[Bugfix] model_max_length should consider max_model_len in tokenizer_config

vllm/config.py

@@ -44,7 +44,8 @@
     ConfigFormat, get_config, get_hf_image_processor_config,
     get_hf_text_config, get_pooling_config,
     get_sentence_transformer_tokenizer_config, is_encoder_decoder,
-    try_get_generation_config, try_get_safetensors_metadata, uses_mrope)
+    try_get_generation_config, try_get_safetensors_metadata,
+    try_get_tokenizer_config, uses_mrope)
 from vllm.transformers_utils.s3_utils import S3Model
 from vllm.transformers_utils.utils import is_s3, maybe_model_redirect
 from vllm.utils import (DEFAULT_MAX_NUM_BATCHED_TOKENS,
@@ -1427,6 +1428,18 @@ def get_and_verify_max_len(self, max_model_len: int):
             sliding_window_len=self.get_hf_config_sliding_window(),
             spec_target_max_model_len=self.spec_target_max_model_len,
             encoder_config=self.encoder_config)
+
+        tokenizer_config = try_get_tokenizer_config(
+            self.tokenizer,
+            trust_remote_code=self.trust_remote_code,
+            revision=self.tokenizer_revision)
+
+        if tokenizer_config is None:
+            return max_model_len
+
+        model_max_length = tokenizer_config.get("model_max_length",
+                                                max_model_len)
+        max_model_len = min(max_model_len, model_max_length)
         return max_model_len
 
 

vllm/transformers_utils/config.py

@@ -23,6 +23,7 @@
     get_image_processor_config)
 from transformers.models.auto.modeling_auto import (
     MODEL_FOR_CAUSAL_LM_MAPPING_NAMES)
+from transformers.models.auto.tokenization_auto import get_tokenizer_config
 from transformers.utils import CONFIG_NAME as HF_CONFIG_NAME
 
 from vllm import envs
@@ -867,3 +868,18 @@ def try_get_safetensors_metadata(
                           "Error retrieving safetensors")
     except Exception:
         return None
+
+
+def try_get_tokenizer_config(
+    pretrained_model_name_or_path: Union[str, os.PathLike],
+    trust_remote_code: bool,
+    revision: Optional[str] = None,
+) -> Optional[dict[str, Any]]:
+    try:
+        return get_tokenizer_config(
+            pretrained_model_name_or_path,
+            trust_remote_code=trust_remote_code,
+            revision=revision,
+        )
+    except Exception:
+        return None