[BUG]: Error in version 0.25.0 - LLama.Exceptions.RuntimeError: Failed to load the native library.

[rockofme1]

Description

After upgrading to version 0.25.0, an error occurs when downloading:

LLama.Exceptions.RuntimeError: Failed to load the native library. Please check the log for more information.
   at LLama.Native.NativeLibraryUtils.TryLoadLibrary(NativeLibraryConfig config, INativeLibrary& loadedLibrary)
   at LLama.Native.NativeLibraryConfig.DryRun(INativeLibrary& loadedLibrary)
   at LLama.Native.NativeLibraryConfigContainer.DryRun(INativeLibrary& loadedLLamaNativeLibrary, INativeLibrary& loadedLLavaNativeLibrary)

Namely in the code:

        NativeLibraryConfig
            .All
            .WithCuda()
            .WithAutoFallback(false)
            .DryRun(out INativeLibrary? _, out INativeLibrary? _)

The project is running in Docker, variables are used:

GGML_CUDA_DISABLE_GRAPHS: "1"
GGML_CUDA_FORCE_MMQ: "1"
CUDA_LAUNCH_BLOCKING: "1"

UPDATE
I found the following line in the logs:

Failed Loading 'runtimes/linux-x64/native/cuda12/libllava_shared.so'

After this application is shutting down.

Reproduction Steps

I tried adding WithAutoFallback(true)

Environment & Configuration

• Operating system: Linux
• .NET runtime version: 8
• LLamaSharp version: 0.25.0
• CUDA version (if you are using cuda backend): 12
• CPU & GPU device: A100

Known Workarounds

No response

[SignalRT]

@rockofme1 ,

Llava libraries was removed from llama.cpp

This PR #1261 should allow to use the supported multimodal models in LlamaSharp