Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CuBLAS error in executión and compilatión with warnings #150

Open
SilvaRaulEnriqueCJM opened this issue Jul 25, 2023 · 7 comments
Open

Comments

@SilvaRaulEnriqueCJM
Copy link

Work fine with OpenBLAS and not acceleration.
But, by separatelly, llama.cpp compile and work fine with CuBLAS.

Executión in Error:

(python_3.8) raul@lesquina:~/Desktop/Proyecto/go-llama.cpp$ CGO_LDFLAGS="-lcublas -lcudart -L/usr/local/cuda/lib64/" LIBRARY_PATH=$PWD C_INCLUDE_PATH=$PWD go run ./examples -m "/home/raul/Desktop/Proyecto/models/llama-2-13b-chat.ggmlv3.q2_K.bin" -t 14

github.com/go-skynet/go-llama.cpp

binding.cpp: In function ‘void llama_binding_free_model(void*)’:
binding.cpp:613:5: warning: possible problem detected in invocation of ‘operator delete’ [-Wdelete-incomplete]
613 | delete ctx->model;
| ^~~~~~~~~~~~~~~~~
binding.cpp:613:17: warning: invalid use of incomplete type ‘struct llama_model’
613 | delete ctx->model;
| ~~~~~^~~~~
In file included from ./llama.cpp/examples/common.h:5,
from binding.cpp:1:
./llama.cpp/llama.h:66:12: note: forward declaration of ‘struct llama_model’
66 | struct llama_model;
| ^~~~~~~~~~~
binding.cpp:613:5: note: neither the destructor nor the class-specific ‘operator delete’ will be called, even if they are declared when the class is defined
613 | delete ctx->model;
| ^~~~~~~~~~~~~~~~~
ggml_init_cublas: found 1 CUDA devices:
Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6
llama.cpp: loading model from /home/raul/Desktop/Proyecto/models/llama-2-13b-chat.ggmlv3.q2_K.bin
llama_model_load_internal: format = ggjt v3 (latest)
llama_model_load_internal: n_vocab = 32000
llama_model_load_internal: n_ctx = 128
llama_model_load_internal: n_embd = 5120
llama_model_load_internal: n_mult = 256
llama_model_load_internal: n_head = 40
llama_model_load_internal: n_head_kv = 40
llama_model_load_internal: n_layer = 40
llama_model_load_internal: n_rot = 128
llama_model_load_internal: n_gqa = 1
llama_model_load_internal: n_ff = 13824
llama_model_load_internal: freq_base = 0.0
llama_model_load_internal: freq_scale = 5.60519e-44
llama_model_load_internal: ftype = 10 (mostly Q2_K)
llama_model_load_internal: model size = 13B
llama_model_load_internal: ggml ctx size = 0.11 MB
llama_model_load_internal: using CUDA for GPU acceleration
llama_model_load_internal: mem required = 5587.01 MB (+ 100.00 MB per state)
llama_model_load_internal: offloading 0 repeating layers to GPU
llama_model_load_internal: offloaded 0/43 layers to GPU
llama_model_load_internal: total VRAM used: 0 MB
llama_new_context_with_model: kv self size = 100.00 MB
fatal error: unexpected signal during runtime execution
[signal SIGSEGV: segmentation violation code=0x1 addr=0x100 pc=0x7f84c7ab39fa]

runtime stack:
runtime.throw({0x56fc86?, 0x0?})
/usr/lib/go-1.20/src/runtime/panic.go:1047 +0x5d fp=0x7fff116a39d8 sp=0x7fff116a39a8 pc=0x4523fd
runtime.sigpanic()
/usr/lib/go-1.20/src/runtime/signal_unix.go:821 +0x3e9 fp=0x7fff116a3a38 sp=0x7fff116a39d8 pc=0x466ac9

goroutine 1 [syscall]:
runtime.cgocall(0x4b5320, 0xc00004ebf0)
/usr/lib/go-1.20/src/runtime/cgocall.go:157 +0x5c fp=0xc00004ebc8 sp=0xc00004eb90 pc=0x4239bc
github.com/go-skynet/go-llama%2ecpp._Cfunc_load_model(0x118dce0, 0x80, 0x0, 0x1, 0x0, 0x1, 0x1, 0x0, 0x0, 0x0, ...)
_cgo_gotypes.go:238 +0x4d fp=0xc00004ebf0 sp=0xc00004ebc8 pc=0x4b20ed
github.com/go-skynet/go-llama%2ecpp.New({0x7fff116c2fed, 0x43}, {0xc00004ee68, 0x4, 0x1?})
/home/raul/Desktop/Proyecto/go-llama.cpp/llama.go:26 +0x257 fp=0xc00004ecf8 sp=0xc00004ebf0 pc=0x4b2637
main.main()
/home/raul/Desktop/Proyecto/go-llama.cpp/examples/main.go:35 +0x38f fp=0xc00004ef80 sp=0xc00004ecf8 pc=0x4b460f
runtime.main()
/usr/lib/go-1.20/src/runtime/proc.go:250 +0x207 fp=0xc00004efe0 sp=0xc00004ef80 pc=0x454ce7
runtime.goexit()
/usr/lib/go-1.20/src/runtime/asm_amd64.s:1598 +0x1 fp=0xc00004efe8 sp=0xc00004efe0 pc=0x47fac1

goroutine 2 [force gc (idle)]:
runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
/usr/lib/go-1.20/src/runtime/proc.go:381 +0xd6 fp=0xc000040fb0 sp=0xc000040f90 pc=0x455116
runtime.goparkunlock(...)
/usr/lib/go-1.20/src/runtime/proc.go:387
runtime.forcegchelper()
/usr/lib/go-1.20/src/runtime/proc.go:305 +0xb0 fp=0xc000040fe0 sp=0xc000040fb0 pc=0x454f50
runtime.goexit()
/usr/lib/go-1.20/src/runtime/asm_amd64.s:1598 +0x1 fp=0xc000040fe8 sp=0xc000040fe0 pc=0x47fac1
created by runtime.init.6
/usr/lib/go-1.20/src/runtime/proc.go:293 +0x25

goroutine 3 [GC sweep wait]:
runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
/usr/lib/go-1.20/src/runtime/proc.go:381 +0xd6 fp=0xc000041780 sp=0xc000041760 pc=0x455116
runtime.goparkunlock(...)
/usr/lib/go-1.20/src/runtime/proc.go:387
runtime.bgsweep(0x0?)
/usr/lib/go-1.20/src/runtime/mgcsweep.go:278 +0x8e fp=0xc0000417c8 sp=0xc000041780 pc=0x441e8e
runtime.gcenable.func1()
/usr/lib/go-1.20/src/runtime/mgc.go:178 +0x26 fp=0xc0000417e0 sp=0xc0000417c8 pc=0x437366
runtime.goexit()
/usr/lib/go-1.20/src/runtime/asm_amd64.s:1598 +0x1 fp=0xc0000417e8 sp=0xc0000417e0 pc=0x47fac1
created by runtime.gcenable
/usr/lib/go-1.20/src/runtime/mgc.go:178 +0x6b

goroutine 4 [GC scavenge wait]:
runtime.gopark(0xc000068000?, 0x5884e0?, 0x1?, 0x0?, 0x0?)
/usr/lib/go-1.20/src/runtime/proc.go:381 +0xd6 fp=0xc000041f70 sp=0xc000041f50 pc=0x455116
runtime.goparkunlock(...)
/usr/lib/go-1.20/src/runtime/proc.go:387
runtime.(*scavengerState).park(0x6a2940)
/usr/lib/go-1.20/src/runtime/mgcscavenge.go:400 +0x53 fp=0xc000041fa0 sp=0xc000041f70 pc=0x43fdd3
runtime.bgscavenge(0x0?)
/usr/lib/go-1.20/src/runtime/mgcscavenge.go:628 +0x45 fp=0xc000041fc8 sp=0xc000041fa0 pc=0x4403a5
runtime.gcenable.func2()
/usr/lib/go-1.20/src/runtime/mgc.go:179 +0x26 fp=0xc000041fe0 sp=0xc000041fc8 pc=0x437306
runtime.goexit()
/usr/lib/go-1.20/src/runtime/asm_amd64.s:1598 +0x1 fp=0xc000041fe8 sp=0xc000041fe0 pc=0x47fac1
created by runtime.gcenable
/usr/lib/go-1.20/src/runtime/mgc.go:179 +0xaa

goroutine 18 [finalizer wait]:
runtime.gopark(0x1a0?, 0x6a2d80?, 0xa0?, 0x61?, 0xc000040770?)
/usr/lib/go-1.20/src/runtime/proc.go:381 +0xd6 fp=0xc000040628 sp=0xc000040608 pc=0x455116
runtime.runfinq()
/usr/lib/go-1.20/src/runtime/mfinal.go:193 +0x107 fp=0xc0000407e0 sp=0xc000040628 pc=0x4363a7
runtime.goexit()
/usr/lib/go-1.20/src/runtime/asm_amd64.s:1598 +0x1 fp=0xc0000407e8 sp=0xc0000407e0 pc=0x47fac1
created by runtime.createfing
/usr/lib/go-1.20/src/runtime/mfinal.go:163 +0x45
exit status 2

@SilvaRaulEnriqueCJM
Copy link
Author

Warnings in compilation:

(python_3.8) raul@lesquina:~/Desktop/Proyecto/go-llama.cpp$ BUILD_TYPE=cublas make libbinding.a
I llama.cpp build info:
I UNAME_S: Linux
I UNAME_P: x86_64
I UNAME_M: x86_64
I CFLAGS: -I./llama.cpp -I. -O3 -DNDEBUG -std=c11 -fPIC -Wall -Wextra -Wpedantic -Wcast-qual -Wdouble-promotion -Wshadow -Wstrict-prototypes -Wpointer-arith -Wno-unused-function -pthread -march=native -mtune=native
I CXXFLAGS: -I./llama.cpp -I. -I./llama.cpp/examples -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -pthread
I CGO_LDFLAGS:
I LDFLAGS:
I BUILD_TYPE: cublas
I CMAKE_ARGS: -DLLAMA_CUBLAS=ON
I EXTRA_TARGETS: llama.cpp/ggml-cuda.o
I CC: cc (Ubuntu 12.2.0-17ubuntu1) 12.2.0
I CXX: g++ (Ubuntu 12.2.0-17ubuntu1) 12.2.0

cd llama.cpp && patch -p1 < ../patches/1902-cuda.patch
patching file examples/common.cpp
patching file examples/common.h
touch prepare
mkdir -p build
cd build && cmake ../llama.cpp -DLLAMA_CUBLAS=ON && VERBOSE=1 cmake --build . --config Release && cp -rf CMakeFiles/ggml.dir/ggml.c.o ../llama.cpp/ggml.o
-- The C compiler identification is GNU 12.2.0
-- The CXX compiler identification is GNU 12.2.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: /usr/bin/git (found version "2.39.2")
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE
-- Found CUDAToolkit: /usr/local/cuda-12.1/include (found version "12.1.105")
-- cuBLAS found
-- The CUDA compiler identification is NVIDIA 12.1.105
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /usr/local/cuda-12.1/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- Using CUDA architectures: 52;61
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- x86 detected
-- Configuring done
-- Generating done
-- Build files have been written to: /home/raul/Desktop/Proyecto/go-llama.cpp/build
gmake[1]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/cmake -S/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp -B/home/raul/Desktop/Proyecto/go-llama.cpp/build --check-build-system CMakeFiles/Makefile.cmake 0
/usr/bin/cmake -E cmake_progress_start /home/raul/Desktop/Proyecto/go-llama.cpp/build/CMakeFiles /home/raul/Desktop/Proyecto/go-llama.cpp/build//CMakeFiles/progress.marks
/usr/bin/gmake -f CMakeFiles/Makefile2 all
gmake[2]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f CMakeFiles/BUILD_INFO.dir/build.make CMakeFiles/BUILD_INFO.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/CMakeFiles/BUILD_INFO.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f CMakeFiles/BUILD_INFO.dir/build.make CMakeFiles/BUILD_INFO.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
gmake[3]: No se hace nada para 'CMakeFiles/BUILD_INFO.dir/build'.
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 2%] Built target BUILD_INFO
/usr/bin/gmake -f CMakeFiles/ggml.dir/build.make CMakeFiles/ggml.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/CMakeFiles/ggml.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f CMakeFiles/ggml.dir/build.make CMakeFiles/ggml.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 4%] Building C object CMakeFiles/ggml.dir/ggml.c.o
/usr/bin/cc -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -isystem /usr/local/cuda-12.1/include -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wdouble-promotion -Wshadow -Wstrict-prototypes -Wpointer-arith -mf16c -mfma -mavx -mavx2 -std=gnu11 -MD -MT CMakeFiles/ggml.dir/ggml.c.o -MF CMakeFiles/ggml.dir/ggml.c.o.d -o CMakeFiles/ggml.dir/ggml.c.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/ggml.c
[ 6%] Building CUDA object CMakeFiles/ggml.dir/ggml-cuda.cu.o
/usr/local/cuda-12.1/bin/nvcc -forward-unknown-to-host-compiler -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 --options-file CMakeFiles/ggml.dir/includes_CUDA.rsp -O3 -DNDEBUG --generate-code=arch=compute_52,code=[compute_52,sm_52] --generate-code=arch=compute_61,code=[compute_61,sm_61] -mf16c -mfma -mavx -mavx2 -std=c++11 -MD -MT CMakeFiles/ggml.dir/ggml-cuda.cu.o -MF CMakeFiles/ggml.dir/ggml-cuda.cu.o.d -x cu -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/ggml-cuda.cu -o CMakeFiles/ggml.dir/ggml-cuda.cu.o
[ 8%] Building C object CMakeFiles/ggml.dir/k_quants.c.o
/usr/bin/cc -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -isystem /usr/local/cuda-12.1/include -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wdouble-promotion -Wshadow -Wstrict-prototypes -Wpointer-arith -mf16c -mfma -mavx -mavx2 -std=gnu11 -MD -MT CMakeFiles/ggml.dir/k_quants.c.o -MF CMakeFiles/ggml.dir/k_quants.c.o.d -o CMakeFiles/ggml.dir/k_quants.c.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/k_quants.c
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 8%] Built target ggml
/usr/bin/gmake -f CMakeFiles/ggml_static.dir/build.make CMakeFiles/ggml_static.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/CMakeFiles/ggml_static.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f CMakeFiles/ggml_static.dir/build.make CMakeFiles/ggml_static.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 10%] Linking CUDA static library libggml_static.a
/usr/bin/cmake -P CMakeFiles/ggml_static.dir/cmake_clean_target.cmake
/usr/bin/cmake -E cmake_link_script CMakeFiles/ggml_static.dir/link.txt --verbose=1
/usr/bin/ar qc libggml_static.a CMakeFiles/ggml.dir/ggml.c.o "CMakeFiles/ggml.dir/ggml-cuda.cu.o" CMakeFiles/ggml.dir/k_quants.c.o
/usr/bin/ranlib libggml_static.a
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 10%] Built target ggml_static
/usr/bin/gmake -f CMakeFiles/llama.dir/build.make CMakeFiles/llama.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/CMakeFiles/llama.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f CMakeFiles/llama.dir/build.make CMakeFiles/llama.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 12%] Building CXX object CMakeFiles/llama.dir/llama.cpp.o
/usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -isystem /usr/local/cuda-12.1/include -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT CMakeFiles/llama.dir/llama.cpp.o -MF CMakeFiles/llama.dir/llama.cpp.o.d -o CMakeFiles/llama.dir/llama.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/llama.cpp
[ 14%] Linking CXX static library libllama.a
/usr/bin/cmake -P CMakeFiles/llama.dir/cmake_clean_target.cmake
/usr/bin/cmake -E cmake_link_script CMakeFiles/llama.dir/link.txt --verbose=1
/usr/bin/ar qc libllama.a CMakeFiles/llama.dir/llama.cpp.o CMakeFiles/ggml.dir/ggml.c.o "CMakeFiles/ggml.dir/ggml-cuda.cu.o" CMakeFiles/ggml.dir/k_quants.c.o
/usr/bin/ranlib libllama.a
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 14%] Built target llama
/usr/bin/gmake -f tests/CMakeFiles/test-quantize-fns.dir/build.make tests/CMakeFiles/test-quantize-fns.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests/CMakeFiles/test-quantize-fns.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f tests/CMakeFiles/test-quantize-fns.dir/build.make tests/CMakeFiles/test-quantize-fns.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 16%] Building CXX object tests/CMakeFiles/test-quantize-fns.dir/test-quantize-fns.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT tests/CMakeFiles/test-quantize-fns.dir/test-quantize-fns.cpp.o -MF CMakeFiles/test-quantize-fns.dir/test-quantize-fns.cpp.o.d -o CMakeFiles/test-quantize-fns.dir/test-quantize-fns.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests/test-quantize-fns.cpp
[ 18%] Linking CXX executable ../bin/test-quantize-fns
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/cmake -E cmake_link_script CMakeFiles/test-quantize-fns.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/test-quantize-fns.dir/test-quantize-fns.cpp.o" -o ../bin/test-quantize-fns -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 18%] Built target test-quantize-fns
/usr/bin/gmake -f tests/CMakeFiles/test-quantize-perf.dir/build.make tests/CMakeFiles/test-quantize-perf.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests/CMakeFiles/test-quantize-perf.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f tests/CMakeFiles/test-quantize-perf.dir/build.make tests/CMakeFiles/test-quantize-perf.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 20%] Building CXX object tests/CMakeFiles/test-quantize-perf.dir/test-quantize-perf.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT tests/CMakeFiles/test-quantize-perf.dir/test-quantize-perf.cpp.o -MF CMakeFiles/test-quantize-perf.dir/test-quantize-perf.cpp.o.d -o CMakeFiles/test-quantize-perf.dir/test-quantize-perf.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests/test-quantize-perf.cpp
[ 22%] Linking CXX executable ../bin/test-quantize-perf
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/cmake -E cmake_link_script CMakeFiles/test-quantize-perf.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/test-quantize-perf.dir/test-quantize-perf.cpp.o" -o ../bin/test-quantize-perf -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 22%] Built target test-quantize-perf
/usr/bin/gmake -f tests/CMakeFiles/test-sampling.dir/build.make tests/CMakeFiles/test-sampling.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests/CMakeFiles/test-sampling.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f tests/CMakeFiles/test-sampling.dir/build.make tests/CMakeFiles/test-sampling.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 24%] Building CXX object tests/CMakeFiles/test-sampling.dir/test-sampling.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT tests/CMakeFiles/test-sampling.dir/test-sampling.cpp.o -MF CMakeFiles/test-sampling.dir/test-sampling.cpp.o.d -o CMakeFiles/test-sampling.dir/test-sampling.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests/test-sampling.cpp
[ 26%] Linking CXX executable ../bin/test-sampling
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/cmake -E cmake_link_script CMakeFiles/test-sampling.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/test-sampling.dir/test-sampling.cpp.o" -o ../bin/test-sampling -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 26%] Built target test-sampling
/usr/bin/gmake -f tests/CMakeFiles/test-tokenizer-0.dir/build.make tests/CMakeFiles/test-tokenizer-0.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests/CMakeFiles/test-tokenizer-0.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f tests/CMakeFiles/test-tokenizer-0.dir/build.make tests/CMakeFiles/test-tokenizer-0.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 28%] Building CXX object tests/CMakeFiles/test-tokenizer-0.dir/test-tokenizer-0.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT tests/CMakeFiles/test-tokenizer-0.dir/test-tokenizer-0.cpp.o -MF CMakeFiles/test-tokenizer-0.dir/test-tokenizer-0.cpp.o.d -o CMakeFiles/test-tokenizer-0.dir/test-tokenizer-0.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests/test-tokenizer-0.cpp
[ 30%] Linking CXX executable ../bin/test-tokenizer-0
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/cmake -E cmake_link_script CMakeFiles/test-tokenizer-0.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/test-tokenizer-0.dir/test-tokenizer-0.cpp.o" -o ../bin/test-tokenizer-0 -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 30%] Built target test-tokenizer-0
/usr/bin/gmake -f tests/CMakeFiles/test-grad0.dir/build.make tests/CMakeFiles/test-grad0.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests/CMakeFiles/test-grad0.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f tests/CMakeFiles/test-grad0.dir/build.make tests/CMakeFiles/test-grad0.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 32%] Building C object tests/CMakeFiles/test-grad0.dir/test-grad0.c.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/cc -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wdouble-promotion -Wshadow -Wstrict-prototypes -Wpointer-arith -mf16c -mfma -mavx -mavx2 -std=gnu11 -MD -MT tests/CMakeFiles/test-grad0.dir/test-grad0.c.o -MF CMakeFiles/test-grad0.dir/test-grad0.c.o.d -o CMakeFiles/test-grad0.dir/test-grad0.c.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/tests/test-grad0.c
[ 34%] Linking CXX executable ../bin/test-grad0
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/tests && /usr/bin/cmake -E cmake_link_script CMakeFiles/test-grad0.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/test-grad0.dir/test-grad0.c.o" -o ../bin/test-grad0 -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 34%] Built target test-grad0
/usr/bin/gmake -f examples/CMakeFiles/common.dir/build.make examples/CMakeFiles/common.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/CMakeFiles/common.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/CMakeFiles/common.dir/build.make examples/CMakeFiles/common.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 36%] Building CXX object examples/CMakeFiles/common.dir/common.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/CMakeFiles/common.dir/common.cpp.o -MF CMakeFiles/common.dir/common.cpp.o.d -o CMakeFiles/common.dir/common.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/common.cpp
[ 38%] Building CXX object examples/CMakeFiles/common.dir/grammar-parser.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/CMakeFiles/common.dir/grammar-parser.cpp.o -MF CMakeFiles/common.dir/grammar-parser.cpp.o.d -o CMakeFiles/common.dir/grammar-parser.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/grammar-parser.cpp
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 38%] Built target common
/usr/bin/gmake -f examples/main/CMakeFiles/main.dir/build.make examples/main/CMakeFiles/main.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/main /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/main /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/main/CMakeFiles/main.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/main/CMakeFiles/main.dir/build.make examples/main/CMakeFiles/main.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 40%] Building CXX object examples/main/CMakeFiles/main.dir/main.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/main && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/main/CMakeFiles/main.dir/main.cpp.o -MF CMakeFiles/main.dir/main.cpp.o.d -o CMakeFiles/main.dir/main.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/main/main.cpp
[ 42%] Linking CXX executable ../../bin/main
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/main && /usr/bin/cmake -E cmake_link_script CMakeFiles/main.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/main.dir/main.cpp.o ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/main -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 42%] Built target main
/usr/bin/gmake -f examples/quantize/CMakeFiles/quantize.dir/build.make examples/quantize/CMakeFiles/quantize.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/quantize /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize/CMakeFiles/quantize.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/quantize/CMakeFiles/quantize.dir/build.make examples/quantize/CMakeFiles/quantize.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 44%] Building CXX object examples/quantize/CMakeFiles/quantize.dir/quantize.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/quantize/CMakeFiles/quantize.dir/quantize.cpp.o -MF CMakeFiles/quantize.dir/quantize.cpp.o.d -o CMakeFiles/quantize.dir/quantize.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/quantize/quantize.cpp
[ 46%] Linking CXX executable ../../bin/quantize
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize && /usr/bin/cmake -E cmake_link_script CMakeFiles/quantize.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/quantize.dir/quantize.cpp.o -o ../../bin/quantize -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 46%] Built target quantize
/usr/bin/gmake -f examples/quantize-stats/CMakeFiles/quantize-stats.dir/build.make examples/quantize-stats/CMakeFiles/quantize-stats.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/quantize-stats /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize-stats /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize-stats/CMakeFiles/quantize-stats.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/quantize-stats/CMakeFiles/quantize-stats.dir/build.make examples/quantize-stats/CMakeFiles/quantize-stats.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 48%] Building CXX object examples/quantize-stats/CMakeFiles/quantize-stats.dir/quantize-stats.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize-stats && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/quantize-stats/CMakeFiles/quantize-stats.dir/quantize-stats.cpp.o -MF CMakeFiles/quantize-stats.dir/quantize-stats.cpp.o.d -o CMakeFiles/quantize-stats.dir/quantize-stats.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/quantize-stats/quantize-stats.cpp
[ 51%] Linking CXX executable ../../bin/quantize-stats
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/quantize-stats && /usr/bin/cmake -E cmake_link_script CMakeFiles/quantize-stats.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/quantize-stats.dir/quantize-stats.cpp.o" -o ../../bin/quantize-stats -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 51%] Built target quantize-stats
/usr/bin/gmake -f examples/perplexity/CMakeFiles/perplexity.dir/build.make examples/perplexity/CMakeFiles/perplexity.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/perplexity /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/perplexity /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/perplexity/CMakeFiles/perplexity.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/perplexity/CMakeFiles/perplexity.dir/build.make examples/perplexity/CMakeFiles/perplexity.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 53%] Building CXX object examples/perplexity/CMakeFiles/perplexity.dir/perplexity.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/perplexity && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/perplexity/CMakeFiles/perplexity.dir/perplexity.cpp.o -MF CMakeFiles/perplexity.dir/perplexity.cpp.o.d -o CMakeFiles/perplexity.dir/perplexity.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/perplexity/perplexity.cpp
[ 55%] Linking CXX executable ../../bin/perplexity
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/perplexity && /usr/bin/cmake -E cmake_link_script CMakeFiles/perplexity.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/perplexity.dir/perplexity.cpp.o ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/perplexity -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 55%] Built target perplexity
/usr/bin/gmake -f examples/embedding/CMakeFiles/embedding.dir/build.make examples/embedding/CMakeFiles/embedding.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/embedding /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embedding /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embedding/CMakeFiles/embedding.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/embedding/CMakeFiles/embedding.dir/build.make examples/embedding/CMakeFiles/embedding.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 57%] Building CXX object examples/embedding/CMakeFiles/embedding.dir/embedding.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embedding && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/embedding/CMakeFiles/embedding.dir/embedding.cpp.o -MF CMakeFiles/embedding.dir/embedding.cpp.o.d -o CMakeFiles/embedding.dir/embedding.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/embedding/embedding.cpp
[ 59%] Linking CXX executable ../../bin/embedding
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embedding && /usr/bin/cmake -E cmake_link_script CMakeFiles/embedding.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/embedding.dir/embedding.cpp.o ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/embedding -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 59%] Built target embedding
/usr/bin/gmake -f examples/save-load-state/CMakeFiles/save-load-state.dir/build.make examples/save-load-state/CMakeFiles/save-load-state.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/save-load-state /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/save-load-state /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/save-load-state/CMakeFiles/save-load-state.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/save-load-state/CMakeFiles/save-load-state.dir/build.make examples/save-load-state/CMakeFiles/save-load-state.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 61%] Building CXX object examples/save-load-state/CMakeFiles/save-load-state.dir/save-load-state.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/save-load-state && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/save-load-state/CMakeFiles/save-load-state.dir/save-load-state.cpp.o -MF CMakeFiles/save-load-state.dir/save-load-state.cpp.o.d -o CMakeFiles/save-load-state.dir/save-load-state.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/save-load-state/save-load-state.cpp
[ 63%] Linking CXX executable ../../bin/save-load-state
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/save-load-state && /usr/bin/cmake -E cmake_link_script CMakeFiles/save-load-state.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/save-load-state.dir/save-load-state.cpp.o" ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/save-load-state -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 63%] Built target save-load-state
/usr/bin/gmake -f examples/benchmark/CMakeFiles/benchmark.dir/build.make examples/benchmark/CMakeFiles/benchmark.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/benchmark /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/benchmark /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/benchmark/CMakeFiles/benchmark.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/benchmark/CMakeFiles/benchmark.dir/build.make examples/benchmark/CMakeFiles/benchmark.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 65%] Building CXX object examples/benchmark/CMakeFiles/benchmark.dir/benchmark-matmult.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/benchmark && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/benchmark/CMakeFiles/benchmark.dir/benchmark-matmult.cpp.o -MF CMakeFiles/benchmark.dir/benchmark-matmult.cpp.o.d -o CMakeFiles/benchmark.dir/benchmark-matmult.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/benchmark/benchmark-matmult.cpp
[ 67%] Linking CXX executable ../../bin/benchmark
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/benchmark && /usr/bin/cmake -E cmake_link_script CMakeFiles/benchmark.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/benchmark.dir/benchmark-matmult.cpp.o" ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/benchmark -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 67%] Built target benchmark
/usr/bin/gmake -f examples/baby-llama/CMakeFiles/baby-llama.dir/build.make examples/baby-llama/CMakeFiles/baby-llama.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/baby-llama /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/baby-llama /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/baby-llama/CMakeFiles/baby-llama.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/baby-llama/CMakeFiles/baby-llama.dir/build.make examples/baby-llama/CMakeFiles/baby-llama.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 69%] Building CXX object examples/baby-llama/CMakeFiles/baby-llama.dir/baby-llama.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/baby-llama && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/baby-llama/CMakeFiles/baby-llama.dir/baby-llama.cpp.o -MF CMakeFiles/baby-llama.dir/baby-llama.cpp.o.d -o CMakeFiles/baby-llama.dir/baby-llama.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/baby-llama/baby-llama.cpp
/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/baby-llama/baby-llama.cpp: In function ‘int main(int, char**)’:
/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/baby-llama/baby-llama.cpp:1614:32: warning: variable ‘opt_params_adam’ set but not used [-Wunused-but-set-variable]
1614 | struct ggml_opt_params opt_params_adam = ggml_opt_default_params(GGML_OPT_ADAM);
| ^~~~~~~~~~~~~~~
[ 71%] Linking CXX executable ../../bin/baby-llama
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/baby-llama && /usr/bin/cmake -E cmake_link_script CMakeFiles/baby-llama.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/baby-llama.dir/baby-llama.cpp.o" ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/baby-llama -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 71%] Built target baby-llama
/usr/bin/gmake -f examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/build.make examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/train-text-from-scratch /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/train-text-from-scratch /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/build.make examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 73%] Building CXX object examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/train-text-from-scratch.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/train-text-from-scratch && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/train-text-from-scratch/CMakeFiles/train-text-from-scratch.dir/train-text-from-scratch.cpp.o -MF CMakeFiles/train-text-from-scratch.dir/train-text-from-scratch.cpp.o.d -o CMakeFiles/train-text-from-scratch.dir/train-text-from-scratch.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/train-text-from-scratch/train-text-from-scratch.cpp
[ 75%] Linking CXX executable ../../bin/train-text-from-scratch
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/train-text-from-scratch && /usr/bin/cmake -E cmake_link_script CMakeFiles/train-text-from-scratch.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/train-text-from-scratch.dir/train-text-from-scratch.cpp.o" ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/train-text-from-scratch -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 75%] Built target train-text-from-scratch
/usr/bin/gmake -f examples/simple/CMakeFiles/simple.dir/build.make examples/simple/CMakeFiles/simple.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/simple /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/simple /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/simple/CMakeFiles/simple.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/simple/CMakeFiles/simple.dir/build.make examples/simple/CMakeFiles/simple.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 77%] Building CXX object examples/simple/CMakeFiles/simple.dir/simple.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/simple && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/simple/CMakeFiles/simple.dir/simple.cpp.o -MF CMakeFiles/simple.dir/simple.cpp.o.d -o CMakeFiles/simple.dir/simple.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/simple/simple.cpp
[ 79%] Linking CXX executable ../../bin/simple
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/simple && /usr/bin/cmake -E cmake_link_script CMakeFiles/simple.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/simple.dir/simple.cpp.o ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/simple -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 79%] Built target simple
/usr/bin/gmake -f examples/embd-input/CMakeFiles/embdinput.dir/build.make examples/embd-input/CMakeFiles/embdinput.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/embd-input /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input/CMakeFiles/embdinput.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/embd-input/CMakeFiles/embdinput.dir/build.make examples/embd-input/CMakeFiles/embdinput.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 81%] Building CXX object examples/embd-input/CMakeFiles/embdinput.dir/embd-input-lib.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/embd-input/CMakeFiles/embdinput.dir/embd-input-lib.cpp.o -MF CMakeFiles/embdinput.dir/embd-input-lib.cpp.o.d -o CMakeFiles/embdinput.dir/embd-input-lib.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/embd-input/embd-input-lib.cpp
[ 83%] Linking CXX static library libembdinput.a
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input && /usr/bin/cmake -P CMakeFiles/embdinput.dir/cmake_clean_target.cmake
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input && /usr/bin/cmake -E cmake_link_script CMakeFiles/embdinput.dir/link.txt --verbose=1
/usr/bin/ar qc libembdinput.a "CMakeFiles/embdinput.dir/embd-input-lib.cpp.o" ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o"
/usr/bin/ranlib libembdinput.a
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 83%] Built target embdinput
/usr/bin/gmake -f examples/embd-input/CMakeFiles/embd-input-test.dir/build.make examples/embd-input/CMakeFiles/embd-input-test.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/embd-input /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input/CMakeFiles/embd-input-test.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/embd-input/CMakeFiles/embd-input-test.dir/build.make examples/embd-input/CMakeFiles/embd-input-test.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 85%] Building CXX object examples/embd-input/CMakeFiles/embd-input-test.dir/embd-input-test.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/embd-input/CMakeFiles/embd-input-test.dir/embd-input-test.cpp.o -MF CMakeFiles/embd-input-test.dir/embd-input-test.cpp.o.d -o CMakeFiles/embd-input-test.dir/embd-input-test.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/embd-input/embd-input-test.cpp
[ 87%] Linking CXX executable ../../bin/embd-input-test
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/embd-input && /usr/bin/cmake -E cmake_link_script CMakeFiles/embd-input-test.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG "CMakeFiles/embd-input-test.dir/embd-input-test.cpp.o" ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/embd-input-test -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a libembdinput.a ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 87%] Built target embd-input-test
/usr/bin/gmake -f examples/server/CMakeFiles/server.dir/build.make examples/server/CMakeFiles/server.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/server /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/server /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/server/CMakeFiles/server.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f examples/server/CMakeFiles/server.dir/build.make examples/server/CMakeFiles/server.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 89%] Building CXX object examples/server/CMakeFiles/server.dir/server.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/server && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -DSERVER_VERBOSE=1 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/server -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT examples/server/CMakeFiles/server.dir/server.cpp.o -MF CMakeFiles/server.dir/server.cpp.o.d -o CMakeFiles/server.dir/server.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/server/server.cpp
[ 91%] Linking CXX executable ../../bin/server
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/examples/server && /usr/bin/cmake -E cmake_link_script CMakeFiles/server.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/server.dir/server.cpp.o ../CMakeFiles/common.dir/common.cpp.o "../CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/server -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64: ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 91%] Built target server
/usr/bin/gmake -f pocs/vdot/CMakeFiles/vdot.dir/build.make pocs/vdot/CMakeFiles/vdot.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/pocs/vdot /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot/CMakeFiles/vdot.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f pocs/vdot/CMakeFiles/vdot.dir/build.make pocs/vdot/CMakeFiles/vdot.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 93%] Building CXX object pocs/vdot/CMakeFiles/vdot.dir/vdot.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/pocs -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT pocs/vdot/CMakeFiles/vdot.dir/vdot.cpp.o -MF CMakeFiles/vdot.dir/vdot.cpp.o.d -o CMakeFiles/vdot.dir/vdot.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/pocs/vdot/vdot.cpp
[ 95%] Linking CXX executable ../../bin/vdot
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot && /usr/bin/cmake -E cmake_link_script CMakeFiles/vdot.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/vdot.dir/vdot.cpp.o ../../examples/CMakeFiles/common.dir/common.cpp.o "../../examples/CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/vdot -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64 ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 95%] Built target vdot
/usr/bin/gmake -f pocs/vdot/CMakeFiles/q8dot.dir/build.make pocs/vdot/CMakeFiles/q8dot.dir/depend
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/pocs/vdot /home/raul/Desktop/Proyecto/go-llama.cpp/build /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot/CMakeFiles/q8dot.dir/DependInfo.cmake --color=
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/gmake -f pocs/vdot/CMakeFiles/q8dot.dir/build.make pocs/vdot/CMakeFiles/q8dot.dir/build
gmake[3]: se entra en el directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[ 97%] Building CXX object pocs/vdot/CMakeFiles/q8dot.dir/q8dot.cpp.o
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot && /usr/bin/c++ -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DGGML_USE_CUBLAS -DGGML_USE_K_QUANTS -DK_QUANTS_PER_ITERATION=2 -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/pocs -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/examples/. -I/home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/. -O3 -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -mf16c -mfma -mavx -mavx2 -std=gnu++11 -MD -MT pocs/vdot/CMakeFiles/q8dot.dir/q8dot.cpp.o -MF CMakeFiles/q8dot.dir/q8dot.cpp.o.d -o CMakeFiles/q8dot.dir/q8dot.cpp.o -c /home/raul/Desktop/Proyecto/go-llama.cpp/llama.cpp/pocs/vdot/q8dot.cpp
[100%] Linking CXX executable ../../bin/q8dot
cd /home/raul/Desktop/Proyecto/go-llama.cpp/build/pocs/vdot && /usr/bin/cmake -E cmake_link_script CMakeFiles/q8dot.dir/link.txt --verbose=1
/usr/bin/c++ -O3 -DNDEBUG CMakeFiles/q8dot.dir/q8dot.cpp.o ../../examples/CMakeFiles/common.dir/common.cpp.o "../../examples/CMakeFiles/common.dir/grammar-parser.cpp.o" -o ../../bin/q8dot -L/usr/local/cuda-12.1/targets/x86_64-linux/lib/stubs -L/usr/local/cuda-12.1/targets/x86_64-linux/lib -Wl,-rpath,/usr/local/cuda-12.1/lib64 ../../libllama.a /usr/local/cuda-12.1/lib64/libcudart.so /usr/local/cuda-12.1/lib64/libcublas.so /usr/local/cuda-12.1/lib64/libculibos.a /usr/local/cuda-12.1/lib64/libcublasLt.so -lcudadevrt -lcudart_static -lrt -lpthread -ldl
gmake[3]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
[100%] Built target q8dot
gmake[2]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
/usr/bin/cmake -E cmake_progress_start /home/raul/Desktop/Proyecto/go-llama.cpp/build/CMakeFiles 0
gmake[1]: se sale del directorio '/home/raul/Desktop/Proyecto/go-llama.cpp/build'
cd build && cp -rf CMakeFiles/llama.dir/llama.cpp.o ../llama.cpp/llama.o
cd build && cp -rf examples/CMakeFiles/common.dir/common.cpp.o ../llama.cpp/common.o
cd build && cp -rf examples/CMakeFiles/common.dir/grammar-parser.cpp.o ../llama.cpp/grammar-parser.o
g++ -I./llama.cpp -I. -I./llama.cpp/examples -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -pthread -I./llama.cpp -I./llama.cpp/examples binding.cpp -o binding.o -c
binding.cpp: In function ‘int llama_predict(void*, void*, char*, bool)’:
binding.cpp:533:42: warning: cast from type ‘const char*’ to type ‘char*’ casts away qualifiers [-Wcast-qual]
533 | if (!tokenCallback(state_pr, (char*)token_str)) {
| ^~~~~~~~~~~~~~~~
binding.cpp:591:1: warning: label ‘end’ defined but not used [-Wunused-label]
591 | end:
| ^~~
binding.cpp: In function ‘void llama_binding_free_model(void*)’:
binding.cpp:613:5: warning: possible problem detected in invocation of ‘operator delete’ [-Wdelete-incomplete]
613 | delete ctx->model;
| ^~~~~~~~~~~~~~~~~
binding.cpp:613:17: warning: invalid use of incomplete type ‘struct llama_model’
613 | delete ctx->model;
| ~~~~~^~~~~
In file included from ./llama.cpp/examples/common.h:5,
from binding.cpp:1:
./llama.cpp/llama.h:66:12: note: forward declaration of ‘struct llama_model’
66 | struct llama_model;
| ^~~~~~~~~~~
binding.cpp:613:5: note: neither the destructor nor the class-specific ‘operator delete’ will be called, even if they are declared when the class is defined
613 | delete ctx->model;
| ^~~~~~~~~~~~~~~~~
cd build && cp -rf CMakeFiles/ggml.dir/k_quants.c.o ../llama.cpp/k_quants.o
cd build && cp -rf CMakeFiles/ggml.dir/ggml-cuda.cu.o ../llama.cpp/ggml-cuda.o
ar src libbinding.a llama.cpp/ggml.o llama.cpp/k_quants.o llama.cpp/ggml-cuda.o llama.cpp/common.o llama.cpp/grammar-parser.o llama.cpp/llama.o binding.o

@gulldan
Copy link

gulldan commented Aug 1, 2023

same problem
any idea for solution?

@johndpope
Copy link

what is your go version? 18 or 21

go version
go version go1.21.0 linux/amd64

@lstep
Copy link

lstep commented Aug 31, 2023

Having the same problem. Using go version 1.21 on Ubuntu Linux on latest version of go-llama.cpp.

$ BUILD_TYPE=cublas make libbinding.a
I llama.cpp build info:
I UNAME_S:  Linux
I UNAME_P:  x86_64
I UNAME_M:  x86_64
I CFLAGS:   -I./llama.cpp -I. -O3 -DNDEBUG -std=c11 -fPIC -Wall -Wextra -Wpedantic -Wcast-qual -Wdouble-promotion -Wshadow -Wstrict-prototypes -Wpointer-arith -Wno-unused-function -pthread -march=native -mtune=native
I CXXFLAGS: -I./llama.cpp -I. -I./llama.cpp/common -I./common -O3 -DNDEBUG -std=c++11 -fPIC -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -pthread
I CGO_LDFLAGS:
I LDFLAGS:
I BUILD_TYPE:  cublas
I CMAKE_ARGS:  -DLLAMA_CUBLAS=ON
I EXTRA_TARGETS:  llama.cpp/ggml-cuda.o
I CC:       cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
I CXX:      g++ (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0

cd llama.cpp && patch -p1 < ../patches/1902-cuda.patch
patching file common/common.cpp
patching file common/common.h
touch prepare
mkdir -p build
cd build && cmake ../llama.cpp -DLLAMA_CUBLAS=ON && VERBOSE=1 cmake --build . --config Release && cp -rf CMakeFiles/ggml.dir/ggml.c.o ../llama.cpp/ggml.o
-- The C compiler identification is GNU 11.4.0
-- The CXX compiler identification is GNU 11.4.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: /usr/bin/git (found version "2.34.1")
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE
-- Found CUDAToolkit: /home/lstep/.local/cuda-11.8/include (found version "11.8.89")
-- cuBLAS found
-- The CUDA compiler identification is NVIDIA 11.8.89
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /home/lstep/.local/cuda-11.8/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- Using CUDA architectures: 52;61;70
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- x86 detected
-- Configuring done
-- Generating done

[...]

cd build && cp -rf CMakeFiles/ggml.dir/k_quants.c.o ../llama.cpp/k_quants.o
cd build && cp -rf CMakeFiles/ggml.dir/ggml-cuda.cu.o ../llama.cpp/ggml-cuda.o
ar src libbinding.a llama.cpp/ggml.o llama.cpp/k_quants.o llama.cpp/ggml-cuda.o llama.cpp/ggml-alloc.o llama.cpp/common.o llama.cpp/grammar-parser.o llama.cpp/llama.o binding.o

Then

CGO_LDFLAGS="-lcublas -lcudart -L/usr/local/cuda/lib64/" LIBRARY_PATH=$PWD C_INCLUDE_PATH=$PWD go run ./examples -m "/home/raul/Desktop/Proyecto/models/llama-2-13b-chat.ggmlv3.q2_K.bin" -t 14

binding.cpp: In function ‘int llama_predict(void*, void*, char*, bool)’:
binding.cpp:323:53: warning: format ‘%zu’ expects argument of type ‘size_t’, but argument 2 has type ‘int’ [-Wformat=]
  323 |                 printf("<<input too long: skipped %zu token%s>>", skipped_tokens, skipped_tokens != 1 ? "s" : "");
      |                                                   ~~^             ~~~~~~~~~~~~~~
      |                                                     |             |
      |                                                     |             int
      |                                                     long unsigned int
      |                                                   %u
binding.cpp: In function ‘void llama_binding_free_model(void*)’:
binding.cpp:632:5: warning: possible problem detected in invocation of ‘operator delete’ [-Wdelete-incomplete]
  632 |     delete ctx->model;
      |     ^~~~~~~~~~~~~~~~~
binding.cpp:632:17: warning: invalid use of incomplete type ‘struct llama_model’
  632 |     delete ctx->model;
      |            ~~~~~^~~~~
In file included from ./llama.cpp/common/common.h:5,
                 from binding.cpp:1:
./llama.cpp/llama.h:60:12: note: forward declaration of ‘struct llama_model’
   60 |     struct llama_model;
      |            ^~~~~~~~~~~
binding.cpp:632:5: note: neither the destructor nor the class-specific ‘operator delete’ will be called, even if they are declared when the class is defined
  632 |     delete ctx->model;
      |     ^~~~~~~~~~~~~~~~~
create_gpt_params: loading model /home/raul/Desktop/Proyecto/models/llama-2-13b-chat.ggmlv3.q2_K.bin
SIGSEGV: segmentation violation
PC=0x7f5de394bfbd m=0 sigcode=1
signal arrived during cgo execution

goroutine 1 [syscall]:
runtime.cgocall(0x49e1c0, 0xc000057ab8)
        /snap/go/10319/src/runtime/cgocall.go:157 +0x4b fp=0xc000057a90 sp=0xc000057a58 pc=0x413eab
github.com/go-skynet/go-llama%2ecpp._Cfunc_load_model(0x1d39bb0, 0x80, 0x0, 0x1, 0x0, 0x1, 0x1, 0x0, 0x0, 0x200, ...)
        _cgo_gotypes.go:266 +0x4c fp=0xc000057ab8 sp=0xc000057a90 pc=0x49ac6c
github.com/go-skynet/go-llama%2ecpp.New({0x7ffda726c211, 0x43}, {0xc000057e28, 0x4, 0x1?})
        /home/lstep/AI/GO-LLAMA.CPP/go-llama.cpp/llama.go:39 +0x3aa fp=0xc000057cb8 sp=0xc000057ab8 pc=0x49b3aa
main.main()
        /home/lstep/AI/GO-LLAMA.CPP/go-llama.cpp/examples/main.go:35 +0x37d fp=0xc000057f40 sp=0xc000057cb8 pc=0x49d47d
runtime.main()
        /snap/go/10319/src/runtime/proc.go:267 +0x2bb fp=0xc000057fe0 sp=0xc000057f40 pc=0x4448fb
runtime.goexit()
        /snap/go/10319/src/runtime/asm_amd64.s:1650 +0x1 fp=0xc000057fe8 sp=0xc000057fe0 pc=0x46e981

goroutine 2 [force gc (idle)]:
runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
        /snap/go/10319/src/runtime/proc.go:398 +0xce fp=0xc000042fa8 sp=0xc000042f88 pc=0x444d4e
runtime.goparkunlock(...)
        /snap/go/10319/src/runtime/proc.go:404
runtime.forcegchelper()
        /snap/go/10319/src/runtime/proc.go:322 +0xb3 fp=0xc000042fe0 sp=0xc000042fa8 pc=0x444bd3
runtime.goexit()
        /snap/go/10319/src/runtime/asm_amd64.s:1650 +0x1 fp=0xc000042fe8 sp=0xc000042fe0 pc=0x46e981
created by runtime.init.6 in goroutine 1
        /snap/go/10319/src/runtime/proc.go:310 +0x1a
        
goroutine 3 [GC sweep wait]:
runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
        /snap/go/10319/src/runtime/proc.go:398 +0xce fp=0xc000043778 sp=0xc000043758 pc=0x444d4e
runtime.goparkunlock(...)
        /snap/go/10319/src/runtime/proc.go:404
runtime.bgsweep(0x0?)
        /snap/go/10319/src/runtime/mgcsweep.go:280 +0x94 fp=0xc0000437c8 sp=0xc000043778 pc=0x431654
runtime.gcenable.func1()
        /snap/go/10319/src/runtime/mgc.go:200 +0x25 fp=0xc0000437e0 sp=0xc0000437c8 pc=0x426985
runtime.goexit()
        /snap/go/10319/src/runtime/asm_amd64.s:1650 +0x1 fp=0xc0000437e8 sp=0xc0000437e0 pc=0x46e981
created by runtime.gcenable in goroutine 1
        /snap/go/10319/src/runtime/mgc.go:200 +0x66

goroutine 4 [GC scavenge wait]:
runtime.gopark(0xc000020070?, 0x596648?, 0x1?, 0x0?, 0xc0000071e0?)
        /snap/go/10319/src/runtime/proc.go:398 +0xce fp=0xc000043f70 sp=0xc000043f50 pc=0x444d4e
runtime.goparkunlock(...)
        /snap/go/10319/src/runtime/proc.go:404
runtime.(*scavengerState).park(0xa0ed40)
        /snap/go/10319/src/runtime/mgcscavenge.go:425 +0x49 fp=0xc000043fa0 sp=0xc000043f70 pc=0x42ee89
runtime.bgscavenge(0x0?)
        /snap/go/10319/src/runtime/mgcscavenge.go:653 +0x3c fp=0xc000043fc8 sp=0xc000043fa0 pc=0x42f41c
runtime.gcenable.func2()
        /snap/go/10319/src/runtime/mgc.go:201 +0x25 fp=0xc000043fe0 sp=0xc000043fc8 pc=0x426925
runtime.goexit()
        /snap/go/10319/src/runtime/asm_amd64.s:1650 +0x1 fp=0xc000043fe8 sp=0xc000043fe0 pc=0x46e981
created by runtime.gcenable in goroutine 1
        /snap/go/10319/src/runtime/mgc.go:201 +0xa5
        
goroutine 5 [finalizer wait]:
runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
        /snap/go/10319/src/runtime/proc.go:398 +0xce fp=0xc000044628 sp=0xc000044608 pc=0x444d4e
runtime.runfinq()
        /snap/go/10319/src/runtime/mfinal.go:193 +0x107 fp=0xc0000447e0 sp=0xc000044628 pc=0x425a07
runtime.goexit()
        /snap/go/10319/src/runtime/asm_amd64.s:1650 +0x1 fp=0xc0000447e8 sp=0xc0000447e0 pc=0x46e981
created by runtime.createfing in goroutine 1
        /snap/go/10319/src/runtime/mfinal.go:163 +0x3d

rax    0x0
rbx    0x1d3c900
rcx    0x4
rdx    0x0
rdi    0x1d3c900
rsi    0x7ffda726b4e0
rbp    0x7ffda726b4e0
rsp    0x7ffda726b1c0
r8     0x65
r9     0x7fffffff
r10    0x7f5de380d258
r11    0x7f5de394bf70
r12    0x0
r13    0x0
r14    0x1d3c910
r15    0x7ffda726b4c0
rip    0x7f5de394bfbd
rflags 0x10246
cs     0x33
fs     0x0
gs     0x0
exit status 2

@amorroxic
Copy link

amorroxic commented Sep 28, 2023

same issue. I'm able to compile (some warnings only):

BUILD_TYPE=cublas make libbinding.a

then able to load the model on GPU:

LIBRARY_PATH=$PWD C_INCLUDE_PATH=$PWD go run -tags cublas ./examples -m "[..]/Vicuna-7B-CoT-GGUF/vicuna-7b-cot.Q5_K_M.gguf" -ngl 35

resulting in:

..
llama_model_loader: - type  f32:   65 tensors
llama_model_loader: - type q5_K:  193 tensors
llama_model_loader: - type q6_K:   33 tensors
llm_load_print_meta: format         = GGUF V2 (latest)
llm_load_print_meta: arch           = llama
llm_load_print_meta: vocab type     = SPM
llm_load_print_meta: n_vocab        = 32000
llm_load_print_meta: n_merges       = 0
llm_load_print_meta: n_ctx_train    = 2048
llm_load_print_meta: n_ctx          = 128
llm_load_print_meta: n_embd         = 4096
llm_load_print_meta: n_head         = 32
llm_load_print_meta: n_head_kv      = 32
llm_load_print_meta: n_layer        = 32
llm_load_print_meta: n_rot          = 128
llm_load_print_meta: n_gqa          = 1
llm_load_print_meta: f_norm_eps     = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-06
llm_load_print_meta: n_ff           = 11008
llm_load_print_meta: freq_base      = 10000.0
llm_load_print_meta: freq_scale     = 1
llm_load_print_meta: model type     = 7B
llm_load_print_meta: model ftype    = mostly Q5_K - Medium
llm_load_print_meta: model params   = 6.74 B
llm_load_print_meta: model size     = 4.45 GiB (5.68 BPW) 
llm_load_print_meta: general.name   = kevinpro_vicuna-7b-cot
llm_load_print_meta: BOS token = 1 '<s>'
llm_load_print_meta: EOS token = 2 '</s>'
llm_load_print_meta: UNK token = 0 '<unk>'
llm_load_print_meta: LF token  = 13 '<0x0A>'
llm_load_tensors: ggml ctx size =    0.09 MB
llm_load_tensors: using CUDA for GPU acceleration
llm_load_tensors: mem required  =   86.03 MB (+   64.00 MB per state)
llm_load_tensors: offloading 32 repeating layers to GPU
llm_load_tensors: offloading non-repeating layers to GPU
llm_load_tensors: offloading v cache to GPU
llm_load_tensors: offloading k cache to GPU
llm_load_tensors: offloaded 35/35 layers to GPU
llm_load_tensors: VRAM used: 4539 MB
..................................................................................................
llama_new_context_with_model: kv self size  =   64.00 MB
llama_new_context_with_model: compute buffer total size =   19.09 MB
llama_new_context_with_model: VRAM scratch buffer: 17.63 MB
Model loaded successfully.

but then segfaults on inference:

>>> hello

Sending hello

 hello
SIGSEGV: segmentation violation
PC=0x521dc0 m=0 sigcode=1
signal arrived during cgo execution

goroutine 1 [syscall]:
runtime.cgocall(0x4a06c0, 0xc000065968)
	/usr/local/go1.21/src/runtime/cgocall.go:157 +0x4b fp=0xc000065940 sp=0xc000065908 pc=0x415fcb
github.com/go-skynet/go-llama%2ecpp._Cfunc_llama_predict(0x274f6e70, 0x18b4d30, 0xc000026200, 0x1)
	_cgo_gotypes.go:238 +0x4b fp=0xc000065968 sp=0xc000065940 pc=0x49cfcb
github.com/go-skynet/go-llama%2ecpp.(*LLama).Predict.func2(0x582fa0?, 0xc000065b70?, {0xc000026200, 0x1?, 0xc000065a28?}, 0xc0000140d6?)
..

@mudler
Copy link
Member

mudler commented Sep 28, 2023

this should be fixed already. can you try master? which version are you running on?

@amorroxic
Copy link

amorroxic commented Sep 28, 2023

same result unfortunately (just cloned again fresh):

git clone --recurse-submodules https://github.com/go-skynet/go-llama.cpp
cd go-llama.cpp
BUILD_TYPE=cublas make libbinding.a
LIBRARY_PATH="/usr/local/cuda/lib64:$PWD" C_INCLUDE_PATH=$PWD go run -tags cublas ./examples -m "/mnt/booterstown/gguf-models/Vicuna-7B-CoT-GGUF/vicuna-7b-cot.Q5_K_M.gguf" -ngl 35

able to load the model (confirmed also by nvidia-smi):

...
llm_load_tensors: offloaded 35/35 layers to GPU
llm_load_tensors: VRAM used: 4539 MB
..................................................................................................
llama_new_context_with_model: kv self size  =   64.00 MB
llama_new_context_with_model: compute buffer total size =   19.09 MB
llama_new_context_with_model: VRAM scratch buffer: 17.63 MB
Model loaded successfully.

with the same segfault on inference:

>>> hello

Sending hello

 hello
SIGSEGV: segmentation violation
PC=0x521dc0 m=0 sigcode=1
signal arrived during cgo execution

goroutine 1 [syscall]:
runtime.cgocall(0x4a06c0, 0xc000065968)
	/usr/local/go1.21/src/runtime/cgocall.go:157 +0x4b fp=0xc000065940 sp=0xc000065908 pc=0x415fcb
github.com/go-skynet/go-llama%2ecpp._Cfunc_llama_predict(0x274f6e70, 0x18b4d30, 0xc000026200, 0x1)
	_cgo_gotypes.go:238 +0x4b fp=0xc000065968 sp=0xc000065940 pc=0x49cfcb
..

I believe (not sure, far from competent in C, used LOG() function to chase this) the problem may originate on this line (both llama_sample_token_binding / llama_sample_token seem to produce the crash, compiled both ways):

const llama_token id = llama_sample_token_binding(ctx, ctx_guidance, grammar, params_p, last_tokens, candidates);

const llama_token id = llama_sample_token_binding(ctx, ctx_guidance, grammar, params_p, last_tokens, candidates);
//const llama_token id = llama_sample_token(ctx, ctx_guidance, grammar, params, last_tokens, candidates);

my env (also tried cuda 11.2 with nvidia driver 470.199.02):

$ lsb_release -a
No LSB modules are available.
Distributor ID:	Ubuntu
Description:	Ubuntu 22.04.3 LTS
Release:	22.04
Codename:	jammy

$ go version
go version go1.21.1 linux/amd64

$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Wed_Sep_21_10:33:58_PDT_2022
Cuda compilation tools, release 11.8, V11.8.89
Build cuda_11.8.r11.8/compiler.31833905_0

$ cat /proc/driver/nvidia/version 
NVRM version: NVIDIA UNIX x86_64 Kernel Module  535.104.05  Sat Aug 19 01:15:15 UTC 2023

$ gcc --version
gcc (Ubuntu 10.5.0-1ubuntu1~22.04) 10.5.0

Hope that's useful somehow, cheers!

[edit]
tried building llama.cpp (via cmake) and and the examples/main binary seems to work on GPU as well

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

6 participants