过程详见blog：使用 keras 复现 MemN2N
参考论文：《End-To-End Memory Networks》
数据集：bAbI-tasks
最终源码：keras-MemN2N（GitHub）
前期工作：
BoW + Adjacent 效果很差，task-1 只能到 0.66+，但是 task-20 能到 1（数据集 1k 还是 10k 都差不多，hop 取 3 和取 1 也差不多）
forked from Rocuku/keras-MemN2N
-
Notifications
You must be signed in to change notification settings - Fork 0
arnavdas88/keras-MemN2N
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
The implementation of End-to-end memory networks in keras
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published
Languages
- Jupyter Notebook 66.6%
- Python 33.4%