apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 08:37:06 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[NewFeatures] support eplb (#3547)

Browse Source 
* [NewFeatures] support eplb

* fix eplb
This commit is contained in:
xiaoxiaohehe001
2025-08-26 16:19:30 +08:00
committed by GitHub
parent 56e2d7e668
commit 9afa236e39
17 changed files with 174 additions and 67 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
fastdeploy/model_executor/load_weight_utils.py
View File

@@ -127,7 +127,11 @@ def load_ep_checkpoint(model_path: str, fd_config: FDConfig, return_numpy: bool
num_local_ffn_keys.append(down_proj_in_scale_key)
# for EP w4a8, we need all expert's activation_scale for up_gate_proj
for j in range(fd_config.model_config.moe_num_experts):
num_experts = fd_config.model_config.moe_num_experts
if isinstance(num_experts, list):
num_experts = num_experts[0]
for j in range(num_experts):
up_gate_proj_in_scale_key = f"ernie.layers.{i}.mlp.experts.{j}.up_gate_proj.activation_scale"
num_local_ffn_keys.append(up_gate_proj_in_scale_key)