Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature] Implement Adalite optimizer #229

Merged
merged 12 commits into from
Apr 7, 2024
Prev Previous commit
Next Next commit
update: recipes
  • Loading branch information
kozistr committed Apr 7, 2024
commit e0acc05373093760219996707e8c7e12682acd93
4 changes: 4 additions & 0 deletions tests/constants.py
Original file line number Diff line number Diff line change
@@ -23,6 +23,7 @@
AdaFactor,
AdaHessian,
Adai,
Adalite,
AdaMax,
AdaMod,
AdamP,
@@ -120,6 +121,8 @@
'padam',
'came',
'aida',
'galore',
'adalite',
]

VALID_LR_SCHEDULER_NAMES: List[str] = [
@@ -434,6 +437,7 @@
{'lr': 5e-1, 'weight_decay': 1e-3, 'rank': 2, 'scale': 1.0, 'update_proj_gap': 2, 'projection_type': 'full'},
5,
),
(Adalite, {'lr': 1e0, 'weight_decay': 1e-3}, 5),
]
ADANORM_SUPPORTED_OPTIMIZERS: List[Tuple[Any, Dict[str, Union[float, bool, int]], int]] = [
(AdaBelief, {'lr': 5e-1, 'weight_decay': 1e-3, 'adanorm': True}, 10),