cturan/llama.cpp

ggml : refactor llamafile_sgemm PPC code (#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 <Shalini.Salomi.Bodapati@ibm.com>

shalinib-ibm 6 months ago

parent

9c9e4fc635

commit

55c509daf5

1 changed files with 172 additions and 691 deletions

Split View Show Diff Stats

				
					File diff suppressed because it is too large
					
						+ 172
						
						- 691
					
ggml/src/ggml-cpu/llamafile/sgemm.cpp

ggml : refactor llamafile_sgemm PPC code (#14673)

File diff suppressed because it is too large + 172 - 691 ggml/src/ggml-cpu/llamafile/sgemm.cpp

Some files were not shown because too many files changed in this diff

File diff suppressed because it is too large
+ 172 - 691
ggml/src/ggml-cpu/llamafile/sgemm.cpp