Skip to content

Actions: ggml-org/llama.cpp

Actions

CI

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build.yml
11,629 workflow run results
11,629 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

CUDA: add fp kernel for larger batch size MoE CI #28790: Pull request #16512 synchronize by am17an
13m 50s am17an:cuda_mmf_mmid_opt
am17an:cuda_mmf_mmid_opt
13m 50s
CUDA: use fastdiv + ggml_cuda_mad for mmvf CI #28789: Pull request #16557 synchronize by am17an
35m 39s am17an:cuda_mmvf_fastdiv
am17an:cuda_mmvf_fastdiv
35m 39s
CANN: Format ggml-cann src code using clang-format CI #28788: Pull request #15863 synchronize by noemotiovon
44m 46s noemotiovon:format
noemotiovon:format
44m 46s
Add experimental ggml-hexagon backend for the Hexagon NPU CI #28783: Pull request #16547 synchronize by max-krasnyansky
1h 2m 46s CodeLinaro:hexagon
CodeLinaro:hexagon
1h 2m 46s
graph : reuse SSM graphs CI #28782: Pull request #16490 synchronize by ggerganov
1h 27m 25s gg/graph-mamba-reuse
gg/graph-mamba-reuse
1h 27m 25s
metal : FA support F32 K and V CI #28781: Pull request #16531 synchronize by ggerganov
1h 30m 51s gg/metal-fa-f32-inp
gg/metal-fa-f32-inp
1h 30m 51s
graph : support cacheless embeddings with FA and iSWA (#16528) CI #28780: Commit e38b7c6 pushed by ggerganov
1h 39m 32s master
master
1h 39m 32s
server : dynamic token limit for prompt cache CI #28779: Pull request #16560 synchronize by ggerganov
1h 32m 56s gg/server-dynamic-token-limit-for-prompt-cache
gg/server-dynamic-token-limit-for-prompt-cache
1h 32m 56s
graph : support cacheless embeddings with FA and iSWA CI #28778: Pull request #16528 synchronize by ggerganov
1h 13m 34s gg/cacheless-embd
gg/cacheless-embd
1h 13m 34s
opencl: fix build targeting CL 2 (#16554) CI #28776: Commit 5016b72 pushed by max-krasnyansky
52m 20s master
master
52m 20s
embedding: add raw option for --embd-output-format CI #28775: Pull request #16541 synchronize by SamMalayek
1h 6m 8s SamMalayek:feature/raw-embedding-output
SamMalayek:feature/raw-embedding-output
1h 6m 8s
Leverage the existing GGML_F32_VEC helpers to vectorize ggml_vec_set_f32 for faster fills CI #28772: Pull request #16522 synchronize by sirus20x6
3h 21m 58s sirus20x6:vectorize-vec-set-f32
sirus20x6:vectorize-vec-set-f32
3h 21m 58s
vulkan: Add ACC_TYPE_VEC2 implementation CI #28771: Pull request #16203 synchronize by SavicStefan
5m 9s SavicStefan:acc_type_vec2
SavicStefan:acc_type_vec2
5m 9s
graph : support cacheless embeddings with FA and iSWA CI #28770: Pull request #16528 synchronize by ggerganov
1h 5m 55s gg/cacheless-embd
gg/cacheless-embd
1h 5m 55s
CUDA: fix numerical issues in tile FA kernel (#16540) CI #28769: Commit 7049736 pushed by ggerganov
51m 27s master
master
51m 27s
CUDA: fix numerical issue in tile FA kernel CI #28768: Pull request #16540 synchronize by JohannesGaessler
57m 53s JohannesGaessler:cuda-fa-fix-numerics
JohannesGaessler:cuda-fa-fix-numerics
57m 53s
graph : support cacheless embeddings with FA and iSWA CI #28767: Pull request #16528 synchronize by JohannesGaessler
58m 22s gg/cacheless-embd
gg/cacheless-embd
58m 22s
metal: optimise GGML_OP_SUM CI #28766: Pull request #16559 synchronize by cern1710
52m 18s cern1710:metal-optimize-op-sum
cern1710:metal-optimize-op-sum
52m 18s
ggml : fix build broken with -march=armv9-a on MacOS (#16520) CI #28765: Commit 01d2bdc pushed by ggerganov
51m 7s master
master
51m 7s
ggml : fix build broken with -march=armv9-a on MacOS CI #28764: Pull request #16520 synchronize by slaren
46m 6s DamonFool:sve-broken
DamonFool:sve-broken
46m 6s
server : dynamic token limit for prompt cache CI #28758: Pull request #16560 synchronize by ggerganov
50m 51s gg/server-dynamic-token-limit-for-prompt-cache
gg/server-dynamic-token-limit-for-prompt-cache
50m 51s
server : dynamic token limit for prompt cache CI #28757: Pull request #16560 opened by ggerganov
2m 9s gg/server-dynamic-token-limit-for-prompt-cache
gg/server-dynamic-token-limit-for-prompt-cache
2m 9s
metal: optimise GGML_OP_SUM CI #28756: Pull request #16559 opened by cern1710
47m 4s cern1710:metal-optimize-op-sum
cern1710:metal-optimize-op-sum
47m 4s
CANN: fix CPU memory leak in CANN backend (#16549) CI #28755: Commit 56fc38b pushed by hipudding
1h 12m 5s master
master
1h 12m 5s
metal: add support for opt_step_sgd (#16539) CI #28754: Commit 3f750f8 pushed by ggerganov
1h 28m 47s master
master
1h 28m 47s