 Georgi Gerganov
|
d4cdd9c1c3
ggml : remove kompute backend (#14501)
|
7 months ago |
 Daniel Bevenius
|
c46944aa25
ggml : add version function to get lib version (ggml/1286)
|
7 months ago |
 Aaron Teo
|
60ef23d6c1
ggml-cpu: enable IBM NNPA Vector Intrinsics (#14317)
|
7 months ago |
|
Daniel Bevenius
|
dd8e59f443
ggml : disable warnings for tests when using MSVC (ggml/1273)
|
7 months ago |
|
Daniel Bevenius
|
c2056ed6d4
examples : include examples in msvc disable warn (ggml/1270)
|
8 months ago |
 uvos
|
7d6d91babf
HIP: disable rocwmma on gfx12 by default until rocm 7.0 (#14202)
|
7 months ago |
 xctan
|
f470bc36be
ggml-cpu : split arch-specific implementations (#13892)
|
8 months ago |
 Diego Devesa
|
3a077146a4
llama : allow using mmap without PrefetchVirtualMemory, apply GGML_WIN_VER to llama.cpp sources (#14013)
|
8 months ago |
 Jeff Bolz
|
bef8176387
vulkan: use timestamp queries for GGML_VULKAN_PERF (#13817)
|
8 months ago |
 xctan
|
05f6ac6283
ggml : riscv: add xtheadvector support (#13720)
|
8 months ago |
 Łukasz Ślusarczyk
|
9c404ed54c
sycl: use oneDNN for matrices multiplication (#12972)
|
8 months ago |
|
Daniel Bevenius
|
13b0a04597
whisper: remove MSVC warnings pragmas (whisper/3090)
|
9 months ago |
|
Daniel Bevenius
|
99881f77d8
whisper : add check that target name exists (whisper/3103)
|
9 months ago |
|
Daniel Bevenius
|
b5769d92b4
ggml : suppress Windows compiler warnings (whisper/3075)
|
9 months ago |
|
Diego Devesa
|
1d735c0b4f
ggml : add SSE 4.2 and x64 base variant for CPUs without AVX (#12871)
|
9 months ago |
 David Huang
|
84778e9770
CUDA/HIP: Share the same unified memory allocation logic. (#12934)
|
9 months ago |
|
Daniel Bevenius
|
e408d4351a
ggml : add logging for native build options/vars (whisper/2935)
|
10 months ago |
|
Georgi Gerganov
|
771d84371c
scripts : update sync + fix cmake merge
|
10 months ago |
|
xctan
|
24feaec057
ggml : riscv: add 128-bit RVV support (#12530)
|
10 months ago |
|
Łukasz Ślusarczyk
|
35cae5ba05
SYCL: using graphs is configurable by environment variable and compile option (#12371)
|
10 months ago |
 Henry Linjamäki
|
8acdacb3ea
opencl: use OpenCL C standard supported by the device (#12221)
|
11 months ago |
 Rémy O
|
07d1572347
ggml-cpu: Faster IQ1 mul_mat_vec on AVX2 using BMI2 instructions (#12154)
|
11 months ago |
|
David Huang
|
becade5de7
HIP: implement FlashAttention via rocWMMA for CDNA and RDNA3+ (#12032)
|
11 months ago |
 petterreinholdtsen
|
4512055792
Told cmake to install ggml-cpp.h as a public header file. (ggml/1126)
|
11 months ago |
 Erik Scholz
|
80c41ddd8f
CUDA: compress mode option and default to size (#12029)
|
11 months ago |
 Vladimir Vuksanovic
|
b95c8af37c
cmake: Fix ggml backend dependencies and installation (#11818)
|
11 months ago |
 Aaron Teo
|
af7747c95a
ggml-cpu: Support s390x SIMD Instruction Set (#12019)
|
11 months ago |
 Johannes Gäßler
|
a28e0d5eb1
CUDA: app option to compile without FlashAttention (#12025)
|
11 months ago |
 Charles Xu
|
c5d91a7400
ggml-cpu: Add CPU backend support for KleidiAI library (#11390)
|
11 months ago |
 Christian Kastner
|
8f8290ada9
cmake: Add ability to pass in GGML_BUILD_NUMBER (ggml/1096)
|
1 year ago |
