4 ms·
WebGPU in Chrome Canary now supports float16 for faster local LLM inference
- NiekvdMaas 3y agoTo test, run Chrome Canary with: --enable-dawn-features=allow_unsafe_apis,disable_robustness Then, open: https://webllm.mlc.ai/#chat-demo https://webllm.mlc.ai/#chat-demo
- hardware2win 3y ago>disable_robustness What a name, especially since it improves perf
- no_time 3y agoI hope they gate webGPU behind a per domain allow/deny notification like they do with location and camera access. This whole thing is just begging to be exploited by cryptominers.
- charcircuit 3y agoIt may be more profitable to use the compute for LLMs or other GPU dependent AI than crypto mining. The downside is that there is no way to verify the client isn't returning random or bad data.
- lostmsu 3y agoCryptomining is mostly gone.