add load_model and delete init

app.py

@@ -10,12 +10,13 @@ from bs4 import BeautifulSoup
 
 class NewsAnalytic():
     @st.cache_resource
-    def __init__(self):
-        self.model = T5ForZeroShotClassification.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
-        self.tokenizer = T5Tokenizer.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
-        self.classifier = ZeroShotClassificationPipeline(model=self.model, tokenizer=self.tokenizer,ypothesis_template = '{}', encoder_decoder = True)
+    def load_model(self):
+        model = T5ForZeroShotClassification.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
+        tokenizer = T5Tokenizer.from_pretrained('knowledgator/comprehend_it-multilingual-t5-base')
+        classifier = ZeroShotClassificationPipeline(model=model, tokenizer=tokenizer,ypothesis_template = '{}', encoder_decoder = True)
         nltk.download("punkt")
         print("load model berhasil!")
+        return classifier
     def anoted_sentence(self,content,candidate_labels):
         sentences = nltk.sent_tokenize(content)
         sentences = [sent for sent in sentences if "simak" not in sent.lower()]
@@ -46,6 +47,7 @@ class NewsAnalytic():
         print("mendapatkan konten berita berhasil!")
         return {'title': title.strip(), 'content': content.strip()}
     def streamlit_run(self):
+        self.classifier=self.load_model()
         st.markdown("<h2 style='text-align: center;'>Indonesian News Analytic </h2>", unsafe_allow_html=True)
         st.markdown("<p style='text-align: center;'>Hafizh Zaki Prasetyo Adi|[email protected]|https://www.linkedin.com/in/hafizhzpa/ </p>", unsafe_allow_html=True)
         part=st.sidebar.radio("input_type",["content", "link"],captions = ["input news content", "input news link"])
@@ -66,10 +68,4 @@ class NewsAnalytic():
             result_category=self.classifier(text, ["Politik","Ekonomi","Hukum dan Kriminal","Teknologi","Pendidikan","Kesehatan","Olahraga","Hiburan","Gaya Hidup","Lingkungan","Transportasi","Pariwisata"])
             st.text(f"category: {result_category['labels'][0]}")
             result_region=self.classifier(text, ["regional","nasional","internasional"])
-            st.text(f"region: {result_region['labels'][0]}")
-
-                
-
-if __name__=="__main__":
-    pipeline=NewsAnalytic()
-    pipeline.streamlit_run()
+            st.text(f"region: {result_region['labels'][0]}")