Under matched training conditions, auto-regressive models slightly outperform flow-matching on music quality and temporal control, while flow-matching offers faster inference and better inpainting flexibility.
To better isolate the runtime overheads, we discard cross attention layers (that can’t use KV-cache) and replace the AR softmax + top-p sampling with argmax sampling
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
Auto-Regressive vs Flow-Matching: a Comparative Study of Modeling Paradigms for Text-to-Music Generation
Under matched training conditions, auto-regressive models slightly outperform flow-matching on music quality and temporal control, while flow-matching offers faster inference and better inpainting flexibility.