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

move load state dict after initialize parallel state in nlp_model #9382

Merged
merged 4 commits into from
Jun 15, 2024
Merged

move load state dict after initialize parallel state in nlp_model #9382

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
f5c481b
move load state dict after initialize parallel state
Jun 5, 2024
db27ef8
delay sharded_state_dict in save_to
Jun 7, 2024
2d3bb2f
Merge branch 'main' into fix_load_dist_ckpt
ryxli Jun 12, 2024
e205623
Merge branch 'main' into fix_load_dist_ckpt
ryxli Jun 13, 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
4 changes: 2 additions & 2 deletions nemo/collections/nlp/models/nlp_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -387,8 +387,6 @@ def load_from_checkpoint(

# if the checkpoint is distributed, we deferred loading the state_dict until now
if checkpoint_dir is not None:
sharded_state_dict = model.sharded_state_dict()
checkpoint['state_dict'] = sharded_state_dict
# dist checkpointing needs torch.distributed to load the checkpoint
if not parallel_state.is_initialized():

Expand All @@ -398,6 +396,8 @@ def dummy():
if model.trainer.strategy.launcher is not None:
model.trainer.strategy.launcher.launch(dummy, trainer=model.trainer)
model.trainer.strategy.setup_environment()
sharded_state_dict = model.sharded_state_dict()
checkpoint['state_dict'] = sharded_state_dict
# load the checkpoint from disk
checkpoint = dist_checkpointing.load(sharded_state_dict=checkpoint, checkpoint_dir=checkpoint_dir)
# restore the weights
Expand Down
3 changes: 1 addition & 2 deletions nemo/collections/nlp/parts/nlp_overrides.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -943,8 +943,6 @@ def save_to(self, model, save_path: str):
if dist_ckpt:
# model weights is a directory
dist_ckpt_dir = ckpt_to_dir(os.path.join(dir_name, self.model_weights_ckpt))

sharded_state_dict = model.sharded_state_dict()
# dist checkpoint needs torch.distributed to save the checkpoint
if not parallel_state.is_initialized():

Expand All @@ -954,6 +952,7 @@ def dummy():
if model.trainer.strategy.launcher is not None:
model.trainer.strategy.launcher.launch(dummy, trainer=model.trainer)
model.trainer.strategy.setup_environment()
sharded_state_dict = model.sharded_state_dict()
checkpoint_io = DistributedCheckpointIO(model.cfg.get('dist_ckpt_format', 'zarr'))
checkpoint_io.save_checkpoint(sharded_state_dict, dist_ckpt_dir)

Expand Down
Loading