datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
TangoPromptBank
Project Links
Github
Web
Huggingface Space
Dataset Description
This dataset was used to Pre-train Tango-Full-FT-Audiocaps. TangoPromptBank is a diverse corpus consisting of textual prompts and audio samples sourced from WavCaps [1], AudioCaps [9], ESC [2], UrbanSound [3], MusicCaps [4], GTZAN [5], and Musical Instruments [6] dataset. The dataset statistics are reported in Table 1. All audio clips longer than 10 seconds were segmented into partitions of successive 10… See the full description on the dataset page: https://huggingface.co/datasets/declare-lab/TangoPromptBank.Nsynth_Test_Split_Tango_Format