 Kunshang Ji
|
7f412dab9c
enable CPU HBM (#2603)
|
пре 2 година |
 Cebtenzzre
|
00d62adb79
fix some warnings from gcc and clang-tidy (#3038)
|
пре 2 година |
|
Cebtenzzre
|
9912b9efc8
build : add LLAMA_METAL_NDEBUG flag (#3033)
|
пре 2 година |
 Georgi Gerganov
|
e36ecdccc8
build : on Mac OS enable Metal by default (#2901)
|
пре 2 година |
|
Cebtenzzre
|
ef15649972
build : fix most gcc and clang warnings (#2861)
|
пре 2 година |
|
Cebtenzzre
|
849408957c
tests : add a C compliance test (#2848)
|
пре 2 година |
|
Georgi Gerganov
|
3a007648f2
metal : add option to disable debug logs (close #2764)
|
пре 2 година |
 Henri Vasserman
|
6bbc598a63
ROCm Port (#1087)
|
пре 2 година |
|
Georgi Gerganov
|
6381d4e110
gguf : new file format with flexible meta data (beta) (#2398)
|
пре 2 година |
 Kolen Cheung
|
0919a0f73d
cmake : install ggml-meta.metal if LLAMA_METAL (#2449)
|
пре 2 година |
 Shouzheng Liu
|
bf83bff674
metal : matrix-matrix multiplication kernel (#2615)
|
пре 2 година |
 Johannes Gäßler
|
f64d44a9b9
CUDA: Fixed OpenLLaMA 3b mmq, reduced compile time (#2590)
|
пре 2 година |
|
Johannes Gäßler
|
4f6b60c776
CUDA: Fix models with output size != 32000 (#2480)
|
пре 2 година |
|
Johannes Gäßler
|
b772bba42e
CUDA: fixed cmake F16 option (#2471)
|
пре 2 година |
|
Johannes Gäßler
|
0728c5a8b9
CUDA: mmq CLI option, fixed mmq build issues (#2453)
|
пре 2 година |
 slaren
|
a113689571
ggml : add graph tensor allocator (#2411)
|
пре 2 година |
|
Johannes Gäßler
|
11f3ca06b8
CUDA: Quantized matrix matrix multiplication (#2160)
|
пре 2 година |
|
Cebtenzzre
|
6df1f5940f
make : build with -Wmissing-prototypes (#2394)
|
пре 2 година |
 wzy
|
78a3d13424
flake : remove intel mkl from flake.nix due to missing files (#2277)
|
пре 2 година |
|
wzy
|
45a1b07e9b
flake : update flake.nix (#2270)
|
пре 2 година |
|
wzy
|
b1f4290953
cmake : install targets (#2256)
|
пре 2 година |
 Howard Su
|
4e7464ef88
FP16 is supported in CM=6.0 (#2177)
|
пре 2 година |
 Evan Miller
|
5656d10599
mpi : add support for distributed inference via MPI (#2099)
|
пре 2 година |
 clyang
|
3bbc1a11f0
ggml : fix buidling with Intel MKL but ask for "cblas.h" issue (#2104) (#2115)
|
пре 2 година |
|
Johannes Gäßler
|
924dd22fd3
Quantized dot products for CUDA mul mat vec (#2067)
|
пре 2 година |
 Tobias Lütke
|
7ee76e45af
Simple webchat for server (#1998)
|
пре 2 година |
 Daniel Drake
|
b213227067
cmake : don't force -mcpu=native on aarch64 (#2063)
|
пре 2 година |
 Kawrakow
|
6769e944c7
k-quants : support for super-block size of 64 (#2001)
|
пре 2 година |
|
Johannes Gäßler
|
bbca06e269
cmake: revert CUDA arch default to 52, 61 if f16 (#1959)
|
пре 2 година |
|
Georgi Gerganov
|
23fc5c219a
cmake : fix trailing whitespaces
|
пре 2 година |
