Closed ThomasVuillaume closed 10 months ago
@ThomasVuillaume Does running this command without --cuda
works for you?
npx --no node-llama-cpp download
Yes, this one's working fine :
vuith@THOMZ-FIXE MINGW64 ~/GitRepositories/node-llama-cpp-experiment (main)
$ npx --no node-llama-cpp download
Repo: ggerganov/llama.cpp
Release: b1378
✔ Fetched llama.cpp info
✔ Removed existing llama.cpp directory
Cloning llama.cpp
Clone ggerganov/llama.cpp (local bundle) 100% ████████████████████████████████████████ 0s
◷ Compiling llama.cpp
Not searching for unused variables given on the command line.
-- The C compiler identification is MSVC 19.37.32825.0
-- The CXX compiler identification is MSVC 19.37.32825.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe - 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: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: C:/Program Files/Git/mingw64/bin/git.exe (found version "2.42.0.windows.2")
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - not found
-- Found Threads: TRUE
-- CMAKE_SYSTEM_PROCESSOR: AMD64
-- CMAKE_GENERATOR_PLATFORM: x64
-- x86 detected
-- Configuring done (4.8s)
-- Generating done (0.1s)
-- Build files have been written to: C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/build
Version MSBuild 17.7.2+d6990bcfa pour .NET Framework
1>Checking Build System
Generating build details from Git
-- Found Git: C:/Program Files/Git/mingw64/bin/git.exe (found version "2.42.0.windows.2")
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
ggml.c
ggml-alloc.c
ggml-backend.c
k_quants.c
Génération de code en cours...
ggml.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llam
a.cpp\ggml.dir\Release\ggml.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
llama.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1054,31)
: warning C4305: 'initialisation' : troncation de 'double' à 'float' [C:\Users\vuith\GitRepositories\node-llama-cpp-exp
eriment\node_modules\node-llama-cpp\llama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1202,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u0120' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1203,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u010A' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1204,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u0120' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1205,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u010A' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(2235,60)
: warning C4566: le caractère représenté par le nom de caractère universel '\u010A' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(9452,28)
: warning C4146: opérateur moins unaire appliqué à un type non signé, le résultat sera non signé [C:\Users\vuith\GitRep
ositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(9482,28)
: warning C4146: opérateur moins unaire appliqué à un type non signé, le résultat sera non signé [C:\Users\vuith\GitRep
ositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\llama.vcxproj]
llama.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\lla
ma.cpp\Release\llama.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/common/CMakeLists.txt
common.cpp
sampling.cpp
console.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\console.cpp
(253,30): warning C4267: 'initialisation' : conversion de 'size_t' en 'DWORD', perte possible de données [C:\Users\vuit
h\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\common\common.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\console.cpp
(407,28): warning C4267: 'initialisation' : conversion de 'size_t' en 'int', perte possible de données [C:\Users\vuith\
GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\common\common.vcxproj]
grammar-parser.cpp
train.cpp
Génération de code en cours...
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\common.cpp(
785): warning C4715: 'gpt_random_prompt' : les chemins de contrôle ne retournent pas tous une valeur [C:\Users\vuith\Gi
tRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\common\common.vcxproj]
common.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\ll
ama.cpp\common\common.dir\Release\common.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
ggml_static.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\bui
ld\llama.cpp\Release\ggml_static.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/CMake
Lists.txt
addon.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\log.h(379,1
9): warning C4996: 'fopen': This function or variable may be unsafe. Consider using fopen_s instead. To disable depreca
tion, use _CRT_SECURE_NO_WARNINGS. See online help for details. [C:\Users\vuith\GitRepositories\node-llama-cpp-experime
nt\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\log.h(387,9
7): warning C4996: 'strerror': This function or variable may be unsafe. Consider using strerror_s instead. To disable d
eprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. [C:\Users\vuith\GitRepositories\node-llama-cpp-ex
periment\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\addon.cpp(339,53): warning C
4267: 'argument' : conversion de 'size_t' en 'int32_t', perte possible de données [C:\Users\vuith\GitRepositories\node-
llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\addon.cpp(341,33): warning C
4267: '=' : conversion de 'size_t' en 'int32_t', perte possible de données [C:\Users\vuith\GitRepositories\node-llama-c
pp-experiment\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\addon.cpp(415,9): warning C4
996: 'llama_sample_temperature': use llama_sample_temp instead [C:\Users\vuith\GitRepositories\node-llama-cpp-experimen
t\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
win_delay_load_hook.cc
Génération de code en cours...
Création de la bibliothèque C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/l
lama/build/Release/llama-addon.lib et de l'objet C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_module
s/node-llama-cpp/llama/build/Release/llama-addon.exp
llama-addon.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\bui
ld\Release\llama-addon.node
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/CMake
Lists.txt
✔ Compiled llama.cpp
Repo: ggerganov/llama.cpp
Release: b1378
Done
Strange this is, as I said in the first message : I tested to compile with CUDA direclty on the llama.ccp project with CMAKE, without error..
Also having this issue. Compilation fails in VS 2019 and VS 2017 as well. This seems to be related to setting the option CMAKE_GENERATOR_TOOLSET
. I was able to successfully compile with CUDA after removing this option.
Thanks for the tip @sk-1982 ! I just tested to remove this line of code in dist\utils\compileLLamaCpp.js
(in my node_modules) and it worked :
vuith@THOMZ-FIXE MINGW64 ~/GitRepositories/node-llama-cpp-experiment (main)
$ npx --no node-llama-cpp download --cuda
Repo: ggerganov/llama.cpp
Release: b1378
CUDA: enabled
✔ Fetched llama.cpp info
✔ Removed existing llama.cpp directory
Cloning llama.cpp
Clone ggerganov/llama.cpp (local bundle) 100% ████████████████████████████████████████ 0s
◷ Compiling llama.cpp
Not searching for unused variables given on the command line.
-- The C compiler identification is MSVC 19.37.32825.0
-- The CXX compiler identification is MSVC 19.37.32825.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe - 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: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.37.32822/bin/Hostx64/x64/cl.exe - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: C:/Program Files/Git/mingw64/bin/git.exe (found version "2.42.0.windows.2")
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - not found
-- Found Threads: TRUE
-- Found CUDAToolkit: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.2/include (found version "12.2.140")
-- cuBLAS found
-- The CUDA compiler identification is NVIDIA 12.2.140
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.2/bin/nvcc.exe - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- Using CUDA architectures: 52;61;70
-- CMAKE_SYSTEM_PROCESSOR: AMD64
-- CMAKE_GENERATOR_PLATFORM: x64
-- x86 detected
-- Configuring done (22.1s)
-- Generating done (0.1s)
-- Build files have been written to: C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/build
Version MSBuild 17.7.2+d6990bcfa pour .NET Framework
1>Checking Build System
Generating build details from Git
-- Found Git: C:/Program Files/Git/mingw64/bin/git.exe (found version "2.42.0.windows.2")
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
Compiling CUDA source file ..\..\llama.cpp\ggml-cuda.cu...
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp>"C:\Progra
m Files\NVIDIA GPU Computing Toolkit\CUDA\v12.2\bin\nvcc.exe" --use-local-env -ccbin "C:\Program Files\Microsoft Vis
ual Studio\2022\Community\VC\Tools\MSVC\14.37.32822\bin\HostX64\x64" -x cu -I"C:\Users\vuith\GitRepositories\node-l
lama-cpp-experiment\node_modules\node-addon-api" -I"C:\Users\vuith\.cmake-js\node-x64\v18.18.2\include\node" -I"C:\Us
ers\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\." -I"C:\Program File
s\NVIDIA GPU Computing Toolkit\CUDA\v12.2\include" -I"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.2\includ
e" --keep-dir x64\Release -use_fast_math -maxrregcount=0 --machine 64 --compile -cudart static --generate-code=
arch=compute_52,code=[compute_52,sm_52] --generate-code=arch=compute_61,code=[compute_61,sm_61] --generate-code=arch=
compute_70,code=[compute_70,sm_70] -Xcompiler="/EHsc -Ob2" -D_WINDOWS -DNDEBUG -DNAPI_VERSION=7 -DGGML_USE_K_QUANTS
-DGGML_USE_CUBLAS -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DK_QUANTS_PER_ITERATION=2 -DGGML_CUDA_PEER_MAX_BATCH_SI
ZE=128 -D_CRT_SECURE_NO_WARNINGS -D_XOPEN_SOURCE=600 -D"CMAKE_INTDIR=\"Release\"" -D_MBCS -DWIN32 -D_WINDOWS -DNDEBUG
-DNAPI_VERSION=7 -DGGML_USE_K_QUANTS -DGGML_USE_CUBLAS -DGGML_CUDA_DMMV_X=32 -DGGML_CUDA_MMV_Y=1 -DK_QUANTS_PER_ITER
ATION=2 -DGGML_CUDA_PEER_MAX_BATCH_SIZE=128 -D_CRT_SECURE_NO_WARNINGS -D_XOPEN_SOURCE=600 -D"CMAKE_INTDIR=\"Release\"
" -Xcompiler "/EHsc /W3 /nologo /O2 /FS /MD /GR" -Xcompiler "/Fdggml.dir\Release\ggml.pdb" -o ggml.dir\Release\ggml
-cuda.obj "C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\ggml-
cuda.cu"
ggml-cuda.cu
tmpxft_00004eb8_00000000-7_ggml-cuda.compute_70.cudafe1.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\ggml-cuda.cu(5668)
: warning C4477: 'fprintf' : la chaîne de format '%ld' nécessite un argument de type 'long', mais l'argument variadique
1 est de type 'int64_t' [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\bu
ild\llama.cpp\ggml.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\ggml-cuda.cu(566
8): note: utilisez '%lld' dans la chaîne de format
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\ggml-cuda.cu(566
8): note: utilisez '%Id' dans la chaîne de format
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\ggml-cuda.cu(566
8): note: utilisez '%I64d' dans la chaîne de format
ggml.c
ggml-alloc.c
ggml-backend.c
k_quants.c
Génération de code en cours...
ggml.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llam
a.cpp\ggml.dir\Release\ggml.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
llama.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1054,31)
: warning C4305: 'initialisation' : troncation de 'double' à 'float' [C:\Users\vuith\GitRepositories\node-llama-cpp-exp
eriment\node_modules\node-llama-cpp\llama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1202,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u0120' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1203,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u010A' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1204,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u0120' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(1205,40)
: warning C4566: le caractère représenté par le nom de caractère universel '\u010A' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(2235,60)
: warning C4566: le caractère représenté par le nom de caractère universel '\u010A' ne peut pas être représenté dans la
page de codes actuelle (1252) [C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\ll
ama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(9452,28)
: warning C4146: opérateur moins unaire appliqué à un type non signé, le résultat sera non signé [C:\Users\vuith\GitRep
ositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\llama.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\llama.cpp(9482,28)
: warning C4146: opérateur moins unaire appliqué à un type non signé, le résultat sera non signé [C:\Users\vuith\GitRep
ositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\llama.vcxproj]
llama.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\lla
ma.cpp\Release\llama.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/common/CMakeLists.txt
common.cpp
sampling.cpp
console.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\console.cpp
(253,30): warning C4267: 'initialisation' : conversion de 'size_t' en 'DWORD', perte possible de données [C:\Users\vuit
h\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\common\common.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\console.cpp
(407,28): warning C4267: 'initialisation' : conversion de 'size_t' en 'int', perte possible de données [C:\Users\vuith\
GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\common\common.vcxproj]
grammar-parser.cpp
train.cpp
Génération de code en cours...
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\common.cpp(
785): warning C4715: 'gpt_random_prompt' : les chemins de contrôle ne retournent pas tous une valeur [C:\Users\vuith\Gi
tRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama.cpp\common\common.vcxproj]
common.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\ll
ama.cpp\common\common.dir\Release\common.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/llama
.cpp/CMakeLists.txt
ggml_static.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\bui
ld\llama.cpp\Release\ggml_static.lib
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/CMake
Lists.txt
addon.cpp
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\log.h(379,1
9): warning C4996: 'fopen': This function or variable may be unsafe. Consider using fopen_s instead. To disable depreca
tion, use _CRT_SECURE_NO_WARNINGS. See online help for details. [C:\Users\vuith\GitRepositories\node-llama-cpp-experime
nt\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\llama.cpp\common\log.h(387,9
7): warning C4996: 'strerror': This function or variable may be unsafe. Consider using strerror_s instead. To disable d
eprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. [C:\Users\vuith\GitRepositories\node-llama-cpp-ex
periment\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\addon.cpp(339,53): warning C
4267: 'argument' : conversion de 'size_t' en 'int32_t', perte possible de données [C:\Users\vuith\GitRepositories\node-
llama-cpp-experiment\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\addon.cpp(341,33): warning C
4267: '=' : conversion de 'size_t' en 'int32_t', perte possible de données [C:\Users\vuith\GitRepositories\node-llama-c
pp-experiment\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\addon.cpp(415,9): warning C4
996: 'llama_sample_temperature': use llama_sample_temp instead [C:\Users\vuith\GitRepositories\node-llama-cpp-experimen
t\node_modules\node-llama-cpp\llama\build\llama-addon.vcxproj]
win_delay_load_hook.cc
Génération de code en cours...
Création de la bibliothèque C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/l
lama/build/Release/llama-addon.lib et de l'objet C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_module
s/node-llama-cpp/llama/build/Release/llama-addon.exp
llama-addon.vcxproj -> C:\Users\vuith\GitRepositories\node-llama-cpp-experiment\node_modules\node-llama-cpp\llama\bui
ld\Release\llama-addon.node
Building Custom Rule C:/Users/vuith/GitRepositories/node-llama-cpp-experiment/node_modules/node-llama-cpp/llama/CMake
Lists.txt
✔ Compiled llama.cpp
Repo: ggerganov/llama.cpp
Release: b1378
Done
@sk-1982 Interesting. I originally added it to solve a compilation issue with CUDA, but since it seems to create another issue I'll remove it then. Thanks for the help :)
:tada: This issue has been resolved in version 2.7.4 :tada:
The release is available on:
Your semantic-release bot :package::rocket:
Issue description
CUDA compilation fails with code
Expected Behavior
Compilation with CUDA should work
Actual Behavior
I have an error MSB8020 whan trying to recompile with CUDA. I tested to compile with CUDA direclty on the llama.ccp project with CMAKE, without error.
Steps to reproduce
Lauching
npx --no node-llama-cpp download --cuda
My Environment
node-llama-cpp
versionAdditional Context
The logs :
Relevant Features Used
Are you willing to resolve this issue by submitting a Pull Request?
No, I don’t have the time and I’m okay to wait for the community / maintainers to resolve this issue.