{
  "architectures": [
    "S3GenModel"
  ],
  "cfm_inference_cfg_rate": 0.7,
  "cfm_sigma_min": 1e-06,
  "cfm_solver": "euler",
  "cfm_t_scheduler": "cosine",
  "decoder_act_fn": "gelu",
  "decoder_attention_head_dim": 64,
  "decoder_channels": [
    256
  ],
  "decoder_in_channels": 320,
  "decoder_n_blocks": 4,
  "decoder_num_heads": 8,
  "decoder_num_mid_blocks": 12,
  "decoder_out_channels": 80,
  "dtype": "float32",
  "encoder_attention_heads": 8,
  "encoder_dropout_rate": 0.1,
  "encoder_linear_units": 2048,
  "encoder_num_blocks": 6,
  "encoder_output_size": 512,
  "fmax": 8000,
  "fmin": 0,
  "hop_length": 480,
  "input_frame_rate": 25,
  "mel_bins": 80,
  "model_type": "s3gen",
  "n_fft": 1920,
  "pre_lookahead_len": 3,
  "sampling_rate": 24000,
  "speaker_embed_dim": 192,
  "speaker_feat_dim": 80,
  "token_embed_dim": 512,
  "token_mel_ratio": 2,
  "transformers_version": "5.0.0.dev0",
  "vocab_size": 6561,
  "win_size": 1920
}