Skip to content
conftrace_

edit-invariant loss

1 paper

Explore in graph

Co-occurring keywords · papers mentioning both

noise-robust learning (1) noise-robust training (1) sequence loss (1) non-autoregressive generation (1) cross-entropy loss (1) edit distance (1) sequence model (1) neural text generation (1) text generation (1) knowledge distillation (1)

Papers

Don’t Take It Literally: An Edit-Invariant Sequence Loss for Text Generation NAACL 2022