Comment by rahimnathwani
1 day ago
Has anyone successfully run this on a Mac? The installation instructions appear to assume an NVIDIA GPU (CUDA, FlashAttention), and I’m not sure whether it works with PyTorch’s Metal/MPS backend.
1 day ago
Has anyone successfully run this on a Mac? The installation instructions appear to assume an NVIDIA GPU (CUDA, FlashAttention), and I’m not sure whether it works with PyTorch’s Metal/MPS backend.
FWIW you can run the demo without FlashAttention using --no-flash-attn command-line parameter, I do that since I'm on Windows and haven't gotten FlashAttention2 to work.
I recommend using modal for renting the metal.
It seems to depend on FlashAttention, so the short answer is no. Hopefully someone does the work of porting the inference code over!