Chandler_or_No_Chandler.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Chandler or No Chandler.ipynb",
      "version": "0.3.2",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "metadata": {
        "id": "GYz4uVAXnEDt",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## IMDb"
      ]
    },
    {
      "metadata": {
        "id": "SlhXfCZEvbKM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "3221049c-9174-44c1-f548-568999b6b34c"
      },
      "cell_type": "code",
      "source": [
        "# http://pytorch.org/\n",
        "from os import path\n",
        "from wheel.pep425tags import get_abbr_impl, get_impl_ver, get_abi_tag\n",
        "platform = '{}{}-{}'.format(get_abbr_impl(), get_impl_ver(), get_abi_tag())\n",
        "\n",
        "accelerator = 'cu80' if path.exists('/opt/bin/nvidia-smi') else 'cpu'\n",
        "\n",
        "!pip install -q http://download.pytorch.org/whl/{accelerator}/torch-0.3.0.post4-{platform}-linux_x86_64.whl torchvision\n",
        "import torch"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "metadata": {
        "id": "1JvApLhWvdQ4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2128
        },
        "collapsed": true,
        "outputId": "0cb6d656-d2bf-4df5-b088-fb2215054df1"
      },
      "cell_type": "code",
      "source": [
        "%time !pip install --upgrade git+https://github.com/fastai/fastai.git"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting git+https://github.com/fastai/fastai.git\r\n",
            "  Cloning https://github.com/fastai/fastai.git to /tmp/pip-req-build-7nnvrxmi\n",
            "Collecting bcolz (from fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/5c/4e/23942de9d5c0fb16f10335fa83e52b431bcb8c0d4a8419c9ac206268c279/bcolz-1.2.1.tar.gz (1.5MB)\n",
            "\u001b[K    100% |████████████████████████████████| 1.5MB 7.1MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: bleach in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.1.3)\n",
            "Requirement not upgraded as not directly required: certifi in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2018.4.16)\n",
            "Requirement not upgraded as not directly required: cycler in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.10.0)\n",
            "Requirement not upgraded as not directly required: decorator in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (4.3.0)\n",
            "Requirement not upgraded as not directly required: entrypoints in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.2.3)\n",
            "Collecting feather-format (from fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/08/55/940b97cc6f19a19f5dab9efef2f68a0ce43a7632f858b272391f0b851a7e/feather-format-0.4.0.tar.gz\n",
            "Collecting graphviz (from fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/84/44/21a7fdd50841aaaef224b943f7d10df87e476e181bb926ccf859bcb53d48/graphviz-0.8.3-py2.py3-none-any.whl\n",
            "Requirement not upgraded as not directly required: html5lib in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (1.0.1)\n",
            "Requirement not upgraded as not directly required: ipykernel in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (4.6.1)\n",
            "Requirement not upgraded as not directly required: ipython in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (5.5.0)\n",
            "Requirement not upgraded as not directly required: ipython-genutils in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.2.0)\n",
            "Collecting ipywidgets (from fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/7d/24/fabc09ad81c6071159a4d12d5bfbddcbea69bd9e3b16c3250ef300c0285f/ipywidgets-7.2.1-py2.py3-none-any.whl (106kB)\n",
            "\u001b[K    100% |████████████████████████████████| 112kB 9.0MB/s \n",
            "\u001b[?25hCollecting isoweek (from fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/c2/d4/fe7e2637975c476734fcbf53776e650a29680194eb0dd21dbdc020ca92de/isoweek-1.3.3-py2.py3-none-any.whl\n",
            "Collecting jedi (from fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/e7/42/074192a165622e645ed4aeade63e76e56b3496a044569b3c6cae3a918352/jedi-0.12.0-py2.py3-none-any.whl (172kB)\n",
            "\u001b[K    100% |████████████████████████████████| 174kB 9.0MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: Jinja2 in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.10)\n",
            "Requirement not upgraded as not directly required: jsonschema in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.6.0)\n",
            "Collecting jupyter (from fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/83/df/0f5dd132200728a86190397e1ea87cd76244e42d39ec5e88efd25b2abd7e/jupyter-1.0.0-py2.py3-none-any.whl\n",
            "Requirement not upgraded as not directly required: MarkupSafe in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (1.0)\n",
            "Requirement not upgraded as not directly required: matplotlib in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.1.2)\n",
            "Requirement not upgraded as not directly required: numpy in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (1.14.5)\n",
            "Requirement not upgraded as not directly required: opencv-python in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (3.4.1.15)\n",
            "Requirement not upgraded as not directly required: pandas in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.22.0)\n",
            "Collecting pandas_summary (from fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/3d/a7/0364272be0c6561c45d67edec8a7bf0532d56b830438168f9078f7720f63/pandas-summary-0.0.41.tar.gz\n",
            "Requirement not upgraded as not directly required: pickleshare in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.7.4)\n",
            "Requirement not upgraded as not directly required: Pillow in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (5.1.0)\n",
            "Collecting plotnine (from fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/3a/68/cf39dfde4e9fd886703621e3393cd8103cb48d5ecc95b8f048ec148e53a6/plotnine-0.3.0-py2.py3-none-any.whl (3.4MB)\n",
            "\u001b[K    100% |████████████████████████████████| 3.4MB 11.2MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: ptyprocess in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.5.2)\n",
            "Requirement not upgraded as not directly required: Pygments in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.1.3)\n",
            "Requirement not upgraded as not directly required: pyparsing in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.2.0)\n",
            "Requirement not upgraded as not directly required: python-dateutil in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2.5.3)\n",
            "Requirement not upgraded as not directly required: pytz in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (2018.4)\n",
            "Requirement not upgraded as not directly required: PyYAML in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (3.12)\n",
            "Requirement not upgraded as not directly required: pyzmq in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (16.0.4)\n",
            "Requirement not upgraded as not directly required: scipy in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.19.1)\n",
            "Requirement not upgraded as not directly required: seaborn in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.7.1)\n",
            "Requirement not upgraded as not directly required: simplegeneric in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.8.1)\n",
            "Collecting sklearn_pandas (from fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/c5/9e/42d7bcddb09a3ff52d0c60c810ba5d0fded28abbe320c85bbf7368192956/sklearn_pandas-1.6.0-py2.py3-none-any.whl\n",
            "Requirement not upgraded as not directly required: testpath in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.3.1)\n",
            "Requirement not upgraded as not directly required: torch<0.4 in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.3.0.post4)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Collecting torchtext (from fastai==0.7.0)\r\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/78/90/474d5944d43001a6e72b9aaed5c3e4f77516fbef2317002da2096fd8b5ea/torchtext-0.2.3.tar.gz (42kB)\r\n",
            "\u001b[K    100% |████████████████████████████████| 51kB 10.7MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: torchvision in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.2.1)\n",
            "Requirement not upgraded as not directly required: tornado in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (4.5.3)\n",
            "Collecting tqdm (from fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/93/24/6ab1df969db228aed36a648a8959d1027099ce45fad67532b9673d533318/tqdm-4.23.4-py2.py3-none-any.whl (42kB)\n",
            "\u001b[K    100% |████████████████████████████████| 51kB 18.6MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: traitlets in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (4.3.2)\n",
            "Requirement not upgraded as not directly required: wcwidth in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.1.7)\n",
            "Requirement not upgraded as not directly required: webencodings in /usr/local/lib/python3.6/dist-packages (from fastai==0.7.0) (0.5.1)\n",
            "Collecting widgetsnbextension (from fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/ff/fa/64acc09fc845a6b2dc0724d6f3f81e829b778ed5e9a7559567b4f19a3f4b/widgetsnbextension-3.2.1-py2.py3-none-any.whl (2.2MB)\n",
            "\u001b[K    100% |████████████████████████████████| 2.2MB 14.5MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: six in /usr/local/lib/python3.6/dist-packages (from bleach->fastai==0.7.0) (1.11.0)\n",
            "Collecting pyarrow>=0.4.0 (from feather-format->fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/fe/27/576e6c979bfa5f63070f809c930aa85d6198b91d17a64866234597ba861f/pyarrow-0.9.0-cp36-cp36m-manylinux1_x86_64.whl (13.7MB)\n",
            "\u001b[K    100% |████████████████████████████████| 13.7MB 3.6MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: jupyter-client in /usr/local/lib/python3.6/dist-packages (from ipykernel->fastai==0.7.0) (5.2.3)\n",
            "Requirement not upgraded as not directly required: setuptools>=18.5 in /usr/local/lib/python3.6/dist-packages (from ipython->fastai==0.7.0) (39.1.0)\n",
            "Requirement not upgraded as not directly required: pexpect; sys_platform != \"win32\" in /usr/local/lib/python3.6/dist-packages (from ipython->fastai==0.7.0) (4.6.0)\n",
            "Requirement not upgraded as not directly required: prompt-toolkit<2.0.0,>=1.0.4 in /usr/local/lib/python3.6/dist-packages (from ipython->fastai==0.7.0) (1.0.15)\n",
            "Requirement not upgraded as not directly required: nbformat>=4.2.0 in /usr/local/lib/python3.6/dist-packages (from ipywidgets->fastai==0.7.0) (4.4.0)\n",
            "Collecting parso>=0.2.0 (from jedi->fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/cd/3e/5908f9577dbd1e5df53e64349bfd11e46b726c1e4d8cd676bbe8aa4de316/parso-0.2.1-py2.py3-none-any.whl (91kB)\n",
            "\u001b[K    100% |████████████████████████████████| 92kB 23.2MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: nbconvert in /usr/local/lib/python3.6/dist-packages (from jupyter->fastai==0.7.0) (5.3.1)\n",
            "Collecting jupyter-console (from jupyter->fastai==0.7.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/77/82/6469cd7fccf7958cbe5dce2e623f1e3c5e27f1bb1ad36d90519bc2d5d370/jupyter_console-5.2.0-py2.py3-none-any.whl\n",
            "Requirement not upgraded as not directly required: notebook in /usr/local/lib/python3.6/dist-packages (from jupyter->fastai==0.7.0) (5.2.2)\n",
            "Collecting qtconsole (from jupyter->fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/90/ff/047e0dca2627b162866920e7aa93f04523c0ae81e5c67060eec85701992d/qtconsole-4.3.1-py2.py3-none-any.whl (108kB)\n",
            "\u001b[K    100% |████████████████████████████████| 112kB 23.1MB/s \n",
            "\u001b[?25hCollecting mizani>=0.4.1 (from plotnine->fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/52/01/8a3b4c6e45749674a1e5241174b4b63cd6435125e124bec275f3e02c96ac/mizani-0.4.6-py2.py3-none-any.whl (65kB)\n",
            "\u001b[K    100% |████████████████████████████████| 71kB 18.4MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: patsy>=0.4.1 in /usr/local/lib/python3.6/dist-packages (from plotnine->fastai==0.7.0) (0.5.0)\n",
            "Requirement not upgraded as not directly required: statsmodels>=0.8.0 in /usr/local/lib/python3.6/dist-packages (from plotnine->fastai==0.7.0) (0.8.0)\n",
            "Requirement not upgraded as not directly required: scikit-learn>=0.15.0 in /usr/local/lib/python3.6/dist-packages (from sklearn_pandas->fastai==0.7.0) (0.19.1)\n",
            "Requirement not upgraded as not directly required: requests in /usr/local/lib/python3.6/dist-packages (from torchtext->fastai==0.7.0) (2.18.4)\n",
            "Requirement not upgraded as not directly required: jupyter-core in /usr/local/lib/python3.6/dist-packages (from jupyter-client->ipykernel->fastai==0.7.0) (4.4.0)\n",
            "Requirement not upgraded as not directly required: pandocfilters>=1.4.1 in /usr/local/lib/python3.6/dist-packages (from nbconvert->jupyter->fastai==0.7.0) (1.4.2)\n",
            "Requirement not upgraded as not directly required: mistune>=0.7.4 in /usr/local/lib/python3.6/dist-packages (from nbconvert->jupyter->fastai==0.7.0) (0.8.3)\n",
            "Requirement not upgraded as not directly required: terminado>=0.3.3; sys_platform != \"win32\" in /usr/local/lib/python3.6/dist-packages (from notebook->jupyter->fastai==0.7.0) (0.8.1)\n",
            "Collecting palettable (from mizani>=0.4.1->plotnine->fastai==0.7.0)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/56/8a/84537c0354f0d1f03bf644b71bf8e0a50db9c1294181905721a5f3efbf66/palettable-3.1.1-py2.py3-none-any.whl (77kB)\n",
            "\u001b[K    100% |████████████████████████████████| 81kB 23.6MB/s \n",
            "\u001b[?25hRequirement not upgraded as not directly required: idna<2.7,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext->fastai==0.7.0) (2.6)\n",
            "Requirement not upgraded as not directly required: urllib3<1.23,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext->fastai==0.7.0) (1.22)\n",
            "Requirement not upgraded as not directly required: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext->fastai==0.7.0) (3.0.4)\n",
            "Building wheels for collected packages: fastai, bcolz, feather-format, pandas-summary, torchtext\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "  Running setup.py bdist_wheel for fastai ... \u001b[?25l-\b \b\\\b \bdone\n",
            "\u001b[?25h  Stored in directory: /tmp/pip-ephem-wheel-cache-2_k9sxhy/wheels/cf/46/39/b2d08762125ed2376861976ab2c4ac30c029b86e375735d9b8\n",
            "  Running setup.py bdist_wheel for bcolz ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/9f/78/26/fb8c0acb91a100dc8914bf236c4eaa4b207cb876893c40b745\n",
            "  Running setup.py bdist_wheel for feather-format ... \u001b[?25l-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/85/7d/12/2dfa5c0195f921ac935f5e8f27deada74972edc0ae9988a9c1\n",
            "  Running setup.py bdist_wheel for pandas-summary ... \u001b[?25l-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/86/a9/6b/36b727a9ec687dac08bd3d0501cddd7b1a223943513eb04a03\n",
            "  Running setup.py bdist_wheel for torchtext ... \u001b[?25l-\b \b\\\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/42/a6/f4/b267328bde6bb680094a0c173e8e5627ccc99543abded97204\n",
            "Successfully built fastai bcolz feather-format pandas-summary torchtext\n",
            "\u001b[31mplotnine 0.3.0 has requirement scipy>=1.0.0, but you'll have scipy 0.19.1 which is incompatible.\u001b[0m\n",
            "Installing collected packages: bcolz, pyarrow, feather-format, graphviz, widgetsnbextension, ipywidgets, isoweek, parso, jedi, jupyter-console, qtconsole, jupyter, pandas-summary, palettable, mizani, plotnine, sklearn-pandas, tqdm, torchtext, fastai\n",
            "Successfully installed bcolz-1.2.1 fastai-0.7.0 feather-format-0.4.0 graphviz-0.8.3 ipywidgets-7.2.1 isoweek-1.3.3 jedi-0.12.0 jupyter-1.0.0 jupyter-console-5.2.0 mizani-0.4.6 palettable-3.1.1 pandas-summary-0.0.41 parso-0.2.1 plotnine-0.3.0 pyarrow-0.9.0 qtconsole-4.3.1 sklearn-pandas-1.6.0 torchtext-0.2.3 tqdm-4.23.4 widgetsnbextension-3.2.1\n",
            "CPU times: user 1.9 s, sys: 624 ms, total: 2.52 s\n",
            "Wall time: 1min 24s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "dhGEi7111PDh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1465
        },
        "collapsed": true,
        "outputId": "526f08da-92a0-44b8-f1cf-717e611e6f8a"
      },
      "cell_type": "code",
      "source": [
        "!pip install spacy\n",
        "!python -m spacy download en"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting spacy\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/3c/31/e60f88751e48851b002f78a35221d12300783d5a43d4ef12fbf10cca96c3/spacy-2.0.11.tar.gz (17.6MB)\n",
            "\u001b[K    100% |████████████████████████████████| 17.6MB 2.0MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.7 in /usr/local/lib/python3.6/dist-packages (from spacy) (1.14.5)\n",
            "Collecting murmurhash<0.29,>=0.28 (from spacy)\n",
            "  Downloading https://files.pythonhosted.org/packages/5e/31/c8c1ecafa44db30579c8c457ac7a0f819e8b1dbc3e58308394fff5ff9ba7/murmurhash-0.28.0.tar.gz\n",
            "Collecting cymem<1.32,>=1.30 (from spacy)\n",
            "  Downloading https://files.pythonhosted.org/packages/f8/9e/273fbea507de99166c11cd0cb3fde1ac01b5bc724d9a407a2f927ede91a1/cymem-1.31.2.tar.gz\n",
            "Collecting preshed<2.0.0,>=1.0.0 (from spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/1b/ac/7c17b1fd54b60972785b646d37da2826311cca70842c011c4ff84fbe95e0/preshed-1.0.0.tar.gz (89kB)\n",
            "\u001b[K    100% |████████████████████████████████| 92kB 21.6MB/s \n",
            "\u001b[?25hCollecting thinc<6.11.0,>=6.10.1 (from spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/55/fd/e9f36081e6f53699943381858848f3b4d759e0dd03c43b98807dde34c252/thinc-6.10.2.tar.gz (1.2MB)\n",
            "\u001b[K    100% |████████████████████████████████| 1.2MB 13.4MB/s \n",
            "\u001b[?25hCollecting plac<1.0.0,>=0.9.6 (from spacy)\n",
            "  Downloading https://files.pythonhosted.org/packages/9e/9b/62c60d2f5bc135d2aa1d8c8a86aaf84edb719a59c7f11a4316259e61a298/plac-0.9.6-py2.py3-none-any.whl\n",
            "Collecting pathlib (from spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/ac/aa/9b065a76b9af472437a0059f77e8f962fe350438b927cb80184c32f075eb/pathlib-1.0.1.tar.gz (49kB)\n",
            "\u001b[K    100% |████████████████████████████████| 51kB 15.8MB/s \n",
            "\u001b[?25hCollecting ujson>=1.35 (from spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/16/c4/79f3409bc710559015464e5f49b9879430d8f87498ecdc335899732e5377/ujson-1.35.tar.gz (192kB)\n",
            "\u001b[K    100% |████████████████████████████████| 194kB 20.4MB/s \n",
            "\u001b[?25hCollecting dill<0.3,>=0.2 (from spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/6f/78/8b96476f4ae426db71c6e86a8e6a81407f015b34547e442291cd397b18f3/dill-0.2.8.2.tar.gz (150kB)\n",
            "\u001b[K    100% |████████████████████████████████| 153kB 21.6MB/s \n",
            "\u001b[?25hCollecting regex==2017.4.5 (from spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/36/62/c0c0d762ffd4ffaf39f372eb8561b8d491a11ace5a7884610424a8b40f95/regex-2017.04.05.tar.gz (601kB)\n",
            "\u001b[K    100% |████████████████████████████████| 604kB 19.8MB/s \n",
            "\u001b[?25hCollecting wrapt (from thinc<6.11.0,>=6.10.1->spacy)\n",
            "  Downloading https://files.pythonhosted.org/packages/a0/47/66897906448185fcb77fc3c2b1bc20ed0ecca81a0f2f88eda3fc5a34fc3d/wrapt-1.10.11.tar.gz\n",
            "Requirement already satisfied: tqdm<5.0.0,>=4.10.0 in /usr/local/lib/python3.6/dist-packages (from thinc<6.11.0,>=6.10.1->spacy) (4.23.4)\n",
            "Collecting cytoolz<0.9,>=0.8 (from thinc<6.11.0,>=6.10.1->spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/0f/e6/ccc124714dcc1bd511e64ddafb4d5d20ada2533b92e3173a4cf09e0d0831/cytoolz-0.8.2.tar.gz (386kB)\n",
            "\u001b[K    100% |████████████████████████████████| 389kB 21.6MB/s \n",
            "\u001b[?25hRequirement already satisfied: six<2.0.0,>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from thinc<6.11.0,>=6.10.1->spacy) (1.11.0)\n",
            "Requirement already satisfied: termcolor in /usr/local/lib/python3.6/dist-packages (from thinc<6.11.0,>=6.10.1->spacy) (1.1.0)\n",
            "Collecting msgpack-python (from thinc<6.11.0,>=6.10.1->spacy)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/8a/20/6eca772d1a5830336f84aca1d8198e5a3f4715cd1c7fc36d3cc7f7185091/msgpack-python-0.5.6.tar.gz (138kB)\n",
            "\u001b[K    100% |████████████████████████████████| 143kB 21.6MB/s \n",
            "\u001b[?25hCollecting msgpack-numpy==0.4.1 (from thinc<6.11.0,>=6.10.1->spacy)\n",
            "  Downloading https://files.pythonhosted.org/packages/2e/43/393e30e2768b0357541ac95891f96b80ccc4d517e0dd2fa3042fc8926538/msgpack_numpy-0.4.1-py2.py3-none-any.whl\n",
            "Requirement already satisfied: toolz>=0.8.0 in /usr/local/lib/python3.6/dist-packages (from cytoolz<0.9,>=0.8->thinc<6.11.0,>=6.10.1->spacy) (0.9.0)\n",
            "Building wheels for collected packages: spacy, murmurhash, cymem, preshed, thinc, pathlib, ujson, dill, regex, wrapt, cytoolz, msgpack-python\n",
            "  Running setup.py bdist_wheel for spacy ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/fb/00/28/75c85d5135e7d9a100639137d1847d41e914ed16c962d467e4\n",
            "  Running setup.py bdist_wheel for murmurhash ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/b8/94/a4/f69f8664cdc1098603df44771b7fec5fd1b3d8364cdd83f512\n",
            "  Running setup.py bdist_wheel for cymem ... \u001b[?25l-\b \b\\\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/55/8d/4a/f6328252aa2aaec0b1cb906fd96a1566d77f0f67701071ad13\n",
            "  Running setup.py bdist_wheel for preshed ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/8f/85/06/2d132fb649a6bbcab22487e4147880a55b0dd0f4b18fdfd6b5\n",
            "  Running setup.py bdist_wheel for thinc ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \b-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/d8/5c/3e/9acf5d9974fb1c9e7b467563ea5429c9325f67306e93147961\n",
            "  Running setup.py bdist_wheel for pathlib ... \u001b[?25l-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/f9/b2/4a/68efdfe5093638a9918bd1bb734af625526e849487200aa171\n",
            "  Running setup.py bdist_wheel for ujson ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/28/77/e4/0311145b9c2e2f01470e744855131f9e34d6919687550f87d1\n",
            "  Running setup.py bdist_wheel for dill ... \u001b[?25l-\b \b\\\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/e2/5d/17/f87cb7751896ac629b435a8696f83ee75b11029f5d6f6bda72\n",
            "  Running setup.py bdist_wheel for regex ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/75/07/38/3c16b529d50cb4e0cd3dbc7b75cece8a09c132692c74450b01\n",
            "  Running setup.py bdist_wheel for wrapt ... \u001b[?25l-\b \b\\\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/48/5d/04/22361a593e70d23b1f7746d932802efe1f0e523376a74f321e\n",
            "  Running setup.py bdist_wheel for cytoolz ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b \b\\\b \b|\b \b/\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/f8/b1/86/c92e4d36b690208fff8471711b85eaa6bc6d19860a86199a09\n",
            "  Running setup.py bdist_wheel for msgpack-python ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \bdone\n",
            "\u001b[?25h  Stored in directory: /content/.cache/pip/wheels/d5/de/86/7fa56fda12511be47ea0808f3502bc879df4e63ab168ec0406\n",
            "Successfully built spacy murmurhash cymem preshed thinc pathlib ujson dill regex wrapt cytoolz msgpack-python\n",
            "\u001b[31mplotnine 0.3.0 has requirement scipy>=1.0.0, but you'll have scipy 0.19.1 which is incompatible.\u001b[0m\n",
            "Installing collected packages: murmurhash, cymem, preshed, wrapt, cytoolz, plac, dill, pathlib, msgpack-python, msgpack-numpy, thinc, ujson, regex, spacy\n",
            "Successfully installed cymem-1.31.2 cytoolz-0.8.2 dill-0.2.8.2 msgpack-numpy-0.4.1 msgpack-python-0.5.6 murmurhash-0.28.0 pathlib-1.0.1 plac-0.9.6 preshed-1.0.0 regex-2017.4.5 spacy-2.0.11 thinc-6.10.2 ujson-1.35 wrapt-1.10.11\n",
            "Collecting https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.0.0/en_core_web_sm-2.0.0.tar.gz\n",
            "\u001b[?25l  Downloading https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.0.0/en_core_web_sm-2.0.0.tar.gz (37.4MB)\n",
            "\u001b[K    100% |████████████████████████████████| 37.4MB 44.0MB/s \n",
            "\u001b[?25hInstalling collected packages: en-core-web-sm\n",
            "  Running setup.py install for en-core-web-sm ... \u001b[?25l-\b \b\\\b \b|\b \bdone\n",
            "\u001b[?25hSuccessfully installed en-core-web-sm-2.0.0\n",
            "\n",
            "\u001b[93m    Linking successful\u001b[0m\n",
            "    /usr/local/lib/python3.6/dist-packages/en_core_web_sm -->\n",
            "    /usr/local/lib/python3.6/dist-packages/spacy/data/en\n",
            "\n",
            "    You can now load the model via spacy.load('en')\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "bADbkspbnEDw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "4f0c3e9a-4356-475c-a492-5c626dc3b649"
      },
      "cell_type": "code",
      "source": [
        "from fastai.text import *\n",
        "import html"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "metadata": {
        "id": "bQQPKR45nED5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "3eb19fce-de23-497f-e316-3e61581a7ce0"
      },
      "cell_type": "code",
      "source": [
        "BOS = 'xbos'  # beginning-of-sentence tag\n",
        "FLD = 'xfld'  # data field tag\n",
        "\n",
        "PATH=Path('/content/data/')"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "metadata": {
        "id": "bwkN783o4x8b",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "582dbdf4-ee98-4d78-b5ea-108ce5e7d0d4"
      },
      "cell_type": "code",
      "source": [
        "!mkdir data/\n",
        "import os\n",
        "os.chdir('data/')\n",
        "# !wget http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "metadata": {
        "id": "AMwVo8Ie3D-e",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "1a3f3580-c1e1-4fc9-8c1c-2c83c4088c60"
      },
      "cell_type": "code",
      "source": [
        "!pip install -U -q PyDrive\n",
        "import os\n",
        "from pydrive.auth import GoogleAuth\n",
        "from pydrive.drive import GoogleDrive\n",
        "from google.colab import auth\n",
        "from oauth2client.client import GoogleCredentials\n",
        "\n",
        "# 1. Authenticate and create the PyDrive client.\n",
        "auth.authenticate_user()\n",
        "gauth = GoogleAuth()\n",
        "gauth.credentials = GoogleCredentials.get_application_default()\n",
        "drive = GoogleDrive(gauth)\n",
        "\n",
        "# choose a local (colab) directory to store the data.\n",
        "local_download_path = os.path.expanduser('~/data')\n",
        "try:\n",
        "  os.makedirs(local_download_path)\n",
        "except: pass\n",
        "\n",
        "# 2. Auto-iterate using the query syntax\n",
        "#    https://developers.google.com/drive/v2/web/search-parameters\n",
        "file_list = drive.ListFile(\n",
        "    {'q': \"'1-_UnohUZPUuIN4CDUYCtTVAeB0mtIO04' in parents\"}).GetList()\n",
        "\n",
        "for f in file_list:\n",
        "  # 3. Create & download by id.\n",
        "  if(f['title']=='train.txt'):\n",
        "    print('title: %s, id: %s' % (f['title'], f['id']))\n",
        "    fname = os.path.join(local_download_path, f['title'])\n",
        "    print('downloading to {}'.format(fname))\n",
        "    f_ = drive.CreateFile({'id': f['id']})\n",
        "    f_.GetContentFile(fname)\n"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[31mplotnine 0.3.0 has requirement scipy>=1.0.0, but you'll have scipy 0.19.1 which is incompatible.\u001b[0m\r\n",
            "title: train.txt, id: 1H_6UEsydwxBfTmOf3XewAzp6xUoDqR1Z\n",
            "downloading to /content/data/train.txt\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "3XMNjZSn9idj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "3ff6d76a-7893-4351-9abc-2ce16cbbe891"
      },
      "cell_type": "code",
      "source": [
        "for f in file_list:\n",
        "  # 3. Create & download by id.\n",
        "  if(f['title']=='labels_test.txt'):\n",
        "    print('title: %s, id: %s' % (f['title'], f['id']))\n",
        "    fname = os.path.join(local_download_path, f['title'])\n",
        "    print('downloading to {}'.format(fname))\n",
        "    f_ = drive.CreateFile({'id': f['id']})\n",
        "    f_.GetContentFile(fname)\n"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "title: labels_test.txt, id: 1G3o4LqgXS4wCixVkVu5e5BIRC61U1wWW\n",
            "downloading to /content/data/labels_test.txt\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "cnMpkn1s5v-b",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "244965f4-0e10-4517-b2e4-f54a5e560c7f"
      },
      "cell_type": "code",
      "source": [
        "texts_val = []\n",
        "fname = Path('/content/data/')\n",
        "texts_val.append((fname/'test.txt').open('r').read())"
      ],
      "execution_count": 131,
      "outputs": []
    },
    {
      "metadata": {
        "id": "rmK6vDzc-OxD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "a45ede55-19cd-4870-84a2-6b2f548e17e4"
      },
      "cell_type": "code",
      "source": [
        "texts_val = texts_val[0].split('\\n')"
      ],
      "execution_count": 132,
      "outputs": []
    },
    {
      "metadata": {
        "id": "btG-SR8F-_NP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "dcac70df-c170-430d-c049-d95013dbaace"
      },
      "cell_type": "code",
      "source": [
        "labels_val = []\n",
        "labels_val.append((fname/'labels_test.txt').open('r').read())\n",
        "len(labels_val)"
      ],
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 133
        }
      ]
    },
    {
      "metadata": {
        "id": "9xmlXema_54j",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "110dec91-c90f-4c65-937d-198995fe7e20"
      },
      "cell_type": "code",
      "source": [
        "labels_val = labels_val[0].split()\n",
        "len(labels_val)"
      ],
      "execution_count": 134,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "21553"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 134
        }
      ]
    },
    {
      "metadata": {
        "id": "Cxh7OVydAJxx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "b4c9b370-13b5-4bdd-e2e3-a15f2759bb47"
      },
      "cell_type": "code",
      "source": [
        "len(texts_val)"
      ],
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "21553"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 135
        }
      ]
    },
    {
      "metadata": {
        "id": "cfuNHdhNnEEA",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Standardize format"
      ]
    },
    {
      "metadata": {
        "id": "quiVQ-rNnEEb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "d047f7ea-f994-4086-c97d-b98189d0208b"
      },
      "cell_type": "code",
      "source": [
        "texts = np.asarray(texts)\n",
        "labels = np.asarray(labels)"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "metadata": {
        "id": "XclEeiJcnEEh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "87e753ff-7493-470d-8847-ed6616227a11"
      },
      "cell_type": "code",
      "source": [
        "np.random.seed(42)\n",
        "trn_idx = np.random.permutation(len(texts))\n",
        "# val_idx = np.random.permutation(len(val_texts))"
      ],
      "execution_count": 63,
      "outputs": []
    },
    {
      "metadata": {
        "id": "HINKLiAXBo27",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "a47d9b90-376d-4084-8cef-3538901c764e"
      },
      "cell_type": "code",
      "source": [
        "trn_idx.shape"
      ],
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(193974,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "metadata": {
        "id": "qFoC8dV4nEEn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "e66a5f16-17e9-4f7c-bed7-2b05f737c639"
      },
      "cell_type": "code",
      "source": [
        "trn_texts = texts[trn_idx]\n",
        "# val_texts = val_texts[val_idx]\n",
        "\n",
        "trn_labels = labels[trn_idx]\n",
        "# val_labels = val_labels[val_idx]"
      ],
      "execution_count": 65,
      "outputs": []
    },
    {
      "metadata": {
        "id": "NbpBEbL_Gzys",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "45188707-ae43-4e44-a30a-f488f9423a0e"
      },
      "cell_type": "code",
      "source": [
        "trn_texts.shape"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(193974,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 66
        }
      ]
    },
    {
      "metadata": {
        "id": "qv0xn9zfnEEt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "3000b101-468e-4109-fd01-38bf4a639171"
      },
      "cell_type": "code",
      "source": [
        "# df_trn = pd.DataFrame({'text':trn_texts, 'labels':trn_labels}, columns=col_names)\n",
        "df_val = pd.DataFrame({'text':texts_val, 'labels':labels_val}, columns=col_names)"
      ],
      "execution_count": 136,
      "outputs": []
    },
    {
      "metadata": {
        "id": "XmWhw9QpnEE2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "72f9cb9e-c330-4a29-f87f-b4ea1055bd86"
      },
      "cell_type": "code",
      "source": [
        "CLAS_PATH = Path('/content/data/CLAS/')\n",
        "CLAS_PATH.mkdir()\n",
        "# df_trn[df_trn['labels']!=2].to_csv((fname/'train.csv'), header=False, index=False)\n",
        "df_val.to_csv(CLAS_PATH/'test.csv', header=False, index=False)\n",
        "\n",
        "# (fname/'classes.txt').open('w').writelines(f'{o}\\n' for o in CLASSES)"
      ],
      "execution_count": 137,
      "outputs": []
    },
    {
      "metadata": {
        "id": "z8jUbp-1nEE8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "fc8c239a-4667-427a-bad7-f8736209c2f1"
      },
      "cell_type": "code",
      "source": [
        "trn_texts,val_texts = sklearn.model_selection.train_test_split(\n",
        "   trn_texts, test_size=0.1)"
      ],
      "execution_count": 69,
      "outputs": []
    },
    {
      "metadata": {
        "id": "6LhLUieHnEFC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "9035aa5b-b903-4595-b093-37d1b10da66b"
      },
      "cell_type": "code",
      "source": [
        "len(trn_texts), len(val_texts)"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(174576, 19398)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 70
        }
      ]
    },
    {
      "metadata": {
        "id": "9TxO9Ta0HZzZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "ac7cdbc0-2585-48c5-b409-6bf2d8f81581"
      },
      "cell_type": "code",
      "source": [
        "LM_PATH = Path('/content/data/LM/')"
      ],
      "execution_count": 75,
      "outputs": []
    },
    {
      "metadata": {
        "id": "4MAl24gEHihx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "e5864ab9-c1af-4e52-e890-c6c216317158"
      },
      "cell_type": "code",
      "source": [
        "LM_PATH.mkdir()"
      ],
      "execution_count": 76,
      "outputs": []
    },
    {
      "metadata": {
        "id": "z1FoHc14nEFK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "6873fd77-c292-4b07-8bc9-887be3ec1f9c"
      },
      "cell_type": "code",
      "source": [
        "df_trn = pd.DataFrame({'text':trn_texts, 'labels':[0]*len(trn_texts)}, columns=col_names)\n",
        "df_val = pd.DataFrame({'text':val_texts, 'labels':[0]*len(val_texts)}, columns=col_names)\n",
        "\n",
        "df_trn.to_csv(LM_PATH/'train.csv', header=False, index=False)\n",
        "df_val.to_csv(LM_PATH/'test.csv', header=False, index=False)"
      ],
      "execution_count": 77,
      "outputs": []
    },
    {
      "metadata": {
        "id": "jxnYKR-FnEFS",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Language model tokens"
      ]
    },
    {
      "metadata": {
        "id": "tJU4HRKKnEFU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "f956becd-a3f4-4b35-d1cb-fb8086d224fd"
      },
      "cell_type": "code",
      "source": [
        "chunksize=10000"
      ],
      "execution_count": 78,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QOq8eRC1nEFb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "4aceb613-a2c5-41db-8394-0b7a132fd04a"
      },
      "cell_type": "code",
      "source": [
        "re1 = re.compile(r'  +')\n",
        "\n",
        "def fixup(x):\n",
        "    x = x.replace('#39;', \"'\").replace('amp;', '&').replace('#146;', \"'\").replace(\n",
        "        'nbsp;', ' ').replace('#36;', '$').replace('\\\\n', \"\\n\").replace('quot;', \"'\").replace(\n",
        "        '<br />', \"\\n\").replace('\\\\\"', '\"').replace('<unk>','u_n').replace(' @.@ ','.').replace(\n",
        "        ' @-@ ','-').replace('\\\\', ' \\\\ ')\n",
        "    return re1.sub(' ', html.unescape(x))"
      ],
      "execution_count": 79,
      "outputs": []
    },
    {
      "metadata": {
        "id": "STrL8SAcnEFg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "c08fb251-b0e1-4b0f-9a90-b16c1958accb"
      },
      "cell_type": "code",
      "source": [
        "def get_texts(df, n_lbls=1):\n",
        "    labels = df.iloc[:,range(n_lbls)].values.astype(np.int64)\n",
        "    texts = f'\\n{BOS} {FLD} 1 ' + df[n_lbls].astype(str)\n",
        "    for i in range(n_lbls+1, len(df.columns)): texts += f' {FLD} {i-n_lbls} ' + df[i].astype(str)\n",
        "    texts = texts.apply(fixup).values.astype(str)\n",
        "\n",
        "    tok = Tokenizer().proc_all_mp(partition_by_cores(texts))\n",
        "    return tok, list(labels)"
      ],
      "execution_count": 80,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ae_eSIm1nEFm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "fcd147f2-2d0f-40c6-c80e-386bfd8a0fb9"
      },
      "cell_type": "code",
      "source": [
        "def get_all(df, n_lbls):\n",
        "    tok, labels = [], []\n",
        "    for i, r in enumerate(df):\n",
        "        print(i)\n",
        "        tok_, labels_ = get_texts(r, n_lbls)\n",
        "        tok += tok_;\n",
        "        labels += labels_\n",
        "    return tok, labels"
      ],
      "execution_count": 81,
      "outputs": []
    },
    {
      "metadata": {
        "id": "AwurwhiMnEFr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "2e1015a6-a1a2-4506-9cae-3055bb252d99"
      },
      "cell_type": "code",
      "source": [
        "df_trn = pd.read_csv(LM_PATH/'train.csv', header=None, chunksize=chunksize)\n",
        "df_val = pd.read_csv(LM_PATH/'test.csv', header=None, chunksize=chunksize)"
      ],
      "execution_count": 82,
      "outputs": []
    },
    {
      "metadata": {
        "id": "i193aqawnEFw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "collapsed": true,
        "outputId": "24766a41-6b4b-4f62-f6ca-85dbb339d863"
      },
      "cell_type": "code",
      "source": [
        "tok_trn, trn_labels = get_all(df_trn, 1)\n",
        "tok_val, val_labels = get_all(df_val, 1)"
      ],
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0\n",
            "1\n",
            "2\n",
            "3\n",
            "4\n",
            "5\n",
            "6\n",
            "7\n",
            "8\n",
            "9\n",
            "10\n",
            "11\n",
            "12\n",
            "13\n",
            "14\n",
            "15\n",
            "16\n",
            "17\n",
            "0\n",
            "1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "1Dk9hMdznEF6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "7d810ed7-56f2-4f2b-af9f-6a0e2ed9c152"
      },
      "cell_type": "code",
      "source": [
        "(LM_PATH/'tmp').mkdir(exist_ok=True)"
      ],
      "execution_count": 84,
      "outputs": []
    },
    {
      "metadata": {
        "id": "OeJfgoUJnEF_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "94c5cf9b-6f3e-4b76-c191-c2fb1de05e0f"
      },
      "cell_type": "code",
      "source": [
        "np.save(LM_PATH/'tmp'/'tok_trn.npy', tok_trn)\n",
        "np.save(LM_PATH/'tmp'/'tok_val.npy', tok_val)"
      ],
      "execution_count": 85,
      "outputs": []
    },
    {
      "metadata": {
        "id": "3SMm3bM9nEGH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "14d8d83a-e198-4afd-f0ea-b982c9501076"
      },
      "cell_type": "code",
      "source": [
        "tok_trn = np.load(LM_PATH/'tmp'/'tok_trn.npy')\n",
        "tok_val = np.load(LM_PATH/'tmp'/'tok_val.npy')"
      ],
      "execution_count": 86,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fiFkygRonEGN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "6a7edf31-8190-492f-cee2-8ad8b03d1804"
      },
      "cell_type": "code",
      "source": [
        "freq = Counter(p for o in tok_trn for p in o)\n",
        "len(freq)"
      ],
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "73883"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 87
        }
      ]
    },
    {
      "metadata": {
        "id": "P_hkpns2nEGa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "84e6546f-0a9b-4e33-e78f-2c6733c66e34"
      },
      "cell_type": "code",
      "source": [
        "max_vocab = 60000\n",
        "min_freq = 3"
      ],
      "execution_count": 88,
      "outputs": []
    },
    {
      "metadata": {
        "id": "aAAU6QA2nEGj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "0d48655b-e1b8-41a5-af7a-e3dee996faa8"
      },
      "cell_type": "code",
      "source": [
        "itos = [o for o,c in freq.most_common(max_vocab) if c>min_freq]\n",
        "itos.insert(0, '_pad_')\n",
        "itos.insert(0, '_unk_')"
      ],
      "execution_count": 89,
      "outputs": []
    },
    {
      "metadata": {
        "id": "J0LA0JPcnEGp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "704aa90e-6e2e-43d6-dcda-0a30bd6398db"
      },
      "cell_type": "code",
      "source": [
        "stoi = collections.defaultdict(lambda:0, {v:k for k,v in enumerate(itos)})\n",
        "len(itos)"
      ],
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "19973"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 90
        }
      ]
    },
    {
      "metadata": {
        "id": "ZM2Ta5SHnEGy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "283eae73-49e2-45cd-b7f6-b957e7a613af"
      },
      "cell_type": "code",
      "source": [
        "trn_lm = np.array([[stoi[o] for o in p] for p in tok_trn])\n",
        "val_lm = np.array([[stoi[o] for o in p] for p in tok_val])"
      ],
      "execution_count": 91,
      "outputs": []
    },
    {
      "metadata": {
        "id": "TCf1vT0_nEG6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "9a2fc6fb-7a4d-47a3-d833-877149d736ca"
      },
      "cell_type": "code",
      "source": [
        "np.save(LM_PATH/'tmp'/'trn_ids.npy', trn_lm)\n",
        "np.save(LM_PATH/'tmp'/'val_ids.npy', val_lm)\n",
        "pickle.dump(itos, open(LM_PATH/'tmp'/'itos.pkl', 'wb'))"
      ],
      "execution_count": 92,
      "outputs": []
    },
    {
      "metadata": {
        "id": "l5nStIqHnEG_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "09bee84e-99db-4568-b68d-80be3a0f51e0"
      },
      "cell_type": "code",
      "source": [
        "trn_lm = np.load(LM_PATH/'tmp'/'trn_ids.npy')\n",
        "val_lm = np.load(LM_PATH/'tmp'/'val_ids.npy')\n",
        "itos = pickle.load(open(LM_PATH/'tmp'/'itos.pkl', 'rb'))"
      ],
      "execution_count": 93,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Y7f8Ipp9nEHE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "3c976de7-5352-4801-9af2-b8480db912e6"
      },
      "cell_type": "code",
      "source": [
        "vs=len(itos)\n",
        "vs,len(trn_lm)"
      ],
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(19973, 174576)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 94
        }
      ]
    },
    {
      "metadata": {
        "id": "yrWMSJiynEHR",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## wikitext103 conversion"
      ]
    },
    {
      "metadata": {
        "id": "ZkH2maK85qDq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "af6d95f5-2cb2-4d7c-f55a-b0236d431b72"
      },
      "cell_type": "code",
      "source": [
        "Path('/content/data/models/').mkdir()\n",
        "os.chdir('/content/data/models/')"
      ],
      "execution_count": 97,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fTE0GnaZnEHT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2587
        },
        "collapsed": true,
        "outputId": "f928044b-a5a6-40d0-c828-8bd3bccad4f7"
      },
      "cell_type": "code",
      "source": [
        "! wget -nH -r -np http://files.fast.ai/models/wt103/"
      ],
      "execution_count": 98,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2018-06-26 08:10:33--  http://files.fast.ai/models/wt103/\n",
            "Resolving files.fast.ai (files.fast.ai)... 67.205.15.147\n",
            "Connecting to files.fast.ai (files.fast.ai)|67.205.15.147|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:33 (116 MB/s) - ‘models/wt103/index.html’ saved [857/857]\n",
            "\n",
            "Loading robots.txt; please ignore errors.\n",
            "--2018-06-26 08:10:33--  http://files.fast.ai/robots.txt\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 404 Not Found\n",
            "2018-06-26 08:10:33 ERROR 404: Not Found.\n",
            "\n",
            "--2018-06-26 08:10:33--  http://files.fast.ai/models/wt103/?C=N;O=D\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=N;O=D’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:33 (126 MB/s) - ‘models/wt103/index.html?C=N;O=D’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:33--  http://files.fast.ai/models/wt103/?C=M;O=A\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=M;O=A’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:33 (120 MB/s) - ‘models/wt103/index.html?C=M;O=A’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:33--  http://files.fast.ai/models/wt103/?C=S;O=A\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=S;O=A’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:33 (129 MB/s) - ‘models/wt103/index.html?C=S;O=A’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:33--  http://files.fast.ai/models/wt103/?C=D;O=A\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=D;O=A’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:33 (123 MB/s) - ‘models/wt103/index.html?C=D;O=A’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:33--  http://files.fast.ai/models/wt103/bwd_wt103.h5\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 462387687 (441M) [text/plain]\n",
            "Saving to: ‘models/wt103/bwd_wt103.h5’\n",
            "\n",
            "models/wt103/bwd_wt 100%[===================>] 440.97M   111MB/s    in 4.0s    \n",
            "\n",
            "2018-06-26 08:10:37 (109 MB/s) - ‘models/wt103/bwd_wt103.h5’ saved [462387687/462387687]\n",
            "\n",
            "--2018-06-26 08:10:37--  http://files.fast.ai/models/wt103/bwd_wt103_enc.h5\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 462387634 (441M) [text/plain]\n",
            "Saving to: ‘models/wt103/bwd_wt103_enc.h5’\n",
            "\n",
            "els/wt103/bwd_wt103  65%[============>       ] 290.46M   112MB/s               "
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "models/wt103/bwd_wt 100%[===================>] 440.97M   112MB/s    in 3.9s    \n",
            "\n",
            "2018-06-26 08:10:41 (112 MB/s) - ‘models/wt103/bwd_wt103_enc.h5’ saved [462387634/462387634]\n",
            "\n",
            "--2018-06-26 08:10:41--  http://files.fast.ai/models/wt103/fwd_wt103.h5\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 462387687 (441M) [text/plain]\n",
            "Saving to: ‘models/wt103/fwd_wt103.h5’\n",
            "\n",
            "models/wt103/fwd_wt 100%[===================>] 440.97M   111MB/s    in 4.0s    \n",
            "\n",
            "2018-06-26 08:10:45 (111 MB/s) - ‘models/wt103/fwd_wt103.h5’ saved [462387687/462387687]\n",
            "\n",
            "--2018-06-26 08:10:45--  http://files.fast.ai/models/wt103/fwd_wt103_enc.h5\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 462387634 (441M) [text/plain]\n",
            "Saving to: ‘models/wt103/fwd_wt103_enc.h5’\n",
            "\n",
            "models/wt103/fwd_wt 100%[===================>] 440.97M   112MB/s    in 3.9s    \n",
            "\n",
            "2018-06-26 08:10:49 (112 MB/s) - ‘models/wt103/fwd_wt103_enc.h5’ saved [462387634/462387634]\n",
            "\n",
            "--2018-06-26 08:10:49--  http://files.fast.ai/models/wt103/itos_wt103.pkl\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 4161252 (4.0M) [text/plain]\n",
            "Saving to: ‘models/wt103/itos_wt103.pkl’\n",
            "\n",
            "models/wt103/itos_w 100%[===================>]   3.97M  --.-KB/s    in 0.04s   \n",
            "\n",
            "2018-06-26 08:10:49 (111 MB/s) - ‘models/wt103/itos_wt103.pkl’ saved [4161252/4161252]\n",
            "\n",
            "--2018-06-26 08:10:49--  http://files.fast.ai/models/wt103/?C=N;O=A\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=N;O=A’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:49 (97.3 MB/s) - ‘models/wt103/index.html?C=N;O=A’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:49--  http://files.fast.ai/models/wt103/?C=M;O=D\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=M;O=D’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:49 (99.9 MB/s) - ‘models/wt103/index.html?C=M;O=D’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:49--  http://files.fast.ai/models/wt103/?C=S;O=D\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=S;O=D’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:49 (112 MB/s) - ‘models/wt103/index.html?C=S;O=D’ saved [857/857]\n",
            "\n",
            "--2018-06-26 08:10:49--  http://files.fast.ai/models/wt103/?C=D;O=D\n",
            "Reusing existing connection to files.fast.ai:80.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 857 [text/html]\n",
            "Saving to: ‘models/wt103/index.html?C=D;O=D’\n",
            "\n",
            "models/wt103/index. 100%[===================>]     857  --.-KB/s    in 0s      \n",
            "\n",
            "2018-06-26 08:10:49 (95.7 MB/s) - ‘models/wt103/index.html?C=D;O=D’ saved [857/857]\n",
            "\n",
            "FINISHED --2018-06-26 08:10:49--\n",
            "Total wall clock time: 16s\n",
            "Downloaded: 14 files, 1.7G in 16s (111 MB/s)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "C7ErkEeoL7nE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "7b9822df-fdb5-4706-d4d4-f239ea3c80af"
      },
      "cell_type": "code",
      "source": [
        "!mv /content/data/models/models/* /content/data/models/"
      ],
      "execution_count": 102,
      "outputs": []
    },
    {
      "metadata": {
        "id": "xy_ZcQtvEopw",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# !mkdir /content/data/aclImdb/models\n",
        "# !mv /content/data/aclImdb/wt103/ /content/data/aclImdb/models/"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "NqbQ0V61nEHb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "c9cc2b49-bc23-44e5-e3f1-4ec83bccdeee"
      },
      "cell_type": "code",
      "source": [
        "em_sz,nh,nl = 400,1150,3"
      ],
      "execution_count": 99,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QnEcFdqYnEHi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "42865544-f1fa-41e9-def9-0c9e418385d9"
      },
      "cell_type": "code",
      "source": [
        "PRE_PATH = fname/'models'/'wt103'\n",
        "PRE_LM_PATH = PRE_PATH/'fwd_wt103.h5'"
      ],
      "execution_count": 100,
      "outputs": []
    },
    {
      "metadata": {
        "id": "eFzNouc_nEHo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "953c3334-b2d9-4c2c-f100-2468aa9cd64c"
      },
      "cell_type": "code",
      "source": [
        "wgts = torch.load(PRE_LM_PATH, map_location=lambda storage, loc: storage)"
      ],
      "execution_count": 103,
      "outputs": []
    },
    {
      "metadata": {
        "id": "AKPwe1L1nEHt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "56797c9e-1be5-48fd-fdad-eb1ff3f1dac0"
      },
      "cell_type": "code",
      "source": [
        "enc_wgts = to_np(wgts['0.encoder.weight'])\n",
        "row_m = enc_wgts.mean(0)"
      ],
      "execution_count": 104,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Oy0JeirOnEHx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "a8f5414c-c568-476c-ab51-f1346cb40d91"
      },
      "cell_type": "code",
      "source": [
        "itos2 = pickle.load((PRE_PATH/'itos_wt103.pkl').open('rb'))\n",
        "stoi2 = collections.defaultdict(lambda:-1, {v:k for k,v in enumerate(itos2)})"
      ],
      "execution_count": 105,
      "outputs": []
    },
    {
      "metadata": {
        "id": "DJ6fsI1lSxVj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "0d51e7d3-785f-4334-b7ae-103fd03ebc57"
      },
      "cell_type": "code",
      "source": [
        "x = {1:\"a\", 2:\"b\", 3:\"c\"}\n",
        "for i,j in enumerate(x):\n",
        "  print(i,j)"
      ],
      "execution_count": 106,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0 1\n",
            "1 2\n",
            "2 3\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "MxW3M1_bnEH2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "d29f4893-5b27-4f0e-bfb0-85be3e4b715c"
      },
      "cell_type": "code",
      "source": [
        "new_w = np.zeros((vs, em_sz), dtype=np.float32)\n",
        "for i,w in enumerate(itos):\n",
        "    r = stoi2[w]\n",
        "    new_w[i] = enc_wgts[r] if r>=0 else row_m"
      ],
      "execution_count": 107,
      "outputs": []
    },
    {
      "metadata": {
        "id": "WVy3NKXdnEH7",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "3c50c561-3c2d-4251-fbea-5d35ccb775f4"
      },
      "cell_type": "code",
      "source": [
        "wgts['0.encoder.weight'] = T(new_w)\n",
        "wgts['0.encoder_with_dropout.embed.weight'] = T(np.copy(new_w))\n",
        "wgts['1.decoder.weight'] = T(np.copy(new_w))"
      ],
      "execution_count": 108,
      "outputs": []
    },
    {
      "metadata": {
        "id": "4TvOXLtgnEH_",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Language model"
      ]
    },
    {
      "metadata": {
        "id": "HIsSyQNunEIA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "94dd11ea-e8cf-44f3-ea50-199289f8f532"
      },
      "cell_type": "code",
      "source": [
        "wd=1e-7\n",
        "bptt=70\n",
        "bs=52\n",
        "opt_fn = partial(optim.Adam, betas=(0.8, 0.99))"
      ],
      "execution_count": 109,
      "outputs": []
    },
    {
      "metadata": {
        "id": "8a4ImahMnEIF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "17d170db-1097-4bee-a821-039c2be09543"
      },
      "cell_type": "code",
      "source": [
        "trn_dl = LanguageModelLoader(np.concatenate(trn_lm), bs, bptt)\n",
        "val_dl = LanguageModelLoader(np.concatenate(val_lm), bs, bptt)\n",
        "md = LanguageModelData(PATH, 1, vs, trn_dl, val_dl, bs=bs, bptt=bptt)"
      ],
      "execution_count": 110,
      "outputs": []
    },
    {
      "metadata": {
        "id": "-YZvOY_AnEIJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "bea37402-7520-475b-c52d-524d12b1264c"
      },
      "cell_type": "code",
      "source": [
        "drops = np.array([0.25, 0.1, 0.2, 0.02, 0.15])*0.7"
      ],
      "execution_count": 111,
      "outputs": []
    },
    {
      "metadata": {
        "id": "dYz-lnNYnEIW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "e713e82b-7aaa-48f7-9233-e920b8969e8b"
      },
      "cell_type": "code",
      "source": [
        "learner= md.get_model(opt_fn, em_sz, nh, nl, \n",
        "    dropouti=drops[0], dropout=drops[1], wdrop=drops[2], dropoute=drops[3], dropouth=drops[4])\n",
        "\n",
        "learner.metrics = [accuracy]\n",
        "learner.unfreeze()"
      ],
      "execution_count": 112,
      "outputs": []
    },
    {
      "metadata": {
        "id": "gRXpJaQHnEIb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "0b04473f-7553-4ad7-c6bb-ee34430b6ab7"
      },
      "cell_type": "code",
      "source": [
        "learner.model.load_state_dict(wgts)"
      ],
      "execution_count": 113,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7pkTlnFjnEIg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "669f912f-0faf-4ba5-bf32-e0af6f7def09"
      },
      "cell_type": "code",
      "source": [
        "lr=1e-3\n",
        "lrs = lr"
      ],
      "execution_count": 114,
      "outputs": []
    },
    {
      "metadata": {
        "id": "MK51ks4MFYwh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "5e89fe2c-37c6-4c4a-ec19-1c1323829939"
      },
      "cell_type": "code",
      "source": [
        "learner.freeze_to(-2)"
      ],
      "execution_count": 115,
      "outputs": []
    },
    {
      "metadata": {
        "id": "X-xat9W0nEIk",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "7b688407-908b-4411-9f82-9c4f1d601316"
      },
      "cell_type": "code",
      "source": [
        "learner.fit(lrs/2, 1, wds=wd, use_clr=(32,2), cycle_len=1, cycle_mult=2)"
      ],
      "execution_count": 116,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "8ed192beaaca4d3b8f1849b4d9ee2e84",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "epoch      trn_loss   val_loss   accuracy   \n",
            "    0      4.179501   3.990243   0.338889  \n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([3.99024]), 0.3388894772529602]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 116
        }
      ]
    },
    {
      "metadata": {
        "id": "nJagPDw4nEIv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "74c53f67-2190-45d2-f4d4-70aeb2cca6b7"
      },
      "cell_type": "code",
      "source": [
        "learner.save('lm_last_ft')"
      ],
      "execution_count": 117,
      "outputs": []
    },
    {
      "metadata": {
        "id": "gscT9yBqnEI2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "collapsed": true,
        "outputId": "cde7e7ac-e834-4b2b-b66b-74da06e93009"
      },
      "cell_type": "code",
      "source": [
        "learner.load('lm_last_ft')"
      ],
      "execution_count": 118,
      "outputs": []
    },
    {
      "metadata": {
        "id": "JDPGmgq3nEI-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "d2987ef6-51cb-41fc-935d-609b26ce65b3"
      },
      "cell_type": "code",
      "source": [
        "learner.unfreeze()"
      ],
      "execution_count": 119,
      "outputs": []
    },
    {
      "metadata": {
        "id": "u6O_TDJSnEJF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "collapsed": true,
        "outputId": "51e6b325-7cf5-437e-f35a-4db6ad6c0ec9"
      },
      "cell_type": "code",
      "source": [
        "learner.lr_find()#start_lr=lrs/10, end_lr=lrs*10, linear=True)"
      ],
      "execution_count": 120,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "d2e3d71fa80642cbae9eedc55aae5950",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            " 72%|███████▏  | 817/1139 [06:13<02:27,  2.19it/s, loss=16]  "
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "GfJfR2kdnEJO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "collapsed": true,
        "outputId": "1eb39d3e-726f-48a4-a8e0-d70de5a50b20"
      },
      "cell_type": "code",
      "source": [
        "learner.sched.plot(n_skip=2, n_skip_end=1)"
      ],
      "execution_count": 121,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFcCAYAAADlIuYrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhDYAAB5BaAMA4BH/HyZYTngTrTPaAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f1063253e80>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "FlCHD-EwnEJU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "collapsed": true,
        "outputId": "73dea893-5236-48c7-b50e-cfdde4d9d2f0"
      },
      "cell_type": "code",
      "source": [
        "learner.fit(2e-5, 1, wds=wd, use_clr=(20,10), cycle_len=2)"
      ],
      "execution_count": 122,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "006c70bae9394b0883bb18efbc3d0bcf",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=2), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            ""
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "epoch      trn_loss   val_loss   accuracy   \n",
            "    0      4.095288   3.92304    0.346232  \n",
            " 29%|██▉       | 334/1139 [02:32<06:07,  2.19it/s, loss=4.09]"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "    1      4.078028   3.906646   0.348019  \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([3.90665]), 0.3480192596912384]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 122
        }
      ]
    },
    {
      "metadata": {
        "id": "m14uvC7nnEJf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "a5d84ed4-9b7a-4a37-84ba-f9aaad5c0441"
      },
      "cell_type": "code",
      "source": [
        "learner.save('lm11')"
      ],
      "execution_count": 123,
      "outputs": []
    },
    {
      "metadata": {
        "id": "PPjrf2U0STKs",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "/content/data/models/"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "HD2eTDR9nEJi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "e9630606-1938-4a8f-9b54-f36d3c41c6ad"
      },
      "cell_type": "code",
      "source": [
        "learner.save_encoder('lm11_enc')"
      ],
      "execution_count": 124,
      "outputs": []
    },
    {
      "metadata": {
        "id": "6noMI-X-5EH6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "collapsed": true,
        "outputId": "a11cf429-35f1-4169-8951-b95042962366"
      },
      "cell_type": "code",
      "source": [
        "learner.load('lm11')"
      ],
      "execution_count": 127,
      "outputs": []
    },
    {
      "metadata": {
        "id": "_q9ba_AtnEJq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        },
        "outputId": "1b43eee4-57cd-40ff-dd6f-ec120cf4d89d"
      },
      "cell_type": "code",
      "source": [
        "learner.sched.plot_loss()"
      ],
      "execution_count": 126,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFKCAYAAAAwrQetAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f10c4a62f98>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "qtUpeaGdnEJw",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Classifier tokens"
      ]
    },
    {
      "metadata": {
        "id": "AmEl5eevnEJy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "38c0f8ef-b8b3-4756-d83a-f641aabf6d80"
      },
      "cell_type": "code",
      "source": [
        "df_trn = pd.read_csv(fname/'train.csv', header=None, chunksize=chunksize)\n",
        "df_val = pd.read_csv(CLAS_PATH/'test.csv', header=None, chunksize=chunksize)"
      ],
      "execution_count": 138,
      "outputs": []
    },
    {
      "metadata": {
        "id": "WdI2wQAPnEJ2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 408
        },
        "collapsed": true,
        "outputId": "f91b16a8-3335-4a1f-f883-4d0b4717999f"
      },
      "cell_type": "code",
      "source": [
        "tok_trn, trn_labels = get_all(df_trn, 1)\n",
        "tok_val, val_labels = get_all(df_val, 1)"
      ],
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0\n",
            "1\n",
            "2\n",
            "3\n",
            "4\n",
            "5\n",
            "6\n",
            "7\n",
            "8\n",
            "9\n",
            "10\n",
            "11\n",
            "12\n",
            "13\n",
            "14\n",
            "15\n",
            "16\n",
            "17\n",
            "18\n",
            "19\n",
            "0\n",
            "1\n",
            "2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "_pNInKOanEJ-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "0cf7c7d6-063d-4ee6-9dcd-70cc018e48f6"
      },
      "cell_type": "code",
      "source": [
        "(CLAS_PATH/'tmp').mkdir(exist_ok=True)\n",
        "\n",
        "np.save(CLAS_PATH/'tmp'/'tok_trn.npy', tok_trn)\n",
        "np.save(CLAS_PATH/'tmp'/'tok_val.npy', tok_val)\n",
        "\n",
        "np.save(CLAS_PATH/'tmp'/'trn_labels.npy', trn_labels)\n",
        "np.save(CLAS_PATH/'tmp'/'val_labels.npy', val_labels)"
      ],
      "execution_count": 140,
      "outputs": []
    },
    {
      "metadata": {
        "id": "3iYroYEXnEKG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "7b21b434-992c-4a90-8315-d469f2f9632e"
      },
      "cell_type": "code",
      "source": [
        "tok_trn = np.load(CLAS_PATH/'tmp'/'tok_trn.npy')\n",
        "tok_val = np.load(CLAS_PATH/'tmp'/'tok_val.npy')"
      ],
      "execution_count": 141,
      "outputs": []
    },
    {
      "metadata": {
        "id": "qDkTFD7tnEKK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "ea6c0339-6a87-4dac-94a6-c0ea02a47d74"
      },
      "cell_type": "code",
      "source": [
        "itos = pickle.load((LM_PATH/'tmp'/'itos.pkl').open('rb'))\n",
        "stoi = collections.defaultdict(lambda:0, {v:k for k,v in enumerate(itos)})\n",
        "len(itos)"
      ],
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "19973"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 142
        }
      ]
    },
    {
      "metadata": {
        "id": "Oth-55KRnEKR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "0f3dc157-fb4c-459b-ba24-d710f7829706"
      },
      "cell_type": "code",
      "source": [
        "trn_clas = np.array([[stoi[o] for o in p] for p in tok_trn])\n",
        "val_clas = np.array([[stoi[o] for o in p] for p in tok_val])"
      ],
      "execution_count": 143,
      "outputs": []
    },
    {
      "metadata": {
        "id": "6cbINedunEKU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "bf3ed834-d0e9-4ef3-e4ec-49053d4062be"
      },
      "cell_type": "code",
      "source": [
        "np.save(CLAS_PATH/'tmp'/'trn_ids.npy', trn_clas)\n",
        "np.save(CLAS_PATH/'tmp'/'val_ids.npy', val_clas)"
      ],
      "execution_count": 144,
      "outputs": []
    },
    {
      "metadata": {
        "id": "PQeB9JVCnEKY",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Classifier"
      ]
    },
    {
      "metadata": {
        "id": "IBf3TgINnEKZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "681186a7-ea42-4fc6-a23f-23732785d38a"
      },
      "cell_type": "code",
      "source": [
        "trn_clas = np.load(CLAS_PATH/'tmp'/'trn_ids.npy')\n",
        "val_clas = np.load(CLAS_PATH/'tmp'/'val_ids.npy')"
      ],
      "execution_count": 145,
      "outputs": []
    },
    {
      "metadata": {
        "id": "wWm3vy69nEKd",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "513e7fca-7d86-4a89-865d-b33d4df77e99"
      },
      "cell_type": "code",
      "source": [
        "trn_labels = np.squeeze(np.load(CLAS_PATH/'tmp'/'trn_labels.npy'))\n",
        "val_labels = np.squeeze(np.load(CLAS_PATH/'tmp'/'val_labels.npy'))"
      ],
      "execution_count": 146,
      "outputs": []
    },
    {
      "metadata": {
        "id": "LqysXfIEnEKm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "2799d5f8-e5c3-470c-831f-03c8cd0de05d"
      },
      "cell_type": "code",
      "source": [
        "bptt,em_sz,nh,nl = 70,400,1150,3\n",
        "vs = len(itos)\n",
        "opt_fn = partial(optim.Adam, betas=(0.8, 0.99))\n",
        "bs = 48"
      ],
      "execution_count": 147,
      "outputs": []
    },
    {
      "metadata": {
        "id": "YaIbrBk2nEKs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "2761129c-b649-44a1-aae9-bf33fc746ce5"
      },
      "cell_type": "code",
      "source": [
        "min_lbl = trn_labels.min()\n",
        "trn_labels -= min_lbl\n",
        "val_labels -= min_lbl\n",
        "c=int(trn_labels.max())+1"
      ],
      "execution_count": 148,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ZCeK2Dh2nEKv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "ab21d373-f70f-41b0-e1c6-07f83c135dfc"
      },
      "cell_type": "code",
      "source": [
        "trn_ds = TextDataset(trn_clas, trn_labels)\n",
        "val_ds = TextDataset(val_clas, val_labels)\n",
        "trn_samp = SortishSampler(trn_clas, key=lambda x: len(trn_clas[x]), bs=bs//2)\n",
        "val_samp = SortSampler(val_clas, key=lambda x: len(val_clas[x]))\n",
        "trn_dl = DataLoader(trn_ds, bs//2, transpose=True, num_workers=1, pad_idx=1, sampler=trn_samp)\n",
        "val_dl = DataLoader(val_ds, bs, transpose=True, num_workers=1, pad_idx=1, sampler=val_samp)\n",
        "md = ModelData(PATH, trn_dl, val_dl)"
      ],
      "execution_count": 149,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7CkWZZxInEK0",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "78bc04cf-391e-4f96-cd14-77ad47a25134"
      },
      "cell_type": "code",
      "source": [
        "# part 1\n",
        "dps = np.array([0.4, 0.5, 0.05, 0.3, 0.1])"
      ],
      "execution_count": 150,
      "outputs": []
    },
    {
      "metadata": {
        "id": "DM35cLwSnEK8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "593bf46f-2d56-4a3e-e19a-ae6b99b23914"
      },
      "cell_type": "code",
      "source": [
        "dps = np.array([0.4,0.5,0.05,0.3,0.4])*0.5"
      ],
      "execution_count": 151,
      "outputs": []
    },
    {
      "metadata": {
        "id": "DhT3QdBenELB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "339d7fb2-63c7-428d-f745-550e55e19902"
      },
      "cell_type": "code",
      "source": [
        "m = get_rnn_classifer(bptt, 20*70, c, vs, emb_sz=em_sz, n_hid=nh, n_layers=nl, pad_token=1,\n",
        "          layers=[em_sz*3, 50, c], drops=[dps[4], 0.1],\n",
        "          dropouti=dps[0], wdrop=dps[1], dropoute=dps[2], dropouth=dps[3])"
      ],
      "execution_count": 152,
      "outputs": []
    },
    {
      "metadata": {
        "id": "MWeQ-6V6nELH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "a0ba3904-dca6-492f-8b1e-4b4bca76ed29"
      },
      "cell_type": "code",
      "source": [
        "opt_fn = partial(optim.Adam, betas=(0.7, 0.99))"
      ],
      "execution_count": 153,
      "outputs": []
    },
    {
      "metadata": {
        "id": "T_yHVLJWnELK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "a84a74ac-21e5-4578-99d4-44b928defee5"
      },
      "cell_type": "code",
      "source": [
        "learn = RNN_Learner(md, TextModel(to_gpu(m)), opt_fn=opt_fn)\n",
        "learn.reg_fn = partial(seq2seq_reg, alpha=2, beta=1)\n",
        "learn.clip=25.\n",
        "learn.metrics = [accuracy]"
      ],
      "execution_count": 154,
      "outputs": []
    },
    {
      "metadata": {
        "id": "tQbel_V0nELN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "62da34fb-e939-47a1-a326-81cbc5e99af8"
      },
      "cell_type": "code",
      "source": [
        "lr=3e-3\n",
        "lrm = 2.6\n",
        "lrs = np.array([lr/(lrm**4), lr/(lrm**3), lr/(lrm**2), lr/lrm, lr])"
      ],
      "execution_count": 155,
      "outputs": []
    },
    {
      "metadata": {
        "id": "uMcLZB5FnELS",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "275db8a0-98e1-44e5-b8c5-2aa99fd07095"
      },
      "cell_type": "code",
      "source": [
        "lrs=np.array([1e-4,1e-4,1e-4,1e-3,1e-2])"
      ],
      "execution_count": 156,
      "outputs": []
    },
    {
      "metadata": {
        "id": "2NINi5lz5kzu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "39747317-ac91-4048-f5a2-d464edfcead7"
      },
      "cell_type": "code",
      "source": [
        "os.chdir('/content/')"
      ],
      "execution_count": 157,
      "outputs": []
    },
    {
      "metadata": {
        "id": "LnSmG35jAy0G",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "1f0223f1-59c4-4d14-e403-ca31b5809c01"
      },
      "cell_type": "code",
      "source": [
        "!pwd"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\r\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "ZBNA0FFRv539",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f9deeb2d-c3aa-4d87-e3c7-d4ce578e139c"
      },
      "cell_type": "code",
      "source": [
        "type(val_clas[0])"
      ],
      "execution_count": 291,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "list"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 291
        }
      ]
    },
    {
      "metadata": {
        "id": "wu9gdFAtnELV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "47db6b93-8878-45ee-cc28-d79bdaf0e801"
      },
      "cell_type": "code",
      "source": [
        "wd = 1e-7\n",
        "wd = 0\n",
        "learn.load_encoder('/content/data/models/lm11_enc')"
      ],
      "execution_count": 161,
      "outputs": []
    },
    {
      "metadata": {
        "id": "2fSOfxkInELa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "e65c9632-fcab-4abb-c950-131729a4330c"
      },
      "cell_type": "code",
      "source": [
        "learn.freeze_to(-1)"
      ],
      "execution_count": 162,
      "outputs": []
    },
    {
      "metadata": {
        "id": "cBZVB18qnELj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "collapsed": true,
        "outputId": "e909c35b-a8b9-492b-fbbc-b93b30137652"
      },
      "cell_type": "code",
      "source": [
        "learn.lr_find(lrs/1000)\n",
        "learn.sched.plot()"
      ],
      "execution_count": 294,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f66ec733df4a473387e9c477951ed656",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            " 69%|██████▉   | 5559/8083 [11:59<05:26,  7.73it/s, loss=0.466]"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfUAAAFcCAYAAADRWyc3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f1027807048>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "I_IaE117nELp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "3232ed8f-3001-4922-d435-6d6360f4a4b2"
      },
      "cell_type": "code",
      "source": [
        "learn.fit(lrs, 1, wds=wd, cycle_len=1, use_clr=(8,3))"
      ],
      "execution_count": 164,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f5fa92b6a8ca4163942d19b7ce73f44b",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "epoch      trn_loss   val_loss   accuracy   \n",
            "    0      0.342574   0.312996   0.87482   \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([0.313]), 0.8748202118824703]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 164
        }
      ]
    },
    {
      "metadata": {
        "id": "ppIJBiHwnELt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "44bac9a6-e2c8-487b-9f96-7479ac0ff785"
      },
      "cell_type": "code",
      "source": [
        "learn.save('clas_0')"
      ],
      "execution_count": 165,
      "outputs": []
    },
    {
      "metadata": {
        "id": "eG4Yt6W2nELx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "2d8aa2c1-8fa7-449e-849f-2460a59c574c"
      },
      "cell_type": "code",
      "source": [
        "learn.load('clas_0')"
      ],
      "execution_count": 166,
      "outputs": []
    },
    {
      "metadata": {
        "id": "5-nQehTtnEL2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "938f9b0d-a55f-41db-b548-5c32bca65a0b"
      },
      "cell_type": "code",
      "source": [
        "learn.freeze_to(-2)"
      ],
      "execution_count": 167,
      "outputs": []
    },
    {
      "metadata": {
        "id": "w6ydMKQKnEL5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "a59c2e79-1943-4ff8-fb8f-ee6e59dc15d2"
      },
      "cell_type": "code",
      "source": [
        "learn.fit(lrs, 1, wds=wd, cycle_len=1, use_clr=(8,3))"
      ],
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "6ab115afa72d455e99c445aeca04098a",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "epoch      trn_loss   val_loss   accuracy   \n",
            "    0      0.297706   0.265753   0.892776  \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([0.26575]), 0.8927759471771718]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 168
        }
      ]
    },
    {
      "metadata": {
        "id": "oftIKFtpnEL9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "2f6d3607-fdf6-4f4b-e794-c93eb377bb3e"
      },
      "cell_type": "code",
      "source": [
        "learn.save('clas_1')"
      ],
      "execution_count": 169,
      "outputs": []
    },
    {
      "metadata": {
        "id": "xOK-v1rAnEME",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "51c427fd-a8a3-4e53-c72a-d3978d84a1a8"
      },
      "cell_type": "code",
      "source": [
        "learn.load('clas_1')"
      ],
      "execution_count": 170,
      "outputs": []
    },
    {
      "metadata": {
        "id": "hDOMrZbVb-Jp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "0b3265fa-f9f7-435f-98db-4bcc935df9fa"
      },
      "cell_type": "code",
      "source": [
        "learn.metrics = [accuracy]"
      ],
      "execution_count": 175,
      "outputs": []
    },
    {
      "metadata": {
        "id": "0QaKSX3rnEMJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "4104c324-22bc-44c4-c25a-172b2ada8f41"
      },
      "cell_type": "code",
      "source": [
        "learn.unfreeze()"
      ],
      "execution_count": 209,
      "outputs": []
    },
    {
      "metadata": {
        "id": "vWDg4i00nEMN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "collapsed": true,
        "outputId": "d1b53256-3472-4613-f734-c3c066cdf4a6"
      },
      "cell_type": "code",
      "source": [
        "learn.fit(lrs, 1, wds=wd, cycle_len=4, use_clr=(32,10))"
      ],
      "execution_count": 210,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "8dafcfde117d4f5d85f7281a5088b2f8",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Epoch', max=4), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            ""
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "epoch      trn_loss   val_loss   accuracy   \n",
            "    0      0.237879   0.220661   0.912912  \n",
            "  4%|▍         | 356/8083 [00:44<16:15,  7.92it/s, loss=0.22] "
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "    1      0.219465   0.216532   0.921357  \n",
            "  0%|          | 0/8083 [00:00<?, ?it/s, loss=0.22]"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "    2      0.188966   0.208992   0.921449  \n",
            "  0%|          | 0/8083 [00:00<?, ?it/s, loss=0.188]"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "    3      0.159816   0.210661   0.923352  \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([0.21066]), 0.9233517379311184]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 210
        }
      ]
    },
    {
      "metadata": {
        "id": "dNcCqCq7nEMU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        },
        "outputId": "05f3d5e3-ae81-478d-bd41-77c859557960"
      },
      "cell_type": "code",
      "source": [
        "learn.sched.plot_loss()"
      ],
      "execution_count": 215,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAFKCAYAAAAnj5dkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYXMmIiKyM2zOREREdobNmYiIyM78Fxy5ftWoK8J0AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f102edfa048>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "HL1N5xawnEMY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "05847050-942d-47a5-97ad-b2dd24fa9585"
      },
      "cell_type": "code",
      "source": [
        "learn.save('clas_2')"
      ],
      "execution_count": 211,
      "outputs": []
    },
    {
      "metadata": {
        "id": "rNRDSlNYjBT5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "d323f6fc-3a67-4a77-c10d-39bc651edc6c"
      },
      "cell_type": "code",
      "source": [
        "preds = learn.predict()\n",
        "preds.shape"
      ],
      "execution_count": 223,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(21553, 2)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 223
        }
      ]
    },
    {
      "metadata": {
        "id": "VtUuy1HZjSOG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "771fe91f-2f15-4254-ef2a-6e658a517ca1"
      },
      "cell_type": "code",
      "source": [
        "y_pred = np.argmax(np.exp(preds), axis=1)"
      ],
      "execution_count": 268,
      "outputs": []
    },
    {
      "metadata": {
        "id": "b2IDORuwSRE_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "38b810f6-1ef7-461d-acd6-adde79e105e8"
      },
      "cell_type": "code",
      "source": [
        "y_pred = [int(i) for i in y_pred]"
      ],
      "execution_count": 269,
      "outputs": []
    },
    {
      "metadata": {
        "id": "_-AwrcffUEss",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "outputId": "0a8c0ee8-95cd-4067-b5f0-6acaf5dec921"
      },
      "cell_type": "code",
      "source": [
        "y_pred = T(y_pred)"
      ],
      "execution_count": 271,
      "outputs": []
    },
    {
      "metadata": {
        "id": "xJ8SFlWpiuw7",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "d556d1f6-6ded-4df7-8568-f5e778c6bae0"
      },
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score"
      ],
      "execution_count": 203,
      "outputs": []
    },
    {
      "metadata": {
        "id": "_ymv9-Lm8R9o",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 748
        },
        "collapsed": true,
        "outputId": "d157bdad-f551-4c88-e5fc-f703cf130c88"
      },
      "cell_type": "code",
      "source": [
        "learn.model.eval()"
      ],
      "execution_count": 222,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SequentialRNN(\n",
              "  (0): MultiBatchRNN(\n",
              "    (encoder): Embedding(19973, 400, padding_idx=1)\n",
              "    (encoder_with_dropout): EmbeddingDropout(\n",
              "      (embed): Embedding(19973, 400, padding_idx=1)\n",
              "    )\n",
              "    (rnns): ModuleList(\n",
              "      (0): WeightDrop(\n",
              "        (module): LSTM(400, 1150)\n",
              "      )\n",
              "      (1): WeightDrop(\n",
              "        (module): LSTM(1150, 1150)\n",
              "      )\n",
              "      (2): WeightDrop(\n",
              "        (module): LSTM(1150, 400)\n",
              "      )\n",
              "    )\n",
              "    (dropouti): LockedDropout(\n",
              "    )\n",
              "    (dropouths): ModuleList(\n",
              "      (0): LockedDropout(\n",
              "      )\n",
              "      (1): LockedDropout(\n",
              "      )\n",
              "      (2): LockedDropout(\n",
              "      )\n",
              "    )\n",
              "  )\n",
              "  (1): PoolingLinearClassifier(\n",
              "    (layers): ModuleList(\n",
              "      (0): LinearBlock(\n",
              "        (lin): Linear(in_features=1200, out_features=50)\n",
              "        (drop): Dropout(p=0.2)\n",
              "        (bn): BatchNorm1d(1200, eps=1e-05, momentum=0.1, affine=True)\n",
              "      )\n",
              "      (1): LinearBlock(\n",
              "        (lin): Linear(in_features=50, out_features=2)\n",
              "        (drop): Dropout(p=0.1)\n",
              "        (bn): BatchNorm1d(50, eps=1e-05, momentum=0.1, affine=True)\n",
              "      )\n",
              "    )\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 222
        }
      ]
    },
    {
      "metadata": {
        "id": "es21DE8YmJWO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 37
        },
        "outputId": "3930198e-adf8-4e81-f1c6-03b53da52cbd"
      },
      "cell_type": "code",
      "source": [
        "int_labels = [int(i) for i in labels_val]\n",
        "int_labels = T(int_labels)"
      ],
      "execution_count": 267,
      "outputs": []
    },
    {
      "metadata": {
        "id": "KnSilAbXnEMk",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        },
        "outputId": "1211a3ac-4dd5-4e08-d6fc-c936b85947be"
      },
      "cell_type": "code",
      "source": [
        "learn.sched.plot_loss()"
      ],
      "execution_count": 221,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAFKCAYAAAAnj5dkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYXMmIiKyM2zOREREdobNmYiIyM78Fxy5ftWoK8J0AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f10287aafd0>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "QeYr1gRKNfNp",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}