[Bugfix] Fix prefix strings for quantized VLMs (#9772)

2024-10-29 19:02:59 -04:00
parent 8d7724104a
commit bc73e9821c
20 changed files with 288 additions and 97 deletions
--- a/vllm/model_executor/models/internvl.py
+++ b/vllm/model_executor/models/internvl.py
@@ -439,7 +439,10 @@ class InternVLChatModel(nn.Module, SupportsMultiModal, SupportsPP):
        )

        self.language_model = init_vllm_registered_model(
-            config.text_config, cache_config, quant_config)
+            config.text_config,
+            cache_config,
+            quant_config,
+            prefix="language_model")

        self.mlp1 = self._init_mlp1(config)