Skip to content

More gpu memory saving for llama#20

Open
sfc-gh-zhwang wants to merge 24 commits intocorvofrom
zhwang/more-mem
Open

More gpu memory saving for llama#20
sfc-gh-zhwang wants to merge 24 commits intocorvofrom
zhwang/more-mem

Conversation

@sfc-gh-zhwang
Copy link
Collaborator

@sfc-gh-zhwang sfc-gh-zhwang commented Oct 1, 2023

Get rid of the qkv_buf_tmp_ -> qkv_buf_ with repeat_kv hack in llamacontextdecoding to save gpu memory.

@sfc-gh-zhwang sfc-gh-zhwang changed the title Zhwang/more mem More gpu memory saving for llama Oct 1, 2023
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.

1 participant