xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-10 07:45:29 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Misc] Slight improve deepgemm print (#24085)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li 2025-09-05 00:06:51 +08:00 committed by GitHub
parent 83609ca91d
commit 94866d7c93
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/model_executor/layers/fused_moe/deep_gemm_moe.py
View File

@ -57,13 +57,14 @@ def _valid_deep_gemm(hidden_states: torch.Tensor, w1: torch.Tensor,
if not _valid_deep_gemm_shape(M, N, K): if not _valid_deep_gemm_shape(M, N, K):
logger.debug_once( logger.debug_once(
"DeepGemm disabled due to unaligned problem size. " "DeepGemm disabled due to unaligned problem size. "
"M: %s, N: %s, K: %s. M should >= align size " "M: %s, N: %s, K: %s. M should >= %s "
"and N and K must be multiples of %s." "and N and K must be multiples of %s. "
"This is not an error and we will fall back to triton.", "This is not an error and we will fall back to triton.",
M, M,
N, N,
K, K,
align, align,
align,
) )
return False return False
elif N <= 512: elif N <= 512: