Skip to content

Actions: ngxson/llama.cpp

Actions

Server

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
server.yml
1,472 workflow run results
1,472 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

imatrix : use GGUF by default (#14842) Server #1956: Commit d31192b pushed by ngxson
29m 6s master
master
29m 6s
imatrix : fix 3d activation handling for hybrid and recurrent models … Server #1955: Commit 0a2f549 pushed by ngxson
26m 46s master
master
26m 46s
memory : handle kv_unified for hybrid models (#15050) Server #1954: Commit 11a3811 pushed by ngxson
26m 7s master
master
26m 7s
vocab : JetBrains Mellum pre-tokenizer (#15045) Server #1953: Commit 97366dc pushed by ngxson
12m 0s master
master
12m 0s
vulkan: Use coopmat2 for conv2d (#14982) Server #1952: Commit 6c7a441 pushed by ngxson
13m 31s master
master
13m 31s
opencl: fix adreno compiler detection logic (#15029) Server #1951: Commit 5c0eb5e pushed by ngxson
12m 15s master
master
12m 15s
CUDA: use mma FA kernel for gqa > 4 on RTX 4000 (#15035) Server #1950: Commit 03d4698 pushed by ngxson
26m 4s master
master
26m 4s
cuda: make im2col a little faster (#15025) Server #1949: Commit 3303c19 pushed by ngxson
24m 39s master
master
24m 39s
llama : enable LLAMA_SET_ROWS=1 by default (#14959) Server #1948: Commit a4569c4 pushed by ngxson
16m 5s master
master
16m 5s
chat : fix multiple tool_calls on hermes-2-pro (#14962) Server #1947: Commit f738989 pushed by ngxson
13m 13s master
master
13m 13s
vulkan: coopmat2 mul_mat optimizations (#14934) Server #1946: Commit 4cb208c pushed by ngxson
15m 26s master
master
15m 26s
vulkan: Support ne[3]>1 in noncontig matrix-vector multiply (#15015) Server #1945: Commit ec0b188 pushed by ngxson
21m 0s master
master
21m 0s
model : support Qwen3-Embedding (#15023) Server #1944: Commit 339bd02 pushed by ngxson
18m 53s master
master
18m 53s
server: enable token array inputs for OAI API (#15001) Server #1943: Commit f906275 pushed by ngxson
14m 4s master
master
14m 4s
vulkan: optimizations for direct convolution (#14933) Server #1942: Commit a9f7541 pushed by ngxson
17m 28s master
master
17m 28s
vendor : update vendored copy of google/minja (#15011) Server #1941: Commit c76b420 pushed by ngxson
12m 27s master
master
12m 27s
model : add hunyuan dense (#14878) Server #1940: Commit 0f5ccd6 pushed by ngxson
12m 42s master
master
12m 42s
opencl: add f16 for add, sub, mul, div (#14984) Server #1939: Commit 1c872f7 pushed by ngxson
11m 46s master
master
11m 46s
ggml : Q2k interleaving implementation - x86/x64 SIMD (#14373) Server #1938: Commit baad948 pushed by ngxson
12m 43s master
master
12m 43s
graph : fix equal_seq() check (#14986) Server #1937: Commit ba42794 pushed by ngxson
12m 15s master
master
12m 15s
docker : add cann build pipline (#14591) Server #1936: Commit 2860d47 pushed by ngxson
12m 39s master
master
12m 39s
quantize : skip tensor override when in fallback mode (#14995) Server #1935: Commit daf2dd7 pushed by ngxson
11m 38s master
master
11m 38s
llama : add simple option to enable CPU for MoE weights (--cpu-moe) (… Server #1934: Commit a06ed5f pushed by ngxson
11m 30s master
master
11m 30s
Fix params bug in diffusion example (#14993) Server #1933: Commit 7845240 pushed by ngxson
12m 3s master
master
12m 3s
llama : allow other bufts when overriding to CPU, add --no-repack opt… Server #1932: Commit d6818d0 pushed by ngxson
11m 45s master
master
11m 45s