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

fix q2k bug (#599)
Heng Guo committed 1 year ago
f41bb3d
update readme (#597)
Wenhua Cheng committed 1 year ago
45b312d
rm unnecessary forward to improve speed (#593)
Wenhua Cheng committed 1 year ago
23d40a0
dc6d389
fix cuda ut (#592)
Wenhua Cheng committed 1 year ago
4c3dc07
6783fc5
add recipes for deepseek-r1-0528 (#588)
Heng Guo committed 1 year ago
9263cc8
Update gguf_accuracy (q3_ks) (#590)
SinpackKonmakan committed 1 year ago
da9efc2
fix device bug at calibration (#587)
Wenhua Cheng committed 1 year ago
23c71e3
e075d43
fix awq exporting at group_size=-1 (#579)
Wenhua Cheng committed 1 year ago
2bb57c2
add torch backend (#555)
Weiwei committed 1 year ago
ce8eedd
switch to transformers API in cpu ut (#580)
Wenhua Cheng committed 1 year ago
6ceaa8a
improve the robustness of parsing vlm config (#577)
Wenhua Cheng committed 1 year ago
69749b9
add gguf accuracy data (#574)
Wenhua Cheng committed 1 year ago
3e79ebc
support for llava-based hf model (#568)
Heng Guo committed 1 year ago
9148737
update readme (#571)
Wenhua Cheng committed 1 year ago
1bad5ee
update readme (#569)
Wenhua Cheng committed 1 year ago
90fef62
update habana docker to 1.21.0 (#566)
Sun, Xuehao committed 1 year ago
9a23061
fix typo in backend (#564)
Wenhua Cheng committed 1 year ago
eee6db4
support xpu in triton backend (#563)
Wenhua Cheng committed 1 year ago
f683f99
add UE5M3 simulation (#562)
Wenhua Cheng committed 1 year ago
fa41c3d
itrex requires torch<2.7 (#548)
Sun, Xuehao committed 1 year ago
2e842a2
add recipes for qwen3 8b and 14b (#552)
Heng Guo committed 1 year ago
6a6b590