Skip to content

fix: is_offload_optimizer_states_in_train_step not passed to batch.meta_info in RLVR and agentic pipelines #478

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

Open
sanmuf wants to merge 1 commit into
base: main
Choose a base branch
from
+2 −0
Open

fix: is_offload_optimizer_states_in_train_step not passed to batch.meta_info in RLVR and agentic pipelines #478

Show file tree
Hide file tree
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
1 change: 1 addition & 0 deletions roll/pipeline/agentic/agentic_pipeline.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -314,6 +314,7 @@ def run(self):
batch.meta_info["global_step"] = global_step
batch.meta_info["_broadcast_non_tensor_batch"] = True
batch.meta_info["loss_mask_keys"] = ["response_mask"]
batch.meta_info["is_offload_optimizer_states_in_train_step"] = self.pipeline_config.is_offload_optimizer_states_in_train_step

if val_future is not None:
val_metrics = val_future.result()
Expand Down
1 change: 1 addition & 0 deletions roll/pipeline/rlvr/rlvr_pipeline.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -570,6 +570,7 @@ def run(self):
batch.meta_info["global_step"] = global_step
batch.meta_info["_broadcast_non_tensor_batch"] = True
batch.meta_info["loss_mask_keys"] = ['response_mask', 'final_response_mask']
batch.meta_info["is_offload_optimizer_states_in_train_step"] = self.pipeline_config.is_offload_optimizer_states_in_train_step
batch.non_tensor_batch['sample_uuid'] = np.array([str(uuid.uuid4()) for _ in range(batch.batch.shape[0])], dtype=object)
batch.batch["prompt_id"] = torch.arange(batch.batch.batch_size[0], device=batch.batch.device)

Expand Down
Loading