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.

Sign up for GitHub

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

Jump to bottom

fix conflict of param grouping and filter bn bias #602

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix conflict of param grouping and filter bn bias #602

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 18 additions & 2 deletions mindcv/optim/optim_factory.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,9 @@


def init_group_params(params, weight_decay):
"""
Filter bias and norm layers including nn.BatchNorm and nn.LayerNorm from weight decay.
"""
decay_params = []
no_decay_params = []

Expand Down Expand Up @@ -60,7 +63,9 @@ def create_optimizer(
momentum: momentum if the optimizer supports. Default: 0.9.
nesterov: Whether to use Nesterov Accelerated Gradient (NAG) algorithm to update the gradients. Default: False.
filter_bias_and_bn: whether to filter batch norm parameters and bias from weight decay.
If True, weight decay will not apply on BN parameters and bias in Conv or Dense layers. Default: True.
If True and `param` is of type list[Parameter] (no param group defined), weight decay
will not be applied on parameters containing "beta", "gamma", and "bias", i.e., BatchNorm
and LayerNorm parameters and bias in Conv or Dense layers. Default: True.
loss_scale: A floating point value for the loss scale, which must be larger than 0.0. Default: 1.0.

Returns:
Expand All @@ -69,8 +74,19 @@ def create_optimizer(

opt = opt.lower()

# check whether param grouping strategy is encoded in `params`
customized_param_group = False
if isinstance(params[0], dict):
customized_param_group = True

if weight_decay and filter_bias_and_bn:
params = init_group_params(params, weight_decay)
if not customized_param_group:
params = init_group_params(params, weight_decay)
else:
print(
"WARNING: Customized param grouping startegy detected in `params`."
"filter_bias_and_bn (default=True) will be disabled"
)

opt_args = dict(**kwargs)
# if lr is not None:
Expand Down