a

app.py

@@ -5,7 +5,7 @@ from huggingface_hub import InferenceClient
 Copied from inference in colab notebook
 """
 
-from transformers import AutoTokenizer , AutoModelForCausalLM , TextStreamer
+from transformers import AutoTokenizer , AutoModelForCausalLM , TextIteratorStreamer
 import torch
 from threading import Thread
 
@@ -101,7 +101,7 @@ def respond(
         return_tensors = "pt",
     )
     # Generate tokens incrementally
-    streamer = TextStreamer(tokenizer, skip_prompt=True)
+    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True)
     generation_kwargs = {
         "input_ids": inputs,
         "max_new_tokens": max_tokens,