Skip to content

Actions: ngxson/llama.cpp

Actions

Server

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
server.yml
1,075 workflow run results
1,075 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

metal : use FA-vec kernel up to batch size 20 (#13496) Server #1510: Commit f0995d2 pushed by ngxson
11m 49s master
master
11m 49s
mtmd : remove libllava, remove clip-quantize-cli (⚠️ breaking change)… Server #1509: Commit b472634 pushed by ngxson
13m 0s master
master
13m 0s
model : Granite MoE shared (#13269) Server #1508: Commit d590cd4 pushed by ngxson
10m 15s master
master
10m 15s
llama-bench : add defrag-thold, check for invalid ranges (#13487) Server #1507: Commit cf0a43b pushed by ngxson
10m 23s master
master
10m 23s
opencl: remove unnecessary assert for add (#13257) Server #1506: Commit f0d46ef pushed by ngxson
8m 44s master
master
8m 44s
clip : cap max image size 1024 for qwen vl model (#13478) Server #1505: Commit de4c07f pushed by ngxson
17m 21s master
master
17m 21s
llama/ggml: add LLM training support (#10544) Server #1504: Commit 10d2af0 pushed by ngxson
11m 29s master
master
11m 29s
context : fix state io for memory-less contexts (#13470) Server #1503: Commit 064cc59 pushed by ngxson
10m 2s master
master
10m 2s
server : allow content to be null in oaicompat_completion_params_pars… Server #1502: Commit 91159ee pushed by ngxson
10m 4s master
master
10m 4s
llama-bench : accept ranges for integer parameters (#13410) Server #1501: Commit 22cdab3 pushed by ngxson
8m 53s master
master
8m 53s
ggml : add mrope kernel for metal (#13457) Server #1500: Commit df84919 pushed by ngxson
8m 50s master
master
8m 50s
enable dpcpp nightly builds with libraries (#13406) Server #1499: Commit 1449214 pushed by ngxson
10m 9s master
master
10m 9s
mtmd : Use RMS norm for InternVL 3 38B and 78B mmproj (#13459) Server #1498: Commit c104023 pushed by ngxson
8m 50s master
master
8m 50s
tools : fix uninitialized llama_batch in server (#13436) Server #1497: Commit 9a390c4 pushed by ngxson
10m 26s master
master
10m 26s
CUDA: fix crash with partial offloading of MoE (#13439) Server #1496: Commit 7474e00 pushed by ngxson
9m 16s master
master
9m 16s
Add --no-op-offload to improve -ot pp perf in MoE models like lla… Server #1495: Commit 7f323a5 pushed by ngxson
18m 11s master
master
18m 11s
mtmd : support InternVL 3 38B and 78B mmproj (#13443) Server #1494: Commit 3eac209 pushed by ngxson
9m 5s master
master
9m 5s
vocab : add ByteDance-Seed/Seed-Coder (#13423) Server #1493: Commit d2a4ef0 pushed by ngxson
8m 52s master
master
8m 52s
mtmd : add hard limit on image resolution for qwen2vl / qwen2.5vl (#1… Server #1492: Commit 15e6125 pushed by ngxson
9m 57s master
master
9m 57s
server : update docs (#13432) Server #1491: Commit 3b24d26 pushed by ngxson
10m 0s master
master
10m 0s
server : update docs Server #1490: Commit 6449e95 pushed by ngxson
9m 30s master
master
9m 30s
llguidance : set tokenizer slices to default (#13424) Server #1489: Commit 43dfd74 pushed by ngxson
18m 32s master
master
18m 32s
mtmd : support InternVL 2.5 and 3 (#13422) Server #1488: Commit 053367d pushed by ngxson
11m 33s master
master
11m 33s
arg : add env var to control mmproj (#13416) Server #1487: Commit 7fef117 pushed by ngxson
15m 56s master
master
15m 56s
vulkan: scalar flash attention implementation (#13324) Server #1486: Commit dc1d2ad pushed by ngxson
8m 33s master
master
8m 33s