[V0 deprecation] Remove more V0 references (#29088)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/entrypoints/llm.py

@@ -339,7 +339,6 @@ class LLM:
 
         log_non_default_args(engine_args)
 
-        # Create the Engine (autoselects V0 vs V1)
         self.llm_engine = LLMEngine.from_engine_args(
             engine_args=engine_args, usage_context=UsageContext.LLM_CLASS
         )

vllm/entrypoints/openai/protocol.py

@@ -377,7 +377,7 @@ class ResponsesRequest(OpenAIBaseModel):
             "environments. The salt should be random, protected from "
             "access by 3rd parties, and long enough to be "
             "unpredictable (e.g., 43 characters base64-encoded, corresponding "
-            "to 256 bit). Not supported by vLLM engine V0."
+            "to 256 bit)."
         ),
     )
 
@@ -763,7 +763,7 @@ class ChatCompletionRequest(OpenAIBaseModel):
             "environments. The salt should be random, protected from "
             "access by 3rd parties, and long enough to be "
             "unpredictable (e.g., 43 characters base64-encoded, corresponding "
-            "to 256 bit). Not supported by vLLM engine V0."
+            "to 256 bit)."
         ),
     )
     kv_transfer_params: dict[str, Any] | None = Field(
@@ -1249,7 +1249,7 @@ class CompletionRequest(OpenAIBaseModel):
             "environments. The salt should be random, protected from "
             "access by 3rd parties, and long enough to be "
             "unpredictable (e.g., 43 characters base64-encoded, corresponding "
-            "to 256 bit). Not supported by vLLM engine V0."
+            "to 256 bit)."
         ),
     )
 

vllm/model_executor/layers/mamba/mamba_mixer2.py

@@ -590,7 +590,6 @@ class MambaMixer2(MambaBase, CustomOp):
             hidden_states, _B, _C = self.split_hidden_states_B_C_fn(hidden_states_B_C)
             return hidden_states
 
-        # NOTE: V0 put prefill before decode, v1 puts decode before prefill
         num_prefills = attn_metadata.num_prefills  # request count
         num_decodes = attn_metadata.num_decode_tokens  # token count (=request)
         num_prefill_tokens = attn_metadata.num_prefill_tokens  # token count

vllm/model_executor/models/interfaces.py

@@ -586,13 +586,11 @@ class IsHybrid(Protocol):
     def get_mamba_state_shape_from_config(
         cls,
         vllm_config: VllmConfig,
-        use_v1: bool = True,
     ) -> tuple[tuple[int, int], tuple[int, int, int]]:
         """Calculate shapes for Mamba's convolutional and state caches.
 
         Args:
             vllm_config: vLLM config
-            use_v1: Get shapes for V1 (or V0)
 
         Returns:
             Tuple containing:

vllm/model_executor/models/plamo2.py

@@ -290,7 +290,6 @@ class Plamo2MambaMixer(MambaBase, CustomOp):
         has_decode = num_decodes > 0
         num_actual_tokens = num_prefill_tokens + num_decodes
 
-        # NOTE: V0 put prefill before decode, v1 puts decode before prefill
         # Separate prefill and decode by splitting varlen input
         # Split along token dimension
         hidden_states_d, hidden_states_p = torch.split(