From dec28688c5060f921594f9a7ce1647cf7a87a39e Mon Sep 17 00:00:00 2001
From: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Date: Mon, 12 Jan 2026 13:08:30 -0800
Subject: [PATCH] [Model Runner V2] Minor refactor for logit_bias (#32209)

Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
---
 vllm/v1/worker/gpu/sample/logit_bias.py | 80 +++++++++++++++++--------
 1 file changed, 54 insertions(+), 26 deletions(-)

diff --git a/vllm/v1/worker/gpu/sample/logit_bias.py b/vllm/v1/worker/gpu/sample/logit_bias.py
index 396e177d08..f959b36e45 100644
--- a/vllm/v1/worker/gpu/sample/logit_bias.py
+++ b/vllm/v1/worker/gpu/sample/logit_bias.py
@@ -119,35 +119,18 @@ class LogitBiasState:
         idx_mapping: torch.Tensor,
         pos: torch.Tensor,
     ) -> None:
-        num_reqs, vocab_size = logits.shape
-        BLOCK_SIZE = triton.next_power_of_2(
-            max(
-                MAX_NUM_ALLOWED_TOKEN_IDS,
-                MAX_NUM_LOGIT_BIAS_TOKENS,
-                MAX_NUM_STOP_TOKEN_IDS,
-            )
-        )
-        LOGITS_BLOCK_SIZE = 8192
-        _bias_kernel[(num_reqs,)](
+        apply_logit_bias(
             logits,
-            logits.stride(0),
-            vocab_size,
             idx_mapping,
-            self.num_allowed_token_ids,
-            self.allowed_token_ids,
-            self.allowed_token_ids.gpu.stride(0),
-            self.num_logit_bias,
-            self.logit_bias_token_ids,
-            self.logit_bias_token_ids.gpu.stride(0),
-            self.logit_bias,
-            self.logit_bias.gpu.stride(0),
             pos,
-            self.min_lens,
-            self.num_stop_token_ids,
-            self.stop_token_ids,
-            self.stop_token_ids.gpu.stride(0),
-            BLOCK_SIZE=BLOCK_SIZE,
-            LOGITS_BLOCK_SIZE=LOGITS_BLOCK_SIZE,
+            self.num_allowed_token_ids.gpu,
+            self.allowed_token_ids.gpu,
+            self.num_logit_bias.gpu,
+            self.logit_bias_token_ids.gpu,
+            self.logit_bias.gpu,
+            self.min_lens.gpu,
+            self.num_stop_token_ids.gpu,
+            self.stop_token_ids.gpu,
         )
 
 
@@ -240,3 +223,48 @@ def _bias_kernel(
             -float("inf"),
             mask=mask,
         )
+
+
+def apply_logit_bias(
+    logits: torch.Tensor,
+    idx_mapping: torch.Tensor,
+    pos: torch.Tensor,
+    num_allowed_token_ids: torch.Tensor,
+    allowed_token_ids: torch.Tensor,
+    num_logit_bias: torch.Tensor,
+    logit_bias_token_ids: torch.Tensor,
+    logit_bias: torch.Tensor,
+    min_lens: torch.Tensor,
+    num_stop_token_ids: torch.Tensor,
+    stop_token_ids: torch.Tensor,
+) -> None:
+    num_reqs, vocab_size = logits.shape
+    BLOCK_SIZE = triton.next_power_of_2(
+        max(
+            allowed_token_ids.shape[-1],
+            logit_bias_token_ids.shape[-1],
+            stop_token_ids.shape[-1],
+        )
+    )
+    LOGITS_BLOCK_SIZE = 8192
+    _bias_kernel[(num_reqs,)](
+        logits,
+        logits.stride(0),
+        vocab_size,
+        idx_mapping,
+        num_allowed_token_ids,
+        allowed_token_ids,
+        allowed_token_ids.stride(0),
+        num_logit_bias,
+        logit_bias_token_ids,
+        logit_bias_token_ids.stride(0),
+        logit_bias,
+        logit_bias.stride(0),
+        pos,
+        min_lens,
+        num_stop_token_ids,
+        stop_token_ids,
+        stop_token_ids.stride(0),
+        BLOCK_SIZE=BLOCK_SIZE,
+        LOGITS_BLOCK_SIZE=LOGITS_BLOCK_SIZE,
+    )