj)2}OJZ z#e~op9ww!YtqnZQVgNHOW#&D{{3CN@vaA5l3wnU2Y^R63ey073WD~%1iG0AyCuh5P z?Z7d151-#BrJlG_i_GoU0|x|q_VJvmzSE$Of)t+6;^#PTie#7+;Dj7UeB^ny9zP}B z$~1W6(eN*FpE~NjEjhuS57Q_ubT0V79q6OZlB!nXCE|cNs=nla+w~;=v1t$W+#de3 zkfoc?@fvf5z~FxE`37Ykz-^<`wVDg=KoEe@n&xAg3&1+F8e73!7Sas`yi8%0QiIIj zD*h^kKw+m)w4S2j8i|%Pe1D*E-|i{;p75y!C;)bl+fthnOVmNRdv76!)IZnTf^IYl z!i@UDH>Y+H@qJ?nlDQN#IbL}#8_;^e0p&?mp!?;e=7;Zp-hGeMPvwLn`fxcC82 zA0S&HA==5$`g96P3%Ys8y$Fi^+d+nlfD39Dmfcr>p}|svvOElO!^z+?F)a24)20xm zNIu8O1R%ZZPCqLnEy@mMQqR;Pz}yiiUSOtx6ONR3!D1|nqIUkh1p5Odix_$yG;Zr4 zA;3q}AOWp!z#>1Kt5*tkLKLqqsuDh!GSp?UA2x(GNaR$m?fNfB<>wL0bfChJtUflS z>Qm82t>>r&;y-RpQIf>E4|rGIeRBiZ1DO!G0~ryyYTO8Qov)>oQWW`_b6(}-4gVZN z5hN_0#Wk(8{Ma#SGLUVJ!So8W4Qhs!LYoG^fe@a@#F}Kz&Qh_Wg)uczWFQAMcA2cM z^)6WzU^*?_TXVT7*aJ4cGf9N-K<;U0H0RS7Fm8HA=0;R~`}I3X-)5>3ITeY|wB2J% z$1iK`-`VD#t$y2=QXWtQxcpS9S{%PtDcT-j%3hK<6xo9|tyrmkKm6C)eH%-nZPOA_ zHJ>6*--sX0k|Zk;qNAaubb?>SEk7R}!pJpnw+W5XaZ)%ZXqznxu+#0^-2p#|_>D$c zV;^U~%wC(`XBq}Lc?`tH_uH-B$U`#{Hd<3<7gCmMy=6uu{lSC|RT{3SgPr%h`B9m+ zre&jbfLlT1Xjk>G1vPFc!BR0zdQ#<^qN=4fTlR1e$ydKfAXeDt{E{?GfFXW%Z@+Tn zTzW(FX*8KeXW=L0WCF_3AlF(q<;X=5`=u@>q9?!8a;uhQvv+lJIVZKO9k3VT7%cef zqbyyfXk_W)f)tQbYt$!6ErYpaMQ)NE1@4lQ?SXhJGPAM!1hwL4;&+-*zlw5kCA2QuDML6;$=#=%WX*OM_*-h z63Uk9H;P(}2z(S@0g?8AI(SxY)u$M-gvYiUaIKr;6V%M17G}-2VsVbLLWV1q7jr%8 z?NN2R@gw;yes(|9yS;y^J`$BE<_(fR%0rDnCDHfu&(C1+UX$GC(GkPY6gU_+1P2Y2 ziba^DO3rtn*OpFZ@U$}!;MCqHxQ|8@benTlmn3d8Q-6BjnR{1{;7A|pca_h@Wo7m8 z?jwV9CRn2dmtnUtR`D}=<%I%J@O+rO?a~Z@z*W7{ad}*#UXh??$|*a|?q}?JvncLH zaPg8jnps{^i5(WiNg8sTxD4vo$|E+LSF`BwDaH9glXg?3%-$ zvo<5u*ts8Pl-9TeMnoZyI1O8V_w{!3P99YMMb7wI(CGTJ{pM*jr`hqPm% zXM0vKU=xenXo2b}Qn@;4+kCEo+qhGDQOWNGE&Bep?6L>VPma9Vej!J_O&9bwXmzDR6CjOSW;YGk zTXbgS8Q5&Q(XX$I41cW;b;q(dMAudt+qBR!05w1^mut=awR~hyUHL5LPVgNgw!ELR~@67<>ZI!s+^mhYcg zY1oh>pQq^SN=ycNMJg0OI;-qy6E+=1k)b;qeE#NOt+rE-J4GW?f65pni$wxR== zIj9WJ3R&4w@?`cEP4(09{d!rzNeUji&$QNa37@QZnYucba z1z%Yc7$F)A)~M;t<$i7T`LmBs-n3->`UVG&X@_6O-pC{llZ7c;bxH7v8_T#PknOr7 z7cBYxp!a$l?e~X#QuaG|-xP^16w;}l%>slgSn}QfuQn1v+BG`Co7@4r#JbN3P#*vt& zPHRMx(}asb2AfyKm8+Ae;)CFX=^cPjdDM#|vf{PuMtdFj{r5t7CB$4nbY*0EZrNGT zP8^EAm06gyBc;-EaD_E>wB`H+H3kjZDw2LA734$WyPcKb`qJGqB{TDRagI$l(Az$ZgKm`;xEL=6%)W>Paw?~C zyCJ_w%Uo;!MTuuRe&3e!jsNq8m*`9_jNdYQwN4lRLFE?pt(o6p-9^b5-rc8TXf44| z;c>qGwjN?!<(!yXc6aw!)V|zhE~Q&po2(LE(^3QrIseV-3}b!Yhb|Lx9iF2i&hsPq zY!rQ<6BsNt~!f%z$ud2BEjRQzZxXL*!PSev`lj3qUEgLn-jdp zNlnGTt&iRn5Dtg12$S0amhTx-+YL~oil+pg=QcP07~mTYKX=Xc(uc&9>~5ixNt@EI z7dIZF@ur@ZE?n?;Ho#fPstAPaSp(J({${NCRUw8-Un?DEj*^oKZW=tl2vUT>!a(zdv=md{y_SEtIq))H6-&zEV%a(4c-rf&HV`5P5hD9#n-7&yW?Jsp)b3 zQQPO1eoXTz^OyoH4J>td;QYlA$0x{;Ou<^uKu53G)!`vg{2yb;7S8+DA;-SF-oYp#!fj637^1av9j!guvQYIOd|FY9Xr ze1wjQW&;PTIk~@QrBDX6R>TN3nd@?^#csu_P_mkg@Q4^Yd{<`XiQ~sbRxR|6^=~!m z8uI!kpO_nrW1d8HQxEQP(ACv!=NIII1)%$jQT>e0oI6krO$H+>y@xN^{S$LRy^H5ldnm!Ce5pJ}n}RCHo)bjopMU;2F)afkM#>~X zpy}cmNnv4thQdloP+>AIiXw~U;gs97!D!+0?SSz@(|@p>e=Q}E^n9#`-O6Rx~ zpnqVS(y{fb#Ys-Kw(fWBM0$prClU>1@*}H&YV&X~jO8M8=zn3K$*?l;Rhhr1%gwnr zRbAx!e*Uz%%zD?W|MuZddyI$4aQC6FR<1zeU9<3uCMoRS~SK965;cecX-eaLpM~UR0YhQQ45L?=os+39@&J@anrfe3^1VKMw_5s{2j)5Mo zvs(NH$fuDS{_CGS)RAWTeN)5)Bz1js%pFvfR_ixK&Pq~CGiYo94#8gqYQ*`px=%<+ zRP|vRYFR%(8HcovMAU~A$5VeWG|42-h|ru!5QGjIF5zJ;EG$ZMI37#JjfkmUbh7A! z`iKav?SQ&)y)@82C`dhefJKV6x-+%02@*-+RIn}`<0O|?tdcO^zFI)*Uu_f=cmelk zp+H%@jNeFi^gTear~zKy1xw{Hhr)9Xi$p=L(CE`Cg7j=&HtaD#NG( zDHpIrL?GFgz>`udL)X^-x)aW&5e7bwd9Wy1?aYQrj&NmEYqGcets_VDp%2WndXc1F zbq_Wexz_p01R;yX0rG?bqsctP(*c<9pq-Udt$yuAD2HgK>UqllY{|IVO~&VENx$04 z0|1@)+xD?jU<9d$3Az61_Hi$REn%X4dQeCpP8&k_z{DZ)gf1I-0h=9(dQ&#o7P5C>qH z)an51>`NI!Lw4`}wzgeNZ${yzCDfB(uUbmEOnq!mlz9NY(eT^Bn=48VK3UW?Scu8( zdf)xkm7I{fqxm9mi5I?*9CCN3&k&EbNl&TtFqgL!q=fx*4j^f??VT}MH}!NsM-6~h zBbwE)oTE*4;`Z8hfww@sT>!_Feaoz~f14e2wuxwonk6KU+PPt@>&JUc@OoZ|N{9nb zVyMJLl{&R(;8FfU6;AilNbsIo(Zwm-mVQ+GdU$a4YZc+jq;#Ubrv;)QaWEDM;cJk4p8bs_Fys zk&C@cf~!VGb84fo4D+~rFXng3 z+Gu_uPBxSP=G-)X=unb4Hnr{xk>wD2zWYKnp^62^aG;pUy}3N8K8=vTUwddqhqBva?>GN`m zZISx*)}~6oN7a^N;Wyc_NEov%j?VTwr}XWv3kOK|vj~OfHXq0tbCWtB5dsLo1G~^l z)sW^=qoSm=l(&cwH8t+yv*m3A)FFE9hi^l41E&P$1JOhys!*2}u7#FML5(qUM2H3Y zEpxGvuQWYBG0{Q7Xz=-4E?+_erGTKU(05(mU(C=B#gjSD!V)~wm3JNbq)lVs08wVK zqmB!1%K&07@uXB9ZAnMzYH8>5W_&y~9J%HL|1;{yvHO+OEaH7oq903PD(Q#-~y+94^ONzALG3oKJo2jkF~Ub}sM%^IPapeBV*dRV6}8AkM$rzmxf z8YYnGHFWi{#M14hQ;b`lw~73PmT^4`AX395=i#!}N&47{z#9)mO63ReRT1Zr##!Am zgt0bl_yK4~Qqeh_ft8W>yOl+y9kARL+N>cgWcH?4Ds@|--X|QFFhi&bEoS)$O3FO8 z2a+)u5|W(1zrXA9a@GMHt|g02xIdYFez}7}hF+P|$cu#mobb&;RUiY5$pxQ zjHz(zvM1>_b|Pb^?~w%t;(B08fA#yXjfSEEF{62BOEz-f;@#1&*S-VR`?f)?Iyo^( zLBIA!x*rjIowHQ)S&a95i@x>NvsR7M_anNf!K}VBK`wZkXzEGT4HWx6>z}X^Pa|F! zX`L9OWf!GJ>f~WT98RrwTid*KrMj91R-J5>6VWuU{B@fcfLGN~g)^E4NH?^b03Q*Hs^0nSK84K-vdK7Wus)WuReP zYM>FYdO#{#Jgw@O;cp)8dUG0hcF1J!`_PFUV&>GagwNmIr&GW%bpG02$0fXTzium%jpIbhs$X%I@ns3LG_aAcdPT$g9) zs_*-{Y4(6^^XBcRxJ8u28xo3mLvh;dy}2$4eMA|6afK!G-&?4zgfwz8=veF1BP=wg z&lEMqQ}w~yGHeeVX_!T$vxujOvvPV@H~ZJYkMZt^!bJ=iuV*Gl)TG+xCZxUB zDR;0xy0N7BN-OQZtY{Lb2=%dv!;)n&2!KEf^bUgCfnRm`$*_IM+xuC2C}sHJa2Aat z2#&z(YQbtAD=3Z^#Pw(zZ(X1wH90?>YS=SuQ(0>3xko~*ULX5E`gv^)<1G%%-y)QP z^yK4ZpD|AdvmiPK55?=nKRF#S3i}G|gb|J?MKtrHH7!=)v9r;JhzUE4)0Qpw=`@d}L!Fwy*mm2ALlOS-;=-IjC zcE4N83Chy;bm=kR7`$@PjDHDoJZMNQG$5EA^j)4~S9kR>yt;+}w8vP_VS z*gP+vMlX-dH%6Fxb_Qj+1S|7K!|f5sxlG5V%v3Wp9!inJ!srlgP>AvU+STJa=R^$-GzN!#1wHHb&Ne=HCr!Oep1AVpE^=ko+H7@lwYsRdY^GQ}6bu z&CQL#9K~41JcZ_tLg#JYI!NeOx*a)*&QruAxVkKcR$v`XT)8hf_}F%mABGwk0R8}P z;3H3-H}F-2f(ED-5fb%-Q5TXl!Xb6Fb5=Zf{)Uoi>+htsv{C;p7RhCJsnz41F}<;I z_0|$GYdb66r3s7Dj;xOgC$M&P%KkT`1rI4pLJ>FILE|mtXs;X-AN5yo_iJ|qLe|s(g zb4xwL@*qHSZs5M9YW(j~2K^0bN7SS;mP!{yfunAbqcN)#mW@CaE#sRqkaga1X+WqWWgZ^!VW8}$q$HJa81;m%?ABeKQikl|k;7Ng=@DCf zqHp}xZ&U>2u2_gsjx1uNg>_nXkCWMXeUL$NG2|3Wa&$t>n5YE`4YWHt{MA_SK@ZfDX zGY9VtlZOJM?yDtNY%&biE+y@%I)3{W2c@=?g^JkN564;{RH7t2Ap$<)SOa!c1yoov z2`r7F4EIVLLQiw1cvMLup?;SG4zy6tOK@>;9>b9$;@#6zd-!l6A7rg+E1~>Wd}Zn@ zOHWbDlifvC!o=R*)pfErIJ!Y3R{{gbfB;n`MgLdz*i3DnyankX#WtY99Pp7;sF}{M zkURkBqXtSJ#EfT3u%-PA`S_p`!QF%)U_fd4g{@~_az>yC_fTpk6)@Td$n+*(#>be? zwaB|umVdyW-3AMMj1%FiLdxrHyzbf^mc_7yMCI&iPFk1>zmhMo0bA~B@WV+xW)R=* z$i2@>cfmmQ9vH3!3RstrJVW;D#C5j@2+7LZ^qZh@kMbGS+qlVJsmuJ``FeOwO@(42 zg&`%732Drw?qewof1uIo8JTboLz`Cmin%wZIH_^MBlaO8h13piPxh>-qymnhWFk z7qi#ARPguLGT^M;`ABb777yoBcjv1BUYRtIY{+JD4EWnIY{B)81gzP-E=occP)nE^ zSPc&Kz8zn-<0RYMixbVXSo3BUtkcQCm}ZQq69!7!K_M{!DG8-bCz7?b1C}b6brnsJ>Lq2L(x*g20xyF&U{MI4!I$!|j zmG)sG(NY{cW)MF-gNIexv%<&Hk8l(XC@dX=Xqlk<*G1hBh#MeQX%XjUvP%q9iq(3^ z&66Jg#{8q=lCBY8v!JzLHd6OQICM6Y@EaZqLHN~QCl1Rb)+i?^2j>`ON@10d9o-z=_PqrH!5nq&T z;YT@V%PHmy=gqs-&CwbBUuKH=I3iv{^cW@Cj>~mLqERaXgjdYUZWVQ(O8FOxSjRF>cw<7(kSOQaDD%j}{ey7am27XWUXhi_M99Ou|O*qN|c22Q#xn z42qxwF5r~peq(Kx(FZ24-eO;os%48XO5jt-xgdoqEsw(M zouLn=#Fjx(l`CquQ#elDoi-nGzm?g7=*<;ZBz)Rg$-5H+Bjv!s4dceEZ{46I$ZWZG z{hQa1-b{osXluo2;q@Ejg1+G=Qe!lATduT%y=YGMPKR}oXD1?Qtx22I?yy0-XbNNi z3yxJ=392JG7nL;tLvgZQmMSpsNb(PFx#DQ`_C8EOmQk6^u0i0l4<2KUn_F7 z%(Mf`VWv^k5wx+yY3l}#b`sMx4#D|SsUAJAYJ1r#1KO-Wm7dA!yWM<*j}bW!T-yJP zPb_B5aU$l42-M|8Gxn6b)P7y4=oxxrAe!_f2AzmgmRmgg|F}96xSaFu{rAi;vW^++ zkUhd=nZd}CB@9`TY%#Kqu@eeWG`1OIgxhXf$-ZU@A+lwytRX6EB??KD_`k33oq2xG zzt=O%W4Q0{^7)+4IoG+a>jXV1m6C-@K%N?Tv)UU}Ne=Eo6$ZKgo|jOwRf;JTJyO~t zn?TtP-AgLo;`mQmEEZEbJYA4M+Nws6wIfa8>uR^cl%)(tZ=8w>BW z?RhtoS%GAohGg^>gZDO4TXo&=vyqMYSM%$?B31-;{Ay{jS&zKj1D#y`&MDKzvf$Je z%BE+Hul52hL8N<_+ijyCs8f8Om9ogk{S%K#u}@YINTvrUZf33!w8Z-&to1R<^1VmD z+_`_mI2898$FiIRqvITLm7IECiRbVe@cZ-{sIQ?RN zLr0l{9oKX{8% z73$sgQ1%Xo`rH43?M4W>MmLcp(A94z)~3UrbD_mQ#XF$$N57%%4a}ta9yK#)k%~WF zVQs7X7bt9?a-gNzOdK76!eW?TY)>15du-wNg2XDGe5B!tq|95fat6BB<6*;NFCBKd z9qmVlrR|f<$t^=OPAyg1QU?cAfMwQ4zRb7jy`x_PWsx+n++!U9bA9g-y{*&a{)VB( zgJ~l!eYkgj-^j&+7^KaDet3gg`~HO~qvj?)dZnhjVq(=<&wps{J(#D=Ig0;xS6p6h z!L6!uY2VuMMn&nLMb=@RhMHF?>&OAOdKJGs!7OU*=8qU10xw->8`TaN41q3=#pWgEmQ3X6 zms{ow#WE-cuPd7>>WTZ+oBHQ-T(hnc4tGmnCxIE7wfnI$vQ_)lDXo=%=?IVKPN`bZ zD~~&+#?(BH_O#2zwWX^Ac_7lhclHCFE>5mW_A?JihuG#)z{z)1Q?i~##3QUqG9)%8 z&yt6=1&z`eDof=BE5B*elkjg(S|!Px7YOBOem{)JuOBMW0HpmHLswK~}F1s&d- z%6PgpI{s#5WcvBi5s$!_P!($yh1fQ-s{k5&e9=cK;*mv7bq4ruydC~x;UPlacAY_{ zrWnK0a3Or>?1!Ra@;K%GZjrz3eD+anx0YU(`R6-LE_9QXIrYc3D-=EC!a#VvR=}et zFNCqhlQHCJ4SCPXZ!DRdy*NAQ(}&JOe6m;J zAPjj-abf!OTfG$u#q0$+F+HBv^Pp`vlFrOyS$^^qOS^G8Lm_75>^gG*OU3u6M5vpV zbYlXRZTUz{!V#5T!H7#P+6=@nP(txBF=hJ&M62vCy6qr0|k z>5@^qui)ye!4n25MFCc?Gf05XLaAH2a@>Abb)#vV{vVc24|rO zvz3Bbx%j;t=3j20=1VzcJmVuv&lvA|#NDIQ_sGfvh>tUjkwi*CSm=yIDnQ({tX2EW z4@pgX0EY3gRBa1m84nstzA=`pN(7QKf$MZ5zH?*JBI-C2B6{=J{frd)DcXxaQ!m~r;bfxX-2y{C9dqPmE1P*|3^MXYi^2m6mOsU{!sAzG1J-|Fmh)A-Sgc3x z+BNmzd1GVWkO7*b$e5Me7#;m~vpW_s&ZlS*0cgy+NN_``%Tsu?@UnjuH`1-`Qi*bC zw%+76c&W!er&IwzC`KBj-I%_b-1A8K`BAUMmyonbydZAx07N+3py_q)u=8BpV8h5! zMTQKZfp1nGJz`$w4O{lI@wna;=|pWu)lqCY>nBb&Z#WLeA0rV47WZk$2 zAk?E#!4n<;7yz4fii&(n9_@N9ZBKtfBR-Px(2m}K@K||@9nfyJ*>L(~QJL5yI8gY) z=_$yzN)(+CbZaXd$gcnpoGYv68+Ht9l6^lWqFM^a590n_*xlyLEz-- zNYl>yv>0I7Gaf-KJK=rC_=RR0&yy!!*ECLA|Ke*GXve#|2*!iVDPL%nF9x-oPXl)c zH5o?-{VsOvyM0CxpT$-M8tZ0~y-cY;ZbDOO?j185!sg>4Ee2S`4)-#DU5ugPbu?)^ z^B2{jf%ww1v33N*y2(5m47{$d9{J=Z?gRWdSO0bQ1p-WQojDm1hmqTyNNVJWk<=n^ z%G%go&pSAd-&gmg>m5Y~|7idA?kS&a25y~ptaOCB-jXkN?b0QV%bWTl*T3{kdhphc z6cX1_976+deW~F-eG@;vb03~BJrk%#hVcBjm!}6pBuPN}Vf<3^vEOcgc&N%Lrq$|^ zBh#8jFhXKgL$};SMw58Hon^`Kz}7gd5m|`%H?r0y0l^ zAjnJmAb^&$EM)Dd#DlTz)yWWtk^i{=u?nFqZO{}^UZR2x_cZcTWmiYkBHso9hu8@XzrlEEerR- zX2W9lKsk%Rx!l5kI_on5Oavc~ap8V#F$AE|zk`4veiFpKsdXAw>>M{>wdbjvCOxOn#vhD)2RP0%=(KLu zA~^Z{Jx{)O4E0bNmvLxZyygwAfW_%`?UhkMTeuJR(7cxd^+C!KcwibM(1&Gq!2N*( z7&G2O7)KGx!Ieg%n!3|w6wdz_Otp=H2*toiL896oS#dM^7wg%-#_#7IlO(_SSo$Wf1z8RJDiM^#VWK<<{(X*7ra% zkpALzlR!6_GKX!pLtSiGE~A+%A-ekD84n7RlyeeZ|1drC{N)E(XjKfCQ_iz%*RDx( zOy#o!tjf!?vYxF)a*0+`Mnqtn!9W9u_yrN2QHXSCdJi64-}p_1yK+bY8{CkQvzO1P zl7ZHggR;3FYW^+DcBC4NpR(@snFV7|EjONU@9@=5l<*V1&TQUIB4Udj8g`)s|UkXacUG(_UtiV_bGmZc+iG7yPAIP~+g`sI7hW;SU7Z;M8bD(wge zfm!eTjH^7kPH_*fN8G6dYM}Ig{Hm`JE}UOlHKOvS$0LZHY;WC~HR`rLZfWoz+WqJ8 z{1dFs9nuSMY~kHXM@d)!)t}U~^u~_{T`9?A255T5?2z9-!TFb=S1V#1MI#cII2GF$bbRWP z0%;ryOA*%|0WIOO>jset+eHn(y)#2uT(O;+Q$ytx^KHLzEWmS1tL>pLwczxFCf~Zs zsDqM0^?g3OC}aA3mN7?|kokgTJm}LzqIfw9sH2*&Lw^X=R0;dM_Kq~$SKZKfA(U+e zFp=MPZB9UNin;UhJuXmhAR0=J43Gsv%>y>a!HJx=6tAXGAi!h=B^ zE=+GWCC$4dO{PJGcsw1BaXgIaKuohloHb^>m(O=L9S!)mOm~07BR%(J~|l1P`XgDuygAY*31}2wq%0#F?hOV0-D}#_#Pto!)tatC~WE zk0BI5F(iW`O6aV2#f+6`zQ9q}aUzwyzbXy}Y23i-SM5YczNK0VL(P!<{MFNJ;~YD0 zPkzS)0~961jy0TtMII2QF#TG*{d%wL@%7-zNk}ZkAvm$p18Y1SE~Ia={f~uz5|5JW z6QeJm0yks-U9W|pvM__RAc^UM+aym(`FJGR5J-p8^e6s(!y(ii4*i&_Lu!0}uEOIL zzhub6aYQihMb$`tLRB6P3Og7`YJX&m_&_I5bTQa(h5sz$&2)37 zh9l?4QIqzi4C(%@EMTnE^1ya&Y6I83?O?vUoSQo6V(RRw#TFdE*AME`ES8IgHH>md zEm)Z5i}Eu>^c%U@(FMdkxMp-xM-Gs}(OT|v5y^9lzkWAXyCTgYw{awkUq_Jp>?i4+Mr0^N&M?Xv>5BYsr$nUAz7T{$s|Qp&ALrcD&ZY z%;Np`A)UW`XJtM|c^ric7NOML!h{i2%jBFT;pM$bzxQ>@#XkDymNVfK99@8rL;tpj z$hr&bTgFS{~Kh<^?3jl(uJ>AFekt@MkiX&IpLB9}MmUu6a zNT+m2WkYcP7jNxv-bn6({P>Uae`Q83-G7t9Q4Ww~5a2R-`LaooEde|vbkHf;wRIq+ z*cY5T51*GxFBCzpT-h!N7c&sc$v=NFmbP*)h!IjCIo>HwqUKwIlY-|YKS?PiDQ{GJ zDmCH2giw?+wc&gxvk&yLYE_QR@Bs>;L|H73gdmfOC4TezG^hmRuPsOCw5hPiIBVQz zzaz^fNbw9s;5+i9ytWIF4Z$O$<>9HAbGv`)ZbZkqC;y2zfbQx!s|d+1KW~{ls&O7r zlZp%3I9%Mfb5Pp+j1q4-{CdHd0zKrd0$knuF8u&_So=qYj9je{z*CJB5M|tY8wawC zk83-!Qzlw6?I5Cthmp3z8{KXHsAmx^UR9+}YzJ3sJk(crBp)3vk^hOpibu#75}s&A z57lMO(~Vzu+7%#jye0_o`I~^X?g!LrT%y*VM%pPzz$T!y27-~Hv{VxG z@ZKFLaya0T{IJwxm!Vb?iFF_yiGnKL572)Dg&LPAN*ph2!~ zQZIY|yV%d~5!@!67`^glT4g~azl^`^8BWifBQcu+0Mb@^d0l9speyboMg9f(4tOh+bvkL^N54vb5F86r`DQ10`7+VSXNn5I1_lxWq3jk>hFnHMchiWYgfB8))zZRBz8k!-e!??Ba`7BGZ&5t+vje2fa6LnqDD zt>N;OS7$sKG-c=d4|JJwGmd6@5E!1mx@kO_r3Tr_xgD(T4r-KhW*!~wWRj9vY7rEs z*XxzM+iLN1#1)*@@)5(3%z^`~4_{NeUjHvLSTlv_P0%S;N+@L#vnw1m%?*l30wR@0 zcvH!uHy3u#Ki}7QBZ?~WtefPD5~`$9;N$yo;p$X>NWlD5Z#(7Oi4#a^F*O~{2}SRN zqtP5|rA5-bF#vn1MZ+gh4<4!PMp~-BtTgB}Ilo26F#1>gOC4>`w?~TB?bH}ACVP$SUCQEW5{WK>t zzSlb&R)uv>0v>E3zr4`+Z`=F*Q|OtC>TeiN3~hLHn8%hTmo>zK%zTK609)IU=z4us z|H1Xrk{l@*Vn%Ves(l2XpsG4jvNIyb+Mh)vzD1%LN>wv5&wpJ9GJ5|hmOhj~Dh|al zV+%>)Mc)gWQ~s(KDa_l^owb{djS@BlWQN(i=&^W~WY{qWu8)Z8B)VV@(?y9^xK?tZ z@sQSHfWaHubqGy%w`^3@-0sEZHQsRJLXQ=?qWc>xSi2;hZ$#nchI%7^HX8^DeK%38 zLjSJRnw+vxqbAW#{Z_;?d_(R8!X7`jkw?YI{D=IB%kBvWkOtd}u>^=g6hU1y(Mt~k zvc$-=m4WNvahKY^dZauVu0vg|o!`oL_SH@m(pWcr;TP}suF7jPb`!>}A`uj5pvhtle@5*Ey6r1AoyjtrSjem)6S8$FGX;FDwB=@X0ceMR4XeBaEe44GAMopE+ev1Xk zC(2RbC~@Yvz=@|m`{h+Q0BbI*nP z+$skUR`{XKv3m^jJNGtbQb<{MIf$w2DEAhTI%giwX!nt&uuFaeJf4}#nPhA9>8qFk zzime6Fjb==*>d+{QerG-;)AMhrCol0-%jtg-;lYkTMbhbUf3A)Mrx&!e^hRxRtoBh zR`;jI+AB4f)!nJNJK^qb(GIBs$w}zn;JS%JgoDeu-LI)J`v$e~@mpV7-YkG>0`&sl zc*TSox`*l~60jXYCDuwPl>OO(YL#`&C4IU-vsK&XZ9)vW-}`g_?G(8`RXrhc zlJ$`)&N%uSUhwk$L}=Mxiim_o*%W$LAv%goMz9^r-Z57;WS{UHWm_!sfju^?Zo6Lu zJJB*6qv4jMcCC6rbCSYQpb&q0T89T)@;(eg2`-`(Eg|4#O1O5_tJmAO|Gl(Kd}&`k zR#Tb|42}B22g#G^qfcP;nKPr+TzUA{_xB3Fi#~0BfXRazvtjN-(8@^KbL43y&2wel zMc<#%AS<2gKEiX_f>f5+c$ADyHwqB8V1HEyqftUq+|9pbD?&B7!_*EOf&w7ko631Dzd~N(3F~jj$o+FTz~1mZvUtj!nZ9(5&znAXPYShI~?LUs{zO_ct zWhe7I-Da)3yza=TIz4Y^9?6W&@6oku%h4~!t^dv4hc3|E^a@ou<`bm_LY`gY}!L+UXhu^ zp(DGo*X~$T{)*zl&@HL@`iQ((6dPo=!`XA{LE{w$8{T(B!%2qoM@y{BjRDuCB2y*; zivcavOUn6rz=~*x6bcfUvQe*&cStQ>-qx~Bt7{<1W_n^*ng^vSIV^4FJOJIVh?%)t`8t&$5G-F9mZ!h`j zRup!X=OiM=9I^|+-2K$0qcie35N?oWmU(^yb%v-&oFy;+To5|~jDClP<-kGO7kbav ziwT`kXO3s)b;WBzGgahRiVQyL%*UvejNL}ByvF)Y_?3cfy7p4#HX6qh)Ap0}0^Ljm zlWP9)P-f}4wV=Rr9?td@kSVe>c_f8eb+Aw#LmUE=rf7^?`1SVRQ+{Xx<@LbgD}AFK zQrWh>arX%+S|aqB2ar9v_x1*uUj`7l0x0=)G_RA`lk;mZJQBsfYBLs zw5X$`xN5Y)1-ErR|BK7Z%yByt1`f_^wYvY!?R_-?Cw1rk7#F4Akc7s9A-<0pL!)*# zj#p}!iFW?`CwnF(seehb6)cG(HP#3hGtzc>Mqq{cv6FZa#t9rs9V^Fh0aET0$9Fi` z;pR=RRr8%v3#>J*8VxV${5HPm{mK5pF|$dy)wtFaAklk%i|wW=*}|+Ja&i^&i@Uc6 zdB#}s^Ge7?e!}(?K&Yi<;2QMC6sNK#zsWpQt!XOd@CnA~BwC@<6pf=!ELq;ABVGNg z2xN=jd*6DR)A+vy$8l_3BiZ%1=4c~fsnk*KAAc`@B`Uvi_+VK1DyGfe@3r!aKYEJWnc>2ASfMFmYyIpp`r9+l&_T0|o(P>b;bOb|teMl^J zH@m?eNeHU6e6!aq*-q}lqx~cFUgpkP%e#~=HUvVHb528!AhN_|DPD!QU(IS+{@*+8 z8gcAOx8YMC7P;Hee$TuGar!e3a~?16{5bDpugv+z>{yEri_D4+?;#)iZ|^fEDE4*D z8H{2M8vbeFY@N1cZCj!t6C`9^#eLYU#5r@Y1{tLSRavPRz9WDAwNMIN%)kFGK{n7$ zgK32=r9M!WLwRt0{M90DBE%^}O%%ERnjYOH219 zv+MqW>uiY*v$h2j&^+pwTdOGfBXYuO!fPPR?9J~pgk)QK>bxad{c9J%R0LYC#X{Ia@sY}Vt5 zi7#hOn-x*F%o?SHirL}WhIXzLNlX#k+1_?lkVZVN>;bd%r=D8h- zG;T2SB~5V%=b3_T&AoqE$# zrX_^0TD5AE%*|=){{uE*{q6G5p$s>gaARr5&bQb=IcQcm`St<u3`YDx3ZxpK{m8#6Zhc^|in&{IR>VXwgCz z3Wt+NIg0(K4w)GGEt%z1ZpqIE0}N7FJdPCg(_+tN^EhO9+Xo8lh}Q)hwQN|A1nU>q zi-N&)+StqSNWkrhhaW5)diChAiVGr!Ja6?MN2#Q;;o*$N0#c3O(jKvS-HYrScKR!+ z?&Y9u{W)jxBe;)+J|GbbCn-nf+=)*EQ(jDs1axlQD0NWPshX`xXE*9xMqZipWr4hX zo?7G7`3>t;j18Y46yhvnvT&}NNf6T@dq;y_b28B!vU$IqcsNdK(22awd)shz^xkFp zWJb5fYzf#5bzW9YG*_5a9d%}wt#=}&xae^@%IBR;3#)Q}|Ni~Wh}HJ5&Xmc8Xf`+C zF|Tx|yymDEKK_7GSwCI>n!7M=6p0?NYPf5MR45k4Zsuvu3>wv9Sr_XKDqWZ$wv0Ks z^&DrteLl>btR$~Zz>!?<8k3DNU=Tm{XtH$|0=KO{j=Y!(i>)7X4bkbz#*Bbimm|b9 zH|7aT$%)qc?c}|ZK1%F zf>sez6nBWVRegTdgnavRO|K-M@OxBO2-KbtRPI?z*7g7Gx6OiM88P!!5aCwi zS_p!jb2XO*?oPSaU!B-P5l4B|)-og^yo$%rR^QfmlW%`I#Fm`~Z@E~+X3i4F;uRYH z`uV0OTXp6n>?IovsHER5{N_e1Aa1x^7$tOJLA>8R3*m??-sB))n!DuYrhD%1Tcr3K zH;BQ&vXKMwj!$pjAr;5{%BLMX-@vmfR>~j zPbftyt27<9aoP*7z!Ik|>)i55`ubD~sB+8no%>}SD5$1L$slqPy$2xTuV-@x0d==b z3ti!t+scUk&MtSttBlWmOJXdMI1ev-hMm=FYuUg#y;ByLm~$mRH8k9`YN~lxNuK^|rIxtn%ODt(iEd*Sg{M@GK>icryuUXXi9to{<;XI(9{7 z-W03OQl|0t)S$fYe!ArQb9Y;LH_Sg`*``U8CWnL4Bc{0Zr;XM5-3zS|3%lImixWq* zj=hwVG_S?o8gEYhbw3kpf}6tp+jX1kU2n?C zWFyy1-p~87)8Lw5=9@ntB^>mK9aH?L-Z?&PZ|{5oXyfeU6jLMUQM(#%Zd-X-e|wdR zdJlt`9FC8BJ^PHD5KR_zbCzp2ppFEf$;gdcwu>i#l>h?Ay2Rz_|E^ z6pG3or?@qDE#0WfvCY%0Y+RFP;~yw?BYjIYW%vY+SjTH{$25QQtHrTj_uW`w%Uo|^ zCqPbDf7G+)9DP4FMQvxq?$lR_TV^K(0ilhn2^88qt#t zFrBU$gD{CNdB;1quX9i+F8NwY`qo5T?h3jYF0ncJLto`8l>u5&Z%RH>@&)7~W#G{g zl-!fl-x$C{iPo2knRn|(m=h-P>6ENjsk(Nv>lzpH(HM=G)s-F^KFeVhd>C zx8Bou+T-Zt)6E?G`?Pi1J{G~&-7sYV)KgIkbl=d9?sVBDKhwHk$FiJh`&Dc%A*s{N$lu^gWvbKYMXLMUv+sg?Gpy?Kpv zz6&8)nUr)VGCP3in3)H*I>l;KP!?`=(?2jZ++d#~_agiGa@vSm1wxQ0p>QGtrg*h~ z5dguQ-gK4Mcc~fWano>hOiJGV>BomGoA_MM^RI=UkFDyGK?^vMOHtLA{*vhUPmQ@M z#V7MKUv6z3FCPgh9j*nfc%zjo%R?5|#@FT@ZdFtZ1dgh#`F5dRTRpPy;6l^Zvxkx{km=|dcx_f_=L zJ}E|DA$qmfPBRgTxYv*EuMC;-9}dq)c1AYO8ZQ*$dAjjm z2ddlOifX_{Bz?F+a@)ToK&c?2Jv(>Ws*w++vo~Fs$6kTmSL`vf{WkS%TInXF%50Y5 z82)?v9#s}~ys>v!hvWDPNNYuQTi*w9jjW#4rnAp`zF*mp^^;pycc)(re=)So5%Qg9 zzCQfJOj~cWJmf=6O4tas z!D#QzyK+Za(#jGGV}H7$XTq9@3f;76j7YkuiTI;7EP{man#=zrWN}_JjC?m-x1}Nn zlmz7Pr^4~kP%mt3I-7($r6O3qCYC_j$rWstoFl?hXaj83=v_6TLfhJ;U0PXhBkNaV zrG`E*3&HzvB(EN+6!Y1&O^cZmqFK*gfuJHF46XE3i4sYfo?$I#vQmZI#UGKm7lwjPh`%ak+ zDnp-}>uQn%Y{0CHBYFNu8CeW|xYs!#HcJFonJ+1iG0Dymlbb`XNoxJKdk|OCOgXf$ zSqY@jW8IIPG`5s&FxTlf;1e|pX?h(*r71e42onOq$(2*-ZN!u3*Kwp0En0rtQu>A{ zF*Uf+y4=n(GoNSO?wH5M+amIq>>&UX=f3nBG-y!iHnK;d*;PPDaUOeiZnA1JJvB^9 z)*tECUt@C$%omfkO!aym;cA}`pCOIP?<(1IIgf_1l^s4&YIxw}T^IN;D*Jeqtz~8= zJ^}@)Vs1g6ELtnPhW{hGsD+cIiB9o)KQ`MnC2Pa{*b$EesnEqN{xqo?iZcyT?)*4k zPt>Za8M9P}zVtu$KC0y6FR3TmySB#OfKV-MFUQL!@A$*JE8DiwM8#|z-y|&psM|c| zt1^#x!u`LEop!p9H|G6kOU0+YO!W0RR8OE^Y!y&=!LkaY~e_DRFpQn_hB_=`hd&g`D( zlq$Ky-l##$&TKmQ;Ta$E=-uQQ*@UQPh@i>loEncpl-Er{Vm0qt_oSU6V;tj@uRrKr z>$i6}j65MCjN>}n-}kxRqxl%dK6HLCK^xDiSQXLQ`EkED)vZ5<1Z+yVgya86^rg`t zWS>uvSpyn;m&kwwfbaNcwrlm%9+e;faLkT zVSmX_o0kGI)wqOANR`xD*Q-^lgG4UNPo=nI44UcWs?Il>s{4<%foB1facQ`m`nwhG)@)4Q$%U#K84XZ%kqCTrtw8pRi%&e?D zQMR)`Z-i^{o4XsWoYD?SNsn`dQpodPM=_rEOb@Fguq-gfK4*~Z9W+pZglhKrj7i4# zTREQXyN15WQbFL>)B`R2HbNuV#&HrAcDgU>BB|ogypo@$eM)6V zw@y8ka~sNUIw*HZKcUn7SbbfpoaLG+<36Fk1k0^=(@G8C03#ZRz})R|#BN!U@K7a) z6A+~yIW!R!t6J@$eOlG)@i&h3S7%n&XO^wT^7>-m5fA9M-_Q^W5=_9i z>P-i0-h3R8`s-(S1U3EP*bz+_?DdO5BBLwp~w8oP*r!LuIcPxQT#tWUDo z)GLvmC8{`{j6bP#nH%T+n1W+#nphH{L+3SjM|UMzWp8R4oXQKM`#8U?^cmf>T~3#i>9nF@#x%=3vrG~jVNKDy3cK)n3mCP&hZ z8OYKhFcT?rdd`~;3t}fRDq9d3LCb_`f#&dOGhZFu)OFb$yMVaLc%ai=DO~(Nx38Mp zMLR_?6eY}h8{FXat-baEt-6M;Z}#n?E-sNh?hX%mxuHIkWIJl^tk^7N zi&CM11B?yHG~Q{uF{O9hp_wyA?HXlK){bC8H^s!^^tF0<*VRYOc@@ZdT~^UhN!vFF zTCt**Bii7;IQ`Gf&2z52Bl_7#O$gNdgwUa-K9aB+c5vKgwI9!GX(p!*JNldP1}s)~ zk>8KfI{Ch6#0xM_@2!(Hyh%phN(viY+IJ(QLlH9J?$;uZZizmXBIIv3xn2IWHkrhI zNF3Iu8+%}8x*f4QJ|n1`+yBul?Pbf56t~eb`Q*plR`{1PS$G}T%81KM)T);{fAH;z zPl(t>%WsMEpz?@8-z9gFO|WIV6BF9yZm zB+!T+qGYrEJLWcyfo}B-$3Uw2X(!W198`I>TNv)j-cHNkjJi^FM@nqVbX(EQOr>(| z+b33PU1`ZPak0=n+or|7^nyQd&acXj-aYM`4OwIGccZI*&be3Gk)G^VH`tIOHv=D) zRnvk}A)`i(D)K4YX9x&vzXiq0y=@Vjd=K}{Tz5Qgba4>`ID+sTVrt&i-k zeB+!twun&{-}sMP?$joSsG{7q_)qB?_d|n9OX>qs1o^yOXC1c8yj%xq7%v5gF({R# z9cdH{rYWbBOnuPugOj}testzpw#@kt-J7zQ12pUAwaCnL#_3!|KUdo0$E7zPgUO{3 z$e5YEsN)-qThbKqsp@JvMp=2Qg$P58)na?Qu8a-ZU5NVB!8j!>+AVr5Wf3R~_O|S* zwL1gb_{g(dN(~CQTONxblt*8(w*LH<_n;q**>y^ft9qs2<|Pl=}PczmuCN zi`EndC7bD*yC9kJlZCM$Wu0bDGw!YH z26Zt$*K&J?RJ(DEK8u({N(j@*9%2q%&HecFV%A?*rsGM9O*(s_hreAfDi=^J5k+>4 zZn|B`G}u=!cH<4QT@HV)Ld{B9mGdsY`@3@SAPI>GlyN;sk{*9-tPA+d?7zpqb@iuG z6t9$yvFtpb*!e+--*7c`*n+qb0q07dj)M(oG$}^~c=mfap!@jOhI<0+s4bT-Urx?~ z7&HX>$x;%^DRHxeO_2;~B*?rGV9cKPlkbeWyf)8YqQ*K`~7c~~OU0~2^k~Wv! z6GJDWQrTkYooV61B_9N7sKsNuu+kBs)x9$!CPq#jUgesR_yaKzBw!AV#@4ox)b?ZGR3tSrg7EtQTA3%e#Yd`rq zsFmM}%zMY@B>l4U%U_T*&z(3Lw5nOQ-j)=8E*}SebU*VfpC?~9=Wj5SP-2yRMdN;a zuU*dZc^|ol=P`zCsr|cqIXySdrwlJx z5bAR}B`hcpa-NJq!zWZ2-D?)_MP6E6FW%qVpkFn4Pfa?Wtr0ag+uUn-pq;k!lcK!d z^!AZVsbi!ZoQvtMBaCS1{4GlnZ_mKM-AWV4seGJ&;Z)j=Tlh>297+7FJQ_DDMhEiB zO*(Vi6v|}T#8xU>p+bd1mp9#8@_^k0$kDVCRXSBDFL;xlohY+cIkpFfI=5!^`fiMF zbIWYdK%vpyc77tA3(QGL%hZ^jY}F&E4J3kbP{tHBmS<$q2B*rdzH{)ADI=B6VejAe z@fq*u>jCOvQqYgqev|Z(s}Xi9iWv}XKN#h!R<7(5I?Cv7C67)R^6nTQf_yj>YPCu@ z&|8l#`64f}VAz_>S=3LR;rB@V6w$Or0b*WtUCWRPq2s)sAA`CwntdPj)SZKr##1jc zn!jF5GM?kfuy2e1quD0$5X#As5iMDN+lvAwTlg@{%2cP>bqjAh+T(hOm>o=f{P=3= zoUJ6+LU%-4r0~!P8kdCqv*I>a7L`vtA(LF{^git7vsXiHRh z^7wHomO8p+hRN1wzHsr)Cx%rFGTIi0RqqR*>*5mTG20B7y5=R%`&u+kjup;+y1=lS zQaN|~U_bKvvn1Ex){_X9&}hbu7;!dcv0J>fj*Ql#b%3wCCT{K9zhAR`VfS~DDs$fC z;Rk26|7o2*-}%RXcj30ZCrZzxQA{Q2A`&p|3$2rDL1mI?tt}KIEcgc;i~;B*2JT}0 ztyS!W3(KwDre0}gJH=cUH&?3=QFvV-Bm+Ir97l~=sL>(S>u9j%<;whBBu0|C`Ej_7 zZlbCh4WKB*cJ_I{XJIej3yfAI|$10E1l27&EE$WsO1W2OL!zJUx zG#<8&cfwP6yR2(rEnaaJinxGDQyjH6n{x+ zv)(~RcK?$Z;rc?QLhzf$-bp!2vpX1HWMNVGAdQXc^E06E*}0kU)DEV+&e+=_RkC4u z-mx~pu?dyrwPUcghOIg6_jB(E!Jgn}O9Dr94jFLP7$68sCZ;P`XT9`d;GXXqsA6)p2k${0X> zvW|JjGhL#eW_k@66|l+O&bwohpzC%QI+(p*K&0yzNuS?_&vI<}y!1?Z3&bPiffy`X z&I&og^FZF#sBn2!;1-2Mu!NFrT4h~sA|rvCCmI=#xpoh>iJ}N4^;lqu_Lj$k1L8&; zJTPXxV74EdG|{9jMKYq?d{FZcM%1KJGH9HeXTpR-a*_t|G0tqUXWD*xzm$S#Gu^|k zxlOzWv-l)2 z;B8=>E+na;8gWFQEb5DQ`CbAt=?-+c7_fE^f+kgAkQd7zF*d&6FT|%iHuGmwhi6rt zLvo)D4l$aSXtGWSpBFPO>=P4~ai^pj9nJIOemHlEo?E;iwhY2|`cq(GJ|}XF&>nn% zfpr1sMC-(60hmDFl8GXQ;?shr<^=8t1hJ*)m@`J`BYZ^A$47RgVC5Su%CB1a5y@&i zE!+@153_umwc&<+AASA(5F?;lIM+YhG9>MY<++9l&&kO)NFF0Ig(?}nwR|8$Sfrlret_zyQq((Y`|;J+ zv;})8rDj>ubD>&z8f_n2+#xk=AVHG|jVj7Po5{M@IK4=l(x0LB;64dQ%S22_fcbFk zfv3YKmYy(#u|Vc|9uszu@RM_IA-aot&2JKw3YzB|+Fe~PmJAv-a-?WK04Cdb3g6nA z%nP~cWGkE7e@*}f5K&EDDkrlR-z{pNx#69zL{U&ZiP#30ip5+=C}}^bc{r*FL0qtQ zZjv|+8DEeMhZE%2OnRl%2?P3pTpefJ z$cjHl`o2$mJ$o8ns!CMkG3p5+5AWuWzK<$2n;)4|XQK426a?z~aFacY_#9FNLudJ_ zh848*eCDiKyQC81#7Z}iNf$FVZGE4yO!6QiN8AFsKZFpv3DU=RgDMmc>ZXe|P#XFa zn%%_2&kFv{p@32MYs?M3f%cuV0c$}YXLo+pLp#Z;TcOWghCy|C7U6dJ#P>ZQf*Yx= zhDLes+%a(HrS)${O{xWgP)*5e)6Is>LQ2iytU`~cHq$fxnq(Mvd=2-en3d?O3UL(L-*^r3tI4>{t9+4MW?ZypFDn`ip?cWa~ zc2ho8B9+Y#HjWhwyV;*Jk&}~0WENy1t2|}|#HHXM1pC#P_%y}lQOVQJ&vzWS9UGR! z0PJQAxsn@nrwqL(-*0PkqEhQzeD1<@CTZMvN+;49j-oDprPU_JHhQ8d%{y6sA zhj*_e%MqiXt4rLr)|39cm(?2fKHW{JGxjRM7rCju&xz?iRd<}F_kB~nC^CSkdo$4a z^VEI`?(syIzjF4@g=U`F$;lhhI!MsV`l9Xt>la_iQJg@OV!J_`tX}52?xic*NSQPQ z*bM}Y!Ykf*P4|IuK^yD3UL4u_HhHx$4VItU7dc>!v0xB3<@3u9XGWoNw$V*1c{ILs zH10S-fogA`t&UNeY&Yd|%L!xkI#4MHay5-NAPFTp+_;8onw0kk`9sNxMat~hKcezU z{PWnJ-z^O5d8k=Iui;wVNj(w)EoIt^(%{^GVwc`LFJu6ObQASgkw=i~=wMe|I4B;@ zFQJd??+q=rX@T}i^=H1%smEah?|HhBK|fBgY6wP>^z#edmfs|gc1A3;I5~OCRuYF5 zsVv{?8iJ`&cBJ|m3}cQcF_PJPKJd;A;bmJ3)_kQoI#|_^CmtP z4uY}n+$9O2_g=mSY~p~TT!BT-7$8fHB>3azgPnFVk>ba_^~F}v_bArLGa9@b51C3~ z^Xr)bnR)B&XN7B`AI=6tPiL>ye%cZ+ZM`v-ol#2Y)>f~q+`8?JpC|qjKK`m#O28J# z&OMZdwHDh9z^cJB<20#T`*XxhKJ2l#bS{#kJW~dg zN6*YmX6jZSZ}Q(L#6-xVWwTp%LfYiQLT05_@NjLkm3&gRz=*Qx~Z>8vlY@J)AeXZw46Z_S*+!rDhV@ucgq%X%}-|4CP zP1?Ea^Nma(em1RS|GeMz&&ba|h5ZtfPcuV_9w_BI5+txpw2eOg8*zotExKQsH!t%7BSH>Zh zRHAL>_?o&N_JhBuN}Qq^4n^mYB_ztX3nFnf^tU0MkA7Fryg>H}`Qe>5OA5O^_YGye z>50}RCzOV5fNSz8q|~ulHxTL27nAsh2YUNdrt!5aJ-pJB@$MN-Xl}nt z+_^;^tsMmfN-oebkmoHTxlDagy;wr7>JvVGK!XAMI;za|PHMZ;S!5UAYfD~iF+LEJ zhhXyEZ$Tm zH)KYjO}*wHA}8K&nNOBo@>B{OV6{jnnRMR)lg|OpL|t!Qu8c84pAb1>@hXy?4z)MgB1C&Uzg^u=XpBx#5;*VO><$@@Z&bdE!4ih zJY#;<^r2PrCjI>LLiKZ`J5D_2MJpr)HHi*-8En&Tt8hI*ucYpT|MD3xCI-N9yAWSA z9}mMgl9{EidP@gr;p{#9L6&@u<={*YET<_sK1un+$Dl~!&2erm<)ZQp+cbzn@Khn2 z&j?F2s#D)<$z41hmM;0G_-QD!Czsw6OAe>t%G3%z!%&2 zBVnX$2Kr@B0PJ0dsJAq`miel78pjCS?71jBfGeclkXFJpxB9!VV$jPW>D|;6Ka#$) zGtD(<*TuEIaadY7mh)Lk{%2}Vz#!R8nlPHvB+5rE3XVQuVpO(Y|70K@?!$ll;M&ai zR;0c|#rlPdNEqqCA!hOrkAw{t{qBOEmPOq*1c}BaAf3WZ>aXd2^|ttLkRT7N<`#=Z zwXCB&BR5>#3Li4Xiz{DIr~6v`Nv!Fa4d@rvMn$Mb6OSY9ldU08jEy5-2E>_u-25#K zUf)Vq*M-b>UUh8ULD-gCi<`yrjp4`-2>e&p}gcI8X`0m^S?UV&98oeJ=Jx~_Bb z5`EUnpZ?8B9mFTqUS{Cf)8av&jk;PdYA|SN$C}D18{#?vMX}K0JDgn;WJ7=L>enGB zI0EIB!I@FZY({C}JQFPSdehqZM`{cjR&JE#{jAnHnlv2JV1X6SxQZZ9r>pD|JaFoI zI?FiUZ5sZ2a6BOU>Mw>84%XEjKZ_HCh8K??(XG?l!G-47c1ZofP;#)5>rd+O4|UC% zGiPf4apddrI&oM#*P=S4hK+jJx+kev5L0@VBp>~q4p6$-g*t(OS|p=J&S;(}%D+91 zGRv(_F4}1=|2ux_PIw*DWIAm#z9sZ^dlGnyu zQ{;8}yupoYO}1OsyadqO%c;Z4AI`y6)zu*t7omQ5-6$*JO1C4Yq6GZW+ygy1d8FuP z%wn~M>(b3c!T*N6Tz;J8{mE92*Eps&v+xTw^Fr_eY5I+!yZO`X@lfz^ES!tII6m5k z%>@o}L+-?s`-N*vx^8>#^)ekQNEuA8ivy4M*W$z?1Olf|51hyVvQi`k> z3DdMdS$A^s6C@JXfvk4(_NZNp414*oR7iBhuK4YJyHRjnm~LBXN%^{Ff^*?v@4QBt z^-=|w<@^xxph<)Z(HrqRzRD>ex(1;royrv!&_Ov%i?e{p;a+C;AFn^>Pg9N_{`bH{ zKqm4d$4}JS-RWb|iHaV!a{c=HjvN8bd6%~YNcyJR%0pl&Q ze@x6l->Soi22p*6@NiPUSvdIz6~X|cVX>N=(#5)_qbRB$bK_5xPZM%v;ne+qmwnd| zIJowFG)38J<{nnEG7`c<0T)()^Hiv?g>LsYUV=C(p%06P1R>N{$K@teT=Hfop(J#X z-yp0K))&aQ396nnRM9Z#_Sx>3@u|A;5UoaZW*DZOAF?|S4&5;%g;T6-ECU$_CNG3h z%=Zxcke=?o+lkn7Se#t+`)A^~@d54XnIOuGh$oK}wJ$1?EvQTI$PV*Mc(8+CPJRB4 zLU1>R=c$moSBIV+_&oB2+&EeLHUSG_}vlnikJ=b=chM3B;8TlY*6I}c1-6JO? z{6xrVaCpnGinJxwkOs$))uE1uf_gIgI2?oV$>Gy2?O7EDE2vz(`BjhB(dLuZ93Kcn zQ^Gc>?Wk4>;Ffj0Ln9(4l&PlZ>FnM|mNozWY--0108wnx9|U=(5yO{66)e}7Qnk2KK60)vXimjt2fbg>XcWeKCnH|J(dpqOgf10LevPYUGCiJ-^=A+R=UGv zQ@>1_5T&_|84^W_2(%~;vK+<<)aR$)r;KEqZO81kE?GF19-eN}56IICh>P^6xRnEi zyRavy?b&C_^h^~8h5d!&N4Z5Ejl~uZ^1NNF0}*MQ?6doL1$HjD0f!bYx&b)xQu#qA z9YF6H!k>x@Qii!x&66bS+Mdqw`Q+Nxu)+U)BX7(7;vI<;Lv&Fd>d4Zs8`5r8w{{6e zzz|Kp-u=pF2wR?Bh2NvPylaegb`XK>NyAS!p4%ynev`quLYea9Xj0U>$VR6dAHbf| z0HUB9eBJg#uqs+rA39r^F=B)*xC*6?E;d@u7SSt!6)<~#u2G4QfT4)KqS3~l_FzMI<1wClBCne(s&A$4hw`aPYT`jLzlNKoq8 zHQ(d9X1_*50{cWgd`l_`aJ-M4da5gESUwGZL5eK6V?8UjxwXrT`;@r}Z%0S4D6bGX zo*z1g6wODZVx=hu7_NYVo5eeZ-Mig2?BUg!wRK6+EyDK~B#6cp)*S{GTK=CT`U~>E zf{-+8IX_GvI zLty01l(AHCMpz|LsqXl`{Lu9>K4LI~V z%(h?=WUNXR`R7chjaP^EJ}lgxPzLw>7H7jZgRoE8^Zfbv0H|lVG@x%Vm$K%Bxkeg3 z%WUud`O4efYk@F2Vb>6j$3+DIO#dmcBGi=1@5Xro)V3jN>yz03Fz`B z&?7fNdvpK}f^LRQuSx-R`+2lGl4r(dX+B+%fH?SEJql(b{WSpLg;-7`JnB#rOegz$ zH~U0WVqVEdbas9Uy#$4x-yuCW1%hvcws56urK}yPR&S#J$mXz> zc)n3xM$ppj&{*;&O?sO&LZ`@oULW&LS~F(7h^jXz9`()SQSILu{jE>=NuLC+Q%^r5 z@FL`mA2fA|^@XNahFUqcxAY-Lib8%{?rE$!7Q8*(w1<*-snHM>oeutK>CL2t!wNM! zMy_Beay{MD#8=Y^B&}Y&XF>d))~A4HWWJS7+93>q?8}$ZbLagciLU2id7SEHX7cYv zB+H@izZIGa@$ky**k&a4`kH2bb$jkmK1LZ0vzMx$j^R|J)A{0~Ol{A;n# zze+aal7pL~p&(DaAsMzr41B*ip~Rxj6w1l4gZwS()a28ROSQ;Q9fUnXSHgrG2TIKt zOf%_cKdWu%)%ux$7iv4|$Q0bq>(89JgfhdeKkr422SqisztNNDx<^ok@phYG_F^`l zzS>KWf`xCbU{VNt-+cj2n=)F-2f;Xo?UVOj7{pHTgib4-BE7HdK9SzdZ*F)?jqp@;y!HH;LsVdv4L&~PEGs@3) zSI3jC$lo8LtL8sa(vduLsVU~be+e2^WFguHS6 zGY{zd>gOGBtbKYSCo=O4LkO{+u9+|L)ulYQC}AD)zX2ejwfwe)x#a6y=`+{ZxI*8We=fKn zlR(e(a12AzBRIFZhVa~eU7y!5%N65ad7~IJo~@#?i*0)HKCZ$ zlEX(9a}dzB>;y+PwMMUh{_$`3xnE%Zm9i@~-YE+IDWB+T zvX0H-40CGnq^v~HQ7J_##~SFA4nxY{m1EYMLh(5UsB zy_4eQ+{@3GgHyZ)mv*WK+$$}jurdxPXEt=H9c2XcntyyB{r`a~dB%b1WC7Z6?C1KH zAxOGGo4<-51V1&(vvAShL8N~%&CS|)b*mzhMuAd>Uz|Sixz+W{FZ##flw;V_18RYp zZqeyz!VDhW^Sg&8?Ew_qK@20SR;`V!LQB~Ze*ZDBbAbRN_HbC$ zb@+HBsA*y<6)(upjgPEu zo()j&@fLNPN)CJE5|oXH#g&EAsYXO)cCBK)+6}e5Z_vs$vy-OJ4bU*LxFkS+ACu)X zBrrvKHl4kS1k{~}4}UD+&g(n;$}Fsw@X#Q#fBwqCtDAez=6sRQj^XS?%%q+*gJ5fZTQE`TiXeB(Ht(!E3Kg8TxD4qeLS5&8eUBK~dD zRSC*qh}CC(A@*qhiep*lhcaL2Qd7@Q{_IX5$X%X|%3@t6=Daovfu@9NY4oVNy&}^~ zHGLYD@!!~GQRIJvdvas$gkJp^b^*XDTF2F51MfLg04N(*7TmZ*(10%g=OTSiKQ$=0 zyd%v*mPH>Wl6n13wi27C!@?51sOHL5sL((82SSqC1_ei$3HK)+7tgSr?=9eANB6w1Q< ztG$^=Pnt(nklf`&9rG9Qci%@=pGCd}Umg|JGGta@|K|a5Ztqw5$|hjGItaXEM4p^dQa0?3-HVKcv4q&-mYk`)1VnS(!nLf4sH4SV&KF z9b%7RuxWLi-ib+tX2P_Pp#PKhTmNq~$}2gEEa1|vd=hn3;Y7|S3 z4SNe!Kw_avu>{SCf+(PXfFeynQJT_;B}P#+gNmY5L5dUsQ9+6|iX$MsNl}5J521Iy zwTGdd_x=C{&8QraKe|qm?mL$2L1D% z5l~qps0h`)zd+&4L`P+M;JbMpi@2E^_s3UCwjEF&`yeDtu}m3fQl*m;ow`Zj(6)+! zJr~`uz|k4UdtWN1WX90yBabD_%0m4DtsDh+AVhUAH>0lXKmPgXF)qfse;pN3a=+~4 zD+```#Q8-1!hA~cp}k$(ogynANP{@(MFTI6;BLX`Wz)tjN6#N89CPk-OK2gWLyHh3 z1bU1JMKM51nzi9)=rxDpI3brJXr7S9ub*-M6U~Sk`BnM*CZJ6%G1q z4beqy#1{x?8W;*WbhWxDFBI#v5kF53&OZ-Zyl_A5kNLb2k&|CxB3*B%_H0$guU= zy~kp%`}aW;*1Ws*robWaTk>ms$=e&}^uaRVZn@4*S0qSUN^9o))xoq5fIl1V6-u%uvV>iF| z#Jko5X4X)Pt4ZwH`o zJBmr@j?|dH4db@l&wmkn_`gzVkxNFa+D8_|xy@(r;*lZ~HlaKZT__Nxp&<{&R}+)d zhqVEvu2L6IKrnp8?nll}h_kXMs%M_&xLCURH#O~9)sCMYe6+RJk9`ye0XeJVe8c>D zcvMmQsY^4tX7GN!`(Dpv?ENE9{z(9;rp*|86MBcy#m?A6VFIb zKHNL$phyf$5>OB=-AGwMZi7g`N8+5{kSV(j6&Z~^jY7Q8f%(|fWf<(L2n9Sdn8I1@ z`DO<{g{)kXp<)&a2=rPwV(fvo2l_#B?je5=GHDn@f5c9nHh>he#MUev@d&~NdP+@< zd3&v(%%cK{Xiq{*h;t}gdBzVy4b(t5HQN-(t&>17NHT-yQe#8r{du^t%K&{lnC2Tg z$Qc?`XSSmGuTVQ|&lZ3-C<#$U-zAc2RD}x}tIq~APego4_)MP}cT&juObm5I_50Mw23b3K`ohn;Bas@Yy%yV=FtsCV@E?Mv5R_djMv5)VD-*2s+cqP;$XRu8j(` zV~&_TL4CaD8`q$F0LDuPy=e|qI5crD&p4glmQ|YIjD{0Z#E+P=edrP&eXle#nKb>> zEo?h3`tDghG5UsRlhD$&zk6ksFL-Mddls1JNE3wwnq+EIu#Rb*6g@kfk>?20LgL8` zMq_XU((NtmR?sI7=js4?mpnRrj^wW7fk&>P@I*YzMyKd4Kyho)l%k1VGJ$c}M8KQX zJd?5E_qH>Hyb&nvM{E{?Lm!F?nwAXHZ3bgl z7_Yc`+ASIaOMmVN4buXW4uJqd65rDHP8UO>&Iqa?)jijf4xQ$#IO6uvAE*L3sOYCN zQyVOKE<)UI!IylT@Fm&MS0D=^tpcQ)Dy#oqIS)~V$Y#(4syK}3%sWo{_lF0u*gt?PlHM)2RIuQpU}fZlYquVczgKSB)#L1;bBnu*m=j_fVH^o+3~x zwgagFh{Xll3x`o|_e(q6Q>c>^+117QC&sEH_nf~boKqryqoLvzn}oM`D3C!y;Y_iV zdV*gjemVK<_pbFBp9SZkHlcc`ndx3~PrBB*?*|MIl z@aVp}XoJwRXc5TjAeE;M9q#f0=?u^{0UpzLFUdqB8FID73F& zBYN~DjC?m54S$=cJKA{_Zt6JFxN^PhghB(dzWlI}@99P(G?S1W1XopsTLv2R2?G&y zKU$3=@dYOWq3khNK?XYm;9b8B)BBYrUUs{_|E2m_H&s5O^5O8Lx}X`q7SB zAdv!}MH0gah;jiH>a(Ds^;o7eUW=4?j5{HxnRZ*U#;uu2p5Mdq=K9kcOZqz%)KQS2 zr%eNyC7gY_5;MC7e)uyLZi~lDiJ+&su!uTh^QKRJcOoJ?`CBo*C7aav#CcQx31osmqHRDl;%7jl`*7Fd^U% z1ep5}2LPe&)J+-&Y!pTGKdat(UdYJJ$szLvYVY7eXTUFi%-n>wQ6MII*qktlMnQ<1 z$_9W(oxv_(n3y!A$pr0P@Y_W>{hptul=6PLf19{RPih()8`si6G?IxRJDL0rbr+m< zQ*B@}8iQ-8gXE1myJ_$Rqt9OBnjEkgn(4Eu(C%Ct4A=1*xMTvj-g34B(BgJT*+Cqu z0bPDOFfNI5pEGRaKXNUyL6nfvWCBt^k*&jts#y0kyBg0#S~%>Y+#1dxyc(zn#q3#eqLK-QR)#QN*o~UyGV?+EL?e9#c|{J3A_Epf8t^DM@a~2 zxRGeXp9i-P{Q1Qh@p)@S4Mn(ZcZ+_^Zh*`eg`zncw_YpT_Myq$wLr(L4t3s7 zNdf!eM*&$|;Z>RPFp|LI0u&Ct8wXj-8cbaBJANZM63wa6}9y z8LcA#P5IUhleFrw5~%XGiP_`4=g`MMwp|CdgNXC9E1z=usHJ&1uyRVT5C*lND5rU$ z)X+pu9O{FZ&1Upo){%Bmr?5qO$;X=E?4TF z#5gOI*YbyoHd%H33sW!)KyH*12P;F~+lIZMM6<$j{I_(b%tx0D9J6sGo*iE+s7-|9 zHpi0Ef^t@D8H}cb>nGk`3iwiJ@)V5UZ{n~K3<^lg3A=U1hnb-0#2Bzq)`rno`!%=z zZvk1`#^adNJ&X5_-O{8zk>mhLLitoYB;st5q0^}z-LQQ9XIPvV_0{>KBcQ(YX$-wu zM&e;>DgaWcxepEoQP36zeB++`m<26Iwt3XJz*VkLyCh0+_{JgRhG5R$K{O@!7K~q1 z6XV>xyVUJRqE-P`oB)HGlE~(UQxv%H&;RrT&KGXLBP;+tELxZ4S;KscbOI#Ua%+XH z!*;gZ8{6~igdMv%mZ7Rj2aCAaPbWizgTv6Ep|bcXpub+0;b4$h^s}vd2Y>FRes)m( z%dBUN|E!bd|MgiqFyljN{|f_lh&lfHSssgPPQc-VW(Pr_G4;zyZbL7n5dlu)p)vP! zBOb=X5C8I2=rjjs_^Y~2`LsK}g*DIU8@bAmPyk2*kY%<5ibRyBdreJE|7boXyQ-ZU zXi-qvm2FiXV+tqT5c;nx3_lPXMEMui3xhkFFUW4Q%#2@t`U5ru17v(M%pP2PO;g@5 z3j3e$ubsxHgVvq8%5iLE$d*qGKW~MwD41tw!5;!n(4=QlsdryHP|ToS)%{R5!lee8 z7+yIv5vgd^dcBx)Xq9bN8|@}&cR)Ty{GK)La1EI@F(I?`k+F*QQj-_`$pgMZM1!e%xmwCJ>-PEMo_fk06@1g_JJ+nj;+96VN0BVD;I`QuNFI4_QVkIdmBVcfwWC=`~ zD>S9c7F`~ipf3+KzWVC|w=yJOZC=+p_sv$mc|AjP>>W=M(MDfl>~aE&31M`1LLyD0 zw8_zeae*%oF9Pfv4clO1%MT}xS_mwwm7gCk?Qsk#p*gzpR=e%jPIJxs%Bf278D>(&*m(>St2BUZr4tn};PMj@Zu zZv>va3DcPCv+({)gZQh{7c?$@Ds^DOk`iXe?Km?S$N8O2r%ysgGwp>Di4zc`xO>2;HMF2s{6 zJ-(qIfQ7pGI?fR#Y`ZLstmFSP_i5>g=aIN~xvJ~i;?QVjdTyrlfoCUR%G+aWSsr0;5i|e#iQ~PfK<&awau4f3~G?E z^WrCqW5=wKaW03G?9^nX^e zg46>;CND2_?f!ChBty>leKY{3zCt6+-$)0l(c_bWIJXZEt+llkDDr{J4SdMppRSagXNH+3*@5#x{c$Lw#;{em%VP1BdUGNTFY(c>TWCK_=y0Q`j*SF0|M~%X- zWbmt;%a)ZXdqj+3iJXXIhGy!lZ8Y2R-uY!JV?nh3qg0fCIu=n0*671Ey|~Q&K&mp= zSq`*n%ul07PWNShvSJ$l}jV%|a>Dn;LEBO!m|igB0K`*ODHbyUd;^XJdUS|Br;4d$N$4q_lx z9h&m6YaZ_3zdtlC&iYFUd`nhA1ilSU;|~Phu4?~bmh4LUJu`1wpd?&bkCL@dTVdLu z6tun{fOFqM!JRd z`UtD&i%YG4Cl6BF+E0gL1OLq`tU%a1=umxAj+5NRoOkilHTn9m8Q{0G4^U{X_*b6y~rAkVQUlKO?Gm1jT+&F zU$nBlba|iLJ?CL`v>NJUrZGnE@HL=_*u|trgSeXkZ=8GB%TX~{+5&Q**tk)TQTB5$ zN(z7UsD_BDkKB6oxQ!dEv_c8musuLAn9x8=Fmj-D6!2GDUIEcd?(@?e>+M-FW~d$WrvoGw38`NspL5fI_l4`WMe4mdBqA1bFVJ-Hdl%oFf+)tK+{LEW|0 z%M!moO70pLwqj2*jlblz{zSGN zI8-cRHJVnqo3KkO2JPut3NWw|C|+r6Klyc%0iqU04#O^D31ja(HLJfS&&{G9)TSq; zYZ-eHUhC`Y!EGl4Am-!n5T!~F##l(M0x#!{6hb*YbX7Yx(S?2Q`7Ug&$@2FHx5d5# zPjTFVkF7Q)VB-bEqu*Y*$99dQLr>_C!r`xPF70+&J@Png#_i2dX>LQzRqa@2xVl~U zWY?i>Uk&#ef3FZHmw^A2PZo6UU@aTOH}aD27ja{yh~aL8;x-MdxpolpJdJpRIrszb8M&6{;K`0Ug=I zycEajada*X%>XO z%;4r9+d_1H9q2~`RTfCM7;Gtx7#y^e1)%-lksNoDEZrV!45D{dq0bT%(_6STt%|8? z{c4)9F=8qo$MEN0B}4Xzh~Kk!g-bsh2*_?Gv9^r3*RwVSpfDJ<*JP_i=eq;{5l@82%XnjqD zfxDLg#TJ~{@i{xis)4$t!wK+NzyAo~Gjoi=5?^LiaArSj7IDp#BTbpL^x3bEc{}X|zxU`nx*jZh-at?9E?(`0!u>iX zHyc5^)D^}UCY|Q^0a1cGOE`T1H7+NE_hyiHuL(>rC{b7oyQ`(A-u%bzK`?J;w7eA@ zO@qZBoB!nUa$Y;${)M5G&A4UqBEU+-AIJ}tVijLmEVTyOBP;9+@4hzPgAF(}TK2lj zFPIs)UrS5k-l5c$e^G*nm>^$>?WJw#(qZ}N<&;ERQvBNmdXKAeM}GW2V?VxHl`jy} z0Q?+4ac)3v_|bQh%w5YkhIEPf{Ta&@WZ{#2R9o+O0%au0$;pDlF0uHanVQ8pqXn&@ z0@N2-=w$Fm(G?sKaqRF!*Spl7edu2wby+dz>#-3kx5O@D0014P_tCfSb%}pbL`kL-AHxbL-J7@ zjb8`Wtg;>cQ}>uDqen}C!TFJ^PVFCippJSQ&-?v00y1oqz@S1Uk^j~jZBgXb`#c#rPW=FvdA*~>-g zJmv`{QyPwfA`ywA7frzwEaiPpfZvDdN?%6UET43wY8xQHS&AO5MVRNs41V%Ny1F{` z7oJAqkjj{dYvV(>Q_G<}yU=S?9UVD_JGq-|9Z}~TdOH@mc{jb8`KIv7$i49X?E2UP z=lzL$)@Xjk`;$rk-~N$%%ynQHArQl7`uoBC!B%f(0Zy?X_BzlRw@&b&cX!D@KHNGn za#%4OdEO6 z;=f>=x>_a@#7=)0Ex#^}D)Dd8xSTHDA;FP8A)|-n7Tyh!qLgMN0f470>Zw-!! z4=D0)<3x@zXaJ*e6JV1x#Upg@-+<^_{F@qY!5r8;vR?u&C;%37fbNdUW@(ly^6sHI zui1~1ossS&R-WavM%hA9bEbhZ=ETs+O^G`!J+C8#C~K((;eBn3Rvfs$2gp$E zj)B3kYE1>r{jlxXj%&;5KJF$%Bkpa|8m>{eiq z0TWM%QLx_7e6%Qq2o0ZjY%QdPSDkR zI1rqjhyLArNN~1urgdNpPbDoIl@92l*(i)NKuF~vq1)b61{a!sG~H%FXx7X(pLZ{} z=8kRr3w7Yt!DI;}ZY3x{*Upo+xiiYsN(ak(B|G8})6OT}m9%KAH z$}`zWK+Uk0^Thjj>-f=`(ZI(lMwcXWrgTw&+HqKjT`hTtIzc4492>SBBK+)uEiSv@ zFGg)Z>0~dSedHh=!?O^uJ7*_$eKgv@TCYMQFtz2$x(O?*P3yT27h9_53SyORtJ zBbP$48;ULCGvZ)yCUxYU#5a^^>EK%Lq9;Xhg>%am`dUht=xZ^_&6h&S&6})d|0{G( zAqQbmeMm(*iVrIHu~Yws9&SDQV-K2|NoGPh+zP2r71lu}u(7=;oY1cjg4J7$Ho9EA z`|9#sti-#d<|AGU(|ahH0uyNrU@U`n*k|5*cAq?YWb3{_HhBkEx(T2fp9O7JQ&y#bs{fhx@th?sck$1ioAh)WkDY|#4aCE*9S_l$zquo!DY2_+4~ z$B#5YwABx*+wy6deb2R_?rTe}->(vtl(c%%#;=8|!iucI(@#gSPXOrJ&T>i{!I%;z zF-xK4g-JnfoHV8oAHnQS2h5)aCeykZ2u5ns{KA z)xa7uu{iPr+B|_nTv;k-eW?p|FVu(da$twu_l)Q&1_+S>NXZ7fy9VkPR81X5oF+6g zr*=eHnSjXaDR`Qy#o2Jnw6!XoLiT#=s`t(Sb6Q1PQN7Vcl*dSMG%~ibq+{& z#G`xlYBL)yO_@*%-@>gKllrDw#2VY@4mP5hf^lgEJ+Jfb11xXT%lgT5#V35aYcuEe@@9n)MdZi9)pV z(GnHo7-Jy{#UvK1Iv$@5kO~v52Gtq(7k`KN?IdBDvVh3;B1(L-k$4s_^_>?tD<(*2QX zgS~Rl%6j)c|GaRhO?k)*Qb{(}nt^3^f7>hs3!s|eYE8gS-@}{tno!sJ)6JhzzC?t% z0y&g%;1$r?&ZF^FqKX^0&*0}0aW2)TxB^I2*^;Z$WjPDHJPU=x9W3Z5jRu&L-z0D! zD`-D%ExT4pO?jYwDJBF`06JHo2#~n=@Wo}V*dbEAjVd(IP59o)t(|z)$-qgkmVCqL z6jjt?HaN?q^$b}ojNQ~Ax1+kH^#VgA;J<%}prf_9U zp%_x=#Zrnc6p0BX(6=uR(MfZz%yJvF>4X7aGS+&BK`xxusa90!co>}&kv@R%B?P_k zDc=%A^1tzpLeqIYSC78V<@sz*GGB$^ZEJiMC^~nyIK~*Zm*cgVq3PE0R{4U3GA9la z#7lXXV9b7PZQ?L~$5V($$Tx+2(6G?}x@>jiSGVCo7W~qsztK#YHJSh5qoM{ZNbVbe zbUYM&N`#d8Hh_K01UjNM@0sNlOr$VDB+9~gjKm@*I(RaqOU0izCs`1@g`RbJkFth< zbFJ5lO#qx)TYN(=j~iHa-5#1dMc(hDz0haOt3FZ@^lb!<8nF z&W?`eBM@T9;YF(&lp-M5z{U-}M1!J|Lus5t{>UuoP`d|uhFA(FJ#z}M4KB@J?|~Ld z1if>)p1>EllUQRVnE__piE`Pu0nNC_xIr7Iq&_U0z)%4vnX4wd3hzoE9fM_6gSQfh zWvvQsz>`Dt$7%^w>1^dw~Ep}oFRU0nD}QG4HQ)Z76&Jx zy!awG#TJU>M)`e}sxGVEAbZk{Nkv{U@YVpDGr6yS{n84bg!%dnc)QTr2Gdbn4AdC- zr8c{lpnHTfaY}aQ{^bV@h{BZuD=44Y`yXCDVYcF5nJh0(Py9p>&hStN)liTaR)&mvNpvIt%QrDEf1hA;4AR( zwAvCRp<1n7DXEYs-Nbb}e#uHoPF6qxk%^!Yk8y|j4i!6+r6DEeUtItap)KLr_NL)<`-Qi<&f#qJk@VGk0hM|hCmm5m zMa&g&25=CMZJC$I=1=QZ$y8|jDA??R)1sQ{y;tvmwK?FtYS^_B4*6zq_Q{ZMaP7_Y zuR$pozJG{l-xG+hv zSDn?>)rpvkxE>6NV6^`RGc`5YFXsi-7IN?rDWQ7T5ZEw~>X3#DrDf0&T}Lgv*6teJ zVbyO(eBu6g_g~|U7WPMgx#3<+`>Hk1*2yKl!4@V_w;o9 zixH6uP@+4ch5iQZ^O?GOSc+hEe}y7vdp5twLAQYpx!Qnp^(loxUpN9PPSXD%&Rs`4 zNJ^N}$;ua~6O*b0npCc?P*$SlNxY+c2ZF9J^0(5xg;{qt2!n2R$gXMyU_z&9NAK1v z%5Y{S_!4jxzmvL)o54_r;n2;Bf#%aH9hBdq4#?e`miG-kZnS>)wfo%>KwvuRinCWd z;UZxHNRd2JQc=RQ@mk+gj+&RE#353YR;hdf>W`H$>lW!*tmm^&O->No2`2(Pg=`S) zWX?zq2b0u$AG}?8@WCY3PLO+-zmJ20@xWIfa+kzjaq%sI)+$fpWkG$4>IU!( zOf#zN91z9o=i{GJQh>7&qmu?v2A{C@XO)oGZd@u*YYDd60?s?-!BZ&6vMVX4@m|7( zcVQ=TIr`>wSxGBo?czO?=TBW`y9abjwBv;=uc{>#-P@_a#0-olYB(sZvgoDdbx@S> zc!eA+ExuzK@}70?mU@wAoI%&xbWr(zknukbO9g3x{oAZVDIMyw_|M?71w+&P(b#=Q z%0k=Au@UxdoQBVA}O>==!p1;lXXr4x`x0<3}p5 zeJBw6iq?7bBKyAu7GdqFYY*1Uuc417KtJe3J+l$TT~D35o35rCmDHFTEMeyoezG2S z=119aM{^+J0;Rzz=Mp_^a5(etwlPF}vVoaPFk<1%5)1rbCJt?V&7svl-t(?O*1A^5 zaO0i|=HdK(77Ogq$~Awt2Ka<5D#c-5TK*09wrJ^nR5R89LO-1ll{nMHV((O}Ar_NlgxiPR00W}jk$&#>vTiz<$y7_GD za!lk-$64M)ZTr_Ro+1&-*zgjcGD4NmF97m3GRzg3^2S<^RZ5)mbJawAdR<_)Vhs74 zhkhG!s%k1!B8F;Y(vE7*zN+snfW08vzU;oq&^=2AnesNEa}T>x!3vJb|eA65S95!N6Ra#BGIO0)fXc%u+v^57Y`RB3CjKRsj4*7NwOY!OggHY{ICH}h1T~%Q9lt#fJD8r za9vQj)2`dUu0mPxaGw~8=$1U5TyGrqo;t}BQCt2RRSSgol*u4gftVDGJ|3}mt{>R% zyHbS*Eufo<-h!u#6YR4;xiK4&h+XxQPuU#K zU_q&4mE*_K1;&!@h`jBVjWDY`$m^w-dfJacAvYYCxpqCTINq8%(`DJM-m=f|YQ3Ac z?MR`wV;cGq%$`6dvAv5<%}{{09r6#AC?Vb6)#`xOQX`EffV*pZht9thl^W__XR;4f;Hp0o7-vBXEjinZaHAg7O-2`vk>fJdtyO`unq?l&> z?;n{cQ>YOp%;4B+yKNioUBW9l1XTqaFxAh+dGuDmGE0c_TwG$UtuiQ}>d%bV3dj}{ ziTC@s?wv6gOLMG<%>MKP{VwpptePyf+VA!Igi3RiRZU_&jOg%V; zl})1LnTn#GCM&n+k2+xkQ9~sr!2tq^lrbZCKF>_5@pJVPKu&3@5f%O=XR_00U?O(M zFjY=F-HARNy4w3;25aV({Crh^2vymb>IGVwQ&o64yyCTB$}C}Zzk`~kU;(*r zUXs=IP{l3U!zoRd33&;Z*0)yX^5tm9uR(nju`dlaP3ysw2u7!nUIsC$AUq11vtk@l z;~vK-RF%4tyTZRC&ig;VY_Bp??$co!1NydWwnD$;ESR^NHU&O;^2E5s2~?YTYdPvE zQ;p-u)C4&aTJ&#Fr+p4yt_@G;?NMU$|55yQWrI(i>bA_@)uK}AN=J*+D?=3UycQwZ zC8pH0>@-4NMv|cmitQbO4H)lRwO-EZRLpBiE1;H>j6jK%z+g*ZS#@2m9&k+%Kp*Q- zYf1mpSUXjeN48bZCm|H&%aO^gA3XF=gOHt{PdX2zQl?F_x$E3r@?z^K>EW*KGSFa{ zg~Q%3R2nA@A%sv8kzKMby99lEIncA1HXQ~83Ft|sQ1{Yqlzt ztHk_AB5~A~* z3R+#$bj#IHz*4qscZ{+TLa~`$-af9BBuApuWd-y4ipL>2hAXL1Q$yF4eBcK{8Og8R zrh-sjE0+en4A;T0LQ}Ku$;GETd6yhHweOm0{@wc*L7Hzov)d%(T7hy5!k1tJmQ zbMRxz@2V_JAPtbi-Nl1%EOd&YT3t@hf+_i-;Spw+2{r)aA})DJ3g@fPLrKz9c|3S6 z$?1AnjS-EmB)Vw>93&Bmew9@u%0ar@@Ocv*Crjnx%&%cNx#~O}mVv&Q9+@06(w|m2 zgxmm?Rx1@IsH>l-9S7?vDrq%%&3Thuz2^GblG~M?K|3e!f+oqJwS4cYLP&X~9OTa& z*&6M^S4}oiSd^@)EPjGIZUKKok0IiG&|Bf(+;DWOcm;Y@r499^*`WND5kdBij-TVv z1wNM8R#Y9_NuDCS+efoC84Dp4@&uKR*-G63j4Tc`uau!R6+zH4zPsHGDB3hi1aeg zQeBT4*5H1pg{6q2&hUoQ`)c6a300MN{gSTYWma3;N__E_cEL$f(;~A^|leVqp1Tb2|!;TXCE$w z;PXMagw$P73MNRj*G4>j({K_JF|i0mlta5q;$s`U-}dEy?L6=qPrepP6biTUlt!39 z@W{7LF=7uGNxC#+O1op5ZY=Bus zs5sX`HeNRe@#TQ#bDVb#{ig%>gxR>D`5$8k?QV@AxRFI;=aD;)!V7YI=ef)VH@`MF zKJDMDR`c=+J{Q`(Vg*#;8pjE+gIuR5KnX(9G%5K5y{M+x2%yLFN=Vrp)(L#e4CLw6 z5)P{pXNIT_rG`|4o7guqJ84$8Mie&jRe3g|=&@G*<%n$$!nC!t!0_!%nZ-e|2%S_+ zHp+UG#vbMVnD*C1NdP_(&9@(c=h1DK3&UIA2T|;yMxS6h(mPW1$R{A}W*T13W|K&GRS!h6en18iGySo~g)!UpzymW5 z#C>%TmTC+jX*iMRO8gfnWqnQMO7|abJ!^({jQI2xEVYEp^A&%}0*YFau$EyD3W1nK zHB2CN=z)Z(tX6TqZAE{LOgy6B{Cx1}153eMU-kPlMSzm%^bP#AfV?!CfO8+u(;+pB z#!rcXkc4Y$Vz`9;{jMbtRB8ka;vuu(ikn}nxa2;rxfl<|qn_k7=U%tqC}KejmOdm=Yoh#G6Htu&4Px85>a{c+JQ{3p~F0z-#9X*vxEo+@BJ3TiE8AF zv4s)9C6W-q?FLCM#p?eIwE!uHd6W*X|L^&pbB zt3Nc<4BYb`CE9I7uNo|G!$ucwDipPDUm_gszL4g7roq#&y~=k1!bmx$5$%i;T`XdV zfK_OaKi9e}`G$Avs&Sk^JuFAWn3PVKO~^Gr`__smE0uH&3#Y$wsK~8>^%%wJU{+Q?({k#UF852W&Y6S0MeV&R;zI^u&H03DuJ>J45BE zj){KtdHBv*^0fg|CYzJ;(+p}`6>2qAghnfUZ=pZTB1I)*`m#G=WDyUXg9bu%)PF&Y z&HxId8{_0%1!3sN?MsephDzh;YfKv&Bsqdl@(>ZF=>G~z?-N%iEdl0zmoNhhfY62F z_#~3o$Mzq1p>TOPyh6QW0WRM9_M#CqOb!AIvwD|ul-;+%X@%(sJ)>!2DhU9sA2XZ= zL7QdbUE(|Ob{hepX3W{@21sKA>?=dcp8^UqMpRMhfQd#m+A~;y)RW(ItiaR;im9ZT zU?|OG$`k-Ry@@(vK|WQKQwGZDnCrnTBm+$jtXioQ-QM+7JOK{*_+ne%5;R$o5H>Gd zOH;E7f2;4u(=$E#ev~i?;MNXMx`c)eA~F~QiQzV~yy>9Sj)cL?eXn5_bQ^0R*}w&r zi!dQ89`zWxxr-#hLD#E;NK!!T1F}m76PQ_H!{@yu$i*6aAaWbih4b1Q3MG*|*dyMQ z1Sn!~o;RY77)kP}kA@J!tIaVXNE?^;5X`oIb3d-vgj75LjoQcu!6}P&pp{#n9%&|h z?Mp!^G_XOAlyVx+>@S!e*>2kav*s286bhx3aK zG|=EH5oYu-i8lu`zmKg89swP<1;Ot9arV%diYKH2wb_6Kw$r+P`0#X}udf3;$j7#G zhxf@O!|fw#Ciw(0Uuh8o&4NGGhJ4AHaALVoG51pq77$shF!W%Q~B==9Q?g}x)Sb*-f& znTstTtG|QFn)-LJ-(~4T2fN&cWN6vbjpKl~`qGOHQA(6deaBJ~%*EnRm5KgpDDFor6>!*%)aj@(l|sswI%n}8Im8fVW3 zjJoxCOj7i&+)9PT+uSIIJ~a4 zB#d%JqD}=+#uF>THt|ofPY;LZhM^9-K`#TtwanWcfZAlt-+wT>wgZmoMUMp4)>~aB zw-_`V?I4+yY1z}jnAgatB{6Ekma;6z7$|YFpe(r#FjSTT0^yOi>KJ;qIBy4nYb#t= z6f645yB0q_p@Ex_gR|`7c}iACDMmt@5ydPdsC9mg!|*d`c{{3UeK98rC>Zatgc|MTc4Dh%K%NTa*`nkaYdmqa9kTY^UR-PC{DQ2E>)h6CG@W zh6EydS*Wgkv6-K4VMl>1MwGH3J*KqYq0_`Kke-aTSVmk$5S5aIx+446hSI1J}<4Gyl z~B> z(3S5@orkCas3i|}b03fbdlBh}RNm|Z%X{E-JQ_sUZi8$$Yj6=JAkrQ1Qum>+s3UAAUoj`}u%zEgKoBwhef zrd&M+YE2duNmcRwpxueNPekV3iWfLvpM!Ww>z4923J$r*iY^pVw;%>U79^%Y&n1Ti zu_T{GB00Pm3CdQe26Kk^hD0(r@bqjs#3qBGkTM^F$xam&M$~VHa4t@uELcp{^5<9( z5Ad>!5j04}O=)LHDJW>eGTC+dvwlCb%&iWBg!j}lj;!JZ<64#Ng67RbpyyDcm zWyQ#6bi>4k*RQtbr~qox2)er5d460HBy6 z-ja4OFq*dsGe06a~6S%$!Y2YzFyJ2M{=wf)##$NJukplJ7kMsMq#sH|WIXvULfb7vCagzBBY6`*Lfi9DWvzu}ad z5@o~B2mdb?4Wx^XFnW-L!=VzyD^}9MC+7k^CdD>zw+s?(LWfp^BA10ELd!Y^>%$aE zH1HgHnOR>wny8vWWr4|JAO@%%mY{bW-y%s}fGyNf*Lk1z82(=8OE)G2r%5j^v#EU> zA5kXmL<3}Cs&|B>Os2&gRxmGT^>(a)2!^Yseu$N6T|1roY3`C@qm40>%r&4RGa$8v zB`Ph6vJcZAqCT$Q}ZmXN{t3oIc3yF>wpQ_=d@+Y>umFYa~e_T8nd^(^=Sr1o263kGB&DAJD%!_nHjWo#!fv9a2;aIB11u!@j z%D6v${FPqPfMhbH9n`mQu=^RfqdJrcgffs~gmkV+fLhP;I#XIqMhVZ6m$;c3Jci4T z{9&cXL^mG%x%+?lFyoKLB=#de_kaKB|DMzTzEo<4eo%6*{PuWTTsJKihKky6(OVCl F{eK@}F`EDY literal 265522 zcmeFZXH=Ev7BxzaQDXs;V529o0tza<+m)_V=|K>rBfVo{gs2FLB3(qJH|ZTTs7UX< ziAXP-zUh4PNi^|{dvgEZ@mmod^lYunEN#tPt&MmdSzorsRkm1NP`9C>Vf%yp?<$XY_DvVjuJS)x2b_02@hOstv`^$ljWcDz+=3 z$z&&P;D7wQ|7Eau;y?f659B4`4YWi5@+T?2ryKtFpVrQ85c%JKIr06!UG7G4-3_|` z{h|A3{_S$_E&uf%_WkSS7MTC-9%%mmNBVOH{x6;-srt#@lB=ng&D=fd*(WCI@6&p_ zQ+M5c({=T$94|ie_Qpw(H0yq8hXBv*ldLp3YL1RMy~2*OLzf%^3`L5#J}=-L!PLZ%zM zSQWo}ci|qpq(sFCP38OQVoTn;Hd{V@^?Qpix1MlgUF3OMW`#dWN}7s2*3D8j)B7yW z3~~tws4hE&U z!_Qs*{A@QiZ~J1So}jbCD3>%bhVU7m+8~nFI5*nD@3v$sK2yo7?zeu@(x5X{->Wy5 zm61^j;UY%%iJ1D{^FEJaVq%7KYo3TFM9BmMKiU2H;xxhmUisVp{*8v_qZq>h*>`>g zf)<^%8O9B1`vhskVyi=iBYWJJkH}an1RXzHR#v9&$Ncwagc-iLJsKKjWF)aYV^#cSzrR2D!{D=jW-w=;h=_)20)K^jyjse{M8G#< z-;RadbbN93Y=fRE|AjwRRg<-h9a)XrGve}PHn4p&Gf>sgUlAZzm1{SuI^JE_(HMFo zpO#&sv%gp=^4YVo0d_%)(2B)e-Omv{=fA^UQ|zUZ0|HLd22s(`!2;)HWxXFCR=>iK zqMjaYY(w+%=F6dAxAAuV)s>l==xY7hAN;g(t>dKqIissTFF=|bp|>{K{{1vGHF276m-X6N9v)OswYIj79@s?F+7f$duRo_s`2mZ@qZ5fCf|e%j8O8!rrED&; zKmPb*bk!Hv^!l%B=Gl*PN=izG_6AdSsC1eJEDD`r%z6IqLLu$I2f{dbnqrcQij*`TTjBce&S2Z2tQ8LTm# z`tU|zs=Lso0zXlmG;D`=b8FQ;*n8e~^8Lv$aTnFNxVQ=K&ztQ2mKDiUe}AEc?b?b{ zl1`z9%luf~+UoM$V7SM~hc|cc`1-1yI`wNoVPXH^V9cpoTjOkp>*Fz@!ThGt7^;ef zhK8>Go&bx^oCvuP0pp6tN7{=$isvakDM~T&91;>bQ&UrE8cF335BmP`#~;cCj@MNZ zROPV2)Dl#acI?>kvA#1#K1|WfY%FYiXy{tW>h}x`h$L)Y)&V@aI(TO8CL_;<1;-;g z1(!{m6YG+-^LMbaCgtP^X%{-HNJvO*{pOpCh?JbFiLqYn?w8FwvQ%FEy0J@R2Agtr zG^N-CL83j=v?Z+Oba#Q1KSqA%;ma3knFW;d?Z+n;ZPKM49@t7x--001@cN)^;M15V z&gH`ef|lKl2qESlFo#OluEjPuh57sWJ;Q|%&{W#fuHL0%mq<)Es^L6#>=Fi+EP$;$ zIiIIpt%!GDT{7v)%Osm_*0rv#PFhyBAl+@Gk)vew!-o$n64?pLaUX28&k0y|C+?Sd zd<}!PG*ctFlYzlza5?tpSHH$9#rQ-kgjTEGVd->Kkvg*7pD~S z=>GkuhzW-*y9>zku{G_X^ySTK(saxdAdFk12(vJ>BAd)8N6grcsb#-->jVoa#>%S^YRBg05sR17* z3uOPkeeX(3&+F;w#b1p0HJH!%S#_9b0@BC?H?^p_*ux`9Ba>5Z49oS-ojdh=eZ@Z5 zc9%UoI66=j;yH6Qo2=2gXJ<9i4e!m5cf>@V_vOR_QA^QLc_QY7weZ|;zf3&h0k-6W z_AK+!Y4M=FKCGhAZRti_oSf2liUuSx+qn^y{;CkQh=_>z%c;+P`st^}h|`AI`4M5OD3o!L?4g~j>Q_8rcxao+H$;Dgssms z#3;02sbgxn-;|cdFxFtLvN;d^m}%T_?!kix7@;#+&Dq8c2hN^7YmAg8VmowSK1?L) z_3JZT1x`v%GXpBW{AZ(bv@GMk*$!h-Q)BY0v^>2Lq!v;#hjE{<6 zuY#>@M#qRrM^=)aoZIqzmER(kwju4-t;a)cI&-cEl|Kj$Hqp+u2%EdLTjzypl17tB z>gBPb?BUfNhc3QcqIj%#I8S{z+n^I(*fS$aDd@ZS1TiL{JJS2i`SZ*JJeb!8VTC7R z#;arLHHk+wGF~h#Es4yDPEAjz*^f3?mRpRc zrlzWb3AinfXPdu8Ae{r+k*^Uos%9w>9;^&{sHbh{UK1*8^L8o4W3@u#>4~qrY9cV( z>ZQKymU8m)_Kw|{9D}ypuqW61-ZHW=Ffs-i7B@C3k77L~`E#jNTNS;0dBRKANLl$& z2_;DUpnT|yqUmz3=3h5%6Pa`0xOubY^6ap_cCK|)1OH+5wAby)+M)>o0Raug*~m2N zni<#hR)=H469c%lkB_fjS2dK>McT;9GHEK0XlZ|Tb{m$uW?1}a?%)&N$~L4IHQlrb z+<|vn+eMdu^&pOH13+jTA<)VdN6tynA(XbCPzxeWtUsvteG=CR(P;2zM3_ z6VBZJ-g7x*kyQwJ-Zb5?Dv^zijjk}DoN8u208#msPBUd(Rf(cDxiyTefEuIx8& z-l$71yx(Ao(0MxCW4O5|NY{1h)*L2VU1RU3W&4qDK&<(1Q1X)Yr@eb=`w;Tv)9Vw} z-B;(-aFw>#Zf@X2}&U;V9@vG(6 z##&S5`)_2puR8Oaw=2K7ySqh(r|>y|kerm1lzhLmynJ(I5bw_KzKcK!wYuvV{M1y`I3Zf24b(gxwAA7Vu>ixB+Ksb z@!fZJ^o++;$Pe1S#^v~~y#5BL zLYm!UB^7fuXyt>=8RxOS7O$C|jD+sLQ6;bRHp9V2ERQ)#kFE=K!p=e19+7cKcYP@U z>p4n6ax4@Kox*^ApKfa0!6}2C$#Uq)2 zn4c)DvPU=u9M!)@31F|IImE)e_62h~&Isiq% zf|kiYp7z|6d^CKg{a9Cz=>IXSzbg#@(}6HGBaNuWO`<%VsZglf4DVOKe~AbGjjq`zVvC&?bmaHzRnK5 zfmnNW$wkn(ejhS@McgK@WhumJ*AWc^1`1#t5LH-w-FJv#7fU>8 z=fc-lN*{Us$E+iZV51gn0lC=fLR={ds1hs2Zi0KiHYV^Qmh~KFII#VN)52u0EQJ|h zh(Ok|)m!QGk!9R~A|oByDcv6HZZXUXN=i!n)*mm}jkVTAY<{4eq@hB~#CMcs8TT9- zzh7K(Z6Q*x01KS+@{6a#GeJ>U=K`t7YcVNG=jdsXDip&d+(OF-e@DuUGpq_G6IhXp zSt(jybHT1x+f2{K7m=qK-X9b0uEA5Z@C-n(*Y*jecwGk{shaj`@@C?fwi{`z!yIc> z+x5MrEu5(qHED>x$E{mrcFK-EE6cnCdHjmN6j?jdAImUNl%-|_)WwB`oJY$5qN*Tv zJDHhBZGBo>FWXIYs!Og+-jvDg3`@s%I1dCF$r_JHx-Pt6bDO>7HrJ$A20QApX61q_u4~w{o^c7mJAj`*DNc^<4k>Z3{O)P0#or zx3&d?ee||IJ$X|$Z1!!1z@x8vwf;u5=It4m)>an7;(5{q%-f%1Vn&eB zqX$-jK|$g60yn+v*Q$|NPBf;Myj`l8l6wmldECHGy*>zVsfIA5=KLJ~YQ>P#Ya5e_Glmj6JxI_d560wdx>2m5R;Bk8! zxAj#j;Q5nY&Oj?!E2|g4^ikx~ic!*A>F881Uc8Tm7+Dd(Gm1e4bk1*1)=ogcZ8c{R zabL-Sz7VyOOZDBARD+oWt8kE!OtUscK+mjv-|Xycjf`s%NVW-%Qy(}lKE1)Mo%bAz zz!*e?$P*wqw7zoE(u&|88BQ~X1leTdfnSrc2yx7{?r+R8*KA5qQ-I2$hFjs*DTsyK zaE(Cc9IJ?~=n&weILn?QgtZ`JJ3}uvap!A8wGqnlDVSnkgi^Pi1d4*_i6 zxp%J#>@m}JSTRYrSfXP@tI#nqCheINN3v-u4y;|`&y7Ga_U-CfAcYDu0 zzn!9yDUGf7ClgaV?rik+ zZ3Z%tyLfcWs1aWuJ9f;ZEiHc=GY-);xQXpD2NpTP%5u!j{K!m*y?khJgo{`*A?@~<~<`hzgq3|1Sj zF3oAY`1u!vHv`8pCnT*Bmzr^a51OpSgdBUlFI{lQZNAEr6bsMBKBi} zkB?|ZBZ;;b9#|dKT~{2gk7|mN*&So@^`}tb@-MV1zasrHhtk^@SqR(Ep+HKv50V9S zJS-Bo;oy^?ou4(=pdgvaCh0ELP#W|6`Q^cya7(-|zN_bak?Ue2gcGZQ{q3vb#Yk3k z{kNh0RPtJM3dM}%X#xxAtzl6Jg^oT{kjS-h1^G*?5vPXQAvZTSHa5x*E+8u;0?xfy z7JXpADr%otk_jMeiUgrogsdXu&m%r>;&2W;q~1+3&w*kg?lNzdC;a{Q-zSol*mNbe zwMr?|L}e5+F6WYjp_Nb9PQ7w=bY*ddw=&y-qGNjI%x_% zQiL5?B;&a*GkpI&OS~mGs3(d1tnNwGl?>m0`!j=SPf_6B+aCqoXRhx(>86|WawB4-ZQS_1t~WP zJi>lRw8SJ+1LCrU8>Nt%2%?@Xod#H07txpgYaulUAyN zVR`lHl|WDS{D}2{$m7sZvm}J&yzl^$^qmn+Y-d6xJ-R3JN5dsSXzX(&KxezMEyMYN z8|-IP@nFI8tU}lnwmmnoi!}tzTJP&UA;c9K%RG6* z`U)ez$)l3FphfLo`Qfg7p=k%-!sP+zxxtEq2xLk4YTi)&53g?rmFrO`XCjwxf3;zg z!N?jxL(Na0{*X`61A|CJ0=He5i;j-g=*TiR#cu^pNtT6kk3q~2lxXed(aE4=7N{H< z9yX@VVPnrUQp)bp2sSi3*P@$n zDe@d$(#OltLGCf~o#I& z9f@&BadZnQE-nU~e-6PX4#H?qH8Vt6!EVRXM5pp*`Q# z2(yo7lBzhh(?(iKs!W;n=4N^U*KoDO%Mm~a2At=e0a8qJ6qE5Ro3`(HZ@o5lG0RN7 zgc5;=G!!XC6v?paJr%Oh=PL;kW}U07>kJwbRTU!0?9*TFFGKD_u*Ws>#*G`cGrxIy zlI6!F>0Tfv=DjY}h}kA|*2|Pdz$}nRRYDL;1OdR0lr+U-fy^Sxn*jT9s>C;jEg)st zE>2(dGCgiRP{~Nkc>J|hC6vP+H)Q4zy;P z27oM%TggFKY<_n3?)Fk53KLcO_3z(p-L|dbs2ZSN(ZYvq$av?8q{8Uy>)W!8r{KNS zy-)kYD&^@v^^?w-rTKBw)<+qZJrV@8_pi(|IheF0Ct+6@iqxbOuf+q#5Xi~h)CUX& z3Aw}k{YtMFJ6oWkks8v04ERwa(rW+s__$&+DNO)>+$|_5FnjwfGICttRHuTXlG0qe zX-XUGItFldwn0JkQTo}tyMjgR;=9hJA%nG>E83|JR)_H;(}NNjEbT$@LN1uE%p$$u z(q4Oc>QJO_JR+sQ)Kp)o(Z~XH1-UShrUTu8=a9NK@6_ght6w@gI(I^m=GRSkhFR)U%2qQ`}(Q_6w$ZC zE4%KV1ukg7ZsYBADD&IzImPb2a1u2a`4+2uAB@JZznac<4h8$knVAO614|wS+o|rk z)WgOMW2KTQ@Xf1v!g*MNB<;&}nq-<(1|ADaE_;HERClx+=r<;L_e4GP+e~k{06{u+U^E zOI22hz`Eppr>WbC;)JDqyoqKsR!zt0?Dgx{1qhg}3Ked!NVjp`y5rDOBBxv}f0W9Y zZ{2@KVr@a+Ncv3=ln;xA?gRVwT|=3PdobI<&E>=GgWE&m!{**HTMh*oJ-I#rs&MplSyww4{*asgERBi?yp9r?}tCwy5o>5i<-Q&2*Y3OxKf)-{%1MlK}&tdJ1C;JV(8yT!mqLw2go+TS> zda%ZlcNYUgBq&OtK+9?(7KbNoUp3L1yEd8fGF`CFgHkqxyy&S<0KAU-;?^}i4z7?Wxh3Dd`3@;I{Wiw!0$J=?OxC|*GsJ_r(6J931z zK$%eP#u6g}j;B5z!s^=2+4WV5hv#*&$-pz?DE?sNBPfO2E#%COb!005dJD2FW7EyH z>iY_ZiT?1z3B2@Y25I$`T#F_0iH+nCW}|EQyh~MI5UHh^AQ*V+8|w6+3Z4j|p(jos z)vUPeZs!WC!x*wjy>s)f;K2e3X)@Uo#f0Jmhla*q@&hf60gh>VodnedggJ z|AiGMhi7)QrRFH$pRJdwP=O|Ptvs}U2I~BLGvfV(tq$+P^71HhA3sZaw)Eb#Zak5B zYiRnFH(LQs1(hGUjU}BL?dy9uA^kjPYQDoV;lwZh@zn2odyjX#koDM=-#{Y5B#IEP zJ;??Z&iGYdUN2+B+{7dvIPg&~B8xY0jFJ}}t88p#D`tn!W@Wt_2$a4!o|AncXu7B?=A77E!{-VmI$zKXexe=t0UmhrJpoSt+FAd3TK*3 zo8Q{HGcEt8@h9Rg044*n{lkrM54Y|*e2qcUX22WCWXRS{DO$D;8P)s)SkiUJeshOy z+qd(%UU&-xb7g^P-56J_I1@0@s9o#`m}@L zuJw~AH-W|48{3VCyDya;)q#WyuG=(R;IND$%=Y!Qvc%oGu5}X*riPRhl0&I_Cy+(* z@TmQ^ZrQ8LJ?oFbRpnd}8bIc)g+JGuKl1(Kh`6zQ3eQ8kery5BC)082O3X~V=E*`x z*wK8)^WybOeLSZi?ClNVdS$+G+b(4=1-TUv2m26C%4UQ+dv@2_dpE}~xBo9Qj`E`jH^zPut-F>&#^*9;1Q%0WrEgA&5kKnME0N;jUT6oacQ7M8n^@q)P3~yLE%x|QA@-#d4XoBJd_Rw^o88wQlt$TAD+z3j^--Z}k#ccd^Ch1;Q_A+wY>tf&)@ zb?XZMQB$KZ-jOZHV)JoHoNxrnFDPmcLFAxoxoT;rM^8^LKGH(0TSR1d)IJ6<=q zSY5ETPK%Zey24;t_Y4bz)4;@S-GJg6@EkEL@VHl@%eM(l@yK8Xjs=BYqkv#K*!^~K z0#eQ&ADmZ4^)O(D7?FT11g2)I#f=)FMnfDDm~t6j=80kYQyDFP#37%QEN zfDOtQ!iY#{4O=?fq&`LG2tU6HYIv2CjxN+2Brz@1>@e!2r{4H9NB7mg+;*OYhI!af zdy+jLKOiM(6jjI^KLb1w4&_PM!=Lo}%^TslU_CmQduOO78>|fzwey>gK7r+cC^V0R z@9f|piB*`dc<-3F(928nsxE>i+w*s~YG!tJOE=BGP=IJVJ>83vc8v z^Wkjn3wk9BzzlL>2f5lr13HZnmpyy({l=27gf@Az9nyo!0_Zr3HEa=k$c5jsyYod` zBO{Dfg5H$&Os7}atNpbWJax;Q*uQI5U%Dnf%Sv!Bl1X?>ld zTbvy3vDVhxVrbbvO9Vl+WGy@^Cv*VMDNqeVoBcWCCs)G8V0JpHlW%XHpYT{T5^tod zqI}~b$KwNc#Zm8`1lpmKCr^&Frbf!GAelGg>CHb7rwr)6*$4H)rA7SX5$g2})TycK z3(NcVp6lWF`!n95M@k%#S?4FZLWafWRkyG~L+a&!`2_iffD)@Q6mY6-pM<^}at;X&j)%vB zdBQS7B(Ko4IB3kxp(RPPrO|0jTk?3G1B8hTNSoJ$3&nU$Z^gEwUEGk?1~ErdA&-vjD5#1#Z-%k9V_jv&B3(c*_2Hq#vv5SXu~?Bn1l ziz}ecP(Bh6dp$ik&l2vHJiV39T$?;sl=-;kJUu-{+{PpqmzSB}Qkpi|4mWe=5b-zo zGE_>0y%?zG&u=p)xli==>gD{D5;VsDJygSUC}u=g??mX{lbxMsC9f25Y9owW4gq>4 zIyco(5cEH&@Wd>6;4#=n;p%(`s?!;a9iX1T1X0BN11W3-8W>@;OM_Kmm1dpW=;_r? zoVd}OIf-PHv3%n}KD=M5U||g)=Jl_Aq0B?&jK{XG>(cX!q2!ITdDP+!HfgA@4@0)eUewgb?gd5khz=<%^=8yxuzExQ4u(9s zky8ruAuuNjwcEt=x{HTVry*sA0xKAJU7l7;7f>N=OAim&gXOCyHM2N*e6HP8U)lJ# zkZonjC^LTefbscG%)MVR5J^3A5d%Rkpqx z*#HU(Z*|}Z& zRm%gFU&qHgFcT$B<`n*OR1$hyuvQyhe@@vWj=V_i)f0E-+UzLb?%o}PqG`K$b@yO> z_h74fvcPTlOprUOtrB5Oi1ygbBt9mS=TL;{hIIxU1Z*R{B^ΜIV3}%F3pi{c%7T zZGIHEe^7_cZp+^+d*Q;-)fr0|m4j=dG^sD)OtGK1^~p|og6`};j`^Gv<#=T&%!>dE zcK4O&8Cfv3qr?mYp}guZUeBqSm8j#`yV1Ner!LBMI5!|fJG}#mq%r5ZzW98rJ}D{7 z$PEnV>Utc+H)Soq_&S|gyP?7X2896YWz>LeS5_+<8k7{orhb!iMW`2XnNRE*??}sg zaLIc0tFH|&q&zrwaLYH}SYb9$4Kmt8Pe)fhAYJd=cTW(gAX|2vnVmfa+;li8yITdQCoH7tRg|XO zw4va&JQ?8K?ZqtU_PDjb8g8C89&N2PBhlGWN z@5b{7Kpl-+9Mdjgb(@rlICYD7^gd@Py9|mRjY)t(65^3DLos*=>X8q7 zr{;cn)b*wqsNVPHae zEVnX0l1u72AX?^cU1ofdzM&>!E~MTmK|MO5Ct{j zFGYHinmyml1)M$Kv~JxN60*G z$K5)DbqrZN_&^rS!cd{*JXWT>^fuhDD<4*WRp%e~U}j-qVb5FCveI(DUq-Dz8Oq-% zGedgxvHHq$(PQhr^}r?hpO^WVL&nyZJkVMjdb;DyV0`K6XfM z1@HZ2YrW#wNi^x*yNwv=X=#zf!dQ~`9yG;_68R9IIf{5LPJ&U?KjKg_^Kl}JJ@Int z;glnhE?(e+4FlHmf6iD>$A#gRzz3nut2btxsS`;=k|ec<4{%h8z|um{wqqxt5iNTn zalioww9JZVFLp9XQI;hNZT!*k_190@rhi%n-WTm)WHfHQAptE66DW<%Eg)ur(0S+f z?bO@Le>ubCn*djKK|+5!;Rrj)>>t~BVo%=OLataDK9;P}+`)CUS^#>e2wAGBTDJnr!Qo;`;w zEiDObHmn?bTQ@s>>(;Fxk=gMX4@MiO-P1CH7zJ%7xZOGeq zk6ls+>V6SibPcbGKN5w#O2c);+#yWeJ3prMQ7zaw9P;Q-r#jH>+qZ9W@v|jtw`0yf zEmAiU?{lef|A~t#bsjF_5$2HJn0$ounLYv`ZCASz@O*t`P%@g&NC#av{Vaylv$K5O za0~?c^?0`|SMU_?dCR2||C~c9I;YWq@bRXnCpUyFAdv9EnH;hJq5E233@|1vmdEqOqz1wh=Bsac7wsME zlGMt#CuUymmd{jx{!3$~TZK-;0wRy?M5l1b0wTsUXzy*W@1jhA6eFTig0dp*dR=~a zcsQ}V#aOQ?p&oJN$`!3@{R1vl9p~=JSLe8>xEQr+u9C8|LiTqY={_vNAFO>`;vO zLRhkPUUNBV$o~A}*oJ`6-v0h5kQzpxt1X46&xMP+^a17yXc@Pw!<}I}?1Y?p31}ei1pn!l}|AWlO70Lr^3# z{y_Z{K4{PUTGV#vz2Za%DJM%magl=eV0#~$>@Go(FrG03jz6jcb4tU&i6!j6CfMG# zQ+YKw#JNavI*N!nE@As0VHsk(=_41cS#^;@D&&L5C7-&Z7|u}vwMVI-S0Rs00faoS z-+l9pi2Y&Na1eOX^R$8h1ylN35L01@ z=Xn3S?JfxqLozw7# z48AJ^(l-sA7J9vC`%vj^K+u&7>I1~E54kW_<(v4Rjx!gqs;#@0k&Vs5fxioZ%E{Liev4FiAUy<=2S4T5^n6OHtCH7gH;l z#2jDft%N6h-BWmF(j4U+QUlQrmWc}c4>_4%TFI874aW>p$!%`kr0 zr4^Lx@FKegH>Vvxngu5LX8}Zv7%}x+ar1dry}wp0_OcPoQ7{NscvIjp{Ry*m$djI> zlB=@~>Z%>)hGE+f@UoUCv%J}Sdi(nHsajRWf6J~`JP}_M;ai?Ke%RvV*t(|s6C7nN z=x?K#DFpLLq1nWclq%v-NS3d$ObM#>ZmNOe_au7ihnR2Z>HkzdEBZUhfI^VcE`Q)( z%pgU217yqgmRX$cM8!xhq~tJi(nF-%-c~?0*&uyL3ifJBM3m;R`N0Oj(ZZo$LkiV_ zglRDGdpo=8b8U=D7Jl2puw(b`fWkM=`*cLI(;pJsXQs(jApt-xlHo(R%9#i;_8xqF zpbqXwQ}las4@zCQPz#gd_EO~j3F|dPnX;CamIDN*p_hU~M5Nt<;fg)qwSljP#f4s9 zNzIvkx!)*?+9i(SH+m5fk@PxI3?4~0vJ3}BxCjCR$N*H(( z<|$w8hDp8sX6+|n`TpeTFKB4W&LbMi#J|_uiXoU-p>}aS>lql4YzOHU@P<9KJ*vQu z5419$D>0wsob7$^xz-2q8flfUb~*Iq7Myxf*i(kBYCAI^7hsJL{HjsKdAtdoDI5(0 zM`d7mMajWfi1d-NiaR$(t!&w{WvOR<%}7Y~*_kaco%TenSapJUreF^-uWg6t=?Fim zFnCU(PmuV(+qCYvQR{QVx}E@vlq?XQiq;g-kc2WO!XCWy`yyzArr@FnS8a=8AwXzZ}tJ5X-!xZe1?^KN;Z|DNhg++w`x3i9AQP8MMaNECmjWM zZP*`=cbJk!2AE$AgyvCY9mdAv?w5q`xXX)Q9IZoeTwL%JCP{kXXu|X7``i}$1Gt+G zO0m06{!$_QT(gI3p2_MwW^LMK3l%bTC~&YL1GxY^>d%NVVloG+Nqb?>Q6d|TNeum% z4lfqJ^X#>fRPZT|qeqXHuz=JExIVp%Qk=xfFhF?_ zSCo{WtD76GniSf%nn8oL%87ge{W9lKUpS!fm7EW~Sf(}~rE^II&s!JBfN%tvT+(}# zPj4{<7T3G{TVfN`fqQJ??USY5HxH$0KpAYxvokG8rI%PZfm^H@xZ{m=*Lc(&ovc{M z1~&Y{GQ<>|tR-BM3fqloTiVq&tp|(Vs`38MK`D)F_6McIwUdAEm!RvACf#Q1&U@;k zu|yHo(1zS~W0=uBr%0CqDR8s6TvU||*a!IRHd7_&*)m2TpG87Ncj0LiEJzL%f_)v7Ce7bjf>ciQ16Vwu z%ECq$I9p(*#FS8$=R?RfMzO~ zO#iC$Al4c_wy&9*oV<`doI%WAnC7g;Y2uJT4;Qip|B*rD`Ambd?@Lm#n*s%V+qz-r z0mHBA?Mp^I2l2ndO{fBr%tsSr2hS%_6UI~RBDD1bOojm7>J?1?PA*zr~&m#3gZ2ZgQk$@ z?bpv80t+TRqr+OniHLDC6(YT|=B+PY^)4O9CwCHc1|9u>{oH5x64JRqZ5nA#3h8YH zsDDn1Syq+ZTTOh(MFEE81nNw(?JdPqt}CHL;=AO6WZ?BQQp_T|-lt-b$E5cO5uLx`Dir zlcVDh^q7>cDPl!&dw)|L`ou_EdJIbH@Jd90)_PfgSevgsP2Xm2g5rB3R26-P%!}ut z_+KWiB&5d}F~?v@2-~bFS-Zn(3^S9Jzjf&5Gu$}2t=?8-g~mj6-h8>vAXGH@(K?E` zJ>0cQ&PqV=izmH(kndo$R*4yWRBx$DyJbrqsQHNES|VwQ28cBw-eZXYOepatkpm8d zbWp-kn#f!;M%!FacQ#fXItcjf+=Z%pdUVN@^pb!HBoPHzCHlw)w>~W3;L9Sc*xlV7 zRNVr{aX!782HHvq67dSIB^4X66ynAXk~ok~Tsir)t7Ib)mGjNpw|Qa;L@`5|9Y`+0 zZnQaCeBai$z^=T7-{M30KaF;Y2SB=n8XTl@BeY2U@Mm{*95&Y(L;)3OzZn41vaAkBQE0NefV`)MpagS@ zR>@AIhIF&G;}|3bL}vjO3ep}-pe*GC9JM{EODlJO@Mn_lE4zBjBuZEMICIn8(51uAX415S`^;I_L9V7jz zq*tTk>kfE9y;LQO$KkN?gXy)m6*)@<{l(kaAERy^k8PfS?-#ND_%195%Kd*fY+y|; zy8uC)^huOmN!{babNR&yEQ1Ebqo$5haw4m9DX=;QXx-c_(%fk$oE)9o2^!GdPObJCMs5f z!Ky!wzrsy`R;8>s<%B zb-a=R^?G#8Dw&kOgXV{C?npbToF16PEE#9$%I%O`C+jy8JI($xQ6A<#BZYP>2%!rMpTZlan^)DAMAvrM}8LKay+WmqF_apZ)bnkJp8-cG6A<3Jfc-mLElTNhx-0|SfG%|vNG)8LMb(74YHANA zfEbg>(gRtyzZT^X$R#GGMT%p1goZA1!V8M<9$AfzfHUY-=d0rPfPsZTMshX?<|r0Q z4e2c=s~BoW5)4}iI?~h`HWZnlPYP?dV9C$IOgjQa@hD3VdaCnqkI1YE_%k^O1a-LR z_wRq+xM>q72gh)=@V8jmu$!5p%fgtA<+CnB|DP{vd>dY&vj8;&R2R-f&r@1b!i{Yp zyNp0om3PQ|gT9FoxKV8nS>hUVu(}#aZ3tk5AkB~d23R^QJ^PSR9GDi5m!SGUI?`c% zmW`v}GmVzTtA-By=OInYp}<3qb+WC--9ZDKDE(eMxjJ?2f=xN)N~ z>MmrL==BxYnr{3<*JD)$`y!!a5eFww65nrWC8q&lP^FP_njy}PPdK3Qku=c3)>`&< zgujzvhf%~T~IaM>DkcU!v9HKJW$h{J` zZ{Mct5?xb5A0gTHxS1GJ77^Q6WE}sAn;tBM+_pGif%LzTH5M~^e7atSr>p`1y@*^N6IQ-l7_++3UteW3I9G+5K46E^nKNgCz0O}wQYEVcA;(rI1{n%ZsGpvf zw-@{py|u^D!4AZE$)0-m_U$^vCVrTUV+KDtp>He)eR<&4JMJKJ&J=_VQXwAc1Sf7L zW@n8eL~dgwDk9+@F;XSbkR0<8FoSB5F**>4#6aFnY8VLC~gi=lTM=jWMFz|LyM12@5E>kx2-v>Y|v z|M7IE_BkK4xH6MsDC$xU4jj?j0ANX*0>p)!{71HMp0D$c|9A)ptqzQt!5v{Ol*kXC16)i1li1jU`A<3HC1N&M%?F>-G zmB(TFpObCz7v?*$qex;Q7N_gY4Z%nii0L0qI7iso*<)-Rot;O>ql1w5kB*vNu~R4L z6VqtY`XZu1m>idcVMxHC0|Ptfv%T1yFJKSZ=Ke_9cJVLbT$b{0a|S1$gWQvVEu}8t z)kZp!;cq?-e6VgL-FF`PUM7&2%ZT`m<&9r|eGR5h(gsf|r_O7Lr+#goaqhh%~q0dsr#VqAr1(&qNJq2W8flyt8FQoBhwS73c;h>b8TYK@DsLwvFhXS za2yEC5cr_O(rPxY)Ka__MRqe1!m->_n#1C)y&xwwsa(~k$uUYGBhhf1Vo2effM2f} zgN+lKgc;)V?Ih42b`FKI+u#IY)x1ByqyYsKs-O&Zz|e#RJA(HM%q=-}U>XwPC6|Kl zluumH%*ol=`TL)M>w+iGzG8Ko&ch>9~R?6&=oLieTsw@s0yT-OB$h#u{cPf z-Pp-;W@bhS)QFP1`PQeHdBX4u12WqooOCqo10Pa6_>sE6I`~5-%EM@m?ut0@Ej+wP zW>Uy?B>jZ;sY$Gp_IhJvHzrsUfk8rt`-ma{3FH!@YDK;O3TeKA$Ulls9ddxy=~C6$ zizms66Hv{uJEgB)y0xy0Y>=T>vWc|8f!pS?!3qd2Bm+aRBF=$~tuRXyqk=RT`lh{j< z$VU5PX2^js#Jq_M{CPMAe#Lc^K-T?%_*tARjIz;m2{}rGbn*fJCf&F}6EjV0kEmfb zA_`_8*VQJp8P_oSvD<)tY66Rtvz%AjKtizO$f00lXS1R%J= z=N?b&XxQ}uI2t7q=bvzPRGYUJ7dasm0@Al>RjsR`STNq1t3rWyqR zJE|!;J?@U^<|L6~ zWpk3l)V6I?f|W@Hun|+G0p`xHb@eFRQmBDs+{W(6Yn-%_%z?v{xT7IXTthMb4vxGu zGfRX(5sgv=2l|14Cdmmn#I{YWo>Mcndsrhx{*br`9&~U49m*u zYCnaNf|W-y3(Awqcybj4`UchdTQfxY!fT{}-j1^<$*4h`BVMIlcl)a;DLBTH*mX&F z5s)JBO>-yFxd7`D^aj#ePwGv;?q|=RC)RRE>gX@=L$3*#-p)OHVsSJRY04yV3#X{m zEgSZalX^XLK~m=jevcu79=?v8xQ8MUnH3xrf^~Nmw^cd6)a^PWcaKH5{s5x19llh& zWAqnORBg#vQS^I z!-PXjhmhc)*{GlmkwleGhYn%*E3@*(q&}UHhYe@mm8arG&)y6_G~ohB&m2V3{N|hc z;It&&;D52OT9K0o&_H~@KOt+c8^H%Ko|%0hb{!C)G=pk1C5=L&&Ad(8{|PUUTY)@7 zT7w>QYsX_6N3fU4(TGraFmJ^C1mY5p@8_aCL&*1fser72Psd`3!9dGc`PZ-ODVXFkD5tPF|^pUT^GNj z2_$Xko>hZn=Rsia#7lquym$Yi+l;(4u?-b)0dMM9n-WCVHZ5A3z(HhbL2pTvVx}kG zhkx}6ozYajS)bni=`!R$H(a_uoe@%!O0*#xewv@3fbJ6M?$2IKQp~Z4{cQEO`VwxC z&r=I<0LMF%%=Xu#jd77!41&sE(+rTd|LuL48`TFtolVPSBJae{%bQboUNDn^Tv%(C)dE~R=uqdze1~;$ot3-+m3(wKgjRzpJaNhyU&faHCWBf5WWSo zw?*t$8m(*?Ey-!I=%+*-lt_71J3mtswjDYX+~m`L|6EW&Spj=v!VFbbZ5)5s0%4Y% zHiEd-G?_!5uH@z%Bh5!GB(=Ye#zc}QYc3e4k>X&;WRzV)^x>34m8WFvFvF)SPu}=K z0wk{qx5LDg05aMVq0i>@@jGO5@w=S=p|L5zXB8w;4;7(HdlO9{Hy9wP{p44WzmcEL!T&nBl;)*d2{5G1K;Hp`)BB0o{7dWLOJB>^z(!5kAC>SV88An?>1bkjijx>&6 zoN-KU;N>NBBaosU&OZg&7oQudyZP3JhPn|-Gx*n^Czb+16Y;hwQN01jEt5~-hwTS) zbh(|w=QaMs=s>db%t?Z&=;sOW$Hg~sCx;QZfz`r-D! zUG8Pie-G~$pS$qIS-)SrjDJoF{_8#X{p;mc?)}?6{J%ZYj5Wq6)CT$^P7K!c+p%lc z0UjQCqyN62zka*nqK$xA9WKcuMItbogTj(#8dA4*Mv@Q-tdp1jWt{lu5(Wg za#~$)$i`>LAydo=|8wGRw5_!%1Z#_>xt;qTH-MK;`26dE^8Xi~eEwf!_5a-?_y0Ux zor;(dj{-i9iX~-3lq^+{l=CYPpZ*Vf?-`c$xoz)~zKMwmR_qZ)W3a>yqGAvW7DSC= zL5!dhQBg#CGizzY0z@oWD59v?Kr9H-j8d$ih@w;lDN+SgK-#(Io0Yv_x&B|y`E;&x zxc1(0C-T0(@;tMQG3IzCbpcusS8zZJDXPR|BdUL!o0tFRk11@yKL$}b&X=;$I8yhf z7E$bz0a{v%z4z_g7X&n#wMLJzoqY+WR@OEje8BZtq@`OApc8Cxj>B6mV>cs8^oG|0 zf8Z~#^dUVkjH8_ij3pHZn@)G~Px5IQV5TxgBYFueV~&MEAs-an$* zg;A68kwQ+YZty7mR7NSW#~1>b>i!2}))7ip_rHY;tbV@WYCa+Q>&S z!-rQ}Y)3Fn?;aX|r-y(4=4ztjmcG0!Fs13&9^I)#Z(Kj$aFe^hCVQql z@XbYk{<|gl`vacmOdUH&X|g;*ef>TC>+9>os(DRHsTURv#zJ!M=F%5az{V^?{;pI0 z{zWCFgOVG52lK3Kv;)HLp~2lxGN8J5D~by>j?7D(@$bL;uRm>B%$oS6m<46m-0<$s z0si%M9jvf&>MKD)BEmX+@EgB$A;hb){Py>kmv6uEG=F8w5MCYsuYOBQ^M_e2E3^2A z&wlxycE%et<5i(t#vgmLW0yy}w0w9M73+KgzU|zpQ~z(i@fsC(93<>U_Nu}~ba-oW&O)1+ z+s^{TeAIKKbKUE&yk6zTxg9pcw7zmvd8_C3uRmUIy8f59Vb(B)>rNoF{ee-f>=PgU zO{FdKB}NRnv3z-PMTg#trs-G|Rdm?$#SIuVaRRz)aphXnzI9hQE&?2|I@#9VzRo4` zSRE~fg15~iHW*PnW-a{Y!w>(Xv!VppD1LIInJGC^SVIv_J2#o0`QoC%c6neuw<|ha zN}~gN3tpD<%rZ6TwoB%gi7Rew)y@Pd`fn@ad8|Swwq+L{Pv>b)v!A_#^gvB7mVaXx ze)i#oD0xafw8d!=lf?JWWZrURr_Q>R4d8J&#c%VQ;)m-FnK#x|F|pwse(%5he%8$K zSoJzJFKfb43RRx`aBuM!Mk_8jg5l9I_iYJ2M}GKk={;T1ZaqJSzt^Zt=0Gja z{9O~bzINk$#kkD$6zIguKxBRsU6w#IO96@sX#sQ2ppX$!(S!HWf`V5t4LX_lwE$Hf zQCP;KaU5Pf^J{vC$3TK0+rx1cm`pg6!UL!)2+{JRLXis6RLU$i9kAH019C6|az%{B z9~Iu7AqP8b>H~%VXq6^RCBO&%K;KhrGk-7OX>|0(3fl_Maxh)lEo%;Uj9Mou$01D&*Zj2-DiuuSwM8UqGQO%GjeL?JKj z1Zd{7Kx3JDLRh^St7SBdaK*t$W26Nl3yvJ$%#Z>ipR-hbq2#q&v<5=P(He+{>EH=o z_nrH+e!7Z3U;n2>&`F*+m=2b+JA~qU5rM>VoJ~m(&AIq=8r)wIwkUPov?)_Wc+ zq+|nT9o%THJz~T>jGgDy9)th|V<}617h#nz6qUYQ=P3MehZsi~3ZkPGPKsV*F&sFV3L2Twm7m2vBQ;vsaEoQ7<{2 zV}O57{fd+K{G+IO0W!~iWjzbS*9xbNndR>>jS-TWfM>30zIsAcd`E^`@A#FArwzkB z_g=@Ov=`qm=@XGL3}UXj!d2uL;J*?Zcn%tm>K`>U0glYTW-JYvnP50a%yJ!y{6=}C zzjEn1cxm`7(ULI&BiP*DQfBGGoSI|-w6{HyiwYeNn($1WvE8hqtb`?ugD$h;KIcVGMC|7N4g+SF ziK`uT-TeBYOunFdJ-I{d&&ux?L82~1z#^lLg7#+@TEBXt#VkjPoS<;hxzM$~pz8uK z@?7T{j}?z>R&y7#iM0^l`{neLr_TO!djXR?k-@mlM;Og5szO*OdIga$0O=W|H>K3y zxBqYZN!rUvo+@ziz}FUt_|QnPq#P?Ei{9KHo4azws>PL9SUr^0H%eoV`6-+R#M2gm zWAdBfEDOhy)2BuuKAP3XjFBiVdM)4Sm(LB1vl%~P43tN~?Y%|G@5odmjep7(&;M6a{)Ce{{M0oPv`x86 zA&O)MW4jk#PQd~UJQ!VUDYj$~^2V0W4Gav#|6Q^uquD;s<^`UV-tvOzs6=EFc+d{I%q6bbxZN>Qy5`HNQ#lDo_e(L)5>w{0P08tk&r|-nvk%+&D)-vNLHh}E*tY>Zjz2faG!r$%7ACbxg@YlbZ90i1T7F1JE zXXbZpn0;m%>9s+H@QWlE?;wy!3%0J^Ch*N!fUE{~u}EB=zUZDE$b7#@d-|_=abp`Q zwG_WF0fAZXuX?wneLeZttDC<)cr;q)uGNXfd#Q_!%cH4z5YP{}%V5oX7!A_V&SS`4Z%~s1pJ)tVQql z1LrJpQqLdA>JM`{z1G_M$q_D$V+XUz&hj|q5Cgcv6U2iOtkP%H{P-KGsW>dD_am8< ztTr!d_;&>S`%k@%rEJ->mB3{IwzbAo^l21S(t3@mf+<*VZqZz1u5zxJ(9+1+7X#C$ zBzw=cj(Rwp#l}S$$}H{fFTlaz2ypxV>A>S388^_sdN`NEgH}G8RRQaO$=n$+UC`Wk zlnb^q$C7L+o=$S^!Xv+Y{_hx%L!x_;9mUm*ssg(1Jj}$H)su$w<)}Ctin5OS)6Z?s z02nT>)O8Sv6J~#c<~Y085|8&DV{Wsfw$vo0W*i^fKjdpI-h8g=C| zybDq9kpJ@cEde)Kxgr8$dHj4>z7z?!=p+toE*%YKw9Fg7&ex;9E=Hhn(ZQ=?0T1I^ zO)h%U*^joE;G(j9d69L>1fP>1HIeM(Rf@joRnNGlcSkVS*f-$ay^$%JKW{%I?%1g7 zPOX{9$pjDaXA^g>j53B2L&b)McPe@@h4K%%V(R$;Zz#goQgos-O5XhSRikA>2%{u^ zE8>TnH~Za}*=SBuaQr>eGN2us=#jiPw)3cbA*MK-g-W-rRPV5 zF=rxYA4yS}osUvYPPP#01v+`nLZ3q?pq@op!O0Ry>qpB|C-CkEyvTOJuICN(#~OAo z<08-f>acDcv2IcyhIL`jQsY-A`<|AH(8-t^@Fm4^nuWWzN1eBf6_#TEwK4ej z%zKXiPN%%08pUq7h@d#ZrhjEUUT{uuoC@Q2Ufjvb(0HL+nyn+2M%)renyQPt>S@$~2I_1POfBbft zCG|ON#!g&c6J=9WQuIapo)5UlNAV<_npvX*au;F|8`Dq`HAbuF;|nWpesv^fWT_~< zCwt@){iNPXjTd*7SB~PjXPv|7YcVh39*h3ae_e)(lk2V?8E0d0xSTTfG{N%F#2oHc zw)nGEt9l?QKJt4-**laf5CelBHP8o)yKhUE!32PrT?6GJhFp zw&@z(gE$83cXUD!0{uI5d$pLa&6)7_=fh8l^}X(3r?QGAof>HN+!rq4ME&4gF$~qn znVBl$`Ij$WYG<R5aO9N9Zfp)S!V^7Kjz6uV$@^$o73LG61@$W9wF?(N zJ410;G4v-YKD%GtsFx{@MukC?Mm21Bc2WI)_@&VLiq9WA$$h4-^TG5U>4!nV55^H2 z50|53)7tE8P}g(fbPTHo4;vP$ok^w@f7#Rtt0@&Z<6-6Dr>Lwh2F2csz!pTzvj!}@ zw0icbn#k&I_(ty(H?lWbEidmyzUF}hJ%dz_o$pbHF#X<-c8>S{Ya_ag!gbF7 z(tK-O`-!UPrnhe0IxF3~4U9Hg%?zjRfx@xl;>sd( zb26R&lpMO+iEo#(_C+k(Sw{30A3msUq~lYOn-Y3F60nnV<-7+Jmeq=h2c-C;co5!w zuky9LKG+pox+ylrV>apWS_`TGPqcpkVqx&yTFI z+&MetpRcvRz&WL#4J z+xV{BY<#z|_rEChzoRz)9Qhy+a`SAIk>ma#^qdsqn!Yknx!e1X&kzcyo;y%BeR6>g zIRB9a^F04hykt3rf%$_KOI#ODPOkTNs)!t=5y*KD3k6ZUdbQS1r0b)2$D@+I=D4o% zZ&pS$)M)6F6MqKd8dQ&9Y1It>bE_+z?k%Qa>sI(-OLA+56n$BUXr0G|A%!{B{lTzh z*H%u|Zx2@E7@p!V>{`@EO=5CYK%X)EY*5d>MA7@U$eSm=FYY&X!6$ab1T%U85AAP^ zG#SMau_$I|Sap@W$|T$E#5~ao%a1Fd_P(uOCOup+m>Q@6oxZ6=k(^ ziOBaq8AqkP?97KK;`O<`h1Gujiw0KGv)54y85B^;drfZ0jd8p%^8QQABR1C^{9-Pc z*rRobakEd4PwwqApW!*p`|qg7vwv33O29gjf_-Mk9upilU(2Kzn1`SArReb8Sxxo( z<RIjqaAEBB7xGTVu+g`$o+< z^qx1rE=H!XCfv--te3wXEoS>ZHQynubUj?o{x1d*XkS!M*MI+qLDPOk4sf^}M}l!p zN}t>D^%X<=E?xP!)R}HsfjGS)bMNafqBXYPjFHV$kt_9kc@nVqAy+;yAl&h-#n=V; zCdlJvUW!5AhF8P(nsZC9CJ(T!efu|E|L?2+?yN}hHbzAqDFvY_+~~J(0k-B^6^ z!i5W?GCjvMTXN_JCxx`{RdZAUOqEl!{K}hcU*PWP8C%)?0R#GLI+AwJ^!UNbLmH9w z(^3+grYwv}rbZ}cN?~@#sF;%a?6Uz~>Bs0bN3Sv&De9d2??|330Clc-B72H}Md|-^ zyykZ!!PEX&t?ky_C?;MruY9nPl{95WaZBdgil&yQQ8gPwlN#NgBMTq5|G*=)jG&-z z=k6IiNO$A_HNyM0Fpu@Wcbjb1BhmZzAFB?}Sh?fD!C~EJs`S;{qdCJVHD}QJQ^}QH zx9jUSADC!f>T5qfSI%8sB(bdNPmtB2r5EYJiH>dp+dLP z>G^M84A_kWv{7)qUAB!;U~S`xae40jq!7~&&SoQo(Ub&awcbY~+S-a@O3rW=&5C(78?L=N$>_$^z}g9A7jAY)M76(SmapdU(~$s&D+A}e zZaxu3P1G@wGTq}+@;D7(+soOJb~z5_(KJpc(iT6<7!!02NPxO#+Bw`1Z>`16VjLmS zDC>zgVCJQ*Rls~bnDcnLqM`EpOxI+pqEX?XJZftL?DKYrL%Xk793P_<3nXm=(|x@cRfwya3+OJ&qpz_VR}FJA)s|9G=vtq-5T@z=-*$? zQLo0%RSC*h_{75Zk*%q{3;q7*=qHASvR_K$vyd$%4o~XId5_)dlSLPq=r%ETRw zlrdi5j|0<^84Wt9ZD$#4eeu$zd>V02Aknz#PR-toIO!M_yWwYQl{R-;0+U_dMZWTZ zpp#w4W^J6Tr^krsskDZyhcbd$`lb{(QjeM~kLkuiX-lLTG6r~i2t?e=ai=5jNXJBg z5Km_8!s2f{E0}_zb%#W(8$9SpqV%4Gpq&m;s>U$?(2o^Rmmxkj`@W#<-1+9so2?o0 ztwzgTY*wybJs>8TIwlVQ(Q9&ht^OgAg@?T@FQv)wpI{$`@-p*(`=YmdSy`ELkqjO` zq_M>@$3VB~;@zQl-=?-sSzCJby>hOH(iV4jM(7gA$6aNK8LVeC&FcC`U+4j_9+yRh zEB-X#98<=ThPyV_+0!-TUrL#{%iWzEDQiBOK>+>A#ng<&jM zmD6zb!G(9+1;_ZvxxB2W2tdzqv@E)|POxpec!%*Cq0#f6-|aYKJ`Z0kWUAPC z1!)NS168$c*Y(75OLka+17spUIXORKv-W1N9i&d)`kLV$I)~!2tcDlF3oHLW{ zu(|j_F`_(aD4x0TZiseQh))h$#c#PqkKMa>+hf0!4{WrH;(FjYn+HFA(Yd}r zSpWb1Nl@R?eo|yr+Wa{QR!P0OQk^EBmczJu+duz}|M}ueUv1`l7lA(peq5}g;?|Iv zn}4}P`d8nL z>zSq)&V=FkA~JF!MZdhf$dL6t*5!1Qe?vSo1^XNhQnqz1y~5Sy(O#xBL0w(l5`;!GfA`1^B?JY~pgl#t#Ib?BCgJbl3ihIn zm}`e|Ix6JOpoT-7A8Fr`2+~rY;&zN?S9#dMMaWRU>NcW3Ct1$>J-b}9tMSI427|gc zzd5C&?(5hLTo(UDALl>Py0e}==qt5Q?9N-VzO{~^Q?Tyhu}VrhXZVgr7#@#~2AX|e zX@NoQ{Gzx0cSEDfE~=H&_TN{9p}Y*1@=6N~W2ZQPdTQpmtCiQ!Oede$kv{S-KQfqE z(hT^H1wgclxu4BKVnpFJK&tcJmSMAGxD0aWwmtKN$8ljxSW*hr;6eZ8Gs{eW95TOr zMglzHzserC*j+}HKB2es>pz2e0;kUfo=R*BuJ82;jfv#)0cj{M)%gNERx1m_Sr-H(GI3hXI* z@?g7?Yhwu^n#cvk5^>O`y2fcVg}X>j{b`$3fz0pM#u5Z0us$_oT|5LQG?gcSxYr() zsws0BRyV&M8_$#PNi1~|_eUaOZw5;3d~KdA9?m7#an93m5}b;P^e9_xu$(v!x3j8B zWyWkmoEZh|@Rhgz;NkZ39`_=KySD;*Pb8MJPPz?eGzxGwr+x95xdga8VwVSpJ0x~9 z_f;WT?ts#4E;HEFT)x4r2LWIQ;?rsTy<-qxd;I--^Im^=mjNt$&&JddA}v38hxeu<9B$QZmWYN-`QjxbI#B8{j1>|><}RfoJaCm z&QdVHVme-_1{6eu z_w3)_nF(`umU5B=5R#lH%jBZvrJvzG$8x+%{vxU_NW{Ts7M|};So5NWN$9)1=RCDx z!@i%{oc)(Hl1!`vwII3r*uaQqNni9wohFpFa7hI zi4oTyr4Qmjks5;=bv~~fgyit;_`JT}30K$-H;PnLe2h0y>&+j56;E&GRECv~x$`s0 zQyI3T)gDqh+W0NF zkL{P~%j}k|qf#|D#sjq&z@?k*r|N5|+p*+%x-|pl!DMQ7>tB35lklxnW(X+YaqiqX z-TRT0n5Hjp?NPqM15Ot!kq%{F-FXr~AbU!!lx)vec#KTeJpPq->)7?&XSz}^rKf?e zg!icgnY@swik;LeUF|E`a9z0)FCD_A@>`0qcnCX_oy%@`UgCwSj{4PbJW^)mw-PWU zC{A70y6}HJbN`YjUn?oaDPA<>>%OWnWa0gw(SH(202IJ@ZcRV{6%H$pddCa) zyC0|XHc_(=GlhhfZMukg6Ft44Q>UhJ9yn0^I8O>QO}_Xzl>VwLt$m(U*m_Km=SAdu z?!x+H6zsv0Cqvr2mv#eW;cSemxVF~Up%B}fL*@r;G%no4D>3%m(tlqO5#CuDh(V`^ z#BZw4i5LCP{QtiwC8huNq<)YN!nN<`PktzE?;81e2>)pF z4z~Dh?fBtq8F;kv4I>A=8P9VVY#i|;yS5{R+FVW%e9EA6-qC-y-w7nQ-)2|l6;k!V zi7t-!Xwu0NDpc{FRKe5d%<-adiT~(}p5KC)6}7JvW1K3unFQ%aMB~UalP|P_MzRcL_fzsx)@jQ=vjysd)xD<5wt6t+qHB&bx;z8tJ>B3yT zqLZZH?{La@(+tq_eM{!Fc}_|fj{p39>F3jPsOIuKwJY3-2?=zl^M9!uuckI~fS8hvl?Qn=)(@+Th%TiY5Vf5-nQ0P!R zH1EJt%N^8uTKMoA7DwV#_LIaw6NiHR9bWy19PiEyi^XHFzr$Wqi$$=1!*%KP!jMo- zrLmG{B1?$E=(f%BsZ*!+WW=lrU#$e{CXEw`nnk)}s?sybW-6;L<*rb4^nK4UlRhw;>#6|8ceGesZ43S1v zl1_8j^a;Crj2&vpOEi89EEm7ZXIU5Nx+!9NE2m3Zh^Q|-Vn#o#{1#X;o|?Zm!~8p- zkGh=hbMzOU;0}7I1ZgM%0C%&y1gz|7`*ejM7(q{@l#*mBW!WNDH{i!Waz%IOq8(tI zU89mnY6-ME-B20L=1$ab`Gj*wbEhU#B=JqR`ilR#Y2{uYxdIdys#29ADU7FtJ5JUN zD-DTCq&~a&sA=7Mz?eK>Q!iM%cvJ*x8X6jKX?<|Oai>U{%0s$)En(9Jd|f2JC-7HH z=`y9 zo?C7b0^&r0f>4Ox8$Fh`*%O8_qR8K=`s>HBAVF4lhg^-PLh~;*w}@|PE(_pNtx(QQ zW4n1*I1vDs4{YU$Cd!`Y{P}s# z?5p}5=_wVR=oIBfJ;}k z?Z8(Ak5C=A`PXQkkwdewdtFQJUf`q=$<)2NO%58x=Be=BiqOu(97?U{v$g_qJP*OZ z9LHf#pY5-U>+x6p;%eFGHgetCf%PqlbG^1)oS)GxwS16kiN`#f`=ireemR3_!wJom zEzP0VU)Oup%88{ZKNbS8bYkfadT7!P)~rKZ^V?7}TCXaMITBBw@?R4C+HA>^C4yfJ zy;gk8IGvICK3B6g%?xz9MoitbEv28DjOmvD&L1pe-aZj5FXX+f1v02fV8ch%?4Aox zc^PIX;Oc*$bqU;*{Uf+sJlJW#Mg`8q@=sKm)>B2SMf_jXxVJP7b`SspNJLSMKh9+L znMq*A$RLrkBEz%)m4*bF1}W=>Grl{)ao`mr6OfKB@T5DTniSxiWWzkZ%Zn^EUXU6| zAbV?br@b8*-$q=&n+$#q{@@U6z|$l-W9OuhmGYdzL#HqDyFfS zF01Po1L=&%BTD&7yP*+BXWcA<^4P&%9sZ6~VvoiCmtO@;(;K|#!gtM8UT*QyY*Jq| z*rvxl|3|wHiNwo$Z;Q_VDCS8&@*6RGzYG9o$_#61Zb;H9^6iLdIHk5oH@>CC-DHBf1bBi z>NQWT!h?%&53sFdx>Ny7U$Hdd1fC9<=y*XZcXF)Y1@$D(+edRj+x)fhMkprUumw&& zu%3CB!lb)k2RzHD2iEj@rJ98Y&BA;9UcsW%oisl5TZfB+Y;1k)h#ctdoA^>{mMIW| z`NV-a?^Ev1DH1Z$Dp{+jlWJj(pcTQ#J!b^3@1ZtFOhY%hw1j^hQB)wHrbJ)P=xG;I z#KIFWv5E?zi(`dXGji(0O4r&b$@6>CTbDdrstuH4dfwIe&IF)00(iUK6$2@bK;2Rn zq-Y#8Swi#lvs!7;i4#B`#(U1jB7P|McWs!Dlv7tSjgbqU8rw=vQO8+5USL#krUm=@ z9k2*pEQWjsbLP@XvO{tN0J{RD;=as4Lj)*M9l3f|NXNZ!b*ixZUf>8xG}FC>veDBM zf%s6)oU+Hh%?vH6)G&4U_~lV9+qlPedLWR2-U8=2IB@9)x`j43z&>o@Fscvgr8oN zzn*BzhJ(s=1dhHJX-DXPh$luJ9=6^&S1(IcwL)i~L^aic1H@lRu`(PWsuMluPN^7p z9r^4T4`TO#tSDZSptGAyCjQ|=LY+oc-O)l6?>p0K!fRaLo?Si#SaH|oh58Z2$97da z2TlTJGbJYqHns?#Dd`aWIZ2>c7!)TVpX#zJU5%bk?Ty=`)`&g>jq`Peo-oz0R*xf;&P2_Uutvym;}R z!-o%R6ff@o$=c3tU+-M~{LICQ?*>U>M{;|4CY?$THfV3@%Yp!M#;#24kjx2fEtt3Y!ylIw-3VLHtN!JqaoQq;8V5|@>7dh_v6B;59 zwOO*YBD}KJy%wBpZb{iMxkIw2?%OP2Vu9(c69@Wjo+o^2eNGDo(eYGuy~Tj|bUEQ^ zxM{rm$8KZv8Y{Pbjn;qlV3ygbC-fKgoPqSEuTO4@Xtv^6qt+9w>(n)>+{XR|$u(xx zwXA;n#GtyRMiIGIow__SC^P@lH=AV8Af@v}FZc4_+tpa#P#&m9V6lF@oE}T8ObD*X67Bp;+p16U`i7%9DgS&HlO((klwP zpXk*mpfbhN-JOJ#0AF`ZcUwzKdLAm@d>WcT8*{k;9mPYCqCz5s2L3Y_y3{`kM%h=A~@I+ELX(s1p*eSn@qA#}7r9ps4t^?5qq z_!fipZWw0wqdE<2-mcTw7uFi0m+XXPr(Kp+mDqqIM)RHD1IEVxMfJ+@M2{Oqt4Ipw21^Bw%|ET#*>U~&GQ zCo^Zx+;y|P_F$t4q2Jj$_~Z_X(fQi5%S%9JcUbzqMlMc2#r)&tXl&bdgn=Hbg(^1= z`=3ne-trtH7<5V1)zLz#7C3c(H-|&i2Z-&A;SosDO&(qT)F0^A2`Xx1vOPIsI0uh^ zTtuksC!5)@2WV_rTuZZlC7z&n>+UXZ_Fn68@nU%ng6JK@kAnKtebG{Mvpk*prwJ$Y z^umhC{BUo={RJXpXfn*m-D#Qw41we%sy@tJ;Y>j!(``2X-=g^paHU3~PP)n6|k*e~?Q4EpqWFW=s|gDSP= z9K~&g{{58FXTZAyI&Zo^$0Ws8+LVN@%R4#Dt~+rSI9lj>{oH_b{acW5zX;>{YIJx^ zEtJ@7N6WeE8s_bxtPj23z^i$&b?-^Qpx%T9%fQ+g4i!F59MZ~{qn;ozjf-t|t`&U; zU66X3fp+<2Avsdl{)}>*aDQ!U2516wLC!6n!$&2g>z5_{bOf1a&ksnecB=o=^ED8F z`fhs3-V(k7pD5OD>^CYZQUb)*f+d4jfqc(!bgX7hHX`=0Z@Ia>EF zh32p{|ICmo0zp>`SG{-=IzhIdl=A|^pmGr)!GGX7$#Kg0NVmF1ybBL35v-l*@<69^ z)J^c;eRC*Id^1Ea06RYP&5{Ixc&4z=b<2INlLx&S2UENwx2eKTc|{%~=t8G6@6bP} zl166QE8it*wi~~#;Z$u?8ZF7(;Oox%4X-L!=*7$S0HF9jt-G-5xs8QmtLVduqAJ2E zeY+YNTTMx>JKBXMo+em{(Z44Os#EAq+Cy)MR}P){JjTBWHjN~78_T|K;|wWq1p?ST-a}s z7UKB({?i&X425Sp*FLMD;h#&FEx^96{KbSNJooSkD;%7_0~h426*a@FNGaQpP{e~3 zngPad%cz}3iE*ozdCj}rMC|lnl3K|020gt|Y`hhXsB1booImCWsnRh~e3@;2a2|m_ zT`#zm4L90N4PWLYLohS{=O|-bEuLUDo$6BmDP<%=F*qiqxiao!x6X zgTtjME3D#bbv}R4uetWCyj`>o7<8rDaq;Az zQZAqvvAQ$pRR6T_3B6pG%AEnE#wl}L)BDbv14Dfh0ZPMe>Y7q3AgXz)b6K<}* znNK}?xAW`f)8riW=wD6&vqZXn6TZMdwiaUbhXjMxw5~}O;XyWwt@pDEE-C?*`Waf`|)~u7%_&gNdRFG!*#|f9o zK*F8g3((OCG=b$6xTCRjdCN|$jmAQzb`&g>h&GL2b0ZXOme#AL8QR}9KiV5(w~qj} z)yl+en)=$zy^pRg(;k-jd2k#R9+l=d)zNkUT9F3%vE+t!&sN>OQ#!KC+irz6)s64` zUr2#0=$`J7xNqI~Xh_8IfwoT;o2&ck2KZFJiJ3Ed_K^R)^*@3JZD+&pHJ=c=anTC# zBa#-)Fw@q!w;XpGY2feSPAM=!;HBw~eyOhEnHk;uhZv5+X4qX4Q{WRiGZAj>Z9X*b zV1SClXEx~c-u%jkrmpQ2#59olMS~SG*to*jH{eiavvaO(d2~;kG&#R=*VbR^iKNF# zQjryaGr!&t2G&dXaQEy3yQ3$X%?(&#;-+I++3jRdke=}NGzoW3i{+LT9ZEfrohQpA1cG5rm4&XJ!(GAIUj#aOP9 zxqV~irXouqRhT@U^U|T(-qrS`xi+26$Dz0y-x^eG_`CEPTdPD}D4j1Sc=vrZ`b+C; z)&w%^$C8va3-- zPH1J*l&;J6r5U$b*1vksxgB|47gMgzc%3x+HUAocl}v z6$BfJj&p+7J1uFH@_=$u`N~QIos`h+eoanodt5&yN3DIiYqHc;I;E`jh&HGg`i-mr zVV`Tp_pL*UxC_(xYY)~shELL$-d*=QEwwy|d=j!~(nJ9CszItO@N9ihoqyy$>xnEv zIiFx34>R*+OfuJCy{(=Jm3?Vh0K(vJ`;CVzPc%{+HdAQfD*r4UX+NwDj3hK zRI>i&BLcZbTE>ZTkjWXUFl=Mk1hKm3t-z8KSOFa@Zu=YdR+&rw(k$||?qDIoPu2^J zau)#*Zd_UaNlPZixLdYXk?h5wdfSpVB;Is?S-3_QhN}soL>)oAi2Qi!IsfSF6SfySSOE8=hY1JCs1Q2=9-1HK8bd z>sSOGEo7?`5Y=DxM#q-VOy2BZ4;nJz6H4{o(I0lHSO5OR3u%eL+WMzepzaV9V((!c ztGxIK+L(?OJQRld96O!yT0gl!h&Wk_MqA>9_M{8ZJ@1mS=2QY`Qh*^HOKBb@;^4G# zI{QUfONqyPk50hjv+Xv5Q-HBRj8Hy4KWWZctc9=QFhHiNDOa^eci^(NFav zZYSjYRei{_D&Sdd$o#GjuB5}&9LKf4S!0LUU08C7v{Q2IP{NWb{0&8Rl?Rt0%|lT7 zVhP2B6RTT&x5tl8H=h5}b!VvvS;&H80vV!@7Z`9b{_(jEsBPVaiorVd#n$%#-70XE z4uK6%$4rU=p%P_K2S?tdE_Bk>2Wx#!qG*etr|oDVznSl;JruxW`SUCY*97H(zTX0y zDZk~xoP99N=^536V9Lr@IImIkwTBF&1V!)loWb}V{5w4slC=eu({5ILlA!HHcWtUK zt-S3m%(E%!(F4g~*S6Czbtk0Q5s&qv+hTh&bgY8xF7R1ic0B0XG0EvuXwA$~rt|$Dx$hbtPDEhBNCIa%|0Xo;Yri*Bdv2! z`V@D}kkl#wCp5t{H+|*~N-J!W1TEd#2V`NX7Zh82x@sTLmq>#i!@fx#S9sqT9JqI z$&kZRGlanyPlcs?MGE<^$7X%YKRtje$H1u?$SE#Umx~UmDk>r@tiv*5-d|NGUd0?2A>b3+}=IDRCJ(WPlklZvlR`p zr1g-sJg!}|TYXa|xz3d->t_0~0ckrM)-kl#3+D8W^|0qfYlH@itW7?zMLpcGxI%zUpqRyz17bvp7oGNdb)LbmAQEm3q=@e z{qyow5zs-qvTlk7A+902(v(H$U;?k>L^FEfR;jSbb5S)4#Gl8*zjYcIs#eISMDXmT zFWZa@ZD~nSE-py>>5E?b&b~wW(|PTEMEoz$jAn_F^^MNi>@uUWQ>b|{8Q8R}6p&(f zFYPfXP-kpTiS{oC7w>4WX2`i{|MhzD;0B9cj{dE>a;HX!yCB^7H8h7l5|xpVSpSA_izTp%!tG2bvn#sX66TNmj(+0;eXKmZ{Fr83%fuPo;3rW{Jy+auQOrXKOvv z7H7f2L~6X?2wVhcz?oZX4GxbQblSysAHZiKaNdz3Sz3k^ck)DGSw_IvG(<=0jvDz8p@n3vlD@+8p>;z3uIL34gT% z9B3Lv>dqK9_9tXe@~xT-VNn$->fR?#L;Z`^AW5Vu9RBAnG^V{9)mxsFlpA8}2Z#*j zhp+oJQ*kbx1QM>OeL zW(Ku^$?78YZzm@w4=o*VtDL2<8%%1fQ4`gQMyhhCWW+KyMOgWUT4XtSU@Sv71xyj$ zS}|E!YPIw$Qr~g1)Z?c-iy5mbiKS2Z5D~E|zr8y&YOug)YUiJEDI($QU;zr^C@}Jr z3TvnTkadjzPg#dv`tJc}B~swCqF@989)juNM`5)j8f`%|0}W;j!-spdaBgNS%`kSe z%$kS400)ngQL?$73^}k?T72*`nJ|+~ z%fV>fgfJl*B(XYs2%uQnW#&D|*L-M?Xw4n zWS$wns9~LU(Du0e6O(;_c`Il#w^SkS>^x>xu?uJQ6_E}fsKhBbDy#~6> zlXDPzi7G9`kxV70^ChewzDDXRQGqPo5VOBjPDT(KZ!Qa>*ONKG!2UlqOF9fNG7&gO zW}()xqyqg99Oy#Dx_-x-*7JW{h4wrIgOmOg=Ewb|MT-uxIv+p-}-9p-MoB_uwq((*8-U*omCAuV{uuWT~bhIaBU6DLAnGh2dWMC}V zmuBg?OG2brJ)F`tK`;w~-!7_LVbb#;40j{(C>rOdt(JG@Y{*Us;fi3dZ&g%eFsb?+ zURnKR{rKGKJ*}6`&$|(38(RPVO`qOh^xAAF;}(Ig)R>A~ZnVs1K~L*@!$sBix3m%6 zcPA-)tYbQTr*;g4r|OhW(`nU!)?h17_Kv60dxN9AHKxDp;`^`~jnhT5Y0p^^-X^7E z9 zeORQ(^39^@8%0Yxuk72_ccG*-BcS60yAdLS!QTCQ6NU&-EDTf-p+X8>3k43fLVzUN zt?rT!!v0x==wj%TMzRM3vMMqln=-y;)>lk_qCZm12Rj3M%ZS0&2HJrRLIC>85U!|D zS>fGm{=*7eT|?c;eER3}k4kxfYnU0i0D7m3EI{rz=3M6c#nOP@^V1$HF%^-ERN^8J z6~*egrMpCL=X-79e~}e*r^_7tkbu6qD7^_$EID}qDRR`7Csr$<}oR0E6do%mzd^V-&%#2S|0&^ z_L{W*OIuRqZBF{Fx!NbL|8Crx)!M!!o%x;bTx%PA31VLiOum2P`sm!gFez zoV*5Z<}NBKdaYwHFQR^iS=%iFX_%N9>*AsXRK~vJ_(4*UNYZ5R5i&`kpr_)76!WX` z8;dQ(DjDtsFi@2%kbVc~rw>PG{Tg)AvL#@yNFoK1oo_t#hf>`Xkr%d=GpmWPN9IVm3}YFXOV6S} z6c9Qe;Pdv_qwnF86J#D6gvwE#R{C4S@0!R#MP}$l8Q8hrT87)J!0Qr7{fd^fHWq3k z9w@()v>;%AT`jy|*JWN;noF6*rqhujDCf#hP^HW3Gsg_&SP+>|7pcIuj7rj@7aZ-& z-6^yr0t)NtF}a(V+x+gTkThmUvmKJ>fRajYAgTdr9s}PE9U`(b3V;D=uFk*9YZ0-U zpr9;nlF5KDPk-b@~zGV8$^72zMd?9tH42#gys|r+6n?_;k{OKRnC8+XYB3KcP_4u1b=nk&M$`l3yD;Zty7(%zL9s<~}7&f_*ZgO{|b8nR} z*1BT!`CrQ8&eu{9(oCVh6hU9r9gWaVs)(fOcp3SO2z${EC&={nn~Ac7Nq-YOA=hNO z-IV_JGi1@Q(IRpc%A3+`VihRHfSBb!$$=?SCMp#9W96I4lZt6wh#yfWc6b(bSvH{O zxRPK&8>_z-O)-Pxifd*)_G6gkMTXR-qNqD!d2LWGLJTG&~X`>k~d|~Mp|NfBKnfRLV0E~Ph0mm)OA5|52f?3VK8aw2l!9!iWp;oSl#_B`5t`694q;_zO;qF0z$!0L^}-H%(3XRT_Ii zVlgWDmb9>wE^Zn_MaM*)DIyiDL7Y+NSFxT4jF*LzAauQq2W8u44tdq&DtM4!q}B!w z7%sUOZ~XJKhYU{$e9@2T5FyG%yoB0L{nO;4`$xJ8`;-ruGw4n1iSmE68>&Eb)zpM| z^I2xQSR`{&4@(Px@!deF$rY@`jZ_Yh-85b}<9oxC^yt5>C(2=}m{&I|zH$B4Y*PZO zSbs=B*bVFP;>7=q&E7m;G|TUxLdxGo3aifpp$(cC7%NXn;3EO;dTzY_(OlNMG(IA_ z5kE_NhMGDW9CbsEb&PKjdy9gMEO(rSsv=fI?OH0{B~x!jgv$j=`5}Nr|@|5 zZTQ8ONN&Yq zND{}J`DbA)MeHq<9=@5#cxXt}x6Wl0;csECMy>;dQ|j6RMFmFJXglb(Q_Ex@o3 z`PvuPmCq_7%Zo&%dkohRR>#tM7av-`ybW#*OO3($1ji0aC;TtIVYe1Rmldh1t7ef* zBfWJ%bl8D2gFVSnQ=l8#`O1;y!kq{LQcykROT#WR-sgJ!wIFn~KUvGf*1oZy7_Dzn zpiIxZo)@+6UgsMa0$dROk;keSRr#v-cuf-j0mG<*150JB4PBWEWy7%7mka+T*X{6+ zP~oFc8=t)HBFy5}*PQd>C|My>GLxcJrXLmaWP>3rHBm!7j}~MEHug(V7HdK)=Aqrv z|DUVh^cXW39Bpo}jjIJGR$6fnMLjNhx<7@kGNS`PB;db(v6jrK@t^Bux=2t-!oW0L z5W+PQdZy~d=OT8J0kL+@LwS`3;>Qv=28Q)%j?-@ngs$&fo>c;yyB!lbsHH!m+Sw`f zQJ^dDAQ7alo$t~NDTX3IpxbqxIBciLo|t~vZz$gW3bPePf!0Zeo=b|xLBr>@oWAnE zKRQck?Dj2ntGn6f81PnROep1=q`{XCBGultENWu3(R`;YY3vR*Ax+o5UQ)n9S2rxyj`rtM{KsR;)78uwALq9B&>(#Xm0T2I<< z@%xP~wt#1~KaYY)q?`u&I&4PSV)~(BTn`8VoF&>E@ZBrmkHaf3MMnE2U*%!{dj5(ok2l;p5w6Br`)J?qLj z(gS01MIR3!VG!K;jWJ~KVe32Zy@VqvTw>-V<{@~^C{yG)ok!XoRPNVhxwi~2ZM70F zgR5~(Y!3R}tAkc+mga;PgP$*Q3oGQZyZs)?HruWU@V#J}Fj&VIhHKv>|tjTEy&WmvD-HZ~()ebgy?QQfuxyspT{;SzF*~-SXPw$6SkMETPI* z8!}}5_Oks(w_;g9+ABBw??&SfPY;PB&`_MBPZ`!0bv7MR(Vgz|R!uBV`7eF1K(5p^ zF@_BV=3IJaj4L0Xz`t2m7PGU%q+w;&9`4c)-q4IQUwU9?gUg^h*3WD`M90GmJ!L22 zesG9+#S>~Sl6j>d-1ABK6b^pP?-ibo*(GvUtoKsyc4~cH&x;HckB^<7U~-%?N&SJo zB>{fOejB%+4Zm*`yA3l>tpO5~Iy_s8ENYwo{0?zr3+7ts7_D*1b-mm#)=qoeme_LT z*C*CwcGZ@XYUa&vny=dWTf!FrEGljBn?2Xy>tVamh7>trr{x-XqX=SE3ef$A4ssi$hr#ZgFCvB3wX2@6M4zOsy_Q2%@k3Es4J+OwzcoD{>b^s6)1HT81n z?yy)WfgAR8hpW=sthId`h?s`cMP;-A%KcFux|2bVU4)L=Y6W_pC(0xM?}0H>*907! z{p#5}CnIWDQCHjzKH@ir_FgeWxtj*Mm*bazVS@j9P-R~!EOIVfT0`;-$)G%*;gZW{ z&|Z;6m8C1AB{CJ|y>3UZUo$JSfiLd{@AZ@8-~V#k1CThkyA24mrFtgD9$(5i^iqPN z7JKnJ^#YTst#Kkey`nUUj|bCL|Vw?Z@0~Dknw^eM9fZq z3bXp7E0eF{?0Di@5gV*yA|Y&OT4PH?T#KNjJ}YOIO_D-1B!gz&_LdA==&On}e0$ka zZ``mufS1k+o`BX?V(W;KHiu{383-iy{Xg)J8@QWZQ94avan{6!xc`eLrBlDG--O zB{Kw596MN%6GUKNhBH1)b6@H|i~^SLX31 z1cp;}XwnUfKiOTSGDl~b)TH%EkWPhaPkaal#1u(~h*soww@|De+9T8v*hS$}AapWC z`wd@UtxI)%!aGSfD&>Oss~D^YMi`FzE&|XnkE+j8Sel}GPI4X-i!4MzcggL`;}vX* z3^wHEoYo6O>BKZP@Xi;l#VcYuXy5{4Avo982bNYMTom(_siH+UuxRF+q-4nSYmpa9 zj_L}3WK1>%UKt=(0q6&2M2x}bO&le6R6gS_1uy6i};!QS4llele%cO_Jv)m{Y=%ffFW75G#a_D_;SOd`0O@ zLE~+5l{6+|`3TnC&c}TUriuJPzyoR5FlsbBcSU}SSM$!c(y99Pz~e%oW}J-*%MV+!ECiYBdjOJ7j#|@Kx%l1P&+>xsE&UW$5WY z>gb1hh~b+rqb+iuNI$?F#FtL2DJT$JFZS8yH~l}o^S+$I>CQt0Gn4R6+t5#jeYum& zifE?us1Lh}6`Z1x5*0jqPhml|?Mv+Nge{)wSQ!_VVQJ28nks0vw2%Z8J*wOp#g8tGTy)iP znaDe&`4{0C0e>nE3!LAV{hmbdV)K}wg9fZ>Ju0fK=@bsIPeVp)KCDtH6*Zxu@&rRZj2o>bS z7@amwLydV0WMQ`bQLK=*|LXs?cb38tN=e4k#T#YV;|Uc2eDjvI5!5^{n#4Z--de;s zBBcDe_CGwop2mJD=v&*Lf`5KMvMI>9eV=fw{BFYRyr$N!ElStl{lABTn=}zIimHl= zq!;dFM8)TLqsi%TDpf5lO|5HzvaC-Z{5HNx^}9Q?BrNg2{_6LHsiIX=+7OekYrfmk z{^e<6i&jevea(onPx;Uvs2_^rU|yE|?$`SHs!vhsb!$G0c2lzxD^hFV(vOv{dok+o zSCdO4J@{tpahxsQ5N)XH5O5_z4fwH-E)2K=NkrMRo?zE@iTOLMQ z@f@8zSob%!P42wIVht`QauZbv`z;g?6JrtiwF4o~GhiO`FiH=MWE%IoJuxX)P(`+!VkpY_5-*5+16XPD z2T7JuE9Fy`{3%iYi ze3*uR+r<4X56EKuEu;wb%D*KiS**XsIbllvS1N{%EpfD4d{o`OaptMusMCkm$`$3rh1sH!YnLdP7if z`{>l|ldoO>Y3uFjtxm4uYAY?i+o3d|e<~=cV^^SaZDO6X^2k8$jDMU@rtn$*@h|%B zV>B&FRC9lQgr+rx%NwGXe}tyR{{aW=V>GS52gPFjJyU4<{r8$;v2JhOm;7mnImlC) zR6KSF4Oj0Vyr5$8w*Y1IR2GGl zO@&UjyE~D+39XEm;Yu5^LheirX}*LD4i;Fr%mo+S839L=NhL6i4ODwUEFS{oJW%(v z78xwQZ%YaS0I_)e(!_0M0k7tJNIAHl;UQKu~`uQ3)NM1mZDtk^!B%;8A zvT7Rrj95EMe{1rmuiB`221vW#Sp14=Ygjqj_mIE)z^eK!k%pTl9GPx41 zRH~Z_s`g>1_Nlj0e2KioY7*LWxR$g6D}pKsy^^`I7}yCm(-UM^0d^pY94&)h?vIpw zPL3)f37T91>pcpiz+>b3Ou0vKKY1rhW+!mKb%a}bvMHa6f|lp4E|gnSFaxKCRHn;~ zWx3yXux6QCP$*SsP|j1q8LB8rt#+jEg*<^+D>9Vv>T(}_3ph5O+60e+UKXRCjV192 zdaq)w$iKvV{4L@Ny2PF##faAVZ%}8VvDliwuN9)GD zh+;sH1`!kvgRkHvy_|cF#Xpyfxm5+sT zkB4HOC_B%XF@qXl5UWq54;-P&$j*LPIU7rerah!u4MOWkpje+> z7%b;{et`ODiVkxR=)S_l7AeXqmRFEsMRRTWmYCO)g-bxQ?lFqro3-MhCr4RfpA%vJ1!ogjgowb$O> zKB@1Gq;e880>T8+$=L1Qg)PxYj1j#|-BWydR5uU5R_2=@NtV$ZcFS1oqi^XSo| z@(K{3TVG_PFAk{jUc1{z2(`hXR#vqX$R2HSr=KD%NgSoki8_Ns4$9U@#Girj|~(^ z(mq|?>-)x0@!v}CLpfBYvU7?Z9#DCZ;5ZPsW8d;|zcWGxA9qDakvN;7vRKud`vO5X zPWofZN9tq!z*cg7xy1^?2m(mEypO6}109Hc5vrlB%?A)KVt-xxOYoGau#&eAM4cla z{YV>hH#<^B{tG4QVjY!Ho~|%_+lI`Ek^ix>>@F{hkH1p_(4=QIeF7wYl3E9FR~n8J z9c=JslC(5#D-S%xFpz0`VC<1!Tn+KiYx`W088s>I1o`I4rm!uFJ_L-#vb$UfkN~#r zsK6l%;)GkX=Up-6pEsW*>=yJZsY!cQ!!|93$*+l$HS#AF>r4COu&r)@yI#>_)>$I% zb$(s?N>~*iOO7<+>({T>?BU9cvsj#~DR_RtLuF8^t3^eGZo$kQ6!A_Yrt>Ka1yS5HSgWlPk8RYWolpd)8k4zt>T zMv%@?s4NUAQL;_Sh{Qw`WWzEjQgnW?;m9|7JD=_*u^w<-mQg9!Pn#?cL6XKiI>;5& z!TO6|&pL-P!KCV8s;DPbP+0};dva4yztTS8c>-a}i=kvmItjjPNIA|Y?gU}g0nWCt zJC-Gi!jC|n73+8j4(PtUw@1f@KWS`iTn{&nq<2Z*o%BARLmZd777}@jb{;&d?K^ht zph{O_y;83{X%J((P;zFU=!}}cSE;|9<>%{C`ty--z6lwF>zPVGG%2@2GBi$g9OR37 zqPKv{`Ve(*zf!q0sofFL1E$NHsH1RW=VMg7GNL#_HzAwM-!?yr5X#trj@ry&FtSQd zMh~w8t(@vfiOOmsrV&Qvxxkhd(qrAv@Kzo#n9wWz6;M(d=W9c`VXd6ZuKL_ukSI-F zm_Q+#aLkJ4Q#F~`2swbn_%-$IkIq}L;FzNT_dP`9by8M@!K~*=F#hu+0tLFL)|#R zbY8)2z{6jhELEbeGaFx>93r{i^Ds`L@LxFKELlEsQ1O5+;9wNOsUGDV7ZI_NO3+C_ zfNn^Cs>P1&4^be^38>%!P(Vei0u?z|8mC29;vQ(BB)?OmnZZJ9xJ#bn{7l_SRX$-;lW z%Ie}r`nP=0R%asvzx#Nz7y|vczCPCL=aUaEE+}|oWz8pLDMaUl;>cp!d0;g!BaQ%h z>VyNqBrj9+p`wFC8HUp@UswfMJmQIP&hMRz?SkTF1v7dQr4n#*%pAu@eAo6V=aA}GAz0CJ6xA>%m zIPygCm4)RIdC=n?@mZ(^yM~WnMZWHoF&iaHq^Qw1I-T)espX=dw19-;L4w|-9mgYG z2WED?!C0fgYj7_;Au5;)+i|yi6#600X{Fy=heH9jHnNu4Q2OzX9aW1%Hd3oXaxQ#d z*%Q;NZmhlTH5l;`a~8)F=AJ_|PO48mY6cBk;`EnAigFK2eqDaX++txkm=wFrk?llC zc@KNn^2u_)GAXAlVrvQtiVP5WvNeqP=D55J78{eMK4j+~zE4eU9-y-Ga5XF2WCQwl zmx>=Nysi~L%Iz1itd#U87ej=43E~hZA;mdH{t7-q>!Hm^J1esj5qrEyB~lV3r&Tu0 za64IPg7Fhus5k;kPBW4yuQL-g!Ko872Ci>@F_ZiDSXNu{Va+%#8`BYM-vPxz{INc~ z+$Ryh7ZbMjGjb!3C)>}LAIAw1rR!SHUw8a=rpNj7V#Xx?a;E)L?rUE(!)pyN`;B#q zKHG7?@gI%%Ld|*g!C}T7AJM@e04wX$G~I#j(3F}s(Bq3De6L&)f9}G4Syt6m%?Z7f z4#?{Rvw+Vu9`qmev~`FaKve!GFGmHxMw&kLu|ohoGojJ|Rhc4{@Nac+c5HmC&-jS6 zrIo}A1Tn1LQa@ZOGc|RSIS$}&O2q}oc(##07z%b0Rm@x16m@rZLCK9H+N86clZeTe%YIi#vdHb8VqT6VX1j}k1y zYND-DvpY`BA1Og0?;)Zh?LCu?FR#PMYf zY5v;6-)UxC>d3FMxvsaNM_j3%N*Spop%?qF)B~D@sw*iIdfnYA??o@*NLxijMC^L* zRQeTL)6Bu5bQqrTtSxUp=X7x)4=wZ__qofGk6W47&+7XPT?s6SblnBQyle+(%)}Yk zzWr$Gz4Xu73jANU5swE0m5W35&vDeK{yO^~lu1*&{xpkVQ*ZeTLv2#71^pbN#@f4z>)$GOBN*CI%He!mco%wK-v{wb?(H4ei*XZBF|xBWSr zw3xe;v*E#8jb@9VIm#HKl{)sw>#(0eSWW!+b-+J| zWyd*ibp>PX9~YR|5JZDRAr?8<-3E;gAZI@<5n|kl*Vnjd5=i0n z#}-l)R4V<(M~JeE?cf=4eeqpku{CL5k{AcNd#Ja<)jV=_1;d!qi1gP zaqjsJ*9{fgC5ZEKU6dtXPi9KT1YvPsi6K&f1ho+9nrcUg#h;7nuVqT5?(P^Do{5km zr)1c|xlN}p;iE1SmAhGPws(Fmekv;xiYB>;esnvTZbO6ry%`}Ech`TVZ;-_&jiXjO z#e5MogHK{0q552E<{0qf9Y}PAB;zEltixVZ@DH~{a*L{EhFRYW)lM!3fx6bz`>&sk?{6!KJmwY=zz-Xyd>D9q*kp7jw?Yiw_YoG+za0&^ z)3b))Q(^y92`kuoLr2DXb(`Yb$6^yfSJCH)qi6gFnkJ@g35(_HRw|X6df(O>`urlp zZH$P%j>Mqx<1Kz?x)_tvw@P?o%eGp^Ppka!d8nbi+|I-Oo9_jm?q=ahDQ@4Y#&Q>! z8bT^^Rf`$V&dq)0|HCsqGxEj_LCU-l<5yg%t<2=dMYitvK}G2i?z?5B83s%XA-+3R zAW$D9ic`8!W6PQWtRCuMksj*j7)#Qy{fqI5o;)uv0bw0=df9nL7n3M9UX4<49 zUt>q-BD)2my1DP~_K+fajVg`{)c=8^tFj0xG~qbU@#E}x%41UJn76R7z5}_QymP6v z9zqN^x7IS8!KIZA@ba@!th0Jtxm4Uo=p2skl}IXR-pZmQkBU4AUgDUxF0|XqA+Q?h zIh#fN+$&{ibE7&d6YGi`hHrdsZg6sed!Dmz=OT5{Q9BsLQwT#yE6i6oMg`Fn&rNG-5Iz z^&v?eXC;Hvp086~P3_7)9k>Ku!Hj!+EQt>zEd!2IfkZGq5FNHTXD1vi9IFzwZAh^f zM)>;qRet;vNgOM$4<%{Isafhc%n}Jg^{{^E)L7V# zB9*gn^5pSTj}SGHL^9Mg%q@c{yNuGI&ygVT4@XlH><|*0HI|d+9+>=v{AiZLc8(g0 zk-tXJ_Y~n4#I+-9FM!Ui>J>#uw$T|)T~u?}U{P>{V6m&Btx6nxgecgo=n3GyD68S> z4b$1TlPF*uEW)~VAIzRxFpi9|ndK;ff#lS?M@T<&LU!mGAY)pnH}Dvrp;(nZh_rmD zN$~R9i+QM8kDztFw}FgLvDylgKnUD%&;$T;FT{PKppQfv>QN;WX*)_P=oDiThXAJT z=+w-6_oNBbM4~rBHz3>yvEHE5;m=1&_O%c~XTJw94zVMeaK@i&12^`DRI}#UM~HO6 zA+SGR=|=V$h8U<_4Cs8bnX;UJ`2_(>GGu}|~s zdV}&ihjS^bfBj^MB8h3{^#wSZAKp6(xk68o5XO@gCrJ|jc zD_0)BC5 zMD^ibq{2Eg*#PnzH-PBOtY9Q|mez?s-B)jO8xQV%P~}P86abh|PsFOlf8htxq#g|L zv!2lV&KWNiC+R%WjvzBnA}KjoED17<&3Av}>gI_pJmU+u)rq^0A-bcQ{nx0J^CbOA z6sjzt(<(Wa;cq<;$24VvgM&lcYvYt>k~2doW7CD6SsREHKvmjQM%jg9=$NGy_Y-$B z`?_#au^ckBr0`B#iHAr1bPLb5Z+nT_gkbVIN_`2C=D^;0S5{Sh!8qp+23!=I&`a4= zq)eXcrICru&J{T5Ef5rbPD8bE_HQv!Vb`K;Hpo=c|CFkms7Io ztd-$tj-qJ+HQ%Mo7|EQ9{nBC}Gharjl;}0}lo+8L1?tcQD7A_p%dpFg`xaTJ;M;aJzT0XF{(R z#Tm3c$WP!Tn4YM(%pt1Av&a_G8lkq@WXc{=_f*g-P^NxI@Vo_Mmt1u0Ug4I-$*EJe zjZ$8KPQ^%2d55K?rPnWIgbJkCp%v= z$fHgpR20W)T#gD$UN}M4UdlUyG2RjMzP!Kz1)wO+k)eV*qm~oH-*C%Gp$t(HLZ^qY z17F}Vp1&33AVNVXj?vNTuN(kc5^PJL>T^IB9N8L-ZS8pr#@SMQD)|X=DFCc2scI4T z!AGfZ;>5B0+XHyEd%Jt?e})O?0r|#`49p!|=MjWH{K8Tg3*)#z%}JK?ehm3Mm9LU+ z7Dp0^gqqPa>y3%Z4g{6R9=s7c$^F)Y0Vyx(j_A?|O$Dey0c9G_e#Scox8l3C*K0gL zNO2=`uO3&cp-1)0FTaq!*OIr~>jhK;$2=+8WBN z8KcQadO4+-dMAE?rW{0qxr3@F_o=u3T605hsYz;w&j#OvQ{DGNZsv=A0XL#Ga; zP_s|U843Z6d!keX8G#n}3=SONFz1LaKrRQt$)N>I zB#zwUEIeFle84Gxg^%g(&KGQp`^d;D5 zM&~1(WjpFWV-CF83IoTlh-bhVI6@T8!0AATaEdLx9Ym3;O`k9fn(5b8E8JYDpDrcM z=g*%nF)MK4!iBkZSv`To`=pK#0W!Rd^TRFP4+fulPg1v z*-5sMnk(=9+^WTiB4N@=rxk=rwK(j}K$CA5WHpDmKaORZ0<8PK0(Touyr#-m?1MB<0;xi9ef^8L^nW@%^a03z@FQUP|NNlN z!13S9X^w$>;oa`^+qb`(wYeq4;Y}vGYfzhHUm}VwLbH{4zcH?P2S6@@)Hzl+NtQi= zh{0+TM3S*7EC(30>tI`x9;{{sHSQQ3?Q}KxAnC_RoIlqFp6m_VhPdYvIrt-Q4dtR! z#sGkw2W9DQf-A5qxNj($jZTqajKFFiB^kgzT zcS0Sh!$DsskBfrz&n zg%DEg)_%sB_jo=+-(cuO__m=P&vqny&b;jdfR~sDSRtC!1kFAUVv`ookR2uipu&`z zXpnb#lYoIsejGmrimsWqZrwWb+nG^qRCzq5hLPY{T0<7xXr;7LDd}(Wo5r^~2rHI+ z9K5c9Tv4L~?CH$pN-PO6>QriUdz3MM32Tw|Ahxk@8*z-OaVRDO$A@~{WubGBS(sAA z2qTs*#QbFtkuiG>rxAMFy?wr(+JN(n+X>qHhjUOv z(6+SVhsnnxJz}*EgqZ^1WPNODZ~|?z+24SD;sBZ=hc%}&j;iNJh)TepME55h3nmuV za0A(RYP^g1%-!|+X=>n{8xMzS^T4{*tIv~^JSm&x$p{MW+P%A;FePhO?g?_!y=ph1 zmt;+qr5dl~?Jt%*TMTu`YU+1!7(#tXtx!C($aNWRr&ca8p9LeOfh2^AOZVlKYu3c1 z*l}MDDS!Kg6wac@`a(RA$a}S4Y+HhC8#0{BXn$g(vn%^oVRY2vL>Yt?YW_{0zwd*7 zuiturO3P7$t0|6*(9}(LA9#GmBsxa27K60|S(;ikUu-jgd-DRrqYfZDp)Pd@3*8&f zm6R@_QYj4)Iv#>U;m72nwjdN?zTl0#bFKV@?bBf`BEfwNZf5HG=rJltt}?Zr(nbEj zH_2;jm)Ez*3i;)Nr4#M!Qtuxxo6)5YmIpPrK1VetJq_F!Vx>iLXgwrzI9sN7FBoqL zyF#QDeew+a!^_9A(#=D~l}N{uX#G2odEy@pWLzO(dHg3WJUsB+In<^pvlr_w*Ufqs z{$)@6*hb%QBjjsWxFDQ=NaTva9LcO_KXL1og8XFzuk0P{>J(v6!pk0c5Q zYYx;v2b8)(aP;O$d6s;uWCJcu4MeB zYkv{HD?w<9ameqRa$&{r^0YAP9QCJdC8mT16ic`|zriyt}r5;h5cyeeA$O%sfSp zVdLZrj>xzz?s66Ov{1Y3HWc2*?|**__r%=J{Mb#My}oIJvxy3?od_eBFF%^^K)K}` zTPqoZ*eeG7jr-axw-^6z)gljVz+^0ydpuBJGW4ycRt`9a3re;#KBMLWOri?Z!^xAz zc~=2S@uf%S7_UlRL}Vqn%9L?cKqzzwol3N>uaqQf(1pwe$rVwihuV`oM=~>%@_AP!N87Wv_vMjoiL;5*5nA|C&YQqSOOa9CR~DT$ku zY!Ntd+7zXtLp?QRGstj(hEEqtbsmBb`?+DDjRpw;BL7x0^|E;_G72X1HWr;la!{B^ z1aL&&fF6RT(6sA``%5ZgBwl}K{Z)i_e#;8q@y9tO)v zinMm71lXH^*+~4mky5$Q{Uczgd4Y($J_vH{3RRA9q@aDPd87;tkCzaibfi@?+XFe7 z$h=Gfj9Mc?U!%pB7t%VC^;OW=CMtN%-eue-NB#`CxyTcv>_D8u2BCN!X>qVY_$sK` zw57H^j^lciNL+xkY1&mqX-w)C2CpM8)eaR+;%vp2a4`UZ@exIZxZWhcN3J!=9ulv8 z5cy;Q#G(gq0EX{t8hRw^!ULi47}eFFBy#GrP()_lRRyESNwR{x%R4YuFE_Rf#b-M4 ziQE}3`LFXcSPoPOhXw%WaI@e{RDcXBeEq~Wl2p&{KR%N~2B$|DI!4)ZNWXdW^-(mR zB>ae)?T~=}=Au-IKWYLwtpO-=N7!P{n|80GD2Js|0#BNp*Qj=i-sg=^q78EcG)MJT zpoW$`1(KDyKXEVC;^W{sfTlhjbZfC{yV{a&EL{`Q(R$A~FjNGt&>U z`wvm4L=6p%unMe>K}euwsXK#>zJSDJ6`bLiQJT8KNP+uvkMFVPB}a}37{qF(BOFVR zZn;~kR{EjnJr)>6RP&oiKBr$SCno%3M1cvamcB{1aL|KpjZud zB`jws>f|)OYxja9jQm=*vUOz+h*1lm&{+^Md(q~*8lyzxq_RSsCE>fAVWOz_TX~dW zmOljMu|Kx(_Eo#oVA&`Dk06vook}OYsZ1cLmEjF}twlIWjQ`I9c8(ZqatifwwYayC zhvqQ{=`rnFGLkPt*JlDT=S@OlnA1Zr+J)jw!Y`2c+#4o~6;01VKRReAD&2yghno6S zq&K4kh!>2@7}WqI&te%+xBTJN1g!U!-$sm~zU@t4Uw$}x8y+bm1wr&gj zTows%49>ZjaTbI7S(CxR1W6Il)Vp{ksiPs*=k3GT%OX3Qf+#c%WH((V-W?!7h0M4* zTdaLg;vN9O_GS0?nsGahk1Y^|)B#USkl7YWCOj$@5!xMPXUTz)rv91*YxVD?R-bUy z%WltEY#@{7;>b?sUi}Rz#B&n;zt&({$Qni9!BG*3r?y!(wT(nfL(0+#L-4RS6ZdS~wqu9Qxa22L1NGSl zL9b}*I;Y=i7*C+@Iz9qM%fKuXb=-xS(QVYlZSz3E!l7<69I+{nh6#ni@~DiG@oH&Jsy7iAxKw9jLlq5SXL4WNTNegPNVyK(dHImKj88 z^>Vo<-YwP#;iSQ6t0rtkV;_pBYnD#vh0v+t+wz|#^rGg>xD5-0rog+PJ!Ne-?LTud(E5y_u90A08%}OJhaKZy5xwTv)Y>+ zOjQ`~avNEJ`64vZxQ(hM|0vH*SxcN+@*_j_D=ji`Let{e&{|WP!kz41nhnBK1aRS< z$UrHK0uza%n&CPi)|>kZgn^Y%GlnB&ZEUlJh>bYQx%~~}pL(Ean%ozms*!^`*iGW*Yxn3%QXsa+OUxj-P@ zIY{0TU&f>#4L3Ho@8gHyum>T&ka0E$8-qk!E^1Cv((4F_0QVtsxgyE($CX|miU4}X z;U#gbIOnDCYVMSe27z%%buS8|=*<$@9C^qsl9`=lW0!buXOC ziL^LzJ&M+nqblL82JLG@$qQP*-BuQ37>dJ0ay0M_f3v_NQUxWg(_pLitk8{~vM`GtXDd;>O z!T_H}jJIWvwQ2NS!vQ7@sGqklc8FnJFV3(cpN99#W+hjh#N^~`FP!W|o~ge4EB|hy zZ~`>o?V}bZ*jd`4i>XG5inc?u;LqoxpK9yoL$$^@&8c4lZ=W2R9(HBajm9a+0L@G! z;6uR~zM}_73uX2nI5twe@I#sy0`<9q-tZe@6f(uyhf{(StQJqV=`IHVXsV5o?g+1%iTW*llGBt&~13Sy% z1VFdd3(UH44{FmL+H8l;h}162M4EA?HF#W~9T3~29)_PL(IJ6`oT*(o{B73v?p2@&np}DULHFUSv9{sKWzq(a=qUdq zIo0OK$VVKXH`yTjd)=!mV_&0vY%1T78WngPruKM+JBRVe=RzAaS&@fo-R0LM@B+FB~Xn@ zY$+K_O-FA)T|mp9QA66Q0={aBA&hmDbgUzeSwhjNCI+$-YiN<-#`C(rFE_z-(YmWS zY`PCp!1QRry=om#v}=B*QQ0?^YrsG2@y(D!?gv9Iddyw;=^im_W$T~~^FQ72V@8>P zs%Diae_VcvfNE#EL*IfgX5G9QzDjRdfxY~WU(bg}uXaD}A-K~0vmZx?Lq#WboUz*U zYwf2?yps-CV%rxrk3MM(R*|!AY5;Mk2qcygq##kEwfPM~+FfC)zPA{=u&KOh;z_;= zFsl&+g;pPqxr(iJwP0gSO!Z;zha!L8M6;dS(}66)%!v$+0?S9prhIWo+9*J%B1jQk zryWE51~CK+Pp?afp(MEw+NS$Od=n5J8FP1NfVhLT#(7qz?z-j+TRI+7ixMgoNbAc< z2dmf`isz!EY1noMW_Ol=(|GI}10xO5NE%{|A_ytOh73f)LMYO+aCcZqNp}36i!Q}m zKW(J*0;OH5>-;c;4e<4^L(aWmGs7X2k2NZtzy@vPCsH#4bwr2HPn7?9ya&#dW*m-1 zSicD;mRCYF&fF~-$omI%X?irHS*Bd`~B?6&peKK(H zcbIBkqtzK#zQcfgcaQ6@^XmcU>IeYt0%vYNi{<8Dvf`6o)T1|pxAz2d=XbXGiQJ(s zG~?Y&X%8Y^6SEkF_6I`;>X6K)sR@%4Ma2cQj|}`xoZb)S-QRBhWw;=1ejsWF>!b}p z7P7{<)d2IFkN*P}**^ag{=hB*`!ZeeXCj5T16FK66LqFN43k!NW8=aQZ$=1nemnEd-Pzc@v9DCGKOTl6~iBIG#FbB>!)pQhNHN^r4ShgE3iMq)I`E`*WFSRIf?_oxSv5FVi_yq zy<1MUT%@LYM)PP7K&o2=Wq4*4)+FLjg!lZ|MWA>UI53VMY|#A^9DKkEF-;4m{3&66 zS_q93-;N{y(g-HD(Jt)!B!F#4;J2AyWmdr$s|U;|_C_|9Y_;N5h1V)FeQKnSKnYW& z>S3JEg{VUOiOpsUX|YS#z>ouR+P@fAXH0=tNmEoSe$xmC;Tx>PZttmN!07$p!v`T_ zDm!Drn&V|b)U|~49^Rt#hIGeDZf9a@FyB&O;0BB`VTz>ixav=0JSgr!F4v0q@K<() zh4HYdDL9!WA_D@M%r;GVX}jA%%%B1>h#wY)Fbc7s;5fTT75%iu*{|)py(^2h*uMP< zxk&1}6zjZ$`!3uxDq$0Z(DTY<1Guf&hPFoE2aPNfK#Q;eF>DV{q$=NZirTNoy%{3$ zt7jg6)L_3sw>35%$p;uByo$AC;9G7{bZBU(!%4Fu1A}b19^XY_t&SbN%Uxy7tBS5Z zY|77RhCt0@?5&Gn;3QQ$a}SIa6}-un}O$PpJhth9dsO4N-nS*J{H?D=D;v zGy=_QUl#)v`u$pI_$R%~sp%s5x8&}DtylA^t$FqILcftWke?!)fO@=RvFb_0F}CJj zb3Bg2?s0a)^8pQEiW*Vo1EABc#)KfwNco&1~$46uSV`GcP*`lK7rzRdx_ zjlF+AN7^L+IP?CvRPhbt74qZHyj`xSMyj>1*Z79{*)bh>fx!*kb()UfPfrraVGLH> z&woa-$=d$7M@t!dBV>7*u3PQi-}OxPF@}1-HKJ)^uWIBK*hu_|RkhcNL69omWxP(_ zKm#cfD|p_`-_KS42675fAMmJ*WA&0gJ^t87v*;Ey--1UPgn7R@C{GrGn6^U*t)rNA zruSPTk4@}F&Vznx_bvH3EeKK$6{53XZ+ng8r;ASRbwM5Ht*or9>dHS?!MkscUp5wz zH#med}gdmU5feA`z^W!#*b(fl-Q@ffK*YI^ublsqm|sU zV&KbeTBvt&-wKtW>vCgV-co2yLq^r(3z?zZ-EXM9G@Rjmjgbt46=jn2S+U0tt-wwY&ZY?u=(2g*e8Kd?m3D^%ajLcZEK5`0pzOqaW|Q+ z@%y!;>Q7q`uN9)c!TQ&kk`2Jubkebh;HpP$;^qDvc{8oGpY)>B$5EiNyvWv2JKHPy zjrcDt7@n8M;619(`atwXNh@|@FGVHG6g(f(X@Zkj=t;@%PF1g*o@m1oSV1_l0m(!a z1aCUd`rM!Drd`tXNiPy2jgr|zN(vVQu0aZFxuS^v76aQ@NWIZE5Cb)nRF7iizT$qf zyqEFi32dOD);PM4QPW4n<5y7NB>6?4@c3s%gkKSrt`0nwgbHnwBsNnfCn*ljC^-J6 zL??j2)N;5s`aZ5xnr7~3hNIy^^(u6Q0JCJRw%H7lZSmPMC$}+pAPoea(z1lIfY1}+ zEL$QcQy*zwbnqnD8A@MS<%SHG>-c%&4e|!4`lmW9YBEc`g@Nt#o_l8RV+x0t-c6ef zs&$DUzPk;UBp6$ZMh>ibJw(K8{}?%NDqAYXp^?{^CMta}^4olo*EmHv_E5a?X_#xU z#RBOd^4IAtfYVNFwtf4&l>ut9VlLo(aYa7JZ!GiY%EC`)3H$^gxfvv|-NCZX+Rn1v zkH@WCuV7M(;?YxC@PeA4G5Z+BrafVz&XFFM8M`88jO-##E^=<+!Pz76)RWoi93RyK zA59vh2lYV4TSCIhb=|O*{l+>0Ag8=wCy3rmX_7AA#7Jg`B;j>gG0WGpJzUIXAS8dAF+-j>u84_Bo-&G{$hK&FIqg>280gk`iZ?MrvLihG4nZLY*^KH_< z=^spLVO(k62R{+wVY(-KBcl5`Z@a6f^YXTO($0{aJ<=|Rl6L;%S|QW!G2MuN&0+cA zD6Gs!>r_b&Wj z=Bs`<(;xigU%J!(=ztU%s{wldyK7p{#zx6ZlsNLbYi_j6#IiM_%pLS!=iq-ZMgNWa z#pfnM-S^8g`1Q`0(J~W~j)fLsJ-z$sL5ar6A7mP*7Bfuj$`5|>jUaQoge+zL(YK{+mlaU0H1`m*O3q(z$%r?YCc0;Semr^-M>Dp8$ z_0;WoKTYrccDpJlZ*d0vR)6T!Y^GM9U|J>rQl|s$cWZR!Xd>JbRT|YEb2k>E+3!!$ zap@vZM|pzOPX=B_zz!RyuJt1^y`?KT+i zGYI}7SeOP8fY5Fu;?$#1sk;6&ST5Ul1vk3oRy1XT13$6l1d!$*8XL$CMo2Hg9tKLE zhvas!rh#+P{1ya##gqlml2RcFG zZ~?4-42+I95U4Kc`(ehyCZzN8CNI>zNJ+f3PU@CQk_lrY&2qPp+*vE@_%&4{LF?n^ zLzxhnPlL#Z6;RJ`-Mh@97e zh^s3CbCau#%yk01)dDQn>zXZz2hA+<%_@L6HD9wr-$$=@dkF- ztIN*KTj~^@oky)qz~wIhc2aZDX{@U*1KG^PM+BC<}k z7^wYi0g_{JrfVnx)kT?vEK-W(ks&4X!Pk9Z%E%YoH@w8kJSzh*W(ex1x_8D%0R)+p zs$K0=F+k!a^V`7VeF~0DWdkpNJ=SLtI61m^T#&GAvJM`8fmb4ns0)t!{5DreWLqK8 zv(bh@#u(f)H>C3R*@<1ajwE%E2<@)|v8e~Gm6Hx-{`^(ua*30{NS!sd8H_uG_Di7# zKn(KwzSRtMkOTG;d-;-pq9bJ*erOcvup)S_e3^BSU{br&58=Q@fL3~8Lws#Op9=xr z8r*R~9Wr1!3FCNVmFKER-#O%r2Y(HG4Ur;7nMkk=F5d@yv;lJ-eSnnoHH;vnQX>kW zGVTCM+TMz^S02FLF+g%}bp0aK*!?`M90S$0gX!8#h~x9CieRGp@2gzhE@sG zY3gwam`tKp&u^QVxvO>#UwRP|qorW7s8{0;7fQN6axi)MXZZSV&1mez=p>-W#( zNMKsg_nNVCS62`XO}#dLJuZZ0Tzpf8KfMTGcj-+T?r`B%;vP5s&@qTfXvCO-6QfjZ z4Qf-|fdI5tf+|DFO=r0R0yS;C97bIGjx37jSF~?It=E*w@y_uvQPfPBl+*@OqSNH7=@#dP*XX8x zg{*%8K=ix29BES(j*oSY9|IOUv+y3^+~(Z|Y*oURs4aOFpmWVP41_ykhrWFLR=NmG z`09#b7`B_vn!Te?vX(TeQ%SrrYSJYOg7s6di!=v4*i@Q8LSa(ijU|F()Ep6dArFGt z8bi|DmMvmAT^PXn@e`S-LsF`?+firm%)THjtRoR4lWe`$luWA;hKbG~z+*pAa{;5u zxS_FtS@F=l6?Op}+k0#a19+Y@!9fXW*kfnfkfygN$pZgO1^GQOSl`A9V2v+Vv=19; zx5}EtCTW~$Tk9O`4YKGy!kkEACJhBArC3A9s_c+6HoeLK5Dy-mR%}v?qnS^)6BhsS zfcZ-K@xh{k*szXLu#Trx@;e3ty*&yjy$EclYfv&WB#fe2b|CE}s!PBb&O_?k52t1W z$h`STj|P%9^8;s;6!{&#AhiPUt4?ATn)Pyh#qMy^KeJ$+?(d ztW-O2`EHprU;6HZ?tv;e1Jl(;T#7>k3 zb#i_-wS8zN00A?c+PzTgJKVg>>t?*Z2pb@PiFd8}5>E{$=FB*>n#y|OD0!E(js5R+ z3L8ZZ$FvNub_@ozY~owE2;?&=2PI-z8oM=mwQM$+UDjkVK#YsNJ3a>|lUd0C_cjW- zQ;m2;B(~+~?@d!5@mC+-A$m^I?`nFHM^;_QfZ`ZW_W0|C*^i+;AtD<9h6B;(76fx& zgXJH<%A68=Cx>rloalkuA)G;a6q>ACYZE-*IXkQ~aXVmU0v%v7Q(h+m^pgv* znsJs}lOS#tY;CM00yOZfk9#LoXh%O8cJX;XA}uPWutHK;GbVeeDS)Onx_~j9f~|OV zd=C4zD1wbq4nqLg7v76&SkXU*)MH5~@0m8R*euGK=W8T+*uE78uA@^b>u^S{XwSIO z8Q*fRsatWN3e`NC_I1WDo9oF#lc}wfrm|36fl1lLKrWokaGSRVNtjgme6C7rW`0nCZRW=T< z7dM?r!(~b(vg6Njr%LEfQwsi&N8i=dB7c*(Ymlyn$!kgg%KZVt>xKE=2N}S~1CG%D z54hV5vmsJJk~p9eZ-lWSjuohjoQuKr!ax1*S{yEiTZU~Ed%yrYTKi)|^r}V&Adw0( zxj;SRyMq924ekmoox``e8Ml)#kX7(? zj7g4|b1DH^#VH28zl;Y9f92nNYq|KzLY$x7@OI)#Tu7Fm>c#1rFJX9Q#!2>iFmmE$ z8zwk-k($|*&%W4uXN+8M-~Nog3Z2#ZZ`J`P1J#L801c9FIQxo1k=`g8atR5B0thpxF9>a~VkaZE`V2ZKf20Jd+`0XcX_nhxG5 zq5$%-cHkDsjb8qtKlc#-UDx4Fm{)Z(G?I}F8jO0QctwU(n!Lo+Vw2}n_)Ra8Fxvb!QU)a=QEb^_s>h=s!lMt=yA1^S?B8j6 zr9HW39gZ!PM{E^WcBwSn^wmkZpsf1@kG?gW-)1meU70{^#}E6eyhR%XsZg2bSGNsEOqZ zyV&7quEIYzpb?u#5i!N@npCpK2jdZ%Ofj&j1I}dy4vp(;I|y|8!3mRtbd;`nCSrrM z*B-0+By!A7>Y# z6xR<;9t}}5OTiQ2d?b(sJAJ5#8X&&2d*N7rO;>fyhsr|6OPv~58$6*-6+F+m!#8=I zFXb&hnT6_M(jG{I7?_AGKQt6_E-cC@acl=v2_c2u{?_*twm`g03{?fX*c17&e))~$ zegLc2f0Ic86!obw@8IDN!*ma!Wb!(eohbZ^_|`Sti@$gtfnTee9qWemDXLsCxe)s= zA0%et&>p*PTCR)!f>OXd1McP`a>5-IjhprkBx4Z}$xqQxSP?!;PppOWYFblSpf-3g zq_vLsOlmQEjP>ISDxf$;O`%zppoCDNG{w8LDYunM5SOvYb*zV4P7VFCFsnGC)8)4r z29Hr9YK;aUH{rgs-;$2CJtzh}AgHgU(k1I0Uc)2c6e*Gck9+Ac?i&#uoe!gH4Qjk0 ztid&CK~wXA&-lb6o{4eaz5LoJII7}>aF{Qmd~Y*k?edBJ3dP+<{2eiv3C>jm#Qm&@ zO(g*St8keW1M&EickilS5?kcGZs4RHHSuU6(r zRvo!)n($-=(Bi?8BskQKM3TUUzBpTM*eZZ?s3LcUOhs8uMsQ z(NiO>%8Z^GM>%G>+N&iWG`V?yyFOEyOsx068QaQaomelEb=)C z92Ey3a^%-$pOcMw^Px4Pq#mXA3`kYXE_M#wW@gX7|JaOSqQQ1 z<>lQ()T3?!_-O%Hb}>8K^+sLIJ^RGg%J_I>KEt6#Y5O)p@$!o0>}v(bN|6+eL0xGy zdZgyzdN=B}V7bw20r!TrG2pa6#_9HS8*6kO#8?b|ID2-*rl4gzXL)_GejAXDVgOP} zPD56pB`N;)+i&cy+$HKz;+)hx0_?*9jI#ZiZKo6fd#Bt*PJU7@(2y8lBpQ&DMThw4 ztD**Rkak(MX3ZW{f@%D`?T0zn6P<@`DVJ>IauqBBZJ??)r|ZYY#xTZU0Q}Yyh6xCVI89}H%n$#SppGhj}*=9x8ucLI!1c=PE3z~SUn6)y{ z^rGC{+)f2go-(Bsq>fTdMjvX)gg4ninp->Jh+6Pf;T8+-8@KJ?pl ze!Br*A3j|VJH7wl!Glrp1E|0g+LY`%?qXJ#C8Jd2r;SF4m1tNxSHQL2OY+E_1FAM_ z&p*0BMY0dNVDH-i60S;zaGBT~0h8j8?>y$M@`*CMG-td@K|!Gvx;DyvN$*Vv*Wa{J zWoFc*T82VP@XD)G*bk*AGhy*pEnT|wL055cvACXS5)?>Rq|^kyIt5U#9Zb3hwIkd=wJ2@Hki@u&Ma}y>!?2QAq?tbX3&lOO4GAz`;#zzfMj=0Qj z)#lpBV1`cOAD?lNho)60*;MhwTi6%Bm*y%x62a(#Fex|LqAsYZEllB1=l-x=z9;p> z?hI5Vo9Xxmg+Z-HL^t)sS=szc%XZG=;kndg!p+ute*<=NCiJc`1(-MOf_lW>FZC*& zK)jqIaNga$7zMK>HZOH7HwRK9gSi6Pi#ih_qMoK3ccds|znqRD8~yQ0Z1mn=itF!J z%3rz#?T`|f6}ZU~riL0R>Q%1QN;q;O*IPi!C$6nF4+0%?XZ}!kVE~_SZtU&r%_t>Z zWWV4||G4Q#4S<^`p6+=%!=)FioURA4RJqdT^5${qLgO2Y<&kZMIm4R*mEmY?{xx_x zXtv#beM;s32yB=1xPY=D#h0R?FX9GP?G2HS>JV0hl|FZNxO8}p zqF%l2KONNT%T=@89{%v-kHQYpfW_aW#}ypgzjyCN*iOFAQwI(lNRM+|ozmtEi*U82 ztD&~p7LU8=az$`ENNSfyM_?A3>Y!hAIoy$Dce<_n(NmX6N7f53nznGlgV6%sj;C*T z$}g~vt~+zW@Y&V+V|N=iu_{Zl=%9YAleHY_&@37#y7zVo3U0=N8YNHA{IQY3_`Wx{*%acH2yqy?giG z$i3_sTr--eRFEii3tWO#V;#`s@Wp`Tk9{*Vp7Gn5<g6kPTz~$?Jz1dZ+0{?(zx>`7-(2`ai+}tF!7r<=9`4kc#j@kC zwt6AI|_w3eFcx34{?dI;Nk-L#!k zwDL)&io@wr#ZhS=&M*r6NKJS~}K`=Nq7}SsVrqcIw`Jx-s=Zgr*$z@MMkRTTn*S5lD z3|x<&U~zV-Ru*{<+|iB|!8v_aE+16FhOSepeGEAKkBiA`khw`Xuto${>va96vls8F zw1K$rm6kuOS*93klwAz?2i4%WfKG^wrD+vR%q13ogMcY@my2{iOM zi1lhYIXS7pTNF2#%Vl)m|0DG>9U0_(pD;y4<>BeMr$<84^!ddb!!Kr?dua$y-)NSz zW2Qd*)>S(O6rb2ZlQ{;SW%wJdn-5)+ApsGAB~;?1#JZmu`OPK4Zih_#C!%`un`3ey z*?lt^E1cC~j{VJgk$D`Vu&Ei?)?8cSo7K3_1CWTue$9BP4c&kLI6IJWw%ff?ek+^u z^`U{Z6Th9N6D+(9KCse4QT;v7oXs^t6=n_%4hn6`I{BtVT;_b~n=*I`rM>^`8EwUZ z55dFhyRNRe_G%DBGo0vHKEltrpZdQ9;Y@45JZ(5wu1k(+eR5|MTy^|2|Jo zZWRcS8gPy03Uq>y6%}+8QJ*%%*VF$=4ExTv{9*e164bj~eKiQdEpgP5%+sp}{|3RY z&guI3B5HdPk(La*M|PueXe*fcN++Bk9RSG*XOB1tNovpJ6JPnD3#7kRqBwRXPMS0+ z!f~Ku+6A;2&<4R>>6yej-P9126bWq#5;c&`1buat%g3$H($!1dTow+ZakCu8zWPBI z_%J4LAD;we9L}9p9=U(#rAwEl$?{7WUx}&MGA{QWIo|2fj-VmkMeKB>NH5DN4Llnu zsldWhxu~>}fG46Rmrrq#YS^yjn`*(e(b4aOYG9OJM)C!il~;)fp|(AE@F0@e;P;QE zvQ&&mK6?hlDt98XQ$8WRVub=$xjJZ8Nhf~WhN1#x1kgM@JldHKNwskC)gx7RfMc}h zggJaMJp9-#U4zrEtkuzg5G~#=D7XzydO6CTDB@csEv<{U!#`m|U8XGE#sp%Ra?idR z1w&2^bbd9t^TENvR7$n<^Ugkp@=QlNTv7=Jh|`h z@BV*(kKg^6+br@~-q&@V^E$8B>zwPo!?w#$1~X?HM|al>E+LGoW-|MwlCgAv4V}l- zsd;Y;wtEBlE|vD6rpjcvOMmwV9#>13aBUK^R+rtf%t_nY>wgZ~fB!>Yt!qu|jr{Rt z_ZVrs$5LmNXGkRtA0^aTbGLr};?RB4&gcT1sP;yFd=ZG5)wg}#;NVPy7%e zF0xt6H3S3xw_W_lyX=bUN@bZS%gsX$WXS`^gBy?n+pWZ6WI)B{bpOZTkQmJs)UDBDci`gY< zK&6h`HNd)3a8P#pVbs55q49ZD4jhQ^Q?Mj+XCCR9n)Lnj&%gXq!bJ>*wC`o!*|F`f zJ$v$Q{XafbdFP*sYxnQp-&|Nb`X7!{(ohcjk?~AFt-p{18L*XhF`AU-QR9;M$i^ep0BUSEqn6%Qe#Du z_xRt+*RCl3>xl=23GW~L=SBGYi(A}Il8)V4@=Xba38O=_8VrgE?ekx)iMw+9LkdCugfYYD3XapD)seOT>fJEoIw_~iFrCDl=1rfiSi_Q?__`EN|T z^dMM#`%<-_k**8hybQRd)zG2w57hQOG46+3)zOVKYerPzXDX&P0lNodaN!a_f>BvLSwkpk=f+nio*xF%$oj@LXzHH?xOS?|WmUz~t!2-&+w>f>5Z>uzobkwITV;BV zw#mcx#|gWfXAhtMN;kWzN5QP~RrDJKOQ~Br6CQ7?9 z#xnhoJ8cK^hJE_!#TN(&P0%e=tpyEFSUGh)m7 zrQD!6)4lic^paz$A{oF5DW~`$ z)Fv64H=CkV+d6MdS$>pm!OxKqdLcQMI^N@*4J2F?CHsWn;l4~JI{o=IwPX2R8M(;p zaXbAqen3t0mWSLuOHt#m`fFAbrH2Qj*MzJ%v^nR-bb6%s?W}v#t=AZv$-s-$)Koz8 zB2xJh?v;{?gdEegOg`=5)U!UkA$D#<76`JD#J!ECr zY+5H%L`$J%YcMe8zZfG z^7Zuwi+( z)tMWG4KS#O8OOCOdmj6IymjkS)Ldl+F{=)B8*a#+TgqkO%<8GDe!gIUy?0JQoVdL)KpLMY8 zX!!CWtb^@1Z-qZ4=qymO&az{qpYXMLT~<>j1@Z+w%c{vZB!`U9jSKjA@}@mTtW)|I zqxo)}L-8>OFui*o8>?gabltF)8^d#H%79}wX+7tur4FP_m7dz{V+$!v1r}4RAH3C@ zCt^#BF4x2?7flPoJn0vs9ox4j)=#ijDY!UKrKT)*8Gq_%NXYQ0n_RM}PXeWMnBAuP zh{-TRI=SX+*RJKY^+)Ay)pYn6b?@Z}iyAImp21}4G@6V8U%YscxV;NCv)<N>?mtJV~!PUWF2rbO2&i{TLVp{7_1B0UK@K$crQ>?BLJ6upRgQ1~~DC|^?>LnBK%rSH+B#*jd|{^V7~Fzea|_B4RS}?Xp%ZQ?fun1KUqFG z{wVrguu+sD@`jE21T|8@TqmapDsNv3{|f=?x=+g-$c9?S)&v@;i4MnBluS$1!s8Wj(Au?#Oso`Zc#_B&YzW%Vx!Rrrg_XmEIg`!6y(JTWRio8Ar zJd@WCeJE9)K4>RYorB6L9klS=E_!HxmkJxGuVAC-Y%UPAT8fsVo$KruvIF|5$HeUs z5~U=B!tXsyT*mUB-I(!&F{Z9v(C*05@7Mvhk*h1ax*`k8b0CMwdOj|5cx)QF;Zq^c z>v@#zn}ZM~7YSRV{x!z=&i1+`@D%dyFg~(Pb*%FO1TPYgOy|kgu#M#E44pR2B>{SZ z0=$ng(qQgRF5vtDY^EXuUtsML9z2-YcHdO~Tsl3IoqP5?V|l58me&w)S_#ipVX`+x z&0z|=aK7C*?{_(&myiTCHE-?CFSgWmFwtBRaMYnZFA$m0q{xq5=Nhqk*wE$ICd148 z-M065vs*ceCEQrKgHEyYjh(eN(sc>dV|i~Q_?t9I-1t-~L`v>}sB=2K=! zbDDC`&9_qcc{c^#KoJCDplJ-dv^OyplSrM zj?hgfDa)od@1?XW_hCQ`<7@rVVl4n8Jf5(XU?J@tv{lJI;+48qtoA$6XM)R3#!LDJ z1Ozx)x$uaET_B=9UOfghoY%Xz`-3a@xRmoMZif;U&D6{KOF9ZZFmGj~?FD$XfJicv z2@VJL*b+uh+P}Elq;bc%2)y3%YFd{e%SZfnbo{t+D}e;rbkW}2R%>ndFNc$lH+%{W zx5%D9!KHd3LabIvnUd1}>5T!clhh~2hp#R8QcKt}=Mp8H7^|Y^*GIhCo7PZMdP$-P zU%T;S9X(Bck<-TjaQ)jO|29&Pa)@6{wAHG4pR*Voce*ZsZl}S+;f7 ztoQGAE(TL8HgSlg3RgXPhbn6pjy!9%p(41G9CDenctLb39C0CY@tXiITw1*WHW4QbEHD5U-ms!<^oO5s_ z+0uohsJ)}!zRSRcllk0iSH#Lmqt7}Q{3Az8ru2PTjesBx{q&!WRZMV6y~<&V`+9s5OYoh zA4N1h9QQ&skPT2@oxWh{`Dray4CxA#;a5_EQS4452H9sLYg+H5w9i{H!zw>`m9AA@ z#K_;~eS`2*U+JOp($23<$dW(>cd?v<2U~4m=^C@h17pPsoHFW^Zx8LozkAmx>6D!3 zk-RWprmnvJGB=PaI&ZG|bpO8hXly@|VwiFwEXBj0D73h8`b^(F)_YQ^9_8ceAB+tE z%SVI?#%3NNgMd)nNmoC7u(9iQLa-6N^nXJuMS2|?x#&spNos}5a%jnYgQ&~o zBSqX3hYW;gTu_2&@f`B!H`H~X)$k(R54(!|tLU7o+lTzA?>Tb)UQZn>7uicnd877N z-z*hx3sY@*AXgK&Gr6eTf`Xz&d;*oB+z!Fzp#!ID*YT}JyMEg)h``{B=XO@wg&8~L ziI2f~ts86ktEO|0fWSw6k0AuRfHxO!cZ=|@ZG33sI=1pv@&2S3P3W&F&tKJ1&hx0p z+$>{X_i@9IqMCrWP*~t==eHR^(P-Pgo<*s*K+(B<`@R&Inhp!)o z=;Ux8SQ?MZp4&O6_1r$I7MJxp{!bf%?XuFWV6{aas^x1K- zX|uK_MTI|+W`%xu;xlP;)@w1Q;LW)JdhMM3n)TAZq0Ys{ZmQVV&`srPb+X9=lX!C+J>K05^Lmrp;@sG}%zG zx_SN^+tcgcP{S>hB8}$$ADcu^WB!>D4+vdRs^Tab1gGu4D*JW(n6cVpm0nmk0&ElK z-tHP}n#koX2>+!}pSQs#swH%_r*j!aU;Uzy>mS@a)}IeGIqTT+=0=K0gZFrwE`#Rh z$bY@--5En2yaHO+si5nzAO2}YppSerdCW^Y>X^1(N$Ma%{vsi<&|VJ~gD9soo{6cD zA=GI72Mkcl@7`_mvzvhc6vxL^Zbn(BishAG*0>~%-NXl+u3Z^CrFh0u8N~7~{ZhNe z=YNhj71}n<&Dj!+p#AoY#=b4K=*sawh{P9$lxW7r#zJdJsXEy3%F#^orS|qWXCE8r zE1?EZZot2)mhiQqCrsCk533dCZDp_&8LB11-@y+jwZDDuWeNdZRsTPAMfrv%N|VnY zZhu?Vt>@UZo=a+7Vr--GO{2a-w0rie&X{|Q!AF`KGm;3F%jV8KOh8gr9c*;FdT6tI z{I9U=M%V7$&Dy5mmROAEMse)MZ#5B76f3)KQfly7fRLON@l~eT_ZqNgnEu-Bm4ET1 zHZ)?v#P+wzVnC8(8Vh=>6JL(Dg~7<8*iPJO6P_D9<*%0dw-<$Ls1Gpz*hyo405nPX z7j(edAg|n+hvJOCZ8NsAH?VTdbG9UQAMYF$_z0fzwf#7GdHZo;>;CnVej3H;5)x(Z zhZ*eJs|=SYh6t`S|HwX_Guvq1*%Ke?u%9N?_rYH9#|jK(edk1Nue-qvGhNxC#42rf zl}=n?_53PTjsOm8O6B&kveGvRMBCLmb3wU7RF{q&`;*70 z55=>WRC{}A^Bu$yh`Rrvl(c^2cmqql%@DJ1-(JE-H_Jd_WntVYfAK00r0+FMFY?CH zcNG=Xv+&bR+S%_P%w1as2zRbB#qzcgF2m#A%Y|$n>}FS>d_??P2pgmb&S^nObhKvL zj2W|UFY;*i^UwZ)>bn2D)eKHxvHi$}Ye?V16zy15K$!IEFqy5g!jQrZnd>rzTM2`- zD#zuz89ysFnY8muL|qQG%T3FiQQDSIN5uVc%j<&r9_#DNGcL^OW(cRy_lVc-b)57R z{AOD}yk<``d57DOADpq-~f5@3D3>3|@KUr)#B7DA%4kb*dn&ZF=OHq|l&MoYZQqEoMd2^>fcyB5`$2&E+hd1)bNr|S+9jdg3=Gb68F&fdH}Vh@^=g-d3$n}%RasBQgd0#WH$Y~|6$EI zz!C(`B^Ix)ZKU5#|DIr;F;KJw_V0qK%cbEheH$7Y27~v}GPCe<8#{b}tW{}RDuN^h z9rAI{=`&_{_S4cA!qLX&8cb0LuHFCuo@UEKon#U0l(j!eXw|&6(_u_cg+DQ@gs?H+ zX14N>)0jS;}OUwl;sKcuE|=9HV|PaeB36ml@RG%qdtgORBn1=!Zrt-O3TFg zE(V#3is9k8R^_*f2><$Ta$fA~1NqQmF?{D^D#2VPS~1OyC}9muO1vS-2Zy{rnNS3pdh@Hfe| z9)RkesvN~89UJP?CRSMe(R9!%MZ&U#O+AMXDUnR8ya z)~mVhlwycB0~po9uTR$PNom%mhd&f+$BFElF3{T)x9YC0u2rXU%d)sao>w{j5=kZc z3%ht_ltZxp)7#&fP3c>~^0@BVx6guCVhWLmVaS@HoIvRHoH7T{hh2IXAkMyfzaQ~h z6>JEjG);W>y!B0peAV375mir-e}k~vE)lk{@lhC9#yXR7Zh|lLK7HdbEr6S%(4}tU zSSOttITfBgVQ+HNrcKjVd-UNoy>zS;oufXyCq5=le+3HmZX~UT#C31|oq8dR>)1la zaMmtEk}=xx3RuztM~^rOxb8kk&U=$4O%%qhtOi_hp!h}%dAfR$j^C7GM&}%f959E4 z)^;pktzW6Z6Ss-IP5R-)`-AkBQtOB~fv^Idl5?@)^XG?eY!fJHnwDbh1#cS=!?oAl zRf0BH8*UlZ7`2^w#4S4yTz4xb{=5w%Q+7+L0@#3KO=o@1E3Khq&?Fsk+3^wrcAcJF z;f~?qf?ib`4`jN?mSSDOAo46`TX(s9Pdz(8tzR{{ zY%)$oW!p&{!e)d&ahqUbNZjcN5eCY&pbAtRGm}&sNihO=f$hg1u)`?YX%Ph~OrzL1 zWcr)lhj?(R6-Elr@y<7dAHe(jAn#2cv=f{dF%T!ipv4yt^Y!{_I#}wJA1S8Hg#XAA zs$W^l^i)_wlCXJ4>oAoCaN@k|vVskRIHi}#KeK1wR$uUVjI)8ZGrI(JBfmP~PF+Lhyl#_&s9^ z-R;+R^y-}3Z71xvKpzs*-~RT)i%%?Xo}Zn)%rIyvfl!PGf9~F$UMo9g=3$3uZZ~U8 zixjk?Ot6zYRADHVice=~IjQHQi{~$()k!cu3yy7RUU5j%9%M zt_-brg(5k!T9wZVK)XSJSg~m{ta*^8m7dRou$^CBP}`Rx;f(AS5V$t>4l?7(B}UDV zY)_|7gmYX1qnC~y^7l_0VbDowv%fQ|6zlACS!6*_h^xq=!%E_(jrBVl!ub^&e^91p zzppRkIEo7m1-I%f-aLgo&;y?kx-u}F#ZZ-m@QF@)q^QBJ&zAyO(3<`xpJHpa?LK%k zxdZpKwlil5@6E>_AMC)$BAkFTq||;Ee*DF6EncPbaKWKgxmTw|7r zR%5y*`_%Baut*Tvmtm2lIyj%)-m!Ns#JS)#$rxCuqK~L*TRmIC)#rhnahp}H>XH_A zeyv#sl6&}GPuKh-ja!-n2hJkb9Y6MM%SXx$Db$}``+fHckL({Z(VR?7HR7@anmzvD zG+v&#gbY>^`s~eiiZ0NU{^&15^9P%BCQ&S2*>v*B@kb`{o_YMWUvkNS0cf&LQU4 zQe`fPC&?*ABvbj!1q7zTAr!W`(rw7{lss+i;N9=yE0j54>C8%93qO?p<-J%4M2cW; z_q7PPCf!Y#Ba4!}x@~^dz$jZk*uOy_rpjl}DO@z+Merva?YwGERKB&=o^SjqF<2SI z4~cgvN>tv#^f^9+JBo)dqQIQl+bfq5X`{(&(!m1*?W3BodbIl0>a5q~T^EL*YHxkF zq$(8E)(o~4Dp<{G4EBxA;k{slZb?~2`nNm$g*%^9A(MbvJ|^G8wtQhfx^T25k%sna z2R)NpK?r&$_U(@=6s>$4>gqHJxC*JGMf_B*2a&$m>F_Q?Q?`_V$st`ITLmUC_xGw? zFSlmi8mJJF2+VQi7QUuF{Ig&!AAXbdu$ykLa!E-zDLmoXNbPn1~nlxzw zgW=>qUp5TJVKO*8S6~Z;14Z&|LR1D6LQ+-Qjjjvw&ij0T#n@?T@4)jz4}QKnGvE07 zK;&O$_#HG&J}XtQ+L&y!-AI2SD4`a_BUx_c?ub4gtE$qO87uurMjV?HWw+pLv3(}} zVF!`LdrHYtDWqbBj)2TbnH{lAy$agZKjaGSTaCfM+yew1xRck?k(8iMg#BpU_0f@z)7=IpvG*=WAYoDdcz_SJk8#X9vUh ziRFn_F|`ZPvp&_-EM?5{=4})jN_nT2%dNeoXiO)uG&qi6r`;hYUxv<`)-0>=*1srt zT^O*K^kj>s&T7-v``h)IxSAW{=BEBL9O};Pw_3h=AT^lsM=V2m7S>nY78?v!iEZ@- zLxBMiJL#`YuWscaoGC_cc+^=6+TQMOPe9p)&8cvbI5>v~fbnuCX zP?p$I6vhoNEl6Y_36Z@-TuLr1H<1p@7%a^2xt{({Kj~RF-+>CO++)gc#Iv)m-<0w8_2-Bl(xxOZL0UJ(wyZEl0h)LwyHwlaBilgp zp7}@Vy!6@rv4y<(5o6VEJl!q2`;sx>%!IC3DD)R?4r+0G)1}RX{OCk=5`Q!=1*sWsq`mMF} zam|8~TD)Y*lIQ~;bPsJ48BZFc8*{zGLN;G%%a$$EsIr8ob{>DW^U#H`|1pEEJGgLy z6~?*s9o1$_Yh}^>+Eu^AK+sK3I;}O+;8nkBd_J=tooavkkH!#XZ#Qoz6gp+s^y$+z z$Bi48Q_+JHSr;`Kvq1(CsHy>*!5X(1>LLB>R!#=%3%9t2i01}wMr?fgl@=eV4<#g@ zSH~kJyZHiT<>}u>5i;>ytdPCF=<|5wZiwMDS};}PhCQ<$k|IlN_j}E;pELbG_p27R zSxT31>-<-LB?tbiiPEg)8!25yG-3hfU(zbrsPA_CL2v@H#in5|JeN=ULef&b5Un2T zQ06Hzo!PvJ>*HG=AF@K>*;C18W?f$WV3MTr?A3=fVmXhOEFq~{cx|H3-eKL zLU4yZTrj6uM29IS{hKuX)r=U|VoLE|biPy5C;UREQDEq%bag<>90SX1dc3)?{g|q5 zb#8M-HUa#<#9aTUe{3D_Sw}Vjb_#<>&9d6y&23~x`}Y^+#;#j0Ncd(ht}af(*uUo8;Q~4I5_Qzh4YAWIEwlO>;M|4(0j@`Vg;8 z{PJNcb>&0L=WY%Zu%6m=lx5YlT$mwDD_&*4rtV0m2W~oILPUsiO%-Y_R7ql_8IM+s zRQ{d!B_n32YkuqQyz-y7zYf@g;9djwmpy)dZW{u`47~}+kD=HwbS+JvjGbY3@p zFk~LOAU0)_HqOL7cJLu_bmCg)d24spU7)EavCDJR12`a(FFujrQ5znaQT;%h&#KzN z?*OY!_lj(@A{xk%d3os^@Sh=hRrSK1LYz_!$1$S8%|r5D6hRQrLW%u5?rTjEjTIX} z{>L8T24Xt9BxHSEL(Ai7-6pyZE?3>A&aKDo2ZM?*vKgT7+%!q(Xt(UR7hiII@}%$4 zLi-VFG;-7*^m^SJ+lbU)`ntbXxcyQj9Xq%^q)==({=+U|P^)^-G_(YIj?v=hVY5z0 zg(@G)B5s>Od`cM?|FthrbWU9G=F*HnKR7J$g@UnD1XmKCmm0kP`8K)3|MTrFx%DVY zm)Ti(Ju-+hRC}~bD@(_}6AHJDh0uM>Tdkp)WfO+vqDn8v-yWlNpLKbsDX+_w5CU%$ zF;^cHrV4j+d*#94$2>^tHPt}m(#wnd_A->Y&6_9Qt1$K-RTU!@#NR@Wn~*wE9f$(S z9TlQZYkiz4`YliQG8;fh-5lLD#0Spz`x9>qt`a1bBEm=&7I7>iXRz*OWqqgColQvq-y3{(H2YSy2I5f<>`Xcu`}k>C-Wn4F{PS>*^{e zhgE7+%N~wvB%tPQAsov~)BHc$ak`en_|Qhbo+zSWxKGGL!;QS(H3|?DpgPV2l?-@| zQ5p}#{dU=iehLgw7#poRc<9Y<_5}=WrCf!`6WXenhg?ubD7w z;u(|R2m2FVwvl|4-?w~<1-Co@$(_j>`athnXpuQq{wF0 zK9pY+4?EL-2+G4kkLbiFsGTl+ig^G=Wy7Omjbl6DgD(Sh58Nu+*{J zPdz1!9J;4(Kx;7B&kvZ5#wei{9tN}DsD}ElU%p@k^CK#XE61EN=)FTjuX=r;PT2IB zGo{pjc4Nb#9EXXxxv6j#-O&D=^s*{q{p;|SXge4l#l|B9|KUGiq*~HmWIFE~lHGk3 zjo3(aUAZuKy=<*aER7f+`sV12<1hj#vM9ik{(5O=QZUdI_myleD0L{G zQGk`cbBsH87q9{u2qH_Tu2>o)rG41r*1jqn8Lq|{E=B-S{hVbRBLUG=-^2Ayo99!$ zV7>qgHj4Io*S<4UdfyRnhks|Nj~3P?E?CeK)}w z)dRdB@K0>#ZD1c?a1u z!c;-^7u|P8ABKtkxABivV8+QQ=c~}BSTE?jAXgBg38j}LUBddD{*3D1xAXv}QjV!w zzS%yv%5wksi1etc$Ssx(%$A2D9uF*RjUV?H+IXvllw||!XT3&EF8Jvhk6Q+RP*^8< zMx&1M8+f9r%);cYxL2nb{!GWh)~_f$$5O2Du)UG#kz3rnEI#tWX7bsK%&B`$yxzzu zR5IyHv04)UWFyV965x6}zwRI|mCv8=tMokS&nFdUXvv-M4pS=rr@FP6zaI}?y`(+z z3(;^$e(g_`v@g*9k@;y$xnH`xwyMg*UsWUe-!Q18piaVSJpofsrNX#URE`f!(n*r^sh%<_ok=H z;9a*3TJE*Kl_pyv5?xcsGu##Ffw6iY-+xlo}0@UsKd&4wthoz ztqhj=KfD-p<}w*nv1XOkeihXaU{bt$pAfN4vm~YydOQ9TMbTGkD<QOoXi^= z=GgsK=T(8L@ar%DBShl~`}M2!taEpSR_{%21Hj565nq80E0G2|l0gfa9Bf(*J2^PP zGjI74ZT^tr*t#ou=3(#f`vL+^2!f(II(6vwgAHGXtxH@=XTS|Wukn+=+@sb-E%Hz) zd_&x=EN@^adH^)FtTfa3e*-%G+oWV=S7j0!;a!GOs1jEvI0CRTU6E!q?Rn2Bx@S|9 zlNV6x3T1N3I^Xl^HqvqIwdF$|dFzGKSoizhIRB@a1!`)yOHX&0*hdiW&3(y8q+&L7 z>pfwms8RmY9F3WkGAg$i!p=-2E>PYIAHAM!rB49(Z(wE?fNIJYw3OpulA;>j5a$5?P<=Yc4215G4gKcEo2#L`6Ba=@BpntrIy3)k_8hl z<~jR{*DWc9E<)Y*7-h0w$eN%`^H0HZ3^=SS2865#jc@I7p{wMZwB&$|_x|~3G*%(v z+{2!&9LMw=wDwN4`usBi`0|Y4^wVXbumlHj5L>oxJ-XBuD?b<~c=`1y#Ew+3iU&cN zzJfq5QJ;&YY#HXk<5(ZGo?5@ZS&mG&@o#^QrI?qCuG)~oMr!0E%q~YHfTe7GS=s6% zM!uWqmp?y`9lO1E`}RlQeTZC?9kME%!v6dRbfm@nKM-s5tUocI;VYNHSpcgpo6kg3 z>-W$yJ#2V&bs#K@xp+{X$Hof0!z1LR?QCM=FQa*&nB8x+}dSRN2G;SMRu59iqV$K0ESYuki3QIhRUwA z+PR9V+8XXQQydjKdJG$5C9bH?J$0@pZuACg9)a|h2>VyCXRMr5rD$}*cTcbTv23S{ zzgk*rM(`7a@-fy?w?f|X( z@9yYhNv0k|GUU{1*YCBijzxc=j?RhUYM^@b=U17pSv?Un5`Wq7N=%|+7DuBK2TswG zH*9M{qey?ORWK}sYEC!n_x9FLV_jpV2*6+>T{99G$nRoI=Vyv1eXnyDllz34jKxJ` zGoWR!{3)YMXI%&Gmh$sVMc9vz!sE-DybsRk!8hs7l{r8zRAxy>s_x(MLC^XDeW(qk z>GLe0vvn)q@07!0DwPRXP!pS2@fK(MkeUPrA(sN0Q(5jLqbJfy49^&VrMAdX%lIv* zU=p8Fyf%P0PoH^b$FCqvL>Pq-UyL?G!U%Yze}3K7cLlyvo3(-0qgTC1>2s3R^~Z10 zxCJecKz!QD(HuVE{=3gOA`xrF+)1cK9jq+hQFV@f&RV~SvfuMQjkfmMfyL<59_t|= z>e(3!u=wLBer^TL4dMHv%a7ePXGSznruMLO0MLn$O?XU9D8~jO9`}wJuhZOmrT_EK zKcYlPM-BUV{^&gu#Vr!B@Z|}s?Z#Up@r^Rn;9zV!C8bv?*d3%E)-PJ_OdIh{tkGSD ztzItEqp+2wb(yHU8sU$-DbGqn4V(^K5$2`Eh&*oL=U`a&Zw6kw0Qs0=x5zH_l)t&{ z`u8gp(u~0HuwoqaGE**bY*}CnUzLsx!s$yLT*<9p59i*TP31gY24jdFffCGf+}6uE z{dR^%yevbWupksD{P}J|c9T5J{aV%}pSKEWRjqlqR2peRdxhrB~M^5e?exel&h_9HQcL_Eiw5IW)T!Gnq)DNxn!H)VY>#!5#q zvd>bmARQK8E~y5ujx>N{@yIPZ;<(2A{)7oA{n`5n>Lvs71g+`SHu8fg!rTH z4rMfo5nJ#?=iIm7Xcg8=@EmVFbTti5bdatT`#*b*y#dooPp&jKVEnp5g3Gg=-$3)J zDq43{MPooXa)Q?FJOjcxS93OITShrZ?*mQN?|(*Xds#-Gz$O}jq;r@7nB;rLLwUZw zYL@U;QhG#y&?AdoDW~veaioTMHb?qm+09}L!7wBB3Y<|P!iRo|!TZ;tzqIQ55U)Gl6*zA2 zU^b-!?U!#C(rhn;)$IReKac5DZKfX;Kjs!+s#w0r#A<9nAh5%9YV`S$z+t>Rg z7hwW_#O_B9WB1dvxT~=P_o&JIX(#RxUs>9qaL5Z)gR`Zw{<=pi->B8Y?7AJN zG?tzc5O|VR3*T_F)G5;Bp?X?YYGokoj)NCie8xcIRaP%)U68zi z<=QEPm<+xXHC^&}y3l|r^LAnwvHSfLzL(R_19AT|xarvWEw;OX+X6;FHi68sD`|_p z#zy!2yPuVm64>Z6`BGvdXqN(>`g!oEQEBWx41{seA$Ig@b1Bng)ZRuZY{kuuGeiyZ zcLO=ln%y`Np;Y`P9B~wvm<`t7)Xqo63c@q%+;6+3v9l_aHWCh#1E=+5cIk$7 zoQvdHrPC@Seh{*~)U~2xF?OwubmMyL?0b4jb6m_ApA{tcQp}80!DqPrbFv!rpG9bUA%5ZMg_q%GAd2; z^143gRjAWl?~;1lsb5Jv2_-iz!am8$tp#lv{%OHeUTA(jqAJ}u3={j?zR+<`1owpE^+GOE(=Qr!Su+(aSZMn+ zH}4Fp&=A&g9>u19WfjC`CeGP8qAuxGL=Vuk>UUKfZ`37P-ORQ`uD-k?^dg_&B4(0@ zJPES3AxAi^`tS^ua@$Bb%3Q_hC?~OVluNYq9ARl#%TF(%Do?iuA(VGPrNa`#TroGM zB3osJDI~oJA&q0lA=>WM-el-OgbyZ&1yZw!+Mn=kXZ!FOA|Z>i0AL_wp9;IJ+Gw+& z)??=J_lM?T2CI0u+ZYxo9$ebdlkJVxljAW-;Br5jU79JC6Yp(~xJ1L8=OouVaNnn{ zZn1)n?=sl1aUn3@Fy>@{`-AnjYQ^(3`Uf2TLkF#8)n`0uVN_hgb@Dy1f=NfA%mC0z zJkgN?BwLzP>F@J;aJZxnOq(HZ304iHTcoI>$&Bv#pP_*~08#|{VCf}CTvJ%@Ne*Q` zEeTHg7nSR^*c_pPdRSWne8%b9-2|^tvh!UQ16EW6srkhB&cjgyXiKNOJi=wcJt3@R z5HEX*jeczGrF;?`MFd(~7VcVUZY#RUV^aQ^FmZS21y%Uv@{Cu7g&rivrj69PdxUoh z2x}ZIu3%|X214YG$VQq1BYvrry^E^lEgHJ+%N*X0IK99%wBYhjfm zk;kebUFBSv!~I_lxz0S5u+7icAPy}^X(xl1#bKrpq$`4Y!|BDHrd$$zy;Y-1)cJtQ z6jykgYPN1ER#nrDk*ZY{a>D5e4`z?}58iIwm%u+czRGF5v`vW_PbLL{W3_T^flvG? z8FxhANjOeL_FJrKNSE-?FJj}Ty+$?ROtoq2E6XTCI2S3h1 zMoFJz-<;Bp8abnP$ByUel96+qxUZ>F-r|ly2_TGy(SM~-R0cDr%nV9LY}4PQz-KQj z{Oz+}zh|K$d>SWbX_|Ka7pbV8QYlEt1K&~W-x%t@OS^E0(pKh&Z+sflXb0aSH8bJO z3=@iA2IW~*DD@8t4}{w}lVs%NQ4r5XYCV2zd;2F(Be$b5D2$iw*Hw?9VG{9r4i4_c z*e#YY#zizMPNKMTx4ydHLZ}RhQ0U)kaqWCusa@En6RgJ5Ixy0u+Dxo=6w%I#YV>fi zR11e_up>!g9eJA30!MUHtXkh##=Jcuwm4(cYu;-FmoZ>x%{Q!<3@-!&0o)$|xiSyar=E3opr zyDy0TR(=0|uMlR2n=TaAUx*A6(X3$B_R?V{*^@wYbJu$%nI0MLuBD9`zg_+1_RHCmBHcy2G|^kxqRU~axsH$xhnffriu3*x8M{TFGHqhDb^E~)KjGSMdDzUc!@ z())*FbrS$cYp#r^U;u)r>yJ!W8PWY>BBm4oCv+f?AjqWp=hoqK6eormH4Y+_vLQObiCt4lckTBQ-t%~Fna+1) z@>vQ5j>X%(Bl=ZI4MUwNBN>t^&HaW7Q^dw%Nrv++ALtDa7K?fI)nF;Y7>G|}C4wa& zMrLzhg`tReWiMjEwm-lY@T;P=;AhC_*++M0vk^tn`OzH?EehxCtYR6{Ap$$b?95wv zn?o<6fD9af%slONl|dQ!YqiYYf`LHjRTVQP=Qj@)(p?&JIC#@OG*v8tJ|6sgu?XgJAlF3vSlR$ z193I#ok7u~abi)pYD_IZndJ@Sh&ZXd{)r-k=qx5`(-H46y2#Kc)eBeE-5c`#`T8y~ zF2%n4GPnpMhJvW%@`J+L&KmZ9ZZggm>#%pfS}e}iCe>^E5J3PMsp9C*iUzsL$R*&6 z7%%nHFe%7z&{A$Ofe$N+h+L!>%a|Prqugrp*R+amwY7gxzi*<95D>uiC0#bpEr3fM zcJlET>%2%uF_x1M(5alK1B#z~?$eT>lNgWymnI!pSb03RTCTMCxTiWP5$hMCAN_!)lP@3cGiePUQK2f=SpY#OnqeYEn)6R=<%@*EKE)eZ zjAxPpG=0?jvz)P{!++EeJ*5LZi!a-c2(ULr#BdNvAX12oVIbbAONaI0Tj?;n+IH>0v*f3G`V zKF$#hMrdW4Ow{6h+WT}B!|uPnTsrGAJNC=RL&&0k`}V6QM!lm#5W1JE9tLB!)0$fF zi=5g;Q!D7sgl9-$3VmT_CZJJlRbrDMc+D}`n{jVb8{duhU+dA1emNrSgY+0lSOjK> z-;Nsd$h={Dtl921EBJ3U^kRxnNja-%+z;QmXprI+$7^DDT~D~TuizF3nbR=VuorLy zZPA?UNteXFx|oL#XL(Q9W+-HSo!%d+TC>#kWm1;Z945`bnqu_Iz*AMtr!Ys}07fo} zPTq}~oX85p7YNZ&|NQ&6OYn+0WuK~J$8fTbE%Ja+iK$MT%nwBlb~?)*p`0PTvQ{dL2Mw=W8Vv*8}*|ms|LcX(1T-g27Eo7%lRdH4mr@PtyaDZ~D4@I=(wwJk|%9;OgV8>vFZw-IRM79D1> zKwXnrOm|q{EJPYfgX9&Yr@yI{@AJHhoeVD_F+=hhunCT#eh~09>E=8>aM3G=XyDNb z!@TzO@OILOJ7@OVd1qHQpI|-joAf)8;!xQ={SUYDP4?DXMuCut(%|uLePuoeh+#sk6qgQYtedAc_yJd40e%HwM5_Q)Czb;re5OhBjOmreGbTTH0%^ZS3+rm(1(U)MDJ${>7dsU$rnW zoP~$i+PmaEJBvYCX%qfQ5uwqUC^<>O{@92hU?!zbtn}&@aGG0a#M`1-Mgode@mI|| z(V+<^uDmPl`mSt5!WQ?{Bqp|Zarvi;N4>Q1QG51eOG7pZ@JHMs6cV~++K=h%=3j}K zVh^}OUU35SO}-*cw9+(@qa(vHWE3;z>rQr95T$qkJ%eoBWhPSldX82G5tT9a_b=V@ zIBO(baBFjsbeABf7$5;Q7WdnQd#u}F`va>ps|xkXxwXb=3%L_b*t*=i$3 zEQI%3j9MZz#Mh6Hhbvw!FpW*`XJ1%=4xd7;lfG5BC5J5*AK4w!l!Uu+5v7@d{Z8Do zFdc5UKI0bVV|l3RQ_ZAPUsKKeZPM^hVb*i;gs=~#VG`z9=ZtP*Bcx9%L|Od zbGP!O#EC$Wlhuf{S1{0!Jj6_K9qBz$<;E@<+C+)M3>}@5a0JrKO!GYt^+wZDvG=hk z;(;Xgtq3;cSH=g;R`5+Oath`sBF(oMcPu@Fz@{*Ejg6=qul}qF4Q27?hbtu6NY5vw zWhN|84Ub~hL95|zpQ;VH!st9u<}q|0=6MuxG7nnfzr7@wK{u8E$6JvaW%C` zwM@Hc=n~3MwaMcVT+j5~i54s#A7YQzqy^?5u}n0$ek#`*_vqE;!q1MY2DLv}J;>;n zC7oNHYo2ytrc0-v7j;;;Xxg>G16{|>s?e@h@83V|Q1j*n{sWix@7um_Kh>Q2=!bUC zEdyHotQk4+t&?ihw(uQV4PK`$ehtZf+_c?zn|tl++44!f(iRUIHtfll@YJz3&QDjr zwBuT#Iadq^LztbJh6m}Vy$rq6+nt4p)p83p(8>3VmWd#cxUO2xCV!Nk(@EZE#rBbv;5@475_SpWXD zHz@g1VCdu)yjLw=&zQG{;cO&v_axH+0|z#r)!J8c4?kzS`9s}y{byNp+`O z|E6Cx>f_XY_IyzjAqJcP#s`iZnZL#7R_c7_Yij7v1JFCaIlip(lx`h5)VXW!;)VLF zf6p}1n0x9jY8@YRsS{>ik3A_NaRakFI!*D>wlF+6wHP1MOTY|^(LbSP+O%n}>#=K0 zW->}$Eo}h5*yJzGr%NNhweN5Yj+s;C0n`S8Dn?-SmnR1V+x)I4TGQy=FIu!{VHzb@ z9sHF7iY)(ahMTwAoURf>2n)a8S$4xDL9fRYr|rqE8|m}Ye`ovSwT8`DRoZ8au6ZtP*G&$rFQhbye_M6I?6<9rzi)7n0f~ou zXdfh1rqYJ%)zTLx)_DE;KL%U;c!h7PdT4jWi0qMRt~PPvS%Pv3#zl(WU+%)CmN4ay zLQ6kutaaxplbFKwOs^DRG)t=w!*r7w@Ndl)kTo+aRm@Th=5D)qF88CZ3(3vI}c` z8gdNtOq@#F$)@iUvGQ&7NHZb}bwE3!*@L%L9s2ZfUF==h(wAGnRFlDmPk?0Cb1Y3& zlsk2rDp{iQ6y0Yj2!($F6TMi>Ci)stgBuT#V;bUSE)FeEkFc=p8w0=iA?gY_fO>Uzti+tm^(_ zFZR~)YO{;@IBNT68^+#E=KI$XA(VA2yJMN5jM~>)dDM%lI$0{4n#!8e%8=OS(DN31 zQwY~V?Qiej@4td{sPlHKiq(1(6V=5R=jtzNo*$6Vt8FvZcEI@YJ>j?H=N|Slt_;-X z-ZcFxCZQ#>*4;;&wTD{K<1=_A?M*LH9W&+>;}C#}b79#v8N7D`cF%<|RKVwhlu%U+ z;yPVcn47tIKjv;!8Qrt#&^h@TflAZ-whIswU-{Uu;xwnSWeyBcr^yr+oPr=l98h)n)oSb&+J6(FO zfC>I=lF1*ReUpcSLIOy;$HPp3)Kv74n6=Y}df?&$pZRDLfYR1&_caaAbxVFbfG|SW z<{m0u1pv{cW>rJ>KM=kffU-p+X9Tw!fgh5)NP+ARPKGWU> zVY3+bIFRE4{^7~S5mqd>>`_XPChdFnoV$8;Pd?c?YK!qO@S2`WIIy;p-~m~Ry)5V{ zHtGks1r6s(U|)vjbn4P24i$wHF!${E)9|8s_HQE^qwTVdt^XwZc*^!?ZSriJ=f}2L z+T4icEN5=>7~9v);e^$B9!)zs&TP@VdEb8h{(z$$$v_9fna_K>i-yg=QgzbC=0>jn z=|i!f`wcEhA+gz}Co@&vHk|7HXPz)SW8xXk!kw1BK|Z{JYAPA)6d{&if%^CDE`5sY zpT@Thq@f{UBypGRd4?Unm?P(9z03pc3USQ|O^;!5Tn4rB14NY;6CVI$4=DN@qi|%q zfs*GHdznpfwnZoP5qwTC#vZ!JHBLd_dJ>Ws(Vo+@Z8MvyxX?vc*LwVAJrxfGF zToSzUIIl?8e)0&bXQP>N`Gf+lN%NL1`;#Og2q@Wm^yX5hdE2{n_ng1U6KXnB&8tDZ>8D0+*50$z`SW=C&e+c%ASJFNaoDY%EGN{KVaAAVwrlsKpcKxB zp>;x7S?bOfRiYr+0FB+c&%`r0=~Pea6;;&%yG(teM@}y&F8%d0cT`X??5nHGRi5HX z+EyO1PKkR3BI}xgA|fC3oZxZ{2J?Bh#u_diD4`{3N71zrc`0VZJ$RsngCEs54zc_C z@*$tMPsGgYOgBS!?gcvNU>w=ud;5B>tVnWvzGLRxxvh`hT;8mCn>K1XI$2+Gi%&5d zcRhQ)r4OI>X!nVIbgg^xL0`kw@Tl!<^16@L{igHllVHQa>?GJGG`7ibBb!TlE228p z!wZ@{yIC^SXbulzjji64;&0y`x4)fadaul3G2M$yRh`37;e!>+qhq;hpbeB6#NT*ThZc~ck3opQiIBsp8AMqTBqXpQI2}oXF z$4Q=P)!Tg^Ck48|#|Cp9x(ce@#>CyBxn)k{&#JC}qr|K>ab!(?w~yw~R$zWrRn-*g ze8%s&99l4xGguMnGo=`YV;FH`W7zlo*bgJ_yxJ%i18$)1DF03# z8FMr?W-DVtJDCe0nU1v5lp?Lg8-Nt^u(ZU0Imf1GQ>COP6;eOc4j@p#>}fU$1o-C{ zU?5@e{Ht?w}qO1aD90$v?HugN-P59tv3+8+5#5~`INh# z_ZQ0sSj-9lF-e@9d$ayX5q9m~eUb}}V!dUYVaT=P#cih=+^1t4$47K~bmv{;nnGcI1lD-qCM42*gYfDL7FBCstsFsC1Qh{K}re}qjI?(~X0KW~qK&U-Jp0!cd4H7UjYR9~l4v*oi*jZt?bNQ?MO}Y~u=>|rLLq;nzUJ$yCwIg8&b$7nv((yt4x=TnkE@=u z=E|bT5gfdS@QB^7NuHrJfElw{5S!NcK8brDqX+|OMiA#U*K}y53bYsFI(M!r!&QG0 z41(akBt97XS=`)4g>sUViIzCS>?&(7ys)nL+o!)NmD=r%Wj(;+o4VHi)?L4TeSvlF zCo5l`nKTiYXd}&JthaV@zl_urhq1b||4q4A$^$-wKv;cYj8%g7HsvYZ6u8KrY5dG z%bP%mes~Wwm*3WFl)O+-iPSnKBs#C;uKErhumVSu7KO{5yUX?yN+a#$;%> zZw_Onjb{`6op(0$fu)H7V$b&oyayErPHp}X>$#lwX=B#5_krc;Z!UxJRYOWrHg)sZ6HQm3$ zYnK7lKc`e4(yFlbo>TlC806*kYQ-_ul zoc(etcx3n_sEx4ylf)mkVoVBz&JbD{BM((&%|;nSEk;<=54Z1Ml5G)|wmu>w|LNhq zU>Lm>$M%K9Zoi=?Nt}+RjejFaJWX>fFId_@#q;Bsno`^AIsX`V;EjW9)a3lOYS->i z(3PR@KA|I`yqLnY{SQbgFOS2V7@21ga`=p8NTJ+?6ys`!^a!i^d~|4I5_C zcqA(y_|JX&_MOn(J}k|Yc;^d|a&gBqBIN{{5+}5n2@~G^!x1r_e^yEQrjZ4+_r75W zO$u31>Z@ME_hDu$^T;o5LU_k?tL|T=QAX!c_a>Dn zo5oUyCpJuF3;hGC=_Y^tamO+$?0N7Vv;M3on2Zi?F7Kgz_pfQ!+Wt(bp(rwmFAZ(= zlk(9CCmQ9He6oz99Z2sUt$t#1bKsNy$;Bny`-3+>Ik-Rjoe@lY`^>!6_P^NOzSfE^HfY7C04YCJhOT{gNFWcIS**qr z1LAF-!1X96JYc+C+A%k)4skU_Da2b3M3G<8?UJ8EbD`Xg=E+e}aEgznssdOLHdaz* zTC{f6nX_OEOSHzfQEMC9{n|KvDjI4iI_^?(p*b{eYO}y`%NBqw#2xIZLd4QT^V^fO zLuH$8t=nSfHhM|0Hw|W>5-5#rJ{}*?(!Y-c53kts@1>rCnl}X){lZ{si@sEDl^!d<@i9+Z=Dp?a)#|^j+NIc9(~`%_Y6#a zR8uG}XKcJ6sZiHX)iKoNOT6~2tn(pw>3C43-eBVs*UXehZenaaFj*E|HEPi%7M8A^ zA@{AymEK*Db!X}W@_Q4X`0W|iF?A5oV1x-$z>{$jGX4pPaZPu5;NAomly?SC&VN&e z5qresrPqf2KDGO%{V)U0=v4cqiiDE5sfxa3@Tdyx?_`!ISMbEVZ2q@nxInfz@2nZtnJdpLqg=M^kkq@M zbu9}`vS+N^!d%2gL=|p*=2#DzZKpG$9PvgHGmLY?ZF^B{4b%VSxmlj!D4T0@=Z0Fg zhXj>PsIvfL=3RPtajrzK{83k37*%CVa0+cZvqwWmy#zR8hScd{R)Z&>#2i$re^hL}dK?!Q(zC@kDe~N^}l# zZ(r<@$KbPQ({A9+%ok475-}3S=l12f$_g`@H*TrP_h75pm)f6u8Ap9Mk2EPU5fE;9 zg6~1u(TtLZftjhCFt?)5cfV6{Q>?0CfK)~PSwFnX@}y%&X<)t{IX+Vwr4{baw)UY; z;$vTT99yXo7C3NeJ-6_-ccvG?@N*y<94HPD?E6KJ+Ck-LqqYJwTm@Gof3J8QtAxO} z7nsw8!dvp?GTt=FXnc~c0)t{|ly90_6o-3+N2?L*9#8qew21L{G&E#+epg7`Vv1z9 z<)0N#%V(9vQBu&t%XC7M0}TVv?RLA?)0z4VVQKDJ$FaI29@84qOu5!n;ZriqH$)Ak zUe#OV4bs(cb?LJ9sTm@r8Q!Hu#f=*E&pfBSj~{Sw|7HzAViTv{ZLK@%9Kh_(8H*2bnNDo983aA$4%+QWG^ z8?=SI*esJUAD^FAK~SIu6dd`nf&pZAPKC1XhRyJtv7CFuz|Vf)!fS(9ci6j~0ZzPH zC&HcoO-zN8#iMzNGM@NB^euPcW1Wc|mlsE68We2O4`IRRLCmP---e`1{1X1oiC=mt zqN9Ug)+Sc6U=SbdfA;cLl%AmhpjjBl2v0(?#nVF{&j#mcj7^;)42Rl4=#)mjZQ!+RWhD@n}ii$UUL_1ThCVe?em z(}_DCxm440m^n@ynit8%k;qKT8N=+@o9io@9#FN=9cx3-rt%m}2PTpUlF5<$E~)yp z?_X&M$hKAiyr(bp=yVH^mmk)^KhbE~v{q>3;Z^s7@we&JX%Api{SUfrTg*}5Xh|~_ z|HwSqdZj$Dx%3#SJd^{&uO{1Gs1%h@A0A+gw%t@M-A1t#Fc<4kc6)6kInh7iD}JyS z{V2Ek#5zC7!Kh5^HVB;psiocLSH9r0LM{aYV3F>YvP`z&T+e$>Jr7vlb{?BfskNFK zjJupm!s22Hk>ekoS0*4?L^N%$b%@=RA7yB)6h!ZsaDNb#}`3dFgfUXP)C96{ssMaH_a`Vy_V-MJe8NZ=Vt z%ygU@y|rF(R+L6a>Lf%Y;b~4l>IBl!xtiZ;j4CMI{R+QUIT!TrL%n;7@H0FhNpo8< zJZ{wN~%+jvrot zO#EwU2AyEzn%|X+`*vvR5TWi9KY?ggmA$Q{RJCO3yQZ?X)9aR6Xf-v=>Ni)r1L{4o z4CRS1CknR!LYquwP3>lMJ?Ud@+LHf^H!%67j1uiW@5hPJVe9xCyZMx>5$SdYk9rTH zTX4$Cs1}{dw+|T+y5x$fufv@UHKc1s&^e@|4Lza(TF}ebbh4R(ddDGRq{-snNUx4! z@(-6VESagKgzbxRp6)I-8b}$|wq3jNWSbvIXhD*nJs3n>3A;~WgstNr;FUStc1{RC z3iyG(LwdVS--}I%N!}?MpKBMK;MPt)HvGAP7F6Po9~+gZlT_%j9sWBdOJBVLXO{+q zXyUu&eGMNNS1ahXUq?bdR#d&YwYD=lw-v+)lTYg}r95Thfsn6=ckWbG9oiQ!ar&Dy zS||>WW3B-w<{zXnif)U+crWTn|B3?ZR};}K(z8IY48o1DLU+5X={>q4+%K!rYMl8} zl_NZ-@T1b}BY50QS54Duf-ZaJ&9lseAMVjT;o#c>i6ZB6q zTmnFn$S}vA)FP-SK+iYuCY%AV@2(LUEi|VX@Ave4PcYBHfN0pmLV{93Fzda1;Ni-t zsRuEEgCcecf8|fJ+v~gnQkf?@3SUBnF;fb{{9)k%`X~4EcmanVSC=fAd_L7Vu1(HcdrlHU9K>(v%OJ zpY9_@{5;|?7R;bvY>7Au$}iMBI&`c+p!^cNF0D1{et$~84<&s6o*`wZw+iAsyyijDu3x0Q!P&#SKId2}ezpF~ahFq)WNVUfE zxfqBelo8v)$LU3*eHacf-dsCk;iSLEexQDuu1EpHszSzofGAQC5=BlA)u-F?eiqt% zyb#R48^|kRI*eEHDuua+U(SYi7Y&$dP}X4G>W&q+Zx4U6Pkj=|&(Tq?h9pUX%XWrv z2BEKKPPLKdj^ws&+E7M=&iMu|zPYM3cv%4u(@M%pq1g}p_0Q8|U$Kgb=6fbO{4(+` z;v3}`XllD>vTVn{#Yx1H9iZ%D@?_m6P>F_ddI^Kfn*BxqjWXbn4R5@XA!W%Kq-+q` z#jVFyRykA2BQ_!W^%y+3{K)6LP2KGSlVgyS%ELz8eEYo>(Iqyem$rKiL}I#q<4C3{ z0e1L$fM|m|UY~AC#cONcAuBOy28GE%{bh`z-n2?>VO(slJKLDwG1NNU7a1n`g%E1; zAg2+@{qeH2)It1WQX2~3_ytBV>Y zN2bO)?hjXc@_8HT(&re|aqWLVNF`p4h=1RMf8{QIYl4+hR}a$ZN`0Px&+f|iHy?IH zJum&!sGlo~6BzV}C4)fA(S4B9I9i0`A@o>?Lk{NtO1rb6kHBt#Q@|4)Z_kR6pz@jJ zB_@Os(6Z>lie4Whf`Qmj)t?3X)HO)vsc!^r*zM`5qy40F_n~<<+n!H~-_k*eaSfC4 zV`fdHSxK;x7N1Bhtw@EPVPNvH@fw{w$72^=1%C(fQ(#?$C+}at;U^a3zxNH{wqPUN4!DH%GInR7FpTG3Ixd5DOBNTu0FzO3$ z@;n0OCE(tMbH(1aYJ+k;Q~;wvUEqQJ7eTg2K@t?F5MRh-ufS3mcoe8AS*3_=?AF( zq;i9lAh5oz9=2)I2R_` zrFuC1wI)(N+`v&ku5{5l)BP(l09$VEET*%vZdlq9Ir3);W}V;l@O%`oY{K43RhK*y z>yV1bLk`CW%)9OmAG{78wBiTb5B`fyy!k>hDHIMi(!MUD3S6+Tnny>#); zOIi@zH)GGQBrOrTLX@&x9GV^tx@^rXkX%~Kf&;I@Qap~Ifspu(V zaNwCt7>CPHr^txAc#@9Vvef-}r|$L&AXRs&q5<9f=E{y|p2_q39=X&ZwRaOAwHULz z8xIQcM1k!48H1E#)UGozm~UUo3o1}sn}bGUBb-&@)(yyoh$s|6UyI**^A`Qu=48L? zZr^hkqueVP5#1DDa^-oANCbp*$9m%hvqCr4rv7o2=eV0z3>4@r(NM@suEo*MRiEGR zlBkUTho*4lIph;zsMeJzm!`FruKVD40Wh>fdg;2C(z{$+){sUfkFQR_)IIJOmBUFw zxij-$?fRSf5=FDV-1G3OrX(kW?_qh8U&#J0dS2{9Ezt@@a+Y7iIor+U1xBg{Ip&2N z{meuR#qXjJ-MB$AkVR-_li3ltYt5n&x__M1t<&RZ=_{}kp@ZllExre+sHn{3zW`WH zUbX5&`@pD7P?yC{mOGyT%JTDdIvz@K*5x})B6_i-Et}Vrk5D6@#8t34^;I>cR`3>+*M@TE)&|uv6a96@eh7OSwN8eURiWIKkPZ>*FopY& z7B+z1CaPa54=#L;YCs{a0Zsc2F#@rn`3BXfqCXXMkZU%li)&;nCX2htTy;vTpYU;2 zfC|!qp7=t1+66Mu6j0fZ^s|RP=V;sa(8vk9KPZr%Wv}y${F^^RAdUIlRIc*Y&O4u> zS}}!xl?0?v%v{I^<}fO8#4=}J=YooI+##~X8LML&y-{t>dnhdlWNzw}k{o_EyRXy5Cs6GlPWt(zO;)AzTHV4k0mypdXuo;K~7PGMNaZenmr1W&aWXarF3F@XcAHWy>_ zsBCZEUtI2E5^&s&gS@B8KF67evs_#!o2h6g6l{3Q^yb~|-Ix`6&PtU3R10|^vH3zp zq1Fs5xx>k?^ut=ii;w@R*dB{2DHcVIM#x;~*y>j~3YO(zubQ1)>{47(qb|u%tl#;2 z|1!$_g~KAm!rL`;$yQy>n+11no1=sqa*{}L$@E*T!b0o1c*6_r$3`8`e_ik-oaotd z(Ux!j&_plqKc7v0rfktJ2(001L&}jdETA3ysIg8 z*zDD91Y}eFRjePGcYS$y&W5@tcBAfzO|4FaiP&1fCM5>6k}#@wo0L<**sQo0%MyzA z-}vUW$uJkh4iWxx;82nbr2qra@M3&;d1y+lLUJ|UYtQc8OUaeZIt|^0%1*G9g|Fc1 zjHp%F$@K?h8}jR|xHujLCGzB-DF!>A*VGqT3ifu!hU`5;Y)#-|pEzOMd*ndKNy^KWlGs9#;^EfN}~h zoI9RNlv6^0a)_jNfB?Nbd1W0smPA%sN}s|x`m)E*fGjU0`CZ^mp&(F~&}+=2rD0`L zN3z00WK?Avxi!!djGvUV`Wn@7PUl67>%RU(MQ->N>)L*-?!KwU#!CQMMGMVOsSo`+ z-r3oJR40Nw*o+Il9R0Eijzu{c-`90id2CMr)Z<82gU%<;QnYSSc?vNUT=K52BR5^e z_0jH|!`{bZL3Bjqit#+LFhyl-JBP5`jXoIRdt>i2dQud!SM?S82Q$6m3 z=pKl>bDE!=h(v&rE$!Pb4*6<~My0-@3##4YCnlp7rg$58<-ft{U>#PGLJyG*6U)}u z`rQRIzl%F$!Gm?FcEoJxtJRrz^$5B>`9>Fh^%8DxeH~l666)`Fc}nK#Q>P>!@HgM{ zJ5;=OBy+5$ka5~MwB-aRx-QAMih~kvf_81+VGpRAI-dNpNx;t3K%*n0__j{2vy zqSocq6BR7q)Iz-rbAmvwAF`rG@otBV5}ESczH^F~*D&EbhzVb(h{-d|FUn+$c6o6KAO&( zR$rKKPzakesV2QTtS%t2#%?(aHiyipgR$9|*RYhk+ot`8dQlAljTm%IZl1jI@;~PG zjf&re?j>GVP*w}tGQ`<>vVF_u%~zm$W)i@xJLQGR%W9R^H1FeB*7rv=Us_c4+9XAU zC(1#?om=cKv7u&01>2ab_vF3pQd+0EebK#Lh2E5NMxU8D#YIB8+UZ+cs4GNbLDu1; z5RL$c0Iy`HO@L=Dt=Ns<{brqS#aL_(2uD;8=N8khEqmNKbMN%{QD36#gCF`MKlatR zxBfUat%+j1#(x@)9-`f5ELwz^crv62dNGf>-=vYJ@Nr>B*ML5M&0!Fuvg-F;>3Kf4*h=P`K^{B!?`R9@-496L8;L`-N; z5@v!*&K(LcS+QIr86mT2g+@=1wEiLnq$CwR1O2b^?zQXqP`&5h{8Kk#r{D8A$Mx(i zh%FN!Y7(OTfhu^YZ{E!H8@Bc-t>Vn0G7{p8cC~)aL`)OQavE{%2ra^t#2qrk;`P>U zNK$(B;T7ly&=?#6;@HpCu?y~^*xEg9Y1dT%Tc8Iys zBf*E6DWw)yOm>)bH!n|<8JUmg0ud|lc4PIBQ8N811(YRYtT za#Zx0FHT5H5UqA$O~HdJhF+Zikv6=_YS>bUC$KP}92{x~;t_4<6I$eHin52+?=d91 zP%j7U$J~Z&87wi2XE%lzkeZYDEFmRsXU_}9aV@S;=btNEPp3*?z-EM;pv8M^{2GFW2 z5(c&rnt)mxb7KG>!-kITCKcb1Mz1^TSUw^l0i6jOg#`K~#~3NsaO!2og^br&{hs4s z+8d|2|3J3HTkjg5B=sNV!p}fKw&31z>0Ot;x&<*rS=4rP?pPb@Xmc=?f<2|W^WFRX z_17Mni!wk(rT>)iD1&tyrFbux`fH~GcPCNxo_!g}DM87r|Bjx7tlsAC&#-Qm5=Kpi zaT%6SazlYR^X|s7m4U6%miVY#Si*YjHF-IBtJjst7Nc0W+S-p~#FK5P3j4{>f`h3cYe=p3d1tLL6@F31(g1x$;VN@x3guvWiw=R;wo_W{O7{`u zjBZOQq16b@4n@$I7W@qqrkx!cP@m9f>Tnq!(?}C!dF^O8$Dz_}(z2yCLH>`M01eeq zqSge{Vb99Ca5Jh!!Ymb@K!!>TwjLqmo_+VAA4eA2G=Gm=tDdkp1QUGr`>FC`ybPEX z5@{c_mMgcqTo7eo*P;N;A7;L^T6)r=+*4?tL8l6 zgVJ!C0vC_?Q=Mr<+dwOZbif%JwsBmIQ;kx=6k!u$&{X$%un5%tzIE6Te3p3??bpahHALvx{P)+}2OSQr7 zz(jer;aAg22x@lPU+Gmufxew|Gv130iaSUgN<-|$m~;8gWOVYb3mz06__6r`g-MPb z^IBsza7|_{D|TOWe%eXRh+E1smsD3jXD$yYlBN>*_Nyxke2}@?f{4=_c{vvxoY-|K zu~s4UuO0Xz*doMPQC1HM-P8x<$Y$A-+h;H9>a>rSQInqvgBmQaKTNNlK@7~X5HyP#$B?8NNawFrJA6%Zdd=W87+u z>qya#HryZc5Ylko9sAw~HkEv!@Fp+F3=Sl8B+A*O&EjIsg0KJBi z-Kq2xRsF!MrzbYqzO$+|5B(Djw(%!X1TD@wr|Ev2aKi4bwWZ}9TD1cEWAbx4j$EYc z7}9v}>8Bo*xM)g@dP>ORR5cT#&WhMsKRD{VmE~mnQNT<(9oI#5@;oeX-1q4K#QeZS zP$m%stBqAjP=BUOA^Sv9j%Dh-*|JD9&wj+#N88tQMcF$mf&dizyBN-ZvP3Ix*+rTP zC2KS9tWU$Mz)^$sJGkm8kh-w_7N?02L9Kzk7hHMWZs)VKo2$pc!EP;@dW+V}q)3*M z@klmIXq@4ik^)|}o;ihCk)^^G=8WJ~jK97-4{+U-iI=8e?#wB4)7geg# z5FA}^`%r)?*r&I&*BiD0RQ4#(s?%OK52w)oGneV7sACRu1n;=V(z}CH_TIc!&4`1Y z*Oo9~pCB_zGj94n_588(8I}K}v198|eDZ%CLQEcAzla|h*SbxcWi@H44RjqQh5Ic~ zpQXEK!QG96VCv5zwOm8@wV|sKpVIp2dl3}}AIaEP*M!iR2l`;lcngdCl!i62s(?K%_oa{7s)w6CA z)X=^#kjlTOe)f;~`~uH#IPbXTBDXSr$~4q&_?2E5-@zClV57roDh|NkzNa=m*|MvF zs(<1Wc2V$+8z`|Ss_q#rd6noEZvK&rL`6R5X4(av^DazgAYoLAMhk=F*;^r<3V#@U ztL>)0$9;e!%qQEdLmHqW`Q^(zp&ZyMoGY{?5B@k&b*{({!sNkvJADw zU;<9)F=f%r-*5R0rH)Jz!9AgX1i!1s%1>lW77+oRJl0nN*<2&j+j-QD_UHGZTLBC* z!8CvR?fA5plBXCki`Q|YzJQ)qSpuMrEYpwH@NDTeZlCd<^Lx<0m~5*XhyN@MUD-xt zBFBcW{}*+sEjKQ$&Dap;FL-s?h}rA6d01%FSoA}|#LgLl!(c<#=n?ee?#Iha2EQGt z(o2L4n1)soNR!93d-v|}kuCp}^LC_D_pj94UhFl=e&SuiPkKS!{Gv4x>*dFx z+F>DculSPn_vk>@3|rvkSB>B+o4-f{`$Fmx)@T zM{hdAmOz!DmF!;2Py-)htGmt)%~RXVzk=4O@?TdI$EL`F*dp3#`( z^6$DD9Vf_{g!L?_`J3$E1nFQ?nI~y+!+(R63;PH!DO4zN^4X^lzWIgX`2uY{rLAQ# zb9Zqc{nqNveb;ld<>6882;M4zk1$7F%M|mnYog_4MYRnJ9yZ6f%}^c^Jz?GVT$v(x&nQ;;D3N<4NWApe|9U*WUh$h7N z`o;ti_1oKj?Kjn4cTL8f!!no3nsfdxqz8raHJr=Z1%SDYuj!0im8OBgG3=3;8Ia#$ z)?Z0j3oh`Vi|+nVzXvFFFe3Ha&nVAMKe!M@3-GsfK}cLJ z;Yl=o(3$41^95359HA%Ch4Q#xI^%@Fku2&^lT1a70Wvb~7+Ai&Ehumd=2OD-pqYYI?v5#0mKc zM=PdeNm--5!O8UAhN9VGe4hQYZKg7)0xyXdtXI&2f_^{hVswj2jhi&dq27v4Ov$IR zgOT$jk^6ssAA4|{FosC5t;P%|>UvkFt)dnr9mCTt2Pt^;_;G6{xf!C7QA)NVb;Gm( zz}Yuu@ulDYLC*3S=ni=C4byLHMIjReHP0u&4J^D>#q8|A{{o7sSWh*TVvw$A97yQU zokD)MP;lVmCyQhU*PS!Gduei&(Dr0lN8QksW;L@L6K_S{_gIW_l8jxAF!eIBvK^&lxopt)sIUY>JMahD^5ns z?)hNaMCnx!yw>PyBq)JQHdB9{dr>FQkOfoV3d>M$)!~!v1F10wiZ%?#*6gDzk9SAL z6CK*orcTEjm6O#WnGqUL)FI$HknhpL_H<6o=X{^yYIYgrl5VEj!rI0ZN1Y_*AjDYV zuPtrm>(#ME2Q?8ni072DqKB%+nq5Yf1_Si+Q0Nv%FB_e}xC#U6Xz-I65PB@+JOG~e zmQv{(#L-F^TU&qG!Of)Un`S#@eUwb!QvhTQbjR{!)$7Y@d~eh1cOpjBesTKyUCDoh zUiOPApP1_rzBUM%*}K(<+4H+zWd(>RktmqArxr0Tv1lpUt7uQ{G|Gnb&#>I06s8v8 z`*krkdvoPK6QA2<{VxF3pM({r0cgZaPcED^!-vAkm-7ZaL(+@OR4=jAaY9foGZPc` zw41Ik0c8fG5Xiy=r@TD(+5=}g3;(7!eyTk4c)V1JRRkg*Y}~S?Zq|R~5D>CGun=di z4OookEpccYK>?~vyCHq5jJBe7zVteuO{FI&NG{CqF2sFbDnDX|NCI9*YMel5Qf;pia z%&P<*McPy9R|qJ8`c{_^*raB(DQ>~IaV9{M;NTp!aS#YmQ zx%chU<>du^kMxgTuF8&edz+6yd8-XQ`(m~w;w zmQxN@EdFu^I70>%VIa&Z&5V9UFdNQyi=pgH%#w152%Knfn#U=i2Rmq97<@0!QbOP( zq6C!CJi;zYb&*~7VIdhaJjX^ws;XZyxW-06CEZdL=Aps^@d6efn?XyWl|XKH;9~$M zs$lLUmC&k`qvU%*6j~lYFbEJ@0?)oC=UyINo*GK}l0j33e8wUm^6FBUr4{q|aT&tr zAt1h35H;4k5_itj--z!W8jtKh$b<*ClpaA?REksWk27^u^9pJ_8qV`Qfa~d{mlUOinczx#HyQK->8U>_BrYXPVqm4BLzWkqKbY7lt#fgZI%JVhckr9coAXj$nxDIe{GqXq(NqYzmfd7^_w*4@_bH|AgAVQmM|pNST^uF>c+!hV;`G%?dh z{YAaeAoz0Jf}5*OW)lg6OlU6UN%5K2=f}1t*70?jAZ9egsk$u2i7vlM)20tU+S`_< z>q(o=m!JiH(jU_t{Hj+0uaaUManfWt-uqI38ut>nXf zIua5K_~->NnmYSUtWEF1(6M4|BnABHq=mFSs@HB(c4WHzS}Rln9&h6j@EjOErG~Qg z8PoLqi*eNU&4m+TG*{7=5LkQUYIDN4aDzYt#N&lNgDYSg&b0jydYRLtWJ*{Ps~972 zkP%i=vjx9t(!Lvlen}|gMdf!*q+)WQp6&uDZw_yyFi>zr@5Jd{$-OZ^4BR36&)e?DV zX-E0&{3+KE;Sd|&3b8zqfeS#^8?F(czM?UNU~k>2(;q>1wN!vn>T`8_vtBHpVza>A zvG#FJ9qfrQ382mc_!8+tlSGJLH@AYiK|CK=HPxj>Ovw+xk?}QcbzcT|GEXI&od&cn z_&3;S!8}YfMO2P6Af>E{?V2@f`Z8IQOpsJ5hC@Y9EaN9GP-;NX^>GM_8UwP`U=O&{Fzf-O$d6vgA)}hNy9cOhz4BhCybpqq7!q zZCM;^OeD7cSzXgMGt4GS#7QzjgY=l;JeWhfiT1UIF2@w6OkTwbp-_q=9h?((|WfVNNtO!j^m~%+XKG?R&(k|4o-Gt=RJ(L-TDv!U8Wu zEZs7}l<%{FZ?SPwUVcr~AeXB59a-a<2*AmSI2L|?&!CP;svViv^fX;kZB@>E<3Ro@ zQx>jGrxM6}gn8Hdz(4q7yEy#tKN;+3m_$Z-gLkr?Nxz{(4@q;4m#^K#4z=dhA%%Pz z63V<=Yt(@A8VuBqR2Jlfnn2c`F3%_eJjkQJ-YG^8M4q!B^horjD~;0bGE%3&4nAkD zZDF}{`}Wop>aZ8aKK(Ovw8JtyAT&fgKm`LksB~cX#?nz$Nm#H|R3)gt@HE-PT|Z48 z64I_+pV`G{U1I89Et6A50g=`jBUttwO9=3Om5)SjFx~m&Z+_L(kxPl>&dBnqFk_rE zhDWtNFl3P(S9W)c+CN$OMKeB3TCh}js zH`s{zD+HN>42lMi9Hvq%5RTjk`ed)ZT0j8me5Jnyp&Yq!?d0Q?wbh~fBCEeK`njF+ zf(KiVDS?KG)J`-fq=e}_Zx(?}LRlaqEg`p&1IVA)$cq@Q;4(YFL=k$Rjt-ylOaEbJ zJIZGA30=b|m|s9!4J7Gf%*>9x)m_X)zBvCtG031qVxHo)MP4K%_BFeGfRt%B=9LIp zq>-ec>h;TjW>2`u=rii^G-`2z+~uC|bB*{(8AFYqb4&qhJP9T$@_*hlozkauqc=Kcv@$La|dsC5WZY;^xhpzvqgZKEGI6qyB}I)=g8h0d@e) zR3u)#O{U}vVPUWv7XhtHkymi&fs(NRB@uPkG{p>`EzSE9=2rOpby^pyl|gv)bLf07 ze@|DPru)u<{+WJV05T$+iRG5O8$K$p0vUf%`tM=dWDd?6clD0=jM_=0U(W9bM44&p~ z-O#`GD%vFIVg4f7W=Rv{8?+@6V-B`I^`IS&8?p0w2kZ3Pt&N)n;^(MDyihriyKu(u zjt;-snkdB!MXSl++4u0vBRWk&(;cio+q$o@_khKQ3cx1V9Cx(KAex*z%1O(PSJ6eo;&)9BcU7jryu?UPk{212xy{O zDsiAe7AYRj?=0|z#r-V<8s1?E^#Jh)A)Lhe=@bRrQ39>)NH zDAj#43+{DsMnp&BHCc%AlEn`)_wjPHK+;45QLH3^iEs~PGp`)0bpgmu5|b4^WRcD< z{HND4Jm6n@@3q_A|6S8HC?PC1CN8^%=JpXOY$A2$j9K2an%9_@b*V0!DE5k64}^JC z*t==N#rFg8-Gm{#Tq4UHYu#X0A2Q{3A*1~wG@hGNM7&KB^)6n=&OR3kxX0A0zQy6I zjTYSA^2;Ing{nQZ)St62zcI}UbnTpmRRX3bF(fY3t;2L3t0Z6nXvuWy_2dnqQ4*u1 zVxc5mQ={IcO9B9^F@@huK4(nK4RX5J{c>2|M!Ak6Q1Z%VFik|hj+0xoZr#9jzx~o{ z@6_SeYyHI6+P1A&l=6G4#GZlm^;LJSSsfJzSnY+3^7Hgga9k-`dPVp~dDFX$XSfB} zvjtL2QH625-%rUp0)5klO&g8476tVNa>nv8z9ErLXN+}uAn8@=l8j{ZY#>i`%hPct z^1q`?cT2Y;AD770oxxoO*J*bFs0j}2jh%{z_Ka88f<>zr+e^W5KjyN#P1+krgEEZ` zdD-vA*wl_GgGjWaDmVQ3p{H<|XfZ8al2)a$YO^gexN*SPRV1{;INLe;QR^4gj;Y+d zV)_WhEoKaB;HOVWe(L!)y}HBQ{R>*p`f-=yrlNtPq5O#64Y^1m14|bDx~i%g69Liw z2qr?*5@TVa^_GkgX;H~+S5Ro-emwy8!PV$a__zzLQ5M;1a;i0b_H`9%9quK?br_#c+&qv+HWL_vRT*yh|X(5$80k2kCWX{`lcffMAC1 zp1BDz-fy7g7sk?6jPLkPCNhLPQn&=kM z>Gto@g3^IyANicx_*UrZmpc$v;(|wsz&BpDY?-gVWB(KC(dG1ilxQ~)W^F@eZ z7%Dk_+0U7~Amd#K{jbrvbG{myU5s1U-)EjTJ~M)nfaWmTmb%tO&TjaE^OmAX*~;M` z0xmCZM-=pl+2POiWJ~Yv!Y5J6+=Qo`-3z=jh7lsH7zp~2{vjJYR=Lf?^*$jE;(L3J z?l}N;GQ(MkaVw$ymb1{CT@ZO^9aKiq`ST4e26oAYX~aD-Yli>Y$@eHizhd(T%;#Ik ze+G}yAek4PO%@<|H^G?HSCB=M4@`(l+m^@CcWSd>ZUivkgAU@T$iySf915&HkH+OH zMZkt49{;UGc+W$opqs^)lRqYVAx-EE3$kFdVA}!KohC#UXJxs7`v4J}fO8!+a2gCO zE+fP>J|FMtH55}SO-d}LsFkf(m;GaY*#ONk??JlJnqj<0NTb}d$qecD*zMt$9tBPOIjV}rJ!FA* z%%evGg|B~g;UPjo?|)cjF9U{D*Gu83NPDoK*0tk=uvR}aZ#_REF8|x6d1!1Yck+3& z1^X`X@sWiAj>UpTVX51o&79reHX10rNSYEk0kaPCsbw!yM1O^ME?GRTUUxt)!Rb47+LV&#(Qvj+{#KQ)O8viuQl?>}hNR=}`^R%UZ$VV%LT~19 z?C<>qPavi^P3rE*^Sha}?3$?CZm46?!0q0LG}A8ULBs5)e$F5z_Qh`w@^l5RUKpz7 zNO4O^04J)cHj4j3ED>Ksk(&7GUm)hj{+^_m^65NqL^vCSCi3CQc~Rkc;W)PnisnB| zCZrD24wn(y2gBK65n#lUIK8brP;AvDvQx!ITpIgfPSZfD5>a=em`-G<(Rt_RX()=W znHo~wae3;uQF*N_4HZ0&ck$KS&*itimq%f@%MCwEcU-WU6Aw3X9lfR7ta(>-z|@9X zy-slMV1CsFTlRt={C-=T=%KT~P(Io4S6?^1FC`8Ut0C*i10a{|BA}+M2c0Gy+)17S z13+Ef>RN(2LJDJ!ze!IbcApz&tuD+B|LUM*M*7iE0Q|<9sfs*+Kbltt{Y=teC`-xC zEd8VXV5}xY49mH}9SGAAKN%pS$rdjuj*r_TrsiV6JUPKR-ct>i@{vLNw=rt;h4mAD zP0w~lSDYdhNKjS$?N{9*AWaif-hqCH-bAFE)KH}Q243LVQF2O<&hBUB~0F`RcM&%1h7WztHLVzTzQzD4sPmq`!=gFuZ9P421yCMtCj~zp>#UrYj z=o8mr7+v&uK@%C{2wf`WCTXxdr%>WI@Q4n9GsGK3G`?->f8af4^O1W%Nr%}ll4v80`AN_^1uuVILNrQvbkHU= zc7mxxqUN&di7&z_Tk|?!vY`|5yZ-^;5Ly;<`DL2N36(L9r3T(`&Wr(OHV|5_i#B(x{SS>(LINEF5 zOZ%0-NWviKcY?B=OyhK0Upcc@=QT4>T$riUlj6#ekld+HD6>lZb!-MgosJFKXh_co zeBX54BxO9z=(81?&}|h@_Fd*v+S++IzyV(`=>I|CyS5kpjmZW|(7{-2u({d>qoG=l zG2i%*1L06`y&Xg2gDZA}{wYEc+@ejIW~{%Brd7)++z1rp>) z_-`pDl~Ub`oB$3x8K_+kn9r-zyYq3t1yW^+VLU(9;);#?+xqqDfibG!+k9Rxn0XRP z0Qd0DmnE#T(Vb_-UWp-Y;WMjvq`u7FtXo~3=IU`*{Cx4(rks#Cje3(xUEAwe8gGd; ztYuJc{)Gbzp5UCngJ9)b81?pd5^0PT5eiCX+UpVr|JtY&_1Bb(F9tJXjx<5`NYv@J z@iX6tDL0*Y0f|Ts$(poVo>Zzgw5Vn+OtEwcktq+mal`V$+^2h=o8+^PZ?Fqw&(EA? zvH^f*3nbVWcRHb_C`PW*lL|dcc7Ja>{nna!D5^<8tGxC(8h4sfQRNJ{v4i4M7Gp&G z<*w+-()WS}ARfoFY05Nzs#~!3M*Tw{)=fqMShYfXM>QlMxN>mg-6P?r41Q#9E66hX zWyn4RUgx2|DqR_!qHy)m%s-CV-mxi#>hnBM_u_;$*+>He7E;g}uQ)3sf2nHk|L225 zVaC{|3?AP5{e1yIjn(nw9}9C97k2D3+k-gl-?l^n_{`iJR~Igf{Z_onKo0yqG<8u= zBX?sFB>W0Ir=b1Df>pU;Sk5(Dqil%K7Hr?OTbD?Av(8GQ#+cUZj|*M8RP^u z9ILg`O_EtnV89hF-z~;++C9mh>nSNTQ!k@uknNcr93_%AG%~GuF=@ba#+(r6?q`5b zcYOF=E!`(^m{Iqhc0vj4V1#8@TRqOvt4X+bfdw;sfd9oD&l{6s^THCXniS&gfEFh$;7p#44ujp z3l)3j*Damvm*I(mDXiwa?)P|cwC|epo{5Hg#uL>aEx_JH>QmPnLIhDn$okS}{5IB4 zNUME~Y6W*+6BZwMRV+-JtDN~pE#rL77DElthWV4_C-LXSO)mJ$TWc2xcS6;9czO5P z(U<9=I`V-F5e?9{>fz7yV^3U3>ayYkmS3+SWF}5oXN?mL(FC!2Jnr&ywVk2gAgvle z`oYxnH=1u@^#kt~3{8YroSvL(FkoD`@B53RPG;W8t?YU`ywx4N z6ae0(IpoXSDkSYrDig}C#~dGo#!xpC$zTHYhfj;Zok~MH*=K+TDR~6|J^AucwW%@^ z0#CmT)s~fuV;5Fr@C{2{6_dt4%Cgi-gb!GpMX_?=P4^l?{{fP|+w%5FPu#ykeWbVV z{$I$h1k+$-Fmd}5stA!xB8yl_thSiA5fPi1at`xPObkS(aC z7y-YxRo)%dOPrZ$T_l2Sq!d!N9=^Ii=^t?>N6|aKLL`}Nqu(a-fBtwCyeB8Yj*5@3 zAnW{F$lU3KXa!BrsA^f^Pog(M9|P^PgfP9jd3|(xVu<9BLf8L>OFX6k_Vm^ zX|+57eDWs_vrdoYyszm$+3xhMH4T_%*Kboag!331urjJl%H$V~ z!qV0T=gx+>5(Mzx&rf~CWCHn3LNb`LH{i^<59x?RS=p2fZAC43yNCNdAVF%Y{V{na z-{OiyPZk@QfLGqNxiV5RFUm#q|44ob;#?wrO{-R|MjIP9mW`QehIBr3G88>A6fx(` z!pqyQz1+I7=p@hw2*64N@~>aS-xQJ&E4OE0+0vTkANR7$V&2_?%3P_y`hv6&6km6m z3r-jtnXLL@*t_{5shcwvK2xHI0ja^2QNEJced3SF4lCX>c+W&gOxnL_iyOVvCPz*8 zXAA&?!VNG)EhyC|7oBI-Ww!7^!^s)+hjou@c8z$6O4GT)-#h64>NC?kE9?RP$8 z%XvOct1PVg@1z9j%()cj-G@#>M#3P&Y|*)L)W-ooqcDcZ>~JY%(@dQz80sHj4Y0K$ zW(ike*5KggJqj{fvqOy0PG`(lXv(@PF`EF4auC41PK+(BcuY{xd09MKSHP{bcj;We z2#N(av*r|eL);Bn<023@jv7$qW(552iSBT5ctgD%J-M6 z7OWlCnWU44E8X9(zXd;DRfKCzBK4ryMSTn0^r^S3*E(b%lL!2XU07(z8q0`_01?{f z$CW>cqw!fBp5;Sq4DF%x+GkE05?Gi^+5f4|-xk zBr*Ube?v5_Z>AY-Uge$)v*D2OrRseu$GZ^*K})iuw^s9dn1h7zJOiOB|LrbIohcq$ zBLY+aRI-Qf+GaQ6)*XkErxb?&+qZ-u;{~=;l-|}^9Ywa>GW_}Q@TmOP`F!8gxRnVU zdogvAiD_3h_P{-g;$Ys(Q9nPNhI#vTu~Lw;%8m*(`R$>*BMH8unl1j4i6oPNtvknF zC1FbWA$9}c8CzF=4efPahD)+fY^^5omlC?k|o@1*lZr7RKuKsITWMn4WFgaQE`ZvE^NIa%)IrDDW z5c!~kYaL!;;3?z?W?fP^K3>+nZ(k3-Seve0r*M=IH&tv76TUNo~{_Zgs9 z(yY-!_n6B8c(H7-gnAy;)C$D?(4f?VkD!O&;N;1Nl0|axEVA-Q)`iq>X|Sa0<|~_i z7@2=$MKDwLLl}i=z%0Mjjkhv!!E>JUy^2s!%Es2jH5y1RJRUDw$_OwqF8f7~0l;f; zA`H~t2RArcTp_8yetlFyz2pufgyp$CYUbOydnMOjTvldK28)1dutmZu>uLr&&?mfS z;K-aRf*l1dZx1yRYKyNB8#v1Klj|7%Q)RbWziLEaD`1plnjBo>z<~p!QNQ+P6iN~W zgo=f_^T-Pey=kNiohvgON9TTS4Ob3XlJIzW-NJlU^1xxNR%BJ#$S%%ooHh0Isvt!Hfp?$zdo}d6# z=mG{(d1t?f-UO>8lmj7d^6RHJ`TnaX0L#xMx1ijL{PYVHGhY~DYD5Ld^xHutdK9h@ zKfjkWHM@8z5LSauZXDwU)@d84BIHj^A0C=G%>mKZT>OmJs~`lRCjaf9Yh>+enn{I0 z%2WX%7wkyviX`8s1q+}ZHn_lNA?9jyJGLkfP}t}hwP=N18?BuaZXs^@B}e}7dH0QO zdLVuSI14=jK2=I+_{K9@e>5*w3T#Zl8rMOT;-1CWFUCL0{XGv#E5!pK+vUBg`J0bh z>cqEb!gX58N1n4K%QCcFIy|t38NAWjHKmm)mI{#>MaPu7p6jHziR}s`#)u>k{Ip)8$R41MG~2T)>=}V`UO6d~wy(RpMGM_8dR~ z=%Tcaeh@=A5cU|Bf%Gwq zM-W-Tl#)z_j(+i_iduFxdyEwWWVoi+6A?JY)-vi<3;&{Z!@q2bWIXala|Z(VdT7dA zQBH;z9#bi2`K{6u*Ns6T^AHyPgbD)^DC zafubcGWI$9(ziY9ZIL`s&5#ibfT$aKW$m#|wB($*OJIwXN& z%_;TwghGWSaOp}~X)>?`ogvy!)YkszMS!TN0%?00dy6w(>2P4h1* zOV!`yK0`8xJSYUjkYN&JJwz|-@{ioGQVlQ$8H@t{xys}o1vP8V6!~q+i=<9_AL?~! zArnL5s3k+0QT05Oz6tED!EW=u;KI+6;))f@JN#m z|0)ZK6J;Ym_~@{;olb}zLi0+UQ|Dj7s21y4z-pmZMeoo&C^P>%1N*wO58~V<(;T?s z_jp@4OQa{fy0tc#-*Ss}wVzZ?b;bvQD?!OFTd}w>{|{O30+wUGzWqNH2_X!!k4acY zB1y6hGc|<4Bq_-zLxfT($u40SQpPSt#y?5FatbtJCK5>*x~m2~ zrR2P~;SKk8C<_Mwm>S#^jcTs{k)p0p#i+V*XgK<1q+v{A3chPRIZ?O$!@{G$3kJe} zqav_$+fH9(H;q7qNa1&QKXx0F7T2pc?nA6a&!s-q)ZZ2VF+H-)JP75&%a>b^gH0hG z3fwq;i7bF3Z(AYcB`_?1ySBAt+dh98bT?W29UT4SsllBqvb!+{PL`BnS#9W=uNON0 zY&vA9MlmagSP9-6$hGhFosTlVl0C&SfCM3q{>v9w=TLml`le1eF9%_o2EgqKCCEjbZe@AWSw0umiFfVjb@N z76a??CtrhtfMEZ^RE(^&Y7j|)bC>gm&M(Or^m3|sDs!r|Du~1u(<1RpwV*??2}$%8y&71D@vaTono5>UI(|u?Gdfcb@W& zC1;0$Y@``kNASY&U?5`Jr8~xqO}kPneK@#^xT_>y+5E`mBbfaeqJ%szWF-p?hZQ9} zdLhXKF+RTgvDZP_cCAjsL-8S|LP>}1aHW#lw{yg2U79exmE3&vkq zz*R3K22VU^IefH8&V;{o=xh|lCKf*iI)9R?P@t6 zen<+W)nl_wYZ?_3FLMyz-K0+#vtXt%M5ng2oG&jfgtDt;U7ViwXo@VJ*x27Kid8_r z<}|GH5P%A3j0WO-J%0rng+Wlo6!0|rZg1COhpGI0YmGxX*{&L#M5aXWGR#}f_Lc!s z0xJD}MZJeNUPHxrF7Vu&(68?gKfgO5RgIP~2^dJ&c||=Vk2J3f%|@x+`^zWl+T{=*$2<3v&$T% z=PeqHz;`tn@Z4w|=R@Ro8R@%^stAe~g1o(`cuubsF@f9Usfu&6CXPonF`Z^09YBjc z4*l7>r;OwGyNl|gljZC$I5(~rohIo}U5DWZ_@Z_!uG&=-M69M(IKk^v)^BrIq03}2 z=HDFz)Exa~&1vAU9xE@q&8eeqr8n`X62K5f){8s{8(}V+B&p(FQH5M9Jj>@=^v5q1 zEaAu%Q7q?XzuHF_H`z_Yzm7X`cDu^hZ@Kir$5E5=wtYuB2I=?^gP7;ha0E^Ya+&=I zI7Ez`;r0hPlCfCGL^X;gk?wT^H824F&a53V{}$>67meTM6B2#_mm0FDIur3n6 zP+&D6h_8|9X~08l5&0<16xwkr@#-)Jdr#?n8;8NDWWtWF92|HLG=R|syIaZ0OojXt9IE%}Q!mm1htoDV-+Jrh= zbkTV2Eb_c@|K2_H7T=L@_C%OtHd|M)GOD*%)vr>U>|Yx{GiLS(I@$T@kIAt72VuBi z8e(2}c8dKB@hw|V5FT0CI9{?lT{wC*A0Jl)e!HLJ5HR!Cx}{ydg&j2LwRz9( zAM1bDo7=OG7Isd0n$5PH)U(}^=8%@=l=dHopPTl~{*R^Er=){fDcih1J8Nq0TOVGQ z;^CcmFvDzFdiCkp_p#m9WSsekHPx)0G;)7=kM7Orm`tv4EW^CXIG zj%qK$n-Xi=SY^$2T#`h${N(D_w~>pkw^fE=C%w8Ta^}}$^-C`tSV)epbn4(qVQB}h za{R=ruWz$(#!dHuFa$DtuG6`*%d-N}sM6x#eAit0%f~4%boa1P{_oic;B#&Ci_NJq z0cI=eq2`Dfw&vJ_Jy@D-q%8&YxN-i^J|8Gu&YTe~Z>33#@Q@T$)5wPh;$tG7Fo>n^ zjlQgi0_z3~29y^-wYHf=-z&bt?4U~~-zbr#EgLWo0~Wb_`Lh1~UeQmcwY6ws%fJuc zS(+^hLKY~Dfd&w8fFj!NZ~kiearMRk;4oxB`lNgi(uHB(0n8}budojLk~^c5g(EmA zv)>|Z2BalcjYY^U3%Izd@?8sdxci0FLJ$P?on`=&yg#A@^Q_+?B^#!)+ZHoZPQeS1 zYHQ+6sjWN8&o^T6M+#ev#?%l)-^~A@I{F{~u*;8UW4+Vew_FYDox84=DoCekxKKz- zZfmI8M-S8RuTderelKy>79e;p`}84l)Gv1@Q?;$d@`lF>MBVkG^am+rWjdC2{0Y z4BMQkW9yanZNUGPgQj-!&Lq=pn$~7j$0Y%)Ja{63x@8D`ASkw&{+U z!DUEKOmlsMVS?^PBu?1f=O(`hx1cLW4%ABC9yBP3_qH6%rRIx2auMj$A5dx{>NOHf zJOGrE-__1+*(VLYf_?KMuPqn{HpVZUwc==^cVldukh?Rpbkei8OL(4tadK3s=+q9S zEG#{95whRqR!zT{t8k79(OoM5zI^ub31;tIjWEi--&XT<4#q?35z%~#l_XOvdq?Ch z=|DK}CKswjiN76`FPjolh@WbT>rPxtSxH7ktA<)f6psx2I(`^eB|H0o^R;20Z|5Cm zjT^;^OX^7Gi^*Ra<=~~%l!h!_M<>yj9gm#G8?QC~<#c+Qn!9(YMT+){|5LwFRuzmJ zgy8<*cAIm{?!TEhY|}e8lj`OJuF6un2BI6y^p{)1QlcK!0QLkP(eWYUNoPu9yH3tt zoIC9*jG-X$MCsQ=HPyregP9(-qN$kL&SEm-)uM`;0}g2QqK?xiLoqM!C3%oU;<9^1 zCvqk_=!F~X+_A$MtqsO<4lAr#GqViTZ%el8LM*jWv8b{4X?=y2h-fl1_{xA|6sm*BEr$GdjT;1*ece2hECta39 zwPwLw=^0OBf0^!fZQ|V96J}#Sl?Je2ZxvY5Je-;snJ+%~e2QZ2TzD3}KjmjL{dZGA z`4A;-ianEb_g0bY!lHh*J=gDw=!|d&X={1%W~=TalPIyvYwB!9guT$YVs_G1H$d}Q zL+&oPsH}Ylw^P|RzaS?2D0|iIS01Sm^a4F+@aaSfnxBcpYGNKJd=8RO-GKE@x-(cv zd%I}4(HsoTl)TNDgG#l8W{%OSjMcmQ&Da5yO|jTk)$fj`ZI4f2L29;LD|pF4gZvT>VdMCby`Ue! zy3To5II9R`f@&dNn#Y2=zE{$#dwi0qamr;4dmRhxnGaWn%9gmaNk+Q!nO{Rmuhw+w za+(~x>wJO|UmM&uLUty}oD!1CKB((-D7L*Gk4IExjU|^ShMHe8nc_Mr(0FVypn>`s zi{iaU8s{iEf??p+T8$X0P^G_m3)Cn(jW*0nK2S?-Sp*dKJ#Qv8HB~=8%7fV+3VSIO zYU)&Cfxo$%&z3nxTby*$qg?`xI$nphs^yJ6fAfbN#u37Y&>zUKI5QYb>mvwCe`Uza z6#pYy!6CoqxlhA*j;r+-4@s475sj(@3A@}bRJl!mi^dks((d+k!(4 zt$F$7KLv|F7TilriAm&KJw_HjK~LxMVFUi^6p`MbEW9>pnMfBdbrwRT;G!F5qcVL8 zd^7Xvs-9y zyNM|z0Fs&Y2a^Va{Ieh_C@)k3PMQpi2FJ;wiA(IZjHO~MHj*|>8u3!kS-Bdp$Vr^6 z`K7)sn`(SefmlH)hMhGi*st+v`2z9Gl%HLhV_+lrD(hmO*pWmyY;(`ur|^A3pt+2>`ccS?^NT;+iI4n`;-a#G(2!5fT8PkO9_OF0 z!iPo>jYcE&xOHdRnuRpw*1(7B-RoYR3Ri4!D!s6Zc!JqzNs(~*NW1h`M4;!tnd+i*H@<4rx9{}a4z&K7IU25NtAWvx8NE%q&n{_w*UW7h?7zC=MQ7g0zW z+9$ib@U>D}-*a@4_YDf@~E zB2cSG&n1%2>lh%GwY2wUWE}w)vuO9z%O2-0t%dOtR|yax(i5e!N=h{;70K-eW;w(K ztwzwTSL;4!%Fr`)J^SRMh7H%%#y*Tx%*EyWM?CpAbncMS9&NYyH8EY#C@3azXIJwd z<)Ls~cqkExF3%#y^WLF_zBPV6B<2IXb1+O1Z(+e2=l53%t7_g*fI%fN?YQine!gyGLW{|L#qpR3s-HKQ|I~2RI{7dF z4voq#vK`=B>WR+_xJSoG(&@}$8BpjCA%OJ6*Q$ek=8_YU(@$9dPjc^L3lkfwi*FRi z36P}=UzmXzaQ;WJYn(XDai>qtAcnbm7cP#3)L9_*56oZTvY^QKMNW=qp?ATqpZc(z z4}8%;M@xTZUVmNfbIWo^iV}z2AF89Sy+6V4rD7YcVwY>IPc20sc;8{0&F%Od+Zn4A zFjA-sk0VGq(858_n^Jq5;BGg$&B)265$_u;5bhaezU16fV{{kE?^HCHC0+*SXN&KI zn5Prt(hXk&FNi$-oMu5v(ej}x;LIk}Zs`m0!#FtFi+Cd2|F{WigIBP)Q-0p-QfSVH zet#M{qm`vya7bkprOAGDZ7>pVH?GCJQ>xU2lDqLF`%hT$a16av_nb)8OA^=iX%q>eJ3e_*_C_Q!5F=LFK@GdDg^FFdDJ z^X4ovZqr8n2G{mJA|5!zHFgRA6#`hFMt5-|F0s_W*a*ky(JocL!kE(7h!Hwe-Gg2= z1Ls(=cyo+CxUA<+ein!yi0 zXeZ>ZwGJC2A$(U!8P2|HR3G9I#n*haI2iub0+enQ(?bKi9nKr1?CBrkQHLD}jZiE7 zrv-(!)kGA9FBmswotgwOF(BfOIvBI$7j04aio6{7rPxg{l*ZvCU%hyzHwY)#_gUalr|5117klK%!?O@tTyF|hRih5z3?J~p6(2cABrPl9}iEA0O^A< za`|3I3oUI-#`&ZA{`K&6YA==(ZuR^3nh{JPD^KmO1#_kmRx^c7b zUef5vVoZut7D6HbSy`NV)9wek)%v}jevjhr(s0y1n3Yp%(JpJ>KFlHfRGPW>rIT8w{PG6M{@FHC1~L>>P6dY{vUI~6b5GmMhetNm;!{m$gZl#~A837oz|_# zkWXa=$?-D&W%+T7hhd3U1rPXHL7?`G{?_NdJP-a z$x^p%LRKYK_&4a>Kh)GwJ`!a_xWM?Q^|t2>s_&Yw8*r0qGwMLrl1y&Q7>?bvUxT|d z45KvhFDW%BCP`cP$jxb&T}arTr;ZB`mNhk~M8~xE9hTrNObT z#m+_Xc!8BvGIq#)v3fkY%Oq0=Lx0!v8CsefWrYL^=3^fmrG086m3)IJxHNt0<>gK; zRTYm|yh{On9$|3Fd2S?gUr@h{5wuaPRj)gjiPV-Kw`)q*`cLWS`85gC9xCQz9Htv_ zU8e(vPvUD&UD<+U0@8=GeQ0crH>v^Z?eBr`L=j2->1r8J z?Rn~0BSvDf2DR}z!!m&VO}edG$Hc~2m*<9hJ$?vdP)ssZV1&ax5O%{(7F9miX0YCw zW7cu-Z-KGXA{2>-Re!!1STglM5vbZ+#Bk`ZMviZuxUX#qw}u_AvJyrlaZ6Ym$#6Xz zQGOV-dQL)q+FDx^QsAt? zwL=_bz46ZnYHvgsvxBSm(k%|SNxn9ml!|yXwN`pS`7sTtaB&Ms-ZOti`lGIE`%<;O zrVz0jX$~B)ezeLmqqHq>nCg_5ADp=X9T8(^!8PMnLvh%0P+K9cZy3bfKCm?0o3pGx)TW`4` z$d3UeHHVs>3F|eXJ%4>8d*4B`UOo0E*HJd#MN^OjoN^QFX~0md!B0I!DNKF+H{|lR zi+P;F7Psfn7r{hsXws$t&!POOQXyhLWPj87qAGDc2PzC_tz99oRK5pA444;GsTqM&3 z@c{s1J((Ec68y&|N*uAG%-&I?x3~=#pwp?-&%VA_J#zSfW)#suo3af*+moBfk_V&; zE(%$rp#W9|i~BZ3;b%JHwoz2MCs>c10Tud=8;T$Q>3QFO|NZ)>W#^o)*lWqA_a=tj zEfW6ukJvW5xz3T3JNnj*VWYXo1L6a>XWWaPb!ms}QiC_26@PN4RU&Pn}RPcQ+F zG@u#Chla;G=*5+9)ukS%QJhmfi|kcHIgwnkYOMWKN|KqE)936J1LBHSzeN6rdM;IsQ<_)%Q)h z`_AdVe4ef*%H=#D@v9MD4F)Q$<2nMY0^Shjn}f$J;Qx2%lJ6N$uRXjX$yi47@{wMj z9~nR0VacUTINw&#!s06|l@7+J-XHbLu2!_7Z_9skKwXqn-Q+6n$uk;B{$ApiZ6R$T(a}`y4o^! zfG*}H4|xC^JT?Uc*wxkX;De^r;OZ@GG2+9StSo7KUxXBHaVKf7A(~&Jx*^@oD*bTx z-b18pVq!~{kSwUyrZ54@W{ZrTqiAD-_uixQ0HKN3UOztox9MkRC|m2f(A@S-*?It@Ndc$b-2{$kQFNr%UI13gDKpy8DKbA65Z^5|<5{s3?@W@E3&Zse&@?CqpB(b2Ls zIiFi5zaEL+Ao_X~yw*tN7zTC-x;Qv)H66aJ)I!j+$odBUVr23P8MlGv(a&w#W^!sp zO`WV{pmb3$+EC7;l_~J_$_5B#oa=6a*Ab1Zs-uDmu3ULB= z<_+Cqpiz3uDaZ^lSZ}a)9Jqh!3OGWwsI%`#nL=(Gm}#E5m)090DgWiMOb`$_Pv8AV zW4$WCi5m!^)`LVZ=8Sq)xz`#BT4Aq{9Ji3ah@&_JysFXT(vyp$zo3)$0w&h5uihafkJk~h29-A0^Xj2Yqr+$O>nTaa zaSJBIN;5rqh$k7O7Ca0K+Xr7AUVY9Op8ncxrukX*^xw-=xN|yTJ|SiE^ZZ|V*{+?0 zDg}U(*9&x_-q*2oIDC^h(1NUh*6*2LbfKbm3<$PtI@t}Dz`-l~E`2*nv+`f- zUvrn_C;Bl)I?1vPmkNSKK^75@SqFh^NOmLj8@K1OfvHA*VUhEQZrYl{GNM_KLQ2Zs zy;q;3csPki6?yEU=X4Y8jbs@)RX(4J1-K;meUGme89e~GVg;kPs-ndo`B3L6Dt~X^ zKO!`pPY@*S4Tj#T!1+}lIbs!z+sf{4pHF=xG0D(0BS1w^n<^%9;1i;Te8c?UCGxV@ z!U)j3)sjlYZVLpdou1W{I^=PF zzb;yK9FAuQhKMOpDDcf?y<7@_OMk%!33H&`^P^`0$())Jt}c}==*CgR(gK8$W@L-~ z^_utQFYT?vfDG{$SL}_)rEM2^mD?lnI1nWq%CJ`)Q331MnKmz=-ygX8^Ac*1MllHWUjl#<+3ARn1IqOFk)nv!2|9BX>n zxBqIOa8mqFm%4A3dyR02Ly zUs6}y6@hnB{TJ%uzhT5sJZk8-_;bdoshgOh zI0qF;2)xa0KixPxIbX%yyeB%H2CO{Y_3d#4K&g;|dQR6vNCg}QX&&TeAIN7u#>*dg zQF4rm!wok(ZE6<_*@gs0@)wzD2J0?zJ_y(#8sT({5}Zhiq`CpZ3?u3E-d~~_Nbf8o z|Ikp-NQLCa^s=_k3&dR#Sn{YMO|UW+{q=+XuaVt}RWp}RW~|R?o6D6@5lRtK2>{3))Tewwd-))qCfa@g%$@uZ8*^PW zVnQs_dk*MmVcO&ZeW;L!F@@)-R_iUdxy`=mt@orO=Hn(L?$RTjjlA19v0CXJ3ATjOw{ELFS zo|JXWTmwXtD-@2ZEh;<$k$fhXkHiK6-Jw4V*}TjHz(7))iOe2=Q6y6oCU>RzfB`*S zvM=^R+?iR#f9^HsdKjB<3|5!%QDy9&y6}c%DV-}bGkZH5>f-f6R}Miskhsp4&jSW) zN0FH_e+BddNT94qvD@-YDUE}>EIglK_>WaP@d|~N-q<^A)vYUG-UqA8DCK20@);Z& zUB-L?Ugi^8&@)Fr2|*B%HPqSK)$r-75359gLG2YkUn_gb1QBJ(3Ik$W(9;4Q#xdr# zWIH-{gQp;&fgVl_bz-C4{j8yAhZxqoB>TXQrvcFk`RiQXrC^TS@sQc8Mx1goQ=qw(B*!p(rN zbXw!eB=)6D8OOj^Go0eJLsn*&RQ{Q+pruG>dg9$S)bwoj{E@-gu6z>_9qt=APl{s0 zy(OLMB$LaJVHs4MuF9Efyc&SP7=;{b4eo9mS;Ku6Ur?FA;o6*!b{g3vuppPNw0-+~yf14| zX~Tck%Tk=j12cW-jK$>?RYX+o{xvV16TDtkFDI-DFM-fx16SWZ^+eb~0=pOvGa7*x z9jma#L76KW^C(^pQ6JRiC%EA#svf>rnUrQZobiW5?1?&LmyFFOngm5ACNFH! zy!pJW%PXEnJ1CwFTaLyRtnSKJa-9pKB}1CD_`&4m4L1*)3}nQ|9dl3mllury)82(q zzKd5Ee@3I}`Zx+QAlIn!xOCyFt<2mTr=#nf%7Ap_Us>$UTRs%8ob;dH6(&%-0@8Ew zjd3*QeEha4rEQbvU0ZH`W@t+41fte&8VMUg5X*Wm(nF-?aX>jyrGQBT?b!eE0{`&B02jB8qh zB3P0+kSz=(A|EukB0)dE>^+uBVzw?Rfr!)i%-3?P-kz1cud0b-=G2>^E^d$}r{G=V zHE*#spSp9)7S$Pbufm~Y+)CjS9+ciP(*9VO%r?~t@GK{PH4gFkKnN1AH?+g|8y6i| zx&&2<=p{^+;aRppC_jGH3O?`4>C~Stm|XmedpD8Y%=CMnEh`;+cKHU{r#EILyh|Bfe4B zUgq$Oez3c+g>u!=&Ay&0UxvF#w}0}lfx}IP1EONW3Fn~XaIu8o;Xm3@#*uqq=DhlS zMYZgL777wJ*V1hwoR{=!!ut}5F1B~Hwi>mHE)RZs0L?GU720SuVslz}?B4{w2c=0- zlezkqfOi?Vi8c0EfmtPEQg-Le!^KpHo13m^AMS!Q!7?mkynsgR1A&-p1!kFk4dL!X z;Bo1zdbzxT*{^`>JL~rPXYuVs{SF$;IhzXcBL3CP)i9cPL8;{8oW&PgVFd29W|q z^60g!g6goOI4EslXTUHKx1uzLNpaHc)SD-5{xjNGj`Ca@0*2z*`x?$*2wzYI)_g}k z7*%_O%ST=!jYwA+0C-kVcug!Q=tuw+C!8y-tv0Eq@xr=8hL(F+^^XyvYg}}%WAxMo z0YBU8u{Jz_uq%=}&P4~%gn5nc8rqzpN_s;3v`)VV&Ivb>mv&?@Kt($5Nl04sF^T$0 z*^pG5@W^A%iK81cjkHmrh|qy{-w8BT;AOFW;^|2X#m8yyGT!USv1rD{wHm2Q^dVBS zg@ga1!U1bkyJdZdUnR|(HVr2%scf{eyJUS<6iqsMR$6tOITZygn9|Z07kJ&;d$X#C z4|FfjVGR~zZO|1c@oE`Pu<2jQJ!8^ShUPdM*ACa4X8KfA!1)R#QI}&s>sjqJg>|9D zbTb}!|4MV-G>JCYKQ(z%2K)U+(>4V8r`Gw60>X_r7BYB|1?9&y`(O>)Vj`4RwnqoJ zs;>O)3uw@hky8M|L19oL3^d`{V*2#y=bo=~ud78xQ`PwXK9e}fJP$(ez1x{XKSB7B z(XcKfe0I|VcVoqlfct_a7Z*)>-|0A6xoShrQ3GGs9D$9gZiL{nE82^g6-v4D6iZfe zz1McOp3QhZN}fIX{;uf->?Yx|8q31Yb^ZfADYs;8DFYJGCk@z0Hb7=`#R-i+cZW*8 zszz3>GcEzj_mN78qHWJRlb81K4lLVhzT+2#hJ3!U4T~Jy#_n(I21CA&brXs{7DP)! z#E@=rI$ycZ>n|YCmld14%xAtDR#1=U5CBl-r}97irbO73YUUPK{*R}N$i&(RL&G#=uTo$)H0?eaVpx|+8yOY=ln{#*CjV+Q90 zC*EcLi=NI)zP!K^Jh2UMX!_M_2#klEVO?Ma$xOJbRmn7K3zd=Hf#nA%N6Qd4i*?cRGEf9zaP=zv+?Q zUZ?M9-@g3v^6r%fD7;HXI{JRt)b6+ju|`9xEf1Jc&b)TG7vGz-IZyXq;~t+F5IB!= zp7^Sze<1<5?F5H##>&a=b8_%L1EOYbXDb833`dxUQxHivIeO^@g~8gA<@dN%IHLL& zZy${u<$s;jC3++GvMVKq6o#Qo5`2E%SLp?-EK`k?Td|(AaL5&sk=cTovL&MDUQb*J zz-Go!eRU-C(&KA@Dj+(f&Y|t2RlZAYYefo{sk?#Nx0@YtXzpyDb8+b`5zMP@-MWRy zP2?T)+cTB17V|{3$AD!s=cPF-StNbKibB@{DtLl1zVO#0BND0TsYK_YM}XVy#stAq zzt5a{F^VgZ%_HMnfBd>fV6QuzV$owVxjTmxpT0O!MbyPEq#P`B1)Sk`j>aprMi6P6 z$qHlpu@JHgpoe7vkzgjw%5R-x^xNrsoyfyLL?Q&|b#x(eN@#|9PZ@xbsKeh6OPm+_ za;;_RO9jY=f`P<=6m3#aKA-3PU>Miwkv64AlvNYud7HcAb=wB5V&EC0q-Tw8C+qki zn1#nV${S3dG=SDt$UF>f{_pz=o7%}@S=!O2_<>w7xWhbezaRXLC9U{EFc!lT1gW&n zikVrIeOwUP?8LK3Y4a|Urte`6Q=AoN%rM#gyyfiBr8j_Hewo!Xve57pLG&nvnRu`< zqj5X$@++Op>Nbun$=ktu`5RLlJ*Vx%`ES&f$U3cQu+s{{c)*vA?DkJZOJW5CVTQDU zw{m{5+xcq=QwzN|2!YSr@5l&)kuH{Ld^ScocNNA;6=$97eDv4J0Z6NxvAEZW&9p!Q z_oyiwq)lKC-~m)Z9Z_*MM6YLbb=^Ae_CLID|-b=V=gBdCAjDS32&fe)&SiP|nj7M636cK67!cTZS_yNEIc+X|dh#!7P>SXIk zqlaYQ@>!J(_-6ijKi>(-wU8>ZH}%E-yJb~vF@~W9V zFLla+thEq2!*P&jh__e9rb^1SV~mY1a%&}@?QeosN9tfOYBAzGXEAT&p)!Q?<7Z+N z=aYBj)H=`?}Jn?=MFhN#ZBcUwu_>HkuRFARskSq}eFsnf|7^znyo+ zbCfA@BRu5W5U&~Hen4$-oWp9f`Fh4A$G+|+b_uyDl(lR>zS}tRil8bqVzlv2rF~GR zxScTevS7F=zcCSzHU~p;mzS9=TVt?&UF3_V+4g#d^fur1nW7qM+Ij$)0GVXTTgQp= z4ajW{*6U8Y4*dnCNP+OO%>{XJId(aUWeTf9!$ zx^C6O*oggczw0lJi;GxI1tiic=2{ohkxSbsgwbwRpaO7+ml_3-IMv9;F&C2=#({h6 zsOFeeLTrD7D%%$DV7aT@wCyq1zdr0_JGrE*c<#9_koZXWS;qr9>se|S=`f^sLg#>+ zd@cbOxU*Kl7Y7d;c7UgLj(<+k5_>z7|3V8IR(VbKAIsbF+&$D^>^y`>A)Z8qrY8Gj zT@X1(?kR_>BL-I3yB7^H{?t-XffCsPU(m(9OFZ*dMBTfK@5=*We6|_}TM{pyDcuaR zZ5{{Q7i|lb^ty;7-&3Tyb4(P7O~VD2dAc*8*Q{k^h(+F50JLVM2HAScs`sz z2JNJE0gsySBw@2eSbeTW+s!G-V?IkjVL(OW@|8cIC%c24x#~ITWY=~(gK%Csg>n!x z_nYO_!Hqlf(K!qEBJTHi)3a2J<@r7xe8cPt<777u09$$r{x3W4yFXKu!VY44h5&)- zR3JTJlaT$&ATq78bLs5GpFI$`?N>&m&|e|O?Pu>CkV-o|ftF(0)wWkMMt%b#5#JWh z#78FnMY_u%T)Rmux-5#0e?J#ah*654)|3+ZI;=Pwni?FL2)UGuoZ(49m`VVUo;`0Z zPGBP)v2Vs2Rz$VlgSSl+=u z^{4H+4U*;s`$FH!enGgTv%~Iiv1K$`wvmAT^uVBq>H#i@J6#A*dyJZdU0v!b#_Tqw zPChvFCU_bHfu@IKE2e~YV2M!R39%N1NLlAUWMo4`m&yzp>a>~+XIjlB7XgM(ZewBE z;qZF@@pFXu!L`2Socz6nc#r?XF*!9 zyyks7>?jXbQ;o;d&Zxhw2I;f4Y@Q=)CKVdYgsM3EAtmO^YD=PLm|Q0AiVCf z4lq!EXjmsy4wN8l3Cv8wUk|MpmqS75}N@%%dCKYA$lXYofeNJU}8w~7_$ z7kfp))<>d1&=y_^Uqu{HRb-ddGDrkB6LvqiJ_NElhjgR+D9b%u{y1e2xeCm~IG}CT)sM2mxo&%Ucd} zl@vlQS_qqiqz;f2Bw6VKXw-i%o0py|!r;IuYnhiA+iHaW?L5o0OXAdVV&GV3170+M zqjVR5jmYrnd8Q4{O-6IJw3q7anTINJaU(R<=x}%woQbi=fd%^3CE~|Voc721wAf{e}84$ z^VzFQBgPg}T^6BpJBls>6T5Z^xxlhgkMU>1G!t^$7b1!y8w^m^4A6Cf}()!&gUdA=441>A-3N;vS z0bMF{^RkCSlpo;VOdxMziZs%j;JwTI1qux3)pdkUks+Psqc;3b#{|DGzSm)d!-jjO zy%oM~T-rhBd=MuSFW4y1fFS=w`uF8yn>jb+-g666Sl9X&ErnPhs`h5B;NL2k6u4>* zg!Kpd<#tjYusDiN7vx|nerb9h%@d`7u(l|si8#ivgoyD;PHCuOM0m}C83*{KI<3`@ z^JM}iCsskh@BCd`Mu8EF2OfNiG3$I(b6Q+IF^-{JQW#3#+npXD+>m8*LIFN{4$><> zai<&a{$B^=n^$!2B9(D_u zT^FmvN5pzee=fXu#Y!MVKl`Wmf7onMcMG5zb4h3vMvW{hD>!5iHW{|FKzhY!hLDb8 z>&i8YC<+1HkS#<+<pN?yGUCp8 zi#geKa-@zA=R!<_J7Ul!$lHZL#m}o*ix$VgN0Yd0U2qw;{e@_?bd0Ibj=1`0!kT|znJk8RQHXoFn#Xuwv= z=Fiau|oWTlk$G#MBeh*kIDUeLyDWs>0b(o3`Wpi4ds9K0NLuUqdy z+P4=sC-lxF|7c()prOl@kqn@{C@n89kGwL&Zj+1yfbyTOthKYm$y|TwCV&4aDEv5! zE5~SMmsOqjtMg&@W!#~WYocD{9@JSj=^&G#=?jJ52P;VmHGh8Px|6O5&5*Ouz=q9F z`*%&ogU~*pj0i9)9-EhyGw&~8)=T>@KdjhCkj=wOiX01oa^Y(|WjuAPBm)Svm!9zr zPx*z&2^o~7hJ>O5dLwQq^{wyHgG!pW7hU1z|R4t{RZM(~W z&v9BXSy(T&4;Z;Jh`LU3yKwk$jgn6n3w(5>nm0;4DRR`LcM&?4=9*1XvzlkfeXK0P zbsEqaF}xX%8`NrMV<<)A+OmRbyu~mi-&wi~EF(|dHE(7sW@EY7j zzDdO|y*xn6<-&aFm`&sR&$=kD)o9Jvm6F6p9!hEUhla?kWGNRcSzW~Y1)nz!U}O-O z-VLr-Um21V4Ikx8l(h(j3%(92DWDPA$YnlD+ZSnIT6B(-$tat2H?oTGK8nO8{r}^R zJD}*kkpD?@B*uM@pB2c&2t5+SqrD<{$$I+lHPQYZ1GM#69l)~?$0^X!0Yy1Xj@cxv zB@nl&Z*xHqwWzo_?RQY|-KG;3zOkUOiHu?r(p5JUVpJ7%*O50VL&PyY zQNy`rfT@SjmN(4oZ`@b;k&JPQ)EK1mwc5m+-UR7k5A>&SQl!6b%mt^2fUab}N!&eL zdzffbya(;wse@nBf8JwqN@1(AU~wc(f|vJCZeE0R=Oe)I6Xdr!YpNVZD?_V~2Abbv zLxHT(A#Kcs-BH=xH)B%q`?dA8qM#|@DOtbxhXTnFu!B$TH>SyT&bYlJ*cKc`8oWDn zn_lA2Wr--isIq7g7>5u(S#)zCxWDb{*bumPO}IQ=tE=U?b&)goh6KL==|Su}$H)=} z8gCsARsa?KFHyDw6;`0gY?Qz2#hAhkrIKnuD{iUas2h1Ea$GSe$Mro_|s@LB2I|&F-b%a>jt{v(Udl9Uy6?7iK^>on0bH?*qn%) zIp49ryU`k6H?djDa?W*Qg$RDIfd7TB1=cL`a#4_QVbz;o!%sT@)fwRny_IiKGwC0N%-AmM3l09%C zU^eb~Vjo{lu|Ms<4>M7;Ym`Q!>{8hnB`xt<$Uc!jTpUZUnz|c05I|(JG<}Lq&639( zStzi(;ll=IXqv5$f7OxQPx8YlxFBl{nAqE^?CwFRDnb z?vz6!_m6$_=Ff8(OZWq%r5j8raOcu*#8JEbb%{$BSGPa%Vn9H^N1z{dtZ43N6iE?f z<3uQ*e;_L}t7%tJqKJpc6h`Iy`9`-~#l`77fB`pfJPq3(D~DE%xK&cg#vYs5tMVFC zIJ@7hF4VjF;=e8>yC+<2`<2ZIiXLUi63an()EdU1+aeW}#iln+l16$e=V7n&dn_k_ z2q^%#vgikBG=P(Vm7ajCMgDP}HQTR3Q|Y@>P#phAu^Clf{Y*1UqT1Wggtk6zGRSrx zafi5RVRx1~KIY$C-|Ouz9+nG}nXT!Dc(^jtP9FQuEkSLFB?5o0AN?|-tP&Pu(As)m zz3?qntA>l<7RsO^A_k$4TW?}aJy9+gDpJpBIaoENgV317BPR;&MV`xy{p)d&Qv+G- z&+S&%pLlRuldBxFmvAeiq>ZKLro>)m0IW zO{p1AP-$|K_al;ed(t^ z2aPSpY*2^qI~T{Hbq$v^xMHkmE`FXG%(6m^$A5+*@#d8tSjA*Fg8A3fKpLnw0xe7nDU6CDlIl!9e65tWRcsY4Ip)M|7 z{~o8-R-I```NgzdqP7uv?cY29!|!`E!O$P9o8f3aec+snpPoVUfBEHqHO7_wSjd9WaVluZZzaPEljp$4U3pi`J(B z*$_H*x9pRH0fzN~!A?}$GJIX?$TKTw>C|#Ddy=1^EES6t{1evuhfXKVqSeQkq#W>i z>fzDr2d~u>=hktquEZ@YIClPvtYcX~6p`)fSlVY5-~TVr*MEOqm5e4*<6!qMABG5t zBN7tM4?m7N!s%K{cIYZwXKfB$GP5Dz;0iNJ+*^6I8sBxH6N+oQue7eAR2V4|ihGI5 z%H9X1ieq71G*;!70Nq=+Z9BiA5aOJGA_ejAx=&BIf4wvMjyQWK?-SXV2rRxbDkPsU zU%S{Yh{)8E^E<7pv&@wpbq!D)_p-9YXcTu0YETTLOvE)4%2(rmimc8zP3GRjz6L$RxxG5F9#DLGnOB`qoYhHy)Wc>-9&r(XJ#8viWry)3ndhEp@t z3P%jOPyDz|XLx!$mFOlnj^XS=Yl0;~YoeP3tp;ReI8{$} z7n`}?yE5rjc{SU^K-8>)Zi@M-x%I29&Or8}XM(@w16?FJ!$}D_EyIe?mHMl5?jAMp z$2oh#dtQf2)e&;v|NHYT1v3ZfIsxK9z?sidZpS6R9Xoau3ZVfYCSVx*F{Wq#q&yl0 z6KLCK9Z|~kX#-8wjryJ_C!5w#?aC~w*aJOY7M+xjK94(zhl!jWkbz^$qVR7uvdbJ9 zDN!$>c%azlvm_;-oZkC~AAT6WF;{u{9EAV=JV=&RkwX5<`Q+)@RU{Kx{StERPG*CD zGmSsmhYD3JAVrle@%h;Z#XC$s@2<>HWrFE8$i_BClca{?df^sLReW=fQQ}!Fp_=_g zVC3OQ$3g33!o`Shy;D%oX{eNBq+Y78WjVu3+?j`N@bXwW4CRMi9-BEnp;THA^Y=SS z))U1e|9@ZLkM<~!1Odjkn)5LsD0zbV`Wub!lS8O=S((c!1o60qwOmyc>LuD;<`R)J z7_MHXoFe7+cs3Y5d?(T!9v=nJNZPDCNGj_#qlwL{uI1<#iK~|I^fZ$})Kea5Kzv=x zfG;IhBCM(4I?vK5QP%Xj^#3@wX3bB+tL{lJ3DjAaY$*57RNdMmhWBQQ4 z{J!uBC>6*@cM;&^=cDwa`0hqKNPEls<^{Q7FXmF41%6Rg1I+UmjktKx%V?IRx7He% zcZs)fJY6k2S){O4jjMSUFUU70{{v4o6{f@bH34oOt&)uT)AyJ2>45_9qd-%~N0Z$^ zylz=vVzX+JvUTH0*HL~4soxHOT%KiVB8O1vVO$y~iN6Mb5%lAL1|=orzmc|!M-{K_ z3Cn?`_!LuwjGE&H{`LPR9GAKi;{5u5e|BYN(Gl!tN>@yr2uux{dA>Q>GA(>;6RwUz zPm~-4Vjjq8g9(xGw$s(lMAf~I#v`(jE>NnAz$>Atq!eFZq&F+&g|U@%GEpr0rXyzJ z%1^D>4N7`drmyE3FY!sbWnd4H)a697{(1 zYc}-XOPLjrH02AyNzUs(OdHdE+%}@>KE^$#B!7Hhp7V}{A=^$I!q%3@@b}>{8YT7s zc8X+1<|={b$8yVpv&!6{k2fpRkgDG>#-Rqe{-ff%D$jhE%UC5{y|IsYGV)m0ue@}x z-x*`2&#(lmVC{8?$r;=>Pl{h87ncFrbDqOY*NQnHA4?qT&}%aa(u|r|8GZ?55@RGp z!{Daj49&Oc-o4cgBsEOXCYsdYiMS;z;<4j=>uu5heqsCWiGPfBSzi z8nsa>kMRl`^B{S``3|I)5x1=q(mo5#XR#f zaQzASg^cZkBo%=Qk4{so5+o6J1NXW4*qr-h6b)=rDf~a+gtwUM65rOvMy!S+6g6gR zD`wpcsxy{MR;%^<);QtcoVV+O)914xvYZ~Nm$-JE_Nle$9`d=S&h%?*4rSWzjxC!= z9V&1kp`<~sOo{dB?X!Lv<=(Fipo7PKOu%= zC6|txvB_wmbIaqt)^h#Z9CVZ+2k1iC{T7wdW`$_756p=jH#-FO=K2Ur~lEY0$h zOPQfdYW$2}un7U}(-BU;>DONxF7CJ>6PPPB3-@?sRPvO3%se+mqH_~$^hf4rP5L-kC}&KRlEsvjgJfj8VL2f<)BnOvM-5*$^)TA5ZhaHk=6)6 z#lQii@XG5Kf1|RK-a+`al*sKpVbV1So$e6|s=#3=weFSc`?<-Zw9#pC_JQLUhm!15q#TQ{bQCy5O< zjQuJAkm&rjvqgcO7VTi}_a(210QLYbeWo6L@Q2?IM4HICnh6l*6L}Q-N36aWJIfos zWG3GRjBT&|;kvnb{F4w_10W<70bTfaRFWRRm13W5c}usNF5wL@IwHsG#Va5;|6-#% zd`8h8JKLosf8oQ(b-RQ}LEI&}&(3wW@(pGvcJ`d>U9+~B?W-d=f0QvFct!;@7Jd9V z{tJ=cD{+10oxBQoA&AM_rJnMl*x4OsAPqceL0!4n1Mi|@V*9v!R7R+G5c;soj2E5? zGKPaowAS3XFZ(hNQx+XekG9X~baGPZ{7WwKK}E?W?lnv$kKy?dmrGmt4N@*%>$oSO zNVCMZ7qK0oXRUtIjr(IS&78;Itrs56TQX%J=VJR!@fnL(JONk*CcaG4m7K?Eynb`C zG>w&_lm%c~P7Am;w_okWewrxt8g}DdK+>iY4vH+EIwUCFO?&UwnXdhpc z4`OvPaV>9A=ID6nr#~!rzuM(Y{nGSeP9X+pdryS<{Ej$~O`Zu+6)qn+@Y7`V7SsHc zip$EMNO-sJ*#8utg&()(1Dl*>m#u^+nQCJj#Wuz{X0P~CE6f`NwrmvOcDIHD$5<`e zgDGV(QIgyabT**oV4-R~RwUYEr3z{NQdF9V1L?ZXV}?$J-w-An%s0Z_71kM;nx&Tj zWI|qwOS!B(A&IrK{NlI#e3Y?I{oyNd`4c&$q5)~)Co@?5#07`qG{L9)$WfogNyr;? z{`!!bykxuKI$Grv^r^UG(=yMmd+&Veu65D#0!lQQB!I@Glv%OykeIW*EzLN;nJ!zS zY3V`C!l7q{o>Fyaafm~CkZSXHs7yby=ZD zh?+U?lWD$}Y-JD05c3v-0SIX9ts)K@NncOLGVLPvx|}?^c@Avq6%UtxQMJ=0*-J{7 z<15a9z3nxhSd+Cf=v-i<2rhU>5BfVp@$>Xs&|-|IeWTYoN|$tFlSYa2ms*OHd|CFZ z=UY!(AuApsMPy(|Mowi;3OMYE-Lny~Aawn%yHtAd8U44uC*kBX#5e~KB>k{#?LsXf zy$<(IQh7PkDbu3u%{Unx2DGdSE42bo0qUoFw90p29UEnJ3C@1peLB{dVm3(aw23-Y zjAJ@~^;J$)Pq(6Gjntw4_KI3}brs8Bz~MDe15v%fmQp|AlLK>SlW=+FHW+YtvxyXN zGh-LGTkih!)p2~7Dfg&J?@*#LoyXmgA(%?~7csDaUEv=SAeg!nccu*unLOdg6C9BEueB_qXCg=z5iJ^rMHL<%PmXz%^H7T08SJU&ki^`rx!RU zcVj?;>Xur!74qsPuF;-f<|+)ASk*J5a8>y%J>6bksB7Rdc~ks(HcqS43-U<$K*KS9 z=2OgG9%ZI?9g>tex{Uy+LwsF5R}w8`zKi1i6t){QAN$FL>ZgIp#W4@L=c5Bh8O7`4aAh!Y=LCuq zp30wiL!CioD`kTK33E23QhY(yGg}=7w8Jol;XfvIKVF`0eH|61bo_3aVnr)u65R$` zJDJOuJ7l1kJR_u$(O^2N=gnf&SiZO&D&8*7y<*im!j`yVh-ow%S)QdI>TXfSTcEMJ z-uHTQK5tf3(g8#m`M7=ngB(R+BLpqTjR+G_Z9k$fk}7|Bv;i!K^iIsHvAMNBY{%MZ z7jEy3NZi|h*6{`^*Lz65^UTi_9P%?YiU-3;WwUjJ2A1TOZ0i2o(W#67O&%8C@xOKMsI5r> zC9e&^&1+;$G&Fy4|4620f8@g47JQcJ8zU5pNN+gUNc?SsKIgDPY%K3z{Da^O zRPbnoMO;AP1g0oMRc`$=8a7VwvMoOb%^HB3Uu8r!IeTrL$2$GcH$~x!8><0@`I!f>7RZk%S{Je|b|(#%$AxOHNVQtW?Kx=rRjuKs6j)%$H=kBsx<7u=Ar9-ytaTST?5T&xDCTI8 zZYs-jL6P1fE|l3d<}B081SrIz%{ybqdvW%~8)E~!_06wIUFKsk-v!K)%LO>N>h+&9 z9zbW>)ywX0*UrGyGQ&fm3z>=f9ZAT=GECGqSK8R{B`7r47+4rT81mwE2UQhQfdi) zd()Ut{lJXZE-1Z?bv=P44G0-)HrZTJT^J&dU$QICTlyx|uu&>8o@c$dLL&Bb`{U&O z1{r;UL0eEb9;nP9d(*O1!*oQ2KFYJ#>MHUoUX@FPxAqO$zr6hAI+LMTvrQ6W4wnz# z#Dz3)D$xsC>i)*0zaE->Wj_Nr6rVGRFZ6%{ra@QgI##T=aed)WQ1GE%U=BR%#e!0| zwoj0?e1QcM_&d;O-BLqCF|q)kS1os#;qm-`?`eCRsy}5kWHLFMbRY{+UG~OqdsYBe zX}3L=FN;&)fp2wXYe%PrrY7(lO4!9DOh8=MP@E^}Xgx55Wa3%Z7U1-X+NNn|u=4Ql;HA!|4e ziGzdlcqwvsecNkqecd^^lU{h}V|;If7hYlAty8CmjP|C82_Z*F=mp2J(9{O*A@s*< z-5M$*(_OUF#(-luSICoNdv7>@Hz-EY1|6TdiKPc(9Cl%qh#@l{(-_FkFkwX3Nw%SgV#08|*GN-_ zv!qIvfsU71sv-LRDbzTvmS@|zH)#kWaY2`UH zUf2EA70PBnMpn2;Vy7=+d7eAK6KVh>-=-BY^?vRYYR^E;ec8N2)S; z#EsN>w1aS9pzz%Bi!Q?=K`1W{r>;k9OgW8*qHN__kP4pI|sbND(exl2e%bdV4tI$>tYM?KjY6PO(yx>0g*EgiGe)*wqafB4RAWw^&k1~Bow*!+Ufif)@R8F z$`22F{_WdXBIX;*g@JHt(#~IuRfLQ7#g}`iHwbdp$8kpTC6n_1k)Cw0Gp!RNAT*>{ z=M4mxa78?Fix-`%pUi%)?(-nT2i$-b>>v|aGHMQ9r3^{0>#uk;xO}vz#6i!KPt4b~ zy2w_OT5_p zga8z7KWZv4RKg`C-gqjkJkm%(#)(e7{PRW{u7K&l)`=!5L29rfxm!W^YR)eq9fwCc zoQtS2GJ5XjEkNn}WsgJNl60&ftc_lLR=R=1m^K(Gie6sB5G-(#`lt)yHY*v znSEdU{f~Xc5#YchgD!XeK8-9EM(RJboL^v{%fSz;eA4*^B-FB9*2}qdE9(O-#pqb_DM$)vM2!UYnH{fwQj6_bj?TPy$Ya<6iZ*OPUOh&fQck(U z2KT2vq{IEO4iGeu-!e2A5P@`=@XPa$?Idow{Au^|z@u39q(1~Z$o`cla|}=VPsOpI zdhWe=@#0;KzH)tB1lTv@Ou#uEcojAzD&>#Rx&O-=5{eQW_Fl~VN+Yj~{!u(AT;oPG zx&_dSpn7e8`*b9UtHgEj^%cA$!KTj$B^tpbMF}{SG%D;nsaH8=35>9DV26*qO0}Z)NgWOy!nFfQ4ZH$o1U9r-wfv{fz&mP(dA&Ag&J5R^UMbx`0D+%e9V6+P-eD~{z3~26CTCBb%8r2wA3A(^ z2lU8seN#``e=jNck84w3bbu(ifdPo_$!u7JrzBcl8~n$GGQrw0l$5RF&--p)|NAsA z+UE;e?`g7k_wKfqUzq3#&jcKCr_CL;xPPH2dH;T0SNs3wlHbF_#y}Ej$by!4y{@rs zr#2H>r6Sk42ftob&}5z5a*y(PuxpGyqYGP;w-t$ZcfJF7ODD|$6JH;4l{wT4AqzXv zM*gWgO&X@k&3|K@{`$x5Abb*P_ZJ)dxNI{)Rw{)8?rg4rpMOE_< z$m2yn4aV23#>hlSXcjKLw$=}#qnT|e6{x9smg+oBHBH1RS?ik7N&E9D*z2Tm>e>+; zUDB|P5}*7|;kGcWTAXO*lour3--*GtMWb=RlQd~&7}vIBTvvRy-(nm+9WJZ=CGB63 zn6a$k$AK7U^9|ZtLEVaiU=*JJ1tmP`Wy>K$|5=l$(Y6GO0R0jUQ%>3Q^7-dU_Yx_9 zI0V%D)DP!0KAtx#wF>P0|2I}{(f0+#pqLfFzhaO=!MBa|lp*HaZ~x4x&Wj)}wgw60 z^zc7fr`}E?v=5HS?KDkjNER~S?9q`L4fzx#FL4R=UgLD|T`d>e@Cg&75kK*r*>T*g zqV&LIdj9^yB!p2Zf#G0DU$`L=&${6!$FfR&zl{_^?u-0;Tj5}z%Aw*Cz%|-aFkleGlJ=!CSqQ7X%C4y9m2E}SCiM`c=El0u#2pgQ<47jd+R>l! zStR-5h$bE4Rn=3T*%{Z8ey)ncHUBKi;M$QlhS&b{`7om94$o8lOs(8w@`i%aI)h6(EaSc!?I}LvJ`gP4dwULVcMzVTNdD9?R(X1+Z_};f{_$K zk|+&ZShz@b+oi+M{?fJ-l9LXDbErIGYkk`DegP@vJUpXuSC%oC$eW9rZ@Q>7l{Use z8=0>vggDXi1%(^Zv_xR9%<2I3(Ff)w!9}{D$?S;rWDuVx1tcl-zGy%dzk9g zIMA^Lj+15x4GT+iyGH0id~h7CD1MPi8njP_1{2jCbP!X0zx64p)Pon*Pf2NXb}7P% zq*E}c9RzZnE|d0A)c3@{G5N8=7aWHJ6E_S8Zu@d3a(0rgObkrVBs)N=AR%Pt6k+B# zS32M4%UPAvsB_h#y_=W`WG>{3iIDomd;ZR-;!C8`rOk1W>8E0*0Q*RLh`t>ZR6$A& zU;4PNT({q;i}AfTZrn(`ev$&gA!-868ldD0{43$x60cWkNBr*^?sM&)$UNUnQn zS(J0bq*S*^0g_0_utH{71uZw3w_d*!(r;0Ya2wfyIT#Yb;nJA?!Uq(~fx}DS^TNi= zAnjgEs7n0e$8KDq9ewfWOMQ+Ea|Zs3G)GYT8=>^)tQgVlyIiLxY=L(O0?vKpYQ9UExsLo+y=7DKpx~9 z@~s^(o8qn|Svs(CRUz$$6r*%EM2YqX(cLbz%iT6uoO0-vAJMOc{9$sS^Q5%8DdGK0*R~GB`+ZzjsMNLjg^VL0uk@9@^7Yph=lW?M@`Tuz zAID>)MN;L;wYK>8vt(`zv354!&;s6~Lvi3m_on@`Qfrz>1gCRi6&!V*=BPPGE^BzG zyYRJky25xU4Y)&e9aWDXVr#f{`mZ&HVKC{Fcr1r(Na^$`V>{K)vI|4nv#AUxU@ItH%_dn$J&8{Ei*0P zn+*cS-7T>@`uYd34>xE8K!fzidEb;)@O2Q&3u+K#VE0Q!F3;vvtC=IHKlGu>tT7`eKtz~O;Z z(URsg(iK#C?zlP>{y>uq9{jam)X zOT9H2C^~EL+gb``r7vCI?XBRzP<*^hj+AcZuY4#IR6r{t^Ko#=5zcdf;;tOPqxK!O<-Gz;CY;vy*A&Et z?JOp9ROm#uf;t~*e{#b1;>e$;4VhsW9GwTnCcLWD)4-HH5?9i7{}>TsvyJMzqMr-} zM8$T#A~&@~&XldLOY3An1R;TlPgun--|QcNBY?;5es9Cn7havw%27FF@a#IBO@nA% z%!WbEm~--e{U7PFJ0vxhiFlQETyt{M9o}zKCry%(REishHefN`hfdae>_{;lB+N?9 ztOTk!joEmFn(icbN6rI%H;X6r8#VXsXd)XaNFZ;ZneRbzBo^`1E7}H8Z)NvG6=@dki7)8n1 z2NqOj$>^^6CxD=>Wzfim!Cxm z+cBbD)X21si+i+h)2975^UnPoUn0jFUc#oK@Wfz_HLyyepI8fndHEZ_l901K*$!z``5V^GV>jo+n3fIlO6`G|LkIP|NriUDVXd@jMrN=^ zp~%{IJUnUV&Ye@Zg>Z6chQGUgm>t`Xd!Cd24L9VGUzT2;D^_^wG zdHaH=rB|21rxjqI*HnIGxZxDzYWA%#Ik4XUEyaQPY{&A(mOh=gt+WXin{l@R`t$G# z8P+xq7`wH0&;C;HiZG0kAuu?dBHAj}b>~h}PJR3QS6rTL8J){tjcLS;IIvW8aZ&=% z?=;NS>?z*Udw)9AL!t1i|EvvvQz-Oza%HRrecnI2{8yjXPqlB1^EXUlHPgsBLKqsa zLDQ4jvJKHV4_DP|b%~%~o!5AKNV`0`qZ%_fg?x1Doi>VF?Hz_~eb2dfrWV`h#h+jH z>Dr;he0A3w>$a@r@+|9CI%W6s&_+JH+yA~T*@*Izz^pSyPN&W>!6>bu8xKu-Lt zvknb^g1Oc2-_J7%ADernx{OUdT3)`g@f9AFN@j|9 ze%PG)JgL68d%I8Mxjz7Lme6EDE3C|9X-9|$o8Txy`YsUk!7T}=BLrrKL5aY9`7w@XjzuIFOe^dDN(%<;}C_boDaIsaehSm2k z#zBpitl-3V3IEAxP@fU=rorwOQ1l!_9qULP9i7~!xnG!iU0cP z*||+NtxR-rh6Ln6g%RJ6U$~$P@M^SKYZI)I^o@sBWUbv?ta%GbD7#S|)&7Ng-Rf=K z+|FZSk7t6@OtOd|#}MH6>7;Hq0@BOo>kkMHx}tc{>DZSe7t9hS9}A77*lbXH$(?a#Wet)Z~M{>Q#g*aoPyf5ZIwO@%G2HS=K1I? zW93!->f8Qz8IbtrUn+Ln4f2g7<&27qoHe^H^6}O?6D@-E9CN;vM%2Ac#x6-b)w5?$ zkMx0Fp&p;81D<#G&QD7EO|OK`$Vbfe?{6yh(|mmQgxp7wrDaCh(<1^bt*l<5g{+45 zvpxI$&q%bVI2ebbV`;$|iwxpK(`=oz2211T*V?9swv}b_tC!wAlW5tzi8ixSrq1+TKfShbHHsj(zRKUTFQ z^fZNDNM_yVu&vj9v+jCH#JpK_Og&0-s?EhcnlG;})8**57NN=uAi9 zMvlLE^Jaz9OS+L%;vdQ5FvXyN#nD@*8JAlFyR{Ny%6^SYj=OU;-=R1Z(Ehkj|Ne)g z>au5dOe6CTAvaPlxt*zEpnN_5VcObCXv{(q$6=Z0Cp^1vD_0aPD631aUe%}qS`@-o z%pJ9$e|6GFy01*-HBQ;y`0;Aze^DY zkMKFU3pkgn`fjI0#XqULysaW@rlX@{$oLi8iKnsb*WA;)U5W}8TH8ml7Smv6rVgJ& zf0juo+@7_?HmuS1<(Gfg&pht|L*|5>XwcJ?;aaoJ%&cG+kR_*9u>@K7wzN|yULHNQ z{f_3U{~c5PXU~&>`FP}ZB3)Alo?qvSCIIz<>Rm+qrg{D$?;o|yors=((J}Wsc)#OaZv?m$?b^BXF0GB@wr*`lph&uSHuvA+ zrUb_jA4Z+$0h1b2ISe>E^GKVsM?h0pf41XBGAL6Q{GIS>?k9?AN8ITRHPOPwCYZ_Y zT<+LDDjBc6bDqBJkgV<%_nKH%i9w`c;l;Lw-0C3qc!sF#b6nA|XHrtV;Zag0!3Hh; za&`N3-=V3LolGvg`MaSFdrI>A>#ja3j`*1Oo%kvN-9E0B2A;h*B!(3Jqi@z#WDnXq zb7{Oa8zN@le|P-7t6Q#0;mz$ZLu;i{smckjLzAX_6(h8MR8f?fJ>2Xs-4V-Z&~6M421sq=8PNukk}QoX4p|;` ze`MJ0+JEx%ob)ydzd$n!ivr97X8@#5u=s)y*}v1vNC@x3M86_Q4sS4ucKqx7=ovF+ z#Bw;Da-)CKNhE*Z!GIDv4+BBe89UBzh2NLdxC|n3EEq}4G_|8T$l9K01 zt88pKhtHfbBPr}3jaAjRUb8(;vQ6$|od}MYa`q+4NG$%*GLBg-=fT z(wy&zP{!P-lU(ixg$z(Ah8(!tM*RiPKZH%XOY^-GSSc>Yuj=ZX{^y^Lu%^j?Dfh_! z24RAK3{@(XE_R%5{U&@X_A|p+y72by(2PGr{mj#&kKmcB?|s!Gjn2L@fVKvy77g!@0H@R7nj2qMPV6g2TEPD08gZz_gg!8n9&Fkg!Q;`Guq+t}FH+H+Y?#>EtMdI#e< zj~a1}`GH({<-?~Om7DLH_SPHA_cLah^2$gm(u1`hK6`ub{OH@^Q%ilH&GA|s(~FR5 zFqZoRJz+j!P8z9X>YDf49x-jmQN;RcwBrISmeKSgy4x+>uuGN2lS+$9DNgy{>JS}F zVZ^y9utWfT=CW$JbFUf6SQ{Q*Bl7%CqAQoc2t zX3mt+{78T9#|IsL0M3=x9ng?S$rYdZqDFtARs%F~HD)xUc;NRhn1(BnJAyRWV*AG~ zd%wr&C7IM}VC7fizW0mV;_Oyobnn^L+I80}^2%9-q}?C$sI9APYW?0Y=<6ggxa{)4 zn}LCWlPH@phM}n)Uol~;3G;T$L-pr#mOMbf5A`iAEp=(VKoFh-Y;a3CzSVSSiAS!v z?d5e5l4IL@<}Xy0oBSHqenSihhbB~l=lEIo2^!f&q{0!}zFkv+B|ccw7NWh+DSOe~ zAwp&*Tl6*ALZt8g`R5T#)<)w%o|Hmw%12g=y*MI~X|pTm9(wz?R;F2iI-Sb}3o}O)N7? z1q{`idZkybsWb&O@#EKvXqgMEq^Ts#pLbaT{Z`OTuXsq+KjsfZv!Y2zEs0foQ|_c; zymfsxMgqbe5}_f=&f84SkIw8CA^7lKXIb%BoHMg$&63u&8cNJ0cVSW*BYwf353GYCsdzeHsKj9*{i)vYbkiFScBnc!*lF z_ja8k%yNVVSmK)W@184N6n<|t%C^47HF<^;K7)X+YtAekeB^Tg4h#7p^uG^I2@CHk zmpDKER4bdp^c@(v;KWfJ24*TLZ@2fUib~&9&?(^^u0tQnplGXO0;zMElZWP-hCnnD zYCU>|MiqEDkK1Fcw(ln60o^6Xd)U>^=dkHZ|w!HU7sj{%P_uP+~c@$_Pi&p#O zOvT8pCQQtd9FGIxL+)QvmvK2`G#BA1=tV&y)ctB}YH9I-a2hW>W7l*rdysvgmZs*a zlNnfFAI+hx;Gzp{M(0zOSnR&-#&WrQ6&2%eY}j0-4zB*oZGz|I=1~yXKwl-UIva@~ zk{RSLv=-XJ2Q5lNx%uic7Fi{8BQCP3p!fGVh8*gUJNLlE%VU-Je&;dz8B`HX*rHN? zTCd@A0wM=A>2MqZ30?CcaD&s!_K_|Q>2Zgcb_Ns2h{=-+d>>S3RD_bkEyzG9)}(>&Cd=owpTN)c^XK$qC9eQnK9~th|(8P5f~WmA`ifTn!c=O0%r5jD0R> zgzaCyS&n~DP;UvlJp1_^{|nGWRE!-tJDjHC$E6kP9P{_oD-80QXrTS>P%+I+d;vjl zkhLAgXEhsW`<7h>*NvJy`Ss`rWB9^upM1>h+H2kMB`=2QRYGg$GN@IP0D$P+ZLoVUgY~u5@U*WPs4i+Xw;AS zV(y$dd4!OoFs?~$Jd}q_)%Y2@e*s;jL$4|o3LQn(8=sp!k`tOT?{$m_AFaep zDKm5Pd@i9Nu4~@$#n$kCMjOwf8OPvwxg1}*p~%Gty595lq(a`O;?P^J$r={10NcOF4p`yt)E{{96pD4H2z*L0W}Ciq5~LvzTJ z(}q0Uey6dnw66B@gP6I4y^W4c?4D%M_9a{cyX7lB1_xsF&gTXSMS49Yn~s<~YD?iN zNXxbq+^Dc@u4|upzW2a^r|AJ;rs{7>lE4s!c>oe@s&n~v{H*&OlXLUc4=qz^12)U@ z$(@7s6;Q?1+=V&(T1j&<`_rlPO4=D}_j(;E>Xb%Ck}E?wk;rg!fXwv-VBCuY69R9NGwH^l53M|7Qpu^t5$TN#RAfcH3Va^@LWS${JV?!;_qi3)m!gq z#=-+aN7^N3!B7r2#^x%KMVEoR$|?WW$s={Lcw;wZvpH9Kd<`A=ZZ8NFx-=PiDxqOC zJuNOQ32fP9!t}hE2x+PuXSbrng@|h$&flN?hYa2?b*INCvN_8A67F>(&3!i8+1ql3hNx>LyUsc|nzI}Nmdoo|rSfq=~ zh3P*|ke5`<0#fajT3oMU7Pb0(rj8W>7P5C$(d+Rm%KI6l^CWG8H8amQ;QK!bJaMA8 z8>aT7wBAs3gA za@*3pHnrlFPl>rB>Bdn6KjY@gZRH?DQGW9N)Y7TogWE{JLW%4BL!Ee^UcCZG6$;KfcC3Lq!Wua~cxN;}H{p}; z2#@nXoguIfO$ijHi?{|-iRl`~mW2jejY?AMKw zX7^(jm8$b4os7qipAX0?WZ-adR;(lfk^phlhQIo$4Ozr%<`X+nC0y%XJg)1-HTNvr zckX@5&HPAc=qZGgzs7atxQwBmN&+=1TpZ>iOUVpEXD-g~D4*3}gqFIBdYu`+qTvfCox=b2XusCxW#9-~=(Qq$?_^31OUe(qkAt5Js z4DuOa^?WqVILX>2z%-!go2eO0mFgIEkZG_~?efXQGHJ*(LOadB`!-YV&_X-+-&*cx z{F68)zAz2~%$X4nwKZAyFY_U)=#rl^<4gNLCIU7e8h!Jr8cN+-9#g&5grY604W^~% zSM3(p5{5h*Rqy`&<3RZ*I0oz$124U{ifi7ZVM_+>)yn;LBPcQQ)R@KD-5tL9X+&?g zWtsoP-d-MZXvfOA7e4Q%UIp(kt1gBgrXDHFzZr9$6RW=0Srsz)nx<2sV+V9s?-rmy z@wui^XD(m&y~?KFnoV+QW|7yaWZa%PVRxu(Riq z9i*`smy2mTkDVs@4HD z-c+uFheoE(YV@t&#iF2-F^L~-los&yxK{stv9UK~MtxK7?mML4TQ!hTHlRk-+*5<6 zc+5fVqcR?ikr)i%{QSD0an^PUi=lp2cKS_W2bNEDOp0dH9 zX$K*mF?CFIYEQU>G)$XUDIw$gy z(KBn=G%Ii49qMtlG@fXs-M8z=9vyF0)XP5*sw6qyyHA5-NRBG`Pcu0QL(dceKU=lv z*j=6ng;M|rz5n3B%7Su;LMA0F;IN~68namQVB%~{WOR^bRH6JaTrswioQs7^C2|dF zJh*3)*rJZJ1!vEW&!{18VrmQjeIOxih~DSq>f8hgCrAR*yGbKFC9LpFtHB+Ixk; zuc-zk4R_FMB(OSR<`rI!W+ZV4pq#K?t)aiXKOY%JE?fO~0u(spr*O;r`oaCSKPt#ccdv>%a1vVDwXeBKRJz~#jyGj{Ea35LMvYzP z^^1$evy#(G^aN+Czo^p7C&b7xe&qw1%YT`& z^Hk9Lzpp&rHAUkS*ZZx zD8Kz?0hs-_eiOfw+??94cxu0Fc%uN1m%lv{=9^FL$XBgRXp(JbYWfRiyJ+-&lAA^! zLK7Nc2}Htq7elLA5_c*-+x)wLr)xHgm;!1{<$@1uu8~o4w?v%|$z#ICV|(t0Z4k_( zs!Huk%=V9*AsGF`*cETyB7H72HNCFAcm0Od(ve}^Vio0bQ^qV!U7T|@EGK)bW5|?E zZ_di6K#semYyOTl%KP~`?(8os4`-^>a2y?78=W4HSrTjtL{k+19z z4vt1(zKrVlvuDg`_b-YKes3VIIMKlrU_6D-cC9^+AypYToO2m0%gi!_H(Vx7IstNS z29}T$8#ak0*q2lg=lz@U_W22!=SQ0#9QO@V%YR{&{?$p|!eh@PKIWQ?mG@S7|H;n7 z4y7Sn8&i~b04Mm1W@?@Gj#sD4B-DMoCYP9h+U=`y|G>ewZAM?E%4HHbOZsJ^s33mU zh+XxfYl$sBoB4=Jx|Dyo6knd-fAX3D4aqpdH_@;FAo45_RM~YP;ANx?U&n${?Lr z{KBl+;i*fRVh7fAFV4Q7cp8R;bitnfS?j!fy}mdNQTfW2|6r3zsiVv5ABqyY-nawm zW#+jdG(us?80sRDn+c_TKL7mDW)cT>O8s3FXI(}tbo)wo7S#WtFT!zGT5?t74fHCG z;KM;}5Z}#ktQFD@=OP_4^rj)F7Jc*V^rZ(m6rXOqUPn`T*+o+MPHpzuT7%Ftl{4L| z*E>Y~v|vFH*2vSe!yk8(G;PrOhQ&>s2-px|x%!M=JP$OP{byWT#gMgA7MgFqx9UI@ zFL40s7{QWx&>|56<~Rl!?>5~G*+h$Zt>V5GDg+qp%qR68Y@E)Tuha7pS$wE4v*p%b)nkytud6T@1Z%UU z@r7X;TX6X1!#4$XxV5F}-!CzMOrsB+$Bidhk#9**P79}6QNJp#m(NWH4^o0xL@2Y_ z%bqVKFHED{w;GzJ<)dsktSIpnWKt~bhi%G+(SF0m!lthdv82E_c1$JBZo3 zJ`JmVVMLjG&}|==pfZQy-^XM&%GMPRGcqzrCqs)oORA36j`lq7Qq(QV;pv4z@CL9w z%sN_RbEdux;&9pDhX${$Uo_dHz2f1r;8j)Vhw4j>Yb%M4OX-pye1jigK4tZ!vXADZ zyOdtBr>)0b3MLw;+`R1Rbu%QDQ7A&5S-dHsRYM^R3}%6K5P?+(@0TQX*9x*QXQczC ze!RNvuxB&?IVEZ`DnpX)mO+6vV2;ly;}iviCgBFucmHJl_tyg8#~Kxp62Q{ zc_|g@;a^Q&a{^sF=}{fo?WDC2?1joO+4Vp&wb})ANi62W}`Dza5>NojP)F&x2d-4 zUxHnRB#5`?WZW4ytv}M1#!F{c{*&1%kJPluVLKbdktEz{kjdTJr#I=m!eFnjOE#AM zCOQ{)Nl?|BJry2D?pikc*Fg?U8NO-ImtV$E+_YcEzWH(?pFnW(wN3$OeG36kfdpyi zD*R)PG#~)=uh-j{xp7bZmKdEuFkNTTUfVVn_e;OxoOfborIO1jU zo5#B~m2gZYp;S56fH18&K4OEDbyZ6}L?)qmHH>lB)c&|8%bDvvYeAr4sBLj<5*@k< zYx2=rjtJk zj%bT)CO>{g*@sd$;{nU3h@zF7x|b+XYc<}m7^8Fm;v|ar00lp(a@s!4vUmNi<&Buvf~l3Qph4=uc%F}&sve3rska1U0IhLH3mRUAGZH& zc>2#!4kVF@FC{r-* zVmWS*DM{oC4Dyh)hFd4~-@I~3DMTt0Ds_@KvHT(`)|&isvIr}x4K%~N)g1y(+a!Dk zbs~ga6eFi4;jrS%AvJU0lfy6(%B)ZbLk+zKiZe<0;?pygbh=UFXTE3fo2 z*}H334qXEhdi%A4*Z2gNB@*ySC;uyp9I)STi2grgmG_ob!V`F-y=o4$s#y~3`FD+I z_WXp_ttsN^r^p?l{7_Ji?>|axxjSV;E1(H34OMf2vZb#dx&R9tM?PS94rdARHHHWp z?pEQcL{PdQiZtt-)TVw<1tbifo>C^YKy}Q?4JxehoO&2^LLtrVgX5>8n+0cA81}!U z=18s?kdBBej0YwR6%XH%zx{ZAsl$%Vn*yxN@f?@!9BQ#~nj#zEUr|;^C$r z_`N|!C!l!HrY{X=rtnVO%p3kHW4?_z(D$mpg~2b&d|RJ^B5M?HUBIy$rECqvlC+6y zb>4`esEilb(nMe8(wdJHzkrJrD)PoST&Ub5&|*c2{Jn*meYs=bm40{^4!04F`!4j^ zk`D3~4@1{ERbxD`VYG2jc7^*ghCM8Z&K@w=(lWyT{s zYBAc6GBB_KB>0{tHK?_FwNw95y?7dPs`ASboK(dR4S8{$HLWTv{}`%A)OIOKh7*FF zkuaZ8hh5FWvblUC!U?#P?YwL6-a_O~A=Sl+tKu$&!C6(5I6Bi!>wcF$BP03dRg_Q_ zvS$`R0bd%{8onwF9e$VQY#LEx^+~3p9jjGlIvEXzjuY|ba^#FpZ>#ku@W#@i1q^(c zM5a}dyw<+#+eu0Io9W5AvhPi$^GgcO!WNm={(XAw{R>K3L0rIa!QB*_!8siA|DTTi zj?s0o_pl_ssw*wE4{)^!4%V?h$%SD|8I)G&#_U^7e?!|v?$$}31*FPl;T!nq&Tu*6 zJnnD84BKPk^M&N8f5zsYMOQ`&>E-tB}Uu;Q9v>JYFThLG?^rZ&(pV;S!I-|GUnRp~u*PLX` z0-huDWko{0nut@BduIL6fS+3pVu%GAGM?D!&e7jH-KelQa1jQ3pn==Bmanf@$gzh+ z;1r>@A0tOt#K-b7lzW~FKHR|59S%*C6avH3DI#Ck#cq$7Wpj+IyOH}opPkYO6hhAc{d{`rX$E-i&cxV;7a@VI4)Hy!Qlz6X7934M!h>CW4HW&7Ri!Wz=HxCs{j zad=x_b%I<;rrN7jjqHj*4G}=04W9Qiwl>cVN-r~bk}Zp(DEA^JPnci{J%!((q~RF| z@GfqcC;ordm83KteD1yB%Ve60-aO6$?V|q4kbr~&;7ZU1X``hkSMyhSh75$Oe5w+XFYPr znNjS0vlRCV889emyMi7VYNEN76a;YCG^5sbuUXSv0Iqol{t{BW z?0HhqgOV|QyNY*HGW}NYsS#E97FvI*bE6mp71Nfk2#JO-IF84{>@T}OR%8kEpylk@ zfA#Y+Iq?2;#!TWck}gS*_JHmXMPjHS7k+*(bs@hW`f1n0eGbYiWwjLsoL~~k+R(H6 zirX^UZIIpA2MkvdHe;j!Ti8S4%80e~_Ga{f%0E8yGxKm0dv5hha&_U^ee!x$8%4kh z{E>upjaY8sDVTBP3eG<!^axfX+%1ltW+OQ0hmj^|*>P%9{6TP07v^`b>Sk7CGG z_b*PsO{9z4g3BVZVMZO=nh*a=A7bv{!DFi$zDzb`BEe~L-Wt?NPrbkt(vb@Sox;=? zv0r_dRa>U3%%>yh33a}5ezch|>yljH?Z5{4jVdd3_y;^Rc@&rNasED?0Vlf%)7I1_ z6>N!V@4r=(0@IEh2nveCtKrCOv1ZIJ6Sn1qs&0qk~}7#WbBlQZ}SPty}hSw z6)PHh!WoUHcG`PU$@x{OGAQpdb4NCembMbMM(u2*iLKJwZvD11mmha-Ja}r(O2h0% z8a>qP)H-qpr^ZUTS}alZCs8JcP&{5+oi$%`6gTlCW?s!Xe1(l(18h-6{x5!gk!;xi z4xB|SyI7!G%ThAP23$6=eh-VW*Axk{9E$~sCWtpx$n)=?f;uF|VZ4a~^6nha^}-WN&^}22MO8 zVr?7@4bL0{h_1QU`V`Q^2P1lvuNmucAdJ+sG#TJbt zWoQNqFXR@oyJCt%!~%I;!1e6eTYfaF3M@4VpG%zw*iJ-FNvUw=RC65ii3ytT`VW~p zzsI|&ADZ?u39a@>QANE{C7cXuWB7A@`K}6`Z-ZBby_}d8y*3}mFljh0xkuaY3>1(Q z6cBI}RPa}#%eEVjtX$hwjFujhDIm*{_?u{ydxC~7;%t)nP{yCh%iZ!iZxy&~{hxi3 z%lr)?(D%k>wEfvnqY7KiI%l_-pEEPKd3JQ`FM3ZZvYa@uag2;t@2go~e_TVi4)^hC z#Dh2IO2@UL7ra*`S*U=o3*ZDGr-iH;i%i?Mqi_hVS&m*_BC=S?x*OfRM{eDm`rI39 z2AgIP~;oAz8#D@+ZPd$SuG^ zmW17(t9zFwiPj8}wX9v+K`E>`n|eb|nS5!*t+oDx+LqG7B5`f$8DN~U+3Q@A)u`}O zKcdK-l~OT?$(2(U%`P2x26+tbd-j|;LCgEXUZfi@yV?FVoKZ7SCJr#f?`|@pHVzy0 zBH&BHh1G-PE&NiSREBIq=;dS_ASD{COfM+ck{VT^_D{0YqgO;VsCfxC6da4!;fn+2 z-mDXjo)6$r9I^?5194IDwrJMP}CfFt6|#izG3?kh8D}cCLBMXsP>vFO4Mjv>H@Nsho9`e6b)4RC(z2f@nA?Q8wZAIr+^`QX4+~Q&17$q z{YKJ*+UfoUN7aPFa&du^^sAwKK&MtOn0F4$h9|sx6%h+8F@Id&1!Ux+dlL8^ItLhb zc{qSdtvraB<7flrNj*e7P~^alfQgsjx4DC|ashRTFEm?~VwmC>Qdn+E$tATUgh70F9Jkh&zq zR+zz*M~7j)_#Si>oG2S9=iG{C-5p+lATt+U-Rvtwy4oAcZo@YLk0ZAimT>$s7d`MJ zh~xgd@0?)NWuJ@q$8%XIqC!KRX(QPrw8aSs(u&GuY7B%X#|M#&9~u;HeSUPK?5=?& zj@NUJ&FF&iV6SByV->fB!&~bB?N$_p%*^Cuym5UjDK7F5%@)YBVS3$?#5H%>vEq5- zNX?}dB~3#emsWl(Jo8pFk8Gvn-DMbC^s^RXASFcDVt4s9G6m~*ZmZH>HI zy5;rz_38(&zgztM)Kc&qR>n$k&v{*I4cZb?p zHVdj~KbzOiE>yCkf%@yFLFBqkJ$!EX;PF3fg;e`a2rR#{k2gT!z_IO2X#eP0m4j>d zB;`gF(fKd!Ikg#R;R4K0SF`~QtAOIU0488HL4xoPwRfS z8+V9quduPN{^LKNi~K?py{gRxWiM`t4tivc{oQzcbM}OHA4CZCZii+Rrj8bnkGJ2E zKb4{vgmfdV57;3133ZlQOe`9*S>W20(T>R_854E75z)zCo~1rXEj+hkrYIHYV+N1B z^oKS;O4MBa@Gf2!04$Tz<$-zb0?b|<$Lo!IJg??XD)Elz!k4RwT+J)-y-n+q*GT0Z zAa)jKZQ&0;Jf?&r#&yGk8m?l-W3R$;I;in&Rs~`b2Qi-#TDcx!cxsK3tpW`_H-48M zm$93oYizAG!Ao2Xx_tPW7rr&Nqq=h8)h2A`a5 zYiDMfwXSxT43-*-@{`7-MuZ@Ns`IGOn(I@1wS~(K`kRR;T*suL2oA2pDR+!}ch<%V z1xGt_@!6y7zYAzwjf)-Dn27C{QjyaV#UuM)Ht6y%EeKA4o1NfXMaJpBnjLzm+VjS^ zV(0ZH(v4p7JpmmaH~|Jt9IjqrX!F<%I(SAx@eH&Iq*g~#QP8l|Gx_y>XCC-ib^HkJ*&54}hNG(^ zsuU6p)T^`X_nn34Ot{X2!92S`RT7ZBiKa&S+J@z z#OVPzn3z`QVdK>s1(~o=CY#vqqhV{Y;aKTgbkHwWIGs zov=-igJkbgc-3+Qz7BNnQUns~x5F68TZl0F*ThLok5n@NFZc6p;~30RdvgstTZ-0T zJtKj#$`AhgX(WPz9@+t&d%EYVix_A{OOVEEcBOAUwdEqA-zcXu1MSRE4F za)<3TIwnS2=R0CY#wheSXi}3HJh3b|^Q$V0v7{IwMrC%mS<+MIsu$`KL^xi?q9SVfXG$1->6X{Gwm z-TG2iQIIy5R)8mW4P_Fn6V5EV1fG`n)+tI3np+2N3>_nF=f z@l|`qE9L3=%m&u*$E4x3jI-w9P#!W_8B^M+BsOKQ^_K9e0mdwy&YHI>2wO#fGf4P^ z1PhTSq<{>tr)0F-4qPiJB$#<253fvY)y_t_wLv%RBnar^w!R zxp#gBKAVVnBZ^9=t54=EY6wKeqn;NVQ}V!@=DBJ&gb?<;-G&<#S%VHbOgR^}{(SNI z(c(stP0Y)SCBTViB1Sw9Zh6cyjVTb3t84`AF=NN=?4J#Ra%WeA0~KiRzsyM@9qrj zwE$q|0A!21tpN7ZV&yYO;^Ig~YehS}zPEUF?VKRPt5gJ>I4S}%3kE_9mYbHqIZH*x z)nYeTcXa@6jqtLM(sNSzo^w1^})Mht(B^gz(w^Oc<5by_Fpr zvI)E_R#4*P^1jK`kZZ1rAO1mdV5&<)$EnAQ;c8muP_szBe&9!7XuY!QF`lIv!D9XN zN{d;4kLyav!NFg}MYAu+`^$fF*PpkjXg)rEWg*k;=-2M&xfL` zI{YpLktiZSDC*V1-7Jk`Ru-Qtp_E|MqB#`Y<WJ5a;$5*6 zz!*|=$vwb8!!&9t#Gwqw+$*AX&(&U1GbgEa@T#OoZ-{r7y7(;eI;!S(AKFomV1IcK zMkRcS0>Z`mik}L;Sh#u#cnEFycMu7fQM#mb`p56{_k@CS6+-xMs2X!|z#-j?Q6^*D zT!3Adid=XfrRtNPrzKv|qUs)EIP`_u6kzl3?Pp!&vHR@_?UAfOuaBjGXOcPCGGX4~ zqhL5djdk6xg}dGvvqGzZJVa!6xr=a^K5^d=YrGHVuBh)EtN)U=TXFagDU~i8PqM;B zY5VGVWuBK}d$|`AT`FuV>dV$|9^6~o@>`;JKA=Zf#jYJ748QONN!R`$9|}hHYqlu) zu;}dKJP`2*O`L2)%2LJaBfWnp?wp0%0&+>&@XQh+E8(4e>jNL9l#K`3_*iEn26r)J zX~a5YA2|s?^MCNsKDcdp-zje8{}I0<`+YUgpUp2clLUNr_z7+w`B#EGz{B(T!*3w2 zSGA|d2|My1@%i!uDy-8FefHa93XA%0f3(5ckfv-!i83`nU4^Oe+f+eQIAN>6 z`te*-S-vH)hhg*7CK0I2JAzC?q6N*vDo6=`%%o4v)%(@^^HHNDY?do6I!c=oF4JGX zflF5Ah>;Q$nbNj8OYyd9%9J%n_CIfJVMALEDHs)J|40AzEiZg0ga~frVWQLnO~oBU z)k=1y!KOserg%!!3XPZ2Hg)4U@~n>7%4JuDeym0Q&{F#00;@sje2-+0rV{kvH<@%- z8=jk!qwCx3p{zO_f5akPm^*x`fe2qK;n-sj&x#>fZN z`lEWbd|rc_iO&MB94Yzv)$!?-^uilBE4!*kbJe=qn(h5vndMq0GIt!mThRbk*0-kg z+yf7)+B`yusNh-lk^bBrm&1;3BDK?qntHWWEG1&OH&x4fC~iHa#xP{Us$GRb#e7C=2F`|4 zP+DOZhI>7^6=wce1C2uIL+2KajBckK3S&(xDLD$z5OEx3Ve0Xg_WYuSW{t}`d4-1> zqciF9)53Q`u9oGbj7KO&SY|-fp(z(acTtEJ`Jya6C5*B5e*O6u@Iq4C=M^K{9aukq znTV$u_}c@Pi1i~fGICCr!nCDpgTt$m{`6in^Skeob zSZc34##FAqXh(d*C&|3TZ!75or%Lbu58OJs`vOD-Y~a0P$PD3xXFFyN-qi2gB!wpywU%|5?Q@|t(tGZu+Tr9 zVNdgfE9Uy<+C%wOxW5hYGTekVbJ1?mhQ%UyIc^hu1e`~v|wUcSC?$yi&w*R81-(N_MN0hFHrRPFi`Z2YhNAGLiw! z=(|pbppE9rrBR-{^p;Zbe&w&uw^K8x>o&VNy@WU!V`c8zcRQ769i%55n)Dsb19r!_ zY6VgfVa^L!^6KRkqhmqIFRdZ^;NQZmI5yq1lBg^To9NR2qdV(HqKo0BvEnhf*7OXT zND6_3SRkGy#Zx1RLxW(j!R$AzRv+FsP8EJOx%y4ZCr5TQ2tkVDftQ%!aMH_zpt?D1MJa}~90EEtFH-n2JQ-3nPTxk= z>3Z67h!EyO?d^V*>anZ8NKj%^HQ10Q-(OL_6$8G~(&s<|Zb`0ZVWwIk-3G(+%BqVA z>1B6Ca>5hmj8$CS9Bq9)tVZ(Tc4a3%oLzIM3xr|?cU(XOgQPHR)SXE^NYLr%%@3~OL$Rbtb z=29Q#AFNgV!F+yzKjTm$6iN?^42U>+gJj%X{gZcumh>|lK4{25P9YP z(z-(G>is)KAe}DDIke->-lZe5O}*rGt~Fufm6}b+AVQ-fH6VXw>Xx4K1^;?rO&ub_F5(OM-3d=^^7$?E2YX@mRAZ;kcg2PfQ zM1Nr=Sk>*6`sRTCUeS6lDJ(N=+u9h}jq>C^qu;WIHYII!jbw7RZT?Y}naqpUUFk1Xb0w;y!LCF3FMej{3p9p2R97dEmIw zwWIA0M2h`h=vmzSqm{M4cqxVAl!|=0#B-|l9(M5h*+|w$;eGoq%}XBKw%PIp!4Z%$ ziuR0Vlako#V=fgRBcnY5*{Z^kGbYn*dZiSfMwgu*wZh!O7ly{qeVe+v(QnQADTf(g z`U)Qp06W-DAtTXOso!3u?nia?t(&r{Z4*z9;&AXK1Eo{cjc_VV zDVwykuxHR;prqM29@!O6~UKu#&>LlyHdNEtDT=59_K_R|29YuWB;l ziRLmUD@IlYk2DnueE$M*!$z$=apMp5SAS^dKG|`9`+D!%O?q%1Ke*EXHn_!=HUy1G^rOK|iq;i<3zW|_4@4lMNS&1%sMu{skdP)f& zoP4kb-Jw?5@Q|f|K-3XsRm9m-?b7j^pC%Vp7a?j;Ye^drZDS2RVxD;14K_8>5$Ljz z=F-O@=jy=w8nJL8$bN_6LB?p;WV`LCZvq&DWeVP<4N-DQ6#$BkucTdt$Cj}x8}G}PR@~~;@5|&%tM8gpv2#E}O8!1GdhSOnHpm>wTRDZAODeJ! zkCj$zq6$D7Ybr%^@9?WC()mo%cXX{3DMfSVSN-G#*DcqdZ*D^Kq>}aN1L4sFgG}_& zQ03Qz6~(gQazqtEv-U;llr6{x0G0kfuFeB4>vDbL)X6=fG&MtWplHsVD1wTWz78Z@ zkmAgh3l&5J)SRh_`etrH6t|+d1;v%(O58X=C*lN;A!J?LEO~VBp095`pAWg z9|fGtNXHtKf_4|6@RwPhx^k;^-L6cXivCyH+^Au9<%L7&RHGJ$#E{#6PafIuLDRA< z2V1J1OV1-K;UH6l+f;29QM$Yf{Zb9HXvEAqy;YWc`Kx70S1d ztPnZ*O}{~bzi+SW+ZNLy7y>J|uMjSgSxjYWGhVMn#KgK%Uf1b(7tJW-07Q1pj>PQ@zk(Rf-qx&MujpR1>RV!=Qw($!wIrqlw8BO42J?;@36XMx4drmQa;o(a# z`%ZEc8hz`+p>ZRh&S`8hthJx;^%!B{Npm>MzNoquy%0!zLlXe;?6B;;Q-Mw}Dl2BX z*mtFf@>2M5g!#%A4s?NhfrCS>0`sHt*&$t7v`X;lmnJyciQ+Md4wE^IE=za0Y207j z=;35vAbfcfVt-XhvH%)(6YVaoFEx$oNMt>wrq3c`c)mR4SoBjAp*p2($|MWQ0_(Hi*nyv2U_+lAmoLKQ$ zkI$GCWB>*=+E4r`6By$u?P1@z9mAI|iu7VxgR*Kw(|>O|r<3kV2W#kZ{kB29CORO3jFs7(Om` zhigE3vq1?kCSSKk2lngh5K4=#D^X5#PPJ|;9kYNcn@RTxAe#}}_oA%FP`J$q&wD^1 zTU9l~R=YboFb*4l7~ak5?R5c0GekB}s?T|Md0By0)}&8I8@&3My{HT);Azuc%}9f8 zw=_1nR}E{g_Fuq+GTD`zA95KWL8{=eUP=usc6Kd?kl*fM^P_A!pu1J_vkU~L6mGy_ zV?ad8o9TBYBo|(z@kTlZy9c`i<}kFNVR~@XwZCl=v6$p%Wa0(q6qn;5s(p94hXMf} z{UCdj!{(7^`PM7nc?seIgIjUqutf+MJZU4&P<{VAG!C8;5b-EL@9|cJ-G9og0BXD@ zfLdyLpB~ zC<6g8C+u+I#>+aeHPN<;CK@&)Dx9n^{`Z%w1C;y&uTmzvp_~--tI>J})ON|cy;@I4%p7Yk<@abXDNx8$L2{;f=iHUarqyx_DV(#B>EKFRY^al9+O<+c-q_g#`$Jc$Dsy*AG$0 zsz2|&{%QeWwXB1$r)D$+z@K#cTI?fS~tC365`=TR-3qhVa*5a8R4-)S-rSHa0J>tY0H<)JT}Kt5B#PMCi z&0mJ)JD0hys`Hn)+R?55+SH~34e=b~UcG8w_^D(d`%*?%b=Kz{n{w`TM4PL$E2h0`5xtjX=WJic1KRFvq5#vDWau{no+Z$$FME%OueJHoj`+ zjK|*@ShxL`P0J|5Pg*Wr1HGIbkOsc;*cXO43(p86CXp-M=72s(n#NcCdv30@6>SaM zNwS{tx$6oKkZMQm+iw+oHq-KT>y0Lv*;d<`1K_ve9xU<|!>PzkW-zRsFfp+_*@{Wk zZmb9RsWIrh`)?@ASBlIW?=z7s6&%1UIjPv{<{zsR9}~x)TbdEEPm$rs%r>)=uDt7c z$o2U*HN8zFR{1o$euqWPv7UVARhydGpfgyJwOS44`+K9aeLSj$GY_cH+i%grx#?EbQvOd#<46HAw<;DJV$^e>yll}cs_+fml|BL;aVweyUHz!zRnO};BO-Od_tVM zE|_@yn40^Wk8=~mWAy0uXn8SQY2?Gq53G(1u6F$MGHdr6AGx&s@LhTR4~z2jmExw|=;XU10l%w!TC&l#+Ep^K#l4XS!A9l3zXJMAs+2Ry$l_ z8r?b>L6~3+ZnY^=!P<2;lJ1#C!07My#+F{Kj6x* zHM0?38(#S>Z0+pQfCHx3?d*A=%bS$&k-X}@oS!_k8N3kvA%1FndC{e0;IliJ+jUB5 zIMuQ~FR!aNXMB}&U)RhgOMLwI?wXH4|Ql1oO?Z1rU zv@&s{CKDDw^(Plr*f%NDTiL;*Wo;X4*A|FcO?(n`$Q2#jb zHKtq$L)KG7X^nY!;h_J8Qnb=v*D+-=pz;4LolF}y5iAA3 zg-ZH&Otza?)Z2NBhU||t9`RviZB$NQ_j2FOSBZ2cIcE9!v0Vp`W9b*IWYLa~Bz{tU zejgAx=?x@LAh(_TCJH;cLCD#@6iU>B!XK`ENbJ_ZP3xJYzgM#p7;`6$6ZpXS6VD|3 z#!rm#`@$&^PK%tYm(93k3z4@4C^ty!<5h9SQtxo(lDj`l+WWW| zuMF~2lsx}&(L=wJ)GU5Ppe;a>R+Ru$SqXt4_J&v}`FzcS1)=p>$cPxitLDd}cpwh(nRkD(`1@ zY>(!Ln>-A&G&^c0RVGQWTH3RbXJl`lB1l%fLAY0GZ>*aWTW341J(;%H4lf;wbgPi0 zavE}lKF6o|!1A0~<_Dt6t4F`}>}uw+IJp=X0D^U3$eH2_nt9Cvc$D4%o%GG*e?p=l zW+Ff5=WVKdX#Z8?61}3KYu!e6u=!HsCi%eZ+%Yt2 z#>MfslSdEGty|1}$w7t1+>*A6tW|V~bdFuGi?sJWb~Mm<1CxTZ7J6a7vY34h!jjN2 zmn$@N&Di+VySHOM<%4#?Q=L&7yHbG?`D@CYJabxoWK5M0f!>1Q5@6lcJT2HYCHOA} z?I3WZxBNYZ)pE_&(Ry@-lf6Y=CyPh0OG?38!&iE)xYW<6E-e>tzmjXnMrQX_&6*u0 zKB4Tjx4EsUFYHl{#=>=bjDp}r6SL}>4o+B=(mcm7GwMhMX9?%S;!ADlmRd8L4r>o2 z0L#7slCYYOX_75fv$Lh(U$*?M&C2UhA#idKR?YTwHiJH z9OmG(SFnETV~kAul)(<2h{7z10gVSK(#||~w3*J*lcOWX^*7VHuU+49X8JJ|sZo2@ zmy@bKc7n%kQ~R}FRJ((7+ocCnAC^i{>?1xpP!`r;ClJcNbP&|}oNQ^&Z)AS(Tsr*zX$!Q9H z;F9cFa=pc>wMLV+(_ZGYu@7KyC&w+_OVcWsv)|@b>C&@Xx5@7UoaA~-B6%7!_V)Ea z5>Z4k@<66Z)aaJOB*Dik8GrJmx2uitllOD$!v+bf!<7cugM?+fJiIdawN7buz8F_U zg4*VKsf5p;C4w-PJ>$w`PWCYzy1f0jM>7}f%{imldm$qyjli#hwRhB%aV9rZklF9U zW`gy!rd7N3R3!HZVp?x6ki7+VwHxRv) z6}aHF`+m$w)G8Zg4+j2iV1dac@jHIr%~(YPPBy!}WyNsGl4PG}t5^g{tt#=yubPwW z#mOFgXf;de6Q-F79ij){8N1uaE*%byOT!ia1}QhB7lVZqAnp>z>~A-?uQE=rnRHuk>LgjPRas>cDUCXePNC_B;r215Gn z{z}ze>msr;;@ap#e`8Z))-M5hSoYaTuwEfNlFqA1$LVrqs&60Tm}x=eg;wA~^~$QH z7e?5=w4nb7a1Kk(KeKv@_aI;=;@7Iw=63hb9_P*^g%U|lYz9~qA^0?1(1t!>@Y$lY?@k&et80-P>-0ip>(alv#xfe0HNZt`w9 z9AJ^V!hpKm)MOy*xIZAenV#0kxec*Cq`(&oN{>jWCum-JJ0Gl;;@cJmJiPgcF6shX z=~HBq^omXNDD!t6B_q?{6wFn{m=wkq_~x5A9E}@kK+q0u{mYYv`zY*|In10^iPR#dQ_-zk@Ek=xe$KDfqh)^YV4p>c=L{6KObAH`g1-EA6q zLlSeSwZB8ZpzsS!H<9xosL)QBkw*i3^iXD&K~w^@os?dR&YK8`lz@oVBrlWg3{^9D z)SFumY!hO$r>?{~jL`L0#bx=3ldEeRt(ix?_~baAT!Ryd8I9Nc7RK&Pp+>zywi@um z?AKRnIe?k7W&oPIyXAk>0mcY!D0Zfc%Vt7b%lmos8WIYzM>Vif{R49grhiKh(7$H3 z@FmTt6uIF^my&^akr`I@0l8k{iKvwWdtl!y8#_zjKEG()%M2O0Vyq$5iY10p5Z(i? zt~BCp{vi5oSYs%X;kgu)RuXWzeGW3zYr``eFHhhV z3|usH=l~I`Vf&b(gE+I+rG3PA0#=x$@|26dF&Ac^+MviJu^%W*98vKV7^`7QgkHv0 zNfUpKKVHb?Z>I^)fSn=1r9wk?)qRH+z+Bd7hbu3hyjF~|viENsojX)&T`C1u%d#O=#`n7+YfbJ` z2K(NzjycSL{i@g2>|dD|;JfNz!){_*KpC=4P5dBRxtv;ga?nZ50`%IZ-Gqp040LK0 z&9q!B0wR)VWrrNXsj@oQxM|aVpLAmtGKBf;!K&H?6Nd|`=ux%v-&pxPB$oIyt<2V zMSVsuJGpDg^+wWT<>kZ~h57k)q|%IMl0r}ncD&1b)vjo8a^j~sG->S=svP0u+;gar($LFsm(zSwr@uz#zZpQn!BoF+ge4U9Mbwhx z`|}pJey~mD03Rz{82grhK=4BI+rZ$MKmHg65kC@EQ$n#^yVI5jm=w?y8}kJ2J3W!D z@B>|no?wc0@lzW;OjDyz`grP0oVcREn{}fDaRi@S-PstTN2btoU0Vn84XSq^GP?Nu z`=vq!i2ysq9vtKNnW0V{@QbBkKcI#(C_pw4PVX7yGbx}bf{gN|o0dpt9 zsrr^cmU1Cs^6;`DA7dws38dk|GPBG;!ky9=XgKlZLZHn7#ih`DyA7Bs$@#DT&cUf! zH&THR=M{$lZ476DtU~Y=3y=+pfc%c*m5iJQaE~D7-K6`w&&FxSUvn$3>dl)5W2~>; zFd_fbDkC=?24g(p zhMw76e_P32-v4p?sCVysv)nr@-B|CJ8t$$?TX%67dGTGxw~XJeS!(IY0v}KR)t6Tv z9a*>Mos6Ryv4yv^03lAME-jN#W+yB{&103K3C`_DHM2*x!0TT2!Fkb5?jL#_8E#wu z(**(w1rM+Xrkzig6_q!vq%@s!!#y~XyYDcP`Jki6{>lPQN#zLq-4^_=Rc>$r9tw~( z4n8V)TXRgNdy2+qAk?vs_AMC*HhpSj=eBdMSonz@Mi2M6-LX?jMDefGNYm(V1Y|JQ zvZ}Oy6=HerwVHn%sNV6+hZou;uaio`Sf>F20oL6YHTq+b&ybd78Ip2-->$)}@41(V zgJbH+hq*C}UW_A`seYCBw14?W1ra!At-l9rODrgW*y<54{JL}O$s~c&WuSTa7fmt3 zZocYBPT(MnS>$`6aRLy5TCBR$eia~!3Kg^M{X?rQ3(;n>-dniFnf}dt* zsi#DNQX+B2mt7mqX>1|`?@I!C3tf-LiqDulBl|fG(jyaT5;JjO)~yeVJS3M-i8%Vy zz2tG^yOItrkGktWw)QKav{7L2tL@hsCQ^;;h=Z0!m>DsdJJ21=-DJJbZ4bi6N+r{{6FVlhA%F@l z+N#E`cJhl8$xfyxxDRXkQq!_`5A_gnq#bUrq3rD9?^=CV>l1+y zxH2~KO3u2|CgN+c;ni)XTXDl=w?2%U*G%+Nc_n^mkx>v0SvQ*~_Y;3a4T>r0@!KVd zqISq99S)aQ4%Zc(@#XCk$(pM~BLl2O(-4|%H@aN*BZj^j3rk$?zlQzF@9Wp^2CMSK zRDhcCI?!e2tsC!tr*)mM#J3dktJZQ zOHy$wA6f#IS%WSbR9*a4^X9$SwtebVyqNQCbh4jWFxAX<2w^u0STxTnpF(B21X|+2 z4KMtjMef?wPy|kGQe6EHAUyrH)+cbcUGyQ09g=1ZQCP#oTDixDo`~bok+AlmP?+`Y zOaBI=;tn4}j-Z*XXQGX}omqPbZFlpe6Z?a{snxMw0M0xe6Qai6Z6>TdIQg?}-^Hf6 z&v}2R^-?oiHCgcIf-gr^cy+9NCFdpljztocQPBB#1qg`Gw4pBb1ph~`nC)1^g8 z6w5Of7JiXW%4>4lo6?TC++48P`{dl`e(wrp<|7nKqo49gZ=|3zS^cF@{# zGfTg;I7gHy7y60of2@+OLPx1t-}kzw85i0;ti3P2K7G*i*QCTdyi{@cnZUF}E({+n zy3y#8IoWTLXX{)?2Z#P5Hi0iTPRzN5T=>je-*q$Y8@OxCe?Z9WgAkBFlnxW(U8YjQ zrQ_~rwri4DD@U#)&{X@u11cSO;!LT~wOh77dp1`3oxQDmhBH@`pAEJb9Wn|b>S}UB zK8fc+?1hI9A3ho+QmRM z)Rev0kL+RGTFWEhdk=d5zr%;a;PInJ=W_x@$7f#ck3Yx!So9{tE?&NTld^FSbCaBv z@gd1+q>`~?0;LW7ui+jYJ7ISGRMF ziKkZOFTe0uZ?M0ttSWZhm3CK+elZ;BiFcfT@HdyXCZM$IXzC|0-_ir~swm8n8^~W< zG^t6GCfS6HsyI+(?QlQ+f>C~d!>P#G66GEotek~*%Ax&aK!a@ zdiO>%J-^*_ZE#J0jz=FVXt`u{?P*MX81B)x>C5V6Lc{>%Mcq2__SK%j6Q_;soWpLF z&=ofD5dmz2?ea)T(nJ}eli+pvfu$>F-ngM=HmzzMi1lw<(-F9#v)d7vtDxPbpTDW~ zi3ArKwNVGAgZd4|pcU44OgzPtyskA(ga!k~6&=TUMhyK{lpqMj%K#14sO6~GJ>aOD zhX)hDyc?rU3r~9rVlBgMpL3s&wAov}{hFHD>^S8cgq>}+5-5NE`R9gKd8JdbMfkagc5N4E!*Vyen`*> zp+(yyY1KrV&b_vD{~nIt0b_Dcf=+PZN3GJZtf-&GBA& z%3M)`q|PeSn5|TB+?S+Rf+b97Ji@H=QCbKJqOSIMzZ6!_cQS$*k>ux9Tk0p~_#g%u z12D4#Bf+Mr&;5eSo+(-Pc3|Sir;|xm+otW>-@@l<@8dZ5m!N+by$9Zp*fwRR^UR?s zwoMKDEhyi?ID%Z=?jEW(Hvcu{AW`gF9^R{|ppv#2a}0z>U;6aJxz7{#o=BS3@l3MW zo5E`cBm)hvBkhs_RMANmwBXK@rl=6Ac?{1 zoDy%%79xq}*VJ#*rp=%~k^ZqI!rG_Bvr6Pae`>_j=!_C;4+;yoc;P~uN0V*p&8fb& zHrb7ljVxX2y3Kd7W7^!x1Gh;XN$qOviIW4%z8e)AY^-AT;ZbVur?km#lNB7C^85#+ zPnahnPaiOHz0V5e>agHj*|)!a?ag2^o12k7V*pBROCEgpxXBD1jIJpr1827CS~@l> zbKJ#B3%+?&5WR2TJ|#gMa+-HTI^k~Jq}s79Q*7+qli(b#90N55TsAnFla^f2be%bR z8s7TRqS%l7Y^uCyfRiaUsR{ZgR{^4 zn~ZI8{-9jVL7N-LOq4@}0uqBwco7B1J9n&mP}6@vRFc!SJ;2zusKezJn^QRWWj-i} zruoX7OwL2_rg%M5C`$WmT|Z@wOWRc21P*c1;$WV(@6|PJZ^xkCUv=g-2bqwHL@7 zIc2UdR^aGx?VouZhU-r5T`Q^z#J<$4_R zJ822QYk3uPvAJ>8Ubd9}okZQ_#1xJKgwT-5^_^)aWgV+g7CriaxZlNOYwlil5WOT% z?m>DD@THO#maSKnp`kCtf~Ior?b=H~AmKqLT=}Xt?cLTJR2ZB7Pa>up-$>`g{ldkI z*XhGaDI%>L8_GN>>Q{sJ4xia*1_}p#;FnxYL(-?wu6@OhD?aSA7^fwEF?V>-=`@xC zPbUK*$hDH~Vi%jWFrcUL6+v^^&lU5Nb2m11H_mn97r z)M5EbQ$csVxc&vfaMfVOijbuTJxuxXUGDtiWvf`vxn1r|z3a2G-Af#1FAAuiQE)#| z!Mc7#*=n+8YHr(Jd)Y5IFNFy`7Vb-a zu!$#=${)B+3+vV8=CPV^xqNy5&QakF?jM@nk=K|n zRBdV|o5jhODsdXQMK*b`Ci*Q?ydC5YP|{z^~I!c)!v&i4S_@$dRPn2dpM3r~bRx#g5he z?p?aksu`%IJHMG=yo-|&uSqZ@d@cWUm?V<~jt2jQ(;o&{vBD-FI zZrSz9U#~3*cpVdY!>L-YkiW@(UV7cISw7B70;{hbB{aDu=c?z_zvhA@cCdC@vS?A6 z)5(Bc-$VN<>JxU|?QZ=>+FjNPh;#SmzNuY-MYADRY~y?=@NT@sv+~ZZN}R>dlC*R7 zmB)UaQofw=CSv@FLO`BnHI76d{*RAytp6J-%Q{7rn<0O+nlstP?oQM31T;e8pxxEe z0tPOB;Wzde9p*7a{ZnrWG}?#}6R&?cCbVScnNF4g!E`6AhyHe-Fa5dtw3M2{j7u(x ziHT7rD!Vm~m*p8%<9ueHO*Z}v>V(|`kN*I|+v3l`Lyxh&pC&{d25-wu$QUdlDkJpd zJbWFdnM`0|RS*ms2FW4u!SsT00?<0AG_kzZYt5C6aD;bsZ%6S63+g5W5lPztBwh30 zK7wIzZn1c$25~I7@|>%iRrvB^U0ODqPqgykx6YX&BVkby;^LR-6ou z<2}*2f0vO13)x($C%a^tDKRDo3`sB^Cu%+$r~KS9Au&!mj2cI#)MC6--EC15@Zn&M zr`J2yNDr&iKl87G{KL~2l_2Pm;XH8!BofPj#C}K;8G+|3gIXF;7Y|R48d8OP3|s^B znU|{+()O{-LjT~(4Tpw+gps%-Yy0u=j`l$Xmq->9&> z{&Lo{oBwQw5l1_W-pg_P?WXVZqW5qR11P4_6z&VtXF#0esN4Z=(Xnatk)U|@#7N}; zn88E)jxkx6FLzfRO1JH(ZMD7OSMI-j!@_Xmg?VH6Xl)$*R_nabJMSU+qN2y70gME( zUYfXHw!?q?Q8e#{Zn*<4uJ0Cm;lecp@2>!zXntMFYS$)c6Z2qSlsr*Kk8Re_c@_%2 zRZPpooY_lr*-N|ms1he=MA6Y0&FF_y&N;`UHc}AM_|dmecddMAPfKIrH?_=@&OGOJ z5r!celnw7@GL>>pDOfw8J;jqZgFiK2oOryP!j@yOUF;3UitHpwl);cIhvdh13k@^d znYFhcd3s&0grs>ZY^eugCW>(-qKy5NBZkA~gyVhSx{>(}wywf)4qM^|)%hUp=n|f} z)g2)G(nR^nUY?}q|raHH-KgICA#nW{>(PA)ssuqdVnjfpU&O6*Jcp)O_p5OSn}5(xyH{7Kq~>J|DA ze#ekY+3@#kQT5N(%Ld|=#`@`TZPWqRn%Oeq0{C8$7Adj^#4uSl6s8S&cU>qs0~Hc; zilR}_YVy!}_3Mv5vG!24@Lzu|YdY$A*)#5UTV)XP_F8t1y|86kc1kF@=@@u?dCn<7 ze37ahvB6B=Fzk<3a~Iu83QjH}AT96eKf!ioaIoO@YT0Y<ljt}G-lJ(+3p@aThH0md)EL<;oV`uYx?7r z^M6Rpsbi8#eMD^VZccQQn~O`33#pV67`18AA`9jm%bFL9-OKAzAvw?P_EKM)vaOz} zBiu5b$dEo%FqtdPxtg<;uQZwha{VN^GnmYnWH4(X=?7hcRnS~=seuAfG=J{WzFAkN z4iSjLfp-yF-1OJ&_H-F0LnB_@Cfcxd9#6L$Y|IvhvR%e|$+5(8dp|c;&4v~i9dZiU zNKf&{=mTA8*1cR;HLe}$h?y&neNorMzckyz`Y!fY7PNqh_5taB178@DJVfxOqcwBb zC#oh%et)+N+UQW~RNp_13$>~jp@D*%w{A^CCf|$D`Ef!1osV7vfmHUcOPLQae1p6d zN2I*M^x)~+UgR;GbVucZQ`I$hgH|EFtu}Mj#l_2He!R89G|mnT;|mAZv~sZEG9IZYh| z2ly#8!#IsQjJDo+H)Pt=`p1b=>ImAJI8biEIoU>GVDsXhdtl&=9<4{ylb30mfva|G z6!P0|wM}Aj0?N(KFL3>0XUx;L9RBX4LQBI2mwyWmaFy2RF3#C6o1IskqI1i{;%c*>*s(zTOrnZAUi-LY~WVkLAX^soNw%x6r zFyebJRwZYu_9EYX_w;+Cw~7?kmPSgQLXZ71PxJO)Y1@qk_|mZ1(|ayGTztuQo1`Wb z=f;%ZJRq3nGpf6%%X_(8h^wQL>10Ib_Yh80RyHd&5R4aT%-9ikg|6!TLt*hR4%@2n3a-#T#Xe$M zA@pWre$XF({1I^QKt_Rcc66-GyYkBmc{VBhohvyl0TqBmG~=hmJf1GiaEBPfX@VGOX7^J{$u znVm?kIxm!hKV`yqVLocJLQt@mi|p-S|kL08OIv;JYdn>8+D4@ z4X;1M7VEamk!g;9MY{ETA>^nS`ND{;f4QC#1*bVt+`KOQFdw9Bou^~NH zGupOT!tFY07kyjx3q8A;=AYPNYwRa(tR9&Cz3gu~=EhI7G3I+9XJikF1tfZHiE0fF z{S|2|+6FT$IyF7~ykNTP4e`>o&TipHvfDkv85sH0Nb~aFe%l1#lNLI^`gK_gkWia0 zmJ1m&g9!uI$aGzO-*wEuVIaZlxf0AV%G=H;Fs9DEpZ{*J}AU3{1uiYLn#r z690Gnb1hFvhq3DlIecNMXv5>QW)Qvqso~-~*)2gf!QtmAE@pmG}Dsj9Y z=Zd&~Od_^g(e~=tG>*l$L#kYw{JbEU>)R82aw^u&;ix+`|f|FBK zJ*(-7f!~ObiO{Y9;YjwUOgV%CYo`W;5hNO~IHALAX6^Zn)>~>!*III`Ap*++0d^_B z1=ef`&Sm=n3L4ods6rVUP^Xi_)1Elk-LCmu?^#AD23W1(dKz|5VPl_S8!+|_>3y*i zD~nn=w!994aX0g!@d-Of-28(3C7rM&NiohGCUW!DAa8T#{Mc$S7ktwnoXBW&`)3K{ zI<{^>{<0>YMBF%+>d$C2!+z+iw=|Dt`G2?Rd(-vNTh~|m)ySGVjOK*_B>?G=ePNN9^orSkaKMH8fjmsR^b{rf; z?E>4!FLR$ch+4N(Yky3Aywk;&VU+xdnYi(>r8wCb$GW+T9DY!Yh}fG`mT3r zU+9f;wtiya!m7qe?B)F)`|Yscuy0AnjK8x{)pv0teTmMHmyS5UVeuCx$NAtMxd*hP z%e@&f*~fU(qu0pyflbdu7am9*lSHEgtw(;r@Q_WGI?#2;+0)4y+k|64U0AWuP9%0w zc8V`Z`6|w_QK|g_d4wc!wlp5JzCCS0P&id-ec7zTtDLLZdfuX2kIgOco`*ru=Exv| zrZCpR@UqdCs{JC*6J35 zu*oo*+MR$6!H$&hQq4{&eo-4>op^|0uF2~hv~#YG1QFB$O}&yI)Bf$m^FdERK^axI zO;!*WJs$_NzUV)9^Eu?NC+LF+>`Ee)GoA-TSbrd*z+`x#TX9GfxhOf_?8UG1y7%Aol-! zE*D9GH?+9X+}Nq03>kA^IB@5z(aiK3008T&)S!W*qbIN@PYO^is~rvhr%P*QAFFnW zs5Md2^vre>LfW>+K0uTL;FS&3*V_@K6#PmC^a^kxmEu+{Dci!&4aJsbGO!+tm-*W` zkLv9vinD@!P8F*xU(HkVoAzll64+V~rOwCidtI(OY<5);QK3#M4}|c%d_0NP4oc_ zgtl|sskd;od9!Ak^PsF!Y3+E%g=(}ZR~pCpu?3jRn4Ux)tOlZK#=O`|XcGvP8itzp zp#f%FE$_?tWV)K5qHDw2NBueoum|MhUdOaV_uK#isk{nX%Cze4zB;+_E&dE_HDN9B zLMvAZx#84WdT!9ooqEpoezp*}38HSJGh9+4tRs6LaY@lZ(0{^l%d~BqnD1E(DvOD! zU%OkKXfG^N4SUnk3Rx=8<<(#|pRQ#w44pXU%DOfMbFQisfACO~Oa{&<^cj~B*~k3j z{V7-d8oeHbHwfB4M&BWrgm_CU+VRv>xU}>E#D3?r&d4cx#C_4wx^lb`!JzTbRaRq- zJqC{cVDc@2wqpGVfLy(LwE##PyOJR>8|y?Q4jZ-(_2M)pK^V-gz2~&kz1|FY+`xEw zR$Vu{+hw!QU~j^WsZm1$Z^zenhGF-RP%~yhk9CO%D)AEpDqWct$w$ivLIdO3_a6<5 zsdU^SNHe&1(lzXu+m14T`K9g7KJw4WmhD)yY4!{Q_FLkKGP;aM`2mO?PIi|5a~(sUtV7C!4xGDPP0 zOc1S5v{z!*BOYTT11wRWo@pvxfh?|1wSl2qCZq}NqjztT+v~_Thys@{V;4(1e$|&b z?V6MfjroZgSF~cp6VeJXf3sV;)VdD;E{$<44ZJoc+6XKtRE8;r2K~$Q8wYoTMF~iR zuCx*Ol@7IXRna^lgEgt&Iw1J>-#24+sq9WqVQ&{D3+ToFn7D3m?L*+F8mh@e$jcwZ zHu=de``Dtt$;T(lf)4Aml1@wjv$vbjI{R3{9k?UZDvhKICm2kBe>B&tPkze-YJBw3 zr^2`_@T*xSOsQagZeW{rOoS@>S-Maa8x+^ynWZIIw{G2rbK~)o6B_pl+>KSF`C{2h zM+kDzE50~Q&ZAvOw=3qE!3VSPeBa-5efn9;In~Q7+=K6g&^Gv(vx{XJ?LAl^E|3NY zdKa%;>HEVEKX?VIcV{`$ofG{l97YQ@iut%oMIaT&%#OY<`Zy6R1?5m53cVubP|QIM308a97I&it zpj}2EP16+>bwC!jLLqWy_(l2fMRPGp;9`aJN^99?&p=PYNtr8(9&*d4w{Y>$^9`G( zfvw%#Gw8&o+%0;MK}A!tNRb*>PT%GA;0~;5I%Jk+g5pDj>hpQ)lEJM59smRjyfOF2F4QR0K`ZG>m!BH4 zao*LO6?DZASjap@MSeyG;i#(X`9O^}j-*ir8~~X#1GyQYaHMq0{3y)dSF7p?%Ux_1m-s3~8^WUCMX5t2GiIqu0x*AY>lW_!2FHt6xF$eX~ zxS&32=nA#Csh{nq%aZ-1Rqxw(1rNGHz#-ofMbw-{$6Dq|%8anETIL!&{bd^_2UFYn zDZPVRf0OPPOkVrYp7u0HyNV-%U6q?T{&C3>KeS^nkf+=vm9BPqv8F3vkW}nmZ>1g` zQe%FyjJ~q6X^q-r2SDBqnv-PS>^h!@S(*R)(E7oiCB-f^K^-7e1R+bnwJkl$ zdW$n}KVp;i!tnDID9o2H-ncK0CyJ^M&5%dGra@3RUoKMo#B4e?`LeVhblhjf>Hx+$FKt5{M+%xrX{dPi zHR&L`n8`uB{4-m2y%;sfyIyh3hGKl%vEM7HQt|$1a(%LX zxt^zcH5qpBXdm`v^favA-eOp>7ALOTFgcIp_SKzXOGa}Rl&JP_aD|uegWoBg+=k5} zA{`JOk{cx1J$jTooqT9qD9K`8a1Zyfp@H@J=p8l&~>_foJI#RvSGO`p7ED*0(AtU;{4O;q2jN7 z^ykwb-2Fz6lYMC~MTbv%ya&bDalYOtEm?Hn5OPyM*BS6@h{jfdpQW+hZ&TB*%5Wds zaT@t6HLhW_jRK`$W2QvQP$h1HUC8ia@prPEi0uXcY&jx*5HgwrL(7~n&H&fzk9W_; zkT#UiT;iY#TGoTDyvW%sai91VkIgcstf+ z;j3U>jMZ}Q&XE*Uv;Cf(0GJl4ujfm?_+RL|2652t>D_&T68sGZ{^32i)bsa+=4Lm@ zPO_E(WI0afdKv>vI=oykH&y_tSeh$_7fZOoHXtdFT-Fmbl|9p|MO8;GDkrWKcmjZ|Am@&=Fvu77q6~n?IHycuie7jXOQ0{l&p{m9=RCb`olK+e-nr-s| zZ`R+s&YWT3ChLy8hm4cvd@8G*)ZHp?XJ-_rG~F$NA@EEvB8|=l(AU9Y z?a1?3TB67{&2FE$=#<=W;DSSJ0Mm@Ws$IJiuTXYLOS+VFY!y7*qgd(Cm~&_)L^?b` zN+J#q)ZU(ch^p=0uZA%Q&DdLu2r5^!g~oa)NJaFd_FCA1_yY zQXdKEQE6vN&b%JUO#sHkHmGW^KaBsC*Y(82!%{Z$Z+ za^iGybcUu8T3PKK^u8RlTxrmM{LUm%IkvAn;%9$~h=q0LMAH&~g#-aYPdHa9^*v{- zhQaV-(~owXM#6VB1Krxz=Z(l7HDcb?YB9@fMKsY=AJ^Y1w~%XQ#~J#inN=)WMawsG z9$$U1=r&NqZox>r2+;(sGh6+6y3xm}*EY@PerwmCA#_yzBEW+iPT+N?aySmsDJSB-8aSvfY$}YAdmRj(#v(V zo)_Z#%lf-;ws8$! zztZ?NhGamMMz_FQtho5QIwnpp*355WvvF0IvW_`Vs8TtLVrU!c)B+m}&&1;jlaVYv z?f%QswHuTE_~w0?XMFgEit5@6@FfQtO@(l3BFcArm#eGW)vqtHnL?GeC5OD}t7z+d zrS4=f-x96dkmpH0LWSodf?`bcT3t>MvdRMZP|J~pD_$*-pza89A4X=m42 zyITfs^h}bN9IP@AOxt!erI_@i3!hh0MzyZW`)FfasMYdXRE}7+gsS1JqKh0cClEzg zo1oezHUav8_Ax<;$Sa=0M#P5sD^uHDXVVMHg!UKd%o`RK+%*>}LwOYdaMaQ5K~U7F zM1Jgt*!<~|qBBP`q=2H7TWLfQfE9un&wgrhuvGdE*rR=&^wH3N+eF3{b~qKbdGly` zH^zt}iy*GX;OYy$nU%$Eb??xw(nm6xRzFyCvf1qYt$Q~)KQu>F7UI9oYXes+efkmI z32}#K`B$9>>q%{UXya)FgB#Q=Mg0%2_%id)oGLz^Bpygs6HU9gEptwZ7^BU%ozqwD zCn|$PYcm_9J#Z^D3|sEEdijYixmf37M!IiYKF3XI-2e+%zo?Fv#nf)WHyT){hMdIY zL_fr`a<~`2l~Y_4PiGDESs>9kWo;q*uKh0SeA3Bng3FOBb5_?aimE@wABhkRSr!P8IxhPr%qdktxhfqfEkPs$KAtpaT)^gBTU70Xhf5h7 zXWS0tVSq=*D?(42xaE+XmGLh4hU6b8pTlsY`i8KA&dn@HiXfltxuX*R* z&{0YxyDm{csaN`c19+XKX^|AH{FX$UY$a>&uX+-6L8G(L!F>f(V&8g?`^x>2sqzn} z{(cSG#BS-0WUPFHInjsJJA&Wm25Og6>P$@JK_iRs+Ss;t$>SWs;)-M3yEn0F`zDDw zuh>axN>jK-&S289T`2ChLA>dl_PZXiP4b;9MUq?Nd2g8WGO)e-*y$8_@`vV9i6bp~ zXr!LpVar!>%4ydV1{AzE;BY4=r^IAIKOo}mQw20Ft=Y-8e#xVi=cGlN!>9o@)`~wk z#nG}ptsr{tqD=Y;o;pL!dqX>Vnn~wGo&gg0Sqjiuk!^2JXgVSB6l-%7hCsaZgJJRI zQWQ2V1h$2rz0b9ctDgC3-AeJn)D2MhrKnR1eT!L)p$f$u1oWv+{5APTi&x^OqID$k z??V*_N?XUocPtMJSd>ua3+MK(oe!OPzQIxAQ9|Tdwlh+)4%0Pg^^7BR{{Ihu_5<^x2cRx*C{;z&k1JoH9o<09j_$>FLi?97g48GzZY-t?6fy)gA+v7YjRO z`_|y=WMOq4PieNqryuSz^XScOVy7j4QQ7$Xic7dQ5cpJd9&wgZdUlV$Nj&z}kLQ9~1#p zX=et0Bp6Q5!7MpSJ$x=SG7#}ZV^fK;AtHz&WPQa=sJDfXudzu{Gd;Rsyv!c+>DbIp zbBvWEo{hJ2MnY8UqAIxF2xbhzrE;a+C9@oZ)17(i4f{55?~rxl?w^|Ht*Xy8bQZ_u z7OHNy!$zc-(z{C0H5>nRc7sT)GQQMs5N&I?R?SVk^I}9%%>jcADz375eAM$pH97=S z+UlJLzWQC0>L|b(usutLsK1qe;$1;|bs^oxyOSK!s7Zkyw89ol;MdyqONM@2y`6n- zeAcnkZ#<0eJ~X+B_2Yd67MQ(>T+8|}iio?UI7U4)i#~@)3l0~lG(R#lpt2>^L z&nZlbnsX7} zpdbT_?rPTL6&BD(Wn#*mv43<#l5TZTv@V;>$h-7GY&QyY8PubVH$heS-571J9pfNVdUm>y{DHG#eaIO(}=_}6k($o z1X?w3Ne(NG`g)qsYsmXd5R&4I5O_J~4Z;;-n8A9oxB+j-eOwrx1S6s^RE^#o#5Dq{ z7pyZ{yEAM5ZxP(#lRr)ivLF||KcQrTkEdlhdz+g58MxA~A-h4|DEL&D?S#$`t1DUF z%zBXPTEF!MQK&TnjJ6}) z;1u)preuuQlszXtWHL1hdCkc&nA)`e;P#=s*zOwpui5nm@0x}w==*DMEnB*(a|ry1 zu#NMtddA(-2z8SY7ue;!T-wyJpVBnKf6LabB|Ps*wISlEST!^*!-#mFT9iD-r@7Sl z-;#Rag=^0svqWpuoC$Lp;+Zfg7kuK9vSu$HS!Et@wo)zP9%NXPpwT)evO8CfpYrdP zyuU=ke(&c0D!z?>ZPSzImPiUoPv@v@v6|j-KAL7AoQpEXBRZ}0#hk(om#DaSOTvU{ zG*+;2xO32KblQmtW|$*=Fs`7SE!Z1uq+P}*pM3dG&%pTS4qrFuQZiwp4H4rfd;p~h z%<}AK&AQ<#ONTe_(F@(Nv`u~c-u_@u7Tx)+ALD7!pkMfMTX<{47m5Dn-&*kRId<~p zS*P3oQ*(Q@yXQa_pedec7eAkKbNed+0NPmnQN&3V2eecTn=b5#`YOLp# zi-kL{?B|p2tiLJ*18Ug-2=H0$2$T%w?Wy%G??Tib$^Mnl)`N-&#X`cg!B*~8K@Yc) ziAxCooAd2>*W%VWT$E8TO&?AR{xgd{scRH?zjBG}xp?UpKSXf?`oTMCNx`En<;v&?Hia#>( zDf1@{uHM>3jTJz}Xwa^ZIS3UVCv&vf7I?s@3E?FN5!qt1 zz%_6}lj2tsKFzYMd;e48fA?p_h{3)cEBXosg44h#0gbWK0Lra2HoA*m@3hFX!QHo~T z+CvTiGIP@QR9yy=7{^^@u1}V!MlV9~O@vM8l9Xr z$BJMJ0K)W@D-PqyWWJ&a8kU!n+wb46b-nkGua0pg_TBwv-!E#J+L|c%t^wyRY$=Fd z!@1$87QlxqT=3l{Fv`EYrk=y?|3$Ug#oO`3i^e;;kaO3^k~(* zt7Qs`WOn-1RWbShUJ?J?)gYCMV9NReda|x4=U-7Z@PiseIWgZR#$fnya*{`BWUE-i zdmVzmi$^K z4AE~gyP4c9t+(MG66X|USdXIq4MSE77rt*gS-yX+b(+%yxfx>f8sY*MBjp!E6ai9b zWC^UV0Ly3-d1;-`?BdM1 zMamid*$wr|hV0(+-*0N+<(xoHK#62{2RGj2WeFEqgH74;hxXdvgv^apk&J#0KRFCa zyF6iw;&*++LaT~;WBi*wG#Hj-kFm7(^kw!}{$Ej31!`ktvftSUrAR_(Gr_yxK~Z1S z{?)HoRV?$rk*4CNP1i~l!V0(s_0ZM?IQ*OD)+i_0Vru{IcUjH-M_^@qPKO!uj^(^P zT5njj?biA9$~8Xa>AJ~Ru4&kXiSlb6+f?E`qy(|F!q#4?XAI8-Mq0j2UAPvQ|bSvNVk? zQem=`Elt^ljIl3SN~JP}(L#~zl%-HovP64hN;5ImNtB9`y%Jhz`##RQccuID{R6%~ zeEo3GJ&oS)*YbRx=Q-DPo$GXW0`~bW$LmDfPf+P+dC|+ji&3JiQ9pHCSsN;T`SG~g znP7FFWN=Itf@u4-5zs4mbZPz;C5^8qUS!W5C147yuJHGo(&_gpFO*IyijP>WT}Ok} zexlAr>4TJsuiCxA@*-D^C)alas_3C+v;}h^a5`}c%3K+P7K6SqGp$-qSn=COlNui; zlNK{DKwLjHk^D+U^G}CS=q;an^%GxSM8kN;claAioX`iiB3jEN_Q{_`ibVbS%K3wU`sVrQpM?YqX>rbGUiMJTp32%LMhUV*6-UD z*3;&Uu9eIyek*ZGQ<4 z1L)+$Q~pjN2+nobcsgGtHlzTFcM@AjXM5eBL>t>9w|fLbe2RFA2C1*-i19@0b))pi zO%;JkgzdDcQ>%z*YRn0NHv}|20vFIIy)YPjV?kBj2P^*BV@lsGAXq-*eSv3Mf913F zc2}7V2{y6Kw&U;gSIU@6Y1iXigdr1DtG*Xo@d2wRG+whQXJelf#`zQ!U5L}r-BQhO zYfSwr>-})C8DRtrf-Npngx);-OSfDaqRr#!^UlMez{M)u{`<-2`2a5wvABO@7nW&Kgtvy>FN?F(FWKux7iPF&e@8Wm+&y)+zuks_TN zw79qbr0k)RFFI<_$#rDp`-bBl2yV2<))5kkag_+9&@w|cutH?sNeJR@@tYI=>%TrP z9sT|3EjEHi^$oP>rJcS$UmBXG(HEG;RY}hXi!~4=GE7_Me_$qP{{1h~or%7j zq`PwGNM?W6{W0iVr{t~uiT5`sVJY0>q+}QO|6Sv>%WvS;?XQzlc8QNf_$3B9;4W$8 zh(ThLICSX8qj@-Z5~64NegMk;^2lwYTf3h-O8cJh2Z9`75GBoX7yyfgUf7eaE+2MP zmnGDg39$+9DGOH$+>Q$OQ|09hYIGcJvRp+2CyJyr$!IAN|2_{R*5!*^?}40Y$r6(G z6x|x)g7Ibwj-<)DXUKZDzwF zidK5i;@cw-B}}0$^=U=*e~KFRr^?v2jj~*e5ZGbicf_Pf#u~xz&?+FUf$)SMgN1?E1ZCpQZ#kYYlVPXOO()-H& z$Q6#_H_`(pJB}{2r9_Wk3&rnWdiT@t)CdGfwdi$ma{2J~_CEdYEG8?bG!!+}cYdD* z%N|V|-~X$m&6&a`@nV=MbN=W8dQjx(5#sy^J+|UC^V>3ouzGHS4Gh6y0CZ71&DaZb zwfVK8PYB3*K*v=@E9s}%E1GmEJ)oWU?A_bb?4y5j-Ftn7!QpMu^wVj)^XBuqMNGP@ z-H0$b`tK{@wSBsF-4eUo*H=vGcBl`v1ee+qSoLLGI$~g%Q6SF_8TIDY(JL|)v@YG6 zvUEm1e+PZy_}fc#MN&bEe-<5=Qv@Z2+~53MVSOhxENzkKYqe|1PQ~RMO=_IUm-p@4htSKTQ69R<@+A(~>F{3~>r7TLQrsBVcrh%S zx~*8KH25GHuCN)(zO56Sm5h}ZdyuA`U(b88Ci*j|qSv?k(>_n!d-&5nek-DX9t5S? z>QG@tA+&vcebE5^`dc=s5*UG^r?yrnq0qmw^Tmw;1bi7eQ`62x^N#$ZwC+&c0$=j5 z(U%-VZp?76i;UT0S<{)dbf2ho=+NP?gBVj%Fl~+A#whIJ zlVJ?g3Byv3Qf1Q>F=MY7?g|ogXYq$~R~7oY+LvL`xpC>x1L?`g@>q0F@=&zRgbx=% z`ZxWrd}lC3y=8}&HYX3nY&kC75||HBN$DzLJMPIq-ZC1a{Y!;prtae#AA>zCLo+v+ zn|lJ2^<-We5e(Af2>F)4bHtq>!NIqul;QE zCjXa8O|1SImmFlVcj(|chUpeL72qhw{^-7DmEPuw(6^-Z)!7bd_&pv4W*PTsz!n?& zk7<(@1(xn!=h{z~{~cjw+6W%9&?08TjDj>c$oOiCTtu%}W6ba9tIcE1j1&qSgx>f{ zSTC|r$?gP9@8*Ep8to%UMB06fYxw-5+O^Y&>V9G_om&Rp;uc*4i9RfE`v8<43p4HG z->awG0b&e=p8XMMEHO;`3r&sSGPc^@Z-G|VA+=7M-ZiHFmywh*PwpUwVk6GeV|bvWGc)06KfD*9h!0q1{-XC#2N zIG@-hw6A;8NHTjhLS{z<+nsA2jr-BoQG{19{*mUj1HU}kJEkuqpg1+~m1b+92?U0b zM91^vj)qS=`g3Cl*?iBBTVjo>oH<*njTdmDn%{A>vt7e=_xhtnz>27bHKUR^<*}np>vwzC{FZ}N zR~rxMX(O|o26=0YlHM)x_GLw^6M~-DC^oBGM;&{-p|4*yZ6Ly;bDWL+-X_})q|x=$ zOC9n4mGL|Xa2+8uDX0cImoA4sS{k6}v#Y(9O?lroI;vME<$vq!hBjAb?8(v)aVpG0 z2IF+dIzk~hf_b9~D-dd03F6z-4ciZ9u;5QzmJDxhCiv!4;3R`0fJ%HEmV<8zjE{xb z!wX~Ka1;>>!hD=BS0(!s4c3m1`dV{WH{jAL-isQolf)@Zt~(}Tn z@g_iAER9*=bRCSB&TiCE+69@Tjh@XD6cPOhNasz=>F8tUrDOeJXy_d9+>C+GOy~@z zNt9uoYGt3Hp8uOFJa!konfT;E%S{CZZW+BR<-fW55qS||`TFpoVp9Jvy+&pP@$-F` z1yrc_6|w<*rl^9BySa~z+BI+bgi=ogJ;DKk8s8;IIRsYsZCDc)T8)o{u{Xr&506sO zaLaU3aA5-8nkIP>7FPl{9+R6p_)BZkaAdwxDAE`|PIvZKM)s|GUvBbAz*HjF@E4nf zL>DAo5)6@NaRIM4X?zA_J>D?7kv2_TRcyhKOhF-^6!$8o>M1OCYinh>0pK<%$%$NH zk29vfFod&;T%njCiA@K3v}ev<8vNINZ5SC-oLH#81TQJ;P#Z#-&Se9sb{o~Gp4rIDYd_Cs40bhs70Er@ ze^o>Gy7+p?;J%Tk4e{m?S)o`u+bgWN%B2V6 z*pJ)L^t{|OzbG~<_1sx z=OzK!r-%b1OfiFs0Ptl;;}L`%dhyO~MX>?Wr6^`Ktc#w#hioa7K|S8;a? zW81XzfB&075=9k)*I*@woj~eu-VDb8_+^WWzqvFn5CTX6Ifzj5iTH_G1R`fZfUs_6 z!vo#awsD-ZST4zxqK}1Lr7ek_G_c1n0A?8TcC7&V$PQn68y4>TuxRXBGtpL3>9iWE z+2{@%i)fENK*vY9i9hZaSFmEkUL{Ct$1EuBs*-TGG)UF{XlSdXa~*1<<p$gL=i7MuU^oU6d=3JXJKGW8xX;KGP91BOJS80as(>?@B2 zd=`<-+y2iLS}|V2m-x!@rY-&1mUo67Ma!`zu_&gL#~$wj;1Cea@j>6F}?WH=E$BxMhG&MT+w6Q`f# z`8yoQA!_>;IqEo*tOK7nASj#eZfTrF4Mw$`PLa@VOEN4EX=51u8RIQ)C&!+nK?1IR z9n7A2YQc8RZz3@vd4Dd&(BiACffL2Tf(F?hpOIpZx77Q-D;|EkLo7g<2}IS#-M=3$ zl|P%FsErgu&8c$wxi~6dU%yqM@J@wmL%)MN9)jQ%ZiyM(XU6zw&N;2$@NAA=Y+NdO zZhh`3Y6!O5WYBrD)B+38Qo)CKXEwf&0vmJm?#9{v=fi_?Uy(&We()rfMkf?mXvPoe zZPa{y{kcPr? zQ(l+Mb)RK~oBn*}nCIhUnvUGw{43S}G}WJS;yX*LzJCuL zIz9K}l5qp|J(;HAI>+D%|12hiT>}rWrH8CX;pG+K zE7HRe>TF;*M|UleNJapkEJhNCGH$5W@#!_nJNBD1Mr+qNtMzS_|DHG`MIZ;zGY>}p z41D+QotB>7DrbW@La<%pf0`wU6&9!K4o>)W711l^7rUQrdmYF26Q@u2LY|v?1sRuC z_wL=>cqEm!b}}`^U}6DUrsDSjHN)9}YNWa5uXIdIOwdcKCSX;%fTK;dIN+@#Y_Tn& zUt}t_&z0|Zr)GT){2luE)_nNNN3?zGgQ^siU?ndI;M3>N$1*gc3be0XtZLB-D>!Y5D3Da{!@dbB>yQTrLb^7PrWF{aB~9b5HHki(1H-2mRZK)-ph z+D3D*=HaL04n&$z%db1@TBNZ7$-kMpp!w9rDf&8f&PvxV$Z-8GB}64N`% zRS%~bkl4xi^@6f?YHG=mZt>!iA^M3o(BcuCwtwBasr8p9J9w<}6Kb zDPxB2&icT*(TjN2pVX@@rO33tGcWe~VZeayyt3xseA7M6r1yjNlYAZw(}+8I^gAsr zDu)ihrk-T%eYB20Fnt$3c~nioDz`33FgL`KfvBnqw_C7ih2fA~leTv+SCvH?HE-Tr zOGihC2&q8|ISH2g>>9Tb$`zIKL;czGw%^4SAs=b?$gQbLsOdn@s}HY`c1O?3%F5HU zj>$Y>rbl-5Fx6t^+?Szx*5gzwC*s}3EG z@3+%`oVG3ZSWJa`jldWa7qFt+`xO;ql1eXS-e*#oQA>lzn`Aqj0LY0cYoxv6v3wWFXk?T0^(E-W>WoX@QSXdoE&2{0RbO62($`1d8MOVuS}; z-H}D_MeQgZEXT+1>u<+nO2J9_1S@L_;(MizTV<2EcKrdHGm5{sLll{$p>coRyy+bq z8;cC-n_RBR3yg1H$IdlQRU9Ha`0qn7d4J|UuH~^-oqo`krKq~AUFV#^-@2rHqBbDJ z=m87Zav4)lCZ<;EE7#7sCiUCDyuCDVpO~BAn%VU-vqvz_?L_tZFn>$&iYLV@3a{jO zU*WMTZWv)xGExwShm%7_M4QzC%`pL!mna|ZJiTca7CWsfsHC;Djsp|!+P%9iimB}= zFHJ{|{G0ABbl*wpQQ&(5lh!TKpPfD7C^C>HwU0u%BTEQS%Dtx}X z)|^*G<$;%4yLP7ZQ<<7>$L7J=g&uEJ)m{7cb>P8o2O2UpU7m3HVjB-rQ`6aw`Fo17 z?i{A&x$QRcv#Jfb9y+P-+bQq2{T|lg`W0DHL%{JAtef_kUcsT~v znE-tH#v-sj;+hZFEx|bO!q^?1)i8XpbPTA=Ens5qEnuvt3xmjvTjU`J4qUQv`HnYjArcIlwMhvu=UCyQyz-3$zCK!i-C^8VF&g_ZN%vbj+1rY*~3~TE8sr>nQ(V`R-q@7e}ojShgmz7n|zs8bBH>a&~Z6Cec@UfRi zAZ@k${dQYKDLev78wr_HEG#U#Jtt*0O}ek*M6fb7-3Yy}c3#oa%+>Ar>_qDkuQ^Ju zS10yowL+-Q7ugtZ&y!k|A%; z*CQ8+mv4K^CEbvqfwGP-zhN_oM=+R5t57iVVddkBgbv5tlCMs!xW6m zTFT~o9?*z-R3<*izQr-N9lI_E+{g|HB88MTGAE(mJb!boT@Q>4Dydc7(FzQyU6ZP+ zK$PCt6QyqGtXv9KN4F>4G}fqFXr)fFM-V%XO}d}Bfqk-h^JaJV7dZ)TExH;H#}=Ut z<@6M?1=igZE|8=mdGm8sRY@Km(e1|>&CY$m*iD&;WP=%=0V$$GO>Ofr4`V`oVA*u! z_uHHVf0f^Kh*v~FHrW_zY_GFf{5r>7v*J@GRlSiPbWpANIsk?}2}}_ z+WHlzqPk0cl%B4xC&zX{I`v8$og?ck*A-+8V#l|mYH)`mPT>=B85-=(pc_&S$qLyQ z@1>ZTq||mm8`s3vm7bO} zsi&0da75>Zhov3U{1`y8PByW6I`+TRi#@!Dy*h1>>)zcK*{|fTi^*Zh+D3DE*2Z;W z)6`xuA+;@KTvWUxBzJr4nQYh5wlC?-w%gscNTO^RHoQZ}e}mTGRoUutCL^lV!x<>b z#YJnEavB0BOp^okr?B7R6tfpp0kR5IRj18aZ~cg(MpZTQ$5$uoD8M%l8*bo8hW-{Z zrQ?$8apru(PVCd?OzJ0Pg>|%YW4?j%v7P$~kjx9OC4H^PNyB7=DMG*%t|>vY2v#%)yL>BOK8}YpkV36*Fbe zS>-2MjA^*?C>NtG{)Y~A;hd1FV*?&(-L<6DT3E-X`B{#+suQbx+9R4CP+p8Ec4*qz z?(WXo{OiUCjgQK|;{4zG$9r@XIEkJ&`_{c-S=_6Sa4LI7)rp(t-D=Eu@mV8GP0J5{ zMtCSbUaK)!!p}B^H+-J*3Rj>~-1aGdI{odp`^ZV6jPCp?m@7I{X#Za9pq2#_XnR zU<$T$E=F)1JfO=~ju0}Z<4k;uKpHKwVQE0K&eE;@&cw@m>++)!J)&q8%R6~!rEakk zJc86CVdvg+j!PX7v7tD(P0>f{JWcd43$`Bl+VBD&v;lV37HL8;RC$UZKOyU>mptJM z#-M*YjyP)yUuj;bO!79ZsK3RKzpyZ004*j_r1`@lS}PwdDOb1O2d;JvUy07*JWM-l z^o`dY`ps;|q3=nLM_*jsG|1=evAh{{2O!)0MdZ8TD6Y*eQ7>AL)TzBKl^~tEpkcA~ znzu%m%USlbw!T4OxWDf0p>gPxMBaDJf1w26iOZ<;z|fmZCu4)=95?mBR>1#Lm!qn<35g2=oDm4xVd;@=-BO1kA* zdSSnz!^#PGcgcXy8E2I@x3k_yD{&>`UREgEf5A&7n{|BwpoXsXr4& zl`1d(=+E^#Jmw)C;ccW^Ry?eG8GQhIddFr1LdYoSjH;1KLYZnLZDfAwQVN11Z`|si z4&g1eu`FY?Xg%fPN?tf^z49wma>gK&-hog2{E9r>my%?J;IrF>RG1nt?Bgy$8IqM4a0Kqyp z$UnfG1%f4Xjr-PbI$R^**dni=f^)K*?r#!ms_)RBfV70*riE? zipAfo7&w;wzfn~ZBghz2uKPIVNO$l@{p)5}tlJNkocPR=wTYz_HiA4v5Fx2v2@AOt z|2yE=re-|salpjQmEoH@()5epsymk@K#L5L_a6FeKT9QfWsyvK@*=wEvB}#>I!^X> znmx0QA*hoKo@~j=R;sp_*B>SpO%ePZ?>KbL&*VBQ!ZGfQyz!q88ak(#Qm7Y?RAbsW zpWs-+ryx^u_{p6c#+98u;q58HnXoIlo@msjY%O7M)Vp$|9bGX8@{wniZ$wcg&$m3h z=2N)rAT8*Rt2c9|km^l1IIuUyNx!0W>L7SQkL+2XO)o%fTCf#vBAGOQ9eiGls{!-& zScS7^AHf${-=4Bm5-L9R8d{?TKySqe=L**RLq{qW6d&*_9+QM)5bct7#gg)i6Nj6b>yQa}FhPYPmAFyarstdL!>NWFOUX0MXy5MNM2-lw!fTwye)tOLVu3D2Pnngdp7 zhh%0uZox?cOs^@^XWWzlAC?ETlqHw++WAi(*AFKQK-lucyRht`p_o(5#@N#C-zj60 z`B~icw~nSsRNA?FXRY-3vE@*0u?w41-TzcxO?6StaRlA^wn^@7{(L@n2%`dJGe%?( z0@YEh=UmTXkC zouv;AQ?Fn5f}|hLf9!y|cgjrX*y>NO?~p@nz=^g`Xn58_@|xbS4JR!N)8)96+92Z5 zT&K5ahOa>*cEuabspWymXKywS0vOJV#A_I zMCx@fgR}d4DQYQtC%=AUC4m|DWgjTuz(GbvV1usj59K6rs#qSIYrGok`LPIZZPC63 z*|5EjbmY9XtO9lH<6t5H9lJ;E45^yy9ZV|M6xEhsiT;!8O7^3k0-Y|I`GKQ+e4J~F zGp|5{s>^okhsb|mvyYSF2*$2R9rBokvx5d%Q@EuClxO|y!b$3kfeh6UE9tK7noyps zm&^L-S1i-1@+9=qku_QXIWo@VFul3)!*3eoZ4XRbIE?KiGjY7xpO=%ICI6Awe;6L= z4m-yHOk@0`*MRKfhkBGB(&gC9|Ki!7Ee-pEA6(IWae6t<0G$=T!L&gnMLq`T@fYws z^opxRPeL1&2BdSg%f!7``EdS%W-$oo-VKvE%N189kODVyg43sW(X0!LD1hbeifno(6H3wb*s&V)+(@VXUJR|izRS!4%>evSM1=?zp}d>2Wm9|6pm zwRN8h=PvlMuu(yU3pnNNK#_xD?=v4setv>Qvg^RdfTDRh#9hg>af8@|!=o}~IWzc>w>3DMBYM9Lj&GgIM(<;tvrT=?IFEFymllw>k{bsu1IzR26CV46 z^bapaMKASLOf`VML^2KLh4yamo(&GvLfznd!8{goF~LrUiML|d);oMfMHU9W%OY~N zN){gBxmv=HDdMW__72?VQ++fU#&I$*_Mo(gP zw5p?{qglG7pYO`zE(+U6q|2(la1)q@ls5%E&k6Q>a`t7d$oL!?dj%7KTuC~`<(?zc zfl?L_-R*D{3`)Dl0k(f&u`xs!PG@&7DpS>YMZ0%BD85(M4yl_2!MDHNVPlN2Jez)v zT1P|j(FkYPk~4FY(aa^>ivvHuTMiq2zRC{pkrj9G!58m{)55oUo~@{>0E0i7(A$yc zq2oU;L_i{@f9@X`=Q0*MTX&X(A$NXc-3oj?>?m6b&u;^-l!IjV)l%>~Ns~Cc$|t2W zkYcE;h$E9aBKsa4d}`KYavsTS0-e~qc@Zg^UVM4s8Xt+9`KASV-iH09FbI6?q$`jO z`lfbgWbpJIQtmln{ldu^Pqz3*&`E(zuiU+*w_`5TQ4~2QjZkVS>T4$)xZj@SdBx?d z_K0id-0je7IAUizMmX>D8P9$gwD>093%Lj*Wp=&+L{7RpfnAE&CDc)vM*)2Gm z%cKB!*FWLkJoIZmFk?OA7L75LxKs@PwRoSA&{a?k*108!33U8n@<~qH+DE0Q8*xhO zbDY{02Ang4i0F@Kj1MTgW!CH<5|hSgrC>uERrIb}ze z%l8HmRXho@7pF_9%Q%&MnM)VcBEaGG{eDu?&H4RScP+8N!6RnE)2C0h=XBAxH4WYz z#JJRe-B*EOWji`_G6odqDB7-JI5@s6uBhN$_M90InW+PCd3w5k=-OX`1k_dAoqhEL&`OjYL&4d8o^ zQza%Zg$~#*3Be8L{AbTjR*v$P{H6YJm$ez8RD7MbHT->0hrZJ-mYaX_mdPiQj5tJ( zbqs307D=duWlZO3sZ;&g5y^|xw--Ih(naIwE;tX{fqzXT7B+cBPii(FYcTe zuW_T_-~4m;$tb2?4)o{ukS>lnP%1v&k{}0HPyEWDD3$4#Kizj|@$a;tbtgv^y^zr^ z>b%b6rIc3Y2WKb5h+#Zb*W6HhN#)@FuUE!qURu^6;p0E6d|y#AKJTA?@>A9uRHd`A z&0P5E2@_yE09p1d`NvzB*udVeWG71n&M6XHHuTN8KP}_S6zscC*STSuZnQ9sWqsY9 zDHFsx4qiUJHg9=vBkcO!e=At`CL%7l=v15c%`2)WTrHBnq`0Jj?lZNEq5X1?vZ4(u z@UZY;n@pWSf0s905{fR6hxeW^JXRR z3}>51KLoh9)+gr-`~*5R^Q7MI<Q#Wqp*4%hbRCNb2 zKR%%%u=&}#6d_I7Yv$ac4jJ8~d!I5Y@a2UbO-Gmz!8T9^d8PG_eS0-f#o4ivGSi3J z*7g$Y&lT#jdl?LZUfTOqSv=tqj&@pZG*O~|@O_Tpi>JPl2`up6B^U1lR-7rOsWfK# zSq^C_2QQXZI88P@$UcrL4P8!|K4}KI(;tN^ejf6I#L}{Wvm<=-q{xHA)-SzVadrMS zjj3T@=L!eB_U6!2)BBOu;+n9dO}YR(qz)J*nghO*>RC|Bz)zff_cHWs zmpyPEi{^Je*`Lmm0tZxEdkh~#(_9Zscn*`=kwPoDY`+xz`}I#u??=&m50$WAjFPe}?O%^S<*39tmwD5Va(Vg+l7VE+4jgRMNSEz+A3YTH+#gtkq<;_87lcJ1wi zfq0wqG6PTD6m(di7otJ~7?{ybiC&y@s3#7}fA2VK-QoxpCpc5E_+&lTVlQ|1jst#~ z5K*!0-_xpCTb{nH3Cm;oWkMPF;hG|GlG90()a+fhRf)(F*zc)nKl0e_&D^KE-C5%4 znSJLEmrV3s>g!^fHtEaB`JxGfKS-twvN(ny?MlR`vOZ^l=}(5vr8j=h8XL3lKhZRd z21K-sqc1Ffm>w%CyN`WYMcPf7u<|HwV;1Lx{f1e| zyDt;ETQ<`l-+bN4<^!fpZ54B#%lG>=Gxa=GSeE_NVJ!jN{Yq%#DR?%?d@zYW2kO|Z zC@u zk_I8)cgGUG!x>zguJzbA+wxf5+ba9M&a3Znm!DtUK1P&VwcM?eWtc=p(oq5&^1Pl| zv+hXkcHTzBWJO1o`uLtSD=MGeU5njiQ>aP0JmuM-80Yg2Q>NTmX%rqRG^}3#f#9_ld?*atePay`G?-I(+M9U(H-+hzQ=dHFH*n|%3cUPm!!*qv6OATW zhr4S6p-g)==;1@-nKjWQr-7jf44;UJ+Wl`Bj?I4Id#bA{QcyUZAm0wkmKkqq4`)V6 zjDcbwH)Ad$*ndN33@C6Je=sz@kGL5fY?+|Aeh8v5WFWcN8$Yl)#91N758%+$H`#K8 zn7D4dt4V|7jD?pKcKRxbntdJR9?! zTyeczIP-Pq!OQnsWo8vkvwi~9IxZyW;5uWn>cE6UOgOY|8W~m7q!`3%iBai(KvB{XIZRUC?L;I_s`_6-*9_m zvmR-pom*3IpzxDbQMZ_isu11T#1A0|Z7&h~?4op|(V z_W7U9qMRP?7B%<6Dk#ABD+>d!AV<_mgN5E{|DY7Sxc92@Y9~1w?%=;2=wuz+5{ni; zD`yP#Q!g!we%&XpDkRXz+$kZXIqS>|Shn36Du8R`&zDVXz>8u@t&$7>Mz{#{{qyAy zwUm0I|M)sQ?s1o%wX~CEZb|B4S4~wrRB%1hNGzv>8R5`dAo|bXn;Flxw4F3AM68tc zrYQ)inXfHNq86NcpBack#||^=9%)j)gAttF>>WusGHry^T9y#^hX54kfd<}Aht~`` z5_2Jrl;M!o8SZoDbIAqstFQjX$)%kJ{&6o`FYWbPsEr-K3m3~Kj0JA-%?=RyN1(Ww z71F9DtTp0eVs)4&wi?qZdi;>!~s)5IU@1(ta3zCc7pR|kCQ!4`O+S3RsxXO{cM#`B0{8C zR^ci1@MScK$dKNNF@UPJZC$SmaBjA2j%gA{f;wr3?$qn#KskKylA@?n71R-382&lN z(gpm?ju!Pmqi>xTNiOZTXr*|}1ZI#z?!7-l%4Ey0qC#bafMRE6R4{1Fq_;&4@jivj zVqdBMOQ9qM1Pyj&c-EewEdt1Q1vbctlSI6C!e6;q{7zVwzzLE>_Pg7vi0zS@KFxak zpr(R3dofv5+hwG%G$$Cx6mzE5!&9YnUPD4^aq5WuOrc931aJri-@QfWrhwypgR7w@ zl8A()yfe>!4ya z9|LNrtX@ub^91w>j4SdlB%th&9ruqO<7t1E509pPSm;vCqfN?2k z*v&7ECCfq#IS>W$e*3OuyaLno3N~i7Yf=_qfIOpr za5g!$0o0m(RCUOiFi&d5t}9sa!{Bw=0>SsOZY9M_i;mwPwJIyTfjcx3LpxJNo^qH} zy;z!zoohIk`buk3p^N<~IVm(@WBrcZ2jLpSR?EW>w{LnUk{r*rlMK3rMLjz>=gW-A zjLZtm;H1zLJH+5yhv4)N0}`nH@JAe}pBX@n^V!+Xq8{EM>bWxl&vIEmb?f#4JMJVZ z5_`wFIq@zhc-m$q7+NIZ1iOIcXgDVU_MzkV^G1jY0!8dN|1dnuMDOeSO3vm3)rF@f z9B7kJnXk9|l>kVvpQp0sE)=g^sqLos8!}`_HeC|7mw?``(3ow%=>O%(t`Txjf{RrR zE>QF{%A+>;|n zno(3$p$3D74cP-Oklduz@SW{$%!M2i@gK8NJF%+z-}9$1C>Fkmdp!Z2s(}6qszb~% zGF50$QdNotGC+an1x}bkB|7(&(Pa>1!A?$KODl0cnF*%WA?syZF@>9&oMcU4ll{ih%wF`v z0=Q2b5r2LTUZ^o|!igtx=w|(^bUt&A-i~Pc^sAn=yF{gQjFb+!yi{bC2)_Ho zO+_(}uXGY~6W_=GyhVEyPj3E<5x~|n_E*vL&}x>y|a8@u>rl9l75XdqO0&WGRIlC z2(Y#pv7B47B$8R8Hp~AZ0pH!zc3*zU4}?g+pKiAj{uwe-whECtKtvvQjRb_FM>;Sm zj!gHNC?-;ukrgZ0DZI@63^-+N1<4bc$I6u}LuuBLE+S48FSqJX#x3WlsI*vVW;)X9 zeLyipW3uSa*+!Y^jBtJ_WXrx+Lm57O?~b5%kL_n%T$N*N!wOR9m7I9x2YTLV)$#l9 zG+oP|_@zT>t9G#2F#j-_Lj>8Cv`lTJQWC^C>55&RdFAHKbpKtG(%JGz0tbQ|kr6Xz zT`27&;C0|tp-D1 zz}ov2H#Q>(H3 zbq}a~s$=-=7gW&3(jvf!m61_5qscJyc%yn{uNO}HV~?G*a*iwRjfi z5Y$($3Yzf;a$AdMf4^2$l_(=@F0LVMXq`z8lR>&tdcBIn6T2fdM##Cc2SOvl<%rli zC^|v6$qURiJ&Ih4oO_3B09HpJcCa*23S>SLjg>z`xjRER?`Vf2n=ltR%{;i#$$fy} z*RLs+3{TFYnXkh+Q2c(1xF>Cx$pEP8qMh1W&?#cQd6|DGS)Tr)n}^FeB61bN zVyO~jpU#P<=ZhwDn&$6&;e&|yIlV? zY&b0rwiuS#kvr~l5*_tymWwD_lE`SfU~p&_eG1-p;VVhsie)~W>)YSD`Yp-wCN<1O z{M-Ad^4JiD9t{!6gx{d1wthB`TI{lO*R@N>|;k* zlc-MyVC0Hx!GDWhl|km0p6&dxrH5eI0%wWNEehoi_Ee&bw4>_S+p)6m4)EI-0QwTg zJ;X8gzlNdmbc$i=et=%tU7i;1s>abc$vOOl_b}a+QTcmi{<$rqb?qp;i0Oikv7EiR z&pj_=DblWSWVIGXcxzL8;9SMAewZDmc0R{jN=GUB9k+DqD3?<(MYg(+0ML~(R4X#X<2@Gy9 zlwTxt6a0q{wc+@gj-|z2BbKdbQ$~}6+rqit0cmbtq-b%Jgr6I4Pn)%CC>B5@|0kr- zBdp8maDq#@1O5HijnK-XU!(`52Hl32yc%#NMSOzw^(&*8(q|^)aY^$Z(0(IAT#nll)CG5VWI~0Wpyx=0Uj`hr*Ii0E<<+1mJPB9b zYMJ3O!5z1-xo@!NeU&`A!yLbkN6XDD%DuGYsP|X#zL|RC(DPf9cq(!C3BZxb$ zk{T7f56K{%%`1nRogRvqEJ0}fGpD=P?xs?!`Z^+T1y5L^YviSg`doC4jE0*h{gxrK z&RBAYil{A_1Ft(7AmOKzl;o5HV0ZFKa9DBMOSob8mmSsIZ~-qcV%^KT55PT4mskDp zD|_gB348MwS?!U;LLQ>`gmCFjXn6OzHn5rSSXrmc2aS;Th*EAE_h`~ z>673YopC{g$efH;q4w-s>0lBupl5X59>{N7mkp_e+#pg1v@@2r4&gwp|GneCh-A;~ z5+|mWQV_rC;{G}vw91G*(!-(d{3b%Q+_ojKF4u@o)`KP%%$F4|IaS0=9${WiO|=85 z%EgB7eYVV7DN0cpcb!bamQ-iS**TEY()np7&=pYO@xw=43_-lgbEtm;Kan`=D9z?Z zWYm{9e1HI-$Q3pSif@CMjJAqL`3KSeZ@|J;b_EiN?SZe!AiYG?JtoHa2<>b`>4BUP zDv6_j%sac}S*6W=K3#RV7tZs97Ifv6HRcGs~IZk|sn42d~=K`)2eH zywl?$hid;-MP=TtwH)UHTK4%l9LhIEN+*Pur*~GB_7r>;V&cuactm)zU4;#ZIo3zH z2d4I{Ym`&52x@6u{&+v68Zm=G(GM3;8ur7vy3WdcMAX8m^^#$gHwVyIcVZ-s4?S5Y zFI6A?t)(ZoVaR7RdGdaZ=vj~!w!Om|4O!V#@pYTxMl^1TS9ud9*?_3qZX6ucM>6B- z0B52+V!L(~g)W^jQdWqfSCDiOLy2HWVi`?lVXsOPim}`2G=6X2ht*dIQgVhEKu_k~ z7;sCpEn~rd&z$>FdAmL9w|w2hqv`UOdcLRbhlxXetHUtsex#z>(9Nv;S>m*Yg{J(G z`&V^zY5niT*UQ4|el)`L|FnO+t(V`|{Uw$EueXiVU1)h@+d%p8|Np3X2%mM>)IX}! anr2;{95=UBX*))KW;S|~Y1pWFoBto&EMm<7 diff --git a/thesis/resources/figures/results/transformer_results_by_input_length.png b/thesis/resources/figures/results/transformer_results_by_input_length.png index 2be874ef67fb8e7f34c6c15e698a9f73607dc41a..015534b36caa62a37181f52be52916eeff1afd92 100644 GIT binary patch delta 52 zcmZqb=43KPKG diff --git a/thesis/sections/background.tex b/thesis/sections/background.tex index d034373..b8ceef1 100644 --- a/thesis/sections/background.tex +++ b/thesis/sections/background.tex @@ -7,25 +7,18 @@ \subsection{Physiological Background}\label{subsec:physiological_background} \subsubsection{Menstrual Cycle}\label{subsec:menstrual_cycle} -The menstrual cycle describes the physiological changes in the female body that prepare it for pregnancy. -It is divided into two phases: the \textbf{follicular phase} and the \textbf{luteal phase}. -\\ -During the follicular phase, the ovarian follicles mature, and the endometrium (the inner lining of the uterus) thickens -in preparation for a potential implantation of a fertilized egg. -Around day 14 of a typical cycle, ovulation occurs, marking the transition to the luteal phase. -Ovulation refers to the rupture of the mature ovarian follicle and the release of an egg cell into the fallopian tube. +The menstrual cycle consists of physiological changes preparing the female body for potential pregnancy, +typically spanning around 28 days but varying considerably among individuals. +It includes two main phases: the follicular phase, beginning with menstruation, and the luteal phase, following ovulation. -Ovulation is triggered by a surge in \textbf{luteinizing hormone (LH)} -and \textbf{follicle-stimulating hormone (FSH)}, following a peak in estradiol levels. -As ovulation occurs, estradiol levels drop, progesterone levels begin to rise and a slight increase in body temperature -(typically around 0.5°C) can be observed. -This marks the beginning of the luteal phase. +During the follicular phase, ovarian follicles mature under the influence of rising estradiol levels, thickening the uterine lining (endometrium). +Around mid-cycle, a surge of luteinizing hormone (LH) and follicle-stimulating hormone (FSH), triggered by peak estradiol, +induces ovulation—the release of a mature egg into the fallopian tube. -During the luteal phase, the endometrium thickens further, creating an optimal environment for embryo implantation. -LH and FSH levels decrease, while progesterone remains elevated to support endometrial maintenance. -If fertilization does not occur, progesterone levels drop, leading to the shedding of the endometrial lining along -with the unfertilized egg. -This process, known as menstruation, marks the beginning of a new cycle. +After ovulation, the luteal phase begins. +Progesterone increases substantially, maintaining endometrial thickness for potential embryo implantation. +In parallel, a subtle rise in body temperature (~0.5°C) occurs due to progesterone elevation. +If fertilization does not happen, progesterone and temperature decline back to baseline levels, resulting in menstruation and initiating a new cycle. \begin{figure}[htbp] \centering @@ -36,27 +29,23 @@ This process, known as menstruation, marks the beginning of a new cycle. \label{fig:background_menstrual_cycle_physiology} \end{figure} -The menstrual cycle typically lasts around 28 days, with ovulation occurring near the midpoint. -However, variations, particularly in the follicular phase length, are common and can be influenced by factors such as stress, diet, exercise and age~\cite{silberstein_physiology_2000}. -Figure~\ref{fig:background_menstrual_cycle_physiology} provides a detailed overview of the hormonal and physiological changes throughout the menstrual cycle. - \begin{figure}[htbp] \centering \includegraphics[width=0.9\textwidth]{resources/figures/background/background_labeled_cycle} \caption{Body core temperature curve across a menstrual cycle. The red line represents a locally smoothed temperature trend. - Menstruation, ovulation, and the fertile phase are indicated in red, blue, and green, respectively.} + Menstruation, ovulation, and the fertile window are indicated in red, blue, and green, respectively.} \label{fig:background_labeled_cycle} \end{figure} +Figure~\ref{fig:background_menstrual_cycle_physiology} illustrates these physiological changes, highlighting hormonal fluctuations and temperature shifts around ovulation. +The menstrual cycle length varies significantly, influenced by factors such as stress, age, diet, and exercise~\cite{silberstein_physiology_2000}. + Figure~\ref{fig:background_labeled_cycle} shows the temperature curve over the course of a menstrual cycle with the -menstruation, fertile phase and ovulation marked. +menstruation, fertile window and ovulation marked. It starts with a menstruation and ends just before the next menstruation. The follicular phase starts at the beginning and goes on until the ovulation. The luteal phase begins at the ovulation and continues until the next menstruation. -Not every cycle results in ovulation—a phenomenon known as anovulation—which leads to a monophasic temperature pattern. -Anovulation can have various causes, including hormonal imbalances, stress, or underlying health conditions~\cite{rosenfield_adolescent_2013}. - \begin{figure}[htbp] \centering \includegraphics[width=0.9\textwidth]{resources/figures/background/background_anovulatory_cycle} @@ -65,25 +54,28 @@ Anovulation can have various causes, including hormonal imbalances, stress, or u \end{figure} -Anovulation is reflected in temperature data as either an absence of a clear temperature rise or a rise -that is insufficient in magnitude or duration to be considered a reliable indicator of ovulation. -Distinguishing between ovulatory and anovulatory cycles is challenging, as the only definitive confirmation of -successful ovulation in a clinical sense is a positive pregnancy test. +While many cycles exhibit a characteristic biphasic pattern, deviations from this norm are common. +Some remain monophasic, which might be an indication for an anovulatory cycle, which is a menstrual cycle, where no ovulation occurs. +Anovulation can have various causes, including hormonal imbalances, stress, or underlying health conditions~\cite{rosenfield_adolescent_2013}. +Monophasic cycles with a confirmed ovulation event have been observed, so there seems to be no clear indication that it is a direct cause of anovulation~\cite{moghissi_accuracy_1976}. +Thus, distinguishing between ovulatory and anovulatory cycles is challenging, as the only definitive confirmation of +successful ovulation in a clinical sense is a pregnancy. Even ultrasound imaging can only confirm that an egg was released from its follicle—not whether it was successfully implanted or fertilized. -Figure~\ref{fig:background_anovulation} shows a cycle that does not have an ovulation, and thus no resulting temperature rise. - +Figure~\ref{fig:background_anovulation} shows a cycle with a monophasic temperature pattern. To illustrate the diversity of real-world menstrual cycles, Figures~\ref{fig:background_long_cycle} and~\ref{fig:background_short_cycle} show examples of cycles that are significantly longer or shorter than a normative 28-day cycle. +\citeauthor{bull_real-world_2019} have done an extensive study on cycle variability, +highlighting that women frequently deviate from the normative cycle, especially with age\cite{bull_real-world_2019}. -\begin{figure}[htb] +\begin{figure}[htbp] \centering \includegraphics[width=0.9\textwidth]{resources/figures/background/background_long_cycle} \caption{Example of a long cycle with a length of 111 days} \label{fig:background_long_cycle} \end{figure} -\begin{figure}[htb] +\begin{figure}[htbp] \centering \includegraphics[width=0.9\textwidth]{resources/figures/background/background_short_cycle} \caption{Example of a short cycle with a length of 22 days} @@ -93,36 +85,28 @@ show examples of cycles that are significantly longer or shorter than a normativ These irregularities appear not only on a per-cycle basis, but also across time within the same individual. Figures~\ref{fig:background_irregular_cycles} and~\ref{fig:background_regular_cycles} show examples of a woman with an irregular and a regular menstrual cycle pattern, respectively. -Raw body core temperature readings are shown in light blue, with a red line indicating smoothing by local regression. -Vertical dotted black lines mark the beginning of each cycle. The irregular example highlights how multiple parameters can vary between individuals: cycle length, timing of ovulation, temperature shift magnitude between phases, and intra-phase temperature variability. -This multidimensional variability underscores the need for adaptive, data-driven models capable of learning personalized patterns--- -rather than relying on population-wide assumptions. -For a regular cycle pattern, sophisticated analysis or predictions are often not necessary, as the last ovulation day -can reliably be used as the next. -\begin{figure}[htb] +\begin{figure}[htbp] \centering \includegraphics[width=0.9\textwidth]{resources/figures/background/background_irregular_cycle_example} - \caption{Example of a woman with irregular menstrual rhythm} + \caption{Example of a woman with irregular menstrual rhythm. The dashed vertical lines indicate the ends of each cycle.} \label{fig:background_irregular_cycles} \end{figure} \begin{figure}[htbp] \centering \includegraphics[width=0.9\textwidth]{resources/figures/background/background_regular_cycle_example} - \caption{Example of a woman with regular menstrual rhythm} + \caption{Example of a woman with regular menstrual rhythm. The dashed vertical lines indicate the ends of each cycle.} \label{fig:background_regular_cycles} \end{figure} \subsubsection{Fertility Prediction}\label{subsubsec:fertility_prediction} -Throughout the menstrual cycle, the chance of fertilization varies significantly. -An egg cell released from the ovary during ovulation, can be fertilized for up to 24 hours. -However, since male sperm cells can survive up to 6 days inside the female reproductive tract, -the fertile window is typically defined as the five days before ovulation until one day after ovulation~\cite{dunson_day-specific_1999}. -Research by~\citeauthor{dunson_day-specific_1999} has shown that the highest chance of fertilization is around one day before ovulation, -as illustrated in Figure~\ref{fig:background_pregnancy_chance}. +Fertility varies throughout the menstrual cycle, centered around the ovulation event. +An egg remains viable for about 24 hours post-ovulation, while sperm can survive up to 6 days in a woman's reproductive tract, +thus the fertile period extends to approximately five days prior to ovulation~\cite{dunson_day-specific_1999}. +Consequently, the whole fertile window generally spans six days: five days preceding ovulation and one day after. \begin{figure}[htbp] \centering @@ -132,11 +116,30 @@ as illustrated in Figure~\ref{fig:background_pregnancy_chance}. \label{fig:background_pregnancy_chance} \end{figure} -It is important to note that fertility prediction is inherently dependent on ovulation prediction. -Since the probability of conception is tightly linked to ovulation timing, the accuracy of fertility prediction methods -is constrained by the precision of ovulation detection. -This relationship underscores the necessity of developing reliable ovulation prediction models, -as even small inaccuracies can significantly impact fertility assessments. +Figure~\ref{fig:background_pregnancy_chance} demonstrates the probability of fertilization peaking one day before ovulation, emphasizing the critical timing for fertility prediction. + +Fertility prediction fundamentally depends on accurate ovulation timing. +However, since the goal is to identify the fertile window before ovulation occurs, detection must be early and precise. +For individuals trying to conceive or avoid pregnancy, knowing the window of fertility is more actionable than identifying the ovulation event itself. + +\subsubsection{Practical Use Cases}\label{subsubsec:practical_use_cases} +In this study, we will focus on \emph{natural family planning} (NFP), which includes preventing and achieving pregnancy. +Individuals aiming to avoid pregnancy identify fertile days to abstain from intercourse, +whereas those seeking pregnancy aim to focus intercourse around days with the highest fertility probability. + +Both use cases revolve around accurately predicting ovulation. +However, the implications of prediction errors differ significantly. +A false-positive prediction indicates high fertility despite actual fertility being low or nonexistent, +whereas a false-negative prediction implies low fertility when fertility is actually high. + +For women aiming to avoid pregnancy, minimizing false-negative predictions is crucial due to the risk of unintended pregnancy. +Although false-positives may lead to unnecessary abstinence, this outcome is generally considered less severe. +Consequently, prediction algorithms should be conservative, erring on the side of higher fertility estimates to prioritize safety. + +Conversely, for women aiming to conceive, false-positive predictions could misdirect efforts toward incorrect cycle days, +causing frustration or delays. +False-negatives have fewer negative consequences. +Therefore, algorithms for this group should prefer cautious fertility estimates, reducing the risk of misdirected effort. \subsubsection{Physiological Signs of Ovulation}\label{subsubsec:physiological_signs} Several physiological signs correlate with ovulation and can be used for prediction. @@ -154,26 +157,7 @@ detecting the slight temperature rise that follows ovulation. Advances in wearable technology have further enabled continuous and automated temperature monitoring, improving accessibility and usability~\cite{alexander_fertilitatsmonitoring_2014, luo_detection_2020, yu_tracking_2022}. -\subsubsection{Use Cases of Fertility Prediction}\label{subsubsec:use_cases_of_ovulation_prediction} -The prediction of ovulation and corresponding fertility within a menstrual cycle serves two distinct use cases. -Specifically, we will focus on \emph{natural family planning} (NFP), which includes preventing and achieving pregnancy. -Individuals aiming to avoid pregnancy identify fertile days to abstain from intercourse, -whereas those seeking pregnancy aim to focus intercourse around days with the highest fertility probability. - -Both use cases revolve around accurately predicting ovulation. -However, the implications of prediction errors differ significantly. -A false-positive prediction indicates high fertility despite actual fertility being low or nonexistent, -whereas a false-negative prediction implies low fertility when fertility is actually high. - -For women aiming to avoid pregnancy, minimizing false-negative predictions is crucial due to the risk of unintended pregnancy. -Although false-positives may lead to unnecessary abstinence, this outcome is generally considered less severe. -Consequently, prediction algorithms should be conservative, erring on the side of higher fertility estimates to prioritize safety. - -Conversely, for women aiming to conceive, false-positive predictions could misdirect efforts toward incorrect cycle days, -causing frustration or delays. -False-negatives have fewer negative consequences. -Therefore, algorithms for this group should prefer cautious fertility estimates, -reducing the risk of misdirected effort. +Among the available methods, temperature-based monitoring, especially when automated and continuous, offers a promising avenue for large-scale cycle analysis. \subsection{Data Source and Characteristics}\label{subsec:data_background} This study is based on a dataset collected from users of the \emph{OvulaRing}~\cite{noauthor_ovularing_nodate}, @@ -202,14 +186,15 @@ Therefore, all user-entered cycle starts undergo manual review to reduce annotat In addition to temperature measurements, the database includes contextual metadata such as age, height, weight, and optional user-entered markers. These markers provide further physiological context and may include information about intermediate bleeding, sexual intercourse, or positive pregnancy tests. -At the time of writing, the dataset contains approximately 65{,}000 annotated cycles, comprising more than 350 million individual temperature measurements. +At the time of writing, the dataset contains approximately 65{,}000 annotated cycles, +comprising more than 350 million individual temperature measurements~\footnote{This is the largest dataset of continuous body core temperature data used in any study so far.}. \subsubsection{Dataset Summary} For the present study, the dataset was reduced to approximately 40{,}000 cycles after filtering out entries that were incomplete, contained hardware-related anomalies, or fell outside a reasonable cycle length range. -Very short cycles typically result from incorrect cycle start entries or premature termination of temperature recordings. -Extremely long cycles are often due to data entry errors or pregnancy-related recordings, +Cycles shorter than 10 days typically result from incorrect cycle start entries or premature termination of temperature recordings. +Long cycles, longer than 150 days, are often due to data entry errors or pregnancy-related recordings, where the sensor was worn continuously throughout gestation—sometimes producing sequences up to nine months long. While such cases may still contain useful information, they were excluded from this analysis to avoid complications in preprocessing and labeling. @@ -222,16 +207,18 @@ The median number of cycles per user is 4 (IQR: 2--8) and the median cycle lengt The average data density—defined as the fraction of available measurements out of the theoretical maximum of 288 measurements per day—is 0.90. This corresponds to an average data availability of 90\% per cycle, with an average loss of 10\%. -It should be noted, that both ovulation day and anovulation estimates are based on retrospective algorithmic inference, not human labels. -Further details are provided in section~\ref{sec:methodology}. +37934 cycles (94\%) were classified as biphasic and 2333 (6\%) as monophasic. + +Users had a median age of 32 years (IQR: 29–36), median weight of 65 kg (IQR: 58–77), and median height of 168 cm (IQR: 163–172). \subsubsection{Irregularities and Confounding Factors} -Core body temperature is influenced by various factors unrelated to the menstrual cycle. +Despite careful data collection, real-world measurements are subject to physiological and behavioral noise. +Especially core body temperature is influenced by various factors unrelated to the menstrual cycle. Illnesses—especially those involving fever—can significantly affect temperature patterns. This poses a challenge for any analysis relying on temperature data, as one of the key physiological indicators of ovulation is a post-ovulatory temperature rise (see Section~\ref{subsubsec:physiological_signs}). Figure~\ref{fig:background_fever_cycle} shows an example of a cycle where an illness caused a marked increase in temperature. -This event is particularly problematic because the fever-induced rise occurs just before the expected ovulatory shift, potentially confounding ovulation detection. +This event is particularly problematic because the fever-induced rise occurs just before the expected ovulatory shift, potentially confounding fertility detection. Distinguishing illness-related changes from cycle-related ones requires models that are sensitive to context and robust to outliers. \begin{figure}[htbp] @@ -254,39 +241,35 @@ Figure~\ref{fig:background_menstruation_data_gap} shows an example cycle with a \label{fig:background_menstruation_data_gap} \end{figure} -%TODO: more stats! - \subsubsection{Privacy} The dataset used in this study contains sensitive personal health information and is handled with strict privacy safeguards. All data is pseudonymized and processed exclusively on encrypted devices, ensuring that no identifiable information can be traced back to individual users. VivoSensMedical does not share user data with third parties; the data is used solely for internal research and product improvement efforts that directly benefit users at no additional cost. \subsection{Technical Background}\label{subsec:technological_background} +With a large, high-resolution dataset of longitudinal temperature measurements and associated metadata available, +the next challenge lies in how to model such sequential data effectively. +Accurate ovulation prediction requires algorithms that can handle temporal dependencies, +irregularities, and physiological variability across users. +To this end, we turn to machine learning techniques designed for time series analysis, +beginning with foundational concepts and progressing to modern neural architectures. \subsubsection{Time Series Analysis}\label{subsubsec:time_series_analysis} -Time series analysis is a fundamental tool for studying sequential data that evolves over time. -Unlike other data types, time series data has an inherent temporal order, where each data point is associated -with a timestamp, capturing its dependence on past values. -Time series analysis typically serves two main goals: -Understanding the underlying mechanisms that lead to the observed data and predicting future data points based on the -historical information and potentially external factors~\cite{cryer_time_2008} -\\ -Time series analysis encompasses various methods, ranging from simple statistical models to complex deep learning architectures. -Classical methods - - -In the following, we will introduce the two most common approaches used for machine learning on time series data, -LSTMs and transformers. +Temperature data recorded by the OvulaRing forms a high-resolution time series, where each measurement carries temporal context. +Time series analysis is essential to uncover meaningful patterns and predict future physiological states from such sequential data. +In machine learning, this often involves models that can learn temporal dependencies—most notably +Recurrent Neural Networks (RNNs) and the more recent Transformer architecture. \subsubsection{RNN and LSTM Networks}\label{subsubsec:lstm_networks} Recurrent Neural Networks (RNNs) are extensions of classical neural networks that incorporate cyclic connections between neurons. These recurrent connections allow the network to retain information from previous inputs by feeding the hidden state from a prior time step into the current one, enabling a form of temporal memory. +They process input \emph{sequentially}, maintaining this hidden state over time. In practice, this means that input data is processed sequentially, one step at a time. At each step \(t\), the input \(x_t\) is combined with the previous hidden state \(h_{t-1}\) to produce a new hidden state \(h_t\), which contributes to the output \(o_t\). -This process allows the network to learn temporal dependencies and model sequential data effectively~\cite{medsker_recurrent_1999}. +This enables the network to learn temporal dependencies in sequential data~\cite{medsker_recurrent_1999}. However, this approach has the downside that the model cannot explicitly control how it remembers or forgets information at each step, limiting its ability to manage long-term dependencies. @@ -366,7 +349,7 @@ as well as the biases for each gate in a cell, are learned through backpropagati To build more expressive models, memory cells can be stacked in multiple layers, and their outputs concatenated or passed sequentially to higher layers. LSTMs are widely used in biomedical applications due to their capacity to handle sequences of variable length and complexity. -In the context of ovulation prediction, where hormonal patterns exhibit periodicity but also irregularity, +In the context of fertility prediction, where hormonal patterns exhibit periodicity but also irregularity, LSTMs are well-suited to learn relevant time-dependent signals from sequential physiological measurements. While powerful, LSTMs can be computationally intensive and sensitive to hyperparameter tuning. @@ -374,16 +357,19 @@ Therefore, they are often compared with alternative architectures, including simpler feedforward networks and more recent attention-based models, to evaluate trade-offs in performance, interpretability, and computational cost. -The next section introduce the \emph{Transformer} architecture, a more recent alternative that forgoes +Given their ability to learn from sequences with noisy periodic structure, +LSTMs offer a natural choice for modeling hormonal and temperature fluctuations across menstrual cycles. + +The next section introduces the \emph{Transformer} architecture, a more recent alternative that forgoes recurrence in favor of attention mechanisms. \subsubsection{Transformer Models}\label{subsubsec:transformer_models} Transformer models are a class of neural architectures that use \emph{self-attention} to model dependencies in sequential data without relying on recurrence~\cite{vaswani_attention_2017}. -Unlike recurrent neural networks (RNNs), Transformers process input sequences in parallel, +Unlike recurrent neural networks (RNNs), Transformers process input sequences \emph{in parallel}, allowing them to model relationships between any pair of input tokens or timesteps directly. -This mitigates the limitations of recurrent models, such as long-term memory constraints +This design mitigates the limitations of recurrent models, such as long-term memory constraints and vanishing gradients. Originally introduced for machine translation, Transformers have proven broadly applicable to @@ -423,13 +409,13 @@ to the model regardless of their location, even if they play different syntactic Positional encodings, often based on sinusoidal functions, inject a unique position-dependent signal into each token, enabling the model to distinguish between identical tokens in different positions. -In this work, we use since and cosine functions of different frequencies: +In this work, we use sine and cosine functions of different frequencies: \begin{align} PE_{\text{pos}, 2i} &= \sin\left(\frac{\text{pos}}{10000^{\frac{2i}{d_{\text{model}}}}}\right), \\ PE_{\text{pos}, 2i+1} &= \cos\left(\frac{\text{pos}}{10000^{\frac{2i}{d_{\text{model}}}}}\right) \end{align} -where \(i\) is the dimension of the input and \(pos\) is the position in the sequence. -This was done according to the original paper~\cite{vaswani_attention_2017}. +where \(i\) is the dimension of the input and \(pos\) is the position in the sequence, +as in the original paper~\cite{vaswani_attention_2017}. \begin{figure}[htbp] \centering @@ -553,6 +539,9 @@ including time-series forecasting and biomedical modeling~\cite{wu_deep_nodate,z Its ability to model complex, long-range dependencies without recurrence makes it well-suited to domains like biomedical time-series, where signals are often irregular and span diverse temporal resolutions. +This makes Transformers well-suited for learning long-range temporal dependencies in physiological data, +such as ovulatory trends spanning multiple days or cycles. + \subsubsection{Convolutional Layers as Temporal Feature Extractors} For high-resolution time-series data, the input dimensionality can become large, especially in models like Transformers that process the entire sequence in parallel. @@ -578,3 +567,7 @@ The stride determines how far the filter moves at each step, affecting both the \label{fig:background_convolution_example} \end{figure} +In summary, time-series modeling offers a range of approaches, each with specific trade-offs. +RNNs and LSTMs provide explicit sequential modeling but suffer from training inefficiencies. +Transformers excel at long-range context capture but demand more memory and parallelization. +Convolutional layers offer efficient local feature extraction and often serve as useful pre-processing stages for both model families. \ No newline at end of file diff --git a/thesis/sections/discussion.tex b/thesis/sections/discussion.tex index d053ac6..4c1ad6a 100644 --- a/thesis/sections/discussion.tex +++ b/thesis/sections/discussion.tex @@ -1,8 +1,17 @@ %! Author = alex %! Date = 3/6/25 + \section{Discussion}\label{sec:discussion} +In this study, we investigated the performance of different machine learning architectures on the task of fertility prediction, +with the aim to find a model that performs well for natural family planning and natural contraception on regular and irregular cycles. + +Our goal was to + +Based on an extensive real-world database and established model architectures for timeseries analysis, +we expect to outperform both rule-based baselines and related studies. +We think, that for regular cycles, the performance difference will be lower than % talk about whether bbt / temperature can be used for such a task, discuss bbt doubt papers diff --git a/thesis/sections/introduction.tex b/thesis/sections/introduction.tex index ce7ce9d..3bbb4c5 100644 --- a/thesis/sections/introduction.tex +++ b/thesis/sections/introduction.tex @@ -3,48 +3,64 @@ \section{Introduction}\label{sec:introduction} -A 2024 report by McKinsey and the World Economic Forum\cite{mckinsey_health_institute_closing_2024} highlights -persistent disparities in women's healthcare and health-related research, particularly in reproductive health. -The Institute for Health Metrics and Evaluation (IHME) has identified reproductive and gynecological health issues as -the most significant factors affecting both life span and health span globally\cite{global_burden_of_disease_collaborative_network_global_2020}. -Despite their widespread impact, many aspects of reproductive health remain under-researched. -Improving our ability to understand the menstrual cycle could have significant implications for fertility tracking, contraception, -and overall reproductive health. -\\ -\\ + +In textbooks, a menstrual cycle is 28 to 30 days in length with its ovulation happening around day 14~\cite{Phy} + + + The average age of pregnant women in developed countries has been increasing over the past few decades. -This combined with the overall chance of conception sharply decreasing with age, especially after 35, -makes it more and more important to understand and predict ovulation accurately\cite{sauer_reproduction_2015}. +Combined with the sharp decline in conception rates after age 35, +this trend underscores the growing need for accurate understanding of the menstrual cycle~\cite{sauer_reproduction_2015}. Predicting the fertile days in a women's menstrual cycle is not only relevant for family planning but also for natural contraception and general health monitoring, as the corresponding hormone levels have a significant impact on the overall health and well-being of a woman. -\\ -Ovulation is the process in which an egg is released from the ovarian follicle, making fertilization possible. + +Textbook cycles usually have a length of 28 days with an ovulation around day 14. +This however, does not represent the real world variability of menstrual cycles. + +Ovulation is the process in which an egg cell is released from the ovaries, making fertilization possible. This process is regulated by hormonal changes, including fluctuations in luteinizing hormone (LH) and -follicle-stimulating hormone (FSH), and is accompanied by an increase in basal body temperature (BBT)\cite{holesh_physiology_2025} -This process is complex and yet not fully understood. -Factors such as stress, diet, and exercise can influence the menstrual cycle and make it hard to predict ovulation. -\\ -\\ -Several physiological signs can be used to predict ovulation. -The most accurate method is ultrasonography, which detects changes in follicle size and rupture. -Other methods include detecting LH and FSH in urine, measuring BBT, and observing cervical mucus, -each with its own advantages and limitations. -These, as well as the interplay of those factors, will be discussed in more detail in Section~\ref{sec:background}. +follicle-stimulating hormone (FSH), and is accompanied by other physiological changes such as an increase in electrical resistance +and viscosity of the cervical mucus or an increase in body temperature~\cite{wallach_prediction_1980}. +These processes remain incompletely understood and are influenced by lifestyle factors such as stress, diet or exercise or +health-related factors such as Polycystic Ovary Syndrome (PCOS), making ovulation difficult to predict. -Many studies have used these physiological signs to predict ovulation and fertility -\cite{noauthor_cervicovaginal_2005, sato_novel_2024, royston_identifying_1991, luo_detection_2020, -alexander_fertilitatsmonitoring_2014, luz_improved_2024, yu_tracking_2022, pratikno_pdf_2024}. -However, most of these methods rely on manual data collection, requiring either daily measurements or invasive procedures. -This not only makes them impractical but also results in small sample sizes, limiting their generalizability. +In theory, these physiological changes provide a basis for predicting ovulation and the surrounding fertile window. +In practice, however, many of these signals are difficult to measure continuously, as they require invasive procedures, +manual tracking, or costly equipment. +Body temperature, by contrast, is a non-invasive marker that can be measured continuously using intravaginal, wrist-worn, or ear-worn thermometers. +While some studies have deemed body temperature unusable for predictive analysis~\cite{bauman_basal_1981}, +others suggest its predictive value for ovulation~\cite{sato_novel_2024, royston_identifying_1991, + luo_detection_2020, alexander_fertilitatsmonitoring_2014, luz_improved_2024, yu_tracking_2022, pratikno_pdf_2024}. -Generalization is crucial for developing a reliable ovulation predictor, given the high variability of the menstrual cycle -\cite{munster_length_1992, bull_real-world_2019}. -This is particularly important for applications where prediction accuracy -is critical, such as natural contraception or high-cost procedures like in-vitro fertilization (IVF), -where false predictions can have severe consequences. +However, most existing studies rely on small, idealized datasets that exclude cycles with irregular lengths or late ovulation. +While such restrictions simplify the prediction task and yield high accuracy, they give a misleading impression of real-world model performance. +It is thus not yet fully clear, whether temperature can reliably be used as a predictive marker for ovulation or fertility. -This work aims to develop an ovulation predictor that is both accurate and generalizable while maintaining interpretability, -allowing for insights into key variables and patterns influencing the prediction. +Additionally, other studies have focused on peripheral temperature measurements of skin or in-ear temperature, +which are subject to many sources of noise that can significantly affect the quality of the resulting predictions. +Intravaginal temperature reflects true core body temperature and offers higher resolution and stability, +as it is largely unaffected by external circumstances. +This allows for more reliable detection of subtle thermal shifts associated with ovulation, especially in irregular cycles. -% research questions \ No newline at end of file +Generalization is critical for reliable ovulation prediction, +especially given the high variability in cycle patterns~\cite{munster_length_1992, bull_real-world_2019}. +This is particularly relevant in high-stakes applications such as natural contraception or in-vitro fertilization (IVF), +where inaccurate predictions can have serious consequences. + + +This study aims to advance ovulation prediction by leveraging an extensive database of more than 40,000 menstrual cycles recorded using +an intravaginal wearable device that continuously measures core body temperature. +The objective is to develop a machine learning model that performs reliably across diverse cycle types, including irregular ones. +To this end, we compare a set of time series-based machine learning architectures and evaluate their performance for +natural family planning and contraception. +Finally, we demonstrate that high predictive accuracy on highly regular, curated datasets, as commonly reported in prior work, +does not reflect general applicability, since such datasets tend to favor even simple, rule-based approaches. + + +The research objectives are: +\begin{itemize} + \item To assess the performance of different model architectures on the use cases of natural family planning and contraception, across both regular and irregular cycles. + \item To compare sophisticated machine learning models with simple rule-based baseline approaches. + \item To study the overall predictive quality of temperature for ovulation and fertility prediction. +\end{itemize} diff --git a/thesis/sections/methodology.tex b/thesis/sections/methodology.tex index a3ded31..10134d4 100644 --- a/thesis/sections/methodology.tex +++ b/thesis/sections/methodology.tex @@ -4,22 +4,18 @@ \section{Methodology}\label{sec:methodology} -While prior studies have demonstrated the promise of physiological signals for ovulation detection and phase classification, -many are limited by small sample sizes, rigid inclusion criteria, or non-transparent methodologies. -Temperature has emerged as a potentially predictive signal, but existing work often lacks scalability or generalizability. -This study extends previous approaches by leveraging a large, heterogeneous real-world dataset of high-resolution core body temperature -readings to develop and evaluate machine learning models for real-time ovulation prediction. -In addition to model development, special emphasis is placed on evaluating performance across irregular cycles and assessing -the predictive value of low-noise, high-resolution temperature data. - -The following section outlines the methodology used, including data preprocessing, -feature extraction, input encoding, and model architectures. - -% why did I select tft over other methods -> include examples of time series and why I belief a complex model could help -% you did I apply it -% implementation details +Despite promising results in earlier studies, ovulation prediction remains constrained by small datasets, +assumptions of cycle regularity, and opaque modeling approaches. +To address these limitations, we develop a data-driven framework based on a large, +heterogeneous dataset of real-world menstrual cycles. +Our approach emphasizes model transparency, adaptability to irregular patterns, and the predictive utility of +high-resolution core body temperature measurements. +This section outlines the methodology used, including preprocessing, labeling, feature extraction, and model architectures. \subsection{Data Preprocessing}\label{subsec:data_preprocessing} +This section outlines the preprocessing steps applied to the raw temperature data, +including cycle filtering and retrospective ovulation labeling. +These steps ensure that only clean, complete, and labeled cycles are used for model training. \subsubsection{Data Filtering}\label{subsubsec:data_filtering} @@ -53,15 +49,15 @@ The algorithm operates in two stages: \end{enumerate} This retrospective labeling provides a practical and scalable proxy for ground truth, enabling training and evaluation across a large, real-world dataset, -especially, as temperature is at least an excellent retrospective marker for ovulation. +particularly given that temperature is a well-established retrospective marker of ovulation. In internal evaluations, the estimated ovulation day fell within a \(\pm\)2-day window of the expert reference in approximately 86\% of labeled cycles. These labels serve as the supervisory signal for model training and evaluation. -We acknowledge the limitations of this method: ambiguous or noisy temperature patterns—due to illness, dropout, -or sensor error—can lead to mislabeled examples, which may propagate to downstream models. +We acknowledge the limitations of this method: ambiguous or noisy temperature patterns, due to illness, dropout, +or sensor error, may result in noisy labels, which can affect downstream model performance. However, label quality is continuously reviewed and may be refined iteratively as model performance improves. -The specific usage of ovulation labels in feature construction is described in the next section. +The next section details how these labels are incorporated into feature representations and model training. \subsection{Feature Engineering}\label{subsec:feature_engineering} @@ -70,8 +66,8 @@ The features used as model inputs have been divided into three categories: \item \textbf{Static features} - Characteristics, that remain constant across a user's cycle, such as age, height, or average ovulation day \item \textbf{Known features} — Inputs known a priori at each time step, such as time of day or calendar-based variables. \item \textbf{Observable features} — Inputs available at the current time step, including raw and derived temperature values. - \item \textbf{Target features} — Outputs the model is trained to predict, such as the fertility probability. \end{itemize} +The target variables predicted by the model—like ovulation status or fertility probability—are described separately. Each feature type can handle categorical and continuous features. This allows for mixed inputs, such as scalar measurements and class labels, within the same category. @@ -84,48 +80,44 @@ may process each feature group differently depending on their architectural desi Static features are the features that do not change over the course of a cycle. They might even be static for all cycles from a specific user, such as age, height and weight. +Static features provide user-specific context that helps the model learn individualized cycle patterns beyond what temperature alone can reveal. -The static features are supposed to create contextual information about the cycle and the user that each model can then -use to learn patterns based not only on the temperature data, but also on this context. +%\begin{table}[htbp] +% \centering +% \begin{tabular}{l>{\raggedright\arraybackslash}p{0.65\linewidth}} +% \toprule +% \textbf{Feature} & \textbf{Description} \\ +% \midrule +% User age & Age in years; mean imputed if missing \\ +% User height & Height in centimeters; mean imputed if missing \\ +% User weight & Weight in kilograms; mean imputed if missing \\ +% Average cycle length & Mean length of all previous cycles for this user \\ +% Cycle length SD & Standard deviation of previous cycle lengths \\ +% Average ovulation day & Mean day of ovulation from previous cycles \\ +% Ovulation SD & Standard deviation of ovulation day of previous cycles \\ +% Ovulatory fraction & Proportion of prior cycles classified as ovulatory \\ +% Cycle count & Number of previous completed cycles available \\ +% Avg. pre-ovulation temperature & Mean temperature in the follicular phase of previous cycles \\ +% Avg. post-ovulation temperature & Mean temperature in the luteal phase of previous cycles \\ +% \bottomrule +% \end{tabular} +% \caption{Static features used as model inputs} +% \label{tab:static_features} +%\end{table} +%Table~\ref{tab:static_features} shows all static features and their descriptions. -\begin{table}[htbp] - \centering - \begin{tabular}{l>{\raggedright\arraybackslash}p{0.65\linewidth}} - \toprule - \textbf{Feature} & \textbf{Description} \\ - \midrule - User age & Age in years; mean imputed if missing \\ - User height & Height in centimeters; mean imputed if missing \\ - User weight & Weight in kilograms; mean imputed if missing \\ - Average cycle length & Mean length of all previous cycles for this user \\ - Cycle length SD & Standard deviation of previous cycle lengths \\ - Average ovulation day & Mean day of ovulation from previous cycles \\ - Ovulation SD & Standard deviation of ovulation day of previous cycles \\ - Ovulatory fraction & Proportion of prior cycles classified as ovulatory \\ - Cycle count & Number of previous completed cycles available \\ - Avg. pre-ovulation temperature & Mean temperature in the follicular phase of previous cycles \\ - Avg. post-ovulation temperature & Mean temperature in the luteal phase of previous cycles \\ - \bottomrule - \end{tabular} - \caption{Static features used as model inputs} - \label{tab:static_features} -\end{table} - -Table~\ref{tab:static_features} shows all static features and their descriptions. Prior research by \citeauthor{li_menstrual_2023} has shown that menstrual cycle characteristics vary significantly with age and BMI~\cite{li_menstrual_2023}. Including such information is therefore expected to improve predictive performance. In addition, summary statistics from previous cycles—such as ovulation timing, temperature levels, or the fraction of ovulatory cycles—provide useful individual context. These features help the model learn subject-specific variability and better estimate the likelihood and timing of ovulation in the current cycle. +Table~\ref{tab:feature_overview} shows the full list of static input features. All historical features are computed using only data available prior to the current cycle, ensuring no data leakage and supporting robust, user-adaptive learning. -% -The idea here is to provide as much context information to the models as possible to help them predict the ovulation. - \subsubsection{Known Features}\label{subsubsec:known_features} -In the context of this study, known features correspond to time-dependent inputs. -These help the model place each observation in temporal context: +Known features encode temporal context that is available at each time step and independent of physiological measurements. +These help the model interpret observations in relation to time-based structure, including circadian and behavioral rhythms. \begin{itemize} \item \textbf{Time since cycle start} — Provides the model with a relative position within the menstrual cycle. @@ -172,11 +164,9 @@ They represent real-time physiological signals from which the model must infer o \item \textbf{Rolling Window Temperature Maximum} — The maximum temperature within a 1-day window, capturing transient peaks or elevated plateaus. \end{itemize} -These derived features are intended to reduce model complexity by providing smoothed or extremal summaries of the raw signal. -The \textit{rolling average} allows the model to capture broader trends without having to learn temporal aggregation from scratch. -The \textit{rolling minimum} and \textit{maximum} support the detection of boundary behavior (e.g., temperature shifts, sustained elevation, extreme values) -without requiring explicit memory or aggregation. +These derived features summarize local trends or extrema in the temperature signal, reducing the burden on the model to learn such patterns from raw data. Special care was taken, so that the sliding window can only look backwards, so that no data leakage can happen. +We extend each windowed feature at the beginning with the starting value, so that the window can be calculated for the first real value already. The 1-day window length reflects the expected circadian cycle and strikes a balance between temporal sensitivity and signal stability. Figure~\ref{fig:methodology_observable_features} illustrates the behavior of all observable features within a single cycle. @@ -260,7 +250,8 @@ a \textit{robust scaler} was used for distributions with outliers, and a \textit \subsubsection{Time-Series Input Representation} \label{subsubsec:time_series_input_representation} -Due to the high temporal resolution of the temperature data (288 measurements per day), raw input sequences can become prohibitively long for most model types. +The high temporal resolution of the temperature data, 288 measurements per day, results in very long input sequences +that are impractical for most deep learning models to process directly. To manage input size and evaluate the impact of temporal resolution on predictive performance, a parameterized resampling strategy is applied. Consecutive time steps are aggregated into bins of configurable size, and each bin is reduced to a single value using a feature-specific aggregation function. @@ -274,17 +265,17 @@ The effect of different sampling resolutions and aggregation strategies is evalu To simulate real-time prediction rather than retrospective analysis, a sliding-window approach is employed. This allows the model to make predictions based only on data available up to a specific point in the cycle. -Each cycle is split into overlapping input windows, where each window includes data from the cycle start up to a defined time step. -The window length is fixed and configurable. -As the cycle progresses, the window slides forward, allowing the model to incorporate increasing historical context over time. +To simulate real-time prediction, each cycle is split into overlapping, +fixed-length input windows that capture all available data up to a given time step. +As the cycle progresses, these windows slide forward, allowing the model to update its prediction based on growing historical context. For the model types used in this study, each window produces a single output vector. -By default, this corresponds to the predicted target values at the final time step of the window, though this can be offset depending on configuration. +By default, this corresponds to the predicted target values at the final time step of the window, +though this can be offset to predict targets several steps into the future, depending on configuration. While the architecture could be extended to produce output sequences (e.g., one prediction per input step), this study focuses on single-vector outputs. -This setup enables temporally resolved predictions at different stages of the cycle and supports analysis of how predictive accuracy evolves with increasing context. -Depending on the configuration, downsampling and windowing can be skipped to allow the raw data to be processed by the models themselves. -This is used primarily in the convolutional flavours of the models, to allow them to learn the best way of reducing the input complexity based on the data itself. +This setup mimics a real-time setting, enabling the model to generate predictions dynamically as new data arrives during the cycle. +For convolutional architectures, downsampling and windowing can be disabled entirely, allowing the model to learn temporal compression directly from the raw input. \begin{figure}[htbp] \centering @@ -296,18 +287,16 @@ This is used primarily in the convolutional flavours of the models, to allow the \label{fig:methodology_padding_example} \end{figure} -Fixed-length input windows would normally prevent early-cycle predictions when insufficient data is available. -To address this, left-padding is applied using masked values. - -In this study, predictions are enabled once at least four days of data are available. -A padding value of 0.0 is used for all features, and the padding length is adjusted accordingly. -This design ensures that the model learns to ignore tokens consisting entirely of padding. - -The feature \textit{hours since start}, which encodes the time elapsed since cycle onset, is also set to 0.0 for all padded tokens— -explicitly indicating that these entries contain no usable information. +Because fixed-length windows require a minimum amount of input data, early-cycle predictions would normally be impossible. +To address this, left-padding is applied with masked tokens until sufficient real data is available—enabled here from day four onward. +Padding values are set to 0.0 across all features. +Since the feature \emph{hours since start} is also set to 0.0 for padded steps it is reinforced, that the section is not relevant +for the prediction as no information is present. Figure~\ref{fig:methodology_padding_example} shows an example of such padding during early-cycle input preparation. +This input strategy supports efficient, temporally-aware learning and allows us to evaluate how predictive accuracy evolves over time within each cycle. + \subsection{Model Architecture and Selection}\label{subsec:model_architecture_and_selection} The primary objective of this study is to find models that accurately predict the features introduced in~\ref{fig:methodology_target_features}, @@ -565,11 +554,9 @@ Due to the heterogeneity of both the models and the trainings, a dynamic batch s the batch size dynamically during training to optimize the resource usage. The learning rate was scaled linearly with the batch size to allow for equivalent convergence behaviour~\cite{goyal_accurate_2018} The batch size was capped at 2048 to avoid OOM errors during data preparation. -%TODO sources, for square rule \subsection{Evaluation}\label{subsec:evaluation} -% TODO: review To meaningfully compare model performance, we define a set of metrics that capture both overall accuracy and behavior at key points in the prediction sequence. This includes metrics for different temporal segments, enabling a more detailed understanding of model strengths and limitations. @@ -698,7 +685,7 @@ Additionally, ovulation must occur no later than cycle day 150, as later values biologically atypical cases that fall outside the scope of this study. \subsubsection{Use Case Evaluation} -We further evaluate the two distinct use cases introduced in Section~\ref{subsubsec:use_cases_of_ovulation_prediction}. +We further evaluate the two distinct use cases introduced in Section~\ref{subsubsec:practical_use_cases}. For this purpose, two specialized evaluation algorithms were developed, enabling comparability between models and providing interpretable performance metrics for each scenario. @@ -721,9 +708,11 @@ A day-specific probability of intercourse is computed for each user based on age We assume, that the users don't have any health-related or non-health-related issues affecting fertility. If a user's age is unknown, it is randomly drawn from the overall dataset distribution. Only users with at least one continuous year of data are included. +To get a representative result, we use 500 randomly selected user years. -Each day of data for a full year is categorized by the algorithm into one of the following outcomes: +Each day of data for a full year we count the following states by the algorithm: \begin{itemize} + \item \emph{Sex}: Intercourse occurred. \item \emph{No Sex}: no intercourse occurred. \item \emph{Correct Denial}: fertility prediction correctly indicated abstinence during a fertile period. \item \emph{Incorrect Denial}: fertility prediction incorrectly indicated abstinence during an infertile period. @@ -753,16 +742,16 @@ Figure~\ref{fig:methodology_use_case_pregnancy_decision_diagram} illustrates the The fertility threshold is adjustable and is explored further in Section~\ref{subsec:use_case_evaluation_results}. Since sexual intercourse frequency differs slightly for couples trying to conceive~\cite{gaskins_predictors_2018}, we assume an average frequency of six times per month. -We derive a daily probability of intercourse based on remaining fertile days predicted for the month, -ensuring that intercourse frequency averages out to this monthly rate. We assume no health-related fertility impairments for comparative simplicity, though we acknowledge that real-world fertility is influenced by numerous complex factors. Similar to the contraception scenario, only users with at least one continuous year of data are considered. +For representative results, we use 500 randomly selected user years. -Each day in a full year is classified into one of the following categories: +For each day in a full year we count occurrences of the following states: \begin{itemize} - \item \emph{No Sex}: Day predicted as fertile, but no intercourse occurred. + \item \emph{Sex}: Intercourse occurred + \item \emph{No Sex}: No intercourse \item \emph{Pregnancy}: Correct fertile prediction, intercourse occurred, resulting in pregnancy. \item \emph{No Pregnancy}: Correct fertile prediction, intercourse occurred, but no pregnancy occurred. \item \emph{Incorrect Deferral}: Incorrect non-fertile prediction, actual fertility was above threshold. diff --git a/thesis/sections/related_work.tex b/thesis/sections/related_work.tex index 7e834f3..9303f30 100644 --- a/thesis/sections/related_work.tex +++ b/thesis/sections/related_work.tex @@ -16,6 +16,8 @@ confirming that LH surges reliably indicate an imminent ovulation event. Despite their diagnostic value, many of these biomarkers are difficult to measure continuously and reliably in everyday settings, limiting their practicality for real-time or large-scale applications. +Among the physiological indicators explored, body temperature has gained particular attention due to its accessibility +and suitability for passive, continuous monitoring. \subsection{Temperature-Based Approaches}\label{subsec:temperature_based_approaches} Body temperature has emerged as a more accessible physiological signal for ovulation tracking, @@ -34,7 +36,7 @@ In contrast, this study, along with several recent works, leverages continuous o This richer signal provides a more robust foundation for detecting ovulatory patterns and addresses many of the limitations historically associated with BBT-based methods. This was further supported by a study from \citeauthor{zhu_accuracy_2021}, who compared the accuracy and sensitivity of traditional BBT measurements with continuous skin temperature recordings from a wrist-worn device~\cite{zhu_accuracy_2021}. -They found that continuous temperature measurements had significantly higher sensitivity in detecting ovulation, though at the cost of increased false positives and lower specificity. +They found that continuous temperature measurements achieved higher sensitivity but at the cost of more false positives and reduced specificity. Importantly, the continuous data showed a greater average temperature difference between the follicular and luteal phases. The authors conclude that for women seeking to optimize their chances of conception, continuous temperature tracking offers measurable benefits—primarily due to improved phase delineation enabled by the richer signal. @@ -43,33 +45,34 @@ who used an in-ear wearable device that measured ear canal temperature every fiv They trained a Hidden Markov Model (HMM) to classify each data point into either a high- or low-temperature state, augmented with biorhythm information from the user. -After filtering, the final dataset consisted of 65 cycles, each with at least 40\% data availability and at least one self-reported ovulation day, as determined by a hormone test kit. +After filtering, the final dataset consisted of 65 cycles, each with at least 40\% data availability and at least one self-reported ovulation day, as determined by a hormone test kit, +a notable contrast to the 40,000 cycles analyzed in this study. However, no information was provided regarding the distribution of cycle lengths or ovulation timing. Ovulation detection was considered successful if the predicted day fell within ±3 days of the self-reported value. The method achieved a sensitivity of 92.31\%, with 54.69\% of ovulation days detected exactly on the reported date. -The model, however, relies on strong assumptions of phase regularity and fixed transition durations—such as a standard -luteal phase length of 14 days—which do not reflect real-world variability. +The model, however, relies on strong assumptions of phase regularity and fixed transition durations, such as a standard +luteal phase length of 14 days, which do not reflect real-world variability. Moreover, HMM predictions are conditioned on either a previous cycle or population-level averages, limiting performance in irregular or anovulatory cycles. As a result, the approach performs well on regular, well-behaved data but lacks robustness in more diverse, real-world scenarios. - -In~\citeyear{yu_tracking_2022}, \citeauthor{yu_tracking_2022} employed an in-ear thermometer along with a fitness tracker +Building on this idea, \citeauthor{yu_tracking_2022} combined temperature with additional physiological signals to improve predictive performance. +In~\citeyear{yu_tracking_2022}, they employed an in-ear thermometer along with a fitness tracker for heart rate monitoring to predict the fertile window using machine learning~\cite{yu_tracking_2022}. Their study population consisted of 153 women, divided into a regular cycle group ($n = 103$) and an irregular group ($n = 50$). After filtering, 89 and 25 participants remained in the regular and irregular groups, respectively. The prediction task was to determine whether a given day falls within the fertile window, based on data from the preceding days. -A second model was trained to predict whether menstruation occurs on a given day, again using preceding data as input. +They developed a probability function based on a changepoint analysis of the smoothed waveforms of the BBT and heart rate. +A second function was developed in a similar way to predict whether menstruation occurs on a given day, again using preceding data as input. For the fertile window prediction, the model achieved a sensitivity of 69.30\% in the regular group and 21.00\% in the irregular group. For menstruation prediction, the model detected 70.70\% of menstruation days in the regular group and 36.30\% in the irregular group. -These results indicate that the model performs reasonably well for individuals with regular cycles, -but struggles significantly in the presence of menstrual irregularity—particularly in detecting the fertile window. +These results indicate that the model performed well in regular cycles but struggled with irregularity, particularly in detecting the fertile window. -In addition to academic research, several commercial products use temperature-based methods for fertility tracking, +Complementing academic efforts, several commercial products have adopted temperature-based tracking, such as \textit{Ava}~\cite{sl_ava_nodate}, \textit{Daysy}~\cite{electronics_zykluscomputer_nodate} or \textit{Trackle}~\cite{noauthor_trackle_nodate}. However, these products typically rely on proprietary algorithms, and no peer-reviewed publications are available detailing their methodology or performance. This lack of transparency limits their scientific evaluation and comparability. @@ -109,17 +112,15 @@ Anovulatory cycles were excluded, along with users meeting the following criteri They further subdivided participants into groups with low and high sleep variability, called HVST and LVST respectively. As a result, the study population and cycle types were highly regular and homogeneous, with 30 cycles (18 women) in the HVST and 26 cycles (16 women) in the LVST category. -No information about the distribution of both lengths or ovulation dates was given. +No information about the distribution of both cycle lengths or ovulation dates was given. Based on selected features—such as minimum sleeping heart rate and single-point basal body temperature (BBT) after waking—they report classification accuracies between 0.843 and 0.864, depending on the feature subset, with very similar numbers for precision, recall, specificity and F1 score. Ovulation day prediction yielded an average absolute error between 3.6 and 4.1 days. - - \paragraph{Summary:} While various physiological signals and modeling strategies have been explored for ovulation prediction, -many existing studies are limited by small, highly selected datasets, assumptions of cycle regularity, or reliance on proprietary algorithms. +many existing studies are limited by small, highly selective datasets, assumptions of cycle regularity, or reliance on proprietary algorithms. The present work extends prior approaches by leveraging a large, heterogeneous dataset of real-world cycles and applying transparent, data-driven modeling to better capture individual variability. \ No newline at end of file diff --git a/thesis/sections/results.tex b/thesis/sections/results.tex index 165166e..de68cd6 100644 --- a/thesis/sections/results.tex +++ b/thesis/sections/results.tex @@ -25,6 +25,32 @@ allowing for a nuanced comparison of approaches and their practical relevance to \subsection{Fertility Probability Prediction Accuracy}\label{subsec:fertility_probability_precition_accuracy} +\begin{landscape} + \begin{table}[ht] + \centering + \caption{Model comparison for fertility prediction using MAE and MSE} + \begin{adjustbox}{max width=\linewidth} + \begin{tabular}{lllcccccc} + \toprule + \textbf{Model} & \textbf{Window} & \textbf{Daily} & + \textbf{MAE$_{fert}$} & \textbf{MAE$_{fert,during}$} & \textbf{MAE$_{fert,non}$} & + \textbf{MSE$_{fert}$} & \textbf{MSE$_{fert,during}$} & \textbf{MSE$_{fert,non}$} \\ + \midrule + ModelA & 7 & Yes & 0.67 & 0.59 & 0.73 & 0.89 & 0.82 & 0.95 \\ + ModelB & 14 & No & 0.65 & 0.58 & 0.71 & 0.87 & 0.80 & 0.93 \\ +% More rows... + \bottomrule + \end{tabular} + \end{adjustbox} + \label{tab:fertility_comparison} + \end{table} +\end{landscape} + + + +% show why I selected the individual input configs for model config training +% selected by best mse fertility, use 2nd best, as it provides basically the same performance, but more input data for more complex model configs + \subsubsection{Performance Across Fertile Window}\label{subsubsec:fert_performance_across_fertile_window} \subsubsection{Impact of Input Resolution}\label{subsubsec:fert_impact_of_input_resolution} @@ -56,6 +82,4 @@ allowing for a nuanced comparison of approaches and their practical relevance to \subsubsection{Pregnancy Use-Case Results}\label{subsubsec:use_case_pregnancy_results} -% show, that past cycles might not directly be included but are indirectly included by the static features - \subsection{Summary of Key Findings}\label{subsec:summary_of_key_findings}