Support pinning adapter via server args. (#9249)

This commit is contained in:
Lifu Huang
2025-08-20 16:25:01 -07:00
committed by GitHub
parent 24eaebeb4b
commit b0980af89f
8 changed files with 162 additions and 55 deletions

View File

@@ -491,7 +491,7 @@ class SRTRunner:
tp_size: int = 1,
model_impl: str = "auto",
port: int = DEFAULT_PORT_FOR_SRT_TEST_RUNNER,
lora_paths: List[str] = None,
lora_paths: Optional[Union[List[str], List[dict[str, str]]]] = None,
max_loras_per_batch: int = 4,
attention_backend: Optional[str] = None,
prefill_attention_backend: Optional[str] = None,