MusicLM

Musiclm tool is a model for creating high loyalty music from text descriptions. The model is a hierarchical task of modeling a sequence to a sequence that creates music at 24 kHz, which remains consistent in minutes. The tool can be conditioned by created music on both the lyrics and the tune, allowing it to convert whistling and hoarse tunes in accordance with the style described in the text inscription. The tool can also be created by music from painting descriptions, instruments, genres, levels of musical experience, places and epoh. In addition, the tool can create different versions of the same text call and the same semantic tokens. Discover more about MusicLM Tool