Update app.py

app.py

@@ -11,7 +11,7 @@ model= AutoModelForCausalLM.from_pretrained(
                                             torch_dtype= torch.bfloat16,
                                             low_cpu_mem_usage= True, 
                                             token=token,
-                                            attn_implementation="flash_attention_2",
+                                            #attn_implementation="flash_attention_2",
                                             device_map= "auto"
      )