Skip to content

Pull requests: ggml-org/llama.cpp

New pull request New
1,164 Open 11,827 Closed
1,164 Open 11,827 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

cuda : CUDA GGML_OP_LIGHTNING_INDEXER implementation (generic vector kernel + wmma kernel) CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning
#25545 opened Jul 10, 2026 by fairydreaming Collaborator • Draft
common : add support for multiple end sequences in the reasoning budget sampler server testing Everything test related
#25544 opened Jul 10, 2026 by aldehir Contributor Loading…
ggml-cpu: bound destination write extent in get_rows / get_rows_back / im2col / pool / pad ggml changes relating to the ggml tensor library for machine learning
#25542 opened Jul 10, 2026 by BitF4rmer Loading…
1 task done
2
server: refactor server_stream merge ready A maintainer can use this label to indicate that they consider the changes final and ready to merge. server
#25541 opened Jul 10, 2026 by ngxson Collaborator Loading…
2
ui: add reasoning effort control to mobile add sheet server/ui
#25539 opened Jul 10, 2026 by ServeurpersoCom Contributor Loading…
1
2
opencl: add int8 dp4 dense and MoE prefill optimization for Adreno GPUs ggml changes relating to the ggml tensor library for machine learning OpenCL Issues specific to the OpenCL backend
#25537 opened Jul 10, 2026 by wanghqc Contributor Loading…
TP: fix Phi3, Bert, Plamo2/3, ChatGLM testing Everything test related
#25536 opened Jul 10, 2026 by JohannesGaessler Contributor Loading…
ui: Re-design recommended MCP Servers UX server/ui
#25535 opened Jul 10, 2026 by allozaur Contributor • Draft
2
3
metal : optimize Mamba-2 SSM scan decode Apple Metal https://en.wikipedia.org/wiki/Metal_(API) ggml changes relating to the ggml tensor library for machine learning testing Everything test related
#25533 opened Jul 10, 2026 by TommyTranX Loading…
1
Backend sampling multi output ggml changes relating to the ggml tensor library for machine learning server testing Everything test related
#25532 opened Jul 10, 2026 by gaugarg-nv Contributor Loading…
webui: parse effective-parameter sizes (E2B, E4B) as params server/ui
#25529 opened Jul 10, 2026 by erusev Member Loading…
server : prevent LRU eviction of models still loading server
#25528 opened Jul 10, 2026 by ggerganov Member • Draft
1 task done
2
server: allow stream for exec_shell_command documentation Improvements or additions to documentation server
#25526 opened Jul 10, 2026 by ngxson Collaborator Loading…
sycl: Increase minimum buffer size for USM system allocations ggml changes relating to the ggml tensor library for machine learning SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language
#25525 opened Jul 10, 2026 by ifdu Contributor Loading…
mtmd : read LFM2 tiling params from GGUF metadata conversion mtmd Related to multimodal functionality (video/image/audio)
#25524 opened Jul 10, 2026 by michael-dm Loading…
DeepseekV4: clear cache only for seq rather than full
#25521 opened Jul 10, 2026 by am17an Contributor Loading…
7
tokenize : align usage by using common args examples
#25516 opened Jul 10, 2026 by angt Member Loading…
1
fix: use text_config.eos_token_id for HunyuanOCR v1.5 conversion
#25514 opened Jul 10, 2026 by wendadawen Contributor Loading…
model : reject n_layer > LLAMA_MAX_LAYERS (fix OOB write in set_swa_pattern)
#25513 opened Jul 10, 2026 by Yuva1l • Draft
1
vocab : validate special-token ids against vocab size (fix OOB read in llama_vocab::impl::load)
#25508 opened Jul 10, 2026 by Yuva1l Loading…
1
Placed MMA gateway over WMMA gateway CUDA Related to the CUDA backend ggml changes relating to the ggml tensor library for machine learning
#25507 opened Jul 9, 2026 by adam-kabbara • Draft
2
model: GraniteSWAForCausalLM / GraniteMoeSWAForCausalLM conversion model Model specific
#25505 opened Jul 9, 2026 by gabe-l-hart Collaborator • Draft
3
tools: Flush log before exiting after usage()
#25504 opened Jul 9, 2026 by ckastner Collaborator • Draft
Disable -ffast-math on HIP ggml changes relating to the ggml tensor library for machine learning
#25495 opened Jul 9, 2026 by Beinsezii Contributor Loading…
1 task done
vulkan : dequant q8_0 KV once in coopmat1 ggml changes relating to the ggml tensor library for machine learning Vulkan Issues specific to the Vulkan backend
#25494 opened Jul 9, 2026 by Nathanw1014 Loading…
1
9
ProTip! Updated in the last three days: updated:>2026-07-07.