Update app.py

app.py

@@ -76,7 +76,7 @@ model = AutoModelForSeq2SeqLM.from_pretrained(
     MODEL_ID,
     device_map="auto",
     quantization_config=quantization_config,
-    attn_implementation="flash_attention_2",
+    trust_remote_code=True
 )
 
 # Create Gradio interface