-
Notifications
You must be signed in to change notification settings - Fork 5.6k
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.
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
segment parallel parameter sync and grad sync #57061
Conversation
你的PR提交成功,感谢你对开源项目的贡献! |
and self._sharding_degree > 1 | ||
): | ||
# sharding may coexist with dp |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
如果sharding_degree > 1 and dp_degree > 1
时,这个逻辑跟之前不太一样?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
嗯, 这个之前写的不优雅;
sharding > 1 and dp > 1 改成在 shardding_parallel里处理
fleet_env.sharding_degree | ||
== fleet_env._hcg.get_sharding_parallel_world_size() | ||
) | ||
broadcast_sharding_parameters(model, fleet_env._hcg) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这个地方删除代码是为了什么?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
sharding > 1 and dp > 1 改成在 shardding_parallel里处理
python/paddle/distributed/fleet/meta_optimizers/dygraph_optimizer/hybrid_parallel_optimizer.py
Outdated
Show resolved
Hide resolved
… sep_grad_reduce
… sep_grad_reduce
… sep_grad_reduce
* sep grad all sync * polish * polish * polish * polish * add test * add test, and align with dp * follow comment * polish
PR types
New features
PR changes
Others
Description
paddle nlp llama 7b sep VS mp