{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": { "id": "KNdNi1HSHfsn" }, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np" ] }, { "cell_type": "code", "source": [ "from google.colab import drive\n", "drive.mount('/content/drive')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "yOonl0NnHg0A", "outputId": "501a5acf-26c0-4e14-8396-e962141df5fb" }, "execution_count": 3, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Mounted at /content/drive\n" ] } ] }, { "cell_type": "markdown", "source": [ "#Data processing" ], "metadata": { "id": "ydyYNFA5HvaK" } }, { "cell_type": "code", "source": [ "data=pd.read_csv('your data path')" ], "metadata": { "id": "IS99jGeIHjVV" }, "execution_count": 5, "outputs": [] }, { "cell_type": "code", "source": [ "data.head()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 206 }, "id": "eraHgXlqHkfS", "outputId": "a917e036-69df-4088-d1d3-2fc24f254687" }, "execution_count": 6, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " timestamp number gas_used gas_limit base_fee_per_gas\n", "0 2023-03-21 00:00:11 UTC 16872313 29872569 30000000 14709149536\n", "1 2023-03-21 00:00:23 UTC 16872314 13937940 30000000 16532173214\n", "2 2023-03-21 00:00:35 UTC 16872315 19559490 30000000 16385855215\n", "3 2023-03-21 00:00:47 UTC 16872316 13826473 30000000 17008448073\n", "4 2023-03-21 00:00:59 UTC 16872317 20820719 30000000 16842115798" ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
timestampnumbergas_usedgas_limitbase_fee_per_gas
02023-03-21 00:00:11 UTC16872313298725693000000014709149536
12023-03-21 00:00:23 UTC16872314139379403000000016532173214
22023-03-21 00:00:35 UTC16872315195594903000000016385855215
32023-03-21 00:00:47 UTC16872316138264733000000017008448073
42023-03-21 00:00:59 UTC16872317208207193000000016842115798
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ] }, "metadata": {}, "execution_count": 6 } ] }, { "cell_type": "code", "source": [ "import pandas as pd\n", "new_df = data.drop_duplicates(subset='number', keep='first')" ], "metadata": { "id": "jiTWj4HLHlnT" }, "execution_count": 7, "outputs": [] }, { "cell_type": "code", "source": [ "new_df=new_df[[\"timestamp\",\"number\",\"gas_used\",\"gas_limit\",\"base_fee_per_gas\"]]" ], "metadata": { "id": "Xod1KfQAHnE5" }, "execution_count": 8, "outputs": [] }, { "cell_type": "code", "source": [ "new_df[\"base_fee_per_gas\"] = new_df[\"base_fee_per_gas\"]*10**-9" ], "metadata": { "id": "mKOc4GVpHo0e" }, "execution_count": 9, "outputs": [] }, { "cell_type": "code", "source": [ "new_df.head()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 206 }, "id": "zgqXTNOuHqEJ", "outputId": "0a2f0f0e-8ccf-4b86-b9fa-17ace8f924c4" }, "execution_count": 10, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " timestamp number gas_used gas_limit base_fee_per_gas\n", "0 2023-03-21 00:00:11 UTC 16872313 29872569 30000000 14.709150\n", "1 2023-03-21 00:00:23 UTC 16872314 13937940 30000000 16.532173\n", "2 2023-03-21 00:00:35 UTC 16872315 19559490 30000000 16.385855\n", "3 2023-03-21 00:00:47 UTC 16872316 13826473 30000000 17.008448\n", "4 2023-03-21 00:00:59 UTC 16872317 20820719 30000000 16.842116" ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
timestampnumbergas_usedgas_limitbase_fee_per_gas
02023-03-21 00:00:11 UTC16872313298725693000000014.709150
12023-03-21 00:00:23 UTC16872314139379403000000016.532173
22023-03-21 00:00:35 UTC16872315195594903000000016.385855
32023-03-21 00:00:47 UTC16872316138264733000000017.008448
42023-03-21 00:00:59 UTC16872317208207193000000016.842116
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ] }, "metadata": {}, "execution_count": 10 } ] }, { "cell_type": "code", "source": [ "new_df['gas_fraction'] = new_df['gas_used'] / new_df['gas_limit']" ], "metadata": { "id": "nKx-8k_NHrRY" }, "execution_count": 11, "outputs": [] }, { "cell_type": "code", "source": [ "new_df['gas_target'] = (new_df['gas_used']-(new_df['gas_limit']/2)) / (new_df['gas_limit']/2)" ], "metadata": { "id": "mAXOeH0iHsTa" }, "execution_count": 12, "outputs": [] }, { "cell_type": "code", "source": [ "org_data=np.array(new_df)" ], "metadata": { "id": "volE2hx-HuK9" }, "execution_count": 13, "outputs": [] }, { "cell_type": "markdown", "source": [ "## 4 Machine learning medhods" ], "metadata": { "id": "jUIk9aW9K3Uw" } }, { "cell_type": "code", "source": [ "from sklearn.model_selection import TimeSeriesSplit\n", "from sklearn import linear_model\n", "from sklearn.metrics import mean_squared_error\n", "import tensorflow as tf\n", "import tensorflow.compat.v2 as tf\n", "import matplotlib.pyplot as plt\n", "import xgboost" ], "metadata": { "id": "-H92WFDWH1tH" }, "execution_count": 14, "outputs": [] }, { "cell_type": "code", "source": [ "import copy\n", "x=[]\n", "y=[]\n", "for i in range(len(org_data)-10):\n", " base=org_data[i:i+11,4].tolist()\n", " gas_frc=org_data[i:i+10,5].tolist()\n", " now=np.concatenate((base, gas_frc),axis=0)\n", " now=now.tolist()\n", " x.append(now)\n", " y.append(org_data[i+10,6])\n", "x=np.array(x)\n", "y=np.array(y)" ], "metadata": { "id": "VL7k8SaiI5ZH" }, "execution_count": 15, "outputs": [] }, { "cell_type": "markdown", "source": [ "## Linear Regression" ], "metadata": { "id": "oSqmXzT4M2LJ" } }, { "cell_type": "code", "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "from sklearn.model_selection import TimeSeriesSplit\n", "from sklearn import linear_model\n", "from sklearn.metrics import mean_squared_error\n", "\n", "reg = linear_model.LinearRegression()\n", "train_indices = []\n", "validation_indices = []\n", "test_indices = []\n", "n_splits = 5\n", "tss = TimeSeriesSplit(n_splits)\n", "tss2 = TimeSeriesSplit(2)\n", "\n", "for train_index, test_index in tss.split(x):\n", " x_train, x_test = x[train_index, :], x[test_index, :]\n", " y_train, y_test = y[train_index], y[test_index]\n", " for train_index, vali_index in tss2.split(x_train):\n", " x_train, x_vali = x[train_index, :], x[vali_index, :]\n", " y_train, y_vali = y[train_index], y[vali_index]\n", " train_indices.append(train_index)\n", " validation_indices.append(vali_index)\n", " test_indices.append(test_index)\n", " reg.fit(x_train, y_train)\n", " pred_reg = reg.predict(x_test)\n", " print(mean_squared_error(y_test, pred_reg))\n", "\n", "offset = 0.1\n", "\n", "plt.figure(figsize=(12, 6))\n", "for i in range(n_splits):\n", "\n", " plt.scatter(train_indices[i], np.full_like(train_indices[i], i + 1) - offset, label=f'Fold {i + 1} - Train', marker='o', color='blue')\n", "\n", " plt.scatter(validation_indices[i], np.full_like(validation_indices[i], i + 1), label=f'Fold {i + 1} - Validation', marker='s', color='orange')\n", "\n", " plt.scatter(test_indices[i], np.full_like(test_indices[i], i + 1) + offset, label=f'Fold {i + 1} - Test', marker='^', color='green')\n", "\n", "plt.yticks(range(1, n_splits + 1), labels=[f'Fold {i + 1}' for i in range(n_splits)], fontsize=14)\n", "plt.xlabel('Data Index', fontsize=16)\n", "plt.title('Position of Training, Validation, and Test Data', fontsize=18)\n", "plt.legend(fontsize=10)\n", "plt.show()\n" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 538 }, "id": "rr6m7lRSIIkA", "outputId": "afbf29a3-f95a-4145-8905-ef673126bdc6" }, "execution_count": 16, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "0.12497945007235127\n", "0.1232660189285254\n", "0.11655808500097173\n", "0.1088873888698169\n", "0.1213640078300685\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "##DNN" ], "metadata": { "id": "p4ctDIuuMzgF" } }, { "cell_type": "code", "source": [ "from sklearn.model_selection import TimeSeriesSplit\n", "from sklearn import linear_model\n", "from sklearn.metrics import mean_squared_error\n", "reg = linear_model.LinearRegression()\n", "train_indices = []\n", "validation_indices = []\n", "test_indices = []\n", "n_splits = 5\n", "result=[]\n", "tss = TimeSeriesSplit(n_splits)\n", "tss2 = TimeSeriesSplit(2)\n", "for train_index, test_index in tss.split(x):\n", " tf.enable_v2_behavior()\n", " model_stability = tf.keras.models.Sequential([\n", " tf.keras.layers.Dense(128, activation = 'tanh',use_bias=True),\n", " tf.keras.layers.Dense(128, activation = 'linear',use_bias=True),\n", " tf.keras.layers.Dense(128, activation = 'tanh',use_bias=True),\n", " tf.keras.layers.Dense(1)\n", " ]\n", " )\n", " model_stability.compile(loss=tf.keras.losses.mean_squared_error,\n", " optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),\n", " metrics=['mse'])\n", "\n", " training_callbacks = [\n", " tf.keras.callbacks.ReduceLROnPlateau(patience = 7, factor = 0.1, min_lr = 0.00001, verbose = 1),\n", " tf.keras.callbacks.EarlyStopping(patience = 12, restore_best_weights = True, verbose=1),\n", " ]\n", "\n", " x_train, x_test = x[train_index, :], x[test_index,:]\n", " y_train, y_test = y[train_index], y[test_index]\n", " for train_index, vali_index in tss2.split(x_train):\n", " x_train, x_vali = x[train_index, :], x[vali_index,:]\n", " y_train, y_vali = y[train_index], y[vali_index]\n", " train_indices.append(train_index)\n", " validation_indices.append(vali_index)\n", " test_indices.append(test_index)\n", " model_stability.fit(x_train, y_train, batch_size=64, epochs=100,callbacks=training_callbacks,validation_data=(x_vali,y_vali))\n", " pred_reg=model_stability.predict(x_test)\n", " result.append(mean_squared_error(y_test, pred_reg))\n", " print(mean_squared_error(y_test, pred_reg))\n" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "XK2yTNGQIR2z", "outputId": "6c467062-f094-4095-90be-b8cbc9d14d1f" }, "execution_count": 17, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Epoch 1/100\n", "136/136 [==============================] - 2s 6ms/step - loss: 0.1843 - mse: 0.1843 - val_loss: 0.1281 - val_mse: 0.1281 - lr: 0.0010\n", "Epoch 2/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1298 - mse: 0.1298 - val_loss: 0.1227 - val_mse: 0.1227 - lr: 0.0010\n", "Epoch 3/100\n", "136/136 [==============================] - 1s 6ms/step - loss: 0.1262 - mse: 0.1262 - val_loss: 0.1199 - val_mse: 0.1199 - lr: 0.0010\n", "Epoch 4/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1272 - mse: 0.1272 - val_loss: 0.1298 - val_mse: 0.1298 - lr: 0.0010\n", "Epoch 5/100\n", "136/136 [==============================] - 1s 6ms/step - loss: 0.1255 - mse: 0.1255 - val_loss: 0.1170 - val_mse: 0.1170 - lr: 0.0010\n", "Epoch 6/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1182 - mse: 0.1182 - val_loss: 0.1185 - val_mse: 0.1185 - lr: 0.0010\n", "Epoch 7/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1205 - mse: 0.1205 - val_loss: 0.1155 - val_mse: 0.1155 - lr: 0.0010\n", "Epoch 8/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1151 - mse: 0.1151 - val_loss: 0.1130 - val_mse: 0.1130 - lr: 0.0010\n", "Epoch 9/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1120 - mse: 0.1120 - val_loss: 0.1087 - val_mse: 0.1087 - lr: 0.0010\n", "Epoch 10/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1113 - mse: 0.1113 - val_loss: 0.1111 - val_mse: 0.1111 - lr: 0.0010\n", "Epoch 11/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1172 - mse: 0.1172 - val_loss: 0.1135 - val_mse: 0.1135 - lr: 0.0010\n", "Epoch 12/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1114 - mse: 0.1114 - val_loss: 0.1164 - val_mse: 0.1164 - lr: 0.0010\n", "Epoch 13/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1170 - mse: 0.1170 - val_loss: 0.1149 - val_mse: 0.1149 - lr: 0.0010\n", "Epoch 14/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1111 - mse: 0.1111 - val_loss: 0.1097 - val_mse: 0.1097 - lr: 0.0010\n", "Epoch 15/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1098 - mse: 0.1098 - val_loss: 0.1076 - val_mse: 0.1076 - lr: 0.0010\n", "Epoch 16/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1100 - mse: 0.1100 - val_loss: 0.1179 - val_mse: 0.1179 - lr: 0.0010\n", "Epoch 17/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1125 - mse: 0.1125 - val_loss: 0.1062 - val_mse: 0.1062 - lr: 0.0010\n", "Epoch 18/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1141 - mse: 0.1141 - val_loss: 0.1094 - val_mse: 0.1094 - lr: 0.0010\n", "Epoch 19/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1147 - mse: 0.1147 - val_loss: 0.1315 - val_mse: 0.1315 - lr: 0.0010\n", "Epoch 20/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1212 - mse: 0.1212 - val_loss: 0.1268 - val_mse: 0.1268 - lr: 0.0010\n", "Epoch 21/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1117 - mse: 0.1117 - val_loss: 0.1093 - val_mse: 0.1093 - lr: 0.0010\n", "Epoch 22/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1151 - mse: 0.1151 - val_loss: 0.1187 - val_mse: 0.1187 - lr: 0.0010\n", "Epoch 23/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1141 - mse: 0.1141 - val_loss: 0.1138 - val_mse: 0.1138 - lr: 0.0010\n", "Epoch 24/100\n", "114/136 [========================>.....] - ETA: 0s - loss: 0.1113 - mse: 0.1113\n", "Epoch 24: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1118 - mse: 0.1118 - val_loss: 0.1118 - val_mse: 0.1118 - lr: 0.0010\n", "Epoch 25/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1081 - mse: 0.1081 - val_loss: 0.1084 - val_mse: 0.1084 - lr: 1.0000e-04\n", "Epoch 26/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1068 - mse: 0.1068 - val_loss: 0.1074 - val_mse: 0.1074 - lr: 1.0000e-04\n", "Epoch 27/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1065 - mse: 0.1065 - val_loss: 0.1076 - val_mse: 0.1076 - lr: 1.0000e-04\n", "Epoch 28/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1063 - mse: 0.1063 - val_loss: 0.1066 - val_mse: 0.1066 - lr: 1.0000e-04\n", "Epoch 29/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1062 - mse: 0.1062 - val_loss: 0.1057 - val_mse: 0.1057 - lr: 1.0000e-04\n", "Epoch 30/100\n", "136/136 [==============================] - 1s 6ms/step - loss: 0.1047 - mse: 0.1047 - val_loss: 0.1069 - val_mse: 0.1069 - lr: 1.0000e-04\n", "Epoch 31/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1051 - mse: 0.1051 - val_loss: 0.1049 - val_mse: 0.1049 - lr: 1.0000e-04\n", "Epoch 32/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1043 - mse: 0.1043 - val_loss: 0.1058 - val_mse: 0.1058 - lr: 1.0000e-04\n", "Epoch 33/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1046 - mse: 0.1046 - val_loss: 0.1048 - val_mse: 0.1048 - lr: 1.0000e-04\n", "Epoch 34/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.1043 - val_mse: 0.1043 - lr: 1.0000e-04\n", "Epoch 35/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1041 - mse: 0.1041 - val_loss: 0.1043 - val_mse: 0.1043 - lr: 1.0000e-04\n", "Epoch 36/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1045 - mse: 0.1045 - val_loss: 0.1043 - val_mse: 0.1043 - lr: 1.0000e-04\n", "Epoch 37/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1041 - mse: 0.1041 - val_loss: 0.1044 - val_mse: 0.1044 - lr: 1.0000e-04\n", "Epoch 38/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1048 - mse: 0.1048 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 1.0000e-04\n", "Epoch 39/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.1042 - val_mse: 0.1042 - lr: 1.0000e-04\n", "Epoch 40/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1044 - mse: 0.1044 - val_loss: 0.1044 - val_mse: 0.1044 - lr: 1.0000e-04\n", "Epoch 41/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1042 - val_mse: 0.1042 - lr: 1.0000e-04\n", "Epoch 42/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1050 - mse: 0.1050 - val_loss: 0.1043 - val_mse: 0.1043 - lr: 1.0000e-04\n", "Epoch 43/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1043 - mse: 0.1043 - val_loss: 0.1059 - val_mse: 0.1059 - lr: 1.0000e-04\n", "Epoch 44/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1041 - val_mse: 0.1041 - lr: 1.0000e-04\n", "Epoch 45/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1044 - mse: 0.1044 - val_loss: 0.1053 - val_mse: 0.1053 - lr: 1.0000e-04\n", "Epoch 46/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1062 - val_mse: 0.1062 - lr: 1.0000e-04\n", "Epoch 47/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1037 - mse: 0.1037 - val_loss: 0.1042 - val_mse: 0.1042 - lr: 1.0000e-04\n", "Epoch 48/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1043 - mse: 0.1043 - val_loss: 0.1039 - val_mse: 0.1039 - lr: 1.0000e-04\n", "Epoch 49/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1045 - mse: 0.1045 - val_loss: 0.1041 - val_mse: 0.1041 - lr: 1.0000e-04\n", "Epoch 50/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.1052 - val_mse: 0.1052 - lr: 1.0000e-04\n", "Epoch 51/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1034 - mse: 0.1034 - val_loss: 0.1042 - val_mse: 0.1042 - lr: 1.0000e-04\n", "Epoch 52/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1036 - mse: 0.1036 - val_loss: 0.1048 - val_mse: 0.1048 - lr: 1.0000e-04\n", "Epoch 53/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1041 - mse: 0.1041 - val_loss: 0.1056 - val_mse: 0.1056 - lr: 1.0000e-04\n", "Epoch 54/100\n", "136/136 [==============================] - 1s 6ms/step - loss: 0.1036 - mse: 0.1036 - val_loss: 0.1038 - val_mse: 0.1038 - lr: 1.0000e-04\n", "Epoch 55/100\n", "136/136 [==============================] - 1s 6ms/step - loss: 0.1032 - mse: 0.1032 - val_loss: 0.1045 - val_mse: 0.1045 - lr: 1.0000e-04\n", "Epoch 56/100\n", "136/136 [==============================] - 1s 10ms/step - loss: 0.1041 - mse: 0.1041 - val_loss: 0.1045 - val_mse: 0.1045 - lr: 1.0000e-04\n", "Epoch 57/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1038 - mse: 0.1038 - val_loss: 0.1042 - val_mse: 0.1042 - lr: 1.0000e-04\n", "Epoch 58/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1033 - mse: 0.1033 - val_loss: 0.1041 - val_mse: 0.1041 - lr: 1.0000e-04\n", "Epoch 59/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1034 - mse: 0.1034 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 1.0000e-04\n", "Epoch 60/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1038 - mse: 0.1038 - val_loss: 0.1045 - val_mse: 0.1045 - lr: 1.0000e-04\n", "Epoch 61/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1039 - mse: 0.1039 - val_loss: 0.1041 - val_mse: 0.1041 - lr: 1.0000e-04\n", "Epoch 62/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1034 - mse: 0.1034 - val_loss: 0.1037 - val_mse: 0.1037 - lr: 1.0000e-04\n", "Epoch 63/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1032 - mse: 0.1032 - val_loss: 0.1040 - val_mse: 0.1040 - lr: 1.0000e-04\n", "Epoch 64/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1033 - mse: 0.1033 - val_loss: 0.1037 - val_mse: 0.1037 - lr: 1.0000e-04\n", "Epoch 65/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1031 - mse: 0.1031 - val_loss: 0.1141 - val_mse: 0.1141 - lr: 1.0000e-04\n", "Epoch 66/100\n", "129/136 [===========================>..] - ETA: 0s - loss: 0.1032 - mse: 0.1032\n", "Epoch 66: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1037 - mse: 0.1037 - val_loss: 0.1042 - val_mse: 0.1042 - lr: 1.0000e-04\n", "Epoch 67/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 1.0000e-05\n", "Epoch 68/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 1.0000e-05\n", "Epoch 69/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 70/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 71/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 1.0000e-05\n", "Epoch 72/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1038 - val_mse: 0.1038 - lr: 1.0000e-05\n", "Epoch 73/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 1.0000e-05\n", "Epoch 74/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 75/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 76/100\n", "116/136 [========================>.....] - ETA: 0s - loss: 0.1012 - mse: 0.1012\n", "Epoch 76: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 77/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 78/100\n", "136/136 [==============================] - 1s 7ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 79/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1038 - val_mse: 0.1038 - lr: 1.0000e-05\n", "Epoch 80/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 81/100\n", "136/136 [==============================] - 1s 5ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 82/100\n", "136/136 [==============================] - 1s 6ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 83/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 84/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 85/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 86/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 87/100\n", "136/136 [==============================] - 0s 4ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 88/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 89/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 90/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 91/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 92/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 93/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 94/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 95/100\n", "136/136 [==============================] - 1s 4ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 1.0000e-05\n", "Epoch 96/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 97/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 98/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "Epoch 99/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-05\n", "Epoch 100/100\n", "136/136 [==============================] - 0s 3ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.1035 - val_mse: 0.1035 - lr: 1.0000e-05\n", "408/408 [==============================] - 1s 1ms/step\n", "0.10360595084384616\n", "Epoch 1/100\n", "272/272 [==============================] - 3s 6ms/step - loss: 0.1670 - mse: 0.1670 - val_loss: 0.1225 - val_mse: 0.1225 - lr: 0.0010\n", "Epoch 2/100\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1276 - mse: 0.1276 - val_loss: 0.1124 - val_mse: 0.1124 - lr: 0.0010\n", "Epoch 3/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1266 - mse: 0.1266 - val_loss: 0.1087 - val_mse: 0.1087 - lr: 0.0010\n", "Epoch 4/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1180 - mse: 0.1180 - val_loss: 0.1131 - val_mse: 0.1131 - lr: 0.0010\n", "Epoch 5/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1142 - mse: 0.1142 - val_loss: 0.1065 - val_mse: 0.1065 - lr: 0.0010\n", "Epoch 6/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1174 - mse: 0.1174 - val_loss: 0.1186 - val_mse: 0.1186 - lr: 0.0010\n", "Epoch 7/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1146 - mse: 0.1146 - val_loss: 0.1056 - val_mse: 0.1056 - lr: 0.0010\n", "Epoch 8/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1101 - mse: 0.1101 - val_loss: 0.1002 - val_mse: 0.1002 - lr: 0.0010\n", "Epoch 9/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1143 - mse: 0.1143 - val_loss: 0.1054 - val_mse: 0.1054 - lr: 0.0010\n", "Epoch 10/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1143 - mse: 0.1143 - val_loss: 0.1142 - val_mse: 0.1142 - lr: 0.0010\n", "Epoch 11/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1122 - mse: 0.1122 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 0.0010\n", "Epoch 12/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1141 - mse: 0.1141 - val_loss: 0.1024 - val_mse: 0.1024 - lr: 0.0010\n", "Epoch 13/100\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1130 - mse: 0.1130 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 14/100\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1157 - mse: 0.1157 - val_loss: 0.1033 - val_mse: 0.1033 - lr: 0.0010\n", "Epoch 15/100\n", "267/272 [============================>.] - ETA: 0s - loss: 0.1129 - mse: 0.1129\n", "Epoch 15: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1129 - mse: 0.1129 - val_loss: 0.1071 - val_mse: 0.1071 - lr: 0.0010\n", "Epoch 16/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1068 - mse: 0.1068 - val_loss: 0.0996 - val_mse: 0.0996 - lr: 1.0000e-04\n", "Epoch 17/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1021 - val_mse: 0.1021 - lr: 1.0000e-04\n", "Epoch 18/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1056 - mse: 0.1056 - val_loss: 0.0994 - val_mse: 0.0994 - lr: 1.0000e-04\n", "Epoch 19/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1059 - mse: 0.1059 - val_loss: 0.0993 - val_mse: 0.0993 - lr: 1.0000e-04\n", "Epoch 20/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1055 - mse: 0.1055 - val_loss: 0.1010 - val_mse: 0.1010 - lr: 1.0000e-04\n", "Epoch 21/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1054 - mse: 0.1054 - val_loss: 0.0993 - val_mse: 0.0993 - lr: 1.0000e-04\n", "Epoch 22/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1054 - mse: 0.1054 - val_loss: 0.0984 - val_mse: 0.0984 - lr: 1.0000e-04\n", "Epoch 23/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1053 - mse: 0.1053 - val_loss: 0.0990 - val_mse: 0.0990 - lr: 1.0000e-04\n", "Epoch 24/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1054 - mse: 0.1054 - val_loss: 0.0985 - val_mse: 0.0985 - lr: 1.0000e-04\n", "Epoch 25/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1049 - mse: 0.1049 - val_loss: 0.0994 - val_mse: 0.0994 - lr: 1.0000e-04\n", "Epoch 26/100\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1055 - mse: 0.1055 - val_loss: 0.1001 - val_mse: 0.1001 - lr: 1.0000e-04\n", "Epoch 27/100\n", "272/272 [==============================] - 2s 9ms/step - loss: 0.1049 - mse: 0.1049 - val_loss: 0.0992 - val_mse: 0.0992 - lr: 1.0000e-04\n", "Epoch 28/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1048 - mse: 0.1048 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-04\n", "Epoch 29/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1048 - mse: 0.1048 - val_loss: 0.0981 - val_mse: 0.0981 - lr: 1.0000e-04\n", "Epoch 30/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1046 - mse: 0.1046 - val_loss: 0.1004 - val_mse: 0.1004 - lr: 1.0000e-04\n", "Epoch 31/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1048 - mse: 0.1048 - val_loss: 0.0995 - val_mse: 0.0995 - lr: 1.0000e-04\n", "Epoch 32/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1043 - mse: 0.1043 - val_loss: 0.1008 - val_mse: 0.1008 - lr: 1.0000e-04\n", "Epoch 33/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.1001 - val_mse: 0.1001 - lr: 1.0000e-04\n", "Epoch 34/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1045 - mse: 0.1045 - val_loss: 0.1001 - val_mse: 0.1001 - lr: 1.0000e-04\n", "Epoch 35/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.1002 - val_mse: 0.1002 - lr: 1.0000e-04\n", "Epoch 36/100\n", "253/272 [==========================>...] - ETA: 0s - loss: 0.1038 - mse: 0.1038\n", "Epoch 36: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.0982 - val_mse: 0.0982 - lr: 1.0000e-04\n", "Epoch 37/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1029 - mse: 0.1029 - val_loss: 0.0983 - val_mse: 0.0983 - lr: 1.0000e-05\n", "Epoch 38/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1029 - mse: 0.1029 - val_loss: 0.0979 - val_mse: 0.0979 - lr: 1.0000e-05\n", "Epoch 39/100\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1029 - mse: 0.1029 - val_loss: 0.0982 - val_mse: 0.0982 - lr: 1.0000e-05\n", "Epoch 40/100\n", "272/272 [==============================] - 2s 6ms/step - loss: 0.1028 - mse: 0.1028 - val_loss: 0.0981 - val_mse: 0.0981 - lr: 1.0000e-05\n", "Epoch 41/100\n", "272/272 [==============================] - 1s 5ms/step - loss: 0.1028 - mse: 0.1028 - val_loss: 0.0982 - val_mse: 0.0982 - lr: 1.0000e-05\n", "Epoch 42/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1028 - mse: 0.1028 - val_loss: 0.0981 - val_mse: 0.0981 - lr: 1.0000e-05\n", "Epoch 43/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1029 - mse: 0.1029 - val_loss: 0.0980 - val_mse: 0.0980 - lr: 1.0000e-05\n", "Epoch 44/100\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0987 - val_mse: 0.0987 - lr: 1.0000e-05\n", "Epoch 45/100\n", "267/272 [============================>.] - ETA: 0s - loss: 0.1027 - mse: 0.1027\n", "Epoch 45: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0988 - val_mse: 0.0988 - lr: 1.0000e-05\n", "Epoch 46/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0994 - val_mse: 0.0994 - lr: 1.0000e-05\n", "Epoch 47/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0985 - val_mse: 0.0985 - lr: 1.0000e-05\n", "Epoch 48/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0994 - val_mse: 0.0994 - lr: 1.0000e-05\n", "Epoch 49/100\n", "272/272 [==============================] - 1s 4ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.0983 - val_mse: 0.0983 - lr: 1.0000e-05\n", "Epoch 50/100\n", "255/272 [===========================>..] - ETA: 0s - loss: 0.1027 - mse: 0.1027Restoring model weights from the end of the best epoch: 38.\n", "272/272 [==============================] - 1s 3ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0986 - val_mse: 0.0986 - lr: 1.0000e-05\n", "Epoch 50: early stopping\n", "408/408 [==============================] - 1s 2ms/step\n", "0.10487520836968875\n", "Epoch 1/100\n", "408/408 [==============================] - 4s 6ms/step - loss: 0.1544 - mse: 0.1544 - val_loss: 0.1251 - val_mse: 0.1251 - lr: 0.0010\n", "Epoch 2/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1221 - mse: 0.1221 - val_loss: 0.1158 - val_mse: 0.1158 - lr: 0.0010\n", "Epoch 3/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1146 - mse: 0.1146 - val_loss: 0.1243 - val_mse: 0.1243 - lr: 0.0010\n", "Epoch 4/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1122 - mse: 0.1122 - val_loss: 0.1119 - val_mse: 0.1119 - lr: 0.0010\n", "Epoch 5/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1117 - mse: 0.1117 - val_loss: 0.1081 - val_mse: 0.1081 - lr: 0.0010\n", "Epoch 6/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1115 - mse: 0.1115 - val_loss: 0.1118 - val_mse: 0.1118 - lr: 0.0010\n", "Epoch 7/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1137 - mse: 0.1137 - val_loss: 0.1088 - val_mse: 0.1088 - lr: 0.0010\n", "Epoch 8/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1150 - mse: 0.1150 - val_loss: 0.1093 - val_mse: 0.1093 - lr: 0.0010\n", "Epoch 9/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1096 - mse: 0.1096 - val_loss: 0.1079 - val_mse: 0.1079 - lr: 0.0010\n", "Epoch 10/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1126 - mse: 0.1126 - val_loss: 0.1148 - val_mse: 0.1148 - lr: 0.0010\n", "Epoch 11/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1126 - mse: 0.1126 - val_loss: 0.1106 - val_mse: 0.1106 - lr: 0.0010\n", "Epoch 12/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1110 - mse: 0.1110 - val_loss: 0.1128 - val_mse: 0.1128 - lr: 0.0010\n", "Epoch 13/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1101 - mse: 0.1101 - val_loss: 0.1180 - val_mse: 0.1180 - lr: 0.0010\n", "Epoch 14/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1108 - mse: 0.1108 - val_loss: 0.1295 - val_mse: 0.1295 - lr: 0.0010\n", "Epoch 15/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1099 - mse: 0.1099 - val_loss: 0.1282 - val_mse: 0.1282 - lr: 0.0010\n", "Epoch 16/100\n", "402/408 [============================>.] - ETA: 0s - loss: 0.1103 - mse: 0.1103\n", "Epoch 16: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1103 - mse: 0.1103 - val_loss: 0.1080 - val_mse: 0.1080 - lr: 0.0010\n", "Epoch 17/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1038 - mse: 0.1038 - val_loss: 0.1061 - val_mse: 0.1061 - lr: 1.0000e-04\n", "Epoch 18/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1030 - mse: 0.1030 - val_loss: 0.1066 - val_mse: 0.1066 - lr: 1.0000e-04\n", "Epoch 19/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1028 - mse: 0.1028 - val_loss: 0.1058 - val_mse: 0.1058 - lr: 1.0000e-04\n", "Epoch 20/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.1065 - val_mse: 0.1065 - lr: 1.0000e-04\n", "Epoch 21/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.1060 - val_mse: 0.1060 - lr: 1.0000e-04\n", "Epoch 22/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.1056 - val_mse: 0.1056 - lr: 1.0000e-04\n", "Epoch 23/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.1099 - val_mse: 0.1099 - lr: 1.0000e-04\n", "Epoch 24/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.1062 - val_mse: 0.1062 - lr: 1.0000e-04\n", "Epoch 25/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1025 - mse: 0.1025 - val_loss: 0.1056 - val_mse: 0.1056 - lr: 1.0000e-04\n", "Epoch 26/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.1055 - val_mse: 0.1055 - lr: 1.0000e-04\n", "Epoch 27/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1046 - val_mse: 0.1046 - lr: 1.0000e-04\n", "Epoch 28/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.1088 - val_mse: 0.1088 - lr: 1.0000e-04\n", "Epoch 29/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.1061 - val_mse: 0.1061 - lr: 1.0000e-04\n", "Epoch 30/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1015 - mse: 0.1015 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-04\n", "Epoch 31/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.1054 - val_mse: 0.1054 - lr: 1.0000e-04\n", "Epoch 32/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1016 - mse: 0.1016 - val_loss: 0.1048 - val_mse: 0.1048 - lr: 1.0000e-04\n", "Epoch 33/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1014 - mse: 0.1014 - val_loss: 0.1064 - val_mse: 0.1064 - lr: 1.0000e-04\n", "Epoch 34/100\n", "388/408 [===========================>..] - ETA: 0s - loss: 0.1017 - mse: 0.1017\n", "Epoch 34: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1015 - mse: 0.1015 - val_loss: 0.1053 - val_mse: 0.1053 - lr: 1.0000e-04\n", "Epoch 35/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1048 - val_mse: 0.1048 - lr: 1.0000e-05\n", "Epoch 36/100\n", "408/408 [==============================] - 2s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1044 - val_mse: 0.1044 - lr: 1.0000e-05\n", "Epoch 37/100\n", "408/408 [==============================] - 2s 6ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1049 - val_mse: 0.1049 - lr: 1.0000e-05\n", "Epoch 38/100\n", "408/408 [==============================] - 1s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1054 - val_mse: 0.1054 - lr: 1.0000e-05\n", "Epoch 39/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 1.0000e-05\n", "Epoch 40/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1044 - val_mse: 0.1044 - lr: 1.0000e-05\n", "Epoch 41/100\n", "408/408 [==============================] - 2s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1052 - val_mse: 0.1052 - lr: 1.0000e-05\n", "Epoch 42/100\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 1.0000e-05\n", "Epoch 43/100\n", "392/408 [===========================>..] - ETA: 0s - loss: 0.1001 - mse: 0.1001\n", "Epoch 43: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 1.0000e-05\n", "Epoch 44/100\n", "408/408 [==============================] - 1s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1046 - val_mse: 0.1046 - lr: 1.0000e-05\n", "Epoch 45/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 1.0000e-05\n", "Epoch 46/100\n", "408/408 [==============================] - 2s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1046 - val_mse: 0.1046 - lr: 1.0000e-05\n", "Epoch 47/100\n", "408/408 [==============================] - 1s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1045 - val_mse: 0.1045 - lr: 1.0000e-05\n", "Epoch 48/100\n", "399/408 [============================>.] - ETA: 0s - loss: 0.1003 - mse: 0.1003Restoring model weights from the end of the best epoch: 36.\n", "408/408 [==============================] - 1s 3ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1046 - val_mse: 0.1046 - lr: 1.0000e-05\n", "Epoch 48: early stopping\n", "408/408 [==============================] - 1s 2ms/step\n", "0.09402643477420265\n", "Epoch 1/100\n", "544/544 [==============================] - 3s 4ms/step - loss: 0.1330 - mse: 0.1330 - val_loss: 0.1184 - val_mse: 0.1184 - lr: 0.0010\n", "Epoch 2/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1140 - mse: 0.1140 - val_loss: 0.1145 - val_mse: 0.1145 - lr: 0.0010\n", "Epoch 3/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1092 - mse: 0.1092 - val_loss: 0.1082 - val_mse: 0.1082 - lr: 0.0010\n", "Epoch 4/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1103 - mse: 0.1103 - val_loss: 0.1121 - val_mse: 0.1121 - lr: 0.0010\n", "Epoch 5/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1110 - mse: 0.1110 - val_loss: 0.1068 - val_mse: 0.1068 - lr: 0.0010\n", "Epoch 6/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1078 - mse: 0.1078 - val_loss: 0.1229 - val_mse: 0.1229 - lr: 0.0010\n", "Epoch 7/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1079 - mse: 0.1079 - val_loss: 0.1130 - val_mse: 0.1130 - lr: 0.0010\n", "Epoch 8/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1102 - mse: 0.1102 - val_loss: 0.1087 - val_mse: 0.1087 - lr: 0.0010\n", "Epoch 9/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1101 - mse: 0.1101 - val_loss: 0.1069 - val_mse: 0.1069 - lr: 0.0010\n", "Epoch 10/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1091 - mse: 0.1091 - val_loss: 0.1191 - val_mse: 0.1191 - lr: 0.0010\n", "Epoch 11/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1124 - mse: 0.1124 - val_loss: 0.1270 - val_mse: 0.1270 - lr: 0.0010\n", "Epoch 12/100\n", "544/544 [==============================] - 2s 5ms/step - loss: 0.1105 - mse: 0.1105 - val_loss: 0.1064 - val_mse: 0.1064 - lr: 0.0010\n", "Epoch 13/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1093 - mse: 0.1093 - val_loss: 0.1201 - val_mse: 0.1201 - lr: 0.0010\n", "Epoch 14/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1103 - mse: 0.1103 - val_loss: 0.1107 - val_mse: 0.1107 - lr: 0.0010\n", "Epoch 15/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1065 - mse: 0.1065 - val_loss: 0.1146 - val_mse: 0.1146 - lr: 0.0010\n", "Epoch 16/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1061 - mse: 0.1061 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 17/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1095 - mse: 0.1095 - val_loss: 0.1065 - val_mse: 0.1065 - lr: 0.0010\n", "Epoch 18/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1090 - mse: 0.1090 - val_loss: 0.1165 - val_mse: 0.1165 - lr: 0.0010\n", "Epoch 19/100\n", "543/544 [============================>.] - ETA: 0s - loss: 0.1074 - mse: 0.1074\n", "Epoch 19: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1074 - mse: 0.1074 - val_loss: 0.1120 - val_mse: 0.1120 - lr: 0.0010\n", "Epoch 20/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1022 - mse: 0.1022 - val_loss: 0.1041 - val_mse: 0.1041 - lr: 1.0000e-04\n", "Epoch 21/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1016 - mse: 0.1016 - val_loss: 0.1034 - val_mse: 0.1034 - lr: 1.0000e-04\n", "Epoch 22/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1011 - mse: 0.1011 - val_loss: 0.1037 - val_mse: 0.1037 - lr: 1.0000e-04\n", "Epoch 23/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1011 - mse: 0.1011 - val_loss: 0.1029 - val_mse: 0.1029 - lr: 1.0000e-04\n", "Epoch 24/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1012 - mse: 0.1012 - val_loss: 0.1056 - val_mse: 0.1056 - lr: 1.0000e-04\n", "Epoch 25/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.1029 - val_mse: 0.1029 - lr: 1.0000e-04\n", "Epoch 26/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.1031 - val_mse: 0.1031 - lr: 1.0000e-04\n", "Epoch 27/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.1027 - val_mse: 0.1027 - lr: 1.0000e-04\n", "Epoch 28/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-04\n", "Epoch 29/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.1021 - val_mse: 0.1021 - lr: 1.0000e-04\n", "Epoch 30/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1031 - val_mse: 0.1031 - lr: 1.0000e-04\n", "Epoch 31/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.1005 - mse: 0.1005 - val_loss: 0.1027 - val_mse: 0.1027 - lr: 1.0000e-04\n", "Epoch 32/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.1032 - val_mse: 0.1032 - lr: 1.0000e-04\n", "Epoch 33/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.1023 - val_mse: 0.1023 - lr: 1.0000e-04\n", "Epoch 34/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1030 - val_mse: 0.1030 - lr: 1.0000e-04\n", "Epoch 35/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1026 - val_mse: 0.1026 - lr: 1.0000e-04\n", "Epoch 36/100\n", "542/544 [============================>.] - ETA: 0s - loss: 0.1002 - mse: 0.1002\n", "Epoch 36: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1022 - val_mse: 0.1022 - lr: 1.0000e-04\n", "Epoch 37/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0992 - mse: 0.0992 - val_loss: 0.1020 - val_mse: 0.1020 - lr: 1.0000e-05\n", "Epoch 38/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0992 - mse: 0.0992 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 39/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0991 - mse: 0.0991 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 40/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0991 - mse: 0.0991 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 41/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0991 - mse: 0.0991 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 42/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 43/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 44/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1021 - val_mse: 0.1021 - lr: 1.0000e-05\n", "Epoch 45/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0991 - mse: 0.0991 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 46/100\n", "540/544 [============================>.] - ETA: 0s - loss: 0.0990 - mse: 0.0990\n", "Epoch 46: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 47/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1020 - val_mse: 0.1020 - lr: 1.0000e-05\n", "Epoch 48/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 49/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 50/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 51/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 52/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-05\n", "Epoch 53/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-05\n", "Epoch 54/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 55/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 56/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 57/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1016 - val_mse: 0.1016 - lr: 1.0000e-05\n", "Epoch 58/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-05\n", "Epoch 59/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 60/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1016 - val_mse: 0.1016 - lr: 1.0000e-05\n", "Epoch 61/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 62/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-05\n", "Epoch 63/100\n", "544/544 [==============================] - 2s 3ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 64/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-05\n", "Epoch 65/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 66/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0988 - mse: 0.0988 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 67/100\n", "544/544 [==============================] - 3s 5ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-05\n", "Epoch 68/100\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0989 - mse: 0.0989 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-05\n", "Epoch 69/100\n", "534/544 [============================>.] - ETA: 0s - loss: 0.0992 - mse: 0.0992Restoring model weights from the end of the best epoch: 57.\n", "544/544 [==============================] - 2s 4ms/step - loss: 0.0988 - mse: 0.0988 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-05\n", "Epoch 69: early stopping\n", "408/408 [==============================] - 1s 2ms/step\n", "0.09130745273371572\n", "Epoch 1/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1441 - mse: 0.1441 - val_loss: 0.1096 - val_mse: 0.1096 - lr: 0.0010\n", "Epoch 2/100\n", "680/680 [==============================] - 4s 5ms/step - loss: 0.1189 - mse: 0.1189 - val_loss: 0.1007 - val_mse: 0.1007 - lr: 0.0010\n", "Epoch 3/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1132 - mse: 0.1132 - val_loss: 0.1013 - val_mse: 0.1013 - lr: 0.0010\n", "Epoch 4/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1121 - mse: 0.1121 - val_loss: 0.1005 - val_mse: 0.1005 - lr: 0.0010\n", "Epoch 5/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1112 - mse: 0.1112 - val_loss: 0.0966 - val_mse: 0.0966 - lr: 0.0010\n", "Epoch 6/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1123 - mse: 0.1123 - val_loss: 0.1009 - val_mse: 0.1009 - lr: 0.0010\n", "Epoch 7/100\n", "680/680 [==============================] - 4s 5ms/step - loss: 0.1104 - mse: 0.1104 - val_loss: 0.1091 - val_mse: 0.1091 - lr: 0.0010\n", "Epoch 8/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1103 - mse: 0.1103 - val_loss: 0.0974 - val_mse: 0.0974 - lr: 0.0010\n", "Epoch 9/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1112 - mse: 0.1112 - val_loss: 0.0975 - val_mse: 0.0975 - lr: 0.0010\n", "Epoch 10/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1093 - mse: 0.1093 - val_loss: 0.0985 - val_mse: 0.0985 - lr: 0.0010\n", "Epoch 11/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1101 - mse: 0.1101 - val_loss: 0.1074 - val_mse: 0.1074 - lr: 0.0010\n", "Epoch 12/100\n", "667/680 [============================>.] - ETA: 0s - loss: 0.1085 - mse: 0.1085\n", "Epoch 12: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1084 - mse: 0.1084 - val_loss: 0.1043 - val_mse: 0.1043 - lr: 0.0010\n", "Epoch 13/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1036 - mse: 0.1036 - val_loss: 0.0947 - val_mse: 0.0947 - lr: 1.0000e-04\n", "Epoch 14/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1030 - mse: 0.1030 - val_loss: 0.0951 - val_mse: 0.0951 - lr: 1.0000e-04\n", "Epoch 15/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.0951 - val_mse: 0.0951 - lr: 1.0000e-04\n", "Epoch 16/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.0959 - val_mse: 0.0959 - lr: 1.0000e-04\n", "Epoch 17/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.0951 - val_mse: 0.0951 - lr: 1.0000e-04\n", "Epoch 18/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.0942 - val_mse: 0.0942 - lr: 1.0000e-04\n", "Epoch 19/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.0940 - val_mse: 0.0940 - lr: 1.0000e-04\n", "Epoch 20/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.0930 - val_mse: 0.0930 - lr: 1.0000e-04\n", "Epoch 21/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.0932 - val_mse: 0.0932 - lr: 1.0000e-04\n", "Epoch 22/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.0946 - val_mse: 0.0946 - lr: 1.0000e-04\n", "Epoch 23/100\n", "680/680 [==============================] - 4s 6ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.0954 - val_mse: 0.0954 - lr: 1.0000e-04\n", "Epoch 24/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1015 - mse: 0.1015 - val_loss: 0.0933 - val_mse: 0.0933 - lr: 1.0000e-04\n", "Epoch 25/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1016 - mse: 0.1016 - val_loss: 0.0931 - val_mse: 0.0931 - lr: 1.0000e-04\n", "Epoch 26/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1016 - mse: 0.1016 - val_loss: 0.0944 - val_mse: 0.0944 - lr: 1.0000e-04\n", "Epoch 27/100\n", "673/680 [============================>.] - ETA: 0s - loss: 0.1014 - mse: 0.1014\n", "Epoch 27: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1014 - mse: 0.1014 - val_loss: 0.0938 - val_mse: 0.0938 - lr: 1.0000e-04\n", "Epoch 28/100\n", "680/680 [==============================] - 4s 6ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 29/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-05\n", "Epoch 30/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 31/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0933 - val_mse: 0.0933 - lr: 1.0000e-05\n", "Epoch 32/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 33/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 34/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 35/100\n", "675/680 [============================>.] - ETA: 0s - loss: 0.1002 - mse: 0.1002\n", "Epoch 35: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-05\n", "Epoch 36/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-05\n", "Epoch 37/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 38/100\n", "680/680 [==============================] - 4s 5ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0935 - val_mse: 0.0935 - lr: 1.0000e-05\n", "Epoch 39/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 40/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 41/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 42/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 43/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0931 - val_mse: 0.0931 - lr: 1.0000e-05\n", "Epoch 44/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-05\n", "Epoch 45/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 46/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-05\n", "Epoch 47/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 48/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 49/100\n", "680/680 [==============================] - 4s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0930 - val_mse: 0.0930 - lr: 1.0000e-05\n", "Epoch 50/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0931 - val_mse: 0.0931 - lr: 1.0000e-05\n", "Epoch 51/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-05\n", "Epoch 52/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-05\n", "Epoch 53/100\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 54/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 55/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 56/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 57/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 58/100\n", "680/680 [==============================] - 2s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-05\n", "Epoch 59/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 60/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-05\n", "Epoch 61/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 62/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 63/100\n", "680/680 [==============================] - 3s 4ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-05\n", "Epoch 64/100\n", "680/680 [==============================] - 4s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-05\n", "Epoch 65/100\n", "680/680 [==============================] - 3s 5ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-05\n", "Epoch 66/100\n", "669/680 [============================>.] - ETA: 0s - loss: 0.1002 - mse: 0.1002Restoring model weights from the end of the best epoch: 54.\n", "680/680 [==============================] - 2s 3ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-05\n", "Epoch 66: early stopping\n", "408/408 [==============================] - 1s 2ms/step\n", "0.10088310862590738\n" ] } ] }, { "cell_type": "markdown", "source": [ "## XGBoost" ], "metadata": { "id": "avDtKnShM455" } }, { "cell_type": "code", "source": [ "from sklearn.model_selection import TimeSeriesSplit\n", "from sklearn import linear_model\n", "from sklearn.metrics import mean_squared_error\n", "reg = linear_model.LinearRegression()\n", "train_indices = []\n", "validation_indices = []\n", "test_indices = []\n", "n_splits = 5\n", "result=[]\n", "tss = TimeSeriesSplit(n_splits)\n", "tss2 = TimeSeriesSplit(2)\n", "for train_index, test_index in tss.split(x):\n", " x_train, x_test = x[train_index, :], x[test_index,:]\n", " y_train, y_test = y[train_index], y[test_index]\n", " for train_index, vali_index in tss2.split(x_train):\n", " x_train, x_vali = x[train_index, :], x[vali_index,:]\n", " y_train, y_vali = y[train_index], y[vali_index]\n", " train_indices.append(train_index)\n", " validation_indices.append(vali_index)\n", " test_indices.append(test_index)\n", " model = xgboost.XGBRegressor()\n", " model.fit(x_train, y_train)\n", " pred_reg=model.predict(x_test)\n", " result.append(mean_squared_error(y_test, pred_reg))\n", " print(mean_squared_error(y_test, pred_reg))\n" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "7bTYxgPLIVoG", "outputId": "dbef8317-8756-4064-9299-a95abc44bad2" }, "execution_count": 18, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "0.12260423842432859\n", "0.1129450415694429\n", "0.09961083337293349\n", "0.09394534402873217\n", "0.10429804165796275\n" ] } ] }, { "cell_type": "markdown", "source": [ "##Lstm" ], "metadata": { "id": "1fGesdH-M7p4" } }, { "cell_type": "code", "source": [ "import copy\n", "x=[]\n", "y=[]\n", "for i in range(len(org_data)-10):\n", " temp=[[0,0]]\n", " base=org_data[i:i+10,4:6].tolist()\n", " temp[0][0]=org_data[i+10,4]\n", " base.extend(temp)\n", " x.append(base)\n", " y.append(org_data[i+10,6])\n", "\n", "x=np.array(x)\n", "y=np.array(y)" ], "metadata": { "id": "xPPVwQM2IX7E" }, "execution_count": 19, "outputs": [] }, { "cell_type": "code", "source": [ "from sklearn.model_selection import TimeSeriesSplit\n", "from sklearn import linear_model\n", "from sklearn.metrics import mean_squared_error\n", "reg = linear_model.LinearRegression()\n", "n_splits = 5\n", "result=[]\n", "tss = TimeSeriesSplit(n_splits)\n", "tss2 = TimeSeriesSplit(2)\n", "for train_index, test_index in tss.split(x):\n", " x_train, x_test = x[train_index, :], x[test_index,:]\n", " y_train, y_test = y[train_index], y[test_index]\n", " for train_index, vali_index in tss2.split(x_train):\n", " x_train, x_vali = x[train_index, :], x[vali_index,:]\n", " y_train, y_vali = y[train_index], y[vali_index]\n", " train_indices.append(train_index)\n", " validation_indices.append(vali_index)\n", " test_indices.append(test_index)\n", " input_shape = (11, 2)\n", "\n", " model = tf.keras.Sequential([\n", " tf.keras.layers.LSTM(24, input_shape=input_shape, return_sequences=True),\n", " tf.keras.layers.LSTM(24, return_sequences=True),\n", " tf.keras.layers.LSTM(24),\n", " tf.keras.layers.Dense(100, activation='relu'),\n", " tf.keras.layers.Dense(1)\n", " ])\n", "\n", " training_callbacks = [\n", " tf.keras.callbacks.ReduceLROnPlateau(patience = 7, factor = 0.1, min_lr = 0.00001, verbose = 1),\n", " tf.keras.callbacks.EarlyStopping(patience = 12, restore_best_weights = True, verbose=1),\n", " ]\n", " model.compile(loss=tf.keras.losses.mean_squared_error,\n", " optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),\n", " metrics=['mse'])\n", "\n", " model.fit(x_train, y_train, batch_size=64, epochs=100,callbacks=training_callbacks,validation_data=(x_vali,y_vali))\n", " pred_reg=model.predict(x_test)\n", " result.append(mean_squared_error(y_test, pred_reg))\n", "\n" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "nENTppQ5IaxL", "outputId": "2a02a6cb-35e6-441d-e6ce-1dde5e34d491" }, "execution_count": 20, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Epoch 1/100\n", "136/136 [==============================] - 10s 31ms/step - loss: 0.1338 - mse: 0.1338 - val_loss: 0.1346 - val_mse: 0.1346 - lr: 0.0010\n", "Epoch 2/100\n", "136/136 [==============================] - 4s 31ms/step - loss: 0.1334 - mse: 0.1334 - val_loss: 0.1347 - val_mse: 0.1347 - lr: 0.0010\n", "Epoch 3/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1335 - mse: 0.1335 - val_loss: 0.1351 - val_mse: 0.1351 - lr: 0.0010\n", "Epoch 4/100\n", "136/136 [==============================] - 3s 24ms/step - loss: 0.1333 - mse: 0.1333 - val_loss: 0.1346 - val_mse: 0.1346 - lr: 0.0010\n", "Epoch 5/100\n", "136/136 [==============================] - 3s 21ms/step - loss: 0.1333 - mse: 0.1333 - val_loss: 0.1348 - val_mse: 0.1348 - lr: 0.0010\n", "Epoch 6/100\n", "136/136 [==============================] - 3s 25ms/step - loss: 0.1334 - mse: 0.1334 - val_loss: 0.1351 - val_mse: 0.1351 - lr: 0.0010\n", "Epoch 7/100\n", "136/136 [==============================] - 4s 26ms/step - loss: 0.1333 - mse: 0.1333 - val_loss: 0.1346 - val_mse: 0.1346 - lr: 0.0010\n", "Epoch 8/100\n", "134/136 [============================>.] - ETA: 0s - loss: 0.1332 - mse: 0.1332\n", "Epoch 8: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1332 - mse: 0.1332 - val_loss: 0.1346 - val_mse: 0.1346 - lr: 0.0010\n", "Epoch 9/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1330 - mse: 0.1330 - val_loss: 0.1346 - val_mse: 0.1346 - lr: 1.0000e-04\n", "Epoch 10/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1330 - mse: 0.1330 - val_loss: 0.1346 - val_mse: 0.1346 - lr: 1.0000e-04\n", "Epoch 11/100\n", "136/136 [==============================] - 4s 30ms/step - loss: 0.1329 - mse: 0.1329 - val_loss: 0.1345 - val_mse: 0.1345 - lr: 1.0000e-04\n", "Epoch 12/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1329 - mse: 0.1329 - val_loss: 0.1345 - val_mse: 0.1345 - lr: 1.0000e-04\n", "Epoch 13/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1329 - mse: 0.1329 - val_loss: 0.1344 - val_mse: 0.1344 - lr: 1.0000e-04\n", "Epoch 14/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1328 - mse: 0.1328 - val_loss: 0.1344 - val_mse: 0.1344 - lr: 1.0000e-04\n", "Epoch 15/100\n", "136/136 [==============================] - 4s 29ms/step - loss: 0.1328 - mse: 0.1328 - val_loss: 0.1345 - val_mse: 0.1345 - lr: 1.0000e-04\n", "Epoch 16/100\n", "136/136 [==============================] - 3s 21ms/step - loss: 0.1328 - mse: 0.1328 - val_loss: 0.1344 - val_mse: 0.1344 - lr: 1.0000e-04\n", "Epoch 17/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1328 - mse: 0.1328 - val_loss: 0.1344 - val_mse: 0.1344 - lr: 1.0000e-04\n", "Epoch 18/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1327 - mse: 0.1327 - val_loss: 0.1343 - val_mse: 0.1343 - lr: 1.0000e-04\n", "Epoch 19/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1327 - mse: 0.1327 - val_loss: 0.1343 - val_mse: 0.1343 - lr: 1.0000e-04\n", "Epoch 20/100\n", "136/136 [==============================] - 4s 32ms/step - loss: 0.1326 - mse: 0.1326 - val_loss: 0.1342 - val_mse: 0.1342 - lr: 1.0000e-04\n", "Epoch 21/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1326 - mse: 0.1326 - val_loss: 0.1341 - val_mse: 0.1341 - lr: 1.0000e-04\n", "Epoch 22/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1326 - mse: 0.1326 - val_loss: 0.1340 - val_mse: 0.1340 - lr: 1.0000e-04\n", "Epoch 23/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1324 - mse: 0.1324 - val_loss: 0.1339 - val_mse: 0.1339 - lr: 1.0000e-04\n", "Epoch 24/100\n", "136/136 [==============================] - 4s 32ms/step - loss: 0.1323 - mse: 0.1323 - val_loss: 0.1340 - val_mse: 0.1340 - lr: 1.0000e-04\n", "Epoch 25/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1323 - mse: 0.1323 - val_loss: 0.1336 - val_mse: 0.1336 - lr: 1.0000e-04\n", "Epoch 26/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1322 - mse: 0.1322 - val_loss: 0.1338 - val_mse: 0.1338 - lr: 1.0000e-04\n", "Epoch 27/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1321 - mse: 0.1321 - val_loss: 0.1332 - val_mse: 0.1332 - lr: 1.0000e-04\n", "Epoch 28/100\n", "136/136 [==============================] - 4s 26ms/step - loss: 0.1319 - mse: 0.1319 - val_loss: 0.1336 - val_mse: 0.1336 - lr: 1.0000e-04\n", "Epoch 29/100\n", "136/136 [==============================] - 3s 25ms/step - loss: 0.1318 - mse: 0.1318 - val_loss: 0.1338 - val_mse: 0.1338 - lr: 1.0000e-04\n", "Epoch 30/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1316 - mse: 0.1316 - val_loss: 0.1328 - val_mse: 0.1328 - lr: 1.0000e-04\n", "Epoch 31/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1313 - mse: 0.1313 - val_loss: 0.1322 - val_mse: 0.1322 - lr: 1.0000e-04\n", "Epoch 32/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1309 - mse: 0.1309 - val_loss: 0.1317 - val_mse: 0.1317 - lr: 1.0000e-04\n", "Epoch 33/100\n", "136/136 [==============================] - 4s 32ms/step - loss: 0.1305 - mse: 0.1305 - val_loss: 0.1316 - val_mse: 0.1316 - lr: 1.0000e-04\n", "Epoch 34/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1297 - mse: 0.1297 - val_loss: 0.1301 - val_mse: 0.1301 - lr: 1.0000e-04\n", "Epoch 35/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1288 - mse: 0.1288 - val_loss: 0.1285 - val_mse: 0.1285 - lr: 1.0000e-04\n", "Epoch 36/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1274 - mse: 0.1274 - val_loss: 0.1295 - val_mse: 0.1295 - lr: 1.0000e-04\n", "Epoch 37/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1260 - mse: 0.1260 - val_loss: 0.1249 - val_mse: 0.1249 - lr: 1.0000e-04\n", "Epoch 38/100\n", "136/136 [==============================] - 4s 27ms/step - loss: 0.1248 - mse: 0.1248 - val_loss: 0.1243 - val_mse: 0.1243 - lr: 1.0000e-04\n", "Epoch 39/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1233 - mse: 0.1233 - val_loss: 0.1236 - val_mse: 0.1236 - lr: 1.0000e-04\n", "Epoch 40/100\n", "136/136 [==============================] - 3s 24ms/step - loss: 0.1228 - mse: 0.1228 - val_loss: 0.1218 - val_mse: 0.1218 - lr: 1.0000e-04\n", "Epoch 41/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1213 - mse: 0.1213 - val_loss: 0.1207 - val_mse: 0.1207 - lr: 1.0000e-04\n", "Epoch 42/100\n", "136/136 [==============================] - 4s 28ms/step - loss: 0.1205 - mse: 0.1205 - val_loss: 0.1208 - val_mse: 0.1208 - lr: 1.0000e-04\n", "Epoch 43/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1196 - mse: 0.1196 - val_loss: 0.1214 - val_mse: 0.1214 - lr: 1.0000e-04\n", "Epoch 44/100\n", "136/136 [==============================] - 3s 21ms/step - loss: 0.1196 - mse: 0.1196 - val_loss: 0.1211 - val_mse: 0.1211 - lr: 1.0000e-04\n", "Epoch 45/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1181 - mse: 0.1181 - val_loss: 0.1176 - val_mse: 0.1176 - lr: 1.0000e-04\n", "Epoch 46/100\n", "136/136 [==============================] - 4s 27ms/step - loss: 0.1176 - mse: 0.1176 - val_loss: 0.1173 - val_mse: 0.1173 - lr: 1.0000e-04\n", "Epoch 47/100\n", "136/136 [==============================] - 3s 24ms/step - loss: 0.1173 - mse: 0.1173 - val_loss: 0.1166 - val_mse: 0.1166 - lr: 1.0000e-04\n", "Epoch 48/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1173 - mse: 0.1173 - val_loss: 0.1181 - val_mse: 0.1181 - lr: 1.0000e-04\n", "Epoch 49/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1159 - mse: 0.1159 - val_loss: 0.1182 - val_mse: 0.1182 - lr: 1.0000e-04\n", "Epoch 50/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1157 - mse: 0.1157 - val_loss: 0.1172 - val_mse: 0.1172 - lr: 1.0000e-04\n", "Epoch 51/100\n", "136/136 [==============================] - 4s 32ms/step - loss: 0.1153 - mse: 0.1153 - val_loss: 0.1152 - val_mse: 0.1152 - lr: 1.0000e-04\n", "Epoch 52/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1150 - mse: 0.1150 - val_loss: 0.1158 - val_mse: 0.1158 - lr: 1.0000e-04\n", "Epoch 53/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1149 - mse: 0.1149 - val_loss: 0.1149 - val_mse: 0.1149 - lr: 1.0000e-04\n", "Epoch 54/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1141 - mse: 0.1141 - val_loss: 0.1159 - val_mse: 0.1159 - lr: 1.0000e-04\n", "Epoch 55/100\n", "136/136 [==============================] - 4s 28ms/step - loss: 0.1138 - mse: 0.1138 - val_loss: 0.1139 - val_mse: 0.1139 - lr: 1.0000e-04\n", "Epoch 56/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1132 - mse: 0.1132 - val_loss: 0.1135 - val_mse: 0.1135 - lr: 1.0000e-04\n", "Epoch 57/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1128 - mse: 0.1128 - val_loss: 0.1124 - val_mse: 0.1124 - lr: 1.0000e-04\n", "Epoch 58/100\n", "136/136 [==============================] - 3s 21ms/step - loss: 0.1126 - mse: 0.1126 - val_loss: 0.1132 - val_mse: 0.1132 - lr: 1.0000e-04\n", "Epoch 59/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1123 - mse: 0.1123 - val_loss: 0.1127 - val_mse: 0.1127 - lr: 1.0000e-04\n", "Epoch 60/100\n", "136/136 [==============================] - 4s 29ms/step - loss: 0.1121 - mse: 0.1121 - val_loss: 0.1123 - val_mse: 0.1123 - lr: 1.0000e-04\n", "Epoch 61/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1121 - mse: 0.1121 - val_loss: 0.1128 - val_mse: 0.1128 - lr: 1.0000e-04\n", "Epoch 62/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1114 - mse: 0.1114 - val_loss: 0.1118 - val_mse: 0.1118 - lr: 1.0000e-04\n", "Epoch 63/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1110 - mse: 0.1110 - val_loss: 0.1128 - val_mse: 0.1128 - lr: 1.0000e-04\n", "Epoch 64/100\n", "136/136 [==============================] - 4s 29ms/step - loss: 0.1109 - mse: 0.1109 - val_loss: 0.1109 - val_mse: 0.1109 - lr: 1.0000e-04\n", "Epoch 65/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1116 - mse: 0.1116 - val_loss: 0.1118 - val_mse: 0.1118 - lr: 1.0000e-04\n", "Epoch 66/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1104 - mse: 0.1104 - val_loss: 0.1145 - val_mse: 0.1145 - lr: 1.0000e-04\n", "Epoch 67/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1106 - mse: 0.1106 - val_loss: 0.1112 - val_mse: 0.1112 - lr: 1.0000e-04\n", "Epoch 68/100\n", "136/136 [==============================] - 4s 31ms/step - loss: 0.1100 - mse: 0.1100 - val_loss: 0.1107 - val_mse: 0.1107 - lr: 1.0000e-04\n", "Epoch 69/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1099 - mse: 0.1099 - val_loss: 0.1102 - val_mse: 0.1102 - lr: 1.0000e-04\n", "Epoch 70/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1101 - mse: 0.1101 - val_loss: 0.1111 - val_mse: 0.1111 - lr: 1.0000e-04\n", "Epoch 71/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1099 - mse: 0.1099 - val_loss: 0.1120 - val_mse: 0.1120 - lr: 1.0000e-04\n", "Epoch 72/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1094 - mse: 0.1094 - val_loss: 0.1107 - val_mse: 0.1107 - lr: 1.0000e-04\n", "Epoch 73/100\n", "136/136 [==============================] - 4s 27ms/step - loss: 0.1091 - mse: 0.1091 - val_loss: 0.1102 - val_mse: 0.1102 - lr: 1.0000e-04\n", "Epoch 74/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1090 - mse: 0.1090 - val_loss: 0.1101 - val_mse: 0.1101 - lr: 1.0000e-04\n", "Epoch 75/100\n", "136/136 [==============================] - 3s 21ms/step - loss: 0.1093 - mse: 0.1093 - val_loss: 0.1100 - val_mse: 0.1100 - lr: 1.0000e-04\n", "Epoch 76/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1085 - mse: 0.1085 - val_loss: 0.1097 - val_mse: 0.1097 - lr: 1.0000e-04\n", "Epoch 77/100\n", "136/136 [==============================] - 4s 31ms/step - loss: 0.1082 - mse: 0.1082 - val_loss: 0.1097 - val_mse: 0.1097 - lr: 1.0000e-04\n", "Epoch 78/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1085 - mse: 0.1085 - val_loss: 0.1102 - val_mse: 0.1102 - lr: 1.0000e-04\n", "Epoch 79/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1081 - mse: 0.1081 - val_loss: 0.1107 - val_mse: 0.1107 - lr: 1.0000e-04\n", "Epoch 80/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1082 - mse: 0.1082 - val_loss: 0.1098 - val_mse: 0.1098 - lr: 1.0000e-04\n", "Epoch 81/100\n", "136/136 [==============================] - 3s 25ms/step - loss: 0.1077 - mse: 0.1077 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-04\n", "Epoch 82/100\n", "136/136 [==============================] - 4s 28ms/step - loss: 0.1076 - mse: 0.1076 - val_loss: 0.1118 - val_mse: 0.1118 - lr: 1.0000e-04\n", "Epoch 83/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1081 - mse: 0.1081 - val_loss: 0.1105 - val_mse: 0.1105 - lr: 1.0000e-04\n", "Epoch 84/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1072 - mse: 0.1072 - val_loss: 0.1086 - val_mse: 0.1086 - lr: 1.0000e-04\n", "Epoch 85/100\n", "136/136 [==============================] - 3s 20ms/step - loss: 0.1076 - mse: 0.1076 - val_loss: 0.1098 - val_mse: 0.1098 - lr: 1.0000e-04\n", "Epoch 86/100\n", "136/136 [==============================] - 4s 32ms/step - loss: 0.1070 - mse: 0.1070 - val_loss: 0.1104 - val_mse: 0.1104 - lr: 1.0000e-04\n", "Epoch 87/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1076 - mse: 0.1076 - val_loss: 0.1102 - val_mse: 0.1102 - lr: 1.0000e-04\n", "Epoch 88/100\n", "135/136 [============================>.] - ETA: 0s - loss: 0.1067 - mse: 0.1067\n", "Epoch 88: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "136/136 [==============================] - 3s 24ms/step - loss: 0.1069 - mse: 0.1069 - val_loss: 0.1108 - val_mse: 0.1108 - lr: 1.0000e-04\n", "Epoch 89/100\n", "136/136 [==============================] - 3s 21ms/step - loss: 0.1064 - mse: 0.1064 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-05\n", "Epoch 90/100\n", "136/136 [==============================] - 4s 28ms/step - loss: 0.1061 - mse: 0.1061 - val_loss: 0.1081 - val_mse: 0.1081 - lr: 1.0000e-05\n", "Epoch 91/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1061 - mse: 0.1061 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-05\n", "Epoch 92/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1061 - mse: 0.1061 - val_loss: 0.1083 - val_mse: 0.1083 - lr: 1.0000e-05\n", "Epoch 93/100\n", "136/136 [==============================] - 3s 19ms/step - loss: 0.1061 - mse: 0.1061 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-05\n", "Epoch 94/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1082 - val_mse: 0.1082 - lr: 1.0000e-05\n", "Epoch 95/100\n", "136/136 [==============================] - 4s 29ms/step - loss: 0.1061 - mse: 0.1061 - val_loss: 0.1083 - val_mse: 0.1083 - lr: 1.0000e-05\n", "Epoch 96/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1083 - val_mse: 0.1083 - lr: 1.0000e-05\n", "Epoch 97/100\n", "136/136 [==============================] - ETA: 0s - loss: 0.1060 - mse: 0.1060\n", "Epoch 97: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-05\n", "Epoch 98/100\n", "136/136 [==============================] - 3s 22ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1080 - val_mse: 0.1080 - lr: 1.0000e-05\n", "Epoch 99/100\n", "136/136 [==============================] - 4s 33ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-05\n", "Epoch 100/100\n", "136/136 [==============================] - 3s 23ms/step - loss: 0.1060 - mse: 0.1060 - val_loss: 0.1084 - val_mse: 0.1084 - lr: 1.0000e-05\n", "408/408 [==============================] - 3s 5ms/step\n", "Epoch 1/100\n", "272/272 [==============================] - 14s 27ms/step - loss: 0.1359 - mse: 0.1359 - val_loss: 0.1254 - val_mse: 0.1254 - lr: 0.0010\n", "Epoch 2/100\n", "272/272 [==============================] - 7s 24ms/step - loss: 0.1356 - mse: 0.1356 - val_loss: 0.1255 - val_mse: 0.1255 - lr: 0.0010\n", "Epoch 3/100\n", "272/272 [==============================] - 5s 20ms/step - loss: 0.1357 - mse: 0.1357 - val_loss: 0.1250 - val_mse: 0.1250 - lr: 0.0010\n", "Epoch 4/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1355 - mse: 0.1355 - val_loss: 0.1256 - val_mse: 0.1256 - lr: 0.0010\n", "Epoch 5/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1356 - mse: 0.1356 - val_loss: 0.1249 - val_mse: 0.1249 - lr: 0.0010\n", "Epoch 6/100\n", "272/272 [==============================] - 8s 29ms/step - loss: 0.1351 - mse: 0.1351 - val_loss: 0.1248 - val_mse: 0.1248 - lr: 0.0010\n", "Epoch 7/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1307 - mse: 0.1307 - val_loss: 0.1214 - val_mse: 0.1214 - lr: 0.0010\n", "Epoch 8/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1191 - mse: 0.1191 - val_loss: 0.1112 - val_mse: 0.1112 - lr: 0.0010\n", "Epoch 9/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1135 - mse: 0.1135 - val_loss: 0.1044 - val_mse: 0.1044 - lr: 0.0010\n", "Epoch 10/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1104 - mse: 0.1104 - val_loss: 0.1048 - val_mse: 0.1048 - lr: 0.0010\n", "Epoch 11/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1086 - mse: 0.1086 - val_loss: 0.1022 - val_mse: 0.1022 - lr: 0.0010\n", "Epoch 12/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1079 - mse: 0.1079 - val_loss: 0.1009 - val_mse: 0.1009 - lr: 0.0010\n", "Epoch 13/100\n", "272/272 [==============================] - 5s 20ms/step - loss: 0.1071 - mse: 0.1071 - val_loss: 0.1005 - val_mse: 0.1005 - lr: 0.0010\n", "Epoch 14/100\n", "272/272 [==============================] - 6s 24ms/step - loss: 0.1064 - mse: 0.1064 - val_loss: 0.0996 - val_mse: 0.0996 - lr: 0.0010\n", "Epoch 15/100\n", "272/272 [==============================] - 6s 24ms/step - loss: 0.1062 - mse: 0.1062 - val_loss: 0.1013 - val_mse: 0.1013 - lr: 0.0010\n", "Epoch 16/100\n", "272/272 [==============================] - 6s 24ms/step - loss: 0.1058 - mse: 0.1058 - val_loss: 0.1005 - val_mse: 0.1005 - lr: 0.0010\n", "Epoch 17/100\n", "272/272 [==============================] - 7s 24ms/step - loss: 0.1058 - mse: 0.1058 - val_loss: 0.0992 - val_mse: 0.0992 - lr: 0.0010\n", "Epoch 18/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1055 - mse: 0.1055 - val_loss: 0.0991 - val_mse: 0.0991 - lr: 0.0010\n", "Epoch 19/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1058 - mse: 0.1058 - val_loss: 0.0991 - val_mse: 0.0991 - lr: 0.0010\n", "Epoch 20/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1053 - mse: 0.1053 - val_loss: 0.0999 - val_mse: 0.0999 - lr: 0.0010\n", "Epoch 21/100\n", "272/272 [==============================] - 8s 28ms/step - loss: 0.1051 - mse: 0.1051 - val_loss: 0.1008 - val_mse: 0.1008 - lr: 0.0010\n", "Epoch 22/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1046 - mse: 0.1046 - val_loss: 0.0998 - val_mse: 0.0998 - lr: 0.0010\n", "Epoch 23/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1047 - mse: 0.1047 - val_loss: 0.0993 - val_mse: 0.0993 - lr: 0.0010\n", "Epoch 24/100\n", "272/272 [==============================] - 5s 20ms/step - loss: 0.1051 - mse: 0.1051 - val_loss: 0.0983 - val_mse: 0.0983 - lr: 0.0010\n", "Epoch 25/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1045 - mse: 0.1045 - val_loss: 0.0989 - val_mse: 0.0989 - lr: 0.0010\n", "Epoch 26/100\n", "272/272 [==============================] - 5s 20ms/step - loss: 0.1047 - mse: 0.1047 - val_loss: 0.1016 - val_mse: 0.1016 - lr: 0.0010\n", "Epoch 27/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.0983 - val_mse: 0.0983 - lr: 0.0010\n", "Epoch 28/100\n", "272/272 [==============================] - 6s 21ms/step - loss: 0.1043 - mse: 0.1043 - val_loss: 0.0997 - val_mse: 0.0997 - lr: 0.0010\n", "Epoch 29/100\n", "272/272 [==============================] - 7s 25ms/step - loss: 0.1044 - mse: 0.1044 - val_loss: 0.1022 - val_mse: 0.1022 - lr: 0.0010\n", "Epoch 30/100\n", "272/272 [==============================] - 6s 21ms/step - loss: 0.1046 - mse: 0.1046 - val_loss: 0.1003 - val_mse: 0.1003 - lr: 0.0010\n", "Epoch 31/100\n", "272/272 [==============================] - 5s 20ms/step - loss: 0.1044 - mse: 0.1044 - val_loss: 0.0982 - val_mse: 0.0982 - lr: 0.0010\n", "Epoch 32/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.1021 - val_mse: 0.1021 - lr: 0.0010\n", "Epoch 33/100\n", "272/272 [==============================] - 6s 24ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1028 - val_mse: 0.1028 - lr: 0.0010\n", "Epoch 34/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1038 - mse: 0.1038 - val_loss: 0.0995 - val_mse: 0.0995 - lr: 0.0010\n", "Epoch 35/100\n", "272/272 [==============================] - 7s 24ms/step - loss: 0.1039 - mse: 0.1039 - val_loss: 0.1006 - val_mse: 0.1006 - lr: 0.0010\n", "Epoch 36/100\n", "272/272 [==============================] - 8s 31ms/step - loss: 0.1042 - mse: 0.1042 - val_loss: 0.0995 - val_mse: 0.0995 - lr: 0.0010\n", "Epoch 37/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1035 - mse: 0.1035 - val_loss: 0.0997 - val_mse: 0.0997 - lr: 0.0010\n", "Epoch 38/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1036 - mse: 0.1036 - val_loss: 0.0976 - val_mse: 0.0976 - lr: 0.0010\n", "Epoch 39/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1036 - mse: 0.1036 - val_loss: 0.1003 - val_mse: 0.1003 - lr: 0.0010\n", "Epoch 40/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1033 - mse: 0.1033 - val_loss: 0.0981 - val_mse: 0.0981 - lr: 0.0010\n", "Epoch 41/100\n", "272/272 [==============================] - 6s 21ms/step - loss: 0.1037 - mse: 0.1037 - val_loss: 0.1007 - val_mse: 0.1007 - lr: 0.0010\n", "Epoch 42/100\n", "272/272 [==============================] - 7s 27ms/step - loss: 0.1035 - mse: 0.1035 - val_loss: 0.0992 - val_mse: 0.0992 - lr: 0.0010\n", "Epoch 43/100\n", "272/272 [==============================] - 5s 20ms/step - loss: 0.1032 - mse: 0.1032 - val_loss: 0.1007 - val_mse: 0.1007 - lr: 0.0010\n", "Epoch 44/100\n", "272/272 [==============================] - 7s 24ms/step - loss: 0.1035 - mse: 0.1035 - val_loss: 0.0986 - val_mse: 0.0986 - lr: 0.0010\n", "Epoch 45/100\n", "270/272 [============================>.] - ETA: 0s - loss: 0.1030 - mse: 0.1030\n", "Epoch 45: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "272/272 [==============================] - 7s 25ms/step - loss: 0.1029 - mse: 0.1029 - val_loss: 0.0982 - val_mse: 0.0982 - lr: 0.0010\n", "Epoch 46/100\n", "272/272 [==============================] - 7s 25ms/step - loss: 0.1015 - mse: 0.1015 - val_loss: 0.0969 - val_mse: 0.0969 - lr: 1.0000e-04\n", "Epoch 47/100\n", "272/272 [==============================] - 7s 26ms/step - loss: 0.1014 - mse: 0.1014 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-04\n", "Epoch 48/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1013 - mse: 0.1013 - val_loss: 0.0974 - val_mse: 0.0974 - lr: 1.0000e-04\n", "Epoch 49/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1013 - mse: 0.1013 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-04\n", "Epoch 50/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1013 - mse: 0.1013 - val_loss: 0.0975 - val_mse: 0.0975 - lr: 1.0000e-04\n", "Epoch 51/100\n", "272/272 [==============================] - 8s 28ms/step - loss: 0.1012 - mse: 0.1012 - val_loss: 0.0974 - val_mse: 0.0974 - lr: 1.0000e-04\n", "Epoch 52/100\n", "272/272 [==============================] - 6s 22ms/step - loss: 0.1013 - mse: 0.1013 - val_loss: 0.0976 - val_mse: 0.0976 - lr: 1.0000e-04\n", "Epoch 53/100\n", "270/272 [============================>.] - ETA: 0s - loss: 0.1013 - mse: 0.1013\n", "Epoch 53: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "272/272 [==============================] - 8s 28ms/step - loss: 0.1012 - mse: 0.1012 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-04\n", "Epoch 54/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1010 - mse: 0.1010 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 55/100\n", "272/272 [==============================] - 8s 29ms/step - loss: 0.1010 - mse: 0.1010 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 56/100\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1010 - mse: 0.1010 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 57/100\n", "272/272 [==============================] - 10s 36ms/step - loss: 0.1010 - mse: 0.1010 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 58/100\n", "272/272 [==============================] - ETA: 0s - loss: 0.1010 - mse: 0.1010Restoring model weights from the end of the best epoch: 46.\n", "272/272 [==============================] - 6s 23ms/step - loss: 0.1010 - mse: 0.1010 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 58: early stopping\n", "408/408 [==============================] - 3s 5ms/step\n", "Epoch 1/100\n", "408/408 [==============================] - 17s 28ms/step - loss: 0.1324 - mse: 0.1324 - val_loss: 0.1324 - val_mse: 0.1324 - lr: 0.0010\n", "Epoch 2/100\n", "408/408 [==============================] - 8s 20ms/step - loss: 0.1322 - mse: 0.1322 - val_loss: 0.1324 - val_mse: 0.1324 - lr: 0.0010\n", "Epoch 3/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1321 - mse: 0.1321 - val_loss: 0.1328 - val_mse: 0.1328 - lr: 0.0010\n", "Epoch 4/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1321 - mse: 0.1321 - val_loss: 0.1324 - val_mse: 0.1324 - lr: 0.0010\n", "Epoch 5/100\n", "408/408 [==============================] - 8s 21ms/step - loss: 0.1319 - mse: 0.1319 - val_loss: 0.1317 - val_mse: 0.1317 - lr: 0.0010\n", "Epoch 6/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1287 - mse: 0.1287 - val_loss: 0.1201 - val_mse: 0.1201 - lr: 0.0010\n", "Epoch 7/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1117 - mse: 0.1117 - val_loss: 0.1127 - val_mse: 0.1127 - lr: 0.0010\n", "Epoch 8/100\n", "408/408 [==============================] - 8s 19ms/step - loss: 0.1073 - mse: 0.1073 - val_loss: 0.1109 - val_mse: 0.1109 - lr: 0.0010\n", "Epoch 9/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1054 - mse: 0.1054 - val_loss: 0.1064 - val_mse: 0.1064 - lr: 0.0010\n", "Epoch 10/100\n", "408/408 [==============================] - 10s 26ms/step - loss: 0.1048 - mse: 0.1048 - val_loss: 0.1070 - val_mse: 0.1070 - lr: 0.0010\n", "Epoch 11/100\n", "408/408 [==============================] - 9s 21ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1065 - val_mse: 0.1065 - lr: 0.0010\n", "Epoch 12/100\n", "408/408 [==============================] - 11s 26ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1064 - val_mse: 0.1064 - lr: 0.0010\n", "Epoch 13/100\n", "408/408 [==============================] - 11s 26ms/step - loss: 0.1034 - mse: 0.1034 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 14/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1033 - mse: 0.1033 - val_loss: 0.1067 - val_mse: 0.1067 - lr: 0.0010\n", "Epoch 15/100\n", "408/408 [==============================] - 9s 22ms/step - loss: 0.1031 - mse: 0.1031 - val_loss: 0.1057 - val_mse: 0.1057 - lr: 0.0010\n", "Epoch 16/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.1055 - val_mse: 0.1055 - lr: 0.0010\n", "Epoch 17/100\n", "408/408 [==============================] - 11s 27ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.1062 - val_mse: 0.1062 - lr: 0.0010\n", "Epoch 18/100\n", "408/408 [==============================] - 9s 22ms/step - loss: 0.1024 - mse: 0.1024 - val_loss: 0.1064 - val_mse: 0.1064 - lr: 0.0010\n", "Epoch 19/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1028 - mse: 0.1028 - val_loss: 0.1072 - val_mse: 0.1072 - lr: 0.0010\n", "Epoch 20/100\n", "408/408 [==============================] - 10s 26ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.1064 - val_mse: 0.1064 - lr: 0.0010\n", "Epoch 21/100\n", "408/408 [==============================] - 9s 22ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 22/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.1075 - val_mse: 0.1075 - lr: 0.0010\n", "Epoch 23/100\n", "406/408 [============================>.] - ETA: 0s - loss: 0.1025 - mse: 0.1025\n", "Epoch 23: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1024 - mse: 0.1024 - val_loss: 0.1078 - val_mse: 0.1078 - lr: 0.0010\n", "Epoch 24/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.1056 - val_mse: 0.1056 - lr: 1.0000e-04\n", "Epoch 25/100\n", "408/408 [==============================] - 9s 22ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1049 - val_mse: 0.1049 - lr: 1.0000e-04\n", "Epoch 26/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1051 - val_mse: 0.1051 - lr: 1.0000e-04\n", "Epoch 27/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1049 - val_mse: 0.1049 - lr: 1.0000e-04\n", "Epoch 28/100\n", "408/408 [==============================] - 8s 21ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-04\n", "Epoch 29/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1051 - val_mse: 0.1051 - lr: 1.0000e-04\n", "Epoch 30/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1051 - val_mse: 0.1051 - lr: 1.0000e-04\n", "Epoch 31/100\n", "408/408 [==============================] - 9s 21ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1053 - val_mse: 0.1053 - lr: 1.0000e-04\n", "Epoch 32/100\n", "408/408 [==============================] - ETA: 0s - loss: 0.1001 - mse: 0.1001\n", "Epoch 32: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.1049 - val_mse: 0.1049 - lr: 1.0000e-04\n", "Epoch 33/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.0999 - mse: 0.0999 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-05\n", "Epoch 34/100\n", "408/408 [==============================] - 9s 22ms/step - loss: 0.0999 - mse: 0.0999 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-05\n", "Epoch 35/100\n", "408/408 [==============================] - 10s 25ms/step - loss: 0.0998 - mse: 0.0998 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-05\n", "Epoch 36/100\n", "408/408 [==============================] - 10s 24ms/step - loss: 0.0998 - mse: 0.0998 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-05\n", "Epoch 37/100\n", "407/408 [============================>.] - ETA: 0s - loss: 0.0999 - mse: 0.0999Restoring model weights from the end of the best epoch: 25.\n", "408/408 [==============================] - 9s 21ms/step - loss: 0.0998 - mse: 0.0998 - val_loss: 0.1050 - val_mse: 0.1050 - lr: 1.0000e-05\n", "Epoch 37: early stopping\n", "408/408 [==============================] - 5s 7ms/step\n", "Epoch 1/100\n", "544/544 [==============================] - 21s 28ms/step - loss: 0.1295 - mse: 0.1295 - val_loss: 0.1293 - val_mse: 0.1293 - lr: 0.0010\n", "Epoch 2/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.1294 - mse: 0.1294 - val_loss: 0.1294 - val_mse: 0.1294 - lr: 0.0010\n", "Epoch 3/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.1292 - mse: 0.1292 - val_loss: 0.1289 - val_mse: 0.1289 - lr: 0.0010\n", "Epoch 4/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.1228 - mse: 0.1228 - val_loss: 0.1137 - val_mse: 0.1137 - lr: 0.0010\n", "Epoch 5/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.1087 - mse: 0.1087 - val_loss: 0.1141 - val_mse: 0.1141 - lr: 0.0010\n", "Epoch 6/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.1053 - mse: 0.1053 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 7/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 8/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.1028 - mse: 0.1028 - val_loss: 0.1063 - val_mse: 0.1063 - lr: 0.0010\n", "Epoch 9/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.1023 - mse: 0.1023 - val_loss: 0.1059 - val_mse: 0.1059 - lr: 0.0010\n", "Epoch 10/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.1030 - val_mse: 0.1030 - lr: 0.0010\n", "Epoch 11/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.1016 - mse: 0.1016 - val_loss: 0.1039 - val_mse: 0.1039 - lr: 0.0010\n", "Epoch 12/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.1015 - mse: 0.1015 - val_loss: 0.1070 - val_mse: 0.1070 - lr: 0.0010\n", "Epoch 13/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1012 - mse: 0.1012 - val_loss: 0.1036 - val_mse: 0.1036 - lr: 0.0010\n", "Epoch 14/100\n", "544/544 [==============================] - 12s 22ms/step - loss: 0.1012 - mse: 0.1012 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 0.0010\n", "Epoch 15/100\n", "544/544 [==============================] - 12s 23ms/step - loss: 0.1011 - mse: 0.1011 - val_loss: 0.1029 - val_mse: 0.1029 - lr: 0.0010\n", "Epoch 16/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.1013 - mse: 0.1013 - val_loss: 0.1028 - val_mse: 0.1028 - lr: 0.0010\n", "Epoch 17/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1010 - mse: 0.1010 - val_loss: 0.1023 - val_mse: 0.1023 - lr: 0.0010\n", "Epoch 18/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.1044 - val_mse: 0.1044 - lr: 0.0010\n", "Epoch 19/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.1030 - val_mse: 0.1030 - lr: 0.0010\n", "Epoch 20/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.1025 - val_mse: 0.1025 - lr: 0.0010\n", "Epoch 21/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1057 - val_mse: 0.1057 - lr: 0.0010\n", "Epoch 22/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.1005 - mse: 0.1005 - val_loss: 0.1039 - val_mse: 0.1039 - lr: 0.0010\n", "Epoch 23/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.1003 - mse: 0.1003 - val_loss: 0.1024 - val_mse: 0.1024 - lr: 0.0010\n", "Epoch 24/100\n", "543/544 [============================>.] - ETA: 0s - loss: 0.1002 - mse: 0.1002\n", "Epoch 24: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "544/544 [==============================] - 12s 22ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.1026 - val_mse: 0.1026 - lr: 0.0010\n", "Epoch 25/100\n", "544/544 [==============================] - 16s 30ms/step - loss: 0.0990 - mse: 0.0990 - val_loss: 0.1019 - val_mse: 0.1019 - lr: 1.0000e-04\n", "Epoch 26/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.0988 - mse: 0.0988 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-04\n", "Epoch 27/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.0987 - mse: 0.0987 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-04\n", "Epoch 28/100\n", "544/544 [==============================] - 13s 25ms/step - loss: 0.0987 - mse: 0.0987 - val_loss: 0.1018 - val_mse: 0.1018 - lr: 1.0000e-04\n", "Epoch 29/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.0987 - mse: 0.0987 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 1.0000e-04\n", "Epoch 30/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.0986 - mse: 0.0986 - val_loss: 0.1016 - val_mse: 0.1016 - lr: 1.0000e-04\n", "Epoch 31/100\n", "544/544 [==============================] - 13s 25ms/step - loss: 0.0986 - mse: 0.0986 - val_loss: 0.1016 - val_mse: 0.1016 - lr: 1.0000e-04\n", "Epoch 32/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.0986 - mse: 0.0986 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-04\n", "Epoch 33/100\n", "543/544 [============================>.] - ETA: 0s - loss: 0.0985 - mse: 0.0985\n", "Epoch 33: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0986 - mse: 0.0986 - val_loss: 0.1016 - val_mse: 0.1016 - lr: 1.0000e-04\n", "Epoch 34/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0984 - mse: 0.0984 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 35/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0984 - mse: 0.0984 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 36/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.0984 - mse: 0.0984 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 37/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.0984 - mse: 0.0984 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 38/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0984 - mse: 0.0984 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 39/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 40/100\n", "544/544 [==============================] - ETA: 0s - loss: 0.0983 - mse: 0.0983\n", "Epoch 40: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "544/544 [==============================] - 15s 27ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 41/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 42/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 43/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 44/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 45/100\n", "544/544 [==============================] - 12s 21ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 46/100\n", "544/544 [==============================] - 12s 23ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 47/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 48/100\n", "544/544 [==============================] - 14s 25ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 49/100\n", "544/544 [==============================] - 15s 27ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 50/100\n", "544/544 [==============================] - 13s 25ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-05\n", "Epoch 51/100\n", "544/544 [==============================] - 13s 25ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 52/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 53/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 54/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 55/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 56/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 57/100\n", "544/544 [==============================] - 12s 22ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 58/100\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 59/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 60/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 61/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 62/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 63/100\n", "544/544 [==============================] - 13s 25ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 64/100\n", "544/544 [==============================] - 14s 26ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 65/100\n", "544/544 [==============================] - 13s 24ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 66/100\n", "543/544 [============================>.] - ETA: 0s - loss: 0.0983 - mse: 0.0983Restoring model weights from the end of the best epoch: 54.\n", "544/544 [==============================] - 13s 23ms/step - loss: 0.0983 - mse: 0.0983 - val_loss: 0.1014 - val_mse: 0.1014 - lr: 1.0000e-05\n", "Epoch 66: early stopping\n", "408/408 [==============================] - 4s 7ms/step\n", "Epoch 1/100\n", "680/680 [==============================] - 22s 24ms/step - loss: 0.1307 - mse: 0.1307 - val_loss: 0.1176 - val_mse: 0.1176 - lr: 0.0010\n", "Epoch 2/100\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.1305 - mse: 0.1305 - val_loss: 0.1174 - val_mse: 0.1174 - lr: 0.0010\n", "Epoch 3/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1273 - mse: 0.1273 - val_loss: 0.1185 - val_mse: 0.1185 - lr: 0.0010\n", "Epoch 4/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1109 - mse: 0.1109 - val_loss: 0.1011 - val_mse: 0.1011 - lr: 0.0010\n", "Epoch 5/100\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.1067 - mse: 0.1067 - val_loss: 0.1017 - val_mse: 0.1017 - lr: 0.0010\n", "Epoch 6/100\n", "680/680 [==============================] - 15s 22ms/step - loss: 0.1050 - mse: 0.1050 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 0.0010\n", "Epoch 7/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1038 - mse: 0.1038 - val_loss: 0.0957 - val_mse: 0.0957 - lr: 0.0010\n", "Epoch 8/100\n", "680/680 [==============================] - 22s 33ms/step - loss: 0.1033 - mse: 0.1033 - val_loss: 0.0966 - val_mse: 0.0966 - lr: 0.0010\n", "Epoch 9/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.1031 - mse: 0.1031 - val_loss: 0.0961 - val_mse: 0.0961 - lr: 0.0010\n", "Epoch 10/100\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.1030 - mse: 0.1030 - val_loss: 0.0959 - val_mse: 0.0959 - lr: 0.0010\n", "Epoch 11/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1027 - mse: 0.1027 - val_loss: 0.0941 - val_mse: 0.0941 - lr: 0.0010\n", "Epoch 12/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.0939 - val_mse: 0.0939 - lr: 0.0010\n", "Epoch 13/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.1022 - mse: 0.1022 - val_loss: 0.0939 - val_mse: 0.0939 - lr: 0.0010\n", "Epoch 14/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 0.0010\n", "Epoch 15/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.0953 - val_mse: 0.0953 - lr: 0.0010\n", "Epoch 16/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1022 - mse: 0.1022 - val_loss: 0.0939 - val_mse: 0.0939 - lr: 0.0010\n", "Epoch 17/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.0954 - val_mse: 0.0954 - lr: 0.0010\n", "Epoch 18/100\n", "680/680 [==============================] - 20s 29ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.0955 - val_mse: 0.0955 - lr: 0.0010\n", "Epoch 19/100\n", "679/680 [============================>.] - ETA: 0s - loss: 0.1017 - mse: 0.1017\n", "Epoch 19: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.0941 - val_mse: 0.0941 - lr: 0.0010\n", "Epoch 20/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1004 - mse: 0.1004 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-04\n", "Epoch 21/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0930 - val_mse: 0.0930 - lr: 1.0000e-04\n", "Epoch 22/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1002 - mse: 0.1002 - val_loss: 0.0931 - val_mse: 0.0931 - lr: 1.0000e-04\n", "Epoch 23/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-04\n", "Epoch 24/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-04\n", "Epoch 25/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1001 - mse: 0.1001 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-04\n", "Epoch 26/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.1000 - mse: 0.1000 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-04\n", "Epoch 27/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.1000 - mse: 0.1000 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-04\n", "Epoch 28/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.0999 - mse: 0.0999 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-04\n", "Epoch 29/100\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.0999 - mse: 0.0999 - val_loss: 0.0929 - val_mse: 0.0929 - lr: 1.0000e-04\n", "Epoch 30/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.0999 - mse: 0.0999 - val_loss: 0.0927 - val_mse: 0.0927 - lr: 1.0000e-04\n", "Epoch 31/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.0998 - mse: 0.0998 - val_loss: 0.0928 - val_mse: 0.0928 - lr: 1.0000e-04\n", "Epoch 32/100\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.0998 - mse: 0.0998 - val_loss: 0.0924 - val_mse: 0.0924 - lr: 1.0000e-04\n", "Epoch 33/100\n", "680/680 [==============================] - 15s 23ms/step - loss: 0.0997 - mse: 0.0997 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-04\n", "Epoch 34/100\n", "678/680 [============================>.] - ETA: 0s - loss: 0.0998 - mse: 0.0998\n", "Epoch 34: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.0997 - mse: 0.0997 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-04\n", "Epoch 35/100\n", "680/680 [==============================] - 16s 23ms/step - loss: 0.0996 - mse: 0.0996 - val_loss: 0.0926 - val_mse: 0.0926 - lr: 1.0000e-05\n", "Epoch 36/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 37/100\n", "680/680 [==============================] - 16s 24ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 38/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 39/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 40/100\n", "680/680 [==============================] - 17s 24ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0924 - val_mse: 0.0924 - lr: 1.0000e-05\n", "Epoch 41/100\n", "679/680 [============================>.] - ETA: 0s - loss: 0.0995 - mse: 0.0995\n", "Epoch 41: ReduceLROnPlateau reducing learning rate to 1e-05.\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 42/100\n", "680/680 [==============================] - 17s 25ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0924 - val_mse: 0.0924 - lr: 1.0000e-05\n", "Epoch 43/100\n", "680/680 [==============================] - 17s 26ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 44/100\n", "678/680 [============================>.] - ETA: 0s - loss: 0.0995 - mse: 0.0995Restoring model weights from the end of the best epoch: 32.\n", "680/680 [==============================] - 18s 26ms/step - loss: 0.0995 - mse: 0.0995 - val_loss: 0.0925 - val_mse: 0.0925 - lr: 1.0000e-05\n", "Epoch 44: early stopping\n", "408/408 [==============================] - 4s 6ms/step\n" ] } ] }, { "cell_type": "markdown", "source": [ "#Training eporch for DNN" ], "metadata": { "id": "KXGshnDRMABM" } }, { "cell_type": "code", "source": [ "import copy\n", "x=[]\n", "y=[]\n", "for i in range(len(org_data)-10):\n", " base=org_data[i:i+11,4].tolist()\n", " gas_frc=org_data[i:i+10,5].tolist()\n", " now=np.concatenate((base, gas_frc),axis=0)\n", " now=now.tolist()\n", " x.append(now)\n", " y.append(org_data[i+10,6])\n", "x=np.array(x)\n", "y=np.array(y)" ], "metadata": { "id": "JAINS3mkMC_S" }, "execution_count": 21, "outputs": [] }, { "cell_type": "code", "source": [ "from sklearn.model_selection import TimeSeriesSplit\n", "from sklearn import linear_model\n", "from sklearn.metrics import mean_squared_error\n", "tss = TimeSeriesSplit(n_splits=2)\n", "for train_index, test_index in tss.split(x):\n", " x_train, x_test = x[train_index, :], x[test_index,:]\n", " y_train, y_test = y[train_index], y[test_index]" ], "metadata": { "id": "BHw223LQMJ5d" }, "execution_count": 22, "outputs": [] }, { "cell_type": "code", "source": [ "import tensorflow as tf\n", "import matplotlib.pyplot as plt\n", "\n", "model_stability = tf.keras.models.Sequential([\n", " tf.keras.layers.Dense(128, activation = 'tanh',use_bias=True),\n", " tf.keras.layers.Dense(128, activation = 'linear',use_bias=True),\n", " tf.keras.layers.Dense(128, activation = 'tanh',use_bias=True),\n", " tf.keras.layers.Dense(1)\n", " ]\n", " )\n", "training_callbacks = [\n", " tf.keras.callbacks.ReduceLROnPlateau(patience=7, factor=0.1, min_lr=0.000001, verbose=1),\n", " tf.keras.callbacks.EarlyStopping(patience=12, restore_best_weights=True, verbose=1),\n", "]\n", "model_stability.compile(loss=tf.keras.losses.mean_squared_error,\n", " optimizer=tf.keras.optimizers.Adam(learning_rate=0.0001),\n", " metrics=['mse'])\n", "\n", "\n", "history = model_stability.fit(x_train, y_train, batch_size=512, epochs=100, callbacks=training_callbacks,validation_data=(x_test,y_test))\n" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "acjIrARUMLMQ", "outputId": "61f613f4-18aa-4c81-9fdd-329919f1eb10" }, "execution_count": 23, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Epoch 1/100\n", "102/102 [==============================] - 2s 9ms/step - loss: 0.1359 - mse: 0.1359 - val_loss: 0.1187 - val_mse: 0.1187 - lr: 1.0000e-04\n", "Epoch 2/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1247 - mse: 0.1247 - val_loss: 0.1164 - val_mse: 0.1164 - lr: 1.0000e-04\n", "Epoch 3/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1215 - mse: 0.1215 - val_loss: 0.1135 - val_mse: 0.1135 - lr: 1.0000e-04\n", "Epoch 4/100\n", "102/102 [==============================] - 1s 13ms/step - loss: 0.1194 - mse: 0.1194 - val_loss: 0.1131 - val_mse: 0.1131 - lr: 1.0000e-04\n", "Epoch 5/100\n", "102/102 [==============================] - 1s 13ms/step - loss: 0.1181 - mse: 0.1181 - val_loss: 0.1116 - val_mse: 0.1116 - lr: 1.0000e-04\n", "Epoch 6/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1173 - mse: 0.1173 - val_loss: 0.1102 - val_mse: 0.1102 - lr: 1.0000e-04\n", "Epoch 7/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1164 - mse: 0.1164 - val_loss: 0.1097 - val_mse: 0.1097 - lr: 1.0000e-04\n", "Epoch 8/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1157 - mse: 0.1157 - val_loss: 0.1084 - val_mse: 0.1084 - lr: 1.0000e-04\n", "Epoch 9/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1152 - mse: 0.1152 - val_loss: 0.1094 - val_mse: 0.1094 - lr: 1.0000e-04\n", "Epoch 10/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1149 - mse: 0.1149 - val_loss: 0.1073 - val_mse: 0.1073 - lr: 1.0000e-04\n", "Epoch 11/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1130 - mse: 0.1130 - val_loss: 0.1085 - val_mse: 0.1085 - lr: 1.0000e-04\n", "Epoch 12/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1123 - mse: 0.1123 - val_loss: 0.1059 - val_mse: 0.1059 - lr: 1.0000e-04\n", "Epoch 13/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1111 - mse: 0.1111 - val_loss: 0.1047 - val_mse: 0.1047 - lr: 1.0000e-04\n", "Epoch 14/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1101 - mse: 0.1101 - val_loss: 0.1030 - val_mse: 0.1030 - lr: 1.0000e-04\n", "Epoch 15/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1090 - mse: 0.1090 - val_loss: 0.1023 - val_mse: 0.1023 - lr: 1.0000e-04\n", "Epoch 16/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1081 - mse: 0.1081 - val_loss: 0.1033 - val_mse: 0.1033 - lr: 1.0000e-04\n", "Epoch 17/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1070 - mse: 0.1070 - val_loss: 0.1039 - val_mse: 0.1039 - lr: 1.0000e-04\n", "Epoch 18/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1064 - mse: 0.1064 - val_loss: 0.1015 - val_mse: 0.1015 - lr: 1.0000e-04\n", "Epoch 19/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1055 - mse: 0.1055 - val_loss: 0.0993 - val_mse: 0.0993 - lr: 1.0000e-04\n", "Epoch 20/100\n", "102/102 [==============================] - 1s 13ms/step - loss: 0.1050 - mse: 0.1050 - val_loss: 0.0996 - val_mse: 0.0996 - lr: 1.0000e-04\n", "Epoch 21/100\n", "102/102 [==============================] - 1s 13ms/step - loss: 0.1041 - mse: 0.1041 - val_loss: 0.0995 - val_mse: 0.0995 - lr: 1.0000e-04\n", "Epoch 22/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1040 - mse: 0.1040 - val_loss: 0.0999 - val_mse: 0.0999 - lr: 1.0000e-04\n", "Epoch 23/100\n", "102/102 [==============================] - 1s 11ms/step - loss: 0.1039 - mse: 0.1039 - val_loss: 0.1009 - val_mse: 0.1009 - lr: 1.0000e-04\n", "Epoch 24/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1032 - mse: 0.1032 - val_loss: 0.0987 - val_mse: 0.0987 - lr: 1.0000e-04\n", "Epoch 25/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1036 - mse: 0.1036 - val_loss: 0.0980 - val_mse: 0.0980 - lr: 1.0000e-04\n", "Epoch 26/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1031 - mse: 0.1031 - val_loss: 0.0997 - val_mse: 0.0997 - lr: 1.0000e-04\n", "Epoch 27/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1032 - mse: 0.1032 - val_loss: 0.0977 - val_mse: 0.0977 - lr: 1.0000e-04\n", "Epoch 28/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1026 - mse: 0.1026 - val_loss: 0.1007 - val_mse: 0.1007 - lr: 1.0000e-04\n", "Epoch 29/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1029 - mse: 0.1029 - val_loss: 0.0980 - val_mse: 0.0980 - lr: 1.0000e-04\n", "Epoch 30/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1034 - mse: 0.1034 - val_loss: 0.1009 - val_mse: 0.1009 - lr: 1.0000e-04\n", "Epoch 31/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1021 - mse: 0.1021 - val_loss: 0.0978 - val_mse: 0.0978 - lr: 1.0000e-04\n", "Epoch 32/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1024 - mse: 0.1024 - val_loss: 0.0987 - val_mse: 0.0987 - lr: 1.0000e-04\n", "Epoch 33/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.0980 - val_mse: 0.0980 - lr: 1.0000e-04\n", "Epoch 34/100\n", "102/102 [==============================] - 1s 14ms/step - loss: 0.1022 - mse: 0.1022 - val_loss: 0.0975 - val_mse: 0.0975 - lr: 1.0000e-04\n", "Epoch 35/100\n", "102/102 [==============================] - 2s 15ms/step - loss: 0.1018 - mse: 0.1018 - val_loss: 0.0974 - val_mse: 0.0974 - lr: 1.0000e-04\n", "Epoch 36/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.0987 - val_mse: 0.0987 - lr: 1.0000e-04\n", "Epoch 37/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1024 - mse: 0.1024 - val_loss: 0.0988 - val_mse: 0.0988 - lr: 1.0000e-04\n", "Epoch 38/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-04\n", "Epoch 39/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.0976 - val_mse: 0.0976 - lr: 1.0000e-04\n", "Epoch 40/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1016 - mse: 0.1016 - val_loss: 0.1002 - val_mse: 0.1002 - lr: 1.0000e-04\n", "Epoch 41/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1020 - mse: 0.1020 - val_loss: 0.0990 - val_mse: 0.0990 - lr: 1.0000e-04\n", "Epoch 42/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1019 - mse: 0.1019 - val_loss: 0.0976 - val_mse: 0.0976 - lr: 1.0000e-04\n", "Epoch 43/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1017 - mse: 0.1017 - val_loss: 0.0977 - val_mse: 0.0977 - lr: 1.0000e-04\n", "Epoch 44/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1022 - mse: 0.1022 - val_loss: 0.1002 - val_mse: 0.1002 - lr: 1.0000e-04\n", "Epoch 45/100\n", " 96/102 [===========================>..] - ETA: 0s - loss: 0.1013 - mse: 0.1013\n", "Epoch 45: ReduceLROnPlateau reducing learning rate to 9.999999747378752e-06.\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1012 - mse: 0.1012 - val_loss: 0.0973 - val_mse: 0.0973 - lr: 1.0000e-04\n", "Epoch 46/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0973 - val_mse: 0.0973 - lr: 1.0000e-05\n", "Epoch 47/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0973 - val_mse: 0.0973 - lr: 1.0000e-05\n", "Epoch 48/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 49/100\n", "102/102 [==============================] - 2s 15ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 50/100\n", "102/102 [==============================] - 2s 15ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 51/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 52/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 53/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0976 - val_mse: 0.0976 - lr: 1.0000e-05\n", "Epoch 54/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 55/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0973 - val_mse: 0.0973 - lr: 1.0000e-05\n", "Epoch 56/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 57/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 58/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0969 - val_mse: 0.0969 - lr: 1.0000e-05\n", "Epoch 59/100\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 60/100\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-05\n", "Epoch 61/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1008 - mse: 0.1008 - val_loss: 0.0974 - val_mse: 0.0974 - lr: 1.0000e-05\n", "Epoch 62/100\n", "102/102 [==============================] - 1s 11ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0974 - val_mse: 0.0974 - lr: 1.0000e-05\n", "Epoch 63/100\n", "102/102 [==============================] - 1s 13ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-05\n", "Epoch 64/100\n", "102/102 [==============================] - 2s 15ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0973 - val_mse: 0.0973 - lr: 1.0000e-05\n", "Epoch 65/100\n", " 96/102 [===========================>..] - ETA: 0s - loss: 0.1008 - mse: 0.1008\n", "Epoch 65: ReduceLROnPlateau reducing learning rate to 1e-06.\n", "102/102 [==============================] - 1s 10ms/step - loss: 0.1007 - mse: 0.1007 - val_loss: 0.0976 - val_mse: 0.0976 - lr: 1.0000e-05\n", "Epoch 66/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.0972 - val_mse: 0.0972 - lr: 1.0000e-06\n", "Epoch 67/100\n", "102/102 [==============================] - 1s 9ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-06\n", "Epoch 68/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-06\n", "Epoch 69/100\n", "102/102 [==============================] - 1s 8ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.0973 - val_mse: 0.0973 - lr: 1.0000e-06\n", "Epoch 70/100\n", " 93/102 [==========================>...] - ETA: 0s - loss: 0.1007 - mse: 0.1007Restoring model weights from the end of the best epoch: 58.\n", "102/102 [==============================] - 1s 7ms/step - loss: 0.1006 - mse: 0.1006 - val_loss: 0.0971 - val_mse: 0.0971 - lr: 1.0000e-06\n", "Epoch 70: early stopping\n" ] } ] }, { "cell_type": "markdown", "source": [ "#Prediction using DNN" ], "metadata": { "id": "GO3zGdD-Mgbe" } }, { "cell_type": "code", "source": [ "plt.plot(history.history['loss'], label='Training Loss')\n", "plt.plot(history.history['val_loss'], label='Validation Loss')\n", "\n", "plt.xlabel('Epoch')\n", "plt.ylabel('Loss')\n", "plt.title('Training Curves')\n", "plt.legend()\n", "plt.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 472 }, "id": "vOiYz_kFMOeh", "outputId": "7cb01d15-75ea-4488-d19f-671d50f436c8" }, "execution_count": 24, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkgAAAHHCAYAAABEEKc/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACIaElEQVR4nOzdeVhUZfvA8e/MAMO+y6KiiPuKhopkqSWFWpZJZWWpZbapZeSb+as0bdHKylLTlldt0TR73bJccalc0lDcNVQUFxZR2feZ8/vjyCgKCggclvtzXXPNmTPPnHOf4+jcPqtOURQFIYQQQghhodc6ACGEEEKI6kYSJCGEEEKIa0iCJIQQQghxDUmQhBBCCCGuIQmSEEIIIcQ1JEESQgghhLiGJEhCCCGEENeQBEkIIYQQ4hqSIAkhhBBCXEMSJCGEpoYNG4a/v3+5PvvOO++g0+kqNiAhhEASJCFECXQ6Xakemzdv1jpUTW3evJmBAwfi4+ODjY0NXl5e9O/fn6VLl2odmhDiFuhkLTYhRHF+/PHHIq+///571q9fzw8//FBk/z333IO3t3e5z5Ofn4/ZbMZoNJb5swUFBRQUFGBra1vu89+KiRMnMnnyZJo3b87jjz9O48aNuXDhAr///jubN29mwYIFPPHEE5rEJoS4NZIgCSFKZdSoUcyaNYub/ZORlZWFvb19FUWlnV9++YVHHnmEhx9+mIULF2JtbV3k/bVr15Kfn8/9999/y+eqK/dUiOpEmtiEEOXWq1cv2rVrR1RUFD169MDe3p7/+7//A2DFihXcd9991K9fH6PRSNOmTXn33XcxmUxFjnFtH6STJ0+i0+mYNm0aX3/9NU2bNsVoNNKlSxd27dpV5LPF9UHS6XSMGjWK5cuX065dO4xGI23btmXNmjXXxb9582Y6d+6Mra0tTZs25auvvip1v6a3334bd3d35s6de11yBBAWFmZJjubPn49Op+PkyZPXnf/aZsqS7un9999PQEBAsbGEhITQuXPnIvt+/PFHgoKCsLOzw93dnccee4zTp08XKRMTE0N4eDg+Pj7Y2trSsGFDHnvsMVJTU296/ULUdlZaByCEqNkuXLhA3759eeyxx3jyySctzW3z58/H0dGRiIgIHB0d2bhxIxMmTCAtLY2PP/74psdduHAh6enpPP/88+h0Oj766CMGDhzIiRMnik1IrvbXX3+xdOlSXnrpJZycnPjiiy8IDw8nLi4ODw8PAPbs2UOfPn3w9fVl0qRJmEwmJk+eTL169W4aW0xMDEeOHOGZZ57BycmpFHepbIq7p0FBQQwZMoRdu3bRpUsXS9lTp06xY8eOIvf0/fff5+233+bRRx/l2Wef5fz588yYMYMePXqwZ88eXF1dycvLIywsjNzcXEaPHo2Pjw9nz55l1apVpKSk4OLiUuHXJUSNogghRCmMHDlSufafjJ49eyqAMmfOnOvKZ2VlXbfv+eefV+zt7ZWcnBzLvqFDhyqNGze2vI6NjVUAxcPDQ7l48aJl/4oVKxRA+fXXXy37Jk6ceF1MgGJjY6McO3bMsm/v3r0KoMyYMcOyr3///oq9vb1y9uxZy76YmBjFysrqumNeqzCWzz777IblCs2bN08BlNjY2CL7N23apADKpk2bLPtKuqepqamK0WhUXnvttSL7P/roI0Wn0ymnTp1SFEVRTp48qRgMBuX9998vUm7//v2KlZWVZf+ePXsUQFmyZEmprkGIukaa2IQQt8RoNPL0009ft9/Ozs6ynZ6eTnJyMnfeeSdZWVkcOXLkpscdNGgQbm5ultd33nknACdOnLjpZ0NDQ2natKnldYcOHXB2drZ81mQysWHDBgYMGED9+vUt5Zo1a0bfvn1vevy0tDSASqk9guLvqbOzM3379uXnn38u0g9s8eLFdOvWjUaNGgGwdOlSzGYzjz76KMnJyZaHj48PzZs3Z9OmTQCWGqK1a9eSlZVVKdchRE0mCZIQ4pY0aNAAGxub6/YfPHiQhx56CBcXF5ydnalXrx5PPvkkQKn6uBT+4BcqTJYuXbpU5s8Wfr7ws0lJSWRnZ9OsWbPryhW371rOzs6AmvhVhpLu6aBBgzh9+jTbt28H4Pjx40RFRTFo0CBLmZiYGBRFoXnz5tSrV6/I4/DhwyQlJQHQpEkTIiIi+Pbbb/H09CQsLIxZs2ZJ/yMhLpM+SEKIW3J1TVGhlJQUevbsibOzM5MnT6Zp06bY2tqye/duxo0bh9lsvulxDQZDsfuVUgy8vZXPlkarVq0A2L9/f6nKl9Tp+9oO64WKu6cA/fv3x97enp9//pnbb7+dn3/+Gb1ezyOPPGIpYzab0el0rF69utj74OjoaNn+5JNPGDZsGCtWrGDdunW8/PLLTJkyhR07dtCwYcNSXZsQtZUkSEKICrd582YuXLjA0qVL6dGjh2V/bGyshlFd4eXlha2tLceOHbvuveL2XatFixa0bNmSFStW8PnnnxdJOopTWPuVkpJSZP+pU6dKHzTg4ODA/fffz5IlS/j0009ZvHgxd955Z5FmwqZNm6IoCk2aNKFFixY3PWb79u1p3749b731Ftu2baN79+7MmTOH9957r0yxCVHbSBObEKLCFdZcXF1jk5eXx5dffqlVSEUYDAZCQ0NZvnw5586ds+w/duwYq1evLtUxJk2axIULF3j22WcpKCi47v1169axatUqAEt/qD/++MPyvslk4uuvvy5z7IMGDeLcuXN8++237N27t0jzGsDAgQMxGAxMmjTpuhozRVG4cOECoPajujbu9u3bo9fryc3NLXNcQtQ2UoMkhKhwt99+O25ubgwdOpSXX34ZnU7HDz/8UGFNXBXhnXfeYd26dXTv3p0XX3wRk8nEzJkzadeuHdHR0Tf9/KBBg9i/fz/vv/8+e/bsKTKT9po1a4iMjGThwoUAtG3blm7dujF+/HguXryIu7s7ixYtKjaxupl+/frh5OTE2LFjMRgMhIeHF3m/adOmvPfee4wfP56TJ08yYMAAnJyciI2NZdmyZTz33HOMHTuWjRs3MmrUKB555BFatGhBQUEBP/zwQ7HHFKIukgRJCFHhPDw8WLVqFa+99hpvvfUWbm5uPPnkk/Tu3ZuwsDCtwwMgKCiI1atXM3bsWN5++238/PyYPHkyhw8fLtUoO4D33nuPu+++my+++ILZs2dz8eJF3Nzc6NatGytWrOCBBx6wlF2wYAHPP/88U6dOxdXVleHDh3PXXXdxzz33lCluW1tbHnjgARYsWEBoaCheXl7XlXnjjTdo0aIFn332GZMmTQLAz8+Pe++91xJTYGAgYWFh/Prrr5w9exZ7e3sCAwNZvXo13bp1K1NMQtRGstSIEEJcZcCAARw8eJCYmBitQxFCaEj6IAkh6qzs7Owir2NiYvj999/p1auXNgEJIaoNqUESQtRZvr6+DBs2jICAAE6dOsXs2bPJzc1lz549NG/eXOvwhBAakj5IQog6q0+fPvz0008kJCRgNBoJCQnhgw8+kORICCE1SEIIIYQQ15I+SEIIIYQQ15AESQghhBDiGtIHqZzMZjPnzp3DycmpxHWWhBBCCFG9KIpCeno69evXR68vuZ5IEqRyOnfuHH5+flqHIYQQQohyOH369A0XZZYEqZycnJwA9QY7OztrHI0QQgghSiMtLQ0/Pz/L73hJJEEqp8JmNWdnZ0mQhBBCiBrmZt1jpJO2EEIIIcQ1JEESQgghhLiGJEhCCCGEENeQPkhCCCE0YTKZyM/P1zoMUctYW1tjMBhu+TiSIAkhhKhSiqKQkJBASkqK1qGIWsrV1RUfH59bmqdQEiQhhBBVqjA58vLywt7eXibbFRVGURSysrJISkoCwNfXt9zHkgRJCCFElTGZTJbkyMPDQ+twRC1kZ2cHQFJSEl5eXuVubpNO2kIIIapMYZ8je3t7jSMRtVnh9+tW+rhJgiSEEKLKSbOaqEwV8f2SBEkIIYQQ4hqSIAkhhBAa8ff3Z/r06aUuv3nzZnQ6nYwArAKSIAkhhBA3odPpbvh45513ynXcXbt28dxzz5W6/O233058fDwuLi7lOl9pSSImo9iqnczcAi5m5uFotMLNwUbrcIQQQgDx8fGW7cWLFzNhwgSOHj1q2efo6GjZVhQFk8mEldXNf2Lr1atXpjhsbGzw8fEp02dE+UgNUjUzYcVB7vxoE4t2ndY6FCGEEJf5+PhYHi4uLuh0OsvrI0eO4OTkxOrVqwkKCsJoNPLXX39x/PhxHnzwQby9vXF0dKRLly5s2LChyHGvbWLT6XR8++23PPTQQ9jb29O8eXNWrlxpef/amp358+fj6urK2rVrad26NY6OjvTp06dIQldQUMDLL7+Mq6srHh4ejBs3jqFDhzJgwIBy349Lly4xZMgQ3NzcsLe3p2/fvsTExFjeP3XqFP3798fNzQ0HBwfatm3L77//bvns4MGDqVevHnZ2djRv3px58+aVO5bKIglSNeNiZw1AarZMvy+EqBsURSErr0CTh6IoFXYdb7zxBlOnTuXw4cN06NCBjIwM+vXrR2RkJHv27KFPnz7079+fuLi4Gx5n0qRJPProo+zbt49+/foxePBgLl68WGL5rKwspk2bxg8//MAff/xBXFwcY8eOtbz/4YcfsmDBAubNm8fWrVtJS0tj+fLlt3Stw4YN459//mHlypVs374dRVHo16+fZVj9yJEjyc3N5Y8//mD//v18+OGHllq2t99+m0OHDrF69WoOHz7M7Nmz8fT0vKV4KoM0sVUzkiAJIeqa7HwTbSas1eTchyaHYW9TMT+FkydP5p577rG8dnd3JzAw0PL63XffZdmyZaxcuZJRo0aVeJxhw4bx+OOPA/DBBx/wxRdfsHPnTvr06VNs+fz8fObMmUPTpk0BGDVqFJMnT7a8P2PGDMaPH89DDz0EwMyZMy21OeURExPDypUr2bp1K7fffjsACxYswM/Pj+XLl/PII48QFxdHeHg47du3ByAgIMDy+bi4ODp16kTnzp0BtRatOpIapGrGxU79i5omCZIQQtQohT/4hTIyMhg7diytW7fG1dUVR0dHDh8+fNMapA4dOli2HRwccHZ2tiydURx7e3tLcgTq8hqF5VNTU0lMTKRr166W9w0GA0FBQWW6tqsdPnwYKysrgoODLfs8PDxo2bIlhw8fBuDll1/mvffeo3v37kycOJF9+/ZZyr744ossWrSIjh078vrrr7Nt27Zyx1KZpAapmnG1Vztmp2TnaRyJEEJUDTtrA4cmh2l27ori4OBQ5PXYsWNZv34906ZNo1mzZtjZ2fHwww+Tl3fjf9+tra2LvNbpdJjN5jKVr8imw/J49tlnCQsL47fffmPdunVMmTKFTz75hNGjR9O3b19OnTrF77//zvr16+nduzcjR45k2rRpmsZ8rWpRgzRr1iz8/f2xtbUlODiYnTt3llj24MGDhIeH4+/vj06nK3b+iNmzZ9OhQwecnZ1xdnYmJCSE1atXFynTq1ev64ZpvvDCCxV9aWUmTWxCiLpGp9Nhb2OlyaMyZ/TeunUrw4YN46GHHqJ9+/b4+Phw8uTJSjtfcVxcXPD29mbXrl2WfSaTid27d5f7mK1bt6agoIC///7bsu/ChQscPXqUNm3aWPb5+fnxwgsvsHTpUl577TW++eYby3v16tVj6NCh/Pjjj0yfPp2vv/663PFUFs1rkBYvXkxERARz5swhODiY6dOnExYWxtGjR/Hy8rqufFZWFgEBATzyyCO8+uqrxR6zYcOGTJ06lebNm6MoCt999x0PPvgge/bsoW3btpZyI0aMKNJOWx3WBnKWBEkIIWqF5s2bs3TpUvr3749Op+Ptt9++YU1QZRk9ejRTpkyhWbNmtGrVihkzZnDp0qVSJYf79+/HycnJ8lqn0xEYGMiDDz7IiBEj+Oqrr3BycuKNN96gQYMGPPjggwCMGTOGvn370qJFCy5dusSmTZto3bo1ABMmTCAoKIi2bduSm5vLqlWrLO9VJ5onSJ9++ikjRozg6aefBmDOnDn89ttvzJ07lzfeeOO68l26dKFLly4Axb4P0L9//yKv33//fWbPns2OHTuKJEj29vbVbj6JwhqklCxJkIQQoib79NNPeeaZZ7j99tvx9PRk3LhxpKWlVXkc48aNIyEhgSFDhmAwGHjuuecICwsr1Sr3PXr0KPLaYDBQUFDAvHnzeOWVV7j//vvJy8ujR48e/P7775bmPpPJxMiRIzlz5gzOzs706dOHzz77DFDncho/fjwnT57Ezs6OO++8k0WLFlX8hd8inaJhQ2VeXh729vb88ssvReZjGDp0KCkpKaxYseKGn/f392fMmDGMGTOmxDImk4klS5YwdOhQ9uzZY6n+69WrFwcPHkRRFHx8fOjfvz9vv/12ibVIubm55ObmWl6npaXh5+dHamoqzs7Opb/om0jOyKXze+o8Gcc/6IdBLws6CiFqj5ycHGJjY2nSpAm2trZah1Mnmc1mWrduzaOPPsq7776rdTiV4kbfs7S0NFxcXG76+61pDVJycjImkwlvb+8i+729vTly5MgtHXv//v2EhISQk5ODo6Mjy5YtK9I2+sQTT9C4cWPq16/Pvn37GDduHEePHmXp0qXFHm/KlClMmjTplmIqjcIaJID0nHxLp20hhBCiPE6dOsW6devo2bMnubm5zJw5k9jYWJ544gmtQ6vWNG9iqywtW7YkOjqa1NRUfvnlF4YOHcqWLVssSdLVa9+0b98eX19fevfuzfHjx4sMlyw0fvx4IiIiLK8La5AqmrVBj72Ngaw8E6nZkiAJIYS4NXq9nvnz5zN27FgURaFdu3Zs2LChWvb7qU40TZA8PT0xGAwkJiYW2Z+YmHjLfYNsbGxo1qwZAEFBQezatYvPP/+cr776qtjyhfM5HDt2rNgEyWg0YjQabymm0nKxsyYrz0RKVj6NParklEIIIWopPz8/tm7dqnUYNY6mw/xtbGwICgoiMjLSss9sNhMZGUlISEiFnstsNhfpQ3St6OhoQJ1gS2sy1F8IIYTQluZNbBEREQwdOpTOnTvTtWtXpk+fTmZmpmVU25AhQ2jQoAFTpkwB1I7dhw4dsmyfPXuW6OhoHB0dLTVG48ePp2/fvjRq1Ij09HQWLlzI5s2bWbtWncr++PHjLFy4kH79+uHh4cG+fft49dVX6dGjR5EZTLUiCZIQQgihLc0TpEGDBnH+/HkmTJhAQkICHTt2ZM2aNZaO23Fxcej1Vyq6zp07R6dOnSyvp02bxrRp0+jZsyebN28GICkpiSFDhhAfH4+LiwsdOnRg7dq1ljVybGxs2LBhgyUZ8/PzIzw8nLfeeqvqLvwGJEESQgghtKV5ggTqwnolLdxXmPQU8vf3v+kU6v/9739v+L6fnx9btmwpU4xVSRIkIYQQQlvVYqkRUZSrvSRIQgghhJYkQaqGLDVIMpu2EEIIoQlJkKohaWITQojaqVevXkVWf/D39y920fWr6XQ6li9ffsvnrqjj1BWSIFVDhQvWpmTnaRyJEEIIUNf47NOnT7Hv/fnnn+h0Ovbt21fm4+7atavIxMUV4Z133qFjx47X7Y+Pj6dv374Veq5rzZ8/H1dX10o9R1WRBKkaKpw9OzW7QONIhBBCAAwfPpz169dz5syZ696bN28enTt3Ltc0MfXq1StxDdCK5uPjU2UTHtcGkiBVQ4VNbGnSxCaEENXC/fffT7169Zg/f36R/RkZGSxZsoThw4dz4cIFHn/8cRo0aIC9vT3t27fnp59+uuFxr21ii4mJoUePHtja2tKmTRvWr19/3WfGjRtHixYtsLe3JyAggLfffpv8fPX3Yv78+UyaNIm9e/ei0+nQ6XSWmK9tYtu/fz933303dnZ2eHh48Nxzz5GRkWF5f9iwYQwYMIBp06bh6+uLh4cHI0eOtJyrPOLi4njwwQdxdHTE2dmZRx99tMhqGnv37uWuu+7CyckJZ2dngoKC+OeffwB1Tbn+/fvj5uaGg4MDbdu25ffffy93LDdTLYb5i6KkD5IQok5RFMjP0ubc1vag0920mJWVFUOGDGH+/Pm8+eab6C5/ZsmSJZhMJh5//HEyMjIICgpi3LhxODs789tvv/HUU0/RtGlTunbtetNzmM1mBg4ciLe3N3///TepqalF+isVcnJyYv78+dSvX5/9+/czYsQInJyceP311xk0aBAHDhxgzZo1bNiwAQAXF5frjpGZmUlYWBghISHs2rWLpKQknn32WUaNGlUkCdy0aRO+vr5s2rSJY8eOMWjQIDp27MiIESNuej3FXV9hcrRlyxYKCgoYOXIkgwYNskzpM3jwYDp16sTs2bMxGAxER0djba3+Jo4cOZK8vDz++OMPHBwcOHToEI6OjmWOo7QkQaqGXC8nSBm5BeSbzFgbpKJPCFGL5WfBB/W1Off/nQMbh1IVfeaZZ/j444/ZsmULvXr1AtTmtfDwcFxcXHBxcWHs2LGW8qNHj2bt2rX8/PPPpUqQNmzYwJEjR1i7di3166v344MPPriu39DVkxr7+/szduxYFi1axOuvv46dnR2Ojo5YWVndcE3ThQsXkpOTw/fff4+Dg3r9M2fOpH///nz44YeWyZrd3NyYOXMmBoOBVq1acd999xEZGVmuBCkyMpL9+/cTGxtrWez9+++/p23btuzatYsuXboQFxfHf/7zH1q1agVA8+bNLZ+Pi4sjPDyc9u3bAxAQEFDmGMpCfnmrocJO2iDNbEIIUV20atWK22+/nblz5wLq4uZ//vknw4cPB8BkMvHuu+/Svn173N3dcXR0ZO3atcTFxZXq+IcPH8bPz8+SHAHFrku6ePFiunfvjo+PD46Ojrz11lulPsfV5woMDLQkRwDdu3fHbDZz9OhRy762bdtiMBgsr319fUlKSirTua4+p5+fnyU5AmjTpg2urq4cPnwYUJcfe/bZZwkNDWXq1KkcP37cUvbll1/mvffeo3v37kycOLFcneLLQmqQqiGDXoeT0Yr03AJSs/PxcJROdUKIWszaXq3J0ercZTB8+HBGjx7NrFmzmDdvHk2bNqVnz54AfPzxx3z++edMnz6d9u3b4+DgwJgxY8jLq7gRydu3b2fw4MFMmjSJsLAwXFxcWLRoEZ988kmFneNqhc1bhXQ6HWazuVLOBeoIvCeeeILffvuN1atXM3HiRBYtWsRDDz3Es88+S1hYGL/99hvr1q1jypQpfPLJJ4wePbpSYpEapGrKWfohCSHqCp1ObebS4lGK/kdXe/TRR9Hr9SxcuJDvv/+eZ555xtIfaevWrTz44IM8+eSTBAYGEhAQwL///lvqY7du3ZrTp08THx9v2bdjx44iZbZt20bjxo1588036dy5M82bN+fUqVNFytjY2GAymW56rr1795KZmWnZt3XrVvR6PS1btix1zGVReH2nT5+27Dt06BApKSm0adPGsq9Fixa8+uqrrFu3joEDBzJv3jzLe35+frzwwgssXbqU1157jW+++aZSYgVJkKqtwuVGUiRBEkKIasPR0ZFBgwYxfvx44uPjGTZsmOW95s2bs379erZt28bhw4d5/vnni4zQupnQ0FBatGjB0KFD2bt3L3/++SdvvvlmkTLNmzcnLi6ORYsWcfz4cb744guWLVtWpIy/vz+xsbFER0eTnJxMbm7udecaPHgwtra2DB06lAMHDrBp0yZGjx7NU089Zel/VF4mk4no6Ogij8OHDxMaGkr79u0ZPHgwu3fvZufOnQwZMoSePXvSuXNnsrOzGTVqFJs3b+bUqVNs3bqVXbt20bp1awDGjBnD2rVriY2NZffu3WzatMnyXmWQBKmakqH+QghRPQ0fPpxLly4RFhZWpL/QW2+9xW233UZYWBi9evXCx8eHAQMGlPq4er2eZcuWkZ2dTdeuXXn22Wd5//33i5R54IEHePXVVxk1ahQdO3Zk27ZtvP3220XKhIeH06dPH+666y7q1atX7FQD9vb2rF27losXL9KlSxcefvhhevfuzcyZM8t2M4qRkZFBp06dijz69++PTqdjxYoVuLm50aNHD0JDQwkICGDx4sUAGAwGLly4wJAhQ2jRogWPPvooffv2ZdKkSYCaeI0cOZLWrVvTp08fWrRowZdffnnL8ZZEpyiKUmlHr8XS0tJwcXEhNTUVZ2fnCj/+iz9GsfpAApMfbMuQEP8KP74QQmghJyeH2NhYmjRpgq2trdbhiFrqRt+z0v5+Sw1SNSUL1gohhBDakQSpmnKRPkhCCCGEZiRBqqZkNm0hhBBCO5IgVVOSIAkhhBDakQSpmpIESQhRm8n4IFGZKuL7JQlSNeVqZwNIJ20hRO1SODNzVpZGi9OKOqHw+3XtTOBlIUuNVFNSgySEqI0MBgOurq6W9bzs7e0tM1ELcasURSErK4ukpCRcXV2LrCNXVpIgVVOSIAkhaqvCVebLu+ipEDfj6upq+Z6VlyRI1VRhgpSdbyK3wITRqvxZsBBCVCc6nQ5fX1+8vLzIz5f/BIqKZW1tfUs1R4UkQaqmnGyt0OlAUdRaJC8nSZCEELWLwWCokB8yISqDdNKupvR6Hc62sh6bEEIIoQVJkKox6YckhBBCaEMSpGpMEiQhhBBCG5IgVWOuheuxyVxIQgghRJWSBKkac5YaJCGEEEITkiBVY9LEJoQQQmhDEqRqTBIkIYQQQhuSIFVjroUJkvRBEkIIIapUtUiQZs2ahb+/P7a2tgQHB7Nz584Syx48eJDw8HD8/f3R6XRMnz79ujKzZ8+mQ4cOODs74+zsTEhICKtXry5SJicnh5EjR+Lh4YGjoyPh4eEkJiZW9KXdEqlBEkIIIbSheYK0ePFiIiIimDhxIrt37yYwMJCwsLAS1+jJysoiICCAqVOnlrjOSsOGDZk6dSpRUVH8888/3H333Tz44IMcPHjQUubVV1/l119/ZcmSJWzZsoVz584xcODASrnG8pIESQghhNCGTlEURcsAgoOD6dKlCzNnzgTAbDbj5+fH6NGjeeONN274WX9/f8aMGcOYMWNueh53d3c+/vhjhg8fTmpqKvXq1WPhwoU8/PDDABw5coTWrVuzfft2unXrdtPjpaWl4eLiQmpqKs7Ozje/0HLYdjyZJ775m+ZejqyP6Fkp5xBCCCHqktL+fmtag5SXl0dUVBShoaGWfXq9ntDQULZv314h5zCZTCxatIjMzExCQkIAiIqKIj8/v8h5W7VqRaNGjUo8b25uLmlpaUUela2wBilFapCEEEKIKqVpgpScnIzJZMLb27vIfm9vbxISEm7p2Pv378fR0RGj0cgLL7zAsmXLaNOmDQAJCQnY2Njg6upa6vNOmTIFFxcXy8PPz++W4isNaWITQgghtKF5H6TK0rJlS6Kjo/n777958cUXGTp0KIcOHSr38caPH09qaqrlcfr06QqMtniFCVJegZmcfFOln08IIYQQKistT+7p6YnBYLhu9FhiYmKJHbBLy8bGhmbNmgEQFBTErl27+Pzzz/nqq6/w8fEhLy+PlJSUIrVINzqv0WjEaDTeUkxl5Wi0wqDXYTIrpGbnY2ttqNLzCyGEEHWVpjVINjY2BAUFERkZadlnNpuJjIy09BeqKGazmdzcXEBNmKytrYuc9+jRo8TFxVX4eW+FTqe70g9J5kISQgghqoymNUgAERERDB06lM6dO9O1a1emT59OZmYmTz/9NABDhgyhQYMGTJkyBVA7dhc2leXl5XH27Fmio6NxdHS01BiNHz+evn370qhRI9LT01m4cCGbN29m7dq1ALi4uDB8+HAiIiJwd3fH2dmZ0aNHExISUqoRbFXJxc6ai5l50g9JCCGEqEKaJ0iDBg3i/PnzTJgwgYSEBDp27MiaNWssHbfj4uLQ669UdJ07d45OnTpZXk+bNo1p06bRs2dPNm/eDEBSUhJDhgwhPj4eFxcXOnTowNq1a7nnnnssn/vss8/Q6/WEh4eTm5tLWFgYX375ZdVcdBnIgrVCCCFE1dN8HqSaqirmQQIYOncnW/49z7RHAnk4qGGlnUcIIYSoC2rEPEji5q70QcrTOBIhhBCi7pAEqZorTJDSpIlNCCGEqDKSIFVzMlmkEEIIUfUkQarmXO0lQRJCCCGqmiRI1ZyzrMcmhBBCVDlJkKo5aWITQgghqp4kSNWcJEhCCCFE1ZMEqZor7IMko9iEEEKIqiMJUjV39VpsMqenEEIIUTUkQarmChOkArNCVp5J42iEEEKIukESpGrOztqAtUEHSD8kIYQQoqpIglTN6XQ6XOxsALWZTQghhBCVTxKkGsDFzgqQGiQhhBCiqkiCVAPIUH8hhBCiakmCVAPIgrVCCCFE1ZIEqQZwtb/cByk7T+NIhBBCiLpBEqQaQJrYhBBCiKolCVIN4CwJkhBCCFGlJEGqAVwtCVKBxpEIIYQQdYMkSDXAleVGpA+SEEIIURUkQaoBZBSbEEIIUbUkQaoBXOylD5IQQghRlSRBqgFcpZO2EEIIUaUkQaoBrh7mbzYrGkcjhBBC1H6SINUAhcP8zQpk5MlINiGEEKKySYJUA9haGzBaqX9UqVnSzCaEEEJUNkmQaghX6agthBBCVBlJkGoIWW5ECCGEqDqSINUQkiAJIYQQVUcSpBpCEiQhhBCi6kiCVEO42NkAkiAJIYQQVUESpBriynpskiAJIYQQla1aJEizZs3C398fW1tbgoOD2blzZ4llDx48SHh4OP7+/uh0OqZPn35dmSlTptClSxecnJzw8vJiwIABHD16tEiZXr16odPpijxeeOGFir60CiNNbEIIIUTV0TxBWrx4MREREUycOJHdu3cTGBhIWFgYSUlJxZbPysoiICCAqVOn4uPjU2yZLVu2MHLkSHbs2MH69evJz8/n3nvvJTMzs0i5ESNGEB8fb3l89NFHFX59FcXFzgqQBWuFEEKIqmCldQCffvopI0aM4OmnnwZgzpw5/Pbbb8ydO5c33njjuvJdunShS5cuAMW+D7BmzZoir+fPn4+XlxdRUVH06NHDst/e3r7EJKu6cbWXPkhCCCFEVdG0BikvL4+oqChCQ0Mt+/R6PaGhoWzfvr3CzpOamgqAu7t7kf0LFizA09OTdu3aMX78eLKysko8Rm5uLmlpaUUeVcnSByk7r0rPK4QQQtRFmtYgJScnYzKZ8Pb2LrLf29ubI0eOVMg5zGYzY8aMoXv37rRr186y/4knnqBx48bUr1+fffv2MW7cOI4ePcrSpUuLPc6UKVOYNGlShcRUHs7SB0kIIYSoMpo3sVW2kSNHcuDAAf76668i+5977jnLdvv27fH19aV3794cP36cpk2bXnec8ePHExERYXmdlpaGn59f5QV+DUsnbRnFJoQQQlQ6TRMkT09PDAYDiYmJRfYnJiZWSN+gUaNGsWrVKv744w8aNmx4w7LBwcEAHDt2rNgEyWg0YjQabzmm8ipciy09t4B8kxlrg+b964UQQohaS9NfWRsbG4KCgoiMjLTsM5vNREZGEhISUu7jKorCqFGjWLZsGRs3bqRJkyY3/Ux0dDQAvr6+5T5vZXK3t8HZ1gpFgUPnqrb/kxBCCFHXaN7EFhERwdChQ+ncuTNdu3Zl+vTpZGZmWka1DRkyhAYNGjBlyhRA7dh96NAhy/bZs2eJjo7G0dGRZs2aAWqz2sKFC1mxYgVOTk4kJCQA4OLigp2dHcePH2fhwoX069cPDw8P9u3bx6uvvkqPHj3o0KGDBnfh5vR6HV383Yk8ksSukxcJ9HPVOiQhhBCi1tI8QRo0aBDnz59nwoQJJCQk0LFjR9asWWPpuB0XF4def6Wi69y5c3Tq1Mnyetq0aUybNo2ePXuyefNmAGbPng2ok0Febd68eQwbNgwbGxs2bNhgScb8/PwIDw/nrbfeqtyLvUVdmqgJ0s7Yizx7Z4DW4QghhBC1lk5RFEXrIGqitLQ0XFxcSE1NxdnZuUrOGXXqEuGzt+HuYEPUW6HodLoqOa8QQghRW5T291t6+tYg7Ru4YGut52JmHsfPZ2gdjhBCCFFrSYJUg9hY6el4ue/RzthL2gYjhBBC1GKSINUwXf3V2cB3xl7QOBIhhBCi9pIEqYbp0kRNkHadlBokIYQQorJIglTD3NbIDYNex9mUbM6mZGsdjhBCCFErSYJUwzgYrWhbX+11vyv2osbRCCGEELWTJEg1kKUf0klJkIQQQojKIAlSDWTphyQ1SEIIIUSlkASpBupyuQYpJimDi5l5GkcjhBBC1D6SINVA7g42NPNyBGCXNLMJIYQQFU4SpBqqsBZJmtmEEEKIiicJUg3VtYkbIDVIQgghRGWQBKmGKqxBOnAujczcAo2jEUIIIWoXSZBqqIZu9tR3scVkVtgTl6J1OEIIIUStIglSDVY43F/mQxJCCCEqliRINVhXmQ9JCCGEqBSSINVghTNq7467RF6BWeNohBBCiNpDEqQarJmXI2721uQWmNl/NlXrcIQQQohaQxKkGkyn09G5cD4k6YckhBBCVBhJkGq4rjJhpBBCCFHhJEGq4QpHsv1z6hJms6JxNEIIIUTtIAlSDde2vjN21gZSs/P5Nyld63CEEEKIWkESpBrO2qDntsauAOyUZjYhhBCiQkiCVAuEBHgAsP5QosaRCCGEELWDJEi1wP0d6gOw9VgySek5GkcjhBBC1HySINUC/p4OdGrkilmBX/fGax2OEEIIUeNJglRLPNSpAQDL95zVOBIhhBCi5pMEqZa4r70vVnod+8+mckxGswkhhBC3RBKkWsLD0UjPFvUAWL7nnMbRCCGEEDWbJEi1yIDCZrboszJppBBCCHELJEGqRUJbe+NotOLMpWyi4i5pHY4QQghRY0mCVIvY2Rjo084HgGXSWVsIIYQot2qRIM2aNQt/f39sbW0JDg5m586dJZY9ePAg4eHh+Pv7o9PpmD59+nVlpkyZQpcuXXBycsLLy4sBAwZw9OjRImVycnIYOXIkHh4eODo6Eh4eTmJizZ9osXA022/74skrMGscjRBCCFEzaZ4gLV68mIiICCZOnMju3bsJDAwkLCyMpKSkYstnZWUREBDA1KlT8fHxKbbMli1bGDlyJDt27GD9+vXk5+dz7733kpmZaSnz6quv8uuvv7JkyRK2bNnCuXPnGDhwYKVcY1XqFuCBt7OR1Ox8Nh8t/h4KIYQQ4sZ0iqJo2ps3ODiYLl26MHPmTADMZjN+fn6MHj2aN95444af9ff3Z8yYMYwZM+aG5c6fP4+XlxdbtmyhR48epKamUq9ePRYuXMjDDz8MwJEjR2jdujXbt2+nW7duN407LS0NFxcXUlNTcXZ2Lt3FVpEPfj/M13+coF97H74cHKR1OEIIIUS1Udrfb01rkPLy8oiKiiI0NNSyT6/XExoayvbt2yvsPKmpqQC4u7sDEBUVRX5+fpHztmrVikaNGpV43tzcXNLS0oo8qqsHO6pLj2w4nERqdr7G0QghhBA1j6YJUnJyMiaTCW9v7yL7vb29SUhIqJBzmM1mxowZQ/fu3WnXrh0ACQkJ2NjY4OrqWurzTpkyBRcXF8vDz8+vQuKrDG18nWnh7UhegZk1B2TpESGEEKKsNO+DVNlGjhzJgQMHWLRo0S0dZ/z48aSmploep0+frqAIK55Op7PMiSSj2YQQQoiy0zRB8vT0xGAwXDd6LDExscQO2GUxatQoVq1axaZNm2jYsKFlv4+PD3l5eaSkpJT6vEajEWdn5yKP6uzBjmqCtOPERc6lZGscjRBCCFGzaJog2djYEBQURGRkpGWf2WwmMjKSkJCQch9XURRGjRrFsmXL2LhxI02aNCnyflBQENbW1kXOe/ToUeLi4m7pvNVJA1c7gpuofa5W7pWlR4QQQoiysNI6gIiICIYOHUrnzp3p2rUr06dPJzMzk6effhqAIUOG0KBBA6ZMmQKoHbsPHTpk2T579izR0dE4OjrSrFkzQG1WW7hwIStWrMDJycnSr8jFxQU7OztcXFwYPnw4ERERuLu74+zszOjRowkJCSnVCLaa4qFODfg79iLLdp/l+R4B6HQ6rUMSQgghagTNh/kDzJw5k48//piEhAQ6duzIF198QXBwMAC9evXC39+f+fPnA3Dy5MnraoQAevbsyebNmwFKTATmzZvHsGHDAHWiyNdee42ffvqJ3NxcwsLC+PLLL0vdtFedh/kXSs3Op8v7G8grMPPtkM6EtvG++YeEEEKIWqy0v9/VIkGqiWpCggTw4ZojzN58HD93O9a/2hNba4PWIQkhhBCaqRHzIInKN+quZvi62HL6YjZzthzXOhwhhBCiRpAEqZZzMFrx5n2tAZi9+TinL2ZpHJEQQghR/UmCVF1VYMvnfe19ub2pB7kFZiavOlRhxxVCCCFqK0mQqpu/v4Ivb4djGyrskDqdjskPtsVKr2P9oUQ2ySK2QgghxA1JglTdJP8LSQfh8MoKPWwzLyeeuUMd/Tdp5UFyC0wVenwhhBCiNpEEqbpp3V99PvIbmCs2iXm5d3O8nIycvJDFN3+cqNBjCyGEELWJJEjVTePuYOcGWRcgbnuFHtrxqg7bMzcd46wsQSKEEEIUSxKk6sZgDS37qduHKraZDeCBwPoEN3EnJ9/Me9JhWwghhCiWJEjVUWEz2+FfwWyu0EOrHbbbYdDrWH0ggc3SYVsIIYS4jiRI1VHAXWDjCOnn4NyeCj98Sx8nhob4A/DyT3s4mpBe4ecQQgghajJJkKoja1tofq+6XcGj2Qr9J6wlQY3dSMspYMjcv2UCSSGEEOIqkiBVV5ZmtpUVOmlkITsbA/8d2pkW3o4kpuXy1H//Jjkjt8LPI4QQQtRE5UqQTp8+zZkzZyyvd+7cyZgxY/j6668rLLA6r/k9YDDCxROQdLhSTuFqb8P3zwTTwNWOkxeyGDZvJ+k5+ZVyLiGEEKImKVeC9MQTT7Bp0yYAEhISuOeee9i5cydvvvkmkydPrtAA6yyjEzS9W90+/GulncbHxZYfnw3Gw8GGA2fTeO77KHLyZRJJIYQQdVu5EqQDBw7QtWtXAH7++WfatWvHtm3bWLBgAfPnz6/I+Oq2q5vZKlETTwe+e6YrjkYrtp+4wJhF0ZjMFd+sJ4QQQtQU5UqQ8vPzMRqNAGzYsIEHHngAgFatWhEfH19x0dV1LfuCzgCJB+DC8Uo9VbsGLnw9JAgbg541BxN4a/l+lEro+ySEEELUBOVKkNq2bcucOXP4888/Wb9+PX369AHg3LlzeHh4VGiAdZq9OzS5U90+sqrST3d7U0++eLwjeh38tPM00zfEVPo5hRBCiOqoXAnShx9+yFdffUWvXr14/PHHCQwMBGDlypWWpjdRQa6eNLIK9Gnny3sD2gPweWQMi3bGVcl5hRBCiOpEp5SzHcVkMpGWloabm5tl38mTJ7G3t8fLy6vCAqyu0tLScHFxITU1FWdn58o7UXoCfNIKUCDiMDjXr7xzXeWTdUeZsfEYBr2Ob4YEcXcr7yo5rxBCCFGZSvv7Xa4apOzsbHJzcy3J0alTp5g+fTpHjx6tE8lRlXLyAb/LtXJHfquy00bc04KHgxpiMiuMXLCHvadTquzcQgghhNbKlSA9+OCDfP/99wCkpKQQHBzMJ598woABA5g9e3aFBii40sx2aEWVnVKn0zFlYHt6tKhHdr6JZ+bv4mRyZpWdXwghhNBSuRKk3bt3c+edaufhX375BW9vb06dOsX333/PF198UaEBCq4kSKe2QuaFKjuttUHPl4Nvo10DZy5k5jF03k6ZbVsIIUSdUK4EKSsrCycnJwDWrVvHwIED0ev1dOvWjVOnTlVogAJw8wefDqCY4ejvVXpqR6MVc4d1oaGbHacuZDF8/i6y8gqqNAYhhBCiqpUrQWrWrBnLly/n9OnTrF27lnvvVRdWTUpKqtwOy3VZa3WuqaoazXY1LydbvnumK2721uw9k8qI7/8hM1eSJCGEELVXuRKkCRMmMHbsWPz9/enatSshISGAWpvUqVOnCg1QXFbYzHZiE+RVfV+gpvUc+XZoFxxsDGw9doGn/vs3qVmybpsQQojaqdzD/BMSEoiPjycwMBC9Xs2zdu7cibOzM61atarQIKujKhvmX0hRYHoHSI2DJ/8HzUIr/5zF2BN3iWHzdpGanU9rX2e+f6Yr9ZyMmsQihBBClFWlDvMH8PHxoVOnTpw7d44zZ84A0LVr1zqRHGlCp4OAHur2iS2ahdGpkRuLn++Gp6ORw/FpDPpqO2dTsjWLRwghhKgM5UqQzGYzkydPxsXFhcaNG9O4cWNcXV159913MZvNFR2jKNSkl/ocq12CBNDKx5klL4TQwNWOE8mZPDpnO7EyBYAQQohapFwJ0ptvvsnMmTOZOnUqe/bsYc+ePXzwwQfMmDGDt99+u6JjFIUK12WL3wdZF7UNxdOBJS+EEODpwNmUbB6Zs53D8WmaxiSEEEJUlHL1Qapfvz5z5szhgQceKLJ/xYoVvPTSS5w9e7bCAqyuqrwPUqFZwXD+CDz6PbR5sOrOW4LkjFye+u9ODsen4WxrxdTwDvRr76t1WEIIIUSxKrUP0sWLF4vta9SqVSsuXtS2ZqPWa9JTfdawH9LVPB2NLBrRjaDGbqTlFPDSgt1E/BxNWo6McBNCCFFzlStBCgwMZObMmdftnzlzJh06dLjloMQNBFxOkDTuh3Q1F3trfhrRjZF3NUWvg6W7z9J3+p/sjJVkWQghRM1UrgTpo48+Yu7cubRp04bhw4czfPhw2rRpw/z585k2bVqZjjVr1iz8/f2xtbUlODiYnTt3llj24MGDhIeH4+/vj06nY/r06deV+eOPP+jfvz/169dHp9OxfPny68oMGzYMnU5X5NGnT58yxa2Zxt1Bp4cLxyC1+jRl2ljp+U9YK35+PgQ/dzvOpmQz6OvtTF19hNwCk9bhCSGEEGVSrgSpZ8+e/Pvvvzz00EOkpKSQkpLCwIEDOXjwID/88EOpj7N48WIiIiKYOHEiu3fvJjAwkLCwMJKSkootn5WVRUBAAFOnTsXHx6fYMpmZmQQGBjJr1qwbnrtPnz7Ex8dbHj/99FOp49aUnSvUvzwZZzWqRSrU2d+d1a/04NHODVEUmLPlOANmbWPb8WRM5nJNuSWEEEJUuXJPFFmcvXv3ctttt2Eyla7GIDg4mC5dulia68xmM35+fowePZo33njjhp/19/dnzJgxjBkzpsQyOp2OZcuWMWDAgCL7hw0bRkpKSrG1S6WlWSdtgA2T4K9PocNjMPCrqj13Gaw5kMD4pfu4dHnG7XpORvq286Ffe1+6+Ltj0Os0jlAIIURdU+kTRd6qvLw8oqKiCA29MiO0Xq8nNDSU7du3V/r5N2/ejJeXFy1btuTFF1/kwoULNyyfm5tLWlpakYdmru6HVHH5bYXr086Hta/24LEufjjbWnE+PZfvt5/isa930G1KJBNWHCD6dIrWYQohhBDX0SxBSk5OxmQy4e3tXWS/t7c3CQkJlXruPn368P333xMZGcmHH37Ili1b6Nu37w1rvqZMmYKLi4vl4efnV6kx3pBfMBiMkB4PyTHaxVEKXk62TA3vwD9v3cO8p7vwcFDDIslS+Oxt0plbCCFEtWOldQBaeOyxxyzb7du3p0OHDjRt2pTNmzfTu3fvYj8zfvx4IiIiLK/T0tK0S5Ks7aBRMMT+odYi1WuhTRxlYGOl566WXtzV0ou8h9qz9XgyszcfZ2fsRb7bfpKuTdy1DlEIIYSwKFOCNHDgwBu+n5KSUupjeXp6YjAYSExMLLI/MTGxxA7YlSUgIABPT0+OHTtWYoJkNBoxGqvRoqxNel5JkLqO0DqaMilMluo5Grl/xl+sP5jIpcw83BxstA5NCCGEAMrYxHZ1E1Nxj8aNGzNkyJBSHcvGxoagoCAiIyMt+8xmM5GRkYSEhJTtKm7RmTNnuHDhAr6+NWgG6IBe6nPsn2CumcPo2zVwoW19Z/JMZpbtqT5TFgghhBBlqkGaN29ehZ48IiKCoUOH0rlzZ7p27cr06dPJzMzk6aefBmDIkCE0aNCAKVOmAGrH7kOHDlm2z549S3R0NI6OjjRr1gyAjIwMjh07ZjlHbGws0dHRuLu706hRIzIyMpg0aRLh4eH4+Phw/PhxXn/9dZo1a0ZYWFiFXl+l8u0IRmfISYGEfVeG/tcwg7r4MWHFQX7+5zRPd1fntxJCCCG0plknbYBBgwYxbdo0JkyYQMeOHYmOjmbNmjWWjttxcXHEx8dbyp87d45OnTrRqVMn4uPjmTZtGp06deLZZ5+1lPnnn38sZUBNwjp16sSECRMAMBgM7Nu3jwceeIAWLVowfPhwgoKC+PPPP6tXE9rNGKzA/w51u5osO1IeDwY2wMZKz5GEdPadSdU6HCGEEAKo4HmQ6hJN50EqtGMOrBkHTe+Gp5ZpE0MFeGXRHlZEn+OJ4EZ88FB7rcMRQghRi1X7eZBEBWjSQ30+tR0KcrWN5RYM6qyOBvw1+hzZeTWzP5UQQojaRRKkmsyrNTh4QUE2nNmldTTl1i3AAz93O9JzC/h9f/zNPyCEEEJUMkmQajKd7kotUg3uh6TX63g0SK1FWvzPaY2jEUIIISRBqvmuXnakBnu4c0P0OtgZe5HY5EytwxFCCFHHSYJU0zW5nCCdjYLcdG1juQW+Lnb0aFEPgJ+lFkkIIYTGJEGq6dwag5s/mAvg5Fato7klhZ21/xd1hgKTWeNohBBC1GWSINUGhbNqLx0Bm6ZATs2cT6h3a288HGxISs9l89HzWocjhBCiDpMEqTbo8R/wbg+5abBlKkzvAH9+ArkZWkdWJjZWeh7q1ACQztpCCCG0JQlSbeDSEJ7/Ax75DjxbqsuPRE6GzwNh2wzIz9Y6wlIb1EVtZtt4JImk9ByNoxFCCFFXSYJUW+j10HYAvLQdBn4D7gGQlQzr3oI5d9aYDtzNvZ3o1MgVk1lh6W5ZwFYIIYQ2JEGqbfQG6PAojNwFD8xUJ5K8EAM7ZmsdWakVdtaevfk40adTtA1GCCFEnSQJUm1lsILbnoK+U9XX22ZA1kVtYyqlAZ0a0KmRK6nZ+Qz+ZgfbjidrHZIQQog6RhKk2q7NQ+BzuQP3X59qHU2p2Fob+HF4MLc39SAzz8SwebvYcChR67CEEELUIZIg1XZ6PfSeqG7v/AbSzmkbTyk5GK2YO6wL97TxJq/AzAs/RrEiWvokCSGEqBqSINUFzUKh0e1QkANbPtQ6mlKztTbw5eDbeKhTAwrMCmMWR7Pg71NahyWEEKIOkASpLtDpIPRyLdLuH+DCcW3jKQNrg55PHgnkyW6NUBR4c9kB5mypOfELIYSomSRBqisadYPmYaCYYON7WkdTJnq9jncfbMdLvZoCMHX1EZbuPqNxVEIIIWozSZDqkt5vq88Hl0L8Xm1jKSOdTsfrfVox8i41SZq86hDJGbkaRyWEEKK2kgSpLvFpD+0eVrcj39U2lnIaE9qC1r7OpGTlM/nXQ1qHI4QQopaSBKmuuev/QG8Fx9bDqW1aR1Nm1gY9H4a3R6+DlXvPsfGIDP8XQghR8SRBqms8msJtQ9TtDZNAUbSNpxw6NHRl+B1NAHhr2QEycgs0jkgIIURtIwlSXdTjdbCyg9M7IGad1tGUy6v3tMDP3Y5zqTlMW3tU63CEEELUMpIg1UXOvhD8nLq94R0wmzQNpzzsbayY8lAHAL7bfpKoU5c0jkgIIURtIglSXXXHq2DrCkmHIHqB1tGUyx3NPXk4qCGKAm/8bx95BWatQxJCCFFLSIJUV9m5QY//qNsb34e8TG3jKac3+7XG09GGmKQMZm+WCSSFEEJUDEmQ6rKuI8C1MWQkwPZZWkdTLm4ONkzs3xaAmZtiiElM1zgiIYQQtYEkSHWZlRF6T1C3t34OGUnaxlNO93fwpXcrL/JNCv/5ZR/5JmlqE0IIcWskQarr2oVD/dsgLwM2T9U6mnLR6XS8O6AdTrZWRJ9OYerqI1qHJIQQooaTBKmu0+ng3strs0XNh/P/ahpOedV3teOTRwIB+O9fsfy+P17jiIQQQtRkkiAJ8O8OLfupC9lueEfraMrt3rY+PN8zAIDXf9nHifMZGkckhBCippIESahCJ4HOAEd/q5FLkBT6z70tCW7iTkZuAS/+uJusPJllWwghRNlJgiRU9VpcWYJk3Vs1cgkSACuDnhlPdKKek5Gjiem8uewASg29FiGEENrRPEGaNWsW/v7+2NraEhwczM6dO0sse/DgQcLDw/H390en0zF9+vTryvzxxx/079+f+vXro9PpWL58+XVlFEVhwoQJ+Pr6YmdnR2hoKDExMRV4VTVUr/Fg7QBno+DgUq2jKTcvJ1tmPt4Jg17Hsj1nWfB3nNYhCSGEqGE0TZAWL15MREQEEydOZPfu3QQGBhIWFkZSUvHDzbOysggICGDq1Kn4+PgUWyYzM5PAwEBmzSp5Xp+PPvqIL774gjlz5vD333/j4OBAWFgYOTk5FXJdNZaTN3R/Rd3e8E6NHfYPEBzgwethLQGY/Osh9p1J0TYgIYQQNYpO0bD9ITg4mC5dujBz5kwAzGYzfn5+jB49mjfeeOOGn/X392fMmDGMGTOmxDI6nY5ly5YxYMAAyz5FUahfvz6vvfYaY8eOBSA1NRVvb2/mz5/PY489VqrY09LScHFxITU1FWdn51J9pkbIy4QZnSH9HDj5wqPfg19XraMqF0VReP6HKNYdSqSBqx2/vXwHrvY2WoclhBBCQ6X9/dasBikvL4+oqChCQ0OvBKPXExoayvbt2yvtvLGxsSQkJBQ5r4uLC8HBwTc8b25uLmlpaUUetZKNAwxZAZ4tIT0e5vWDv7+ukX2SdDodHz8SSGMPe86mZPPW8gNahySEEKKG0CxBSk5OxmQy4e3tXWS/t7c3CQkJlXbewmOX9bxTpkzBxcXF8vDz86u0GDVXrwWMiIQ2A8CcD6v/A8ueh7wsrSMrMxc7a2Y+fhsGvY5V++L5bZ/MjySEEOLmNO+kXVOMHz+e1NRUy+P06dNah1S5jE7wyHy49311+P++xfDfe+BCzVsQtn1DF0b2agrAW8v3cz49V+OIhBBCVHeaJUienp4YDAYSExOL7E9MTCyxA3ZFKDx2Wc9rNBpxdnYu8qj1dDq4fRQM/RUcvCDxAHx9F5zYonVkZTbq7ua08XXmUlY+by7bL0P/hRBC3JBmCZKNjQ1BQUFERkZa9pnNZiIjIwkJCam08zZp0gQfH58i501LS+Pvv/+u1PPWaP7d4fk/wC8YclPV5rbcmjVLtY2Vnk8eDcTaoGPdoUSWR5/VOiQhhBDVmKZNbBEREXzzzTd89913HD58mBdffJHMzEyefvppAIYMGcL48eMt5fPy8oiOjiY6Opq8vDzOnj1LdHQ0x44ds5TJyMiwlAG1U3Z0dDRxcepcODqdjjFjxvDee++xcuVK9u/fz5AhQ6hfv36R0W7iGs6+MGQluPmrnbf/+rT0n02LhwLtm7Va+zrzSu/mAExccZCE1Do+rYMQQoiSKRqbMWOG0qhRI8XGxkbp2rWrsmPHDst7PXv2VIYOHWp5HRsbqwDXPXr27Gkps2nTpmLLXH0cs9msvP3224q3t7diNBqV3r17K0ePHi1T3KmpqQqgpKamlvfSa6ZDvyrKRGdFmVxPUS6cuHn5g8sV5R1XRfl5aKWHVhr5BSblgRl/Ko3HrVKGzv1bMZvNWockhBCiCpX291vTeZBqslo7D9LNKAr8MABObIZW98NjC0ouezEWvuoBuWmg08NrR8HRq6oiLdGxpHT6ffEXeQVmPgxvz6AujbQOSQghRBWp9vMgiRpKp4M+H6oj246sguObii9XkAu/PK0mRwCKGQ78r+rivIFmXk6MvbcFAO+uOsyZSzVv+gIhhBCVSxIkUXZeraDrCHV7zRtgyr++zIZ34NwesHOD7mPUfXsXVVWENzX8jgCCGruRkVvAf5bso8Bk1jokIYQQ1YgkSKJ8er0Bdu5w/gj8M7foe0d+gx1fqtsDZsPto0FvBfHRcP5olYdaHINex7RHArGzNrD9xAWmrj6idUhCCCGqEUmQRPnYucHdb6nbm96HzAvqdkocLH9R3Q4ZBS37goMnNLu8tMu+n6s+1hI08XRg2iOBAHz7Vyy/RJ3ROCIhhBDVhSRIovyChoF3e8hJhU3vqU1tvzyjvm4QBL0nXinbYZD6vO9nMFef5qz7Ovjy8t3NAPi/pfvZHXdJ44iEEEJUB5IgifLTG6Dvh+p21Hz433A4swuMLvDwXLCyuVK2ZV8wOkNqHJzecePjms1qf6WE/ZUW+tXGhLbg3jbe5JnMPP9DFPGp2VVyXiGEENWXJEji1vh3h7YPqaPUDq1Q9z04U51Q8mrWdtDmAXX7Zp21d32rzta9aLA6rUAl0+t1fDaoI618nDifnstz30eRk2+q9PMKIYSoviRBErfunslgZatud33uSiJ0rcJmtoPLIb+EWazTzkHkZHU75RQkHarQUEviYLTimyGdcbO3Zv/ZVMb9b5+s1yaEEHWYJEji1rk2gofnwZ1j4d73Si7X+A5wbqCu5xazrvgyq1+HvPQrr4+urthYb8DP3Z4vBwdhpdexIvocc7acqLJzCyGEqF4kQRIVo1U/6P02WBlLLqPXQ/tH1O19i69//8jvcPhXdUqALs+q+/5dU/Gx3kBIUw8mPtAWgI/WHmHz0aQqPb8QQojqQRIkUbUKm9n+XQtZF6/sz02H38eq2yGj4I4IdfvMP5BxvkpDfKpbY54IboSiwNgl+7iYmVel5xdCCKE9SZBE1fJuo04NYM6Hg8uu7N/0AaSdBdfG0HMcuDQAnw6AUnJzXCWacH8bmns5kpyRy/8t3S/9kYQQoo6RBElUvcCr5kQCdUmSv+eo2/d/Cjb26naLPupzFTezAdhaG/hsUEes9DrWHExg6e6zVR6DEEII7UiCJKpeu4cBnTof0oXj8Osr6jQB7R6+MuM2QMvLCdLxjerit1UdZgMXXr1HXdT2nZUHZVFbIYSoQyRBElXP2RcCeqrbPz0G8XvB1gX6TClazrcTOHpDXgac/Kvq4wSe7xHAbY1cSc8tYOySvZjN0tQmhBB1gSRIQhsdHlOfk/9Vn++ZDI5eRcvo9dD8XnX737VVF9tVrAx6Pn20I/Y2BnacuMjcrbGaxCGEEKJqSYIktNH6frCyU7cbhUCnIcWXa9lXff53dZXMql0cf08H3rqvDQAfrTnK0YT0m3xCCCFETScJktCG0Qm6vwLuTaH/F2ptUXECeoHBCClxcP5IlYZ4tce7+nF3Ky/yTGbGLI4mr6D6LLgrhBCi4kmCJLRz13h4eTfUa1FyGRsHaNJD3a7CWbWvpdPpmBreHjd7aw7Hp/Hp+n81i0UIIUTlkwRJVH8ttRvufzUvJ1umDGwPwJwtx5n86yEKTFKTJIQQtZEkSKL6K5wP6fROyLygaSh92vkScXno/9ytsQyZu1Nm2hZCiFpIEiRR/bk0VGff1mhW7Wu93Ls5c54MwsHGwLbjF+g/4y8OnE3VOiwhhBAVSBIkUTNYmtm064d0tT7tfFg2sjv+HvacTcnm4TnbWBEts20LIURtIQmSqBlaXB7uf2wjFFSPJq0W3k6sGHkHPVvUIyffzCuLonn/N+mXJIQQtYEkSKJmqN8JHLwgLx1ObdU6GgsXe2vmDuvCi72aAvDNn7G8/ss+WdxWCCFqOEmQRM2g10MLbWfVLolBr2Ncn1bMeLwTBr2OpXvO8kXkMa3DEkIIcQskQRI1RwvtZ9W+kf6B9XlvQDsAPtvwL8v2nNE4IiGEEOUlCZKoOQpn1b50Es4f1TqaYj3etRHP9wwA4PVf9vH3CW2nJRBCCFE+kiCJmsPoqCZJAH99pmkoNzIurBX92vuQb1J47ocojp/P0DokIYQQZSQJkqhZeo0DdLBvEZzapnU0xdLrdXz6aEc6+rmSmp3PM/N3cSEjV+uwhBBClIEkSKJmaRAEQUPV7d/GgqlA23hKYGtt4NuhnWnoZsepC1k890MUOfkmrcMSQghRSpIgiZqn90Swc4Okg7DrG62jKZGno5H5T3fB2daKqFOXiPg5muw8SZKEEKImqBYJ0qxZs/D398fW1pbg4GB27txZYtmDBw8SHh6Ov78/Op2O6dOnl+uYvXr1QqfTFXm88MILFXlZorLYu6tJEsCmDyA9Udt4bqCZlxNzngrCSq/j9/0JhE3/g63HkrUOSwghxE1oniAtXryYiIgIJk6cyO7duwkMDCQsLIykpKRiy2dlZREQEMDUqVPx8fG5pWOOGDGC+Ph4y+Ojjz6q8OsTleS2IVD/NshNg/UTtI7mhm5v6sncYV3wdbEl7mIWg7/9m/8s2UtqVr7WoQkhhCiB5gnSp59+yogRI3j66adp06YNc+bMwd7enrlz5xZbvkuXLnz88cc89thjGI3GWzqmvb09Pj4+loezs3OFX5+oJHoD3DcNS4ftk9Vndu3i9GhRj3Wv9mBISGMAlkSdofenW/h9f7zMui2EENWQpglSXl4eUVFRhIaGWvbp9XpCQ0PZvn17pR9zwYIFeHp60q5dO8aPH09WVlb5LkRo4+oO27//p9p22C7kZGvN5Afb8csLITSt50ByRi4vLdjN8z9EcT5dRrkJIUR1ommClJycjMlkwtvbu8h+b29vEhISKvWYTzzxBD/++CObNm1i/Pjx/PDDDzz55JMlHjc3N5e0tLQiD1EN1JAO21fr7O/Oby/fyct3N8NKr2PdoUSe+GYHKVnVYxFeIYQQ1aCJTSvPPfccYWFhtG/fnsGDB/P999+zbNkyjh8/Xmz5KVOm4OLiYnn4+flVccSiWNd12C5fYl3VbK0NRNzbkl9H34G3s5GYpAye/e4fmQpACCGqCU0TJE9PTwwGA4mJRUchJSYmltgBu7KOGRwcDMCxY8UvMjp+/HhSU1Mtj9OnT5crPlEJinTYnqh1NGXS2teZ758JxtnWin9OXWLUwj0UmMxahyWEEHWepgmSjY0NQUFBREZGWvaZzWYiIyMJCQmp0mNGR0cD4OvrW+z7RqMRZ2fnIg9RTVg6bAP7f4aLsdrGU0YtfZz4dmgXbKz0bDicyJvLDkjHbSGE0JjmTWwRERF88803fPfddxw+fJgXX3yRzMxMnn76aQCGDBnC+PHjLeXz8vKIjo4mOjqavLw8zp49S3R0dJGan5sd8/jx47z77rtERUVx8uRJVq5cyZAhQ+jRowcdOnSo2hsgKkaDIGjaGxQz/P2V1tGUWdcm7sx4vBN6HSz+5zSfrv9X65CEEKJuU6qBGTNmKI0aNVJsbGyUrl27Kjt27LC817NnT2Xo0KGW17GxsQpw3aNnz56lPmZcXJzSo0cPxd3dXTEajUqzZs2U//znP0pqamqpY05NTVWAMn1GVLKYDYoy0VlR3q+vKNkpWkdTLgt2nFIaj1ulNB63Svl+W6zW4QghRK1T2t9vnaJIXX55pKWl4eLiQmpqqjS3VReKAl+GwPnDcO97cPtorSMql883xPDZhn/R6eCLxzpxfwdfdDqd1mEJIUStUNrfb82b2ISoMDodhLykbu+YU+3nRSrJy72bMTi4EYoCo3/aw+1TN/LKoj0s+PsUMYnp0j9JCCGqgNQglZPUIFVT+TnwWVvISoaH50K7cK0jKheTWeGt5ftZ8s8ZCsxF/4q6O9jQubEbTb0c8XG2xdvZFl8XW3xcbPF0NGLQS22TEEKUpLS/35IglZMkSNXYpimwZaracfvZSLVmqaJcjIUVo8DaDp74GfSVWwmblVfAnrgUdsZeZGfsRfacvkROfsnTABj0Olp6O/HJo4G09pXvpRBCXEsSpEomCVI1lpEEn7UDUy48sw4aBVfMcY/8BstehNxU9fXw9eDXtWKOXUp5BWYOnEtl96lLnLmUTUJqDglpOSSk5pCUnkNhZZOTrRXfDulMcIBHlcYnhBDVXWl/v62qMCYhqoajF3R4BPb8CDtm3XqCZMqHyMmw7Qv1td4azPlqwlTFCZKNlZ7bGrlxWyO368M0K5xLySbi52h2nbzEU3N38sVjHenTrvi5vQrFp2YTk5hB92ae0jwnhBCXSSdtUTt1G6k+H/4VLp0s/3HS4uG7/leSo24jYcCX6vaR324pxIpm0Ovwc7fnh+HB3NvGm7wCMy8t2M2PO04VWz4pLYeJKw7Q86PNDJm7k4Gzt3HgbGoVRy2EENWTJEiidvJuAwF33drEkSc2w1d3Qtx2MDrDo99Dnw+gRR8w2MCFGDhf/SZ0tLU28OXg23i8ayPMCry1/ACfrf/XMvotOSOX91Yd4s6PNvHd9lPkmcxYG3TsPZ3CAzP/4t1Vh8jIrdoRgMeSMjibkl2l5xRCiBuRPkjlJH2QaoCYDbAgHGycIOIg2LqU/rPRC2HFSDXB8m6nJkceTa+8/+PDcGy9ulDunREVH3sFUBSF6Rti+DwyBoDHu/rham/Dd9tOkpWnLoob1NiN1+5pQVMvR95ddYhV++IB8HG2ZWL/NvRp51OpczDFXcjiw7VH+G1fPM62VqyP6Im3s22lnU8IIaSTdiWTBKkGUBT4shucPwL3vg+3jyrd585Gwdw+YMqDjoPhvk/UUWtX+2curHoVGnaBZzdUfOwV6Mcdp5iw4gBXzxbQoaELEfe0oGeLekUSoC3/nuft5QeIu5gFwF0t63FPGx+S0nNISs8lKS1X3U7LxaDX8eZ9renX/sZ9nIqTmpXPzE0xfLdNrcEq1D+wPjMe71T+ixVCiJuQBKmSSYJUQ0TNh19fARc/eDkaDDcZl5B5Ab7uCamnoeV9MOjH4ofypyfAJy3V7deOgpNPRUdeodYciOe1n/fS2MOBV+9pQWhrrxJrhnLyTczadIw5W46Tb7r5Pw//CWvJS72alqqmKa/AzI87TvHFxhhSsvIBuLO5J+G3NSTi52jMCix4NpjuzTzLdoFCCFFKkiBVMkmQaoj87MsTR16AnuOg1/iS50Uym+DHcDixCdwD4LnNN26W+zYUzuyC+z+Dzs9USvgVqcBkxspQ+m6Hx5IymLExhszcAuo52eLlZMTL2YjX5e3l0WeZt/UkAOG3NeSDge0wWhmKPZbJrLBq3zk+W/8vJy+otVMtvB35v36tLbVYE1cc4Lvtpwio58DqV+4s8VhCCHErZJi/EKA2jd05FtaOhy0fQkoc9P8crIzXl930gZocWdurNUc367PUsp+aIB35rUYkSGVJjgCaeTny+WMlN3cF+rkSUM+Rd1Ye5H+7z3D6YhZzngrC3cHGUqYwMfoiMobj5zMB8HQ08tq9LXgkqGGRmCLubclv+xM4cT6Tb/+MZeRdzcp4hUIIUXFkFJuo/UJegn7TQGeAvT/Bdw9AZnLRMkd+hz+nqdsPzADvtjc/bqv71efYPyAnrWJjriGe6taYucO64GS0YufJizz05VaOn8+gwGRm2Z4z3PPZFl5ZFM3x85m42Fnz2j0t2PyfXjzetdF1CZuLnTVv3tcKgBkbYzh9uR+UEEJoQZrYykma2Gqg4xvh52HqTNiujeDxxep0ABeOw9e9IDcNgl+EvlNLf8wZndXh/g/Pg3YDKyvyai8mMZ1nvtvF6YvZONta4eFoJDZZrTFytbdmxJ0BDAlpjJOt9Q2PoygKj329g79jLxLa2ptvh3auivCFEHVIaX+/pQZJ1B1N71ZHnLk1UZva/nsvHFwOi59SkyO/bnDvu2U7Zqt+6nM1mzSyqjX3dmL5S90JauxGWk4BscmZuNlb85+wlvw17m5G3tXspskRgE6n470B7bDS69hwOJENhxKrIPryk/9fClF7SQ1SOUkNUg2WdRF+HgIn/7yyz9Ebnv+j7KPRTu+E/96jTiT5n+NgZXPzz9RiOfkmZm8+jpOtFY91bYSjsXzdHKesPsxXW07Q0M2O9a/2xM7mSoft7DwTG48k8fuBeBxtrJjQvw0O5TxPeaXl5PPVluN8v+0UXZu4M3lAOxq42t38g0IIzckotkomCVINV5AHv78Gu79X+yYNWwWNby/7ccxmdbh/ZhI8tUytpRK3LDO3gNBPtxCfmsPou5sx6u5mbDl6nlX74tlwONEy0SVAG19n5g7rgo/LzSeYzCswU2A2Y29TvoQqt8DEgh1xzNgYw6XL0xQAONgYGNe3FU8GN0Zfzdezyy0wEZOYQdv6zpU6CagQ1ZUkSJVMEqRaQFHg6Gqwc4PGIeU/zsqXYfd30OVZdVJJUSHWHIjnhR93Y23QYWtlIP2q5U8autlxbxsfVu49S3JGHr4utswd1oXWvsX/XVQUhd/3JzB51UFSs/MZcWcAz/UIKFWzH4DZrPDrvnNMW3eU0xfVJVEC6jnwQo+m/PzPaf45dQmALv5uTBnYgWZejrd49ZVj/5lUIn6OJiYpg2G3+/POA6UYjCBELSMJUiWTBElY/LsOFj4CTvXh1YPFTywpykxRFIbN28WWf88D6vIn93XwpX9gfQIbuqDT6Th9MYth83Zy/HwmjkYrZg2+jZ4t6hU5TmxyJhNWHODPmKIjFz0cbHgltDmPd22EdQlTIKTl5PPHv+eZs+U4B86qIxXrORl5NbQFj3ZWpykwmxUW/H2KqauPkJlnwsag55XQ5jzXI6DE45ZFVl4Bv+2LJzU7H3sbKxyMBuxtrLC3MWBvY8DdwYZG7vY3rA3KN5n5ctNxZmyMoeCqKdWnD+rIgE4NbjlGIWoSSZAqmSRIwiI/Bz5uCnkZMGIjNAjSOqJa41JmHr9EnSHQz5XOjd2Kbb5Kzcrn+R//YceJixj0Ot59sB1PBDciJ9/El5uPM2fzcfJMZmys9LzYsyktvJ2Ytu6oZZRdE08HXg9rSZ92av+zE8mZbDycxMYjSew6edGSUDgarXi+RwDD72xSbBPd2ZRs3ly2n81H1YSuuZcjA29ryD1tvGhaz7HMzVlJaTl8t/0kP+6IIzU7/4Zlm3s58tBtDXiwY4Pr+kIdS0on4ue97DuTCsB97X3xcbHlv3/FYmutZ9lL3UuseROiNpIEqZJJgiSK+HkIHFqhTkrZ+22to6lz8grMvPG/fSzdcxaARzs35O/Yi5y6PGv3nc09mfxgO5p4OgBqjcqinXF8HhlDckYeAO0aOJOeU2D5TKGAeg6EtfXh2Tua4OFYzASjV1EUheXRZ5n866EifZSaeDoQ2tqL0NbeBDV2u+GknUcT0vnmzxOsjD5nWaeusYc9Hf1cycozkZVXQGau+pyVZyIpLbfIenbBTdwZeFsD+rT15ZfdZ/hozRFyC8w421rx7oB2PBBYH7MCT8/fxR//nsffw54Vo+7Axa50zY1C1HSSIFUySZBEEft+hqUjoF5rGLlD62jqJEVR+DwyhukbYiz7vJ2NTLi/Lf3a+xRbg5ORW8DXf5zgmz9OkJ2vdvy2MegJDnDnrpZe3N3KC//LSVVZXMrMY9X+eDYcSmT78QtFEhhXe2uaeDpgb2PAzvpKU5mttYETyZn8cblJESCosRsj7gzgnjbeGEro/J2Wk8+a/Qks3XOGHScuWvbrdGo3O4AeLerxUXiHIh3ZL2Xmcf+Mvzibkk1oay++fqpzte9gLkRFkASpkkmCJIrIvgQfNwNzAYzeDR5NtY6ozlq6+wzTN8RwbxtvxtzTolRTDSSl5bBqXzz1Xe24o7lnuacnKE5GbgF//nue9YcS2Xg0ybJIb0n0OtQaqzsDCGrsVqZznU3JZmX0OZbtOcO/iRnYWRt46/7WPNG1UbEJ4v4zqYTP2UZegZn/hLWU5V1EnSAJUiWTBElc5/sH4cRmuGcydH9F62hENVRgMrP3TCoXMnLJzjdZmsqy80xk5asdvAfe1oDGHmWvtbqaoigcS8rA2c4ab+cbT3/w867TvP6/feh08N3TXelxTSd3IWobWaxWiKrW6n41QTqwVBIkUSwrg77MtULlodPpaO7tVKqyj3bxY8/pS/y08zQvL9rDr6PuwM/dvpIjFKL6k/HIQlSUtg+B3grio+H8Ua2jEaLUJvZvS4eGLqRk5fPM/F3svzziTYi6TBIkISqKgyc07a1u7/tZ21iEKANbawNfDr4NDwcbYpIyeGDWX7y9/ACpN+kvJURtJgmSEBWpw6Pq8/6frwwhqkq5GZAcc/Ny1Y3ZrD6EZhq62fP7K3fyQGB9FAV+2HGKuz7ZzM//nMZslq6qou6RTtrlJJ20RbHysmBac3XSyGfWQqNuVXfunDT4NhQuxKgTVtbvVHXnvhX52TDnDjA6wbMbZSbyamDb8WQmrDjIsaQMQJ1uYPKDbWlb30XjyCpegcnM+YxcEtNySUjNISk9h/ScAuo5GvFxscXHxRZvZ1ucba1uee26nHwT8ak5nEvJ5uylbC5k5lHf1ZYAT0ea1HO44ejJwjiT0/OwtdbjZGuNo60V9tYGmZ6hjKSTthBasLGH1v1h709qM1tVJUiKAstfhOTLfZ+iF2qXIF2MBRtHcCzlaKi4HXDhmLqdfBS8WldebKJUbm/qye8v38m8rbF8HhlD1KlL3PfFX9hY6bHW67Ay6LE26LDS67Ey6LA26DHodVjpdRiufuh05JsV8grM5BWYyDdd3jaZKTCZMStgVpTLD3X0nVkBK70OGys9Ris9RisDNlZ6bAx69HodeQUmcgvMluPk5qvPBp36GWuD+qxu67HW6ykur1EUuJiVR3JGbqkqe+1tDPg42+JsZ42ttR5bawO2VgbLto2VnnyTQoHJTL7JTL5Jufxs5mJmHmdTckjOyL3hObycjATUc6CJpyNGKz0JqTnEp+WQkJrN+fRciqvI0+nA0cYKJ1srjNaGq66vaGF7Gyuc7axwsrXG2dYaJ1sry7WYzQomM5gURd1WFBQF3OytLQmij7OaLNpedY7aTmqQyklqkESJjm+EHx5SF8F97V+wsqn8c/75KUROuvLa3hNeOwqGKv4/UOpZmNkZnBvAqF0U+8t0rQ3vwF+fqdv9pkHXEZUaoiib+NRs3lt1mN/2x2sdSqWx0uvwcjLidTkRcLS14nx6LolpOcSn5tx0qZeysLM20MDNjvqudrjbW3M2JZvY5EzLjO43i9PdwYY8k5n0nAJMGjR9uthZ4+Fog9HKgPFyImq8/LCx0mPQ6zHoQK/TodPpMOivbIMab2HWoSigcDk5M5vJNyuYTAoFZjMFZoUCk8LwO5twV0uvCr0GqUESQitNeoKjN2QkwvFIaNm3cs93fCNsfFfd7jcNNk+BrGSI3QzNQiv33Nc6sQnys9RmvsQD4NP+5p+J/ePK9qltkiBVM74udswafBvvZeaRmVdAweUfsHyTgsms1pIUmNVtk1m5vG22/OhZ6fVFanSu/JCqNUx6nQ795R9R9YdUXQomr8B8XU2RWVEsP8g2l2uXCrcLYyn8XOF2wQ2SCJfL80R5ONjcsJkqO89kSZYycgvIyTeRnW8iN99ETr6ZnHwT+SYzVga1Rs3GoMdKr8PaSq3BcrazpqGbHQ1c7XC1ty62qS41K5/YC5mcOJ9BbHImBWbFUmvj66Imbh6ORsuM6oqikFugJkrpOflk5BaQW2Dm6iMXnsasQGZuAek5BaTl5KvP2epzTr4Jg16HXq9Dr0P9M9Hr0KHjYmYuCWk5JKblEp+aTU6+mdTs/ApNGG+mb3ufKjvXtapFgjRr1iw+/vhjEhISCAwMZMaMGXTt2rXYsgcPHmTChAlERUVx6tQpPvvsM8aMGVPmY+bk5PDaa6+xaNEicnNzCQsL48svv8Tb27uyLlPUFXoDtHsYdsyCfYsrN0FKiYNfhoNihk5PQpdn4fwR2PUt7P+l6hOkk39d2T624eYJUnYKnNtz5fWpbep/K2+xr4eoeG4ONrg5VEFtaDVkZ2PA39OhXMvOlJaLvTUd7V3p6OdaqvI6nU5t5rM2UM/pxmsEVgRFUUjLLiAxPYcLGXnkXU5A1YTUdFXTqdpkqiiXm+wuN9uZFdBx5a/21UliYfOs1eXm26ufA0t5PyqD5gnS4sWLiYiIYM6cOQQHBzN9+nTCwsI4evQoXl7XV6tlZWUREBDAI488wquvvlruY7766qv89ttvLFmyBBcXF0aNGsXAgQPZunVrpV6vqCM6PKomSEdXq52nbSuhGTY/BxY/BdkX1f5G/T5R//Vp/4iaIB1eBfdng7XdzY9VERQFYv+88vpYJNxR/N9Ri1Nb1eTOtRGkJ0BGAlw8IUu1CFHN6HQ6XOytcbG3hjpSj6D5cJFPP/2UESNG8PTTT9OmTRvmzJmDvb09c+fOLbZ8ly5d+Pjjj3nssccwGovPmm92zNTUVP773//y6aefcvfddxMUFMS8efPYtm0bO3bIQqOiAvgGgmcLKMiBw79W/PEVBX5/TZ2U0s4dHv0erC8vKdGwK7g0grx0+HdtxZ+7JJdOQtoZKKzkj9sOuek3/kxh81qze6BBkLp9qo7+J+XYBlj2AmScv3nZW3F8ozpqMH5v5Z5HiBpO0wQpLy+PqKgoQkOvNAPo9XpCQ0PZvn17pR0zKiqK/Pz8ImVatWpFo0aNyn1eIYrQ6a7MibRvccUfP2o+7PkRdHp4eK5aA1NIr4f24er2/iUVf+6SnLxce+QXDG5N1IV7r65RKs6JLepzkx7QuLu6fWpb5cVYna0Zr45+/C2ics/zxzRI2A875lTueYSo4TRNkJKTkzGZTNf1+/H29iYhIaHSjpmQkICNjQ2urq6lPm9ubi5paWlFHkLcUPtH1OfYPyCtgkYAmU2w8xtY/br6uvcEaHpXyeeOWaf286kKhf2P/O+40vfp2IaSy2ckwfnDgO5ygnS7ur8u1iBdPAHJ/6rbh1fCkd8q5zyZF9SaPYDYLdpMZipEDaF5E1tNMWXKFFxcXCwPPz8/rUMS1Z2bP/h1AxQ48MutH+9MFHxzF/w+Fkx50GYAdB9TfFnvtuDVRi1XGU1811KUKwlSkzuLJkgl/QgXNq/5tAd7d/DrCjqD2vE85XTlx1yd/LtOfdZf7hb621i171pFi1mn9vkCSDt7Zf4pIcR1NE2QPD09MRgMJCYmFtmfmJiIj0/5hvaV5pg+Pj7k5eWRkpJS6vOOHz+e1NRUy+P06Tr2D7gon4poZsu6CL+OgW97q/1GbF3gvk/UprUbjfZqd7mZrbzJWWYyfHM3rHv75mUvnlB/cPXWah8o/zvU7ZRT6nvFObFZfW7SQ302Oql9t+BKLUdd8e8a9bnXG+AeAOnnIHJyxZ/n6O9FXxf+GQghrqNpgmRjY0NQUBCRkZGWfWazmcjISEJCQirtmEFBQVhbWxcpc/ToUeLi4ko8r9FoxNnZuchDiJtq+5BaK5CwH5IOX9lvNsOp7fD76zArGL69B5Y+D5unwt7FcHqn2ll3z4/qxItR8wAFAp+AUVHqcH79TWa0bf+w+hz7hzpCrKz+ngNno2D7zJt/vrD2qGFndTZxoyM0vvx3qaRmtsIapIBeV/b5dy96vLogN/3K9bZ5CO6frm7v+lb9HlSU/Bx1ZCFA6wfUZ0mQhCiR5sP8IyIiGDp0KJ07d6Zr165Mnz6dzMxMnn76aQCGDBlCgwYNmDJlCqB2wj506JBl++zZs0RHR+Po6EizZs1KdUwXFxeGDx9OREQE7u7uODs7M3r0aEJCQujWrQrXzhK1n707NL9X/Z/7vsXQog8cXAaHVkD6Nf2Sztzgx7Bea7XWqDCBKA03f7U258xO9ZzdXiz9Z/My1R9oUJtk9v0M3V8uubyl/9GdV/Y1C1WToGMbIPj5ouUvnVRrl/RW0Oiq/5Q07g7bZtStjtrHN4E5X6058mymPjo+CdE/wsqX4fk/KmY29pN/Qn4mOPmqTbOHV6qd6E0FVT/juhA1gOZ/KwYNGsT58+eZMGECCQkJdOzYkTVr1lg6WcfFxaG/avHKc+fO0anTlTWmpk2bxrRp0+jZsyebN28u1TEBPvvsM/R6PeHh4UUmihSiwrV/RE2Q/vrsypIaAEZnaHWf+jDlw6VYdR2zSyfV57SzYOMAPcepyY3BunznPrNTHc1WlgQpeiFkX1JHySlmdXTV7aOLb9JTlCsj2PzvuLK/WSisn6D+COfnXJmGAK7UHjXorNY2FWrUDdCpM3FnJIFjxS4xUC0VTsXQos+Vffe+qza7nT8MWz+Hnv+59fMUNq+17Av1O6pNtTmp6lQRDTvf+vGFqGVkLbZykrXYRKnlZ8NnbSHrgpoUtewHbQdA07vB6gYz4ObnqAnKrdQeZCTBJ61AMcHo3aWbgNFsghm3qYna3W/Blo/BlAvPbVF/WK914bha3mADb8RdmZhSUeDT1mpN2VPLi462+2W42jeqx+tw95tFjze7u7pMySPfqfepNjOb4ZMWkHkehqwo2ty4/xf433D1vr64DTyb39p5Pmuj/lkM/gWa3wOLn1Q78N/9FvSogARMiBqitL/fMopNiMpmbQfPrIUnl8LYGBj4lfq/+BslR6DWuNxq04qj15Uf3QNLS/eZI6vU5MjOHbqNhFb91P17FxVfvrD2qGGXorN263TQtLe6fXU/JEW5qv9Rz+uPV5eG+5/boyZHNk7Q6Pai77ULV2vhTHnw6ytqklNe8dFqcmTjeKVTfOH3onAuKiFEEZIgCVEVPJtDs95Fm5mqSmFn7f0/33zeG0WBrV+o212eVTtcBz5x+fNL1KbAa8UW07xWqFlhgnRlQATnj0BmEljZqUnVterShJExl5vXmt19fTKs08F9n4K1vZos7vm+/Oc5ulp9vrrWMuByjd7pv9U+Z0KIIiRBEqK2a3U/GIzqRIQJ+29cNm4HnP1HLd91hLqv6d3g4AVZyRCzvmj5q+c/Ki5BCuilNhOePwypZ9V9hTUWjboVX4tWWIOUeFCd4qA2Kxzef3X/o6u5NYa7LjdBrp9Y/vth6X/U78o+9wBw8VNrqOratApClIIkSELUdrbO0PLyD3BJzWSFts1QnwMfu9JB2mB1ZT6nvQuLlr9wXF1g1mBUR8xdy979yhprxy/XIsVeTpCKa14D9bwezQFFrd0oSW6G+qgqOWlwLhoO/A/++BiWvQjf9S990+W10s5dXg9Np65FV5LgF8CrLeSkwJYPy36eS6fUPl06PbQIu7Jfp7vyZyDD/YW4jiRIQtQFHQapzzu+hH/mFV8mOeZKTUPIqKLvBT6uPh9dU7QW4+TlvkQNu5TcfHj1rNqmgqtm3O5RcryFtUglzYeUEqfOD/Vp68rvQ/PXZ/BxM5jqB1/3hF+egY3vqcli7B+w6tXyzXodc3n27AZB4Fiv5HIGK+jzgbq98xs4f7Rs5ymspWoUoiasVytsZpMESYjrSIIkRF3Qsh90Hg4osGpM0ekGCm2fpb7foi/Ua1H0PZ926pIg5ny1BqXQjZrXChUmSMc3w7ndkJumDjH37VjyZwqPV1w/pLxM+OkJtdNxbhr8GF7+WpybOfkXbHhH7UgN4FBPXT6m42B19Jd7U7VmZ9c3ZT92ccP7SxLQS/0zVEyw7q2ynae45rVChUlqwn515nQhhIUkSELUBTqdOtHkHZdXit/wjtqnpbDTdsZ5da4jUOc7Kk5hZ+3oy81sinKlg3aTO4v/DED9TmDnBrmp8Oen6j7/O288E3hhDVL8XnWm6UJmMyx7ARL3q8lKy/vUpO2XZ+Dvr0o+XnmY8uG319TtTk/CG6fhP8dg+FoY8KU6NL7nOPX97bPK1tE5P+dKrc3VzV43cu976vItMesg5gaLAF8tO+VKEtuy7/XvO3qBdzt1O1ZGswlxNUmQhKgrdDoInQj3XF7ja+t0tXnIbFJnzS7IUZt7Gt9e/OfbP6LOfH1ut9rMkxyjjkYzGNUJH0uiN1xpyvn38miqGzWvAbg0BNdGao3J1ctt/PGxOgO03hoGLYBBP0CXEYACq19X1y+rqKnddnypjriz94B73lX7cl2rXTi4NVHnuPpnbumPffIvyM8Cp/pqzVxpeDS9MiP52v9Tmytv5tgGMBeAZ8uS58CyDPffXLo4hKgjJEESoq7p/gr0/xzQqWu8/W/4lSaikmbLBrWfTGFn4r0/XZn/yK/rzacvKGxmK9SkhA7aV7t2uP+hlbD5cl+c+z+DRsFq8tXvY7W5C+DPT2DlqNIlDzeSekZdFw/U5OjavjuFDFZw5+Vapq1fqJOCloZl9FrYjRccvlaP/6jzUyUfvbw+300UDu9vVUzzWqHCBOn45opLLoWoBSRBEqIuChoGD89Va2IOLlNrQFwbQ6v+N/5cx8udtfcuvtIk43+D5rVChfMhATh6Q72WN/+MJUHaqvaRWXa59qTbS3DbU1fK6XRq4tD/C3Wk1p4fYfFgyMu6+TlKsma8WsPTKORKB/WSBD4GLo3U2rSo725+bEUpW/+jq9m5Xpl5fNMH6nIwJTHlX5mWobj+R4Uahag1g6lx6nI3QghAEiQh6q52A+HxReqEjQAhI2++aGmLPmDrCunn1GUq4MYdtAs5+Vzp69KkR+lqTQqb+s5GqZ2y87PUprp73i2+fNBQtdnNylatofm61+Vh9GUUs15txtMZ1H5b+pv8M2mwhjtfVbe3Tlf7F91I0mE1GbGyvXlTY3FuG6YuXpx9UV0GpiSntqr9vhzqXZlqoThGxytTNEgzmxAWkiAJUZc1D4Xh6yBsyuVRbjdhZVT73YC6iK2VbekXOu38DKBTR4CVhnsAOPqoExmmxqmvH5l34ySuVT91TTNHb7UZ6pveasdws6l058zPht/HqtvdXgTvtqX7XMfB4NxAHVkX/eONyxY2rzXpqc5UXlZFhv1/BcnHii9X2LzWos+NO8SDLDsiRDEkQRKirvPtACEv3bz2qFDHJ65s+3W9+ZpyhTo/A++kFF209kZ0uiu1SDZOam2XndvNP9eoG7y4HVr3V0e4RU6C+ferEybezF/T1XXonHyh1xulixPUe9B9zJVjFOSVXNbSvHZv6Y9/raZ3q4mPuQDWvQlp8Woz5PGNsG8JbP8SDi5Xy96oea1QYYIUu+XW1nwTohaRBEkIUTYNgi7PdE3p+h8VKktn5ELdXlTnHXrsx9L1Wyrk4AGP/gAPzlIXaI3bBrO7Q/RPJXdEvnD8yvxQfaaA0alssd72lFpzlXoa9pUwY/m5aDhzeVRe81IO7y/Jve+pfYf+XQOftoI5d8APD8HSZ2HteHWGc2v7K8nPjTS4TU1Csy9Bwr5bi6u6MOWrf56LBkN8LbmmqpBwANa+qf59WTlaHbBQR+kURYYtlEdaWhouLi6kpqbi7FzM8F8harPYP9QZuftNU5OR6uxirDp30ukd6uvGd6gTYTp4qfMAOXqrj80fqDUwTe+GJ5eWL6HbNlOt0XHzh1FRV2rl4vepy4QcWaW+bhAEIzbe+rVtfE+d+kCnB3tPcCh81FNft+yjXk9pLHxMnYYhdBLcMebWY9NSwn5Y/tKVZE9/ebThnWOvXxRYQHqiuhj13kXqHGNXMxjV6SXueLXk0Zw1TGl/vyVBKidJkISoQcwmtTZh8xS1WaokBht4aUfJcwbdTF4mTO+gLuw7YI7ah+nqxAid2jk+dBK4+pXvHNfKTQdrh5t3Jr+ZHXNgzTjwC1Y7p9u5q02a1/aTKsiDjET1kZ6g1lQpitoHy6UBODdUf0jLk2DeqoI8+HOaOt2DuUCNv/5tV9YB9G6nTvLpG1j85/Nz1GbG5H/Vmd4bdgZruyoL/4YunVLntTr9t/o9K8gFU676XJCr9tWzc1MHMjS7G3w7lfydMJvVazy9Aw6vUv9joFzup2ewUZtvW4Spk8Ke2qrut3VRk6TgFyr2nhTkqqNpb/X7WwaSIFUySZCEqIHOH1Xnb8o4f/lHPkkdnp+RCNmp0GucOprvVvz1mTpTubUD5BfOrq1TO7f3fL1sTYVVKekIfBl8/X4rWzVZMjqq00FkXbj5sazswLm+OuGnZ3Oo1wq8WqvPDp5Fy5oK1E74F47DhWNqImDvAe5N1ETVPUD9cb6Zc3tg+UhIOqi+bt0f7vtUrU07uEztfJ91QR2deOdr6tQQVjbqOnox69QkNmY95F21ALLeWm1+bHy7Ou2EX3DxE4aWR2ayOgmqjb1a22fvoSaWhX368rPh5FY1KTq2AS7ElO34du5q7WGz3mrsKXFqQnR6p/rISSlavmFXdcqKtg9dqSlSFPWebHjnyn118lXXavRoqt5bB081fhuH0iXFWRchbrs6v9mpbepIUytb9T8TPu0vL2vUAbzalG8QQylIglTJJEESQhQrNx2mt788R5EO2j+s/hhX18SokHJ5NvITm9XYsy+VXNumt1anbnD0Vp8B0s5C6lk14bwRe081WbJxhIvH1SZQc/7NP+PRVK2lsrZTazmsbNVkwsqoJhtR89VaEHsPtem37UNFf7AzzsPvr8GhFeprr7ZqEhe7Ra19KeTkqy6Pc26POirxajq92nzq4qfWALo0uvzsp8787tLwxiMGsy6q02McXKou06MUM7rSxklNUDIS1dntLec2qIMiAnqpTcOGy9duuRc2cPEEHItURyPmpV9/7KtZ2ak1ZI27Q4dHb1xrajapTXAb31P72JV0PAdPNZk1Oql/vkYnNbE2Oqu1XnE74PzhG8dluV49eDRTm0UDB5XuM6UkCVIlkwRJCFGiE1vU//V3eur6hX9rCkVRk73si2qylJuu1ko4+apNOSU1iRTkQto5tXNv6mm11u78EXX+p5QSRhJa2aoL/3o0BbfGaiJx4bj6g3+zhOtqbQeqM6tfW0t1tYPL1DX2rq4J82gOre9XJ0qtf7lpSlHUEY2FNR2ntt58Ik2DUf1R92x++dFCvabzR9UFlU9sKpp01mulPmddUK/52oTJuaFaA9QsFAJ6lq4mDdQO6md2qcnSsQ0QH60ua9MoWB304NdVra0xWJfueIUKctW+h8c2qM3ImcnqQs5XJ3Kl4dkSGodAo9vVUacFuWp/sYT9lx/7riwQ/dDXkiDVNJIgCSFEGeVlXkmY8rMuJ0XN1JqhkhKunDQ1Ubp4XO1MfHW/m8J+OKZ8td/MjZZUuVpmMuz8Wq19ad2/9LV76QlqM2DKaTX5S4lTn1PPqNumG0zvUMi7PbQdoNZwXV1rYzarE3tmXVTjs3NTk6yK6MtVkFd5ndMVRf1zLUyYclLVZsrc9MuPDMhNU6+jYRd15vYbJbCF0hPVZMm3g1pjVoEkQapkkiAJIYSwMJvUJCk5Ru0vlPyvup0coyYEbQaoSVFNrVGsRUr7+13KmeGEEEIIUSK9Qe1Y7t4EuIVJQEW1IRNFCiGEEEJcQxIkIYQQQohrSIIkhBBCCHENSZCEEEIIIa4hCZIQQgghxDUkQRJCCCGEuIYkSEIIIYQQ15AESQghhBDiGpIgCSGEEEJcQxIkIYQQQohrVIsEadasWfj7+2Nra0twcDA7d+68YfklS5bQqlUrbG1tad++Pb///nuR9xMTExk2bBj169fH3t6ePn36EBMTU6RMr1690Ol0RR4vvPBChV+bEEIIIWoezROkxYsXExERwcSJE9m9ezeBgYGEhYWRlJRUbPlt27bx+OOPM3z4cPbs2cOAAQMYMGAABw4cAEBRFAYMGMCJEydYsWIFe/bsoXHjxoSGhpKZmVnkWCNGjCA+Pt7y+Oijjyr9eoUQQghR/ekURVG0DCA4OJguXbowc+ZMAMxmM35+fowePZo33njjuvKDBg0iMzOTVatWWfZ169aNjh07MmfOHP79919atmzJgQMHaNu2reWYPj4+fPDBBzz77LOAWoPUsWNHpk+fXq64S7sasBBCCCGqj9L+fmtag5SXl0dUVBShoaGWfXq9ntDQULZv317sZ7Zv316kPEBYWJilfG5uLgC2trZFjmk0Gvnrr7+KfG7BggV4enrSrl07xo8fT1ZWVoVclxBCCCFqNistT56cnIzJZMLb27vIfm9vb44cOVLsZxISEootn5CQAECrVq1o1KgR48eP56uvvsLBwYHPPvuMM2fOEB8fb/nME088QePGjalfvz779u1j3LhxHD16lKVLlxZ73tzcXEvyBZCamgqomagQQgghaobC3+2bNaBpmiBVBmtra5YuXcrw4cNxd3fHYDAQGhpK3759i9yM5557zrLdvn17fH196d27N8ePH6dp06bXHXfKlClMmjTpuv1+fn6VcyFCCCGEqDTp6em4uLiU+L6mCZKnpycGg4HExMQi+xMTE/Hx8Sn2Mz4+PjctHxQURHR0NKmpqeTl5VGvXj2Cg4Pp3LlzibEEBwcDcOzYsWITpPHjxxMREWF5bTabuXjxIh4eHuh0uptfbCmlpaXh5+fH6dOn63TfJrkPcg9A7kEhuQ9yDwrJfbj1e6AoCunp6dSvX/+G5TRNkGxsbAgKCiIyMpIBAwYAauIRGRnJqFGjiv1MSEgIkZGRjBkzxrJv/fr1hISEXFe2MDOMiYnhn3/+4d133y0xlujoaAB8fX2Lfd9oNGI0Govsc3V1LfF4t8rZ2bnOfvmvJvdB7gHIPSgk90HuQSG5D7d2D25Uc1RI8ya2iIgIhg4dSufOnenatSvTp08nMzOTp59+GoAhQ4bQoEEDpkyZAsArr7xCz549+eSTT7jvvvtYtGgR//zzD19//bXlmEuWLKFevXo0atSI/fv388orrzBgwADuvfdeAI4fP87ChQvp168fHh4e7Nu3j1dffZUePXrQoUOHqr8JQgghhKhWNE+QBg0axPnz55kwYQIJCQl07NiRNWvWWDpix8XFoddfGWx3++23s3DhQt566y3+7//+j+bNm7N8+XLatWtnKRMfH09ERASJiYn4+voyZMgQ3n77bcv7NjY2bNiwwZKM+fn5ER4ezltvvVV1Fy6EEEKIakvzBAlg1KhRJTapbd68+bp9jzzyCI888kiJx3v55Zd5+eWXS3zfz8+PLVu2lDnOqmA0Gpk4ceJ1zXl1jdwHuQcg96CQ3Ae5B4XkPlTdPdB8okghhBBCiOpG86VGhBBCCCGqG0mQhBBCCCGuIQmSEEIIIcQ1JEESQgghhLiGJEjVzKxZs/D398fW1pbg4GB27typdUiV6o8//qB///7Ur18fnU7H8uXLi7yvKAoTJkzA19cXOzs7QkNDiYmJ0SbYSjBlyhS6dOmCk5MTXl5eDBgwgKNHjxYpk5OTw8iRI/Hw8MDR0ZHw8PDrZpOv6WbPnk2HDh0sE7+FhISwevVqy/t14R5ca+rUqeh0uiKT4taF+/DOO++g0+mKPFq1amV5vy7cA4CzZ8/y5JNP4uHhgZ2dHe3bt+eff/6xvF/b/2309/e/7nug0+kYOXIkUDXfA0mQqpHFixcTERHBxIkT2b17N4GBgYSFhZGUlKR1aJUmMzOTwMBAZs2aVez7H330EV988QVz5szh77//xsHBgbCwMHJycqo40sqxZcsWRo4cyY4dO1i/fj35+fnce++9ZGZmWsq8+uqr/PrrryxZsoQtW7Zw7tw5Bg4cqGHUFa9hw4ZMnTqVqKgo/vnnH+6++24efPBBDh48CNSNe3C1Xbt28dVXX103cW1duQ9t27YlPj7e8vjrr78s79WFe3Dp0iW6d++OtbU1q1ev5tChQ3zyySe4ublZytT2fxt37dpV5Duwfv16AMsUP1XyPVBEtdG1a1dl5MiRltcmk0mpX7++MmXKFA2jqjqAsmzZMstrs9ms+Pj4KB9//LFlX0pKimI0GpWffvpJgwgrX1JSkgIoW7ZsURRFvV5ra2tlyZIlljKHDx9WAGX79u1ahVkl3NzclG+//bbO3YP09HSlefPmyvr165WePXsqr7zyiqIodee7MHHiRCUwMLDY9+rKPRg3bpxyxx13lPh+Xfy38ZVXXlGaNm2qmM3mKvseSA1SNZGXl0dUVBShoaGWfXq9ntDQULZv365hZNqJjY0lISGhyD1xcXEhODi41t6T1NRUANzd3QGIiooiPz+/yD1o1aoVjRo1qrX3wGQysWjRIjIzMwkJCalz92DkyJHcd999Ra4X6tZ3ISYmhvr16xMQEMDgwYOJi4sD6s49WLlyJZ07d+aRRx7By8uLTp068c0331jer2v/Nubl5fHjjz/yzDPPoNPpqux7IAlSNZGcnIzJZLIssVLI29ubhIQEjaLSVuF115V7YjabGTNmDN27d7csnZOQkICNjc11CyPXxnuwf/9+HB0dMRqNvPDCCyxbtow2bdrUqXuwaNEidu/ebVl78mp15T4EBwczf/581qxZw+zZs4mNjeXOO+8kPT29ztyDEydOMHv2bJo3b87atWt58cUXefnll/nuu++Auvdv4/Lly0lJSWHYsGFA1f1dqBZLjQgh1JqDAwcOFOlvUZe0bNmS6OhoUlNT+eWXXxg6dGi1XRKoMpw+fZpXXnmF9evXY2trq3U4munbt69lu0OHDgQHB9O4cWN+/vln7OzsNIys6pjNZjp37swHH3wAQKdOnThw4ABz5sxh6NChGkdX9f773//St29f6tevX6XnlRqkasLT0xODwXBdL/zExER8fHw0ikpbhdddF+7JqFGjWLVqFZs2baJhw4aW/T4+PuTl5ZGSklKkfG28BzY2NjRr1oygoCCmTJlCYGAgn3/+eZ25B1FRUSQlJXHbbbdhZWWFlZUVW7Zs4YsvvsDKygpvb+86cR+u5erqSosWLTh27Fid+S74+vrSpk2bIvtat25taWqsS/82njp1ig0bNvDss89a9lXV90ASpGrCxsaGoKAgIiMjLfvMZjORkZGEhIRoGJl2mjRpgo+PT5F7kpaWxt9//11r7omiKIwaNYply5axceNGmjRpUuT9oKAgrK2ti9yDo0ePEhcXV2vuQUnMZjO5ubl15h707t2b/fv3Ex0dbXl07tyZwYMHW7brwn24VkZGBsePH8fX17fOfBe6d+9+3XQf//77L40bNwbqxr+NhebNm4eXlxf33XefZV+VfQ8qrLu3uGWLFi1SjEajMn/+fOXQoUPKc889p7i6uioJCQlah1Zp0tPTlT179ih79uxRAOXTTz9V9uzZo5w6dUpRFEWZOnWq4urqqqxYsULZt2+f8uCDDypNmjRRsrOzNY68Yrz44ouKi4uLsnnzZiU+Pt7yyMrKspR54YUXlEaNGikbN25U/vnnHyUkJEQJCQnRMOqK98YbbyhbtmxRYmNjlX379ilvvPGGotPplHXr1imKUjfuQXGuHsWmKHXjPrz22mvK5s2bldjYWGXr1q1KaGio4unpqSQlJSmKUjfuwc6dOxUrKyvl/fffV2JiYpQFCxYo9vb2yo8//mgpU9v/bVQUdSR3o0aNlHHjxl33XlV8DyRBqmZmzJihNGrUSLGxsVG6du2q7NixQ+uQKtWmTZsU4LrH0KFDFUVRh7O+/fbbire3t2I0GpXevXsrR48e1TboClTctQPKvHnzLGWys7OVl156SXFzc1Ps7e2Vhx56SImPj9cu6ErwzDPPKI0bN1ZsbGyUevXqKb1797YkR4pSN+5Bca5NkOrCfRg0aJDi6+ur2NjYKA0aNFAGDRqkHDt2zPJ+XbgHiqIov/76q9KuXTvFaDQqrVq1Ur7++usi79f2fxsVRVHWrl2rAMVeV1V8D3SKoigVVx8lhBBCCFHzSR8kIYQQQohrSIIkhBBCCHENSZCEEEIIIa4hCZIQQgghxDUkQRJCCCGEuIYkSEIIIYQQ15AESQghhBDiGpIgCSFEBdHpdCxfvlzrMIQQFUASJCFErTBs2DB0Ot11jz59+mgdmhCiBrLSOgAhhKgoffr0Yd68eUX2GY1GjaIRQtRkUoMkhKg1jEYjPj4+RR5ubm6A2vw1e/Zs+vbti52dHQEBAfzyyy9FPr9//37uvvtu7Ozs8PDw4LnnniMjI6NImblz59K2bVuMRiO+vr6MGjWqyPvJyck89NBD2Nvb07x5c1auXFm5Fy2EqBSSIAkh6oy3336b8PBw9u7dy+DBg3nsscc4fPgwAJmZmYSFheHm5sauXbtYsmQJGzZsKJIAzZ49m5EjR/Lcc8+xf/9+Vq5cSbNmzYqcY9KkSTz66KPs27ePfv36MXjwYC5evFil1ymEqAAVuvStEEJoZOjQoYrBYFAcHByKPN5//31FURQFUF544YUinwkODlZefPFFRVEU5euvv1bc3NyUjIwMy/u//fabotfrlYSEBEVRFKV+/frKm2++WWIMgPLWW29ZXmdkZCiAsnr16gq7TiFE1ZA+SEKIWuOuu+5i9uzZRfa5u7tbtkNCQoq8FxISQnR0NACHDx8mMDAQBwcHy/vdu3fHbDZz9OhRdDod586do3fv3jeMoUOHDpZtBwcHnJ2dSUpKKu8lCSE0IgmSEKLWcHBwuK7Jq6LY2dmVqpy1tXWR1zqdDrPZXBkhCSEqkfRBEkLUGTt27LjudevWrQFo3bo1e/fuJTMz0/L+1q1b0ev1tGzZEicnJ/z9/YmMjKzSmIUQ2pAaJCFErZGbm0tCQkKRfVZWVnh6egKwZMkSOnfuzB133MGCBQvYuXMn//3vfwEYPHgwEydOZOjQobzzzjucP3+e0aNH89RTT+Ht7Q3AO++8wwsvvICXlxd9+/YlPT2drVu3Mnr06Kq9UCFEpZMESQhRa6xZswZfX98i+1q2bMmRI0cAdYTZokWLeOmll/D19eWnn36iTZs2ANjb27N27VpeeeUVunTpgr29PeHh4Xz66aeWYw0dOpScnBw+++wzxo4di6enJw8//HDVXaAQosroFEVRtA5CCCEqm06nY9myZQwYMEDrUIQQNYD0QRJCCCGEuIYkSEIIIYQQ15A+SEKIOkF6EwghykJqkIQQQgghriEJkhBCCCHENSRBEkIIIYS4hiRIQgghhBD/324dCwAAAAAM8rfeP4aiaAQJAGAECQBgBAkAYAQJAGAECQBgAiJZ7cRaVxBTAAAAAElFTkSuQmCC\n" }, "metadata": {} } ] }, { "cell_type": "code", "source": [ "pred = model_stability.predict(x_test)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "n3TbLP1VMQ1u", "outputId": "4ec3c5dc-bd9d-4098-fc63-47ca00d4c34f" }, "execution_count": 25, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "816/816 [==============================] - 1s 2ms/step\n" ] } ] }, { "cell_type": "code", "source": [ "plt.plot(pred[0:100], label='pred gas used')\n", "plt.plot(y_test[0:100], label='true gas used')\n", "plt.title('Prediction')\n", "plt.xlabel('0-100 datapoint')\n", "plt.ylabel('gas used')\n", "plt.legend()\n", "plt.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 472 }, "id": "2Q9tgIuLMaEV", "outputId": "ddcdb4e4-2f43-4e47-8b35-f0dfe42dd030" }, "execution_count": 26, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "#Prediction Using Linear Regression" ], "metadata": { "id": "5p1zZuGeRYbc" } }, { "cell_type": "code", "source": [ "reg = linear_model.LinearRegression()\n", "reg.fit(x_train, y_train)\n", "pred = reg.predict(x_test)" ], "metadata": { "id": "J9g7ztBERRmi" }, "execution_count": 27, "outputs": [] }, { "cell_type": "code", "source": [ "plt.plot(pred[0:100], label='pred gas used')\n", "plt.plot(y_test[0:100], label='true gas used')\n", "plt.title('Prediction of NFT-airdrip period using None-NFT-airdrop model')\n", "plt.xlabel('0-100 datapoint')\n", "plt.ylabel('gas used')\n", "plt.legend()\n", "plt.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 472 }, "id": "aHJFj2TORTeV", "outputId": "41c73eeb-db69-461d-85db-6d617b0490b5" }, "execution_count": 28, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": {} } ] } ] }