| allow_cache: true | |
| batch_size: 16 | |
| config: ./examples/fastspeech2/conf/fastspeech2.v1.yaml | |
| delay_f0_energy_steps: 3 | |
| dev_dir: ./dump/valid | |
| energy_stat: ./dump/stats_energy.npy | |
| eval_interval_steps: 500 | |
| f0_stat: ./dump/stats_f0.npy | |
| model_type: fastspeech2 | |
| fastspeech2_params: | |
| n_speakers: 1 | |
| encoder_hidden_size: 384 | |
| encoder_num_hidden_layers: 4 | |
| encoder_num_attention_heads: 2 | |
| encoder_attention_head_size: 192 | |
| encoder_intermediate_size: 1024 | |
| encoder_intermediate_kernel_size: 3 | |
| encoder_hidden_act: "mish" | |
| decoder_hidden_size: 384 | |
| decoder_num_hidden_layers: 4 | |
| decoder_num_attention_heads: 2 | |
| decoder_attention_head_size: 192 | |
| decoder_intermediate_size: 1024 | |
| decoder_intermediate_kernel_size: 3 | |
| decoder_hidden_act: "mish" | |
| variant_prediction_num_conv_layers: 2 | |
| variant_predictor_filter: 256 | |
| variant_predictor_kernel_size: 3 | |
| variant_predictor_dropout_rate: 0.5 | |
| num_mels: 80 | |
| hidden_dropout_prob: 0.2 | |
| attention_probs_dropout_prob: 0.1 | |
| max_position_embeddings: 2048 | |
| initializer_range: 0.02 | |
| output_attentions: False | |
| output_hidden_states: False | |
| format: npy | |
| hop_size: 256 | |
| is_shuffle: true | |
| log_interval_steps: 200 | |
| mel_length_threshold: 32 | |
| mixed_precision: true | |
| num_save_intermediate_results: 1 | |
| optimizer_params: | |
| decay_steps: 150000 | |
| end_learning_rate: 5.0e-05 | |
| initial_learning_rate: 0.001 | |
| warmup_proportion: 0.02 | |
| weight_decay: 0.001 | |
| outdir: ./examples/fastspeech2/exp/train.fastspeech2.v4 | |
| pretrained_model: null | |
| remove_short_samples: true | |
| resume: '' | |
| save_interval_steps: 5000 | |
| start_training_f0_energy_steps: 25000 | |
| train_dir: ./dump/train | |
| train_max_steps: 200000 | |
| use_norm: true | |
| verbose: 1 | |
| version: 0.6.1 | |