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.

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

Set sequence_parallel strategy true in default. #846

Merged

Conversation

GhostScreaming
Copy link
Contributor

  1. GPT with sequence_parallel has an improvement of performance in 8.76% ~ 20.37%.

image

image

  1. GPT with sequence_parallel has the same training loss curve compared with its model-parallel and pipeline-parallel peer.
    image

  2. Sequence_parallel strategy needs GPT mp_degree > 1. If mp_degree <= 1, it will be set false automatically in hybrid_model.py

Copy link
Collaborator

@sneaxiy sneaxiy left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

Copy link
Member

@ForFishes ForFishes left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@ForFishes ForFishes merged commit b6ed5ed into PaddlePaddle:develop Oct 25, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants