Update app.py

app.py

@@ -49,7 +49,7 @@ with gr.Blocks(theme=theme) as demo:
         outputs = ""
         for token in llm2.generate(tokens, top_k=50, top_p=0.73, temp=0.72, repeat_penalty=1.1):
             text = llm2.detokenize([token])
-            outputs += text.decode()
+            outputs += text.decode(errors='ignore')
             count += 1
             if count >= 500 or (token == llm2.token_eos()):
                 break