From 17e899207330081a35023f8d67e8825fbf19b055 Mon Sep 17 00:00:00 2001 From: karllzy Date: Wed, 11 May 2022 11:00:55 +0800 Subject: [PATCH] First Commit --- .gitignore | 219 + README.md | 21 + model_evaluating.ipynb | 155 + model_training.ipynb | 7124 +++++++++++++++++++++++++++ models.py | 264 + preprocess.ipynb | 127 + preprocess/draw_pics_origin.m | 45 + preprocess/draw_pics_preprocessed.m | 48 + preprocess/pics/preprocessed.png | Bin 0 -> 91504 bytes preprocess/pics/raw.png | Bin 0 -> 179507 bytes preprocess/preprocess.m | 8 + preprocess/preprocess_mango.m | 15 + preprocess/train_test_split.m | 15 + utils.py | 153 + 14 files changed, 8194 insertions(+) create mode 100644 .gitignore create mode 100644 README.md create mode 100644 model_evaluating.ipynb create mode 100644 model_training.ipynb create mode 100644 models.py create mode 100644 preprocess.ipynb create mode 100755 preprocess/draw_pics_origin.m create mode 100755 preprocess/draw_pics_preprocessed.m create mode 100644 preprocess/pics/preprocessed.png create mode 100644 preprocess/pics/raw.png create mode 100755 preprocess/preprocess.m create mode 100755 preprocess/preprocess_mango.m create mode 100755 preprocess/train_test_split.m create mode 100755 utils.py diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..187847f --- /dev/null +++ b/.gitignore @@ -0,0 +1,219 @@ +preprocess/dataset/* +checkpoints/* +.idea +### JetBrains template +# Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio, WebStorm and Rider +# Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839 + +# User-specific stuff +.idea/**/workspace.xml +.idea/**/tasks.xml +.idea/**/usage.statistics.xml +.idea/**/dictionaries +.idea/**/shelf + +# Generated files +.idea/**/contentModel.xml + +# Sensitive or high-churn files +.idea/**/dataSources/ +.idea/**/dataSources.ids +.idea/**/dataSources.local.xml +.idea/**/sqlDataSources.xml +.idea/**/dynamic.xml +.idea/**/uiDesigner.xml +.idea/**/dbnavigator.xml + +# Gradle +.idea/**/gradle.xml +.idea/**/libraries + +# Gradle and Maven with auto-import +# When using Gradle or Maven with auto-import, you should exclude module files, +# since they will be recreated, and may cause churn. Uncomment if using +# auto-import. +# .idea/artifacts +# .idea/compiler.xml +# .idea/jarRepositories.xml +# .idea/modules.xml +# .idea/*.iml +# .idea/modules +# *.iml +# *.ipr + +# CMake +cmake-build-*/ + +# Mongo Explorer plugin +.idea/**/mongoSettings.xml + +# File-based project format +*.iws + +# IntelliJ +out/ + +# mpeltonen/sbt-idea plugin +.idea_modules/ + +# JIRA plugin +atlassian-ide-plugin.xml + +# Cursive Clojure plugin +.idea/replstate.xml + +# Crashlytics plugin (for Android Studio and IntelliJ) +com_crashlytics_export_strings.xml +crashlytics.properties +crashlytics-build.properties +fabric.properties + +# Editor-based Rest Client +.idea/httpRequests + +# Android studio 3.1+ serialized cache file +.idea/caches/build_file_checksums.ser + +### Python template +# Byte-compiled / optimized / DLL files +__pycache__/ +*.py[cod] +*$py.class + +# C extensions +*.so + +# Distribution / packaging +.Python +build/ +develop-eggs/ +dist/ +downloads/ +eggs/ +.eggs/ +lib/ +lib64/ +parts/ +sdist/ +var/ +wheels/ +share/python-wheels/ +*.egg-info/ +.installed.cfg +*.egg +MANIFEST + +# PyInstaller +# Usually these files are written by a python script from a template +# before PyInstaller builds the exe, so as to inject date/other infos into it. +*.manifest +*.spec + +# Installer logs +pip-log.txt +pip-delete-this-directory.txt + +# Unit test / coverage reports +htmlcov/ +.tox/ +.nox/ +.coverage +.coverage.* +.cache +nosetests.xml +coverage.xml +*.cover +*.py,cover +.hypothesis/ +.pytest_cache/ +cover/ + +# Translations +*.mo +*.pot + +# Django stuff: +*.log +local_settings.py +db.sqlite3 +db.sqlite3-journal + +# Flask stuff: +instance/ +.webassets-cache + +# Scrapy stuff: +.scrapy + +# Sphinx documentation +docs/_build/ + +# PyBuilder +.pybuilder/ +target/ + +# Jupyter Notebook +.ipynb_checkpoints + +# IPython +profile_default/ +ipython_config.py + +# pyenv +# For a library or package, you might want to ignore these files since the code is +# intended to run in multiple environments; otherwise, check them in: +# .python-version + +# pipenv +# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. +# However, in case of collaboration, if having platform-specific dependencies or dependencies +# having no cross-platform support, pipenv may install dependencies that don't work, or not +# install all needed dependencies. +#Pipfile.lock + +# PEP 582; used by e.g. github.com/David-OConnor/pyflow +__pypackages__/ + +# Celery stuff +celerybeat-schedule +celerybeat.pid + +# SageMath parsed files +*.sage.py + +# Environments +.env +.venv +env/ +venv/ +ENV/ +env.bak/ +venv.bak/ + +# Spyder project settings +.spyderproject +.spyproject + +# Rope project settings +.ropeproject + +# mkdocs documentation +/site + +# mypy +.mypy_cache/ +.dmypy.json +dmypy.json + +# Pyre type checker +.pyre/ + +# pytype static type analyzer +.pytype/ + +# Cython debug symbols +cython_debug/ + +!/checkpoints/ +!/preprocess/dataset/ +!/preprocess/dataset/ diff --git a/README.md b/README.md new file mode 100644 index 0000000..0157bab --- /dev/null +++ b/README.md @@ -0,0 +1,21 @@ +# SCNet: A deep learning network framework for analyzing near-infrared spectroscopy using short-cut +## Pre-processing + +Since the method we proposed is a regression model, the classification dataset weat kernel is not used in this work. + +The other three dataset (corn, marzipan, soil) were preprocessed manually with Matlab and saved in the sub dictionary of `./preprocess` dir. The original dataset of these three dataset were stored in the `./preprocess/dataset/`. + +The mango dataset is not in Matlab .m file format, so we save them with the `process.py`. +Meanwhile, we drop the useless part and only save the data between  684 and 900 nm. + +> The data set used in this study comprises a total of 11,691 NIR spectra (684–990 nm in 3 nm sampling with a total 103 variables) and DM measurements performed on 4675 mango fruit across 4 harvest seasons 2015, 2016, 2017 and 2018 [24]. + +The detailed preprocessing progress can be found in [./preprocess.ipynb](./preprocess.ipynb) + +## Network Training + +In order to show our network can prevent degration problem, we hold the experiment which contains the training loss curve of four models. The detailed information can be found in [model_training.ipynb](./model_training.ipynb). + +## Network evaluation +After training our model on training set, we evaluate the models on testing dataset that spared before. The evaluation is done with [model_evaluation.ipynb](model_evaluating.ipynb). + diff --git a/model_evaluating.ipynb b/model_evaluating.ipynb new file mode 100644 index 0000000..2c5fd23 --- /dev/null +++ b/model_evaluating.ipynb @@ -0,0 +1,155 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "collapsed": true, + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "# Experiment 2: Model Evaluating" + ] + }, + { + "cell_type": "code", + "execution_count": 29, + "outputs": [], + "source": [ + "import numpy as np\n", + "from keras.models import load_model\n", + "from matplotlib import ticker\n", + "from scipy.io import loadmat\n", + "from sklearn.model_selection import train_test_split\n", + "from sklearn.metrics import mean_squared_error\n", + "import matplotlib.pyplot as plt\n", + "%matplotlib inline" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "markdown", + "source": [ + "In this experiment, we load model weights from the experiment1 and evaluate them on test dataset." + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + }, + { + "cell_type": "markdown", + "source": [], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 30, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "shape of data:\n", + "x_train: (5728, 1, 102), y_train: (5728, 1),\n", + "x_val: (2455, 1, 102), y_val: (2455, 1)\n", + "x_test: (3508, 1, 102), y_test: (3508, 1)\n" + ] + } + ], + "source": [ + "data = loadmat('./preprocess/dataset/mango/mango_dm_split.mat')\n", + "x_train, y_train, x_test, y_test = data['x_train'], data['y_train'], data['x_test'], data['y_test']\n", + "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=12, shuffle=True)\n", + "x_train, x_val, x_test = x_train[:, np.newaxis, :], x_val[:, np.newaxis, :], x_test[:, np.newaxis, :]\n", + "print(f\"shape of data:\\n\"\n", + " f\"x_train: {x_train.shape}, y_train: {y_train.shape},\\n\"\n", + " f\"x_val: {x_val.shape}, y_val: {y_val.shape}\\n\"\n", + " f\"x_test: {x_test.shape}, y_test: {y_test.shape}\")" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "source": [ + "## Build model and load weights\n", + "plain_5, plain_11 = load_model('./checkpoints/plain5.hdf5'), load_model('./checkpoints/plain11.hdf5')\n", + "shortcut5, shortcut11 = load_model('./checkpoints/shortcut5.hdf5'), load_model('./checkpoints/shortcut11.hdf5')\n", + "models = {'plain 5': plain_5, 'plain 11': plain_11, 'shortcut 5': shortcut5, 'shortcut11': shortcut11}\n", + "results = {model_name: model.predict(x_test).reshape((-1, )) for model_name, model in models.items()}\n", + "for model_name, model_result in results.items():\n", + " print(model_name, \" : \", mean_squared_error(y_test, model_result)*100, \"%\")" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + }, + "execution_count": 31, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "plain 5 : 0.2707851525589865 %\n", + "plain 11 : 0.26240810192725905 %\n", + "shortcut 5 : 0.28330442301217196 %\n", + "shortcut11 : 0.25743312483685266 %\n" + ] + } + ] + }, + { + "cell_type": "code", + "execution_count": 31, + "outputs": [], + "source": [], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 2 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython2", + "version": "2.7.6" + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} \ No newline at end of file diff --git a/model_training.ipynb b/model_training.ipynb new file mode 100644 index 0000000..1e0aecd --- /dev/null +++ b/model_training.ipynb @@ -0,0 +1,7124 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "# Model training Experiment" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "In order to provide the experimental evidence to back up the statement in the section 2.3.2 that our shortcut block can prevent degration problem.\n", + "\n", + "Plain Networks. We first evaluate 5-layer and 11-layer plain nets. Comparing to the 5-layer one, the 11-layer one has three times convolutional layers." + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "import pickle\n", + "import time\n", + "\n", + "import numpy as np\n", + "from matplotlib import ticker\n", + "from scipy.io import loadmat\n", + "from models import Plain5, Plain11, ShortCut5, ShortCut11\n", + "from sklearn.model_selection import train_test_split\n", + "import matplotlib.pyplot as plt\n", + "%matplotlib notebook" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "## Load data\n", + "load data and split them into train, val, test" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "import random\n", + "from numpy.random import seed\n", + "import tensorflow\n", + "import time\n", + "seed(4750)\n", + "tensorflow.random.set_seed(4750)\n", + "time1 = time.time()\n", + "data = loadmat('./preprocess/dataset/mango/mango_dm_split.mat')\n", + "x_train, y_train, x_test, y_test = data['x_train'], data['y_train'], data['x_test'], data['y_test']\n", + "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=12, shuffle=True)\n", + "x_train, x_val, x_test = x_train[:, np.newaxis, :], x_val[:, np.newaxis, :], x_test[:, np.newaxis, :]\n", + "print(f\"shape of data:\\n\"\n", + " f\"x_train: {x_train.shape}, y_train: {y_train.shape},\\n\"\n", + " f\"x_val: {x_val.shape}, y_val: {y_val.shape}\\n\"\n", + " f\"x_test: {x_test.shape}, y_test: {y_test.shape}\")" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "## Build Plain networks and Training" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "plain_5, plain_11 = Plain5(input_shape=(1, 102)), Plain11(input_shape=(1, 102))" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "2022-05-10 17:09:33.906952: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch 1/1024\n", + "90/90 [==============================] - 1s 4ms/step - loss: 0.0236 - val_loss: 0.0272 - lr: 0.0025\n", + "Epoch 2/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0188 - val_loss: 0.0297 - lr: 0.0025\n", + "Epoch 3/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0127 - val_loss: 0.0301 - lr: 0.0025\n", + "Epoch 4/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0088 - val_loss: 0.0315 - lr: 0.0025\n", + "Epoch 5/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0081 - val_loss: 0.0813 - lr: 0.0025\n", + "Epoch 6/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0072 - val_loss: 0.0309 - lr: 0.0025\n", + "Epoch 7/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0069 - val_loss: 0.0930 - lr: 0.0025\n", + "Epoch 8/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0066 - val_loss: 0.0313 - lr: 0.0025\n", + "Epoch 9/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0058 - val_loss: 0.0597 - lr: 0.0025\n", + "Epoch 10/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0050 - val_loss: 0.0376 - lr: 0.0025\n", + "Epoch 11/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0048 - val_loss: 0.0731 - lr: 0.0025\n", + "Epoch 12/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0047 - val_loss: 0.3068 - lr: 0.0025\n", + "Epoch 13/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0046 - val_loss: 0.0852 - lr: 0.0025\n", + "Epoch 14/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0048 - val_loss: 0.2993 - lr: 0.0025\n", + "Epoch 15/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0046 - val_loss: 0.3298 - lr: 0.0025\n", + "Epoch 16/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0046 - val_loss: 0.0272 - lr: 0.0025\n", + "Epoch 17/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0045 - val_loss: 0.0434 - lr: 0.0025\n", + "Epoch 18/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0044 - val_loss: 0.0611 - lr: 0.0025\n", + "Epoch 19/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0043 - val_loss: 0.3292 - lr: 0.0025\n", + "Epoch 20/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0042 - val_loss: 0.0732 - lr: 0.0025\n", + "Epoch 21/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0041 - val_loss: 0.0785 - lr: 0.0025\n", + "Epoch 22/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0043 - val_loss: 0.0719 - lr: 0.0025\n", + "Epoch 23/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0042 - val_loss: 0.2986 - lr: 0.0025\n", + "Epoch 24/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0040 - val_loss: 0.3209 - lr: 0.0025\n", + "Epoch 25/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0041 - val_loss: 0.1807 - lr: 0.0025\n", + "Epoch 26/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0041 - val_loss: 0.0513 - lr: 0.0025\n", + "Epoch 27/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.0834 - lr: 0.0012\n", + "Epoch 28/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.1304 - lr: 0.0012\n", + "Epoch 29/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0039 - val_loss: 0.2374 - lr: 0.0012\n", + "Epoch 30/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.1049 - lr: 0.0012\n", + "Epoch 31/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0041 - val_loss: 0.0522 - lr: 0.0012\n", + "Epoch 32/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0039 - val_loss: 0.2504 - lr: 0.0012\n", + "Epoch 33/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0040 - val_loss: 0.1557 - lr: 0.0012\n", + "Epoch 34/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0037 - val_loss: 0.0697 - lr: 0.0012\n", + "Epoch 35/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0039 - val_loss: 0.0616 - lr: 0.0012\n", + "Epoch 36/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0040 - val_loss: 0.3263 - lr: 0.0012\n", + "Epoch 37/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.1540 - lr: 0.0012\n", + "Epoch 38/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0037 - val_loss: 0.0840 - lr: 0.0012\n", + "Epoch 39/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0039 - val_loss: 0.3193 - lr: 0.0012\n", + "Epoch 40/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0037 - val_loss: 0.1026 - lr: 0.0012\n", + "Epoch 41/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0041 - val_loss: 0.2626 - lr: 0.0012\n", + "Epoch 42/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.1396 - lr: 0.0012\n", + "Epoch 43/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.0453 - lr: 0.0012\n", + "Epoch 44/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.0897 - lr: 0.0012\n", + "Epoch 45/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0038 - val_loss: 0.3225 - lr: 0.0012\n", + "Epoch 46/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.3101 - lr: 0.0012\n", + "Epoch 47/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.3089 - lr: 0.0012\n", + "Epoch 48/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.0290 - lr: 0.0012\n", + "Epoch 49/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.0591 - lr: 0.0012\n", + "Epoch 50/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0036 - val_loss: 0.1132 - lr: 0.0012\n", + "Epoch 51/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.3202 - lr: 0.0012\n", + "Epoch 52/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.0190 - lr: 6.2500e-04\n", + "Epoch 53/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0036 - val_loss: 0.0558 - lr: 6.2500e-04\n", + "Epoch 54/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.1351 - lr: 6.2500e-04\n", + "Epoch 55/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.1235 - lr: 6.2500e-04\n", + "Epoch 56/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.2803 - lr: 6.2500e-04\n", + "Epoch 57/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.0929 - lr: 6.2500e-04\n", + "Epoch 58/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0037 - val_loss: 0.0510 - lr: 6.2500e-04\n", + "Epoch 59/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.0590 - lr: 6.2500e-04\n", + "Epoch 60/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0036 - val_loss: 0.3023 - lr: 6.2500e-04\n", + "Epoch 61/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.1735 - lr: 6.2500e-04\n", + "Epoch 62/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.1487 - lr: 6.2500e-04\n", + "Epoch 63/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.1859 - lr: 6.2500e-04\n", + "Epoch 64/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.3022 - lr: 6.2500e-04\n", + "Epoch 65/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.2015 - lr: 6.2500e-04\n", + "Epoch 66/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.3198 - lr: 6.2500e-04\n", + "Epoch 67/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.2997 - lr: 6.2500e-04\n", + "Epoch 68/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.1298 - lr: 6.2500e-04\n", + "Epoch 69/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.1149 - lr: 6.2500e-04\n", + "Epoch 70/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.1068 - lr: 6.2500e-04\n", + "Epoch 71/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0951 - lr: 6.2500e-04\n", + "Epoch 72/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0429 - lr: 6.2500e-04\n", + "Epoch 73/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.2853 - lr: 6.2500e-04\n", + "Epoch 74/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0702 - lr: 6.2500e-04\n", + "Epoch 75/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2109 - lr: 6.2500e-04\n", + "Epoch 76/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.1927 - lr: 6.2500e-04\n", + "Epoch 77/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.0741 - lr: 6.2500e-04\n", + "Epoch 78/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0521 - lr: 3.1250e-04\n", + "Epoch 79/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.1454 - lr: 3.1250e-04\n", + "Epoch 80/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0716 - lr: 3.1250e-04\n", + "Epoch 81/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0251 - lr: 3.1250e-04\n", + "Epoch 82/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1286 - lr: 3.1250e-04\n", + "Epoch 83/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0273 - lr: 3.1250e-04\n", + "Epoch 84/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0988 - lr: 3.1250e-04\n", + "Epoch 85/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0657 - lr: 3.1250e-04\n", + "Epoch 86/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.0328 - lr: 3.1250e-04\n", + "Epoch 87/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0158 - lr: 3.1250e-04\n", + "Epoch 88/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0611 - lr: 3.1250e-04\n", + "Epoch 89/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0537 - lr: 3.1250e-04\n", + "Epoch 90/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.1233 - lr: 3.1250e-04\n", + "Epoch 91/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1027 - lr: 3.1250e-04\n", + "Epoch 92/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0456 - lr: 3.1250e-04\n", + "Epoch 93/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0112 - lr: 3.1250e-04\n", + "Epoch 94/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0546 - lr: 3.1250e-04\n", + "Epoch 95/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0120 - lr: 3.1250e-04\n", + "Epoch 96/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0978 - lr: 3.1250e-04\n", + "Epoch 97/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0384 - lr: 3.1250e-04\n", + "Epoch 98/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0324 - lr: 3.1250e-04\n", + "Epoch 99/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.1121 - lr: 3.1250e-04\n", + "Epoch 100/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0442 - lr: 3.1250e-04\n", + "Epoch 101/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1185 - lr: 3.1250e-04\n", + "Epoch 102/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0731 - lr: 3.1250e-04\n", + "Epoch 103/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1457 - lr: 3.1250e-04\n", + "Epoch 104/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.2414 - lr: 3.1250e-04\n", + "Epoch 105/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0337 - lr: 3.1250e-04\n", + "Epoch 106/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.1618 - lr: 3.1250e-04\n", + "Epoch 107/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0374 - lr: 3.1250e-04\n", + "Epoch 108/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0780 - lr: 3.1250e-04\n", + "Epoch 109/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0158 - lr: 3.1250e-04\n", + "Epoch 110/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.2529 - lr: 3.1250e-04\n", + "Epoch 111/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.2905 - lr: 3.1250e-04\n", + "Epoch 112/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.1713 - lr: 3.1250e-04\n", + "Epoch 113/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.1680 - lr: 3.1250e-04\n", + "Epoch 114/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0264 - lr: 3.1250e-04\n", + "Epoch 115/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1308 - lr: 3.1250e-04\n", + "Epoch 116/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0473 - lr: 3.1250e-04\n", + "Epoch 117/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.1393 - lr: 3.1250e-04\n", + "Epoch 118/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0336 - lr: 3.1250e-04\n", + "Epoch 119/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0409 - lr: 1.5625e-04\n", + "Epoch 120/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0932 - lr: 1.5625e-04\n", + "Epoch 121/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0227 - lr: 1.5625e-04\n", + "Epoch 122/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0491 - lr: 1.5625e-04\n", + "Epoch 123/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0734 - lr: 1.5625e-04\n", + "Epoch 124/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.1225 - lr: 1.5625e-04\n", + "Epoch 125/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0082 - lr: 1.5625e-04\n", + "Epoch 126/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.1061 - lr: 1.5625e-04\n", + "Epoch 127/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0511 - lr: 1.5625e-04\n", + "Epoch 128/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0280 - lr: 1.5625e-04\n", + "Epoch 129/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0899 - lr: 1.5625e-04\n", + "Epoch 130/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0167 - lr: 1.5625e-04\n", + "Epoch 131/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0133 - lr: 1.5625e-04\n", + "Epoch 132/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0519 - lr: 1.5625e-04\n", + "Epoch 133/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1900 - lr: 1.5625e-04\n", + "Epoch 134/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.0392 - lr: 1.5625e-04\n", + "Epoch 135/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0222 - lr: 1.5625e-04\n", + "Epoch 136/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0120 - lr: 1.5625e-04\n", + "Epoch 137/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0755 - lr: 1.5625e-04\n", + "Epoch 138/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0820 - lr: 1.5625e-04\n", + "Epoch 139/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0921 - lr: 1.5625e-04\n", + "Epoch 140/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0145 - lr: 1.5625e-04\n", + "Epoch 141/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0545 - lr: 1.5625e-04\n", + "Epoch 142/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.1179 - lr: 1.5625e-04\n", + "Epoch 143/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.1231 - lr: 1.5625e-04\n", + "Epoch 144/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0159 - lr: 1.5625e-04\n", + "Epoch 145/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0107 - lr: 1.5625e-04\n", + "Epoch 146/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0575 - lr: 1.5625e-04\n", + "Epoch 147/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.2076 - lr: 1.5625e-04\n", + "Epoch 148/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0351 - lr: 1.5625e-04\n", + "Epoch 149/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.1428 - lr: 1.5625e-04\n", + "Epoch 150/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.1004 - lr: 1.5625e-04\n", + "Epoch 151/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0082 - lr: 7.8125e-05\n", + "Epoch 152/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0117 - lr: 7.8125e-05\n", + "Epoch 153/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0295 - lr: 7.8125e-05\n", + "Epoch 154/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0128 - lr: 7.8125e-05\n", + "Epoch 155/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0414 - lr: 7.8125e-05\n", + "Epoch 156/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0187 - lr: 7.8125e-05\n", + "Epoch 157/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0387 - lr: 7.8125e-05\n", + "Epoch 158/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0032 - val_loss: 0.0540 - lr: 7.8125e-05\n", + "Epoch 159/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0196 - lr: 7.8125e-05\n", + "Epoch 160/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0126 - lr: 7.8125e-05\n", + "Epoch 161/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0174 - lr: 7.8125e-05\n", + "Epoch 162/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0150 - lr: 7.8125e-05\n", + "Epoch 163/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0521 - lr: 7.8125e-05\n", + "Epoch 164/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0038 - lr: 7.8125e-05\n", + "Epoch 165/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0056 - lr: 7.8125e-05\n", + "Epoch 166/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0418 - lr: 7.8125e-05\n", + "Epoch 167/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0794 - lr: 7.8125e-05\n", + "Epoch 168/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0140 - lr: 7.8125e-05\n", + "Epoch 169/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0390 - lr: 7.8125e-05\n", + "Epoch 170/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0498 - lr: 7.8125e-05\n", + "Epoch 171/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0114 - lr: 7.8125e-05\n", + "Epoch 172/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0130 - lr: 7.8125e-05\n", + "Epoch 173/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0497 - lr: 7.8125e-05\n", + "Epoch 174/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0415 - lr: 7.8125e-05\n", + "Epoch 175/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0331 - lr: 7.8125e-05\n", + "Epoch 176/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0074 - lr: 7.8125e-05\n", + "Epoch 177/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0507 - lr: 7.8125e-05\n", + "Epoch 178/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0291 - lr: 7.8125e-05\n", + "Epoch 179/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0252 - lr: 7.8125e-05\n", + "Epoch 180/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0259 - lr: 7.8125e-05\n", + "Epoch 181/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0340 - lr: 7.8125e-05\n", + "Epoch 182/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0094 - lr: 7.8125e-05\n", + "Epoch 183/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0213 - lr: 7.8125e-05\n", + "Epoch 184/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0055 - lr: 7.8125e-05\n", + "Epoch 185/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0140 - lr: 7.8125e-05\n", + "Epoch 186/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0448 - lr: 7.8125e-05\n", + "Epoch 187/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0061 - lr: 7.8125e-05\n", + "Epoch 188/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0218 - lr: 7.8125e-05\n", + "Epoch 189/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0128 - lr: 7.8125e-05\n", + "Epoch 190/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0624 - lr: 3.9062e-05\n", + "Epoch 191/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0200 - lr: 3.9062e-05\n", + "Epoch 192/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0434 - lr: 3.9062e-05\n", + "Epoch 193/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0101 - lr: 3.9062e-05\n", + "Epoch 194/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0087 - lr: 3.9062e-05\n", + "Epoch 195/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0212 - lr: 3.9062e-05\n", + "Epoch 196/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0185 - lr: 3.9062e-05\n", + "Epoch 197/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0654 - lr: 3.9062e-05\n", + "Epoch 198/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0122 - lr: 3.9062e-05\n", + "Epoch 199/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0077 - lr: 3.9062e-05\n", + "Epoch 200/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0086 - lr: 3.9062e-05\n", + "Epoch 201/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0058 - lr: 3.9062e-05\n", + "Epoch 202/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0134 - lr: 3.9062e-05\n", + "Epoch 203/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0553 - lr: 3.9062e-05\n", + "Epoch 204/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0033 - val_loss: 0.0058 - lr: 3.9062e-05\n", + "Epoch 205/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0239 - lr: 3.9062e-05\n", + "Epoch 206/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0095 - lr: 3.9062e-05\n", + "Epoch 207/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0299 - lr: 3.9062e-05\n", + "Epoch 208/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0111 - lr: 3.9062e-05\n", + "Epoch 209/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0288 - lr: 3.9062e-05\n", + "Epoch 210/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0236 - lr: 3.9062e-05\n", + "Epoch 211/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0035 - lr: 3.9062e-05\n", + "Epoch 212/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0315 - lr: 3.9062e-05\n", + "Epoch 213/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0076 - lr: 3.9062e-05\n", + "Epoch 214/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0043 - lr: 3.9062e-05\n", + "Epoch 215/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0141 - lr: 3.9062e-05\n", + "Epoch 216/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0274 - lr: 3.9062e-05\n", + "Epoch 217/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0216 - lr: 3.9062e-05\n", + "Epoch 218/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0063 - lr: 3.9062e-05\n", + "Epoch 219/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0076 - lr: 3.9062e-05\n", + "Epoch 220/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0061 - lr: 3.9062e-05\n", + "Epoch 221/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0179 - lr: 3.9062e-05\n", + "Epoch 222/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0463 - lr: 3.9062e-05\n", + "Epoch 223/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0545 - lr: 3.9062e-05\n", + "Epoch 224/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0094 - lr: 3.9062e-05\n", + "Epoch 225/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0534 - lr: 3.9062e-05\n", + "Epoch 226/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0254 - lr: 3.9062e-05\n", + "Epoch 227/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0067 - lr: 3.9062e-05\n", + "Epoch 228/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0080 - lr: 3.9062e-05\n", + "Epoch 229/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0116 - lr: 3.9062e-05\n", + "Epoch 230/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0381 - lr: 3.9062e-05\n", + "Epoch 231/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0471 - lr: 3.9062e-05\n", + "Epoch 232/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0042 - lr: 3.9062e-05\n", + "Epoch 233/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0097 - lr: 3.9062e-05\n", + "Epoch 234/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0033 - lr: 3.9062e-05\n", + "Epoch 235/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0030 - lr: 3.9062e-05\n", + "Epoch 236/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0184 - lr: 3.9062e-05\n", + "Epoch 237/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0214 - lr: 3.9062e-05\n", + "Epoch 238/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0520 - lr: 3.9062e-05\n", + "Epoch 239/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0182 - lr: 3.9062e-05\n", + "Epoch 240/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0103 - lr: 3.9062e-05\n", + "Epoch 241/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0041 - lr: 3.9062e-05\n", + "Epoch 242/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0509 - lr: 3.9062e-05\n", + "Epoch 243/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0191 - lr: 3.9062e-05\n", + "Epoch 244/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0159 - lr: 3.9062e-05\n", + "Epoch 245/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0174 - lr: 3.9062e-05\n", + "Epoch 246/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0035 - lr: 3.9062e-05\n", + "Epoch 247/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0113 - lr: 3.9062e-05\n", + "Epoch 248/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0075 - lr: 3.9062e-05\n", + "Epoch 249/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0139 - lr: 3.9062e-05\n", + "Epoch 250/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0072 - lr: 3.9062e-05\n", + "Epoch 251/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0197 - lr: 3.9062e-05\n", + "Epoch 252/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0061 - lr: 3.9062e-05\n", + "Epoch 253/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0181 - lr: 3.9062e-05\n", + "Epoch 254/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0148 - lr: 3.9062e-05\n", + "Epoch 255/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0328 - lr: 3.9062e-05\n", + "Epoch 256/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0056 - lr: 3.9062e-05\n", + "Epoch 257/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0374 - lr: 3.9062e-05\n", + "Epoch 258/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0133 - lr: 3.9062e-05\n", + "Epoch 259/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0173 - lr: 3.9062e-05\n", + "Epoch 260/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0105 - lr: 3.9062e-05\n", + "Epoch 261/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0116 - lr: 1.9531e-05\n", + "Epoch 262/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 263/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0093 - lr: 1.9531e-05\n", + "Epoch 264/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0160 - lr: 1.9531e-05\n", + "Epoch 265/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 266/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0045 - lr: 1.9531e-05\n", + "Epoch 267/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0038 - lr: 1.9531e-05\n", + "Epoch 268/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 269/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0036 - lr: 1.9531e-05\n", + "Epoch 270/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0131 - lr: 1.9531e-05\n", + "Epoch 271/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0034 - lr: 1.9531e-05\n", + "Epoch 272/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0053 - lr: 1.9531e-05\n", + "Epoch 273/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0034 - lr: 1.9531e-05\n", + "Epoch 274/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0035 - lr: 1.9531e-05\n", + "Epoch 275/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0062 - lr: 1.9531e-05\n", + "Epoch 276/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 277/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0064 - lr: 1.9531e-05\n", + "Epoch 278/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0050 - lr: 1.9531e-05\n", + "Epoch 279/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0121 - lr: 1.9531e-05\n", + "Epoch 280/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 281/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0051 - lr: 1.9531e-05\n", + "Epoch 282/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 283/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0075 - lr: 1.9531e-05\n", + "Epoch 284/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 285/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0041 - lr: 1.9531e-05\n", + "Epoch 286/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0052 - lr: 1.9531e-05\n", + "Epoch 287/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0064 - lr: 1.9531e-05\n", + "Epoch 288/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 289/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0057 - lr: 1.9531e-05\n", + "Epoch 290/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0056 - lr: 1.9531e-05\n", + "Epoch 291/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0096 - lr: 1.9531e-05\n", + "Epoch 292/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0036 - lr: 1.9531e-05\n", + "Epoch 293/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0075 - lr: 1.9531e-05\n", + "Epoch 294/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0066 - lr: 1.9531e-05\n", + "Epoch 295/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0039 - lr: 1.9531e-05\n", + "Epoch 296/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0050 - lr: 1.9531e-05\n", + "Epoch 297/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0056 - lr: 1.9531e-05\n", + "Epoch 298/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0056 - lr: 1.9531e-05\n", + "Epoch 299/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0140 - lr: 1.9531e-05\n", + "Epoch 300/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0034 - lr: 1.9531e-05\n", + "Epoch 301/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0038 - lr: 1.9531e-05\n", + "Epoch 302/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0045 - lr: 1.9531e-05\n", + "Epoch 303/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0078 - lr: 1.9531e-05\n", + "Epoch 304/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0113 - lr: 1.9531e-05\n", + "Epoch 305/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 306/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0074 - lr: 1.9531e-05\n", + "Epoch 307/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0069 - lr: 1.9531e-05\n", + "Epoch 308/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 309/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0054 - lr: 1.9531e-05\n", + "Epoch 310/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0034 - lr: 1.9531e-05\n", + "Epoch 311/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0106 - lr: 1.9531e-05\n", + "Epoch 312/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0080 - lr: 1.9531e-05\n", + "Epoch 313/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0068 - lr: 1.9531e-05\n", + "Epoch 314/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0057 - lr: 1.9531e-05\n", + "Epoch 315/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0036 - lr: 1.9531e-05\n", + "Epoch 316/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0038 - lr: 1.9531e-05\n", + "Epoch 317/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0172 - lr: 1.9531e-05\n", + "Epoch 318/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0070 - lr: 1.9531e-05\n", + "Epoch 319/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0079 - lr: 1.9531e-05\n", + "Epoch 320/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0048 - lr: 1.9531e-05\n", + "Epoch 321/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0073 - lr: 1.9531e-05\n", + "Epoch 322/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 323/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0166 - lr: 1.9531e-05\n", + "Epoch 324/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 325/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0047 - lr: 1.9531e-05\n", + "Epoch 326/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0104 - lr: 1.9531e-05\n", + "Epoch 327/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0161 - lr: 1.9531e-05\n", + "Epoch 328/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0036 - lr: 1.9531e-05\n", + "Epoch 329/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0146 - lr: 1.9531e-05\n", + "Epoch 330/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0042 - lr: 1.9531e-05\n", + "Epoch 331/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 332/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0058 - lr: 9.7656e-06\n", + "Epoch 333/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 334/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0055 - lr: 9.7656e-06\n", + "Epoch 335/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0038 - lr: 9.7656e-06\n", + "Epoch 336/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 337/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 338/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 339/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 340/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 341/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0072 - lr: 9.7656e-06\n", + "Epoch 342/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0038 - lr: 9.7656e-06\n", + "Epoch 343/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0039 - lr: 9.7656e-06\n", + "Epoch 344/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0051 - lr: 9.7656e-06\n", + "Epoch 345/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 346/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0034 - lr: 9.7656e-06\n", + "Epoch 347/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0053 - lr: 9.7656e-06\n", + "Epoch 348/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0062 - lr: 9.7656e-06\n", + "Epoch 349/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0036 - lr: 9.7656e-06\n", + "Epoch 350/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 351/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0045 - lr: 9.7656e-06\n", + "Epoch 352/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 353/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 354/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 355/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 356/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0034 - lr: 9.7656e-06\n", + "Epoch 357/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0049 - lr: 9.7656e-06\n", + "Epoch 358/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0060 - lr: 9.7656e-06\n", + "Epoch 359/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 360/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 361/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 362/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 363/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 364/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 365/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 366/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 367/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 368/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 369/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 370/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 371/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 372/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 373/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0036 - lr: 9.7656e-06\n", + "Epoch 374/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 375/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 376/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 377/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 378/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0041 - lr: 9.7656e-06\n", + "Epoch 379/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0045 - lr: 9.7656e-06\n", + "Epoch 380/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0058 - lr: 9.7656e-06\n", + "Epoch 381/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0055 - lr: 9.7656e-06\n", + "Epoch 382/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 383/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 384/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0031 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 385/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0040 - lr: 9.7656e-06\n", + "Epoch 386/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0048 - lr: 9.7656e-06\n", + "Epoch 387/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0036 - lr: 9.7656e-06\n", + "Epoch 388/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0074 - lr: 9.7656e-06\n", + "Epoch 389/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0072 - lr: 9.7656e-06\n", + "Epoch 390/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0037 - lr: 4.8828e-06\n", + "Epoch 391/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 392/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0037 - lr: 4.8828e-06\n", + "Epoch 393/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 394/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 395/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0034 - lr: 4.8828e-06\n", + "Epoch 396/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 397/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 398/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 399/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 400/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0031 - lr: 4.8828e-06\n", + "Epoch 401/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 402/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 403/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 404/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0037 - lr: 4.8828e-06\n", + "Epoch 405/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 406/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 407/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 408/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 409/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 410/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 411/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 412/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 413/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 414/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 415/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 416/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0033 - lr: 4.8828e-06\n", + "Epoch 417/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 418/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0041 - lr: 4.8828e-06\n", + "Epoch 419/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 420/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 421/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 422/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 423/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 424/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 425/1024\n", + "90/90 [==============================] - 0s 4ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 426/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0033 - lr: 4.8828e-06\n", + "Epoch 427/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 428/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 429/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 430/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 431/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 432/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 433/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 434/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 435/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 436/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 437/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0028 - lr: 2.4414e-06\n", + "Epoch 438/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 2.4414e-06\n", + "Epoch 439/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 440/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 441/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 442/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 443/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0030 - lr: 2.4414e-06\n", + "Epoch 444/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 445/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 446/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 447/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 448/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 449/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 450/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 451/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0029 - lr: 2.4414e-06\n", + "Epoch 452/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 453/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 454/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 455/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 456/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 457/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 458/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 459/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 460/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 461/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 462/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 463/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 464/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 465/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 466/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 467/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 468/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 469/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 470/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 471/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 472/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 473/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 474/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 475/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 476/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 477/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 478/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 479/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 480/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 481/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 482/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 483/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 1.2207e-06\n", + "Epoch 484/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 485/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 486/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 487/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 488/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 489/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 490/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 491/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 492/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 493/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 494/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 6.1035e-07\n", + "Epoch 495/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 496/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 497/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 498/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 499/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 500/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 501/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 502/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 503/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 504/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 505/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0027 - lr: 6.1035e-07\n", + "Epoch 506/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 507/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 508/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 509/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 510/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 511/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 512/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 513/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 514/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 515/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 516/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 517/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 518/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 519/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 520/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 521/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 522/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 523/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 524/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 525/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 526/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 527/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 528/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 529/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 530/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 531/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 532/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 533/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 534/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.0518e-07\n", + "Epoch 535/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 536/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 537/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 538/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 539/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 540/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 541/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 542/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 543/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 544/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 545/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 546/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 547/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 548/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 549/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 550/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 551/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 552/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 553/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 554/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 555/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 556/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 557/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 558/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 559/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.5259e-07\n", + "Epoch 560/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 561/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 562/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 563/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 564/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 565/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 566/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 567/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 568/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 569/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 570/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 571/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 572/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 573/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 574/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 575/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 576/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 577/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 578/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 579/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 580/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 581/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 582/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 583/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 584/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 7.6294e-08\n", + "Epoch 585/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 586/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 587/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 588/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 589/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 590/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 591/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 592/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 593/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 594/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 595/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 596/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 597/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 598/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 599/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 600/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 601/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 602/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 603/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 604/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 605/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 606/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 607/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 608/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 609/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 3.8147e-08\n", + "Epoch 610/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 611/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 612/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 613/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 614/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 615/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 616/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 617/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 618/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 619/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 620/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 621/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 622/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 623/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 624/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 625/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 626/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 627/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 628/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 629/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 630/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 631/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 632/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 633/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 634/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 1.9073e-08\n", + "Epoch 635/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 636/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 637/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 638/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 639/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 640/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 641/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 642/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 643/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 644/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 645/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 646/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 647/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 648/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 649/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 650/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 651/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 652/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 653/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 654/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 655/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 656/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 657/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 658/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 659/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0026 - lr: 9.5367e-09\n", + "Epoch 1/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0162 - val_loss: 0.0337 - lr: 0.0025\n", + "Epoch 2/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0086 - val_loss: 0.0453 - lr: 0.0025\n", + "Epoch 3/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0070 - val_loss: 0.1119 - lr: 0.0025\n", + "Epoch 4/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0061 - val_loss: 0.2422 - lr: 0.0025\n", + "Epoch 5/1024\n", + "90/90 [==============================] - 0s 831us/step - loss: 0.0055 - val_loss: 0.1867 - lr: 0.0025\n", + "Epoch 6/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0051 - val_loss: 0.0551 - lr: 0.0025\n", + "Epoch 7/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0046 - val_loss: 0.1507 - lr: 0.0025\n", + "Epoch 8/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0043 - val_loss: 0.1627 - lr: 0.0025\n", + "Epoch 9/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0045 - val_loss: 0.2282 - lr: 0.0025\n", + "Epoch 10/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0042 - val_loss: 0.0914 - lr: 0.0025\n", + "Epoch 11/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0043 - val_loss: 0.3210 - lr: 0.0025\n", + "Epoch 12/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0040 - val_loss: 0.1090 - lr: 0.0025\n", + "Epoch 13/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0040 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 14/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0041 - val_loss: 0.0339 - lr: 0.0025\n", + "Epoch 15/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0042 - val_loss: 0.0163 - lr: 0.0025\n", + "Epoch 16/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0041 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 17/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0039 - val_loss: 0.0743 - lr: 0.0025\n", + "Epoch 18/1024\n", + "90/90 [==============================] - 0s 980us/step - loss: 0.0039 - val_loss: 0.2028 - lr: 0.0025\n", + "Epoch 19/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.3298 - lr: 0.0025\n", + "Epoch 20/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.2000 - lr: 0.0025\n", + "Epoch 21/1024\n", + "90/90 [==============================] - 0s 837us/step - loss: 0.0036 - val_loss: 0.0345 - lr: 0.0025\n", + "Epoch 22/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0037 - val_loss: 0.3248 - lr: 0.0025\n", + "Epoch 23/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0037 - val_loss: 0.3253 - lr: 0.0025\n", + "Epoch 24/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0035 - val_loss: 0.3070 - lr: 0.0025\n", + "Epoch 25/1024\n", + "90/90 [==============================] - 0s 833us/step - loss: 0.0036 - val_loss: 0.3283 - lr: 0.0025\n", + "Epoch 26/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0036 - val_loss: 0.0853 - lr: 0.0025\n", + "Epoch 27/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0036 - val_loss: 0.2262 - lr: 0.0025\n", + "Epoch 28/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0035 - val_loss: 0.3298 - lr: 0.0025\n", + "Epoch 29/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0038 - val_loss: 0.1514 - lr: 0.0025\n", + "Epoch 30/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0035 - val_loss: 0.0532 - lr: 0.0025\n", + "Epoch 31/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0039 - val_loss: 0.1852 - lr: 0.0025\n", + "Epoch 32/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0035 - val_loss: 0.1469 - lr: 0.0025\n", + "Epoch 33/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0036 - val_loss: 0.1926 - lr: 0.0025\n", + "Epoch 34/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0034 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 35/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0035 - val_loss: 0.2885 - lr: 0.0025\n", + "Epoch 36/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0037 - val_loss: 0.2894 - lr: 0.0025\n", + "Epoch 37/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0036 - val_loss: 0.2274 - lr: 0.0025\n", + "Epoch 38/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0035 - val_loss: 0.1992 - lr: 0.0025\n", + "Epoch 39/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0035 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 40/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0034 - val_loss: 0.3245 - lr: 0.0025\n", + "Epoch 41/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0035 - val_loss: 0.3290 - lr: 0.0012\n", + "Epoch 42/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0034 - val_loss: 0.1089 - lr: 0.0012\n", + "Epoch 43/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0033 - val_loss: 0.0756 - lr: 0.0012\n", + "Epoch 44/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0034 - val_loss: 0.0401 - lr: 0.0012\n", + "Epoch 45/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0033 - val_loss: 0.3272 - lr: 0.0012\n", + "Epoch 46/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0033 - val_loss: 0.3168 - lr: 0.0012\n", + "Epoch 47/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0033 - val_loss: 0.3233 - lr: 0.0012\n", + "Epoch 48/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0031 - val_loss: 0.1908 - lr: 0.0012\n", + "Epoch 49/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0034 - val_loss: 0.2271 - lr: 0.0012\n", + "Epoch 50/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0032 - val_loss: 0.0582 - lr: 0.0012\n", + "Epoch 51/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0033 - val_loss: 0.2781 - lr: 0.0012\n", + "Epoch 52/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0031 - val_loss: 0.1596 - lr: 0.0012\n", + "Epoch 53/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0032 - val_loss: 0.0435 - lr: 0.0012\n", + "Epoch 54/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0033 - val_loss: 0.0385 - lr: 0.0012\n", + "Epoch 55/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0032 - val_loss: 0.0780 - lr: 0.0012\n", + "Epoch 56/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0033 - val_loss: 0.3291 - lr: 0.0012\n", + "Epoch 57/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0032 - val_loss: 0.1991 - lr: 0.0012\n", + "Epoch 58/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0033 - val_loss: 0.2066 - lr: 0.0012\n", + "Epoch 59/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0031 - val_loss: 0.2353 - lr: 0.0012\n", + "Epoch 60/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0033 - val_loss: 0.1137 - lr: 0.0012\n", + "Epoch 61/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0032 - val_loss: 0.3180 - lr: 0.0012\n", + "Epoch 62/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0033 - val_loss: 0.2062 - lr: 0.0012\n", + "Epoch 63/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0031 - val_loss: 0.2061 - lr: 0.0012\n", + "Epoch 64/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.3123 - lr: 0.0012\n", + "Epoch 65/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.3288 - lr: 0.0012\n", + "Epoch 66/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0032 - val_loss: 0.3214 - lr: 6.2500e-04\n", + "Epoch 67/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0031 - val_loss: 0.1355 - lr: 6.2500e-04\n", + "Epoch 68/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0030 - val_loss: 0.0484 - lr: 6.2500e-04\n", + "Epoch 69/1024\n", + "90/90 [==============================] - 0s 838us/step - loss: 0.0031 - val_loss: 0.2270 - lr: 6.2500e-04\n", + "Epoch 70/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0030 - val_loss: 0.0456 - lr: 6.2500e-04\n", + "Epoch 71/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0031 - val_loss: 0.1166 - lr: 6.2500e-04\n", + "Epoch 72/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0031 - val_loss: 0.0350 - lr: 6.2500e-04\n", + "Epoch 73/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0030 - val_loss: 0.2658 - lr: 6.2500e-04\n", + "Epoch 74/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0030 - val_loss: 0.0387 - lr: 6.2500e-04\n", + "Epoch 75/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0031 - val_loss: 0.2661 - lr: 6.2500e-04\n", + "Epoch 76/1024\n", + "90/90 [==============================] - 0s 837us/step - loss: 0.0031 - val_loss: 0.0306 - lr: 6.2500e-04\n", + "Epoch 77/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0031 - val_loss: 0.0508 - lr: 6.2500e-04\n", + "Epoch 78/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0030 - val_loss: 0.0662 - lr: 6.2500e-04\n", + "Epoch 79/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0032 - val_loss: 0.1038 - lr: 6.2500e-04\n", + "Epoch 80/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0031 - val_loss: 0.0455 - lr: 6.2500e-04\n", + "Epoch 81/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0032 - val_loss: 0.2104 - lr: 6.2500e-04\n", + "Epoch 82/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0030 - val_loss: 0.0288 - lr: 6.2500e-04\n", + "Epoch 83/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0030 - val_loss: 0.1063 - lr: 6.2500e-04\n", + "Epoch 84/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0031 - val_loss: 0.3056 - lr: 6.2500e-04\n", + "Epoch 85/1024\n", + "90/90 [==============================] - 0s 838us/step - loss: 0.0030 - val_loss: 0.3213 - lr: 6.2500e-04\n", + "Epoch 86/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0032 - val_loss: 0.0759 - lr: 6.2500e-04\n", + "Epoch 87/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.2254 - lr: 6.2500e-04\n", + "Epoch 88/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0031 - val_loss: 0.0319 - lr: 6.2500e-04\n", + "Epoch 89/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0031 - val_loss: 0.1777 - lr: 6.2500e-04\n", + "Epoch 90/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0031 - val_loss: 0.1783 - lr: 6.2500e-04\n", + "Epoch 91/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0056 - lr: 3.1250e-04\n", + "Epoch 92/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0030 - val_loss: 0.0405 - lr: 3.1250e-04\n", + "Epoch 93/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0030 - val_loss: 0.1284 - lr: 3.1250e-04\n", + "Epoch 94/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0030 - val_loss: 0.0306 - lr: 3.1250e-04\n", + "Epoch 95/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0030 - val_loss: 0.0071 - lr: 3.1250e-04\n", + "Epoch 96/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0030 - val_loss: 0.1139 - lr: 3.1250e-04\n", + "Epoch 97/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0031 - val_loss: 0.0524 - lr: 3.1250e-04\n", + "Epoch 98/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0031 - val_loss: 0.0218 - lr: 3.1250e-04\n", + "Epoch 99/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0030 - val_loss: 0.2167 - lr: 3.1250e-04\n", + "Epoch 100/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0029 - val_loss: 0.0393 - lr: 3.1250e-04\n", + "Epoch 101/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0030 - val_loss: 0.0430 - lr: 3.1250e-04\n", + "Epoch 102/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0031 - val_loss: 0.0761 - lr: 3.1250e-04\n", + "Epoch 103/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0030 - val_loss: 0.0112 - lr: 3.1250e-04\n", + "Epoch 104/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.1880 - lr: 3.1250e-04\n", + "Epoch 105/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0030 - val_loss: 0.0084 - lr: 3.1250e-04\n", + "Epoch 106/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0030 - val_loss: 0.0204 - lr: 3.1250e-04\n", + "Epoch 107/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0030 - val_loss: 0.0173 - lr: 3.1250e-04\n", + "Epoch 108/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2058 - lr: 3.1250e-04\n", + "Epoch 109/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0875 - lr: 3.1250e-04\n", + "Epoch 110/1024\n", + "90/90 [==============================] - 0s 964us/step - loss: 0.0029 - val_loss: 0.0100 - lr: 3.1250e-04\n", + "Epoch 111/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.1672 - lr: 3.1250e-04\n", + "Epoch 112/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0030 - val_loss: 0.1869 - lr: 3.1250e-04\n", + "Epoch 113/1024\n", + "90/90 [==============================] - 0s 837us/step - loss: 0.0029 - val_loss: 0.0400 - lr: 3.1250e-04\n", + "Epoch 114/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0029 - val_loss: 0.2014 - lr: 3.1250e-04\n", + "Epoch 115/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0030 - val_loss: 0.1869 - lr: 3.1250e-04\n", + "Epoch 116/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0030 - val_loss: 0.0204 - lr: 3.1250e-04\n", + "Epoch 117/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0048 - lr: 1.5625e-04\n", + "Epoch 118/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0038 - lr: 1.5625e-04\n", + "Epoch 119/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0644 - lr: 1.5625e-04\n", + "Epoch 120/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0028 - val_loss: 0.0170 - lr: 1.5625e-04\n", + "Epoch 121/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0029 - val_loss: 0.0230 - lr: 1.5625e-04\n", + "Epoch 122/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0171 - lr: 1.5625e-04\n", + "Epoch 123/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0114 - lr: 1.5625e-04\n", + "Epoch 124/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0030 - val_loss: 0.0643 - lr: 1.5625e-04\n", + "Epoch 125/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0198 - lr: 1.5625e-04\n", + "Epoch 126/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0527 - lr: 1.5625e-04\n", + "Epoch 127/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0031 - val_loss: 0.1066 - lr: 1.5625e-04\n", + "Epoch 128/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0029 - val_loss: 0.0126 - lr: 1.5625e-04\n", + "Epoch 129/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0029 - val_loss: 0.0094 - lr: 1.5625e-04\n", + "Epoch 130/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0237 - lr: 1.5625e-04\n", + "Epoch 131/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0030 - val_loss: 0.0102 - lr: 1.5625e-04\n", + "Epoch 132/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0029 - val_loss: 0.0119 - lr: 1.5625e-04\n", + "Epoch 133/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0029 - val_loss: 0.0440 - lr: 1.5625e-04\n", + "Epoch 134/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0030 - val_loss: 0.1435 - lr: 1.5625e-04\n", + "Epoch 135/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0172 - lr: 1.5625e-04\n", + "Epoch 136/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0030 - val_loss: 0.1439 - lr: 1.5625e-04\n", + "Epoch 137/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0572 - lr: 1.5625e-04\n", + "Epoch 138/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0409 - lr: 1.5625e-04\n", + "Epoch 139/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0104 - lr: 1.5625e-04\n", + "Epoch 140/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0067 - lr: 1.5625e-04\n", + "Epoch 141/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0443 - lr: 1.5625e-04\n", + "Epoch 142/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.1467 - lr: 1.5625e-04\n", + "Epoch 143/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0030 - val_loss: 0.2135 - lr: 1.5625e-04\n", + "Epoch 144/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0078 - lr: 7.8125e-05\n", + "Epoch 145/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0030 - val_loss: 0.0148 - lr: 7.8125e-05\n", + "Epoch 146/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0030 - val_loss: 0.0695 - lr: 7.8125e-05\n", + "Epoch 147/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0059 - lr: 7.8125e-05\n", + "Epoch 148/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0314 - lr: 7.8125e-05\n", + "Epoch 149/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.0089 - lr: 7.8125e-05\n", + "Epoch 150/1024\n", + "90/90 [==============================] - 0s 992us/step - loss: 0.0029 - val_loss: 0.0090 - lr: 7.8125e-05\n", + "Epoch 151/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0410 - lr: 7.8125e-05\n", + "Epoch 152/1024\n", + "90/90 [==============================] - 0s 978us/step - loss: 0.0029 - val_loss: 0.0308 - lr: 7.8125e-05\n", + "Epoch 153/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0352 - lr: 7.8125e-05\n", + "Epoch 154/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0195 - lr: 7.8125e-05\n", + "Epoch 155/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0083 - lr: 7.8125e-05\n", + "Epoch 156/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0035 - lr: 7.8125e-05\n", + "Epoch 157/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0030 - val_loss: 0.0118 - lr: 7.8125e-05\n", + "Epoch 158/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0365 - lr: 7.8125e-05\n", + "Epoch 159/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0038 - lr: 7.8125e-05\n", + "Epoch 160/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.0081 - lr: 7.8125e-05\n", + "Epoch 161/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0292 - lr: 7.8125e-05\n", + "Epoch 162/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0029 - val_loss: 0.0325 - lr: 7.8125e-05\n", + "Epoch 163/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0128 - lr: 7.8125e-05\n", + "Epoch 164/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0098 - lr: 7.8125e-05\n", + "Epoch 165/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0058 - lr: 7.8125e-05\n", + "Epoch 166/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0283 - lr: 7.8125e-05\n", + "Epoch 167/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0385 - lr: 7.8125e-05\n", + "Epoch 168/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0030 - val_loss: 0.0039 - lr: 7.8125e-05\n", + "Epoch 169/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0042 - lr: 7.8125e-05\n", + "Epoch 170/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0095 - lr: 7.8125e-05\n", + "Epoch 171/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0312 - lr: 7.8125e-05\n", + "Epoch 172/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0030 - val_loss: 0.0207 - lr: 7.8125e-05\n", + "Epoch 173/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0326 - lr: 7.8125e-05\n", + "Epoch 174/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0063 - lr: 7.8125e-05\n", + "Epoch 175/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0125 - lr: 7.8125e-05\n", + "Epoch 176/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0036 - lr: 7.8125e-05\n", + "Epoch 177/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0107 - lr: 7.8125e-05\n", + "Epoch 178/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0442 - lr: 7.8125e-05\n", + "Epoch 179/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0030 - val_loss: 0.0142 - lr: 7.8125e-05\n", + "Epoch 180/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0171 - lr: 7.8125e-05\n", + "Epoch 181/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0116 - lr: 7.8125e-05\n", + "Epoch 182/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0040 - lr: 3.9062e-05\n", + "Epoch 183/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0030 - val_loss: 0.0042 - lr: 3.9062e-05\n", + "Epoch 184/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0028 - val_loss: 0.0112 - lr: 3.9062e-05\n", + "Epoch 185/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0047 - lr: 3.9062e-05\n", + "Epoch 186/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0156 - lr: 3.9062e-05\n", + "Epoch 187/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0030 - val_loss: 0.0150 - lr: 3.9062e-05\n", + "Epoch 188/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0223 - lr: 3.9062e-05\n", + "Epoch 189/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0086 - lr: 3.9062e-05\n", + "Epoch 190/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0118 - lr: 3.9062e-05\n", + "Epoch 191/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0030 - val_loss: 0.0062 - lr: 3.9062e-05\n", + "Epoch 192/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0065 - lr: 3.9062e-05\n", + "Epoch 193/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0084 - lr: 3.9062e-05\n", + "Epoch 194/1024\n", + "90/90 [==============================] - 0s 987us/step - loss: 0.0030 - val_loss: 0.0120 - lr: 3.9062e-05\n", + "Epoch 195/1024\n", + "90/90 [==============================] - 0s 832us/step - loss: 0.0030 - val_loss: 0.0056 - lr: 3.9062e-05\n", + "Epoch 196/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0063 - lr: 3.9062e-05\n", + "Epoch 197/1024\n", + "90/90 [==============================] - 0s 836us/step - loss: 0.0030 - val_loss: 0.0045 - lr: 3.9062e-05\n", + "Epoch 198/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0028 - val_loss: 0.0089 - lr: 3.9062e-05\n", + "Epoch 199/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0028 - val_loss: 0.0049 - lr: 3.9062e-05\n", + "Epoch 200/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0059 - lr: 3.9062e-05\n", + "Epoch 201/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0151 - lr: 3.9062e-05\n", + "Epoch 202/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0030 - lr: 3.9062e-05\n", + "Epoch 203/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0047 - lr: 3.9062e-05\n", + "Epoch 204/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0030 - val_loss: 0.0040 - lr: 3.9062e-05\n", + "Epoch 205/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0070 - lr: 3.9062e-05\n", + "Epoch 206/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0029 - val_loss: 0.0060 - lr: 3.9062e-05\n", + "Epoch 207/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0029 - val_loss: 0.0052 - lr: 3.9062e-05\n", + "Epoch 208/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0135 - lr: 3.9062e-05\n", + "Epoch 209/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0029 - val_loss: 0.0046 - lr: 3.9062e-05\n", + "Epoch 210/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0057 - lr: 3.9062e-05\n", + "Epoch 211/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 3.9062e-05\n", + "Epoch 212/1024\n", + "90/90 [==============================] - 0s 836us/step - loss: 0.0029 - val_loss: 0.0089 - lr: 3.9062e-05\n", + "Epoch 213/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0028 - val_loss: 0.0048 - lr: 3.9062e-05\n", + "Epoch 214/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0029 - val_loss: 0.0050 - lr: 3.9062e-05\n", + "Epoch 215/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0030 - val_loss: 0.0058 - lr: 3.9062e-05\n", + "Epoch 216/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0037 - lr: 3.9062e-05\n", + "Epoch 217/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0059 - lr: 3.9062e-05\n", + "Epoch 218/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0202 - lr: 3.9062e-05\n", + "Epoch 219/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0313 - lr: 3.9062e-05\n", + "Epoch 220/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0030 - val_loss: 0.0050 - lr: 3.9062e-05\n", + "Epoch 221/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0115 - lr: 3.9062e-05\n", + "Epoch 222/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0029 - val_loss: 0.0046 - lr: 3.9062e-05\n", + "Epoch 223/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0065 - lr: 3.9062e-05\n", + "Epoch 224/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0029 - val_loss: 0.0044 - lr: 3.9062e-05\n", + "Epoch 225/1024\n", + "90/90 [==============================] - 0s 980us/step - loss: 0.0029 - val_loss: 0.0489 - lr: 3.9062e-05\n", + "Epoch 226/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0027 - val_loss: 0.0035 - lr: 3.9062e-05\n", + "Epoch 227/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0069 - lr: 3.9062e-05\n", + "Epoch 228/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0030 - val_loss: 0.0081 - lr: 3.9062e-05\n", + "Epoch 229/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0028 - lr: 3.9062e-05\n", + "Epoch 230/1024\n", + "90/90 [==============================] - 0s 893us/step - loss: 0.0029 - val_loss: 0.0077 - lr: 3.9062e-05\n", + "Epoch 231/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 3.9062e-05\n", + "Epoch 232/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0028 - val_loss: 0.0095 - lr: 3.9062e-05\n", + "Epoch 233/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0103 - lr: 3.9062e-05\n", + "Epoch 234/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0034 - lr: 3.9062e-05\n", + "Epoch 235/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0029 - val_loss: 0.0053 - lr: 3.9062e-05\n", + "Epoch 236/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0030 - val_loss: 0.0069 - lr: 3.9062e-05\n", + "Epoch 237/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0173 - lr: 3.9062e-05\n", + "Epoch 238/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0028 - val_loss: 0.0057 - lr: 3.9062e-05\n", + "Epoch 239/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0029 - val_loss: 0.0139 - lr: 3.9062e-05\n", + "Epoch 240/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0029 - val_loss: 0.0090 - lr: 3.9062e-05\n", + "Epoch 241/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0029 - val_loss: 0.0045 - lr: 3.9062e-05\n", + "Epoch 242/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0173 - lr: 3.9062e-05\n", + "Epoch 243/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0029 - val_loss: 0.0029 - lr: 3.9062e-05\n", + "Epoch 244/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0029 - val_loss: 0.0130 - lr: 3.9062e-05\n", + "Epoch 245/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0190 - lr: 3.9062e-05\n", + "Epoch 246/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0061 - lr: 3.9062e-05\n", + "Epoch 247/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0029 - val_loss: 0.0034 - lr: 3.9062e-05\n", + "Epoch 248/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0028 - val_loss: 0.0163 - lr: 3.9062e-05\n", + "Epoch 249/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0027 - val_loss: 0.0129 - lr: 3.9062e-05\n", + "Epoch 250/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0031 - val_loss: 0.0044 - lr: 3.9062e-05\n", + "Epoch 251/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0320 - lr: 3.9062e-05\n", + "Epoch 252/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0029 - val_loss: 0.0176 - lr: 3.9062e-05\n", + "Epoch 253/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0254 - lr: 3.9062e-05\n", + "Epoch 254/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0027 - val_loss: 0.0031 - lr: 3.9062e-05\n", + "Epoch 255/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0029 - val_loss: 0.0085 - lr: 1.9531e-05\n", + "Epoch 256/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0069 - lr: 1.9531e-05\n", + "Epoch 257/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 258/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0030 - lr: 1.9531e-05\n", + "Epoch 259/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0029 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 260/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0027 - lr: 1.9531e-05\n", + "Epoch 261/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 262/1024\n", + "90/90 [==============================] - 0s 828us/step - loss: 0.0029 - val_loss: 0.0027 - lr: 1.9531e-05\n", + "Epoch 263/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0035 - lr: 1.9531e-05\n", + "Epoch 264/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0036 - lr: 1.9531e-05\n", + "Epoch 265/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 266/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0033 - lr: 1.9531e-05\n", + "Epoch 267/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0041 - lr: 1.9531e-05\n", + "Epoch 268/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 269/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 270/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0027 - val_loss: 0.0050 - lr: 1.9531e-05\n", + "Epoch 271/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0029 - val_loss: 0.0029 - lr: 1.9531e-05\n", + "Epoch 272/1024\n", + "90/90 [==============================] - 0s 978us/step - loss: 0.0028 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 273/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0035 - lr: 1.9531e-05\n", + "Epoch 274/1024\n", + "90/90 [==============================] - 0s 978us/step - loss: 0.0029 - val_loss: 0.0039 - lr: 1.9531e-05\n", + "Epoch 275/1024\n", + "90/90 [==============================] - 0s 838us/step - loss: 0.0029 - val_loss: 0.0039 - lr: 1.9531e-05\n", + "Epoch 276/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0053 - lr: 1.9531e-05\n", + "Epoch 277/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0033 - lr: 1.9531e-05\n", + "Epoch 278/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0061 - lr: 1.9531e-05\n", + "Epoch 279/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0072 - lr: 1.9531e-05\n", + "Epoch 280/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0055 - lr: 1.9531e-05\n", + "Epoch 281/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0029 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 282/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0029 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 283/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 284/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0028 - val_loss: 0.0052 - lr: 1.9531e-05\n", + "Epoch 285/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0030 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 286/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0029 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 287/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.7656e-06\n", + "Epoch 288/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0030 - val_loss: 0.0055 - lr: 9.7656e-06\n", + "Epoch 289/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 290/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.7656e-06\n", + "Epoch 291/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 292/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 293/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 294/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0038 - lr: 9.7656e-06\n", + "Epoch 295/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 296/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 297/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0027 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 298/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 299/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 300/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0035 - lr: 9.7656e-06\n", + "Epoch 301/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 302/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 303/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0029 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 304/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 305/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 306/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0052 - lr: 9.7656e-06\n", + "Epoch 307/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 308/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 309/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 310/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 311/1024\n", + "90/90 [==============================] - 0s 907us/step - loss: 0.0028 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 312/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 313/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 314/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 315/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 316/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 317/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 318/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 319/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 320/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 321/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 322/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 323/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 324/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 325/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 326/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0028 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 327/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 328/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 329/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 330/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 331/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 332/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 333/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 334/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 335/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 336/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 337/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 338/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 339/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 340/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 341/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 342/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 343/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0030 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 344/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0029 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 345/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 346/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 347/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0029 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 348/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 349/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 350/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 351/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 352/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 353/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 354/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 355/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 356/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 357/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 358/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 359/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 360/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 361/1024\n", + "90/90 [==============================] - 0s 828us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 362/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 363/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 364/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 365/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 366/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 367/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 368/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 369/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 370/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 371/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 372/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 373/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 374/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 375/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 376/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 377/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 378/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 379/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 380/1024\n", + "90/90 [==============================] - 0s 827us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 381/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 382/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 383/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 384/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 385/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 386/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 387/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 388/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 389/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 390/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 391/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 392/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 393/1024\n", + "90/90 [==============================] - 0s 899us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 394/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 395/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 396/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 397/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 398/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 399/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 400/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 401/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 402/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 403/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 404/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 405/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 406/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 407/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 408/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 409/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 410/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 411/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 412/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 413/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 414/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 415/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 416/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 417/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 418/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 419/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 420/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 421/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 422/1024\n", + "90/90 [==============================] - 0s 836us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 423/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 424/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 425/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 426/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 427/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 428/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 429/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 430/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 431/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 432/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 433/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 434/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 435/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 436/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 437/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 438/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 439/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0031 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 440/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 441/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 442/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 443/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 444/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 445/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 446/1024\n", + "90/90 [==============================] - 0s 893us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 447/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 448/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 449/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 450/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 451/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 452/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 453/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 454/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 455/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 456/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 457/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 458/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 459/1024\n", + "90/90 [==============================] - 0s 834us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 460/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 461/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 462/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 463/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 464/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 465/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 466/1024\n", + "90/90 [==============================] - 0s 837us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 467/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 468/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 469/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 470/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 471/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 472/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 473/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 474/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 475/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 476/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 477/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 478/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 479/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 480/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 481/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 482/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 483/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 484/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 485/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 486/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 487/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 488/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 489/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 490/1024\n", + "90/90 [==============================] - 0s 907us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 491/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 492/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 493/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 494/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 495/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 496/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 497/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 498/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 499/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 500/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 501/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 502/1024\n", + "90/90 [==============================] - 0s 891us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 503/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 504/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 505/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 506/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 507/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 508/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 509/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 510/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 511/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 512/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 513/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 514/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 515/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 516/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 517/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 518/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 519/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 520/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 521/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 522/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 523/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 524/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 525/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 526/1024\n", + "90/90 [==============================] - 0s 838us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 527/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 528/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 529/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 530/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 531/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 532/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 533/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 534/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 535/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 536/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 537/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 538/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 539/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 540/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 541/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 542/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 543/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 544/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 545/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 546/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 547/1024\n", + "90/90 [==============================] - 0s 834us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 548/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 549/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 550/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 551/1024\n", + "90/90 [==============================] - 0s 884us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 552/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 553/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 554/1024\n", + "90/90 [==============================] - 0s 884us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 555/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 556/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 557/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 558/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 559/1024\n", + "90/90 [==============================] - 0s 829us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 560/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 561/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 562/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 563/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 564/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 565/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 566/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 567/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 568/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 569/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 570/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 571/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 572/1024\n", + "90/90 [==============================] - 0s 833us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 573/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 574/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 575/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 576/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 577/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 578/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 579/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 580/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 581/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 582/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 583/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 584/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 585/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 586/1024\n", + "90/90 [==============================] - 0s 933us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 587/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 588/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 589/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 590/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 591/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 592/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 593/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 594/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 595/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 596/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 597/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 598/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 599/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 600/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 601/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 602/1024\n", + "90/90 [==============================] - 0s 897us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 603/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 604/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 605/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 606/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 607/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 608/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 609/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 610/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 611/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 612/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 613/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 614/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 615/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 616/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 617/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 618/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 619/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 620/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 621/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 622/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 623/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 624/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 625/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 626/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 627/1024\n", + "90/90 [==============================] - 0s 893us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 628/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 629/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 630/1024\n", + "90/90 [==============================] - 0s 891us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 631/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 632/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 633/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 634/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 635/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 636/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 637/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 638/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 639/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 640/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 641/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 642/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 643/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 644/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 645/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 646/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 647/1024\n", + "90/90 [==============================] - 0s 946us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 648/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 649/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 650/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 651/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 652/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 653/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 654/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 655/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 656/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 657/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 658/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 659/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 660/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 661/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 662/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 663/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 664/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 665/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 666/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 667/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 668/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 669/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 670/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 671/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 672/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 673/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 674/1024\n", + "90/90 [==============================] - 0s 982us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 675/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 676/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 677/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 678/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 679/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 680/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 681/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 682/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 683/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 684/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 685/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 686/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 687/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 688/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 689/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 690/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 691/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 692/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 693/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 694/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 695/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 696/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 697/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 698/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 699/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 700/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 701/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 702/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 703/1024\n", + "90/90 [==============================] - 0s 834us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 704/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 705/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 706/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 707/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 708/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 709/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 710/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 711/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 712/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 713/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 714/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 715/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 716/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 717/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 718/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 719/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 720/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 721/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 722/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 723/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 724/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 725/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 726/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 727/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 728/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 729/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 730/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 731/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 732/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 733/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 734/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 735/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 736/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 737/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 738/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 739/1024\n", + "90/90 [==============================] - 0s 891us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 740/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 741/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 742/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 743/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 744/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 745/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 746/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 747/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 748/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 749/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 750/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 751/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 752/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 753/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 754/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 755/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 756/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 757/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 758/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 759/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 760/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 761/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 762/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 763/1024\n", + "90/90 [==============================] - 0s 957us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 764/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 765/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 766/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 767/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 768/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 769/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 770/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 771/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 772/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 773/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 774/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 775/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 776/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 777/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 778/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 779/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 780/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 781/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 782/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 783/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 784/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 785/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 786/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 787/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 788/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 789/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 790/1024\n", + "90/90 [==============================] - 0s 915us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 791/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 792/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 793/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 794/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 795/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 796/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 797/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 798/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 799/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 800/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 801/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 802/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 803/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 804/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 805/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 806/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 807/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 808/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 809/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 810/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 811/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 812/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 813/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 814/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 815/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 816/1024\n", + "90/90 [==============================] - 0s 996us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 817/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 818/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 819/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 820/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 821/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 822/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 823/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 824/1024\n", + "90/90 [==============================] - 0s 846us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 825/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 826/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 827/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 828/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 829/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 830/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 831/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 832/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 833/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 834/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 835/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 836/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 837/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 838/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 839/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 840/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 841/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 842/1024\n", + "90/90 [==============================] - 0s 970us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 843/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 844/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 845/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 846/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 847/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 848/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 849/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 850/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 851/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 852/1024\n", + "90/90 [==============================] - 0s 842us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 853/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 854/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 855/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 856/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 857/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 858/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 859/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 860/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 861/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 862/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 863/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 864/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 865/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 866/1024\n", + "90/90 [==============================] - 0s 913us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 867/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 868/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 869/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 870/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 871/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 872/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 873/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 874/1024\n", + "90/90 [==============================] - 0s 915us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 875/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 876/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 877/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 878/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 879/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 880/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 881/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 882/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 883/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 884/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 885/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 886/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 887/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 888/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 889/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 890/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 891/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 892/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 893/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 894/1024\n", + "90/90 [==============================] - 0s 833us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 895/1024\n", + "90/90 [==============================] - 0s 965us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 896/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 897/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 898/1024\n", + "90/90 [==============================] - 0s 844us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 899/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 900/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 901/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 902/1024\n", + "90/90 [==============================] - 0s 950us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 903/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 904/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 905/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 906/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 907/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 908/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 909/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 910/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 911/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 912/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 913/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 914/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 915/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 916/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 917/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 918/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 919/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 920/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 921/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 922/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 923/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 924/1024\n", + "90/90 [==============================] - 0s 832us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 925/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 926/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 927/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 928/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 929/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 930/1024\n", + "90/90 [==============================] - 0s 899us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 931/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 932/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 933/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 934/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 935/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 936/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 937/1024\n", + "90/90 [==============================] - 0s 829us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 938/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 939/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 940/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 941/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 942/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 943/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 944/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 945/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 946/1024\n", + "90/90 [==============================] - 0s 962us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 947/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 948/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 949/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 950/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 951/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 952/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 953/1024\n", + "90/90 [==============================] - 0s 848us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 954/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 955/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 956/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 957/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 958/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 959/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 960/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 961/1024\n", + "90/90 [==============================] - 0s 891us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 962/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 963/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 964/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 965/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 966/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 967/1024\n", + "90/90 [==============================] - 0s 855us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 968/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 969/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 970/1024\n", + "90/90 [==============================] - 0s 835us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 971/1024\n", + "90/90 [==============================] - 0s 984us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 972/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 973/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 974/1024\n", + "90/90 [==============================] - 0s 832us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 975/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 976/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 977/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 978/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 979/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 980/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 981/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 982/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 983/1024\n", + "90/90 [==============================] - 0s 897us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 984/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 985/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 986/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 987/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 988/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 989/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 990/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 991/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 992/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 993/1024\n", + "90/90 [==============================] - 0s 849us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 994/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 995/1024\n", + "90/90 [==============================] - 0s 988us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 996/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 997/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 998/1024\n", + "90/90 [==============================] - 0s 857us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 999/1024\n", + "90/90 [==============================] - 0s 838us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1000/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1001/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1002/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1003/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1004/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1005/1024\n", + "90/90 [==============================] - 0s 843us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1006/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0030 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1007/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1008/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1009/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1010/1024\n", + "90/90 [==============================] - 0s 841us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1011/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1012/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1013/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1014/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1015/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1016/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1017/1024\n", + "90/90 [==============================] - 0s 853us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1018/1024\n", + "90/90 [==============================] - 0s 845us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1019/1024\n", + "90/90 [==============================] - 0s 840us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1020/1024\n", + "90/90 [==============================] - 0s 851us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1021/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1022/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1023/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n", + "Epoch 1024/1024\n", + "90/90 [==============================] - 0s 847us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.2760e-14\n" + ] + } + ], + "source": [ + "epoch, batch_size = 1024, 64\n", + "history_plain_11 = plain_11.fit(x_train, y_train, x_val, y_val, epoch=epoch, batch_size=batch_size)\n", + "history_plain_5 = plain_5.fit(x_train, y_train, x_val, y_val, epoch=epoch, batch_size=batch_size)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "Save the result to history.p" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "# import pickle\n", + "# with open('history.p', 'wb') as f:\n", + "# pickle.dump({'history_11': history_plain_11, 'history_5':history_plain_5}, f)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "pycharm": { + "name": "#%% md\n" + } + }, + "source": [ + "## Post analysis and drawing" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "# with open(\"history.p\", 'rb') as f:\n", + "# data = pickle.load(f)\n", + "# history_plain_11, history_plain_5 = data['history_11'], data['history_5']" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "smoothing_windows = 16\n", + "def moving_average(x, w):\n", + " return np.convolve(x, np.ones(w), 'valid') / w" + ] + }, + { + "cell_type": "code", + "execution_count": 20, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [ + { + "data": { + "text/plain": "
", + "image/png": "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\n" + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "%matplotlib inline\n", + "fig, ax = plt.subplots()\n", + "x = np.array(range(len(history_plain_11.history['loss']) - smoothing_windows + 1))\n", + "ax.plot(x, 100*moving_average(history_plain_11.history[\"val_loss\"], smoothing_windows), linewidth=3.0, label='plain11', c='b')\n", + "ax.plot(x, 100*moving_average(history_plain_11.history[\"loss\"], smoothing_windows), linewidth=.5, c='b')\n", + "x = np.array(range(len(history_plain_5.history['loss']) - smoothing_windows + 1))\n", + "ax.plot(x, 100*moving_average(history_plain_5.history[\"val_loss\"], smoothing_windows), linewidth=3.0, label='plain5', c='r')\n", + "ax.plot(x, 100*moving_average(history_plain_5.history[\"loss\"], smoothing_windows), linewidth=.5, c='r')\n", + "ax.set_xlabel(\"iter\")\n", + "ax.set_ylabel(\"error(%)\")\n", + "ax.set_xlim(0, min(len(history_plain_5.history['loss']), len(history_plain_11.history['loss']) - smoothing_windows + 1))\n", + "ax.set_ylim(0.20, 0.35)\n", + "ax.spines[\"top\"].set_visible(False)\n", + "ax.spines[\"right\"].set_visible(False)\n", + "ax.yaxis.set_major_locator(ticker.LinearLocator(numticks=6))\n", + "ax.grid(axis='y', linestyle='--')\n", + "ax.annotate(\"11-layer\", (540, 0.267), c='b')\n", + "ax.annotate(\"5-layer\", (540, 0.245), c='r')\n", + "plt.legend(loc=3)\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "# plt.savefig(fname=\"fig1.png\", dpi=300)" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "shortcut5, shortcut11 = ShortCut5(input_shape=(1, 102)), ShortCut11(input_shape=(1, 102))" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch 1/1024\n", + "90/90 [==============================] - 1s 3ms/step - loss: 0.0219 - val_loss: 0.0283 - lr: 0.0025\n", + "Epoch 2/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0196 - val_loss: 0.0270 - lr: 0.0025\n", + "Epoch 3/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0137 - val_loss: 0.0267 - lr: 0.0025\n", + "Epoch 4/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0089 - val_loss: 0.0280 - lr: 0.0025\n", + "Epoch 5/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0071 - val_loss: 0.0266 - lr: 0.0025\n", + "Epoch 6/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0059 - val_loss: 0.1138 - lr: 0.0025\n", + "Epoch 7/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0048 - val_loss: 0.1172 - lr: 0.0025\n", + "Epoch 8/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0047 - val_loss: 0.0186 - lr: 0.0025\n", + "Epoch 9/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0046 - val_loss: 0.2067 - lr: 0.0025\n", + "Epoch 10/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0044 - val_loss: 0.2280 - lr: 0.0025\n", + "Epoch 11/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0044 - val_loss: 0.2274 - lr: 0.0025\n", + "Epoch 12/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0043 - val_loss: 0.3251 - lr: 0.0025\n", + "Epoch 13/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0043 - val_loss: 0.2281 - lr: 0.0025\n", + "Epoch 14/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0043 - val_loss: 0.1210 - lr: 0.0025\n", + "Epoch 15/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0041 - val_loss: 0.0582 - lr: 0.0025\n", + "Epoch 16/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0043 - val_loss: 0.2277 - lr: 0.0025\n", + "Epoch 17/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0040 - val_loss: 0.2268 - lr: 0.0025\n", + "Epoch 18/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0044 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 19/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0040 - val_loss: 0.3142 - lr: 0.0025\n", + "Epoch 20/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0039 - val_loss: 0.0752 - lr: 0.0025\n", + "Epoch 21/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.0617 - lr: 0.0025\n", + "Epoch 22/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0040 - val_loss: 0.0920 - lr: 0.0025\n", + "Epoch 23/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0040 - val_loss: 0.3005 - lr: 0.0025\n", + "Epoch 24/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.2490 - lr: 0.0025\n", + "Epoch 25/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.0489 - lr: 0.0025\n", + "Epoch 26/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0039 - val_loss: 0.2725 - lr: 0.0025\n", + "Epoch 27/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0038 - val_loss: 0.2280 - lr: 0.0025\n", + "Epoch 28/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.3246 - lr: 0.0025\n", + "Epoch 29/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0040 - val_loss: 0.2171 - lr: 0.0025\n", + "Epoch 30/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.3298 - lr: 0.0025\n", + "Epoch 31/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0042 - val_loss: 0.2163 - lr: 0.0025\n", + "Epoch 32/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 33/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0039 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 34/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2266 - lr: 0.0012\n", + "Epoch 35/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.1462 - lr: 0.0012\n", + "Epoch 36/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.2875 - lr: 0.0012\n", + "Epoch 37/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.3281 - lr: 0.0012\n", + "Epoch 38/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2047 - lr: 0.0012\n", + "Epoch 39/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.2271 - lr: 0.0012\n", + "Epoch 40/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0597 - lr: 0.0012\n", + "Epoch 41/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.1329 - lr: 0.0012\n", + "Epoch 42/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.1875 - lr: 0.0012\n", + "Epoch 43/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2277 - lr: 0.0012\n", + "Epoch 44/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.1507 - lr: 0.0012\n", + "Epoch 45/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2337 - lr: 0.0012\n", + "Epoch 46/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.2683 - lr: 0.0012\n", + "Epoch 47/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2474 - lr: 0.0012\n", + "Epoch 48/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.1367 - lr: 0.0012\n", + "Epoch 49/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0035 - val_loss: 0.2283 - lr: 0.0012\n", + "Epoch 50/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0999 - lr: 0.0012\n", + "Epoch 51/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2481 - lr: 0.0012\n", + "Epoch 52/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0795 - lr: 0.0012\n", + "Epoch 53/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0034 - val_loss: 0.3165 - lr: 0.0012\n", + "Epoch 54/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0367 - lr: 0.0012\n", + "Epoch 55/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0336 - lr: 0.0012\n", + "Epoch 56/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.3071 - lr: 0.0012\n", + "Epoch 57/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0255 - lr: 0.0012\n", + "Epoch 58/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.2279 - lr: 0.0012\n", + "Epoch 59/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0216 - lr: 6.2500e-04\n", + "Epoch 60/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.2756 - lr: 6.2500e-04\n", + "Epoch 61/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2596 - lr: 6.2500e-04\n", + "Epoch 62/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.1604 - lr: 6.2500e-04\n", + "Epoch 63/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0691 - lr: 6.2500e-04\n", + "Epoch 64/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.2389 - lr: 6.2500e-04\n", + "Epoch 65/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2540 - lr: 6.2500e-04\n", + "Epoch 66/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.2570 - lr: 6.2500e-04\n", + "Epoch 67/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0667 - lr: 6.2500e-04\n", + "Epoch 68/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2504 - lr: 6.2500e-04\n", + "Epoch 69/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2053 - lr: 6.2500e-04\n", + "Epoch 70/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2035 - lr: 6.2500e-04\n", + "Epoch 71/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1357 - lr: 6.2500e-04\n", + "Epoch 72/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.1904 - lr: 6.2500e-04\n", + "Epoch 73/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.3200 - lr: 6.2500e-04\n", + "Epoch 74/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1841 - lr: 6.2500e-04\n", + "Epoch 75/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0314 - lr: 6.2500e-04\n", + "Epoch 76/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.1188 - lr: 6.2500e-04\n", + "Epoch 77/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.2098 - lr: 6.2500e-04\n", + "Epoch 78/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2265 - lr: 6.2500e-04\n", + "Epoch 79/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.1782 - lr: 6.2500e-04\n", + "Epoch 80/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1006 - lr: 6.2500e-04\n", + "Epoch 81/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2163 - lr: 6.2500e-04\n", + "Epoch 82/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0857 - lr: 6.2500e-04\n", + "Epoch 83/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2199 - lr: 6.2500e-04\n", + "Epoch 84/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.2611 - lr: 3.1250e-04\n", + "Epoch 85/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0029 - val_loss: 0.0573 - lr: 3.1250e-04\n", + "Epoch 86/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.1257 - lr: 3.1250e-04\n", + "Epoch 87/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2239 - lr: 3.1250e-04\n", + "Epoch 88/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1547 - lr: 3.1250e-04\n", + "Epoch 89/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2397 - lr: 3.1250e-04\n", + "Epoch 90/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0815 - lr: 3.1250e-04\n", + "Epoch 91/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0359 - lr: 3.1250e-04\n", + "Epoch 92/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1996 - lr: 3.1250e-04\n", + "Epoch 93/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0720 - lr: 3.1250e-04\n", + "Epoch 94/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.1951 - lr: 3.1250e-04\n", + "Epoch 95/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.1836 - lr: 3.1250e-04\n", + "Epoch 96/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0294 - lr: 3.1250e-04\n", + "Epoch 97/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0181 - lr: 3.1250e-04\n", + "Epoch 98/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.1766 - lr: 3.1250e-04\n", + "Epoch 99/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0591 - lr: 3.1250e-04\n", + "Epoch 100/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1220 - lr: 3.1250e-04\n", + "Epoch 101/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2105 - lr: 3.1250e-04\n", + "Epoch 102/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0453 - lr: 3.1250e-04\n", + "Epoch 103/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.1782 - lr: 3.1250e-04\n", + "Epoch 104/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0238 - lr: 3.1250e-04\n", + "Epoch 105/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2880 - lr: 3.1250e-04\n", + "Epoch 106/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2095 - lr: 3.1250e-04\n", + "Epoch 107/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.1675 - lr: 3.1250e-04\n", + "Epoch 108/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0127 - lr: 3.1250e-04\n", + "Epoch 109/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0940 - lr: 3.1250e-04\n", + "Epoch 110/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2401 - lr: 3.1250e-04\n", + "Epoch 111/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1006 - lr: 3.1250e-04\n", + "Epoch 112/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2042 - lr: 3.1250e-04\n", + "Epoch 113/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.3052 - lr: 3.1250e-04\n", + "Epoch 114/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2821 - lr: 3.1250e-04\n", + "Epoch 115/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0383 - lr: 3.1250e-04\n", + "Epoch 116/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1079 - lr: 3.1250e-04\n", + "Epoch 117/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1174 - lr: 3.1250e-04\n", + "Epoch 118/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2065 - lr: 3.1250e-04\n", + "Epoch 119/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.2121 - lr: 3.1250e-04\n", + "Epoch 120/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2459 - lr: 3.1250e-04\n", + "Epoch 121/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1985 - lr: 3.1250e-04\n", + "Epoch 122/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0073 - lr: 3.1250e-04\n", + "Epoch 123/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.2513 - lr: 3.1250e-04\n", + "Epoch 124/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0388 - lr: 3.1250e-04\n", + "Epoch 125/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1103 - lr: 3.1250e-04\n", + "Epoch 126/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0519 - lr: 3.1250e-04\n", + "Epoch 127/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.0992 - lr: 3.1250e-04\n", + "Epoch 128/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.1512 - lr: 3.1250e-04\n", + "Epoch 129/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0933 - lr: 3.1250e-04\n", + "Epoch 130/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.2168 - lr: 3.1250e-04\n", + "Epoch 131/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.2016 - lr: 3.1250e-04\n", + "Epoch 132/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1200 - lr: 3.1250e-04\n", + "Epoch 133/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0228 - lr: 3.1250e-04\n", + "Epoch 134/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0439 - lr: 3.1250e-04\n", + "Epoch 135/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2147 - lr: 3.1250e-04\n", + "Epoch 136/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1953 - lr: 3.1250e-04\n", + "Epoch 137/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0456 - lr: 3.1250e-04\n", + "Epoch 138/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2198 - lr: 3.1250e-04\n", + "Epoch 139/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0155 - lr: 3.1250e-04\n", + "Epoch 140/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0063 - lr: 3.1250e-04\n", + "Epoch 141/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.2162 - lr: 3.1250e-04\n", + "Epoch 142/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1195 - lr: 3.1250e-04\n", + "Epoch 143/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0230 - lr: 3.1250e-04\n", + "Epoch 144/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2873 - lr: 3.1250e-04\n", + "Epoch 145/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.1387 - lr: 3.1250e-04\n", + "Epoch 146/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0527 - lr: 3.1250e-04\n", + "Epoch 147/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1454 - lr: 3.1250e-04\n", + "Epoch 148/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0317 - lr: 3.1250e-04\n", + "Epoch 149/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1015 - lr: 3.1250e-04\n", + "Epoch 150/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1067 - lr: 3.1250e-04\n", + "Epoch 151/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2105 - lr: 3.1250e-04\n", + "Epoch 152/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.1155 - lr: 3.1250e-04\n", + "Epoch 153/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.2155 - lr: 3.1250e-04\n", + "Epoch 154/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1054 - lr: 3.1250e-04\n", + "Epoch 155/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0111 - lr: 3.1250e-04\n", + "Epoch 156/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2359 - lr: 3.1250e-04\n", + "Epoch 157/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2450 - lr: 3.1250e-04\n", + "Epoch 158/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1439 - lr: 3.1250e-04\n", + "Epoch 159/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0100 - lr: 3.1250e-04\n", + "Epoch 160/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0506 - lr: 3.1250e-04\n", + "Epoch 161/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.2382 - lr: 3.1250e-04\n", + "Epoch 162/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.1973 - lr: 3.1250e-04\n", + "Epoch 163/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2399 - lr: 3.1250e-04\n", + "Epoch 164/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2276 - lr: 3.1250e-04\n", + "Epoch 165/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2351 - lr: 3.1250e-04\n", + "Epoch 166/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0165 - lr: 1.5625e-04\n", + "Epoch 167/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0790 - lr: 1.5625e-04\n", + "Epoch 168/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0259 - lr: 1.5625e-04\n", + "Epoch 169/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0053 - lr: 1.5625e-04\n", + "Epoch 170/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.2109 - lr: 1.5625e-04\n", + "Epoch 171/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1137 - lr: 1.5625e-04\n", + "Epoch 172/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0062 - lr: 1.5625e-04\n", + "Epoch 173/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.1020 - lr: 1.5625e-04\n", + "Epoch 174/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0978 - lr: 1.5625e-04\n", + "Epoch 175/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0829 - lr: 1.5625e-04\n", + "Epoch 176/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0343 - lr: 1.5625e-04\n", + "Epoch 177/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0330 - lr: 1.5625e-04\n", + "Epoch 178/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0645 - lr: 1.5625e-04\n", + "Epoch 179/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0073 - lr: 1.5625e-04\n", + "Epoch 180/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0157 - lr: 1.5625e-04\n", + "Epoch 181/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0310 - lr: 1.5625e-04\n", + "Epoch 182/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0674 - lr: 1.5625e-04\n", + "Epoch 183/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0094 - lr: 1.5625e-04\n", + "Epoch 184/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1829 - lr: 1.5625e-04\n", + "Epoch 185/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0138 - lr: 1.5625e-04\n", + "Epoch 186/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1026 - lr: 1.5625e-04\n", + "Epoch 187/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1747 - lr: 1.5625e-04\n", + "Epoch 188/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.1049 - lr: 1.5625e-04\n", + "Epoch 189/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.1235 - lr: 1.5625e-04\n", + "Epoch 190/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0160 - lr: 1.5625e-04\n", + "Epoch 191/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1550 - lr: 1.5625e-04\n", + "Epoch 192/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.1186 - lr: 1.5625e-04\n", + "Epoch 193/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.1048 - lr: 1.5625e-04\n", + "Epoch 194/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1281 - lr: 1.5625e-04\n", + "Epoch 195/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0403 - lr: 7.8125e-05\n", + "Epoch 196/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0055 - lr: 7.8125e-05\n", + "Epoch 197/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0401 - lr: 7.8125e-05\n", + "Epoch 198/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0092 - lr: 7.8125e-05\n", + "Epoch 199/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0259 - lr: 7.8125e-05\n", + "Epoch 200/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0162 - lr: 7.8125e-05\n", + "Epoch 201/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.2130 - lr: 7.8125e-05\n", + "Epoch 202/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0112 - lr: 7.8125e-05\n", + "Epoch 203/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0100 - lr: 7.8125e-05\n", + "Epoch 204/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0776 - lr: 7.8125e-05\n", + "Epoch 205/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.1450 - lr: 7.8125e-05\n", + "Epoch 206/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0288 - lr: 7.8125e-05\n", + "Epoch 207/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0676 - lr: 7.8125e-05\n", + "Epoch 208/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0220 - lr: 7.8125e-05\n", + "Epoch 209/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0327 - lr: 7.8125e-05\n", + "Epoch 210/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0070 - lr: 7.8125e-05\n", + "Epoch 211/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0160 - lr: 7.8125e-05\n", + "Epoch 212/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0820 - lr: 7.8125e-05\n", + "Epoch 213/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0207 - lr: 7.8125e-05\n", + "Epoch 214/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0066 - lr: 7.8125e-05\n", + "Epoch 215/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0233 - lr: 7.8125e-05\n", + "Epoch 216/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0210 - lr: 7.8125e-05\n", + "Epoch 217/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0495 - lr: 7.8125e-05\n", + "Epoch 218/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0075 - lr: 7.8125e-05\n", + "Epoch 219/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0057 - lr: 7.8125e-05\n", + "Epoch 220/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0293 - lr: 3.9062e-05\n", + "Epoch 221/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0143 - lr: 3.9062e-05\n", + "Epoch 222/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0068 - lr: 3.9062e-05\n", + "Epoch 223/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0158 - lr: 3.9062e-05\n", + "Epoch 224/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0050 - lr: 3.9062e-05\n", + "Epoch 225/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0089 - lr: 3.9062e-05\n", + "Epoch 226/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0126 - lr: 3.9062e-05\n", + "Epoch 227/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0092 - lr: 3.9062e-05\n", + "Epoch 228/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0032 - lr: 3.9062e-05\n", + "Epoch 229/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0037 - lr: 3.9062e-05\n", + "Epoch 230/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0038 - lr: 3.9062e-05\n", + "Epoch 231/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0033 - lr: 3.9062e-05\n", + "Epoch 232/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0077 - lr: 3.9062e-05\n", + "Epoch 233/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0274 - lr: 3.9062e-05\n", + "Epoch 234/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0078 - lr: 3.9062e-05\n", + "Epoch 235/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0660 - lr: 3.9062e-05\n", + "Epoch 236/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0066 - lr: 3.9062e-05\n", + "Epoch 237/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0219 - lr: 3.9062e-05\n", + "Epoch 238/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0559 - lr: 3.9062e-05\n", + "Epoch 239/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0174 - lr: 3.9062e-05\n", + "Epoch 240/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0114 - lr: 3.9062e-05\n", + "Epoch 241/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0109 - lr: 3.9062e-05\n", + "Epoch 242/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0045 - lr: 3.9062e-05\n", + "Epoch 243/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0084 - lr: 3.9062e-05\n", + "Epoch 244/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0225 - lr: 3.9062e-05\n", + "Epoch 245/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0151 - lr: 3.9062e-05\n", + "Epoch 246/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0138 - lr: 3.9062e-05\n", + "Epoch 247/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0136 - lr: 3.9062e-05\n", + "Epoch 248/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0480 - lr: 3.9062e-05\n", + "Epoch 249/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0104 - lr: 3.9062e-05\n", + "Epoch 250/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0088 - lr: 3.9062e-05\n", + "Epoch 251/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0052 - lr: 3.9062e-05\n", + "Epoch 252/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0063 - lr: 3.9062e-05\n", + "Epoch 253/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0412 - lr: 3.9062e-05\n", + "Epoch 254/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0184 - lr: 1.9531e-05\n", + "Epoch 255/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0332 - lr: 1.9531e-05\n", + "Epoch 256/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0105 - lr: 1.9531e-05\n", + "Epoch 257/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0033 - lr: 1.9531e-05\n", + "Epoch 258/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0062 - lr: 1.9531e-05\n", + "Epoch 259/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0085 - lr: 1.9531e-05\n", + "Epoch 260/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 261/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0042 - lr: 1.9531e-05\n", + "Epoch 262/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0057 - lr: 1.9531e-05\n", + "Epoch 263/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0080 - lr: 1.9531e-05\n", + "Epoch 264/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0110 - lr: 1.9531e-05\n", + "Epoch 265/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 266/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0086 - lr: 1.9531e-05\n", + "Epoch 267/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 268/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0049 - lr: 1.9531e-05\n", + "Epoch 269/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0076 - lr: 1.9531e-05\n", + "Epoch 270/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0071 - lr: 1.9531e-05\n", + "Epoch 271/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0062 - lr: 1.9531e-05\n", + "Epoch 272/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0118 - lr: 1.9531e-05\n", + "Epoch 273/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0044 - lr: 1.9531e-05\n", + "Epoch 274/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0045 - lr: 1.9531e-05\n", + "Epoch 275/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0030 - lr: 1.9531e-05\n", + "Epoch 276/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 277/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0033 - lr: 1.9531e-05\n", + "Epoch 278/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0068 - lr: 1.9531e-05\n", + "Epoch 279/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 280/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0058 - lr: 1.9531e-05\n", + "Epoch 281/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0038 - lr: 1.9531e-05\n", + "Epoch 282/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0206 - lr: 1.9531e-05\n", + "Epoch 283/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0037 - lr: 1.9531e-05\n", + "Epoch 284/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0148 - lr: 1.9531e-05\n", + "Epoch 285/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0068 - lr: 1.9531e-05\n", + "Epoch 286/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 287/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 288/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 289/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 290/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 9.7656e-06\n", + "Epoch 291/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0039 - lr: 9.7656e-06\n", + "Epoch 292/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0045 - lr: 9.7656e-06\n", + "Epoch 293/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 294/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 295/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 296/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 297/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0026 - lr: 9.7656e-06\n", + "Epoch 298/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 299/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 300/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0035 - lr: 9.7656e-06\n", + "Epoch 301/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 302/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 303/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 304/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 305/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0029 - lr: 9.7656e-06\n", + "Epoch 306/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 307/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0045 - lr: 9.7656e-06\n", + "Epoch 308/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0035 - lr: 9.7656e-06\n", + "Epoch 309/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 310/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0041 - lr: 9.7656e-06\n", + "Epoch 311/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0072 - lr: 9.7656e-06\n", + "Epoch 312/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 313/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 314/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0045 - lr: 9.7656e-06\n", + "Epoch 315/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 316/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 317/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 318/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 319/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 320/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 321/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 322/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 323/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 324/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0035 - lr: 4.8828e-06\n", + "Epoch 325/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 326/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 327/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 328/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 329/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 330/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0031 - lr: 4.8828e-06\n", + "Epoch 331/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 332/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 333/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 334/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0033 - lr: 4.8828e-06\n", + "Epoch 335/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 336/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 337/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0031 - lr: 4.8828e-06\n", + "Epoch 338/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 339/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0031 - lr: 4.8828e-06\n", + "Epoch 340/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 341/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 342/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 343/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 344/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 345/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 346/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 347/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 348/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 349/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0040 - lr: 4.8828e-06\n", + "Epoch 350/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 351/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 352/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 353/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 354/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 355/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 356/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0034 - lr: 4.8828e-06\n", + "Epoch 357/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 358/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 359/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 360/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 361/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 362/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 363/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 364/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 365/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 366/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 367/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 368/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 369/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 370/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 371/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 372/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0029 - lr: 2.4414e-06\n", + "Epoch 373/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 374/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0028 - lr: 2.4414e-06\n", + "Epoch 375/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 376/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 377/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 378/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 379/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 380/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 381/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 382/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 383/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 384/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 385/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 386/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 387/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 388/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 389/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 390/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 391/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 392/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 393/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 394/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 395/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 396/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 397/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 398/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 399/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 400/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 401/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 402/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 403/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 404/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 405/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 406/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 407/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 408/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 409/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 410/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 411/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 412/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 413/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 414/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 415/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 416/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 417/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 418/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 419/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 420/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 421/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 422/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 423/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 424/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 425/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 426/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 427/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 428/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 429/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 430/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 431/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 432/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 433/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 434/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 435/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 436/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 437/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 438/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 439/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 440/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 441/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 442/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 443/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 444/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 445/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 446/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 447/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 448/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 449/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 450/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 451/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 452/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 453/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 454/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 455/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 456/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 457/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 458/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 459/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 460/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 461/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 462/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 463/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 464/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 465/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 466/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 467/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 468/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 469/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 470/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 471/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 472/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 473/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 474/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 475/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 476/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 477/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 478/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 479/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 480/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 481/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 482/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 483/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 484/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 485/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 486/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 487/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 488/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 489/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 490/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 491/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 492/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 493/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 494/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 495/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 496/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 497/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 498/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 499/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 500/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 501/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 502/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 503/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 504/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 505/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 506/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 507/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 508/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 509/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 510/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 511/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 512/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 513/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 514/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 515/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 516/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 517/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 518/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 519/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 520/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 521/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 522/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 523/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 524/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 525/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 526/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 527/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 528/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 529/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 530/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 531/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 532/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 533/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 534/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 535/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 536/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 537/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 538/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 539/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 540/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 541/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 542/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 543/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 544/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 545/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 546/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 547/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 548/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 549/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 550/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 551/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 552/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 553/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 554/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 555/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 556/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 557/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 558/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 559/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 560/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 561/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 562/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 563/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 564/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 565/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 566/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 567/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 568/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 569/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 570/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 571/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 572/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 573/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 574/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 575/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 576/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 577/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 578/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 579/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 580/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 581/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 582/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 583/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 584/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 585/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 586/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 587/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 588/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 589/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 590/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 591/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 592/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 593/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 594/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 595/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 596/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 597/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 598/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 599/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 600/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 601/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 602/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 603/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 604/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 605/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 606/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 607/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 608/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 609/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 610/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 611/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 612/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 613/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 614/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 615/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 616/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 617/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 618/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 619/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 620/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 621/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 622/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 623/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 624/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 625/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 626/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 627/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 628/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 629/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 630/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 631/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 632/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 633/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 634/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 635/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 636/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 637/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 638/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 639/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 640/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 641/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 642/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 643/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 644/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 645/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 646/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 647/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 648/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 649/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 650/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 651/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 652/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 653/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 1/1024\n", + "90/90 [==============================] - 1s 2ms/step - loss: 0.0198 - val_loss: 0.0598 - lr: 0.0025\n", + "Epoch 2/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0112 - val_loss: 0.0525 - lr: 0.0025\n", + "Epoch 3/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0071 - val_loss: 0.0330 - lr: 0.0025\n", + "Epoch 4/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0060 - val_loss: 0.0307 - lr: 0.0025\n", + "Epoch 5/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0052 - val_loss: 0.0281 - lr: 0.0025\n", + "Epoch 6/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0046 - val_loss: 0.0257 - lr: 0.0025\n", + "Epoch 7/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0045 - val_loss: 0.0653 - lr: 0.0025\n", + "Epoch 8/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0044 - val_loss: 0.0471 - lr: 0.0025\n", + "Epoch 9/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0044 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 10/1024\n", + "90/90 [==============================] - 0s 869us/step - loss: 0.0042 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 11/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0042 - val_loss: 0.1118 - lr: 0.0025\n", + "Epoch 12/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0042 - val_loss: 0.1824 - lr: 0.0025\n", + "Epoch 13/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0042 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 14/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0042 - val_loss: 0.0377 - lr: 0.0025\n", + "Epoch 15/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0040 - val_loss: 0.1167 - lr: 0.0025\n", + "Epoch 16/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0041 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 17/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0040 - val_loss: 0.2175 - lr: 0.0025\n", + "Epoch 18/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0041 - val_loss: 0.2283 - lr: 0.0025\n", + "Epoch 19/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0039 - val_loss: 0.0520 - lr: 0.0025\n", + "Epoch 20/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0039 - val_loss: 0.1469 - lr: 0.0025\n", + "Epoch 21/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0038 - val_loss: 0.0476 - lr: 0.0025\n", + "Epoch 22/1024\n", + "90/90 [==============================] - 0s 970us/step - loss: 0.0039 - val_loss: 0.1398 - lr: 0.0025\n", + "Epoch 23/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0040 - val_loss: 0.0864 - lr: 0.0025\n", + "Epoch 24/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0037 - val_loss: 0.2282 - lr: 0.0025\n", + "Epoch 25/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0038 - val_loss: 0.2278 - lr: 0.0025\n", + "Epoch 26/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0039 - val_loss: 0.1589 - lr: 0.0025\n", + "Epoch 27/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.0401 - lr: 0.0025\n", + "Epoch 28/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0037 - val_loss: 0.0640 - lr: 0.0025\n", + "Epoch 29/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0038 - val_loss: 0.3123 - lr: 0.0025\n", + "Epoch 30/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.1451 - lr: 0.0025\n", + "Epoch 31/1024\n", + "90/90 [==============================] - 0s 979us/step - loss: 0.0041 - val_loss: 0.2279 - lr: 0.0025\n", + "Epoch 32/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.0576 - lr: 0.0012\n", + "Epoch 33/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0036 - val_loss: 0.0103 - lr: 0.0012\n", + "Epoch 34/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0034 - val_loss: 0.1927 - lr: 0.0012\n", + "Epoch 35/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.1350 - lr: 0.0012\n", + "Epoch 36/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0037 - val_loss: 0.0297 - lr: 0.0012\n", + "Epoch 37/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.0853 - lr: 0.0012\n", + "Epoch 38/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.2278 - lr: 0.0012\n", + "Epoch 39/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2013 - lr: 0.0012\n", + "Epoch 40/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.2283 - lr: 0.0012\n", + "Epoch 41/1024\n", + "90/90 [==============================] - 0s 965us/step - loss: 0.0036 - val_loss: 0.0482 - lr: 0.0012\n", + "Epoch 42/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0035 - val_loss: 0.2170 - lr: 0.0012\n", + "Epoch 43/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.2271 - lr: 0.0012\n", + "Epoch 44/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.0587 - lr: 0.0012\n", + "Epoch 45/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0034 - val_loss: 0.1143 - lr: 0.0012\n", + "Epoch 46/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0034 - val_loss: 0.2379 - lr: 0.0012\n", + "Epoch 47/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0034 - val_loss: 0.1721 - lr: 0.0012\n", + "Epoch 48/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0033 - val_loss: 0.1126 - lr: 0.0012\n", + "Epoch 49/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0035 - val_loss: 0.0890 - lr: 0.0012\n", + "Epoch 50/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0033 - val_loss: 0.1434 - lr: 0.0012\n", + "Epoch 51/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.2669 - lr: 0.0012\n", + "Epoch 52/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0032 - val_loss: 0.0172 - lr: 0.0012\n", + "Epoch 53/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.2013 - lr: 0.0012\n", + "Epoch 54/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0034 - val_loss: 0.0816 - lr: 0.0012\n", + "Epoch 55/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0034 - val_loss: 0.2064 - lr: 0.0012\n", + "Epoch 56/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0034 - val_loss: 0.2262 - lr: 0.0012\n", + "Epoch 57/1024\n", + "90/90 [==============================] - 0s 959us/step - loss: 0.0033 - val_loss: 0.0168 - lr: 0.0012\n", + "Epoch 58/1024\n", + "90/90 [==============================] - 0s 897us/step - loss: 0.0034 - val_loss: 0.2105 - lr: 0.0012\n", + "Epoch 59/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1548 - lr: 6.2500e-04\n", + "Epoch 60/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0033 - val_loss: 0.0658 - lr: 6.2500e-04\n", + "Epoch 61/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0032 - val_loss: 0.0978 - lr: 6.2500e-04\n", + "Epoch 62/1024\n", + "90/90 [==============================] - 0s 960us/step - loss: 0.0033 - val_loss: 0.0730 - lr: 6.2500e-04\n", + "Epoch 63/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0031 - val_loss: 0.0170 - lr: 6.2500e-04\n", + "Epoch 64/1024\n", + "90/90 [==============================] - 0s 905us/step - loss: 0.0031 - val_loss: 0.2857 - lr: 6.2500e-04\n", + "Epoch 65/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0030 - val_loss: 0.2169 - lr: 6.2500e-04\n", + "Epoch 66/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0032 - val_loss: 0.2722 - lr: 6.2500e-04\n", + "Epoch 67/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2282 - lr: 6.2500e-04\n", + "Epoch 68/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2139 - lr: 6.2500e-04\n", + "Epoch 69/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0031 - val_loss: 0.0982 - lr: 6.2500e-04\n", + "Epoch 70/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2279 - lr: 6.2500e-04\n", + "Epoch 71/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0761 - lr: 6.2500e-04\n", + "Epoch 72/1024\n", + "90/90 [==============================] - 0s 978us/step - loss: 0.0031 - val_loss: 0.1284 - lr: 6.2500e-04\n", + "Epoch 73/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2601 - lr: 6.2500e-04\n", + "Epoch 74/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0350 - lr: 6.2500e-04\n", + "Epoch 75/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0031 - val_loss: 0.1648 - lr: 6.2500e-04\n", + "Epoch 76/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0031 - val_loss: 0.2015 - lr: 6.2500e-04\n", + "Epoch 77/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0031 - val_loss: 0.0225 - lr: 6.2500e-04\n", + "Epoch 78/1024\n", + "90/90 [==============================] - 0s 970us/step - loss: 0.0030 - val_loss: 0.2144 - lr: 6.2500e-04\n", + "Epoch 79/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0032 - val_loss: 0.0431 - lr: 6.2500e-04\n", + "Epoch 80/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0031 - val_loss: 0.0944 - lr: 6.2500e-04\n", + "Epoch 81/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0031 - val_loss: 0.1827 - lr: 6.2500e-04\n", + "Epoch 82/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0031 - val_loss: 0.1524 - lr: 6.2500e-04\n", + "Epoch 83/1024\n", + "90/90 [==============================] - 0s 905us/step - loss: 0.0030 - val_loss: 0.0833 - lr: 6.2500e-04\n", + "Epoch 84/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0031 - val_loss: 0.0131 - lr: 3.1250e-04\n", + "Epoch 85/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0030 - val_loss: 0.0984 - lr: 3.1250e-04\n", + "Epoch 86/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0031 - val_loss: 0.0298 - lr: 3.1250e-04\n", + "Epoch 87/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0029 - val_loss: 0.2283 - lr: 3.1250e-04\n", + "Epoch 88/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0030 - val_loss: 0.2270 - lr: 3.1250e-04\n", + "Epoch 89/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.1353 - lr: 3.1250e-04\n", + "Epoch 90/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.2282 - lr: 3.1250e-04\n", + "Epoch 91/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1818 - lr: 3.1250e-04\n", + "Epoch 92/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0031 - val_loss: 0.0185 - lr: 3.1250e-04\n", + "Epoch 93/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0030 - val_loss: 0.0603 - lr: 3.1250e-04\n", + "Epoch 94/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0085 - lr: 3.1250e-04\n", + "Epoch 95/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0055 - lr: 3.1250e-04\n", + "Epoch 96/1024\n", + "90/90 [==============================] - 0s 968us/step - loss: 0.0029 - val_loss: 0.1777 - lr: 3.1250e-04\n", + "Epoch 97/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2107 - lr: 3.1250e-04\n", + "Epoch 98/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0031 - val_loss: 0.2022 - lr: 3.1250e-04\n", + "Epoch 99/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0030 - val_loss: 0.0119 - lr: 3.1250e-04\n", + "Epoch 100/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0247 - lr: 3.1250e-04\n", + "Epoch 101/1024\n", + "90/90 [==============================] - 0s 962us/step - loss: 0.0030 - val_loss: 0.1989 - lr: 3.1250e-04\n", + "Epoch 102/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0322 - lr: 3.1250e-04\n", + "Epoch 103/1024\n", + "90/90 [==============================] - 0s 972us/step - loss: 0.0030 - val_loss: 0.2276 - lr: 3.1250e-04\n", + "Epoch 104/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0326 - lr: 3.1250e-04\n", + "Epoch 105/1024\n", + "90/90 [==============================] - 0s 1000us/step - loss: 0.0031 - val_loss: 0.2399 - lr: 3.1250e-04\n", + "Epoch 106/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0030 - val_loss: 0.0445 - lr: 3.1250e-04\n", + "Epoch 107/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0030 - val_loss: 0.2216 - lr: 3.1250e-04\n", + "Epoch 108/1024\n", + "90/90 [==============================] - 0s 913us/step - loss: 0.0030 - val_loss: 0.0257 - lr: 3.1250e-04\n", + "Epoch 109/1024\n", + "90/90 [==============================] - 0s 927us/step - loss: 0.0029 - val_loss: 0.1788 - lr: 3.1250e-04\n", + "Epoch 110/1024\n", + "90/90 [==============================] - 0s 961us/step - loss: 0.0029 - val_loss: 0.0195 - lr: 3.1250e-04\n", + "Epoch 111/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0308 - lr: 3.1250e-04\n", + "Epoch 112/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0169 - lr: 3.1250e-04\n", + "Epoch 113/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2607 - lr: 3.1250e-04\n", + "Epoch 114/1024\n", + "90/90 [==============================] - 0s 965us/step - loss: 0.0029 - val_loss: 0.2670 - lr: 3.1250e-04\n", + "Epoch 115/1024\n", + "90/90 [==============================] - 0s 985us/step - loss: 0.0030 - val_loss: 0.0159 - lr: 3.1250e-04\n", + "Epoch 116/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0030 - val_loss: 0.1088 - lr: 3.1250e-04\n", + "Epoch 117/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.2283 - lr: 3.1250e-04\n", + "Epoch 118/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.1138 - lr: 3.1250e-04\n", + "Epoch 119/1024\n", + "90/90 [==============================] - 0s 982us/step - loss: 0.0029 - val_loss: 0.0120 - lr: 3.1250e-04\n", + "Epoch 120/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0029 - val_loss: 0.1762 - lr: 3.1250e-04\n", + "Epoch 121/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0073 - lr: 1.5625e-04\n", + "Epoch 122/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0924 - lr: 1.5625e-04\n", + "Epoch 123/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0188 - lr: 1.5625e-04\n", + "Epoch 124/1024\n", + "90/90 [==============================] - 0s 972us/step - loss: 0.0030 - val_loss: 0.0138 - lr: 1.5625e-04\n", + "Epoch 125/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0028 - val_loss: 0.0270 - lr: 1.5625e-04\n", + "Epoch 126/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0028 - val_loss: 0.0375 - lr: 1.5625e-04\n", + "Epoch 127/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0031 - val_loss: 0.0750 - lr: 1.5625e-04\n", + "Epoch 128/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0109 - lr: 1.5625e-04\n", + "Epoch 129/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0467 - lr: 1.5625e-04\n", + "Epoch 130/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1844 - lr: 1.5625e-04\n", + "Epoch 131/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0444 - lr: 1.5625e-04\n", + "Epoch 132/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0101 - lr: 1.5625e-04\n", + "Epoch 133/1024\n", + "90/90 [==============================] - 0s 963us/step - loss: 0.0029 - val_loss: 0.1559 - lr: 1.5625e-04\n", + "Epoch 134/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.2269 - lr: 1.5625e-04\n", + "Epoch 135/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1741 - lr: 1.5625e-04\n", + "Epoch 136/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0029 - val_loss: 0.0634 - lr: 1.5625e-04\n", + "Epoch 137/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.1282 - lr: 1.5625e-04\n", + "Epoch 138/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1617 - lr: 1.5625e-04\n", + "Epoch 139/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0028 - val_loss: 0.1587 - lr: 1.5625e-04\n", + "Epoch 140/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0652 - lr: 1.5625e-04\n", + "Epoch 141/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.1194 - lr: 1.5625e-04\n", + "Epoch 142/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0069 - lr: 1.5625e-04\n", + "Epoch 143/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0030 - val_loss: 0.2232 - lr: 1.5625e-04\n", + "Epoch 144/1024\n", + "90/90 [==============================] - 0s 964us/step - loss: 0.0028 - val_loss: 0.0199 - lr: 1.5625e-04\n", + "Epoch 145/1024\n", + "90/90 [==============================] - 0s 964us/step - loss: 0.0029 - val_loss: 0.0634 - lr: 1.5625e-04\n", + "Epoch 146/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0029 - val_loss: 0.0338 - lr: 7.8125e-05\n", + "Epoch 147/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0029 - val_loss: 0.0403 - lr: 7.8125e-05\n", + "Epoch 148/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0028 - val_loss: 0.0079 - lr: 7.8125e-05\n", + "Epoch 149/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0402 - lr: 7.8125e-05\n", + "Epoch 150/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0441 - lr: 7.8125e-05\n", + "Epoch 151/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0071 - lr: 7.8125e-05\n", + "Epoch 152/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0029 - val_loss: 0.0945 - lr: 7.8125e-05\n", + "Epoch 153/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0028 - val_loss: 0.0241 - lr: 7.8125e-05\n", + "Epoch 154/1024\n", + "90/90 [==============================] - 0s 950us/step - loss: 0.0028 - val_loss: 0.0100 - lr: 7.8125e-05\n", + "Epoch 155/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0289 - lr: 7.8125e-05\n", + "Epoch 156/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0029 - val_loss: 0.0380 - lr: 7.8125e-05\n", + "Epoch 157/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0557 - lr: 7.8125e-05\n", + "Epoch 158/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0217 - lr: 7.8125e-05\n", + "Epoch 159/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0129 - lr: 7.8125e-05\n", + "Epoch 160/1024\n", + "90/90 [==============================] - 0s 936us/step - loss: 0.0028 - val_loss: 0.0220 - lr: 7.8125e-05\n", + "Epoch 161/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0950 - lr: 7.8125e-05\n", + "Epoch 162/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0053 - lr: 7.8125e-05\n", + "Epoch 163/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0113 - lr: 7.8125e-05\n", + "Epoch 164/1024\n", + "90/90 [==============================] - 0s 953us/step - loss: 0.0028 - val_loss: 0.0521 - lr: 7.8125e-05\n", + "Epoch 165/1024\n", + "90/90 [==============================] - 0s 993us/step - loss: 0.0029 - val_loss: 0.0089 - lr: 7.8125e-05\n", + "Epoch 166/1024\n", + "90/90 [==============================] - 0s 963us/step - loss: 0.0028 - val_loss: 0.0064 - lr: 7.8125e-05\n", + "Epoch 167/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0489 - lr: 7.8125e-05\n", + "Epoch 168/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0030 - val_loss: 0.0217 - lr: 7.8125e-05\n", + "Epoch 169/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0157 - lr: 7.8125e-05\n", + "Epoch 170/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0189 - lr: 7.8125e-05\n", + "Epoch 171/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0028 - val_loss: 0.0087 - lr: 7.8125e-05\n", + "Epoch 172/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.1240 - lr: 7.8125e-05\n", + "Epoch 173/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0183 - lr: 7.8125e-05\n", + "Epoch 174/1024\n", + "90/90 [==============================] - 0s 976us/step - loss: 0.0028 - val_loss: 0.1380 - lr: 7.8125e-05\n", + "Epoch 175/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0104 - lr: 7.8125e-05\n", + "Epoch 176/1024\n", + "90/90 [==============================] - 0s 984us/step - loss: 0.0028 - val_loss: 0.0084 - lr: 7.8125e-05\n", + "Epoch 177/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0049 - lr: 7.8125e-05\n", + "Epoch 178/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0705 - lr: 7.8125e-05\n", + "Epoch 179/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0030 - val_loss: 0.0130 - lr: 7.8125e-05\n", + "Epoch 180/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0066 - lr: 7.8125e-05\n", + "Epoch 181/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0028 - val_loss: 0.1058 - lr: 7.8125e-05\n", + "Epoch 182/1024\n", + "90/90 [==============================] - 0s 944us/step - loss: 0.0029 - val_loss: 0.0491 - lr: 7.8125e-05\n", + "Epoch 183/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0029 - val_loss: 0.0252 - lr: 7.8125e-05\n", + "Epoch 184/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0028 - val_loss: 0.0125 - lr: 7.8125e-05\n", + "Epoch 185/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0027 - val_loss: 0.0052 - lr: 7.8125e-05\n", + "Epoch 186/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0028 - val_loss: 0.0645 - lr: 7.8125e-05\n", + "Epoch 187/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0029 - val_loss: 0.0465 - lr: 7.8125e-05\n", + "Epoch 188/1024\n", + "90/90 [==============================] - 0s 968us/step - loss: 0.0027 - val_loss: 0.0429 - lr: 7.8125e-05\n", + "Epoch 189/1024\n", + "90/90 [==============================] - 0s 984us/step - loss: 0.0028 - val_loss: 0.0272 - lr: 7.8125e-05\n", + "Epoch 190/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0577 - lr: 7.8125e-05\n", + "Epoch 191/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0040 - lr: 7.8125e-05\n", + "Epoch 192/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0028 - val_loss: 0.0437 - lr: 7.8125e-05\n", + "Epoch 193/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0036 - lr: 7.8125e-05\n", + "Epoch 194/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0055 - lr: 7.8125e-05\n", + "Epoch 195/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0131 - lr: 7.8125e-05\n", + "Epoch 196/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0290 - lr: 7.8125e-05\n", + "Epoch 197/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0913 - lr: 7.8125e-05\n", + "Epoch 198/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0460 - lr: 7.8125e-05\n", + "Epoch 199/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0028 - val_loss: 0.1060 - lr: 7.8125e-05\n", + "Epoch 200/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0051 - lr: 7.8125e-05\n", + "Epoch 201/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0334 - lr: 7.8125e-05\n", + "Epoch 202/1024\n", + "90/90 [==============================] - 0s 990us/step - loss: 0.0028 - val_loss: 0.0753 - lr: 7.8125e-05\n", + "Epoch 203/1024\n", + "90/90 [==============================] - 0s 964us/step - loss: 0.0029 - val_loss: 0.0238 - lr: 7.8125e-05\n", + "Epoch 204/1024\n", + "90/90 [==============================] - 0s 953us/step - loss: 0.0029 - val_loss: 0.0136 - lr: 7.8125e-05\n", + "Epoch 205/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.2139 - lr: 7.8125e-05\n", + "Epoch 206/1024\n", + "90/90 [==============================] - 0s 968us/step - loss: 0.0028 - val_loss: 0.0143 - lr: 7.8125e-05\n", + "Epoch 207/1024\n", + "90/90 [==============================] - 0s 947us/step - loss: 0.0029 - val_loss: 0.1455 - lr: 7.8125e-05\n", + "Epoch 208/1024\n", + "90/90 [==============================] - 0s 964us/step - loss: 0.0028 - val_loss: 0.0175 - lr: 7.8125e-05\n", + "Epoch 209/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0451 - lr: 7.8125e-05\n", + "Epoch 210/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1155 - lr: 7.8125e-05\n", + "Epoch 211/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0095 - lr: 7.8125e-05\n", + "Epoch 212/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1050 - lr: 7.8125e-05\n", + "Epoch 213/1024\n", + "90/90 [==============================] - 0s 973us/step - loss: 0.0028 - val_loss: 0.0601 - lr: 7.8125e-05\n", + "Epoch 214/1024\n", + "90/90 [==============================] - 0s 957us/step - loss: 0.0028 - val_loss: 0.0070 - lr: 7.8125e-05\n", + "Epoch 215/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0028 - val_loss: 0.0444 - lr: 7.8125e-05\n", + "Epoch 216/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0029 - val_loss: 0.1996 - lr: 7.8125e-05\n", + "Epoch 217/1024\n", + "90/90 [==============================] - 0s 929us/step - loss: 0.0028 - val_loss: 0.0211 - lr: 7.8125e-05\n", + "Epoch 218/1024\n", + "90/90 [==============================] - 0s 953us/step - loss: 0.0029 - val_loss: 0.0245 - lr: 7.8125e-05\n", + "Epoch 219/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0029 - val_loss: 0.0044 - lr: 3.9062e-05\n", + "Epoch 220/1024\n", + "90/90 [==============================] - 0s 975us/step - loss: 0.0029 - val_loss: 0.0474 - lr: 3.9062e-05\n", + "Epoch 221/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0457 - lr: 3.9062e-05\n", + "Epoch 222/1024\n", + "90/90 [==============================] - 0s 978us/step - loss: 0.0029 - val_loss: 0.0045 - lr: 3.9062e-05\n", + "Epoch 223/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0230 - lr: 3.9062e-05\n", + "Epoch 224/1024\n", + "90/90 [==============================] - 0s 965us/step - loss: 0.0028 - val_loss: 0.0180 - lr: 3.9062e-05\n", + "Epoch 225/1024\n", + "90/90 [==============================] - 0s 947us/step - loss: 0.0028 - val_loss: 0.0337 - lr: 3.9062e-05\n", + "Epoch 226/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0289 - lr: 3.9062e-05\n", + "Epoch 227/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0036 - lr: 3.9062e-05\n", + "Epoch 228/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0044 - lr: 3.9062e-05\n", + "Epoch 229/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0188 - lr: 3.9062e-05\n", + "Epoch 230/1024\n", + "90/90 [==============================] - 0s 997us/step - loss: 0.0028 - val_loss: 0.0352 - lr: 3.9062e-05\n", + "Epoch 231/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0164 - lr: 3.9062e-05\n", + "Epoch 232/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0092 - lr: 3.9062e-05\n", + "Epoch 233/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0099 - lr: 3.9062e-05\n", + "Epoch 234/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0028 - val_loss: 0.0159 - lr: 3.9062e-05\n", + "Epoch 235/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0028 - val_loss: 0.0126 - lr: 3.9062e-05\n", + "Epoch 236/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0029 - val_loss: 0.0117 - lr: 3.9062e-05\n", + "Epoch 237/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0290 - lr: 3.9062e-05\n", + "Epoch 238/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0027 - val_loss: 0.0302 - lr: 3.9062e-05\n", + "Epoch 239/1024\n", + "90/90 [==============================] - 0s 987us/step - loss: 0.0028 - val_loss: 0.0401 - lr: 3.9062e-05\n", + "Epoch 240/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.1169 - lr: 3.9062e-05\n", + "Epoch 241/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0076 - lr: 3.9062e-05\n", + "Epoch 242/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0303 - lr: 3.9062e-05\n", + "Epoch 243/1024\n", + "90/90 [==============================] - 0s 987us/step - loss: 0.0028 - val_loss: 0.0102 - lr: 3.9062e-05\n", + "Epoch 244/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0545 - lr: 3.9062e-05\n", + "Epoch 245/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0034 - lr: 3.9062e-05\n", + "Epoch 246/1024\n", + "90/90 [==============================] - 0s 976us/step - loss: 0.0029 - val_loss: 0.0135 - lr: 3.9062e-05\n", + "Epoch 247/1024\n", + "90/90 [==============================] - 0s 971us/step - loss: 0.0028 - val_loss: 0.0045 - lr: 3.9062e-05\n", + "Epoch 248/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0028 - val_loss: 0.0153 - lr: 3.9062e-05\n", + "Epoch 249/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0092 - lr: 3.9062e-05\n", + "Epoch 250/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0030 - val_loss: 0.0083 - lr: 3.9062e-05\n", + "Epoch 251/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0028 - val_loss: 0.0221 - lr: 3.9062e-05\n", + "Epoch 252/1024\n", + "90/90 [==============================] - 0s 963us/step - loss: 0.0028 - val_loss: 0.0097 - lr: 3.9062e-05\n", + "Epoch 253/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0028 - val_loss: 0.0062 - lr: 3.9062e-05\n", + "Epoch 254/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0032 - lr: 3.9062e-05\n", + "Epoch 255/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0028 - val_loss: 0.0411 - lr: 3.9062e-05\n", + "Epoch 256/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0529 - lr: 3.9062e-05\n", + "Epoch 257/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0028 - val_loss: 0.0063 - lr: 3.9062e-05\n", + "Epoch 258/1024\n", + "90/90 [==============================] - 0s 973us/step - loss: 0.0028 - val_loss: 0.0036 - lr: 3.9062e-05\n", + "Epoch 259/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0308 - lr: 3.9062e-05\n", + "Epoch 260/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0085 - lr: 3.9062e-05\n", + "Epoch 261/1024\n", + "90/90 [==============================] - 0s 979us/step - loss: 0.0028 - val_loss: 0.0068 - lr: 3.9062e-05\n", + "Epoch 262/1024\n", + "90/90 [==============================] - 0s 982us/step - loss: 0.0028 - val_loss: 0.0074 - lr: 3.9062e-05\n", + "Epoch 263/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0028 - val_loss: 0.0079 - lr: 3.9062e-05\n", + "Epoch 264/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0109 - lr: 3.9062e-05\n", + "Epoch 265/1024\n", + "90/90 [==============================] - 0s 963us/step - loss: 0.0027 - val_loss: 0.0214 - lr: 3.9062e-05\n", + "Epoch 266/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0027 - val_loss: 0.0143 - lr: 3.9062e-05\n", + "Epoch 267/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0102 - lr: 3.9062e-05\n", + "Epoch 268/1024\n", + "90/90 [==============================] - 0s 991us/step - loss: 0.0028 - val_loss: 0.0262 - lr: 3.9062e-05\n", + "Epoch 269/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0028 - val_loss: 0.0217 - lr: 3.9062e-05\n", + "Epoch 270/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0027 - val_loss: 0.0221 - lr: 3.9062e-05\n", + "Epoch 271/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0056 - lr: 3.9062e-05\n", + "Epoch 272/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0146 - lr: 3.9062e-05\n", + "Epoch 273/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0307 - lr: 3.9062e-05\n", + "Epoch 274/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0197 - lr: 3.9062e-05\n", + "Epoch 275/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0028 - val_loss: 0.0057 - lr: 3.9062e-05\n", + "Epoch 276/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0027 - val_loss: 0.0204 - lr: 3.9062e-05\n", + "Epoch 277/1024\n", + "90/90 [==============================] - 0s 897us/step - loss: 0.0028 - val_loss: 0.0130 - lr: 3.9062e-05\n", + "Epoch 278/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0027 - val_loss: 0.0049 - lr: 3.9062e-05\n", + "Epoch 279/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0096 - lr: 3.9062e-05\n", + "Epoch 280/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0027 - val_loss: 0.0226 - lr: 1.9531e-05\n", + "Epoch 281/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 1.9531e-05\n", + "Epoch 282/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 283/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0027 - val_loss: 0.0043 - lr: 1.9531e-05\n", + "Epoch 284/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0044 - lr: 1.9531e-05\n", + "Epoch 285/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0151 - lr: 1.9531e-05\n", + "Epoch 286/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0028 - val_loss: 0.0076 - lr: 1.9531e-05\n", + "Epoch 287/1024\n", + "90/90 [==============================] - 0s 905us/step - loss: 0.0028 - val_loss: 0.0042 - lr: 1.9531e-05\n", + "Epoch 288/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0029 - val_loss: 0.0032 - lr: 1.9531e-05\n", + "Epoch 289/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0045 - lr: 1.9531e-05\n", + "Epoch 290/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0027 - val_loss: 0.0042 - lr: 1.9531e-05\n", + "Epoch 291/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0037 - lr: 1.9531e-05\n", + "Epoch 292/1024\n", + "90/90 [==============================] - 0s 946us/step - loss: 0.0028 - val_loss: 0.0048 - lr: 1.9531e-05\n", + "Epoch 293/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0063 - lr: 1.9531e-05\n", + "Epoch 294/1024\n", + "90/90 [==============================] - 0s 946us/step - loss: 0.0028 - val_loss: 0.0030 - lr: 1.9531e-05\n", + "Epoch 295/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 1.9531e-05\n", + "Epoch 296/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0126 - lr: 1.9531e-05\n", + "Epoch 297/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0034 - lr: 1.9531e-05\n", + "Epoch 298/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0035 - lr: 1.9531e-05\n", + "Epoch 299/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0027 - val_loss: 0.0056 - lr: 1.9531e-05\n", + "Epoch 300/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0028 - val_loss: 0.0107 - lr: 1.9531e-05\n", + "Epoch 301/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0028 - val_loss: 0.0060 - lr: 1.9531e-05\n", + "Epoch 302/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0028 - val_loss: 0.0133 - lr: 1.9531e-05\n", + "Epoch 303/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.0037 - lr: 1.9531e-05\n", + "Epoch 304/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0028 - val_loss: 0.0066 - lr: 1.9531e-05\n", + "Epoch 305/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0027 - val_loss: 0.0032 - lr: 1.9531e-05\n", + "Epoch 306/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0027 - val_loss: 0.0032 - lr: 1.9531e-05\n", + "Epoch 307/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0027 - val_loss: 0.0033 - lr: 1.9531e-05\n", + "Epoch 308/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0027 - val_loss: 0.0039 - lr: 1.9531e-05\n", + "Epoch 309/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0067 - lr: 1.9531e-05\n", + "Epoch 310/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0045 - lr: 1.9531e-05\n", + "Epoch 311/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0039 - lr: 1.9531e-05\n", + "Epoch 312/1024\n", + "90/90 [==============================] - 0s 978us/step - loss: 0.0028 - val_loss: 0.0043 - lr: 1.9531e-05\n", + "Epoch 313/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0028 - val_loss: 0.0040 - lr: 1.9531e-05\n", + "Epoch 314/1024\n", + "90/90 [==============================] - 0s 906us/step - loss: 0.0028 - val_loss: 0.0108 - lr: 1.9531e-05\n", + "Epoch 315/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.9531e-05\n", + "Epoch 316/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0028 - val_loss: 0.0041 - lr: 1.9531e-05\n", + "Epoch 317/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0028 - val_loss: 0.0030 - lr: 1.9531e-05\n", + "Epoch 318/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0052 - lr: 1.9531e-05\n", + "Epoch 319/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0112 - lr: 1.9531e-05\n", + "Epoch 320/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0205 - lr: 1.9531e-05\n", + "Epoch 321/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0102 - lr: 1.9531e-05\n", + "Epoch 322/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.0042 - lr: 1.9531e-05\n", + "Epoch 323/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0027 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 324/1024\n", + "90/90 [==============================] - 0s 947us/step - loss: 0.0028 - val_loss: 0.0031 - lr: 1.9531e-05\n", + "Epoch 325/1024\n", + "90/90 [==============================] - 0s 904us/step - loss: 0.0027 - val_loss: 0.0063 - lr: 1.9531e-05\n", + "Epoch 326/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0166 - lr: 1.9531e-05\n", + "Epoch 327/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0035 - lr: 1.9531e-05\n", + "Epoch 328/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0027 - val_loss: 0.0055 - lr: 1.9531e-05\n", + "Epoch 329/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0028 - val_loss: 0.0091 - lr: 1.9531e-05\n", + "Epoch 330/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0027 - val_loss: 0.0143 - lr: 1.9531e-05\n", + "Epoch 331/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0056 - lr: 1.9531e-05\n", + "Epoch 332/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0045 - lr: 1.9531e-05\n", + "Epoch 333/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0048 - lr: 1.9531e-05\n", + "Epoch 334/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0027 - val_loss: 0.0057 - lr: 1.9531e-05\n", + "Epoch 335/1024\n", + "90/90 [==============================] - 0s 863us/step - loss: 0.0028 - val_loss: 0.0142 - lr: 1.9531e-05\n", + "Epoch 336/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0028 - val_loss: 0.0052 - lr: 1.9531e-05\n", + "Epoch 337/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 1.9531e-05\n", + "Epoch 338/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0027 - val_loss: 0.0054 - lr: 1.9531e-05\n", + "Epoch 339/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0032 - lr: 1.9531e-05\n", + "Epoch 340/1024\n", + "90/90 [==============================] - 0s 901us/step - loss: 0.0027 - val_loss: 0.0034 - lr: 1.9531e-05\n", + "Epoch 341/1024\n", + "90/90 [==============================] - 0s 850us/step - loss: 0.0028 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 342/1024\n", + "90/90 [==============================] - 0s 854us/step - loss: 0.0028 - val_loss: 0.0034 - lr: 9.7656e-06\n", + "Epoch 343/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0029 - val_loss: 0.0063 - lr: 9.7656e-06\n", + "Epoch 344/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0028 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 345/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0037 - lr: 9.7656e-06\n", + "Epoch 346/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0050 - lr: 9.7656e-06\n", + "Epoch 347/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0042 - lr: 9.7656e-06\n", + "Epoch 348/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 349/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0034 - lr: 9.7656e-06\n", + "Epoch 350/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0027 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 351/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0028 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 352/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0051 - lr: 9.7656e-06\n", + "Epoch 353/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0027 - val_loss: 0.0028 - lr: 9.7656e-06\n", + "Epoch 354/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0027 - val_loss: 0.0040 - lr: 9.7656e-06\n", + "Epoch 355/1024\n", + "90/90 [==============================] - 0s 904us/step - loss: 0.0028 - val_loss: 0.0032 - lr: 9.7656e-06\n", + "Epoch 356/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 357/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0027 - val_loss: 0.0036 - lr: 9.7656e-06\n", + "Epoch 358/1024\n", + "90/90 [==============================] - 0s 893us/step - loss: 0.0028 - val_loss: 0.0048 - lr: 9.7656e-06\n", + "Epoch 359/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0027 - val_loss: 0.0062 - lr: 9.7656e-06\n", + "Epoch 360/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0027 - val_loss: 0.0030 - lr: 9.7656e-06\n", + "Epoch 361/1024\n", + "90/90 [==============================] - 0s 992us/step - loss: 0.0028 - val_loss: 0.0054 - lr: 9.7656e-06\n", + "Epoch 362/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0029 - val_loss: 0.0031 - lr: 9.7656e-06\n", + "Epoch 363/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 9.7656e-06\n", + "Epoch 364/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0028 - val_loss: 0.0034 - lr: 9.7656e-06\n", + "Epoch 365/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0026 - val_loss: 0.0033 - lr: 9.7656e-06\n", + "Epoch 366/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0041 - lr: 4.8828e-06\n", + "Epoch 367/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 368/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0035 - lr: 4.8828e-06\n", + "Epoch 369/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 370/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 371/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 372/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0028 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 373/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0027 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 374/1024\n", + "90/90 [==============================] - 0s 982us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 375/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0032 - lr: 4.8828e-06\n", + "Epoch 376/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 377/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 378/1024\n", + "90/90 [==============================] - 0s 897us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 379/1024\n", + "90/90 [==============================] - 0s 886us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 380/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 381/1024\n", + "90/90 [==============================] - 0s 893us/step - loss: 0.0027 - val_loss: 0.0041 - lr: 4.8828e-06\n", + "Epoch 382/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0028 - val_loss: 0.0033 - lr: 4.8828e-06\n", + "Epoch 383/1024\n", + "90/90 [==============================] - 0s 914us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 384/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0027 - val_loss: 0.0031 - lr: 4.8828e-06\n", + "Epoch 385/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 386/1024\n", + "90/90 [==============================] - 0s 905us/step - loss: 0.0029 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 387/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0029 - lr: 4.8828e-06\n", + "Epoch 388/1024\n", + "90/90 [==============================] - 0s 936us/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 389/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 390/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0031 - lr: 4.8828e-06\n", + "Epoch 391/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0038 - lr: 4.8828e-06\n", + "Epoch 392/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0028 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 393/1024\n", + "90/90 [==============================] - 0s 907us/step - loss: 0.0027 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 394/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.8828e-06\n", + "Epoch 395/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0027 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 396/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0033 - lr: 4.8828e-06\n", + "Epoch 397/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 398/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 399/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 400/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0027 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 401/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 402/1024\n", + "90/90 [==============================] - 0s 914us/step - loss: 0.0029 - val_loss: 0.0027 - lr: 4.8828e-06\n", + "Epoch 403/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0029 - val_loss: 0.0030 - lr: 4.8828e-06\n", + "Epoch 404/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0029 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 405/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0027 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 406/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 407/1024\n", + "90/90 [==============================] - 0s 897us/step - loss: 0.0027 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 408/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0033 - lr: 4.8828e-06\n", + "Epoch 409/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0028 - lr: 4.8828e-06\n", + "Epoch 410/1024\n", + "90/90 [==============================] - 0s 970us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 4.8828e-06\n", + "Epoch 411/1024\n", + "90/90 [==============================] - 0s 861us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 412/1024\n", + "90/90 [==============================] - 0s 907us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 413/1024\n", + "90/90 [==============================] - 0s 899us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 414/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 415/1024\n", + "90/90 [==============================] - 0s 867us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 416/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 417/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 418/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 419/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 420/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 421/1024\n", + "90/90 [==============================] - 0s 965us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 422/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 423/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 424/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 425/1024\n", + "90/90 [==============================] - 0s 981us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 426/1024\n", + "90/90 [==============================] - 0s 990us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 427/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 428/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 429/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 430/1024\n", + "90/90 [==============================] - 0s 986us/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 431/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 432/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 433/1024\n", + "90/90 [==============================] - 0s 982us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 434/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.4414e-06\n", + "Epoch 435/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 436/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0027 - lr: 2.4414e-06\n", + "Epoch 437/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 438/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0031 - lr: 2.4414e-06\n", + "Epoch 439/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0030 - val_loss: 0.0028 - lr: 2.4414e-06\n", + "Epoch 440/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 441/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 442/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 2.4414e-06\n", + "Epoch 443/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 444/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 445/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 446/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 447/1024\n", + "90/90 [==============================] - 0s 963us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 448/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 449/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 450/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 451/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 452/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 453/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 454/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 455/1024\n", + "90/90 [==============================] - 0s 997us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 456/1024\n", + "90/90 [==============================] - 0s 940us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 457/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 458/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 459/1024\n", + "90/90 [==============================] - 0s 958us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 460/1024\n", + "90/90 [==============================] - 0s 913us/step - loss: 0.0027 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 461/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 462/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 463/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 464/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 465/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 1.2207e-06\n", + "Epoch 466/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 467/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.2207e-06\n", + "Epoch 468/1024\n", + "90/90 [==============================] - 0s 878us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 469/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 470/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 471/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 472/1024\n", + "90/90 [==============================] - 0s 856us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 473/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 474/1024\n", + "90/90 [==============================] - 0s 898us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 475/1024\n", + "90/90 [==============================] - 0s 888us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 476/1024\n", + "90/90 [==============================] - 0s 884us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 477/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 478/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 479/1024\n", + "90/90 [==============================] - 0s 870us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 480/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 481/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 482/1024\n", + "90/90 [==============================] - 0s 959us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 483/1024\n", + "90/90 [==============================] - 0s 864us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 484/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 485/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 486/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 487/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 488/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 489/1024\n", + "90/90 [==============================] - 0s 986us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 490/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0026 - lr: 6.1035e-07\n", + "Epoch 491/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 492/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 493/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 494/1024\n", + "90/90 [==============================] - 0s 875us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 495/1024\n", + "90/90 [==============================] - 0s 872us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 6.1035e-07\n", + "Epoch 496/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 497/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 498/1024\n", + "90/90 [==============================] - 0s 962us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 499/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 500/1024\n", + "90/90 [==============================] - 0s 905us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 501/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 502/1024\n", + "90/90 [==============================] - 0s 984us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 503/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 504/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 505/1024\n", + "90/90 [==============================] - 0s 906us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 506/1024\n", + "90/90 [==============================] - 0s 935us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 507/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 508/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 509/1024\n", + "90/90 [==============================] - 0s 920us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 510/1024\n", + "90/90 [==============================] - 0s 913us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 511/1024\n", + "90/90 [==============================] - 0s 894us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 512/1024\n", + "90/90 [==============================] - 0s 933us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 513/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 514/1024\n", + "90/90 [==============================] - 0s 913us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 515/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 516/1024\n", + "90/90 [==============================] - 0s 962us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 517/1024\n", + "90/90 [==============================] - 0s 929us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 518/1024\n", + "90/90 [==============================] - 0s 962us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 519/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 520/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.0518e-07\n", + "Epoch 521/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 522/1024\n", + "90/90 [==============================] - 0s 914us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 523/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 524/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 525/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 526/1024\n", + "90/90 [==============================] - 0s 997us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 527/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 528/1024\n", + "90/90 [==============================] - 0s 992us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 529/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 530/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 531/1024\n", + "90/90 [==============================] - 0s 979us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 532/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 533/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 534/1024\n", + "90/90 [==============================] - 0s 946us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 535/1024\n", + "90/90 [==============================] - 0s 899us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 536/1024\n", + "90/90 [==============================] - 0s 947us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 537/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 538/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 539/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 540/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 541/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 542/1024\n", + "90/90 [==============================] - 0s 936us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 543/1024\n", + "90/90 [==============================] - 0s 987us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 544/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 545/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.5259e-07\n", + "Epoch 546/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 547/1024\n", + "90/90 [==============================] - 0s 862us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 548/1024\n", + "90/90 [==============================] - 0s 883us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 549/1024\n", + "90/90 [==============================] - 0s 859us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 550/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 551/1024\n", + "90/90 [==============================] - 0s 974us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 552/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 553/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 554/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 555/1024\n", + "90/90 [==============================] - 0s 968us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 556/1024\n", + "90/90 [==============================] - 0s 933us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 557/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 558/1024\n", + "90/90 [==============================] - 0s 893us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 559/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 560/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 561/1024\n", + "90/90 [==============================] - 0s 991us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 562/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 563/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 564/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 565/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 566/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 567/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 568/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 569/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 570/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.6294e-08\n", + "Epoch 571/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 572/1024\n", + "90/90 [==============================] - 0s 962us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 573/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 574/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 575/1024\n", + "90/90 [==============================] - 0s 984us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 576/1024\n", + "90/90 [==============================] - 0s 975us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 577/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 578/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 579/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 580/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 581/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 582/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 583/1024\n", + "90/90 [==============================] - 0s 981us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 584/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 585/1024\n", + "90/90 [==============================] - 0s 871us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 586/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 587/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 588/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 589/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 590/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 591/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 592/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 593/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 594/1024\n", + "90/90 [==============================] - 0s 960us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 595/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.8147e-08\n", + "Epoch 596/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 597/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 598/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 599/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 600/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 601/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 602/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 603/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 604/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 605/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 606/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 607/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 608/1024\n", + "90/90 [==============================] - 0s 994us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 609/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 610/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 611/1024\n", + "90/90 [==============================] - 0s 992us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 612/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 613/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 614/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 615/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 616/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 617/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 618/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 619/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 620/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.9073e-08\n", + "Epoch 621/1024\n", + "90/90 [==============================] - 0s 953us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 622/1024\n", + "90/90 [==============================] - 0s 960us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 623/1024\n", + "90/90 [==============================] - 0s 989us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 624/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 625/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 626/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 627/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 628/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 629/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 630/1024\n", + "90/90 [==============================] - 0s 993us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 631/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 632/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 633/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 634/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 635/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 636/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 637/1024\n", + "90/90 [==============================] - 0s 954us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 638/1024\n", + "90/90 [==============================] - 0s 957us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 639/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 640/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 641/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 642/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 643/1024\n", + "90/90 [==============================] - 0s 904us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 644/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 645/1024\n", + "90/90 [==============================] - 0s 944us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.5367e-09\n", + "Epoch 646/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 647/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 648/1024\n", + "90/90 [==============================] - 0s 996us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 649/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 650/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 651/1024\n", + "90/90 [==============================] - 0s 987us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 652/1024\n", + "90/90 [==============================] - 0s 991us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 653/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 654/1024\n", + "90/90 [==============================] - 0s 927us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 655/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 656/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 657/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 658/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 659/1024\n", + "90/90 [==============================] - 0s 874us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 660/1024\n", + "90/90 [==============================] - 0s 929us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 661/1024\n", + "90/90 [==============================] - 0s 960us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 662/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 663/1024\n", + "90/90 [==============================] - 0s 927us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 664/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 665/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 666/1024\n", + "90/90 [==============================] - 0s 995us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 667/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 668/1024\n", + "90/90 [==============================] - 0s 944us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 669/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 670/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.7684e-09\n", + "Epoch 671/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 672/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 673/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 674/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 675/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 676/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 677/1024\n", + "90/90 [==============================] - 0s 940us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 678/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 679/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 680/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 681/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 682/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 683/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 684/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 685/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 686/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 687/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 688/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 689/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 690/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 691/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 692/1024\n", + "90/90 [==============================] - 0s 967us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 693/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 694/1024\n", + "90/90 [==============================] - 0s 1000us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 695/1024\n", + "90/90 [==============================] - 0s 975us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3842e-09\n", + "Epoch 696/1024\n", + "90/90 [==============================] - 0s 947us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 697/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 698/1024\n", + "90/90 [==============================] - 0s 976us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 699/1024\n", + "90/90 [==============================] - 0s 961us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 700/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 701/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 702/1024\n", + "90/90 [==============================] - 0s 944us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 703/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 704/1024\n", + "90/90 [==============================] - 0s 905us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 705/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 706/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 707/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 708/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 709/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 710/1024\n", + "90/90 [==============================] - 0s 958us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 711/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 712/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 713/1024\n", + "90/90 [==============================] - 0s 985us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 714/1024\n", + "90/90 [==============================] - 0s 959us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 715/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 716/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 717/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 718/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 719/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 720/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1921e-09\n", + "Epoch 721/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 722/1024\n", + "90/90 [==============================] - 0s 951us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 723/1024\n", + "90/90 [==============================] - 0s 929us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 724/1024\n", + "90/90 [==============================] - 0s 910us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 725/1024\n", + "90/90 [==============================] - 0s 972us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 726/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 727/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 728/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 729/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 730/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 731/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 732/1024\n", + "90/90 [==============================] - 0s 957us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 733/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 734/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 735/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 736/1024\n", + "90/90 [==============================] - 0s 920us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 737/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 738/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 739/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 740/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 741/1024\n", + "90/90 [==============================] - 0s 915us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 742/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 743/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 744/1024\n", + "90/90 [==============================] - 0s 940us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 745/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.9605e-10\n", + "Epoch 746/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 747/1024\n", + "90/90 [==============================] - 0s 953us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 748/1024\n", + "90/90 [==============================] - 0s 940us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 749/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 750/1024\n", + "90/90 [==============================] - 0s 973us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 751/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 752/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 753/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 754/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 755/1024\n", + "90/90 [==============================] - 0s 4ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 756/1024\n", + "90/90 [==============================] - 0s 4ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 757/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 758/1024\n", + "90/90 [==============================] - 0s 4ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 759/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 760/1024\n", + "90/90 [==============================] - 0s 929us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 761/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 762/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 763/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 764/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 765/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 766/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 767/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 768/1024\n", + "90/90 [==============================] - 0s 944us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 769/1024\n", + "90/90 [==============================] - 0s 948us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 770/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9802e-10\n", + "Epoch 771/1024\n", + "90/90 [==============================] - 0s 986us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 772/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 773/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 774/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 775/1024\n", + "90/90 [==============================] - 0s 936us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 776/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 777/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 778/1024\n", + "90/90 [==============================] - 0s 920us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 779/1024\n", + "90/90 [==============================] - 0s 956us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 780/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 781/1024\n", + "90/90 [==============================] - 0s 892us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 782/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 783/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 784/1024\n", + "90/90 [==============================] - 0s 991us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 785/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 786/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 787/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 788/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 789/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 790/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 791/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 792/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 793/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 794/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 795/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4901e-10\n", + "Epoch 796/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 797/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 798/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 799/1024\n", + "90/90 [==============================] - 0s 933us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 800/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 801/1024\n", + "90/90 [==============================] - 0s 954us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 802/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 803/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 804/1024\n", + "90/90 [==============================] - 0s 927us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 805/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 806/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 807/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 808/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 809/1024\n", + "90/90 [==============================] - 0s 903us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 810/1024\n", + "90/90 [==============================] - 0s 977us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 811/1024\n", + "90/90 [==============================] - 0s 956us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 812/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 813/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 814/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 815/1024\n", + "90/90 [==============================] - 0s 956us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 816/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 817/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 818/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 819/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 820/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 7.4506e-11\n", + "Epoch 821/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 822/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 823/1024\n", + "90/90 [==============================] - 0s 914us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 824/1024\n", + "90/90 [==============================] - 0s 973us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 825/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 826/1024\n", + "90/90 [==============================] - 0s 947us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 827/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 828/1024\n", + "90/90 [==============================] - 0s 971us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 829/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 830/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 831/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 832/1024\n", + "90/90 [==============================] - 0s 967us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 833/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 834/1024\n", + "90/90 [==============================] - 0s 936us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 835/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 836/1024\n", + "90/90 [==============================] - 0s 950us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 837/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 838/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 839/1024\n", + "90/90 [==============================] - 0s 920us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 840/1024\n", + "90/90 [==============================] - 0s 926us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 841/1024\n", + "90/90 [==============================] - 0s 928us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 842/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 843/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 844/1024\n", + "90/90 [==============================] - 0s 920us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 845/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 3.7253e-11\n", + "Epoch 846/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 847/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 848/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 849/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 850/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 851/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 852/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 853/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 854/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 855/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 856/1024\n", + "90/90 [==============================] - 0s 929us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 857/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 858/1024\n", + "90/90 [==============================] - 0s 945us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 859/1024\n", + "90/90 [==============================] - 0s 938us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 860/1024\n", + "90/90 [==============================] - 0s 922us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 861/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 862/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 863/1024\n", + "90/90 [==============================] - 0s 969us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 864/1024\n", + "90/90 [==============================] - 0s 944us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 865/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 866/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 867/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 868/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 869/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 870/1024\n", + "90/90 [==============================] - 0s 953us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.8626e-11\n", + "Epoch 871/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 872/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 873/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 874/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 875/1024\n", + "90/90 [==============================] - 0s 915us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 876/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 877/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 878/1024\n", + "90/90 [==============================] - 0s 957us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 879/1024\n", + "90/90 [==============================] - 0s 912us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 880/1024\n", + "90/90 [==============================] - 0s 907us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 881/1024\n", + "90/90 [==============================] - 0s 935us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 882/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 883/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 884/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 885/1024\n", + "90/90 [==============================] - 0s 904us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 886/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 887/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 888/1024\n", + "90/90 [==============================] - 0s 968us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 889/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 890/1024\n", + "90/90 [==============================] - 0s 940us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 891/1024\n", + "90/90 [==============================] - 0s 967us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 892/1024\n", + "90/90 [==============================] - 0s 949us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 893/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 894/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 895/1024\n", + "90/90 [==============================] - 0s 924us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 9.3132e-12\n", + "Epoch 896/1024\n", + "90/90 [==============================] - 0s 915us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 897/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 898/1024\n", + "90/90 [==============================] - 0s 955us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 899/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 900/1024\n", + "90/90 [==============================] - 0s 954us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 901/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 902/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 903/1024\n", + "90/90 [==============================] - 0s 3ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 904/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 905/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0026 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 906/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 907/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 908/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 909/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 910/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 911/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 912/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 913/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 914/1024\n", + "90/90 [==============================] - 0s 950us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 915/1024\n", + "90/90 [==============================] - 0s 936us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 916/1024\n", + "90/90 [==============================] - 0s 990us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 917/1024\n", + "90/90 [==============================] - 0s 921us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 918/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 919/1024\n", + "90/90 [==============================] - 0s 967us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 920/1024\n", + "90/90 [==============================] - 0s 925us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 4.6566e-12\n", + "Epoch 921/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 922/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 923/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 924/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 925/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 926/1024\n", + "90/90 [==============================] - 0s 906us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 927/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 928/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 929/1024\n", + "90/90 [==============================] - 0s 989us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 930/1024\n", + "90/90 [==============================] - 0s 900us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 931/1024\n", + "90/90 [==============================] - 0s 902us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 932/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 933/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 934/1024\n", + "90/90 [==============================] - 0s 990us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 935/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 936/1024\n", + "90/90 [==============================] - 0s 987us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 937/1024\n", + "90/90 [==============================] - 0s 911us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 938/1024\n", + "90/90 [==============================] - 0s 889us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 939/1024\n", + "90/90 [==============================] - 0s 896us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 940/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 941/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 942/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 943/1024\n", + "90/90 [==============================] - 0s 899us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 944/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 945/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.3283e-12\n", + "Epoch 946/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 947/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 948/1024\n", + "90/90 [==============================] - 0s 880us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 949/1024\n", + "90/90 [==============================] - 0s 884us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 950/1024\n", + "90/90 [==============================] - 0s 915us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 951/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 952/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 953/1024\n", + "90/90 [==============================] - 0s 982us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 954/1024\n", + "90/90 [==============================] - 0s 941us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 955/1024\n", + "90/90 [==============================] - 0s 994us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 956/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 957/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 958/1024\n", + "90/90 [==============================] - 0s 973us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 959/1024\n", + "90/90 [==============================] - 0s 943us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 960/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 961/1024\n", + "90/90 [==============================] - 0s 965us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 962/1024\n", + "90/90 [==============================] - 0s 957us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 963/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 964/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 965/1024\n", + "90/90 [==============================] - 0s 973us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 966/1024\n", + "90/90 [==============================] - 0s 917us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 967/1024\n", + "90/90 [==============================] - 0s 884us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 968/1024\n", + "90/90 [==============================] - 0s 881us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 969/1024\n", + "90/90 [==============================] - 0s 887us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 970/1024\n", + "90/90 [==============================] - 0s 918us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.1642e-12\n", + "Epoch 971/1024\n", + "90/90 [==============================] - 0s 858us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 972/1024\n", + "90/90 [==============================] - 0s 939us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 973/1024\n", + "90/90 [==============================] - 0s 884us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 974/1024\n", + "90/90 [==============================] - 0s 879us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 975/1024\n", + "90/90 [==============================] - 0s 852us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 976/1024\n", + "90/90 [==============================] - 0s 976us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 977/1024\n", + "90/90 [==============================] - 0s 868us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 978/1024\n", + "90/90 [==============================] - 0s 890us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 979/1024\n", + "90/90 [==============================] - 0s 860us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 980/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 981/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 982/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 983/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 984/1024\n", + "90/90 [==============================] - 0s 919us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 985/1024\n", + "90/90 [==============================] - 0s 909us/step - loss: 0.0026 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 986/1024\n", + "90/90 [==============================] - 0s 882us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 987/1024\n", + "90/90 [==============================] - 0s 873us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 988/1024\n", + "90/90 [==============================] - 0s 923us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 989/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0029 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 990/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 991/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 992/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 993/1024\n", + "90/90 [==============================] - 0s 952us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 994/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 995/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 5.8208e-13\n", + "Epoch 996/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 997/1024\n", + "90/90 [==============================] - 0s 899us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 998/1024\n", + "90/90 [==============================] - 0s 930us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 999/1024\n", + "90/90 [==============================] - 0s 998us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1000/1024\n", + "90/90 [==============================] - 0s 956us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1001/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1002/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1003/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1004/1024\n", + "90/90 [==============================] - 0s 914us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1005/1024\n", + "90/90 [==============================] - 0s 934us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1006/1024\n", + "90/90 [==============================] - 0s 932us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1007/1024\n", + "90/90 [==============================] - 0s 992us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1008/1024\n", + "90/90 [==============================] - 0s 942us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1009/1024\n", + "90/90 [==============================] - 0s 963us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1010/1024\n", + "90/90 [==============================] - 0s 916us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1011/1024\n", + "90/90 [==============================] - 0s 966us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1012/1024\n", + "90/90 [==============================] - 0s 908us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1013/1024\n", + "90/90 [==============================] - 0s 865us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1014/1024\n", + "90/90 [==============================] - 0s 876us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1015/1024\n", + "90/90 [==============================] - 0s 895us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1016/1024\n", + "90/90 [==============================] - 0s 885us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1017/1024\n", + "90/90 [==============================] - 0s 866us/step - loss: 0.0029 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1018/1024\n", + "90/90 [==============================] - 0s 937us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1019/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0028 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1020/1024\n", + "90/90 [==============================] - 0s 2ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 2.9104e-13\n", + "Epoch 1021/1024\n", + "90/90 [==============================] - 0s 1ms/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1022/1024\n", + "90/90 [==============================] - 0s 931us/step - loss: 0.0028 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1023/1024\n", + "90/90 [==============================] - 0s 901us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4552e-13\n", + "Epoch 1024/1024\n", + "90/90 [==============================] - 0s 877us/step - loss: 0.0027 - val_loss: 0.0025 - lr: 1.4552e-13\n" + ] + } + ], + "source": [ + "epoch, batch_size = 1024, 64\n", + "history_shortcut_11 = shortcut11.fit(x_train, y_train, x_val, y_val, epoch=epoch, batch_size=batch_size)\n", + "history_shortcut_5 = shortcut5.fit(x_train, y_train, x_val, y_val, epoch=epoch, batch_size=batch_size)" + ] + }, + { + "cell_type": "code", + "execution_count": 21, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [ + { + "data": { + "text/plain": "
", + "image/png": "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\n" + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "fig, ax = plt.subplots()\n", + "x = np.array(range(len(history_shortcut_11.history['loss']) - smoothing_windows + 1))\n", + "ax.plot(x, 100*moving_average(history_shortcut_11.history[\"val_loss\"], smoothing_windows), linewidth=3.0, label='SCCNN11', c='b')\n", + "ax.plot(x, 100*moving_average(history_shortcut_11.history[\"loss\"], smoothing_windows), linewidth=.5, c='b')\n", + "x = np.array(range(len(history_shortcut_5.history['loss']) - smoothing_windows + 1))\n", + "ax.plot(x, 100*moving_average(history_shortcut_5.history[\"val_loss\"], smoothing_windows), linewidth=3.0, label='SCCNN5', c='r')\n", + "ax.plot(x, 100*moving_average(history_shortcut_5.history[\"loss\"], smoothing_windows), linewidth=.5, c='r')\n", + "ax.set_xlabel(\"iter\")\n", + "ax.set_ylabel(\"error(%)\")\n", + "ax.set_xlim(0, min(len(history_shortcut_5.history['loss']), len(history_shortcut_11.history['loss']) - smoothing_windows + 1))\n", + "ax.set_ylim(0.20, 0.35)\n", + "ax.spines[\"top\"].set_visible(False)\n", + "ax.spines[\"right\"].set_visible(False)\n", + "ax.yaxis.set_major_locator(ticker.LinearLocator(numticks=6))\n", + "ax.grid(axis='y', linestyle='--')\n", + "ax.annotate(\"11-layer\", (500, 0.24), c='b')\n", + "ax.annotate(\"5-layer\", (500, 0.255), c='r')\n", + "plt.legend(loc=3)\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "# plt.savefig(fname=\"fig2.png\", dpi=300)" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "import os\n", + "time_spent = (time.time() - time1) /60\n" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "interpreter": { + "hash": "7f619fc91ee8bdab81d49e7c14228037474662e3f2d607687ae505108922fa06" + }, + "kernelspec": { + "display_name": "Python 3.9.7 64-bit ('base': conda)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.7" + } + }, + "nbformat": 4, + "nbformat_minor": 1 +} \ No newline at end of file diff --git a/models.py b/models.py new file mode 100644 index 0000000..2dae6dd --- /dev/null +++ b/models.py @@ -0,0 +1,264 @@ +import keras.callbacks +import keras.layers as KL +from keras import Model +from keras.optimizers import adam_v2 + + +class Plain5(object): + def __init__(self, model_path=None, input_shape=None): + self.model = None + self.input_shape = input_shape + if model_path is not None: + # TODO: loading from the file + pass + else: + self.model = self.build_model() + + def build_model(self): + input_layer = KL.Input(self.input_shape, name='input') + x = KL.Conv1D(8, 3, padding='same', name='Conv1')(input_layer) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + + x = KL.Conv1D(8, 3, padding='same', name='Conv2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + + x = KL.Conv1D(8, 3, padding='same', name='Conv3')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + + x = KL.Dense(20, activation='relu', name='dense')(x) + x = KL.Dense(1, activation='sigmoid', name='output')(x) + model = Model(input_layer, x) + return model + + def fit(self, x, y, x_val, y_val, epoch, batch_size): + self.model.compile(loss='mse', optimizer=adam_v2.Adam(learning_rate=0.01 * (batch_size / 256))) + checkpoint = keras.callbacks.ModelCheckpoint(filepath='checkpoints/plain5.hdf5', monitor='val_loss', + mode="min", save_best_only=True) + early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', min_delta=0, + patience=1000, verbose=0, mode='auto') + lr_decay = keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=25, min_delta=1e-6) + callbacks = [checkpoint, early_stop, lr_decay] + history = self.model.fit(x, y, validation_data=(x_val, y_val), epochs=epoch, verbose=1, + callbacks=callbacks, batch_size=batch_size) + return history + + +class Residual5(object): + def __init__(self, model_path=None, input_shape=None): + self.model = None + self.input_shape = input_shape + if model_path is not None: + # TODO: loading from the file + pass + else: + self.model = self.build_model() + + def build_model(self): + input_layer = KL.Input(self.input_shape, name='input') + fx = KL.Conv1D(8, 3, padding='same', name='Conv1')(input_layer) + fx = KL.BatchNormalization()(fx) + x = KL.Activation('relu')(fx) + + fx = KL.Conv1D(8, 3, padding='same', name='Conv2')(x) + fx = KL.BatchNormalization()(fx) + fx = KL.Activation('relu')(fx) + x = fx + x + + fx = KL.Conv1D(8, 3, padding='same', name='Conv3')(x) + fx = KL.BatchNormalization()(fx) + fx = KL.Activation('relu')(fx) + x = fx + x + + x = KL.Dense(20, activation='relu', name='dense')(x) + x = KL.Dense(1, activation='sigmoid', name='output')(x) + model = Model(input_layer, x) + return model + + def fit(self, x, y, x_val, y_val, epoch, batch_size): + self.model.compile(loss='mse', optimizer=adam_v2.Adam(learning_rate=0.01 * (batch_size / 256))) + checkpoint = keras.callbacks.ModelCheckpoint(filepath='checkpoints/res5.hdf5', monitor='val_loss', + mode="min", save_best_only=True) + early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', min_delta=0, + patience=1000, verbose=0, mode='auto') + lr_decay = keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=25, min_delta=1e-6) + callbacks = [checkpoint, early_stop, lr_decay] + history = self.model.fit(x, y, validation_data=(x_val, y_val), epochs=epoch, verbose=1, + callbacks=callbacks, batch_size=batch_size) + return history + + +class ShortCut5(object): + def __init__(self, model_path=None, input_shape=None): + self.model = None + self.input_shape = input_shape + if model_path is not None: + # TODO: loading from the file + pass + else: + self.model = self.build_model() + + def build_model(self): + input_layer = KL.Input(self.input_shape, name='input') + x_raw = KL.Conv1D(8, 3, padding='same', name='Conv1')(input_layer) + fx1 = KL.BatchNormalization()(x_raw) + fx1 = KL.Activation('relu')(fx1) + + fx2 = KL.Conv1D(8, 3, padding='same', name='Conv2')(fx1) + fx2 = KL.BatchNormalization()(fx2) + fx2 = KL.Activation('relu')(fx2) + + fx3 = KL.Conv1D(8, 3, padding='same', name='Conv3')(fx2) + fx3 = KL.BatchNormalization()(fx3) + fx3 = KL.Activation('relu')(fx3) + x = KL.Concatenate(axis=2)([x_raw, fx1, fx2, fx3]) + + x = KL.Dense(20, activation='relu', name='dense')(x) + x = KL.Dense(1, activation='sigmoid', name='output')(x) + model = Model(input_layer, x) + return model + + def fit(self, x, y, x_val, y_val, epoch, batch_size): + self.model.compile(loss='mse', optimizer=adam_v2.Adam(learning_rate=0.01 * (batch_size / 256))) + + checkpoint = keras.callbacks.ModelCheckpoint(filepath='checkpoints/shortcut5.hdf5', monitor='val_loss', + mode="min", save_best_only=True) + early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', min_delta=0, + patience=1000, verbose=0, mode='auto') + lr_decay = keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=25, min_delta=1e-6) + callbacks = [checkpoint, early_stop, lr_decay] + history = self.model.fit(x, y, validation_data=(x_val, y_val), epochs=epoch, verbose=1, + callbacks=callbacks, batch_size=batch_size) + return history + + +class ShortCut11(object): + def __init__(self, model_path=None, input_shape=None): + self.model = None + self.input_shape = input_shape + if model_path is not None: + # TODO: loading from the file + pass + else: + self.model = self.build_model() + + def build_model(self): + input_layer = KL.Input(self.input_shape, name='input') + x_raw = KL.Conv1D(8, 3, padding='same', name='Conv1_1')(input_layer) + x = KL.BatchNormalization()(x_raw) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv1_2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv1_3')(x) + x = KL.BatchNormalization()(x) + fx1 = KL.Activation('relu')(x) + + x = KL.Conv1D(8, 3, padding='same', name='Conv2_1')(fx1) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv2_2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv2_3')(x) + x = KL.BatchNormalization()(x) + fx2 = KL.Activation('relu')(x) + + x = KL.Conv1D(8, 3, padding='same', name='Conv3_1')(fx2) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv3_2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv3_3')(x) + x = KL.BatchNormalization()(x) + fx3 = KL.Activation('relu')(x) + x = KL.Concatenate(axis=2)([x_raw, fx1, fx2, fx3]) + + x = KL.Dense(200, activation='relu', name='dense1')(x) + x = KL.Dense(1, activation='sigmoid', name='output')(x) + model = Model(input_layer, x) + return model + + def fit(self, x, y, x_val, y_val, epoch, batch_size): + self.model.compile(loss='mse', optimizer=adam_v2.Adam(learning_rate=0.01 * (batch_size / 256))) + checkpoint = keras.callbacks.ModelCheckpoint(filepath='checkpoints/shortcut11.hdf5', monitor='val_loss', + mode="min", save_best_only=True) + early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', min_delta=1e-6, + patience=200, verbose=0, mode='auto') + lr_decay = keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, + patience=25, min_delta=1e-6) + callbacks = [checkpoint, early_stop, lr_decay] + history = self.model.fit(x, y, validation_data=(x_val, y_val), epochs=epoch, verbose=1, + callbacks=callbacks, batch_size=batch_size) + return history + + +class Plain11(object): + def __init__(self, model_path=None, input_shape=None): + self.model = None + self.input_shape = input_shape + if model_path is not None: + # TODO: loading from the file + pass + else: + self.model = self.build_model() + + def build_model(self): + input_layer = KL.Input(self.input_shape, name='input') + x = KL.Conv1D(8, 3, padding='same', name='Conv1_1')(input_layer) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv1_2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv1_3')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + + x = KL.Conv1D(8, 3, padding='same', name='Conv2_1')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv2_2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv2_3')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + + x = KL.Conv1D(8, 3, padding='same', name='Conv3_1')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv3_2')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + x = KL.Conv1D(8, 3, padding='same', name='Conv3_3')(x) + x = KL.BatchNormalization()(x) + x = KL.Activation('relu')(x) + + x = KL.Dense(200, activation='relu', name='dense1')(x) + x = KL.Dense(1, activation='sigmoid', name='output')(x) + model = Model(input_layer, x) + return model + + def fit(self, x, y, x_val, y_val, epoch, batch_size): + self.model.compile(loss='mse', optimizer=adam_v2.Adam(learning_rate=0.01 * (batch_size / 256))) + checkpoint = keras.callbacks.ModelCheckpoint(filepath='checkpoints/plain11.hdf5', monitor='val_loss', + mode="min", save_best_only=True) + early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', min_delta=1e-6, + patience=200, verbose=0, mode='auto') + lr_decay = keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.5, + patience=25, min_delta=1e-6) + callbacks = [checkpoint, early_stop, lr_decay] + history = self.model.fit(x, y, validation_data=(x_val, y_val), epochs=epoch, verbose=1, + callbacks=callbacks, batch_size=batch_size) + return history + + +if __name__ == '__main__': + # plain5 = Plain5(model_path=None, input_shape=(1, 102)) + # plain11 = Plain11(model_path=None, input_shape=(1, 102)) + residual5 = Residual5(model_path=None, input_shape=(1, 102)) + short5 = ShortCut5(model_path=None, input_shape=(1, 102)) diff --git a/preprocess.ipynb b/preprocess.ipynb new file mode 100644 index 0000000..e87add3 --- /dev/null +++ b/preprocess.ipynb @@ -0,0 +1,127 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "dd2c8c55", + "metadata": {}, + "source": [ + "# Preprocessing" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "716880ac", + "metadata": {}, + "outputs": [], + "source": [ + "import pandas as pd\n", + "from sklearn.model_selection import train_test_split\n", + "from scipy.io import savemat, loadmat\n", + "import os" + ] + }, + { + "cell_type": "markdown", + "id": "4d7dc4a0", + "metadata": {}, + "source": [ + "## Step 1: \n", + "Convert the dataset to mat format for Matlab." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "711356a2", + "metadata": {}, + "outputs": [], + "source": [ + "dataset = pd.read_csv('preprocess/dataset/mango/NAnderson2020MendeleyMangoNIRData.csv')\n", + "y = dataset.DM\n", + "x = dataset.loc[:, '684': '990']\n", + "savemat('preprocess/dataset/mango/mango_origin.mat', {'x': x.values, 'y': y.values})" + ] + }, + { + "cell_type": "markdown", + "id": "3e41e8e6", + "metadata": {}, + "source": [] + }, + { + "cell_type": "markdown", + "id": "ea5e54fd", + "metadata": {}, + "source": [ + "## Step3:\n", + "Data split with train test split." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "6eac026e", + "metadata": {}, + "outputs": [], + "source": [ + "data = loadmat('preprocess/dataset/mango/mango_preprocessed.mat')\n", + "x, y = data['x'], data['y']\n", + "x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.3, random_state=24)\n", + "if not os.path.exists('mango'):\n", + " os.makedirs('mango')\n", + "savemat('preprocess/dataset/mango/mango_dm_split.mat',{'x_train':x_train, 'y_train':y_train, 'x_test':x_test, 'y_test':y_test,\n", + " 'max_y': data['max_y'], 'min_y': data['min_y'],\n", + " 'min_x':data['min_x'], 'max_x':data['max_x']})" + ] + }, + { + "cell_type": "markdown", + "id": "b2977dae", + "metadata": {}, + "source": [ + "## Step 4:\n", + "Show data with pictures\n", + "use `draw_pics_origin` to draw original spectra\n", + "![img](./preprocess/pics/raw.png)" + ] + }, + { + "cell_type": "markdown", + "source": [ + "use `draw_pics_preprocessed.m` to draw proprecessed spectra\n", + "![img](./preprocess/pics/preprocessed.png)" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + } + ], + "metadata": { + "interpreter": { + "hash": "7f619fc91ee8bdab81d49e7c14228037474662e3f2d607687ae505108922fa06" + }, + "kernelspec": { + "display_name": "Python 3.9.7 64-bit ('base': conda)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.7" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} \ No newline at end of file diff --git a/preprocess/draw_pics_origin.m b/preprocess/draw_pics_origin.m new file mode 100755 index 0000000..0e65e53 --- /dev/null +++ b/preprocess/draw_pics_origin.m @@ -0,0 +1,45 @@ +set(gca,'LooseInset',get(gca,'TightInset')) +f = figure; +f.Position(3:4) = [1331 331]; +%%% draw the pic of corn spectra +load('dataset/corn.mat'); +x = m5spec.data; +wave_length = m5spec.axisscale{2, 1}; +subplot(1, 4, 1) +plot(wave_length, x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Absorbance'); +clear + +%%% draw the pic of Marzipan spectra +load('dataset/marzipan.mat'); +x = NIRS1; +wave_length = NIRS1_axis; +subplot(1, 4, 2) +plot(wave_length, x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Absorbance'); +clear + +%%% draw the pic of Marzipan spectra +load('dataset/soil.mat'); +x = soil.data; +wave_length = soil.axisscale{2, 1}; +subplot(1, 4, 3) +plot(wave_length, x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Absorbance'); +clear + +% draw the pic of Mango spectra +load('dataset/mango/mango_origin.mat'); +wave_length = 684: 3: 990; +subplot(1, 4, 4) +plot(wave_length, x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Signal intensity'); +clear \ No newline at end of file diff --git a/preprocess/draw_pics_preprocessed.m b/preprocess/draw_pics_preprocessed.m new file mode 100755 index 0000000..271b9fb --- /dev/null +++ b/preprocess/draw_pics_preprocessed.m @@ -0,0 +1,48 @@ +set(gca,'LooseInset',get(gca,'TightInset')) +f = figure; +f.Position(3:4) = [1331 331]; +%%% draw the pic of corn spectra +load('dataset/corn.mat'); +x = m5spec.data; +wave_length = m5spec.axisscale{2, 1}; +preprocess; +subplot(1, 4, 1) +plot(wave_length(1, 1:end-1), x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Absorbance'); +clear + +%%% draw the pic of Marzipan spectra +load('dataset/marzipan.mat'); +x = NIRS1; +wave_length = NIRS1_axis; +preprocess; +subplot(1, 4, 2) +plot(wave_length(1, 1:end-1), x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Absorbance'); +clear + +%%% draw the pic of Marzipan spectra +load('dataset/soil.mat'); +x = soil.data; +wave_length = soil.axisscale{2, 1}; +preprocess; +subplot(1, 4, 3) +plot(wave_length(1, 1:end-1), x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Absorbance'); +clear + +% draw the pic of Mango spectra +load('dataset/mango/mango_preprocessed.mat'); +wave_length = 687: 3: 990; +subplot(1, 4, 4) +plot(wave_length, x'); +xlim([wave_length(1) wave_length(end)]); +xlabel('Wavelength(nm)'); +ylabel('Signal intensity'); +clear \ No newline at end of file diff --git a/preprocess/pics/preprocessed.png b/preprocess/pics/preprocessed.png new file mode 100644 index 0000000000000000000000000000000000000000..516150da7e705a97381114b26a62feda317ab3ed GIT binary patch literal 91504 zcmeEuWmJ@5*X}4Ff&vZ*NF&|dsWO0ccXxM5D~dEoHz?gmGjvFobT`b<-8J8X?|Rqz zex0+{`F)-R%wh(3?%8qO*S+_(2~m`n#6TxR2Z2BsQXe78AkZ^85a>zi^QXWwT>CjO zzyq4&M=ci+=mpN>=Mzv$+8f|WR97iEaa7n-d{kt<6Bvp)@D!D+xTdR$gSo4Rk+T^{ z?!Bt)`wvXys+MNtaz?I}8V=4jE==T7_9iSsnv0Krg5vS##hlHIT&*1J$yKfF%s?ER z%o1kC%p7bS9OUfW0&H9Y+&ow+QSU$?a*!0{z3Qj5y+u#WSc^HHqZc1%IR9qhq7fIIk|qnZ7;aagjWHc3;G4m#=nfomf~qZZ$d8P=mYBMu>te zrKpaE{5lv zt*@<-dhfC4eFe4Ezl%VU^`vw8T!_Jq2~tVw?CpI%PGD%gv$K=J>!7cuhCBWqWT^YU zZWa_$UXfi=QRQPa_$=>hn`wYsSlj2@!`9a7>bULKM_(6~cb&V3#s+KINldY2JLLMc7L*m3wZjdJ1bVA^7W(ZU7I7pCJ@RIGhHe z(q<*wrW}8|?wFy?%Hoae0p}YfD;RCLJOP1Z?X&DD$H9hP$FVOZVHh0gK-}l$0l&{W z=YA$xF?Y$-JVylHnv#YT!mM+>whsm=Ur1@4> z)7}}@Acv)^y12E3M*83_cX6jOIR#s=+kp)&Us}Z;4+IU?b(;>pH0avXcqN%R7c|v3 zARwtv;s~C(K$-HsQuk7p+;gV5DHgGbxaiQ66qX;QeTzBel(G|0_~iC45)#eg@0;q} z*hJhCpEz0>(yj*adV4~@PK8bt^S=kpi75}uA)@pd|M6A}9HIKR+HSsfn39$AzwR18 zi8G=oSZGjj`E+~4`h6(qo>%a0jGEb5s#*EhPras@n-Gc!?vbOae1 zTU{K}p~Kn#c{~#Ec)b5&-*{Iaci65jpGa}SmSBV&WiG1f5We)eTY`xUi-_JzQdycm zLf4)|Bk3mQR+#L;?R!mC()J1)_}2&uXjvQ^mnvj~6ZnHIK?U=FT#=LXH=gPGlYl_o zgU2a)(+PsMN^t#iEGep%v@VC%Dd&&T6@~E6^r0K2{6ystx8^9dtP08Z*7>#S^$;?z?r@PNXU6#@5bgF@l~H>ac{*2 zQ>X1nZ;O@(#`dX$mFNN}sdXHAa)tnpjuGJ0^!563>sLwB@hBCnC! zxE>nGUBVlRD+_t3Va|(|?*M*+lIr=?0~^&=;k0G~&w*qnFCw6aFw5oxmXR81xxXu0 z-WaxLzYYKRp{er(P$XR?}v&R60@|99RC_?l{$VzD7)#U^LTb z$$uKbQ!8#>=8U`AoD^E->PX{OEs7`FKhle67PsW@|8Df?4L@!l_Hi%}rCGtGEEWO)D9L)o>v*W!FM9r|HNT)% z(ew2VPTWyWEKwLF8&>^p;B_1wU6nff9ZGlze26NqDliF}%A1Ru+uz0@=Tj6iQ6ZSZ z;6FAJIe#Dmv#ZP81;;IraGw853YJ9Cmkm6B9y1KMlB~~e^_^oYLvvtZ$7%T(sFg1Y zg=vlWRFZW)&SrxJu7BICs==$;g@-=eGdh$j)d)T#Q7PdR@#M}8^SvE(3`7;I%x!5| z92u!%4M4Qd8{1g0kJBRKpaAze68>YCJ2fYervhrL8^?H_9Qk<1qxG%J=8e7@S8u6v zGjPp)F%DxxnhNEfTv_;$xmx48wJ+ljt6E=Fphg&Ar@00PvD5TgVzXj}nMJs;RT?T8dvI zcy27dOJm<|7melfMG0wy6U%1=wy5<(%{L=A$D-DP3m@+5NDz-F-^l~PV z^o!z$9tP5nB$flwfg98J-Z{Hb*K)F%II^#S?WD~EO-NwO#?c(WFOV#g^Ve_NKvc46 zZC^H@*>9a_mgUf;AquX?X_;%9Uy!I{9-uo_FY&u>7iPFMAUP$=3Fx9!!L0&`IU#PMJK!@ge$_(rX(U0wEPfIoA!b-&?*KoT|uD_#3b9wnqjOz3b?JHA)A zxEnORiTugyh4dOw=y&|%p9tvkpTykK^Uzj*{k!{9QWO4Hqp4HTG!_2V?`MTl*RpoA z-R@d4a7uok6+iI>6NgnM;J@ABDIye165&sJbp+=F)&F{R_>5?r;4xdkUWJgw`+oq? zC)s+KD}r>p=8UPQEuovTzi}k~&&|$Rb-IaYubY*W-lI4XoqL(IDRup*2@T${U?$JP z9l#mb`Pxray;pC%WEYx^Hlizt7_D-Xk1Dbp`;_BbPC%8m>d5rZS1%Xs-P)*Ys*xHlie|m?(nltR38mL(4nvCZfPDk zTFvVyDcu2!V2tkx{#etNqU$lClhhCe1c$(W8On(#j~Ubsf{k09KVZE7hedCkWQzG| zo5w1&hU1Y7Jl?u)C$r4#lw{6xSjdGDyzJOvb;Yx3Y!k9TUbx#Z^xAAgJ>w(|-?|Ws z;9#7kT8&u~TA6w8a%=JxU0*oPf`6;&OYq%Ux3+I(_j|#{_q#E8*YE8Xs*3LZfJX7X z%-+RE24^dwhTRbes~HkLLX4iYo{75Bi?TW(^`_nGOKAmq{&5E`0M<@UM0KKH!zb+v zU5|!rCmMzZ{QqWApfWhhmTyW1+eiz;1G^FDE>S_s>P+D;u{3WDq6|*Mf6L(o)f0*ObNuC`pQCdGmA~JcKYE{ zP~?!O`EbCz#?qE&Wkyo0_}P~^pN&uBchB~}#653~c~(~qHF%l0oZXQ2hMYk|$+FG~ z%hR;5q3VxF-N8$vfc6K#7Wa;5o*+=Y{DOLlby@o}Iv_WjpA`)QOHkb*JWbf;fW9B& zwE1HHjt=cX7G|2jo70DZ4Snl%{LasIb_Cz{imu8ibXz7(MH744D6)xTAV`#RUt87BCr&_VzmwD8&fT>;EF zN&d-rSpEKo6e!F#>f1IsfQ+Lx53BMBR9)pE?)OLN`z>y$C>{eC zqd?$&i1*G(&D4BTwTdcYVQ+df$5z+9Kplo86y@d^#)`hR57LJp3@f{J_I;l3DDFfrW+Gckmw=9@ek696eZUOn&j47eGzt zgds#`wM7!<)4_KU3V5RexB&O8{E&F}X|CUG$|Fw48r>+ViScZ_0pw}!Z_N`caQDSL$e}^NY zL6%dV76pgw0v@1LR8TNeYc~%pLs&?N)NAY4%*;%#Oyb<(%~2x(&wYEr935o#+UMvk3}OzR1-CV}_LRk1w#-g0Ciy9x!v+imp36$}YQeJX$C+Qg0ErSV^?!MPL8DP< z6W+kL^@2r%`Xj*Z&A;(FSPd7uNEf+D>uc0*ny2Wx10iSB*B(>wv*G3Rl70rtttoeGrh@Qs9(tb6a<)&o7%P&GQw2WE6o?MpmF4)7oAv|H2#lAwp zmX|?^4mYZX${&EG9o|qNqNAhBS*}_STW(qOn?8*QUv6w}y1BVk%9;} zuUpP4VPmLidM+x8MX|%p$C+++GTY+Ycz>I=O`aV@h=3IHVOEOMPRNF&60H;vO2f zYrb3>!oFR|DgntR!MG#qoZkfQ4&a7JU|Qm<3tYim@kIC#=gpz>OR&;dt&!p72LX6g zU)X8Jn_|HHLmQR)UKGc}&7>(iCLv%nZ``_ji^MW4R*QS{B@ZF@%0zx2#>d#CrlG;Z z!*h3cPm#@25WVN3r!Q=6Mf65fHa9n;prBxhTp4O>r}UXJFftDG_Ztf^g=6RYZP@S{ z-u&+ZO5Myk&2(<$EOi-s|1~t~7{fn!*{h!F;OFW09?=;NjT%b7o_@Ee7VhtIERD&! zQ)JE^+FpMEUw!pc|6WbUo_m2B>DoT=RRlj=S^eMT?p`ZPgM%JetWnOiV1~ajZF`5@s#N_-+SbJ4E?m*$M2~SuJXBt zVj7C4r{~)!V+}c7`U8JxmYL2xN$uVaW;5dVaZBfEi{fTuGpx}T78Y(jXQ8E~wSYU! zPD@cl8@}>dv)QBbcHe&pB30COt?i_CY83Sf+=j$X<~)3tj8JeyQiJR3YQ)?_1i)s( z>_OXnj?9u{Q4bt&+(d_v(8*@6IvXgES#4T z2&to{q3KP#^@h(BZlSuQjt;-i5>24$$2w~Ei_nUpMoOMl`9w;Wmzn;5D^(481n`)J z?^V|>TZ~q{{TP4^%~54xVL`$PUzUQyK0S?U|I18}U&;qyY4>Vy_7n_Ws|hpx9FCtO zC~U{8rNrHn?Mx1~5bCzNzZlO` zeZ7EP4cOMlBjxq;mM@jgd*#Cy;kSr|P!uz&Vck@0?`1(6H}iWgp} zCfFRW=%tn{EpWTfhl&L|@|@FWX`X?>!vw8&oVWw5!ZLf5htUfejd7!{q5<;yQ)jT+ z77;fL1zEXvJ9d56RE^}IEQ8ErR?n)OE2LqcMX3xXt%hCZjPNf8WK;Ktc?pU73u2xk zWmqz?1Fe=Tp$!plwo}yql^$Z8cC{rnKcE)lPj|TD0cS+I0C79~kPnR!z)2XCM7$Sv zOCb<;nUV`Q`t<nu0f{5`4Xu?`rAvXdGWt!MAt6WN8+|c|`g+@Sd<(aM#GDC|{i8Qv`YwN_ zzl-;v@xGOWcR-n?CH**^gTlF-Hf9E2;z=8_+2~v_agzKor7r))s8!9%uezk+C)v52 zFI0xO$*v4{B|ukTbhX7s(WVkX4LllmDX3%NT}iD{FOBna%yBLQdOU+>MTW#$RDtEh zm{;Pgqd*+ zOD)+$;20@}#Qz?@g~^k~<7o}Pxt-tIEO+aop0Vw*G0GdUz@;LfjMkXx{m{TK;UvRJJqRQ5!i6tK!-4$gqShMI(o~3GVB*7ge4wZ&{G>|8Sjv7wT70t`;d0toUX`?&Iqb^y#^ABI zQO7A2EM2&29YZcF1_A0*QQIGZI-%xy*RREP+5FF=Cbzj*?7$(12E{EIc6jLT#LV-y z*R1U)8!37$INHl9a=Ep9bN>)-)#{|REJ-!F8h3>){1(?!nqT1;QB;)`176MPXc zLA@SiqX2fipgmGWw#)r{vKSjzNhT#2&y+<+EzKaqTI&Kq>Pu^?NOaz6MXGjZBL>jR zWq}RAPx8F4K0p#*7}!*n{XHJ>=CW+|DZ5s~R5#VB7qpdhv7a_rn?&y_%;jcqlCC}O zOg`8zGNV^`li{v-n(C7<|G{l{4wJ9;*!GtII!$8x&l>y8nzKrnKGE8zl2DSOdLdRi z>I;IxJq(tF6+sNrqV_;uEU9m=B7v7a1SL9E*$=F(;%Q!JB99YaX40m8Zgn^F|@zPBGbhoZv8UTNLXU7Th3= z?)IJA!S;psomF0TJ0HNsK`jOV&x+Z82|D5i-s^=%VV@J`_qRDgZ^ zVEH&=L47Z#4@&-WBtH5n*AFo{LaZuDGvZz_w-Gk`sr3=xOL?8uCw;Gf$c{p^6cBNi zKMy8v;}CpQm3(pWsZVp~550b;6bvtrE^e?->umQitMGCsDQIY*DR*c|66?bCn|VG7 z82)W8U6tFFppX{U7_YMY(x?XJ;50rxQeXW1*Yr4iZF2VY?3dHT-J8G{ez+2lWgiT2 zg9ilFo9UY`WBv^%@YY*9cEK;+k|tY%YbK*(mthwocp5zP5@Uw0G*hA&AU1X@q*+-> z&}W44K2z;h821j9zV0q}sJXy!6drM&Qlb^4&+?O;aUjN`zb5fx3&(H?;A9`-+vU8s ze|S@XXB*@_(}@cmt&%&0V5Gvj@<;8>#khR*k z@#I|wJIPjM?h2|btAPU6?tgf{hr8nOXls0lESYXoFPGkBfe$j-tf)oorBJ#^h4~0>psITCOndIom(jQap-|=~RZbra9_noTs`_}479&ytP{zb#|Zl`hu)9{6r)a@jhej=4YE zxl{f?e)mPTBC6N$rK=5wE804=20Pk&A&tG&@mEciW;Uk1*E;v|!#%|Hr|)RvuL0i} z+1AxK@iZH`_|2}CDi|!v*x{3>naJsGgT`&VJtxLw3hTpgq^%lt&Z~6FtOtUz+(}eC ztxG?=lbn(d+ps>XaK!c^qpn)OL{N)J8MmCEppaH&4$rwWkpoqg^Tlc~(&!jQifB3= zntoF4w{1%{mj4(Qzn#|l>R|-wvrcaoYsHo=WP_A9=$V^r&o#Itz3Xf; z3zdeQAgW5`r9yZq1O?2klv<*$vlYqknL!ViaS@k4f0KxQJ8yl?{G+`a2XW<3WQmQy ztjsKMlSVI4PfEiP3ay@$kw?b#Kqa)+y>#T4GkU z)s;Vfg3^bQGTsN7h_L>xIY&B1nq0N#BHn}_^6xHS4j+Agxxjk>uNLguB{p(SMV z*JF?LS3N?YUi?}xn_|iI?oYw%+OAJiCmCMLZ|SL&lKA5O+vyPFmGGd6aEjlioW8j6fkX z1gH%!rAg5@JW-Ru&yG~u1kn4_pEh_r%kRgDWg(d?^1Tss zA?(fSZLw#m*bl4s7YFNAuD|T4^OfNharwDi2x-*77`r6Vx+W$LhrhcSzk=Ifou!`m z3kjE1Igra!zkAQ9B*lztA!#_}?gJ zlAu0wqw2+CPFH)CQB}|4Uw7<0ZDEoI6~kzW(;_T> z^RRv^hXb}!4Vh^irK?qwFWlFWLqgu(1K#6~#)w_W1enoxXH6jFi~q+>ama%ozmn1} z(z6RlqyM0jddfFt9Kl~T7E(1!mI{6n%NDJ_5SNl!ZVZG}N_7<%hm;V=Yr?h3m33tq z7T4)8|F;y@6Ru5UvXEM~h$!-U|sZ`;5WPaRpr{i;S)?`@yKedBU z@6^3tDH$iRbON=jxATJ~bu@D=S4Z42TMBgmcD3G|EK zuu87o7WsKO=u~-6>R~l#3oj15zPM1I?<3A!q!gYd(C(Ai#>BOfD;`nVH^Ia4iO1Dq9W%_|H9!Wk2N>Q69J=!&IkiIh%qu;QIQcswhNO>CY_qOr$l zJH>^@;szYJSl7o!Q_}vJU$0I;^Lw6~i-G-q*b3FP4+LQSnw?O6;6jHjX-?wUq zf9HVc@}JYu<$}^+9A1wEHy9LdZ%5bni!Y8P4l!2AYx^x%-?B1Z0fAZZ?Xg@XxDVxM z___w&)0Q@uVjbb)BzX>cnBTN7&ZB2y%#Z)r>zY8u%qP1(X?F)~gaDsV{lP;%PAMCW zV=y5Nv8s!|y5Q1rr6kP#nPY|#0pN;PL-fvI>*Nkfc>ZEwha^OlMlY0^w z!3PCSI>43l_2U)E0xHf|UVW!?w5mn-AxW^Qma9+g+091blF!vy2d z5+qOinHr*lsz<`+S%lyxL0f^=qHwFFwY5|mkzN!4t&x4%!2jcNB#a&a379YCAw#MJCstM!G7M$?N}isE96;u zrb9M)mmg)*e!#s%p^U8hH_FkyEB|r{R~=Vhn`;z^rGuO_6T-D`Fzx^;^O9-jUxZ2~ z6kQz+uQc!u3-tpVn9P4vW72EI_&fCB+ppP3fQnHmifj<*^Ksmq#M^xqtGYhsIq>q5 z`2c~AlSX&Vmsb>_X-Z2rh6e08e=ODZWpWId>Rrw+mWM2GGcs@~*QaV`H1GG2d0IDo zrD?JTCG7mF&+J#p1QwFLqkg`nsV%*;x4_VC!2nVLlZU%Cb#2jfin?o9ud(%(l{DV^!r*uj$ELw5&m| z7VtWrT)dojim0YI?ivtXjitsLA1H1+h9_+)k5RrDnmn-admElA2C;V{P+I0&EZbfQ zf9FXlj@3x;l5aq3B!=ngw$3JT`;b_gjNYojN~b*tGe0?thfg?m2<=*a3?D2kbNi+? z+AA3CX({89u|PiRA{(G3O>vW1M8ZnbqNTCi!m;jZu{|)?h^_o^E-O>cM`uvsH%aPt z{BrRw4RU!|mr!258gl5$&mb>CB5ab=#=ZzWH$HiIQ(R`#*3(6Ztz&`BHN;Wqt7!$B zTFz`%{BbfNSvDJKF?;|nJvt%P&Tb}umCSQaOt;x%8VW6*z6%J>`t6D5gEAoMn;g~Y zsUAQt9j#LABj`HoKqN}wxQc4kRCKu)rq}v?`!m}!n&pQYkmEpay_?8)#edoupeG;v zk(PBlozX<#g)A$cFZ*M)^ig(nyrxxNCfbW=@)-j52%y*Poao0TcyXGCU5C9+2{G-3 zd5Acst|wPv-Je3M2W{+z4vL_|rNUsBW*Z#|%~bKAIxaXEoUdo+4jaP7>m?pC81&ls zElrLCQ^{6sq+3L) z?y!Ek??krT{#$7WEd5Sq3`fy)U^5+laqf*@J6=Up{In?EEcd23=7%KehQCT{s&(00 zidOE{`om1xK>wQ=W)>vDKb^c#DL;k^%d_LVS!U-R^9=g_FGC)x?n_RSqK`zKhYe8H z0TLRklmH|-2kg}zj{4j?s!pI2rk&>ks9~?0d?ElgLY$|Se(%((Z17-jnF!ZCp>imz zJgGRxvq?Pa(u0=8XQh4V_c<7ytl;)>%Br#nuG{&_vfX6AzKBK(m;<9$1|`{yq*~eGb-!}?SF z#?yK}>J5$0$G_hkc^WMhC!(0WLaIkH4^A|MVpr-RcJYW9bytlXobA-$c!8&I^cS!i zXFyE(qYl>l?V7}XSI`(|?P8yl$?6qB_;Je#AW{hEoU_*I3sMbejkE={9rgTs0{I&U zj3hFu+|#ML&Z)cJEtP_Lku8b{qg1g-9@@9tGYQ`>9KZB(Wlzvv%XUgCiMnJaN1tE0 zSUR_3fzuk5j7O8Q=qj}Q?B3DwLJ(9Ymk&Q*=ee}cJl3~?lGsV(F8(7?eOMYv`wa9R zQwi6vmhW;!zQTTNBoetrWD4~veWI$Pk)oW% zXkHy_59h1Bu0A~28IkQsLhdvl{nhVt2#^+?@9?+c-H{EY2g>=M5MiO((+T>=cSgD7 zC2|vXacHMzkubpdJG(U(Ml2NQZ$9ygUh{g;roF_gNtcTg*BQ{~Dc9C*5W`b`b**Cd z#pLE4;@FErMPsm7xyGL4w}>?Efg7P^NqB&~)70`f$!F zDcR$F*v!k%@44C$IG7ih`b!Db`klR1Cdb6a z0)MJ`cQ`C=b1SYR3(-p~P?P@J!%CZmM0@NM;lu?tvYHMR?&2DZ-yBFqs$jK(Uj3C+ z=NB~7E4~Q<7ekjt?%zW67II5&#LDt=Hw?NmV;!A6M-L&UFOSs&`LtUdrE3od)%M+` za~3FFxGrIgV;QUfvUqA~>M&CeV8n=qzIe8XvRTF-`dPr}ZR>1LXT>ZZZJi4u^;ZK5 zA%)kkhtB@8@cy_S^XvZgi$f)hr*e{HIU# znoJibCzZ1_$A8;hy9>##4-uve&)JZ;McNCI=?UG~MoPJF_a&2uPu49rHz8bw@X@^R zRHjq&vYvk`kGY|VdWXzPUR^IH_z0fOba(^a_-agc*tGct9bhc)c&SbHzR3~lgNl21 zK&vb#jEb(zgRR8O7!qu{FbCx0s6hN&59J>jYS-VS>S%aa8<$l$!M)9X8)OgV`4)*> zxFdA2Oa}b*11-i*GxsKlzP@7WKypRuq;!K6?G~D<4zwS@Phf zT~anEpTu~MdU)u_5EDwws=8R?XP1<^oMN!hF^!uSxv~&3s!<|uH4WVE5KYf$(v3a+ zdH1s$T{da2W)V8NUOz!Px{u}42Ca;5-@d1aX40u;0@~-~J*F&Dl}hrd(SzNZ_WM9-$9*o!8Ye^^%Q4p zsa;sLR*FGTr&G9~fJZI!1P9CWu@Ehl<>_~_R!E!PbA4L%!C>{1VRCbRc@L4(X(+!x^`enbN@@alKGqI zCPq7-${;&($iq-zAPPoX!9ij)D0|95OYu+5($Ia6@ph>Q)$gQnop{?ZpwqhnzsY3= zF{n_@3k4|fVMEG!#sP{)VQga@Zli;O-K8BfC6gq#sOE&Zf50n;coWW6!0IhX_wRHf zw_-zon)ILTVMhR>nAQ%F!5S@BJ{{%d)fRnw7rA=Vrj$A60=tkw5+;#&bwKF4C86F( zT)NlqC?CeSTO@Pso<=F|XV1ya9 zYVc0i$W)%8fi%T!wn;fSAr_r-5ZN%|J`}FO_gZM({Wpb|xk`(mCr@eumq&eRVjWD# zZst7%rI1Ji^FUzh2u0@z5(-h|VMw@_H>dW!++CGmZ+UWou@TljdaIuQ~ty zZxQC(8vjWlXWt~7g9RTOHk30)qLG}T6K1|F7_&3e_C?IIl*=Z9hj+hLKPSL-edY7? zV%QM-t-Ado_`xEa5B9YV%Na!&yVKvH*JdVN6br9s$1)3Q@V!_P-IMq`b8P_$&Z}^g z|C@5>%a<=*U0r~Lr?ayYpH+Y3_|t!6GGH(PC0mQ29H4d!#G6m_FK19DNRoRRJ35r; z`K#y_&*IrXge#96`gm-QnjSeWByjHMP9Bh4a%;*jYeC<7ly z5RKJ&$~hoNDokJ7W5i1jHLrnFF|jPZEwD41#UC_+t`)*|8orwC-j+q_~i%)fu_s!yf0=gxCw`RFBkENiPNH^-^TUz zJM?*-9dgNiX5i=q_@L{n`dr0z-otds%i;m9_2a=Bc_~s1ZMj%ONRACjvT62ggNs=` z!xZwDM-8BTUiV0DkI0t)0eIxH^p}0(<=}~&)t;t&NTS@~oL{Qn9fo)8fXE?23Fk<&UrvXw*6)IS+br9P(yIZZGN9ClzO~#b zK$*VjT^0}|2H+R&F)2{7nxjTLKc6$%P1J%jribPL*>tfFci105 zQ^6Y$)%zq_)FNwZECa~YO-y3$9be)Eef`>*z^wCd)g!81s;;1*Kmp*&^fYVJ`IOH6 z{k?kec`?!$8+77`l4`>9UX+&r^VMRPA!&UVU=am9pwUzl8uPVhy)QdxKnN-Z8HW6E zalqKWAd1QkN0p0hdmB(iC9A?ABKPiURS^bKnID?CBGW=Se!%y}7(Rp}^83_m8ah|Q zs#0HOzDlfh(vS(pO#Gv^Tm!_&!E6XTOu{F9Lg-{4`ap%6&O0%wk;L&4-bl5bIlG^C zZWR^Zv~Zon+gMjMU`Yl$*@?VgXk@>RsUU|J*0=8S-`viBT~cpKB)O|z1%dT;`s)*7 zf7iPPRg`tHjYPXi?93Na88CcY{Pd<%m(4xEg;zrK@~@)>_oz=^Rok03A^e5cyu7@( zjWAwlW8*sQI=xsno}>5iNSEt3Lb4jCl#K>^yXDghqJ-f!C*fDJp1dlLddsVCPd~GB zHV{T#wiYOPH)UoTIyj@~5$!nou0(Z)8I&*Xbdj748-+DQRNP)bY}5dnOZeZR;Z zUkP0hfWH3#aS22j#Zae+K|2*(n8Fz~^!~6B?i@DkK4-E|nQ`c#2EPeAgG}tI*pb1; zxLhb)P-WzE^DGj5R_PWeTvl?JrV2k=lfiac1k=>oYaB$Qj9A{d0L0_~R29N%Ebu%W z<8BR@IAPOI>CT@w8}*-^=cvdz+-Bi*kBjJ0>Q*b1c{vVVG0(-EVgsb)-C+{@PB~N; zUav6ehW*xibXC026bBK8m#{S352e-F~r8LGrI%dYSuapuKxtJSs3O`AFF8iBRGYP4Led8Vo($liY zfe|crQ|1CwN4R-zhbl2XGiE9yUp!F47Rzod6RDV5j&2T5{0_{LA^@{H%!s=7w~f#N zg79$Thy2L;6r#2$Ha|fZuu^zUQ>m0;E6JKANq%-XSq2loM6s~|O8Ao+tRP@`h3$gx z+|?Jw;=X(f`0p5oT&XA&?T37V=SX`b6r3ykbih zFdK83?Hxk58Z!#)+>5u7QP-?<9*wjw0V`uY4!HOT$_tO9fmXV8QQf8;fN8KdvL_~< zPEMotpwDQflq~PbMqKLXg16`{yV_$IQ{BqFoq`xXtYh*!d^Av0_V2B^DLe1VA6vUP z4C&__uq%@nSLVJP==KP%>#s;Q*o_1B#vKbxntq%*K(|%o zF1h%bZENnw&su5@kKa168n3%!(N3G|#SYQ&=2@%iuR3-isyl}Vs#NFuDuzifKz`59 zprDoczNX!|l$JD!5O^j3IprYiNs=oIQMbpK^!V#R7aNid2R3p3@ZB%nKjihhVS+g+ z73Dafwm)wBcS<@1EL(GmK;Yn*rIEP#*EkHQ&Bj+~xzf2WlR}5oV5S%QDi9dv+%#$T zZGT8X^6q=cTYZMikD|>>bZ7Q#6**5H)iD|5vL>T*dtrEjK`aAlgLCiT2^qn-m?+nU z3M8(cLUaa3>5HcjEdnsLE{7Lub?7{c?Pbfld--ro2iLN`nPY(z^+(*gc#gXrVD>Wz z@WXaSJz2y+ zl`L#-X>);feX7}5|2jYe3EXE_#@PDT@72S*g$Lne7V0U;*Otgh(ycrEC|Y3Wsgdtn zt5u$MUyj1dH6Vk9UK-{0!#K|?L2agAfOP}qqJnYBt==S4 z1Bf75^|u=4WkA&f0Fir}xnk^`M>{x68s7w9_aLcpN88@AC8owvletq<eG?)vLp*G zsiKd8GRIG4-`@Kn3~L-8ZNj~Z0F)fo6pd|+q;^Ovv|3!M`-C}cmf8bd&p;p*K=82QEBV@ zJvF*u{hj_5XiX6^J#IwPtk3y%o?_j5y)32Sg|H`}SS(=6SvF4bsBckor!I?n#`YFy zkedi}yjkf}PI?@pTxJ<&Od=_!&Kim62%FqyDFlypn%!t8%%j@?2vUFVz1OL`|7Dos zEBoJ4Gos7LzxLVlKNBANe)qP8`VQ5i@w0&rh%0L?ZzH`8W%*qVeC7?`(lV zK+K%m)_zTyhKe{jZDUe2Fr3sLzGHor{M!-?^5X+=u0`C$;&H(K@BU4Zc1zM@td4xL zecuJ+$Bvf#sy_ajm}yM5>>lcR!a)J-inJ7A?%%fg*u3!7Fgff=9{)jtE|-db(o8M| z*FmVdtwF_`Mu(^cnD)WVb3~GH!l8d&G2ly!j;rJd?dbKu$a-YxdyeVGGD@CyzlyFe zwZu`+N?oN+fzGDe`5gANrbDYVMa;Y|CwiKUo=KKlQD82-j8!}Cf*Tk&Nt8-pZ8;JE zQ>$a$UNE3?8oyA^=*SWWU9J9;g-qkOKaSk`NM&R4NF3Qwq~G(5J4s}>>?Z=7dj)vT z&J=hxTo?dZ`$k8s`<>_~P{dD5X?%Ff4}MC&k;gxl^G)o975`+1wsy^Vb*(lp+y@3*fWdZrd|c;GZI`EaXIC#5 z7bbk`GJ_=j$JKbtgYjT*We|<}v|Mu$ntLId_|HIc;h6U9H-5x|j zN~yGt4gDd`gF?(V*u_j|s3|6P8X=j@nSv(}n9eTR_* z3X+9^grr)Bf?kx%t%40-?gQ2weam*qoFTnX<{T$X$jB>JR}{--3#Of-?6&f*h5Tzg zXFT>la+F48qXNVb)VlcD-@2<_;T{UJKTfP3Z(OTM{*B$-Il8;f$2-arn85gr%_P+6 zJ{K~x?{H=eC^Hw;XRz7MD#2kMCwe9n8Ya4_7ugica$?rQ0xcnb85PGFMH$2n7z4kF zn7NA>kIScWyc^EXIIS`mu_UE^SV=~CGk%bPj_?op>nuz=jtTTW)EXI*lL)omna$4k zL`8v;bOIjcsBmF^#4GE(y9i*mQ(i05nKw;gxxW`A`-S z!_ob7(f}NK0d0PV1b4b&{bF_3rhVTUeUhintQv#1w8p#+SWEAaIn8vGj_(TF@`6!9|%da#Cnp z4%PgpGs4cX;f(?(>=EqsJQfE-=rnLw3o%pOuIY!oo>L00u_Ts|6pVj++Y&X(vyD(b8#g8Pj zvG*fqgIaZ@xvSFI->mava1h)?jbAk;{HhZB{2|LuK}wZe8nt|uQl2Xk{%~*i61H8& zIPl*KoYmN!PmQdy*B64?x7`B3@ruU53ka(C#;LPYXoYOCj|#}nlrlX+ z{pu}`(jtEs4cQcBYX+0m-Je-3q&SkRix=P1t~f=S312kfMox&t;<$#T%MRPZkk*(j z^WX|qZyj&lH8~FqHI5N|OeyU*mE$DDVR;psgPuYqtfJ6=Cm%W_Tg`RzEng@$j7kzd zNZnv?lc@KRV*c}wF0L{Dr~`C#q2SBW+;yev>jhpgS!m-K^n0{VHk4rSfb2rO^-Yc8 zUG6HsgSk$9oR~+858fgc(f-&qDa0JS+y7P;>~Va9OpM~MY|1QfQr1haO>d`l{}s)Q z-S9zHsyx;hVrNb+>o-_!UI8NpLAqKPDR6JM&d&xF$Qoni@&qsgaLR>Ol>1&!5k*N# zrj9l}(A|pC7ph7!6yq8{|AxI(^7V=X5 zuEKP?w-TLVHntKGJv??*{2q}(>}kj8xwIi@u8eXXBfm*K!ma6GC1+R;hOcx-}s8T3y!9z1M!!_L9OBS zc}kKaDJ;~)s0~#Q*1Atd=VH+bn~Um&cpx1@l{FC5Y}kx9)Nr$h?&FF)XapBvfd4S2aJpO z^M7ioHZwcgEg3EtMmsrD=C6fsZ0xo13O#$;#yHsAgX5%jVEk;bH?8%* zTFZ%IZ?YD=hJK8>vZ($OCG=d7$r+ou&|1)@>?hO=_(kk`dMcE+?tCsoFm`l$qMVK`#WNaQH~8n{+Y+Iqu5Mo5#xq?;t8eiZ+4_^bJV^%M7A%A+Vch7s z|2kXIB4COjV!${)70N^&tG?d_p+wSj*MIdkpNb)+>s->*Qo{!pF;g_${$3cp@#}n( zYuok4cM6FGpWFRN!3VBzv20@B)o{)QkKNKOeHj@U!Mo*`<5-4a1?vp{AJwA|KB7xz zqol8frg*h8#Sp^Hu4xB}DsW_ue-PI27YfCebqrwDv~u){lV>;&Wtz{*c(i@gHAeOF~S3wo`1a zzsi5WU$(B^7FS83^*Vk&j?${AOkn;)#D(-F#gc<2Cg0>dTTsCpHF6Zuuinvrm%PZye zax9$edtC%qZ`p3zx+o*^_%f#Qb>6PQK=ZU!8>dqm=X)0LkA_?x}4-`WDZ}A+xc1Aii=D__Revtq$}6bBjZ7}R&Uy}$?3xI8nx-_a` zo$lXX{r5se?ZBqecINMD%7XcE$3p5WPTwZ0l{nuKJO&}#-XbDnGp;qGYSGei6o`1O z1qQ-_DfcN+w5lmBt!af}ziXCjF|BcxMuGx(KU$$=@XbYmQyPz-Ta6m(NE%@9bythp z1@s_$kbCg%bmQtagXYw0uZNP&qDhNdjr%#rVM@*}F2ulQGeu?*^{Qf@O80|?%>v)e zZ8Qp(Wd5Gm{;X;#tRbT?uP!#`)o50SN{;4O>QvCt)|mZm05f9)?9abOqkL`+q=Y>} z!3A>N>$}feCT0BGFMmy#Fj~+Zi&bmZ-o+_0(0%7xulV+9fDRw-_tth5->Z4DEDkSF zED4oW$n}QeWFkOC5>oeWn+yGX&`~ksdN;T~bBt=M)6e`Q0Mqns5J!H@xSdU0wG-Sq za)FG)W4D4?;z-lqBp=%rACG%T7VUe8rg`Sg*zm)DYd5)2a`tl-NQK&`SvKAr4><)@ z^zQwu4ZU$qKX`8tckIYF!&7ZxrV^dgWj@7?$yH*Y7jUxyL8<-cqdk|aNp>Z=MZx82UqwvKH!o+`Xj18}2S!pYeawTfj z^vzmuyv=7mFIES(Q?=F{Ajd{|y&BQitoViJh2pO#MK=K+j#3_{X14IQfGR{^&surp z2R1p(UbP1bSU#-v?3H9s`d}}g-+Z6Ml_j*%L(p#cGV-f2=n0w}*dmZZYIoy=gGkd2 za$!>?>W*u@(XFkmpcnoA_Hue+A}k`}V7A`Y%nT7a^WYEaq+w$EtFOSW^zdH%4K9P`4f7er-nvXoap0_Lok&w{&htJd^1# zkB~D7b&nu%KX!9^m+10El%Ja92)EK2WM9$wmI}_zEDbU2%`9RTqe&6**Mn~n01I4I~@a&lOn%Nr6XB;Q1B31zw& z-4TBJETB{WS>Ew})gVSj`T@>W4QJ+TZstv{O@P+&nb?<=`%F#VyO#ui3sH_&dhw=nNchUrG+*GUcEfPZFgp_#>+o6NCWM&FV zL+3z1vl+%s<_D8EK~ZsB_`i98ELUCS5552CpSKeapad4mHux7HF15#dhSFiGN6rBB ze(9(86UiR+IsLv^Q&;D6vz=ewd?E?@$%QhR@xzrgSHR4^M<(BEsYb&vjC#72!iliB zj*m3Z`$*Husn%d|Dwcsw;>2&sHy^IHXMktoOMY%R5)Fm)bTL|Q*GG*}3 zsrqNupx#pwu>#R1#MWYA(kk3$!y+e=MPNWv`$s9sZcYr&AQhJ_?cpMZSD%*wcRnpL zgc9~o;E)+|yZtJjgimYBvW8Z_5%=x|k2H2B$%!#;RYD64J5;fW{9GD)lLrX#>5cUY z!t>}p*T)o(9_T`N$lx?;{k3wjv9W2){#ZdASyqv)%`cEeOd{b*Z~ks~M<*Zg@{pbz zkGn_lt%e=vXOoG{mjl{p)m+QNF;*{WVkUzzyP?iCTVcRM6Opd8 zQl44e4LwY~S z#_#siEwQs>0^#618j|C%M}qpb=7YVnD3+1Gm?HE#3a~yw`wGMgePkiLE&A-25R$m= zYQR6uJ!&Y;Sdy)HH8brw!AedhNb4P-5u zfn6BAocDT>%u{%`L0&UL?C!@OQ5#(3`Z-?=M~H_ya?=}J{IvACUn^t$@4*qC9-O!J zjoF*Hxm<1cwEm!wB9hL8P{hG_(&DL@VVQ+1wCVthgVxvj#${one!jTq{UYJdvK!<6 z#-5@>6Fvr*;dZLHW-xw|8haj*lQar10-FIOOoMIFNh_imhM+3G*f<<~eG8=Fdd1Z^ znXMXu+OZjvQ0CvlGo1o&m9Z(-kS^hKKm`H|Eu$e#=8V(pM>fR|ar}4GiO4f@w{yt@ zSYe)?^OXxWDAxN*91C)9VSaFPQ#NvCUTnLo3*f z+~UkeoeVG%X{l?lG-n3W4s~OVqLAf^Pk{5rYBKo`A>OG`N@3*)A20lwq5V=v^!jdg z@EuxcSIV&E>)jG!4j4aOQ9FtI>Mv<%p=t^%z1RWLD?A_b>@)^|OgI|~s>#F)S z9V--Ag{T<`tL|?Q24hK!6POS|gc?f04y=7aB3PL8{C;RjYLrPzT&DHxw?DejmW8Qz z4yPL$bTC2URAhYMS5vg4W+}@ZU*o248}T9MIfOBL(NsyKWY1WB>2Rq`OR_>A2l-jc z^`X{|mv-TrBM9Ob(;6{UBmRVr2J^~C>B-tc%=k+0EARM&)LA|9BAF0V5~fRsOmkL6 z!SX$Avy+TN1_|VGBb&<`#b$=}nkb=Lbq+r>(BA}uo$mBrvLnlp!tO7*BC@db7aVUp zv5@0uJ?Lexkldj+jP|_%ZqnUC-xKKUb*TZ4zQ9MNg{MiBu6%<07F zL{A)ZZwmmUR=?*sDWsUe1%o!Q1BO{6)G~!2~{vC)d%Pw(~g z^fR?p{rQQcYAJcTpB&Y?q@OLapr_DT3LHM(uFW1ul~WiVZAm&k!8_c+DXDsuP3F)- z2U(@>z3$u=Cw-4TNW~w(idW$g)x-L!=I;%c!uv6ep~-&InDcTx0EAe!M+{BY%-6UH zocYoCH`&q98?6LQj!alT08jv_!Z6ncLRXOYpY#4$Cp!7AI&jZQ_THRpm6r6?GFKXO z)q9L5Ol?+j;R;d9UX@~JCTHos+3n$}6j4=@2xproeN&TUC*iHold)iHkHO#!)Be{E z51W~LxGeMprp>E8vMC7X+psiS9-^QaSzPkpm6sZOngz3?Io&BCt2*ScQ2MLA9G{7P zT3ako>4Q{Ym%Rr|W9<-HIZlR=$We@B-RpE1YGrABp``0zmL;rk*>i^R5`&=$&QK$0H%Sr>1Qo(?x;QkYX zlL5fgbEtgUmO$y5A8GMd{P$UQ&-|wHaS(?gKd*K{wqCx$hBVrVbM{H^|Ja?5-Sd}P zN0O3+mUkLYtV(@ky6m{67@{$n)~1jd4!qEH?-H+I;N$m$_hWgAIg+l3hztfeX@+ry zq;tQFO@F@sN(yuDd809~>!D?4FK zfouQ=qjF!#r&Fxj)4KcE$8ak6dz-dfD-qOzgzF6FRE6lBE+IzMqVos{s_Fe_P-Y67 z(WH~f+*8Oh-s`qZs*z}U=htkG?i(SijjekPnvAw2c11;sY!iMUodnp>b-_y&R$Mj!uR82)uq}dC07kF@k5~kT={<-_Zo* zs$y3iOndUfc`)bGHJHiacKA1WBRc1q#cv5%EFK|P@S=Vo2Hg9zQtfJ{!DH=A$ zoNjn_>X2IJ+lTb$ai!-t_V&VCJA*+y{hEC$nbTE3q+!GBY zJ;!*YWCkXzW)UD#nik^ZhHQw5VjSpckbN*aWz&0Ez;h4>&jUyk zPt_K1-P0veIVaG@LmXL%^7k>|m(kE}lUm~Oi)L0%ro2)_*BV8ubNAu$`q31vB{bC6 zpQxSdDh#3IXHRT5+~m;$vA*6rngBVgtTfG6y;+VQCCk*-kBbq5j9CD+g+lkDXtz11 zi>3S_IOkGi?28Z!kHtwY@ACk;FZUl?U9)Fo=Fwb2RB+9>C&Q3E9c@ZC>iiiS+lu%$Zr2I3d?Yu&4(vmi$2 zt4n)Z|BzlJu8nG_<$;!B_LbV-vg8#QJ>yo97N%bP8RYuxJwQ*;>j6@}!7pW(YVgqx zkN-TgyBlgQN(B?zF8hzXSzChv3(3-+%pyAzFodHoryl+Udg_v5`H2~XK7rceT~=5P%T@LOS~5ypUZ{RRMX6P8(afr?47RcI@bQHzit;$~3jG zFh46>FIGadQgt`Gsd$#?);vt%H)~iwf9IEAorpt6$OAC{+sCx&b_mkPbDt-Fhna6F zpYZU(?1Q;7DN5`QLTEhZ2|QbPFmvpC}|TBy1H#BB493YB%4x{?K;hm6lWoI@l1 zw5Hm$-`+B`=3QmU^;#;W5l()(``x%N9?}B%&}UGox24(vJS3}X6emwKs5x;nVVqjO zts;M~wi(6bnS%wr@8=ZFSbt=m7Sm*TDy}n5a6|s5I2pkPa03QX6d_?q^9pEE+8xnF z-9;s5bM^XgR6Rs+@o;)@6~uG55I-J$IF*C!QwiDP52M_?o?+H8B>=ilP%ls|Q*%?lk>r}Fr zXhX?VG%^e)vpx_HT$P{+#YoR^6udS~0+ShdaKggo>UVRh?ZLqopTX%9S4Znc)l%?|rqC_~#L__sU>Cu7~UU z9p)a*&7mwuY75GR`gE4y-K*;sB?Q;jH*w!SjtfghdS{e;v+lHzLq?r(WqnGkYtquv z6wY@5-A+b2D@cT{rx2HamKR)@=`TW0h$dB4BQTjoP?m9r7gi~q-c3A~QNZWE@2C)2 z#wwDSjgk~2htcazWKfD#sut}aqCY@|VT2E6{$DSEpZ7;|0r@|~u8Q3a|G`y;LM9~A z8mA>7J*Y0Zh+cisyLY3?x2*C;5klA10`frGYe?=E6JR21!nFl;clk4HZ;o+y>z}I&t z_vw7+8AI%kOM>c$>gCEYg{+m*>8Gd&kP4_Art%YxWdL;z>FG^~e98}T{7@z;Ne*BK z04j>>wa^Ij)J#7uNjk02A*QkBLCM!b(K0DbVwvBT9}pQn{1<0Sx!0vJw09ND3puAE zUGYVVoWujrHA+gZ`GQ1IR5_$mnby1O0I;iE#f9423)Z7NdX(Bk+d}M;D1DUev!2N& zzup<4O7LnYxgLpHc+L7A&_sduE{)mm&H5UazY-jPd32EiH)gB9!z1}lz<@Y$%%;wP zlwn7N93VMizXSLox>CA#HR@(;-!S_o*jHVJ6DjO#34Re`3;i%fcI;m&=nd=4UWW8Qe_ zj#E$+2fUE`xxZQwN@l}%`TDJ+P^zcs!V+dU#eGNH@t#^!eVtVGN+ZafmYoG99cRMX zr${96DO?ZQq;~CZR&5a`m^n=Up5AKe`t zPY-=Ao1_vk70aT%SsuB6X`jG6%8c;?u@#g``9g}0I^x)yI}}e__gNkpy7+({Ur(Ja zbHIf|_x3x30{t}~1lAP?rvA@-%V1u{g9y1|MlE-7#W>L8^#4P_M2GTQFMAVB0=0`M zW{#8tv%sToBO+vW_u!aFr_uuwnS&M%vb*Yj#q!d4lv{*e(^B>t+1>V;AO9b4ULik4 zci8tKH(BsJ5$=l(`M$OisjDNeOCfZJKtmh!Yi>I4Qy|wE7BU5lWN~nDxm_JCzZATy zXlwJGZ*&Gm&?!w@YisLpUnBcq#dv;vHG%i@xm81^SYX<-z+nYjIRUhX@f-2f0mCCR z#R`ue=u$udli+nVi;{3=PFzn;H3`-grDm$z=kzVr`vLse20y9`x<+MUv~weZMKor9XUH4qPUAQ^lU36Ow z*R*L^AO9riy;GR~a1P&Z$XglwR$=pJJ0#5wWq{7$KEeXFipt9HTL8N=1m zk`uAd(X_^NtKY$CrwNbKp81&AuetZqP(Wth79B)?bQ{8icfiqR3pG<77J#mH zLv8_$VG+yvKM*_fb!|*!xC^<{RCl$EV>tb*<)&C zBRQFOredi=zE;XyAFBjyE1=^6FfN%$$=Vb93y|oNsc6-I0Z*ZVRDhtVj?IF) z4tPUI@6D$JW~rNNhn?d+5hgzG*2vULko4vG@lMjiObLV3g<1%ciiX);Lq11uHM}4t zIa-}!emXJ=U;yUid92uYY}grmx*so6%w$6pV%mkH7N zLI=#TKkBb0Sbbb;5L>L}UO+mX*b-+-L>`&}AL=oMTdguvoRTO(vYw}9HJW7ep6bvt z&)0l`(F05rq@T8}Wb}Bo1JFClaHz#}%Ikl+RduS`+gFa7g+25dRB{nUe79+&;pOF} zquZTS)eYA%)}SZEsRVb>f)eNm)=AI#7xO>5Vo|5DdPY!v=kFLnyo2=mHpAhK*6xAA$`PbaCjSxnYl+d~3%Q=zOYa;k zqq8^C5cGUg-8NLafp}@33--M{pgT33z6J#e-k*(MTwECTM7($&o2c;YbNg<2+XKnt zG08i$2=@r_HO5gIg8Wqb<;TH>$A`sI+>?H*#wa?k!f!7ig`4Nf0mKkTz&R_=b@b=W z!bxbW5K|lIp~+EH+0hgO0(*uRQ)WvPY+=qG#cHIk&(+@7fBU=2*cj+%ph#D&Mg2$S z9}@~`4c4*`BO7mI&nbxA9(?J)e*Ll}dL?rbDMB{>>H+ridt2MX&#(w+L_9SXlSPR- z9@{&PPl`;b50GCliQVrCN~Jp`$=Ko2aii5a-7_u!uPkHaA_WEvLOqBO`urbf-Cw(O zKJN`NYpx>UFG?SOdEi1}+KJE`dRN7K zH(EAHQ-XOi%tWpvvPK4nYegrUdXSu9p}cQ48ah3c=hg=@4-7WSl`oz=1FD;=;x=&{ z$aX2&SHOIc(oF(l;Ks&=II;U0=xO6z^g87|Y<4_n@e){EJ(jEICmueNdVhU=4Xy&{ zS;`$X3l9$mNn~dnh|HsAbG4SLF*-tg;J+THS5F(3n~)qQlJc-?GI5WhvHIv1p8luK za++~x(Y+Bixf?TMweUkO4hP*SjNi^X?02WZ>ps}O?r(?=+Sk5g#zW7b;*(?u-SM`= z{Y66usjZgtmc*e1WX7-tbjKo>R*_}4;dhR5yL6#cU$=$I+yvJk0q!+T~Sw->;Qcn-275b!H?HpWX`uu%-vTG8hK_D!|NUXi9ib0oZT}JlS`NOxJ z^%n0nrz{^q#KGxjsTmpzWsZ@%}KwW2dl~fT2{qdIP{F zZ*OmMa`M~D!}7vHJR+j`dfUa35jjIcLru-u_V)I!uC7esp#FZ$i!=&$lmLet?YXW1 zR+ORRhNISVl-fud6SV5CnBmZcNtYg-gam$pZ3oG^U9qtDDcN z?zzLtxZPugworS0;4UeQ+5zucnf9^NE*Lw4qGe1vrXR?qM`h>a0dg^H?CkvDMo;t8 zZ*coGo%<#|Z&^$d^+v%dqG}d{kNwg1r*U6r#s@m`xMyvZ_{}L{P63615htdzh}kJS zELrD;+?g&G1(_f6D#Ac5G{*NzIkxpDr`p(%rH(Hagf_+hhx9z}x@J)nSlEE;jcvTa zag)Pl?)7|QgU`JOPnu1onhD|5|3{SqPdM)|z)(N$GW!*k;LhjKwe5CZsqL!>DAMpn z`qeKv8Dyj?D10}5aSPyP6ZPCIc)Z;h&ki4BxYn|Gm5{GMF3+{u^d11kE=sf{5Al;? z#L%Hkr#hXqXEP~^#EvzN6Eq5ebNW?G7Yy289Bu--a(y38GsMNj3YtKiIKB27#ZQoD!{6s@yB-|&xp)^t z;ppgC_$r$~;BmL@VKj9Kc*w!yrZo=Fn(YV5UF2Zc$OCvrk4s6)~pgTwdt zQ3ZOAP3;M3#mtap;XeexJRz3`5J2(4G0d8MRzxepB^@gY?oXC zF-rE79TdTjx%u`~siw@^@z_rS*B~UEBM4r>d>>3d*W>sub3fUPAKRyuqwB{ak73M_ zx>~r5_7R+!B7tc^&s|3iMi-OGKd$JZLx3tihc{9}S}U^T?4fpb?RmBD45)0^1)Rf2 zn;`WQ^Yr%uOZDP!cO3&a#%(KoN=|cyUAypLMOm21XzVNKSX~fco-8%W_D?@7L7jNS z#2nuT2NIclkGpYwFRC-No84IG=q#rc4U$&dYHe;yoi!VZZ@UH;gK;)jl#KC2B-PXV zqLeN78Yu?8|Iu7?`kb%ONqIgl{^PjRsrZ|uQjUcr&gIuE<+uA7V)nKbz`{Vib}sSA zP_~FDgs2TSJv1r}&dT*>hM?2d2|cph+|bQcWR+)b>F!)x>qy( zlp*_^1t(5D%;Uxh)OVQgh)qMb<{14toVo%q5jL;1e!?nhA1%P?G$;1p!-?fN%>Pt=u5LIN^>9Na z4W#@aNka61ug62=dCr6OzrmbNQrDygdJI?wzW>zK`T6a0+j)!W8c%f`@3r38B{~#>kUK^(4@b3?@#a$pz_yebinH?tEyerWq4zReLA( z{Ue}-5Dqqlsuc^BKnc91+IkXCIcqpsP6oXpKUFu#J7*a6MZtTqd9}$qyLc6SN>`a? zImLbT1znn5vu5LTO9kVPu}e2u zUPz++Wa#Dq*-{?@W|Ss!xD!JPK9?~8XEpL}l= zi6c2JbtnF6{q@}d2;JC5(VRy9c*+tpFhem74!W>QAT9pAd(J5-`1ikp#keLaU#OFE z){~CC*#=ipWcp@eVTrpj9)f~IpFyaGX=PbfMKcC;#a^k-FQAKo*4fE*+P$P6_Ll+Z zDae3sekUU>&3Cor-{b-)^->j2kY0(o?B2|O$PMqL*T8<-|06|*9y5rz|7K^W%XL4+ORDK07WKB4#t+#Dy?Z+6BA##pJV)7V>FIHjbR1eXO)H$5KERmxVr5^c|%x)Yj2 zu)lC!=fo5<)GK@4?`iwKpV$@FZ=8PgX2Sj;w@UmqSzXId&NWp`Ao^&nBuy91fI3pX zDF|@X;?k`FD7DZsn>ckfJ3BiOw}a7aoi(s%ig(5aR^hj`K+a+$o%-Q2Er zOxExUd$LzuK4iC>^%mWydJIc-?acVV{uLB}rC;YxXS@vc`A5zCTQDL^Wrta+p!`)K zu0znefMD0Zn-NPR2pS~wkhPl@tWr%J^nIx@f-Spe{1|H7u#t3;$KXYizHzGEDT>%# zn3J=1KCSzB0S0-T3^0k4MS~AD0?XsLS&c~eqvmd%YjHhNU^7ry0GkbnR1)rD0ywcO zPA>I;blZREQv+K$)4f}z1sgH@+0v+xocj&5RPk*aN8YJjMTJ|wPnAcraNy$;BgJeq zNys0Ls)->K9JH%tmKPHfGc0ghQeGa}SUkOZ@8he#98<noLodUK$3rK@r;;-@lpfz-guPy{(&nnE5eHkar3{zEBuJyikf%j+O;<{>weaW2s z=Ocx_F#-wUtZ)7UY-W_|-%LOj&cMedBA}}Kb7JJS!4y8rt&ue z{(-6mLP2GP<&M4RextB}mLYmbyF=lnnNQ%7!;%B{rP7K>F@SD0Suc3_xFM9v zIogJ@q^PgE&?I;1DCoTYH#+fN4wQ@t;$qdAqnFT7Y;{4*Atn8vVX_ z=nGt+L#f>F?te{9O@Xg&1cdp@t2sYNaUNXV7XJq+UaPP+Xv-y&Fqf2U?>Mka3P!Cb z8?lW!1*1XflHt?E49}OGn}B_j6W$W}o>nuUvXRs5*ckcy$I?&tYmK3|RTcw%i5lwI z?5U5)&>ebyi})D{dwRh0Fm$2ur@9GD%-N0fy;}t|&eXqu{d&wkSDhq`x~+Ve!Jr%M zEX#0bTsfB)J^&ML^UApamQ|u}7R@mXH<{dlbN|GGTP@GU@I0cs%I-JC7V|knd(5zN z66NQWlewiooxqX~Qxd9w*YKJK&hA%33Ig_q@NpOr+B=O)WWuU!4sM}#%Go|OPr2Xa z`pu^azhLLPKjqN*a7=?ubQ?Nh6OHie^f^db|w=SscB_Jrj`e-SF#0(D! zv2$O22$&J!<*GTaKYy8K!1s~p4X`M%jVZ^WLV&4#?43=AM?EY+?t&mEWSI3>VW0~C zJ0VXDsS++bLNc9Q&Z33PjN>y@loIe72d!MZ4YRVu5Hls7f`#M-T$bnudEV>HU*?q+*Hk)# zIQjv9Rs_tQEY%+L{hmoxyyrg5Tw&=iK`Z#S$n%q_>(VxW39(err;ljl+=^_EGK@gw zkF7sJO4Ee2V_wB)iIHE=NDp85HNE0u71}zDrchT@tOhgyMR;4SqU~A%sQKW&0Bw{J zdakW6_ddhb&API%yERw76&L!cjR6&qqmh^{3 zdiA8dB8I!odMdfZd$F|^?2P+1A7mN-Y3p+R!x)egv9(jK*1I83BU>L-jW;|jEs*_H zV|`qhcFk8kG|KnrD7^4$0#&rs$S@Ku<(EEA;-Bj?^0$jc=74RHh8-o2t9;Se(H>#y z2Z#D(?y~Ouiptb>=iuqtB=+u5D9BO&P22-n){%3I$nx|Jt*%LFi)x*&I^l!yg|N<{ zdGf#K5Gzk%nqD{yyH5Eds6c9$KzUMobGzeZxc9@XmlM60l!7wibMpwkrMRn#CS5Th z6yoWhza9^w4rk{T3Wuf$A|;HpI2rRT#Ro}S5ZIe9wytCdn&?C+)U{7_s& z<0@VkQB=Px>-N+@IWQ+B8M;5{>(&)ia;zaqNfH9y&aI=@tuS@td^JrYg9neCJfJMM zxOiao7?^MQ0fon|2=D?XyN_&Ve@I)KMx*y$hbO(c4kr{aml!%X`!!B!Aq%+pOAAy$ z9_VqDVT9D)uEoowVpi^VKDn7Kyp5qHbeqkX{n3uZxAtk%yCnFjy1`-q6p=rHBI~~u z%X1Bm`}_NVi$ndX9W5&a2Njpo?`sLx)!d-?3>C>mZYt4|OMH9R<}I&z%utV$*!Y%V zom;>)kf`#}=cjfx-C z>w7!f1U{bxawk>AvzZdvn2$}IbEeGZ8 zvr&VtmZ$?noUquElUcRU6>IeG{hr}!uq3V#IISUkJ-w(G> ztw*o8yBrrJ>$aAj;L-8*9jNh)ZZGx)ulqh(am)g7$C~4Nb=8V>FadjiflW@N^MA$$ zrAqhx=X5uU>N@pdV*ffo>j~8Uk;X18Rw5L7zh94S|F*p6FMRxjF)?FdWbZs#L`%H# z5XBbsrXxo*K+ZwWo0096^JlZ){!*oQ*A+!v9<%xw??k2Oa*II&np&6gtckntk_%;d zuBO7z+D734x?#;{^wQd?Uw|k2>dTK=%GxzY?Bj65QSO7Q^oDt5e(%0Lg3FSJBRdq0_-P}va>qAsuUYmK- zld27-r?&N;x;b#8$S<2GLwjxqjodas%&+%GH5}Y3>>JH-hr)PR|Oyq3BKIIAR`n16lj41xQPna z#+a0Ss@Qw7(H2umRk`K>H^e2^O7g`2v9bD!Sy{Yjs0&9*|IlSE!qlnnVO=flFI?+llp)e5KSD5hz*qm>rk_S1`JL^X}tCJ zt75etRcue1&$d%zw}HgO)q2?h&?gVF&!j}9p%K}&kLEf5RY(I@dFJk+7@q>aPVZ+8 z8H8+UWTZ@c_Yv!lFpz;$->n}AoN$1*r%khNF&x{`R}|DmgX&0?Ygkx-l!5c~#(_ea zkN0hl^*ZG*gF3-bf0j{DxlT+_;4sDS>;ad@NqIg&{x3wpm(5VK&JVdE#x z!#U6!MKGyZZ|3|{nAk3YDp_=M$>p!d&(joL_4*of2!@j*8Y__(LBTcvGYzO^LQ;HG(&xm&&lr{^a*g679ATO3}Ns z5|`@*31f<5M@#3Osr+Ho01Ah_Ar5nz;bU(6eXfJT64|jOknVzs?34xGRl*MSmyj#RYTGFy+T>h8C`#9uW&oWK6p3n02hx;~z|$D5oJ zjUiNRVPQebzyL(i;yxpflRl~^n)qP}K~S%HCKz8=Yb6I*iA0@dqpBqI__{r!XoVj0 z_UkI<7;>ZM7||IA&OPq+8$1KlQ5DY{?r6#AUB^o&d>ADd?7NiXN}cK&U==!uKw-No zGti>rt4T#wp@Wkl8*nGu%o}x&=%$JWdyT>ThGOLv`;ACX19`5DNY{ zZt2ID;3Mf&V1*M}zb+4%dP`2?&Og=TI+u!w(F|Z>O=B*cBQ4njlt_ zzlg2ewk3ZiW0bwLbaWIx!gtU<8-pO2_9GcgFLF7UYXChK@pP)`0^U4SR6&C#ylE^r zMtvxNp4dhs&E*l5kn)w}_R&Fv>JW6FLG>3(jh*V~b`iSMJ&7Swl(#T?HLGeU`9bSt z;V+7~J6_ZdAAj8O6pcA*)E+&do7ySw;rKS41P6^5R{|?%{ZCndC_F0}47@vHw1moz zfUB*mPOq)G63Y*?Ct#~Qsp#InK+^6KHm*mNnqOB?#3dmXMyT^E@jnnqG1P6=ik<@ zL)Y}_A^OlEMtLsdv*2CEx6Oj(`L=q*4%LpX(Yld;V zI>2YK7)mkaIiyTSI)Bw#9u|D7cPHF_b!)`ODRjl3vh%V zZg!RUTLY+f%n%^Y0CN&0_7`E&82lOm4>GgX6D)z7=b-kI+1h+fo{l6_W`5#;=&Wt@ zw+Fa*Zlfd=n6g@yZ>bP z`Z_OkN>Jc#LG-;t}pQQ$AJeXhm8+A1~S=9-lSWEebP zLmAAfufGIp^8VJA9&0ixDyl;R-#kXHQdwW^PCw5Xyih}qK+@&ewo1XfwZYF!h&M$X zi}%GO-wz5cD<0ui9&A$a60Qi_7;~$`6>Fup9)EBi$#wG$*KX|v673sO4hAEcEZ1n8 zVmdK*@qVMsxXk2^O8Ix4VcXlow@XgIkv4Zb{AaGF=|^6}IK9AKej1%E%DaSI7Nj^N zQ=eNhNa=p73E*kNLZzaol#!cjJe{p#A2lhAHY|T z0`(ihY-7egY6e8JJ6F`+?2gV|2pk^y7c+&xQ&o>z`}fzbQwcGO+S`s$_M4Q|0LOrf z6oVh#DH1tZ01nz8O_6u*EeoK3Z+PEj(RI;V_ihDO5ZF`>eIIYP(P(qOzB8s%{F8DL zhBmOLN#Jo4qXtX8T+rOjV=Yxsuj^hJJ{y1FhXbFwnsHbX`P>f~*K+H$ujqS(s`aPd z6|8z>hy088-mOOULq|#$z_6r`1N$Bi&oGS_KN}lj990Lxr!PAK1dP@Z31b`Vh_x@I zb`^VYI-NGX)^rCA;e@B0zsrgt2N1a*Uk&?Qm0uxzYQp;D==$De-Y_CTk31?@=y4(x z#K=>?<2(s^K5cEbDIeET?yEM7{}>yDnALsA4O^5s(|M2*G5!%Wr#~ANV^Q~!UG=}5 zj#m1qMLLQ1N*?ElE8>rz)DOHpSW3<;e5i8@-6}D)UcPo23KS6yK^Lm-jbb36{;Z*) zAwo8q@o+}PlZFO?1#>kt%%0uH2{!Q&ns4#E{BGMS5wo^mzX+gb%WdoZ75!Yobgrr$26Jf5}>=xS8VM9_TSkY46_d%W$ux}Cno z)o2jN`$n_C3Avp!_omRj^ZkuIS4(WiJL_Hw*wdi(W*bMGqp&R z(6nVL3hC(7BP(^~X6+e!Q!!y1zdCqhI^W&)ZTAN4ZuQ^>?q&lxa#4~`m2oKhudaTI zkYx(s!9p3srqwgq+1c;Vc)f4Vz!z0-cZ$nT+?o-H-F{?ctqN zfOI!VHwYpnEiIjbG}6)`UD73>(p}Qs-5?Dj-QV&a&pW;|_;LO@+-L2%=Dcby(sOsx z)6@TxW!0_vrgyx9APJVVBejK2@jIx0N-WGqVol_At1K9v%LO%pFv%H%R!qF zS1bF2njZLZj^{S*Atx@`B&_Ql5R{BYmWMlX>7h}n8Y157ggq;B^0(1IgXDTH=_4qh za`5O-m_0_~7l9&nZiD0*$-CcgchuzP^x|gw{7WW$EWq~3io1Q5gwT1SL!nv8^xrO4 zePP&XP5&R&n!Ho`b8}DfPpc&5^8ymLSzXY*wS$1-y!j+M>Dz4Qce1~c(CcHa3oRaJ z$)h}d49y$%Eljv}-MC5z4=1%N<8)PQY&_ltJ7eq;Pj@+}hi_~k~K z-;6LMk&Rbq8gTi_e* zh`Pn|BJH!GE%K!bWZ-z*@$5$eerf-^hB0%X#Jd6~4G*GH_zU#1!vh1Tp;AT!}g^?f)n!otLij*0oVd>m{CF7c*51+oq069SnX zau64iEMG{BY~Co9f&`a?j>)BMPs;xdv;m`sry07=Hc8r@KpX&w7OZ!3L|!?rtyK7a z9tII3KfJ33o#ajWcbF>N9s3iT!lP3VarIVb@%~g0a`ZHsDriRnFT}$n;g^uaph*^( zZtu?GF-NH*Xj-cE5OA9fW<-V&-tBGr>+3|m^RyL#BkOtet6&MpyKzBUx>=(Z-M#}} zfnl|U58)bgDRn7kK->25W;=OtagjG|baCm~85!k~z$4cInD{PX{edfImbO*<@@J&% zjPcA#qy)t+hyI^ zi@N!lhVtdo1DT_Rg$3*K@{2dRW4H~(L1E$nC@8ezl40=3@bdD$eq@m;;$cf2XWr-I zNLd+Yg4wh?nJy!D&8w9}`HFNfFC7+2Fvy$3grzOQzgQ5Y;mYKIkS>QVqnG6E%H!@> zp1#~NjgF2U+40VUE%M-S?#E`sPF6%@q^*NP z-QrQ8(X)x-xpKEX=LSS(qmgv}6VOF}k*8HDI!kW5-!97by#<(^hb<{3B^Crq)M2gN zV$Y48GCFk)q^AB7B{DteVe@H#<&%R`R(#sV37Y7Gb?VB}XYob3hjKCp(7w9{=@ZiJ`6Ohv(bNy}o*!DMxYcPCZDVLh7rY}>eE~W|i z2cbV7mO}3I_Ynl-Lg~f`e>wid7(Q=LA@q_rC3)R}{pt(a@wj&OzBQ3b88?0p>qs?* zOXLs#?FNpDK(}ik@Vo+I+(+Qx_o*Z!B(&hcM6OL&F6=1MB*WzQd-R6dH&4-s*lcZW z&sV(vz!mHT28=dXz+7@CW1SSha`YhI7=0T({JVZ^`ZPrndr$rt1ZbBWq z5$hQ;ei=9{)?&Om{YsL(2N7;QQXTxJcfMF+9qlBNS=N++bogr;BJ6@nbBs2^aPp%1 zPrVqV&R_8dEtHP1mv;&B%BtuQwbw7pxZ(w4vF-TYAPUSQ5-a{1qYOhAiBeXGCT93j z^_S%KWYh8D95cJXrOkVDM_b#Ga+Tvzk?o0B~SD(i%m0BCQ{AsYX()Y zch(3RtF>HqT7C>55`7AG+0_pU!CWzr)ue-%^$pmUq#Lx3NEOfuw6l9I#$bg09i9m? zJI!9!fkMJW#246mth?40+%wf-oKY#Y#?hRF-|bn@Rlya|Z}u07(tFKO!24V6yN&?x z0nAlJMFr3_pNZy~s=?{+wh%eM2a` zuB7g1d2qm5dH!)OrX5vX^}@9ft*s@K+izqFP!QrK19_V#18`JBh*BU(vQumAl2l02P*^HL) zT}yM^lIkCQ;AJE=33b2gh3Q1AF>x=Xctl}^yg7Pa8O6zz<^8uGyyuO@#k&&hApS!{ zi{eA%jyi?Ln8$%9^RQw7qna6?-l?|N&#pw}jJyto_wRR7k|b^0HSwj0r^DM`y1U6|HZEF?e6}0%L+p1!H1t;n5tF7v$gbqg&k(a^^Wab3My*Vmw6l_J zuHYZC-agK^ovhw(uLVaE-o2qnWW!1DjFMe^rPLUV?y%`-=j}(3_8I;D4aM}60UpUL zK^V~A3Ec)VDulT`d(-*;%a0OQ5C2L__=<2cuEpJuWsB|V58zi+%5}8=l0uY<2q70a zMY85z&X`E}&8MDQ&LH?8K?wVNJS}gZWe9B4B3LVJ!L`ZMd#-Y4OVYCrp~R(13yn}} zn`lfvGo1-^^Cle~l-0*Gbv;~aIQz$C#?vXOd|Wpe<=rQaDNRx;R+Mu6N<;N(%!odR zDbse-5(Dn27zOczQZ6>@pycFRGvjMpLInC4C+TpHjMO$lg`DGzn8Y-rkgcs`q;lgt|ZIuRzm9unW@jm)K@jt$Vw`vCaH#^%J&hG`mvkHpfOLjTH|jhh1k9RY(GCyggIVh)P52Q z;~&U<3Oc!HVQm_I)_c^*)kC68JIOKBv=iln9!cUq-XmS_S*=cBO5=Di091XeHl4&K z8H6W&ZdsTm2<7sw`s8&ShqP!r?}E#8K899{3FH&{!whj~@~cTxR1@@?d|=}UXZ_G3 z4*3hkU67PYTB<0Wln9cLkIUdYZ6U$Du}mt#K}76PhA-Rc{6kAaBgsO@&%=`;;_(-} zi2FxJ|JpBPVj&4}nz*s5LjLvjb?#Hdlw3a7zov-n<2G@r)8F{r!qA%yN%^O%h>@I` z%jh(=OG8=?OXX;+a+ub=-KST|hgc~#?RmIb7(+ejA4pji?ubz^!n|SH@avhoAUfRNSUsz$8dzCdwYZsezc4--UUs3C~dU z8*3Eb%mofQ+f_)Rw9%u9??gN;EF|HVD&!XDe0Z>E(Rn8(9dNow766qMFo>Qt`mAS6H;_v<`8G5Oh9p zNVuaF?Zl7Z6`}Evj73b;Tw);nv7x1gxWYVnEaEP`o5}JPds=rOZ=p2&o}{Ql;VX*0 zBn&ERgQSjp#HF8j+(6irF)`P@598Sumqgc>=lJ8BX3)dUYoYyF5BE>|}yfPjJ7%0OwkXk5KAFZGRWmy_$o_hAfpLGk0^>^SCiODh=yxtE0OLWApy~!f? zla+_;r~93MnoIy_+L@korLIqn{lF)N#H}1VHB{H>Uvqq%Uq@}>W3$~u!!3Oiu0RC` z8jIQ27jfE(cQf8D#RoBx~7-sM`i6)H0Myu#olXju)!J5(J;)4U@u#1p^Uw){(=Ye_3N3eHd_=};;=6EHVdr_FH_|OVFXy^Fzy#Foh#__ zd-tO1X<_Q4;cqDgcWJvloi31TG(7$Zw=`aU5FVmoRhQJXxTa1LIcP99_2#I|oIx&y zHM)0aEMYXUTH~zC6_hb|Os2+`n@%L;iyurm>5;{cQjT%FkR^4Ia^7M1{f@GVD2J3M zly@#OFQJ#1MaF@jXmOa%H~P1+R6ox1sAur@EDO(+-KRa=<*1>rd*$G${ghgcI-EoBdV1M}vBL(P59{S*@+M7Mo)c0h{*>${>?! zam14`7k{vPZiS!zBlfdPGwCCmYb+=^P>|Xk`uHbEJWnmRlzGgK6Vw5E{KEYuE zuxFWNnwZ-~iD#l9->_@^w|St^&w8xVoedp4MBXX-W(68$20(%D(7g&hV z>PL2yN+B)`G%OICKp2=DcDE~@KK{+Sr>5oOr8A*#>+XR2#q(%9#hSq6bXLVM15trg z-V{Poy!CyY3Q<1?wYSn}3t+Qsb6+)%!80FI{`w5t=a7D5V@j^IY`y1H&d<@cY8?AB z%OFg&9}f3_KW~myqq=BVmZx)ImRb`}k|6|!K0+ow5qQ1eJ+pKu>R!JmlzSCQ#}T<; zsfbIi-TR|cgC)b%F_mJWx(`)7!NvgMcrpx}!gWx@mzj`+@t4hm{1fcLtbXLFswxE_ z&`%T!D?Ke+*P6VFASNbGfPUA7t%J2S3a?(Z0K_y{(%8YFQlIUgn%S~?;e2?bHIGy| z|JV;2Y)&54`qn&tUGj=K@=YqOe#valFbw1`n#axWOE_H`1;f&Vc1f7IJ+b4B=N(r` zT^=5pz$Z0vq1{#N_H%Wf&f!@`kZ_MTPG#cWiK31D*wQchW%zZ?s1dD;V6ED>q&Si< z_9AQV1WhxL{ACMYEYCR$ToX-<>*{M|%6x5jBlk>>WC{za?jsF~6*MV)Ty^phmoZ;TRjn>NplClA0lJ-6l5*k)Kl&P7%b3xPW0eFsn@jD@PyF;_ zO6TqsHZqv)-ffV{mwq8k(HJj7F9}YP&Y8Cj?)$_`jfuL5pjL{yfEwP?kvhxl%~K2u zrR9AVSw_&vKy&JwcG?Rk>z#p2e#TL1zlS!xyU?T(qt^sQD+hXIL( zWxNMZxs~~CIui_F}P;AkRwc9&}2rxlhFz@WY)KgIn$L!tF za0%eHw-Y?S@ro;BJnyp^EStOx+K$PPvx&w;CwZm5N6Wd~DUw%2U)X{ryJ{QG$^HHZ zXAKUYR*I)Ls}X1OkHsV4(+$yQa~H~l##UD=-;{uvYICEfmpxr&0!IAeAx4hSH)~kq zxYQ_NSmgOGyYO*t$Q0}ll0`~sIs**=WIC;aJb$!*r4igusfR%#AnFaPB+6}(suuim z7!kaUkJ3Ii-2~1y@g|zvK{O{k0RaJ9@&O>)G|HwOfRh!}j6zS@grI;=sizVKWh^c3 zyJj;(6UHURf55B~NS6eDNOgg1t8P}(O$OdqiO}o2S7S>Ir?9@YQ_c_V$R!_0?39mq zwFQK%P&ImLxL$Tns|y2P&*4k)N!1>EdPv{Xg}u{?Oz!i!{RYOxEBdy?&E>Z@=M!u2 zCl!t@h!RnaAze}x#M5YJ_Ss|zTB1^wQFji|^nhiMp*op*Z$WPwJ4O*vNWj3{Yy=-8 z@}#b!$8C_{v)9G2g<=&5#N0%I)gTRey)<2M(ljyvXeI$J{mrnYv7+MmM=ymbS8ADd zB}FAI5lBB4vS~v>gi7eL2nt6EnUK7u(&DA5F$Hz@xadAhP9rvrf4*dmKO!)?Sbok? zZ2KuP#EqhJJXGyYF%b9dx;4a6r)x}GM)(g1ax(5&`e~y444AUwFR(RQKzui9e@;e9ML}x8qe6Z=tOM3-QCnPT5nHTk$p7)-XK~@DPI@W;3LlDbz za*cME7y0UU!S~-(tYu_qYis-}sd3yU)+@=yxB*(`-y zbH$fujHw&t`qpK>hW|ydj-XCMhPw7PuznZKnVr9SrTTvZEQL4yOFK=Oy_dYZy#=!- zJycbxw_(^?DbI{Xzjg~3da_kz^!4=tZ;kth55wvFmCel}3=Cn?G`xJ#=~%Wr3Pm4& zr4DD}mf}L9$gQOR`qo9;xJkZP^}GaEX`$Bd24ft(1D%49zou^*OTN`}hHo6aQ1K>V z9J%v{!`$Ni@h5vz(9|yB|5u8W2W6cy5zXcMu@ktK4I84IoN;o_`1dtf7%?wAXvS{|xf0SW$Kj-tO zhd6|%mv!d2Zb9pC?oYp!d&OT&Y*gAPe7#<^;EhP_B+kZs|D55F&&EgWXB^(4;qFW} zYQcg6KU4jx22Ei4yd9cJ{dP-;=`X{+*oE;qQQaH#sjC(`wfA2J6a(?>tz8^!HFxxE zzcr|}Y7{RCljHn-J#kIOS><#MOPLo%(%WK~^-lRMPNw5}!Y}R0jK^>c)5ZvU1>8Me zAO(#y7p_c|4|zbAyR1BGu>(ypB!6&&ZO?@vl)AJ5;6F{X^O;}7caNjDhG>OSVjZs; zzWlpPu7EcbfGv-=vu%pB@qp8c@83)^OcWrHpfr&D zR~qW|k6&ba|1=*=0=WY1+s-J#ogl?@)3N^Zii6>>=&HNAx>67N$?RG7Gep1AA`VfW z>d|<9u484*NL4)4OTy>+FHKJ|+W7Q4>;9pswQecBZ$>q>G7)^kV(J%PhDbl3e6HUz z-qXyO-P`faKw`h|DSFwWMl34`@Pmw1!JCibh4a0n6MQI<-wd02Og3y#LW$?mo51=LTh6v z`1d<2pT&#pA4@usTFlC-U>H-2x>*;Gp#iTadUl!d5Nvevhn!nPZiDT2vf}g!smc~M zHa5Dt3t&s2iK(eUg_8^S)XEE3fBqVqIXkVZF7y|R#Bb0zc8BEENW#$djhosHc7(65 zGAQ#9{cp}l7%wPYt5t1YNrC%J2Qg`yI( z+WpmnP@4V+=}+ORM)_<{!(AGVN*-%iN2l-@lUR`On%2##i{4-BOAiMMlgexDH9i@C z;``!}bSChOeO!a{X1u5|Z-P`H)3n$Xb*X2iAMs4Rf9RG2frg*1Zkic3fUh^}(KLcu zv&>hhZsQ13MC!#?`uXY)zK&1AN6Df%_33=_%*-={zXy%x2xChR7-l}wh#0=JCA4y-}BZv6wLuFK-gxD zy5rB;DqpNf!|GCM$$6AgN}NsQmNkATl6UH2u998@xE)6a`I28wq;?Wr^lk(&8?YZ1 zKe01~Dp`Is>u|1%cI>-JAdWX^ofYojp;6N$uxvmNDyyWUku8Z79~AMpzgyQDS=C!2 z9J$QwWWFJ>tmbU7qglhO51cI(0wcaMlB%;3pq-4_V7qs+-L<+xvAXd4_2*_x>j!ffP^@;9Lkk^y z>7|`(CV($@do{n-y5h)`EL_2Wc0RRGpojJvQF5N=9FPKtAFOb}>W(X2tU~QL>z% zl_PWhIbjaUNhwBw1j-vFk$)Hn#y(FSf`pWC%h|0>IdwRM)GSWw}rjCore$zI=wiDIc^1SpRc+MMR z(Tr68bWA0xnL)k-iM9=?XL>IaO=4BHx0k;4t~&w_i0;J;btKN{mi`~Vm1=hEge)Pn z9q~ItwW{vjQdITa;^`$4O;#j-foMZX&9p@L>SU2pMbQ_nZ31Fl-5_Udc^4Kk+(f9G zr!nC2p;H9NhXvlh7A0Y2#SmWN&-EwM=PWFd(lmf-Qa$(WuL1fJp?tSA3!3`|-c0gc zShoEdr+z8My&)F1+S-row(r|C|0arp7YfOLkg(T$@MGF=8@8+RYXSzazhdTr@pT;- z8z0i!yfv$LG{>{Rk-g*h>$k7EWG8R*Ks`za#192U=9Z}Gwk8TABeUPh_sihpEMYzD zE>jmQ0h3>HN5sHFuCe| z|M6TfR2^SYHA$egQm(Hr!Z4xe#jSY_x_$E@I$xxec^ZJ)Zv_ArSEL_oJ008xe!TsY zlau3Pu;igGRhd4tD|IY$Ce@6(rlPD=NDupYzT1*Ck6Z_bV!!gLrb-?DBsb39E5DMR zOzya()aUhF$HEI9vfOqEo*`z6CS(N@8v^-1qBi|4-4p@5g62B1QwY^0z5d}Nf!hk^1me#e|{%w058 za~UiVPYsVRy((T9dxfu4teY1lm@y~9$rbqA;#WS#Ig~5h zI0nOaf#gGus@F<^c!rE8OKK}95NFX={`Jx5qi?^9-`m`91G6KZyaU&aW%Y-?gQx!6 z;Wc;N?{&q4w-=$6ZQKIW=AzABProZmu_$Ua2{e;kPIP)2j4@{ih9wUPpfLkY*Imc< zX^!=24&(mOdI!I;F$Ez2){0^I(B&{i^Re?MoK$*d+f{RtuwG(&$?&V~Zepm2f_ZMQ zCHylfnQ7r`vG_%no{wt(N9I%EqfS;pP)iEoe;yl7eMe;@s0(8tcs2~Ie|!!TVNUJbKY6}2d0fgvd^~@%l#9$WDC1v zLS&e|AP@=;7B)krXmA^(vS(&yZqA*FkukcA7}??|pBsiq&5pp;*Zle|C*kQ3#hzCm z_gypd1%~=0=kd_zlUbY1H?vEQcP~vWhZ#8he(0_0seeB!!!ful%wUW;9kjbt6x(mO z7g4bt@Y6C(Y_4y0RV{d5a_AYbw{?1AH<0I)QP12HOV?5}P|0E-ox`~I8R2N+u@L!; zPp}<>^%EJVNj2vVIycn^fUKhX~e#1`1ha1jVecuHUf@|IVxaR-A*dzRwQ-%}`)MC7~opF(Po&g`B2K zn;%1N$d+8O3{3dL0w3Ep)K4uxcM|(&lP$DmQDupkLa~=q~3%09jYYr^YdU=$^IIh0GT}*Tysw5 ziyWp1p2I^yMemp7*Mu(?=M31b6?jsHwZPW`T?sW36J11&IdE1Dq}3QS7BJXoi*TqE zaVsLN_SAb>Fh=AEa! zz~~zvHS3^ZyheV&2PUA`xVt_9Mt})+LgiAHs65*aH8wNfAncJ+A4hVGq{x(7h;l%UJr#nVfuaaDJ|lLO(5G0t$qmr zzcq@tJpJ3kU(ipMRrAfGQ+7agsao^raX&!Ii6`RIZM$7x^C)MDcrL3#@zna~AUF% zF;S~9lsBvF$Ktmdl8lt*%x5|aj#BFfdw{fVZG^4Q;xc5^ko=&D)Bfh# zt^Xln&7qrWSScaUv zuMTtOQLL<%>Wd}_ya4}2=>)R;e3rbawYPmCWwHOLy}nqMI3h~8zDAq%b#=|d^_68> zZp|X+61tTGaZ)f)Aw%y^F*T=u9Z{O@0|auyYeln{Ef)q7_kl;?4oDSHR8o5So80E= z<~Ff+J~A?Le0)4NHOIeq0#?=C|RRKmbzUGe@6I0SaqNC zYU(%T_WCzhW3D|61uv%OTkpnTvptw+@JQ-s4TVskrb|8Lalx4TsGGx+WV~>~zJ5 zJo`&zmr-;lzu2oQ)s%-O~-(^{FI)KH$R-aOA8$%TR9og!!sx-#YOn zghIrgE2)0$1jyBp^$8ern7~h#nO}v#`IvleO$BE}De(h=^O`_=$%vEDij$KI531LB z)wUw?&3loSd%MWlLDyFKiu&k1YihjUJ7rk=OzA0y)1TeibzD8fsoa*B0EYp2jR5v$ z)35&y))A$oqyz-OR?UmBgW_iWj#eM!r|3}ht%W44vTl`F;ha{dZ0t{ z9HmlZf2*IRvc^ih*u#f`i=hK;la4^>qAkwkPG@_23J(|~da%zgE;0q2Hh?zGsMY&s zJWp=+n-2H*h5mbcC6Dgc&x<C2Tza&jC2w6hGnGg@tl=Ur z1=QGWKN4aII)q)ZLij7Jw6s{SYC?V7r7>SIas@0n3^zg^F21MOQl(Mfy97B^j>|F-fBbM;og%QK(W@mWkj==DAN zkq4v+TlVB1ji}ukme(4o`ehB+Pp*B_4Fg45slmNmGa?hn3LC`kF!#9Q7x!E2sY~ZE z%)uo-wVLCFv<2>7J$0TbN8!}Uc~%;8ixh1drv;KMX<3$HL!E=sVFL(uwKIy~jQ*08 z6yCiN6a)`ZVI!-)ywH%5k@53;O5?Qw!B9P-zC;ppYqT6rm|FD!z!m>DXfi>LKcdXG zFDxhCD0wSweE8I3|E{?=X@zm_bosQDJcd1!#jkCB@3 z`4tiu{m|92mL|`tXX3I&=GW7gYetV#KK!B*WQ^0>zh(ii)Fk`pEGT>$TWcxl&d;W7 z5jlC)8u8TN_y<=(TopCZKYSLmpF-|itLRp!;rdmtDrTLEVL#)w#QNKJBul<2O}sV* zk`=SR%qx36>gYh<{(*j;I?C>-S((x>cm2qvnFx?(01%s0M}L$3sjnR)mEf;k&Mr)v z(5X1{S?~T5zeBfY_`o|wTp4ZccHRUDNFe`6F#pGPCIc))wlen9 zdJcz8j|@K%W}Oeyiub^`&E8A#_6qSTWcSJai+e@CgI>ZaBIKJ?RDMz>^8Ka^%yBHl z8vO3_CT$PKXSbb3m<=d$k&eImUJeRifvFoqGEa@c3$2gQt*N7-{GqtqNoMQ0N}u{6 z!^t7Dra_odw+#ydgg|}dJAi$Gd7LyY+M^DAzulm&AO_{Dbo=(xI(4yX*ic55EMDz- z@GN`e%FB04sio>Sj?xLP_7`k_>0hT4OVjWvDkDSvpZ7Zo{Ev!QrsbKx1DGS=Xm3{TEEv&}IlJJE`-yecyCH@z}Nmzy> zIL}iTPqvY0*`L~WUh?&q&0odj`nA0=Hb|=fAaBkFi8H(Eb5*$$5Inm^E;_X?9{Wa> zA@bPLi%K^gZsRb&N>3hBN>{F|`eJWC%Mvg%hU@(bw5y0QTZCelc zF}uauOrU<->QAtlK`%;l$K$Eg$KHs|o9;y}Qrn@~UAQTv1p+3^;IbwZN%al!1Csy= zb3Q$o_HVS{S>+p=mNg`hRG6%jLH@H7!V7Q&t}D6M89zo*{}J@6-_>z?KIpR%QN&=O zpya$}01O@E-}P#L+Pf=&05(roV&_${w6=kb&W>*?ASBgR8$@)8q%;vv_?W+(qm(%} ze%SqvEv14Y%<@T{Zd@j8NSTB3S>-sUl~-}%jWpt8MAOrsKN+gD8rwQG-AxD9|2m~d z+q-eHBjYPEX4dGMv!C%LLF>6g2E3p^vxSl#vEL#PpipN-cvUx1m?e!QW9ZA-HPJ&#w+7h zCNw|nf*Q7f*2wyB^n`j8E~shf^Jh?7#>-d7y0n=Y*Jt4*y+FiTXAIBs?%1I9w)e3h z`PcdV3+okl|HviWH_evE6Bw1J0Zq}))a;7`mS)4#0~^|mVe|Nj)PFl<@HM$Rf!$jo zD^)Lk_B)~W=vqFf4;c!>dqS2kLTk@|KLX6c!mP=Arw@$HVu_|pKl87q@~>vF_pTQl z0zSKvu|@F#z3DJEa?roHW{}@(V`8%L5B=F4%VE{4Wznzy1I`%BjDW@~Sxrpomns0?MQ2vq{GR@94}v7fRo^?i zWSg*|skX<9%48$o`D}yP)B%XfZ_-meUjChiYPG1ge7Vx~gw}zt-&O7`&8~vyKq~z9 zPn6uI(&8@MVznz-pB*k~B$a8NF|?mf)W3(sTsCZA#MPC5KkoarnfgiPZ_2S6@(F^Q z&^_w`YS|(AJ53$4#m_e3vX9<Vs?s7`W3hq7 z>3|eX(x;j^Hzi(Ls&-aJvmb^kSP|LiH!|9ZOG*UNsGU_N3}h>Tu5q=jDqy(W1YkDK zB4%5ONDMLW(pjpd^wH#T1&qL^ps?zpsYG3kR-aG4ZRPvkLnVJ_Rcu5J`>GJeaqKs% zY#za|X@Zu!x@gdv*L)UR+xxH@{deRG@`Sf)|Nm(LU?FKuP4tp)F^=cZYh}{<&vY-$ z*AGZSdHAA~pB9u}7-&8M#XBPWC!0nDucqV)k zX&yoe<4>H*xuKjyCxQTZyo9^3-#(;0$9i*rIkoz@oPEU=th`>b#4oK|C}cnWtER_P zS+Jl3ZX#Se5k?v@Yc^=lXky zG~>Dy!Mwj(N_xs#UL}opP2Dt zShq0rH3~VW7`>LNQBEaPWx@ahk4FCifT9C2<3dC@`{`v^nD`|vmlOSK6RRRu;dqdI zOGuicmk+%8FCP!E30$hg&DC8rP3R$KTJ$xGLilR5<1?$%XkYrn>HE@+-EnYsxhOpj z3Z{j}e%5?n^iB9K&dpwX_-A{JpD+lM!t8nS)NKQc_dIg0?s)j5J*hN6QRDVZRkWu* ze;cPYSZhY?wR5J`Od~a7jxh46Ju&p-a;L6YGa=QM^D!2zxeh)m$7 z9=QApGYni>alMJ{L4?4h?f1M*)oe}MNaPpJb0V9cjszz;tF6Wb$PltSUEZy}Z@g@s z#!#Wc+(zmz=suKnfQ+h&4D!be~u53N;7$C z`0+)J9QmffnUit`+^}N@6)SSx0F#XQni(HKE!t}*&#L&>U2B%yqxsm+02usMAar@n zY@KX>PNrtZ3*y%M1?EQm|K>(`M8pcimRpb&@8RJAu6N;&f4$%|qJ9*|R{VR0Eeh3);NDSA^NTngF1F9JqMIcX@d13&{4Qyrk~i zmcV4s6%ELprK0G3y9MAPjSwatM%anSfJBwr9JSSJP^KtHeZ(3LrF8g{FWVhpzD{g_ z`}Mvd`f6rnV&HsSlb}J5TR{Q1AudCLnRDgJMyE|o8+z+VGr>doKA;LXw?SK0Nc=#D z4mBK0MJ2aeSpIae|8s=ZDW+R`%{zy2S(Wd1Z8M|nX1|jdkWE&@udi}{iZy;jP1!88`LZWE;SS+a8*94uQ%Fb2 zDr;LOIerdX@t%!hH3*#8D$k)Wu5_$xmXzAPdu6IRH|QI~=cKs`JW$`d+UH0Tk0#On zB4^6w2k-y^l4;L|2sJfq{a!`w8ER)p%nAMUr=T2qeYy8aqrv2Q6yyLsUH6Ky>HORZ zBBG|w^4L=At$#pOU!aJ|SG4-1aPduR%61%RY=oJXR7+ob*?&?a--xYpbgMR7Vc`^* zm!%&WCnN8BzXu0>lCt2_Gze*H?t!ZNf;S2a(DhZp!rj)!N;x4M@5 zw};fhcBy_mR2;fp+5<^cAJ{!oee`i+L-wasibaVU2*EhNRIZ~f6|A;uuo@Rs+N^lW zaw#j3y_I*ts4kzWKzOmNoY-~p9p0JC_WmoNQ&(Me>ATJ%7Cy-xed@^sEH=~f$|%=N z>yThYu2>+OviG6F`R<3J#uc9{gL8rGFu_P7Jfh9YmP`I~ERD{RUuo6w-m=*1u#oFt zVgP+|B?5(zt@o!3)j#qT!AfM0<BSC`Mic+&#$>Kc zQl`s)jS`ZWJLqzPg6BhOaNUKsg$3l+{eq*g5O8A&fAE#}V0Y-_w{{>nWK~RkZr3_9 zlz#iSl98#4;g(RXvc2(*u#hM$<4@1D5)1A9qfduEabamSYV4u3~Kbw0-_4@r=u=k$6vWdeAa6G6lk~o+-!Ev0h}; z|E1rmTv!pgyAi7WT-~327 z>)4*!IK1Y-OLTo3m2cerpxazzkrN$#Di^vX0cn~)nkQF9*TQh3Q}*%_&Iz$x&Rg1C zjx>&+kH;k4Am0v>b-Cx547pnGdee|wYGDs4!p22c#-w1>cujKjn40|ieA8kEt2NGU z0EGis`SJHWA{6XY?dlp%Q=Amihc@qbVEjSAx(~0pB7Q0Z)6c)R-Btf7d`_P!5x|&K zM`w*sgyt&(!~@au>+Df1ZG+ zmWQ9;0UW;G6vYA%6>AT?nH9XaX-Pk05bO%Nh4Eb}l;2Lrd%YC`$ni2J6ZY>#I4>c2 zb{W5&@Y@mI96AG_tz*H5 zw0`q2Yd0K`6 ztgo6c5;R5{o6UG~E#_fpYUK=%15I!n+BUj0eG>f{ODPvg#>=wFYj7=2nJr0|f^~9S z!Nj$Y@?GEmF-u#Qj4yv@V!9icqVkyQ&X<3Dj0hFHUCCbN!%pe?P=obI&T;v!uBnQw zr{-toxs~HIk6WQ}WL<5oJ%}CzzGrx-MB{c$3XA-u=Q-Fx{_o<8J%|wk0lo!Iu1d4M z4+@$ZH9$&;dp-z!85VRobZT>C;D`9- zb)s9KI!-L#IFrHnYJ(^3=RxZf;}1qbQvspd8;-j*PV40AOO0=4!wH*SZLukH0j`e( zKQ5^Rje)Ye@1_LBjJjj4AzFUWJJu(+$=Zaj(VAG~Bv@Lvr=INJsXTc-Ykocz8w~HR zPO#9JLTc3X<9Kb(os*o{G_nXxXIz)Lsb4p>!_^VZGvl9!9V_jCI{D<|#!z{_b%($V zvzE2J^l7-i+!G^S+qrb5-aFZCc&Q0NrFMMLbUYmp6Ts+@Ub-boO~;z`q?zCz&`hF}J#ZcPGI>9vU}{nGyb4Do($`4v>P~O*f0F zC%E*IoO8>ztJ}9cIL%+x6AZ(2;3rDl{}Fd^CJr97ucBG%@hE%C%0~!&xsOD{!Oef@ z_-sVszplAx1&gFx5}JlZ&o=gJX{H!i9?su|QFZJ+oRB%_7y1jZz;rHFR&4LkOa??+ zIiBo-j~)isM%c&LI(%5p8mKW~13>O2qjMJ1CDm=IjXM(Dp2I*kob*Ka!6URHPnxO(5b#u8SNE@# zhyH-D23C-8%g%oCFF^$Obeb<3WXU&8e#-99&=!BNuaeOP;v!q~{~;&s3L+g}H~0;q zf7}|FA}SEN<{=;+|9>=nV_2S#|Mp$WSXkz&Rm-+*+qUi1vX*VPbhm6S+b!I+uxz{M z`u>jR-?0~aTi11caDGlaMNAz2l#)0=gzgIsAo-7TFZQ2UU2-%S78wZ%Nmoy=qrF`* zhdVJXZ3zN-1tvOAcXuFM{us#h2)v%;BK18eOi4)rBFjw&W@iY%d`vHEj_#9D=l6g= zP(wjQ(NJ%A)v)5L#!DkZ4gmlW=z4CPw4zd4VJEZCJWGnaYWt53(C%KWw;*Uaaq3N4 zRuyd4Zp+4HteBdGsXQ-DjGGTQzKw_VL#M*l@Lh`l7Y9HqGxs;tuY<8n84);H>Y+nU z;m|c0&Z!XQ{VJh8kFU39>;&77@}nt%mh-q@s^|NX45>qm+Ao3k$0=q=C9U2VrO9o;_t`J=7Q`k9e2PPA}fXy}P# z%-fJtYQ~9+*Pyk;03#%Y^jAU>zI{uEjja{ z&Hayv46F+&U$R!mVXr0wIe;OxLb-f-$YlUQSDms+{vlaFw=+b#198RXeKgrHV9E7n zVo`ViCw&WM|?Jrg(wn&yCZiennM|Y$ri~9MzmL_0c!&M1Anio1r0-~?G zv(;7@xV*0LLb6e>-AsTQHjM0Z@7ampS>~bBueRX_Ai`eKW>tCDYPUK0-N%8~9dBH( z%X#%B;vb7hzkiItCKQ;C&kxRZKy0HXu)`=wtA!x!sO?-^uFLH+QKV<9Bc zz8&va$7k~&2LWggxmca(Vke&5=3jq}s7W6Svr^-lraWnE6wy~yk#dbG^# z?m=|Q8y+wfo`y#aE3bT{gMu_e%A8NPnlch^AePtGWQh-b7U(~jhB)uHLVuy&gAQX! z7;LfIGWM@z2DF65he|Te&@!R-U-zJc0B$L8vzcp!7 zPv^m`a5&F@dAjR3bL+V|u}XJsEdMs@^SSGNhk~@mkNz?BqnxqMpQUvpBPD9!44Zr~ ziAo$EAsa7#vH>C%+Y}|8`$I)u_b)uB5$~_u`M-PmonF?s=X`fn;1Ye!CS<|I4dwX2 zQiw=&?F*;McHLu3@ai5BEImF!$AcZhR48yk?pn8XrUo z=x~)l3kos}nMtXWS@yBWo1P%9xNpzYRo4hB=%K$H|3K#Xm>F@$tf)~iUQ3H$`NY=z zxg(!+YEY9xi? U{t+-*mhiWb!OwUqq5e(l*V0Ygx3e*vB7wD1&nog^qbO{d^}7} z1K~i(4^{k%Wl+M|G`l15&_8v>^vGJ0d$0Ig^=Qb!7x6gHA-s7!v@M@5xGlv&cBBP| zZ)pxw%BrfWK)4t?05bq7V}QDZ-8G(=Z_k-~Zz}h=?&_ym=M3gF((~s!#*#b-$Dc}@ zC9g|KS>q^b9-ky~Y<+$Qr@H71fG7a^Um+ZPZY|j4BKPDe;%!5tHSXDaniBMj(7@Kv zS0SIb@8{4rzyD+s_|-80nF_OA4-Mm-t;zL#1a4OISfO8+7$To@I7| z8Iwy1x`7rDY%v)@1*#*>m0jztV#(RxEbZWA;8NQMEl?^VHjW*S`$_A{vWgeLdIE>v zhKe(gOHZ52f$@-Zbx&d8j5=3k*t@!@m|Lh=Aif|@?tFK+Rn7+36uDDmWplT=QnXx; zPyjxHgUN~Se$gH_c|3+TavT`7!PC(YQ}qRd#s>jXr6T77p?9IE5Km0Paq=dcf(#vG zrx7+q^U{Bpg;c8G(LW*5eDe$6`3o!%pk5iXaFtV%Bs~~iDv`HzmS*sji?Lf%m}PoN~JW!4D z**Aa>uK})D9dMZd(Ct}S+lHi>6lTZeh_{C8*uCZ2Z>i7uG08p0GXMD+(J?TjN`C?J zSU%YvGy>uZVm?p%&F)-rf_8^MES9+1@(+OKu+yCVe*pfi+%~5Cj8T<)X}?+7{169dHz2)?+{|Tp=>LR~Xkd?U7PJaw zw*abVRKh_eTPk9#ZNtiHWu{tlXuh#BRFDcSaC|wDEI$XQt;@}eYJ8zzcgz@(47%C4 z49xZ@V{LNIdsap4LLkY#&&+?ga3fMJdMlHhI;bE4Q(WDZw^>q}X#DmwTNykNwsj~^ zb_cWzy)M7EXc4u(xee1rO&I@PmN`gzxwfliFU8l4=4aK6Kk{wYGSr>gfCsrz2wW3e z4{vkJx}YNFE&#qsKl8v&)@U!DSX5C2nS!QaR=8UaqLC95TM{!Q-icT}0$z5O0CDf+ zRxG6qXWm*j)6r3V6)=`rI-Zu$&{|;j7F?%>0%_!WsoCh_X;4ygNR)Vfj-u2AHagj{ z=t?ki-o08rGN8uu>JLo3W}kL8)O>j;!y&XAK%s~Zt>mBfi~d6XvcEZvmd1}(de;_c zC0Y^CTuJUFM=?}l^1ru!J;@obA(h!S|6WurY5zlP6xH^~KA`Vu>E(7;xIZxl++KcO z_ycZD;s6^eO8g2?>^RMrDeSc3e!1XW;4XC-Kwb^x5`hpC0)a>dC9S9aj98TuZ{x& zFEvIu7zP?fCP=!KQaJqL*^IeVzyAvi^i}8k1vZcpnnz8RocYMVdA^Sfx_#9S(6f&S z+sH|nn*bm6@0y47_j=FeR~gxxYuqvQGJs~V;loueuzoOaL*>>%r%H_N)a{qFRc`+f z{*3+!YOW2vTWYyT)3wo4t;e_`gyy~_n@B-!@DoVnsWXs_I6+bM;s!3CkSc`MdU z(01#g)?xE;e_zgSEMUV4ws5U%t25SZItSl|osg!=2yYU@2?|9>ZC(lm4#GMJw8~Mm zs$(&rdhB1wj;y&r6Nsc)w!E|WF^XwVqVe*7EAE!t!k?=3EGD(8ROW)p9=c zf0)pXWyD)7idknIQ3na@Dup4wGB`@eH~-ZANa*-9WA@_CtC-A=rFg|&ZGoHhNv%Kb z8$U_$M3MlHlf$zU>o0d(Ah<|XO^t#ajsnF;_)N8(@VUZP#}&87cS(K5^6ddAz6@ zkVSd~q==Bp3@;Rd#$F{@MII6PgsCJ;WwFdvr$q?0YbAoAAQ=Jvx&?b z|9XE#iB>vi^kOkRt{CZ@YIccZg_sO6{DMUSp6G{Oe1{mfXZS~}y9hzuWFbPe1(xKK zzEmx3y|YTwT>ZPCvX&`bB?W{+S69|wAe7BkWYxaA* z;%O!ZyZ=@K_E~J%h?6UT%E+NU3Ja}mtdMHfZ6z@)t}i!%$rMX8M37ZplG{rMKGjs* z8&~!BTony0^-E)LYsRVU?)psd>mX4h@JI6F2v470Y0F zeCDl$jd5&4utH4=pg@(lk(D9H#w|IzpUl=$IAr&i1P~tvUods$v};4EEj2`!>FXf7 z#$zl$0&V}c3$orw;#W2i4u7Sv&K=?Bi#{Nq_wGbw?=)6X{*DAe`%0lE2w$aWBY<7X z{2i|8dS3Ep-l2%@it~t`4=+Kl@A*mXeow+GyT6m4-N0X? zJo=Ot?!*Hg3Iy#bJo9?`o!zMe924Dz3+xupk8eyg0XQj2wkORp9#NSSMhOM`GKk4z z$*z(mBTcM7#pOF%$$yrZPp>fr>h0Y6C?cOoO2HIwU^rCTr8l;~vQv(v`@dIEtPZrp zY06K5S(9X9et3hbFuhgDY1N18LOCR1++BF$etm{*$jS_^pjTCqPOaTUg8rTnRY3T0 z0L!#l*7}8h*3tdKJaiOCq6+&L?eI(4fhLmE4lcFe&K1$Ti&f?R7Kn-olGAQCVC((k5ft&(%&qaTaC{TTI6wYaNsmJ1f zI%;awDS{(6IlZ4tVBHkZxr-TnXFe1{c9|CtDy;4l(8FNglbZr&^Yiw%`=p?Zw|q4E zm!9<-YyZ`+Xh_bTMbkS_F$yApOv9MXbo$m8?~^9AQ;mOOsLH>@svziz6(1XQYPlY` zc>O-WiLkBBttQ08;rf%D$9S`<^S)nMvu0F?MfD# zWSWn9KtX51P$1@03DB>MVu4Mn4?#nbAWKSh1k++scekVChJ8!zLz=>Jhjz(8kg`sR zPF;)ZsT@!bEkt;6`|oqA>OsB#q9tD7vBC++GTTQEST+#Syk6$BgFo6t_(!)9fDC18 z614u03-BF&VsEOQQ|3?b6jDx7>yA#Y2vdJ$sKNFpKTYD(=ub-T2jebzV z8mWSoZEzL`GX)ZplcaO0O#Ip1=6_+T(WQ4gnxQxbkP#566x@)5JkJozwYXnpBbTq)b)?8Qt zY;3sTkq#FMMR<^6pn;nyQtfq0YMyCn`LC*2U#kuBa$FFcX$Sd1Nv5TVvA@mD4z+vY zT&vFSQw|IS7*kB)|DMuUo~&#P)S@Zv@1s<6f0ID;u@hOBV1V98#7`QauaSB=-Mica z&Mn`ycEII*)f{6t6Ij5%NzCyy%41rbv_)kEnps$w2WSYNM2Z7LOnhV5}iNvletLv6AgFpw|En3TIPEU70OEvP9Bt&oyf4X$n zIyv&9R>^UYq2!1(SL% zQxkgM*zXdv0B@bd)+RR7>dry3FIb*QDmHjRArm)PbmE7Iti^mQVxc^nWe_DQVy z26wqVUmiP>>2XG9qRTbWXG%-J|h&*%WRaFTFQujub6?~*EGj-h_kEg7*>*# zy{BBx6YbRF7{77sOFEBrXtCvG07|{_f3<68mwMbvmUENZ>ENetD4?S?)8n--lZZ6( zMHv;RPO=FNRk4JYxo{v7Av{oZ20yL1RMp_9kZ)ROLwv*5T4Q-hKsqSMLW=4J0j3}R zb1pKA2J?!zS1o!BwrIecJ*AW%$#16t81H50+*0C5?h#%6k(6uCT0u)L9!{xT_J6>X z;}8~D+3;n3L40(U;3m7ELAmKZvq58(xSNN!RwBA{<_ zd_3-e>c#o$%z&nXTWJ5CTVgaew%zw|jUe0eyh$xz^?uKYc8gi(aO`gPxp2wgQo|xC-C!*CtfXB9-kO!_7%N$dS! zDR?yJw|=`%AYk;^6M$~D0E~6lK7FKz4n3z63Z)mVYW37o?xf}Q^@$CIPm_Og+nS}4 ze_weXFS8^+nn(%dJ+vwDgl>levxkno%K#jpIO?n(A%L&8OGqlS(Cjbzy54r5x8Zen zusk@3SoToE0sXyI<9`J*^5yAd==-yAov3+z)Ctw_4uo3nwUsQ#kacB-?m%~eoWO*O z=YaditXa_xJFnt%THRCgx!^w;#EwyoGClc9^iE>^J;oBBv{5Q(uk8y+QM-?*>Q8p) z(F7^Y*nAZ{EGJrTKl1PoLHz{}WO=QM{MTiaql-;(hsE@o2-%-@OL4e5Tbx+tqQsU= zMRD8EK}oH6ppsAUP{R0?90wE>W&|Ua*W4MQ`MN~5Ba`@R)l#*borx{)t}|w}6FsGv zqNhXOn99kRw9g%haH3e00v1I(olAEfn*8T%#u@Hkbs4isFgX36d2$C!RM}{}n(Y0; z54E?4Yp3!$o(zw!j2O}(%50)$z)R;lj}(nHE3JaYk|`d4r50PzAMgXWR-8J|#yg0l zk8eKrBT9m$&PSkjCp0H4eB3` zVfeo}d;26ZprbJ{H4i-^nxVMolwB&p0(~#&qMn?UuHTzYH*?s`j_!x~qFhQ#r0{2d zXke=o9I7BnjAv&-0t&zMuucA`7L$-yGX(Rmflbs+3d;e-`?*n%plYN3PJX+bfAD%~ zwt(bag5{UwicxX=N{62qp0Y65!G93z4+z)-KY@-`A#X(DWMXvw^FD{Nad%w>e2~+z zmJ%Mri1BMo@8BfTU!x&_5}pKFTr0vQrWax|HOdq9p7BaVSE`}3>q#_Pwzwck(Z^IUV2q9?hFIfR$vq@&lf)Dh)D0yAz6Q%CQ|V&i|; zsw`K99jjxH0TSgFTDhc8bvEvMB6A9LUvGg$a+V|>%rti}x|yr8J37+>{wE&r223pP z_P(dmb-H4#k+Y@-RzVjl|MTvV}Xy;e;r}E)kk@g z=}qGk^qpOv$H&&wF=U894Y|UB5>BW{V8>GBSwC+(pMU!JdRXidrXUe`A9z}WQI92E z)9Rkn!AN%<(aPPIc#J6{wv{2O@!ZDcS`zpQ>7sBNyDy%}A=JcLC5zk2sJUE{Lz-IaCD}>xl&&5)SJMLVA*gtq23xzD zsQhLqj2O)EAr{k<)_&Ea5_hJKQ(=L($18G%-X>eyhH%JRFiVn$w4uqt!nc2+mK#w; zuR5GGSh5!ez%j{az=x|A<3m)46Om`zcR9+VF*Y2xMz~kJLdvtJnYFPi00@{h zBkIK1v6^uq$c*B_h`mcUr818zcPy)H$dzqaa-wMZ9zYV$3XwSmf9)zn0b(*h=Gw|{ zD>t=i?+?m7s~8R?*Q`^A^w9r$A|B9UX7k*Uu6I8u*i_EIp8GMUI;?E(<_mKz6%+WE4LM7@2YQmN~jTAddnU2-{qJtVN`X%0I!khw-iG#KgZf= zGrpKOdfqQK`V=X=PGu)#dEKMFr8VdK?QHeDd1Ll-4RztVQsiX@H}QK3zxKKEoo~WT zUL(^kcp%TUIW)2D$DEo=DA$eBLI>RrmV*lTltCZ>X9A)Blge>QmIyl_pn&clkD-I9 zQz7Fu*cD}!V2KJzq%5~6qc!S=0v3^m|Ml4mb8%q2Jly=Ei4dZ`Jnu`3*GF(sNr_X1 znL^~yw{kPINK-q3w&$dZm7@R;qLf-Q|D^Ex*B00)nj(|`xqQ;CwIvsfXv{-$Z3<-m zWZ;_E9>w7?LJk5496){E7CA=~_+d=-M!$n%wZK>OV_EUFqfqi|tW9rfbmJM4)#A7h zk-ZyAF`MS!NY0PvN5xrZu^vncUOPUx?t>yBhTIsN8eTwKPFiG$F`Bf2wAxJE{r*sF z{~fygF)MlZSLY2SxEifXG%d8_WK;sY6Hc;`ax?c{oq3Qsp*;bJ3C{sbr-ws#s2XPH zlaup`(Z9%Pg_7KR^)uWnX6Pr4M9-dEwI6T<@AA9+Ry2kLP&d$8?~s`CD}k2}FQ*X! z)T-LYp4KkNthL-e?@F(#c8|nw-1TiTaD|s7KA^R1?L$S?>GQQb&(cN_%zpsY5C?vf zelx?@gZjUx{k;!&ia=Z_>C0}F*WsfdJb9#ph0l4nhCc(Ftgn<>4i%?6w!vH1-qf|0 zR=^txSpCcSs-TdkOIqWcY&w^8anUK;ap9pX=33-D)ZZj$GoY%@8`_tavs0J-s)A!& zGH~a4>{tGT2(m#GDxJ9YwzpZY-V04}Us{yO=Z5|+3<4(KEw}KsiDb>2(n;i{(sUc= zs20(mmD!8t(d5MC0(hTbKwh=6EDH!VgD{$Q$Fx+urdDm-KsTGsq5ttu9k1UQI^dvd z)4@y}L|p#kfuGrh5|{dGTgX@l5$IuYJLwp=c1oS@DVX&2N&Le{;xLg*j96ZfHJ`<_ zP`2Qg>i92FAs(}8u5RJ9d2Kwnv0f$XN#t5)14=g@}8`lp^*)il( z`JT%YdGh?En1CB+h2MtsAK(0COCDY(a5R42sdO15uRxpf+_pHIX=633I^UOEh8nB( zRovA{Sk6#BN{HZpf0EjyX7nq)+98fmWXRBBUmK8y1Q>| zDrYVD(8mwQ#e4#>76Ra0_lHA$giWzAB}KAPe5I5nwb@2>l#NeDDMU4p-4UVe4ew$9 z)=UZ?3ifu^APrx!6@0!s)?LQ(fSy0@Wm`~^ z@v|5X_F(Jr zwWCqJFQaze$^vRn!?@e_EB z#ryeh=icn!<-k7$R#3tSpIj?Xm@vl#qhVaJ{m@do`V z&vXN9P+I%zsGw@u&*M~_Og}y{F?pV8=#*&sMzfFyX&^3fAN21|MXvgAIqj5x z%Yy;-ny}11CDTZnkfh>Mj<+wQ?n6i@9Q`F1<4fD!9)}hi1cNE%VDy2cqYyUHP#~F= z@*}+M^3IKmHs23YUPZcYEZXD3%MutnQ&!yLRZ3#48^IQYt(&;kT>3U{5koVT!glNN zk--zUDt=r#ja-M{6q?~;Y04ROp~M2LogamHpI>`*1{*@{=HV$Bc*&5Nc(G5}u+Lwr zi~^stZd=TmaZwiZh!8SZ7ZA!tTdj|W6s|U%hEtJLBi*%fo(8WNA*TcCYAI8L2np27 zyFql=>JtDHk4sS}bQ`YF+uHRErY^t1B>jOKBNMxtzjI8*IjppE?OacTQ3+Z3c&v1z0vt;m=u#IvLR`IS{%Lv|!TvsTg#WfCZZWhmprR61SEEll}pB&Bphz zbZj2GKmZk%3uLqlPsb(u76C02F`f!kgRTCmFq$*UM4aG=Wm$;@mxI_t~gp@Ls2 zpq3A5eWFG+1v=Q&cPL@2YU2%U-gFUC{MEDl+=|wCummbd2a^%M)`TeS{16+MGLyK? z*3(;$cN`Tz61YHhf=P>a8sV8Yx_bYO~i;bq3)wDQ# z0D9QUmp2QN_(}bxmqGllmT+$85_YcrV1`(=IZ49 z_4S@(K9HshzW({RY^$Z~H5lpk*++!LfnmDA=3a6(a<6qi!}a8N$Ea@82OmT-sf|zV zXFG9^2fSLwhO>>DHjqIwP5_C;L000R9I*&EdipjIJ(~SV4O-&|kR8(i1$=}2G%7$! zq|pIU=hS!dHb4?Lohj0@9BCtNRb5ahnL%0g#i6Hbj+*8ttQNkE8{@vPrmrk8@q8Tp zBO=dIQFOE?XP6laZ+)Q+I*7%7G4;U(aWBtwA{z+1<)#2JH1MrH`NP#GRyP^pPWhj} zKhKBr{=PVpcMfHIQp0D<^75Ad&v4t%uN|#@%$d`SAXYbqY3dm;1HCc5CLP{rBU?GSql8z{a>bONdvVmkb@wN2NkN+RS4bTcZyN~2Z-tR-5jM3r2-q2D!8oo?q&>~AK; zmuQ5jQt{Joiq94=8^YjkSQ2pPq( zU^wG*J!oudby92%o*LJlaV|*v?Me0n)oS+q!_RflI_m_}srq8Ap{6br2)N$FS${ev zV5cq_xUbN_p{c@b-wx72G@0xEi4LLM8X}9}qF&Fs%K{MG_MlzQ(D=FwBF>=Q zS_3)XC|LvWCmmfwWu7bb7Pw=VM^%@L@~A1Yz}%5dLXVTvJH{eR>lWSvttUSu7~^u0 zzVhgX2wq8=Mue9maPbtql>G?D#I!{t*0fJF|7di{^3o)#ynQ=Co|p19%0STT&z&K# zqY^$G$kW~?Y%7};RmDaXYkr-;;dW~%-3z6&N3AZI_)SY zoXT|sGj>v_w3eT)DhTXWQg6*Y}t6rz9Wmio5jKvolXQQ;!DUUtM0Kc|^aX zUh#y?Qm+J}%>n|p=_I!ZJ4QcuA!9Evgd*&A4#?QVTYvPJRfw5_<0 z09pkq*lAqn)KQmCWl#ngt2W6%+2-e#LTm3G(@>xz3@H5v8DnG6h2k;&E@0}#zvag_ zxlY1Og-g${FOm>y>CYT4B@Oft93dxK3PrZ~C7!cByp7?^0R=Ldb38&7QdU2!ijU`0lyp{|++ zUl&v1^{uL2y+11c$$b+SO{&Fn?%l{&w`GidA|Dw5Uw)mm;e+tm$TkZhqSK2yH@tHG z(<#jRMc1tC!AmX8uPtqrmTI!k>x-mnJho@V_eFS&1L(GGC!r$m9D!H54C$%N^hiUR zo={)B6!!+os7i=8#>;I1hJa@?u^)=?=_VNNSag5Lr3e^uzNS>Asx`K24avfZVseo zdI=Y1n@_4MYnafZoIw>GB$k566|fD{x^o>_6W{*OX~_xfCq*-E#5lYl1TzBLLz>B` z*OSWcMWb65gK^f$*vZ+%xt+4ocke5kcsy=o=m`04$3lvq%V?0_tzNe_^&#xc6ia4_L*md|ofJIy`l6oKU zIbeo&Saz%0QKf@0nALQH93*X_bVUJgDVmOT*4_^5aHSFSdzcqaOawP&=&0 z_pHZ<8=A`wr{K32cigd5nTGG}PW%5W_iB!?q(wE;HC!kaxZ z{vC(U{1vCToS^cQJE1%ne48@`4QkVcBM?-X{;K@tzZm z%?Ez9QwCUm{K7s;l|o)Y#&@&*?0Bh+y^R+dlDs&rzdagdPi#t({u<9hZ9{XWcE9vKAL*NY#AOI%#{F|?s;ZfT1_7ps68F%Yr$?v8# z#CE^Cr{8)X{=SK{q5fJ7-@~A|NHrL81oc@#&!vrSN9mbY{{|RhhYi(VFyfefFm?fn zx)hI+?E3gM^Rge+XQWvO?t{grk*rUux17c45@LLmo1CK$5U+}cB8@j35-#Aq+WR`jX3aC3j_ zK_ozg2TRvrEEE$Pno) z{|CQvpQ@R1^D^ngdVY4PSC%g6ySx^F@7O1gRZ#*!jppYknBt;k2;XW}`4NSCKdHBO@!L>8)Qo%g`anL&#yF z>KHgkftEhnhR?e6kWH4+P^Iefe^rT7udmh$#8KAA!^&SWFrn*6SW+W{l0F9iac!Xe zr}&*zAPf#fmED^5o6$k%vouQPaNpiF>t5#d{_vaQfCw;%k9}xa#OkEn$*Z%uKWYl% zKqpQM8r%8aZ&k!3@a}wc_CZ|ko8JagoxC+CZ;`x^aVC^9df|~>{6hrUyH@;9`Z+CZ za{acAK|Q_Xfc?Dv#m=d}Et_q%=J9QXWQxd;I2Y`b69`N|K%PaQdg6*-=Gv=kw zE<$r4aOdd)N#ulSD-3=K?2Q2eoo%fb(y8s}4If9jAb-NHd~RppOd_555=}fibfK!o zW`^7`hEN9p<62_hIX(^;;ke|H0T5K{Db=8!9k2jqe4cSa?;lp@suFWhF%7eR0C#9b z=8x+jTx4p&=`NPkWfSelVDzU(SG74sb+Ho0kF?vcI?y}quS2**-2D{<{*Mb#9Gl6Z zxlH9kmlvaf_}fE1iDxT%``bxzxK0P;FkvpzOoc2GRr=bgqzyxuy{e)m)^+Evu?xs4 zX>3}Fe{C+Yrp=P1Q5Wh9bOJ18R)7E@2y~%GW2Z_uHHl&V z=t~u8g9KI&I;_t7P0(4_>uhTPJz0H=yRi@=%`+jw_GBEr6T}{B`tga0xK}r=#bogM zSXBqR6$X^V82Br!n3!N;2^J(gT>z2yLu0qA+k^4(_PNw}Unp=QFta+$y9%UbWV1eW zSce}J`B13A=(fQX69PVaI&}+V@3DUVG8PYD93=RIv7veQN(_E}YEw;G`@l)!_Hny^ z9-p`_IK|Bf=z zGl!FMc_Uy(^8BUhAJgpkGH!0L#@Z+s$D83K&v{0g5A5p7+~j zqxY+Erq`DjPKVV-hc(Xlh>YRB6DoM8ltJMm;LZ765)e*cqMEY5=@TNDG=J70&u&TS zZjK1|zN}TE!uVS}m}9s9zyfCP0-H}R4!V1TNvcT1t zIJD1=zv*)qf9yRK1lSJWTxvI?-BY5ucP^G}zKMk(S~ND|{BVn@p+C2Gec|1>5AzIpfozae0t`j$ zI-0X4&-?nGCEbpv%QB}7WB;<>QfhHENB8RqZi(-^bGkeeW0l`Ku18@9o&H%kX%mds zU6B@_RkWBU#Qs~o8;4!L(={Y#DLIOfDQ&j7M;gp+CO18Hcysh9ADg zv)oPFW>US{XhE5JI4a-nK+Ewq71;VpQYMs_yYuo{knc-%6rG89UXhJ6$XZc8YP<&l zR5Vk0QC~#mfTScV9(HgshORu3SG<3LWfRm@C9t1D1nS#g~--$0lVuem7h7T zaJ(Xxkae%Rw87cbl-+N#zAfXNH7ro6sgmQTDM{!uA2mGUy3GBPGt>+B=qd%J)IP;e zyGS)+fMrLPfsLYJoDo<;#`E+W9T!6cy$RKl&>fRUsp<)z`5rd7j&G}zATd>@po9?) z;5{iN0y&rV=iIZ!fGM;>tTIn}KlBwOb(lBc{z?b$S<>fR@#0oEUzQ9MZQg&Kc$_+Q z&#`{K@QNw zjxmdJy>LAqi1}@VCK|lbXz?#!H_8G&9<2XPA?6!I7RL8}4;Jz#z%$YqRHOGkDWn8} z66!M1K5uZS3NM<^G}~8n-6V)Kk!q0yW#?1RCJAwMl()2=$0TStIe72|a>(^hCOMmE zpuF4_>KoXxa)e2|J)c6~$Ix#=d;7;NOJ@EakES0T2aJmxOmorMg))+AeDy36bq{F_ zT*`s74DcZf-M}*#qAH#kWv-hpRi{wfZEai~1p0kZ~-6v^+6KyZJ|<#fd*=GsuI*3uaeA`ni;=to1|Y zT)Zhc-=A`2lz(u{xO#Z0=x=&ft(NJ;@W5x40l(StA{;^z$|0&x8rl%%cF^(XSe6GV zb|`O!ElLp%bc}_|rPB4$zd=S)jWh~sBEvQwuYYnx93}|pW!G~*QRZ`Lh_#QOXv)!7`nZ!yR0(M73S&Ch8Y_{|I`58arFFHTbuj8cvC(o~36yxNC5eQZ;D{YE4Ea zD&IJj8C3_Os?$vUeMKnL6m zEebfB>K4dYAe7ylEDpSaqY3~}h-L|642?VtTd~jMu=~tuyHu^Cd#V(F|sGxu`lPOgNwO)Uh=&|e|keIeuu_tF-$9oTz5pm#u)BmsN5uYn$mIFLZ zHrmR5bQ}@hXK(EP(Tn5_xrU}iO^iE_l`F^)^vA?XumeO!g$AN~TIZ`Gm_Om#*eXrJ z{v~<#9<}4_sEEaV`Ycf>f?r9w`t##LP6q2Ud`95CrQJV^>GESWe!=061Dd6=@ww9; zxS$e_o`>TvQ$7S_u`D+P#OPbcKA(q%-+1X=qckoJ)n~E>e&`6>bbPj}@`dY43g_1p${VFgH3> zVf_fF!Wf~*XR1iOzwA;d!&YUhk%x-X($a3%!nwcp_VyAI(D%W`Qt^iE?lgjeZ$Ng* zA0WM(3Ar-LRo z^S5q_O#g2hT2~D{8eD9hlm!6LIkpkfd^aPvX~j$C6@^tw@{|jbDLwmz@<9IYh~AKB zHA;QT%$wAHe7-H953%En2L{H(2{y713*P z|J*?fyJqEz{t<4z9a1&SYBrkWd>hA}qjK-tLEh`E<0GMwGQfK$XZ=+L%=EUj+>ng{ z6kA+?k3(I?9A&D@Kw@aPp7PmB3*SXD06?vUOK7%+0 z%iBkNYYe(4xSflsdKV?))1%)J0?mB*lw=d&$PGiC9-JpQ@q~UeoR+bWgFyOxzZP!I z0G^84xtsg}^y(w07BVYUDzE&K)r?eDlXluoR!mw*=H|sTO}FcrL8ogRg=7gt6g~*d zjj)727_Gvg17lird1ezwhYe^16d_0RL0`&<8u&4|OYoUK*YVJ~ODK<$G^{(iYL7D} z(2*w#1GOBtV|BXAW##Ke@${Q5k5}D+x~DsZjb!d-=Oi#_|p1Moo_!SIY5!hkg86&*`<*i$8z zB+{Q3x6N5@&X``T$}aJNg>C!4`#G7;|9${JCpFqs@E%E@h$y5uR(M>Hc3FUjP_jtYj!?o97?Gf5yIL2s4^RurZ8+82i zrMDS2=pjFuuSmhb-N(%|diM#Sql9Tnq~Nmo$^#hv(>MEUl``>)t$nNOtP()e-T?0nSHi5#c-&u#4%;ku$Dj~a>F6pA zK62WmrOPtz9qA4hpD+H!q#Kx9v+kmb_^dW-S5#Ts#wXT*lYGepC0~Ge8M>Y726T)j zea{ZE-V7|zkBt)?tA*n}cu{`Sr)|i*enJIvAe_f3uzDNQjU%7r8IXF1NRisc8scFI z#b`Xp4IwibJaSfUkyn>3KO5~SIiC&!vMc*he37Yc%W80WwV2Yw z=eYLOS7|1pw+oMUvnjwk1&M-XzU;$)JD8`0@9jyKhl~O7`}6zLZnEYetK9O~V5znN zg(Kq+x#u6X`y6T)4@6r3kF2kb$|`E!eE|`a2I-PUy1NAFlI{ipDd`SL>5^1Hx?8$I zy1PNTyZdhZzTb^?&mSyY>pgPLerL}-^UO1|m$C2->pU`1Aa64o)myxpN6bLqfvVf< zhFyJeM2L`%pwm{SQo$({P=+A-tk$8f`L^jtsSU39`}==vWOYiZ7P)l`V-=VGr@HKN z>Cg|A@`;%yp(w_(hr3xmvr=s6HCVxP@%)ASStFwaDeyn2}5W15G6#Ep%W1D&k~N1mgQombePlS zE606J5(mDGho`I+gnboqK`7PkgYsdckU1)W_yCVhHrG?+pN%V0rIv&B$zl(1XDx?s z2w;0N;5j<2~^XhUp0tuEB^na`kEU8k_3INCx*>mhfB2 z)sd$8Tr?DY$}sxRBd)jwB^)+Lcs3K^JFefSblz9*uqzB|8TT>sFs9tHg~*F`VQjwS!Zb-l7R+kjS^rrPS~ins!VpnE>M2n z#m=Hx6%wLee?dd+-BX3 zkbLCZvwZajHlFaHt`H67uO9p9q4!YL3!^E+(l^sN&Izohjk@w^+WFKMYh76?X+Qjy zN3I#4*4~rb$s*uP*Qv@dT<7%EQTQsLK=_E}L0+T-lNyG4gZ7D@CeK&sZO$`3yT*y7 z@2f(2Ks{4K(N}?%7hBL5533}GpwfOAbXpcHN6FXOpufH_x*Lg|sFJ3xev?QY(8B$o z-3To&l}AU`kPsdHfikXs5gh{AqF3BK(ayZ!#uK0$2~NBn`Fz44W1WZQcH-_EqDztk zZEPIRqAcBQKIoi4LG@)ov7uh)5*`DFCjL4um5^lcG^XS~v7q{XzZoS_L@YD$g6 zRw8;BJ!;x7nx-12AJs)gqZh`KT=Fev#MJku^-R2bED-Hwj!@9mASYv>hPwkT7$=IS ztNP+u%QwaKS27qMZ~d#^|11xB$YI>WXwJzxy=qM!d+i@u>C#Wdg1L)urm~J_*>b(Z zEUzt2O=|*Nxox7yoj;<1e~Ow;{JMF#rL8uMw?rlqnUR1nhF z%@zx4sC$$bpj|`11!sn)u3iXq%V3DTr9`eUL&Uu(YEy)ITmNbF_|w*W)>*CU*v2hr z(%V_YbjtLL80T}!2e$W2{3yZnZ)_E|G#jo$(x~rGczXwwn30A0jkU<3n|D=JF;U>pKDc zH2Q%pirkB&*I&9*`kvv9iSE^3`m_Pfo}WH=oFWr!r9vA-)WLqBp)Mq5JaK(jGn2k# zmtGa#5UO>89vVUs6qD8x_T?@T2y`n1zSmr{HeOIQ!;ZkUpdCFXe3tpROWGa;${<_q zFZ*>7A!KgIfDX~yi{*fy*fRXmdo?3HOLdUrmF-=_+}7~+IQ zE3Wdi-+L@SXTa~p6%xf1o@y~OP+H7M*_}Rxf|g;lYO&B=%_1Mck5J@wB<_IFI^@2gr=)u`5#3Y<6DNA(0t{vpjrsfeIA-M!_$YG zUSi2~dTX6Xm4>UaGnAlHWeB9L>P(vVDxgGAyXMm7VJFC9dg{VRq5UXz*u&ar!JU@9 z_rtnkNXMw{oM8hR`~o1W|kD-4Z@F3SGWnj}T}1Mt>&=WM?u)b~M+0w)yoM{8{L$Uk>C;Iqr83ISYuK zRlfAGOGU(6pt(Ekn|;(y14l}-J?4K1b!K5U@l3!oOo*(Wwzjj3;OUJ;SK%M%YqSmt zfQ@1B{Cmg-78{~nHQ-QX@s1@N54LoRel_eCUM$7=?!(&!;ZI{X1_6+_r)0`lJ7&{| z{^Ck=^icRA&D_|2({{57v{(VCGNA3UKP0trCi>~6+LUFYBhP*+@0+d^n$S0ErhiPF zP4{=(uW^BXOX;w*K<|P1UPbh-Ki9YA!2D@FJJ$88d6CAe$y?^=rtOWs7I@liF4>L8 zRXAwo@LvdK6r(Mqdd{O8F@d4$dw*%jZZevQF@EVY0EqUl4aXF4&aR5SAuDL1P0gDz zTL`&U2Czs=tnM5Z|kFl)Wv8n0)2v)&`d8P|i}?o3$9u)yb> zj2Q);2n~LCW_Y#?*1K)GQfR>8C3ZRC0lNH_srs7Gn}Q#DxjeN(ia2i=L3hWLp^r|M zz(~y%AE+;O82c}-Ujtchf@k@Ygy<(Hv+27O6eu7JumqYvpRf8QaK6L~=G;^AS}XGb z##6#L>u=(J*c;S%=~&us^tG|Bp{O&7;-JjWPQTh-sGtGHw*+^H%U zUU4DN4hx)Z*dH{HeABlT&C*~ffaW0%ceQo7a=WXoV;D9su%tix$s8Ei>SsJ8c17W4yl${LOIef9+(a{MsjJjp0-`+_8 zDV3h(ZoT;sW?LQ{1jxq7L!O&OKv?hjiSSiOMq@-2Z-uo-ot$P#%lP+ z_6xUiP8ud!gKryF@%|Q^+|Z|9ZsZazO~)%J=9=RXUx)F!^V!rTit71%Ssg{ePgVe~OYm1&#Ok^te(pu3p-hXd)8vpAt7T%L` z12-`Q$5RkqTbDNOadBm%GbmLTOM67Z`i9RA!?nnw2SFQNuQL{%#6#sIl>lOfQm9^{ zac`^N%{&W(TTgg2oUAwJYCu_Wn1BZ~0odXrReU9cTG#>>bTZ0^VxYegL)M^8 z;Np$9xekoI7t97L$QQ1Br@HhigAy=3uO=`+6^-#$trVVy!-9F5Lvw8hx74>tWBfQj>&+VxX3 zi@zze6d2(FIt)IpvBL%pNu#WAxE6ATfPf8EZE}nSngLp>=_s1F^o|~AK&yvqD4VpZ zMt~qDm3s&k1v0@qM`w7W)Ow}Ws(m>sho54jZBT2bK)}^B$25dy;_oIAv4!d;i`*Uf z2k7$YUe$CiKtbww`V$wi6d9Pg6ZPe6@;eXPzCLHCga+iF3g{FS3t#6pyxz<{8(T)u z1ZuZ=6EpH@s)0>86_t7*hwM}UjW6;0a+>FNCAIt3cG7~;IicXd&+g&m#XFZPwa|Ei ziuf@rq(bLv&&$6_tACGfO9zZ}^1GaW6Rs^jEnF|_8+?RbpkwtFT4)DpE|!8Z;cxQH zbu;z2k)Kh}5JzN0Os(gH0{1zgc$u%IP#$#kR&U@HE_;`^etI`QB9;F*Z<6S_#8wf> z)R7y*-o@bs-@CRVe?g10U|1{jMa(QUPftNgd8;%1UnHtm40+bJWMaX;_p^Pw1XZYs zN{84fvz!r=;6RcA#eP%_i}aLrHnvFEA4nfe^^tE{{$unTQ|Kyf4$(N`uZg;JV1|o6 zC-w|n-DqS_XrWQ6*A_}>MGY8y*sM}{>5*1TSKQE!i_C!+h;hg4B?g9~EDS`G58TMO zpG3qgYY)qAz>=4+m^;qaKcwT|4^zxTzYvw&AJ!6>1Ea=!U?Jv<8$d@$TxZAIso4bo zhpqD$Hcy6Zk;TZ*MA9&3Q2(z7oxmR}>DX4eOkOVubAhnewXlOZ-&z5t^(HNWX*lM~4)omNoFzr^-B$|YD9KRM-O4S0 z+GMG;rMVhClGeu9NWyUlVzLSCAETdU726faoJQV3P68$VDI)v(J9uw@oUcuXj=OHS47^ zYSt`G+PhfE8ph+}VsyOug6B9NBtfXR5XM<-wz{pA(Sse%2YQKXKn&mBY@fO|bcR9m` z*&p*I2@oN-@mPSjA^PxBp;U@{Gln;~m~OcEm9l<5(g}$*Aqzo6osda%-DiCVFv;Nb zt3O`&(~KigsU|#y_}+VM5>b>#tsm{bURcb+>J1zAk?bigkKaXntm`4c7cox|TO{u)_`4^}w@;(xlp1s+_ z1Em&?y~qKe0{)}?^0}b`U@joPkKn#ivJ?zDGB9%hs;5ItUFveSM7meg+ve{@Q1n$V zX}RZC=C_t?kL>qQHqvC537o3Tk8)<{r{qCk2wlqabE z7RQ8^#2T!T#X<=U6H>F+OggKfpOO;+6L z-;f%(iJt=WY0yeGO)!-|5X3Ay0ALmR;>B6H!EeEw9;!|4@JCbl7QuHF`}hP~t@}B! z7Z*J>rT(BfB^!`*DQir1?WmCUSU@=1r8JgJ1Q-(=V9cDi$mPNRvP)a*q+wHsPNX+l z`ypo{lA*u{(>zxlhnd7)a4iInv241M(iclJ7WSrXw0&jBYgL^7w#eajg_HW*#@Xuf z_`opW7<4u1@n00iD#$0Vs$CU7;MaM|)e37F2eUscR{c0aU8uOpeT2kLiAbW|<(t}k z*LKiBK|m>}!NS<noj2Kq%9`T9DUbBRQ1c^66vU^hc35*k{2>qX zWF9NDnSC~ce424eCAXfT&V_^Vu8?3O>DC?0Vgf)F&=?4Qt4H4(hrzs_#D~B)gcge_ zPXe}|4sLEIOF8-P=6?;}2D34CfmsGSw4{PWicq*;YcIZCXQYK~fwp`yBSe}|WiY6y zeusZJ7X4Fk3*%N0D8Vv3_rq}}Dt|5)cDYM-&m(S+t+yTDCxC?z%;GQbkFY!F2DUv_ zINv$EcdWdXM-an$p{I>WrV1~6fPdB^pLQ{bx#eVVi41XBiHf)dG5S5_xx!aNJ~hXF zibYG#7_r`%(7Jln-Jq~8#Kd(K0Su;OvLEAcqGMasLmgR}5UI52Ras+?F zyq)0W16t%LO=zy8p0F{FGJ zwc!0+tdmkmn6krBm6D{F+r)aXuQu6rQ*C5%Ff(Ohx!C!-FGmPdHcH>;mk|65 zaym+q7D2iGk5lVW0T|6`E$xy+u4&|3Uys-MhxZi85TkH*WxjJ_X}$0{L3;B1YNVYr zVO>wMNsZKfkG1o#!wySh@(uIH$T<47la8!Lw=CX>zn9N4Khwb#icJ69b(-qOUWH(m z=?PQ*W}+Ld6Q}`8V=>qvysC5+;>_wu#Uu!)t^y&ZAl$7Z zO76X}rq^8d7^G+tEpE9rnSN)lGm7VQ)Uc4Kx9XJsYsh#58xGR?ZgJUOL)SGFwS8@w z*}D|B9KYQ)gfw!dIc_m9OED0>UQoe*fKK+9#D5|=?~5QTXHNir)O*;Wj6%ys*Wr&| zE#{9|-MgX;$!h2f)4s0^PABb>m}uC9<4WVe4RY3YM-yTtBiYpnufH+>62yjtndaGl zg>exbew)dTK>VwmH>^!4{01?1B`hrUE#+@D{5dA%*tj1`+8k9Di>VinO*ai;d=_E2 zdeV=>B>u`oc6uG9Wo|eRqw2qj*{DfBmj4s$O zmA;|*P9mM6c)Qb=t^93qIc?aMvuO_nHqzj;?K%4{ssAbFz3)a9+djJ~9bNkvjlNs6oue{Gc)d^A{`?He@0MXT!l)yJ-yB6%3< zSd+J#YkrBvWA51OP-n87Y4!+64%-x26YXini!jK~>dDQg-!8ud=HJ2y6 z56W~AFyUf!M8LHeG?Q?+uu@K_a(nqu*Ut5lHBykgDMw#xumX)J3Uh~13c=G=;l+cI zOr|gD82k3(fXeSBY>#%OQ-|$}!5!Qd6M{&FWO&jq(fufPn4022ObpZyp2~fwaV*B` z>==+4iT*v=qv(`#boKhRwMwtrb#1cT^W^q6l;e#(8(F8b67sP;VG2yfjs86w|8)xr zMVlpczv>r@@B^u2&@+0qUE-8s`&ls_gk)u8D z+@!FZgs5aSGWue)8MBBkzw@!jXCv05Gz8j;**KzLNU3HC>g?5Yl1o9lKd;sNOe5Uy zHH+)$8Y=6#U3D|rj`FVaPpv`{uK7Dcp_8-KEXc#&bT`-ZbL|QgX(J&;)^?&v49Oz! zuYx~rSWukfg+igjsBN$6nE&rr-4p-^B?Ta0e=;=%lC?!+mJ2J+B$=beGi1-{d7Bn~ zmu}41Id;4LPqqGMZ{h=U>GCTV(^lX_i4v+du^8=~b!v}i5K6>#ja7H?E@&E8Y8}_V z1TbtmEWXCfr@Q?umx-Gl^#NfEht4ze70LS~^T|PM`o+Q)m6ICW{mDc&wf^aG z!N8E+J(-D_Cr`r*)AwE6$eAqa>a}oXXob0TY81029=z4*A@7+q6dwJf=A+;H9kH-s zo2@J3h(Gml`radBCo!>TDb|TP9ODkdY%)S2GC?s9*}9c?epKKm`5;bFHvxYU-sB$z z365^d<%K}p+o@pm^M3z!04H2~gCo%e?~gk=f>6A{F$12?go0ikPmd2r&3A|HM@=F^>_pV-@fXLeA<)ip4jKjhIVyz*>m^cq>rP#B=f95W2VE;SQR zbNb$}Cmd_5r@EPbEZwag*=RPz@SsvCec3?4gkX2-myXBH-feTvl9cV3J@eAOKY~v% zG?&GE{rkLYAO%Fk{(ZBmp~BhPIT0>NR^_U&kyr1uWT;8Igj<8I(nn3>hpv0NcCKzH zA9~B^&1J^x4kP~I>Jsh2w?q`nxjke~wE=jOzqE45}VWnwjSD~1{-(u8TZsCt=kkc8feIDMFo{_$%o!i3y z>t?m|V7ckU?iAmFUZ(WjNYA^pZcE-c%8F*Qb-4Sh6MWe&&y|FTl-?c|diTC@HK%+U zn{}O+W)^!fx7W-F!`YPI5~tDLg&CEZm5Z{fDKr*MYRv_uq6)))dLhx}u$YmjV*2Q3 zu~|CaL*10~cW=ED`zYjAa1n(*JdT2vELIiQOHw3R5H(OV!RZz!Gdi%0(^;=^?>coJ%@9Uo%w2S@y3;p_UeO~$6 zMx^b_k^1|K<(*c=Xku*WNKTB$z9RXFXJE?q-uu6fBajVG_RAM0PhKuA^~}-p#VT_! z&GiQdn((^a&CS`?`1ttUk&E7lWcbh9o1aza!NyxdDVvcqC64%qCGk&-XJ<)4hDuW2 zb*DB|y6$V47f%H|^(`gS6%>T@=)<|jCHYUgc-ae;8h4YvWq!IbxsMiyca`=WSc;J! z*fo#87fwq$!T6fTW`)T8P&SrLt2%fRAR_~MMPA~v+A&T! zrXkzp52WR6-B9X5a`wUg#rA?$(VO^9>{p+(O`J58RI|K|)Nr00cA^NmCv7FGtfCbK zxk>!-nL+_}fH#Tj_5ge3Y|e0-xT!Q25}aNqje%;U?a-aW`Z;ZuEFZt5jEdL{ zFFG2iI<165mny+Esz{}!4PTSmp8mj{4fskhuiF#Vx<#2rxO2DLbss}ZxE0x8d~d3m zWFs4Pl5r!ujFc?wX~a0)m@yVr`xQ6ricrp%LUwkrrM)~-Ob)=Z2d(xW@}=(v;^ZqG z7~Oh11MfrIWStdc^&iP9mte*>|BIs#5r^6U{mVC*+mokcT8psmF2jtHY8x`v+i zI@%9|>uaY+yrvuoCPSIus~>9>zj~Njhy~gQSNdO(_=d0$Jw>Qdv^j?UHU9Z_(0~an z8wpWlu4>`V{O79V?z`cFrI&sNzz+4iA1+e)-HAMo+n{(qua_hga3hMnfy-H}>?%1w>#Fpr$dh#Eo8OCI{cjjxg(ONyi>^&+k3*~5fuYw-6Lp~jK`lDJ$g<9eT zgknMI-;1Om;7CbI_6`nOSzAx-IEY6ONXp6{?CDPA&#wp3QQ^k!@svzVk}MQY=HWEMUCD-rRd4juqpUUuO>@k+2+ z{MI7p`m6{Hf387Vu{fm8R8SaQ0Fd%?5zD(jy><@aGtBnjUUVdit%D4-adlgm8lwse zlPce628T(VHI55;Q=gU?oAfI{63qa_39R_fRau#>Xs)h41J1Ypk(}&xH@{q-HJ;`uUJZuWd|e)B0OKY=CRzLSIcJB1?*2Zy)rPo{S*tjToW7KPx^5XMw3 z-SwKjR~pXWg{37Y-QV*Gi(Yi-_^TgJef5{T+1e~6IM(o942#Fd-W&M2Xf*Q1*g=m_ zI)SxY2j{d3r5z?cK#?mJ^d7#NTTpq>-ueP0W8<_mZY`KWq%hA4f z1W8kDarzXY4fzMwV4?NGX7utgKd&8cPUZQEe8k5Lk|9*OScrD{WYOIzw; z_5@;am^Z%xC&4*g%iW_S5LA|-v7fq3YX8X$uEs*?p?ySg)?LEV$-C5J zKyrFfk1h=2cbzbH97`5!a+pcnD+uIHCFwP|UX)R35G`@QA`VK>C97HD1GkMKX z@86GlrQOXVHNVnzm3~DlVd)U&XAUG=V4}7jFJI5R3}`OrR5rFv3Elakf`49G;<4X9 zpc1&?TH37X!gyhwQUz=f;v)(+H`!k3oz2dk2=NBq(XU?+%J}NolJc%lS1aZ!%D#pe z7d zn1b3WkwZtdP<^2`YY7JoH(UK(96w8bBJ`mfdET-G9c)#BN0YHdc2lQ^kN2-JC$}u< zfA}!eY6|QcfzXZXy+-vt2gJUYG9vgYYMrj1ItD{eBk%6MdZ)za5V=@P1rQFqc|1rA z1k407tKdM1v7h7j2szaz5&d7ay;mu zg}ps)W-_+2qw%RJOchn`c*=9OHa~+0M%Mk9mWy{(gHaiix3oL-eKGZ4Qcc(iS+{uj zM>?_+_=?4diHmwG&9F;GDD-Fzj@gYqewr%?!uG1T%f^3FcUmLvmvg$e%U$-=qq_;s z?uKI03vWRIGy_hA*yiW!9T{UHJ-uzTP!lYueUZL|@ns6>hjj)i5|XTP+MlSaDm#B$ zA&`NuJlJJj=wxXlvdrES6ZU?z4z~NbOZ|hapF-pM&{g!H4)3fQ>W`#~`L>yBxv>wX zW3fQ~TI{g_AEXkW2Uh&ZSaT&m4pTJffh$cKB#k``GSrO(dJYt}D6y@ON7&FNvOzhg5RCnaLyL zt2fF`V-1xHWqCDaHd1OEqp`1IPhXMv{jeW(7$vW`uNPnWiWafZrxwxlil3f#OmBHc z+M8Lf#{7!D@#29wRcze$VcSbsLd+?BH6FI*buxR+pzDlDar3)~*_dE|)K$}CkY0;< zFUWVYQ5Pu+Bw}wYx17RG#VmOavJSmAt-|2KCt0WaK&$8T^@DkOxQ+qLtv#CXgstR) z4`Pu23nr)=dZsbG&wY7pCvBO#JH335ggR_nny4KVpy0Cq3cO0}$zuW$*fi=vwP4IvN&*Khe2S*K-|`0UHp7 zFP@30Jo005^>!EvvWx(KOyLd=>IKrIdR;Z)p7p{F;b9(O37>Y#T6{X&HM{Q4|y)-4vf%-bFCTN`WoJAOs-WiA1il_cIZk@T!+n>otvh7>H}dB@`aXuB57e| z=g-=HP0w25GYW*YoA0@O<72Al8Vb||VZ3>pz_otJPv2$M1}wi(H|0-kKp zuZ#iDq^}H|nSit`g9f=Up2Mj4wubamV&O;(aj0HImDsw%F2$lfoe7+`$rv#bFG7rQ zdk?_GI0gwIK1aLnC+UdjeV?ZAF05S8FWv<_^AxE)rrVTk{n`Cg@=raorNj<;I1OKy@^qzCzZ1Kc9skWJF+!w_aE+`kbI8kk zZ2hi+m~qfC?jB%^7=H69uD{-}4h0Z(5T3!w=E$s}R{f z-Ol`^cx&@&(OSC*e<5RUpu`u}01iA|&1#i@y;(-V$rAp}7fTV9@g@Sdwf&#r2bwh? z>@p;G{7vM0Oi>^{6Ou>rBfB~Y=^vPul5$^W-fGKM**ZsRO0~sen}CD(*h=$12qrKT zti`+{QPq&ch76$56j1q3BvHQ?#T}L}4W&#OCFic52kQ?>gas}Iqa_~E8kLHv^(wxQ z3V^m~g-zw}`$E*RWd1VEnoDGNF{m-MCTl&B*a%~q#%~^QGg?5X{bYh|UEHgo0LC6Y zT*PL^5an|dXn64t$de$;9E85Vp%YhGOJURe?boZVv@@9#WG`y*R`AiZ9f$Las+4TN zgvmLIirVi$oVCvW4`HRaP>`7+R%@pxfph<;p4$wtbrlMS6iA?CgcO^##j??a=jLly z7y>mhCYSiSN8kVV(bFO?b*2bsMHs8iLSOCX1 z^DBpUy}Q&A3eqL`Y`?&#fI3f{YDZ1Ba6L&ACKZ3U!fJf@tKxiw!zA`yV+KuAHgm2$ z()EX201=Wb;GUYB88lVXLn(K&e->~b!sJ-fe_JebdbmTNBc}a2T&3$kcEhy2{9e|1 zTlEjwOyPcrRi0FY)Ngs9SJbz7xE;a%-9s(h-dRt=?UbhaAVv`s<4RzgfNap1UI8TF= z_pk&3Fkc{4HuNFV_C;Pej#j)be@oxY4qTU~>C37A(E^koWpq9|YuNr0NT>=-6cgq1 zE<&2{8g{Sj;lEt3au#=y0X|w(^hog;wM1meRVJSH57-gWS*XRIv=z;@fk$g#MJQ3j7gVwlgRPWNg%usd*;qoJ3sz-t!1q%+ zOy9U3m`+{m<@0T~T?*DTM`j1~eAc6He)>uqAKXGAM$L_bC#zlOAGKSQSNj9@XjhuUi9Jv9YXgs}-}?hAHZivq#m zlSe6)wPM}Ah+gf?`XWN}>I6|}7!SYPGw=W|RQ0caM;24IXi?pFJkhp1eRy|e?>CJ6 z=-D#mrIP&DuM-Oj@>9%6E%8UeZ~Y%NQkT#qa~BvAQ?@s$S8rAX&le$9}x4o0Wd)G%F9dR z@7HGFpdmgd9fdL1Qiw=f(#fkanQXar*Tnd(XpJC{0fri1>dxzLm;LDvWNDJ_>fz+O>@a?5{E^I4LiZXg#bXF)O0N~F^@GXWI zHxD%gVh=2O!1H`m908kQU}$Km*|VXSt!$@OEHhdvTa{QDTb6CG22?$C*lc~JC;#~V zp6%zPNi)%i2EO}%pm}i&v-m5j;xv(Y@rqXpCD(TKSJa9s%B|QTJwYlh_u{EFW1XR3 zD!nl?p5(&~uC^p50dDecsMgdjtK`C5DpAj45bDjEJ!QOmqk)0Wlha`qwWaMdU`oGP7*;O@d620OjNj>iV{V8tvxKJ0UZGqB!ATW=IAK+!Z6I? zTwPs)ljUv>x)#Ut_X4h|xL0mk_E}u5S0&r50?2wsHQ$%o7)i{;20$&=9OP;8&>`8Ers6 zX>f3`VL2BS6%`|6=@A1%0zNkORlnZj7}nZuMz3;l1S%<3*4*kfa$7yATC*zpTV_OO z8(O`j(zp?f#bT529wSq;c~*-fsf7v!I*B{u0~k}0ha(Sr{{7Kjx@5R+hKN2}NZAkD zADg7bB}df*>~YT~<19fADe3|bm&<`J>@ez&I<`Tyu4A6?1qxF10^lWy%+VKcaN{+8 zetztDvdcQw*4Cgov*GTrL3~DsG+Kf@CWY5olLvS;wCn(;O4-~hK=j-uO290tQ?Eo_}xiNZ8GnP~NE)Yi6t4F3S2Z(aCOHH&j>3MW4kMVhAB1|*r+ zU~_QLZ+wGE*{T=a5|MX+Q?*x)0Ghn8Gd3G8Jz@CZB1Q%~aZstE#M!Eq!j*zQDIhl> z5J?H{3;7U-3z>QTDyRY5%gI?&SErFXPRMOX&dy#{S$TH6(gFH6tBK~W%GoXyFE4&4 zb1_(4%r5JNZW+~6H)!y8+hdw(qA)dc9n0VJ;5D?9?uEW00eNmtt+DX8k;aoJ{wkeM?~9xkf4&xQ&*5K#()#kBcbsR#&JGykynUQ!>#l-dvx0%TKgqI_xNX-J#?@x=v&|F;hrkHtKcvdPK{3I?I+ zPoSojXV1L=-~X=qG;Y#PySB$|YS30@6B;Lo;5v7K0BDkMW;=(>SS}QUG7Ta6>|NYT z4vgZ?;TR;$dEZIMWmwy3<^i%r-L!8@ba8VV^JYJICCU&Kt6;9ZQr5!n@dl}evSF5$ zsEjaZoeMHE?jm=_lH~qzJcSemY*_t>_Yn=$0ZVa|Ksi&AAPt`m%?xxy&)*NEOg+qC zCKRjG3<(sQyb!Q_;6M}s&&|>>b`fTJg?Tty4w?i;5d{9Wp^LFHpCSD36 zQ3CR@AMl6}MD|S-=0e(k+-cF}MYT5=RQA{Sdw(ooCFeT}2?*g^|FG0$$Y2lJBG@Us z6Y*DMLxh6FzfVDcB=$^81zJks(G(0@ger7X%nFi2z=gxR5wgzUY=G1Ku@uqKfez$@ zpjZQ;!x|=b%AW!WyQN4RzaLaJm)7L&HY&u2J!HYi5AVp#1B;gj)hC#GMaH(fT})Ww zo5aSpIs`%&2;P;6WTQW5p@dFhCquFUYmdHNXor8garF}~&ufT&?7TvoKWe&ZaV88@ zqWT->x&V7b8WRaXE;ecGZhry)#q)A8rQ7TiHiA+=su7*-T?r?AF~Q&5|3aZ65-9|$ zedN$a4T1$1P-LH?6&N2ygOZ_RkTUoJDpfYc$n(R;G6F1$t@t+*-r<_-1D6)Z#X|GEK^*_Sxvyo|DPpH{B}j0f%(9# z7&ndOCvXmdFo}sCmIDe#TpWt&eXu4krTY}Pr`4DN9E|<^q`}-?r(nUK=z*T4O}rnT zq-5vTz{&SjAFZ{M+m#Bq4WK~1!kIFzB-Cf-ZFB`9L+6_qGXftcg-eR1A1DARL)q1a zwEn@p`K{g|U!sL5pqU)4qwX}w9~H`eq5rFLTLc9XkD48T>H3F(=FvS}yhx4Ub_8Zd z1r)<*9rD3xA?qa(5wMU}!+-A{5_SG4c&OsUiF;*L+#7wqJ@OL)0x8PwfqAHd>=#$l z1GNIFBCJH%|8JPCC)%nYCa-T#x(xPx6({==zU0}{jf9&v zfEBY7S(Flk_W)YdmjT7GrQ1{7!QAb@bBMnD!rNyPc*eCpB>%n?lhs?m_jG-%xDrcP zes@~Tk{&(}1TIX6dc;QNL|_ADQcu^;UG}t(?_>(TffQNk&EB6)Dm`d2B_Tl60ksLy z7Y1}I{EeqZw_iOKlEM0XvVX5>T*^k*sp;d&_6W!qFMuha2_v5M1?W10rPObK&b)z0 zrfu{qFdINuLztplb5G|Cbx-e^?xq4Fo0E9xbGJ5rF-8pEPu}0Y77uy}zLzb`gm<40 zdP=njcqv9l%ONsc7o5T}8Daw?aFBT6XCLsq7VK2O53obZN==| z-kQ}^cga4^SOE8{azlWv)R(Af|7}b~jZxxp-dt1n0#{ftp3Xzd{7^3$P#m zt*VPa=eT>ne>AuAl5BpySu)ia<&gJoNKodQ?@2$ab@ekbHHGRnC1-UYaTM*kq6X)) z=!)Cd!4+5}(fofNLyq+*7}KPir$4-V+&gYXDZ17qGn4H3@BhV3kc2nieVvIv2IIr; zs3_@=N`txwpfD|fx*R89N*uj*EfX)V``zUMWo$1^_d=b0u1bVoc~N#ahTcQ9xuVuQ z(4CStv*QW~?g$MpZLxcG;SF0xudQJfE<32g$o-7Kd%!U zeQ$DIwLv4>Kb*mSktzQ>{eZRpEc@m|;ElGvr>mgw@Bl9M-lbf3x{4*i~-m;yQ zfz`h2tJ=}r&FShllX0d>Pw4j%FkDS1QJXu?NrvaH=dfWAeVxqB>-2V)pcgM8Oitf3 zVIWjlqh`K-emnDRZEbdS9<4BAxjt&K$;T67YG zQn&3x@=3I^%GhuLFqvkC`P6-V#u@eByD|Oe-FUEP{CB!T?{{4NU=AO^gh2DFZ>^L- zHF)9?Vu^G8gu=!DlDj!3D#3~fMCn1U8>QwkfewfaAKCh~Z&(Aiv9guz2f#uZ1?^$A z-TvQjZ)?V-)oJx6FHZhLk1ZrOhz;(GHTdNV@SR^Tva_<*H#a#rIrkju$;2bFa&mZh z0P>++=kggBNrvI#1c-x+d7u-r*WdT=8X9tPbOZxXbn4Kj2_RgU|L+Hf{stmIR7WQ# zO_*>zJUs~s;Q8b+UTv4G5WGpKY1NwAbM-eJr6^nJu>(S;MJe)s=mE`&3&N3~g zOVBPGwPEbK#kMNL5#dwmk9_&PV|`@HH@;_A=t)A$ zJq0bT#qIgdc{~_*^vT5J-J1}wqyqy37*ulF|Nn0Dt6{iSDI zK5u1f%|+qY7WMOI(MRgf|1LK-H}w-cq#?rjlwl|MRMG#u^Fi76cO#>UO6Kn9xuGlo zsdpy})k<~g85z^SWV+|m#@(Bp{lq!9?rZu+s_)(mS~@iF+s5|(HXTlNKdckaR@7BRjjpjbhJD-M-$ttsig&Hc{4lg)%CyHyYhIb*EO!?qLR^u zvY(37M7Hdd5>cF^LXM)yzHfuEonu5Ls~Ju=-c zT~DXe1D1!wy;X>>Zsh&Z*oY9(aLOun8Lah#Yvg1EsfdaWMM~pZJ35f3PnW$j?HCkt zjLtUBbM|uK5_W8kI$W-*n2I&Z?Pza@7%bg5?@(mZ?Ch+QUlsdUjPf09hJ>uBFK^u* z#@f~G<&Dp;#T3aAZYHr-#|VXT8fqH1H>^_YrFkPWd`EVn(_Oz$)a2GgnV6WYs<~^m z&Jt_n5Lw-8Pb_=$rk3%= zk)DX#9d9lLVpzMsR{+PH%7df*t=!y-%%xDCo&rT-1XD~b{x;hAeTqVy_|HOH8>YVh^e%%Cquu?UXw5FHcVHu&yXSKnDd*&% zT4qXmjM5k{yFsdwJH%xQj-R>siM&?c-rfod3PS4sbUOLt{QNwyaSJKUCt*lR-mLMV zFc+sC9UNkW)g79Olj~HC1=}x?nnRBE(+1qO1J`YJt`VWRIs-+xh;R%hfJx55lu0L1 zI&FWo!f!p@(HWI-b}OP}%R@fFT#7Z|jrZpI63thQsiRttTGvset%;~U4SHyE2DDyD zY%DWexY~={0}ABn;bCTGR%t#PYejJOLTiA4le_*g8sA{K{N6Pv%Q$aohHyYg=ql&% zyl-@LG`)^R!yDZ@_=i)zp~lL^Y$H~9TUS?CP(bMl-YBIsFDxx($XMO~Mk_j*AgmFf zq^wN1WR<4k?La8GhtOO*!_u>LXcu?4xbLKu3>uZMkVxr_ZqO+wRSoY{669X5uBxo; zNS4XM;clNeSTa9eLR@j;K>G8|tUTT66t0~!VCkeb-n_Xd=97=CkdRQ0K^k+7)2gw^x&r*XNRgW=U2ZyPl(;dxUfVByFRPWqh&LI+T z2H@g>M0yd?)F}5&Wo2bqS+AXZx<;UixOlUV)zx&hs%Ouh)zuO07Yjm2?=Rwl&PtdloU`;J*gyB%RQyVFRwh&`ug*UgBltd*Ir)P#m!ybZ3-A)1qm%` z6-k(nkFVIY)_wF+aB#3d^x}sLh(o-*Ha0fQi3KquiY(Dst=Xm+p0(`xRKNr9R6 z*YZY?_2<_Xn`3c?df7&CZSuCAsh}=DA8cw)g7gU}K&epC;57k-ZYO;Fm^3l zBHtbM)5(*ss5+h{%|72O-97XEz=t+xt0N;L17USrcR$zu8bAZapJ#P-HDGnp)AH)Umh(BZVEX36;uxqZBk<`-qUE>*-#$S87qM>s5C;dU0+c}B1W zhnh&;G``bi*gX;DQXyd=Lg?-71=1#_ct*KeJwsJ#8+h;E$J@iKjhXL5X~*CG0jY}b zqZ4>4KR&y$v!cqa5h4d9ueU)#w5MMK7JE(TAib#@dB$lPfhils&!692oa%*kdmS<6 zH*c;ik7BXve~gTbl$V$H^z=X?4}rtm$xGd;rZ-0m@CwlYI-!>wbm+4}Ab@PT_b3m@T!nwJ|Z=hK7bVIyh)oIo_VsB0OB$rGXxQ zmjD+6`^=47di&ds^XJd++qaKb=p!DlcGtz80`kJnli`@EC~%Z2>JR-G0O6RaTmwOxUjlaO+K8y z>K@y#6`_}1?K^+hzyOk@s2D+I4@g(W$H#*UvvGk7mzSA2yfEHAJUk35%^iiZsPZ_D zZ8?hZ^72Ym^IL$-rYEWMM(%ZcdwbxqPfty$pczhjdb`EOK;1PxC%CXJStyw!(ZC)L zpp@b7tBKahAdyH#Ma45`+?pO9f*+ldmgb9W(oaPCjXlx<{lMj<`@h5E@tv=gL(;_F zKbjaCh#hTpck|cve4$()xQKgAoy&!dTSCnjc5)BpV}49`!bo+ft|1{jXndfPt~pWTDBPJvF}b&o5=|;^G2E zzNDliI3843!(?*irs#W+BCLbL!a{SYC6FBWWd0)$Bve#XYU}E-BAOdL`qF{xi+Jcj z41FZS^75wPAbjDi)z+if4=p0I@WVndH*W9ZCAGD+u^LP}S66C7D2J-5Dppv1YGR`N zZqq{!Vc!s*1?Kc?5{bl`On5Pho%}4JG-p!^YD2U#W1s(1457$SCyy1( zz$%;Q5W{plkt0Dd3VW#7@#9S&Kgvl;=2llT3APyp`T6+;1rWicDSN&=C;E(}g+iga zo+?s%O)Lru3U0lz+ZXsYJw5%^tCLIyxWD1GjtgULFqdK&&M+--sQUX)#kCn-yG9%s z;1q)$lF%A1uEBB8aL~X;qQ2`v&Yc(@3*fHxnZlX1BCuFPG+I^68fMAPY1e_BJ%Vqm z#9Fqo?meUzHbaZ@qN1V{0Z!o*0%x5>F}m43004j-Jth-WLMf?kX>b)B-PG9Fx%Y}+ zL6sgi4kImaYhZ4pqoT}9taoYug2Uk?Gk>Cav9awuefi?~>ynb-ty9T6KI8E^Iy&v+ z9$KxjsD$|V%Iazcw!vB!;^tmF)KQgJRKQIlB0T4$goTC83VbTj95bE6ZtPCv$w{#4 zGMi!BEu~*5hj?Cqa9aI^GFx7z=GJWMXB*wPe!U|_0XF&s1mF@~>FWBMz4?WOypc@+ z48Va7M~Ix|=fcDSXo2)aRJ=||NQCpqx(?PZEG)p{fpV9G#6*@e3uPG^%bcjau?Y#@ z5JL?N3@{iB@BkLq37+omfEbzeD~Cmq$Qh7t=;pZ;Xs`@RPNpAj<*xvj4>f`bko-IY-UB5T1DY*TmS^ z4?R7PlaqBhFYD{Gf{ZMzhJ*Z(#>U1?dX8o{Em~QylM}Uw{jG7>+S(d0Ku#LBd>6_B z)RzWEsZ>KlOlL>OK*_y!2tU}VkJHm#WS=-VI)Cb>NU1_R{W9Qb_XHoT|v=!;vx!<+BB6wIGaA3SozVd6F*4PbR(du4JK zAP0nefsW9fj3~O53HX*0;0NRS!|JAP`}`{~)wYIh2qY7~kN?eP|M)d4ukhpZ5Enf6 ik1zkf1Y3yQ^5(YFhA!gH2HLD?xvXt)u~5tM_kRJF5ZTB8 literal 0 HcmV?d00001 diff --git a/preprocess/pics/raw.png b/preprocess/pics/raw.png new file mode 100644 index 0000000000000000000000000000000000000000..88a4944cedc318b43a8aa126c98f48a67a0f7378 GIT binary patch literal 179507 zcmeEuWmHvB+wB2GL?jiEmZNk@HwYX+N$EaFcXy+R97I5*yQI526zT5n?(Vt^zxVs@ zxMSRZ_wU`~h@S1<+qG9b&z$p_%YYBE5*Sa2o-+tZQco zkr7ss78YT6sbpgKQbyOoMAh2P+@9g(dnkV`P29%KDOpotJ})mxYsnRM!duc?o$B6IODG-=1+*3o{-S+J8Xx?&ZTbZ{L1K z7H&>O##Ps(z;7AIvDDVE*cvoTH_IJ0oYa{MzU-aK9%0?J)Y32~$j+J}5N7;n$U&3! zSU3_Q{EUJD=kZTSqYzmfuO#Ud9nOUp6LX>ebe~<~yEi=3XFT{bGcylCL;oF~LFv^G z{`-rM&;wJ_|9tTgMEn1L27U4W+z;y$4W3u6A>_;vl5_5R^}D-PLrVz>348TtiAr>& zLSENxg(xoT8!ItVMBbXi@Hf6{D|H}fm+Y2(-5KC=&qICB%!&vIq$(=@4Nd6 zbUZ>MBTpxclQ=jz8_pL4Ux|PJ@$U6%5Pq_dl#Gmk$N9cLx!3r_1dsJBzqj|@`1tr8 z??Xs^UU~U_bF#CuWx!RtHmBPYA%d4C+vCOcBrhP-iMQf%xZn}t7Q5}KNtcLyQf4Dh*VAVDVc8c4F@tW^fLc} z09!sjINRX5vTxrfaMsP%$A8q##>mJxR-mn4I#Ee-etp!pwz6^pPllWco0zO;CV83d z((I}@J3A{V3`R0)suaQn%!adSNuYHa3n~4T@O%MYr|t0wdevF4tDW%ha5Cp6DLgVv z&-IZ@cWd9H!^35}i^G+!`@XQ4usdEKpjFJ6_IL@;-^X4TWmbM8tbyv*v~<8ugi)%k zs@k@6AT0`7D0Fi?RBgXDJw457J-fZV?a@{2>gvjMr?F&(Xjv3;YftjYou~6Qxw|6h zt~X>S((ybQsi~>y6TGrG-(PU%`CGLzGRb#`e}=EOi-P6S^shOB7t7BgBsn}U9TH^6 zrl+s_yl;>4D`#t5{qHWxw;tSDJFtCm6FQFgaMhW60_KohJU5Hu<}i=_Dt4%BrquiQ z)s*@6T-LrlhkJg3f(8Wjg{s^1%*s`SW?OHuKKuF(4GnE=ZGp+)eZJuT_CfXENWa{z z-P;9IcBwU})VPm~$8I^0xxr+&A97g^M%T&KwxgqCV8a2-6yx4_oW~DWG4UZjk4;I} z<)!bB?@Dw*-@YC0??(%IUXJ9ds^)8?D`ZniN56djgxqyGbS=@{>hgHw_wV1{H+F?} zhn+e?Xa%2hqi^VpLPJC4KYVy&)Ezzl@9;&%iiPJ_V$EDAc^;^Xbk`^Mp-ihr!*jr2 zoF3dCV708-RfDg$X_=d%cjAvvyw*iiPSKoQh_9#Z`_TOUw70isWo3Q+`c< zvf*;jTcA<>w%%sIe>>G>q?wVy3eNn*hOHiM%}%tlz3oOq^yZ^aM8wFBsHY*-n@r<4)8BbkB%C#6IsLK4Ouzz4>YrvZbC5U1vfRBl|O#0 zz`SO{^dDOCHtPNm!t*bI_%!T-LAA87aO>W{me<^fL6>R0A|!yk#riRd^e@AlOR>is zzJH%YrM%fg`WSM|_ipRIGS4-o7_YLq)`DyID3TA{zYG)7E8ouAM*A~ zHn?^0=@TY|V``znko$1u^F5%Mk)8LPiJo-{PX)Uex%!M9S1? z%BFxS$vpCb`+`1$C=YkkCx~>DqJWEzm2)JWiA6uTzc2S29KpYTU4r}9^$}CMZ)B7W z?=XRJFM2)PW9FlwIRk9O0J6>j#^bHGbDfK7p7-bKG;p1dV0Xu4hkB#Vn1P|@WKd2A zq(AO|k7j4V)Eninqr98x0;gm%R>(0CXk>C@ZF+e%m7-PaNAQ5;kZ-(k>IB%!aD^U^ zfcyLT?(aVjqEducy+0MgAo)*+q9ye?)=ui`(>$(+Yt3x?-d}32whG>=Dc*`H&)zES zJK`FdxRf}yhv+$Y@bP;FYH#fgv9VmzNBRpFR7;OhJtmQTt}shMKY69e$TGb6ed77u za`)|hE0~Kpp5unnpRwcMDPK*#c*4X^)J^&va*Tfe#*HfS)MrjlC{pc-4JViX%A4b6 zbWh)fJiuQKMu*HZ>t>K6zbzbC7w~fH9#GcLLVpZ}R_2m&wu;!JqCjqt!360q{MYrA z8O%Yu1KDy6FLK+$O(E;ey_VtGrOV*NTMQQ;c$&j9E>2%cb9L5y%PD!lI=>{&aOn>3wgVZk%tY*+Nslb#7 zupN|m{xttV{CAJ~+~%eVCcWP>VO`<(4-UT}uNIfAm_P&?^|CpefJa5VloU}5Lg}wP z%)5b=yQYiJ`@j(5Q-AK{ZyUyz~`=4 zElGbDo9u&!PMS6L#G{hoDK4DQmy(|&)gWha=?eY4@vOdS;AW5F;_(q9b%Ncg|*4g>2H-WG1AN;GdoHE+Af$UJw;*PB1gsS|eBg9Ts zE-HEqLJ?BeG>8t-5@HF{N^1THt+dyKD>7S)b$YPi>p^nFt>rG{E@yanxY-};K0|^8 zcC^}abaoaJ6vW~D0dnW3P;eOjSZ2z!#A4`26fBym3&$;64;MO0koS@7dlGxRi60@d z=LeQjJIP01Y`TW$ux8J=d#vf%pQOGn3t*42%izL?Ok?Fpo|z>ev^oP2YyAsTF*<9U z4&iovei;(t-1BS%kZFl?955G@Kr7w1$u3V6KNL3^pUJh%1T1xvjtMBLsBr4HKDD;C zhCX@nM1d}1HTdy-6S=Glq`|;&Of;4{zi_L#Nb58u3SIov}vDG)G%pWB3Z( zP5%#k80xx&RK=FAB?jxW-v7ws*Y+BwCCNj=dMH(Mn=TwTWck#Rp2LhDlZbtW!7Ezk z?Q8pYn)0LXresUSEg!{7yKX*1M^>f&t})b%%r27=zt(>+6i`$d>|WJSPf5)v&6gt< z|C8t1P7l1Ife({4bnUkNQ}o8a}A!Tm?UqAf&c=A zSKFog`T3ceobArMfBznkyl2$1*H?3QUPst(b>!uf`hGbgX!F>J44=vWG@AD}y?ABG7hr)I$7aL!(T;bM!4-?y6biKq+D(0S ztS1?CHaAOn{W~~CRAYN-{4zBU9)z2$V+X=~wzWKZ9t@L|NeqR(GXXLAt%SxbuUhnh zk3ZocZ@(|VWma-d`lpVCiE zyoraflI@O@mB+GgU*l{9mVh=)eJeq3 zu=*ic=4@8!vcXPob%B@JOyZ4bCO>{huMB#eU}J?K4~xwSF+d+ScP7VVAUJF~k$*A| zLhwqe{WMGH!z0KAi0FfT+am}T*Dt_RGG!BpxU5KyYHv)3G7@+lJHtmeHaAPk%TG>E z$=-Z~gtTpKZnB#Hf_VSThO;WFt@Xd{bF>{@#oQ#5l9V*FvXYgQ)GOXkBO}w=D6xRX zHacg(izEJap~`3*5e-sNk5W^|Mr!f&Xqaa8m7^j0=_~C1c0HY(Hjg=&h~w>kRvA;a z9flR>rd>GRxkO;XN-0o~m6QM6uE6{Y+u|sX>x#wIq@J{o7>~gr=VQXvK+uN&T8$1euZ~dMU1D!<)32s3`uZB@L;Bi?y<(vpq5odsf~Nz+s+AT*kxsiGkj^ zI-+*^kDGtPwK}>=BD{Eni7sH}c|mT7Sz5#%bIDeZ5Bb9w?NN8e)XA@N(H1V~`kt6L z2PC|`S|xQZvo!NcTsA1>5h8I-NGBZu^<)fhg%nud>ZLq);jGZIErvq#KjW<=(>UP; zn90TQu|r(i7gVeIOYiypSJODe;s2eD*)ileHa+N7Vvc#grFDWe3}TGz2w zq>xE0C+}O3^djg|5pIJ;>UWb8L-?nE0FJ-!!XG5wY%w-6a^S*0)F1GSIzu*r*SMDi zI=JCLOG{gs>G&c40htZgm=qN>zk-r#@)H6?m>P!-1(}TNi?t*!9S2A~qjscaR-dQQ zh+(AS7dQ)FMl!2*WWDw86ru2G-AlEb0fP^*3~ML8kuU>-h+i_YWoRTH2_pX5O-jNn zyk2zCUWsa-3YxPKgcMrZRqT@z}05letIn$wSAS!Fy+a2)K~Ga z3{&-_DpgXMb{OT(LxjJTdQ(f(?eUtMq#~MBx`_?iQ^kinb>DRAPby|B6XEpL_ z5@ED`FT61BX42~ zl)2Vy!|ftWMPsphN&Y++**$pr-9f#AokA;)KTya#qJc)r*{-y$ra4x+`sJp$YXZQz zTMG30FwBh=8L!O}eb{H3_u;;OdP_&BY+lBZlQm}^c{KY0NbP@n=+#;g{`q<`(5brH zWznOdB_F*%nh*gyt~S89(rb=bz*CUYcaET+z`?nx!R#uI(totBx!cbJOB-O&Y4B+N ziXkW{7)Zd3iH1ff?_PZdNV}gYmojrcOm7hMBg#PV}EZ+{nEe-eT$f zJANr__#izH5fRCdOY#B=hjOk8&>&QSs-k6OWo2OSI;!jB`1pEF=+1QaBhu{~#d-tr z`cooT^=AXsK_via&IrziIdA^DNPV!byiu}kW6C5^7r((N20LB)Szy>YB}UmZ$HxX{ z8>dMrJH@o(;_vMwsxB(1smK@F5YaoH_x?Oz5WmnC2ERN`ucOkAoY6O;mWe#tyxS^y z(QQW0UV%q>Xk5U5s2{i#r9(1qai)>QU0=r{zFcMXnE*$CV@Ksl5umTl8E}BEj4)z*dvDFg%PuullmO`6d4LkXwONYy$z5rV zl2eKtqt}uNjl?1U$A*K}xub|KZczuocofl`E=h3Hz zgbNX7Wy6S%?fr-R=0e$`#dhafaZ)g**Ev0}{NJ%nKZP7)-K->FUsK@Y;{yd{x5jbn zyLehc0x6EGn_Fz8%Faw}Wqv-$q_cJ0=*d{wSz+!WGyfRaH=|y|CtKL}#N^^vLQY`s z^j8}Dw#WTag9bdgP8u(UwKsRzq_HL+ekI9N*QGykI@?yYctjc@mLk+d0l;Pa$gCf;7dBA@2pYD z^>XW#3_c*xKjP#RnojP~?KR57GCpeV<01^vMvs5B@ZIKPz_Z@zHsdvJ9>t+0FNxdHWmO<4H4Sp%L zMBVmju8ZUsNmmSbW)g3| zpmdoC1@oFwT6l4Lk4i%65?IdXq`oy^>^frw{iUPH8_-m9MidO8&M%5l zlMhLZs2tWy7ExvL>uzZDJ%fega`{3xif`ZGj8IdL65Ra3kB35M$C=>A-=?st4R0|^ zQ}SQNy;U*4T|dc=(DAVIpjYK%s7ZJQ@3qq_j&EPxc^B*w!^u-j%h6>z+F6mhJxJ)w zgM>rQe{j(Fc?W1$v1-Y`H z8Rmj#Ph7pkufqHlpOW2Ld@{6`taxdLn?v>BIuR^fTL%Xc0XKHGFabKc&Dy<&+KVPS z)B^NS!-%aP$%pi(+hE-$div|Qs*9J(iN(dw#hpR++gK7r*T*ti`z!y`VMUmVnq|F0 z2-iBBu)GDoo8rW+@2{eI+8H^Wt>yv4j)Nz=|uHe>jq0 z`BmcuPrC`f%&<;PclcPBA4~(D%oMGW)0#zRj~YO1CEE}1^s*St;rwzbCTw=hOcS23 zu|4nsKEpsHc~V1l8sw$M_z37qUVWsz{xC7I`is@Lpr9arvc0t=@6(f$`ir;upE)K= zM;BqwQOdF9P23OO<7#0%&M|eO%rs-0!T68LbNA=@Xm#ii+GB-qct9}5s6poz^5J8=#s`E~b^c z`7J$!Oyeo95~oLRy=~6m{t74NkOP2F0Ke>!p*yKR9SGx2zA?g<*|%Qeu;kWQ^=-e| z>7IHLv66g2~%7ap2dgZ0r(voUC59foP!-SVbpM zV5~3I5%VJ8hCr@{R7j7*&e29vUiRf$jd%~%(~H;Zv~a2@sP^+REw3_*MtdIH&{zt{N4S@1@6 z2YFR*9t6;cFT3m_u5-NSU&LMNx72RW5*~FCObGM=hQ0n`%l>KXZfcDEmYw&YnPw{L zyU5!n^stb>U;?kbDYahB|M65n;f8nz3-ME!mvZ-<8oyh zLK_zPG&-EhUqkh8oFWzPBN&*CoIC?mFF(i3=s&}L53>|gmn~H_4V$x9RjR%EMYzSt zv$*m;!Yyr~j8`#{-F~*Ji{Lu&Ng6Xgw)_%$U-dInRyp+4*1A{UN!)0JEXHx_QyssY zl~;;>L*U15O>7wgEw-XoA~0mDOQ-M3Vfj(eY;#EOY+dt9TVu-yIH6Zr-KP{nMk;8( z76)HNyb2?F@T;M9cgQSA(1)fnpC+!Vc(oeU2X!`q&%f@Vl{i-=zsBo^H>#_Cc zB)>#s9Qnco@wrDq52XTD>1M(_QnGgmyEBj~>)U{~lQL5A7eRt}mq2Me((fwaIv@LG z(`X<&7Hu^UCpqd`gPBJnrn#EieimQ%F!Pg;QMF8t6&*j1RZZfFiRcC42`U>1ie*K! zy3g86$j=bX59#3sKRJ2ib;2zv!})p$O(?^s9+vQrlCD04RW+2vG8PR}QRCvok&DCJ zTJY=j=oPQ$qWo@V` ztx$L|V6)u3uC9z(kt0c=W098?#^!3aYOS39VWXYYvti_BFTrxHjPk*2vdGp=E0t*W zC_+l;!ZAyCFJ#}Hn*=j#sG2mcr$Hc76z24a17@I3`QlG)`EJCdUp{kZ&6+SW`FuK3?suRN129ZQ$~s3{+^VbxyvArVsMR|LQ{H|LWkwK}=43RNv; zDCBfnF{UtqRych`l+bgACVFP5K)2qn=al{&L~ z=FNJgjbl*1{QQb4<84UWWxEdf_t#L+mq#l_raL6IzuGQez3)674DTGXp9&9tj3Ht1 z-(G;`00gXgjEuhc#U8PC_C%#iStcx+sVn$B;zOr$TW9bdP5tCT&AMced-?7a&=7Di zZx(O9%g}1C4P1Ps8I8%{X%@WdFpdOFT_Bz3Bmm`E!>swF<>SI>rj^buUHU{`&tbW= z>npay^oXHa-v*1Kz(p-p^J6ln^oh?|0iZ}>V@ztXzn=zCVL`vmKu8+lbgEHbvEsIyUe=O!#>l-)i{tkG3k%v30zu`%VjOr; zE>zQB%6V|8ag`?IU+%rY_E%FWP69Ku{(AVN;TtaxPImY%A^h`q;ZHG{EffzBh;u#t zjitO{YW&BkuT^|el!K?u60(2{+;K?l&+3FQhtsKEhhTn~)Lxdn@ zAk9Lf7B1vAxqdWkj=%%diC#){mY#+w*lNGUHjKi-QS|AaMnZ0CzBgQ~J6jb*)j17R z*`vT?hirrQv`0`%wjwG9A2!f{>FBVY*2B_oaB!6xS+mA$7AP-XYf9cL@& z``L2M?FT`B;t-rsqjPg`B5}p0ft;s0+z}Fmxjh;%!+T=BScj;~Gvv8h3&Wio%evGW zryL<6tnfmy#%pX_Ay>Za-v16&lq); zJS*{7+ySn>@PQawkxSV})&oR|;*=mpuT9nxYG{rX{>ang1s(@(0zB>mZAY2yic|){ z6dQ8LG{L_LmhHa^7V?d8y#Z9gxI?+_4b-jVjUhH^-xC}(3$yMQ)Nl8|$^sZx?O{nY zhf3v(ouhW@>b(|sJF0}5A4jQk)j_O${yQz3Km4qp?w%ifX&XlasyQwL&N<0M!vet; z7kQGdDz`$@VvshAWRYRep(eBhhxrxsq=S8S8;`V{ry@;_A#-2fRpLm^J|LrEP*66e znUFJo$2@obm21vcb<}99tGvEc6}#%kDDz>bJX zbWJ5VM3rC;`*qq8X_kpy^Fj63e`5luq;_F2n6~H$K$dk!S8bZN`E3+lwoY%7y&BkF-7Rg4)YObc8HSXcM8~LH79*%H`=^OT@ z@E-M+&1>UoNI1;&zK9DkGKTNEtLz?D{E_Xby+R&){n&Gklym12l)YBTX*jcLXRi^3 zGG5GO_2&_%wuKY7jlmiLFKYs;Bew=hO#>zR+}#W`<(&?Eeqb?@5&oFZ3q~hm+anH| zO>@=i<{n$n6c6{Ncf3~FY-d!gbsuper;lkf-a&W7F|yuGr`|%c(ITJWP?CX}2iG~} z7<~>Z_CY+2Ehsh2)VslNxuhOed6G*~-4J^yEGK=Ynda&ikeWpA3)Ab& zfc?+PQXHGtFM~oZbnqV-eMm{GQQbGV%3L`GXsYonVeghvFy6eN{^UoNg!V2JDFh9O z?Mb|-2~#v|C+`4oWO7!wYv_0j^UY9k>$y|nb_`(Bx39j6?k`e?IWWz4Cuc{5bQiS? zRdzGc(b^p@6h30otJUDVi=R+2Sw=*$B@}_Ggqn|PnGnCJzJ*aS(7%?D zU7HNQHICeP)*1qt7W%j5fktm!6URnw^bIA|B-EEm(Q8EL!-ff;*T)6StISDrYNpG$ zImMoQ;O3aWWi9^=O~t{y59}l)ciSn;We+4hfu;eLy%Ei9;8D;Blv5~gL6f;&FBe{{ z?0h(vBCoG3nbI#+E|{)xc^ji^%l3Veint(DaoU6aDq|*^`6GXNKuJ+(C9O1>8 zvHkoS(xmeV)Mt;;xd=u<5XKHwnZkW#rv&0Gqt<<#^+Q;EM@hibNA+tknl(Xklj-pJ z0fdmbw(ms*+Rk+z{CDz6-^=61(p6X|`oJLI1aFHn3e%>2<5@JEtKGYrG{eE6UEz>K zEhx<3udmx^4`H0dRap>M2Sqf%!Cjbp8~y9UiE4i37lplow+2cN(pNlIsA-a_ABmSE81GjYKdoaQs!0aawQsLh9pg(iBAiy1XIQZs?Wk zgBzFAN6^EOa2<1Tu1An69blpGxfKhe9pg~P>xnnNQ~)7)O2O{TLf~aAn~V})?I`DE z_;{T#^0ak?U4)NfGrn3${2wfull-7qZnJS>W6+CU)l)Fk6Sps zecSZM58axaj-Eb_MgMDMxRHh~g|>mY<87KgK%UkCg)Zpk!H_$^p(J^UPa#nDpR%@>J zy6d-8^TuxuCq;ozVNoD|Ng%&5g16OO1AwcYiAY7Em zag4w6moSwdu9W-B?q1#Ir&0j#${2z5D4kke`)eSFG10@L@aFdO$I821W;(p1YC{4@ zJytxzij$9qfJ>v;E08trZ%x^rGdeZz>pXJ6gK3MTHrfq^*sSkNLXUrXdHd?lPBj8W zWwenqdzwMzwyycMEm8K})#CAJD2Ak#(l>Ag#`dTLP87C8SjR@Pd_bBdwK7o@N>k0rFC$tgZ=lT;z zH}$34N6jGemp6#cdRkUo(#{fm(k}I$wVKU5NL=6D6||XeI_eW@bUWFUPjfTjW5N&g zWb&BhP|_aSZeFveR!y|AI>*tr*=~7XkfEtboVtz2?~l2bik#m>9dezO+nEQb zr6@XZ!X?X1dfF`Zt6#f@f`-CHF)mZeL0oNRrJ`*h@<< z*tX++<<4#bh~~+pq73f>KK};FdlbwhHv2c<04UE;LsyjszaUsEmHQ5K#=)bB+5B62 zA{h+S!RO1!po=jdlxU+V$8PyDyQ@I9xQe%BQSE@qQAH4^9M8}+>=}q_k~+afUPlLo zi9rJRH3rMc-7d_*E>&9%EWJXEcbX=Q0y|V5b-_o+@dvQwyt%%=*sNw-;w$VrGOo!+ z=k3DkfPccR{}#w)Kh+Auer0D8e=9w?IB&fAH=~i{+8b%4Tv8JnyMF>t57XO9zSXu~nDAVX4kw4Z&=mz2-LEr>T^L7+` zd-j)A>(Hil^0%!VSWJU-Y81B?Sgq*uzG*zBi}|B&S2ClqFa_%9u{2GFGXkz<$1?3A9Fl)HNW{$ zIf^cz**wW?R$KWY{8m`u7lkXH4bgtvibYEo3 zI&L-lfaaV82pkR^_c4k{K7@*m*u#z|>_ZI0xo%#}G&p@8L)|hiNSMca zWv9jWTMC;XeI`pB~3s)Ns?I20ga zKYYpG;M_JHrdys+o{X#(aajz4ZiP;BHw>7YzqcelJ9J{D49hG}93e2&c{e8u8?+}V zbQS7JDrX(F9;kD^O96(AVs`jZY)jpXt*~eJD6KiETmg(9<>=8}4+)hg&*F%Ag(}sf zzB#_qmbM54yQ<*xa@pQ`wZp~`6x;#X7I9E>0IJ&^r{f@cOzvjbGxJedE)*?!YA_M# z4eunTFMa}KK%9}}X}X=Lb6PPsWqi^=#BTCIULL1(XJaG#1I_i#4KSN&Dk(Xfs>*?F z1UslIK>n!=QRcNXj`FtSDH}N0RwEYGHF38sm;GxiXKIKNfkie1wHx6q<38Al7F{F_ zG^;ANQ0)Sxim4GyO?}~k?&?QDk+g-1GcACf9+H5}GY>aAu+2w*41_}Io}w2Gu_d^l z5Qu+v>a%;dYJVjQh67^VDq{ybv>-;C0NWrZYMMidbP}$xE7*FnK7r=SH&atU%&vU^ z2n|-9gE}X0HO69xQgz8j8G_t#bk-{!c60Xr*0dzPNTSwJMun0gAU~xEFrMXPY0ypa zKJVlL_D}knqm?yGOSVC`d{v$0hS%jfw9bRFAaETSOW3u%5(zLdaMs^t#-F-1%$AW= zD!htEgkx$YRdj%U4ylTyNsQ@d2r{x`zps2dNX=ZfHyAr-F!z&rShpBwr%n#Z>Jb9~ zqEG-Fe~O)0*zy5QYt!QQL=?0^wp1lYJYW(ySwbqwo%#d=e8$qZ50{d|Mh ze-$CIjVHZ9617+JK$Tj{+$hlO=-gx1v|K5-zID49uw}EO@q$Q_1G``s*R%2zdCYHn zHPjCitXHZsH}kcm)Nq{(C%ZSfjl5-6O-pC6nnWd(;+#i>`m3|v0?W5N<;zqiISw3P zcVff4v5@K^p@8t`-LntD)m1#@r$(!oC*c#iWV+&gmf>jr6`Qen-%z+>6}9>pj})Wk zX!4eJYRWF8Jmp8%4+0Y#$YzM0k!_ek`5L6|bgh=|>4Tq-hxVYQ`hTALJa#Xe0tFpJ zyitChob31ah^+8j1dVni=~g8|$lUNmXiVkK-m;XPchovmas$57j%Q;=W$7&g9HXlh zo%IWfeXWrX@F~D}#1OPRfprA_K7U=k>33BZz>O$qoZ;jlX4F<#QUw>?v>8z21Cah#}pq(wn zY(EnzB&fU=uq$<207Ch|$`EXLCA*+sJ0fY+DZ_J$Ly0j*05~T}YVglbn#2x*#Y9~U z0htAJUEUHN1&R&Zc9Q{FY&9R!&fXk>DxB8gqstMLht*mPc~_44m7sh#`&I~A!h(Nz zF4#V`Rrn0z|je z<4$@R9o*2ld3kcLF`ja&GS%HjmwpcFr%Ch{9>(2I7>xL+DD-Y;V={Imt2Z%YNHZ{0 zH03#m=HoRpBB{P?1(xt2g#rfxsGD0+*m_e*+JS_i4Y|dJ*%>V(lS}q><)X zx$JGc2*SiNUNu~T4ksY~wVc)T5FORp9<7bzXWn$G-WZP@ zWbt&6DU|~;C1a%;W2nzbh>$i%fRSuBQk|4Kf}Yt$1`yxoGNHMwcqc=j^3!>D;9chW zqT1_JeyzZW#yhu}I;X>KZ^b~Jrs4dmyr)zm#NNl9>e!8?RJY#&t9YG6PX|7;-ORt3 zvGFoB)<~WhRO*{N?``PE7+~wi3^f7hbL*HK`_i~A8aD5`1QF>#jwf6lfR_s|2QimF ziS$Y~4{P~R_sI*$t=+!HOzN^p-UjQno)$QfD!E|8P|eN3Os%t2`e1Z)GzKxJgL7Ja z6k+rreK&8myQ2mb$B}!Q^tZJ$-1ptc^w7mCBB@93EQ0)iH6-5WF|e24voFE&pD{(E zuzUQ2o1y@3HgEzWC2+3Vdl$b=4r4E++^5NuW)EGeR%HKXJ$Aw?_K+Sqmvb0|n=8n@YDyfvt12C|%=ht><)*qG(iU1rq*%Xv zU9d}yu83F%WuN}7mNg=1^RMdkIgKmoy+)l%+rsjSiiV?JJ~pFn6Em|K!>r`wk0627 zI_*r>I`2j4c)2eAMJt7kC$VG)MSpobWD)BgF~O2<s_MQ-S%+Ln0Z=*e-o3-&!kla3)$B<%-MS3xMQQ8RFm;9QX7%G)>^ zMZDsYi43Fa(hY{DMz!wpcpX%t`~<4VD^fMQlZz3Fr!WJOneE^7_gQ#&2QOn_gwf5T z$Q!75F%|kZ_;}yTf(p-n8N5YE1YWJ-Q@Hie4y|8?clKu$TX?C_)4%a*$D(ccRIbda zn2Rp3-F3sh(k*!JYtqrTo_o;LEY)VV6v)RjHUdmS{aZq~ThgwiQwrB!D3D$BdxVPh zaTIAW=x4{0wNHDdpgB2p9|{o`jPy^;psmV~(onTsqCy2j1N3l};cb zA%RVib)H$dxm}~9qf=9wuW=7t>Iw=9X0Zx|*j5~*L*7iC+4S}%ute3?-|(;vN$7;P z5D@3qsz)jpbpK6y5&8-z9{@#0yK&vS=WM!dicrvz@1Fe`GTvmM==4fX*skVMrq6*<5Bj0ap}Z1!za>bTfPl49Ck@|R8-)5@Cik&}Ks z(uG|XxO`(Cn>)`_r|Ww-a0%n*MgT+@h3Da`&D0fzBd*z3-j9T(J<>We!8yFplHb|D zTv(?G2{#T-=hI<0cl-DU8(uvaYEHsB4kD}RMg;INcyg?a^rRM~5ZTY5H;K}MvP8fZ zE0h(l01TGc1U=%Nl6!358a}#5Hz7HMzwVz<*OISD&S>`%swuh<=9djKpv%*L);?&O zb}Xj;Cnet~_^?1-U0ntkz!(1pRAK=%9MH1Z5k}qC(!#^X2O`ES7M7K5AuC_^_qf{X zcy@|E?B-vsHj~oLE1*!Rbf1omydY4R$soZCn`6;^vYyARhI49Cc{i=z?Ym1k@+tS( zU>EUF7qLYCd+w7}e;~k=+rmpCg+aM`OM!0ek>zQ)3%KX^^Gg7IzcBGMNlA@SQ5!Lw zc9-xA8MfZCs|UAcP?pt?TdIm*-OZ?dIw44k@gM0)dQ_i42^=ie#vS+Tf5()Vi(JSt z##p~gd_^gvwz^_Tgw)vT?P1#*>5xn=x7Qrlm z-s!hv_8tcI=frOj{Sz2?a7f9xf3Ef0<_E~YfCI(tzoUT9Eh3xVVwdLDXR21`Mt+l_ z8sI+=+TC0gzqiYu>;#W4C>au#H{a3rL52#9k_Vh6s|jM?IRkLQSe}=&T4Y z=@avZGpp1V&GEbM0QXW)=~#4ku#aNCa*<%=W7i5k@57j1AZRfNrc(ps*$~URDy4Y#5y0Pp&A*=Hej+I!XzsV6TrCJpONZ&Xy_P||t0j;sF6#(z z9ldpy-FhR9My1%y5O#c+(SN!qS59sj?G>4=-LmYQ(Rk2zsnZ9TRh~NE3^1*2gyBtk z$txRPtsFQH;fKg-^4IX$PG@m2kEu+&32oPS_ zcQB8J8P!j=OoW|Paza>^%mp^&OL=p0_s(iH(k}JJJAVphtc8fE$FNSVU;qWNuIzLi z)yrjpn9w851o3hIY_BZ?1AZz;84GAh9Bf8~7 zUC#tL+SBlltXRVJX%4Tq$pS=CO=7HD&|s$;r2XuuJeqpke?&+;*AmT=xdP3=&`Eu& zC+csh#%Pf>qtLW}S$VnV*^D!9LT^h8B{j7~=8&V46ELKUzr>E_W3T!YBV5X9a+DNT zsvnyJwo=BwsrbHQ=E_Bm8{OH{UC}_(rNqippBIXVUX=u!oATyT0Vczh2LV{zb?9T{YN_rC&ZgWUmLXV8iAih&{k{-cpddZzxxEzK}s2P6Y$J8E|DI zw|e@O9Cr2I{?^E}hEKXXE#Ak3fN^bsVi}mO)r?4t!kTyM3!X}29~@2Gi?f?CMH!30 zHaUTj=}YC0)8FuFx2_)qWCgQv%?`yPpqZ58MCc&!?Dthl@X!FQ8*onZMkOHWh9Gj{ zJe49u998DdLm>0NGjpC_I<3wfCTAl?g3$QPz&0!&cPvvwYEkR)q|Adk9Rp`; z>l7~SdN;?B9A&zQ&FZzpXHWd@_kqjHV_;waV?3v`$xZvraD%!|(PE#YfbrxhiUOTOrQS|~I3bvEdOuyELD*}PC;Bg+m%ZiZK(URQUZ9Cez zB}Yt(;Mr*@*p9lhDrJ^yA-?MU{FU8D6(b5S+51Yj!yAFe()pO?H!U(D7&A%v#lBq&MWTU$k!9jCm+SDd#90aS;uDPEDCKls{$ z0&lvhYeWPVkApV+@3G83pzJkN-Bq2{)T(BJ9ejJYo#3vdXgqmp{SxU93-(K4S+ta| zzSuu#{MAqb#|iT%H5^~eW$z$jD|pMxCpR2cl^aG!-LlFZC)LV>Pb^8xHLy+~>7Pe! zo{D^UDJ+8gQ08$c_Ddw(`KzT*mJeY{ys|47`FRCBvwU&X>Qt1f1zR;k$bE?juL}Ml zOUErV1Ep=l9HgcoGcS)2G+)yQbGlThhWc6jV#ZFF+-e7NzrLUbHUG^3Rr{ly3 z1VK&Ng2Au4aHXj*>GNM4$~U-f$9l?1K1$XgsTGf`C4;kIOhvd8IeojU+{P=OI=7&}`%770I1T z9C)3d8JGk{#5@Q%m5y|QCMaEJt@)r-n%VPsThBq`FF|~t*F(+vj>z;Pv1ZC!9g5nn zGnAKLQP&z5QzJ#XA{WwFv;BJU`s&eGbpx`i%O{l)Z8Vu`gnBgB0=ffpydhHTw{J++ z_ltxX^N)th7P1j4BX)j0pJ_`lKUNjEDfH0Srk^H+Rq zU~tO%`uf3Q{pI=ji<|bli^VGqyT!T#b{4EtOs%KMzKr}OI;C@-8L%sjOx+VbY$Nec zBR}fkC#A2eA}mwVLQ4b+Sd)5VvfIp{#0B+@Zff;(xAjHWWGY-aO}sGEt8>yBKV({3 z+LE2yK!V)705Q3za^6;BJ$1)*Q6*&}g#2x4$r>1m#y=a2g9 z?lsjUCz4l&Z(23GGD8)lq~j?xm!qQ!#R9Q5QR|qPs2#u)(&zU%I@P_*{0is0J`r>G zC~Ntm9Je`BlR-6GQLvj@da8Hoq%*~P$t-%g$oa-yyH*=I69k=#|k-f5;(yDI?2o8z|87T^>DESGYmK3MyQR%_0wjl?m zwhhyCeR|2GQ+ng#ZSS<0Ml|TBL*Qv))v0TgnqaX@lbbzXeOuhpn7nUG1ZsdsG+Rw{ zrXN;3A!N6^jc~}#rjk1}=7RXWBV#MHLHB7uM<}<}Ob42Twe=22Yqy#9h2Py=rh|ze zw^yg#nV?shS+~w^zG~y{as_k*V_;zTh3<6=J(fUJZ0Dz1{2Yefn@hepQhG(DnMhhI zF2?76=!z$hivWe=aZ^F-ni_8PnyY^yklv?roIobk8UH1>ygPqR zm|G(E&z71w6d9E(#(+jx2cG@x`hkinymUz*Fb=+)7Z63VewnARR|qGWx`?meArp82 zYjE!=v!WKYCzV(=EeBr!Rc#2M5HU zIS3cm4U95L;K*`7?LH(1CgsShIBg-@BBvjh!vkz2t~P zF?h+3I?i|V;$AKedg}1%gW$6qOpLuYvn;Lrk^%OynNJpXO(z(%CH16>f&K)epsrdl z%eKS%d8-aWm#a;Se#n@H${HLi=|xALQb3W`j!JsXH-v8Adi0|A#!(vgo)AXiw;Ggy zorM`yV=>q{!DuTw;=#Pxtl6eXf19Lss7-`jEB=11r2Ab|LiYvaFMxp;*Gc-N;{JXO zcXgP&*=98Eda6)g^<9axK8(L)|FJ(8?i1)!tA{RbeBS3$9ey8Hr`=47El^TBmRmr* zsvV~OL0aQ6~W8%0PQ#dG!R&FR}y^2uiqZ!}`SYEy_yl#J}N}hVFjk|*ejo?1} z{azAX^6fyYQ#-x1Di=hDO83^)tJSIuQdzOw>fgDF5XDB@?9T~!rN1JBa+39gLqkKj67Tls-Q8WV9wpk<7IV$c zAO#&*4LYI?-QyAiO;j?h(XMt+Q1(XcOkL;2lb^#!Lr$>cq?^n%57^6n-AA@yU?)7! zAr*}D4OMabmLg51u@wYsc1)@>v6rk-n&k;ltUZ)xaYB<&`yfE`+CQu(qh9pR`Frm+ z+%F@v=sQk8Bda#*#BkH6pwlMgakIzKkDsAb+N>XI%{*h}U_Si8JKuPMTi@N4qNeTZmc(xQ?aOaS;-7`R+3vIW(mt9*oDmt;>Q1|c~mQM#) zGk*Cof(RMv8@yfA`?TN}rTdh+A`;PDk9==+w!Uhdhuk%j){y9hPsQbtK$}2!agmVK z6NZ&aH?|-`RrIttC?l74oG(6}nn_xXH*8ZCW+WOO_^CsvvSH@uYodj3+X5l&qVi85 zrFf3G2~s2)<8dn%8oi{@4x?Vee`^%DJ-M^DSA94mj{Kz)Ox-#*_(IH)q(Bo>YS4N; zt530+`8Te~RfWT3Xm(uVCsx6xWkqtgaUrWQjRd3ooI~>)p zC->-4m6>3p@n(J1n+N>p&`LelE^ush>j$^jl%4Iq=QUC!u`Aa#wz3qbG0rfJP z3-@E6rGlmHu-Rr+)ph)fx%h7#MrkS)B>@CTyu7@n8Xq(gw4zyT=7p9&L&5rc?gsy$qs*d7U#aEyfi06)(NOipRD?EkaL`Z^Jwb^%ed8&*-f-x!TM+#rN zv)_%XPJvXFA|G9#RHS{f_>O9JgNa*iWTazzlUE6eC&G3AF2eUL#t*P6i%GFrC#*$N zQwCJ28t9i`E`6Ak_yPZXTxaZi@e1CEv2WuSf2SOg*oRyFT8IY+90>M@ovUO4h~{F& z^M$?Jd!z2Z$W}}2B2vEm{ep*BZ>%+tbNuq@pO13HYPG2&g>x-eh9j7dpZ1PFP2+!4&4o14?h z9{;d8L=eycL0CcE{r&VzXnxawh>)Jd_O7dLhg@eSePd@m6ertL95`B=ObiE`#`<{@ zch;&z6C%Z;ZB5U=qv;3ECS$1y)6YH-B-Aj@yniK%*eKxrO{OIy^VtfAg zAxN1R0;|hF#jI;fWRi*;zxfGZZP!-wZ2?fhpsL~v%y`VK6N|Ba<>BAGQz5c7e_t7k zW2M*-|9ZXnRpVwriq$LSIR#Fl)uUdt&Jc4{s7-Onsl70Y#64A`V_C@3S^H_wAd}z+ z>i5?n!~NL3T$axBV~mjp36OtT^CxNN@qXDF71Xd9QSvl}K0g!_HdBcL>9Vu;2>yBR z9pF8z-+QsPXJF_wuUDX7`s5yV^4$Pe(EWYbNHms}`|nFhY4h~ze_F7ejg9BcLA^Qy z4)kN=1k7jaYn2}A^2gFD_=RYm>`Q(81RBHJk8qe1Hy7I#svQ65NOq?c&bBfp=Mel2 z$^Xr}i3to%+xyXp8wW?l>x4dniH7_eaB}c>-o7%v2S*T5w)=>DDeRqqCX){C4ruS0 z-dO^*c|ylCQRg`#X9Xtw`56Jq@F3SjNO|hqZO3IqgWXU=Pt#3r$&Pb6|1j0!G>shP zOCwmEPL|U~f9SC+39>`M;!AdrV z8rSf-$UHDR3@S%(-#q0f{6w9ZoP0DxC7?~krqae4`&C5MN;!xpwU{Y`tX_##O?W>E zA8nJIP3iNWPB+U_{USvPTjYe<;JW!RDFd!q*r!cG$JXbdNUHr;0f~rGKtFRsLBTd~ zKIYr~+FM(NWBOzJ&8$+dJy8qVNu2p;7BR(*4HMxa$Em*yr=cd8Gu0HXanc+*jx%AcW zPc3@*f&SY%x1C9-;x?UsMozKL@he8I$Y`FPy!M6oaU9&E*?sfiKI8Ae(;iK%?6x6R zKOAUtqozf3;ljXr5x3`{P+vWeC6@c)tr&xA!o(5wnH+%~lyT}|tspa%9}uHKiD|W{ zMzF7}{+!u?%vvzfNU-Y+*!H1J(Xu(*x1*eE^I)8~%m zk2vEq(g^!`jHi$gc1Q5LX4W%9n z9QnFuh#cX{xA&XfM=(!j1K#7C6i_-vL@$;M8Xn{96`Xcca8SOR6h%UXez+`N8%%Ux z!5O}+2o6p9#23ma&1!0{!7A7>o)vK+LDhK)Llwd^BXJ8o(bkTn_z8G84z!c2(65~c z$|!w@$6v!tUf{J;%?1{2K$s8Z0WKmDQ9rzFd^L@g^sb(4s+~3u)cYm@CPgjS?$r@; zHw<5iV}y|_*pxYO51>L~!^3lC;zHN0;@573eJA%a?9d8^x+p|T(zwZd3 zZ+3%gdJGl<-YLCVIhAyr#BK1MKlTP!#Y&qTu*r_eeiEyPk%u?us2nU;JGz^(P+aFq z3FE(?25Sp`RlkTX1cgebxhA4Iv%07Xnh1{dI^A#;E5=$_vxt@ULw}}L+Hw@5Rzy59 z=_wZGITYFu;|TbZ2E7%Y^1t*c#W+VQbTsEc;-vX(vmPEp!+jPt7NOcaCq^Xi@1+Qf zM9L+OO4X6VU|)40MnmGvGtzzUS99Xg2@U~eu#?nmI@S)H~ znJO+eDp&SCi;Gg;arP+rW%-MsaqRVSmU6xv!9UX*IMJoA4?R1m8%c2XtRz>1s<|Z% zPf}uI(0!=hl~7JNuSTPCj(ASSm)w+DCzf}!Cyr;2yyUbvOGEdL>(ox!wRv-h>w~|h zBVlMCeUSBjCfW5#|IOr4@jhCZ$p$hiK>H9|ZFTTaMeY{`BuLIuvuDF4wHW`G#WQrp}Czl(j3K&UJdB$q5njR3((i(dqR z170D`@7qDn;8Wk*jK4p2_4m~6M`BsH|JB~3IQ&knF82fk|LR7j@4c^om*l=*X6tuD z5Ub#29m&nA5T=+Dbo;axD2;_Dd@4b{$jTm_jo>fse3i53!G%_wi6c|Qo@mSm0=9r- zMaRU{1I5mTj0y4vb99M#?C@K7wI)371fb74a?p3Fvai1l%#D+ci-y@j9e(@0>Nyf; z+t|0@9q(nlXE3<#X+Py=h4d-+4JDylWUDMCR&!$tx5sKypb0o6vU!ks%o(oTh93K` zvzB4e0wJ&a>JzfK;Zbu>H#$6Bck&TL7^Ow@qI9zMrBA-{|FT8gkj8WD1VJtXNc8m`QeBSD#bLd_IM}^FEybJGTq)q(B zr;({pv0_(|YT%4RGG7THw(SG!S0Vj}*7D8>_QwJN;eKnelYU2Yu(Pk;y%iP#i+0hx-)fQBx~G z&{h6tD6-e3O6zl=dZbcd5%GWA(EY-c_At?gb%@jv z?bOVxx4Il+|HwbR=;HhPAo3xymQM#NpHU!%DrNdqULXxXW3ahOS;+q|r7V}}l!v%w z_B5zOSTWXI3HKT8_;^NrL^T6mml50J+}zxBzxylT>0KPH_-zg)fk|Q*!oT;*+Yk5m zU)8K=yuoluK=kLKAf%uDCX%=LS)eD*#)21j(u(D2K-x_$=_7D%!v5QqGE+7^O4ILe zqz{}6UCU}^B2j;3qH|x??vnKu#Whk7&bvun1M~u(3(xRB8uU&>^;_U@QwV`JN^F1d z0vQ<&UPckIi|+Mr8|X*e9&cV17lor5U%mG1*X!unZDk0m+D&|g@7VuEH}2G%ShTyf zG?AUWCg03E--c1TQvSWU_xB@GR75gDag+;}Vqeb-uk1?01>Qx}fttelU%l|#tKDk0tA(`nxS z*7jaN+N8hW-GQkpsW#j>gagF{;Bs-;sCz;o_Prd&Rw&bxm@Vb3o~JvD**q-^qC4Kj zDnBAb)`^)L(~-ZB1v>#6@4>TiKNL>A;Z}+_S`_OGPJBa)mcQn`A>?i^Vp-4wNJCy~ zWs46ODIb(X-zX9=%s=+(nX&0fG1YUXk!%>jFT}P(O3{HC38oCD4x6hAk}ywwl5C|> z%c#VvuQy?HZ*bZe7)s&{%!_oH@*_Zb z(M!!1adrym3AACD;Nz!QRP&fG_nYOi6Th8)-h=>=BW?`nx-?y|)xJ8XM)ZG~K(*_b z^F;>|%h~#a$Pv)$QRj-LAAXt5gwL7w_M3XRKA~A4`+%EbWSs~8uAXg#*?re-{NT!j zR`|_gf=)8(kD9*?X#vW^L0P9qYxBk@x10mZ3b<5ciccY@4bM@P5G84+Q!H0oPawAd zj@htxa}H!jnH^+~emBQi)U*V%N6^Xs7;d|;t#lYrK6UptL1S)KBjoQ$0g0X9%&usQ8G>} zQjME~f#pu3GcP_l{1%zfDTD2iHtd@-PhJ%$sr!rUob`F;bJ`=1Nod2Zy0U}X%tX{@ z#$-@xYBRDmbtDkTa!i%N+5E5w@?Wz>e5TE(t;we<)Hu*7#Y_bDvSOkmh0JEoz?x?9oDB;6JpVLM zF0adK;CPDlZ=qNRQ#}CtXxZS6YjhGW znw8{@*ZIF$fFwLk$HF`*N-P-A;5p^GyQlzV5NAg?>)d%in z94Sw|;qVuQdh->g zuU?@_eMY_flIco@hUu>LBOuS~vNuy&Q&VGn)YGho2!U0VV0}aMLb0e_>J`|wKy%4H zU=SdJ6dJe*P7QlOj*1^BkX;@>Q9FdtfYNW&i&xIm`Q?S@A0?icSkOvbzwfTn|KD+}c?EaN)U zmE+*wd^M^Rwc6CRdXVS&QsP7$bzqUnD@5D_QdxK~+mlcu*z##8mRthO>BD`pe(hE# zYk2|oeXF=l>UTona|s>IAWwKEtYX{n)0;Zm$STf&gWvk&hUM8|-=qcikn%E(+3-K7 zq@-vROD+Cyraj-;Fz*Uo+`t!9JTMOh)L4SmNb4pY`we$ums( z9wTIxs?kXm8QqrW=)tQFn6c6VMAX7@E3Lx(W($>WWpC;RnUDV*8UC?ptvL{C1*)nDI6*NN9>-k zvjRMb3p>Wu`R1VD})faaLuLQ?GF!@ zEl6ztvd|~*XG{Y!PYGjM zq3s+hv*~9O_04AEF;9+NM3lZ5k>E^~ngSW%bE!b2?>@jB6S;IV1MO3n(c7^|ft@i< z>9g{-z6cNi7w(DqU1`om8#MYM12hRE+L+jW7V?%>NB3s9G zI4EM!`?9ZJMV!?5WB)PxKw>hc!Kl{pkztSV(zy@j|0#IG4xaItf(Kp!_guB@}AI)9Wrp}IT@s+o-V8d zDc;#2MZrp-XtG(#OuArG$&e$-L3c$ynXg?f2cva@aKO$w-^ChfM-6{S!rwq`Kc$B! zCS5erSAkxGzNZK%jIw>N%6s-gqe)`+CqKXsa{@97-rqP>#$Vh<<&5oe6Z_res33Kn zhhtDEFv~0n&>m<4kVNYd2}CcN`Fue4)8X&MHpgrMO`fn((`~N`5E2B21n>%PIEltx zRj!d-J(Lcjl@3bKe(;qEHc;SPruDFXkZPw`$|H$>L{{nAiKQ*_91{XzrsqF`9)?)n@a_ ztomVh#_-c=y^;`Dt?`p`A>=6D_poAmAOx(Lv<92q9ebjH8>pH}g?Xg^CaQ~@{261&$ z{jFP;YKf=$N>*qUFW8Vk&gO3x0^t0LlPKju|J6&5(ywg-fd8eGPwX)6T2N1{dOmOa zjm3N7EwF0X#`B5j4fV85Pt?2Usqny5L2IUg`R|&#tVf?e>Y1r6a{Au;(#)4FI_Loz zVc?VUCC+?h&m=Pm1ddL9wwJN9T9|CU?&33gD3{W;bROWK48v%UNQyQX+AVfRG@F83XPH z*GJtgV1dTQWG18~)taUec;dx)JtS-efqWVhgfP_U4T+KdM&1*z(A<~6cboDs;k9wcSu1NA z)AJa-awhmHFDT>j;$4ZFvMuil@PP@SD#3j4V%x*qRhy<4h->}}N`m{(A zMh+~o(H~|VP7!{^`GkyN7E_6LGTz`4%Zkq6wse>5grN@}-Tn$4Vu>Gl<2Z3BTC(r5 zL7EmPd!s+b&E>aifENM8Qg%C6jtNDb*cd4%t$FQYoW9QrF*hw~y)LPe)KS)sxf6#x zZwVnudrY9y;a*=CDqj+?9Nn&-8v>?(W=sM-T3UxNh0n{JB*D&hkpYv&ie^24_p@fF zVXXA?B zHMMM9z-XX1EB`AV2zX59%kmQPh2T{SrL@agJIO*Q}7dYUh)qxVk4T=RI_IK^tlvd{}th;IJ$>q!yO>krLm{Y2%tlm;hS$RV&|N z1`z=t2Bp-~^I`Tf5c2?kK=Fhu9o>6cD=AQ|Yjf2~IiAP)o)_FhA12<3zTmsP@|Wu4V?KPVLwtgFC8n)?3l(ZGBX@~Tga>lw{hGq4Mrmqi1mH3Q)$V<5Y zF2e)Ps{U1=^(bkhTVcZ&r7i(O$5GjXdb7weU6)1mwyIkD$-G z`I)d++UabGh8XsIco6tVp4gqcs)%C0{0Zgr7MoOuw-;;K!dj5)X-wG?D-j0;@Q3(1 z8=FSM7SJI;0s)l9?xeUGlRiAb|Lj`%99l;?O^G-}){)I)HmEl@-T{Fl@V=bb4M;fH zj}m+{5l%G@-2PB^!HoF*xY~XrNxV<9eZ}SWbWte;Fi$QZKzKb2det77;Tb^wPohyX z2^&eP*PM5OjGlMdeiJ2E2w_f5_{zK10ckW$DoK)>ptwBz{#l0+k~nIu+3ct#IxT2E zt0McoVBhw3B74m5_s>&ps9tyr+@*v7&Y zQrVt!6T^p|L#abppp*%+zEmB_@FWI`W)cry!4vmlQCSLio`n9pJf(x8Yjw@=j!;Gu zqKH|dIp77DD6fgxR>_AEw$AIAc9#AVj6(lcn8>KG&Juj0BoZ-l5lC!QaiYElpA0dE zxx9I%oBJYYupY%{MoH1jh*l^ODeK0TohtR2=*w+AQunO5xu+w`IHW z5QH{ixYSWwwzV$b&clX-rmx$*UT?)f_$ak`PnNkRO%ajKjji}wLnr^R4HsicJ*VH} z6G^cwwQej{^n<;>X{Cq!@M%UxwJ{A<2}XfiebGA7R=cDXMS8l~9&f0mzfO@kuD9_r zFVtv`Uw;BDob17ek36`*=iq&*`PA!ls{X=K69qv`UVGD2;I9wSrPOSd+C?!CRf7aM zusi}=vJt+!Z8(99C1Jo5Xg-g?MJ}M$Kw`ZD)E~|8^TOEsA0Dxcv%2N$06r_iJj!eA z&m!PhQTUOLLS$=yD`Wun17_AX_czZ^V@NhKG!EGhuF7ZzOo9+{MJ<5(PX@a{POmBr zCiT07x^B4X^+e=7al>MT+RA`im8`&z@$R?E&!0CNvzdyus^>&4X_@&HL==kOH&FDc zxS2-r2rXpLKIG2^f-r~Pf4mhh@~TINLYzo3&wsq02FT-dAHS3cfOG~Z$soqr&-0tR zbwe<>zmX8FpOZpZ3fRbjj7l`VZ~Y!I=7@fb)Bo_hZivNP=@C@2>2Y)6lBRp*`TfQa z7nl-;@j;>dGiGM=z?Fy#G4M5r=59BrTD*Hwu`3^bJ|1$Gnh}4oOcoYj#S|t*Fs&BU z@skK-dDgUY+$h5im9VI4p0E~*^#4UHyn8G`cwL)Aw)gS9f!ho6XeLGv1iKeaUFmUpu&;!8 z>u)7@5nwH1cTRC<67qX07W(Yhz8DuE=+6rWV+p{rftZ+cwp(D~#(v5Nd|V)voBmHv zKn{8gDM^a%Nm#S-vq{e<+bt!8emfb;W;U1w&OVSTcKdzgMfF-YLU^l3akYZ`y9nK+ zle((zr3x1KHJ49DXje$CHfiaD-1T82I*|r#>J7BB+AH;Yml3U6zdJIKY_%+ zsRL{>KVcHi0(Bd*VDliZGhgzDb|8ixUpPj@lhFs^El|cQCq}PoiA+_*Fv6ZqPL1*5 z+Dp1vDZlC4z;OEfqzCp8iNzxO?HVgd8(@kN-2O&ftugA`wlMjVVA6%wPfbqJAIRte zd)BY2KiuY|Q%{g!Q9 z)9@Q%C2&3q{f-I;;)-C~%o($zV2;KW>@(kgq$80%bBCYEMDVW#`+^v^&3^A{M!efz z3Ir1;$pKfzz?e%8#nSpwL^;o3Ms=dP`SXgYV~I=!Tf2xjHOkm> z%tyvBwPa+^a1@Q_{A`wJ(u;@rNN5&&B)3%Gw3&h!;-WaVNxRhj2(0zw>av@Ro{eJO z0G85cO*NZ~$|5ShclnmN(IASlc<)60j?z6h8I`z8#gL;@dL`z5nd|<()$ov)C>Hzui~KgTD7ujYq94NM_!m`&|G*=T_2T#P(L@ z8Iqg4t;lm`jok0tX*|7G*gC}za>sT<*+Q2v&`F^S>(mzPbRFY5yLW&S{W6Qf42<*) z?=T{HV%>*&Dc};K*O?ub%7Mxj;ELWSX~c-tIjSmb&{&9D|5h;;fZ@cuby` zVhQ@B8eP~e@osXuJm`fP5FzYsL3^(BJ)mZHq0bQgJ82s)gcF?cJJtK7u2|k*SA+{ zbU_o>@zrhS4Y(2%w9s9HyGKE=J{OEg6K$_>M9%C48DR04A|+ZcyKqNg_&x&k<>1AP zMy=)rGee==k1I=z5WgM2iAw`ES^I&u*o7Wo;#p~>)Ed)*sKJ_EBW?axPFK;BC9f=< z8?|><>l$Tj0F68hv`NXXKD*fL0en4$+v@5)UQVP7RE@+6*bkJ(ZOA3fAOe)ImrOj5Skf1n_mR5$U^kW3t z)smqM;D$CaDD3>Ml7hMasoPi~A7b`we#biqh-q)ouT2MOweLS%h^7(%K#;t?$KaI4 zy5zm?sLIY>%Z?E7Jnq3e_q@xHAPloOSg0LLU=;T#1(PSU_$&R7#E%~d?-ZKj~L24lnjXl<|x4$od zkP!X(tNyKP>ywR4Ov&ei@1m3&1+%z+6a|1h4hCTlp$2&;T8Os$>>npkxL&JcxV&(| zi@w7G9!wVZKR)?v>7E^g-ROP(%9@g@p(6&6Vbc8nE1?f71cVAvp3;(tq%b{*HY+{X z#ahMoZh z6CBDYOL*6(8ywo~#~M+lT39Kb&3RPjO0NkgD=&+E6rJrUbWLF175$TB;sfDsWy34^ zOu+2rrZ`T+HUzkQLfsgsQ9OMOb_(Wi>YfGlwMqMZ8mcJetq5vcCqN17A@yz60AcP| zFJ4rd^aN)HvgkFi`COj?nYXR24cyGy+k-Aj`joS-K*}V6(6v-UA@+UtSK_k)Ia7gq zpMjb#lo(fopnlZ%DfgRU)#pt2xec9jB=ub;BYK7UCFq58Uyk@atUt_rIs*|96g?N? zWWviNXNYWy!Y!^i=E)uDNzCJC{hxd{s6i%}l1U|49s08zYYli5Woua#QOLv6TJqvM znq9}oAird|^V_5;O>-TeCyXKBqxA;cm`ebNsdA~W6EU8^zkcxeYK|4h)gNHz^W}!4 z!-CPCUCW|QYqtd8Ob zSy`c5qk@vz@Uk<@k?Dsdh`_T&JO`=WL`MWHl`$^5Fj;SMuBPe=BqW^}1gMrK0_j~z zK%I~8ezks5T2$^FSIQR({cm73w`OYA2XuiVs$?w|GZKViCdOalg|ow)*JZG4V#d>rJv~+ z=Q8Pz)jkEP+{U(gp6Yew2ZX!HfuwMwT;S)I>n?ekj7bp98%pRIbFGeKYq^4}w!`kv zMSd>%Nq``V`WYtlzD^dYr=(=h70L02YtJCi$Ou1P&6TuLW@=eLdauW*msKE%ILy=$ zf{2Tbr@Ci1$K<>_8 z%$>NOTL$PtmT#s-00|p``GMcnev~V1j|t?-_4W$`4{HaxxgwD3dMwYP^zDaEWVl6o z!hd~0qavCaa8!^v^z-zHal(31ZuKfewE9?fXJ|QN#k+E(l`TmDTuF`x_TY!(e^*vq z>=X0`Qpfb0am<5>Vh^W-UaS3O*dSK|932g%l@zQx25P%$Aelfa#W>1Y(ITO4;QJ&0 z1;8fzTb$_|VATDjtHbv}YOtEO#-~rdds6^Y3aa}%TzNVp`!YX7INb8{cAnF~5 z)*Jo+u@T~TRh&U`ARZ^}X$Ze4f4EX;{Y%_szy+0QFqVP3(TrJ^#Fe*pj>~CGgYx6A zNRT4+W^x2lr_ZTa9sH^J`R??#;ZgY|nEm`Y4EWeJiai|*w+%l81mdE}38>_EM!=My z{IsG7cB3Q?C6$7{X;uELM*}dOxx%Rxn4kKC00w-*f20wXR4Jfr=VCNQkN(fuJzC=h zi#kkFvI9XJM;UauQC(FmN^P*`@^D%=XI@SA*IX2pJQO-Uw(SQpYBkRG;AN4fC^uev zy);e?L@Ss3*sp#qp0?KsZ#_j1^il?-1{v^}Zj2CR62Gj%4^b?3flXFD%3~Ok&txQ)B@Fb{VMra=F%#`CiwGXRD27gRi zJolwHF9F0HON(Y)xg=M7${maGssGf^e1(GR=oMu0IxlLc&W7~HZ!>SP_>zm^cHAe3 zYUIrpzU{06VThBK@#{sPii2hAt3-n>-|Vi(wUz|i3Nqi#0mm~u>W@P!ik zHEIXkl7Aux04g9-FDY>axGE}v8|2@qE&qD~ZX{dIuKE4@AEwW8zm?rv7z6m7q(-)% zYHANkNHmt!3mLPv8b%cA#}yp=&{@%Ca$D$Uf_pNB8!;dk4lD?5!spv|y!!@xVBPBq z6>OL7#y270B+Df9X#KcElta#WEo%9qM@#k$HhEbrDiv-l{WJMrpVy=Rh{30YO)SeS z>5f{Kxs+Ki%Gt$c#JHt*?5`KjExreebv^|2tVheh+-ek z!EZ$Rp5)Wk*_OHlZ%fv?se^R*B@LZagtgz*$>k>V*e`$+@8;j_ zub^oY_?eGB{&Hlse?@+o(#d}k1OAGh2P!auR=4Rg_aQNWVWGc!$N2w?LXYOMA!O{~ z&i1)rLP`S2B}=%#$+b9#i=QW~UnkJ)Ic)?4W{?9hP_ADk(^%psKt!kssVKV54Y6d5|$96^dSR?CVYHrd1>?mn-^*nQtSX0`)22 zmJg`!4_D=gVpB*T%~Hn3hIsqk(1{aq_J{WptpZ}KAfn%*{#lM@d2#7vrYFa2xXXZl}mOpKh#o=;QzCj?Qq`aqG`t8aUdk2AT( zN@>*5zq}t~!}36J45&uK4{8xF6#yRqSj30h&rL=_3h8Guu0VQzRcX)}Hu`{Z$NQds z*~#&5-84^oBPy*SqZyn>{(<67x9Oue>Kan<-adJ(=U5@WK~kc`Kcohp|1;n8kT+N^ zLJX@voG14bZy~WJTGDQ}RkUx7vSt{+1Qi|fKVk{^*tWh*b{+$kb7x2U?YO8q$w!j%D3I{d@Kb!Z+l2eszj$0+jR-7gMq*^qO)dE(P1hWZ%M<65`Tlnj# z*)bx=xcSqMXo5^>qRH-u1ERN@Ybfs;RTr7cV672?lt9(Ane$L#wlM2@Zo$Q;iTE$* z*F_?=aXgdg`lLRWYPG;D<-JNZM?d3o2_y1FFt^fUAW=;fP5L`)fg|Wt;><28O=9iL zb{<5}?IpQ5h7rVIk+B^9}YjW{#xM$!8GN z51&3zoxWuc5B-dKlKxykUl%KdFs}>n7(TJr3(l1zbnTJdg(5$}IO=U>mblLH zh(ZrlVDtYls|p$XLyF998lw8sUs|nkV7gf`93w6H4sON_D}+Kba_M^I3d<^^(&kV2UN)r&=rbkLAhj6+#k_uoGxjq)j!( zelhP+XUfwf$+-z_(U*CzJcWj|{PxCOyBJcs;@5<%h(@RwK)+s60%zzhS8_NB+Q(`s zj(y)~%@zS=g_aEvNp#^*o%-rq8aT2AvWN`rECJ2gT4p*eQks7mSqTCN+&};8!N@Bk z$Vpk88b|zT`)TR})wo?ok{l^+)-Wh`N6&H>hET$Mjg6lrUFUy~ zhtX7*FKrsqSrW1w{ru$~(k#xlHzjhTT({^~-#?0i)6{qIrLq*wSOn^k+v7KQpu zNw`PWLYwg)64|ryAg(+JSTB_F?961S!*c^=3o-h&mTDIEK8%ak)7|S=l)2BODi8P5 zv5fB4TG99dT6i#KzXF6P0N-bSH&5UrGs>4=r;P@IAP|TB`$woBO*E_)OBi3d5W+{* zwt{qBZ^ggDK^nGC@;%RGu@v{ue2`EPBK*g__LUyz=RX`QrOzWntqQJ-oIGYHq4u(ur4e8Y4APbl*+qNK1~ z$}>39xbJ%Mh^*^5VYbrkkNg1vnn#erLG*6;A=@4bNMO|O(-9;EbBe|j61-J*9!}75 zi4r;U?qB`)p!2Gt=SToVw32Xl=}T&R=R(iXr7^W2x}=x0HKbyNHahpq4P&TWGRhQx zoc2pjs(!OZ`m>#c*j`rc^a0|<^znEw-JTzEt|`Q$DO$XD}r zb&#TpJ@f6L&-?dY`h6zYIg1(#`TxNDRk` z&rw^XasQA>3gu!K0S;d_Rfv2d)2+Z=;2BjDv6?ffnloq$Tw}O?O49evUHRVrkWfJM zgDxcfVhgDQAvB!15}J>2zb|i(oSWZy2c_#z`iTh zQJ^ZP()Py*iPJ^*TGOd+_>6Rf{<56_oGA^@C09yb7_mviI9Fe~nm1$sWwjtlAWUf`iz_W}O8D>=`l9}_1Y6N9%6fQ)XYX@hsK3}HGX{}P zZfh|eBUk_z0Pi*;4h_&upVy%s0!uR~$sWuWlzvfR^I}W6k>hJ^9LQtwWVkKGm>T>@ zM~C~L85(R@Hb6W6Kax(!!^EpV;N(VsJQ)})>WgI%db~R|(xKV>3ygt9B|=<-25yft z636}T`4rv@@fBR4j3)6*&XQY%j)Of?4aG%#=I zj-6fFwiFHAos%GQ7}LzM;+v^>ol`deOVW*QU*0`ck=P(u-Ic^*@igCchF<-&dpNqTA@_(@GXKF5sAFLd{Yq-w zp57I-NNQA3)DI4IDlF-%fH^k1{bqvY@@R;bL`sAqq!HX1AAle0gzPc^4^j2Q`h@qf z_sHFC^{$R8&JQh`&{D*?$i9Gr_8#FWNeCrA2i6NNn% z=08R7_v&9(@y_SwU@-Ida9)=pgaLC!JMQ^!A8#@d4SfWNTJ(96)ip`)7^bw@01VWL z`4My`y5GEG4Fb~Z>%YuZo4Z~_D)WMNLGuL+Q{v67|3@Fd(9;`r&qTF z^A)^WptD0!WO8Q~Pb0cVtx=GL=#V~ImsP^lfLxpft@8NjrQy3kr)awqIn>7G>pOYD$CVaWUH-%RlmYM|4~G@9BzbjL2t#d z!Ln)jni`3EzKVsT;~C%x19bR0I}6v;Y>mev3HqHR1hfbPw=aclqjO4``Bg5mS%1Ju z>0pX(7J!hJc|@lcF|D(0m(MBDV;CedM-LUETWU*N_$UYt@hhOCt)}QCGeUaP0%T9wF$QT^$VbUo?l!p z-u<$)*BSjVa*_A>dc4k13`v}E4?H8pO0wuCp62+TP2h*CWDR-XsbHWo6_T6W$!z#* zsnpn52fXwElI)H?^HZbU6|HnWHev#uS>aD)kk8m)ln!zOx?6=CV@ig8jQ79acSt!L z$0wVjXWP9fLCFWda*@1DNwbFZ-Oh`AbDWL|R+AdW1hC+fwht);C9oDfd4!)sn3y{B z@t?<)MWn*f1&`|H%+r~AR>uQRCBg}u-VzR*(<(!f`^s23#Om<4ALS6deS?sF-HvQt z2yTt+f$b*6MFja3Ik#bMndGpr5`1$Vi`?xxt&^seI423dZMG;HJExB`RoIjf1#36j zT1r@PD~=m+A^vlk$WE_bwHwF9<&uE?lWdycy)_!xTzp>l-J-R-7(_bP`M+zVOh5Zz zcNLx)z5KyDnZsfCF@Dn;!KdAT_%)1BRc)zePmQCC9Kl-Wu6cRji{1WWrHbW=&?9_X zLoha&z=sg;#40L!wevPbUfqVIX>n;}#1=%ef}7Hg?H(@Vw9SAR0O1SZi!@>w)@P598~XRj$(typanf{3{`T#J+lS_VRVE*pH(xV~#-?yP zm-KJ$z^wdidn`1Gb~hc6zF%8kBB)x2U7l>t+A$$hz@p*|rj^8t^qW^}v0%NXm-E%Et&{YSQdXJdPZt1TY` zNCJkCAhx)RPN5d>cyA3JwgNdBVP3{YvmD*qR7(*G1a{fb)V7isDu(*+<@v;w>8&V` zNUwn5k{iV8a8sjxR{DTTnix?F=MLocgr@(NZROy=szz(HdeY+j;(}TY455Ln(b4|? zkoOe5#+6A)N$pLRHCiQVO&-@Lw;VbHR=)2wj`$xQrOj8FiZ#zO%YD{{7##E!tp=iy zHA>`!df!TbSGbC3J$_+KR)Db=@;ND7Es_ujm&{c_8rUKH+cZ-#0n58-h%mrd$wC!5kvEYFy3Gfph0#}aFfI!2P)jc<$@3u+|?D0iH^ z_v$^Em`VQmF`h*~T(ITxjwXTt6+f0bm1ANOGa6H1*4Y5deW|dWs7z`re(aCfO>X42xf*_~C5`Jx-6=Sn7^JJG-FvAWSK zvc`3izJR{nR0^aH3w)0Px_T4>Gl)rg#mrn<54FX&#=WD>7SbYGfq~EhdpX3f!o1uX`mC2dp-M;LhZDv zM(M#iwaj^YP`)#G7-#KBl7UV|0VJmcCbEQiIlUw`nDCsAHs^^O!xy|=MFYvpiy6Bq zvn@=&l)oGYwxp_VzUMiZjAuA~8K*m6X=laj{Wz1=B=T^q&>fh#IRCJ7WqYf9iKfZ| z{Y;~We7+CJL+LvIcI1?t6saz|m7lTCZ6nuBa3VuKq$$LWhyCRJyihr{Z(C=MA>!-$ zb|1ouxk8k`lhKy-QW0;F9fsK4=u>UA!>VN_r%t&O*HBz-C(3IGe76Gw$lQ>z^&m|pA2C*h6h474!AP_oRq+r{qagQeW zdL281@;A6kjpAY~NZB^$-)?XSR}UhRU>kIp}1DTey+Gp4{Ee;ePOR4y7JDLREt^j}k7 zt|)yr!{YG_FmwcSv1l#!^7vVyh0&1oaCXO!&>Gi(R&ir zgyK*PImwiPJIS+?DH!N9@$>)_!*~b5=WOiqKFRVf$rAD*&)_IwOPjlq??zV*?{c{9 z-Zvcf=8VUj&afYFziy#P6FTnp-v2_mxAVFm*Cn6W6{#QprL?JG%Rq_4G-*(1jw(mk zoto_cTs!~vsYmo(A5O!s%j1FWkTbbjxa8fbfs3@d7ZCU!@HK#>(ahxe@}KAc?De1h z>Nf5GQqF&s!*!=6Ke-`L?2ylRK)_57kJ~D$Q(D;SC-bX7+YCpt*v_@@seAGEQMG=z z90P>}dS_Mjs^Wv^Wp}p$YiW8qL-d=sq3iE>i;2{Uf@;3qD2Hi#L_MHU{sXr70O;xb zc}Ed7Ri>}6rF8>F{J=GjSg2MHIUE(-Me~!K4XpuRcj#rgNTDu4%MsicFGNC*Nopk z)V$(Ur{z*7hgUB|iLgcFX`M|BLBJl9w@aEW9OjX(1>AUW&UYFQ$%~@fe5QAtG|XUSJZm~$N`?0&*!VuE zntx@@7O+rVDEAA1NM{%3q}WCA2@(X~_B#r&FB)Du@-FuGTlT1`&`*w!p9)4GI&{}3 zS;mb0VMszl+U>6_yrgn2j8GW(mexp(ZoU{8>J(X1n8fhR=NE<1#sWN-HW74HqxN~T zjrie!X12<(s`!N_;nA{#H2+tT(!Udx^PJc_Y3LJhK03cvskP%5TM8KK!}Jxf#;_c_-oGxI z+8`KV?0X*wGi>kWLCYpDKbWc$=`dBchU}RC)E>Em_ukht9`k(%8r00;#HVbesvR34 zZqgKv=)M<9uj8b_o$>d%wTW{;yMc9JspSu{7i%!TSSs;jTYMk~O1T}JE&#VuLY z&WOSb#b{<3{WT@Zc?!;O{`E9Th21{E1O8o*?(STW)XmB>=)0vltD`|!a1Cut4`-ri znM2LA@GGMfh9>lFUn~w{;b0%9!n0@4dp0x#V&9)!7okC+JIykv0-w^%KXjQmKQJ_^ z4dMjErd3ayzHPizAJScM`R7pYkA{rEiAWGbEfO*`Axy{g5;}Pq=XrYJU7%CR9FG7Y zHY9nRbA>Z{6yO-R?U}( z{EdSVVq(0m3#K8AkKA5>pVm`qoKE45s2) z*_0{)*a6cRD1GzAjxO#Ylb;e~FyfLC^oxR0+2FcHjX2Jjk@s{iZY-Z?bJ{ap+L=0K zvnSh}Lse-K)h)iWYFEqs8QoZ>1TSNQk5e?-UM^@!E|eYMp;uh$H4;7nl(PHE3dH)y zfee;3++2DByQBZ83TicfrPf&_b9{Xq3x@BfE;i$J+{g+Rxep5sS>0+$9j>Q+km;83 z-ucp-IatPx5O}lQ(Srtk5#X;|Z+Z`k2!WcSFUp6l^_IQjTKdz;`7Q)W{iMeGgV_zE z`!yTYSpY{RVoAP%=c+~HPe!&|p2F95qd*dCdbd?lXsz;+qAa|xDs+^#c%tR_ajy<0o}kJ6G&<8lE+n4 zF9|Nnk~n^s#PwHZJ`;m20&n_)v1`^`!rCXy5t#qu1rRtuLTiy9#OZ@)G~6)Fx0c(z!8zl^`2q@tvP`6xB%UJRwPww}K#4+`D4= z7FHzCpU}C!cPxa+-r(I=9LchAM4E3hjSV{4mP2{X%H^{0G7d8Py{I9Hd7t1(9s2xR zuXnA$gZiju>OeUL8*DK(ZJasX#%%lG^tGz$U^YA?Y%A{t=m@73hTJfOoCeI=GZibv z(89Fa_;&}-;yH=Q=~thTI(WUraJLhcpJ&b~8Vq9q&sb->DMRr4b}M;kbhNSzv!23G znfJ|ZFU!VkG;)#kLhDvV5jOHyUeCkLl#is&0iitV4t^Sb{_P8>wDc3PyfiiB(jwo> zp#VdQc|y#!V12Xt#K9rX-~<0f;9y}UzK zP+)DFn7(QYa#yFaQPQfr8IaD>Ew;oRX<}T{jOlAYF7?WxIAj7FGAT?mB~0E}Fe_9u z#Y!dVbFEr6eSQX!6r6{)N=PvGHj(UdYcthDb?fz68((#@r;uQOYTYi%wet0|VFSHe z%!qi{kkDyP>yHjNZt&Oydt!!~#nQUWR^1GU;Msp{Gam5Fsuhc*)+78ffCT-Rt{gOt z0mnQ{cZw+2KA^cAALgPFGH(%_Hqrx9wwBORVzf>I>pDOK-iZ6PXzNYJe5HD71?ooC z8p!{cxQzEJEZIV|3A#r8INrE84`*+z9Q;W2TQ6`&amN?yWz$4{36Lfhu<+GyY5dMj zpnsxD9tjbo*R*v=4f(Hu{rcPlv-b)a2~-~K|6Xqj_gxez^rLW)+sGdLUU1Nvh2(sbH1m z{G8)Mz|AWtKLX{n?K2nsUL+br+xnkKUPk=<{cN&X7zn(XBa|!@i$BRTnn`Lit|xhr z+%h| zWom~@7YCvP#>5zkN=hk;*-Hbu0Ud^-Ib~-xLpR3#7r!_VxirW0Et=)HGW05D{aqF9 z3UPrQFn**6$K52NkQ3H@A&?3ANS6Qe!NUA`8L8YCHQg-x-k_AEIBI$=Woaz*M76I^ z?=t$+yPO4^i#DNSC~KJb1ZIO^gF~^1A5_qIDr86oEKRr5J&7?&dm#grB&sqK{e?&) zC0Zp!)W7j4P!w(pCoe8tIe3@nekXB?D%$Y<8}je3z+>*7Lu;<;u{_%UGr}ZTHD-oq zsUbVe&pc*`#31CG;izF2nO`aTj<#G{;15)OGLtsmFZfop5#N@Wf{o9qfkT%$ECbHQ zTWndsAB}ooae9jMLS-%ll_>4suqaE#dROSAH$B;KOoP)LM+`=4!4k#-o^_X3Z-QY? z`EGZkEa=g8B%JfPHvbe4$=|Ph`|}cD3F?LKLhe+il5}dpJ~f@DxR}=d^lvBCa^gQP z%W;xGm z^p%(0dOf@}Ipn-?ON(i+O4E${$aM)Q?WOKwN#`BsT!5*~hXIH^LFtb=R$8vT6c>}p zCke6TKt60a3i`lMo#+En9rbDBtV%Q-O%)=Cv55V5vGFgH@S#+0(mL^Ogc%q2l-2A} z6>ZN?C7UEE7lRJ*gfwGW(a8_h%R7?QBwtMR)^2+uCqpJhkGR;~=;^nZ{|P7xXatIJ z?7sq7Bdb4i^F`+-{rXv`v=aQARjaXQPi}u-TU%RAhcohFsDG1XyTG-yX`wk{h8o|54K=>UeR2V_i}6}n zuS?~Uo&XIZAY2jHgUi#bWcF1glFp~h6K{?TyG6ls@* z9sV5jMN1~s)3$Szk~b@V?%hv7oj(|eZM6-t>_FpHFNzEyC1iI3kja=)&d$O3@@S8= z--BPPmExhA6r%X#D5Yk4Y9gjCNN>9RYS6Xs{^S8b5vL@uAT8@1<1%V3DGk?cqY%c^ z!YEP^a0XLjmKC&=#fNBvY+M+=%4L=Y#pRfpf=}-irfi={4kwIv#UXn};TIar{sC?E z;jw}?wx~OeP~7#ET;fK4<>}v@*OO`{NnZ*q8+qn_c`wmFx)#lxs=djiSuRW@^S8p} zBpEOZpp0%D+U=$S6}o`f{>mkdKcC!?wT&b=BqUER85awS-FbgzU|?W%_2WlXRn^I# zAg}}^t64A9{}<1@-AGxLC$dHj55;K8<8MPyLjcxhi3HZCa6Usz+N+Gv!3)BNOFt0t z`xf+5?V-3Mmq>s=mP1mIaYGTv&jZXo#>9wKSonWXAGEB0wM>O$*iE2UcsHr^0`0c> zGx3aoo+}ZodXFg#sLPm$%jSvoe6qUPJ`F`}ci+xd6|5?XlA7EY>WfcK002j#6#*i( z%{L8i*X9+8+B*|}io?x-6oBo%#^Qh|9>{e?u5R}HBeII?Zdmgy zxZE-gC-7%Qa3ezt7!X%a1XPEiRDh+xhJ%zopb0uXA4&+Ft+l?LjE?`20hlyJY=eC z1RuppQNi53D6%COpG4hx!gF{N;doH|4XOlW%s(3c9M~>M_?6kCF_$gR$y*#i$y!Ha zyX^$M`&pWGGH`sw2$j>u9AIYuVdjjLQR4_%{gXl-wYQTk*QhgxYJ*hx499zCo#oE> z#PUGpk?K|{7+u@$L4O=nO7I06!BZ5s?r6rnXK(%*UaCi<`@9GT!2$gZtz5G(H76jC zg=Ch;Y@c!O6uxCEn9-@p`{`K}kV4L6A6bNkY2x3bZsSmbN}iUY&uLFp`seNw`J`>Y zP7fDT=jazSC;<&e>l^#f8~kT~^VS@Bbz7W9egIRr;N^ltS<8i%tn7dsUoBe9XUL}j zo+7i|7zM#+4i^UcvO5Sm0MWA7weR)20{(g$u(o|G(XI=APZ9Mz{)$Zcf52WN%LwLO z*I2DbQGL?hvD=r@OercO^A-&ONXSmkc`%=TfCm}!(`SR8&A6~Ttqktcn>e!xgY2l__9sX_ za7a}es~^5tGDHyIO9-&#Rk4Ajbn_h85G#Q%-eCg6FbD+f)aTvFjKySc)cRbhSZVb4 zgB;z5R=LEJk`L8ERpsAFcIr!=9K64N3w4?sV^X{M#z=bL^|<{e3qf6GjZTmdsu#T7 zY5)7-G|d|u`fLK5Negg=?os~q!^Gb&th5LWkJ9oFvBc9Gb@p5hshq?%52DrP6}@PD4Gn;Em13~tK+Wd`)k__qF4ga$sW$e71CTc z6O)~dFI1|L8qVPiHA*QMq34u5ip!2x%MsGH!kz@mM;}?ye*#JUiO*cbk%qD{#8u7^SY%ns2| zC47j(Ny4@wrI;X~ObW3PDX_8KJZN2j`kr>6akmrx8bJU2$~7UINk3#VR99Szj0HVc zHR0A6+*kpwce`^Z;ICBI6yI!dD_IE)H;o3kZLqf4Ko5$UvvemUUK3-l>@u$YdDs7fqH#avuZ!g5&zeib$%GgVLXQ1N*f8)0Kg^TW;*wF`%m2p0N|bHo#LMfwlA`b)~->`qmf(@xPyZh0%z zn9~70*n9*5P{bhB6aVyOyxhMJ*qc8&>E+;gwg5$MQHmiY1iSAJy(;+7;?HU6dYZBN zy3AvdRO>`{H8?KfPq-0M(m`brVwm)JK+-_Z(y0pbwcV#h+%!$ezzt2O5`k~n=A+)x zx176}Z?XAW>S8R=93^5?2zF{VWD}DZIMR5~p(Q~9Y-Y_-s-)jz0HH1)ef|{UR^;`) zY`5eyPc)~m;97FobY|aBX{Y{(7y}qX%e$c=z@|eG>zmYx01FpUoThj#>hZiyu#T6i z)4=?1_ow3>^!2R>g4zGcGETjl-pxS$!rn=~AqlTZR+0-$`|S&O}*|Yy_^ILPA2a`|opmdE~MlmZb`E)KR`oQ&rao=Vq(7Qc+9u^@EcoG9(b%60jTyS#gAy0g80 zxxv^b%{muNM>jgF?Fyg3AF$${8{W5O!3o#=JRWd=A45o?by!?(p40;0 z)wcxD7u6#~pI?11L=IF%uoMi4CGzQ?;C#wKGb z+y^QT^TXIqN^=AY!;jXhlCw@5uvlr;jVY@_y$ig2ex_rMopA4NF%+g*6WkH9@+f-7 zL9wVA84oqVK=Z>%LQD_JO#8DG2>;N_|K z&~iA%%ZTk8>;0PMUp@OuJ(jA2qRLdwJD2Kq+cwRC4m1hsoJt!1AH<=gI29P@pFgr@ z5xfkl_PW0T;{Q?GqZdHEhlhv!qzXlo;$mV4GZk5tm6aYoz!T7P`2D>GYoy;sa4*K) z+Hn{$c7m9qG!~7s9ziYl6NVD}4?}t2^S-n0mP%2F(kWr(H;Ph9{^D3ffI4kF5ollV ze0p6O{TMh@#wdsU`8Ma!J~y(qpO#a>K@?0kXqZ6-d+7N|vn_H(?4S&NKmMSTc7Mh- z`jb(5F7+SH-P}nU-C|2Cb#x!O#6>Eg6ye@^8f*NrqX|`29n!TuXTTXV#Mw=@9!OfQ zm6`pS1Pg_3_vZhp63WGUVpM`scESH6InKVK{0s111h+ZLOEbav9J-mWEsbMO1yglf zH#xJ@i-5ELEZNZ3GZvfbu=c!1L#qy%5xul9g)n4%@J)v_64MXcu{Q+1DapAdQd@k4 z!Guy!TF*uoy@ou$4CJ$@(2`*%E1vh^ax5mxzth8QO%xI2!n1u{8La(N6 z!}=O8p1+)5wtP(WF#hSG@uQtjIm-Q<7|0~>*UAlF;UFM=$dUksq9^N5ZielToPybC z*D*!ci*>YEqqO#iN{?RsyRP6Q3*( zFoY4s*u1FXoIZ0U%ThN@&OP332^FcAwWh5Rq$D6tFp$I#W1xRhx<+^{MHEYYmU#B# zpvsaPES8<|*k^@g%?w5jE=;hgmw|?`MTYsZo4B|`%lf z*yNhAtxiK+HT>|@05n`3e8Gly`_(76nr;?fOKsmgIZy^%(vPaBk(v0zqJJK zBd|3`Cs>9*G+iL%j-%?VTt$QB-dja8Uvp%|`Y9KmF&qQmwS*Upc-XX@Pk~M>DJkGV zTtiRqVN`+)+~m2m^)*anCJOD~oqmfYVykXbBJ)ieSXX!+Q}70;%w~kBQP9N!9Y0Tn z=JXSD>vzFBk>yGoYY3v`b#?lFZR)52`DBiB5(_82C7N_pL{qO(b>(x_pcehtd|cvk z8u#Y%tIeFK2_sXd_n3eXn*_BOnq|)t4g~g)9|Eo~0fB5PKM#=eg@lA&gMgc09Djd* zmxZ>|5FqcRs(u3MZ$ymB}=vN#QWhqy5{o8RyrUbWk~$FFJ%07+m0RxG4YwVcoG5 zm-RmLu}e((AR>Pgt^TQ@=Cv-14t`6*t`MDS_UXrSaO~2@%Wjw7X)(akGXO#*pPLQ& zoO>!8WVC@|;joasJDeqn5R}9agX~Gkbg%E3Pi$}JK*cu}1qjrexJPqUV<#Xh2eIrq zPCqP=s7Y+V{YrH>_&V__d|J54xr*PNO@WYNI5kwy8QjAehBcZ|JX-P8AzlUw>?;;b zY;SMBmWe}Ys0Budr=-tHlm5fgd~qW29Hd@|08JIYK!^f3MkrHfGnI#alQ{64X>o)c zjwyA>-$aqnTJhR!u42Q;bYQA!$AY@sW6^#dA{abdNis8`M zoJ4A2ORErl)9>pj1XK-*5W;{cO5fWu2W1H2-Zv0`2^_~@ZfG9_9y$Y4VGtHgsdk+_ zb9~$FVbgS}ZqwFQWf39ROYQwcnMOeyPJ<$lS$UvaRpZe$^hAEaKapRU;g%<0Du`D; zurF#RCpAh1$yXluX}2SxAt7jdL*+}S9+p^eI*_RK&}2{bfpGkvFPUz{>vAW z-pY}3$M<{jf(scu?o@es$2tiSt>pgPMXT>4a>|@Pz8o{m8D(a3GPqp9;AMm&kv05p za`&pXggUYw8~z!sY67KuOXLeja5V6yb|2R4cywGhmi>OnA5TPnaKC8S3$|(FQI^$d;GL)V*D*5Ezd*NPGZ`>FU}4NIcJe4+IBh&9!=(K8eDL@TR8QPy+T`g*3tS zIqsz|l;CY%ccdRdpDTB|R`zG^6HKh=#mDoz0cMcn zDed6kAS~Qo@FA~MzfI75EE`yzx=G!OtE*1FXQ2_=s9nC}r{O8@xnk*!57tXs(K$}^#p}A)qo8!*9qQ`~ zX8249jxX1`(*mRiyb)C6fcyKvK9W_f0^)32IeE=^Dy znd_qudkXM#Hx~?+HBhr0n?U&OUA%Rx69pf@%iQ1BcavyU4GdsKx~(49oQiJb2IFvsm|x+CBrPnbG~NlZKFxJgPc`UR5`Ave7@9cPWf+Oobu6+lJS< z$eR!qIBIscM)sKS^XJ{HX__4HFy0=uE!W$vZ4M@HfolpDU6xUoT}Qc@IWN1}1Wu-e z&-}IXbt<`wh)jM`+II0g3A_IlM;supY|}-`i6Xt`X9@jh7vRnv@f7&$nC)SI1nWegmgLNT>x5AM*oiMUwxI1 zXh`)ok@A;&UPd&!ojhJfOUL3FW!3yPa=a@Fe?0R^7g|)U&!rX-uL^j6|B-h>DiKP2 z9Z#Ts_4kHlo@%)|5S&n6OofM7d}R7P`ZI*~E0yXbv1RD;D#Y|Cn}QqrI^lZ!z5A4h zGETe>!R@P~Z(zfmyi!5tDt=hkY?W!-=@6gkaO(Y*(BpNw1jk1p$8GE4*7juBg|M7s zsB6bk8`&25%u}z=@pHvtQ|a}87lIrlOKRm9#o(EM3AnNaKZ^8>4=-tcNE%n5}ZmbvAy2 zfvP1U&)S0|18b;LSnzI22Uh@h#JO`Pa*vu@+O`3clUzkfLV|Z8<4!u;R+-uuwIwp|1q8R>+_#fVl6QUU0=sdryl8m4PxzMS%^j}-0vQ= z`vgDWtFa1sx9Lli&Pw?L;FX~FY<8>fKlXxmPoZg1(0bu5X@rZ-u!+a+rjJ7|F7%6# zDZ?dEsSa&7YC`em{#-k0Vbm`)M{a*?Vg49!t)u^wnu&iC#cp@gS%>fMEDi4~(IvhA z9MzGVW44fu5~QuXg#-Ci&`*4AhpwWTCK|36-Ubo(Ce9BcSFm;c}0G7;p{9IZ4^C6=+D z-USY6r8hmf8RcT(H7z@0&918_Whr?aUL`oF3P0@K>s*;*3b{5J-5;Y0#v5A&$KnnT zAfJJ7`>reFcb8+RYPLu&WT+e*9L+boMby#e0|{?LK=>{;0wSWjn_D;m`=&6uRW+c} zp?A#2oM|bLQ>mI50;+Y|aK|cV&0nhGWx?YV zz9^bGTc5u0XEpM{+(n2X<*byxQ&UrAZ4Wm+QIsG${oc(7WU_eO?BscRdUn`#Ia?GT z7u0i+*DYau99K4GEJ$3!Dz5L=A^Ddw2JLtqpyE zoeRyDMwLi6d-;aSy5}uV!Rd>Jfd1`*a`8@at+VBIuduEldT`sV`Dk5)?;m;8*zu#j zVzZZFEJgOIV=*2ICzeW_{KBw49LHn8-k^t^gX?cUXT*+IUU19eI=h<6v^rRnbz)^V zJf?qr^~FYnX zLuB4}$H1hTD7$IQU$iMQZe3W*EL@$Ol&1qGylDoltQT>uv#Y62v$yNPa?25NRDiwA+-9S%mroBUdXgL7a=NpX6_l*BnYP_;KEHAK zOSJr$qR)`}u0~vnRNRQfqek#OMqRtWrXXay^#jIW=}HZ)C7cD7DI-WXgw5uU>qd<- z*97AKm{ze^lB_X5`w^fbV+Z)_LF#=jC~)mhof7MRvRU9M;$F6esz<)uFP_TGA04-3m!XXnS;%QYwAI{}T>ibBA`4 zteMNWZfo-Iv{YCgJg@1_xfncQ=B|{ZDS0;A;ys47)1bw+sa96+-*YGrzo#(Hj)Z!2 zm6piQtgAS0{<2TCtM7^W37ZE$!l|#}l1bY?(KQX?%m%SznVXMvGr%6RmKd>I-Jfiy z=gJh5E%8&`+577YUj4E!b;rMQ9pI&aUO>U*rXoY4EEZd~6o(rd_OX0_?ovlCK595z zEaN>YpBa@e`pb!tFccEfB~V1s&~I}rp2l|2KH{4ET(c$__-@!2LkIj1L-#y*E&Z~x z+F5|V|1F7@-N8oS^~0}oy8wdR>Kgvi7;4Ns#d|Fv)fclj)_YvrOG^h!Qi)Xc_CDqB zm~jY>2pb+E9$->Un0=U*ctapUUGkI1d}F}PNRqa9lvG$`9(YnZRfblM&S>d#VyGXm z?!R|2dAVFgmj!MC^PgRZcI8F(-)rO>ItMAW?k!Z^fP^>0cTBA?kYK;5z44emr$#HhP!fD5MLvMzW;7)2%l;TtmKQg70RzPPNJF5_DYTy|N&IiE(=m=+e1e zcFTW-$Be-`zZ&COPkPUibVJ(A_{(>|pCWZ5sH?Js0T^7MjB}A`VvI zgUD%Q3tC0Ig;Dr%fyR2g>_()0k*L~f_^!SA+9$_Y!3$hy!0UJgujEHr*M48f?G=0( z@<9En>0b}SNmTJr@2xj=tYiY(>t7P@5i&YXE_OiUBTHhOI}S!K5>)MAUD?a}+4*yL z%{}OACq#eF6g=3?1Vqwrssqq*W6A5mTd`|acoh1L2q=s)ty?X)(yBxbBYV+48k1c= z&nWqd72W~L@xhQ`Z>8wm&~NsG@C-g2WaLr#a7_Zd89F!`&on*Um?@bZ7u!w!iM-o3 zhBNliDz{2jNA~tfe*zW`yN00d1znl7ihMmYybr0i)N%aKk8f1d0V$jWccK;JNF-A$ z6ljK7)@*)2s=i$HjIao}RR)7Bhm#oy9Z^zIXmvDHX!m`bh>f>FZgzEOpPrRlFE+M4 z-XBHqpY*8v;#1p@Jk{=Ccqp1YyL*hr$eaK!Q+!?5pEg4b7K#;>-d7oK8gNE0O>dQ5 z!`0sgl5_nPUa&P1floU+@ z@6^88y4TveUpPz!gJlCpzBN6Z03%p5^56L!1e&RT80#|Tjh`cBPv<<>h`<}wK^798GEmnmF?5^)%r@oXn0?21${O247*9O%}PuR{7J)A_;}a zoF*`tGwZr!@i}4Mw4Oz^Hs9syNBMJnIO}ClqW>gG*Y+_3a`^!Q-&_qV^fxj8_Xx?oh_mvq86Z?5p}-`S5g;?O2Kw@*}< zkDjU@<9Szt3k6{7E5B4nE}6VRt*54O(#x>#VMszZe+%mtpCj*&OeQxlD%XpyZsZ{929BoM9MeN}l? z?L>nLmI{m$I9%w<-LJp`b>WG9t6_o#=nYJirr7Prng$U@_Tn#4uP339_i)aRei9&k6z;>hEYl+A$rw$;Dqbe~3gwjvaz|byAoWF|9v> z^dEx^T&3%TXjOTzp2NZ$q{5!Fq?=?fipuMrOz>E?nx-5S*e}#eb*_Q6080bF>p^(8jvdQBUj38eOs7d40kpgA@^1dxEq zOWwxwz6HFesWiX*khhhW%|N;jA>IpP*V`gloW1$1SSFA&*L*vAWQlgsgBsV}VX2mu zwi&${Sy7#)epfmr1qK|&q9xaGBY1ry(fUgyUm7f1p~jFsE()myl7%IVo_h{Z1c=6p zH%VqEPZNpE`RS-s!uV)UnpIw%{ixjrC_g^e+S`2hqRuqFO`z$H+-j2i7g#|5y+K=` zMB!cWa?3dk<~JfW{Nm1^ZavI4G=Jb!dRE8BYVc616HV^Nv>R5vt#ok^xeumKn9n%l z5>sW@jIUE%A1=e{WEULYIuCf_2$55Jv#ArHqu6 zX7iV0V3d4Z4PkjWu<<@UJ1Z_IAP9HxzuxwqFeLDeLMs{qrx%MFddsbZ=tk?m2N87X z%y%UMY;jx_GVH`~VA7V76f|ABXBQXDTGeLPCmT?@BCXeD)AG}Fdk>S#!QAA@)jd&jQ92sK1y^vT}(T)wy zmHZu1^>p~;0NqVCh5d)!2q8{?JWh)DJ^bnNBZk2;>Nb~Z_WbM(W96G#cUuqwTVw1w z7&4gFV{0nG{Q-71`pB4J0>^bF(xN;=Fx5XwPJ47Ffo1q!3#GF z?qFDXgFmL1=cN94U)$kqeINw~oy`pK5DqKNMsrvngF!f2^bpGE%zzYND2$uHUVQ}| z(b`D`7N!44n_UO~=@)X9%sUO1|9^bFbx;;w)HV9h4T6ZI(%mH?jdXW+3DPOuEg>K+ z-3`(u-6h>fcZc*n{NDTCnfu50F^=Pm%rkIepS{;wd+mryVaZ46MIdELkhW~Z4AqTe zA%iKHl+04l($b=(r3FK8iO|KhwNt@;vrChne0m+=sZ#nG1Nj5O^#_S;tdEuvlL~jQ zM<$p&g}jP-8QKuR!2pd9(234sZ-MRrIN46-%R)|HoXEho5`3^BY|#WM{0|0g;8FI= zac2}*8F*cceCRF))Wc+f{4@B@6+nT2Z-DjJ=9lrN2HYS40r4jeB3K7pB1!q0k?@P| zM}iu0Q`W%CGkUo6vG`w~E^p{4 z-2`IMvVya%3kRVhrKq@wDz6wI)K`1sqm_}ux5sVO1D>v_LrJ7rVI>q2_4iAZYd{7I zQk*q=^Q35RpkUDPVone-tKozCy;zXCVtPHiky3G_!-xhta{)DU8-|@?ECUDq*DQ}#~8@y-M{Prg~R?#mN?bx zQLgJjwpr+94#JLanitQURnLYeNFL6TLf>_pZ&Q(@25|ov(?R*x;yt8sa@vPm5xP3X zSe9VrrC!+lGHl=U=(N)>p5=%)be7GprhZbg@ro-UJdb;p&U!+t*@)Z;BXrxS7S0To8g1DuR4?XOpbeYl?_lq@C7@Bd&2P;huC z4ex((dPtg{k5e0sEK!VPDiKEGEjXm1&j@H2-v2NGouDzTQJ|sP-PiL&Kz9S{QljPX{^8`M;TRVN+4ClLM+(kiN+cd0 zFj#z<#5wStg4gBHjwc-}%mh#u@U!RE2E6c+1s4P3o8#dlhkr9>_-f!IwjCk5G%Lh+kR-Z{spc~c zC@q{Bi|0YWy%BBHk3gPwDlU^UQ3nF^9+r5pndDtz&h?a2acgb04>nTl{{{! z%ec@P`ZF?0CmfV)aG*{ejFnWJUmmuXKUA9KL>cxw7W+{!fP0=95)1}@fOF2u#zun; zn7510Ae17Nec5a1n-~{M`jo;CftmPCDRh-{B4ra;?uqlx(hg{SAzl-y(`?;}FR+HX zKiHQD8s+>*f)9)g4AuY&ALF}Oda;1Am>{K$<93r?$M`w9=k&Po0Z6RVd0oM%)tk90 zEXsrACC?pCPQhT~aTpH*{^K2vA2Q_AnFrBWEVb9hMJx(gfROKJ|{2phGZo`LDbs z7xc6A_8_tgc%B<2cLq2Jf7O+L0tOR2d|dTKzneVjY?i};TftwQ_WmRNvLuAjy!esHdsXw4VvAw{BwM>PJ#o>b_MRpLnr|ay}y{)wL2;07)xlla+j)v$Z*XJLA;uP z5yTwj0$tnf|79x;jg8i;ZC-%V03YyD3+`n2v>9~xcSO^-w6>P1(WTUBm4R2Sn-(h! z8xs=>V#4|rc!ykFT?LF~Pg~Y%F`P$xx=8|L01Aya@Ev)PGSR%plh{f$B?mPoAq2?k zo@ej*I`s;emXwy=Rtnj-^auWv)5Q#uBv)Meyq3zal$K(+eLigctP^Lx@Z0OPNR{SU&Q zJrH+XY{~Ya;rZTcM1*JG9a4DCD1ydo%TcO9a4%iYHd%&~Lce*L`P#rZEp0qNCdF@| zWWL}~O)|wy^)q7#TsEOoYvdu397HgF_VnF$n8C&N$ZmxzR=Kmbm_}JrGZcAu16Beg zrt<KbvC5aM~|;6Gb#BRfv9(5>GdZv|uO zG<%?Z;hR7^$(8P|a&l3Vh5rm_c{XF#n5klFYHFGRq6rX2UqH$i>T9O7@h?LJ7&z(g z?>{r6mNVf74n#u(SKNTp;apWzRFp^vdh7Lq)#KwMWn2*`Us+ilq`YJ7J6xcpwW)Qnuhrk2l)ilG8%AnZ5=1eY08v0elgbsq*no%T+o2z(l((XOJP$se)0Yw}nF8_Bz;PPQ( z>B_2j{imn)5eCG}iv{&|o9oC80WwaF15qH0WB-Ij!`97X6V@n_?@`1mDhx4*3#Idx zL*i2kffvgOk}>=)$9r!6o0k|~FpBLDcv#rj>`CZS z0{r$>kI~P>cs#6S9P`HzM6VIE@W5#iWBnNxNlO!@{k|=IJ&uwD`s`Y_jA8^x^`C}{ z43+GYquCAoMgzh`>-u3|rP?E6g|8EY__$v%fgpT4DXjjff>mY>z~>MV;n-JQBItnPc0lPhI)_z?y?-${>+h(J z@?z=K#v#YrE303jWng$bs$T_O_eYzX-GA*i9d4Hz9a~#k+8(bLfz!oLelK@$pRfl_ zB70?EuXvCKImBgoq@y>xzl^x4s2i3XW)kIJg)H37uwDvwWWYcyO?`tdnvM0@({)UV z3tJZ+w{Ff)tj34>vGW5odBFx{+N3*48%?#~w{PIku)#Wy2VN!aZEdmzGvgG@pVs8K zikl9>kP|t3@g!P$EK~D*?Mb_((2eyz(D<9uB!#3ans~#-MueYKUK=yaj|7>;r3!(& z3&eb2z#AsdipTRes$`$yW{$VvFJmj~?KN33=w2dG`#=bL{tcgv=3`U1oRwO&0+-Ec z>FP5{wc4-7THGE3DBq+V)l6n=>V@u)Jm>RSH0)c>=Ht@qub37k=pnz)#5RUdsZ|F{ zMIneMOv=1CZH*7H{BsUNcJpgUBgHk#=<7Q&ot=;kQs54;z%UMvu*S9s! zQ&)S0iH2`wcN%sJ4}j1u9u|*k$g?vPxyZ3eeZ+ESqo1qMUeCxVI({6L&|pIL{Dv_` zo9`ona-gVFBge8QmbbJkLEP5>Oh>9h^8IJSE3FJIXL(J666ym z5f%-R+3a?2B^vZItPr{=9@rbxemLPBa)O0x%pqr^~N)>^X7)c(ft@6ng1A6A`} zxTIMeT<~V1T~XQ$63H!Df33GGM-wf08s$pV=rq4zbx>Y`S(3L{Q&x5ocJG@?2*ojo z{#Og&laz6t_V*Qdm(mA*R+l*R5hmViSGd5{I=@kC@%Ac-N^oMV&80h+x;V?MD<>OG z#Zs2`K9&>8(c2Pog!-*AzdEwNvgBRZO|dR`o89IAi&n}CtaqmqHGAM9qrLhp#ypzPm?c4K5!Q&}sEglTIt_7V#8v+t}T5j0pjGHXg(EKhuGqLiO~))@g0rrLMxQ za@AZ$%71wY-c6&m_u$uXkhyNbi)F#$DP`Vo5_s8ukz;{j4u-B(&ryTx-~Po^FeKu2 zJdv-g!mwh)vzY<+qIJ2uWLiwlC*7hIEga*x4EW&L$vyqQuWtPYVC8 z{~Dfa;ah7T4Jty6AIGu|7Y&PsD*@O8%tSms-6(r{8%NDg3axTj&V5v(N|X~?^yr#; zcKwxg3Rv`5ZH>i*5Et(|ya}y!C5PM&E56z7S0!oPnAOAj+VQ5chLWbL!MBP|4#?3@ z3~Ztv=eczX*)-oFzL6aP?wZKa-hFgE!+}b9|ArB`Op)P&Xh^d5%LtACd_kE<_^s1N zspV5QN^i_O84Egj0Z#jl$7z?z=N9FgmP)-FKFSy5f@wr0kaOLRR&6MF+NGNE3 z{m6xxs07TGtRAt?O@FV3v+6S_;BwZ#&t@is9+HBr?Hpr6?9?qhXRD-|NNs0|148rT zoMWNQp9iTWm%+1Z1=p_QymL=f59>|~&4(?3JFi-Hp>rDrZf^)Ye5}&EU_sVrCQRM@ z#MtKaK3;YQvWE%>uCm1Xi=}G0E)z-n2TEYU<)@o5>VaO9r0wAVVxpv%w?k zFLH2^v$MaoDTu`~E!}0%E9dO&EF~qSp@Hk35+A?0wg!SW<+_$EPBq6q6lAIx!XA<^ zggBWpH}mboN3}&F3ojjokdR97_U`@$Z?7R|YQ6R1b)UeKJ=sg-KfebH1?o0%?A zFL2j&B`TEf<0-4O{|i*(x|53YLRL}ZQ;N9g%cGG8vX-Ug1K@#`=kn8ui{`R+*y5$U zl(=U~=ebCy$P(WlbSC)C&V&u3ty`vxiCP?#M%yOwe9!{< ziT4y`g6KDFypq9EhInE>Gd&5`rS3NI;T-iIKFRCk_p${^R}=w`8Xw&k1FT{{c=~0d z6{JR!v1O{1X@1Thx2hwX-Fy^x$TyG$!*@IW+R@U{eG?;d*V5IsX2yMZe56{}`gZgY z!W?_sRR;nJJ>fUKRV0WXIqA5F;y#w&ln-(jNtQaT=_%xQ-YF@5XEjfvF<4Z@mxLxZ zF)=YUwHXXwySuvs65PlUGbZgidg8U-8#FWK!aN;i@LCY{Afn8_L5`LY>kNUs9c zpgzR6&uRxP}=Oy!M5 zB~eGWUu)K#VgquvS^KzpNV6A{H5r}%-eN#}^;SO@AL^$&aZln_H$Ey9Ne9wp=uGpB zjZTh0sUNTlQG8xU)&wzxwK_)4HE;aYw$MoqGvTpL;zJaTN(pr6j1fz;cX9bhF>A!G zL$a}sw+l;nM-xfEl&0p)r96~%Ml}Hq?+kM(bdcc>4B0XhL)&zFL$ml|JvnqI;(cf= z@a)qvrEJi&O>=NRU;!It8Djt|i-?F!{`rHA+yk^AxBX<%AAt?B%Dy#!>k?Ao0h%8) z8YU7SYeSuOND;>e&3?$*Ldl+sS;OQ=%eG_q{MqYzP5?nFzbvvnz?I|U;|qSehmY;_ zb$1^wHL?i^thys~t&HP5F&-SwpDzDITYY1aQQO8-YS?F~$#c|AWCcAsQrScaro?RlrQtHq&OWb^T0nDWQA2DZA}M+`dUHGK(jY#Bw=f{zwGsP2Lb^euXK= z;aFKx5DDJ9^W(a;!9sp|k+RZTFquCkfT#5GuH2-F@Q`bn(5_aFPT@Cq#3cRpL>b$6QtpVTCBom?)XDHWW*)N|GbEm>f0ZW!l-^~04Cd+8;lG&jF(k4O*lReVgDLu3WWzodM22pD0bq;U2nQ)Lg!Uz^g zi@J`?6X*j|7XusUmt;h<#!OXRo#0z9a59chOyuR|RW6zWa=0SBSrg?|t!*`{s%a-B zHk@2DdSKpWYOc{GS9y|Yy2p;?y6<~=mlD->c~KT|D@U&SuL?=}c|x`}L?VoFmn?;ts0t9$PjfgE znnaUoafJ0-JU;Ig!>So*N`h=ERdmj=iF?g5T`kZGd&65U0Yq36=b=>*M~%W+u|ZU! z3uRiS4`;Hd(Xp-Q;{B{v^n=xukkIq&f5Q%mEN&^JVfhlIronDnV8)|1Wi$$i3V5Vh zdSYgAi0T&xi9*Wa5;;L>b7ToO;g=4q6Ivf@K@)xNa8JbB_S@~$PZ2n+r=2Lt$;pa} ziavn3Xl`x>;i6?|=Sv1Bo#w7z{J^A3P1z!hr067M9u%SK7QrrvP>}YN)kV+$wA*-Z zUT}ir)|(YDKnFM~ynG_~;1abmR{U@DI934AhqFK)im|m2y4Ea@>A|e#68`Z|3#|1V zw~>8!{%5LVV$!>zH8B|3Jsrz17UV|C2tp4?<`hM?Am#-=seM7JO`NQU|w|@P@6c&ibcnEjXqU)+o-|Y1(qr! z2XHJu1oz=kFBHZCJ*x#nRDN8^;;_-X)}xM)TWwYoRJ~AR?4Q5!nanhbwAe<8@Nv~2 z((rfLjj7Y<2;bp>DawydE-p`}T^Qm~gkVK%y>ZvqIsyRaPK1k_6bev3iPaQjh~S63 zOIC;K0{R$HRM^j7{|A`)@~;3Nd9ov51pz{IguNQ^CfpbHVC^FiqeX&|o^x;6@sD;V zaw=NR2La4_QQ`9_O^)RStQ5f5DdpC38Wp7j!$tsTDP@uz33=sn!^dR015iRg(%HCFiP z$lV+l65hVTK^-wfgZ=Ek(1eU;NH2GBD~O3@6N3`=3n1%yyH%uINhE!*@owDP8!t^uv|h^z06`w0`N!Tj|!9NGMm) z=cF4*rDsNF@%yTcWRM2Tgk8EV8b%n1s1u(1^pX~9LIg+@0@X?$|7}G309`D&%+VqL z0B@^SrnR-b&55aQ;z50RzS>m@@`cNPcY-2jC%1=Vy(_;f(%L3 zLIKDHhHF;w0^GC>d<0OR`NDJ!6fp>J zPqjH3V90YRebgMRYgKiCQDyFMOh55FgNulW5=h{Jg6%lN1l>mveqyB1B0)~ivfw_% zl=6by03vt;iL`?raz|Pz<(ig)s#SMXJYsN+JV-fNANP%nyMOJT`+od=%js*Xx+c2) z#8C{s_k>M99`jSl>}ppY3&|~UMs{Dz^%Dirm*anCv7F|DO)EIvU9PG4u}Oth24Io5 zY&7LZ9ktdNR((ot6t>Cv%NPjKjKqM0 z59RwX{59nEGr(<8qp~y083fp_XZp{r#;xP^ac^^j^sbRkJ5`1NZIqM?C9o?Vzg zFYqytLvA=%i6aQQT1M^bgamDH?-&v!GRy85^<8JDj5^3L>TOq7rl+SZ1A16Vq$n0e zpk^rzlR04eU!nQsmyA1EhxoC+%C)yJ;G(P}zm_QSO0l8JSDJ`ua(?MJR58%(YvXd< zMh)or_i!-=fR7B%^MRM~TGiD`CDU?OO_T(e(=+ci3{i6` zn4hfpDmU(e(+Q%CNN~iz3^uz>s-M8`|8uly#t6*^4z!Hl?VyINoXUMxn~<|lhWw_) zWX=_tFRLFdELN?UjS$T?Bw)Z90tvw{lY#G3Q&T{d;j^CjWfs8b7KlU3lZV}eAvbC) zV52q1O0v~T|YiWO%0!~NirpUdVt0~t}2hj^` z*1vQnESMacY0#UK>AuwFbzfZ@v|0RfJm_W~xwD(C}eKv)|3Ji=B<#cj#io6Z1rv~R?^4&XJ9e9xy9_)OWIg{`E z+sFW)6IAlaEa((O{>7&-xCSM-u0Z>PGRjE zytI9lxa=W8k#UdmYEEZK1r5VC=G%EWqkDQB>bP>!X>{^%Y*M+^>E67T4%Ig?O(q$8 z+sx~h9Qfa!?`GT8D)g+GQ8T*6Mn)b_e4d*@_wozF`~36>?5^UfY&?W16deQUg()wP z$V{T_H@u%nW=+&;{Tj^fM@GJf5b3RBRbBV^+O8rIYeh^}+mt*ZulM_|dJh8y=L~Rq01KL8u_Ap)M9!Zk^2R%@8AM!~6dd zs}|hYi65tPKALvoWn0=a1L<7P;XmlZmr3VhFfTde=H}+$;2`(`irU*(O;`b6mOf(U z$dv|i$Xuh=t-|<1v6<21@R z2c6(d-1@HYeC3?4j*-8G^Jn9DnWP#0ii;pYJE7FZ6*;O9v+ZS{s94BKmG(57p&>!624Oj*|;CXcj=$`f5jiz z2lX~%V}vnWeLEfh2wFKrVIM1Yj;sd{46HEQ1o2PLZ9CbV zl#DEqUNd{b;$=t$=vDrDu?85oHVyRF;v`9S9;eiR4DKNchE3AVd3vFw zWCrXS?mzm1w)kBM+0T5HWE{rcek0U4cT@MZ6>Guka+MW!gdS6hA*1V|vIK9v&4^w# zzc$^OnQNj1WhNe(C=*r0?m5+z&wYme+h8j|nIHfZ!ao9UY0ZXR;-ntmiB{5-+R^TM%U*jXlTay>e;__o}HZy#x)OU zwDM(rXSWZcKcG5cJNnzN0R`!HCA33_XntN&))50`SgoHR8m}rmGypAgc)zam;h3~W z{!F=|u%wXwlo3OW+iawgS3G+Fn7RpLA0MAqJzFnH;#Yw|LCZeRk3b#@CPD)*!MGbX zoe(;fxw3%H%7IP}0W( z^#4!>Kgt^L9{-w1`LX3u)!7#k{NhqITv*vxA#*(dob&I%^E8&iQPSGF{DpI|ja`T) z+Lhs|%oLItU-($~O;w2CWBm@{)7W3@_lY^`4B`%L*Kgc#OlNll(*DJ4jW|<)Y}A*MQA>`hglOVexHJi?E7DzR4;F?37wz+@_?fwa&lfqh14jLpv+HKkj1Q z@9MQI^S#&BPx*JqizXQ1z0485 zn0B!m0XpCTtp*f<04`SI>Bi@Iwx?$c^x-~&mdYkr1(a>7OXrkmrH?_O|x@=J*h zBtrQ!j)T?;{NUCD4oz~LUE}7f|H1NUtx-cb%_u0VM*!v5@nGs|> z2t1cogW3atSPGeUf8~KWdnse-A$GO2knady@eG6vL4RY$$-Dw??SvhPAgACzlM>N) z`1$jg*II;!O&mV7R2lW6f+!1z69ESi02YbBk}2QB=GjJsLUAm@f)$$^u(h?*wVE#U z(7v;d3Au^hk1l;Ki45FC2MS%ILvZuv)(J$Fn69*|17ci_NaKU=9J=4RAX8uoQ*I z?e3L$lv;$)deTKFU8j@^!C6`_jP=!t8&N`$&60epn7C74=28I|33NTO5lJCOjoajl zDJix32fiE4ixWFZSB%sTwqno-e|Yl)6%&d>LPDm+H47%^z4IP`uO5gUPjkl6P=ey( zeFJe+6^YK2O#mD)oX;RKw6IXee03tu4kh?PL-Vlf{e|iWi*?*=zWsn?NcpX(OwCX5 z^mZo$W07l{H{~VG$=qUv=JO8Vw=1bl+SYT-v|`ZU!q>&f(3&5IiG{3fh**8Wz$3mO zU`gXsII@I&AqLlskpi!!%6&<3(bF14G8bmT|4N{)o-d-GR>|b?t_}_Y8Q^|Gr z^k_;?19yM$6z1gQw6{M$++LuZ5 zf-%N_spJ^o#W4Y_W9^}ac&>XuIDeV^^>rsbDJdy8Ht2Qbrxa zQ3@}~Zg0P#x7O_2?*u%DsAOhV!TA!|4Dr8-<_?HmO-wQ!56m(Yzp^z?`y;fD;Y#64 z%22Yl;jLMusM2HVu6*w-dR+NhJ6RGWAHX6f@{TO)! zlw`I3`njrR!4hs?n`?@`@BRaW#1=mObNOk5?y6A6z`(%N{!pz|#_znZ0yMgwo}Mq> zZeI?c9`C6P=dXkHR`-Zzn8;8*JF7$1p8V`9r|^QgjJ^NECixD2#82Sfgnre7bGH67 zq&gz?;xyme%d;_vD6LA56VbCiHKh_O(QQ_vUZ!QrFfH)7Cr^)qLgWDmI)HG|+zR#A zY8(S6{{E!P->3p-@wrbIKJDX=+f8XRUNr;+g*|M>_0ymHH#7* zX_Snd`a{O8e~{Xiq2x5kpOdSecSS>x zwax*f)-=y)-QPPiZ+56$f|!pumkGr*Uo>MJ9086lgy;w(R-2n^n<+>IgMxt=e2naY zWvGNZ7dbS=K~gxsO|xq?C_Y=B;@8&EqMs>+|0tyBwYl{$dCVLbE&N3?%z$fH^j#vi ze#PBtvCbM+sG+$zK(+{4P);stz=RahXxRJx23)Erdnz5|9Teovvb76LKqf5#uUnI8 zuC#bM(Y6ovVY9iVdME(SO6u(;B>LUbFYbivddg<7Om=D618)N{i?hDD<8gC;Qrm3@ z$i-{OE-9Q!89c<(Y(mO|YUIlgt;|G_4DlHyn_8eEu(F9upH& zmIBa20YFON>F%JSwY3!mhu+S~DGgA2Xsf61V0;M{Cak45-eE2EaVsg}(pmnON(w!V z5CAuob_$qpM&At4>ExT}=G3_ylFb4W3uh&Dtg=$>0kTj1&k^9yk=@0;xX4p_vaaS_ z5Dm34r5_5;`pPhR$6ncSx)z8qkmYK7yP!|K11sqXPgHcpP8Wa+2iFH=L9J=StKU0T zo5F%ZDgl;{pn~NL?aT}h_c$S^zyu@S|A)7IlEFuwLNPjPf-*?l_S8X{PIFW9B5Y)0o=b%b_o@T4(MSy3! zXw)Q;#%nLoT)sp7f`z%8_0RT%6~g=-P8J6l(r|6rl#aI#{{bSnQO~tO_`i52NI|3d z?{-JidB2Gn0|hE@4sku4xWK*rUoF5K%XCD_mtR5OZOkcOglZNqVzL3<#tzQLk5)8^ z$=C(9M^m^P5roA{2f)ITzTt{mZHsrF`Hs8PJv~>CsxcX*+S#zqN(6e`y22m2Jv+XS zl-(Jr^-t(i^&%r>x-@sMzT4IRH}&(Q{F}zK<7LIqxroP} z%(VZJeeV(f{F_lpV||VlvL?OkSsoo0cYed7&BmDiU<*P_Uu*pyL&_O|i%e9ou5hEm z$ER)sU(3RngaogeD%XCm<+OBbyPA7-(SnQ(pG)5yajn>;=bIWjQO}12HpHb|*c_MX z;63VarjjWo+KMi7@OOwH=Bo@)rhe=Jy_lQIp2pM-Gjj1RVI9K|Xudb4Nx}z{<*)L! z-mJZTYCiZ;mpWp_2nPXa7-T%MC9HGypCXof9`(jjz{>1|9F%W2rp!2C_)wu=f&@{p_E~}Y?{YO(a4ANeZP;{oOQDfLfKcNJEQ=23 zl(Z~02&)0oAy1T#v)-5`T22e=Cr}#?`!M8E{5AUM$KgGkq+}Wzm>9)VBN<48P1=61 z#KQaYFRMk((W=)TcH(DLiapa!t$A_fW;xXD^1Ldq@wfB#Gbh{{W+T3#jQ?SsFVe77 zy3A`!7|F+SQ8);|@MEo5A(>-&MDSDqOnCoZOSs}PzLR-R02V$V`Ns<;XJKiHg&g$r zXTaYuXusc!i`r%HhWPgPBCHskcv*^C@$mu@YvWU!9lD@!9%r|^&)KR%%1b{Bw7xE) z0tm0=90|?sd98FxwVHik=`5WdtS>=0?*ju+{N()6dC7Y1R2JG_~~kTK+vAhNYfFeIK$y=5xRo&F4JYiF2gxNy$%(%V6wttzgOw}ZPdj~m;D zgj=5`J|V;Vv;X253Hxm>^EM)M&gD|c$Wl^o5*bCDjqzI@Y_0Z$@-Sb6P6JEutn*{B96;Tna1SeBBgfrCP4LPN1%qAnjQatb@sz;Yp#AWI64 zc1+AT9dJR{4v6A(IRya%Aho0+Ka2E;;n_$zZI*CdUQ<#t?Zfpyf_9d7rQGW+G=*g)bb;J@ zMC=nz)v=`G3}l~mnEFvt4pSc zc@VRhd>wLO2DX+5|HcY3F=-)gq~>UmwG+L?XnYSwzGDlCy&m6OMj~lcgC#;c-DR@Y zv@e?3LPcOUG9B3=Q`e!F4h~)iWh?Gr^k4{}Vr^&54VDeDX38{=oIZ}%p!2y<&;qfG zCB={nXiNh@nPmn%njuM(a1u4rkd%1yq$oW`K*zAb3$7n3nho+tfrq~sz|_#^-`|iA z)(e+_qT%7?l?R|R#-(ArkTtz?*xZlaGEXY)delzgM;d*GD|d$cX6_LzU8VK3z^T_$ zm5#x~g-jgxA1{@DI=xkL`7O3IU-OsuV4$pe2r(fYFfu)kKC?D4ap*i8SM7E!7-OcB-IR@`lO37;i8Dw0i}2Gw=WLgbqa;8nUL%O&-D9=Noo(Ps~VR0h8{ zQ}49|sm{n7etvfLDywY7@_*c4-L4fqYM13?^c-mhxtMS& zV`U|!>yy>zDP?)tt87YwdaVeA@~oxGryc>*4ADX zz7(!y22?2^$%M%`no*qXGxrXNC(^d3;xP78ov&Irh$O)n+Rv@Ilo&Bq?P=So#0@ML%Pft}lJ`0W`K=#n9fcv)H1hBTrM@0ah za*iCym&)dw)!Dqt49aPJJYyr$DT{;0vgoxlNbr009}cn^X+YNyX*-I&U)vbRmU%^N zT_|Yy`(=5D=~S_U;UM9@ZoyV@z|$M@ury@O`)YYq=1WCDvR=~2w)JA&<5{$iPL7 ztjF*rh>(NDJuWOPtkj?fJ%B3I*nmm729IZf`=>bM!4PO)HLt`v1=4g8l8o&F1@9ss z`GeDne^CW=8@gAJ%xJT0MS7Mo>EWCV5;oe^r*2WqZgX#CKrx2%_Lq0^d<%+Cj}2Bu z!#mJ9$#rB4<~UqB6t43Utot!g4>x~c7gTB|s~dY$d>v68t=X476QmN10Xl zRQFCQmkBeqPBpR#kGIbk3C{!J?}aB1J)uzyccyt-Eixw|zXNP$Qsz+i3qu-hzLin4 znZRA{K%US1+);p;0KG0ayO5X=;wW!Bv_WSOi} z(u_z}tFVaXjcRI9m0P4k$#Q=NLYnuxyAjj>qn-lBxkSX9G~m{vJa z_8yf_im=_O958Ro!+|_$S@|cvYY`kC70UD4*0iGHIl;6cdce7#XMYX38tNBz?OAJ8tJgx2k@0l%-fo$dr$yOcsQ%&5=K{Yu z;kaY0lVk{3z)!A}bL5@-DZ`nZ0MCZ<_03jFMAYA>-)s*!a%P?k8cJj;$j$wDHKh!s z(3F&Bpl|v1_EszhXhrkDeO8=(i{I(h39r2ZA{YdS!l}XU_{3jhk;jM$BZU7wJj;;l zt6Pe$2M4$tFOXk+lPLN18Z$j8G)Z4#Hul@M;Pb!CRsOAC_4WW$81=!JO#E*%F{bWKTd0p01XlSmDUr&88 zT`TZxexlDmSFo>1STz8)WSy;jPvVXGTTck22oL4FX3W3L{=Bq7+h@Nvg+d)nqjQi& z5Syt4H=7bKb#-rWt-410uc ziJV()@ILDqXQeiLxb9)Hw5o7(U1_kLUrE}ve0)-3xqs$=S5YHt6;nToi!R!$sz!Mi zVFNY-P`0iv!5R?U8s@-861Kx{2_LQaQ5z}gpw|HnkEX{S0w26IVJ;Ue1aC&2f8}hKb45#U)>ZTWc`U#T~fv z7XnOpFNxao+j}?&D6%ap@)wf_toUyPY-9oQ|DQd%*b5(s644_}83#P+Upkh*e@FD+ z-QRa`sOD{dM2Df!DS<$O52;4m8Bk*@eaH?3ROa{D1& z@6{tKs-hBrj(b5caMNUX<>r_v+Un1Df=`D!FAWBLrb`tZtzZL)a5{qV7-+*5(s}2u zv3@7N`#b|u*cqqZ^yJIl@Fr(!{)>6??8Be0RwDui(IJ7azq^D>ZhaED=E=E=?lQ;^ z_c&>-FDRg5V{-&eg+OMvU94IXDMl8-fkWd%4oZPKb0WyU=JcB7pNBLs5Nwn32y6up zj{gczi!VD+km^~<-2|y*5(}U9lyI+>tJl9Iz>&Xri9NAtX*voDmB9IjGnEJn>l_sD z_y7LM^byXm1UUUD(J4TbO{c*wR5JS$9*Kd^X!_);H&B68o*@>(cyh&u?j2nzN!yz^ zF;QrU5PWR=9w(QtxTNxpD{0S|s^|pB&gVG~EyBet(@JT+FAd9!Qz-5LcgzuuQ}><- za09r6>&gN!*go62qm*0CRdLL4VU-Lwg8vZwHS$4_vw9a3Ack#0KY|#^=hM26$@>~<_f}mvx zMwWH01ZiBqY)Pm4DvDmFVdpd zPTU`|fsRMh0uiVyA(=0bmgVs0L&PdH)_RcF!G8W`(($H!wblS&i&nkdz!wPEBD^4r zaEU{psyvyL7f6qbGp^G3=lKK-G9_{eH1JLkr+wJ$TS3d4fM3H zHw_6tQB@?dwdc}RZ$V-CNW=R`_g2*8{ArvDYnZt-+MMR$jM0bhsatKAk!p}_bgE4I(Q&1je{?JQ*whq#&YB~&whoQ$Ff7xIeZz!T^hFY~ef%VZjw znyTjyfj1z)g|7}~$A0qN2WNcg`ztI;!9o0CT*N-Y3nYbSDbi5HF;gc}DiB{#5V2ZF z@5T<|emof+wAopG;oSb;Eavp|w88FgXD}MUJNMnJr>7@^l?rMmrf1+q2dMWdAW;LB z1}_S$*^?Iv;O6pDX7GDoMwE+GW=Lll4m!Z-1-P4m5EmpQOM&K429EbxT@UO#+3OM* zIFSDBm@+h8gAGX`_dRv;V}O)+XDsp^WF=~099FLE*P}}^{7kj42~G#dj8*jJ9TI)Q zi$W~{tCq{SqcX7+O(+zg{0{0ig3$q9K^<+KAk%kGCqFT|1~lt}Fp|Tv)#2I%b_(Vl z1=%Xpw>lzMCsA~?RgWO#X4G;!agbM?FZ*#$D_)kCBSP~XSu$mGoG zzbNtISR{0;&sdl{8t%RZ{Tssne=Ueicdd*`TS+4{-*v0TUnv!Ug_h}ZLjqKD{(}dg zK%L6#-j0j-Cy5$V>PqnPiz8xqfC^*(pWd1O-hfuUcAI4$FwPzHTOt6S?s9MP#mf4s zIi1_l48%Vc78cMk4V3Lb)sv!Ul{4{g!K%X4%uH^kxbL;BVRDFh1}U-MQITC49As{- z?*9JGUINPd@szocKhHnM$!(sG&B#nc*7>1dwZJ zZkEeSCv7c8krlHE6f4G6qlP3DyJo!uYV9W406P3(U=Ac&4~I42+JFHFDcSj@sz8!; ziQwWii#^ZtkN8VMdJksmOa65*qPk^)60kt7iR=Fb6Jqe|h_g;v_q9o>F<{X?4Zd*Z za~cRNot8b3D4;{{9k1T;XkH0bpL44Dam^RU#1$c$CcTHAG8*&oE5!6fKWMZnK-sEn8 zzBJd8t|z<$e*3pg9DqJ1xjRTNhMQyUv?O8!@sQ;Hb4ImCiyT)Out#t@Cu5u(+=eb>qxrFGx6;A(?-5#D$|T6{N?3 zu!uBdZ=oupiVo2qOyY?M56@m?7Bc?ce>*4C0}H8NAUMe}L zbC}yQhd2XIUc+9VeiXbSZbSfyg3S`5of0D5-6crJk!}!>?v`%p?*6v#z4trAU;c17&OU4J zx#pZ}E*jX>%W+K+jtJA#JVv^mAsp}K77Hy>H5i3IgF^gQVnW-9az*;@%J~R9!RoW= z&#o5|Z+$63FW4-nyML=bYfLcb4=8(3e*#U^8rqNzaOs2i4B#pRwyIcL)|w|P=4B{n zB`>)A5KsC_<~-X`|05n|cydCgR#;4A^<=T?<?;JO1Z{^P@C ztK#fxH@bwVP+#O()+H1Cs@lbYz-rYtK=^U@!aGQeKv)`10o4wo*g;9OiydnQA3 z^RMr)NA2qni)653W=>yw<)1RN(1lWbu7L6>v!KGmh74kSUj+ulpOQDXh2}k%$4hWg zZw+H{R5K% zi&t1#>F_7a4k)HCU`=DQvog6K6w~;Qf8#4A$>o8Cd~|el8(@f8jF$?m4DbXtFEE>8 zZtj)wQVH~DlbU8o@_ydwFA#-<21hH0DD1p=cn$Dj*VafDHLpaP+D2Lz@ZdNkK&*+*j|UgCIftJ6se?-~xRp*L`|1OyhV-4=A`gZ0@=sM- zDV7b4^FwaT5<={G*zWe?zm==@KYJ=#_L|;+9uOBS3%7k+eL}oouJ0S$lFA6~N#Q(Q z0@K<$of8sc1@V}#aZn-9%8xBZDGv<`C-zr)?)m9%BZnauHv5Yr5?9Q&{X6o%HGrVf zbv>U{%k+M@-mS&Pz|hW|p#pNYp&c-|=HwV@YtMu3iA}Es1Ol*A2!C+}!69Z8{$g)( zMoj>sT#==)8KPr&M&|K}yy@al7M#J9P-KEj=w6p48G)*nGa?puNwc^{{P1c43BG~~deq2Q;NHOV*T57O$- zsVt6SUqXapxYdTD%@#0wx5EJRdp#b}Ub*C`lDv^>xVZI^ls%ax=pr z(zf6w1$GEGpyVL+EY$YAyoZv(8 zyDJvhvioIp$TJ0XLp$a>c32HA1`asrDQUaN zp2)h4a!St^LbZhw3@{oL%xBLNOvy(Y<^RtKY6xAqj+3&VOcaGiq`!Qk)cVWK2Fkkl ztUPoK$IXBa?<_n{LvXqUWSf`67GKmt!5{;kD@wnvez`jCv{=rlLWWRsbhH5%Y2dyb zb@Zb9ZYfyO1%f%PfUc*Ypzs_^iY-wKY#e?kE9(IgAsH?Mi7CEV{2L1mCj%qx`?#C( zva%w0eTwVcut5vmRG^wYvU%>1)-MN*=V|OB0c&?k2Ryr1=J1hs$oR1#y3>6;S6iX+ zwU*j-HxhY~m%L?6YhXzn@-t^m!3FcLG{#IYZUrQWE9AVd?GLbFQLC->xm1+i3xfd_ zXt~>?*~KXhX3nX_o!;TgH747d{fLD=MyON9$CJt~nNHB$o@+>J#P98!eN2xF1#)rc5}9o0P51v4*YymF9R@>{q{d9?7n8k# z$+Q!MGJj~r&ZXd9T|rQ#GiA*z<28k|DN1aom2+_?fc2k!YHMqYi+7`Gq!hr-2Py$)N_ZJ1gG4=l+W;Zv)#Zt&0f0p4ZK$+YISPN2SaA%RgIGEYxW28A@&yW?))QMCt|&|nAT z(q0~Qyo{4eW`$gKa8^t8Kq(<0*P$qP16y~uPfKN)UvzKnb$hy&ZzGp(#Q-3u1C0dk zRhWi_-9~J*%PS|@h=>N^4=2=+HtT(@df~p_#p~RmqrS;jAmLKhrUNe$Rhwr$YVyJP ztpQ`e{M4{3u*ZQ?Y^^Vv=b04}Y2=%2EYM|e>$ZbMG%Wr(OLB%KH_&Z%I$b^|{6R2L zy8Yx_GZ2sB)z?BtC}!(F%;hL`n{npaVtnn@Se| zse(c{BOAF3zNpUs^8x_4391sOe$x*h54svTZ8JL*OP4E%bjD(!Wjdb)DFM$||L7uf zVluL_5}^6x;^IQ0^szvLQdUwL1tR{Ph^6tW4W?6O~;b< z_9p<~FemndImS_tivt!xzh{h_qY(P0U&rNdK|p>ctR_68{HCc?^EjvQAvz|9qIc<_ zN5t!G)Rybnh|I@p+qOM)df>f3FHIk{vJ=?3sRt_bf`*;J#!zwYV@nKu(qq1h z{TCW3AKRG$Bh&afc7)5mC7!KXW{z|Gd27a*`dI&3rsX32ZnB zAH-G@fJD9i3jNa`PWv?b28YuajE>Gq)78R`^HQiIqB;pGESqDa5PuvdpS_{l0Po|lA{4% zS6_cWX22VbA9Zy}V&7d|U3-jQ5hdF3fPFyM)<9ARKZPw}B1=8Y!^EFtL0>6Zfss@e z4*ZugHr65zkXcN;u_4xlJwNioIFhmb0R-s(WOt1Je^MS)FJ4#x0f;$%2Mo*Y0ssn< zbV22nel2lEY&ciy#*XLjg8PMhhWt3Z(bMs41KTE!U9se)j@f86A0Xi3*e`m2A2=%(%=cRsc39-(>;PM` zY2Ne>h3w;h^fr5WyK$A?+|3{5UY2+mRgOov=w{JT$}?Cnx%_!3pxOeI`9McX-Koff zbdqR02@1!XGEsQcZ2EV6T+I3ecayhUMf){>R9>jm23l3?G-@?UX_e24TOAi^YL(XL z7VMUqq1q!*G19()hl8i}6ZLs68nAxZt)Eajm7^tLpo<-;_$t|;B{uL_@HF{y^FHsd zincBNQZ<3a#12MN6aw`3M5uXDy)*I`cI-0Ks$E7Y?&nkb*2k{8&3DEG?vK^TWX#f@ z45_G4u|~J2p8aiwCn{ z!k#y0U}+Vwzw0>biIz!(r}^Np=NeI!YK^}?)r%F2ZLPGf^O-8`TzW-(;I#j|c*;cV zO8#NUR&Z^8b>PQe6~-rPtQ)^Op1{NV-g3NS6Gk}*U9e_|?X|YH23aseLqm`#_5DTI zkNl^#GL4=7z2l4v%=>Fw_WacIYMZRtFEY~Cb~$spOeF{J66s-BgHsL3&MApubEn-O zsbJYvH$1Qrky?stMl1iz$rAsIw_bbPWUOQv*Mw>Z%8}~g*8ABChw;`uGKS-ySi*H3 zSkU}T$Rq_xEGPB)$^|6%vTRr`qM(F@g!f&6hMY|^QIXJxA6F1>}a!ikkM3YoGb zzgM|`qY~CNb*@gJSFwmh$=f1$$AL<-=Y*1Bo0gVFub6QMhKvEa*QUgQe*HA)0Lc&D zmt_7mB=KL5uXg91bvb&`pt9Gb)yzLtR(|iYc8QE-4Us#(D3x0K!_gvZf;|cEGljP& zUE8O`Q!3?996V^vsEpL5qeObdQL^Ou3jqP)8&Y-&r{A)DJvb|%cFHg?!%<>@ z?x|otsv4+o+9c21V!xMDMX)->?yfw zFS*Kw4e!?IgSg8?sTMo6L*wW3G|9d(!Sp5h(cRuP5vfv!ne(R_i-b)ToIiF1bHI+5 zYBvRS?jhNSWi5V?vrxNnY#o2GuXjy*C2G8x*d2rZnl0i}5Ezx$qZJKtv^agRTfGf;HJkBZLIk>ez8u47E z=`b*B|GPC^4j1!Z2cEatl_Dsz7aPNOmyShD3r_iwi&XCB+138&SPl#EgByH3Hc$ee zxo+w8m&znf*CtK#@;tIscC+Qnj}oU6#`XEVF<({ehnnexQbDzkO(wWy^DiBw=|g@| z{fW-8Ol-23y{_gkh?4mbc}PltdMvx{=>`8FNK)`FDb{H)Q+4`tR?|U7)WETViN~X1 z)flFRLH3Jg7{a}f-N7*aCJ#e-$TSN$k~;nF$EQI?Z+RVfa;gOC#D9LIQ4&mk0f}}Q z3ua*93ORwrGu+fxx+o;68$!F4O{^BO0*@MTf=`(Rx$&x_@s~SxtIaLTMvWj*2^WX7 znoG+3zU7Yncp5QYxt*|>PNahz3YNCme##8>&-2p~juM}8C1JeQ7bzWw%1E&sP!N8o z-L!GLZs2n8MhCXY-LBrF09bnV(PjDX-wVcz0*o+T`y_CM*|Re9*ycRk!pp%#-#SX@ zAZx8D1|QxGe?$m~^w1I1Dn)Ztdtg|b#|_v}Q*x)g%T-%Ar^qOEdul{aNlE-1=TrK3 z+4t{9p4l`-$HuO0Z}X2BxydQ-hS;?l3oU&4VI#d^SM(@-dnAIIcVJHRuuUz{cHlk6 zHaK7&{6#P|vwp?`RE8SeLG~^$;_z#lGQJygvsAmr@wrM_bYC&~kN7RxE@c77U47C@ z{dcS$PF0TDKVt{-rfVzXzuG-i-K2O;5-QU^uNGL7&?)SgB+i{Y{3NevGJfbNCCqCf zeTlHvw|LJ#v)zS>v4+%jLl%w8*_ivR;<+zb!QaFPM~nJYeV|wND=7r!k43L394Vi+ zrj@!k>(b&axN!gGeMyxodU7%9Z&20wCWP)yNNb$qLtdM(JQz@quhojn3n6?>2)Sx* zc=E#ekOrp^OmF7nAH3jFTZeb>fbz}KaHBO==~$i#;;2hP83&d0 zUbC6mo49huTyAo;tQVrBcELP*hl+9_Ci?H=GitVO%l8~Bzq0Za#OZ+%J}@xW+qJt!8e-k@%yqrW0HymilOSmc{LvOgBy|X5b=^ zqu21sCEaWSKmbTch|a8K3#;6lbH64=c-BI5&O>HJko;40T@MvwC3A>+g=;Fy=jm0aI4{Jw|5AjP8h2C36%?DIJ=G_nn`chl75ix2g<|r zS+B3yP7+|@x6>IFBe|0et9mHT_7!E}XSi#KEYSO6_~QT;-~9Bz*IM%Q>^a(x1_`l0 zyEsB$tE#F7GDSqVxe1`(hR1+w_g?Rb{%0Mzzd4UIIHbx@ftEw0ojM+iNATs^_gF9w zrXpLNI(|SnBlAW^pHn|?=h8gp&}Lt8-u21GcdUxBAJAyWqN1Q2v{BZdP*S>zDn#fK zgXh{c={YXFgUv#p^LK@Aw5C|x&1)A`i6oDEio(!Y6jQbyKNoB#T0vj?s4 zQ`B+WJL!+NnJ(c+H`cfg^4HS{Cq5D>_0(4%b?hFXJ}|Ul#`fWUXTq~xTPOYGQe9Tp zYInl^Gfpq`!R}DWip>_65|;XQM0d-T-Z+6W3N!+BN*AoH=#E1fC;rKzT0 zku3`3e1WZ5&^z|7Gd@!ky@W+EKUq^!Vo@xn{qKOhFtbas^aj7fFmwy=Gwo1KaWw8yWGYxH^5))omh@(~rS+>@#ZzdGtanGl0z_*53w}pVyc zT^pjHL3>tOi3PvLH1x~mO#`zh*EP?EoOWjUuM_db-l@lWXUbRf&B_zF242&CdO;Ag#iVgQ9lV=mm|3- zz?m%D0Kkhz1LvKd^a01#nFHNfImSJK!z?XvzNp%XIPLn-7zD7--A|J;*x2NnpnkKD zv%+9gHI?G<6khzQQX1=PTM&j^h>|)$*-S344vdDAwSZvD@#K=19+cT9BeJJ zUE|C!98dLsCkFrT#Kh`Ud^`P_r1&G~8b14IufPmG4ZB|ClMFv_^*x?SV65x$AOHE8 zJUISNhav4(4o|X6@zh-NT8yjeB~RB~E!=Yd%cUgQ^81QTaBAR&yPf$NHM>PNV|3PN zRr~eQK%>kF6ss3&VP6lb+2}Jk(FUsGRqSn#vNDu}_i$O}opCv@yN^?_+{R42_NqmM zKc01)hFleMf@exiSMxdDZ$!af%R{Hn22uzgzg#E-#KOAa)=OffyLg@UFe2I9fdY5vWsWRwzTg9}bsSLAll74o38smBB>F-Zref4lxpTxggB++GDk+EA z3+RF86%*WFSqw&+YpJhv)PXkpFfBOpgye2HJ{|kX<6`!qF~>Z%U+YsCr^S0yNS|YV zgRyZjLjC;OuNkW?whOGp#D2~)>D!61r2`D_0=t-(}@l za+W82R_7==qciP@;tZP5;_9bR(%|cJXX6h&cAOG8tA$~QF_Z3%PoT!{Bd}~`5@KGO zDc1ZDvv%sM%%5?a=Z_>?D>!!@~*_hz|-mUj`nox~aX8JKomC(udPi?vC6QhT@ z%~dsUdZ6Y~7w@(ti_P$=OLkL60W^4{sm|sfj7Wd4`GKHRi$-duQg&E!U`zPi=d1IF)rG)3k5&TM|V+pieOg@RBauq~NiRegaP&WBMx zKDwhR5CFeI0D_wq4eQ6fw^xocq;S&pj-o-ea2|dq3W+)CRWv$&7W9Upl1sU1suU+{M9Q?a~v_7Rq=6 z3CVph%`ZV6B;Tf)avo|N`MI1U-cfXiCGlfWi`xTs-=wWWKPFFeonH0mQ*V08$@?}} zy}BmBrtv&b_}*dqx5mO87S?UT%g258?k^7adHM&LSc%mUorr?26smEOOj0wo4?0sz zS4b4%fC}(Cb_R6XN>PYw*kw6I3c>23V`=k$H zGx6^IW#~ogJGoj!!3+#L4HH60TbJD7Ek|x`P>fW|cabV3Z?iST8J@%@U`etq*PD04 z!x7>OvN2S*Jfb}4D2gt4-M*!i!VK}T&pAJzGdZfdK5{36qr2>EmRVCQ?$%S=4WXD6 zMMXh+s!VIC!Uy-uxb~ZSbGZ+xKRPCSipNh8dt*5)D9ED_?RpB1oDr_g8=ss!!h=?? z<%vdIZySiC)Eqs9nbHmLvnOZeeYQAK zv~3cfrbyAxH1qfcI~~O&(Vg6^-ni4c=vu=0eEk*%6x4Lld0dkdGVM`Oarc>O7FfT-hXGK1vF=AGELUpT^IVL zUJP)MP|B&5LpC6MD+Wr8Yj_LN;~qz5+KHdhn+lhPRF38MSOAxKzQcw=Lu z)k=%>@?3Q4e!L}9*P4_AiD>*q)a(pAU=^WkE)^C7oC7Z>HPA-H*EaUZQc3wm_ybbS z0DLd%c7JwR&RPIs-#wS43Dnry?6a()*O_6Knz_>cwtb51nuUPiBAPR5Xk-MO8ppMv zgsdUqp6ii~axQiB&-mp4DfEa}f!0ZJxWe&e9=$*b;y7_8o~m1RXhrFO_c7*4N9mw; zwf%Q)2XrDg2g4}`t*i<)Mpd7?5qV|w`;%X+6fkJ*^d`{L=)s8-YT9?POkP?SgHx1v z>XM-wRDsp@zO8iZK>83hZhFP$9PRgWR8%HJ5lfUGw`uOBKgKQD#;VxiVxviVr)fo=tMl=ZI@g3DnSgt3Pb`?Z8gOVGsP1Pv(egI6+Hl zfyxv04er-dR9E%*u-?(6)#qKjP*lvzzZ>bp$^EAwe=Ijyqm&#hRg)L#*X5L0#;v%H zM-^B5k^f#t zc(3};A++41V1|!f3ZbdYw;xrQ?8DHI8iNFD>eK45R)LO7_IJ+Zh0FG0Vlr=B3me0F zc_e)Io3lMgb5Wy$rROeUXL#Q&=6pFZ(^PseHmQ2kd zaY4eMW~^r|ZLV;UOTQi^WgyQ;`U5yhGS+)=jtOV)pAmgZORe?`xW9{G zzI-0T;(qx%H55BIt>LYLPH{#(^|?NBxttfnK7_+1ljhA&WoW{i5#s;io7_1&zrWsV zZS~=CX8CB^KwlSHP*q2%zixjbl{YJdw#2q|&YtJoWYVY=9XO4a#1yQ3cE59~gGKoQ z(#aaIB(c`4j=_c`Vazs{Qi5jizTH^l4p@+dSmYBQd-XIGBXlaGvZ=8luSiVhU||;D zwsx=q2P^?di-|b_nJlyW&fWHn3oTB&s&_dc+2CXd)2+FaJ_L9rzH4^3|6TFbE0O$g zp3Y~_V1Pp6;JUlq+9rof%a>_aGVmWstHE7UKkN?Hp(0?nONlI!R)1E^Iw+}^n@+|V zNygC+@uT_EzXR^O1Oq zhSJsSU&)S$bs2ZN^g_R$qe4;M{BW`{0k>TNx2PTWezDmJ-?Qsb4x(w~V@pYIzqCor z8+4cA3Ao6jeTkXQ4cLw)lp5Gu%;a#OKks1Gb?DRVo<*0dY^Wu3&?QXxLp*Epqet`h z8N-_NCaJe+;Im|aFN^Bnp{B^^C85$!e0aOds;rjWtf%loZe*S=LC!b)v~G);5fkKL zy~6za-8E`unt4qx9!%Q*2t-uQ{2fGqXc&CfH9Xp4{8IOpeaZe#YZ2Q#Tr-Mb)}lvq z&RoT+!23QTeqUW*uAFXj=*y`B5Lw29WRk?P}OqQ%cQBe~@Bgk+5x1xP5=(YmT<)R;%oVj#^&$pva)J*+uQCDix zkf-;Kqyy2}7{bWCteV=gw(z(WX(SdZWH>9?l#O*FU0rm{|7-`Vb$VX~(Tjqb)|5R8 zYES*R+tP(7$hrIDK0G$r$f&>4A?1S}Y)YDW6SMeH{{V5OrdIbsaIx?Wx{zQ6>r{=W z>0ktV&Q-wf_iVB8=k=#`lf}E%G7pC*m9jtZ^VZg;<#6q-n)UwA3ot(t?o7ZZpPjHB z`Umi~mcCULuQd~&77Z=p#6>1K;LMwmQ-qP%$`sHT4xu8U0}C=pae2;duI%c11=7$Q9303fR8&+z*Uv5_ zBt%UuMu}6PT=ej4YHR_H5^Yxy{Wp@XZ7V$sT5L%o4c7F!9_S{gt1DCQnDzq1@Y`Xv zAG>Ow0en`fwHDMJ(tUvkO8oNgZ^U@3TejNh*XgKmdJ0Q*$MtyPfBLCt2f~hyMQGh; za%bFN14ADEs)V|s`8k7w6f0sUvSb+U=*f|>Q zqka2Ca4?baoOP8(dOIv*Z*aX^x7eBq%sO@r3~Z%S?k|PVMP_kGh~d^3n}%9^;!#OS zK7WW;>*U$F&8Y^Of{5?6YH9kydoyEBCdOoH3q;6)77*K5f$2M9d*I2=8=l+sZug4I z3@CQGQxJP!(7Uu}9fn~<$3v*Ca5Y^^VfFbc&`*%{V&KL}otgKnk>#KX{Bcq@Hy@oc zVk0T}Eg6bV8b$HxZ%kIo(yP5}Yd#8#-suoJiXQ)D|DBAX& z2IBNIp9oDB(4I3lG-AVQhXNqcZ_GW7;h$Q_^kM*1Jb}@^ndi`Vg>R5j!Ed2WUJGws zz91~N@HI!ZA@glRFXP4;CBUL%f|f4?6x6O(bm<5H<+^$y^} zk)goMhz0Az!_sf;Gk|~%WpB9}5s3d7Ez-)ezLqTvOa`&)uyb?WAaBr4DOperVpy!ML%p9 zuo9bkV-5%mQVtoOK1cB`e88hZ^SGfrxIqa(143u&%J@fX<>gMSP;|}Red`d~PfL9^ z-*ee9az5p|KKZp#w#nX81NZChtap(tx6b8NeS(|pOViAYzonLmK_m;*_K89B1nCBe zCMNsii`t2#0rWGSNM*VeZhI7$q?p zNK^!VYVbkz7gJPJOjihGpWb!^ap@7092i<)GJP~hs`d7u5iH+(AZq;$Y#ct*9xisp z*x0A+>`qN|)7h`+{~?89c=31YKTTD11m_5gFza=FM2xbsjKfxN!u3oCqc?#MkWgm>JW+L5~p^%gQYrbiJ-Ads&_%p@$u~QW% zMY96H63#&K3SZr{vnOr+RVTt)8 z*&M$Q4}Em04(-u7)2RWE27t(V*x^|9y84rU;yPB}IwD%9u|$ei8A`vJ^U%@!BeJW_ z{DS*vm0IG{IJnYF)#0IsjodXN!t`VLS>C%RgHPcazi&+%g?81Ddl_AIuWg;+`ty2B zw3Amyb$RA|Nqr=pvbzcI=XK0!lz>xz_p1)C9CalXr-f}fyA$z~ZF$82OQ|~bLKFdi zPUxaSj9@3;_Fsfn^jP0W!8n$0!GIP^rnaq2&~=05UHqIZAE6mQhYqY^A14EN?PHYX zyl1@&$Le^gTL1Uh&}))Mws5HQch=s7v{~G0j`k$k!w>jrzsdR za+t5P1Jhjm{QO{I353f{6n-o(FE^>Y28)!N6W*sJ z{BRJ?YOdd$q-DdFaHqwWT8zXauCCe@D)RfnUHx4trDKi~S)GjIg;&~eAzENDM7_`< zt`|CjBInouj3%k?u2)v_M(2^iYBZU%x>E3!Rj3dEzK)@-w{8=|71LZ2JNymrc+w$@S-jH_?njr&eRN_{jhN2T@h9D%F#2c+dR9ge|GwHmj?G zBW%jsAJ=jHzRoVOLGO>3({@raNebh}_OAEQf4oeK`zZfoqsy$&^@hfe)8{(}IgSvD z>8yZDjQ6kkk9F~o&!u$zNxI3DG@NJzO5r#-o)YX29fVC>+l(Un|12V7?IwvCRvD5f zf~eqyDLWgW53Eqqyf5x1%q6sX4g5`^TeJIdmu}`rYxT8Tg+sjfl7tN|XuPi}OwSk{ zgLmfqfAy}m%Ly{_I7pj>P<~`b!PkLe;8~q952i-~QrndxT$2-71^Fkntcx zA}X$>s?%)ohC4Hv)0q&c(|ZM7*?MdQjbQG2Y^D6pTU)}grjr%r2m;N*;|u032E_33 z;0q*Yz!JcbDY1V!H7|2|AiOkfWmAKyO%KCsb%APKN$v&DWx286DH zEr4?WJ_BvvkZ)~>m+0`H)jf@S`{8_McfID+x0e{O5!|ko1Z5<{R-5K`Oc`h7nzcv} zxNb6%<45RrN6$KSg178t-lm83Zgaec7aw%}B=uI!>?8!$gUna&eL`~X^t1jHb1KkntePhm1m?|{OAaW*8@Quu~r5pH$~E~2a?>`Kq*9YLXFxn+XnZ7 zUJ1W%#6Z})ruRyuiyxj}vfMZ3_&nU!1KFca#$wJm`J8<{_s1dE_fy8TC%N783fJj= z>$rDm_+wETea+t=6Te?=r*%ob>HTgc=`5~9Vo_+F^4n@0I2`UMRj%Hct4CC@ChzO0 zqCq&nU%~lAcNixURy+&E@@qY_`N#pwxf(B9@XFoaGCZusDs=gfxQgw-D0gBfj0cO` zZfq+m?@jl6xchJ^W3gHzw!vo_MDS%K`XPEtCJALKy_gD9Txk1sSXcHgw~@e%pJw1P zn^|WJY65aPe7oVdg%@0ZliG}6z{fJ}-gsKox9b%P6vnSneJ0Mz(v%3m1cCYFLbofa zUt=1GDD&(*14^WRwR%Q3U&>b&&Mj@W>oc&vIf%$kKl^i32J!_FFBD_{`^*T*U@RQf-@V-kZ!EZHNz3%^Q&c!%AI)Wg= z(A@WokUq$Vxcu|8WBD$W4m?xz&bVob=XyOrse+^@UEK1@P(v7usC0p9NR){jmDKSE z)8F6gOL(iX?rHdwEqk9L6IqUyid>;ce&@kM$1H?zf#{DznNQ7cguQOBy26Qu{D#u* z2@d0`wO;guN@ypIMRhns;S9|`_Y~PO&Zr5Z0BC{JhJ3U`v#Z=A`=}e=2;2lj$2RJa z5ZfF=NrEGRZ;zsI%*lMI;K)VfcBhuDhX1~!oPL$+_R))K((Y%J#3#U-wZun;tollZ zV88VLG-yNr<7tV(r^o&5r;xdK1zWtrL{ICPOmfONOL)U4)@UVgd-zW?(WMSlADmt4AdJw>`^-psS$6K*LAc zIO9}kNsG@+dx;d=e_Ttm?YX2|tK5uh^CE8VDnHW zrc?w<-j$<_-I{UeYUo#TdN`nmgiCf<2w*Fe%J-Oa?wm_<`rDvSR?I$(CtIv@*a;dp z!>^nZUS2cqhBuhOvHkhl66O#=gVA49{rN5(kP4b=ImpcnH{H#ftE%eN7oQU!6ae3N zme0cZqKk0~bMwison8FQ$Ojqae(tJV0(sPSUkFkzey5gbdy?2$e6>Bzr`0c(I8}WB zJE~V+aVT_OZv=m3^W*n!Ufay&yh`!AuS{+^HaPMdiWF}(0`*a-Oe>17fFHlC6D*~! zK56IB;&`k+RdqC?)KcW6HuOXMj|U)&>N0)&gd10e zuRx_MgC;Yz8}2&BUr$JuJ3A^TnW%~# zmg);#CY)9dv)qR`#brtw#4Le?gqNm)!k>?~s}zB8%ml^{9?s^)ps}4+m)j7YC*rJ0 zTR(d-V@dhpc!%nLvQt(4s$3ZGe;UUkb)$?(6GYGA+o%-!$2Kw_*+A7r6`l1yI%4x> z$2c?Gudt&9j>1DtbAWresfSy>0v0gP?X znSe6XaoPW6 z?^A0?K%CRzX|$Gcj+HJov&x;{cQ_KRsw5`UcI7W2yJ@XC;I(=lQhU?)qPxH{K@2gY~ z=-15{e`2`@9Hm&J;=$8Y&39L-c=Mr}GKV1c|HupNQ~xrmDmbh+EKbd2a8L{yCf_BZ z{c8Bbi+L`6+r|n9{@oOr51@qq{@r(U;I&|qDMki52^rCoJ|zXSR3!-)kGpO$(IY;) zVuS^*F<=r2U1;r245M=i_wnv>0qzd~0zPF{Hb10xIw ztKZcg({gA}@=iZiuZ0pfTk9CLLZCzL`rzWQhIh- zy-H2{8~^5d2uK}nKfjOP_;hEQL;DckUOt?vOQu)TKd4!>A>Sq|^aL0gurdGH7?`uF z1X}G2(|Z4&6V|4qbJ-_da^rPrl?%7loP(}j_xVByo%D2isRk_Ncyfvtjg7%EU^w$i z^7w=xv-KzpNa*ss;S)~n(m>wLeqpO-q-vNW{n{J1Fw?E;qNzI&g@sy&z&dA(Wt@&NL>lgbd-MB)KnD%!J=*ur-Z)JmxJ4w&^57nnApx07r)U!zhGv^uAu7?|6;@BBQHnwYqEXp*-=T+Bjs-| z5YNy+W|jolSetMGY!{v+2h-x+xZqn7PX>3kll-!dNcOqe1oxY#^dt~o1G6o~2B6_N zYC`qutKJY0i?tiJS8q58K_gd!YX<7TnKx+^V*DQ8S^Ao<8!3FP<|p^kY`Y5ICguOx{UhmpHPl^N~V3qM%Gv6JQ^mYOTDZX^RP11lLBF4-5oZg3b?Gd zl1>|f<hCJxL zj}{_3z){XCZgs3cz`wB-?Ujolj!AmybjR$Fg?!Su(y<%Q-`QaJ!CSNW1;uv`Ja7@u zGXd=l$fv1v4T8LdC6*y$%5Xk_-N_p!dREu_{P4f-9@(W&?k`T40K&KD%Cu*KpLr$V z6^ps-;j0q#A=~N%P_LRALv?{&K@z8;4-DDBt`BZ@R|mF?a^gJ|Iv!iHUOQ-{ z$hN)Xr)|nVD?|824NV2c1_IaRnVBWvbN`(*ayjV$E`Vu?Esm_kT_%%6LV4oUt`)(; zRozZ;Fj;t0>j!Mv1_dg$Yr0qFd>g+xcI&!EZBE=nFC8N?Cey3?ui7!T&GWuoVr=i! zr?|L$$PdBP=I@(&fm8mqSdB8dh@>{6k)v!c<|V%y+;Bi&B<5K(t`OJ1*Z1ABCMuz6 zUWZ)ws_(2u3i8azXowRS#Q8As3i|0dq|w6vMZi=1k4NlOdg;s;tLWlipA~+&t)o#8 z?JZYwH}A!W|Lm}4e5Hg?Zu~(U{bG$S`j6z~=opE%Tm;I;(3aU%ZU!lk!aOb2kSh`! z&)TECr30$eIYYdZl(S5nJ;Y8EH!Aaevb>cXOP$mFC%~NmE)SzqQDfeXebJQ&il{dZ zGbtKU8mIoW2Z)3qncrn!)&tlAd}AiL-v9%j)LAK>(Eh{+OU?3iz~#*sr$j7~3XOKU zeno<^-RYi68a^&(Fv?G^wBHEw^1mPVPYjS|rhid$IDj^C;Ygh3H0o>jueTe4fQ4PS zL}Nk2P1xbd5!>>LFhnUeBb@%B0k>n9RKIsG8TR{D%B z%gN$s3`4Q_>9aa2TDflmg5>Sd=A$@@bc+`3F9XBRnuoI9PS4}^il(^nCa)h}2tfn0bQnLfZOfYqoxJ~Uq#i<~;a4-+u`uzGdvr1UhO zTONZI7Gos_%wfLwMr|(yGY44Vpobj{+0C>|ksS;Vo3Xr{?ULVJZzY}yYwZ%elQyf; zwE@;DLu=B7l3YnFPO~`IKW8-ZpM8^_N&cJp__Nw*oVxA6i0iaHY;1)o_)24D@@E^J z$hdrU$XR;~S&6)JYwSq9R{I5kLWHK|v3x-#P2KP=0ec%j;}vzMx=q(*4fIEwN}*Dm zDqe$X4=wv{FfO!%-*=3yO1}BzvTln~D<)w20KgeA<57Qj6J)dcV|mTB_Hi6IMa#~e ziAD*~7`}rNN2CLd(o;z>6jX=KANKj`6uRv5HPd#s<386aaW}RecAWM=0sTku?GU0u zJ@0EY2@10D*}fsP_QB*!ps2*e%XV~jq0<3Q5Cs!FX!5K8SD4Z@@5i5Cle^(>Sh%6z zy8~pxub>eRG_!@2*^Vi9L26)F-{V6?A?F09+W1<*#cUL)=DmDS;tA3276Hn)+st0X zOH#FI1pg@a!yv z8JK;X+Ots`TojaB1jhS$UC*M)Z>=fum=L%qo_N~+&x2lv4#-Jh8eXikTLW8`6w>(Y zKx#y{)aG!eh{wequyv;gWt2cgF6or+5|HlhZt4D(_w#=H*oQw1 zkbSK+GiS^UTxxLdLQuXUc;|&ITP)C@N+(XOB!joAR-ehm#$Dc2u|~-ynDX;)Cn=EYftL4Pr-{>R-3{H^HTCm*3J7R`AU`a)-+!vTt>Wc}IB(0# z73BzdwxN;{p-xB856Qj+nbli^3@Fn*{uFhtAvZQMRRMW zNn{x~cYy3SS@=N=fxA8qZ|x^976s%L(8g={nGgcLRwoS#y)u@V_FV4G;aryCm{L7M zGPN5gjt|*ZCsL)!v}X+tPds+U5#nK$@!D8atZhsl8dJEV1nWAQJT}YQr_MY(&Yl-^ zS{pVG50O-#Cw(Kn#D^-y2_w@o)V+GLn!n81gWofK24QOV_hO(Q_)zJsa>p^Gat-T(9OrDDCqQZ*z_x!7KtWJ% zzmV%h^W)xT0ZGc?{-QamF!$5blD4gjW2j_SOBO5Cwpo0yiuQe3qLI#f%`w+OROrTd zru;vDX1E34fle;hR?{ECm!LV2LrFCU$1*n~$1mvvng7ij93fa(G@)0jya`&-cv=8Y zK_BE3k~+jcxM9I`PJF@EFcx#GYvwKa8#FeeTkb=?l|^4G#;uQU-cx&t00=vw^Q+IECx1?K_O^ z_E2a@NJwZXQ3ew~?6tNDVEennBY3)9lue1rU+l}C^vnJj&V^ya5p&N6viemh3O)r6 zffKL?=F{E-uX_JFL#Z1TBu+oqkb6opY!uKv*;loUhiBWlGs{gV#|IBN7$`88jYWk9 zSWFaIDe9(c3 z8O4NtE*MHoQ{zS3b~NZ0H|FbD5pI$>l#RJbd^*iEYZ2m$T~zn2HO4}}#|LUL2uT{K z7GOCw_I^t;GUfD_v^+#e`qg&=)gTVN)o{MWa>XCf5$&BuQo}Sf$*2?Rs09rW^fay4 zi*-cRMIZgl>*rxRNjfI_f)6+&G_jH;s&YxJreJj2+4k`9&JOOs&uY~}NHpaem3^U( z)z>q`?nC>+=j%@k9vh&K1GF|b>1vAstpJ$`v*9*}z}Up(9Fl0CY?ZLMq)@=b~J>OqD0f&=ATqpRqbd|99KS# z*DucT<)_iv&rtp|xYwtbw5`#XpZMl~i}i=I%G~lgTPoK00rY{yz3|)X^ZQMwb@SuI zcm}>Ei;5$oV(F+kx~KJU;|DqK-cl_a_oUuj`i-Lc5KhA{oSVo!#Q;5?!UfW&@jw8*^<}oo3~*oVDBI9XKCc(VZzuEx%iZT|3S`-dSrN5+>j-mKF2Es`AQA!i9X7) zL6UzH9acP>VfZ3?1)1mP{{%EWxj0(4tUbbIP!Y!Fw6H{KSc&~XAZg8cJkgTrSEGPd zbaI&#V1W@ETM8g3naeKdbI<0ll)_(LI!JOwb|!Atg_=PyNv?4>H;6Rd8((pi1BK926*(|SPFrHZJIH+@r(h9}| zfRiz1Yidad(b8X9uul{f%W_+9HG$dhc?q4@4*cMoVNIso~Bf|gzL^3gp zPD~j%p16M?t{92~DKA+9y2ZV;51%0a-nB!=__tzQ;^ZFbj+5Uj>tCZRZ=RsayYOc77XGNnz>S@2fehTpwMY(4gKe7yfd*$Jcba zrIqp0m_kl-({XHx=7D^s$}9sd^8W%%`}nw0{yt0W$o~Yb7H#gJpK^@kDfjU z!%TZ z<@+v5i|IGY?eKJ1G723By0q!&kFtr{Mu`Y zyQ-l|!(=zy!z4VVV4gnl6=h_Nd#^hou6ff@&r*WiPrP{MT|WCkYgOv>`9dA>jyCvf>vLC67nce2I8FHcUu6!^E1 z1;2iQ4==qsF!o7S^k7(}d!1|F+<0hj)~}c4jPd1?zdfI@@cxu^P(S*WB7M;N7cS+n zFn>l8HGvV|R&qFCCuqTexU6;trx4D|gd-j~Y}pAl-{SeN6%{EJDT6Vu@%2E*!U&K2OWz^O9vma}sCK zWBm?n0&ynnZggcbGBMc)powyj6BR<9-lTN94?H0;Zjj}S6{t0tkX2ga>5;H4G?K=8 zFMVgGH>bvu7nq)CSnW~!;lvXTLVs9avWflkc<2EuhXkPR3PV>{7g*d0x)i}YGo^gl z?w+2c!7X6T0$W?(y$S+ej1Ci42YdV2m>9-Hxlr7$-+EK_<<9EM&HKlS2W`PT2BPi{ zH#H@FgjQPR>l5#H7f5Hg_L`6(Q507XX7`^5eixv7Y9g!b@hc}Xv~4(l;w@W5)jBn= z&4rPC_gC-EnX8N{Ubv-_%>Du+^hCt&a176t`x6TwgG$oeo}wMXEg+YnSkK-Bf}USjO-;m&fRXG%(+q#P zlXIK!KoxRKF#4~irsm@EQm4T#uy=j&X&QyKpx_uUaHcKS-rRm1#hW0Ho5%Y~FLm>j zZCL{J(eD(RR#x42~`Jx_pL}LpDtORGRHZ zq?qcR;tL!Lp((Q}6@z%Syf9w)-}~ZMF^c$kSFQG&$=qgLIV*Nke->M;>Fi1Tu<^{0 zSdg;6|Kdp9D7~EgR*==X@uCq9m`sNk|4x7Bm;Ar9^aC6P@QR6f6)FCTkdW8$XaVdE z2e3x7M78yAKP}jG1FV|_0v=7E8E~yT7~ZA)3jRuJ>c4RfP=+epTK~*)@DIAeegmZ% zS#KoJRwNya6hsvXoC`0!CbCMBsJ*j?2H3TwZ>xh`$-fzytxf?EiIxNms`!oFlARG9 zp%|f@9r;yq{gdZXWWy~u57-~~>sS9wcuDIHq_rv-gb*6ZCp5Hk96rtRT-D_r-95G1 zlD&Qo`+P(AIXSF2#P()2tKzw8Zi#B4vZgD>w5nmrqN;_(-68wD)wD%}!(<`RubL^# zB9$VziI+jo<9>I)2}B_O{^7acH&XOB5g&y9-mRPTPdg-Es)?NNlH6m+ev{>=LSyRN zZ_+8_xGE#QPgk7i-gr%zKPOz_qpC>6#Zdp{u1am$DehuQ)iSanK%c7W>*((iwEK>_ zvXTCCXe8mD(Dov()FOO2OV2o|hB8~(q&RW(`_3UkQ}C8zN_##$V{X>;-(lWMZv~q9 zU%v5uXv+7wAFA0%AZrcZ&Pk~f?>{>q?k0nMRMTW@w8X$xDasIjw%z(7um7cNAQ6JI4l(RBTC!Mg%c_|MKI z+~0i=UoK@bWXX$kS&7U#kf{2$`UU;*lve1VkKuVb-AI8g5wcihdqqW*FOCpKb)BSk zd;5*vr#`NIQ)WtQV1?G*9kb(tn6>7x+!;@qP>bWIap0*^Tg$QDJpw18C77i7&lfPT z`S+p6L-$Sky(H21kJh`p zud`I1l1+l?;DvI3H~pE>5FVP#C>`(4br;Z1(TG;kRYz}ocv)l z8fUsv`vlK4@eQ~?fkng{qdn*v6tnH`NP1QZewzslZ)TOw4e2)S>k4W;qYl8 zd91Cu)&_)=iYCGz`TxnU-Ym!>~d+1J7Y;SRcYmv4h{}p2j~H zreW8L3Dg;IyilciIegd9;z&|`)jtDC;s~%Dt+y_mvxg4f2KNoH<9Vi5rPtC{eBPaDBiYdCVhpwfkF22aI_KguKh1g zeq1;0px%rOebN4m{F`S*_6xcOIThk`O`cv@P|6F(Jrg9=GTT-R5MOouqF=^6C|(S- z35sMUM_G?yH7?ui-9TWBEiMFQygiGb0rR^b*pdx5R@C%fymPTyvpPR&B@&xpE)g}K}P5AX?p%86EOQr$;Y#8CS%f{+K&2V-!1CUFoVONFX%VzeW4GsLtlR~M1)O~LMvw&;G&H(oAYK;;`_y=vn|g~9Lg3^|hmZn$fm&Vqd2w^Z9I%;=32TRqb^{Wa3Kzums-@z9h|16J`)QQ>${S04zW|H+Xm zd5)?UZW)3wqKty@W1ZR7EvpRQ+xN+skxMqXEfjCtiHNpC*PvpC8~!@g=f1L(5JKo} zoXKNCo1WqN{aQXJUUTakHvDi7P}|6YY>IWZTe51-3A%L=lhigzDgH91J!%B;Y>;=+ zk*K^@(2omm;XHW-DT|VZQ%jnd4gbwm{HSIS@1HQwGc!}M#=Lp+@xDf`BUx8>y_Jo| z!(92^xiQ~lq&DdH-287E0&G6s{jZtBJ=yO2a8HE4(fB!Lw}WivKUbpmoaZlUp+S{! zZpU#vqi>u1#7caw4_eD65ORi=-4Uu2ZU0F zuLev{$_upJJ+d-kDzSO^YYGB{vPH>)cCWPmXf=zN5tJ7MI1&9F->NfMpqj~wl)n$D zt#XdV-$&(J>$P6$*XdfZTdn(AILvt*vD@k7LLqptA7xv&xMV}FSo9_YG!3Cwgc>9f zH`VUhUrxEj34c{PEVSHy4(a%Zuxi6X)o<$IR*nDSxme(;5%P0w`+R3|7QQx*q4O}u zNThsQE@N=`f)`>;w6l1V`qrY7Lpy|6hlVKsGZ$UR-FNl;C0S9Sq!vfJtdc&(BIwY; zP7J@1hm%z18wKenP92i(ws`8~$04@RhgdK@)3#dPI0mtIU&Z$e6aiasu13CP$ocd^ z=^$b7&WX<7Uf0G@OmMuAmxax$d#1&6t;?RYdL%k+hEU9$seV}(7FzHi;Lbpwv=gV} zRSMF|;}tNun002)nbzlNyLrg#7ICF$^Y$6I^W-C>zFNx z`Ar(q-!W&yxThjkFL+$ORZZ8`pTB+HsvDpg{Ci+LX^DUQbXE*I54luC;%h>mjTrvp zKo--H0nOa4#i+Ory6HS7F}+@?fWaT@)40-t;VhaxoKlJKa>VbfFoM) zB%+kXq(na81kV}XN-pn`>YtS}9>M=P4TL_+ipMZ2D;8NNIGs_T+j=8F(s&CIMk0*S zOI&SY_9Rlr5%l<@=AE1ime2=(J?g+U2DHeo4l>GohOBrCe1b4Icd-QPCgQ^>uy@(< zhLDLeILLc1lC~KX7T^k37tf6h#A20j&({n*^Gd;~5>miVW--Gvf9};2hMkHr>sgxe zNd$N!e_fj1GC9{)9BXhq!U%OSeVkk7CgEQu#gC$}m(&fX^Z5IHSU0`4yEED11VnPahHV91(7~+Nw8j zp_ztgE07AjjnZH6|3e-`<@(YQCphT$>aLjF#dC;zKgdbY_c8ROO#3;sSD;Xa63kNl z>_cI;h#B^peepiakQfT&HC{nXl@9AwF@=z%#4r7d-eLNkznX(fz$ic2*Kyp$lAW{g z4M*KE-&|0|guUxYjnkm>XFL4)u0Zu*X1bate91_i>6cX_LvIr@rG`CCHag_AUuzEH zyI%;Q9sc5@`>Fj$+ggVq=hOCvX=9HGsls0O@qJC2Z}%+y&EN7a+k1P|{*YIktk&7Q zT+&Zxw7h`z4*R=g_B*8}|xsDH7Jh!jJSZ)eJ)vg{NIwz;5iu zWcXq>?k|qywb^G>8h+BIhCUZwgA9ERveyuAIQ)E8WOHaS#Q{;hocd*%gBYv{e|>pN z#qy^kr$TTZ@^p+@x3MuYFvh~PS1O=4F>@07aFcGz0pd|aYGp>@F%F}Rn!$b0XR0b* ztv@9!52hw2GtI77@FZ^Ko==YeTc_~5&$-5HF+6QV%YlZst6Af8A_0&4h5D!H7#w`o z$;pzPLuCH!FwAXFb1u5VmFbSt+n;HmdZ!>wu)gD2X#nH;IipdRPl=Z@sy*S1aKoD4|4jbMc{H7QfhClDt8 z1dm;mi$ia&QsE)v(_QF8(5`-RPXljPYm$e7=GS6+xsPc2R)sITrpbb7D%LhkLyKzj znldO&#Fw_?#jRL!3@0Uf`vj`@isfV2OZu}J9POP0K*`BBv|T7({2;X_)DTom@SBv1 zFGj4F#7(4OGd$k*p*31~I>HE52g^?v%g?s%SMDAB$abaWo9wlroHV8$xo6S#jCov# z%@-|Zb^Wf&9rB=j+v(w%=JdgL;H=73EQ+AMUnvX7%*i#3hslvGq|z`htTHdRbSWOHLfSw+S9_+P(jON)Yn zf*b!bcpW;(Ehu!zw=8%N64^dtC5)ByX(Xm~9_H1*qBEQBIxMj87dfnq{=`R@J{s(d zO?k7KzyW!^N|F8gc_s_PC7Cp4PDFoI;?F)waV(t&t6m<1HBIcI{^xrwJCu`A0Kg)C z<&F#c267d1OKOdkS4*^a1w|j*%FE=Nj)wH<{7w~tC!0_pYaWCk6&NAZGJ-|o)iOcj z6!;$4eI8ChZyS3ixWFs1GTMZLRjECGQIem>D>+;mP)5o70^5jxXRUX?^OOPX%{=AR z=`|C)_60V$q1McjK4Bw`JrgMrf~~2hD@21uXW+Jv zT*um_MClXH;@l}<2ddD`VU%ano#((m(jm+TFdF+Fa#h`5Kh{6p8>{Yb2UJrwmaeb9 z7t_%hZ(NTis)|8K@2sJ}F%W>%IU=a;a+oX$q^H{J-oKekJ|4mv{DLkPkR|Q$*)=D`jF#JRcH#teM~k=^tz> z{fF?=mBoT7=_m$Ag}}aCa`M_?aXj2y7d znOPYc+QN9@P(x2dnhFoCc#wvD41&Pp@Ag1dB36&?uY?wRA+%$@*uf5VBuZv5yF~l- z$YWj&35sr#9p~151~z%Q8H*tV&O1TQcwvcIFExAkPm`N_Uq3vg77;uuB%NF6|C(;{ zVfxYbg($={Ou;3P0t2RP2F3{*G4BMF%O9Ryy-%IlQ*)R&bcC@QKOT{(%w|qie42jz zb@?)Sd9Z%Qm0QxiaZ0~_iz++w)l@m)dgB_kEaNPFnmt zIxtS_froRX z_e~k&yONTUlt+^$_Af(|POWDi7fJy&5Axs*{DL9^Ql6A~WtkMQB9VYV+(EOyE`@Puv_?OV*y{PT~G5az915iP`CKq)w@ zj=Dm|DY}*t;cfNEnfy4xC5GhjOH(k=d!DO2ZjFV=9+ULQ((gNw>Th_AGA_8G71laX zs&(wy=TTt8P10;GWsq`ZVx4HfLik2wB#JcTPX4a#ZpRQ^o#mvB_~I)t0CwJ<3*>D_ za3m%anyN8$*-Bng2RtbS z6={51(G;~|X})u>G{v)AiQbSTMuKF+jz73YS}Ayc>dbv}upO*no+hPReSj7$a2ZGt zBzo{AXESKOKg^q!2hoh5=HTpgrm01UX*k$%H2YQ5DtXYz)6?@<8*~$Bwz%EQ)tIZ5 z={zjh4#&pD$u5kV+Ofcux^Rr>Z_`P67KhO_b>nR_(rZM=`hH2lAtkK~T;zU7C;7U0 zaHNUmJr(7H+V99yWRXz%Oo?_vE#K@~ye5NkD5F^|PxxBBVkoEFbeu94Y$DNn|42^e zLhqS--|%Ij$t=|}#m6iagEOAPpQ6D!>p&R=WOo6O-5Jbhx5HS$(dMMuQe8I8d-wQE zR)MZlRunoM7d&wyLV`MA<_cf-lfC39*0%M^KeJEO*%J%WK?H4_cjv_V`ZA(?wlQC0KXnHggqr$%B6K=H%5MGLI1 z>89uhCk`0JhY*Bf8Wi5Ft^mn_#$ zz5AY%lM@sa1azd;RnV<1OF=~y9ufi!LHgF#jI6BHd3i?G*4A!r_kTAB=%$`*e9&Ih zynER<%p>c(t+&WMp2v2RC-oDOXM1C(VOFOv>n|-JB&qeP)&?b9AE|er=**a$ z9W~te9U(uRsqy9Yb2yDM1FJ$*on}K}rnD_?POj|`n_kYlC4aRl@uh(=6Q#e1-FPpT zNCQMlYxhpn8QTkOFE@Oc+baQTf>v@vCr}3YIx!gp9eAxD)fdu{(Hn99QO8lX2L^_8pAr>B!O(Q4NLIvVNOwx~Nr!`ywl>dr>5BXAnZdjGQq;M59 zj0fd_Gvx%wFhzy;{ZQo#`lt@EcdvCwZN9PpNIO5j3D&|x1k3|U{I|y5E71x--L|^9 zsN57gkqy3p^t6(nI&vZTJ_9D7of*5cyYGMgdleVc1qx#l6Q?C7n=>csfE}>SI!}LC zhOIeMoX@r-$fH2X_gdG@W-Uyiy-zX=*SiEK7{#(0jv%*IuPqUMDgKHZ zT*{n+%D>x{Qk#bPh0i`KEc85t-*1Hnp07T=GceTe1~7(J#}FP8bwj66;5E6>OYSj| zt@q6ovQ?Mp9II4f3tl^8>Ku9y{WQQ2MXV6nFRv!?ho&5CE_|Y+>A1MHF8a_RAIr=# zdf17jmHEpgr_txkM5Ahg=Fd^d`4H!o0q0MHY50UZE)7MZf64uqX~?9acZ2;2cO1`H z!l>Q>;fAKNuxkn?ir)8HPP_p<8M2{Az3qSwA(hG&Y{T@R-*Ur+BFO-7Q0n&_@S|6` zool9t0?2POp4go`?7xHrW!FxVMGx$B0fdSywc6b$unKG0GqP!|3ZACL)9O(!^Gxy z)p3}wzdL3|mmhi0d*=-xLUK;G_n5TK#X5Tp>7OSB zL4>CZ8Wxz)vpO#EbO_r-;GwY9Q7}mFT2dZd-KEWr2!IBMU1V{u-u+Kru(!7dLqQiG zAMV`U-0HMSC3Azq!}Aq?f^9FKw6zyOqiW0JtvO{hmrd`G!4LxnsO}ZSP}E9F=kWvC zP(YLyX43!GfxlqXdZll3^TEFw3=~7eAlWOglG;qXu|1S|z3Qiq z3<$4eh6s;w8%D+>+9jP17$~-(7bldpq~DzWZ8SRXJnJvrds(hnPbd5kq_)(RA^D>eklRhsDwJS62`h$ zqvuZZoF!psbq|pCPDH`W?(SxSoJ+GJ0fltkEXRTyEQMy~c4iCunq}P_)vGqx zqOfXO9t*uiL}zM_L4K2qn+5^WuLrPZW22?w4VUko!4b%~g4ucYGFgpqp=s?ia^quQ zm#rLmTJo?wMOCrfeF-eJX+YyGD$SAfrMcJtcw4&{sg_5>)~LPDp`98>-tU{;<}(zAlcbo7xF7yH?*VF%^IDiTdetO3Gs6`cWYb1=S4u$w&`F0s7lSnzu`p3iG=wBHK zMmJ~i1V>?`6p)AwJcBTi1iHjdxN%(YAQguZ=}?kR)7=^o4GT~(KHqMLn+r$n%nHLd z8yJ6UUE146MURslm?UP~KZ--F%O>pb!q!Nt{^6IR#;;N0rA{H9J97!Trty`YZ|dVd zLEdKFVeWirfkH_eP{gR-IVu~KYLLm>tcHMvYMNtzRifAi_-Ra2wK^wLGG%|Cztw&< z@%CTG8&d_@AJsMq=4CB)Y)%sM1&GN_zHMlx7Yqz!Jw9UG`wecv3s{5{Z;*y<+sl3A zGxQ`^92-%)xcJhFj9ZHWC|YSxRy}H)7bt=DYELz;C_KYM71i{17x2QM+7#C0`1ND+ zhm8ev5{;#fCAc5$sCo<*`*zp7%#El@uE4Z61u)pyMD|BiL{rw?i4&*ttA(ZRv!C-x z#j*C8XJT36RLunQgUshz00Ra7sO8Y!2;!1^eAqP5os~g48&$v{l|h8cIkcOo8`#n_ zy$O`6!3F%Nj#;5V^U*fI;UjkQpclTw^;eFhPQ+Y?%#MM zkua-SNsO`Sd8#U zI-?E?W?~PiTAmtYr*~~|o@+Er$bjIgbhowdz+8DN-tFdx_-XWft!=!*&s|JCc$sMa z0>`anj4ykP?+nLnFpv-xe`I#E6h%P_gw*N!ZZu+B>Ay(dZSMZ*DF5yw1FcG>jEU~I z8T6;4$s`?)17UdY^j{XMPKS-cf-V$ePQ{5XuTG7F3HWFhQQ%BD*K!pGL){R zt~*tEOc@(QyVXSIDN=23a6HZIlHBgOI`-$jo=j%anyjIlJ#x1t~-<=V0eP>0Az)E)+wIglr57{X1CDG6Y}J9d4D37g9WPUQCX1{tw`-ZNq?(9 z367y9FB7Ws4hke@9On#@VxIc}?TtkVLSJ20Wqb1~Y`{+~^sq(NOd~S6KHUEb>7w|*5{YtjLU$<@?JZFb?*lzq-`>b#s%uHU| zri!9n8nfk^(m!m?STTI6iU=sU_05LAyT`n{$7+`8>ypHH|A22cS&vyy45_}8Bq4a{ zYC`@q(R7`yTa6QWi1=DYIHrNxrKC(XCZ8nBnCPC8zUVT&ZbM*#KdJ0CFS@06t($y^ zXPNgb+_qv_EBIk+j^&B7*8NOrzS0@IQ_q{Ld>Ht>m%Z4VhztTE{?>?{hPA&-^a-# zcP@VbsFc~+8ox3f@uw1Jap|0^U+HPG>R0*1#g zqxEX%F-e4g8hEhjtvj#NOP+(s7cN>Wk0Rb7zstjkm7iwtdBNYwR$UTj87H({YE!1@ z?}SKkYE;(2`ypMu*R>I1 z8A=z{XY(qG!lmY`&5GoUuO>OfGo;C2Bk|K zTFn$r7dT~gEL>&Jd!FDEihYbn7g)(FB?iCl+}W^{h3 zrkDh}|LHkaqkWcl^!#8Y$)KX)*02azp7$X?+4VSzQTfDY$q+f@E;Nvf&aN(vi+|MV zS+8+ps_4~Rq~|q!2%hOLTKBWKi9^1en)gBT@9dSzk2Xw0XxD>ak3VSIv!f8^2O`Y- zESv5mY~euFAF1rCyBmLW?x#7$$G^!O$a)Jw&4&%C`BW3V0&NmRY|K9x?AWIUV0!6s z2=uH`2C3m8+os>Lq``iyN8+8*pOT}xuT*?SD3tC-QeQY$+q~CE3Hh0;gi#``t0PY| zr8(DRwivXh0tlnzHMPa|Sx_SU*gyeq@xZ32Ud*matT1BM_b_X47>Z*nHGAit?wr6L z;yl@KURK1sjs2cq$U|Uj>~c_Uv*o5#ZBWo>dt2^xmcWu>#_iDL#`aJ5Jash$d`)^CRh<67+Bs!}W1C3<$Tr1J%g~P6imLDGb}R=99AUs?!64!p7=9^_aE+O{%0L z_Mmb|kiUV4dtGa}9mtMvtr;ddl*9sIpd={;@gXth1>jBapASXklf-FN>+yfb1%dJl zXhDxiI(VGAKfp6YD$BFoD@wHrDRxYGb9qDIcecOXrIy~EZ#X}UPAhVjWw-Yh5ppWw z1}*}iIc13zgb2wOh1sP%hd}6vz;t#|J$x_9@jm_Ifcu%YSqCpF(ub2Jf?TqcUFi({@+?6GeqEKZzU;;n*#y;G9Gc|?-)683<;6ZF49iq}$oJJjJtjB??Ki1QMg^am zmgEDss}|#hOfdZr*$!AQ2Q|N$zIUm~9{<>tr9K{28RyQzyw|Ln%J+gQH>7Ev0Q`lx0D|`OvYEKQUV~$`M3KC`ri_b1OZ0}*MJ?dXT0#$3m zLv)gFWqX4VMj)?l{j=j><9Jq3vFvD5ZvpxWE~yiro^nEQ%|*yv^wqk|m~moJt&`SI zEZZ*JE6#$vXhL(Yrt96JS(&Z zNOGCdNNO>raEL;)zPmt0Oj{ZNNDU)W+Mdg1&XWleOzv7M4}41m?(~O;+Ac{2kBva^ zfu~-dZ7=xU|HgtC6&BhPzWpmus=MFCR*bwsECxhkiR^Gy0YvhsI~G@}snvvM;@bah0}vQ}=j(l`p0<;% z@(}@afzWmnw6f2`B4x!+C_EaDh^=>u(?3}JtHZ#5zZ*O(%x$VoK$3v!_S=uF^Jyyf z-VN!qQ*NL)xr^Zj{@B)S%$fF+z@wRDaP?h+#t26U?oD|LL;iCCI@*&aIeC2s`# zI{}Ap6u|b)ailWp{Z2jDb&;KjmW1|V_@jiXPF{IZX|A{JAM)Gzq+g*(Ov9EfF+1r5 zfs5n~g#5qWX6f?l47%-Wzuht7rH1EN-bhBKsocq>;NIIsF87PM3WLyiy)FZmcm7rj zIUI)@vvqu8H@ewzlchfOE?`FdMEX9o&qjQH49akxdWSEx9KNI?*wYQf0(>K zq}Bsg-}*LHb17AHxxIeLZ9$VnUMJ8}n23AlN)hv~)iMdoO$aqUUuRogHYh1*?B(Pf zfQGeH?V@dt@kz*93o$d7TjL(4=fz74!$moU`twubL}%X^T&1QZWJ zx@0Ua-b0W8(ZdGnru3k4nS{n5T;Kp&Eem9+2B8=7@;9GwwDyKw7cwC5pC_8N+9zCD zuU_sZt@abr!R{(1$Zs&jqc@lhs~J zLlqqxc!U1`I2bIe3fWI+uAG;5hXMm9KhxW!Uoe4MB2tfybgv`p+D+Q6z3#P>M@2C0 zR=^8RNg-yX6Q!Kt-iM8vP8QG0AB#jylTLVbM>{eiNY{%I5Na(CAak2Gn++tBy*JXb30ugM0-(aD`A(-Q06MFRthL@EMf?l?_@>(~MGuVts1|8Hg zm8Jk%XBmh!bU|!`<+P5^KUg^917mV(t9t`4W$!Ah3%8k(^yS_W9iS%uS!7I_!+bLf z@zJBu33Ym(dvSgw00uR~PGJo0bwVh{_sh$VlWgejizQeqfCD}JGF8*h1#Bj@I>x>+ z;KWSM4$D_{!mt{#G{<3bXx{BcR%$hxQOO9iDr1(vo8k*r%@a83 zQ1GdKTwohs?Us8Vw&0i>Gp`{{oGXN@mOjU$xKU-}agC>+8$fNBYdBdL{Ck<+jSxJh z5BH<*$&n}BWPc_m_1?y3)&vmt5K&f~;9!rYMC|NGHEtpEHn~yTUf{4Q)Bu&Twy`75 zwsy2Sz$W}ZJp`*--Jb$OlkOL%k6%|uy#1zA_E2;OK0_`itu;z`kSU-SpUY;uqAtTg z3Z#8I{vsN2^}|p3cAL`l`=CtSwr&I-IOxk?7*N*bh$crz>xId|dI67&#cWgUupQJU z!&OXTF%NQeHF7aC9t;fR1%x@TRee6lcLpu1yN``PNug8rnC@F)b~oXB=CQA#?oj|q z@v7znzO_FIiP6v)|Fha5cm^uB2mb?FAT_DIlkBYPl+E5&KK8CWqM~yBUfT~gE#kb_ zNr{cUc)UAyH<4PMp_M)YZww8 zL=@Ry*|e%$6VpWZY>2{|q^+2v626k?kb0IetjkK4-BL^o4~=-L{Z9pf5}4~0IVg;z zBD5ICGJlP|>QCW-(a-|7DIh7Zfd zJvbn~Ov_znLD10_Ap>&AQ^@#^!^ITftcv5J(j9!c?r+G1Bc~gfu?5+Fp0QwlE%NMR z56x_iN-(Xyb|#0lO3XM2VO7a??Xln84luGpN0bk1%Yt;OPc9Q}c85u&f zhyxrYrJ0Tta$w$zj|FJp>V0jtoDuX-%~mJ|m`%NJi1kCsLXqkbkad{nQJ|HC(bnXO zzlQrZP=S>!2dxq%c!ZK4KX!s1VS(!vpNG4vbWWSK^WAX)w>B`-@yL3CkapkdP-I;3 z>-x+{D4{B><-J%sh4{{%5=M;hDf1fKF=KJf#G;A$-;CH-vv_prn@AAgtN+PV&pzb) zbGCdls-033vJNqR7j;_4GC{$buk|XhQN3oWuFv}50|R8zp(tx^tP$1}PR(``n!~b_@y}u@+sALiO+W%IYF5Ojx5~TA_tppP;{w?4|%@bO;OEjU0g@#-6y!q zPxzz;{((L92$TBPabQh09v!P^!JEcPGI_G0XBu{7$AFsp!#01zo9Hs{ohIhg04zLS zPOhJ+0C-6|)V~U31Uds_CE_u)5o9N&OP@B_4s~5nc#4qYN8U|k)rA72{gL5zrJF6G zl0~GV$f1H>McAkT@K|DDuCp$@Av|umRyqdh!mq_iDs|Hi43{v$lxC#Y3@LoB4xnif zY)My9ne1})_xA^DN|#%&*MbEWB0Kkqhvj!_2r<2DMaUvmlP72pdUZX6h4=00aFQK+ z>Pb3ZLcoKw9)@wKp@@({ zv6?$xBsY=!|I>~Li3GX>Ap^MVYYBs*!cE(3E1O(RmPnQ0aRl!82i zg)b2+VzgNi46R2IM%@L10B3M5SlVOY)nJz4&gUPmfXtAPq5!q+M;s4C^-|ZNMmmT@ zX)K;qUT}hz4G`i$8AzVa-m?hz&tOCf3`RgKtg7AeS8NpT$_R8E?D)fs3q0xanY(3V zy_vPfOya4;Zv%fAD>Pb4&62*z7hu<5#NM~>uwgja?G9|cH!D9=KftkFDC;6g~?>5V8^s=RID@NA7V6n;JeSE8J zefP#GDFD!r1=)T*{C!E6??>O^0KLlmH4)2eY3zyqW`{|JMzi>$S|jlWQQG}Iy&)?r zj_EWxI4>k#SR$`UtiACWuyms|p~|ZCOQo8-lsX*_dvfTKQPnI?u;E(sar4zu5}Wzk zw{Jb_$hdv9%&-x9TjL+^VuY(TjnyF&^GfR`kAU!<8oiIIm=GfzLw^DWv{B_YUF}J2 zr@j2TWcuaRGz6{pJ7cj~RsXV2Am2MjCpudj{Yu7G%!(Jz^KlT>G|G5hYMT{R?L=K| zLiMLg9dGl`$JOa8G(L;#`NAsCnNJGx4_KuOpYNZ%3c2Pn1%%;=kYGnYxoL<;yH$(9LZ*pelX?&rY2=pT8;ZY4e_1CA&r06iUHpFg zGdYWbJj1XLLjQ&-efsu^8O?06eNF9(YDP>k#5CYasG<31Ll`E5$hyU7Z3)Lrr7%9jb1a1!I8r&e0m7B7hb?bZP_CU z{r+;q+Dnz!awRetg(90hZ$2YvOm07+Xxm4nU?^r~^Q=v`pv+Fx==JT_kTW;y`jWkU zV@E5g`|_-ehO7+ys*?pSdA99wY|WvTE7D`*sDEMUa5R8vb9^mjQaDx*ymFh(oC zBn1GTDM_v5Ov4`3As$J2U|%d)ne_%8!Q&_%cHQ#7`U$;S9UkKTq7B^bU94c;-}zj5 zUx{>2TV4L|4frX=HOsfadNM4%_arc)gWs&3YVn?+}H-=N;&x3oss|V1z20| zSU+X{@71MW@jquv?f*GjI$Q4lN|mT4Cnw*5%1cd6?cW?`EHohkw(I{#)K>-7v2e`} zE+G&cf?FWCy9Rf6cY?bI4elPC;7)M2;O_1k++FXSuj;QmZ<_};Rm`&PUcE|5NCkzB zwjgqXVk2$A#0L=nI>a`&5RS^dP5cJ|4B3aU4nY9#r>a0x@^Z?Ng}(uf%P1(0EzX0j z)BCsQzw>e?ZGkN9y!G>|o{1tV{xzCb#NZYuHkd4GHe3x2E2}=wnrz5I4S@M#p_T7S zF)5x(&-by9Iiz^feCNn72FnNauZ$B;{%ttH^SZ9eX_7l5_Qbn@R;o`fN65cgj$mCP zGy3+`QN9K<4(fxp%e9Q9o`uhENopMtRSS;6#@%g5@s&qwLc2e9&cQD2b;5)6f_{fA z^yenk&5{O{Cp>r?2NqbY)a9Ql&9l<=RMuoxA_ZH=RC(*KR!fak{O?7XeTA~=Dxi-+ z!4sGM=g3x=eVwjfw(h!6U}xAO_{8`D{EcI~l?YHfS-VG!h9XRI5gUC;_}g8sB2 z6mr<2XPBZqybXH3^G5^2)};Vs>tsfi12|_Sq{y&Py%GK9OL^Z>A4%2WTvJL=Td;mjvvp1 zgYbY(ll||~(xbr?FAce9VT|T&1Av7 zKOlCp@G-^^snzFgX01YWcjVOw0EHJeoLK)2h}RpsyR~Zs+{*`N;Sil7)LoNCI})_y z*+ydTmN#>8DWqo7hc>WJxM~C5I|t|yB?jsN5g~oSslz$b>Dsq{wK4@ zDx`Ei{%a(RO7embvjYu5Wstk{)WpA%e6h}{b^(H5L*BZKUl2ws0-4q>u#N%0XrY*k zPWrcTWj~3ECd#iW$lgY*;_a4=!PJK92ie3M|ZNvRlllGil2=kPP}pVK>QP5NWIQ~Y$DMS{EdM>=`UN?fq`SL2Xwfrd;(0r#e(U)P9yA3{jDkgcldY&OV?><;UfL_F6*;rppHKPRTI3&<0Q-Iz2*0m-tek?5RT*5a zrP#}&JGkL&X@S@^j!T71>X{RLCU3%0qT4z97bR}UBSa?QZ09KN=PNtXh(Flxe7LxC z?Uz(}#zy?eR^OiRT%1tG5-&bDW#>}6o+zIzhQQWO(i5IAT!y&oa#{2rop=#_nOG0{ z?6lvj)q)#ye(rv`n6?J?O~;;FB?EIj?`ML*i6{XX%(RX0W_5Ik6xx)I)Mz=hDh&T&Z282@!R~Lg^YiS{K!Uvyh5vsn; zoUg$<;6WIypsn+mbAvCPzZ9>)Ppf*Dh@SIxM<&KXc-kBD!Jn))iTYZj$2QQ5kz4pU#T-#?s3^(JwFZEsuou7q7%*-wtjuhMKaL zspYEHFtE7-CCx{L`HJt2xx~1d3b7GxqfHpBMmTn~`e&@}-%7Z10_bD2F!K~fbfpcd z4i$B-8_|x6{WI9#OIyr$)K|zVdfHiv#IOiMdTcg2UQZi0T$^=#e0)H2CPzD|LKE}< zo{c)P&Sx~wuYy${(=n(2*u9{9qk;mE8(Ya%!xyS2t)f)bx_SNfsgOHXznZYBeCmxs zxa&)5sDHQ=E@94s(o#W!hvi9~Tk!vroalV)t&hyx!DFx|Q?!~QY5sm2L6g2<+Hoz_ z*MOq#Ow@om%$BZ9QMkj1wExh1K>e8X?}^?zm@-|ipj6xSzH+tlI#b}^=m%lnS*RDf zt9s}A4RXX*F1B+~Z_m{nTKsb{CNaf%y#r29bCx}+vne2HYIl+j2T**AqmS*_d|`SI8sTHSXy#!L(RRs^?p+gd@u7?vby>Qh~@#C1eR|#w;C?u zM1|NQmNBQviM$Z1$iw?~D6L3rYEpvz9VX&%3EY>e>1!W-IcaXjo&-ao!JR#tWs^5vWpN<1%Oyt1 z^s%by-zLQYr=2ga76_1_mV#3_b?U#9VjU`Q6*gXdX8{=$cj#eLBGCgWvBD> zpgL4r=ei@KsqHJc8-Psk`x$K%DeYu$!-K{joU=Tu z-pFXEYbv->R!$>J|(BcObv z(}1?bp(!g5$NQ6bah<41e28Rj?3YKj{S%SjMi1nD%p-vdkNjzdmWqs<`7s{2^!Z9| zE{f#daG+qOLtKYus+yT6a#{v)^e6cr@b-_qfEMYp_Cal2vIK3ubvYb3RKNp6fzRL) z)fR)fLw}>dwNL36+8}$Kt&B^J)94?7f}R4)y954LPMfzK&h53Qq61Ny87os=0uIY8 zMYyProu`mwA|FqogoXS~qIay#lZrVj@FMZFn|<33^GuB;gh5A_gi)ck3N-rrk31pQ zE*i<5^It^=S)5n-H8~@ucSQQapNRh6*J1&ipe^9PmmvLAgb5Ugd zz4oBzo-g#ur<~x>JyWPauj=yd)_dH95X0*Hi(H~nS0;2Q68YyPXUyerrY~FKkG!=c z0n@s5N;CGCaXUp`_XnnnZ0(7a1#;L*+&+qZghe9O{Mxeg?GN7@&<-&i4}$o4A(v1o zQ*e#xk$0{&VFS5Wfvt$>mn9M%gm;_h@oYSC`O~>$+^C<;*TNJ8dhtzJ%liAS%Z2V` zXG;B%qy5!ZMXPeg~78Y!h$U}@wmVjbknarHj@ZE2Mo4n3pQ0j55Pl)0A{as8T)1U0 z>g0%kFun%og(XcJ?waPgN!q~QZ{|7ze*_#h8j@l%WB1r_YK!@88a%|72d2k2+n5CS z0Y*k-3jexC*TXYEQB5fM`e6NoTOORL&^mfuw&xMq(xy3oZ8b&v0Ww$mrV?1q{4ih+-7s~u3kP|9*&2y^k7+l}1`@-W5)0D7& zY|#)eww=eVEg5wGg(kYwpbqDWK=Qj0R$;iWt@PefnI`h~PidMA7i-7$ct*8V*Gags7;iPo%NYkdT00n}~>|Wo48mnRj=0 z6w=8eqM{Ifg2a;+texAu?@7^FV)1APq+3Y>3{173V8?Kl`7BXVR$=jhUS(eFHaL_W znVZO!xB;}3>;0Qqvr7H%zB7A#92(c#_h&A{cnRKN*l%|`F0*LDN&&wia-P^@+*7Qq z725tTYmw!)DanVTy7+2G+#@IabT0ac1+WC3G8uYY*xMbD{%$2o&KTbp&n^A43zM>; z6x42Vi+yl1?Nkd{i;~V%k}p1#UHql}mV@F)ffnXCvq~vZ==V-Ee}|VV1m;fDlYOl! ztC7O*We)NMVFPuYGTqU#mksWtl#o02H?yj~$wuk3Wwr*j)XtWqJU@#^ke^3wogo1v z;dc2^@a^QW$Arpn*k5O@(&GmbVPi*;Tc~oq#r0G~&E@=1=tbfYG;6ZGmODw!ygr~4 zL+&s#Ms3Aq?z-*_Nsj!60*HBQW?qN5d%OM8?94R9mXrrBzW8xFcu|uFX!=6|=J`y2 z^8-*PAz%O*2S^cy+m=y*S`+tyrO&+BWq1biR;#{D6k?YSUVg%^Z1>f3`jtEa<_wQp z;3MXKp9-x^y7pj;WEt5&@iGi~}E7`#w>AiaPP!&HLW8`MrXO zwOQKT62Tcyi0CVlIYN)GR@gsVHuJ0e{t%*RrDkMam$YiJ<6kP9Cb?(Vt~3TOm_G>g z^$j%WGU?I!W98TQh;_-K0TtdiOnB@JM1UP~+bEY?HC`;Y(c+VGc^6gEFC!R}|VFbKz-)GBZokg}kgCPDtSZ|J#C){B}r z>|)~e@i?Q$G2>#|KE6u#^PCR5R{C)j3;}w`)Z~tnxXIK#jXY|iRG!moJPcW1J+`tx zOtv8CHq7;VA|$KyjTIMb1H&*xKluXx7XzdtAWs_Ps#x#n=e z=(WeQJ@){U1$;#zJR6H-0~d9=K_ink;(-?}a>GFwL8~OqTx=`y;Pbtnv|+hMb|;#Z zhRRpF2Q@>55Qb3&SzZ~3#MKT3*+^;hM=IBcURtpUl?KfBR;a^+?-(MKvyC|1bB3HT zP$PjMb`)gXe{L!_clS|5nm>9;@AF}PPdU?H6_KY0byy_-pZwgVY=;Wl zEq&*O?IGZgf|_BPx?;sd<}zD<(CtamsW&6bgBzxB2*kYV6{~0iI^HcY6#_f+_W_XnOKLqYF5H` z{E^rt?rWvmBibqWhMQm|%=s4QBnz2ukE>#OvyiekIYPKV4ck@X;*s(rF!avrn?mn) zTt3qXi_ZKN?Gk56MCTA$ntxsYDW*Gy&v3}sp0_4C#B0Z~`3u&UWXdcq$K63~yY^df zFSeP+vjZY!1jY~KpzN0G4zYh7t>N^cxuz+QO(krfvM}J^rx^~+R6Aza%NW;;3KhDS zy6}3_fF|oppp`}yUxa=rRT?8F;s+r0q)UdqFPi)Gr&h{TnBh0Uh8?4+xw&(OS3Rh! z{?8%sFYj$@@?-Z(Ijm^H-7j(n(7*wsWYd8~ru^s##2nF&W(+<;e_Bh39MP$GjQcA? zSp%E)(zLokdln?B4OVlHi_I=Ra($)cj)V9|MSb)e$5yv2T4F0hsp|G^ex ziJzXfShu(BFV85MT;Kh2?^$C!@8Yr*9oda|i5OD@1-Lq`RY*&!(t4_5D7njGY9tY3 zN&1`x5Qxj$h{upCr8rAt;}elrHue=lhp8HZm&DJkKT(DQf+eifBpjysHbIv}*<*!H zVj-iZ*Rzn))26Db7`k%i5v0Siy?;*x+sb#uTt~$LyBBZtH(A5UpIdo!rSdi>aMR|i zo;KJ@f4n8?uvUF}DNfr1A50ZmBPbJb?tw$Cwv+wfp_>fqNX2^`bzi=hHwc zgr8&~?8k#&mc~!@ZZ)xdrIJJ3Vx_$7+NVk6Dj5)B2x8*cETj4<7eD*x;MYyXGhdwK zTPX?K-^#WCUwZ@yrTIW1USD4QFRCaQ#zE)g7aABiA)Pn2;%i& z+*ZV_MupDl>_a}JnNb%>AIydV)TB|2UGCO#2W+iDbJj7cq+@&LC|BnftS_T7umGMR zZ>;}WUb_2t{Q>$2+FXWd9zyQr*n<)bkG4jwtX`kQTt=EuyiFbBuxxzF{CWwi9T7Sj zL`hdol&>*Uoit|}2+cpR0nXI^Ecbj<7FYl@sK3^o2KDQ}i?n#BE4p3Pt&CR({RO_?Y!)SAaXKRR`t1Yh!~*RPnxM z$mHt&K3ikYo-@mR8zEun=5m0@i8I6N>@A8gc0DLFXm%hkFk?@WPbLk$J{M6HR)lD8Kr>n)|+UPdH3sK$1i zgSivgqIcbHIa8}CC#K`xJxf$klCyVV9ZEvp7lSVW2`*44y*w~H;HcxV5fIpYqweBb zV$1Z{Eu~2?JgDj3Hyx>x__<8O`#09mewEkxl`)Yaei^2n*Qy5kg@P((M8{Q|Xx3#d z_Q~Ry3o6b4o!>E^@{&R}^J>~B6(=f(4*}8P=%Tq_s+oFum=JGO1DG(luou$Zdze@I z#KDk&+OG7TLLemqnlco?mZ>6hrnd2o>5Pl#ftwQik8-9K3!yjgUcRinPFUwr_107g z(EibS`ZVw>UtnFzmnB8DOf@7V&VQaXX#nB7{a482F3VWY>|cp-?AfTi6{4O_%5QbU zEG_X|`!|Sc9{|Bjc^Ut^SM(1XG{v~(*2J>egi|FScNPU!SxVehZ`h<+|5{L!S&igz zRyp_DNxQMViQcNj%C(j}S~-L;DPwjiEWzK5XuR$3Pan}LKCxXwCFYK`D3Kylw#^?T zm8LZ1rtRkfHET&fFaTJit_7dfPWf5xZ9=+J&e#t?wU z1(86Bve@!Ps^fi)5Q`wb5YL>UUJ(+U1MV5+nv&eg6>n*Q_wc1IIR%i|T#)gb7wiC# zMsbMpObrMCg_-X1P~e0>||b z+RNoCpV!immF90?X#qd<_2aIs<8=nSy- zAGaAt#oyhWR%kY0%0BQ4eqUoF+{kPojriS^t!x!oVL}zKWHX{QUv$J-J zLDu1oK+c^}sem>tfFQ&M5L=>^eW8d*>)jSn3@PL9FaDf{LQWx5ih_i5%YN$>a+?VE zk@z`l0?rgW1)ch;v`AALN4wEvpuG0$ddE}D6p;_&8y;giovd6Xl|+9Rr;f}lZLNaw z5FnS&ro}ng{`q39`+aZ=q83hy>exvtQKM!K`BYK-HH6W1fE*akm;p_PorA zaRAflnx`2Jg$w`_#oJ%VR9o6+V;dw6g_P|L-K!K!Sf?K!Y3Fc%P@NU{C%TDk%uwtX zz3IXlM9veI-+awTQpeVsuWh)mht`eKUS+!1T>j)5sR$!-kx3Taj@-dJV&?H(baoMjFAIXC?nNYK`D|E*R0-_?EQjUItJKLB%Tz}9`fR&mR= zKUjEDIp*E(g#)>f0B`ZBCyC*MBEGXo!tKy049cr|Y3-EC%X(=W=%1WM8IIm5+K!OG z3~wX_TnEYCK?Z00hkHyjEsp9>g@EK?cw%zbD@7ed30?4+DjWTQh27l)cC)=Z#2f!W zWT-MF0-jL}0dEDOm2OM86h*Bif8VWgE~O-O1#U`SY|sEQf3`zuC9Ud-40w#5ZLSk6 zb=;pkhM(p%E~EnL#)?e8{~qw)+p^{5XdyfkGW*}7rk0$NqOPPg4chJ5t+m{q)J|k^ zI^>p?N>L|$p`>hWTy3yf0S!5PFFvR?V2wRdNy2`t6D*FT8;^cel;k^WGr@!Sjee_- zEIJC=MW)d@^}UGzYn<7_XLJ&*LWCZljlVNW0>lXfzvMJo;pkhlInwEpEktS?PrRb; zKsyW_+EN!QZu_E^XQipR(c76$Q|zeTPvgx0l~LoU?YEc7vnkJP98^(51r}+~(<9hG zctH_G6=cfEn2rwDf0y_40jFzTIm72;3Ok~xrTb31Y;CopPHE$HI{=XC^uvfOJBLY=^!o z%?8F9tY1q+3E;Pka}U>!0$>i|B{7wf;A`Td&l*>nx1Z(S&m4;A;>=i*xkD8a71@af zUI!C~DpR!remz21NLcziHcmIQ$l(lI0Y$RDDy0&&T7NfesSx4lG9wR#=;A>9 zxzL?r{@7U~Z`JJodo6D!l>wXGcZ73HVWWzye#K9Rp4q z-})fhzg|u4$9+@=zpT~)lHRNm(;F zJM!Gzl_O=~pJ>I%!E-at`4GJoM<;nsTeWYP6J@K#SaC;36fGI8KK3Cu5tpZhUsS40 zn7b`2n#;nWlPaRRKDd}-a|kds`c$>5h_7Z@qQR$iH9e*^edS-@>AC_4H3xe5x(ygx zT~C**jYE32Eez`;GS!y@HyGU#_sRtcHK*Z~qv4q_*`VqC8$5lAcc? z9r|Fpx5W|+wiH=@r-{KRHF014AF?g?!LUWPICBF8F>F@ZDM%kz#_DlTO7@g=tYN`K zXAvgWQH2u0q2^(Hkz(QGQ*vNA|K!_}Fr*{!3P$jI@rlgj6wCTrn>s=82l85=&Y!%H z8^7?cfq>7SjTP^k+xn3^w_oH*V9WxxrUR!qs!Fc&kG>A*gnOe zpIO4}U3J$Ss&J9GCW+g;!k(Lo(L7|5F~BIoNjmh8>9&G2k5vke)EcJ(JVBf9`4(n81Ci zQ20JF*z!=F#g>)Umf`s(@yx4CD{$3TkoMO&o(H09NMz4h(a5Sk%5Sj9L?EL+@64!_ zhfVV=i9DC}2zD{|>)z9~wLn~!JN|hQvn$4~d>L>6V)bsere*R(Etw(_0w1j3@ zbWKu^Q0xz{XuCI>74w3&!yo$mt_A03wYBUSqrT{06Q)rT<5=__c`1E~95#Yy@ThCT z<OndELE7y&p-@L|9{?48vC%XzPzOv^O=CK2H0 zOTBezZh^aB_?5g-G(vqLajtzJ<=kc3D^nsTR{6SKPjcjY-?koo`|FI(^i<7erul$SByUTZ>r)sG^ z0)M=681j{cA^q?a@9({t?eB@50l&-P&)k>C0D*7#>)BZF{$n$?H=yC!zB*pGI{ER& z>(~yZRM-gTD>WX;-^b8Mf7~4de57*YSJR!P%Z`S~8x zgZ^j+f>|B3vJnqG*`PtA5l6E z-^aO*g3Y;Qd?N_cJ}?Q4v7Kyn{*P$e_(qQAdzJ#vwIlUl&j)~ZUR!L|EAC&0j^OOfMt)9MOVCQj#(`VY%p}W#{06=IwYCXCl!!b&uKdSiMoBaXz;DSMRr#+5uyU;d7sDoY zcdMV(PA=@V9;39g!puiaZ$l@0;2NgZfN4HZsovR`j>C}|?t~K^V$7HM2PEu}$*pxe zErz$#q8Rj)ZvV-Bwsj)~F`~ov0rYiFcc&StP{3XSO&J3*5eewt#AzQIgkGbV@6OyS zTOA&co6p7MlolC7By}4hnu(oDU5#*Fp`3$Tmpat`#(7O|)iZpJse+aaoSf3bk{<_ygD9fmaY+k9zvAX~*IrlmfX@8=Hge+K(|&y&gM#nJ5%A zD4;^L;rx)UdO+&$5fsGc)|n$7M4YjNOXV+|=HdM)3rI;4HOG)w6frGOM(^-faWykUG52!NYX?%&_$W3IJ1?(7@Nf{>JrX;NS^UP`J8 zdb{lkw(4I!MvV{wp;#~snJDl_TAMhR>d$BB926j!Q_w0-iA$vxUmWs4A7R*X8l)dv zp|B)(B*W&gw3Q?=jo*<_neG<5`J~n92+W$a7%+PEQp;Abm$a3bny~ivpwfs+>3)yV zKJpgXgW;~T#!nBoB`S&ROj|2Sr~c+z@RBD}LSG9oZ*l7wIY)Am8x zpwG|p?r?oUf(9_)_@OA}mTsXEXK6tu&MBq=8+Myc!FveUhEc+NS1u^T@JfONh`zI? z6KpbIyIr^&oB#TOh2FgOWBreo2)ab{Zz?qL!B`k;&o4mJTt;)tj}c^!CuK23?cQ6g zhA2PhtqZNsspurhGkFjyN`kHwbcao#>cZ_0Bl1k z=$5CJF-#2;6E=sfT#mAQ1f+_qrf>ARl@31JEc*f_4+&0997*E0lqr=iJb%W6Tuu|E z=jqqq6Lh)=-^wlrxe0yPOscT0)ht0-o+#EmOc)t1%s+scJ(bAZT~k z=?5M4GBPs42#OT!|7|~{*5%WeyP0Y;jeBYGk(eQ_5;x*uj+ZD8kiY)@XO}45soovU z;P!!4<$ch1_dH<5-$jJQTf`=!Xfy4Zz*qVv{ad$bMfTk(43A2sa5Q>wgLGNys)$9V zyN?(BaeOHXAc7%^OAp1@X4OA(euw{qVDomHErMFIAd7IM(j(FHB+C!M|z7 zOn|uwZ$y1Rah}zzkU~}P?6dILQrprR5vxc@DXr=B9VOxqxO~PzlhrCEZy~wBH{#pD zPgz+7{t=pR_Y&-w<^pLS?eps`o)cBix>5uZYjzGwFz=lfq{;Of(Wt#dmu|Dag;Se| zRDC3$DSg;B9eb}d7$|gqHhW|ub^G8&aAff;a8yjTUD%U@P^&yh0EDE?ehSJDNf9_@ z({W?E{a7S_q!A}^i3NlXc;{ODZP{@JB;ZSN5ANT18JzCcUwb&Yo!)?@qvO$A7{xp%eMStxNGke$^1UT$5@Y4)Qz1VJL*>!7$MQH@$`F4OehZPdV39uh9we;MExF;s2!mxG zj)KHrSI?6)rO|kJCoU_e7%F1PDOZCU+l`NLrgF*nqRK%O4C$Vd|C0oNV`^(n5!)SE z^9!8o@4E=%GYymH{0nHNqlw8@ds!LRPIhLtkU(5>!jE3;g{~-?T;yfvy!P(}sbau7;dpp!l z(D+A}WL+&1`1U(p1KtM!$O|IC5KaUXnE~QPHwbX;UUWJaGHusyJ;am8qyyR*=AY=U zUmZ&JNHQBvJfpi=a|BlbyC)qmX-%;>jo1EB5$|e!5DQuJa(5UOO5rrV3q7an`KAdEufyd~4c zf%#QlFFwP64%*VP6H^O;DcvGCB-F_|O1Gk(nsxq(s@5?GAg_Af>)tJ=g$8W8^NQJ| zI`6-N zeT1Us0sOMdL$v*kahY+4qs6VJaWM-T#8VgI0~ItwQhr`Dy#9slqnVBJ1OxvYe`V$5 z&g_e)tVYSxF^3qDA2rtNT=8RW=+!^-ET3hoUxRC;u1}z{i1g%-m{D;j=Z0?SwTIr9 z3RN+s7Npnw@lD<_lPgAu-`XVhf=4_?(0(Fi8-J*hSxxKqkH*xCTnP_lf9Tnd6uJiu zuKnGzc@loyZwoNSBlMm*^bCLi9jc3{=m7wz-PnH1>9#xsy5JQc{MLhNJ;>(1OzXrt z?xS$HBdoFQlZJoe=#on&oJv@nZ_;~G7qfEu)WYkmYBiwF6GukD(2mK z@DR(Cuj4an8a^U#M<@AsRmG5RJ2biHdAhq(b5z)`_XGCx5nR#v1VjFklvq2_*?Xyd zcU_eFo(v^A8|sGwDI)6Zll<|S-cfBEAD^JoUW}0m8}~+Kmf2H`NFNWD-AGgb)XV@~ z!~P`5z@WWi*HEyCX;2lKoM<_IF-w;4_8te(`e)D=U>gM*j(6T-lJSOKXNHVQE!{82 zwM9AR^x(HJm!n4_C_;b&0K)FKV=iZDE+EHvJHJ-4m~L`N1wyc%pJPHz$g4+mk6S_t ze%rF78X5RQ|Hs=HI*=P52>#XY^%nKi`lmnkl9?eDYDs@`_LadQ9DR_Wh$QI8-RGg-6?af-&%q z^3%wT`GYUX5d(9{PS;m&6I?}_D2;UEQ{oLJl0*)cFw1;U{`JwN-ls~JrW z%CNu8J#cK2_6#qPCoH(vF+hd|+p`%#Y8`_RNu&$Pr1 zt>FWn<;t|Frl(&}L~`Trsg&iV?=zHD2B@|3^6zd+iCRVm9(WMMk){U+4o$7;c--8i z=j24IcOQuE`-b0kSo z(EFmxb+Q8Ato2JI{RflDSRH4y|5Yg*VhUhiscR5l4!F-Ew%33?NpDL}(j~Qmy zI*_EwNY`O`2P!r$p>tIW0tA3oGNQz#>$z=$`+RE*j^6U{IjDcJ0g@_2?-{J0ljK&a zpjvCYF`h%4v9M7&){mdkW=!v;5MxfD=?-jbGi5M*sbZL}6-o?t63p+?l9MTG7FGE| zA9cVw7FfvX&oiqpqL`6QbR(eDnrcKIjS~{r>#5IH zf(r9Xlr(P5W*?X__L87x!2jTmw2=^)J^Hu1x3$9fNVl3H>p?6K z7C@2HBLI|Oz6F6bym_McC*_}Q7{=Sj5<>5a5>ZCQJV^D*i4saoW;-8HR^=rDDO33!LHLCigzxj^`}yDtJKiPF7J)m1DN5gEbop)yYh9F!mWqO4oD?(ur>BZl$|$2So;b<`f`@B@ zybwF{8E}V_-3Ai8A^2=H(bwZfYIyu~Yxhz=7~hgr&P_6y?s8bZ623Q)%+?sHq^)1# zWtfh@5!&g^9Lh7lCO}oZJm)3*iHOXn^5Yvs>iIb@yO*t=3XaZEGI$#RYng`dp-1FY zL#*F1E!$(P=C#)NMg-ZH(6rx9kebd_K%l?NhThK3_dBIYovV6X(Vz z;6FF5uBWV-iwbsscw=TCfP(>X0>cLB)+=ayzsW$VFO*K|iv>5Kh_L+*8XV@Xn2`T* zYx))kB<4D|&aOpV*?xoZm32iiD$32#$Hq?D$@_TMFo0mQkC5ver|EwFKWn>611`eU z0~;x7H?VQD#iC*#3{N?GeJ_<)X#GwxZy{1(H#6q+Ii(AC9> zL^D#Y#og1n7LF~L7%7EupwonuCA?0V=tVK7C$l$Z?R?*}Jo)Cs_SKI4-?vVDY|BKg zTvTlVqbB-ohtfo%KJt@^ z(a70nCjR5G#2>}Xj9wc#v(BlPf_u6osbJQMm<|6oqm@r$4xCZ%GK~c~7yEs@2v8HV z5tF}glCNjfd?x&z?lV`9_SBP^xgU zqditcE|=o5wKaS{uIl6@d|OVe4r4s4Yui_Iy6I1dmCYN5I)BXiM64KtUI{Hs}Qmhx>9-~*QDl|TkpQ-N^yuBvjP3$kg$d3J<-2Av`^yEU;uHH|1Y+QD!*)H(JyaJ_Z!G3A-&3kQk`z9-_I)o)ocwUy zft3xfoP;=iqI&fMuIfvf3|U}BcvC{8Q+{?#*|zz2|D_^Gpmr(ucWj;V(ph##K*`RL z>(D0w;iaVCVqhBZq24*Eilv@+TCWw)lV7U$XGBq6$B;)}+DYT}-SrmdCvw8q4xuW< zj9-a5#cFzV<{w`U_wm62EdRzgYkoDVHH8_e=1VbEZj;@$Mcv?hsDo)m|359juz)Wj zC|npAy?ZN|SYwdpg)!%U5?z|^GXRBi4xyR++dj6R`yqe{7Wva@EM6Rr^$#*~F<5%G zMfJ`Wh3%+Wf8bbU3RAYK3qOiCt*Cdm0y%gFJgFA`a%ya3+>geoHqZ#0$6|^tTb^kz zh7RS)M-g#j2$Sl^PBY3FkIctO@DY`|r7DOPfUj4RnDfcza!;$=#Wnl&f*_6re0G_Y z@J7c&Z(VuI1MwxK&!yyFZk!4}+}3f74;|48WwAtFCk9p7emoa-QGy{<-#^vjmlY%= z!9Pc!#iPkMylh$I^I&NuV?hFR$tVCxMNKTrbXboY_SGW1L39x)TNQ-vvUV4uxmFEe z&nU-1MGc7}r(h-ZZbGoXCZvqr+TaiTA}vPmx^^kaWi||Z0eP@F(|HnLA%fnhbWOb2 z1}1GdTnNA2SV%wLfd|%A7@^($tJaWPW!&@fS4#6_FJo(#OwIsg2fuvi4RnJ&fT|mV z5@-85a{&N``DnQjXuZPb!VckHRm}Iw?U8sa_ zS7vmf>ODYKBXeSZ+BaLeH28{7-ss72(LRc(Q{L7~knEknnutp z(`Jzy2?-_W3@6sOH;1nMpGnICMocSB&ep$V!Xl`&jTq&;SQWc`uN!)Q%pw!V&%%%k z>}9!E=o5nm|1ndNJxHOQE4qI-nw*CxobC~t^d`)|{)J{TGVw0YMylQD2Rr|RVSV{* zq=YqWJ9(>oPEXtH6hYFyiksub9p9kzF`T4LoX%EduXTgz0g|kNn18XBE`JLU8_#fY zXic~G{42R}`{*#O#LcT}A1|RGiAa8DvLbpU@{Hsc3-+a+ZZ>f)h-Po)Yo-T5JE&?9 zs}gCKF_*&6sLwHz%svGTuzrXG?QqpJ-=>Z8v0M6yevtBsN87|EfQg8cQF+Y!A;3kf3s^J$f4pH3-}SkT=k2$xq+2Ok$> zq`nf6zZM`ZN-D(o_iakDb7mp9NyjiS%X^3B-qZb5y`qB zVIA79f0&~s4I>IQKFlkZzotVJ^_zsGRYQ~wdU701_>w~rz zIY~N2XI3Y_?(u5R!uZ|GkW(;PP3o8E2~Lq*$_fRfMEsoze|aKZKZcho?Ax*{{eZZD7El0{T3g7%S^dE)9(LBJBp9F4`BzqTR5N#;@;p z@{-ytSca~bTzQu&Ey#hqf^5idbZ*xoQKCf9m`@=oBh%__?|D?Z=cAjtk4{o9q@OMK zGU?Uo5GTDq3Y&zQ3r}!VFYYM{U{HuVA(Sz@sxPogS|AX(F2# zt2_GpbY7U*=dY*sGTi8wYGqfRT#mwDlGqw>5WLFk;RbvWyd`M>xf_gm7(P137dhIA z7~Me``3Aa;i;i;PN0jb!!NZ|Yq5JDw9+p>Y))rE$V7JKZ8i-P9kU zmyz%En$9OB%xHjw{P36%CP>b(N00r5yZ!Qc73%;^9>lV^casA$M0Vx3;f^ezU-#pJ zC7AR{el9->nWAp8=dA7<#4xC{$O5f%s7`%SuyxuPG2`oPP6&Olph@A4dbD zEHA-`Tgj`Ku8qkg~4#(H~=yhN~ z#R9WiqJltxYkrM5YBv!BHt^I$0M@N+L1^B_I4z@{4H_B&d$$Aly91PhgN`lgv?R`P zXKFR0q3g?Ai??a9aL2Q8S_tyrD%tt66b&=674A>jP^fEF7-)oH0q>$01i~-I%=+`D zg*Sv{2&dc|ezd#kXeD`x3Kms(La%8DC|OD(2>#AC+( zck-AbHSHIMf>8N;N>VIZ0tJBbO1pGglf86B@aP{Y5aI^^8i{q}! z8t({AB&$u%bo9w_8!x2Z7xhT1vEje4pt)`d!Ug{aEjYWdVn4kT94#h_&U5B%eb}4_ zlUY8!uJ?>|0Q-MGu3U%2Cd3Lmd?(Wj!Pi~U8YSyNnM1N2esYiif&~_7`8Qp?CGvuS z@4w|}LG($N@fByr&N+!e<1R9;8w*B@zTXn5FbPN|x_(q)H2|`|m8l;641v*B6B5F~ z64r$I!9J2f#$pcbmy1MM$D*^e>#cJn@vbW2LBYqLdZKMb{N{8toEXqJ-XsswYO95; zLcerq_uxGiJAw9+yIFTAM~S^&W#>wUf;)qLek8;mA#XJmug1|@r8lH%e2ky+;pw5q z0^F?mDm}(J3=3efl!RwDq9BVfE=zL=S_k1 zjL*=cBKtANmCdkG5CCUXcEn$4;-6puswyT|Z5pbhmkD=BVUa!!L>XZP-7b2dD3OtJ z7)V6mj_&D`abNRX#&kJ)aDk`O4558pZ?d7D5)-=-+wX2RD1%BxrA_0IGx>+Zr@A!K zw?l-c4ZC{*EP%}3&xQ{OAgj`dBZ;DE^RNA8*g9faLEWk2v8nyaXBxJo#2d|=0HM!W zKv=#{GkZRp?p3u~Ac-l-?8`#@ZvqOGq@pcAKAqa$nein$JXSCNx0Vi+UHGBk)*ULY z&yZBEzLV(|8>V|2O27^m>RMDt@^Dx|TNs6udSZ^)8;srj5-acL`(l#Rm5SPe&_0uH+{ByEv@;k(E`hb+QG841MN`JIE z80o3uh*juF4z1t64lw#;MCzbz@ZpM_hi*{k&>Mh|{qS|>cRtxdXO=w2BUitIZnn=; zS!4TQhA^r{sIT51kW;&0nuB>r#^P;#cG=-INXAlr{11XzuA|+qrqa;X1=n1zSy#i zie8Q`8J-dx9*#UBBKOzR`)=xY;S0OA$ClMwpTScUfu3!T3DN4?gLgmwmlL$`L;oa@ z>rH~(2-B$1y1iK7J!=pF2Qzi> zk;m3(63An-MsM{qBg{4)PfmF1>?AVx<$GPJ3-UgTz)RLA{)~67giJU(Czhl<|5U3` zxka!5T{c%q5p#jAtOzb-!dJ!6>7w-Cut5KYr<1biA|?MENPB^!YtK-#aB4%SghYp= z!_4LZMZSNR78)4tCqYgOgsGN~S(Y!`YSQ|S9kktnouW6WO2h@?z#eALqF)&!!cR4U z6XKu9LgVpXQTq7i@sJ6rc4n)3j+xn+NXP1S(Q&(JyJVJZ_wCp~;{i9vR6_tGnevT^ z?oFgx%YsowJViN-X1^T^VD=?KZ01Q_p`qQJa#~?W!w)MsiMCR zrHmEcZ%IT0=gJcC*!+z2dqxHd@OBkJYew`|)D<@g=c7BH6+7mXp_Vy3!Q$ z(M2jIh&>SeVH*f<@)>1_UsvNx)af_Qobo8t`4Oqt z@mnvW!@JcdajQHNG5`i9*6wGmEjzsBiks%q(+%sq-mdCUK-_39AukW6V6m-B< zUtztc0-9acNh(i`@c|qTND|jX(66@N8F-ZV;zaS_w*4LG% zGRoaKcUyMD&rNUTV%WE@ZWxPqzI{!xvw_=Lg?`tt@$RQUh= zxU0=#a_e4vQEKgt47f3~pe40yla6+FkG6J?renM3PZEdDEAueqO$rjCso03$YI}m# z#4HHSbmYN3HCexPF?a(aee@6kTyhCN&#*vSNv)bW1X?Fd9pGrQaHsi(7sY^YW@=b| zv5GTDKgx5UJ&QYiuJW<>eS@6`O-~ zOQerUCV&MX^kyp$pKAi^d64P|w}EonDH4m|y5wf!L&MB;z~bc?3G zyu09n0ejhSf2t|X5$LtZ)i;*{_HRZyXV86CU4D`^UqGKxlS849intJRPon_EhJ1u( z(%~~WfXoke0QZ)hMU{Xr1ewwv3JA*;U(}gC2}mG@`|Gl6t86CLb{C!3P(-HxEL1Qk zc<{oXjwCr5)aS-Cexukn@d1t=)k$$yIyF^wdJ$1m5K9Rwn$LJ0GT)LoV>#ITyV)&N z;OG89XtM89E+uvoL;Yee>ZzBwX7sbxbF;+FY8CEouP|O=Xhbrs=-u0?4Guce(@kSZH9}? zOAAxWEZ{y``%N|sW~ z^iCJT=qKMr`>h!-YtJMx5x*s}6pVuDCehD-qyD=|Mdhy;!(l&j*+o%3=lS7m8*{(>Q- z*0!aKv|DjaHYgMc{KjH?RLR1@B?BaN^r_s-$ zv5DEgWjW}Tj04L*Bi%ukP}Qm$=50Dx{K~q4*&$QLDG|~tg>r|wY+vQ{QfgW16KD?a zE5{dt0)&QEI1wh;)%bCh$+a#DkqEf-N_uFy$)?hksf`(28it}}C1}8l>Z+~E*45I2 zzRmG+V(z=dJ~MI8EdivT6c%A1^E|R1AHZ9`*I1@LA3CFEzzGUK5doQ;zeyz~pXKBI z!1c?y_NBxPzGNL^WY|}b3C2TbZzEXRVr_Oy*~_IX)%evu(36Jki1{eo2o`_=vQD-f z#`;$8y<9XQAAQG_!X8&t1iq2iNQpPWeT1jS+O`)StW*R&WBI-SN;13C62}yt#ET%Nb$WVW1c#_N`|szx|F5#72>h=`{|=`U$b_5>n5LHJ+yAVKE`aX znN_bfKZ6~Qb&o-;xg5Z_{OcU)X@z2QRx@~NnwxKf+Z}=*UQIdn_pg;Aa`O%eJoi=a z(z?0|A)*Ym=)L-scG8iNRPx#&jIQ@imEFcIbee>ev*u`g-PlQ6|57ohj;7CXF)!!X z(vUT{U=@qS!mwsB66kb)A+OjV;qyY0e5NbFaVhiq9Wt0OD*`6aBkmnB4@_M2{mSZm zWh?` z3L2fDcKeGpuXpD6fmORm()_m!KY!JG+TXjGM20kDBJ-{@pBX8MPiaZs();CwLv}b~ zABjABqK+`b8h(>rX4*ag@D)dkS4Gp9PEa6Rikf6Hjn)7)OMANUlL<#FsE0Xh;?-|J%=!k{t|v5%S88_HPWcH#p=yxpK$j`b(LIS!tEpU)U?%Pu&Jb=X?$GXWr9LE=)F0?_oEQ8Q}f*z-!Jsv z?PnEDxb^k54+k=t(lH#$zpgrGoN*EBIl)(@t8FqI!jhl+W|Ac_z!Gd%(ZdBMEqXGi zPhTkF-4{RcT4-UjfbVD0(lJfy-cgnxx%);aM|#LQuTH8Yx0k(dwM*twJ$cm^@#hHr ze1MtJ9^h>Fhnc4DddU*Y(lu=~s-Q+dsdJISlkK_lTf8Nuw}1W&0fa=IC-zbd7^jq$ z-GG3Ed+d$9)%|NDGX=#ZYky%?kXzdzkOrM|KG#n(_~C zxOmDvGOXh&YcjMj?Y8zvaaJjbIf{E_9E;r(}wr1eFBx7au$oA z@68w;7@+HZ9JE^MPlzBoRH0HDT?h;yguJ}xQxIxNp8z`ZQZLVcb#^soM!vxc%miM3 zlbeWx<&hzjO6lD~k9FK6Fl275<};&DV7Ux}_(kysAts-#16$R<* z08TClSQT#255^&sgegzNwJ(0|aeN+&gD}Y_^;d*dQ|4kTbuKl-D+U36VIOcySU|-6 znpD>%lMx-3Ls5dj(X)czLO)Uuk2uvA#KutU_k_9xM;JS3zt+sg&+m1;)4!>|ouDj~Cg6Ti)ezbxD1#X( zp7A{MTR2YHpru>i`|tLhPcEtunK%HQ&4eBtc&uE7Hc8Brkr#Kjf0-}L`_m@R6-_d` zfCxUPaQY!qaD1c0S;=U%+CXap-?PH%Vmn!zX7{FKF38$(a9P_vXGGJt{Ol=QX_>_V z=5J7%f$x*eh72ZNQIX7ydHkPZK14acLMg$Y@AdcJDCrF6v^T4kE_0ktw1>_4q7u*< z-PuI9r{*DG(Ge#XO|lIvzc}%IJyml4mjkUm3nMku%nZb7gaQG=rh

5q~>*iM4}f z5`O-AuHNiNq9w&*NZ&5?F)u-B>3R{|s9nka_XK(lA5Unj=iw!GT_Jh7F(I~Mr*QMX zG2uRIn?$h*$geQ6yiN_?Fy8g05)-hBDUQkBRGg9c`B6L*?p`m1KB0HDLC%lxDPI^M z1iHf~DZ$ALJZXd8%~X3LvHkQpGr&3A$g)zYcx*nN;!U6v*XZ~3*Sl~8r6(VItW)2_ z6wwR1)NbHiO#M2ZX}!;Ea({E9JbF(?0|r8iv5Gk%Aq`E@Vr9!#yMl6y&$mx{YGmvj zQQyl?qm}#PPapYMN70`w46iDF(LLc-iFMzrl?bf3yQV1dHqRZO<-}$PU`QL#@hf1& zw*dgjK5@=ZQd5h$P8{G^653>*yZD5FO6TJCpsLz$Ab-#(rks5wZH%~9+rtW!bM%Wz z$lvRq-p4$p_-mmUi$Q_E$WERZ0q97TO@Y8l1>*>8I*JisGIizTxEu1DHnwb8vjzZ! zl&D?i z{Pnpu^14*8bL7MnRF^oS09HN)%B}yemH3{0C}~$JV^F_Mm1et-_%ke5CFOOqVKcS0aE4DE9k~MWy_;DdQ09XmoRDgXo zN~X`kYlfL4M&OHo7h}Ju!i689;qO;=0ll62@z(UcxNZ%Okw-W-Jv|+ibkpD84<-W2 zs!~)O53HRS8XC&^d^-e1j(tpIZ7x_xs(U z(e^kNXjz=-7E~!I@^}OTbd6RL8fY&%c8$yn=%{fKP*vQ+!DPYEsYjMVg#%b zjm{d|QieK*-Rs-%__JWN#&-M6gY%qX1PZT9_N z{OBq{yNwr$CklWribzO0K2L4P5~cU$5Z>GufS<9}?4;IHK9fOmSA+ALWMV-p%=; zF7KLWF%JJP9u_5hO?_(F9XTb7-)p+AD0fQ6ln;vua^@*vAjD(~~ z+c|smyr6Giso%s*OMBEreVQAX;BKYYc=6&rzwp#RfZdpqf;B#&H-kxxqk*8PlKqr@ zd=Z|x+R*++ft7L!{^DZ4SgZccUrCWpI1&U`#tqu`;uR0t8;yq)>7ZG1IqBOI0*V(YjwFPPm6>GPzxoYFT)VAwd=9KO(Y}UA9DDtPdClITAz8$GD2P^c9ulOH;Py9 z$o9IayAr7WFy)tSe>GbAg_c&_>)xnf`&e9`kOoot=J1KUscv##Q|lY4<8DF;|3-)W ziwYLddr=S3P&}b2s!&MoqtoPC}mU_!pxoQC*Ib_cdTS~@V%bO z!3#jx3rS`No7%glQt>`K$i- zH964TLye4h9X0&|nGw>Y)9cm58L>JDNa3n?2UNfCF)CF|sZ~bOAsEJkD?YS_jk<^E zd_^D-2|B|XEIrxAm&!M(E5j;)DL>N&vG zG_XK=9`L0=L1Uq*TM^9YFWYl9T<>P4jOAABPn94u{f7D)@@p8dAj#}$^nBuhW(p;b80IcD&q{Pxt-;; zDcC5o*~84w`QfQ5LKh9QNa<4Qsbpli5b0XL>Z+wOePnY8~Xc}z- z=UGSk$%KW>H5wy1zFJQHkD~PEHY%NsoGgg<_OJFr!+!aU$DZ!GwHf{q|8+o-`6mhN zflxJ~4X-`u>b;LNu1g^WM_znfL6+>yn z0liFjXu?z6=2dK(vL^gAOAEm(nvry(p^|I!-vQ(^L>{N;mkm=@Es=e_DWHeSk3d;h zI0-uY)5&hIJA+DkI};taA__TyPEh=BHA1CZd&^R%*QWMt%H+HBdD8)yf`YOXf^cvZeU zWElhB&WBKWc~YE9A^&F=fFMPy>;r~R#$1YxpQqGtV0G;{229w+8&|EwS&ZO2RV=^!SpZPZ_ed8t<3a za7J{O^g^GBr9JwHci!NLrtyd`oFpG1&AjR{86v-O>w7yr%UPfRh;_7e z_`}^0t&)9;#JE?zLG{DcmE(WrTd+Fc-onspg3o(5jI`C3tmB3H!zB%rgvpLTpCikc z$Ny-S&*N$f6d?TJOmpIoBK~a^x!)233Cb!`Py6r}cu#K@FvS(_=d)A1#zA&}E#-Cn zg6E9gB8V(2l#|l&bh$-(<|9jffto(>>_;ESu`#pR2Z7Mn<Bc^vNg5>)b-0EMnM>?BjQ`%m|5^_ph1G!(oow(=~t-$d~7xTdbSX*Q4 zYHVOV98pl#A-usQw~+9ei2+_I%@ah=DZ%qCI&!Xoo5J(rrJlSjf-DsFAEZX&TVMgx z?h;Ct;N3R_79fZOY~p+V3G-0Xch>&S0SLOE>j(rxJvgD++-2R-$fi*1hx5+behmLZ zvW&!P7d2O*jBX6HJYg(C-vg;y3Q&na%U$48dRaXqiPbf^>%c~-PKtZ9*sM^%L{3lE zGH!L?En=V5Fa}F+fYk3t0VGFULw`T6^d%~uw-frgTnuuWoDPo`73+v88n}diJzNp= z_cCGq*JdfSINjO&;gNK!;@?i7mxD8f_-eO{X0xK9KPTE$>Jqb-1v}9?_u+J z$I>%QUl6?ag`e46D4od`3u6Nss0n^wEM~l#)zRJNz~AxF`vML_1qXNkuBGQ*o#jYz zU=&K`^&G5yd7wO4#5~0edNHczd0I~hgV151-JJ3{kJ}@p!nS~cd24Aj0r4jOXkr-`d^=sHox~RhvmU{5N z>6UU9A-dyq35@WksDJ4UJ(BY8QpEyM{$9Ivu-^dEP`i&5B!Q&hd%S-A$HX!#w5gH+ zwW8W~g4}Q7hal38ztUdnqp8ZSHGn@dV4=J>l`$g!z2@o)z0?!F>CPgX7Ucn}m;>Sz ziv;twaZ3PE<9D_?A1OH^aaUbX&HZ5IDu^8R9+^v;oP-Vl6&~%4KhpS#S-$u8Nj)de zn2)%i<_O2%-@jlwj4?xj+TZ{4xJ@aOH*fY>>#qT5t2rkx@6$nsedqN6!KU}q1orT4kBRSM^ZF>hsLK;-fpFHqzz zL*2?Lw>PIzeF9eeClO71R~XAZ)A!!39|qa^y(<@z%&b}h5^8|wx!`mXyc_%i8>9g# z#{!G;`SSUq&olfi8v#D_JoNkOElFM6z2gmTx8T<7LP7|vNpDrA%n!adOCxz%kWlR0 zt~)-wE^RB_;8}tk=tG~f1S=~mDDH4=hh;F44n{w$BuBrttixLMv!^MEXsE=6dsK#yb$hTr2(FV`bG9Z@+RcWT-{?sIaB5 z7Io#2jUN_{0^MQ-&Qc($YI~MGNf=qEwtB#-8l7dV9RID>mv2(Td~fS*?;CjDYekVN zCA`&9DPqT6Fq61bDR`fjOGH(Wzh0ZrxIyG8&AfJ|Gal%|*j zSGstH-I4Ei$Hhk00qeGHbfPW{0K4vhY=+T8Ho~A4(;!ab>8}upgjLgHkR}LIR@X11 zNNG^Ai6*f>h|1s!%tYX&KKvWuy-OgHevvvJI(leMj=+6V(eiv$nj1LsyfdGc8#V_mF3R+ z=E-8q;(r*jlAZR<59-b9Jnx(Yf{V06!MgFqDag?kW97SIFC*Gg?$a;dT5!jlqsuJ1u4D?#dC$@y-a5;p=4_aa|a>l;xvk3XZYCMA>LS zC&akM=6~3LoYa=676v~gl1gdKrf*=YNAq2HVrvuP`RF15ky{$QL(PjMBEpA#-deXR zw1*W>w>+)vG$gXVP~K#9+tqld=e-mw=9Yz;Ff}Kn7Q?&`e(vC-q@&p$)i?rQ5qqUZ*ftz9-4<-IL&tbO!52U}L zQIBEamjy}vRnFH70no!-dbXGRtLJ#W54%w~ug)~=6+@Um3r}af_KZ#&LIgh{fSQ_t z<5QHtBkA4%9iV`XRA8uy1)^FBDMY~{aANRMsOH=xg*KpwDeK+D12Rq@-t>uJhM^r zir5by3>C^TRC&^Cyv2R($cVilY}Ub-U>iUtL^U2tOz0W;w;{M?FFH@b*eo;3 zcnAUP=;WrmlC_IX6&ww0vtqC0IhYz6`dn0ifqhY6>^{u-@Sm*yY%uTaN2hgy{HG)| zU=lWrQUJ4z_3COv?;&Cn(0-CFJGMZ1uR^%t429XNV~iHgI`Ov`z8Aox0?IPDXoHCyA-3F9vW0IH@o5=?$RuyNAcBNnXW93oh!@ z0dWrE1CDZ_CTX9E=%CI*qH(rD#4LrsO`xml{h2r2Hz{Cm5d;O(^EY`iDnYR7-Id50 ziRc^?dj^b>oc74(6eEG>9{*1m>0((KhY2PRwV#7}xLKvJk4;ZXlmbhFUWO^`uQ2vK z{b(Aa66EivwQ4+?104w2U@2e&=Go#ZB`2Qo8XOI7e+R}p@C2wf8Lt%1ZW zw;2DE!w?SW(dkl7$+UrO_5VXi#iMOe>~8zO0!l3@M~>mpN?$T!tLxp1h&Hnd2vc%E zWoP8j?ymqpAC1+?dBrTHWCcHZIIo2|Dpd8N#NMinc289p`Th1FE^f+A?)8MlI|M=_Td4%xg6xY_%uz&CtlYg0R-7MGz-3;`Z+nnx-rkgq%$4I=XnEfvw7o-sd(g*CQ z8Mf`@ATvWIICUxe``H#D!t$M@gtNP6aMg1VXP_%K-t3g1{Nrz-CkmN4;Z@aj+Maz@V(L3 zsw|NHxy72^Tio*TI@!EcrtdLG-|OY125%?Aoj(>1T&?ss?)fJ zWZHde5W18i9<{Z+gl1S?LdBpx6ed6>9$1M5HBFAsp6uu(k8LR|Vxm5?BrNFgH^(Zu zpeS#K$er>0FjyzFJhJSx5q7_&Wg4VM2c^Xk{xOt~#P^||+8T;!B*`D&o%odKKnuxg zpwiE}HuP#t(zvAVmSdr-;)XO!WFyW}^=H8VwQ@!W6<0IpsN%l9Or}38_nooo(7^9J z473k=LmX8J2d7N6mu{ylU5RLhHvZN*AF@S;@T7JrF(A$x-oTgXU6RR5O=~KSQO~Fl zk`bW95P}B41L#7k;N|b2n_d0p7N_43&+XmJqBFvk_LSl8d4r7%-FbUMyf8*S0 zR!ZaiIWm>v=pn5_KEbHjSnYa+>1dhJ6RSX-)?X?F-_*rlsMjX!3+1*33*b16?emNb z@z^wnAb+hkw&z3vKy^k-vI_?w%MMH&56x9i=Aqtw=*zgvA+#b#55jF)Pq1YYzF-Ah6xST3XKb-SBee4|Dn2Y@J zY$m1QKr5Ja4M;3iAsLiu=SnnZ5i{Nf4azj6u zfq^PcnbL6X&A1zPJx+*4?=&#Gx(v@8S{gh%O&XL58gGXn5!BK7k>T z$A-&h=SpE&eY0~EWD%~Kr6l^$Xl12|WklD&T2RC}b)YT`v6)N#>3&GZF>JJ_WYXez z%W4uPM&GctM$-z#=Pc&EzmZ_r)MI4$&5x)hg)H|48@#huMM)N$KhlmaVROVB==Ohv zqwgy3{-6cTe$;h+?i`II??S`z0-}ammGiGtb3yPW7@5ef+DKV*4x}4FV>YxW+x|608lXXW4@;D z|6W})$RAZ(EXozP&9Xv{JMKXBCt7mq?QzCvtWOI|1-G>Yd}aai>`K@3?vKq;{%G5W zA1To#-~qYY2fYIB2eb0|lbFjEO2?$@*tv4i3>ec=k2cOm;eJ+Gf5K$?25n-MJ~5e= z%dNF?;hJcN&E-3NE51A41b`bJrlHPIGaxiE;w=0>Ex@)#)|W(CVHc0(YZ)sJ=8?!U z`*v9pgui*VX6N5qvYguV%+1y;3^q@0=l8?{Ml;~UZE18V9qweq2oF(^W#9w2N2gOw z$U;(W)qOqmB3|3gc-UvC1@Q9@{4QmNef{Vd?P!&sNL#-W-?bHcL;G6M=T z<8&A<=iA$;MD|63mH*ZJOtp=_^AP_{n58l{x)StB4qgsV;_TK$?CinY`=$6zR0j93 zE@nH#KCT;C;>jJfghfisx@>!YTRxLg|2NABU3Pc&*$9X^0{uS6l8u=4IDox`MVxKPf zY?$zEbywYAq0&+)HtbMg4bp_vIURK`o4_8d)C^;;dUlZR7Requ0K<~O$SU`%OFqyq z;SWpkTs$NWj_81Ec*5dqtLo5Z?}p%eE0Z{g>hP}a@fW$b2U+|^aQNOm%*!=3wO8m@ zUjz7_TQNtC&ZtadB~K(sbHK6j^)E-FqeQxhV0?nuf!oho&WEdiZtx5P*0wYpU~pl- zbvFgOUCdecRA6_E#V&sdcbksl9#fc5uwJVS{+&1X4_>*2=I&Va0q;N< z-Q5&iHB9y}AH%zf|L3&i88|5_K8Q?bslHuS%{%=i$KJ{Bu|n!$*ZV`3s=jhqGugvx zT55v2gVkaljlGq6ddsh}<$3V`Vj>yio7MMQ6Dh@@&^kh<>RQ@{>UgykTb(|h|8UXY zvzF(_I4YN5OgziC4kmMKk|IpZkptoZFv_m<(E6eQ==qC2Cz29FsR)gRzkPd%vJGZ) zL9cWFQxKajCH135E^CA+QOy4a0=7Ijt!i}MmJ59EkG9NfWxAu8xRbax*J|lFwsd8U zI|h+m`4g(y>E2@cpOvMvKSx7|qBX46`A12USFCZmHkKvEHx7;xhNfGdkOLiFYyIy^ zquTcptfU_WS6khOXD3>!rdq=Q%C@imS?NC_4;^XX1!$%@0Ty_8zq{Rgl7J||MaUOB z!B#KjmX-fd;blN?%a_7WNHkHl>aXWFw`#9kI8?fs)$D#foH&q}6^pyeBb}?YSnVf_OflPRm-+*FMG>gwzXW#Ubd}e+wOP2 zzsK)$9R2?s)pMQK`QpW4*3E$C5iL0V{M%CZ2i(|xb>R}a9~;}GS#tsRp$Fi8_268D zlVv4txP+sx<)1+ouATE;Qx7Z zwr&Cj1W;1Q8*I>el{ZV>frJXiRDLd^55j+q#sFkZ7tC`(@sP_`ky2pFk-_a29fQ`6yGigDlZ_F+avZ{9(K|u>5_Z3@|XXbOz_i?jIIc`v^J0 zyS}Rrz#_lr()jnS<0+nQHKk%qmxw&1-gpT_1n&YZ&Nu2imR=F(fDF*wKmSt+4&ICqz+_T4d6NszJFNQ{C$+1MIXGmk=QTSx%N8Xv2(}(NW z)K47v@jgf+?@|IEOguper%=u2o;htEQ|6e=wR!SwG^s|fs(8@PNFvB?(l>GGc~fkx z{QI4*Ny+2To+gWV=gZBlTTunCU|XuJ57!zKFPBij|Lr`67}BB(^l<55Gl#|*dJN;; zG9>+?r=zthMNsLaJ-gV8RYeQi=BAo>9Ns2CO3!KvE-$uGmJ0r*yA6C{^Q_3 zs|Rai8IP@N106|t-WnkVCVAKh7k-7fZH~ z@Y+MTNt-)#ODtqp0K$lTwLU@4db@r%9qn8U)>IWAfEJikOzaw*JTR|?AP!*%>F=qI zl^=)u5fHDz=vwtC9#sbeOLRi^>Xi9~ISy{kPmm@M{!RG6MvN)(_$X0NjlS=Nu>zvj#_nTsa%1v6SK024h*>g&@lIM!vdSB6lF*aVr0WvAO zWUur#8eW$aUhAE~Sc}JRxUVerZpGHuIeo|`5_Go{%FgxPLzytNqV`0>_eE2`aF2ZM z_cdf!Y`=a8(j&YZc(ZB`4qtTs5p1gU!=BOydlDj*UyNeQCxg}bUv*V`n|0`rqm+x0 z)jf&wR|OK;A+jGo;1WHNTv8-OI9qq-Y4UwMk4%xSQ0zuOl8D3oFiX(}NCXvSt>Qmf z%0;5?z7rA-!1*6#eT~7_iJ~CRq#PAvTV{5CeQp;2sBaPbjW(zN?)MUU2j@T_(^Urt zOLOJ51zHUudgxW^F5HG?T}ZNiei^Z)n7%m+csDzc`&SR`Yq&9?PkrZ%(i=bF^;VC* z9r}W@=pGuh~{Jga%Ez_jas5$_S6If9Uc(#rb+6CV^i}NgL zpB>b#Yjj!k-U<(OU>4dNDo(?m@U-ZW8_f8Occ_iRbE3mA_xQOsDf@N0lAHXuR9Qam z{9cKteW!zRLb}ssX473eU?6y1`8e7NlGA^!T{n!Biijh=Q;nb}>yLJ2P48q@gxytH z5qhpb>Dzw%>%#Z??30*t^GsDCj1o7F7)XDy=vZ;}=zfn#&t>=FO%!o& zMbxG|xYoPe{%B`7f&iU%#%?NKvV32qsox77roJiP=$$f65YdTAm+tpFZqT&>r5a4x z9O!OMHSjhfxgD8lJGZH}Ra~N~h!S`G^~Zd?K?A=5D#4*~z;>9A?$$3a_oL$_Q!yAp-J53>c9 zDT2_VW;f+kG$^_wDt^2Mc?BepeYe1n{+mPlPZ3nb(~{g}Fn?70f496E?k7iR)&@4b z`_`{a(-A7Z0vmX|pz^FT-l#^{Td>V|`%wHqOB!j$M;5l!)Z`GmF)4AaU+QP}JiJ11 z?jlLKvUzc@Sm3tj$&Fv>Rd_XQ)$#{x`nK-)Z#Hy;8*QeI@c{i<)Qb@0fY?Os zhouf*-}n`JNGLWegXhcGvBxG$F(E?B>7{UV2Qr&*`zIOli6Q*K0|SuUdtMn&+<#ja z12&Em95bdJP^hG&KRNKye1JwJA>3PD&V+qZ3tqm_pfk$1`wt$0gij&h~4^7%g{ z*}`KETu*K3f4EoOFgEjbX&}umc}JF=6u_&wTvRh63bm7bC@~6OCnOS1>SX{M_4aFf zP@EZK)~&^B5t#f4yahhfd6zBA;mTUK##7rKcR zBT(L3PPA!5^ZDi_ad4L0@%vNf5qQL$9T+O%w$PDVFN`^9|EEpio(vm>c5ZAK?LBCB zeDEI^Jgh9Q_YxXf0_Su5FRtQi+26kq;HXg^ZQ38v1o1mIWMa=h`il1;A5X`H5V%(U zI%1?6A0EO*mp2eAnX0VZXb*5VGe!yZXZ#2Yv~EG|Z~XhBFZri_ZW)6Z=&Gv~#a2jW z%*p20fz{pz{HKV{(p{$my@WbW+-E~RnV-jHk+OI7PVfmGjtaXKM&omJ0-&U16NrM9 zb*V+IQ}?3k?-P`q5POeRj%M%={RmdGfN_>9kW>1MiAfr^~pTYeS|gpZ0M>om$yi zOGQL|8gIImkyXkDrYr^qx(Ws+G{oioz{=o^)_%jEOsp88qwSE1%n8^BexZql>pmkC zU0m&_^+=?iKx!aojI?!WC*cRnW+KABQEw4qNtu&ht`c7vjLcC+Y=1usj>Y(V#^ZMi zrCVwG+>q+@f@vswnXc^l89=4!MBiGshW`Md0QN9RB)!?O0!Pb&knUZaY{l*|jVGGS zvlgmvs_IxJY|vRoeq!M%dsr-)~c3lWaRCBO`|qv94lnIpjqw)wDX32S9Il+&1F13OG{ zti=&vt%pCUZ`!L*Kp)@g8~zaI|?G+ z91k70iXdm_b*z`>AFjL|w*DbVM;=v=y$j1Jx%=!Q<;2bEbaJtA{E{`9|Ft{M6Khkv@{DoF$6@Aa#_WpIb4;*V( z`ybS~j|_tl{+JgHs^;%a<~9E=Lg2k+e)(w55>y$!(E8;iNhlh4x%3`Hci@j&Gq}4@ zH{1Hd0+4Q>0g7-;V1oe%8qi&GXJB_*--R*zRt4#nikoMTfv8NXU=gu>)D-#A-|){} ziMEUW;bgodR|Nudp?1&szj0ca6z7Tizi({k6HRI4EEj#VmpW_(8-TxUMs;f4h6HR?Mo9d8&8mj<(^!3xvS1z)60BtnJl&QR~ z&(s%Wvev=^G?NS8XQSR~#uEoDw^wPy|0<>(LRt>zOjnJ>$*qL#uZ#6K9IWT)@h0j9kVGft=>RP*tEDnW-FhJJY%T zx$j3+1KTYVdwZ#g)n0+OT)(TE>pE@q535ob8EutEqwovW(lc+oZig#b(9(re#bsGd zjkd+i$6Dj%b+U^Y&Xv~CoQDiSDAJ1%L`V95u$kN*Ig(o*NKWk2Y-tHMmKddwQA>E$ zR!zPvtS33oQ31g^SCI^5EnjJ~_OwRa>}5xm_Bi!j+IEN;iE&*IX7=>seeY3__@d#< z`-0m#uyR$+2hwoj!!RE9lFyAI;~_d&A=Yp`pFk6K5?ZrgmDt$3Eau`w>Wy)O1T@;^xiaJ6BP`ni%yUw4Ussg<}mB}()&B#JjgP) zS*GF8;{GN7=3Mx_ccut|8z>211b|Q3gu5;(i2l-m`2mi7;qTuc$|`IBxD-P10~;Yv zpI(y~-mHk#Z0A^YVDu@^Qm(Bmp1zT*TE|%~+-5_#LwDsL#y#|7-1nyH4_1DJ$SHb| z9k^Rr+P=*)CgN?4!OB+3y&;1VM*aLAoEBhLuv=POV@3IM_DtFep?49d)V|zQe+gNn zrV0W0U})RqejmRlcDzz0dEgDUEsOX_Ev7-KtC9C0E0eXPLdjRAATguKNyAvw-9~av zlO2p)%v?gu7)ExIc2n(TAyovv=tAFW8m;aY5e$u8S0dJYbN03GO6LsAOxc?K{wBuX zlvQ}j`p+rnFzSP!ja#6DaHZVCrok?o^+!bdfG{L>7#*Y&R1V+!xoNk&sssY%32n~D z8B%%|uZz^vqwhE-&l9B>LPqO*cZjYEBYFa4Rxf8an}zF8S5u!q<$Ne}jqK|EhWLZq zSu~`}`0s4|2g;F2E8G?T)Ez)_h! zvB@{O3o_5law^=1N;Gxc=8B*w6(Wkuhkz4YyP^akVqo0k{lF~VP1}?i`bt(}$(9|f znEazPg2`fRFO-7)9O0{mqr{dZ;m|R>B%U6L3{FKx0r6~|=LZi?^nAXXCz1HHCC=aw zEC6sR+1A$RwPUIM!Xdc+bYtLLXg|!X=PqjB>#I&yEtwr%h3<@B(PCf!i7|SW8t)&c z;x7n*9T8AlYbxKyapu~IX5w4r_0bK6Q|to3;yMIU)Rz`|Ufm5;L8(_P!QJmU+Mg4LpkW6cG1_qc-9{}$Cv|LN1YUl4AMfr9&=o(1! zDuxPr+AMHgJTr>JZ1=h)}fxifUM;})Vk?|5}QMLQR&k!VzYm~pHHXt2lm z+#b#NF*F6KuGl*nb@zv3J(}A(^OfAhAjKuW{=IF@+e^XA?sOOQW4V1QzNa6Y&VD(7 zPflH@ghdAY4eD5{?p3Eoq1Fms$KmOzR3=WXMexq#h1XWDb?TwT8p2wTioP#aE2pRiKdvZ8H2R=D zl|F8d^d_WQU$`s)?283c(Rn!0)RVJK5xxHurY9t3s4mg@iD`~IYn_c~u#{Z2Myw}- z3ku|=s8{_Mc*p~~x{jnTKTu+&8O~3*ak*{v4%s7wy=zgP438$8>@}H%it&ckJ@&nt z2QkcGo6<{fr*ZQ{$u2*Z3fXUE>{po;-J{o4P_{fG48b@)nCjX~V{|n&{oy9BvK9>X zzOuHVo_ye3S8bhN=p$cNnnF+Nu3SHEHqrZM)`j}&up$GC(i_XNHS7n8l48D`(B+c*_Ywl%G9H7Bw|{X#0LxDv zr>|SW6MG~IFp3FjR+p;f4lt^fS57G5N-&C7D}bHur(ypX#j&q$v-71a0QJRZYU6~g z#_kZMb~|=UbWIN+MwVZ2yHEpZiD^pgS2}B{o5|X!V`hS&k=1G;Ln36CaE!)-nYf3W zl62$E+hsq+n>E5i6(+k$`ko&vQ#`^a`)z(Fl!S?S+(ZD&JR)F}k=oTIm*7sJ z=G*)w$f`d0QZeAfthKl@J(it-B=B1fG2wDs z!Z(@e9m>xP4Z?|8XesC)8}3-zYSGJfn{mGX(h43y)%(b-$Ab)r;NSy*m(CJtZ2u{b zXdty)nM^*supf+(jXy-A9ZvHSdT5PHiN>3nChB~VrK#c;W1#hKT_{N(q zgUX`=eXISner}FAOQgHrJ81!0&a4>C`F}c)uacmVmZSB3-ZT@rP&SG(7t$&TYFgU< zeJ4mz5J7wrdWc(1!6;T!@&-fnd^%2PWJwcEZ<=dNacRJ_|49aX*vSmzVzm&ANwM^Kq$SKZq ztOg;mK@$@cn~X*`WI5}`Yt@`@917Jo#>770XVgZ=zXxXXMfsVp)~EYD?WH_;j;An; zHyYmG$9M|7Zv8cAIjD%~e9IqLsq3SfG+v+3=jZq~LM!`t-O(YX{eBj=9j~Ld)!0lZ z!>w2AdmZVQqV}iIidm$JsWUb;f#r1H4ilU^5W+CQX?8<$R#UWmHNI7ShrB47E_Wk8 z;trxD8iqNMrIOmgtT1NvFpK&ooigm_k$xM;1WkM@M%mAE7-qWUZ?grLlNG!s8bgr|QNBI^P$QesC+*8>aZj&=&Twr( z3kf?#xZrz5VxjhinQu&{MM{g7Fy-6?PZGA7$+4G=BfAW@%CE znynL8S$Tl~Ccs3|BjJ{P-X`B4@b~jN^5r;KO0xx+~6%iB=3(tycAE52GEre z#>w7B(?5o2BG!Ba0`1M1d-MEM6GUdPXU%hM_TGbwu3LrTUC<8YPLECtiGAe!7{0^% zomaTAT@x`d-KoBQ`op}66Izpbq(6Sp;FS~ivT;EJhi0(|0%lYd#j@bbFTmt+l%1NU>TzcLWXRqoj<{{-ROgK+ zZo8bBt7%+%{SB`NlUV2p4}hAj)1jM*44^^j{|haiiE?e={H+QL;9&L`?lg~Q4{W(+ zIBP;R_7S?JXDldw8m(h%rRVSDgH}M6kLvAWqM9zs==`4-AQ(zGwEQu4$d400cG z&^nLZne61AZ6v@3UojK zQui{ruZ4dQ*iW0{3?3-Oag82lY7EDSH8(%5wt{IL+fVbxYMrn^G_3W=mSq2Z9)0qq ztb`=^^ioGj1Q7}_nNC&oq4xY{G#AY=IKV+@%fcyey_|j;g6&(|z~DJvzf?wbI>NBW zO4csjXrK0VtFz%z9Yy+WO~15L`oCkUpOXDd@(@SVC)C*ba0AOkL8Ioe%FW7sJ0NeA z#^)iOEhM_V0`GQdvXkObM|W2Uv9gM6N2JHl{w+*xVJ-Z4WxH8p8uWB-b+z5TnmhLW z(k&v14&9w{GHO&;|R!&CCWvYj=+ zpOq=eIZP6F#&g2?9@WpAD)D&n%BawAL?!d>BtGGAfYQ@dMQJR{Rvp%3bXUjdC6Xon z>To8=PWg#OvtojnM&~t~Sd^-h2c>83iTPyP=V69xKu({P@z*Z=cPZc?Ery3q#`LLu z5XJPW<&Fgpd~icXV4jnSEed?`hE&1#gbr#QX$}5Uv*z)7V9R>s{+}cK%XXqKjt)~Y ze__uTd_EDHHHhIK^jQHpa5mI`^4pm8jHn>}!=>g`%FIPofb;gj*0X& z-ycfFl58l>0HCE0+8^Esr`GaC3C^C{RfWPCSF73J0d_0Wt8|>^+f069)9{5r~P`{g1#T&&%fMyv?PXn|UdHs~OWOVXI zR1HWM(Q_SS$fN1`vFiLDk$OzW$i(UdgJW8J62<4Z;Ol=@`Qn(&I9nmAEVFth6q5(H z;(&$C(d=;)Wajiqw+6)oTXS@ypAxe>MFcWZS=*pP>n4U0?G$iC+t)VBbJBfNk-WJK z@fZf#gojluYP6Ajg8&$WSZyMamjV>{q`H^@!4P+}6-^VN zak+hOx(+U3Mft;yQ%cU>Rk3B5?F~)S z3Py3e)SX=>mJxMcHduMSqSGbEm}tAWmd$qzLQ2JO1AtK7sjBe6ey;#i5iPyD8eE&F z#f7Dwgjhy%ZbkO}{3ap5mjF>CK; zK?HRIh=M?D)tZWz3n!~PumL{Wxx5QmF`ZJ{SDM<;xJ$tS1)SIrmfClh;%acP!2tP2 zky8|vGR;L=s}@;uwFfGCJ4!I%Z~go+5Aer&hQ*6oO~r=>h}7$LdbHjPhAATFIhEhA z_`s6CvY86f+30yd)d|JlwWwXL!px|nO0_H%q3ImyDy1JpAH7vs-&>76H*B%xbMyqH zOk6c}zWn_a`4c&M=r+1Nbhd=8nY9a3BO{un;jAgjW>SI{$|z(tubLKyRNl1JNqDV<5DvR096$jG?vXK*hUQ^>VF8W1DI4y*ehbg|7|sL0Zc4ci;zX~! zV)Ujxs{RK+#Q5>X*(bYZx$t;A zYJTU*I90M5ceKCWn{jz+J|!B5C*sgNs>_Adv=K!ttM%WcxZ5gI zCs*GVEV(_mJNB?yQb77!r76918EN#+v0P>kK6>ku^rxyh-Vdf&dtiuZRNz)O^DFH; za~ac9IF9v~alB3)TmS=_4L#K-6A>Gal&rw|U#gp)X8X61 zu@TnLu!Om)A0I36oJmDWX%)*N2$x&EQF{0Et~57~%DbGdMy_J{iyjW`-VW1^Pn4Ac z3Io~moM#Q!Ocw)I5@j3bA3WYPcFrF^|BZUieBYNO)e9GUZXD)Zw^ns4;aKd5#X9{a z=|N?+A8BmIyorl3{hfFJE4a%t;0^Wbe{vlUBfI31IX>JB)qutxskCcuP;Ho+ZK7lA zbQx2gSalhx;poH!=F^EuFFUeIPPm2H7rMUn+pz3uKcdMcpmk)44brwgMD5? za*prKqK}O{i14oGwk;nE*nLa%EvV5w%!d)SN8q<3nj?3Ti%s|H;lB?rRc==#Y>N|= za-3Z1piM4v7KS`7>${#0`nd3f1kR$~RLPOO=JLDQ4d$BlYsHvV*g4R6YUxcUX%UQm zrNlGKisU2#cYs#%yq4}DfCQ&vr#~Gq=n@71&rAZKg**^kk=+7O-HMXX?eUnKK#~GN zI3PlCLk!623M<#*tIFT1%v6Qx{gkpe)>D@3{pJhSI`D5!O0%K7t^VS1Fy#QaDJ>3@ z8g5j01=XjrkqKduWL_@5M>=CziRF8(`=_=_GUWq%OX8*nwyW-7>~1f zYCS&^&rw+D;Ts-IG`7Gw7JI6`Gj7eC02!KM7x2%l=+XbQY&N{<3QY{;XxOF1uG^X)+jaIjME)(swUdpN zmxA)Y^UWHLh-~8@9PAlQgz(qTitL+ar#D~AK z=`T%1#-6yHd<29u1;XKzS8bf!j+@+IFiG=15SS&K8s$4;V6uM=wIc38*$e`Sf%5R0 z!MHVjf8~v@^?9*hcLhHtaWQ*@Wkm(x;Frp#QxC7+Qfk$?N|;=sZk~G%acq5U9FWkg zzmp3_C^dCH;089#5~IfB_>eIwHMnI$$vmm63D$Q4m}oe;cDkYbxt*c#aMYdcTJu2d zfgLYZk0~8Q0MXu2?5sazct6vGU1v8%c9Pi)llD_3FM#|&RTta8=V5c6=aHcmf?aj{ zi0`Ax@~>)-YAC~nMo%!=RaQ&^hFRUy7Jy200zI+*`dTjzmFz z+8vxX9LTOMt5Rp4h`=T2SizHS{?!&vBqzuw!ZYKC+^s>vO{!NJtf=MosQZFLt->rr z6H9m)(dN&U2tZDedPc0y+o{EuVf*$!avA(h6qjvUp#Jrx{z8f#dg~jX59s<- z6USCW=;8B4JZ4 zZwh0Yd)c)dE!X1=zGFtBV^r%I%`qU1TDa71Gm03Cu_ZI-xV0nlqu0-tkGSDVTYaaH zA!#4{5pbqu+UD`$aFbRPJx12yrLpZH^L6#UqJj&$G)?~=gbOF8fo0|j4_JJjzE5;J zJ5Ivbe?s;OEGnC5uGAbSh3GotARdLkoMwoF^uHL3H0OM!@%t72g>RNWB9pzaXBpjZ z*K_|yQ@@RWv9|NGvplx5fI2=NUf*cE7|J*HViT8`s@}}s6*iGh<=qPV%Y>i$T9tC4 zWc*tqyX&ZJYuet98CrE=kUrRpm_p`Tb#1VUI)wH&LhMFQ{&Zh;^WUB2X zfyFe=SaiC=)sLD(+w46OUy_&X>R5;-<(cfyde0B8hT4@&` zkcTMi?YbY(JDjKdgtI7UaHD3k$=uKwmR7~Z5~t;Ihic2tM+{(<;UX&W0Xf04km)M< z@YNGD5dMv6hLHrx_|;aywng3r-=SB>qalE)*?EYc5v{UVR+Le9vmmC&bTKrtG7bqh zHE+yQz=ycD>A_n&-|4E!~6Gy(~ zkBU7+#0&WqH;`QEU;)O#%$WY6Sv<;i(2m(DnY611DOAp%!f}@k#Yhegp|v~pkI*dR zZCJ`ng2DEGwy$g80RM#i+!a%`Z<)CJ<14I{3)~KkL(H_4fID@)hCg?7QD@Gu&~G$g zRz!fA@Wf`VB<-)=wc>`+@{wpa)N_%u92B)X;R2mquqMQf!7O^u2d%;%LtaTWraAw_ z{%$}(DUh)$Ig{PJX&S^kZOKw}hw4QO*=LpFBOv+}9V8aCX^Vk-p2Rdk=gsWX-)6(d znzGmNp*S|QH>a5!Opbx;U;nyF|G1s6gGlbhs?7OnWGlfX<3%dvNQ!Sp2M)2^Kt)Rr zCnnh9wR zO)tOIn#y;X0=L8U;oRnqVQyv7=!I^fKK;d`+P%SzS>j!mvoGFUmW4B=|DgkPZl zKF~jNYwI+uq^GiDTI9cl**GyuLkS&ln7PnQeno(_W z!o{tm%R<$8u;&e>Pp)jcv8cM9F`>IZ`i*>MzOs(mc;Wt?{^*}+%k)fbT94%yY0pbbS{U&1ja@sMw)n)R zVjLq>mF*I~>jDRF#T}cfJk8Mj7}V|nUtMK2$S!8R!H5oIpgS7g1RsJcM}cA3U-feZ zM8BmBu;HRZd{_eLGaph6{o_Q$`$#nVNywiYK%a8quAAFf`(j}5;zRd3aG}9Lt~mCA za6=A$PR|5$E>rwkK`620j%>na3R2jx#wM=RiHUS$9Cl#u_E;M8W(1yS*u`{UKI5VGekpH>k0}Ku}O{0*0a{+F(k~r>T}(m z`P=N@B*9Y_ChhAbip|-WMXJo*fWJ6-vB$m#!anyb0SFrbFPsbE%NoQ%-SPC|Bbk6Y zjorwS9~6EghZD(&Be#FB^%awyxJt8QBqe2NO&bSGL z^B87{q?Y2J=Ws+*ekdNEFyQ~%QpVKOw65-pKVD~w&I+T0W(s zZlxsHUSxCmEIaTVZ>}m!ntvHRFfx?AzKAx%QRq3QyEFvk@C~j@hQ9Tv*J^h-CZyuO z8%nrRp?vuN>|*Ee{LB1ko0`cAEx||33OO96dWjz0ZwmEIRiYN?H-Fi>DKQkE|M9|r0e{8JF0RZ!E7o_zlvNR#m4n8IS6ze zH=TtlxKJ!GDiak>4N_}*fJcsPnvm2RynEhIuDU5S`a2$`+k{>P*^}RQ*)($%a%}Ub zEHE)KL5|@0?i8rDUZ}jgb5)6j238vF&(F@p{jzvnj=F;o^t~O68{0wLW3}&<{b`?p zN|Hq5Ol>IZj=(p4<*(#`+jtfKd_ADdvE-A2%CT*8U1xv^G{Ie-Bb&If>a14O9^h@8 z(uwB!|JGq}Ty`g>b6au8Bt6=RvZ3j5&(*V!WaBE(`v(iOMIVE#8OapsxOeg5Bt$@% z>V{FmTMghJhYAXVz4)I2X0g{){bXnW29fb6%n#T(ctE!DtcjZrP?2?o1zs>oSLLMC zynexY$lyiU{I+lnLlF6_WraUNSCKE6a{FEOPJudy^OV}rW3S|ec20R_7(@}`rSPSY z_}vm7)B}XIadS{Ecx6soHmw<4k;KP9x)aezd-WR>Z}q`9-e}q=@uGzxH>QV@8p()A9uW*v z_RPi8bc}=F-|g|B{6m!`pgDgtO??E|17xeZ+)CC!6(jfyOP$8&D0}qnQ5=qK(D`9F zkHAv4TqWKynXpX@B1#vX4DrG^V>oaxxtsQ24!JvD|hg*CDo(KI7l}ymk z)EFJ_<&a(8l$)CiW!)6--4~rXS9fX;+n#1qEBLLbmP}HiIH!+W6`J6n3;S3Avu?6Eh(RX17{UJkR%2CkwXqqQ0BV z?jj%VIL9v#;AT7-8+w0n@(9w!@XU^K06hUhe-~14IF(D|th~vK49E2le63r0s{gz? zJkG-fgKGSEidyTb)j(R}0PmCD#jmd!{7y!9Gs<65v`z69737@y>%vHfLENCCgaqXL z8c>ONwo%G1MsxLZm}45jyGFK&_KpbX0D${9ME6h_+fB+BSL2w^y3vuJBd6A0m*BE(<~idW`{FC*bH?-Xyn-~s0d zI!^+@Bo+TP>09^(_vic{MDv_o*E5>H21?~W2!E_}6acQh9VDbR1oE}f5CGSPRt40K zop~2MiN41hQ_A{YdGqG4h-ElP`7BC4-2w!pe7n8X`_diGVK&zI8l-nNFFqE$@7o_g zKn4i=2ZJvn?~SE*ETh zjAM;>;>HH6;VB90?=wE1y)LB>f=FpZSsT9*`%6s5;UoQ!I0EnPTLfm09W#c~2}n>G zNy);w!)b08#@6ICc5#ge+OkY-l7xC?97&L9?dcJ=$2ds%T^$2%MPOIzM*ct<;Ga8a z>Z(bu3VUk>^K+JwtX_Q;5uX|L%|R6c=nHVdGXF+L^8Kl$8uW<)$y6qe9BKUdAn^|r ztsC_EV8wRX3K2cgNEh=61vmb2*@Q~_Ixoi8#ad1|sr{c^nQhCgSGdXJBYM|G@bN zSw!|2DC{j8?V8tv2nVjvXDD@^OQY?6L}?$|v={O}n6;sV&pRC|v|^@+(uNxT?(pb+ zWPoUd&b;cLHEb4Ke^=-($`!4!J~TIh1Oh6ixdqy0y$%bEn@e7fRIGVyXYk##1NHxv zZtsW;9L_sBp`68If=N*)XXhyK?dS7OeRcKw84thG^Sm~bxO&R(ztj$iBPi}#%XvSB z7#4Spxj|5NQu@NPZX=G^jPh&xo@kwmEam-zGE)}^X=>hNAC9Y_L(<(DFL{tPZ$-F# z#y%6uJ{cWxxI(x9slujxyJgQ|L46;qBMPsF;elg4Z9Co0CBsF@x0`;)?yYP*S1Dr~ zTH42$tVF$Hd57$Gr$&6gTI|1$E(>@zfJ6Bl`%(mRP`Dd=sp#0aHDUwEgj@>3^$0SV zbX>j?1o2xwJx+eVrj5ZL)mhS2BTPZ{I`J~3_h-X+I%NQu=Mv+{iujf<^GwKTD(6+R+G@r;qLP2@o_=z%}Efmvd^z-LWx6|d_)6>(G znr}<@Lty(NSG`Dl{@%`evzK6#YAvy5<-a}G>T{hl{N!~cbUaTo-HCWJq-z#CHC%wc z2cJRfxiJZvhtV1!VNa0cBDA$RN~kLaNf?)lpK7H{1*m0yEx({(u%YACcNqnTv*fFi z$r0dy51yQSMSn|Et!Q~JTy^7mne{ZuCuw{vcFr!S^(2CXi|NzbwPb8nHjF?rFJ@0J zGMuM%%bzLZe_nv&_Y_^MOqy5`EA!Z8x==%i@CIO9!7&FLphnjP{MW?64B3>2Ys0)_ z1NlolEuF4Db1<_c76Vq9SS9XXi`{~0P8}WWwBRMvR1MT9wCI(A3y>F-weG@^D872o zVAdKMtvF)pT3OPP@UPKN4w}X%T6OeEAp!4PSpP6E&8pHAEF1Z4Ar*2?h4sb9?y1*O z;h3e+ubh=6T92uf+XeT|N)qZPh!3#axMT|uZ>2%C?%)z}Kx^UGI9~JiWZkM^i4HYmxCZI(-F7C(74c{3`SFY zu;a~@l5i7E((_A790z%0dL~0}cNb`$!uJ`KI$Wf7C@wT;Zo|HjAuAl3Yo8R?HWfj3 zh4)?P*+<{CV8_KAIw}4=BTUl~rRRQ0ptB(&!3RmdQ&SJ8^R@lpEfYbtp;<^!JIq%b zgXa0|X6*8uxOvMW3>u*Qy*K?7Q}tq+Tfw#AJpWF(Xz6LTo$ZU5;tihuAkP8RmPqMR zmPM=Qat;`LD-FS*aQ*(zm>l9F2Z=BEfv z&K>`OF}(L?gs7NI4 z$yTy}X<`ja;d}cL?ynOAW+0ez84DS$+sJlJ6&aR`vsd|5TUDS>g6kp{uX;Wh=a<4^ zxjFdNasTRmzuL;oI^mP56X_|89rn*i$L|?9^P}9I!jTm@f)NypIW6nas%+Yx1v}os zprAEu?5?+O7jHN>Dnf+rCx$cJve_kQ{0>tAucXy z{Q2gSj>h8t<@P$i(lWk&Rzv0Bd#-g~^UfNdlYaZtATf@nmp;3Gq|7mKQjBZKYT4ix zOJxQ-j}gu8k(|AKd72|M-3I`Te5!y>5^$3NXF75PwTBPFW$&k$?^3Sey@_hY4V3c1L!xxhjeHJnD5-QpQS>q{qb8%-loU z@O)pl#MAPz$}s!rQ#pRmVDE3H>4M4_GBy6G)V-~qn3_Xd08v+LH* z(q5YX&vS`?ud*QKZFg(mc}J>B+OL(U>&E{D_d)iu<&6#5aiFvwcBpisrTkLC%183_ z#{2}7zhhO4$y71$8pdgkxgW*ct2-vnvIC)kR=4wc3wA!&lP2qziBn1|XF;!#{!ONT zvyuh*`W+CBetI@z_H)?})NHS0rcW#(=KI&!@`lyw$VGNi4GC?1f}OTeyF-|N$hk~O zvd2wm^{9SjJ#2+gWT~G|Q^mXM)x&7V_@V+0d-Yb*bZpAagdv@I3qvjBd1|jU+{F0r zP@K8{kEyo|t0P#VMF+P83ldy|yE_DeTkzl{I0SdM0KwfoxVvj`cXxMp=S|MN_q_9K z|KMZx?C$ESRjXE2r_`{xZbdaO=4R`qh57NOC$zYf@+jhb>wQ9@rpCN zRW-Sah_aigPKbxqgCUy}p^kM}aWMd3)^)u9*L~DsImH~%O(*HW54Yn4+^S-s9TW$Z@t$<)TB9Q`%kWAqrN!-Q>_5j!%qUn>5Csc z58(ljI3YgRE9O^+)V51hEmUZh*WeP)(l~sSu+aZC?!OBvV96-)ij6*sfcF%M3;uQm zFDxLfohTx38!WCXU}kHTMXvt@%O?Xb`reQP?#Is${)AX+*&RWUT(~_a3|n+%u>aagz(X}6Y&lCx4Y%M9iuD!P zMNXjja&|Lck3gxorL;0+DZ2Z+XwNpGVrXx4NGPDQ#z{*Wq!h9srfF6u_ym6c`O-S( z#}~(Db;(3RGp@uPiamK`c&f_XDhK=5)+nLJ8v)e&fiUPRqyjhdJ4x=;wyeL(^yqUY z3^7V4h973|0c&`kJ*_#morcHRA?kFjxyi>>6|`Z7p}AsYnZ*R!k8uLG_?#77&l+o3 z8=CSjj3zwis9OINcMIU@gYXxgIOr9jD6$kC&G6OOf|@N*-#IeO+Yt|cW?<@6y@ zZBIk8PNm7NkJRR?Q)+T}Okl(l`FCy#F8BNM;Uw0#2ScMg+atGIOe1YKv#rOCCa_11 znYXrhf-%%du~Bp5X81#e=v_`WE-knIV(!fxAp#X&PblUcC&j7vpHu;Ee#8|i70H}; zl~krp@WA4S=mhwwagENJktqlxf9JnRwiiUA={0D@ZTG21$-Z^h2d(+VZiZC$H~?(0 zQGq*botQb+a{+$bAJa)JE`kv9LRKYr zEJv(l7j+$y)y-z&kFywLH{BGPN-?3n1MAcFVDkKOAxCouQtz|w3}(K_!1;G9pd-*p zKB-6Z)kEsc!!lK|b2hVgL(@EhylOZoJ=mW+rQ62^Sf~JEnhvTD3Djk|n}w%Fz=o!# zmD%Tu3En&-rq_#DQ1qslzv`|3M}-OW_E;NfrQ&hofs2wSjdYj1br5j zWc(`UqbPW@%*fXei}Job2i9n0HRrtKOJ60zryi8pm^XB23e`*%if8Muf)47HNoZm} zV22LLC+fof6dQw8`ajVuobh7aPGj53H&+lsf8o{pxMQ}cwZuv-l;kIBVAn$e_rK32 zp~V!`-~t7(_(K#hz&UhkszlELbdyoEOOEfyfQfJl5fQ)SB377%g+(2m>GNN^qV?K( zddB86_n>5~>tkom+<+vR2D!og?9hP&3*wCLHX%xp152%AB3E*^)g#&4PZ;K(9qW@F z?;2gCW;@;ud@n)Wj}@pXT|vt3<|r*zAcZflIqaWdpZeQAHnpC%hSh9&{tn-dLTziD zSnO!2?AGCL_(r8!9!~+BuWQ9;S6)p#kkJa?5MjDR#zb=HhXju9-T~)3TBk6?Fw@{` zP5^mnrG}A<*$gPPDChQhLe$<-eWAZIQ4_y;L@4Mi!daZ-*rW^PD>FF*N~R(P02Rq4 z1mGW~dn0R$A9lbNSZ;<6@EsJktRV8^CfOY03$F1brQBRI5&0p9uEt!+XKEG)~_mWR!r5OQ==@EWjx5-C?#p21)d^s)Bs*@tLC_M2%W0nRs2lu7Zz zWic=-|0Yv---wEDFz>Q>&r;s4DDrpaTCNWhDWvh;+2;q>uB7TtHbo492_*`)pxxa=OxdmV77B-1x@Gh`uzr zE6``*wTHqnOSh#t^}Vy2%;48U_R|vBCnB*xz2tEEG^S})mZqPz9wAu7$$sj32CHAR z=gzG1Hl+Zn-Et3d^0;Y@K1Q7lFVim*>oUkB^*+qA~e zDrLf&fASjwOhzGi4VM=nx?Q0D{@H7_X*cQ&$UKiLG%@W{-W^rJvHL|PlJ%9{|Jb=f zJLecLmQFgzYxtxCW5;OlMn0i>rlw?eF9M_Nbqu8~<_(50v-oCh_xpUGF8(!hT+$p( z9!G#_%1zi73JYE$2_h^b?@MB~$EqMSKc9GA8bOgiC`In>c#b`Px4rz?hZj*xMh*!H z7?;NYGAH^$_y$Rzh2M^T_8}t2KvP~lYLYmQ1B?1;(?T4x;UkPr&KdKhd%wMQR3gtS zL&BEY?Ck=_iu1Pnrbe`0m(|nqEB3aN7X^0g9^$1MHhX};&^+ll^_sh?*M%~)pt~;c zUj7F~UHk){%WzI99{jtKHncSYG!|;vLsk8|^ooKHEqHn4R|SBs{S5Y!S{m z{E*UM_~6>%ARSaYXN}uLgcg!n9G}sqe03kw=5kfY3Z^`L(2@xmAEBTOGkIz1RY?a% zQ32&jvTkMnZel$=HujDeFw6p@h*JRZL>}JO;3TtsvE~foNgOErOmwND)!N-!Jjze3 z3?G?g(E@~yr1!5KY)TfnxW{2%j<+ zs_Sscg?c%T>$?}FeTOZ$zuSYqmiVj3C1Pkrc0|R{hK3nvm5jlu2!jyb6l!-owE*Tf zaD1&?1{zM3R+r*#-HFrpAC#_kEpR(vs=2=r5;G}Ir5rLTs5+VI&8r!b+P~kL`L$Gs zghc9XUhyn08#g$`|Y<19HFoDjZ@Hn^3Kx?Ot`n;lI$3X zF-un)9mFyI-Cm=C@48nL>y5ro(oCSC`978uuD~#>#iH|8Mvy`Sa(lk~Ecu!{zA=ve1L6~_~{Y3e__2qUchEo3a_O{LbV3OaEfNfo(({{a}Yw~o( zcMb#2^2IUV@us^^T=Pn^l;6;*a*Ws3F@=`X0e29ss~Y zwl4<@;FCoqi)>omSGkOvaie`v*er)v>YT->{^aoh(fRq=E`uK#Xx(d^LE}W{$CC|f zmU?{PRYtlCkOqNVzSf1hkxsEAf+%GaXr_B3?jx z8)x6#;Ke8L`h(RCj5O`np*A!*PeZ~Y@)EfH4+mb6q=vecmvU!vNBu#>@;mJsv&!R< zE0gq?`XipLpO$70KJ*ZPoFUULeZ_+7AcWA`Z~fn*!v~}drH0w>smxBL5igQ^vOA&c zGfVS_ix}%~V%2X9|6w2wQw7UYr0sGy*Dzq+E!+7b+^;=F?i)&15}oYE*MfLsaD4vn z?_*$5`DYw<4x4I@0b}XqiqB6Vb8d)m!UGxL9Bby;Epl_u)vn2?2M>kl$q`qX2pid3-4B0221WT$r&3s?c^7oU`Og@+&4`gw`<{A*pr}pSInS_>Bk3@TMHf>E{b0Bh- zN)9F(nJug294}>Qzyto}GNKV{`M1WmdnSoTYjldYX>A4|W{~7FQ~aGER9r)i0aDsf z^Sx(8H;#P4eJ!v0pqL^}+48dphrGZ#+wBpDC1nq&(+iz zk0g-6emIc!Js=IVqYT&egkx(gz zEmNKzNbJcbje~2%l?}@PJLKE{o^-^T2h{HEqQ7CEUU*Uy2C%wZ{ruaMB+_-MAn-(s z8}{tJETp7l9FcoVf9oQp>m2n%vFn(-&7o>|T^sDQVc@Rw{L|p~lK?UW11RBn^8TC4 zOIk#$HBW%HGf_f-=>Hd05F{`se}~%?#u!Yera8W2SCQ-18W4D33-^94)e1L z*yEz)ZaIFm13EB4veE2`eu(UhHAHEob5fmc-B2Ef>$&y7lj6kF|vrV#f?Uq1lZPR9Vw&yyLeMh zluYeVZ*z7$mB>gGIrAuSlZBUd@>3vw7nbY#FLZ>lKhyCpe)hwSHnRkPhSfBOn~fG9 zOz-Unt_fUShj_)jhqfET_#a+NlKMlWCKmx!9V*BV!bpbU3y==D`e^ku)y%|sT`pBT zZDI^oo|{?mi4E}42Q6JO>agPjj0R*m#mFDUln6YmEeWm@hDS!=40uZ)<~kcn_d4%) zg7R8q-SC#%U9Iz=03uk;cBBlV_N(_bXFd-VgWr^|Q`Cd=%69oTriXlb2Yf7h2yp8$ zG)tKV8%|^&A2A7`s*B+d(%^=F-MrkbNs~u|wKq;sfy=?s_V)U8)_SLd!a20Jn0Aqz zB_%_P$z-sDB^UWq{h9@corW*)YWkVHb1^%j;y%$XQPLH@d9g(f(>R#_ZB%SA6*2C( z>s?u{lu2(A;<+L%|G0RiZti^fA=kA9=G=_fvGOQ~95VDeq&JVoJnDv`MvBTx+hH!_ znoK1EIg%dr7v4Pzuk00OZb>boYF<15T{qR;WQe^j!q-OM;{I1Zj;5nhyPFj^T*t=X zE7>=U0k8WXlQ{ER>M879R;7-8*Mi6i7I0lnN3oOeB-izP>MeWfEl4nKifl$sY(^=@ zEN8hDuSfWy)jD5JHniOEscf!%Dj znGw44r>$A^^|NDz6xy!b)sp=thR}Z<#!n)$nRaSQ62Gg^y-!q9H91~7?it|Ng#eP; zgDSoe+vgSnLb(Qw9}8m@WtIF8>S?-4RM-C~(w?xDwIfJ)8XOSqznX@JS@ga4lnde9 zz7nYZiU5&mWQq(_U_1?V3|7I@1$R_Q94VdlmRzq*ylyE_TUpu|7B;jB z9XcTGFG03n^HaeBd1Il*OL9;h+dP^&I&@jnVT=OLSGGOGG# zat$92Zf9^A+u=Ly#xALs>TpC6aqjrbwlk5#ag5w%CZ*Md)A&cl9uV{Uw>pdiN6}}m z9f(yc1Ru4W2S_8^j^1<*W>~fj*y3}g_KVRQ#+4O%Ei{Mpm8SEi*o1-ef|70MW^V^> z$C=MiauJ?+&K)w}Mn*P8M4!ue@EF=vCnR5h@Wxx(tPV~j4V>GNgHW6aIf*ny3qF3!8lj2V#LM=>Qrd(x9nL@!E4p(#W>qoX?sc|7S+dUAhBe7VI4y) zGf?bnrlvJR(NQn+XfB*EsJ>_3jbZa4EFNHjq>han6*oa*b2?HBIn;sL(m|kC99@ zR9xB)^Pb8Qi8Q?dLN~vJ4r|$8KO11$A;pc$f3jv)gz$D7^<=q~qMn)#3YApPDMtlo zGHJ^SG!MLGX)+R0-o?LT6>O8A;N{@eZMi7l${QQn8OgCiP8vb9SIr`w(stmsgu!QQ z{4JM0V4xwZGB`|HN`UcXzT*^Eucc4>t4)zP$hKQawxCkY0F!0~Sqc~zMmessxs>fv zG`?0xBI|FuM_y`~ENfi++cFuKS7o@muI$7#F9yzqQq5|)m+PO;!;cE8R_#(?Ooj&( zbWC&f!&@TDVoe7pM+#4sdlSl8PD%$L7!TLW)N9K*yhwpmByCDZ#GTX;8<~FU`fA%f z?s6TD!r@U{Y2kyM5-T-koYhK2v+zzdmgiumUwJ%_dL@3aiKlpf!z?cLkZwyn?ZhOzqZc$hF2OWvr{v1LBd%g(+;^_m zF1Pip-ywvAg*y+X^4-|sL;77Q55;bZ?J{orb=x6<#E4dfR&Q9ZS9WS6r`xem(dnXZ z|R&6;eS5C>!(e8iO?|5z!4_6LBl)+yt^dn0SP6TREhKk)sCP| z2%$1kiRJy&Fo$QD_mQW>Gu%M7K3wyv^RH8Gih z(?h`Pzznp=m-(AnZce&b?h;-U?nJe4`9fk^FQ^XFIcz{L+T+%L@EgkvQjjgsQS{Bs zey8Y>n$gHkxESSqlWdI4?y_kXT-BvNpr@WwOV@OV8n;-JWGjU=3ITwCrCb95NG#UA zVL+(oYC0y(PU}V2+hZ;>uKA4q?mIGT9{KnQCpYf{AYR(iXw;d}iS3rhPAe#W6;VhMO*ZrT89Dd`gjku&SS&s^QH7u4c57ICo1i)4 zX-EKT1A=$qQ7RV%r75A%Y||gOW(X)!1z!rEs1PQKq|!i8B{i5jATwzZ_2J{sY^rj$ z(-WCXHq$Bv^(JoMcUtE+7j746HHG3J?UaYquu|LH2QNEwN~n&;aR-1{sp0%fJ}%>U zp-^Aip8Nh2(vZKHV86gM9(HERf&0r#`?83(m|TE4ysX3Z-*0g5D;O$~0SqdEG@O!U zrw36%2i2}e50Bq^$nmuhz(FIcQW`8!CRT(|s~*3AJ8VC$H{Lu}4UvRUY;}sr_y4p2 zya#Xb>3n-US1WU6e0+j@O^?baC6og0)K144uUXOLWp8uxf*DDkNy)r8faT8+`B3@d zY6Nn>hZhRwUYQIyAn3w0rZ+FrfuU%SE(KQ>g96{58V|t8!*@z&jXMB@RuCe3qRFN; z)S#1y_Uc6D^Ml3zNE;r#y)-K3QMYo8E{JrxY+yHebjRD}PGbB;inj%B!~iTWTQI7MKQnUz znf`aQ|#FmnCqSf~*MWsBG%0l31G7Q$*v#1I(WNG-SCZxjzG_ zleb|ZQG>!stkN+IvK(e+OobnCbdkS7fm7nDly~=cdB<@sxKy3E26-g9p#ad>oV84H z^axR{8imNHAR-R?!OkvVlO2x zvB@fwIGFZz0=^bzK`j@P0m^aQQE_z=MW3kc932xdCeLbRS$v|(X^MLV_oHr|ZA)TV z!q^t5x~9iAY;k6o``!{>l!>U!Egd|Ei>q1MuT_jOaRWg=>=bD6L%G*8M_=Cc2iHN# z7Z@G`h``wvbI??k>E{nN9YP1~#n#VhMNEpaQF?hyybs#Yo;wh2Tb?2BhVIVz@T1Nx zWBV?xWD|^Ux&0?j8ubg&Ra44?35a}>K$i=93A)^wlI5EF*vd5Zo}c-P(8l%g%dm0U zTZ!s@?G(D7Gp{ry5*lNutH4?E71^)X^2s^(=DCmvH~Da53u$12q?{~bkK-e{;7op> z+j1sNv7i<^HXZ3e9r z!L~PPe#JN%#$A=190hm4D`HX_@U>4NGe-v|Iu!U z_J#d)x1Yxh(Ivm8dgU;s@FY@z3`12%j->Bs^qAQH*sL1656nP##I!8hisA_Oi$@^= z-P+9YL`rIh!zw}TI+3prazYPkAXq3sEH(v^e_a@xTrErAk94uG39q-Bsy&lNjB+4A zvfrKvbGBrqG-Q)+OgO_ZrPpqA;*D2geTz`Ek0BFRhqE7TkRkjuw0QVO)%uKOK6q0~ zNff8Paa@_aEefn2e(41PJ1a2GLa6c4%RsYqm2E{3qGN)ky6a(eOmw;Eovs6xpRC^~ zJ{(X0gnZ$i{>EY3#ig7Cnk{>RM`@>g(=()<@Ed3n*^lFwdjpVBXpO62>Q--je)RgfS9 zw>qBTxWuT{yiYsvVJk6d0?jE^n;X(oWo>u9r)i(qef9CuHLdSfpSk3Xr^&-Po%R}I z7t}8s9oyTVsh1w%B^`e_8uzKsnC8jSuROUZcUkTXS&f}*YStJmw`=QW>(;L%+%J0{ zlj9=1%c{i{CU}n@;0UZXw97Y;GA!2fCTrTj*F@Xo)_E+c@3dqU zJt?q&Or8o7nFQycOOD4FcKzFmv(ITURzK3g?ghsZopL2MOe!<5#a&mS^%+L1W>U*m7bNmMF(gd!Z$^de`}dO6 zCqdwhD49_}%ibI69;M=GUlOpS5FSv7IAl~-o@562p*i8pXtY8A8ZBp{!Um#_x^cvV z4Pi0+CB=p@v+ zxQ)L@_3#E5rat~8Yy7>s`MxxixHK?Q6}NwIh}ytzW-5=xV|v%Z!q(+#a26^@9~tP! zU;x$=BlT72=+iai9*5j z=*b5Y8lYd*dttMZ`kDHCj;t0lJ*K4A9F7K84v&cRwr71tEbKKVBrxWxD7%8pyR`?4 zwW}?mdimnDKy+`EIy1$0cq9?s@LDSG#e-HO>lB?PFm+NS(smu`wV_2duhevBcZEAz znsTKdUa}hfb?~lAnx+iSo7_fZGvsej$X#oLCFMe3;>tpbk=r*3JA$zj+XdS%p3Z&P zzRBI15J05;@rCCp2`=~lv7;!Gk2h+zNkW%f?VR2GIFrc<$)5g3Z}WHtE8yI=YOkhp zEpizE@TU%3F^fC?%TQgtxbC^>P_4gdla~R6(*Qr_bEvCr%HRKAUNl~s7U#{xF37=RZwykk!iRkz< z!YEBvefTrmy*RR`+LQREX7MA5$msjqv!b%!YS|1X_rR2Ciu!}Z<`8yhs%f-C*kc|K z&39uswvNk9Q(%2yQ}#mhXkfDtJqCwXcIoty2Tlj8x*ZwsbVUUICiq3)YWQJIAA4K> z=a1dB#~0=Kt*P!@`gVM|;U~kjtu+MV&DAgH?qw^>^NpG@qMBctOP&ah_7~kAxLH`k zjV)&VpYIpcc#5*a)+qp?&5)S-e~ImqUueOiZQ1nJ2@a)aJY7b`(4qk8Mb7}^&&M8l z4Sed+Dz28T&V!nF^$!Ylq`ls!gu?J7UmC7bt9xpWN%d9wag!Hn7rS_EkN%MUm@Iy% zbJt(eXL5^T+PDaP!GgH8T)2wnpAS6B&v(A(!l!X_wH#G}HO^joaf`;n4ZJU^W>D^h z6k3Jh1R`U_lIo>BPERRK<1^woVTQy?{hT(gd6+s5@>>Hel&s)yXt4oJbvu2uW@!Z;$8EQyC*V0ewSvD~~^)i~!3~V!O&8li-Dlzlqxm?du9m&ae ze(p2s=8NsySjN#h9QAE1Fy2#tn5Tv4C$oYo=q5oPW-=|Kze)lLV2^-SUv1(nuJLCH z{30E9+^k3Rh1Nf3q}W|Wozw^@=AmxjW6Um^O*M5`yn(rh{aXB^P{n6jgN}gq4HF2D z4(I0EJ5RJ13^X2~x&-sHsX*DBLfwVX1CWQ_D9}{5v(%1toteznXTkW>H{yY;*zOjG zK<>sU`1GtqR+fQ0sXMO;_nv9RT@5_`{@2wEQ+?N7KL~yWMEg9;OmvxHfgyugkKv%= z!@UuHD-8}X>R%d@Ymj{6xv-!8wdl=y`1E*gIK1zEc`{9)5#>AUy^h@%^ywDV4K4X%Kv`$^+d-bflFCrA6jpSCkj%txUE0^P@>w^9Q9UB>}^SBx2^ zHVC{|b&X@Lp3AYX$Aq{F|IY`3$>2g?pYfRaQ*7-4>)TDy^LaQuLhh7QZli<30O9fZ z0nXAO7GwwM_fFF9!f6hXow21%;OK(tz%Po2_-HZ6fEu1VkXdxR>Tah{=N~;OxSLMq zVb_1cb0lY+yV(a8*D;yU<2PCgr&vH2gCqi z_D7-C>aptD@iB>wW4V;{J4$ma1;0iqedCl`>8k`#N!8fm;_obY&?5qCCW=!FArxzq z!nRPL^??R^0!HAfNxmLmF|y}cf!YwaNGaN%MI7v`+u4g$&?|X%dqA$InVM42*IxqT zmttWS_M}MF@NA9q)UL|2RXjylXouMaY~ybo)(j&B=+ad^1$)MFA3>+LEF<>3uhK0s zg;4`1OT!>m>rhL}jWwgGwV$L8dRIZ(YTFMzGyRbku)Zuvp4cYTN@uaUbDw11>#e)j z_o z7LvS`U|R8MUAzD>^D8cblLVC=o%JnfDb*JFJEX2BOBmndxB`x@Q<-ylcHakY@&n~E{ z#hf+cP$6=Oaf0Z|3h2VHnj^*c8RZ(|AOoSK2iliri)ra<2TyKpuVmPb0s&DyT}dkd zBnlM8!e&&|LFzip64W68saBqJC8dD9SS5szlwZWrM`m?W?PGf@44-*w44ye$GN=+j zh{1|N0U1bi=*JqXb)9`8GT7FD2$X(b4e1SH=YU#F5dMHi@08CtsnhUK5x4K~w_t@_g+_9N z$Msl2dr4P|Ijs1wNe}1k{oW?HWg5)x@XROzU#}NSwse)+xuW!T)SkZho8*hGx7xHB zxC(>j3Dp6$3a4Caia5bwmvIvsp&_Y8uu*rvKloUaG$z-cMoC>?GF}H;oAWyLDCo}E zJd(lZz7`f~SfPc21c~nmj;tHmJO+VM^8KcPda&L_?r>mu&Bozwb|^`Wsz?DD3p5Cx zl^mvJ9pN7~2m~p1FzYp+=WZwV(GRLm+OaDd-#z`7T;K&)><@4ATcu1+PIW!J*H$und;@bd&DGsM*ZTog9^aXDDlRY!E)a3X`NPwd%`(!Kd*J14wL& z-9_7!p-?`I-bK~*dSe8|OuvkI^gjMSHtEl&t^P2^UcN`&CGnjJbH|r8TCil7;#a-} zPnW|8CRZhUGdVoR!|r#re^5B=NL|PmBo~`&Z}V+?qQ0)2QAut-Ca%nJj@>#JZcqW= z?w!J&{}>OVSQ!i4=DSQuiWBDYy$3*!q4BObqgD_S$SEE@j!1^$=w>Ez%eI}`#2+fK zwJ4s5D+Xr>2Y7Vr_V8rJ9N(J?+~{su+FdthK(T}0(Cjp{KWCxl)e5b~NhJF}VM)FI zpNFyh^!&KpS2lfy1Dc_~iE*^a^N#~DAW?}*Ik9371JUAV+<0RZx`L%K43JYTnUto| zm}dA*6Jn_a9dOUxe8;->Vu_TDE6Wu{0#y7}=V7(|X}-!1+gHHOhzZa=lMOh=Du(&L zVdTo_pAsPe;0tY!{-!K!Nblw+Kcd!idl(3|=%soyf1&~++xXfhe*~)10>t`J1;62d z)Rkfuk6mszWdct1BZlR*PI&jr=Dli4b?)Z+$lrZ;@xce#RdkTpWSCRpFaT&kg@3k!>DQlJuh%qUFhzmMHeY+RZIh(vUP3;vPHI!FdktbggBvKf> zdYx@M>#F=grvo~R-Cv4Yu;#JMzh=$g9liu~pq6bjIM98NXmh(w01MhHV|3`JKj|al zK26;=(*tR8YzFhn-!w^Fj7S2J00R@1fFxarrQ*VgoXC+KI}6T{+pjZu^24vaEZ$|Y z`I3XFAi9}OTF&HZR#T`a7^cM;KEJqHiB){xC9vGW`j}Of4*o8nLLzbfc-FR!pQ(jD znA`;LB2$1|IEZ#t z!o0ppv;jcT3)D&kf`e17qPf-<5&2oac@~^T(XSjgIn8b86lMbQkxH8Q%P;vQJ&KZr ztOMr#$wvmFG``*6fv>dfOFPpQm;U}g8E*2+Wn_ezb(0duYRR;1flTRg{QD~qrw|j^ z{n*UoYL>?Mqh0z?i1mT=;1=zj(Y*&;N(>A?gm=lo!V3afBnjKh1e`trTSt7`6%U}y zW-Gg|@8!0i7Dm0sQ*+W6QA{GnWG z39Y1-yh9kt=P_`{C(;Od5H2mE4{5=^2LqttYenCE;RyuZrmaYNRIUNNCbu8o;?;vq zfkEIQtk-Vs&cU?`%Kon3b=3cai?#nag?`kn(+%W*#l4Akj+CK_h{!y6iBq8${LZCG8>d>)J2of-aL;xbW0!Bmi z<^y?MlwE#tDMdj)T$6jaNUYYZS^O=!jA=qcnF$F@4uS=+1nstkMN3X(Zd%z#uk1kL zKVB;9=>i{bDIDCy=^rYktK$UhZ#J}Kzcu47R$?^Sr}7|yxLpj~t5aL<9Y@s>9ejle zf_<$Y_&|ZnZ_puab5I`YcyH3S-k~B`8c;(5dg2#|Tw3ZN|sTmgcy>fXf zm1b1VG+6RYPyfiE)Bt9Ql70Af>Yu{WKaZq9dGR}B3c&LG`tO#x-(gG-p7q)m&whKM zO5j8iNE=zI;hc9)yXs5`aGOItq=tInJF;MpaK9$ZFi+9MZsg8l+x4dou|)?u%66W0 z9p)R$>Gz}xp#zDlfyk1;L2pMsf|m8N<2%tnPxG4VxFBwY7TIb(18QISYLrBeEi5|& z2YFaU3vM-yWa=P>IDwZNMSqhypD;`=-OK&|MU7r6eIT_92z90vN}DL7W){gR4Nj@l zno7AOE9Q|Lykq|rY%(JK3Cdj}HZwMVKvs|hFNc*3v{(R8*Kgh%ghr>ipvIH{KaEj> z+5Z=)vsmU;?qw@bfSpv#e zaqNs1$tO{o*oCMkQV3x;>g^7alczbYnn=GK>8t(a`IfQ`_N<~DuFZy32mLYtZ*WUM zF4GG9A3DnhhugV!4cMc$TY^EA*A}abDRE8zqVn*T@T*S|0s-S1@3X@DQ9pLZJ+Jdb z(@`)B(34WEEE{_yShA3BJ`~S^RWA_wnFejpxM*ZdrNU+UrfVX!M^C7nk}#ZRjq-;& z_uCJ}guT|J$DRt6LT;h@jJX#zJdGv~mk!2KYZ)1yj+Ta8s~{nUxi}fY#Vx4^1)qYh zDYYN9EkV2Q=uO7sLVIPDeYa@7 zk3y+*3Tg+(J?ewE0$dZy;jhzBII zJ&OL+ePNEYU0Ge04=ngT{?{Qs?&&vK$zXubesej??ak&qB&CNsc!;xL?b_vPtk>d7 zQ`L1pDi8Q^GaH`oN4duV_w#qPJc??YMOJKn{*+o<#eo_aRs%Gb>f-5-AD+3uM|q6B z@?^y8Z4foecK!`=5F`3eVL{vg{{I#aqFRXFZcWM8vYG|}Z_r%t{>2mJOE!Sm1}p_R zC)?>3E!IrZNvagaSax?+pyWT|I0QM{UlHggeS3;11WzhDqov7SRI% z{oh-hpgC~I6K#STW$NN-xt;2`2ayn#31KEmQRQiWp~EEc-@|bNz1=hZ7i-B43d&Ci88cLNuQZMfS*2Z^^dTVC31S~VoIQ?OP& zmDA&cw;3ILRBc{N;`#_TifM-}K&R9|^M!W`y(_|K8L0*6NE3%jgoM!m&e#2wmVS;j zW4)q)?e$z5t&dvTd3DkJ)-o+tH_4mmSY=@k*<4G)E??&4d70I~jHn=BMj1Lh~)lYxj{cu0`*eTr?(*yaG(-}e&9IE|Pxlmqf)9r>@P!%D zxXXgERzp5Ek*K2pALFV)o!V&@gP7q@;N0Niu|f@m>5@rG7Yn)&#PrO|3nDC@x9BTb z1_n(#YhF+MUBeNFJy2fh)389IO!1RthK!HQ?g;g|GiFsI$durv=8}&{hPI2rWqObV3vk~l?5w)P#Y;bw%;!k44!GZrDhcH`8_8_?>?AfBg?e^Fe5;{2V zfuz(#h9*+ELmie<5{M=u876e6U!NnEiFbYN*w>%KT958YB3&dcl~%am+dHNM34p0P zSZ@eOA_{UAGOnbdIhqhvt&l?Jv?dklAKWA#QJBs9rksxp0C>)lUp_0VNfwp87Ui;% zP}rDQ?MV37@6p#BD$u$0&@@09yeJ;mNNeU?ukNO?Uc<6jFGu8~$ralBEZ2ke8K z3v~FD2d7+FRid2eSBB5pQox+&_ifddodx}Ke&i`P2=$g)Y^&DT-`G#bleXUdbj%KU zSHUt&C7$qK;50dV4*^W6!2r7gM>g7aIT8O-MOn9Ha>08diR0d>%4FvnE4;_%n+AZK1+;?qOSMXa>zlwAPELG_0nq zw$#=)f-ECO!Vb1TslooRXj};H#hkkj9}Qx&hyEKh>jiLhP4GJhKXAxs@*2++ceV1u z-}r+8^SuQ7+T*o&y>8VH9hRc!T$1wr<3_XC!x)wPR6=5 zChdpt_F5iyj@SGBVa!sSggC=j=C_(&?b(N6K^7KYo{H&nYmsnUdmLd<$jmhR8S1zz zH{I&xjPm3DL@E6HPaqb?i!`GMI;!SqG0Ofvs`XVi;D-VV^`+v-3t2OtZd5|@Qf9J& zG&Zk)lpDq3WY(1$@8z?&am&pOZJ-lKskEknQv`_r=bRJ9IUKY64*kN%k6$24nm&qW z3Uug+fr*G?MpvSY=W{)6h|<+cV;AnVWsr6T6icmBwqR39r`yEv_0^K>y>#dc?#(~F+GxN;D449tX>x2U3`VAb!Eqn zDB^ezNX%v*{S8r!Qek)4!-O0*fgLGm4;BCPXnh8N#ip0r^Ug*q%(>>5=$+SSg{@p| z&brSD1|V_mY+B^~Vc@{n`ZK5WAuyz^2Pz;7-vw48ga7xs`1trid=bg()!^q+ z@{Wifaq44cY%$W&q_iu%f9sXc4A5+f)$J+kx}tCdEdqYT1JcU!?9U7~QS zh_SxM1Yev@F^Ylpjx^nuA3)V@>)5v?Qky>XwI2foM$Tea={HZn6#g&5clrdsP=%qq zALfs?uHvcuZS?b|_Ie1ZA1$x}4BLibag_{PO%NI&)~A7zS^1nNC?S#*yq#z9Hyv13 z_tpj4S7bOVgX4O5ckDsPG+n0Z8=(G?-dJL2jm3 zbdQ>rZzN-LX$%U~Yp|mLNE7W2{v8KdyJU<&vI=m;N%-xoVzHKwdam+g^WBjEp@&xX zmF5OcFw;bV(9+4C%5$#{8rxwyVjVFdSVnoi}V33xOOeQ0TIb!#3G zXvsmTz@RfcPZta&947}aBpBKQuZXn)N6#1-Y(HG>HMi7Bv)>fVVl3~GFm*aP|2iWy z1X9eL9&J%;8jAfybRAW6Z?7OqZys-sNJvOXqxQk+Md0{450g@U_fhYzf3M-E@u!VV zB?rmvD08*Z_2%&UUKXUtb>-O!F+e{fhkTCKN-*z@SrQ5CqOWzrWb>^YtAZ z91O7HL}Rg^S4Nw}MlmVH4eOnpoSd498QNK#pEqGk{=O{#3-!eJZY}apr4vi?jBEyLH8?+=<3lQw($Aki3kV2Y?T$>D)zV?n zQc&a#nY-PyJAWI+#^M=7Bxpgu$t@(+{&-Hv?$*{HpgYpi(!x8mudi7e z;Cjwj8{&4&Ei5Fg9}gfCB{!N99j%W>qphuz(A1WemUG3Q8fYVY(4;~khN3RHb!^TJ zi+%FsiLs9lM4I{p6r`)zG)pVtrlYeIso7CHBd=ecs3 z$~)T7RW-cU-UgyAp1i%`OKPwOacJMzh_;%8+K={t@U#YF2 zaHGEqU+NOEGzvK&pylAh0sdnlL*d5LpK8LjCXRQlKExy=m!PPRx>t`%XjW!+w(jh! zEy6nUZEs{F^V)Gdxp!zp*6UZVW|meU%m4R+vQHE%_C;~F<2%>;tIHRXBxj2C@Odt^ zx@FeKk8{OOj(o^E`tR>Cnmtp^aeU|O+r8wv z&*bc7@!lkX7i69lPy7(S%w*cLPR3MU`I}d--rqkOFjNyxZ+V$N5iTHeIpD@TN&NB4 zkNy2Cb4$Lfytr-6if3dpWN1%iq-*OB+!L$Z%a<>|!uo}+A4KBvbsgP-^afYwoam|s z`8Awy^9wFU?d?d#iBr*?8QC`!Z7gi| zTv>EsKoenNVscOAO#S4i;?$Gpf`iK{E5lBlc-vFxMkX-!YiYe;v)T3a^>9{VY0tOm z(-bx$z`z0ZSo9OjEVPm|zsAGV<~!8J(2& z_I5!*!G{kYW@l&1FMSU@FQoISb9i7$NlTmgmu38rUg{Pdw~g*hSsx}xG`h&)pm-eM zh)WoOkndQ0jxRrFI%bhxKke~2dRDDeI7SK_IM|Fe)hDrS8^_Ukd3g&RoSmGUY;Dzg z)TS^w85tw4-CHN*IoeEW^yJBt;!4gAka!UGttJw4dJK1oW@TsRpMJ3jWh!poyh%4V zHa5O(;*Q03jDneLK*&aXAHE@W;Pp)z;@lTASh^hL#-JP>Qw`)&54l~9jg2iVoUI98 z=E=ANsjp>a$zgL}&X=OaK0R^78S~3mk7-HyVA5;2_D6nq?uL?Zlg!HoeS0&q7}N zd|xj73IOAoN1gac_Qcbt4+)HB^`TvBE#1L4H|(Z0P7^H(0KtE>JkUB|mp5)TU3d41 z<-TVW3UDfMq9ZGI%iifv#U3iCw~dV=eE+i0o_$r^s_TToL;$vP?a%V@ICBiOZKinW zY`T8H;zBo2Ae045Cr7X+z-G+N%>gUq$2(`Y#R~+@jQ`QTBQ1pcqXofFG3cpscZZ2? zbq{jE{^G@z#NDPjSy1Uubwa$iLEwcR&x9x52i!ePcZr?^3B)0LC@CRAkt(Rd;^K7eu%gmZHH3ko;REfk!~6E(>P#!)G8k6-eCJIFGc&VIn>K;*t{#NG zK3>cE_iIRI-u(FJGCR&F<2LtMcX!n=CjHqnh2PVu%5OYWls+C;`-fRug%AQ26M7or z6m25a)YV<8Lubdxk$WPAn$I(=a286F5qC8tU96Wa$FFSK!t){2T|#`U?`wWrw|-C)mm{Mc`W(#f^UG+u;wHr9=;)-xM7*b`6B_a@=*ozS zc5slcDClcfK@A%ZC7C#Y4N%dQ=XgXi3wnC?+_|pqZb~P^G(~BN%@#7+FC-)cxf`Lg z_)<+Rt@P$|tlB#@4`P0#9G3K`UH<*rm&vaM8OEd^Lu;#KH1&eFw;2k>WU;z4Mvh2I zFdj)Q8541Xf|hFuE1tPpI#qX#CWrqJe^z@1`8XyfhD?@S3c#6Lr~CzUul@$Qv6cr; zG(u2lXsD@a+_B$2&n@X&ewdCCI&Su}qr3O+9i!8)T)mp8r+%UTdygAwc4_6_JxMpW zkz2QJ!T#1GlJQEc`La8HBd>2uV+(uUy~Foc_<=tIT7As59s{}ynRb?!m$$HJt*Fod z9OP0S1Nb~I zWKe74LBm{|vhs2#M@K>Ye%N=Ykc=vK8FjtWk>t&23fq=%HE-zVmXVs8nwz^nh_pk7 zJX~eDd{UNFHYnrH9o60Os3W6+Vr(onA$9k|M-Jm&{o=_O64(5fn(e-=;23-M>=E9F z<>TkL*Xm_X)b&$c+``mX2L03-@ZP-WYt8?BB!VPtT8|BXzQNHQ|Ge)Pg#uBFWdV)( z?tn3A$L?}d6zZ;`Q}9H`Zl(tYGsqlm65}@Kg7C30&4h4OW6dZ()U^HNY&oRb7-EY%I`&q+;Bcd77=>aV%D z!G%9we$V|JDc5o?f`oGd0S-Phjk$DW^neCWVr zz^e@L_SK!aHfibUUm1+Fd-rg7d{yWy78;1aUAR!=J6sQB{M0Oc&ux=X?Lpw`t^9`m z{{FtcDhO}@`9p`=U%$RuZ9s*bG=E5NxGY@VWIE8#?*s}}R9@~Yn9~7HPYX-sXrRQ?6{Mn`mkS|j;+r?Ox3yVXSv|2) zfZ1^&PL;CRghGcrJ4$hJF}NDQF4oO0DJiM1uTKx0M@B2yx#nV6cm8QQELg9tg4-se zM!}0Bek_fF^zFtFr)t+$l}lHbb!{9RzIQqK(AW&h;nc~xDDJOaRPIY%U0tJ1Dc^=_ zYqPVzWTjvTkhBn@EC{JUAwx-;AkMKsDX8gfYdc2EeDVaTkbE^JW*{UQ&_`PR!bP|q z@Uh%ys0P>m1O~3VyStV(NaPjX0rLP*dcw?1=4$j!OR}ryFMq$CvF5yP8+tjY$+ys3 z;Wq+;zo@7PqBa_hMj~1G`ugJW;J-K@pS`kTAF|E*s)8tmg@>YFHa0h(2Kkt#<_UAC zsHoWbKntGu4F)5RMq8a3ZUAQ^X)_zQSZEjGI~9snYY8TgEicYBtHbu;rp87G8lluE zV`-{<2n-O&*b|E_g2rc=rfjGK<$OOeky$UGye}D5iVpu=$2E|$RYIcplb7a$h#|oZ z$3TvPH~|rr^mYZpHY+QulV)bN3StwD3_3j!qCwzv;l6g`9QEg=SWiRt2B!mMb{x;a zS{urXfNL)p8W>zT_FF~x@*?0Umn-8VM4h<`$qE`8J^lUKDk{%2GwBQ6F3JiD&vJ73 zfFlJ3&lWml3gbb81qROZ%3icQb?T63*TJHfkV)syKe7H=dU|U$!bdeU;-%Gv5ItUH zXey+Cgc*VW9AF0_A|lcjt`fjXxpVifzsCJq!kzxN*g`~IZ*OmFYwO34ko+Pq4>Tsi z6)ari^!5(rUw!dBEJfxh*EcjktO^pcDtHQwxPIeCZ?OjnP2IzNNqD=)bEs#-q5h7{ z6HvHhIYvMNr`jt|()R!Iew$K?md~Izlqm4_hQMK^MeExwK|5R9Fd{L|Kn_vyO@BjF zT~Cigh-?X(8WUq^Yb&-?3_}7K+!PTU8Jn8wsITAPqSHG%swzco(gBPC#L>~So-kJ^uDbMTW`#cym6_GSdA9ebybmc9}y1C5zr+^PW6^% zW=XG5WQ*urih%Vbq zqG>@Q3<2@D(TR!k9v&WN&u&D(H&|z%tn?p!2O0w5I$R$Eb)O(X(m54LkfeH{GoPVh zpsqE}cVvP)U%G@qe4Y&j#|3ULMN@-=f>v3zYi8`Q4qOSZcIFFVRVhVn?UmAQOkahc z8$>Yxk5iMACcGaZ8WxoP9YT)0Wm5=NO3>bhhJS>s_pF)^7z3E>7L6?^_;yLg_zI6K zEJ4g~i-Q)fN7u{G}>f z$ei+DxXxgGQSm4DW~TV={%=bF{~HzmP3V~^;5~m6dWY%y3h2)}Cw9mGKQaIR<#8Qr Y+!hg6oks4?A~?qhLvw?Cr0eDX0^I6MkpKVy literal 0 HcmV?d00001 diff --git a/preprocess/preprocess.m b/preprocess/preprocess.m new file mode 100755 index 0000000..359ca95 --- /dev/null +++ b/preprocess/preprocess.m @@ -0,0 +1,8 @@ +%% x preprocessing +x = x'; +x = sgolayfilt(x,2,17); +x =diff(x); +max_x=max(max(x)); +min_x=min(min(x)); +x=(x-min_x)/(max_x-min_x); +x = x'; diff --git a/preprocess/preprocess_mango.m b/preprocess/preprocess_mango.m new file mode 100755 index 0000000..6c903c5 --- /dev/null +++ b/preprocess/preprocess_mango.m @@ -0,0 +1,15 @@ +%% x preprocessing +clear; +load('dataset/mango/mango_origin.mat') +x = x'; +x = sgolayfilt(x,2,17); +x =diff(x); +max_x=max(max(x)); +min_x=min(min(x)); +x=(x-min_x)/(max_x-min_x); +x = x'; +y = y'; +min_y = min(min(y)); +max_y = max(max(y)); +y = (y-min_y)/(max_y-min_y); +save('dataset/mango/mango_preprocessed.mat') diff --git a/preprocess/train_test_split.m b/preprocess/train_test_split.m new file mode 100755 index 0000000..8c07bbc --- /dev/null +++ b/preprocess/train_test_split.m @@ -0,0 +1,15 @@ +data=[x,y]; +test_rate = 0.3; +data_num = size(x, 1); +train_num = round((1-test_rate) * data_num); +idx=randperm(data_num); +train_idx=idx(1:train_num); +test_idx=idx(train_num+1:data_num); +data_train=data(train_idx,:); +x_train=data_train(:,1:size(x, 2)); +y_train=data_train(:,size(x, 2)+1); +test_data=data(test_idx,:); +x_test=test_data(:,1:size(x, 2)); +y_test=test_data(:,size(x, 2)+1); +clear data_num train_num idx train_idx test_idx test_data train_data x y; +clear data data_train test_rate; diff --git a/utils.py b/utils.py new file mode 100755 index 0000000..b0147f0 --- /dev/null +++ b/utils.py @@ -0,0 +1,153 @@ +from scipy.io import loadmat +import numpy as np +from sklearn.model_selection import train_test_split +import os +import shutil + + +def load_data(data_path='./pine_water_cc.mat', validation_rate=0.25): + if data_path == './pine_water_cc.mat': + data = loadmat(data_path) + y_train, y_test = data['value_train'], data['value_test'] + print('Value train shape: ', y_train.shape, 'Value test shape', y_test.shape) + y_max_value, y_min_value = data['value_max'], data['value_min'] + x_train, x_test = data['DL_train'], data['DL_test'] + elif data_path == './N_100_leaf_cc.mat': + data = loadmat(data_path) + y_train, y_test = data['y_train'], data['y_test'] + x_train, x_test = data['x_train'], data['x_test'] + y_max_value, y_min_value = data['max_y'], data['min_y'] + x_train = np.expand_dims(x_train, axis=1) + x_test = np.expand_dims(x_test, axis=1) + x_validation, y_validation = x_test, y_test + return x_train, x_test, x_validation, y_train, y_test, y_validation, y_max_value, y_min_value + else: + data = loadmat(data_path) + y_train, y_test = data['y_train'], data['y_test'] + x_train, x_test = data['x_train'], data['x_test'] + y_max_value, y_min_value = data['max_y'], data['min_y'] + x_train = np.expand_dims(x_train, axis=1) + x_test = np.expand_dims(x_test, axis=1) + print('SG17 DATA train shape: ', x_train.shape, 'SG17 DATA test shape', x_test.shape) + + print('Mini value: %s, Max value %s.' % (y_min_value, y_max_value)) + + x_train, x_validation, y_train, y_validation = train_test_split(x_train, y_train, test_size=validation_rate, + random_state=8) + + return x_train, x_test, x_validation, y_train, y_test, y_validation, y_max_value, y_min_value + + +def mkdir_if_not_exist(dir_name, is_delete=False): + """ + 创建文件夹 + :param dir_name: 文件夹 + :param is_delete: 是否删除 + :return: 是否成功 + """ + try: + if is_delete: + if os.path.exists(dir_name): + shutil.rmtree(dir_name) + print('[Info] 文件夹 "%s" 存在, 删除文件夹.' % dir_name) + + if not os.path.exists(dir_name): + os.makedirs(dir_name) + print('[Info] 文件夹 "%s" 不存在, 创建文件夹.' % dir_name) + return True + except Exception as e: + print('[Exception] %s' % e) + return False + + +class Config: + def __init__(self): + # 数据有关的参数 + self.validation_rate = 0.2 + # 训练有关参数 + self.train_epoch = 20000 + self.batch_size = 20 + # 是否训练的参数 + self.train_cnn = True + self.train_ms_cnn = True + self.train_ms_sc_cnn = True + # 是否评估参数 + self.evaluate_cnn = True + self.evaluate_ms_cnn = True + self.evaluate_ms_sc_cnn = True + # 要评估的保存好的模型列表 + self.evaluate_cnn_name_list = [] + self.evaluate_ms_cnn_name_list = [] + self.evaluate_ms_sc_cnn_name_list = [] + + # 存储训练出的模型和图片的文件夹 + self.img_dir = './pictures0331' + self.checkpoint_dir = './check_points0331' + + # 数据集选择 + self.data_set = './dataset_preprocess/corn/corn_mositure.mat' + + def show_yourself(self, to_text_file=None): + line_width = 36 + content = '\n' + # create line + line_text = 'Data Parameters' + line = '='*((line_width-len(line_text))//2) + line_text + '='*((line_width-len(line_text))//2) + line.ljust(line_width, '=') + content += line + '\n' + content += 'Validation Rate: ' + str(self.validation_rate) + '\n' + # create line + line_text = 'Training Parameters' + line = '=' * ((line_width - len(line_text)) // 2) + line_text + '=' * ((line_width - len(line_text)) // 2) + line.ljust(line_width, '=') + content += line + '\n' + content += 'Train CNN: ' + str(self.train_cnn) + '\n' + content += 'Train Ms CNN: ' + str(self.train_ms_cnn) + '\n' + content += 'Train Ms Sc CNN: ' + str(self.train_ms_sc_cnn) + '\n' + # create line + line_text = 'Evaluate Parameters' + line = '=' * ((line_width - len(line_text)) // 2) + line_text + '=' * ((line_width - len(line_text)) // 2) + line.ljust(line_width, '=') + content += line + '\n' + content += 'Train Epoch: ' + str(self.train_epoch) + '\n' + content += 'Train Batch Size: ' + str(self.batch_size) + '\n' + + content += 'Evaluate CNN: ' + str(self.evaluate_cnn) + '\n' + if len(self.evaluate_cnn_name_list) >=1: + content += 'Saved CNNs to Evaluate:\n' + for models in self.evaluate_cnn_name_list: + content += models + '\n' + + content += 'Evaluate Ms CNN: ' + str(self.evaluate_ms_cnn) + '\n' + if len(self.evaluate_ms_cnn_name_list) >= 1: + content += 'Saved Ms CNNs to Evaluate:\n' + for models in self.evaluate_ms_cnn_name_list: + content += models + '\n' + + content += 'Evaluate Ms Sc CNN: ' + str(self.evaluate_ms_cnn) + '\n' + if len(self.evaluate_ms_sc_cnn_name_list) >= 1: + content += 'Saved Ms Sc CNNs to Evaluate:\n' + for models in self.evaluate_ms_sc_cnn_name_list: + content += models + '\n' + + # create line + line_text = 'Saving Dir' + line = '=' * ((line_width - len(line_text)) // 2) + line_text + '=' * ((line_width - len(line_text)) // 2) + line.ljust(line_width, '=') + content += line + '\n' + content += 'Image Dir: ' + str(self.img_dir) + '\n' + content += 'Check Point Dir: ' + str(self.img_dir) + '\n' + print(content) + if to_text_file: + with open(to_text_file, 'w') as f: + f.write(content) + return content + + +if __name__ == '__main__': + config = Config() + config.show_yourself(to_text_file='name.txt') + x_train, x_test, x_validation, y_train, y_test, y_validation, y_max_value, y_min_value = \ + load_data(data_path='./yaowan_calibrate.mat', validation_rate=0.25) + print(x_train.shape, x_test.shape, y_train.shape, y_test.shape, x_validation.shape, y_validation.shape, + y_max_value, y_min_value)