Skip to content

Pull requests: ggml-org/llama.cpp

New pull request New
1,362 Open 12,782 Closed
1,362 Open 12,782 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

llama : fix K/V and recurrent state cleanup after failed restores testing Everything test related
#27530 opened Aug 22, 2026 by CHIPMUNK-T0T Contributor Loading…
1
@ggerganov
Thread swizzling in kernel_mul_mm (Metal) for better cache locality. Apple Metal https://en.wikipedia.org/wiki/Metal_(API) ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#27529 opened Aug 22, 2026 by skoulik • Draft
1
vulkan: combine duplicated fastdiv functions, rename the one optimizing small divs ggml changes relating to the ggml tensor library for machine learning Vulkan Issues specific to the Vulkan backend
#27526 opened Aug 22, 2026 by jeffbolznv Contributor Loading…
cuda : fuse RWKV7 recurrent input and state paths CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#27523 opened Aug 21, 2026 by 123123213weqw Contributor • Draft
server: observe resumable stream cancellation during first result wait (#27481) server
#27522 opened Aug 21, 2026 by loulanyue • Draft
1
1
mtmd: use ggml_rope_set_offset mtmd Related to multimodal functionality (video/image/audio)
#27521 opened Aug 21, 2026 by ngxson Collaborator • Draft
2
common: handle empty forced_tokens gracefully in reasoning-budget sampler testing Everything test related
#27514 opened Aug 21, 2026 by arnavahire19 • Draft
1
mamba2 : Flatten in/out projections to dispatch GEMM instead of GEMV model Model specific
#27513 opened Aug 21, 2026 by pskrunner14 Loading…
1
5
Quant: OCP FP8 E4M3 support conversion ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#27512 opened Aug 21, 2026 by ORippler Collaborator • Draft
1
common: add json.h abstraction documentation Improvements or additions to documentation jinja parser Issues related to the jinja parser server testing Everything test related
#27511 opened Aug 21, 2026 by ngxson Collaborator Loading…
5 tasks done
3
Add Q2_K reordered MMVQ and ESIMD kernels ggml changes relating to the ggml tensor library for machine learning SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language
#27509 opened Aug 21, 2026 by malsbat Contributor Loading…
1
hexagon: the ggml_hexagon_supported_mul_mat function has incorrect logic when checking quantization types ggml changes relating to the ggml tensor library for machine learning Hexagon
#27502 opened Aug 21, 2026 by zhouwg-jeffzhou Loading…
1
mtmd: fix hanging with specific video-vision mtmd in Windows documentation Improvements or additions to documentation mtmd Related to multimodal functionality (video/image/audio)
#27500 opened Aug 21, 2026 by craftingmod Loading…
1
fit: also take into account n_streams server
#27496 opened Aug 21, 2026 by ngxson Collaborator Loading…
3
Optimize Krea Vulkan - fusion and misc. changes CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning testing Everything test related Vulkan Issues specific to the Vulkan backend
#27495 opened Aug 21, 2026 by pwilkin Member Loading…
2
Optimize Krea Vulkan - Flash Attention CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning testing Everything test related Vulkan Issues specific to the Vulkan backend
#27494 opened Aug 21, 2026 by pwilkin Member Loading…
1
Optimize Krea Vulkan - MUL_MAT CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning testing Everything test related Vulkan Issues specific to the Vulkan backend
#27493 opened Aug 21, 2026 by pwilkin Member Loading…
8
SVE 128 bit Implementation of gemm_q4_k_8x8_q8_k kernel ggml changes relating to the ggml tensor library for machine learning
#27491 opened Aug 21, 2026 by anubhavfujitsu Loading…
1
ggml : reuse compute buffers for MTP (#27282) ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#27489 opened Aug 21, 2026 by mushang0 Loading…
1
2
llama : return 0 on invalid ON_DEVICE seq restore testing Everything test related
#27487 opened Aug 21, 2026 by xiehuanyi • Draft
1
3
misc : prevent RAM peaking at model loading stage
#27483 opened Aug 21, 2026 by tdakhran Contributor Loading…
server: cancel resumable streams before first token server
#27482 opened Aug 21, 2026 by NitayRabi • Draft
1
1
ggml : speed up batch-1 CPU decode, align large allocations ggml changes relating to the ggml tensor library for machine learning
#27478 opened Aug 21, 2026 by matevz-kovacic Loading…
3
common: set Muse Glimmer thinking tags in common_chat_params
#27475 opened Aug 21, 2026 by paralin Loading…
4
SVE 256 bit implementation of gemm_q6_K_8x8_q8_K kernel ggml changes relating to the ggml tensor library for machine learning
#27472 opened Aug 21, 2026 by anubhavfujitsu Loading…
1
ProTip! Find all pull requests that aren't related to any open issues with -linked:issue.