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

Update llama.cpp

Browse source
This commit is contained in:
Andrei Betlen 2023-04-25 19:03:41 -04:00
parent 3cab3ef4cb
commit cbd26fdcc1
2 changed files with 2 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
llama_cpp/llama_cpp.py
View file

@ -119,6 +119,7 @@ LLAMA_FTYPE_MOSTLY_Q4_1_SOME_F16 = ctypes.c_int(
) # tok_embeddings.weight and output.weight are F16
LLAMA_FTYPE_MOSTLY_Q4_2 = ctypes.c_int(5) # except 1d tensors
LLAMA_FTYPE_MOSTYL_Q4_3 = ctypes.c_int(6) # except 1d tensors
LLAMA_FTYPE_MOSTYL_Q8_0 = ctypes.c_int(7) # except 1d tensors
# Functions

2
vendor/llama.cpp vendored

@ -1 +1 @@
Subproject commit 54bb60e26858be251a0eb3cb70f80322aff804a0
Subproject commit 4afcc378698e057fcde64e23eb664e5af8dd6956