Update app.py

app.py

@@ -11,10 +11,11 @@ pipe = DiffusionPipeline.from_pretrained(
     "Tongyi-MAI/Z-Image-Turbo",#"T5B/Z-Image-Turbo-FP8",
     torch_dtype=torch.bfloat16,
     low_cpu_mem_usage=False,
+    attn_implementation="kernels-community/vllm-flash-attn3",
 )
 
 
-pipe.transformer.set_attention_backend(attention_backend)
+#pipe.transformer.set_attention_backend(attention_backend)
 
 if enable_compile:
     print("Compiling transformer...")