Шығын ландшафты
otherentity
ID: losslandscape
Related Papers
Masking as an Efficient Alternative to Finetuning for Pretrained Language Models
AbstractWe present an efficient method of utilizing pretrained language models, where we learn selective binary masks for pretrained weights in lieu of modifying them through finetuning. Extensive evaluations of masking …...
On the Complementarity between Pre-Training and Random-Initialization for Resource-Rich Machine Translation
AbstractPre-Training (PT) of text representations has been successfully applied to low-resource Neural Machine Translation (NMT). However, it usually fails to achieve notable gains (some- times, even worse) on resource-rich NMT …...
On the Interaction of Regularization Factors in Low-resource Neural Machine Translation
AbstractWe explore the roles and interactions of the hyper-parameters governing regularization, and propose a range of values applicable to low-resource neural machine translation. We demonstrate that default or recommended values …...