English

ESPnet-EZ: Python-only ESPnet for Easy Fine-tuning and Integration

Sound 2024-09-17 v1 Artificial Intelligence Audio and Speech Processing

Abstract

We introduce ESPnet-EZ, an extension of the open-source speech processing toolkit ESPnet, aimed at quick and easy development of speech models. ESPnet-EZ focuses on two major aspects: (i) easy fine-tuning and inference of existing ESPnet models on various tasks and (ii) easy integration with popular deep neural network frameworks such as PyTorch-Lightning, Hugging Face transformers and datasets, and Lhotse. By replacing ESPnet design choices inherited from Kaldi with a Python-only, Bash-free interface, we dramatically reduce the effort required to build, debug, and use a new model. For example, to fine-tune a speech foundation model, ESPnet-EZ, compared to ESPnet, reduces the number of newly written code by 2.7x and the amount of dependent code by 6.7x while dramatically reducing the Bash script dependencies. The codebase of ESPnet-EZ is publicly available.

Keywords

Cite

@article{arxiv.2409.09506,
  title  = {ESPnet-EZ: Python-only ESPnet for Easy Fine-tuning and Integration},
  author = {Masao Someki and Kwanghee Choi and Siddhant Arora and William Chen and Samuele Cornell and Jionghao Han and Yifan Peng and Jiatong Shi and Vaibhav Srivastav and Shinji Watanabe},
  journal= {arXiv preprint arXiv:2409.09506},
  year   = {2024}
}

Comments

Accepted to SLT 2024

R2 v1 2026-06-28T18:44:50.148Z