Skip to content

Actions: ngxson/llama.cpp

Actions

Server

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
server.yml
1,584 workflow run results
1,584 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

graph : fix graph reuse reset of params (#14760) Server #1852: Commit 9fb1042 pushed by ngxson
11m 36s master
master
11m 36s
parallel : add option for different RNG seeds (#14757) Server #1851: Commit 2adf8d8 pushed by ngxson
12m 44s master
master
12m 44s
cuda : Fix Gemma3n not executed as CUDA_GRAPH on NVGPUs (#14741) Server #1850: Commit 021cc28 pushed by ngxson
14m 39s master
master
14m 39s
graph : avoid huge warm-up graphs for MoE models (#14753) Server #1849: Commit d498af3 pushed by ngxson
11m 59s master
master
11m 59s
model : fix build after merge conflict (#14754) Server #1848: Commit eacdeb5 pushed by ngxson
23m 53s master
master
23m 53s
model : add EXAONE 4.0 support (#14630) Server #1847: Commit e0cb5c5 pushed by ngxson
4m 29s master
master
4m 29s
CUDA: set_rows + cpy.cu refactor (#14712) Server #1846: Commit f9a31ee pushed by ngxson
13m 0s master
master
13m 0s
graph : refactor context to not pass gf explicitly (#14629) Server #1845: Commit 8f974bc pushed by ngxson
11m 41s master
master
11m 41s
graph : Pass the graph placeholder message in debug mode (#14748) Server #1844: Commit 09651d0 pushed by ngxson
11m 57s master
master
11m 57s
use max work group size for device to replace the magic number (#14732) Server #1843: Commit 349ea79 pushed by ngxson
12m 26s master
master
12m 26s
model: add Ernie 4.5 MoE support (#14658) Server #1842: Commit cb887f1 pushed by ngxson
12m 7s master
master
12m 7s
kv-cache : fix k-shift for multiple streams (#14742) Server #1841: Commit d6fb3f6 pushed by ngxson
11m 23s master
master
11m 23s
llama : reuse compute graphs (#14482) Server #1840: Commit 01612b7 pushed by ngxson
11m 58s master
master
11m 58s
llama : fix parallel processing for lfm2 (#14705) Server #1839: Commit 086cf81 pushed by ngxson
13m 11s master
master
13m 11s
kv-cache : opt mask set input (#14600) Server #1838: Commit d9b6910 pushed by ngxson
28m 35s master
master
28m 35s
batch : fix uninitialized has_cpl flag (#14733) Server #1837: Commit ad57d3e pushed by ngxson
11m 24s master
master
11m 24s
llama : fix parameter order for hybrid memory initialization (#14725) Server #1836: Commit 496957e pushed by ngxson
11m 35s master
master
11m 35s
ggml: Add initial WebGPU backend (#14521) Server #1835: Commit 21c0217 pushed by ngxson
11m 21s master
master
11m 21s
model : support output bias for qwen2 (#14711) Server #1834: Commit b0f0ecc pushed by ngxson
11m 41s master
master
11m 41s
llama : add high-throughput mode (#14363) Server #1833: Commit 225e7a1 pushed by ngxson
13m 16s master
master
13m 16s
Support diffusion models: Add Dream 7B (#14644) Server #1832: Commit ab14019 pushed by ngxson
11m 47s master
master
11m 47s
ggml : add asserts (#14720) Server #1831: Commit 6497834 pushed by ngxson
11m 44s master
master
11m 44s
server : pre-calculate EOG logit biases (#14721) Server #1830: Commit 6ffd4e9 pushed by ngxson
11m 7s master
master
11m 7s
llama : fix parallel processing for plamo2 (#14716) Server #1829: Commit e4841d2 pushed by ngxson
13m 3s master
master
13m 3s
server : fix handling of the ignore_eos flag (#14710) Server #1828: Commit 538cc77 pushed by ngxson
11m 14s master
master
11m 14s
ProTip! You can narrow down the results and go further in time using created:<2025-07-16 or the other filters available.