minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

Update logging for inplace setting in MoE layer (#25499)

Browse Source
This commit is contained in:
Cheng Wan
2026-05-16 17:12:11 -07:00
committed by GitHub
parent 0c017db916
commit 229cadec04
1 changed files with 4 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/sglang/srt/layers/moe/fused_moe_triton/layer.py
+4 -3
View File
@@ -312,9 +312,10 @@ class FusedMoE(torch.nn.Module):
get_moe_runner_backend().is_flashinfer_trtllm_routed() get_moe_runner_backend().is_flashinfer_trtllm_routed()
or get_moe_runner_backend().is_flashinfer_trtllm() or get_moe_runner_backend().is_flashinfer_trtllm()
): ):
logging.warning( if self.moe_runner_config.inplace:
"Setting inplace to False for FlashInfer TRTLLM MoE backend." logging.info(
) "Setting inplace to False for FlashInfer TRTLLM MoE backend."
)
self.moe_runner_config.inplace = False self.moe_runner_config.inplace = False
self.should_fuse_routed_scaling_factor_in_topk = ( self.should_fuse_routed_scaling_factor_in_topk = (