[Feature] overlap LoRA weight loading with compute (#15512)

This commit is contained in:
Glen Liu
2026-01-19 10:43:17 +08:00
committed by GitHub
parent 51f147ada3
commit ad1b4e4728
13 changed files with 302 additions and 21 deletions
+2
View File
@@ -552,6 +552,7 @@ class SRTRunner:
max_lora_rank: Optional[int] = None,
lora_target_modules: Optional[List[str]] = None,
enable_lora: Optional[bool] = None,
enable_lora_overlap_loading: Optional[bool] = None,
max_loaded_loras: Optional[int] = None,
json_model_override_args: Optional[dict[str, Any]] = None,
lora_eviction_policy: str = "lru",
@@ -612,6 +613,7 @@ class SRTRunner:
max_lora_rank=max_lora_rank,
lora_target_modules=lora_target_modules,
enable_lora=enable_lora,
enable_lora_overlap_loading=enable_lora_overlap_loading,
max_loaded_loras=max_loaded_loras,
json_model_override_args=(
json.dumps(json_model_override_args)