memory : add llama_memory_hybrid_iswa#18601
Merged
ggerganov merged 2 commits intoggml-org:masterfrom Jan 21, 2026
Merged
Conversation
5 tasks
Contributor
Author
|
@ggerganov , #18641 works on top of this PR, and shows that it works. Please let me know if something else is needed to merge. UPD: PR doesn't affect any existing models or functionality. |
ggerganov
approved these changes
Jan 15, 2026
Member
ggerganov
left a comment
ggerganov
left a comment
There was a problem hiding this comment.
Rebase and lets wait for CI before merge
tdakhran
force-pushed
the
tarek/feat/memory-hybrid-iswa
branch
2 times, most recently
from
b40e4f2 to
99e3092
Compare
tdakhran
force-pushed
the
tarek/feat/memory-hybrid-iswa
branch
from
99e3092 to
d6a45a4
Compare
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Contributor
Author
|
Addressed the feedback, rebased, and tested with #18641. |
CISC
approved these changes
Jan 21, 2026
shaofeiqi
pushed a commit
to qualcomm/llama.cpp
that referenced
this pull request
Feb 6, 2026
* memory : add llama_memory_hybrid_iswa * Update src/llama-memory-hybrid-iswa.cpp Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> --------- Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Upcoming models use hybrid caches with sliding window attention.
Tried modifying the existing
llama_memory_hybridto usellama_kv_cache_iswa, but the change was very intrusive.I don't like the code much, it's essentially a copy-paste done by Claude. But it works! Not very confident with the memory subsystem and open to suggestions and feedback on how to improve it.