LLamaSharp
LLamaSharp copied to clipboard
CentOS x86_64 Failed Loading 'libllama.so'
[llama Info]: NativeLibraryConfig Description:
- LibraryName: Llama
- Path: ''
- PreferCuda: True
- PreferredAvxLevel: AVX
- AllowFallback: True
- SkipCheck: False
- SearchDirectories and Priorities: { ./ } [llama Debug]: Loading library: 'llama' [llama Info]: Detected OS Platform: 'LINUX' [llama Debug]: Detected OS string: 'linux-x64' [llama Debug]: Detected extension string: '.so' [llama Debug]: Detected prefix string: 'lib' [llama Info]: Detected cuda major version 12. [llama Debug]: Trying './runtimes/linux-x64/native/cuda12/libllama.so' [llama Info]: Failed Loading './runtimes/linux-x64/native/cuda12/libllama.so' [llama Debug]: Trying './runtimes/linux-x64/native/avx/libllama.so' [llama Info]: Loaded './runtimes/linux-x64/native/avx/libllama.so'
Sorry for the late reply, is this problem already been solved?
BTW, just for curiosity, are you using OpenVINO along with LLamaSharp?
Sorry for the late reply, is this problem already been solved?
BTW, just for curiosity, are you using OpenVINO along with LLamaSharp?
yeap bt a other questions that we use the GPU model but it's actually using a CPU...
Sorry for the late reply, is this problem already been solved?
BTW, just for curiosity, are you using OpenVINO along with LLamaSharp?
Yes, for this sln we also use OpenVINO to process the images.
Is avx
the highest avx level supported by your CPU?