diff --git a/Knee Osteoarthritis Prediction/Readme.md b/Knee Osteoarthritis Prediction/Readme.md new file mode 100644 index 0000000..57cff34 --- /dev/null +++ b/Knee Osteoarthritis Prediction/Readme.md @@ -0,0 +1,21 @@ +# Knee Osteoarthritis Prediction Dataset πŸ©ΈπŸ’Ό + +Welcome to the Knee osteoarthrits Xray Dataset! πŸŽ‰ The objective of this project is to classify xrays of Knee Xrays into 5 distinct categories based on severity of the disease. + +- [Knee Osteoarthritis](https://www.kaggle.com/datasets/shashwatwork/knee-osteoarthritis-dataset-with-severity/data) + +## Context πŸ“‹ + +Knee osteoarthritis is defined by degeneration of the knee’s articular cartilage the flexible, slippery material that normally protects bones from joint friction and impact. The condition also involves changes to the bone underneath the cartilage and can affect nearby soft tissues. Knee osteoarthritis is by far the most common type of arthritis to cause knee pain and often referred to as simply knee arthritis. Many other less common types of arthritis can also cause knee pain, including rheumatoid arthritis, pseudogout, and reactive arthritis. + +## Image Descriptors πŸ“ + +The dataset consists of four subdirectories: train, test, auto test and val (test and auto_test are the same). All four contain 5 subdirectories, each representing a severity of osteoarthritis. The test and auto_test subdirectory contains 1346 images in total, while the train subdirectory contains 5778 images in total and the val subdirectory contains 826 images in total. + +Each subdirectory has 5 severity grades: + +0. **Grade 0**: Healthy knee image. +1. **Grade 1(Doubtful)**: Doubtful joint narrowing with possible osteophytic lipping +2. **Grade 2(Minimal)**: Definite presence of osteophytes and possible joint space narrowing +3. **Grade 3(Moderate)**: Multiple osteophytes, definite joint space narrowing, with mild sclerosis. +4. **Grade 4(Severe)**: Large osteophytes, significant joint narrowing, and severe sclerosis. \ No newline at end of file diff --git a/Knee Osteoarthritis Prediction/knee-osteoarthritis-prediction.ipynb b/Knee Osteoarthritis Prediction/knee-osteoarthritis-prediction.ipynb new file mode 100644 index 0000000..36a95e0 --- /dev/null +++ b/Knee Osteoarthritis Prediction/knee-osteoarthritis-prediction.ipynb @@ -0,0 +1 @@ +{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"nvidiaTeslaT4","dataSources":[{"sourceId":2097406,"sourceType":"datasetVersion","datasetId":1257880}],"dockerImageVersionId":30733,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import numpy as np\nimport keras\nimport tensorflow as tf\n\n# Loading train, val, and test datasets\ntrain_dataset = keras.utils.image_dataset_from_directory(\n \"/kaggle/input/knee-osteoarthritis-dataset-with-severity/train\",\n image_size=(128, 128),\n batch_size=32\n)\n\nval_dataset = keras.utils.image_dataset_from_directory(\n \"/kaggle/input/knee-osteoarthritis-dataset-with-severity/val\",\n image_size=(128, 128),\n batch_size=32\n)\n\ntest_dataset = keras.utils.image_dataset_from_directory(\n \"/kaggle/input/knee-osteoarthritis-dataset-with-severity/test\",\n image_size=(128, 128),\n batch_size=32\n)\n\n# Normalizing datasets\nnormalization_layer = keras.layers.Rescaling(1./255)\ntrain_dataset = train_dataset.map(lambda x, y: (normalization_layer(x), y))\nval_dataset = val_dataset.map(lambda x, y: (normalization_layer(x), y))\ntest_dataset = test_dataset.map(lambda x, y: (normalization_layer(x), y))\n\n# Function to convert dataset to lists of negative images and labels\ndef process_dataset(dataset):\n images = []\n labels = []\n for imgs, lbls in dataset:\n for img, lbl in zip(imgs, lbls):\n img = tf.image.rgb_to_grayscale(img)\n img = img.numpy()\n negative_img = 1 - img\n images.append(negative_img)\n labels.append(lbl.numpy())\n return np.array(images), np.array(labels)\n\n# Processing the datasets\ntrain_images, train_labels = process_dataset(train_dataset)\nval_images, val_labels = process_dataset(val_dataset)\ntest_images, test_labels = process_dataset(test_dataset)\n\n# Converting labels to one-hot encoded format\ntrain_labels = keras.utils.to_categorical(train_labels, 5)\nval_labels = keras.utils.to_categorical(val_labels, 5)\ntest_labels = keras.utils.to_categorical(test_labels, 5)\n\n# Check the shapes of the datasets\nprint(f\"Train images shape: {train_images.shape}\")\nprint(f\"Train labels shape: {train_labels.shape}\")\nprint(f\"Validation images shape: {val_images.shape}\")\nprint(f\"Validation labels shape: {val_labels.shape}\")\nprint(f\"Test images shape: {test_images.shape}\")\nprint(f\"Test labels shape: {test_labels.shape}\")\n","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2024-06-16T04:26:28.800930Z","iopub.execute_input":"2024-06-16T04:26:28.801902Z","iopub.status.idle":"2024-06-16T04:27:08.204198Z","shell.execute_reply.started":"2024-06-16T04:26:28.801857Z","shell.execute_reply":"2024-06-16T04:27:08.203231Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stderr","text":"2024-06-16 04:26:30.409384: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n2024-06-16 04:26:30.409482: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n2024-06-16 04:26:30.537087: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n","output_type":"stream"},{"name":"stdout","text":"Found 5778 files belonging to 5 classes.\nFound 826 files belonging to 5 classes.\nFound 1656 files belonging to 5 classes.\nTrain images shape: (5778, 128, 128, 1)\nTrain labels shape: (5778, 5)\nValidation images shape: (826, 128, 128, 1)\nValidation labels shape: (826, 5)\nTest images shape: (1656, 128, 128, 1)\nTest labels shape: (1656, 5)\n","output_type":"stream"}]},{"cell_type":"code","source":"#combining into one\n\nimages = np.concatenate((train_images, val_images, test_images), axis=0)\nlabels = np.concatenate((train_labels, val_labels, test_labels), axis=0)\n\nimages.shape","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:27:08.206168Z","iopub.execute_input":"2024-06-16T04:27:08.206494Z","iopub.status.idle":"2024-06-16T04:27:08.414967Z","shell.execute_reply.started":"2024-06-16T04:27:08.206467Z","shell.execute_reply":"2024-06-16T04:27:08.414217Z"},"trusted":true},"execution_count":2,"outputs":[{"execution_count":2,"output_type":"execute_result","data":{"text/plain":"(8260, 128, 128, 1)"},"metadata":{}}]},{"cell_type":"code","source":"from keras.models import Sequential \nfrom keras.layers import Dense, Dropout, GlobalAvgPool2D, BatchNormalization\nfrom keras.layers import Conv2D, MaxPooling2D \nfrom keras.callbacks import ModelCheckpoint \n\nmodel = Sequential()\n\n#first cnn layers followed by a relu and a max pooling layer\nmodel.add(Conv2D(64, (3,3), padding = 'same', input_shape = (128, 128, 1), activation = 'relu'))\nmodel.add(BatchNormalization())\nmodel.add(MaxPooling2D(pool_size=(2,2)))\n\n#second layer similarly\nmodel.add(Conv2D(64, (3,3), padding = 'same', activation = 'relu'))\nmodel.add(BatchNormalization())\nmodel.add(MaxPooling2D(pool_size=(2,2)))\n\n#third layer\nmodel.add(Conv2D(64, (3,3), padding = 'same', activation = 'relu'))\nmodel.add(BatchNormalization())\nmodel.add(MaxPooling2D(pool_size=(2,2)))\n\n#brings everything to a simple shape, basically (128,1)\nmodel.add(GlobalAvgPool2D())\n\nmodel.add(Dense(1024, activation='relu'))\n\n#final layer\nmodel.add(Dense(5,activation='softmax'))\n\nmodel.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])\n\nmodel.summary()","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:27:08.416083Z","iopub.execute_input":"2024-06-16T04:27:08.416430Z","iopub.status.idle":"2024-06-16T04:27:08.613846Z","shell.execute_reply.started":"2024-06-16T04:27:08.416393Z","shell.execute_reply":"2024-06-16T04:27:08.612937Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stderr","text":"/opt/conda/lib/python3.10/site-packages/keras/src/layers/convolutional/base_conv.py:107: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"\u001b[1mModel: \"sequential\"\u001b[0m\n","text/html":"
Model: \"sequential\"\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n┑━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\nβ”‚ conv2d (\u001b[38;5;33mConv2D\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m, \u001b[38;5;34m128\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m640\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ batch_normalization β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m, \u001b[38;5;34m128\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m256\u001b[0m β”‚\nβ”‚ (\u001b[38;5;33mBatchNormalization\u001b[0m) β”‚ β”‚ β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m, \u001b[38;5;34m64\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m0\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m, \u001b[38;5;34m64\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m36,928\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ batch_normalization_1 β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m, \u001b[38;5;34m64\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m256\u001b[0m β”‚\nβ”‚ (\u001b[38;5;33mBatchNormalization\u001b[0m) β”‚ β”‚ β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ max_pooling2d_1 (\u001b[38;5;33mMaxPooling2D\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m0\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m36,928\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ batch_normalization_2 β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m256\u001b[0m β”‚\nβ”‚ (\u001b[38;5;33mBatchNormalization\u001b[0m) β”‚ β”‚ β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ max_pooling2d_2 (\u001b[38;5;33mMaxPooling2D\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m0\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ global_average_pooling2d β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) β”‚ \u001b[38;5;34m0\u001b[0m β”‚\nβ”‚ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m) β”‚ β”‚ β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ dense (\u001b[38;5;33mDense\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1024\u001b[0m) β”‚ \u001b[38;5;34m66,560\u001b[0m β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ dense_1 (\u001b[38;5;33mDense\u001b[0m) β”‚ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m5\u001b[0m) β”‚ \u001b[38;5;34m5,125\u001b[0m β”‚\nβ””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜\n","text/html":"
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n┃ Layer (type)                    ┃ Output Shape           ┃       Param # ┃\n┑━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\nβ”‚ conv2d (Conv2D)                 β”‚ (None, 128, 128, 64)   β”‚           640 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ batch_normalization             β”‚ (None, 128, 128, 64)   β”‚           256 β”‚\nβ”‚ (BatchNormalization)            β”‚                        β”‚               β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ max_pooling2d (MaxPooling2D)    β”‚ (None, 64, 64, 64)     β”‚             0 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ conv2d_1 (Conv2D)               β”‚ (None, 64, 64, 64)     β”‚        36,928 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ batch_normalization_1           β”‚ (None, 64, 64, 64)     β”‚           256 β”‚\nβ”‚ (BatchNormalization)            β”‚                        β”‚               β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ max_pooling2d_1 (MaxPooling2D)  β”‚ (None, 32, 32, 64)     β”‚             0 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ conv2d_2 (Conv2D)               β”‚ (None, 32, 32, 64)     β”‚        36,928 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ batch_normalization_2           β”‚ (None, 32, 32, 64)     β”‚           256 β”‚\nβ”‚ (BatchNormalization)            β”‚                        β”‚               β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ max_pooling2d_2 (MaxPooling2D)  β”‚ (None, 16, 16, 64)     β”‚             0 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ global_average_pooling2d        β”‚ (None, 64)             β”‚             0 β”‚\nβ”‚ (GlobalAveragePooling2D)        β”‚                        β”‚               β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ dense (Dense)                   β”‚ (None, 1024)           β”‚        66,560 β”‚\nβ”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€\nβ”‚ dense_1 (Dense)                 β”‚ (None, 5)              β”‚         5,125 β”‚\nβ””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m146,949\u001b[0m (574.02 KB)\n","text/html":"
 Total params: 146,949 (574.02 KB)\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m146,565\u001b[0m (572.52 KB)\n","text/html":"
 Trainable params: 146,565 (572.52 KB)\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m384\u001b[0m (1.50 KB)\n","text/html":"
 Non-trainable params: 384 (1.50 KB)\n
\n"},"metadata":{}}]},{"cell_type":"code","source":"from sklearn.model_selection import train_test_split\nx_train, x_test, y_train, y_test = train_test_split(images, labels, test_size=0.1)","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:27:08.615005Z","iopub.execute_input":"2024-06-16T04:27:08.615308Z","iopub.status.idle":"2024-06-16T04:27:09.187191Z","shell.execute_reply.started":"2024-06-16T04:27:08.615283Z","shell.execute_reply":"2024-06-16T04:27:09.186387Z"},"trusted":true},"execution_count":4,"outputs":[]},{"cell_type":"code","source":"from keras.callbacks import ModelCheckpoint\n\nmodel_checkpoint = ModelCheckpoint('model.keras', monitor='val_accuracy', save_best_only=True, verbose=1, mode='max')","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:27:09.189221Z","iopub.execute_input":"2024-06-16T04:27:09.189510Z","iopub.status.idle":"2024-06-16T04:27:09.195130Z","shell.execute_reply.started":"2024-06-16T04:27:09.189486Z","shell.execute_reply":"2024-06-16T04:27:09.194242Z"},"trusted":true},"execution_count":5,"outputs":[]},{"cell_type":"code","source":"history=model.fit(x_train,y_train,epochs=75,validation_split=0.2, batch_size=40,callbacks=[model_checkpoint])","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:27:09.196127Z","iopub.execute_input":"2024-06-16T04:27:09.197406Z","iopub.status.idle":"2024-06-16T04:32:31.868546Z","shell.execute_reply.started":"2024-06-16T04:27:09.197382Z","shell.execute_reply":"2024-06-16T04:32:31.867745Z"},"trusted":true},"execution_count":6,"outputs":[{"name":"stdout","text":"Epoch 1/75\n\u001b[1m 7/149\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m3s\u001b[0m 26ms/step - accuracy: 0.2644 - loss: 1.5503","output_type":"stream"},{"name":"stderr","text":"WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\nI0000 00:00:1718512039.785344 128 device_compiler.h:186] Compiled cluster using XLA! This line is logged at most once for the lifetime of the process.\nW0000 00:00:1718512039.805165 128 graph_launch.cc:671] Fallback to op-by-op mode because memset node breaks graph update\n","output_type":"stream"},{"name":"stdout","text":"\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 60ms/step - accuracy: 0.3631 - loss: 1.4462","output_type":"stream"},{"name":"stderr","text":"W0000 00:00:1718512048.629996 130 graph_launch.cc:671] Fallback to op-by-op mode because memset node breaks graph update\nW0000 00:00:1718512049.594612 128 graph_launch.cc:671] Fallback to op-by-op mode because memset node breaks graph update\n","output_type":"stream"},{"name":"stdout","text":"\nEpoch 1: val_accuracy improved from -inf to 0.25824, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m21s\u001b[0m 75ms/step - accuracy: 0.3632 - loss: 1.4461 - val_accuracy: 0.2582 - val_loss: 1.6645\nEpoch 2/75\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.3872 - loss: 1.4107\nEpoch 2: val_accuracy did not improve from 0.25824\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.3872 - loss: 1.4106 - val_accuracy: 0.1305 - val_loss: 1.6405\nEpoch 3/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.3998 - loss: 1.3953\nEpoch 3: val_accuracy improved from 0.25824 to 0.27102, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.3998 - loss: 1.3952 - val_accuracy: 0.2710 - val_loss: 1.7419\nEpoch 4/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.3927 - loss: 1.3856\nEpoch 4: val_accuracy did not improve from 0.27102\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.3927 - loss: 1.3855 - val_accuracy: 0.2670 - val_loss: 1.5633\nEpoch 5/75\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.4174 - loss: 1.3410\nEpoch 5: val_accuracy did not improve from 0.27102\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.4173 - loss: 1.3410 - val_accuracy: 0.1479 - val_loss: 2.5093\nEpoch 6/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.4328 - loss: 1.3063\nEpoch 6: val_accuracy improved from 0.27102 to 0.34499, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.4329 - loss: 1.3062 - val_accuracy: 0.3450 - val_loss: 1.4576\nEpoch 7/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.4550 - loss: 1.2223\nEpoch 7: val_accuracy did not improve from 0.34499\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.4552 - loss: 1.2221 - val_accuracy: 0.3416 - val_loss: 1.3725\nEpoch 8/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.4964 - loss: 1.1741\nEpoch 8: val_accuracy did not improve from 0.34499\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.4963 - loss: 1.1741 - val_accuracy: 0.1231 - val_loss: 2.2042\nEpoch 9/75\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.5115 - loss: 1.1327\nEpoch 9: val_accuracy did not improve from 0.34499\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5115 - loss: 1.1327 - val_accuracy: 0.1372 - val_loss: 2.9342\nEpoch 10/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.5230 - loss: 1.0828\nEpoch 10: val_accuracy improved from 0.34499 to 0.46738, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5232 - loss: 1.0825 - val_accuracy: 0.4674 - val_loss: 1.1848\nEpoch 11/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.5552 - loss: 1.0373\nEpoch 11: val_accuracy improved from 0.46738 to 0.50908, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5552 - loss: 1.0373 - val_accuracy: 0.5091 - val_loss: 1.1329\nEpoch 12/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.5605 - loss: 1.0064\nEpoch 12: val_accuracy did not improve from 0.50908\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5606 - loss: 1.0063 - val_accuracy: 0.4956 - val_loss: 1.1555\nEpoch 13/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.5852 - loss: 0.9698\nEpoch 13: val_accuracy did not improve from 0.50908\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5852 - loss: 0.9699 - val_accuracy: 0.1856 - val_loss: 2.7812\nEpoch 14/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.5885 - loss: 0.9539\nEpoch 14: val_accuracy improved from 0.50908 to 0.51379, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5887 - loss: 0.9537 - val_accuracy: 0.5138 - val_loss: 1.1554\nEpoch 15/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.5931 - loss: 0.9369\nEpoch 15: val_accuracy did not improve from 0.51379\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.5932 - loss: 0.9370 - val_accuracy: 0.1870 - val_loss: 2.3389\nEpoch 16/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6153 - loss: 0.9089\nEpoch 16: val_accuracy improved from 0.51379 to 0.55481, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6153 - loss: 0.9089 - val_accuracy: 0.5548 - val_loss: 1.0603\nEpoch 17/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6049 - loss: 0.9182\nEpoch 17: val_accuracy did not improve from 0.55481\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6050 - loss: 0.9180 - val_accuracy: 0.4983 - val_loss: 1.1813\nEpoch 18/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6302 - loss: 0.8710\nEpoch 18: val_accuracy did not improve from 0.55481\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6301 - loss: 0.8712 - val_accuracy: 0.1950 - val_loss: 3.7831\nEpoch 19/75\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6315 - loss: 0.8709\nEpoch 19: val_accuracy did not improve from 0.55481\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6315 - loss: 0.8709 - val_accuracy: 0.4970 - val_loss: 1.1695\nEpoch 20/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6362 - loss: 0.8539\nEpoch 20: val_accuracy did not improve from 0.55481\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6362 - loss: 0.8538 - val_accuracy: 0.5010 - val_loss: 1.3112\nEpoch 21/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6452 - loss: 0.8280\nEpoch 21: val_accuracy improved from 0.55481 to 0.55952, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6453 - loss: 0.8281 - val_accuracy: 0.5595 - val_loss: 1.0547\nEpoch 22/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6482 - loss: 0.8222\nEpoch 22: val_accuracy did not improve from 0.55952\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6483 - loss: 0.8220 - val_accuracy: 0.3631 - val_loss: 2.1385\nEpoch 23/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6629 - loss: 0.7819\nEpoch 23: val_accuracy did not improve from 0.55952\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6628 - loss: 0.7822 - val_accuracy: 0.2919 - val_loss: 1.9053\nEpoch 24/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6660 - loss: 0.7872\nEpoch 24: val_accuracy did not improve from 0.55952\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6660 - loss: 0.7873 - val_accuracy: 0.5528 - val_loss: 1.1515\nEpoch 25/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.6678 - loss: 0.7780\nEpoch 25: val_accuracy did not improve from 0.55952\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6677 - loss: 0.7781 - val_accuracy: 0.2434 - val_loss: 2.5526\nEpoch 26/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6791 - loss: 0.7644\nEpoch 26: val_accuracy did not improve from 0.55952\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6789 - loss: 0.7646 - val_accuracy: 0.5293 - val_loss: 1.1253\nEpoch 27/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.6855 - loss: 0.7431\nEpoch 27: val_accuracy improved from 0.55952 to 0.59987, saving model to model.keras\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6853 - loss: 0.7434 - val_accuracy: 0.5999 - val_loss: 1.0301\nEpoch 28/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.6887 - loss: 0.7218\nEpoch 28: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6886 - loss: 0.7219 - val_accuracy: 0.5817 - val_loss: 1.0336\nEpoch 29/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.6957 - loss: 0.7090\nEpoch 29: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.6958 - loss: 0.7090 - val_accuracy: 0.5205 - val_loss: 1.3257\nEpoch 30/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.7126 - loss: 0.6794\nEpoch 30: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7123 - loss: 0.6800 - val_accuracy: 0.5696 - val_loss: 1.1282\nEpoch 31/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7136 - loss: 0.6690\nEpoch 31: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7135 - loss: 0.6693 - val_accuracy: 0.4936 - val_loss: 1.3762\nEpoch 32/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7219 - loss: 0.6512\nEpoch 32: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7219 - loss: 0.6514 - val_accuracy: 0.5192 - val_loss: 1.2084\nEpoch 33/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7212 - loss: 0.6475\nEpoch 33: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7211 - loss: 0.6478 - val_accuracy: 0.5548 - val_loss: 1.1369\nEpoch 34/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7308 - loss: 0.6205\nEpoch 34: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7308 - loss: 0.6206 - val_accuracy: 0.5837 - val_loss: 1.0790\nEpoch 35/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7495 - loss: 0.6090\nEpoch 35: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7493 - loss: 0.6093 - val_accuracy: 0.4445 - val_loss: 1.4558\nEpoch 36/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7545 - loss: 0.5843\nEpoch 36: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7542 - loss: 0.5848 - val_accuracy: 0.5837 - val_loss: 1.3020\nEpoch 37/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7594 - loss: 0.5855\nEpoch 37: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7593 - loss: 0.5855 - val_accuracy: 0.5192 - val_loss: 1.3373\nEpoch 38/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7555 - loss: 0.5754\nEpoch 38: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7553 - loss: 0.5757 - val_accuracy: 0.5743 - val_loss: 1.2139\nEpoch 39/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.7704 - loss: 0.5642\nEpoch 39: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7704 - loss: 0.5642 - val_accuracy: 0.5118 - val_loss: 1.5321\nEpoch 40/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7894 - loss: 0.5162\nEpoch 40: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7890 - loss: 0.5170 - val_accuracy: 0.5084 - val_loss: 1.6940\nEpoch 41/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.7732 - loss: 0.5409\nEpoch 41: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7733 - loss: 0.5408 - val_accuracy: 0.4418 - val_loss: 1.8387\nEpoch 42/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.7806 - loss: 0.5054\nEpoch 42: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.7805 - loss: 0.5057 - val_accuracy: 0.5077 - val_loss: 1.5920\nEpoch 43/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8035 - loss: 0.4706\nEpoch 43: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8035 - loss: 0.4708 - val_accuracy: 0.4728 - val_loss: 1.6184\nEpoch 44/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8009 - loss: 0.4862\nEpoch 44: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8009 - loss: 0.4862 - val_accuracy: 0.5434 - val_loss: 1.5181\nEpoch 45/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8031 - loss: 0.4821\nEpoch 45: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8031 - loss: 0.4820 - val_accuracy: 0.5965 - val_loss: 1.3539\nEpoch 46/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8259 - loss: 0.4282\nEpoch 46: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8256 - loss: 0.4288 - val_accuracy: 0.5864 - val_loss: 1.2449\nEpoch 47/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8145 - loss: 0.4512\nEpoch 47: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8146 - loss: 0.4512 - val_accuracy: 0.5165 - val_loss: 1.3564\nEpoch 48/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8460 - loss: 0.3860\nEpoch 48: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8457 - loss: 0.3865 - val_accuracy: 0.5723 - val_loss: 1.4342\nEpoch 49/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8235 - loss: 0.4180\nEpoch 49: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8238 - loss: 0.4176 - val_accuracy: 0.4438 - val_loss: 2.5214\nEpoch 50/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8465 - loss: 0.3844\nEpoch 50: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8463 - loss: 0.3847 - val_accuracy: 0.5905 - val_loss: 1.3839\nEpoch 51/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8392 - loss: 0.3784\nEpoch 51: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8391 - loss: 0.3785 - val_accuracy: 0.4983 - val_loss: 1.8856\nEpoch 52/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8598 - loss: 0.3584\nEpoch 52: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8597 - loss: 0.3586 - val_accuracy: 0.5656 - val_loss: 1.4128\nEpoch 53/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8737 - loss: 0.3193\nEpoch 53: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8735 - loss: 0.3198 - val_accuracy: 0.5387 - val_loss: 1.7037\nEpoch 54/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8794 - loss: 0.3036\nEpoch 54: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8794 - loss: 0.3037 - val_accuracy: 0.5010 - val_loss: 2.2785\nEpoch 55/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8783 - loss: 0.3137\nEpoch 55: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8782 - loss: 0.3140 - val_accuracy: 0.4391 - val_loss: 3.6478\nEpoch 56/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8692 - loss: 0.3228\nEpoch 56: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8693 - loss: 0.3228 - val_accuracy: 0.4607 - val_loss: 2.8096\nEpoch 57/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8786 - loss: 0.3067\nEpoch 57: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8788 - loss: 0.3064 - val_accuracy: 0.5777 - val_loss: 1.8906\nEpoch 58/75\n\u001b[1m148/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.8950 - loss: 0.2734\nEpoch 58: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8950 - loss: 0.2735 - val_accuracy: 0.5454 - val_loss: 1.9054\nEpoch 59/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.8888 - loss: 0.2805\nEpoch 59: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.8887 - loss: 0.2809 - val_accuracy: 0.4533 - val_loss: 2.2852\nEpoch 60/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9143 - loss: 0.2416\nEpoch 60: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9141 - loss: 0.2417 - val_accuracy: 0.5367 - val_loss: 2.0425\nEpoch 61/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.9099 - loss: 0.2284\nEpoch 61: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9098 - loss: 0.2287 - val_accuracy: 0.5857 - val_loss: 1.9333\nEpoch 62/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9116 - loss: 0.2333\nEpoch 62: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9115 - loss: 0.2335 - val_accuracy: 0.3813 - val_loss: 3.6938\nEpoch 63/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.9173 - loss: 0.2175\nEpoch 63: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9170 - loss: 0.2180 - val_accuracy: 0.5783 - val_loss: 2.1910\nEpoch 64/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9166 - loss: 0.2192\nEpoch 64: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9165 - loss: 0.2194 - val_accuracy: 0.5286 - val_loss: 2.5215\nEpoch 65/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9231 - loss: 0.2050\nEpoch 65: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9229 - loss: 0.2053 - val_accuracy: 0.3853 - val_loss: 4.4304\nEpoch 66/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.9188 - loss: 0.2161\nEpoch 66: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9188 - loss: 0.2162 - val_accuracy: 0.4869 - val_loss: 2.1539\nEpoch 67/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9355 - loss: 0.1746\nEpoch 67: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9354 - loss: 0.1748 - val_accuracy: 0.4795 - val_loss: 2.9539\nEpoch 68/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9357 - loss: 0.1790\nEpoch 68: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9353 - loss: 0.1798 - val_accuracy: 0.4217 - val_loss: 2.9449\nEpoch 69/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9406 - loss: 0.1720\nEpoch 69: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9406 - loss: 0.1719 - val_accuracy: 0.5535 - val_loss: 1.9233\nEpoch 70/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9380 - loss: 0.1754\nEpoch 70: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9379 - loss: 0.1755 - val_accuracy: 0.4613 - val_loss: 2.6149\nEpoch 71/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.9403 - loss: 0.1562\nEpoch 71: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9402 - loss: 0.1564 - val_accuracy: 0.5595 - val_loss: 1.9777\nEpoch 72/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.9463 - loss: 0.1483\nEpoch 72: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9462 - loss: 0.1485 - val_accuracy: 0.5757 - val_loss: 2.1103\nEpoch 73/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9327 - loss: 0.1758\nEpoch 73: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9328 - loss: 0.1755 - val_accuracy: 0.5649 - val_loss: 2.0644\nEpoch 74/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - accuracy: 0.9438 - loss: 0.1502\nEpoch 74: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - accuracy: 0.9438 - loss: 0.1502 - val_accuracy: 0.5306 - val_loss: 2.7951\nEpoch 75/75\n\u001b[1m147/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 26ms/step - accuracy: 0.9416 - loss: 0.1568\nEpoch 75: val_accuracy did not improve from 0.59987\n\u001b[1m149/149\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 27ms/step - accuracy: 0.9415 - loss: 0.1568 - val_accuracy: 0.5266 - val_loss: 2.2239\n","output_type":"stream"}]},{"cell_type":"code","source":"model.save('model.keras')","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:32:31.869841Z","iopub.execute_input":"2024-06-16T04:32:31.870127Z","iopub.status.idle":"2024-06-16T04:32:31.922999Z","shell.execute_reply.started":"2024-06-16T04:32:31.870102Z","shell.execute_reply":"2024-06-16T04:32:31.922317Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"code","source":"from matplotlib import pyplot as plt\n\n# Assuming 'history' is the variable holding the training history object\nN = len(history.history[\"loss\"]) # Dynamically set N to match the number of epochs\n\n# Plot the training loss and accuracy\nplt.style.use(\"ggplot\")\nplt.figure()\nplt.plot(np.arange(0, N), history.history[\"loss\"], label=\"train_loss\")\nplt.plot(np.arange(0, N), history.history[\"val_loss\"], label=\"val_loss\")\nplt.plot(np.arange(0, N), history.history[\"accuracy\"], label=\"train_acc\")\nplt.plot(np.arange(0, N), history.history[\"val_accuracy\"], label=\"val_acc\")\nplt.title(\"Training Loss and Accuracy\")\nplt.xlabel(\"Epoch #\")\nplt.ylabel(\"Loss/Accuracy\")\nplt.legend(loc=\"center right\")\nplt.savefig(\"CNN_Model.png\") # Save the plot as a PNG file\nplt.show() # Display the plot\n","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:32:31.924168Z","iopub.execute_input":"2024-06-16T04:32:31.924804Z","iopub.status.idle":"2024-06-16T04:32:32.397957Z","shell.execute_reply.started":"2024-06-16T04:32:31.924772Z","shell.execute_reply":"2024-06-16T04:32:32.397042Z"},"trusted":true},"execution_count":8,"outputs":[{"output_type":"display_data","data":{"text/plain":"
","image/png":"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"},"metadata":{}}]},{"cell_type":"markdown","source":"### Prediction","metadata":{}},{"cell_type":"code","source":"# X = 32\n# img_size = 128\n# img_single = x_test[X]\n# img_single = cv2.resize(img_single, (img_size, img_size))\n# img_single = (np.expand_dims(img_single, 0))\n# img_single = img_single.reshape(img_single.shape[0],256,256,1)\n\n# predictions_single = model.predict(img_single)\n# print('A.I predicts:',categories[np.argmax(predictions_single)])\n# print(\"Correct prediction for label\",np.argmax(y_test[X]),'is',categories[np.argmax(y_test[X])])\n# plt.imshow(np.squeeze(img_single))\n# plt.grid(False)\n# plt.show()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Confusion Matrix","metadata":{}},{"cell_type":"code","source":"from sklearn.metrics import confusion_matrix\nfrom mlxtend.plotting import plot_confusion_matrix\n\ntest_labels = np.argmax(y_test, axis=1)\npredictions = model.predict(x_test)\npredictions = np.argmax(predictions, axis=-1)\n\ncm = confusion_matrix(test_labels, predictions)\nplt.figure()\nplot_confusion_matrix(cm,figsize=(12,8), hide_ticks=True,cmap=plt.cm.Blues)\nplt.xticks(range(5), ['Normal','Doubtful','Mid','Moderate','Severe'], fontsize=16)\nplt.yticks(range(5), ['Normal','Doubtful','Mid','Moderate','Severe'], fontsize=16)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-06-16T04:32:32.399037Z","iopub.execute_input":"2024-06-16T04:32:32.399319Z","iopub.status.idle":"2024-06-16T04:32:36.567101Z","shell.execute_reply.started":"2024-06-16T04:32:32.399294Z","shell.execute_reply":"2024-06-16T04:32:36.566230Z"},"trusted":true},"execution_count":9,"outputs":[{"name":"stdout","text":"\u001b[1m16/26\u001b[0m \u001b[32m━━━━━━━━━━━━\u001b[0m\u001b[37m━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 3ms/step","output_type":"stream"},{"name":"stderr","text":"W0000 00:00:1718512354.516577 129 graph_launch.cc:671] Fallback to op-by-op mode because memset node breaks graph update\n","output_type":"stream"},{"name":"stdout","text":"\u001b[1m26/26\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 70ms/step\n","output_type":"stream"},{"name":"stderr","text":"W0000 00:00:1718512356.254444 129 graph_launch.cc:671] Fallback to op-by-op mode because memset node breaks graph update\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"
","image/png":"iVBORw0KGgoAAAANSUhEUgAAAwcAAAKxCAYAAAAGvNCTAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB+MklEQVR4nOzdd1jVdf/H8dcBGaIgIIIrEcSJe+DInWmWmZrS1sbvbumtLbWhZaWlLbUby7RMy0rRysw9EkXFDQ6cKbhBHIiyx/n9weHkCXAgcECej+vyujzfdd5ncl7fz/gajEajUQAAAADKPBtrFwAAAACgZCAcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAElSOWsXgDvTkLe+1+HoWGuXgSL2xluDrV0CiknDys7WLgHFyNezorVLQDExGAzWLgHFwCDJ/iZ/9RMOUCQOR8cq4uApa5eBInY+Kc3aJaCYJDlnWbsEFCOjtQsAYDV0KwIAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEBSGQ0HQ4cOVWBgoAIDA7Vly5Z8t/vwww8VGBiokJCQ4iuuhJg2bVqZfewAAABlVZkMB9f65ZdflJmZae0yAAAAAKsr0+HAwcFBZ8+e1dq1a61dCkqgu1vW0cIpL+jYqglKDg/Sg12bWqz3dHfWjPef1LFVE3Rh8xf6I+hl1alVJddx2jb10fJv/qvzmz9XbOinWv3dK3J0sCuuh4ECeLt/R73Y3ifXv18+HWve5tjeXZo87HEN79ZIr9zTRJ+9FKi0lBQrVo3CMPvryWrj66rPP3gz1zqj0ajhzwxUG19XhaxaYoXqUNg2hm7Qw/36yte7hpzsbbT4j0XWLglFbPpX01Tfr7ZcKzqqU4e22r5tm7VLKnHKdDjo3bu3JOnXX39VamqqlatBSVOhvIP2Hj6tVz6en+f64MnPy6emhwa98o3aPTZRJ85e1LLp/5WTo715m7ZNffRH0Mtau+WgOj35qTo++ammz1uvrCxjcT0MFMBbs/7QpCXbzP9GTP1RktTyngckZQeDL199Wg0DOunN7xbpzVmL1PXhwTLYGKxZNm5T5O5d+v2X71W3gX+e63+Z9ZUM4jW+kyQmJqpJ06aaPDXI2qWgGCwInq/RI1/TO2PeU9i2XWratJn6PtBL586ds3ZpJUo5axdgTS1bttThw4e1f/9+LV26VAMGDLjpfTdt2qS1a9cqKipKqampcnV1VePGjdWvXz9Vr1491/ZDhw5VXFycgoKCdPz4cS1btkzR0dFKTEzUe++9J39/fwUGBkqSgoODtWHDBi1fvlynTp2Svb29mjRpoieffFIeHh4yGo1auXKl1q5dq5iYGNnb26tly5Z68sknValSJYv7zcjI0ObNmxUREaFjx47p0qVLysjIkIeHh5o1a6Z+/frJ3d399p7IO9SqTfu1atP+PNf51fJU26Y+avnweB04FiNJGv7RfEWv+UiBvVtp9u9hkqRPXh+gr+aF6LPvV5v3PXKcL6GSztmtssXtlT98rSo1vFWvRVtJ0oKpH6r7oCG6b/BL5m2qetcp1hpRuJISr+rdV/+jtz/6UrOmfZpr/aH9e/TTd9M054916t22vhUqRFHodV9v9bqvt7XLQDH5csoXeua5/2jw089Ikv731XQtX75Uc2bP0shRuVsLy6oy3XIgSU888YQk6Y8//tCVK1duuL3RaFRQUJCmTp2qAwcOyMfHRwEBAbKzs1NISIhGjx6tiIiIfPdfsmSJPv30U6WkpKh58+Zq1KiRbGwsX4aff/5ZX3/9tcqXL68WLVrIwcFBmzdv1tixY3X16lVNmTJFP/74o9zc3NSsWTPZ2Nho/fr1Gj9+vDIyMiyOdfnyZQUFBWnXrl2qUKGCmjVrpsaNGyslJUUrVqzQqFGjFBMTc+tPXBnnYJ+dq1PS/nm+jUaj0tIy1KF59o/EKm4VFdDUR3EXr2rd7NcUveYjrfp2hDo097VKzSiYjPQ0bV25SB36DJLBYFDCxfOKioyQs3tlffKfhzXy/tb6/KVH9Pfu7dYuFbfhk/fe0N3deqptx6651qUkJ2nsK//RqPc/lUcVr+IvDsBtS0tLU/iunep+Tw/zMhsbG3Xv3kPbtoRZsbKSp0y3HEhS3bp1FRAQoG3btum3337TkCFDrrv96tWrtWHDBjk7O2vs2LGqXbu2pOwfhgsWLNDChQs1depUTZ06VS4uLrn2X7VqlUaNGqXWrVvnex9r167Vxx9/bD52Wlqaxo8fr4MHD2rcuHFKTU3VlClTVKVKdv/2hIQEjRkzRsePH1dYWJg6depkPpaTk5NGjRql5s2bq1y5f17ujIwMBQcHa9GiRfr+++/11ltv3exTZiE9PV3p6enm2waDQeXLly/QsUqTQ9ExOnH2oj78b18NG/+LEpPTNPzJbqpZ1U1VPbJbb3xqekiS3nnhfr01+XftOXRKT/QJ0LJv/qtWgz7S0RNx1nwIuEkR61cp+WqC2j8wUJJ0/sxJSdKSb6fq4f++rZp1G2nL8t805b9PauxPK+R1l481y0UBrPrzVx3ct0dz/vgrz/VfjH9bTVsGqMu9DxRzZQAKy/nz55WZmSlPT8uA7+nlpUOHDlqpqpKpzLccSNLjjz8uW1tbrVq1SnFx1//B9ueff0qSBg4caP7xLmX/KB40aJC8vb2VmJioNWvW5Ll/ly5drhsMJCkwMNDi2Pb29urTp48k6cSJE3rmmWfMwUCSXFxc1LNnT0nS3r17LY5Vvnx5tW7d2iIYSFK5cuX0+OOPy83NTREREUpOTr5uTfn5/fff9fTTT5v/jRs3rkDHKW0yMrL06Osz5eftqbMbPtXFsC/UuXU9rdgYqSxjliTJxtT//LtfN+rHxVu0+9Apjfr8Nx2OPqchD7W3Zvm4BZuXBMu/XRe5ms4YG7OyX99O/R5Xhz6DVKu+vwJfGSuvWj7a/OcCa5aKAog5c0qff/CmPpw8Qw4OjrnWr1+zTDs2b9BrYz+2QnUAUPzKfMuBJFWvXl3dunXTmjVrNH/+fA0bNizP7S5cuKDY2FhJ2T/y/81gMKhr166aM2eOIiMj8xzD0K5duxvW06JFi1zLqlatKkmytbVVs2bN8l1/6dKlPI8ZHR2tffv26dy5c0pJSZHRmD0gNisrS0ajUTExMfLxufUznv379zcHFyn7OSgrwg+cVLtHJ8qloqPs7crp/KWr2vDDG9q5/4Qk6WxcgiSZxyTkOBQVo7uquhV7vbh1F86e0oHtm/TCx1+bl1Xy8JQkVfPxs9i2am0/XYw9U6z14fYd3Behixfi9FTff77TMzMzFb5tsxb8OFMPP/GsTp2IUvfm3hb7jX55sJq3aa9vflla3CUDKAAPDw/Z2trq3LlYi+XnYmPNv6GQjXBgMmjQIIWGhio0NFQPPvigvL29c21z8eJFSZKzs7OcnJzyPI6Xl5fFtv927Rn//Hh4eORa5uiYfUbL1dVVtra2udbndOW5touPJKWkpCgoKEjbbjBVV0FbDuzs7GRnV7an5Uy4mj19ZZ1aVdSyUS29/1X2FIfHz1zQmXPxqlfb02J7P2/PfAc6o2TZvHShnN0qq0mH7uZllavVVCUPL8UeP2ax7bkTUfJv37WYK8TtatOhi35Zvtli2Qejhqp2nboa/MIrcnWvrP6PPWOx/rHeHfTqmI/U6Z77irNUALfB3t5eLVq20rq/1qrvQ/0kZZ8gXbdurV58Oe+TwmUV4cDEzc1NvXv31qJFi/Tzzz8XuA/+jdjb299wm38PUL7ZdXn5+eeftW3bNtWoUUOPP/646tSpIxcXF3M3ozFjxujw4cPmlgT8o0J5e9W5658wV7tGZTWtV0OXEpJ0MuaSBvRoobhLV3Uy5qIa162uz0YO1J8he7R2yz99FyfPWaMxLz6gvYdPa/ehU3rywbaqX9tLj4/8zhoPCbcgKytLYUsXqP39D8v2mm55BoNBPZ94Xn9+O0U16jbUXXUbacuyXxVz/Kie/+grK1aMgqhQ0Vl+9RtZLCvv5KRKru7m5XkNQq5avaZq3FW7OEpEEbp69aqO/v23+fbx6CjtjoiQu7u77qpVy4qVoSgMf+U1/efZIWrVqrVatwlQ0JdTlJSYqMFDnrnxzmUI4eAa/fr109q1axUeHq79+3Of2c2Z8vPKlStKSkrKs/UgZ67ckjI9aFhY9gj8V155Jc/WkLNnzxZ3SaVGy0beWvXtCPPtT954WJL04+Itev69uapaxUWTXh8gz8rOijmfoJ+WbNXHM1ZYHCPo5xA5Otjpk9cfllslJ+09fFp9XgpS1KnzxfpYcOsObt+oizFn1KHPoFzr7nn0WaWnpWrh1PFKTIhXTb+GGvHlj6pSM/dnDEDJtWvnDt137z8tg6NHvi5JevKpIZrx3ffWKgtFZFDgIzofF6cP3n9XsTExatqsuf5YssLc6wPZCAfXcHJyUv/+/fXDDz/op59+MnflyVG5cmV5eXkpNjZWISEhuv/++y3WG41GhYSESJL8/fO+iE5xu3r1qqS8uzNFRETc1PStZVXoziMq3yL/psavflmvr35Zf8PjfPb9aovrHKB0aNS2s6aHReW7/r7BL1lc5wB3jhuNI9h+LL54CkGR69ylq5LSsqxdBorRS0OH6aWhdCO6HmYr+pdevXrJw8NDR44c0eHDh3Otf/DBByVlX1U5OjravNxoNJqXVahQQT169Mi1rzXUrFlTkrR8+XKL5WfOnNHMmTOtURIAAABKKFoO/sXOzk6BgYH66quvlJqammv9vffeq8OHD2vDhg1666231KhRI7m4uCgqKkpnzpyRvb29hg8fnuc1Dqxh4MCB+uKLLzR//nyFhYWpZs2aSkhI0IEDB9SwYUO5u7vr0KFD1i4TAAAAJQAtB3no3LmzauUzEMlgMGjYsGEaPny4GjRooGPHjmnLli1KTU1V165d9cknn+Q5Fam1tG3bVuPGjVOTJk106dIl7dy5U5cvX9agQYP09ttv5znzEQAAAMomg5FpalAE2j82UREHT1m7DBSxKV+/Ye0SUExaeXFtjrLEv2bJaP1G0StL1ycqywySHG6yvxAtBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwMRiNRqO1i8CdJ+zYJV1JybB2GShiLvZ21i4BxcTBjnNJZUndqhWtXQKKSTlbPttlgUGSQ7mb25Z3BAAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJEnlrF1AQQwdOlRxcXHm2waDQQ4ODnJyclK1atXk4+OjDh06yM/Pz4pVZjt37pyGDRumKlWqaNq0aYV+fKPRqD///FPr169XTEyM0tPTJUnBwcEFPmZR1wwAAICSqVSGgxz169dX1apVJUlpaWm6cuWKoqKiFBkZqSVLlqhRo0Z66aWX5OXlZeVKCyYnBAUFBcnT0zPPbVatWqW5c+fKyclJLVq0UPny5Yu5SgAAANwpSnW3onvuuUdDhw7V0KFD9eqrr+rdd9/VrFmz9Oabb6patWrav3+/xowZo3Pnzlm71CITFhYmSXrttdf0xhtvmJ8PFL4fp32iXv6eFv+e69NBkhRz+kSudTn/NqxcbOXKcbt++Gay2td10+Txb5mXXYiL1ftvvKAH2tdXt6Y1NOShLlq3gte6tPv+qy/UsnYlffr+m+ZlqSkp+njs6+rWvLbublRdb7z4pC7E3bl/V8qyzz+dJGdHW41+41Vrl4IiMv2raarvV1uuFR3VqUNbbd+2zdollTilOhzkxWAwqGXLlvroo49UrVo1Xb58WdOnT7d2WUXm/PnzkmRuQUHR8vZroF9C9pr/ffHjn5KkKlVrWCz/JWSvnho6SuWdKqhNx+5Wrhq3Y/+eXVo0b7b8GvhbLP9g5Es6HvW3Ppn+s+Yu2aSuPR/UmBHP6FDkHitVitsVuXunfv35e9Vt0Nhi+ecfvqXQtSs06as5mjl/qeJiY/TGi09aqUoUlZ07tuv7b2eocZOm1i4FRWRB8HyNHvma3hnznsK27VLTps3U94Fed/RJ5IK448JBjgoVKmjIkCGSpH379unYsWMW669evaqff/5Zr732mp588kkNHjxYo0eP1h9//KG0tLRcxwsJCVFgYGC+ffDPnTunwMDA6561z8zM1B9//KHXXntNTzzxhJ577jl98cUXOn36dJ73lTOuYtiwYQoMDDT/i4yM1Lhx4xQYGGh+Q1+7Tc54g+DgYIvb/xYZGanAwECNGzcu35phydbWVu5VvMz/KrlVznO5exUvbV67TJ3ve0jlK1S0ctUoqKTEqxr3+vN6c/xUObu4WqzbG75Ng576j/ybtVKNWrX1zNA3VNGlkg5FRlilVtyepMSreueV/2jsxC/lUsnVvPxKwmUtCv5Rr42ZoIAOXdSoSQuN+/Qr7d65VXt2bbdewShUV69e1XNPP6X/ffWNXF3drF0OisiXU77QM8/9R4OffkYNGzXS/76arvJOTpoze5a1SytR7thwIEktWrRQxYrZP8z27PnnbF5sbKxGjx6tRYsWKSEhQS1atFDjxo0VExOjn376SWPHjtXVq1cLvZ7Jkydr3rx5cnNzU5s2beTk5KQtW7borbfe0uHDh83bVa1aVV26dJGDg4MkqW3bturSpYv5n6urq5o3b57vNrVr1y702pHt9IkoPda1iYb0aq2Jo17UuTOn8tzuSORuHT24T70GPF7MFaIwffb+SHXo2lMBd3fNta5JiwCtWfq7LsdfUlZWllYv+VVpqalq0bZj8ReK2zZx7Bvq2K2X2nbsZrH8wL4IZaSnq+017wEfv3qqWuMu7dlFd4Q7xWsjhum+3ver2z09rF0KikhaWprCd+1U92teYxsbG3Xv3kPbtoRZsbKSp1QPSL4Rg8EgHx8f7d27VydPnjQv//LLLxUXF6fWrVtr+PDhcnR0lCQlJCRowoQJioqK0qxZszR8+PBCqyUuLk4pKSmaOHGivL29JUlZWVmaPXu2VqxYoalTp2rKlCmys7NTgwYN1KBBA+3fv19xcXF66qmncg1IrlGjhiRddxsUrgZNW+mNCV+qZu06uhgXq7lff6bXB/fVN39skNO/WgdW/PqTavnWk3+LACtVi9u1esmvOhS5W7N++yvP9eO//F5jRzyr+9r4yrZcOTk6ltfEaT/qLm/fYq4Ut2vl4oU6GLlbP/6xLte6C3HnZGdvL+drWhMkqbJHFV2Iiy2mClGUFgbP0+6IcK3ftNXapaAInT9/XpmZmfL0tJykxtPLS4cOHbRSVSXTHd1yIEnOzs6SZG4JOHjwoI4cOSIHBwc9//zz5mAgSS4uLnrhhRckSZs2bdKFCxcKtZaHH37YHAyk7MT61FNPyd3dXXFxcdq6tfR9MaWnpyspKcn8Lzk52dolFZk2ne5R51595VvfX607dtf4r3/R1SuXtWHFHxbbpaYka92y39TrYVoNSqvYs6c0efxbev/zGXJwcMxzmxlTJuhKwmV9OWeRvv/tLz327FCNGfGM/j4UWczV4nbEnDmlTz94U+OnzJSDY96vNe5cp06e1Kg3XtV3s3+0+D0AlGV3dMuBlH0dgGtFRmb/4W7WrJlcXV1zbe/r6ytvb28dP35c+/fvV6dOnQqtli5duuRaZmdnpw4dOmjJkiWKjIxUx46lq0vC77//roULF5pv+/j4aNKkSVasqPhUdKmkmt51dOZElMXy0FV/KjU5WT36BlqpMtyug/t269KFOD3dr6t5WWZmpiK2b9avc2dq3srtWvjjTP20bLN86zaUJNVt2EQRO8L069xvNfrDyVaqHLfqwN4IXTwfpyf6dDYvy8zM1K5tmxT8wwwF/fCb0tPSdOVyvEXrwYXzcapcpXROk41/hIfvVNy5c+rYrrV5WWZmpjZt3KBvvp6mCwnJsrW1tWKFKCweHh6ytbXVuXOWLX7nYmOZ1OVf7vhwcOXKFUkyjz24ePGiJF23C46Xl5eOHz9u3rYwVKhQQRUqVMhzXU4thXl/xaV///7q06eP+bbBYLBiNcUrOfGqzpyM1j19B1ksX/nbz2rXrZdc3T2sVBluV+v2nTV36SaLZRPeHCZv37p68vkRSklJkiTZGCwbX21tbHOdkEDJFnB3FwWvtOxvPG7ky6pdp56efvEVeVWroXJ2dtq2eb3u6f2QJCn66BHFnD6ppi3pNljade12j7bu3G2x7KXnn1O9evX16hujCAZ3EHt7e7Vo2Urr/lqrvg/1k5TdvXvdurV68eVh1i2uhLmjw4HRaFRUVPZZ3Vq1ahX5fZWEYxT3/dnZ2cnOzq4Qqin5Znz6ntp17SXP6jV14VyMfpz2iWxtbdX1/v7mbU4fP6a9O8L04de/WLFS3K4KFZ1Vp14ji2WO5Z3k4uquOvUaKSM9XTW9fTVp7Ksa9uaHquTqrg1rlmrbpnX6bMY8K1WNgqhQ0Vl+9S1f6/LlK6iSq7t5eb/Ap/T5+HfkUslNFZyd9cl7o9S0ZYCatmxjjZJRiJydndXI33LqWienCnKvXDnXcpR+w195Tf95dohatWqt1m0CFPTlFCUlJmrwkGesXVqJckeHg/DwcCUmJkrK7kYkSe7u7pJ03Tltc9blbCtJ5cplP1X59anPmXY0P4mJiUpMTMyz9SCv+ysMOTWnpKTkuf5GNcPS+diz+njkC7oSf0mV3CvLv2VbTfl5mUULwcrff5GHV3W1ymN2G9w5ytnZ6Ytvg/XVp+9r5AuPKTkpUTW9fTT2k6/UoWtPa5eHQvb62I9lsLHRyJeeUlpamtp37q63PvzC2mUBuEWDAh/R+bg4ffD+u4qNiVHTZs31x5IV8vKii+C17thwkJSUpDlz5kiSmjZtap7e098/+0JGERERio+PzzXuICoqStHR0TIYDGrYsKF5ec4P9zNnzuR5f7t27bphTRs2bFDv3r0tlmVkZJivcpxTW46cH/dZWVk3PHZecmr+93UUctxMzfjH25/NuOE2z77yjp595Z1iqAbF7auflljcvqt2HX087QcrVYOiNHP+UovbDo6OeuvDz/XWh59bqSIUp+Wr856hDHeGl4YO00tD6UZ0PXfcbEVGo1Hh4eF66623dPbsWbm5uZlnIJKkBg0aqG7dukpLS9OMGTOUmppqXpeQkKAZM7J/AN59993y8PjnjLCfn5/Kly+vU6dOacOGDRb3GRYWpuXLl9+wtl9//VUnTpww387KytLcuXN14cIFVa5cWW3btrXYPufH/bXTsN6Kxo0by2AwKCIiQvv37zcvNxqNWrZsWamcHQkAAABFp1S3HKxdu9Y8+1B6erquXLmiqKgo87Sl/v7+eumll1SlShWL/YYPH64PPvhAO3bs0LBhw9SwYUNlZGQoMjJSycnJ8vHx0bPPPmuxj729vQIDAzVnzhwFBQVp1apVcnd31+nTp3Xq1CkNGDBAv/76a761enh4yNfXV6NHj5a/v78qVqyoo0ePKjY2Vg4ODhoxYoTs7e0t9mnbtq0iIyP1v//9T02bNjUPqu7bt6+qV69+w+fHw8NDvXv31rJly/TBBx+oYcOGqlChgo4fP67z58+rX79+WrRo0Q2PAwAAgLKhVIeDQ4cO6dChQ5IkBwcHOTk5qVatWvL19VWHDh3k5+eX535eXl6aNGmSFi9erO3bt2vnzp2ysbFR9erV1b59e91///25fqhL0gMPPKCKFStq2bJlioqK0smTJ1WnTh0NGTJEVatWvW44MBgMevXVV/XHH38oNDRUBw4ckIODg9q2batHHnlENWvWzLVPz549lZKSotDQUIWHhys9PV2S1KlTp5sKB5I0ZMgQeXh4aO3atTp06JAcHR1Vv359vfrqq0pOTiYcAAAAwMxgZN49FIGwY5d0JSXD2mWgiLnYl42ZqiA52N1xvVBxHXWrVrzxRrgjlLPls10WGCQ53GSTAO8IAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJiUs3YBuDPVqOSo1ApGa5eBInY47oq1S0Ax+XXHOWuXgGL0bo+61i4BxcTTxcHaJaAYGAySZLipbWk5AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAm5W5mo0ceeeSWD2wwGDRv3rxb3g8AAACAddxUOHj44YdlMBiKuhYAAAAAVnRT4SAwMLCo6wAAAABgZYw5AAAAACDpJlsO8nL+/Hn99ttvioyMVEJCgkaOHKlGjRopISFBCxcuVLdu3eTj41OYtQIAAAAoQgVqOTh16pRGjRqlsLAweXp6KikpSVlZWZIkFxcXHTp0SCtWrCjUQgEAAAAUrQKFg7lz56pChQqaOnWq/vvf/+Za36JFCx08ePC2iwMAAABQfAoUDg4cOKB7771XLi4uec5i5OHhoYsXL952cQAAAACKT4HCQVZWlhwcHPJdn5CQoHLlCjycAQAAAIAVFCgc+Pr6ateuXXmuy8zM1ObNm1WvXr3bKgwAAABA8SpQOOjXr58iIiI0c+ZMnTx5UpIUHx+vPXv2aPz48Tp9+rQeeuihQi0UAAAAQNEyGI1GY0F23LBhg77//nslJSVZLC9fvrz+7//+Tx07diyUAlE6nbiQrNSMAr21UIocjrti7RJQTH7dd87aJaAYvdujrrVLQDHxdMm/mzjuHAaDVN4u9zjhvBR4YEDnzp0VEBCgPXv2KCYmRllZWapataqaNWum8uXLF/SwAAAAAKzktkYNOzo6KiAgoLBqAQAAAGBFtxUOdu7cqfDwcMXFxUmSqlSpohYtWqhVq1aFUhwAAACA4lOgcJCYmKjPPvtM+/fvl42Njdzc3CRJe/bs0erVq9WwYUONHDlSFSpUKNRiAQAAABSdAoWD77//XgcOHNATTzyhnj17ytHRUZKUkpKiVatW6eeff9b333+vYcOGFWqxAAAAAIpOgcLB9u3b1bNnT/Xt29diuaOjo/r27avz589r/fr1hVIgAAAAgOJRoOsclCtXTtWrV893ffXq1blCMgAAAFDKFCgctG3bVlu2bFFWVlaudZmZmQoLC1O7du1uuzgAAAAAxeemTu8fO3bM4nanTp00a9YsjRkzRj169FDVqlUlSWfPntWaNWuUkZGhTp06FX61AAAAAIrMTV0h+ZFHHinQwefPn1+g/VD6cYXksoErJJcdXCG5bOEKyWUHV0guGwr9CskvvfTSbRUEAAAAoOS7qXDQtWvXIi4DAAAAgLUVaEAyAAAAgDtPgecbTUtL09atWxUVFaWkpKRcMxcZDAa6IwEAAAClSIHCQVxcnN5//33FxcXJyclJSUlJqlixojkkODs7m6+aDAAAAKB0KFC3oh9//FFJSUmaMGGCpk6dKkl69dVX9cMPP+iJJ56Qvb293nnnnUItFAAAAEDRKlA4iIyMVM+ePeXn5ycbm+xDGI1G2dnZqW/fvmrcuLFmz55dmHUCAAAAKGIFCgepqany9PSUJJUvX16SlJSUZF5fr149HTx4sBDKK7uGDh2qwMBABQYG6vvvv7/utosXLzZv++ijj1qsGzdunAIDAxUZGXlL9x8cHKzAwEAFBwffcu0AAAAonQoUDjw8PHThwgVJkq2trdzd3XXkyBHz+lOnTsne3r5wKoQ2btyojIyMfNevW7euGKsBAADAnapA4aBx48basWOH+XbXrl21dOlSTZ8+XV9//bVWrlypVq1aFVqRZVmdOnV05coVbd++Pc/1hw4d0unTp1WnTp081w8bNkyTJ0+Wn59fUZZZJkz/8lMN6NVJLep4qZ2/t156+hEd+/uwxTYnoo/p5WceVdtG3mrhV1Uj/vOUzsfFWqli3I7zsWf12ZtD9VjHhhrQuraG9u+qI5ER5vVGo1FzgybpqW5NNaB1bb3zf4N0+vgxq9WLm1OvipOGd/LWFw810KxHm6hFDZdc2/Rr7KkvHmqg6QP99UZXH3lWzPtkVzkbg8b18tOsR5voLlcm4SjpvpryqR669241rl1FrRvW0vODB+nov77D3359mLq0aaQGd7mpVYO79J+nBunokUNWqhiF6dNPPlanDgHyquwi75peemRgfx0+xGublwKFg379+ql///5KT0+XJPXv319dunTR1q1btWPHDnXs2FGDBw8u1ELLqm7duknKv3Xgr7/+stju3zw8PFSjRg05OHB59Nu1PWyjnnzmeQUvXafvg/9URnq6nn2kr5ISEyVJSYmJeuaRvjIYDPrh16Wa9+capaWn6YWnBuWa6hcl29XL8Ro1+EGVK1dO477+SV8tWq/nRo5TRRdX8za/zgrSnz9/p6FjP9HnPy2TY3knvfvCo0pLTbFe4bghh3I2Ohmfork7zuS5vncDD/Wo56EfdpzW+NVHlZqRpde7+qicjSHXtoOaVVV8cv6tuihZtm4O1VPPvqjfVqzXDwuWKCM9Q4MH9TF/h0tS42Yt9MnUGVqzKUJz5i+W0WjU4EF9lJmZacXKURg2btig5198WetCw/TnslVKT09X3z69lHjN649sBZrK1MPDQx4eHubb9vb2evHFF/Xiiy8WWmHIVqtWLdWpU0e7d+/WxYsX5e7ubl6XkpKisLAwVa5cWc2aNctz/3Hjxmn//v1677335O/vb7EuLS1NixYt0saNG3X+/HlVrFhRTZs2zTVuAdm+++UPi9uTpn6jdo1rK3JPuNq076hd28N0+uRx/bFmsyo6Z5+N/OTLGWpdv4bCNobo7s7drVE2CmDhrCB5VK2hV8ZPNS+rWtPb/H+j0ag/5s7UI8+/onbd75MkvfbR//Rk1yYK+2uFuvTuV9wl4ybtPXtVe89ezXf9vfU99GfkOUWcviJJ+nbrSU3p11Ata7po24nL5u2aVKso/6oV9dWmE2pa3bnI68btmxO82OL2p/+bodYNa2nv7nC17dBRkvT44OfM62vW8tbrb72n+7sG6NSJ4/L28S3WelG4/liy3OL2N99+r9o1vRS+a6c6dupspapKJq6QXAp069ZNRqNRISEhFss3b96slJQUdenSRQZD7rNa15Oamqr3339fCxcu1KVLl9S0aVM1aNBAu3fv1ujRo3Xu3LlCfAR3pitXEiRJlVzdJGWHLYPBIHv7f1ppHBwcZWNjo51bw6xSIwpma8hK1W3UTB+/9n96oou/hg/qoRUL55rXx546oUvnz6l5u3/+oFRwdlH9Ji10cPeOvA6JUqBKBTu5lrfT/th/wkNyepaOXUhSncpO5mUuDuU0pE1NfbvllFIzaRUsra4kZH+Hu7q55bk+KTFRC3/5QXd511a1GjWLszQUg4TL2WHf7ZqTrsh2Uy0HCxcuLNDBBw4cWKD9YKljx4764YcfFBISogEDBpiXr1u3TgaDId8uRdcTHBysI0eOqEaNGho7dqy5RSI1NVVTp07Vhg0bCq3+O1FWVpYmjB2llgHtVa9hdotM85ZtVN6pgj4dP0avvfW+jEajPpvwrjIzMxV3LsbKFeNWxJw6oWXBc9Rv8AsK/M8IHdkXoRkTx8jOzk73PPSILl3IDs+ulatY7OdauYrizxOsSysXRztJUkKKZVehhJQMVSr/z5/L59rWVMjfFxR9KVmVK9gVa40oHFlZWfpwzEi1Dmiv+g0tW9V/nPWNJr7/jpKSEuXrV08/LljKJCt3mKysLI1641W173C3/P0bW7ucEuemwsGCBQsKdHDCQeFwcnJSQECANm7cqP3796tRo0Y6c+aMDh06JH9/f3l5ed3Smf60tDStWbNGkjRkyBCLrkoODg56/vnntXv3bvOYkutJT0+32M5gMJint72Tvf/mqzpycL9+WbzGvMzdo4q+nPmj3hv9in749mvZ2Njogf6D5N+0uWwMNNKVJsasLPn5N9OQEW9Lkuo0bKLjfx/UsuAfdM9Dj1i5OlhTj7qV5Whno6UH4qxdCm7Du6Nf0aGDkVqwZG2udQ8NfFQdu9yjc7ExmvnVFA37vye1cOlfcnBk0Pmd4tXhQ7V//z6t+SvU2qWUSDcVDubPn1/UdeAGunfvro0bN+qvv/5So0aNbjgQ+XqOHTum5ORkOTs7q3nz5rnWu7q6qlmzZhYzUuXn999/t2hZ8vHx0aRJk265ptLk/bde07o1y/XT76tUtXoNi3Udu/bQ2q37dPHCeZUrV04ulVzVoYmP7nqotnWKRYG4VfFUrTr1LJbd5VtXm9YszV5fOfs6L/EX4uRexcu8TfyFOPk04CxUaZWQkn2iw8WxnC5f03rg4lhOJy5lDzRv4FVBdSo7acYgy9f53Z5+2nI8Xt9tPVV8BaNA3h39iv5atUzzF69Rteq5uwu5uFSSi0sl+dTxU4vWAWpet5pWLvtDfQdwYuBO8NqIYVq+fKlWrVmvGjXpLpaXAg1IRvHz9/eXp6entm7dqqefflobNmxQ+fLl1a5du1s+1sWLFyXJfCG7vFxv3bX69++vPn36mG/f6tiH0sRoNOqDt1/X6uWLNfe3FbrLu3a+27pXzh6wH7YxRBfOx6l7rweKqUoUhkbNA3Qq+qjFstPRx+RZLfsPiVfNWnLz8FTE1lD5msJA0tUrOrQ3XL0febq4y0UhiUtMV3xyuhp5VdTJ+Oww4FjORr6VnbTu7+zvzZ93ndXve/+Znti1vJ1e7+qj6ZtP6NiFpDyPi5LBaDTqvTdf1apli/XLolXX/Q6/dh+j0ai01LSiLxBFymg06vVX/qvFixdpxap1qu3jY+2SSizCQSlhMBjUtWtXBQcHa9q0aYqPj1ePHj2s3g/Szs5OdnZlo8/t+2++qj9/D9bXs+erQsWK5nEEzs6V5GjqSvXrLz+oTr0Gcq/sofAdWzVh7Cg9/fww+frVu96hUcI8NPh5jXzqQQXPnKqOvfrq8N5wrfj1Rw179zNJ2Z/Hh578j+Z/M0U1avnKq0YtzQ2aJPcqXmpvmr0IJZNDORuL6xZ4VLDTXa6OSkzL1MWkdK0+dF59/D0VeyVVcYlp6t/ES/HJGdp1Knvw6sUky+6WKRnZA5LPXU3TJaY1LdHeHf2K/vh1vmb8sEAVK1ZUXKzpO9wl+zv8RHSUlixaqE7d7pF7ZQ/FnDmtr7/8XI6O5dW1Ry8rV4/b9erwoQqe/4vmL1ykis7OionJfv0rVapUJrpD3wrCQSnStWtXLViwQDt37pRUsC5FksxjDK43ToHZinL7ec5MSdKTAyx//E2cMl0DHn1KknTs6BF9/tF7uhx/STXu8taLI0bqmRf+W+y14vbUa9xC70yZpTlTPtIv07+QV41a+s+oD9Wtz8PmbR5+dphSkpP0v/ffUOKVBDVqEaAPpv8iewf6JZdktd3La3T3f6akfKxldUnSxqhLmrX1lJYfPC+HcjYa0qaGnOxtdSQuSV+sj1JGltFaJaOQzP1+hiTpsX49LZZ/+uUMDXzsKTk4Omj7lk2aNSNICfGX5FHFUwHtO2rhsnXyqHJzrekouWbOmC5Juu9ey99O02fO0lODn7ZCRSUX4aAU8fDwUJs2bXTgwAFVrVpVdevWLdBxfH195ejoqCtXrmj37t25rpEQHx+vPXv2FEbJd5TDMTe+UMrIMR9q5JgPi6EaFLWALj0V0KVnvusNBoOeHDZaTw4bXYxV4XYdOpeoZ+ftve42i/ad06J9N3eC5EJi+g2Ph5IhKi75uuu9qlbX9/MWFU8xKHaJqUw7fLMIB6XMG2+8cdvHsLe3V48ePbRkyRLNmTNHY8eOlZvbP3P1f/vtt0pLo38lAABAWUM4KKMCAwN18OBB/f333xoxYoT8/f1lZ2engwcPKjMzU507d+ZaBwAAAGXMbU++funSJUVHRyslJaUw6kExcXR01HvvvaeHH35YlSpV0u7du3XgwAE1btxYH3/88U3PVgQAAIA7h8FoNBZolNX27dv1008/6ezZs5KksWPHqnHjxkpISND48eM1cOBABQQEFGqxKD1OXEhWagYD+O50h+OuWLsEFJNfb7IPPu4M7/Yo2Jg2lD6eLg7WLgHFwGCQytvd3HTzBWo52LFjhz777DM5Oztr0KBBFutcXFzk7u6ukJCQghwaAAAAgJUUKBz8+uuvatSokT788EP16pV77t969eopKirqtosDAAAAUHwKFA5OnDih9u3b57u+UqVKSkhIKHBRAAAAAIpfgcKBg4PDdQcgx8bGqmLFigUuCgAAAEDxK1A48Pf31/r165WZmZlrXXx8vNauXZvrwloAAAAASrYChYPHHntMFy9e1FtvvaXVq1dLkiIiIjRv3jy9/vrrkqSBAwcWXpUAAAAAilyBpzI9efKkZs+erX379lksb9SokZ577jnVrFmzUApE6cRUpmUDU5mWHUxlWrYwlWnZwVSmZcOtTGVa4Csk33XXXRo7dqyuXr2qmJgYGY1GeXl5ycXFpaCHBAAAAGBFBQ4HOSpWrCg/P7/CqAUAAACAFRUoHKxfv/6mtuvSpUtBDg8AAADACgoUDr766qub2o5wAAAAAJQeBQoHQUFBuZZlZWUpLi5OK1eu1Pnz5zV06NDbLg4AAABA8SnQVKZVqlTJ9c/Ly0uNGzfW66+/LhcXF61YsaKwawUAAABQhAoUDm6kVatWCgsLK4pDAwAAACgiRRIOYmJilJ6eXhSHBgAAAFBECjTmYP/+/XkuT0pK0v79+7V8+XK1adPmtgoDAAAAULwKFA7ef//9fNfZ2NioXbt2evbZZwtcFAAAAIDiV6Bw8N577+W5vGLFivLw8JCTk9NtFQUAAACg+N1yOEhPT1dSUpKqVKkib2/voqgJAAAAgBXc8oDkcuXK6YsvvtChQ4eKoh4AAAAAVnLL4cBgMKhatWq6cuVKUdQDAAAAwEoKNJVp//79tWLFCp05c6aw6wEAAABgJQUakHz48GE5Ozvr9ddfV6NGjVSlShXZ29tbbGMwGPTMM88USpEAAAAAil6BwsHKlSvN/9+3b1++2xEOAAAAgNKjQOFg/vz5hV0HAAAAACsr0JiD8+fPKy0tLd/1aWlpOn/+fIGLAgAAAFD8ChQOhg4dqm3btuW7fseOHRo6dGiBiwIAAABQ/AoUDm4kIyNDNjZFcmgAAAAAReSmxxwkJSUpKSnJfPvKlSt5dh1KTEzU5s2b5erqWigFAgAAACgeNx0Oli5dqoULF5pvz549W7Nnz853+0ceeeS2CgMAAABQvG46HDRr1kyOjo4yGo366aefdPfdd8vHx8diG4PBIAcHB/n6+qpOnTqFXiwAAACAonPT4aBevXqqV6+eJCk1NVVt27ZVrVq1iqwwAAAAAMWrQNc5GDRoUGHXAQAAAMDKmFIIAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJksFoNBqtXQTuPKkZEm+sO19yWqa1S0AxyeJPRZny7bbj1i4BxWTY3b7WLgHFwMYglbcz3Ny2RVwLAAAAgFKCcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAk3LWLuDfhg4dqri4OElS79699cwzz+S77eLFizV37lxJko2NjebNm1ektZ07d07Dhg1TlSpVNG3atCK9LwAAAKC4leiWg40bNyojIyPf9evWrSvGakqX4OBgBQYGKjg42NqlAAAAoJQoseGgTp06unLlirZv357n+kOHDun06dOqU6dOMVcGWJr+1TTV96st14qO6tShrbZv22btklAErly5ordGvqamDXxVvXJF9ereUbt25v39hNJl88ZQPT6on/z9asmjop2W/fmHxXqj0aiPPxynRnXuUk0PZw3o00tH/z5ipWpRUFmZmVo1a7ImPd5VY+7z1ydPdNPaH4NkNBrz3P73yWP1Znc/bVz4fTFXiqIw85uv1bZVM1XzqKRqHpXUvXMHrVqx3NpllUglNhx069ZNUv6tA3/99ZfFdoA1LAier9EjX9M7Y95T2LZdatq0mfo+0Evnzp2zdmkoZCOGPq+QdWs0/dvZ2rgtQt3uuVf9+/TSmTOnrV0ablNSUqIaN26qT774Ms/1/5v8mWZOD9JnU6dpZcgmOVWooMB+DyglJaWYK8XtWD/vG21Z/LMeGv6eXpu9Ur2fH6X182Zq8+8/5Np2X+gqndgfIZfKXlaoFEWhRo2a+mD8xwoN26ENm7erc9duemRgP+3fH2nt0kqcEjfmIEetWrVUp04d7d69WxcvXpS7u7t5XUpKisLCwlS5cmU1a9Ys32NcvXpVixcv1o4dO3Tu3DnZ2NioWrVq6tChg3r37i17e/s899u5c6cWL16sqKgoGQwGeXt768EHH5S3t/d1a7569aqWLVum7du3KzY2VllZWfLy8lKHDh3Up08fOTg4WGwfHByshQsXauDAgerevbsWLFigPXv2KD4+Xh07dtTQoUMlSVu3blV4eLiOHDmiixcvKi0tTa6urvL391e/fv1UvXp1i+MGBgaa/79w4UItXLjQfLtLly7m40pSZmamQkJCFBoaquPHjys1NVVubm5q3ry5+vfvLw8Pj+s+5rLuyylf6Jnn/qPBT2ePjfnfV9O1fPlSzZk9SyNHvWnl6lBYkpOT9eei3/RT8G/q0LGzJOnNd97TymVL9f3M6XrnvQ+tXCFuR4+e96lHz/vyXGc0GjV92pd6bdTbur9PX0nSVzO+V0PfGlr25x8aMOiR4iwVt+F4ZLga3X2PGrTLPqnoXrWmIv5aopMHd1tsdzkuRov/976em/S9vn/7P9YoFUXg/j4PWtwe98EEfTdjurZv3aJGjfytVFXJVGJbDqTsVgGj0aiQkBCL5Zs3b1ZKSoq6dOkig8GQ576xsbEaPXq0Fi1apISEBLVo0UKNGzdWTEyMfvrpJ40dO1ZXr17Ntd+SJUs0adIkHThwQDVr1lTLli2Vnp6uTz/9VCtWrMi31lOnTmnkyJFauHChEhIS1KBBAzVp0kQJCQmaP3++xo4dq6SkpDz3jYmJ0ahRoxQeHi4/Pz+1atVKzs7O5vWTJ0/Wpk2bZG9vL39/fzVr1kw2NjYKCQnR6NGjdejQIYvjdenSxRxkvL291aVLF/O/Bg0amLdLTk7W+PHj9c033+jYsWPy9vZW69atZWdnp9WrV2v06NGKiorK9zGXdWlpaQrftVPd7+lhXmZjY6Pu3Xto25YwK1aGwpaRkaHMzEw5ODhaLHcs76gtYZusVBWKw/HoKJ2LjVGXbt3Ny1wqVVLL1gHasW2LFSvDrfL2b6G/d4Up7mT237UzRw/o+L4dqh/QxbxNVlaW5n/8hjo/8h95+dSzVqkoYpmZmVoQPE+JiYkKaNfe2uWUOCW25UCSOnbsqB9++EEhISEaMGCAefm6detkMBiu26Xoyy+/VFxcnFq3bq3hw4fL0TH7j3pCQoImTJigqKgozZo1S8OHDzfvc/z4cc2dO1cGg0Gvvvqq2rVrZ14XGhqqoKCgPO8rLS1NkyZN0oULFzRgwAANHDhQ5cplP7WpqamaPn26Nm3apNmzZ+vll1/Otf/GjRvVqVMnvfjii7Kzs8u1fvjw4WrZsqX5MUjZZ7NWrVql7777TjNmzNBnn31mDkpDhw5VcHCwjh8/rjZt2li0JFxr5syZioyMVMuWLfXSSy+pUqVK5nVLly7VnDlzNGXKFE2ePFk2NiU6R1rF+fPnlZmZKU9Py2ZnTy8vHTp00EpVoSg4OzurTdt2+mzSBNVr0FCenl76NXietm/dIt86ftYuD0XoXGyMJKnKvz/nnl6KjY21RkkooC6PvaiUxKv64umeMtjYypiVqZ7PvaYWPR4yb7N+3jeytbXV3QOGWLFSFJV9+/bqns4dlJKSoooVK+qX4N/UsGEja5dV4pToX3xOTk4KCAhQTEyM9u/fL0k6c+aMDh06pEaNGsnLK+++gAcPHtSRI0fk4OCg559/3uJHtYuLi1544QVJ0qZNm3ThwgXzuuXLlysrK0vt27e3CAaS1KlTJ7Vq1SrP+wsJCVFsbKxatmypRx991BwMJMnBwUEvvPCCKlWqpNDQ0DxbKypWrKjnnnsuz2AgSR06dLB4DJJkMBjUq1cv1atXTydPntTp07fW7/nUqVPatGmT3NzcNGLECItgIEkPPPCAWrRoobNnzyo8PDzf46SnpyspKcn8Lzk5+ZbqAEqL6d/OkdFolL9fLVV1c9KMr/+nhwc9KgPBGSgV9oYsU8TaxXr0ncka/s0fGjT6E4UGf6edK3+TJJ06vE+bfp2jQaM/ybdXAkq3evXqa/O2cIVs3KL/e/5FPf9/T+vAgf3WLqvEKdEtB5LUvXt3bdy4UX/99ZcaNWp0UwORIyOzB5c0a9ZMrq6uudb7+vrK29tbx48f1/79+9WpUydJMgeQnNv/1rVrV+3YsSPX8l27dknK/hGfF0dHR/n6+io8PFxHjx7NNU6iSZMmcnJyyvfxSNldjyIiIhQTE6Pk5GRlZWVJki5fviwpOzTVrFnzuse4Vnh4uIxGo1q0aKHy5cvnuY2/v7/Cw8N16NChfIPR77//bjGmwcfHR5MmTbrpOkozDw8P2dra6tw5y7OH52JjVbVqVStVhaLi41tHS1auU2Jioq4kJKhqtWp6dvBjql3bx9qloQh5emV/luPOxapq1Wrm5efOxapJ0/zHvKHkWfbNRHV97AU1695HklTVt74uxZ5RyM/T1arXAEXv2a7E+Aua+Ghn8z5ZWZlaOv1jbfx1tt78Zb21Skchsbe3Vx2/7NbeFi1baeeOHfrqf1P1v6++sXJlJUuJDwf+/v7y9PTU1q1b9fTTT2vDhg0qX758rjP717p48aIkydPTM99tvLy8dPz4cfO2ksytCPntl9/ynJlpgoKC8u16lCMhISHXsipVquS7fVZWlr777jutWbMm3+nWJOU7niE/Oc3hf/31lzlw5SevmnP0799fffr0Md8uS2db7O3t1aJlK637a636PtRPUvbrtW7dWr348jDrFociU6FCBVWoUEHxly7przWrNG78RGuXhCLkXdtHnl5VtSFknZo0bS5JupKQoF07tumZ/3vBusXhlqSnpshgsGzps7G1kdGYfbKtxb395Nfqbov1s0Y9oxb3PqTW9w0stjpRfLKMWUpLS7N2GSVOiQ8HBoNBXbt2VXBwsKZNm6b4+Hj16NEj35mGrCHnR3vz5s1zdc/5t7yCwPUey7Jly7R69Wq5urpq8ODBql+/vipVqmTeZ+rUqdq06dYHRObUXLt27RvOwlS3bt1819nZ2eXbHaosGP7Ka/rPs0PUqlVrtW4ToKAvpygpMVGDh+R/ZW+UTmtXr5TRaFTdevV17Ojfeu+dN1W3Xn098dTT1i4Nt+nq1auKOva3+fbx41HauydCbm7uqnlXLb04dLi++OQj+dbxk7d3bX08fpyqVquu+x986DpHRUnToH13/fXTV3L1qi7P2nV15sh+bVwwS617D5IkVajkpgqV3Cz2sSlXTs7uVVSllq81SkYhem/MW7q3V2/ddVctXbl6RQvm/azQ9SH6Y0n+k82UVSU+HEjZ3XkWLFignTt3SrrxtQ1ypj293lzzOeuunSLV3d1dsbGxiouL01133ZXvPv9WuXJlnT59Wt27d79ui0ZBhIVlz3rz/PPPq3Xr1rnWx8TEFOi4lStXliTVr19fzz33XMELLOMGBT6i83Fx+uD9dxUbE6OmzZrrjyUr8h0Pg9IrISFBH773js6cPiU3N3c92G+Axrz3YZkOx3eKiF071e/+f2YdG/vmSEnSo088paBvZum/r76hxMREvf7fl3T5crzatr9b839fkmssGEq2h/77rlbNmqJFU97T1fgLcqnsqYA+j+mewbT0lgVxcef0/HNDFHP2rFwqVVLjxk31x5IV6t7jXmuXVuKUinDg4eGhNm3a6MCBA6patep1z2RL2V2RJCkiIkLx8fG5xh1ERUUpOjpaBoNBDRs2NC9v1KiRYmNjFRoaqpYtW+Y67oYNG/K8v+bNm2vPnj3avHlzoYeDnAHMeV1v4OTJk4qOjs5zv5xB0TljE/6tRYsWmjdvnnbs2KGnnnqqRLXElDYvDR2ml4byx+VO1//hQer/8CBrl4Ei0LFzF52/mp7veoPBoLfGjtNbY8cVX1EodA5OFfXgsDF6cNiYm96HcQZ3jq+++c7aJZQapWaajTfeeEPfffedJkyYcMNtGzRooLp16yotLU0zZsxQamqqeV1CQoJmzJghSbr77rstfnT37t1bNjY2CgsL07Zt2yyOuWnTJm3fvj3P++vRo4eqVKmiLVu2aO7cuXnO2BMfH681a9bc1GO9Vo0aNSRJK1eutPihf+nSJQUFBSkzMzPP/XJaBk6ePJnneh8fH7Vt21YXLlzQ559/nmerSEpKikJDQxUfH3/LdQMAAKD0KRUtBwUxfPhwffDBB9qxY4eGDRumhg0bKiMjQ5GRkUpOTpaPj4+effZZi31q166txx9/XHPnztVnn32munXrysvLS2fPntXRo0f1wAMPaOnSpbnuy9HRUW+++aYmTpyoxYsXa+3atapVq5YqV66s1NRUnT17VqdPn5aLi4t69OiRa//r6d+/vyIiIrR27VpFRkbKx8dHycnJ2r9/v7y8vBQQEJAryEjZMzU5ODho+/btGjt2rKpVqyYbGxvVr1/f3C3r5ZdfVlJSksLDw/XKK6/I29tbnp6eMhqNiouL0/Hjx5WRkaHJkyfnOesTAAAA7ix3bDjw8vLSpEmTtHjxYm3fvl07d+6UjY2Nqlevrvbt2+v+++/PsytN3759Vb16dS1evFhRUVE6efKkvL299dprr8nX1zfPcCBJd911lz777DOtXr1a27Zt04kTJ3T48GG5uLjI3d1dDz74oAICAm75cdStW1cTJ07UvHnzdPToUe3YsUMeHh6677779PDDD2vWrFl57ufq6qq3335bCxcu1LFjx3T48GEZjUZlZmaaw0H58uX1zjvvaPPmzQoNDdWxY8cUHR0tJycnubq6qmPHjmrdujX95wEAAMoIg/F682MCBZSaIfHGuvMlp+XdrQ13niz+VJQp3247bu0SUEyG3c1MTGWBjUEqb3dz082XmjEHAAAAAIoW4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJgaj0Wi0dhG486RmSLyxgDtHVhaf6LIkMTXD2iWgmCSnZ1m7BBQDO1uDqlWyv6ltaTkAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgEk5axdQUpw9e1bLli3Tvn37dP78eWVlZcnZ2Vlubm7y8/OTv7+/2rVrZ+0yAQAAgCJjMBqNRmsXYW1bt27Vl19+qfT0dDk7O8vHx0cuLi5KTExUdHS0Ll26JGdnZ3333XfWLrXUSM2QyvwbC7iDZGXxiS5LElMzrF0Ciklyepa1S0AxsLM1qFol+5vatsx3K4qPj9e0adOUnp6uPn366Ouvv9aYMWM0fPhwvfXWW/rmm280ceJE3XPPPdYuFSXU9K+mqb5fbblWdFSnDm21fds2a5eEIsJrXTZ8+snH6tQhQF6VXeRd00uPDOyvw4cOWbssFIKwTaF6MrCfmtTzlqeLvZYt+cNi/ScffaAOrRqrdlVX1a3lqYf73qed2/mcl0bTJn+iB++5W41qeahl/bv0nycH6eiRwxbb/DznWz3S9175e1eRd2VHXb4cb51iS5gyHw527dqllJQUubm5afDgwbK3z52qfH199fjjj1uhOpR0C4Lna/TI1/TOmPcUtm2XmjZtpr4P9NK5c+esXRoKGa912bFxwwY9/+LLWhcapj+XrVJ6err69umlxMREa5eG25SUmCj/xk018fOpea6v41dXH382VSFhu/TnynWqVctbgf3v1/nzccVcKW7X1s2hGvzcC1q0aoPm/rpU6RnpemrgA0q65nOcnJysLt17auiro6xYaclT5rsV/fbbb5o3b568vb316aef3tK+mZmZCgkJUWhoqI4fP67U1FS5ubmpefPm6t+/vzw8PMzbrl69WjNnzlSzZs30zjvv5Hm8K1eu6IUXXpDRaNQ333wjFxcX87qrV69q2bJl2r59u2JjY5WVlSUvLy916NBBffr0kYODg8WxgoODtXDhQg0cOFDdu3fXggULtGfPHsXHx6tjx44aOnSoedstW7Zo7dq1OnbsmJKSkuTi4iJ/f38NGDBANWvWvKXnJEdZ6VbUqUNbtWrdRlO+DJIkZWVlyc/nLr009L8aOepNK1eHwlTWX+uy3K0oLi5OtWt6aeWaEHXs1Nna5RSLstCtyNPFXrN/XqD7+zyU7zZXEhJUp6aHFi5eoc5duxdjdcWnrHQrunA+Ti3r36XgP1erbYdOFuvCNq7Xow/10p5jMapUydU6BRYxuhXdgpwf8CdPntTevXtver/k5GSNHz9e33zzjY4dOyZvb2+1bt1adnZ2Wr16tUaPHq2oqCjz9nfffbfs7e21Z88eXbx4Mc9jbty4URkZGWrVqpVFMDh16pRGjhyphQsXKiEhQQ0aNFCTJk2UkJCg+fPna+zYsUpKSsrzmDExMRo1apTCw8Pl5+enVq1aydnZWVJ2uJk8ebK++OIL7d+/X9WrV1ebNm3k4uKijRs36s0331RERMRNPydlTVpamsJ37VT3e3qYl9nY2Kh79x7atiXMipWhsPFal20Jly9Lktzc3a1cCYpTWlqafpj9rVwqVZJ/k6bWLge36UpCgiTJ1Y3P8Y2U+dmKAgIC5O7urosXL2r8+PFq1KiRGjduLB8fH/n5+Vn8SL/WzJkzFRkZqZYtW+qll15SpUqVzOuWLl2qOXPmaMqUKZo8ebJsbGzk5OSkgIAAbdy4URs2bFC/fv1yHTMkJESS1K1bN/OytLQ0TZo0SRcuXNCAAQM0cOBAlSuX/bKlpqZq+vTp2rRpk2bPnq2XX3451zE3btyoTp066cUXX5SdnZ3FuuDgYIWFhalu3boaMWKEPD09zeu2bNmiKVOmaOrUqQoKClKFChVu+jktK86fP6/MzEx5enpZLPf08tKhQwetVBWKAq912ZWVlaVRb7yq9h3ulr9/Y2uXg2KwavlSPf/sk0pOSpJX1WpasGi5Klf2uPGOKLGysrL0/jtvqHXb9qrf0N/a5ZR4Zb7lwNHRUe+++67q1q0ro9GoyMhIzZ8/XxMnTtT//d//adSoUVq1apWysv5pdjt16pQ2bdokNzc3jRgxwiIYSNIDDzygFi1a6OzZswoPDzcv7949u0kyJwRcKzo6WlFRUXJ1dVXz5s3Ny0NCQhQbG6uWLVvq0UcfNQcDSXJwcNALL7ygSpUqKTQ0VFevXs113IoVK+q5557LFQyuXr2qpUuXys7OTq+//rpFMJCkdu3aqUePHkpMTFRoaGi+z196erqSkpLM/5KTk/PdFgBKm1eHD9X+/fs058dfrF0Kisndnbvqr43btXT1BnXv0VP/efpxxcUxtqg0GztyhA4fiFTQzB+tXUqpUObDgSRVr15dEyZM0EcffaSBAweqWbNm5haD6Ohoffvtt5owYYIyMrL7YIaHh8toNKpFixYqX758nsf0989OpoeumeHC399fVapU0ZkzZ3T4sOWI+ZzA0LlzZ9na2pqX79q1S5LUoUOHPO/H0dFRvr6+yszM1NGjR3Otb9KkiZycnHIt37dvn9LS0tSgQQO559NUntdj+Lfff/9dTz/9tPnfuHHj8t32TuPh4SFbW1udOxdrsfxcbKyqVq1qpapQFHity6bXRgzT8uVLtXzlX6pRwPFXKH0qVKgg3zp+ah3QVlOmzZCtbTn9/MP31i4LBTR21Ctau2qZfvljparV4HN8M8p8t6Jr+fn5yc/PT5JkNBoVFRWlxYsXa/Pmzdq7d6+WLVumvn37KjY2+wfCX3/9pb/++uu6x0ww9XGTJIPBoK5du2rBggVat26d6tWrJ0nKyMgwn52/tkuRJPNMKEFBQQoKCrrp+8pRpUqVPLfNOe7evXsVGBh4y8fN0b9/f/Xp08d822AwXPdYdxJ7e3u1aNlK6/5aq74P9ZOU3XS5bt1avfjyMOsWh0LFa122GI1Gvf7Kf7V48SKtWLVOtX18rF0SrCgrK0upqanWLgO3yGg06t3Rr2rl0sWav3iVannzOb5ZhIN8GAwG+fr66pVXXlFaWpp27Nih7du3q2/fvsqZ4Kl27dry9va+7nHq1q1rcbtr165auHChwsLC9Mwzz8je3l47d+7UlStXVLduXdWoUcNi+5z7at68ea7uS/+WVxDIa2pWSeZuUlWrVlX9+vWve9x/13QtOzu7XF2WypLhr7ym/zw7RK1atVbrNgEK+nKKkhITNXjIM9YuDYWM17rseHX4UAXP/0XzFy5SRWdnxcTESJIqVaqUb2sxSoerV68q6tjf5tsnoqO1d0+E3Nzc5eZeWVM++1i9ej8or6pVdfHCBc2a+bVizp5W3/4PW7FqFMSYkSO0+Nf5mjl3gSpUrKhzsdmfYxeXSnI0fY7PxcYo7lysoqOye14c2r9PFSo6q0bNu8r0wGXCwU1o2rSpduzYYT6DXrlyZUlS/fr19dxzz93SsapUqaLGjRtr79692rp1qzp16pTnQOQclStX1unTp9W9e3e1a9fu9h7INXJmaapevbrFtKa4NYMCH9H5uDh98P67io2JUdNmzfXHkhXy8vK68c4oVXity46ZM6ZLku671/I7efrMWXpq8NNWqAiFZXf4TvV/4F7z7XffHilJeuTxp/TplGk6cviQ5v88VxcvnJebe2W1aNlKi1esUwMGsZY6c7+fIUl6pG9Pi+Wf/W+GBj0+WJL00+yZmvLJBPO6QX165NqmLCrz4cBoNN6wK8z58+cl/RMKWrRooXnz5mnHjh166qmn8j07n59u3bpp7969CgkJUZMmTRQRESF7e/s8xxU0b95ce/bs0ebNmws1HDRu3FjlypVTZGSkLl++fMNWCeTvpaHD9NJQupaUBbzWZUNiatmY970surtTF51LSMt3/eyfFhRjNShKxy+k3HCbV0eP1aujxxZDNaVLmR+QvHLlSgUFBeU56NZoNGrr1q1auXKlpH8GBfv4+Kht27a6cOGCPv/88zyvkJqSkqLQ0FDFx8fnWhcQEKAKFSpo3759+u2335SZmam2bdvmOXC4R48eqlKlirZs2aK5c+fmORtQfHy81qxZc0uP29XVVffdd59SU1M1adIknThxItc26enp2rFjh06fPn1LxwYAAEDpVOZbDjIzM7VhwwZt2LBBLi4u8vHxkbOzsxITE3Xq1CnFxWVfMr1Tp07mqUgl6eWXX1ZSUpLCw8P1yiuvyNvbW56enjIajYqLi9Px48eVkZGhyZMny9XV1eI+c1oJVq9erRUrVkjKu0uRlD0b0ZtvvqmJEydq8eLFWrt2rWrVqqXKlSsrNTVVZ8+e1enTp+Xi4qIePXrkeYz8PPHEE4qPj9fGjRs1cuRI1a5dW56enrK1tdWFCxcUHR2t1NRUvf3229cddwAAAIA7g8GYM+K1jEpOTta+ffu0d+9e/f3337p06ZLi4+Nla2srNzc3+fn5qUuXLhbXHsiRlZWlzZs3KzQ0VMeOHdPVq1fl5OQkV1dX1alTR61bt1bLli0trk2Q4++//9bbb78tKXscQlBQ0HW7NyUnJ2v16tXatm2bTp8+rZSUFLm4uMjd3V3+/v4KCAgwz34kZV/gbOHChRo4cOANZyMKDw/X2rVr9ffff+vy5ctydHSUq6ur+arPbdq0kYODw00+o9lSM6Qy/cYC7jBZWXyiy5LE1Axrl4BikpxON7qywM7WoGqVbq4bfJkPBygahAPgzkI4KFsIB2UH4aBsuJVwUObHHAAAAADIRjgAAAAAIIlwAAAAAMCEcAAAAABAEuEAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAAAAmBAOAAAAAEgiHAAAAAAwIRwAAAAAkEQ4AAAAAGBCOAAAAAAgiXAAAAAAwIRwAAAAAEAS4QAAAACACeEAAAAAgCTCAQAAAAATwgEAAAAASYQDAAAAACaEAwAAAACSCAcAAAAATAgHAAAAACQRDgAAAACYEA4AAAAASCIcAAAAADAhHAAAAACQRDgAAAAAYEI4AAAAACCJcAAAAADAhHAAAAAAQBLhAAAAAIAJ4QAAAACAJMIBAAAAABPCAQAAAABJhAMAAAAAJoQDAAAAAJIIBwAAAABMCAcAAAAAJBEOAAAAAJgQDgAAAABIIhwAAAAAMCEcAAAAAJBEOAAAAABgQjgAAAAAIIlwAAAAAMCEcAAAAABAklTO2gXgzmSwdgEACpWBD3WZUo5Th2WGnS0f7rKgnM3Nv84Go9FoLMJaAAAAAJQSnBsACkFycrJGjx6t5ORka5eCIsZrXXbwWpcdvNZlB6/1jREOgEJgNBoVFRUlGuLufLzWZQevddnBa1128FrfGOEAAAAAgCTCAQAAAAATwgFQCOzs7DRw4EDZ2dlZuxQUMV7rsoPXuuzgtS47eK1vjNmKAAAAAEii5QAAAACACeEAAAAAgCTCAQAAAACTctYuALhVQ4cOVVxcnCTptddeU7t27fLc7sMPP9TevXv18ssvq2vXrsVYofVNmzZN69evL/GP/drXUpIMBoMcHBzk5OSkatWqycfHRx06dJCfn58Vq8x27tw5DRs2TFWqVNG0adMK/fhGo1F//vmn1q9fr5iYGKWnp0uSgoODC3zMoq65JLv2vdW7d28988wz+W67ePFizZ07V5JkY2OjefPmmdeNGzdO+/fv13vvvSd/f/+bvv/g4GAtXLhQAwcOVGBgYAEfRclXWM9zUSjL7/+S6OzZs1q2bJn27dun8+fPKysrS87OznJzc5Ofn5/8/f3z/XuO4kU4QKn2yy+/qE2bNrK1tbV2KbgN9evXV9WqVSVJaWlpunLliqKiohQZGaklS5aoUaNGeumll+Tl5WXlSgsm5wdUUFCQPD0989xm1apVmjt3rpycnNSiRQuVL1++mKu8c23cuFFPPfWUypXL+0/eunXrirmiOxPPc8Hd6WFy69at+vLLL5Weni5nZ2fVr19fLi4uSkxMVHR0tFauXKnNmzcTDkoIwgFKLQcHB509e1Zr165Vz549rV0ObsM999yTq4XDaDQqPDxcc+bM0f79+zVmzBhNmDAh3x/XpV1YWJik7Nawpk2bWrmaO0edOnV09OhRbd++Xe3bt8+1/tChQzp9+rR5u38bNmyYUlNT5eHhURzlllq3+zzjzhUfH69p06YpPT1dffr00aOPPip7e3uLbY4dO6YtW7ZYqUL8G2MOUGr17t1bkvTrr78qNTXVytWgsBkMBrVs2VIfffSRqlWrpsuXL2v69OnWLqvInD9/XpLMLSgoHN26dZOU/1nrv/76y2K7f/Pw8FCNGjXk4OBQNAXeIW73ecada9euXUpJSZGbm5sGDx6cKxhIkq+vrx5//HErVIe80HKAUqtly5Y6fPiw9u/fr6VLl2rAgAE3ve+mTZu0du1aRUVFKTU1Va6urmrcuLH69eun6tWr59r+2m4hx48f17JlyxQdHa3ExERzX+ScpuDg4GBt2LBBy5cv16lTp2Rvb68mTZroySeflIeHh4xGo1auXKm1a9cqJiZG9vb2atmypZ588klVqlTJ4n4zMjK0efNmRURE6NixY7p06ZIyMjLk4eGhZs2aqV+/fnJ3d7+9J7KEq1ChgoYMGaKJEydq3759OnbsmHx9fc3rr169qsWLF2vHjh06d+6cbGxsVK1aNXXo0EG9e/fO9YcoJCREX331lbp06aKhQ4fmur+b6aecmZmpJUuWaP369YqNjZWjo6P8/f31yCOPqEaNGrnuK8ewYcMsjvPee+9pwYIF2r9/f57b5HQxuFGXg8jISL3//vtq1KiRxo0bl88zWTbVqlVLderU0e7du3Xx4kWLz0tKSorCwsJUuXJlNWvWLM/9rzfmIC0tTYsWLdLGjRt1/vx5VaxYUU2bNtWjjz5apI+pJLrd51m69c9yjp07d2rx4sWKioqSwWCQt7e3HnzwQXl7e1+35qtXr2rZsmXavn27YmNjlZWVJS8vL3Xo0EF9+vTJFQiv/Rx2795dCxYs0J49exQfH6+OHTuav0+2bt2q8PBwHTlyRBcvXlRaWppcXV3l7++f59+Yaz/TCxcu1MKFC823//09lZmZqZCQEIWGhur48eNKTU2Vm5ubmjdvrv79+5fIFq74+HhJkouLyy3veyuPd/Xq1Zo5c6aaNWumd955J8/jXblyRS+88IKMRqO++eYbi5qK6v0gSVu2bNHatWt17NgxJSUlycXFRf7+/howYIBq1qx5y89LUSMcoFR74okn9M477+iPP/7QvffeK2dn5+tubzQaNW3aNG3YsEG2trZq2LChXFxcFBUVpZCQEIWFhen1119X8+bN89x/yZIlWrFiherUqaPmzZvr0qVLsrGxbID7+eef9eeff6phw4Zq0aKF/v77b23evFmHDh3Sp59+qpkzZ2rHjh3y9/eXl5eXDh06pPXr1ys6Oloff/yxRX/dy5cvKygoSE5OTqpRo4Zq1aql1NRURUdHa8WKFdq8ebPGjx9/x59tbtGihSpWrKirV69qz5495nAQGxurDz74QHFxcXJxcVGLFi2UmZmpyMhI/fTTT9q8ebPGjh2rihUrFmo9kydP1s6dO9WoUSPVqlVLR48e1ZYtWxQREaExY8aoXr16krJbAbp06aItW7YoNTVVbdu2laOjo/k4rq6uat68uapUqZLnNrVr1y7Uusuqbt266ejRowoJCbE4ibB582alpKTo/vvvl8FguKVjpqam6oMPPtCRI0fk4OCgpk2byt7eXrt371Z4eLhatGhR2A+jxLud57mgn+UlS5bohx9+kCT5+fnJy8tLMTEx+vTTT9WnT598az116pQmTJigCxcuyM3NTQ0aNJCtra3+/vtvzZ8/X1u3btW4cePk5OSUa9+YmBiNGjVK5cqVU/369WU0Gi3+9kyePFl2dnaqWbOm/P39lZWVpZMnT5r/xowZM0b169c3b9+lSxdFR0fr+PHj8vb2tvjcN2jQwPz/5ORkffLJJ4qMjJSjo6N8fX3l4uKiEydOaPXq1dqyZYvGjBkjHx+ffB+3NeT8gD958qT27t2rJk2a3NR+t/p47777bs2ZM0d79uzJFVBzbNy4URkZGQoICLAIBkX1fsjMzNSXX36psLAw2dnZydfXV25ubjp79qw2btyobdu26Y033sj3N4e1EA5QqtWtW1cBAQHatm2bfvvtNw0ZMuS6269evVobNmyQs7Ozxo4da/4SNhqNWrBggRYuXKipU6dq6tSpeZ7lWLVqlUaNGqXWrVvnex9r167Vxx9/bD52Wlqaxo8fr4MHD2rcuHFKTU3VlClTVKVKFUlSQkKCxowZo+PHjyssLEydOnUyH8vJyUmjRo1S8+bNLUJDRkaGgoODtWjRIn3//fd66623bvYpK5UMBoN8fHy0d+9enTx50rz8yy+/VFxcnFq3bq3hw4ebf1QnJCRowoQJioqK0qxZszR8+PBCqyUuLk4pKSmaOHGi+cxkVlaWZs+erRUrVmjq1KmaMmWK7Ozs1KBBAzVo0ED79+9XXFycnnrqqVxjJnJaGq63DW5Px44d9cMPP+T60bpu3ToZDIYCdXUJDg7WkSNHVKNGDY0dO9b8QyQ1NVVTp07Vhg0bCq3+0uJ2nueCfJaPHz+uuXPnymAw6NVXX7UYzBoaGqqgoKA87ystLU2TJk3ShQsXNGDAAA0cOND8/Zqamqrp06dr06ZNmj17tl5++eVc+2/cuFGdOnXSiy++KDs7u1zrhw8frpYtW1qcCDAajVq1apW+++47zZgxQ5999pk5KA0dOlTBwcE6fvy42rRpk++A5JkzZyoyMlItW7bUSy+9ZNHSvHTpUs2ZM0dTpkzR5MmTc520sqaAgAC5u7vr4sWLGj9+vBo1aqTGjRvLx8dHfn5++bYo3OrjdXJyUkBAgDZu3KgNGzaoX79+uY4ZEhIiybJ7W1G+H4KDgxUWFqa6detqxIgRFt/tW7Zs0ZQpUzR16lQFBQWpQoUKN/2cFrWS8+4BCujxxx+Xra2tVq1aZTEtZl7+/PNPSdndNa49O2MwGDRo0CB5e3srMTFRa9asyXP/Ll26XDcYSNlNxNce297e3nwG68SJE3rmmWfMwUDKbmrNGVC9d+9ei2OVL19erVu3zjX7R7ly5fT444/Lzc1NERERSk5Ovm5Nd4KcMzFXr16VJB08eNB81vb555+3+EPs4uKiF154QVJ2F7ILFy4Uai0PP/ywRZcFGxsbPfXUU3J3d1dcXJy2bt1aqPeH25PzoyEmJsbchevMmTM6dOiQGjVqdMuzYKWlpZm/I4YMGWJxhjLn/ZjXj8Y7XUGf54J+lpcvX66srCy1b98+1yw3nTp1UqtWrfK8v5CQEMXGxqply5Z69NFHLb5fHRwc9MILL6hSpUoKDQ01f99cq2LFinruuefyfY07dOhg8Rik7L8xvXr1Ur169XTy5EmdPn06z33zc+rUKW3atElubm4aMWJEri6oDzzwgFq0aKGzZ88qPDz8lo5d1BwdHfXuu++qbt26MhqNioyM1Pz58zVx4kT93//9n0aNGqVVq1YpKyvLvE9BH2/37t0l/RMCrhUdHa2oqChzi22Oono/XL16VUuXLpWdnZ1ef/31XCd92rVrpx49eigxMVGhoaE3fiKLEeEApV716tXVrVs3paena/78+flud+HCBcXGxkrK/pH/bwaDwTxjTmRkZJ7HuJlp1vLqTpDT7cfW1jbPPrc56y9dupTnMaOjo7VkyRLNmjVLX331laZNm6Zp06YpKytLRqNRMTExN6yrtDMajRa3c16jZs2aydXVNdf2vr6+8vb2ltFotOjTXxjyev/Y2dmpQ4cOFrWh5Mj50ZAzMPZ2BsgeO3ZMycnJcnZ2zrM7gKur63X71t/JCvI8F/SznPP/a1tbr5XfNV527dolSebP67/ldGHJzMzMc2alJk2a5Nm95FoxMTFasWKFZs+era+//tr8nX358mVJ2aHpVoSHh8toNF53muOcMTGHDh26pWMXh+rVq2vChAn66KOPNHDgQDVr1szcYhAdHa1vv/1WEyZMUEZGhqSCP15/f39VqVJFZ86c0eHDhy22zwkMnTt3tpj+vKjeD/v27VNaWpoaNGiQ79jAkvqa0a0Id4RBgwYpNDRUoaGh+Q5Eu3jxoqTsM9D5fbHnnNnK2fbfrj3jn5+8BoTlnEVydXXN85oMOV9+ORe+ypGSkqKgoCBt27btuvdZFloOrly5IknmPsc5r9H1uuB4eXnp+PHj+b6eBVGhQoV8m39zainM+0Ph8Pf3l6enp7Zu3aqnn35aGzZsUPny5Qs0r/rNvPfKatewgjzPBf0s57Qi5LdffsvPnTsnSQoKCsq361GOhISEXMuu93cgKytL3333ndasWZPrhMa1kpKSrnu//5ZzYuuvv/4yB6785FVzSeHn52e+qKXRaFRUVJQWL16szZs3a+/evVq2bJn69u1b4Mebc5JvwYIFWrdunXn8V0ZGhvns/L+DalG9H3KOu3fv3hteu6KkvWaEA9wR3Nzc1Lt3by1atEg///xzkfXBz2+2jGtdr6/nrfYD/fnnn7Vt2zbVqFFDjz/+uOrUqSMXFxdzs+eYMWN0+PDh6/4RuhPk/BGRsmdFKer7KgnHKMn3Vxrl/GgIDg7WtGnTFB8frx49etzUZxo3rzQ8zzmfl+bNm+fqrvJvef3wu95jWbZsmVavXi1XV1cNHjxY9evXV6VKlcz7TJ06VZs2bSpwzbVr177hLEx169a95eNbg8FgkK+vr1555RWlpaVpx44d2r59u/r27Xtbj7dr165auHChwsLC9Mwzz8je3l47d+7UlStXVLduXYsZ5aSiez/kdJOqWrWqxQD0vPy7JmsjHOCO0a9fP61du1bh4eF5diPJada7cuWKkpKS8mw9yEn6JWV60JwLY73yyit5fkGePXu2uEuyivDwcCUmJkqSubtGzmuU85rlJa/XMydY5dfacqNxK4mJiUpMTMyz9aCo3j85NaekpOS5/kY1I1vOGcWdO3dKKvic+7fy3iuLbvV5Luhn2d3dXbGxsYqLi9Ndd92V7z7/VrlyZZ0+fVrdu3cv9Cvy5nxnP//883mOTytoF9DKlStLyr6a/HPPPVfwAkuopk2baseOHeYz6LfzeKtUqaLGjRtr79692rp1qzp16pTnQOQcRfV+yOlFUL169TynzS7JGHOAO4aTk5P69+8vSfrpp59yra9cubK521Beg5WMRqN5+b/nM7eWnMFPeZ2tiIiIMHe1uZMlJSVpzpw5krL/gOQM9s55jSIiIszzaF8rKipK0dHRMhgMatiwoXl5zo+L/Pr85vQ/vZ68ZqLJyMgw/zD49/sn58f9tQPubkVOzfkNYryZmpH9x7pNmzZydnZW3bp1C3yG1dfXV46Ojrpy5Yp2796da318fLz27Nlzu+WWWrf6PBf0s9yoUSNJyncwZ34zRuWME9m8efONHsoty/nOzqt76cmTJxUdHZ3nfjf6jsgZy7Zjxw6lpaUVQqXF52ZaNnMuApkTCm738eaEgJCQEMXHxysiIkL29vZ5jisoqvdD48aNVa5cOUVGRprHmpQWhAPcUXr16iUPDw8dOXIk12AkSXrwwQclZV9V+dovaaPRaF5WoUIF9ejRo7hKvq6ci6MsX77cYvmZM2c0c+ZMa5RUbIxGo8LDw/XWW2/p7NmzcnNzM89aImXP/123bl2lpaVpxowZFlfJTkhI0IwZMyRlz3197R9qPz8/lS9fXqdOncr14yEsLCzXc52XX3/9VSdOnDDfzsrK0ty5c3XhwgVVrlxZbdu2tdg+58f9tdOw3orGjRvLYDAoIiLColXMaDRq2bJlzI50C9544w199913mjBhQoGPYW9vb/6OmDNnjsVEAmlpafr2229L3Q+4wnYrz3NBP8u9e/eWjY2NwsLCco3L2rRpk7Zv357n/fXo0cN8bZG5c+fm2YoYHx+f76x115PTPWTlypUWP/QvXbqkoKAgZWZm5rlfzo/i/L4jfHx81LZtW124cEGff/55nq0iKSkpCg0NzTNgWdPKlSsVFBSU56Bbo9GorVu3auXKlZL+GRR8u483ICBAFSpU0L59+/Tbb78pMzNTbdu2zbPHQFG9H1xdXXXfffcpNTVVkyZNsvibkSM9PV07duy45dmrihrdinBHsbOzU2BgoL766iuLPzA57r33Xh0+fFgbNmzQW2+9pUaNGpkvgnbmzBnZ29tr+PDhBbqSY1EYOHCgvvjiC82fP19hYWGqWbOmEhISdODAATVs2FDu7u4lbpaDgli7dq15xpL09HRduXJFUVFR5rNw/v7+eumll3K1oAwfPlwffPCBduzYoWHDhqlhw4bKyMhQZGSkkpOT5ePjo2effdZiH3t7ewUGBmrOnDkKCgrSqlWr5O7urtOnT+vUqVMaMGCAfv3113xr9fDwkK+vr0aPHi1/f39VrFhRR48eVWxsrBwcHDRixIhcfVDbtm2ryMhI/e9//1PTpk3Ng6r79u2b5xW587rP3r17a9myZfrggw/UsGFDVahQQcePH9f58+fVr18/LVq06IbHQeEJDAzUwYMH9ffff2vEiBHy9/eXnZ2dDh48qMzMTHXu3LlMXuugoAryWa5du7Yef/xxzZ07V5999pnq1q0rLy8vnT17VkePHtUDDzygpUuX5rovR0dHvfnmm5o4caIWL16stWvXqlatWqpcubJSU1N19uxZnT59Wi4uLrd8oqh///6KiIgwf6f5+PgoOTlZ+/fvl5eXl/m6PP/WrFkzOTg4aPv27Ro7dqyqVasmGxsb1a9f33wW/OWXX1ZSUpLCw8PNXU09PT1lNBoVFxen48ePKyMjQ5MnT85z1idryczM1IYNG7Rhwwa5uLjIx8dHzs7OSkxM1KlTp8zdIjt16mSe7Uq6vceb00qwevVqrVixQlL+3duK8v3wxBNPKD4+Xhs3btTIkSNVu3ZteXp6ytbWVhcuXFB0dLRSU1P19ttvl6hxB4QD3HE6d+6sJUuW5JnSDQaDhg0bpubNm5svZZ6SkiJXV1d17do1z0vbW1Pbtm01btw4LVy4UNHR0YqNjZWnp6cGDRqkvn37avz48dYusVAcOnTIHHIcHBzk5OSkWrVqydfXVx06dDDPbvFvXl5emjRpkhYvXqzt27dr586dsrGxUfXq1dW+fXvdf//9eQ4We+CBB1SxYkUtW7ZMUVFROnnypOrUqaMhQ4aoatWq1w0HORdc+uOPPxQaGqoDBw7IwcFBbdu21SOPPGJu7blWz549zWe5wsPDzbNSderU6abfb0OGDJGHh4fWrl2rQ4cOydHRUfXr19err76q5ORkwkExc3R01HvvvadFixZp48aN2r17typUqKAmTZro0UcfzbPrIvJX0M9yTsBevHix+bPs7e2t1157Tb6+vnmGA0m666679Nlnn2n16tXatm2bTpw4ocOHD8vFxUXu7u568MEHFRAQcMuPo27dupo4caLmzZuno0ePaseOHfLw8NB9992nhx9+WLNmzcpzP1dXV7399ttauHChjh07Zp5oIjMz0/yjtnz58nrnnXe0efNmhYaG6tixY4qOjpaTk5NcXV3VsWNHtW7d+pav21HUunfvLk9PT+3du1d///23Tp06pfj4eNna2srNzU133323unTpkmta4Nt9vN26ddPq1aslZXfNvV534aJ6P9ja2mr48OHq1KmT1q5dq7///lsnTpyQo6OjXF1d1apVK7Vu3dqiu1xJYDAyzQUAAAAAMeYAAAAAgAnhAAAAAIAkwgEAAAAAE8IBAAAAAEmEAwAAAAAmhAMAAAAAkggHAAAAAEwIBwAAAAAkEQ4AAKVIZGSkAgMDFRkZaV42bdo0DR061IpVWcqrxryEhIQoMDBQ586du+X7GDdunF5//fWClpinoUOHatq0aYV6TAClD+EAAFAm/fbbb9q2bZu1ywCAEqWctQsAAOB2vPDCCzIajbe83++//6527dopICCgCKoCgNKJcAAAKHJZWVnKyMiQvb19oR+7XDn+lAFAYeEbFQBwU4KDg7Vw4UJNnjxZ8+fP1+7du2Vra6tOnTrpiSeesPjhHxgYqF69eqlevXr6/fffdfbsWb366qsKCAjQxYsXNW/ePIWHhysxMVFVq1ZVnz591L17d4v7u3Dhgr777jvt3btXDg4O6tixo5o3b56rrmnTpmn//v0W/eWzsrK0YsUKrV27VjExMXJ0dJSvr68effRR1alTR4GBgZKk9evXa/369ZKkLl26mMcuFHaNN2v79u1as2aNoqOjdeXKFVWuXFldunTRgAEDZGOTuyfwsWPHNGvWLEVFRcnV1VUPPfSQevbsabFNenq6fv/9d4WGhurChQuqVKmS7r77bj3yyCOys7MrcK0A7kyEAwDALZk8ebKqVKmixx57TEeOHNHy5cuVmJioYcOGWWy3b98+hYWF6b777pOzs7M8PT0VHx+vd955R5LUq1cvubi4KCIiQtOnT1dycrIeeOABSVJaWpo++OADnT9/Xr1795a7u7s2bNhww0G+OaZPn66QkBC1aNFC99xzjzIzM3XgwAEdOXJEderU0bBhw/TNN9/Iz89P99xzjySpatWqklRsNeYlJCREjo6OeuCBB+To6Kh9+/YpODhYycnJeuqppyy2vXr1qj7++GO1b99ed999t8LCwvTtt9+qXLly5hCTlZWlTz75RAcPHtQ999yjmjVr6sSJE1q6dKnOnDmjUaNGFbhWAHcmwgEA4JZ4enqaf1Ted999Kl++vFatWqUHH3xQ3t7e5u3OnDmjzz//XDVr1jQvmz59urKysvTZZ5/J2dlZktSzZ09NmTJFCxYs0L333it7e3utWbPG3NrQvn17SdI999yjkSNH3rC+ffv2KSQkRL1799YzzzxjXv7ggw+axyZ07txZM2fOlKenpzp37myx/7x584q8xvyMGDHCogWmZ8+emjFjhlatWqVHH33U4kz/pUuXNHjwYPXp00eSdO+99+rtt9/WL7/8os6dO6tcuXLauHGj9uzZo/fff18NGjQw73vXXXdp5syZOnTokOrXr1/gegHceZitCABwS3r16mVxu3fv3pKk8PBwi+WNGjWyCAZGo1Fbt25Vq1atZDQalZCQYP7XvHlzJSUl6dixY+Zjubm5qV27dub9HRwc1KNHjxvWt3XrVhkMBg0aNCjXOoPBcN19i6vG/FwbDJKTk5WQkKCGDRsqNTVVp0+fttjW1tbW4r7KlSunHj166PLly+Yat2zZopo1a6p69eoWj6Vx48aSdFutHADuTLQcAABuSbVq1Sxue3l5yWAw5Jqv39PT0+J2QkKCEhMTtWbNGq1ZsybPYyckJEiS4uLiVLVq1Vw/5qtXr37D+mJjY+Xm5qaKFSvecNu87r84aszPyZMnNW/ePO3bt0/JyckW65KSkixuu7m5ydHRMc/7jouLU716/9/evbxC38ZxHP88SpNuJlMOZSGNHBKGlZTTUv4BhShJslJSpGSlLBUrZOGQkiRyio0pw0YJKWWccmgKiSmH4XcvbibD4OHuoTzv125+v+vq+ja77+97Xd8rXkdHRzo4OFBFRYXf9c7Pzz8dK4CfieQAAPBXXvsa/7wz0eOWnuzsbOXm5vqd83Rb0nf4zhjdbream5sVFBSkwsJCRUZGKjAwUNvb2+rv7/9Uu1bDMBQdHa3S0lK/78PCwv42bAA/DMkBAOBDjo6OfKoCx8fHMgzjRaXgObPZrKCgIN3f3ys1NfXNseHh4drb25NhGD7Jx+Hh4bvxRUZGamVlRZeXl29WD/wlNV8Voz/r6+u6uLhQbW2tkpKSvM9fu0H57OxMV1dXPtWDx7XDw8Ml/fkvdnd3lZKS8u6WKgCQOHMAAPig6elpn9+Tk5OS9G4Lz4CAAGVkZGhpaUl7e3sv3j9u15Gk9PR0nZ2daXFx0fvs+vr61a0+T2VkZMgwDA0NDb149/Tru8lkktvt/pYY/fHXqtTj8WhmZsbv+Lu7O5+1PB6PZmdnZTabZbVaJUmZmZk6PT3V3Nzci/k3Nze6urr6VKwAfi4qBwCAD3G5XGptbVVaWpo2Nzdlt9uVlZWlmJiYd+cWFRVpfX1djY2N3taal5eXcjqdWl1dVU9Pj6Q/XX+mpqbU3t4up9Mpi8Wi+fl5mUymd9dITk5WTk6OJicndXx8LJvNJsMwtLGxoeTkZOXn50uSrFarVldXNT4+LovFooiICMXFxX1JjP4kJCTo169f6ujo8B7yttvtr24nslgsGh0dlcvlUlRUlBYWFrSzs6PKykrvxXA5OTlyOBzq7OzU2tqaEhMTdX9/r4ODAzkcDjU2Nio2NvZT8QL4magcAAA+pKamRoGBgRoYGNDy8rLy8/NVVVX1r+aGhoaqpaVFeXl5WlpaUnd3tyYmJuR2u1VcXOwdZzKZ1NTUJJvNpqmpKQ0PDysxMdFnzFuqq6tVUlIil8ulvr4+jYyM6Pb2VvHx8d4xZWVlslqtGhwcVFtbm/cL/VfF+FxISIjq6+sVGhqqwcFBjY2NKSUlRSUlJX7HBwcHq6GhQU6nU729vTo5OVF5eblPB6OAgADV1dWpqKhI+/v76u3t1dDQkLa2tlRQUPDicDkA/GN85oQTAOB/5/GG5K6uLpnN5u8OBwDwH6ByAAAAAEASyQEAAACAByQHAAAAACRx5gAAAADAAyoHAAAAACSRHAAAAAB4QHIAAAAAQBLJAQAAAIAHJAcAAAAAJJEcAAAAAHhAcgAAAABAEskBAAAAgAckBwAAAAAkSb8BGxfOCYB8BfwAAAAASUVORK5CYII="},"metadata":{}}]}]} \ No newline at end of file