metadata
language: id
datasets:
- common_voice_id_6.1
tags:
- audio
- automatic-speech-recognition
- speech
- bahasa-indonesia
license: apache-2.0
Dataset used for training:
- Name: Common Voice
- Language: Indonesian [id]
- Version: 6.1
Test WER: 19.3 %
Repo for training:
https://github.com/bagustris/wav2vec2-indonesian
NEWEST VERSION AVAILABLE HERE WITH SMALLER MODEL AND SMALLER WER (5.9%):
https://huggingface.co/Bagus/whisper-small-id-cv17
Contact:
[email protected]