Skip to content

[megatron, recipe] fix: error of megatron init while detached actor and rollout #4179

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

Merged
ISEEKYAN merged 6 commits into from
Nov 19, 2025
Merged

[megatron, recipe] fix: error of megatron init while detached actor and rollout #4179

Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
657d0db
fix error of megatron init while async train
lalala-2 Nov 17, 2025
7679cc9
fix error of param_sync in offload param
lalala-2 Nov 18, 2025
f73ce84
unified with the above
lalala-2 Nov 18, 2025
f9772e1
fix
lalala-2 Nov 18, 2025
3c555c1
fix error of one_step_off_policy
lalala-2 Nov 18, 2025
c88cc88
one_step_off support param offload
lalala-2 Nov 18, 2025
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
Prev Previous commit
Next Next commit
unified with the above
  • Loading branch information
@lalala-2
lalala-2 committed Nov 18, 2025
commit f73ce849890f74a60f2585da75eb1deefc54c87b
2 changes: 1 addition & 1 deletion recipe/fully_async_policy/megatron_worker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,7 +91,7 @@ def sync_rollout_weights(self):
if self._is_rollout:
inference_model.load_weights([(key, tensor)])
if self._is_actor and self._is_offload_param:
offload_megatron_model_to_cpu(self.actor.actor_module)
offload_megatron_model_to_cpu(self.actor_module)

@register(dispatch_mode=Dispatch.ONE_TO_ALL)
def save_model_to_cpu(self, n):
Expand Down
Loading