arXiv - CSCL: "Vcc: Scaling Transformers to 128K Tokens or More …" - CleverLibre Social

arXiv - CSCL @arxiv_cscl@social.cleverlibre.org

Vcc: Scaling Transformers to 128K Tokens or More by Prioritizing Important Tokens. (arXiv:2305.04241v2 [cs.CL] UPDATED)

http://arxiv.org/abs/2305.04241 #arXiv #NLProc

May 30, 2023, 03:06 · · arxiv-cscl · · ·

Sign in to participate in the conversation