{
  "model_type": "audex_causal_speech_decoder",
  "architectures": [
    "AudexCausalSpeechDecoderModel"
  ],
  "auto_map": {
    "AutoConfig": "configuration_audex_causal_speech_decoder.AudexCausalSpeechDecoderConfig",
    "AutoModel": "modeling_audex_causal_speech_decoder.AudexCausalSpeechDecoderModel"
  },
  "hidden_dim": 2048,
  "depth": 12,
  "heads": 32,
  "pos_meb_dim": 64,
  "hop_length": 320,
  "vq_dim": 2048,
  "lookahead_steps": 4,
  "sample_rate": 16000,
  "codebook_levels": [4, 4, 4, 4, 4, 4, 4, 4],
  "codebook_size": 65536,
  "token_embed_dim": 8,
  "embed_tokens_from_codes": true
}