[Refactor] Make Renderer an abstract class (#33479)

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

vllm/v1/engine/async_llm.py

@@ -24,7 +24,7 @@ from vllm.multimodal import MULTIMODAL_REGISTRY, MultiModalRegistry
 from vllm.outputs import STREAM_FINISHED, PoolingRequestOutput, RequestOutput
 from vllm.plugins.io_processors import get_io_processor
 from vllm.pooling_params import PoolingParams
-from vllm.renderers import RendererLike, merge_kwargs
+from vllm.renderers import BaseRenderer, merge_kwargs
 from vllm.sampling_params import RequestOutputKind, SamplingParams
 from vllm.tasks import SupportedTask
 from vllm.tokenizers import TokenizerLike
@@ -844,7 +844,7 @@ class AsyncLLM(EngineClient):
         return self.input_processor.get_tokenizer()
 
     @property
-    def renderer(self) -> RendererLike:
+    def renderer(self) -> BaseRenderer:
         return self.input_processor.renderer
 
     async def is_tracing_enabled(self) -> bool:

vllm/v1/engine/input_processor.py

@@ -29,7 +29,7 @@ from vllm.multimodal.parse import ModalityDataItems, MultiModalDataItems
 from vllm.multimodal.processing.context import set_request_id
 from vllm.multimodal.utils import argsort_mm_positions
 from vllm.pooling_params import PoolingParams
-from vllm.renderers import RendererLike
+from vllm.renderers import BaseRenderer
 from vllm.sampling_params import _SAMPLING_EPS, SamplingParams
 from vllm.tokenizers import TokenizerLike
 from vllm.tokenizers.mistral import MistralTokenizer
@@ -96,7 +96,7 @@ class InputProcessor:
         return self.input_preprocessor.get_tokenizer()
 
     @property
-    def renderer(self) -> RendererLike:
+    def renderer(self) -> BaseRenderer:
         return self.input_preprocessor.renderer
 
     def _validate_logprobs(

vllm/v1/engine/llm_engine.py

@@ -21,7 +21,7 @@ from vllm.multimodal import MULTIMODAL_REGISTRY, MultiModalRegistry
 from vllm.outputs import PoolingRequestOutput, RequestOutput
 from vllm.plugins.io_processors import get_io_processor
 from vllm.pooling_params import PoolingParams
-from vllm.renderers import RendererLike
+from vllm.renderers import BaseRenderer
 from vllm.sampling_params import SamplingParams
 from vllm.tasks import SupportedTask
 from vllm.tokenizers import TokenizerLike
@@ -367,7 +367,7 @@ class LLMEngine:
         return self.input_processor.get_tokenizer()
 
     @property
-    def renderer(self) -> RendererLike:
+    def renderer(self) -> BaseRenderer:
         return self.input_processor.renderer
 
     def do_log_stats(self) -> None: