Add Multimodal Processor Benchmark (#29105)

Signed-off-by: Reagan Lee <reaganjlee@gmail.com>
Signed-off-by: Reagan <reaganjlee@gmail.com>
This commit is contained in:
Reagan Lee
2026-01-01 19:26:53 -08:00
committed by GitHub
parent adcf682fc7
commit 1f5b7c41c3
13 changed files with 1035 additions and 222 deletions
+10 -6
View File
@@ -15,7 +15,7 @@ from vllm.multimodal import MULTIMODAL_REGISTRY, MultiModalRegistry
from vllm.multimodal.cache import processor_cache_from_config
from vllm.multimodal.inputs import MultiModalFeatureSpec, MultiModalUUIDDict
from vllm.multimodal.parse import MultiModalDataParser
from vllm.multimodal.processing import EncDecMultiModalProcessor
from vllm.multimodal.processing import EncDecMultiModalProcessor, set_request_id
from vllm.multimodal.utils import argsort_mm_positions
from vllm.pooling_params import PoolingParams
from vllm.sampling_params import SamplingParams
@@ -60,6 +60,7 @@ class InputProcessor:
self.input_preprocessor = InputPreprocessor(
self.model_config,
tokenizer,
self.vllm_config.observability_config,
mm_registry,
mm_processor_cache=self.mm_processor_cache,
)
@@ -493,11 +494,13 @@ class InputProcessor:
# 1. Tokenize text prompt, with LoRA request if one exists.
# 2. For multimodal models with a merged preprocessor, preprocess
# multimodal data and expand prompt token ids accordingly.
processed_inputs: ProcessorInputs = self.input_preprocessor.preprocess(
prompt,
tokenization_kwargs=tokenization_kwargs,
mm_uuids=mm_uuids,
)
with set_request_id(request_id):
processed_inputs: ProcessorInputs = self.input_preprocessor.preprocess(
prompt,
tokenization_kwargs=tokenization_kwargs,
mm_uuids=mm_uuids,
)
from vllm.platforms import current_platform
current_platform.validate_request(
@@ -641,6 +644,7 @@ class InputProcessor:
mm_registry = self.input_preprocessor.mm_registry
mm_processor = mm_registry.create_processor(
model_config,
self.vllm_config.observability_config,
tokenizer=tokenizer,
)
assert isinstance(mm_processor, EncDecMultiModalProcessor)