Skip to content

Actions: ngxson/llama.cpp

Actions

Server

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
server.yml
1,415 workflow run results
1,415 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

CUDA: fix overflow in FA, tune performance (#14840) Server #1878: Commit a86f52b pushed by ngxson
15m 1s master
master
15m 1s
CUDA: fix compilation with GGML_CUDA_F16 (#14837) Server #1877: Commit b284197 pushed by ngxson
12m 53s master
master
12m 53s
CUDA: fix quantized KV cache + multiple sequences (#14822) Server #1876: Commit 07a19e2 pushed by ngxson
11m 41s master
master
11m 41s
memory : handle saving/loading null layers in recurrent memory (#14675) Server #1875: Commit 7233358 pushed by ngxson
18m 14s master
master
18m 14s
ggml: fix loongarch quantize_row_q8_1 error (#14827) Server #1874: Commit 6c88b3b pushed by ngxson
12m 46s master
master
12m 46s
CANN: weight format to NZ for Ascend310P3 (#14407) Server #1873: Commit 14c28df pushed by ngxson
11m 47s master
master
11m 47s
CUDA: add fused rms norm (#14800) Server #1872: Commit 8c988fa pushed by ngxson
12m 54s master
master
12m 54s
vulkan: fix rms_norm_mul to handle broadcasting dim0 (#14817) Server #1871: Commit 84712b6 pushed by ngxson
11m 44s master
master
11m 44s
llama : add model type detection for rwkv7 7B&14B (#14816) Server #1870: Commit d4d1522 pushed by ngxson
20m 50s master
master
20m 50s
imatrix: add option to display importance score statistics for a give… Server #1869: Commit d1aa0cc pushed by ngxson
12m 26s master
master
12m 26s
Mtmd: add a way to select device for vision encoder (#14236) Server #1868: Commit c8ade30 pushed by ngxson
13m 50s master
master
13m 50s
cuda : implement bf16 cpy ops and enable bf16 cont (#14763) Server #1867: Commit e28c0b8 pushed by ngxson
11m 56s master
master
11m 56s
opencl: remove unreachable return (#14806) Server #1866: Commit 8e6f8bc pushed by ngxson
13m 18s master
master
13m 18s
server : allow setting --reverse-prompt arg (#14799) Server #1865: Commit adef817 pushed by ngxson
12m 20s master
master
12m 20s
cuda: remove linking to cublasLt (#14790) Server #1864: Commit 48b86c4 pushed by ngxson
18m 54s master
master
18m 54s
opencl: add conv2d kernel (#14403) Server #1863: Commit 6c9ee3b pushed by ngxson
21m 19s master
master
21m 19s
sycl: Fix im2col (#14797) Server #1862: Commit cd465d8 pushed by ngxson
13m 31s master
master
13m 31s
kleidiai: add support for get_rows (#14676) Server #1861: Commit 9220426 pushed by ngxson
19m 52s master
master
19m 52s
vulkan/cuda: Fix im2col when KW!=KH (#14789) Server #1860: Commit c2e058f pushed by ngxson
11m 25s master
master
11m 25s
llama : fix --reverse-prompt crashing issue (#14794) Server #1859: Commit c82d48e pushed by ngxson
13m 4s master
master
13m 4s
server : add parse_special option to /tokenize endpoint (#14783) Server #1858: Commit b4efd77 pushed by ngxson
21m 52s master
master
21m 52s
ggml: adds CONV_2D op and direct GEMM Vulkan implementation (#14316) Server #1857: Commit a979ca2 pushed by ngxson
12m 11s master
master
12m 11s
imatrix : use GGUF to store importance matrices (#9400) Server #1856: Commit 9008328 pushed by ngxson
15m 16s master
master
15m 16s
vulkan: Add logging for bf16 features to ggml_vk_print_gpu_info (#132… Server #1855: Commit d4b91ea pushed by ngxson
23m 7s master
master
23m 7s
Vulkan: Fix fprintf format-security warning (#14770) Server #1854: Commit 83f5872 pushed by ngxson
12m 25s master
master
12m 25s