Skip to content

Actions: ggml-org/llama.cpp

Actions

CI

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build.yml
13,872 workflow runs
13,872 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Override SSM_A op for Qwen3 Next to reduce splits CI #31324: Pull request #17587 opened by pwilkin
Queued pwilkin:qwen3_next_ssm_a_override
pwilkin:qwen3_next_ssm_a_override
Queued
CUDA: add stream-based concurrency CI #31323: Pull request #16991 synchronize by am17an
Queued am17an:fused-qkv-stream
am17an:fused-qkv-stream
Queued
Improve Qwen3-Next Speed CI #31322: Pull request #17585 opened by lovedheart
Action required lovedheart:lovedheart-qwen3next-fix
lovedheart:lovedheart-qwen3next-fix
Action required
Add support for CUMSUM and TRI for CUDA. CI #31321: Pull request #17584 synchronize by pwilkin
Queued pwilkin:tri_cumsum_cuda
pwilkin:tri_cumsum_cuda
Queued
Add support for CUMSUM and TRI for CUDA. CI #31320: Pull request #17584 synchronize by pwilkin
23m 46s pwilkin:tri_cumsum_cuda
pwilkin:tri_cumsum_cuda
23m 46s
Add support for CUMSUM and TRI for CUDA. CI #31319: Pull request #17584 opened by pwilkin
2m 12s pwilkin:tri_cumsum_cuda
pwilkin:tri_cumsum_cuda
2m 12s
vulkan: improve topk perf for large k, fix overflow in unit tests CI #31318: Pull request #17582 opened by jeffbolznv
Queued jeffbolznv:topk_largek
jeffbolznv:topk_largek
Queued
cmake: fix macOS build with -DGGML_BACKEND_DL=ON CI #31317: Pull request #17581 opened by giladgd
Queued giladgd:fixBackendDlOnMacos
giladgd:fixBackendDlOnMacos
Queued
Add PagedAttention support (experimental, CUDA only) CI #31316: Pull request #17579 synchronize by ericcurtin
Queued ericcurtin:add-pagedattention
ericcurtin:add-pagedattention
Queued
Add safetensors support CI #31315: Pull request #17580 opened by ericcurtin
Queued ericcurtin:support-safetensors
ericcurtin:support-safetensors
Queued
Add PagedAttention support (experimental, CUDA only) CI #31314: Pull request #17579 synchronize by ericcurtin
20m 25s ericcurtin:add-pagedattention
ericcurtin:add-pagedattention
20m 25s
Add PagedAttention support (experimental, CUDA only) CI #31313: Pull request #17579 synchronize by ericcurtin
19m 31s ericcurtin:add-pagedattention
ericcurtin:add-pagedattention
19m 31s
Add PagedAttention support (experimental, CUDA only) CI #31312: Pull request #17579 synchronize by ericcurtin
4m 1s ericcurtin:add-pagedattention
ericcurtin:add-pagedattention
4m 1s
Add PagedAttention support (experimental, CUDA only) CI #31311: Pull request #17579 synchronize by ericcurtin
8m 12s ericcurtin:add-pagedattention
ericcurtin:add-pagedattention
8m 12s
Add PagedAttention support (experimental, CUDA only) CI #31310: Pull request #17579 opened by ericcurtin
10m 5s ericcurtin:add-pagedattention
ericcurtin:add-pagedattention
10m 5s
common : move all common_chat_parse_* to chat-parser.cpp. (#17481) CI #31309: Commit 03914c7 pushed by ngxson
Queued master
master
Queued
server: fix: /metrics endpoint returning JSON-escaped Prometheus form… CI #31308: Commit 3ce7a65 pushed by ngxson
Queued master
master
Queued
model: LFM2-VL fixes CI #31307: Pull request #17577 opened by tdakhran
Queued Liquid4All:tarek/feat/upstream_17290
Liquid4All:tarek/feat/upstream_17290
Queued
CUDA: generalized (mma) FA, add Volta support CI #31306: Pull request #17505 synchronize by JohannesGaessler
Queued JohannesGaessler:cuda-fa-mma-update-5
JohannesGaessler:cuda-fa-mma-update-5
Queued
graph : reuse SSM graphs CI #31305: Pull request #16490 synchronize by ggerganov
Queued gg/graph-mamba-reuse
gg/graph-mamba-reuse
Queued
rpc : fix alloc size logic CI #31304: Pull request #17116 synchronize by ggerganov
Queued gg/rpc-fix-alloc-size
gg/rpc-fix-alloc-size
Queued
ggml : add GGML_SCHED_NO_REALLOC option to disable reallocations in g… CI #31303: Commit e072b20 pushed by ggerganov
Queued master
master
Queued
sampling : add support for backend sampling CI #31302: Pull request #17004 synchronize by ggerganov
Queued danbev:gpu-sampling
danbev:gpu-sampling
Queued
sampling : add support for backend sampling CI #31301: Pull request #17004 synchronize by danbev
6m 16s danbev:gpu-sampling
danbev:gpu-sampling
6m 16s
common : throttle download progress output to reduce IO flush CI #31300: Pull request #17427 synchronize by angt
Queued angt:common-throttle-download-progress-output-to-reduce-io-flush
angt:common-throttle-download-progress-output-to-reduce-io-flush
Queued