notebooks/unsupervised_models_check.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from collections import defaultdict\n",
    "\n",
    "import fasttext\n",
    "import fasttext.util\n",
    "\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-10_loss-hs_nongrams', '../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-10_loss-ns_nongrams', '../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-7_loss-hs_lr-0.05_ngrams-1', '../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-7_loss-hs_nongrams', '../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-7_loss-ns_nongrams']\n",
      "['sunset', 'song', 'music', 'ibiza', 'people', 'school', 'laptop', 'computer', 'sand', 'sky', 'player']\n"
     ]
    }
   ],
   "source": [
    "training_path = '../semantic_learning/training_experiments/unsupervised/'\n",
    "models = sorted([training_path + file_name for file_name in os.listdir(training_path) if file_name != '.DS_Store'])\n",
    "print(models)\n",
    "words = ['sunset', 'song', 'music', 'ibiza', 'people', 'school', 'laptop', 'computer', 'sand', 'sky', 'player']\n",
    "print(words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning : `load_model` does not return WordVectorModel or SupervisedModel any more, but a `FastText` object which is very similar.\n"
     ]
    }
   ],
   "source": [
    "neighbors = defaultdict(lambda: defaultdict(list))\n",
    "for model_name in ['../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-7_loss-hs_nongrams']:\n",
    "    model = fasttext.load_model(model_name + '/unsupervised_model.bin')\n",
    "    for word in words:\n",
    "        neighbors[model_name][word] = model.get_nearest_neighbors(word)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fasttext_dim-100_epoch-7_loss-hs_nongrams\n",
      "sunset [(0.8617753386497498, 'sunrise'), (0.7498640418052673, 'dusk'), (0.7480682134628296, 'twilight'), (0.7005783915519714, 'sea'), (0.6984926462173462, 'cloudy'), (0.6844289898872375, 'jetty'), (0.6735334396362305, 'lake'), (0.6686261892318726, 'ocean'), (0.6632625460624695, 'foreground'), (0.6624375581741333, 'pier')]\n",
      "song [(0.746880829334259, 'songs'), (0.7412995100021362, 'sings'), (0.6545684933662415, 'sing'), (0.6517430543899536, 'voice'), (0.6449910998344421, 'singing'), (0.6435425281524658, 'tunes'), (0.6390900015830994, 'melody'), (0.6193773150444031, 'sang'), (0.6168822646141052, 'lyrics'), (0.6165723204612732, 'microphone')]\n",
      "music [(0.7306813597679138, 'performer'), (0.6663976311683655, 'video'), (0.6407639980316162, 'vj'), (0.6342215538024902, 'program'), (0.6087695360183716, 'appears')]\n",
      "ibiza []\n",
      "people [(0.6680805683135986, 'group'), (0.6508774161338806, 'diverse'), (0.6133853793144226, 'friends')]\n",
      "school [(0.736855685710907, 'elementary'), (0.7273381352424622, 'university'), (0.7137252688407898, 'students'), (0.6949441432952881, 'classroom'), (0.6763275265693665, 'teacher'), (0.6444289684295654, 'schoolchildren'), (0.642188549041748, 'pupils'), (0.6136353015899658, 'student'), (0.6027256846427917, 'schools'), (0.6016047596931458, 'college')]\n",
      "laptop [(0.8530187010765076, 'computer'), (0.837763786315918, 'tablet'), (0.7415210604667664, 'smartphone'), (0.7195461392402649, 'typing'), (0.7171635627746582, 'verifying'), (0.7094386219978333, 'desk'), (0.6939656138420105, 'businesswoman'), (0.6918912529945374, 'keyboard'), (0.6856878995895386, 'cellphone'), (0.6752386093139648, 'pc')]\n",
      "computer [(0.8530187010765076, 'laptop'), (0.7452898025512695, 'tablet'), (0.7167372703552246, 'monitor'), (0.7093794345855713, 'pc'), (0.7012506723403931, 'screen'), (0.6933917999267578, 'wireless'), (0.6929288506507874, 'smartphone'), (0.6890875697135925, 'keyboard'), (0.6849894523620605, 'typing'), (0.6804766058921814, 'desktop')]\n",
      "sand [(0.7576702833175659, 'sandy'), (0.7253162860870361, 'beach'), (0.7154739499092102, 'sands'), (0.7112962603569031, 'seashore'), (0.6977460384368896, 'dunes'), (0.673485279083252, 'footprints'), (0.6436341404914856, 'starfish'), (0.632897138595581, 'sea'), (0.6079244613647461, 'powdery')]\n",
      "sky [(0.8030616641044617, 'clouds'), (0.755169689655304, 'skies'), (0.7173047065734863, 'cloudless'), (0.7117840051651001, 'cumulus'), (0.6881585717201233, 'cloudy'), (0.6747150421142578, 'horizon'), (0.6646609902381897, 'cloud'), (0.6548331379890442, 'flying'), (0.6292660236358643, 'starry'), (0.623878002166748, 'twilight')]\n",
      "player [(0.7726331949234009, 'football'), (0.7651488780975342, 'team'), (0.729665219783783, 'goalkeeper'), (0.7084017992019653, 'soccer'), (0.7071017026901245, 'players'), (0.6963939666748047, 'match'), (0.6594505310058594, 'teammate'), (0.6285793781280518, 'association'), (0.6211039423942566, 'ball'), (0.6179524660110474, 'midfielder')]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for model_name in neighbors:\n",
    "    print(model_name.split('/')[-1])\n",
    "    for word in words:\n",
    "        print(word, [x for x in neighbors[model_name][word] if x[0] >= 0.6])\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning : `load_model` does not return WordVectorModel or SupervisedModel any more, but a `FastText` object which is very similar.\n"
     ]
    }
   ],
   "source": [
    "selected_model = '../semantic_learning/training_experiments/unsupervised/fasttext_dim-100_epoch-7_loss-hs_nongrams/unsupervised_model.bin'\n",
    "reduced_model = fasttext.util.reduce_model(fasttext.load_model(selected_model), 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning : `load_model` does not return WordVectorModel or SupervisedModel any more, but a `FastText` object which is very similar.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x1152 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = fasttext.load_model(selected_model)\n",
    "plt.figure(figsize=(16, 16))\n",
    "for word in words:\n",
    "    plt.scatter(*reduced_model[word].tolist())\n",
    "    plt.annotate(word, reduced_model[word].tolist())\n",
    "    for confidence, neighbor in model.get_nearest_neighbors(word):\n",
    "        if confidence >= 0.7:\n",
    "            plt.scatter(*reduced_model[neighbor].tolist())\n",
    "            plt.annotate(neighbor, reduced_model[neighbor].tolist())\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0.8521948456764221, 'the'),\n",
       " (0.8002354502677917, 'a'),\n",
       " (0.7639672756195068, 'in'),\n",
       " (0.7474198341369629, 'of'),\n",
       " (0.7459055185317993, 'and'),\n",
       " (0.7275305986404419, 'on'),\n",
       " (0.6877672076225281, 'with'),\n",
       " (0.6597943305969238, 'at'),\n",
       " (0.599723756313324, 'for'),\n",
       " (0.5637527704238892, 'from'),\n",
       " (0.5094008445739746, 'person'),\n",
       " (0.5019296407699585, 'an'),\n",
       " (0.4852863848209381, 'to'),\n",
       " (0.4801138937473297, 'next'),\n",
       " (0.4738605320453644, 's'),\n",
       " (0.4646811783313751, 'near'),\n",
       " (0.45565688610076904, 'by'),\n",
       " (0.45337536931037903, 'alongside'),\n",
       " (0.45287424325942993, 'red'),\n",
       " (0.45203423500061035, 'as')]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.get_nearest_neighbors('</s>', k=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.48 ms ± 2.97 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "model.get_nearest_neighbors(word, k=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "daa06ee645250b8a2dc707d71e9800f0b1cc0a90be827c1c2df03a597f042bff"
  },
  "kernelspec": {
   "display_name": "Python 3.8.12 64-bit ('typing-assistant': conda)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.11"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}