List your company
ResearchFormerly WizWand

Machine Translation (Chinese-to-English) on NIST MT-05 2005

50.72BLEUAdvAug
30.64835.85941.0746.281Jun 10, 2018Oct 11, 2018Feb 12, 2019Jun 16, 2019Oct 17, 2019Feb 18, 2020Jun 21, 2020
Updated 8mo ago

Evaluation results

MethodLinks
AdvAugLoss Config.=L_aut + L_adv, Extra corpora=true2020.0650.72
AdvAugLoss Config.=L_aut + L_adv, Extra corpora=false2020.0649.88
AdvAugLoss Config.=L_clean + L_adv, Extra corpora=false2020.0648.76
AdvAugLoss Config.=L_aut, Extra corpora=false2020.0646.88
Edunov et al.Loss Config.=not elaborated, Extra corpora=true2020.0646.81
Cheng et al.Loss Config.=not elaborated, Extra corpora=false2020.0646.58
AdvAugLoss Config.=L_mixup, Extra corpora=false2020.0646.08
Sato et al.Loss Config.=not elaborated, Extra corpora=false2020.0645.88
Sennrich et al.Loss Config.=not elaborated, Extra corpora=true2020.0645.69
Miyato et al.Loss Config.=not elaborated, Extra corpora=false2020.0645.32
Vaswani et al.Loss Config.=L_clean, Extra corpora=false2020.0644.96
RNNSEARCH-COVERAGE + D_RNN + O_CDR# Para.=+1.20M, Speed=0.7K2018.1139.39
RNNSEARCH-COVERAGE# Para.=+1.03M, Speed=1.2K2018.1138.73
SAWRsyntax-integrated=true2019.0538.27
Tree-RNNsyntax-integrated=true2019.0537.86
Tree-Linearizationsyntax-integrated=true2019.0537.64
RNNSEARCH + D_RNN + O_CDR# Para.=+0.17M, Speed=0.8K2018.1137.58
RNNSEARCH + O_CDR# Para.=+0M, Speed=1.0K2018.1136.86
Seq2Seq + SACTMechanism=Self-Adaptive Control of Temperature2018.0836.81
RNNSEARCH + MRT_BLEU# Para.=+0M, Speed=0.3K2018.1136.78
Seq2Seq+Attention+DeconvDec2018.0636.77
RNNSEARCH + O_BLEU# Para.=+0M, Speed=1.1K2018.1136.62
RNNSEARCH + O_CHRF3# Para.=+0M, Speed=1.1K2018.1136.57
Baseline2019.0536.26
RNNSEARCH + D_RNN# Para.=+0.17M, Speed=1.2K2018.1136.13
RNNSEARCH + D_CNN# Para.=+0.23M, Speed=1.0K2018.1135.97
MemDec2018.0635.91
MemDecType=Attention-based Seq2Seq with external memory2018.0835.91
RNNSEARCH# Para.=86.7M, Speed=1.4K2018.1135.88
InterAtten2018.0635.53
Li et al. (2017)2019.0535.5
Coverage2018.0634.91
CoverageType=Attention-based Seq2Seq with coverage model2018.0834.91
Chen et al. (2017a)2019.0534.35
Chen et al. (2017b)2019.0534.18
Seq2Seq+Attention2018.0633.52
Seq2Seq2018.0833.52
Lattice2018.0632.4
Moses2018.0631.47
MosesType=Phrase-based translation, Language Model=4-gram2018.0831.47
RNNSearch2018.0631.42
RNNSearchType=Attention-based Seq2Seq2018.0831.42