Machine Translation on IWSLT2015 German-English

Metric: BLEU score (higher is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	BLEU score▼	Extra Data	Paper	Date↕	Code
1	PS-KD	36.2	No	Self-Knowledge Distillation with Progressive Ref...	2020-06-22	Code
2	Pervasive Attention	34.18	No	Pervasive Attention: 2D Convolutional Neural Net...	2018-08-11	Code
3	Transformer with FRAGE	33.97	No	FRAGE: Frequency-Agnostic Word Representation	2018-09-18	Code
4	ConvS2S+Risk	32.93	No	Classical Structured Prediction Losses for Seque...	2017-11-14	Code
5	Denoising autoencoders (non-autoregressive)	32.43	No	Deterministic Non-Autoregressive Neural Sequence...	2018-02-19	Code
6	ConvS2S	32.31	No	Convolutional Sequence to Sequence Learning	2017-05-08	Code
7	Conv-LSTM (deep+pos)	30.4	No	A Convolutional Encoder Model for Neural Machine...	2016-11-07	Code
8	NPMT + language model	30.08	No	Towards Neural Phrase-based Machine Translation	2017-06-17	Code
9	RNNsearch	29.98	No	An Actor-Critic Algorithm for Sequence Prediction	2016-07-24	Code
10	DCCL	29.56	No	Compressing Word Embeddings via Deep Composition...	2017-11-03	Code
11	Bi-GRU (MLE+SLE)	28.53	No	Neural Machine Translation by Jointly Learning t...	2014-09-01	Code
12	FlowSeq-base	24.75	No	FlowSeq: Non-Autoregressive Conditional Sequence...	2019-09-05	Code
13	Word-level CNN w/attn, input feeding	24	No	Sequence-to-Sequence Learning as Beam-Search Opt...	2016-06-09	Code
14	Word-level LSTM w/attn	20.2	No	Sequence Level Training with Recurrent Neural Ne...	2015-11-20	Code
15	QRNN	19.41	No	Quasi-Recurrent Neural Networks	2016-11-05	Code

#1PS-KDSOTA
36.2
BLEU score· 2020-06-22
Self-Knowledge Distillation with Progressive Refinement of Targets Code
#2Pervasive AttentionSOTA
34.18
BLEU score· 2018-08-11
Pervasive Attention: 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction Code
#3Transformer with FRAGE
33.97
BLEU score· 2018-09-18
FRAGE: Frequency-Agnostic Word Representation Code
#4ConvS2S+RiskSOTA
32.93
BLEU score· 2017-11-14
Classical Structured Prediction Losses for Sequence to Sequence Learning Code
#5Denoising autoencoders (non-autoregressive)
32.43
BLEU score· 2018-02-19
Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinement Code
#6ConvS2SSOTA
32.31
BLEU score· 2017-05-08
Convolutional Sequence to Sequence Learning Code
#7Conv-LSTM (deep+pos)SOTA
30.4
BLEU score· 2016-11-07
A Convolutional Encoder Model for Neural Machine Translation Code
#8NPMT + language model
30.08
BLEU score· 2017-06-17
Towards Neural Phrase-based Machine Translation Code
#9RNNsearchSOTA
29.98
BLEU score· 2016-07-24
An Actor-Critic Algorithm for Sequence Prediction Code
#10DCCL
29.56
BLEU score· 2017-11-03
Compressing Word Embeddings via Deep Compositional Code Learning Code
#11Bi-GRU (MLE+SLE)SOTA
28.53
BLEU score· 2014-09-01
Neural Machine Translation by Jointly Learning to Align and Translate Code
#12FlowSeq-base
24.75
BLEU score· 2019-09-05
FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow Code
#13Word-level CNN w/attn, input feeding
24
BLEU score· 2016-06-09
Sequence-to-Sequence Learning as Beam-Search Optimization Code
#14Word-level LSTM w/attn
20.2
BLEU score· 2015-11-20
Sequence Level Training with Recurrent Neural Networks Code
#15QRNN
19.41
BLEU score· 2016-11-05
Quasi-Recurrent Neural Networks Code