]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/commit
overview / pkg / ggml / sources / whisper.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 1550be7)
metal : optimize FA kernels (llama/10171)
authorGeorgi Gerganov <redacted>
Fri, 8 Nov 2024 11:47:22 +0000 (13:47 +0200)
committerGeorgi Gerganov <redacted>
Fri, 15 Nov 2024 13:21:04 +0000 (15:21 +0200)
commitd0b83357895ae4196331e1c4f5513637a0f6dfce
treedc991988a55ffe79c32a8b05027969acd1294c4ftree
parent1550be79f149b4e7780683b585fc28f7bf1d9c6ccommit | diff
metal : optimize FA kernels (llama/10171)

* ggml : add ggml_flash_attn_ext_get_prec

* metal : use F16 precision in FA kernels

ggml-ci

* metal : minor clean-up

* metal : compile-guard bf16 FA kernels

ggml-ci

* build : remove obsolete compile flag [no ci]

* metal : prevent int overflows [no ci]

* cuda : disable BF16 FA

ggml-ci

* metal : fix BF16 requirement for FA kernels

ggml-ci

* make : clean-up [no ci]
ggml/include/ggml.h diff | blob | history
ggml/src/ggml-cuda.cu diff | blob | history
ggml/src/ggml-cuda/fattn.cu diff | blob | history
ggml/src/ggml-metal.m diff | blob | history
ggml/src/ggml-metal.metal diff | blob | history
ggml/src/ggml.c diff | blob | history
Packaging of ggerganov/whisper.cpp