[VLM] Calculate maximum number of multi-modal tokens by model (#6121)

2024-07-05 07:37:23 +08:00
parent 69ec3ca14c
commit ae96ef8fbd
12 changed files with 265 additions and 95 deletions
--- a/vllm/inputs/registry.py
+++ b/vllm/inputs/registry.py
@@ -51,7 +51,7 @@ class InputContext:
        additionally checking its type.

        Raises:
-            ValueError: If the model is not of the specified type.
+            TypeError: If the model is not of the specified type.
        """

        hf_config = self.model_config.hf_config