

To support future research, we publicly release MusicCaps, a dataset composed of 5.5k music-text pairs, with rich text descriptions provided by human experts. Moreover, we demonstrate that MusicLM can be conditioned on both text and a melody in that it can transform whistled and hummed melodies according to the style described in a text caption. Our experiments show that MusicLM outperforms previous systems both in audio quality and adherence to the text description. MusicLM casts the process of conditional music generation as a hierarchical sequence-to-sequence modeling task, and it generates music at 24 kHz that remains consistent over several minutes. Download this game from Microsoft Store for Windows 10, Windows 10 Mobile, Windows 10 Team (Surface Hub), HoloLens. Download and installation of this PC software is free and 1.0 is the latest version last time we checked. We introduce MusicLM, a model generating high-fidelity music from text descriptions such as "a calming violin melody backed by a distorted guitar riff". Virtual Piano is provided under a freeware license on Windows from instrument software with no restrictions on usage.


Denk, Zalán Borsos, Jesse Engel, Mauro Verzetti, Antoine Caillon, Qingqing Huang, Aren Jansen, Adam Roberts, Marco Tagliasacchi, Matt Sharifi, Neil Zeghidour, Christian Frank
