Abstract page for arXiv paper 2211.17192: Fast Inference from Transformers via Speculative Decoding
Press ? anytime to show this help