PPOGAN

Reported on 6 benchmarks across 2 tasks · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Adversarial5 results

Text GenerationonEMNLP2017 WMT
NLLgen· 2020-06-12
2.265
SOTA
Improving GAN Training with Probability Ratio Clipping and Sample Reweighting arXiv:2006.06900
Text GenerationonEMNLP2017 WMT
BLEU-2· 2020-06-12
0.905
best: 0.956 (LeakGAN)
Improving GAN Training with Probability Ratio Clipping and Sample Reweighting arXiv:2006.06900
Text GenerationonEMNLP2017 WMT
BLEU-3· 2020-06-12
0.692
best: 0.819 (LeakGAN)
Improving GAN Training with Probability Ratio Clipping and Sample Reweighting arXiv:2006.06900
Text GenerationonEMNLP2017 WMT
BLEU-4· 2020-06-12
0.47
best: 0.627 (LeakGAN)
Improving GAN Training with Probability Ratio Clipping and Sample Reweighting arXiv:2006.06900
Text GenerationonEMNLP2017 WMT
BLEU-5· 2020-06-12
0.322
best: 0.498 (LeakGAN)
Improving GAN Training with Probability Ratio Clipping and Sample Reweighting arXiv:2006.06900

Medical1 result

Image GenerationonCIFAR-10
FID· 2020-06-12
10.7
best: 1.22 (GMem)
Improving GAN Training with Probability Ratio Clipping and Sample Reweighting arXiv:2006.06900