PaLM 2 Technical Reportwith Rohan Anil, Andrew M. Dai, Orhan Firat, Melvin Johnson +1231 versionW_cag8ztjxv1 · current
SPIRe: Boosting LLM Inference Throughput with Speculative DecodingSanjit Neelam, Daniel Heinlein, Vaclav Cvicek et al.0 citations2504.06419v12025arXiv
PaLM 2 Technical ReportRohan Anil, Andrew M. Dai, Orhan Firat et al.158 citations2305.10403v32023arXivon Valency
Efficiently Scaling Transformer InferenceReiner Pope, Sholto Douglas, Aakanksha Chowdhery et al.113 citations2211.05102v12022arXiv
PaLM: Scaling Language Modeling with PathwaysAakanksha Chowdhery, Sharan Narang, Jacob Devlin et al.2,555 citations2204.02311v52022arXiv