Skip to content

Actions: ngxson/llama.cpp

Actions

CI

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build.yml
1,687 workflow run results
1,687 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

imatrix : use GGUF by default (#14842) CI #1859: Commit d31192b pushed by ngxson
1h 37m 58s master
master
1h 37m 58s
imatrix : fix 3d activation handling for hybrid and recurrent models … CI #1858: Commit 0a2f549 pushed by ngxson
1h 35m 0s master
master
1h 35m 0s
memory : handle kv_unified for hybrid models (#15050) CI #1857: Commit 11a3811 pushed by ngxson
1h 41m 20s master
master
1h 41m 20s
vocab : JetBrains Mellum pre-tokenizer (#15045) CI #1856: Commit 97366dc pushed by ngxson
1h 4m 21s master
master
1h 4m 21s
vulkan: Use coopmat2 for conv2d (#14982) CI #1855: Commit 6c7a441 pushed by ngxson
58m 8s master
master
58m 8s
opencl: fix adreno compiler detection logic (#15029) CI #1854: Commit 5c0eb5e pushed by ngxson
1h 7m 16s master
master
1h 7m 16s
CUDA: use mma FA kernel for gqa > 4 on RTX 4000 (#15035) CI #1853: Commit 03d4698 pushed by ngxson
1h 1m 57s master
master
1h 1m 57s
cuda: make im2col a little faster (#15025) CI #1852: Commit 3303c19 pushed by ngxson
1h 15m 44s master
master
1h 15m 44s
llama : enable LLAMA_SET_ROWS=1 by default (#14959) CI #1851: Commit a4569c4 pushed by ngxson
56m 4s master
master
56m 4s
chat : fix multiple tool_calls on hermes-2-pro (#14962) CI #1850: Commit f738989 pushed by ngxson
59m 20s master
master
59m 20s
vulkan: coopmat2 mul_mat optimizations (#14934) CI #1849: Commit 4cb208c pushed by ngxson
1h 2m 35s master
master
1h 2m 35s
vulkan: Support ne[3]>1 in noncontig matrix-vector multiply (#15015) CI #1848: Commit ec0b188 pushed by ngxson
1h 23m 50s master
master
1h 23m 50s
model : support Qwen3-Embedding (#15023) CI #1847: Commit 339bd02 pushed by ngxson
1h 10m 48s master
master
1h 10m 48s
server: enable token array inputs for OAI API (#15001) CI #1846: Commit f906275 pushed by ngxson
1h 11m 40s master
master
1h 11m 40s
vulkan: optimizations for direct convolution (#14933) CI #1845: Commit a9f7541 pushed by ngxson
57m 29s master
master
57m 29s
CUDA: fix MMQ nwarps for AMD with warp_size==32 (#15014) CI #1844: Commit 9c35706 pushed by ngxson
1h 6m 18s master
master
1h 6m 18s
vendor : update vendored copy of google/minja (#15011) CI #1843: Commit c76b420 pushed by ngxson
1h 10m 4s master
master
1h 10m 4s
model : add hunyuan dense (#14878) CI #1842: Commit 0f5ccd6 pushed by ngxson
1h 6m 50s master
master
1h 6m 50s
opencl: add f16 for add, sub, mul, div (#14984) CI #1841: Commit 1c872f7 pushed by ngxson
1h 4m 42s master
master
1h 4m 42s
ggml : Q2k interleaving implementation - x86/x64 SIMD (#14373) CI #1840: Commit baad948 pushed by ngxson
1h 4m 57s master
master
1h 4m 57s
graph : fix equal_seq() check (#14986) CI #1839: Commit ba42794 pushed by ngxson
1h 17m 56s master
master
1h 17m 56s
docker : add cann build pipline (#14591) CI #1838: Commit 2860d47 pushed by ngxson
56m 43s master
master
56m 43s
quantize : skip tensor override when in fallback mode (#14995) CI #1837: Commit daf2dd7 pushed by ngxson
1h 7m 45s master
master
1h 7m 45s
llama : add simple option to enable CPU for MoE weights (--cpu-moe) (… CI #1836: Commit a06ed5f pushed by ngxson
1h 6m 45s master
master
1h 6m 45s
Fix params bug in diffusion example (#14993) CI #1835: Commit 7845240 pushed by ngxson
1h 9m 27s master
master
1h 9m 27s
ProTip! You can narrow down the results and go further in time using created:<2025-07-31 or the other filters available.