Update app.py

app.py

@@ -129,8 +129,7 @@ try:
         MODEL_ID,
         device_map="auto",
         quantization_config=quantization_config,
-        torch_dtype=torch.bfloat16,
-        attn_implementation="flash_attention_2"  # Enable Flash Attention 2
+        torch_dtype=torch.bfloat16
     )
     
     # Apply Better Transformer optimization