Fetching the paper…

Self-Distillation Mixup Training for Non-autoregressive Neural Machine Translation · Around