[Core] Implement sharded state loader (#4690)

Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Aurick Qiao
2024-05-15 22:11:54 -07:00
committed by GitHub
co-authored by Woosuk Kwon
parent 52f8107cf2
commit 30e754390c
7 changed files with 351 additions and 0 deletions
+11
View File
@@ -77,6 +77,17 @@ class DistributedGPUExecutor(GPUExecutor):
def list_loras(self) -> Set[int]:
return self._run_workers("list_loras")
def save_sharded_state(
self,
path: str,
pattern: Optional[str] = None,
max_size: Optional[int] = None,
) -> None:
self._run_workers("save_sharded_state",
path=path,
pattern=pattern,
max_size=max_size)
@abstractmethod
def _run_workers(
self,