Datasets:

Modalities:
Audio
Text
Formats:
parquet
ArXiv:
License:

Nepali Split

#11
by shreeyad - opened

Hi I downloaded Nepali split but I only see 467.37 hours of data in the train split with 1076 speakers for Nepali. According to the paper it is supposed to be 100h ans 496 speakers. Is it because the paper is outdated?

AI4Bharat org

Hi yes.
This was an ongoing effort and we were regularly updating the dataset with newly collected data.

safikhan changed discussion status to closed

Sign up or log in to comment