transformers
FEAT / Optim: Add GaLore optimizer
#29588
Merged
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Overview
Commits
44
Changes
View On
GitHub
FEAT / Optim: Add GaLore optimizer
#29588
younesbelkada
merged 44 commits into
huggingface:main
from
younesbelkada:add-galore-optimizer
add galore v1
b31ce79c
muellerzr
commented on 2024-03-11
add import
58169f15
add tests and doc
9032635c
fix doctest
136f104c
younesbelkada
marked this pull request as ready for review
2 years ago
forward contrib credits from discussions
a5483b36
forward contrib credits from discussions
887d3adc
younesbelkada
requested a review
from
muellerzr
2 years ago
younesbelkada
requested a review
from
pacman100
2 years ago
younesbelkada
changed the title
DRAFT / Optim: Add GaLore optimizer
FEAT / Optim: Add GaLore optimizer
2 years ago
muellerzr
approved these changes on 2024-03-11
Apply suggestions from code review
d6f119fb
Merge remote-tracking branch 'upstream/main' into HEAD
3fae2290
fix failing tests'
c8c50f80
younesbelkada
requested a review
from
amyeroberts
2 years ago
BenjaminBossan
commented on 2024-03-11
amyeroberts
commented on 2024-03-11
pacman100
commented on 2024-03-12
Merge remote-tracking branch 'upstream/main' into add-galore-optimizer
2bdda681
switch to `optim_target_modules` and clarify docs
630bd13c
more clarification
a871b75a
Merge remote-tracking branch 'upstream/main' into add-galore-optimizer
cb6cd7e9
enhance lookup logic
51b7b292
update a test to add peak memory
3da3b90d
add regex, all-linear and single string support
9115c94d
add layer-wise optimization through DummyOptimizers and LRSchedulers
0b4ba838
forward contrib credits from discussions and original idea
3e5930ef
younesbelkada
requested a review
from
BenjaminBossan
2 years ago
younesbelkada
requested a review
from
pacman100
2 years ago
younesbelkada
requested a review
from
amyeroberts
2 years ago
younesbelkada
requested a review
from
muellerzr
2 years ago
add a section about DDP not supported in layerwise
a16d3a87
muellerzr
commented on 2024-03-13
Update src/transformers/trainer.py
29e7e94f
fix self
18ea144a
check only if layer_wise
7800bf1f
peterjc123
commented on 2024-03-13
peterjc123
commented on 2024-03-13
amyeroberts
commented on 2024-03-13
BenjaminBossan
commented on 2024-03-13
PenutChen
commented on 2024-03-14
Update src/transformers/training_args.py
e022bdda
oops
830c68df
make use of intervals
b640e980
clarify comment
14a89b2f
add matching tests
6f7102db
GaLoRe -> GaLore
c11cb63e
move to `get_scheduler`
36782019
add note on docs
fdc4b2a2
add a warning
e7ce9b7d
younesbelkada
requested a review
from
amyeroberts
2 years ago
younesbelkada
requested a review
from
BenjaminBossan
2 years ago
younesbelkada
requested a review
from
muellerzr
2 years ago
BenjaminBossan
approved these changes on 2024-03-14
adapt a bit the docs
91d64368
update docstring
b9e338a0
support original API
6ff37620
hiyouga
commented on 2024-03-17
younesbelkada
commented on 2024-03-17
Update docs/source/en/trainer.md
0d0440a1
BenjaminBossan
commented on 2024-03-18
slightly refactor
832f2be9
matthewdouglas
commented on 2024-03-18
Update docs/source/en/trainer.md
898a3c5a
amyeroberts
approved these changes on 2024-03-18
Update src/transformers/training_args.py
ed3ad4ad
fix args parsing and add tests
57e7096e
remove warning for regex
64ccfa6b
Merge remote-tracking branch 'upstream/main' into add-galore-optimizer
4413f074
fix type hint
73dcabb8
add note about extra args
1987b7ae
make `is_regex` return optional
db2bf219
younesbelkada
merged
f6261d7d
into main
2 years ago
Login to write a write a comment.
Login via GitHub
Reviewers
BenjaminBossan
amyeroberts
muellerzr
matthewdouglas
hiyouga
felladrin
PenutChen
kiddyboots216
peterjc123
pacman100
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub