[1/N] Initial prototype for multi-modal processor (#10044)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2024-11-13 12:39:03 +00:00
committed by GitHub
parent bb7991aa29
commit 0b8bb86bf1
48 changed files with 1132 additions and 436 deletions
+3 -1
View File
@@ -7,6 +7,7 @@ from vllm.envs import VLLM_ENABLE_V1_MULTIPROCESSING
from vllm.inputs import INPUT_REGISTRY, InputRegistry, PromptType
from vllm.logger import init_logger
from vllm.lora.request import LoRARequest
from vllm.multimodal import MULTIMODAL_REGISTRY, MultiModalRegistry
from vllm.outputs import RequestOutput
from vllm.pooling_params import PoolingParams
from vllm.prompt_adapter.request import PromptAdapterRequest
@@ -32,6 +33,7 @@ class LLMEngine:
usage_context: UsageContext = UsageContext.ENGINE_CONTEXT,
stat_loggers: Optional[Dict[str, StatLoggerBase]] = None,
input_registry: InputRegistry = INPUT_REGISTRY,
mm_registry: MultiModalRegistry = MULTIMODAL_REGISTRY,
use_cached_outputs: bool = False,
multiprocess_mode: bool = False,
) -> None:
@@ -50,7 +52,7 @@ class LLMEngine:
# Processor (convert Inputs --> EngineCoreRequests)
self.processor = Processor(vllm_config.model_config,
vllm_config.lora_config, self.tokenizer,
input_registry)
input_registry, mm_registry)
# Detokenizer (converts EngineCoreOutputs --> RequestOutput)
self.detokenizer = Detokenizer(