[feat] support fa3 backend for pd disaggregated (#2695)
Some checks failed
Deploy GitHub Pages / deploy (push) Has been cancelled

* support fa3 backend run in pd disaggregated

* support fa3 backend run in pd disaggregated

* support fa3 backend run in pd disaggregated

* support fa3 backend run in pd disaggregated

* delete use_fast_ffn
This commit is contained in:
Yuanle Liu
2025-07-03 22:33:27 +08:00
committed by GitHub
parent 00863c43fd
commit 240bdac2a4
26 changed files with 455 additions and 139 deletions

View File

@@ -86,7 +86,7 @@ class DefaultModelLoader(BaseModelLoader):
if isinstance(v, paddle.Tensor):
v.value().get_tensor()._clear()
paddle.device.cuda.empty_cache()
paddle.device.cuda.synchronize()
paddle.device.synchronize()
def load_model(self, fd_config: FDConfig) -> nn.Layer:
context = paddle.LazyGuard()