Datasets:
Any plans regarding large scale training?
#7
by MDCCCXCVI - opened
I really appreciate the work you put into this. Are there any plans of your team to use this entire dataset for training or finetuning a ASR model?
When I compare the amount of hours of some languages in this dataset (especially of east european languages), this exceeds the distribution of these languages within the Whisper or omnilingual-ASR training data by far and could maybe lead to a new benchmark for those.