Our code is built based on lit-gpt, please refer to original repo to build the conda environment.

• Model code: lit_gpt/robust_ger.py;
• Training script: finetune.sh;
• Inference script: infer.sh;

To run the training or inference script, you need to enter the scripts and modify the absolute paths of data, model, and experiment directory. Then, directly run the .sh script using bash command.

• For LLMs, please refer to tutorial for details, which support many mainstream LLMs like LLaMA-2;
• For trained adapter weights, please refer to our HuggingFace repo.

We have released our Robust HyPoradise dataset at HuggingFace.

@article{hu2024large,
  title={Large Language Models are Efficient Learners of Noise-Robust Speech Recognition},
  author={Hu, Yuchen and Chen, Chen and Yang, Chao-Han Huck and Li, Ruizhe and Zhang, Chao and Chen, Pin-Yu and Chng, EnSiong},
  journal={arXiv preprint arXiv:2401.10446},
  year={2024}
}