File size: 109,625 Bytes
b81c7ad |
1 |
{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[{"file_id":"10ug8ACYtGx_f_qOyoWrBmRjQJSvlRig2","timestamp":1687945981523}],"gpuType":"T4"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU"},"cells":[{"cell_type":"code","execution_count":1,"metadata":{"id":"3RLcLALzXHvF","executionInfo":{"status":"ok","timestamp":1687947267413,"user_tz":-330,"elapsed":7847,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"outputs":[],"source":["import tensorflow\n","from tensorflow import keras\n","from keras import models\n","from keras import Sequential\n","from keras.layers import Dense,Flatten, Dropout\n","from keras.applications.vgg16 import VGG16"]},{"cell_type":"code","source":["from google.colab import drive\n","drive.mount('/content/drive')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"nJS2cnYWdL06","executionInfo":{"status":"ok","timestamp":1687947292567,"user_tz":-330,"elapsed":25172,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"ee27cb4e-6e0d-495e-c3c2-0661810e7c7b"},"execution_count":2,"outputs":[{"output_type":"stream","name":"stdout","text":["Mounted at /content/drive\n"]}]},{"cell_type":"code","source":["vgg_base = VGG16(weights='imagenet', # use weights for ImageNet\n"," include_top=False, # drop the Dense layers!\n"," input_shape=(300, 300, 3))\n","print(vgg_base.summary())"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"a-n70dmOweOF","executionInfo":{"status":"ok","timestamp":1687947298519,"user_tz":-330,"elapsed":5956,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"7bf24647-3248-495c-b0a3-45dd49fba541"},"execution_count":3,"outputs":[{"output_type":"stream","name":"stdout","text":["Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n","58889256/58889256 [==============================] - 2s 0us/step\n","Model: \"vgg16\"\n","_________________________________________________________________\n"," Layer (type) Output Shape Param # \n","=================================================================\n"," input_1 (InputLayer) [(None, 300, 300, 3)] 0 \n"," \n"," block1_conv1 (Conv2D) (None, 300, 300, 64) 1792 \n"," \n"," block1_conv2 (Conv2D) (None, 300, 300, 64) 36928 \n"," \n"," block1_pool (MaxPooling2D) (None, 150, 150, 64) 0 \n"," \n"," block2_conv1 (Conv2D) (None, 150, 150, 128) 73856 \n"," \n"," block2_conv2 (Conv2D) (None, 150, 150, 128) 147584 \n"," \n"," block2_pool (MaxPooling2D) (None, 75, 75, 128) 0 \n"," \n"," block3_conv1 (Conv2D) (None, 75, 75, 256) 295168 \n"," \n"," block3_conv2 (Conv2D) (None, 75, 75, 256) 590080 \n"," \n"," block3_conv3 (Conv2D) (None, 75, 75, 256) 590080 \n"," \n"," block3_pool (MaxPooling2D) (None, 37, 37, 256) 0 \n"," \n"," block4_conv1 (Conv2D) (None, 37, 37, 512) 1180160 \n"," \n"," block4_conv2 (Conv2D) (None, 37, 37, 512) 2359808 \n"," \n"," block4_conv3 (Conv2D) (None, 37, 37, 512) 2359808 \n"," \n"," block4_pool (MaxPooling2D) (None, 18, 18, 512) 0 \n"," \n"," block5_conv1 (Conv2D) (None, 18, 18, 512) 2359808 \n"," \n"," block5_conv2 (Conv2D) (None, 18, 18, 512) 2359808 \n"," \n"," block5_conv3 (Conv2D) (None, 18, 18, 512) 2359808 \n"," \n"," block5_pool (MaxPooling2D) (None, 9, 9, 512) 0 \n"," \n","=================================================================\n","Total params: 14,714,688\n","Trainable params: 14,714,688\n","Non-trainable params: 0\n","_________________________________________________________________\n","None\n"]}]},{"cell_type":"code","source":["model = Sequential([\n"," # our vgg16_base model added as a layer\n"," vgg_base,\n"," # here is our custom prediction layer\n"," Flatten(),\n"," Dropout(0.50),\n"," Dense(1024, activation='relu'),\n"," Dropout(0.20),\n"," Dense(512, activation='relu'),\n"," Dropout(0.10),\n"," Dense(1, activation='sigmoid')\n"," ])"],"metadata":{"id":"AeRPxkDovZFU","executionInfo":{"status":"ok","timestamp":1687947298519,"user_tz":-330,"elapsed":13,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":4,"outputs":[]},{"cell_type":"code","source":["model.summary()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"f8ViMxe0xuSx","executionInfo":{"status":"ok","timestamp":1687947298519,"user_tz":-330,"elapsed":12,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"e4a3f43a-740c-468e-bccc-2149c4fce63b"},"execution_count":5,"outputs":[{"output_type":"stream","name":"stdout","text":["Model: \"sequential\"\n","_________________________________________________________________\n"," Layer (type) Output Shape Param # \n","=================================================================\n"," vgg16 (Functional) (None, 9, 9, 512) 14714688 \n"," \n"," flatten (Flatten) (None, 41472) 0 \n"," \n"," dropout (Dropout) (None, 41472) 0 \n"," \n"," dense (Dense) (None, 1024) 42468352 \n"," \n"," dropout_1 (Dropout) (None, 1024) 0 \n"," \n"," dense_1 (Dense) (None, 512) 524800 \n"," \n"," dropout_2 (Dropout) (None, 512) 0 \n"," \n"," dense_2 (Dense) (None, 1) 513 \n"," \n","=================================================================\n","Total params: 57,708,353\n","Trainable params: 57,708,353\n","Non-trainable params: 0\n","_________________________________________________________________\n"]}]},{"cell_type":"code","source":["vgg_base.trainable = False"],"metadata":{"id":"ZAGXpQnMx541","executionInfo":{"status":"ok","timestamp":1687947327325,"user_tz":-330,"elapsed":2,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":6,"outputs":[]},{"cell_type":"code","source":["import zipfile\n","zip_ref = zipfile.ZipFile('/content/drive/MyDrive/ai_resized.zip')\n","zip_ref.extractall('/content')\n","zip_ref.close()"],"metadata":{"id":"vbUl-2mU3TK-","executionInfo":{"status":"ok","timestamp":1687947356364,"user_tz":-330,"elapsed":5433,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":7,"outputs":[]},{"cell_type":"code","source":["import zipfile\n","zip_ref = zipfile.ZipFile('/content/drive/MyDrive/real_resized.zip')\n","zip_ref.extractall('/content')\n","zip_ref.close()"],"metadata":{"id":"BTSBjeNoL4rE","executionInfo":{"status":"ok","timestamp":1687947359818,"user_tz":-330,"elapsed":3461,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":8,"outputs":[]},{"cell_type":"code","source":["import zipfile\n","zip_ref = zipfile.ZipFile('/content/drive/MyDrive/real_test.zip')\n","zip_ref.extractall('/content')\n","zip_ref.close()"],"metadata":{"id":"KTUVVc6QMUA8","executionInfo":{"status":"ok","timestamp":1687947364079,"user_tz":-330,"elapsed":1612,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":9,"outputs":[]},{"cell_type":"code","source":["import zipfile\n","zip_ref = zipfile.ZipFile('/content/drive/MyDrive/Real.zip')\n","zip_ref.extractall('/content')\n","zip_ref.close()"],"metadata":{"id":"MphEg3XWR3Oy","executionInfo":{"status":"ok","timestamp":1687947373016,"user_tz":-330,"elapsed":3041,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":10,"outputs":[]},{"cell_type":"code","source":["import zipfile\n","zip_ref = zipfile.ZipFile('/content/drive/MyDrive/ai_test.zip')\n","zip_ref.extractall('/content')\n","zip_ref.close()"],"metadata":{"id":"BeVdgiXqSCR3","executionInfo":{"status":"ok","timestamp":1687947376956,"user_tz":-330,"elapsed":1477,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":11,"outputs":[]},{"cell_type":"code","source":["import zipfile\n","zip_ref = zipfile.ZipFile('/content/drive/MyDrive/AI generated.zip')\n","zip_ref.extractall('/content')\n","zip_ref.close()"],"metadata":{"id":"8dfHDMHqSUGx","executionInfo":{"status":"ok","timestamp":1687947381669,"user_tz":-330,"elapsed":1687,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":12,"outputs":[]},{"cell_type":"code","source":["import os\n","from keras.preprocessing.image import ImageDataGenerator\n","from keras.optimizers import Adam\n","from keras.metrics import categorical_crossentropy"],"metadata":{"id":"hIBim07njtqv","executionInfo":{"status":"ok","timestamp":1687947385375,"user_tz":-330,"elapsed":5,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":13,"outputs":[]},{"cell_type":"code","source":["train_dir = \"/content/train\"\n","test_dir = \"/content/test\"\n","eval_dir = \"/content/eval\""],"metadata":{"id":"gRGUwKzZjlkK","executionInfo":{"status":"ok","timestamp":1687947446822,"user_tz":-330,"elapsed":4,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":14,"outputs":[]},{"cell_type":"code","source":["model.compile(optimizer='adam',\n"," loss='binary_crossentropy',\n"," metrics=['accuracy'])"],"metadata":{"id":"4lvofCtPx_CN","executionInfo":{"status":"ok","timestamp":1687947459636,"user_tz":-330,"elapsed":5,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":15,"outputs":[]},{"cell_type":"code","source":["train_datagen = ImageDataGenerator(\n"," rescale=1.0/255,\n"," rotation_range=40,\n"," width_shift_range=0.2,\n"," height_shift_range=0.2,\n"," shear_range=0.2,\n"," zoom_range=0.2,\n"," horizontal_flip=True,\n"," fill_mode='nearest')"],"metadata":{"id":"ddtPeP74mr5T","executionInfo":{"status":"ok","timestamp":1687947461756,"user_tz":-330,"elapsed":5,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":16,"outputs":[]},{"cell_type":"code","source":["eval_datagen = ImageDataGenerator(rescale=1.0/255)\n","test_datagen = ImageDataGenerator(rescale=1.0/255)"],"metadata":{"id":"RiW5RpUonaHx","executionInfo":{"status":"ok","timestamp":1687947464395,"user_tz":-330,"elapsed":7,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":17,"outputs":[]},{"cell_type":"code","source":["train_generator = train_datagen.flow_from_directory(\n"," train_dir,\n"," target_size=(300,300),\n"," batch_size=32,\n"," class_mode='binary')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"5Vpe5itvndH2","executionInfo":{"status":"ok","timestamp":1687947467096,"user_tz":-330,"elapsed":13,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"b3ed110a-bcb1-494a-86cb-f94646a5bc41"},"execution_count":18,"outputs":[{"output_type":"stream","name":"stdout","text":["Found 13550 images belonging to 2 classes.\n"]}]},{"cell_type":"code","source":["eval_generator = eval_datagen.flow_from_directory(\n"," eval_dir,\n"," target_size=(300,300),\n"," batch_size=32,\n"," class_mode='binary')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"3xX6Us95oIWc","executionInfo":{"status":"ok","timestamp":1687947469571,"user_tz":-330,"elapsed":4,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"b850dbb8-2a75-419f-e2b9-b0cd645223a7"},"execution_count":19,"outputs":[{"output_type":"stream","name":"stdout","text":["Found 2000 images belonging to 2 classes.\n"]}]},{"cell_type":"code","source":["test_generator = test_datagen.flow_from_directory(\n"," test_dir,\n"," target_size=(300,300),\n"," batch_size=32,\n"," class_mode='binary')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"JxDcGs4LqFkN","executionInfo":{"status":"ok","timestamp":1687947471883,"user_tz":-330,"elapsed":5,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"d82c91e7-c8a7-44a4-91df-e5a0f0b32592"},"execution_count":20,"outputs":[{"output_type":"stream","name":"stdout","text":["Found 640 images belonging to 2 classes.\n"]}]},{"cell_type":"code","source":["train_steps = train_generator.n // 32\n","eval_steps = eval_generator.n // 32\n","test_steps = test_generator.n // 32"],"metadata":{"id":"AsSTeIqgumEH","executionInfo":{"status":"ok","timestamp":1687947473678,"user_tz":-330,"elapsed":5,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":21,"outputs":[]},{"cell_type":"code","source":["checkpoint_path = \"/content/drive/MyDrive/checkpoints/cp.ckpt\"\n","checkpoint_dir = os.path.dirname(checkpoint_path)\n","\n","# Create a callback that saves the model's weights\n","cp_callback = keras.callbacks.ModelCheckpoint(filepath=checkpoint_path,\n"," save_weights_only=True,\n"," verbose=1)\n"],"metadata":{"id":"U6WYsavRhEiH","executionInfo":{"status":"ok","timestamp":1687947828206,"user_tz":-330,"elapsed":908,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":23,"outputs":[]},{"cell_type":"code","source":["history = model.fit(\n"," train_generator,\n"," steps_per_epoch=train_steps,\n"," epochs=43,\n"," validation_data=eval_generator,\n"," validation_steps=eval_steps,callbacks=[cp_callback])"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"8_2xVqITOKdk","executionInfo":{"status":"ok","timestamp":1687963768109,"user_tz":-330,"elapsed":15933697,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"5834e57d-d72e-48e6-ac66-108c33c98d13"},"execution_count":24,"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/43\n","423/423 [==============================] - ETA: 0s - loss: 0.6162 - accuracy: 0.7296\n","Epoch 1: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 366s 865ms/step - loss: 0.6162 - accuracy: 0.7296 - val_loss: 0.4409 - val_accuracy: 0.7989\n","Epoch 2/43\n","423/423 [==============================] - ETA: 0s - loss: 0.4615 - accuracy: 0.7818\n","Epoch 2: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 359s 849ms/step - loss: 0.4615 - accuracy: 0.7818 - val_loss: 0.3662 - val_accuracy: 0.8392\n","Epoch 3/43\n","423/423 [==============================] - ETA: 0s - loss: 0.4454 - accuracy: 0.7921\n","Epoch 3: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 356s 842ms/step - loss: 0.4454 - accuracy: 0.7921 - val_loss: 0.4337 - val_accuracy: 0.7913\n","Epoch 4/43\n","423/423 [==============================] - ETA: 0s - loss: 0.4238 - accuracy: 0.8020\n","Epoch 4: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 370s 874ms/step - loss: 0.4238 - accuracy: 0.8020 - val_loss: 0.3559 - val_accuracy: 0.8397\n","Epoch 5/43\n","423/423 [==============================] - ETA: 0s - loss: 0.4175 - accuracy: 0.8043\n","Epoch 5: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 359s 849ms/step - loss: 0.4175 - accuracy: 0.8043 - val_loss: 0.3549 - val_accuracy: 0.8387\n","Epoch 6/43\n","423/423 [==============================] - ETA: 0s - loss: 0.4074 - accuracy: 0.8095\n","Epoch 6: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 360s 851ms/step - loss: 0.4074 - accuracy: 0.8095 - val_loss: 0.3687 - val_accuracy: 0.8513\n","Epoch 7/43\n","423/423 [==============================] - ETA: 0s - loss: 0.4039 - accuracy: 0.8108\n","Epoch 7: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 350s 827ms/step - loss: 0.4039 - accuracy: 0.8108 - val_loss: 0.3475 - val_accuracy: 0.8493\n","Epoch 8/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3949 - accuracy: 0.8174\n","Epoch 8: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 356s 841ms/step - loss: 0.3949 - accuracy: 0.8174 - val_loss: 0.3212 - val_accuracy: 0.8740\n","Epoch 9/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3938 - accuracy: 0.8168\n","Epoch 9: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 354s 837ms/step - loss: 0.3938 - accuracy: 0.8168 - val_loss: 0.3240 - val_accuracy: 0.8725\n","Epoch 10/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3927 - accuracy: 0.8155\n","Epoch 10: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 355s 837ms/step - loss: 0.3927 - accuracy: 0.8155 - val_loss: 0.3361 - val_accuracy: 0.8634\n","Epoch 11/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3811 - accuracy: 0.8218\n","Epoch 11: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 344s 813ms/step - loss: 0.3811 - accuracy: 0.8218 - val_loss: 0.3429 - val_accuracy: 0.8548\n","Epoch 12/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3840 - accuracy: 0.8222\n","Epoch 12: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 350s 826ms/step - loss: 0.3840 - accuracy: 0.8222 - val_loss: 0.3290 - val_accuracy: 0.8579\n","Epoch 13/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3815 - accuracy: 0.8225\n","Epoch 13: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 351s 829ms/step - loss: 0.3815 - accuracy: 0.8225 - val_loss: 0.3698 - val_accuracy: 0.8498\n","Epoch 14/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3819 - accuracy: 0.8257\n","Epoch 14: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 353s 835ms/step - loss: 0.3819 - accuracy: 0.8257 - val_loss: 0.3510 - val_accuracy: 0.8478\n","Epoch 15/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3745 - accuracy: 0.8267\n","Epoch 15: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 360s 852ms/step - loss: 0.3745 - accuracy: 0.8267 - val_loss: 0.3348 - val_accuracy: 0.8644\n","Epoch 16/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3730 - accuracy: 0.8283\n","Epoch 16: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 360s 851ms/step - loss: 0.3730 - accuracy: 0.8283 - val_loss: 0.3229 - val_accuracy: 0.8710\n","Epoch 17/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3664 - accuracy: 0.8327\n","Epoch 17: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 353s 835ms/step - loss: 0.3664 - accuracy: 0.8327 - val_loss: 0.3550 - val_accuracy: 0.8392\n","Epoch 18/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3723 - accuracy: 0.8263\n","Epoch 18: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 359s 848ms/step - loss: 0.3723 - accuracy: 0.8263 - val_loss: 0.3145 - val_accuracy: 0.8735\n","Epoch 19/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3749 - accuracy: 0.8280\n","Epoch 19: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 343s 810ms/step - loss: 0.3749 - accuracy: 0.8280 - val_loss: 0.3247 - val_accuracy: 0.8624\n","Epoch 20/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3673 - accuracy: 0.8302\n","Epoch 20: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 368s 870ms/step - loss: 0.3673 - accuracy: 0.8302 - val_loss: 0.3095 - val_accuracy: 0.8765\n","Epoch 21/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3648 - accuracy: 0.8339\n","Epoch 21: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 365s 864ms/step - loss: 0.3648 - accuracy: 0.8339 - val_loss: 0.3052 - val_accuracy: 0.8664\n","Epoch 22/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3610 - accuracy: 0.8338\n","Epoch 22: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 365s 864ms/step - loss: 0.3610 - accuracy: 0.8338 - val_loss: 0.2791 - val_accuracy: 0.8921\n","Epoch 23/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3661 - accuracy: 0.8313\n","Epoch 23: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 361s 853ms/step - loss: 0.3661 - accuracy: 0.8313 - val_loss: 0.3252 - val_accuracy: 0.8679\n","Epoch 24/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3638 - accuracy: 0.8341\n","Epoch 24: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 374s 883ms/step - loss: 0.3638 - accuracy: 0.8341 - val_loss: 0.3151 - val_accuracy: 0.8624\n","Epoch 25/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3633 - accuracy: 0.8336\n","Epoch 25: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 366s 865ms/step - loss: 0.3633 - accuracy: 0.8336 - val_loss: 0.3152 - val_accuracy: 0.8735\n","Epoch 26/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3611 - accuracy: 0.8342\n","Epoch 26: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 350s 825ms/step - loss: 0.3611 - accuracy: 0.8342 - val_loss: 0.3175 - val_accuracy: 0.8831\n","Epoch 27/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3557 - accuracy: 0.8412\n","Epoch 27: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 358s 846ms/step - loss: 0.3557 - accuracy: 0.8412 - val_loss: 0.3037 - val_accuracy: 0.8906\n","Epoch 28/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3549 - accuracy: 0.8382\n","Epoch 28: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 348s 821ms/step - loss: 0.3549 - accuracy: 0.8382 - val_loss: 0.2730 - val_accuracy: 0.8942\n","Epoch 29/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3554 - accuracy: 0.8369\n","Epoch 29: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 363s 857ms/step - loss: 0.3554 - accuracy: 0.8369 - val_loss: 0.3479 - val_accuracy: 0.8574\n","Epoch 30/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3578 - accuracy: 0.8349\n","Epoch 30: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 345s 817ms/step - loss: 0.3578 - accuracy: 0.8349 - val_loss: 0.3022 - val_accuracy: 0.8770\n","Epoch 31/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3525 - accuracy: 0.8404\n","Epoch 31: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 366s 864ms/step - loss: 0.3525 - accuracy: 0.8404 - val_loss: 0.2862 - val_accuracy: 0.8775\n","Epoch 32/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3544 - accuracy: 0.8386\n","Epoch 32: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 379s 895ms/step - loss: 0.3544 - accuracy: 0.8386 - val_loss: 0.2864 - val_accuracy: 0.8821\n","Epoch 33/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3491 - accuracy: 0.8423\n","Epoch 33: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 360s 851ms/step - loss: 0.3491 - accuracy: 0.8423 - val_loss: 0.2561 - val_accuracy: 0.9007\n","Epoch 34/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3588 - accuracy: 0.8360\n","Epoch 34: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 375s 887ms/step - loss: 0.3588 - accuracy: 0.8360 - val_loss: 0.2983 - val_accuracy: 0.8629\n","Epoch 35/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3467 - accuracy: 0.8396\n","Epoch 35: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 366s 863ms/step - loss: 0.3467 - accuracy: 0.8396 - val_loss: 0.3170 - val_accuracy: 0.8634\n","Epoch 36/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3500 - accuracy: 0.8390\n","Epoch 36: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 359s 846ms/step - loss: 0.3500 - accuracy: 0.8390 - val_loss: 0.2749 - val_accuracy: 0.8936\n","Epoch 37/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3521 - accuracy: 0.8372\n","Epoch 37: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 355s 839ms/step - loss: 0.3521 - accuracy: 0.8372 - val_loss: 0.2809 - val_accuracy: 0.8982\n","Epoch 38/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3497 - accuracy: 0.8424\n","Epoch 38: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 353s 834ms/step - loss: 0.3497 - accuracy: 0.8424 - val_loss: 0.3034 - val_accuracy: 0.8690\n","Epoch 39/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3498 - accuracy: 0.8438\n","Epoch 39: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 352s 831ms/step - loss: 0.3498 - accuracy: 0.8438 - val_loss: 0.2824 - val_accuracy: 0.8906\n","Epoch 40/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3465 - accuracy: 0.8445\n","Epoch 40: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 368s 869ms/step - loss: 0.3465 - accuracy: 0.8445 - val_loss: 0.2741 - val_accuracy: 0.8795\n","Epoch 41/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3415 - accuracy: 0.8438\n","Epoch 41: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 357s 845ms/step - loss: 0.3415 - accuracy: 0.8438 - val_loss: 0.2749 - val_accuracy: 0.8931\n","Epoch 42/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3489 - accuracy: 0.8423\n","Epoch 42: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 354s 835ms/step - loss: 0.3489 - accuracy: 0.8423 - val_loss: 0.2868 - val_accuracy: 0.8967\n","Epoch 43/43\n","423/423 [==============================] - ETA: 0s - loss: 0.3433 - accuracy: 0.8466\n","Epoch 43: saving model to /content/drive/MyDrive/checkpoints/cp.ckpt\n","423/423 [==============================] - 350s 827ms/step - loss: 0.3433 - accuracy: 0.8466 - val_loss: 0.2597 - val_accuracy: 0.9057\n"]}]},{"cell_type":"code","source":["model.save('my_model.keras')"],"metadata":{"id":"eII0lQ5ORsZO","executionInfo":{"status":"ok","timestamp":1687963784960,"user_tz":-330,"elapsed":1917,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":25,"outputs":[]},{"cell_type":"code","source":["model.save('/content/drive/MyDrive/image_classify.keras')"],"metadata":{"id":"oMQ-DS3tRuZ3","executionInfo":{"status":"ok","timestamp":1687963794108,"user_tz":-330,"elapsed":5009,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":26,"outputs":[]},{"cell_type":"code","source":["loss, acc = model.evaluate(train_generator,\n"," steps=train_steps, verbose=1)\n","print('Training data -> loss: %.3f, acc: %.3f' % (loss, acc))\n","loss, acc = model.evaluate(eval_generator,\n"," steps=eval_steps, verbose=1)\n","print('Cross-val data -> loss: %.3f, acc: %.3f' % (loss, acc))\n","loss, acc = model.evaluate(test_generator,\n"," steps=test_steps, verbose=1)\n","print('Testing data -> loss: %.3f, acc: %.3f' % (loss, acc))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"fHxcoDRNPbeV","executionInfo":{"status":"ok","timestamp":1687964202626,"user_tz":-330,"elapsed":402542,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"186491c8-5c9b-45c5-acda-8afbd89b646d"},"execution_count":27,"outputs":[{"output_type":"stream","name":"stdout","text":["423/423 [==============================] - 322s 760ms/step - loss: 0.3042 - accuracy: 0.8744\n","Training data -> loss: 0.304, acc: 0.874\n","62/62 [==============================] - 14s 219ms/step - loss: 0.2596 - accuracy: 0.9057\n","Cross-val data -> loss: 0.260, acc: 0.906\n","20/20 [==============================] - 5s 226ms/step - loss: 0.4725 - accuracy: 0.8062\n","Testing data -> loss: 0.473, acc: 0.806\n"]}]},{"cell_type":"code","source":["model.save('/content/drive/MyDrive/my_model.h5')"],"metadata":{"id":"qqg4Jjp2Us_R","executionInfo":{"status":"ok","timestamp":1687964216228,"user_tz":-330,"elapsed":6550,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}}},"execution_count":28,"outputs":[]},{"cell_type":"code","source":["import matplotlib.pyplot as plt\n","\n","plt.plot(history.history['accuracy'],color='red',label='train')\n","plt.plot(history.history['val_accuracy'],color='blue',label='validation')\n","plt.legend()\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":430},"id":"0euQKTDPgcr0","executionInfo":{"status":"ok","timestamp":1687964256036,"user_tz":-330,"elapsed":911,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"81461c16-3666-4245-9a86-c181cf895977"},"execution_count":29,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAjUAAAGdCAYAAADqsoKGAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB1lUlEQVR4nO3deVxU1fsH8M+ArLIoLiBumJpmrqEQ6q/MKEozNTNLc9es0FKyXMq1EtvMFpf6pt/qm6ZZtmlZRmpZuISZuZH7CrgCgsoynN8fT5dhWGdgdj7v1+u+uDNz594DA8wz5zznOTqllAIRERGRk3OzdwOIiIiILIFBDREREbkEBjVERETkEhjUEBERkUtgUENEREQugUENERERuQQGNUREROQSGNQQERGRS6hh7wbYSkFBAc6ePQt/f3/odDp7N4eIiIhMoJTClStXEBoaCje38vtiqk1Qc/bsWTRu3NjezSAiIqJKOHXqFBo1alTuMdUmqPH39wcgP5SAgAA7t4aIiIhMkZmZicaNGxe+j5en2gQ12pBTQEAAgxoiIiInY0rqCBOFiYiIyCUwqCEiIiKXwKCGiIiIXEK1yakxhVIK+fn50Ov19m4KWYC7uztq1KjBKfxERNUEg5p/5ebmIiUlBVevXrV3U8iCfH190aBBA3h6etq7KUREZGUMaiCF+Y4dOwZ3d3eEhobC09OTn+6dnFIKubm5OH/+PI4dO4aWLVtWWLSJiIicG4MaSC9NQUEBGjduDF9fX3s3hyzEx8cHHh4eOHHiBHJzc+Ht7W3vJhERkRXxo2sR/CTveviaEhFVH/yPT0RERC6BQQ0RERG5BAY1VCgsLAwLFy60dzOIiIgqhYnCTq5Hjx7o2LGjRYKRnTt3ombNmlVvFBERkR2wp8bFaQUFTVGvXj3O/iIiIrNlZgJ33QX88ot928GgpixKAdnZ9tmUMqmJI0aMwJYtW/DWW29Bp9NBp9Phww8/hE6nw/fff4/w8HB4eXlh69atOHLkCPr27Yvg4GD4+fmhS5cu+Omnn4zOV3z4SafT4YMPPkD//v3h6+uLli1b4ptvvrHkT5mIiFzAjBnATz8BY8YAJn6OtgoGNWW5ehXw87PPZmJV47feegtRUVEYO3YsUlJSkJKSgsaNGwMApk6divnz5+PAgQNo3749srKy0KtXLyQkJODPP//EPffcgz59+uDkyZPlXmPOnDl46KGHsGfPHvTq1QtDhgzBpUuXqvzjJSIi15CUBLz7ruwvXgzUsGNiC4MaJxYYGAhPT0/4+voiJCQEISEhcHd3BwDMnTsXd911F5o3b46goCB06NAB48aNQ9u2bdGyZUu8+OKLaN68eYU9LyNGjMAjjzyCFi1aYN68ecjKysKOHTts8e0REZGD0+uBceOAggJg8GAgOtq+7WGicFl8fYGsLPtdu4o6d+5sdDsrKwuzZ8/G+vXrkZKSgvz8fFy7dq3Cnpr27dsX7tesWRMBAQE4d+5cldtHRETOb9Ei6akJDAQWLLB3axjUlE2nA5x4JlDxWUyTJ0/Gxo0b8frrr6NFixbw8fHBgw8+iNzc3HLP4+HhYXRbp9OhoKDA4u0lIiLncuYM8MILsj9/PhAcbN/2AAxqnJ6npyf0en2Fx/32228YMWIE+vfvD0B6bo4fP27l1hERkauaOBG4cgW49Vbgscfs3RrBnBonFxYWhu3bt+P48eO4cOFCmb0oLVu2xNq1a7F792789ddfGDx4MHtciIioUr77Dvj8c8DdHXjvPcBRltmrVDMWLVqEsLAweHt7IzIystzE0by8PMydOxfNmzeHt7c3OnTogA0bNph9zuvXryM2NhZ16tSBn58fBgwYgLS0tMo036VMnjwZ7u7uaNOmDerVq1dmjsyCBQtQu3ZtdO3aFX369EFMTAxuueUWG7eWiIic3dWrQGys7E+aBBRJvbQ/ZaZVq1YpT09PtXz5crVv3z41duxYVatWLZWWllbq8c8995wKDQ1V69evV0eOHFGLFy9W3t7eateuXWad8/HHH1eNGzdWCQkJ6o8//lC33nqr6tq1q8ntzsjIUABURkZGiceuXbum9u/fr65du2bGT4KcAV9bIiLLmjpVKUCpxo2VunLF+tcr7/27OLODmoiICBUbG1t4W6/Xq9DQUBUfH1/q8Q0aNFDvvvuu0X0PPPCAGjJkiMnnTE9PVx4eHmrNmjWFxxw4cEABUImJiSa1m0FN9cTXlsh5vfeeUu3bK3XypL1bQpq//1aqRg0Jar7+2jbXNCeoMWv4KTc3F0lJSYguMhHdzc0N0dHRSExMLPU5OTk58Pb2NrrPx8cHW7duNfmcSUlJyMvLMzqmdevWaNKkSbnXzczMNNqIiMg5KAXMmQPs2QN8+aW9W0OA1KJ5/HGpGNyvH3D//fZuUUlmBTUXLlyAXq9HcLF5W8HBwUhNTS31OTExMViwYAEOHTqEgoICbNy4EWvXrkVKSorJ50xNTYWnpydq1apl8nXj4+MRGBhYuGmVdomIyPHt2gWcPSv7//xj37aQWL4c+O03qXby9tv2bk3prJ6v/NZbb6Fly5Zo3bo1PD09MX78eIwcORJuVk6VnjZtGjIyMgq3U6dOWfV6RERkOd9+a9hnUGN/588Dzz0n+3PnAo7aT2BWZFG3bl24u7uXmHWUlpaGkJCQUp9Tr149fPXVV8jOzsaJEydw8OBB+Pn54YYbbjD5nCEhIcjNzUV6errJ1/Xy8kJAQIDRRkREzsHZghqlZKVqVzV5MnD5MtCxI/DUU/ZuTdnMCmo8PT0RHh6OhISEwvsKCgqQkJCAqKiocp/r7e2Nhg0bIj8/H1988QX69u1r8jnDw8Ph4eFhdExycjJOnjxZ4XWJiMi5nD4tw0+akyeBa9fs156KnDolBejq1gX277d3ayxv0ybg44+l0P5779l3wcqKmN20uLg4DB8+HJ07d0ZERAQWLlyI7OxsjBw5EgAwbNgwNGzYEPHx8QCA7du348yZM+jYsSPOnDmD2bNno6CgAM9p/VgmnDMwMBCjR49GXFwcgoKCEBAQgAkTJiAqKgq33nqrJX4ORETkINatk69RUcCBA0B6OnDkCNC2rV2bVapffwUefBDQlsT78UegTRv7tsmScnIkORgAnngCiIiwb3sqYnZQM2jQIJw/fx4zZ85EamoqOnbsiA0bNhQm+p48edIoX+b69et44YUXcPToUfj5+aFXr1743//+Z5T0W9E5AeDNN9+Em5sbBgwYgJycHMTExGDx4sVV+NaJiMgRaUNPffrIjJvt22UIypGCGqWk12LCBJkN5OkJ5OYCf/5p75ZZ1iuvyM8+JASYN8/erTGBDaaYOwTWqSld06ZN1Ztvvll4G4D68ssvyzz+2LFjCoD6888/q3RdS52nItX5tSVyRllZSnl5SR2Uv/9WauhQ2Z83z94tM7h+XamxY6VdgFKDBin16aey37atvVtnGVlZSk2erJS7u3xfn35qv7aYU6fGgUfGyB5SUlJQu3Zti55zxIgRSE9Px1dffVV4X+PGjZGSkoK6deta9FpE5Nw2bpQhj2bNgJtvBm68Ue53lGThlBRgwAAgMVFyTOLjZVbQmTPy+IEDkv/j42PfdlbFd98BTz4JnDght0ePBgYNsm+bTOUgS1CRowgJCYGXl5fVr+Pu7o6QkBDUcOSMMyKyuaJDTzqdYwU127cD4eES0AQGypv/lCnSzoYNJVFYrwf27rV3SysnJQV46CGgd28JaJo2lfymDz6Q79EZMKhxYu+//z5CQ0NLrLbdt29fjBo1CkeOHEHfvn0RHBwMPz8/dOnSBT/99FO559TpdEY9Kjt27ECnTp3g7e2Nzp07489iA8Z6vR6jR49Gs2bN4OPjg1atWuGtt94qfHz27Nn46KOP8PXXX0On00Gn02Hz5s04fvw4dDoddu/eXXjsli1bEBERAS8vLzRo0ABTp05Ffn5+4eM9evTAU089heeeew5BQUEICQnB7Nmzzf/BEZFDKigA1q+X/T595KujBDX//S9w223yxn/TTcDOncA99xge1+mATp1kv+jMLWdQUAAsWQK0bg2sWSMrb0+eDOzbJwGOM2FQUwalgOxs+2xKmdbGgQMH4uLFi9i0aVPhfZcuXcKGDRswZMgQZGVloVevXkhISMCff/6Je+65B3369ClzJe/isrKycN9996FNmzZISkrC7NmzMXnyZKNjCgoK0KhRI6xZswb79+/HzJkzMX36dHz22WcAZBXxhx56CPfccw9SUlKQkpKCrl27lrjWmTNn0KtXL3Tp0gV//fUXlixZgmXLluGll14yOu6jjz5CzZo1sX37drz66quYO3cuNm7caNoPjMhFZWUB//sfUEaBdaexcyeQlgYEBEgAAQAtWsjXCxeAS5ds36a8PEkGHjVKEoH79ZMem5YtSx57yy3y1ZmShf/+G+jWTYabMjOBLl2AP/4AXntNKgc7HRvk+DgEcxOFs7IMSWC23rKyTP+++vbtq0aNGlV4+7333lOhoaFKr9eXevzNN9+s3nnnncLb5SUKv/fee6pOnTpGP5clS5ZUmOAbGxurBgwYUHh7+PDhqm/fvkbHFE8Unj59umrVqpUqKCgoPGbRokXKz8+v8Hu5/fbbVffu3Y3O06VLFzVlypQy28JEYXJ133wjqyUDSg0ebO/WVM306fJ9PPSQ8f0NG8r927bZtj2XLyt1++2G/82zZytVxr9WpZRSq1bJcRERtmph5WVnKzVlimFxSn9/pd5+W6n8fHu3rCSrLWhJjmfIkCH44osvkJOTAwBYsWIFHn74Ybi5uSErKwuTJ0/GTTfdhFq1asHPzw8HDhwwuafmwIEDaN++vdGCpKUVO1y0aBHCw8NRr149+Pn54f333zf5GkWvFRUVBV2Rgdtu3bohKysLp0+fLryvffv2Rs9r0KABzmkFIoiqkbNnpT7K/fdL8TcAKDKa65SK5tMUZa8hqIULgS1bAD8/4KuvgFmzgPJW+NGGn/bskWnejur332V6/CuvSDv795eigRMmyNCTM2OWZhl8faVL117XNlWfPn2glML69evRpUsX/Prrr3jzzTcByNDPxo0b8frrr6NFixbw8fHBgw8+iNzcXIu1ddWqVZg8eTLeeOMNREVFwd/fH6+99hq2b99usWsU5eHhYXRbp9OVyCkicmV6PbB0KTBtGnDlirwJjRgBLFsGHD4sjzvjG9OJEzIU4uYG3Huv8WM33ihVbW0d1Gj/xuLjgX+L4JerRQsJgLKygIMHHauujiYvT4LhlBSgUSPg3XdN+96cBYOaMuh0zjGe6O3tjQceeAArVqzA4cOH0apVK9zy78Dub7/9hhEjRqB///4AJEfm+PHjJp/7pptuwv/+9z9cv369sLdm27ZtRsf89ttv6Nq1K5588snC+44cOWJ0jKenJ/R6fYXX+uKLL6CUKuyt+e233+Dv749GjRqZ3GaqnrZtkzegp55ynlkalfHXX8C4cYY328hI4P33ZerzJ5/IVOjjx4Hmze3azErRemm6dQPq1DF+rFUr+WrroOavv+RreLhpx7u5ydpIW7dKXo0jBjVffSUBTUiI9M74+9u7RZbF4ScXMGTIEKxfvx7Lly/HkCFDCu9v2bIl1q5di927d+Ovv/7C4MGDzerVGDx4MHQ6HcaOHYv9+/fju+++w+uvv250TMuWLfHHH3/ghx9+wD///IMZM2Zg586dRseEhYVhz549SE5OxoULF5CXl1fiWk8++SROnTqFCRMm4ODBg/j6668xa9YsxMXFWX1Fd3J+w4cDEycaZs64muxsqYUSHi4BTUAAsGgR8NtvQPv20jOjDdEkJ9u3rZX1zTfy9f77Sz5mj+8tLU3e/HU6oF0705+nDUE5arKwVoh/7FjXC2gABjUuoWfPnggKCkJycjIGDx5ceP+CBQtQu3ZtdO3aFX369EFMTExhL44p/Pz88O233+Lvv/9Gp06d8Pzzz+OVV14xOmbcuHF44IEHMGjQIERGRuLixYtGvTYAMHbsWLRq1QqdO3dGvXr18Ntvv5W4VsOGDfHdd99hx44d6NChAx5//HGMHj0aL7zwgpk/Dapuzp83fIL/5Rf7tsUavv9ePvG/9poMLT34oBR4e/JJ42EmrTfDGYOazExg82bZL55PAxiCmkOHZPqxLWi9NNqQkqkcOajZv19+zu7uwGOP2bs1VmL1tGUHwWUSqie+tq7v228Ns1OiouzdGsvJzVXq0UcN31uTJvK9luX55+W4ceNs10ZLWbNG2t6yZemP5+YaZumcOmWbNr36qlxv4EDznrd7tzwvMFCpIpM5K+XqVaUef1yp9eurdh7N+PHStn79LHM+W+HsJyKqNoqmef3xh5SodwUzZkiejJsb8MwzUgjtvvvKPt6Ze2rKmvWk8fAAbrhB9m2VV6PNJOvQwbzntWkji1tmZADHjlWtDStWSFL4kCFyvqrIygI+/lj2i3WmuxQGNUTk1IoGNXl5Etg4uw0bZLotAKxaBbz+esVDIFpQc/CgddtmaXq9IReqtHwaja2ndWtBTceO5j3Pw8OQIFzVIaiff5av6ekyvbwqVq6UYb6WLYE776zauRwZgxoiclp6vWEmkPamV0rKllM5cwYYOlT2Y2OBgQNNe54W1KSmypuXs0hMBC5eBGrXlplPZbFlUHPtmqHHy9ygBrDMcglKGYIaAFiwALh8ufLnWrRI9p94ovxaO87Ohb81InJ1+/dLt3rNmobEx61b7dumqsjPBwYPliUBOnaUHhpTBQYCwcGy70xDUNrQ0733AuWtb2vLoGbfPgmY69QBQkPNf74llkvYv19mYPn4yJT9zEwJbCojMVEKAvr4SE0jV8aghoicljb0FBFhWCvo999tN0PG0ubOlRlcfn7AZ58BRYp5m6R1a/nqTEGNNpW7rHwajS2DGm3mU8eOlat7ZIkZUAkJ8rV7d+DFF2X/rbekV8tc2jTuRx6RHjFXxqCmCGXqSpLkNPiaujYtqLn1VnkD8vWVLvoDB+zarEpJSAC09Vvff7/0BRMr4mzJwocPSw5QjRrGK16XRgtqjh6V3Clrqmw+jaZ9ewmGUlOl1k1laENPd94pi2h26iQVpN94w7zznDsnK28Drp0grGFQA0Pp/atXr9q5JWRp2mtafHkFcg1FgxoPD6mwCzhfXk1qqsxwUUqKoj3ySOXO42xBjTb0dNttQK1a5R8bGirDjHp91WcVVaSyM580NWsaXovK9Nbk5xvq9vTsKQHSnDly++23pTaTqZYvl9XFIyJMr4zszLhMAgB3d3fUqlWrcGFEX19fo4UVyfkopXD16lWcO3cOtWrVgrszLoZD5UpPl7wDwBDMdO8uawRt3eo8xcX0euDRRyV/om3bqs1ycbYZUBVN5S5Kp5Pemj//lKBN67mxtIIC4+GnyrrlFnkd/vwT6NXLvOf++adM4Q4MNOTn3Hcf0LmzzO577TXg1VcrPo9eD7z3nuw/8YR5bXBWDGr+FRISAgBc8dnF1KpVq/C1JdeircbRrJkhQVabPeNMPTXx8TL05OsreTTmLGhbnBbUaJV3HXmWy+XLhgrQpgQ1gCGosWZezfHjMszj6WnIUaqMTp1kGnVlemq0fJoePQxVo7Xemt69ZRHKZ54x/N6XZcMG+X5q1wYGDTK/Hc6IQc2/dDodGjRogPr165e6NhE5Hw8PD/bQuLCiQ0+aqCh5Iz96FDh7tnIzV2xpyxZg1izZX7wYuOmmqp0vLEyG4a5fB06elNuOasMG6Ulo08b0BThtkSysDT3dfLP8LCurKsnCRfNpirr3XumV3L5d6hhVNBtKSxAeNUpmPlUHDGqKcXd35xshkRMoLagJCJDFB//6S3prTK3xYg/nz8v07YICWZBz+PCqn7NGDUkw3r9fhmgcOagxZ+hJY4ugxhJDT4AhqDl6VIZKK8oZ0uTkGMoS9Oxp/JhOJzPkYmKAJUuAyZPLDtyPHpV1wwDg8cfNbb3zcuDOSSKi0ilVelADSF4N4NhDUAUFwLBh0pvUurWhMJolOEOycF6e4Q3X0YKaqs580gQFAU2bGp/TFNu2SfG/4GDpxSrurrtkmPX6dWD+/LLP89578ncSEyOLclYXDGqIyOkcPgxcugR4eZV889GCGkcuwvf66zL84u0teTQ1a1ru3M4Q1GzdKr0XdeuWDErLo01zP3tWii5aQ1VnPhVVmSEoLZ9Gm/VUnNZbA0jgcvp0yWOuXweWLZP96jCNuygGNUTkdLRemvBwSegsSksW3r3bem98VfH778D06bL/zjsyXGZJzhDUaENPvXsbEmFNUbs2UK+e7B86ZPl2Xb4suUiAZYMac5ZLKCufpqg77gBuv12mas+bV/LxNWukSF+TJvIzrk4Y1BCR0ylr6AkAGjeWf+ZF14VyFNeuyfRtvV5q0YwebflrOPq0bqUql0+jseYQlJZPExZmeg5MecxdLiEry/A7WzyfpqiidWs++MAQiGmWLJGv48aZFzS6AgY1ROR0ygtqAMed2v3661I4rmFDYOnSypXgr4gW1Jw545g9VcnJMnzo6Qncfbf5z9e+P2sGNZbopQEMPTUHD0pAW5Fff5XCe82ayVae22+X3py8PODllw33//mnrPXk4WGdoNnRMaghIqdy9arhzaesoMYR82pOnpSaNIAENwEB1rlOUJDkqgC2WSfJHCkpwMiRst+jB+Dvb/45tJ4aawyvWSpJWBMaKsNlej3w998VH180n8YUWm/N8uWGKstaL82AARXXsXFFDGqIyKkkJcmbRGgo0KhR6cdoPTWJifLJ1xE8+6x8Wr/tNusXQnPEhS23bZMcqG3bZGhHS3Y1lzWHnywd1Oh05uXVmJJPU1S3bjK7KT9f1g1LTwdWrJDHqluCsIZBDRE5laJDT2UN37RtKz0hWVmmfUK2tk2bZJaTm5us3WPtVVgcLVl4+XIZLklJkaJ2O3calrYwV9GgxpLr1ebmGpbdsNTwE2B6Xs3Fi4ag6o47TD+/1lvz0UfAzJnSk9m2raG3srphUENETqWifBpAkiOjomTf3kNQ+fnAU0/J/hNPWPYNsyyOEtTk5QETJkhuR24u0L+/9J5VpW5K8+YSFGZkmLewY0UOHpQ2BgRYtmihqdO6N2+WIO3mmwFzVnaJjJS1pfR6mU0HSC9NdV2+kEENETkNpeRNEai4vomjFOFbsgTYu1dyXSo75GIuR5gBdf68FIp79125PXcu8PnnlcujKcrb21DUzpJDUEXr01gyINCCmj17JMgrizb0ZGo+TVFabw0A+PnJDLvqikENETmN06dlCMPdXfIzyqPl1WzdatlhCnOcPy9DAoDMUAkKss11i84QKiiwzTWL2rVLVpTeskWCmK+/BmbMsNwCm9bIq7HU8gjFNW8uP4OcnPKDTC1J2NR8mqI6dwbuv1/2hw2reuDozBjUEJHT0HppOnSoeDXriAhZC+nMmZJ1PGzl+eclebNjR2DsWNtd94Yb5Hu/elW+f1tauVICypMnpQLw9u2GN1xLsUZQY+kkYY2bm+GcZQ1BnTkjQ4VubpJ7VBnLlwMLF5a/dEJ1wKCGiJyGKfk0mpo1DUma9sirSUqSwmiA5DrYsgiah4dh5Wtb5dXk58sMryFDpEx/r17Ajh1VX3m8NJYOapSyfI2aoirKq9GGnsLDK1/0r04d4Omnq3cvDcBVuonIiZgT1ADSY7Bjh+TVDBlivXYVV1AgCbJKyXXtMROlVSsJaJKTgejoqp9PKel1SksruZ07J2/Yf/whx06fLjk01grkLF2A78wZmX3k7i6JupZW0bTuquTTkDEGNUTkFHJyDG8KpgY13bsDb75p+56aFStkqKxmTeCVV2x7bY0lZkAtXAh88okhcMnNLf94X1/gww+BgQMrf01TaD01hw/LrJ+qBk/a0NNNN0kisqVpPYa7d0vAWzS3SKmq5dOQsUoNPy1atAhhYWHw9vZGZGQkduzYUe7xCxcuRKtWreDj44PGjRtj0qRJuH79euHjYWFh0Ol0JbbY2NjCY3r06FHi8ccff7wyzSciO0lPBx56CPj4Y/Of+9dfEtgEBZk+JVhLFt67V65tC5mZwHPPyf6MGbIkgj1UNajJyJDhpKQkSdDWApqAAMmV6d5dqtY++aTMvlm6VGoCWTugAWR9Ly8v+X2wRL6UNYeeAAmWvLzkd0Or/Ks5cgQ4dUqWjdB+X6nyzO6pWb16NeLi4rB06VJERkZi4cKFiImJQXJyMurXr1/i+JUrV2Lq1KlYvnw5unbtin/++QcjRoyATqfDggULAAA7d+6EXq8vfM7evXtx1113YWCxv46xY8dibpE5kb4VZQoSkUNZvFhWEP7mG0mI1KbmmsKUonvFBQdLAHT4sPSc3Huv+W0210svAamp8sY/caL1r1eWqk7r/uknyZNp3lySf4ODZbNGT4a53N3ldd23T4agKlonqSLWShLWeHhIQbykJBmm0/KdAEMvTVRUxcnvVDGze2oWLFiAsWPHYuTIkWjTpg2WLl0KX19fLF++vNTjf//9d3Tr1g2DBw9GWFgY7r77bjzyyCNGvTv16tVDSEhI4bZu3To0b94ctxdLA/f19TU6LsBai6cQkcUVFAD/+Y/s5+RI3oU5zM2n0RSd2m1tyckyZAPIVy8v61+zLFpQc/KkzIIy1/ffy9c+fWQmWdOmjhHQaCyZLGztoAYoO6+G+TSWZVZQk5ubi6SkJEQXyTpzc3NDdHQ0ErW5lsV07doVSUlJhUHM0aNH8d1336FXr15lXuOTTz7BqFGjoCv2cWzFihWoW7cu2rZti2nTpuFqOX+pOTk5yMzMNNqIyH5++gk4ftzwaXTlSkniNVVlgxpbFeFTSmaf5OUBvXvL7B97qlsXqF1b9g8dMu+5ShmCGlv0blWGpYKaK1dkCAiwbrXn0pZLKCgwf70nKp9Zw08XLlyAXq9HcLGlP4ODg3GwjD7OwYMH48KFC+jevTuUUsjPz8fjjz+O6WV8TPvqq6+Qnp6OESNGlDhP06ZNERoaij179mDKlClITk7G2rVrSz1PfHw85hQts0hEdvX++/J11CjJ1/jf/4DJk6VAW0XDSWlpkoug00mvgTm0nprt2yUvxNPT/LabYt064Icf5Pxab4096XSysGViovQgmfOGvWcPcPasBKC33Wa9NlaFpYKav/+WIE5bUdtaivbUKCWvz969wIULklDepYv1rl2dWL1OzebNmzFv3jwsXrwYu3btwtq1a7F+/Xq8+OKLpR6/bNky3HvvvQgNDTW6/7HHHkNMTAzatWuHIUOG4OOPP8aXX36JI1qIXcy0adOQkZFRuJ06dcri3xuZ5/p1+dTcvDnw2GOSW3Hxor1bRbaQmipVZQF57V9+WYYyfv0V+Oqrip+/fbt8vekmIDDQvGu3bi01PK5fr3j9ncq6ft2QPxMXV7W1jSypssnCWi9Nz56ONeRUlKWCGlsMPQFA+/Yy6+ncOamKDRjyaW67zXrBdnVjVlBTt25duLu7Iy0tzej+tLQ0hJSxAteMGTMwdOhQjBkzBu3atUP//v0xb948xMfHo6BY/e4TJ07gp59+wpgxYypsS+S/S7wePny41Me9vLwQEBBgtJF9JSbKEMDRo5Jb8dBD8smoSxdg2jTphs3JsXcrSTN3rswy0VYurooPP5Sk06gooF07Oe8zz8hjzz1X8VThyg49AfKJuGtX2bdWXs3ChfJ7HRoqVYQdRVWDGkcdegIMQc2JExJUVpa1Zz5pfH0lwAYMwTXzaSzPrKDG09MT4eHhSNDCSwAFBQVISEhAlLYkbjFXr16FW7EFP9z/LSqgii3I8t///hf169dH7969K2zL7n/D6wYNGpjzLZAd/f67fO3WTT7V3nyzdMP+8YeU9r7zTskBiIkBXn9d/tnYa80ekmnXp0+bn9BbXNEE4cceM9w/ZYrMpjl8WBZ9LE9VghrAunk1GRnAq6/K/vz5sqCgo6jMDKj0dMPPyZGDmnr1pPquUoacmMqwVU8NYFxZOD9fhl4B5tNYlDLTqlWrlJeXl/rwww/V/v371WOPPaZq1aqlUlNTlVJKDR06VE2dOrXw+FmzZil/f3/16aefqqNHj6off/xRNW/eXD300ENG59Xr9apJkyZqypQpJa55+PBhNXfuXPXHH3+oY8eOqa+//lrdcMMN6rbbbjO53RkZGQqAysjIMPdbJgu5916lAKXeestw35kzSn30kVJDhyoVEiKPF93i4uzX3urs6lWldDrD6/DHH5U/18aNco7AQKWys40fe+89eSwoSKlLl0p/fn6+UjVrynF//125Nvz2mzy/Xj2lCgoqd46yzJkj577pJmmrI9m3T9rm52f6971mjTyndWvrts0SIiKkrWvXVu75+flK+fjIOZKTLdu20rz+ulzrgQeUSkyU/dq1ldLrrX9tZ2bO+7fZQY1SSr3zzjuqSZMmytPTU0VERKht27YVPnb77ber4cOHF97Oy8tTs2fPVs2bN1fe3t6qcePG6sknn1SXL182OucPP/ygAKjkUn6zTp48qW677TYVFBSkvLy8VIsWLdSzzz5rVoDCoMa+9HqlatWSP+KdO0s/pqBA3rQWLDAEQIBSv/5q27aSUn/+aRxc9u5d+XMNHCjniI0t+VhenlI331x+APvXX/K4v3/lg4br15Xy8pLz/PNP5c5RmkuXJFgDlFq92nLntZTr15Vyc5P2nTlj2nNGjZLjJ02ybtss4dFHpa3x8ZV7/oED8nxfX9sEpAkJcr2wMKVeeskQ4FD5rB7UOCMGNfalfWL09VUqN9e054wZI8+5+WalcnKs2z5nd/asUteuWe58K1bIz75ZM6Xc3WW/yGcXk6WmKlWjhjx/9+7Sj/n+e3ncw0Opw4dLPq715tx5p/nXL6p7dznP8uVVO09RM2bIOdu2ddxP2y1aSBt//rniYwsKlGrQQI7/8Ufrt62q5s6Vto4cWbnnf/qpPP/WWy3brrJcumT4oNCpk3xdtMg213Zm5rx/c5VusgltjD4iQqprmuKVV6TWxr59wL/Fp6kU33wDNGkC9O9vuXNqycF33w0MGyb7s2aZf56PPpLcgcjIshMx77lHrpOXB0ydWvLxqubTaLSp3ZbKq7l40TB1e84c4/V8HIk5ycJ//SUzcxx5KndRVZ0BpSUJ2yKfBpCcwbAw2deShZkkbFkO+mdIrkZLEtZmoZgiKAh44w3Znzu35JopJGXXH3lEAodNm+SrJWhBTZs2wAsvADVqSA0WcwKCshKES/P66xIUfP654XdFY6mgRksWttQMqDfekMJtHTsC/fpZ5pzWYE5Qo816uvNO+1ZDNlVVgxotSdjaM5+K0pKFAZktp70+ZBkMasgmKhPUAMDQocAddwDXrgGxsZwNVdSpU1LCXiusnZNjfuXYshQNam64ARg5Um6b01uzebPMbPL3BwYNKv/Ydu2kKB8gU7211zk9HThwQPb/reJQadrvXnIycP581c51/jzw9tuy78i9NEDlghpHnvVUVMuW8vX8eeDyZfOfb8uZT5qiQU3PnqavY0amceA/RXIV588bPkmVMfO/TDqdTPf19JR/uF98Yfn2OaPMTCnFn5IiC+W1ayf3//131c+dkyPBCCBBDSC1Vzw8pFiYNg21IloF4UcflYqpFZk7V47btg347DO5T1tGoXnzqld7DQoyfD/Fe4PM9dprQHY2EB4ugaUjM3Vad3q64efiLEGNn5/0dgDmB/RpaVIUUqcz/P3YgrZcAsCp3NbAoIasTlsW7Kab5I3FXK1aGXItnn5a3tCrs/x86fn4+2+p87JunaEXY8+eqp//0CFAr5fKvVoZqKZNgdGjZX/mzIp7zM6fB7QVTCoaetI0aCCF+AB5va9ft9zQk8YSi1umpQHvviv7c+c6/idtLag5frz8InUbN8rrftNNhrwPZ1DZISgtn6ZlS9OCbkspGtQwn8byGNSQ1VV26KmoadOk9PzZs5LjUV0pBUyYAGzYAPj4AN9+KwFH+/byuCV6aooOPRV9w54+XXrMfvlF8nfK89FHkvjbpYt5XfvPPCOfvI8fB955xxAQWyqosUQRvldekeHQyEjn6NEIDgYCAuR3p4wC7ACcb+hJU9WgxpZDT4AE72+/LVuTJra9dnXAoIasTnsD0T4lV4a3t6Hq7LvvShXi6ujNN4GlSyXYWLHCsAie1n1uiZ6aokFNUY0bG3pdyuutUcow9DRunHnXrllT1oUC5Ku1gppt22RBTXOlpBh+D52hlwYwLGwJlJ1XU1DgvEGN1hNlblBjjyRhzYQJspHlMaghq8rNBXbulP2q9NQAQHQ0MHiwvGmOGydd5dXJV1/JqtaAzBYqOoVbC2qOH5cZOVVRVlADSI+Zl5cEqhs3lv78LVtkCMuUBOHSDBsmn54zMmTz9jb0RFXVDTcATz4pv0PDh8uaVOaYP1+GcLp1A+66yzJtsoWKkoX/+kvyS2rWBP7v/2zXLkuobE+NPZKEyfoY1JBV/fmnJJ7WqWP451MVCxbIei+7dgGLFlX9fM7ijz8MAd0TTwCTJhk/XqeOIWFy796qXau8oCY0VK4PlN1bo/XSDBlSuXWQ3NwMU/kBSca15ArG77wDPP64tH3UKGDZMtOed/o08N57sj9njnP00mgqCmqcbSp3UUWDGlNnR167ZvhZMKhxLQxqyKq0oaeuXS3zJhAcLJ+WAcmtOXOm6ud0dCdOyAyba9ekUN3bb5f+s7TEEFRenuETb2lBDSALUfr4ANu3G94MNRcuGGaomZogXJqePYH77pN9S/ccuLkBixcD48fLm+CYMYZArDzx8RKg33ab8yV4VjQDylmHngCgWTPA3V1mo509a9pz9u2Tnt66dQ3J8OQaGNSQVVkiSbi4sWMlx+LKFZkN5coyMuTNPTVVhmBWr5ZCeKWxRLLwkSMS2Pj5SQ5NaUJCpGYQULK35qOPZMixc2fjehyV8fHHEsBNmVK185RGp5Nza78/48ZJoFOWkycNhQSdJZemqKI9NcV7My5fdr6p3EV5eMiwImD6EFTRoSdney2pfAxqyGqUskyScHFubjIM4O4uvQLr11vu3I4kLw8YOFCGkxo0kKnbAQFlH2+JWjXa0NNNN5X/z/655yT/IilJZmABxgnCVeml0dSuLcmUtWpV/Vyl0ekk8fqZZ+R2bKyhoF5xL78sr0fPnsDtt1unPdbUooV8vxkZwLlzxo9t3CiJwm3ayEw6Z2RuXo29Zj6R9TGoIas5cUJ6GGrUkE/ultS+vSGvJDZWup5dzaRJ8obj6yuBQ1k9J5qiw0+VrbxcXj5NUfXqGWZvzJwpb4q//CJvKn5+wMMPV+76tqbTSSE9rTfo6acl0Cnq2DFg+XLZnzPHtu2zFB8fQ+2Z4nk1zjz0pDE3qLHnzCeyLgY1ZDVaL80tt8g/VUubPVvqPJw4Abz4ouXPb09nzxqGQz79VJJlK3LTTdJ7lZ5e+VwjU4MaQGZi+fnJp96vvjL00gweLDOfnIVOJ/ky06fL7bg4mV2meeklKXh4992GKeHOqLRkYWeeyl2UOUFNQQF7alxZGaPzRFWnjdNbcuipqJo1ZSZL374yW2bIENuWO7emL76Q3paoKOD++017jpeXvHHt3y+9NY0amX9dc4KaOnWAiRPlTX/6dMOCo5YYerI1nU6+jxo1JGfm2WclkHnwQckTApy3l0bTqpUUbSwa1OzeLRWS/fycO2ArPgMqM1O+r9K2M2ckH0/7eyHXwqCGrMYaScLF3X+/rJD81VdSxv/338tOpHUmq1fLV3PrvLRrJ4HJ338DvXqZ91y93jA7xpSgBpBejbffNrxR3nKLab1Kjkink8DF3V0W7pw2TZKD9Xr5WVqqAKC9lNZT48xTuYsqWoDPx0dmqVWka1dJMibX4gL//skRXblimFpszaAGkHo1mzZJkb8335RP2c7s1CkZutPppKfAHNoMqcokCx89Km8GPj6mJ4zWri2BzezZctsZe2mKmzlTAuPnn5efCeD8vTRA6dO6XWHoCZD6Sc2aSW+hFtD4+0sJiKJb/fqGfWeblk+mYVBDVrF9u4xdh4UZisJZS2ioBDOjRgEzZkjvjTN3K3/+uXzt3h1o2NC851alVo029NS6tfRWmGriROnRKCgAHnnE/Os6ounTJbCZMkVmoFk60d0etL+JY8dk2n1WlmEZCmcPanQ6KVB5+LAhcLFGHh85PiYKk1XYYuipqBEjJJEzJ0eGoQoKbHNda6js0BNgqFVz8KBMQTaHOfk0RQUGSs/Qvn3lTzl3Ns89J0nolVkjyhGFhkrujF4v9Yi0qdw33+waCysGBQEREfJBigFN9cWghqyiaCVhW9DpZPaNn59c21mXUDh+XHq53NyAAQPMf36TJhJY5OWVXRK/LJUNagAZhqpd2/znObomTZw716Qonc44r8ZVhp6IimJQQxan18sqyID1Zj6VpmlT4NVXZX/qVEM+hDNZs0a+3n67VO41l04HtG0r++YOQVUlqCHnUDSvhkENuSIGNWRx+/fLlEo/P8MbrK2MGwf06AFcvSpr+lS2CJ29VGXoSVOZ5RIKCoADB2SfQY3r0oKaTz+VysLOPpWbqDgGNWRx2tBTZKTtp1e7uQEffCBj6ps2GdbrcQaHD8uyA+7uwAMPVP48lUkWPnFCFsz09DSso0OuRwtqtN+N6GjLroBOZG8MasjirF10ryLNmwPz5sn+5MkyRdoZaENPPXvKMgSVVZk1oLShp1atXKPOD5Wu+KxADj2Rq2FQQxZn65lPpZkwQarxXrkiQ1KVGYZKTZVpr7ZiiaEnwBDUnDolSyaYgvk01UPLlsa3GdSQq2FQQxaVlibTRXU6+1ZgdXeXRQi9vCQh0pxpuSdOAMOHyxTYLl1sE9gkJ8t6NDVqAP37V+1ctWoZFr80tbeGQU31ULOm4XejbduKF0klcjYMasiitF6atm2lfok9tW5tqHT79NNASkr5x1+4INVxb7wR+Phj6d05eNCwGrg1ffaZfL3rLqm3UVXmDkFpQc3NN1f92uTYWreWr+ylIVfEoIYsyhGGnoqaPFnWIkpPB558svRhqOxs4OWXJRfnzTel2uodd0itG51OEo+/+ca67dSGnh56yDLnM2cGlFLsqalOnnsO6N0beOope7eEyPIY1JBF2broXkVq1AD++19ZuO6rrwzJuIAUqHvvPaBFC+CFF2QaeseOspJxQoIEQXFxcuyYMTK0Zg379snm6SmLc1qCOTOgTp+WIbYaNeRnQa4tOhpYt65yq7gTOToGNWQx16/LlGTAfjOfStOunSxOCADjxwPnz0twc/PNwOOPS0Jws2bAihXS/pgY6aEBpAenXTt5ztix1ql7ow09xcRIPowlFO2pqajNWi/NjTdy1WIicm4Mashidu2SoZv69R2v1sm0aYbgpGVLGeY5dEimTr/9tuTODB4sdW6K8vKSYMfTE/j2WxmKsiSlLD/0BMjUXQ8Pmf114kT5x3LoiYhcBYMaspii+TRaT4ej8PSUYSh3dyAjQ2aBzJolM7UmTCi/AFm7doa6N5MmSZE8S/n7b5n55OUlq4tbioeHISG0orwaBjVE5CoY1JDFaPk0jjT0VFR4uPSKzJ4twczs2YC/v2nPnTRJll/IzgaGDgXy8y3TJq2Xplcvy69wbWqyMIMaInIVDGrIIpRyvJlPpRkwQHpogoPNe56bG/DRRzJNfds2ID6+6m2x1tCTxpRkYc58IiJXwqCGLOLoUVkgz9MTuOUWe7fGOpo0kWneADBnDrBzZ9XO9+ef0mPk4wPcd1/V21ecKbVqUlNlurubmyQKExE5MwY1ZBHa0FPnzoC3t33bYk2DB0uvil4PPPqorAZeWVovTe/eslqypWnDT8nJQE5O6cdovTQtWkheDxGRM2NQQxbhDENPlqDTAUuWAA0bAv/8Azz7bOXOo5RhKndV13oqS8OGMkVcrwcOHCj9GA49EZErqVRQs2jRIoSFhcHb2xuRkZHYsWNHuccvXLgQrVq1go+PDxo3boxJkybh+vXrhY/Pnj0bOp3OaGutTd341/Xr1xEbG4s6derAz88PAwYMQJq1qqER9uyRnoilS4GLFys+vroENYAsY/Dhh7K/eLGsLWWunTuB48dlFlavXpZsnYFOV3GyMIMaInIlZgc1q1evRlxcHGbNmoVdu3ahQ4cOiImJwblz50o9fuXKlZg6dSpmzZqFAwcOYNmyZVi9ejWmT59udNzNN9+MlJSUwm3r1q1Gj0+aNAnffvst1qxZgy1btuDs2bN44IEHzG0+mejxx6U+yxNPAA0aSKXbNWuAa9dKHpueDuzdK/vVIagBpCrr00/L/qhRsm6UObShpz59AF9fy7atqIqShRnUEJFLUWaKiIhQsbGxhbf1er0KDQ1V8fHxpR4fGxurevbsaXRfXFyc6tatW+HtWbNmqQ4dOpR5zfT0dOXh4aHWrFlTeN+BAwcUAJWYmGhSuzMyMhQAlZGRYdLx1dnvvysFKOXpqVSHDrKvbQEBSo0apdTPPyul18vxGzbIY82b27XZNnf1qlJt2sj3/sADShUUmPY8vV6pxo3leV9+adUmqqVL5ToxMaU/XreuPL5rl3XbQURUWea8f5vVU5Obm4ukpCRER0cX3ufm5obo6GgkJiaW+pyuXbsiKSmpcIjq6NGj+O6779CrWJ/7oUOHEBoaihtuuAFDhgzByZMnCx9LSkpCXl6e0XVbt26NJk2alHndnJwcZGZmGm1kmgUL5OuQIcDu3TJ0MXWqzP7JzASWLwd69gSaNpXF8VatkuOrSy+NxscH+OQTKXS3dq2sD7VtW8XLEmzbBpw6JTVy7rnHum0sb/jp/HnpYdLppAIxEZGzMyuouXDhAvR6PYKLFfkIDg5Gampqqc8ZPHgw5s6di+7du8PDwwPNmzdHjx49jIafIiMj8eGHH2LDhg1YsmQJjh07hv/7v//DlStXAACpqanw9PRErWIL45R33fj4eAQGBhZujRs3NudbrbaOHZM3aMCwmGPbtlKX5dgxYMsWWQOpVi1ZCPG11wz5JY5adM+aOnWS9aEACfaiomQZhtmzZRmG0mhDT337Wn+mWNu28vXs2ZK5UdrQU7Nm1h0CIyKyFavPftq8eTPmzZuHxYsXY9euXVi7di3Wr1+PF198sfCYe++9FwMHDkT79u0RExOD7777Dunp6fhMmx5SCdOmTUNGRkbhdurUKUt8Oy7v7beBggLg7rsNb4gaNzfgttuA99+X+iZr1wIPPCC1aXx8ZEHG6ujZZ4Eff5SeLV9fqT0zZ47UfYmMBN55R2r4APKz1VYKt9asp6L8/YGwMNkv3lvDfBoicjU1zDm4bt26cHd3LzHrKC0tDSEhIaU+Z8aMGRg6dCjGjBkDAGjXrh2ys7Px2GOP4fnnn4db8RUEAdSqVQs33ngjDv+7yE5ISAhyc3ORnp5u1FtT3nW9vLzgxcIbZklPNyzYqPXSlMXLC+jfX7aMDFnIsl49qzfRYd11l2xZWcDXX8uw1I8/Ajt2yDZpkgSKnTsDKSlSmfiuu2zTtvbtZabV33/LUg8aBjVE5GrM6qnx9PREeHg4EhISCu8rKChAQkICoqKiSn3O1atXSwQu7u7uAABVRvJBVlYWjhw5ggYNGgAAwsPD4eHhYXTd5ORknDx5sszrkvk++EDelG++Wd6ATRUYWL0DmqL8/KTH5vvvZcjnrbeALl2kVsz33wNaB2X//rYrdlfWDCgGNUTkaszqqQGAuLg4DB8+HJ07d0ZERAQWLlyI7OxsjBw5EgAwbNgwNGzYEPH/Lo7Tp08fLFiwAJ06dUJkZCQOHz6MGTNmoE+fPoXBzeTJk9GnTx80bdoUZ8+exaxZs+Du7o5HHnkEABAYGIjRo0cjLi4OQUFBCAgIwIQJExAVFYVbb73VUj+Lai0vT4aeAOmlcbRVtp1RcDDw1FOyJSfLFPkVKyRJ+IknbNeOspZLYFBDRBal1wP/vq/bTWWmV73zzjuqSZMmytPTU0VERKht27YVPnb77ber4cOHF97Oy8tTs2fPVs2bN1fe3t6qcePG6sknn1SXL18uPGbQoEGqQYMGytPTUzVs2FANGjRIHT582Oia165dU08++aSqXbu28vX1Vf3791cpKSkmt5lTusu3cqVM7a1fX6lr1+zdGtdVUKBUXp5tr7l/v7y2NWsapuFfvGiYpp+Zadv2EJELOnZMqc6dlfrsM4uf2pz3b51SFU1AdQ2ZmZkIDAxERkYGAgIC7N0ch6IUEBEB/PEHMHcuMGOGvVtElpSfL8NiOTnA4cNA8+ayVlf37jJN/8QJe7eQiJzat98Cw4ZJYmZYmKwh4+FhsdOb8/7NtZ8IW7dKQOPtLZWEybXUqGEYYtKGoDj0RERVlp8PTJkC3H+/BDSRkVL3w4IBjbkY1BDeeEO+DhvGhF9XVTxZmEENEVXJ2bNShfXVV+X2008Dv/wi3b92ZHaiMLmWQ4eAb76R/UmT7NsWsp7iycL79slXBjVETujqVZlxcPKkbCdOyNfz52WGQpMmsjVtKl8bN7Zspc+ffwYeeUQKcPn7A8uWAQMHWu78VcCgppp76y3JqendGyi2MDq5kOLLJbCnhsjBKSUrBW/ZIslwWgCjBS/mKhrsNGki1UHvvBNo0cL06a4FBcC8ecCsWbLfvr1UE73xRvPbYyUMaqqxS5eA//5X9isqtkfOTeupOXRIqkGfOSO3b7rJfm0iomKOHwcSEgybVoq8NH5+0hOj9cY0aSL5A2lpxj04J05Iz05ammw7dxqfp1kzKQd/990ynBQYWPr1LlwAhg4FNmyQ2yNHAu++63BrrDCoqcbee09+1zt2BO64w96tIWsKCQHq1JH1n774Qu4LDZU1vIjITs6fBzZtkgDmp5+Ao0eNH/f1Bf7v/4AOHUoOKQUGmtbDohRw+bJhiErbkpJklsixY8DSpbK5uwO33ipBTkwMEB4u923bJsNLp0/LMNbixRLUOCAGNdVUbq6sSQSw2F51oNNJT/GmTYZV1Tn0RA5v715g+3bg0UdtV4Lb2goKgAULZC2Vv/4yfszdXWYQRUfL0NCtt8rielWh0wFBQbJ16mT8WFYWsHmzrOnyww8yFfu332SbOVOeExUlj+Xny2q9a9ZIkOWgGNRUU6tWyRpEDRrYZmFFsr927SSo2bpVbjOoIYeVkwO8/DIQHy9vphs2yD8te1erraq8PGDUKAloNO3aSQATHS0rBvv72649fn7AfffJBsjwlxbgJCRIjsL69fLYgw9KQrCD13ljUFMNKSUfFABgwoSqfxAg56Dl1WgY1Dg4pYDVq6W7//77gVIW/3VJO3bIG782RU+nAz7/XKZnvvWW83YrX70KPPSQBAnu7lJL4+GHJYHXUYSFAY89Jlt+vvSSbd4sycQPPeQUP3sGNdXQpk3S6+nrC4wbZ+/WkK1oM6A0DGoc3OLFwPjxsh8eDrzyinyid1XXrsmsmjfekCGaevXkZ5CfL9OH33kHaNhQir05m/R0oE8f6Sb19pYgrXdve7eqfDVqAN26yeZEqknoT0VpxfZGjpQhU6oebr7Z+IMWgxoH9ssvwMSJsu/pKUmd0dGSvPnnn3ZtWqkOHJBfqE6dgJdeAg4eNO/5W7dKnsZrr0lAM2SI1B148EHpzdC6lqdOBf73P8u335pSU4Hbb5fvMTAQ2LjR8QMaZ2bxlacclCsvaLlihVLTpsmilHv3KpWbW/ax2uKGOp1Shw7Zro3kGFq0MCxcSg7q5El5gQClHnlEqbQ0pZ56SikPD8MqpIMHK3XkiL1bKvbuNbS36HbzzUrNni2PFxSU/tysLKUmTJB/SIBSDRoo9fXXpR87ebIcU6OGUhs2VL69+flyjU8/Veq335Q6fVrus4YjR5S64QZpd0iIUn/9ZZ3ruDhz3r8Z1NjQiRNKPfOMUsePW+6cFy4o5eZm/L/E01Opjh2VGjZMqddfV+qHH5RKSZH/K489Jsf062e5NpDz6N9fXv877rB3S6hU164p1aWLvEgdOiiVnW147MgRCWa0P3QPDwl2zp2zW3PVnj1K1a0r7enYUan331fq3nsl8Cj6T6l1a6VeeEGp3bsNAU5CglLNmhmOGTVKqcuXy76WXq/UkCGGJed37jS/vXv3KnXrrSUDsBo1lAoLU+r225UaOlSp559X6r33JHg6eLDsoKw8f/0lgQwggY2jBKFOiEFNKRwhqImLk9/vuDjLnfOPP+Scvr5Kde2qlJ9fyb9XbatXz/Bh75dfLNcGch7x8fL6P/OMvVtCJRQUKDVihLxAQUFKHT1a+nG7dikVE2P4w/bzU2rOHKWuXLFte3fvVqpOHWnDLbcodfGi4bFLl5T68EOl7rtPPmUV/UfUooVSffsabjdpIp+8TJGTo1R0tKG78fBh0553/bpSs2YZ/gH6+yvVvbtSTZsq5e5e9j9NbWvYUKnYWKV++qn8rnDNr78qFRhoCE5TUkxrJ5WKQU0pHCGo6d1bfsf797fcOdeulXNGRMhtvV7+F371lVJz5yr14INK3XijoXcXUCoysnIfPMj5ZWUp9ckn5X8gJjt55x35A3VzU2rjxoqP/+knpcLDDX/YwcFKPf20UmvWKJWaat22JiVJ4AVIz9KlS2Ufm54uv3T9+inl5WUcLDzxhFKZmeZdOyNDeoUApZo3l+G58iQmKtWmjeGaffoodeqU4fH8fBny27pVxvDnz5d29e6tVLt2Svn4GLe5dm3pBv/yS+OeNM26dUp5e8ux3bvzj80CGNSUwhGCmlat5Pc8PNxy53zrLTnngw+Wf1x2tvTWrlyp1Jkzlrs+EVnAli2GIZvXXzf9eXq9UqtXy5t78d6Fli2VGjlSqeXLlfrnH8t9ktm5U6lateQat94qQYupMjOVWrVK8mM2b658G1JSZLgIUKpz59J7qa5ckeE57RNd/fryszL353DtmlLffqvU6NGGoTZt8/GRYO2jj6Sn6uOPDT0/991XetBDZmNQUwp7BzX5+YZe2Hr1LHfeZ56x/JAWEdnQyZPyT0FLAK5M8JGbq9Rnnyn15JNKtW9v3DVbtCdnwAClFi6UIazKXGf7dsOwSteu0mtiL8nJhuGve+4xHhbasEGGlrTvfdgwSUCsqvx8CUAnTjQ+P2A8jDV0qGnDVGQSBjWlsHdQc/y48e//1auWOe/AgXK+hQstcz4isqGrV6WnQUu0tdQn+0uXlFq/XqmpU2UIpHhei5bbMnOmBAem+P13pQICDMMq5g4bWcO2bYbhoeHDlTp/XgIK7XsMCzM9X8dcBQVK/fmn/AzbtTNcc+JE6UEjizHn/VunlFL2mEpua5mZmQgMDERGRgYC7FDmOSFBykxoDh4EWrWq+nlvvVWKPq5dC/TvX/XzEZGNKCXFoj76SFYb/eMPqehqDdevy+rMW7cCv/4KbNkiFW414eFSG+bhh2XtlOJ++w245x5ZK+j224F166TEviNYtw7o1w/Q6wEfHynip9MBTz8NvPii7dp55Iisgh0V5RSVd52JOe/fLL5nI4cPG98+edIy5z11Sr42bmyZ8xGRjbz7rgQ0bm6yHIK1AhpAqtj+3/8B06YB330nb74rVgC9eknJ/qQkWdm2USPgrruA//4XyMiQ5/7yixT9y8oCevaUMv+OEtAAsm7R++/L/rVrUmUyMRF4803btrN5c6BrVwY0dsZlEmzkyBHj25YIanJzZVFKQFaiJyInsWWLrGUESBVdWy9/4OcHDB4s2/nzwGefSZCTmAj89JNsTzwB3HuvLHB49ap0NX/9tayv4mhGjZKFIM+fB8aM4YJ21RiDGhvRempq1JClTCwR1Jw5Iz3YXl6yTAoROYGjR4GBA2W4ZMgQQ3BjL/XqAbGxsh09CqxcKQHOwYPAV1/JMTExwJdfyvCOoxo40N4tIAfAoMZGtKAmMlKGpy0R1GjnaNKEPZ7kgg4dAj7+WD4FeHnJp28vL+Ot6H0NGgAdO1rmjyEjA1i4EFi6VD45hITIasohIcb7Re8LCADOnZNPG+VtWVlyjY4dZdjEkf54b7gBeOEF4Pnngd27gU8/lbWYXnpJhrCIHByDGhtQyjD81LOn5YIa5tOQS9LrJR9ixgxJcDVHeLjkhgwcCHh4mH/trCzJdXn1VeDyZcP9aWnmn6s8bdtKz4cjDuUAEmh16iQbkRNhUGMDqakyJO3uDtx2m9xn6Z4aIpPo9cBff8mKyO7u9m5NSfv2SX7Ejh1yu0cPaWtuLpCTI1tZ+/v2ScLrkCHAlCnAU08BY8cCtWpVfN1r14AlS4D58yUvAwBatwZmz5Zpimlp8oecmmrYL/r14kV5jru79Bg1bGi8hYYa33akRFsiF8Kgxga0oaemTYEWLWT/1Cnp1XWrwvwzBjVklnPnZMrupk0yY+TLLyXJyxHk5UnvyNy5EqgEBgILFsiUZ1OHZy5ckOGid98FTp8GnntOzjd6tEzvbdas5HNycoAPPgBeftmQdd+8OTBrliTRmhr45eYCmZlA7dqOGSwSVROc0m0DWlDTvLl8SNPp5H+p9oGwshjUkMl27JChmU2b5Pa6ddKT4QhlqnbvBiIiJJcjN1cCLq3Hxpx8k7p15RwnTgDLl8sQT1YW8NZb8mli4ECZ3QNIEPWf/wAtWwLjx0tA06SJBDgHDgBDh5oXnHh6yvUZ0BDZFYMaG9DyaVq0kGH+0FC5XdUhKO35zKmhcn3wgdQoOX1ahlLeeEOChSVLgNdft1+7cnKAmTOBLl0ksAkKAj75BPjmG4n+K8vLS3p49uwBfvgBuPtu6Rb9/HOpIxIVJUNLjz0mXaahocCiRcA//0ivTmVycYjIITCosQGtp0YbetJ6Vqoa1GiJwuypoVLl5Mgb99ix0gPSr5/02MTFydAOIEM0q1fbvm1az9GLL8rspgEDgP37JR/GUrOBdDoJaH74Afj7b+n58fQEtm2Tqcv160tC8uHDwJNPSjBERE6NQY0NFB1+AiwT1GRkyBA+wJ4aKsWpU5KV/p//yJv7vHnAF1/ItGMAmDhR8kwAYNgwKZ1vCykpwOTJ0luyb5/USFmzRnpRgoOtd922bYFly2Ro6tVXgXfekcBm4kTHrr1CRGZxkCxB16VU2T01J05U/rxaQFSnDlCzZuXPQy5o0yZg0CBJ2goKklojd99d8rg33pBfpC+/BPr2BX7/XYZlLE2vB77/XgKs9evlNiC9MgsXSi6KrYSEAM8+a7vrEZFNsafGyi5dMiyhcsMN8tUSPTVMEqYSlJJhpbvukoCmY0dZJLG0gAaQpNZPPpGKkJcvyzpAlqzHcuKE5Mw0bQr06SO5Mno90K0b8O23cm1bBjRE5PLYU2NlWi9No0aGXm5LBDUsvEdGsrJkzRstP2boUOC99yoeWvH1lQAjKkoy2vv0kZ6eynb/5eVJ8PKf/8iaQdrsqjp1ZJhrzBigTZvKnZuIqAIMaqyseD4NwJ4aspD8fGD7dgkePv1UlhWoUUOGdJ580vSE23r1ZOXmrl2BnTuBRx6RISlzpicfOAB8+KFs584Z7u/ZUxKV+/dnIi4RWR2DGisrnk8DGAKR8+elkGll8hQZ1FRTx4/LbJ4ffwQSEgxjm4Dki3z+uQzvmOvGG6WHpWdP6bl56ikpYldWYJSfL+t9fPONHH/okHE7Ro6U6dFFo3kiIitjUGNlRWvUaGrXlt797GwZRrrxRvPPy6CmmsjKAjZvlkDmhx+MgwdAEoGjo2UV5X795HZlde0qqzMPHAgsXiwVeCdPNjyekQFs2CCBzPffG6+N5OEhuTtjxgC9e7PWCxHZRaUShRctWoSwsDB4e3sjMjISO7R1WsqwcOFCtGrVCj4+PmjcuDEmTZqE60UWqouPj0eXLl3g7++P+vXro1+/fkhOTjY6R48ePaDT6Yy2xx9/vDLNt6nSemp0OsmdBCo/BMXCey7oyhWZgbRkCfD445LnEhQkeS7vvisBjbs70L27lP/fvl2GelavlhosVQloNAMGyKwoQGYJLV0qFXmjoyWp9+GHgZUrJaDR8mQ+/1zWPlq3TgIrBjREZCdm99SsXr0acXFxWLp0KSIjI7Fw4ULExMQgOTkZ9evXL3H8ypUrMXXqVCxfvhxdu3bFP//8gxEjRkCn02HBvwXAtmzZgtjYWHTp0gX5+fmYPn067r77buzfvx81iyQsjh07FnPnzi287euoK9wWUVpODSA9LPv3Vy6o0euBM2cM5yEno5S88H/9Zdh27zZ06xXXrJn0xNx9twwPBQZat30TJwLHjkktlyeeMH7sppskyLr/fuDWW7ksABE5FLODmgULFmDs2LEYOXIkAGDp0qVYv349li9fjqlTp5Y4/vfff0e3bt0wePBgAEBYWBgeeeQRbN++vfCYDRs2GD3nww8/RP369ZGUlITbtGWtIUFMSEiIuU22m8xMw/pOpQU1QOWCmtRUSWnQFgQmJ3HpkvSwfPyx8dBNUQ0ayKrU2tali/zyWKrKril0Oqm0e+6c9MLcdpsEMn36GHc5EhE5GLOCmtzcXCQlJWHatGmF97m5uSE6OhqJ2kJxxXTt2hWffPIJduzYgYiICBw9ehTfffcdhg4dWuZ1Mv5Nfgwq1p2+YsUKfPLJJwgJCUGfPn0wY8YMh+6t0T54169vKOSqqUpQoz2nUSN+UHYK+fkyxXnGDBmmAWSWUps2xgFMhw4yE8kRuLvLjKpPPnGclbyJiCpg1n+rCxcuQK/XI7hYOfPg4GAcPHiw1OcMHjwYFy5cQPfu3aGUQn5+Ph5//HFMnz691OMLCgowceJEdOvWDW3btjU6T9OmTREaGoo9e/ZgypQpSE5Oxtq1a0s9T05ODnJycgpvZ2prCthQWUNPgGWCGg49OYGff5bhnL//lts33yyLSPbsKesQOTKdjgENETkVq//H2rx5M+bNm4fFixcjMjIShw8fxtNPP40XX3wRM2bMKHF8bGws9u7di61btxrd/9hjjxXut2vXDg0aNMCdd96JI0eOoHkpUUN8fDzmzJlj+W/IDKUlCWuqEtSw8J4TOHZMZg5pQXft2rJ447hxDBSIiKzErNlPdevWhbu7O9KKlVJPS0srM9dlxowZGDp0KMaMGYN27dqhf//+mDdvHuLj41FQUGB07Pjx47Fu3Tps2rQJjRo1KrctkZGRAIDDWuRQzLRp05CRkVG4ndIiARsqbTq3pmhQU+zHUCH21DiwrCzg+ecloXbtWhnGGT9eZi7FxjKgISKyIrOCGk9PT4SHhyMhIaHwvoKCAiQkJCAqKqrU51y9ehVubsaXcf83EUT9W0JdKYXx48fjyy+/xM8//4xmzZpV2Jbdu3cDABqUkSnr5eWFgIAAo83WyuupadhQevdzcgzJxKZiUOOACgqA//1Pig7NmycvbM+eMqvpnXdk+jMREVmV2R8b4+LiMHz4cHTu3BkRERFYuHAhsrOzC2dDDRs2DA0bNkR8fDwAoE+fPliwYAE6depUOPw0Y8YM9OnTpzC4iY2NxcqVK/H111/D398fqampAIDAwED4+PjgyJEjWLlyJXr16oU6depgz549mDRpEm677Ta0b9/eUj8Liysvp8bTUya6nD0rQUqxNKVyMaixgZwcmaGUmSnblSuyafvFv+7ZA+zaJc+94Qap9dK3r21nLRERVXNmBzWDBg3C+fPnMXPmTKSmpqJjx47YsGFDYfLwyZMnjXpmXnjhBeh0Orzwwgs4c+YM6tWrhz59+uDll18uPGbJkiUApMBeUf/9738xYsQIeHp64qeffioMoBo3bowBAwbghRdeqMz3bBPXrhlqyZQ1C7ZJE0NQ06WL6edm4b0qOHRI1jm6fLnklp5u2L92zfxz16wJvPCCJAZ7e1u65UREVAGd0saAXFxmZiYCAwORkZFhk6GoffuAtm2BWrWkPElpH9gHDQI++wxYsACYNMm08169alhA+fJlOT+Z6I8/gB49ZH0KU+h0Mhff37/k1+L31aolBelYOIiIyKLMef9m1qKVFB16KmsEojIzoLR8Z39/6xeWdSlHjsiaRNnZQMeOUg23dm3DVquW8e3atSVYcTMr7YyIiOyIQY2VlJckrKnM+k9F82mYrmGic+dkmYFz5ySg2bKlZDVEIiJyevwYaiXlTefWVKanhknCZsrKkh6aI0eAsDBZXZoBDRGRS2JQYyWm9NRUZfiJScImyMsDBg6UXJo6dYAffgCcaO0wIiIyD4MaKylvOrdGC2rOnTN9sg17akykFDBmDLBhA+DrC6xfLzVkiIjIZTGosYLcXODECdkvr6emdm3DTKbTp007N4MaEz3/vKyG7e4uU8z+rUBNRESui0GNFZw4IQVmfX3LH+3Q6QzBiRYEVcTpg5q8POD6dete4913gX+LP+L99yWnhoiIXB5nP1lB0XyaimYoNWkCHDhgWl6NUk6eU3P4MHDbbUBKiuS4NGwoW2ioYb/oVreu+VO8vvgCeOop2Z87Fxg1yvLfBxEROSQGNVZgSj6Nxpxk4QsXpJNDp5P3fKdy6ZL0mKSkyO2LF2Xbs6fs53h5yRTs7t2B//s/+VreGkq//goMGSLR37hxUt2XiIiqDQY1VmDKdG6NOUGNdkxIiLzfO43cXGDAAOCff6SLaeNGue/MmbK38+dl/aXt22V74w051003SYCjBTlNm0qUt2+fVPTNyZE1lxYtYiEfIqJqhkGNFZgynVtTmaDGqfJplAIeewzYvFnKIK9fD7RqJY+1a1f283JyZKwtMRHYulV6YQ4cMGzvvy/HNWokwc3WrbJ2U9euwKefSoIwERFVKwxqrMBaw09OGdTExwMffWSYhVReIFOUl5dEhS1aAEOHyn0XLgC//WYIcpKSZNrYqlXyeOvWwLffAj4+1vleiIjIoTGosTC9Hjh6VPbN7alRqvwRE6dLEl69WqZWA8A77wD33FO189WtK0NLffvK7atXZWhq61b5Ac6cCQQFVe0aRETktBjUWNjp0zJr2dNTRkYq0qiRBDI5OZJGUr9+2cc6VU9NYiIwfLjsT5oEPPGE5a/h6wvccYdsRERU7bFOjYVpQ0833GBaWoenJ9CggexXNATlNEHN0aPSm5KTI8m7r71m7xYREVE1wKDGwszJp9GYmlfjFEFNerpM3T5/HujUCVixgkm7RERkEwxqLMycmU8aU4Ka3FwgNVX2HTanJi8PePBB4OBBKaTz7beAn5+9W0VERNUEgxoLM6dGjcaUoObMGUkk9vIC6tWrfPusRinJm0lIkAWt1q1zwgqBRETkzBjUWFhVhp/KW/+p6NCTQ9aUe+01YNkywM1NZj117GjvFhERUTXD2U8WpJT1emocMp9GKakS/PXXwJQpct+bb3IBSSIisgsGNRaUmiqlU9zdpXq/qZwmqMnPB/78Uwrfbd0q2/nzhsfHjzcsJklERGRjDGosSBt6atpUpmqbSgtUzp0Drl0rvSCuXQrvZWdLcbtff5Vt2za5rygvLyAyErjvPqlHQ0REZCcMaiyoMvk0gBTB9fWVXp7Tp4GWLUseY9OeGqWAV18FZsyQGU1F1a4NdOtmWFAyPNzJVtckIiJXxaDGgioznRuQxN8mTWQm9MmTdg5qlAKefdawKnbjxoYA5v/+D2jTRpKBiYiIHAyDGguqTJKwpmhQUxqbBDV6vayovXy53H7zTWDiRCtekIiIyHIY1FhQZXtqAENicWlBTUYGcOWK7FstpyYnBxgyBPjiC+mJWbYMGDHCShcjIiKyPAY1FqJU5XNqgPJnQGn31akjuTcWl50NPPAA8OOPkuH86adym4iIyIkwqLGQixelRwWQxSzNZUpQY5Whp8uXpa5MYqJUAv7qKyA62goXIiIisi4GNRai5dM0alT6lOyK2CWoSU0FYmKAPXtkVtN33wG33mrhixAREdkGgxoLqcrQE2Ac1ChlvBSCVYKa48eBu+6ShoeEyNBTu3YWvAAREZFtcW6uhVQlSRiQtR91OuD6deMivYAVCu/t3y9TtA8fBpo1k8rADGiIiMjJMaixkKpM5wakfl1IiOwXH4KyaE/NH38At90my363aSOVgivbvURERORAGNRYSFV7aoCy82osEtT88w/wzDNAjx6S1dylC/DLL9JFRERE5AKYU2MhVc2pASRo2b7dOKjR62XpBO1xs+TlyQraS5YAP/9suP/OO4EvvwT8/SvfWCIiIgfDoMYCMjMNeTBVDWoA46AmNVUCmxo1DMNTFTp5Enj/fSmgl5oq9+l0MnX78ceBe+6RpcSJiIhcCIMaC9DyaerXBwICKn+e0oIabb9hwwriEL0e2LABWLpUpmYXFMj9ISHAmDGyaWWLiYiIXBCDGguwxNATUPpSCSbl0/znP8DLLwMnThjuu/NO6ZXp2xfw8Khaw4iIiJxApRKFFy1ahLCwMHh7eyMyMhI7duwo9/iFCxeiVatW8PHxQePGjTFp0iRcv37drHNev34dsbGxqFOnDvz8/DBgwACkpaVVpvkWZ4kkYaD8npoyg5rNm2URyhMnpIBeXJysjPnTT8CDDzKgISKiasPsoGb16tWIi4vDrFmzsGvXLnTo0AExMTE4d+5cqcevXLkSU6dOxaxZs3DgwAEsW7YMq1evxvTp080656RJk/Dtt99izZo12LJlC86ePYsHHGR9oqpO59ZogUtamtSrAQw1akoNavR6YNIk2R8+XKZpv/EG0KpV1RpCRETkjJSZIiIiVGxsbOFtvV6vQkNDVXx8fKnHx8bGqp49exrdFxcXp7p162byOdPT05WHh4das2ZN4TEHDhxQAFRiYqJJ7c7IyFAAVEZGhknHm+P225UClFqxomrnKShQytdXznXokNzXt6/cXry4lCcsWyYPBgYqdf581S5ORETkgMx5/zarpyY3NxdJSUmILrLgoZubG6Kjo5GYmFjqc7p27YqkpKTC4aSjR4/iu+++Q69evUw+Z1JSEvLy8oyOad26NZo0aVLmdXNycpCZmWm0WYulcmp0upJDUGUOP125Ami9XTNnAnXrVu3iRERETs6soObChQvQ6/UIDg42uj84OBip2tThYgYPHoy5c+eie/fu8PDwQPPmzdGjR4/C4SdTzpmamgpPT0/UqlXL5OvGx8cjMDCwcGtssTUGjF27JqM+QNWHnwAzgpr582WcqkULYPz4ql+YiIjIyVm9ovDmzZsxb948LF68GLt27cLatWuxfv16vPjii1a97rRp05CRkVG4ndKSUyzs6FH5WqsWEBRU9fNpwcuJE8DVq1L8t+j9hQ++8Ybsv/464OlZ9QsTERE5ObOmdNetWxfu7u4lZh2lpaUhpIzKcDNmzMDQoUMxZswYAEC7du2QnZ2Nxx57DM8//7xJ5wwJCUFubi7S09ONemvKu66Xlxe8vLzM+fYqpejQU9GVtSuraE+NFof5+wOBgUUOmjIFyMkB7rgDuP/+ql+UiIjIBZjVU+Pp6Ynw8HAkJCQU3ldQUICEhARERUWV+pyrV6/Czc34Mu7/VpFTSpl0zvDwcHh4eBgdk5ycjJMnT5Z5XVuJiABWrTKkt1RV0aCm1KGn338HVq+WCGrBAstEUkRERC7A7OJ7cXFxGD58ODp37oyIiAgsXLgQ2dnZGDlyJABg2LBhaNiwIeLj4wEAffr0wYIFC9CpUydERkbi8OHDmDFjBvr06VMY3FR0zsDAQIwePRpxcXEICgpCQEAAJkyYgKioKNx6662W+llUSoMGwKBBljtfuUFNQYFhCvfo0UDHjpa7MBERkZMzO6gZNGgQzp8/j5kzZyI1NRUdO3bEhg0bChN9T548adQz88ILL0Cn0+GFF17AmTNnUK9ePfTp0wcvv/yyyecEgDfffBNubm4YMGAAcnJyEBMTg8WLF1fle3dIRYMarUBwYVCzciWwYwfg5wdYOSeJiIjI2eiUUsrejbCFzMxMBAYGIiMjAwFVWaDJynJyAG9v2b/vPmDdOlkBYfrEq1JU7/RpID4emDrVvg0lIiKyAXPev60++4nM4+UlQ1oA8Ntv8rVxY8gsp9OngbAwYOJEO7WOiIjIcTGocUDacNPly//e9jkPvPKK3HjlFUNXDhERERViUOOAihfaa7JyvhSt6dYNGDjQPo0iIiJycAxqHFDRoEanU2j45Tty4803OYWbiIioDAxqHFDRoCakxkV4Ig8YOhTo0sV+jSIiInJwDGocUNGgpkneYcDXF5g3z34NIiIicgIMahyQUVCDk8BzzwGNGtmvQURERE6AQY0DMgpq/C4BkyfbrzFEREROgkGNA6pTcB4+uAoAaNIvHKhZ084tIiIicnwMahyQ7qUX0RSyRkKT/uF2bg0REZFzYFDjaA4fBpYswQy8iAdvO4e7YvgSERERmcLsBS3JyqZPB/LzMfjedAz+rr69W0NEROQ02A3gSLZvB9askQJ72rIIREREZBIGNY5CKeDZZ2V/xAigXTu7NoeIiMjZMKhxFN9+C/z6K+DjA8yda+/WEBEROR0GNY4gPx+YMkX2J01ioT0iIqJKYFDjCJYtAw4eBOrWlerBREREZDYGNfaWlQXMmiX7M2cCgYH2bQ8REZGTYlBjb2+8AaSlAc2bA+PG2bs1RERETotBjT2lpgKvvSb78fGAp6d920NEROTEGNTY05w5QHY2EBEBPPigvVtDRETk1BjU2MvBg8B//iP7r70mBfeIiIio0hjU2Mu0aYBeD9x/P3DbbfZuDRERkdNjUGMPW7cCX30FuLkB8+fbuzVEREQugUGNrRVdDmHMGOCmm+zbHiIiIhfBoMbWvvgC2LYNqFlTEoWJiIjIIhjU2FJuruTSAMDkyUBIiH3bQ0RE5EIY1NjS//4HHD4MBAcDzzxj79YQERG5FAY1trR9u3wdMwbw97dvW4iIiFwMgxpbOn1avjZrZt92EBERuSAGNbZ05ox8bdTIvu0gIiJyQQxqbEnrqWFQQ0REZHEMamzl2jXg0iXZb9jQvm0hIiJyQQxqbEUbeqpZEwgMtG9biIiIXBCDGlvRhp4aNuTilURERFbAoMZWmCRMRERkVQxqbIVJwkRERFZVqaBm0aJFCAsLg7e3NyIjI7Fjx44yj+3Rowd0Ol2JrXfv3oXHlPa4TqfDa6+9VnhMWFhYicfnO9MK10WHn4iIiMjiapj7hNWrVyMuLg5Lly5FZGQkFi5ciJiYGCQnJ6N+/foljl+7di1yc3MLb1+8eBEdOnTAwIEDC+9LSUkxes7333+P0aNHY8CAAUb3z507F2PHji287e9MVXk5/ERERGRVZgc1CxYswNixYzFy5EgAwNKlS7F+/XosX74cU6dOLXF8UFCQ0e1Vq1bB19fXKKgJKbaw49dff4077rgDN9xwg9H9/v7+JY51Ghx+IiIisiqzhp9yc3ORlJSE6Ohowwnc3BAdHY3ExESTzrFs2TI8/PDDqFmzZqmPp6WlYf369Rg9enSJx+bPn486deqgU6dOeO2115Cfn1/mdXJycpCZmWm02ZXWU8PhJyIiIqswq6fmwoUL0Ov1CA4ONro/ODgYBw8erPD5O3bswN69e7Fs2bIyj/noo4/g7++PBx54wOj+p556CrfccguCgoLw+++/Y9q0aUhJScGCBQtKPU98fDzmzJljwndlA3l5gDbExp4aIiIiqzB7+Kkqli1bhnbt2iEiIqLMY5YvX44hQ4bA29vb6P64uLjC/fbt28PT0xPjxo1DfHw8vLy8Spxn2rRpRs/JzMxE48aNLfBdVEJqKqAU4OEB1KtnnzYQERG5OLOGn+rWrQt3d3ekpaUZ3Z+WllZhrkt2djZWrVpV6rCS5tdff0VycjLGjBlTYVsiIyORn5+P48ePl/q4l5cXAgICjDa70YaeQkMBN86iJyIisgaz3mE9PT0RHh6OhISEwvsKCgqQkJCAqKiocp+7Zs0a5OTk4NFHHy3zmGXLliE8PBwdOnSosC27d++Gm5tbqTOuHA6ThImIiKzO7OGnuLg4DB8+HJ07d0ZERAQWLlyI7OzswtlQw4YNQ8OGDREfH2/0vGXLlqFfv36oU6dOqefNzMzEmjVr8MYbb5R4LDExEdu3b8cdd9wBf39/JCYmYtKkSXj00UdRu3Ztc78F22ONGiIiIqszO6gZNGgQzp8/j5kzZyI1NRUdO3bEhg0bCpOHT548CbdiQyzJycnYunUrfvzxxzLPu2rVKiil8Mgjj5R4zMvLC6tWrcLs2bORk5ODZs2aYdKkSUY5Mw6NNWqIiIisTqeUUvZuhC1kZmYiMDAQGRkZts+veeQRYNUqYMECYNIk216biIjIiZnz/s2sVVtgjRoiIiKrY1BjC0wUJiIisjoGNdamFHtqiIiIbIBBjbVduADk5gI6HdCggb1bQ0RE5LIY1FibNvQUHAx4etq3LURERC6MQY21sUYNERGRTTCosTbWqCEiIrIJBjXWxp4aIiIim2BQY23sqSEiIrIJBjXWxho1RERENsGgxto4/ERERGQTDGqsjcNPRERENsGgxpoyM4ErV2SfPTVERERWxaDGmrShp1q1gJo17doUIiIiV8egxpo49ERERGQzDGqsiUnCRERENsOgxprYU0NERGQzDGqsiTVqiIiIbIZBjTVx+ImIiMhmGNRYE4efiIiIbIZBjTVx+ImIiMhmGNRYy/XrwIULss/hJyIiIqtjUGMtZ8/KVx8foHZt+7aFiIioGmBQYy1Fk4R1Ovu2hYiIqBpgUGMtTBImIiKyKQY11sIkYSIiIptiUGMtrFFDRERkUwxqrIXDT0RERDbFoMZaOPxERERkUwxqrIXDT0RERDbFoMYa8vOB1FTZZ08NERGRTTCosYa0NECvB9zdgfr17d0aIiKiaoFBjTVoScKhoRLYEBERkdUxqLEGJgkTERHZHIMaa2CSMBERkc0xqLEG1qghIiKyuUoFNYsWLUJYWBi8vb0RGRmJHTt2lHlsjx49oNPpSmy9e/cuPGbEiBElHr/nnnuMznPp0iUMGTIEAQEBqFWrFkaPHo2srKzKNN/6OPxERERkc2YHNatXr0ZcXBxmzZqFXbt2oUOHDoiJicG5c+dKPX7t2rVISUkp3Pbu3Qt3d3cMHDjQ6Lh77rnH6LhPP/3U6PEhQ4Zg37592LhxI9atW4dffvkFjz32mLnNtw0OPxEREdmc2UHNggULMHbsWIwcORJt2rTB0qVL4evri+XLl5d6fFBQEEJCQgq3jRs3wtfXt0RQ4+XlZXRc7dq1Cx87cOAANmzYgA8++ACRkZHo3r073nnnHaxatQpnz54191uwPg4/ERER2ZxZQU1ubi6SkpIQHR1tOIGbG6Kjo5GYmGjSOZYtW4aHH34YNWvWNLp/8+bNqF+/Plq1aoUnnngCFy9eLHwsMTERtWrVQufOnQvvi46OhpubG7Zv327Ot2B9SrGnhoiIyA5qmHPwhQsXoNfrERwcbHR/cHAwDh48WOHzd+zYgb1792LZsmVG999zzz144IEH0KxZMxw5cgTTp0/Hvffei8TERLi7uyM1NRX1ixWxq1GjBoKCgpCqVe4tJicnBzk5OYW3MzMzTf02q+bSJUC7bmioba5JRERE5gU1VbVs2TK0a9cOERERRvc//PDDhfvt2rVD+/bt0bx5c2zevBl33nlnpa4VHx+POXPmVKm9laL10tSvD3h52f76RERE1ZRZw09169aFu7s70tLSjO5PS0tDSEhIuc/Nzs7GqlWrMHr06Aqvc8MNN6Bu3bo4fPgwACAkJKREInJ+fj4uXbpU5nWnTZuGjIyMwu3UqVMVXtciOPRERERkF2YFNZ6enggPD0dCQkLhfQUFBUhISEBUVFS5z12zZg1ycnLw6KOPVnid06dP4+LFi2jQoAEAICoqCunp6UhKSio85ueff0ZBQQEiIyNLPYeXlxcCAgKMNptgkjAREZFdmD37KS4uDv/5z3/w0Ucf4cCBA3jiiSeQnZ2NkSNHAgCGDRuGadOmlXjesmXL0K9fP9SpU8fo/qysLDz77LPYtm0bjh8/joSEBPTt2xctWrRATEwMAOCmm27CPffcg7Fjx2LHjh347bffMH78eDz88MMIdbS8FfbUEBER2YXZOTWDBg3C+fPnMXPmTKSmpqJjx47YsGFDYfLwyZMn4eZmHCslJydj69at+PHHH0ucz93dHXv27MFHH32E9PR0hIaG4u6778aLL74IryI5KStWrMD48eNx5513ws3NDQMGDMDbb79tbvOtj4X3iIiI7EKnlFL2boQtZGZmIjAwEBkZGdYdioqJAX78EfjwQ2D4cOtdh4iIqBow5/2baz9ZGoefiIiI7IJBjaUxUZiIiMguGNRY0pUrQEaG7LOnhoiIyKYY1FiS1ksTEAD4+9u3LURERNUMgxpL4tATERGR3TCosSQmCRMREdkNgxpLYo0aIiIiu2FQY0kcfiIiIrIbBjWWxOEnIiIiu2FQY0nsqSEiIrIbBjWWxJwaIiIiu2FQYyk5OcC5c7LP4SciIiKbY1BjKSkp8tXLC6hTx75tISIiqoYY1FhK0SRhnc6+bSEiIqqGGNRYCvNpiIiI7IpBjaVw5hMREZFdMaixFNaoISIisisGNZbCnhoiIiK7YlBjKcypISIisisGNZbC4SciIiK7YlBjCXq9oU4Ne2qIiIjsgkGNJZw7B+TnA25uQHCwvVtDRERULTGosQRt6KlBA6BGDfu2hYiIqJpiUGMJnPlERERkdwxqLIFJwkRERHbHoMYS2FNDRERkdwxqLIE1aoiIiOyOQY0lcPiJiIjI7hjUWAKHn4iIiOyOQU1VKcWeGiIiIgfAoKaqLl8Grl2TfQY1REREdsOgpqq0oae6dQFvb/u2hYiIqBpjUFNVHHoiIiJyCKzpX1U33ADMng3Urm3vlhAREVVrDGqqqlUrYNYse7eCiIio2uPwExEREbkEBjVERETkEioV1CxatAhhYWHw9vZGZGQkduzYUeaxPXr0gE6nK7H17t0bAJCXl4cpU6agXbt2qFmzJkJDQzFs2DCcPXvW6DxhYWElzjF//vzKNJ+IiIhckNlBzerVqxEXF4dZs2Zh165d6NChA2JiYnDu3LlSj1+7di1SUlIKt71798Ld3R0DBw4EAFy9ehW7du3CjBkzsGvXLqxduxbJycm4//77S5xr7ty5RueaMGGCuc0nIiIiF2V2ovCCBQswduxYjBw5EgCwdOlSrF+/HsuXL8fUqVNLHB8UFGR0e9WqVfD19S0MagIDA7Fx40ajY959911ERETg5MmTaNKkSeH9/v7+CAkJMbfJREREVA2Y1VOTm5uLpKQkREdHG07g5obo6GgkJiaadI5ly5bh4YcfRs2aNcs8JiMjAzqdDrVq1TK6f/78+ahTpw46deqE1157Dfn5+WWeIycnB5mZmUYbERERuS6zemouXLgAvV6P4OBgo/uDg4Nx8ODBCp+/Y8cO7N27F8uWLSvzmOvXr2PKlCl45JFHEBAQUHj/U089hVtuuQVBQUH4/fffMW3aNKSkpGDBggWlnic+Ph5z5swx8TsjIiIiZ2fTOjXLli1Du3btEBERUerjeXl5eOihh6CUwpIlS4wei4uLK9xv3749PD09MW7cOMTHx8PLy6vEuaZNm2b0nMzMTDRu3NhC3wkRERE5GrOGn+rWrQt3d3ekpaUZ3Z+WllZhrkt2djZWrVqF0aNHl/q4FtCcOHECGzduNOqlKU1kZCTy8/Nx/PjxUh/38vJCQECA0UZERESuy6ygxtPTE+Hh4UhISCi8r6CgAAkJCYiKiir3uWvWrEFOTg4effTREo9pAc2hQ4fw008/oU6dOhW2Zffu3XBzc0P9+vXN+RaIiIjIRZk9/BQXF4fhw4ejc+fOiIiIwMKFC5GdnV04G2rYsGFo2LAh4uPjjZ63bNky9OvXr0TAkpeXhwcffBC7du3CunXroNfrkZqaCkBmTnl6eiIxMRHbt2/HHXfcAX9/fyQmJmLSpEl49NFHUZtrLhEREREqEdQMGjQI58+fx8yZM5GamoqOHTtiw4YNhcnDJ0+ehJubcQdQcnIytm7dih9//LHE+c6cOYNvvvkGANCxY0ejxzZt2oQePXrAy8sLq1atwuzZs5GTk4NmzZph0qRJRjkzREREVL3plFLK3o2whczMTAQGBiIjI4P5NURERE7CnPfvarNKtxa7sV4NERGR89Det03pg6k2Qc2VK1cAgNO6iYiInNCVK1cQGBhY7jHVZvipoKAAZ8+ehb+/P3Q6nUXPrdXAOXXqFIe2HBxfK+fB18q58PVyHs72WimlcOXKFYSGhpbI2S2u2vTUuLm5oVGjRla9BuvhOA++Vs6Dr5Vz4evlPJzptaqoh0Zj9irdRERERI6IQQ0RERG5BAY1FuDl5YVZs2aVugYVORa+Vs6Dr5Vz4evlPFz5tao2icJERETk2thTQ0RERC6BQQ0RERG5BAY1RERE5BIY1BAREZFLYFBTRYsWLUJYWBi8vb0RGRmJHTt22LtJBOCXX35Bnz59EBoaCp1Oh6+++srocaUUZs6ciQYNGsDHxwfR0dE4dOiQfRpbzcXHx6NLly7w9/dH/fr10a9fPyQnJxsdc/36dcTGxqJOnTrw8/PDgAEDkJaWZqcWV19LlixB+/btC4u2RUVF4fvvvy98nK+T45o/fz50Oh0mTpxYeJ8rvl4Maqpg9erViIuLw6xZs7Br1y506NABMTExOHfunL2bVu1lZ2ejQ4cOWLRoUamPv/rqq3j77bexdOlSbN++HTVr1kRMTAyuX79u45bSli1bEBsbi23btmHjxo3Iy8vD3Xffjezs7MJjJk2ahG+//RZr1qzBli1bcPbsWTzwwAN2bHX11KhRI8yfPx9JSUn4448/0LNnT/Tt2xf79u0DwNfJUe3cuRPvvfce2rdvb3S/S75eiiotIiJCxcbGFt7W6/UqNDRUxcfH27FVVBwA9eWXXxbeLigoUCEhIeq1114rvC89PV15eXmpTz/91A4tpKLOnTunAKgtW7YopeS18fDwUGvWrCk85sCBAwqASkxMtFcz6V+1a9dWH3zwAV8nB3XlyhXVsmVLtXHjRnX77berp59+Winlun9X7KmppNzcXCQlJSE6OrrwPjc3N0RHRyMxMdGOLaOKHDt2DKmpqUavXWBgICIjI/naOYCMjAwAQFBQEAAgKSkJeXl5Rq9X69at0aRJE75edqTX67Fq1SpkZ2cjKiqKr5ODio2NRe/evY1eF8B1/66qzYKWlnbhwgXo9XoEBwcb3R8cHIyDBw/aqVVkitTUVAAo9bXTHiP7KCgowMSJE9GtWze0bdsWgLxenp6eqFWrltGxfL3s4++//0ZUVBSuX78OPz8/fPnll2jTpg12797N18nBrFq1Crt27cLOnTtLPOaqf1cMaojIYcTGxmLv3r3YunWrvZtCZWjVqhV2796NjIwMfP755xg+fDi2bNli72ZRMadOncLTTz+NjRs3wtvb297NsRkOP1VS3bp14e7uXiJTPC0tDSEhIXZqFZlCe3342jmW8ePHY926ddi0aRMaNWpUeH9ISAhyc3ORnp5udDxfL/vw9PREixYtEB4ejvj4eHTo0AFvvfUWXycHk5SUhHPnzuGWW25BjRo1UKNGDWzZsgVvv/02atSogeDgYJd8vRjUVJKnpyfCw8ORkJBQeF9BQQESEhIQFRVlx5ZRRZo1a4aQkBCj1y4zMxPbt2/na2cHSimMHz8eX375JX7++Wc0a9bM6PHw8HB4eHgYvV7Jyck4efIkXy8HUFBQgJycHL5ODubOO+/E33//jd27dxdunTt3xpAhQwr3XfH14vBTFcTFxWH48OHo3LkzIiIisHDhQmRnZ2PkyJH2blq1l5WVhcOHDxfePnbsGHbv3o2goCA0adIEEydOxEsvvYSWLVuiWbNmmDFjBkJDQ9GvXz/7Nbqaio2NxcqVK/H111/D39+/cDw/MDAQPj4+CAwMxOjRoxEXF4egoCAEBARgwoQJiIqKwq233mrn1lcv06ZNw7333osmTZrgypUrWLlyJTZv3owffviBr5OD8ff3L8xL09SsWRN16tQpvN8lXy97T79ydu+8845q0qSJ8vT0VBEREWrbtm32bhIppTZt2qQAlNiGDx+ulJJp3TNmzFDBwcHKy8tL3XnnnSo5Odm+ja6mSnudAKj//ve/hcdcu3ZNPfnkk6p27drK19dX9e/fX6WkpNiv0dXUqFGjVNOmTZWnp6eqV6+euvPOO9WPP/5Y+DhfJ8dWdEq3Uq75eumUUspO8RQRERGRxTCnhoiIiFwCgxoiIiJyCQxqiIiIyCUwqCEiIiKXwKCGiIiIXAKDGiIiInIJDGqIiIjIJTCoISIiIpfAoIaIiIhcAoMaIiIicgkMaoiIiMglMKghIiIil/D/Z9yVwVIfmQgAAAAASUVORK5CYII=\n"},"metadata":{}}]},{"cell_type":"code","source":["plt.plot(history.history['loss'],color='red',label='train')\n","plt.plot(history.history['val_loss'],color='blue',label='validation')\n","plt.legend()\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":430},"id":"c_Lo1auZgnzY","executionInfo":{"status":"ok","timestamp":1687964277958,"user_tz":-330,"elapsed":8,"user":{"displayName":"Anuj Khandelwal","userId":"13654942296751135894"}},"outputId":"7e577462-5450-4b24-a5fc-109d0884a248"},"execution_count":30,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":[],"metadata":{"id":"b9hGvFokgtX5"},"execution_count":null,"outputs":[]}]} |