processor_config.json

{
  "audio_padding": "longest",
  "audio_placeholder": "<|audio|>",
  "encoder_ds_factor": 320,
  "processor_class": "UltravoxProcessor",
  "stack_factor": 8
}