fix Lint style

Li-Z-Q · Li-Z-Q · commit 305835cc97ae · 2025-07-17T16:34:02.000+08:00
diff --git a/paddlenlp/experimental/transformers/fused_transformer_layers.py b/paddlenlp/experimental/transformers/fused_transformer_layers.py
@@ -3172,7 +3172,7 @@ def compute_attn(
             k_dequant_scales = kwargs.get("k_dequant_scales", None)
             v_dequant_scales = kwargs.get("v_dequant_scales", None)
 
-            if kwargs['kv_cache_reuse']:
+            if kwargs["kv_cache_reuse"]:
                 k_cache_index = 0
                 v_cache_index = 1
             else:
