Update app.py

app.py

@@ -7,11 +7,11 @@ import random
 # Load model for chat
 client = InferenceClient("Qwen/Qwen2.5-72B-Instruct")
 
-# Read the TV show database
+# Step 2: Load TV show database
 with open("reconext_file.txt", "r", encoding="utf-8") as file:
     reconext_file_text = file.read()
 
-# Preprocess text chunks
+# Step 3: Preprocess the text
 def preprocess_text(text):
     cleaned_text = text.strip()
     chunks = cleaned_text.split("\n")
@@ -24,7 +24,7 @@ def preprocess_text(text):
 
 cleaned_chunks = preprocess_text(reconext_file_text)
 
-# Create embeddings
+# Step 4: Create embeddings
 model = SentenceTransformer('all-MiniLM-L6-v2')
 
 def create_embeddings(text_chunks):
@@ -33,7 +33,7 @@ def create_embeddings(text_chunks):
 
 chunk_embeddings = create_embeddings(cleaned_chunks)
 
-# Semantic search function
+# Step 5: Semantic search
 def get_top_chunks(query, chunk_embeddings, text_chunks):
     query_embedding = model.encode(query, convert_to_tensor=True)
     query_embedding_normalized = query_embedding / query_embedding.norm()
@@ -43,17 +43,18 @@ def get_top_chunks(query, chunk_embeddings, text_chunks):
     top_chunks = [text_chunks[i] for i in top_indices]
     return top_chunks
 
-# Chat response function
+# Response function
 def respond(message, history):
     best_next_watch = get_top_chunks(message, chunk_embeddings, cleaned_chunks)
     str_watch_chunks = "\n".join(best_next_watch)
+    
     messages = [
         {
             "role": "system",
             "content": (
                 "You are a Gen Z and Gen Alpha-friendly chatbot that helps teenagers find their next best TV show to watch. "
-                "Speak naturally and casually, like someone from Gen Z. Only recommend TV shows, never movies. Use only the shows in our database "
-                "YOU CAN NEVER USE OUTSIDE DATA ONLY TAKE DATA FROM OUR DATABASE! Match show suggestions to the user's age using TV ratings: "
+                "Speak naturally and casually, like someone from Gen Z. Only recommend TV shows, never movies. Use only the shows in our database. "
+                "YOU CAN NEVER USE OUTSIDE DATA — ONLY TAKE DATA FROM OUR DATABASE! Match show suggestions to the user's age using TV ratings: "
                 "TV-G is for all ages, TV-PG is for ages 6 and up, TV-14 is for 14 and up, and TV-MA is for 18 and up. "
                 "If they don’t share their age, assume they’re Gen Z or Gen Alpha and use those guidelines. "
                 "If the user is not Gen Z or Gen Alpha, you can recommend any show from the database. "
@@ -65,18 +66,21 @@ def respond(message, history):
             )
         }
     ]
+
     if history:
         messages.extend(history)
+
     messages.append({"role": "user", "content": message})
 
     response = client.chat_completion(
         messages, max_tokens=700, temperature=1.3, top_p=0.6
     )
+
     return response['choices'][0]['message']['content'].strip()
 
 # Initial chatbot message
 initial_message = [("🤖", "Hey! I’m your Gen-Z watch buddy.\nI help you find your next favorite TV show based on what you like, your age, or your favorite genre. Just tell me what you're into!")]
 
-# Gradio chat interface
+# Launch Gradio interface with initial message
 chatbot = gr.ChatInterface(respond, type="messages", chatbot=initial_message)
 chatbot.launch()