error loading model: invalid model: tensor 'blk.0.attn_norm.weight' is duplicated

[gomidas]

I HAVE AN ISSUE ABOUT LOADING MODEL. I AM NOT FAMILIAR WITH LLAMA GGML OR GGUF.

MY STEPS:

1. I DOWNLOADED REPO.

2. MY CMAKE COMMAND:

cmake .. -G "Visual Studio 17 2022" -A x64 -T version=14.38.33130 ^
 -DCMAKE_BUILD_TYPE=Release ^
 -DCMAKE_CXX_FLAGS="-march=native -mno-avx512" ^
 -DBUILD_SHARED_LIBS=OFF ^
 -DLLAMA_BUILD_TESTS=OFF ^
 -DLLAMA_BUILD_EXAMPLES=OFF ^
 -DLLAMA_BUILD_COMMON=OFF ^
 -DLLAMA_BUILD_SERVER=OFF ^
 -DLLAMA_NATIVE=ON ^
 -DCURL_INCLUDE_DIR="C:/Users/gomi/Downloads/curl-8.13.0/include/" ^
 -DCURL_LIBRARY="C:/Users/gomi/Downloads/curl-8.13.0/build/lib/Release/libcurl.lib" ^
 -DCURL_STATICLIB=ON ^
 -DCURL_USE_SCHANNEL=ON ^
 -DCMAKE_USE_PTHREADS=OFF ^
 -DCMAKE_CXX_FLAGS_RELEASE="/MD /O2 /DNDEBUG" ^
 -DCMAKE_CXX_FLAGS_DEBUG="/MDd /Zi /Ob0 /Od"

I DID IT LIKE THIS TO MAKE .lib FILES COMPATIBLE WITH UNREAL ENGINE 5.

3. I GET FOLLOWING LIB FILES:

ggml.lib (3MB)
ggml-base.lib (7MB)
ggml-cpu.lib (6MB)
llama.lib (50MB)

I MAKE A NEW VS2022 C++ CONSOLE APP TO TEST ".libs"

3. MY PROJECT CONFIG:

C/C++:
/permissive- /ifcOutput "x64\Release\" /GS /GL /W3 /Gy /Zc:wchar_t /I"C:\Users\gomi\source\repos\llamaTest\include\ggml" /I"C:\Users\gomi\source\repos\llamaTest\include\llama" /Gm- /O2 /sdl /Fd"x64\Release\vc143.pdb" /Zc:inline /fp:precise /D "NDEBUG" /D "_CONSOLE" /D "_CRT_SECURE_NO_WARNINGS" /D "_UNICODE" /D "UNICODE" /errorReport:prompt /WX- /Zc:forScope /Gd /Oi /MD /std:c++17 /FC /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Fp"x64\Release\llamaTest.pch" /diagnostics:column

LINKER:
/OUT:"C:\Users\gomi\source\repos\llamaTest\x64\Release\llamaTest.exe" /MANIFEST /LTCG:incremental /NXCOMPAT /PDB:"C:\Users\gomi\source\repos\llamaTest\x64\Release\llamaTest.pdb" /DYNAMICBASE "ggml.lib" "ggml-base.lib" "ggml-cpu.lib" "llama.lib" "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib" "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib" "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib" /DEBUG /MACHINE:X64 /OPT:REF /PGD:"C:\Users\gomi\source\repos\llamaTest\x64\Release\llamaTest.pgd" /SUBSYSTEM:CONSOLE /MANIFESTUAC:"level='asInvoker' uiAccess='false'" /ManifestFile:"x64\Release\llamaTest.exe.intermediate.manifest" /LTCGOUT:"x64\Release\llamaTest.iobj" /OPT:ICF /ERRORREPORT:PROMPT /ILK:"x64\Release\llamaTest.ilk" /NOLOGO /LIBPATH:"C:\Users\gomi\source\repos\llamaTest\lib" /TLBID:1

I DOWNLOADED THIS MODEL: https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF

4. I USE THIS SIMPLE CODE TO LOAD MODEL:

#include "llama.h"
#include <stdio.h>

int main(int argc, char** argv) {

    // load dynamic backends
    llama_backend_init();
    ggml_backend_t backendggml = ggml_backend_init_best();

    // initialize the model
    llama_model_params model_params = llama_model_default_params();
    llama_model* model = llama_model_load_from_file("G:/GGUF/llama-2-13b-chat.Q8_0.gguf", model_params);
    const llama_vocab* vocab = llama_model_get_vocab(model);

    return 0;
}

OUTPUT:

llama_model_load: error loading model: invalid model: tensor 'blk.0.attn_norm.weight' is duplicated
llama_model_load_from_file_impl: failed to load model

PROJECT FILES: llamaTest.zip

I GET SAME ERROR FOR ANY MODEL DOWNLOADED.

WHAT IS THE WAY I SOLVE THIS PROBLEM?

THANK YOU.

[gomidas]

I directly downloaded from website. Today I downloaded with command and compiled. Now I do not have this problem.