ModernKoreanSubword_Train.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "view-in-github"
   },
   "source": [
    "<a href=\"https://colab.research.google.com/github/ByungjunKim/ModernKoreanSubword/blob/main/ModernKoreanSubword_Train.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "J0SDHrHgdyTN"
   },
   "source": [
    "# 근대 한국어 서브워드 형태소 분석기 학습"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2e7SnyYRinXo"
   },
   "source": [
    "### 서브워드 학습 준비"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "JXLd3XvGhn0H"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# kiwi 로드\n",
    "from kiwipiepy import Kiwi\n",
    "kiwi = Kiwi()\n",
    "from kiwipiepy.sw_tokenizer import SwTokenizer\n",
    "from kiwipiepy.sw_tokenizer import SwTokenizerConfig\n",
    "import gc\n",
    "import hanja\n",
    "import pickle\n",
    "import bz2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 한국근대잡지 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>level3id</th>\n",
       "      <th>mainTitle</th>\n",
       "      <th>subjectClass</th>\n",
       "      <th>pTitle</th>\n",
       "      <th>paragraph</th>\n",
       "      <th>level2id</th>\n",
       "      <th>level1id</th>\n",
       "      <th>mainTitle_권호</th>\n",
       "      <th>publisher</th>\n",
       "      <th>dateIssued</th>\n",
       "      <th>mainTitle_잡지</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ma_001_0010_0010</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>논설</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>\\n大韓自强會月報序［三編］\\n</td>\n",
       "      <td>ma_001_0010</td>\n",
       "      <td>ma_001</td>\n",
       "      <td>대한자강회월보 제1호</td>\n",
       "      <td>대한자강회</td>\n",
       "      <td>1906-07-31</td>\n",
       "      <td>대한자강회월보</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ma_001_0010_0010</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>논설</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>＜1-4＞</td>\n",
       "      <td>ma_001_0010</td>\n",
       "      <td>ma_001</td>\n",
       "      <td>대한자강회월보 제1호</td>\n",
       "      <td>대한자강회</td>\n",
       "      <td>1906-07-31</td>\n",
       "      <td>대한자강회월보</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ma_001_0010_0010</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>논설</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>\\n南嵩山人曰誰道日本之少文章哉此篇文字明確正大其組織極精緻極宏博至其論國家正義處尤切時病令...</td>\n",
       "      <td>ma_001_0010</td>\n",
       "      <td>ma_001</td>\n",
       "      <td>대한자강회월보 제1호</td>\n",
       "      <td>대한자강회</td>\n",
       "      <td>1906-07-31</td>\n",
       "      <td>대한자강회월보</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ma_001_0010_0010</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>논설</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>\\n金陵\\n大垣丈夫\\n</td>\n",
       "      <td>ma_001_0010</td>\n",
       "      <td>ma_001</td>\n",
       "      <td>대한자강회월보 제1호</td>\n",
       "      <td>대한자강회</td>\n",
       "      <td>1906-07-31</td>\n",
       "      <td>대한자강회월보</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ma_001_0010_0010</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>논설</td>\n",
       "      <td>大韓自强會月報序［三編］</td>\n",
       "      <td>蓋韓國二千萬人中豈無達者也哉余嘗結交其士君子知有老練精達之人傑見有少壯有爲之人士而今則如此何...</td>\n",
       "      <td>ma_001_0010</td>\n",
       "      <td>ma_001</td>\n",
       "      <td>대한자강회월보 제1호</td>\n",
       "      <td>대한자강회</td>\n",
       "      <td>1906-07-31</td>\n",
       "      <td>대한자강회월보</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588569</th>\n",
       "      <td>ma_095_0260_0140</td>\n",
       "      <td>버드나무 그늘의 少女</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>버드나무 그늘의 少女</td>\n",
       "      <td>\\n버드나무 그늘의 少女\\n尹影\\n</td>\n",
       "      <td>ma_095_0260</td>\n",
       "      <td>ma_095</td>\n",
       "      <td>호남평론 제3권 제8호</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>1937-08-15</td>\n",
       "      <td>호남평론</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588570</th>\n",
       "      <td>ma_095_0260_0150</td>\n",
       "      <td>詩: 바다가에서</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>詩: 바다가에서</td>\n",
       "      <td>\\n詩: 바다가에서\\n一石\\n</td>\n",
       "      <td>ma_095_0260</td>\n",
       "      <td>ma_095</td>\n",
       "      <td>호남평론 제3권 제8호</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>1937-08-15</td>\n",
       "      <td>호남평론</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588571</th>\n",
       "      <td>ma_095_0260_0160</td>\n",
       "      <td>戱曲: 水害後</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>戱曲: 水害後</td>\n",
       "      <td>\\n戱曲: 水害後\\n朴東和\\n</td>\n",
       "      <td>ma_095_0260</td>\n",
       "      <td>ma_095</td>\n",
       "      <td>호남평론 제3권 제8호</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>1937-08-15</td>\n",
       "      <td>호남평론</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588572</th>\n",
       "      <td>ma_095_0260_0170</td>\n",
       "      <td>詩: H孃의 半生記</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>詩: H孃의 半生記</td>\n",
       "      <td>\\n詩: H孃의 半生記\\n溝隱\\n</td>\n",
       "      <td>ma_095_0260</td>\n",
       "      <td>ma_095</td>\n",
       "      <td>호남평론 제3권 제8호</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>1937-08-15</td>\n",
       "      <td>호남평론</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588573</th>\n",
       "      <td>ma_095_0260_0180</td>\n",
       "      <td>創作: 不貞女</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>創作: 不貞女</td>\n",
       "      <td>\\n創作: 不貞女\\n野人\\n</td>\n",
       "      <td>ma_095_0260</td>\n",
       "      <td>ma_095</td>\n",
       "      <td>호남평론 제3권 제8호</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>1937-08-15</td>\n",
       "      <td>호남평론</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>588574 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                level3id     mainTitle subjectClass        pTitle  \\\n",
       "0       ma_001_0010_0010  大韓自强會月報序［三編］           논설  大韓自强會月報序［三編］   \n",
       "1       ma_001_0010_0010  大韓自强會月報序［三編］           논설  大韓自强會月報序［三編］   \n",
       "2       ma_001_0010_0010  大韓自强會月報序［三編］           논설  大韓自强會月報序［三編］   \n",
       "3       ma_001_0010_0010  大韓自强會月報序［三編］           논설  大韓自强會月報序［三編］   \n",
       "4       ma_001_0010_0010  大韓自强會月報序［三編］           논설  大韓自强會月報序［三編］   \n",
       "...                  ...           ...          ...           ...   \n",
       "588569  ma_095_0260_0140   버드나무 그늘의 少女         <NA>   버드나무 그늘의 少女   \n",
       "588570  ma_095_0260_0150      詩: 바다가에서         <NA>      詩: 바다가에서   \n",
       "588571  ma_095_0260_0160       戱曲: 水害後         <NA>       戱曲: 水害後   \n",
       "588572  ma_095_0260_0170    詩: H孃의 半生記         <NA>    詩: H孃의 半生記   \n",
       "588573  ma_095_0260_0180       創作: 不貞女         <NA>       創作: 不貞女   \n",
       "\n",
       "                                                paragraph     level2id  \\\n",
       "0                                        \\n大韓自强會月報序［三編］\\n  ma_001_0010   \n",
       "1                                                   ＜1-4＞  ma_001_0010   \n",
       "2       \\n南嵩山人曰誰道日本之少文章哉此篇文字明確正大其組織極精緻極宏博至其論國家正義處尤切時病令...  ma_001_0010   \n",
       "3                                            \\n金陵\\n大垣丈夫\\n  ma_001_0010   \n",
       "4       蓋韓國二千萬人中豈無達者也哉余嘗結交其士君子知有老練精達之人傑見有少壯有爲之人士而今則如此何...  ma_001_0010   \n",
       "...                                                   ...          ...   \n",
       "588569                                \\n버드나무 그늘의 少女\\n尹影\\n  ma_095_0260   \n",
       "588570                                   \\n詩: 바다가에서\\n一石\\n  ma_095_0260   \n",
       "588571                                   \\n戱曲: 水害後\\n朴東和\\n  ma_095_0260   \n",
       "588572                                 \\n詩: H孃의 半生記\\n溝隱\\n  ma_095_0260   \n",
       "588573                                    \\n創作: 不貞女\\n野人\\n  ma_095_0260   \n",
       "\n",
       "       level1id  mainTitle_권호 publisher  dateIssued mainTitle_잡지  \n",
       "0        ma_001   대한자강회월보 제1호     대한자강회  1906-07-31      대한자강회월보  \n",
       "1        ma_001   대한자강회월보 제1호     대한자강회  1906-07-31      대한자강회월보  \n",
       "2        ma_001   대한자강회월보 제1호     대한자강회  1906-07-31      대한자강회월보  \n",
       "3        ma_001   대한자강회월보 제1호     대한자강회  1906-07-31      대한자강회월보  \n",
       "4        ma_001   대한자강회월보 제1호     대한자강회  1906-07-31      대한자강회월보  \n",
       "...         ...           ...       ...         ...          ...  \n",
       "588569   ma_095  호남평론 제3권 제8호      <NA>  1937-08-15         호남평론  \n",
       "588570   ma_095  호남평론 제3권 제8호      <NA>  1937-08-15         호남평론  \n",
       "588571   ma_095  호남평론 제3권 제8호      <NA>  1937-08-15         호남평론  \n",
       "588572   ma_095  호남평론 제3권 제8호      <NA>  1937-08-15         호남평론  \n",
       "588573   ma_095  호남평론 제3권 제8호      <NA>  1937-08-15         호남평론  \n",
       "\n",
       "[588574 rows x 11 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 근대 잡지\n",
    "with bz2.BZ2File('./한국근현대잡지자료.pkl.bz2', 'rb') as f:\n",
    "   mag = pickle.load(f)\n",
    "mag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 연도 추출\n",
    "mag['year'] = mag['dateIssued'].str.slice(0,4).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1940까지 활용\n",
    "mag = mag[mag['year']<=1940].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='year'>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mag['year'].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1890년대      1664\n",
      "1900년대     47261\n",
      "1910년대      8198\n",
      "1920년대    171235\n",
      "1930년대    302766\n",
      "1940년대     26357\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# 10년 단위로 문단수 통계 내기\n",
    "# Create a new column for 10-year intervals\n",
    "mag['decade'] = (mag['year'] // 10) * 10\n",
    "\n",
    "# Calculate the frequency of each decade\n",
    "decade_counts = mag['decade'].value_counts().sort_index()\n",
    "\n",
    "# Format the index to display as \"1880년대\", \"1890년대\", etc.\n",
    "decade_counts.index = [f\"{decade}년대\" for decade in decade_counts.index]\n",
    "\n",
    "# Display the result\n",
    "print(decade_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "mag_paragraph = mag['paragraph'].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "del mag"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 조선일보"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 조선일보\n",
    "chosun = pd.read_parquet('/mnt/hdd/bjkim/ModernConcept/chosun_preprocessed.parquet',columns=['year','type','body_archaic'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>year</th>\n",
       "      <th>type</th>\n",
       "      <th>body_archaic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>smallad</td>\n",
       "      <td>祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1920</td>\n",
       "      <td>etc</td>\n",
       "      <td>祝詞와祝電\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4679014</th>\n",
       "      <td>1999</td>\n",
       "      <td>article</td>\n",
       "      <td>폐품 할머니\\n40년간 고물 수집\\n외대 1억 기증후 타계\\n 지난 7월 장경자(張...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4679015</th>\n",
       "      <td>1999</td>\n",
       "      <td>article</td>\n",
       "      <td>고시준비생 장학금\\n외대에 3억원 전달\\n商街주인 할머니\\n 지난 14일힐튼호텔 컨...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4679016</th>\n",
       "      <td>1999</td>\n",
       "      <td>article</td>\n",
       "      <td>“1900년에 출두하시오”\\n수원지법 컴퓨터‘Y2K’\\n피고인에 엉뚱한 소환장\\n ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4679017</th>\n",
       "      <td>1999</td>\n",
       "      <td>photo</td>\n",
       "      <td>◇故장경자씨\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4679018</th>\n",
       "      <td>1999</td>\n",
       "      <td>article</td>\n",
       "      <td>콩나물 장사 30여년\\n충북대 12억건물 기증\\n콩나물 할머니\\n “모든 걸주고 나...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3419546 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         year     type                                       body_archaic\n",
       "0        1920  smallad                     祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n\n",
       "4        1920  article  朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...\n",
       "5        1920  article  朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...\n",
       "6        1920      etc                                            祝詞와祝電\\n\n",
       "7        1920  article  大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...\n",
       "...       ...      ...                                                ...\n",
       "4679014  1999  article  폐품 할머니\\n40년간 고물 수집\\n외대 1억 기증후 타계\\n 지난 7월 장경자(張...\n",
       "4679015  1999  article  고시준비생 장학금\\n외대에 3억원 전달\\n商街주인 할머니\\n 지난 14일힐튼호텔 컨...\n",
       "4679016  1999  article  “1900년에 출두하시오”\\n수원지법 컴퓨터‘Y2K’\\n피고인에 엉뚱한 소환장\\n ...\n",
       "4679017  1999    photo                                           ◇故장경자씨\\n\n",
       "4679018  1999  article  콩나물 장사 30여년\\n충북대 12억건물 기증\\n콩나물 할머니\\n “모든 걸주고 나...\n",
       "\n",
       "[3419546 rows x 3 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chosun = chosun.dropna()\n",
    "chosun"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>year</th>\n",
       "      <th>type</th>\n",
       "      <th>body_archaic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>smallad</td>\n",
       "      <td>祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1920</td>\n",
       "      <td>etc</td>\n",
       "      <td>祝詞와祝電\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>930051</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>各界人物招致\\n首相,意見交換豫定\\n【東京電話同盟】新體制確立에關하여 그根本原理組織方法...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>930052</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>臨時家族手當\\n十月부터實施决定\\n【東京電話同盟】政府는 九日閣議에서 下級官吏에 臨時家...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>930053</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>外相,首相과會談\\n【東京電話同盟】松岡外相은 九日定例閣議散會後 首相官邸에남어 近衞首相...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>930054</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>人事·集會\\n▲朱□宜氏(前普成中學校長)退任人事次十日來社\\n▲鈴木文次郞氏(全朝鮮洋服商...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>930055</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>洞庭湖銃砲擊\\n【漢口九日發同盟】中支艦隊報道部 八月九日午後四時發表=\\n一,江上艦隊一部...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>930056 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        year     type                                       body_archaic\n",
       "0       1920  smallad                     祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n\n",
       "1       1920  article  朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...\n",
       "2       1920  article  朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...\n",
       "3       1920      etc                                            祝詞와祝電\\n\n",
       "4       1920  article  大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...\n",
       "...      ...      ...                                                ...\n",
       "930051  1940  article  各界人物招致\\n首相,意見交換豫定\\n【東京電話同盟】新體制確立에關하여 그根本原理組織方法...\n",
       "930052  1940  article  臨時家族手當\\n十月부터實施决定\\n【東京電話同盟】政府는 九日閣議에서 下級官吏에 臨時家...\n",
       "930053  1940  article  外相,首相과會談\\n【東京電話同盟】松岡外相은 九日定例閣議散會後 首相官邸에남어 近衞首相...\n",
       "930054  1940  article  人事·集會\\n▲朱□宜氏(前普成中學校長)退任人事次十日來社\\n▲鈴木文次郞氏(全朝鮮洋服商...\n",
       "930055  1940  article  洞庭湖銃砲擊\\n【漢口九日發同盟】中支艦隊報道部 八月九日午後四時發表=\\n一,江上艦隊一部...\n",
       "\n",
       "[930056 rows x 3 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chosun = chosun[chosun['year']<=1940].reset_index(drop=True)\n",
    "chosun"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>year</th>\n",
       "      <th>type</th>\n",
       "      <th>body_archaic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>smallad</td>\n",
       "      <td>祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1920</td>\n",
       "      <td>etc</td>\n",
       "      <td>祝詞와祝電\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900622</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>各界人物招致\\n首相,意見交換豫定\\n【東京電話同盟】新體制確立에關하여 그根本原理組織方法...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900623</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>臨時家族手當\\n十月부터實施决定\\n【東京電話同盟】政府는 九日閣議에서 下級官吏에 臨時家...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900624</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>外相,首相과會談\\n【東京電話同盟】松岡外相은 九日定例閣議散會後 首相官邸에남어 近衞首相...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900625</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>人事·集會\\n▲朱□宜氏(前普成中學校長)退任人事次十日來社\\n▲鈴木文次郞氏(全朝鮮洋服商...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900626</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>洞庭湖銃砲擊\\n【漢口九日發同盟】中支艦隊報道部 八月九日午後四時發表=\\n一,江上艦隊一部...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900627 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        year     type                                       body_archaic\n",
       "0       1920  smallad                     祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n\n",
       "1       1920  article  朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...\n",
       "2       1920  article  朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...\n",
       "3       1920      etc                                            祝詞와祝電\\n\n",
       "4       1920  article  大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...\n",
       "...      ...      ...                                                ...\n",
       "900622  1940  article  各界人物招致\\n首相,意見交換豫定\\n【東京電話同盟】新體制確立에關하여 그根本原理組織方法...\n",
       "900623  1940  article  臨時家族手當\\n十月부터實施决定\\n【東京電話同盟】政府는 九日閣議에서 下級官吏에 臨時家...\n",
       "900624  1940  article  外相,首相과會談\\n【東京電話同盟】松岡外相은 九日定例閣議散會後 首相官邸에남어 近衞首相...\n",
       "900625  1940  article  人事·集會\\n▲朱□宜氏(前普成中學校長)退任人事次十日來社\\n▲鈴木文次郞氏(全朝鮮洋服商...\n",
       "900626  1940  article  洞庭湖銃砲擊\\n【漢口九日發同盟】中支艦隊報道部 八月九日午後四時發表=\\n一,江上艦隊一部...\n",
       "\n",
       "[900627 rows x 3 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 중복제거\n",
    "chosun = chosun.drop_duplicates('body_archaic').reset_index(drop=True)\n",
    "chosun"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# br 태그 삭제\n",
    "chosun['body_archaic'] = chosun['body_archaic'].str.replace('</br>','')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 【三十日東京特電】 처럼 특파원 삭제\n",
    "chosun['body_archaic'] = chosun['body_archaic'].str.replace('\\【.+\\】','',regex=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>year</th>\n",
       "      <th>type</th>\n",
       "      <th>body_archaic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>smallad</td>\n",
       "      <td>祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1920</td>\n",
       "      <td>etc</td>\n",
       "      <td>祝詞와祝電\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>article</td>\n",
       "      <td>大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900471</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>各界人物招致\\n首相,意見交換豫定\\n新體制確立에關하여 그根本原理組織方法等에關하여 愼重...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900472</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>臨時家族手當\\n十月부터實施决定\\n政府는 九日閣議에서 下級官吏에 臨時家族手當支給을 决...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900473</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>外相,首相과會談\\n松岡外相은 九日定例閣議散會後 首相官邸에남어 近衞首相과 會見하고 當...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900474</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>人事·集會\\n▲朱□宜氏(前普成中學校長)退任人事次十日來社\\n▲鈴木文次郞氏(全朝鮮洋服商...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>900475</th>\n",
       "      <td>1940</td>\n",
       "      <td>article</td>\n",
       "      <td>洞庭湖銃砲擊\\n中支艦隊報道部 八月九日午後四時發表=\\n一,江上艦隊一部는 八日岳州南 方...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>900476 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        year     type                                       body_archaic\n",
       "0       1920  smallad                     祝 …朝；蘚 n 報創 刊\\n趙… . … . 命· 九\\n\n",
       "1       1920  article  朝鮮日報\\n創刊을祝함\\n候爵 大隈重信\\n朝鮮合邦以來歲를閱한지玆에十年이되鮮人의自營하난...\n",
       "2       1920  article  朝鮮現下의\\n工業狀態\\n를鑑하야振興策\\n을一二를言함\\n朝鮮의工業으로言하면從來此를賤視...\n",
       "3       1920      etc                                            祝詞와祝電\\n\n",
       "4       1920  article  大正八年朝鮮貿易槪况(一)\\n(朝鮮總督府財務局調査)\\n一,總况\\n大正八年朝鮮貿易額은輸...\n",
       "...      ...      ...                                                ...\n",
       "900471  1940  article  各界人物招致\\n首相,意見交換豫定\\n新體制確立에關하여 그根本原理組織方法等에關하여 愼重...\n",
       "900472  1940  article  臨時家族手當\\n十月부터實施决定\\n政府는 九日閣議에서 下級官吏에 臨時家族手當支給을 决...\n",
       "900473  1940  article  外相,首相과會談\\n松岡外相은 九日定例閣議散會後 首相官邸에남어 近衞首相과 會見하고 當...\n",
       "900474  1940  article  人事·集會\\n▲朱□宜氏(前普成中學校長)退任人事次十日來社\\n▲鈴木文次郞氏(全朝鮮洋服商...\n",
       "900475  1940  article  洞庭湖銃砲擊\\n中支艦隊報道部 八月九日午後四時發表=\\n一,江上艦隊一部는 八日岳州南 方...\n",
       "\n",
       "[900476 rows x 3 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chosun = chosun.dropna(subset=['body_archaic']).reset_index(drop=True)\n",
    "chosun = chosun.drop_duplicates('body_archaic').reset_index(drop=True)\n",
    "chosun"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1920년대    340320\n",
      "1930년대    525253\n",
      "1940년대     34903\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# 10년 단위로 문단수 통계 내기\n",
    "# Create a new column for 10-year intervals\n",
    "chosun['decade'] = (chosun['year'] // 10) * 10\n",
    "\n",
    "# Calculate the frequency of each decade\n",
    "decade_counts = chosun['decade'].value_counts().sort_index()\n",
    "\n",
    "# Format the index to display as \"1880년대\", \"1890년대\", etc.\n",
    "decade_counts.index = [f\"{decade}년대\" for decade in decade_counts.index]\n",
    "\n",
    "# Display the result\n",
    "print(decade_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "chosun_paragraph = chosun['body_archaic'].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "del chosun"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 동아일보 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>year</th>\n",
       "      <th>body_ori</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>主旨를宣明하노라\\n蒼天에太陽이빗나고大地에淸風이불도다山靜水流하며草木昌茂하며百花爛發하며...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1920</td>\n",
       "      <td>九十度난太甚\\n혹독한형벌인우에\\n따리는데도불공도\\n스코필트씨밀\\n관텽에서 허가하는 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1920</td>\n",
       "      <td>肉體보다도精神\\n몸을앗김보다\\n명예를위하라\\n방하춍독무의원장의말사월초하ㄹ날로무터 태...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1920</td>\n",
       "      <td>四十七人의公判\\n셰길이넘는긔록을조사하기에\\n판검사아홉명이매여달녀잇다\\n느저도금년안에...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>姜宇奎公判은\\n금월오일개정\\n남대문역두에서 재등춍독에게폭탄을던지여 한세상을 떠들니든...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3125897</th>\n",
       "      <td>1999</td>\n",
       "      <td>키노코너\\n독립영화협의회가 주최하는 제26회 독립영화 워크숍이 2000년 1월3일부...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3125898</th>\n",
       "      <td>1999</td>\n",
       "      <td>한국영화 대약진“올해만 같아라”\\n점유율 40% 육박…작품성은 미흡\\n외화 상대적 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3125899</th>\n",
       "      <td>1999</td>\n",
       "      <td>1999년 영화계 결산\\n99 한국영화 흥행 베스트10\\n99 외국영화 흥행 베스트10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3125900</th>\n",
       "      <td>1999</td>\n",
       "      <td>Y2K D-1\\n업무 본격시작 1월3,4일 더 위험\\n‘Y2K D—1’\\n컴퓨터의 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3125901</th>\n",
       "      <td>1999</td>\n",
       "      <td>만화속의 요리</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3125902 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         year                                           body_ori\n",
       "0        1920  主旨를宣明하노라\\n蒼天에太陽이빗나고大地에淸風이불도다山靜水流하며草木昌茂하며百花爛發하며...\n",
       "1        1920  九十度난太甚\\n혹독한형벌인우에\\n따리는데도불공도\\n스코필트씨밀\\n관텽에서 허가하는 ...\n",
       "2        1920  肉體보다도精神\\n몸을앗김보다\\n명예를위하라\\n방하춍독무의원장의말사월초하ㄹ날로무터 태...\n",
       "3        1920  四十七人의公判\\n셰길이넘는긔록을조사하기에\\n판검사아홉명이매여달녀잇다\\n느저도금년안에...\n",
       "4        1920  姜宇奎公判은\\n금월오일개정\\n남대문역두에서 재등춍독에게폭탄을던지여 한세상을 떠들니든...\n",
       "...       ...                                                ...\n",
       "3125897  1999  키노코너\\n독립영화협의회가 주최하는 제26회 독립영화 워크숍이 2000년 1월3일부...\n",
       "3125898  1999  한국영화 대약진“올해만 같아라”\\n점유율 40% 육박…작품성은 미흡\\n외화 상대적 ...\n",
       "3125899  1999   1999년 영화계 결산\\n99 한국영화 흥행 베스트10\\n99 외국영화 흥행 베스트10\n",
       "3125900  1999  Y2K D-1\\n업무 본격시작 1월3,4일 더 위험\\n‘Y2K D—1’\\n컴퓨터의 ...\n",
       "3125901  1999                                            만화속의 요리\n",
       "\n",
       "[3125902 rows x 2 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 동아일보\n",
    "donga = pd.read_parquet('/mnt/hdd/bjkim/ModernConcept/donga_preprocessed.parquet',columns=['year','body_ori'])\n",
    "donga"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>year</th>\n",
       "      <th>body_ori</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>主旨를宣明하노라\\n蒼天에太陽이빗나고大地에淸風이불도다山靜水流하며草木昌茂하며百花爛發하며...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1920</td>\n",
       "      <td>九十度난太甚\\n혹독한형벌인우에\\n따리는데도불공도\\n스코필트씨밀\\n관텽에서 허가하는 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1920</td>\n",
       "      <td>肉體보다도精神\\n몸을앗김보다\\n명예를위하라\\n방하춍독무의원장의말사월초하ㄹ날로무터 태...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1920</td>\n",
       "      <td>四十七人의公判\\n셰길이넘는긔록을조사하기에\\n판검사아홉명이매여달녀잇다\\n느저도금년안에...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>姜宇奎公判은\\n금월오일개정\\n남대문역두에서 재등춍독에게폭탄을던지여 한세상을 떠들니든...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1001312</th>\n",
       "      <td>1940</td>\n",
       "      <td>手術界의驚異\\n醫學朝鮮에凱旋歌\\n없는男子生殖 를붙이기에 드디어成功\\n世專李鶴松博士의...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1001313</th>\n",
       "      <td>1940</td>\n",
       "      <td>三年後에오는大京城\\n二百萬人口를收容\\n十八年度 지七個地區에新市街地를形成\\n京城府에서...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1001314</th>\n",
       "      <td>1940</td>\n",
       "      <td>張皷峯事件二 年記念\\n今日、嚴肅히擧行\\n正午를期해一齊默禱\\n황군전사(皇軍戰史)에 최...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1001315</th>\n",
       "      <td>1940</td>\n",
       "      <td>辯護士試驗延長\\n受驗生代表들이再次當局에陳情\\n누보한바와같이 조선에잇어서의 변호사시험...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1001316</th>\n",
       "      <td>1940</td>\n",
       "      <td>第二次中等敎育擴充計劃\\n十八年度부터七個年間\\n志願의三分一을收容\\n中等校、實業校를半半...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1001317 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         year                                           body_ori\n",
       "0        1920  主旨를宣明하노라\\n蒼天에太陽이빗나고大地에淸風이불도다山靜水流하며草木昌茂하며百花爛發하며...\n",
       "1        1920  九十度난太甚\\n혹독한형벌인우에\\n따리는데도불공도\\n스코필트씨밀\\n관텽에서 허가하는 ...\n",
       "2        1920  肉體보다도精神\\n몸을앗김보다\\n명예를위하라\\n방하춍독무의원장의말사월초하ㄹ날로무터 태...\n",
       "3        1920  四十七人의公判\\n셰길이넘는긔록을조사하기에\\n판검사아홉명이매여달녀잇다\\n느저도금년안에...\n",
       "4        1920  姜宇奎公判은\\n금월오일개정\\n남대문역두에서 재등춍독에게폭탄을던지여 한세상을 떠들니든...\n",
       "...       ...                                                ...\n",
       "1001312  1940  手術界의驚異\\n醫學朝鮮에凱旋歌\\n없는男子生殖 를붙이기에 드디어成功\\n世專李鶴松博士의...\n",
       "1001313  1940  三年後에오는大京城\\n二百萬人口를收容\\n十八年度 지七個地區에新市街地를形成\\n京城府에서...\n",
       "1001314  1940  張皷峯事件二 年記念\\n今日、嚴肅히擧行\\n正午를期해一齊默禱\\n황군전사(皇軍戰史)에 최...\n",
       "1001315  1940  辯護士試驗延長\\n受驗生代表들이再次當局에陳情\\n누보한바와같이 조선에잇어서의 변호사시험...\n",
       "1001316  1940  第二次中等敎育擴充計劃\\n十八年度부터七個年間\\n志願의三分一을收容\\n中等校、實業校를半半...\n",
       "\n",
       "[1001317 rows x 2 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "donga = donga[donga['year']<=1940].dropna().reset_index(drop=True)\n",
    "donga"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>year</th>\n",
       "      <th>body_ori</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1920</td>\n",
       "      <td>主旨를宣明하노라\\n蒼天에太陽이빗나고大地에淸風이불도다山靜水流하며草木昌茂하며百花爛發하며...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1920</td>\n",
       "      <td>九十度난太甚\\n혹독한형벌인우에\\n따리는데도불공도\\n스코필트씨밀\\n관텽에서 허가하는 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1920</td>\n",
       "      <td>肉體보다도精神\\n몸을앗김보다\\n명예를위하라\\n방하춍독무의원장의말사월초하ㄹ날로무터 태...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1920</td>\n",
       "      <td>四十七人의公判\\n셰길이넘는긔록을조사하기에\\n판검사아홉명이매여달녀잇다\\n느저도금년안에...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1920</td>\n",
       "      <td>姜宇奎公判은\\n금월오일개정\\n남대문역두에서 재등춍독에게폭탄을던지여 한세상을 떠들니든...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922467</th>\n",
       "      <td>1940</td>\n",
       "      <td>手術界의驚異\\n醫學朝鮮에凱旋歌\\n없는男子生殖 를붙이기에 드디어成功\\n世專李鶴松博士의...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922468</th>\n",
       "      <td>1940</td>\n",
       "      <td>三年後에오는大京城\\n二百萬人口를收容\\n十八年度 지七個地區에新市街地를形成\\n京城府에서...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922469</th>\n",
       "      <td>1940</td>\n",
       "      <td>張皷峯事件二 年記念\\n今日、嚴肅히擧行\\n正午를期해一齊默禱\\n황군전사(皇軍戰史)에 최...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922470</th>\n",
       "      <td>1940</td>\n",
       "      <td>辯護士試驗延長\\n受驗生代表들이再次當局에陳情\\n누보한바와같이 조선에잇어서의 변호사시험...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922471</th>\n",
       "      <td>1940</td>\n",
       "      <td>第二次中等敎育擴充計劃\\n十八年度부터七個年間\\n志願의三分一을收容\\n中等校、實業校를半半...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>922472 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        year                                           body_ori\n",
       "0       1920  主旨를宣明하노라\\n蒼天에太陽이빗나고大地에淸風이불도다山靜水流하며草木昌茂하며百花爛發하며...\n",
       "1       1920  九十度난太甚\\n혹독한형벌인우에\\n따리는데도불공도\\n스코필트씨밀\\n관텽에서 허가하는 ...\n",
       "2       1920  肉體보다도精神\\n몸을앗김보다\\n명예를위하라\\n방하춍독무의원장의말사월초하ㄹ날로무터 태...\n",
       "3       1920  四十七人의公判\\n셰길이넘는긔록을조사하기에\\n판검사아홉명이매여달녀잇다\\n느저도금년안에...\n",
       "4       1920  姜宇奎公判은\\n금월오일개정\\n남대문역두에서 재등춍독에게폭탄을던지여 한세상을 떠들니든...\n",
       "...      ...                                                ...\n",
       "922467  1940  手術界의驚異\\n醫學朝鮮에凱旋歌\\n없는男子生殖 를붙이기에 드디어成功\\n世專李鶴松博士의...\n",
       "922468  1940  三年後에오는大京城\\n二百萬人口를收容\\n十八年度 지七個地區에新市街地를形成\\n京城府에서...\n",
       "922469  1940  張皷峯事件二 年記念\\n今日、嚴肅히擧行\\n正午를期해一齊默禱\\n황군전사(皇軍戰史)에 최...\n",
       "922470  1940  辯護士試驗延長\\n受驗生代表들이再次當局에陳情\\n누보한바와같이 조선에잇어서의 변호사시험...\n",
       "922471  1940  第二次中等敎育擴充計劃\\n十八年度부터七個年間\\n志願의三分一을收容\\n中等校、實業校를半半...\n",
       "\n",
       "[922472 rows x 2 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "donga = donga.drop_duplicates('body_ori').reset_index(drop=True)\n",
    "donga"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1920년대    372290\n",
      "1930년대    511036\n",
      "1940년대     39146\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# 10년 단위로 문단수 통계 내기\n",
    "# Create a new column for 10-year intervals\n",
    "donga['decade'] = (donga['year'] // 10) * 10\n",
    "\n",
    "# Calculate the frequency of each decade\n",
    "decade_counts = donga['decade'].value_counts().sort_index()\n",
    "\n",
    "# Format the index to display as \"1880년대\", \"1890년대\", etc.\n",
    "decade_counts.index = [f\"{decade}년대\" for decade in decade_counts.index]\n",
    "\n",
    "# Display the result\n",
    "print(decade_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "donga_paragraph = donga['body_ori'].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "del donga"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2380429"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "paragraph_list = mag_paragraph + chosun_paragraph + donga_paragraph\n",
    "len(paragraph_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "del mag_paragraph, chosun_paragraph, donga_paragraph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "184"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2380429"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(paragraph_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2287985"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 중복삭제\n",
    "len(list(dict.fromkeys(paragraph_list)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "paragraph_list = list(dict.fromkeys(paragraph_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save\n",
    "# with open('./paragraph_list.pickle', 'wb') as f:\n",
    "#     pickle.dump(paragraph_list, f, pickle.HIGHEST_PROTOCOL)\n",
    "with bz2.BZ2File('./paragraph_list.pkl.bz2', 'wb') as f:\n",
    "    pickle.dump(paragraph_list, f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "n3MamwhJilVM"
   },
   "source": [
    "### 서브워드 학습"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load\n",
    "# with open('./paragraph_list.pickle', 'rb') as f:\n",
    "#     paragraph_list = pickle.load(f)\n",
    "with bz2.BZ2File('paragraph_list.pkl.bz2', 'rb') as f:\n",
    "   paragraph_list = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "cKEWkhlRif06"
   },
   "outputs": [],
   "source": [
    "# https://bab2min.github.io/kiwipiepy/v0.15.2/kr/sw_tokenizer.html#kiwipiepy.sw_tokenizer.SwTokenizerConfig\n",
    "config = SwTokenizerConfig(split_chinese=False) # split_chinese 제외하고 모두 기본 옵션 사용"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "kRKKDEvhiy61",
    "outputId": "b073aa0c-2ea7-4165-fc10-0647c7300e39"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Tokenizing: 100%|██████████| 2287985/2287985 [15:56<00:00, 2391.81it/s] \n",
      "Reducing #1:   0%|          | 0/1000 [00:00<?, ?it/s, vocab_size=8557579, loss=8.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 0 VocabSize: 8557579 Loss: 8.7211\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 1/1000 [00:30<8:29:12, 30.58s/it, vocab_size=6155199, loss=8.58]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 1 VocabSize: 6155199 Loss: 8.5833\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 2/1000 [01:01<8:34:34, 30.94s/it, vocab_size=5490883, loss=8.57]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 2 VocabSize: 5490883 Loss: 8.5747\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 3/1000 [01:33<8:38:04, 31.18s/it, vocab_size=4936841, loss=8.57]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 3 VocabSize: 4936841 Loss: 8.5652\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 4/1000 [02:04<8:36:38, 31.12s/it, vocab_size=4442669, loss=8.56]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 4 VocabSize: 4442669 Loss: 8.5562\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 5/1000 [02:34<8:33:44, 30.98s/it, vocab_size=4e+6, loss=8.55]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 5 VocabSize: 3999138 Loss: 8.5454\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 6/1000 [03:06<8:35:08, 31.10s/it, vocab_size=3.6e+6, loss=8.54]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 6 VocabSize: 3601074 Loss: 8.5361\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 7/1000 [03:37<8:36:45, 31.22s/it, vocab_size=3242848, loss=8.53]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 7 VocabSize: 3242848 Loss: 8.5255\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 8/1000 [04:08<8:34:29, 31.12s/it, vocab_size=2920929, loss=8.52]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 8 VocabSize: 2920929 Loss: 8.5165\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 9/1000 [04:39<8:29:46, 30.86s/it, vocab_size=2631673, loss=8.51]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 9 VocabSize: 2631673 Loss: 8.5080\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 10/1000 [05:09<8:29:04, 30.85s/it, vocab_size=2371448, loss=8.5] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 10 VocabSize: 2371448 Loss: 8.5000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 11/1000 [05:39<8:24:39, 30.62s/it, vocab_size=2137257, loss=8.49]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 11 VocabSize: 2137257 Loss: 8.4915\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 12/1000 [06:10<8:24:42, 30.65s/it, vocab_size=1926056, loss=8.48]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 12 VocabSize: 1926056 Loss: 8.4824\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|▏         | 13/1000 [06:40<8:20:17, 30.41s/it, vocab_size=1735912, loss=8.47]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 13 VocabSize: 1735912 Loss: 8.4731\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|▏         | 14/1000 [07:10<8:17:06, 30.25s/it, vocab_size=1565023, loss=8.46]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 14 VocabSize: 1565023 Loss: 8.4637\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 15/1000 [07:39<8:11:16, 29.93s/it, vocab_size=1411498, loss=8.45]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 15 VocabSize: 1411498 Loss: 8.4540\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 16/1000 [08:10<8:13:22, 30.08s/it, vocab_size=1273485, loss=8.44]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 16 VocabSize: 1273485 Loss: 8.4440\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 17/1000 [08:40<8:13:11, 30.10s/it, vocab_size=1149786, loss=8.43]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 17 VocabSize: 1149786 Loss: 8.4338\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 18/1000 [09:09<8:11:01, 30.00s/it, vocab_size=1038390, loss=8.42]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 18 VocabSize: 1038390 Loss: 8.4233\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 19/1000 [09:41<8:19:48, 30.57s/it, vocab_size=938666, loss=8.41] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 19 VocabSize: 938666 Loss: 8.4127\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 20/1000 [10:11<8:15:29, 30.34s/it, vocab_size=849321, loss=8.4] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 20 VocabSize: 849321 Loss: 8.4022\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 21/1000 [10:40<8:06:45, 29.83s/it, vocab_size=769092, loss=8.39]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 21 VocabSize: 769092 Loss: 8.3914\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 22/1000 [11:09<8:01:58, 29.57s/it, vocab_size=697188, loss=8.38]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 22 VocabSize: 697188 Loss: 8.3805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 23/1000 [11:39<8:04:54, 29.78s/it, vocab_size=632514, loss=8.37]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 23 VocabSize: 632514 Loss: 8.3695\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 24/1000 [12:09<8:07:44, 29.98s/it, vocab_size=575185, loss=8.36]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 24 VocabSize: 575185 Loss: 8.3583\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▎         | 25/1000 [12:40<8:12:00, 30.28s/it, vocab_size=523651, loss=8.35]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 25 VocabSize: 523651 Loss: 8.3471\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 26/1000 [13:11<8:15:27, 30.52s/it, vocab_size=477575, loss=8.34]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 26 VocabSize: 477575 Loss: 8.3359\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 27/1000 [13:43<8:17:53, 30.70s/it, vocab_size=436805, loss=8.32]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 27 VocabSize: 436805 Loss: 8.3247\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 28/1000 [14:14<8:22:09, 31.00s/it, vocab_size=400578, loss=8.31]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 28 VocabSize: 400578 Loss: 8.3135\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 29/1000 [14:45<8:22:08, 31.03s/it, vocab_size=368006, loss=8.3] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 29 VocabSize: 368006 Loss: 8.3023\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 30/1000 [15:18<8:31:01, 31.61s/it, vocab_size=339151, loss=8.29]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 30 VocabSize: 339151 Loss: 8.2912\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 31/1000 [15:52<8:38:40, 32.12s/it, vocab_size=313415, loss=8.28]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 31 VocabSize: 313415 Loss: 8.2800\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 32/1000 [16:25<8:44:12, 32.49s/it, vocab_size=290430, loss=8.27]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 32 VocabSize: 290430 Loss: 8.2690\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 33/1000 [16:58<8:44:47, 32.56s/it, vocab_size=270412, loss=8.26]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 33 VocabSize: 270412 Loss: 8.2581\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 34/1000 [17:30<8:44:14, 32.56s/it, vocab_size=252266, loss=8.25]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 34 VocabSize: 252266 Loss: 8.2477\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 35/1000 [18:04<8:47:13, 32.78s/it, vocab_size=236045, loss=8.24]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 35 VocabSize: 236045 Loss: 8.2372\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 36/1000 [18:36<8:42:21, 32.51s/it, vocab_size=221704, loss=8.23]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 36 VocabSize: 221704 Loss: 8.2270\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 37/1000 [19:09<8:44:13, 32.66s/it, vocab_size=208867, loss=8.22]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 37 VocabSize: 208867 Loss: 8.2172\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 38/1000 [19:40<8:35:59, 32.18s/it, vocab_size=197542, loss=8.21]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 38 VocabSize: 197542 Loss: 8.2077\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 39/1000 [20:11<8:31:02, 31.91s/it, vocab_size=186980, loss=8.2] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 39 VocabSize: 186980 Loss: 8.1984\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 40/1000 [20:43<8:32:13, 32.01s/it, vocab_size=177863, loss=8.19]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 40 VocabSize: 177863 Loss: 8.1891\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 41/1000 [21:14<8:27:15, 31.74s/it, vocab_size=169290, loss=8.18]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 41 VocabSize: 169290 Loss: 8.1805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 42/1000 [21:44<8:15:14, 31.02s/it, vocab_size=161743, loss=8.17]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 42 VocabSize: 161743 Loss: 8.1719\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 43/1000 [22:12<8:00:15, 30.11s/it, vocab_size=154592, loss=8.16]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 43 VocabSize: 154592 Loss: 8.1637\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 44/1000 [22:40<7:50:01, 29.50s/it, vocab_size=148047, loss=8.16]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 44 VocabSize: 148047 Loss: 8.1555\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 45/1000 [23:12<8:01:49, 30.27s/it, vocab_size=142164, loss=8.15]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 45 VocabSize: 142164 Loss: 8.1475\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 46/1000 [23:40<7:49:43, 29.54s/it, vocab_size=136511, loss=8.14]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 46 VocabSize: 136511 Loss: 8.1398\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 47/1000 [24:08<7:42:50, 29.14s/it, vocab_size=131997, loss=8.13]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 47 VocabSize: 131997 Loss: 8.1319\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 48/1000 [24:39<7:51:01, 29.69s/it, vocab_size=126529, loss=8.13]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 48 VocabSize: 126529 Loss: 8.1252\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 49/1000 [25:13<8:11:37, 31.02s/it, vocab_size=123560, loss=8.12]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 49 VocabSize: 123560 Loss: 8.1167\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 50/1000 [25:45<8:18:25, 31.48s/it, vocab_size=117321, loss=8.11]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 50 VocabSize: 117321 Loss: 8.1118\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 51/1000 [26:13<7:59:58, 30.35s/it, vocab_size=117166, loss=8.1] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 51 VocabSize: 117166 Loss: 8.1009\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 52/1000 [26:43<7:56:33, 30.16s/it, vocab_size=106803, loss=8.1]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 52 VocabSize: 106803 Loss: 8.1005\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 53/1000 [27:11<7:44:31, 29.43s/it, vocab_size=117570, loss=8.08]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 53 VocabSize: 117570 Loss: 8.0804\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 54/1000 [27:39<7:39:45, 29.16s/it, vocab_size=106669, loss=8.08]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 54 VocabSize: 106669 Loss: 8.0805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 55/1000 [28:08<7:38:16, 29.10s/it, vocab_size=97966, loss=8.08] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 55 VocabSize: 97966 Loss: 8.0797\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 56/1000 [28:35<7:28:40, 28.52s/it, vocab_size=108795, loss=8.06]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 56 VocabSize: 108795 Loss: 8.0606\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 57/1000 [29:04<7:28:24, 28.53s/it, vocab_size=98727, loss=8.06] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 57 VocabSize: 98727 Loss: 8.0607\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 58/1000 [29:31<7:20:04, 28.03s/it, vocab_size=89706, loss=8.06]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 58 VocabSize: 89706 Loss: 8.0602\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 59/1000 [30:01<7:30:23, 28.72s/it, vocab_size=103695, loss=8.04]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 59 VocabSize: 103695 Loss: 8.0392\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 60/1000 [30:28<7:21:16, 28.17s/it, vocab_size=94232, loss=8.04] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 60 VocabSize: 94232 Loss: 8.0394\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 61/1000 [30:55<7:15:29, 27.83s/it, vocab_size=85530, loss=8.04]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 61 VocabSize: 85530 Loss: 8.0390\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 62/1000 [31:22<7:11:23, 27.59s/it, vocab_size=91179, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 62 VocabSize: 91179 Loss: 8.0271\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 63/1000 [31:49<7:07:20, 27.36s/it, vocab_size=82676, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 63 VocabSize: 82676 Loss: 8.0271\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 64/1000 [32:16<7:06:17, 27.33s/it, vocab_size=85217, loss=8.02]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 64 VocabSize: 85217 Loss: 8.0184\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 65/1000 [32:45<7:15:32, 27.95s/it, vocab_size=77338, loss=8.02]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 65 VocabSize: 77338 Loss: 8.0183\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 66/1000 [33:13<7:15:17, 27.96s/it, vocab_size=89207, loss=8]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 66 VocabSize: 89207 Loss: 8.0008\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 67/1000 [33:42<7:16:30, 28.07s/it, vocab_size=80986, loss=8]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 67 VocabSize: 80986 Loss: 8.0010\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 68/1000 [34:10<7:15:50, 28.06s/it, vocab_size=73432, loss=8]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 68 VocabSize: 73432 Loss: 8.0006\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 69/1000 [34:37<7:13:10, 27.92s/it, vocab_size=81851, loss=7.99]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 69 VocabSize: 81851 Loss: 7.9865\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 70/1000 [35:06<7:15:10, 28.08s/it, vocab_size=74256, loss=7.99]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 70 VocabSize: 74256 Loss: 7.9866\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 71/1000 [35:34<7:17:37, 28.26s/it, vocab_size=67407, loss=7.99]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 71 VocabSize: 67407 Loss: 7.9862\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 72/1000 [36:04<7:24:44, 28.76s/it, vocab_size=88427, loss=7.96]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 72 VocabSize: 88427 Loss: 7.9627\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 73/1000 [36:35<7:32:08, 29.27s/it, vocab_size=80326, loss=7.96]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 73 VocabSize: 80326 Loss: 7.9631\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 74/1000 [37:05<7:35:24, 29.51s/it, vocab_size=72726, loss=7.96]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 74 VocabSize: 72726 Loss: 7.9628\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 75/1000 [37:34<7:34:47, 29.50s/it, vocab_size=65960, loss=7.96]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 75 VocabSize: 65960 Loss: 7.9624\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 76/1000 [38:04<7:33:31, 29.45s/it, vocab_size=67812, loss=7.96]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 76 VocabSize: 67812 Loss: 7.9561\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 77/1000 [38:32<7:28:41, 29.17s/it, vocab_size=61549, loss=7.96]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 77 VocabSize: 61549 Loss: 7.9560\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 78/1000 [39:01<7:25:37, 29.00s/it, vocab_size=77492, loss=7.94]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 78 VocabSize: 77492 Loss: 7.9353\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 79/1000 [39:29<7:22:30, 28.83s/it, vocab_size=70432, loss=7.94]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 79 VocabSize: 70432 Loss: 7.9357\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 80/1000 [39:57<7:19:25, 28.66s/it, vocab_size=63785, loss=7.94]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 80 VocabSize: 63785 Loss: 7.9354\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 81/1000 [40:26<7:19:21, 28.68s/it, vocab_size=57927, loss=7.93]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 81 VocabSize: 57927 Loss: 7.9350\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 82/1000 [40:57<7:26:46, 29.20s/it, vocab_size=73502, loss=7.92]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 82 VocabSize: 73502 Loss: 7.9160\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 83/1000 [41:27<7:30:07, 29.45s/it, vocab_size=66637, loss=7.92]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 83 VocabSize: 66637 Loss: 7.9163\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 84/1000 [41:57<7:33:06, 29.68s/it, vocab_size=60320, loss=7.92]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 84 VocabSize: 60320 Loss: 7.9161\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 85/1000 [42:28<7:37:14, 29.98s/it, vocab_size=54745, loss=7.92]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 85 VocabSize: 54745 Loss: 7.9157\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▊         | 86/1000 [42:59<7:41:42, 30.31s/it, vocab_size=71520, loss=7.9] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 86 VocabSize: 71520 Loss: 7.8973\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▊         | 87/1000 [43:29<7:43:35, 30.47s/it, vocab_size=64886, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 87 VocabSize: 64886 Loss: 7.8976\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 88/1000 [44:01<7:46:14, 30.67s/it, vocab_size=58736, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 88 VocabSize: 58736 Loss: 7.8974\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 89/1000 [44:31<7:44:28, 30.59s/it, vocab_size=53251, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 89 VocabSize: 53251 Loss: 7.8970\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 90/1000 [45:02<7:45:42, 30.71s/it, vocab_size=58800, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 90 VocabSize: 58800 Loss: 7.8870\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 91/1000 [45:31<7:36:52, 30.16s/it, vocab_size=53251, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 91 VocabSize: 53251 Loss: 7.8870\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 92/1000 [46:01<7:38:19, 30.29s/it, vocab_size=48434, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 92 VocabSize: 48434 Loss: 7.8866\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 93/1000 [46:31<7:36:20, 30.19s/it, vocab_size=77175, loss=7.85]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 93 VocabSize: 77175 Loss: 7.8533\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 94/1000 [47:01<7:31:02, 29.87s/it, vocab_size=7e+4, loss=7.85] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 94 VocabSize: 70049 Loss: 7.8541\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|▉         | 95/1000 [47:31<7:31:32, 29.94s/it, vocab_size=63312, loss=7.85]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 95 VocabSize: 63312 Loss: 7.8538\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|▉         | 96/1000 [47:58<7:20:55, 29.26s/it, vocab_size=57251, loss=7.85]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 96 VocabSize: 57251 Loss: 7.8536\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|▉         | 97/1000 [48:26<7:13:02, 28.77s/it, vocab_size=51807, loss=7.85]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 97 VocabSize: 51807 Loss: 7.8532\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|▉         | 98/1000 [48:54<7:07:24, 28.43s/it, vocab_size=47021, loss=7.85]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 98 VocabSize: 47021 Loss: 7.8527\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|▉         | 99/1000 [49:21<7:03:53, 28.23s/it, vocab_size=53171, loss=7.84]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 99 VocabSize: 53171 Loss: 7.8413\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|█         | 100/1000 [49:49<7:01:06, 28.07s/it, vocab_size=48173, loss=7.84]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 100 VocabSize: 48173 Loss: 7.8415\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|█         | 101/1000 [50:17<6:59:15, 27.98s/it, vocab_size=43623, loss=7.84]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 101 VocabSize: 43623 Loss: 7.8411\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|█         | 102/1000 [50:45<6:57:40, 27.91s/it, vocab_size=67110, loss=7.81]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 102 VocabSize: 67110 Loss: 7.8111\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|█         | 103/1000 [51:13<6:57:49, 27.95s/it, vocab_size=60947, loss=7.81]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 103 VocabSize: 60947 Loss: 7.8118\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|█         | 104/1000 [51:40<6:55:52, 27.85s/it, vocab_size=55075, loss=7.81]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 104 VocabSize: 55075 Loss: 7.8115\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|█         | 105/1000 [52:08<6:53:39, 27.73s/it, vocab_size=49759, loss=7.81]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 105 VocabSize: 49759 Loss: 7.8112\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 106/1000 [52:35<6:50:48, 27.57s/it, vocab_size=45031, loss=7.81]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 106 VocabSize: 45031 Loss: 7.8108\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 107/1000 [53:02<6:49:34, 27.52s/it, vocab_size=40899, loss=7.81]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 107 VocabSize: 40899 Loss: 7.8103\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 108/1000 [53:29<6:46:44, 27.36s/it, vocab_size=63133, loss=7.78]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 108 VocabSize: 63133 Loss: 7.7832\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 109/1000 [53:56<6:44:43, 27.25s/it, vocab_size=57292, loss=7.78]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 109 VocabSize: 57292 Loss: 7.7839\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 110/1000 [54:24<6:46:14, 27.39s/it, vocab_size=51766, loss=7.78]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 110 VocabSize: 51766 Loss: 7.7837\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 111/1000 [54:52<6:47:06, 27.48s/it, vocab_size=46781, loss=7.78]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 111 VocabSize: 46781 Loss: 7.7834\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█         | 112/1000 [55:19<6:46:17, 27.45s/it, vocab_size=42345, loss=7.78]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 112 VocabSize: 42345 Loss: 7.7831\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█▏        | 113/1000 [55:46<6:45:45, 27.45s/it, vocab_size=38503, loss=7.78]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 113 VocabSize: 38503 Loss: 7.7825\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  11%|█▏        | 114/1000 [56:14<6:44:53, 27.42s/it, vocab_size=63286, loss=7.75]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 114 VocabSize: 63286 Loss: 7.7545\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 115/1000 [56:41<6:43:04, 27.33s/it, vocab_size=57355, loss=7.76]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 115 VocabSize: 57355 Loss: 7.7553\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 116/1000 [57:08<6:40:47, 27.20s/it, vocab_size=51827, loss=7.76]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 116 VocabSize: 51827 Loss: 7.7551\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 117/1000 [57:34<6:36:34, 26.95s/it, vocab_size=46869, loss=7.75]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 117 VocabSize: 46869 Loss: 7.7548\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 118/1000 [58:01<6:34:54, 26.86s/it, vocab_size=42426, loss=7.75]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 118 VocabSize: 42426 Loss: 7.7545\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 119/1000 [58:27<6:32:27, 26.73s/it, vocab_size=38448, loss=7.75]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 119 VocabSize: 38448 Loss: 7.7540\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 120/1000 [58:53<6:29:07, 26.53s/it, vocab_size=36160, loss=7.75]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 120 VocabSize: 36160 Loss: 7.7524\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 121/1000 [59:19<6:24:49, 26.27s/it, vocab_size=63922, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 121 VocabSize: 63922 Loss: 7.7224\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 122/1000 [59:44<6:20:15, 25.99s/it, vocab_size=57880, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 122 VocabSize: 57880 Loss: 7.7233\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 123/1000 [1:00:10<6:20:04, 26.00s/it, vocab_size=52268, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 123 VocabSize: 52268 Loss: 7.7231\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▏        | 124/1000 [1:00:37<6:21:13, 26.11s/it, vocab_size=47202, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 124 VocabSize: 47202 Loss: 7.7228\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  12%|█▎        | 125/1000 [1:01:02<6:18:07, 25.93s/it, vocab_size=42657, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 125 VocabSize: 42657 Loss: 7.7225\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 126/1000 [1:01:29<6:21:13, 26.17s/it, vocab_size=38626, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 126 VocabSize: 38626 Loss: 7.7221\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 127/1000 [1:01:55<6:18:58, 26.05s/it, vocab_size=35047, loss=7.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 127 VocabSize: 35047 Loss: 7.7215\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 128/1000 [1:02:21<6:18:47, 26.06s/it, vocab_size=42308, loss=7.71]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 128 VocabSize: 42308 Loss: 7.7126\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 129/1000 [1:02:47<6:17:57, 26.04s/it, vocab_size=38275, loss=7.71]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 129 VocabSize: 38275 Loss: 7.7128\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 130/1000 [1:03:13<6:16:18, 25.95s/it, vocab_size=34603, loss=7.71]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 130 VocabSize: 34603 Loss: 7.7125\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 131/1000 [1:03:39<6:16:28, 25.99s/it, vocab_size=32249, loss=7.71]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 131 VocabSize: 32249 Loss: 7.7120\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 132/1000 [1:04:05<6:15:58, 25.99s/it, vocab_size=32011, loss=7.71]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 132 VocabSize: 32011 Loss: 7.7119\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 133/1000 [1:04:31<6:16:27, 26.05s/it, vocab_size=32001, loss=7.71]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 133 VocabSize: 32001 Loss: 7.7119\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  13%|█▎        | 134/1000 [1:04:57<6:59:48, 29.09s/it, vocab_size=32000, loss=7.71]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 134 VocabSize: 32000 Loss: 7.7119\n",
      "Finished. Iteration: 134 VocabSize: 32000 Loss: 7.7119\n"
     ]
    }
   ],
   "source": [
    "# 학습\n",
    "# save_path = 토크나이저.json 파일 저장 위치\n",
    "# texts = 학습할 데이터(리스트)\n",
    "# vocab_size = 단어수\n",
    "# config = 학습 옵션\n",
    "tokenizer = SwTokenizer.train(save_path='./241112_vo32000_tokenizer.json', texts=paragraph_list, vocab_size=32000, config=config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Tokenizing: 100%|██████████| 2287985/2287985 [15:54<00:00, 2397.02it/s] \n",
      "Reducing #1:   0%|          | 0/1000 [00:00<?, ?it/s, vocab_size=8557558, loss=8.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 0 VocabSize: 8557558 Loss: 8.7211\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 1/1000 [00:30<8:23:51, 30.26s/it, vocab_size=6155175, loss=8.58]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 1 VocabSize: 6155175 Loss: 8.5833\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 2/1000 [01:00<8:20:37, 30.10s/it, vocab_size=5490866, loss=8.57]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 2 VocabSize: 5490866 Loss: 8.5747\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 3/1000 [01:31<8:27:00, 30.51s/it, vocab_size=4936824, loss=8.57]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 3 VocabSize: 4936824 Loss: 8.5652\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 4/1000 [02:01<8:21:47, 30.23s/it, vocab_size=4442647, loss=8.56]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 4 VocabSize: 4442647 Loss: 8.5562\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 5/1000 [02:34<8:42:38, 31.52s/it, vocab_size=4e+6, loss=8.55]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 5 VocabSize: 3999119 Loss: 8.5454\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 6/1000 [03:05<8:38:14, 31.28s/it, vocab_size=3.6e+6, loss=8.54]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 6 VocabSize: 3601054 Loss: 8.5361\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 7/1000 [03:36<8:37:16, 31.26s/it, vocab_size=3242826, loss=8.53]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 7 VocabSize: 3242826 Loss: 8.5255\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 8/1000 [04:06<8:28:58, 30.79s/it, vocab_size=2920910, loss=8.52]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 8 VocabSize: 2920910 Loss: 8.5165\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 9/1000 [04:36<8:25:13, 30.59s/it, vocab_size=2631656, loss=8.51]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 9 VocabSize: 2631656 Loss: 8.5080\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 10/1000 [05:06<8:21:28, 30.39s/it, vocab_size=2371437, loss=8.5] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 10 VocabSize: 2371437 Loss: 8.5000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 11/1000 [05:36<8:15:57, 30.09s/it, vocab_size=2137254, loss=8.49]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 11 VocabSize: 2137254 Loss: 8.4915\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 12/1000 [06:05<8:10:38, 29.80s/it, vocab_size=1926054, loss=8.48]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 12 VocabSize: 1926054 Loss: 8.4824\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|▏         | 13/1000 [06:41<8:40:54, 31.67s/it, vocab_size=1735905, loss=8.47]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 13 VocabSize: 1735905 Loss: 8.4731\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|▏         | 14/1000 [07:11<8:35:11, 31.35s/it, vocab_size=1565017, loss=8.46]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 14 VocabSize: 1565017 Loss: 8.4636\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 15/1000 [07:43<8:37:40, 31.53s/it, vocab_size=1411491, loss=8.45]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 15 VocabSize: 1411491 Loss: 8.4540\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 16/1000 [08:12<8:24:44, 30.78s/it, vocab_size=1273481, loss=8.44]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 16 VocabSize: 1273481 Loss: 8.4440\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 17/1000 [08:44<8:30:34, 31.16s/it, vocab_size=1149782, loss=8.43]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 17 VocabSize: 1149782 Loss: 8.4338\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 18/1000 [09:15<8:25:11, 30.87s/it, vocab_size=1038377, loss=8.42]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 18 VocabSize: 1038377 Loss: 8.4233\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 19/1000 [09:46<8:26:57, 31.01s/it, vocab_size=938663, loss=8.41] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 19 VocabSize: 938663 Loss: 8.4127\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 20/1000 [10:16<8:22:10, 30.75s/it, vocab_size=849313, loss=8.4] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 20 VocabSize: 849313 Loss: 8.4022\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 21/1000 [10:46<8:17:30, 30.49s/it, vocab_size=769090, loss=8.39]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 21 VocabSize: 769090 Loss: 8.3914\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 22/1000 [11:17<8:20:28, 30.70s/it, vocab_size=697170, loss=8.38]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 22 VocabSize: 697170 Loss: 8.3805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 23/1000 [11:49<8:24:37, 30.99s/it, vocab_size=632504, loss=8.37]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 23 VocabSize: 632504 Loss: 8.3695\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 24/1000 [12:20<8:23:50, 30.97s/it, vocab_size=575178, loss=8.36]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 24 VocabSize: 575178 Loss: 8.3583\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▎         | 25/1000 [12:53<8:34:18, 31.65s/it, vocab_size=523652, loss=8.35]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 25 VocabSize: 523652 Loss: 8.3471\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 26/1000 [13:25<8:33:12, 31.61s/it, vocab_size=477579, loss=8.34]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 26 VocabSize: 477579 Loss: 8.3359\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 27/1000 [13:56<8:33:40, 31.68s/it, vocab_size=436795, loss=8.32]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 27 VocabSize: 436795 Loss: 8.3246\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 28/1000 [14:28<8:33:50, 31.72s/it, vocab_size=400573, loss=8.31]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 28 VocabSize: 400573 Loss: 8.3135\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 29/1000 [15:00<8:35:14, 31.84s/it, vocab_size=368003, loss=8.3] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 29 VocabSize: 368003 Loss: 8.3023\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 30/1000 [15:33<8:36:54, 31.97s/it, vocab_size=339147, loss=8.29]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 30 VocabSize: 339147 Loss: 8.2912\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 31/1000 [16:04<8:34:58, 31.89s/it, vocab_size=313410, loss=8.28]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 31 VocabSize: 313410 Loss: 8.2800\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 32/1000 [16:36<8:33:07, 31.81s/it, vocab_size=290433, loss=8.27]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 32 VocabSize: 290433 Loss: 8.2690\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 33/1000 [17:08<8:33:26, 31.86s/it, vocab_size=270425, loss=8.26]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 33 VocabSize: 270425 Loss: 8.2581\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 34/1000 [17:39<8:29:53, 31.67s/it, vocab_size=252242, loss=8.25]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 34 VocabSize: 252242 Loss: 8.2477\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 35/1000 [18:10<8:27:21, 31.55s/it, vocab_size=236060, loss=8.24]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 35 VocabSize: 236060 Loss: 8.2372\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 36/1000 [18:42<8:27:33, 31.59s/it, vocab_size=221665, loss=8.23]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 36 VocabSize: 221665 Loss: 8.2270\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 37/1000 [19:13<8:26:28, 31.56s/it, vocab_size=208890, loss=8.22]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 37 VocabSize: 208890 Loss: 8.2172\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 38/1000 [19:45<8:27:59, 31.68s/it, vocab_size=197554, loss=8.21]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 38 VocabSize: 197554 Loss: 8.2077\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 39/1000 [20:17<8:25:43, 31.57s/it, vocab_size=187026, loss=8.2] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 39 VocabSize: 187026 Loss: 8.1984\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 40/1000 [20:47<8:20:07, 31.26s/it, vocab_size=177905, loss=8.19]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 40 VocabSize: 177905 Loss: 8.1892\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 41/1000 [21:18<8:19:01, 31.22s/it, vocab_size=169349, loss=8.18]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 41 VocabSize: 169349 Loss: 8.1805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 42/1000 [21:50<8:18:57, 31.25s/it, vocab_size=161802, loss=8.17]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 42 VocabSize: 161802 Loss: 8.1719\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 43/1000 [22:21<8:17:59, 31.22s/it, vocab_size=154539, loss=8.16]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 43 VocabSize: 154539 Loss: 8.1637\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 44/1000 [22:52<8:16:03, 31.13s/it, vocab_size=148161, loss=8.16]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 44 VocabSize: 148161 Loss: 8.1554\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 45/1000 [23:23<8:16:06, 31.17s/it, vocab_size=142040, loss=8.15]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 45 VocabSize: 142040 Loss: 8.1476\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 46/1000 [23:54<8:14:33, 31.10s/it, vocab_size=136756, loss=8.14]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 46 VocabSize: 136756 Loss: 8.1396\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 47/1000 [24:25<8:11:59, 30.98s/it, vocab_size=131640, loss=8.13]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 47 VocabSize: 131640 Loss: 8.1322\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 48/1000 [24:55<8:08:47, 30.81s/it, vocab_size=127118, loss=8.12]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 48 VocabSize: 127118 Loss: 8.1246\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 49/1000 [25:26<8:06:48, 30.71s/it, vocab_size=122552, loss=8.12]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 49 VocabSize: 122552 Loss: 8.1176\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 50/1000 [25:56<8:04:47, 30.62s/it, vocab_size=119070, loss=8.11]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 50 VocabSize: 119070 Loss: 8.1100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 51/1000 [26:26<8:02:31, 30.51s/it, vocab_size=114002, loss=8.1] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 51 VocabSize: 114002 Loss: 8.1040\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 52/1000 [26:57<8:04:02, 30.64s/it, vocab_size=112609, loss=8.09]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 52 VocabSize: 112609 Loss: 8.0949\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 53/1000 [27:28<8:04:09, 30.67s/it, vocab_size=105173, loss=8.09]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 53 VocabSize: 105173 Loss: 8.0921\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 54/1000 [27:59<8:05:00, 30.76s/it, vocab_size=111178, loss=8.08]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 54 VocabSize: 111178 Loss: 8.0770\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 55/1000 [28:29<8:00:26, 30.50s/it, vocab_size=100867, loss=8.08]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 55 VocabSize: 100867 Loss: 8.0770\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 56/1000 [28:59<7:57:56, 30.38s/it, vocab_size=101606, loss=8.07]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 56 VocabSize: 101606 Loss: 8.0672\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 57/1000 [29:29<7:55:48, 30.27s/it, vocab_size=92271, loss=8.07] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 57 VocabSize: 92271 Loss: 8.0670\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 58/1000 [29:59<7:53:21, 30.15s/it, vocab_size=105479, loss=8.05]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 58 VocabSize: 105479 Loss: 8.0460\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 59/1000 [30:29<7:51:44, 30.08s/it, vocab_size=95775, loss=8.05] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 59 VocabSize: 95775 Loss: 8.0461\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 60/1000 [30:59<7:50:58, 30.06s/it, vocab_size=86893, loss=8.05]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 60 VocabSize: 86893 Loss: 8.0457\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 61/1000 [31:29<7:51:15, 30.11s/it, vocab_size=95176, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 61 VocabSize: 95176 Loss: 8.0311\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 62/1000 [31:59<7:49:39, 30.04s/it, vocab_size=86362, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 62 VocabSize: 86362 Loss: 8.0312\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 63/1000 [32:29<7:47:25, 29.93s/it, vocab_size=80672, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 63 VocabSize: 80672 Loss: 8.0291\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 64/1000 [32:59<7:47:25, 29.96s/it, vocab_size=92413, loss=8.01]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 64 VocabSize: 92413 Loss: 8.0122\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 65/1000 [33:28<7:45:39, 29.88s/it, vocab_size=83869, loss=8.01]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 65 VocabSize: 83869 Loss: 8.0123\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 66/1000 [33:58<7:43:19, 29.76s/it, vocab_size=76125, loss=8.01]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 66 VocabSize: 76125 Loss: 8.0119\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 67/1000 [34:26<7:36:40, 29.37s/it, vocab_size=85978, loss=8]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 67 VocabSize: 85978 Loss: 7.9964\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 68/1000 [34:55<7:33:54, 29.22s/it, vocab_size=78016, loss=8]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 68 VocabSize: 78016 Loss: 7.9966\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 69/1000 [35:24<7:29:49, 28.99s/it, vocab_size=70741, loss=8]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 69 VocabSize: 70741 Loss: 7.9962\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 70/1000 [35:52<7:28:39, 28.95s/it, vocab_size=86559, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 70 VocabSize: 86559 Loss: 7.9762\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 71/1000 [36:24<7:39:30, 29.68s/it, vocab_size=78586, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 71 VocabSize: 78586 Loss: 7.9765\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 72/1000 [36:54<7:39:22, 29.70s/it, vocab_size=71135, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 72 VocabSize: 71135 Loss: 7.9762\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 73/1000 [37:25<7:46:13, 30.18s/it, vocab_size=64737, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 73 VocabSize: 64737 Loss: 7.9756\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 74/1000 [37:55<7:44:39, 30.11s/it, vocab_size=87574, loss=7.95]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 74 VocabSize: 87574 Loss: 7.9510\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 75/1000 [38:27<7:51:50, 30.61s/it, vocab_size=79544, loss=7.95]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 75 VocabSize: 79544 Loss: 7.9514\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 76/1000 [38:56<7:47:45, 30.37s/it, vocab_size=72015, loss=7.95]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 76 VocabSize: 72015 Loss: 7.9512\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 77/1000 [39:26<7:44:53, 30.22s/it, vocab_size=65291, loss=7.95]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 77 VocabSize: 65291 Loss: 7.9508\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 78/1000 [39:57<7:47:11, 30.40s/it, vocab_size=59637, loss=7.95]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 78 VocabSize: 59637 Loss: 7.9502\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 79/1000 [40:28<7:47:26, 30.45s/it, vocab_size=81011, loss=7.93]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 79 VocabSize: 81011 Loss: 7.9256\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 80/1000 [40:59<7:52:13, 30.80s/it, vocab_size=73599, loss=7.93]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 80 VocabSize: 73599 Loss: 7.9261\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 81/1000 [41:31<7:57:17, 31.16s/it, vocab_size=66627, loss=7.93]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 81 VocabSize: 66627 Loss: 7.9258\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 82/1000 [42:00<7:43:55, 30.32s/it, vocab_size=60381, loss=7.93]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 82 VocabSize: 60381 Loss: 7.9255\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 83/1000 [42:28<7:34:54, 29.77s/it, vocab_size=55054, loss=7.92]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 83 VocabSize: 55054 Loss: 7.9249\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 84/1000 [42:56<7:27:30, 29.31s/it, vocab_size=79382, loss=7.9] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 84 VocabSize: 79382 Loss: 7.8992\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 85/1000 [43:25<7:22:54, 29.04s/it, vocab_size=72016, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 85 VocabSize: 72016 Loss: 7.8997\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▊         | 86/1000 [43:54<7:24:34, 29.18s/it, vocab_size=65130, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 86 VocabSize: 65130 Loss: 7.8995\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▊         | 87/1000 [44:22<7:18:33, 28.82s/it, vocab_size=58978, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 87 VocabSize: 58978 Loss: 7.8991\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 88/1000 [44:50<7:15:22, 28.64s/it, vocab_size=53544, loss=7.9]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 88 VocabSize: 53544 Loss: 7.8987\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 89/1000 [45:19<7:12:28, 28.48s/it, vocab_size=59853, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 89 VocabSize: 59853 Loss: 7.8889\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 90/1000 [45:47<7:11:28, 28.45s/it, vocab_size=54227, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 90 VocabSize: 54227 Loss: 7.8890\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 91/1000 [46:15<7:09:37, 28.36s/it, vocab_size=49226, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 91 VocabSize: 49226 Loss: 7.8886\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 92/1000 [46:43<7:06:54, 28.21s/it, vocab_size=5e+4, loss=7.89] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 92 VocabSize: 49984 Loss: 7.8870\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 93/1000 [47:11<7:05:59, 28.18s/it, vocab_size=48111, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 93 VocabSize: 48111 Loss: 7.8869\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   9%|▉         | 94/1000 [47:39<7:03:02, 28.02s/it, vocab_size=48006, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 94 VocabSize: 48006 Loss: 7.8869\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:  10%|▉         | 95/1000 [48:07<7:38:22, 30.39s/it, vocab_size=48000, loss=7.89]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 95 VocabSize: 48000 Loss: 7.8869\n",
      "Finished. Iteration: 95 VocabSize: 48000 Loss: 7.8869\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "tokenizer = SwTokenizer.train(save_path='./241112_vo48000_tokenizer.json', texts=paragraph_list, vocab_size=48000, config=config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Tokenizing: 100%|██████████| 2287985/2287985 [15:49<00:00, 2409.30it/s] \n",
      "Reducing #1:   0%|          | 0/1000 [00:00<?, ?it/s, vocab_size=8557558, loss=8.72]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 0 VocabSize: 8557558 Loss: 8.7211\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 1/1000 [00:36<10:01:57, 36.15s/it, vocab_size=6155165, loss=8.58]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 1 VocabSize: 6155165 Loss: 8.5833\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 2/1000 [01:07<9:13:07, 33.25s/it, vocab_size=5490862, loss=8.57] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 2 VocabSize: 5490862 Loss: 8.5747\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 3/1000 [01:40<9:10:03, 33.10s/it, vocab_size=4936820, loss=8.57]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 3 VocabSize: 4936820 Loss: 8.5652\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 4/1000 [02:11<8:57:59, 32.41s/it, vocab_size=4442649, loss=8.56]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 4 VocabSize: 4442649 Loss: 8.5562\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   0%|          | 5/1000 [02:44<8:57:16, 32.40s/it, vocab_size=4e+6, loss=8.55]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 5 VocabSize: 3999122 Loss: 8.5454\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 6/1000 [03:15<8:51:10, 32.06s/it, vocab_size=3.6e+6, loss=8.54]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 6 VocabSize: 3601052 Loss: 8.5361\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 7/1000 [03:46<8:47:10, 31.85s/it, vocab_size=3242822, loss=8.53]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 7 VocabSize: 3242822 Loss: 8.5255\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 8/1000 [04:18<8:44:26, 31.72s/it, vocab_size=2920905, loss=8.52]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 8 VocabSize: 2920905 Loss: 8.5165\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 9/1000 [04:50<8:46:57, 31.90s/it, vocab_size=2631653, loss=8.51]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 9 VocabSize: 2631653 Loss: 8.5080\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 10/1000 [05:21<8:42:14, 31.65s/it, vocab_size=2371431, loss=8.5] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 10 VocabSize: 2371431 Loss: 8.5000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 11/1000 [05:57<9:00:52, 32.81s/it, vocab_size=2137248, loss=8.49]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 11 VocabSize: 2137248 Loss: 8.4915\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|          | 12/1000 [06:30<9:00:52, 32.85s/it, vocab_size=1926047, loss=8.48]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 12 VocabSize: 1926047 Loss: 8.4824\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|▏         | 13/1000 [07:01<8:53:33, 32.44s/it, vocab_size=1735902, loss=8.47]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 13 VocabSize: 1735902 Loss: 8.4731\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   1%|▏         | 14/1000 [07:33<8:48:12, 32.14s/it, vocab_size=1565015, loss=8.46]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 14 VocabSize: 1565015 Loss: 8.4636\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 15/1000 [08:06<8:52:03, 32.41s/it, vocab_size=1411489, loss=8.45]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 15 VocabSize: 1411489 Loss: 8.4540\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 16/1000 [08:41<9:04:55, 33.23s/it, vocab_size=1273476, loss=8.44]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 16 VocabSize: 1273476 Loss: 8.4440\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 17/1000 [09:15<9:10:26, 33.60s/it, vocab_size=1149781, loss=8.43]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 17 VocabSize: 1149781 Loss: 8.4338\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 18/1000 [09:51<9:19:04, 34.16s/it, vocab_size=1038374, loss=8.42]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 18 VocabSize: 1038374 Loss: 8.4233\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 19/1000 [10:25<9:21:11, 34.32s/it, vocab_size=938653, loss=8.41] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 19 VocabSize: 938653 Loss: 8.4127\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 20/1000 [10:58<9:11:26, 33.76s/it, vocab_size=849310, loss=8.4] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 20 VocabSize: 849310 Loss: 8.4022\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 21/1000 [11:34<9:24:24, 34.59s/it, vocab_size=769089, loss=8.39]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 21 VocabSize: 769089 Loss: 8.3914\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 22/1000 [12:11<9:33:25, 35.18s/it, vocab_size=697168, loss=8.38]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 22 VocabSize: 697168 Loss: 8.3805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 23/1000 [12:48<9:41:22, 35.70s/it, vocab_size=632507, loss=8.37]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 23 VocabSize: 632507 Loss: 8.3695\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▏         | 24/1000 [13:19<9:20:26, 34.45s/it, vocab_size=575179, loss=8.36]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 24 VocabSize: 575179 Loss: 8.3583\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   2%|▎         | 25/1000 [13:51<9:05:02, 33.54s/it, vocab_size=523656, loss=8.35]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 25 VocabSize: 523656 Loss: 8.3471\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 26/1000 [14:22<8:54:45, 32.94s/it, vocab_size=477581, loss=8.34]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 26 VocabSize: 477581 Loss: 8.3359\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 27/1000 [14:54<8:46:28, 32.46s/it, vocab_size=436798, loss=8.32]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 27 VocabSize: 436798 Loss: 8.3246\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 28/1000 [15:26<8:46:45, 32.52s/it, vocab_size=400582, loss=8.31]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 28 VocabSize: 400582 Loss: 8.3135\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 29/1000 [15:58<8:42:51, 32.31s/it, vocab_size=368001, loss=8.3] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 29 VocabSize: 368001 Loss: 8.3023\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 30/1000 [16:31<8:47:00, 32.60s/it, vocab_size=339135, loss=8.29]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 30 VocabSize: 339135 Loss: 8.2912\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 31/1000 [17:05<8:50:26, 32.84s/it, vocab_size=313412, loss=8.28]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 31 VocabSize: 313412 Loss: 8.2800\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 32/1000 [17:37<8:47:12, 32.68s/it, vocab_size=290421, loss=8.27]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 32 VocabSize: 290421 Loss: 8.2690\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 33/1000 [18:07<8:34:21, 31.91s/it, vocab_size=270423, loss=8.26]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 33 VocabSize: 270423 Loss: 8.2581\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   3%|▎         | 34/1000 [18:37<8:25:22, 31.39s/it, vocab_size=252260, loss=8.25]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 34 VocabSize: 252260 Loss: 8.2477\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 35/1000 [19:08<8:19:28, 31.05s/it, vocab_size=236052, loss=8.24]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 35 VocabSize: 236052 Loss: 8.2372\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 36/1000 [19:38<8:14:20, 30.77s/it, vocab_size=221675, loss=8.23]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 36 VocabSize: 221675 Loss: 8.2270\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▎         | 37/1000 [20:08<8:09:44, 30.51s/it, vocab_size=208889, loss=8.22]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 37 VocabSize: 208889 Loss: 8.2172\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 38/1000 [20:37<8:03:35, 30.16s/it, vocab_size=197553, loss=8.21]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 38 VocabSize: 197553 Loss: 8.2077\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 39/1000 [21:06<7:58:33, 29.88s/it, vocab_size=187026, loss=8.2] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 39 VocabSize: 187026 Loss: 8.1984\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 40/1000 [21:36<7:55:37, 29.73s/it, vocab_size=177902, loss=8.19]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 40 VocabSize: 177902 Loss: 8.1892\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 41/1000 [22:05<7:52:53, 29.59s/it, vocab_size=169363, loss=8.18]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 41 VocabSize: 169363 Loss: 8.1805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 42/1000 [22:34<7:50:49, 29.49s/it, vocab_size=161813, loss=8.17]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 42 VocabSize: 161813 Loss: 8.1719\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 43/1000 [23:03<7:48:32, 29.38s/it, vocab_size=154537, loss=8.16]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 43 VocabSize: 154537 Loss: 8.1637\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 44/1000 [23:33<7:49:54, 29.49s/it, vocab_size=148162, loss=8.16]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 44 VocabSize: 148162 Loss: 8.1554\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   4%|▍         | 45/1000 [24:01<7:42:12, 29.04s/it, vocab_size=142053, loss=8.15]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 45 VocabSize: 142053 Loss: 8.1476\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 46/1000 [24:31<7:46:34, 29.34s/it, vocab_size=136757, loss=8.14]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 46 VocabSize: 136757 Loss: 8.1396\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 47/1000 [25:00<7:42:42, 29.13s/it, vocab_size=131651, loss=8.13]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 47 VocabSize: 131651 Loss: 8.1322\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 48/1000 [25:27<7:31:40, 28.47s/it, vocab_size=127110, loss=8.12]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 48 VocabSize: 127110 Loss: 8.1247\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▍         | 49/1000 [25:54<7:24:35, 28.05s/it, vocab_size=122552, loss=8.12]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 49 VocabSize: 122552 Loss: 8.1176\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 50/1000 [26:21<7:19:37, 27.77s/it, vocab_size=119087, loss=8.11]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 50 VocabSize: 119087 Loss: 8.1100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 51/1000 [26:48<7:17:05, 27.63s/it, vocab_size=113963, loss=8.1] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 51 VocabSize: 113963 Loss: 8.1041\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 52/1000 [27:15<7:15:03, 27.53s/it, vocab_size=112728, loss=8.09]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 52 VocabSize: 112728 Loss: 8.0948\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 53/1000 [27:43<7:13:37, 27.47s/it, vocab_size=104864, loss=8.09]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 53 VocabSize: 104864 Loss: 8.0924\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   5%|▌         | 54/1000 [28:09<7:09:44, 27.26s/it, vocab_size=111742, loss=8.08]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 54 VocabSize: 111742 Loss: 8.0763\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 55/1000 [28:37<7:09:02, 27.24s/it, vocab_size=101380, loss=8.08]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 55 VocabSize: 101380 Loss: 8.0763\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 56/1000 [29:03<7:05:25, 27.04s/it, vocab_size=1e+5, loss=8.07]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 56 VocabSize: 100438 Loss: 8.0683\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 57/1000 [29:30<7:03:10, 26.92s/it, vocab_size=93519, loss=8.07]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 57 VocabSize: 93519 Loss: 8.0663\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 58/1000 [29:56<7:00:49, 26.80s/it, vocab_size=101854, loss=8.05]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 58 VocabSize: 101854 Loss: 8.0493\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 59/1000 [30:23<7:00:41, 26.82s/it, vocab_size=92444, loss=8.05] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 59 VocabSize: 92444 Loss: 8.0494\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 60/1000 [30:50<6:59:01, 26.75s/it, vocab_size=87436, loss=8.05]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 60 VocabSize: 87436 Loss: 8.0461\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 61/1000 [31:16<6:57:15, 26.66s/it, vocab_size=93201, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 61 VocabSize: 93201 Loss: 8.0328\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▌         | 62/1000 [31:43<6:55:21, 26.57s/it, vocab_size=84562, loss=8.03]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 62 VocabSize: 84562 Loss: 8.0328\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 63/1000 [32:10<6:58:09, 26.78s/it, vocab_size=86547, loss=8.02]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 63 VocabSize: 86547 Loss: 8.0241\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 64/1000 [32:37<6:57:37, 26.77s/it, vocab_size=78528, loss=8.02]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 64 VocabSize: 78528 Loss: 8.0240\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   6%|▋         | 65/1000 [33:03<6:56:09, 26.71s/it, vocab_size=93181, loss=8.01]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 65 VocabSize: 93181 Loss: 8.0050\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 66/1000 [33:30<6:55:29, 26.69s/it, vocab_size=84620, loss=8.01]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 66 VocabSize: 84620 Loss: 8.0052\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 67/1000 [33:58<7:04:02, 27.27s/it, vocab_size=76688, loss=8]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 67 VocabSize: 76688 Loss: 8.0049\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 68/1000 [34:26<7:06:11, 27.44s/it, vocab_size=77134, loss=8]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 68 VocabSize: 77134 Loss: 7.9983\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 69/1000 [34:56<7:15:09, 28.04s/it, vocab_size=7e+4, loss=8] "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 69 VocabSize: 70005 Loss: 7.9982\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 70/1000 [35:24<7:17:03, 28.20s/it, vocab_size=86754, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 70 VocabSize: 86754 Loss: 7.9776\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 71/1000 [35:51<7:08:17, 27.66s/it, vocab_size=78716, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 71 VocabSize: 78716 Loss: 7.9779\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 72/1000 [36:17<7:02:50, 27.34s/it, vocab_size=71256, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 72 VocabSize: 71256 Loss: 7.9776\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 73/1000 [36:44<6:59:54, 27.18s/it, vocab_size=64864, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 73 VocabSize: 64864 Loss: 7.9770\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   7%|▋         | 74/1000 [37:12<7:02:42, 27.39s/it, vocab_size=64109, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 74 VocabSize: 64109 Loss: 7.9769\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 75/1000 [37:43<7:20:44, 28.59s/it, vocab_size=64004, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 75 VocabSize: 64004 Loss: 7.9769\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 76/1000 [38:12<7:20:07, 28.58s/it, vocab_size=64001, loss=7.98]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 76 VocabSize: 64001 Loss: 7.9769\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Reducing #1:   8%|▊         | 77/1000 [38:41<7:43:49, 30.15s/it, vocab_size=64000, loss=7.98]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 77 VocabSize: 64000 Loss: 7.9769\n",
      "Finished. Iteration: 77 VocabSize: 64000 Loss: 7.9769\n"
     ]
    }
   ],
   "source": [
    "tokenizer = SwTokenizer.train(save_path='./241112_vo64000_tokenizer.json', texts=paragraph_list, vocab_size=64000, config=config)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyM7R1Qlv0w+dP881fM9OVxS",
   "include_colab_link": true,
   "provenance": []
  },
  "kernelspec": {
   "display_name": "python311",
   "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": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}