Machine Translation On Iwslt2015 German
평가 지표
BLEU score
평가 결과
이 벤치마크에서 각 모델의 성능 결과
비교 표
| 모델 이름 | BLEU score |
|---|---|
| sequence-to-sequence-learning-as-beam-search | 24.0 |
| a-convolutional-encoder-model-for-neural | 30.4 |
| quasi-recurrent-neural-networks | 19.41 |
| flowseq-non-autoregressive-conditional | 24.75 |
| deterministic-non-autoregressive-neural | 32.43 |
| sequence-level-training-with-recurrent-neural | 20.2 |
| convolutional-sequence-to-sequence-learning | 32.31 |
| frage-frequency-agnostic-word-representation | 33.97 |
| classical-structured-prediction-losses-for | 32.93 |
| an-actor-critic-algorithm-for-sequence | 29.98 |
| neural-machine-translation-by-jointly | 28.53 |
| pervasive-attention-2d-convolutional-neural-1 | 34.18 |
| towards-neural-phrase-based-machine | 30.08 |
| self-knowledge-distillation-a-simple-way-for | 36.20 |
| compressing-word-embeddings-via-deep | 29.56 |