Simple bug fix and more commenting

README.md

@@ -1,10 +1,13 @@
 ---
-title: Test
-emoji: 📈
+title: KairosNews
+emoji: 📲📰
 colorFrom: red
 colorTo: indigo
 sdk: docker
-pinned: false
+pinned: true
+preload_from_hub:
+  - unicamp-dl/ptt5-base-portuguese-vocab
+  - recogna-nlp/ptt5-base-summ
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -11,7 +11,7 @@ from models.embedding import EmbeddingModel
 from models.summarization import SummarizationModel
 from models.nlp import NLPModel
 from database.query import DatabaseService
-from KairosNews.main import QueryProcessor
+from main import QueryProcessor
 
 # Configure logging
 logging.basicConfig(

database/query.py

@@ -6,7 +6,7 @@ from psycopg2 import sql
 
 class DatabaseService:
     def __init__(self):
-        # Connection parameters
+        # Supabase Connection Parameters
         self.DB_HOST = os.getenv("SUPABASE_HOST", "aws-0-eu-west-3.pooler.supabase.com")
         self.DB_PORT = os.getenv("DB_PORT", "6543")
         self.DB_NAME = os.getenv("DB_NAME", "postgres")
@@ -14,6 +14,7 @@ class DatabaseService:
         self.DB_PASSWORD = os.getenv("DB_PASSWORD")
 
     async def semantic_search(
+        #Query parameters
         self,
         query_embedding: List[float],
         start_date: Optional[datetime] = None,
@@ -23,7 +24,9 @@ class DatabaseService:
         limit: int = 10
     ) -> List[Dict[str, any]]:
         
+        # Entity log Checking
         print(f"Extracted entities2: {entities}")
+        
         try:
             with psycopg2.connect(
                 user=self.DB_USER,
@@ -33,7 +36,7 @@ class DatabaseService:
                 dbname=self.DB_NAME
             ) as conn:
                 with conn.cursor() as cursor:
-                    # Base query with date range and topic filters
+                    # Base query
                     base_query = sql.SQL('''
                         WITH filtered_articles AS (
                             SELECT article_id
@@ -74,7 +77,8 @@ class DatabaseService:
                                 sql.Literal(e[1])   # Original entity label (case-sensitive)
                             ) for e in entities
                         )
-
+                        
+                        # Final query with entity conditions and all filters
                         final_query = sql.SQL('''
                             {base_query},
                             target_articles AS (
@@ -99,7 +103,10 @@ class DatabaseService:
                             embedding=sql.Literal(query_embedding),
                             limit=sql.Literal(limit)
                         )
+
+                    # Final query with all filters but no entities
                     else:
+                        print("No articles found with entities...")
                         final_query = sql.SQL('''
                             {base_query}
                             SELECT
@@ -121,9 +128,9 @@ class DatabaseService:
                     cursor.execute(final_query)
                     articles = cursor.fetchall()
 
-                    # Fallback: Retry with fewer filters if no results
+                    # Fallback: Retry with no filters if no results, only semantic search
                     if not articles:
-                        print("No articles found with entities...")
+                        print("No articles found with the filters applied. Trying fallback query...")
                         fallback_query = sql.SQL('''
                             SELECT
                                 content,
@@ -160,5 +167,5 @@ class DatabaseService:
             return []
 
     async def close(self):
-        # No persistent connection to close in psycopg2
+        # No explicit close needed with context manager
         pass

models/summarization.py

@@ -8,6 +8,8 @@ class SummarizationModel:
         self.model = T5ForConditionalGeneration.from_pretrained('recogna-nlp/ptt5-base-summ').to(self.device)
     
     def summarize(self, text: str, max_length: int = 256, min_length: int = 128) -> str:
+        """Summarize the input text using T5 model"""
+        # Model and tokenization parameters
         inputs = self.tokenizer.encode(
             text,
             max_length=512,