Spaces:
Running
Running
Commit History
examples : fix compile warnings [no ci] (#0)
32cfce9
sync : ggml
6ef5667
ggml : sync sycl (skip) (#0)
bf6ccee
scripts : fix sync scripts
e2461ca
ggml : remove unnecessary UNUSED macro call (ggml/880)
ab9a7d0
cmake : add GGML_BUILD and GGML_SHARED macro definitions (llama/8281)
a8f9bda
Enabled more data types for oneMKL gemm_batch (llama/8236)
08501f8
Ouadie EL FAROUKI
commited on
CUDA: MMQ support for iq4_nl, iq4_xs (llama/8278)
8411e3c
CUDA: revert part of the RDNA1 optimizations (llama/8309)
fcd0c52
Daniele
commited on
CUDA: fix MMQ stream-k rounding if ne00 % 128 != 0 (llama/8311)
04d4209
Fix WARP_SIZE=16 bug of Intel GPU (llama/8266)
1ce11e2
rm get_work_group_size() by local cache for performance (llama/8286)
08fd758
Neo Zhang Jianyu
arthw
commited on
Define and optimize RDNA1 (llama/8085)
6aa5a89
Daniele
commited on
fix typo (llama/8267)
0c9c7c8
Judd
Judd
commited on
Removes multiple newlines at the end of files that is breaking the editorconfig step of CI. (llama/8258)
cc49462
cuda : update supports_op for matrix multiplication (llama/8245)
2314334
slaren
commited on
Fix win build conflict of math library (llama/8230)
5a33963
Fix the sub group size of Intel (llama/8106)
2dd429e
CUDA: refactor and optimize IQ MMVQ (llama/8215)
afa1447
Update SYCL-Rope op and Refactor (llama/8157)
06acee2
CUDA: fix MMQ stream-k for --split-mode row (llama/8167)
ef3d018
feat: cuda implementation for `ggml_conv_transpose_1d` (ggml/854)
025493b
John Balis
slaren
commited on
ci : disable java build
b5bb445
unverified
server : add inference path to make OAI API compatible (#2270)
66a3eb1
unverified
sync : ggml + fix sync script
bce6859
unverified
make : disable CUDA graphs
ab5ee59
unverified
ggml : add GGML_CUDA_USE_GRAPHS option, restore GGML_CUDA_FORCE_CUBLAS (cmake) (llama/8140)
e83fdad
unverified
slaren
commited on
make : disable CUDA mel build
7d13d39
unverified
cmake : minor fixes
369b16c
unverified
make : fix missing -O3
9cccc55
unverified
whisper : disable CUDA mel + fix FFMPEG
2831df8
unverified
sync : ggml
cd6e534
unverified
whisper : reorganize source code + improve CMake (#2256)
f75c2e3
unverified
whisper : optimize fft() function (#2242)
cc603fa
unverified
mky_coder
Mike Fan
commited on
talk-llama : sync llama.cpp
e8e18fb
unverified
whisper : use ggml_backend_sched (#2239)
bfa5a95
fix : remove extra files
1b0dec0
scripts : sync ggml-blas
463e11c
build : update make / cmake
0b4241c
sync : ggml
89ada87
move BLAS to a separate backend (cont) (llama/6210)
4b26445
slaren
commited on
