Accessibility: sharing the processed audio segments and transcriptions
Hey 👋 . Thanks for releasing this dataset.
I worry some users might struggle and be put off by using the url_list + EmiliaPipe.
EmiliaPipe is a handy package, but I think if the dataset format was more accessible it could get more usage. For example, do you think it would be possible to release the row-wise audio segments and transcriptions? So that the rows contain references to audio files of a segment and their transcription. Similarly to this example: https://huggingface.co/datasets/mozilla-foundation/common_voice_11_0.
Thank you so much for your attention to our work! We are preparing for uploading a processed dataset with permitted license in the following months! We will keep you posted!
Thank you so much for your attention to our work! We are preparing for uploading a processed dataset with permitted license in the following months! We will keep you posted!
It's great！ Looking forward to the progress of your dataset, thank you very much.