fix: fix mm processor double bos (#26505)

python/sglang/srt/multimodal/processors/base_processor.py

@@ -204,6 +204,12 @@ class BaseMultimodalProcessor(ABC):
         else:
             self._tokenizer = self._processor
 
+        # Same guard as in serving_chat.py against double BOS.
+        try:
+            self._tokenizer_auto_adds_specials = len(self._tokenizer.encode("")) > 0
+        except Exception:
+            self._tokenizer_auto_adds_specials = False
+
         # FIXME: not accurate, model and image specific
         self.NUM_TOKEN_PER_FRAME = 330
 
@@ -467,6 +473,12 @@ class BaseMultimodalProcessor(ABC):
                 npu_apply_glm46v_image_preprocess_patch()
                 kwargs["device"] = "npu"
 
+        # Avoid double BOS when the chat template already wrote one.
+        if self._tokenizer_auto_adds_specials and isinstance(input_text, str):
+            bos = getattr(self._tokenizer, "bos_token", None)
+            if bos and input_text.startswith(bos):
+                kwargs.setdefault("add_special_tokens", False)
+
         result = processor.__call__(
             text=[input_text],
             padding=True,

test/registered/unit/managers/test_mm_process_config.py

@@ -138,6 +138,8 @@ class TestProcessMmDataKwargs(unittest.TestCase):
         proc.disable_fast_image_processor = server_args.disable_fast_image_processor
         proc.skip_tokenizer_init = server_args.skip_tokenizer_init
         proc._processor = mock_processor
+        proc._tokenizer = MagicMock()
+        proc._tokenizer_auto_adds_specials = False
         proc.image_config = mm_process_config.get("image", {})
         proc.video_config = mm_process_config.get("video", {})
         proc.audio_config = mm_process_config.get("audio", {})
@@ -296,5 +298,48 @@ class TestOverrideProcessorsConfigInjection(unittest.TestCase):
         self.assertTrue(audio_kw.get("truncation"))
 
 
+class TestDoubleBosGuard(unittest.TestCase):
+    """Regression test for the multimodal double-BOS bug.
+
+    Repro condition (Cohere2 / Llama3-LLaVA-Next family):
+      - tokenizer.encode("") returns [bos_id]  (auto-adds specials), AND
+      - chat template renders the BOS string as a literal at the start.
+
+    Without the guard in BaseMultimodalProcessor, the inner processor.__call__
+    on the rendered prompt would auto-prepend a second BOS, producing 2 leading
+    BOS tokens vs the HF reference's 1.
+    """
+
+    def test_guard_passes_add_special_tokens_false_on_bug_condition(self):
+        from sglang.srt.multimodal.processors.base_processor import (
+            BaseMultimodalProcessor,
+        )
+
+        server_args = MagicMock()
+        server_args.mm_process_config = {}
+        server_args.disable_fast_image_processor = True
+        server_args.keep_mm_feature_on_device = True
+
+        mock_hf_processor = MagicMock()
+        mock_hf_processor.__class__.__name__ = "TestProcessor"
+        mock_hf_processor.__call__ = MagicMock(return_value={})
+        mock_hf_processor.tokenizer.encode = MagicMock(return_value=[2])
+        mock_hf_processor.tokenizer.bos_token = "<BOS>"
+
+        with patch.object(BaseMultimodalProcessor, "__abstractmethods__", set()):
+            proc = BaseMultimodalProcessor(
+                hf_config=MagicMock(),
+                server_args=server_args,
+                _processor=mock_hf_processor,
+                transport_mode=None,
+            )
+        proc.FEATURE_NAMES = []
+
+        proc.process_mm_data("<BOS>hello", images=["img1"])
+
+        call_kwargs = mock_hf_processor.__call__.call_args.kwargs
+        self.assertEqual(call_kwargs.get("add_special_tokens"), False)
+
+
 if __name__ == "__main__":
     unittest.main()