receipt_whisperer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "g0lzPVtAkK-L",
    "outputId": "7b8e5a07-2801-463d-f371-4a9f80709473"
   },
   "outputs": [],
   "source": [
    "import tensorflow\n",
    "from tensorflow.keras.layers import Reshape\n",
    "from keras.layers import Input, Embedding, Reshape, LSTM, Dense, Flatten, Bidirectional, Concatenate\n",
    "from keras.models import Model\n",
    "import os\n",
    "import numpy as np\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "import re\n",
    "import pickle\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "dPoU8viTkYAK"
   },
   "outputs": [],
   "source": [
    "batch_size = 16\n",
    "max_sequence_length = 384\n",
    "max_word_length = 20\n",
    "embedding_size = 128\n",
    "lstm_units = 256\n",
    "num_classes = 14\n",
    "\n",
    "directory_train='./training-data'\n",
    "directory_test='./testing-data'\n",
    "directory_all='./all-data'\n",
    "model_path='bi_best_model.keras' #modelin konumu\n",
    "model_test_path='./deneme-veriler'#modelin testi için kullanılacak verinin konumu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "EoSXRADHCYxw"
   },
   "outputs": [],
   "source": [
    "train_test_texts=[]\n",
    "\n",
    "for filename in os.listdir(directory_all):\n",
    "    if filename.endswith('.txt'):\n",
    "        file_path = os.path.join(directory_all, filename)\n",
    "        with open(file_path, 'r', encoding='utf-8') as file:\n",
    "            content = file.read()\n",
    "            content = re.sub(r'<start_\".*?\">', '', content)\n",
    "            content = re.sub(r'<end_\".*?\">', '', content)\n",
    "            txt_data = content.split()\n",
    "            train_test_texts.extend(txt_data)\n",
    "\n",
    "tokenizer = Tokenizer(char_level=True, lower=False)\n",
    "tokenizer.fit_on_texts(train_test_texts)\n",
    "\n",
    "vocab_len = len(tokenizer.word_index) + 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'0': 1, 'A': 2, '1': 3, 'E': 4, 'T': 5, 'I': 6, '2': 7, 'R': 8, 'N': 9, 'L': 10, 'K': 11, '*': 12, 'S': 13, 'O': 14, 'M': 15, '5': 16, ':': 17, '.': 18, '3': 19, '8': 20, '4': 21, '9': 22, 'D': 23, '6': 24, 'U': 25, 'İ': 26, '7': 27, ',': 28, 'B': 29, 'C': 30, 'Y': 31, 'P': 32, 'V': 33, 'e': 34, 'a': 35, 'i': 36, 'H': 37, 'G': 38, 'Z': 39, 'r': 40, 'F': 41, '/': 42, 'Ş': 43, 'n': 44, 'o': 45, 'l': 46, '%': 47, 't': 48, 's': 49, 'Ü': 50, '-': 51, 'k': 52, 'd': 53, 'm': 54, 'u': 55, 'X': 56, 'y': 57, 'c': 58, 'g': 59, 'h': 60, 'z': 61, '#': 62, 'ı': 63, 'b': 64, 'w': 65, 'Ç': 66, 'Ğ': 67, 'Ö': 68, 'p': 69, 'v': 70, 'ş': 71, 'W': 72, 'ü': 73, ')': 74, '(': 75, 'J': 76, 'f': 77, 'x': 78, \"'\": 79, '+': 80, '$': 81, 'ğ': 82, '»': 83, 'ç': 84, '!': 85, 'Q': 86, '=': 87, 'ö': 88, 'Í': 89, 'ж': 90, '&': 91, 'j': 92, '\"': 93, '>': 94, '\\\\': 95, 'Т': 96, 'О': 97, 'Р': 98, 'q': 99, '×': 100, '@': 101, '<': 102, ';': 103, 'Ø': 104, 'К': 105, '|': 106, 'Á': 107, 'Ú': 108, 'Ș': 109, '[': 110, '_': 111, '•': 112, '€': 113, 'É': 114, 'Č': 115, 'Ł': 116, 'ł': 117, 'Ž': 118, 'à': 119, 'Ì': 120, '※': 121, '}': 122, 'ä': 123, 'ó': 124, 'č': 125, 'Ó': 126, '?': 127, 'Ė': 128, '¥': 129, 'Ē': 130, 'ć': 131, 'Ī': 132, 'Ā': 133, ']': 134, 'Å': 135, 'о': 136, 'á': 137, 'ś': 138, 'ø': 139, '到': 140, 'У': 141, '米': 142, '…': 143, 'Š': 144, '^': 145, 'À': 146, '✰': 147, 'Ữ': 148, 'Î': 149, '⁹': 150, 'é': 151, '{': 152, 'Ệ': 153, 'и': 154, 'শ': 155, 'Ř': 156, '三': 157, '风': 158, '紫': 159, '头': 160, 'ń': 161, 'Ś': 162, '·': 163, 'Ő': 164, 'Ị': 165, '`': 166, 'Ý': 167, '¨': 168, '®': 169, 'ə': 170, 'Đ': 171, '₁': 172, 'Ļ': 173, 'Ģ': 174, 'ž': 175, 'ě': 176, '▸': 177, '„': 178, 'Ż': 179, 'å': 180, 'л': 181, 'с': 182, 'ē': 183, 'õ': 184, 'Ể': 185, 'Ů': 186, '₂': 187, 'ș': 188, 'š': 189, 'û': 190, 'Ć': 191, 'Ð': 192, 'Ä': 193, 'ộ': 194, '¡': 195}\n"
     ]
    }
   ],
   "source": [
    "#SAVE TOKENIZER\n",
    "with open('tokenizer.pickle', 'wb') as handle:\n",
    "    pickle.dump(tokenizer, handle, protocol=pickle.HIGHEST_PROTOCOL)\n",
    "\n",
    "print(tokenizer.word_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Xdd7tSD_kcQs",
    "outputId": "fc5c41a4-7644-40ff-e95d-f1c605d091e2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(547, 384, 20)\n",
      "(547, 384, 20)\n"
     ]
    }
   ],
   "source": [
    "def process_text_files(directory):\n",
    "    all_texts = []\n",
    "    all_texts_extend = []\n",
    "\n",
    "    # Iterate over all files in the directory\n",
    "    for filename in os.listdir(directory):\n",
    "        if filename.endswith('.txt'):\n",
    "            file_path = os.path.join(directory, filename)\n",
    "            with open(file_path, 'r', encoding='utf-8') as file:\n",
    "                content = file.read()\n",
    "                content = re.sub(r'<start_\".*?\">', '', content)\n",
    "                content = re.sub(r'<end_\".*?\">', '', content)\n",
    "                txt_data = content.split()\n",
    "                all_texts.append(txt_data)\n",
    "                all_texts_extend.extend(txt_data)\n",
    "\n",
    "    #tokenizer = Tokenizer(char_level=True, lower=False)\n",
    "    #tokenizer.fit_on_texts(all_texts_extend)\n",
    "\n",
    "    all_sequences=[]\n",
    "    for i in range(len(all_texts)):\n",
    "        all_sequences.append(tokenizer.texts_to_sequences(all_texts[i]))\n",
    "        \n",
    "\n",
    "    max_word_length = 20\n",
    "    trunc_type = 'post'\n",
    "    padding_type = 'post'\n",
    "\n",
    "    # Pad each word sequence to the length of the longest word\n",
    "    all_padded_sequences=[]\n",
    "    for i in range(len(all_sequences)):\n",
    "        padded_sequence = pad_sequences(all_sequences[i], maxlen=max_word_length, padding=padding_type, truncating=trunc_type)\n",
    "        all_padded_sequences.append(padded_sequence)\n",
    "\n",
    "    return all_padded_sequences, tokenizer.word_index\n",
    "\n",
    "\n",
    "#   TRAIN\n",
    "directory_path = directory_train\n",
    "padded_sequences, word_index = process_text_files(directory_path)\n",
    "padded_arrays = []\n",
    "for arr in padded_sequences:\n",
    "    pad_width = ((0, max_sequence_length - len(arr)), (0, 0))\n",
    "    padded_arr = np.pad(arr, pad_width, mode='constant', constant_values=0)\n",
    "    padded_arrays.append(padded_arr)\n",
    "\n",
    "#  TEST\n",
    "directory_path_test = directory_test\n",
    "padded_sequences_test, word_index_test = process_text_files(directory_path_test)\n",
    "padded_arrays_test = []\n",
    "for arr in padded_sequences_test:\n",
    "    pad_width = ((0, max_sequence_length - len(arr)), (0, 0))\n",
    "    padded_arr = np.pad(arr, pad_width, mode='constant', constant_values=0)\n",
    "    padded_arrays_test.append(padded_arr)\n",
    "\n",
    "\n",
    "np.set_printoptions(threshold=np.inf)\n",
    "\n",
    "#son array'i numpy array'ine çevir\n",
    "x_train=np.array(padded_arrays)\n",
    "x_test=np.array(padded_arrays_test)\n",
    "\n",
    "print(x_train.shape)\n",
    "print(x_test.shape)\n",
    "\n",
    "np.set_printoptions(threshold=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "dVT6bwgCkeWQ",
    "outputId": "badc6c95-dd66-41b3-981d-43b9ee7f1084"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(547, 384, 14)\n",
      "(547, 384, 14)\n"
     ]
    }
   ],
   "source": [
    "global wordsList,wordsIndextList\n",
    "\n",
    "wordsList = []\n",
    "wordsIndextList = []\n",
    "max_sequence_length = 384\n",
    "\n",
    "\n",
    "label_list = [\"<start_\\\"company_name\\\">\",\n",
    "              \"<end_\\\"company_name\\\">\",\n",
    "                  \"<start_\\\"date\\\">\",\n",
    "                  \"<end_\\\"date\\\">\",\n",
    "                  \"<start_\\\"time\\\">\",\n",
    "                  \"<end_\\\"time\\\">\",\n",
    "                  \"<start_\\\"receipt_number\\\">\",\n",
    "                  \"<end_\\\"receipt_number\\\">\",\n",
    "                  \"<start_\\\"tax\\\">\",\n",
    "                  \"<end_\\\"tax\\\">\",\n",
    "                  \"<start_\\\"amount\\\">\",\n",
    "                  \"<end_\\\"amount\\\">\"]\n",
    "\n",
    "label_list2= [\"Pad\",\"Others\",\"B_Comp\",\"l_Comp\",\"B_Date\",\"l_Date\",\"B_Time\",\"l_Time\",\n",
    "                  \"B_Receipt\",\"l_Receipt\",\"B_Tax\",\"l_Tax\",\"B_Amount\",\"l_Amount\"]\n",
    "\n",
    "\n",
    "\n",
    "def one_hot_encode_index_list():\n",
    "    oneHotEncodedList = []\n",
    "    unique_labels_len = len(sorted(set(label_list2)))\n",
    "    for i in wordsIndextList:\n",
    "        bitlist = list(0 for i in range(unique_labels_len))\n",
    "        bitlist[i] = 1\n",
    "        oneHotEncodedList.append(bitlist)\n",
    "\n",
    "    for i in  range(max_sequence_length-len(oneHotEncodedList)):\n",
    "        oneHotEncodedList.append([1,0,0,0,0,0,0,0,0,0,0,0,0,0])\n",
    "    return oneHotEncodedList\n",
    "\n",
    "\n",
    "def splitfunction(text:str):\n",
    "    global wordsList,wordsIndextList\n",
    "    wordsList = []\n",
    "    wordsIndextList = []\n",
    "    words = text.split()\n",
    "    for word in words:\n",
    "        wordsList.append(word)\n",
    "        wordsIndextList.append(1)\n",
    "\n",
    "\n",
    "\n",
    "def indexAssignment(tag,last):\n",
    "    if(tag==None):\n",
    "        return None\n",
    "    index = label_list.index(tag)+2\n",
    "    if(last == \"l\"):\n",
    "       index +=1\n",
    "    return index\n",
    "\n",
    "\n",
    "def checkTag(word:str):\n",
    "    for tag in label_list:\n",
    "        index = word.find(tag)\n",
    "        if index != -1:\n",
    "            return tag\n",
    "    return None\n",
    "\n",
    "\n",
    "\n",
    "def tagingWords():\n",
    "    control = None\n",
    "    counter = 0\n",
    "    for word in wordsList:\n",
    "        tag = checkTag(word)\n",
    "        if tag != None:\n",
    "            if(word.find(\"<start_\") != -1):\n",
    "                control = tag\n",
    "            if(word.find(\"<start_\") != -1 and word.find(\"<end_\") == -1):\n",
    "                wordsIndextList[counter] = indexAssignment(tag,\"B\")\n",
    "                counter+=1\n",
    "            elif(word.find(\"<start_\") != -1 and word.find(\"<end_\") != -1):\n",
    "                wordsIndextList[counter] = indexAssignment(tag,\"B\")\n",
    "                counter+=1\n",
    "            else:\n",
    "                wordsIndextList[counter] = indexAssignment(tag,\"B\")\n",
    "                counter+=1\n",
    "            if(word.find(\"<end_\") != -1):\n",
    "                control = None\n",
    "        elif control !=None:\n",
    "            wordsIndextList[counter] = indexAssignment(control,\"l\")\n",
    "            counter+=1\n",
    "        else:\n",
    "            counter+=1\n",
    "\n",
    "\n",
    "def oneHotEncodedFunction(texts):\n",
    "    splitfunction(texts)\n",
    "    tagingWords()\n",
    "    return np.array(one_hot_encode_index_list())\n",
    "\n",
    "\n",
    "\n",
    "# TRAIN\n",
    "folder_path_train =  directory_train\n",
    "file_list_train = os.listdir(folder_path_train)\n",
    "y_train= []\n",
    "for file_name in file_list_train:\n",
    "    if file_name.endswith(\".txt\"):\n",
    "        file_path = os.path.join(folder_path_train, file_name)\n",
    "        with open(file_path, 'r', encoding='utf-8') as f:\n",
    "            text = f.read()\n",
    "            result = oneHotEncodedFunction(text)\n",
    "\n",
    "            if len(y_train) == 0:\n",
    "                y_train = np.array([result])\n",
    "            else:\n",
    "                y_train = np.concatenate((y_train, [result]), axis=0)\n",
    "\n",
    "\n",
    "# TEST\n",
    "folder_path_test =  directory_test\n",
    "file_list_test = os.listdir(folder_path_test)\n",
    "y_test= []\n",
    "for file_name in file_list_test:\n",
    "    if file_name.endswith(\".txt\"):\n",
    "        file_path = os.path.join(folder_path_test, file_name)\n",
    "        with open(file_path, 'r', encoding='utf-8') as f:\n",
    "            text = f.read()\n",
    "            result = oneHotEncodedFunction(text)\n",
    "\n",
    "            if len(y_test) == 0:\n",
    "                y_test = np.array([result])\n",
    "            else:\n",
    "                y_test = np.concatenate((y_test, [result]), axis=0)\n",
    "\n",
    "\n",
    "print(y_train.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Cn98EjXx-yjY",
    "outputId": "0209eb6d-bd0e-4300-af8b-fbb6efdc1cfc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(None, 384, 20)\n"
     ]
    }
   ],
   "source": [
    "# LSTM Modeli\n",
    "\n",
    "input_shape = (max_sequence_length, max_word_length)\n",
    "input_layer = Input(shape=(max_sequence_length,max_word_length))\n",
    "print(input_layer.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "mdptHORKERmx",
    "outputId": "72270e50-638f-4f63-f838-689bd1a941c6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(None, 384, 20, 128)\n"
     ]
    }
   ],
   "source": [
    "char_embedding = Embedding(input_dim=vocab_len, output_dim=embedding_size)(input_layer)\n",
    "print(char_embedding.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "W7FoMTlUES5H",
    "outputId": "fff0bfb8-4eb5-434c-8ad9-e81d08007e99"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(None, 20, 128)\n"
     ]
    }
   ],
   "source": [
    "reshaped_embedding = tensorflow.reshape(char_embedding, (-1, max_word_length, embedding_size))\n",
    "print(reshaped_embedding.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "j7icA8cmETxy",
    "outputId": "586c56d5-e82a-45b6-e2c8-075fa3387bc8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KerasTensor(type_spec=TensorSpec(shape=(None, 20, 128), dtype=tf.float32, name=None), name='bidirectional/concat:0', description=\"created by layer 'bidirectional'\")\n",
      "KerasTensor(type_spec=TensorSpec(shape=(None, 20, 128), dtype=tf.float32, name=None), name='bidirectional_1/concat:0', description=\"created by layer 'bidirectional_1'\")\n"
     ]
    }
   ],
   "source": [
    "lstm_units = 64\n",
    "char_lstm, forward_h, forward_c, backward_h, backward_c = Bidirectional(LSTM(units=lstm_units, return_sequences=True, return_state=True))(reshaped_embedding)\n",
    "print(char_lstm)\n",
    "\n",
    "char_lstm2, forward_h2, forward_c2, backward_h2, backward_c2 = Bidirectional(LSTM(units=lstm_units, return_sequences=True, return_state=True))(\n",
    "    char_lstm, initial_state=[forward_h, forward_c, backward_h, backward_c])\n",
    "\n",
    "state_h = Concatenate()([forward_h2, backward_h2])\n",
    "state_c = Concatenate()([forward_c2, backward_c2])\n",
    "\n",
    "print(char_lstm2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "YGtn72ZpEUh4",
    "outputId": "c9619696-5622-47f6-a121-2ef16f66086a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(None, 384, 128)\n"
     ]
    }
   ],
   "source": [
    "reshaped = tensorflow.reshape(state_h, (-1, max_sequence_length, 2 * lstm_units))\n",
    "print(reshaped.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "BcEFTZt2EVUr",
    "outputId": "d6b5882e-16bc-42a7-a70b-4fbb2ef0ea54"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KerasTensor(type_spec=TensorSpec(shape=(None, 384, 128), dtype=tf.float32, name=None), name='bidirectional_2/concat:0', description=\"created by layer 'bidirectional_2'\")\n",
      "KerasTensor(type_spec=TensorSpec(shape=(None, 384, 128), dtype=tf.float32, name=None), name='bidirectional_3/concat:0', description=\"created by layer 'bidirectional_3'\")\n"
     ]
    }
   ],
   "source": [
    "word_lstm1, forward_h, forward_c, backward_h, backward_c = Bidirectional(LSTM(units=lstm_units, return_sequences=True, return_state=True))(reshaped)\n",
    "print(word_lstm1)\n",
    "\n",
    "word_lstm2, forward_h2, forward_c2, backward_h2, backward_c2 = Bidirectional(LSTM(units=lstm_units, return_sequences=True, return_state=True))(word_lstm1, initial_state=[forward_h, forward_c, backward_h, backward_c])\n",
    "print(word_lstm2)\n",
    "\n",
    "state_h = Concatenate()([forward_h2, backward_h2])\n",
    "state_c = Concatenate()([forward_c2, backward_c2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3x_Rxbx0EWXU",
    "outputId": "3bf8b225-10d5-4d96-f816-8a6fd0efd758"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KerasTensor(type_spec=TensorSpec(shape=(None, 384, 128), dtype=tf.float32, name=None), name='bidirectional_3/concat:0', description=\"created by layer 'bidirectional_3'\")\n"
     ]
    }
   ],
   "source": [
    "classification_output = Dense(units=num_classes, activation='softmax')(word_lstm2)\n",
    "print(word_lstm2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "-uURGeTIEXGw"
   },
   "outputs": [],
   "source": [
    "model = Model(inputs=input_layer, outputs=classification_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "VZyQlIvNEYC7",
    "outputId": "ecae5f3d-b295-4f1d-85a8-50f8f48f6870"
   },
   "outputs": [],
   "source": [
    "#model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "#model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 384, 20)]    0                                            \n",
      "__________________________________________________________________________________________________\n",
      "embedding (Embedding)           (None, 384, 20, 128) 25088       input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "tf.reshape (TFOpLambda)         (None, 20, 128)      0           embedding[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional (Bidirectional)   [(None, 20, 128), (N 98816       tf.reshape[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_1 (Bidirectional) [(None, 20, 128), (N 98816       bidirectional[0][0]              \n",
      "                                                                 bidirectional[0][1]              \n",
      "                                                                 bidirectional[0][2]              \n",
      "                                                                 bidirectional[0][3]              \n",
      "                                                                 bidirectional[0][4]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 128)          0           bidirectional_1[0][1]            \n",
      "                                                                 bidirectional_1[0][3]            \n",
      "__________________________________________________________________________________________________\n",
      "tf.reshape_1 (TFOpLambda)       (None, 384, 128)     0           concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_2 (Bidirectional) [(None, 384, 128), ( 98816       tf.reshape_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_3 (Bidirectional) [(None, 384, 128), ( 98816       bidirectional_2[0][0]            \n",
      "                                                                 bidirectional_2[0][1]            \n",
      "                                                                 bidirectional_2[0][2]            \n",
      "                                                                 bidirectional_2[0][3]            \n",
      "                                                                 bidirectional_2[0][4]            \n",
      "__________________________________________________________________________________________________\n",
      "dense (Dense)                   (None, 384, 14)      1806        bidirectional_3[0][0]            \n",
      "==================================================================================================\n",
      "Total params: 422,158\n",
      "Trainable params: 422,158\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.callbacks import ModelCheckpoint\n",
    "import tensorflow.keras as keras\n",
    "\n",
    "checkpoint = ModelCheckpoint('bi_best_model.keras', monitor='val_accuracy', save_best_only=True)\n",
    "\n",
    "opt = keras.optimizers.Adam(learning_rate=0.0001)\n",
    "\n",
    "model.compile(optimizer=opt, loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "model.summary()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "X1GwVzzxEZCh",
    "outputId": "1e1b9b95-266a-46f6-9662-ea7cb81a328b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "35/35 [==============================] - 18s 234ms/step - loss: 2.2554 - accuracy: 0.6677 - val_loss: 1.2587 - val_accuracy: 0.7306\n",
      "Epoch 2/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.8725 - accuracy: 0.7167 - val_loss: 0.6981 - val_accuracy: 0.7306\n",
      "Epoch 3/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.6211 - accuracy: 0.7470 - val_loss: 0.4143 - val_accuracy: 0.9249\n",
      "Epoch 4/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.2731 - accuracy: 0.9544 - val_loss: 0.2145 - val_accuracy: 0.9629\n",
      "Epoch 5/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.2001 - accuracy: 0.9602 - val_loss: 0.1796 - val_accuracy: 0.9647\n",
      "Epoch 6/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.1768 - accuracy: 0.9650 - val_loss: 0.1662 - val_accuracy: 0.9668\n",
      "Epoch 7/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1666 - accuracy: 0.9658 - val_loss: 0.1595 - val_accuracy: 0.9660\n",
      "Epoch 8/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.1611 - accuracy: 0.9652 - val_loss: 0.1543 - val_accuracy: 0.9665\n",
      "Epoch 9/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1564 - accuracy: 0.9657 - val_loss: 0.1504 - val_accuracy: 0.9667\n",
      "Epoch 10/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1531 - accuracy: 0.9659 - val_loss: 0.1474 - val_accuracy: 0.9669\n",
      "Epoch 11/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1503 - accuracy: 0.9658 - val_loss: 0.1448 - val_accuracy: 0.9670\n",
      "Epoch 12/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1477 - accuracy: 0.9661 - val_loss: 0.1437 - val_accuracy: 0.9661\n",
      "Epoch 13/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.1461 - accuracy: 0.9658 - val_loss: 0.1410 - val_accuracy: 0.9670\n",
      "Epoch 14/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.1442 - accuracy: 0.9661 - val_loss: 0.1389 - val_accuracy: 0.9676\n",
      "Epoch 15/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1425 - accuracy: 0.9668 - val_loss: 0.1374 - val_accuracy: 0.9674\n",
      "Epoch 16/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1406 - accuracy: 0.9663 - val_loss: 0.1370 - val_accuracy: 0.9661\n",
      "Epoch 17/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1393 - accuracy: 0.9661 - val_loss: 0.1341 - val_accuracy: 0.9673\n",
      "Epoch 18/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.1378 - accuracy: 0.9668 - val_loss: 0.1333 - val_accuracy: 0.9676\n",
      "Epoch 19/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1365 - accuracy: 0.9677 - val_loss: 0.1315 - val_accuracy: 0.9683\n",
      "Epoch 20/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1354 - accuracy: 0.9678 - val_loss: 0.1304 - val_accuracy: 0.9684\n",
      "Epoch 21/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.1341 - accuracy: 0.9680 - val_loss: 0.1298 - val_accuracy: 0.9678\n",
      "Epoch 22/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1332 - accuracy: 0.9677 - val_loss: 0.1283 - val_accuracy: 0.9687\n",
      "Epoch 23/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1320 - accuracy: 0.9683 - val_loss: 0.1274 - val_accuracy: 0.9684\n",
      "Epoch 24/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1318 - accuracy: 0.9677 - val_loss: 0.1262 - val_accuracy: 0.9689\n",
      "Epoch 25/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.1298 - accuracy: 0.9688 - val_loss: 0.1250 - val_accuracy: 0.9690\n",
      "Epoch 26/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1287 - accuracy: 0.9688 - val_loss: 0.1236 - val_accuracy: 0.9692\n",
      "Epoch 27/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.1272 - accuracy: 0.9689 - val_loss: 0.1226 - val_accuracy: 0.9688\n",
      "Epoch 28/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.1262 - accuracy: 0.9687 - val_loss: 0.1212 - val_accuracy: 0.9689\n",
      "Epoch 29/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.1247 - accuracy: 0.9689 - val_loss: 0.1198 - val_accuracy: 0.9688\n",
      "Epoch 30/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.1232 - accuracy: 0.9687 - val_loss: 0.1186 - val_accuracy: 0.9687\n",
      "Epoch 31/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.1218 - accuracy: 0.9687 - val_loss: 0.1174 - val_accuracy: 0.9686\n",
      "Epoch 32/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1203 - accuracy: 0.9687 - val_loss: 0.1154 - val_accuracy: 0.9689\n",
      "Epoch 33/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1195 - accuracy: 0.9685 - val_loss: 0.1152 - val_accuracy: 0.9686\n",
      "Epoch 34/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.1179 - accuracy: 0.9686 - val_loss: 0.1130 - val_accuracy: 0.9686\n",
      "Epoch 35/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1167 - accuracy: 0.9685 - val_loss: 0.1117 - val_accuracy: 0.9685\n",
      "Epoch 36/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.1148 - accuracy: 0.9687 - val_loss: 0.1098 - val_accuracy: 0.9688\n",
      "Epoch 37/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1134 - accuracy: 0.9687 - val_loss: 0.1089 - val_accuracy: 0.9685\n",
      "Epoch 38/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1128 - accuracy: 0.9683 - val_loss: 0.1077 - val_accuracy: 0.9690\n",
      "Epoch 39/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1108 - accuracy: 0.9688 - val_loss: 0.1059 - val_accuracy: 0.9690\n",
      "Epoch 40/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1090 - accuracy: 0.9690 - val_loss: 0.1043 - val_accuracy: 0.9690\n",
      "Epoch 41/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.1076 - accuracy: 0.9689 - val_loss: 0.1030 - val_accuracy: 0.9690\n",
      "Epoch 42/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1063 - accuracy: 0.9689 - val_loss: 0.1012 - val_accuracy: 0.9693\n",
      "Epoch 43/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1057 - accuracy: 0.9685 - val_loss: 0.1017 - val_accuracy: 0.9687\n",
      "Epoch 44/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1039 - accuracy: 0.9691 - val_loss: 0.0990 - val_accuracy: 0.9692\n",
      "Epoch 45/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.1023 - accuracy: 0.9692 - val_loss: 0.0985 - val_accuracy: 0.9687\n",
      "Epoch 46/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1009 - accuracy: 0.9694 - val_loss: 0.0989 - val_accuracy: 0.9680\n",
      "Epoch 47/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.1016 - accuracy: 0.9690 - val_loss: 0.0958 - val_accuracy: 0.9699\n",
      "Epoch 48/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0982 - accuracy: 0.9701 - val_loss: 0.0948 - val_accuracy: 0.9702\n",
      "Epoch 49/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0967 - accuracy: 0.9702 - val_loss: 0.0932 - val_accuracy: 0.9705\n",
      "Epoch 50/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0956 - accuracy: 0.9705 - val_loss: 0.0922 - val_accuracy: 0.9705\n",
      "Epoch 51/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0945 - accuracy: 0.9708 - val_loss: 0.0906 - val_accuracy: 0.9709\n",
      "Epoch 52/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0930 - accuracy: 0.9713 - val_loss: 0.0896 - val_accuracy: 0.9712\n",
      "Epoch 53/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0916 - accuracy: 0.9714 - val_loss: 0.0884 - val_accuracy: 0.9716\n",
      "Epoch 54/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0908 - accuracy: 0.9717 - val_loss: 0.0875 - val_accuracy: 0.9719\n",
      "Epoch 55/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0894 - accuracy: 0.9721 - val_loss: 0.0864 - val_accuracy: 0.9725\n",
      "Epoch 56/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0890 - accuracy: 0.9723 - val_loss: 0.0866 - val_accuracy: 0.9722\n",
      "Epoch 57/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0933 - accuracy: 0.9709 - val_loss: 0.0980 - val_accuracy: 0.9675\n",
      "Epoch 58/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0919 - accuracy: 0.9716 - val_loss: 0.0866 - val_accuracy: 0.9722\n",
      "Epoch 59/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0882 - accuracy: 0.9727 - val_loss: 0.0846 - val_accuracy: 0.9728\n",
      "Epoch 60/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0862 - accuracy: 0.9733 - val_loss: 0.0834 - val_accuracy: 0.9730\n",
      "Epoch 61/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0850 - accuracy: 0.9735 - val_loss: 0.0831 - val_accuracy: 0.9731\n",
      "Epoch 62/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0840 - accuracy: 0.9740 - val_loss: 0.0816 - val_accuracy: 0.9737\n",
      "Epoch 63/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0833 - accuracy: 0.9742 - val_loss: 0.0808 - val_accuracy: 0.9740\n",
      "Epoch 64/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0823 - accuracy: 0.9744 - val_loss: 0.0797 - val_accuracy: 0.9748\n",
      "Epoch 65/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0813 - accuracy: 0.9748 - val_loss: 0.0791 - val_accuracy: 0.9750\n",
      "Epoch 66/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0809 - accuracy: 0.9751 - val_loss: 0.0786 - val_accuracy: 0.9754\n",
      "Epoch 67/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0796 - accuracy: 0.9756 - val_loss: 0.0778 - val_accuracy: 0.9755\n",
      "Epoch 68/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0788 - accuracy: 0.9756 - val_loss: 0.0770 - val_accuracy: 0.9754\n",
      "Epoch 69/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0782 - accuracy: 0.9759 - val_loss: 0.0766 - val_accuracy: 0.9756\n",
      "Epoch 70/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0778 - accuracy: 0.9760 - val_loss: 0.0756 - val_accuracy: 0.9761\n",
      "Epoch 71/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0770 - accuracy: 0.9763 - val_loss: 0.0760 - val_accuracy: 0.9757\n",
      "Epoch 72/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0766 - accuracy: 0.9761 - val_loss: 0.0745 - val_accuracy: 0.9764\n",
      "Epoch 73/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0757 - accuracy: 0.9764 - val_loss: 0.0767 - val_accuracy: 0.9754\n",
      "Epoch 74/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0758 - accuracy: 0.9765 - val_loss: 0.0734 - val_accuracy: 0.9769\n",
      "Epoch 75/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0740 - accuracy: 0.9771 - val_loss: 0.0727 - val_accuracy: 0.9768\n",
      "Epoch 76/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0733 - accuracy: 0.9774 - val_loss: 0.0721 - val_accuracy: 0.9771\n",
      "Epoch 77/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0732 - accuracy: 0.9771 - val_loss: 0.0713 - val_accuracy: 0.9775\n",
      "Epoch 78/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0718 - accuracy: 0.9779 - val_loss: 0.0714 - val_accuracy: 0.9772\n",
      "Epoch 79/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0722 - accuracy: 0.9776 - val_loss: 0.0716 - val_accuracy: 0.9773\n",
      "Epoch 80/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0710 - accuracy: 0.9780 - val_loss: 0.0702 - val_accuracy: 0.9777\n",
      "Epoch 81/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0702 - accuracy: 0.9782 - val_loss: 0.0693 - val_accuracy: 0.9781\n",
      "Epoch 82/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0697 - accuracy: 0.9786 - val_loss: 0.0702 - val_accuracy: 0.9773\n",
      "Epoch 83/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0689 - accuracy: 0.9789 - val_loss: 0.0683 - val_accuracy: 0.9785\n",
      "Epoch 84/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0684 - accuracy: 0.9788 - val_loss: 0.0681 - val_accuracy: 0.9784\n",
      "Epoch 85/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0676 - accuracy: 0.9791 - val_loss: 0.0672 - val_accuracy: 0.9786\n",
      "Epoch 86/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0673 - accuracy: 0.9792 - val_loss: 0.0671 - val_accuracy: 0.9785\n",
      "Epoch 87/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0672 - accuracy: 0.9793 - val_loss: 0.0666 - val_accuracy: 0.9789\n",
      "Epoch 88/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0665 - accuracy: 0.9793 - val_loss: 0.0663 - val_accuracy: 0.9787\n",
      "Epoch 89/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0658 - accuracy: 0.9798 - val_loss: 0.0662 - val_accuracy: 0.9788\n",
      "Epoch 90/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0655 - accuracy: 0.9799 - val_loss: 0.0657 - val_accuracy: 0.9791\n",
      "Epoch 91/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0653 - accuracy: 0.9797 - val_loss: 0.0651 - val_accuracy: 0.9797\n",
      "Epoch 92/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0653 - accuracy: 0.9797 - val_loss: 0.0659 - val_accuracy: 0.9791\n",
      "Epoch 93/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0644 - accuracy: 0.9803 - val_loss: 0.0646 - val_accuracy: 0.9796\n",
      "Epoch 94/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0643 - accuracy: 0.9802 - val_loss: 0.0638 - val_accuracy: 0.9798\n",
      "Epoch 95/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0632 - accuracy: 0.9807 - val_loss: 0.0638 - val_accuracy: 0.9796\n",
      "Epoch 96/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0629 - accuracy: 0.9806 - val_loss: 0.0633 - val_accuracy: 0.9800\n",
      "Epoch 97/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0624 - accuracy: 0.9807 - val_loss: 0.0627 - val_accuracy: 0.9804\n",
      "Epoch 98/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0614 - accuracy: 0.9810 - val_loss: 0.0626 - val_accuracy: 0.9801\n",
      "Epoch 99/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0616 - accuracy: 0.9810 - val_loss: 0.0617 - val_accuracy: 0.9805\n",
      "Epoch 100/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0608 - accuracy: 0.9814 - val_loss: 0.0618 - val_accuracy: 0.9803\n",
      "Epoch 101/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0600 - accuracy: 0.9815 - val_loss: 0.0611 - val_accuracy: 0.9806\n",
      "Epoch 102/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0596 - accuracy: 0.9817 - val_loss: 0.0613 - val_accuracy: 0.9810\n",
      "Epoch 103/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0596 - accuracy: 0.9819 - val_loss: 0.0602 - val_accuracy: 0.9813\n",
      "Epoch 104/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0587 - accuracy: 0.9820 - val_loss: 0.0599 - val_accuracy: 0.9814\n",
      "Epoch 105/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0584 - accuracy: 0.9822 - val_loss: 0.0597 - val_accuracy: 0.9811\n",
      "Epoch 106/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0578 - accuracy: 0.9823 - val_loss: 0.0593 - val_accuracy: 0.9815\n",
      "Epoch 107/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0578 - accuracy: 0.9822 - val_loss: 0.0592 - val_accuracy: 0.9815\n",
      "Epoch 108/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0572 - accuracy: 0.9825 - val_loss: 0.0598 - val_accuracy: 0.9809\n",
      "Epoch 109/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0568 - accuracy: 0.9825 - val_loss: 0.0579 - val_accuracy: 0.9822\n",
      "Epoch 110/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0566 - accuracy: 0.9827 - val_loss: 0.0585 - val_accuracy: 0.9814\n",
      "Epoch 111/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0564 - accuracy: 0.9826 - val_loss: 0.0572 - val_accuracy: 0.9823\n",
      "Epoch 112/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0556 - accuracy: 0.9831 - val_loss: 0.0568 - val_accuracy: 0.9824\n",
      "Epoch 113/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0546 - accuracy: 0.9835 - val_loss: 0.0569 - val_accuracy: 0.9822\n",
      "Epoch 114/1000\n",
      "35/35 [==============================] - 6s 181ms/step - loss: 0.0555 - accuracy: 0.9828 - val_loss: 0.0565 - val_accuracy: 0.9827\n",
      "Epoch 115/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0543 - accuracy: 0.9836 - val_loss: 0.0559 - val_accuracy: 0.9825\n",
      "Epoch 116/1000\n",
      "35/35 [==============================] - 6s 183ms/step - loss: 0.0536 - accuracy: 0.9836 - val_loss: 0.0555 - val_accuracy: 0.9826\n",
      "Epoch 117/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0527 - accuracy: 0.9840 - val_loss: 0.0556 - val_accuracy: 0.9823\n",
      "Epoch 118/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0526 - accuracy: 0.9839 - val_loss: 0.0552 - val_accuracy: 0.9826\n",
      "Epoch 119/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0531 - accuracy: 0.9838 - val_loss: 0.0552 - val_accuracy: 0.9824\n",
      "Epoch 120/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0524 - accuracy: 0.9838 - val_loss: 0.0540 - val_accuracy: 0.9832\n",
      "Epoch 121/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0515 - accuracy: 0.9844 - val_loss: 0.0545 - val_accuracy: 0.9828\n",
      "Epoch 122/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0513 - accuracy: 0.9844 - val_loss: 0.0532 - val_accuracy: 0.9835\n",
      "Epoch 123/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0506 - accuracy: 0.9848 - val_loss: 0.0531 - val_accuracy: 0.9838\n",
      "Epoch 124/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0506 - accuracy: 0.9848 - val_loss: 0.0532 - val_accuracy: 0.9835\n",
      "Epoch 125/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0498 - accuracy: 0.9851 - val_loss: 0.0527 - val_accuracy: 0.9836\n",
      "Epoch 126/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0492 - accuracy: 0.9851 - val_loss: 0.0521 - val_accuracy: 0.9836\n",
      "Epoch 127/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0486 - accuracy: 0.9854 - val_loss: 0.0515 - val_accuracy: 0.9842\n",
      "Epoch 128/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0486 - accuracy: 0.9853 - val_loss: 0.0512 - val_accuracy: 0.9843\n",
      "Epoch 129/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0481 - accuracy: 0.9855 - val_loss: 0.0522 - val_accuracy: 0.9837\n",
      "Epoch 130/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0482 - accuracy: 0.9853 - val_loss: 0.0510 - val_accuracy: 0.9842\n",
      "Epoch 131/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0473 - accuracy: 0.9858 - val_loss: 0.0510 - val_accuracy: 0.9841\n",
      "Epoch 132/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0468 - accuracy: 0.9859 - val_loss: 0.0521 - val_accuracy: 0.9837\n",
      "Epoch 133/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0473 - accuracy: 0.9858 - val_loss: 0.0504 - val_accuracy: 0.9844\n",
      "Epoch 134/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0462 - accuracy: 0.9860 - val_loss: 0.0500 - val_accuracy: 0.9846\n",
      "Epoch 135/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0463 - accuracy: 0.9860 - val_loss: 0.0490 - val_accuracy: 0.9848\n",
      "Epoch 136/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0453 - accuracy: 0.9863 - val_loss: 0.0495 - val_accuracy: 0.9846\n",
      "Epoch 137/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0457 - accuracy: 0.9861 - val_loss: 0.0503 - val_accuracy: 0.9845\n",
      "Epoch 138/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0446 - accuracy: 0.9865 - val_loss: 0.0487 - val_accuracy: 0.9848\n",
      "Epoch 139/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0445 - accuracy: 0.9865 - val_loss: 0.0488 - val_accuracy: 0.9849\n",
      "Epoch 140/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0443 - accuracy: 0.9865 - val_loss: 0.0516 - val_accuracy: 0.9838\n",
      "Epoch 141/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0444 - accuracy: 0.9864 - val_loss: 0.0476 - val_accuracy: 0.9854\n",
      "Epoch 142/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0429 - accuracy: 0.9872 - val_loss: 0.0480 - val_accuracy: 0.9853\n",
      "Epoch 143/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0429 - accuracy: 0.9870 - val_loss: 0.0473 - val_accuracy: 0.9853\n",
      "Epoch 144/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0432 - accuracy: 0.9870 - val_loss: 0.0469 - val_accuracy: 0.9856\n",
      "Epoch 145/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0426 - accuracy: 0.9870 - val_loss: 0.0470 - val_accuracy: 0.9854\n",
      "Epoch 146/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0420 - accuracy: 0.9873 - val_loss: 0.0461 - val_accuracy: 0.9857\n",
      "Epoch 147/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0412 - accuracy: 0.9875 - val_loss: 0.0463 - val_accuracy: 0.9854\n",
      "Epoch 148/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0411 - accuracy: 0.9873 - val_loss: 0.0457 - val_accuracy: 0.9860\n",
      "Epoch 149/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0409 - accuracy: 0.9876 - val_loss: 0.0454 - val_accuracy: 0.9858\n",
      "Epoch 150/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0407 - accuracy: 0.9875 - val_loss: 0.0450 - val_accuracy: 0.9861\n",
      "Epoch 151/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0398 - accuracy: 0.9880 - val_loss: 0.0448 - val_accuracy: 0.9863\n",
      "Epoch 152/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0398 - accuracy: 0.9879 - val_loss: 0.0449 - val_accuracy: 0.9860\n",
      "Epoch 153/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0398 - accuracy: 0.9879 - val_loss: 0.0446 - val_accuracy: 0.9862\n",
      "Epoch 154/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0390 - accuracy: 0.9882 - val_loss: 0.0447 - val_accuracy: 0.9860\n",
      "Epoch 155/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0388 - accuracy: 0.9882 - val_loss: 0.0452 - val_accuracy: 0.9857\n",
      "Epoch 156/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0389 - accuracy: 0.9882 - val_loss: 0.0434 - val_accuracy: 0.9865\n",
      "Epoch 157/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0388 - accuracy: 0.9881 - val_loss: 0.0437 - val_accuracy: 0.9866\n",
      "Epoch 158/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0383 - accuracy: 0.9883 - val_loss: 0.0446 - val_accuracy: 0.9865\n",
      "Epoch 159/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0383 - accuracy: 0.9885 - val_loss: 0.0442 - val_accuracy: 0.9865\n",
      "Epoch 160/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0376 - accuracy: 0.9886 - val_loss: 0.0435 - val_accuracy: 0.9867\n",
      "Epoch 161/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0381 - accuracy: 0.9884 - val_loss: 0.0435 - val_accuracy: 0.9869\n",
      "Epoch 162/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0371 - accuracy: 0.9887 - val_loss: 0.0423 - val_accuracy: 0.9871\n",
      "Epoch 163/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0393 - accuracy: 0.9880 - val_loss: 0.0435 - val_accuracy: 0.9866\n",
      "Epoch 164/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0379 - accuracy: 0.9885 - val_loss: 0.0424 - val_accuracy: 0.9870\n",
      "Epoch 165/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0366 - accuracy: 0.9888 - val_loss: 0.0424 - val_accuracy: 0.9868\n",
      "Epoch 166/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0373 - accuracy: 0.9887 - val_loss: 0.0450 - val_accuracy: 0.9861\n",
      "Epoch 167/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0378 - accuracy: 0.9885 - val_loss: 0.0426 - val_accuracy: 0.9871\n",
      "Epoch 168/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0370 - accuracy: 0.9888 - val_loss: 0.0411 - val_accuracy: 0.9874\n",
      "Epoch 169/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0354 - accuracy: 0.9894 - val_loss: 0.0417 - val_accuracy: 0.9873\n",
      "Epoch 170/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0357 - accuracy: 0.9893 - val_loss: 0.0414 - val_accuracy: 0.9872\n",
      "Epoch 171/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0351 - accuracy: 0.9895 - val_loss: 0.0413 - val_accuracy: 0.9873\n",
      "Epoch 172/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0343 - accuracy: 0.9895 - val_loss: 0.0407 - val_accuracy: 0.9875\n",
      "Epoch 173/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0339 - accuracy: 0.9897 - val_loss: 0.0413 - val_accuracy: 0.9875\n",
      "Epoch 174/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0365 - accuracy: 0.9889 - val_loss: 0.0414 - val_accuracy: 0.9874\n",
      "Epoch 175/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0343 - accuracy: 0.9894 - val_loss: 0.0399 - val_accuracy: 0.9878\n",
      "Epoch 176/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0337 - accuracy: 0.9898 - val_loss: 0.0403 - val_accuracy: 0.9875\n",
      "Epoch 177/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0331 - accuracy: 0.9899 - val_loss: 0.0403 - val_accuracy: 0.9877\n",
      "Epoch 178/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0332 - accuracy: 0.9898 - val_loss: 0.0409 - val_accuracy: 0.9871\n",
      "Epoch 179/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0330 - accuracy: 0.9899 - val_loss: 0.0403 - val_accuracy: 0.9876\n",
      "Epoch 180/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0325 - accuracy: 0.9901 - val_loss: 0.0396 - val_accuracy: 0.9879\n",
      "Epoch 181/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0325 - accuracy: 0.9902 - val_loss: 0.0404 - val_accuracy: 0.9875\n",
      "Epoch 182/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0320 - accuracy: 0.9902 - val_loss: 0.0399 - val_accuracy: 0.9876\n",
      "Epoch 183/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0320 - accuracy: 0.9903 - val_loss: 0.0400 - val_accuracy: 0.9878\n",
      "Epoch 184/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0315 - accuracy: 0.9905 - val_loss: 0.0385 - val_accuracy: 0.9882\n",
      "Epoch 185/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0317 - accuracy: 0.9904 - val_loss: 0.0387 - val_accuracy: 0.9881\n",
      "Epoch 186/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0312 - accuracy: 0.9907 - val_loss: 0.0392 - val_accuracy: 0.9881\n",
      "Epoch 187/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0312 - accuracy: 0.9905 - val_loss: 0.0385 - val_accuracy: 0.9882\n",
      "Epoch 188/1000\n",
      "35/35 [==============================] - 6s 168ms/step - loss: 0.0308 - accuracy: 0.9907 - val_loss: 0.0388 - val_accuracy: 0.9880\n",
      "Epoch 189/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0307 - accuracy: 0.9908 - val_loss: 0.0388 - val_accuracy: 0.9882\n",
      "Epoch 190/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0302 - accuracy: 0.9910 - val_loss: 0.0394 - val_accuracy: 0.9877\n",
      "Epoch 191/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0316 - accuracy: 0.9904 - val_loss: 0.0388 - val_accuracy: 0.9882\n",
      "Epoch 192/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0303 - accuracy: 0.9909 - val_loss: 0.0384 - val_accuracy: 0.9881\n",
      "Epoch 193/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0297 - accuracy: 0.9912 - val_loss: 0.0389 - val_accuracy: 0.9883\n",
      "Epoch 194/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0303 - accuracy: 0.9909 - val_loss: 0.0392 - val_accuracy: 0.9882\n",
      "Epoch 195/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0299 - accuracy: 0.9909 - val_loss: 0.0389 - val_accuracy: 0.9881\n",
      "Epoch 196/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0299 - accuracy: 0.9909 - val_loss: 0.0382 - val_accuracy: 0.9882\n",
      "Epoch 197/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0291 - accuracy: 0.9912 - val_loss: 0.0378 - val_accuracy: 0.9884\n",
      "Epoch 198/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0291 - accuracy: 0.9913 - val_loss: 0.0384 - val_accuracy: 0.9883\n",
      "Epoch 199/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0290 - accuracy: 0.9913 - val_loss: 0.0383 - val_accuracy: 0.9884\n",
      "Epoch 200/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0289 - accuracy: 0.9912 - val_loss: 0.0377 - val_accuracy: 0.9884\n",
      "Epoch 201/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0283 - accuracy: 0.9915 - val_loss: 0.0382 - val_accuracy: 0.9883\n",
      "Epoch 202/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0285 - accuracy: 0.9915 - val_loss: 0.0380 - val_accuracy: 0.9887\n",
      "Epoch 203/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0281 - accuracy: 0.9916 - val_loss: 0.0376 - val_accuracy: 0.9885\n",
      "Epoch 204/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0281 - accuracy: 0.9917 - val_loss: 0.0377 - val_accuracy: 0.9885\n",
      "Epoch 205/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0278 - accuracy: 0.9917 - val_loss: 0.0374 - val_accuracy: 0.9887\n",
      "Epoch 206/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0279 - accuracy: 0.9917 - val_loss: 0.0375 - val_accuracy: 0.9888\n",
      "Epoch 207/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0276 - accuracy: 0.9919 - val_loss: 0.0374 - val_accuracy: 0.9886\n",
      "Epoch 208/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0285 - accuracy: 0.9915 - val_loss: 0.0369 - val_accuracy: 0.9886\n",
      "Epoch 209/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0267 - accuracy: 0.9923 - val_loss: 0.0377 - val_accuracy: 0.9887\n",
      "Epoch 210/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0273 - accuracy: 0.9919 - val_loss: 0.0387 - val_accuracy: 0.9884\n",
      "Epoch 211/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0265 - accuracy: 0.9921 - val_loss: 0.0370 - val_accuracy: 0.9889\n",
      "Epoch 212/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0266 - accuracy: 0.9922 - val_loss: 0.0371 - val_accuracy: 0.9888\n",
      "Epoch 213/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0276 - accuracy: 0.9917 - val_loss: 0.0369 - val_accuracy: 0.9887\n",
      "Epoch 214/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0261 - accuracy: 0.9923 - val_loss: 0.0366 - val_accuracy: 0.9890\n",
      "Epoch 215/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0258 - accuracy: 0.9925 - val_loss: 0.0375 - val_accuracy: 0.9888\n",
      "Epoch 216/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0258 - accuracy: 0.9925 - val_loss: 0.0366 - val_accuracy: 0.9893\n",
      "Epoch 217/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0260 - accuracy: 0.9922 - val_loss: 0.0358 - val_accuracy: 0.9893\n",
      "Epoch 218/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0254 - accuracy: 0.9926 - val_loss: 0.0371 - val_accuracy: 0.9885\n",
      "Epoch 219/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0254 - accuracy: 0.9926 - val_loss: 0.0363 - val_accuracy: 0.9894\n",
      "Epoch 220/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0250 - accuracy: 0.9926 - val_loss: 0.0365 - val_accuracy: 0.9892\n",
      "Epoch 221/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0248 - accuracy: 0.9928 - val_loss: 0.0373 - val_accuracy: 0.9892\n",
      "Epoch 222/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0265 - accuracy: 0.9921 - val_loss: 0.0363 - val_accuracy: 0.9892\n",
      "Epoch 223/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0246 - accuracy: 0.9928 - val_loss: 0.0359 - val_accuracy: 0.9895\n",
      "Epoch 224/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0243 - accuracy: 0.9930 - val_loss: 0.0364 - val_accuracy: 0.9893\n",
      "Epoch 225/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0243 - accuracy: 0.9929 - val_loss: 0.0362 - val_accuracy: 0.9892\n",
      "Epoch 226/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0240 - accuracy: 0.9930 - val_loss: 0.0370 - val_accuracy: 0.9890\n",
      "Epoch 227/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0246 - accuracy: 0.9928 - val_loss: 0.0361 - val_accuracy: 0.9892\n",
      "Epoch 228/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0239 - accuracy: 0.9930 - val_loss: 0.0369 - val_accuracy: 0.9888\n",
      "Epoch 229/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0238 - accuracy: 0.9930 - val_loss: 0.0360 - val_accuracy: 0.9894\n",
      "Epoch 230/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0237 - accuracy: 0.9931 - val_loss: 0.0363 - val_accuracy: 0.9892\n",
      "Epoch 231/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0262 - accuracy: 0.9920 - val_loss: 0.0362 - val_accuracy: 0.9892\n",
      "Epoch 232/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0238 - accuracy: 0.9932 - val_loss: 0.0360 - val_accuracy: 0.9895\n",
      "Epoch 233/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0232 - accuracy: 0.9933 - val_loss: 0.0359 - val_accuracy: 0.9896\n",
      "Epoch 234/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0231 - accuracy: 0.9933 - val_loss: 0.0361 - val_accuracy: 0.9895\n",
      "Epoch 235/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0228 - accuracy: 0.9934 - val_loss: 0.0352 - val_accuracy: 0.9897\n",
      "Epoch 236/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0224 - accuracy: 0.9935 - val_loss: 0.0358 - val_accuracy: 0.9895\n",
      "Epoch 237/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0223 - accuracy: 0.9936 - val_loss: 0.0366 - val_accuracy: 0.9891\n",
      "Epoch 238/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0230 - accuracy: 0.9934 - val_loss: 0.0374 - val_accuracy: 0.9890\n",
      "Epoch 239/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0232 - accuracy: 0.9932 - val_loss: 0.0354 - val_accuracy: 0.9896\n",
      "Epoch 240/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0219 - accuracy: 0.9937 - val_loss: 0.0354 - val_accuracy: 0.9898\n",
      "Epoch 241/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0223 - accuracy: 0.9936 - val_loss: 0.0362 - val_accuracy: 0.9893\n",
      "Epoch 242/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0223 - accuracy: 0.9936 - val_loss: 0.0358 - val_accuracy: 0.9895\n",
      "Epoch 243/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0221 - accuracy: 0.9935 - val_loss: 0.0354 - val_accuracy: 0.9895\n",
      "Epoch 244/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0218 - accuracy: 0.9939 - val_loss: 0.0361 - val_accuracy: 0.9893\n",
      "Epoch 245/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0217 - accuracy: 0.9938 - val_loss: 0.0370 - val_accuracy: 0.9892\n",
      "Epoch 246/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0217 - accuracy: 0.9938 - val_loss: 0.0358 - val_accuracy: 0.9895\n",
      "Epoch 247/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0213 - accuracy: 0.9939 - val_loss: 0.0359 - val_accuracy: 0.9897\n",
      "Epoch 248/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0227 - accuracy: 0.9934 - val_loss: 0.0358 - val_accuracy: 0.9895\n",
      "Epoch 249/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0214 - accuracy: 0.9937 - val_loss: 0.0357 - val_accuracy: 0.9899\n",
      "Epoch 250/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0221 - accuracy: 0.9935 - val_loss: 0.0354 - val_accuracy: 0.9896\n",
      "Epoch 251/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0207 - accuracy: 0.9941 - val_loss: 0.0355 - val_accuracy: 0.9897\n",
      "Epoch 252/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0207 - accuracy: 0.9940 - val_loss: 0.0364 - val_accuracy: 0.9895\n",
      "Epoch 253/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0205 - accuracy: 0.9942 - val_loss: 0.0353 - val_accuracy: 0.9897\n",
      "Epoch 254/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0201 - accuracy: 0.9943 - val_loss: 0.0362 - val_accuracy: 0.9896\n",
      "Epoch 255/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0200 - accuracy: 0.9944 - val_loss: 0.0360 - val_accuracy: 0.9897\n",
      "Epoch 256/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0202 - accuracy: 0.9942 - val_loss: 0.0360 - val_accuracy: 0.9897\n",
      "Epoch 257/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0200 - accuracy: 0.9942 - val_loss: 0.0352 - val_accuracy: 0.9897\n",
      "Epoch 258/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0201 - accuracy: 0.9942 - val_loss: 0.0354 - val_accuracy: 0.9898\n",
      "Epoch 259/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0198 - accuracy: 0.9944 - val_loss: 0.0355 - val_accuracy: 0.9896\n",
      "Epoch 260/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0199 - accuracy: 0.9943 - val_loss: 0.0361 - val_accuracy: 0.9897\n",
      "Epoch 261/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0196 - accuracy: 0.9945 - val_loss: 0.0354 - val_accuracy: 0.9897\n",
      "Epoch 262/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0199 - accuracy: 0.9942 - val_loss: 0.0349 - val_accuracy: 0.9901\n",
      "Epoch 263/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0195 - accuracy: 0.9945 - val_loss: 0.0364 - val_accuracy: 0.9897\n",
      "Epoch 264/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0192 - accuracy: 0.9946 - val_loss: 0.0362 - val_accuracy: 0.9897\n",
      "Epoch 265/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0193 - accuracy: 0.9945 - val_loss: 0.0368 - val_accuracy: 0.9893\n",
      "Epoch 266/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0200 - accuracy: 0.9941 - val_loss: 0.0367 - val_accuracy: 0.9893\n",
      "Epoch 267/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0193 - accuracy: 0.9946 - val_loss: 0.0356 - val_accuracy: 0.9899\n",
      "Epoch 268/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0192 - accuracy: 0.9945 - val_loss: 0.0362 - val_accuracy: 0.9895\n",
      "Epoch 269/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0186 - accuracy: 0.9948 - val_loss: 0.0355 - val_accuracy: 0.9900\n",
      "Epoch 270/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0186 - accuracy: 0.9949 - val_loss: 0.0355 - val_accuracy: 0.9900\n",
      "Epoch 271/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0193 - accuracy: 0.9944 - val_loss: 0.0358 - val_accuracy: 0.9898\n",
      "Epoch 272/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0191 - accuracy: 0.9945 - val_loss: 0.0353 - val_accuracy: 0.9901\n",
      "Epoch 273/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0183 - accuracy: 0.9948 - val_loss: 0.0368 - val_accuracy: 0.9893\n",
      "Epoch 274/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0194 - accuracy: 0.9944 - val_loss: 0.0381 - val_accuracy: 0.9888\n",
      "Epoch 275/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0422 - accuracy: 0.9871 - val_loss: 0.0379 - val_accuracy: 0.9886\n",
      "Epoch 276/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0245 - accuracy: 0.9926 - val_loss: 0.0362 - val_accuracy: 0.9893\n",
      "Epoch 277/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0213 - accuracy: 0.9938 - val_loss: 0.0360 - val_accuracy: 0.9896\n",
      "Epoch 278/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0198 - accuracy: 0.9942 - val_loss: 0.0349 - val_accuracy: 0.9900\n",
      "Epoch 279/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0194 - accuracy: 0.9944 - val_loss: 0.0350 - val_accuracy: 0.9901\n",
      "Epoch 280/1000\n",
      "35/35 [==============================] - 6s 181ms/step - loss: 0.0186 - accuracy: 0.9948 - val_loss: 0.0355 - val_accuracy: 0.9898\n",
      "Epoch 281/1000\n",
      "35/35 [==============================] - 6s 184ms/step - loss: 0.0177 - accuracy: 0.9951 - val_loss: 0.0356 - val_accuracy: 0.9898\n",
      "Epoch 282/1000\n",
      "35/35 [==============================] - 6s 185ms/step - loss: 0.0175 - accuracy: 0.9951 - val_loss: 0.0355 - val_accuracy: 0.9899\n",
      "Epoch 283/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0173 - accuracy: 0.9952 - val_loss: 0.0360 - val_accuracy: 0.9897\n",
      "Epoch 284/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0173 - accuracy: 0.9952 - val_loss: 0.0356 - val_accuracy: 0.9901\n",
      "Epoch 285/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0171 - accuracy: 0.9953 - val_loss: 0.0357 - val_accuracy: 0.9898\n",
      "Epoch 286/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0169 - accuracy: 0.9953 - val_loss: 0.0358 - val_accuracy: 0.9900\n",
      "Epoch 287/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0172 - accuracy: 0.9953 - val_loss: 0.0354 - val_accuracy: 0.9899\n",
      "Epoch 288/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0175 - accuracy: 0.9949 - val_loss: 0.0355 - val_accuracy: 0.9900\n",
      "Epoch 289/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0167 - accuracy: 0.9954 - val_loss: 0.0352 - val_accuracy: 0.9901\n",
      "Epoch 290/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0164 - accuracy: 0.9955 - val_loss: 0.0358 - val_accuracy: 0.9899\n",
      "Epoch 291/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0169 - accuracy: 0.9952 - val_loss: 0.0358 - val_accuracy: 0.9900\n",
      "Epoch 292/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0168 - accuracy: 0.9954 - val_loss: 0.0361 - val_accuracy: 0.9900\n",
      "Epoch 293/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0164 - accuracy: 0.9955 - val_loss: 0.0365 - val_accuracy: 0.9899\n",
      "Epoch 294/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0163 - accuracy: 0.9955 - val_loss: 0.0355 - val_accuracy: 0.9903\n",
      "Epoch 295/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0160 - accuracy: 0.9956 - val_loss: 0.0367 - val_accuracy: 0.9901\n",
      "Epoch 296/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0170 - accuracy: 0.9952 - val_loss: 0.0363 - val_accuracy: 0.9896\n",
      "Epoch 297/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0174 - accuracy: 0.9949 - val_loss: 0.0358 - val_accuracy: 0.9902\n",
      "Epoch 298/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0157 - accuracy: 0.9958 - val_loss: 0.0357 - val_accuracy: 0.9902\n",
      "Epoch 299/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0155 - accuracy: 0.9958 - val_loss: 0.0364 - val_accuracy: 0.9900\n",
      "Epoch 300/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0158 - accuracy: 0.9956 - val_loss: 0.0367 - val_accuracy: 0.9898\n",
      "Epoch 301/1000\n",
      "35/35 [==============================] - 6s 183ms/step - loss: 0.0160 - accuracy: 0.9955 - val_loss: 0.0362 - val_accuracy: 0.9901\n",
      "Epoch 302/1000\n",
      "35/35 [==============================] - 7s 189ms/step - loss: 0.0160 - accuracy: 0.9956 - val_loss: 0.0355 - val_accuracy: 0.9904\n",
      "Epoch 303/1000\n",
      "35/35 [==============================] - 7s 191ms/step - loss: 0.0157 - accuracy: 0.9957 - val_loss: 0.0366 - val_accuracy: 0.9900\n",
      "Epoch 304/1000\n",
      "35/35 [==============================] - 7s 194ms/step - loss: 0.0154 - accuracy: 0.9958 - val_loss: 0.0359 - val_accuracy: 0.9902\n",
      "Epoch 305/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0149 - accuracy: 0.9959 - val_loss: 0.0361 - val_accuracy: 0.9901\n",
      "Epoch 306/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0150 - accuracy: 0.9960 - val_loss: 0.0361 - val_accuracy: 0.9901\n",
      "Epoch 307/1000\n",
      "35/35 [==============================] - 7s 190ms/step - loss: 0.0154 - accuracy: 0.9957 - val_loss: 0.0360 - val_accuracy: 0.9901\n",
      "Epoch 308/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0152 - accuracy: 0.9958 - val_loss: 0.0362 - val_accuracy: 0.9901\n",
      "Epoch 309/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0153 - accuracy: 0.9958 - val_loss: 0.0373 - val_accuracy: 0.9895\n",
      "Epoch 310/1000\n",
      "35/35 [==============================] - 6s 182ms/step - loss: 0.0159 - accuracy: 0.9955 - val_loss: 0.0367 - val_accuracy: 0.9901\n",
      "Epoch 311/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0147 - accuracy: 0.9961 - val_loss: 0.0362 - val_accuracy: 0.9902\n",
      "Epoch 312/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0146 - accuracy: 0.9961 - val_loss: 0.0358 - val_accuracy: 0.9903\n",
      "Epoch 313/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0144 - accuracy: 0.9961 - val_loss: 0.0361 - val_accuracy: 0.9901\n",
      "Epoch 314/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0153 - accuracy: 0.9958 - val_loss: 0.0367 - val_accuracy: 0.9899\n",
      "Epoch 315/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0147 - accuracy: 0.9960 - val_loss: 0.0365 - val_accuracy: 0.9900\n",
      "Epoch 316/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0143 - accuracy: 0.9961 - val_loss: 0.0369 - val_accuracy: 0.9901\n",
      "Epoch 317/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0142 - accuracy: 0.9962 - val_loss: 0.0371 - val_accuracy: 0.9898\n",
      "Epoch 318/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0143 - accuracy: 0.9961 - val_loss: 0.0368 - val_accuracy: 0.9901\n",
      "Epoch 319/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0139 - accuracy: 0.9963 - val_loss: 0.0369 - val_accuracy: 0.9901\n",
      "Epoch 320/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0140 - accuracy: 0.9962 - val_loss: 0.0366 - val_accuracy: 0.9902\n",
      "Epoch 321/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0146 - accuracy: 0.9959 - val_loss: 0.0368 - val_accuracy: 0.9901\n",
      "Epoch 322/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0144 - accuracy: 0.9961 - val_loss: 0.0362 - val_accuracy: 0.9904\n",
      "Epoch 323/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0141 - accuracy: 0.9962 - val_loss: 0.0374 - val_accuracy: 0.9901\n",
      "Epoch 324/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0138 - accuracy: 0.9962 - val_loss: 0.0373 - val_accuracy: 0.9899\n",
      "Epoch 325/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0134 - accuracy: 0.9963 - val_loss: 0.0373 - val_accuracy: 0.9899\n",
      "Epoch 326/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0134 - accuracy: 0.9965 - val_loss: 0.0371 - val_accuracy: 0.9902\n",
      "Epoch 327/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0131 - accuracy: 0.9965 - val_loss: 0.0376 - val_accuracy: 0.9901\n",
      "Epoch 328/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0130 - accuracy: 0.9966 - val_loss: 0.0380 - val_accuracy: 0.9898\n",
      "Epoch 329/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0131 - accuracy: 0.9965 - val_loss: 0.0380 - val_accuracy: 0.9897\n",
      "Epoch 330/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0137 - accuracy: 0.9962 - val_loss: 0.0385 - val_accuracy: 0.9896\n",
      "Epoch 331/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0138 - accuracy: 0.9962 - val_loss: 0.0371 - val_accuracy: 0.9898\n",
      "Epoch 332/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0138 - accuracy: 0.9962 - val_loss: 0.0382 - val_accuracy: 0.9898\n",
      "Epoch 333/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0140 - accuracy: 0.9961 - val_loss: 0.0374 - val_accuracy: 0.9902\n",
      "Epoch 334/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0129 - accuracy: 0.9965 - val_loss: 0.0375 - val_accuracy: 0.9900\n",
      "Epoch 335/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0126 - accuracy: 0.9966 - val_loss: 0.0378 - val_accuracy: 0.9900\n",
      "Epoch 336/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0125 - accuracy: 0.9967 - val_loss: 0.0374 - val_accuracy: 0.9901\n",
      "Epoch 337/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0124 - accuracy: 0.9967 - val_loss: 0.0372 - val_accuracy: 0.9906\n",
      "Epoch 338/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0124 - accuracy: 0.9967 - val_loss: 0.0376 - val_accuracy: 0.9903\n",
      "Epoch 339/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0124 - accuracy: 0.9967 - val_loss: 0.0368 - val_accuracy: 0.9903\n",
      "Epoch 340/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0122 - accuracy: 0.9968 - val_loss: 0.0375 - val_accuracy: 0.9899\n",
      "Epoch 341/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0128 - accuracy: 0.9965 - val_loss: 0.0392 - val_accuracy: 0.9899\n",
      "Epoch 342/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0149 - accuracy: 0.9958 - val_loss: 0.0445 - val_accuracy: 0.9881\n",
      "Epoch 343/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0230 - accuracy: 0.9926 - val_loss: 0.0376 - val_accuracy: 0.9898\n",
      "Epoch 344/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0155 - accuracy: 0.9954 - val_loss: 0.0375 - val_accuracy: 0.9902\n",
      "Epoch 345/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0130 - accuracy: 0.9963 - val_loss: 0.0374 - val_accuracy: 0.9902\n",
      "Epoch 346/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0122 - accuracy: 0.9968 - val_loss: 0.0371 - val_accuracy: 0.9900\n",
      "Epoch 347/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0119 - accuracy: 0.9969 - val_loss: 0.0378 - val_accuracy: 0.9905\n",
      "Epoch 348/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0120 - accuracy: 0.9968 - val_loss: 0.0377 - val_accuracy: 0.9900\n",
      "Epoch 349/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0119 - accuracy: 0.9969 - val_loss: 0.0376 - val_accuracy: 0.9902\n",
      "Epoch 350/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0118 - accuracy: 0.9968 - val_loss: 0.0378 - val_accuracy: 0.9903\n",
      "Epoch 351/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0114 - accuracy: 0.9970 - val_loss: 0.0372 - val_accuracy: 0.9905\n",
      "Epoch 352/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0113 - accuracy: 0.9970 - val_loss: 0.0379 - val_accuracy: 0.9902\n",
      "Epoch 353/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0118 - accuracy: 0.9969 - val_loss: 0.0380 - val_accuracy: 0.9903\n",
      "Epoch 354/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0115 - accuracy: 0.9969 - val_loss: 0.0375 - val_accuracy: 0.9903\n",
      "Epoch 355/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0114 - accuracy: 0.9971 - val_loss: 0.0397 - val_accuracy: 0.9898\n",
      "Epoch 356/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0114 - accuracy: 0.9970 - val_loss: 0.0386 - val_accuracy: 0.9901\n",
      "Epoch 357/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0114 - accuracy: 0.9970 - val_loss: 0.0381 - val_accuracy: 0.9903\n",
      "Epoch 358/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0115 - accuracy: 0.9970 - val_loss: 0.0385 - val_accuracy: 0.9902\n",
      "Epoch 359/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0109 - accuracy: 0.9972 - val_loss: 0.0383 - val_accuracy: 0.9904\n",
      "Epoch 360/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0110 - accuracy: 0.9971 - val_loss: 0.0378 - val_accuracy: 0.9904\n",
      "Epoch 361/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0109 - accuracy: 0.9971 - val_loss: 0.0381 - val_accuracy: 0.9902\n",
      "Epoch 362/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0106 - accuracy: 0.9972 - val_loss: 0.0388 - val_accuracy: 0.9901\n",
      "Epoch 363/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0108 - accuracy: 0.9972 - val_loss: 0.0391 - val_accuracy: 0.9899\n",
      "Epoch 364/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0109 - accuracy: 0.9970 - val_loss: 0.0390 - val_accuracy: 0.9903\n",
      "Epoch 365/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0107 - accuracy: 0.9972 - val_loss: 0.0399 - val_accuracy: 0.9899\n",
      "Epoch 366/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0104 - accuracy: 0.9974 - val_loss: 0.0390 - val_accuracy: 0.9901\n",
      "Epoch 367/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0108 - accuracy: 0.9972 - val_loss: 0.0389 - val_accuracy: 0.9904\n",
      "Epoch 368/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0116 - accuracy: 0.9967 - val_loss: 0.0388 - val_accuracy: 0.9903\n",
      "Epoch 369/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0108 - accuracy: 0.9970 - val_loss: 0.0393 - val_accuracy: 0.9900\n",
      "Epoch 370/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0105 - accuracy: 0.9972 - val_loss: 0.0403 - val_accuracy: 0.9899\n",
      "Epoch 371/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0103 - accuracy: 0.9974 - val_loss: 0.0390 - val_accuracy: 0.9903\n",
      "Epoch 372/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0100 - accuracy: 0.9975 - val_loss: 0.0381 - val_accuracy: 0.9905\n",
      "Epoch 373/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0104 - accuracy: 0.9973 - val_loss: 0.0396 - val_accuracy: 0.9903\n",
      "Epoch 374/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0137 - accuracy: 0.9958 - val_loss: 0.0401 - val_accuracy: 0.9901\n",
      "Epoch 375/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0117 - accuracy: 0.9967 - val_loss: 0.0386 - val_accuracy: 0.9905\n",
      "Epoch 376/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0102 - accuracy: 0.9972 - val_loss: 0.0392 - val_accuracy: 0.9904\n",
      "Epoch 377/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0100 - accuracy: 0.9974 - val_loss: 0.0398 - val_accuracy: 0.9901\n",
      "Epoch 378/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0098 - accuracy: 0.9974 - val_loss: 0.0405 - val_accuracy: 0.9899\n",
      "Epoch 379/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0097 - accuracy: 0.9976 - val_loss: 0.0395 - val_accuracy: 0.9901\n",
      "Epoch 380/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0112 - accuracy: 0.9969 - val_loss: 0.0394 - val_accuracy: 0.9902\n",
      "Epoch 381/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0104 - accuracy: 0.9972 - val_loss: 0.0403 - val_accuracy: 0.9900\n",
      "Epoch 382/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0112 - accuracy: 0.9967 - val_loss: 0.0403 - val_accuracy: 0.9901\n",
      "Epoch 383/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0122 - accuracy: 0.9964 - val_loss: 0.0394 - val_accuracy: 0.9901\n",
      "Epoch 384/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0108 - accuracy: 0.9969 - val_loss: 0.0394 - val_accuracy: 0.9903\n",
      "Epoch 385/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0100 - accuracy: 0.9973 - val_loss: 0.0395 - val_accuracy: 0.9902\n",
      "Epoch 386/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0163 - accuracy: 0.9949 - val_loss: 0.0419 - val_accuracy: 0.9896\n",
      "Epoch 387/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0142 - accuracy: 0.9955 - val_loss: 0.0407 - val_accuracy: 0.9898\n",
      "Epoch 388/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0124 - accuracy: 0.9963 - val_loss: 0.0387 - val_accuracy: 0.9905\n",
      "Epoch 389/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0114 - accuracy: 0.9967 - val_loss: 0.0394 - val_accuracy: 0.9901\n",
      "Epoch 390/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0100 - accuracy: 0.9973 - val_loss: 0.0393 - val_accuracy: 0.9905\n",
      "Epoch 391/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0093 - accuracy: 0.9977 - val_loss: 0.0392 - val_accuracy: 0.9904\n",
      "Epoch 392/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0091 - accuracy: 0.9978 - val_loss: 0.0396 - val_accuracy: 0.9903\n",
      "Epoch 393/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0089 - accuracy: 0.9978 - val_loss: 0.0395 - val_accuracy: 0.9904\n",
      "Epoch 394/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0088 - accuracy: 0.9978 - val_loss: 0.0399 - val_accuracy: 0.9903\n",
      "Epoch 395/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0089 - accuracy: 0.9978 - val_loss: 0.0398 - val_accuracy: 0.9905\n",
      "Epoch 396/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0088 - accuracy: 0.9978 - val_loss: 0.0398 - val_accuracy: 0.9904\n",
      "Epoch 397/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0088 - accuracy: 0.9978 - val_loss: 0.0402 - val_accuracy: 0.9905\n",
      "Epoch 398/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0086 - accuracy: 0.9978 - val_loss: 0.0406 - val_accuracy: 0.9902\n",
      "Epoch 399/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0088 - accuracy: 0.9978 - val_loss: 0.0402 - val_accuracy: 0.9903\n",
      "Epoch 400/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0084 - accuracy: 0.9980 - val_loss: 0.0408 - val_accuracy: 0.9902\n",
      "Epoch 401/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0091 - accuracy: 0.9976 - val_loss: 0.0408 - val_accuracy: 0.9902\n",
      "Epoch 402/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0090 - accuracy: 0.9976 - val_loss: 0.0404 - val_accuracy: 0.9901\n",
      "Epoch 403/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0086 - accuracy: 0.9978 - val_loss: 0.0416 - val_accuracy: 0.9899\n",
      "Epoch 404/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0132 - accuracy: 0.9959 - val_loss: 0.0417 - val_accuracy: 0.9898\n",
      "Epoch 405/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0124 - accuracy: 0.9962 - val_loss: 0.0391 - val_accuracy: 0.9906\n",
      "Epoch 406/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0100 - accuracy: 0.9972 - val_loss: 0.0402 - val_accuracy: 0.9902\n",
      "Epoch 407/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0090 - accuracy: 0.9977 - val_loss: 0.0406 - val_accuracy: 0.9902\n",
      "Epoch 408/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0108 - accuracy: 0.9968 - val_loss: 0.0410 - val_accuracy: 0.9899\n",
      "Epoch 409/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0092 - accuracy: 0.9976 - val_loss: 0.0409 - val_accuracy: 0.9904\n",
      "Epoch 410/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0087 - accuracy: 0.9977 - val_loss: 0.0400 - val_accuracy: 0.9904\n",
      "Epoch 411/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0085 - accuracy: 0.9979 - val_loss: 0.0402 - val_accuracy: 0.9904\n",
      "Epoch 412/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0081 - accuracy: 0.9981 - val_loss: 0.0405 - val_accuracy: 0.9905\n",
      "Epoch 413/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0079 - accuracy: 0.9981 - val_loss: 0.0406 - val_accuracy: 0.9904\n",
      "Epoch 414/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0078 - accuracy: 0.9980 - val_loss: 0.0405 - val_accuracy: 0.9905\n",
      "Epoch 415/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0078 - accuracy: 0.9981 - val_loss: 0.0420 - val_accuracy: 0.9902\n",
      "Epoch 416/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0076 - accuracy: 0.9982 - val_loss: 0.0416 - val_accuracy: 0.9903\n",
      "Epoch 417/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0075 - accuracy: 0.9983 - val_loss: 0.0419 - val_accuracy: 0.9902\n",
      "Epoch 418/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0076 - accuracy: 0.9982 - val_loss: 0.0424 - val_accuracy: 0.9901\n",
      "Epoch 419/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0077 - accuracy: 0.9981 - val_loss: 0.0417 - val_accuracy: 0.9902\n",
      "Epoch 420/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0076 - accuracy: 0.9982 - val_loss: 0.0418 - val_accuracy: 0.9902\n",
      "Epoch 421/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0075 - accuracy: 0.9982 - val_loss: 0.0418 - val_accuracy: 0.9904\n",
      "Epoch 422/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0078 - accuracy: 0.9981 - val_loss: 0.0429 - val_accuracy: 0.9900\n",
      "Epoch 423/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0082 - accuracy: 0.9979 - val_loss: 0.0437 - val_accuracy: 0.9900\n",
      "Epoch 424/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0106 - accuracy: 0.9968 - val_loss: 0.0450 - val_accuracy: 0.9896\n",
      "Epoch 425/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0098 - accuracy: 0.9972 - val_loss: 0.0410 - val_accuracy: 0.9906\n",
      "Epoch 426/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0086 - accuracy: 0.9976 - val_loss: 0.0429 - val_accuracy: 0.9901\n",
      "Epoch 427/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0076 - accuracy: 0.9982 - val_loss: 0.0419 - val_accuracy: 0.9904\n",
      "Epoch 428/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0075 - accuracy: 0.9982 - val_loss: 0.0428 - val_accuracy: 0.9902\n",
      "Epoch 429/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0073 - accuracy: 0.9983 - val_loss: 0.0420 - val_accuracy: 0.9903\n",
      "Epoch 430/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0071 - accuracy: 0.9982 - val_loss: 0.0428 - val_accuracy: 0.9903\n",
      "Epoch 431/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0070 - accuracy: 0.9984 - val_loss: 0.0421 - val_accuracy: 0.9904\n",
      "Epoch 432/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0071 - accuracy: 0.9983 - val_loss: 0.0428 - val_accuracy: 0.9902\n",
      "Epoch 433/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0069 - accuracy: 0.9983 - val_loss: 0.0428 - val_accuracy: 0.9900\n",
      "Epoch 434/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0072 - accuracy: 0.9982 - val_loss: 0.0428 - val_accuracy: 0.9904\n",
      "Epoch 435/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0075 - accuracy: 0.9981 - val_loss: 0.0433 - val_accuracy: 0.9903\n",
      "Epoch 436/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0073 - accuracy: 0.9982 - val_loss: 0.0429 - val_accuracy: 0.9900\n",
      "Epoch 437/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0068 - accuracy: 0.9984 - val_loss: 0.0431 - val_accuracy: 0.9903\n",
      "Epoch 438/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0065 - accuracy: 0.9986 - val_loss: 0.0436 - val_accuracy: 0.9903\n",
      "Epoch 439/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0066 - accuracy: 0.9985 - val_loss: 0.0440 - val_accuracy: 0.9901\n",
      "Epoch 440/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0074 - accuracy: 0.9981 - val_loss: 0.0440 - val_accuracy: 0.9903\n",
      "Epoch 441/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0069 - accuracy: 0.9983 - val_loss: 0.0440 - val_accuracy: 0.9902\n",
      "Epoch 442/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0069 - accuracy: 0.9984 - val_loss: 0.0435 - val_accuracy: 0.9903\n",
      "Epoch 443/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0067 - accuracy: 0.9984 - val_loss: 0.0442 - val_accuracy: 0.9901\n",
      "Epoch 444/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0066 - accuracy: 0.9984 - val_loss: 0.0441 - val_accuracy: 0.9903\n",
      "Epoch 445/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0064 - accuracy: 0.9985 - val_loss: 0.0441 - val_accuracy: 0.9902\n",
      "Epoch 446/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0066 - accuracy: 0.9984 - val_loss: 0.0446 - val_accuracy: 0.9902\n",
      "Epoch 447/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0075 - accuracy: 0.9979 - val_loss: 0.0443 - val_accuracy: 0.9898\n",
      "Epoch 448/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0084 - accuracy: 0.9975 - val_loss: 0.0421 - val_accuracy: 0.9901\n",
      "Epoch 449/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0091 - accuracy: 0.9974 - val_loss: 0.0447 - val_accuracy: 0.9904\n",
      "Epoch 450/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0081 - accuracy: 0.9978 - val_loss: 0.0442 - val_accuracy: 0.9903\n",
      "Epoch 451/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0070 - accuracy: 0.9982 - val_loss: 0.0436 - val_accuracy: 0.9905\n",
      "Epoch 452/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0063 - accuracy: 0.9985 - val_loss: 0.0440 - val_accuracy: 0.9903\n",
      "Epoch 453/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0062 - accuracy: 0.9985 - val_loss: 0.0458 - val_accuracy: 0.9901\n",
      "Epoch 454/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0074 - accuracy: 0.9980 - val_loss: 0.0435 - val_accuracy: 0.9904\n",
      "Epoch 455/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0069 - accuracy: 0.9982 - val_loss: 0.0442 - val_accuracy: 0.9903\n",
      "Epoch 456/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0064 - accuracy: 0.9985 - val_loss: 0.0446 - val_accuracy: 0.9901\n",
      "Epoch 457/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0061 - accuracy: 0.9986 - val_loss: 0.0449 - val_accuracy: 0.9902\n",
      "Epoch 458/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0059 - accuracy: 0.9986 - val_loss: 0.0453 - val_accuracy: 0.9902\n",
      "Epoch 459/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0069 - accuracy: 0.9982 - val_loss: 0.0449 - val_accuracy: 0.9904\n",
      "Epoch 460/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0066 - accuracy: 0.9984 - val_loss: 0.0444 - val_accuracy: 0.9904\n",
      "Epoch 461/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0060 - accuracy: 0.9986 - val_loss: 0.0458 - val_accuracy: 0.9901\n",
      "Epoch 462/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0062 - accuracy: 0.9985 - val_loss: 0.0444 - val_accuracy: 0.9902\n",
      "Epoch 463/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0058 - accuracy: 0.9987 - val_loss: 0.0459 - val_accuracy: 0.9903\n",
      "Epoch 464/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0058 - accuracy: 0.9987 - val_loss: 0.0459 - val_accuracy: 0.9900\n",
      "Epoch 465/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0057 - accuracy: 0.9988 - val_loss: 0.0464 - val_accuracy: 0.9901\n",
      "Epoch 466/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0057 - accuracy: 0.9988 - val_loss: 0.0456 - val_accuracy: 0.9902\n",
      "Epoch 467/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0056 - accuracy: 0.9987 - val_loss: 0.0449 - val_accuracy: 0.9903\n",
      "Epoch 468/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0055 - accuracy: 0.9988 - val_loss: 0.0461 - val_accuracy: 0.9904\n",
      "Epoch 469/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0055 - accuracy: 0.9987 - val_loss: 0.0464 - val_accuracy: 0.9899\n",
      "Epoch 470/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0056 - accuracy: 0.9987 - val_loss: 0.0468 - val_accuracy: 0.9902\n",
      "Epoch 471/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0057 - accuracy: 0.9987 - val_loss: 0.0465 - val_accuracy: 0.9903\n",
      "Epoch 472/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0056 - accuracy: 0.9988 - val_loss: 0.0459 - val_accuracy: 0.9904\n",
      "Epoch 473/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0055 - accuracy: 0.9988 - val_loss: 0.0469 - val_accuracy: 0.9903\n",
      "Epoch 474/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0052 - accuracy: 0.9989 - val_loss: 0.0469 - val_accuracy: 0.9901\n",
      "Epoch 475/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0054 - accuracy: 0.9988 - val_loss: 0.0464 - val_accuracy: 0.9904\n",
      "Epoch 476/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0055 - accuracy: 0.9987 - val_loss: 0.0462 - val_accuracy: 0.9903\n",
      "Epoch 477/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0057 - accuracy: 0.9987 - val_loss: 0.0474 - val_accuracy: 0.9900\n",
      "Epoch 478/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0062 - accuracy: 0.9983 - val_loss: 0.0481 - val_accuracy: 0.9902\n",
      "Epoch 479/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0061 - accuracy: 0.9985 - val_loss: 0.0481 - val_accuracy: 0.9898\n",
      "Epoch 480/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0086 - accuracy: 0.9975 - val_loss: 0.0497 - val_accuracy: 0.9898\n",
      "Epoch 481/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0176 - accuracy: 0.9947 - val_loss: 0.0519 - val_accuracy: 0.9885\n",
      "Epoch 482/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0158 - accuracy: 0.9947 - val_loss: 0.0446 - val_accuracy: 0.9889\n",
      "Epoch 483/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0123 - accuracy: 0.9960 - val_loss: 0.0441 - val_accuracy: 0.9903\n",
      "Epoch 484/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0076 - accuracy: 0.9978 - val_loss: 0.0433 - val_accuracy: 0.9904\n",
      "Epoch 485/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0058 - accuracy: 0.9986 - val_loss: 0.0445 - val_accuracy: 0.9903\n",
      "Epoch 486/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0055 - accuracy: 0.9987 - val_loss: 0.0449 - val_accuracy: 0.9903\n",
      "Epoch 487/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0052 - accuracy: 0.9988 - val_loss: 0.0454 - val_accuracy: 0.9903\n",
      "Epoch 488/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0052 - accuracy: 0.9988 - val_loss: 0.0454 - val_accuracy: 0.9904\n",
      "Epoch 489/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0050 - accuracy: 0.9989 - val_loss: 0.0456 - val_accuracy: 0.9903\n",
      "Epoch 490/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0048 - accuracy: 0.9991 - val_loss: 0.0460 - val_accuracy: 0.9903\n",
      "Epoch 491/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0047 - accuracy: 0.9991 - val_loss: 0.0460 - val_accuracy: 0.9905\n",
      "Epoch 492/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0046 - accuracy: 0.9991 - val_loss: 0.0460 - val_accuracy: 0.9904\n",
      "Epoch 493/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0047 - accuracy: 0.9991 - val_loss: 0.0464 - val_accuracy: 0.9905\n",
      "Epoch 494/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0046 - accuracy: 0.9991 - val_loss: 0.0468 - val_accuracy: 0.9904\n",
      "Epoch 495/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0046 - accuracy: 0.9991 - val_loss: 0.0472 - val_accuracy: 0.9901\n",
      "Epoch 496/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0047 - accuracy: 0.9990 - val_loss: 0.0466 - val_accuracy: 0.9903\n",
      "Epoch 497/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0046 - accuracy: 0.9991 - val_loss: 0.0475 - val_accuracy: 0.9902\n",
      "Epoch 498/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0045 - accuracy: 0.9991 - val_loss: 0.0470 - val_accuracy: 0.9903\n",
      "Epoch 499/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0044 - accuracy: 0.9992 - val_loss: 0.0472 - val_accuracy: 0.9903\n",
      "Epoch 500/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0045 - accuracy: 0.9992 - val_loss: 0.0469 - val_accuracy: 0.9903\n",
      "Epoch 501/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0046 - accuracy: 0.9991 - val_loss: 0.0479 - val_accuracy: 0.9902\n",
      "Epoch 502/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0045 - accuracy: 0.9992 - val_loss: 0.0480 - val_accuracy: 0.9904\n",
      "Epoch 503/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0043 - accuracy: 0.9992 - val_loss: 0.0479 - val_accuracy: 0.9903\n",
      "Epoch 504/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0043 - accuracy: 0.9992 - val_loss: 0.0479 - val_accuracy: 0.9903\n",
      "Epoch 505/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0044 - accuracy: 0.9991 - val_loss: 0.0475 - val_accuracy: 0.9904\n",
      "Epoch 506/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0043 - accuracy: 0.9992 - val_loss: 0.0486 - val_accuracy: 0.9902\n",
      "Epoch 507/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0043 - accuracy: 0.9992 - val_loss: 0.0484 - val_accuracy: 0.9902\n",
      "Epoch 508/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0046 - accuracy: 0.9990 - val_loss: 0.0491 - val_accuracy: 0.9900\n",
      "Epoch 509/1000\n",
      "35/35 [==============================] - 6s 168ms/step - loss: 0.0048 - accuracy: 0.9989 - val_loss: 0.0482 - val_accuracy: 0.9906\n",
      "Epoch 510/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0047 - accuracy: 0.9990 - val_loss: 0.0489 - val_accuracy: 0.9903\n",
      "Epoch 511/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0051 - accuracy: 0.9987 - val_loss: 0.0473 - val_accuracy: 0.9903\n",
      "Epoch 512/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0046 - accuracy: 0.9989 - val_loss: 0.0493 - val_accuracy: 0.9900\n",
      "Epoch 513/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0050 - accuracy: 0.9988 - val_loss: 0.0504 - val_accuracy: 0.9899\n",
      "Epoch 514/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0057 - accuracy: 0.9984 - val_loss: 0.0493 - val_accuracy: 0.9900\n",
      "Epoch 515/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0051 - accuracy: 0.9987 - val_loss: 0.0482 - val_accuracy: 0.9904\n",
      "Epoch 516/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0047 - accuracy: 0.9990 - val_loss: 0.0493 - val_accuracy: 0.9902\n",
      "Epoch 517/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0042 - accuracy: 0.9992 - val_loss: 0.0493 - val_accuracy: 0.9902\n",
      "Epoch 518/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0041 - accuracy: 0.9992 - val_loss: 0.0495 - val_accuracy: 0.9903\n",
      "Epoch 519/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0039 - accuracy: 0.9993 - val_loss: 0.0492 - val_accuracy: 0.9904\n",
      "Epoch 520/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0038 - accuracy: 0.9993 - val_loss: 0.0497 - val_accuracy: 0.9902\n",
      "Epoch 521/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0039 - accuracy: 0.9993 - val_loss: 0.0495 - val_accuracy: 0.9904\n",
      "Epoch 522/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0038 - accuracy: 0.9993 - val_loss: 0.0496 - val_accuracy: 0.9904\n",
      "Epoch 523/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0038 - accuracy: 0.9993 - val_loss: 0.0497 - val_accuracy: 0.9902\n",
      "Epoch 524/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0040 - accuracy: 0.9992 - val_loss: 0.0497 - val_accuracy: 0.9903\n",
      "Epoch 525/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0045 - accuracy: 0.9989 - val_loss: 0.0526 - val_accuracy: 0.9898\n",
      "Epoch 526/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0063 - accuracy: 0.9982 - val_loss: 0.0493 - val_accuracy: 0.9901\n",
      "Epoch 527/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0057 - accuracy: 0.9985 - val_loss: 0.0492 - val_accuracy: 0.9904\n",
      "Epoch 528/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0055 - accuracy: 0.9985 - val_loss: 0.0489 - val_accuracy: 0.9903\n",
      "Epoch 529/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0044 - accuracy: 0.9989 - val_loss: 0.0493 - val_accuracy: 0.9904\n",
      "Epoch 530/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0040 - accuracy: 0.9992 - val_loss: 0.0496 - val_accuracy: 0.9903\n",
      "Epoch 531/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0039 - accuracy: 0.9992 - val_loss: 0.0492 - val_accuracy: 0.9903\n",
      "Epoch 532/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0037 - accuracy: 0.9994 - val_loss: 0.0496 - val_accuracy: 0.9902\n",
      "Epoch 533/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0037 - accuracy: 0.9993 - val_loss: 0.0501 - val_accuracy: 0.9903\n",
      "Epoch 534/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0038 - accuracy: 0.9993 - val_loss: 0.0507 - val_accuracy: 0.9903\n",
      "Epoch 535/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0037 - accuracy: 0.9993 - val_loss: 0.0498 - val_accuracy: 0.9904\n",
      "Epoch 536/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0034 - accuracy: 0.9995 - val_loss: 0.0504 - val_accuracy: 0.9903\n",
      "Epoch 537/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0035 - accuracy: 0.9994 - val_loss: 0.0507 - val_accuracy: 0.9903\n",
      "Epoch 538/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0512 - val_accuracy: 0.9902\n",
      "Epoch 539/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0512 - val_accuracy: 0.9905\n",
      "Epoch 540/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0518 - val_accuracy: 0.9902\n",
      "Epoch 541/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0034 - accuracy: 0.9995 - val_loss: 0.0514 - val_accuracy: 0.9905\n",
      "Epoch 542/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0509 - val_accuracy: 0.9904\n",
      "Epoch 543/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0033 - accuracy: 0.9994 - val_loss: 0.0523 - val_accuracy: 0.9903\n",
      "Epoch 544/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0078 - accuracy: 0.9978 - val_loss: 0.0533 - val_accuracy: 0.9893\n",
      "Epoch 545/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0147 - accuracy: 0.9954 - val_loss: 0.0542 - val_accuracy: 0.9885\n",
      "Epoch 546/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0121 - accuracy: 0.9961 - val_loss: 0.0489 - val_accuracy: 0.9903\n",
      "Epoch 547/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0199 - accuracy: 0.9940 - val_loss: 0.0482 - val_accuracy: 0.9895\n",
      "Epoch 548/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0128 - accuracy: 0.9959 - val_loss: 0.0453 - val_accuracy: 0.9902\n",
      "Epoch 549/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0080 - accuracy: 0.9974 - val_loss: 0.0460 - val_accuracy: 0.9904\n",
      "Epoch 550/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0061 - accuracy: 0.9983 - val_loss: 0.0468 - val_accuracy: 0.9902\n",
      "Epoch 551/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0055 - accuracy: 0.9985 - val_loss: 0.0484 - val_accuracy: 0.9902\n",
      "Epoch 552/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0050 - accuracy: 0.9987 - val_loss: 0.0479 - val_accuracy: 0.9905\n",
      "Epoch 553/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0042 - accuracy: 0.9991 - val_loss: 0.0485 - val_accuracy: 0.9902\n",
      "Epoch 554/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0038 - accuracy: 0.9992 - val_loss: 0.0487 - val_accuracy: 0.9903\n",
      "Epoch 555/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0037 - accuracy: 0.9993 - val_loss: 0.0493 - val_accuracy: 0.9903\n",
      "Epoch 556/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0035 - accuracy: 0.9994 - val_loss: 0.0494 - val_accuracy: 0.9903\n",
      "Epoch 557/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0035 - accuracy: 0.9994 - val_loss: 0.0496 - val_accuracy: 0.9903\n",
      "Epoch 558/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0034 - accuracy: 0.9994 - val_loss: 0.0498 - val_accuracy: 0.9905\n",
      "Epoch 559/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0500 - val_accuracy: 0.9903\n",
      "Epoch 560/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0033 - accuracy: 0.9995 - val_loss: 0.0502 - val_accuracy: 0.9904\n",
      "Epoch 561/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0032 - accuracy: 0.9995 - val_loss: 0.0504 - val_accuracy: 0.9903\n",
      "Epoch 562/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0031 - accuracy: 0.9996 - val_loss: 0.0505 - val_accuracy: 0.9904\n",
      "Epoch 563/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0031 - accuracy: 0.9995 - val_loss: 0.0508 - val_accuracy: 0.9904\n",
      "Epoch 564/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0031 - accuracy: 0.9995 - val_loss: 0.0515 - val_accuracy: 0.9902\n",
      "Epoch 565/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0030 - accuracy: 0.9995 - val_loss: 0.0513 - val_accuracy: 0.9904\n",
      "Epoch 566/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0031 - accuracy: 0.9995 - val_loss: 0.0514 - val_accuracy: 0.9904\n",
      "Epoch 567/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0030 - accuracy: 0.9996 - val_loss: 0.0518 - val_accuracy: 0.9903\n",
      "Epoch 568/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0030 - accuracy: 0.9996 - val_loss: 0.0518 - val_accuracy: 0.9903\n",
      "Epoch 569/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0029 - accuracy: 0.9996 - val_loss: 0.0515 - val_accuracy: 0.9904\n",
      "Epoch 570/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0029 - accuracy: 0.9996 - val_loss: 0.0517 - val_accuracy: 0.9903\n",
      "Epoch 571/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0028 - accuracy: 0.9997 - val_loss: 0.0519 - val_accuracy: 0.9904\n",
      "Epoch 572/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0523 - val_accuracy: 0.9903\n",
      "Epoch 573/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0522 - val_accuracy: 0.9904\n",
      "Epoch 574/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0029 - accuracy: 0.9996 - val_loss: 0.0522 - val_accuracy: 0.9902\n",
      "Epoch 575/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0029 - accuracy: 0.9995 - val_loss: 0.0523 - val_accuracy: 0.9902\n",
      "Epoch 576/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0035 - accuracy: 0.9994 - val_loss: 0.0520 - val_accuracy: 0.9903\n",
      "Epoch 577/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0031 - accuracy: 0.9995 - val_loss: 0.0524 - val_accuracy: 0.9903\n",
      "Epoch 578/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0527 - val_accuracy: 0.9904\n",
      "Epoch 579/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0523 - val_accuracy: 0.9904\n",
      "Epoch 580/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0027 - accuracy: 0.9996 - val_loss: 0.0522 - val_accuracy: 0.9903\n",
      "Epoch 581/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0532 - val_accuracy: 0.9903\n",
      "Epoch 582/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0533 - val_accuracy: 0.9902\n",
      "Epoch 583/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0028 - accuracy: 0.9996 - val_loss: 0.0535 - val_accuracy: 0.9903\n",
      "Epoch 584/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0027 - accuracy: 0.9996 - val_loss: 0.0530 - val_accuracy: 0.9904\n",
      "Epoch 585/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0532 - val_accuracy: 0.9905\n",
      "Epoch 586/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0535 - val_accuracy: 0.9902\n",
      "Epoch 587/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0541 - val_accuracy: 0.9900\n",
      "Epoch 588/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0544 - val_accuracy: 0.9902\n",
      "Epoch 589/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0536 - val_accuracy: 0.9903\n",
      "Epoch 590/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0025 - accuracy: 0.9996 - val_loss: 0.0537 - val_accuracy: 0.9904\n",
      "Epoch 591/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0537 - val_accuracy: 0.9905\n",
      "Epoch 592/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0547 - val_accuracy: 0.9902\n",
      "Epoch 593/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0029 - accuracy: 0.9995 - val_loss: 0.0534 - val_accuracy: 0.9903\n",
      "Epoch 594/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0030 - accuracy: 0.9994 - val_loss: 0.0550 - val_accuracy: 0.9902\n",
      "Epoch 595/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0075 - accuracy: 0.9979 - val_loss: 0.0538 - val_accuracy: 0.9896\n",
      "Epoch 596/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0084 - accuracy: 0.9972 - val_loss: 0.0522 - val_accuracy: 0.9901\n",
      "Epoch 597/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0065 - accuracy: 0.9979 - val_loss: 0.0500 - val_accuracy: 0.9905\n",
      "Epoch 598/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0058 - accuracy: 0.9982 - val_loss: 0.0505 - val_accuracy: 0.9905\n",
      "Epoch 599/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0040 - accuracy: 0.9990 - val_loss: 0.0512 - val_accuracy: 0.9904\n",
      "Epoch 600/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0034 - accuracy: 0.9992 - val_loss: 0.0513 - val_accuracy: 0.9905\n",
      "Epoch 601/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0029 - accuracy: 0.9995 - val_loss: 0.0516 - val_accuracy: 0.9906\n",
      "Epoch 602/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0028 - accuracy: 0.9995 - val_loss: 0.0522 - val_accuracy: 0.9905\n",
      "Epoch 603/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0026 - accuracy: 0.9996 - val_loss: 0.0532 - val_accuracy: 0.9904\n",
      "Epoch 604/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0025 - accuracy: 0.9997 - val_loss: 0.0533 - val_accuracy: 0.9903\n",
      "Epoch 605/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0024 - accuracy: 0.9997 - val_loss: 0.0534 - val_accuracy: 0.9904\n",
      "Epoch 606/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0532 - val_accuracy: 0.9905\n",
      "Epoch 607/1000\n",
      "35/35 [==============================] - 6s 183ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0539 - val_accuracy: 0.9903\n",
      "Epoch 608/1000\n",
      "35/35 [==============================] - 6s 185ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0541 - val_accuracy: 0.9904\n",
      "Epoch 609/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0024 - accuracy: 0.9996 - val_loss: 0.0541 - val_accuracy: 0.9904\n",
      "Epoch 610/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0541 - val_accuracy: 0.9904\n",
      "Epoch 611/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0541 - val_accuracy: 0.9905\n",
      "Epoch 612/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0548 - val_accuracy: 0.9903\n",
      "Epoch 613/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0550 - val_accuracy: 0.9904\n",
      "Epoch 614/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0022 - accuracy: 0.9997 - val_loss: 0.0554 - val_accuracy: 0.9902\n",
      "Epoch 615/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0022 - accuracy: 0.9997 - val_loss: 0.0555 - val_accuracy: 0.9903\n",
      "Epoch 616/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0022 - accuracy: 0.9997 - val_loss: 0.0550 - val_accuracy: 0.9903\n",
      "Epoch 617/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0022 - accuracy: 0.9997 - val_loss: 0.0548 - val_accuracy: 0.9903\n",
      "Epoch 618/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0551 - val_accuracy: 0.9905\n",
      "Epoch 619/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0559 - val_accuracy: 0.9903\n",
      "Epoch 620/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0559 - val_accuracy: 0.9903\n",
      "Epoch 621/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0559 - val_accuracy: 0.9902\n",
      "Epoch 622/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0561 - val_accuracy: 0.9903\n",
      "Epoch 623/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0027 - accuracy: 0.9994 - val_loss: 0.0560 - val_accuracy: 0.9903\n",
      "Epoch 624/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0034 - accuracy: 0.9992 - val_loss: 0.0551 - val_accuracy: 0.9902\n",
      "Epoch 625/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0033 - accuracy: 0.9992 - val_loss: 0.0582 - val_accuracy: 0.9901\n",
      "Epoch 626/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0083 - accuracy: 0.9975 - val_loss: 0.0606 - val_accuracy: 0.9888\n",
      "Epoch 627/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0153 - accuracy: 0.9952 - val_loss: 0.0549 - val_accuracy: 0.9893\n",
      "Epoch 628/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0080 - accuracy: 0.9974 - val_loss: 0.0516 - val_accuracy: 0.9905\n",
      "Epoch 629/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0054 - accuracy: 0.9984 - val_loss: 0.0519 - val_accuracy: 0.9905\n",
      "Epoch 630/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0033 - accuracy: 0.9992 - val_loss: 0.0528 - val_accuracy: 0.9904\n",
      "Epoch 631/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0028 - accuracy: 0.9995 - val_loss: 0.0523 - val_accuracy: 0.9907\n",
      "Epoch 632/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0025 - accuracy: 0.9996 - val_loss: 0.0542 - val_accuracy: 0.9904\n",
      "Epoch 633/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0023 - accuracy: 0.9997 - val_loss: 0.0544 - val_accuracy: 0.9904\n",
      "Epoch 634/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0022 - accuracy: 0.9997 - val_loss: 0.0544 - val_accuracy: 0.9904\n",
      "Epoch 635/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0547 - val_accuracy: 0.9904\n",
      "Epoch 636/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0020 - accuracy: 0.9997 - val_loss: 0.0552 - val_accuracy: 0.9904\n",
      "Epoch 637/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0020 - accuracy: 0.9997 - val_loss: 0.0550 - val_accuracy: 0.9905\n",
      "Epoch 638/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0020 - accuracy: 0.9997 - val_loss: 0.0552 - val_accuracy: 0.9905\n",
      "Epoch 639/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0020 - accuracy: 0.9998 - val_loss: 0.0554 - val_accuracy: 0.9905\n",
      "Epoch 640/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0019 - accuracy: 0.9998 - val_loss: 0.0554 - val_accuracy: 0.9904\n",
      "Epoch 641/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0019 - accuracy: 0.9997 - val_loss: 0.0559 - val_accuracy: 0.9904\n",
      "Epoch 642/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0019 - accuracy: 0.9997 - val_loss: 0.0555 - val_accuracy: 0.9906\n",
      "Epoch 643/1000\n",
      "35/35 [==============================] - 7s 197ms/step - loss: 0.0019 - accuracy: 0.9998 - val_loss: 0.0558 - val_accuracy: 0.9905\n",
      "Epoch 644/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0019 - accuracy: 0.9998 - val_loss: 0.0561 - val_accuracy: 0.9904\n",
      "Epoch 645/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0019 - accuracy: 0.9998 - val_loss: 0.0562 - val_accuracy: 0.9905\n",
      "Epoch 646/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0566 - val_accuracy: 0.9904\n",
      "Epoch 647/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0563 - val_accuracy: 0.9904\n",
      "Epoch 648/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0568 - val_accuracy: 0.9905\n",
      "Epoch 649/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0569 - val_accuracy: 0.9904\n",
      "Epoch 650/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0567 - val_accuracy: 0.9904\n",
      "Epoch 651/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0574 - val_accuracy: 0.9904\n",
      "Epoch 652/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0018 - accuracy: 0.9998 - val_loss: 0.0575 - val_accuracy: 0.9903\n",
      "Epoch 653/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0018 - accuracy: 0.9997 - val_loss: 0.0574 - val_accuracy: 0.9904\n",
      "Epoch 654/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 0.0018 - accuracy: 0.9997 - val_loss: 0.0574 - val_accuracy: 0.9904\n",
      "Epoch 655/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0018 - accuracy: 0.9997 - val_loss: 0.0579 - val_accuracy: 0.9904\n",
      "Epoch 656/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0575 - val_accuracy: 0.9903\n",
      "Epoch 657/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0575 - val_accuracy: 0.9904\n",
      "Epoch 658/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0575 - val_accuracy: 0.9903\n",
      "Epoch 659/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0581 - val_accuracy: 0.9904\n",
      "Epoch 660/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0586 - val_accuracy: 0.9904\n",
      "Epoch 661/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0581 - val_accuracy: 0.9903\n",
      "Epoch 662/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0583 - val_accuracy: 0.9903\n",
      "Epoch 663/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0018 - accuracy: 0.9997 - val_loss: 0.0580 - val_accuracy: 0.9903\n",
      "Epoch 664/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0021 - accuracy: 0.9996 - val_loss: 0.0583 - val_accuracy: 0.9905\n",
      "Epoch 665/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0035 - accuracy: 0.9991 - val_loss: 0.0591 - val_accuracy: 0.9900\n",
      "Epoch 666/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0093 - accuracy: 0.9971 - val_loss: 0.0625 - val_accuracy: 0.9892\n",
      "Epoch 667/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0123 - accuracy: 0.9959 - val_loss: 0.0544 - val_accuracy: 0.9897\n",
      "Epoch 668/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0090 - accuracy: 0.9971 - val_loss: 0.0539 - val_accuracy: 0.9898\n",
      "Epoch 669/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0052 - accuracy: 0.9983 - val_loss: 0.0524 - val_accuracy: 0.9905\n",
      "Epoch 670/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0047 - accuracy: 0.9986 - val_loss: 0.0525 - val_accuracy: 0.9903\n",
      "Epoch 671/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0032 - accuracy: 0.9992 - val_loss: 0.0538 - val_accuracy: 0.9906\n",
      "Epoch 672/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0024 - accuracy: 0.9995 - val_loss: 0.0542 - val_accuracy: 0.9907\n",
      "Epoch 673/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0020 - accuracy: 0.9997 - val_loss: 0.0547 - val_accuracy: 0.9908\n",
      "Epoch 674/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0020 - accuracy: 0.9997 - val_loss: 0.0555 - val_accuracy: 0.9904\n",
      "Epoch 675/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0018 - accuracy: 0.9997 - val_loss: 0.0556 - val_accuracy: 0.9905\n",
      "Epoch 676/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0555 - val_accuracy: 0.9905\n",
      "Epoch 677/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0017 - accuracy: 0.9998 - val_loss: 0.0559 - val_accuracy: 0.9906\n",
      "Epoch 678/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0016 - accuracy: 0.9998 - val_loss: 0.0565 - val_accuracy: 0.9905\n",
      "Epoch 679/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0016 - accuracy: 0.9998 - val_loss: 0.0562 - val_accuracy: 0.9906\n",
      "Epoch 680/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0016 - accuracy: 0.9998 - val_loss: 0.0562 - val_accuracy: 0.9906\n",
      "Epoch 681/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0016 - accuracy: 0.9998 - val_loss: 0.0570 - val_accuracy: 0.9906\n",
      "Epoch 682/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0569 - val_accuracy: 0.9906\n",
      "Epoch 683/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0572 - val_accuracy: 0.9905\n",
      "Epoch 684/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0572 - val_accuracy: 0.9905\n",
      "Epoch 685/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0576 - val_accuracy: 0.9905\n",
      "Epoch 686/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0569 - val_accuracy: 0.9906\n",
      "Epoch 687/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0577 - val_accuracy: 0.9905\n",
      "Epoch 688/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0578 - val_accuracy: 0.9905\n",
      "Epoch 689/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0580 - val_accuracy: 0.9905\n",
      "Epoch 690/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0581 - val_accuracy: 0.9904\n",
      "Epoch 691/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0575 - val_accuracy: 0.9905\n",
      "Epoch 692/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0580 - val_accuracy: 0.9905\n",
      "Epoch 693/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0580 - val_accuracy: 0.9904\n",
      "Epoch 694/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0583 - val_accuracy: 0.9904\n",
      "Epoch 695/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0588 - val_accuracy: 0.9904\n",
      "Epoch 696/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0583 - val_accuracy: 0.9905\n",
      "Epoch 697/1000\n",
      "35/35 [==============================] - 7s 194ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0590 - val_accuracy: 0.9903\n",
      "Epoch 698/1000\n",
      "35/35 [==============================] - 6s 186ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0588 - val_accuracy: 0.9903\n",
      "Epoch 699/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0589 - val_accuracy: 0.9904\n",
      "Epoch 700/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0590 - val_accuracy: 0.9904\n",
      "Epoch 701/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0590 - val_accuracy: 0.9904\n",
      "Epoch 702/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0592 - val_accuracy: 0.9905\n",
      "Epoch 703/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0594 - val_accuracy: 0.9905\n",
      "Epoch 704/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0016 - accuracy: 0.9997 - val_loss: 0.0588 - val_accuracy: 0.9906\n",
      "Epoch 705/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0020 - accuracy: 0.9996 - val_loss: 0.0622 - val_accuracy: 0.9899\n",
      "Epoch 706/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0117 - accuracy: 0.9965 - val_loss: 0.0627 - val_accuracy: 0.9883\n",
      "Epoch 707/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0111 - accuracy: 0.9964 - val_loss: 0.0584 - val_accuracy: 0.9892\n",
      "Epoch 708/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0069 - accuracy: 0.9978 - val_loss: 0.0544 - val_accuracy: 0.9902\n",
      "Epoch 709/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0107 - accuracy: 0.9965 - val_loss: 0.0531 - val_accuracy: 0.9908\n",
      "Epoch 710/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0035 - accuracy: 0.9989 - val_loss: 0.0559 - val_accuracy: 0.9905\n",
      "Epoch 711/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0024 - accuracy: 0.9994 - val_loss: 0.0546 - val_accuracy: 0.9908\n",
      "Epoch 712/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0019 - accuracy: 0.9997 - val_loss: 0.0554 - val_accuracy: 0.9907\n",
      "Epoch 713/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0016 - accuracy: 0.9998 - val_loss: 0.0561 - val_accuracy: 0.9906\n",
      "Epoch 714/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0016 - accuracy: 0.9998 - val_loss: 0.0565 - val_accuracy: 0.9905\n",
      "Epoch 715/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0015 - accuracy: 0.9998 - val_loss: 0.0567 - val_accuracy: 0.9906\n",
      "Epoch 716/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0566 - val_accuracy: 0.9906\n",
      "Epoch 717/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0571 - val_accuracy: 0.9905\n",
      "Epoch 718/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0014 - accuracy: 0.9998 - val_loss: 0.0573 - val_accuracy: 0.9905\n",
      "Epoch 719/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0574 - val_accuracy: 0.9906\n",
      "Epoch 720/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0576 - val_accuracy: 0.9905\n",
      "Epoch 721/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0013 - accuracy: 0.9999 - val_loss: 0.0578 - val_accuracy: 0.9906\n",
      "Epoch 722/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0580 - val_accuracy: 0.9905\n",
      "Epoch 723/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0013 - accuracy: 0.9999 - val_loss: 0.0579 - val_accuracy: 0.9906\n",
      "Epoch 724/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0581 - val_accuracy: 0.9905\n",
      "Epoch 725/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0580 - val_accuracy: 0.9906\n",
      "Epoch 726/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0584 - val_accuracy: 0.9906\n",
      "Epoch 727/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0012 - accuracy: 0.9998 - val_loss: 0.0586 - val_accuracy: 0.9905\n",
      "Epoch 728/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0587 - val_accuracy: 0.9905\n",
      "Epoch 729/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0590 - val_accuracy: 0.9906\n",
      "Epoch 730/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0589 - val_accuracy: 0.9905\n",
      "Epoch 731/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0590 - val_accuracy: 0.9905\n",
      "Epoch 732/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0594 - val_accuracy: 0.9905\n",
      "Epoch 733/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0594 - val_accuracy: 0.9904\n",
      "Epoch 734/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0598 - val_accuracy: 0.9905\n",
      "Epoch 735/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0594 - val_accuracy: 0.9905\n",
      "Epoch 736/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0602 - val_accuracy: 0.9905\n",
      "Epoch 737/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0597 - val_accuracy: 0.9906\n",
      "Epoch 738/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0601 - val_accuracy: 0.9905\n",
      "Epoch 739/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0598 - val_accuracy: 0.9905\n",
      "Epoch 740/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0600 - val_accuracy: 0.9905\n",
      "Epoch 741/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0604 - val_accuracy: 0.9904\n",
      "Epoch 742/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0605 - val_accuracy: 0.9904\n",
      "Epoch 743/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0605 - val_accuracy: 0.9905\n",
      "Epoch 744/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0606 - val_accuracy: 0.9906\n",
      "Epoch 745/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0612 - val_accuracy: 0.9904\n",
      "Epoch 746/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0610 - val_accuracy: 0.9904\n",
      "Epoch 747/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0010 - accuracy: 0.9999 - val_loss: 0.0610 - val_accuracy: 0.9904\n",
      "Epoch 748/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0010 - accuracy: 0.9999 - val_loss: 0.0607 - val_accuracy: 0.9905\n",
      "Epoch 749/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0612 - val_accuracy: 0.9905\n",
      "Epoch 750/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0622 - val_accuracy: 0.9904\n",
      "Epoch 751/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0615 - val_accuracy: 0.9904\n",
      "Epoch 752/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0012 - accuracy: 0.9998 - val_loss: 0.0619 - val_accuracy: 0.9905\n",
      "Epoch 753/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0625 - val_accuracy: 0.9903\n",
      "Epoch 754/1000\n",
      "35/35 [==============================] - 6s 183ms/step - loss: 0.0015 - accuracy: 0.9997 - val_loss: 0.0611 - val_accuracy: 0.9904\n",
      "Epoch 755/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0613 - val_accuracy: 0.9906\n",
      "Epoch 756/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0012 - accuracy: 0.9998 - val_loss: 0.0613 - val_accuracy: 0.9906\n",
      "Epoch 757/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0012 - accuracy: 0.9998 - val_loss: 0.0629 - val_accuracy: 0.9904\n",
      "Epoch 758/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0015 - accuracy: 0.9997 - val_loss: 0.0637 - val_accuracy: 0.9900\n",
      "Epoch 759/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0041 - accuracy: 0.9988 - val_loss: 0.0615 - val_accuracy: 0.9899\n",
      "Epoch 760/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0086 - accuracy: 0.9971 - val_loss: 0.0610 - val_accuracy: 0.9895\n",
      "Epoch 761/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0090 - accuracy: 0.9970 - val_loss: 0.0626 - val_accuracy: 0.9897\n",
      "Epoch 762/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0054 - accuracy: 0.9982 - val_loss: 0.0579 - val_accuracy: 0.9902\n",
      "Epoch 763/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0026 - accuracy: 0.9993 - val_loss: 0.0586 - val_accuracy: 0.9906\n",
      "Epoch 764/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0016 - accuracy: 0.9997 - val_loss: 0.0586 - val_accuracy: 0.9907\n",
      "Epoch 765/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0587 - val_accuracy: 0.9906\n",
      "Epoch 766/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0012 - accuracy: 0.9998 - val_loss: 0.0592 - val_accuracy: 0.9907\n",
      "Epoch 767/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0596 - val_accuracy: 0.9907\n",
      "Epoch 768/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0599 - val_accuracy: 0.9907\n",
      "Epoch 769/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0010 - accuracy: 0.9999 - val_loss: 0.0600 - val_accuracy: 0.9907\n",
      "Epoch 770/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0010 - accuracy: 0.9999 - val_loss: 0.0600 - val_accuracy: 0.9907\n",
      "Epoch 771/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 9.9830e-04 - accuracy: 0.9999 - val_loss: 0.0601 - val_accuracy: 0.9907\n",
      "Epoch 772/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 9.7151e-04 - accuracy: 0.9999 - val_loss: 0.0604 - val_accuracy: 0.9907\n",
      "Epoch 773/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 9.5978e-04 - accuracy: 0.9999 - val_loss: 0.0608 - val_accuracy: 0.9907\n",
      "Epoch 774/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 9.4701e-04 - accuracy: 0.9999 - val_loss: 0.0607 - val_accuracy: 0.9906\n",
      "Epoch 775/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 9.3569e-04 - accuracy: 0.9999 - val_loss: 0.0613 - val_accuracy: 0.9906\n",
      "Epoch 776/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 9.2402e-04 - accuracy: 0.9999 - val_loss: 0.0614 - val_accuracy: 0.9906\n",
      "Epoch 777/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 9.1856e-04 - accuracy: 0.9999 - val_loss: 0.0612 - val_accuracy: 0.9906\n",
      "Epoch 778/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 9.1070e-04 - accuracy: 0.9999 - val_loss: 0.0618 - val_accuracy: 0.9906\n",
      "Epoch 779/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 8.9957e-04 - accuracy: 0.9999 - val_loss: 0.0620 - val_accuracy: 0.9906\n",
      "Epoch 780/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 8.8521e-04 - accuracy: 0.9999 - val_loss: 0.0617 - val_accuracy: 0.9906\n",
      "Epoch 781/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 8.7820e-04 - accuracy: 0.9999 - val_loss: 0.0617 - val_accuracy: 0.9905\n",
      "Epoch 782/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 8.7206e-04 - accuracy: 0.9999 - val_loss: 0.0619 - val_accuracy: 0.9906\n",
      "Epoch 783/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 8.5818e-04 - accuracy: 0.9999 - val_loss: 0.0622 - val_accuracy: 0.9906\n",
      "Epoch 784/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 8.4875e-04 - accuracy: 0.9999 - val_loss: 0.0621 - val_accuracy: 0.9906\n",
      "Epoch 785/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 8.4476e-04 - accuracy: 0.9999 - val_loss: 0.0627 - val_accuracy: 0.9905\n",
      "Epoch 786/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 8.8712e-04 - accuracy: 0.9999 - val_loss: 0.0624 - val_accuracy: 0.9905\n",
      "Epoch 787/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 8.4498e-04 - accuracy: 0.9999 - val_loss: 0.0632 - val_accuracy: 0.9905\n",
      "Epoch 788/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 8.4642e-04 - accuracy: 0.9999 - val_loss: 0.0628 - val_accuracy: 0.9905\n",
      "Epoch 789/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 8.3154e-04 - accuracy: 0.9999 - val_loss: 0.0628 - val_accuracy: 0.9906\n",
      "Epoch 790/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 8.1817e-04 - accuracy: 0.9999 - val_loss: 0.0629 - val_accuracy: 0.9906\n",
      "Epoch 791/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 8.0449e-04 - accuracy: 0.9999 - val_loss: 0.0630 - val_accuracy: 0.9906\n",
      "Epoch 792/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.9658e-04 - accuracy: 0.9999 - val_loss: 0.0631 - val_accuracy: 0.9905\n",
      "Epoch 793/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.8420e-04 - accuracy: 0.9999 - val_loss: 0.0629 - val_accuracy: 0.9906\n",
      "Epoch 794/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 7.8316e-04 - accuracy: 0.9999 - val_loss: 0.0634 - val_accuracy: 0.9906\n",
      "Epoch 795/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 8.5643e-04 - accuracy: 0.9999 - val_loss: 0.0642 - val_accuracy: 0.9906\n",
      "Epoch 796/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 9.4859e-04 - accuracy: 0.9999 - val_loss: 0.0633 - val_accuracy: 0.9906\n",
      "Epoch 797/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 9.8231e-04 - accuracy: 0.9999 - val_loss: 0.0633 - val_accuracy: 0.9905\n",
      "Epoch 798/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0014 - accuracy: 0.9997 - val_loss: 0.0634 - val_accuracy: 0.9903\n",
      "Epoch 799/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0083 - accuracy: 0.9977 - val_loss: 0.0718 - val_accuracy: 0.9881\n",
      "Epoch 800/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0204 - accuracy: 0.9943 - val_loss: 0.0583 - val_accuracy: 0.9896\n",
      "Epoch 801/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0158 - accuracy: 0.9951 - val_loss: 0.0553 - val_accuracy: 0.9902\n",
      "Epoch 802/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0068 - accuracy: 0.9977 - val_loss: 0.0548 - val_accuracy: 0.9904\n",
      "Epoch 803/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0033 - accuracy: 0.9990 - val_loss: 0.0549 - val_accuracy: 0.9907\n",
      "Epoch 804/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 0.0023 - accuracy: 0.9995 - val_loss: 0.0555 - val_accuracy: 0.9908\n",
      "Epoch 805/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0017 - accuracy: 0.9997 - val_loss: 0.0570 - val_accuracy: 0.9907\n",
      "Epoch 806/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0013 - accuracy: 0.9998 - val_loss: 0.0571 - val_accuracy: 0.9907\n",
      "Epoch 807/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0012 - accuracy: 0.9999 - val_loss: 0.0576 - val_accuracy: 0.9906\n",
      "Epoch 808/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0578 - val_accuracy: 0.9907\n",
      "Epoch 809/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0580 - val_accuracy: 0.9906\n",
      "Epoch 810/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0010 - accuracy: 0.9999 - val_loss: 0.0586 - val_accuracy: 0.9907\n",
      "Epoch 811/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 9.7414e-04 - accuracy: 0.9999 - val_loss: 0.0588 - val_accuracy: 0.9906\n",
      "Epoch 812/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 9.4802e-04 - accuracy: 0.9999 - val_loss: 0.0591 - val_accuracy: 0.9905\n",
      "Epoch 813/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 9.2455e-04 - accuracy: 0.9999 - val_loss: 0.0596 - val_accuracy: 0.9906\n",
      "Epoch 814/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 8.9707e-04 - accuracy: 0.9999 - val_loss: 0.0597 - val_accuracy: 0.9906\n",
      "Epoch 815/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 9.0085e-04 - accuracy: 0.9999 - val_loss: 0.0595 - val_accuracy: 0.9907\n",
      "Epoch 816/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 8.6066e-04 - accuracy: 0.9999 - val_loss: 0.0600 - val_accuracy: 0.9906\n",
      "Epoch 817/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 8.4425e-04 - accuracy: 0.9999 - val_loss: 0.0602 - val_accuracy: 0.9906\n",
      "Epoch 818/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 8.2703e-04 - accuracy: 0.9999 - val_loss: 0.0602 - val_accuracy: 0.9906\n",
      "Epoch 819/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 8.1168e-04 - accuracy: 0.9999 - val_loss: 0.0604 - val_accuracy: 0.9906\n",
      "Epoch 820/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 8.0235e-04 - accuracy: 0.9999 - val_loss: 0.0604 - val_accuracy: 0.9907\n",
      "Epoch 821/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.9167e-04 - accuracy: 0.9999 - val_loss: 0.0609 - val_accuracy: 0.9906\n",
      "Epoch 822/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 7.8239e-04 - accuracy: 0.9999 - val_loss: 0.0608 - val_accuracy: 0.9907\n",
      "Epoch 823/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 7.7243e-04 - accuracy: 0.9999 - val_loss: 0.0611 - val_accuracy: 0.9906\n",
      "Epoch 824/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 7.6726e-04 - accuracy: 0.9999 - val_loss: 0.0611 - val_accuracy: 0.9906\n",
      "Epoch 825/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 7.5590e-04 - accuracy: 0.9999 - val_loss: 0.0616 - val_accuracy: 0.9906\n",
      "Epoch 826/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 7.4555e-04 - accuracy: 0.9999 - val_loss: 0.0617 - val_accuracy: 0.9906\n",
      "Epoch 827/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 7.3873e-04 - accuracy: 0.9999 - val_loss: 0.0618 - val_accuracy: 0.9907\n",
      "Epoch 828/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 7.3094e-04 - accuracy: 0.9999 - val_loss: 0.0619 - val_accuracy: 0.9906\n",
      "Epoch 829/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 7.2339e-04 - accuracy: 0.9999 - val_loss: 0.0621 - val_accuracy: 0.9906\n",
      "Epoch 830/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.1574e-04 - accuracy: 1.0000 - val_loss: 0.0620 - val_accuracy: 0.9906\n",
      "Epoch 831/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 7.1032e-04 - accuracy: 1.0000 - val_loss: 0.0624 - val_accuracy: 0.9906\n",
      "Epoch 832/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 7.3813e-04 - accuracy: 1.0000 - val_loss: 0.0624 - val_accuracy: 0.9906\n",
      "Epoch 833/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 7.1525e-04 - accuracy: 1.0000 - val_loss: 0.0624 - val_accuracy: 0.9906\n",
      "Epoch 834/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 6.9506e-04 - accuracy: 1.0000 - val_loss: 0.0627 - val_accuracy: 0.9905\n",
      "Epoch 835/1000\n",
      "35/35 [==============================] - 7s 205ms/step - loss: 6.8740e-04 - accuracy: 1.0000 - val_loss: 0.0627 - val_accuracy: 0.9906\n",
      "Epoch 836/1000\n",
      "35/35 [==============================] - 7s 204ms/step - loss: 6.8496e-04 - accuracy: 1.0000 - val_loss: 0.0627 - val_accuracy: 0.9906\n",
      "Epoch 837/1000\n",
      "35/35 [==============================] - 7s 199ms/step - loss: 6.7595e-04 - accuracy: 1.0000 - val_loss: 0.0631 - val_accuracy: 0.9905\n",
      "Epoch 838/1000\n",
      "35/35 [==============================] - 7s 203ms/step - loss: 6.6956e-04 - accuracy: 1.0000 - val_loss: 0.0631 - val_accuracy: 0.9906\n",
      "Epoch 839/1000\n",
      "35/35 [==============================] - 7s 193ms/step - loss: 6.6312e-04 - accuracy: 1.0000 - val_loss: 0.0628 - val_accuracy: 0.9906\n",
      "Epoch 840/1000\n",
      "35/35 [==============================] - 8s 228ms/step - loss: 6.6433e-04 - accuracy: 1.0000 - val_loss: 0.0629 - val_accuracy: 0.9906\n",
      "Epoch 841/1000\n",
      "35/35 [==============================] - 10s 285ms/step - loss: 6.6273e-04 - accuracy: 1.0000 - val_loss: 0.0629 - val_accuracy: 0.9906\n",
      "Epoch 842/1000\n",
      "35/35 [==============================] - 8s 216ms/step - loss: 6.5752e-04 - accuracy: 1.0000 - val_loss: 0.0633 - val_accuracy: 0.9906\n",
      "Epoch 843/1000\n",
      "35/35 [==============================] - 7s 196ms/step - loss: 6.5072e-04 - accuracy: 1.0000 - val_loss: 0.0635 - val_accuracy: 0.9906\n",
      "Epoch 844/1000\n",
      "35/35 [==============================] - 7s 191ms/step - loss: 6.4523e-04 - accuracy: 1.0000 - val_loss: 0.0636 - val_accuracy: 0.9906\n",
      "Epoch 845/1000\n",
      "35/35 [==============================] - 7s 190ms/step - loss: 6.4105e-04 - accuracy: 1.0000 - val_loss: 0.0636 - val_accuracy: 0.9906\n",
      "Epoch 846/1000\n",
      "35/35 [==============================] - 6s 180ms/step - loss: 7.0525e-04 - accuracy: 0.9999 - val_loss: 0.0636 - val_accuracy: 0.9907\n",
      "Epoch 847/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 6.8457e-04 - accuracy: 1.0000 - val_loss: 0.0635 - val_accuracy: 0.9905\n",
      "Epoch 848/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 6.7555e-04 - accuracy: 1.0000 - val_loss: 0.0645 - val_accuracy: 0.9906\n",
      "Epoch 849/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 6.7085e-04 - accuracy: 0.9999 - val_loss: 0.0637 - val_accuracy: 0.9907\n",
      "Epoch 850/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 6.5517e-04 - accuracy: 1.0000 - val_loss: 0.0645 - val_accuracy: 0.9906\n",
      "Epoch 851/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 6.4279e-04 - accuracy: 1.0000 - val_loss: 0.0644 - val_accuracy: 0.9905\n",
      "Epoch 852/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 6.7801e-04 - accuracy: 0.9999 - val_loss: 0.0642 - val_accuracy: 0.9905\n",
      "Epoch 853/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 7.4352e-04 - accuracy: 0.9999 - val_loss: 0.0658 - val_accuracy: 0.9903\n",
      "Epoch 854/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 7.1592e-04 - accuracy: 0.9999 - val_loss: 0.0640 - val_accuracy: 0.9907\n",
      "Epoch 855/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 6.9075e-04 - accuracy: 0.9999 - val_loss: 0.0642 - val_accuracy: 0.9906\n",
      "Epoch 856/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.4088e-04 - accuracy: 0.9999 - val_loss: 0.0640 - val_accuracy: 0.9906\n",
      "Epoch 857/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 7.5860e-04 - accuracy: 0.9999 - val_loss: 0.0642 - val_accuracy: 0.9907\n",
      "Epoch 858/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 7.1894e-04 - accuracy: 0.9999 - val_loss: 0.0656 - val_accuracy: 0.9906\n",
      "Epoch 859/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0013 - accuracy: 0.9997 - val_loss: 0.0637 - val_accuracy: 0.9906\n",
      "Epoch 860/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0118 - accuracy: 0.9968 - val_loss: 0.0692 - val_accuracy: 0.9885\n",
      "Epoch 861/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0129 - accuracy: 0.9960 - val_loss: 0.0563 - val_accuracy: 0.9904\n",
      "Epoch 862/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0039 - accuracy: 0.9987 - val_loss: 0.0580 - val_accuracy: 0.9906\n",
      "Epoch 863/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 0.0021 - accuracy: 0.9994 - val_loss: 0.0583 - val_accuracy: 0.9908\n",
      "Epoch 864/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 0.0013 - accuracy: 0.9997 - val_loss: 0.0578 - val_accuracy: 0.9909\n",
      "Epoch 865/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 0.0011 - accuracy: 0.9999 - val_loss: 0.0580 - val_accuracy: 0.9911\n",
      "Epoch 866/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 9.2368e-04 - accuracy: 0.9999 - val_loss: 0.0594 - val_accuracy: 0.9908\n",
      "Epoch 867/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 8.1344e-04 - accuracy: 0.9999 - val_loss: 0.0596 - val_accuracy: 0.9908\n",
      "Epoch 868/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.5175e-04 - accuracy: 1.0000 - val_loss: 0.0601 - val_accuracy: 0.9909\n",
      "Epoch 869/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 7.2032e-04 - accuracy: 1.0000 - val_loss: 0.0603 - val_accuracy: 0.9908\n",
      "Epoch 870/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 7.0044e-04 - accuracy: 1.0000 - val_loss: 0.0606 - val_accuracy: 0.9909\n",
      "Epoch 871/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 6.7180e-04 - accuracy: 1.0000 - val_loss: 0.0611 - val_accuracy: 0.9908\n",
      "Epoch 872/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 6.5757e-04 - accuracy: 1.0000 - val_loss: 0.0610 - val_accuracy: 0.9908\n",
      "Epoch 873/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 6.5270e-04 - accuracy: 1.0000 - val_loss: 0.0614 - val_accuracy: 0.9907\n",
      "Epoch 874/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 6.3512e-04 - accuracy: 1.0000 - val_loss: 0.0616 - val_accuracy: 0.9908\n",
      "Epoch 875/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 6.2672e-04 - accuracy: 1.0000 - val_loss: 0.0622 - val_accuracy: 0.9907\n",
      "Epoch 876/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 6.1168e-04 - accuracy: 1.0000 - val_loss: 0.0621 - val_accuracy: 0.9908\n",
      "Epoch 877/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 5.9924e-04 - accuracy: 1.0000 - val_loss: 0.0621 - val_accuracy: 0.9909\n",
      "Epoch 878/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.9212e-04 - accuracy: 1.0000 - val_loss: 0.0626 - val_accuracy: 0.9908\n",
      "Epoch 879/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.8465e-04 - accuracy: 1.0000 - val_loss: 0.0626 - val_accuracy: 0.9908\n",
      "Epoch 880/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 5.7547e-04 - accuracy: 1.0000 - val_loss: 0.0626 - val_accuracy: 0.9908\n",
      "Epoch 881/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.7029e-04 - accuracy: 1.0000 - val_loss: 0.0630 - val_accuracy: 0.9908\n",
      "Epoch 882/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 5.6511e-04 - accuracy: 1.0000 - val_loss: 0.0629 - val_accuracy: 0.9908\n",
      "Epoch 883/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.5912e-04 - accuracy: 1.0000 - val_loss: 0.0629 - val_accuracy: 0.9909\n",
      "Epoch 884/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 5.5445e-04 - accuracy: 1.0000 - val_loss: 0.0633 - val_accuracy: 0.9907\n",
      "Epoch 885/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.4737e-04 - accuracy: 1.0000 - val_loss: 0.0634 - val_accuracy: 0.9908\n",
      "Epoch 886/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.4057e-04 - accuracy: 1.0000 - val_loss: 0.0635 - val_accuracy: 0.9907\n",
      "Epoch 887/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 5.3806e-04 - accuracy: 1.0000 - val_loss: 0.0636 - val_accuracy: 0.9908\n",
      "Epoch 888/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.3203e-04 - accuracy: 1.0000 - val_loss: 0.0640 - val_accuracy: 0.9907\n",
      "Epoch 889/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.3158e-04 - accuracy: 1.0000 - val_loss: 0.0641 - val_accuracy: 0.9907\n",
      "Epoch 890/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.2698e-04 - accuracy: 1.0000 - val_loss: 0.0641 - val_accuracy: 0.9908\n",
      "Epoch 891/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.1946e-04 - accuracy: 1.0000 - val_loss: 0.0643 - val_accuracy: 0.9907\n",
      "Epoch 892/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.1606e-04 - accuracy: 1.0000 - val_loss: 0.0645 - val_accuracy: 0.9907\n",
      "Epoch 893/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.1732e-04 - accuracy: 1.0000 - val_loss: 0.0645 - val_accuracy: 0.9906\n",
      "Epoch 894/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 5.0930e-04 - accuracy: 1.0000 - val_loss: 0.0646 - val_accuracy: 0.9907\n",
      "Epoch 895/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.0275e-04 - accuracy: 1.0000 - val_loss: 0.0646 - val_accuracy: 0.9908\n",
      "Epoch 896/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.0072e-04 - accuracy: 1.0000 - val_loss: 0.0650 - val_accuracy: 0.9906\n",
      "Epoch 897/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.9914e-04 - accuracy: 1.0000 - val_loss: 0.0651 - val_accuracy: 0.9906\n",
      "Epoch 898/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.0279e-04 - accuracy: 1.0000 - val_loss: 0.0646 - val_accuracy: 0.9907\n",
      "Epoch 899/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.1531e-04 - accuracy: 1.0000 - val_loss: 0.0649 - val_accuracy: 0.9907\n",
      "Epoch 900/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.0559e-04 - accuracy: 1.0000 - val_loss: 0.0658 - val_accuracy: 0.9907\n",
      "Epoch 901/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 5.0194e-04 - accuracy: 1.0000 - val_loss: 0.0656 - val_accuracy: 0.9907\n",
      "Epoch 902/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 5.0364e-04 - accuracy: 1.0000 - val_loss: 0.0657 - val_accuracy: 0.9906\n",
      "Epoch 903/1000\n",
      "35/35 [==============================] - 6s 169ms/step - loss: 5.1565e-04 - accuracy: 1.0000 - val_loss: 0.0655 - val_accuracy: 0.9907\n",
      "Epoch 904/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.1144e-04 - accuracy: 1.0000 - val_loss: 0.0659 - val_accuracy: 0.9906\n",
      "Epoch 905/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 5.0310e-04 - accuracy: 1.0000 - val_loss: 0.0659 - val_accuracy: 0.9906\n",
      "Epoch 906/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.8928e-04 - accuracy: 0.9999 - val_loss: 0.0662 - val_accuracy: 0.9908\n",
      "Epoch 907/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 6.0455e-04 - accuracy: 0.9999 - val_loss: 0.0662 - val_accuracy: 0.9908\n",
      "Epoch 908/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0015 - accuracy: 0.9996 - val_loss: 0.0668 - val_accuracy: 0.9906\n",
      "Epoch 909/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0063 - accuracy: 0.9980 - val_loss: 0.0672 - val_accuracy: 0.9897\n",
      "Epoch 910/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0222 - accuracy: 0.9937 - val_loss: 0.0614 - val_accuracy: 0.9894\n",
      "Epoch 911/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0122 - accuracy: 0.9961 - val_loss: 0.0583 - val_accuracy: 0.9904\n",
      "Epoch 912/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 0.0045 - accuracy: 0.9985 - val_loss: 0.0574 - val_accuracy: 0.9909\n",
      "Epoch 913/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0023 - accuracy: 0.9993 - val_loss: 0.0575 - val_accuracy: 0.9909\n",
      "Epoch 914/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0014 - accuracy: 0.9997 - val_loss: 0.0581 - val_accuracy: 0.9908\n",
      "Epoch 915/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0011 - accuracy: 0.9998 - val_loss: 0.0594 - val_accuracy: 0.9908\n",
      "Epoch 916/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 9.2564e-04 - accuracy: 0.9999 - val_loss: 0.0592 - val_accuracy: 0.9910\n",
      "Epoch 917/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 8.1087e-04 - accuracy: 0.9999 - val_loss: 0.0598 - val_accuracy: 0.9910\n",
      "Epoch 918/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 7.7881e-04 - accuracy: 1.0000 - val_loss: 0.0606 - val_accuracy: 0.9908\n",
      "Epoch 919/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 7.2119e-04 - accuracy: 1.0000 - val_loss: 0.0605 - val_accuracy: 0.9909\n",
      "Epoch 920/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 6.9056e-04 - accuracy: 1.0000 - val_loss: 0.0606 - val_accuracy: 0.9909\n",
      "Epoch 921/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 6.6439e-04 - accuracy: 1.0000 - val_loss: 0.0614 - val_accuracy: 0.9909\n",
      "Epoch 922/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 6.4403e-04 - accuracy: 1.0000 - val_loss: 0.0612 - val_accuracy: 0.9909\n",
      "Epoch 923/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 6.3754e-04 - accuracy: 1.0000 - val_loss: 0.0615 - val_accuracy: 0.9908\n",
      "Epoch 924/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 6.1068e-04 - accuracy: 1.0000 - val_loss: 0.0621 - val_accuracy: 0.9908\n",
      "Epoch 925/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 5.9381e-04 - accuracy: 1.0000 - val_loss: 0.0620 - val_accuracy: 0.9908\n",
      "Epoch 926/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.7789e-04 - accuracy: 1.0000 - val_loss: 0.0625 - val_accuracy: 0.9907\n",
      "Epoch 927/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 5.6421e-04 - accuracy: 1.0000 - val_loss: 0.0626 - val_accuracy: 0.9908\n",
      "Epoch 928/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.5731e-04 - accuracy: 1.0000 - val_loss: 0.0627 - val_accuracy: 0.9907\n",
      "Epoch 929/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.4936e-04 - accuracy: 1.0000 - val_loss: 0.0628 - val_accuracy: 0.9908\n",
      "Epoch 930/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 5.3553e-04 - accuracy: 1.0000 - val_loss: 0.0631 - val_accuracy: 0.9908\n",
      "Epoch 931/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 5.3123e-04 - accuracy: 1.0000 - val_loss: 0.0634 - val_accuracy: 0.9907\n",
      "Epoch 932/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 5.2337e-04 - accuracy: 1.0000 - val_loss: 0.0637 - val_accuracy: 0.9908\n",
      "Epoch 933/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 5.2116e-04 - accuracy: 1.0000 - val_loss: 0.0639 - val_accuracy: 0.9907\n",
      "Epoch 934/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.1257e-04 - accuracy: 1.0000 - val_loss: 0.0640 - val_accuracy: 0.9907\n",
      "Epoch 935/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.0728e-04 - accuracy: 1.0000 - val_loss: 0.0640 - val_accuracy: 0.9907\n",
      "Epoch 936/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.9856e-04 - accuracy: 1.0000 - val_loss: 0.0644 - val_accuracy: 0.9907\n",
      "Epoch 937/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.3596e-04 - accuracy: 1.0000 - val_loss: 0.0639 - val_accuracy: 0.9907\n",
      "Epoch 938/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 5.6040e-04 - accuracy: 1.0000 - val_loss: 0.0648 - val_accuracy: 0.9907\n",
      "Epoch 939/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 5.0885e-04 - accuracy: 1.0000 - val_loss: 0.0643 - val_accuracy: 0.9907\n",
      "Epoch 940/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.9154e-04 - accuracy: 1.0000 - val_loss: 0.0644 - val_accuracy: 0.9907\n",
      "Epoch 941/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 4.7264e-04 - accuracy: 1.0000 - val_loss: 0.0647 - val_accuracy: 0.9907\n",
      "Epoch 942/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.6708e-04 - accuracy: 1.0000 - val_loss: 0.0649 - val_accuracy: 0.9907\n",
      "Epoch 943/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 4.6150e-04 - accuracy: 1.0000 - val_loss: 0.0650 - val_accuracy: 0.9907\n",
      "Epoch 944/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.5813e-04 - accuracy: 1.0000 - val_loss: 0.0652 - val_accuracy: 0.9907\n",
      "Epoch 945/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.6049e-04 - accuracy: 1.0000 - val_loss: 0.0651 - val_accuracy: 0.9907\n",
      "Epoch 946/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 4.6589e-04 - accuracy: 1.0000 - val_loss: 0.0654 - val_accuracy: 0.9907\n",
      "Epoch 947/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.6453e-04 - accuracy: 1.0000 - val_loss: 0.0652 - val_accuracy: 0.9907\n",
      "Epoch 948/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.6941e-04 - accuracy: 1.0000 - val_loss: 0.0656 - val_accuracy: 0.9908\n",
      "Epoch 949/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.7212e-04 - accuracy: 1.0000 - val_loss: 0.0660 - val_accuracy: 0.9907\n",
      "Epoch 950/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 4.7654e-04 - accuracy: 1.0000 - val_loss: 0.0656 - val_accuracy: 0.9907\n",
      "Epoch 951/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 4.4996e-04 - accuracy: 1.0000 - val_loss: 0.0658 - val_accuracy: 0.9908\n",
      "Epoch 952/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 4.3545e-04 - accuracy: 1.0000 - val_loss: 0.0660 - val_accuracy: 0.9907\n",
      "Epoch 953/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.3363e-04 - accuracy: 1.0000 - val_loss: 0.0661 - val_accuracy: 0.9907\n",
      "Epoch 954/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.2869e-04 - accuracy: 1.0000 - val_loss: 0.0662 - val_accuracy: 0.9907\n",
      "Epoch 955/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 4.2344e-04 - accuracy: 1.0000 - val_loss: 0.0663 - val_accuracy: 0.9907\n",
      "Epoch 956/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.2399e-04 - accuracy: 1.0000 - val_loss: 0.0666 - val_accuracy: 0.9907\n",
      "Epoch 957/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.2358e-04 - accuracy: 1.0000 - val_loss: 0.0666 - val_accuracy: 0.9907\n",
      "Epoch 958/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 4.1861e-04 - accuracy: 1.0000 - val_loss: 0.0667 - val_accuracy: 0.9907\n",
      "Epoch 959/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.1675e-04 - accuracy: 1.0000 - val_loss: 0.0670 - val_accuracy: 0.9907\n",
      "Epoch 960/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 4.1203e-04 - accuracy: 1.0000 - val_loss: 0.0664 - val_accuracy: 0.9907\n",
      "Epoch 961/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 4.1403e-04 - accuracy: 1.0000 - val_loss: 0.0667 - val_accuracy: 0.9907\n",
      "Epoch 962/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 4.1196e-04 - accuracy: 1.0000 - val_loss: 0.0669 - val_accuracy: 0.9907\n",
      "Epoch 963/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 4.0920e-04 - accuracy: 1.0000 - val_loss: 0.0671 - val_accuracy: 0.9907\n",
      "Epoch 964/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 4.0588e-04 - accuracy: 1.0000 - val_loss: 0.0665 - val_accuracy: 0.9908\n",
      "Epoch 965/1000\n",
      "35/35 [==============================] - 6s 178ms/step - loss: 4.1614e-04 - accuracy: 1.0000 - val_loss: 0.0674 - val_accuracy: 0.9907\n",
      "Epoch 966/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.3394e-04 - accuracy: 1.0000 - val_loss: 0.0669 - val_accuracy: 0.9907\n",
      "Epoch 967/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 4.6977e-04 - accuracy: 1.0000 - val_loss: 0.0670 - val_accuracy: 0.9908\n",
      "Epoch 968/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 8.3210e-04 - accuracy: 0.9998 - val_loss: 0.0663 - val_accuracy: 0.9906\n",
      "Epoch 969/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0044 - accuracy: 0.9986 - val_loss: 0.0683 - val_accuracy: 0.9905\n",
      "Epoch 970/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0094 - accuracy: 0.9971 - val_loss: 0.0716 - val_accuracy: 0.9895\n",
      "Epoch 971/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 0.0075 - accuracy: 0.9975 - val_loss: 0.0600 - val_accuracy: 0.9902\n",
      "Epoch 972/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 0.0045 - accuracy: 0.9985 - val_loss: 0.0598 - val_accuracy: 0.9909\n",
      "Epoch 973/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 0.0023 - accuracy: 0.9993 - val_loss: 0.0613 - val_accuracy: 0.9907\n",
      "Epoch 974/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 0.0019 - accuracy: 0.9995 - val_loss: 0.0620 - val_accuracy: 0.9908\n",
      "Epoch 975/1000\n",
      "35/35 [==============================] - 6s 179ms/step - loss: 0.0013 - accuracy: 0.9997 - val_loss: 0.0621 - val_accuracy: 0.9906\n",
      "Epoch 976/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 7.9677e-04 - accuracy: 0.9999 - val_loss: 0.0618 - val_accuracy: 0.9908\n",
      "Epoch 977/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 5.9157e-04 - accuracy: 1.0000 - val_loss: 0.0624 - val_accuracy: 0.9910\n",
      "Epoch 978/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 5.3486e-04 - accuracy: 1.0000 - val_loss: 0.0628 - val_accuracy: 0.9909\n",
      "Epoch 979/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.9841e-04 - accuracy: 1.0000 - val_loss: 0.0633 - val_accuracy: 0.9908\n",
      "Epoch 980/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 4.7631e-04 - accuracy: 1.0000 - val_loss: 0.0635 - val_accuracy: 0.9910\n",
      "Epoch 981/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.6692e-04 - accuracy: 1.0000 - val_loss: 0.0638 - val_accuracy: 0.9909\n",
      "Epoch 982/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.5193e-04 - accuracy: 1.0000 - val_loss: 0.0640 - val_accuracy: 0.9909\n",
      "Epoch 983/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.3919e-04 - accuracy: 1.0000 - val_loss: 0.0644 - val_accuracy: 0.9909\n",
      "Epoch 984/1000\n",
      "35/35 [==============================] - 6s 170ms/step - loss: 4.3118e-04 - accuracy: 1.0000 - val_loss: 0.0645 - val_accuracy: 0.9909\n",
      "Epoch 985/1000\n",
      "35/35 [==============================] - 6s 171ms/step - loss: 4.2320e-04 - accuracy: 1.0000 - val_loss: 0.0649 - val_accuracy: 0.9908\n",
      "Epoch 986/1000\n",
      "35/35 [==============================] - 6s 177ms/step - loss: 4.1862e-04 - accuracy: 1.0000 - val_loss: 0.0651 - val_accuracy: 0.9909\n",
      "Epoch 987/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 4.1172e-04 - accuracy: 1.0000 - val_loss: 0.0651 - val_accuracy: 0.9909\n",
      "Epoch 988/1000\n",
      "35/35 [==============================] - 6s 172ms/step - loss: 4.0539e-04 - accuracy: 1.0000 - val_loss: 0.0654 - val_accuracy: 0.9908\n",
      "Epoch 989/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 3.9935e-04 - accuracy: 1.0000 - val_loss: 0.0654 - val_accuracy: 0.9908\n",
      "Epoch 990/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 3.9745e-04 - accuracy: 1.0000 - val_loss: 0.0657 - val_accuracy: 0.9909\n",
      "Epoch 991/1000\n",
      "35/35 [==============================] - 6s 173ms/step - loss: 3.8942e-04 - accuracy: 1.0000 - val_loss: 0.0659 - val_accuracy: 0.9908\n",
      "Epoch 992/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 3.8541e-04 - accuracy: 1.0000 - val_loss: 0.0660 - val_accuracy: 0.9908\n",
      "Epoch 993/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 3.8303e-04 - accuracy: 1.0000 - val_loss: 0.0662 - val_accuracy: 0.9908\n",
      "Epoch 994/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 3.8239e-04 - accuracy: 1.0000 - val_loss: 0.0661 - val_accuracy: 0.9908\n",
      "Epoch 995/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 3.7628e-04 - accuracy: 1.0000 - val_loss: 0.0667 - val_accuracy: 0.9908\n",
      "Epoch 996/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 3.7296e-04 - accuracy: 1.0000 - val_loss: 0.0665 - val_accuracy: 0.9908\n",
      "Epoch 997/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 3.7682e-04 - accuracy: 1.0000 - val_loss: 0.0670 - val_accuracy: 0.9907\n",
      "Epoch 998/1000\n",
      "35/35 [==============================] - 6s 174ms/step - loss: 3.6899e-04 - accuracy: 1.0000 - val_loss: 0.0667 - val_accuracy: 0.9908\n",
      "Epoch 999/1000\n",
      "35/35 [==============================] - 6s 175ms/step - loss: 3.6450e-04 - accuracy: 1.0000 - val_loss: 0.0668 - val_accuracy: 0.9908\n",
      "Epoch 1000/1000\n",
      "35/35 [==============================] - 6s 176ms/step - loss: 3.6743e-04 - accuracy: 1.0000 - val_loss: 0.0671 - val_accuracy: 0.9908\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x220cef7ab48>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "model.fit(x_train, y_train, epochs=1000, batch_size=16, validation_data=(x_test, y_test), callbacks=[ checkpoint])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(\"bi_model.keras\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#MODELİ ÇALIŞTIR\n",
    "def model_run(directory):\n",
    "    all_texts = []\n",
    "\n",
    "    for filename in os.listdir(directory):\n",
    "        if filename.endswith('.txt'):\n",
    "            file_path = os.path.join(directory, filename)\n",
    "            with open(file_path, 'r', encoding='utf-8') as file:\n",
    "                content = file.read()\n",
    "                txt_data = content.split()\n",
    "                all_texts.append(txt_data)\n",
    "                \n",
    "    #TOKENIZER YÜKLE\n",
    "    with open('tokenizer.pickle', 'rb') as handle:\n",
    "        tokenizer_model = pickle.load(handle)     \n",
    "    print(tokenizer_model.word_index)        \n",
    "\n",
    "    all_sequences=[]\n",
    "    for i in range(len(all_texts)):\n",
    "        all_sequences.append(tokenizer_model.texts_to_sequences(all_texts[i]))\n",
    "        \n",
    "    max_word_length,trunc_type,padding_type = 20,'post','post'\n",
    "\n",
    "    all_padded_sequences=[]\n",
    "    for i in range(len(all_sequences)):\n",
    "        padded_sequence = pad_sequences(all_sequences[i], maxlen=max_word_length, padding=padding_type, truncating=trunc_type)\n",
    "        all_padded_sequences.append(padded_sequence)\n",
    "\n",
    "    padded_arrays = []\n",
    "    for arr in all_padded_sequences:\n",
    "        pad_width = ((0, max_sequence_length - len(arr)), (0, 0))\n",
    "        padded_arr = np.pad(arr, pad_width, mode='constant', constant_values=0)\n",
    "        padded_arrays.append(padded_arr)\n",
    "\n",
    "    xtrain=np.array(padded_arrays)\n",
    "    model = tensorflow.keras.models.load_model(model_path)\n",
    "\n",
    "    prediction=model.predict(xtrain)\n",
    "    print(prediction.shape)\n",
    "    #------------------------------------------------------------\n",
    "    label_list = [\"Pad\", \"Others\", \"B_Comp\", \"I_Comp\", \"B_Date\", \"I_Date\", \"B_Time\", \"I_Time\", \"B_Receipt\", \"I_Receipt\", \"B_Tax\", \"I_Tax\", \"B_Amount\", \"I_Amount\"]\n",
    "\n",
    "    #deneme yapılacak fişin konumu\n",
    "    with open(\"./deneme-veriler/deneme-veri.txt\", \"r\", encoding=\"utf-8\") as file:\n",
    "        words = file.read().split()\n",
    "    \n",
    "    predicted_labels = np.argmax(prediction, axis=-1)\n",
    "    print(predicted_labels)\n",
    "    \n",
    "    #klasördeki ilk fişi temsil eder\n",
    "    fis_1=predicted_labels[0]\n",
    "\n",
    "    truncated_values = fis_1[:len(words)]#uzun olan diziyi kısalt ve iki dizi boyutunu eşle\n",
    "    my_map = dict(zip(words, truncated_values))#kelimelere ve karşılarına predictionları 0 1 2 formatında maple\n",
    "\n",
    "    result_map = {}\n",
    "    for key, value in my_map.items():#anlamlı gözükmesi için 0 1 2 yerine etiket isimleri olacak şekilde maple\n",
    "        index = min(value, len(label_list) - 1)\n",
    "        label = label_list[index]\n",
    "        result_map[key] = label\n",
    "\n",
    "    print(result_map)\n",
    "    for key, value in result_map.items():\n",
    "        print(f'{key}: {value}')\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "id": "un1sqF0VGhlQ"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'0': 1, 'A': 2, '1': 3, 'E': 4, 'T': 5, 'I': 6, '2': 7, 'R': 8, 'N': 9, 'L': 10, 'K': 11, '*': 12, 'S': 13, 'O': 14, 'M': 15, '5': 16, ':': 17, '.': 18, '3': 19, '8': 20, '4': 21, '9': 22, 'D': 23, '6': 24, 'U': 25, 'İ': 26, '7': 27, ',': 28, 'B': 29, 'C': 30, 'Y': 31, 'P': 32, 'V': 33, 'e': 34, 'a': 35, 'i': 36, 'H': 37, 'G': 38, 'Z': 39, 'r': 40, 'F': 41, '/': 42, 'Ş': 43, 'n': 44, 'o': 45, 'l': 46, '%': 47, 't': 48, 's': 49, 'Ü': 50, '-': 51, 'k': 52, 'd': 53, 'm': 54, 'u': 55, 'X': 56, 'y': 57, 'c': 58, 'g': 59, 'h': 60, 'z': 61, '#': 62, 'ı': 63, 'b': 64, 'w': 65, 'Ç': 66, 'Ğ': 67, 'Ö': 68, 'p': 69, 'v': 70, 'ş': 71, 'W': 72, 'ü': 73, ')': 74, '(': 75, 'J': 76, 'f': 77, 'x': 78, \"'\": 79, '+': 80, '$': 81, 'ğ': 82, '»': 83, 'ç': 84, '!': 85, 'Q': 86, '=': 87, 'ö': 88, 'Í': 89, 'ж': 90, '&': 91, 'j': 92, '\"': 93, '>': 94, '\\\\': 95, 'Т': 96, 'О': 97, 'Р': 98, 'q': 99, '×': 100, '@': 101, '<': 102, ';': 103, 'Ø': 104, 'К': 105, '|': 106, 'Á': 107, 'Ú': 108, 'Ș': 109, '[': 110, '_': 111, '•': 112, '€': 113, 'É': 114, 'Č': 115, 'Ł': 116, 'ł': 117, 'Ž': 118, 'à': 119, 'Ì': 120, '※': 121, '}': 122, 'ä': 123, 'ó': 124, 'č': 125, 'Ó': 126, '?': 127, 'Ė': 128, '¥': 129, 'Ē': 130, 'ć': 131, 'Ī': 132, 'Ā': 133, ']': 134, 'Å': 135, 'о': 136, 'á': 137, 'ś': 138, 'ø': 139, '到': 140, 'У': 141, '米': 142, '…': 143, 'Š': 144, '^': 145, 'À': 146, '✰': 147, 'Ữ': 148, 'Î': 149, '⁹': 150, 'é': 151, '{': 152, 'Ệ': 153, 'и': 154, 'শ': 155, 'Ř': 156, '三': 157, '风': 158, '紫': 159, '头': 160, 'ń': 161, 'Ś': 162, '·': 163, 'Ő': 164, 'Ị': 165, '`': 166, 'Ý': 167, '¨': 168, '®': 169, 'ə': 170, 'Đ': 171, '₁': 172, 'Ļ': 173, 'Ģ': 174, 'ž': 175, 'ě': 176, '▸': 177, '„': 178, 'Ż': 179, 'å': 180, 'л': 181, 'с': 182, 'ē': 183, 'õ': 184, 'Ể': 185, 'Ů': 186, '₂': 187, 'ș': 188, 'š': 189, 'û': 190, 'Ć': 191, 'Ð': 192, 'Ä': 193, 'ộ': 194, '¡': 195}\n",
      "(1, 384, 14)\n",
      "[[ 2  3  3  1  1  1  1  1  1  1  1  1  1  1  1  1  4  1  6  1  8  1  1  1\n",
      "   1  1  1 10 12  1  1  1  1  1  1 12  1  1  1  1  1  1  1  1  1  1  1  1\n",
      "   1  4  1  1  1  6  1  1  1  1  1  1  1  1  1  1  1 12 13  1  1  1  1  1\n",
      "   1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1\n",
      "   1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
      "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0]]\n",
      "{'VAYRO': 'B_Comp', 'ÖMER': 'Others', 'ALBAYRAM': 'Others', 'KURUKÖPRÜ': 'Others', 'MH.': 'Others', 'ÇAKMAK': 'Others', 'CD.': 'Others', 'ÇUKUROVA': 'Others', 'İŞH.NO:105/1': 'Others', 'SEYHAN': 'Others', '/': 'Others', 'ADANA': 'Others', 'SUBURCU': 'Others', 'V.D.': 'Others', '39937304660': 'Others', ':': 'Others', '27/08/2021': 'B_Date', '16:56:55': 'B_Time', '0049': 'B_Receipt', 'TARIH': 'Others', 'SAAT': 'Others', 'FİŞ': 'Others', 'NO': 'Others', 'TEKSTİL': 'Others', '*32,99': 'B_Amount', '*2,44': 'B_Tax', 'TOPKDV': 'Others', 'TOPLAM': 'Others', 'KREDİ': 'Others', 'İŞ': 'Others', 'BANKASI': 'Others', '**1069': 'Others', 'T.SİCİL': 'Others', 'NO:': 'Others', '.': 'Others', 'WEB': 'Others', 'EKÜ': 'Others', 'NO:0001': 'Others', 'Z': 'Others', 'NO:0691': 'Others', 'İŞYERİ': 'Others', 'NO:667118459': 'Others', 'BÖLÜM:': 'Others', '0001/VAYRO': 'Others', 'GiV': 'Others', 'TERH': 'Others', 'SOMKG601': 'Others', '16:57': 'B_Time', 'SATIŞ': 'Others', 'AID:': 'Others', 'A0000000041010': 'Others', 'APP': 'Others', 'LABEL:': 'Others', 'KREDI': 'Others', 'KA': 'Others', '****': 'Others', '1069': 'Others', '32,': 'B_Amount', '99': 'I_Amount', 'TL': 'Others', '#': 'Others', 'YURTICI': 'Others', 'MC': 'Others', 'KARTI': 'Others', 'ONAY': 'Others', 'KODU:': 'Others', '338472': 'Others', 'VERSİYON:': 'Others', 'IIWE280': 'Others', 'EINGE2M68IS214C77': 'Others', 'RC-00': 'Others', 'SIRA': 'Others', 'IA000041': 'Others', 'REF:': 'Others', '123916503972': 'Others', 'BATCH': 'Others', '000694': 'Others', 'UUKH:': 'Others', '00682900': 'Others', 'BU': 'Others', 'BELGEVİ': 'Others', 'SAKLAY': 'Others', 'INIZ': 'Others', 'KARI': 'Others', 'HAMİLİ': 'Others', 'NÜSHASI': 'Others', 'TiC.SICIL': 'Others', 'NO:-1': 'Others', 'TÜRKİYE': 'Others', '$': 'Others', 'MASTERCARD': 'Others', 'PAYPASS': 'Others', 'NF': 'Others', '2A': 'Others', '20144856': 'Others'}\n",
      "VAYRO: B_Comp\n",
      "ÖMER: Others\n",
      "ALBAYRAM: Others\n",
      "KURUKÖPRÜ: Others\n",
      "MH.: Others\n",
      "ÇAKMAK: Others\n",
      "CD.: Others\n",
      "ÇUKUROVA: Others\n",
      "İŞH.NO:105/1: Others\n",
      "SEYHAN: Others\n",
      "/: Others\n",
      "ADANA: Others\n",
      "SUBURCU: Others\n",
      "V.D.: Others\n",
      "39937304660: Others\n",
      ":: Others\n",
      "27/08/2021: B_Date\n",
      "16:56:55: B_Time\n",
      "0049: B_Receipt\n",
      "TARIH: Others\n",
      "SAAT: Others\n",
      "FİŞ: Others\n",
      "NO: Others\n",
      "TEKSTİL: Others\n",
      "*32,99: B_Amount\n",
      "*2,44: B_Tax\n",
      "TOPKDV: Others\n",
      "TOPLAM: Others\n",
      "KREDİ: Others\n",
      "İŞ: Others\n",
      "BANKASI: Others\n",
      "**1069: Others\n",
      "T.SİCİL: Others\n",
      "NO:: Others\n",
      ".: Others\n",
      "WEB: Others\n",
      "EKÜ: Others\n",
      "NO:0001: Others\n",
      "Z: Others\n",
      "NO:0691: Others\n",
      "İŞYERİ: Others\n",
      "NO:667118459: Others\n",
      "BÖLÜM:: Others\n",
      "0001/VAYRO: Others\n",
      "GiV: Others\n",
      "TERH: Others\n",
      "SOMKG601: Others\n",
      "16:57: B_Time\n",
      "SATIŞ: Others\n",
      "AID:: Others\n",
      "A0000000041010: Others\n",
      "APP: Others\n",
      "LABEL:: Others\n",
      "KREDI: Others\n",
      "KA: Others\n",
      "****: Others\n",
      "1069: Others\n",
      "32,: B_Amount\n",
      "99: I_Amount\n",
      "TL: Others\n",
      "#: Others\n",
      "YURTICI: Others\n",
      "MC: Others\n",
      "KARTI: Others\n",
      "ONAY: Others\n",
      "KODU:: Others\n",
      "338472: Others\n",
      "VERSİYON:: Others\n",
      "IIWE280: Others\n",
      "EINGE2M68IS214C77: Others\n",
      "RC-00: Others\n",
      "SIRA: Others\n",
      "IA000041: Others\n",
      "REF:: Others\n",
      "123916503972: Others\n",
      "BATCH: Others\n",
      "000694: Others\n",
      "UUKH:: Others\n",
      "00682900: Others\n",
      "BU: Others\n",
      "BELGEVİ: Others\n",
      "SAKLAY: Others\n",
      "INIZ: Others\n",
      "KARI: Others\n",
      "HAMİLİ: Others\n",
      "NÜSHASI: Others\n",
      "TiC.SICIL: Others\n",
      "NO:-1: Others\n",
      "TÜRKİYE: Others\n",
      "$: Others\n",
      "MASTERCARD: Others\n",
      "PAYPASS: Others\n",
      "NF: Others\n",
      "2A: Others\n",
      "20144856: Others\n"
     ]
    }
   ],
   "source": [
    "\n",
    "#model_test_path içinde test edilecek veriler txt halde bulunmalıdır\n",
    "model_run(model_test_path)"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "gpuType": "T4",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.7.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}