Due to the complex rhyme structure of rap, standard rhyming models are not suitable for rap generation. Due to the lack of datasets with rap beat-lyric alignment, no rhythmic modeling method for rap has been created before.
A recent study on arXiv.org proposes a transformer-based rap generation model for both rhymes and rhythms.
Firstly, a data mining pipeline is developed to create rap datasets with aligned rhythmic beats. In order to generate rap lyrics with rhyme constraint, an autoregressive language model is created. Beat information is modeled by inserting a beat token besides the corresponding word.
The model is pre-trained using non-rap songs with aligned beats and pure lyrics. Then, it is fine-tuned on the rap songs with aligned beats. Objective and subjective evaluations confirm that the model generates high-quality raps with good rhymes and rhythms.
Rap generation, which aims to produce lyrics and corresponding singing beats, needs to model both rhymes and rhythms. Previous works for rap generation focused on rhyming lyrics but ignored rhythmic beats, which are important for rap performance. In this paper, we develop DeepRapper, a Transformer-based rap generation system that can model both rhymes and rhythms. Since there is no available rap dataset with rhythmic beats, we develop a data mining pipeline to collect a large-scale rap dataset, which includes a large number of rap songs with aligned lyrics and rhythmic beats. Second, we design a Transformer-based autoregressive language model which carefully models rhymes and rhythms. Specifically, we generate lyrics in the reverse order with rhyme representation and constraint for rhyme enhancement and insert a beat symbol into lyrics for rhythm/beat modeling. To our knowledge, DeepRapper is the first system to generate rap with both rhymes and rhythms. Both objective and subjective evaluations demonstrate that DeepRapper generates creative and high-quality raps with rhymes and rhythms. Code will be released on GitHub.
Research paper: Xue, L., “DeepRapper: Neural Rap Generation with Rhyme and Rhythm Modeling”, 2021. Link: https://arxiv.org/abs/2107.01875
Discover more from Today Headline
Subscribe to get the latest posts to your email.