This is the code for the paper "Adversarial Mixing Policy for Relaxing Locally Linear Constraints in Mixup" accepted at EMNLP'21
pip install -r requeriments.txt
Download link and unzip all the datasets into data fold.
Create fold bert-base-uncased and enter the fold. Download
pytorch_model.bin
config.json
vocab.txt
Enter the project root directory. Download GloVe embeddings glove.840B.300d.zip from link
Detailed descriptions of arguments are provided in run_main.py
. For run the default parameters,
python run_main.py
MIT License