Support DeepSeek V3.2 Exp (#11061)

Co-authored-by: Stefan He <11166516+hebiao064@users.noreply.github.com>
Co-authored-by: Liangsheng Yin <95566987+hnyls2002@users.noreply.github.com>
Co-authored-by: Baizhou Zhang <56809903+fridge003@users.noreply.github.com>
Co-authored-by: DarkSharpness <76582120+darksharpness@users.noreply.github.com>
Co-authored-by: ZhengdQin <46387172+zhengdqin@users.noreply.github.com>
Co-authored-by: DarkSharpness <2040703891@qq.com>
Co-authored-by: hnyls2002 <lsyincs@gmail.com>
Co-authored-by: Zhengda Qin <zhengdqin@gmail.com>
Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com>
Co-authored-by: HAI <hixiao@gmail.com>
Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com>
This commit is contained in:
fzyzcjy
2025-10-06 15:24:15 +08:00
committed by GitHub
parent 292a867ad9
commit efbc687c28
29 changed files with 4540 additions and 139 deletions

View File

@@ -471,7 +471,7 @@ def is_pin_memory_available() -> bool:
class LayerFn(Protocol):
def __call__(self, layer_id: int, prefix: str) -> torch.nn.Module: ...
def __call__(self, idx: int, prefix: str) -> torch.nn.Module: ...
def make_layers(
@@ -482,7 +482,7 @@ def make_layers(
prefix: str = "",
return_tuple: bool = False,
offloader_kwargs: Dict[str, Any] = {},
) -> Tuple[int, int, torch.nn.ModuleList]:
) -> Tuple[torch.nn.Module, int, int]:
"""Make a list of layers with the given layer function"""
# circula imports
from sglang.srt.distributed import get_pp_indices