Skip to content

Pull requests: ggml-org/llama.cpp

New pull request New
1,408 Open 12,932 Closed
1,408 Open 12,932 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

snapdragon: python SDK setup (Windows) documentation Improvements or additions to documentation
#27903 opened Aug 28, 2026 by kurquhar • Draft
1
Blackwell IQ Failures issue Fixes CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning
#27902 opened Aug 28, 2026 by matteius Loading…
1
vulkan : fix target-env for non-_cm2-named shaders needing vulkan1.3 ggml changes relating to the ggml tensor library for machine learning Vulkan Issues specific to the Vulkan backend
#27900 opened Aug 28, 2026 by skywalk1411 • Draft
2
speculative: fix combined draft-mtp + external draft (-md) init crash server
#27897 opened Aug 28, 2026 by simongonzalezdc • Draft
1
2
cuda: enable GGML_CUDA_FA_ALL_QUANTS by default ggml changes relating to the ggml tensor library for machine learning
#27895 opened Aug 28, 2026 by 47Hunter47 • Draft
2 of 4 tasks
1
1
ggml: avoid KleidiAI init mutex on dispatch (#27078) ggml changes relating to the ggml tensor library for machine learning
#27891 opened Aug 28, 2026 by ac-mmi Loading…
1
1
grammar : expand \d/\w/\s regex class escapes in GBNF conversion testing Everything test related
#27889 opened Aug 28, 2026 by yo3hcv • Draft
1
llama : fix out-of-bounds layer split when all devices report zero free memory
#27887 opened Aug 28, 2026 by 0xm0w • Draft
1
1
ci : check for missing autoreleasepools devops improvements to build systems and github actions
#27884 opened Aug 28, 2026 by nikwen Member • 2/2 Loading…
1
metal : fix more leaks due to missing autoreleasepools Apple Metal https://en.wikipedia.org/wiki/Metal_(API) ggml changes relating to the ggml tensor library for machine learning merge ready A maintainer can use this label to indicate that they consider the changes final and ready to merge.
#27883 opened Aug 28, 2026 by nikwen Member • 1/2 Loading…
8
bench: add --tensor-read-lazy documentation Improvements or additions to documentation examples
#27881 opened Aug 28, 2026 by ngxson Collaborator Loading…
2
Qwen4exp correctness fixes Apple Metal https://en.wikipedia.org/wiki/Metal_(API) ggml changes relating to the ggml tensor library for machine learning model Model specific testing Everything test related
#27879 opened Aug 28, 2026 by tarruda Contributor • Draft
ggml-cuda: fix divergent barrier in f16 flash attention CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning
#27870 opened Aug 28, 2026 by siavashnorouzi Loading…
Add Spark2_5 Model conversion model Model specific testing Everything test related
#27868 opened Aug 28, 2026 by KnightYao • Draft
1
mtmd: reject zero granite4-vision downsample sides mtmd Related to multimodal functionality (video/image/audio)
#27867 opened Aug 28, 2026 by 01xjw • Draft
1
spec : reset ngram-cache state between requests testing Everything test related
#27866 opened Aug 28, 2026 by Beatrice0377 Loading…
1
llama: GPU-resident LRU cache for host-offloaded MoE expert weights ggml changes relating to the ggml tensor library for machine learning
#27861 opened Aug 28, 2026 by csantiago78 • Draft
2
fix assertion DFlash2 with --split-mode tensor on CUDA model Model specific
#27858 opened Aug 28, 2026 by art-den Loading…
1
ggml-backend: check return value of ggml_gallocr_reserve_n in alloc_splits ggml changes relating to the ggml tensor library for machine learning
#27855 opened Aug 28, 2026 by kritikagarg Loading…
1
1
ggml-cpu: tiled mul_mat for k-quants ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#27851 opened Aug 28, 2026 by jbooth Loading…
4
sycl: split long rows in TOP_K instead of one workgroup per row ggml changes relating to the ggml tensor library for machine learning SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language
#27847 opened Aug 28, 2026 by Titaniumtown Contributor Loading…
1
ggml-cuda: hip: add missing AMD GCN MMQ config CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#27841 opened Aug 28, 2026 by thelittlefireman Loading…
@IMbackK
1
clamp video fps to the value of new --video-max-tokens cli flag mtmd Related to multimodal functionality (video/image/audio) server
#27838 opened Aug 28, 2026 by DipCrai Loading…
llama: improve TENSOR_READ_LAZY handling
#27837 opened Aug 27, 2026 by ngxson Collaborator Loading…
2
qwen4exp : add NextN/MTP draft head (--spec-type draft-mtp) for Qwen3.8-Flash-Next conversion model Model specific
#27836 opened Aug 27, 2026 by rmonsurate • Draft
1 task done
7
ProTip! Mix and match filters to narrow down what you’re looking for.