paper:Improving Text Semantic Similarity Modeling through a 3D Siamese Network
- python3.10
- tensorflow2.10
- cuda11.6
- transformers4.17
- RTX3090
python train.py
Process your own dataset into a csv file with three fields for header sentence1, sentence2, label, as shown below:
Split the data into train.csv, dev.csv, test.csv and put them in . /inputs/dataset folder.