]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
overview / pkg / ggml / sources / ggml / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 35c6325)
ggml-cpu : add chunking support to mul_mat_id (llama/11666)
authorDiego Devesa <redacted>
Thu, 13 Feb 2025 00:02:38 +0000 (01:02 +0100)
committerGeorgi Gerganov <redacted>
Tue, 25 Feb 2025 11:33:09 +0000 (13:33 +0200)
commit7e4f40c7e31fa050788672dfb12f6faeeb97d0cc
treeba1ff1614f2598385869459c5a7651e2c9708fb8tree
parent35c6325aa1d55b0b74dbab7c3eb4d5d316c6ed84commit | diff
ggml-cpu : add chunking support to mul_mat_id (llama/11666)

* ggml-cpu : add chunking support to mul_mat_id

* allocate chunk counter in wdata
parallelize src1 quantization by column to allows parallelization even when there is only one row

* disable for arm

* cleanup

* better way to disable for arm

* fix uninitialized counter when using 1 thread only

* revert test-backend-ops changes
src/ggml-cpu/ggml-cpu.c diff | blob | history
Packaging of ggml-org/ggml