Skip broken AutoModel mapping entries when resolving Llava submodules (#21892)

python/sglang/srt/models/llava.py

@@ -53,6 +53,9 @@ from sglang.srt.multimodal.mm_utils import (
 )
 from sglang.srt.utils import add_prefix, flatten_nested_list, logger
 
+_KNOWN_BROKEN_AUTOMODEL_CONFIG = "VoxtralRealtimeTextConfig"
+_KNOWN_BROKEN_AUTOMODEL_ERROR = "Could not find VoxtralRealtimeTextModel"
+
 
 class LlavaBaseForCausalLM(nn.Module):
     def pad_input_ids(self, input_ids: List[int], image_inputs: MultimodalInputs):
@@ -657,7 +660,22 @@ class LlavaForConditionalGeneration(LlavaBaseForCausalLM):
     ) -> Dict[str, str]:
         mapping = {}
         for config_cls in auto_model_type._model_mapping.keys():
-            archs = auto_model_type._model_mapping.get(config_cls, None)
+            try:
+                archs = auto_model_type._model_mapping.get(config_cls, None)
+            except ValueError as exc:
+                if (
+                    auto_model_type is not AutoModel
+                    or config_cls.__name__ != _KNOWN_BROKEN_AUTOMODEL_CONFIG
+                    or _KNOWN_BROKEN_AUTOMODEL_ERROR not in str(exc)
+                ):
+                    raise
+                logger.warning(
+                    "Skipping broken %s mapping for config %s: %s",
+                    auto_model_type.__name__,
+                    config_cls.__name__,
+                    exc,
+                )
+                continue
             if archs is not None:
                 if isinstance(archs, tuple):
                     mapping[config_cls.__name__] = tuple(