Text Generation On Yahoo Questions
評価指標
KL
NLL
Perplexity
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | KL | NLL | Perplexity | Paper Title | Repository |
|---|---|---|---|---|---|
| SA-VAE | 7.19 | 327.5 | 60.4 | Semi-Amortized Variational Autoencoders | |
| CNN-VAE | 10.0 | 332.1 | 63.9 | Improved Variational Autoencoders for Text Modeling using Dilated Convolutions | |
| Aggressive VAE | 5.7 | 326.7 | 59.7 | Lagging Inference Networks and Posterior Collapse in Variational Autoencoders |
0 of 3 row(s) selected.