 Piotr Wilkin
|
43eb7a7757
Now that eval's running move delta net stuff back to llama-model, add cbs
|
3 ヶ月 前 |
|
Piotr Wilkin
|
8152df60f3
Getting closer (graph builds for bs=1 but tensor shaping is still wrong for bigger sizes)
|
3 ヶ月 前 |
 Georgi Gerganov
|
0320ac5264
metal : refactor + optimize v2 (#15995)
|
4 ヶ月 前 |
 Aaron Teo
|
186415d595
ggml-cpu: drop support for nnpa intrinsics (#15821)
|
4 ヶ月 前 |
 xctan
|
05c0380f2a
ggml-cpu : optimize RVV kernels (#15720)
|
4 ヶ月 前 |
 Charles Xu
|
4d74393bcc
ggml: update kleidiai to v1.13.0 (#15663)
|
4 ヶ月 前 |
 Johannes Gäßler
|
7a6e91ad26
CUDA: replace GGML_CUDA_F16 with CUDA arch checks (#15433)
|
4 ヶ月 前 |
|
Aaron Teo
|
ff27f80a74
ggml: initial IBM zDNN backend (#14975)
|
5 ヶ月 前 |
 uvos
|
7ad67ba9fe
HIP: add cmake option to enable compiler output of kernel resource usage metrics (#15103)
|
5 ヶ月 前 |
 Christian Kastner
|
41613437ff
cmake: Add GGML_BACKEND_DIR option (#15074)
|
5 ヶ月 前 |
|
uvos
|
b77d11179d
HIP: add GGML_HIP_MMQ_MFMA option to allow disableing the MFMA path. (#14930)
|
5 ヶ月 前 |
|
Aaron Teo
|
c7f3169cd5
ggml-cpu : disable GGML_NNPA by default due to instability (#14880)
|
5 ヶ月 前 |
 R0CKSTAR
|
3f4fc97f1d
musa: upgrade musa sdk to rc4.2.0 (#14498)
|
5 ヶ月 前 |
 Reese Levine
|
21c021745d
ggml: Add initial WebGPU backend (#14521)
|
6 ヶ月 前 |
|
Georgi Gerganov
|
d4cdd9c1c3
ggml : remove kompute backend (#14501)
|
6 ヶ月 前 |
 Daniel Bevenius
|
c46944aa25
ggml : add version function to get lib version (ggml/1286)
|
6 ヶ月 前 |
|
Aaron Teo
|
60ef23d6c1
ggml-cpu: enable IBM NNPA Vector Intrinsics (#14317)
|
6 ヶ月 前 |
|
Daniel Bevenius
|
dd8e59f443
ggml : disable warnings for tests when using MSVC (ggml/1273)
|
7 ヶ月 前 |
|
Daniel Bevenius
|
c2056ed6d4
examples : include examples in msvc disable warn (ggml/1270)
|
7 ヶ月 前 |
 uvos
|
7d6d91babf
HIP: disable rocwmma on gfx12 by default until rocm 7.0 (#14202)
|
7 ヶ月 前 |
 xctan
|
f470bc36be
ggml-cpu : split arch-specific implementations (#13892)
|
7 ヶ月 前 |
 Diego Devesa
|
3a077146a4
llama : allow using mmap without PrefetchVirtualMemory, apply GGML_WIN_VER to llama.cpp sources (#14013)
|
7 ヶ月 前 |
 Jeff Bolz
|
bef8176387
vulkan: use timestamp queries for GGML_VULKAN_PERF (#13817)
|
7 ヶ月 前 |
|
xctan
|
05f6ac6283
ggml : riscv: add xtheadvector support (#13720)
|
7 ヶ月 前 |
 Łukasz Ślusarczyk
|
9c404ed54c
sycl: use oneDNN for matrices multiplication (#12972)
|
8 ヶ月 前 |
|
Daniel Bevenius
|
13b0a04597
whisper: remove MSVC warnings pragmas (whisper/3090)
|
8 ヶ月 前 |
|
Daniel Bevenius
|
99881f77d8
whisper : add check that target name exists (whisper/3103)
|
8 ヶ月 前 |
|
Daniel Bevenius
|
b5769d92b4
ggml : suppress Windows compiler warnings (whisper/3075)
|
8 ヶ月 前 |
|
Diego Devesa
|
1d735c0b4f
ggml : add SSE 4.2 and x64 base variant for CPUs without AVX (#12871)
|
9 ヶ月 前 |
 David Huang
|
84778e9770
CUDA/HIP: Share the same unified memory allocation logic. (#12934)
|
9 ヶ月 前 |
