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

[LoRA] Roadmap of LoRA operators #199

Open
3 tasks
yzh119 opened this issue Apr 8, 2024 · 1 comment
Open
3 tasks

[LoRA] Roadmap of LoRA operators #199

yzh119 opened this issue Apr 8, 2024 · 1 comment

Comments

@yzh119
Copy link
Collaborator

yzh119 commented Apr 8, 2024

  1. Reducing the latency of LoRA operators (per lorax feedback, lora operators introduce ~20% overhead).
  2. Numerical issue of LoRA operators for large batch size.
  3. Using fp8 tensor cores for LoRA operators.
@tgaddair
Copy link

Thanks for filing this issue @yzh119! Happy to help out in any way I can.

@yzh119 yzh119 mentioned this issue Jun 4, 2024
Merged
yzh119 added a commit that referenced this issue Jun 5, 2024
@yzh119
feat: add group gemm operators (#282)
e08ba42 
First step towards #199 .

Group gemm should also be helpful for MoE.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@tgaddair @yzh119