All Stories


This story seems to be about:

Abstract We introduce MusicLM, a model generating high-fidelity music from text descriptions such as "a calming violin melody backed by a distorted guitar riff". 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. Our experiments show that MusicLM outperforms previous systems both in audio quality and adherence to the text description. 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. 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.

'We got cheated': LA Lakers furious after missed foul in loss to Boston Celtics


LeBron James was left hunched on the court in frustration after referees missed a foul on his attempted game-winning layup, and the Los Angeles Lakers succumbed to a 125-121 overtime loss against the Boston Celtics.

Biden's approval drops as he takes the spotlight from Trump


President Joe Biden's political standing has benefited greatly over the past couple of years from his predecessor Donald Trump's inability to cede the limelight. It's part of the reason that Biden's Democratic Party had one of the best midterms last fall for a party that controlled the White House, despite the president's low approval rating.