vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

ggml : add TQ1_0 and TQ2_0 ternary quantization types

Browse source
This commit is contained in:
Francis Couture-Harpin 2024-07-30 17:55:54 -04:00
parent 79a278e922
commit 77b8f84ae7
10 changed files with 563 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

6
ggml/include/ggml.h
View file

@ -390,8 +390,10 @@ extern "C" {
GGML_TYPE_Q4_0_4_4 = 31,
GGML_TYPE_Q4_0_4_8 = 32,
GGML_TYPE_Q4_0_8_8 = 33,
GGML_TYPE_Q2_2 = 34,
GGML_TYPE_Q1_3 = 35,
GGML_TYPE_TQ1_0 = 34,
GGML_TYPE_TQ2_0 = 35,
GGML_TYPE_Q2_2 = 36,
GGML_TYPE_Q1_3 = 37,
GGML_TYPE_COUNT,
};