aditya/llama.cpp
1
0
Fork 0
mirror of https://git.adityakumar.xyz/llama.cpp.git synced 2024-11-14 08:59:45 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

FP16 is supported in CM=6.0 (#2177)

Browse source 
* FP16 is supported in CM=6.0

* Building PTX code for both of 60 and 61

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
Howard Su 2023-07-12 20:18:40 +08:00 committed by GitHub
parent 2b5eb72e10
commit 4e7464ef88
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
CMakeLists.txt
View file

@ -272,7 +272,7 @@ if (LLAMA_CUBLAS)
if (NOT DEFINED CMAKE_CUDA_ARCHITECTURES) if (NOT DEFINED CMAKE_CUDA_ARCHITECTURES)
if (LLAMA_CUDA_DMMV_F16) if (LLAMA_CUDA_DMMV_F16)
set(CMAKE_CUDA_ARCHITECTURES "61") # needed for f16 CUDA intrinsics set(CMAKE_CUDA_ARCHITECTURES "60;61") # needed for f16 CUDA intrinsics
else() else()
set(CMAKE_CUDA_ARCHITECTURES "52;61") # lowest CUDA 12 standard + lowest for integer intrinsics set(CMAKE_CUDA_ARCHITECTURES "52;61") # lowest CUDA 12 standard + lowest for integer intrinsics
endif() endif()