ktrain_bert_jp_text_ Classification_20200608.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install -U pip\n",
    "!pip install ktrain\n",
    "!pip install mecab-python3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline\n",
    "import os\n",
    "os.environ[\"CUDA_DEVICE_ORDER\"]=\"PCI_BUS_ID\";\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"]=\"0\";   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!wget https://www.rondhuit.com/download/ldcc-20140209.tar.gz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "!tar -zxvf ldcc-20140209.tar.gz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "datapath = Path('text/')\n",
    "topics = [x.parts[-1] for x in datapath.iterdir() if x.is_dir()]\n",
    "topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_files\n",
    "dataset = load_files(datapath, categories=topics, encoding='utf8')\n",
    "dataset.data = [doc.split('\\n', 2)[2] for doc in dataset.data]\n",
    "dataset.data[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X, y = dataset.data, dataset.target\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, \n",
    "                                                    test_size=0.1, random_state=42)\n",
    "len(y_train), len(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ktrain\n",
    "from ktrain import text\n",
    "MODEL_NAME = 'cl-tohoku/bert-base-japanese-whole-word-masking'\n",
    "t = text.Transformer(MODEL_NAME, maxlen=128, classes=topics)\n",
    "trn = t.preprocess_train(X_train, y_train)\n",
    "val = t.preprocess_test(X_test, y_test)\n",
    "model = t.get_classifier()\n",
    "learner = ktrain.get_learner(model, train_data=trn, val_data=val, batch_size=32) # lower bs if OOM occurs\n",
    "#learner.fit_onecycle(5e-5, 3)\n",
    "learner.autofit(5e-5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learner.validate(class_names=t.get_classes())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learner.view_top_losses()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictor = ktrain.get_predictor(learner.model, t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictor.predict(X_test[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictor.save('model/livedoor')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "p = ktrain.load_predictor('model/livedoor')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "p.predict(X_test[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "conda_tensorflow_p36",
   "language": "python",
   "name": "conda_tensorflow_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}