Question Answering On Cnn Daily Mail
評価指標
CNN
Daily Mail
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
比較表
| モデル名 | CNN | Daily Mail |
|---|---|---|
| teaching-machines-to-read-and-comprehend | 63.8 | 68.0 |
| a-thorough-examination-of-the-cnndaily-mail | 67.9 | 68.3 |
| gated-attention-readers-for-text | 77.9 | 80.9 |
| attention-over-attention-neural-networks-for | 74.4 | - |
| a-thorough-examination-of-the-cnndaily-mail | 77.6 | 79.2 |
| iterative-alternating-neural-attention-for | 76.1 | - |
| text-understanding-with-the-attention-sum | 75.4 | 77.7 |
| text-understanding-with-the-attention-sum | 69.5 | 73.9 |
| natural-language-comprehension-with-the | 74 | - |
| linguistic-knowledge-as-memory-for-recurrent | 78.6 | - |
| teaching-machines-to-read-and-comprehend | 63 | 69 |
| a-thorough-examination-of-the-cnndaily-mail | 72.4 | 75.8 |
| teaching-machines-to-read-and-comprehend | 69.4 | - |
| reasonet-learning-to-stop-reading-in-machine | 74.7 | 76.6 |
| bidirectional-attention-flow-for-machine | 76.9 | 79.6 |
| dynamic-entity-representation-with-max | 72.9 | - |