single source · model architecture
Causal audio encoder with sliding-window attention and short convolutions
Combines sliding-window attention with short convolutions in a fully causal audio encoder.
The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See multimodal architecture for the methods filed beside it.
Also called Causal audio encoding, fully causal.
Filed under model architecture › multimodal architecture