The authors of Multiple-hypothesis RNN-T Loss for Unsupervised Fine-tuning and Self-training of Neural Transducer have not publicly listed the code yet.

Request code directly from the authors: Ask Authors for Code

Get an expert to implement this paper: Request Implementation

(OR if you have code to share with the community, please add it here 😊🙏 )