[Core] Support Lora lineage and base model metadata management (#6315)

This commit is contained in:
Jiaxin Shan
2024-09-20 06:20:56 +00:00
committed by GitHub
parent 9e5ec35b1f
commit 260d40b5ea
15 changed files with 337 additions and 45 deletions
+3 -3
View File
@@ -14,7 +14,7 @@ from vllm.entrypoints.openai.protocol import (EmbeddingRequest,
EmbeddingResponse,
EmbeddingResponseData,
ErrorResponse, UsageInfo)
from vllm.entrypoints.openai.serving_engine import OpenAIServing
from vllm.entrypoints.openai.serving_engine import BaseModelPath, OpenAIServing
from vllm.logger import init_logger
from vllm.outputs import EmbeddingOutput, EmbeddingRequestOutput
from vllm.utils import merge_async_iterators, random_uuid
@@ -73,13 +73,13 @@ class OpenAIServingEmbedding(OpenAIServing):
self,
engine_client: EngineClient,
model_config: ModelConfig,
served_model_names: List[str],
base_model_paths: List[BaseModelPath],
*,
request_logger: Optional[RequestLogger],
):
super().__init__(engine_client=engine_client,
model_config=model_config,
served_model_names=served_model_names,
base_model_paths=base_model_paths,
lora_modules=None,
prompt_adapters=None,
request_logger=request_logger)