This software repository hosts the self-contained implementation of the state-of-the-art models used in Machine Reading and Comprehension Task.
| Folder | Reference |
|---|---|
| watson/ | Text Understanding with the Attention Sum Reader Network, Kadlec et al., ACL 2016. |
| stanford/ | A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task, Chen et al., ACL 2016. |
| fair/ | The Goldilocks Principle: Reading Children's Books with Explicit Memory Representations, Hill et al., ICLR 2016. |
| mins per batch (mins per epoch) | watson/ | stanford/ | fair/ |
|---|---|---|---|
| GPU\Batch Size | 32 | 32 | 1 |
| K40 | 806 ms (46m 16s) |
800 ms (2h 40m) |
18ms (34m 8s) |
| Titan X | 746 ms (42m 38s) |
- | 13ms (24m 45s) |
| 1080 | 889 ms (51m 8s) |
- | 13ms (25m 29s) |
This repository would not be possible without the efforts of the maintainers of the following libraries:
- Element-Research/rnn
- MemNN
- Torch (Ofcourse!)
MIT