Text Generation On Coco Captions
평가 지표
BLEU-2
BLEU-3
BLEU-4
BLEU-5
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | BLEU-2 | BLEU-3 | BLEU-4 | BLEU-5 | Paper Title | Repository |
|---|---|---|---|---|---|---|
| LeakGAN | 0.950 | 0.880 | 0.778 | 0.686 | Long Text Generation via Adversarial Training with Leaked Information | |
| SeqGAN | 0.831 | 0.642 | 0.521 | 0.427 | SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient | |
| partGAN | 0.910 | 0.713 | O.753 | 0.590 | Long Text Generation via Adversarial Training with Leaked Information | |
| RelGAN (100) | 0.849 | 0.687 | 0.502 | - | RelGAN: Relational Generative Adversarial Networks for Text Generation | |
| RankGAN | 0.850 | 0.672 | 0.557 | 0.544 | Adversarial Ranking for Language Generation |
0 of 5 row(s) selected.