Update app.py

app.py

@@ -114,21 +114,22 @@ def talk(prompt, history):
     # the chat template structure should be based on text generation model format
 
 # indicates the end of a sequence
-    stream = model.create_chat_completion(messages = messages, max_tokens=1000, stop=["</s>"], stream=False)
-    print(f"{stream}")
+    stream = model.create_chat_completion(messages = messages, max_tokens=1000, stop=["</s>"], stream=True)
+#    print(f"{stream}")
     print("check 7")
-    print(stream['choices'][0]['message']['content'])
-    return(stream['choices'][0]['message']['content'])
-    #    text = []
-#    for output in stream:
-#       text += output['choices'][0]  
-#        print(f"{output}")
-#        print("check3H")
-
-#        text.append(output['choices'][0])
-#        print(f"{text}")
- #       yield "".join(text)
- #       print(text)
+#    print(stream['choices'][0]['message']['content'])
+#    return(stream['choices'][0]['message']['content'])
+    text = ""
+    for output in stream:
+       text += output['choices'][0]['message']['content']  
+       print(f"{output}")
+       print("check3H")
+       print(text)
+       return(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