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

ggml : remove K_QUANTS_PER_ITERATION (#8306)

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-07-10 15:23:12 +03:00 • committed by GitHub
parent 91deef4606
commit 117f7adbd9
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
13 changed files with 110 additions and 282 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/CMakeLists.txt
View file

@ -297,7 +297,6 @@ if (GGML_CUDA)
add_compile_definitions(GGML_CUDA_DMMV_X=${GGML_CUDA_DMMV_X})
add_compile_definitions(GGML_CUDA_MMV_Y=${GGML_CUDA_MMV_Y})
add_compile_definitions(K_QUANTS_PER_ITERATION=${GGML_CUDA_KQUANTS_ITER})
add_compile_definitions(GGML_CUDA_PEER_MAX_BATCH_SIZE=${GGML_CUDA_PEER_MAX_BATCH_SIZE})
if (GGML_CUDA_USE_GRAPHS)
@ -426,7 +425,6 @@ if (GGML_HIPBLAS)
add_compile_definitions(GGML_USE_HIPBLAS)
add_compile_definitions(GGML_CUDA_DMMV_X=${GGML_CUDA_DMMV_X})
add_compile_definitions(GGML_CUDA_MMV_Y=${GGML_CUDA_MMV_Y})
add_compile_definitions(K_QUANTS_PER_ITERATION=${GGML_CUDA_KQUANTS_ITER})
if (GGML_HIP_UMA)
add_compile_definitions(GGML_HIP_UMA)