News for hackers. Stuff that matters.
Re:Anthropic’s $5B, 4-year plan to take on OpenAI by anentropic on Wednesday April 12, 2023 @05:40PM (#35547373)
I think the GPU performance for inference is probably limited currently by immaturity of PyTorch MPS (Metal) backend

before I found the repo above I had a naive attempt to get llama running with mps and it didn't "just work" - bunch of ops not supported etc

There are no comments on this comment yet.