Update app.py

app.py

@@ -443,22 +443,22 @@ def send_message(message, history):
     #history.append({"role": "assistant", "content": f"This is a response about: {message}"})
     #return history
     tokenizer = generator.tokenizer
-    streamer = TextIteratorStreamer(self.tokenizer, skip_prompt=True)
+    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True)
 
     gen_kwargs = {
         "inputs": input_ids,
         "streamer": streamer,
-        "pad_token_id": self.tokenizer.eos_token_id,
+        "pad_token_id": tokenizer.eos_token_id,
         "max_length": 8192,
         "temperature": 0.1,
         "top_p": 0.8,
         "repetition_penalty": 1.25,
     }
-
-    thread = Thread(target=self.model.generate, kwargs=gen_kwargs)
+    partial = ""
+    thread = Thread(target=pipeline.generate, kwargs=gen_kwargs)
     thread.start()    
     #for token in generator(message, max_new_tokens=200):
-    for t in response_generator:
+    for t in streamer:
         partial += t#token["generated_text"][len(message):]
         yield history + [{"role": "assistant", "content": partial}]