Update app.py

app.py

@@ -132,25 +132,27 @@ 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}]  
     # 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
     import pprint
-    stream = model.create_chat_completion(messages = [{"role":"system","content":SYS_PROMPT},{"role":"user","content":formatted_prompt}], max_tokens=1000, stop=["</s>"], stream=True)
-#   print(output['choices'][0]['message']['content'])
-    text = []
-    for output in stream:
+    stream = model.create_chat_completion(messages = messages, max_tokens=1000,echo=True, stop=["</s>"], stream=False)
+    print(f"{stream}")
+    print("check 7")
+    print(stream['choices'][0]['message']['content'])
+#    text = []
+#    for output in stream:
 #       text += output['choices'][0]  
-        print(f"{output}")
-        print("check3H")
+#        print(f"{output}")
+#        print("check3H")
 
-        text.append(output['choices'][0])
-        print(f"{text}")
-        yield "".join(text)
-        print(text)
+#        text.append(output['choices'][0])
+#        print(f"{text}")
+ #       yield "".join(text)
+ #       print(text)
     
     # preparing tokens for model input
     # add_generation_prompt argument tells the template to add tokens that indicate the start of a bot response