minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 13 Packages Projects Releases Wiki Activity

Fix logging for inplace setting in the flashInfer-trtllm backend (#25522)

Browse Source
This commit is contained in:
Cheng Wan
2026-05-17 02:20:14 -07:00
committed by GitHub
parent e547f3f804
commit 568ba7216a
1 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/sglang/srt/layers/moe/fused_moe_triton/layer.py
+2 -1
View File
@@ -70,6 +70,7 @@ from sglang.srt.utils import (
get_bool_env_var, get_bool_env_var,
is_cpu, is_cpu,
is_hip, is_hip,
print_info_once,
round_up, round_up,
) )
from sglang.srt.utils.custom_op import register_custom_op from sglang.srt.utils.custom_op import register_custom_op
@@ -313,7 +314,7 @@ class FusedMoE(torch.nn.Module):
or get_moe_runner_backend().is_flashinfer_trtllm() or get_moe_runner_backend().is_flashinfer_trtllm()
): ):
if self.moe_runner_config.inplace: if self.moe_runner_config.inplace:
logging.info( print_info_once(
"Setting inplace to False for FlashInfer TRTLLM MoE backend." "Setting inplace to False for FlashInfer TRTLLM MoE backend."
) )
self.moe_runner_config.inplace = False self.moe_runner_config.inplace = False