hparams.yaml

################################
# Experiment Parameters        #
################################
epochs: 4
iters_per_checkpoint: 64
seed: 1234
dynamic_loss_scaling: True
cudnn_enabled: True
cudnn_benchmark: False
ignore_layers: ['embedding.weight']

################################
# Data Parameters             #
################################
load_mel_from_disk: False
training_files: 'filelists/train.txt'
validation_files: 'filelists/validation.txt'
text_cleaners: ['english_cleaners']

################################
# Audio Parameters             #
################################
max_wav_value: 32768.0
sampling_rate: 22050
filter_length: 1024
hop_length: 256
win_length: 1024
n_mel_channels: 80
mel_fmin: 0.0
mel_fmax: 8000.0

################################
# Model Parameters             #
################################
n_symbols: 148
symbols_embedding_dim: 512

# Encoder parameters
encoder_kernel_size: 5
encoder_n_convolutions: 3
encoder_embedding_dim: 512

# Decoder parameters
n_frames_per_step: 1  # currently only 1 is supported
decoder_rnn_dim: 1024
prenet_dim: 256
max_decoder_steps: 1000
gate_threshold: 0.5
p_attention_dropout: 0.1
p_decoder_dropout: 0.1

# Attention parameters
attention_rnn_dim: 1024
attention_dim: 128

# Location Layer parameters
attention_location_n_filters: 32
attention_location_kernel_size: 31

# Mel-post processing network parameters
postnet_embedding_dim: 512
postnet_kernel_size: 5
postnet_n_convolutions: 5

################################
# Optimization Hyperparameters #
################################
use_saved_learning_rate: False
learning_rate: 0.001
weight_decay: 0.000001
grad_clip_thresh: 1.0
batch_size: 16
mask_padding: True  # set model's padded outputs to padded values