update m // 8

cf975626 · zhuwenwen · 3fcdea26 · cf975626
Commit cf975626 authored Nov 20, 2025 by zhuwenwen
Hide whitespace changes
Inline Side-by-side

Showing with 1 addition and 1 deletion

vllm/model_executor/layers/quantization/slimquant_w4a8.py vllm/model_executor/layers/quantization/slimquant_w4a8.py +1 -1

No files found.
--- a/vllm/model_executor/layers/quantization/slimquant_w4a8.py
+++ b/vllm/model_executor/layers/quantization/slimquant_w4a8.py
@@ -178,7 +178,7 @@ class SlimQuantW4A8Int8LinearMethod(LinearMethodBase):
                elif m<=64:
                    m_= (m + 3) & -4 #取值到最近的4的倍数
                elif m<=160:
-                    m_=(m + 7) & -8
+                    m_ = (m // 8) * 8
                    
                elif m<200: #256
                    m_=160