In this project, we developed a framework "MT2ST" for embedding training acceleration in Natural Language Processing.

execute !git clone https://github.com/stanfordnlp/GloVe.git to download GloVe.

To run the project, execute the training.py file.

Citation

If you find MT2ST useful or relevant to your project and research, please kindly cite our paper:

@article{liu2024mt2stadaptivemultitasksingletask,
      title={MT2ST: Adaptive Multi-Task to Single-Task Learning}, 
      author={Dong Liu and Meng Jiang},
      year={2024},
      eprint={2406.18038},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2406.18038}, 
}