{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"\n### Initial Comments\n\nFirst of all thankyou to G-Research and Cambridge Spark for hosting this competition and to the Kaggle staff for making everything run relatively smoothly. I found it so interesting and more open ended than other comps I've competed in, in my short Kaggler career.\n\nIt goes without saying that while I'm in 1st Place now there is no guarantee it will stay that way in the next 6 weeks, but I'm confident I'll be there or thereabouts. Hopefully you find this solution useful regardless of the final outcome.\n\nElephant in the room - there has been a lot of concern about the validity of the final LB results and requests to release the final test data after the competition ends so that competitors can validate their final scores. I think this would be a good idea and should put any concerns to bed. If it's possible & not too much effort for Kaggle to accomodate the request please consider it. However, as far as im concerned my current score seems reasonable based on all the scores I achieved in testing. I have no reason to believe any of the LB scores are incorrect until there is evidence to the contrary.\n\n**I would encourage those who feel their score is unexpected share their solutions as the Kaggle community might be able to debug any issues/misunderstandings.**\n\nIt is obviously difficult to accurately predict crypto prices, but it's possible to find *some* signal - hopefully this notebook will shed some light on how I got mine.\n\n*Disclaimers*\n- *this was not my exact final submission but very similar*\n- *I have removed the dataset containing my pretrained models and scalers - so some commands will error*\n\n### Summary\n\n- 17 features with lagged and timestamp averages for all models.\n- Ensembles of LGBM and Keras NN models.\n- Target Engineering and prediction switching.\n- Some training on extra cryptos outside competition scope.\n\nTest Scores (no lookahead bias):\n- 0.0565 best score on original testing period (Jun 13 2021 - Sep 21 2021)\n- 0.0475 best score on supplemental_train.csv update. (Sep 21 2021 - Jan 24 2022)\n- 0.0465 best single model score (Jun 13 2021 - Sep 21 2021)\n\n### External Data\nI downloaded free public data from multiple exchange apis (Binance, FTX, Coinbase, Kucoin... etc.) to see if this extra data would improve my models. Some of my final models were trained on extra data. Specifically, using some currencies not included in the competition (e.g. XRP, ZEC...) from Binance seemed to provide a consistent small improvement over multiple timeframes.\n\n### Feature Engineering\nI settled on a group of 17 features for all models. There are 8 lagged features - **a simple mixture of EMA's, historical returns and historical volatility over various lookback periods** - and these features were averaged across timestamps to produce 8 more. Asset_ID was also included. It was tricky to find a group that performed the best consistently across different time periods, adding and removing features from this set seemed to reduce performance. Im sure this isnt the optimal feature set, but this group seemed to work well enough.\n\nIt was also important to perform some kind on binning on the features, especially for training the LGBM model. The commonly used reduce_mem_usage function and some rounding functions seemed to provide a suitable amount of bins. I found binning to 500-1000 unique values worked well for any given continuous feature.\n\n### Target Engineering\nI think a crucial part of this competition was manipulating the target. Thanks to the work of **@alexfir** and other kagglers [we found out how the target was being calculated](https://www.kaggle.com/alexfir/recreating-target).\nI thought it would be useful to split the target into two components:\n- The forward 15 minute return of an asset\n- The beta component, where we calculate the mean of forward 15 minute returns for all assets and incorporate past 15 minute returns for the previous 3750 timestamps\n\nAs **@gengdaiziwang** illustrates [in his helpful notebook](https://www.kaggle.com/gengdaiziwang/are-we-training-models-on-an-inaccurate-target), we can see that in the case where an observation is missing for a given asset, the beta component is automatically set to 0.\n\n![download.jpeg](attachment:3935f637-5324-464e-b014-1a21702d4adb.jpeg)\n\nBut this makes it a completely different target! A model trained on the target given as standard will have to deal with training on a target that randomly switches between a 15 minute forward return for one asset vs what is essentially a 15 minute forward return *relative to the other assets* - very different things.\n\nThe approach I took was to create recreate two targets to represent these two different cases. Then I could train separate models on these recreated targets which would learn the two cases much more effectively.\n[See an implementation of this target engineering](https://www.kaggle.com/tomforbes/target-engineering).\n\nNow, having two models trained on different targets, I can alter my predictions to match the target produced by the api. The api provides all the information i need to figure out whether the beta component will be == 0 because if an asset is missing from an iteration i know that for the next 3750 iterations the beta component == 0 for this asset. **Therefore my next 3750 predictions will come from a model that was training on a target constructued without a beta component.**\n\nThis target engineering and dual model method added roughly 0.01 to my score on the original test period, although this did vary quite significantly for other testing periods.\n\nTo summarise as simply as possible:\n\n- TargetZero = A target based on 15 minute Forward Return only\n- TargetBeta = TargetZero + Beta Component\n- ModelZero = Model(s) trained on TargetZero\n- ModelBeta = Model(s) trained on TargetBeta\n\n![gres.png](attachment:026ae3c5-ee03-4532-b11e-191b37105d21.png)\n\n### Models\n\nI found LightGBM worked well and was easy to experiment with, i used the weighted correlation evaluation metric and fairly out of the box hyperparameters, tuning didnt seem to add much and wasn't consistent across time periods.\n\nI also used a Keras NN heavily influenced by **@lucasmorin's**\n[excellent notebook](http://https://www.kaggle.com/lucasmorin/online-fe-nn-loss-shap). This just worked really well for me out of the box and i couldnt find many ways to improve the architecture or hyperparameters.\n\nEnsembling these two models worked well and added an extra 0.005 to my score (roughly) although the improvement varies in different market regimes. I also experimented with ensembles using catboost and XGboost but the predictions from most gradient boosting models were too highly correlated to provide much ensembling benefit. I found the diversity in predictions between NN and GB were a good match.\n\nI used a fairly simple CV method for training and testing. My folds were based on timestamp values similar to the below:\n- Training Fold 0: 1514764860 - 1570000000 \n- Training Fold 1: 1514764860 - 1580000000 \n- Training Fold 2: 1514764860 - 1590000000 \n- Training Fold 3: 1514764860 - 1600000000 \n- Training Fold 4: 1514764860 - 1610000000\n- Validation Fold 0: 1570000000 - 1580000000\n- Validation Fold 1: 1580000000 - 1590000000 \n- Validation Fold 2: 1590000000 - 1600000000 \n- Validation Fold 3: 1600000000 - 1610000000 \n- Validation Fold 4: 1610000000 - 1620000000 \n- Test Fold A: 1615000000 - 1623542400\n- Test Fold B (Supplemental Train Update): 1623542400 - 1643000000\n\nSince this public lb was useless in this competition, it was essential to construct a robust CV framework to have a reliable benchmark for making improvements to models. It was also important to have test folds in several different market regimes to avoid overfitting to one period.\n\n**My best single model was a NN trained on fold 0 data only (surprisingly).**\n\nI was fairly selective on picking models to include in the final submission as some folds just trained much better than others and had more consistent outperformance on unseen data.\n\n### Submission\n\nThis is almost the exact notebook I submitted for my final predictions. I used lists to store historical data for feature calculations and did most calculations using numpy or lists, avoiding pandas at all costs.\n\nThe submission completes in roughly 7 hours.\n\n\n### What didnt work\n- Layering extra data from different exchanges for the same asset. E.g training on BTC data from Binance + Coinbase + FTX + G-Research version. Didnt harm but didnt improve models significantly either.\n- Framing as a classification problem - poor results\n- Hyperparameter tuning LGBM\n- PCA features\n- Using all 14 assets features per row / predicting on 14 targets.\n- Meta modelling GB and NN predictions\n\n### With more time id try\n- Extra Feature engineering\n- Pytorch implementation of keras NN for ensembling\n- LSTM\n- Other model architectures\n\n","metadata":{},"attachments":{"3935f637-5324-464e-b014-1a21702d4adb.jpeg":{"image/jpeg":"/9j/4AAQSkZJRgABAQEAkACQAAD/4RDoRXhpZgAATU0AKgAAAAgABAE7AAIAAAAKAAAISodpAAQAAAABAAAIVJydAAEAAAAUAAAQzOocAAcAAAgMAAAAPgAAAAAc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAE5pbmEgV2FuZwAABZADAAIAAAAUAAAQopAEAAIAAAAUAAAQtpKRAAIAAAADMTIAAJKSAAIAAAADMTIAAOocAAcAAAgMAAAIlgAAAAAc6gAAAAgAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAADIwMjE6MTI6MjIgMTY6MzY6MDUAMjAyMToxMjoyMiAxNjozNjowNQAAAE4AaQBuAGEAIABXAGEAbgBnAAAA/+ELHGh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8APD94cGFja2V0IGJlZ2luPSfvu78nIGlkPSdXNU0wTXBDZWhpSHpyZVN6TlRjemtjOWQnPz4NCjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iPjxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+PHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9InV1aWQ6ZmFmNWJkZDUtYmEzZC0xMWRhLWFkMzEtZDMzZDc1MTgyZjFiIiB4bWxuczpkYz0iaHR0cDovL3B1cmwub3JnL2RjL2VsZW1lbnRzLzEuMS8iLz48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOnhtcD0iaHR0cDovL25zLmFkb2JlLmNvbS94YXAvMS4wLyI+PHhtcDpDcmVhdGVEYXRlPjIwMjEtMTItMjJUMTY6MzY6MDUuMTE4PC94bXA6Q3JlYXRlRGF0ZT48L3JkZjpEZXNjcmlwdGlvbj48cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0idXVpZDpmYWY1YmRkNS1iYTNkLTExZGEtYWQzMS1kMzNkNzUxODJmMWIiIHhtbG5zOmRjPSJodHRwOi8vcHVybC5vcmcvZGMvZWxlbWVudHMvMS4xLyI+PGRjOmNyZWF0b3I+PHJkZjpTZXEgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5LzAyLzIyLXJkZi1zeW50YXgtbnMjIj48cmRmOmxpPk5pbmEgV2FuZzwvcmRmOmxpPjwvcmRmOlNlcT4NCgkJCTwvZGM6Y3JlYXRvcj48L3JkZjpEZXNjcmlwdGlvbj48L3JkZjpSREY+PC94OnhtcG1ldGE+DQogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgIDw/eHBhY2tldCBlbmQ9J3cnPz7/2wBDAAcFBQYFBAcGBQYIBwcIChELCgkJChUPEAwRGBUaGRgVGBcbHichGx0lHRcYIi4iJSgpKywrGiAvMy8qMicqKyr/2wBDAQcICAoJChQLCxQqHBgcKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKioqKir/wAARCAHhAx0DASIAAhEBAxEB/8QAHwAAAQUBAQEBAQEAAAAAAAAAAAECAwQFBgcICQoL/8QAtRAAAgEDAwIEAwUFBAQAAAF9AQIDAAQRBRIhMUEGE1FhByJxFDKBkaEII0KxwRVS0fAkM2JyggkKFhcYGRolJicoKSo0NTY3ODk6Q0RFRkdISUpTVFVWV1hZWmNkZWZnaGlqc3R1dnd4eXqDhIWGh4iJipKTlJWWl5iZmqKjpKWmp6ipqrKztLW2t7i5usLDxMXGx8jJytLT1NXW19jZ2uHi4+Tl5ufo6erx8vP09fb3+Pn6/8QAHwEAAwEBAQEBAQEBAQAAAAAAAAECAwQFBgcICQoL/8QAtREAAgECBAQDBAcFBAQAAQJ3AAECAxEEBSExBhJBUQdhcRMiMoEIFEKRobHBCSMzUvAVYnLRChYkNOEl8RcYGRomJygpKjU2Nzg5OkNERUZHSElKU1RVVldYWVpjZGVmZ2hpanN0dXZ3eHl6goOEhYaHiImKkpOUlZaXmJmaoqOkpaanqKmqsrO0tba3uLm6wsPExcbHyMnK0tPU1dbX2Nna4uPk5ebn6Onq8vP09fb3+Pn6/9oADAMBAAIRAxEAPwD6RooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigArL16fWoLND4dtLW6uC+GW6lKKF9cgGtSsnxDc6SumyWmu3i2tvdLsLNN5RP0bPBoAzfCnii81fUNQ0zWLSG2v7B1WRbdy6HIzwTVfX/ABXq8OvHSfC+nW99cwx+ZcfaJGRU4yACM8kVh+AZNO0fxDq9tpdwJdJMi7LqWTfltvQOfvd+9Whqtn4b+JWr3WsTpb29/FG8EshwpCJggE9+aAOp8LeII/EuiC+RPLdZGhlQdFkQ4YfnWvJIkMTySsERFLMx6ADqa4v4V20sHhe6lmQoLjUbmaMHujSEqfxBrS8X6bq95pl1JputGwiW3ffF9nWTeNp4yelACr8Q/CLKCviLTyD0PnCj/hYPhL/oYbD/AL/Cs/4daNpsnw30J5tPtJJGs0LO0Ckk4+ldL/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8KP+Fg+Ev8AoYbD/v8ACtX+w9J/6Bdl/wCA6f4Uf2HpP/QLsv8AwHT/AAoAyv8AhYPhL/oYbD/v8KP+Fg+Ev+hhsP8Av8K1f7D0n/oF2X/gOn+FH9h6T/0C7L/wHT/CgDK/4WD4S/6GGw/7/Cj/AIWD4S/6GGw/7/CtX+w9J/6Bdl/4Dp/hR/Yek/8AQLsv/AdP8KAMr/hYPhL/AKGGw/7/AAo/4WD4S/6GGw/7/CtX+w9J/wCgXZf+A6f4Uf2HpP8A0C7L/wAB0/woAyv+Fg+Ev+hhsP8Av8K2dN1Oy1ixS80u6iurZyQssTZU44PNR/2HpP8A0C7L/wAB0/wrnPhaqp4JVUUKovLgAAYA/etQB2NFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFV7zT7LUECX9pBcqpyFmjDgfnViigCpFpWnwQeTBY20cWd2xIlC59cYou9K0/UAovrG3uQn3fOiV9v0yKt0UANjijhjWOFFRFGFVRgAfSqms/8gK//AOvaT/0E1dqlrP8AyAr/AP69pP8A0E0AY3w3/wCSa6D/ANeafyrpq5n4b/8AJNdB/wCvNP5V01ABRRRQAUUUUAFFFQyXcENxHDJKiyyHCITyfwo3C9iaiiigAooooAKKKKACiiigAooooAKQkKpLHAAyTS1X1D/kGXX/AFxf/wBBNNasT0QWl9bXyO9pMkqo20lCCM1Yrz34OEt4UuWPJN038hXoVaVqfs6jguhlQqOrTU31CiiisjYKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACuP+F/8AyJY/6/Ln/wBGtXYVx/wv/wCRLH/X5c/+jWoA7CiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKpaz/wAgK/8A+vaT/wBBNXapaz/yAr//AK9pP/QTQBjfDf8A5JroP/Xmn8q6auZ+G/8AyTXQf+vNP5V01ABRRRQAUUUUAFeaeKHY/GTQUz8uwHHvhq9LrzPxN/yWnQf9wf8As1deE+N+j/I48X8EfVfmemUUUVyHYFFFFABRRRQAUUUUAFFFFABVfUf+QXdf9cX/APQTViq+o/8AILuv+uL/APoJpx3QpbM4X4N/8ilcf9fTfyFehV578G/+RSuP+vpv5CvQq6cX/Hl6nNg/93h6BRRRXKdQUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFcf8L/+RLH/AF+XP/o1q7CuP+F//Ilj/r8uf/RrUAdhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRWVruvw6DAkk1peXRkbaqWkXmN+VAGrRWN4f8UWPiITi0SeCa3YLLBcpskQkZGRVPXfHFhoWsw6XLZahd3UyNIqWlv5nAxnPPvQB0tUtZ/5AV//ANe0n/oJp+m341KxS5WCe3Df8s7hNjj6imaz/wAgK/8A+vaT/wBBNAGN8N/+Sa6D/wBeafyrpq5n4b/8k10H/rzT+VdNQAUUUUAFFFFABXmfif8A5LToP+4P/Zq9MrzPxP8A8lo0H/cH/s1deE+N+j/I48Z8EfVfmemUUUVyHYFFFFABRRRQAUUUUAFFFFABVfUf+QXdf9cX/wDQTViq+o/8gu6/64v/AOgmnHdClszgPg1cIfDlzbg/vBOXI9sCvR68w+DDp/Z14gZfM35K98cV6fXVjFavI5ME74eIUUUVyHYFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUVm+ItSfR/Dl9qESB3t4i6qTgGnFOTSQpSUU2+hpUjMq43HGax/CmsSa/4atdRnjWN5gSVU5AwcVyXxGup4fFnhuKKVkR5ssqnGfmFbQouVT2b8/wMaldRpe0Wq0/E9GooByBRWBuFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAVx/wv/wCRLH/X5c/+jWrsK4/4X/8AIlj/AK/Ln/0a1AHYUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAVna3q9loWmSX+osFjiGQMZZj6AdzWjWB4q8G6X4xt4INYNz5cD+YggnMfPvjr1oAy/A2mXcl3qXiHUo/Il1SQPHCP4FAxz9cA07xD4T1K+8TQ67o+srp89tC8e1oFkDBsdc9OlXdO8D6dpdu8Nveam6MQf3t67kYHbPQUus+CNO1y4Wa6u9SiKqFxb3jxgjHcDrQAzwH4iuPEnh57m9VRPBcy2zlejmNtu78cU7xfp+sXemXT6XrIsYlt3Lxm3WTeNp4yela+k6RZ6JpyWWnR+XCpJx1JJ6knuT60az/AMgK/wD+vaT/ANBNAGL8Nsj4aaBk5P2JOfwrp65n4b/8k10H/rzT+VdNQAUUUUAFFFFABXmfif8A5LRoP/XMf+zV6ZXmfif/AJLRoP8AuD/2auvCfG/R/kceM+CPqvzPTKKKK5DsCiiigAooooAKKKKACiiigAqvqP8AyC7r/ri//oJqxVfUf+QXdf8AXF//AEE047oUtmeS/BQf8TjUTu/5dxx/wIV7FXjXwUP/ABPdS/69h/6FXstd2Yf7w/kcGXf7uvmFFFFcB6AUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABWB47/5EPV/+vZq36wPHf8AyIer/wDXs1a0f4kfVGVb+FL0ZU+Gv/Igad/ut/6Ea534mf8AI4eGf+uv/sy10Xw1/wCRA07/AHW/9CNc78Tf+Rv8M/8AXX/2Za7Kf+9y+f6nFV/3OP8A27+h6Wv3B9KWkT7i/Slrzj0gooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACuP+F//Ilj/r8uf/RrV2Fcf8L/APkSx/1+XP8A6NagDsKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAqlrP/ICv/8Ar2k/9BNXapaz/wAgK/8A+vaT/wBBNAGN8N/+Sa6D/wBeafyrpq5n4b/8k10H/rzT+VdNQAUUUUAFFFFABXmfif8A5LRoP+4P/Zq9MryD4lahLpfxI0u9twpkiiUru6dSP6124NOVRpdmcONko0030aPX6KZC5kgjc9WUE0+uI7gooooAKKKKACiiigAooooAKr6j/wAgu6/64v8A+gmrFV78btNuQOpicfoacdxS2PFPhFqsNh4qkt5QS97GIo8eoOf6V7nXzt4CiaH4jaZG/DLMQf8Avk19E16OZRSrJrqjzMsk3RafRhRRRXmnqBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFYHjv/kQ9X/69mrfrnvHhI8Catj/AJ92rWj/ABI+qMq38KXoyt8Nf+RA07/db/0I1znxNIPi/wANAHkS8/8AfS10Pwz/AORBsPo3/oRrj/iGxPxE0gZ6Sp/MV3UlfFy+ZwVnbBx/7dPWo/8AVr9KdTY/9Uv0FOrzD1AooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACuP+F//ACJY/wCvy5/9GtXYVx/wv/5Esf8AX5c/+jWoA7CiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKpaz/yAr/8A69pP/QTV2qWs/wDICv8A/r2k/wDQTQBjfDf/AJJroP8A15p/Kumrmfhv/wAk10H/AK80/lXTUAFIzBRliAPelrhfivdvbeF4jFI0bfaEIKnHrWlKn7Sah3MqtT2cHPsd1RVXS2L6PZuxyWgQk+vyirVQ1Z2NE7q4V4r8Xf8AkdrH/rin/oRr2qvE/jC2zxjaP/dt1P8A48a78v8A4/yZ5+Y/wPmj2azObKH/AHB/Kpqy/DWoDVfDVjehdgmiB2+nb+lalcMk1JpnfFpxTQUUUVJQUUUUAFFFFABRRRQAVDd/8eU//XNv5VNUF823T7hvSJj+hprcT2PBPCH/ACVOx/6+W/8AQTX0FXzx4HmE/wAStOlAwHnJx/wE19D16WZaVI+h5eWa05eoUUUV5h6oUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABXPePP+RE1b/r3auhrnvHv/Iiat/17tWtH+JH1RlW/hS9GVPhn/wAiFYfRv/QjXEeP5kf4madGrAtHLHuHpkiu2+GX/Ih2P0b/ANCNeeeN1I+L0ZI4MsOP0r0aCvip/M8zEO2Ep/I9ui/1Kf7op1Mi/wBSn+6KfXknsBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFcf8L/8AkSx/1+XP/o1q7CuP+F//ACJY/wCvy5/9GtQB2FFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFZWueILfQYEluLa7ufMbaqWkPmN+VatZ+t6vY6HpsmoaiwWOIZHGWJ9AO5oAraB4osPEaz/AGJZ4pbdgssNxHskQkZ5Wodf8Zad4duYra5iurm4lUsIbSHzGAHUkZrL8C6bdyXmpeItRi8iTVJFeOEfwKBt5+uAagstg+Lepfa9uTAvkb/7uz5sUAdZpGr2muacl7p7l4mJHIwVI6gjsRS6z/yAr/8A69pP/QTXI/CvP9jal5f+p/tO52fXzDmtfxfYa1daZdPpWsJYwrbv5kbWwk3jac8npQA34b/8k10H/rzT+VdNXMfDbI+GmgZOT9iTn8K6egArz34uxCXw/ACcEzKP516FXm/xiTGi28oYgrIBgH3rqwn8eJyYz+BI73SV2aLZL6W8Y/8AHRVvIzjPNU9HOdDsT620f/oIrhvBupzXXxA1qKW4Z1VtqoW4GCegqFTc+aXY0dVQ5I9z0WvEfjL/AMjZbf8AXsP5mvbq8Q+Mh/4qy3/69h/M105d/H+RyZl/u/zR6L4CvbaPwVo1tJPGs7wZWMty3zHtXVV4dodwzeJPBaKxAWMKQD1+dq9xrLF0uSd+93+Jtg6vtKdu1l+CKGu3Mln4fvriBtskUDuh9CBWH8PtaudV8GR3+rThpTK6lyAOAeK1vFDBfCupZ4zbOP0NeL6frM7fCnUbCMtGLW4jZXRsE725rShR9rSa80Z4iv7Ksn5PQ98VldQyEMpGQR3payvDDFvCumMxJJtkySevFatcUlaTR3RfNFMKKKKkoKKKKACoL4Z0+4B6GJv5Gp6gvf8AkH3H/XJv5U1uJ7Hz/wCBkWP4laciDCicgD/gJr6Hr558Egj4laaWBGZyRnv8pr6Gr08y/iR9Dy8r/hy9Qoooryz1QooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACue8e/wDIiat/17tXQ1z3j3/kRNW/692rWj/Ej6oyrfwpejKfwy/5EOx+jf8AoRrz/wAdN/xdiAekkP8ASuj0DxKPC3wostQMHn/PsCZxySf8K4zxpqUcnxDivycR7YJW284+UEivVw9OX1icraO55GIqR+rQjfVWPYPD3iiw1+a6trLzPMsiEl3rgZ6cflW7XlvwhmW41bXpk5WSQMPoSa9SrzcTTVOq4o9PC1HVpKbCiiiuc6QooooAKKKKACiiigAooooA4Hxf4gvtL8caLbW0u2Gdtjrjrkiu+ryz4g/8lD8P/wDXZf8A0Ja9TrrrxSp02u36nJQk3UqJ9H+gUUUVyHWFFFFABRRRQAUUUUAFFFFABXH/AAv/AORLH/X5c/8Ao1q7CuP+F/8AyJY/6/Ln/wBGtQB2FFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFYHirwbpnjG2gg1d7oRwP5ii3mMfPvjr1rfooA5zTvBNjplu8MF9qjqxB/e3jMRgdBnoKdrvgrS/EMkEl491DNAmxZbacxuR3BI5PSuhooApaTpNpomnJZaemyFOeTkknqSe5PrRrP8AyAr/AP69pP8A0E1dqlrP/ICv/wDr2k/9BNAGN8N/+Sa6D/15p/Kumrmfhv8A8k10H/rzT+VdNQAV5z8YufDsOP8AnqtejV558Xf+QHbf9dl/nXVhP48Tkxn8CR2+j/8AICsP+vaP/wBBFeVeBWx8U9THrNL/ADNeraV/yBrL/r3T/wBBFeSeCWx8XL4es039a2w+sKvoY4jSdH1PUvEmtDw9oFxqTQmYQgfIDjOTivE/iLqg1rUNN1FY/KFxZq+wnO3JPFerfEv/AJJ/qP0X/wBCFeJ+ISfsejZ/58Vx+Zrpy6EdJ9btfgcuZVJaw6WT/E1vC0vmeL/C6/3GA/U19AV84eBXLeO9GBJIW4AGe3Wvo+ssyVqkV5fqbZY705Pz/QxPGP8AyKWof9cW/ka8M0v/AJEHXf8Artb/APoRr2j4g3JtfBl2y/xDZ+YNeL6X/wAiDrv/AF2t/wD0I1rgV+5b81+aMMe/3yX91/kz1r4U3U114Gia4kZykzopJ6KMYFdpXDfCH/kQ1/6+ZP6V3NedidK0vU9PC60IegUUUVznSFFFFABUF9/yDrn/AK5N/I1PVLWZxb6LdyN0ETD9KqOrRMtIs8f8LrGfEHh+VQpf7SVLDr9w8V7bXgngP/kLaKf+n9v/AEWa97rvzBWqJf1ueflzvTb/AK2QUUUV5x6QUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABXO+Pf+RE1b/r3auirnfHv/Iiar/17tWtH+JH1RlW/hS9GeZX1xFN8E7WONwzxXShwOxO41xesQzQXwS4JL+WhyTngqMVq2k0Y+HGowlwJGv4mC55I2msbUL6TUbrz5sBtipwOwGK+loxcZP1f6Hy1aalGPov1PTfgl/rNV+kf9a9arwbwB4k/4RrT9TuhH5jM8Khc9QWINe42F0L7T4LpV2iZA4B7ZrxcfCSrOXR/5Hu5fOLoqHVf5liiiivPPRCiiigAopk0qQQvLKdqIpZj6AVW0vVrPWbP7Vp0wmh3lNwGOR1p2dri5lexcooopDCiiigDyrx+xPxG0Eeky/8AoQr1WvLviEAPiD4eIHPnDP8A30K9RrsxH8Kn6fqcWH/i1fVfkFFFFcZ2hRRRQAUUUUAFFFFABRRRQAVx/wAL/wDkSx/1+XP/AKNauwrj/hf/AMiWP+vy5/8ARrUAdhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFUtZ/wCQFf8A/XtJ/wCgmrtUtZ/5AV//ANe0n/oJoAxvhv8A8k10H/rzT+VdNXM/Df8A5JroP/Xmn8q6agArzz4u/wDIEtv+uy/zr0OvPPi7/wAgS2/67L/OurCfx4nJjP4Eix4b8YXNz4qg0CS3VYVs0dH78IDzXK+CLaSX4tX8yfcimm3fiTVzw1/yV63/AOwcv/osVW8D3nk/FbUbbbnz5pefTBJrv5FGM+VbxR5/O5yhzvaTO4+Jf/JP9R+i/wDoQrxXxH/x5aJ/14L/ADNe1fEv/kn+o/Rf/QhXi/iBd1roY/6cF/marL/gXq/yIzL436L8xfAhx470g/8ATwK938U+I4fDGjG/njaUbwgVepJr588NXYsPE1hdN0imDGvXfi2+/wACQuP4p4z+hqsZTU8RBS2YsFUcMNUcd0aPjsHU/h/JLGRHuRZcN6YzivINL/5EHXf+u1v/AOhGvXfFCJP8Mxk8C3QjHqFryLS/+RB13/rtb/8AoRpYP+E1/eX5oeN1qp/3X+TPVfhD/wAiGv8A18yf0rua8w+EPiKzGl/2G25boSNKCejA9h+Vd8de00awNLN3GLwjIizya87FQl7eWnmelhKkPYQ18jRooorkOwKKKKACs/XUEmhXat08pv5VoVR1r/kCXf8A1yb+VVD4kTP4WeJeBf8AkJ6Kf+og3/os173XhdldW2g2fhvUGiOxZfMkCdWJUjNe3Wlyl5aR3EX3JF3CvQx93JS6a/medl9lFx66fkTUUUV5p6YUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABXO+Pf+RE1X/r3auirnfHv/ACImq/8AXu1a0f4kfVGVb+FL0Z84UUUV9cfGD42YfKH2gkZr6a8M/wDIrablt/8Ao6fN68V8xV9NeFP+RR0v/r1T+VeTmnwR9T2cq+OXoaMd3BLcPBHKrSx/eUHkVNXnHgyQ/wDC1PFCsx68An/ar0evIrU/Zyt5I9mjU9rHm83+AUUUVibHE/E/X5tH8OGG2yJLr5C3ovQj9ab8I/8AkQ4/+viT+YrI+M+7+zrLA4ycn8RWr8Ij/wAUMg/6byfzr0nFLBJrqzzFJvHNPojuqKKK809MKKwdU8U2+meJtO0WSJ2lvvuuOi845/Kt6qlGUUm+pMZxk2l0PLviF/yUHw9/12H/AKEK9Rry34g/8lC8Pf8AXZf/AEIV6lXVX/hU/R/mcmH/AItX1X5BRRRXGdoUUUUAFFFFABRRRQAUUUUAFcf8L/8AkSx/1+XP/o1q7CuP+F//ACJY/wCvy5/9GtQB2FFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUVS1bVrXRNNkvr9ykKYBwMkk8AAepriF+MujSTBU0zVxH5/kGR7JgqvnoT260AeiUUA5ANc74p8b6Z4QCHVYbx0ddxe3gMioM4yT2oA6KimQTLcW8c0edkih1z6EZrlpviLpUetT6ZHZapcT27KsjQWZdFJ6cigDrKpaz/yAr/8A69pP/QTVuKQSwpIAyh1DAMMEfUVU1n/kBX//AF7Sf+gmgDG+G/8AyTXQf+vNP5V01cz8N/8Akmug/wDXmn8q6agArx34keL9O1fGm2e9pYZhlsfKcdea9ir5y8ZxrH43u1RQo3jgCvRy+EZVG30PMzKpKFJJdTs/DX/JXrf/ALBy/wDosUnhK2iPxav5NvzI7lT6Zzml8Nf8let/+wcv/osVmeG7v7H8Vb+6uJAkUTzswJwSBnGPWuuSbUrfyo5ItLlv/O/0PQ/iX/yT/Ufov/oQrxnWGV49DCkHFioPtya9P8UeIbXxL8K9RvrEMI9wT5xg5DCvF7Rme7iDMTjgZPQYq8BBqm77pv8AIjMKilUXLs0vzLWg263fiKzgk4WSUA16h8SdWstS8EmCwmWU2tzGj7T0+U14+GZJNyMVYHgg4IrqNJBk8B6szHd/pCnJ5/hNdNemnONRvY5cPVahKmlvcS68Y3R8GwaAikBSWkkY8nngfTmq+mNjwJrg9Zbf/wBCNZN/dLe3okRNq7VXA9hinaj9qim8q5O3dGh2pwCMcZFaKmkrLS7uZOpJu71srfgdJ8LY5H8eWjopKorbiOg+U1qaczH47Pkk4vJB19jWT8L2I8fWQBIBD5Gevymun8Nxo/xk1ZmUEpOxU+nNcld2qTf907MOr04L+9+h63RVSXU7ODUI7GWdEuZRlIyeW+lc14k8R32neOdB0m1YLBeMDLxkkbsYrwoU5Tdl6nvzqxgrv0OwormovF6S+OpfDphVfLjLmUt1OBgfrXS0pQlC3N1KhOM78vTQKo61/wAgS7/65N/Kr1Uda/5Al3/1yb+VKPxIc/hZ4XrfHg/Qj7j+Ve0aRe29l4Ptru4lVII4dzOTwBXhevXEp8O6PbmIiJYQ4k9TzxXpUcqyfAd8OGZbEhuc4O6vZxVPmhBP+b8zxMJU5Zza6R/I7uz1G1v7WK4tJ0kilGUZTw1WScda8S02eWO88CxxyuqMDuUMQD+8PUV2PxVu7y10O3NlIYz5yncrYOc8VwywtqkYJ73/ADO+GLvSlNra35XO8oqrpTO+j2bSnLtAhY+p2jNWq42rOx2p3VwooopDCiiigAooooAKKKKACiiigAooooAKKKKACud8ff8AIiar/wBe5roq53x9/wAiJqv/AF7mtaP8SPqjKt/Cl6M+cKKK2fDWhtrt5PCoYmO3kkVVGSWAyBX1spKKuz46MXOXKjGr6b8Kf8ijpf8A17J/Kvm6+0650y8FtexmOXaGKnqAa+kvC4x4V0wf9OyfyryczacItHr5WmqkkzzDSNRnsvjZfQwn5Lq6dJB6gZNera1qaaNotzqMylkt03kDvXjtqcfHST/r+f8Aka9O8ff8iDq//Xuf5iubExTqU13SOrCzapVWujZo6Dqya5odtqMaFFnUsFPbnFaNeUTa5faD8HNGutMl8qVn2FtoPGWr1CxkabT7eSQ5Z4lZj6kgVyVqPI+ZbXa+47KNbnSi97J/ecB8Ysf8I7D67xj86t/CE/8AFEqP+m7/AM6zvjN5n9k2u0/Ju+YfiKv/AAg/5E3/ALbP/Out/wC4r1OOP+/v0Or1fX9N0OEvqN1HCSCVVmwW+lZ2jeONH1kWyQzhJ7g4WFjhs4z0rhfjRCBcWU2TliVx26Vx/wAPyf8AhPtIyf8Alt/7Kaqng4Tw/tL6k1cbUhifZJaaHo3i/wD5K74X+g/9CNej15x4v/5K54X+g/8AQjT/ABlrEifEPw9ZWt1+78wedGjd93f8KxlTdSNOK/lf6m0aqpSqSf8AMvyRQ+IBB+ImgD0mX/0IV6nXlfxA/wCSiaB/12X/ANCFeqVGI/hU/T9TTD/xavqvyCiiiuM7QooooAKKKKACiiozcRCYQmRRIRkJnmgCSiiigArj/hf/AMiWP+vy5/8ARrV2Fcf8L/8AkSx/1+XP/o1qAOwooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAOb8d2lre+GhFfXP2aL7VA2/bu5EgIGPc8VyfiPwtqOnw3yHUYodFu7wXUzGIF0dnHA79QBWp8TvDFlrGjfatQ1W7sYoZEbEUjBSQwwNo7+lc3p+meHNR0GTWL7VNclsbe6EMkVxduRvBGMqR6kUAevr90fSuN+K6Kfhtq7FRuFuQDjmuyH3Rjpiud8V+CdO8YwCHVbm/jh27Wjtbkxq49wOtAGxpX/IHsv+vdP/AEEVwmq6PrvhSbWfEmlavB9mcCaW0eBSWCA8ByeK6rSvCdppEFvFb3uoSJb42+fcl8+x9ao3vw+0zULx57q91R1dstB9rPlH229MUAbujX/9qaHZX5XYbmBJSvpuUH+tZHi+9122026TSNJhvYWt38yR7kRlBtOeCOa6CCCO2t44IFCRxKERR2AGAKraz/yAr/8A69pP/QTQBi/DbJ+GegZGD9iTj8K6euZ+G/8AyTXQf+vNP5V01ABXzp42/wCR6u/98fyr6Lr5z8b8eObv/fH8q9TLf4kvQ8nNP4cfU7Dw9IkXxZhkkYKi6aCST0HliuB12Y3Pii/ktHJDTOQynqMmur3Y8bTn/qDH/wBE1wcb+QqyRuC5yCuOgr06Efe5vJfqeXXn7vL5v9D0PTfl+Bl+D3uD/Na5DwhZpfeKrOCZC0TM2/HptNdfZnHwPux6zn+a1U+E0KS65fM6hmS2JXjoc1mpclOrLzZo489SlHyRwkw2zyAdAxH611Oi/wDJPtX/AOu6/wDoJrLutMEWhz3s0Ukc32zyxuBAK4z0rU0X/kn+r/8AXdf/AEE10VJJxVu6OalFxm79n+Rzt55a3Y8rG3avT6Vs+OIUh1y3WMYBsYG/NBXOV0/j3/kPWv8A2D7b/wBFiqelSK8n+hK1pyfmv1KHhXVJdI8TWd1AAXEgTn0bg/zr0Dwz/wAli1j/AK7t/OvLbOYW99BM3IjkVz+BzXo/gXUI9V+JuoX0IKx3Dl1B7AmufFR92UvI6cJP3ox/vfoaWu3PnfGrRlDZ8tthwenWrXjU4+KvhY/T/wBDNcdrWqf2d8YZrxz8tved+ccf/Xrq/Gcwf4meFpOg2g5/4FXJyOMof4X+R2e0Uoz/AMS/NGDrkjp8b4yjMu65iBwcZGBxXtleIa4wb42QlSCPtMPIPsK9vrlxnw0/Q68F8VT/ABBVHW/+QJd/9cm/lWX4p8Wp4ZudOie1af7dL5YKtjbyBn9a1Nb/AOQJd/8AXJv5VyKElyyezOxzjLmit0eFeIf+RG0T/P8ACaLDWLmx+Gd9BG+UurryGDc4XZnj0o8Q/wDIi6J/n+E1mqf+KCcf9P4/9Ar6KKTgk+/6nzMpOM21/L+h0mn/APIR8CfQ/wDow1o/FLxI0+tx6Gke1IHRnb+8eorlddmkt9B8MSwO0ciWzlXU4IPmGofF159o8UfaS4kPlREsDnJCCojSUqkZvpf8zSVZxpSguvL+R3vwq1q6vdWv4byd5OFVFJOFC56CvU68Z+D25/EF3IB8u3J9s5r2avIx0VGu0j2cBJyoJvzCiiiuI7gooooAKKKKACiiigArj/iVrV7oXhlLnTZBHK06oSRnjBrsK8t+L+vWUmmppEb7rpZhIwH8IAP+NdOFhz1oq1zlxk+ShJ3sz0fS7pLzTLeZJUlLRKWZGBGcc1bryf4P6zDb2t3ZXt1h5ZF8hGPbBzj8a9YqcRSdKo4lYesq1JTCiiisDoCud8ff8iJqv/Xua6Kud8ff8iJqv/XA1rR/iR9UZVv4UvRnzvZ2kt/exWtsN0szBEBOMk13XwmjaDxvJDJw6RyKw9xwa5nwahfxnpe0ZxcoT+ddV8OGx8UrtfUz/wA6+ixUm6c4+R8zhIpVIS8zO+KK4+IEo9Y4/wCVe1eGxjwxpw/6d0/lXjPxTGPiI4/6Zxfyr2jw+MeHLD/rgv8AKvNxf+70z1cH/vNU8a+0x2nxrmmmbai3z5P4GvVPHpz4A1Yj/n3/AKivHtVtPt/xXu7cNt33x5+nNevePPl+H+pj/p3x+oqsQl7Si/QjDN+zrL1PNdduo/8AhTujWxP7wybgMdstXe+JPF58KeHdOlSDznkjj+UnHGBmvJNfun/4RjQbYEeWbdmI995rtPiomfDekN5ijEKDZ3PA5rWdKMpQjLZuRjCtKMZyjulEm+LN4tz4b02U4RrhFkCZ6Zwf61xXh/xTeafHpdhYyGMrdEyejBiBiuu+JyKfB+iMVG5YEAOOnyrXlUcjRSpIhwyMGU+hFb4WnGVDlfdmGMqShiOZdkeqfGk/NYj/AGj/ACri/AP/ACPuk/8AXb/2U1rePtTkvtL0RZ9zSm1SV5D0YkVkeAzjx3pX/Xb/ANlNOlFxwri+zFWkp4tSXdHpPi//AJK54X+g/wDQjXIanLt+MqO7cLdpyT04Fbes6v8A2l8a9Lt/L2fYZxDnP3u+f1rm9YdD8Vi0pAT7Su4k8YwKxoRaST/l/U3ryTba/n/Q6rx6d3xC8Pkcgyqf/Hlr1WvJfG0iv4/8PBMFRImCDweVr1qvOxH8On6fqelhv4tT1X5BRRRXEdwUUUUAFUdavTp+i3d0jBXiiZkJ/vAcVerC8ZqG8J3wP/PMn9DV00nNJkVG1BtHmSfFzWbmSzgiSKNzKqyuUB3gkdB2rfu7qWT41aYPMcI9qpKBjj7rdq8h0/nU7X/rsn/oQr1WRt3xp0nvttVB/wC+Wr3a1GnTfuK3us8ChWqVV77v70T1Siiivnz6IK4/4X/8iWP+vy5/9GtXYVx/wv8A+RLH/X5c/wDo1qAOwooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAMHxhpcet6C1g11FBKZEljMjYyyMGA/MV55p+ieIL9r7QtSuLGG3vL9rud1YbmyQQFGcjlaZ461HQx4+1m28Taibf7Np0UunqyscSkNkrgHnIFc7HqPglfBNhqFreRP4ubyWecRv5zSFgGycY6ZoA+hQMKB6ClpF+4PpS0AFFFFABVLWf+QFf/APXtJ/6Cau1S1n/kBX//AF7Sf+gmgDG+G/8AyTXQf+vNP5V01cz8N/8Akmug/wDXmn8q6agAr5w8ef8AI533+8P5V9H184ePP+Ryvf8AeH8q9XLP4j9DyM1/hL1J/Dl1Jd69cSTOXYadKuT6CPArl88Guh8I/wDIYuf+vCb/ANArJ/sy5OkjUFhY2+4qZOwNexGym/l+p4srygvn+h2tvK4+C0qheGumBPp92pPg1z4jvP8ArgP/AEKntx8Fz/12P/stN+DX/IyXn/XAf+hVxT/gVfVnfD/eKXojb+McMcOj2giQKDLk4Hsa4jRnUeAdWUsATOuAT1+U13Xxn/5BFn/11/oa8dDsEKAnaTkiqwcefDpeZGNlyYlvy/QbXo/xK0+2j0PQ75UAuJbVFZv7wCrivOK9R+Jig+EPD5xyLdcf98rW1ZtVqfzMKCTo1Pl+Z5dXcfCuZYPEskjfwxFsZ64ya4+8sp7GSNLlCjSRrIoPdT0NP0vUZtK1CO8twDImQAenIx/WtqsfaU3FdTGjL2VVSfQveIdYXXPFE9+kAhWWbIXv17+9dX8VLsprWky2s2HjtQQyNypzXnp3B9zAjnmrOoXL3UqyOGC7cKD6VPslzRa2VylWfJNPd2L3h+9ZvGGn3V9MWIuEZ5HPYGvpdHWSNXQ5VgCD6ivk+vqXSP8AkB2P/XtH/wCgivMzSKXLI9XKpt80Tzv4pyu3iDQIz91blSPxK16Frf8AyBLv/rk38q89+Kf/ACMGgf8AXyv81r0HW/8AkCXf/XJv5VyVP4dL5/mdlP8AiVfl+R4X4h/5EXRP8/wmsZbhP+ESa33fvDeB9vttxWt4gb/ijdFX/ZB/Q1yuflx717tKN4/N/mfP1pWn8l+R0XiL/kXPDf8A16yf+jDXOhvmBPNb2vTJJoGgIjAtHbOGA7fOawK0pfD9/wCZnV+L5L8kel/CC6x4guIQjfvI857DANezV4/8I+Nduf8Arkv8q9gr57H/AMdn0eX/AMBBRRRXCegFFFFABRRRQAUUUUAFfOvxGBHj/U8/89B/6CK9+vtWsdNkhS+uEhaZtsYb+I14X8TyD40uCvQ4P6CvUy26qvzR5OaWdJeTMXww2fFWlD0uU/8AQq+mq+UoZpbadJ4HMciNuRh1BHevqDRrj7VollMZPMZoELNnqdozWuaR1jIyyqWkol2iiivGPbCud8ff8iJqv/XA10Vc74+/5ETVf+uBrWj/ABY+qMq38KXozxj4dWz3PjSz8sZ8s7z9ARW38Of+SrXP/bx/OqXwo/5HZP8Ari38xV34c/8AJVrn/t4/nXvYh61F/d/zPnsOtKT/AL3+RW+Kf/JR2/65w/yr1mz1mw0nQtKjv7hYWuIlSIEH5jjpXk3xTIHxGYnoI4c/lVj4g65GYdDsoUJktYUuN5+6wI4H6VhKj7anSh5fodEa3salafn+pUc/8Xkm/wCv1q9N+JEhTwNeAHG5MH3rybRdT/tTxtbXs6Ik892ZHI9x0r1H4oPt8FyD+9x+lTXi1WpJ+RWHknQqyXmeHXF3LdWVukrjbbL5ca47E5rt/ipPun0Nf+edoP5LXBhm+xsuBt3gk984rrPiPP52pWI/uWyj/wAdFehOP76HzPOhL9zP5Gt4q1dPE3hvw7DBE0JuG+zguc4KkLmvP7+0aw1Ce0dgzQSNGSO5BxXVWrA6T4QAIyL2TI9P3gqleaFdaz4m1p7YYjt5pHkc9F5OPzxUUWqfu9NfzKrKVX3t3p+Rm6xrEuqR2cbEiO2t1iC9sjvTdB1MaNr1pqBQuLd9xUd+MVnUV1ckeXl6HLzy5ubqdNP4mh/4WIfEMcLNELgTCPPJ4xisnXNSXV9auL5EMYmbIUnJFZ/TrRSjTjFprtYcqkpJp9Xc7C0upbu+8LSXDl3E+3JPYOAK+ga+anu3sdP0W6iwXgZ3UH1DZr3Twh4jOv8AhaLU7sLCS7I3OBkHFeNj6b5YyW2q/E9vLqi5pRe+j/A3bmYW9rLOw3CNC5A74GayfC3iWDxTpJv7aF4VDlCjnJBFY3xF8WS+HdFg+yRpMb7fHlugGOv61ynw78Yad4e8NxWt6WL3F4w+U/cGBgmuWGGlKg5pa30/U654qMa6pt6W1/Q9eorNsNfsdS1G5sbaUNPbHEienv8ArWlXG4uLsztjJSV0FYPjVtvhG9/65n+RrermviBMIfBt2xBORt4960o61I+pnW0py9D54tpDDdRSKMlHDAeuDXo2i3w1H4r6TcA/M1uN4/una2RXnVowS8hkbhFkUsfQZrs/BMsU3xRS5SQeV5rsG7EEGvpMQrxb8mfL4aTU4rzR7xRQCGGQcj1FFfLH1oVx/wAL/wDkSx/1+XP/AKNauwrj/hf/AMiWP+vy5/8ARrUAdhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUARSWtvK26WCN29WQE00WNoOlrCP8AtmKnooAKKKKACiiigAqlrP8AyAr/AP69pP8A0E1dqlrP/ICv/wDr2k/9BNAGN8N/+Sa6D/15p/Kumrmfhv8A8k10H/rzT+VdNQAV84eO/wDkcr7/AHh/Kvo+vnDx3/yOV9/vD+Verln8R+h5Ga/wl6ieEf8AkMXP/XjN/wCgVpp/yST/ALeW/wDQhWZ4R/5DFz/14zf+gVqJ/wAkjP8A18t/6EK9Kr8a9V+p5lL4H6P9CGx1Saf4dX1hJgxwsHU9+SBitf4M/wDIx3n/AFwH/oVc3pf/ACKOq/7q/wDoQrpPgz/yMd5/1wH/AKFUV0lRqW/rY0w7br0r/wBbm98Z/wDkD2f/AF1/oa8b617J8Z/+QPZ/9df6GvHEOHH1p4D+Aicx/wB4Zdh07zdEub/fgwSpHt9d2f8ACvV/F3h7Udb0Hw1JpsBnEEcfmIPQhefpXnFsuPA+on1uov619CaF/wAi9p//AF7R/wDoIrnxlWVNxkujf5I6MDRjUUoPql+bPDvidC1v4rihcBWSziUheg4NcjF/rk/3hXbfF3/kfG/69o/61xEX+uT/AHhXdhnejF+RwYlWryXmdffabJeeGbu5tbYyeRqG6V1H3E2Dr7ZqzP4aGuXlpa2rLCy2Zkz6kNWvpZ/4tt4rx/fP/stY3wvd5PEdxvdji1YDJzjkVzc8lGUl9n/JHVyRc4xa+L/NnFSxmKZ4zyUYqfwNfUWj/wDIDsf+vaP/ANBFfMN7/wAhC4/66t/Ovp7R/wDkB2P/AF7R/wDoIrDM/hgdGVfHM85+LEqwaxokshwkc4Zj6AFTXe6hdRXvhia5t23RSwFkYdwRXnPxq/1mnfVv5Cu00z/kmdr/ANeI/wDQa5ZxXsKUjrpyf1irHyR434gb/imNGX/pkD+hrBsLKXUL6K0t+ZJW2rn1rY19/wDiS6Qn/Tup/nTfBADeNdMDf89uPyNe3F8tJv1PCklOqovyMORWSRo3PKEj6U2p7wE6hOAMnzW/nTJ7eW2l8u4jaN8A7WGDg1smc7R1XhDxUnhbXvtFzEzwPEAyr16cV73YXiahp1veRAqk8YkUHqARmvmS/Cb8k/N5aYH4V9HeGP8AkVdM/wCvWP8A9BrxcxpxSjNbs93LKkm5U3sjVooorxz2gooooAKKKKACiiigDxz4xTyp4k06NXYIId2B67jzXN/EPJ8UPn/nmn/oIrQ+KerR6j40ECKVNkPJYnuc5/rVDxl/xM/Ght7bDMQiD0J2ivo8NFxjTv2f6HzOJkpSqW7r9Tm5nja1gVB8yg7vzr1T4NapcSJeWM0jvEmGjU9E65ryq5jjikKRsxKkhsjGDV/THntdMvL21uZIXjKxlV/iDcVviKaq0nHuc+HqulVU+x9OZor5+1jxzNqXhvSrWJ5Ybqzc+YVJwygDac9zxXsHg7xFba5odsFuklvEhUzoDyp968GthJ0YczPoaOMp1p8iOhrnfH3/ACIuq/8AXA10Vc74+/5EXVP+uBrCj/Fj6o6K38KXozyT4VMF8ax5OMxMP1FaHw7Xb8WLke1x/Ok8CadCPiV5MQKJFD5gA9cKak8AEH4t3O05GJ/517lZp+0a/l/zPAoJpU0/5v8AIq/Fq2lTxyZ5UZYJIo1VyODgc1keOQo1PTxGcr/ZsGD68Guo+M8qNqVnGGBdVJI9OBXB61qn9qz20gXHk20cH/fIrTC3lTpy7Iyxdo1Kke7KVtJ5V1FJnG1wSR9a9n8fXg1X4b217agvFKA2fQYxXkEek3k2lPqEVvI9vG215Avyr9TXpupX0EPwT05HYbni8tR/tcnFLFJOdOS3TKwrap1IvZq55UwaOJo3G1twJBrT8T3yX+seZDKJY1iRVI6DCjNZ17OtzdvKg2q2OPwqXTNLutXuHt7CEyzKhk2jqQOtdjSXvM4k27wj1JdBYnxFpoJOBdR4GenzCvbb3wnBotjrt9YvJJJfRlnjIGF6kkfnXieho0fiewSRSrLdxhgex3CvpTVv+QLe/wDXvJ/6Ca8vHzcZxt1/zPWy6EZU5t9P8j5aJJ60gzkY61IYWECzcbGYqKjr1zxR8xkaZjNnf3zTKCc9aKAJpLmSW3ihY/JFnb+NehIzJ8AiUJU/bOoOP4683r0cf8kBP/X7/wCz1y4j7H+JHZhn8f8AhY74hOX8A+FyeTsP/oC153BL5c0TNkqjhsfjXd+OryCTwX4ctVlUzxx7njB5UFRg1wGDjOOPWjCr91Z93+YsW/3t12X5HrXw11KLVPHus3cCsqTRb1DdQMivTTqVmNQFibhPtRXcIs/Nj1r5t8OeI73wzqX2uwYZYbZEPR19K6rwn4ml1j4mQ3t7lGuWCKi8henH04rgxWDcpOfRL8j0MLjYxgodW/zPcHdUXc5AA7mua+IDD/hDLojkEDFZ3xamkg8EFoXZGNygypwcc1F4hvY7z4VLLFIsn7qNSVOeQuK4KNJrkqd3Y9GtVT56fZXPMtMhgb4faxIYh5yvF85HbdXPwzCGOQxuysQMHODnNdTpssbfC/Vo1I8xZYyw9Bv4rjq+ip6uV+/6I+bq6KFu36s+lfBUrzeCdJklYu7W6lmPetysDwL/AMiJpH/Xstb9fLVf4kvVn1dH+HH0QVx/wv8A+RLH/X5c/wDo1q7CuP8Ahf8A8iWP+vy5/wDRrVmanYUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAVR1DW9L0lkXU9QtrRn+6JpQu76Zq9XBfF7S7O58B6lez26SXEFuRG7KCU57elAHeKyugZCGVhkEdxVG513SrK8W0u9RtobhiAsUkoDHPtUmlf8gey/690/9BFeYWWn22s+FfE+sahEkl8vm+XK4y0WzcFKnt0HSgD1oEEAg5B6GqWs/wDICv8A/r2k/wDQTVLwhcSXXhHTZJiWf7OgLH+L5RzVXxffa3baZdJpWjx3sLW7+ZI1yI9g2nPB60AN+G//ACTXQf8ArzT+VdNXMfDbJ+GmgZGD9iTj8K6egAr5w8d/8jle/wC8P5V9H183+O/+Ryvf94fyr1cs/iP0PIzX+FH1E8I/8ha5/wCvGb/0CtVP+SRn/r5P/oQrK8If8ha5/wCvGf8A9ArVT/kkZ/6+T/6EK9Kr8a9V+p5lL4H6P9DM0v8A5FHVf91f/QhXSfBn/kY73/r3H/oQrmtL/wCRR1T/AHV/9CFdN8Gf+Rivf+vcf+hVGI/g1P67F4b+NT/rub3xn/5A9n/11/oa8dhAaZAehNexfGf/AJA9n/11/oa8eg/4+E/3qMB/u6DMP95fyPS/GXhyz8P/AA/hFkDm4eOSQk9Tj/69ep6F/wAi9p//AF7R/wDoIrg/if8A8iDY/SL+Vd5oX/Ivaf8A9e0f/oIryq8nKhFvuz16EVGvJR2sjxb4u/8AI+N/17R/1rhq9U+M2kQRz2mqpnz5sxPzxhRx/OvK69rByUqEbHhY2LjiJXPRvDBLfCPxIWJJJ6n/AIDVD4V/8jJcf9ezfzFX/C3/ACSHxH9f/iao/Cv/AJGW4/69m/mK55fBW9f0R0x/iUfT9Wcdff8AIQuP+urfzNfT2j/8gOx/69o//QRXzDff8hC4/wCurfzNfT2j/wDIDsf+vaP/ANBFYZn8EDfK/jmeX/Gr/Wad9W/kK7XTP+SZ2v8A14j/ANBrivjV/rdP+rfyFdrpn/JM7X/rxH/oNYT/AN2p+p0U/wDeqvoeNeIAP+Ea0Y458oc/gah8Df8AI8aV/wBdx/I1uRaSdVt/DkEsEklvIoVygPoe9afhjSbWy+LN/ZwR/u7U/ug3JU4H+NelKrFU5R62Z5kaUpVYy6XX5XPPJv8AkMS/9dm/nXW/E3SFsdSsL1ZNxvbZSU2427VArK1zwvqmiXv2nUIVjjlmO0hwc5Jrpvi5/q/D/wD16H/2WqdROrT5Xo7kqm1Sqcy1VjgdQ/4+V/65r/KvpLwx/wAirpn/AF7J/wCg1833YzfxA91T+VfSmgDb4dsAOgt0/lXHmXwQO3K/4k2aFFFFeIe6FFFFABRRRQAUUUUAfOPj04+IGrZ/5+P6Cm6eRe+Lra5tVkeIzooJHfFX/idAkPjS6ZBgyHc3ua6b4NQRTJemaJHKSBlLKDtOOor6WVRQwyn5Hy8abninT8zznXdPn0zWri2ugBIG3EA54PIqS0/5FrUP+ukX8zXb6tp8F98ZJ7O6TfFIqhh9QKpePLO30+91W2tIliijFuFVRjtVRrqXLF7tJ/kTKg4801sm1+ZwQBPQZr034J/8hjVP+uCf+hVxOu2S21xa/ZlI8y1RiFHfHJrtvgn/AMhjVP8Argn/AKFSxclLDSa/rUMHFxxUU/60PY653x9/yI2qf9cDXRVzvj7/AJEbVP8Arga+eo/xI+qPpa38KXox/hawtV0izvFhQXBi2mQKMkU3TPBel6T4hm1e1RhPICAMnC56/nVvwv8A8i1Zf9cxWtTnUkpSSe5MKcHGLa2PF/jJarHrlvcgndKm0j6CvNq9V+NUZE2nP2bcP5V5VX0WCd6ET5rHK2Ike3WWmfbfguba1VFkmtlYnpk8Hn8q89u5pJPhjbpJJvEWobFH90bOleveF4g3w1sVA+9ZZP5V42//ACTxv+wqf/QK48NK8pL+8d2KjaEX3j/kc26lGKnqK7j4Q/8AI9D/AK9pP6Vx+op5d/Io7Y/lXYfCH/kex/17Sf0ruxLvQk/I8/Cq2IivMwof+SgL/wBhL/2pX0Vqg3aPeAnAMD8/8BNc3H8NdBTWDqPlymbzfNH7w4DZz0+tdHq3/IFvf+veT/0E14uKrwrShy9D3cLh50Iz5up82vZs+is8ZDJBKSxz1HArPgieaQKiM59FGTV+zLw6TO0vMExCYDdCDk8V7F4Z8MaNZ+DzqVhDve6t94eTkj6Z6V7FbEKgtdbs8Wjh3XlppZHh8EXm3KRn+I4rWh0uB/Bt7qLA+fDeJCpz/CVJP8qo2X/Iaj/66mtuD/km+qf9hOL/ANBatakmrW8vzMacU737P8jC0/TbjU5pY7UKWiiaZtxx8qjJrvMY+AJ/6/P/AGeuc8Ef8hO//wCwbcf+g1QPiK//AOEaGhb1+xCTzNu3knOetZ1IyqTSXRpmlKUacHJ9U1+RW1LUG1GWF3TZ5UKxDBzkL3rR8JeHl8S6ylg9w0AbncFzUGmQo+has7KC0aR7SR0y1dB8Kv8Akcovp/jVVZclKXLpYmlDnqx5tbnNeINLGieILzTVlMotpCgcjG78KZpF+2l6gl9C+yeD54jtzlh0zWx8QonTxzqbMjKGmJBIxniuZHvV03z01fqiKi9nVaj0Z614t1efXfg5a6hd482W5Xdj2LCuZ8NSSSeAtcjZ2YCWIKpbgZzTNb1WeL4c6Fpa48icPK3HOVcgfzrmLK7ngkEcUrLHIwDpnhvqK5qVH924rvf7mdVat+9Un/Kl96O4HhufQ/hdqF3exqHu2iIAbtu4P6158vLD617h46GPhEn+5B/SvD1OGB96eDqOpGUn3JxtNU5xjHsfSvg1dvgvSwOgt1rbrE8GNv8ABelNjGbdeK26+dq/xJerPpqX8OPogrj/AIX/APIlj/r8uf8A0a1dhXH/AAv/AORLH/X5c/8Ao1qzNDsKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAK4/4g+Gdf8V6S+maNqtnYW08ZScT25kZuexBGK7CigDntE07xLY2dpBqeqWNx5KhZDFbFNwHHHPHFYWpeAtXe9vYtG1a3tdJ1Aj7VbSQlnI/i2sCMZye1d9RQBW06yj03Tbeyg/1dvGsa/QDFR6z/AMgK/wD+vaT/ANBNXapaz/yAr/8A69pP/QTQBjfDf/kmug/9eafyrpq5n4b/APJNdB/680/lXTUAFfN/jr/kcr7/AHh/KvpCvm/x1/yOV9/vD+Verln8R+h5Ga/wo+o3wh/yFrn/AK8Zv/QK1k/5JGf+vk/+hCsnwh/yFrn/AK8p/wD0CtaP/kkTf9fJ/wDQhXpVfjXqv1PMo/A/R/oZel/8ijqn+6v/AKEK6b4M/wDIxXv/AF7j/wBCrmdM/wCRS1T/AHV/9CFdN8Gf+Rivf+vcf+hCpxH8Gp/XYrDfx6f9dze+M/8AyBrP/rr/AENePW//AB8J/vV7D8Z/+QNZ/wDXX+hrx63/AOPiP/epYD/d18x5h/vL+R7F8T/+RBsfpF/Ku80L/kXtP/69o/8A0EVwfxP/AORBsvpF/Ku80L/kXtP/AOvaP/0EV5FX/d4+rPZpf7xL0RwXxp/5A+n/APXV/wCQrxqvZfjT/wAgfT/+ur/yFeNV7GX/AO7r5ni5j/vD+R6N4W/5JD4j+v8A8TVH4Vf8jLcf9ezfzFXvC/8AySDxH9f/AImqPwq/5Gaf/r2P8xWcvgrev6I0j/Eo+n6s46+/5CNz/wBdW/ma+ntH/wCQHY/9e0f/AKCK+Yb7/kI3P/XVv5mvp7R/+QHY/wDXtH/6CKwzP4IG+V/HM8v+NX+t0/6t/IV2mm/8kytf+vEf+g1xfxq/1un/AFb+QrtNN/5Jla/9eI/9BrCf+7UvU6Kf+9VfQj+HUaP4LsGZQSqAqSOnFcvon/JbNY/3/wDCuq+HH/Ik2X+4P5Vyuh/8ls1j/f8A8KI/HW9H+Y5fBR9V+Ra+L/8Ax62P++P61g/Fz/V+H/8Ar0P/ALLW98X/APj0sf8AroP61g/Fz/V+H/8Ar0P/ALLW+F/5df8Abxz4v/l7/wBunAX5xcoR18tf5V7n8LdVn1Lwigun3yQyMgOf4RjArwy//wCPhf8Armv8q9k+D0YHheRx1aVga3x6Tw9zny5tYiyPQ6KKK+ePpAooooAKKKKACiiigDyLxdawXGu+JHmjV2hsC8ZYZ2ncORU3wW/1V/8A74/lVvU9Ok1bxd4hsYCFknsCqk9vmFVvg0hi/tKNuqS7T+Ar2ZS/2WS9DxIx/wBri/UgvQR8cmJHBC4/IVn/ABI/5DOsf9u/8q3NX/5LHa/7v+FYnxGRpNe1ZEGWY24A9Tirou9SH+FfmiKytTn/AIn+TM25iM2t6dGgyz2Kqo9SUrofg9ZT2PiDV4LqMxyRxKrA9iGqhJbNa+NtDikGGFrFkeh212nhf/koWv8A+6P/AEI0V6n7lxXVfqGHp/vlJ9H+h29c74+/5EbVP+uBroq5zx9/yI2qf9cDXkUf4sfVHs1v4UvRl3wv/wAi1Zf9cxWtWT4Y/wCRas/+uYrWqanxsqn8C9Dyj42/6vSvrJ/SvK4rdpYJpV6RAE/icV7F8XdGv9UtrCSwtmmS3EjSlSPlGB/hXklpPHHp97G5w8iqEHrg19Dgpf7Orf1qfN4+P+0tv+tD6B8Kf8k50/8A68f6GvGH/wCSen/sLf8Aslez+FP+Sc6f/wBeP9DXiLXcX/CGm03fvf7S8zb7bcVy4RXnP/F/mdeMdqcP8L/QJtLW+1TVSzbfstsZh7kAf410Xwfs5T4ra6MbeWsDrvxxziqEa7dX8RD/AKhzfyWvTPhaijwPbMAASWyfXk1riqrjQa72X4GOFoqVdPtd/idlVbUYnn0u6ijGXkhdVHqSpFWaK8FOzufQtXVj5yPhu/03TZpNThaNDMIVBP8AECM/zr3Gxto7TwRFbwghI7baoPpXP/Fb/kCWf/Xcf0rpoefCi/8AXvXo4itKtThJ9zzMPRjRqTgux842f/IaT/rqa24P+Sb6p/2E4v8A0FqxbIZ1yMf9NT/Ou98UeFT4X+HkymcTfa7yKXgdPlPFevVnFSjF7u35njUYScZSWyT/ACKvw9022l8N+INQYMbiC2aNDngBlOf5VwjxKtnFIPvMSD+Fej/DRd3g3xMCCQY+g6n5WrzeV/lEYBGxjwaVJt1Z+q/IdZJUab8n+Zq6T/yLutf7kX/oVb3wtRk8YQ7hjcuR9Oals9OtR8Grq/WFBctcFGlxyQGGB+tHwyjaPxhbb23bogR7DnioqzUqVS3n+RpSg41aV/L8z13WvDOl65Cy31qjuRw4ADD8a+f/ABL4Zv8Aw1f+TfRbUky0TA53LmvpeuI+Jvhy517SIPsEPmTRyDJ9F5ry8FiZU58rejPWx2FjUg5xWqPD5r24vLe1tXO9LcFYlA55OTUMJ8u5QvxtYZ9ua7L4XafBc+O/IvYUlEUTsA4zhgRzUXxB8MHw1rizK6SRXTNIqAdOckH869r20FV9l8zwvYzdL23nY9a1HTrfWvh0ILjcYvsayjaccquRXzxAM3EYP98fzr6a0RVufC9ikijZJaoGXtgr0rz3x58PrDT9NXVNJC24hkBkjx94EgDFeXgsRGnKVOXV6HrY7DyqRjUj0Wp6H4eVU8O2KoAFEQwB2rSrP0D/AJF+y/65CtCvLn8TPWh8CCuP+F//ACJY/wCvy5/9GtXYVx/wv/5Esf8AX5c/+jWqCzsKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAqlrP8AyAr/AP69pP8A0E1dqlrP/ICv/wDr2k/9BNAGN8N/+Sa6D/15p/Kumrmfhv8A8k10H/rzT+VdNQAV83+Ov+Ryvf8AeH8q+kK+b/HX/I5Xv+8P5V6uWfxH6HkZr/Cj6jPCH/IVuf8Aryn/APQK2I/+SRN/18n/ANCFY/hD/kK3P/XlP/6BWxH/AMkhb/r5P/oQr0qvxr1X6nl0fgfo/wBDJ0z/AJFPVP8AdX/0IV0/wY/5GK9/69x/6EK5jTP+RT1T6L/6EK6f4Mf8jFff9e4/9CFRif4NT+uxphv49P8Arub3xn/5Atn/ANdf6GvH7f8A4+Y/96vYPjP/AMgWz/67f0NeP23/AB8x/wC9RgP93QZh/vL+R7F8T/8AkQLL6Rfyru9C/wCRe0//AK9o/wD0EVwnxP8A+Sf2X0i/lXd6F/yL2n/9e0f/AKCK8ir/ALvH1Z7NL/eJeiOC+NP/ACB9P/66v/IV41Xs3xp/5A2n/wDXV/5CvGa9jL/93XzPFzH/AHh/I9G8L/8AJH/Ef1/+Jqj8Kf8AkZp/+vZv5ir3hf8A5I/4j+v/AMTVL4U/8jNP/wBex/mKzl8Fb1/RGkf4lH0/VnG33/IRuf8Arq38zX09o/8AyA7H/r2j/wDQRXzDff8AIRuf+urfzNfT2j/8gOx/69o//QRWGZ/BA3yv45nl/wAav9dp/wDwL+QrtNN/5Jla/wDXiP8A0GuL+NX+u0/6t/IV2mm/8kytf+vEf+g1hP8A3al6nRT/AN6q+g34cf8AIk2X+4P5Vyuh/wDJbNY/3v8ACuq+HH/Ik2X+4K5bQ/8Aktms/wC9/hSj8db0f5jl8FH1X5Fn4v8A/HpY/wDXQf1rB+Ln+r8P/wDXof8A2Wt74v8A/HnY/wDXQf1rB+Ln+r8P/wDXof8A2WujC/8ALr/t458X/wAvf+3Tz+//AOPhf+ua/wAq9n+D/wDyKbf9dmrxi/8A+Phf+ua/yr2f4Pf8io//AF2aujH/AO7nNl/+8noNFFFfOn0oUUUUAFFFFABRRRQBw2nf8lR1X/r0P/oQrI+EX/Hzq/8A18GtfTv+So6r/wBeh/8AQhWR8Iv+PnV/+vg16kv4M/SJ5Mf40PWQur/8litf93/Csnx3/wAjZf8A/XS1rW1j/ksVp/u/4Vk+O/8Akbb/AP66WtbUfih/h/UxrfBP/F+hPrX/ACUnSf8Ar3i/lXT+F/8Akoev/wC4P/QjXMaz/wAlJ0n/AK94v5V0/hf/AJKHr/8AuD/0I1nV/hf9u/qaUf4v/b36Hb1znj7/AJEbVP8Arga6Ouc8ff8AIjap/wBcDXnUf4sfVHp1/wCFL0Zd8Mf8i3Z/9cxWtWT4Y/5Fuz/65itapqfGyqfwL0KWs/8AICv/APr2k/8AQTXzXY2gn02/lMZZolTaQOmWxX0prP8AyAb/AP69pP8A0E1xHwo0yzuvBDPcQJI0lwwbI6gYIr0MJW9jSlLzR52Mo+3rRh5M6TwpGV+H+nRyAqfsYDD04rwO/aCCaWxgDNtutyue4xX00sarH5aqAuMAD0rwt7GB/jYbNowYDfbSnbG2rwNRc1ST9TPH0nyU4r0GSLt17xKPTTm/9BWvSfhd/wAiJafVv/QjXAatGI/GPi1AMBbFwB+C13/wu/5ES0+rf+hGqxbvQT9PyJwatiGvX8zsKKKK8c9oyfEPh638RWSW90zIEcOpXrn/ACKtTwLBo8kMf3UiIH5VcqG8GbKb/cP8qtSdlHoQ4xTcurPmOx/5D0f/AF2/rXuHj/RbvXPA8cFgm+WMpJsHVgB0FeH2P/Iej/67f1r6dtv+POH/AK5r/KvZx83TnCa6HiZdBVIVIPqeefDjwzfaf4U1L7Wmx74ELEfvLgEc149qERh1K5ibgpKyn8DX1SAAMAYHtXy7rv8AyMOof9fMn/oRp4Gq6tScn1JzCjGlShFdLnd2ClvgLdgDJ+1Hp/vLUHw6Xb4wssjB+zr/AFrsvhPDHP8AD8RzIHQ3MmVI47VkaXGkXxhnSNQqrwAO3WodT+LT9WaKl/BqeiPU6CARg9KKK8U9w5rRfA2maFrs2q2ZfzpQRtOMLnriuQ+Llm19rWh2y8GZimfTJAr1Sq9xYWt1cRTXEKySQ/cY/wANdNPESjVVSWtjlq4aMqTpx0uJptqbHS7a0J3GGJY8+uBisL4hf8iVd/7yf+hCumrmfiF/yJV3/vJ/6EKii71ovzLrK1GSXY1tA/5F+y/65CtCs/QP+Rfsv+uQrQrOfxM0h8CCuP8Ahf8A8iWP+vy5/wDRrV2Fcf8AC/8A5Esf9flz/wCjWqSzsKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKhuru3sbdp7yeOCJfvPIwUD8TQBNRVey1Cz1GHzdPuobmPpvicMPzFQXuu6Vpswi1DUrW2kYZCTTKpP4E0AX6paz/AMgK/wD+vaT/ANBNWbe5huoVmtpUmjbo6NkH8araz/yAr/8A69pP/QTQBjfDf/kmug/9eafyrpq5n4b/APJNdB/680/lXTUAFfN/jr/kcb7/AHh/KvpCvm/xz/yON7/vD+Verln8R+h5Ga/wo+ozwj/yFLn/AK8pv/QK2I/+SQt/18n/ANCFY3hH/kKXP/XlN/6DWzH/AMkhb/r5P/oQr0qvxr1X6nl0fgfo/wBDJ0z/AJFPU/ov/oQrp/gx/wAjFff9e4/9CFcxpn/Ip6n9F/8AQhXUfBj/AJGK+/69x/6EKjE/wan9djTDfx6f9dze+M//ACBbP/rt/Q149bf8fMf+8K9h+M//ACBLP/rt/Q149bf8fUf+8KMB/u6DMP8AeX8j2L4n/wDJP7L6Rfyru9C/5F7T/wDr2j/9BFcJ8T/+Sf2X0i/lXd6F/wAi9p//AF7R/wDoIryKv+7x9WezS/3iXojg/jT/AMgbT/8Arq/8hXjNezfGj/kDaf8A9dW/kK8Zr2Mv/wB3XzPFzH/eH8j0fwv/AMkf8R/X/wCJqj8Kf+Rnn/69j/MVe8L/APJHvEf1/wDiao/Cj/kZ5/8Ar2P8xWcvgrev6I0j/Eo+n6s42+/5CNz/ANdW/ma+ntH/AOQHY/8AXtH/AOgivmG//wCQjc/9dW/ma+ndG/5Adj/17x/+gisMy+CBvlfxzPMPjT/rtP8Aq38hXaab/wAkytf+vEf+g1xfxp/12n/Vv5Cu003/AJJla/8AXiP/AEGsJ/7tS9Top/71V9Bvw4/5Emx/3BXLaH/yWzWf97/Cup+HH/Ik2P8AuCuW0L/ktms/73+FKPx1vR/mOXwUfVfkWvi//wAedl/10H9awPi5/q/D/wD16H/2Wt/4v/8AHlZf9dB/WsD4uf6vw/8A9eh/9lrowv8Ay6+Zz4v/AJe/9unn99/r1/65r/KvZvg9/wAiq/8A12avGb7/AF6/9c1/lXs3we/5Fd/+urV0Y/8A3c5sv/3k9Cooor50+lCiiigAooooAKKKKAOG07/kqOq/9eh/9CFZHwi/4+NX/wCvg1r6d/yVHVf+vQ/+hCsj4Rf8fGr/APXwa9SX8GfpE8qP8aHrIXWP+SxWn+7/APE1k+PP+Ruvv+ulrWvrH/JYbT/d/wDiayPHn/I333/XS1raj8cP8P6mFb4J/wCL9CxrP/JSNJ/694/5V03hj/komvf7g/8AQjXNaz/yUjSf+veP+VdL4Y/5KJr/AP1zX/0I1nV/hf8Abv6mtL+L/wBvfodvXOePv+RG1P8A64GujrnPH3/Ijan/ANcDXnUf4sfVHo1/4UvRl3wx/wAi3Z/9cxWtWV4Y/wCRbs/+uYrVqanxsun8CEdFkjZHGVYEEeoqGzsrewg8m0iWKPOdqjvU9FTd2sVZXuFeMXFo9t8doHcqRNeb1wegx3r2evLvENqlv8aNDdOszb2+vNdmDlZyXdM4cbG8YPtJD/EvhW6s7nxRr07p5NxbMsSqckjA5P5VvfC7/kRLT6t/M1oeO/8AkR9V/wCvdqz/AIXf8iHafVv/AEI1Uqkp4a77pfgKNONPFWj2b/E7CiiiuA9AKiuf+PSb/cb+VS1Fc/8AHpN/uN/KmtxPY+aLF0/tNEx8/wBpznHavpe1/wCPOH/rmv8AKvmSxT/idI//AE3xX03a/wDHnD/1zX+Vevmf2Txsr+0S14j4c0az1PWfFE95F5jW3ntHnsfm/wAK9urgPCfhO+sdW8Qi/ASK+LCNlOeG3f41yYaooQnrbb8zsxNN1Jw0utfyJvhD/wAiGv8A18yf0rIsP+Sy3P1/xruPDHh6Lwzoq6fBIZFDl9xGMk1xFh/yWS5+v+NaRkp1Ksls0zKUHCnSjLdNHp9FFFecemFFFFABXM/EL/kSrv8A3k/9CFdNXM/EL/kSrv8A3k/9CFbUP4sfVGNf+FL0ZraB/wAi/Zf9chWhWfoH/Iv2X/XIVoVnP4mXD4EFcf8AC/8A5Esf9flz/wCjWrsK4/4X/wDIlj/r8uf/AEa1SWdhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABXD+O1F94g8O6Zcc2s14rSoekgwRtPqPau4rnvFnhybXre1l0+5S1v7KYT28zqWUMAfvAdRz0oAwvD8Eei/ErUNN05BFZ3EfmmJBhUKgAYHbrVTxVomo/8J0msR+HbfWrOOBldZZFG3gdAQc9K2NE8Ka1p091qN7qVrcatdSKXmEJEaoBgqq54zx3q3reneL7m6P9iaxp9rbsgDLPas7ZxyQQRigC54U1ew1rQUudKhWCFXaJolXaEdThl/A1T8Y+KtP0PTbm2vUu2ea3cKYbZ5Byp6kDirnhTw7H4Y0MWMcnmM8rzyv/AHpHOWI9s1d1n/kBX/8A17Sf+gmgDF+Gx3fDPQCO9kn8q6euZ+G//JNdB/680/lXTUAFfN/jn/kcb3/eH8q+kK+bvHP/ACON9/vD+Verln8R+h5Ga/wo+ozwl/yFLn/rym/9Araj/wCSQt/18n/0IVi+Ev8AkKXH/XlN/wCg1tR/8kgf/r5P/oQr0qvxr1X6nl0fgfo/0MfTP+RU1P6L/wChCup+C/8AyMV9/wBe4/8AQhXLaZ/yKmp/Rf8A0IV1PwX/AORhv/8Ar3H/AKEKnE/wan9djTDfx6f9dze+M/8AyBLP/rt/Q149bf8AH1H/AL1ew/Gf/kCWf/Xb+hrx61/4+o/94UsB/u6DMP8AeX8j2P4n/wDJP7L/ALZfyrutC/5F7T/+vaP/ANBFcL8T/wDkn9n/ANsv5V3Whf8AIvaf/wBe0f8A6CK8ir/u8fVns0v94l6I4P40f8gbT/8Arq38hXjNezfGj/kDWH/XVv5CvGa9jL/93XzPFzH/AHh/I9H8Lf8AJHvEf1/+Jqj8KP8AkaJ/+vc/zFXvC3/JHfEf+9/8TVL4T/8AI0T/APXuf5is5fBW9f0RpH+JR9P1Zxt//wAhK5/66t/M19OaN/yArH/r3j/9BFfMd/8A8hK5/wCurfzNfTmjf8gKx/694/8A0EVhmXwQN8r+OZ5h8af9dp//AAL+QrtNN/5Jla/9eI/9BrivjT/rtP8A+BfyFdrpv/JMrX/rxH/oNYT/AN2pep0U/wDeqvoN+HH/ACJNj/uCuW0L/ktms/73+FdT8OP+RJsf9wVy+hf8ls1n/e/wpR+Ot6P8xy+Cj6r8iz8X/wDjxsv+ug/rWB8XP9X4f/69D/7LW/8AF/8A48bL/roP61gfFz/V+H/+vQ/+y10YX/l18znxn/L3/t08/vv9ev8A1zX+Vey/B7/kV5P+urfzrxm9/wBev+4v8q9m+Dv/ACLEn/XVv510Y/8A3c5sv/3k9Dooor50+lCiiigAooooAKKKKAOH07/kqOq/9eh/9CFY/wAIv+PjV/8Ar4NbGnf8lR1X/r0P/oQrH+EX/Hxq/wD18GvUl/Bn6RPKj/Gh6yHax/yWG0/3P/iayPHv/I333/XS1rX1n/ksFn/uf/E1k+Pf+Rwvf+ulrW1H44f4f1MK3wT/AMX6FjWf+SkaR/17x/yrpfDH/JRde/65r/6Ea5vWP+SkaR/17x/yrpPDH/JRte/65r/6Eayq/wAL/t39TWl/F/7e/Q7euc8ff8iNqf8A1wNdHXOeP/8AkRtT/wCuJrz6P8WPqj0a/wDCl6MveGP+Rbs/+uYrVrK8Mf8AIt2f/XMVq1NT42XT+BBRRRUFhXm3ir/ksfhv/dH82r0mvOfFMUjfGDw46oxVVGWA4HLV1YX436P8jkxfwL1X5nTeOv8AkRtV/wCvdqz/AIX/APIh2f1b/wBCNaHjr/kRtW/692qh8L/+RDs/q3/oRpr/AHV/4v0E/wDe1/h/U6+iiiuQ7AqK5/49Jv8Acb+VS02RPMidOm5SKa3E9j5vsEHlh8DP2/Gce1fRtr/x5w/9c1/lXi3iDwq/hWS3he4WZbi7Eq4GNowRg17Ta/8AHnD/ANc1/lXpY6amoyjtqeXl8JU3KMt9CWiiivMPVCvMLD/kslz/AJ7mvT68xsP+SyXP+e5rrw20/Q48VvD1PTqKKK5DsCiiigArmfiF/wAiVd/7yf8AoQrpq5n4hf8AIlXf+8n/AKEK2ofxY+qMa/8ACl6M1tA/5F+y/wCuQrQrP0D/AJF+y/65CtCs5/Ey4fAgrj/hf/yJY/6/Ln/0a1dhXH/C/wD5Esf9flz/AOjWqSzsKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAqlrP/ACAr/wD69pP/AEE1dqlrP/ICv/8Ar2k/9BNAGN8N/wDkmug/9eafyrpq5n4b/wDJNdB/680/lXTUAFfN3jn/AJHG+/3h/KvpGvm7xx/yOF9/vD+Verln8R+h5Ga/w4+ozwl/yFLn/rym/wDQa2o/+SQP/wBfJ/8AQhWJ4T/5Clx/15zf+g1tx/8AJIH/AOvk/wDoQr0q3xr1X6nl0fgfpL9DH03/AJFXU/ov/oQrqfgv/wAjDf8A/XuP/QhXLab/AMirqX0X/wBCFdV8F/8AkYb/AP69x/6EKnE/wan9djTC/wAen/Xc3fjP/wAgOz/67f0NePWv/H1H/vCvYfjP/wAgO0/67f0NePWv/H1H/vClgP8Ad18wzD/eX8j2P4n/APJPrP8A7ZfyrutC/wCRe0//AK9o/wD0EVw3xP8A+SfWf/bL+VdzoX/Ivaf/ANe0f/oIryKv+7x9WezS/wB4l6I4P40f8gWw/wCurfyFeM17N8aP+QLYf9dW/kK8Zr2Mv/3dfM8XMf8AeH8j0fwt/wAkd8R/73/xNUvhP/yNE3/Xuf5ir3hb/kjniP8A3v8A4mqPwn/5Gmb/AK9z/MVnL+HW9f0RpH+JR9P1Zxt//wAhK5/66t/M19OaL/yArH/r3j/9BFfMd/8A8hK5/wCurfzNfTei/wDICsf+veP/ANBFY5l8EDfK/jmeYfGn/X6f/wAC/kK7XTf+SZWv/XiP/Qa4r40/6/T/APgX8hXa6d/yTK1/68R/6DXPP/dqXqdFP/eqvoN+HH/Ik2P+4K5fQv8Aktetf73+FdR8OP8AkSbH/cFcvoP/ACWvWv8Ae/wpR+Ot6P8AMcvgo+q/Is/F/wD48LL/AK6D+tYHxc/1fh//AK9D/wCy1v8Axf8A+QfZf9dR/WsD4uf6vw//ANeh/wDZa6ML/wAuvmc+M/5e/wDbp59e/wCvX/cX+Vey/B3/AJFmT/rq3868avf9cv8AuL/KvZfg7/yLcv8A11b+ddGP/wB3ObL/APeT0SiiivnT6UKKKKACiiigAooooA4fTv8AkqOq/wDXof8A0IVj/CL/AF+r/wDXwa2NO/5Kjqv/AF6H/wBCFZHwi/1+r/8AXwa9SX8GfpE8qP8AGh6yHaz/AMlgs/8Ac/8Aiax/Hv8AyON7/wBdLWtjWf8AksFn/uf/ABNY/j7/AJHK9/37WtqPxw/w/qYVvgn/AIv0Lesf8lH0j/r3j/lXR+GP+Sja9/1zX/0I1zmsf8lH0j/r3j/9Bro/DP8AyUbXv+ua/wDoRrKr/C/7d/U1pfxf+3v0O3rnPH//ACI2p/8AXE10dc54/wD+RG1P/ria8+j/ABY+qPRr/wAKXoy94Y/5Fuz/AOuYrVrK8M/8i3Z/9cxWrU1PjZdP4EFFFFQWFRtbxPMsrRqZF6MRyKkooAwPHP8AyI+rf9ezVn/C/wD5EOz+rf8AoRrQ8c/8iPq3/Xs1Z/wv/wCRCs/q3/oRrrX+6v8Axfocb/3tf4f1OvooorkOwKKKKAPNvip/x+aV/wBdB/WvRLX/AI84f+ua/wAq87+Kn/H5pX/XQf1r0S1/484f+ua/yrrq/wACn8zjo/x6nyJaKKK5DsCvMrD/AJLJc/57mvTa8ysP+Sx3P0/qa68NtP0OPFbw9T02iiiuQ7AooooAK5n4hf8AIlXf+8n/AKEK6auZ+IX/ACJV3/vJ/wChCtqH8WPqjGv/AApejNbQP+Rfsv8ArkK0Kz9A/wCRfsv+uQrQrOfxMuHwIK4/4X/8iWP+vy5/9GtXYVx/wv8A+RLH/X5c/wDo1qks7CiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKpaz/wAgK/8A+vaT/wBBNXapaz/yAr//AK9pP/QTQBjfDf8A5JroP/Xmn8q6auZ+G/8AyTXQf+vNP5V01ABXzd44/wCRwvv94fyr6Rr5u8cf8jhe/wC8P5V6uWfxH6HkZr/Dj6kfhP8A5Clx/wBec3/oNbcX/JH3/wCvk/8AoQrE8J/8hO4/685v/Qa3Iv8Akj7/APXyf/QhXpVvjXqv1PLo/A/R/oYum/8AIq6l9F/9CFdX8Fv+Rhv/APr2H/oQrlNN/wCRW1L6L/6EK6v4Lf8AIw3/AP17D/0IVOJ/g1P67GmF/j0/67m98Z/+QFaf9dv6GvHbX/j6j/3hXsXxn/5AVp/12H8jXj1r/wAfcf8AvClgP93XzDMP95fyPY/if/yT6z/7ZfyrudC/5F7T/wDr2j/9BFcN8T/+Se2f/bL+VdzoX/Ivaf8A9e0f/oIryKn+7x9WezS/3iXojhPjR/yBbD/rq38hXjFez/Gf/kC2H/XVv5CvGK9jL/8Ad18zxcx/3h/I9I8Lf8kc8R/73/xNUfhN/wAjTN/17n+Yq94X/wCSOeI/97/4mqPwm/5Gqb/r3P8AMVnL+HW9f0RrD+JR9P1Zxt//AMhK5/67P/M19N6L/wAgGx/694//AEEV8yX/APyErn/rs/8AM19N6J/yAbH/AK94/wD0EVjmXwQNsr+OZ5h8af8AX6f/AMC/kK7XTv8AkmNr/wBeI/8AQa4r40/6/T/+BfyFdrp3/JMbX/rxH/oNc8/92pep0U/96q+g34cf8iTY/wC4K5jQf+S161/vf4V0/wAOP+RJsf8AcFcxoP8AyWvWv97/AApR+Ot6P8xy+Cj6r8iz8X/+QfZf9dR/Wuf+Ln+r8P8A/Xof/Za6D4v/APIPsv8ArqP61z/xc/1fh/8A69D/AOy10YT/AJdfM58Z/wAvf+3Tz69/1y/7i/yr2T4O/wDIty/9dG/nXjd5/rl/3F/lXsfwc/5F2X/rof5104//AHc5cv8A95PRqKKK+cPpgooooAKKKKACiiigDh9O/wCSo6r/ANeh/wDQhWR8Iv8AXav/ANfBrX07/kqOq/8AXof/AEIVkfCL/Xav/wBfJr1JfwZ+kTyo/wAaHrIdrP8AyWCy/wBz/Csjx/8A8jle/wC/a1r61/yV+y/3P/iax/H/APyOd5/v2tbUfjh/h/UwrfBP/F+hc1j/AJKPpH/XvH/6DXReGf8Ako+vf9cl/wDQjXO6v/yUfSP+veP/ANBrovDP/JSNe/65L/6Eayqfwv8At39TWl/F/wC3v0O3rnPH/wDyI2p/9cTXR1zfj/8A5EbU/wDria8+j/Fj6o9Gv/Cl6Mv+Gf8AkW7P/rmK1ayvDP8AyLdn/wBcxWrU1PjZdP4EFFFFQWFFFFAGB45/5EbVv+vZqz/hf/yIVl9W/wDQjWj44/5EbVv+vZqz/hh/yINl9W/9CNda/wB1f+L9Djf+9r/D+p11FFFch2BRRRQB5t8VP+PzSv8ArqP616Ja/wDHnD/1zX+VeefFT/j80r/rqP616Ha/8ecP/XNf5V11f4FP5nHR/j1PkS0UUVyHYFeZ2H/JY7n6f1NemV5nY/8AJY7n6f1NdeG2n6HHid4ep6ZRRRXIdgUUUUAFcz8Qv+RKu/8AeT/0IV01cz8Qv+RKu/8AeT/0IVtQ/ix9UY1/4UvRmtoH/Iv2X/XIVoVn6B/yL9l/1yFaFZz+Jlw+BBXH/C//AJEsf9flz/6Nauwrj/hf/wAiWP8Ar8uf/RrVJZ2FFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFRXN1b2cDTXc8cES9XkYKB+JqWuH8eKL7X/D2l3HzWk14pljPSQYI2n1HtQB2Npf2moRebY3UNzH/AHoXDD8xTL3VLDTtv9oXtva7vu+dKEz+dcX4egi0T4k6hpenIIrOePzTCgwqFQBwO3Wmf2ZaeJviZrFvrMK3MFjFGkMcgyBvTJIB78UAd/HKk0ayQurowyrKcgiqms/8gK//AOvaT/0E1y3wruZZ/C91HM5cW+o3MMZPZFkIUfgBWh4w8U6domm3NtfC5Mk1u4Xyrd5Byp6kDigA+G//ACTXQf8ArzT+VdNXMfDY7vhnoBHQ2Sfyrp6ACvm7xx/yOF7/ALw/lX0dNIYoWcLuKjOB3r5z8UWmpX3iK6uf7NuUDtwBGW6e4Ferluk5NnkZprTil3IfCn/ITuP+vOb/ANBrci/5I/J/18n/ANCFZejaXqunTS3LabO6vbumNhB+ZceldJbaFqMnwmeFbWTzTcFghGDjcO1ejWlHmTv1X6nm0Yy5WrdH+hyum/8AIral9F/9CFdX8Fv+Rg1D/r2H/oQrmrLTdVj0m9s/7NnJkwN2wjGDn0rsPg9pd7aaxfz3NtJDGYQgLqVydw9ajEyXsZ6/1oXhYy9vT0/rU0/jP/yArT/rt/Q149af8fcX+8K9t+K2k3mq6Pax2MJlYTZIHbg15HbaBqi6pFAbGcPvAz5Zx+dLAyiqFr9ysfCTxF0ux6l8T/8Akntp/wBsv5V3Ohf8i9p//XtH/wCgiuT+IWlXmoeB7W2tITJKpiBUdsCuv0iJ4NEsopBh0t0Vh6EKK8io17CK82evTT+sSfkjgfjP/wAgWx/66t/IV4xXufxX0q81TRrQWMLSmORiwHbOK8ibwvrKzJEbCbc5wPkOK9fATiqCTfc8fMISddtLsdl4X/5I34j/AN7/AOJql8Jv+Rqm/wCvc/zFdB4d8P6lb/C3XbCW3ZbmZj5aHv0rO+Fuh6jaeIpZrm1kij8grlxjnIrOU48lXXr+iNYwl7Sjp0/Vnnt//wAhK5/67P8AzNfTeif8gGx/690/9BFfO2q+HtVh1KfdYzkNK2CsZPevWNP8b3tppttbnw/dMYolTOeuBj0pY6Lqwjyajy+SpTnz6GD8af8AX6f/AMC/kK7XTv8AkmNr/wBeI/8AQa84+IN/qHiU20iaPcwCHOflLZz9BXQ2Xim+j8IwaYNCuSy2wi38jPGM4xWM6cnQpx6pm0KsViKkujR0vw5/5Emx/wCuYrl9B/5LXrX1/wAK67wHZz2PhCyhukMcgjGVPauf0bRr6D4tatfSwFbaXlH7HpWEZLnq69H+ZvKL5KOnVfkRfF//AJB9l/11H9a5/wCLn+r8P/8AXof/AGWut+J2k3mqWNmllCZCJRnHbrWD8UtD1C8ttGktbd5RBb+W+0ZIJx/hXRhZRXs7vuc+LhJ+1sux5def65f9xf5V7J8HP+Rdm/66H+deYS+GdZnkBWwmGFA+6fSu38D6rqPhbTZLabRbiYsxOV4/pXXjLVKPLF6nHgr06/NJWR69RXE/8J/ef9C9df8AfX/1qP8AhP7z/oXrr/vr/wCtXifVqvb8Ue99Zpd/wZ21FcT/AMJ/ef8AQvXX/fX/ANaj/hP7z/oXrr/vr/61H1ar2/FB9Zpd/wAGdtRXE/8ACf3n/QvXX/fX/wBaj/hP7z/oXrr/AL6/+tR9Wq9vxQfWaXf8GdtRXE/8J/ef9C9df99f/Wo/4T+8/wCheuv++v8A61H1ar2/FB9Zpd/wY3Tv+So6r/16H/0IVkfCL/Xav/18mtDwu95qXjXUNTnsZLWKS2KKH9cim/DHSL3TJdV+2wmLfcErnuK65tKnOL3tE4oJyqwktryKmtf8lesv9z/4msb4gf8AI6Xn+/a10uraTey/FCzvI4SYFQ5fsOlYvjrStRufG0rW1o8kUzW5EgHA29a1oyjzx1+z+plXjLknp9r9B2r/APJR9H/690/9BrofDP8AyUjXv+uS/wDoRrO1TRL+Tx5pVykBMSQqrN6ELzSte33hzxzql4NLmu47pAqlDjGGJ9KiVpw5Y78v6mkbwnzS25v0PR65v4gf8iNqf/XE1m/8J/ef9C9df99f/WrL8R+Kb/W9AutPj0K5jadCoYnOP0rmpUKkakW117o6auIpypySfTsztfDP/It2f/XMVq1naBBJb6FaxTLtdYxkelaNc0/jZ1U/gQUUUVBYUUUUAYPjj/kRtX/69mrP+GH/ACINl9W/9CNavi62lvPB+p29uu+WS3ZVUdzVL4e2Vxp/gu0t7uMxyqWyp7cmupNfVmvP9Dkaf1pP+7+p01FFFcp1hRRRQB5v8VP+PvSv+uo/rXodr/x5w/8AXNf5VxHxG0m81K60w2cJkCyjdjt1ruLdStrErdQgB/Kuqq06MF6nJSTVeb9CSiiiuU6wrzSx/wCSxXP+7/U16XXA2ek3ifFS4vGhIgK5D9uprqw7SU79jkxCbcLdzvqKKK5TrCiiigArmfiF/wAiVd/7yf8AoQrpq57x1bS3fhG6ht0LyMyYUDr8wrWjpVj6oxr60pejNHQP+Rfsv+uQrQqjokbRaHaJINrLGAQe1XqifxMuHwoK4/4X/wDIlj/r8uf/AEa1dhXH/C//AJEsf9flz/6NapLOwooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigArnvFnhyfXYbSbT7hLW/sZhPbyyKSoYA/eA6jmuhooA47RPCutafcXWpX2o2lxqt1IpeVYmEaoBgqq5yM8Ua74U1mXW/7W8NalbWV1NF5dz58TOr8YBABHIrsaKAMbwt4fTw1ogsUfzHaRppXHRpHOWP51c1n/kBX/8A17Sf+gmrtUtZ/wCQFf8A/XtJ/wCgmgDG+G//ACTXQf8ArzT+VdNXmvgf4g+FtK+H+j22o6vHbzQWqJIjxv8AKfQ8V0X/AAsrwj/0Gov+/b//ABNAHUUVy/8Awsnwj/0Gov8Av2//AMTR/wALJ8I/9BqL/v2//wATQB1FFcv/AMLJ8I/9BqL/AL9v/wDE0f8ACyfCP/Qai/79v/8AE0AdRRXL/wDCyfCP/Qai/wC/b/8AxNH/AAsnwj/0Gov+/b//ABNAHUUVy/8Awsnwj/0Gov8Av2//AMTR/wALJ8I/9BqL/v2//wATQB1FFcv/AMLJ8I/9BqL/AL9v/wDE0f8ACyfCP/Qai/79v/8AE0AdRRXKt8TfByuqtrkIZvugxvz/AOO07/hZPhH/AKDUX/ft/wD4mgDqKK5f/hZPhH/oNRf9+3/+Jo/4WT4R/wCg1F/37f8A+JoA6iiuX/4WT4R/6DUX/ft//iaP+Fk+Ef8AoNRf9+3/APiaAOoorl/+Fk+Ef+g1F/37f/4mj/hZPhH/AKDUX/ft/wD4mgDqKK5f/hZPhH/oNRf9+3/+Jo/4WT4R/wCg1F/37f8A+JoA6iiuX/4WT4R/6DUX/ft//iaP+Fk+Ef8AoNRf9+3/APiaAOoorlY/id4OlXdHrkLDOMhH6/8AfNO/4WT4R/6DUX/ft/8A4mgDqKK5f/hZPhH/AKDUX/ft/wD4mj/hZPhH/oNRf9+3/wDiaAOoorl/+Fk+Ef8AoNRf9+3/APiaP+Fk+Ef+g1F/37f/AOJoA6iiuX/4WT4R/wCg1F/37f8A+Jo/4WT4R/6DUX/ft/8A4mgDqKK5f/hZPhH/AKDUX/ft/wD4mj/hZPhH/oNRf9+3/wDiaAOoorl/+Fk+Ef8AoNRf9+3/APiaa/xN8HJt365Cu44GY35P/fNAHVUVy/8Awsnwj/0Gov8Av2//AMTR/wALJ8I/9BqL/v2//wATQB1FFcv/AMLJ8I/9BqL/AL9v/wDE0f8ACyfCP/Qai/79v/8AE0AdRRXL/wDCyfCP/Qai/wC/b/8AxNH/AAsnwj/0Gov+/b//ABNAHUUVy/8Awsnwj/0Gov8Av2//AMTR/wALJ8I/9BqL/v2//wATQB1FFcv/AMLJ8I/9BqL/AL9v/wDE0f8ACyfCP/Qai/79v/8AE0AdRRXL/wDCyfCP/Qai/wC/b/8AxNNX4m+DnLBNchYqcMBG/B/75oA6qiuX/wCFk+Ef+g1F/wB+3/8AiaP+Fk+Ef+g1F/37f/4mgDqKK5f/AIWT4R/6DUX/AH7f/wCJo/4WT4R/6DUX/ft//iaAOoorl/8AhZPhH/oNRf8Aft//AImj/hZPhH/oNRf9+3/+JoA6iiuX/wCFk+Ef+g1F/wB+3/8AiaP+Fk+Ef+g1F/37f/4mgDqKK5f/AIWT4R/6DUX/AH7f/wCJo/4WT4R/6DUX/ft//iaAOoorlj8SvCABJ1uIAdT5b/8AxNInxM8HyIHj1yFlYZBEb8/+O0AdVRXL/wDCyfCP/Qai/wC/b/8AxNH/AAsnwj/0Gov+/b//ABNAHUUVy/8Awsnwj/0Gov8Av2//AMTR/wALJ8I/9BqL/v2//wATQB1FFcv/AMLJ8I/9BqL/AL9v/wDE0f8ACyfCP/Qai/79v/8AE0AdRRXL/wDCyfCP/Qai/wC/b/8AxNH/AAsnwj/0Gov+/b//ABNAHUVx/wAL/wDkSx/1+XP/AKNarH/CyfCP/Qai/wC/b/8AxNVfhXIs3gWOWMko91cMpIIyDI2DzQB2VFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUEZGDyKztetr670eWLS7xrK4OCsyqCVHfg8Vg/DXUNS1Dw5df2xetfXFvfzW/nuoUsqkAcDigC5400I6t4Su7Kwt4/Pk2bcIOzgn9BW3HZwLEgMEWQoB+QVwXxG12703X9Fs4/EJ0K1upSs04VTxtJ/i9xXQWNre3nhZk07xI99LMP3eobEyPcAcUAdB9kt/+eEX/AHwKPslv/wA8Iv8AvgVw+mXes6H4+ttCv9Zk1mK6tWnLSoqNEQwHRR0rvaAIvslv/wA8Iv8AvgUfZLf/AJ4Rf98CpaKAIvslv/zwi/74FH2S3/54Rf8AfAqWigCL7Jb/APPCL/vgUfZLf/nhF/3wKlooAi+yW/8Azwi/74FH2S3/AOeEX/fArk/iMutw6DLqGiavJp32SF5HEaK3mEDIHIrotBuJbvw7p1zcNvlmtY3dvVioJNAGTregtd+KvD95Bbx+RZyTGb5B0ZMD9a6D7Jb/APPCL/vgV5be6/dXHxJ1DTLvxo2i28MUbQQBIz5hLEEfMM9hXba/rR8P+G1ZZTd3ki7IMj5pm9QB+dAG59kt/wDnhF/3wKPslv8A88Iv++BXHfDLU9Y1HTNRTX7s3VzbXZi34AwNoOOPrXbUARfZLf8A54Rf98Cj7Jb/APPCL/vgVLRQBF9kt/8AnhF/3wKPslv/AM8Iv++BUtFAEX2S3/54Rf8AfAo+yW//ADwi/wC+BUtB6HHWgCL7Jb/88Iv++BR9kt/+eEX/AHwK4LSLnX9P+Ka6VqesyX1pcWUlwsLxqBGQwAAIGT1rofHWo3OleDb68srg200cZKygA7eDzzQBX8DeH30fQri31C3jEr31xMuUH3WkJX9DXSfZLf8A54Rf98CuK8BXhvreO7bxm2tkwK0tvsQCNiuT93niseDxNL4h1u4ZvFy6FHDdNbxWahGabY2MncM80Aem/ZLf/nhF/wB8Cj7Jb/8APCL/AL4FOhDLCodt7Act60+gCL7Jb/8APCL/AL4FH2S3/wCeEX/fAqWigCL7Jb/88Iv++BR9kt/+eEX/AHwKlooAi+yW/wDzwi/74FH2S3/54Rf98Cpa5/xjbarcaKzaNqcmmyRAu8kaKxIx0waANv7Jb/8APCL/AL4Fc/4p0FtSm0Y2lvGRa6jHPLhB9wZzT/AOo3ereAtJvtRmM91NDukkIA3HcRnj6VyPizX7qL4iRaXN4tbQLM2rOPkQ72BGPvCgD0v7Jb/88Iv++BR9kt/+eEX/AHwK4vxfq9/oXgyzls9UaXzbiOObUiq5jjbOXx04HNV/AevXeoeIL6zh1z/hIdOiRGW+2qMMRkr8vHXigDvPslv/AM8Iv++BR9kt/wDnhF/3wKlooAi+yW//ADwi/wC+BR9kt/8AnhF/3wKlooAi+yW//PCL/vgUfZLf/nhF/wB8CpaKAIvslv8A88Iv++BR9kt/+eEX/fAqWvPrqfX9K+J2kwz61LPp+oySj7GY1CoFXI5xmgDvPslv/wA8Iv8AvgVznhjQH07VvEM13bx7LzUDNDlB9zYB/MVtazqSaRo9zfS/dhTP49q4Twlq11rWoWt3c+MVM7nzJdHjCFUH93ON1AHon2S3/wCeEX/fAo+yW/8Azwi/74FebX+q+JNZudfv9H1V7K30d9kFsqKVuflB+YkZHPHFd34d1T+2dBtr3GGdcN/vDg/rQBe+yW//ADwi/wC+BR9kt/8AnhF/3wKlooAi+yW//PCL/vgUfZLf/nhF/wB8CpaKAIvslv8A88Iv++BR9kt/+eEX/fAqWigCL7Jb/wDPCL/vgUfZLf8A54Rf98CqHiG1v7zR5ItKvmsLjIInRQSo78HisT4ZanqOq+EXl1i7a8uYr2eAzMoBYI5A4H0oA6HUtPiuNKu4YrePfJC6r8g6lSBWd4O0X+yvBulWN7bxi4t7ZUkyg6isPxx4ne01e00a31VNIEgL3F62MxLjK4B4OSMVo6ba3994VePTvFD38sr5j1EImVHHAAGKAOl+yW//ADwi/wC+BR9kt/8AnhF/3wK4nRbzWNI8eHQNQ1aTWIpLTz/MlRVaM7sYwvau7oAi+yW//PCL/vgUfZLf/nhF/wB8CpaKAIvslv8A88Iv++BR9kt/+eEX/fAqWigCL7Jb/wDPCL/vgUfZLf8A54Rf98CpaKAIvslv/wA8Iv8AvgVIiLGu2NQq+ijApaKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAo6zqUelaVLdSw3E6qMbLeMyOc+gFcB8NfES2+nalbXWk6tA76hNcL5tkygo5GPx4r02igDlvEGu6VbLA+paJeXwkTdH5dj5xT6jtXN+HrnUPDPhjxHrMemXTxXFyZ7CxCMXCbQMbeq854r02igDzX4eXsM+pS3epWOq/2zefNLLc2bJHGOm1SeAOlelUUUAFFFFABRRRQAUUUUAcb8TNVFn4Tu7JLG+u5ruB0jFrbmTBxjnHSpfCPiWGbwzpkLafqUMkcEcLLNashDKoB69s9662igDz/AMV6tpUn22wbw7e3F/IhRJ0sCVycgHzO2DVa18A61deG9AaXXZLC/wBOtQrL5ImBfaQfvfWvSaKAPPPhbous6VNrZ1e9lmWS7O1ZLcR7zgfOP5V6HRRQAUUUUAFFFFABQTgZoooA8qk8Sq3xhtb0aPrH2aOzktWl+wtt3lgRz6cda7q91u0k0meefT7uaGMgPCbYlmB9F71t0UAeaaYia78QrC/0TSrrTLC3t5o7rzrcweYWXC4Xocc1hf2Wmiafquhz6Jd3OoXd/JPBcx25ZArvlf3nbFe0UUAZ3h+1ubLw9Y218++5ihVZGznLVo0UUAFFFFABRRRQAVg+MNXj0nQJjJa3lyZlZFS0gMrZx3AreooA89+G3iFLfwHp1jdaZqkE9qnlyLLaMvOSePUc1oeJdc0e2upIdQ8P3t9MyfLJFYGUY9N3auyooA8iOi6zY/DvSlvbWW5ihv8Az57VQXfyjIWAx1OFPStrwpF9v8dXGqaTpk+maZ5Ko0c0BhMjbcfc9jXodFABRRRQAUUUUAFFFFABXl3inxIo+JGhyppGryxafLKJpI7JmX5l2gg9xmvUaKAOd1cx+KPD9/psUNzEzxdZYigPPQV55YWT3UfhvSdP0a7s9Q064zd3LwFE24YcP/H1HWvZaKAPJp5tR8MN4k0oaZd3UmpOGsZIYS6H5APnYfd5zXoPhXTH0jw3a2kv3wu5h6E8kfrWxRQAUUUUAFFFFABRRRQBna7qcek6RLdSwXE6j5dltEZHOfYVwXwu18Wfh67tLvStVgl+3TzASWbKCruSuPwr06igDznxbp0UHi/TfE9zp1xfWhjKSwxwmRlG3j5O/JqrpF/qHhzw3r2tWmlXbR3Fy8lnZCJt4BIwdn8P09q9QooA85+HN3DJdTT31nqn9sXnz3E91aNGi/7Kk9uK9GoooAKKKKACiiqlzqun2cyw3l/bW8rfdSWZVY/QE0AW6KKZNPFbQtLcSpFGoyzuwUD8TQA+iq9pqFnfqzWN3BchTgmGQPj8qsUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAV5EfHGuR3Hi1dQ0W5lsbKRQXgu1RoV2Z4Pv14r12vJte0bxDbzeLbGx0Oe+i1ra0NxHIqqmI9uCCc9aAOhufH0tpf2Gk6ZotxqFzcWa3CL54U7dgPJI681Uh+Kj3Ni15b6BO9vDP9nun84DyX3bT2+bBz0q1pGg6jbeMtMvZrYrBDpywu+R8r+Wox+YrO0vwxq1v4E1ixls2W5uNSnmjTI+ZGkJB/KgDZ1f4hW1leJZ6bbLfXLQrOyNcrDtRuhy3Xoagf4mWv9i293b2Rmup7g2wthMoAkAyRv6dO9czdeB57LWbbWb3wyniBZdPhtZLQ7d0LJkk5Y45z+lbU+lxJ4Tjih8BEJJOXNhE0YaPj72c4z2oAj8XeMPENt4RW+t9Ea1uDcBPL+1I+VxndkcdeK3ZPFGrWmgW93caDtuHC7oZL2NMcddx4rj4/DHiS48H6qsltcrm482ysLiQM6JgDZkcdc1c8Q2eta2mk3d34ZuLm0t0CT6U8qEu2fvZzjp/OgDaj+JNs3hm+1V7JhLYOFntkmD7SQSPmHB4FVR8S7pLu0tbrw3cQzX8Xm2afaFbzhgHsPl6jrXOaR4N1iHRfE0K6ItgL24he2to9oG0IwPQ+9dZc6FqD+L/Cd2tsTBY2MsVw+R+7YooA/MUAMk+JUdv4bvdRu9KlhubKYRS2fmgsMttB3YxVjSvHsl5q1lZ6no02mrqA/wBDleUOJeM4wBxx61x3jLSNRsNN8Q3c1uUjuJ7fyGJGHIl//VWzp9pr3iTXPD0upaTJp9tozNKZnZSLgshX5QDx680Aek0UUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAV518W9GsJdDXU5LZGvEnhVJSASo8xen516LXFeP/C/iTxTbLZ6PqWn2lpuR2FxCzPuU56g9OBQBn+OPHt1pHiNNE025WzmFuLh5ns2uQQSRjavT61jXWvan4s8IxzXVvNLDp98FvtsDRfaotuSVQ84yRxz0rpJvCfif+0YtbttV0+PWvJEE7+S3kyIDnG3OevvWzHZeKl0SRX1KwbU2bKy+S3lDjptzmgDiPDGsWB+K9pZeHLN7PTptNkkmjMRjDyBlAO0gdBmvWK5bQPDGowa2+t+JLyC81LyzFE1uhVI0OMgA+4rqaACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigCtf6daapam21CBLiEsGKOOMg5B/OrCIsaBEG1VGAB2paKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooA//Z"},"026ae3c5-ee03-4532-b11e-191b37105d21.png":{"image/png":"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"}}},{"cell_type":"code","source":"# importing and installing tensorflow version 2.7.0 offline\n! ls ../input/tensor270/gresearch2/gresearch","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:32:47.424882Z","iopub.execute_input":"2022-03-16T19:32:47.425822Z","iopub.status.idle":"2022-03-16T19:32:48.268449Z","shell.execute_reply.started":"2022-03-16T19:32:47.425713Z","shell.execute_reply":"2022-03-16T19:32:48.2671Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# importing and installing tensorflow version 2.7.0 offline\n!mkdir -p /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/Keras_Preprocessing-1.1.2-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/Markdown-3.3.6-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/Werkzeug-2.0.2-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/absl_py-1.0.0-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/astunparse-1.6.3-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/cached_property-1.5.2-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/cachetools-4.2.4-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/certifi-2021.10.8-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/charset_normalizer-2.0.10-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/flatbuffers-2.0-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/gast-0.4.0-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/google_auth-2.3.3-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/google_auth_oauthlib-0.4.6-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/google_pasta-0.2.0-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/grpcio-1.43.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/h5py-3.6.0-cp37-cp37m-manylinux_2_12_x86_64.manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/idna-3.3-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/importlib_metadata-4.10.1-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/keras-2.7.0-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/libclang-12.0.0-2-py2.py3-none-manylinux1_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/numpy-1.21.5-cp37-cp37m-manylinux_2_12_x86_64.manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/oauthlib-3.1.1-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/opt_einsum-3.3.0-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/protobuf-3.19.3-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/pyasn1-0.4.8-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/pyasn1_modules-0.2.8-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/requests-2.27.1-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/requests_oauthlib-1.3.0-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/rsa-4.8-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/setuptools-60.5.0-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/six-1.16.0-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorboard-2.7.0-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorboard_data_server-0.6.1-py3-none-manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorboard_plugin_wit-1.8.1-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorflow-2.7.0-cp37-cp37m-manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorflow_estimator-2.7.0-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorboard_data_server-0.6.1-py3-none-manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorboard_plugin_wit-1.8.1-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorflow-2.7.0-cp37-cp37m-manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorflow_estimator-2.7.0-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/tensorflow_io_gcs_filesystem-0.23.1-cp37-cp37m-manylinux_2_12_x86_64.manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/termcolor-1.1.0 /tmp/pip/cache/termcolor-1.1.0.tar.gz\n!cp ../input/tensor270/gresearch2/gresearch/typing_extensions-4.0.1-py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/urllib3-1.26.8-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/wheel-0.37.1-py2.py3-none-any.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/wrapt-1.13.3-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl /tmp/pip/cache/\n!cp ../input/tensor270/gresearch2/gresearch/zipp-3.7.0-py3-none-any.whl /tmp/pip/cache/\n\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:32:58.69248Z","iopub.execute_input":"2022-03-16T19:32:58.692822Z","iopub.status.idle":"2022-03-16T19:33:49.687514Z","shell.execute_reply.started":"2022-03-16T19:32:58.692785Z","shell.execute_reply":"2022-03-16T19:33:49.686163Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# importing and installing tensorflow version 2.7.0 offline\n!pip install --no-index --find-links /tmp/pip/cache/ tensorflow","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:33:49.690381Z","iopub.execute_input":"2022-03-16T19:33:49.690661Z","iopub.status.idle":"2022-03-16T19:34:45.231631Z","shell.execute_reply.started":"2022-03-16T19:33:49.690626Z","shell.execute_reply":"2022-03-16T19:34:45.229935Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nimport random\nimport pandas as pd\nimport numpy as np\nimport lightgbm as lgb\nimport gresearch_crypto\nimport time\nimport datetime\nimport gc\nimport joblib\n\nfrom tensorflow import keras\nfrom keras import backend as K\nimport tensorflow as tf\nfrom sklearn.preprocessing import MinMaxScaler\n\nTRAIN_CSV = '/kaggle/input/g-research-crypto-forecasting/train.csv'\nASSET_DETAILS_CSV = '/kaggle/input/g-research-crypto-forecasting/asset_details.csv'\ntf.__version__","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:34:45.235968Z","iopub.execute_input":"2022-03-16T19:34:45.236411Z","iopub.status.idle":"2022-03-16T19:34:54.172471Z","shell.execute_reply.started":"2022-03-16T19:34:45.236357Z","shell.execute_reply":"2022-03-16T19:34:54.171431Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def m(x, w):\n    \"\"\"Weighted Mean\"\"\"\n    return np.sum(x * w) / np.sum(w)\n\ndef cov(x, y, w):\n    \"\"\"Weighted Covariance\"\"\"\n    return np.sum(w * (x - m(x, w)) * (y - m(y, w))) / np.sum(w)\n\ndef corr(x, y, w):\n    \"\"\"Weighted Correlation\"\"\"\n    return cov(x, y, w) / np.sqrt(cov(x, x, w) * cov(y, y, w))\n\n\ndef reduce_mem_usage(df):\n    \"\"\" iterate through all the columns of a dataframe and modify the data type\n        to reduce memory usage.        \n    \"\"\"\n    start_mem = df.memory_usage().sum() / 1024**2\n    print('Memory usage of dataframe is {:.2f} MB'.format(start_mem))\n    \n    for col in df.columns:\n        col_type = df[col].dtype\n        \n        if col_type != object:\n            c_min = df[col].min()\n            c_max = df[col].max()\n            if str(col_type)[:3] == 'int':\n                if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:\n                    df[col] = df[col].astype(np.int8)\n                elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:\n                    df[col] = df[col].astype(np.int16)\n                elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:\n                    df[col] = df[col].astype(np.int32)\n                elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:\n                    df[col] = df[col].astype(np.int64)  \n            else:\n                if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:\n                    df[col] = df[col].astype(np.float16)\n                elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:\n                    df[col] = df[col].astype(np.float32)\n                else:\n                    df[col] = df[col].astype(np.float64)\n#         else:\n#             df[col] = df[col].astype('category')\n\n    end_mem = df.memory_usage().sum() / 1024**2\n    print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))\n    print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))\n    \n    return df\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:34:54.176409Z","iopub.execute_input":"2022-03-16T19:34:54.1767Z","iopub.status.idle":"2022-03-16T19:34:54.196641Z","shell.execute_reply.started":"2022-03-16T19:34:54.176666Z","shell.execute_reply":"2022-03-16T19:34:54.195294Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class getFeatures():\n    \"\"\"\n    Class for creating and storing features.\n    \"\"\"\n    def __init__(self,windows=[]):\n        self.max_length=max(windows)\n        self.dataframes=[]\n    def compute(self,current):\n        \"\"\"\n        Function to append current close price to dataframes variable\n        and returns list of features.\n        \"\"\"\n        if len(self.dataframes)>2010:\n            self.dataframes = self.dataframes[-2010:]\n        self.dataframes.append(float(current['Close']))\n           \n        return30 = current['Close']/np.array(self.dataframes)[max([-1*len(self.dataframes),-30])]\n        return120 = current['Close']/np.array(self.dataframes)[max([-1*len(self.dataframes),-120])]\n        \n        volatility15 = np.std(np.array(self.dataframes)[max([-1*len(self.dataframes),-15]):])/np.std(np.array(self.dataframes)[max([-1*len(self.dataframes),-500]):])\n        volatility200 = np.std(np.array(self.dataframes)[max([-1*len(self.dataframes),-200]):])/np.std(np.array(self.dataframes)[max([-1*len(self.dataframes),-2000]):])\n        \n        #ema7 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-7]):]).ewm(span=7, min_periods=1, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        ema21 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-21]):]).ewm(span=21, min_periods=5, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        ema35 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-35]):]).ewm(span=35, min_periods=10, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        #ema55 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-55]):]).ewm(span=55, min_periods=15, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        ema80 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-80]):]).ewm(span=80, min_periods=20, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        #ema110 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-110]):]).ewm(span=110, min_periods=25, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        ema250 =  pd.Series(self.dataframes[max([-1*len(self.dataframes),-250]):]).ewm(span=250, min_periods=30, adjust=False).mean().iloc[-1]/current['Close'] # 15min window average\n        \n        if return30>2:\n            return30 = 2\n        if return30<0.5:\n            return30 = 0.5\n            \n        if return120>2:\n            return120 = 2\n        if return120<0.5:\n            return120 = 0.5\n            \n        if volatility15>5:\n            volatility15 = 5\n        if volatility15<0:\n            volatility15 = 0\n            \n        if volatility200>5:\n            volatility200 = 5\n        if volatility200<0:\n            volatility200 = 0\n            \n        if ema21>2:\n            ema21 = 2\n        if ema21<0.5:\n            ema21 = 0.5\n            \n        if ema35>2:\n            ema35 = 2\n        if ema35<0.5:\n            ema35 = 0.5\n            \n        if ema80>2:\n            ema80 = 2\n        if ema80<0.5:\n            ema80 = 0.5\n            \n        if ema250>2:\n            ema250 = 2\n        if ema250<0.5:\n            ema250 = 0.5\n        \n        \n        if str(return30) in ['nan','inf','-inf']:\n            return30 = 1\n        if str(return120) in ['nan','inf','-inf']:\n            return120 = 1\n        if str(volatility15) in ['nan','inf','-inf']:\n            volatility15 = 0.2\n        if str(volatility200) in ['nan','inf','-inf']:\n            volatility200 = 0.2\n        if str(ema21) in ['nan','inf','-inf']:\n            ema21 = 1\n        if str(ema35) in ['nan','inf','-inf']:\n            ema35 = 1\n        if str(ema80) in ['nan','inf','-inf']:\n            ema80 = 1\n        if str(ema250) in ['nan','inf','-inf']:\n            ema250 = 1\n        \n        if len(self.dataframes)<5:\n            return [1, 1, 0.2, 0.2, 1, 1, 1, 1]\n        else:\n            return [return30, return120, volatility15, volatility200, ema21, ema35, ema80, ema250]\n    ","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:34:54.200346Z","iopub.execute_input":"2022-03-16T19:34:54.201563Z","iopub.status.idle":"2022-03-16T19:34:54.232867Z","shell.execute_reply.started":"2022-03-16T19:34:54.20151Z","shell.execute_reply":"2022-03-16T19:34:54.231689Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"feats=[getFeatures([2010]) for _ in range(14)]","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:36:06.668904Z","iopub.execute_input":"2022-03-16T19:36:06.669568Z","iopub.status.idle":"2022-03-16T19:36:06.67485Z","shell.execute_reply.started":"2022-03-16T19:36:06.66953Z","shell.execute_reply":"2022-03-16T19:36:06.673824Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# lgb models trained on zerobeta target\n\nmodelzero0a = joblib.load('../input/gresmodels/ema14_other_feval_fullzero/lgb_model_val_beta0_0.pkl')\nmodelzero1a = joblib.load('../input/gresmodels/ema14_other_feval_fullzero/lgb_model_val_beta1_0.pkl')\nmodelzero2a = joblib.load('../input/gresmodels/ema14_other_feval_fullzero/lgb_model_val_beta2_0.pkl')\n\n# lgb models trained on fullbeta target\n\nmodelfull2b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta2_0.pkl')\nmodelfull3b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta3_0.pkl')\nmodelfull5b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta5_0.pkl')\nmodelfull8b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta8_0.pkl')\nmodelfull10b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta10_0.pkl')\nmodelfull11b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta11_0.pkl')\nmodelfull12b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta12_0.pkl')\nmodelfull13b = joblib.load('../input/gresmodels/ema014otherfevalbeta14fold/lgb_model_val_beta13_0.pkl')\n\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:36:07.229682Z","iopub.execute_input":"2022-03-16T19:36:07.230595Z","iopub.status.idle":"2022-03-16T19:36:07.710387Z","shell.execute_reply.started":"2022-03-16T19:36:07.23054Z","shell.execute_reply":"2022-03-16T19:36:07.708983Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# various minmax scalers for transforming test data to appropriate distribution\nscaler1 = joblib.load('../input/gresmodels/scalers/scaler1.save')\nscaler4 = joblib.load('../input/gresmodels/scalers/scaler4.save')\nscaler01 = joblib.load('../input/gresmodels/scalers/scaler01.save')\nscaler02 = joblib.load('../input/gresmodels/scalers/scaler02.save')\nscaler03 = joblib.load('../input/gresmodels/scalers/scaler03.save')\nscaler01best = joblib.load('../input/gresmodels/scalers/scaler01best.save')\nscaler02best = joblib.load('../input/gresmodels/scalers/scaler02best.save')\nscalerzero = joblib.load('../input/gresmodels/scalers/scalerzero.save')","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:36:15.143834Z","iopub.execute_input":"2022-03-16T19:36:15.144196Z","iopub.status.idle":"2022-03-16T19:36:15.180861Z","shell.execute_reply.started":"2022-03-16T19:36:15.144144Z","shell.execute_reply":"2022-03-16T19:36:15.179518Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# loading keras neural net models\nmodel1 = keras.models.load_model('../input/gresmodels/emadnnbetamodel0seed0/emadnnbetamodel0seed0', compile=False)\nmodel4 = keras.models.load_model('../input/gresmodels/emadnnfeval014extrafevalmodel4seed0lr1', compile=False)\nmodel01 = keras.models.load_model('../input/gresmodels/emadnnfeval0onlybetafold1seed0lr1', compile=False)\nmodel02 = keras.models.load_model('../input/gresmodels/emadnnfeval0onlybetafold2seed0lr1', compile=False)\nmodel03 = keras.models.load_model('../input/gresmodels/emadnnfeval0onlybetafold3seed0lr1', compile=False)\nmodel01best = keras.models.load_model('../input/gresmodels/emadnnfeval0onlybetafold1seed0lr1best', compile=False)\nmodel02best = keras.models.load_model('../input/gresmodels/emadnnfeval0onlybetafold2seed0lr1best', compile=False)\nmodeldnnzero = keras.models.load_model('../input/gresmodels/emadnnfeval0onlyzerofold0seed0', compile=False)\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:37:11.798764Z","iopub.execute_input":"2022-03-16T19:37:11.799139Z","iopub.status.idle":"2022-03-16T19:37:11.842415Z","shell.execute_reply.started":"2022-03-16T19:37:11.799045Z","shell.execute_reply":"2022-03-16T19:37:11.84108Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# defining mean and stdev variables for all models\n# we use these to normalise the predictions from the LGBM and DNN models so they can be ensembled\n\ndnnmean1 = 0.016834533105211127\ndnnstd1 = 5.367454277045551 \ndnnmean4 = -0.16693439553133801\ndnnstd4 = 6.459708512850307 \ndnnmean01 = 0.055314195869729785\ndnnstd01 = 6.723390661480986 \ndnnmean02 = 0.10034501108790424\ndnnstd02 = 7.438331336036615 \n\ndnnmean03 = 0.019139631916902544\ndnnstd03 = 5.675169602420376 \ndnnmean01best = 0.09889716583318675\ndnnstd01best = 6.784986628806578 \ndnnmean02best = 0.015836443924659535\ndnnstd02best = 7.490301098140777 \nlgbmmean1 = -6.812755955898616e-06\nlgbmstd1 = 0.0006158765594927516 \n\nlgbmmean2 = -1.0658401603886789e-05 \nlgbmstd2 = 0.0006463366961800999  \nlgbmmean3 = -1.2944577273296175e-05\nlgbmstd3 = 0.0005934272368661652 \nxgbmmean = 0.002491480874423726\nxgbmstd = 0.0010478100911640877  \nlgbmzeromean = 1.814303436247774e-05\nlgbmzerostd = 0.0010088461763049226 \ndnnzeromean = 0.02286166896039839\ndnnzerostd = 0.8958637238915896\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:37:26.433181Z","iopub.execute_input":"2022-03-16T19:37:26.433547Z","iopub.status.idle":"2022-03-16T19:37:26.450341Z","shell.execute_reply.started":"2022-03-16T19:37:26.433516Z","shell.execute_reply":"2022-03-16T19:37:26.449301Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# missingdict will store assets that have previously missed a timestamp\n# this is important because if an asset has been missing from a timestamp in the last 3750 iterations\n# then we know that the beta component of the target will == 0.\nmissingdict = {}\n\ntry:\n    # create a running start for missingdict relative to the supplemental train period\n    train = pd.read_csv('../input/g-research-crypto-forecasting/supplemental_train.csv')\n    train = train[train['timestamp']<1643760000]\n    train = train[train['timestamp']>1643260000]\n    timestamps = sorted(train['timestamp'].unique())\n    train = train[['timestamp','Asset_ID','Close','Target']]\n\n    \n    for z, timestamp in enumerate(timestamps):\n        df_test = train[train['timestamp']==timestamp] \n        ids = list(df_test['Asset_ID'].unique())\n\n        missing = [i for i in range(14) if i not in ids]\n        missing = list(set(missing))\n\n        if len(missing)>0:\n            for i in missing:\n                missingdict[i] = 0\n\n        keys = list(missingdict.keys())\n        for i in keys:\n            missingdict[i] = missingdict[i]+1\n            if missingdict[i] == 3750:\n                del missingdict[i]\n        \n        # calculate features for running start\n        for j, row in df_test.iterrows():\n            asset_id = int(row['Asset_ID'])\n            feat=feats[asset_id]\n            row_features1=feat.compute(row)\n\n        if z%100==0:\n            print(missingdict)\nexcept:\n    missingdict = {}\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:43:12.951253Z","iopub.execute_input":"2022-03-16T19:43:12.951588Z","iopub.status.idle":"2022-03-16T19:43:20.967457Z","shell.execute_reply.started":"2022-03-16T19:43:12.951555Z","shell.execute_reply":"2022-03-16T19:43:20.966388Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"start = time.time()\nenv = gresearch_crypto.make_env()\niter_test = env.iter_test()\nfor i, (df_test, df_pred) in enumerate(iter_test):\n    \n    \n    ids = list(df_test['Asset_ID'].unique())\n    \n    # create a list of missing asset_ids\n    missing = [i for i in range(14) if i not in ids]\n    missing = list(set(missing))\n    \n    # for each missing asset, add to the missing dictionary\n    if len(missing)>0:\n        for i in missing:\n            missingdict[i] = 0\n    \n    # we increment the value assigned to each asset per iteration\n    # we remove the asset from the missing dictionary after 3750 iterations\n    # this is because after valid 3750 iterations beta is being calculated again\n    keys = list(missingdict.keys())\n    for i in keys:\n        missingdict[i] = missingdict[i]+1\n        if missingdict[i] == 3750:\n            del missingdict[i]\n    \n    # compute features for each asset\n    dict_ = {0:[],1:[], 2:[],3:[], 4:[], 5:[],6:[], 7:[],8:[],9:[],10:[],11:[],12:[],13:[]}\n    arravg = []\n    for j , row in df_test.iterrows():\n        asset_id = int(row['Asset_ID'])\n        row_id = int(row['row_id'])\n        feat=feats[asset_id]\n        row_features1=feat.compute(row)\n        if arravg==[]:\n            arravg = np.array([row_features1])\n        else:\n            arravg = np.append(arravg, [row_features1], axis=0)\n        row_features1 = [row_id, asset_id] + row_features1\n        \n        dict_[asset_id] = row_features1 \n    \n    # calculate timestamp means for all features\n    means = list(arravg.mean(axis=0))\n    rows,fullbeta,zerobeta,rowidsbeta,rowidszero = [], [], [], [], []\n    for i in ids:\n        rowid = dict_[i][0]\n        rowsbeta = dict_[i][1:]\n        rowsbeta = rowsbeta + means\n        # add asset features to zerobeta list or fullbeta list depending on whether asset is in missingdict\n        if i in missingdict.keys():\n            zerobeta.append(rowsbeta)\n            rowidszero.append(rowid)\n        else:\n            fullbeta.append(rowsbeta)\n            rowidsbeta.append(rowid)\n    \n    # create normalised predictions for dnn & lgbm models\n    \n    dnnrowsbeta = scaler1.transform(np.array(fullbeta))\n    dnn_pred = model1.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred1 = (dnn_pred-dnnmean1)/dnnstd1\n    \n    dnnrowsbeta = scaler4.transform(np.array(fullbeta))\n    dnn_pred = model4.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred4 = (dnn_pred-dnnmean4)/dnnstd4\n    \n    dnnrowsbeta = scaler01.transform(np.array(fullbeta))\n    dnn_pred = model01.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred01 = (dnn_pred-dnnmean01)/dnnstd01\n    \n    dnnrowsbeta = scaler02.transform(np.array(fullbeta))\n    dnn_pred = model02.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred02 = (dnn_pred-dnnmean02)/dnnstd02\n    \n    dnnrowsbeta = scaler03.transform(np.array(fullbeta))\n    dnn_pred = model03.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred03 = (dnn_pred-dnnmean03)/dnnstd03\n    \n    dnnrowsbeta = scaler01best.transform(np.array(fullbeta))\n    dnn_pred = model01best.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred01best = (dnn_pred-dnnmean01best)/dnnstd01best\n    \n    dnnrowsbeta = scaler02best.transform(np.array(fullbeta))\n    dnn_pred = model02best.predict_on_batch(np.array(dnnrowsbeta)).reshape(1,-1)[0]\n    dnn_pred02best = (dnn_pred-dnnmean02best)/dnnstd02best\n    \n    dnn_pred = (dnn_pred1 + dnn_pred4 + dnn_pred01 + dnn_pred02 \n                + dnn_pred03 + dnn_pred01best + dnn_pred02best)/7\n    \n    lgbm_pred2b = modelfull2b.predict(np.array(fullbeta))\n    lgbm_pred3b = modelfull3b.predict(np.array(fullbeta))\n    lgbm_pred5b = modelfull5b.predict(np.array(fullbeta))\n    lgbm_pred8b = modelfull8b.predict(np.array(fullbeta))\n    lgbm_pred10b = modelfull10b.predict(np.array(fullbeta))\n    lgbm_pred11b = modelfull11b.predict(np.array(fullbeta))\n    lgbm_pred12b = modelfull12b.predict(np.array(fullbeta))\n    lgbm_pred13b = modelfull13b.predict(np.array(fullbeta))\n    \n    lgbm_predbeta = (lgbm_pred2b + lgbm_pred3b  + lgbm_pred5b  + lgbm_pred8b  \n              + lgbm_pred10b + lgbm_pred11b + lgbm_pred12b + lgbm_pred13b)/8\n\n    lgbm_pred = (lgbm_predbeta-lgbmmean1)/lgbmstd1\n\n    # ensemble lgbm and dnn predictions and write to df_pred\n    for ind, rowid in enumerate(rowidsbeta):\n        df_pred.loc[df_pred['row_id'] == int(rowid), 'Target'] = lgbm_pred[ind]*0.35 + dnn_pred[ind]*0.65\n    \n    # predicitions for assets in missingdict using models trained on zerobeta targets\n    if zerobeta == []:\n        pass\n    else:\n        dnnrowszero = scalerzero.transform(np.array(zerobeta))\n\n        dnn_predzero = modeldnnzero.predict_on_batch(np.array(dnnrowszero)).reshape(1,-1)[0]\n        dnn_predzero = (dnn_predzero-dnnzeromean)/dnnzerostd\n        \n        lgbm_pred0a = modelzero0a.predict(np.array(zerobeta))\n        lgbm_pred1a = modelzero1a.predict(np.array(zerobeta))\n        lgbm_pred2a = modelzero2a.predict(np.array(zerobeta))\n\n        lgbm_predzero = (lgbm_pred0a  + lgbm_pred1a  + lgbm_pred2a)/3\n        lgbm_predzero = (lgbm_predzero - lgbmzeromean) / lgbmzerostd\n        \n        for ind, rowid in enumerate(rowidszero):\n            df_pred.loc[df_pred['row_id'] == int(rowid), 'Target'] = lgbm_predzero[ind]*0.5 + dnn_predzero[ind]*0.5\n    \n    # Send submissions\n    env.predict(df_pred)\nstop = time.time()\nprint(stop-start)\n","metadata":{"execution":{"iopub.status.busy":"2022-03-16T19:44:24.047323Z","iopub.execute_input":"2022-03-16T19:44:24.048149Z","iopub.status.idle":"2022-03-16T19:44:24.2132Z","shell.execute_reply.started":"2022-03-16T19:44:24.048044Z","shell.execute_reply":"2022-03-16T19:44:24.207532Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}