Text Generation on Yahoo Questions
Metric: KL (higher is better)
Metric: KL (higher is better)
| # | Model↕ | KL▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | CNN-VAE | 10 | No | Improved Variational Autoencoders for Text Model... | 2017-02-27 | Code |
| 2 | SA-VAE | 7.19 | No | Semi-Amortized Variational Autoencoders | 2018-02-07 | Code |
| 3 | Aggressive VAE | 5.7 | No | Lagging Inference Networks and Posterior Collaps... | 2019-01-16 | Code |