[V0 Deprecation] Remove pooling model support in V0 (#23434)

Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>

vllm/engine/protocol.py

@@ -3,7 +3,7 @@
 
 import asyncio
 from abc import ABC, abstractmethod
-from typing import AsyncGenerator, Iterable, Mapping, Optional, Union
+from typing import Any, AsyncGenerator, Iterable, Mapping, Optional, Union
 
 from vllm.beam_search import BeamSearchSequence, create_sort_beams_key_function
 from vllm.config import DecodingConfig, ModelConfig, VllmConfig
@@ -224,6 +224,7 @@ class EngineClient(ABC):
         lora_request: Optional[LoRARequest] = None,
         trace_headers: Optional[Mapping[str, str]] = None,
         priority: int = 0,
+        tokenization_kwargs: Optional[dict[str, Any]] = None,
     ) -> AsyncGenerator[PoolingRequestOutput, None]:
         """Generate outputs for a request from a pooling model."""
         ...
@@ -320,7 +321,7 @@ class EngineClient(ABC):
         ...
 
     @abstractmethod
-    async def add_lora(self, lora_request: LoRARequest) -> None:
+    async def add_lora(self, lora_request: LoRARequest) -> bool:
         """Load a new LoRA adapter into the engine for future requests."""
         ...