Skip to content

Actions: ggml-org/llama.cpp

Actions

Build on RISCV Linux Machine by Cloud-V

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build-riscv-native.yml
816 workflow run results
816 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

server : bring back timings_per_token Build on RISCV Linux Machine by Cloud-V #870: Pull request #15879 opened by ngxson
16m 21s ngxson:xsn/revert_rm_timings
ngxson:xsn/revert_rm_timings
16m 21s
vulkan: sort graph to allow more parallel execution Build on RISCV Linux Machine by Cloud-V #869: Pull request #15850 synchronize by jeffbolznv
6m 24s jeffbolznv:optimize_graph
jeffbolznv:optimize_graph
6m 24s
ggml-backend : add GGML_BACKEND_DEVICE_TYPE_IGPU device type Build on RISCV Linux Machine by Cloud-V #868: Pull request #15797 synchronize by slaren
46m 27s sl/ggml-backend-dev-ids-ext
sl/ggml-backend-dev-ids-ext
46m 27s
feat: Extra debugging support for model conversion Build on RISCV Linux Machine by Cloud-V #866: Pull request #15877 synchronize by pwilkin
1h 24m 14s pwilkin:extra-debug
pwilkin:extra-debug
1h 24m 14s
feat: Extra debugging support for model conversion Build on RISCV Linux Machine by Cloud-V #865: Pull request #15877 opened by pwilkin
1h 55m 21s pwilkin:extra-debug
pwilkin:extra-debug
1h 55m 21s
Deepseek V3.1 native tool calling support (OpenAI Style) Build on RISCV Linux Machine by Cloud-V #864: Pull request #15533 synchronize by createthis
1h 42m 16s createthis:deepseek_3_1_thinking_mode
createthis:deepseek_3_1_thinking_mode
1h 42m 16s
Deepseek V3.1 native tool calling support (OpenAI Style) Build on RISCV Linux Machine by Cloud-V #863: Pull request #15533 synchronize by createthis
1h 37m 53s createthis:deepseek_3_1_thinking_mode
createthis:deepseek_3_1_thinking_mode
1h 37m 53s
Feat: Apertus model implementation Build on RISCV Linux Machine by Cloud-V #860: Pull request #15852 synchronize by pwilkin
1h 24m 14s pwilkin:apertus-implementation
pwilkin:apertus-implementation
1h 24m 14s
metal : make the backend async Build on RISCV Linux Machine by Cloud-V #859: Pull request #15832 synchronize by ggerganov
59m 0s gg/metal-async
gg/metal-async
59m 0s
Feat: Apertus model implementation Build on RISCV Linux Machine by Cloud-V #858: Pull request #15852 synchronize by pwilkin
29m 11s pwilkin:apertus-implementation
pwilkin:apertus-implementation
29m 11s
Feat: Apertus model implementation Build on RISCV Linux Machine by Cloud-V #857: Pull request #15852 synchronize by pwilkin
27m 46s pwilkin:apertus-implementation
pwilkin:apertus-implementation
27m 46s
webgpu : fix build on emscripten Build on RISCV Linux Machine by Cloud-V #856: Pull request #15826 synchronize by ngxson
4m 26s ngxson:xsn/emscripten_webgpu
ngxson:xsn/emscripten_webgpu
4m 26s
metal : refactor + optimize Build on RISCV Linux Machine by Cloud-V #855: Pull request #15857 synchronize by ggerganov
3m 4s gg/metal-refactor
gg/metal-refactor
3m 4s
metal : refactor + optimize Build on RISCV Linux Machine by Cloud-V #854: Pull request #15857 synchronize by ggerganov
4m 55s gg/metal-refactor
gg/metal-refactor
4m 55s
cuda : fix supports_op condition for get_rows when number of blocks is too large Build on RISCV Linux Machine by Cloud-V #853: Pull request #15868 synchronize by ggerganov
8m 31s gg/cuda-fix-supports-get-rows
gg/cuda-fix-supports-get-rows
8m 31s
CUDA: Add fastdiv to k_bin_bcast*, giving 1-3% E2E performance Build on RISCV Linux Machine by Cloud-V #852: Pull request #15872 opened by ORippler
9m 3s ORippler:osimons/add_fastdiv_to_k_bin_bcast
ORippler:osimons/add_fastdiv_to_k_bin_bcast
9m 3s
cuda : fix supports_op condition for get_rows when number of blocks is too large Build on RISCV Linux Machine by Cloud-V #851: Pull request #15868 synchronize by ggerganov
29m 5s gg/cuda-fix-supports-get-rows
gg/cuda-fix-supports-get-rows
29m 5s
model-conversion : add embedding prompt file support Build on RISCV Linux Machine by Cloud-V #850: Pull request #15871 opened by danbev
4m 56s danbev:model-conversion-prompt-file-support
danbev:model-conversion-prompt-file-support
4m 56s
cuda : non-contiguous src0 not supported for PAD Build on RISCV Linux Machine by Cloud-V #849: Pull request #15869 opened by CISC
5m 16s cisc/cuda-pad-contiguous-src0-check
cisc/cuda-pad-contiguous-src0-check
5m 16s
cuda : fix supports_op condition for get_rows when number of blocks is too large Build on RISCV Linux Machine by Cloud-V #848: Pull request #15868 opened by ggerganov
10m 6s gg/cuda-fix-supports-get-rows
gg/cuda-fix-supports-get-rows
10m 6s
convert : force setting sliding_window from original config Build on RISCV Linux Machine by Cloud-V #847: Pull request #15867 synchronize by danbev
19m 33s danbev:embeddinggemma-sliding-window-conversion
danbev:embeddinggemma-sliding-window-conversion
19m 33s
model : avoid ggml_cont_3d for fused QKV weights Build on RISCV Linux Machine by Cloud-V #846: Pull request #15662 synchronize by ggerganov
22m 54s gg/model-avoid-cont3d
gg/model-avoid-cont3d
22m 54s
convert : force setting sliding_window from original config Build on RISCV Linux Machine by Cloud-V #845: Pull request #15867 opened by danbev
6m 47s danbev:embeddinggemma-sliding-window-conversion
danbev:embeddinggemma-sliding-window-conversion
6m 47s
model : avoid ggml_cont_3d for fused QKV weights Build on RISCV Linux Machine by Cloud-V #844: Pull request #15662 synchronize by ggerganov
17m 48s gg/model-avoid-cont3d
gg/model-avoid-cont3d
17m 48s
model : avoid ggml_cont_3d for fused QKV weights Build on RISCV Linux Machine by Cloud-V #843: Pull request #15662 synchronize by ggerganov
16m 24s gg/model-avoid-cont3d
gg/model-avoid-cont3d
16m 24s