Skip to content

Conversation

@gshtras
Copy link
Collaborator

@gshtras gshtras commented Aug 12, 2024

Adding changes required to support LLama 3.1:
vllm-project#6553
vllm-project#6693

@gshtras gshtras requested a review from shajrawi August 12, 2024 17:24
Copy link
Collaborator

@shajrawi shajrawi left a comment

Choose a reason for hiding this comment

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

ship it

@gshtras gshtras merged commit dd1a208 into main Aug 12, 2024
@gshtras gshtras deleted the llama3.1 branch August 12, 2024 20:45
@JArnoldAMD JArnoldAMD mentioned this pull request Aug 26, 2024
shajrawi pushed a commit that referenced this pull request Aug 26, 2024
* Add support for a rope extension method (vllm-project#6553)

* [BugFix] Fix RoPE error in Llama 3.1 (vllm-project#6693)

---------

Co-authored-by: Simon Mo <[email protected]>
Co-authored-by: Woosuk Kwon <[email protected]>
shajrawi pushed a commit that referenced this pull request Aug 27, 2024
* optimizations for process output step

* Llama3.1 (#129)

* Add support for a rope extension method (vllm-project#6553)

* [BugFix] Fix RoPE error in Llama 3.1 (vllm-project#6693)

---------

Co-authored-by: Simon Mo <[email protected]>
Co-authored-by: Woosuk Kwon <[email protected]>

* Update hipblaslt and FA revs to match what was used for MLPerf

* Switch to "unified docker" with a ROCm 6.2 base image

This base image includes current libraries, so there is no need for
us to rebuild hipblaslt, RCCL, and Flash Attention.

---------

Co-authored-by: Shomy <[email protected]>
Co-authored-by: Gregory Shtrasberg <[email protected]>
Co-authored-by: Simon Mo <[email protected]>
Co-authored-by: Woosuk Kwon <[email protected]>
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.

5 participants