Natural Language Generation on TV RNNLG benchmark (test)
0.5287BLEU ScoreCNN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CNN2018.05 | 0.5287 | 1.87 | |
| SCLSTM2018.05 | 0.5265 | 2.31 | |
| Deep Ensemble Modeltype=ensemble2018.05 | 0.5226 | 1.67 | |
| LSTM2018.05 | 0.5012 | 3.86 |
Loading the SOTA2 catalog…