A Pytorch implementation of paper An Efficient and Robust Semantic Hashing Framework for Similar Text Search (TOIS2023). Our code refers to VDSH (https://github.com/bayesquant/VDSH)
- pytorch 1.13.0
- numpy 1.22.3
- pandas 1.5.2
- faiss-cpu 1.7.3
We have provided the processed dataset for ng20 in dataset/ng20/. You can refer to (https://github.com/bayesquant/VDSH) for more datasets.
Run following command to run the code in ng20 dataset:
sh scripts/run.sh
If you find our code useful and use SMASH in your work, please cite our paper.