from datasets import load_dataset, Audio


ds_i = load_dataset("language-and-voice-lab/samromur_asr",split='train+validation+test')
ds_i = ds_i.cast_column("audio", Audio(sampling_rate=16_000))
ds_i = ds_i.data.to_pandas()

#ds_f = load_dataset("carlosdanielhernandezmena/ravnursson_asr",split='train+validation+test')
#ds_f = ds_f.cast_column("audio", Audio(sampling_rate=16_000))
#ds_f = ds_f.data.to_pandas()