Spaces:
Running
Running
Commit History
examples : expose language detection probabilities to server example (#3044)
6b8d348
unverified
whisper : remove empty .gitmodules file [no ci] (#3085)
aa54166
unverified
talk-llama : sync llama.cpp (#3084)
511930c
unverified
ci : disable publishing of java binding [no ci] (#3086)
4b6e041
unverified
build : Add Moore Threads GPU support and update GitHub workflow for MUSA build (#3069)
8ede9a1
unverified
R0CKSTAR
commited on
examples : fix deprecated FFmpeg functions (#3073)
0aa41e8
unverified
ruby : add encoder begin callback related methods (#3076)
855927b
unverified
ci : enable bindings java job (#3070)
469f43c
unverified
ruby : add cmake option (#0)
4a21ad6
cuda : fix unused variable compile warning (#0)
a1f4201
sync : ggml
5222212
opencl : remove obsolete files (skip) (ggml/1200)
adc6542
sync : ggml
cac9245
opencl: split ggml-opencl.cl into multiple files and cleanup (llama/12886)
291a5b7
lhez
Shangqing Gu
commited on
ggml : fix trailing whitespaces (llama/0)
5d27bbf
CUDA: use switch statements in constexpr functions (llama/13095)
f5cd546
metal : fix floating-point range of attention scores in FA kernels (llama/13090)
e093044
vulkan: matmul gcn tuning (llama/13016)
ac537d2
CUDA: noncont MMVQ + batched bs1 MUL_MAT_ID (llama/13014)
285a334
ggml : add SSE 4.2 and x64 base variant for CPUs without AVX (llama/12871)
f8795d3
Diego Devesa
commited on
SYCL: Add non-contiguous support in ROPE (llama/12993)
a29a2c3
Akarshan Biswas
commited on
vulkan: support noncontiguous rms_norm (llama/13031)
e4d1f59
metal: add neg operator (llama/13029)
42283e1
SYCL: Refactor and enable FP16 in binary broadcast OPs (llama/12975)
1377b05
Akarshan Biswas
commited on
rpc : add RPC_CMD_HELLO (llama/12955)
ff22836
graph : make FA compatible with MLA + add initial Metal kernels (llama/12953)
fb0d243
ggml: Re-enable CUDA graphs in presence of CONT and DUP nodes (llama/12970)
3944ae5
Alan Gray
commited on
CANN: Add support for async operator submission (llama/12864)
1b9d0f0
opencl: fix incorrect local_size index in profiling log (llama/12868)
8f5d919
kimminsu
commited on
vulkan: enable coopmat2 FA gqa and split_k optimizations more often (llama/12931)
f844153
CANN: Add 310P operator support check (llama/12962)
14d0d7c
Chenguang Li
commited on
metal : add FA-vec kernels for head size 96 (llama/12952)
f1f88b8
CANN: Add x86 build ci (llama/12950)
f4c9b36
CUDA/HIP: Share the same unified memory allocation logic. (llama/12934)
143cb70
David Huang
commited on
SYCL: Add ROPE vision kernel (llama/12887)
5c44879
Akarshan Biswas
commited on
ggml : Add AVX512 implementation of GEMM - Q4_Kx8 (llama/12829)
2457b99
Srihari-mcw
commited on
CANN: Opt ROPE optimization (llama/12865)
3773a09
Chenguang Li
commited on
CANN: Optimize CANN buffer pool memory management (llama/12875)
66b93b3
SYCL: Fix im2col (llama/12910)
a33d74f
Akarshan Biswas
commited on
rpc : use ggml_context_ptr (llama/12938)
24b9742
ggml : Depthwise 2D convolution (ggml/1152)
0c950d5
ggml: use _mm[512/256]_dpbusd[_avx]_epi32 to directly accumulate into the result register (llama/12773)
acb674d
ggml: disable CUDA graphs for unsupported DUP and CONT node types (llama/12891)
9e42c4d
Alan Gray
commited on
vulkan: use aligned loads for flash attention mask (llama/12853)
825889e
sycl: Support sycl_ext_oneapi_limited_graph (llama/12873)
5db8b21
Ewan Crawford
commited on
SYCL: Add fp16 type support to unary op kernels (llama/12788)
b5c106e
Akarshan Biswas
commited on
ggml: fix compilation error s390x (llama/12848)
2458d68
Aaron Teo
Aleksei Nikiforov
commited on
cpu: fix cpu backend's supports-op for GET_ROWS_BACK. fixes a fatal when running test-backend-ops with only the CPU backend (ggml/1190)
ee7706c
cmdr2
commited on
CANN: Support more ops (llama/12841)
6aecea5
Chenguang Li
commited on