Skip to content

[PageKVCache] Add support of sinks params and fix sliding windnow - #20104

Open
krishnaraj36 wants to merge 2 commits into
apache:mainfrom
krishnaraj36:gpt_oss_kvCache
Open

[PageKVCache] Add support of sinks params and fix sliding windnow#20104
krishnaraj36 wants to merge 2 commits into
apache:mainfrom
krishnaraj36:gpt_oss_kvCache

Conversation

@krishnaraj36

Copy link
Copy Markdown
Contributor
  1. Added support of sinks param in attention funtion which is introduced in GPT_OSS model.
  2. Fix the layerwise sliding window logic in PageKVCache manager.

1. Added support of sinks param in attention funtion which is introduced in
GPT_OSS model.
2. Fix the layerwise sliding window logic in PageKVCache manager.
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