Skip to content

Actions: ngxson/llama.cpp

Actions

Server

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
server.yml
1,724 workflow run results
1,724 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

CANN: ROPE cache sin/cos repeat (#15501) Server #2088: Commit c247d06 pushed by ngxson
14m 48s master
master
14m 48s
vulkan: apply MUL_MAT_ID subgroup optimization to non-coopmat devices… Server #2087: Commit 043fb27 pushed by ngxson
15m 12s master
master
15m 12s
vulkan: Support FA with any multiple of 8 head sizes (#15537) Server #2086: Commit c9a24fb pushed by ngxson
12m 27s master
master
12m 27s
vulkan: enable Conv2D for Apple after MoltenVK fixed the bug (#15526) Server #2085: Commit a9c6ffc pushed by ngxson
12m 34s master
master
12m 34s
CUDA: fix half2 -> half conversion for HIP (#15529) Server #2084: Commit 710dfc4 pushed by ngxson
12m 37s master
master
12m 37s
vulkan: optimize rms_norm, and allow the work to spread across multip… Server #2083: Commit 611f419 pushed by ngxson
12m 26s master
master
12m 26s
model : add support for Seed-OSS (#15490) Server #2082: Commit b1afcab pushed by ngxson
15m 7s master
master
15m 7s
chat : fix debug build assertion in trim function (#15520) Server #2081: Commit 21dc4dd pushed by ngxson
13m 41s master
master
13m 41s
vulkan: Rewrite synchronization to allow some overlap between nodes (… Server #2080: Commit 289bf41 pushed by ngxson
14m 16s master
master
14m 16s
vulkan : support ggml_mean (#15393) Server #2079: Commit 0a9b43e pushed by ngxson
12m 35s master
master
12m 35s
test-opt: allow slight inprecision (#15503) Server #2078: Commit e92734d pushed by ngxson
13m 29s master
master
13m 29s
ggml WebGPU: add support for quantization types (#15440) Server #2077: Commit 4536363 pushed by ngxson
16m 7s master
master
16m 7s
model : gpt-oss add response_format support (#15494) Server #2076: Commit 32732f2 pushed by ngxson
13m 45s master
master
13m 45s
ggml: add conv3d op (#15182) Server #2075: Commit 92f7f0a pushed by ngxson
13m 18s master
master
13m 18s
cuda : add Pad Reflect 1D support (#14659) Server #2074: Commit b1ab918 pushed by ngxson
12m 19s master
master
12m 19s
llama : remove KV cache defragmentation logic (#15473) Server #2073: Commit 9ebebef pushed by ngxson
13m 20s master
master
13m 20s
ggml-cpu: Support Q5_0 and Q5_1 on s390x (#15486) Server #2072: Commit ad5c975 pushed by ngxson
12m 23s master
master
12m 23s
readme : model : mtdm : lfm2 improvements (#15476) Server #2071: Commit e288693 pushed by ngxson
12m 52s master
master
12m 52s
CANN: Optimize RMS_NORM using cache (#15419) Server #2070: Commit a0f98dd pushed by ngxson
14m 58s master
master
14m 58s
sched : fix possible use of wrong ids tensor when offloading moe prom… Server #2069: Commit 54a241f pushed by ngxson
12m 55s master
master
12m 55s
llama : remove deprecated llama_kv_self API (#15472) Server #2068: Commit cd36b5e pushed by ngxson
25m 41s master
master
25m 41s
graph : remove build_attn_with_sinks overload (#15469) Server #2067: Commit 3f196be pushed by ngxson
16m 24s master
master
16m 24s
vulkan : support conv_2d_dw with f16 weights (#15392) Server #2066: Commit 97ae596 pushed by ngxson
26m 53s master
master
26m 53s
vulkan: Reuse conversion results in prealloc_y (#15410) Server #2065: Commit 96452a3 pushed by ngxson
14m 28s master
master
14m 28s
examples : add model conversion tool/example (#15455) Server #2064: Commit 2758fa1 pushed by ngxson
22m 4s master
master
22m 4s