Skip to content

Actions: ngxson/llama.cpp

Actions

CI (AMD)

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build-amd.yml
99 workflow runs
99 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

vulkan: fix debug build (add_rms_len/data not found) (#16624) CI (AMD) #99: Commit b194915 pushed by ngxson
Queued master
master
Queued
metal : add CONV_TRANSPOSE_2D (#16542) CI (AMD) #98: Commit 9ad4f19 pushed by ngxson
Queued master
master
Queued
grammar : use int64_t to avoid int overflows in int schema to grammar… CI (AMD) #97: Commit 79967ec pushed by ngxson
Queued master
master
Queued
SYCL SET operator optimized for F32 tensors (#16350) CI (AMD) #96: Commit ceff6bb pushed by ngxson
Queued master
master
Queued
mtmd : support home-cooked Mistral Small Omni (#14928) CI (AMD) #95: Commit 1bb4f43 pushed by ngxson
Queued master
master
Queued
sycl : add ARANGE operator (#16362) CI (AMD) #94: Commit b22572e pushed by ngxson
Queued master
master
Queued
CANN: format code using .clang-format (#15863) CI (AMD) #93: Commit 7a50cf3 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
common : Update the docs on -t --threads (#16236) CI (AMD) #92: Commit 6f5d924 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
SYCL: Add GGML_OP_MEAN operator support (#16009) CI (AMD) #91: Commit ee50ee1 pushed by ngxson
1d 0h 0m 1s master
master
1d 0h 0m 1s
cpu : add FLOOR, CEIL, ROUND and TRUNC unary operators (#16083) CI (AMD) #90: Commit 466c191 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
opencl: add q8_0 mm support (#16469) CI (AMD) #89: Commit 0cb7a06 pushed by ngxson
1d 0h 0m 1s master
master
1d 0h 0m 1s
metal: optimise GGML_OP_SUM (#16559) CI (AMD) #88: Commit f4ce81c pushed by ngxson
1d 0h 0m 7s master
master
1d 0h 0m 7s
server : fix img token logs (#16595) CI (AMD) #87: Commit 17304cb pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
llama-quant: add support for mmproj (#16592) CI (AMD) #86: Commit 3e3cb19 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
CUDA: Changing the CUDA scheduling strategy to spin (#16585) CI (AMD) #85: Commit 5acd455 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
server : fix mtmd checkpoints (#16591) CI (AMD) #84: Commit 554fd57 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
metal : avoid using Metal's gpuAddress property (#16576) CI (AMD) #83: Commit fa882fd pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
vulkan: Add ACC_TYPE_VEC2 implementation (#16203) CI (AMD) #82: Commit ffa0590 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
CUDA + openCL: fix bug in accessing rms_norm->src while doing fusion … CI (AMD) #81: Commit 120bf70 pushed by ngxson
1d 0h 0m 3s master
master
1d 0h 0m 3s
vulkan: Support FA with K/V in F32 (#16543) CI (AMD) #80: Commit 4258e0c pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
vulkan: Improve build time for MSVC (#16545) CI (AMD) #79: Commit 7ea15bb pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
CUDA: enable FA for FP32 KV cache (#16546) CI (AMD) #78: Commit 9c7185d pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
CUDA: use fastdiv + ggml_cuda_mad for mmvf (#16557) CI (AMD) #77: Commit 1ee9d0b pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s
cuda : remove legacy copy-op pointer indirection code (#16485) CI (AMD) #76: Commit 5b6913c pushed by ngxson
1d 0h 0m 1s master
master
1d 0h 0m 1s
server : dynamic token limit for prompt cache (#16560) CI (AMD) #75: Commit bc07349 pushed by ngxson
1d 0h 0m 2s master
master
1d 0h 0m 2s