Wiki103_from_scratch_fastai_v1_raw_articles.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"]=\"0\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai import *\n",
    "from fastai.text import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Python code Examples (testing out tokenizer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reading the texts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "PATH = Path('/home/farzin/rnn_python_code/wiki103_from_download')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/farzin/anaconda3/envs/fastaiv1_dev/lib/python3.7/site-packages/ipykernel_launcher.py:1: ParserWarning: Falling back to the 'python' engine because the 'c' engine does not support regex separators (separators > 1 char and different from '\\s+' are interpreted as regex); you can avoid this warning by specifying engine='python'.\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "all_texts_df = pd.read_csv('/home/farzin/fast_ai/fastai_v1/courses/dl2/imdb_scripts/data/wiki/en/train.csv',sep='%$#')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_texts = all_texts_df.values.squeeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#all_texts = np.concatenate([train_texts,valid_texts])\n",
    "idx = np.random.permutation(len(all_texts))\n",
    "cut = int(0.1 * len(idx))\n",
    "train_df = pd.DataFrame({'text':all_texts[idx[cut:]], 'labels':[0] * (len(all_texts)-cut)}, columns=['labels','text'])\n",
    "valid_df = pd.DataFrame({'text':all_texts[idx[:cut]], 'labels':[0] * cut}, columns=['labels','text'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Does any of this apply to Python code?  What do I need to include here that is relevant?\n",
    "\n",
    "#export\n",
    "def sub_br(t:str) -> str:\n",
    "    \"Replaces the <br /> by \\n\"\n",
    "    re_br = re.compile(r'<\\s*br\\s*/?>', re.IGNORECASE)\n",
    "    return re_br.sub(\"\\n\", t)\n",
    "\n",
    "def spec_add_spaces(t:str) -> str:\n",
    "    \"Add spaces between special characters\"\n",
    "    return re.sub(r'([/#])', r' \\1 ', t)\n",
    "\n",
    "def rm_useless_spaces(t:str) -> str:\n",
    "    \"Remove multiple spaces\"\n",
    "    return re.sub(' {2,}', ' ', t)\n",
    "\n",
    "def replace_rep(t:str) -> str:\n",
    "    \"Replace repetitions at the character level\"\n",
    "    def _replace_rep(m:Collection[str]) -> str:\n",
    "        c,cc = m.groups()\n",
    "        return f' {TK_REP} {len(cc)+1} {c} '\n",
    "    re_rep = re.compile(r'(\\S)(\\1{3,})')\n",
    "    return re_rep.sub(_replace_rep, t)\n",
    "    \n",
    "def replace_wrep(t:str) -> str:\n",
    "    \"Replace word repetitions\"\n",
    "    def _replace_wrep(m:Collection[str]) -> str:\n",
    "        c,cc = m.groups()\n",
    "        return f' {TK_WREP} {len(cc.split())+1} {c} '\n",
    "    re_wrep = re.compile(r'(\\b\\w+\\W+)(\\1{3,})')\n",
    "    return re_wrep.sub(_replace_wrep, t)\n",
    "\n",
    "def deal_caps(t:str) -> str:\n",
    "    \"Replace words in all caps\"\n",
    "    res = []\n",
    "    for s in re.findall(r'\\w+|\\W+', t):\n",
    "        res += ([f' {TK_UP} ',s.lower()] if (s.isupper() and (len(s)>2)) else [s.lower()])\n",
    "    return ''.join(res)\n",
    "\n",
    "def fixup(x:str) -> str:\n",
    "    \"List of replacements from html strings\"\n",
    "    re1 = re.compile(r'  +')\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>',UNK).replace(' @.@ ','.').replace(\n",
    "        ' @-@ ','-').replace('\\\\', ' \\\\ ')\n",
    "    return re1.sub(' ', html.unescape(x))\n",
    "\n",
    "default_rules = [fixup, replace_rep, replace_wrep, deal_caps, spec_add_spaces, rm_useless_spaces, sub_br]\n",
    "default_spec_tok = [BOS, FLD, UNK, PAD]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokenizer = Tokenizer(pre_rules=default_rules, special_cases=[BOS, FLD, 'xxunk', 'xxpad'], n_cpus=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Tokenizer SpacyTokenizer in en with the following rules:\n",
       " - fixup\n",
       " - replace_rep\n",
       " - replace_wrep\n",
       " - deal_caps\n",
       " - spec_add_spaces\n",
       " - rm_useless_spaces\n",
       " - sub_br\n",
       " - replace_all_caps\n",
       " - deal_caps"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = train_df.dropna()\n",
    "valid_df = valid_df.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_vocab = 60000 #can we make this the full possible vocab at 260k?\n",
    "batch_size = 32\n",
    "\n",
    "keyword_args = {'max_vocab':max_vocab,'bs':batch_size}\n",
    "data = TextLMDataBunch.from_df(PATH, train_df, valid_df, tokenizer=tokenizer,\n",
    "                               text_cols='text', label_cols='labels',**keyword_args)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60004"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data.vocab.itos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.save()"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#Save/load data object. \n",
    "data.save()\n",
    "data.load(PATH)\n",
    "#!ls -lhart {PATH}/tmp/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 1.2G\r\n",
      "drwxrwxr-x 3 farzin farzin 4.0K Jan 17 13:23 ..\r\n",
      "-rw-rw-r-- 1 farzin farzin 999K Jan 17 13:23 itos.pkl\r\n",
      "-rw-rw-r-- 1 farzin farzin 1.1G Jan 17 13:23 train_ids.npy\r\n",
      "-rw-rw-r-- 1 farzin farzin  20M Jan 17 13:23 train_lbl.npy\r\n",
      "-rw-rw-r-- 1 farzin farzin 118M Jan 17 13:24 valid_ids.npy\r\n",
      "-rw-rw-r-- 1 farzin farzin 2.2M Jan 17 13:24 valid_lbl.npy\r\n",
      "-rw-rw-r-- 1 farzin farzin    2 Jan 17 13:24 classes.txt\r\n",
      "drwxrwxr-x 2 farzin farzin 4.0K Jan 17 13:24 .\r\n"
     ]
    }
   ],
   "source": [
    "!ls -lhart {PATH}/tmp/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SequentialRNN(\n",
       "  (0): RNNCore(\n",
       "    (encoder): Embedding(60004, 400, padding_idx=1)\n",
       "    (encoder_dp): EmbeddingDropout(\n",
       "      (emb): Embedding(60004, 400, padding_idx=1)\n",
       "    )\n",
       "    (rnns): ModuleList(\n",
       "      (0): QRNNLayer(\n",
       "        (linear): WeightDropout(\n",
       "          (module): Linear(in_features=800, out_features=3333, bias=True)\n",
       "        )\n",
       "      )\n",
       "      (1): QRNNLayer(\n",
       "        (linear): WeightDropout(\n",
       "          (module): Linear(in_features=1111, out_features=3333, bias=True)\n",
       "        )\n",
       "      )\n",
       "      (2): QRNNLayer(\n",
       "        (linear): WeightDropout(\n",
       "          (module): Linear(in_features=1111, out_features=3333, bias=True)\n",
       "        )\n",
       "      )\n",
       "      (3): QRNNLayer(\n",
       "        (linear): WeightDropout(\n",
       "          (module): Linear(in_features=1111, out_features=1200, bias=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (input_dp): RNNDropout()\n",
       "    (hidden_dps): ModuleList(\n",
       "      (0): RNNDropout()\n",
       "      (1): RNNDropout()\n",
       "      (2): RNNDropout()\n",
       "      (3): RNNDropout()\n",
       "    )\n",
       "  )\n",
       "  (1): LinearDecoder(\n",
       "    (decoder): Linear(in_features=400, out_features=60004, bias=True)\n",
       "    (output_dp): RNNDropout()\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bptt = 140\n",
    "emb_sz,nh,nl = 400,1111,4\n",
    "\n",
    "learn = language_model_learner(data,bptt,emb_sz,nh,nl,drop_mult=0.5,qrnn=True)\n",
    "learn.unfreeze()\n",
    "learn.model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR Finder is complete, type {learner_name}.recorder.plot() to see the graph.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "lr_find(learn)\n",
    "learn.recorder.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "Total time: 53:45:14 <p><table style='width:300px; margin-bottom:10px'>\n",
       "  <tr>\n",
       "    <th>epoch</th>\n",
       "    <th>train_loss</th>\n",
       "    <th>valid_loss</th>\n",
       "    <th>accuracy</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>1</th>\n",
       "    <th>4.556215</th>\n",
       "    <th>4.435592</th>\n",
       "    <th>0.305401</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>2</th>\n",
       "    <th>4.568444</th>\n",
       "    <th>4.433817</th>\n",
       "    <th>0.302561</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>3</th>\n",
       "    <th>4.663723</th>\n",
       "    <th>4.550586</th>\n",
       "    <th>0.292597</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>4</th>\n",
       "    <th>4.731622</th>\n",
       "    <th>4.635006</th>\n",
       "    <th>0.284766</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>5</th>\n",
       "    <th>4.789747</th>\n",
       "    <th>4.663291</th>\n",
       "    <th>0.282145</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>6</th>\n",
       "    <th>4.782752</th>\n",
       "    <th>4.651239</th>\n",
       "    <th>0.283699</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>7</th>\n",
       "    <th>4.756957</th>\n",
       "    <th>4.639008</th>\n",
       "    <th>0.284986</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>8</th>\n",
       "    <th>4.752146</th>\n",
       "    <th>4.618721</th>\n",
       "    <th>0.286460</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>9</th>\n",
       "    <th>4.718633</th>\n",
       "    <th>4.594611</th>\n",
       "    <th>0.288452</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>10</th>\n",
       "    <th>4.657288</th>\n",
       "    <th>4.555625</th>\n",
       "    <th>0.291437</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>11</th>\n",
       "    <th>4.638156</th>\n",
       "    <th>4.509020</th>\n",
       "    <th>0.295237</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>12</th>\n",
       "    <th>4.598953</th>\n",
       "    <th>4.458078</th>\n",
       "    <th>0.299499</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>13</th>\n",
       "    <th>4.541291</th>\n",
       "    <th>4.409991</th>\n",
       "    <th>0.303776</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>14</th>\n",
       "    <th>4.472717</th>\n",
       "    <th>4.350044</th>\n",
       "    <th>0.309074</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>15</th>\n",
       "    <th>4.415161</th>\n",
       "    <th>4.288860</th>\n",
       "    <th>0.314878</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>16</th>\n",
       "    <th>4.359806</th>\n",
       "    <th>4.226478</th>\n",
       "    <th>0.321052</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>17</th>\n",
       "    <th>4.306016</th>\n",
       "    <th>4.170397</th>\n",
       "    <th>0.327303</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>18</th>\n",
       "    <th>4.289517</th>\n",
       "    <th>4.122715</th>\n",
       "    <th>0.332922</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>19</th>\n",
       "    <th>4.211309</th>\n",
       "    <th>4.090364</th>\n",
       "    <th>0.337469</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>20</th>\n",
       "    <th>4.206548</th>\n",
       "    <th>4.080127</th>\n",
       "    <th>0.338790</th>\n",
       "  </tr>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(20, 1e-2, moms=(0.8,0.7), wd=0.03, pct_start=0.25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.recorder.plot_losses()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.recorder.plot_lr(show_moms=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "59.14546984988227"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.exp(4.08)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "Total time: 10:43:11 <p><table style='width:300px; margin-bottom:10px'>\n",
       "  <tr>\n",
       "    <th>epoch</th>\n",
       "    <th>train_loss</th>\n",
       "    <th>valid_loss</th>\n",
       "    <th>accuracy</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>1</th>\n",
       "    <th>4.408109</th>\n",
       "    <th>4.277699</th>\n",
       "    <th>0.316112</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>2</th>\n",
       "    <th>4.365920</th>\n",
       "    <th>4.211289</th>\n",
       "    <th>0.322720</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>3</th>\n",
       "    <th>4.263391</th>\n",
       "    <th>4.117272</th>\n",
       "    <th>0.333146</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>4</th>\n",
       "    <th>4.188361</th>\n",
       "    <th>4.072366</th>\n",
       "    <th>0.338838</th>\n",
       "  </tr>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(4, 1e-2/4, moms=(0.95,0.8), wd=0.03, pct_start=0.15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.recorder.plot_lr(show_moms=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "Total time: 2:50:01 <p><table style='width:300px; margin-bottom:10px'>\n",
       "  <tr>\n",
       "    <th>epoch</th>\n",
       "    <th>train_loss</th>\n",
       "    <th>valid_loss</th>\n",
       "    <th>accuracy</th>\n",
       "  </tr>\n",
       "  <tr>\n",
       "    <th>1</th>\n",
       "    <th>4.239735</th>\n",
       "    <th>4.084991</th>\n",
       "    <th>0.337239</th>\n",
       "  </tr>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit(1,0.00025,wd=0.03)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save(\"wiki103_raw_articles_spacy_20190123\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save_encoder('wiki103_spacy_enc')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7 fasta.ai1 DEV",
   "language": "python",
   "name": "fastai1_dev"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}