Update app.py

app.py

@@ -24,7 +24,7 @@ def predict(message, history):
 
     inputs = tokenizer([messages], return_tensors="pt").to("cpu")
 
-    streamer = TextStreamer(tokenizer)
+    # streamer = TextStreamer(tokenizer)
     
     generation_config = GenerationConfig(
       penalty_alpha=0.6,
@@ -34,12 +34,12 @@ def predict(message, history):
       top_k=5,
       temperature=0.5,
       repetition_penalty=1.2,
-      max_new_tokens=64,
+      # max_new_tokens=64,
       eos_token_id=tokenizer.eos_token_id,
       pad_token_id=tokenizer.eos_token_id
   )
     
-    outputs = model.generate(**inputs,streamer=streamer, generation_config=generation_config)
+    outputs = model.generate(**inputs, generation_config=generation_config)
     response = tokenizer.decode(outputs[0], skip_special_tokens=True)
     return response[len(formatted_prompt(message)):]