arXiv - CSCL: "CodeBPE: Investigating Subtokenization Options fo…" - CleverLibre Social

arXiv - CSCL @arxiv_cscl@social.cleverlibre.org

CodeBPE: Investigating Subtokenization Options for Large Language Model Pretraining on Source Code. (arXiv:2308.00683v1 [cs.LG])

http://arxiv.org/abs/2308.00683 #arXiv #NLProc

Aug 02, 2023, 03:17 · · arxiv-cscl · · ·

Sign in to participate in the conversation