llama : refactor `src/llama.cpp` #10902
ggerganov
merged
25 commits into
master
from
gg/llama-refactor-0
llama : scatter llama.cpp into multiple modules (wip)
498b68f9
llama : control-vector -> adapter
844660ba
llama : arch
cf899ea0
llama : mmap
6b24e6eb
ci : remove BUILD_SHARED_LIBS=OFF
e9c9209e
llama : arch (cont)
6c22ce10
llama : chat
a2dc93ed
llama : model
7a3065f3
llama : hparams
a25ff12f
llama : adapter
30e0c889
examples : fix
2ebe8fe6
rebase
2a3aa05c
minor
55791c17
llama : kv cache
8ab668e1
llama : impl
5f794937
llama : batch
add3bfe0
cont
5bf9dc57
llama : context
007064f5
minor
4b39d702
llama : context (cont)
736e6922
llama : model loader
8d117a51
common : update lora
272cd0ea
llama : quant
e06d267a
llama : quant (cont)
69dd1e85
ngxson
approved these changes
on 2025-01-02
minor [no ci]
9d0156bf
Login to write a write a comment.
Login via GitHub