support WOQ model input, such as kimi2.5 #1642
xin3he
merged
18 commits into
main
from
xinhe/3-31a
support WOQ model input, such as kimi2.5
0fbe8d82
Update auto_round/utils/weight_handler.py
62fe35ee
:qsupport compressed-tensors=0.14.0.1
2f863cc5
![pre-commit-ci[bot]](https://avatars.githubusercontent.com/in/68672?v=4&s=60)
[pre-commit.ci] auto fixes from pre-commit.com hooks
f5980b45
update xpu quantization cost (#1618)
20f90260
Refine CI (#1644)
253d1d3f
[Feature] Enhance dataset preprocessing memory management and fix has…
062e2752
Enhance quantization configuration support for mixed precision and sc…
017ac536
support weight.shape checking
e6f3009c
fix auto decompress during forward
2c7a49a2
support CompressedLinear layer type and sync serialization attributes…
886860de
Merge branch 'main' into xinhe/3-31a
995145c4
Merge branch 'main' into xinhe/3-31a
d16501f0
remove existed quantization_config
94c34ff0
add auto fallback
cb6c1507
fix line-too-long
4be61a71
yiliu30
commented
on 2026-04-14
yiliu30
commented
on 2026-04-14
update to fix CI
8c0f328d
lvliang-intel
commented
on 2026-04-14
lvliang-intel
commented
on 2026-04-14
yiliu30
approved these changes
on 2026-04-15
Merge branch 'main' into xinhe/3-31a
50f7dfd1
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub