mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-05-23 13:52:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

Add support for new GGJT v2 quantizers

Browse source 
This change makes quantized models (e.g. q4_0) go 10% faster on Macs
however doesn't offer much improvement for Intel PC hardware.

This change syncs llama.cpp 699b1ad7fe6f7b9e41d3cb41e61a8cc3ea5fc6b5
which recently made a breaking change to nearly all its file formats
without any migration. Since that'll break hundreds upon hundreds of
models on websites like HuggingFace llama.com will support both file
formats because llama.com will never ever break the GGJT file format
This commit is contained in:
Justine Tunney 2023-05-13 08:08:32 -07:00
parent ba49e86e20
commit 5a4cf9560f
No known key found for this signature in database
GPG key ID: BE714B4575D6E328
24 changed files with 4074 additions and 1805 deletions
Download patch file Download diff file Expand all files Collapse all files

2
tool/emacs/cosmo-cpp-constants.el
View file

@ -47,6 +47,8 @@
"__AES__"
"__AVX__"
"__AVX2__"
"__AVX512F__"
"__AVXVNNI__"
"__ABM__"
"__BMI__"
"__BMI2__"