diff --git a/python/sglang/srt/layers/attention/hybrid_linear_attn_backend.py b/python/sglang/srt/layers/attention/hybrid_linear_attn_backend.py
index 5cfaac9e4418..c121e663ae83 100644
--- a/python/sglang/srt/layers/attention/hybrid_linear_attn_backend.py
+++ b/python/sglang/srt/layers/attention/hybrid_linear_attn_backend.py
@@ -649,11 +649,9 @@ def forward_extend(
         if is_target_verify:
             batch_size = seq_len // forward_batch.spec_info.draft_token_num
             draft_token_num = forward_batch.spec_info.draft_token_num
-            mixed_qkv_reshaped = (
-                mixed_qkv.view(batch_size, draft_token_num, -1)
-                .transpose(1, 2)
-                .contiguous()
-            )
+            mixed_qkv_reshaped = mixed_qkv.view(
+                batch_size, draft_token_num, -1
+            ).transpose(1, 2)
             mixed_qkv_processed = causal_conv1d_update(
                 mixed_qkv_reshaped,
                 conv_states_to_use,
@@ -666,9 +664,7 @@ def forward_extend(
                 retrieve_next_sibling=retrieve_next_sibling,
                 retrieve_parent_token=retrieve_parent_token,
             )
-            mixed_qkv = (
-                mixed_qkv_processed.transpose(1, 2).contiguous().view(seq_len, -1)
-            )
+            mixed_qkv = mixed_qkv_processed.transpose(1, 2).view(seq_len, -1)
         else:
             mixed_qkv = causal_conv1d_fn(
                 mixed_qkv.transpose(0, 1),