Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# A model for predicting the category of news article
|
2 |
+
## Usage:
|
3 |
+
|
4 |
+
```
|
5 |
+
import re
|
6 |
+
from transformers import pipeline
|
7 |
+
from html import unescape
|
8 |
+
from unicodedata import normalize
|
9 |
+
|
10 |
+
re_multispace = re.compile(r"\s+")
|
11 |
+
|
12 |
+
def normalize_text(text):
|
13 |
+
if text == None:
|
14 |
+
return None
|
15 |
+
|
16 |
+
text = text.strip()
|
17 |
+
text = text.replace("\n", " ")
|
18 |
+
text = text.replace("\t", " ")
|
19 |
+
text = text.replace("\r", " ")
|
20 |
+
text = re_multispace.sub(" ", text)
|
21 |
+
text = unescape(text)
|
22 |
+
text = normalize("NFKC", text)
|
23 |
+
return text
|
24 |
+
|
25 |
+
|
26 |
+
model = pipeline(task="text-classification",
|
27 |
+
model=f"hynky/Category", tokenizer="ufal/robeczech-base",
|
28 |
+
truncation=True, max_length=512,
|
29 |
+
top_k=5
|
30 |
+
)
|
31 |
+
|
32 |
+
|
33 |
+
def predict(article):
|
34 |
+
article = normalize_text(article)
|
35 |
+
predictions = model(article)
|
36 |
+
|
37 |
+
predict("Dnes v noci bude pršet.")
|