[Ascend] Native Integration of MegatronAdaptor (FlagScale Module) on FlagOS#1226
Open
ltllt1 wants to merge 11 commits into
Open
[Ascend] Native Integration of MegatronAdaptor (FlagScale Module) on FlagOS#1226ltllt1 wants to merge 11 commits into
ltllt1 wants to merge 11 commits into
Conversation
ltllt1
requested review from
aoyulong,
heavyrain-lzy and
zhaoyinglia
as code owners
|
|
lxd-cumt
reviewed
Jun 18, 2026
| logger = logging.getLogger(__name__) | ||
|
|
||
|
|
||
| def get_device(local_rank=None): |
Collaborator
There was a problem hiding this comment.
Can this function be overridden?
Author
There was a problem hiding this comment.
thanks, I rewrote it separately and did not overwrite the original get_device implementation.
| @@ -0,0 +1,16 @@ | |||
| import time | |||
Collaborator
There was a problem hiding this comment.
I recommend creating a unified .py file under plugin_flagscale that contains all the Huawei-NPU specific function overrides, making them easier to manage centrally.
| @@ -0,0 +1,2 @@ | |||
| def get_device_arch_version(): | |||
Collaborator
There was a problem hiding this comment.
I recommend creating a unified .py file under plugin_flagscale that contains all the Huawei-NPU specific function overrides, making them easier to manage centrally.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
PR Category
Others
PR Types
New Features
PR Description
This PR optimizes the training entry of FlagScale, enabling native NPU training capability without introducing MegatronAdaptor dependencies.
1、Retain complete distributed training ability on NPU
2、No breaking changes to existing FlagScale training workflows