4 ms·
HuggingFace/Kernels: 200 WebGPU Kernels for Local AI
- verdverm 1mo agoI'm not sure why I want WebGPU kernels for running local models. I can see why people who want to spend cycles on my hardware would like to deliver and run models in the browser