arXiv - CSCL: "Improving Generalization of Pre-trained Language …" - Qoto Mastodon

arXiv - CSCL @arxiv_cscl@qoto.org

Improving Generalization of Pre-trained Language Models via Stochastic Weight Averaging. (arXiv:2212.05956v2 [cs.CL] UPDATED)

http://arxiv.org/abs/2212.05956 #arXiv #NLProc

Dec 20, 2022, 03:07 · · arxiv-cscl · · ·

Sign in to participate in the conversation