{
  "model_type": "F5-TTS",
  "architecture": "DiT",
  "dim": 1024,
  "depth": 22,
  "heads": 16,
  "ff_mult": 2,
  "text_dim": 512,
  "conv_layers": 4,
  "mel_spec_type": "vocos",
  "target_sample_rate": 24000,
  "hop_length": 256,
  "vocab_size": 2626,
  "tokenizer": "custom",
  "vocab_file": "vocab.txt"
}