6_1_word_embedding_2_colab.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.7"
    },
    "colab": {
      "name": "6.1 word-embedding-2.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hxf829/Ironcms/blob/master/6_1_word_embedding_2_colab.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HjRqFScenAqO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "263d4857-d9e2-4177-b31b-3c1c6b61dd02"
      },
      "source": [
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "import tensorflow.keras as keras\n",
        "import tensorflow.keras.models as models\n",
        "import tensorflow.keras.layers as layers\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow.keras.optimizers as optimizers\n",
        "from tensorflow.keras.utils import *\n",
        "\n",
        "tf.__version__\n",
        "keras.__version__"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'2.4.0'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ny23RMKqnAqS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.layers import Embedding"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zU4C8MY2nAqV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embedding_layer = Embedding(1000, 64)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sQMngFrEnAqX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "0fbe7a1c-f403-4064-ab10-b27339165c02"
      },
      "source": [
        "import tensorflow.keras.datasets.imdb as imdb\n",
        "from tensorflow.keras import preprocessing\n",
        "max_words = 10000\n",
        "maxlen = 100\n",
        "(x_train, y_train), (x_test, y_test) = imdb.load_data(\n",
        "    num_words=max_words)\n",
        "x_train = preprocessing.sequence.pad_sequences(x_train, maxlen=maxlen)\n",
        "x_test = preprocessing.sequence.pad_sequences(x_test, maxlen=maxlen)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/imdb.npz\n",
            "17465344/17464789 [==============================] - 0s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qFy5OFuVnAqa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "61c5d4a5-f57d-458b-b71d-3081cecdddb7"
      },
      "source": [
        "len(x_train[0])"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "100"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gRTyuNbenAqc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "abdd0734-11a5-4ddb-beb7-c2148c582cb3"
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Flatten, Dense\n",
        "model = Sequential()\n",
        "model.add(Embedding(10000, 32, input_length=maxlen))\n",
        "model.add(Flatten())\n",
        "model.add(Dense(1, activation='sigmoid'))\n",
        "model.compile(optimizer='rmsprop', loss='binary_crossentropy',\n",
        "             metrics=['acc'])\n",
        "model.summary()\n",
        "history = model.fit(x_train, y_train, \n",
        "                    epochs=30, batch_size=32, \n",
        "                    validation_split=0.2)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_1 (Embedding)      (None, 100, 32)           320000    \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 3200)              0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 1)                 3201      \n",
            "=================================================================\n",
            "Total params: 323,201\n",
            "Trainable params: 323,201\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Epoch 1/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.5330 - acc: 0.7430 - val_loss: 0.3709 - val_acc: 0.8366\n",
            "Epoch 2/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.2972 - acc: 0.8769 - val_loss: 0.3285 - val_acc: 0.8562\n",
            "Epoch 3/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.2321 - acc: 0.9069 - val_loss: 0.3310 - val_acc: 0.8572\n",
            "Epoch 4/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.1859 - acc: 0.9292 - val_loss: 0.3390 - val_acc: 0.8544\n",
            "Epoch 5/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.1413 - acc: 0.9513 - val_loss: 0.3562 - val_acc: 0.8556\n",
            "Epoch 6/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0997 - acc: 0.9692 - val_loss: 0.3724 - val_acc: 0.8530\n",
            "Epoch 7/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0646 - acc: 0.9829 - val_loss: 0.3979 - val_acc: 0.8496\n",
            "Epoch 8/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0386 - acc: 0.9924 - val_loss: 0.4321 - val_acc: 0.8416\n",
            "Epoch 9/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0212 - acc: 0.9968 - val_loss: 0.4608 - val_acc: 0.8456\n",
            "Epoch 10/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0108 - acc: 0.9987 - val_loss: 0.4991 - val_acc: 0.8416\n",
            "Epoch 11/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0052 - acc: 0.9995 - val_loss: 0.5443 - val_acc: 0.8408\n",
            "Epoch 12/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0026 - acc: 0.9996 - val_loss: 0.5922 - val_acc: 0.8414\n",
            "Epoch 13/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 0.0012 - acc: 0.9998 - val_loss: 0.6454 - val_acc: 0.8378\n",
            "Epoch 14/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 5.9389e-04 - acc: 0.9999 - val_loss: 0.7026 - val_acc: 0.8378\n",
            "Epoch 15/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 4.1809e-04 - acc: 0.9999 - val_loss: 0.7554 - val_acc: 0.8390\n",
            "Epoch 16/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 1.3953e-04 - acc: 0.9999 - val_loss: 0.8074 - val_acc: 0.8360\n",
            "Epoch 17/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 5.7958e-05 - acc: 1.0000 - val_loss: 0.8616 - val_acc: 0.8346\n",
            "Epoch 18/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 2.5735e-05 - acc: 1.0000 - val_loss: 0.9098 - val_acc: 0.8356\n",
            "Epoch 19/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 3.1217e-05 - acc: 1.0000 - val_loss: 0.9575 - val_acc: 0.8362\n",
            "Epoch 20/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 1.3454e-05 - acc: 1.0000 - val_loss: 0.9875 - val_acc: 0.8370\n",
            "Epoch 21/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 8.1484e-06 - acc: 1.0000 - val_loss: 1.0206 - val_acc: 0.8360\n",
            "Epoch 22/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 1.9968e-06 - acc: 1.0000 - val_loss: 1.0459 - val_acc: 0.8362\n",
            "Epoch 23/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 4.9448e-06 - acc: 1.0000 - val_loss: 1.0672 - val_acc: 0.8366\n",
            "Epoch 24/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 4.4374e-07 - acc: 1.0000 - val_loss: 1.0853 - val_acc: 0.8346\n",
            "Epoch 25/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 3.5534e-07 - acc: 1.0000 - val_loss: 1.0990 - val_acc: 0.8356\n",
            "Epoch 26/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 2.1962e-07 - acc: 1.0000 - val_loss: 1.1120 - val_acc: 0.8350\n",
            "Epoch 27/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 1.1887e-07 - acc: 1.0000 - val_loss: 1.1235 - val_acc: 0.8352\n",
            "Epoch 28/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 1.0516e-07 - acc: 1.0000 - val_loss: 1.1345 - val_acc: 0.8350\n",
            "Epoch 29/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 9.1636e-08 - acc: 1.0000 - val_loss: 1.1425 - val_acc: 0.8350\n",
            "Epoch 30/30\n",
            "625/625 [==============================] - 2s 3ms/step - loss: 7.5096e-08 - acc: 1.0000 - val_loss: 1.1500 - val_acc: 0.8352\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DSM0RebOnAqe",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "b33ffcb4-cdca-432e-a943-ad0d8a8316f5"
      },
      "source": [
        "history.history"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'acc': [0.7429500222206116,\n",
              "  0.8768500089645386,\n",
              "  0.9068999886512756,\n",
              "  0.9291999936103821,\n",
              "  0.9513000249862671,\n",
              "  0.9691500067710876,\n",
              "  0.9829000234603882,\n",
              "  0.9923999905586243,\n",
              "  0.9968000054359436,\n",
              "  0.9987499713897705,\n",
              "  0.9995499849319458,\n",
              "  0.9995999932289124,\n",
              "  0.9998000264167786,\n",
              "  0.9998999834060669,\n",
              "  0.9998999834060669,\n",
              "  0.9999499917030334,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0,\n",
              "  1.0],\n",
              " 'loss': [0.5329963564872742,\n",
              "  0.29716625809669495,\n",
              "  0.2321140319108963,\n",
              "  0.1858804076910019,\n",
              "  0.1412935107946396,\n",
              "  0.09970460832118988,\n",
              "  0.06457670032978058,\n",
              "  0.038604214787483215,\n",
              "  0.02120722271502018,\n",
              "  0.010792842134833336,\n",
              "  0.0052018724381923676,\n",
              "  0.0025652276817709208,\n",
              "  0.0012472775997593999,\n",
              "  0.0005938942194916308,\n",
              "  0.000418085721321404,\n",
              "  0.00013952846347820014,\n",
              "  5.795844117528759e-05,\n",
              "  2.573458550614305e-05,\n",
              "  3.121709596598521e-05,\n",
              "  1.3453558494802564e-05,\n",
              "  8.148392225848511e-06,\n",
              "  1.9968406377302017e-06,\n",
              "  4.9447876335761975e-06,\n",
              "  4.437430334292003e-07,\n",
              "  3.5533685149857774e-07,\n",
              "  2.1961957941130095e-07,\n",
              "  1.1886935880056626e-07,\n",
              "  1.0516047410646934e-07,\n",
              "  9.163577630033615e-08,\n",
              "  7.509617461209928e-08],\n",
              " 'val_acc': [0.8366000056266785,\n",
              "  0.8561999797821045,\n",
              "  0.857200026512146,\n",
              "  0.8543999791145325,\n",
              "  0.8555999994277954,\n",
              "  0.8529999852180481,\n",
              "  0.8496000170707703,\n",
              "  0.8416000008583069,\n",
              "  0.8456000089645386,\n",
              "  0.8416000008583069,\n",
              "  0.8407999873161316,\n",
              "  0.8414000272750854,\n",
              "  0.8378000259399414,\n",
              "  0.8378000259399414,\n",
              "  0.8389999866485596,\n",
              "  0.8360000252723694,\n",
              "  0.8345999717712402,\n",
              "  0.8356000185012817,\n",
              "  0.8361999988555908,\n",
              "  0.8370000123977661,\n",
              "  0.8360000252723694,\n",
              "  0.8361999988555908,\n",
              "  0.8366000056266785,\n",
              "  0.8345999717712402,\n",
              "  0.8356000185012817,\n",
              "  0.8349999785423279,\n",
              "  0.8352000117301941,\n",
              "  0.8349999785423279,\n",
              "  0.8349999785423279,\n",
              "  0.8352000117301941],\n",
              " 'val_loss': [0.3709366023540497,\n",
              "  0.3285408318042755,\n",
              "  0.33098304271698,\n",
              "  0.33899587392807007,\n",
              "  0.356233149766922,\n",
              "  0.37241706252098083,\n",
              "  0.39793258905410767,\n",
              "  0.43205714225769043,\n",
              "  0.4607752859592438,\n",
              "  0.4990922510623932,\n",
              "  0.5442782640457153,\n",
              "  0.5921686291694641,\n",
              "  0.6453970074653625,\n",
              "  0.7026268243789673,\n",
              "  0.7553741335868835,\n",
              "  0.8073741793632507,\n",
              "  0.861644446849823,\n",
              "  0.9098478555679321,\n",
              "  0.9574939608573914,\n",
              "  0.9874571561813354,\n",
              "  1.0205860137939453,\n",
              "  1.0458768606185913,\n",
              "  1.0672441720962524,\n",
              "  1.0852863788604736,\n",
              "  1.0990432500839233,\n",
              "  1.1119798421859741,\n",
              "  1.1235020160675049,\n",
              "  1.1345031261444092,\n",
              "  1.1425271034240723,\n",
              "  1.1499966382980347]}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GA43AVubYAvE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "?plt"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MgFzaYMJnAqi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "d9713710-5552-4129-8e9b-fb4ba9c07d31"
      },
      "source": [
        "val_acc = history.history['val_acc']\n",
        "epochs = range(1, len(val_acc)+1)\n",
        "#可使用axis设定坐标轴取值范围\n",
        "#plt.axis\n",
        "plt.plot(epochs, val_acc, 'bo-', color='green', label='val_acc_haha')\n",
        "plt.xlabel('epochs')\n",
        "plt.ylabel('val_acc')\n",
        "plt.legend()\n",
        "plt.title('embeding_test')\n",
        "plt.show()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ozxc3iLZnAqk",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 527
        },
        "outputId": "3af22fc5-5f2d-4588-8838-167e0bcec332"
      },
      "source": [
        "val_acc"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.8366000056266785,\n",
              " 0.8561999797821045,\n",
              " 0.857200026512146,\n",
              " 0.8543999791145325,\n",
              " 0.8555999994277954,\n",
              " 0.8529999852180481,\n",
              " 0.8496000170707703,\n",
              " 0.8416000008583069,\n",
              " 0.8456000089645386,\n",
              " 0.8416000008583069,\n",
              " 0.8407999873161316,\n",
              " 0.8414000272750854,\n",
              " 0.8378000259399414,\n",
              " 0.8378000259399414,\n",
              " 0.8389999866485596,\n",
              " 0.8360000252723694,\n",
              " 0.8345999717712402,\n",
              " 0.8356000185012817,\n",
              " 0.8361999988555908,\n",
              " 0.8370000123977661,\n",
              " 0.8360000252723694,\n",
              " 0.8361999988555908,\n",
              " 0.8366000056266785,\n",
              " 0.8345999717712402,\n",
              " 0.8356000185012817,\n",
              " 0.8349999785423279,\n",
              " 0.8352000117301941,\n",
              " 0.8349999785423279,\n",
              " 0.8349999785423279,\n",
              " 0.8352000117301941]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MkbG9vminFZb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 629
        },
        "outputId": "9bb15338-70d9-4ccc-9031-81b373064e20"
      },
      "source": [
        "!wget http://ai.stanford.edu/\\~amaas/data/sentiment/aclImdb_v1.tar.gz\n",
        "!tar zxf aclImdb_v1.tar.gz\n",
        "!mv aclImdb imdb\n",
        "!wget http://nlp.stanford.edu/data/glove.6B.zip\n",
        "!unzip glove.6B.zip -d glove"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-08-26 15:23:52--  http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz\n",
            "Resolving ai.stanford.edu (ai.stanford.edu)... 171.64.68.10\n",
            "Connecting to ai.stanford.edu (ai.stanford.edu)|171.64.68.10|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 84125825 (80M) [application/x-gzip]\n",
            "Saving to: ‘aclImdb_v1.tar.gz’\n",
            "\n",
            "aclImdb_v1.tar.gz   100%[===================>]  80.23M  33.0MB/s    in 2.4s    \n",
            "\n",
            "2020-08-26 15:23:55 (33.0 MB/s) - ‘aclImdb_v1.tar.gz’ saved [84125825/84125825]\n",
            "\n",
            "--2020-08-26 15:24:02--  http://nlp.stanford.edu/data/glove.6B.zip\n",
            "Resolving nlp.stanford.edu (nlp.stanford.edu)... 171.64.67.140\n",
            "Connecting to nlp.stanford.edu (nlp.stanford.edu)|171.64.67.140|:80... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://nlp.stanford.edu/data/glove.6B.zip [following]\n",
            "--2020-08-26 15:24:02--  https://nlp.stanford.edu/data/glove.6B.zip\n",
            "Connecting to nlp.stanford.edu (nlp.stanford.edu)|171.64.67.140|:443... connected.\n",
            "HTTP request sent, awaiting response... 301 Moved Permanently\n",
            "Location: http://downloads.cs.stanford.edu/nlp/data/glove.6B.zip [following]\n",
            "--2020-08-26 15:24:02--  http://downloads.cs.stanford.edu/nlp/data/glove.6B.zip\n",
            "Resolving downloads.cs.stanford.edu (downloads.cs.stanford.edu)... 171.64.64.22\n",
            "Connecting to downloads.cs.stanford.edu (downloads.cs.stanford.edu)|171.64.64.22|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 862182613 (822M) [application/zip]\n",
            "Saving to: ‘glove.6B.zip’\n",
            "\n",
            "glove.6B.zip        100%[===================>] 822.24M  1.91MB/s    in 6m 28s  \n",
            "\n",
            "2020-08-26 15:30:30 (2.12 MB/s) - ‘glove.6B.zip’ saved [862182613/862182613]\n",
            "\n",
            "Archive:  glove.6B.zip\n",
            "  inflating: glove/glove.6B.50d.txt  \n",
            "  inflating: glove/glove.6B.100d.txt  \n",
            "  inflating: glove/glove.6B.200d.txt  \n",
            "  inflating: glove/glove.6B.300d.txt  \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xYWf-zvhnAqm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os\n",
        "imdb_dir = '/content/imdb'\n",
        "train_dir = os.path.join(imdb_dir, 'train')\n",
        "labels = []\n",
        "texts = []\n",
        "for label_type in ['neg', 'pos']:\n",
        "    dir_name = os.path.join(train_dir, label_type)\n",
        "    for fname in os.listdir(dir_name):\n",
        "        if fname[-4:] == '.txt':\n",
        "            f = open(os.path.join(dir_name, fname))\n",
        "            texts.append(f.read())\n",
        "            f.close()\n",
        "            if label_type == 'neg':\n",
        "                labels.append(0)\n",
        "            else:\n",
        "                labels.append(1)"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "epVTzEopnAqo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 188
        },
        "outputId": "f0dcc896-05e9-4810-a666-00214cff78b2"
      },
      "source": [
        "len(texts)\n",
        "texts[0]"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'What was Franco Zeffirelli thinking? Was Hollywood responsible for this travesty, or can I take comfort in the idea that someone who didn\\'t speak English as a first language just completely missed the point of Charlotte Bronte\\'s classic? I don\\'t think I can improve on a comment I read below, so I\\'ll just paraphrase it: \"Jane Eyre is a great great book, the screenwriter should read it sometime.\" It\\'s true that this movie\\'s two leads were sadly miscast. But pity the actors, because the screenwriter left out all of the best scenes. The dialog that makes you understand the Jane and Rochester have a meeting of minds and a shared sense of fun...deleted from the script. The marriage proposal, the fortune teller...gone. The allusions, half joking, half sincerely felt, to Jane as a fairy sprite from olde England come to rescue Rochester in his despair...eliminated.<br /><br />It is unfortunate that Zeffirelli felt the need to completely rewrite the end of the novel and Jane\\'s interactions with the Rivers family. But it is unforgivable that he has surgically removed the love from one of the best love stories ever written.<br /><br />Do yourself a favor and go find the 1983 (?) mini series with Timothy Dalton.'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eDSsxVy4nAqq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "\n",
        "maxlen = 100 #截取前100个词\n",
        "training_samples = 200 \n",
        "validation_samples = 10000\n",
        "max_words = 20000 #只考虑topx的词汇\n",
        "#初始化token分析器\n",
        "tokenizer = Tokenizer(num_words = max_words)\n",
        "#分析文本数组\n",
        "tokenizer.fit_on_texts(texts)\n",
        "#将文本数组转化为word_index序列\n",
        "sequences = tokenizer.texts_to_sequences(texts)"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tZMnTx7xnAqs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "word_index = tokenizer.word_index\n",
        "data = pad_sequences(sequences, maxlen=maxlen)\n",
        "labels = np.asarray(labels)\n",
        "indices = np.arange(data.shape[0])\n",
        "np.random.shuffle(indices)\n",
        "data = data[indices]\n",
        "labels = labels[indices]\n",
        "\n",
        "x_train = data[:training_samples]\n",
        "y_train = labels[:training_samples]\n",
        "x_test = data[training_samples:training_samples+validation_samples]\n",
        "y_test = labels[training_samples:training_samples+validation_samples]\n"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q_z9Sr1anAqu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "glove_dir='/content/glove'\n",
        "embeddings_index = {}\n",
        "f = open(os.path.join(glove_dir, 'glove.6B.100d.txt'))\n",
        "for line in f:\n",
        "    values = line.split()\n",
        "    word = values[0]\n",
        "    coefs = np.asarray(values[1:], dtype='float32')\n",
        "    embeddings_index[word] = coefs\n",
        "f.close()"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AXSkqVDwnAqw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "4fdca756-b6cf-422e-c2ff-39644fb554e9"
      },
      "source": [
        "len(embeddings_index)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "400000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nfDL83b5nAqy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embedding_dim = 100\n",
        "embedding_matrix = np.zeros((max_words, embedding_dim))\n",
        "for word, i in word_index.items():\n",
        "    embedding_vector = embeddings_index.get(word)\n",
        "    if i < max_words:\n",
        "        if embedding_vector is not None:\n",
        "            embedding_matrix[i] = embedding_vector"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BtFS9ZIGnAq1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "8e7d3dd3-f5f4-4653-ead8-70d9b5012d80"
      },
      "source": [
        "embedding_matrix"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 0.        ,  0.        ,  0.        , ...,  0.        ,\n",
              "         0.        ,  0.        ],\n",
              "       [-0.038194  , -0.24487001,  0.72812003, ..., -0.1459    ,\n",
              "         0.82779998,  0.27061999],\n",
              "       [-0.071953  ,  0.23127   ,  0.023731  , ..., -0.71894997,\n",
              "         0.86894   ,  0.19539   ],\n",
              "       ...,\n",
              "       [-0.44376999,  0.03826   ,  0.20826   , ..., -0.51973999,\n",
              "        -0.87071002,  0.28297001],\n",
              "       [-0.090716  ,  0.39684001, -0.24558   , ..., -0.9386    ,\n",
              "         0.43085   , -0.24685   ],\n",
              "       [ 0.050756  , -0.91547   , -0.54949999, ...,  0.59113997,\n",
              "        -0.32069001,  0.34022999]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4NRnOfBTnAq3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        },
        "outputId": "15af78b3-75ae-4572-80dc-4d7d25d9667d"
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Embedding, Flatten, Dense\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Embedding(max_words, embedding_dim, input_length=maxlen))\n",
        "model.add(Flatten())\n",
        "model.add(Dense(32, activation='selu'))\n",
        "model.add(Dense(1, activation='sigmoid'))\n",
        "model.summary()"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_2 (Embedding)      (None, 100, 100)          2000000   \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 10000)             0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 32)                320032    \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 33        \n",
            "=================================================================\n",
            "Total params: 2,320,065\n",
            "Trainable params: 2,320,065\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DKvWdKfVnAq6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 377
        },
        "outputId": "4fbbb0d8-eda2-44df-8ce9-0dfe5adfb3f6"
      },
      "source": [
        "model.layers[0].set_weights([embedding_matrix])\n",
        "model.layers[0].trainable = False\n",
        "model.compile(optimizer='rmsprop',\n",
        "             loss='binary_crossentropy',\n",
        "             metrics=['acc'])\n",
        "history = model.fit(x_train, y_train, epochs=10, batch_size=10,\n",
        "                    validation_data=(x_test, y_test))"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "20/20 [==============================] - 1s 60ms/step - loss: 1.6109 - acc: 0.5200 - val_loss: 0.6931 - val_acc: 0.5302\n",
            "Epoch 2/10\n",
            "20/20 [==============================] - 1s 55ms/step - loss: 0.6041 - acc: 0.7050 - val_loss: 0.7246 - val_acc: 0.5285\n",
            "Epoch 3/10\n",
            "20/20 [==============================] - 1s 56ms/step - loss: 0.4729 - acc: 0.7800 - val_loss: 0.7829 - val_acc: 0.5355\n",
            "Epoch 4/10\n",
            "20/20 [==============================] - 1s 54ms/step - loss: 0.1993 - acc: 0.9350 - val_loss: 0.8708 - val_acc: 0.5365\n",
            "Epoch 5/10\n",
            "20/20 [==============================] - 1s 55ms/step - loss: 0.0601 - acc: 0.9900 - val_loss: 0.8785 - val_acc: 0.5544\n",
            "Epoch 6/10\n",
            "20/20 [==============================] - 1s 55ms/step - loss: 0.0654 - acc: 0.9650 - val_loss: 4.6627 - val_acc: 0.4996\n",
            "Epoch 7/10\n",
            "20/20 [==============================] - 1s 53ms/step - loss: 0.0852 - acc: 0.9700 - val_loss: 0.9587 - val_acc: 0.5595\n",
            "Epoch 8/10\n",
            "20/20 [==============================] - 1s 54ms/step - loss: 0.2234 - acc: 0.9250 - val_loss: 0.9698 - val_acc: 0.5663\n",
            "Epoch 9/10\n",
            "20/20 [==============================] - 1s 57ms/step - loss: 0.0031 - acc: 1.0000 - val_loss: 1.0090 - val_acc: 0.5677\n",
            "Epoch 10/10\n",
            "20/20 [==============================] - 1s 57ms/step - loss: 0.0017 - acc: 1.0000 - val_loss: 1.0768 - val_acc: 0.5669\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "81_vHSVKnAq8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.save_weights(os.path.join(\n",
        "    '/content/',\n",
        "    'pre_trained_glove_model.h5'))"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UQ3PpB67nAq-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "outputId": "2ae9f08f-85c8-485d-fcbf-6fb84fa132c1"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "acc = history.history['acc']\n",
        "val_acc = history.history['val_acc']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "epochs = range(1, len(acc)+1)\n",
        "plt.plot(epochs, acc, 'bo', label='acc')\n",
        "plt.plot(epochs, val_acc, 'b', label='val acc')\n",
        "plt.title('train result')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FNMGN8s0nArB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 649
        },
        "outputId": "d4e01b34-a0fc-4d2f-b875-66caa80b0869"
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Embedding, Flatten, Dense\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Embedding(max_words, embedding_dim, input_length=maxlen))\n",
        "model.add(Flatten())\n",
        "model.add(Dense(32, activation='relu'))\n",
        "model.add(Dense(1, activation='sigmoid'))\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer='rmsprop',\n",
        "              loss='binary_crossentropy',\n",
        "              metrics=['acc'])\n",
        "history = model.fit(x_train, y_train,\n",
        "                    epochs=10,\n",
        "                    batch_size=32,\n",
        "                    validation_data=(x_test, y_test))"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_3 (Embedding)      (None, 100, 100)          2000000   \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 10000)             0         \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 32)                320032    \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 1)                 33        \n",
            "=================================================================\n",
            "Total params: 2,320,065\n",
            "Trainable params: 2,320,065\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Epoch 1/10\n",
            "7/7 [==============================] - 1s 93ms/step - loss: 0.6950 - acc: 0.4950 - val_loss: 0.6915 - val_acc: 0.5160\n",
            "Epoch 2/10\n",
            "7/7 [==============================] - 1s 79ms/step - loss: 0.5211 - acc: 0.9750 - val_loss: 0.6925 - val_acc: 0.5255\n",
            "Epoch 3/10\n",
            "7/7 [==============================] - 1s 82ms/step - loss: 0.2994 - acc: 0.9950 - val_loss: 0.6954 - val_acc: 0.5181\n",
            "Epoch 4/10\n",
            "7/7 [==============================] - 1s 78ms/step - loss: 0.1394 - acc: 1.0000 - val_loss: 0.6987 - val_acc: 0.5331\n",
            "Epoch 5/10\n",
            "7/7 [==============================] - 1s 78ms/step - loss: 0.0675 - acc: 1.0000 - val_loss: 0.7020 - val_acc: 0.5334\n",
            "Epoch 6/10\n",
            "7/7 [==============================] - 1s 81ms/step - loss: 0.0345 - acc: 1.0000 - val_loss: 0.7044 - val_acc: 0.5289\n",
            "Epoch 7/10\n",
            "7/7 [==============================] - 1s 80ms/step - loss: 0.0188 - acc: 1.0000 - val_loss: 0.7127 - val_acc: 0.5362\n",
            "Epoch 8/10\n",
            "7/7 [==============================] - 1s 80ms/step - loss: 0.0109 - acc: 1.0000 - val_loss: 0.7147 - val_acc: 0.5286\n",
            "Epoch 9/10\n",
            "7/7 [==============================] - 1s 82ms/step - loss: 0.0064 - acc: 1.0000 - val_loss: 0.7205 - val_acc: 0.5297\n",
            "Epoch 10/10\n",
            "7/7 [==============================] - 1s 80ms/step - loss: 0.0040 - acc: 1.0000 - val_loss: 0.7261 - val_acc: 0.5289\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0dRsqz5rnArD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.load_weights('/content/pre_trained_glove_model.h5')"
      ],
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HR1fokg2nArG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "eb28a945-089e-4967-b1a0-bc2451f75c71"
      },
      "source": [
        "model.evaluate(x_test, y_test)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "313/313 [==============================] - 1s 2ms/step - loss: 0.9941 - acc: 0.5503\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.9940569400787354, 0.5503000020980835]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UTtHG7U_nArJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}