AudioLMStage

Enums · AiDotNet.Audio.Generation

Which of AudioLM’s three Transformers a training step updates.

Fields

Field Summary
AudioLMStage CoarseAcoustic Coarse acoustic modeling: the first Q′ levels given the semantic tokens.
AudioLMStage FineAcoustic Fine acoustic modeling: the remaining levels given the coarse ones, on 3-second chunks.
AudioLMStage Semantic Semantic modeling: p(z_t | z_<t).
int value__