cmake -S . -B build -DLLAMA_CUBLAS=ON

[hungptit123]

(powerinfer) root@ubuntu:~/hungdv/LLM/PowerInfer# cmake -S . -B build -DLLAMA_CUBLAS=ON -- cuBLAS found -- Using CUDA architectures: 52;61;70 GNU ld (GNU Binutils for Ubuntu) 2.30 -- CMAKE_SYSTEM_PROCESSOR: x86_64 -- x86 detected -- Configuring done (0.2s) CMake Error at CMakeLists.txt:661 (target_link_libraries): Target "ggml" links to:

CUDA::cublasLt

but the target was not found. Possible reasons include:

* There is a typo in the target name.
* A find_package call is missing for an IMPORTED target.
* An ALIAS target is missing.

CMake Error at CMakeLists.txt:683 (target_link_libraries): Target "llama" links to:

CUDA::cublasLt

but the target was not found. Possible reasons include:

* There is a typo in the target name.
* A find_package call is missing for an IMPORTED target.
* An ALIAS target is missing.

-- Generating done (0.3s) CMake Generate step failed. Build files cannot be regenerated correctly.

[hodlen]

It seems there might be an incompatibility between your building toolchain and our CMake settings. Could you please provide the versions of your cmake, g++, and nvcc? This information will assist us in identifying the issue. For reference, our environment uses:

• cmake: 3.25.1
• g++: 10.5.0
• nvcc: 11.8.89

Also, please try to prune the /build directory and reconfigure CMake to see if there is a difference. Sometimes CMake cache can make trouble.

After trying these steps, we'd appreciate it if you could let us know the outcome. Your feedback is crucial for us to better understand and address such issues.