[Refactor] [7/N] to simplify the vLLM lora serving architecture (#32251)
Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
This commit is contained in:
@@ -1247,16 +1247,6 @@ StreamingResponsesResponse: TypeAlias = (
|
||||
)
|
||||
|
||||
|
||||
class LoadLoRAAdapterRequest(BaseModel):
|
||||
lora_name: str
|
||||
lora_path: str
|
||||
|
||||
|
||||
class UnloadLoRAAdapterRequest(BaseModel):
|
||||
lora_name: str
|
||||
lora_int_id: int | None = Field(default=None)
|
||||
|
||||
|
||||
## Protocols for Audio
|
||||
AudioResponseFormat: TypeAlias = Literal["json", "text", "srt", "verbose_json", "vtt"]
|
||||
|
||||
|
||||
@@ -10,10 +10,12 @@ from vllm.engine.protocol import EngineClient
|
||||
from vllm.entrypoints.openai.engine.protocol import (
|
||||
ErrorInfo,
|
||||
ErrorResponse,
|
||||
LoadLoRAAdapterRequest,
|
||||
ModelCard,
|
||||
ModelList,
|
||||
ModelPermission,
|
||||
)
|
||||
from vllm.entrypoints.serve.lora.protocol import (
|
||||
LoadLoRAAdapterRequest,
|
||||
UnloadLoRAAdapterRequest,
|
||||
)
|
||||
from vllm.logger import init_logger
|
||||
|
||||
Reference in New Issue
Block a user