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
2,736 workflow run results
2,736 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

mtmd: Add JinaCLIP v2 vision projector + GGUF support for jina-bert-v3 (merged-LoRA or adapter) Build on RISCV Linux Machine by Cloud-V #3156: Pull request #16574 synchronize by pockers21
15m 8s pockers21:feature/jinaclip-v2-projector
pockers21:feature/jinaclip-v2-projector
15m 8s
vulkan: Fix crash when FP16 mul_mat accumulation is not supported Build on RISCV Linux Machine by Cloud-V #3155: Pull request #16796 synchronize by rillomas
14m 32s rillomas:fix-accf16-capability-crash
rillomas:fix-accf16-capability-crash
14m 32s
CUDA: fuse rope + set_rows Build on RISCV Linux Machine by Cloud-V #3154: Pull request #16884 opened by am17an
7m 19s am17an:cuda-add-rope-fusion
am17an:cuda-add-rope-fusion
7m 19s
vulkan: Fix crash when FP16 mul_mat accumulation is not supported Build on RISCV Linux Machine by Cloud-V #3153: Pull request #16796 synchronize by rillomas
6m 43s rillomas:fix-accf16-capability-crash
rillomas:fix-accf16-capability-crash
6m 43s
Disable NUMA-specific chunking for high-core-count HPC systems Build on RISCV Linux Machine by Cloud-V #3152: Pull request #16882 opened by rageshh-fj
25m 14s rageshh-fj:rageshh-fj-llama-ggml-cpu-chunking
rageshh-fj:rageshh-fj-llama-ggml-cpu-chunking
25m 14s
mtmd: Add JinaCLIP v2 vision projector + GGUF support for jina-bert-v3 (merged-LoRA or adapter) Build on RISCV Linux Machine by Cloud-V #3151: Pull request #16574 synchronize by pockers21
3m 3s pockers21:feature/jinaclip-v2-projector
pockers21:feature/jinaclip-v2-projector
3m 3s
vulkan: fuse mul_mat+add and mul_mat_id+add_id Build on RISCV Linux Machine by Cloud-V #3150: Pull request #16868 synchronize by jeffbolznv
8m 32s jeffbolznv:mmv_add_fusion
jeffbolznv:mmv_add_fusion
8m 32s
mtmd: Add JinaCLIP v2 vision projector + GGUF support for jina-bert-v3 (merged-LoRA or adapter) Build on RISCV Linux Machine by Cloud-V #3149: Pull request #16574 synchronize by pockers21
7m 28s pockers21:feature/jinaclip-v2-projector
pockers21:feature/jinaclip-v2-projector
7m 28s
ggml-hexagon: respect input size when getting/setting tensor data Build on RISCV Linux Machine by Cloud-V #3148: Pull request #16836 synchronize by l3utterfly
6m 39s l3utterfly:patch-1
l3utterfly:patch-1
6m 39s
server: add support for local image path loading for server Build on RISCV Linux Machine by Cloud-V #3147: Pull request #16874 synchronize by cchadowitz
Action required Pixel-Forensics-Inc:server-local-image-loading
Pixel-Forensics-Inc:server-local-image-loading
Action required
improve CUDA cpy memory bandwidth when copying transposed tensor Build on RISCV Linux Machine by Cloud-V #3146: Pull request #16841 synchronize by bssrdf
31m 55s bssrdf:cuda-transpose-cpy
bssrdf:cuda-transpose-cpy
31m 55s
Implement SparseK Attention mechanism — new GGML operator with CPU backend (GPU planned next) Build on RISCV Linux Machine by Cloud-V #3145: Pull request #16817 synchronize by GittyBurstein
37m 27s yael-works:feature/sparsek-attn-sycl
yael-works:feature/sparsek-attn-sycl
37m 27s
Implement SparseK Attention mechanism — new GGML operator with CPU backend (GPU planned next) Build on RISCV Linux Machine by Cloud-V #3144: Pull request #16817 synchronize by GittyBurstein
24m 7s yael-works:feature/sparsek-attn-sycl
yael-works:feature/sparsek-attn-sycl
24m 7s
Implement SparseK Attention mechanism — new GGML operator with CPU backend (GPU planned next) Build on RISCV Linux Machine by Cloud-V #3143: Pull request #16817 synchronize by GittyBurstein
17m 34s yael-works:feature/sparsek-attn-sycl
yael-works:feature/sparsek-attn-sycl
17m 34s
improve CUDA cpy memory bandwidth when copying transposed tensor Build on RISCV Linux Machine by Cloud-V #3142: Pull request #16841 synchronize by bssrdf
43m 36s bssrdf:cuda-transpose-cpy
bssrdf:cuda-transpose-cpy
43m 36s
vulkan: fuse mul_mat+add and mul_mat_id+add_id Build on RISCV Linux Machine by Cloud-V #3141: Pull request #16868 synchronize by jeffbolznv
27m 53s jeffbolznv:mmv_add_fusion
jeffbolznv:mmv_add_fusion
27m 53s
Implement SparseK Attention mechanism — new GGML operator with CPU backend (GPU planned next) Build on RISCV Linux Machine by Cloud-V #3140: Pull request #16817 synchronize by GittyBurstein
22m 56s yael-works:feature/sparsek-attn-sycl
yael-works:feature/sparsek-attn-sycl
22m 56s
mtmd: refactor preprocessing + support max/min pixels Build on RISCV Linux Machine by Cloud-V #3139: Pull request #16878 opened by ngxson
7m 26s ngxson:xsn/mtmd-max-min-pixels
ngxson:xsn/mtmd-max-min-pixels
7m 26s
ci : enable free-disk-space on cuda docker build Build on RISCV Linux Machine by Cloud-V #3138: Pull request #16877 opened by CISC
6m 45s cisc/ci-docker-cuda-free-disk-space
cisc/ci-docker-cuda-free-disk-space
6m 45s
rpc: use XXHash64 instead of FNV-1a for hashing tensors Build on RISCV Linux Machine by Cloud-V #3137: Pull request #16753 synchronize by jukofyork
9m 43s jukofyork:rpc-xxhash
jukofyork:rpc-xxhash
9m 43s
rpc: use XXHash64 instead of FNV-1a for hashing tensors Build on RISCV Linux Machine by Cloud-V #3136: Pull request #16753 synchronize by jukofyork
6m 44s jukofyork:rpc-xxhash
jukofyork:rpc-xxhash
6m 44s
Update llama-quant.cpp llama_tensor_get_type with DeepSeek friendly modifications Build on RISCV Linux Machine by Cloud-V #3135: Pull request #12727 synchronize by bartowski1182
7m 56s bartowski1182:quant_types
bartowski1182:quant_types
7m 56s
vulkan: fix shmem overrun in mmq id shader Build on RISCV Linux Machine by Cloud-V #3134: Pull request #16873 synchronize by ggerganov
3m 49s 0cc4m/vulkan-workaround-mmqid-shmem-bug
0cc4m/vulkan-workaround-mmqid-shmem-bug
3m 49s
Model: Minimax M2 Build on RISCV Linux Machine by Cloud-V #3132: Pull request #16831 synchronize by pwilkin
5m 51s pwilkin:minimax-m2
pwilkin:minimax-m2
5m 51s
Model: Minimax M2 Build on RISCV Linux Machine by Cloud-V #3133: Pull request #16831 reopened by pwilkin
6m 58s pwilkin:minimax-m2
pwilkin:minimax-m2
6m 58s