]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
ggml : parallelize FP32 conversion when using BLAS (llama/5045)
authorReinforce-II <redacted>
Mon, 22 Jan 2024 13:15:08 +0000 (21:15 +0800)
committerGeorgi Gerganov <redacted>
Sat, 27 Jan 2024 15:00:47 +0000 (17:00 +0200)
commit14518adb387d0394c0de5a864918ab14fb544fb4
treed38935fb4127c2353eafe157712b01df15330aa8
parent17a2e7d9cdf0ee9fef0379c63065288d2e02a1fe
ggml : parallelize FP32 conversion when using BLAS (llama/5045)

* make GGML_TASK_INIT phase can be run in multithread

* multithreaded dequantize in mul_mat when using blas library

* minor fixes

* update outdated comment
* fix coding style

* simplify code

Co-authored-by: Georgi Gerganov <redacted>
---------

Co-authored-by: Georgi Gerganov <redacted>
src/ggml.c