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.

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

Jump to bottom

Align GLU implementation with LLaMa #829

Merged
merged 2 commits into from
Jan 1, 2024
Merged

Align GLU implementation with LLaMa #829

merged 2 commits into from
Jan 1, 2024

Conversation

vchiley
Copy link
Contributor

@vchiley vchiley commented Dec 31, 2023 •
edited
Loading

Align LLM-Foundry GLU implementation with LLaMa implementation
ie

self.down_proj(self.act(self.gate_proj(x)) * self.up_proj(x))

instead of

self.down_proj(self.act(self.up_proj(x)) * self.gate(x))

Additionally, with #805, we are no longer constricted to using gelus. Therefore it's no longer GeGLU, but ANY_ACT_FN_GLU, so I'm changing the name to generic GLU.

@vchiley vchiley self-assigned this Dec 31, 2023
@vchiley vchiley merged commit 1ac5c4b into mosaicml:main Jan 1, 2024
10 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@dakinggg dakinggg dakinggg approved these changes

@sashaDoubov sashaDoubov Awaiting requested review from sashaDoubov

@mvpatel2000 mvpatel2000 Awaiting requested review from mvpatel2000

Assignees

@vchiley vchiley

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@vchiley @dakinggg