- Prepare the dataset and pre-trained embeddings (FastText is used here) in
./data:
./data/train.json
./data/valid.json
./data/test.json
./data/crawl-300d-2M.vec
- Preprocess the data
cd src/
python make_dataset.py ../data/
- Then you can train a poooor example model as follow:
python train.py ../models/example/
- To predict, run
python predict.py ../models/example/ --epoch 3
where --epoch specifies the save model of which epoch to use.