paper:dual_learning_for_machine_translation
Differences
This shows you the differences between two versions of the page.
| Next revision | Previous revision | ||
| paper:dual_learning_for_machine_translation [2020/11/19 02:18] – Create paper summary Dual Learning For Machine Translation chrisliu298 | paper:dual_learning_for_machine_translation [2023/06/15 07:36] (current) – external edit 127.0.0.1 | ||
|---|---|---|---|
| Line 5: | Line 5: | ||
| **TLDR;** The authors cast machine translation as a game of a primal task and a dual task, which allows two agents to teach each other in a reinforcement learning fashion (i.e., policy gradient method). With this implementation, | **TLDR;** The authors cast machine translation as a game of a primal task and a dual task, which allows two agents to teach each other in a reinforcement learning fashion (i.e., policy gradient method). With this implementation, | ||
| - | ===== Key Points | + | ==== Key Points ==== |
| * The motivation is that parallel data are scarce and costly to collect. | * The motivation is that parallel data are scarce and costly to collect. | ||
| Line 19: | Line 19: | ||
| * It maximizes towards the expected sum of the two rewards over the middle translation outputs. | * It maximizes towards the expected sum of the two rewards over the middle translation outputs. | ||
| * It uses beam search to obtain more meaningful results. | * It uses beam search to obtain more meaningful results. | ||
| - | </ | ||
paper/dual_learning_for_machine_translation.1605752284.txt.gz · Last modified: 2023/06/15 07:36 (external edit)