Skip to content

Actions: ggml-org/llama.cpp

Actions

CI

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build.yml
11,981 workflow run results
11,981 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

metal : initial Metal4 tensor API support CI #28977: Pull request #16634 synchronize by ggerganov
58m 54s gg/metal-mul-mm-rework
gg/metal-mul-mm-rework
58m 54s
metal : initial Metal4 tensor API support CI #28976: Pull request #16634 synchronize by ggerganov
52s gg/metal-mul-mm-rework
gg/metal-mul-mm-rework
52s
metal : initial Metal4 tensor API support CI #28975: Pull request #16634 synchronize by ggerganov
8m 53s gg/metal-mul-mm-rework
gg/metal-mul-mm-rework
8m 53s
CUDA: fuse gate + up for mmvq, and mmvf CI #28974: Pull request #16630 synchronize by am17an
1h 1m 24s am17an:cuda_fuse_gate
am17an:cuda_fuse_gate
1h 1m 24s
HIP: fix GPU_TARGETS (#16642) CI #28973: Commit ee09828 pushed by JohannesGaessler
1h 52m 28s master
master
1h 52m 28s
CUDA: topk-moe: add optional parameter for gpt-oss CI #28972: Pull request #16649 opened by am17an
22m 27s am17an:cuda_topk_moe_gpt_oss
am17an:cuda_topk_moe_gpt_oss
22m 27s
model: add Granite Hybrid models CI #28971: Pull request #16635 synchronize by giuseppe
1h 7m 19s giuseppe:model-report-granite
giuseppe:model-report-granite
1h 7m 19s
vulkan: Implement topk_moe fused shader, ported from CUDA (#16641) CI #28970: Commit e56abd2 pushed by 0cc4m
1h 9m 5s master
master
1h 9m 5s
CUDA: use registers instead of smem in topk-moe (#16647) CI #28969: Commit 38355c6 pushed by JohannesGaessler
1h 11m 41s master
master
1h 11m 41s
CUDA: use registers instead of smem in topk-moe CI #28968: Pull request #16647 opened by am17an
11m 30s am17an:cuda_topk_optimize
am17an:cuda_topk_optimize
11m 30s
CUDA: fuse gate + up for mmvq, and mmvf CI #28967: Pull request #16630 synchronize by am17an
51m 5s am17an:cuda_fuse_gate
am17an:cuda_fuse_gate
51m 5s
vulkan: Optimize SSM_SCAN CI #28966: Pull request #16645 opened by jeffbolznv
2h 0m 19s jeffbolznv:ssm_scan_opt
jeffbolznv:ssm_scan_opt
2h 0m 19s
CUDA: fuse gate + up for mmvq, and mmvf CI #28965: Pull request #16630 synchronize by am17an
1h 32m 7s am17an:cuda_fuse_gate
am17an:cuda_fuse_gate
1h 32m 7s
Add experimental ggml-hexagon backend for the Hexagon NPU CI #28964: Pull request #16547 synchronize by max-krasnyansky
1h 13m 2s CodeLinaro:hexagon
CodeLinaro:hexagon
1h 13m 2s
opencl: transposed gemm/gemv moe kernel with mxfp4,f32 (#16602) CI #28963: Commit 8138785 pushed by max-krasnyansky
1h 21m 35s master
master
1h 21m 35s
opencl: transposed gemm/gemv moe kernel with mxfp4,f32 CI #28961: Pull request #16602 synchronize by shawngu-quic
6h 53m 30s CodeLinaro:moe-mxfp4-trans-reorder
CodeLinaro:moe-mxfp4-trans-reorder
6h 53m 30s
CUDA: better error for FA kernel with 0 occupancy CI #28960: Pull request #16643 opened by JohannesGaessler
2h 51m 6s JohannesGaessler:cuda-fa-better-error
JohannesGaessler:cuda-fa-better-error
2h 51m 6s
opencl: transposed gemm/gemv moe kernel with mxfp4,f32 CI #28959: Pull request #16602 synchronize by shawngu-quic
59m 51s CodeLinaro:moe-mxfp4-trans-reorder
CodeLinaro:moe-mxfp4-trans-reorder
59m 51s
HIP: fix GPU_TARGETS CI #28958: Pull request #16642 opened by JohannesGaessler
2h 38m 21s JohannesGaessler:hip-fix-gpu-target
JohannesGaessler:hip-fix-gpu-target
2h 38m 21s
opencl: transposed gemm/gemv moe kernel with mxfp4,f32 CI #28957: Pull request #16602 synchronize by shawngu-quic
34m 32s CodeLinaro:moe-mxfp4-trans-reorder
CodeLinaro:moe-mxfp4-trans-reorder
34m 32s
vulkan: Implement topk_moe fused shader, ported from CUDA CI #28956: Pull request #16641 opened by jeffbolznv
4h 38m 43s jeffbolznv:topk
jeffbolznv:topk
4h 38m 43s
Add experimental ggml-hexagon backend for the Hexagon NPU CI #28953: Pull request #16547 synchronize by max-krasnyansky
3h 43m 32s CodeLinaro:hexagon
CodeLinaro:hexagon
3h 43m 32s
Add experimental ggml-hexagon backend for the Hexagon NPU CI #28951: Pull request #16547 synchronize by max-krasnyansky
50m 55s CodeLinaro:hexagon
CodeLinaro:hexagon
50m 55s
CUDA: fuse gate + up for mmvq, and mmvf CI #28950: Pull request #16630 synchronize by am17an
4h 23m 17s am17an:cuda_fuse_gate
am17an:cuda_fuse_gate
4h 23m 17s
vulkan: Increase BK to 32; use BK/4 for non-CM mul_mm.comp CI #28949: Pull request #16636 opened by SavicStefan
3h 8m 22s SavicStefan:BK/4_and_BK=32
SavicStefan:BK/4_and_BK=32
3h 8m 22s
ProTip! You can narrow down the results and go further in time using created:<2025-10-17 or the other filters available.