git.djapps.eu Git - pkg/ggml/sources/ggml/commit

]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit

overview / pkg / ggml / sources / ggml / commit

author	Georgi Gerganov <redacted>
	Wed, 15 May 2024 06:38:19 +0000 (09:38 +0300)
committer	Georgi Gerganov <redacted>
	Wed, 15 May 2024 07:37:24 +0000 (10:37 +0300)
commit	da79cd67548704dd5bead501b8ec2b9b265d0bbb
tree	6535663f13525bb75066daf20bb5565968ef8fa5	tree
parent	e87c0557b012350005269c49e1c2b5a8631da59a	commit \| diff

whisper : use flash attention (whisper/2152)

* whisper : use flash attention in the encoder

* whisper : add kv_pad

* whisper : remove extra backend instance (huh?)

* whisper : use FA for cross-attention

* whisper : use FA for self-attention

* whisper : simplify encoder FA

* whisper : add flash_attn runtime parameter

* scripts : add bench log

* scripts : add M1 Pro bench log

examples/whisper/main.cpp		diff \| blob \| history
examples/whisper/whisper.cpp		diff \| blob \| history
examples/whisper/whisper.h		diff \| blob \| history

Packaging of ggml-org/ggml

RSS Atom