Any problem in improving the model's performance on Mandarin Chinese?
Opened this issue · 0 comments
2502128021 commented
Why it is late to improve the model's performance on Mandarin Chinese?
Is that because wav2vec can not embed Chinese speech well?