melgan-stft/melgan_stft.v1.yaml

# This is the hyperparameter configuration file for MelGAN with Multi Resolution STFT.
# Please make sure this is adjusted for the LJSpeech dataset. If you want to
# apply to the other dataset, you might need to carefully change some parameters.
# This configuration performs 4000k iters.

###########################################################
#                FEATURE EXTRACTION SETTING               #
###########################################################
sampling_rate: 22050
hop_size: 256            # Hop size.
format: "npy"


###########################################################
#         GENERATOR NETWORK ARCHITECTURE SETTING          #
###########################################################
model_type: "melgan_generator"

melgan_generator_params:
    out_channels: 1               # Number of output channels.
    kernel_size: 7                # Kernel size of initial and final conv layers.
    filters: 512                  # Initial number of channels for conv layers.
    upsample_scales: [8, 8, 2, 2] # List of Upsampling scales.
    stack_kernel_size: 3          # Kernel size of dilated conv layers in residual stack.
    stacks: 3                     # Number of stacks in a single residual stack module.
    is_weight_norm: false

###########################################################
#       DISCRIMINATOR NETWORK ARCHITECTURE SETTING        #
###########################################################
melgan_discriminator_params:
    out_channels: 1                   # Number of output channels.
    scales: 3                         # Number of multi-scales.
    downsample_pooling: "AveragePooling1D"   # Pooling type for the input downsampling.
    downsample_pooling_params:        # Parameters of the above pooling function.
        pool_size: 4
        strides: 2
    kernel_sizes: [5, 3]              # List of kernel size.
    filters: 16                       # Number of channels of the initial conv layer.
    max_downsample_filters: 1024      # Maximum number of channels of downsampling layers.
    downsample_scales: [4, 4, 4, 4]   # List of downsampling scales.
    nonlinear_activation: "LeakyReLU" # Nonlinear activation function.
    nonlinear_activation_params:      # Parameters of nonlinear activation function.
        alpha: 0.2
    is_weight_norm: false

###########################################################
#                   STFT LOSS SETTING                     #
###########################################################
stft_loss_params:
    fft_lengths: [1024, 2048, 512]  # List of FFT size for STFT-based loss.
    frame_steps: [120, 240, 50]     # List of hop size for STFT-based loss
    frame_lengths: [600, 1200, 240] # List of window length for STFT-based loss.


###########################################################
#               ADVERSARIAL LOSS SETTING                  #
###########################################################
lambda_feat_match: 10.0
lambda_adv: 4.0

###########################################################
#                  DATA LOADER SETTING                    #
###########################################################
batch_size: 16                 # Batch size for each GPU with assuming that gradient_accumulation_steps == 1.
batch_max_steps: 8192          # Length of each audio in batch for training. Make sure dividable by hop_size.
batch_max_steps_valid: 81920   # Length of each audio for validation. Make sure dividable by hope_size.
remove_short_samples: true     # Whether to remove samples the length of which are less than batch_max_steps.
allow_cache: true              # Whether to allow cache in dataset. If true, it requires cpu memory.
is_shuffle: true               # shuffle dataset after each epoch.

###########################################################
#             OPTIMIZER & SCHEDULER SETTING               #
###########################################################
generator_optimizer_params:
    lr_fn: "PiecewiseConstantDecay"
    lr_params: 
        boundaries: [100000]       # = discriminator_train_start_steps.
        values: [0.0005, 0.0001]    # learning rate each interval.

    
discriminator_optimizer_params:
    lr_fn: "PiecewiseConstantDecay"
    lr_params: 
        boundaries: [0]             # after resume and start training discriminator, global steps is 100k, but local discriminator step is 0
        values: [0.0001, 0.0001]    # learning rate each interval.

gradient_accumulation_steps: 1
###########################################################
#                    INTERVAL SETTING                     #
###########################################################
discriminator_train_start_steps: 100000  # steps begin training discriminator
train_max_steps: 4000000                 # Number of training steps.
save_interval_steps: 20000               # Interval steps to save checkpoint.
eval_interval_steps: 5000                # Interval steps to evaluate the network.
log_interval_steps: 200                  # Interval steps to record the training log.

###########################################################
#                     OTHER SETTING                       #
###########################################################
num_save_intermediate_results: 1  # Number of batch to be saved as intermediate results.