biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Feature][Quantization] MXFP4 support for MOE models (#17888)

Browse Source 
Signed-off-by: Felix Marty <felmarty@amd.com>
Signed-off-by: Bowen Bao <bowenbao@amd.com>
Signed-off-by: Felix Marty <Felix.Marty@amd.com>
Co-authored-by: Bowen Bao <bowenbao@amd.com>
This commit is contained in:
fxmarty-amd
2025-07-09 22:19:02 +02:00
committed by GitHub
parent bf03ff3575
commit 332d4cb17b
15 changed files with 873 additions and 104 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
tests/kernels/moe/test_moe.py
View File

@@ -174,6 +174,7 @@ def test_fused_moe(
use_int8_w8a8=False,
use_int8_w8a16=False,
use_int4_w4a16=False,
use_mxfp4_w4a4=False,
per_act_token_quant=False,
block_shape=None)