Skip to content

PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度

Commits on main

417d87d
add opt_rtn support for mxfp4 and nvfp4 (#2100)
Wenhua Cheng committed 3 months ago
41cca72
2a267e7
378187c
ca9524a
212c8bd
8b0923c
refine alg and support lfq (#2082)
Wenhua Cheng committed 3 months ago
f1af646
ark: add SpargeAttn support (#2027)
Yi Liu committed 3 months ago
5ce9733
Decoupling oneDNN (#2049)
Zhenzhong Xu committed 3 months ago
54b4b84
0212c16
Temporarily skip the issue in the CT (#2093)
Xin He committed 3 months ago
33ec4c7
8f8c8e9
Add CUDA CI coverage (#2074)
Sun, Xuehao committed 4 months ago
6ff414b
Stabilize MoE prefill accuracy test (#2060)
Dong, Bo committed 4 months ago
98ad266
c774c83
Remove transformers version limit (#2051)
chen, suyue committed 4 months ago
0ea53be
6d1c6db
Update torch version to 2.13.0 (#2045)
Sun, Xuehao committed 4 months ago
32da6a9