accelerate
[`core`] Add Quantization support for `dispatch_model`
#1237

Merged

[`core`] Add Quantization support for `dispatch_model` #1237

sgugger merged 8 commits into huggingface:main from younesbelkada:fix-dispatch-int9

add quantization support for `dispatch_model`

c0b02479

fix multi-gpu

06d6e11c

more chaecks

972a426f

younesbelkada requested a review from

sgugger 3 years ago

younesbelkada requested a review from

pacman100 3 years ago

BlackSamorez commented on 2023-03-23

fix bias issue

f7e9fd34

BlackSamorez commented on 2023-03-23

sgugger commented on 2023-03-23

BlackSamorez commented on 2023-03-24

Update src/accelerate/utils/modeling.py

bfd7e1af

make style

22dae580

add tests

bba597ce

younesbelkada requested a review from

sgugger 3 years ago

left some todos

e8326ba6

sgugger approved these changes on 2023-03-27

sgugger merged 23ac55fc into main 3 years ago

younesbelkada deleted the fix-dispatch-int9 branch 3 years ago

Reviewers

sgugger

BlackSamorez

pacman100

Assignees

No one assigned

Labels

None yet

Milestone

No milestone