EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

Fix refactor error - fp8.py (#5106)

Browse Source 
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
This commit is contained in:
HAI
2025-04-07 00:34:08 -07:00
committed by GitHub
parent f04c80dc42
commit 819924748a
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
python/sglang/srt/layers/quantization/fp8.py
View File

@@ -860,7 +860,7 @@ class Fp8MoEMethod:
layer.w13_weight_scale1[expert_id] *= max_w13_scales[expert_id] layer.w13_weight_scale1[expert_id] *= max_w13_scales[expert_id]
layer.w2_weight_scale1[expert_id] *= layer.w2_weight_scale[expert_id] layer.w2_weight_scale1[expert_id] *= layer.w2_weight_scale[expert_id]
def process_weights_hip_scale_padding(self, layer: Module, padding_size: int): def process_weights_hip_scale_padding(self, layer: Module):
from sglang.srt.layers.moe.fused_moe_triton.fused_moe import ( from sglang.srt.layers.moe.fused_moe_triton.fused_moe import (
padding_size, # Avoid circular import padding_size, # Avoid circular import
) )