Sequoia: Speculative decoding boosting LLM inference by 8-10x(infini-ai-lab.github.io)3 points by fgfm 2 years ago | 0 commentsNo comments yet