]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
overview / pkg / ggml / sources / ggml / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 4467f47)
ggml : refactor llamafile_sgemm PPC code (llama/14673)
authorshalinib-ibm <redacted>
Mon, 14 Jul 2025 13:16:42 +0000 (18:46 +0530)
committerGeorgi Gerganov <redacted>
Sat, 19 Jul 2025 14:47:23 +0000 (17:47 +0300)
commit20673b6ef2110a0009dbd28f5df6f65a6ab36c9e
tree68639ed0a3ab081fb80a8c8e7bf07cfc17206954tree
parent4467f4732b4d74fb179fd91dcaa21cd0a17c0158commit | diff
ggml : refactor llamafile_sgemm PPC code (llama/14673)

Remove un-necessary templates from class definition and packing functions
Reduce deeply nested conditionals, if-else switching in mnapck function
Replace repetitive code with inline functions in Packing functions

2 ~ 7% improvement in Q8 Model
15 ~ 50% improvement in Q4 Model

Signed-off-by: Shalini Salomi Bodapati <redacted>
src/ggml-cpu/llamafile/sgemm.cpp diff | blob | history
Packaging of ggml-org/ggml