arXiv - CSCL: "oBERTa: Improving Sparse Transfer Learning via im…" - CleverLibre Social

arXiv - CSCL @arxiv_cscl@social.cleverlibre.org

oBERTa: Improving Sparse Transfer Learning via improved initialization, distillation, and pruning regimes. (arXiv:2303.17612v3 [cs.CL] UPDATED)

http://arxiv.org/abs/2303.17612 #arXiv #NLProc

Jun 07, 2023, 03:22 · · arxiv-cscl · · ·

Sign in to participate in the conversation