Hacker Neus
Real-time LLM Inference on Standard GPUs: 3k tokens/s per request
(blog.kog.ai)
116 points
by NicoConstant
6 hours ago |
59 comments
