Skip to content

b3274

Compare
Choose a tag to compare
@github-actions github-actions released this 01 Jul 17:37
49122a8
gemma2: add sliding window mask (#8227)

* gemma2: add sliding window mask

* fix data_swa uninitialized

* better naming

* add co-author

Co-authored-by: Arlo Phoenix <[email protected]>

* replace list with single tensor

* update

* llama : minor styling

* convert : add sanity check for query_pre_attn_scalar

* fix small typo in README

---------

Co-authored-by: Arlo Phoenix <[email protected]>
Co-authored-by: Georgi Gerganov <[email protected]>