| .. |
|
ggml-blas
|
5931c1f233
ggml : add support for dynamic loading of backends (#10469)
|
1 year ago |
|
ggml-cann
|
ba1cb19cdd
llama : add Qwen2VL support + multimodal RoPE (#10361)
|
1 year ago |
|
ggml-cpu
|
68ff663a04
repo : update links to new url (#11886)
|
11 months ago |
|
ggml-cuda
|
73e2ed3ce3
CUDA: use async data loading for FlashAttention (#11894)
|
11 months ago |
|
ggml-hip
|
3ec9fd4b77
HIP: force max threads per block to be 1024 (#11621)
|
11 months ago |
|
ggml-kompute
|
ba1cb19cdd
llama : add Qwen2VL support + multimodal RoPE (#10361)
|
1 year ago |
|
ggml-metal
|
c2ea16f260
metal : fix the crash caused by the lack of residency set support on Intel Macs. (#11904)
|
11 months ago |
|
ggml-musa
|
864a0b67a6
CUDA: use mma PTX instructions for FlashAttention (#11583)
|
11 months ago |
|
ggml-opencl
|
300907b211
opencl: Fix rope and softmax (#11833)
|
11 months ago |
|
ggml-rpc
|
1d20e53c40
rpc: fix known RCE in rpc-server (ggml/1103)
|
11 months ago |
|
ggml-sycl
|
ec3bc8270b
SYCL: remove XMX info from print devices (#11712)
|
11 months ago |
|
ggml-vulkan
|
2eea03d86a
vulkan: implement several ops relevant for ggml_opt (#11769)
|
11 months ago |
|
CMakeLists.txt
|
aa6fb13213
`ci`: use sccache on windows instead of ccache (#11545)
|
11 months ago |
|
ggml-alloc.c
|
1b598b3058
vulkan: use smaller combined allocations to avoid fragmentation (#11551)
|
11 months ago |
|
ggml-backend-impl.h
|
667d72846c
rpc : early register backend devices (#11262)
|
1 year ago |
|
ggml-backend-reg.cpp
|
c792dcf488
ggml : allow loading backend with env variable (ggml/1059)
|
1 year ago |
|
ggml-backend.cpp
|
017cc5f446
ggml-backend : only offload from host buffers (fix) (#11124)
|
1 year ago |
|
ggml-common.h
|
b9ab0a4d0b
CUDA: use arch list for compatibility check (#11775)
|
11 months ago |
|
ggml-impl.h
|
53ff6b9b9f
GGUF: C++ refactor, backend support, misc fixes (#11030)
|
1 year ago |
|
ggml-opt.cpp
|
02e4eaf22f
ggml-opt: fix data corruption (ggml/1022)
|
1 year ago |
|
ggml-quants.c
|
19d8762ab6
ggml : refactor online repacking (#10446)
|
1 year ago |
|
ggml-quants.h
|
ae8de6d50a
ggml : build backends as libraries (#10256)
|
1 year ago |
|
ggml-threading.cpp
|
ae8de6d50a
ggml : build backends as libraries (#10256)
|
1 year ago |
|
ggml-threading.h
|
cb13ef85a4
remove CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS (#10797)
|
1 year ago |
|
ggml.c
|
7b891bdc86
fix: typos in documentation files (#11791)
|
11 months ago |
|
gguf.cpp
|
4dd34ff831
cmake : add sanitizer flags for llama.cpp (#11279)
|
1 year ago |
Johannes Gäßler
73e2ed3ce3
CUDA: use async data loading for FlashAttention (#11894)