AQLM quantizer support #28928
ArthurZucker
merged
24 commits into
huggingface:main
from
BlackSamorez:aqlm
aqlm init
30075e26
calibration and dtypes
bd15281a
docs
d5ab4bc5
younesbelkada
commented
on 2024-02-08
younesbelkada
commented
on 2024-02-08
younesbelkada
commented
on 2024-02-08
younesbelkada
commented
on 2024-02-08
younesbelkada
commented
on 2024-02-08
younesbelkada
commented
on 2024-02-08
younesbelkada
commented
on 2024-02-08
Readme update
b7051b6e
is_aqlm_available
fa8ca68f
Simpler link in docs
e701def6
Test TODO real reference
fa45f370
init _import_structure fix
1c632a5d
AqlmConfig autodoc
0fb92a6a
integration aqlm
92e72504
integrations in tests
8fe4d208
docstring fix
8934fed2
legacy typing
2ea73c45
Less typings
95a4de4b
More kernels information
0f2165dc
Performance -> Accuracy
35153ebc
correct tests
d2b53c06
remoced multi-gpu test
4bda42c0
younesbelkada
approved these changes
on 2024-02-12
ArthurZucker
approved these changes
on 2024-02-13
younesbelkada
approved these changes
on 2024-02-13
Update docs/source/en/quantization.md
579e110a
Update src/transformers/utils/quantization_config.py
d8058e78
Brought back multi-gpu tests
19d23576
Merge branch 'aqlm' of github.com:BlackSamorez/transformers into aqlm
c8621cc1
Update src/transformers/integrations/aqlm.py
dac3a67d
Update tests/quantization/aqlm_integration/test_aqlm.py
f2e0ed3a
younesbelkada
approved these changes
on 2024-02-14
Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub