IndusNet_7 classes_100_100_4th Experiment.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "resized_images_100_100_train_dir='D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Train'\n",
    "resized_images_100_100_test_dir='D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test'\n",
    "resized_images_100_100_validation_dir='D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Validation'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_10 (Conv2D)           (None, 96, 96, 64)        4864      \n",
      "_________________________________________________________________\n",
      "conv2d_11 (Conv2D)           (None, 92, 92, 64)        102464    \n",
      "_________________________________________________________________\n",
      "conv2d_12 (Conv2D)           (None, 90, 90, 128)       73856     \n",
      "_________________________________________________________________\n",
      "conv2d_13 (Conv2D)           (None, 88, 88, 128)       147584    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 44, 44, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 44, 44, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_14 (Conv2D)           (None, 42, 42, 128)       147584    \n",
      "_________________________________________________________________\n",
      "conv2d_15 (Conv2D)           (None, 40, 40, 128)       147584    \n",
      "_________________________________________________________________\n",
      "conv2d_16 (Conv2D)           (None, 38, 38, 256)       295168    \n",
      "_________________________________________________________________\n",
      "conv2d_17 (Conv2D)           (None, 36, 36, 256)       590080    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 18, 18, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 18, 18, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_18 (Conv2D)           (None, 16, 16, 256)       590080    \n",
      "_________________________________________________________________\n",
      "conv2d_19 (Conv2D)           (None, 14, 14, 512)       1180160   \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 7, 7, 512)         0         \n",
      "_________________________________________________________________\n",
      "dropout_6 (Dropout)          (None, 7, 7, 512)         0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 25088)             0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 32)                802848    \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 7)                 231       \n",
      "=================================================================\n",
      "Total params: 4,082,503\n",
      "Trainable params: 4,082,503\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import keras\n",
    "from tensorflow.python.keras import models\n",
    "from tensorflow.python.keras import layers\n",
    "from tensorflow.python.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.python.keras import regularizers\n",
    "from tensorflow.python.keras import optimizers\n",
    "from tensorflow.python.keras.callbacks import ModelCheckpoint\n",
    "\n",
    "indusNetModel=models.Sequential()\n",
    "\n",
    "indusNetModel.add(layers.Conv2D(64, (5, 5), activation='relu', input_shape=(100, 100, 3)))\n",
    "indusNetModel.add(layers.Conv2D(64, (5, 5), activation='relu'))\n",
    "indusNetModel.add(layers.Conv2D(128, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.Conv2D(128, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.MaxPooling2D(pool_size=(2, 2)))\n",
    "indusNetModel.add(layers.Dropout(0.2))\n",
    "\n",
    "indusNetModel.add(layers.Conv2D(128, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.Conv2D(128, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.Conv2D(256, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.Conv2D(256, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.MaxPooling2D(pool_size=(2, 2)))\n",
    "indusNetModel.add(layers.Dropout(0.2))\n",
    "\n",
    "indusNetModel.add(layers.Conv2D(256, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.Conv2D(512, (3, 3), activation='relu'))\n",
    "indusNetModel.add(layers.MaxPooling2D(pool_size=(2, 2)))\n",
    "indusNetModel.add(layers.Dropout(0.2))\n",
    "\n",
    "indusNetModel.add(layers.Flatten())\n",
    "indusNetModel.add(layers.Dense(32, kernel_regularizer=regularizers.l2(0.001), activation='relu'))\n",
    "indusNetModel.add(layers.Dropout(0.2))\n",
    "indusNetModel.add(layers.Dense(7, activation='softmax'))\n",
    "\n",
    "indusNetModel.compile(loss='categorical_crossentropy',optimizer=optimizers.Adam(lr=1e-4),metrics=['categorical_accuracy'])\n",
    "indusNetModel.summary()\n",
    "\n",
    "filepath=\"IndusNet_7_classes_100_100_4th Experiment.h5\"\n",
    "checkpoint = ModelCheckpoint(filepath, save_best_only=True)\n",
    "callbacks_list = [checkpoint]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 2800 images belonging to 7 classes.\n",
      "Found 280 images belonging to 7 classes.\n",
      "Found 560 images belonging to 7 classes.\n"
     ]
    }
   ],
   "source": [
    "batch_size=16\n",
    "\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1./255,\n",
    "    rotation_range=40,\n",
    "    width_shift_range=0.2,\n",
    "    height_shift_range=0.2,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,)\n",
    "\n",
    "# Note that the validation  and test data should not be augmented!\n",
    "validation_datagen=ImageDataGenerator(rescale=1./255)\n",
    "\n",
    "test_datagen = ImageDataGenerator(rescale=1./255)\n",
    "\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "        # This is the target directory\n",
    "        resized_images_100_100_train_dir,\n",
    "        target_size=(100, 100),\n",
    "        batch_size=batch_size)\n",
    "\n",
    "validation_generator = validation_datagen.flow_from_directory(\n",
    "        resized_images_100_100_validation_dir,\n",
    "        target_size=(100, 100),\n",
    "        batch_size=batch_size, shuffle=False)\n",
    "\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "        resized_images_100_100_test_dir,\n",
    "        target_size=(100, 100),\n",
    "        batch_size=batch_size, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_of_training_samples=2800\n",
    "number_of_validation_samples=280\n",
    "number_of_test_samples=560"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Raheel\\Anaconda3\\envs\\EnvName\\lib\\site-packages\\tensorflow\\python\\ops\\math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "Epoch 1/400\n",
      "18/18 [==============================] - 2s 114ms/step - loss: 1.5785 - categorical_accuracy: 0.3107\n",
      "175/175 [==============================] - 45s 258ms/step - loss: 1.8822 - categorical_accuracy: 0.1989 - val_loss: 1.5785 - val_categorical_accuracy: 0.3107\n",
      "Epoch 2/400\n",
      "18/18 [==============================] - 1s 72ms/step - loss: 1.3311 - categorical_accuracy: 0.4536\n",
      "175/175 [==============================] - 40s 227ms/step - loss: 1.5438 - categorical_accuracy: 0.3496 - val_loss: 1.3311 - val_categorical_accuracy: 0.4536\n",
      "Epoch 3/400\n",
      "18/18 [==============================] - 1s 72ms/step - loss: 1.3223 - categorical_accuracy: 0.4607\n",
      "175/175 [==============================] - 40s 228ms/step - loss: 1.3982 - categorical_accuracy: 0.4089 - val_loss: 1.3223 - val_categorical_accuracy: 0.4607\n",
      "Epoch 4/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.2782 - categorical_accuracy: 0.4786\n",
      "175/175 [==============================] - 39s 224ms/step - loss: 1.3714 - categorical_accuracy: 0.4211 - val_loss: 1.2782 - val_categorical_accuracy: 0.4786\n",
      "Epoch 5/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.2855 - categorical_accuracy: 0.4464\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 1.3987 - categorical_accuracy: 0.4125 - val_loss: 1.2855 - val_categorical_accuracy: 0.4464\n",
      "Epoch 6/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.2115 - categorical_accuracy: 0.5464\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 1.3278 - categorical_accuracy: 0.4504 - val_loss: 1.2115 - val_categorical_accuracy: 0.5464\n",
      "Epoch 7/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.1548 - categorical_accuracy: 0.5393\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 1.3062 - categorical_accuracy: 0.4479 - val_loss: 1.1548 - val_categorical_accuracy: 0.5393\n",
      "Epoch 8/400\n",
      "18/18 [==============================] - 1s 73ms/step - loss: 1.2744 - categorical_accuracy: 0.4786\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 1.2623 - categorical_accuracy: 0.5014 - val_loss: 1.2744 - val_categorical_accuracy: 0.4786\n",
      "Epoch 9/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.2095 - categorical_accuracy: 0.5143\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 1.2825 - categorical_accuracy: 0.4779 - val_loss: 1.2095 - val_categorical_accuracy: 0.5143\n",
      "Epoch 10/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.0813 - categorical_accuracy: 0.5571\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 1.2326 - categorical_accuracy: 0.4979 - val_loss: 1.0813 - val_categorical_accuracy: 0.5571\n",
      "Epoch 11/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.0714 - categorical_accuracy: 0.5714\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 1.2240 - categorical_accuracy: 0.5161 - val_loss: 1.0714 - val_categorical_accuracy: 0.5714\n",
      "Epoch 12/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.0775 - categorical_accuracy: 0.5964\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 1.1860 - categorical_accuracy: 0.5250 - val_loss: 1.0775 - val_categorical_accuracy: 0.5964\n",
      "Epoch 13/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 1.0621 - categorical_accuracy: 0.5464ETA: 21s - loss: \n",
      "175/175 [==============================] - 33s 191ms/step - loss: 1.1585 - categorical_accuracy: 0.5271 - val_loss: 1.0621 - val_categorical_accuracy: 0.5464\n",
      "Epoch 14/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.9990 - categorical_accuracy: 0.6036\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 1.1621 - categorical_accuracy: 0.5425 - val_loss: 0.9990 - val_categorical_accuracy: 0.6036\n",
      "Epoch 15/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 1.0520 - categorical_accuracy: 0.5929\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 1.1256 - categorical_accuracy: 0.5529 - val_loss: 1.0520 - val_categorical_accuracy: 0.5929\n",
      "Epoch 16/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.0152 - categorical_accuracy: 0.6000\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 1.1266 - categorical_accuracy: 0.5407 - val_loss: 1.0152 - val_categorical_accuracy: 0.6000\n",
      "Epoch 17/400\n",
      "18/18 [==============================] - 1s 73ms/step - loss: 0.9444 - categorical_accuracy: 0.6250\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 1.0828 - categorical_accuracy: 0.5625 - val_loss: 0.9444 - val_categorical_accuracy: 0.6250\n",
      "Epoch 18/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.9350 - categorical_accuracy: 0.6179\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 1.0666 - categorical_accuracy: 0.5811 - val_loss: 0.9350 - val_categorical_accuracy: 0.6179\n",
      "Epoch 19/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.9695 - categorical_accuracy: 0.6250\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 1.0311 - categorical_accuracy: 0.6025 - val_loss: 0.9695 - val_categorical_accuracy: 0.6250\n",
      "Epoch 20/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.9456 - categorical_accuracy: 0.6357\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 1.0095 - categorical_accuracy: 0.6061 - val_loss: 0.9456 - val_categorical_accuracy: 0.6357\n",
      "Epoch 21/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 1.0133 - categorical_accuracy: 0.6321\n",
      "175/175 [==============================] - 32s 186ms/step - loss: 0.9969 - categorical_accuracy: 0.6125 - val_loss: 1.0133 - val_categorical_accuracy: 0.6321\n",
      "Epoch 22/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.9117 - categorical_accuracy: 0.6750ETA:\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.9803 - categorical_accuracy: 0.6257 - val_loss: 0.9117 - val_categorical_accuracy: 0.6750\n",
      "Epoch 23/400\n",
      "18/18 [==============================] - 1s 64ms/step - loss: 0.8606 - categorical_accuracy: 0.7179\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.9484 - categorical_accuracy: 0.6350 - val_loss: 0.8606 - val_categorical_accuracy: 0.7179\n",
      "Epoch 24/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.8311 - categorical_accuracy: 0.6679\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.9492 - categorical_accuracy: 0.6457 - val_loss: 0.8311 - val_categorical_accuracy: 0.6679\n",
      "Epoch 25/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.8485 - categorical_accuracy: 0.6571\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.9796 - categorical_accuracy: 0.6264 - val_loss: 0.8485 - val_categorical_accuracy: 0.6571\n",
      "Epoch 26/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.9138 - categorical_accuracy: 0.6536A: 0s - loss: 0\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.9613 - categorical_accuracy: 0.6193 - val_loss: 0.9138 - val_categorical_accuracy: 0.6536\n",
      "Epoch 27/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.8088 - categorical_accuracy: 0.6679\n",
      "175/175 [==============================] - 34s 192ms/step - loss: 0.9158 - categorical_accuracy: 0.6554 - val_loss: 0.8088 - val_categorical_accuracy: 0.6679\n",
      "Epoch 28/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.7940 - categorical_accuracy: 0.6893\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.8992 - categorical_accuracy: 0.6621 - val_loss: 0.7940 - val_categorical_accuracy: 0.6893\n",
      "Epoch 29/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.7437 - categorical_accuracy: 0.7143\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.8831 - categorical_accuracy: 0.6632 - val_loss: 0.7437 - val_categorical_accuracy: 0.7143\n",
      "Epoch 30/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6909 - categorical_accuracy: 0.6929\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.8637 - categorical_accuracy: 0.6864 - val_loss: 0.6909 - val_categorical_accuracy: 0.6929\n",
      "Epoch 31/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.7387 - categorical_accuracy: 0.6857\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.8514 - categorical_accuracy: 0.6786 - val_loss: 0.7387 - val_categorical_accuracy: 0.6857\n",
      "Epoch 32/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.6889 - categorical_accuracy: 0.7393\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.8224 - categorical_accuracy: 0.6871 - val_loss: 0.6889 - val_categorical_accuracy: 0.7393\n",
      "Epoch 33/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.7051 - categorical_accuracy: 0.7143\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.8144 - categorical_accuracy: 0.6946 - val_loss: 0.7051 - val_categorical_accuracy: 0.7143\n",
      "Epoch 34/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.7183 - categorical_accuracy: 0.7107\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.8049 - categorical_accuracy: 0.6989 - val_loss: 0.7183 - val_categorical_accuracy: 0.7107\n",
      "Epoch 35/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.6783 - categorical_accuracy: 0.7143\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.7937 - categorical_accuracy: 0.7032 - val_loss: 0.6783 - val_categorical_accuracy: 0.7143\n",
      "Epoch 36/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.7679 - categorical_accuracy: 0.6893\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.7642 - categorical_accuracy: 0.7186 - val_loss: 0.7679 - val_categorical_accuracy: 0.6893\n",
      "Epoch 37/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6630 - categorical_accuracy: 0.7357\n",
      "175/175 [==============================] - 34s 193ms/step - loss: 0.8187 - categorical_accuracy: 0.7086 - val_loss: 0.6630 - val_categorical_accuracy: 0.7357\n",
      "Epoch 38/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.6610 - categorical_accuracy: 0.7036\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.7541 - categorical_accuracy: 0.7232 - val_loss: 0.6610 - val_categorical_accuracy: 0.7036\n",
      "Epoch 39/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.6454 - categorical_accuracy: 0.7571 0s - loss: 0.9917 - cate\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.7487 - categorical_accuracy: 0.7189 - val_loss: 0.6454 - val_categorical_accuracy: 0.7571\n",
      "Epoch 40/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.6801 - categorical_accuracy: 0.7321\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.7333 - categorical_accuracy: 0.7339 - val_loss: 0.6801 - val_categorical_accuracy: 0.7321\n",
      "Epoch 41/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6477 - categorical_accuracy: 0.7500\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.7417 - categorical_accuracy: 0.7382 - val_loss: 0.6477 - val_categorical_accuracy: 0.7500\n",
      "Epoch 42/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.7489 - categorical_accuracy: 0.7000\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.7269 - categorical_accuracy: 0.7404 - val_loss: 0.7489 - val_categorical_accuracy: 0.7000\n",
      "Epoch 43/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.7648 - categorical_accuracy: 0.7214\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.6875 - categorical_accuracy: 0.7564 - val_loss: 0.7648 - val_categorical_accuracy: 0.7214\n",
      "Epoch 44/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6836 - categorical_accuracy: 0.7393\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.7135 - categorical_accuracy: 0.7514 - val_loss: 0.6836 - val_categorical_accuracy: 0.7393\n",
      "Epoch 45/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.7120 - categorical_accuracy: 0.7036\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.6679 - categorical_accuracy: 0.7596 - val_loss: 0.7120 - val_categorical_accuracy: 0.7036\n",
      "Epoch 46/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6367 - categorical_accuracy: 0.7321\n",
      "175/175 [==============================] - 34s 195ms/step - loss: 0.6905 - categorical_accuracy: 0.7607 - val_loss: 0.6367 - val_categorical_accuracy: 0.7321\n",
      "Epoch 47/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.6829 - categorical_accuracy: 0.7536A: 0s - loss: 0.6788 - categorical_ - ETA: 0s - loss: 1.3465 - ca\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.6820 - categorical_accuracy: 0.7704 - val_loss: 0.6829 - val_categorical_accuracy: 0.7536\n",
      "Epoch 48/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.6743 - categorical_accuracy: 0.7536\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.6748 - categorical_accuracy: 0.7604 - val_loss: 0.6743 - val_categorical_accuracy: 0.7536\n",
      "Epoch 49/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.6103 - categorical_accuracy: 0.7821\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.6411 - categorical_accuracy: 0.7750 - val_loss: 0.6103 - val_categorical_accuracy: 0.7821\n",
      "Epoch 50/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.5996 - categorical_accuracy: 0.7643\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.6595 - categorical_accuracy: 0.7718 - val_loss: 0.5996 - val_categorical_accuracy: 0.7643\n",
      "Epoch 51/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.6083 - categorical_accuracy: 0.7571\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.6478 - categorical_accuracy: 0.7657 - val_loss: 0.6083 - val_categorical_accuracy: 0.7571\n",
      "Epoch 52/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.6219 - categorical_accuracy: 0.7536\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.6332 - categorical_accuracy: 0.7739 - val_loss: 0.6219 - val_categorical_accuracy: 0.7536\n",
      "Epoch 53/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.6110 - categorical_accuracy: 0.7679\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.6196 - categorical_accuracy: 0.7789 - val_loss: 0.6110 - val_categorical_accuracy: 0.7679\n",
      "Epoch 54/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.6320 - categorical_accuracy: 0.7429\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.6408 - categorical_accuracy: 0.7771 - val_loss: 0.6320 - val_categorical_accuracy: 0.7429\n",
      "Epoch 55/400\n",
      "18/18 [==============================] - 1s 64ms/step - loss: 0.6435 - categorical_accuracy: 0.7679\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.5991 - categorical_accuracy: 0.7982 - val_loss: 0.6435 - val_categorical_accuracy: 0.7679\n",
      "Epoch 56/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.6083 - categorical_accuracy: 0.7821\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.5953 - categorical_accuracy: 0.7825 - val_loss: 0.6083 - val_categorical_accuracy: 0.7821\n",
      "Epoch 57/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6234 - categorical_accuracy: 0.7821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.6097 - categorical_accuracy: 0.7939 - val_loss: 0.6234 - val_categorical_accuracy: 0.7821\n",
      "Epoch 58/400\n",
      "18/18 [==============================] - 1s 73ms/step - loss: 0.7201 - categorical_accuracy: 0.7607\n",
      "175/175 [==============================] - 34s 191ms/step - loss: 0.5836 - categorical_accuracy: 0.7929 - val_loss: 0.7201 - val_categorical_accuracy: 0.7607\n",
      "Epoch 59/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.6597 - categorical_accuracy: 0.8286\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.5724 - categorical_accuracy: 0.8004 - val_loss: 0.6597 - val_categorical_accuracy: 0.8286\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 60/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6266 - categorical_accuracy: 0.7821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.5682 - categorical_accuracy: 0.8029 - val_loss: 0.6266 - val_categorical_accuracy: 0.7821\n",
      "Epoch 61/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5142 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 34s 195ms/step - loss: 0.5928 - categorical_accuracy: 0.7986 - val_loss: 0.5142 - val_categorical_accuracy: 0.8036\n",
      "Epoch 62/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.5609 - categorical_accuracy: 0.7893\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.5811 - categorical_accuracy: 0.7986 - val_loss: 0.5609 - val_categorical_accuracy: 0.7893\n",
      "Epoch 63/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.6658 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.5436 - categorical_accuracy: 0.8161 - val_loss: 0.6658 - val_categorical_accuracy: 0.8036\n",
      "Epoch 64/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.6231 - categorical_accuracy: 0.7857\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.5261 - categorical_accuracy: 0.8268 - val_loss: 0.6231 - val_categorical_accuracy: 0.7857\n",
      "Epoch 65/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.5630 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.5583 - categorical_accuracy: 0.8021 - val_loss: 0.5630 - val_categorical_accuracy: 0.8036\n",
      "Epoch 66/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.5394 - categorical_accuracy: 0.8143\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.5570 - categorical_accuracy: 0.8082 - val_loss: 0.5394 - val_categorical_accuracy: 0.8143\n",
      "Epoch 67/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.5429 - categorical_accuracy: 0.8250\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.5140 - categorical_accuracy: 0.8225 - val_loss: 0.5429 - val_categorical_accuracy: 0.8250\n",
      "Epoch 68/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.5235 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.5233 - categorical_accuracy: 0.8218 - val_loss: 0.5235 - val_categorical_accuracy: 0.8036\n",
      "Epoch 69/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5640 - categorical_accuracy: 0.8143\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.5179 - categorical_accuracy: 0.8264 - val_loss: 0.5640 - val_categorical_accuracy: 0.8143\n",
      "Epoch 70/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5118 - categorical_accuracy: 0.8143\n",
      "175/175 [==============================] - 34s 195ms/step - loss: 0.4885 - categorical_accuracy: 0.8339 - val_loss: 0.5118 - val_categorical_accuracy: 0.8143\n",
      "Epoch 71/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6445 - categorical_accuracy: 0.7964\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4976 - categorical_accuracy: 0.8296 - val_loss: 0.6445 - val_categorical_accuracy: 0.7964\n",
      "Epoch 72/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5743 - categorical_accuracy: 0.8250\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4676 - categorical_accuracy: 0.8443 - val_loss: 0.5743 - val_categorical_accuracy: 0.8250\n",
      "Epoch 73/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5691 - categorical_accuracy: 0.8214\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4604 - categorical_accuracy: 0.8425 - val_loss: 0.5691 - val_categorical_accuracy: 0.8214\n",
      "Epoch 74/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5744 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.5023 - categorical_accuracy: 0.8296 - val_loss: 0.5744 - val_categorical_accuracy: 0.8036\n",
      "Epoch 75/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5886 - categorical_accuracy: 0.7857\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4696 - categorical_accuracy: 0.8393 - val_loss: 0.5886 - val_categorical_accuracy: 0.7857\n",
      "Epoch 76/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5209 - categorical_accuracy: 0.8250\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.4616 - categorical_accuracy: 0.8382 - val_loss: 0.5209 - val_categorical_accuracy: 0.8250\n",
      "Epoch 77/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5500 - categorical_accuracy: 0.8286\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4649 - categorical_accuracy: 0.8507 - val_loss: 0.5500 - val_categorical_accuracy: 0.8286\n",
      "Epoch 78/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5473 - categorical_accuracy: 0.8357\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4463 - categorical_accuracy: 0.8568 - val_loss: 0.5473 - val_categorical_accuracy: 0.8357\n",
      "Epoch 79/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5367 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.4547 - categorical_accuracy: 0.8436 - val_loss: 0.5367 - val_categorical_accuracy: 0.8036\n",
      "Epoch 80/400\n",
      "18/18 [==============================] - 1s 71ms/step - loss: 0.5615 - categorical_accuracy: 0.8000\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4243 - categorical_accuracy: 0.8582 - val_loss: 0.5615 - val_categorical_accuracy: 0.8000\n",
      "Epoch 81/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4775 - categorical_accuracy: 0.8179\n",
      "175/175 [==============================] - 34s 193ms/step - loss: 0.4450 - categorical_accuracy: 0.8564 - val_loss: 0.4775 - val_categorical_accuracy: 0.8179\n",
      "Epoch 82/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.5536 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.4761 - categorical_accuracy: 0.8404 - val_loss: 0.5536 - val_categorical_accuracy: 0.8036\n",
      "Epoch 83/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.5663 - categorical_accuracy: 0.8250\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.4551 - categorical_accuracy: 0.8479 - val_loss: 0.5663 - val_categorical_accuracy: 0.8250\n",
      "Epoch 84/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.5171 - categorical_accuracy: 0.8321\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.4070 - categorical_accuracy: 0.8611 - val_loss: 0.5171 - val_categorical_accuracy: 0.8321\n",
      "Epoch 85/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4948 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.4096 - categorical_accuracy: 0.8629 - val_loss: 0.4948 - val_categorical_accuracy: 0.8464\n",
      "Epoch 86/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5038 - categorical_accuracy: 0.8250\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.4231 - categorical_accuracy: 0.8575 - val_loss: 0.5038 - val_categorical_accuracy: 0.8250\n",
      "Epoch 87/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5772 - categorical_accuracy: 0.8143\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3952 - categorical_accuracy: 0.8739 - val_loss: 0.5772 - val_categorical_accuracy: 0.8143\n",
      "Epoch 88/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5207 - categorical_accuracy: 0.8071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.4373 - categorical_accuracy: 0.8507 - val_loss: 0.5207 - val_categorical_accuracy: 0.8071\n",
      "Epoch 89/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.7042 - categorical_accuracy: 0.8036\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4193 - categorical_accuracy: 0.8629 - val_loss: 0.7042 - val_categorical_accuracy: 0.8036\n",
      "Epoch 90/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4837 - categorical_accuracy: 0.8536\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4190 - categorical_accuracy: 0.8700 - val_loss: 0.4837 - val_categorical_accuracy: 0.8536\n",
      "Epoch 91/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6108 - categorical_accuracy: 0.8107\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4077 - categorical_accuracy: 0.8582 - val_loss: 0.6108 - val_categorical_accuracy: 0.8107\n",
      "Epoch 92/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4824 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.4045 - categorical_accuracy: 0.8664 - val_loss: 0.4824 - val_categorical_accuracy: 0.8464\n",
      "Epoch 93/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4476 - categorical_accuracy: 0.8393\n",
      "175/175 [==============================] - 34s 195ms/step - loss: 0.3960 - categorical_accuracy: 0.8721 - val_loss: 0.4476 - val_categorical_accuracy: 0.8393\n",
      "Epoch 94/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4341 - categorical_accuracy: 0.8536\n",
      "175/175 [==============================] - 34s 193ms/step - loss: 0.3649 - categorical_accuracy: 0.8757 - val_loss: 0.4341 - val_categorical_accuracy: 0.8536\n",
      "Epoch 95/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5903 - categorical_accuracy: 0.8179\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3676 - categorical_accuracy: 0.8818 - val_loss: 0.5903 - val_categorical_accuracy: 0.8179\n",
      "Epoch 96/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5553 - categorical_accuracy: 0.8393\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3608 - categorical_accuracy: 0.8800 - val_loss: 0.5553 - val_categorical_accuracy: 0.8393\n",
      "Epoch 97/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4581 - categorical_accuracy: 0.8429\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3823 - categorical_accuracy: 0.8711 - val_loss: 0.4581 - val_categorical_accuracy: 0.8429\n",
      "Epoch 98/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4903 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.3731 - categorical_accuracy: 0.8757 - val_loss: 0.4903 - val_categorical_accuracy: 0.8464\n",
      "Epoch 99/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4810 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3997 - categorical_accuracy: 0.8750 - val_loss: 0.4810 - val_categorical_accuracy: 0.8750\n",
      "Epoch 100/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5820 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3544 - categorical_accuracy: 0.8879 - val_loss: 0.5820 - val_categorical_accuracy: 0.8571\n",
      "Epoch 101/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4835 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3530 - categorical_accuracy: 0.8825 - val_loss: 0.4835 - val_categorical_accuracy: 0.8464\n",
      "Epoch 102/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4400 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3600 - categorical_accuracy: 0.8846 - val_loss: 0.4400 - val_categorical_accuracy: 0.8607\n",
      "Epoch 103/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4785 - categorical_accuracy: 0.8429\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3533 - categorical_accuracy: 0.8868 - val_loss: 0.4785 - val_categorical_accuracy: 0.8429\n",
      "Epoch 104/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4925 - categorical_accuracy: 0.8429\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3472 - categorical_accuracy: 0.8843 - val_loss: 0.4925 - val_categorical_accuracy: 0.8429\n",
      "Epoch 105/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5845 - categorical_accuracy: 0.8286\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3470 - categorical_accuracy: 0.8864 - val_loss: 0.5845 - val_categorical_accuracy: 0.8286\n",
      "Epoch 106/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5462 - categorical_accuracy: 0.8286\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3597 - categorical_accuracy: 0.8871 - val_loss: 0.5462 - val_categorical_accuracy: 0.8286\n",
      "Epoch 107/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4136 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 34s 194ms/step - loss: 0.3595 - categorical_accuracy: 0.8807 - val_loss: 0.4136 - val_categorical_accuracy: 0.8643\n",
      "Epoch 108/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4694 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3521 - categorical_accuracy: 0.8921 - val_loss: 0.4694 - val_categorical_accuracy: 0.8500\n",
      "Epoch 109/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4286 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3431 - categorical_accuracy: 0.8968 - val_loss: 0.4286 - val_categorical_accuracy: 0.8750\n",
      "Epoch 110/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4984 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3193 - categorical_accuracy: 0.9071 - val_loss: 0.4984 - val_categorical_accuracy: 0.8500\n",
      "Epoch 111/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4816 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3172 - categorical_accuracy: 0.8954 - val_loss: 0.4816 - val_categorical_accuracy: 0.8571\n",
      "Epoch 112/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4403 - categorical_accuracy: 0.8536\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3523 - categorical_accuracy: 0.8836 - val_loss: 0.4403 - val_categorical_accuracy: 0.8536\n",
      "Epoch 113/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4651 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3316 - categorical_accuracy: 0.8950 - val_loss: 0.4651 - val_categorical_accuracy: 0.8714\n",
      "Epoch 114/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4261 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3290 - categorical_accuracy: 0.8957 - val_loss: 0.4261 - val_categorical_accuracy: 0.8643\n",
      "Epoch 115/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5423 - categorical_accuracy: 0.8393\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3307 - categorical_accuracy: 0.8889 - val_loss: 0.5423 - val_categorical_accuracy: 0.8393\n",
      "Epoch 116/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4466 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3237 - categorical_accuracy: 0.9046 - val_loss: 0.4466 - val_categorical_accuracy: 0.8714\n",
      "Epoch 117/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5645 - categorical_accuracy: 0.8393\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3356 - categorical_accuracy: 0.8946 - val_loss: 0.5645 - val_categorical_accuracy: 0.8393\n",
      "Epoch 118/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5374 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2966 - categorical_accuracy: 0.9100 - val_loss: 0.5374 - val_categorical_accuracy: 0.8464\n",
      "Epoch 119/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4865 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3053 - categorical_accuracy: 0.9025 - val_loss: 0.4865 - val_categorical_accuracy: 0.8607\n",
      "Epoch 120/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4600 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3107 - categorical_accuracy: 0.9000 - val_loss: 0.4600 - val_categorical_accuracy: 0.8464\n",
      "Epoch 121/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4397 - categorical_accuracy: 0.8536\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3000 - categorical_accuracy: 0.9096 - val_loss: 0.4397 - val_categorical_accuracy: 0.8536\n",
      "Epoch 122/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4391 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2845 - categorical_accuracy: 0.9082 - val_loss: 0.4391 - val_categorical_accuracy: 0.8464\n",
      "Epoch 123/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5126 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2868 - categorical_accuracy: 0.9093 - val_loss: 0.5126 - val_categorical_accuracy: 0.8571\n",
      "Epoch 124/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4971 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2700 - categorical_accuracy: 0.9089 - val_loss: 0.4971 - val_categorical_accuracy: 0.8643\n",
      "Epoch 125/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4291 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2873 - categorical_accuracy: 0.9118 - val_loss: 0.4291 - val_categorical_accuracy: 0.8643\n",
      "Epoch 126/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4953 - categorical_accuracy: 0.8393\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.3354 - categorical_accuracy: 0.8979 - val_loss: 0.4953 - val_categorical_accuracy: 0.8393\n",
      "Epoch 127/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4608 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2936 - categorical_accuracy: 0.9050 - val_loss: 0.4608 - val_categorical_accuracy: 0.8750\n",
      "Epoch 128/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4740 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3203 - categorical_accuracy: 0.8996 - val_loss: 0.4740 - val_categorical_accuracy: 0.8464\n",
      "Epoch 129/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4364 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2810 - categorical_accuracy: 0.9089 - val_loss: 0.4364 - val_categorical_accuracy: 0.8750\n",
      "Epoch 130/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4806 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2986 - categorical_accuracy: 0.9086 - val_loss: 0.4806 - val_categorical_accuracy: 0.8643\n",
      "Epoch 131/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5434 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2892 - categorical_accuracy: 0.9093 - val_loss: 0.5434 - val_categorical_accuracy: 0.8643\n",
      "Epoch 132/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4029 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 34s 194ms/step - loss: 0.3007 - categorical_accuracy: 0.9039 - val_loss: 0.4029 - val_categorical_accuracy: 0.8857\n",
      "Epoch 133/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.4259 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.2711 - categorical_accuracy: 0.9186 - val_loss: 0.4259 - val_categorical_accuracy: 0.8786\n",
      "Epoch 134/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4815 - categorical_accuracy: 0.8429\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.2628 - categorical_accuracy: 0.9193 - val_loss: 0.4815 - val_categorical_accuracy: 0.8429\n",
      "Epoch 135/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4336 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2846 - categorical_accuracy: 0.9136 - val_loss: 0.4336 - val_categorical_accuracy: 0.8679\n",
      "Epoch 136/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5669 - categorical_accuracy: 0.8214\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.3000 - categorical_accuracy: 0.9046 - val_loss: 0.5669 - val_categorical_accuracy: 0.8214\n",
      "Epoch 137/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5141 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2586 - categorical_accuracy: 0.9214 - val_loss: 0.5141 - val_categorical_accuracy: 0.8500\n",
      "Epoch 138/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.3769 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 34s 195ms/step - loss: 0.2390 - categorical_accuracy: 0.9243 - val_loss: 0.3769 - val_categorical_accuracy: 0.8750\n",
      "Epoch 139/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5055 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2697 - categorical_accuracy: 0.9175 - val_loss: 0.5055 - val_categorical_accuracy: 0.8714\n",
      "Epoch 140/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5248 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2312 - categorical_accuracy: 0.9289 - val_loss: 0.5248 - val_categorical_accuracy: 0.8500\n",
      "Epoch 141/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4581 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2633 - categorical_accuracy: 0.9218 - val_loss: 0.4581 - val_categorical_accuracy: 0.8786\n",
      "Epoch 142/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5775 - categorical_accuracy: 0.8393\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2796 - categorical_accuracy: 0.9150 - val_loss: 0.5775 - val_categorical_accuracy: 0.8393\n",
      "Epoch 143/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4179 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2803 - categorical_accuracy: 0.9104 - val_loss: 0.4179 - val_categorical_accuracy: 0.8679\n",
      "Epoch 144/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5395 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2386 - categorical_accuracy: 0.9286 - val_loss: 0.5395 - val_categorical_accuracy: 0.8643\n",
      "Epoch 145/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4253 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2523 - categorical_accuracy: 0.9293 - val_loss: 0.4253 - val_categorical_accuracy: 0.8964\n",
      "Epoch 146/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.3486 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 34s 194ms/step - loss: 0.2608 - categorical_accuracy: 0.9193 - val_loss: 0.3486 - val_categorical_accuracy: 0.8786\n",
      "Epoch 147/400\n",
      "18/18 [==============================] - 1s 61ms/step - loss: 0.4472 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.2342 - categorical_accuracy: 0.9329 - val_loss: 0.4472 - val_categorical_accuracy: 0.8750\n",
      "Epoch 148/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.4916 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.2635 - categorical_accuracy: 0.9168 - val_loss: 0.4916 - val_categorical_accuracy: 0.8786\n",
      "Epoch 149/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5168 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.2376 - categorical_accuracy: 0.9296 - val_loss: 0.5168 - val_categorical_accuracy: 0.8643\n",
      "Epoch 150/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 69ms/step - loss: 0.3966 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2279 - categorical_accuracy: 0.9289 - val_loss: 0.3966 - val_categorical_accuracy: 0.8893\n",
      "Epoch 151/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4230 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2495 - categorical_accuracy: 0.9239 - val_loss: 0.4230 - val_categorical_accuracy: 0.8786\n",
      "Epoch 152/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4689 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2634 - categorical_accuracy: 0.9232 - val_loss: 0.4689 - val_categorical_accuracy: 0.8714\n",
      "Epoch 153/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4890 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2226 - categorical_accuracy: 0.9300 - val_loss: 0.4890 - val_categorical_accuracy: 0.8607\n",
      "Epoch 154/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5721 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2518 - categorical_accuracy: 0.9232 - val_loss: 0.5721 - val_categorical_accuracy: 0.8571\n",
      "Epoch 155/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5381 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2180 - categorical_accuracy: 0.9329 - val_loss: 0.5381 - val_categorical_accuracy: 0.8679\n",
      "Epoch 156/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4829 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2597 - categorical_accuracy: 0.9257 - val_loss: 0.4829 - val_categorical_accuracy: 0.8750\n",
      "Epoch 157/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4213 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2485 - categorical_accuracy: 0.9225 - val_loss: 0.4213 - val_categorical_accuracy: 0.8679\n",
      "Epoch 158/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4895 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1987 - categorical_accuracy: 0.9404 - val_loss: 0.4895 - val_categorical_accuracy: 0.8643\n",
      "Epoch 159/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4522 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2660 - categorical_accuracy: 0.9243 - val_loss: 0.4522 - val_categorical_accuracy: 0.8786\n",
      "Epoch 160/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4851 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2198 - categorical_accuracy: 0.9300 - val_loss: 0.4851 - val_categorical_accuracy: 0.8500\n",
      "Epoch 161/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4728 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2273 - categorical_accuracy: 0.9293 - val_loss: 0.4728 - val_categorical_accuracy: 0.8679\n",
      "Epoch 162/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4500 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2286 - categorical_accuracy: 0.9304 - val_loss: 0.4500 - val_categorical_accuracy: 0.8750\n",
      "Epoch 163/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4424 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2493 - categorical_accuracy: 0.9243 - val_loss: 0.4424 - val_categorical_accuracy: 0.8750\n",
      "Epoch 164/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4005 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2395 - categorical_accuracy: 0.9257 - val_loss: 0.4005 - val_categorical_accuracy: 0.8821\n",
      "Epoch 165/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5262 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1995 - categorical_accuracy: 0.9400 - val_loss: 0.5262 - val_categorical_accuracy: 0.8750\n",
      "Epoch 166/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5427 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2111 - categorical_accuracy: 0.9368 - val_loss: 0.5427 - val_categorical_accuracy: 0.8714\n",
      "Epoch 167/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.3989 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2345 - categorical_accuracy: 0.9332 - val_loss: 0.3989 - val_categorical_accuracy: 0.8786\n",
      "Epoch 168/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4983 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2070 - categorical_accuracy: 0.9400 - val_loss: 0.4983 - val_categorical_accuracy: 0.8607\n",
      "Epoch 169/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.3983 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2225 - categorical_accuracy: 0.9336 - val_loss: 0.3983 - val_categorical_accuracy: 0.9000\n",
      "Epoch 170/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4660 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2097 - categorical_accuracy: 0.9318 - val_loss: 0.4660 - val_categorical_accuracy: 0.8643\n",
      "Epoch 171/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4863 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2030 - categorical_accuracy: 0.9436 - val_loss: 0.4863 - val_categorical_accuracy: 0.8643\n",
      "Epoch 172/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.4786 - categorical_accuracy: 0.8929 0s - loss: 0.2493 - cate\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.2053 - categorical_accuracy: 0.9379 - val_loss: 0.4786 - val_categorical_accuracy: 0.8929\n",
      "Epoch 173/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4180 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.2480 - categorical_accuracy: 0.9261 - val_loss: 0.4180 - val_categorical_accuracy: 0.8750\n",
      "Epoch 174/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5662 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2012 - categorical_accuracy: 0.9414 - val_loss: 0.5662 - val_categorical_accuracy: 0.8679\n",
      "Epoch 175/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4852 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1897 - categorical_accuracy: 0.9443 - val_loss: 0.4852 - val_categorical_accuracy: 0.8821\n",
      "Epoch 176/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4884 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2174 - categorical_accuracy: 0.9318 - val_loss: 0.4884 - val_categorical_accuracy: 0.8643\n",
      "Epoch 177/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5349 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1803 - categorical_accuracy: 0.9496 - val_loss: 0.5349 - val_categorical_accuracy: 0.8643\n",
      "Epoch 178/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4158 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2247 - categorical_accuracy: 0.9321 - val_loss: 0.4158 - val_categorical_accuracy: 0.8750\n",
      "Epoch 179/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.3487 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1835 - categorical_accuracy: 0.9461 - val_loss: 0.3487 - val_categorical_accuracy: 0.8893\n",
      "Epoch 180/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 66ms/step - loss: 0.3895 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1951 - categorical_accuracy: 0.9464 - val_loss: 0.3895 - val_categorical_accuracy: 0.8821\n",
      "Epoch 181/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5896 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1811 - categorical_accuracy: 0.9479 - val_loss: 0.5896 - val_categorical_accuracy: 0.8750\n",
      "Epoch 182/400\n",
      "18/18 [==============================] - 1s 71ms/step - loss: 0.4000 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2032 - categorical_accuracy: 0.9400 - val_loss: 0.4000 - val_categorical_accuracy: 0.8964\n",
      "Epoch 183/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.3807 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1957 - categorical_accuracy: 0.9396 - val_loss: 0.3807 - val_categorical_accuracy: 0.8929\n",
      "Epoch 184/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5293 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2193 - categorical_accuracy: 0.9329 - val_loss: 0.5293 - val_categorical_accuracy: 0.8679\n",
      "Epoch 185/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.3857 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2377 - categorical_accuracy: 0.9325 - val_loss: 0.3857 - val_categorical_accuracy: 0.8857\n",
      "Epoch 186/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6405 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1651 - categorical_accuracy: 0.9511 - val_loss: 0.6405 - val_categorical_accuracy: 0.8607\n",
      "Epoch 187/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4953 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2019 - categorical_accuracy: 0.9346 - val_loss: 0.4953 - val_categorical_accuracy: 0.8857\n",
      "Epoch 188/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5230 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1753 - categorical_accuracy: 0.9525 - val_loss: 0.5230 - val_categorical_accuracy: 0.8857\n",
      "Epoch 189/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4308 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1863 - categorical_accuracy: 0.9486 - val_loss: 0.4308 - val_categorical_accuracy: 0.8643\n",
      "Epoch 190/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4613 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2234 - categorical_accuracy: 0.9314 - val_loss: 0.4613 - val_categorical_accuracy: 0.8786\n",
      "Epoch 191/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4724 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2015 - categorical_accuracy: 0.9450 - val_loss: 0.4724 - val_categorical_accuracy: 0.8643\n",
      "Epoch 192/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4241 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1634 - categorical_accuracy: 0.9525 - val_loss: 0.4241 - val_categorical_accuracy: 0.9000\n",
      "Epoch 193/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4463 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1925 - categorical_accuracy: 0.9446 - val_loss: 0.4463 - val_categorical_accuracy: 0.8571\n",
      "Epoch 194/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5351 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1873 - categorical_accuracy: 0.9464 - val_loss: 0.5351 - val_categorical_accuracy: 0.8857\n",
      "Epoch 195/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4489 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.1942 - categorical_accuracy: 0.9436 - val_loss: 0.4489 - val_categorical_accuracy: 0.8679\n",
      "Epoch 196/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4731 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1808 - categorical_accuracy: 0.9471 - val_loss: 0.4731 - val_categorical_accuracy: 0.8929\n",
      "Epoch 197/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6541 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1785 - categorical_accuracy: 0.9471 - val_loss: 0.6541 - val_categorical_accuracy: 0.8643\n",
      "Epoch 198/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6073 - categorical_accuracy: 0.8536\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.2245 - categorical_accuracy: 0.9321 - val_loss: 0.6073 - val_categorical_accuracy: 0.8536\n",
      "Epoch 199/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6752 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1734 - categorical_accuracy: 0.9511 - val_loss: 0.6752 - val_categorical_accuracy: 0.8500\n",
      "Epoch 200/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4626 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1829 - categorical_accuracy: 0.9500 - val_loss: 0.4626 - val_categorical_accuracy: 0.8786\n",
      "Epoch 201/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4349 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1581 - categorical_accuracy: 0.9532 - val_loss: 0.4349 - val_categorical_accuracy: 0.8750\n",
      "Epoch 202/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5140 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1794 - categorical_accuracy: 0.9511 - val_loss: 0.5140 - val_categorical_accuracy: 0.8679\n",
      "Epoch 203/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4324 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1643 - categorical_accuracy: 0.9518 - val_loss: 0.4324 - val_categorical_accuracy: 0.8964\n",
      "Epoch 204/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.3821 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1869 - categorical_accuracy: 0.9468 - val_loss: 0.3821 - val_categorical_accuracy: 0.8893\n",
      "Epoch 205/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4470 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1842 - categorical_accuracy: 0.9475 - val_loss: 0.4470 - val_categorical_accuracy: 0.8857\n",
      "Epoch 206/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5887 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1443 - categorical_accuracy: 0.9614 - val_loss: 0.5887 - val_categorical_accuracy: 0.8714\n",
      "Epoch 207/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5278 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.1737 - categorical_accuracy: 0.9518 - val_loss: 0.5278 - val_categorical_accuracy: 0.8750\n",
      "Epoch 208/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6247 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1646 - categorical_accuracy: 0.9518 - val_loss: 0.6247 - val_categorical_accuracy: 0.8929\n",
      "Epoch 209/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4088 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1922 - categorical_accuracy: 0.9446 - val_loss: 0.4088 - val_categorical_accuracy: 0.8643\n",
      "Epoch 210/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5740 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1516 - categorical_accuracy: 0.9582 - val_loss: 0.5740 - val_categorical_accuracy: 0.8643\n",
      "Epoch 211/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5255 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.1747 - categorical_accuracy: 0.9446 - val_loss: 0.5255 - val_categorical_accuracy: 0.8821\n",
      "Epoch 212/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5946 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1470 - categorical_accuracy: 0.9593 - val_loss: 0.5946 - val_categorical_accuracy: 0.8643\n",
      "Epoch 213/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6265 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1525 - categorical_accuracy: 0.9593 - val_loss: 0.6265 - val_categorical_accuracy: 0.8821\n",
      "Epoch 214/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5142 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1639 - categorical_accuracy: 0.9514 - val_loss: 0.5142 - val_categorical_accuracy: 0.8821\n",
      "Epoch 215/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6225 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1861 - categorical_accuracy: 0.9496 - val_loss: 0.6225 - val_categorical_accuracy: 0.8893\n",
      "Epoch 216/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4961 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1662 - categorical_accuracy: 0.9532 - val_loss: 0.4961 - val_categorical_accuracy: 0.8714\n",
      "Epoch 217/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4059 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1819 - categorical_accuracy: 0.9443 - val_loss: 0.4059 - val_categorical_accuracy: 0.8929\n",
      "Epoch 218/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5287 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1633 - categorical_accuracy: 0.9557 - val_loss: 0.5287 - val_categorical_accuracy: 0.8750\n",
      "Epoch 219/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5778 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1650 - categorical_accuracy: 0.9571 - val_loss: 0.5778 - val_categorical_accuracy: 0.8714\n",
      "Epoch 220/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4992 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1638 - categorical_accuracy: 0.9529 - val_loss: 0.4992 - val_categorical_accuracy: 0.8786\n",
      "Epoch 221/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5266 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1482 - categorical_accuracy: 0.9593 - val_loss: 0.5266 - val_categorical_accuracy: 0.8821\n",
      "Epoch 222/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4723 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1619 - categorical_accuracy: 0.9532 - val_loss: 0.4723 - val_categorical_accuracy: 0.8786\n",
      "Epoch 223/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5554 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1942 - categorical_accuracy: 0.9461 - val_loss: 0.5554 - val_categorical_accuracy: 0.8964\n",
      "Epoch 224/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4876 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1967 - categorical_accuracy: 0.9425 - val_loss: 0.4876 - val_categorical_accuracy: 0.8714\n",
      "Epoch 225/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.3898 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1560 - categorical_accuracy: 0.9554 - val_loss: 0.3898 - val_categorical_accuracy: 0.8893\n",
      "Epoch 226/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4189 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1492 - categorical_accuracy: 0.9607 - val_loss: 0.4189 - val_categorical_accuracy: 0.9071\n",
      "Epoch 227/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5027 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1642 - categorical_accuracy: 0.9564 - val_loss: 0.5027 - val_categorical_accuracy: 0.8714\n",
      "Epoch 228/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4805 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1454 - categorical_accuracy: 0.9582 - val_loss: 0.4805 - val_categorical_accuracy: 0.9000\n",
      "Epoch 229/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5300 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1493 - categorical_accuracy: 0.9564 - val_loss: 0.5300 - val_categorical_accuracy: 0.8750\n",
      "Epoch 230/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.9334 - categorical_accuracy: 0.7893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1574 - categorical_accuracy: 0.9539 - val_loss: 0.9334 - val_categorical_accuracy: 0.7893\n",
      "Epoch 231/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4323 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1787 - categorical_accuracy: 0.9500 - val_loss: 0.4323 - val_categorical_accuracy: 0.8929\n",
      "Epoch 232/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4474 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1389 - categorical_accuracy: 0.9614 - val_loss: 0.4474 - val_categorical_accuracy: 0.8893\n",
      "Epoch 233/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6013 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1410 - categorical_accuracy: 0.9593 - val_loss: 0.6013 - val_categorical_accuracy: 0.8893\n",
      "Epoch 234/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.6464 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1577 - categorical_accuracy: 0.9561 - val_loss: 0.6464 - val_categorical_accuracy: 0.8714\n",
      "Epoch 235/400\n",
      "18/18 [==============================] - 1s 72ms/step - loss: 0.6450 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1578 - categorical_accuracy: 0.9554 - val_loss: 0.6450 - val_categorical_accuracy: 0.8929\n",
      "Epoch 236/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4761 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1593 - categorical_accuracy: 0.9564 - val_loss: 0.4761 - val_categorical_accuracy: 0.8929\n",
      "Epoch 237/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6404 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1499 - categorical_accuracy: 0.9582 - val_loss: 0.6404 - val_categorical_accuracy: 0.8714\n",
      "Epoch 238/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6452 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1321 - categorical_accuracy: 0.9614 - val_loss: 0.6452 - val_categorical_accuracy: 0.8821\n",
      "Epoch 239/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5781 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.1276 - categorical_accuracy: 0.9639 - val_loss: 0.5781 - val_categorical_accuracy: 0.8893\n",
      "Epoch 240/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5632 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1607 - categorical_accuracy: 0.9571 - val_loss: 0.5632 - val_categorical_accuracy: 0.8929\n",
      "Epoch 241/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5484 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1364 - categorical_accuracy: 0.9664 - val_loss: 0.5484 - val_categorical_accuracy: 0.8893\n",
      "Epoch 242/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5933 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1255 - categorical_accuracy: 0.9646 - val_loss: 0.5933 - val_categorical_accuracy: 0.8786\n",
      "Epoch 243/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5057 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1268 - categorical_accuracy: 0.9643 - val_loss: 0.5057 - val_categorical_accuracy: 0.8786\n",
      "Epoch 244/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4221 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1526 - categorical_accuracy: 0.9564 - val_loss: 0.4221 - val_categorical_accuracy: 0.8964\n",
      "Epoch 245/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5584 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1876 - categorical_accuracy: 0.9479 - val_loss: 0.5584 - val_categorical_accuracy: 0.8679\n",
      "Epoch 246/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.6051 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1397 - categorical_accuracy: 0.9632 - val_loss: 0.6051 - val_categorical_accuracy: 0.8714\n",
      "Epoch 247/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5592 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.1353 - categorical_accuracy: 0.9607 - val_loss: 0.5592 - val_categorical_accuracy: 0.8571\n",
      "Epoch 248/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5899 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1226 - categorical_accuracy: 0.9671 - val_loss: 0.5899 - val_categorical_accuracy: 0.8679\n",
      "Epoch 249/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6266 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1281 - categorical_accuracy: 0.9654 - val_loss: 0.6266 - val_categorical_accuracy: 0.8750\n",
      "Epoch 250/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6575 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1271 - categorical_accuracy: 0.9643 - val_loss: 0.6575 - val_categorical_accuracy: 0.8607\n",
      "Epoch 251/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.6432 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1441 - categorical_accuracy: 0.9593 - val_loss: 0.6432 - val_categorical_accuracy: 0.8821\n",
      "Epoch 252/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.6955 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1553 - categorical_accuracy: 0.9568 - val_loss: 0.6955 - val_categorical_accuracy: 0.8643\n",
      "Epoch 253/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.9420 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1294 - categorical_accuracy: 0.9693 - val_loss: 0.9420 - val_categorical_accuracy: 0.8571\n",
      "Epoch 254/400\n",
      "18/18 [==============================] - ETA: 0s - loss: 0.5842 - categorical_accuracy: 0.88 - 1s 67ms/step - loss: 0.5554 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1556 - categorical_accuracy: 0.9554 - val_loss: 0.5554 - val_categorical_accuracy: 0.8857\n",
      "Epoch 255/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5688 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1387 - categorical_accuracy: 0.9600 - val_loss: 0.5688 - val_categorical_accuracy: 0.8786\n",
      "Epoch 256/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4179 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1239 - categorical_accuracy: 0.9636 - val_loss: 0.4179 - val_categorical_accuracy: 0.9071\n",
      "Epoch 257/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5043 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1506 - categorical_accuracy: 0.9575 - val_loss: 0.5043 - val_categorical_accuracy: 0.8786\n",
      "Epoch 258/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6470 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1254 - categorical_accuracy: 0.9704 - val_loss: 0.6470 - val_categorical_accuracy: 0.8750\n",
      "Epoch 259/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.7347 - categorical_accuracy: 0.8464\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1722 - categorical_accuracy: 0.9468 - val_loss: 0.7347 - val_categorical_accuracy: 0.8464\n",
      "Epoch 260/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5068 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1678 - categorical_accuracy: 0.9554 - val_loss: 0.5068 - val_categorical_accuracy: 0.8750\n",
      "Epoch 261/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5848 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1233 - categorical_accuracy: 0.9664 - val_loss: 0.5848 - val_categorical_accuracy: 0.8786\n",
      "Epoch 262/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5207 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1362 - categorical_accuracy: 0.9646 - val_loss: 0.5207 - val_categorical_accuracy: 0.8893\n",
      "Epoch 263/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.6953 - categorical_accuracy: 0.8214\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1527 - categorical_accuracy: 0.9618 - val_loss: 0.6953 - val_categorical_accuracy: 0.8214\n",
      "Epoch 264/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6442 - categorical_accuracy: 0.8571\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1228 - categorical_accuracy: 0.9693 - val_loss: 0.6442 - val_categorical_accuracy: 0.8571\n",
      "Epoch 265/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6642 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1215 - categorical_accuracy: 0.9689 - val_loss: 0.6642 - val_categorical_accuracy: 0.8821\n",
      "Epoch 266/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5328 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1496 - categorical_accuracy: 0.9604 - val_loss: 0.5328 - val_categorical_accuracy: 0.8893\n",
      "Epoch 267/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5983 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1284 - categorical_accuracy: 0.9679 - val_loss: 0.5983 - val_categorical_accuracy: 0.8893\n",
      "Epoch 268/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5393 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1237 - categorical_accuracy: 0.9693 - val_loss: 0.5393 - val_categorical_accuracy: 0.8786\n",
      "Epoch 269/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5590 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1270 - categorical_accuracy: 0.9632 - val_loss: 0.5590 - val_categorical_accuracy: 0.8929\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 270/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6291 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1158 - categorical_accuracy: 0.9704 - val_loss: 0.6291 - val_categorical_accuracy: 0.8857\n",
      "Epoch 271/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5457 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1393 - categorical_accuracy: 0.9614 - val_loss: 0.5457 - val_categorical_accuracy: 0.8786\n",
      "Epoch 272/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5019 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0980 - categorical_accuracy: 0.9725 - val_loss: 0.5019 - val_categorical_accuracy: 0.9036\n",
      "Epoch 273/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5076 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1273 - categorical_accuracy: 0.9643 - val_loss: 0.5076 - val_categorical_accuracy: 0.8964\n",
      "Epoch 274/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4405 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1408 - categorical_accuracy: 0.9629 - val_loss: 0.4405 - val_categorical_accuracy: 0.8893\n",
      "Epoch 275/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4670 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1420 - categorical_accuracy: 0.9632 - val_loss: 0.4670 - val_categorical_accuracy: 0.8750\n",
      "Epoch 276/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5747 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1212 - categorical_accuracy: 0.9682 - val_loss: 0.5747 - val_categorical_accuracy: 0.8964\n",
      "Epoch 277/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5947 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1327 - categorical_accuracy: 0.9654 - val_loss: 0.5947 - val_categorical_accuracy: 0.8929\n",
      "Epoch 278/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5363 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1373 - categorical_accuracy: 0.9632 - val_loss: 0.5363 - val_categorical_accuracy: 0.9036\n",
      "Epoch 279/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4537 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1039 - categorical_accuracy: 0.9711 - val_loss: 0.4537 - val_categorical_accuracy: 0.8857\n",
      "Epoch 280/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5460 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1286 - categorical_accuracy: 0.9636 - val_loss: 0.5460 - val_categorical_accuracy: 0.8786\n",
      "Epoch 281/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4200 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1245 - categorical_accuracy: 0.9696 - val_loss: 0.4200 - val_categorical_accuracy: 0.8821\n",
      "Epoch 282/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5334 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1093 - categorical_accuracy: 0.9696 - val_loss: 0.5334 - val_categorical_accuracy: 0.8929\n",
      "Epoch 283/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5269 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.1369 - categorical_accuracy: 0.9636 - val_loss: 0.5269 - val_categorical_accuracy: 0.8750\n",
      "Epoch 284/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4589 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1358 - categorical_accuracy: 0.9650 - val_loss: 0.4589 - val_categorical_accuracy: 0.8821\n",
      "Epoch 285/400\n",
      "18/18 [==============================] - 1s 72ms/step - loss: 0.5741 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1199 - categorical_accuracy: 0.9707 - val_loss: 0.5741 - val_categorical_accuracy: 0.8893\n",
      "Epoch 286/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4915 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1198 - categorical_accuracy: 0.9693 - val_loss: 0.4915 - val_categorical_accuracy: 0.9071\n",
      "Epoch 287/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4704 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1197 - categorical_accuracy: 0.9671 - val_loss: 0.4704 - val_categorical_accuracy: 0.9036\n",
      "Epoch 288/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6258 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1090 - categorical_accuracy: 0.9704 - val_loss: 0.6258 - val_categorical_accuracy: 0.8786\n",
      "Epoch 289/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4867 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1194 - categorical_accuracy: 0.9668 - val_loss: 0.4867 - val_categorical_accuracy: 0.8893\n",
      "Epoch 290/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4493 - categorical_accuracy: 0.9214\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1116 - categorical_accuracy: 0.9725 - val_loss: 0.4493 - val_categorical_accuracy: 0.9214\n",
      "Epoch 291/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.7094 - categorical_accuracy: 0.8607\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1011 - categorical_accuracy: 0.9757 - val_loss: 0.7094 - val_categorical_accuracy: 0.8607\n",
      "Epoch 292/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5157 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1318 - categorical_accuracy: 0.9693 - val_loss: 0.5157 - val_categorical_accuracy: 0.8964\n",
      "Epoch 293/400\n",
      "18/18 [==============================] - 1s 64ms/step - loss: 0.5355 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.1135 - categorical_accuracy: 0.9693 - val_loss: 0.5355 - val_categorical_accuracy: 0.8857\n",
      "Epoch 294/400\n",
      "18/18 [==============================] - 1s 64ms/step - loss: 0.5482 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.1597 - categorical_accuracy: 0.9536 - val_loss: 0.5482 - val_categorical_accuracy: 0.8679\n",
      "Epoch 295/400\n",
      "18/18 [==============================] - 1s 62ms/step - loss: 0.6443 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.1232 - categorical_accuracy: 0.9689 - val_loss: 0.6443 - val_categorical_accuracy: 0.8750\n",
      "Epoch 296/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.4837 - categorical_accuracy: 0.8964A: 1s - l - ETA: 0s - loss: 0.4714 - categorical_accuracy: 0.\n",
      "175/175 [==============================] - 33s 186ms/step - loss: 0.1192 - categorical_accuracy: 0.9689 - val_loss: 0.4837 - val_categorical_accuracy: 0.8964\n",
      "Epoch 297/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.5854 - categorical_accuracy: 0.9036 0s - loss: 0.6054 - categorica\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.0870 - categorical_accuracy: 0.9800 - val_loss: 0.5854 - val_categorical_accuracy: 0.9036\n",
      "Epoch 298/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.5027 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.1442 - categorical_accuracy: 0.9636 - val_loss: 0.5027 - val_categorical_accuracy: 0.9036\n",
      "Epoch 299/400\n",
      "18/18 [==============================] - 1s 64ms/step - loss: 0.4865 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.0975 - categorical_accuracy: 0.9739 - val_loss: 0.4865 - val_categorical_accuracy: 0.9000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 300/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.4515 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.1020 - categorical_accuracy: 0.9779 - val_loss: 0.4515 - val_categorical_accuracy: 0.8893\n",
      "Epoch 301/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.4750 - categorical_accuracy: 0.8857 0s - loss: 0.3762 - cate\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.1162 - categorical_accuracy: 0.9711 - val_loss: 0.4750 - val_categorical_accuracy: 0.8857\n",
      "Epoch 302/400\n",
      "18/18 [==============================] - 1s 63ms/step - loss: 0.5528 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 187ms/step - loss: 0.0902 - categorical_accuracy: 0.9761 - val_loss: 0.5528 - val_categorical_accuracy: 0.8929\n",
      "Epoch 303/400\n",
      "18/18 [==============================] - 1s 65ms/step - loss: 0.5912 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 188ms/step - loss: 0.1290 - categorical_accuracy: 0.9618 - val_loss: 0.5912 - val_categorical_accuracy: 0.8929\n",
      "Epoch 304/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5951 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1284 - categorical_accuracy: 0.9632 - val_loss: 0.5951 - val_categorical_accuracy: 0.8857\n",
      "Epoch 305/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6102 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 34s 191ms/step - loss: 0.1231 - categorical_accuracy: 0.9657 - val_loss: 0.6102 - val_categorical_accuracy: 0.8821\n",
      "Epoch 306/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.3974 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 34s 192ms/step - loss: 0.1170 - categorical_accuracy: 0.9700 - val_loss: 0.3974 - val_categorical_accuracy: 0.8929\n",
      "Epoch 307/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6104 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1138 - categorical_accuracy: 0.9682 - val_loss: 0.6104 - val_categorical_accuracy: 0.8786\n",
      "Epoch 308/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6542 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1127 - categorical_accuracy: 0.9675 - val_loss: 0.6542 - val_categorical_accuracy: 0.8786\n",
      "Epoch 309/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4963 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1208 - categorical_accuracy: 0.9679 - val_loss: 0.4963 - val_categorical_accuracy: 0.8964\n",
      "Epoch 310/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5739 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0845 - categorical_accuracy: 0.9800 - val_loss: 0.5739 - val_categorical_accuracy: 0.8893\n",
      "Epoch 311/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5847 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1000 - categorical_accuracy: 0.9743 - val_loss: 0.5847 - val_categorical_accuracy: 0.8714\n",
      "Epoch 312/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6870 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1267 - categorical_accuracy: 0.9664 - val_loss: 0.6870 - val_categorical_accuracy: 0.8679\n",
      "Epoch 313/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5417 - categorical_accuracy: 0.8821A: 0s - loss: 0.1\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1253 - categorical_accuracy: 0.9675 - val_loss: 0.5417 - val_categorical_accuracy: 0.8821\n",
      "Epoch 314/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5523 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0798 - categorical_accuracy: 0.9829 - val_loss: 0.5523 - val_categorical_accuracy: 0.8964\n",
      "Epoch 315/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4576 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1502 - categorical_accuracy: 0.9625 - val_loss: 0.4576 - val_categorical_accuracy: 0.8786\n",
      "Epoch 316/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4489 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1175 - categorical_accuracy: 0.9732 - val_loss: 0.4489 - val_categorical_accuracy: 0.9071\n",
      "Epoch 317/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4115 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1210 - categorical_accuracy: 0.9700 - val_loss: 0.4115 - val_categorical_accuracy: 0.9036\n",
      "Epoch 318/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.6581 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0847 - categorical_accuracy: 0.9786 - val_loss: 0.6581 - val_categorical_accuracy: 0.8679\n",
      "Epoch 319/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6722 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1164 - categorical_accuracy: 0.9721 - val_loss: 0.6722 - val_categorical_accuracy: 0.8821\n",
      "Epoch 320/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5405 - categorical_accuracy: 0.9143\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0852 - categorical_accuracy: 0.9786 - val_loss: 0.5405 - val_categorical_accuracy: 0.9143\n",
      "Epoch 321/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.7247 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1113 - categorical_accuracy: 0.9700 - val_loss: 0.7247 - val_categorical_accuracy: 0.8643\n",
      "Epoch 322/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6146 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0985 - categorical_accuracy: 0.9743 - val_loss: 0.6146 - val_categorical_accuracy: 0.8714\n",
      "Epoch 323/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5176 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1047 - categorical_accuracy: 0.9707 - val_loss: 0.5176 - val_categorical_accuracy: 0.9071\n",
      "Epoch 324/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4799 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1037 - categorical_accuracy: 0.9725 - val_loss: 0.4799 - val_categorical_accuracy: 0.8964\n",
      "Epoch 325/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4823 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1490 - categorical_accuracy: 0.9611 - val_loss: 0.4823 - val_categorical_accuracy: 0.8786\n",
      "Epoch 326/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4737 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1139 - categorical_accuracy: 0.9664 - val_loss: 0.4737 - val_categorical_accuracy: 0.9000\n",
      "Epoch 327/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5982 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0852 - categorical_accuracy: 0.9821 - val_loss: 0.5982 - val_categorical_accuracy: 0.8929\n",
      "Epoch 328/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5277 - categorical_accuracy: 0.9107\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0998 - categorical_accuracy: 0.9707 - val_loss: 0.5277 - val_categorical_accuracy: 0.9107\n",
      "Epoch 329/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5806 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1038 - categorical_accuracy: 0.9764 - val_loss: 0.5806 - val_categorical_accuracy: 0.8929\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 330/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.6050 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1087 - categorical_accuracy: 0.9732 - val_loss: 0.6050 - val_categorical_accuracy: 0.8857\n",
      "Epoch 331/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4992 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1083 - categorical_accuracy: 0.9732 - val_loss: 0.4992 - val_categorical_accuracy: 0.8929\n",
      "Epoch 332/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4806 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0792 - categorical_accuracy: 0.9793 - val_loss: 0.4806 - val_categorical_accuracy: 0.8857\n",
      "Epoch 333/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4911 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1171 - categorical_accuracy: 0.9671 - val_loss: 0.4911 - val_categorical_accuracy: 0.8857\n",
      "Epoch 334/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5653 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0967 - categorical_accuracy: 0.9757 - val_loss: 0.5653 - val_categorical_accuracy: 0.8857\n",
      "Epoch 335/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5442 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1036 - categorical_accuracy: 0.9721 - val_loss: 0.5442 - val_categorical_accuracy: 0.8893\n",
      "Epoch 336/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5351 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0988 - categorical_accuracy: 0.9736 - val_loss: 0.5351 - val_categorical_accuracy: 0.8893\n",
      "Epoch 337/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.8161 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0900 - categorical_accuracy: 0.9789 - val_loss: 0.8161 - val_categorical_accuracy: 0.8500\n",
      "Epoch 338/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5830 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1017 - categorical_accuracy: 0.9736 - val_loss: 0.5830 - val_categorical_accuracy: 0.8786\n",
      "Epoch 339/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5585 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0804 - categorical_accuracy: 0.9807 - val_loss: 0.5585 - val_categorical_accuracy: 0.9000\n",
      "Epoch 340/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6067 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1065 - categorical_accuracy: 0.9729 - val_loss: 0.6067 - val_categorical_accuracy: 0.8893\n",
      "Epoch 341/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4504 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1038 - categorical_accuracy: 0.9736 - val_loss: 0.4504 - val_categorical_accuracy: 0.8893\n",
      "Epoch 342/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5831 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.0939 - categorical_accuracy: 0.9743 - val_loss: 0.5831 - val_categorical_accuracy: 0.8857\n",
      "Epoch 343/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4396 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0905 - categorical_accuracy: 0.9775 - val_loss: 0.4396 - val_categorical_accuracy: 0.8786\n",
      "Epoch 344/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4908 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0895 - categorical_accuracy: 0.9754 - val_loss: 0.4908 - val_categorical_accuracy: 0.9000\n",
      "Epoch 345/400\n",
      "18/18 [==============================] - 1s 71ms/step - loss: 0.4840 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0883 - categorical_accuracy: 0.9757 - val_loss: 0.4840 - val_categorical_accuracy: 0.9036\n",
      "Epoch 346/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4213 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1162 - categorical_accuracy: 0.9693 - val_loss: 0.4213 - val_categorical_accuracy: 0.8893\n",
      "Epoch 347/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6077 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0756 - categorical_accuracy: 0.9796 - val_loss: 0.6077 - val_categorical_accuracy: 0.8786\n",
      "Epoch 348/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4439 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0776 - categorical_accuracy: 0.9779 - val_loss: 0.4439 - val_categorical_accuracy: 0.9071\n",
      "Epoch 349/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4942 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1042 - categorical_accuracy: 0.9729 - val_loss: 0.4942 - val_categorical_accuracy: 0.8893\n",
      "Epoch 350/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4397 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1254 - categorical_accuracy: 0.9643 - val_loss: 0.4397 - val_categorical_accuracy: 0.8893\n",
      "Epoch 351/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4324 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1041 - categorical_accuracy: 0.9707 - val_loss: 0.4324 - val_categorical_accuracy: 0.9000\n",
      "Epoch 352/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6596 - categorical_accuracy: 0.8714\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.0757 - categorical_accuracy: 0.9836 - val_loss: 0.6596 - val_categorical_accuracy: 0.8714\n",
      "Epoch 353/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.3834 - categorical_accuracy: 0.9214\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1140 - categorical_accuracy: 0.9714 - val_loss: 0.3834 - val_categorical_accuracy: 0.9214\n",
      "Epoch 354/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4553 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0961 - categorical_accuracy: 0.9779 - val_loss: 0.4553 - val_categorical_accuracy: 0.8929\n",
      "Epoch 355/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5911 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0971 - categorical_accuracy: 0.9725 - val_loss: 0.5911 - val_categorical_accuracy: 0.8893\n",
      "Epoch 356/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4432 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0938 - categorical_accuracy: 0.9793 - val_loss: 0.4432 - val_categorical_accuracy: 0.8821\n",
      "Epoch 357/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4996 - categorical_accuracy: 0.9107\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1024 - categorical_accuracy: 0.9768 - val_loss: 0.4996 - val_categorical_accuracy: 0.9107\n",
      "Epoch 358/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5385 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0858 - categorical_accuracy: 0.9771 - val_loss: 0.5385 - val_categorical_accuracy: 0.8893\n",
      "Epoch 359/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5260 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0932 - categorical_accuracy: 0.9764 - val_loss: 0.5260 - val_categorical_accuracy: 0.8821\n",
      "Epoch 360/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 68ms/step - loss: 0.3941 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1223 - categorical_accuracy: 0.9657 - val_loss: 0.3941 - val_categorical_accuracy: 0.8964\n",
      "Epoch 361/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5581 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0775 - categorical_accuracy: 0.9807 - val_loss: 0.5581 - val_categorical_accuracy: 0.8821\n",
      "Epoch 362/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5707 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1120 - categorical_accuracy: 0.9696 - val_loss: 0.5707 - val_categorical_accuracy: 0.8893\n",
      "Epoch 363/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.6509 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0802 - categorical_accuracy: 0.9750 - val_loss: 0.6509 - val_categorical_accuracy: 0.8929\n",
      "Epoch 364/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5748 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1128 - categorical_accuracy: 0.9721 - val_loss: 0.5748 - val_categorical_accuracy: 0.8857\n",
      "Epoch 365/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4592 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.0913 - categorical_accuracy: 0.9771 - val_loss: 0.4592 - val_categorical_accuracy: 0.8893\n",
      "Epoch 366/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5839 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 189ms/step - loss: 0.0799 - categorical_accuracy: 0.9786 - val_loss: 0.5839 - val_categorical_accuracy: 0.8893\n",
      "Epoch 367/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.5400 - categorical_accuracy: 0.8750\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0995 - categorical_accuracy: 0.9725 - val_loss: 0.5400 - val_categorical_accuracy: 0.8750\n",
      "Epoch 368/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4790 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0903 - categorical_accuracy: 0.9771 - val_loss: 0.4790 - val_categorical_accuracy: 0.9071\n",
      "Epoch 369/400\n",
      "18/18 [==============================] - 1s 71ms/step - loss: 0.6308 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.0850 - categorical_accuracy: 0.9786 - val_loss: 0.6308 - val_categorical_accuracy: 0.8821\n",
      "Epoch 370/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4699 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1071 - categorical_accuracy: 0.9736 - val_loss: 0.4699 - val_categorical_accuracy: 0.8964\n",
      "Epoch 371/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4445 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0878 - categorical_accuracy: 0.9750 - val_loss: 0.4445 - val_categorical_accuracy: 0.8929\n",
      "Epoch 372/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5972 - categorical_accuracy: 0.8679\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0868 - categorical_accuracy: 0.9804 - val_loss: 0.5972 - val_categorical_accuracy: 0.8679\n",
      "Epoch 373/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4121 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1202 - categorical_accuracy: 0.9693 - val_loss: 0.4121 - val_categorical_accuracy: 0.8964\n",
      "Epoch 374/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5008 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1179 - categorical_accuracy: 0.9711 - val_loss: 0.5008 - val_categorical_accuracy: 0.8929\n",
      "Epoch 375/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.8308 - categorical_accuracy: 0.8536\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0936 - categorical_accuracy: 0.9793 - val_loss: 0.8308 - val_categorical_accuracy: 0.8536\n",
      "Epoch 376/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.6643 - categorical_accuracy: 0.8500\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.1062 - categorical_accuracy: 0.9757 - val_loss: 0.6643 - val_categorical_accuracy: 0.8500\n",
      "Epoch 377/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4362 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0814 - categorical_accuracy: 0.9818 - val_loss: 0.4362 - val_categorical_accuracy: 0.9000\n",
      "Epoch 378/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5861 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0740 - categorical_accuracy: 0.9807 - val_loss: 0.5861 - val_categorical_accuracy: 0.8929\n",
      "Epoch 379/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5636 - categorical_accuracy: 0.8893 0s - loss: 0.6413 - categori\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0926 - categorical_accuracy: 0.9807 - val_loss: 0.5636 - val_categorical_accuracy: 0.8893\n",
      "Epoch 380/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.4623 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0843 - categorical_accuracy: 0.9804 - val_loss: 0.4623 - val_categorical_accuracy: 0.9036\n",
      "Epoch 381/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5302 - categorical_accuracy: 0.8821\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0836 - categorical_accuracy: 0.9736 - val_loss: 0.5302 - val_categorical_accuracy: 0.8821\n",
      "Epoch 382/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6419 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0803 - categorical_accuracy: 0.9796 - val_loss: 0.6419 - val_categorical_accuracy: 0.8929\n",
      "Epoch 383/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.6572 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0762 - categorical_accuracy: 0.9821 - val_loss: 0.6572 - val_categorical_accuracy: 0.8857\n",
      "Epoch 384/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5626 - categorical_accuracy: 0.9071\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1016 - categorical_accuracy: 0.9707 - val_loss: 0.5626 - val_categorical_accuracy: 0.9071\n",
      "Epoch 385/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4998 - categorical_accuracy: 0.9107\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0718 - categorical_accuracy: 0.9839 - val_loss: 0.4998 - val_categorical_accuracy: 0.9107\n",
      "Epoch 386/400\n",
      "18/18 [==============================] - 1s 70ms/step - loss: 0.4872 - categorical_accuracy: 0.8893\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0892 - categorical_accuracy: 0.9732 - val_loss: 0.4872 - val_categorical_accuracy: 0.8893\n",
      "Epoch 387/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5072 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0884 - categorical_accuracy: 0.9782 - val_loss: 0.5072 - val_categorical_accuracy: 0.8929\n",
      "Epoch 388/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5843 - categorical_accuracy: 0.8643\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1008 - categorical_accuracy: 0.9739 - val_loss: 0.5843 - val_categorical_accuracy: 0.8643\n",
      "Epoch 389/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4443 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1153 - categorical_accuracy: 0.9700 - val_loss: 0.4443 - val_categorical_accuracy: 0.9000\n",
      "Epoch 390/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5033 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0644 - categorical_accuracy: 0.9846 - val_loss: 0.5033 - val_categorical_accuracy: 0.8964\n",
      "Epoch 391/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.4471 - categorical_accuracy: 0.8786\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0885 - categorical_accuracy: 0.9789 - val_loss: 0.4471 - val_categorical_accuracy: 0.8786\n",
      "Epoch 392/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.5770 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0815 - categorical_accuracy: 0.9800 - val_loss: 0.5770 - val_categorical_accuracy: 0.8929\n",
      "Epoch 393/400\n",
      "18/18 [==============================] - 1s 68ms/step - loss: 0.4861 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0792 - categorical_accuracy: 0.9800 - val_loss: 0.4861 - val_categorical_accuracy: 0.8929\n",
      "Epoch 394/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5802 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.0946 - categorical_accuracy: 0.9725 - val_loss: 0.5802 - val_categorical_accuracy: 0.8929\n",
      "Epoch 395/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.5184 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.1014 - categorical_accuracy: 0.9732 - val_loss: 0.5184 - val_categorical_accuracy: 0.8964\n",
      "Epoch 396/400\n",
      "18/18 [==============================] - 1s 67ms/step - loss: 0.5153 - categorical_accuracy: 0.8964\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0816 - categorical_accuracy: 0.9796 - val_loss: 0.5153 - val_categorical_accuracy: 0.8964\n",
      "Epoch 397/400\n",
      "18/18 [==============================] - 1s 66ms/step - loss: 0.5852 - categorical_accuracy: 0.9000\n",
      "175/175 [==============================] - 33s 190ms/step - loss: 0.0929 - categorical_accuracy: 0.9786 - val_loss: 0.5852 - val_categorical_accuracy: 0.9000\n",
      "Epoch 398/400\n",
      "18/18 [==============================] - 1s 71ms/step - loss: 0.5578 - categorical_accuracy: 0.8929\n",
      "175/175 [==============================] - 33s 191ms/step - loss: 0.0750 - categorical_accuracy: 0.9796 - val_loss: 0.5578 - val_categorical_accuracy: 0.8929\n",
      "Epoch 399/400\n",
      "18/18 [==============================] - 1s 69ms/step - loss: 0.4546 - categorical_accuracy: 0.9036\n",
      "175/175 [==============================] - 34s 192ms/step - loss: 0.0952 - categorical_accuracy: 0.9739 - val_loss: 0.4546 - val_categorical_accuracy: 0.9036\n",
      "Epoch 400/400\n",
      "18/18 [==============================] - 1s 80ms/step - loss: 0.5047 - categorical_accuracy: 0.8857\n",
      "175/175 [==============================] - 37s 212ms/step - loss: 0.0975 - categorical_accuracy: 0.9754 - val_loss: 0.5047 - val_categorical_accuracy: 0.8857\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "\n",
    "history = indusNetModel.fit_generator(\n",
    "      train_generator,\n",
    "      steps_per_epoch=math.ceil(number_of_training_samples//batch_size),\n",
    "      epochs=400,\n",
    "      callbacks=callbacks_list,\n",
    "      validation_data=validation_generator,\n",
    "      validation_steps=math.ceil(number_of_validation_samples//batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "acc=history.history['categorical_accuracy']\n",
    "val_acc=history.history['val_categorical_accuracy']\n",
    "loss=history.history['loss']\n",
    "val_loss=history.history['val_loss']\n",
    "epochs=range(1,len(acc)+1)\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.python.keras.models import load_model\n",
    "\n",
    "best_model = load_model('IndusNet_7_classes_100_100_4th Experiment.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "35/35 [==============================] - 3s 77ms/step - loss: 0.4592 - categorical_accuracy: 0.8750\n",
      "Test-set accuracy: 87.50%\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.python.keras.models import load_model\n",
    "\n",
    "steps_test=int(number_of_test_samples/batch_size)\n",
    "result = best_model.evaluate_generator(test_generator, steps=steps_test,verbose=1)\n",
    "print(\"Test-set accuracy: {0:.2%}\".format(result[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of test examples:  (560, 100, 100, 3)\n",
      "Total number of labels: (560, 1)\n"
     ]
    }
   ],
   "source": [
    "# Preparing test data\n",
    "import glob\n",
    "from pathlib import Path\n",
    "from tensorflow.python.keras.utils import to_categorical\n",
    "import cv2\n",
    "\n",
    "apple_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\apple')\n",
    "banana_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\banana')\n",
    "orange_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\orange')\n",
    "watermelon_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\watermelon')\n",
    "mango_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\mango')\n",
    "grape_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\grape')\n",
    "strawberry_cases_dir = Path('D:\\\\TensorFlow Programs\\\\fruit_image_classification_Indus Fruit Dataset\\\\Indus Fruit Dataset_Resized Images_100_100\\\\Test\\\\strawberry')\n",
    "\n",
    "apple_cases = apple_cases_dir.glob('*.jpg')\n",
    "banana_cases = banana_cases_dir.glob('*.jpg')\n",
    "orange_cases = orange_cases_dir.glob('*.jpg')\n",
    "watermelon_cases = watermelon_cases_dir.glob('*.jpg')\n",
    "mango_cases = mango_cases_dir.glob('*.jpg')\n",
    "grape_cases = grape_cases_dir.glob('*.jpg')\n",
    "strawberry_cases = strawberry_cases_dir.glob('*.jpg')\n",
    "\n",
    "test_data = []\n",
    "test_labels = []\n",
    "\n",
    "for img in apple_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [0]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "\n",
    "for img in banana_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [1]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "\n",
    "for img in grape_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [2]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "    \n",
    "for img in mango_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [3]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "\n",
    "for img in orange_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [4]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "\n",
    "for img in strawberry_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [5]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "\n",
    "for img in watermelon_cases:\n",
    "    img = cv2.imread(str(img))\n",
    "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "    img = img.astype(np.float32)/255.\n",
    "    label = [6]\n",
    "    test_data.append(img)\n",
    "    test_labels.append(label)\n",
    "\n",
    "test_data = np.array(test_data)\n",
    "test_labels = np.array(test_labels)\n",
    "test_labels_one_hot_encoding=to_categorical(test_labels)\n",
    "\n",
    "print(\"Total number of test examples: \", test_data.shape)\n",
    "print(\"Total number of labels:\", test_labels.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "560/560 [==============================] - 3s 6ms/sample - loss: 0.4592 - categorical_accuracy: 0.8750\n",
      "Loss on test set:  0.4591952153614589\n",
      "Accuracy on test set:  0.875\n"
     ]
    }
   ],
   "source": [
    "# Evaluation on test dataset\n",
    "\n",
    "test_loss, test_score = best_model.evaluate(test_data, test_labels_one_hot_encoding, batch_size=8)\n",
    "print(\"Loss on test set: \", test_loss)\n",
    "print(\"Accuracy on test set: \", test_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "560/560 [==============================] - 3s 5ms/sample\n",
      "(560, 1)\n",
      "(560, 1)\n"
     ]
    }
   ],
   "source": [
    "# Get predictions\n",
    "preds = best_model.predict(test_data, batch_size=16,verbose=1)\n",
    "preds=np.around(preds)\n",
    "orig_test_labels=test_labels\n",
    "predicted_labels=np.argmax(preds,axis=1)\n",
    "predicted_labels=np.reshape(predicted_labels,(560,1))\n",
    "print(predicted_labels.shape)\n",
    "print(orig_test_labels.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 3600x2400 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "# Get the confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from mlxtend.plotting import plot_confusion_matrix\n",
    "\n",
    "cm  = confusion_matrix(orig_test_labels, predicted_labels)\n",
    "plt.figure(dpi=600)\n",
    "plot_confusion_matrix(cm,figsize=(12,10), hide_ticks=True,cmap=plt.cm.Blues)\n",
    "plt.xticks(range(7), ['apple', 'banana','grape','mango','orange','strawberry','watermelon'], fontsize=12)\n",
    "plt.yticks(range(7), ['apple', 'banana','grape','mango','orange','strawberry','watermelon'], fontsize=12)\n",
    "plt.xlabel('Predicted Label',fontsize=14)\n",
    "plt.ylabel('True Label',fontsize=14)\n",
    "plt.savefig('IndusNet_7 classes_100_100_4th Experiment.jpeg',dpi=1200)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}