Default Branch

master

4c2f924553 · cmake : update CMakeLists.txt to add correct flags (#26) · Updated 1 year ago

Branches

llama

7f32376b70 · llama : initial working FP16 + 4-bit Q4_0 · Updated 1 year ago

0
38
gq

3adf02e311 · utils : print quantization histograms · Updated 1 year ago

0
36
4bit

baeb88b858 · tests : add 4-bit Clover-based quantization · Updated 1 year ago

5
1
t5

1d38a69d7c · t5 : initial load in ggml · Updated 1 year ago

23
3
experiments/blocking

3afb833f84 · wip : unsuccessful attempts speeding mul_mat using blocking · Updated 2 years ago

38
1