News for hackers. Stuff that matters.

A comment on WebGPU GPT Model Demo. Parent comment

Re:WebGPU GPT Model Demo by lxe on Friday April 21, 2023 @01:08AM (#35650000)
Interesting, I'm getting 100ms/token on plain old wasm with 4 threads via ggml, using a 1.7B quantized cerebras model.
[ Parent | on HN ]

There are no comments on this comment yet.