Spaces:
Build error
Build error
add status
Browse files
app.py
CHANGED
|
@@ -13,7 +13,7 @@ class NewsAnalytic():
|
|
| 13 |
self.model = T5ForZeroShotClassification.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
|
| 14 |
self.tokenizer = T5Tokenizer.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
|
| 15 |
self.classifier = ZeroShotClassificationPipeline(model=self.model, tokenizer=self.tokenizer,ypothesis_template = '{}', encoder_decoder = True)
|
| 16 |
-
|
| 17 |
def anoted_sentence(self,content,candidate_labels):
|
| 18 |
sentences = nltk.sent_tokenize(content)
|
| 19 |
sentences = [sent for sent in sentences if "simak" not in sent.lower()]
|
|
@@ -24,6 +24,7 @@ class NewsAnalytic():
|
|
| 24 |
else:
|
| 25 |
summary_result="neutral"
|
| 26 |
pair_result=[(res['sequence'],res['labels'][0]) if res['labels'][0]!="neutral" else res['sequence'] for res in result]
|
|
|
|
| 27 |
return pair_result, summary_result
|
| 28 |
def get_news_content(self,url):
|
| 29 |
response = requests.get(url)
|
|
@@ -40,7 +41,7 @@ class NewsAnalytic():
|
|
| 40 |
paragraphs = soup.find_all('p')
|
| 41 |
content = ' '.join([p.get_text() for p in paragraphs])
|
| 42 |
content = ' '.join([text for text in content.split("\n") if len(text)>100])
|
| 43 |
-
|
| 44 |
return {'title': title.strip(), 'content': content.strip()}
|
| 45 |
def streamlit_run(self):
|
| 46 |
st.markdown("<h2 style='text-align: center;'>Indonesian News Analytic </h2>", unsafe_allow_html=True)
|
|
|
|
| 13 |
self.model = T5ForZeroShotClassification.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
|
| 14 |
self.tokenizer = T5Tokenizer.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
|
| 15 |
self.classifier = ZeroShotClassificationPipeline(model=self.model, tokenizer=self.tokenizer,ypothesis_template = '{}', encoder_decoder = True)
|
| 16 |
+
print("load model berhasil!")
|
| 17 |
def anoted_sentence(self,content,candidate_labels):
|
| 18 |
sentences = nltk.sent_tokenize(content)
|
| 19 |
sentences = [sent for sent in sentences if "simak" not in sent.lower()]
|
|
|
|
| 24 |
else:
|
| 25 |
summary_result="neutral"
|
| 26 |
pair_result=[(res['sequence'],res['labels'][0]) if res['labels'][0]!="neutral" else res['sequence'] for res in result]
|
| 27 |
+
print("sentiment per sentence berhasil!")
|
| 28 |
return pair_result, summary_result
|
| 29 |
def get_news_content(self,url):
|
| 30 |
response = requests.get(url)
|
|
|
|
| 41 |
paragraphs = soup.find_all('p')
|
| 42 |
content = ' '.join([p.get_text() for p in paragraphs])
|
| 43 |
content = ' '.join([text for text in content.split("\n") if len(text)>100])
|
| 44 |
+
print("mendapatkan konten berita berhasil!")
|
| 45 |
return {'title': title.strip(), 'content': content.strip()}
|
| 46 |
def streamlit_run(self):
|
| 47 |
st.markdown("<h2 style='text-align: center;'>Indonesian News Analytic </h2>", unsafe_allow_html=True)
|