├── Data
    └── spam.csv
├── Notebooks
    └── spam-ham-classifier-sentence-bert-pos.ipynb
├── UI
    ├── Dockerfile
    ├── app.py
    └── requirements.txt
├── docker-compose.yml
├── images
    └── spam.png
├── readme.md
├── requirements.txt
└── service
    ├── Dockerfile
    ├── main.py
    └── requirements.txt


/Data/spam.csv:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/subhasisj/FastAPI-Streamlit-Docker-NLP/cd7706d3d431a09e3f96cd9f22aeadda2d3b2b83/Data/spam.csv


--------------------------------------------------------------------------------
/Notebooks/spam-ham-classifier-sentence-bert-pos.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "metadata": {
  3 |   "language_info": {
  4 |    "codemirror_mode": {
  5 |     "name": "ipython",
  6 |     "version": 3
  7 |    },
  8 |    "file_extension": ".py",
  9 |    "mimetype": "text/x-python",
 10 |    "name": "python",
 11 |    "nbconvert_exporter": "python",
 12 |    "pygments_lexer": "ipython3",
 13 |    "version": "3.6.9-final"
 14 |   },
 15 |   "orig_nbformat": 2,
 16 |   "kernelspec": {
 17 |    "name": "python3",
 18 |    "display_name": "Python 3",
 19 |    "language": "python"
 20 |   }
 21 |  },
 22 |  "nbformat": 4,
 23 |  "nbformat_minor": 2,
 24 |  "cells": [
 25 |   {
 26 |    "cell_type": "code",
 27 |    "execution_count": 98,
 28 |    "metadata": {},
 29 |    "outputs": [],
 30 |    "source": [
 31 |     "import pandas as pd \n",
 32 |     "import numpy as np \n",
 33 |     "import spacy \n",
 34 |     "from sentence_transformers import SentenceTransformer\n",
 35 |     "from tqdm import tqdm\n"
 36 |    ]
 37 |   },
 38 |   {
 39 |    "cell_type": "code",
 40 |    "execution_count": 99,
 41 |    "metadata": {},
 42 |    "outputs": [
 43 |     {
 44 |      "output_type": "execute_result",
 45 |      "data": {
 46 |       "text/plain": [
 47 |        "     v1                                                 v2 Unnamed: 2  \\\n",
 48 |        "0   ham  Go until jurong point, crazy.. Available only ...        NaN   \n",
 49 |        "1   ham                      Ok lar... Joking wif u oni...        NaN   \n",
 50 |        "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...        NaN   \n",
 51 |        "3   ham  U dun say so early hor... U c already then say...        NaN   \n",
 52 |        "4   ham  Nah I don't think he goes to usf, he lives aro...        NaN   \n",
 53 |        "\n",
 54 |        "  Unnamed: 3 Unnamed: 4  \n",
 55 |        "0        NaN        NaN  \n",
 56 |        "1        NaN        NaN  \n",
 57 |        "2        NaN        NaN  \n",
 58 |        "3        NaN        NaN  \n",
 59 |        "4        NaN        NaN  "
 60 |       ],
 61 |       "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>v1</th>\n      <th>v2</th>\n      <th>Unnamed: 2</th>\n      <th>Unnamed: 3</th>\n      <th>Unnamed: 4</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>ham</td>\n      <td>Go until jurong point, crazy.. Available only ...</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>ham</td>\n      <td>Ok lar... Joking wif u oni...</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>spam</td>\n      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>ham</td>\n      <td>U dun say so early hor... U c already then say...</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>ham</td>\n      <td>Nah I don't think he goes to usf, he lives aro...</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
 62 |      },
 63 |      "metadata": {},
 64 |      "execution_count": 99
 65 |     }
 66 |    ],
 67 |    "source": [
 68 |     "df = pd.read_csv('../Data/spam.csv',encoding='latin-1')\n",
 69 |     "df.head()"
 70 |    ]
 71 |   },
 72 |   {
 73 |    "cell_type": "code",
 74 |    "execution_count": 100,
 75 |    "metadata": {},
 76 |    "outputs": [],
 77 |    "source": [
 78 |     "df = df.iloc[:,0:2]"
 79 |    ]
 80 |   },
 81 |   {
 82 |    "cell_type": "code",
 83 |    "execution_count": 101,
 84 |    "metadata": {},
 85 |    "outputs": [
 86 |     {
 87 |      "output_type": "execute_result",
 88 |      "data": {
 89 |       "text/plain": [
 90 |        "  label                                               text\n",
 91 |        "0   ham  Go until jurong point, crazy.. Available only ...\n",
 92 |        "1   ham                      Ok lar... Joking wif u oni...\n",
 93 |        "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
 94 |        "3   ham  U dun say so early hor... U c already then say...\n",
 95 |        "4   ham  Nah I don't think he goes to usf, he lives aro..."
 96 |       ],
 97 |       "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>label</th>\n      <th>text</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>ham</td>\n      <td>Go until jurong point, crazy.. Available only ...</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>ham</td>\n      <td>Ok lar... Joking wif u oni...</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>spam</td>\n      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>ham</td>\n      <td>U dun say so early hor... U c already then say...</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>ham</td>\n      <td>Nah I don't think he goes to usf, he lives aro...</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
 98 |      },
 99 |      "metadata": {},
100 |      "execution_count": 101
101 |     }
102 |    ],
103 |    "source": [
104 |     "df.columns = ['label','text']\n",
105 |     "df.head()"
106 |    ]
107 |   },
108 |   {
109 |    "cell_type": "code",
110 |    "execution_count": 102,
111 |    "metadata": {},
112 |    "outputs": [
113 |     {
114 |      "output_type": "stream",
115 |      "name": "stderr",
116 |      "text": [
117 |       "Processing with spaCy: 100%|██████████| 5572/5572 [00:29<00:00, 186.72it/s]\n"
118 |      ]
119 |     }
120 |    ],
121 |    "source": [
122 |     "nlp = spacy.load('en_core_web_sm')\n",
123 |     "tqdm.pandas(desc='Processing with spaCy')\n",
124 |     "spacy_results = df['text'].progress_map(nlp)"
125 |    ]
126 |   },
127 |   {
128 |    "cell_type": "code",
129 |    "execution_count": 103,
130 |    "metadata": {},
131 |    "outputs": [],
132 |    "source": [
133 |     "# Encode with Sentence Transformers\n",
134 |     "\n",
135 |     "sentence_bert = SentenceTransformer('paraphrase-distilroberta-base-v1')\n",
136 |     "# tqdm.pandas(desc='Applying sentence-bert')\n",
137 |     "# vectors = df['text'].progress_map(model.encode)\n"
138 |    ]
139 |   },
140 |   {
141 |    "cell_type": "code",
142 |    "execution_count": 104,
143 |    "metadata": {},
144 |    "outputs": [
145 |     {
146 |      "output_type": "display_data",
147 |      "data": {
148 |       "text/plain": "Pandas Apply:   0%|          | 0/5572 [00:00<?, ?it/s]",
149 |       "application/vnd.jupyter.widget-view+json": {
150 |        "version_major": 2,
151 |        "version_minor": 0,
152 |        "model_id": "b7839732dc5f4e9f843f8325238729bc"
153 |       }
154 |      },
155 |      "metadata": {}
156 |     },
157 |     {
158 |      "output_type": "stream",
159 |      "name": "stdout",
160 |      "text": [
161 |       "CPU times: user 12min 25s, sys: 6.08 s, total: 12min 31s\nWall time: 2min 36s\n"
162 |      ]
163 |     }
164 |    ],
165 |    "source": [
166 |     "import swifter\n",
167 |     "\n",
168 |     "%time vectors_swifter = df['text'].swifter.apply(sentence_bert.encode)"
169 |    ]
170 |   },
171 |   {
172 |    "cell_type": "code",
173 |    "execution_count": 105,
174 |    "metadata": {},
175 |    "outputs": [
176 |     {
177 |      "output_type": "display_data",
178 |      "data": {
179 |       "text/plain": "Pandas Apply:   0%|          | 0/5572 [00:00<?, ?it/s]",
180 |       "application/vnd.jupyter.widget-view+json": {
181 |        "version_major": 2,
182 |        "version_minor": 0,
183 |        "model_id": "3ed475b895b34021b0a7f087117e9bac"
184 |       }
185 |      },
186 |      "metadata": {}
187 |     }
188 |    ],
189 |    "source": [
190 |     "df['raw_spacy'] = spacy_results\n",
191 |     "df['raw_pos'] = df['raw_spacy'].swifter.apply(lambda x: ' '.join([t.pos_ for t in x]))"
192 |    ]
193 |   },
194 |   {
195 |    "cell_type": "code",
196 |    "execution_count": 106,
197 |    "metadata": {},
198 |    "outputs": [
199 |     {
200 |      "output_type": "execute_result",
201 |      "data": {
202 |       "text/plain": [
203 |        "  label                                               text  \\\n",
204 |        "0   ham  Go until jurong point, crazy.. Available only ...   \n",
205 |        "1   ham                      Ok lar... Joking wif u oni...   \n",
206 |        "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...   \n",
207 |        "3   ham  U dun say so early hor... U c already then say...   \n",
208 |        "4   ham  Nah I don't think he goes to usf, he lives aro...   \n",
209 |        "\n",
210 |        "                                           raw_spacy  \\\n",
211 |        "0  (Go, until, jurong, point, ,, crazy, .., Avail...   \n",
212 |        "1           (Ok, lar, ..., Joking, wif, u, oni, ...)   \n",
213 |        "2  (Free, entry, in, 2, a, wkly, comp, to, win, F...   \n",
214 |        "3  (U, dun, say, so, early, hor, ..., U, c, alrea...   \n",
215 |        "4  (Nah, I, do, n't, think, he, goes, to, usf, ,,...   \n",
216 |        "\n",
217 |        "                                             raw_pos  \\\n",
218 |        "0  VERB ADP ADJ NOUN PUNCT ADJ PUNCT ADJ ADV ADP ...   \n",
219 |        "1            INTJ ADJ PUNCT NOUN VERB NOUN ADV PUNCT   \n",
220 |        "2  ADJ NOUN ADP NUM DET ADJ NOUN PART VERB PROPN ...   \n",
221 |        "3  NOUN NOUN VERB ADV ADJ NOUN PUNCT NOUN AUX ADV...   \n",
222 |        "4  PROPN PRON AUX PART VERB PRON VERB ADP NOUN PU...   \n",
223 |        "\n",
224 |        "                                       sentence-bert  \n",
225 |        "0  [0.076579936, -0.3930265, 0.27844715, 0.371942...  \n",
226 |        "1  [0.022812596, 0.17678502, 0.12619068, -0.65074...  \n",
227 |        "2  [0.15409197, 0.06857502, -0.13811308, -0.40663...  \n",
228 |        "3  [0.09308915, -0.12710004, -0.033977684, -0.630...  \n",
229 |        "4  [-0.036661543, 0.19233567, -0.27760535, 0.3999...  "
230 |       ],
231 |       "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>label</th>\n      <th>text</th>\n      <th>raw_spacy</th>\n      <th>raw_pos</th>\n      <th>sentence-bert</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>ham</td>\n      <td>Go until jurong point, crazy.. Available only ...</td>\n      <td>(Go, until, jurong, point, ,, crazy, .., Avail...</td>\n      <td>VERB ADP ADJ NOUN PUNCT ADJ PUNCT ADJ ADV ADP ...</td>\n      <td>[0.076579936, -0.3930265, 0.27844715, 0.371942...</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>ham</td>\n      <td>Ok lar... Joking wif u oni...</td>\n      <td>(Ok, lar, ..., Joking, wif, u, oni, ...)</td>\n      <td>INTJ ADJ PUNCT NOUN VERB NOUN ADV PUNCT</td>\n      <td>[0.022812596, 0.17678502, 0.12619068, -0.65074...</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>spam</td>\n      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n      <td>(Free, entry, in, 2, a, wkly, comp, to, win, F...</td>\n      <td>ADJ NOUN ADP NUM DET ADJ NOUN PART VERB PROPN ...</td>\n      <td>[0.15409197, 0.06857502, -0.13811308, -0.40663...</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>ham</td>\n      <td>U dun say so early hor... U c already then say...</td>\n      <td>(U, dun, say, so, early, hor, ..., U, c, alrea...</td>\n      <td>NOUN NOUN VERB ADV ADJ NOUN PUNCT NOUN AUX ADV...</td>\n      <td>[0.09308915, -0.12710004, -0.033977684, -0.630...</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>ham</td>\n      <td>Nah I don't think he goes to usf, he lives aro...</td>\n      <td>(Nah, I, do, n't, think, he, goes, to, usf, ,,...</td>\n      <td>PROPN PRON AUX PART VERB PRON VERB ADP NOUN PU...</td>\n      <td>[-0.036661543, 0.19233567, -0.27760535, 0.3999...</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
232 |      },
233 |      "metadata": {},
234 |      "execution_count": 106
235 |     }
236 |    ],
237 |    "source": [
238 |     "df['sentence-bert'] = vectors_swifter\n",
239 |     "df.head()"
240 |    ]
241 |   },
242 |   {
243 |    "cell_type": "code",
244 |    "execution_count": 107,
245 |    "metadata": {},
246 |    "outputs": [
247 |     {
248 |      "output_type": "display_data",
249 |      "data": {
250 |       "text/plain": "Pandas Apply:   0%|          | 0/5572 [00:00<?, ?it/s]",
251 |       "application/vnd.jupyter.widget-view+json": {
252 |        "version_major": 2,
253 |        "version_minor": 0,
254 |        "model_id": "799e80b9326b48ab839658a3d009e915"
255 |       }
256 |      },
257 |      "metadata": {}
258 |     },
259 |     {
260 |      "output_type": "execute_result",
261 |      "data": {
262 |       "text/plain": [
263 |        "   label                                               text  \\\n",
264 |        "0      0  Go until jurong point, crazy.. Available only ...   \n",
265 |        "1      0                      Ok lar... Joking wif u oni...   \n",
266 |        "2      1  Free entry in 2 a wkly comp to win FA Cup fina...   \n",
267 |        "3      0  U dun say so early hor... U c already then say...   \n",
268 |        "4      0  Nah I don't think he goes to usf, he lives aro...   \n",
269 |        "\n",
270 |        "                                           raw_spacy  \\\n",
271 |        "0  (Go, until, jurong, point, ,, crazy, .., Avail...   \n",
272 |        "1           (Ok, lar, ..., Joking, wif, u, oni, ...)   \n",
273 |        "2  (Free, entry, in, 2, a, wkly, comp, to, win, F...   \n",
274 |        "3  (U, dun, say, so, early, hor, ..., U, c, alrea...   \n",
275 |        "4  (Nah, I, do, n't, think, he, goes, to, usf, ,,...   \n",
276 |        "\n",
277 |        "                                             raw_pos  \\\n",
278 |        "0  VERB ADP ADJ NOUN PUNCT ADJ PUNCT ADJ ADV ADP ...   \n",
279 |        "1            INTJ ADJ PUNCT NOUN VERB NOUN ADV PUNCT   \n",
280 |        "2  ADJ NOUN ADP NUM DET ADJ NOUN PART VERB PROPN ...   \n",
281 |        "3  NOUN NOUN VERB ADV ADJ NOUN PUNCT NOUN AUX ADV...   \n",
282 |        "4  PROPN PRON AUX PART VERB PRON VERB ADP NOUN PU...   \n",
283 |        "\n",
284 |        "                                       sentence-bert  \n",
285 |        "0  [0.076579936, -0.3930265, 0.27844715, 0.371942...  \n",
286 |        "1  [0.022812596, 0.17678502, 0.12619068, -0.65074...  \n",
287 |        "2  [0.15409197, 0.06857502, -0.13811308, -0.40663...  \n",
288 |        "3  [0.09308915, -0.12710004, -0.033977684, -0.630...  \n",
289 |        "4  [-0.036661543, 0.19233567, -0.27760535, 0.3999...  "
290 |       ],
291 |       "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>label</th>\n      <th>text</th>\n      <th>raw_spacy</th>\n      <th>raw_pos</th>\n      <th>sentence-bert</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>Go until jurong point, crazy.. Available only ...</td>\n      <td>(Go, until, jurong, point, ,, crazy, .., Avail...</td>\n      <td>VERB ADP ADJ NOUN PUNCT ADJ PUNCT ADJ ADV ADP ...</td>\n      <td>[0.076579936, -0.3930265, 0.27844715, 0.371942...</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0</td>\n      <td>Ok lar... Joking wif u oni...</td>\n      <td>(Ok, lar, ..., Joking, wif, u, oni, ...)</td>\n      <td>INTJ ADJ PUNCT NOUN VERB NOUN ADV PUNCT</td>\n      <td>[0.022812596, 0.17678502, 0.12619068, -0.65074...</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>1</td>\n      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n      <td>(Free, entry, in, 2, a, wkly, comp, to, win, F...</td>\n      <td>ADJ NOUN ADP NUM DET ADJ NOUN PART VERB PROPN ...</td>\n      <td>[0.15409197, 0.06857502, -0.13811308, -0.40663...</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0</td>\n      <td>U dun say so early hor... U c already then say...</td>\n      <td>(U, dun, say, so, early, hor, ..., U, c, alrea...</td>\n      <td>NOUN NOUN VERB ADV ADJ NOUN PUNCT NOUN AUX ADV...</td>\n      <td>[0.09308915, -0.12710004, -0.033977684, -0.630...</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0</td>\n      <td>Nah I don't think he goes to usf, he lives aro...</td>\n      <td>(Nah, I, do, n't, think, he, goes, to, usf, ,,...</td>\n      <td>PROPN PRON AUX PART VERB PRON VERB ADP NOUN PU...</td>\n      <td>[-0.036661543, 0.19233567, -0.27760535, 0.3999...</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
292 |      },
293 |      "metadata": {},
294 |      "execution_count": 107
295 |     }
296 |    ],
297 |    "source": [
298 |     "df['label'] = df.label.swifter.apply(lambda x : 1 if x =='spam' else 0)\n",
299 |     "df.head()"
300 |    ]
301 |   },
302 |   {
303 |    "cell_type": "code",
304 |    "execution_count": 108,
305 |    "metadata": {},
306 |    "outputs": [],
307 |    "source": [
308 |     "df.drop(columns=['raw_spacy'],inplace=True)"
309 |    ]
310 |   },
311 |   {
312 |    "cell_type": "code",
313 |    "execution_count": 171,
314 |    "metadata": {},
315 |    "outputs": [],
316 |    "source": [
317 |     "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
318 |     "from sklearn.compose import ColumnTransformer\n",
319 |     "from sklearn.preprocessing import FunctionTransformer\n",
320 |     "from sklearn.model_selection import train_test_split\n",
321 |     "from sklearn import metrics\n",
322 |     "from sklearn.pipeline import Pipeline\n",
323 |     "from sklearn.linear_model import LogisticRegression\n",
324 |     "from xgboost import XGBClassifier\n",
325 |     "\n"
326 |    ]
327 |   },
328 |   {
329 |    "cell_type": "code",
330 |    "execution_count": 289,
331 |    "metadata": {},
332 |    "outputs": [],
333 |    "source": [
334 |     "train_df = df.copy()"
335 |    ]
336 |   },
337 |   {
338 |    "cell_type": "code",
339 |    "execution_count": 290,
340 |    "metadata": {},
341 |    "outputs": [],
342 |    "source": [
343 |     "def stack_embeddings(embeddings):\n",
344 |     "    import numpy as np\n",
345 |     "    return np.vstack(embeddings.values)\n",
346 |     "\n",
347 |     "ct = ColumnTransformer([\n",
348 |     "    ('bag of ngrams', TfidfVectorizer(ngram_range=(1, 2), max_features=3000), 'text'),\n",
349 |     "    ('bag of POS', CountVectorizer(ngram_range=(1, 2)), 'raw_pos'),\n",
350 |     "    # Lambda functions cannot be pickled\n",
351 |     "    ('sentence bert', FunctionTransformer(stack_embeddings), 'sentence-bert'),\n",
352 |     "    # ('bag of NER types', CountVectorizer(ngram_range=(1, 2)), 'raw_ner'),\n",
353 |     "    # ('ngrams before', TfidfVectorizer(ngram_range=(1, 2), max_features=3000), 'raw_before'),\n",
354 |     "    # ('ngrams after', TfidfVectorizer(ngram_range=(1, 2), max_features=3000), 'raw_after')    \n",
355 |     "],remainder='passthrough')\n",
356 |     "\n",
357 |     "# lm = LogisticRegression()\n",
358 |     "xgb = XGBClassifier(random_state=0)\n",
359 |     "\n",
360 |     "# pipeline = Pipeline([('transformer', ct), ('classifier', lm)])\n",
361 |     "pipeline = Pipeline([('transformer', ct), ('classifier', xgb)])\n",
362 |     "\n",
363 |     "\n",
364 |     "y,X = train_df.pop('label'),train_df\n",
365 |     "\n",
366 |     "X_train, X_test, y_train, y_test = train_test_split(X,y, test_size=0.3, random_state=42,stratify=y)"
367 |    ]
368 |   },
369 |   {
370 |    "cell_type": "code",
371 |    "execution_count": 291,
372 |    "metadata": {},
373 |    "outputs": [
374 |     {
375 |      "output_type": "stream",
376 |      "name": "stdout",
377 |      "text": [
378 |       "[15:13:14] WARNING: ../src/learner.cc:1061: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
379 |       "CPU times: user 53.3 s, sys: 354 ms, total: 53.7 s\n",
380 |       "Wall time: 5.52 s\n"
381 |      ]
382 |     }
383 |    ],
384 |    "source": [
385 |     "\n",
386 |     "\n",
387 |     "%time model = pipeline.fit(X_train, y_train)"
388 |    ]
389 |   },
390 |   {
391 |    "cell_type": "code",
392 |    "execution_count": 292,
393 |    "metadata": {},
394 |    "outputs": [],
395 |    "source": [
396 |     "y_pred = model.predict(X_test)"
397 |    ]
398 |   },
399 |   {
400 |    "cell_type": "code",
401 |    "execution_count": 293,
402 |    "metadata": {},
403 |    "outputs": [
404 |     {
405 |      "output_type": "stream",
406 |      "name": "stdout",
407 |      "text": [
408 |       "              precision    recall  f1-score   support\n\n           0       0.99      1.00      0.99      1448\n           1       0.99      0.91      0.95       224\n\n    accuracy                           0.99      1672\n   macro avg       0.99      0.95      0.97      1672\nweighted avg       0.99      0.99      0.99      1672\n\n"
409 |      ]
410 |     }
411 |    ],
412 |    "source": [
413 |     "print(metrics.classification_report(y_test,y_pred))"
414 |    ]
415 |   },
416 |   {
417 |    "cell_type": "code",
418 |    "execution_count": 294,
419 |    "metadata": {},
420 |    "outputs": [],
421 |    "source": [
422 |     "# import joblib\n",
423 |     "# filename = 'model.sav'\n",
424 |     "# joblib.dump(model, filename)\n",
425 |     "import dill\n",
426 |     "\n",
427 |     "\n",
428 |     "pkl_filename = \"../Models/model.pkl\"\n",
429 |     "with open(pkl_filename, 'wb') as file:\n",
430 |     "    dill.dump(model, file)"
431 |    ]
432 |   },
433 |   {
434 |    "cell_type": "code",
435 |    "execution_count": 295,
436 |    "metadata": {},
437 |    "outputs": [
438 |     {
439 |      "output_type": "stream",
440 |      "name": "stdout",
441 |      "text": [
442 |       "\u001b[0m\u001b[01;32mmodel.pkl\u001b[0m*\n"
443 |      ]
444 |     }
445 |    ],
446 |    "source": [
447 |     "ls ../Models"
448 |    ]
449 |   },
450 |   {
451 |    "cell_type": "code",
452 |    "execution_count": 296,
453 |    "metadata": {},
454 |    "outputs": [],
455 |    "source": [
456 |     "with open(pkl_filename,'rb') as file:\n",
457 |     "    loaded_model = dill.load(file)"
458 |    ]
459 |   },
460 |   {
461 |    "cell_type": "code",
462 |    "execution_count": 297,
463 |    "metadata": {},
464 |    "outputs": [],
465 |    "source": [
466 |     "def make_inference_df(input_text):\n",
467 |     "\n",
468 |     "    model_input_dict = {}\n",
469 |     "    input_row_list = []\n",
470 |     "    \n",
471 |     "\n",
472 |     "    spacy_raw = nlp(input_text)\n",
473 |     "    # pos_tags = [t.pos_ for t in spacy_raw]\n",
474 |     "\n",
475 |     "    model_input_dict['text'] = input_text\n",
476 |     "    model_input_dict['raw_pos'] =  ' '.join([t.pos_ for t in spacy_raw])\n",
477 |     "    model_input_dict['sentence-bert'] = sentence_bert.encode(input_text)\n",
478 |     "\n",
479 |     "    input_row_list.append(model_input_dict)\n",
480 |     "\n",
481 |     "    model_input_df = pd.DataFrame(input_row_list)\n",
482 |     "    return model_input_df"
483 |    ]
484 |   },
485 |   {
486 |    "cell_type": "code",
487 |    "execution_count": 298,
488 |    "metadata": {},
489 |    "outputs": [
490 |     {
491 |      "output_type": "execute_result",
492 |      "data": {
493 |       "text/plain": [
494 |        "array([1])"
495 |       ]
496 |      },
497 |      "metadata": {},
498 |      "execution_count": 298
499 |     }
500 |    ],
501 |    "source": [
502 |     "sample_text = 'Free entry in 2 a wkly comp to win FA Cup final tkts 21st May 2005. Text FA to 87121 to receive entry question(std txt rate)T&C\\'s'\n",
503 |     "# make_inference_df(sample_text)\n",
504 |     "loaded_model.predict(make_inference_df(sample_text))"
505 |    ]
506 |   },
507 |   {
508 |    "cell_type": "code",
509 |    "execution_count": 299,
510 |    "metadata": {},
511 |    "outputs": [
512 |     {
513 |      "output_type": "stream",
514 |      "name": "stdout",
515 |      "text": [
516 |       "Nah I don't think he goes to usf, he lives around here though\n"
517 |      ]
518 |     },
519 |     {
520 |      "output_type": "execute_result",
521 |      "data": {
522 |       "text/plain": [
523 |        "array([1])"
524 |       ]
525 |      },
526 |      "metadata": {},
527 |      "execution_count": 299
528 |     }
529 |    ],
530 |    "source": [
531 |     "sample_text_2 = 'Nah I don\\'t think he goes to usf, he lives around here though'\n",
532 |     "print(sample_text_2)\n",
533 |     "loaded_model.predict(make_inference_df(sample_text))"
534 |    ]
535 |   },
536 |   {
537 |    "cell_type": "code",
538 |    "execution_count": null,
539 |    "metadata": {},
540 |    "outputs": [],
541 |    "source": []
542 |   },
543 |   {
544 |    "cell_type": "code",
545 |    "execution_count": null,
546 |    "metadata": {},
547 |    "outputs": [],
548 |    "source": []
549 |   },
550 |   {
551 |    "cell_type": "code",
552 |    "execution_count": null,
553 |    "metadata": {},
554 |    "outputs": [],
555 |    "source": []
556 |   }
557 |  ]
558 | }


--------------------------------------------------------------------------------
/UI/Dockerfile:
--------------------------------------------------------------------------------
 1 | FROM python:3.8-slim
 2 | 
 3 | WORKDIR /app
 4 | 
 5 | COPY requirements.txt .
 6 | RUN pip install -r requirements.txt
 7 | 
 8 | # COPY . .
 9 | 
10 | EXPOSE 8501
11 | 
12 | CMD ["streamlit", "run", "app.py"]


--------------------------------------------------------------------------------
/UI/app.py:
--------------------------------------------------------------------------------
 1 | import streamlit as st
 2 | import requests
 3 | 
 4 | 
 5 | def main():
 6 | 
 7 |     st.title("Spam Classification")
 8 |     message = st.text_input('Enter Text to Classify')
 9 | 
10 |     if st.button('Predict'):
11 |         payload = {
12 |             "text": message
13 |         }
14 |         res = requests.post(f"http://service:8000/predict/",json=payload )
15 |         with st.spinner('Classifying, please wait....'):
16 |             st.write(res.json())
17 | 
18 | 
19 | 
20 | 
21 | if __name__ == '__main__':
22 |     main()


--------------------------------------------------------------------------------
/UI/requirements.txt:
--------------------------------------------------------------------------------
  1 | aiohttp==3.7.3
  2 | aiohttp-cors==0.7.0
  3 | aioredis==1.3.1
  4 | altair==4.1.0
  5 | argon2-cffi==20.1.0
  6 | astor==0.8.1
  7 | async-generator==1.10
  8 | async-timeout==3.0.1
  9 | attrs==20.3.0
 10 | backcall==0.2.0
 11 | base58==2.1.0
 12 | bleach==3.3.0
 13 | blessings==1.7
 14 | blinker==1.4
 15 | blis==0.7.4
 16 | cachetools==4.2.1
 17 | catalogue==2.0.1
 18 | certifi==2020.12.5
 19 | cffi==1.14.4
 20 | chardet==3.0.4
 21 | click==7.1.2
 22 | colorama==0.4.4
 23 | colorful==0.5.4
 24 | contextvars==2.4
 25 | cymem==2.0.5
 26 | # dask==2021.2.0
 27 | # dataclasses==0.8
 28 | decorator==4.4.2
 29 | defusedxml==0.6.0
 30 | dill==0.3.3
 31 | en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.0.0/en_core_web_sm-3.0.0-py3-none-any.whl
 32 | entrypoints==0.3
 33 | fastapi==0.63.0
 34 | filelock==3.0.12
 35 | fsspec==0.8.5
 36 | gitdb==4.0.5
 37 | GitPython==3.1.13
 38 | google-api-core==1.26.0
 39 | google-auth==1.26.0
 40 | googleapis-common-protos==1.52.0
 41 | gpustat==0.6.0
 42 | grpcio==1.35.0
 43 | h11==0.12.0
 44 | hiredis==1.1.0
 45 | idna==2.10
 46 | idna-ssl==1.1.0
 47 | immutables==0.15
 48 | importlib-metadata==3.4.0
 49 | ipykernel==5.4.3
 50 | ipython==7.16.1
 51 | ipython-genutils==0.2.0
 52 | ipywidgets==7.6.3
 53 | jedi==0.18.0
 54 | Jinja2==2.11.3
 55 | joblib==1.0.1
 56 | jsonschema==3.2.0
 57 | jupyter-client==6.1.11
 58 | jupyter-core==4.7.1
 59 | jupyterlab-pygments==0.1.2
 60 | jupyterlab-widgets==1.0.0
 61 | locket==0.2.1
 62 | MarkupSafe==1.1.1
 63 | mistune==0.8.4
 64 | modin==0.8.3
 65 | msgpack==1.0.2
 66 | multidict==5.1.0
 67 | murmurhash==1.0.5
 68 | nbclient==0.5.2
 69 | nbconvert==6.0.7
 70 | nbformat==5.1.2
 71 | nest-asyncio==1.5.1
 72 | nltk==3.5
 73 | notebook==6.2.0
 74 | numpy==1.19.5
 75 | nvidia-ml-py3==7.352.0
 76 | opencensus==0.7.12
 77 | opencensus-context==0.1.2
 78 | packaging==20.9
 79 | pandarallel==1.5.2
 80 | pandas==1.1.5
 81 | pandocfilters==1.4.3
 82 | parso==0.8.1
 83 | partd==1.1.0
 84 | pathy==0.3.6
 85 | pexpect==4.8.0
 86 | pickleshare==0.7.5
 87 | Pillow==8.1.0
 88 | preshed==3.0.5
 89 | prometheus-client==0.9.0
 90 | prompt-toolkit==3.0.15
 91 | protobuf==3.14.0
 92 | psutil==5.8.0
 93 | ptyprocess==0.7.0
 94 | py-spy==0.3.4
 95 | pyarrow==1.0.0
 96 | pyasn1==0.4.8
 97 | pyasn1-modules==0.2.8
 98 | pycparser==2.20
 99 | pydantic==1.7.3
100 | pydeck==0.6.0
101 | Pygments==2.7.4
102 | pyparsing==2.4.7
103 | pyrsistent==0.17.3
104 | python-dateutil==2.8.1
105 | pytz==2021.1
106 | PyYAML==5.4.1
107 | pyzmq==22.0.2
108 | ray==1.1.0
109 | redis==3.5.3
110 | regex==2020.11.13
111 | requests==2.25.1
112 | rsa==4.7
113 | sacremoses==0.0.43
114 | scikit-learn==0.24.1
115 | scipy==1.5.4
116 | Send2Trash==1.5.0
117 | sentence-transformers==0.4.1.2
118 | sentencepiece==0.1.95
119 | six==1.15.0
120 | smart-open==3.0.0
121 | smmap==3.0.5
122 | spacy==3.0.1
123 | spacy-legacy==3.0.1
124 | srsly==2.4.0
125 | starlette==0.13.6
126 | streamlit==0.76.0
127 | # swifter==1.0.7
128 | terminado==0.9.2
129 | testpath==0.4.4
130 | thinc==8.0.1
131 | threadpoolctl==2.1.0
132 | tokenizers==0.10.1
133 | toml==0.10.2
134 | toolz==0.11.1
135 | torch==1.7.1
136 | tornado==6.1
137 | tqdm==4.56.2
138 | traitlets==4.3.3
139 | transformers==4.3.2
140 | typer==0.3.2
141 | typing-extensions==3.7.4.3
142 | tzlocal==2.1
143 | urllib3==1.26.3
144 | uvicorn==0.13.3
145 | validators==0.18.2
146 | wasabi==0.8.2
147 | watchdog==2.0.0
148 | wcwidth==0.2.5
149 | webencodings==0.5.1
150 | widgetsnbextension==3.5.1
151 | xgboost==1.3.3
152 | yarl==1.6.3
153 | zipp==3.4.0
154 | 


--------------------------------------------------------------------------------
/docker-compose.yml:
--------------------------------------------------------------------------------
 1 | version: '3'
 2 | 
 3 | services:
 4 |   frontendui:
 5 |     build: UI
 6 |     ports:
 7 |       - 8501:8501
 8 |     volumes:
 9 |       - './UI:/app:delegated'
10 |     depends_on:
11 |       - service
12 |     
13 |   service:
14 |     build: service
15 |     ports:
16 |       - 8000:8000  


--------------------------------------------------------------------------------
/images/spam.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/subhasisj/FastAPI-Streamlit-Docker-NLP/cd7706d3d431a09e3f96cd9f22aeadda2d3b2b83/images/spam.png


--------------------------------------------------------------------------------
/readme.md:
--------------------------------------------------------------------------------
 1 | # Text-Classification with FastAPI / Streamlit / Docker Compose
 2 | 
 3 | This project showcases the use of `FastAPI` and `Streamlit` in tandem.
 4 | The trained model is deployed using a FastAPI rest service containerized using `Docker`.
 5 | The front end UI is built on Streamlit which is hosted on its own Docker container.
 6 | 
 7 | We spin both the containers together using `Docker Compose` .
 8 | 
 9 | 
10 | ![GitHub Logo](/images/spam.png)
11 | 
12 | 
13 | ## How to use
14 | 
15 | Clone this repo and run the below docker command:
16 | 
17 | `To Start Application:`
18 | ```docker
19 | docker-compose up -d --build
20 | ```
21 | and navigate to http://localhost:8501/
22 | 
23 | `To Stop Application:`
24 | ```docker
25 | docker-compose down
26 | ```
27 | 
28 | ### Trivia: 
29 | 
30 | _Using the volume tag in the compose file, we can mount the local folders from your computer to the Docker container. Now you can develop your app while using Docker and save changes._


--------------------------------------------------------------------------------
/requirements.txt:
--------------------------------------------------------------------------------
  1 | aiohttp==3.7.3
  2 | aiohttp-cors==0.7.0
  3 | aioredis==1.3.1
  4 | altair==4.1.0
  5 | argon2-cffi==20.1.0
  6 | astor==0.8.1
  7 | async-generator==1.10
  8 | async-timeout==3.0.1
  9 | attrs==20.3.0
 10 | backcall==0.2.0
 11 | base58==2.1.0
 12 | bleach==3.3.0
 13 | blessings==1.7
 14 | blinker==1.4
 15 | blis==0.7.4
 16 | cachetools==4.2.1
 17 | catalogue==2.0.1
 18 | certifi==2020.12.5
 19 | cffi==1.14.4
 20 | chardet==3.0.4
 21 | click==7.1.2
 22 | colorama==0.4.4
 23 | colorful==0.5.4
 24 | contextvars==2.4
 25 | cymem==2.0.5
 26 | dask==2021.2.0
 27 | dataclasses==0.8
 28 | decorator==4.4.2
 29 | defusedxml==0.6.0
 30 | dill==0.3.3
 31 | en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.0.0/en_core_web_sm-3.0.0-py3-none-any.whl
 32 | entrypoints==0.3
 33 | fastapi==0.63.0
 34 | filelock==3.0.12
 35 | fsspec==0.8.5
 36 | gitdb==4.0.5
 37 | GitPython==3.1.13
 38 | google-api-core==1.26.0
 39 | google-auth==1.26.0
 40 | googleapis-common-protos==1.52.0
 41 | gpustat==0.6.0
 42 | grpcio==1.35.0
 43 | h11==0.12.0
 44 | hiredis==1.1.0
 45 | idna==2.10
 46 | idna-ssl==1.1.0
 47 | immutables==0.15
 48 | importlib-metadata==3.4.0
 49 | ipykernel==5.4.3
 50 | ipython==7.16.1
 51 | ipython-genutils==0.2.0
 52 | ipywidgets==7.6.3
 53 | jedi==0.18.0
 54 | Jinja2==2.11.3
 55 | joblib==1.0.1
 56 | jsonschema==3.2.0
 57 | jupyter-client==6.1.11
 58 | jupyter-core==4.7.1
 59 | jupyterlab-pygments==0.1.2
 60 | jupyterlab-widgets==1.0.0
 61 | locket==0.2.1
 62 | MarkupSafe==1.1.1
 63 | mistune==0.8.4
 64 | modin==0.8.3
 65 | msgpack==1.0.2
 66 | multidict==5.1.0
 67 | murmurhash==1.0.5
 68 | nbclient==0.5.2
 69 | nbconvert==6.0.7
 70 | nbformat==5.1.2
 71 | nest-asyncio==1.5.1
 72 | nltk==3.5
 73 | notebook==6.2.0
 74 | numpy==1.19.5
 75 | nvidia-ml-py3==7.352.0
 76 | opencensus==0.7.12
 77 | opencensus-context==0.1.2
 78 | packaging==20.9
 79 | pandarallel==1.5.2
 80 | pandas==1.1.5
 81 | pandocfilters==1.4.3
 82 | parso==0.8.1
 83 | partd==1.1.0
 84 | pathy==0.3.6
 85 | pexpect==4.8.0
 86 | pickleshare==0.7.5
 87 | Pillow==8.1.0
 88 | preshed==3.0.5
 89 | prometheus-client==0.9.0
 90 | prompt-toolkit==3.0.15
 91 | protobuf==3.14.0
 92 | psutil==5.8.0
 93 | ptyprocess==0.7.0
 94 | py-spy==0.3.4
 95 | pyarrow==1.0.0
 96 | pyasn1==0.4.8
 97 | pyasn1-modules==0.2.8
 98 | pycparser==2.20
 99 | pydantic==1.7.3
100 | pydeck==0.6.0
101 | Pygments==2.7.4
102 | pyparsing==2.4.7
103 | pyrsistent==0.17.3
104 | python-dateutil==2.8.1
105 | pytz==2021.1
106 | PyYAML==5.4.1
107 | pyzmq==22.0.2
108 | ray==1.1.0
109 | redis==3.5.3
110 | regex==2020.11.13
111 | requests==2.25.1
112 | rsa==4.7
113 | sacremoses==0.0.43
114 | scikit-learn==0.24.1
115 | scipy==1.5.4
116 | Send2Trash==1.5.0
117 | sentence-transformers==0.4.1.2
118 | sentencepiece==0.1.95
119 | six==1.15.0
120 | smart-open==3.0.0
121 | smmap==3.0.5
122 | spacy==3.0.1
123 | spacy-legacy==3.0.1
124 | srsly==2.4.0
125 | starlette==0.13.6
126 | streamlit==0.76.0
127 | swifter==1.0.7
128 | terminado==0.9.2
129 | testpath==0.4.4
130 | thinc==8.0.1
131 | threadpoolctl==2.1.0
132 | tokenizers==0.10.1
133 | toml==0.10.2
134 | toolz==0.11.1
135 | torch==1.7.1
136 | tornado==6.1
137 | tqdm==4.56.2
138 | traitlets==4.3.3
139 | transformers==4.3.2
140 | typer==0.3.2
141 | typing-extensions==3.7.4.3
142 | tzlocal==2.1
143 | urllib3==1.26.3
144 | uvicorn==0.13.3
145 | validators==0.18.2
146 | wasabi==0.8.2
147 | watchdog==2.0.0
148 | wcwidth==0.2.5
149 | webencodings==0.5.1
150 | widgetsnbextension==3.5.1
151 | xgboost==1.3.3
152 | yarl==1.6.3
153 | zipp==3.4.0
154 | 


--------------------------------------------------------------------------------
/service/Dockerfile:
--------------------------------------------------------------------------------
 1 | FROM python:3.8-slim
 2 | 
 3 | WORKDIR /app
 4 | 
 5 | RUN apt-get update
 6 | RUN apt-get install \
 7 |     'ffmpeg'\
 8 |     'libsm6'\
 9 |     'libxext6'  -y
10 | 
11 | COPY ./requirements.txt .
12 | RUN pip install -r requirements.txt
13 | 
14 | COPY . .
15 | 
16 | EXPOSE 8000
17 | 
18 | CMD ["python", "main.py"]


--------------------------------------------------------------------------------
/service/main.py:
--------------------------------------------------------------------------------
 1 | # https://github.com/kumarvc/Fastapi-docker-kubernetes-streamlit/blob/master/app/main.py
 2 | 
 3 | # https://medium.com/analytics-vidhya/deploying-a-nlp-model-with-docker-and-fastapi-d972779d8008
 4 | 
 5 | # https://www.analyticsvidhya.com/blog/2020/12/streamlit-web-api-for-nlp-tweet-sentiment-analysis/
 6 | 
 7 | # https://testdriven.io/blog/fastapi-streamlit/#docker-compose
 8 | 
 9 | from fastapi import FastAPI
10 | import dill
11 | import os
12 | import numpy as np
13 | import uvicorn
14 | from pydantic import BaseModel
15 | import pandas as pd
16 | import spacy 
17 | from sentence_transformers import SentenceTransformer
18 | 
19 | 
20 | PKL_FILENAME = "model.pkl"
21 | MODELS_PATH = "./Models/"
22 | MODEL_FILE_PATH = os.path.join(MODELS_PATH,PKL_FILENAME)
23 | SENTENCE_BERT = SentenceTransformer('paraphrase-distilroberta-base-v1')
24 | nlp = spacy.load('en_core_web_sm')
25 | 
26 | 
27 | def load_model():
28 |     with open(MODEL_FILE_PATH,'rb') as file:
29 |         return dill.load(file)
30 | 
31 | LOADED_MODEL = load_model()
32 | 
33 | app = FastAPI(title="Ham or Spam API", description="API to predict if a SMS is ham or spam")
34 | 
35 | 
36 | class Data(BaseModel):
37 |     text:str
38 | 
39 | 
40 | 
41 | def make_inference_df(input_text):
42 | 
43 |     model_input_dict = {}
44 |     input_row_list = []
45 |     
46 | 
47 |     spacy_raw = nlp(input_text)
48 |     # pos_tags = [t.pos_ for t in spacy_raw]
49 | 
50 |     model_input_dict['text'] = input_text
51 |     model_input_dict['raw_pos'] =  ' '.join([t.pos_ for t in spacy_raw])
52 |     model_input_dict['sentence-bert'] = SENTENCE_BERT.encode(input_text)
53 | 
54 |     input_row_list.append(model_input_dict)
55 | 
56 |     model_input_df = pd.DataFrame(input_row_list)
57 |     return model_input_df
58 | 
59 | @app.get("/")
60 | def read_root():
61 |     return {"message": "Welcome from the API"}
62 | 
63 | @app.post("/predict")
64 | def predict(data:Data):
65 | 
66 |     model_input_df = make_inference_df(data.text)
67 |     prediction = LOADED_MODEL.predict_proba(model_input_df)
68 |     print(data.text)
69 |     ham_probability = 1 - prediction[:,1][0]
70 |     spam_probability = prediction[:,1][0]
71 |     print(spam_probability)
72 |     return {
73 |         
74 |         "Text" : data.text,
75 |         "prediction": "SPAM" if  prediction[:,1][0] > 0.5 else "HAM",
76 |         "SPAM" : float(spam_probability),
77 |         "HAM" : float(ham_probability),
78 | 
79 |     }
80 | 
81 | 
82 | if __name__ == '__main__':
83 |     uvicorn.run("main:app", host="0.0.0.0", port=8000, reload=True)


--------------------------------------------------------------------------------
/service/requirements.txt:
--------------------------------------------------------------------------------
  1 | aiohttp==3.7.3
  2 | aiohttp-cors==0.7.0
  3 | aioredis==1.3.1
  4 | altair==4.1.0
  5 | argon2-cffi==20.1.0
  6 | astor==0.8.1
  7 | async-generator==1.10
  8 | async-timeout==3.0.1
  9 | attrs==20.3.0
 10 | backcall==0.2.0
 11 | base58==2.1.0
 12 | bleach==3.3.0
 13 | blessings==1.7
 14 | blinker==1.4
 15 | blis==0.7.4
 16 | cachetools==4.2.1
 17 | catalogue==2.0.1
 18 | certifi==2020.12.5
 19 | cffi==1.14.4
 20 | chardet==3.0.4
 21 | click==7.1.2
 22 | colorama==0.4.4
 23 | colorful==0.5.4
 24 | contextvars==2.4
 25 | cymem==2.0.5
 26 | # dask==2021.2.0
 27 | # dataclasses==0.8
 28 | decorator==4.4.2
 29 | defusedxml==0.6.0
 30 | dill==0.3.3
 31 | en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.0.0/en_core_web_sm-3.0.0-py3-none-any.whl
 32 | entrypoints==0.3
 33 | fastapi==0.63.0
 34 | filelock==3.0.12
 35 | fsspec==0.8.5
 36 | gitdb==4.0.5
 37 | GitPython==3.1.13
 38 | google-api-core==1.26.0
 39 | google-auth==1.26.0
 40 | googleapis-common-protos==1.52.0
 41 | gpustat==0.6.0
 42 | grpcio==1.35.0
 43 | h11==0.12.0
 44 | hiredis==1.1.0
 45 | idna==2.10
 46 | idna-ssl==1.1.0
 47 | immutables==0.15
 48 | importlib-metadata==3.4.0
 49 | ipykernel==5.4.3
 50 | ipython==7.16.1
 51 | ipython-genutils==0.2.0
 52 | ipywidgets==7.6.3
 53 | jedi==0.18.0
 54 | Jinja2==2.11.3
 55 | joblib==1.0.1
 56 | jsonschema==3.2.0
 57 | jupyter-client==6.1.11
 58 | jupyter-core==4.7.1
 59 | jupyterlab-pygments==0.1.2
 60 | jupyterlab-widgets==1.0.0
 61 | locket==0.2.1
 62 | MarkupSafe==1.1.1
 63 | mistune==0.8.4
 64 | modin==0.8.3
 65 | msgpack==1.0.2
 66 | multidict==5.1.0
 67 | murmurhash==1.0.5
 68 | nbclient==0.5.2
 69 | nbconvert==6.0.7
 70 | nbformat==5.1.2
 71 | nest-asyncio==1.5.1
 72 | nltk==3.5
 73 | notebook==6.2.0
 74 | numpy==1.19.5
 75 | nvidia-ml-py3==7.352.0
 76 | opencensus==0.7.12
 77 | opencensus-context==0.1.2
 78 | packaging==20.9
 79 | pandarallel==1.5.2
 80 | pandas==1.1.5
 81 | pandocfilters==1.4.3
 82 | parso==0.8.1
 83 | partd==1.1.0
 84 | pathy==0.3.6
 85 | pexpect==4.8.0
 86 | pickleshare==0.7.5
 87 | Pillow==8.1.0
 88 | preshed==3.0.5
 89 | prometheus-client==0.9.0
 90 | prompt-toolkit==3.0.15
 91 | protobuf==3.14.0
 92 | psutil==5.8.0
 93 | ptyprocess==0.7.0
 94 | py-spy==0.3.4
 95 | pyarrow==1.0.0
 96 | pyasn1==0.4.8
 97 | pyasn1-modules==0.2.8
 98 | pycparser==2.20
 99 | pydantic==1.7.3
100 | pydeck==0.6.0
101 | Pygments==2.7.4
102 | pyparsing==2.4.7
103 | pyrsistent==0.17.3
104 | python-dateutil==2.8.1
105 | pytz==2021.1
106 | PyYAML==5.4.1
107 | pyzmq==22.0.2
108 | ray==1.1.0
109 | redis==3.5.3
110 | regex==2020.11.13
111 | requests==2.25.1
112 | rsa==4.7
113 | sacremoses==0.0.43
114 | scikit-learn==0.24.1
115 | scipy==1.5.4
116 | Send2Trash==1.5.0
117 | sentence-transformers==0.4.1.2
118 | sentencepiece==0.1.95
119 | six==1.15.0
120 | smart-open==3.0.0
121 | smmap==3.0.5
122 | spacy==3.0.1
123 | spacy-legacy==3.0.1
124 | srsly==2.4.0
125 | starlette==0.13.6
126 | streamlit==0.76.0
127 | # swifter==1.0.7
128 | terminado==0.9.2
129 | testpath==0.4.4
130 | thinc==8.0.1
131 | threadpoolctl==2.1.0
132 | tokenizers==0.10.1
133 | toml==0.10.2
134 | toolz==0.11.1
135 | torch==1.7.1
136 | tornado==6.1
137 | tqdm==4.56.2
138 | traitlets==4.3.3
139 | transformers==4.3.2
140 | typer==0.3.2
141 | typing-extensions==3.7.4.3
142 | tzlocal==2.1
143 | urllib3==1.26.3
144 | uvicorn==0.13.3
145 | validators==0.18.2
146 | wasabi==0.8.2
147 | watchdog==2.0.0
148 | wcwidth==0.2.5
149 | webencodings==0.5.1
150 | widgetsnbextension==3.5.1
151 | xgboost==1.3.3
152 | yarl==1.6.3
153 | zipp==3.4.0
154 | 


--------------------------------------------------------------------------------