| .. |
|
ggml-cann
|
c8a0090922
cann: support q8_0 for Ascend backend (#8805)
|
1 year ago |
|
ggml-cuda
|
7a11eb3a26
cuda : fix dmmv cols requirement to 2*GGML_CUDA_DMMV_X (#8800)
|
1 year ago |
|
ggml-sycl
|
c887d8b017
[SYCL] Add `TIMESTEP_EMBEDDING` OP (#8707)
|
1 year ago |
|
kompute @ 4565194ed7
|
f3f65429c4
llama : reorganize source code + improve CMake (#8006)
|
1 year ago |
|
kompute-shaders
|
f3f65429c4
llama : reorganize source code + improve CMake (#8006)
|
1 year ago |
|
llamafile
|
6b2a849d1f
ggml : move sgemm sources to llamafile subfolder (#8394)
|
1 year ago |
|
vulkan-shaders
|
4730faca61
chore : Fix vulkan related compiler warnings, add help text, improve CLI options (#8477)
|
1 year ago |
|
CMakeLists.txt
|
6e2b6000e5
cann: update cmake (#8765)
|
1 year ago |
|
ggml-aarch64.c
|
bf5a81df37
ggml : fix build on Windows with Snapdragon X (#8531)
|
1 year ago |
|
ggml-aarch64.h
|
370b1f7e7a
ggml : minor naming changes (#8433)
|
1 year ago |
|
ggml-alloc.c
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
ggml-backend-impl.h
|
f3f65429c4
llama : reorganize source code + improve CMake (#8006)
|
1 year ago |
|
ggml-backend.c
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
ggml-blas.cpp
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
ggml-cann.cpp
|
bfb4c74981
cann: Fix Multi-NPU execution error (#8710)
|
1 year ago |
|
ggml-common.h
|
e54c35e4fb
feat: Support Moore Threads GPU (#8383)
|
1 year ago |
|
ggml-cuda.cu
|
afbb4c1322
ggml-cuda: Adding support for unified memory (#8035)
|
1 year ago |
|
ggml-impl.h
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
ggml-kompute.cpp
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
ggml-metal.m
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
ggml-metal.metal
|
87e397d00b
ggml : fix quant dot product with odd number of blocks (#8549)
|
1 year ago |
|
ggml-quants.c
|
75af08c475
ggml: bugfix: fix the inactive elements is agnostic for risc-v vector (#8748)
|
1 year ago |
|
ggml-quants.h
|
370b1f7e7a
ggml : minor naming changes (#8433)
|
1 year ago |
|
ggml-rpc.cpp
|
f3f65429c4
llama : reorganize source code + improve CMake (#8006)
|
1 year ago |
|
ggml-sycl.cpp
|
c887d8b017
[SYCL] Add `TIMESTEP_EMBEDDING` OP (#8707)
|
1 year ago |
|
ggml-vulkan.cpp
|
203b7f1531
vulkan : initialize vk_buffer_struct members to VK_NULL_HANDLE (ggml/893)
|
1 year ago |
|
ggml.c
|
b7a08fd5e0
Build: Only include execinfo.h on linux systems that support it (#8783)
|
1 year ago |
matteo
afbb4c1322
ggml-cuda: Adding support for unified memory (#8035)