Distribute and run LLMs with a single file.
C++ cross-platformggufllama-cpplocal-ailocal-inferencelocal-llmopen-source-aisingle-file-executablespeech-to-text
Update README.md -GPU windows (#186)
ThePersonOfFuna193fd559fcc
diff --git a/README.md b/README.mdindex 0c0574d9..54af00fc 100644--- a/README.md+++ b/README.md@@ -344,7 +344,7 @@ binaries, and (2) you pass the `-ngl 35` flag. You also need an NVIDIAgraphics card that supports CUDA. There's no support yet for AMD GPUs.You can also use CUDA via WSL by enabling [Nvidia CUDA on WSL](https://learn.microsoft.com/en-us/windows/ai/directml/gpu-cuda-in-wsl)and running your llamafiles inside of WSL. This will also allow you to use-llamafiles greater than 4GB on Windows.+llamafiles greater than 4GB on Windows. NOTE: If you have two different GPUs(integrated AMD, and dedicated NVIDIA like a laptop) then you need to add -ngl 35 --gpu NVIDIAOn Linux, Nvidia cuBLAS GPU support will be compiled on the fly if (1)you have the `cc` compiler installed, (2) you pass the `-ngl 35` flag to