single mention · model architecture
Fixed 6:12 text-to-audio token interleaving
Generates six text tokens followed by twelve audio tokens repeatedly during interleaved generation.
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 6 text tokens followed by 12 audio tokens.
Filed under model architecture › multimodal architecture