Skip to content
Merged
Changes from all commits
Commits
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
7 changes: 7 additions & 0 deletions swift/megatron/trainers/base.py
Original file line number Diff line number Diff line change
Expand Up @@ -66,6 +66,13 @@ def __init__(self, args, template: Template):
self.args = args
self.template = template
self.prepare_model()
# Sync template.padding_free after prepare_model(), because _check_padding_free
# may override args.padding_free for certain models (e.g. DSA attention).
if template.padding_free != args.padding_free:
logger.warning(
f'template.padding_free({template.padding_free}) != args.padding_free({args.padding_free}), '
f'syncing template.padding_free to {args.padding_free}.')
template.padding_free = args.padding_free
self.optimizer, self.opt_param_scheduler = self.get_optimizer_and_scheduler()
self.data_collator = self._get_data_collator()

Expand Down
Loading