apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-04 08:16:42 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Fix w4a8 scale load (#3334)

Browse Source 
* fix_eplb

* fix eplb part3

* support_fp8_rope3d

* fix w4a8 scale
This commit is contained in:
xiaoxiaohehe001
2025-08-11 21:02:42 +08:00
committed by GitHub
parent c7993d35cb
commit 4dbaa3d74c
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
fastdeploy/model_executor/layers/moe/fused_moe_cutlass_backend.py
View File

@@ -361,7 +361,7 @@ class CutlassW4A8MoEMethod(CutlassMoEMethod):
if layer.ep_size > 1:
for expert_idx in ep_rank_to_expert_id_list:
scale_tensor = get_tensor(
get_tensor(
(
state_dict[up_gate_proj_expert_in_scale_key.format(expert_idx)]
if up_gate_proj_expert_in_scale_key.format(expert_idx) in state_dict
else up_gate_proj_expert_in_scale_key.format(expert_idx)