-
Notifications
You must be signed in to change notification settings - Fork 350
Pull requests: ikawrakow/ik_llama.cpp
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Print details when we end up with mmq_x_best = 0
#1954
opened Jun 11, 2026 by
ikawrakow
Owner
Loading…
delta-net: fix np>1 hybrid recurrent-state corruption (batched multi-seq)
#1933
opened Jun 7, 2026 by
poisonxa16
Loading…
Fix misc. expiring logit/sparam bias bugs
#1914
opened Jun 2, 2026 by
dungquixote42
Contributor
•
Draft
1 of 4 tasks
Fix MTP token-generation performance regression
#1894
opened May 29, 2026 by
sayap
Contributor
Loading…
2 of 4 tasks
server: enable checkpoint reuse for recurrent/hybrid models (qwen3next, Mamba)
#1888
opened May 26, 2026 by
localweights
Loading…
docs: Complete rewrite of build.md – CMake-only, focused on supported backends
#1853
opened May 21, 2026 by
maddes8cht
Loading…
2 of 4 tasks
cuda: add get_rows CUDA kernels for Q4_K, Q5_K, Q6_K
#1830
opened May 18, 2026 by
localweights
Loading…
Slightly expand the usage of VNNI256
#1764
opened May 9, 2026 by
XZiar
Contributor
Loading…
2 of 4 tasks
runtime : add
--run-time-repack auto mode for swap-bound MoE safety
#1738
opened May 4, 2026 by
AndrewMoryakov
Contributor
Loading…
2 of 4 tasks
Change signature of llama_set_draft_input_hidden_state
#1727
opened May 3, 2026 by
ikawrakow
Owner
Loading…
convert_hf_to_gguf: add Qwen3.5 / Qwen3.6 support (+ Qwen3-Next scaffolding, not e2e-verified)
#1654
opened Apr 18, 2026 by
markaalonzo
Contributor
•
Draft
Mamba-2 + Nemotron-H MoE backport (Phase 3.x)
#1593
opened Apr 6, 2026 by
AIdevsmartdata
Loading…
5 tasks
ProTip!
Updated in the last three days: updated:>2026-06-11.