---
title: ESPNet2 ASR Librispeech word vs bpe tokens
emoji: 💩
colorFrom: red
colorTo: green
sdk: gradio
app_file: gradio_asr_en_libri100_word_vs_bpe.py
pinned: false
license: mit
---


- 2022-02-23
    - audio reading was not successful
        - live recording data seemed not be correctly saved or transferred to the server
        - different audio modules were tried, but failed
        - librosa.load() surprizingly worked again 
- 2022-02-28
    - three models added (bpe, char, word)
    - example files were renames for the easier use


Check out the configuration reference at https://huggingface.co/docs/hub/spaces#reference