search

0hq / WebGPT

Run GPT model on the browser with WebGPU. An implementation of GPT inference in less than ~1500 lines of vanilla Javascript.
https://kmeans.org
Other
3.61k stars 206 forks source link

Embeddings now on GPU #13

Closed 0hq closed 1 year ago

0hq commented 1 year ago

Instead of doing the embedding selection on the CPU via Javascript, I now do it via copyBufferToBuffer commands. This is not a massive speed-up but means that everything is now on the GPU!

vercel[bot] commented 1 year ago

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name Status Preview Comments Updated (UTC)
web-gpt ✅ Ready (Inspect) Visit Preview 💬 Add feedback Apr 22, 2023 1:13am