return

app.py

@@ -123,10 +123,10 @@ chunk_embeddings = create_embeddings(cleaned_chunks)
 client= InferenceClient("Qwen/Qwen2.5-7B-Instruct-1M")
 response=""
 #defining role of AI and user
-messages = [{"role": "assistant", "content": f"You are a friendly chatbot that gives advice to disadvantaged students about their education based on their question. When you give advice, keep in mind the following infromation {information}"}]
 
 def respond(message,history):
     information = get_top_chunks(message, chunk_embeddings, cleaned_chunks)
+    messages = [{"role": "assistant", "content": f"You are a friendly chatbot that gives advice to disadvantaged students about their education based on their question. When you give advice, keep in mind the following infromation {information}"}]
 
 
     if history:
@@ -134,11 +134,9 @@ def respond(message,history):
 
     messages.append({"role":"user", "content": message})
 
-    response=client.chat_completion(messages, stream=True, max_tokens=100) #capping how many words the LLM is allowed to generate as a respond (100 words)
-    for message in client.chat_completion(messages):
-        token = message.choices[0].delta.content
-        response+=token
-    yield response['choices'][0]['message']['content'].strip() #storing value of response in a readable format to display
+    response=client.chat_completion(messages, max_tokens=100) #capping how many words the LLM is allowed to generate as a respond (100 words)
+    
+    return response['choices'][0]['message']['content'].strip() #storing value of response in a readable format to display
 
 ### STEP 6
 # Call the preprocess_text function and store the result in a cleaned_chunks variable