[V0 Deprecation][Models] Remove all V0 condition for mm embeddings merge (#25331)

Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: isotr0py <2037008807@qq.com>

vllm/model_executor/models/ultravox.py

@@ -597,18 +597,6 @@ class UltravoxModel(nn.Module, SupportsMultiModal, SupportsPP, SupportsLoRA):
         if intermediate_tensors is not None:
             inputs_embeds = None
 
-        # NOTE: In v1, inputs_embeds is always generated at model runner, this
-        # condition is for v0 compatibility.
-        elif inputs_embeds is None:
-            multimodal_embeddings = self.get_multimodal_embeddings(**kwargs)
-
-            inputs_embeds = self.get_input_embeddings(
-                input_ids,
-                multimodal_embeddings,
-                is_multimodal=input_ids == self.config.audio_token_index,
-            )
-            input_ids = None
-
         language_model = self.language_model
         if hasattr(language_model, "language_model"):
             language_model = language_model.language_model

vllm/model_executor/models/voxtral.py

@@ -371,19 +371,6 @@ class VoxtralForConditionalGeneration(nn.Module, SupportsMultiModal,
         if intermediate_tensors is not None:
             inputs_embeds = None
 
-        # NOTE: In v1, inputs_embeds is always generated at model runner, this
-        # condition is for v0 compatibility.
-        elif inputs_embeds is None:
-            audio_encoder = self.tokenizer.instruct.audio_encoder
-            audio_tok_id = audio_encoder.audio_token
-            audio_embeddings = self.get_multimodal_embeddings(**kwargs)
-            inputs_embeds = self.get_input_embeddings(
-                input_ids,
-                audio_embeddings,
-                is_multimodal=input_ids == audio_tok_id,
-            )
-            input_ids = None
-
         hidden_states = self.language_model.model(input_ids,
                                                   positions,
                                                   intermediate_tensors,