| # CMAKE generated file: DO NOT EDIT! | |
| # Generated by "Unix Makefiles" Generator, CMake Version 3.31 | |
| # compile C with /usr/bin/cc | |
| # compile CUDA with /usr/local/cuda/bin/nvcc | |
| # compile CXX with /usr/bin/c++ | |
| C_DEFINES = -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MIN_BATCH_OFFLOAD=32 -DGGML_CUDA_MMV_Y=1 -DGGML_CUDA_PEER_MAX_BATCH_SIZE=128 -DGGML_CUDA_USE_GRAPHS -DGGML_IQK_FLASH_ATTENTION -DGGML_SCHED_MAX_COPIES=4 -DGGML_USE_CUDA -DGGML_USE_IQK_MULMAT -DGGML_USE_LLAMAFILE -DGGML_USE_OPENMP -DK_QUANTS_PER_ITERATION=2 -DNDEBUG -D_GNU_SOURCE -D_XOPEN_SOURCE=600 | |
| C_INCLUDES = -I/content/tmp/ggml/src/../include -I/content/tmp/ggml/src/. -isystem /usr/local/cuda/targets/x86_64-linux/include | |
| C_FLAGS = -O3 -DNDEBUG -std=gnu11 -Wshadow -Wstrict-prototypes -Wpointer-arith -Wmissing-prototypes -Werror=implicit-int -Werror=implicit-function-declaration -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wdouble-promotion -march=native -fopenmp | |
| CUDA_DEFINES = -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MIN_BATCH_OFFLOAD=32 -DGGML_CUDA_MMV_Y=1 -DGGML_CUDA_PEER_MAX_BATCH_SIZE=128 -DGGML_CUDA_USE_GRAPHS -DGGML_IQK_FLASH_ATTENTION -DGGML_SCHED_MAX_COPIES=4 -DGGML_USE_CUDA -DGGML_USE_IQK_MULMAT -DGGML_USE_LLAMAFILE -DGGML_USE_OPENMP -DK_QUANTS_PER_ITERATION=2 -DNDEBUG -D_GNU_SOURCE -D_XOPEN_SOURCE=600 | |
| CUDA_INCLUDES = -I/content/tmp/ggml/src/../include -I/content/tmp/ggml/src/. -isystem /usr/local/cuda/targets/x86_64-linux/include | |
| CUDA_FLAGS = -O3 -DNDEBUG -std=c++17 "--generate-code=arch=compute_75,code=[compute_75,sm_75]" -use_fast_math -extended-lambda -Xcompiler "-Wmissing-declarations -Wmissing-noreturn -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-array-bounds -Wno-format-truncation -Wextra-semi -Wno-pedantic -march=native" | |
| CXX_DEFINES = -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MIN_BATCH_OFFLOAD=32 -DGGML_CUDA_MMV_Y=1 -DGGML_CUDA_PEER_MAX_BATCH_SIZE=128 -DGGML_CUDA_USE_GRAPHS -DGGML_IQK_FLASH_ATTENTION -DGGML_SCHED_MAX_COPIES=4 -DGGML_USE_CUDA -DGGML_USE_IQK_MULMAT -DGGML_USE_LLAMAFILE -DGGML_USE_OPENMP -DK_QUANTS_PER_ITERATION=2 -DNDEBUG -D_GNU_SOURCE -D_XOPEN_SOURCE=600 | |
| CXX_INCLUDES = -I/content/tmp/ggml/src/../include -I/content/tmp/ggml/src/. -isystem /usr/local/cuda/targets/x86_64-linux/include | |
| CXX_FLAGS = -O3 -DNDEBUG -std=gnu++17 -Wmissing-declarations -Wmissing-noreturn -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-array-bounds -Wno-format-truncation -Wextra-semi -march=native -fopenmp | |