DeepSeek/EPLB
1
0
Fork 0
mirror of https://github.com/deepseek-ai/EPLB synced 2025-06-26 18:15:49 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

fix function return type

Browse Source
This commit is contained in:
Shaoyuan CHEN
2025-02-27 17:35:49 +08:00
committed by GitHub
parent f9bc62e841 48c86cb16b
commit cf221658a7
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
eplb.py
View File

@@ -42,7 +42,7 @@ def balanced_packing(weight: torch.Tensor, num_packs: int) -> Tuple[torch.Tensor
return pack_index, rank_in_pack
def replicate_experts(weight: torch.Tensor, num_phy: int) -> torch.Tensor:
def replicate_experts(weight: torch.Tensor, num_phy: int) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
"""
Replicate `num_log` experts to `num_phy` replicas, such that the maximum load of all replicas is minimized.