apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-26 18:10:32 +08:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

[Feature] Support_eplb (#2997)
Some checks failed
Deploy GitHub Pages / deploy (push) Has been cancelled
Details

Browse Source 
* [Feature] support_eplb

* [Feature] support_eplb

* [Fix] fix mm ep
This commit is contained in:
xiaoxiaohehe001
2025-07-24 20:22:45 +08:00
committed by GitHub
parent f37d00e856
commit 2970b00dfa
11 changed files with 118 additions and 50 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
fastdeploy/model_executor/layers/backends/gcu/moe/fused_moe_method_gcu_backend.py
View File

@@ -276,7 +276,7 @@ class GCUWeightOnlyMoEMethod(GCUFusedMoeMethod):
up_gate_proj_expert_weight_scale_key = layer.weight_key_map.get("up_gate_proj_expert_weight_scale_key", None)
down_proj_expert_weight_scale_key = layer.weight_key_map.get("down_proj_expert_weight_scale_key", None)
up_gate_proj_weights, down_proj_weights = layer.load_experts_weight(
up_gate_proj_weights, down_proj_weights, _ = layer.load_experts_weight(
state_dict,
up_gate_proj_expert_weight_key,
down_proj_expert_weight_key,