Update app.py

app.py

@@ -136,14 +136,15 @@ def talk(prompt, history):
    # formatted_prompt_with_history = formatted_prompt_with_history[:600] # to avoid memory issue
 #    print(formatted_prompt_with_history)
     messages = [{"role":"system","content":SYS_PROMPT},{"role":"user","content":formatted_prompt}]  
+#    messages = "role":"system","content":SYS_PROMPT},{"role":"user","content":formatted_prompt}]
     print(messages) 
     # binding the system context and new prompt for LLM
     # the chat template structure should be based on text generation model format
     print("check6")    
 
 # indicates the end of a sequence
-#    stream = model.create_chat_completion(messages = messages, max_tokens=1000, stop=["</s>"], stream=False)
-    stream = model.create_completion(prompt = messages, max_tokens=1000, stop=["</s>"],echo=True, stream=False) 
+    stream = model.create_chat_completion(messages = messages, max_tokens=1000, stop=["</s>"], stream=False)
+#    stream = model(prompt = messages, max_tokens=1000, stop=["</s>"],echo=True, stream=False) 
     print(f"{stream}")
     print("check 7")
     print(stream['choices'][0]['message']['content'])