Effective Approaches to Attention-based Neural Machine Translation

Benchmark Model Rank Results
image-guided-story-ending-generation-on-lsmdcSeq2Seq#3BLEU-1: 14.21BLEU-2: 4.56BLEU-3: 1.70BLEU-4: 0.70CIDEr: 8.69
image-guided-story-ending-generation-on-vistSeq2Seq#4BLEU-1: 13.96BLEU-2: 5.57BLEU-3: 2.94BLEU-4: 1.69CIDEr: 12.04
machine-translation-on-wmt2014-english-germanRNN Enc-Dec Att#68BLEU score: 20.9
machine-translation-on-wmt2014-english-germanReverse RNN Enc-Dec#79BLEU score: 14.0
machine-translation-on-wmt2014-english-germanRNN Enc-Dec#80BLEU score: 11.3