[DP Attention] Refactor: adding some utility functions (#9136)

python/sglang/srt/models/llama4.py

@@ -32,6 +32,7 @@ from sglang.srt.layers.dp_attention import (
     get_attention_tp_rank,
     get_attention_tp_size,
     get_local_attention_dp_size,
+    is_dp_attention_enabled,
 )
 from sglang.srt.layers.layernorm import RMSNorm
 from sglang.srt.layers.linear import (
@@ -45,7 +46,6 @@ from sglang.srt.layers.quantization.base_config import QuantizationConfig
 from sglang.srt.layers.radix_attention import RadixAttention
 from sglang.srt.layers.rotary_embedding import get_rope
 from sglang.srt.layers.vocab_parallel_embedding import VocabParallelEmbedding
-from sglang.srt.managers.schedule_batch import global_server_args_dict
 from sglang.srt.model_executor.forward_batch_info import (
     ForwardBatch,
     ForwardMode,
@@ -466,7 +466,7 @@ class Llama4Model(nn.Module):
             config.hidden_size,
             quant_config=quant_config,
             prefix=add_prefix("embed_tokens", prefix),
-            enable_tp=not global_server_args_dict["enable_dp_attention"],
+            enable_tp=not is_dp_attention_enabled(),
         )
         self.layers = make_layers(
             config.num_hidden_layers,