xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-27 20:30:54 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] Fix Machete zero point issue for GPTQ models on SM90 (#21066)

Browse Source 
Signed-off-by: mgoin <mgoin64@gmail.com>
This commit is contained in:
Michael Goin 2025-07-16 22:54:45 -04:00 committed by GitHub
parent 58760e12b1
commit 28a6d5423d
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/model_executor/layers/quantization/kernels/mixed_precision/machete.py
View File

@ -126,6 +126,11 @@ class MacheteLinearKernel(MPLinearKernel):
if c.has_g_idx:
x_2d = self.act_perm(x_2d)
if c.zero_points:
assert w_zp is not None
else:
w_zp = None
output = ops.machete_mm(a=x_2d,
b_q=w_q,
b_type=c.weight_type,