CSLM-SFT

💡 Model Description

CSLM-SFT is a supervised fine-tuned checkpoint of CSLM from the ACL 2026 Findings project Efficient Training for Cross-lingual Speech Language Models.

It is built on top of the CSLM-base model and further tuned with instruction-style data for speech-to-speech conversation, including monolingual and cross-lingual scenarios.

Paper: https://arxiv.org/abs/2604.11096

📖 Citation

If you use this model, please cite:

@misc{zhou2026efficienttrainingcrosslingualspeech,
      title={Efficient Training for Cross-lingual Speech Language Models}, 
      author={Yan Zhou and Qingkai Fang and Yun Hong and Yang Feng},
      year={2026},
      eprint={2604.11096},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2604.11096}, 
}

✉️ Contact

For questions, contact: zhouyan23z@ict.ac.cn