Spaces:
Sleeping
Sleeping
Commit
·
ab2f83e
1
Parent(s):
7bdad43
Update app.py
Browse files
app.py
CHANGED
|
@@ -18,8 +18,6 @@ import functions
|
|
| 18 |
torch.manual_seed(1)
|
| 19 |
|
| 20 |
# Preprocess function
|
| 21 |
-
import re
|
| 22 |
-
from contractions import contractions_dict
|
| 23 |
from nltk.corpus import stopwords
|
| 24 |
from nltk.tokenize import word_tokenize
|
| 25 |
|
|
@@ -32,9 +30,6 @@ def preprocess_text(text):
|
|
| 32 |
html_pattern = re.compile(r'<[^<>]+>')
|
| 33 |
text = html_pattern.sub(' ', text)
|
| 34 |
|
| 35 |
-
# Expand contractions
|
| 36 |
-
text = ' '.join([contractions_dict.get(word, word) for word in text.split()])
|
| 37 |
-
|
| 38 |
# Remove punctuation and digits
|
| 39 |
text = re.sub(r'[^\w\s]', ' ', text)
|
| 40 |
|
|
|
|
| 18 |
torch.manual_seed(1)
|
| 19 |
|
| 20 |
# Preprocess function
|
|
|
|
|
|
|
| 21 |
from nltk.corpus import stopwords
|
| 22 |
from nltk.tokenize import word_tokenize
|
| 23 |
|
|
|
|
| 30 |
html_pattern = re.compile(r'<[^<>]+>')
|
| 31 |
text = html_pattern.sub(' ', text)
|
| 32 |
|
|
|
|
|
|
|
|
|
|
| 33 |
# Remove punctuation and digits
|
| 34 |
text = re.sub(r'[^\w\s]', ' ', text)
|
| 35 |
|