config.json

{
  "_name_or_path": "./cache/model",
  "architectures": [
    "Pop2PianoForConditionalGeneration"
  ],
  "composer_vocab_size": 21,
  "d_ff": 2048,
  "d_kv": 64,
  "d_model": 512,
  "dataset_input_length": 1024,
  "dataset_mel_is_conditioned": true,
  "dataset_n_bars": 2,
  "dataset_sampling_rate": 22050,
  "dataset_target_length": 256,
  "dataset_use_mel": true,
  "decoder_start_token_id": 0,
  "dense_act_fn": "relu",
  "dropout_rate": 0.1,
  "eos_token_id": 1,
  "f_min": 10.0,
  "feed_forward_proj": "gated-gelu",
  "hidden_size": 512,
  "hop_length": 1024,
  "initializer_factor": 1.0,
  "is_encoder_decoder": true,
  "is_gated_act": true,
  "layer_norm_epsilon": 1e-06,
  "model_type": "pop2piano",
  "n_fft": 4096,
  "n_mels": 512,
  "n_positions": 1024,
  "num_attention_heads": 8,
  "num_decoder_layers": 6,
  "num_heads": 8,
  "num_hidden_layers": 6,
  "num_layers": 6,
  "output_past": true,
  "pad_token_id": 0,
  "relative_attention_max_distance": 128,
  "relative_attention_num_buckets": 32,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.40.1",
  "use_cache": true,
  "vocab_size": 2400
}