- A Tensorflow implementation of R-NET: MACHINE READING COMPREHENSION WITH SELF-MATCHING NETWORKS. This project is specially designed for the SQuAD dataset.
- Should you have any question, please contact Wenxuan Zhou ([email protected]).
There have been a lot of known problems caused by using different software versions. Please check your versions before opening issues or emailing me.
- Python >= 3.4
- unzip, wget
- tensorflow-gpu >= 1.5.0
- spaCy >= 2.0.0
- tqdm
- ujson
To download and preprocess the data, run
# download SQuAD and Glove
sh download.sh
# preprocess the data
python config.py --mode preproHyper parameters are stored in config.py. To debug/train/test the model, run
python config.py --mode debug/train/testTo get the official score, run
python evaluate-v1.1.py ~/data/squad/dev-v1.1.json log/answer/answer.jsonThe default directory for tensorboard log file is log/event
See release for trained model.
- The original paper uses additive attention, which consumes lots of memory. This project adopts scaled multiplicative attention presented in Attention Is All You Need.
- This project adopts variational dropout presented in A Theoretically Grounded Application of Dropout in Recurrent Neural Networks.
- To solve the degradation problem in stacked RNN, outputs of each layer are concatenated to produce the final output.
- When the loss on dev set increases in a certain period, the learning rate is halved.
- During prediction, the project adopts search method presented in Machine Comprehension Using Match-LSTM and Answer Pointer.
- To address efficiency issue, this implementation uses bucketing method (contributed by xiongyifan) and CudnnGRU. The bucketing method can speedup training, but will lower the F1 score by 0.3%.
| EM | F1 | |
|---|---|---|
| original paper | 71.1 | 79.5 |
| this project | 71.07 | 79.51 |
| Native | Native + Bucket | Cudnn | Cudnn + Bucket | |
|---|---|---|---|---|
| E5-2640 | 6.21 | 3.56 | - | - |
| TITAN X | 2.56 | 1.31 | 0.41 | 0.28 |
These settings may increase the score but not used in the model by default. You can turn these settings on in config.py.
- Pretrained GloVe character embedding. Contributed by yanghanxy.
- Fasttext Embedding. Contributed by xiongyifan. May increase the F1 by 1% (reported by xiongyifan).