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

Avoid using lr from checkpoint. #1781

Merged
merged 2 commits into from
Oct 27, 2024
Merged

Avoid using lr from checkpoint. #1781

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
49a30f7
Avoid using lr from checkpoint.
csukuangfj Oct 24, 2024
28791cd
fix a typo
csukuangfj Oct 24, 2024
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
9 changes: 8 additions & 1 deletion egs/librispeech/ASR/zipformer/optim.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -787,7 +787,9 @@ def state_dict(self):
is not the optimizer.
"""
return {
"base_lrs": self.base_lrs,
# the user might try to override the base_lr, so don't include this in the state.
# previously they were included.
# "base_lrs": self.base_lrs,
"epoch": self.epoch,
"batch": self.batch,
}
Expand All @@ -799,7 +801,12 @@ def load_state_dict(self, state_dict):
state_dict (dict): scheduler state. Should be an object returned
from a call to :meth:`state_dict`.
"""
# the things with base_lrs are a work-around for a previous problem
# where base_lrs were written with the state dict.
base_lrs = self.base_lrs
self.__dict__.update(state_dict)
self.base_lrs = base_lrs


def get_last_lr(self) -> List[float]:
"""Return last computed learning rate by current scheduler. Will be a list of float."""
Expand Down
Loading