{
  "architectures": [
    "IndicCanaryForConditionalGeneration"
  ],
  "auto_map": {
    "AutoConfig": "configuration_indic_canary.IndicCanaryConfig",
    "AutoModel": "modeling_indic_canary.IndicCanaryForConditionalGeneration",
    "AutoModelForSpeechSeq2Seq": "modeling_indic_canary.IndicCanaryForConditionalGeneration"
  },
  "bos_token_id": 4,
  "conv_kernel_size": 9,
  "d_model": 1024,
  "decoder_attention_heads": 8,
  "decoder_ffn_dim": 4096,
  "decoder_layers": 24,
  "decoder_start_token_id": 7,
  "encoder_attention_heads": 8,
  "encoder_ffn_dim": 4096,
  "encoder_layers": 32,
  "eos_token_id": 3,
  "is_encoder_decoder": true,
  "max_generation_delta": 50,
  "max_target_positions": 1024,
  "model_type": "indic_canary",
  "num_mel_bins": 128,
  "pad_token_id": 2,
  "subsampling_conv_channels": 256,
  "subsampling_factor": 8,
  "tie_word_embeddings": true,
  "torch_dtype": "float32",
  "vocab_size": 7152
}