Update hparams.yaml

fbec0fa verified about 2 months ago

No virus

5.92 kB

	############################################################################
	# Model: TransformerTTS
	# Training: LJSpeech
	# Author: Kasturi Saha
	# ############################################################################


	###################################
	# Experiment Parameters and setup #
	###################################
	seed: 1986
	__set_seed: !apply:torch.manual_seed [!ref <seed>]
	output_folder: !ref ./results/transformerTTS/<seed>
	save_folder: !ref <output_folder>/save
	train_log: !ref <output_folder>/train_log.txt
	epochs: 5
	keep_checkpoint_interval: 50

	###################################
	# Progress Samples #
	###################################
	# Progress samples are used to monitor the progress
	# of an ongoing training session by outputting samples
	# of spectrograms, alignments, etc at regular intervals

	# Whether to enable progress samples
	progress_samples: False

	# The path where the samples will be stored
	progress_sample_path: !ref <output_folder>/samples
	# The interval, in epochs. For instance, if it is set to 5,
	# progress samples will be output every 5 epochs
	progress_samples_interval: 1
	# The sample size for raw batch samples saved in batch.pth
	# (useful mostly for model debugging)
	progress_batch_sample_size: 3

	#################################
	# Data files and pre-processing #
	#################################
	data_folder: !ref ./data/LJSpeech-1.1 # e.g, /localscratch/ljspeech
	preprocessed_data_folder: !ref ./data/LJSpeech-1.1/preprocessed/phone_seq # e.g, /localscratch/ljspeech
	preprocessed_melspectrogram_folder: !ref ./data/LJSpeech-1.1/preprocessed/melspectrogram # e.g, /localscratch/ljspeech

	train_json: !ref ./save/train.json
	valid_json: !ref ./save/valid.json
	test_json: !ref ./save/test.json

	splits: ["train", "valid", "test"]
	split_ratio: [70, 10, 20]

	skip_prep: False

	################################
	# Audio Parameters #
	################################
	sample_rate: 22050
	hop_length: 256
	win_length: 1024
	n_mel_channels: 80
	n_fft: 1024
	mel_fmin: 0.0
	mel_fmax: 8000.0
	mel_normalized: False
	power: 1.2
	norm: "slaney"
	mel_scale: "slaney"
	dynamic_range_compression: True

	################################
	# Optimization Hyperparameters #
	################################
	learning_rate: 0.001
	weight_decay: 0.000006
	batch_size: 8 #minimum 2
	num_workers: 0
	mask_padding: True

	train_dataloader_opts:
	batch_size: !ref <batch_size>
	drop_last: False #True #False
	num_workers: !ref <num_workers>
	collate_fn: !new:speechbrain.lobes.models.Tacotron2.TextMelCollate

	valid_dataloader_opts:
	batch_size: !ref <batch_size>
	num_workers: !ref <num_workers>
	collate_fn: !new:speechbrain.lobes.models.Tacotron2.TextMelCollate

	test_dataloader_opts:
	batch_size: !ref <batch_size>
	num_workers: !ref <num_workers>
	collate_fn: !new:speechbrain.lobes.models.Tacotron2.TextMelCollate

	################################
	# Model Parameters and model #
	################################
	n_symbols: 148 #fixed depending on symbols in textToSequence
	symbols_embedding_dim: 512
	hidden_dim: 256
	eprenet_dim: 512
	n_prenet_layers: 3
	dprenet_dim: 256
	postnet_dim: 256
	ff_dim: 1024
	n_heads: 8
	n_layers: 6
	n_postnet_layers: 5

	# Decoder parameters
	# The number of frames in the target per encoder step
	n_frames_per_step: 1
	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
	decoder_no_early_stopping: False

	# 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: 256
	postnet_kernel_size: 5
	postnet_n_convolutions: 5

	#model
	model: !new:TransformerTTS.TransformerTTS
	n_mel_channels: !ref <n_mel_channels>
	# symbols
	n_symbols: !ref <n_symbols>
	symbols_embedding_dim: !ref <symbols_embedding_dim>
	eprenet_dim: 512
	n_prenet_layers: 3
	# decoder
	dprenet_dim: !ref <prenet_dim>
	# postnet
	postnet_dim: !ref <postnet_dim>
	hidden_dim: !ref <hidden_dim>
	n_postnet_layers: !ref <n_postnet_layers>
	nhead: !ref <n_heads>

	guided_attention_sigma: 0.2
	guided_attention_weight: 50.0
	guided_attention_weight_half_life: 10.
	guided_attention_hard_stop: 50
	gate_loss_weight: 1.0

	guided_attention_scheduler: !new:speechbrain.nnet.schedulers.StepScheduler
	initial_value: !ref <guided_attention_weight>
	half_life: !ref <guided_attention_weight_half_life>

	criterion: !new:TransformerTTS.Loss
	gate_loss_weight: !ref <gate_loss_weight>
	guided_attention_weight: !ref <guided_attention_weight>
	guided_attention_sigma: !ref <guided_attention_sigma>
	guided_attention_scheduler: !ref <guided_attention_scheduler>
	guided_attention_hard_stop: !ref <guided_attention_hard_stop>

	modules:
	model: !ref <model>

	#optimizer
	opt_class: !name:torch.optim.Adam
	lr: !ref <learning_rate>
	weight_decay: !ref <weight_decay>

	#epoch object
	epoch_counter: !new:speechbrain.utils.epoch_loop.EpochCounter
	limit: !ref <epochs>

	train_logger: !new:speechbrain.utils.train_logger.FileTrainLogger
	save_file: !ref <train_log>

	#annealing_function
	lr_annealing: !new:speechbrain.nnet.schedulers.IntervalScheduler
	intervals:
	- steps: 6000
	lr: 0.0005
	- steps: 8000
	lr: 0.0003
	- steps: 10000
	lr: 0.0001

	#checkpointer
	checkpointer: !new:speechbrain.utils.checkpoints.Checkpointer
	checkpoints_dir: !ref <save_folder>
	recoverables:
	model: !ref <model>
	counter: !ref <epoch_counter>
	scheduler: !ref <lr_annealing>

	progress_sample_logger: !new:speechbrain.utils.train_logger.ProgressSampleLogger
	output_path: !ref <progress_sample_path>
	batch_sample_size: !ref <progress_batch_sample_size>
	formats:
	raw_batch: raw

	max_grad_norm: 1.0

	pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
	loadables:
	model: !ref <model>