datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
spectrogram-captionsDataset of captioned spectrograms (text describing the sound).
Amharic_Audio_and_Spectrograms
Amharic Audio Spectrogram Dataset
Dataset Info
Total samples in full dataset: 662,611
Samples in this preview: 1,000
Audio duration: 2.49 ± 1.60 seconds
Sample rate: 16kHz
Spectrogram dimensions: 80 mel bins × variable time steps
Sample Data
Audio Sample
Spectrogram
License
Apache 2.0
soundsCaps-Spectrograms_to_Base64