Ex08.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "21dec411",
   "metadata": {},
   "source": [
    "# Ex_08. 네이버 영화리뷰 감성분석 도전하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "da74737a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.4.4\n",
      "0.6.0\n",
      "4.1.2\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import konlpy\n",
    "import gensim\n",
    "import os\n",
    "os.environ['KMP_DUPLICATE_LIB_OK']='True'\n",
    "print(pd.__version__)\n",
    "print(konlpy.__version__)\n",
    "print(gensim.__version__)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8a14850",
   "metadata": {},
   "source": [
    "## 1. 데이터 준비하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "7410da86",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>document</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9976970</td>\n",
       "      <td>아 더빙.. 진짜 짜증나네요 목소리</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3819312</td>\n",
       "      <td>흠...포스터보고 초딩영화줄....오버연기조차 가볍지 않구나</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10265843</td>\n",
       "      <td>너무재밓었다그래서보는것을추천한다</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9045019</td>\n",
       "      <td>교도소 이야기구먼 ..솔직히 재미는 없다..평점 조정</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6483659</td>\n",
       "      <td>사이몬페그의 익살스런 연기가 돋보였던 영화!스파이더맨에서 늙어보이기만 했던 커스틴 ...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149995</th>\n",
       "      <td>6222902</td>\n",
       "      <td>인간이 문제지.. 소는 뭔죄인가..</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149996</th>\n",
       "      <td>8549745</td>\n",
       "      <td>평점이 너무 낮아서...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149997</th>\n",
       "      <td>9311800</td>\n",
       "      <td>이게 뭐요? 한국인은 거들먹거리고 필리핀 혼혈은 착하다?</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149998</th>\n",
       "      <td>2376369</td>\n",
       "      <td>청춘 영화의 최고봉.방황과 우울했던 날들의 자화상</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149999</th>\n",
       "      <td>9619869</td>\n",
       "      <td>한국 영화 최초로 수간하는 내용이 담긴 영화</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>150000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              id                                           document  label\n",
       "0        9976970                                아 더빙.. 진짜 짜증나네요 목소리      0\n",
       "1        3819312                  흠...포스터보고 초딩영화줄....오버연기조차 가볍지 않구나      1\n",
       "2       10265843                                  너무재밓었다그래서보는것을추천한다      0\n",
       "3        9045019                      교도소 이야기구먼 ..솔직히 재미는 없다..평점 조정      0\n",
       "4        6483659  사이몬페그의 익살스런 연기가 돋보였던 영화!스파이더맨에서 늙어보이기만 했던 커스틴 ...      1\n",
       "...          ...                                                ...    ...\n",
       "149995   6222902                                인간이 문제지.. 소는 뭔죄인가..      0\n",
       "149996   8549745                                      평점이 너무 낮아서...      1\n",
       "149997   9311800                    이게 뭐요? 한국인은 거들먹거리고 필리핀 혼혈은 착하다?      0\n",
       "149998   2376369                        청춘 영화의 최고봉.방황과 우울했던 날들의 자화상      1\n",
       "149999   9619869                           한국 영화 최초로 수간하는 내용이 담긴 영화      0\n",
       "\n",
       "[150000 rows x 3 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data = pd.read_table('C:/users/juwon/practice_aiffel/sentiment_classification/data/ratings_train.txt')\n",
    "test_data = pd.read_table('C:/users/juwon/practice_aiffel/sentiment_classification/data/ratings_test.txt')\n",
    "train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4f48b7d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>document</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6270596</td>\n",
       "      <td>굳 ㅋ</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9274899</td>\n",
       "      <td>GDNTOPCLASSINTHECLUB</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8544678</td>\n",
       "      <td>뭐야 이 평점들은.... 나쁘진 않지만 10점 짜리는 더더욱 아니잖아</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6825595</td>\n",
       "      <td>지루하지는 않은데 완전 막장임... 돈주고 보기에는....</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6723715</td>\n",
       "      <td>3D만 아니었어도 별 다섯 개 줬을텐데.. 왜 3D로 나와서 제 심기를 불편하게 하죠??</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                                           document  label\n",
       "0  6270596                                                굳 ㅋ      1\n",
       "1  9274899                               GDNTOPCLASSINTHECLUB      0\n",
       "2  8544678             뭐야 이 평점들은.... 나쁘진 않지만 10점 짜리는 더더욱 아니잖아      0\n",
       "3  6825595                   지루하지는 않은데 완전 막장임... 돈주고 보기에는....      0\n",
       "4  6723715  3D만 아니었어도 별 다섯 개 줬을텐데.. 왜 3D로 나와서 제 심기를 불편하게 하죠??      0"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e42735d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 150000 entries, 0 to 149999\n",
      "Data columns (total 3 columns):\n",
      " #   Column    Non-Null Count   Dtype \n",
      "---  ------    --------------   ----- \n",
      " 0   id        150000 non-null  int64 \n",
      " 1   document  149995 non-null  object\n",
      " 2   label     150000 non-null  int64 \n",
      "dtypes: int64(2), object(1)\n",
      "memory usage: 3.4+ MB\n"
     ]
    }
   ],
   "source": [
    "train_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "411be475",
   "metadata": {},
   "outputs": [],
   "source": [
    "from eunjeon import Mecab\n",
    "import numpy as np\n",
    "from collections import Counter\n",
    "\n",
    "tokenizer = Mecab()\n",
    "stopwords = ['의','가','이','은','들','는','좀','잘','걍','과','도','를','으로','자','에','와','한','하다']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95b83f04",
   "metadata": {},
   "source": [
    "## 2. load_data\n",
    "* 데이터의 중복 제거\n",
    "* NaN 결측치 제거\n",
    "* 한국어 토크나이저로 토큰화\n",
    "* 불용어(Stopwords) 제거\n",
    "* 사전word_to_index 구성\n",
    "* 텍스트 스트링을 사전 인덱스 스트링으로 변환\n",
    "* X_train, y_train, X_test, y_test, word_to_index 리턴"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cec8b354",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data(train_data, test_data, num_words=10000 ) :\n",
    "    train_data.drop_duplicates(subset=['document'], inplace=True) # 중복데이터, 결측치 제거\n",
    "    train_data = train_data.dropna(how = 'any') \n",
    "    test_data.drop_duplicates(subset=['document'], inplace=True)\n",
    "    test_data = test_data.dropna(how = 'any') \n",
    "    \n",
    "    X_train = []\n",
    "    for sentence in train_data['document'] :\n",
    "        temp_X = tokenizer.morphs(sentence)\n",
    "        temp_X = [word for word in temp_X if not word in stopwords] # 불용어 제거\n",
    "        X_train.append(temp_X) # 토큰화해서 저장\n",
    "    \n",
    "    X_test = []\n",
    "    for sentence in test_data['document'] :\n",
    "        temp_X = tokenizer.morphs(sentence)\n",
    "        temp_X = [word for word in temp_X if not word in stopwords]\n",
    "        X_test.append(temp_X)\n",
    "    \n",
    "    words = np.concatenate(X_train).tolist()  \n",
    "    counter = Counter(words)\n",
    "    counter = counter.most_common(10000-4)\n",
    "    vocab = ['', '', '', ''] + [key for key, _ in counter]\n",
    "    word_to_index = {word:index for index, word in enumerate(vocab)} # 사전 word_to_index 구성\n",
    "\n",
    "    def wordlist_to_indexlist(wordlist): \n",
    "        return [word_to_index[word] if word in word_to_index else word_to_index[''] for word in wordlist] \n",
    "    \n",
    "    X_train = list(map(wordlist_to_indexlist, X_train))\n",
    "    X_test = list(map(wordlist_to_indexlist, X_test))\n",
    "    \n",
    "    return X_train, np.array(list(train_data['label'])), X_test, np.array(list(test_data['label'])), word_to_index\n",
    "X_train, y_train, X_test, y_test, word_to_index = load_data(train_data, test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7f7da1f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "word_to_index['<PAD>']=0\n",
    "word_to_index['<BOS>']=1\n",
    "word_to_index['<UNK>']=2\n",
    "index_to_word = {index:word for word, index in word_to_index.items()} # index_to_word 구성"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0cc669f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 문장 1개를 활용할 딕셔너리와 함께 주면, 단어 인덱스 리스트 벡터로 변환해 주는 함수입니다. \n",
    "# 단, 모든 문장은 <BOS>로 시작하는 것으로 합니다. \n",
    "def get_encoded_sentence(sentence, word_to_index):\n",
    "    return [word_to_index['<BOS>']]+ sentence  # 이미 load_data로 벡터화를 했으므로 시작 벡터만 붙여준다.\n",
    "\n",
    "# 여러 개의 문장 리스트를 한꺼번에 단어 인덱스 리스트 벡터로 encode해 주는 함수입니다. \n",
    "def get_encoded_sentences(sentences, word_to_index):\n",
    "    return [get_encoded_sentence(sentence, word_to_index) for sentence in sentences]\n",
    "\n",
    "# 숫자 벡터로 encode된 문장을 원래대로 decode하는 함수입니다. \n",
    "def get_decoded_sentence(encoded_sentence, index_to_word):\n",
    "    return ' '.join(index_to_word[index] if index in index_to_word else '<UNK>' for index in encoded_sentence[1:])  #[1:]를 통해 <BOS>를 제외\n",
    "\n",
    "# 여러 개의 숫자 벡터로 encode된 문장을 한꺼번에 원래대로 decode하는 함수입니다. \n",
    "def get_decoded_sentences(encoded_sentences, index_to_word):\n",
    "    return [get_decoded_sentence(encoded_sentence, index_to_word) for encoded_sentence in encoded_sentences]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "697d660e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[32, 600, 4, 4, 39, 229, 20, 33, 745]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "48bff302",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'더빙 . . 진짜 짜증 나 네요 목소리'"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_decoded_sentence(X_train[0], index_to_word)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "5cb4cdaa",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = get_encoded_sentences(X_train, word_to_index) # <BOS> 벡터를 문장앞에 붙여준다.\n",
    "X_test = get_encoded_sentences(X_test, word_to_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d4b8440a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 32, 600, 4, 4, 39, 229, 20, 33, 745]\n",
      "[1, 828, 131]\n"
     ]
    }
   ],
   "source": [
    "print(X_train[0])\n",
    "print(X_test[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5bb8d8b1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "아 더빙 . . 진짜 짜증 나 네요 목소리\n",
      "굳 ㅋ\n"
     ]
    }
   ],
   "source": [
    "print(get_decoded_sentence(X_train[0], index_to_word))\n",
    "print(get_decoded_sentence(X_test[0], index_to_word))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46f80b19",
   "metadata": {},
   "source": [
    "## 3) 모델 구성을 위한 데이터 분석 및 가공\n",
    "* 데이터셋 내 문장 길이 분포\n",
    "* 적절한 최대 문장 길이 지정\n",
    "* keras.preprocessing.sequence.pad_sequences 을 활용한 패딩 추가\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "081b4794",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "문장길이 평균 :  16.972842084785935\n",
      "문장길이 최대 :  117\n",
      "문장길이 표준편차 :  12.843357707480491\n",
      "pad_sequences maxlen :  55\n",
      "전체 문장의 0.9720946661956905%가 maxlen 설정값 이내에 포함됩니다. \n"
     ]
    }
   ],
   "source": [
    "total_list = X_train + X_test\n",
    "len(total_list)\n",
    "num_token = [len(sentence) for sentence in total_list]\n",
    "num_token = np.array(num_token)\n",
    "print('문장길이 평균 : ', np.mean(num_token)) # 데이터셋 내 문장 길이 분포 확인해보기\n",
    "print('문장길이 최대 : ', np.max(num_token))\n",
    "print('문장길이 표준편차 : ', np.std(num_token))\n",
    "maxtokens = np.mean(num_token)+3*np.std(num_token) # 최대 문장 길이 지정\n",
    "maxlen = int(maxtokens)\n",
    "print('pad_sequences maxlen : ', maxlen)\n",
    "print('전체 문장의 {}%가 maxlen 설정값 이내에 포함됩니다. '.format(np.sum(num_token < maxtokens) / len(num_token)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2b6c21d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "X_train= tf.keras.preprocessing.sequence.pad_sequences(X_train,\n",
    "                                              value = word_to_index['<PAD>'],\n",
    "                                             padding='post',\n",
    "                                             maxlen=maxlen) \n",
    "X_test= tf.keras.preprocessing.sequence.pad_sequences(X_test,\n",
    "                                              value = word_to_index['<PAD>'],\n",
    "                                             padding='post',\n",
    "                                             maxlen=maxlen)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a1b3fd7",
   "metadata": {},
   "source": [
    "### 데이터 분리"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "2cf0cff4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(76182, 55)\n",
      "(76182,)\n"
     ]
    }
   ],
   "source": [
    "# validation set 70000건 분리\n",
    "X_val = X_train[:70000]   \n",
    "y_val = y_train[:70000]\n",
    "\n",
    "# validation set을 제외한 나머지\n",
    "partial_x_train = X_train[70000:]  \n",
    "partial_y_train = y_train[70000:]\n",
    "\n",
    "print(partial_x_train.shape)\n",
    "print(partial_y_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "c9411f77",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "146182\n",
      "49157\n"
     ]
    }
   ],
   "source": [
    "print(len(X_train))\n",
    "print(len(X_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b45bc209",
   "metadata": {},
   "source": [
    "## 4) 모델구성\n",
    "### 4.1) LSTM 활용"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "01dca4b4",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'tf' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_6940\\347569248.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[1;31m# model 설계 - 딥러닝 모델 코드를 직접 작성해 주세요.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 5\u001b[1;33m \u001b[0mmodel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mkeras\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mSequential\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      6\u001b[0m \u001b[1;31m# [[YOUR CODE]]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      7\u001b[0m \u001b[0mmodel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mkeras\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mSequential\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'tf' is not defined"
     ]
    }
   ],
   "source": [
    "vocab_size = 10000    # 어휘 사전의 크기입니다(10,000개의 단어)\n",
    "word_vector_dim = 64  # 워드 벡터의 차원 수 (변경 가능한 하이퍼파라미터)\n",
    "\n",
    "# model 설계 - 딥러닝 모델 코드를 직접 작성해 주세요.\n",
    "model = tf.keras.Sequential()\n",
    "# [[YOUR CODE]]\n",
    "model = tf.keras.Sequential()\n",
    "model.add(tf.keras.layers.Embedding(vocab_size\n",
    "                                    , word_vector_dim,input_shape=(None,)))\n",
    "model.add(tf.keras.layers.LSTM(4))   # 가장 널리 쓰이는 RNN인 LSTM 레이어를 사용하였습니다. 이때 LSTM state 벡터의 차원수는 8로 하였습니다. (변경 가능)\n",
    "model.add(tf.keras.layers.Dense(4, activation='relu'))\n",
    "# model.add(tf.keras.layers.Dense(8, activation='relu'))\n",
    "# model.add(tf.keras.layers.Dense(128, activation='relu'))\n",
    "\n",
    "model.add(tf.keras.layers.Dense(1, activation='sigmoid'))  # 최종 출력은 긍정/부정을 나타내는 1dim 입니다.\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "edb90182",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'model' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_6940\\666354579.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m model.compile(optimizer='adam',\n\u001b[0m\u001b[0;32m      2\u001b[0m              \u001b[0mloss\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'binary_crossentropy'\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m              metrics=['accuracy'])\n\u001b[0;32m      4\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mepochs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m7\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'model' is not defined"
     ]
    }
   ],
   "source": [
    "model.compile(optimizer='adam',\n",
    "             loss='binary_crossentropy',\n",
    "             metrics=['accuracy'])\n",
    "\n",
    "epochs=7\n",
    "with tf.device('/device:GPU:0') :\n",
    "    history = model.fit(partial_x_train,\n",
    "         partial_y_train,\n",
    "         epochs = epochs,\n",
    "         batch_size = 512,\n",
    "         validation_data = (X_val, y_val),\n",
    "         verbose=1)\n",
    "    results = model.evaluate(X_test, y_test)\n",
    "    print(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "62a995e8",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'history' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_6940\\3210914784.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mhistory_dict\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mhistory\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mhistory\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mhistory_dict\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mkeys\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'history' is not defined"
     ]
    }
   ],
   "source": [
    "history_dict = history.history\n",
    "history_dict.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3a46a62a",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'history_dict' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_6940\\1041020209.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mmatplotlib\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpyplot\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mplt\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 2\u001b[1;33m \u001b[0mloss\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mhistory_dict\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'loss'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      3\u001b[0m \u001b[0mval_loss\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mhistory_dict\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'val_loss'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0macc\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mhistory_dict\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'accuracy'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mval_acc\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mhistory_dict\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'val_accuracy'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'history_dict' is not defined"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "loss = history_dict['loss']\n",
    "val_loss = history_dict['val_loss']\n",
    "acc = history_dict['accuracy']\n",
    "val_acc = history_dict['val_accuracy']\n",
    "\n",
    "epochs = range(1,len(acc)+1)\n",
    "\n",
    "plt.plot(epochs, loss, c='r', label='Training loss')\n",
    "plt.plot(epochs,val_loss, c='b', label = 'Val loss')\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "20016b46",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.plot(epochs, acc ,'r', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label = 'Validation acc')\n",
    "plt.title('Training and Validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('acc')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f0901a0",
   "metadata": {},
   "source": [
    "검증셋의 정확도가 0.83에 수렴한다. epoch=4 부터 과대적합이 일어났다."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebeb01d1",
   "metadata": {},
   "source": [
    "dim = 64 / 4 4  0.7931\n",
    "dim = 32 / 16 32 64 128  0.834     \n",
    "dim = 32 / 16 16  0.834    \n",
    "dim = 32 / 8 8 8  0.836    \n",
    "dim = 32 / 8 8  0.837    \n",
    "dim = 32 / 4 4  0.83    \n",
    "dim = 16 / 8 8  0.834    \n",
    "dim = 8 / 8 8  0.835    \n",
    "dim = 8 / 8 8 8  0.832    \n",
    "\n",
    "evaluate 해본결과 acc는 0.83에 수렴했고, 더 올라가질 않는다. 워드벡터의 차원수가 풍부한 표현을 할 순 있지만 너무 많다고 해서 성능이 좋아지지는 않는 것 같다. 오히려 떨어진다.  또한 레이어를 깊게 쌓는것은 이번 데이터셋에서는 의미가 없어보인다. 평균적으로 epoch=3~4 부터 오버피팅이 일어났다. 은닉층이 얕고 노드수가 적을수록 좀 더 높은 acc가 나왔다."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60613d27",
   "metadata": {},
   "source": [
    "## 4.2 1-D Convolution Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "d4dd9954",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_41\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_41 (Embedding)    (None, None, 32)          320000    \n",
      "                                                                 \n",
      " dropout_24 (Dropout)        (None, None, 32)          0         \n",
      "                                                                 \n",
      " conv1d_58 (Conv1D)          (None, None, 32)          3104      \n",
      "                                                                 \n",
      " max_pooling1d_29 (MaxPoolin  (None, None, 32)         0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " conv1d_59 (Conv1D)          (None, None, 64)          6208      \n",
      "                                                                 \n",
      " global_max_pooling1d_41 (Gl  (None, 64)               0         \n",
      " obalMaxPooling1D)                                               \n",
      "                                                                 \n",
      " dense_82 (Dense)            (None, 64)                4160      \n",
      "                                                                 \n",
      " dropout_25 (Dropout)        (None, 64)                0         \n",
      "                                                                 \n",
      " dense_83 (Dense)            (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 333,537\n",
      "Trainable params: 333,537\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "vocab_size = 10000  # 어휘 사전의 크기입니다(10개의 단어)\n",
    "word_vector_dim = 32   # 단어 하나를 표현하는 임베딩 벡터의 차원 수입니다. \n",
    "\n",
    "model = tf.keras.Sequential()\n",
    "model.add(tf.keras.layers.Embedding(vocab_size, word_vector_dim, input_shape=(None,)))\n",
    "model.add(tf.keras.layers.Dropout(0.3))\n",
    "model.add(tf.keras.layers.Conv1D(32, 3, activation='relu'))\n",
    "model.add(tf.keras.layers.MaxPooling1D(5))\n",
    "model.add(tf.keras.layers.Conv1D(64, 3, activation='relu'))\n",
    "model.add(tf.keras.layers.GlobalMaxPooling1D())\n",
    "model.add(tf.keras.layers.Dense(64, activation='relu'))\n",
    "model.add(tf.keras.layers.Dropout(0.2))\n",
    "model.add(tf.keras.layers.Dense(1, activation='sigmoid'))  # 최종 출력은 긍정/부정을 나타내는 1dim 입니다.\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "id": "decea01d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/3\n",
      "149/149 [==============================] - 7s 45ms/step - loss: 0.5190 - accuracy: 0.7212 - val_loss: 0.3598 - val_accuracy: 0.8435\n",
      "Epoch 2/3\n",
      "149/149 [==============================] - 6s 40ms/step - loss: 0.3400 - accuracy: 0.8545 - val_loss: 0.3446 - val_accuracy: 0.8493\n",
      "Epoch 3/3\n",
      "149/149 [==============================] - 6s 41ms/step - loss: 0.3007 - accuracy: 0.8747 - val_loss: 0.3505 - val_accuracy: 0.8474\n",
      "1537/1537 [==============================] - 3s 2ms/step - loss: 0.3546 - accuracy: 0.8450\n",
      "[0.35460802912712097, 0.8449661135673523]\n"
     ]
    }
   ],
   "source": [
    "with tf.device('/device:CPU:0'):\n",
    "    model.compile(optimizer='adam',\n",
    "             loss='binary_crossentropy',\n",
    "             metrics=['accuracy'])\n",
    "\n",
    "    epochs=3\n",
    "\n",
    "    history = model.fit(partial_x_train,\n",
    "                    partial_y_train,\n",
    "                    epochs = epochs,\n",
    "                    batch_size = 512,\n",
    "                    validation_data = (X_val, y_val),\n",
    "                    verbose=1)\n",
    "    results = model.evaluate(X_test, y_test)\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "id": "376287ca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history_dict = history.history\n",
    "history_dict.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "id": "62a68058",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "loss = history_dict['loss']\n",
    "val_loss = history_dict['val_loss']\n",
    "acc = history_dict['accuracy']\n",
    "val_acc = history_dict['val_accuracy']\n",
    "\n",
    "epochs = range(1,len(acc)+1)\n",
    "\n",
    "plt.plot(epochs, loss, c='r', label='Training loss')\n",
    "plt.plot(epochs,val_loss, c='b', label = 'Val loss')\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "id": "d91182f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(epochs, acc ,'r', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label = 'Validation acc')\n",
    "plt.title('Training and Validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('acc')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7f235374",
   "metadata": {},
   "source": [
    "dim = 16 (8,7) (8,7) 8 0.8418    \n",
    "dim = 16 (16,7) (16,7) 8 0.8359    \n",
    "dim = 16 (16,7) (32,7) 0.8347    \n",
    "dim =16 (32,7) (64,7) 0.8367    \n",
    "dim= 16 (4,5) (4,50 0.8402    \n",
    "dim = 16 (8,3) (8,3) 8 0.8406    \n",
    "dim = 16 (16,3) (16,3) 16 0.8404    \n",
    "dim = 16 (16,7) (32,7) 16 0.8368\n",
    "\n",
    "dim = 32 (16,4) (32,4) 16 0.845\n",
    "dim = 32 (32,4) (64,4) 16 0.842\n",
    "dim = 32 (32,4) (64,4) 16 0.842\n",
    "dim = 32 (32,4) (64,4) 32 (dropout = 0.2) 0.848\n",
    "dim = 32 (32,4) (64,4) 64 (dropout = 0.3, 0.2) 0.8521\n",
    "\n",
    "acc는 Dense를 사용했을 때 보다 전체적으로 높았다. filter가 많을수록 마지막 은닉층의 Dense의 노드가 많을수록 정확도가 올라갔다.\n",
    "kernel의 경우 7보단 3,4일때 보다 높게 나왔다. 과적합이 적은 epoch에서 일어나서 epoch을 줄였다. 과적합 해결을 위해 dropout레이어를 추가했다. 임베딩레이어 다음과 마지막 층으로 들어가기전에 한 번 더 걸러주었다. Dropout을 사용하였더니 acc가 증가하였다. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e89a19ff",
   "metadata": {},
   "source": [
    "### 4.3 GlobalMaxPooling1D 만 사용해보기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "a8bedc9a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_29\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_29 (Embedding)    (None, None, 8)           80000     \n",
      "                                                                 \n",
      " dropout_3 (Dropout)         (None, None, 8)           0         \n",
      "                                                                 \n",
      " global_max_pooling1d_29 (Gl  (None, 8)                0         \n",
      " obalMaxPooling1D)                                               \n",
      "                                                                 \n",
      " dense_58 (Dense)            (None, 16)                144       \n",
      "                                                                 \n",
      " dense_59 (Dense)            (None, 1)                 17        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 80,161\n",
      "Trainable params: 80,161\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "vocab_size = 10000  # 어휘 사전의 크기입니다(10개의 단어)\n",
    "word_vector_dim = 8   # 단어 하나를 표현하는 임베딩 벡터의 차원 수입니다. \n",
    "\n",
    "model = tf.keras.Sequential()\n",
    "model.add(tf.keras.layers.Embedding(vocab_size, word_vector_dim, input_shape=(None,)))\n",
    "model.add(tf.keras.layers.Dropout(0.2))\n",
    "model.add(tf.keras.layers.GlobalMaxPooling1D())\n",
    "model.add(tf.keras.layers.Dense(16, activation='relu'))\n",
    "model.add(tf.keras.layers.Dense(1, activation='sigmoid'))  # 최종 출력은 긍정/부정을 나타내는 1dim 입니다.\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "ba93ec46",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "149/149 [==============================] - 2s 11ms/step - loss: 0.6579 - accuracy: 0.7004 - val_loss: 0.5922 - val_accuracy: 0.7610\n",
      "Epoch 2/5\n",
      "149/149 [==============================] - 1s 8ms/step - loss: 0.5101 - accuracy: 0.7788 - val_loss: 0.4566 - val_accuracy: 0.7996\n",
      "Epoch 3/5\n",
      "149/149 [==============================] - 1s 8ms/step - loss: 0.4246 - accuracy: 0.8107 - val_loss: 0.4126 - val_accuracy: 0.8176\n",
      "Epoch 4/5\n",
      "149/149 [==============================] - 1s 7ms/step - loss: 0.3955 - accuracy: 0.8257 - val_loss: 0.4011 - val_accuracy: 0.8225\n",
      "Epoch 5/5\n",
      "149/149 [==============================] - 1s 7ms/step - loss: 0.3817 - accuracy: 0.8322 - val_loss: 0.3971 - val_accuracy: 0.8245\n",
      "1537/1537 [==============================] - 2s 1ms/step - loss: 0.4024 - accuracy: 0.8189\n",
      "[0.40238264203071594, 0.818886399269104]\n"
     ]
    }
   ],
   "source": [
    "with tf.device('/device:CPU:0'):\n",
    "    model.compile(optimizer='adam',\n",
    "             loss='binary_crossentropy',\n",
    "             metrics=['accuracy'])\n",
    "\n",
    "    epochs=5\n",
    "\n",
    "    history = model.fit(partial_x_train,\n",
    "                    partial_y_train,\n",
    "                    epochs = epochs,\n",
    "                    batch_size = 512,\n",
    "                    validation_data = (X_val, y_val),\n",
    "                    verbose=1)\n",
    "    results = model.evaluate(X_test, y_test)\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "83301091",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history_dict = history.history\n",
    "history_dict.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "1d285e8a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "loss = history_dict['loss']\n",
    "val_loss = history_dict['val_loss']\n",
    "acc = history_dict['accuracy']\n",
    "val_acc = history_dict['val_accuracy']\n",
    "\n",
    "epochs = range(1,len(acc)+1)\n",
    "\n",
    "plt.plot(epochs, loss, c='r', label='Training loss')\n",
    "plt.plot(epochs,val_loss, c='b', label = 'Val loss')\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "id": "a795928a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(epochs, acc ,'r', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label = 'Validation acc')\n",
    "plt.title('Training and Validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('acc')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6b1451d",
   "metadata": {},
   "source": [
    "dim = 32 16 0.842    \n",
    "dim = 16 16 0.839    \n",
    "dim = 8 16 0.823    \n",
    "dim = 32 32 0.84    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7f0ccb4b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.9.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}